[01-Aug-2009 00:00:46] [connected at Sat Aug 1 00:00:46 2009]
[01-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[01-Aug-2009 00:48:57] <Diddi> hm, I have a switch under monitoring... is it possible to change stuff in it using zenoss, like interface description etc..?
[01-Aug-2009 00:49:08] <Diddi> link up/down,,
[01-Aug-2009 00:51:47] <Diddi> ah.. could always create my own commands for that
[01-Aug-2009 00:51:51] <Diddi> i guess
[01-Aug-2009 05:47:42] <Diddi> is it possible to draw map links only between devices that are actually connected to each other, instead of drawing links between all devices in the same network?
[01-Aug-2009 05:48:16] <Diddi> got a couple of switches in the same subnet... but they're not all connected to eachother (:
[02-Aug-2009 00:00:46] [disconnected at Sun Aug 2 00:00:46 2009]
[02-Aug-2009 00:00:46] [connected at Sun Aug 2 00:00:46 2009]
[02-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[02-Aug-2009 05:38:37] <Diddi> I'm using Paging to send SMS instead of Page (calling another program only).. Is it possible to add newlines in the Paging message?
[02-Aug-2009 05:49:00] <Diddi> I've tried \n and \r in different combinations without success.. they're just sent as-is
[02-Aug-2009 15:06:48] ke4qqq_ is now known as ke4qqq
[02-Aug-2009 21:30:12] <sergeymasushko> how can force remodeling of all devices?
[02-Aug-2009 22:13:52] edwardam1 is now known as edwardam
[03-Aug-2009 00:00:46] [disconnected at Mon Aug 3 00:00:46 2009]
[03-Aug-2009 00:00:46] [connected at Mon Aug 3 00:00:46 2009]
[03-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[03-Aug-2009 03:12:09] <Dieterbe> when i create a datasource for a template and put a nagios command in it. I can select the severity, but is this a "catch all" severity? because the nagios plugin will return different exit codes from warning/error/unknown states. it seems a bit weird that all these conditions map to only 1 severity in zenoss
[03-Aug-2009 08:52:28] edwardam1 is now known as edwardam
[03-Aug-2009 10:42:56] <eidolon> hi folks - i'm configuring up windows clients on our zenoss enterprise installation. i can use WMI or SNMP - we're mosty a linux shop, so SNMP is the preferred method. is there a reason to use one ove rthe other? do all zenpacks work with both?
[03-Aug-2009 10:44:17] <ckrough> I've heard WMI is more reliable for Windows nodes
[03-Aug-2009 10:46:17] <mrayzenoss> so officially we try to make them equally, but the SNMP from MS is somewhat limited
[03-Aug-2009 10:46:31] <mrayzenoss> but the WMI performance with Server 2008 and Vista reportedly sucks
[03-Aug-2009 10:46:41] <mrayzenoss> not us, MS's implementation
[03-Aug-2009 10:46:59] <mrayzenoss> but WMI uses TCP so it should be more network reliable
[03-Aug-2009 10:47:13] <mrayzenoss> but since it's MicroSoft, it also tends to have more memory leaks server-side
[03-Aug-2009 10:47:36] <mrayzenoss> WMI gives you event log and service monitoring, which aren't exposed via SNMP
[03-Aug-2009 10:49:19] <Disconnect> so in the end, the answer is.. don't use windows?
[03-Aug-2009 10:49:20] <mrayzenoss> so the short answer is "use WMI"
[03-Aug-2009 10:49:24] <mrayzenoss> heh
[03-Aug-2009 10:49:43] <mrayzenoss> Disconnect: you said it, not me
[03-Aug-2009 10:49:47] <Disconnect> heh
[03-Aug-2009 10:50:10] <Disconnect> (admittedly, *nix doesn't expose syslog via snmp either..)
[03-Aug-2009 10:50:46] <mrayzenoss> well, there is a tool for exporting Windows Event Log with syslog
[03-Aug-2009 10:51:42] <mrayzenoss> http://www.intersectalliance.com/projects/SnareWindows/
[03-Aug-2009 10:52:02] <mrayzenoss> again, short answer is "use WMI and stay up on your patches"
[03-Aug-2009 11:11:08] <eidolon> mray: that's a very good answer, thank you - oh crap, he left. but... my issue with WMI is this we also seem to be getting system event log traffic into zenoss. stuff we have no interest whatsoever in tracking
[03-Aug-2009 11:11:11] <eidolon> how can we turn that off?
[03-Aug-2009 11:19:21] <rmatte> you want to eliminate all syslog messages via WMI?
[03-Aug-2009 11:19:44] <rmatte> or just certain ones?
[03-Aug-2009 11:22:08] <rmatte> If you want to disable all of them, then you need to go in to zProperties for the Windows class and change zWinEventlog to false. If you want to keep monitoring it but only monitor higher level events then you need to change the value for zWinEventlogMinSeverity
[03-Aug-2009 11:22:29] <eidolon> ooooo
[03-Aug-2009 11:22:35] <eidolon> that sounds awesome rmatte. *makes notes*
[03-Aug-2009 11:22:52] <eidolon> i need to remodel some of these machines that were set up as SNMP, and make them WMI, then i'll set those properties.
[03-Aug-2009 11:23:49] <rmatte> if you want to block certain messages then you need to go to events on the left, go to a class (you can optionally make one called "Suppressed" or something, or you can use an existing class), then go to the dropdown click Add Mapping, name it whatever you want (WinEventlogSuppress1 or whatever), then click on it to access it's properties...
[03-Aug-2009 11:25:10] <eidolon> hum
[03-Aug-2009 11:25:13] <rmatte> Go to the Edit tab, set the event class key to "defaultmapping" and in the Regex box, type in part of the message that you're seeing, in the example box put the whole message you're seeing, then click save and go to the zProperties tab, set zEventAction to "drop" and click save
[03-Aug-2009 11:25:25] <eidolon> rmatte, how can i tell how polling is happening? because on my path, I have: /Devices /Server /Windows /WMI_Active Directory /BOS-ADS01
[03-Aug-2009 11:25:25] <rmatte> that's for suppressing individual messages
[03-Aug-2009 11:25:35] <eidolon> butunder Attributes, it's got all SNMP settings.
[03-Aug-2009 11:26:03] <rmatte> each class can have local attributes assigned to it, it all depends on how you setup the zProperties after creating that class
[03-Aug-2009 11:26:13] * eidolon didn't create this class
[03-Aug-2009 11:26:22] <eidolon> i'm tempted to delete it and remodel it.
[03-Aug-2009 11:26:27] <rmatte> you didn't create the WMI_Active Directory class?
[03-Aug-2009 11:26:37] <eidolon> no. i inherited this setup.
[03-Aug-2009 11:26:47] <rmatte> ah, well the person who created that class then
[03-Aug-2009 11:26:59] <rmatte> you need to go in to the zProperties and take a look at what's enabled/disabled
[03-Aug-2009 11:27:01] * eidolon sees zAxlPassword nad Username in the zproperties
[03-Aug-2009 11:27:15] <rmatte> Also, if you scroll to the bottom of zProperties, you'll see what fields are set locally
[03-Aug-2009 11:27:19] <rmatte> there's a dropdown with a list
[03-Aug-2009 11:27:20] <eidolon> zdevice templates - Device_WMI and ActiveDirectory
[03-Aug-2009 11:27:31] <eidolon> ok
[03-Aug-2009 11:27:45] <eidolon> local properties have... great.
[03-Aug-2009 11:27:46] <rmatte> mrayzenoss: FYI, I'm on vacation for the next 2 weeks, so I probably won't be around much, but I'll stop in a bit
[03-Aug-2009 11:28:05] <eidolon> zwinpassword, zwinusername, zsnmp port...
[03-Aug-2009 11:28:09] <eidolon> mixed signals
[03-Aug-2009 11:28:23] <rmatte> the fields are always going to be there, they are global
[03-Aug-2009 11:28:30] <rmatte> it all depends on what they are actually set to
[03-Aug-2009 11:28:46] <eidolon> ok. i see.
[03-Aug-2009 11:28:59] <eidolon> zwinsuer and zwinpassword are set. zwmimonitorignore is flase
[03-Aug-2009 11:29:03] <rmatte> and it is possible to do both WMI and SNMP monitoring on the same device
[03-Aug-2009 11:29:10] <eidolon> (sorry if i'm sounding stupid here)
[03-Aug-2009 11:29:31] <rmatte> then WMI monitoring is enabled for that device
[03-Aug-2009 11:29:37] <eidolon> ok
[03-Aug-2009 11:29:55] <rmatte> there's zWinEventlog which determines whether the event log is monitored or not
[03-Aug-2009 11:30:02] <rmatte> there's also the severity setting for it
[03-Aug-2009 11:30:14] <rmatte> and there will be zSnmpMonitorIgnore, and such
[03-Aug-2009 11:30:21] <eidolon> 'zwineventlog' 'trus' 'zwineventlogminseverity' '2'
[03-Aug-2009 11:30:24] <rmatte> which should be set to true if you're doing straight WMI
[03-Aug-2009 11:30:36] <rmatte> yeh, that min severity is the default
[03-Aug-2009 11:30:42] <eidolon> zsnmpmonitorignore is set to false
[03-Aug-2009 11:30:45] <eidolon> so we're doing both
[03-Aug-2009 11:30:58] <rmatte> is there an snmp community string specified?
[03-Aug-2009 11:31:02] <eidolon> yes
[03-Aug-2009 11:31:24] <eidolon> okay hang on.
[03-Aug-2009 11:31:28] * eidolon backs up a few steps.
[03-Aug-2009 11:31:43] <rmatte> Under zSnmpCommunity?
[03-Aug-2009 11:31:49] <eidolon> yeah, hang on a sec.
[03-Aug-2009 11:33:09] <straterra> My new netbook has arrived..Mwuahahaha
[03-Aug-2009 11:34:35] <rmatte> lol
[03-Aug-2009 11:35:44] <eidolon> okee. turned off wineventlog for the WMI and WMI_Active Directory classes...
[03-Aug-2009 11:35:49] * eidolon continues fiddling.
[03-Aug-2009 11:38:12] <eidolon> i love classes and organizers with no devices or sub-devices.
[03-Aug-2009 11:38:15] * eidolon cleans up.
[03-Aug-2009 11:39:17] <rmatte> lol
[03-Aug-2009 11:39:25] <rmatte> some are there by default
[03-Aug-2009 11:39:54] <rmatte> I wouldn't go too crazy deleting stuff, you might need some of it in the future
[03-Aug-2009 11:40:01] <Disconnect> yah i discovered that removing some of them breaks zenmigrate or whatever the update app is :/
[03-Aug-2009 11:40:02] <eidolon> so i already see a problem. i suspect thes guys are going to want us to monitor their apps and event logs for system errors.
[03-Aug-2009 11:40:09] <eidolon> but i dont know if they're using any consistent naming scheme.
[03-Aug-2009 11:40:29] <eidolon> like, i have 38,515 occurances of:
[03-Aug-2009 11:40:30] <eidolon> Windows Service 'SQLSERVERAGENT' is down
[03-Aug-2009 11:40:30] <rmatte> naming scheme for?
[03-Aug-2009 11:40:35] <eidolon> application errors.
[03-Aug-2009 11:40:44] <rmatte> I see
[03-Aug-2009 11:41:27] * eidolon has so much to learn.
[03-Aug-2009 11:42:33] <rmatte> Yeh, it takes a while
[03-Aug-2009 11:43:26] <eidolon> well, i keep wanting to go "See that error message? add an ignore rule for it."
[03-Aug-2009 11:43:29] <eidolon> and zenoss doesn't work that way.
[03-Aug-2009 11:43:45] <rmatte> I explained how to ignore errors
[03-Aug-2009 11:43:48] <eidolon> you did.
[03-Aug-2009 11:43:50] <eidolon> i saw
[03-Aug-2009 11:44:33] <rmatte> it's a bit more complex than just point and click, but it's fairly simple
[03-Aug-2009 11:44:57] <eidolon> yah, i see it.
[03-Aug-2009 11:45:08] <eidolon> well, right now what i've done is shut down event logging.
[03-Aug-2009 11:45:34] <eidolon> since i templated to the users "i will be monitoring system up, load, cpu, memory, and remote sockets. if youw ant more, tell me what you want."
[03-Aug-2009 11:45:37] <eidolon> and they never got back to me.
[03-Aug-2009 11:45:40] <eidolon> so. f'em.
[03-Aug-2009 11:45:49] * eidolon BOFH's.
[03-Aug-2009 11:46:00] <rmatte> hehe
[03-Aug-2009 11:46:25] <rmatte> with your sidekick, the PFY?
[03-Aug-2009 11:46:27] <eidolon> can i have no pre-production machines show upin the event list?
[03-Aug-2009 11:46:39] <eidolon> ala, i like having a very clean dashboard.
[03-Aug-2009 11:46:49] <eidolon> well, event console,sorry
[03-Aug-2009 11:47:04] <rmatte> eidolon: you effectively don't want to monitor them at all?
[03-Aug-2009 11:47:09] <rmatte> or just not have events show up?
[03-Aug-2009 11:47:13] <eidolon> let me think a moment.
[03-Aug-2009 11:47:53] <eidolon> i dont' want events to show up. i want to be able to... hmm. basically i want to be able to go to event console, and only see production machines. but if i look up a specific pre-production box, i want to see events / performance stats from it.
[03-Aug-2009 11:48:44] * eidolon may be asking for too much.
[03-Aug-2009 11:48:51] <rmatte> well, there's no real way to filter it like that, what you can do is make an event mapping much in the way that I described above, but do a transform that checks to see the production state of a device, and if it's "Pre-Production" sends the event to history immediately
[03-Aug-2009 11:49:12] <rmatte> then you could always go to the events tab of the pre-prod machines and check the event history to see what has come in
[03-Aug-2009 11:49:18] <mloven> event views?
[03-Aug-2009 11:49:55] <rmatte> mloven: hunh?
[03-Aug-2009 11:50:26] <mloven> wouldn't be exactly what he wanted, but he could create an event view that doesn't have any pre-production devices in it...
[03-Aug-2009 11:53:23] <eidolon> that sounds better.
[03-Aug-2009 11:53:46] <eidolon> not sure what that is though
[03-Aug-2009 11:54:06] <rmatte> mloven: you talking about a portlet basically?
[03-Aug-2009 11:54:11] <eidolon> (what i've been doing is setting up groups. a group of 'production' machines, and having a view of htem)
[03-Aug-2009 11:54:22] <rmatte> The transform would be:
[03-Aug-2009 11:54:22] <rmatte> if evt.prodState == 'Pre-Production':
[03-Aug-2009 11:54:23] <rmatte> evt._action = 'history'
[03-Aug-2009 11:54:25] <rmatte> for my method
[03-Aug-2009 11:54:53] <mloven> It doesn't completely solve the issue because the events from Pre-Production devices will still show up in the event console, but you could have a dedicated event view that only shows Production devices' events
[03-Aug-2009 11:55:09] <eidolon> how do you build event views?
[03-Aug-2009 11:55:46] <mloven> Settings>Users>yourusername>Event Views
[03-Aug-2009 11:57:44] <mloven> you would just specify Production State > Pre-Production
[03-Aug-2009 11:58:23] <rmatte> or Production-State = Production if you only care about Production alerts
[03-Aug-2009 11:58:35] <eidolon> hmmmmmmmmmmmmmm.
[03-Aug-2009 11:59:03] <rmatte> my method would resolve the issue for the main event console
[03-Aug-2009 11:59:06] <rmatte> up to you
[03-Aug-2009 12:01:40] <eidolon> actually i like hte custom views.
[03-Aug-2009 12:01:45] <mloven> yep. both work. my way is just like flipping a filter that doesn't show devices that aren't in Production. the events are still there....lurking.... rmatte's way actually gets rid of the events. Depends on what you want.
[03-Aug-2009 12:01:45] * eidolon is playing now.
[03-Aug-2009 12:02:59] <eidolon> awesome.
[03-Aug-2009 12:03:00] <eidolon> look at that
[03-Aug-2009 12:53:16] <konfoo> anyone here monitoring using the dell pack?
[03-Aug-2009 12:53:45] <konfoo> degrading the raid array raises an alert in the hardware properties page but the overall status is OK
[03-Aug-2009 14:14:14] <eidolon> hmm. how long should zenbackup take to run?
[03-Aug-2009 14:14:28] <eidolon> we have about 100 machines.
[03-Aug-2009 15:12:28] <rhett> eidolon that all depends on how fast the machine is that you're doing the backup on
[03-Aug-2009 15:13:02] <rhett> it tgz's everything into the backup archive
[03-Aug-2009 16:05:49] <mrayzenoss> New WMI Data Source ZenPack: http://www.zenoss.com/community/projects/zenpacks/wmidatasource
[03-Aug-2009 16:33:03] <malbon> excellent. A 3rd way of getting WMI data
[03-Aug-2009 16:44:21] <rmatte> actually, these new WMI packs that are coming out are going to be way nicer than the existing Dastrup ones
[03-Aug-2009 22:28:13] <ironpaw> anyone around?
[03-Aug-2009 22:28:23] <ironpaw> where can i find the conary migrate binary?
[04-Aug-2009 00:00:46] [disconnected at Tue Aug 4 00:00:46 2009]
[04-Aug-2009 00:00:46] [connected at Tue Aug 4 00:00:46 2009]
[04-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[04-Aug-2009 00:28:41] <raddy> Hello Everybody
[04-Aug-2009 00:28:48] <raddy> Anybody out there?
[04-Aug-2009 03:19:52] <raddy> Hello Everybody
[04-Aug-2009 03:20:14] <raddy> Can anybody suggest a regexp for monitoring postgres process?
[04-Aug-2009 03:20:38] <raddy> just using postgres shows 50 processes.
[04-Aug-2009 03:29:37] <raddy> Anybody there???
[04-Aug-2009 03:29:53] <raddy> How do i monitor postgres process?
[04-Aug-2009 03:35:17] <raddy> Anybody there??
[04-Aug-2009 03:41:55] <raddy> Anybody out there???
[04-Aug-2009 04:26:14] <raddy> Hello Everybody
[04-Aug-2009 04:26:19] <raddy> Anybody there?????
[04-Aug-2009 04:39:26] <raddy> Anybody there?
[04-Aug-2009 04:40:01] <raddy> Can anybody suggest a good regex for postgres
[04-Aug-2009 04:57:03] <raddy> Hi folke
[04-Aug-2009 04:57:21] <folke> Hi raddy.
[04-Aug-2009 05:02:06] <raddy> folke : Can you suggest a good regex for postgres.
[04-Aug-2009 05:02:39] <folke> raddy, regex testing?
[04-Aug-2009 05:02:55] <raddy> folke : Nope, real usage.
[04-Aug-2009 05:03:28] <raddy> folke : zenoss is showing too much of processes when i just use postgres
[04-Aug-2009 05:04:10] <folke> raddy: Ah, ok.. Sry I had the same problem before when I monitored a few linux nodes.
[04-Aug-2009 05:04:31] <folke> Do you see the right precesses atleas? Or every process on the server?
[04-Aug-2009 05:04:59] <raddy> folke : right process alias i think
[04-Aug-2009 05:05:44] <folke> Do you run the latest 2.4.2, I know that there has been trouble with processes and linux historically.
[04-Aug-2009 05:06:03] <raddy> folke : Yup, latest version.
[04-Aug-2009 05:07:25] <folke> Ok, I am for the moment trying to "upgrade" our old 2.1.3 env to 2.4.2
[04-Aug-2009 05:07:39] <folke> So I will try to mess with processes in nix again.
[04-Aug-2009 05:07:43] <raddy> folke : ok
[04-Aug-2009 05:07:56] <raddy> folke : Do you have solution?
[04-Aug-2009 05:08:06] <folke> raddy: Sry, not for the momen..
[04-Aug-2009 05:08:18] <raddy> folke : ok
[04-Aug-2009 05:47:48] <raddy> Anybody there?
[04-Aug-2009 05:52:52] baffle_ is now known as baffle
[04-Aug-2009 08:11:32] <klinstifen> morning all
[04-Aug-2009 08:11:49] <klinstifen> can someone confirm for me the size of their data.fs file in $ZENHOME/var
[04-Aug-2009 08:12:16] <klinstifen> mine is 14G
[04-Aug-2009 08:12:31] <klinstifen> that seems high to me especially because I only monitor 100 or so devices
[04-Aug-2009 08:18:59] <Dieterbe> klinstifen: just a sec..
[04-Aug-2009 08:19:46] <Dieterbe> 53M Aug 4 14:07 Data.fs
[04-Aug-2009 08:20:17] <Dieterbe> that's an installation that monitors about 12 devices, and it's up for .. 2 months or so (rough estimate)
[04-Aug-2009 08:20:34] <Dieterbe> 1-2 months
[04-Aug-2009 08:20:47] <klinstifen> hmmm
[04-Aug-2009 08:20:55] <klinstifen> i am monitoring 39 devices
[04-Aug-2009 08:20:56] <Dieterbe> how should one debug "non working" event transformations?
[04-Aug-2009 08:22:00] <klinstifen> and it's been up for just over a year
[04-Aug-2009 08:22:43] <klinstifen> what's not working with your transform?
[04-Aug-2009 08:24:13] <Dieterbe> well, to monitor my boxes i need to go through 2 vpn's. one endpoint is vpn0, the one after that is called firewall. i have this rule:
[04-Aug-2009 08:24:16] <Dieterbe> http://pastebin.ca/1517859
[04-Aug-2009 08:24:51] <Dieterbe> but it doesn't work, but at this point the firewall box is not reachable by ping anymore, however i'm getting notifications about all devices being down
[04-Aug-2009 08:26:24] <klinstifen> so what event is this transforming? the firewall down event?
[04-Aug-2009 08:27:04] <klinstifen> can you ping firewall box from the zenoss server cli?
[04-Aug-2009 08:27:30] <Dieterbe> no, it should work on all events for all other boxes who are "behind" firewall. i've put this rule in /Events
[04-Aug-2009 08:27:31] <raddy> Hello Everybody
[04-Aug-2009 08:27:50] <raddy> Can anybody suggest a good regex for postgres?
[04-Aug-2009 08:28:33] <Dieterbe> klinstifen: well, now it pings again, but a few minutes ago it didn't
[04-Aug-2009 08:30:20] <raddy> When i just use postgres it shows lot of processes
[04-Aug-2009 08:30:47] <raddy> Can anyone suggest?
[04-Aug-2009 08:31:00] <ckrough> What is the question?
[04-Aug-2009 08:31:52] <raddy> Can anybody suggest a good regex for postgres
[04-Aug-2009 08:32:18] <raddy> It shows lot of postgres processes when i use just "postgres" as regexp
[04-Aug-2009 08:32:40] <ckrough> oh, I see what your doing. Lemme see what I have in mine.
[04-Aug-2009 08:34:03] <ckrough> maybe match on an expression based around 'postmaster'?
[04-Aug-2009 08:34:20] <ckrough> Im not running postgres to check atm, but I think postmaster is the parent process
[04-Aug-2009 08:34:25] <raddy> ckrough : postmaster is useless
[04-Aug-2009 08:35:22] <cgibbons> hmmm
[04-Aug-2009 08:39:10] <Dieterbe> raddy: what about 'postgres: '
[04-Aug-2009 08:39:24] <Dieterbe> or /usr/bin/postgres for the 'main one' ?
[04-Aug-2009 08:39:59] <ckrough> good site for testing this stuff out: http://gskinner.com/RegExr/
[04-Aug-2009 08:40:11] <Dieterbe> so, how can i debug non-working event transforms?
[04-Aug-2009 08:40:24] <raddy> Dieterbe : /user/bin/postgres didn't work, but i didn't try postgres:
[04-Aug-2009 08:45:00] klinstifen_ is now known as klinstifen
[04-Aug-2009 08:52:44] <Dieterbe> so, klinstifen .. any idea?
[04-Aug-2009 08:52:56] <ckrough> DO you most of you run your ethernet templates with COUNTER or DERIVE datasources? Im dealing with spikes a little too often using COUNTER with a max value of one petabit
[04-Aug-2009 08:53:19] <klinstifen> i don't know if i am reading your transfer code correctly, but it looks like it will only report the status of either ping vpn0 or ping firewall
[04-Aug-2009 08:53:21] <klinstifen> is that correct?
[04-Aug-2009 08:53:22] <ckrough> actually, a max value of 141 terabits since IOS returns bytes...
[04-Aug-2009 08:54:58] <raddy> Dieterbe : postgres: shows all postgres processses
[04-Aug-2009 08:55:27] <ckrough> can you put your process list in a pastebin?
[04-Aug-2009 08:55:35] <Dieterbe> klinstifen: well i handle 2 different cases separately. first i check if it's a server behind vpn0 (that includes firewall). any such event depends on the ping state of vpn0. then i check for all devices behind firewall. these depend on the ping state of firewall
[04-Aug-2009 08:57:33] <klinstifen> well one way you could check it is to just change the code so that the text of the event changes
[04-Aug-2009 08:57:36] <klinstifen> know what i mean?
[04-Aug-2009 08:57:47] <klinstifen> you're not sure which part of the transform is breaking, right?
[04-Aug-2009 08:58:12] <klinstifen> so why not split it up and just have it to something innocuous like changing the text of the event.
[04-Aug-2009 08:58:33] <raddy> ckrough : Is there any other route?
[04-Aug-2009 08:58:36] <klinstifen> then then you drill down further and see if the problem is with the ping check or device name check etc
[04-Aug-2009 09:00:43] <klinstifen> and another thing...can you ping the devices behind firewall from your zenoss server...from the command line?
[04-Aug-2009 09:01:24] <raddy> basically using postgres shows all the idle postgres processes
[04-Aug-2009 09:02:39] <klinstifen> dieterbe: maybe it's just a route issue
[04-Aug-2009 09:08:05] <Dieterbe> klinstifen: ?? no it's just my event transforms that don't work.
[04-Aug-2009 09:08:36] <klinstifen> ah...well then yeah i would suggest breaking em down and testing it that way...
[04-Aug-2009 09:15:23] <mrayzenoss> New WMI Windows Performance ZenPack: http://www.zenoss.com/community/projects/zenpacks/wmiperf_windows
[04-Aug-2009 09:17:36] <Dieterbe> klinstifen: but then i have to bring down my vpn just to be able to test :/ unless i fire multiple events myself
[04-Aug-2009 09:18:40] <klinstifen> you could create a test event.. i think there are docs on that
[04-Aug-2009 09:19:11] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch05s02.html
[04-Aug-2009 09:19:20] <mrayzenoss> or just zensendevent
[04-Aug-2009 09:19:30] <klinstifen> or you could briefly change the ip within zenoss for one of the devices and then zenoss will think it is down
[04-Aug-2009 09:19:46] <klinstifen> even easier, thanks mray
[04-Aug-2009 09:19:58] <mrayzenoss> it's what I'm here for
[04-Aug-2009 09:25:52] <raddy> Hello Everybody
[04-Aug-2009 09:26:07] <raddy> Please suggest good regexp for postgres
[04-Aug-2009 09:38:04] <sto6ma9ch> Hello.
[04-Aug-2009 09:38:18] <sto6ma9ch> I have an issue with zenperfsnmp
[04-Aug-2009 09:38:45] <sto6ma9ch> The log shows "Error loading config for devices ['..." for every device
[04-Aug-2009 09:39:01] <sto6ma9ch> Then I receive a zenperfsnmp heartbeat failure
[04-Aug-2009 09:39:38] <cgibbons> that would first make me think zenhub was sick; have you checked its log file?
[04-Aug-2009 09:39:55] <sto6ma9ch> Not yet. Checking now.
[04-Aug-2009 09:41:06] <sto6ma9ch> Just restarted zenperfsnmp. Tailing zenhub.log
[04-Aug-2009 09:43:30] <sto6ma9ch> This is the only error thus far:
[04-Aug-2009 09:43:33] <sto6ma9ch> ERROR zen.MinMaxCheck: User-supplied Python expression ((here.totalBlocks * here.zFileSystemSizeOffset) * .9) for maximum value caused error: ['usedBlocks_usedBlocks']
[04-Aug-2009 09:44:58] <sto6ma9ch> Ahh, just found one:
[04-Aug-2009 09:45:01] <sto6ma9ch> CRITICAL zen.hub: ZenPing 'cml-zensvr.as.local' not found, ignoring network topology.
[04-Aug-2009 09:49:11] <sto6ma9ch> That's the ZenOSS server itself.
[04-Aug-2009 09:49:26] <sto6ma9ch> Just added it. Restarting zenperfsnmp and tailing zenhub.log
[04-Aug-2009 09:51:20] <ckrough> What type of datasource should I be using with ethernet templates? COUNTER or DERIVE? datasources? Im dealing with spikes a little too often using COUNTER
[04-Aug-2009 09:56:56] <sto6ma9ch> OK, got that error resolved. Now I see "INFO zen.hub: building pingtree from cml-zensvr.as.local"
[04-Aug-2009 09:58:03] <sto6ma9ch> After that, I receive these messages:
[04-Aug-2009 09:58:21] <sto6ma9ch> "'router1.as.local' network '10.90.16.0/20' not in topology"
[04-Aug-2009 10:00:06] <sto6ma9ch> Each of the networks that are "not in the topology" are listed in the NEtworks section of Zenoss
[04-Aug-2009 10:02:17] <sto6ma9ch> Not sure how to add these networks to the topology. Any advice?
[04-Aug-2009 10:15:41] <sto6ma9ch> I just received the zenperfsnmp heartbeat failure message again after restarting the service
[04-Aug-2009 10:26:49] <DesignsEdg1> Hello all !
[04-Aug-2009 10:30:19] <Dieterbe> hi
[04-Aug-2009 10:53:13] <mrayzenoss> gotta go do the Getting Started Webinar, bbiab
[04-Aug-2009 10:57:21] <Dieterbe> if there are syntax errors in event transforms, are these logged somewhere?
[04-Aug-2009 11:04:54] <jsm> I applied a set of Collector Plugins to /Servers/Linux, but they don't seem to get inherited by its sub-classes. why?
[04-Aug-2009 11:14:44] <al_x> how can i change the links in the emails zenoss sends out so they contain an actual ip, and not localhost:8080
[04-Aug-2009 11:16:10] <chudler> al_x: set zopeurl in zenactions.conf
[04-Aug-2009 11:21:52] <al_x> ever seen this before
[04-Aug-2009 11:21:53] <al_x> snmp_build: unknown failuresnmpwalk: Error building ASN.1 representation (Can't build OID for variable
[04-Aug-2009 11:22:50] <al_x> oh nm heh
[04-Aug-2009 11:28:00] <syntaxcollector> We're running Zenoss open source version at our company, we have two locations our office and colo. I want to deploy a second collector at the colo and be able to read the data off of our primary zenoss instance at the office.
[04-Aug-2009 11:28:12] <syntaxcollector> can we do this with the open source edition? the manual seems to say no
[04-Aug-2009 11:29:36] <al_x> chudler after i set zopeurl do i need to restart zenoss?
[04-Aug-2009 11:29:53] <chudler> probably, at least zenactions.
[04-Aug-2009 11:38:05] <al_x> is there a way to test-fire an email
[04-Aug-2009 11:38:15] <al_x> or is it easier just to block snmp at the firewall and have it trip 'naturally'
[04-Aug-2009 12:16:32] <mrayzenoss> jsm: I'm not sure collector plugins get applied post-creation of the subclass
[04-Aug-2009 12:17:33] <mrayzenoss> syntaxcollector: you can manually do it, there are assorted docs in the community for it: http://delicious.com/mray/collectors/
[04-Aug-2009 12:20:42] <syntaxcollector> mrayzenoss: thanks, I was looking at that first link earlier today
[04-Aug-2009 12:51:39] <jsm> mrayzenoss: yeah, they aren't, grumble.. i'm just deleting the devices and re-adding them
[04-Aug-2009 12:51:51] <jsm> you'd think it would be a feature
[04-Aug-2009 12:51:56] <jsm> sounds like a bug to me
[04-Aug-2009 12:55:56] <brice> What does an Authentication Trap Failure received on the snmptrap service mean?
[04-Aug-2009 12:56:20] <jsm> that your snmp trap failed to authenticate i'd guess
[04-Aug-2009 12:57:06] <brice> so mismatched community names between myd evice and trap service?
[04-Aug-2009 12:59:08] <jb> anybody here running IBM xseries boxes?
[04-Aug-2009 13:00:15] <brice> older ones
[04-Aug-2009 13:00:24] <jb> do you run director on them?
[04-Aug-2009 13:00:29] <brice> i dont
[04-Aug-2009 13:00:39] <jb> we are a dell shop (we run openmanage).. and i'm trying to figure out how director works
[04-Aug-2009 13:20:14] <brice> nm I figured it out
[04-Aug-2009 13:20:36] <brice> hrm was hoping the few cheap dell printers we have would send an snmp trap back that it completed a print job of x pages
[04-Aug-2009 13:20:53] <mrayzenoss> jsm: open a ticket for that, I can see arguments on both sides for it
[04-Aug-2009 13:20:54] <brice> to confirm against whose i'm pulling from WMI
[04-Aug-2009 14:20:21] <gwb2353> is this the command you would use to figure out why zenoss says SNMP isn't responding, but you can snmpwalk successfully from the command line of the collector?
[04-Aug-2009 14:20:35] <gwb2353> [zenoss@eye ~]$ zenperfsnmp run -v 10 --monitor cornea --device com10-chicken-z1r3
[04-Aug-2009 14:20:59] <gwb2353> (where eye is the hub and cornea is a collector)
[04-Aug-2009 14:21:38] <gwb2353> tcpdump is telling me it's not quite right (no traffic from collector to device when command is issued)
[04-Aug-2009 14:24:44] <gwb2353> note that when I model the device, everything works OK.
[04-Aug-2009 14:25:29] <gwb2353> just can't get the "/Status/SNMP SNMP agent down" even to quit showing up
[04-Aug-2009 14:28:41] <jb> twm?
[04-Aug-2009 14:28:47] <jb> anybody here know of a good windows based MIB browser?
[04-Aug-2009 14:39:58] <rhett> hmm. not really. there is one i used thats old but i dont remember what its called
[04-Aug-2009 14:45:05] <chudler> I hadn't noticed the "watchdog" and "watchdogPath" options in the daemon configs before. I can't find documentation for it, either. Anyone can provide a link?
[04-Aug-2009 14:45:14] <gwb2353> [zenoss@cornea etc]$ zenmodeler run -d com10-chicken-z1r3 -v 10 --hubhost=eye runs properly (no errors)
[04-Aug-2009 14:54:51] <gwb2353> getting further (I think) "[zenoss@cornea etc]$ zenperfsnmp run -d com10-chicken-z1r3 -v 10 --hubhost=eye --configfile=/opt/zenoss/etc/cornea_zenperfsnmp.conf"
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:Fetching config for ['com10-chicken-z1r3']
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:Getting device ping issues
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:unresponsive devices: [['com10-chicken-z2r2', 1, 3]]
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:Sent a 'stop' event
[04-Aug-2009 14:56:46] <jb> can you snmpwalk it?
[04-Aug-2009 14:56:53] <gwb2353> (note: ping from the command line works, and snmpwalk is fine)
[04-Aug-2009 14:57:17] <jb> and you have the SNMP community set properly?
[04-Aug-2009 14:57:42] <gwb2353> yes - i can use "zenmodeler" and everything is good and double-checked zproperties
[04-Aug-2009 14:58:20] <jb> hmm
[04-Aug-2009 15:01:51] <gwb2353> works (5/5): zenmodeler run -d com10-chicken-z1r3 -v 10 --hubhost=eye --configfile=/opt/zenoss/etc/cornea_zenperfsnmp.conf
[04-Aug-2009 15:02:16] <gwb2353> fails (5/5): zenperfsnmp run -d com10-chicken-z1r3 -v 10 --hubhost=eye --configfile=/opt/zenoss/etc/cornea_zenperfsnmp.conf
[04-Aug-2009 15:02:39] <gwb2353> ...with the "device ping issues"
[04-Aug-2009 15:17:19] <twm1010> afternoon
[04-Aug-2009 15:17:29] <twm1010> anytone know how to test the apache monitor via command line?
[04-Aug-2009 15:36:47] <jb> hey matt
[04-Aug-2009 15:37:21] <mrayzenoss> howdy
[04-Aug-2009 15:37:28] <jb> not much.. working on a IBM zenpack
[04-Aug-2009 15:37:46] <jb> quick q.. when I edit a modeler plugin .py.. what daemons need to be restarted?
[04-Aug-2009 15:37:59] <mrayzenoss> Egor has an IBMMon he's working on as well
[04-Aug-2009 15:38:04] <jb> oh crap
[04-Aug-2009 15:38:04] <jb> hah
[04-Aug-2009 15:38:06] <mrayzenoss> not sure what it's for
[04-Aug-2009 15:38:27] <jb> is egor on IRC?
[04-Aug-2009 15:38:34] <mrayzenoss> yeah, he's bigegor_
[04-Aug-2009 15:38:48] <jb> wonder if thats the best way to get ahold of him
[04-Aug-2009 15:46:37] <twm1010> i'd think the _ means he is away
[04-Aug-2009 15:50:45] <twm1010> mrayzenoss: is there a built-in method to track latency?
[04-Aug-2009 15:51:09] <twm1010> I see the fping zenpack, that seems like overkill, more datapoints than i'm after
[04-Aug-2009 15:51:25] <mrayzenoss> twm1010: not that I'm aware of...
[04-Aug-2009 15:51:51] <gwb2353> i'm using the ZenPacks.Nagios.CheckPing; pretty simple to setup
[04-Aug-2009 15:51:58] <mrayzenoss> there was/is an fping collector associated with the Dev Guide that never got finished
[04-Aug-2009 15:52:17] <mrayzenoss> it's the example for writing a new collector
[04-Aug-2009 15:52:27] <mrayzenoss> but it never made it into the product
[04-Aug-2009 15:52:42] <mrayzenoss> it would be lighter weight than the command fping or the Nagios CheckPing
[04-Aug-2009 15:53:56] <twm1010> Hrmm.. i think the nagios one will cut it just fine.
[04-Aug-2009 16:05:31] <gwb2353> anybody use the SentryCDU zenpack (under development) here? https://webshare.uchicago.edu/users/chudler/zenoss/
[04-Aug-2009 16:05:46] <chudler> yes :-p
[04-Aug-2009 16:06:08] <gwb2353> doh! it's that annoying guy from UT that keeps on emailing you!
[04-Aug-2009 16:06:25] <mrayzenoss> chudler: just 10 more days until the end of the ZenPack contest
[04-Aug-2009 16:06:42] <jb> mrayzenoss: whats the current URL to the zenpack documentation?
[04-Aug-2009 16:06:47] <chudler> no worries. I am <-> this close to having something to release. I am overworked lately
[04-Aug-2009 16:06:49] <jb> notes on developing zenpacks, etc
[04-Aug-2009 16:07:13] <mrayzenoss> jb: http://www.zenoss.com/community/docs/ Dev Guide and Extended Monitoring Guide
[04-Aug-2009 16:07:31] <jb> thanks
[04-Aug-2009 16:07:37] <mrayzenoss> jb: more here: trac-zenpacks/
[04-Aug-2009 16:08:21] <gwb2353> if i install what you currently have posted, will that mess things up later? (just curious to see what I can get from all of our Sentry devices)
[04-Aug-2009 16:08:36] <gwb2353> (non-production monitoring system, under deployment right now)
[04-Aug-2009 16:09:08] <mrayzenoss> gwb2353: I can bump the version on the Sentry ZenPack when chudler sends it in, so you "upgrade" to whatever is the latest
[04-Aug-2009 16:09:33] <chudler> I suppose it would work. I have no idea.
[04-Aug-2009 16:10:16] <chudler> I dont know what actually happens when a zenpack is updated. If you install the next test one, it probably is better to install clean, otherwise its hard to know waht we are testing.
[04-Aug-2009 16:10:34] <chudler> But in the release version, do whatever matt says :-)
[04-Aug-2009 16:10:56] <gwb2353> no worries; i'm just itching to see what the power draw is in our DC
[04-Aug-2009 16:11:04] <gwb2353> (current draw)
[04-Aug-2009 16:14:39] <chudler> gwb2353: could you tell me the vendor of those?
[04-Aug-2009 16:16:54] <gwb2353> http://products.servertech.com/products/smartpower/
[04-Aug-2009 16:17:51] <chudler> thats good. some resellers puts the sentry3 mib under their oid, which the pack handles but it has to be added if I don't already know about it. I think this vendor is fine.
[04-Aug-2009 16:49:21] <jb> that was odd
[04-Aug-2009 16:49:21] <jb>
[04-Aug-2009 21:47:55] <duffyd> hi all, I'm getting this error when trying to run the stack installer: bash: ./zenoss-stack-2.4.2-linux.bin: No such file or directory
[04-Aug-2009 21:48:03] <duffyd> which is really strange given that the file is there and the perms are all fine
[04-Aug-2009 21:48:10] <duffyd> is there some particular pre-reqs I need
[04-Aug-2009 21:48:15] <duffyd> I'm running ubuntu 8.04
[04-Aug-2009 22:02:07] edwardam1 is now known as edwardam
[04-Aug-2009 22:10:37] <duffyd> what are the pre-reqs for the stack installer?
[05-Aug-2009 00:00:46] [disconnected at Wed Aug 5 00:00:46 2009]
[05-Aug-2009 00:00:46] [connected at Wed Aug 5 00:00:46 2009]
[05-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[05-Aug-2009 02:37:47] <duffyd> hi what would be the best way to setup a snmp service via an ssh tunnel?
[05-Aug-2009 02:39:56] <duffyd> i.e. zenoss uses the ssh tunnel to check the snmp services on a remote box
[05-Aug-2009 03:07:23] <duffyd> does zenoss use snmp for all its linux device communication?
[05-Aug-2009 03:24:24] <rhett> you can do monitoring over ssh
[05-Aug-2009 03:24:36] <rhett> i personally use ssh though for all my hosts
[05-Aug-2009 03:30:28] <duffyd> rhett: really
[05-Aug-2009 03:30:34] <duffyd> how do you set that up
[05-Aug-2009 03:30:44] <duffyd> all the stuff I've read indicates its a maintenance nightmare
[05-Aug-2009 03:34:37] <rhett> oh sorry. i mean to say i use snmp for all my hosts
[05-Aug-2009 03:35:04] <rhett> i've only used ssh monitoring once on a test vm and didnt really go much further with it
[05-Aug-2009 03:42:17] <duffyd> rhett: ok thanks for that
[05-Aug-2009 08:12:09] <twm1010> duffyd: still around?
[05-Aug-2009 08:12:17] <duffyd> aye
[05-Aug-2009 08:12:53] <duffyd> twm1010:
[05-Aug-2009 08:12:55] <twm1010> as far as pre-reqs for stack, not really any different than from source
[05-Aug-2009 08:13:24] <twm1010> the whole point of stack, is that it comes with everything ya need
[05-Aug-2009 08:13:46] <twm1010> now, hardware pre-reqs are something else, a zenoss box should have some decent CPU and ram, and fast disk
[05-Aug-2009 08:17:19] <Dieterbe> is it just me or don't event transformation rules get inherited from the root class if your class is /Unknown?
[05-Aug-2009 08:20:57] <duffyd> twm1010: yeah my issue was that I was using the 32-bit installer on a 64-bit machine
[05-Aug-2009 08:31:49] <klinstifen> morning all
[05-Aug-2009 08:32:15] <klinstifen> my events db has grown way too big
[05-Aug-2009 08:32:28] <klinstifen> the back up last night has events.sql at ~12GB
[05-Aug-2009 08:32:42] <twm1010> that'll do it every time
[05-Aug-2009 08:32:43] <klinstifen> does the "clear event cache" in the zenoss console flush the entire db?
[05-Aug-2009 08:32:51] <twm1010> no, i don't think so
[05-Aug-2009 08:32:54] <klinstifen> hrm
[05-Aug-2009 08:33:20] <twm1010> there is a line: Delete Historical Events Older Than (days)
[05-Aug-2009 08:33:41] <twm1010> what i don't know, is when it does that, and if it truncates the DB afterwards
[05-Aug-2009 08:33:47] <twm1010> probably not
[05-Aug-2009 08:36:25] <klinstifen> my previous backups were only 300MB
[05-Aug-2009 08:36:29] <klinstifen> but that's the zipped file
[05-Aug-2009 08:36:39] <klinstifen> i am restoring one now to see how big it is uncompressed
[05-Aug-2009 08:36:57] <klinstifen> i can't complete the backup job now because i run out of disk space
[05-Aug-2009 08:37:12] <klinstifen> and then zenoss ends up dying later because it can't write to the temp file : /
[05-Aug-2009 08:37:52] <twm1010> let me see how big mine is
[05-Aug-2009 08:38:03] <jb> klinstifen: you can specify the temp file..
[05-Aug-2009 08:38:05] <klinstifen> cool thx
[05-Aug-2009 08:38:07] <klinstifen> yeah
[05-Aug-2009 08:38:08] <jb> if it is filling /tmp up
[05-Aug-2009 08:38:16] <klinstifen> it is
[05-Aug-2009 08:38:22] <jb> but it still uses /tmp in my experience
[05-Aug-2009 08:38:26] <jb> i think it may be a bug in 2.4.2
[05-Aug-2009 08:38:47] <klinstifen> the zenbackup is writing to /tmp
[05-Aug-2009 08:39:02] <klinstifen> "/tmp/tmpmnP5Ux/zenbackup
[05-Aug-2009 08:39:14] <klinstifen> and in there the events.sql is ~12GB
[05-Aug-2009 08:39:27] <jb> are you specifying --temp-dir on zenbackup?
[05-Aug-2009 08:39:32] <klinstifen> let me check
[05-Aug-2009 08:39:53] <klinstifen> no i am not
[05-Aug-2009 08:39:58] <jb> give that a try
[05-Aug-2009 08:40:04] <klinstifen> just the standard "zenbackup --save-mysql-access"
[05-Aug-2009 08:40:08] <twm1010> hrmm... I installed stack to /apps/zenoss
[05-Aug-2009 08:40:12] <twm1010> where would i find my events.sql?
[05-Aug-2009 08:40:54] <twm1010> found this: -rw-rw---- 1 mysql mysql 1881145344 Aug 5 08:40 ibdata1
[05-Aug-2009 08:41:04] <jb> events.sql is just a dump
[05-Aug-2009 08:41:06] <jb> that zenbackup creates
[05-Aug-2009 08:41:22] <klinstifen> you could see how big the events db is by going into mysql
[05-Aug-2009 08:41:28] <klinstifen> assuming you are running mysql
[05-Aug-2009 08:41:39] <twm1010> well im running the mysql instance thats part of the stack
[05-Aug-2009 08:41:56] <klinstifen> yeah
[05-Aug-2009 08:42:04] <klinstifen> so am i
[05-Aug-2009 08:42:14] <twm1010> running a backup right now... let's see
[05-Aug-2009 08:42:52] <klinstifen> twm1010: mysql -u zenoss -p
[05-Aug-2009 08:42:58] <klinstifen> and then the password you created during the install
[05-Aug-2009 08:44:06] <klinstifen> http://forums.mysql.com/read.php?108,201578,201578
[05-Aug-2009 08:44:13] <klinstifen> just ran that command against my events db
[05-Aug-2009 08:44:21] <klinstifen> and it is only ~3GB
[05-Aug-2009 08:44:29] <klinstifen> yet the dumb is ~12GB?
[05-Aug-2009 08:45:03] <klinstifen> restore complete
[05-Aug-2009 08:45:15] <klinstifen> untar'ing
[05-Aug-2009 08:48:31] <mrayzenoss> Good morning
[05-Aug-2009 08:48:35] <ckrough> sup
[05-Aug-2009 08:49:02] <mrayzenoss> in case you missed it, another new WMI ZenPack last night: http://www.zenoss.com/community/projects/zenpacks/wmiperf_iis6
[05-Aug-2009 08:49:27] <twm1010> excellent, very nice.
[05-Aug-2009 08:49:39] <twm1010> I might have a chance to use that... I spent yesterday getting our Apache monitoring going
[05-Aug-2009 08:50:43] <twm1010> Matt: I've never seen WMI as a data source type?
[05-Aug-2009 08:50:49] <twm1010> will that zenpack work on core?
[05-Aug-2009 08:50:54] <jb> yes
[05-Aug-2009 08:51:00] <jb> egor created the WMI datasource type
[05-Aug-2009 08:51:07] <jb> it depends on his WMI pack
[05-Aug-2009 08:51:54] <twm1010> wait, so if i install his WMI zenpack I don't have to write all these damn perl scripts for getting WMI information?
[05-Aug-2009 08:52:00] <twm1010> *bangs head on desk*
[05-Aug-2009 08:52:12] <jb> http://www.zenoss.com/community/projects/zenpacks/wmidatasource
[05-Aug-2009 08:52:18] <jb> correct
[05-Aug-2009 08:52:20] <jb> just the WMI queries
[05-Aug-2009 08:52:23] <twm1010> <-- jackass
[05-Aug-2009 08:52:42] <twm1010> sigh... well, at least i learned something
[05-Aug-2009 08:52:46] <jb>
[05-Aug-2009 08:52:47] <ckrough> there ya go
[05-Aug-2009 08:53:04] <ckrough> ... if I could take make all of the things I've done the hard way...
[05-Aug-2009 08:54:31] <twm1010> klinstifen: how long did yuor backup take?
[05-Aug-2009 08:54:39] <twm1010> my timeout is set to 180seconds and it still times out
[05-Aug-2009 08:55:16] <klinstifen> let me check some time stamps
[05-Aug-2009 08:55:39] <klinstifen> let's see...it starts at 12:20AM
[05-Aug-2009 08:56:13] <klinstifen> haha i mean 8:00PM
[05-Aug-2009 08:56:24] <klinstifen> and the modify stamp says 9:03
[05-Aug-2009 08:56:27] <klinstifen> so...about an hour
[05-Aug-2009 08:57:22] <twm1010> wow
[05-Aug-2009 08:57:23] <twm1010> ok
[05-Aug-2009 08:57:32] <twm1010> WMI datasource zenpack fails install: ERROR: zenpack command failed. Reason: [Errno 20] Not a directory: '/apps/zenoss/zenoss/Products/COPYRIGHT.txt/skins'
[05-Aug-2009 08:59:00] <ckrough> anyone here running rancid+zenoss integration?
[05-Aug-2009 09:04:48] <klinstifen> mray: my events db has grown too big..if I use the "clear event cache" will it flush the entire db?
[05-Aug-2009 09:05:24] <klinstifen> jb: i would specify another path but there is only one disk on my zenoss box
[05-Aug-2009 09:05:50] <jb> oh.. /tmp is the same filesystem as everything else?
[05-Aug-2009 09:06:12] <klinstifen> yeah
[05-Aug-2009 09:06:15] <jb> gotcha
[05-Aug-2009 09:06:23] <jb> then you have a problem
[05-Aug-2009 09:06:39] <klinstifen> yeah
[05-Aug-2009 09:06:47] <twm1010> can you do an NFS mount to something else?
[05-Aug-2009 09:07:20] <klinstifen> i could
[05-Aug-2009 09:07:25] <klinstifen> and for the sake of the backups that might work
[05-Aug-2009 09:07:41] <klinstifen> i'm just surprised that the events.sql is so large
[05-Aug-2009 09:07:48] <klinstifen> when the events db itself doesnt' seem to be so big
[05-Aug-2009 09:07:50] <mrayzenoss> klinstifen: you can get into your events DB and clean it up
[05-Aug-2009 09:07:56] <mrayzenoss> http://www.zenoss.com/Members/netdata/how-to-clean-your-events-database/
[05-Aug-2009 09:08:27] <klinstifen> mray: ok that's what i was thinking
[05-Aug-2009 09:09:53] <klinstifen> should the events.sql be about the same size as the actual events db?
[05-Aug-2009 09:10:58] <mrayzenoss> well, the events DB has several tables (http://dev.zenoss.org/trac/browser/branches/zenoss-2.4.x/Products/ZenEvents/db/zenevents.sql)
[05-Aug-2009 09:11:13] <mrayzenoss> and there will probably be a lot more in history than in status
[05-Aug-2009 09:17:04] <Dieterbe> how should i check the memory consumption of processes matching a certain regex on a windows box? i can't use zenoss process monitoring because zenoss will add up the usages of all processes that match the regex together. and i want to use WMI
[05-Aug-2009 09:17:33] <klinstifen> mray: ok...i was using this to calculate size of entire events db
[05-Aug-2009 09:17:35] <klinstifen> http://forums.mysql.com/read.php?108,201578,201578#msg-201578
[05-Aug-2009 09:18:23] <klinstifen> maybe that isn't counting the right information?
[05-Aug-2009 09:19:24] <klinstifen> jb & twm1010: looks like i have to use an nfs mount (or find other things to delete from this box) because the restore of my backup from a week ago is pretty large as well
[05-Aug-2009 09:19:48] <klinstifen> the tar file is small (300MB) but untarred it's pretty big
[05-Aug-2009 09:23:13] <klinstifen> mray: woops, bad math on my part...the events db is ~3gb and the events.sql is ~1.2GB
[05-Aug-2009 09:23:31] <klinstifen> the big file is the .fs file
[05-Aug-2009 09:27:26] <mrayzenoss> the .fs is the config DB, zope's object database
[05-Aug-2009 09:32:18] <Dieterbe> anyone knows the answer to my question?
[05-Aug-2009 09:34:11] <mrayzenoss> Dieterbe: there is no process monitoring via WMI yet, only service monitoring
[05-Aug-2009 09:35:28] <Dieterbe> well, i'm okay with writing my own script / datasource command
[05-Aug-2009 09:35:55] <Dieterbe> as long as i can pull the values from the windows box, and use the data in zenoss for graphing and alerting
[05-Aug-2009 09:35:58] <mrayzenoss> then you'd probably want to check out the new WMI data source
[05-Aug-2009 09:36:31] <jb> http://www.zenoss.com/community/projects/zenpacks/wmidatasource
[05-Aug-2009 09:36:43] <klinstifen> or use wmic?
[05-Aug-2009 09:37:13] <mrayzenoss> klinstifen: the new WMI data source should replace the need for wmic
[05-Aug-2009 09:37:26] <mrayzenoss> I assume you can craft WQL queries to get the processes
[05-Aug-2009 09:37:43] <klinstifen> it's that so? hmm...must download
[05-Aug-2009 09:38:06] <mrayzenoss> yeah, it's supposedly faster and handles credentials better than wmic-based stuff
[05-Aug-2009 09:38:11] <mrayzenoss> same calls under the covers
[05-Aug-2009 09:38:13] <klinstifen> awesome
[05-Aug-2009 09:38:18] <klinstifen> yeah wmic was odd
[05-Aug-2009 09:38:26] <klinstifen> you had to flip the slash in domain/username
[05-Aug-2009 09:38:46] <klinstifen> even though zenoss used it the other way for grabbing sys logs
[05-Aug-2009 09:38:48] <klinstifen> cool
[05-Aug-2009 09:38:52] <klinstifen> screenshots look promising
[05-Aug-2009 09:39:01] <klinstifen> i'll have to try this out and convert some of my personal zenpacks
[05-Aug-2009 09:40:15] <Dieterbe> hmm i already use 2 wmi zenpacks ( http://blog.dastrup.com/?p=13 and http://www.zenoss.com/community/projects/zenpacks/windowswmidevicetemplate ) won't this new one conflict/break anyone in my current setup?
[05-Aug-2009 09:40:17] <mrayzenoss> gotta run, be back in a bit
[05-Aug-2009 09:40:48] <mrayzenoss> Dieterbe: I believe the new WMI ones from Egor Puzanov are meant to replace those
[05-Aug-2009 09:41:08] <mrayzenoss> so far there's a Windows Perf and an IIS Perf
[05-Aug-2009 09:41:30] <mrayzenoss> in addition to the WMI data source they both use
[05-Aug-2009 09:41:44] <mrayzenoss> so someone could do an Exchange, AD, whatever else you run on MS
[05-Aug-2009 09:42:31] <mrayzenoss> and unlike the 2 you mentioned, the new WMI-based ones should be faster and more secure (ie. no passwords in logs)
[05-Aug-2009 09:43:06] <mrayzenoss> gotta go
[05-Aug-2009 10:13:10] <Dieterbe> hmm odd, i'm googling with terms like 'wql memory usage', 'wql measure memory etc' but i don't have any good results :/ can anyone tell me where i should look?
[05-Aug-2009 10:13:20] <Dieterbe> (to measure memory usage of single processes with wmi/wql)
[05-Aug-2009 10:51:11] <venturaville> does anyone remember who was working on the IBM/LSI storage array module?
[05-Aug-2009 10:51:30] <ckrough> Egor?
[05-Aug-2009 10:51:54] <ckrough> jb as well
[05-Aug-2009 10:53:37] <venturaville> actually it was the IBM SVC we were discussing
[05-Aug-2009 10:53:43] <venturaville> and it was zenChild
[05-Aug-2009 10:53:48] <venturaville> doesn't look like he is on now though
[05-Aug-2009 10:56:33] <jb> ventura
[05-Aug-2009 10:56:39] <jb> im TRYING to work on IBM stuff right now
[05-Aug-2009 10:56:49] <jb> im about to scream though.. IBM director is driving me bonkers.
[05-Aug-2009 10:57:12] <jb> but i'd be interested in what you are tryign to do with SVC as well
[05-Aug-2009 10:58:24] <venturaville> zenChild is working on an IBM SVC zenpack: http://github.com/zenchild/ZenPacks.ndgov.IBMsvc/tree/master
[05-Aug-2009 10:58:49] <venturaville> I was hoping to get caught up with someone's efforts now that I have the time
[05-Aug-2009 10:58:51] <jb> nice!
[05-Aug-2009 10:59:05] <jb> well, egor and I are working on an IBMMon zenpack
[05-Aug-2009 10:59:11] <venturaville> I have a list of commands to extract data from SMcli
[05-Aug-2009 10:59:14] <jb> do you have any experience with IBM Director Agents?
[05-Aug-2009 10:59:44] <venturaville> zilch
[05-Aug-2009 10:59:50] <venturaville> we mostly use Storage Manager
[05-Aug-2009 11:00:07] <jb> yeah
[05-Aug-2009 11:00:10] <jb> we are a SVC shop here
[05-Aug-2009 11:00:13] <jb> that pack looks promising
[05-Aug-2009 11:00:17] <jb> it uses SNMP?
[05-Aug-2009 11:00:17] <venturaville> we have SVCs
[05-Aug-2009 11:00:21] <venturaville> sadly no
[05-Aug-2009 11:00:35] <venturaville> the SVCs and LSI based arrays are SNMP dumb
[05-Aug-2009 11:00:48] <jb> cimon?
[05-Aug-2009 11:00:56] <venturaville> and it looks as though going through the CIMON agents doesn't give me performance data
[05-Aug-2009 11:01:03] <venturaville> only structural and config data
[05-Aug-2009 11:01:15] <jb> how is zenchild modeling the existing data?
[05-Aug-2009 11:01:32] <venturaville> svcinfo commands from the looks of things
[05-Aug-2009 11:01:43] <jb> via ssh?
[05-Aug-2009 11:01:53] <venturaville> not sure
[05-Aug-2009 11:01:59] <venturaville> I'll download the update and take a look
[05-Aug-2009 11:02:03] <jb> k
[05-Aug-2009 11:02:11] <jb> i despise this IBM director bs
[05-Aug-2009 11:02:15] <venturaville> +1
[05-Aug-2009 11:02:22] <jb> why can't they have a simple solution like HP/Dell
[05-Aug-2009 11:02:23] <jb> GRR
[05-Aug-2009 11:02:33] <jb> anything IBM has to be so damn complicated..
[05-Aug-2009 11:02:39] <venturaville> I got some SMcli commands for the arrays themselves from the LSI folk
[05-Aug-2009 11:02:45] <venturaville> if you have access to storage manager
[05-Aug-2009 11:02:53] <jb> i actually was just upgrading my SM
[05-Aug-2009 11:02:56] <venturaville> as for the SVCs we might be stuck with direct ssh access for perf data
[05-Aug-2009 11:03:29] <venturaville> take a look at yawn if you have not used it before
[05-Aug-2009 11:03:36] <venturaville> I think zenChild was the one who pointed me at it
[05-Aug-2009 11:03:40] <jb> yawn?
[05-Aug-2009 11:03:51] <venturaville> http://sourceforge.net/apps/mediawiki/pywbem/index.php?title=YAWN
[05-Aug-2009 11:04:05] <jb> ahh
[05-Aug-2009 11:04:12] <venturaville> works pretty well
[05-Aug-2009 11:04:24] <venturaville> use root/lsissi
[05-Aug-2009 11:06:36] <venturaville> just need to point it at one of your CIMOM servers
[05-Aug-2009 11:09:25] <jb> k.. i'll try that in a bit
[05-Aug-2009 11:28:14] <jb> hi egor..
[05-Aug-2009 11:42:02] <otakup0pe> anyone good with the web services around
[05-Aug-2009 11:42:17] <otakup0pe> how does one easily query devices for os/status/hw info
[05-Aug-2009 11:53:56] <twm1010> anyone had issues installing the WMI datasource ZP on stack?
[05-Aug-2009 12:01:49] <klinstifen> mine went through without issue
[05-Aug-2009 12:02:02] <klinstifen> gonna test it out as soon
[05-Aug-2009 12:10:09] <twm1010> hrmm, i'll try it on my dev machine
[05-Aug-2009 12:11:31] <twm1010> same error on stack install on SLES as on RHEL
[05-Aug-2009 12:50:34] <venturaville> jb: which director version are you using?
[05-Aug-2009 12:50:46] <jb> trying to use 6.11
[05-Aug-2009 12:50:49] <jb> platform agent
[05-Aug-2009 12:50:59] <venturaville> SNMPv2-SMI::enterprises.2.6.159.200.3.1.1.1.1.1 = STRING: "International Business Machines Corporation"
[05-Aug-2009 12:50:59] <venturaville> SNMPv2-SMI::enterprises.2.6.159.200.3.1.1.1.2.1 = STRING: "IBM Director Server"
[05-Aug-2009 12:50:59] <venturaville> SNMPv2-SMI::enterprises.2.6.159.200.3.1.1.1.3.1 = STRING: "4.20"
[05-Aug-2009 12:51:03] <jb> its not registering alerts/subscription with the cimom automatically
[05-Aug-2009 12:51:34] <jb> and a bunch of other issues
[05-Aug-2009 12:56:09] <venturaville> apparently our CIMOM agent is running on a nonstandard port for SVC
[05-Aug-2009 13:02:09] <mrayzenoss> otakup0pe: I tried to do stuff with web services, but it's easier to just put your own Python scripts on Zope
[05-Aug-2009 13:02:30] <mrayzenoss> you can easily expose whatever you want as blocks of HTML or whatever
[05-Aug-2009 13:04:22] <mrayzenoss> examples are here: http://zenossdeki.svn.sourceforge.net/viewvc/zenossdeki/zenoss/
[05-Aug-2009 13:04:57] <Garn32> When ever I attempt to create a backup using the GUI it fails with raise child_exception
[05-Aug-2009 13:04:57] <Garn32> OSError
[05-Aug-2009 13:04:57] <Garn32> :
[05-Aug-2009 13:04:57] <Garn32> [Errno 2] No such file or directory
[05-Aug-2009 13:05:24] <Garn32> is the backup function broken in the GUI of what?
[05-Aug-2009 13:07:23] <mrayzenoss> Garn32: which version?
[05-Aug-2009 13:08:03] <Garn32> 2.4.2
[05-Aug-2009 13:08:26] <mrayzenoss> not that I'm aware of, nothing in trac for that
[05-Aug-2009 13:08:37] <mrayzenoss> which installer?
[05-Aug-2009 13:08:45] <Garn32> its from source
[05-Aug-2009 13:09:24] <mrayzenoss> ahh, let me try my source installs
[05-Aug-2009 13:12:02] <mrayzenoss> mine works, Debian 5 x86 from source: http://pastebin.com/d8d19eec
[05-Aug-2009 13:12:11] <mrayzenoss> 2.4.2 from SVN
[05-Aug-2009 13:12:20] <mrayzenoss> pastebin the whole stack
[05-Aug-2009 13:12:34] <Garn32> Output:
[05-Aug-2009 13:12:35] <Garn32> /usr/local/zenoss/bin/zenbackup -v10 --save-mysql-access
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:Use /tmp/tmpt1BGMx/zenbackup as a staging directory for the backup
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:Can we access ZODB through Zeo?
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:/usr/local/zenoss/bin/python /usr/local/zenoss/bin/zeoup.py -p 8100 -h localhost
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:Elapsed time: 0.50
[05-Aug-2009 13:12:37] <Garn32> INFO:zenbackup:Getting MySQL dbname, user, password from ZODB.
[05-Aug-2009 13:12:39] <Garn32> DEBUG:zenbackup:Writing MySQL credentials to /tmp/tmpt1BGMx/zenbackup/backup.settings
[05-Aug-2009 13:12:41] <Garn32> INFO:zenbackup:Backing up events database.
[05-Aug-2009 13:12:43] <Garn32> DEBUG:zenbackup:mysqldump -uzenoss ***************** --single-transaction --routines events --result-file=/tmp/tmpt1BGMx/zenbackup/events.sql
[05-Aug-2009 13:12:46] <Garn32> Traceback (most recent call last):
[05-Aug-2009 13:12:48] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 420, in ?
[05-Aug-2009 13:12:50] <Garn32> if zb.makeBackup():
[05-Aug-2009 13:12:54] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 388, in makeBackup
[05-Aug-2009 13:12:56] <Garn32> self.backupEventsDatabase()
[05-Aug-2009 13:12:58] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 280, in backupEventsDatabase
[05-Aug-2009 13:13:00] <Garn32> (output, warnings, returncode) = self.runCommand(cmd, obfuscated_cmd)
[05-Aug-2009 13:13:02] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 62, in runCommand
[05-Aug-2009 13:13:04] <Garn32> proc = Popen(cmd, stdout=PIPE, stderr=PIPE)
[05-Aug-2009 13:13:06] <Garn32> File "/usr/local/zenoss/lib/python2.4/subprocess.py", line 543, in __init__
[05-Aug-2009 13:13:08] <Garn32> errread, errwrite)
[05-Aug-2009 13:13:10] <Garn32> File "/usr/local/zenoss/lib/python2.4/subprocess.py", line 975, in _execute_child
[05-Aug-2009 13:13:12] <Garn32> raise child_exception
[05-Aug-2009 13:13:14] <Garn32> OSError
[05-Aug-2009 13:13:16] <Garn32> :
[05-Aug-2009 13:13:18] <Garn32> [Errno 2] No such file or directory
[05-Aug-2009 13:13:20] <Garn32> DONE
[05-Aug-2009 13:13:24] <Garn32> sorry
[05-Aug-2009 13:14:29] <Garn32> http://pastebin.com/m3b9abe6b
[05-Aug-2009 13:14:40] <Garn32> for the output given to me
[05-Aug-2009 13:19:46] <mrayzenoss> hmm... is there a $ZENHOME/backups directory?
[05-Aug-2009 13:19:57] <Garn32> yes that directory exists
[05-Aug-2009 13:20:06] <mrayzenoss> and it's owned by the 'zenoss' user?
[05-Aug-2009 13:20:23] <Garn32> yes
[05-Aug-2009 13:20:57] <Garn32> the other odd thing is that the command line zenbackup works just fine
[05-Aug-2009 13:21:04] <mrayzenoss> that is odd
[05-Aug-2009 13:21:37] <mrayzenoss> so if you use the same /usr/local/zenoss/bin/zenbackup -v10 --save-mysql-access call as the zenoss user, everything works?
[05-Aug-2009 13:21:58] <Garn32> let me check again real quick
[05-Aug-2009 13:22:40] <Garn32> yep ran just fine
[05-Aug-2009 13:23:15] <mrayzenoss> I'll run it by the devs, but maybe open a ticket.
[05-Aug-2009 13:23:47] <Garn32> ok I'll open a ticket
[05-Aug-2009 13:23:57] <mrayzenoss> put that stack trace in there as well
[05-Aug-2009 13:24:06] <mrayzenoss> you are running 2.4.2 right, not trunk?
[05-Aug-2009 13:24:27] <mrayzenoss> just have to check with source installs
[05-Aug-2009 13:25:05] <twm1010> mrayzenoss: is there a more verbose way to troubleshoot a zenpack install?
[05-Aug-2009 13:25:06] <Garn32> not trunk
[05-Aug-2009 13:25:22] <mrayzenoss> if you go to Settings->Versions you should see the SVN number after the Zenoss version
[05-Aug-2009 13:25:25] <mrayzenoss> put that in the ticket too
[05-Aug-2009 13:25:28] <Garn32> ok
[05-Aug-2009 13:26:39] <mrayzenoss> twm1010: zenpack -v10 ZenPacks.community.yourzenpack.egg
[05-Aug-2009 13:27:40] <twm1010> i mean to say, that won't install
[05-Aug-2009 13:27:52] <mrayzenoss> pastebin the error
[05-Aug-2009 13:28:25] <mrayzenoss> Garn32: SELinux?
[05-Aug-2009 13:28:36] <Garn32> its on a debain system
[05-Aug-2009 13:28:44] <mrayzenoss> yeah, me too
[05-Aug-2009 13:28:46] <twm1010> http://pastebin.com/d11f04ac9
[05-Aug-2009 13:29:08] <Garn32> no SELinux installed
[05-Aug-2009 13:29:32] <mrayzenoss> twm1010: your .egg got renamed as a .zip
[05-Aug-2009 13:32:30] <twm1010> let me test with the different extension
[05-Aug-2009 13:32:38] <twm1010> thats the file as it came down from the zenpack site
[05-Aug-2009 13:33:41] <mrayzenoss> Garn32: any chance the path changed between starting zenoss and running the zenbackup from the command line?
[05-Aug-2009 13:33:58] <mrayzenoss> twm1010: are you using IE? it sometimes rewrites the extension
[05-Aug-2009 13:34:34] <Garn32> none that I can think of
[05-Aug-2009 13:35:04] <mrayzenoss> just passing on suggestions from kells and chet
[05-Aug-2009 13:35:19] <mrayzenoss> they surmise that mysqldump isn't in the zenoss application's path
[05-Aug-2009 13:36:40] <Garn32> ok, how would I go about fixing that?
[05-Aug-2009 13:37:27] <Garn32> should I still put the ticket in?
[05-Aug-2009 13:37:38] <mrayzenoss> here's my path settings
[05-Aug-2009 13:37:47] <mrayzenoss> export ZENHOME=/usr/local/zenoss
[05-Aug-2009 13:37:47] <mrayzenoss> export PYTHONPATH=$ZENHOME/lib/python
[05-Aug-2009 13:37:47] <mrayzenoss> export PATH=$ZENHOME/bin:$PATH
[05-Aug-2009 13:37:56] <twm1010> jeebus
[05-Aug-2009 13:37:56] <mrayzenoss> for my zenoss account
[05-Aug-2009 13:38:11] <twm1010> you are right, renamed to .egg, i guess did somehow end up using IE to get that...
[05-Aug-2009 13:38:25] <twm1010> sigh... all that... for that.
[05-Aug-2009 13:38:51] <mrayzenoss> Garn32: I followed this a bit as well: http://dev.zenoss.org/trac/browser/trunk/inst/docs/INSTALL_DEB-ETCH.txt
[05-Aug-2009 13:38:58] <mrayzenoss> it's a bit dated
[05-Aug-2009 13:39:25] <Garn32> mysqldump is in the zenoss users path
[05-Aug-2009 13:39:26] <mrayzenoss> but steps 4, 5, 6, 9, 10 and 11
[05-Aug-2009 13:39:34] <mrayzenoss> are worth double-checking
[05-Aug-2009 13:42:47] <Garn32> checking the sudo bit
[05-Aug-2009 13:45:14] <Garn32> nope everything looks fine from those steps
[05-Aug-2009 13:50:53] <Garn32> just put the trouble ticket it #5395
[05-Aug-2009 13:52:23] <mrayzenoss> thanks
[05-Aug-2009 13:55:49] <venturaville> mrayzenoss: sounds like the zenoss guys are swinging by our office next week :-)
[05-Aug-2009 13:55:59] <otakup0pe> mrayzenoss: yeah i realize that
[05-Aug-2009 13:56:08] <otakup0pe> i'll probably end up mixing it between that and submitting patches to the zenoss cats
[05-Aug-2009 13:56:10] <mrayzenoss> venturaville: who's that?
[05-Aug-2009 13:56:23] <ckrough> demand swag!
[05-Aug-2009 13:56:44] <venturaville> mrayzenoss: not sure who all is coming out ... sounds as though they are doing an enterprise demo then
[05-Aug-2009 13:57:08] <mrayzenoss> well, good luck to us!
[05-Aug-2009 13:58:02] <mrayzenoss> otakup0pe: yeah, I just found it easier to cheat and have Zope serve up exactly what I wanted rather than make a real REST call, parse it on the other side and then format it
[05-Aug-2009 14:01:21] <otakup0pe> mrayzenoss: yeah i'm kionda of the same opinion
[05-Aug-2009 14:01:40] <otakup0pe> when i say REST i really just mean "machine friendly web services with semanticly correct http response codes"
[05-Aug-2009 14:39:29] <straterra> How do I stop paging for one machine entirely?
[05-Aug-2009 14:39:53] <mrayzenoss> put it in a maintenance window
[05-Aug-2009 14:40:13] <mrayzenoss> create a Group for it, put the device in there and put a "Testing" maintenance window on just that device
[05-Aug-2009 14:40:19] <mrayzenoss> that never ends
[05-Aug-2009 14:40:31] <mrayzenoss> it'll still get monitored, just no alerts
[05-Aug-2009 14:40:35] <straterra> Hmm
[05-Aug-2009 14:40:52] <straterra> So..new group, drop the machine in there..and put a maintenance window on the group?
[05-Aug-2009 14:40:57] <mrayzenoss> yeah
[05-Aug-2009 14:41:22] <mrayzenoss> there may be other ways, that just popped into my head first
[05-Aug-2009 14:41:31] <mrayzenoss> you still want monitoring right?
[05-Aug-2009 14:42:01] <mrayzenoss> do you want events from it?
[05-Aug-2009 14:42:24] <straterra> I want monitoring..just no emails/pages
[05-Aug-2009 14:42:27] <straterra> At all
[05-Aug-2009 14:44:18] <mrayzenoss> I believe "Pre-Production" is monitoring with events with no alerts
[05-Aug-2009 14:44:52] <straterra> whats 'maintenance' ?
[05-Aug-2009 14:45:05] <mrayzenoss> heh, I knew you'd ask
[05-Aug-2009 14:45:20] <mrayzenoss> the Admin Guide has 3 of the 5
[05-Aug-2009 14:45:26] <mrayzenoss> I know I've opened a ticket for that somewhere
[05-Aug-2009 14:45:26] <straterra> Ah, ok
[05-Aug-2009 14:46:00] <mrayzenoss> there we go: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch08s02.html
[05-Aug-2009 14:46:56] <straterra> so..maintenance is it showing on the dashboard, monitoring, but just no paging
[05-Aug-2009 14:47:08] <mrayzenoss> yes
[05-Aug-2009 14:47:19] <straterra> (That wording is a bit..weird on the page)
[05-Aug-2009 14:47:25] <mrayzenoss> agreed
[05-Aug-2009 14:48:14] <straterra> So..for a maintenance window that never expires, I just make it reoccure daily..and last exactly 1 day long, right?
[05-Aug-2009 14:49:29] <mrayzenoss> yeah, that works. Seems there used to be a "Stop Production State" pulldown where you could set it to go into the same state
[05-Aug-2009 14:49:34] <mrayzenoss> maybe that was causing issues
[05-Aug-2009 14:50:52] <straterra> Thanks a lot
[05-Aug-2009 14:51:08] <mrayzenoss> no problem.
[05-Aug-2009 14:51:18] <straterra> Oh..one last thing. We have ZenOSS in a VM right now. Is it pretty easy to migrate everything to a real machine?
[05-Aug-2009 14:51:40] <mrayzenoss> yeah
[05-Aug-2009 14:52:08] <mrayzenoss> we should add that to the Install Guide
[05-Aug-2009 14:52:19] <straterra> Restoring from a backup?
[05-Aug-2009 14:52:31] <mrayzenoss> no, migrating from VM to other installers
[05-Aug-2009 14:52:35] <mrayzenoss> restoring is covered
[05-Aug-2009 14:52:40] <straterra> Ah, ok
[05-Aug-2009 14:52:45] <straterra> I'm not using the appliance
[05-Aug-2009 14:52:59] <straterra> I'm using a CentOS 5.3 VM...and the real machine will be CentOS 5.3 too
[05-Aug-2009 14:54:55] <ckrough> zenhub is finicky
[05-Aug-2009 14:56:10] <mrayzenoss> straterra: so that should be similar
[05-Aug-2009 14:56:16] <mrayzenoss> and we don't have great docs for that use case, but we should
[05-Aug-2009 14:56:21] <mrayzenoss> opening a ticket
[05-Aug-2009 14:56:37] <straterra> One last thing..for real this time, heh..Are the installed zenpacks located somewhere that I can easily copy them off for reinstallation before I import my backup?
[05-Aug-2009 14:56:59] <mrayzenoss> $ZENHOME/ZenPacks
[05-Aug-2009 14:57:56] <straterra> Oh, nifty. I do notice one of my zenpacks isn't in there though
[05-Aug-2009 14:58:02] <straterra> Not that I care..I don't use that one
[05-Aug-2009 14:58:11] <mrayzenoss> is it a .zip ZenPack?
[05-Aug-2009 14:58:18] <straterra> It may have been
[05-Aug-2009 14:58:39] <mrayzenoss> yeah, it probably installed it into $ZENHOME/Products, which is one of the reasons we stopped using them
[05-Aug-2009 14:59:19] <straterra> gotcha
[05-Aug-2009 14:59:37] <straterra> Would it be bad if I just didn't reinstall that zenpack, then?
[05-Aug-2009 14:59:56] <mrayzenoss> only if you have stuff that uses it
[05-Aug-2009 15:00:17] <venturaville> jb: if you get a chance to play with yawn, let me know if you come across any actual performance data... as near as I can tell it is getting it all from the array CIMOMs and not from SVC (darned if I can find it in the director CIMOM anyways)
[05-Aug-2009 15:21:29] <mrayzenoss> straterra: http://dev.zenoss.org/trac/ticket/5396
[05-Aug-2009 15:22:24] <mrayzenoss> straterra: it's quite involved if you really want to migrate everything
[05-Aug-2009 15:23:19] <ckrough> How can I avoid spikes in graphs when an interface counter rolls over due to a power outage or other abnormal event?
[05-Aug-2009 15:23:29] <straterra> Hmm :/
[05-Aug-2009 15:23:57] <straterra> Its involved to migrate from rpm install to rpm install?
[05-Aug-2009 15:24:21] <mrayzenoss> no, the most pain is 32 to 64 bit conversions
[05-Aug-2009 15:24:37] <straterra> Ah..that won't be happening
[05-Aug-2009 15:24:37] <mrayzenoss> which is typically what happens when people go VMware -> RPM
[05-Aug-2009 15:25:01] <straterra> I'm going from a heavily used (read as slow) VM server to a lowend 1U server
[05-Aug-2009 15:25:02] <twm1010> finally got that WMI datasource going, but queries don't work against any box
[05-Aug-2009 15:26:08] <mrayzenoss> straterra: here are the notes from professional services
[05-Aug-2009 15:26:17] <mrayzenoss> Here is a brief overview of some of the tools used for data import/export:
[05-Aug-2009 15:26:17] <mrayzenoss> • zendevicedump / zendeviceload Creates a XML file of device configuration information from the ZODB. Requires the same device classes on the target system as on the source system.
[05-Aug-2009 15:26:17] <mrayzenoss> • zendump / zenload Creates a XML file of the entire ZODB, including menu items. Requires the same version of Zenoss as well as the same ZenPacks on the target system as on the source system. Some XSL (or editing) is required to modify some paths that point to on-disk files.
[05-Aug-2009 15:26:17] <mrayzenoss> • zenbackup / zenrestore Used to recover in the event of a system failure. Requires the same version of Zenoss as well as the same ZenPacks on the target system as on the source system. This is capable of being used to migrate between OS versions on the same architecture (ie 32-bit to 32-bit, 64-bit to 64-bit)
[05-Aug-2009 15:26:21] <mrayzenoss> • rrdtool dump / import Used to convert RRD data to XML and back again.
[05-Aug-2009 15:26:23] <mrayzenoss> • mysqldump / mysql Used to export MySQL data from one system to another.
[05-Aug-2009 15:31:43] <straterra> I think I'd just use zenbackup
[05-Aug-2009 15:32:55] <mrayzenoss> yeah, same OS, same architecture, same path and same ZenPacks installed and you should be fine
[05-Aug-2009 15:33:30] <straterra> Yup yup yup and yup!
[05-Aug-2009 15:46:09] <chudler> I am curious, has there been any zenpacks or solutions for people with web environments running virtual hosts? I remember seeing an example zenpack a while back... did anyone run with it?
[05-Aug-2009 15:47:55] <chudler> it used to come up on the forum sometimes. its not trivial to monitor a large virtual hosting environment, since the IP addresses and hostnames are all shared by a single box
[05-Aug-2009 15:48:29] <mrayzenoss> yeah, I actually had a discussion with the dev working on VMware stuff regarding that
[05-Aug-2009 15:49:01] <mrayzenoss> it's not something we do well, so there may be some changes in the next release to make non-device things fit the model better
[05-Aug-2009 15:49:57] <chudler> that sounds really nice. for other things also. I might have a chance in the next few months to expand monitoring to our 1000+ virtual hosts, so I will either build it or keep waiting ;-)
[05-Aug-2009 15:50:18] <mrayzenoss> well, this would be for the Stone Crab release next Spring
[05-Aug-2009 15:50:31] <mrayzenoss> King Crab will be in beta in about a month
[05-Aug-2009 15:51:07] <chudler> can't wait. You guys don't sit still, thats for sure!
[05-Aug-2009 15:51:26] <mrayzenoss> the good news is we now have funding for another developer
[05-Aug-2009 15:53:06] <straterra> My software has releaes names of STD's..but crabs work too
[05-Aug-2009 15:53:38] <ckrough> you guys talking about VirtualHosts (httpd) or Virtual Hosts (vm/xen)?
[05-Aug-2009 15:54:06] <chudler> httpd
[05-Aug-2009 15:54:23] <mrayzenoss> chudler was talking httpd, but we realize we need to have a common way of representing multiple things under a single IP
[05-Aug-2009 15:54:56] <mrayzenoss> whether it's httpd, a cluster of databases or virtual hosts under ESX
[05-Aug-2009 15:55:16] <ckrough> gotcha
[05-Aug-2009 15:55:22] <chudler> ideally the devices could have templates, alert rules, and the whole nine "as if it was a device". But I am not holding my breath on that much of an overhaul
[05-Aug-2009 15:55:35] <ckrough> mac addresses are always unique...
[05-Aug-2009 15:56:00] <ckrough> nm
[05-Aug-2009 15:56:01] <mrayzenoss> ckrough: yeah, I think we tried that with VMware
[05-Aug-2009 15:56:08] <mrayzenoss> ckrough: but they don't have to be
[05-Aug-2009 15:56:23] <ckrough> same problem. I dont NAT with my vms... if you nat still have the same issue
[05-Aug-2009 15:56:29] <ckrough> and it does nothing for VIrtualHosts
[05-Aug-2009 15:56:34] <ckrough> fingers faster than brain
[05-Aug-2009 15:58:22] <mrayzenoss> yeah, I think we fixed an issue with VMware where we now support VMs with the same name and same IP address and same MACs under different ESX servers
[05-Aug-2009 16:08:37] <victor-> i've got a server that uses VPN to connect, each time launching an interface called 'tun0'. the problem is that zenoss is creating a unique 'tun0' interface in the device's "OS" tab. is there some way to only have zenoss recognize one 'tun0' device?
[05-Aug-2009 16:11:23] edwardam is now known as edwardam_phone
[05-Aug-2009 16:21:41] <rhett> hm, dont know about that one
[05-Aug-2009 16:22:01] <mrayzenoss> victor-: you could lock it from getting remodeled
[05-Aug-2009 16:23:44] <victor-> Ok, I was hoping that'll work. Thanks.
[05-Aug-2009 16:42:24] edwardam_phone is now known as edwardam
[05-Aug-2009 17:19:22] <mrayzenoss> just hit a very nice milestone: http://blog.zenoss.com/2009/08/05/zenoss-community-posts-100th-zenpack/
[05-Aug-2009 17:20:18] <rhett> nice
[05-Aug-2009 17:25:08] <locohost> gratz
[05-Aug-2009 22:10:06] <otakup0pe> awesome i managed to add new rest methods :3
[05-Aug-2009 22:10:11] <otakup0pe> who wants to know the secret
[05-Aug-2009 22:29:59] <rhett> :O
[05-Aug-2009 22:51:49] <otakup0pe> hehe zenoss doesn't like having the ZenModel product refreshed
[05-Aug-2009 22:51:54] <otakup0pe> but it makes testing web services sooo easy
[05-Aug-2009 23:25:23] <otakup0pe> http://i30.tinypic.com/v6n792.png from left to right load average, cpu load and memory
[06-Aug-2009 00:00:46] [disconnected at Thu Aug 6 00:00:46 2009]
[06-Aug-2009 00:00:46] [connected at Thu Aug 6 00:00:46 2009]
[06-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[06-Aug-2009 00:46:31] <rhett> is this second life
[06-Aug-2009 00:54:11] <sergeymasushko> rhett: what do you mean? 0o
[06-Aug-2009 01:50:37] <rhett> sergeymasushko [09:24] <otakup0pe> http://i30.tinypic.com/v6n792.png from left to right load average, cpu load and memory this
[06-Aug-2009 07:01:47] <roadies> what plugin am I missing to get the network map "Select Device or Map" to display
[06-Aug-2009 08:08:25] <otakup0pe> rhett: opensim, but close enough
[06-Aug-2009 08:27:09] <TBKDan> I'm trying to monitor a winservice on a particular box. When I go to add it on the device's OS tab, it finds it just fine in the "Add WinService" box, but when I click ok, it says "Could not find a WinService named ..." The service is definitely there on the server, and in /Services/WinServices. Any idea why it's saying it can't find it?
[06-Aug-2009 08:37:02] <twm1010> you want to add it to just the one box, right?
[06-Aug-2009 08:37:22] <TBKDan> Correct
[06-Aug-2009 08:37:51] <twm1010> I've been wondering how to do that, because if you change the zprop on the service, it monitors every instance
[06-Aug-2009 08:38:05] <TBKDan> Yup.
[06-Aug-2009 08:38:32] <TBKDan> I don't understand what the purpose of "Add WinService" is if it doesn't let you add ones that it has in it's list...
[06-Aug-2009 08:38:41] <twm1010> yeah, seems odd, but try this
[06-Aug-2009 08:38:47] <twm1010> go to services/winservice, find the one you want
[06-Aug-2009 08:38:54] <twm1010> click on it so it shows all instances
[06-Aug-2009 08:39:33] <twm1010> find the server you want, but click the name of the service under the "name" column
[06-Aug-2009 08:39:47] <TBKDan> Ahhhh
[06-Aug-2009 08:40:07] <TBKDan> That works for me
[06-Aug-2009 08:40:20] <twm1010> seems a little odd, maybe ask matt ray when he drops in, i've been meaning to figure that out myself
[06-Aug-2009 08:40:34] <TBKDan> As long as I have a way to add it to one specific device, I'm happy (even if it's somewhat of a roundabout way)
[06-Aug-2009 08:40:35] <TBKDan> Thanks!
[06-Aug-2009 08:41:00] <twm1010> dude, i feel like a doof
[06-Aug-2009 08:41:05] <twm1010> check this, go to that add winservice dialog
[06-Aug-2009 08:41:10] <twm1010> and double click in the entry box
[06-Aug-2009 08:41:27] <TBKDan> Well, that's what I've been doing.
[06-Aug-2009 08:41:39] <TBKDan> I select the service from that list, click ok, then it says that it can't find it
[06-Aug-2009 08:41:48] <twm1010> huh, weird
[06-Aug-2009 08:42:14] <TBKDan> My thoughts
[06-Aug-2009 08:42:19] <twm1010> mine does the same thing
[06-Aug-2009 08:42:22] <twm1010> if it makes you feel better
[06-Aug-2009 08:42:32] <twm1010> probably has to do with characters in the name of the service or something
[06-Aug-2009 08:42:39] <TBKDan> Spaces? lol
[06-Aug-2009 08:43:07] <twm1010> perhaps, i think the other method is easier anyway with a tabbed browser you could set 5 servers pretty quick
[06-Aug-2009 08:44:04] <TBKDan> But if I'm already on that device setting it up, it's kind of annoying.
[06-Aug-2009 08:44:19] <TBKDan> I should just be able to type in right there.
[06-Aug-2009 08:44:30] <TBKDan> And to add several services is annoying
[06-Aug-2009 08:48:15] <twm1010> indeed...
[06-Aug-2009 08:49:09] <twm1010> i think it does have to do with the name of the service
[06-Aug-2009 08:49:13] <twm1010> some work, some don't
[06-Aug-2009 08:52:11] <TBKDan> If the service name has spaces, it spazzes.
[06-Aug-2009 08:52:22] <TBKDan> The description doesn't matter, the service name does.
[06-Aug-2009 08:52:33] <TBKDan> I was able to add IISADMIN service.
[06-Aug-2009 08:53:05] <twm1010> I wonder can you enclose it in ''
[06-Aug-2009 08:54:14] <twm1010> I would wager there is a tracking bug/ticket for it already
[06-Aug-2009 08:54:42] <twm1010> I would rather see an entirely separate tab for services, that showed all services on the box
[06-Aug-2009 08:55:20] <twm1010> hah, there already is
[06-Aug-2009 08:55:26] <TBKDan> Details? lol
[06-Aug-2009 08:55:42] <twm1010> in the OS tab, under win services, there is a checkbox called "monitored"
[06-Aug-2009 08:55:45] <twm1010> its jsut a filter
[06-Aug-2009 08:55:49] <twm1010> uncheck it, and watch
[06-Aug-2009 08:55:57] <TBKDan> Ahhh
[06-Aug-2009 08:56:14] <TBKDan> Doesn't let you know if they're up or down though
[06-Aug-2009 08:56:24] <twm1010> right, because they're not monitored
[06-Aug-2009 08:56:25] <TBKDan> Though it does gives you startmode which is nice
[06-Aug-2009 08:56:31] <jb> morning..
[06-Aug-2009 08:56:36] <twm1010> once you monitor it, the checkbox should be checked
[06-Aug-2009 08:56:38] <TBKDan> Goooooooood morning
[06-Aug-2009 08:56:41] <TBKDan> yup
[06-Aug-2009 08:56:55] <TBKDan> So you can do it that way then
[06-Aug-2009 08:57:07] <TBKDan> Show all unmonitored and then click that service from there
[06-Aug-2009 08:57:26] <TBKDan> Not as convenient as the "Add Service" dialog, but still better than hunting it down for each service
[06-Aug-2009 08:57:26] <TBKDan>
[06-Aug-2009 08:57:38] <twm1010> right, hit show all as well, put checkboxes next to the ones you want to monitor hit the drop down and choose monitor enable
[06-Aug-2009 08:57:42] <twm1010> you could do all of them at once
[06-Aug-2009 08:58:22] <TBKDan> Ahhhhh there we go
[06-Aug-2009 08:58:26] <TBKDan> Ok, that makes me a happy camper
[06-Aug-2009 08:58:27] <TBKDan> Danke! lol
[06-Aug-2009 08:58:53] <twm1010> i wonder, if that works for osprocesses too
[06-Aug-2009 08:59:36] <twm1010> no, it does not, it needs a regex for that, but it does work for ports
[06-Aug-2009 09:03:29] <jb> what do you guys transform snmp_coldstart/alertsystemup to?
[06-Aug-2009 09:03:32] <jb> just curious
[06-Aug-2009 09:07:33] <roadies> what plugin is required for Firefox to allow the network map to work, aka "Select Device or Network" display
[06-Aug-2009 09:09:20] <twm1010> I haven't done anything with it yet.
[06-Aug-2009 09:09:33] <twm1010> I'd rather just move it to history
[06-Aug-2009 09:09:41] <twm1010> roadies: probably flash?
[06-Aug-2009 09:16:11] <roadies> twm1010: Thanks, got that on, I guess it doesnt work with Fedora running on a PPC *sign*
[06-Aug-2009 09:16:35] <TBKDan> roadies: It doesn't work for me on Windows either. Can't get it to display any nodes =/
[06-Aug-2009 09:17:35] <roadies> TBKDan: double *sign*
[06-Aug-2009 09:18:17] <roadies> TBKDan: Fedora on Lenovo laptops works great however :-D
[06-Aug-2009 09:19:09] <TBKDan> heh, it works fine for 192.168.0.0....
[06-Aug-2009 09:19:16] <TBKDan> Any other subnet, it doesn't.
[06-Aug-2009 09:21:22] <roadies> it only works to first set the device/net within the subnet view of zenoss, then U can browse by clicking on the device/nodes to get that device/node view
[06-Aug-2009 09:22:01] <TBKDan> Yeah, I've got it seeing a few more now, but nowhere near all the subnets that I have configured.
[06-Aug-2009 09:26:26] <roadies> I assume you slide the "number of hops" bar over... this is a limit I wish they would fix
[06-Aug-2009 09:29:08] <TBKDan> I have a device that has all these vlans connected to it, yet it only shows a few
[06-Aug-2009 09:41:10] <unkwon> hi
[06-Aug-2009 09:42:15] <unkwon> i need help
[06-Aug-2009 09:42:36] <jb> um.. you are going to need to be a bit more specific
[06-Aug-2009 09:43:32] <unkwon> i have installed zenoss server and i need monitoring BackupExec server with snmp trap send by the backup server
[06-Aug-2009 09:43:49] <unkwon> that's work perfeclty but i have problem with Jobsuccess
[06-Aug-2009 09:44:25] <unkwon> when the server zenoss receive the message and i have selected severity 0 (clear) the message as deleted from the history and status
[06-Aug-2009 09:44:25] <jb> whats the problem? is zenoss receiving the trap?
[06-Aug-2009 09:44:47] <unkwon> how i can leave the message with green color in the history
[06-Aug-2009 09:45:51] <unkwon> i have tested diffent solution with transformation but the message as deleted finaly
[06-Aug-2009 09:45:59] <jb> a clear will move to Event History..
[06-Aug-2009 09:47:43] <unkwon> if i leave the jobsuccess in /unknow class but i have the ZProperties in /unknown i clear the message stay in the history with green color background
[06-Aug-2009 09:48:00] <unkwon> it's not possible change just the background color
[06-Aug-2009 09:48:12] <unkwon> i need green color for this message
[06-Aug-2009 09:49:14] <jb> it will appear green in the event history.. you may be able to get the clear message not to move to history with a transform, but I'm not quite sure.
[06-Aug-2009 09:50:02] <unkwon> my server delete the message with clear severity if i assign in class
[06-Aug-2009 09:50:25] <unkwon> i have tested with 2.3 2.4 of zenoss and it's same
[06-Aug-2009 09:55:06] <otakup0pe> oh i forgot about this lil' party
[06-Aug-2009 09:59:39] <otakup0pe> so who likes performance metrics
[06-Aug-2009 09:59:56] <mrayzenoss> oh yeah!
[06-Aug-2009 10:00:07] <mrayzenoss> JP and I will be hosting the session today, he should be here in a second
[06-Aug-2009 10:00:34] <otakup0pe> the only question is what's the patch submission process
[06-Aug-2009 10:00:46] <otakup0pe> i meant the only question -i have- is
[06-Aug-2009 10:00:48] <mrayzenoss> other Zenoss devs may come and go, like cgibbons and npmccallum
[06-Aug-2009 10:02:11] <npmccallum> If I'm on IRC, I'm usually idling in here...
[06-Aug-2009 10:02:12] <mrayzenoss> patch submission is easy. Open a ticket, attach the patch. Fill out the contributor agreement and we patch it
[06-Aug-2009 10:02:22] <npmccallum> can't say I'm always around though
[06-Aug-2009 10:02:27] <explosiv0SX> I have a question: shall I just fire it off?
[06-Aug-2009 10:02:32] <jb> sure.
[06-Aug-2009 10:02:45] <otakup0pe> mrayzenoss: word i'll check out the contrib agreement
[06-Aug-2009 10:02:55] <npmccallum> explosiv0SX: go ahead
[06-Aug-2009 10:02:59] <mrayzenoss> otakup0pe: http://www.zenoss.com/forms/contribute
[06-Aug-2009 10:03:08] <explosiv0SX> What's are zenoss' plans re: rest/api or web services going forward?
[06-Aug-2009 10:03:24] <explosiv0SX> I see some xml methods in the source and I see methods that return JSON.
[06-Aug-2009 10:03:30] <explosiv0SX> which is the way forward?
[06-Aug-2009 10:03:34] <mrayzenoss> explosiv0SX: so far it's been kinda ad-hoc, we add things as needed
[06-Aug-2009 10:03:52] <mrayzenoss> we have talked about making it more rigorous and formal, which would be more useful for external users
[06-Aug-2009 10:04:17] <mrayzenoss> not SOAP formal, just standardizing it
[06-Aug-2009 10:04:24] <mrayzenoss> and being consistent
[06-Aug-2009 10:04:26] <explosiv0SX> right -- SOAP is a little heavy-handed.
[06-Aug-2009 10:04:35] <otakup0pe> consistency++
[06-Aug-2009 10:04:38] <explosiv0SX> REST -> XML or JSON is more than fine.
[06-Aug-2009 10:05:07] <explosiv0SX> so no plans to deprecate any of the JSON or XML methods?
[06-Aug-2009 10:05:20] <mrayzenoss> no, everybody uses them
[06-Aug-2009 10:05:33] <mrayzenoss> well, I don't know if every single one is in use
[06-Aug-2009 10:05:39] <explosiv0SX> great.
[06-Aug-2009 10:05:40] <mrayzenoss> but I haven't heard anything about deprecating
[06-Aug-2009 10:05:44] <npmccallum> explosiv0SX: if we do deprecate them, they are marked as such before there removal...
[06-Aug-2009 10:05:56] <otakup0pe> yeah that makes more sense than just leaving them in for ever.
[06-Aug-2009 10:06:03] <mrayzenoss> so you'll have at least a release to get ready
[06-Aug-2009 10:06:07] <explosiv0SX> great.
[06-Aug-2009 10:06:09] <npmccallum> explosiv0SX: but I don't know that we've ever actually removed a method
[06-Aug-2009 10:06:10] <explosiv0SX> thank you.
[06-Aug-2009 10:06:17] <explosiv0SX> doesn't appear so ;-)
[06-Aug-2009 10:06:37] <npmccallum> explosiv0SX: we're a bit like GNOME in that regard
[06-Aug-2009 10:06:43] <explosiv0SX> right on.
[06-Aug-2009 10:06:56] <mrayzenoss> we are planning on deprecating RHEL 4 with the next release, and Zip-style ZenPacks
[06-Aug-2009 10:07:05] <mrayzenoss> not web services, but deprecations
[06-Aug-2009 10:07:57] <npmccallum> and I assume when that happens some code will probably die off at some point, but most of it isn't really external stuff
[06-Aug-2009 10:08:22] <npmccallum> the most important APIs that people care about are usually device and event management
[06-Aug-2009 10:08:46] <otakup0pe> oh here's a question. was there already rest methods for total memory/swap and i just couldn't find them ?
[06-Aug-2009 10:09:26] <npmccallum> otakup0pe: I'm not sure what you mean... are you trying to get the current memory usage on a device?
[06-Aug-2009 10:09:38] <otakup0pe> no the total memory/swap installed on a device
[06-Aug-2009 10:09:59] <otakup0pe> i couldn't find rest methods so i ended up adding them myself heh
[06-Aug-2009 10:10:27] <otakup0pe> i.e. totalSwapString from operatingsystem.py (iirc)
[06-Aug-2009 10:11:53] <explosiv0SX> otakup0pe: you can get it like: find('myserver').os.exportXml(sys.stdout)
[06-Aug-2009 10:12:09] <explosiv0SX> but not sure if that helps because of that pesky output stream parameter.
[06-Aug-2009 10:13:00] <explosiv0SX> btw -- that was written for zendmd.
[06-Aug-2009 10:13:20] explosiv0SX is now known as rak
[06-Aug-2009 10:13:45] <ckrough> What is the recommended datasource type for ethernet traffic? COUNTER or DERIVE? I'm trying to avoid spikes when interface counters roll over unexpectedly.
[06-Aug-2009 10:15:22] <otakup0pe> rak: that answer didn't make sense ? i meant a rest method.
[06-Aug-2009 10:15:38] <otakup0pe> let me just make a patch and it will make sense :3
[06-Aug-2009 10:17:46] <rak> well, this leads to another question:
[06-Aug-2009 10:18:03] <rak> I'd love to get device info with: http://zenoss.corp.goldlasso.com/zport/dmd/Devices/Server/Linux/devices/myserver/exportXml
[06-Aug-2009 10:18:10] <rak> the method is accessed.
[06-Aug-2009 10:18:29] <rak> but an error is thrown saying it's missing a param.
[06-Aug-2009 10:18:36] <rak> This parameter is the output stream.
[06-Aug-2009 10:18:58] <rak> is there anything I can pass to this method so say -- use the current response stream?
[06-Aug-2009 10:19:22] <otakup0pe> http://pastebin.ca/1520574 was all i added btw
[06-Aug-2009 10:20:40] <otakup0pe> hmm exportXml throws not found on some devices and throws an error on other devices for me. but i'm at work where i only have access to the (outdated) vmware appliance
[06-Aug-2009 10:20:53] <rak> cool. I probably could do the same to expose the exportXml function..
[06-Aug-2009 10:21:06] <otakup0pe> yeah the trick is the freakin @rtype tag
[06-Aug-2009 10:21:20] <otakup0pe> that seems to let zope know about the method
[06-Aug-2009 10:22:43] <jplouis> ckrough: I believe zenoss uses DERIVE with a RRD Min of 0 for interfaces
[06-Aug-2009 10:23:32] <ckrough> jplouis: actually, reading up on it just now, the recommended method is COUNTER with a max.
[06-Aug-2009 10:24:02] <ckrough> jplouis: device with a min will get a log of unknowns during normal counter rollovers apparently
[06-Aug-2009 10:24:26] <ckrough> jplouis: derive I mean, not device
[06-Aug-2009 10:27:19] <jplouis> from the RRD docs "so for high bandwidth interfaces and a 32bit counter, DERIVE with min=0 is probably preferable"
[06-Aug-2009 10:31:10] <daMaestro> does anyone know how to create a report that takes interface stats from specific devices and give a total number of bits transferred in a given time period
[06-Aug-2009 10:31:24] <daMaestro> so, it will actually total the rrd data
[06-Aug-2009 10:33:02] <npmccallum> otakup0pe, rak: there is a better way to get things like total memory
[06-Aug-2009 10:33:10] <otakup0pe> npmccallum: hooray !
[06-Aug-2009 10:33:28] <npmccallum> otakup0pe, rak: http://<hostname>:8080/zport/dmd/Devices/Server/Linux/devices/<device>/hw/getProperty?id=totalMemory
[06-Aug-2009 10:33:35] <otakup0pe> nice
[06-Aug-2009 10:33:41] <otakup0pe> getProperty is that a zope thing ?
[06-Aug-2009 10:33:42] <npmccallum> you can do that with any property
[06-Aug-2009 10:33:56] <npmccallum> I don't recall if it is a zope thing or not
[06-Aug-2009 10:34:16] <rak> ok, that'll do nicely.
[06-Aug-2009 10:34:22] <npmccallum> jplouis: do you know if getProperty comes from zope? or is it a zenoss thing?
[06-Aug-2009 10:34:37] <otakup0pe> looks like a zope-cmf thing ?
[06-Aug-2009 10:34:48] <otakup0pe> word thanks man if we are ever at the same conference i'll buy you a beer
[06-Aug-2009 10:34:54] <jplouis> I believe from Zope
[06-Aug-2009 10:35:08] <npmccallum> otakup0pe: let me know what conference to show up to and I'll take you up on it
[06-Aug-2009 10:37:26] <ckrough> jplouis: Im running 64bit counters, take a look at the line in the doc right after the one you pasted, do you take that to mean 64bit with DERIVE and a max, or 64bit with COUNTER and a max
[06-Aug-2009 10:37:57] <npmccallum> otakup0pe: I'll be at OLF
[06-Aug-2009 10:39:00] <jplouis> I read it as COUNTER with a max
[06-Aug-2009 10:39:15] <mrayzenoss> otakup0pe: me too
[06-Aug-2009 10:39:32] <jb> where is OLF?
[06-Aug-2009 10:39:55] <jb> orlando?
[06-Aug-2009 10:40:06] <mrayzenoss> Ohio
[06-Aug-2009 10:40:09] <jb> ah
[06-Aug-2009 10:40:11] <mrayzenoss> Ohio Linux Fest
[06-Aug-2009 10:40:13] <jb> you guys coming to chi anytime soon?
[06-Aug-2009 10:40:18] <mrayzenoss> actually yes
[06-Aug-2009 10:40:26] <mrayzenoss> our Product Manager is headed up there
[06-Aug-2009 10:40:37] <mrayzenoss> not sure who else is going, he's talking to some customers
[06-Aug-2009 10:40:46] <mrayzenoss> jb: are you on the list?
[06-Aug-2009 10:41:08] <jb> hm, I haven't heard anything about it..
[06-Aug-2009 10:41:15] <jb> i went to a campit a few months ago
[06-Aug-2009 10:41:38] <jb> i'd be interested though..
[06-Aug-2009 10:42:09] <otakup0pe> eh i'm in montreal and in the middle of a release cycle so i won't be going to con's any time soon
[06-Aug-2009 10:42:16] <otakup0pe> heh that should have been a sad emoticon
[06-Aug-2009 10:42:26] <mrayzenoss> jb: I'll hook you up, he was looking for customers and core users in Chicago
[06-Aug-2009 10:42:33] <otakup0pe> btw i should have some video of my third gen sl/zenoss mashups once i figure out fraps
[06-Aug-2009 10:42:40] <otakup0pe> and remember how2machinima
[06-Aug-2009 10:42:43] <mrayzenoss> otakup0pe: awesome
[06-Aug-2009 10:42:46] <jb> mrayzenoss: yeah, pass him my contact info if you don't mind..
[06-Aug-2009 10:42:49] <jb> is it brandon?
[06-Aug-2009 10:43:28] <mrayzenoss> yeah
[06-Aug-2009 10:43:44] <jb> argh, i think i actually accidentally deleted an email about vmware from him yesterday
[06-Aug-2009 10:44:01] <mrayzenoss> I'll ask him to resend that too
[06-Aug-2009 10:44:18] <jb> thanks
[06-Aug-2009 10:46:47] <raddy> Hello Everybody
[06-Aug-2009 10:47:15] <raddy> Can anybody suggest a good regexp for monitoring postgres process?
[06-Aug-2009 10:48:10] <raddy> When i just use postgres, it shows lot of processes.
[06-Aug-2009 10:49:21] <npmccallum> isn't postmaster the name of the master process? (I'm just guessing here, its been a while since I used postgres...)
[06-Aug-2009 10:49:45] <npmccallum> woah, I was right: http://www.postgresql.org/docs/8.1/static/app-postmaster.html
[06-Aug-2009 10:53:52] <raddy> Monitoring master process was useless
[06-Aug-2009 10:54:08] <raddy> As the memory usage n'all very 2-5 MB
[06-Aug-2009 10:54:25] <raddy> Monitoring postgres process is important,
[06-Aug-2009 10:54:29] <ckrough> what are you trying to do?
[06-Aug-2009 10:57:52] <jb> however, if postmaster dies, you are SOL
[06-Aug-2009 10:59:47] <raddy> ckrough : I want to monitor postgres memory usage.
[06-Aug-2009 11:00:07] <ckrough> so your not looking for availability, your looking for usage. gotcha
[06-Aug-2009 11:04:27] <mrayzenoss> Just a heads up, the ZenPack Contest only has 9 days left
[06-Aug-2009 11:04:42] <mrayzenoss> so if you were planning on entering, please send it them in soon
[06-Aug-2009 11:04:56] <chudler_> thanks, I have one or two to put in at the last minute.
[06-Aug-2009 11:05:27] chudler_ is now known as chudler
[06-Aug-2009 11:12:11] <raddy> ckrough : I can monitor availability seperately
[06-Aug-2009 11:12:26] <jb> mrayzenoss: is there any interest in zenpacks that use zenwinperf?
[06-Aug-2009 11:12:40] <raddy> ckrough : I want to monitor performance metrics for postgres process.
[06-Aug-2009 11:13:02] <mrayzenoss> jb: not sure what to do with Enterprise-dependendent zenpacks
[06-Aug-2009 11:13:09] <jb> yeah..
[06-Aug-2009 11:19:20] <otakup0pe> here's another questin actually. what specifically does it mean when getRRDValue returns 204. that the last snmp query failed ?
[06-Aug-2009 11:20:03] <otakup0pe> oh wait it does that for 0 responses nm >_<
[06-Aug-2009 11:25:07] <mrayzenoss> thanks to everyone who stopped by, I'll have a transcript up later on today
[06-Aug-2009 11:25:23] <mrayzenoss> I'll still be around, but the devs are off the hook
[06-Aug-2009 11:31:37] <Matir> I'm having a weird issue with my Zenoss dashboard... all of the groups added show an event status of "Green", even though some devices in those groups have events that are neither clear nor acknowledged
[06-Aug-2009 11:36:27] <rhett> do all the device shave teh same production state
[06-Aug-2009 11:38:09] <Matir> no, not all devices do
[06-Aug-2009 11:38:16] <Matir> does it only show events if they are production?
[06-Aug-2009 11:41:21] <mrayzenoss> Matir: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch08s02.html has which production states are on the dashboard
[06-Aug-2009 11:41:37] <Matir> ok, thanks
[06-Aug-2009 11:41:42] <Matir> I'll see if that addresses it
[06-Aug-2009 11:43:22] <Matir> ok, I get it... Production States are more zenoss-centric and don't necessarily align with the production states we use internally
[06-Aug-2009 11:43:27] <Matir> thanks!
[06-Aug-2009 12:09:23] <rhett> i just noticed i typed out my last line terribly
[06-Aug-2009 12:25:10] <Diddi> rhett: quite funny, though (:
[06-Aug-2009 13:41:46] <daMaestro> what is the recommended upgrade path from a 2.2.4 instance?
[06-Aug-2009 13:41:56] <daMaestro> is it to upgrade to 2.3 then 2.4?
[06-Aug-2009 13:42:16] <daMaestro> because i just tried the 2.2.4 -> latest.. and that fails
[06-Aug-2009 13:43:06] <magnachef> I think so...there are docs somewhere on the zenoss website
[06-Aug-2009 13:47:18] <chudler> anyone heard of using Zenoss with either modbus or bacnet?
[06-Aug-2009 13:51:19] <chudler> I've got PHB's asking about it. Here comes a time sink
[06-Aug-2009 13:55:11] <rhett> haha
[06-Aug-2009 13:56:55] <mrayzenoss> daMaestro: http://www.zenoss.com/community/docs/installation-guide/2.4.2/ch08s01.html#d4e869
[06-Aug-2009 15:31:05] edwardam is now known as edwardam_Fooz
[06-Aug-2009 15:31:30] edwardam_Fooz is now known as edwardamFoodz
[06-Aug-2009 15:39:49] daMaestro|isBack is now known as daMaestro
[06-Aug-2009 16:01:21] edwardamFoodz is now known as edwardam
[06-Aug-2009 16:04:43] edwardam is now known as edwardam_FoodRea
[06-Aug-2009 17:06:49] edwardam_FoodRea is now known as edwardam
[06-Aug-2009 18:22:40] daMaestro|isBack is now known as daMaestro
[06-Aug-2009 18:22:46] <daMaestro> mrayzenoss, ping
[06-Aug-2009 18:22:55] <daMaestro> mrayzenoss, i see you were the first to comment on my bug: http://dev.zenoss.com/trac/ticket/3740
[06-Aug-2009 18:22:58] <mrayzenoss> pong
[06-Aug-2009 18:23:12] <daMaestro> mrayzenoss, it got closed.. even though (so far i'm upgraded to 2.3.3) it's not fixed
[06-Aug-2009 18:23:33] <daMaestro> it might be in 2.4 ... just about to make that move, but the point of the bug was not addressed in the bug that was marked as related
[06-Aug-2009 18:23:52] <mrayzenoss> hmmm... chet verified it, but build 309 may have been 2.4
[06-Aug-2009 18:23:59] <daMaestro> mrayzenoss, i'll include an updated patch soon, but for example.. lighttpd requires you to send \r\n not just \n
[06-Aug-2009 18:24:18] <daMaestro> k, i'll wait to get up to date on the latest release.. just wanted to ping you 11 months later with a reply
[06-Aug-2009 18:24:19] <daMaestro> lol
[06-Aug-2009 18:24:46] <mrayzenoss> try pulling down a 2.4.2 VM and see if it has it
[06-Aug-2009 18:24:57] <mrayzenoss> that way you don't have to install anything
[06-Aug-2009 18:25:06] <daMaestro> i'm literally doing a backup of the 2.3.3 and will be going to 2.4 shortly
[06-Aug-2009 18:25:15] <mrayzenoss> cool
[06-Aug-2009 18:25:29] <daMaestro> finding and merging in all my local changes has been.... fun.
[06-Aug-2009 18:26:09] <mrayzenoss> yeah, whenever professional services makes a change for a customer, it gets pushed into the next maintenance release to avoid that problem
[06-Aug-2009 18:36:16] <sacpinball> mrayzenoss i just updated zenoss in my development system to 2.4.2
[06-Aug-2009 18:36:30] <sacpinball> and i am seeing a weird problem you might know the answer to
[06-Aug-2009 18:36:47] <sacpinball> a solaris 8 server being monitored via netsnmp
[06-Aug-2009 18:36:58] <sacpinball> i add to monitor the sshd daemon
[06-Aug-2009 18:37:02] <sacpinball> in the os process field
[06-Aug-2009 18:37:08] <mrayzenoss> sacpinball: I can't say I've heard good things about SNMP on Solaris
[06-Aug-2009 18:37:30] <sacpinball> a few hours later, zenoss removes that from being monitored
[06-Aug-2009 18:37:32] <sacpinball> any ideas
[06-Aug-2009 18:37:52] <sacpinball> removing object sshd from rel processes on device os
[06-Aug-2009 18:37:54] <mrayzenoss> take a look at Jane Curry's process monitoring paper: http://www.skills-1st.co.uk/papers/jcurry.html It's great
[06-Aug-2009 18:38:06] <mrayzenoss> I was just headed home
[06-Aug-2009 18:38:19] <sacpinball> no problem
[06-Aug-2009 18:38:20] <sacpinball> thanks
[06-Aug-2009 18:38:25] <sacpinball> i will look at that paper
[06-Aug-2009 18:38:28] <mrayzenoss> later
[06-Aug-2009 18:38:28] <locohost> doesnt solaris have its own non-net-snmp snmp daemon
[06-Aug-2009 18:38:40] <sacpinball> it sure does, but its horrible
[06-Aug-2009 18:39:08] <sacpinball> i looked around and found that a lot of people are using netsnmp to monitor their solaris hosts
[06-Aug-2009 18:43:27] <locohost> yeah, i guess thats pre-solaris 9
[06-Aug-2009 18:43:47] <sacpinball> yeah; i only have 6 of these solaris 8 dogs left
[06-Aug-2009 18:43:56] <sacpinball> so i might just go with ssh monitoring for them
[06-Aug-2009 18:44:13] <sacpinball> seems like net-snmp works, but i have seen two wierd things already
[06-Aug-2009 18:46:05] <locohost> sounds exciting
[06-Aug-2009 18:46:16] <sacpinball> lol
[06-Aug-2009 18:46:16] <sacpinball> yep
[06-Aug-2009 18:46:49] <locohost> you think that is bad, once i made a cisco catalyst 6509E reboot with an snmpwalk, cisco said it was an unpublished bug and suggested a firmware upgrade
[06-Aug-2009 18:47:03] <locohost> i mean, ios upgrade
[06-Aug-2009 18:47:04] <sacpinball> yikes
[06-Aug-2009 18:49:11] <daMaestro> @mray yeah... you guys implemented it not how it should be implemented
[06-Aug-2009 18:49:21] <daMaestro> @mray we need more then just \n
[06-Aug-2009 18:57:27] <moqq> hey, how can i adjust the parameters for alerts? e.g. i have a linux system that's getting memory swap alerts, but for that specific system i don't really care. how can i adjust the threshold or severity of that event?
[06-Aug-2009 19:48:47] bzed_ is now known as bzed
[07-Aug-2009 00:00:46] [disconnected at Fri Aug 7 00:00:46 2009]
[07-Aug-2009 00:00:46] [connected at Fri Aug 7 00:00:46 2009]
[07-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[07-Aug-2009 01:12:32] <KeithBilly> Hey, i'm trying to install 2.4.2 over 2.3 and it keeps flaking on the rrddeps build of glib, looking for iconv on freebsd 6.3
[07-Aug-2009 01:13:15] <KeithBilly> i'm following the posts by http://www.caseyhillman.com/?p=88 to deal with this but can't seem to get glib to see iconv or get iconv to build before glib
[07-Aug-2009 01:13:23] <KeithBilly> can anyone help me fix this problem
[07-Aug-2009 01:16:26] edwardam is now known as edwardamZzZz
[07-Aug-2009 02:45:39] <Dieterbe> is it just me or can't you do 'limit <int>' in WQL ?
[07-Aug-2009 08:52:59] <klinstifen> morning all
[07-Aug-2009 08:53:11] <klinstifen> so my zenoss install has grown to big for me to do a backup to my local disk
[07-Aug-2009 08:53:46] <klinstifen> i want to use an nfs share as the temp-dir for zenbackup
[07-Aug-2009 08:53:54] <klinstifen> has anyone done this successfully?
[07-Aug-2009 08:54:05] <klinstifen> i am getting permission errors
[07-Aug-2009 08:54:19] <klinstifen> no matter how i mount the share (mount, smbmount, sudo mount, etc
[07-Aug-2009 08:56:39] <npmccallum> klinstifen: I think this is the wrong channel to ask your question (since it has nothing to do with Zenoss)
[07-Aug-2009 08:59:54] <LW> Question, I am trying to add an ip service to be monitored, which happens to be in the 18000 range. It was never autodetected when I modeled, so I manually added it manually, yet it still does not find it. I noticed that it was bound via a ipv6 port, and most other apps were binding with ipv4 that were showing up, so I added the ipv6 snmp map, and it found ssh on the box, but not the other service, the other thing I noticed that it's b
[07-Aug-2009 10:02:19] <djjay> hi. I've seen some references to zenoss honoring ./.ssh/config but so far it doesn't seem to be true. anyone know?
[07-Aug-2009 10:37:44] edwardam1 is now known as edwardam
[07-Aug-2009 10:38:05] edwardam is now known as edwardamExer
[07-Aug-2009 10:44:57] <pat2man> Quick question, I want to run the nagios check_ping command against a specific IP address for one of our routers. Can I apply an interface template that will run check_ping? The interface template section of the documentation is rather sparse.
[07-Aug-2009 10:48:08] <mrayzenoss> pat2man: if you create a template that utilizes the nagios command, you can bind it to whichever device classes you need
[07-Aug-2009 10:48:19] <mrayzenoss> I've got it attached on /Devices
[07-Aug-2009 10:48:34] <pat2man> right i can bind it to devices but I want to bind it to a specific interface
[07-Aug-2009 10:48:46] <mrayzenoss> hmmm...
[07-Aug-2009 10:49:00] <pat2man> or can I specify a specific interface in the command section of the template?
[07-Aug-2009 10:49:16] <mrayzenoss> I see what you're asking, I'm looking
[07-Aug-2009 10:53:22] <mrayzenoss> pat2man: the interface that will be tested is the one from the zenoss server to that machine
[07-Aug-2009 10:53:42] <mrayzenoss> is the machine reachable from multiple interfaces from the zenoss server?
[07-Aug-2009 10:53:53] <pat2man> yes
[07-Aug-2009 10:53:56] <mrayzenoss> ugh
[07-Aug-2009 10:55:10] <pat2man> it seems like I should be able to apply a template to the specific interface that I want to monitor
[07-Aug-2009 10:55:26] <pat2man> and just have performance metrics for that interface instead of for the host
[07-Aug-2009 10:56:28] <mrayzenoss> yeah, there's nothing in the man page for check_ping that makes me think it would take any sort of guidance
[07-Aug-2009 10:56:48] <pat2man> well check_ping just runs against an IP
[07-Aug-2009 10:56:49] <mrayzenoss> I mean really, we just call the shell and execute the script, passing the IP address
[07-Aug-2009 10:57:00] <pat2man> in my case I run this: check_ping -H ${dev/id} -w100,80% -c500,100% -p1 | sed 's/, RTA = / RTA=/;s/ loss = / |loss=/'
[07-Aug-2009 10:57:08] <mrayzenoss> ahh, sorry, brainfart
[07-Aug-2009 10:57:20] <pat2man> but ${dev/id} is always the primary IP of the device
[07-Aug-2009 10:57:24] <mrayzenoss> change -H ${dev/id} to -H ${here/manageIp}
[07-Aug-2009 10:58:02] <pat2man> would that be for a device template or an interface template?
[07-Aug-2009 10:58:05] <mrayzenoss> hmm... I wonder if there's a ${here/interfaces[0] or something
[07-Aug-2009 10:58:11] <mrayzenoss> device template
[07-Aug-2009 10:58:34] <pat2man> well I mean I could always create a local copy of my template for each device and manually put in the IP too
[07-Aug-2009 10:58:46] <pat2man> but I would rather have an interface template that I can just bind to any interfaces I need it on
[07-Aug-2009 11:00:38] <pat2man> if its not obvious I can just submit a new feature request
[07-Aug-2009 11:01:49] <mrayzenoss> I'm just not sure where it would get bound
[07-Aug-2009 11:03:12] <venturaville> jb: I still cannot find performance data for the SVCs in CIM.....
[07-Aug-2009 11:03:16] <pat2man> well the admin guide has this to say:
[07-Aug-2009 11:03:17] <pat2man> 2.1.3. Interface Templates
[07-Aug-2009 11:03:17] <pat2man> Interface templates are applied to network interfaces by using a special type of binding. Instead of using the name of the underlying class, Zenoss looks for a template with the same name as the interface type. You can find this type in the details information for any network interface (from the OS tab of its containing device).
[07-Aug-2009 11:03:17] <pat2man> If Zenoss cannot locate a template that matches the interface type, then it uses the ethernetCsmacd template.
[07-Aug-2009 11:05:32] <mrayzenoss> ahh, so if you were to name the check_ping template "eth1" and bind it to the interfaces, it would work on just that interface
[07-Aug-2009 11:05:48] <mrayzenoss> at least, that's what it sounds like to me, I haven't messed with that
[07-Aug-2009 11:07:12] edwardamExer is now known as edwardam
[07-Aug-2009 11:08:18] <pat2man> mrayzenoss: well I think they are talking about the interface type not the name
[07-Aug-2009 11:08:22] <pat2man> type is ethernetCsmacd_64
[07-Aug-2009 11:08:29] <pat2man> name is ethernet0/0
[07-Aug-2009 11:09:21] <mrayzenoss> ahh, correct
[07-Aug-2009 11:09:38] <pat2man> well I will just make a local copy for now and manually enter the IP
[07-Aug-2009 11:09:44] <pat2man> and put in a feature request
[07-Aug-2009 11:10:20] <mrayzenoss> yeah, that would probably be useful
[07-Aug-2009 11:37:08] <jb> venturaville:
[07-Aug-2009 11:54:42] <daMaestro> anyone know what needs to be done to have zencommand timeouts send clears if on the next poll the command correctly executes?
[07-Aug-2009 11:54:45] <daMaestro> it used to do that with 2.2.x
[07-Aug-2009 11:56:23] <daMaestro> basically, we have having zencommand events being raised, but they never clear without manual intervention
[07-Aug-2009 11:56:34] <daMaestro> (now running 2.4.2)
[07-Aug-2009 12:19:19] <mrayzenoss> daMaestro: http://dev.zenoss.org/trac/ticket/5359
[07-Aug-2009 12:19:41] <mrayzenoss> awaiting verification for the 2.4.3 release
[07-Aug-2009 12:19:46] <mrayzenoss> but there's a patch
[07-Aug-2009 12:20:09] <mrayzenoss> hmm... this one too: http://dev.zenoss.org/trac/ticket/5100
[07-Aug-2009 12:21:41] <mrayzenoss> here's the tickets currently slated for 2.4.3: http://dev.zenoss.org/trac/report/6
[07-Aug-2009 12:26:59] <daMaestro> mrayzenoss, cool; i updated my bug with a new patch for 2.4.2
[07-Aug-2009 12:28:17] <daMaestro> http://dev.zenoss.com/trac/ticket/3740
[07-Aug-2009 12:29:01] <daMaestro> mrayzenoss, that patch makes zenoss act like most of the hardware based load balancers/healthchecks i've worked with
[07-Aug-2009 12:29:43] <daMaestro> i think most people would expect that behavior, but would recommend a release note point if it's implemented that if you need to send literal \n, etc it would need to be escaped
[07-Aug-2009 12:32:29] <daMaestro> actually ... on a brief ipython session it doesn't look like a simple \\n prints \n so yeah.
[07-Aug-2009 12:32:45] <daMaestro> i'd still recommend it, we use it and it allows you to do advanced testing of tcp sessions ;-)
[07-Aug-2009 12:33:02] <daMaestro> mrayzenoss, thanks for the link to the patches, applying now
[07-Aug-2009 12:33:31] <mrayzenoss> yeah, 2.4.3 is a few weeks away
[07-Aug-2009 12:33:48] <mrayzenoss> should be last 2.4.x release before King Crab
[07-Aug-2009 12:35:48] <daMaestro> odd... that changeset doesn't apply against 2.4.2 cleanly
[07-Aug-2009 12:36:14] <mrayzenoss> perhaps other patches are interfering?
[07-Aug-2009 12:36:18] <mrayzenoss> or missing
[07-Aug-2009 12:36:23] <daMaestro> http://dev.zenoss.org/trac/changeset/14424#file0
[07-Aug-2009 12:36:36] <daMaestro> that doesn't apply to 2.4.2 cleanly
[07-Aug-2009 12:37:29] <mrayzenoss> there are quite a few patches to zencommand.py
[07-Aug-2009 12:46:47] <daMaestro> http://fpaste.org/ZtKX/ is what i'll give a shot
[07-Aug-2009 13:02:47] <folke> I'm trying to get an snmp perf template working.. But fails miserably on the first counter...
[07-Aug-2009 13:03:17] <folke> Everything looks good in the datasource. I'm getting the right values.
[07-Aug-2009 13:03:39] <folke> And a good ol snmpget gets the right stuff SNMPv2-SMI::enterprises.3711.7.1.1.1.1.2.0 = INTEGER: 358
[07-Aug-2009 13:03:52] <folke> But nuthin shows up in the graphs..
[07-Aug-2009 13:04:43] <daMaestro> you have to add a data point
[07-Aug-2009 13:05:09] <daMaestro> so for custom OIDs, you have to setup a datasource for each OID and then assign a data point to each OID
[07-Aug-2009 13:05:21] <daMaestro> at least that is how it was in 2.2, not sure about 2.4... have not looked yet
[07-Aug-2009 13:06:10] <folke> daMaestro, I am so lost. I can't find that anywhare in 2.4.2
[07-Aug-2009 13:06:13] <folke>
[07-Aug-2009 13:07:15] <folke> In the datasource I can only add an datapoint alias..
[07-Aug-2009 13:08:40] <folke> I had it working in our 2.1.x enviroment.. But no cigar now...
[07-Aug-2009 13:11:07] <folke> I think that they are called Graph Points nowdays.. But still nuthin.
[07-Aug-2009 13:13:49] <folke> when I do a zenperfsnmp run -v 10 -d jadda.jadda.net there do not look like it asks my perftemplate..
[07-Aug-2009 13:13:54] <daMaestro> ahhh
[07-Aug-2009 13:14:05] <daMaestro> no, it's still data source
[07-Aug-2009 13:14:13] <daMaestro> but you specify the name and the oid in the same object now
[07-Aug-2009 13:14:18] <daMaestro> rather then multiple representations
[07-Aug-2009 13:15:11] <folke> daMaestro, Humm, I'll take a look at it..
[07-Aug-2009 13:16:59] <folke> I have the name and the oid, there.. Enabled and everything. And when I test it against a device it returns the right value..
[07-Aug-2009 13:18:32] <daMaestro> k, so add those data points to a graph
[07-Aug-2009 13:18:38] <daMaestro> and then bind the template to a device
[07-Aug-2009 13:19:43] <folke> Aye, done. THe graph shows up with the right legend. But it gets no data.
[07-Aug-2009 13:20:27] <folke> Looks like it auint grabbing it when I run zenperfsmp.
[07-Aug-2009 13:20:49] <mrayzenoss> folke: that'll take a few minutes to get some data going
[07-Aug-2009 13:21:29] <folke> I started it a few hours ago at work.. And now when I got home I checked it out..
[07-Aug-2009 13:21:55] <folke> I have just missed something trivial.. Just need to find it
[07-Aug-2009 13:22:24] <folke> It's my first shot at 2.4.x
[07-Aug-2009 13:26:47] <folke> here is the output from a zenperfsnmp run: http://pastebin.com/d228fe20d
[07-Aug-2009 13:27:20] <kisielk> I'm really starting to hate zenprocess
[07-Aug-2009 13:28:27] <folke> I think that INFO:zen.zenperfsnmp:Sent 0 OID requests
[07-Aug-2009 13:28:27] <folke> INFO:zen.zenperfsnmp:Queried 0 devices is not optimal if you want to hace the graph filled
[07-Aug-2009 13:45:15] <folke> Humm, starting to think that I have broken sumfink in my lab vmware guest enviroment.. The snmp is really working flaky.. Evern with the ordinary linux snmp counters...
[07-Aug-2009 13:45:45] <mrayzenoss> folke: I'm testing adding my own snmp datasource to see if I can figure what's blocking you
[07-Aug-2009 13:47:50] <folke> mrayzenoss, Perhaps it's only pebkac this time I will take the vmware guest back a snap and see if I can get stability in the linux prefs..
[07-Aug-2009 13:51:02] <mrayzenoss> well, I'm trying to add a simple dhcp client count and I'm getting a blank graph. I figure once I get that working I'll throw up a quick howto
[07-Aug-2009 13:57:19] <folke> That sounds great. I will continue over here in sweden for a few hours more
[07-Aug-2009 14:40:12] <mrayzenoss> took awhile, but my perf graphs just started showing up
[07-Aug-2009 14:40:50] <mrayzenoss> I had to add a .0 to the end of the OID I was collecting (as noted in the Admin Guide)
[07-Aug-2009 14:47:59] <folke> mrayzenoss, when I add .0 the test in the Data Source tab fails. But I try to run it like that a few minutes.
[07-Aug-2009 14:48:18] <mrayzenoss> yeah, I'm testing different variations of the template now
[07-Aug-2009 14:48:22] <mrayzenoss> to see what works
[07-Aug-2009 14:50:30] <folke> Just to test I made a local copy of Device snmp template and inserted my oids in it.. To see what happens with it.
[07-Aug-2009 14:52:07] <mrayzenoss> it does appear that once you add a custom template to a device, it seems to not bind any other changes made to the device class templates
[07-Aug-2009 14:52:43] <mrayzenoss> so far, only the templates with a .0 added seem to be working
[07-Aug-2009 14:56:40] <folke> The local device modded template is starting to view graphs now.
[07-Aug-2009 14:57:34] <folke> But my "home made" one with the added zero is empty..
[07-Aug-2009 14:58:25] <folke> I think that I already have the end value in thati oid, and get the right intger from it.
[07-Aug-2009 15:00:49] <folke> It just seems that it wont "parse" the template even if looks like it is binded in the template tab..
[07-Aug-2009 15:26:48] edwardam is now known as edwardamAFK
[07-Aug-2009 15:58:45] edwardamAFK is now known as edwardam
[07-Aug-2009 15:59:31] <mrayzenoss> I'll be out the next week on vacation, if anyone's looking for me
[07-Aug-2009 16:12:39] <folke> Now I'll getting somewhere.. Both graphs getting data now.
[07-Aug-2009 16:17:28] <folke> After I deleted the rrd file and restarted zenoss things starting to light up..
[07-Aug-2009 17:37:36] <mrayzenoss> New WBEM-based ZenPacks
[07-Aug-2009 17:38:04] <mrayzenoss> WBEM DataSource http://www.zenoss.com/community/projects/zenpacks/wbemdatasource
[07-Aug-2009 17:38:17] <mrayzenoss> WBEM LinuxMonitor http://www.zenoss.com/community/projects/zenpacks/wbem_linux
[07-Aug-2009 17:43:29] <mrayzenoss> Last Week for the Summer of ZenPacks Contest: http://tinyurl.com/optb2d
[07-Aug-2009 17:46:11] <mrayzenoss> later all!
[07-Aug-2009 18:17:10] daMaestro|isBack is now known as daMaestro
[07-Aug-2009 19:44:17] <malibu> Hi there, I just installed zenoss for the first time from the CentOS 5 stack installer.. But the admin:zenoss ID isn't working
[07-Aug-2009 19:44:29] <malibu> Anyone know what the default ID/pass might be?
[07-Aug-2009 23:34:44] [disconnected at Fri Aug 7 23:34:44 2009]
[08-Aug-2009 00:00:46] [connected at Sat Aug 8 00:00:46 2009]
[08-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[08-Aug-2009 06:24:05] <raddy> Hello Everybody
[08-Aug-2009 06:24:12] <raddy> Anybod out there????
[08-Aug-2009 06:28:20] <raddy> Anybody there???
[08-Aug-2009 08:59:14] <straterra> Indeed
[08-Aug-2009 08:59:22] <straterra> Patience..
[08-Aug-2009 09:37:24] <brice> Did zenoss write their own WMI module or is there one available out there?
[08-Aug-2009 09:38:51] <brice> only WMI module i've found for python is one that needs to be ran on a windows box
[08-Aug-2009 11:20:58] <brice> or does it use wbemcli
[09-Aug-2009 00:00:46] [disconnected at Sun Aug 9 00:00:46 2009]
[09-Aug-2009 00:00:46] [connected at Sun Aug 9 00:00:46 2009]
[09-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[09-Aug-2009 20:13:57] <Ford> greetings people
[09-Aug-2009 20:17:23] <Ford> what's happening ?
[10-Aug-2009 00:00:46] [disconnected at Mon Aug 10 00:00:46 2009]
[10-Aug-2009 00:00:46] [connected at Mon Aug 10 00:00:46 2009]
[10-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[10-Aug-2009 10:05:12] <dollarbang> question for the zenoss community, we are using ldap security for signon verification. I have added someone, but for some reason, the person is unable to logon. What log file on the server can I read to find the error.
[10-Aug-2009 10:06:55] <venturaville1> that WBEM module is nice, but it is probably something that needs to make it into base zenoss at some point
[10-Aug-2009 10:09:44] <jb> yeah, i havent had a chance to try it yet
[10-Aug-2009 10:16:18] <venturaville1> jb: once the dev server is up I will see if it can be used against our arrays, though it obviously won't be able to get the structure from them the way it is built
[10-Aug-2009 10:16:51] <venturaville1> i.e. the volumes on it
[10-Aug-2009 10:17:42] <venturaville1> looks like he is just using pywbem so it should be possible to replicate the same work
[10-Aug-2009 13:57:59] willwh1 is now known as willwh
[10-Aug-2009 15:38:28] <twm1010> bigegor: I'm having trouble getting any WMI queries to work with the datasource ZenPack
[11-Aug-2009 00:00:46] [disconnected at Tue Aug 11 00:00:46 2009]
[11-Aug-2009 00:00:46] [connected at Tue Aug 11 00:00:46 2009]
[11-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[11-Aug-2009 05:14:26] <Dieterbe> If you make a threshold for a performance template and you select multiple datapoints. will it trigger if any of them hits the specified value, or when the sum of them hits the value?
[11-Aug-2009 07:05:50] <Dieterbe> twm1010: what's your problem? i use it and it works for me
[11-Aug-2009 09:49:15] <raddy> Is it possible to monitor Tomcat via ZenOSS without installing the enterprice edition?
[11-Aug-2009 09:58:51] <twm1010> hrmm.. probably
[11-Aug-2009 09:59:17] <raddy> twm1010 : How?
[11-Aug-2009 09:59:34] zenNoob is now known as chickenandbeans
[11-Aug-2009 10:05:06] <chickenandbeans> hi, need help with a zenoss problem, anyone got a minute or two to spare?
[11-Aug-2009 10:06:26] <jb> anybody here transform savy?
[11-Aug-2009 10:07:04] <chickenandbeans> Just want to know if this has been solved: http://forums.zenoss.com/viewtopic.php?t=6706&highlight=
[11-Aug-2009 10:07:04] <jb> http://pastebin.ca/1525153
[11-Aug-2009 10:07:10] <jb> trying to figure out how I can fix this..
[11-Aug-2009 10:14:19] <chickenandbeans> Any takers on my problem?
[11-Aug-2009 10:17:45] <chickenandbeans> It's the sudden lack of graphing issue: http://forums.zenoss.com/viewtopic.php?t=6706&highlight=
[11-Aug-2009 10:51:57] <TreeStump> hey all
[11-Aug-2009 10:53:23] <TreeStump> im just wondering if someone can provide some assitance
[11-Aug-2009 10:53:46] <TreeStump> i have a zenoss install
[11-Aug-2009 10:53:50] <TreeStump> and for some reason
[11-Aug-2009 10:55:02] <TreeStump> my graphs for bandwidth and performance are wierd
[11-Aug-2009 10:55:24] <TreeStump> isee they drop out for hours at a time, and show absolutely no data
[11-Aug-2009 10:56:36] <TreeStump> when an mrtg graph of the same connections for example are viewed
[11-Aug-2009 10:56:46] <TreeStump> it shows data throughout the day
[11-Aug-2009 10:57:36] <TreeStump> any ideas?
[11-Aug-2009 11:18:42] <twm1010> TreeStump: I would say that if the bandwidth monitoring of interfaces AND the performance graphs for CPU/memory are showing gaps in the same spots
[11-Aug-2009 11:18:53] <twm1010> that your zenoss box is having difficulty with something regarding polling
[11-Aug-2009 11:19:43] <twm1010> or i should say that something is affecting its ability to poll
[11-Aug-2009 11:21:50] <TreeStump> maybe because its behind NAT?
[11-Aug-2009 11:23:00] <twm1010> mmm... perhaps, but is the MRTG box also behind NAT?
[11-Aug-2009 11:23:53] <twm1010> if there are gaps in the data, it means A) the collect did not get to the SNMP agents of the device to retrieve it
[11-Aug-2009 11:24:16] <twm1010> or B) the collector did not attempt to
[11-Aug-2009 11:25:01] <twm1010> how long are the gaps?
[11-Aug-2009 11:39:08] <TreeStump> maybe 6-8 hours
[11-Aug-2009 11:44:01] <twm1010> thats a long time
[11-Aug-2009 11:44:39] <twm1010> is it simultaneous across all devices, just those outside the NAT?
[11-Aug-2009 11:45:37] <TreeStump> well i only have one device inside the NAT at the moment being monitored - i just added another but it wont have data yet
[11-Aug-2009 11:45:47] <TreeStump> that one devise is a cisco switch
[11-Aug-2009 11:46:21] <TreeStump> I havent received data on it since 12:00AM yesterday morning
[11-Aug-2009 11:46:32] <TreeStump> just over 24 hours ago
[11-Aug-2009 11:46:37] <TreeStump> hang on
[11-Aug-2009 11:46:52] <TreeStump> the clock on the machine keeps resetting
[11-Aug-2009 11:47:01] <TreeStump> could that have anything to do with it haha
[11-Aug-2009 12:04:21] richard is now known as Guest39185
[11-Aug-2009 12:04:28] <Guest39185> hey all, started using zenoss today and im wondering how i get it to display a list of software which is installed on a linux box?
[11-Aug-2009 12:24:02] <TreeStump> too far outta my depth mate
[11-Aug-2009 14:09:37] <TreeStump> so after some investigating
[11-Aug-2009 14:10:22] <TreeStump> i noticed, that on each device it reports data as being collected last houurs ago
[11-Aug-2009 14:10:36] <TreeStump> under management i did edit collectors
[11-Aug-2009 14:10:41] <TreeStump> changed some settings
[11-Aug-2009 14:10:47] <TreeStump> was hoping it would check more often
[11-Aug-2009 14:11:10] <TreeStump> i noticed even graphs are missing on the localhost collector
[11-Aug-2009 14:11:25] <TreeStump> is there a way i can reset the settings of the collector?
[11-Aug-2009 14:20:20] <twm1010> TreeStump: the collector picks up changes automatically after some time, but it seems like you're having more "general" problems then settings issues
[11-Aug-2009 14:23:20] <twm1010> Is zenperfsnmp running?
[11-Aug-2009 15:38:30] <TreeStump> twm1010: how do I know if its running?
[11-Aug-2009 15:42:51] <twm1010> go settings look at daemons
[11-Aug-2009 16:00:32] <TreeStump> yeah it says its running
[11-Aug-2009 16:00:48] <TreeStump> zentrap and zenjmx are not though
[11-Aug-2009 18:04:35] <jsm> what format is this supposed to produce? http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch08s05.html? it's seems to be a string of a python list... which i'm not quite sure how to treat as a real list.. I would guess it should be JSON format, but i can't seem to use the JSON module inside of that function.
[12-Aug-2009 00:00:46] [disconnected at Wed Aug 12 00:00:46 2009]
[12-Aug-2009 00:00:46] [connected at Wed Aug 12 00:00:46 2009]
[12-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[12-Aug-2009 08:06:33] <jb> anybody have a reccomendation for a python book?
[12-Aug-2009 08:31:21] <brice> http://diveintopython.org/
[12-Aug-2009 08:57:22] <jb> anybody here have any modeler experience?
[12-Aug-2009 09:22:55] <jb> hrm, can't figure out why my modelers are not gettign compiled.
[12-Aug-2009 09:24:57] <jb> relm -l
[12-Aug-2009 09:38:44] <jb> ah.. no __init__.py's
[12-Aug-2009 11:22:29] <x-ip> hi all
[12-Aug-2009 11:22:52] <x-ip> i'm trying and learning zenoss right now .... its ... awsome :,)
[12-Aug-2009 11:39:51] <jsm> what wire format does zenoss use when it returns data as described here? http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch08s05.html
[12-Aug-2009 11:39:55] <jsm> it is not JSON format.
[12-Aug-2009 12:43:08] burlyscudd__ is now known as burlyscudd
[12-Aug-2009 13:04:33] burlyscudd_ is now known as burlyscudd
[12-Aug-2009 19:41:58] <tpdd> Sorry for the newbie question, but can someone point me to a document that can explain how to use zenoss to monitor a linux dhcp server?
[13-Aug-2009 00:00:46] [disconnected at Thu Aug 13 00:00:46 2009]
[13-Aug-2009 00:00:46] [connected at Thu Aug 13 00:00:46 2009]
[13-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[13-Aug-2009 06:42:11] <etank> anyone had expierence with getting zenoss to monitor a windows cluster?
[13-Aug-2009 06:42:24] <etank> I have 4 that I need to monitor
[13-Aug-2009 06:43:32] <etank> the problem is that i can add each node but when i add the cluster node it says that the IP is already assigned to another node
[13-Aug-2009 07:09:00] <etank> setting zInterfaceMapIgnoreTypes to ethernetCsmacd removes all nics but the loopback :/
[13-Aug-2009 07:15:08] <etank> hmm .. i can still ping it though. just not get stats on the nic
[13-Aug-2009 08:02:28] <jb> morning..
[13-Aug-2009 08:23:45] sakura_ is now known as x-ip
[13-Aug-2009 08:33:24] <x-ip> morning all =)
[13-Aug-2009 09:31:54] <jb> yawn, I wish I could figure out why these .py's aren't compiling.
[13-Aug-2009 09:53:46] <x-ip> anybody knows where is zTransportPreference ?
[13-Aug-2009 09:54:03] <x-ip> in vs 2.4.2 seems isnt at zPreferences
[13-Aug-2009 09:54:04] <x-ip> =)
[13-Aug-2009 09:54:24] <x-ip> also, i want to monitor a remote machine that runs ssh in a different port
[13-Aug-2009 09:55:02] <x-ip> than default 22, i set zCommandPort to the correct port number and seems that zenoss isnt using that value
[13-Aug-2009 10:06:16] <x-ip> hmmm surely i'm wrong ... i'm looking at zenoss/Products/DataCollector/zenmodeler.py, line 268, method collectDevice and the self.cmdCollect taks three args , device, ip and timeout but seems to not take a port number
[13-Aug-2009 10:06:57] <chudler> it probably inherits it from some its parent class, or even higher up the chain
[13-Aug-2009 10:07:12] <x-ip> maybe the use from another port different from 22 (ssh's default) isnt still coded ?
[13-Aug-2009 10:07:59] <x-ip> surely charlieS ... only trying to know why zCommandPort isnt working for me
[13-Aug-2009 10:14:57] <Jane_Curry> <jb> - not sure what .py files you are struggling with but I have just put a paper on "creating Zenoss ZenPacks" on the forum and our website
[13-Aug-2009 10:15:17] <Jane_Curry> It includes some debugging hints and tips
[13-Aug-2009 10:15:34] <Jane_Curry> Get the paper either from http://www.zenoss.com/Members/jcurry/creating-zenoss-zenpacks/
[13-Aug-2009 10:15:54] <Jane_Curry> or from http://www.skills-1st.co.uk/papers/jane/zenpacks/
[13-Aug-2009 10:22:47] <jb> ah one sec
[13-Aug-2009 10:23:01] <jb> i didn't know you had that paper..
[13-Aug-2009 10:23:08] <jb> basically, I have made my new modelers
[13-Aug-2009 10:23:28] <jb> but I'm trying to now define a new class (I guess)
[13-Aug-2009 10:24:22] <jb> the modelers are compiling fine
[13-Aug-2009 10:24:51] <jb> but my classes arent.. and im not sure im doing it correctly.. hopefully this paper will enlighten me)
[13-Aug-2009 10:26:29] <Jane_Curry> <jb> I've only finished the draft of the paper today so it only went public 3 hours back - I'm looking for reviews and comments
[13-Aug-2009 10:26:42] <jb> awesome.. I'm reading it now!
[13-Aug-2009 10:27:06] * x-ip reading it too
[13-Aug-2009 10:43:05] <Jane_Curry> Anyone know of a Zenoss feature or ZenPack that collec6ts "Serial Number" from a Windows box?
[13-Aug-2009 10:43:41] <Jane_Curry> (Don't speak "Doze" very well so I'm not sure whether this is information that is available)
[13-Aug-2009 10:46:30] <jb> certain systems management agents (Dell OpenManage, HP Insight, IBM Director, etc) offer this in their MIB's
[13-Aug-2009 11:06:17] <Jane_Curry> Anyone got an easy method for creating pdf copies of reports?
[13-Aug-2009 11:07:20] <Jane_Curry> There seemed to be a forum thread doing serious discussion of possibilities last October but I don't see any obvious ZenPacks
[13-Aug-2009 11:07:53] <Jane_Curry> I also see a note from Matt Ray saying pdf reports is NOT currently in plan for King Crab - unless someone knows better?
[13-Aug-2009 11:15:19] <ali3> hallow
[13-Aug-2009 11:17:36] <ali3> please help
[13-Aug-2009 11:42:42] <jb> jane?
[13-Aug-2009 11:52:07] <tpdd> can anyone explain how to monitor a dhcpd service (running on a linux box with the command tools). When I model the server it shows up but it won't let me turn on monitoring
[13-Aug-2009 12:34:54] edwardam1 is now known as edwardam
[13-Aug-2009 12:44:00] <twm1010> tpdd: it shows up on the OS tab under services?
[13-Aug-2009 12:46:43] <korozion> Hi there, I have many towers, each with some access points on them. Each AP has many customers. Each tower also has a router and/or a switch and some other snmp equipment. Currently I have an old version of What's Up Gold as seen here ... http://www.linuxgeneration.org/pics/WUG1.jpg. I require the map and the ability to browse to the web page of my APs and customer devices from it. Is this something ZenOSS can help me with?
[13-Aug-2009 12:48:11] <rhett> yes
[13-Aug-2009 12:49:03] <rhett> i wish i would have used zenoss when i was in the wisp/wireless world
[13-Aug-2009 12:49:13] <rhett> much easier to manage by geographical location
[13-Aug-2009 12:52:10] <korozion> hmm
[13-Aug-2009 12:52:16] <korozion> nice. Are the maps easy to edit?
[13-Aug-2009 12:53:30] <rhett> you add locations in zenoss then tag devices to that location. the lines between locations are drawn based on the routing tables
[13-Aug-2009 12:53:43] <rhett> and its all quite easy to do
[13-Aug-2009 12:54:33] <rmatte> korozion: when you create a location, you need to click on that location, you'll see a field where you can click an "edit" link to specify the address of the location, that needs to be filled out properly for the location to show up on the map
[13-Aug-2009 12:59:57] <_mingau_> hello people
[13-Aug-2009 13:00:02] <rmatte> hi
[13-Aug-2009 13:00:05] <_mingau_> i need a little help
[13-Aug-2009 13:00:18] <korozion> rmatte: very nice, thanks
[13-Aug-2009 13:00:48] <twm1010> _mingau_: go ahead
[13-Aug-2009 13:00:58] <_mingau_> i am trying to install default mibs...
[13-Aug-2009 13:01:06] <_mingau_> but i got an error
[13-Aug-2009 13:02:09] <rmatte> default Mibs?
[13-Aug-2009 13:02:45] <rmatte> for what type of OS/Device?
[13-Aug-2009 13:02:54] <_mingau_> yes
[13-Aug-2009 13:02:57] <_mingau_> freebsd 7.2
[13-Aug-2009 13:03:26] <_mingau_> its a fresh install of zenoss core...
[13-Aug-2009 13:03:55] <_mingau_> so i was tryinf to load mibs from share/mibs/iana/
[13-Aug-2009 13:04:15] <rmatte> I never had to install Mibs for Linux, I think that Zenoss already comes with all of the Mibs that you would need to translate traps
[13-Aug-2009 13:04:22] <rmatte> I assume you're using net-snmp?
[13-Aug-2009 13:04:56] <_mingau_> yes
[13-Aug-2009 13:05:03] <_mingau_> hey
[13-Aug-2009 13:05:07] <_mingau_> now i got it...
[13-Aug-2009 13:05:12] <_mingau_> hehehhee
[13-Aug-2009 13:05:27] <_mingau_> SMIPATH was wrong
[13-Aug-2009 13:06:13] <rmatte> ummmm, you aren't installing them as the root user are you?
[13-Aug-2009 13:06:23] <_mingau_> nonono
[13-Aug-2009 13:06:32] <_mingau_> i wrote wrong the SMIPATH
[13-Aug-2009 13:06:38] <_mingau_> God
[13-Aug-2009 13:06:42] <twm1010> rmatte: I'm still moving those mibs, a little at a time...
[13-Aug-2009 13:06:42] <twm1010> haha
[13-Aug-2009 13:06:55] <twm1010> a zenDMD line to just move them all and let it crank, would be nice
[13-Aug-2009 13:07:05] <rmatte> twm1010: brutal
[13-Aug-2009 13:07:25] <rmatte> twm1010: I'm sure that'd be possible somehow, I just don't know how unfortunately hehe
[13-Aug-2009 13:07:43] <_mingau_>
[13-Aug-2009 13:07:46] <twm1010> Im thinking the slowness is filesystem performance on my zenoss server
[13-Aug-2009 13:07:52] <twm1010> disk throughput really
[13-Aug-2009 13:08:07] <twm1010> though CPU on that guy is pretty high alot of the time as well
[13-Aug-2009 13:08:12] <rmatte> twm1010: When I was moving them it was on a 15ghz server, and I had all of the processing power to the one box at the time so it went fairly quickly
[13-Aug-2009 13:09:24] <twm1010> well, no big deal really, its a convenience thing, only really a pain when i add new ones and gotta skim for them
[13-Aug-2009 13:12:43] <rmatte> yeh
[13-Aug-2009 13:13:46] <_mingau_> thx people
[13-Aug-2009 13:20:08] <TreeStump> hheya guys
[13-Aug-2009 13:20:26] <TreeStump> i came in here a few days back wondeing if anyone could help
[13-Aug-2009 13:20:47] <TreeStump> i think now, if i can't find a solution i'll just reformat haha so sick of this damn thign
[13-Aug-2009 13:21:30] <TreeStump> i started having issues where snmp was only collecting data for part of the day, there would be massive gaps in my performance/bandwidth graphs
[13-Aug-2009 13:21:49] <TreeStump> now no data is collected at allby the looks of it
[13-Aug-2009 13:22:05] <TreeStump> any ideas anyone what the prob could be?
[13-Aug-2009 13:23:17] <twm1010> It honestly sounds like the SNMP collector (zenperfsnmp) is not running
[13-Aug-2009 13:29:22] <TreeStump> hmm it id
[13-Aug-2009 13:29:24] <TreeStump> is*
[13-Aug-2009 13:29:30] <rmatte> TreeStump: become the zenoss user and issue the command: zenoss status
[13-Aug-2009 13:29:36] <TreeStump> funny thing though when I checked it out today
[13-Aug-2009 13:29:55] <TreeStump> it said last collection was an hour ago(first time it said that in ages)
[13-Aug-2009 13:30:00] <TreeStump> bu still no data
[13-Aug-2009 13:30:32] <TreeStump> rmatte: how do i become the zenoss user? give it a passwd, and login through SSH?
[13-Aug-2009 13:30:51] <rmatte> TreeStump: nah, login, become root, then "su - zenoss"
[13-Aug-2009 13:35:49] <twm1010> either way will do.
[13-Aug-2009 13:36:13] <twm1010> I think you need to do some basic troubleshooting here. 1. can zenoss ping the devices
[13-Aug-2009 13:36:19] <twm1010> 2. can zenoss SNMPwalk the device
[13-Aug-2009 13:36:26] <rmatte> yeh, I just find it best not to allow direct logins to the Zenoss account, one less security issue to worry about
[13-Aug-2009 13:36:37] <twm1010> 3. are there events being logged that might indicate what your problem is
[13-Aug-2009 13:37:28] <rmatte> there's a bunch of things that could be wrong, you might need to remodel devices, reindex the database, make sure that zenperfsnmp isn't erroring out somehow, stuff like that.
[13-Aug-2009 13:37:31] <jb> rmatte: have you by chance developed any new object classes for your zenpacks?
[13-Aug-2009 13:37:50] <TreeStump> hmm as much as I hate to say it, i think i have much larger problems to worry about, now that I look at it
[13-Aug-2009 13:37:53] <rmatte> jb: no, I haven't
[13-Aug-2009 13:38:12] <tpdd> twn1010: it does show up but not as monitored
[13-Aug-2009 13:38:18] <TreeStump> seems like some major major network conjestion
[13-Aug-2009 13:38:21] <rmatte> TreeStump: what are you seeing?
[13-Aug-2009 13:38:29] <rmatte> ah
[13-Aug-2009 13:38:53] <TreeStump> actually no not network conjestion
[13-Aug-2009 13:38:58] <rmatte> that'd do it, I'd recommend upping your snmp timeout value in zproperties, and upping your collector cycle times in properties for the collector
[13-Aug-2009 13:39:19] <TreeStump> jsut some really wierd network issues
[13-Aug-2009 13:39:25] <TreeStump> hmm ok
[13-Aug-2009 13:39:28] <TreeStump> thanks for the tips
[13-Aug-2009 13:39:31] <rmatte> np
[13-Aug-2009 13:39:49] <TreeStump> ill ook into this network issue and then come back to you if i don't get anywhere after that
[13-Aug-2009 13:39:51] <TreeStump> thanks gain
[13-Aug-2009 13:39:54] <TreeStump> again*
[13-Aug-2009 13:44:07] <jb> any devs around?
[13-Aug-2009 13:46:35] <rmatte> Have there been any announcements as to when 2.4.3 is going to be released?
[13-Aug-2009 13:47:38] <twm1010> haven't seen anything
[13-Aug-2009 14:15:59] <TreeStump> ok so network issues were nothing
[13-Aug-2009 14:16:13] <TreeStump> still cant seem to get anything
[13-Aug-2009 14:16:18] <TreeStump> i can model devides
[13-Aug-2009 14:16:21] <TreeStump> devices*
[13-Aug-2009 14:16:34] <TreeStump> and ping them
[13-Aug-2009 14:17:43] <tpdd> so if you have to monitor upd based services is the prefered way to do that by using nagios plugins?
[13-Aug-2009 14:22:14] <TreeStump> zenperfsnmp is running
[13-Aug-2009 14:22:27] <rmatte> check $ZENHOME/logs/zenperfsnmp.log
[13-Aug-2009 14:22:33] <rmatte> see if there are any errors
[13-Aug-2009 14:22:46] <rmatte> you might also want to try doing "zenperfsnmp start -v10"
[13-Aug-2009 14:22:58] <rmatte> and see what it spits out
[13-Aug-2009 14:24:38] <jb> argh.. what daemon actually does something with object classes?
[13-Aug-2009 14:24:45] <jb> i cannot figure out why these won't compile.
[13-Aug-2009 14:24:57] <rmatte> jb: what issue are you experiencing relating to them?
[13-Aug-2009 14:25:06] <TreeStump> zenperfsnmp is running
[13-Aug-2009 14:25:07] <jb> they aren't compiling
[13-Aug-2009 14:25:08] <jb>
[13-Aug-2009 14:25:13] <TreeStump> is what i reports
[13-Aug-2009 14:25:33] <rmatte> jb: when starting Zenoss, or what?
[13-Aug-2009 14:25:35] <jb> yes
[13-Aug-2009 14:25:39] <jb> or zenhub..
[13-Aug-2009 14:25:44] <rmatte> jb: hmmm, what's the exact error?
[13-Aug-2009 14:25:48] <jb> there is none..
[13-Aug-2009 14:25:56] <rmatte> zenhub run -v10
[13-Aug-2009 14:26:00] <jb> yeah, nothing..
[13-Aug-2009 14:26:04] <TreeStump> the log reports that the cycle completes with success... but i see there are rrdtool errors
[13-Aug-2009 14:26:06] <rmatte> (how do you know it's not compiling them?)
[13-Aug-2009 14:26:13] <jb> because its not creating the .pyc's
[13-Aug-2009 14:26:18] <rmatte> ah
[13-Aug-2009 14:26:20] <rmatte> gotcha
[13-Aug-2009 14:26:34] <rmatte> hmmm
[13-Aug-2009 14:26:39] <rmatte> when did this start happening?
[13-Aug-2009 14:26:46] <jb> this is a custom zenpack im working on..
[13-Aug-2009 14:26:54] <rmatte> oh
[13-Aug-2009 14:26:56] <jb> im not entirely sure I have the relationships and classes setup properly
[13-Aug-2009 14:27:00] <jb> which could be part of the problem
[13-Aug-2009 14:27:06] <rmatte> probably is lol
[13-Aug-2009 14:27:28] <jb> this stuff is really confusing
[13-Aug-2009 14:27:31] <rmatte> I don't have any real experience with that, I'm just starting to learn the complexities of ZenPack editing
[13-Aug-2009 14:27:37] <jb> even after reading jane's paper and looking at egor's packs..
[13-Aug-2009 14:27:46] <jb> yeah, me too
[13-Aug-2009 14:27:50] <kgoedtel> does anyone know if there is a way or if someone has attempted to make a threshold/alert based on multiple conditions being true?
[13-Aug-2009 14:27:55] <jb> i have my modelers built!
[13-Aug-2009 14:28:16] <rmatte> kgoedtel: you
[13-Aug-2009 14:28:18] <rmatte> oops
[13-Aug-2009 14:28:34] <rmatte> kgoedtel: you'd have to make the threshold and then apply a transform to it
[13-Aug-2009 14:29:01] <kgoedtel> yeah i was hoping i could somehow do it all within a template
[13-Aug-2009 14:29:33] <rmatte> kgoedtel: nope, Zenoss isn't setup for that at the moment
[13-Aug-2009 14:29:47] <kgoedtel> damn, logical operators for that would be nice :]
[13-Aug-2009 14:29:53] <rmatte> Here's an example of a threshold transform: http://www.zenoss.com/Members/netdata/show-the-usage-in-a-filesytem-threshold/
[13-Aug-2009 14:30:25] <kgoedtel> thanks
[13-Aug-2009 14:30:31] <rmatte> np
[13-Aug-2009 14:31:57] <rmatte> well, I'm on vacation and I've been lounging around all day so far, time to shower & get dressed lol
[13-Aug-2009 14:32:02] <rmatte> afk
[13-Aug-2009 15:15:04] <rmatte> rmatte test
[13-Aug-2009 15:15:08] <rmatte> oops
[13-Aug-2009 16:23:04] <cwj> i have an instance of zenoss-stack core which i have configured for monitoring a bunch of devices via SNMP and Nagios/NRPE checks. the monitoring was working fine for like two weeks, and now i am seeing a ton of failures for both SNMP agents and NRPE checks, even though the same checks succeed every time I run them from the command line on the server where this instance of zenoss is running.
[13-Aug-2009 16:23:18] <cwj> i've been reviewing the logs but still can't make sense of why things are failing
[13-Aug-2009 16:23:36] <cwj> the NRPE checks in zenoss are returning exit code 2
[13-Aug-2009 16:23:47] <cwj> but they return exit code 0 when run from the command line
[13-Aug-2009 16:24:07] <cwj> can anyone suggest something i might have missed?
[13-Aug-2009 21:07:58] <alley> hello
[13-Aug-2009 21:17:47] <alley> ello
[13-Aug-2009 21:17:48] <alley> allow
[13-Aug-2009 22:26:10] <alley> alow
[13-Aug-2009 22:42:50] <alley> halow all
[13-Aug-2009 22:44:07] <korozion> you ok?
[14-Aug-2009 00:00:46] [disconnected at Fri Aug 14 00:00:46 2009]
[14-Aug-2009 00:00:46] [connected at Fri Aug 14 00:00:46 2009]
[14-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[14-Aug-2009 05:18:59] [2]folke is now known as folke
[14-Aug-2009 07:24:05] <baffle> I'm experiencing huge gaps in everyting that is polled over SNMP; What are the common cause for this?
[14-Aug-2009 07:32:37] <folke> baffle, is the server under heavy load?
[14-Aug-2009 07:40:00] <baffle> folke: Not really.
[14-Aug-2009 07:41:19] <baffle> folke: Or, well, it sometimes has some processes that uses a lot of CPU, but it is a quad core 2.33 with 22G ram. 4G is used by ZenOss, rest is fscache.
[14-Aug-2009 07:41:36] <baffle> folke: 5 minute load is 0.18.
[14-Aug-2009 08:12:23] <folke> baffle, then it should not be a problem..
[14-Aug-2009 08:13:37] <baffle> folke: But it is. :-)
[14-Aug-2009 08:14:39] <baffle> folke: Hmm, I have a lot of backtraces in the perfsnmp log, due to some devices reporting the wrong values. I.e. something that usd to be an int is now a string.. Does a crash like that impact all other devices that happend to be polled by the same thread?
[14-Aug-2009 08:38:57] <folke> baffle, Are there some device that is always having trouble..
[14-Aug-2009 08:39:52] <folke> I had one that messed up the zenperfsnmp deamon.. But that was in 2.0.x i think.
[14-Aug-2009 08:44:32] <baffle> folke: Yes, there is a few devices that bakes it barf.
[14-Aug-2009 08:45:51] <folke> if you disable snmp for a moment on those efvil ones, and see if the gaps stops on the working ones?
[14-Aug-2009 16:38:08] <otakup0pe> hmm how do i get zenoss to track network utilization
[14-Aug-2009 16:38:18] <otakup0pe> like total down/upload per interface per "time period"
[14-Aug-2009 16:39:03] <otakup0pe> rather than bw/sec i want to know N GB per day
[14-Aug-2009 17:46:56] <otakup0pe> oh come on there has to be someone in here who is still sober
[14-Aug-2009 17:46:57] <otakup0pe>
[14-Aug-2009 18:03:37] <rhett> haha
[14-Aug-2009 18:04:26] <rhett> i may be able to be of some help
[14-Aug-2009 18:47:19] <otakup0pe> :3
[14-Aug-2009 18:47:27] <otakup0pe> did my question make sense ?
[14-Aug-2009 18:58:35] <rhett> yeah it makes perfect sense. i monitor all that in cacti. just busy and dont know if anyone has already done that or not
[14-Aug-2009 18:58:48] <rhett> for cacti that is
[14-Aug-2009 18:58:59] <rhett> errrr zenoss
[14-Aug-2009 19:15:54] <otakup0pe> what snmp oid's do you use
[14-Aug-2009 19:19:04] <rhett> you use the in/out octets oid's and then have a comment print the ammount of bytes
[14-Aug-2009 19:19:48] <rhett> in cacti for example i use COMMENT:Total Out: |sum:auto:current:2:auto|
[14-Aug-2009 19:19:56] <otakup0pe> rather than do any fancy math on it to get the kbps
[14-Aug-2009 19:19:56] <otakup0pe> ok
[14-Aug-2009 19:19:59] <otakup0pe> i'll poke at this
[15-Aug-2009 00:00:46] [disconnected at Sat Aug 15 00:00:46 2009]
[15-Aug-2009 00:00:46] [connected at Sat Aug 15 00:00:46 2009]
[15-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[15-Aug-2009 16:50:54] <dcoderhed> Howdy
[16-Aug-2009 00:00:46] [disconnected at Sun Aug 16 00:00:46 2009]
[16-Aug-2009 00:00:46] [connected at Sun Aug 16 00:00:46 2009]
[16-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[16-Aug-2009 04:24:38] <philjray> Hi All
[16-Aug-2009 04:25:00] <philjray> I am new to Zenoss and have a question regarding MIBs
[16-Aug-2009 04:25:46] <philjray> Having looked at the documentation and wiki etc. more info is still required - by me anyway!
[16-Aug-2009 04:39:47] <philjray> Hi All
[16-Aug-2009 04:39:54] <philjray> I am new to Zenoss and have a question regarding MIBs
[16-Aug-2009 04:40:02] <philjray> Having looked at the documentation and wiki etc. more info is still required - by me anyway!
[16-Aug-2009 04:41:04] <philjray> I have the 3rd party MIB installed on Zenoss via the web interface
[16-Aug-2009 04:41:31] <philjray> What I need to know is how to "apply" that MIB to the hardware that I want to monitor
[16-Aug-2009 20:08:39] <ironpaw> hi
[16-Aug-2009 20:08:40] <ironpaw> anyone here?
[16-Aug-2009 20:08:50] <ironpaw> why wont my snmpignore=true work?
[16-Aug-2009 20:08:53] <ironpaw> even when pushing changes?
[16-Aug-2009 20:08:59] <ironpaw> as per this thread:
[16-Aug-2009 20:09:01] <ironpaw> http://forums.zenoss.com/viewtopic.php?p=28410
[16-Aug-2009 20:09:05] <ironpaw> this is 2.4.2
[17-Aug-2009 00:00:46] [disconnected at Mon Aug 17 00:00:46 2009]
[17-Aug-2009 00:00:46] [connected at Mon Aug 17 00:00:46 2009]
[17-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[17-Aug-2009 07:32:41] <_mingau_> hi people
[17-Aug-2009 07:35:08] <_mingau_> does anybody has info about wmi monitoring in zenoss?
[17-Aug-2009 07:43:49] <_mingau_> hello?
[17-Aug-2009 09:10:37] * rmatte is back from vacation
[17-Aug-2009 09:11:54] <straterra> how was it?
[17-Aug-2009 09:32:14] <twm1010> rmatte: you around?
[17-Aug-2009 09:34:56] <rmatte> yeh
[17-Aug-2009 09:35:03] <rmatte> fairly busy today, but what's up?
[17-Aug-2009 09:35:16] <rmatte> straterra: quite relaxing
[17-Aug-2009 09:37:16] <twm1010> you done any transformation of poller results?
[17-Aug-2009 09:37:36] <twm1010> take a value from an OID, pump it through a formula to change, like celsius to fahrenheit, etc
[17-Aug-2009 09:47:53] <rmatte> no, I haven't
[17-Aug-2009 09:55:11] <rmatte> The way I'd do it is to write a script to collect the data and convert it, then just use that as a command based datapoint
[17-Aug-2009 09:55:51] <rmatte> You can convert the data in the graph using RPN, but that won't actually change the values stored in the RRD itself
[17-Aug-2009 09:59:09] <twm1010> that's fine, actually
[17-Aug-2009 10:52:08] <fuzz> i recently upgraded to zenoss-stack 2.4.2, and it seems that gaps started to appear in a lot of graphs after this. changing workers from 2 to 4 in zenhub seems to have helped somewhat, but many hosts still have ~ 1-5 hours gaps between every 0.1-1 hours of data.. could this be related to a known issue in 2.4.2?
[17-Aug-2009 10:52:36] <fuzz> it also seems that this only applies to graphs generated by zenperfsnmp
[17-Aug-2009 10:52:40] <fuzz> ping/process graphs seems fine
[17-Aug-2009 10:54:49] <baffle> And also graphs done by executing commands.
[17-Aug-2009 10:56:30] <rmatte> fuzz: network latency?
[17-Aug-2009 10:59:07] <fuzz> rmatte: i doubt it.. one of the affected boxes is on the same subnet as the zenoss box
[17-Aug-2009 11:00:27] <rmatte> hmmm
[17-Aug-2009 11:00:48] <rmatte> how many devices are you monitoring, and what kind of hardware is on the Zenoss server?
[17-Aug-2009 11:02:10] <fuzz> about 500 devices
[17-Aug-2009 11:02:13] <fuzz> let me check the hw
[17-Aug-2009 11:02:15] <baffle> HW is IBM x3650, Quad Core 2.33, 20G ram.
[17-Aug-2009 11:02:19] <fuzz> right
[17-Aug-2009 11:03:23] <baffle> RAID 10 15k disks if memory serves me right.
[17-Aug-2009 11:03:37] <baffle> fuzz: How many hosts are there again?
[17-Aug-2009 11:04:07] <fuzz> baffle: there are 425 dirs under perf/Devices so i guess the same number of hosts
[17-Aug-2009 11:05:17] <rmatte> hmmm, definitely not a hardware issue
[17-Aug-2009 11:05:52] <rmatte> Have you checked the zenperfsnmp log file?
[17-Aug-2009 11:06:10] <fuzz> yes.. i've grepped for a host i know has gaps in the graphs, and it is not mentioned
[17-Aug-2009 11:06:32] <fuzz> i also tried to log each rrd update and it does not seem to try to save rrd data for that host
[17-Aug-2009 11:06:40] <fuzz> don't know if it tries to poll it yet
[17-Aug-2009 11:07:36] <rmatte> have you played around with any collector settings?
[17-Aug-2009 11:07:40] <rmatte> such as cycle time?
[17-Aug-2009 11:09:28] <baffle> collector settings: http://paste2.org/p/383519
[17-Aug-2009 11:09:38] <baffle> SNMP is 300.
[17-Aug-2009 11:09:56] <baffle> I think it is the default.
[17-Aug-2009 11:10:01] <rmatte> yeh it is
[17-Aug-2009 11:10:13] <rmatte> are you monitoring a lot of servers, or mostly network?
[17-Aug-2009 11:10:15] <fuzz> btw.. the host i mentioned turned out to not be on the subnet after all. but i'm still pretty convinced it's not a network issue. i can snmpwalk the box fine from the zenoss box..
[17-Aug-2009 11:10:35] <baffle> rmatte: Most is servers, but there are some switches/routers as well.
[17-Aug-2009 11:10:45] <rmatte> estimate of how many?
[17-Aug-2009 11:11:03] <baffle> rmatte: Like.. 4 big switcher/routers, 10-20 switches.
[17-Aug-2009 11:11:11] <rmatte> servers I meant
[17-Aug-2009 11:11:12] <rmatte> sorry
[17-Aug-2009 11:11:25] <baffle> Around 400 I guess.
[17-Aug-2009 11:11:27] <rmatte> (servers are the most intensive to collect data for)
[17-Aug-2009 11:11:31] <rmatte> ah, quite a few
[17-Aug-2009 11:11:54] <rmatte> you upgraded to 2.4.2 from which version?
[17-Aug-2009 11:12:01] <baffle> 2.4.0 or 2.4.1.
[17-Aug-2009 11:12:19] <rmatte> hmmm
[17-Aug-2009 11:12:46] <baffle> rmatte: Actually, 4 big switches are Cisco 6513 wich loads of 48 port switchmodules, so.. I think those are the worst. :-)
[17-Aug-2009 11:13:22] <rmatte> well, possibly, are you monitoring port operational status via a threshold?
[17-Aug-2009 11:13:50] <rmatte> such as: http://www.zenoss.com/Members/cluther/polling-interface-status/
[17-Aug-2009 11:13:50] <baffle> Graphs for the switches are OK btw.
[17-Aug-2009 11:14:29] <rmatte> the servers that aren't graphing, if you load up a list of servers, is it the servers near the end of the list that aren't graphing
[17-Aug-2009 11:14:33] <rmatte> or does it seem random?
[17-Aug-2009 11:14:53] <baffle> rmatte: Hmm, no, I don't think we do.
[17-Aug-2009 11:15:08] <baffle> rmatte: What is the best way to get the list of servers?
[17-Aug-2009 11:16:12] <rmatte> hmmm, I guess you don't group them the same way we do lol...
[17-Aug-2009 11:16:38] <rmatte> we have organizers under Groups for Network, Security, and Server
[17-Aug-2009 11:16:53] <rmatte> not sure, really depends on how you organize them
[17-Aug-2009 11:17:00] <rmatte> are they all under the same devices organizer?
[17-Aug-2009 11:17:33] <baffle> rmatte: Hmm, we have /Network /Server etc, and under /Server we have OS and Vendor etc etc.
[17-Aug-2009 11:18:10] <baffle> But if servers are polled alphabetically, then there is no diffence from servers starting with a or z.
[17-Aug-2009 11:18:53] <rmatte> yeh, they are polled alphabetically in my experience
[17-Aug-2009 11:20:58] <fuzz> it may be worth mentioning that we had an issue (around the same time as the update) with modeling some devices that had a network interface with ip in 10.0.0.0/24. some python function recursed infinitely and eventually hit the max recursion limit and died, apparently because the 10.0.0.0 network object was a child of itself (looking at the object in the zope db showed 10.0.0.0/10.0.0.0/10.0.0.0/10.0.0.0/etc..). i had to delete the object from the zope db..
[17-Aug-2009 11:21:08] <fuzz> i wouldn't think it should be related, but who knows..
[17-Aug-2009 11:25:34] <rmatte> hmmmm
[17-Aug-2009 11:25:58] <rmatte> so increasing the hub workers did help a bit?
[17-Aug-2009 11:26:00] <fuzz> btw.. there are devices with gaps that does not have a network interface in 10.0.0.0
[17-Aug-2009 11:26:06] <baffle> rmatte: Not, not really.
[17-Aug-2009 11:26:31] <rmatte> fuzz: gaps?
[17-Aug-2009 11:26:41] <baffle> Actually, most servers have public IPs.
[17-Aug-2009 11:26:46] <fuzz> rmatte: gaps in the graphs.. missing data in the rrds..
[17-Aug-2009 11:27:02] <rmatte> ah
[17-Aug-2009 11:27:21] <rmatte> well, if the gaps are in 10.0.0.0, it may be related to that change
[17-Aug-2009 11:27:30] <baffle> No, graphs are everywhere.
[17-Aug-2009 11:27:54] <baffle> We have a few /16 and /17 nets, so servers are spread all over.
[17-Aug-2009 11:28:00] <rmatte> all I can think of is that snmp is timing out while polling
[17-Aug-2009 11:28:21] <rmatte> try increasing the cycle and snmp values for the collector
[17-Aug-2009 11:28:31] <rmatte> increase the snmp timeout value in zPropeties under /Devices as well
[17-Aug-2009 11:29:06] <fuzz> rmatte: would such a timeout be visible in the log? is there somewhere in the code i can add a debug statement to check it?
[17-Aug-2009 11:29:52] <rmatte> it would be visible I believe
[17-Aug-2009 11:30:10] <rmatte> but I don't have any logs as reference to check since I've got my stuff pretty well tuned
[17-Aug-2009 11:30:21] <rmatte> try this...
[17-Aug-2009 11:30:30] <rmatte> as zenoss user: zenperfsnmp run -v10
[17-Aug-2009 11:30:35] <rmatte> and watch it run
[17-Aug-2009 11:30:46] <rmatte> see if there are any errors
[17-Aug-2009 11:33:19] <fuzz> seems fine
[17-Aug-2009 11:33:20] <fuzz> http://pastebin.com/d105833dc
[17-Aug-2009 11:33:35] <fuzz> it completed suspiciously quickly
[17-Aug-2009 11:34:08] <fuzz> success:0 fail:0 pending:0 todo:0
[17-Aug-2009 11:34:09] <fuzz> hm?
[17-Aug-2009 11:36:44] <rmatte> yeh, it does the same for me, hmmm
[17-Aug-2009 11:37:22] <baffle> Maybe you should stop zenperfsnmp first.
[17-Aug-2009 11:37:26] <baffle> I.e. the daemon.
[17-Aug-2009 11:37:33] <rmatte> zenperfsnmp run -v 10 --cycle
[17-Aug-2009 11:37:36] <rmatte> try that
[17-Aug-2009 11:37:46] <rmatte> yeh, you might have to stop the daemon that's running
[17-Aug-2009 11:38:27] <rmatte> yeh --cycle works
[17-Aug-2009 11:38:58] <rmatte> or seems to anyways
[17-Aug-2009 11:39:17] <fuzz> yes.. that's better. running now.. no errors so far.. lots of "Updating threshold..."
[17-Aug-2009 11:39:25] <rmatte> yeh, that's normal
[17-Aug-2009 11:45:20] <fuzz> still running..
[17-Aug-2009 11:45:50] <rmatte> yeh, could take a while, you have a lot of servers lol
[17-Aug-2009 11:45:52] <fuzz> some warnings about '38 degrees C' cannot be parsed as float etc..
[17-Aug-2009 11:46:04] <fuzz> but not for (all) the relevant hosts
[17-Aug-2009 11:46:08] <rmatte> that wouldn't cause this
[17-Aug-2009 11:46:22] <rmatte> that's just complaing that a string can't be parsed as a numeric
[17-Aug-2009 11:46:31] <fuzz> right..
[17-Aug-2009 11:46:45] <fuzz> other warnings about hosts down and bad oids, but still not for (all) the relevant hosts
[17-Aug-2009 11:47:29] <rmatte> yeh, you're really just looking for failures, errors, or timeouts
[17-Aug-2009 11:50:31] <fuzz> mm
[17-Aug-2009 11:50:37] <fuzz> i'll get back to you when it's done :-)
[17-Aug-2009 11:50:58] <TreeStump> heya guys
[17-Aug-2009 11:51:11] <twm1010> Hello
[17-Aug-2009 11:51:49] <TreeStump> me again haha
[17-Aug-2009 11:52:02] <twm1010> had any luck?
[17-Aug-2009 11:52:17] <TreeStump> i reinstalled the os and zenoss
[17-Aug-2009 12:02:19] <TreeStump> so far so good
[17-Aug-2009 12:02:19] <TreeStump> im wodering though
[17-Aug-2009 12:02:20] <TreeStump> wondering*
[17-Aug-2009 12:02:26] <TreeStump> is it possible to add a zenuser who can only view the devices they are assigned to administer?
[17-Aug-2009 12:02:40] <rmatte> not in Core
[17-Aug-2009 12:02:53] <rmatte> device ACLs are a feature of enterprise
[17-Aug-2009 12:03:02] <TreeStump> nope?
[17-Aug-2009 12:03:06] <TreeStump> only in the enterprise version?
[17-Aug-2009 12:03:10] <rmatte> correct
[17-Aug-2009 12:03:10] <TreeStump> tim to see how much that costs then
[17-Aug-2009 12:03:28] <TreeStump> time*
[17-Aug-2009 12:03:30] <TreeStump> thanks
[17-Aug-2009 12:03:43] <rmatte> http://www.zenoss.com/product/pricing
[17-Aug-2009 12:03:45] <rmatte> np
[17-Aug-2009 12:07:09] <TreeStump> i thought it was cheap
[17-Aug-2009 12:07:16] <TreeStump> then I just realised
[17-Aug-2009 12:07:25] <TreeStump> its per resource
[17-Aug-2009 12:07:45] <TreeStump> haha
[17-Aug-2009 12:07:48] <TreeStump> oh man
[17-Aug-2009 12:09:16] <rmatte> yeh, it's quite pricy
[17-Aug-2009 12:10:31] <rmatte> You can assign admin access to specific devices per user in Core
[17-Aug-2009 12:10:38] <rmatte> but they'll still be able to view everything
[17-Aug-2009 12:11:18] <rmatte> If you go to Settings -> Users -> <click on a user> -> Administered Objects
[17-Aug-2009 12:13:02] <rmatte> http://www.zenoss.com/community/docs/zenoss-guide/2.2.4/ch06s02.html
[17-Aug-2009 12:13:39] <straterra> I'm about to try a zenoss migration
[17-Aug-2009 12:17:44] <TreeStump> matte: thanks, I already tried that because I assumed that it would only allow them access to those devices... thanks anyways...
[17-Aug-2009 12:18:41] <TreeStump> actually for the functionality I need, its probably cheaper for me to buy heaps of VPS's and just create multiple installs of zenoss
[17-Aug-2009 12:19:45] <fuzz> rmatte: the --cycle flag makes zenperfsnmp run over and over again? i have multiple "*** Cycle completed ***, Sent xxx oid requests, Queried 297 devices" entries in the output now..
[17-Aug-2009 12:21:52] <rmatte> fuzz: I believe it simulates running as a daemon, but in the foreground
[17-Aug-2009 12:22:02] <rmatte> ctrl-c out of it and the background daemon will take over
[17-Aug-2009 12:22:03] <fuzz> rmatte: grepped for CRITICAL in the output now, and i have see no relevant lines, only misc (other) oids that could not be polled/saved
[17-Aug-2009 12:22:13] <rmatte> grep for fail:
[17-Aug-2009 12:22:15] <rmatte> and error
[17-Aug-2009 12:23:57] <fuzz> well.. there are quite a few matches
[17-Aug-2009 12:24:05] <fuzz> but not much relevant it seems
[17-Aug-2009 12:24:11] <fuzz> INFO:zen.zenperfsnmp:success:289 fail:2 pending:6 todo:0
[17-Aug-2009 12:25:14] <fuzz> nothing for |grep -i fail|grep <troubledevice> or |grep -i error|grep <troubledevice>
[17-Aug-2009 12:25:21] <fuzz> however..
[17-Aug-2009 12:25:35] <fuzz> grepping for trouble devices now.. it suddenly seems to have updated some rrds
[17-Aug-2009 12:27:34] <TreeStump> do you think it would be possible to write a zenpack(plugin) which would allow the extra ACL functionality?
[17-Aug-2009 12:28:30] <chudler> yes, it is possible. It is usually how the enterprise features work
[17-Aug-2009 12:29:44] <TreeStump> because there is no way im gonna have enough devices to even qualify for buying enterprise
[17-Aug-2009 12:32:01] <baffle> TreeStump: We have a custom portal for our customers, where they get access to their graphs via a magic proxy.
[17-Aug-2009 12:33:47] <malbon> yo folks.
[17-Aug-2009 12:34:39] <malbon> has anyone used sha or md5 authentication digests with the web front end? I would like to use rest calls with some of our apps but don't want to give the developers and users the real password.
[17-Aug-2009 12:35:48] <rmatte> fuzz: well, you said the problem is intermittent, so checking logs during the times that the data wasn't being collected would be your best bet, you'd have to make sure that the logs haven't rolled over since then as well
[17-Aug-2009 12:40:41] <fuzz> rmatte: i grepped for a few trouble devices in the log earlier (the log should not have rolled over by then), and there were no matches..
[17-Aug-2009 12:40:42] <baffle> malbon: I.e. set up anonymous auth and allow people in with apache auth? Might work.
[17-Aug-2009 12:40:53] <baffle> fuzz: But did the RRDs update?
[17-Aug-2009 12:42:04] <fuzz> baffle: for appdev (a trouble device) they did now yes..
[17-Aug-2009 12:47:16] <rmatte> fuzz: I'm out of ideas then
[17-Aug-2009 12:47:22] <rmatte> maybe check zenhub log?
[17-Aug-2009 12:49:45] <fuzz> actually.. it seems it has updated almost all the rrds now, after the zenperf.. -cycle.. had run for a while
[17-Aug-2009 12:50:11] <rmatte> sounds like a timeout issue then
[17-Aug-2009 12:50:16] <fuzz> baffle: /srv/zenoss/check-dead-rrds.py [maxage] - shows rrds without data since maxage
[17-Aug-2009 12:50:27] <rmatte> increase your timeout values and cycle time by 100 or so
[17-Aug-2009 12:50:43] <fuzz> i guess.. strange that there was no entries in the log though..
[17-Aug-2009 12:50:44] <fuzz> will do
[17-Aug-2009 12:51:59] <malbon> baffle: yes, I was thinking about that to kerberise it, but I'd prefer not to go down that route at this time.
[17-Aug-2009 12:52:35] <straterra> When I migrate to a new zenoss install..do I have to apply templates to their groups before I do the import..or will the import that care of that?
[17-Aug-2009 12:52:59] <baffle> fuzz: Maybe we should update zenperfsnmp threads?
[17-Aug-2009 12:53:35] <rmatte> straterra: importing devices?
[17-Aug-2009 12:53:48] <straterra> No
[17-Aug-2009 12:53:58] <straterra> I'm doing a full backup of one zenoss and importing in a fresh install of another
[17-Aug-2009 12:54:14] <straterra> Basically from a VM to a real machine
[17-Aug-2009 12:54:17] <rmatte> I see, I'd apply them beforehand just to be safe
[17-Aug-2009 12:54:23] <rmatte> you'll have to apply them either way
[17-Aug-2009 12:55:24] <straterra> by hand again?
[17-Aug-2009 12:55:43] <rmatte> I don't know what setup you had before
[17-Aug-2009 12:55:49] <rmatte> explain?
[17-Aug-2009 12:56:00] <straterra> I have a Centos 5.3 vm with everything configured how I want it
[17-Aug-2009 12:56:15] <straterra> I want to move it to a real machine..so..I have a machine installed with centos 5.3
[17-Aug-2009 12:56:26] <rmatte> why not just copy the whole $ZENHOME directory over?
[17-Aug-2009 12:56:29] <straterra> I was going to do a backup in zenoss for the vm and import it
[17-Aug-2009 12:56:47] <straterra> I have to dump the db and such too, right?
[17-Aug-2009 12:57:07] <rmatte> you have to backup anything in the mysql DB, but not in Zope
[17-Aug-2009 12:57:19] <rmatte> if you copy $ZENHOME the zope db will be copied along with it
[17-Aug-2009 12:57:21] <rmatte> it's just a flat file
[17-Aug-2009 12:57:32] <straterra> If I rsync $ZENHOME over, and then do a db dump/import..I should be fine?
[17-Aug-2009 12:57:43] <rmatte> yes
[17-Aug-2009 12:57:56] <rmatte> If it doesn't work I'll be really surprised
[17-Aug-2009 12:58:03] <rmatte> did you use stack installer to install though?
[17-Aug-2009 12:58:06] <straterra> Might as well try it
[17-Aug-2009 12:58:09] <straterra> I used the RHEL RPM's
[17-Aug-2009 12:58:13] <rmatte> hmmmm
[17-Aug-2009 12:58:21] <straterra> Same OS..same RPM's
[17-Aug-2009 12:58:29] <rmatte> you might want to do a clean install of Zenoss, then copy $ZENHOME over, then dump/import
[17-Aug-2009 12:58:39] <straterra> I have a clean install of zenoss
[17-Aug-2009 12:58:43] <straterra> Just installed it
[17-Aug-2009 12:58:50] <rmatte> same version?
[17-Aug-2009 12:58:54] <straterra> Yes
[17-Aug-2009 12:58:58] <straterra> Everything is the same
[17-Aug-2009 12:59:05] <rmatte> then yes, copy $ZENHOME over and dump/import should work fine
[17-Aug-2009 12:59:18] <straterra> Alrighty..waiting on this slow VM to shut down zenoss
[17-Aug-2009 12:59:23] <rmatte> you should essentially have the exact same settings that you had with the old install
[17-Aug-2009 12:59:30] <rmatte> lol
[17-Aug-2009 12:59:40] <straterra> Thats why we're moving to a dedicated machine
[17-Aug-2009 12:59:43] <straterra> The VM is ungodly slow
[17-Aug-2009 12:59:54] <rmatte> we have VMs, but they are on fast servers
[17-Aug-2009 13:00:15] <straterra> Our servers are old..and are supposed to be upgraded extremely soon
[17-Aug-2009 13:00:28] <rmatte> hehe
[17-Aug-2009 13:00:40] <rmatte> some companies don't understand the words "end of life cycle"
[17-Aug-2009 13:01:32] <straterra> God, I know
[17-Aug-2009 13:01:50] <straterra> zenoss still isn't shut down
[17-Aug-2009 13:02:01] <venturaville> ditto (said while staring at a long list of 10 year old boxes in inventory)
[17-Aug-2009 13:02:03] <rmatte> that's brutal
[17-Aug-2009 13:02:11] <rmatte> I'm surprised it ran at all on that thing
[17-Aug-2009 13:02:24] <scp2004> Hi guys ..Im a Zenoss user from Brazil . And monitoring 475 hosts .. I need some help with a problem . Somebody ?
[17-Aug-2009 13:02:25] <straterra> It wasn't pretty
[17-Aug-2009 13:02:36] <straterra> scp2004: sure..just state your issue
[17-Aug-2009 13:03:16] <rmatte> scp2004: just ask and we'll help you as best we can
[17-Aug-2009 13:03:49] <scp2004> Thanks .. I m using zenoss 2.4.1 on Freebsd 7.2 following the casey tutorial on Zenoss Forum ... But the daemon zentrap never come up !
[17-Aug-2009 13:04:11] <straterra> Thank jesus for slow zopectl
[17-Aug-2009 13:04:26] <scp2004> I tried all .. Install on Bsd 7.0 , 7.1 .. Installing zenoss 2.4.0 , 2.4.1 but still same
[17-Aug-2009 13:05:09] <scp2004> Matt Ray told me to speak with you
[17-Aug-2009 13:07:16] <scp2004> straterra Thanks .. I m using zenoss 2.4.1 on Freebsd 7.2 following the casey tutorial on Zenoss Forum ... But the daemon zentrap never come up !
[17-Aug-2009 13:07:39] <scp2004> [straterra] I tried all .. Install on Bsd 7.0 , 7.1 .. Installing zenoss 2.4.0 , 2.4.1 but still same
[17-Aug-2009 13:12:02] <scp2004> Somebody help ?
[17-Aug-2009 13:13:38] <straterra> rmatte: do you know what db zenoss uses by default
[17-Aug-2009 13:13:40] <straterra> ?^
[17-Aug-2009 13:14:11] <scp2004> events
[17-Aug-2009 13:14:12] <baffle> scp2004: You have ofcourse checked the logfile?
[17-Aug-2009 13:14:30] <scp2004> yes . but in this case , dont make any log
[17-Aug-2009 13:14:52] <scp2004> when i try start zentrap manually with the command ./zentrap start its the same
[17-Aug-2009 13:15:01] <scp2004> dont come up ..and dont make any log
[17-Aug-2009 13:15:24] <baffle> scp2004: Did you try the debug flags?
[17-Aug-2009 13:15:40] <scp2004> its not problem with same port cause i just checked
[17-Aug-2009 13:16:24] <scp2004> <baffle> are you talking about ./zentrap run or ./zentrap --v10 ..something like this ?
[17-Aug-2009 13:17:31] <scp2004> scp2004 Are you ?
[17-Aug-2009 13:17:37] <scp2004> [scp2004] Are you ?
[17-Aug-2009 13:18:42] <rmatte> straterra: MySQL for events, Zope for everything else
[17-Aug-2009 13:19:05] <straterra> Right..but just the events db?
[17-Aug-2009 13:19:11] <rmatte> correct
[17-Aug-2009 13:19:22] <rmatte> everything else is in Zope
[17-Aug-2009 13:19:34] <scp2004> the name of db is events man
[17-Aug-2009 13:19:55] <scp2004> Somebody knows why my zentrap dont come up in freeBsd ?
[17-Aug-2009 13:24:33] <rmatte> zentrap doesn't start?
[17-Aug-2009 13:28:37] <rmatte> try: zentrap run -v10
[17-Aug-2009 13:28:39] <rmatte> check for errors
[17-Aug-2009 13:28:55] <rmatte> Has anything been announced as to when 2.4.3 is going to be released?
[17-Aug-2009 13:43:36] <scp2004> [rmatte] here the error : Segmentation fault: 11 (core dumped)
[17-Aug-2009 13:47:53] <straterra> Wow..the new monitoring server for zenoss is now...200 megs in to swap
[17-Aug-2009 13:53:40] <scp2004> when i try ./zentrap run --v10 here the erro : Segmentation fault: 11 (core dumped) .. Any idea ?
[17-Aug-2009 14:04:02] <baffle> scp2004: zentrap run -v10 or something, yes.
[17-Aug-2009 14:04:29] <baffle> scp2004: Strange, it's made in Python..
[17-Aug-2009 14:04:43] <baffle> scp2004: It should produce a coredump, then.
[17-Aug-2009 14:12:45] <scp2004> how i debug it in Pyton to test ?
[17-Aug-2009 14:12:58] <scp2004> [baffle] how i debug it in Pyton to test ?
[17-Aug-2009 14:16:27] <baffle> scp2004: To be honest, I have no idea why it crashes.. Maybe there is a problem with one of the bundled modules? Maybe you can use truss or something to see what happens?
[17-Aug-2009 14:22:53] <rmatte> scp2004: more importantly, what did you change recently?
[17-Aug-2009 14:23:03] <rmatte> something must have changed to cause that, things don't generally break on their own
[17-Aug-2009 14:23:50] <scp2004> i dont change anything man .. I just install following the casey tutorial in zenoss forum and it doesnt work !
[17-Aug-2009 14:24:29] <rmatte> ah, you're a BSD user?
[17-Aug-2009 14:25:01] <scp2004> [rmatte] i tried install it again more than 10 times and its the same .. Only zentrap doesnt run
[17-Aug-2009 14:25:01] <scp2004> Yes man .. I m using freeBsd 7.2
[17-Aug-2009 14:25:15] <scp2004> any idea ?
[17-Aug-2009 14:26:08] <rmatte> nope, Zenoss with FreeBSD is quite shaky at the moment
[17-Aug-2009 14:26:15] <rmatte> very hard to support
[17-Aug-2009 14:26:34] <rmatte> just looking at this guide makes my eyes water
[17-Aug-2009 14:26:37] <rmatte> so many tweaks
[17-Aug-2009 14:27:18] <rmatte> you're literally editing code to get it to compile, any one of the changes could cause segfaults
[17-Aug-2009 14:28:27] <rmatte> I'd go through and double/triple check your edits
[17-Aug-2009 14:30:04] <rmatte> also, the guide was written for 2.4.1, it's possible that it won't work properly with 2.4.2 due to changes made between versions
[17-Aug-2009 14:30:06] <rmatte> but no idea
[17-Aug-2009 14:30:52] <scp2004> Im using 2.4.1 ..Ok ..thank you ill try ask to casey
[17-Aug-2009 14:30:53] <scp2004> ty
[17-Aug-2009 14:31:24] <rmatte> Yeh, Casey would be the one to speak with, I'll have to setup a virtual BSD server and try this out sometime so that I'm more helpful with FreeBSD related problems
[17-Aug-2009 14:32:07] <rmatte> I used FreeBSD for 5 years, it's nice and stable but installing apps can be a huge chore
[17-Aug-2009 14:32:13] <rmatte> which is ultimately why I switched to Linux
[17-Aug-2009 14:32:51] <scp2004> [rmatte] Will help me a lot if u do this .. Cause i have no idea that other user that have sucess in run zenoss on FreeBsd 7.2 ..
[17-Aug-2009 14:33:11] <rmatte> there aren't many, email Casey and see if they can help
[17-Aug-2009 14:33:16] <scp2004> If you can .. Try install it following the casey steps .. To verify if we have the same problem
[17-Aug-2009 14:33:23] <rmatte> k
[17-Aug-2009 14:33:32] <scp2004> ty ...bye
[17-Aug-2009 14:37:14] <rmatte> or he could just use Linux for his Zenoss server lol
[17-Aug-2009 14:38:14] <cwj> and violate the purity of his GNU-free network? why the very thought..
[17-Aug-2009 14:38:28] <rmatte> lol
[17-Aug-2009 14:38:40] <rmatte> they use RedHat there as well apparently
[17-Aug-2009 14:38:45] <rmatte> so it's not exactly GNU free
[17-Aug-2009 14:39:05] <cwj> well im not really one to pass up an opportunity for poking fun
[17-Aug-2009 14:39:20] <rmatte> hehe
[17-Aug-2009 15:03:48] <sacpinball> Error collecting performance data: NT_STATUS_NET_WRITE_FAULT
[17-Aug-2009 15:04:02] <sacpinball> WMI on a windows box
[17-Aug-2009 15:13:40] <rmatte> sacpinball: http://forums.zenoss.com/viewtopic.php?p=37591
[17-Aug-2009 15:13:43] <rmatte> that thread may help you
[17-Aug-2009 15:13:52] <sacpinball> thx.. will give it a look
[17-Aug-2009 15:21:46] <rmatte> oh great, I was playing with some report plugins on a server and now it's messed up
[17-Aug-2009 15:21:51] <rmatte> grrr
[17-Aug-2009 15:22:04] <sacpinball> just restore from backup
[17-Aug-2009 15:22:04] <sacpinball> lol
[17-Aug-2009 15:22:40] <rmatte> nah, I need to fix this, it's a production box
[17-Aug-2009 15:24:16] <rmatte> finally
[17-Aug-2009 15:38:09] <rmatte> man, these default reports are horrible
[17-Aug-2009 15:38:23] <rmatte> the CPU report divides values by 100 even though they are already collected as a percentage
[17-Aug-2009 15:39:00] <rmatte> so they end up being like 0.1 instead of 10 percent
[17-Aug-2009 16:10:37] <rmatte> nice, I just found a new feature that I didn't know about lol
[17-Aug-2009 16:11:00] <rmatte> if you go to a device class, then go to More -> All Templates it lists all the local templates
[17-Aug-2009 16:11:15] <rmatte> I never used to be able to keep track of all the local ones I had created
[17-Aug-2009 16:11:56] <cwj> someone in this channel pointed that out to me recently
[17-Aug-2009 16:12:00] <cwj> very very helpful
[17-Aug-2009 16:12:05] <rmatte> yeh, very
[17-Aug-2009 16:12:21] <cwj> now if only i could copy templates up the device tree without using the zope manager interface
[17-Aug-2009 16:12:24] <cwj> maybe someday
[17-Aug-2009 16:12:58] <rmatte> yeh
[17-Aug-2009 16:13:10] <rmatte> zope manager interface can be super handy for copying templates between classes
[17-Aug-2009 16:13:24] <cwj> yep
[17-Aug-2009 16:13:38] <cwj> thats actually the only reason ive needed to use it thus far
[17-Aug-2009 16:13:43] <rmatte> actually, thanks for reminding me about that, I need to use that in a few hehe
[17-Aug-2009 16:34:50] <rmatte> anyone who's good at python...
[17-Aug-2009 16:34:51] <rmatte> memFree = d.getRRDValue('mem5minFree', **summary)
[17-Aug-2009 16:34:52] <rmatte> memUsed = d.getRRDValue('mem5minUsed', **summary)
[17-Aug-2009 16:34:52] <rmatte> percentUsed = memFree / memUsed * 100
[17-Aug-2009 16:35:07] <rmatte> that last line, it's not right since I'm getting an error
[17-Aug-2009 16:35:18] <rmatte> I need memFree divided by memUsed then multiplied by 100
[17-Aug-2009 16:35:33] <rmatte> I also need to cut the number down so that it doesn't show a bunch of decimals
[17-Aug-2009 16:35:36] <rmatte> ideas?
[17-Aug-2009 16:53:06] <rmatte> hmmm, after testing it seems that dividing a smaller number by a larger number is python results in "0"
[17-Aug-2009 16:53:11] <rmatte> any way around this?
[17-Aug-2009 16:58:01] <rmatte> think I got it
[17-Aug-2009 16:58:02] <rmatte> round((1.0*memUsed/memFree)*100)
[17-Aug-2009 17:02:04] <rmatte> this code should be working but it's not
[17-Aug-2009 17:02:04] <rmatte> if percentUsed is None:
[17-Aug-2009 17:02:04] <rmatte> memFree = round(d.getRRDValue('mem5minFree', **summary))
[17-Aug-2009 17:02:04] <rmatte> memUsed = round(d.getRRDValue('mem5minUsed', **summary))
[17-Aug-2009 17:02:05] <rmatte> percentUsed = round((1.0*memUsed/memFree)*100)
[17-Aug-2009 17:02:08] <rmatte> no idea why
[17-Aug-2009 17:09:09] <rmatte> anyways, I'm out of here
[18-Aug-2009 00:00:46] [disconnected at Tue Aug 18 00:00:46 2009]
[18-Aug-2009 00:00:46] [connected at Tue Aug 18 00:00:46 2009]
[18-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[18-Aug-2009 03:12:06] [disconnected at Tue Aug 18 03:12:06 2009]
[19-Aug-2009 00:00:46] [connected at Wed Aug 19 00:00:46 2009]
[19-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[19-Aug-2009 06:28:25] <varbouk> Hello all
[19-Aug-2009 08:18:13] <jb> morning!
[19-Aug-2009 08:23:53] <jb> cgibbons: around?
[19-Aug-2009 08:25:22] <DogWater> can zenoss make systems fire off yum updates, etc?
[19-Aug-2009 08:26:08] <jb> not out of the box..
[19-Aug-2009 08:35:22] <cgibbons> yeah
[19-Aug-2009 08:37:13] <jb> cgibbons: i'm trying to link to a new product name in a skin using productLink like I have seen in several other zenpacks..
[19-Aug-2009 08:37:29] <jb> but it is not working (no data is displayed)
[19-Aug-2009 08:37:35] <jb> can you think of anything that I would be missing?
[19-Aug-2009 08:38:08] <jb> <td class="tablevalues" tal:content="structure frame/getProductLink"/>
[19-Aug-2009 08:38:12] <jb> is what I'm trying to use..
[19-Aug-2009 08:38:21] <jb> is it because maybe the model is not in the database?
[19-Aug-2009 08:39:39] <cgibbons> no clue really but lemme poke around a bit
[19-Aug-2009 08:39:49] <jb> thanks!
[19-Aug-2009 08:47:00] <jb> sup venturaville..
[19-Aug-2009 08:47:07] <venturaville> jb: howdy :-)
[19-Aug-2009 08:47:36] <ckrough> jb: what does productLink do?
[19-Aug-2009 08:48:17] <jb> ckrough: it looks like it links you to related products in the zopedb.. ie: http://fc-zenoss01:8080/zport/dmd/Manufacturers/Unknown/products/DELL%20%20%20%20%20ST373455SS
[19-Aug-2009 08:48:40] <jb> basically, I have a table that displays information about a product, one field being the "Model"
[19-Aug-2009 08:48:51] <jb> i want to be able to click on that model, and show all other products with the same model
[19-Aug-2009 08:50:35] <jb> if that makes any since..
[19-Aug-2009 08:50:37] <jb> sense
[19-Aug-2009 08:50:58] <ckrough> yup
[19-Aug-2009 09:01:19] <Mituc> hi guys!
[19-Aug-2009 09:01:37] <Mituc> for some reason the zencommand process crashes every few days
[19-Aug-2009 09:01:45] <Mituc> I can't find anything into the logs
[19-Aug-2009 09:02:08] <Mituc> and I can't find a list with the configuration directive it supports appart of what it already has set by default (monitor localhost)
[19-Aug-2009 09:02:29] <mrayzenoss> go to Settings->Daemons->Edit Config and you can see the options
[19-Aug-2009 09:02:37] <Mituc> is there any way of adding a debug directive in the configuration file instead of launching it manually with -v10?
[19-Aug-2009 09:02:50] <Mituc> mrayzenoss: that's what I'm looking at..
[19-Aug-2009 09:02:50] <mrayzenoss> what I just said will work
[19-Aug-2009 09:03:09] <Mituc> but I can only see the already existing options there
[19-Aug-2009 09:03:22] <mrayzenoss> ahh, are you using pre-2.4?
[19-Aug-2009 09:03:23] <Mituc> not all the configuration directives that it supports
[19-Aug-2009 09:03:26] <Mituc> yep
[19-Aug-2009 09:03:29] <Mituc> pre 2.4
[19-Aug-2009 09:03:56] <Mituc> 2.3.0 to be more precise
[19-Aug-2009 09:04:07] <Mituc> I'm using the RPM version
[19-Aug-2009 09:04:27] <Mituc> and I'm not sure if upgrading the RPM to a higher version will also do all the migration stuff that it's supposed to
[19-Aug-2009 09:04:58] <Mituc> not that I'm lazy to check, but I didn't have time to do that since I didn't need to touch the current setup
[19-Aug-2009 09:06:10] <Mituc> ok
[19-Aug-2009 09:06:20] <Mituc> so I'll probably have to look into upgrading it
[19-Aug-2009 09:14:26] <rmatte> good morning folks
[19-Aug-2009 09:15:10] <jb> sup
[19-Aug-2009 09:15:15] <mrayzenoss> morning
[19-Aug-2009 09:16:52] <rmatte> I get to setup 2 new Zenoss boxes today, yay
[19-Aug-2009 09:17:30] * ckrough is setting up irc services... longest. config. evar.
[19-Aug-2009 09:17:43] <rmatte> hehe, I haven't setup IRC services in years
[19-Aug-2009 09:17:45] <jb> using what ircd?
[19-Aug-2009 09:17:53] <ckrough> inspircd and anope
[19-Aug-2009 09:17:57] <rmatte> I've done it for ratbox, bahamut, and unrealircd
[19-Aug-2009 09:18:02] <jb> ratbox ftw
[19-Aug-2009 09:18:08] * jb = efnet
[19-Aug-2009 09:18:08] <rmatte> yeh, ratbox is awesome
[19-Aug-2009 09:18:13] <jb> i know the ratbox devs..
[19-Aug-2009 09:18:15] <ckrough> is ratbox good? was going to try it, but inspircd has been working well
[19-Aug-2009 09:18:30] <rmatte> ckrough: ratbox is excellent
[19-Aug-2009 09:18:33] <ckrough> hmm
[19-Aug-2009 09:18:37] <jb> yeah, ratbox is very stable..
[19-Aug-2009 09:19:04] <rmatte> yeh, we've been running it on our IRC network for like 3 or 4 years now, and haven't had any problems with it
[19-Aug-2009 09:19:47] <jb> def needsCleaningString(self):
[19-Aug-2009 09:19:47] <jb> clstatusmap = {1: 'Yes', 2: 'No'}
[19-Aug-2009 09:19:47] <jb> clstatus = self.cacheRRDValue('needsCleaning', 'Unknown')
[19-Aug-2009 09:19:48] <jb> return clstatusmap.get(clstatus, 'Unknown')
[19-Aug-2009 09:19:58] <jb> argh, why is this always returning 'Unknown'
[19-Aug-2009 09:20:05] <jb> even though needsCleaning != 0
[19-Aug-2009 09:20:51] <jb> oh.
[19-Aug-2009 09:20:53] <jb> stupid typo
[19-Aug-2009 09:21:13] <rmatte> ?
[19-Aug-2009 09:21:26] <rmatte> Values or something?
[19-Aug-2009 09:21:30] <jb> my modeler was using "needCleaning," not "needsCleaning"
[19-Aug-2009 09:21:34] <rmatte> ah
[19-Aug-2009 09:21:35] <rmatte> lol
[19-Aug-2009 09:21:41] <jb> hopefully thats it
[19-Aug-2009 09:21:47] <rmatte> I'd imagine it is
[19-Aug-2009 09:23:00] <ckrough> does this just return 1,0,or unknown? self.cacheRRDValue('needsCleaning', 'Unknown')
[19-Aug-2009 09:23:16] <jb> 1, 2, unknown
[19-Aug-2009 09:23:19] <ckrough> 1,2,Unknown I mean
[19-Aug-2009 09:23:21] <jb> yeah
[19-Aug-2009 09:23:31] <jb> im a bit confused as to what the cacheRRDValue is for..
[19-Aug-2009 09:23:36] <jb> but im trying to follow an example
[19-Aug-2009 09:23:54] <rmatte> it probably pulls from cached data
[19-Aug-2009 09:23:59] <rmatte> rather than from the RRD itself
[19-Aug-2009 09:24:02] <rmatte> which speeds it up
[19-Aug-2009 09:24:13] <rmatte> just a guess
[19-Aug-2009 09:24:38] <ckrough> "read an RRDValue with and cache it"
[19-Aug-2009 09:24:44] <ckrough> no sure that helps
[19-Aug-2009 09:24:54] <rmatte> yeh, just saw that hehe
[19-Aug-2009 09:24:59] <jb> i think it has something to do with a threshold
[19-Aug-2009 09:25:00] <rmatte> on huihoo.com
[19-Aug-2009 09:25:29] <rmatte> it reads the RRD value then caches it
[19-Aug-2009 09:25:52] <jb> to be checked against a threshold? .. i think
[19-Aug-2009 09:26:11] <ckrough> it looks like it tries to read value from _cache as well?
[19-Aug-2009 09:26:15] <jb> like, if it needs cleaned (ie, needsCleaningString = 1) trigger a threshold
[19-Aug-2009 09:26:18] <ckrough> and it its not there it creates teh cache
[19-Aug-2009 09:27:00] <rmatte> jb: I highly doubt that function directly relates to thresholds
[19-Aug-2009 09:27:13] <jb> i can't figure out why rrdCache was used then
[19-Aug-2009 09:27:14] <jb> for th is
[19-Aug-2009 09:27:16] <jb> this..
[19-Aug-2009 09:27:21] <rmatte> jb: It reads an RRD value and then commits it to cache if it's not already there, that's all the function does
[19-Aug-2009 09:27:23] <jb> other values/vars do not use it
[19-Aug-2009 09:27:28] <ckrough> it looks like it reads/caches rrd values
[19-Aug-2009 09:27:32] <rmatte> if you want it to apply to a threshold, you need to script it
[19-Aug-2009 09:27:43] <ckrough> what rmatte said.
[19-Aug-2009 09:27:47] <jb> yeah
[19-Aug-2009 09:29:29] <ckrough> the four stages of zenoss admins: "MYEYES!", "Why would they do it that way??", "Detective", "Chet"
[19-Aug-2009 09:29:39] <rmatte> lol
[19-Aug-2009 09:29:54] <rmatte> yeh, Chet does generally have the answers for sure
[19-Aug-2009 09:31:20] <Disconnect> ckrough: you skipped "omg what did i ever do to you?" and "where did my charts go??" (not necessarily in that order)
[19-Aug-2009 09:31:28] <ckrough> lol
[19-Aug-2009 09:31:29] <jb> argh wtf
[19-Aug-2009 09:31:41] <ckrough> "wait, that interface cant push 2.5 petabits per second"
[19-Aug-2009 09:31:57] <mrayzenoss> I'm pretty sure 'Chet' is an unattainable goal for mortals
[19-Aug-2009 09:32:03] <Disconnect> and step one is usually "hey, that looks neat!" otherwise you'd never get to step 2.
[19-Aug-2009 09:32:44] <jb> still returning unknown..
[19-Aug-2009 09:33:29] <rmatte> mrayzenoss: Attaining Chethood seems like a valid challenge, I shall attempt it
[19-Aug-2009 09:34:17] <jb> def getRRDNames(self):
[19-Aug-2009 09:34:17] <jb> """
[19-Aug-2009 09:34:17] <jb> Return the datapoint name of this access device
[19-Aug-2009 09:34:17] <jb> 'needsCleaning_needsCleaning' and 'status_status'
[19-Aug-2009 09:34:17] <jb> """
[19-Aug-2009 09:34:18] <jb> return ['needsCleaning_needsCleaning', 'status_status'
[19-Aug-2009 09:34:22] <jb> hm
[19-Aug-2009 09:34:25] <rmatte> (Apparently the secret is to sell your soul to a penguin, eat an apple, and then feed a mouse to a python)
[19-Aug-2009 09:34:42] <rmatte> but shhhhh
[19-Aug-2009 09:35:14] <rmatte> ummm, do you have to specify the datapoint name and the data source?
[19-Aug-2009 09:35:36] <rmatte> datapoint names would technically just be needsCleaning and status
[19-Aug-2009 09:36:02] <rmatte> but I may be wrong in this case
[19-Aug-2009 09:36:20] <rmatte> are you pulling that format straight for an example?
[19-Aug-2009 09:36:25] <rmatte> from*
[19-Aug-2009 09:36:26] <jb> yeah
[19-Aug-2009 09:36:29] <rmatte> ah, ok
[19-Aug-2009 09:36:33] <rmatte> probably right then hehe
[19-Aug-2009 09:36:41] <jb> still trying to figure out why needsCleaningString is returning unknown though
[19-Aug-2009 09:42:36] <rmatte> did you restart Zenoss after modifying your collector?
[19-Aug-2009 09:43:51] <jb> yeah..
[19-Aug-2009 09:47:13] <rmatte> no idea then, it should be able to pull the value if it's being set properly
[19-Aug-2009 09:48:00] <twm1010> morning
[19-Aug-2009 09:48:26] <rmatte> good morning
[19-Aug-2009 09:51:05] <twm1010> ever get an answer to when 2.4.3 is expected?
[19-Aug-2009 09:52:29] <straterra> rmatte: is adding mibs for a new type of device hard?
[19-Aug-2009 09:52:47] <straterra> I want to create a new type of device for our UPS..and graph its information
[19-Aug-2009 09:54:50] <rmatte> you don't need Mibs to graph information
[19-Aug-2009 09:54:53] <rmatte> just for traps
[19-Aug-2009 09:54:57] <rmatte> will the UPS be sending traps?
[19-Aug-2009 09:55:26] <straterra> No
[19-Aug-2009 09:55:27] <mrayzenoss> twm1010: should be early next week
[19-Aug-2009 09:55:41] <straterra> Why don't I need mibs?
[19-Aug-2009 09:55:42] <twm1010> thanks matt
[19-Aug-2009 09:55:47] <rmatte> straterra: then you don't need Mibs, you just need to find what OIDs correspond to the information that you want to gather
[19-Aug-2009 09:55:54] <straterra> Oh..thats what I man
[19-Aug-2009 09:55:56] <straterra> mean^
[19-Aug-2009 09:55:58] <rmatte> straterra: then create a device template and apply it to that class
[19-Aug-2009 09:55:58] <twm1010> straterra: MIBs are for classifying traps, not gathering peformance data
[19-Aug-2009 09:56:00] <straterra> I have the OID's
[19-Aug-2009 09:56:06] <twm1010> what brand UPS curiously?
[19-Aug-2009 09:56:12] <rmatte> straterra: ok, then just create a new Device template
[19-Aug-2009 09:56:14] <straterra> liebert
[19-Aug-2009 09:56:27] <straterra> I went through and mapped damn near every OID heh
[19-Aug-2009 09:56:32] <twm1010> does it use the standard UPS mib or a custom one?
[19-Aug-2009 09:56:37] <straterra> No idea
[19-Aug-2009 09:57:00] <rmatte> mapped every OID?
[19-Aug-2009 09:57:03] <straterra> yeha
[19-Aug-2009 09:57:05] <straterra> yeah
[19-Aug-2009 09:57:09] <rmatte> like, to a device template?
[19-Aug-2009 09:57:20] <straterra> Did an snmp walk..and matched up the OID information to information about the UPS
[19-Aug-2009 09:57:28] <rmatte> ah, gotcha
[19-Aug-2009 09:57:36] <rmatte> have you not created a device template before?
[19-Aug-2009 09:57:38] <straterra> like charge percentage, time remaining, input lines, output lines, frequency, etc etc
[19-Aug-2009 09:57:41] <straterra> Nope
[19-Aug-2009 09:57:47] <twm1010> there is a standard, RFC1628 mib that shows you the OID's you want
[19-Aug-2009 09:57:51] <twm1010> is is three-phase
[19-Aug-2009 09:57:54] <twm1010> ?
[19-Aug-2009 09:57:58] <straterra> Er..no idea o.O
[19-Aug-2009 09:58:06] <twm1010> how many kVa ?
[19-Aug-2009 09:58:11] <rmatte> straterra: it's very straightforward...
[19-Aug-2009 09:58:20] <straterra> twm1010: I dunno I'll have to look it up
[19-Aug-2009 09:58:22] <rmatte> create a new device organizer, or use an existing one
[19-Aug-2009 09:58:23] <straterra> We just got it
[19-Aug-2009 09:58:25] <rmatte> go in to the organizer
[19-Aug-2009 09:58:27] <twm1010> k
[19-Aug-2009 09:58:44] <twm1010> i am monitoring tripp-lites, APC's, and chlorites
[19-Aug-2009 09:58:46] <rmatte> /Power/UPS will probably be what you want
[19-Aug-2009 09:58:59] <twm1010> the tripp-lites diverted from the RFC standard OID's, the others were pretty close
[19-Aug-2009 09:59:02] <rmatte> then maybe add a sub-organizer for liebert
[19-Aug-2009 09:59:14] <rmatte> then go to the organizer where you plan to put the devices
[19-Aug-2009 09:59:19] <rmatte> and click on the "Templates" tab
[19-Aug-2009 09:59:33] <twm1010> i'm monitoring input/output voltage, output load, and some environment sensors
[19-Aug-2009 10:00:15] <rmatte> you can either copy the "Device" template down from /Devices or create a new template called whatever
[19-Aug-2009 10:00:34] <straterra> Then I just go about adding the OIDs, right?
[19-Aug-2009 10:00:36] <rmatte> then click on the dropdown arrow and select Bind Templates
[19-Aug-2009 10:00:42] <rmatte> highlight the template you created and click ok
[19-Aug-2009 10:00:46] <rmatte> then click on the template name
[19-Aug-2009 10:00:59] <rmatte> add snmp type data sources
[19-Aug-2009 10:01:05] <rmatte> it's pretty much self explanatory from that point
[19-Aug-2009 10:01:30] <rmatte> you can then create thresholds and graphs from the data points
[19-Aug-2009 10:01:37] <straterra> So like..if a line from snmpwalk is SNMPv2-SMI::mib-2.33.1.1.1.0 = STRING: "Blablabla", 2.33.1.1.1.0 is the full OID, right?
[19-Aug-2009 10:01:43] <rmatte> no
[19-Aug-2009 10:01:48] <rmatte> that's not the full OID
[19-Aug-2009 10:01:50] <twm1010> that's the RFC standard MIB.
[19-Aug-2009 10:01:51] <straterra> no?
[19-Aug-2009 10:02:06] <rmatte> the "SNMPv2-SMI::mib-" part is masking part of it
[19-Aug-2009 10:02:15] <rmatte> since your standard RFC Mibs are translating that part
[19-Aug-2009 10:02:28] <rmatte> there's some flag for snmpwalk to keep it from translating the OIDs
[19-Aug-2009 10:02:56] <twm1010> 1.3.6.1.2.1.33.1.3.3.1.3.1 is input voltage
[19-Aug-2009 10:03:19] <straterra> Yeah
[19-Aug-2009 10:03:20] <twm1010> (for single phase)
[19-Aug-2009 10:03:30] <straterra> Thats close to what I have written down
[19-Aug-2009 10:04:08] <straterra> I have mib-2.33.1.3.3.1.3.1 for volts input
[19-Aug-2009 10:05:12] <rmatte> you need to add -0n to your snmpwalk options
[19-Aug-2009 10:05:18] <rmatte> that'll display OIDs numerically
[19-Aug-2009 10:05:29] <straterra> aha
[19-Aug-2009 10:05:57] <straterra> twm1010: So..you're telling me that these OID's are standard?
[19-Aug-2009 10:06:31] <rmatte> no, just the first part of them
[19-Aug-2009 10:06:54] <straterra> well..when I did that -On switch, I get this for the input volts..
[19-Aug-2009 10:07:01] <rmatte> the first part specifies the version of SNMP, and some other standard stuff
[19-Aug-2009 10:07:06] <rmatte> the second part is the non-standard
[19-Aug-2009 10:07:16] <straterra> .1.3.6.1.2.1.33.1.3.3.1.3.1 = INTEGER: 246
[19-Aug-2009 10:07:29] <twm1010> that would be 246volts
[19-Aug-2009 10:07:32] <straterra> Right
[19-Aug-2009 10:07:33] <rmatte> yeh, that's the raw data for the input volts
[19-Aug-2009 10:07:50] <straterra> But my question is..if there OID are standard for UPS's..is there an existing template/zenpack?
[19-Aug-2009 10:08:13] <rmatte> I doubt they use the same OIDs between different companies
[19-Aug-2009 10:08:20] <twm1010> No, I don't think so, some manufacturers diverge from these, some don't, also depends on if it's single-phase or three-phase
[19-Aug-2009 10:08:28] <straterra> Ok
[19-Aug-2009 10:08:33] <rmatte> you're better off to taylor it to your needs
[19-Aug-2009 10:09:02] <twm1010> For our ginormous UPS at HQ I use manufacturer specific stuff
[19-Aug-2009 10:09:11] <rmatte> Zenoss is great, but be prepared to spend some development time hehe
[19-Aug-2009 10:09:35] <straterra> That's fine
[19-Aug-2009 10:10:01] <twm1010> rmatte: Yes it is, if only I could get some enthusiasm from my team. You'd think a group of Linux geeks/scripters would be gung-ho, but they're not, they like their Hobbit monitor..
[19-Aug-2009 10:10:03] <straterra> So..I go to add a data source..say I want to add the input volts..
[19-Aug-2009 10:10:33] <straterra> I'd put in .1.3.6.1.2.1.33.1.3.3.1.3.1 and select snmp as the type?
[19-Aug-2009 10:10:45] <twm1010> no, give it a logical name, like input volts
[19-Aug-2009 10:10:53] <straterra> Oh..doh
[19-Aug-2009 10:11:00] <twm1010> i do it ala brand.... so tl_inputvolts for tripp-lite input volts
[19-Aug-2009 10:11:09] <straterra> then .1.3.6.1.2.1.33.1.3.3.1.3.1 as the OID
[19-Aug-2009 10:11:18] <twm1010> then on the next screen you put in the OID.
[19-Aug-2009 10:11:22] <rmatte> I'm pretty sure Zenoss can easily trump that hobbit system
[19-Aug-2009 10:11:54] <rmatte> try to keep spaces out of the datapoint names for good measure
[19-Aug-2009 10:12:02] <rmatte> call it inputVolts or something
[19-Aug-2009 10:12:03] <twm1010> rmatte: oh yes, it can woop it three ways from sunday, but they hate change... it took me a WEEK to get their apache mod_status URL out of them just so i could demonstrate that the Apache Zenpack was every bit if not more powerful.
[19-Aug-2009 10:12:20] <straterra> Gauge type will be ok for this kind of stuff, right?
[19-Aug-2009 10:12:21] <rmatte> twm1010: haha, nice
[19-Aug-2009 10:12:34] <twm1010> Right... you're wanting the values, not rate of change or anything like that
[19-Aug-2009 10:12:40] <rmatte> yes, gauge is usually what you'll be using
[19-Aug-2009 10:12:47] <rmatte> the other types are for very special circumstances
[19-Aug-2009 10:12:51] <rmatte> I've yet to use any of them
[19-Aug-2009 10:14:12] <rmatte> eugh, I'm waiting for an 11gig transfer to finish between our lab nfs share and our SAN in a remote data center
[19-Aug-2009 10:14:17] <rmatte> it's taking forevar
[19-Aug-2009 10:15:20] <straterra> So like.. there is a string for the name and model and stuff too..where do I put that?
[19-Aug-2009 10:15:38] <rmatte> where are you seeing this?
[19-Aug-2009 10:15:47] <rmatte> oh, an snmp string you mean?
[19-Aug-2009 10:15:48] <twm1010> well, it sounds like you did a walk of the whole device
[19-Aug-2009 10:15:59] <rmatte> that would go under "Products" section, not in the template
[19-Aug-2009 10:16:08] <straterra> Yeah
[19-Aug-2009 10:16:10] <straterra> twm1010: I did
[19-Aug-2009 10:16:13] <rmatte> I wouldn't load every single OID from the walk in to the template
[19-Aug-2009 10:16:18] <straterra> Right
[19-Aug-2009 10:16:19] <rmatte> only load stuff you know you'll want to monitor
[19-Aug-2009 10:17:08] <straterra> lunch time..ill inquire about the products section when I get back
[19-Aug-2009 10:17:36] <twm1010> im leaving shortly myself.
[19-Aug-2009 10:17:42] <rmatte> the products section is pretty straightforward
[19-Aug-2009 10:18:47] <DogWater> Where do you specify the username/password Zenoss uses to get the event log on a windows host?
[19-Aug-2009 10:19:01] <rmatte> zProperties, zWinUser zWinPass
[19-Aug-2009 10:19:16] <DogWater> righto
[19-Aug-2009 10:19:20] <rmatte>
[19-Aug-2009 10:20:00] <rmatte> If it's a local admin account you may have to put ./ in front of the username
[19-Aug-2009 10:20:14] <rmatte> if it's a domain account you need to put the domain then a slash either way, then the username
[19-Aug-2009 10:20:19] <twm1010> DogWater: if you have a common account, do it as the zProperties of the class, and not the device
[19-Aug-2009 10:20:28] <rmatte> yeh
[19-Aug-2009 10:20:32] <rmatte> good point
[19-Aug-2009 10:20:42] <DogWater> Is it a good idea to make a specific domain user for Zenoss?
[19-Aug-2009 10:20:51] <DogWater> or do you need a domain admin account?
[19-Aug-2009 10:21:08] <rmatte> the account needs to be full admin, so if it's a domain account it has to be domain admin
[19-Aug-2009 10:21:20] * ckrough shudders
[19-Aug-2009 10:21:23] <rmatte> well, it needs to have admin on all boxes
[19-Aug-2009 10:21:31] <rmatte> not necessarily full admin access to AD
[19-Aug-2009 10:21:52] <rmatte> so it doesn't have to be a "true" domain admin account
[19-Aug-2009 10:22:26] <DogWater> So if I create a user called Zenoss and make it a local admin on the windows boxes that would work?
[19-Aug-2009 10:22:37] <rmatte> just make sure that you limit ssh access to the Zenoss box since anyone with root access will be able to dump that username and password if they know how
[19-Aug-2009 10:22:45] <rmatte> DogWater: correct
[19-Aug-2009 10:22:57] <DogWater> it doesn't store it in any sort of hash/encryption?
[19-Aug-2009 10:23:10] <rmatte> nope, I've complained about that before
[19-Aug-2009 10:23:15] <DogWater> neat-o
[19-Aug-2009 10:23:16] <DogWater> ...
[19-Aug-2009 10:24:11] <twm1010> Yes, and if one was to gain access to zenoss and create a simple command using zWinUser and zWinPassword and monitor the output, there they are.
[19-Aug-2009 10:24:16] <rmatte> dumping the password is as simple as becoming Zenoss user, going in to zendmd, and doing something like...
[19-Aug-2009 10:24:23] <rmatte> >>> mydev = find("172.30.30.8")
[19-Aug-2009 10:24:23] <rmatte> >>> mydev.zWinPassword
[19-Aug-2009 10:24:27] <rmatte> and it'll spit it out
[19-Aug-2009 10:24:28] <twm1010> or does it *** them now?
[19-Aug-2009 10:26:01] <rmatte> You can also open $ZENHOME/var/Data.fs and search for "Password" and you'll find it
[19-Aug-2009 10:26:18] <rmatte> takes a while to open the file since it's huge
[19-Aug-2009 10:27:14] <rmatte> but either way, just keep the server as secure as possible and you should be fine
[19-Aug-2009 10:27:23] <rmatte> they can't really do anything via the web interface
[19-Aug-2009 10:31:59] <cgibbons> You can configure a user that doesn't have admin rights for remote monitoring, but it's rather involved.
[19-Aug-2009 10:32:17] <cgibbons> If you're running Server 2008 Microsoft has created a new local group that has the right access rights to make it easier, too.
[19-Aug-2009 10:32:39] x-ip is now known as xip|comiendo
[19-Aug-2009 10:32:58] <rmatte> cgibbons: We've done that before and still ran in to problems
[19-Aug-2009 10:33:16] <cgibbons> it's easy to miss stuff, very non-trivial
[19-Aug-2009 10:33:45] <rmatte> well, we had an MCSE try it, and he couldn't get it to work 100%
[19-Aug-2009 10:33:52] <rmatte> lol
[19-Aug-2009 10:35:10] <cgibbons> we've got it documented in the knowledge base but there's still a lot of edge cases, especially if you're trying to monitor a domain controller isself
[19-Aug-2009 10:35:53] <cgibbons> I never went through setting it up via domain group policy, either, which is really the right solution
[19-Aug-2009 10:36:05] <rmatte> yeh
[19-Aug-2009 10:40:06] <rmatte> hmmm, I'm getting pretty frequent heartbeat failures on one of our zenoss boxes, but it doesn't seem to affect collection
[19-Aug-2009 10:45:23] <DogWater> is there some manual way to test from the Zenoss box to see if the domain credentials provided should work? Its kind of odd, I assigned a winuser/password to the class, and 3 out of 4 of the machines work but one doesnt.
[19-Aug-2009 10:47:03] <rmatte> I used to use a perl script that came with one of James Dastrup's zenpacks for testing, but those ZenPacks are obsolete now
[19-Aug-2009 10:47:29] <rmatte> If it's not working and has the exact same credentials as the other servers, then the problem is on the server side
[19-Aug-2009 10:49:24] <rmatte> cgibbons: Any news on when 2.4.3 is due for release?
[19-Aug-2009 10:49:33] <cgibbons> real soon now
[19-Aug-2009 10:49:39] <jb> yawn
[19-Aug-2009 10:49:40] <mrayzenoss> rmatte: tentatively scheduled for next week
[19-Aug-2009 10:50:12] <rmatte> ok, I have an upgrade change on Tuesday, so if it's out by then I'll be ecstatic, if not, then I'll hold off on the update until the week after
[19-Aug-2009 10:50:28] <cgibbons> dogwater: what I do to test connectivity for Windows is this: wmic -U 'domain\account' //host 'select name from win32_computersystem'
[19-Aug-2009 10:51:08] xip|comiendo is now known as x-ip
[19-Aug-2009 10:51:37] <DogWater> hm, that gives invalid global switch ;D
[19-Aug-2009 10:53:06] <cgibbons> from the zenoss server? got the right case & quotes and everything?
[19-Aug-2009 10:53:47] <DogWater> Zenoss is running on linux in my environment, i'm trying to do that from the box itself that isn't working just to see what it says.
[19-Aug-2009 10:53:55] <jb> can you guys reccomend any python beginner books?
[19-Aug-2009 10:54:22] <rmatte> jb: you might have to do something like $ZENHOME/bin/wmic
[19-Aug-2009 10:54:33] <jb>
[19-Aug-2009 10:55:06] <rmatte> sorry...
[19-Aug-2009 10:55:27] <rmatte> /usr/local/zenoss/common/bin/wmic
[19-Aug-2009 10:55:32] <rmatte> that's where mine is
[19-Aug-2009 10:56:23] <DogWater> is it supposed to return a prompt for a password or do I need to specify that on the cmmd line?
[19-Aug-2009 10:56:58] <rmatte> specify it as cgibbons stated
[19-Aug-2009 10:57:07] <rmatte> wmic -U 'domain\account' //host 'select name from win32_computersystem'
[19-Aug-2009 10:57:25] <rmatte> I think you need to put a : after account and put the password
[19-Aug-2009 10:57:31] <cgibbons> it'll prompt you for the password
[19-Aug-2009 10:57:38] <rmatte> ah
[19-Aug-2009 10:57:40] <DogWater> ah, it just says: NTSTATUS: NT_STATUS_ACCESS_DENIED - Access denied
[19-Aug-2009 10:57:42] <cgibbons> if you want to include it in the command line it would be 'domain\user%password'
[19-Aug-2009 10:57:56] <cgibbons> then there you go, bad username, bad password, or account locked out
[19-Aug-2009 10:58:07] <jb> TypeError: 'int' object is not callable
[19-Aug-2009 10:58:07] <DogWater> well i'm using the domain administrator to test..
[19-Aug-2009 10:58:07] <jb> hm
[19-Aug-2009 10:58:11] <DogWater> ;-)
[19-Aug-2009 10:58:28] <rmatte> then something is wrong on that server's config
[19-Aug-2009 10:58:35] <mrayzenoss> jb: I like http://oreilly.com/catalog/9780596513986/
[19-Aug-2009 10:58:49] <jb> thats the one I had my eye on..
[19-Aug-2009 10:59:06] <rmatte> jb: that's the book sitting right next to me right now, it's quite good
[19-Aug-2009 10:59:12] <jb> ok nice
[19-Aug-2009 10:59:18] <jb> i'm going to pick it up..
[19-Aug-2009 10:59:36] <mrayzenoss> next version comes out Oct. 2
[19-Aug-2009 10:59:53] <jb> hmm
[19-Aug-2009 11:00:07] <rmatte> just buy the third edition jb, I doubt there'll be a huge difference
[19-Aug-2009 11:00:08] <rmatte> hehe
[19-Aug-2009 11:00:20] <jb> probably not
[19-Aug-2009 11:00:22] <jb> python 3 maybe?
[19-Aug-2009 11:00:30] <mrayzenoss> Zenoss is supposed to move to Python 2.6 in the April release, which will still be the version covered by the 3rd edition
[19-Aug-2009 11:00:39] <jb> right
[19-Aug-2009 11:01:38] <DogWater> yeah, something goofy with this machine.
[19-Aug-2009 11:04:04] <straterra> Ok..back from lunch
[19-Aug-2009 11:04:45] <straterra> rmatte: So..where would I add the OID for like..device name and such?
[19-Aug-2009 11:06:41] <DogWater> hrmmm.
[19-Aug-2009 11:07:26] <cgibbons> can you remote desktop into that machine with those credentials, dogwater?
[19-Aug-2009 11:09:16] <rmatte> ok, so if I'm getting single heartbeat failures every 15 minutes or so, any thoughts on what I could adjust to make this stop?
[19-Aug-2009 11:09:25] <rmatte> I already have 3 zenhub workers on that box
[19-Aug-2009 11:10:05] <rmatte> oh wait, I only have 2, I'll try bumping that up during our next change window
[19-Aug-2009 11:11:37] <straterra> I want to like..fill out the OS information for this new template/device
[19-Aug-2009 11:12:44] <rmatte> If you want it populated automatically when modelled you need to add entries in the Products section
[19-Aug-2009 11:13:00] <rmatte> If you want to manually select it each time, you still need to add entries in to the Products section
[19-Aug-2009 11:13:04] <straterra> Erhm..wheres that?
[19-Aug-2009 11:13:15] <rmatte> right in the left hand meny
[19-Aug-2009 11:13:16] <rmatte> menu*
[19-Aug-2009 11:13:24] <straterra> Oh..duh
[19-Aug-2009 11:13:28] <rmatte>
[19-Aug-2009 11:13:42] <rmatte> lol
[19-Aug-2009 11:14:25] <straterra> So, I add liebert, then an actual product
[19-Aug-2009 11:15:21] <straterra> So, what do I add to make it pick that up upon modeling?
[19-Aug-2009 11:16:14] <DogWater> cgibbons: yep
[19-Aug-2009 11:16:27] <DogWater> cgibbons: I also checked the dcom settings
[19-Aug-2009 11:20:29] <DogWater> cgibbons: I can also connect to the remote host with compmgmt.msc via my vista desktop
[19-Aug-2009 11:20:36] <DogWater> using my account
[19-Aug-2009 11:20:37] <cgibbons> interesting
[19-Aug-2009 11:20:59] <cgibbons> can you add -d99 argument to the start of that wmic command-line ? that'll dump the protocol communications and we can see where it's actually failing and it might provide more clue
[19-Aug-2009 11:22:07] <DogWater> rpc fault: DCERPC_FAULT_ACCESS_DENIED
[19-Aug-2009 11:22:31] <cgibbons> can you paste the whole output to pastebin or e-mail for me?
[19-Aug-2009 11:22:40] <straterra> I'm looking at the product stuff for IBM..and thats very confusing
[19-Aug-2009 11:22:56] <DogWater> sure im me your email
[19-Aug-2009 11:22:59] <straterra> I put the OID of the manufacturer as the hadward name..and the product key?
[19-Aug-2009 11:25:37] <straterra> Is there a guide on this?
[19-Aug-2009 11:26:11] <rmatte> straterra: you need to put the appropriate OIDs for that model
[19-Aug-2009 11:26:13] <cgibbons> that's the earliest place in the protocol communication that you can get an error, so we're likely looking at DCOM settings. you said you checked that though?
[19-Aug-2009 11:26:35] <rmatte> straterra: there are guides to that and other things in the Zenoss Admin guide
[19-Aug-2009 11:26:40] <straterra> kk
[19-Aug-2009 11:26:43] <rmatte> I'd really recommend that you read through it once
[19-Aug-2009 11:27:03] <rmatte> You'll learn a lot from it
[19-Aug-2009 11:27:24] <straterra> I've glanced over it..but not a full read
[19-Aug-2009 11:27:41] <straterra> Erhm..I'm not finding the section I'm looking for..
[19-Aug-2009 11:27:47] * straterra goes to take another look
[19-Aug-2009 11:28:24] <DogWater> cgibbons: I just made sure the dcom settings on the broken one matched a working one.
[19-Aug-2009 11:29:01] <cgibbons> another thing you can try is to use wbemtest from another Windows box in that domain to the one that's not working; i'd expect that to fail, too.
[19-Aug-2009 11:29:29] <cgibbons> what kind of box is the one that isn't working?
[19-Aug-2009 11:29:40] <straterra> rmatte: do you know what section, per chance?
[19-Aug-2009 11:29:45] <DogWater> 2k3 x64 domain controller
[19-Aug-2009 11:30:41] <rmatte> straterra: not sure, ctrl-f
[19-Aug-2009 11:30:46] <straterra> I did that
[19-Aug-2009 11:30:59] <cgibbons> ooh, it's a domain controller, are any of the other working boxes DCs, too?
[19-Aug-2009 11:31:01] <straterra> I see nothing about it in the guide
[19-Aug-2009 11:31:12] <DogWater> yeah, one of them is, its 2k8 though
[19-Aug-2009 11:31:41] <cgibbons> okay, DCs throw some more complexity into the mix (more rules). the account is a domain admin too?
[19-Aug-2009 11:31:51] <DogWater> yeah, domain\administrator
[19-Aug-2009 11:31:54] <rmatte> straterra: just go to the Products section and look at some of the products already there
[19-Aug-2009 11:32:00] <straterra> I did that
[19-Aug-2009 11:32:00] <rmatte> straterra: check the Cisco one
[19-Aug-2009 11:32:07] <rmatte> straterra: it's not overcomplicated
[19-Aug-2009 11:32:35] <rmatte> you said you had a product key that corresponds to the model
[19-Aug-2009 11:32:40] <straterra> Ok..so..the Product Key is the OID that holds the string, right?
[19-Aug-2009 11:32:52] <rmatte> so create an organize for lietberg and then add the model number under that container with the associated product OID
[19-Aug-2009 11:32:55] <rmatte> and then you're done
[19-Aug-2009 11:33:16] <rmatte> I'm not sure what the OID would be, you'll have to hunt it down
[19-Aug-2009 11:33:23] <cgibbons> A domain admin should 'just work' unless the server has been locked down before. Hrm. Can you double-check these settings, too? http://technet.microsoft.com/en-us/library/cc787533(WS.10).aspx
[19-Aug-2009 11:33:24] <rmatte> google for it if need be
[19-Aug-2009 11:33:28] <straterra> I have the OID
[19-Aug-2009 11:33:45] <rmatte> then just add it in with the corresponding model name and you're done
[19-Aug-2009 11:33:50] <straterra> .1.3.6.1.2.1.33.1.1.1.0 = STRING: "Liebert Corporation", .1.3.6.1.2.1.33.1.1.2.0 = STRING: "Nfinity Power System", .1.3.6.1.2.1.33.1.1.3.0 = STRING: "V5007", .1.3.6.1.2.1.33.1.1.4.0 = STRING: "3.000.0"
[19-Aug-2009 11:33:50] <DogWater> i'm thinking i may have run that security wizard thing on the server in the past, its an old box
[19-Aug-2009 11:34:02] <rmatte> v5007 is what you're looking for
[19-Aug-2009 11:34:07] <cgibbons> probably
[19-Aug-2009 11:34:32] <straterra> So.. I put in that OID and the string V5007
[19-Aug-2009 11:34:38] <rmatte> add a lietbert container, then under that container add V5007 with .1.3.6.1.2.1.33.1.1.3.0
[19-Aug-2009 11:34:51] <straterra> Right
[19-Aug-2009 11:35:00] <rmatte> you name the object V5007 and specify that OID for it
[19-Aug-2009 11:35:03] <rmatte> and then you're done
[19-Aug-2009 11:35:26] <rmatte> when you model and Zenoss finds that OID it should automatically assign the make as Lietbert and the model as V5007
[19-Aug-2009 11:35:38] <rmatte> If you have other models you can add the OIDs for those too
[19-Aug-2009 11:35:51] <rmatte> if it's doesn't automatically assign the model info, you can just do it manually from the edit tab for the device
[19-Aug-2009 11:36:00] <straterra> Alright..let me give it a shot
[19-Aug-2009 11:36:01] <straterra> Thanks
[19-Aug-2009 11:36:14] <rmatte> no prob
[19-Aug-2009 11:38:13] <DogWater> cgibbons: the 'root' namespace has the same security configuration as the other 3 machines
[19-Aug-2009 11:38:17] <straterra> Didn't model it correctly..but oh well I guess
[19-Aug-2009 11:38:34] <straterra> it shows the hw make as unknown and hw model as
[19-Aug-2009 11:38:34] <straterra> it shows the hw make as unknown and hw model as .1.3.6.1.4.1.476.1.42
[19-Aug-2009 11:38:41] <straterra> Stupid irssi..sorry
[19-Aug-2009 11:39:14] <rmatte> I guess .1.3.6.1.4.1.476.1.42 is the string it's looking for then
[19-Aug-2009 11:39:19] <rmatte> add that in to products instead
[19-Aug-2009 11:39:31] <rmatte> also, remove the .1.3.6.1.4.1.476.1.42 product that it will have automatically added
[19-Aug-2009 11:40:04] <rmatte> it'll be under Unknown
[19-Aug-2009 11:40:35] <straterra> It got that value from .1.3.6.1.2.1.1.2.0 for some reason
[19-Aug-2009 11:40:37] <straterra> Oh well
[19-Aug-2009 11:40:41] <rmatte> and if that doesn't work then just manually assign it
[19-Aug-2009 11:40:53] <rmatte> what version of Zenoss are you using?
[19-Aug-2009 11:41:06] <jb> cgibbons: were you ever able to hunt anything down about getProductLink?
[19-Aug-2009 11:41:19] <straterra> Let me check
[19-Aug-2009 11:41:32] <straterra> 2.4.2
[19-Aug-2009 11:41:44] <rmatte> k, it should be able to auto-assign make/model then
[19-Aug-2009 11:41:54] <rmatte> there was a bug where it wouldn't do it properly, but that was in older versions
[19-Aug-2009 11:42:57] <straterra> Now to wait and see if my input voltage gets graphed properly
[19-Aug-2009 11:43:11] <rmatte> it should be graphing by now
[19-Aug-2009 11:43:14] <cgibbons> nope sorry
[19-Aug-2009 11:43:17] <rmatte> it's been well over 3 collection cycles
[19-Aug-2009 11:43:24] <jb> k
[19-Aug-2009 11:43:34] <straterra> I have nan in the graph
[19-Aug-2009 11:43:50] <rmatte> when did you finish the template?
[19-Aug-2009 11:44:03] <straterra> before I left for lunch
[19-Aug-2009 11:44:10] <straterra> I just added the device a couple minutes ago
[19-Aug-2009 11:44:13] <rmatte> then if it's not graphing by now there's a problem
[19-Aug-2009 11:44:21] <rmatte> click on the data source and run a test on a UPS device
[19-Aug-2009 11:44:29] <rmatte> input the UPS device name in and click test
[19-Aug-2009 11:44:34] <rmatte> see what you get
[19-Aug-2009 11:44:58] <otakup0p1> is it possible to manually edit the software for a device
[19-Aug-2009 11:45:05] <otakup0p1> or is that only only only for windows wmi voodoo
[19-Aug-2009 11:45:25] <rmatte> otakup0p1: the software list in Zenoss you mean?
[19-Aug-2009 11:45:30] <otakup0p1> i think
[19-Aug-2009 11:45:30] <otakup0p1> ack
[19-Aug-2009 11:45:33] otakup0p1 is now known as otakup0pe
[19-Aug-2009 11:45:38] <twm1010> i thought the software list came from SNMP actually
[19-Aug-2009 11:45:39] <straterra> SNMPv2-SMI::mib-2.33.1.3.3.1.3.1 = INTEGER: 245
[19-Aug-2009 11:45:44] <otakup0pe> twm1010: oh !
[19-Aug-2009 11:45:51] <straterra> Thats the result of the test..so..it works
[19-Aug-2009 11:45:56] <rmatte> straterra: hmmm, give it a bit more time then, but it really should be working by now
[19-Aug-2009 11:46:44] <straterra> Any way to force collection?
[19-Aug-2009 11:47:04] <rmatte> from commandline as zenoss user: zenperfsnmp run --cycle
[19-Aug-2009 11:47:46] <rmatte> actually
[19-Aug-2009 11:47:48] <straterra> So..if I run that a few times real fast..
[19-Aug-2009 11:47:51] <rmatte> from commandline as zenoss user: zenperfsnmp run -d <devicename>
[19-Aug-2009 11:47:53] <rmatte> try that
[19-Aug-2009 11:48:02] <straterra> i already ran the other one lol
[19-Aug-2009 11:48:09] <straterra> I'll wait for it to finish
[19-Aug-2009 11:48:19] <rmatte> well ctrl-c out of it, the cycle one just keeps cycling
[19-Aug-2009 11:48:22] <straterra> graph is populated now \O/
[19-Aug-2009 11:48:28] <rmatte> good
[19-Aug-2009 11:48:35] <straterra> Kickass..thanks
[19-Aug-2009 11:48:36] <rmatte> then ctrl-c out of it and leave it be
[19-Aug-2009 11:48:45] <straterra> yup
[19-Aug-2009 11:48:46] <rmatte> do a zenoss status to make sure all your Zenoss daemons are running
[19-Aug-2009 11:48:52] <rmatte> "zenoss status"
[19-Aug-2009 11:49:02] <straterra> yup
[19-Aug-2009 11:49:22] <rmatte> good
[19-Aug-2009 11:49:36] <rmatte> congrats, you've just created your first device template
[19-Aug-2009 11:50:33] <rmatte> it usually takes 3 polling cycles, so about 15-20 minutes for a graph to populate with data
[19-Aug-2009 11:50:41] <straterra> Gotcha
[19-Aug-2009 11:50:52] <straterra> Now to add all of these data points and graphs
[19-Aug-2009 11:51:42] <rmatte> hehe
[19-Aug-2009 11:51:50] <straterra> and maybe save it as a zenpack when I'm done
[19-Aug-2009 11:51:57] <rmatte> yeh, you could do that
[19-Aug-2009 11:52:14] <rmatte> hmmm, time for me to grab something to eat, afk for a few
[19-Aug-2009 11:52:18] <straterra> Is there anyway to do some simple math on the return stuff?
[19-Aug-2009 11:52:21] <straterra> Enjoy
[19-Aug-2009 11:52:38] <straterra> I need to divide/multiply the output by 10
[19-Aug-2009 11:56:39] <jb> you guys know if there is a routine to set "OS Make" like there is for OS Versino (setOSProductKey)?
[19-Aug-2009 11:59:27] <rmatte> jb, I believe the Make gets set to the container name that the version info is in in Products
[19-Aug-2009 11:59:35] <rmatte> same as product make/model
[19-Aug-2009 12:00:11] <rmatte> so you'd make a container called Microsoft for instance and then put a version item in it with the correct OID and name of the version
[19-Aug-2009 12:00:44] <jb> ah ha
[19-Aug-2009 12:00:54] <jb> thanks
[19-Aug-2009 12:00:57] <rmatte> np
[19-Aug-2009 12:02:46] <scp2004> Hi Guys ! .. Can someone help me with a question ?
[19-Aug-2009 12:02:57] <rmatte> scp2004: just ask and we'll help as best we can
[19-Aug-2009 12:03:13] <rmatte> (asking to ask is kind of counter-productive)
[19-Aug-2009 12:03:46] <scp2004> [rmatte] Thanks. .. I need a report in zenoss that give me the Name of device and Ip ...
[19-Aug-2009 12:03:57] <rmatte> that's it?
[19-Aug-2009 12:04:01] <rmatte> just name and IP?
[19-Aug-2009 12:04:09] <scp2004> yep
[19-Aug-2009 12:04:14] <rmatte> Easy...
[19-Aug-2009 12:04:16] <rmatte> click on Reports
[19-Aug-2009 12:04:21] <rmatte> then click on Custom Device Reports
[19-Aug-2009 12:04:35] <rmatte> go to the dropdown and select Add Device Report
[19-Aug-2009 12:04:43] <rmatte> name it something like "List of Devices" or whatever
[19-Aug-2009 12:04:45] <rmatte> click on it
[19-Aug-2009 12:04:56] <rmatte> click the Edit tab
[19-Aug-2009 12:05:06] <scp2004> done
[19-Aug-2009 12:05:10] <scp2004> and..
[19-Aug-2009 12:05:20] <rmatte> Under Sort Column put: getId
[19-Aug-2009 12:05:26] <rmatte> Under Columns put:
[19-Aug-2009 12:05:27] <rmatte> getId
[19-Aug-2009 12:05:27] <rmatte> getManageIp
[19-Aug-2009 12:05:40] <rmatte> Under Column Names put:
[19-Aug-2009 12:05:41] <rmatte> Device
[19-Aug-2009 12:05:41] <rmatte> IP Address
[19-Aug-2009 12:05:49] <rmatte> Sort Sense should be: asc
[19-Aug-2009 12:05:55] <rmatte> Path should be /
[19-Aug-2009 12:05:58] <rmatte> then click Save
[19-Aug-2009 12:06:00] <rmatte> and you're done
[19-Aug-2009 12:06:57] <rmatte> then just click the Report tab and it will generate the report
[19-Aug-2009 12:07:07] <scp2004> doing ..
[19-Aug-2009 12:07:36] <scp2004> [rmatte] Ooooow ..Very Cool ...Thank you man ..
[19-Aug-2009 12:07:40] <rmatte> no problem
[19-Aug-2009 12:08:13] <rmatte> there are other things you can add to columns too, like:
[19-Aug-2009 12:08:14] <rmatte> getHWManufacturerName
[19-Aug-2009 12:08:14] <rmatte> getHWProductName
[19-Aug-2009 12:08:14] <rmatte> getHWSerialNumber
[19-Aug-2009 12:08:14] <rmatte> getPingStatusString
[19-Aug-2009 12:08:14] <rmatte> getSnmpStatusString
[19-Aug-2009 12:08:18] <rmatte> getPriorityString
[19-Aug-2009 12:08:26] <twm1010> rmatte: are those documented?
[19-Aug-2009 12:08:37] <rmatte> yeh, somewhere, lol
[19-Aug-2009 12:08:58] <rmatte> I have this bookmarked: http://www.zenoss.com/community/docs/zenoss-api-docs/2.1/ZenModel.Device.Device-class.html
[19-Aug-2009 12:09:02] <rmatte> some of them are listed in there
[19-Aug-2009 12:09:35] <scp2004> [rmatte] Maybe you can help me with this ...
[19-Aug-2009 12:09:35] <rmatte> There are also:
[19-Aug-2009 12:09:36] <rmatte> getLocationName
[19-Aug-2009 12:09:36] <rmatte> getCreatedTimeString
[19-Aug-2009 12:09:36] <rmatte> getProdState
[19-Aug-2009 12:09:45] <rmatte> save the list I just pasted to a text file somewhere for future use
[19-Aug-2009 12:10:00] <twm1010> can you do multiple sorts?
[19-Aug-2009 12:10:25] <scp2004> [rmatte] Look .. I need this report . Cause my zentrap doesn't run on FreeBsd , so I m using Centos now
[19-Aug-2009 12:10:53] <rmatte> twm1010: you can sort by any column, but I don't know about actual multiple sorts
[19-Aug-2009 12:10:59] <twm1010> Are you going to use the report to import the devices to a new install?
[19-Aug-2009 12:10:59] <rmatte> twm1010: the sorting can be a bit flaky
[19-Aug-2009 12:11:14] <rmatte> scp2004: oh, there are better ways to back up devices
[19-Aug-2009 12:11:24] <twm1010> rmatte: was just curious, my solarwinds reports have multiple sorts, and filtering, which is really nice
[19-Aug-2009 12:11:29] <scp2004> [rmatte] But , i have 485 Devices listed . And when i used zenbackup / zenrestore to Bsd from Centos. Doesnt work ...
[19-Aug-2009 12:12:07] <rmatte> twm1010: yeh, we're going to probably be using solarwinds for reporting since Zenoss' reporting features are a bit weak at the moment
[19-Aug-2009 12:12:14] <rmatte> I'm still designing a bunch of custom reports for us
[19-Aug-2009 12:12:39] <rmatte> scp2004: one sec, I have to check some docs on what the proper procedure is
[19-Aug-2009 12:12:40] <scp2004> [rmatte] Cause the zenrestore , restores all confs . And its very different confs in Bsd to Centos ..
[19-Aug-2009 12:13:11] <scp2004> [rmatte] So .. I need a command to restore only devices .. Not configs ... You know ?
[19-Aug-2009 12:13:39] <rmatte> scp2004: yeh, working on it...
[19-Aug-2009 12:14:06] <twm1010> scp2004: there may be a command to do just that
[19-Aug-2009 12:14:28] <rmatte> twm1010: yeh, there's some option for zenbackup to just do devices
[19-Aug-2009 12:14:31] <rmatte> I'm seaching for it
[19-Aug-2009 12:14:34] <rmatte> searching*
[19-Aug-2009 12:15:50] <scp2004> [rmatte] Will help me a lot ..Thanks ...
[19-Aug-2009 12:16:49] <straterra> All UPS values added \O/
[19-Aug-2009 12:16:54] <straterra> and graphed..sweeeeet
[19-Aug-2009 12:17:05] <rmatte> hmmm, I thought there was but looks like maybe not
[19-Aug-2009 12:17:39] <rmatte> I'll keep looking
[19-Aug-2009 12:18:17] <straterra> Weird
[19-Aug-2009 12:18:34] <straterra> I have datapoints labeled like.. Input_Volts
[19-Aug-2009 12:18:49] <straterra> Yet..in the graph its called volts_inputvolts
[19-Aug-2009 12:19:14] <straterra> Oh..I can just rename it
[19-Aug-2009 12:19:17] <rmatte> you probably made a data source called volts, then made a datapoint below it called inputvolts?
[19-Aug-2009 12:19:25] <rmatte> so it combines them
[19-Aug-2009 12:19:31] <rmatte> datasource_datapoint
[19-Aug-2009 12:19:54] <straterra> Gotcha
[19-Aug-2009 12:20:11] <rmatte> also, I wouldn't make it a habbit of using underscore for datapoint or data source names
[19-Aug-2009 12:20:19] <rmatte> try to just do stuff like inputVolts
[19-Aug-2009 12:20:21] <rmatte> or whatever
[19-Aug-2009 12:20:40] <rmatte> quoting: "[01:12pm] <straterra> I have datapoints labeled like.. Input_Volts"
[19-Aug-2009 12:21:17] <straterra> Gotcha
[19-Aug-2009 12:22:56] <rmatte> scp2004: yeh, looks like you might have to just resort to using that report
[19-Aug-2009 12:23:02] <rmatte> but at least you'll only have to do it once
[19-Aug-2009 12:33:20] <scp2004> [rmatte] i ll have to go for now ..but if you find .PLease post here .. I ll see ..Thank you man ..
[19-Aug-2009 12:33:48] <scp2004> [rmatte] About this command to restore only devices ..
[19-Aug-2009 12:47:09] <DogWater> cgibbons: yeah so I have no clue about this, lol
[19-Aug-2009 12:59:25] <jb> hey
[19-Aug-2009 12:59:27] <jb> you core users
[19-Aug-2009 12:59:36] <jb> can core monitor windows services, etc out of the box?
[19-Aug-2009 12:59:50] <twm1010> Yes, indeed.
[19-Aug-2009 12:59:50] <jb> been a while since I have used core..
[19-Aug-2009 13:03:05] <rmatte> jb: it can monitor services via WMI, you can monitor processes via SNMP, you can monitor event log via WMI or via SNMP traps
[19-Aug-2009 13:03:06] <jb> without an agent on the windows box?
[19-Aug-2009 13:03:18] <rmatte> jb: Performance data can be collected via either, but WMI is easier to do it with
[19-Aug-2009 13:03:27] <rmatte> jb: Zenoss is 100% agentless
[19-Aug-2009 13:03:36] <jb> well, core used to use a winagent
[19-Aug-2009 13:03:40] <jb> before your time
[19-Aug-2009 13:03:40] <rmatte> Unless you consider SNMP or WMI services an agent
[19-Aug-2009 13:03:41] <jb>
[19-Aug-2009 13:03:44] <rmatte> which they technically are not
[19-Aug-2009 13:03:55] <rmatte> probably
[19-Aug-2009 13:04:04] <rmatte> I've only been using Zenoss for like 8 or 9 months now
[19-Aug-2009 13:04:09] <jb> yeah
[19-Aug-2009 13:04:20] <rmatte> I've learned it pretty inside and out though
[19-Aug-2009 13:33:28] <dranch> Anyone have some time to help with a command script that won't even throw errors in the zencommand.log file?
[19-Aug-2009 13:34:03] <rmatte> this is something in Event Manager -> Commands?
[19-Aug-2009 13:34:34] <dranch> This thread covers what I see but even putting the script in /usr/local/zenoss/libexec doesn't resolve: http://forums.zenoss.com/viewtopic.php?t=5467&postdays=0&postorder=asc&start=30
[19-Aug-2009 13:35:07] <dranch> If I do the manual test via the webui, it works as expected BUT the command won't run automatically
[19-Aug-2009 13:35:32] <rmatte> what's the command?
[19-Aug-2009 13:36:09] <dranch> It's a little bash script I wrote that will return a 1 if there is a failiure
[19-Aug-2009 13:36:16] <mrayzenoss> dranch: I ran into that awhile ago http://dev.zenoss.org/trac/ticket/5175
[19-Aug-2009 13:36:17] <rmatte> I see
[19-Aug-2009 13:37:14] <dranch> ok.. let me give it an absolute path
[19-Aug-2009 13:37:17] <dranch> hold
[19-Aug-2009 13:37:34] <rmatte> well, how do you have it right now?
[19-Aug-2009 13:37:54] <rmatte> the proper way to specify zenhome in a command based datasource is $$ZENHOME
[19-Aug-2009 13:37:57] <rmatte> not just $ZENHOME
[19-Aug-2009 13:38:13] <twm1010> rmatte: why?
[19-Aug-2009 13:38:16] <rmatte> for example, I use the following on many different zenoss boxes
[19-Aug-2009 13:38:17] <rmatte> $$ZENHOME/Products/SNMPWindowsPerfMonitor/libexec/wincpu.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"
[19-Aug-2009 13:38:20] <rmatte> and it works great
[19-Aug-2009 13:38:43] <rmatte> twm1010: I think the extra $ tells Zenoss to look for an environment variable, instead of a variable within Zope
[19-Aug-2009 13:38:50] <dranch> no path.. I just specify use SSH: no; event class /cmd/fail; severity: critical; cycle time: 60; command template: check-vip.sh database-fail
[19-Aug-2009 13:39:30] <rmatte> yeh, command template should be a full path
[19-Aug-2009 13:39:37] <rmatte> you can use $$ZENHOME to shorten it
[19-Aug-2009 13:39:52] <rmatte> when you're doing the test it's running it as though you were running it from a shell I believe, so it works
[19-Aug-2009 13:39:55] <dranch> I have other commands that work find though they DO use the SSH transport
[19-Aug-2009 13:40:23] <twm1010> rmatte: that makes sense, that would be why some of my perl scripts mess up getting to $ZENHOME.
[19-Aug-2009 13:40:32] <dranch> So I should use: $$ZENHOME/check-vip.sh database-fail ?
[19-Aug-2009 13:40:33] <rmatte> twm1010: probably
[19-Aug-2009 13:40:43] <rmatte> well, where is the check-vip.sh file stored?
[19-Aug-2009 13:40:54] <rmatte> what's the absolute path to it?
[19-Aug-2009 13:41:09] <dranch> usr/local/zenoss/libexec
[19-Aug-2009 13:41:50] <rmatte> ok, and if you do echo "$ZENHOME"
[19-Aug-2009 13:41:52] <rmatte> what do you get?
[19-Aug-2009 13:42:11] <dranch> I'm configuring this from the WebUI. let me try from the CMD line
[19-Aug-2009 13:42:15] <rmatte> k
[19-Aug-2009 13:42:33] <rmatte> zenhome paths are different depending on what installation method you used
[19-Aug-2009 13:42:36] <rmatte> that's why I'm checking
[19-Aug-2009 13:42:47] <dranch> $ echo $ZENHOME
[19-Aug-2009 13:42:49] <dranch> /usr/local/zenoss
[19-Aug-2009 13:42:55] <rmatte> ok...
[19-Aug-2009 13:43:08] <rmatte> so you'd do $$ZENHOME/libexec/check-vip.sh
[19-Aug-2009 13:43:13] <dranch> So I guess the proper setting is $$ZENHOME/libexec/check-vip.sh
[19-Aug-2009 13:43:18] <rmatte> correct
[19-Aug-2009 13:43:25] <scp2004> [rmatte] Hi man ..Do you found ? The command ? To restore only devices ?
[19-Aug-2009 13:43:39] <dranch> What I don't understand is why I wouldn't see errors in the zencommand.log file with a -v 10 setting
[19-Aug-2009 13:44:53] <mrayzenoss> that's why I opened the ticket
[19-Aug-2009 13:45:06] <mrayzenoss> it needs to have logging and the same path as the test button
[19-Aug-2009 13:45:40] <dranch> That previous URL I have continues on into the weeds but it seems it's intermittant for some people
[19-Aug-2009 13:46:27] <scp2004> [mrayzenoss] Hi man .. rmatte are trying help me .. But lets see if you know too .. I need a command like zenrestore. But to restore only devices , without configs ..Do you know ?
[19-Aug-2009 13:47:22] <mrayzenoss> zendevicedump?
[19-Aug-2009 13:48:00] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch04s02.html#d4e1761
[19-Aug-2009 13:48:33] <scp2004> i ll see ..thank you !
[19-Aug-2009 13:49:18] <dranch> Nope.. no change on my side
[19-Aug-2009 13:49:38] <mrayzenoss> dranch: try restarting zenhub and zencommand
[19-Aug-2009 13:50:03] <rmatte> scp2004: no, I don't think it's possible, unless Matt knows of a way...
[19-Aug-2009 13:50:16] <rmatte> Matt, is there some way to only export/import devices between Zenoss instances
[19-Aug-2009 13:50:22] <mrayzenoss> rmatte: that's the link http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch04s02.html#d4e1761
[19-Aug-2009 13:50:23] <rmatte> without exporting/importing anything else?
[19-Aug-2009 13:50:37] <rmatte> ah sorry, I was away from my desk for a few minutes, boss called me away
[19-Aug-2009 13:50:43] <dranch> I've restarted zencommand now with -v 10 and I'm greping for check-vip".. I'll try zenhub in a moment
[19-Aug-2009 13:50:48] <scp2004> [rmatte] I m reading this now ... Its true .. Only devices ..
[19-Aug-2009 13:50:58] <rmatte> ah right, zendevicedump
[19-Aug-2009 13:51:03] <rmatte> I totally forgot about that
[19-Aug-2009 13:51:04] <scp2004> yep ..
[19-Aug-2009 13:52:08] <dranch> Hmmmm.. still nothing showing up in the log
[19-Aug-2009 13:54:24] <dranch> As usual, the manual test works via the webui when I specify the zenoss machine as the "test against device"
[19-Aug-2009 13:54:50] <dranch> The only other thing to note is that I'm running this via a template and not a specific machine
[19-Aug-2009 13:56:10] <dranch> brb
[19-Aug-2009 13:59:14] <scp2004> [mrayzenoss] when i use zendevicedump : WARNING:ZEO.zrpc:(32152) CW: error connecting to ('localhost', 8100): EINVAL
[19-Aug-2009 14:00:20] <mrayzenoss> weird, haven't seen that
[19-Aug-2009 14:02:49] <rmatte> dranch: you do know the command needs to output in a certain format right?
[19-Aug-2009 14:03:08] <rmatte> scp2004: is Zenoss running when you're using the command?
[19-Aug-2009 14:10:39] <rmatte> man, King crab is only like 2-3 months away, can't wait
[19-Aug-2009 14:11:35] <mrayzenoss> I think the beta is supposed to start soon
[19-Aug-2009 14:11:51] <DogWater> crap, no matter what I do i can't get wmic to work on this remote host, although i've gotten a bit further.
[19-Aug-2009 14:12:06] <rmatte> I'll most likely be setting up a lab box with the beta, but I really can't wait until we can get the final version on to our Prod boxes
[19-Aug-2009 14:12:16] <rmatte> especially since it'll accurately reflect interface status on the OS tab
[19-Aug-2009 14:12:26] <rmatte> so no more confusion
[19-Aug-2009 14:18:54] <rmatte> haha
[19-Aug-2009 14:19:11] <rmatte> I went through and added a bunch of comments on the new event console demo site
[19-Aug-2009 14:19:16] <rmatte> I'm reading the responses
[19-Aug-2009 14:21:58] <mrayzenoss> supposedly Ian has been putting patches into the demo
[19-Aug-2009 14:22:06] <mrayzenoss> he's integrating it in trunk right now
[19-Aug-2009 14:22:16] <rmatte> not much really seems to have changed
[19-Aug-2009 14:22:26] <mrayzenoss> some IE bugs
[19-Aug-2009 14:22:44] <rmatte> it's still slow opening the details portion in firefox, when you click Show more details... it just closes the details window lol
[19-Aug-2009 14:23:23] <rmatte> and even with those problems it's still nicer than the current event console
[19-Aug-2009 14:23:24] <rmatte> lol
[19-Aug-2009 14:27:37] <rmatte> does the UI consultant you guys hired actually code, or just assists to plan out the layout?
[19-Aug-2009 14:27:48] <mrayzenoss> she does not code
[19-Aug-2009 14:27:53] <rmatte> ah
[19-Aug-2009 14:28:14] <rmatte> so it's more about the aesthetic aspect of it
[19-Aug-2009 14:28:17] <mrayzenoss> she's spent a lot of time on workflow analysis
[19-Aug-2009 14:28:25] <rmatte> ah
[19-Aug-2009 14:28:33] <mrayzenoss> so things won't be buried 7 clicks away
[19-Aug-2009 14:28:37] <rmatte> so it's more about efficiency
[19-Aug-2009 14:28:39] <rmatte> yeh
[19-Aug-2009 14:28:39] <jb> i havent seen updated screeshots in quite some time
[19-Aug-2009 14:28:45] <jb> wonder if alot has changed..
[19-Aug-2009 14:28:55] <rmatte> jb: http://public-demo.zenoss.com/evconsole
[19-Aug-2009 14:28:56] <mrayzenoss> jb: http://public-demo.zenoss.com/evconsole
[19-Aug-2009 14:28:57] <rmatte> seen that?
[19-Aug-2009 14:29:01] <jb> yeah i have seen that
[19-Aug-2009 14:29:07] <mrayzenoss> that's all that's public so far
[19-Aug-2009 14:29:08] <jb> im talking others
[19-Aug-2009 14:29:08] <rmatte> that's basically it
[19-Aug-2009 14:29:18] <jb>
[19-Aug-2009 14:29:27] <rmatte>
[19-Aug-2009 14:30:29] <mrayzenoss> so the Add Device wizard in 2.4 was the first of the new screens, then the Event Console in King Crab and then everything else in Stone Crab
[19-Aug-2009 14:30:32] <mrayzenoss> is the plan at least
[19-Aug-2009 14:30:41] <mrayzenoss> which is why we're hiring another developer
[19-Aug-2009 14:31:02] <jb> obviously some things will need to change in blue crab (the navigation, etc)
[19-Aug-2009 14:32:14] <mrayzenoss> yeah, Ian's putting the Event Console into Subversion trunk right now, so if you're a glutton for punishment...
[19-Aug-2009 14:32:53] <mrayzenoss> or you can wait for the beta
[19-Aug-2009 14:33:01] <mrayzenoss> which will have installers and VMs
[19-Aug-2009 14:40:25] <eidolon> hey folks - i'm enabling Microsoft SQL server monitoring via a template - but i'm not getting any actual data (it shows up int he performance details). do i need to specify a login/password in zproperties for sql server to get performance stats from it?
[19-Aug-2009 14:41:26] <jb> ok cool.. I'll figure it out
[19-Aug-2009 14:42:53] <mrayzenoss> eidolon: depends which ZenPack you're using to monitor
[19-Aug-2009 14:42:59] <magnachef> hey mray, I have a Zenoss scalability question
[19-Aug-2009 14:43:04] <eidolon> let me check.
[19-Aug-2009 14:43:40] <eidolon> actiually i'm not sure how to check
[19-Aug-2009 14:44:00] <eidolon> i see:
[19-Aug-2009 14:44:01] <eidolon> ${here/ZenPackManager/packs/ZenPacks.community.Perfmon/path}/libexec/perfmon.pl 1 "${dev/manageIp}" "${dev/zWinUser}" "${dev/zWinPassword}" '\${dev/zSQLInstance}:Access Methods\Full Scans/sec' "SQLServer_FullTableScansSec"
[19-Aug-2009 14:44:10] <magnachef> let's say that I have around 1000 devices, and I want to be able to assign a variable polling time to certain groups of servers. Some polling every 15 sec, some 30 sec, some 1 min, and some 5 min. Does this sound reasonable to do in Zenoss?
[19-Aug-2009 14:45:31] <mrayzenoss> eidolon: I think that's this one: http://www.zenoss.com/community/projects/zenpacks/sql-server-performance
[19-Aug-2009 14:45:58] <mrayzenoss> magnachef: right now polling is a function of which protocol you're using
[19-Aug-2009 14:46:17] <mrayzenoss> magnachef: so SNMP is managed at the collector level and WMI and SSH are at the device class
[19-Aug-2009 14:46:35] <eidolon> looks like it's using the zWinUser to log in via WMI, but not actually collecting any data. i'mwondering if the 'zenoss' login i'm using needs to be granted something internally on the machine
[19-Aug-2009 14:47:06] <mrayzenoss> magnachef: ping is also set on the collector
[19-Aug-2009 14:47:16] <mrayzenoss> magnachef: hmm... WMI is actually on the collector too
[19-Aug-2009 14:47:54] <mrayzenoss> eidolon: you may need to look at the Perl script and see what it's calling, then try those by hand to check your WMI privileges
[19-Aug-2009 14:48:13] <mrayzenoss> magnachef: I believe in King Crab all collector intervals will move to the device class
[19-Aug-2009 14:50:23] <chudler_> that.... WOULD BE NIIIICE!!!
[19-Aug-2009 14:50:55] <chudler_> not that it matters, if what one wants more resolution on graphs.
[19-Aug-2009 14:51:23] <magnachef> yeah, that's nice
[19-Aug-2009 14:52:02] <dranch> rmatte, you still around?
[19-Aug-2009 14:52:13] <dranch> You mentioned before: dranch: you do know the command needs to output in a certain format right?
[19-Aug-2009 14:52:20] <magnachef> I'm assuming that all the rrd magic is handled properly as well? or is that something the user will need to tweak?
[19-Aug-2009 14:53:03] <twm1010> mrayzenoss: collector intervals are set at the collector now?
[19-Aug-2009 14:53:53] <mrayzenoss> twm1010: yeah Collectors->localhost->Edit
[19-Aug-2009 14:54:48] <twm1010> mrayzenoss: so this is moving to a zproperty?
[19-Aug-2009 14:55:04] <mrayzenoss> unsure where it's going
[19-Aug-2009 14:55:39] <twm1010> Off topic: the collected RRD data, is it ever summarized?
[19-Aug-2009 14:56:06] <mrayzenoss> twm1010: yeah, the RRD Create command has the summarization schedules
[19-Aug-2009 14:56:46] <mrayzenoss> twm1010: older data gets summarized the farther out it gets
[19-Aug-2009 14:57:30] <twm1010> I see entries like: RRA:AVERAGE:0.5:1:600
[19-Aug-2009 14:57:39] <mrayzenoss> those are them
[19-Aug-2009 14:57:54] <mrayzenoss> http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[19-Aug-2009 14:57:57] <magnachef> that is unless you have enough steps created to hold enough single poll values
[19-Aug-2009 14:58:36] <magnachef> you'd need to bump up that 600 value
[19-Aug-2009 14:59:16] <magnachef> mray: have you guys been following reconnoiter?
[19-Aug-2009 14:59:59] <twm1010> So is summarization done on a routine? Or does RRD do it as new data is written?
[19-Aug-2009 15:00:13] <magnachef> as new data is written
[19-Aug-2009 15:00:38] <twm1010> ah, very good, but with a lot of datapoints, thats a lot disk constant disk i/o i'd imagine
[19-Aug-2009 15:00:51] <magnachef> so in the above example, after 600, the last value is gone from that RRA
[19-Aug-2009 15:01:17] <magnachef> s/600/600 data points/
[19-Aug-2009 15:01:27] <twm1010> so if you collect every 5 minutes
[19-Aug-2009 15:01:56] <magnachef> after you 601st collection, you will lose your first collected value
[19-Aug-2009 15:02:15] <twm1010> right, that value is summarized into an archive
[19-Aug-2009 15:02:40] <magnachef> yeah, but is probably already in the other RRAs
[19-Aug-2009 15:02:47] <mrayzenoss> right
[19-Aug-2009 15:02:52] <magnachef> if I remember correctly, it'll update all RRAs at the same time
[19-Aug-2009 15:02:58] <mrayzenoss> so there are multiple RRAs being used
[19-Aug-2009 15:03:15] <magnachef> but some RRAs will require multiple data points before making a datapoint for that RRA
[19-Aug-2009 15:03:43] <twm1010> ideally i only have to fool with it if i need to retain accuracy for a longer or shorter period
[19-Aug-2009 15:04:18] <magnachef> yeah, I try to keep 5 minute averages for about 2 years
[19-Aug-2009 15:04:24] <mrayzenoss> magnachef: I looked into Reconnoiter a little
[19-Aug-2009 15:05:12] <magnachef> I just watched a demo from oscon
[19-Aug-2009 15:05:35] <magnachef> the aggregate graph building is something I'd definately like to see in Zenoss
[19-Aug-2009 15:05:51] <mrayzenoss> I'll watch the demo, I read about it before OSCON
[19-Aug-2009 15:06:34] <magnachef> yeah, my company has done some work with the OmniTi guys, and I've talked briefly with Theo about it
[19-Aug-2009 15:07:31] <mrayzenoss> it looks a lot like BMC Patrol Express
[19-Aug-2009 15:08:07] <magnachef> I've never seen BMC Patrol before ($$$$)
[19-Aug-2009 15:08:21] <mrayzenoss> a few of us at Zenoss were on that team
[19-Aug-2009 15:08:27] <magnachef> ahh
[19-Aug-2009 15:08:33] <jb> oh wow.. i can't believe I accidentally just /quit
[19-Aug-2009 15:08:34] <jb> ... Total uptime : 552d 4h 39m 34s
[19-Aug-2009 15:08:35] <jb>
[19-Aug-2009 15:09:04] <mrayzenoss> at least architecturally
[19-Aug-2009 15:09:27] <jb> are the devs here tommorow?
[19-Aug-2009 15:09:37] <mrayzenoss> yup, mrchippy is on call
[19-Aug-2009 15:09:39] <magnachef> a big issue I'm having is that now that we're gathering a lot of data through Zenoss and Cacti, management is wanting more and more reports, which can be quite tedious in both systems (IMHO)
[19-Aug-2009 15:09:42] <jb> k
[19-Aug-2009 15:09:52] <mrayzenoss> yeah, our reports suck
[19-Aug-2009 15:10:04] <twm1010> mrayzenoss: Yeah, they do
[19-Aug-2009 15:10:09] <magnachef> yeah, so that's a HUGE thing with my company now
[19-Aug-2009 15:10:12] <mrayzenoss> some guy from BIRT said he was going to do an extractor, but I haven't heard back from him
[19-Aug-2009 15:10:38] <magnachef> ...the other issue is the variable polling stuff and allowing it to scale
[19-Aug-2009 15:11:27] <magnachef> so is there a big initative to fix the polling suckiness?
[19-Aug-2009 15:11:46] <mrayzenoss> yeah
[19-Aug-2009 15:11:50] <mrayzenoss> so the scheduling is now asynchronous in trunk, which will provide scaling improvements.
[19-Aug-2009 15:12:25] <mrayzenoss> but the work to move the configuration hasn't been scheduled yet
[19-Aug-2009 15:12:30] <mrayzenoss> we need more devs
[19-Aug-2009 15:12:35] <twm1010> mrayzenoss: to tell you the truth, the reporting is what stopped us from getting enterprise.
[19-Aug-2009 15:12:53] <magnachef> how much are you hiring devs for? :-)
[19-Aug-2009 15:13:05] <mrayzenoss> pay is good
[19-Aug-2009 15:13:26] <magnachef> can't find enough quality guys?
[19-Aug-2009 15:13:36] <mrayzenoss> just opened up the spot recently
[19-Aug-2009 15:13:54] <magnachef> I basically program in Perl...I'd need to tighten up my Python-Fu!
[19-Aug-2009 15:14:03] <mrayzenoss> http://www.zenoss.com/about/jobs#position_senior-web-developer
[19-Aug-2009 15:14:13] <twm1010> it might be easier to put some money into research to clone cluther
[19-Aug-2009 15:14:25] <magnachef> heh
[19-Aug-2009 15:16:49] <mrayzenoss> if you know anyone who would be a good fit, send them our way
[19-Aug-2009 15:17:07] <magnachef> hmm...where is the position out of
[19-Aug-2009 15:17:15] <mrayzenoss> Austin or Annapolis
[19-Aug-2009 15:18:08] <mrayzenoss> but if you're badass enough like cgibbons they'll let you work remote
[19-Aug-2009 15:18:30] <cgibbons> ugh
[19-Aug-2009 15:18:34] <mrayzenoss> heh
[19-Aug-2009 15:19:30] <magnachef> so will you guys be attending Ohio Linux Fest?
[19-Aug-2009 15:19:35] <mrayzenoss> I'll be there
[19-Aug-2009 15:19:42] <mrayzenoss> we're gonna have a free training day
[19-Aug-2009 15:19:53] <twm1010> when is that?
[19-Aug-2009 15:19:55] <magnachef> that would be awesome...I'm about 3 hrs away and was planning on going
[19-Aug-2009 15:20:02] <mrayzenoss> training will be on Friday
[19-Aug-2009 15:20:03] <magnachef> ...that would seal the deal
[19-Aug-2009 15:20:21] <mrayzenoss> condensed version of the 2-day training we offer for $
[19-Aug-2009 15:20:32] <magnachef> yeah, I'd like to get me some of that
[19-Aug-2009 15:21:08] <twm1010> hrmm, i MIGHT swing that...
[19-Aug-2009 15:22:03] <magnachef> so who do I need to pressure to make sure that happens? @mrhinkle?
[19-Aug-2009 15:22:07] <magnachef> :-)
[19-Aug-2009 15:22:10] <mrayzenoss> it's done
[19-Aug-2009 15:22:20] <mrayzenoss> I'm making my plane reservations today
[19-Aug-2009 15:22:24] <mrayzenoss> to get there on Thursday
[19-Aug-2009 15:22:25] <magnachef> oh...well that was easy!
[19-Aug-2009 15:23:35] <magnachef> so class on Friday?
[19-Aug-2009 15:23:41] <mrayzenoss> yeah
[19-Aug-2009 15:24:11] <magnachef> alright, cool. I'll keep my eyes open for the updates
[19-Aug-2009 15:24:12] <dranch> mrayzenoss: what would be the reason for data source "commands" not being logged at all?
[19-Aug-2009 15:24:50] <mrayzenoss> dranch: I'm not sure, that's why it took me so long to discover that bug
[19-Aug-2009 15:25:09] <mrayzenoss> so you're using zencommand run -v10 -d yourdevice ?
[19-Aug-2009 15:25:11] <dranch> So I'm at a loss of how to get this working. I can't see it fail and it doesn't work
[19-Aug-2009 15:25:37] <dranch> the manual webui test works fine.. nothing happens otherwise
[19-Aug-2009 15:26:33] <dranch> If I run that command, I see
[19-Aug-2009 15:26:38] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 15:26:39] <dranch> DEBUG:zen.zencommand:Finished config fetch
[19-Aug-2009 15:26:41] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 15:26:42] <dranch> INFO:zen.zencommand:Daemon zencommand shutting down
[19-Aug-2009 15:26:44] <dranch> DEBUG:zen.zencommand:removing service EventService
[19-Aug-2009 15:26:45] <dranch> DEBUG:zen.zencommand:removing service CommandConfig
[19-Aug-2009 15:26:47] <dranch> INFO:zen.zencommand:zencommand shutting down
[19-Aug-2009 15:26:50] <dranch> Seems it thinks there aren't any commands to run
[19-Aug-2009 15:26:59] <mrayzenoss> try 'zencommand stop'
[19-Aug-2009 15:27:05] <mrayzenoss> which will stop the zencommand daemon
[19-Aug-2009 15:27:11] <dranch> done
[19-Aug-2009 15:27:51] <dranch> and then what.. ./zencommand start -v 10 ?
[19-Aug-2009 15:27:56] <jb> so.. I have a object class that is not compiling. does anyone know where I can look for any debugging? zenhub.log is not showing any pertienent information
[19-Aug-2009 15:28:19] <mrayzenoss> no, just the zencommand run -v 10
[19-Aug-2009 15:28:34] <dranch> ok.. done
[19-Aug-2009 15:28:44] <mrayzenoss> anything relevant in the output?
[19-Aug-2009 15:29:58] <dranch> nope..nothing for the expected hostname or it's resolved IP address
[19-Aug-2009 15:31:28] <mrayzenoss> crap, my home machine crashed
[19-Aug-2009 15:31:44] <mrayzenoss> I was going to log in there and see what my zencommands are doing
[19-Aug-2009 15:32:02] <magnachef> hopefully not crashing your machine ;-)
[19-Aug-2009 15:32:17] <dranch> It doesn't seem to be logging any commands for this template or machine
[19-Aug-2009 15:32:17] <rmatte> too bad you don't have some UPS that you can login to and reboot power
[19-Aug-2009 15:32:50] <mrayzenoss> dranch: do you have an empty graph?
[19-Aug-2009 15:32:53] <mrayzenoss> or nothing?
[19-Aug-2009 15:33:07] <dranch> I'm not looking to graph, only send alerts
[19-Aug-2009 15:33:16] <mrayzenoss> is the template bound?
[19-Aug-2009 15:33:25] <dranch> how do I confrim that?
[19-Aug-2009 15:33:45] <mrayzenoss> go to the device class with the template
[19-Aug-2009 15:33:53] <mrayzenoss> click "Templates"
[19-Aug-2009 15:33:54] <dranch> ok
[19-Aug-2009 15:34:06] <dranch> ok
[19-Aug-2009 15:34:07] <mrayzenoss> go to the page menu and select "Bind Templates"
[19-Aug-2009 15:34:26] <dranch> ok..
[19-Aug-2009 15:34:41] <mrayzenoss> is your custom template highlighted like Device?
[19-Aug-2009 15:34:51] <dranch> I see device, ethernetcdmad*, filesystem, ipservice, mysql, osprocesses, and winservice
[19-Aug-2009 15:35:01] <dranch> device is highlighed
[19-Aug-2009 15:35:12] <dranch> that's where I created my data source
[19-Aug-2009 15:36:01] <mrayzenoss> no, on the Templates page, there's a menu next to "Available Performance Templates"
[19-Aug-2009 15:36:28] <mrayzenoss> and Bind Templates is a choice there
[19-Aug-2009 15:36:45] <dranch> right.. under the down arrow right?
[19-Aug-2009 15:36:52] <mrayzenoss> yeah
[19-Aug-2009 15:37:12] <dranch> right.. and that pops up a window with the list of templates I listed above
[19-Aug-2009 15:37:18] <mrayzenoss> ahh, ok
[19-Aug-2009 15:37:27] <mrayzenoss> and your custom template is bound?
[19-Aug-2009 15:37:36] <dranch> Device [ /devices/server/linux/database/replication] is the top entry and it's highlighed
[19-Aug-2009 15:38:01] <mrayzenoss> so you modified the Device template itself, instead of a custom template?
[19-Aug-2009 15:38:03] <dranch> Should I click on "ok" to re-bind it?
[19-Aug-2009 15:38:09] <dranch> yes
[19-Aug-2009 15:38:27] <dranch> should I place it somewhere else?
[19-Aug-2009 15:38:27] <mrayzenoss> that should be ok
[19-Aug-2009 15:39:01] <dranch> I have three database servers as part of this template and I'd like to run this command on all three machines
[19-Aug-2009 15:40:18] <mrayzenoss> I assume you've pushed your changes and restarted everything for grins?
[19-Aug-2009 15:41:19] <dranch> This is a single machine and I have restarted a few times (zencommand and zenhub). Should I do the whole thing?
[19-Aug-2009 15:42:02] <dranch> I don't understand why zenoss things there aren't any commands to run. There are a total of three commands on that page and it seems NONE of them are being run
[19-Aug-2009 15:42:43] <mrayzenoss> dranch: I'm with you, it sounds like everything should be in the right places
[19-Aug-2009 15:43:07] <mrayzenoss> perhaps remove the custom device template and try adding the commands manually
[19-Aug-2009 15:43:18] <mrayzenoss> I mean, a separate template for each
[19-Aug-2009 15:43:19] <dranch> This has been enlightening in the fact that now I know that Zenoss isn't running anything that's been configured
[19-Aug-2009 15:43:30] <dranch> Uh... ok, let me try
[19-Aug-2009 15:43:30] <mrayzenoss> just to narrow it down
[19-Aug-2009 15:43:46] <dranch> There are also a ton of SNMP values
[19-Aug-2009 15:44:13] <mrayzenoss> ok, just copy and paste the command stuff into a new template
[19-Aug-2009 15:46:06] <mrayzenoss> man, I was working on a bug related to this right before I left on vacation 2 weeks ago
[19-Aug-2009 15:46:13] <mrayzenoss> on my home box that crashed
[19-Aug-2009 15:46:20] <dranch> figures huh?
[19-Aug-2009 15:46:29] <dranch> (for me at least)
[19-Aug-2009 15:46:39] <mrayzenoss> I noticed it had crashed this morning and I was going to take a screenshot so I could debug it
[19-Aug-2009 15:46:51] <mrayzenoss> some weird kernel issue on my Dell mini
[19-Aug-2009 15:46:59] <mrayzenoss> it
[19-Aug-2009 15:47:06] <mrayzenoss> it's been crashing after 2 weeks
[19-Aug-2009 15:47:11] <mrayzenoss> fairly consistently
[19-Aug-2009 15:47:12] <dranch> Ok.. so I added the new template
[19-Aug-2009 15:47:19] <jb> how can I verify relationships in zope?
[19-Aug-2009 15:47:21] <dranch> do I need to bind it?
[19-Aug-2009 15:47:24] <jb> somehow thru zport?
[19-Aug-2009 15:47:33] <mrayzenoss> dranch: yes
[19-Aug-2009 15:47:37] <scp2004> [mrayzenoss] sorry man ..I was afk .. About the error using zendevicedump .. I must use it with zenoss running ?
[19-Aug-2009 15:47:46] <mrayzenoss> scp2004: yes
[19-Aug-2009 15:47:54] <scp2004> oh ..ok ..i ll try
[19-Aug-2009 15:48:25] <dranch> In the bind performance templates, do I only need to select this new "DBmonitor" entry and select ok or do multiple things need to be selected?
[19-Aug-2009 15:50:23] <rmatte> I can't wait until the day when http://dev.zenoss.org/trac/ticket/5289 gets fixed. It's usually not a big deal but it's absolutely driving me up the wall right now
[19-Aug-2009 15:51:01] <jb> hmm
[19-Aug-2009 15:51:01] <jb> you know
[19-Aug-2009 15:51:05] <jb> i think i need to manually rebuild my relations.
[19-Aug-2009 15:51:48] <scp2004> [mrayzenoss] Waf ! ... .. File "/usr/local/zenoss/Products/ZenRelations/ExportDevices.py", line 229, in ?
[19-Aug-2009 15:51:48] <scp2004> ex.export()
[19-Aug-2009 15:51:53] <dranch> Ray? thoughts?
[19-Aug-2009 15:52:16] <mrayzenoss> dranch: yeah, select the new one and the previous ones
[19-Aug-2009 15:52:21] <rmatte> dranch: depends on what templates you want assigned to it
[19-Aug-2009 15:52:26] <dranch> So ALL should be selected?
[19-Aug-2009 15:52:34] <mrayzenoss> all that you want to use
[19-Aug-2009 15:52:35] <rmatte> I wouldn't select ALL
[19-Aug-2009 15:52:38] <rmatte> just the relevant ones
[19-Aug-2009 15:53:03] <mrayzenoss> templates are available hierarchally, so there may be some defined farther up the hierarchy that you don't care for
[19-Aug-2009 15:53:19] <dranch> How can I tell?
[19-Aug-2009 15:53:53] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 15:54:15] <rmatte> dranch: you just kind of have to know
[19-Aug-2009 15:55:54] <dranch> restarted zencommand and zenhub, still schedule has 0 commands
[19-Aug-2009 15:57:06] <scp2004> [mrayzenoss] any idea ? .. File "/usr/local/zenoss/Products/ZenRelations/ExportDevices.py", line 229, in ?
[19-Aug-2009 15:57:33] <mrayzenoss> scp2004: use pastebin.com and post the whole stack trace
[19-Aug-2009 15:57:51] <mrayzenoss> dranch: try "Push Changes"
[19-Aug-2009 15:58:00] <dranch> Where's that?
[19-Aug-2009 15:58:04] <mrayzenoss> from the device in question, go to the page menu
[19-Aug-2009 15:58:43] <mrayzenoss> on the status page, below "Manage->Model Device"
[19-Aug-2009 15:59:37] <dranch> ok.. done
[19-Aug-2009 16:00:12] <rmatte> yeh, Push Changes is your friend
[19-Aug-2009 16:00:20] <mrayzenoss> did that work?
[19-Aug-2009 16:00:26] <rmatte> since the collector takes upwards of an hour to grab updated info from zenhub
[19-Aug-2009 16:00:44] <mrayzenoss> lazy zenhub
[19-Aug-2009 16:00:54] <rmatte> yeh lol
[19-Aug-2009 16:01:06] <mrayzenoss> it needs more workers
[19-Aug-2009 16:01:17] <dranch> :zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:01:27] <rmatte> is the command enabled?
[19-Aug-2009 16:01:54] <dranch> enabled: true
[19-Aug-2009 16:02:01] <rmatte> ok
[19-Aug-2009 16:02:03] <rmatte> hmmm
[19-Aug-2009 16:02:36] <rmatte> but other command based datapoints are working?
[19-Aug-2009 16:03:14] <dranch> name: DB VIP monitor; source type: command; enabled true; use ssh: false; component: DB VIP heartbeat; event class: /Cmd/Fail; Event key: DB heartbeat; severity: critical; cycle time: 60l command template: $$ZENHOME/libexec/check-vip.sh database-fail
[19-Aug-2009 16:03:45] <dranch> No, there are two other commands that I "THOUGHT" were running but it seems they aren't being issued either
[19-Aug-2009 16:04:01] <rmatte> cycle time: 60l?
[19-Aug-2009 16:04:14] <dranch> sorry: 60
[19-Aug-2009 16:04:18] <rmatte> k
[19-Aug-2009 16:04:53] <rmatte> what's the output of that script look like?
[19-Aug-2009 16:05:11] <dranch> In the device template, there is a SSH commend to run mysql, and another command to run /opt/zenoss/libexec/check_mysql_stat -u zenoss -p xxx -H ${device/manageIp} -S
[19-Aug-2009 16:05:21] <dranch> seems those aren't running now either.
[19-Aug-2009 16:05:42] <rmatte> but they were working fine previously I assume?
[19-Aug-2009 16:06:07] <dranch> test against device: database5 ---> Executing command against database5.admin.zvents.com
[19-Aug-2009 16:06:08] <dranch> 1
[19-Aug-2009 16:06:10] <dranch> DONE in 1 seconds
[19-Aug-2009 16:06:24] <rmatte> ummmm, that's not really how a command based datapoint works
[19-Aug-2009 16:06:26] <dranch> I've inherited this setup.. I don't have much history with it
[19-Aug-2009 16:06:45] <rmatte> you need the output to be something like OK|status=1 or OK|status=0 or Unknown
[19-Aug-2009 16:06:58] <rmatte> then you need to make a datapoint named "status" under the data source
[19-Aug-2009 16:07:09] <dranch> hmmm.. ok. let me change the output
[19-Aug-2009 16:07:12] <rmatte> otherwise Zenoss doesn't know how to interpret the data
[19-Aug-2009 16:07:44] <rmatte> you can even just do |status=1 or |status=2 or Unknown
[19-Aug-2009 16:07:52] <rmatte> Nagios format output
[19-Aug-2009 16:07:56] <dranch> So if the test is /Cmd/fail, and it's a fail, the result should be a OK|status=0?
[19-Aug-2009 16:08:09] <rmatte> yes
[19-Aug-2009 16:08:27] <rmatte> if 0 is what indicated a failure
[19-Aug-2009 16:08:31] <rmatte> indicates*
[19-Aug-2009 16:08:50] <rmatte> and if it succeeds it should be OK|status=1
[19-Aug-2009 16:08:59] <rmatte> the OK| just represents that it was able to gather the data
[19-Aug-2009 16:09:14] <rmatte> if you whatever reason it can't check (which is unlikely in this case) it should output Unknown
[19-Aug-2009 16:09:19] <rmatte> but I wouldn't worry about that part
[19-Aug-2009 16:09:37] <dranch> But it from the above zencommand logs, the commands don't seem to be running in the first place
[19-Aug-2009 16:09:41] <dranch> or is that a logging issue?
[19-Aug-2009 16:09:48] <mrayzenoss> dranch: the instructions with screenshots for http://www.zenoss.com/community/projects/zenpacks/nagios-check-ping work well
[19-Aug-2009 16:09:58] <rmatte> well, if you haven't created any datapoints under the command then it's not going to try to bother running
[19-Aug-2009 16:10:28] <rmatte> the datapoint name needs to match what's in the output
[19-Aug-2009 16:10:46] <rmatte> so if you had OK|whatever=1 your datapoint name would be whatever
[19-Aug-2009 16:11:20] <dranch> Ok, for a /Cmd/Fail setup, and and an test FAILURE, it's now returning: OK|status=1
[19-Aug-2009 16:11:24] <rmatte> that way you can have scripts outputting multiple info like OK|CPU1=7 CPU2=10 CPU3=20 CPU4=80
[19-Aug-2009 16:11:42] <rmatte> and then create the datapoints CPU1, CPU2, CPU3, and CPU4 under the command
[19-Aug-2009 16:11:52] <rmatte> and zenoss will assign the corresponding value to each datapoint
[19-Aug-2009 16:12:09] <rmatte> ok, so 1 indicates a failure I assume?
[19-Aug-2009 16:12:13] <dranch> Again, I'm just looking for a failure alert...
[19-Aug-2009 16:12:15] <dranch> correct
[19-Aug-2009 16:12:20] <dranch> I don't need to graph, etc
[19-Aug-2009 16:12:29] <rmatte> ok, so now, under the command options, you'll see a DataPoints section
[19-Aug-2009 16:12:41] <dranch> yup..
[19-Aug-2009 16:12:44] <rmatte> you need to add a datapoint that's called status
[19-Aug-2009 16:12:51] <rmatte> you can just use gauge for the type
[19-Aug-2009 16:12:58] <dranch> ok
[19-Aug-2009 16:13:07] <rmatte> then you can add your threshold against that datapoint and set it to 1, so it triggers when it hits 1
[19-Aug-2009 16:13:44] <dranch> there is name: status; type: gauge; RRD min <null>;l RRD max <null> and what do I send in create cmd: ?
[19-Aug-2009 16:14:12] <rmatte> just leave the datapoint options at default
[19-Aug-2009 16:14:20] <rmatte> you don't need to change anything, the datapoint just has to exist
[19-Aug-2009 16:14:58] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:15:03] <rmatte> once it's there just create a threshold against it, and hopefully Zenoss will run it
[19-Aug-2009 16:15:35] <dranch> ok, and that's not done in the template? if not, where?
[19-Aug-2009 16:15:44] <rmatte> the threshold is done right in the template
[19-Aug-2009 16:15:52] <rmatte> in the thresholds section
[19-Aug-2009 16:16:11] <rmatte> add a min-max threshold, name it whatever
[19-Aug-2009 16:16:20] <rmatte> select that datapoint
[19-Aug-2009 16:16:26] <rmatte> it'll be whatever_status
[19-Aug-2009 16:16:37] <rmatte> set the max value to 1, and set the event class
[19-Aug-2009 16:16:43] <rmatte> severity as well if you wish
[19-Aug-2009 16:16:45] <rmatte> then click save
[19-Aug-2009 16:16:47] <dranch> I'm back at the templates tab, under the down arrow, I don't see any option for threshold
[19-Aug-2009 16:17:01] <rmatte> no, go in to the template itself
[19-Aug-2009 16:17:08] <rmatte> and there's a threshold section
[19-Aug-2009 16:17:12] <dranch> Ah.. I see it. sorry
[19-Aug-2009 16:17:16] <rmatte> no prob
[19-Aug-2009 16:17:21] <rmatte> follow my above instructions
[19-Aug-2009 16:17:47] <rmatte> make sure the threshold is enabled as well
[19-Aug-2009 16:18:02] <rmatte> then at least you're done for the config end of things
[19-Aug-2009 16:18:43] <rmatte> then all we need to do is figure out why the heck the command isn't running
[19-Aug-2009 16:18:47] <dranch> ok.. done but do I have to restart zencommand, etc?
[19-Aug-2009 16:19:07] <rmatte> At this point i'd just restart Zenoss for the heck of it to see if that clears it up
[19-Aug-2009 16:19:33] <dranch> restarting
[19-Aug-2009 16:19:40] <rmatte> k
[19-Aug-2009 16:20:13] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:20:20] <rmatte> gah
[19-Aug-2009 16:20:40] <rmatte> what options are you using with zencommand?
[19-Aug-2009 16:21:03] <dranch> cd /usr/local/zenoss/bin; ./zencommand run -v10 -d database5.admin
[19-Aug-2009 16:21:14] <rmatte> try: zencommand run -v10 --cycle
[19-Aug-2009 16:21:21] <rmatte> see if any commands run through at all
[19-Aug-2009 16:21:36] <dranch> without the cycle, I do see:
[19-Aug-2009 16:21:40] <dranch> DEBUG:zen.zencommand:run
[19-Aug-2009 16:21:41] <dranch> DEBUG:zen.zencommand:Connecting to localhost
[19-Aug-2009 16:21:43] <dranch> DEBUG:zen.zencommand:Logging in as admin
[19-Aug-2009 16:21:44] <dranch> WARNING:zen.zencommand:Reconnected to ZenHub
[19-Aug-2009 16:21:46] <dranch> DEBUG:zen.zencommand:setting up services EventService, CommandConfig
[19-Aug-2009 16:21:47] <dranch> DEBUG:zen.zencommand:chaining getInitialServices with d2
[19-Aug-2009 16:21:49] <dranch> DEBUG:zen.zencommand:callback after getting service EventService
[19-Aug-2009 16:21:50] <dranch> DEBUG:zen.zencommand:callback after getting service CommandConfig
[19-Aug-2009 16:21:52] <dranch> DEBUG:zen.zencommand:Calling connected.
[19-Aug-2009 16:21:53] <dranch> DEBUG:zen.zencommand:connected
[19-Aug-2009 16:21:55] <dranch> DEBUG:zen.zencommand:Sending event {'severity': 0, 'component': 'zencommand', 'agent': 'zencommand', 'summary': 'started', 'manager': 'localhost', 'device': 'admin2.admin.zvents.com', 'eventClass': '/App/Start'}
[19-Aug-2009 16:21:56] <dranch> DEBUG:zen.zencommand:Fetching config
[19-Aug-2009 16:21:58] <dranch> DEBUG:zen.zencommand:Updated configCycleInterval config to 360
[19-Aug-2009 16:21:59] <dranch> DEBUG:zen.zencommand:Loading classes ['Products.ZenModel.MinMaxThreshold']
[19-Aug-2009 16:22:01] <dranch> DEBUG:zen.thresholds:Updating threshold ('zeneventlog cycle time', ('localhost', ''))
[19-Aug-2009 16:22:02] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenmodeler cycle time', ('localhost', ''))
[19-Aug-2009 16:22:04] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenperfsnmp cycle time', ('localhost', ''))
[19-Aug-2009 16:22:05] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenping cycle time', ('localhost', ''))
[19-Aug-2009 16:22:07] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenprocess cycle time', ('localhost', ''))
[19-Aug-2009 16:22:09] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenwin cycle time', ('localhost', ''))
[19-Aug-2009 16:22:11] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:22:13] <dranch> DEBUG:zen.zencommand:Finished config fetch
[19-Aug-2009 16:22:15] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:22:17] <dranch> INFO:zen.zencommand:Daemon zencommand shutting down
[19-Aug-2009 16:22:19] <dranch> DEBUG:zen.zencommand:removing service EventService
[19-Aug-2009 16:22:21] <dranch> DEBUG:zen.zencommand:removing service CommandConfig
[19-Aug-2009 16:22:23] <dranch> INFO:zen.zencommand:zencommand shutting down
[19-Aug-2009 16:22:25] <dranch> --cycle didn't change anything
[19-Aug-2009 16:22:32] <rmatte> strange
[19-Aug-2009 16:22:49] <rmatte> so it's running the zenoss internal commands but not a single template command
[19-Aug-2009 16:23:14] <dranch> seems like it
[19-Aug-2009 16:23:34] <rmatte> ah
[19-Aug-2009 16:23:36] <rmatte> http://forums.zenoss.com/viewtopic.php?p=36653
[19-Aug-2009 16:23:43] <rmatte> the upgrade caused a "null-value" for zSshConcurrentSessions. zencommand doesnt
[19-Aug-2009 16:23:43] <rmatte> run at all without any value. Setting a value will fix it. I set it to 10.
[19-Aug-2009 16:23:58] <dranch> could be that the others are misconfigured though. There are only 4 thresholds. CPU utilizatio, raw user 400, raw user 500, and this new DB VIP change
[19-Aug-2009 16:23:58] <rmatte> that's apparently the solution
[19-Aug-2009 16:24:09] <rmatte> go in to zProperties and set that value
[19-Aug-2009 16:24:48] <dranch> It's not run via SSH
[19-Aug-2009 16:25:03] <rmatte> doesn't matter, that seems to cause general issues with zencommand
[19-Aug-2009 16:25:05] <rmatte> regardless
[19-Aug-2009 16:25:13] <rmatte> just check to make sure a value is set
[19-Aug-2009 16:25:18] <mrayzenoss> is that error in the zenhub log?
[19-Aug-2009 16:25:34] <jb> yawn
[19-Aug-2009 16:26:43] <dranch> I don't have that option available in this zProperties
[19-Aug-2009 16:26:44] <rmatte> just check in zProperties for /Devices, if it's already set then great, if not, then that's your problem
[19-Aug-2009 16:26:47] <rmatte> default value is 10
[19-Aug-2009 16:26:53] <rmatte> what version of Zenoss?
[19-Aug-2009 16:27:13] <dranch> Zenoss 2.2.4
[19-Aug-2009 16:27:18] <rmatte> I have a 2.3.3 box that doesn't have that option
[19-Aug-2009 16:27:21] <rmatte> ah, you're super oldschool
[19-Aug-2009 16:27:25] <dranch> Upgrading isn't an option
[19-Aug-2009 16:27:25] <rmatte> ok, that changes things
[19-Aug-2009 16:28:14] <rmatte> well, i'm pretty much out of ideas, if we don't have any output or logs to go on there's not much that can be done
[19-Aug-2009 16:28:23] <rmatte> need to have some idea of what may be wrong
[19-Aug-2009 16:28:31] <dranch> What do you need?
[19-Aug-2009 16:28:52] <rmatte> you've checked $ZENHOME/logs to see if any of the files contain errors?
[19-Aug-2009 16:29:03] <rmatte> (errors that may be related to the cause of this?)
[19-Aug-2009 16:29:14] <dranch> let me see if anything new is in there
[19-Aug-2009 16:29:26] <dranch> nothing in zencommand.log
[19-Aug-2009 16:29:26] <robos> hi: for the disk i/o performance graph, i have a solaris box that shows the graph climbing.. which is inaccurate. However the linux box works as expected. Any suggestions on what to check?
[19-Aug-2009 16:30:29] <mrayzenoss> robos: SNMP or SSH?
[19-Aug-2009 16:30:30] <rmatte> robos: just climbing infinitely?
[19-Aug-2009 16:30:42] <robos> snmp. And it climbs infinitely... just keeps going up
[19-Aug-2009 16:30:46] <robos> never goes down
[19-Aug-2009 16:31:21] <mrayzenoss> sounds like it's probably got the wrong OID
[19-Aug-2009 16:31:30] <robos> wrong OID?
[19-Aug-2009 16:31:32] <rmatte> yeh, does sound that way
[19-Aug-2009 16:31:44] <rmatte> sounds like it's got an OID for uptime or something
[19-Aug-2009 16:32:00] <robos> hmm
[19-Aug-2009 16:32:01] <mrayzenoss> robos: so SNMP uses individual datapoints called OIDs that expose a metric for collection
[19-Aug-2009 16:32:02] <rmatte> (or something else that increments)
[19-Aug-2009 16:32:15] <robos> ah
[19-Aug-2009 16:32:26] <robos> gotcha
[19-Aug-2009 16:32:29] <mrayzenoss> which version of Solaris?
[19-Aug-2009 16:33:06] <rmatte> mrayzenoss: is it maybe because of the datapoint type?
[19-Aug-2009 16:33:09] <rmatte> might not be a gauge?
[19-Aug-2009 16:33:21] <robos> solaris 10
[19-Aug-2009 16:33:28] <robos> i have a bunch added and they all do that
[19-Aug-2009 16:34:37] <rmatte> actually yeh, the type wouldn't affect it
[19-Aug-2009 16:34:56] <dranch> rmatte: so looking in the zencommand.log and there ARE other zencommands being run for other database servers that are under a different template. Btw, those tests DON'T have any datapoints or threasholds configured
[19-Aug-2009 16:35:32] <rmatte> dranch: hmmm, so their output is just 1 or whatever?
[19-Aug-2009 16:36:41] <rmatte> I'm leaving for the day, gotta go to catch my ride
[19-Aug-2009 16:37:03] <dranch> <sigh>.. well, thanks for the help you've given me.. though I'm still foobared
[19-Aug-2009 16:39:35] <mrayzenoss> robos: this is all I could find: http://dev.zenoss.org/trac/ticket/5237
[19-Aug-2009 16:39:47] <mrayzenoss> and I asked the dev about I/O and he said he hadn't seen anything like that
[19-Aug-2009 16:40:27] <robos> hmm, so what does this mean mrayzenoss?
[19-Aug-2009 16:40:35] <robos> i'm looking at that ticket..
[19-Aug-2009 16:41:14] <mrayzenoss> it means we fixed some other stuff for Solaris 10, but not your bug
[19-Aug-2009 16:41:20] <robos> ah
[19-Aug-2009 16:41:31] <robos> then i wonder why nobody else has this problem
[19-Aug-2009 16:41:35] <mrayzenoss> but I know we've got Solaris users
[19-Aug-2009 16:41:47] <robos> this is solaris 10 on SPARC... a T1000 i believe
[19-Aug-2009 16:43:03] <robos> *shrugs, guess I'll look into this tomorrow
[19-Aug-2009 16:43:08] <robos> cool; ty for your time guys
[19-Aug-2009 16:45:10] <dranch> myray: when do you get in tomorrow?
[19-Aug-2009 16:45:42] <dranch> Since some templates run commands and others don't, I think something uglier is afoot
[19-Aug-2009 16:46:40] <mrayzenoss> I'll be in for the IRC session tomorrow
[19-Aug-2009 16:46:43] <mrayzenoss> and afterwards
[19-Aug-2009 16:46:57] <dranch> sure.. but what hours PST?
[19-Aug-2009 16:47:07] * mrayzenoss does the math
[19-Aug-2009 16:47:12] <mrayzenoss> 8am
[19-Aug-2009 16:47:47] <mrayzenoss> from the blog: Zenoss Core Reaches 1 Million Downloads http://blog.zenoss.com/2009/08/19/network-monitoring/
[19-Aug-2009 16:47:52] <dranch> Ok.. I'll keep hacking around to see if I can get anything to run. By seeing other commands in other templates running WITHOUTH thresholds, etc. this is something else
[19-Aug-2009 16:48:09] <mrayzenoss> dranch: yeah, I suspect something is hosed in the binding
[19-Aug-2009 16:48:17] <mrayzenoss> just hard to find that
[19-Aug-2009 16:48:29] <dranch> yeah..
[19-Aug-2009 16:48:39] <mrayzenoss> try this... go to the status page for the machine being monitored
[19-Aug-2009 16:48:54] <dranch> ok
[19-Aug-2009 16:49:10] <mrayzenoss> select More->Templates
[19-Aug-2009 16:49:20] <dranch> uhuh
[19-Aug-2009 16:49:30] <mrayzenoss> is your template bound there?
[19-Aug-2009 16:49:59] <dranch> yup.. both the modified Device template and the new "DBmonitor" one. Neither seem to run
[19-Aug-2009 16:50:04] <mrayzenoss> crap
[19-Aug-2009 16:50:22] <mrayzenoss> I saw templates bound at the device class not get bound to the device before
[19-Aug-2009 16:50:26] <dranch> I even made the DBMonitor one to be a "local" copy. No change
[19-Aug-2009 16:50:41] <dranch> The path for that template is: /Devices /Server /Linux /database /replication /database5.admin.zvents.com
[19-Aug-2009 16:50:57] <mrayzenoss> my home box will be back up tomorrow, I'll have a better testbed
[19-Aug-2009 16:50:58] <dranch> This one DOES work: /Devices /Server /Linux /database /non-replication /load-database2.admin.zvents.com
[19-Aug-2009 16:51:24] <dranch> It shows only one template: device
[19-Aug-2009 16:52:06] <dranch> ha.. it also shows: INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:52:13] <dranch> I don't trust the logging now
[19-Aug-2009 16:52:36] <mrayzenoss> well, the schedule is things that are the queue
[19-Aug-2009 16:52:41] <mrayzenoss> so the queue isn't backed up
[19-Aug-2009 16:52:55] <dranch> that's not an atomic test?
[19-Aug-2009 16:53:05] <mrayzenoss> I'm not sure
[19-Aug-2009 16:53:15] <dranch> Sure seems like it is
[19-Aug-2009 16:53:16] * mrayzenoss wishes he could get to his test box
[19-Aug-2009 16:53:32] <mrayzenoss> I've got a bunch of zencommands on my home box
[19-Aug-2009 16:54:37] <dranch> Hmmm.. seems that this load-database2 WAS working.. but stopped around 2:18.. I broke something
[19-Aug-2009 16:54:57] <dranch> 2009-08-19 14:18:49 DEBUG zen.zencommand: The result of "/opt/zenoss/libexec/check_mysql_stat -u zenoss -p xxx -H 10.0.30.174" was "Uptime: 30365675 Threads: 1
[19-Aug-2009 16:56:37] <dranch> ack.. false alarm.. zencommand was missing the -v10 option.. it's still running
[19-Aug-2009 16:57:27] <dranch> even still, the manual zencommand shows INFO:zen.zencommand:---------- - schedule has 0 commands :: which is wrong
[19-Aug-2009 18:10:46] kisielk_ is now known as kisielk
[19-Aug-2009 18:58:15] <straterra> rmatte: you around?
[20-Aug-2009 00:00:46] [disconnected at Thu Aug 20 00:00:46 2009]
[20-Aug-2009 00:00:46] [connected at Thu Aug 20 00:00:46 2009]
[20-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[20-Aug-2009 05:15:05] cwj_ is now known as cwj
[20-Aug-2009 09:16:35] <ckrough> "INFO ZODB.Conflict database read conflict error" when I try to run a /viewGraphReport , and I get no graph report.
[20-Aug-2009 09:16:39] <rmatte> has anyone else noticed that they've removed the --debug option from zenmib in latest Zenoss versions?
[20-Aug-2009 09:16:56] <ckrough> nah, kinda load mib once and never look back for me
[20-Aug-2009 09:17:09] <rmatte> that option was pure gold, I can't believe that
[20-Aug-2009 09:17:21] <rmatte> I would have never been able to get the Cisco Mibs loaded without it
[20-Aug-2009 09:17:44] <ckrough> which version?
[20-Aug-2009 09:17:57] <rmatte> and now it appears to be gone, since the script isn't accepting the option, and there's absolutely no reference to it in zenmib.py
[20-Aug-2009 09:18:02] <rmatte> 2.4.1
[20-Aug-2009 09:18:07] <rmatte> and later versions
[20-Aug-2009 09:19:36] <jb> morning
[20-Aug-2009 09:19:41] <rmatte> good morning
[20-Aug-2009 09:19:49] <ckrough> morning
[20-Aug-2009 09:22:12] <rmatte> yeh, the option is available on my 2.3.3 boxes
[20-Aug-2009 09:22:16] <rmatte> but not on my 2.4.1 boxes
[20-Aug-2009 09:22:24] <rmatte> no idea why they'd remove that option
[20-Aug-2009 09:23:04] <rmatte> I'll ask Matt about it when he signs on
[20-Aug-2009 09:24:57] <rmatte> hopefully there's some alternative
[20-Aug-2009 09:25:09] <ckrough> does it provide more than V10
[20-Aug-2009 09:25:27] <rmatte> ckrough: hmmm, that's actually a good question
[20-Aug-2009 09:25:34] <ckrough> Maybe it was redundna
[20-Aug-2009 09:25:37] <ckrough> redundant
[20-Aug-2009 09:25:54] <rmatte> yeh, I never thought of that (old habbits die hard)
[20-Aug-2009 09:25:58] <rmatte> testing now...
[20-Aug-2009 09:29:05] <rmatte> seems to
[20-Aug-2009 09:29:13] <rmatte> I guess it was redundant
[20-Aug-2009 09:29:33] <rmatte> I was just used to using it, I'll have to update my article on loading cisco Mibs to include that instead
[20-Aug-2009 09:30:02] <ckrough> cool
[20-Aug-2009 09:30:21] <ckrough> so, your writeup for the mibs is strictly to get all the traps recognized, right?
[20-Aug-2009 09:30:46] <rmatte> yeh, Mibs translate traps
[20-Aug-2009 09:30:51] <rmatte> Zenoss doesn't use them for anything else
[20-Aug-2009 09:32:07] <twm1010> a MIB walker to help design performance templates would be fantastic.
[20-Aug-2009 09:32:26] <ckrough> What do you mean
[20-Aug-2009 09:33:21] <twm1010> well... for instance in the SolarWinds universal device polar, you can browse the MIB database built-in, find the vendor, device, etc, and test the value you want to monitor, right there in the window, and then turn that into an assignable poller.
[20-Aug-2009 09:33:39] <twm1010> Zenoss could do the same thing, except you could just say "create a datasource" and choose which template to drop it into.
[20-Aug-2009 09:33:48] <rmatte> twm1010: install the Mib Browser ZenPack, it's sort of like that
[20-Aug-2009 09:33:49] <ckrough> that would be nice
[20-Aug-2009 09:33:55] <twm1010> i have it, can it do that?
[20-Aug-2009 09:34:00] <rmatte> well, in terms of browsing the Mib tree
[20-Aug-2009 09:34:07] <rmatte> you still need to add the OIDs to templates by hand
[20-Aug-2009 09:34:11] <rmatte> but that takes seconds to do anyways
[20-Aug-2009 09:34:24] <rmatte> the Mib browser window opens in a new window too, which is nice
[20-Aug-2009 09:34:37] <rmatte> so you can browse the Mibs tree while editing the template
[20-Aug-2009 09:35:04] <rmatte> I'm sure it's not as flashy as solarwinds' system though
[20-Aug-2009 09:35:22] <rmatte> but it runs on Linux, so it gets points for that
[20-Aug-2009 09:35:48] <twm1010> and it doens't cost you 20 grand a year either.
[20-Aug-2009 09:35:54] <jb> yes it does
[20-Aug-2009 09:35:55] <jb>
[20-Aug-2009 09:36:01] <ckrough> heh
[20-Aug-2009 09:36:02] <twm1010> err.. well SW isn't that much, its 20 initial, 3 a year
[20-Aug-2009 09:36:13] <rmatte> if you buy enterprise
[20-Aug-2009 09:36:13] <ckrough> Enterprise...
[20-Aug-2009 09:36:16] <jb> zenoss is the same amount every year.. its subscription based.
[20-Aug-2009 09:36:19] <jb> right.
[20-Aug-2009 09:36:28] <jb> the only thing that solarwinds excels at is reporting and network device monitoring..
[20-Aug-2009 09:36:35] <ckrough> and performance with large installs
[20-Aug-2009 09:36:40] <jb> yeah
[20-Aug-2009 09:36:46] <rmatte> Zenoss excels at network device monitoring as well in my opinion
[20-Aug-2009 09:36:50] <ckrough> we've outscaled RRD... just throwing hardware at it now
[20-Aug-2009 09:36:56] <rmatte> network device monitoring isn't crazy complex
[20-Aug-2009 09:36:59] <jb> rmatte: with customization, yeah
[20-Aug-2009 09:37:07] <twm1010> ckrough: are you using distributed collectors?
[20-Aug-2009 09:37:18] <jb> im looking to do a core installation to monitor 900 windows boxes
[20-Aug-2009 09:37:19] <rmatte> jb: well yeh, it's not one of these "ready to go out of the box" type things, but I actually prefer to customize it myself
[20-Aug-2009 09:37:23] <ckrough> yeah, last count I was at 26
[20-Aug-2009 09:37:33] <rmatte> more flexibility, so provided you spend the development time you can monitor literally anything
[20-Aug-2009 09:37:37] <twm1010> wow, 26 collectors? how many nodes?
[20-Aug-2009 09:37:40] <rmatte> even some obscure POS that no one has heard of before
[20-Aug-2009 09:37:40] <jb> you have 26 collectors?
[20-Aug-2009 09:38:11] <ckrough> thats total over 5 regions, so 5 installs.. largest is 10k devices, total is around 18k devices
[20-Aug-2009 09:38:25] <ckrough> just network gear
[20-Aug-2009 09:38:30] <ckrough> dont do much systems monitoring
[20-Aug-2009 09:38:30] <jb> nice
[20-Aug-2009 09:38:36] <twm1010> 5 regions 5 installs, 5 collectors at each?
[20-Aug-2009 09:38:57] <ckrough> 1 master at each, then sets of collectors. largest has 10 collectors
[20-Aug-2009 09:39:21] <ckrough> lots and lots of OIDs since Im grabbing info for each interface on each switch
[20-Aug-2009 09:39:29] <twm1010> i'd imagine so.
[20-Aug-2009 09:40:03] <twm1010> you on core or enterprise?
[20-Aug-2009 09:40:06] <ckrough> ent
[20-Aug-2009 09:40:40] <jb> ckrough: are you pleased with the ent distributed collectors?
[20-Aug-2009 09:40:44] <jb> i'll be throwing a couple up soon
[20-Aug-2009 09:40:54] <jb> do you have to maintain things such as zenpacks, etc on every collector
[20-Aug-2009 09:41:08] <ckrough> not in the later versions. the master pushes out just about everything
[20-Aug-2009 09:41:13] <jb> awesome
[20-Aug-2009 09:41:16] <ckrough> very
[20-Aug-2009 09:41:17] <jb> i was concerned about that
[20-Aug-2009 09:41:28] <jb> so whats your upgrade process like?
[20-Aug-2009 09:41:30] <ckrough> Used to do it that way. ClusterSSH ftw
[20-Aug-2009 09:41:36] <rmatte> yeh, that'd be nice, I have multiple core instances that I have to keep up to date by hand
[20-Aug-2009 09:41:36] <jb> yeah
[20-Aug-2009 09:41:39] <rmatte> it's fun fun
[20-Aug-2009 09:41:43] <rmatte> though I have scripts that help me
[20-Aug-2009 09:41:44] <rmatte> hehe
[20-Aug-2009 09:41:49] <ckrough> upgrade process involves Chet
[20-Aug-2009 09:41:56] <jb> thats what i thought
[20-Aug-2009 09:42:10] <twm1010> I am unforunately stuck with only my production server right now :|
[20-Aug-2009 09:42:17] <jb> oh, are you RS?
[20-Aug-2009 09:42:23] <ckrough> yeah, it's got to. Quite a few customisations rolled in now that get goofy on upgrades
[20-Aug-2009 09:42:30] <ckrough> yes
[20-Aug-2009 09:42:33] <jb> ok
[20-Aug-2009 09:42:52] <jb> where are you based out of if you dont mind me asking?
[20-Aug-2009 09:43:01] <ckrough> company or me? Im in San Antonio
[20-Aug-2009 09:43:05] <jb> ah ok
[20-Aug-2009 09:43:07] <rmatte> oh, rackspace, we monitor some servers that are hosted by you guys
[20-Aug-2009 09:43:13] <ckrough> oh yeah? nice
[20-Aug-2009 09:43:18] <rmatte> yup
[20-Aug-2009 09:43:30] <ckrough> rmatte: what region?
[20-Aug-2009 09:43:46] <rmatte> and some cisco gear looks like
[20-Aug-2009 09:43:52] <rmatte> looks like they are in europe
[20-Aug-2009 09:44:27] <rmatte> well, they are named rackspace
[20-Aug-2009 09:44:49] <ckrough> just availability I assume
[20-Aug-2009 09:44:56] <ckrough> /I hope/
[20-Aug-2009 09:44:59] <rmatte> I think this company we're monitoring has stuff hosted by you guys in the U.S too
[20-Aug-2009 09:45:08] <rmatte> probably
[20-Aug-2009 09:46:25] <jb> "{'firmware': '85P1'}"
[20-Aug-2009 09:46:41] <jb> whats the best way to just get "85P1" out of that?
[20-Aug-2009 09:46:52] <ckrough> as a string or as a tuple?
[20-Aug-2009 09:46:56] <jb> string
[20-Aug-2009 09:47:08] <jb> ... markeriv [n=markeriv@67.103.221.82] has joined #zenoss
[20-Aug-2009 09:47:10] <jb> oops
[20-Aug-2009 09:47:11] <ckrough> http://gskinner.com/RegExr/
[20-Aug-2009 09:47:14] <rmatte> ckrough: ah, it looks like they are actually hosted by one of your partners or something for you guys
[20-Aug-2009 09:47:14] <jb> om.firmware = firmware.get(om.snmpindex)
[20-Aug-2009 09:47:19] <jb> thats om.firmware
[20-Aug-2009 09:47:20] <rmatte> IXEUROPE
[20-Aug-2009 09:47:22] <markeriv> Greetings fellow Zenoss Junkies
[20-Aug-2009 09:47:49] <jb> ckrough: oh nice, ive never seen that
[20-Aug-2009 09:47:57] <twm1010> IX?
[20-Aug-2009 09:48:03] <ckrough> I love that site. I should send him a donation at this point
[20-Aug-2009 09:48:09] <twm1010> markeriv: welcome.
[20-Aug-2009 09:48:11] <_mingau_> hello all people
[20-Aug-2009 09:49:06] <jb> hmmm
[20-Aug-2009 09:49:36] * jb tries to figure it out
[20-Aug-2009 09:49:40] <rmatte> yeh, RegExr is awesome
[20-Aug-2009 09:50:16] <markeriv> Quick question, has anyone else had problems in 2.4.2 with rrd files not being generated?
[20-Aug-2009 09:50:31] <markeriv> I am having to manually go create em O.o
[20-Aug-2009 09:50:52] <chudler_> that sucks. It hasn't happened to me.
[20-Aug-2009 09:50:53] <ckrough> jb: match group 1 in : ^{'[\w]*':\s'([\w]*)'}$
[20-Aug-2009 09:51:11] <twm1010> markeriv: any trouble I had with that, was related to the datasource setup
[20-Aug-2009 09:51:18] <ckrough> jb: my regex foo is ugly thought, it could probably be more elegant, but that should be accurate
[20-Aug-2009 09:51:25] <markeriv> I would typically think it would be a permissions issue, but all but 1 of the rrd files was correctly populating on device install
[20-Aug-2009 09:51:26] <twm1010> for instance I was using incomplete OID's forgetting to put a .1 on the end
[20-Aug-2009 09:51:31] <jb> ckrough: how did you get that?
[20-Aug-2009 09:51:34] <jb> using the site
[20-Aug-2009 09:51:43] <_mingau_> does anybody use freebsd here?
[20-Aug-2009 09:51:52] <ckrough> jb: entered your string in the main window, then toyed with the match line up top
[20-Aug-2009 09:52:13] <rmatte> haha
[20-Aug-2009 09:52:17] <ckrough> jb: enter it then mouse over the regex and it will explain each portion, then mouse over the highlight match it will show the match groups
[20-Aug-2009 09:52:20] <rmatte> he tried to paste something to be a flooded off
[20-Aug-2009 09:52:26] <rmatte> _mingau_: pastebin.com
[20-Aug-2009 09:52:30] <rmatte> then paste the URL
[20-Aug-2009 09:52:31] <rmatte> much easier
[20-Aug-2009 09:52:36] <_mingau_> kkkkkkkkkk
[20-Aug-2009 09:52:49] <markeriv> hmm, i am using the zenpack ZenPacks.BlakeDrager.fping
[20-Aug-2009 09:53:15] <rmatte> wow, my fingers seems to be going faster than my brain this morning, that should have read: "he tried to paste something to me and flooded off"
[20-Aug-2009 09:53:18] <markeriv> so I have not edited any OIDs
[20-Aug-2009 09:53:26] <twm1010> markeriv: i would think then you need to look at debug events or try and execute the command used to collect the fping data manually, to see if it succeeds
[20-Aug-2009 09:53:56] <jb> hmm
[20-Aug-2009 09:54:07] <twm1010> generally if the RRD files is not created it's 1) a permissions issue or 2) the data is not being collected correctly
[20-Aug-2009 09:54:17] <markeriv> hmm
[20-Aug-2009 09:54:26] <markeriv> i got that folder set to -R 777
[20-Aug-2009 09:54:39] <jb> group 1: 85P1
[20-Aug-2009 09:54:49] <twm1010> what im saying, it is more likely to be a collection issue, and not a permissions issue
[20-Aug-2009 09:54:58] <markeriv> hmm ok
[20-Aug-2009 09:55:29] <markeriv> so check the OID of the fping_min.rrd
[20-Aug-2009 09:55:41] <twm1010> fping probably doesn't have an OID, its not an SNMP based datasource
[20-Aug-2009 09:55:42] <markeriv> thats the one that i cannot get to auto populate on install
[20-Aug-2009 09:55:48] <markeriv> doh
[20-Aug-2009 09:55:53] <markeriv> thats the one thats breaking lol
[20-Aug-2009 09:56:11] <markeriv> not having any similar issues with SNMP devices... just fping
[20-Aug-2009 09:56:12] <twm1010> well, did you assign the fping template to any devices?
[20-Aug-2009 09:56:34] <markeriv> aye, 3 at this point. 1 linux server (fedora) and 2 cisco 1721s
[20-Aug-2009 09:56:44] <twm1010> gotcha...
[20-Aug-2009 09:56:53] <twm1010> well.... look at that template, and see what command its trying to run
[20-Aug-2009 09:56:56] <markeriv> all utilization graphs and charts work correctly by default
[20-Aug-2009 09:56:59] <twm1010> login as the zenoss user, and try and run it yourself
[20-Aug-2009 09:56:59] <markeriv> nod
[20-Aug-2009 09:58:11] <markeriv> kk
[20-Aug-2009 10:01:01] <twm1010> let me start up my dev box, i think i installed the fping zenpack there
[20-Aug-2009 10:01:26] <mrayzenoss> mrchippy and I will be taking questions today, he should be here in a second
[20-Aug-2009 10:01:36] <twm1010> mrayzenoss: welcome.
[20-Aug-2009 10:01:42] <straterra> I added a process to monitor.. It's a Windows machine and the executable is JAMES.exe. I added that as the regex, and its running..but Zenoss says it isnt
[20-Aug-2009 10:01:45] <rmatte> oh right, developer day
[20-Aug-2009 10:01:50] <straterra> Oh, cool
[20-Aug-2009 10:01:57] <twm1010> straterra: how much time did you give it to find the process?
[20-Aug-2009 10:02:05] <straterra> About..8 hours
[20-Aug-2009 10:02:16] <twm1010> hrmm, that should be plenty, and SNMP is working fine to the server?
[20-Aug-2009 10:02:29] <straterra> yup
[20-Aug-2009 10:02:43] <markeriv> <twm1010> thanks a bunch for the help btw
[20-Aug-2009 10:02:48] <straterra> I'm sitting next to the machine..I see the process running
[20-Aug-2009 10:02:51] <twm1010> try changing your regex to.... ^JAMES and then remodel the device and see what happens
[20-Aug-2009 10:02:56] <straterra> no .exe?
[20-Aug-2009 10:03:08] <twm1010> give it a shot... see how it goes.
[20-Aug-2009 10:03:11] <straterra> ok
[20-Aug-2009 10:03:14] <_mingau_> people...
[20-Aug-2009 10:03:27] <_mingau_> does anybody use freebsd here?
[20-Aug-2009 10:03:58] <straterra> Indeed
[20-Aug-2009 10:06:05] <twm1010> markeriv: i had a thought, is fping installed? as well as it's corresponding zenpack?
[20-Aug-2009 10:06:13] <markeriv> yup
[20-Aug-2009 10:06:52] <markeriv> after I manually created the file the graphs increment correctly
[20-Aug-2009 10:07:33] <markeriv> manually created it by coping cp -b fping_max.rrd fping_min.rrd
[20-Aug-2009 10:08:40] <twm1010> i've never had to manually create an RRD file
[20-Aug-2009 10:08:53] <markeriv> *jealous*
[20-Aug-2009 10:09:05] <twm1010> well, you shouldn't have to.
[20-Aug-2009 10:09:09] <markeriv> ahh that was worth a good laugh
[20-Aug-2009 10:09:30] <twm1010> the collector has to put the data somewhere, so if its not putting it anywhere, its not collecting it
[20-Aug-2009 10:09:56] <markeriv> hmm, that makes sense...
[20-Aug-2009 10:10:26] <ckrough> I've never had to manually create on either.
[20-Aug-2009 10:11:50] <markeriv> hmm now that is very curious
[20-Aug-2009 10:12:13] <markeriv> when clicking on /Devices/rrdTemplates/b_fping/datasources/fping/datapoints/min
[20-Aug-2009 10:12:17] <markeriv> under the Zenpack
[20-Aug-2009 10:12:35] <markeriv> i get An error was encountered while publishing this resource. The requested resource does not exist.
[20-Aug-2009 10:12:46] <markeriv> so its creation file is missing
[20-Aug-2009 10:13:02] <markeriv> just adding the rrd file did not "fix the graph so to speak"
[20-Aug-2009 10:13:43] <twm1010> don't go under the zenpack
[20-Aug-2009 10:13:47] <twm1010> go to the template itself
[20-Aug-2009 10:14:12] <twm1010> see what command is being run to get the datasource min
[20-Aug-2009 10:14:43] <twm1010> it will probably something like scriptnamehere.py $zmanageIP or something like that
[20-Aug-2009 10:15:30] <markeriv> /usr/sbin/fping -c 3 -sq ${dev/manageIp} 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /' ?
[20-Aug-2009 10:16:41] <markeriv> hmm
[20-Aug-2009 10:16:52] <markeriv> when I use the Test against device feature i get good output
[20-Aug-2009 10:16:58] <rmatte> hmmm, there's no "RDP" IpService by default?
[20-Aug-2009 10:16:59] <rmatte> weird
[20-Aug-2009 10:18:51] <rmatte> hehe, I just found logs of the day I was building the Cisco Mibs and running in to problems
[20-Aug-2009 10:18:51] <rmatte> back in January
[20-Aug-2009 10:20:42] <rmatte> ah, it's MS WBT Server
[20-Aug-2009 10:22:13] <twm1010> rmatte: why would they monitor the port when they can monitor the service itself over WMI?
[20-Aug-2009 10:22:31] <twm1010> perhaps to see if a firewall is stopping things from listening.
[20-Aug-2009 10:22:31] <rmatte> twm1010: because we're not monitoring windows servers using WMI
[20-Aug-2009 10:22:44] <rmatte> and also, we're more interested in connectivity rather than the service
[20-Aug-2009 10:22:47] <twm1010> markeriv: that's a hell of a command
[20-Aug-2009 10:22:56] <rmatte> connectivity could be interrupted, but the service might show as good
[20-Aug-2009 10:23:09] <twm1010> login as zenoss user and execute /usr/sbin/fpin -c 3 -sq TESTIPHERE
[20-Aug-2009 10:23:15] <twm1010> and see if the data comes back
[20-Aug-2009 10:23:26] <twm1010> it looks like the script depends on both fping, and sed to work properly
[20-Aug-2009 10:24:03] <twm1010> well, thats not technically a script its just a massive command
[20-Aug-2009 10:24:22] <jb> 2009-08-20T10:19:13 ERROR RenderServer Failed to generate a graph
[20-Aug-2009 10:24:25] <jb> File "/opt/zenoss/Products/ZenRRD/RenderServer.py", line 190, in render
[20-Aug-2009 10:24:25] <jb> rrdtool.graph(*gopts)
[20-Aug-2009 10:24:26] <jb> hm
[20-Aug-2009 10:24:48] <twm1010> <-- out to lunch, back in an hour
[20-Aug-2009 10:25:03] <markeriv> good output to 4.2.2.1
[20-Aug-2009 10:25:55] <mrchippy> jb, is that the extent of the traceback, or is there more?
[20-Aug-2009 10:26:00] <jb> nothing more
[20-Aug-2009 10:26:04] <markeriv> using that command
[20-Aug-2009 10:26:25] <jb> yes, thats a traceback
[20-Aug-2009 10:26:34] <jb> any ideas?
[20-Aug-2009 10:27:07] <jb> oh.
[20-Aug-2009 10:27:54] <jb> debug saying the OIDs are bad..
[20-Aug-2009 10:33:41] <mrayzenoss> kinda a quiet day
[20-Aug-2009 10:34:07] <ckrough> yup
[20-Aug-2009 10:34:17] <ckrough> I guess all the issues are resolved. Time for beers/
[20-Aug-2009 10:34:22] <jb> erm, wonder why zen is telling me these OID's are bad..
[20-Aug-2009 10:34:22] <jb> .1.3.6.1.4.1.2.6.182.3.7.1.0 = INTEGER: 709
[20-Aug-2009 10:34:38] <rmatte> yeh it is
[20-Aug-2009 10:37:01] <jb> ok, i have a question..
[20-Aug-2009 10:37:19] <jb> I have events (SNMP) traps that come in..
[20-Aug-2009 10:37:42] <jb> with a summary of "Dell OpenManage: Memory Device Failure"
[20-Aug-2009 10:37:52] <jb> or "Dell OpenManage: Cooling device Failure"
[20-Aug-2009 10:38:12] <jb> but when they get emailed, the text gets mangled..
[20-Aug-2009 10:38:15] <jb> example: "Dell OpenManage: Cooling deviceFailure"
[20-Aug-2009 10:38:23] <jb> (no spaces)
[20-Aug-2009 10:38:29] <jb> Dell OpenManage: SystemRedundancy has been lost
[20-Aug-2009 10:38:36] <jb> anybody notice this?
[20-Aug-2009 10:42:09] <mrayzenoss> jb: is your OID message like this? Error reading value for "dhcp0" on planet-express (oid .1.3.6.1.4.1.63.501.3.3.1 is bad)
[20-Aug-2009 10:42:27] <jb> yeah, sorry, I fixed that.
[20-Aug-2009 10:42:31] <jb> it needed a .0
[20-Aug-2009 10:42:44] <mrayzenoss> yeah, that's what I was about to recommend
[20-Aug-2009 10:42:49] <rmatte> yeh, without the .0 you'll still get the output but Zenoss won't like it
[20-Aug-2009 10:43:07] <jb> yeah, i forgot about that
[20-Aug-2009 10:44:03] x-ip is now known as x-ip|launch
[20-Aug-2009 10:44:14] <jb> ah, another graph related question though..
[20-Aug-2009 10:45:50] <jb> i have one graph that is showing "0" for every value
[20-Aug-2009 10:45:57] <_mingau_> testing
[20-Aug-2009 10:46:05] <rmatte> jb: then that's what it's getting from snmp it would seem
[20-Aug-2009 10:46:11] <_mingau_> hello matt
[20-Aug-2009 10:46:13] <rmatte> jb: you sure you are polling the right OIDs?
[20-Aug-2009 10:46:17] <jb> yes, one would seem
[20-Aug-2009 10:46:21] <jb> it just randomly stopped working though
[20-Aug-2009 10:46:26] <mrayzenoss> _mingau_: greetings
[20-Aug-2009 10:46:28] <jb> one would think rather
[20-Aug-2009 10:46:36] <rmatte> jb: might have to restart snmp or something
[20-Aug-2009 10:46:39] <jb> sec.. i had an event generated from it
[20-Aug-2009 10:46:46] <jb> its not all graphs for that host, and the OID's are fine
[20-Aug-2009 10:47:21] <rmatte> jb: I've had it where CPU monitoring just died, but all other monitoring worked fine on a windows host before, so we had to restart snmp on it and then everything was fine
[20-Aug-2009 10:47:33] <magnachef> I have a question on something that's been a pain for a while now. If I have a device with multiple interfaces, that have multiple IPs that are bound to apache, lighttpd, and mysql, there are times when I can discover and add the device with our admin interface and IP, and it'll see the services and their associated IPs, but the IP Services check will fail unless I change it to an IP that's shared between the services
[20-Aug-2009 10:47:38] <jb> k, i'll try that
[20-Aug-2009 10:48:10] <magnachef> sometimes with lighttpd, http, and mysql, this doesn't work, and I need to drop monitoring of the mysql IP Service
[20-Aug-2009 10:48:10] <jb> g$ snmpwalk -v2c -c fcsnmp1ro fc-iftpcln01 1.3.6.1.4.1.8072.1.3.2.3.1.1.14.112.111.115.116.102.105.120.66
[20-Aug-2009 10:48:11] <jb> NET-SNMP-EXTEND-MIB::nsExtendOutput1Line."postfixBounced" = STRING: 253
[20-Aug-2009 10:48:18] <jb> just odd, cause i don't have any "bad OID" events
[20-Aug-2009 10:49:04] <magnachef> is there a way that I can add the device with my admin IP, have it find the various other IPs and their services, then have them test the services via those IPs and not the admin IP?
[20-Aug-2009 10:49:19] <rmatte> jb: you might have to kill your RRD files and let them regenerate
[20-Aug-2009 10:49:33] <jb> yeah, thats what I was thinking
[20-Aug-2009 10:49:40] <rmatte> it's worth a try
[20-Aug-2009 10:50:13] <jb> ok, did that.. we'll just wait a bit now
[20-Aug-2009 10:50:26] <rmatte> yup, 15-20 mins and we should know
[20-Aug-2009 11:06:00] <jb> so can any devs give me tips on how to troubleshoot object classes that wont compile?
[20-Aug-2009 11:11:22] <mrayzenoss> from mrchippy: magnachef: you can click on the service on the device, choose the ip for the service to checked on, then go back to the os tab, check the service, and lock the service.
[20-Aug-2009 11:13:25] <mrayzenoss> mrchippy's battery died, he should be back soon
[20-Aug-2009 11:16:16] <magnachef> let me find one with multiple IPs. So it'll only give me an option for choosing an IP if there's more than one IP?
[20-Aug-2009 11:16:54] <magnachef> hmmm...looks like it doesn't...it just lists the IPs
[20-Aug-2009 11:17:05] <magnachef> I'm still running 2.3.3 if that matters
[20-Aug-2009 11:20:32] <rmatte> does anyone know of a way to install the SNMP service on a windows 2000 box without a windows 2000 cd?
[20-Aug-2009 11:20:54] <jb> make the media available via the network?
[20-Aug-2009 11:21:33] <rmatte> I'm more looking for some download link to a package that I could use instead or something (we're trying to dig up a 2000 cd but haven't been able to find one so far, and I just want to get this done and over with lol)
[20-Aug-2009 11:29:22] <mrchippy> sorry, magnachef. disconnected for a bit there
[20-Aug-2009 11:36:23] <rmatte> mrchippy: recap...
[20-Aug-2009 11:36:24] <rmatte> [12:10pm] <magnachef> let me find one with multiple IPs. So it'll only give me an option for choosing an IP if there's more than one IP?
[20-Aug-2009 11:36:24] <rmatte> [12:10pm] <magnachef> hmmm...looks like it doesn't...it just lists the IPs
[20-Aug-2009 11:36:24] <rmatte> [12:11pm] <magnachef> I'm still running 2.3.3 if that matters
[20-Aug-2009 11:38:17] <mrchippy> thanks, rmatte. yeah, it's not even in 2.4. it has been added and will be available in King Crab.
[20-Aug-2009 11:39:11] <rmatte> yeh, King crab seems to have a lot of great enhancements, I can't wait
[20-Aug-2009 11:39:36] <rmatte> (I'm most looking forward to port status actually reflecting the current status of the port)
[20-Aug-2009 11:40:04] <jb> so can any devs give me tips on how to troubleshoot object classes that wont compile?
[20-Aug-2009 11:40:06] <jb>
[20-Aug-2009 11:41:06] <rmatte> jb: are you using something like eclipse to author your python?
[20-Aug-2009 11:41:32] <jb> no
[20-Aug-2009 11:41:35] <jb> im using vi
[20-Aug-2009 11:41:40] <jb>
[20-Aug-2009 11:41:44] <rmatte> you might try running it through an IDE to clean up any spacing problems first
[20-Aug-2009 11:41:45] <jb> eclipse.. hmm
[20-Aug-2009 11:41:52] <rmatte> python is very very particular about spacing
[20-Aug-2009 11:41:55] <jb> yeah
[20-Aug-2009 11:42:22] <rmatte> that reminds me, I need to install IDLE on here
[20-Aug-2009 11:42:31] <jb> i saw one reccomended on the forums
[20-Aug-2009 11:42:39] <jb> preferrably, i'd like one that i can forward over X11
[20-Aug-2009 11:42:42] <rmatte> idle is the simplest
[20-Aug-2009 11:42:48] <rmatte> and idle can run on X11
[20-Aug-2009 11:43:21] <jb> ah it's packaged with python
[20-Aug-2009 11:43:31] <rmatte> generally
[20-Aug-2009 11:44:30] <rmatte> eclipse is a bit more robust, but idle is good for beginners
[20-Aug-2009 11:44:48] <rmatte> and it's very lightweight
[20-Aug-2009 11:46:15] <rmatte> general rule, indents in python should be 4 spaces
[20-Aug-2009 11:46:21] <rmatte> no tabs
[20-Aug-2009 11:46:27] <jb> yeah
[20-Aug-2009 11:46:35] <rmatte> unless you're using something like zendmd, then it's single spaces
[20-Aug-2009 11:47:07] <jb> ok i'll take a look at idle
[20-Aug-2009 11:47:11] <rmatte> cool
[20-Aug-2009 11:47:53] <straterra> rmatte: I have server 2k disk
[20-Aug-2009 11:48:34] <rmatte> straterra: it's alright, I'll let the admin responsible for the server dig one up
[20-Aug-2009 11:48:41] <straterra> Alright
[20-Aug-2009 11:48:44] <rmatte> we're a microsoft partner so I'm sure we can dig one up somehow
[20-Aug-2009 11:48:50] <rmatte> we'll download it if need be
[20-Aug-2009 11:48:52] <straterra> I have pretty much all Windows disks..so..if you need anything..lemme know
[20-Aug-2009 11:49:00] <rmatte> cool
[20-Aug-2009 11:49:02] <rmatte> thanks
[20-Aug-2009 11:49:04] <straterra> Yup
[20-Aug-2009 11:49:19] <straterra> My work needs an 'admin nap time'
[20-Aug-2009 11:49:56] <rmatte> hehe
[20-Aug-2009 11:53:06] <straterra> I didnt get to sleep until after 6AM..and I was back in here at 9:30 AM
[20-Aug-2009 11:56:39] <rmatte> brutal
[20-Aug-2009 11:57:02] <rmatte> I've pulled a 25 hour shift here when we were setting up a Zenoss box for a large client
[20-Aug-2009 11:57:04] <rmatte> that was fun
[20-Aug-2009 11:57:48] <rmatte> It was our first production Zenoss box ever, 350 devices spread across 3 continents
[20-Aug-2009 11:58:05] <rmatte> and not much went right, or we would have been done in half the time
[20-Aug-2009 11:58:06] <rmatte> lol
[20-Aug-2009 12:01:09] <straterra> I was moving a VM across storage arrays
[20-Aug-2009 12:02:34] <rmatte> ah
[20-Aug-2009 12:02:58] <straterra> It's still going
[20-Aug-2009 12:03:08] <rmatte> brutal
[20-Aug-2009 12:03:14] <rmatte> big vm?
[20-Aug-2009 12:03:19] <straterra> Yeah
[20-Aug-2009 12:03:26] <straterra> I wish we had 10ge :/
[20-Aug-2009 12:03:33] <rmatte> hehe
[20-Aug-2009 12:03:38] <rmatte> yeh, I'd bet that'd help
[20-Aug-2009 12:03:57] <rmatte> too bad the storage arrays aren't linked with fiber or something
[20-Aug-2009 12:04:22] <straterra> I don't think it'd help..the problem is vmware
[20-Aug-2009 12:04:32] <rmatte> ah
[20-Aug-2009 12:04:33] <straterra> It's not even close to maxing out gigabit
[20-Aug-2009 12:04:41] <straterra> ESX is just..transfering very slowly
[20-Aug-2009 12:04:46] <rmatte> crappy
[20-Aug-2009 12:05:45] <straterra> Good news is, my new work phone came in
[20-Aug-2009 12:09:50] <markeriv> anyone here know much about sed?
[20-Aug-2009 12:11:00] <chudler_> I just wish the spam on the zenoss forum could be dealt with once and for all :-(
[20-Aug-2009 12:11:16] <markeriv> yea i saw that phone offer come through the other day lol
[20-Aug-2009 12:11:30] <mrayzenoss> I've been deleting and banning them
[20-Aug-2009 12:11:49] <mrayzenoss> we're switching to a new platform soon, so please be patient
[20-Aug-2009 12:12:06] <chudler_> thats good. They are persistent little twerps
[20-Aug-2009 12:13:05] <markeriv> thx mray
[20-Aug-2009 12:13:16] <mrayzenoss> I wonder if I could outsource some of this, I'll ask the sysadmins if we can crowdsource the ban and delete
[20-Aug-2009 12:13:27] <straterra> Whats the best way to monitor an SSL website on port 8443?
[20-Aug-2009 12:13:29] <mrayzenoss> I figure some of you guys are trustworthy
[20-Aug-2009 12:15:25] <straterra> A new ip service?
[20-Aug-2009 12:17:43] <mrayzenoss> so anytime you see spam, just PM me and I'll get it
[20-Aug-2009 12:18:24] <straterra> Erhm..I created a new IP service..but can't find where to set the port o.O
[20-Aug-2009 12:18:57] <straterra> oh..nevermind heh
[20-Aug-2009 12:24:39] <erin> Is it cool for a a newbie zenoss to ask a question about a zenpack here? and yea I've searched the forums, RTFM & googled it all...
[20-Aug-2009 12:24:48] <erin> I'm a DBA & not the sys admin
[20-Aug-2009 12:25:00] <erin> thought I speak unix not win
[20-Aug-2009 12:25:53] <mrayzenoss> yup, this is a place of Q&A
[20-Aug-2009 12:27:44] <erin> Ok the zenpack was installed by my sys admin. I find it in zenoss. I bind it to my device (mysql server). I configure it with mysql user/password. I even test the user/pwd connection from the zenoss server.
[20-Aug-2009 12:28:03] <erin> I see the MySQL - Select Statistics title & a broken image ...
[20-Aug-2009 12:28:23] <erin> I search error logs & there's no error about a bad MySQL connection ...
[20-Aug-2009 12:28:34] <erin> I might be looking in the wrong error log tho....
[20-Aug-2009 12:28:48] <erin> how to get a real graph?
[20-Aug-2009 12:30:31] <mrayzenoss> how long has it been monitoring?
[20-Aug-2009 12:30:44] <mrayzenoss> which MySQL ZenPack is installed?
[20-Aug-2009 12:31:03] <scp2004> Hi Guys.. I need some help ...
[20-Aug-2009 12:31:15] <erin> I had the default MySQLMonitoring ZenPack installed.
[20-Aug-2009 12:31:24] <erin> It's been monitoring for about a week ...
[20-Aug-2009 12:31:36] <erin> long enough for there to be data ...
[20-Aug-2009 12:31:38] <mrayzenoss> and no graphs for anything on it?
[20-Aug-2009 12:31:54] <erin> well none on the MySQL zenpack.
[20-Aug-2009 12:31:55] <scp2004> I need to import a Mikrotik Mib to Zenoss 2.4.2 . How ? I m using CentOS 5
[20-Aug-2009 12:32:07] <scp2004> [mrayzenoss] I need to import a Mikrotik Mib to Zenoss 2.4.2 . How ? I m using CentOS 5
[20-Aug-2009 12:32:30] <erin> I've got graphs on IO, Mem, CPU, Load, NumTasksGraph
[20-Aug-2009 12:32:43] <erin> they were all there before I had the mysql zenpack installed.
[20-Aug-2009 12:32:54] <mrayzenoss> erin: double check the mysql settings: http://www.zenoss.com/community/docs/zenoss-extended-monitoring/2.4.2/ch11s03.html
[20-Aug-2009 12:33:42] <erin> ok. I'll be back in a second
[20-Aug-2009 12:33:47] <erin> ok more like 5 min
[20-Aug-2009 12:34:19] <mrayzenoss> scp2004: why don't you install the mib-browser ZenPack, it makes it easy to load MIBs from the UI
[20-Aug-2009 12:34:28] <mrayzenoss> http://www.zenoss.com/community/projects/zenpacks/mib-browser
[20-Aug-2009 12:36:15] <scp2004> [mrayzenoss] thanks .. But i need to import just one mib .. I think its more easy i do this for a command like zenmibload, not ?
[20-Aug-2009 12:36:16] <erin> ok this is a stupid question but the UI sux big time here. WTF is the "page menu"??? where I'm gonna select more--> templates?
[20-Aug-2009 12:36:58] <mrayzenoss> scp2004: sure, the zenpack just makes it simple
[20-Aug-2009 12:37:12] <mrayzenoss> scp2004: look at the command zenmib
[20-Aug-2009 12:37:38] <mrayzenoss> erin: yeah, it's non-intuitive at times
[20-Aug-2009 12:37:56] <erin> at times? gee I wish zenoss would hire a UX person.
[20-Aug-2009 12:38:17] <scp2004> [mrayzenoss] im reading about this . Just dont know where put the mib file to import ... The guide says /$ZENHOME/Share.. Dont have this folder in zenoss Centos Installation
[20-Aug-2009 12:38:22] <erin> I'm lost wheres the page menu?
[20-Aug-2009 12:38:25] <mrayzenoss> erin: we have and the UI revamp is in-progress
[20-Aug-2009 12:38:35] <erin> yea!
[20-Aug-2009 12:38:36] <cgibbons> it's that little drop-down triangle you see on the upper left of a table
[20-Aug-2009 12:38:37] <mrayzenoss> erin: it's the down-arrow to the left of "Status"
[20-Aug-2009 12:38:48] <cgibbons> also known as Zenoss New User Question #2
[20-Aug-2009 12:39:03] <cgibbons> horrible horrible menu
[20-Aug-2009 12:39:11] <mrayzenoss> yeah, you should fix it
[20-Aug-2009 12:39:17] <ckrough> zing
[20-Aug-2009 12:39:32] <erin> uh what place? dashboard?
[20-Aug-2009 12:39:41] <erin> at the page for the device?
[20-Aug-2009 12:40:07] <mrayzenoss> erin: which template are you trying to change? The one for the device class or the actual device?
[20-Aug-2009 12:40:29] <erin> I'm following the instructions on getting MySQLMonitor to work
[20-Aug-2009 12:40:49] <erin> item #2 under 3.2 Zenoss: From the page menu, select More → zProperties.
[20-Aug-2009 12:41:10] <erin> shite found it.
[20-Aug-2009 12:41:16] <erin> damn you guys this ui sux
[20-Aug-2009 12:41:56] <mrayzenoss> erin: may I suggest http://downloads.sourceforge.net/zenoss/Getting_Started_with_Zenoss.pdf
[20-Aug-2009 12:42:34] <erin> I already did. ANd I've already followed the instructions but my graphs are broken images.
[20-Aug-2009 12:43:07] <rhett> there are plenty of worse ui's
[20-Aug-2009 12:43:30] <rhett> its not bad. especially after you've used it
[20-Aug-2009 12:44:19] <erin> yea I know but my frustration is with this product.
[20-Aug-2009 12:44:38] <erin> so I see MySQL in the templates that have been "bind to device".
[20-Aug-2009 12:44:52] <erin> and the headers for these are showing up with broken graphs.
[20-Aug-2009 12:45:24] <scp2004> [mrayzenoss] look : Download and copy these files into /opt/zenoss/share/mibs/site/ .. dont have this folder in my zenoss centOS installation
[20-Aug-2009 12:45:41] <erin> I really can follow directions. I was very excited to see the title & area for the graphs when I first did this last thurs.
[20-Aug-2009 12:46:34] <erin> but alas I still have broken graphs. suggestion of error log to dig into?
[20-Aug-2009 12:46:42] <rhett> well if there are broken graphs it maybe the datasources are not there or do not have data
[20-Aug-2009 12:46:43] <erin> could be how the zenpack was installed?
[20-Aug-2009 12:46:54] <rhett> which could be a permissions error
[20-Aug-2009 12:47:09] <erin> nope -- db has data & is there. I can connect from the zenoss server to the db using the zenoss user/pwd
[20-Aug-2009 12:47:15] <erin> that part was a walk in the park!
[20-Aug-2009 12:47:40] <cgibbons> erin: if you want to get a taste of what the new UI will look like, http://public-demo.zenoss.com/
[20-Aug-2009 12:47:57] <rhett> no events under the host then i assume
[20-Aug-2009 12:48:23] <erin> which log?
[20-Aug-2009 12:48:55] <erin> OH! event.log!
[20-Aug-2009 12:49:06] <rhett> the log on the box you're trying to get the stats from
[20-Aug-2009 12:49:11] <erin> I'm getting this error (not sure if it's my graph but prolly is!)
[20-Aug-2009 12:49:22] <erin> I'm on zenoss server & in the log dir.
[20-Aug-2009 12:49:26] <rhett> ok
[20-Aug-2009 12:49:33] <erin> here's the error:
[20-Aug-2009 12:49:34] <erin> 2009-08-20T10:50:28 ERROR RenderServer Failed to generate a graph
[20-Aug-2009 12:49:34] <erin> Traceback (most recent call last):
[20-Aug-2009 12:49:34] <erin> File "/usr/local/zenoss/zenoss/Products/ZenRRD/RenderServer.py", line 190, in render
[20-Aug-2009 12:49:35] <erin> rrdtool.graph(*gopts)
[20-Aug-2009 12:50:09] <rhett> ok well thats for sure a graph error
[20-Aug-2009 12:50:25] <erin> how to fix?
[20-Aug-2009 12:50:39] <rhett> are you running zenoss or zenoss-stack?
[20-Aug-2009 12:50:43] <erin> I wonder why I'm getting other graphs?? is it the zenpack for MySQLMonitor
[20-Aug-2009 12:50:44] <rhett> i assume zenoss
[20-Aug-2009 12:50:54] <rhett> hm i dont know
[20-Aug-2009 12:51:06] <rhett> ive never run across this personally
[20-Aug-2009 12:51:28] <erin> running ubuntu so we've got teh zenoss-stack
[20-Aug-2009 12:51:43] <erin> I hate ubuntu nothing goes smoothly!
[20-Aug-2009 12:51:50] <erin>
[20-Aug-2009 12:51:51] <rhett> do other graphs work for the same host?
[20-Aug-2009 12:51:55] <erin> yup
[20-Aug-2009 12:52:05] <erin> the default cpu, mem, blah all do
[20-Aug-2009 12:52:08] <rhett> ok so its for sure only the mysql stuff
[20-Aug-2009 12:52:18] <erin> yup
[20-Aug-2009 12:52:31] <erin> and no I didn't install the mysql stuff. I'm the "end user".
[20-Aug-2009 12:53:46] <erin> should I have the zenpack de-installed & re-installed?
[20-Aug-2009 12:53:54] <erin> or is there an ubuntu zenpack for MySQL?
[20-Aug-2009 12:54:03] <scp2004> [mrayzenoss] I dont have perf graphs ! I need it .. Its cause i dont have the Mikrotik Mib installed ?
[20-Aug-2009 12:54:22] <rhett> i was curious if that had been done or if they tried to remove the zenpack from the host and recreate the graphs
[20-Aug-2009 12:54:34] <rhett> that would probably be the first step erin
[20-Aug-2009 12:54:40] <rhett> do you know if any other mysql graphs work?
[20-Aug-2009 12:54:49] <scp2004> [rmatte] I dont have perf graphs ! I need it .. Its cause i dont have the Mikrotik Mib installed ?
[20-Aug-2009 12:55:09] <scp2004> [straterra] I dont have perf graphs ! I need it .. Its cause i dont have the Mikrotik Mib installed ?
[20-Aug-2009 12:55:23] <erin> HAHA! on the re-install. They find the ui & setup as daunting. It's my test!
[20-Aug-2009 12:55:33] <erin> I'll get it re-installed & check back
[20-Aug-2009 12:55:37] <rhett> cool
[20-Aug-2009 12:56:06] <rhett> scp2004 install the mib isnt too difficult. As stated earlier look at the zenmib command
[20-Aug-2009 12:56:28] <markeriv> hey rhett
[20-Aug-2009 12:56:34] <markeriv> what type of perf graphs you looking for?
[20-Aug-2009 12:56:38] <markeriv> and are you monitoring via SNMP?
[20-Aug-2009 12:56:42] <rhett> i.e. zenmib run ~/mikrotik.mib
[20-Aug-2009 12:56:43] <mrayzenoss> scp2004: trac-zenpacks/wiki/CommunityFAQGeneral#HowdoesZenossuseMIBs
[20-Aug-2009 12:56:48] <markeriv> err scp2004 i mean
[20-Aug-2009 12:56:53] <scp2004> [rhett] Yes ..I just installed the Mikrotik mib few minutes ago .. Its ok now .. But still without graphs .. I need to wait next traps ?
[20-Aug-2009 12:57:04] <mrayzenoss> scp2004: you won't use the MIB to monitor the device
[20-Aug-2009 12:57:13] <rhett> you need to create templates for that mib
[20-Aug-2009 12:57:25] <mrayzenoss> or really just from the OIDs of interest
[20-Aug-2009 12:57:25] <scp2004> [mrayzenoss] so ..why i dont have perf graphs ?
[20-Aug-2009 12:58:26] <mrayzenoss> is your Mikrotik device a router? Do you have it under /Devices/Network/Router?
[20-Aug-2009 12:58:38] <scp2004> yep
[20-Aug-2009 12:59:13] <scp2004> [mrayzenoss] I created the class /Devices/Network/Router/Mikrotik ..
[20-Aug-2009 12:59:27] <scp2004> [mrayzenoss] And put it them
[20-Aug-2009 13:00:32] <mrayzenoss> scp2004: and you have SNMP enabled and monitoring?
[20-Aug-2009 13:00:45] <scp2004> [mrayzenoss] yes .. snmp its ok ..
[20-Aug-2009 13:01:03] <scp2004> [mrayzenoss] are collecting ok .. Only dont have graphs
[20-Aug-2009 13:01:15] <mrayzenoss> broken graphs or nothing?
[20-Aug-2009 13:01:29] <Belial_> Hi there, Not sure if distributed collectors on core are supported, but have been following a few guides without much sucess. I was wondering how to point things to the zope server (In this paste, pthouzenoss is a CNAME for pthouap004 and is the 2nd collector & ptabzzenoss is the one with zope etc running) http://pastebin.com/d2a4d9345 I though changing the configuration files as pasted would point it, along with the zope.conf change, but no sucess
[20-Aug-2009 13:02:03] <mrayzenoss> scp2004: just checked, there are no perf graphs defined in /Devices/Network/Router
[20-Aug-2009 13:02:13] <Belial_> If anyone could point me in a direction it would be most appriciated!
[20-Aug-2009 13:02:41] <mrayzenoss> scp2004: at least not under the "Perf" tab
[20-Aug-2009 13:02:54] <mrayzenoss> scp2004: there are some if you go to the interfaces
[20-Aug-2009 13:03:19] <mrayzenoss> Belial_: distributed collectors aren't supported in Core, but there are a few community docs for implementing them
[20-Aug-2009 13:03:33] <mrayzenoss> Belial_: http://delicious.com/mray/collectors/
[20-Aug-2009 13:03:33] <scp2004> [mrayzenoss] so i need to habilite this ? where ? in /Devices /Network /Router /Mikrotik zproperties ?
[20-Aug-2009 13:03:53] <mrayzenoss> scp2004: on the OS tab, do you have interfaces listed?
[20-Aug-2009 13:04:02] <scp2004> [mrayzenoss] yes
[20-Aug-2009 13:04:26] <mrayzenoss> click on one, like "wlan0" or whatever
[20-Aug-2009 13:04:47] <Belial_> Thanks mrayzenoss, I've been over those links, but guess a 13th read wont hurt Grats on the 1mil DL's and thanks for a great product!
[20-Aug-2009 13:05:10] <mrayzenoss> scp2004: you should see Throughput, Packets and Errors
[20-Aug-2009 13:05:34] <scp2004> [mrayzenoss] oh ..i can see the graphs ..
[20-Aug-2009 13:06:14] <mrayzenoss> so that's what you get with the vanilla Router device class
[20-Aug-2009 13:06:20] <rmatte> oh you guys actually hit mil?
[20-Aug-2009 13:06:27] <mrayzenoss> rmatte: yup
[20-Aug-2009 13:06:40] <scp2004> [mrayzenoss] but when i click just in /device perf i cant see anything
[20-Aug-2009 13:06:43] <mrayzenoss> rmatte: and put your name in the press release
[20-Aug-2009 13:06:46] <rmatte> nice, congrats
[20-Aug-2009 13:06:51] <rmatte> haha, thanks
[20-Aug-2009 13:07:01] <rmatte> is the press release on blog?
[20-Aug-2009 13:07:18] <mrayzenoss> rmatte: http://www.zenoss.com/news/archive/zenosspressrelease.2009-08-20.6091752351
[20-Aug-2009 13:07:55] <rmatte> thanks
[20-Aug-2009 13:07:56] <mrayzenoss> scp2004: there won't be anything on the perf tab unless you define some custom data sources for them
[20-Aug-2009 13:08:46] <scp2004> [mrayzenoss] please .how i do this ?
[20-Aug-2009 13:12:11] <erin> Ok during the remove of the zenpack we got errors! it was installed as ROOT & not zenoss. we reinstalled as zenoss.
[20-Aug-2009 13:12:36] <erin> and now that I've got the binding, setup all memorized it's ready to go.
[20-Aug-2009 13:14:10] <mrayzenoss> oops, something crashed the matrix
[20-Aug-2009 13:14:51] <mrayzenoss> scp2004 seems to have disappeared
[20-Aug-2009 13:15:03] <mrayzenoss> didn't want my link to the Admin Guide I guess
[20-Aug-2009 13:15:08] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch06s02.html
[20-Aug-2009 13:15:32] <rmatte> hehe
[20-Aug-2009 13:15:50] <rhett> he'll be back
[20-Aug-2009 13:16:02] <mrayzenoss> if there's anything that's gotten substantially better in the last year, it's the docs
[20-Aug-2009 13:19:51] <rmatte> yeh, the admin guide is definitely a worthwhile read for anyone starting out with Zenoss
[20-Aug-2009 13:34:05] <Belial_> Its a well thumbed through printout on my desk
[20-Aug-2009 13:34:20] <erin> Ok event.log says this:
[20-Aug-2009 13:34:31] <erin> 2009-08-20T11:10:56 ERROR zen.Plugins Could not load plugins from ZenPacks. One of the ZenPacks is missing or broken.
[20-Aug-2009 13:34:43] <erin> The zenpack was installed as zenoss & not root.
[20-Aug-2009 13:34:55] <rmatte> erin: which pack?
[20-Aug-2009 13:34:58] <erin> ZenPacks.zenoss.MySqlMonitor-2.0.0-py2.4.egg
[20-Aug-2009 13:35:08] <rmatte> hmmm
[20-Aug-2009 13:35:09] <erin> is this the correct version?
[20-Aug-2009 13:35:13] <erin> weird huh?
[20-Aug-2009 13:35:18] <rmatte> if it's the latest version then it's the correct one
[20-Aug-2009 13:35:21] <mrayzenoss> did it install?
[20-Aug-2009 13:35:24] <erin> yea
[20-Aug-2009 13:35:33] <rmatte> oh wait, did you restart Zenoss after the install?
[20-Aug-2009 13:35:40] <erin> oh no?
[20-Aug-2009 13:35:43] <rmatte> do that
[20-Aug-2009 13:35:49] <rmatte> some packs don't play nice until after a restart
[20-Aug-2009 13:35:53] <erin> how do I restart Zenoss & need to do it as zenoss?
[20-Aug-2009 13:36:05] <rmatte> as zenoss user just do "zenoss restart"
[20-Aug-2009 13:36:18] <rmatte> then do "zenoss status" to make sure all the daemons are running after the restart
[20-Aug-2009 13:36:23] <rmatte> if not then do zenoss start
[20-Aug-2009 13:36:46] <rmatte> I find that doing zenoss stop, waiting a few seconds, then doing zenoss start tends to work better
[20-Aug-2009 13:36:59] <erin> well now you tell me!
[20-Aug-2009 13:37:01] <erin>
[20-Aug-2009 13:37:04] <rmatte> hehe
[20-Aug-2009 13:38:23] <erin> yea yr right zenjobs not running & zenactions not running did not start
[20-Aug-2009 13:38:36] <ckrough> I do the same, I also add in the step of killing python processes that wont die after the stop
[20-Aug-2009 13:38:45] <erin> mysql used to be that way so I always do a stop/start
[20-Aug-2009 13:39:22] x-ip|launch is now known as x-ip
[20-Aug-2009 13:39:26] <rmatte> erin: yeh, I might just log a trac ticket to suggest that some sort of delay be put in to the restart command
[20-Aug-2009 13:39:36] <rmatte> so that it waits like 3 seconds before starting it back up
[20-Aug-2009 13:40:07] <erin> got this warning:
[20-Aug-2009 13:40:09] <erin> 2009-08-20T11:40:17 WARNING OFS.Application Duplicate Product name
[20-Aug-2009 13:40:09] <erin> After loading Product 'Five' from '/usr/local/zenoss/zenoss/Products',
[20-Aug-2009 13:40:10] <erin> I skipped the one in '/usr/local/zenoss/zenoss/lib/python/Products'.
[20-Aug-2009 13:40:15] <rmatte> that's fine
[20-Aug-2009 13:40:24] <rmatte> I get that too after upgrade
[20-Aug-2009 13:40:31] <rmatte> doesn't hurt anything
[20-Aug-2009 13:40:34] <erin> ok so now that I've done the restart do I have to go thru the rebind process again?
[20-Aug-2009 13:40:47] <rmatte> nah, hopefully that should fix it
[20-Aug-2009 13:40:56] <erin> yea!
[20-Aug-2009 13:41:07] <rmatte> some zenpacks show as broken until a restart
[20-Aug-2009 13:41:26] <erin> oh what does that mean then? gotta wait the 15 min or?
[20-Aug-2009 13:41:35] <rmatte> yeh
[20-Aug-2009 13:41:38] <rmatte> do you see graphs?
[20-Aug-2009 13:42:00] <erin> ah HA! it works!!!
[20-Aug-2009 13:42:02] <erin> yippee.
[20-Aug-2009 13:42:04] <rmatte> perfect
[20-Aug-2009 13:42:31] <erin> now I wonder if I can get it to monitor my infobright server? hmmm I'll let you know!
[20-Aug-2009 13:42:49] <erin> infobright is a datawarehouse built on mysql
[20-Aug-2009 13:42:49] <rmatte> cool
[20-Aug-2009 13:43:07] <rmatte> ah, opensource too
[20-Aug-2009 13:43:08] <rmatte> nice
[20-Aug-2009 13:43:38] <erin> oh wait? whats this error mean?
[20-Aug-2009 13:43:39] <erin> 2009-08-20T11:43:31 INFO ZODB.Conflict database conflict error (oid 0x2c, class Products.ZenUtils.PObjectCache.PObjectCache) at /zport/RenderServer/render (5 conflicts, of which 0 were unresolved, since startup at Thu Aug 20 11:40:21 2009)
[20-Aug-2009 13:43:52] <rmatte> lovely
[20-Aug-2009 13:43:59] <rmatte> zope database conflicts, those can get messy
[20-Aug-2009 13:44:06] <erin> whats that mean?
[20-Aug-2009 13:44:12] <rhett> ewe. thats gross
[20-Aug-2009 13:44:17] <rmatte> first off stop zenoss and tar up the $ZENHOME directory to back it up
[20-Aug-2009 13:44:22] <rmatte> and throw it somewhere else on the drive
[20-Aug-2009 13:44:24] <ckrough> Those can be safely ignored, so I've been told
[20-Aug-2009 13:44:32] <rmatte> oh?
[20-Aug-2009 13:44:37] <ckrough> "0 were unresolved", meaning all good
[20-Aug-2009 13:44:40] <rmatte> I thought you had to attempt a reindex to fix them?
[20-Aug-2009 13:44:43] <rmatte> oh right
[20-Aug-2009 13:44:49] <rmatte> it must have auto-corrected them
[20-Aug-2009 13:44:54] <rmatte> good stuff then
[20-Aug-2009 13:44:58] <erin> ok it was labeled info & not error. whew
[20-Aug-2009 13:45:02] <ckrough> I was told by someone at zenoss (chet?) that they are more informational than actual errors
[20-Aug-2009 13:45:14] <rmatte> yeh, seems that way
[20-Aug-2009 13:45:36] <erin> I'll go back to lurking until my mysql test is over & then flip to infobright & see if zenoss monitors it.
[20-Aug-2009 13:45:52] <erin> then I'll let you all know (as well as infobright!)
[20-Aug-2009 13:53:10] <chudler_> :Q
[20-Aug-2009 13:53:16] <chudler_> wrong window -- sorry!
[20-Aug-2009 13:53:28] <chudler_> plus a typo! bad day
[20-Aug-2009 13:53:40] <ckrough> yay vi
[20-Aug-2009 13:53:53] <ckrough> and the bad caps...
[20-Aug-2009 13:55:52] <erin> well getting data in the mysql graphs! yippee so its really working.
[20-Aug-2009 13:55:57] <erin> still have to test infobright
[20-Aug-2009 13:56:34] <rhett> cool
[20-Aug-2009 13:57:05] <rhett> i have a tendency to check the datasources for data even if i dont have graphs appearing
[20-Aug-2009 13:57:30] <rhett> just to try and troubleshoot where the problem may actually be
[20-Aug-2009 14:15:21] <markeriv> hey guys
[20-Aug-2009 14:15:29] <markeriv> i am still poking around a problem that I mentioned earlier
[20-Aug-2009 14:15:40] <markeriv> and was hoping now that a few different heads are in here that we might make some headway
[20-Aug-2009 14:15:46] <markeriv> i am running ver 2.4.2
[20-Aug-2009 14:15:57] <markeriv> and having problems with Zenpacks
[20-Aug-2009 14:16:14] <markeriv> I have tried multiple zenpacks and seem to have the issue with several types
[20-Aug-2009 14:16:24] <markeriv> where data is not collecting and I have missing rrd files
[20-Aug-2009 14:16:38] <markeriv> I have gone in and manually added the files, but the data still is not collecting.
[20-Aug-2009 14:16:56] <markeriv> when I go into the zenpack template and run the test it appears to give good #s.
[20-Aug-2009 14:16:58] <erin> markeriv -- you installed as zenoss & did a zenoss stop, zenoss start?
[20-Aug-2009 14:17:22] <erin> I had a zenpack installed as root so I got no graphs
[20-Aug-2009 14:17:38] <erin> for that pack only.
[20-Aug-2009 14:17:41] <markeriv> the zenpack? I believe so, I have checked and all zenpack folders and items are owned by zenoss
[20-Aug-2009 14:17:48] <erin> the zenpacks installed as zenoss worked.
[20-Aug-2009 14:17:51] <erin> ok
[20-Aug-2009 14:17:59] <erin> then you stopped & started the zenoss?
[20-Aug-2009 14:18:05] <markeriv> will give that a go here
[20-Aug-2009 14:18:17] <markeriv> have rebooted the machine a few times tho
[20-Aug-2009 14:18:19] <markeriv> would that not do it?
[20-Aug-2009 14:18:27] <erin> dunno I'm a newbie!
[20-Aug-2009 14:18:37] <erin> thats just what I did & it fixed my problem
[20-Aug-2009 14:18:53] <markeriv>
[20-Aug-2009 14:19:00] <markeriv> i am a newb too
[20-Aug-2009 14:20:59] <markeriv> still getting nan's
[20-Aug-2009 14:21:13] <erin> check the event.log what's it complaining about?
[20-Aug-2009 14:21:14] <rmatte> markeriv: it takes 3 polling cycles for the nans to dissapear
[20-Aug-2009 14:21:20] <rmatte> so like 15 mins
[20-Aug-2009 14:21:25] <markeriv> ahh
[20-Aug-2009 14:21:27] <markeriv> kewl
[20-Aug-2009 14:21:32] <markeriv> so now its time to chill and dance
[20-Aug-2009 14:21:35] <rmatte> so even if it is working, you won't know for a bit
[20-Aug-2009 14:21:39] <erin> yup
[20-Aug-2009 14:21:54] <rmatte> seeing "nan" is actually a good thing because it means that the RRD files have been created
[20-Aug-2009 14:22:04] <rhett> another reason why i fetch the datasources
[20-Aug-2009 14:22:09] <markeriv> well, i went and manually made an rdd file
[20-Aug-2009 14:22:13] <markeriv> with the proper name
[20-Aug-2009 14:22:16] <markeriv> and that problem went away
[20-Aug-2009 14:22:16] <rmatte> oh
[20-Aug-2009 14:22:26] <rmatte> I've never had to manually make one
[20-Aug-2009 14:22:37] <markeriv> yea, from what I have heard mine's a unique experience
[20-Aug-2009 14:22:43] <rmatte> apparently
[20-Aug-2009 14:22:48] <erin> haha! don't you just love those!
[20-Aug-2009 14:22:53] <markeriv> for some reason I always end up doing things the hard way
[20-Aug-2009 14:23:01] <markeriv> you shoulda seen me when I was installing gentoo last year
[20-Aug-2009 14:23:06] <erin> not me if I can help it.
[20-Aug-2009 14:23:08] <rmatte> haha
[20-Aug-2009 14:23:14] <rmatte> gentoo is always a fun learning experience
[20-Aug-2009 14:23:16] <erin> I like centos/redhat myself
[20-Aug-2009 14:23:21] <erin> suse is the worst!
[20-Aug-2009 14:23:25] <rmatte> hey, at least last year the gentoo installer was actually around
[20-Aug-2009 14:23:31] <markeriv> my mboard had a chip that was specifically not supported by the (at the time, latest linux version)
[20-Aug-2009 14:23:31] <rmatte> the first time I installed it there was no installer
[20-Aug-2009 14:23:36] <rmatte> you did the whole installation by hand
[20-Aug-2009 14:23:42] <rmatte> now THAT was fun
[20-Aug-2009 14:23:43] <rmatte>
[20-Aug-2009 14:23:44] <erin> ubuntu is a pain for mysql dbs
[20-Aug-2009 14:23:45] <markeriv> but I found a patch for a different linux version
[20-Aug-2009 14:23:58] <markeriv> so i took the patch apart
[20-Aug-2009 14:24:05] <markeriv> figured out which part of the kernel it was patching
[20-Aug-2009 14:24:09] <markeriv> and went in and did it manually
[20-Aug-2009 14:24:11] <markeriv> recompiled
[20-Aug-2009 14:24:16] <markeriv> and was able to listen to sound
[20-Aug-2009 14:24:17] <rmatte> you probably had an nforce chip?
[20-Aug-2009 14:24:19] <markeriv> aye
[20-Aug-2009 14:24:36] <markeriv> and then last weekend, i plugged in fedora 11...
[20-Aug-2009 14:24:43] <markeriv> boom everything worked
[20-Aug-2009 14:24:44] <rmatte> (they don't play nice with linux, or never used to anyways, the latest kernel versions have been great since it looks like they finally added support for them)
[20-Aug-2009 14:24:46] <erin> sounds like I'm surrounded by sys admin geeks!
[20-Aug-2009 14:24:55] <rmatte> I upgraded to ubuntu 9.04 and my system stopped whining
[20-Aug-2009 14:25:01] <markeriv> meh, i am network engineer who likes to tinker
[20-Aug-2009 14:25:15] <markeriv> yea ubuntu is where I jacked the patch for when i fixed the sound for gentoo
[20-Aug-2009 14:25:15] <rmatte> hehe
[20-Aug-2009 14:25:20] <erin> aah I'm a dba who did a side trip as a sys admin in the 90s.
[20-Aug-2009 14:25:22] <rmatte> yeh
[20-Aug-2009 14:25:29] <rmatte> I had that sound problem too
[20-Aug-2009 14:25:40] <jb> gentoo?
[20-Aug-2009 14:25:40] <jb> yuck.
[20-Aug-2009 14:25:50] <rmatte> I'm running gentoo on my home server
[20-Aug-2009 14:25:53] <markeriv> its actually fairly nice once you get used to it
[20-Aug-2009 14:25:57] <markeriv> i still have it
[20-Aug-2009 14:26:05] <rhett> i do everything
[20-Aug-2009 14:26:29] <rmatte> markeriv: was your mobo a M3N-HT Deluxe/Mempipe per chance?
[20-Aug-2009 14:26:35] <rhett> buntu variants or centos for me
[20-Aug-2009 14:26:42] <rmatte> (Asus)
[20-Aug-2009 14:27:08] <jb> im a rhel/cent guy
[20-Aug-2009 14:27:47] <rmatte> I have the M3N-HT and it did not play nice with Linux at all until Ubuntu 9.04
[20-Aug-2009 14:28:07] <rmatte> sound corruption issues (I actually went out and bought a soundcard because of it)
[20-Aug-2009 14:28:32] <erin> Why not use a mac for yr desktop & centos for dev?
[20-Aug-2009 14:28:33] <rmatte> I couldn't boot the system in multicore mode since it wouldn't boot and even if it did it would continuously beep from the PC speaker
[20-Aug-2009 14:28:53] <rmatte> fun times
[20-Aug-2009 14:29:15] <rmatte> plus the bios version that came with the board initially would only allow the system to boot right after the CMOS was reset
[20-Aug-2009 14:29:23] <rmatte> so I finally upgraded the BIOS to fix that problem
[20-Aug-2009 14:29:27] <rmatte> lol
[20-Aug-2009 14:29:34] <chudler_> I had a system like that once. Was an ABIT board
[20-Aug-2009 14:30:15] <chudler_> BIOS self-corrupted right after boot, so it also required a reset on each boot.
[20-Aug-2009 14:30:48] <rmatte> lovely
[20-Aug-2009 14:31:21] <erin> zenoss question? does the reset button refresh the graphs?
[20-Aug-2009 14:31:53] <scp2004> Someone to help me how i can see the perf graphs on devices ?
[20-Aug-2009 14:32:03] <magnachef> resets time frames, zooming, etc
[20-Aug-2009 14:32:06] <rmatte> erin: it refreshes your view of the graphs
[20-Aug-2009 14:32:13] <erin> cool!
[20-Aug-2009 14:32:14] <jb> rmatte: hm, looks like restarting snmp and deleting the RRD's did nothing :/
[20-Aug-2009 14:32:17] <scp2004> Snmp are running ok ...bur i cant see the graphs in perf
[20-Aug-2009 14:32:34] <mrayzenoss> everybody be nice to Gentoo, rocket worked on the installer
[20-Aug-2009 14:32:38] <erin> ok zenoss's MySQL Monitor is monitoring my Infobright instance!!!! yippeee!
[20-Aug-2009 14:32:38] <rmatte> jb: I'm stumped then, you're saying that when you walk the OIDs it is giving you the correct data right?
[20-Aug-2009 14:32:50] <jb> rmatte: yep
[20-Aug-2009 14:32:58] <mrayzenoss> scp2004: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch06s02.html
[20-Aug-2009 14:33:11] <rmatte> mrayzenoss: really? I'll have to give him kudos for that next time I see him, it's a nice installer
[20-Aug-2009 14:33:35] <rmatte> mrayzenoss: and having previously installed gentoo completely by hand, I'm quite appreciative of it
[20-Aug-2009 14:33:36] <scp2004> [mrayzenoss] One more time thank you vm man ..sr ... ..
[20-Aug-2009 14:34:28] <mrayzenoss> yeah, he was rocket@gentoo.org for awhile (I think he's no longer active)
[20-Aug-2009 14:37:19] <markeriv> rmatte no it wasnt, was an ASUS nforc
[20-Aug-2009 14:37:20] <markeriv> nforce
[20-Aug-2009 14:37:28] <rmatte> nforce is the chipset
[20-Aug-2009 14:37:32] <rmatte> I meant the model of the board
[20-Aug-2009 14:37:36] <markeriv> ahh
[20-Aug-2009 14:37:39] <markeriv> hmm
[20-Aug-2009 14:37:47] <markeriv> how can i check my mboard ver in linux?
[20-Aug-2009 14:37:50] <markeriv> its the same one
[20-Aug-2009 14:37:58] <rmatte> does it looks like: http://computersrus.co.za/catalog/images/M3N-HT%20DeluxeMempipe.JPG
[20-Aug-2009 14:37:59] <rmatte> ?
[20-Aug-2009 14:38:13] <markeriv> no
[20-Aug-2009 14:38:20] <rmatte> ok, it's some other nforce board then
[20-Aug-2009 14:38:23] <markeriv> aye
[20-Aug-2009 14:38:25] <markeriv> yours looks newer
[20-Aug-2009 14:38:42] <rmatte> yeh, it was the newest one available when I bought it, now I'm sure there's newer out there
[20-Aug-2009 14:38:43] <markeriv> hey scp is back
[20-Aug-2009 14:38:47] <markeriv> hey SCP, i had a thought
[20-Aug-2009 14:38:55] <markeriv> under the device, under OS
[20-Aug-2009 14:38:59] <markeriv> do you see the interfaces?
[20-Aug-2009 14:39:03] <markeriv> if so click the interface name
[20-Aug-2009 14:39:15] <rmatte> mrayzenoss: that explains his awesome coding skills
[20-Aug-2009 14:39:15] <markeriv> one that has an IP address assigned to it
[20-Aug-2009 14:39:32] <markeriv> anyhow
[20-Aug-2009 14:40:10] <markeriv> hmm has it been 15 mins?
[20-Aug-2009 14:40:34] <rmatte> just about
[20-Aug-2009 14:40:38] <markeriv> hope not, still getting nans
[20-Aug-2009 14:40:50] <rmatte> you can always cheat and do zenperfsnmp run -d <devicename>
[20-Aug-2009 14:40:52] <markeriv> should I have removed the device first?
[20-Aug-2009 14:40:58] <rmatte> or zenperfsnmp run --cycle
[20-Aug-2009 14:41:00] <markeriv> am getting SNMP data just fine
[20-Aug-2009 14:41:07] <markeriv> and utilization ect
[20-Aug-2009 14:41:14] <markeriv> the only graphs i cant seem to get working are ping related
[20-Aug-2009 14:41:16] <rmatte> no, no need to remove the device
[20-Aug-2009 14:41:34] <markeriv> the company i work for monitors around 1000 network devices
[20-Aug-2009 14:41:35] <rmatte> ping related?
[20-Aug-2009 14:41:38] <rmatte> command based?
[20-Aug-2009 14:41:42] <markeriv> err fping
[20-Aug-2009 14:41:52] <rmatte> yeh, so command based
[20-Aug-2009 14:41:52] <markeriv> i need basic uptime over time data
[20-Aug-2009 14:42:04] <markeriv> like that device, 6 months ago was it up and online
[20-Aug-2009 14:42:13] <markeriv> i have to be able to prove or disprove SLAs
[20-Aug-2009 14:42:19] <rmatte> I see
[20-Aug-2009 14:42:31] <markeriv> so latency and jitter info is critical for for that
[20-Aug-2009 14:42:34] <markeriv> hence fping
[20-Aug-2009 14:42:36] <rmatte> why not just make an uptime graph?
[20-Aug-2009 14:42:47] <markeriv> would love that too, however I am as they say, a true newb
[20-Aug-2009 14:42:54] <rmatte> lol
[20-Aug-2009 14:42:59] <markeriv> so far parts of this zenpack are working
[20-Aug-2009 14:43:08] <markeriv> but the parts that dont make it look ugly
[20-Aug-2009 14:43:14] <rmatte> right
[20-Aug-2009 14:43:25] <markeriv> eventually this is gonna get put into a portal
[20-Aug-2009 14:43:35] <markeriv> where our customers should be able to view it dynamically
[20-Aug-2009 14:43:50] <rmatte> yeh, we're doing something similar with our graphed data eventually
[20-Aug-2009 14:44:02] <rmatte> you guys are an MSP?
[20-Aug-2009 14:44:05] <markeriv> anyhow, cause of that the graphs have to look pretty
[20-Aug-2009 14:44:10] <markeriv> and an ISP
[20-Aug-2009 14:44:15] <rmatte> ah, cool
[20-Aug-2009 14:44:26] <markeriv> looking to replace Silverback
[20-Aug-2009 14:44:43] <markeriv> microsofts former 10,000$ an update gig
[20-Aug-2009 14:44:50] <rmatte> if you haven't already, you should download the zenoss admin guide pdf and read it front to back
[20-Aug-2009 14:44:50] <markeriv> although i think they sold it to IBM
[20-Aug-2009 14:45:00] <rmatte> that's what I did when I was starting out, and it really opens your eyes
[20-Aug-2009 14:45:02] <markeriv> have, still some parts are confusing
[20-Aug-2009 14:45:03] <mrayzenoss> I think Dell bought Silverback
[20-Aug-2009 14:45:08] <rmatte> you'll have plenty of "wow, I can do that?" moments
[20-Aug-2009 14:45:11] <markeriv> might have been them, it got sold
[20-Aug-2009 14:45:19] <markeriv> hehe
[20-Aug-2009 14:45:32] <markeriv> i am half debating running 2.3.3
[20-Aug-2009 14:45:36] <mrayzenoss> http://www.silverbacktech.com/
[20-Aug-2009 14:45:45] <markeriv> right now I just have zenoss running on my pc
[20-Aug-2009 14:45:45] <mrayzenoss> well, 2.4.3 is scheduled for next week
[20-Aug-2009 14:45:50] <markeriv> O.o
[20-Aug-2009 14:45:59] <rmatte> I have 2.3.3 running on some servers here, waiting for 2.4.3 release to upgrade
[20-Aug-2009 14:46:02] <markeriv> but we have a server that should be arriving this week
[20-Aug-2009 14:46:20] <markeriv> is it easy to upgrade or is it better to do a fresh install
[20-Aug-2009 14:46:27] <markeriv> since I am going to be installing like next week
[20-Aug-2009 14:46:40] <mrayzenoss> fresh install is probably easiest
[20-Aug-2009 14:46:50] <mrayzenoss> discovery is usually pretty fast
[20-Aug-2009 14:46:56] <mrayzenoss> unless you have a lot of customization
[20-Aug-2009 14:47:14] <markeriv> well, unfortunately the network aint so clean
[20-Aug-2009 14:47:31] <markeriv> best i could hope for is some way to add devices off of a spreadsheet
[20-Aug-2009 14:47:32] <rmatte> yeh, we can't really do discovery ever with our clients
[20-Aug-2009 14:47:39] <rmatte> unless we want like a thousand workstations in inventory
[20-Aug-2009 14:47:46] <jb> question about disovery
[20-Aug-2009 14:47:48] <markeriv> cause there is a few thousand devices
[20-Aug-2009 14:47:54] <jb> i have an entire 10.0.0.0/8 network
[20-Aug-2009 14:47:56] <markeriv> and they are not easily identifable as a block
[20-Aug-2009 14:48:02] <jb> all the hosts I want to monitor end in a .15
[20-Aug-2009 14:48:18] <jb> is it possible to tell zenoss just to scan/add the .15's?
[20-Aug-2009 14:48:28] <jb> (there is about 900 servers)
[20-Aug-2009 14:48:33] <rmatte> jb: I believe you can do something like that in the add device wizard
[20-Aug-2009 14:48:35] <markeriv> nice network
[20-Aug-2009 14:48:39] <jb> haha
[20-Aug-2009 14:48:43] <jb> its a LEGACY network.
[20-Aug-2009 14:48:43] <rmatte> I know it allows you to do specific ranges and stuff
[20-Aug-2009 14:48:52] <jb> for remote locations
[20-Aug-2009 14:48:58] <mrayzenoss> yeah, but that's not a range
[20-Aug-2009 14:49:03] <jb> yeah its not
[20-Aug-2009 14:49:07] <rmatte> mrayzenoss: yeh, true
[20-Aug-2009 14:49:17] <rmatte> jb: you might have to write a script to do that
[20-Aug-2009 14:49:23] <mrayzenoss> I wonder if the code supports something like 10.0.0-255.15
[20-Aug-2009 14:49:43] <mrayzenoss> I doubt it works that way
[20-Aug-2009 14:49:56] <rmatte> mrayzenoss: I doubt it, that's a pretty specific case, I've never heard of anyone wanting to add all .15's or something
[20-Aug-2009 14:49:57] <mrayzenoss> probably easiest to do a bulk load
[20-Aug-2009 14:50:33] <rmatte> I'd just do an auto-discovery, then select all, and uncheck all the .15s, then delete the rest
[20-Aug-2009 14:50:53] <mrayzenoss> and then turn off auto-discovery
[20-Aug-2009 14:50:55] <rmatte> or write a script to do it
[20-Aug-2009 14:51:06] <jb> i have a feeling that auto discovery on an entire /8 would blow zenoss up
[20-Aug-2009 14:51:17] <rmatte> jb: it is pretty intensive
[20-Aug-2009 14:51:26] <jb> well..
[20-Aug-2009 14:51:30] <rmatte> jb: you should just write a bash script to do it or something
[20-Aug-2009 14:51:32] <jb> yeah
[20-Aug-2009 14:51:36] <jb> that would be the easiest
[20-Aug-2009 14:51:38] <mrayzenoss> yeah, you can tweak how many discoveries to run in parallel
[20-Aug-2009 14:51:51] <mrayzenoss> I know Chet messes with that
[20-Aug-2009 14:51:53] <jb> rmatte: you said you were working on CPU stats without snmp informant?
[20-Aug-2009 14:52:09] <rmatte> jb: yeh, I have it working, it's just not very clean
[20-Aug-2009 14:52:19] <rmatte> like clean in terms of the script being bash
[20-Aug-2009 14:52:24] <rmatte> when it really should be python
[20-Aug-2009 14:52:24] <rmatte> lol
[20-Aug-2009 14:52:35] <rmatte> I can send you the pack anyways if you'd like
[20-Aug-2009 14:52:38] <jb> out of the box with core.. i should be able to monitor things like windows services, os processes, filesystems, routing table, etc?
[20-Aug-2009 14:52:48] <mrayzenoss> yeah
[20-Aug-2009 14:52:48] <rmatte> jb: correct
[20-Aug-2009 14:52:55] <jb> and i should be able to monitor 900 windows devices?
[20-Aug-2009 14:52:57] <jb> on a beefy box?
[20-Aug-2009 14:53:00] <rmatte> jb: but not performance data, since Microsoft's snmp agent completely blows
[20-Aug-2009 14:53:02] <mrayzenoss> probably
[20-Aug-2009 14:53:09] <jb> (quad core at least, 32GB ram, raid10, etc)
[20-Aug-2009 14:53:16] <rmatte> my scripts allow performance data to be monitored
[20-Aug-2009 14:55:29] <markeriv> does microsoft have a product that doesnt blow?
[20-Aug-2009 14:55:40] <mrayzenoss> their mice are nice
[20-Aug-2009 14:55:42] <markeriv> you know... I am really starting to like this xchat program...
[20-Aug-2009 14:55:50] <rmatte> markeriv: there are a team of experts trying to find one, but so far, no luck
[20-Aug-2009 14:55:52] <markeriv> it says the word "microsoft" is mispelled
[20-Aug-2009 14:56:02] <markeriv> hahaha
[20-Aug-2009 14:56:05] <markeriv> nice rmatte
[20-Aug-2009 14:56:06] <mrayzenoss> I like my XBox
[20-Aug-2009 14:56:13] <markeriv> yea xboxs are ok
[20-Aug-2009 14:56:15] <markeriv> so are 360s
[20-Aug-2009 14:56:27] <markeriv> but i am trying to push our whole office Open source
[20-Aug-2009 14:56:30] <mrayzenoss> I'm just saying, they'd be a great company if they stuck to hardware and game consoles
[20-Aug-2009 14:56:36] <rmatte> I'm a ps3 guy myself
[20-Aug-2009 14:56:38] <markeriv> hard to get people to give up outlook
[20-Aug-2009 14:56:43] <ckrough> our office went to OpenOffice from MS Office
[20-Aug-2009 14:56:50] <markeriv> yea, only problem with that
[20-Aug-2009 14:56:55] <markeriv> is converged calendar and email
[20-Aug-2009 14:57:01] <markeriv> outlook has it and does it great
[20-Aug-2009 14:57:05] <ckrough> yeah
[20-Aug-2009 14:57:09] <ckrough> outlook is fairly pimp
[20-Aug-2009 14:57:11] <markeriv> so far every linux comparison does it poorly
[20-Aug-2009 14:57:15] <markeriv> its sad
[20-Aug-2009 14:57:39] <markeriv> any linux mail program that tries to compete seriously on business markets needs at least the ability to accept a outlook invitation
[20-Aug-2009 14:57:48] <markeriv> for a conference
[20-Aug-2009 14:57:51] <chudler_> I am not rabidly anti-MS, but exchange has failed to impress me. meh
[20-Aug-2009 14:58:01] <ckrough> client based email is dead. long live SaaS
[20-Aug-2009 14:58:08] <rmatte> actually evolution is quite good
[20-Aug-2009 14:58:16] <rmatte> it's almost as good as outlook
[20-Aug-2009 14:58:18] <mrayzenoss> yeah, I used that for a long time
[20-Aug-2009 14:58:51] <markeriv> i personally am liking gmail
[20-Aug-2009 14:58:58] <markeriv> i also work for mechaps.com
[20-Aug-2009 14:59:07] <markeriv> and they use google as a host for their domain
[20-Aug-2009 14:59:10] <markeriv> and it works very well
[20-Aug-2009 14:59:12] <rmatte> I do use Outlook here, but only because we're a microsoft shop for internal stuff
[20-Aug-2009 14:59:20] <rmatte> I use it inside of virtualbox
[20-Aug-2009 14:59:25] <markeriv> hehe
[20-Aug-2009 14:59:25] <mrayzenoss> Zenoss uses Zimbra
[20-Aug-2009 14:59:28] <markeriv> now that i need to get
[20-Aug-2009 14:59:35] <markeriv> a virtual box for outlook
[20-Aug-2009 14:59:38] <rmatte> with windows vista which is magically eating up 11 gigs of hard drive space, and I can't figure out how
[20-Aug-2009 14:59:39] <markeriv> if i could set that up
[20-Aug-2009 14:59:43] <rmatte> since the files on disk are less than that
[20-Aug-2009 14:59:55] <markeriv> i could theoretically convert my whole office
[20-Aug-2009 15:00:11] <markeriv> and have no one notice... well... not too much
[20-Aug-2009 15:00:23] <rmatte> markeriv: virtualbox is just like vmware, but free
[20-Aug-2009 15:00:28] <rmatte> very good
[20-Aug-2009 15:00:30] <markeriv> bah still getting nans
[20-Aug-2009 15:00:36] <rmatte> crappy
[20-Aug-2009 15:00:54] <rmatte> I'm out of ideas for you, are you sure you had to create the RRDs by hand?
[20-Aug-2009 15:00:58] <rmatte> that might be throwing it off
[20-Aug-2009 15:01:43] <markeriv> yea was no file
[20-Aug-2009 15:01:48] <markeriv> i can try adding a new site
[20-Aug-2009 15:03:29] <rhett> i've been thinking of setting up zimbra but i really only care about getting caldav working on my iphone
[20-Aug-2009 15:03:53] <rhett> cosmo (chandler server) and bedework both have issues w/ ical on the damn iphone from my experience
[20-Aug-2009 15:04:03] <rhett> stupid iphone
[20-Aug-2009 15:04:04] <mrayzenoss> markeriv: look at zencommand logging, maybe there's something in there?
[20-Aug-2009 15:04:39] <mrayzenoss> rhett: I use 14 calendars on my iPhone
[20-Aug-2009 15:05:00] <mrayzenoss> 5 cal dav from Zimbra, 1 ActiveSync from Zimbra
[20-Aug-2009 15:05:12] <mrayzenoss> and a bunch from Gooogle
[20-Aug-2009 15:05:58] <rhett> ah cool. the zimbra stuff work well over caldav?
[20-Aug-2009 15:06:29] <mrayzenoss> it's gotten a lot better
[20-Aug-2009 15:07:04] <mrayzenoss> we have occasional hiccups
[20-Aug-2009 15:07:17] <jb> ..
[20-Aug-2009 15:07:21] <rmatte> If anyone is interested in my Windows SNMP performance ZenPack, here it is: http://dmon.org/downloads/zenoss/ZenPacks.Nova.Windows.SNMPPerfMonitor-1.0-py2.4.egg
[20-Aug-2009 15:07:34] <rmatte> you need to have bc and snmpwalk installed for it to work
[20-Aug-2009 15:08:02] <rmatte> it also creates folders under /Servers/Windows
[20-Aug-2009 15:08:23] <rmatte> so you'll have like /Servers/Windows/1 Processor, 2 Processors, 4 Processors, 8 Processors, and 16 Processors
[20-Aug-2009 15:08:43] <rmatte> you need to do: snmpwalk -v1 -c CommunityString <DeviceName> hrProcessorLoad
[20-Aug-2009 15:08:51] <rmatte> then count how many lines, that tell you which group it should go in
[20-Aug-2009 15:09:02] <rmatte> so if you see 4 lines, drop it in to the 4 processors group
[20-Aug-2009 15:09:44] <rmatte> like I said, it's really rough right now, but it does work
[20-Aug-2009 15:10:01] <rmatte> I'm using it to monitor over 70 servers on one Zenoss instance
[20-Aug-2009 15:11:02] <rhett> thats good to hear though mrayzenoss. cosmo seems to have some odd auth issue w/ the long hashed url's and bedework requires a trailing slash on the url which the iphone removes
[20-Aug-2009 15:12:06] <markeriv> trying adding a new device on the off chance that my adding the rdd file manually screwed it up
[20-Aug-2009 15:14:27] <markeriv> yea by default it is still missing
[20-Aug-2009 15:14:28] <markeriv> grr
[20-Aug-2009 15:14:41] <mrayzenoss> is the template bound?
[20-Aug-2009 15:15:33] <markeriv> the only line that references it in zencommand is this one
[20-Aug-2009 15:15:34] <markeriv> 2009-08-18 09:54:27 INFO zen.zencommand: Deleting command Emser Tile Oakbrook from check_ping -H Emser Tile Oakbrook -w 180,100% -c 300,100% | sed -e 's# - Packet loss#|LOSS#;s#%,##;s#ms##;s# = #=#g'
[20-Aug-2009 15:15:44] <markeriv> aye the graphs are physically there
[20-Aug-2009 15:15:52] <markeriv> i just get nans and missing rrd files
[20-Aug-2009 15:15:59] <markeriv> i bound the template to devices/routers/cisco
[20-Aug-2009 15:16:20] <markeriv> and that aye is for the template question mray, they are bound
[20-Aug-2009 15:16:58] <mrayzenoss> I thought you'd said you had manually added the RRD files, which is odd
[20-Aug-2009 15:17:51] <rmatte> he manually created them I believe he said
[20-Aug-2009 15:17:59] <rmatte> not sure what method was used?
[20-Aug-2009 15:18:14] <rmatte> did you just do something like "touch RRD-FILE-NAME.rrd"
[20-Aug-2009 15:18:15] <rmatte> ?
[20-Aug-2009 15:18:30] <markeriv> i copied the max file
[20-Aug-2009 15:18:41] <rmatte> oh
[20-Aug-2009 15:18:47] <markeriv> and renamed it min
[20-Aug-2009 15:19:10] <rmatte> hmmm
[20-Aug-2009 15:19:19] <markeriv> the file that isnt populating is fping_min
[20-Aug-2009 15:19:33] <rmatte> quite honestly, if one of my Zenoss instances got that messed up I'd be reinstalling lol
[20-Aug-2009 15:19:47] <rmatte> you obviously shouldn't have to create the file by hand
[20-Aug-2009 15:19:56] <rmatte> and if it's not doing it automatically, that's the real issue
[20-Aug-2009 15:19:57] <mrayzenoss> I guess I wasn't paying close enough attention, but what else is hosed?
[20-Aug-2009 15:20:13] <markeriv> just ping graphs
[20-Aug-2009 15:20:14] <markeriv> oie
[20-Aug-2009 15:20:19] <markeriv> 2nd reinstall already
[20-Aug-2009 15:20:23] <rmatte> ah
[20-Aug-2009 15:20:26] <markeriv> here is the confusing bit
[20-Aug-2009 15:20:31] <rmatte> hmmm, maybe it's the pack then
[20-Aug-2009 15:20:34] <markeriv> Executing command
[20-Aug-2009 15:20:34] <markeriv> /usr/sbin/fping -c 3 -sq 69.3.230.36 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /'
[20-Aug-2009 15:20:34] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:20:34] <markeriv> PING OK|xmt=3 rcv=3 loss=0, min/avg/max =min= 115 avg=115 max=115
[20-Aug-2009 15:20:35] <markeriv> DONE in 2 seconds
[20-Aug-2009 15:20:47] <markeriv> it appears to work when I run the command from the template
[20-Aug-2009 15:20:54] <markeriv> and it even pulls data
[20-Aug-2009 15:20:54] <rmatte> are those datapoints present under the datasource?
[20-Aug-2009 15:21:08] <markeriv> avg GAUGE
[20-Aug-2009 15:21:08] <markeriv> loss GAUGE
[20-Aug-2009 15:21:08] <markeriv> max GAUGE
[20-Aug-2009 15:21:08] <markeriv> min GAUGE
[20-Aug-2009 15:21:08] <markeriv> rcv GAUGE
[20-Aug-2009 15:21:09] <markeriv> xmt GAUGE
[20-Aug-2009 15:21:12] <markeriv> those are my data points
[20-Aug-2009 15:21:23] <rmatte> "min/avg/max =min= 115"
[20-Aug-2009 15:21:29] <rmatte> odd
[20-Aug-2009 15:21:38] <markeriv> 115ms for 3 pings
[20-Aug-2009 15:21:40] <rhett> yeah
[20-Aug-2009 15:21:49] <rhett> that seems a bit odd
[20-Aug-2009 15:21:59] <rhett> but possible
[20-Aug-2009 15:22:00] <rmatte> yeh, I just don't know why it's outputting it like that
[20-Aug-2009 15:22:01] <rmatte> bizarre
[20-Aug-2009 15:22:12] <markeriv> hmm i am noticing a little bit more jitter than that is suggesting
[20-Aug-2009 15:22:18] <rmatte> 115ms for 3 pings is plausible
[20-Aug-2009 15:22:23] <markeriv> enough that i am disbelieving 3 115s
[20-Aug-2009 15:22:40] <rhett> bump it up to 10 15 or 20 pings and see what happens
[20-Aug-2009 15:22:44] <markeriv> Executing command
[20-Aug-2009 15:22:44] <markeriv> /usr/sbin/fping -c 3 -sq 69.3.230.36 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /'
[20-Aug-2009 15:22:44] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:22:44] <markeriv> PING OK|xmt=3 rcv=3 loss=0, min/avg/max =min= 114 avg=117 max=120
[20-Aug-2009 15:22:47] <markeriv> DONE in 2 seconds
[20-Aug-2009 15:22:49] <markeriv> meh
[20-Aug-2009 15:22:51] <markeriv> that time was more regular
[20-Aug-2009 15:23:00] <rmatte> yeh
[20-Aug-2009 15:23:21] <rmatte> it's just this: min/avg/max =min= 114
[20-Aug-2009 15:23:24] <rmatte> that I'm wondering about
[20-Aug-2009 15:23:46] <rmatte> the output should be more like: PING OK|xmt=3 rcv=3 loss=0 min=114 avg=117 max=120
[20-Aug-2009 15:23:48] <rmatte> from my experience
[20-Aug-2009 15:24:19] <mrayzenoss> yeah, that's what I'm thinking
[20-Aug-2009 15:24:54] <markeriv> here is the command itself
[20-Aug-2009 15:24:59] <markeriv> should I delete the = at the end?
[20-Aug-2009 15:24:59] <markeriv> /usr/sbin/fping -c 3 -sq ${dev/manageIp} 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /'
[20-Aug-2009 15:25:14] * mrayzenoss eyes glazed over immediately
[20-Aug-2009 15:25:19] <markeriv> uhh oh
[20-Aug-2009 15:25:22] <rmatte> that's quite the command
[20-Aug-2009 15:25:25] <markeriv> yea
[20-Aug-2009 15:25:27] <markeriv> i didnt build it
[20-Aug-2009 15:25:51] <rmatte> I'd get rid of "min\/avg\/max ="
[20-Aug-2009 15:25:53] <rmatte> and then try it
[20-Aug-2009 15:25:56] <markeriv> blame BlakeDrager
[20-Aug-2009 15:26:01] <rmatte> it seems to be inserting that as plain text
[20-Aug-2009 15:26:26] <markeriv> min\/avg\/max =
[20-Aug-2009 15:26:30] <markeriv> pasted to preserve
[20-Aug-2009 15:26:35] <rmatte> yup
[20-Aug-2009 15:26:46] <rmatte> well, you have the whole command pasted to preserve
[20-Aug-2009 15:26:57] <markeriv> PING OK|xmt=3 rcv=3 loss=0 min/avg/max =min= 116 avg=119 max=122
[20-Aug-2009 15:27:05] <markeriv> with that line changed
[20-Aug-2009 15:27:08] <rmatte> so it didn't change anything
[20-Aug-2009 15:27:08] <markeriv> hmm
[20-Aug-2009 15:27:10] <markeriv> ya
[20-Aug-2009 15:27:38] <rmatte> that sed makes my eyes bleed
[20-Aug-2009 15:27:51] <markeriv> lol
[20-Aug-2009 15:28:00] <markeriv> think we could rewrite it, or would that be paid time?
[20-Aug-2009 15:28:09] <rmatte> but yeh, you need to mess around with it until it looks like: PING OK|xmt=3 rcv=3 loss=0 min=114 avg=117 max=120
[20-Aug-2009 15:28:29] <rmatte> since zenoss is probably tripping up on it because of the extra junk
[20-Aug-2009 15:28:40] <rhett> against Cypress
[20-Aug-2009 15:28:40] <rhett> PING OK|loss=0 min=0.141 avg=0.181 max=0.215 mdev=0.027
[20-Aug-2009 15:28:40] <rhett> DONE in 4 seconds
[20-Aug-2009 15:28:50] <rhett> well that didnt work
[20-Aug-2009 15:28:55] <rmatte> lol
[20-Aug-2009 15:28:58] <rhett> /bin/ping -i .5 -n -q -c 10 Cypress | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:29:00] <rhett> there
[20-Aug-2009 15:29:22] <rhett> thats what I use
[20-Aug-2009 15:29:34] <markeriv> should that be fping or just ping?
[20-Aug-2009 15:30:11] <rmatte> markeriv: doesn't matter...
[20-Aug-2009 15:30:20] <rmatte> fping allows you to ping multiple hosts in parallel
[20-Aug-2009 15:30:30] <rmatte> but you're not doing that, you're just pinging a single host per command
[20-Aug-2009 15:30:38] <markeriv> ahh but i am going to be
[20-Aug-2009 15:30:46] <markeriv> got 1000 or so device to track >.>
[20-Aug-2009 15:30:53] <markeriv> meh
[20-Aug-2009 15:30:55] <rmatte> yes, but Zenoss is going to do that for you
[20-Aug-2009 15:30:56] <markeriv> that gave me zenoss error
[20-Aug-2009 15:30:58] <markeriv> ahh nod
[20-Aug-2009 15:31:19] <rmatte> fping just lets you be like ping <host1> <host2> <host3> or whatever
[20-Aug-2009 15:31:24] <rmatte> fping rather
[20-Aug-2009 15:31:30] <rmatte> but you don't need that
[20-Aug-2009 15:31:33] <markeriv> ahh nod
[20-Aug-2009 15:31:48] <rmatte> the command rhett pasted need different datapoints
[20-Aug-2009 15:31:53] <rmatte> they need to match the output
[20-Aug-2009 15:31:53] <rhett> yes
[20-Aug-2009 15:32:03] <markeriv> yikes
[20-Aug-2009 15:32:04] <markeriv> ok
[20-Aug-2009 15:32:10] <rmatte> so loss min avg max and mdev
[20-Aug-2009 15:32:24] <rmatte> you also need to adjust your graph to add the new datapoints
[20-Aug-2009 15:32:26] <rmatte> then test it
[20-Aug-2009 15:32:36] <markeriv> type is gauge?
[20-Aug-2009 15:32:40] <rhett> avg GAUGE
[20-Aug-2009 15:32:40] <rhett> loss GAUGE
[20-Aug-2009 15:32:40] <rhett> max GAUGE
[20-Aug-2009 15:32:40] <rhett> mdev GAUGE
[20-Aug-2009 15:32:40] <rhett> min GAUGE
[20-Aug-2009 15:32:42] <rhett> yes
[20-Aug-2009 15:32:44] <rmatte> correct
[20-Aug-2009 15:33:13] <markeriv> hmm
[20-Aug-2009 15:33:16] <markeriv> added that one data point
[20-Aug-2009 15:33:21] <markeriv> still getting an error when i test
[20-Aug-2009 15:33:49] <markeriv> /bin/ping -i .5 -n -q -c 10 ${dev/manageIp} | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:33:53] <markeriv> thats what I am running
[20-Aug-2009 15:33:57] <rmatte> lol, my friend's IM status is "The artist formerly known as #!/bin/sh"
[20-Aug-2009 15:34:03] <markeriv> lol
[20-Aug-2009 15:34:05] <rhett> /bin/ping -i .5 -n -q -c 10 $devname | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$$/ {;N;s/\n//}'
[20-Aug-2009 15:34:11] <rhett> thats mine
[20-Aug-2009 15:34:36] <rmatte> same thing
[20-Aug-2009 15:34:47] <markeriv> well
[20-Aug-2009 15:34:51] <markeriv> aint pumping out the same error
[20-Aug-2009 15:34:52] <kisielk> hm, is there a faster way to access custom event views?
[20-Aug-2009 15:34:55] <markeriv> now its this one
[20-Aug-2009 15:34:58] <kisielk> than going through Settings/Users/ etc
[20-Aug-2009 15:35:04] <markeriv> Executing command
[20-Aug-2009 15:35:04] <markeriv> /bin/ping -i .5 -n -q -c 10 Emser Tile Louisville | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:35:04] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:35:04] <markeriv> ping: unknown host Emser
[20-Aug-2009 15:35:05] <markeriv> DONE in 0 seconds
[20-Aug-2009 15:35:07] <kisielk> ideally, is there a way to add one to the dashboard?
[20-Aug-2009 15:35:08] <markeriv> least thats better
[20-Aug-2009 15:35:15] <markeriv> now its just not using its own internal dns check
[20-Aug-2009 15:35:17] <rmatte> kisielk: unfortunately, no
[20-Aug-2009 15:35:19] <markeriv> will try by IP
[20-Aug-2009 15:35:30] <rmatte> kisielk: bookmark it
[20-Aug-2009 15:35:30] <kisielk> damn, that sucks
[20-Aug-2009 15:35:39] <kisielk> how hard would it be for them to be dashboardable
[20-Aug-2009 15:35:43] <rmatte> map it to a hotkey sequence or something
[20-Aug-2009 15:35:47] <kisielk> it would make the feature infinitely more useful
[20-Aug-2009 15:35:49] <markeriv> nope, puts the name back in there
[20-Aug-2009 15:36:26] <mrayzenoss> kisielk: http://public-demo.zenoss.com/evconsole
[20-Aug-2009 15:36:29] <markeriv> looks like it is not converting the device name back to its IP address
[20-Aug-2009 15:36:42] <rmatte> kisielk: the new event console has filters for everything
[20-Aug-2009 15:36:48] <mrayzenoss> and saved views
[20-Aug-2009 15:36:48] <rmatte> so it's going to effectively replace those anyways
[20-Aug-2009 15:36:54] <kisielk> right
[20-Aug-2009 15:37:16] <kisielk> how far are w from seeing that?
[20-Aug-2009 15:37:17] <rhett> markeriv yeah. from the error you posted thats the problem
[20-Aug-2009 15:37:24] <mrayzenoss> beta is probably 3 weeks out
[20-Aug-2009 15:37:38] <rmatte> kisielk: final release is due around October
[20-Aug-2009 15:37:41] <rmatte> so it's not that far off
[20-Aug-2009 15:37:45] <kisielk> nice
[20-Aug-2009 15:37:50] <rmatte> King Crab is going to be awesome
[20-Aug-2009 15:38:01] <mrayzenoss> markeriv: in the datasource, what's the command line?
[20-Aug-2009 15:38:03] * rmatte rubs his hands together
[20-Aug-2009 15:38:22] <rhett> you cant go much further w/ the crab names after king crab
[20-Aug-2009 15:38:46] <markeriv> you mean the command you just had me enter?
[20-Aug-2009 15:38:46] <mrayzenoss> rhett: http://en.wikipedia.org/wiki/Crab
[20-Aug-2009 15:38:50] <markeriv> /bin/ping -i .5 -n -q -c 10 $devname | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$$/ {;N;s/\n//}'
[20-Aug-2009 15:38:52] <markeriv> or elsewhere?
[20-Aug-2009 15:39:17] <mrayzenoss> replace $devname with ${here/manageIp}
[20-Aug-2009 15:39:57] <markeriv> Executing command
[20-Aug-2009 15:39:57] <markeriv> /bin/ping -i .5 -n -q -c 10 69.3.230.36 | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:39:57] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:39:57] <markeriv> PING OK|loss=0 min=116.007 avg=118.862 max=125.374 mdev=2.845
[20-Aug-2009 15:39:58] <markeriv> DONE in 4 seconds
[20-Aug-2009 15:39:58] <rmatte> mrayzenoss: hehe
[20-Aug-2009 15:40:00] <markeriv> wootah
[20-Aug-2009 15:40:06] <rhett> nice
[20-Aug-2009 15:40:06] <rmatte> rhett: next one will be stone crab
[20-Aug-2009 15:40:20] <markeriv> stop talking about crabs, you are making me hungry
[20-Aug-2009 15:40:26] <rmatte> omg, please name the next one after that "Sally Lightfoot Crab"
[20-Aug-2009 15:40:38] <rmatte> Arrow Crab would be a good one actually
[20-Aug-2009 15:40:38] <rhett> ha. i just mean because you know its king crab and you dont get much further up the hiearchy than the king
[20-Aug-2009 15:40:44] <mrayzenoss> ahh, true
[20-Aug-2009 15:40:46] <rmatte> Ghost Crab, oooh
[20-Aug-2009 15:40:53] <markeriv> emperor crab?
[20-Aug-2009 15:40:57] <markeriv> darth crab?
[20-Aug-2009 15:40:58] <mrayzenoss> Blue Crab, King Crab and Stone Crab taken so far
[20-Aug-2009 15:41:02] <markeriv> lord crab?
[20-Aug-2009 15:41:17] <rhett> coconut crab
[20-Aug-2009 15:41:25] <markeriv> does anyone have any lemon?
[20-Aug-2009 15:41:28] <rhett> ha
[20-Aug-2009 15:41:33] <mrayzenoss> basically if we release twice a year, we have a good 847 more names according to Wikipedia
[20-Aug-2009 15:41:40] <rmatte> yup
[20-Aug-2009 15:41:55] <markeriv> ok so now i have good points
[20-Aug-2009 15:41:55] <mrayzenoss> so just 423 years of Zenoss releases
[20-Aug-2009 15:41:58] <markeriv> now to edit the graph
[20-Aug-2009 15:42:10] <rmatte> Although "Liocarcinus vernalis Crab" doesn't really have much of a ring to it
[20-Aug-2009 15:42:13] <rmatte> hard to pronounce too
[20-Aug-2009 15:42:21] <mrayzenoss> trust me, in 400 years you won't care
[20-Aug-2009 15:42:22] <rmatte>
[20-Aug-2009 15:42:28] <rmatte> haha, true
[20-Aug-2009 15:42:50] <rmatte> well I might, what is going to monitor my status pod status?
[20-Aug-2009 15:42:55] <markeriv> just call it LVcrab
[20-Aug-2009 15:43:10] <rmatte> erm, stasis
[20-Aug-2009 15:43:10] <markeriv> your status pod? What about my cryo tube
[20-Aug-2009 15:43:26] <markeriv> i want something making sure i get let out on time
[20-Aug-2009 15:43:38] <chudler_> "Open the pod bay door please, Zen. I'm sorry Matt, I cannot do that."
[20-Aug-2009 15:43:48] <rmatte> lol
[20-Aug-2009 15:43:56] <markeriv> so erm
[20-Aug-2009 15:43:58] <markeriv> on your graph
[20-Aug-2009 15:44:04] <markeriv> what becomes pings sent/pings recieved
[20-Aug-2009 15:44:12] <markeriv> need to replace those
[20-Aug-2009 15:44:18] <rhett> http://www.allthepages.org/images/blog/coconut_crab.jpg such a lovely picture
[20-Aug-2009 15:44:19] <markeriv> we have avg, loss, max, mdev, and min
[20-Aug-2009 15:44:29] <markeriv> so do i just delete that chart?
[20-Aug-2009 15:44:53] <rmatte> that's a CRAB!?
[20-Aug-2009 15:44:56] <rmatte> crazy
[20-Aug-2009 15:45:09] <rhett> coconut crab
[20-Aug-2009 15:45:18] <rhett> supposed to be super delicious
[20-Aug-2009 15:45:21] <rmatte> that's a frigging monster
[20-Aug-2009 15:46:29] <rmatte> "some coconut crabs are rumored to steal shiny items such as pots and silverware from houses and tents."
[20-Aug-2009 15:46:35] <rmatte> how the heck would they even get in?
[20-Aug-2009 15:46:42] <rmatte> it's not like they can slip under the door
[20-Aug-2009 15:46:56] <chudler_> THEY HAVE THEIR WAYS. XMAS ISLAND. http://www.jurgenfreund.com/stories/xmascrabs/content/bin/images/large/1001785.jpg
[20-Aug-2009 15:47:19] <rmatte> hehe
[20-Aug-2009 15:48:16] <markeriv> so what does the point mdev do?
[20-Aug-2009 15:48:19] <markeriv> or how should I graph it
[20-Aug-2009 15:49:18] <rhett> it graphs deviation
[20-Aug-2009 15:49:19] <rhett> jitter
[20-Aug-2009 15:49:41] <rmatte> IPSLA style stuff
[20-Aug-2009 15:49:51] <markeriv> so latency and loss then
[20-Aug-2009 15:50:50] <rmatte> jitter is the variation in the time between packets arriving, caused by network congestion, timing drift, or route changes. A jitter buffer can be used to handle jitter.
[20-Aug-2009 15:50:57] <markeriv> nod
[20-Aug-2009 15:51:49] <rhett> i still like smokeping or the advanced ping templates for cacti more
[20-Aug-2009 15:51:52] <rhett> but it gives a good idea
[20-Aug-2009 15:53:02] <rmatte> I like actual IPSLA implemented on a core Cisco router
[20-Aug-2009 15:53:02] <markeriv> oie
[20-Aug-2009 15:53:05] <rmatte> that's the ultimate
[20-Aug-2009 15:53:08] <markeriv> i have really managed to screw this graph up lol
[20-Aug-2009 15:53:25] <rhett> too rich for my blood rmatte
[20-Aug-2009 15:53:45] <rmatte> hehe
[20-Aug-2009 15:54:03] <rmatte> we have it implemented for the links to our clients
[20-Aug-2009 15:54:16] <rmatte> so that we can provide them with stats on it in their monthly report
[20-Aug-2009 15:54:46] <rhett> cool
[20-Aug-2009 15:55:17] <rmatte> I wrote a bash script that collects the data from the router via SNMP and feeds it in to Zenoss
[20-Aug-2009 15:55:40] <rhett> cool
[20-Aug-2009 15:57:23] <rmatte> now I just need to figure out how to write proper snmp scripts in python
[20-Aug-2009 15:57:44] <rmatte> If I can modify my SNMP performance monitor zenpack to python it'll be perfect
[20-Aug-2009 16:00:05] <rmatte> I wonder if there's something similar to the cut command available in python
[20-Aug-2009 16:05:17] <rmatte> markeriv: but is it actually displaying data now?
[20-Aug-2009 16:06:13] <rmatte> afk
[20-Aug-2009 16:10:08] <markeriv> lemmie refresh
[20-Aug-2009 16:10:55] <markeriv> mostly
[20-Aug-2009 16:11:03] <markeriv> a couple fields still are not populating
[20-Aug-2009 16:11:05] <markeriv> but i can remove those
[20-Aug-2009 16:11:09] <markeriv> to get a clean looking graph
[20-Aug-2009 16:11:18] <markeriv> round trip minimum in particular
[20-Aug-2009 16:13:50] <rmatte> back
[20-Aug-2009 16:14:11] <rmatte> weird that min is not populating
[20-Aug-2009 16:14:17] <rmatte> the output is correct now with that command
[20-Aug-2009 16:14:19] <markeriv> at this point i just want to delete the round trip minimum and min difference in the graphs i have
[20-Aug-2009 16:14:44] <markeriv> well i have round trip maximum and round trip average
[20-Aug-2009 16:14:54] <markeriv> but they are listed twice in the graph description field
[20-Aug-2009 16:15:09] <markeriv> and the minimum still isnt populating
[20-Aug-2009 16:15:27] <rmatte> did you adjust your graph datapoints?
[20-Aug-2009 16:15:38] <rmatte> they shouldn't be showing up twice
[20-Aug-2009 16:15:48] <rmatte> you need to remove any duplicate datapoints in the graph
[20-Aug-2009 16:15:50] <markeriv> i believe so, but again one claims newbacy
[20-Aug-2009 16:15:51] <rmatte> start fresh
[20-Aug-2009 16:16:29] <markeriv> there we go
[20-Aug-2009 16:16:31] <markeriv> that did it
[20-Aug-2009 16:16:38] <markeriv> dude had some custom stuff in there
[20-Aug-2009 16:16:39] <rmatte> good
[20-Aug-2009 16:16:42] <markeriv> and i found it and nixed it
[20-Aug-2009 16:16:45] <markeriv> now it looks much cleaner
[20-Aug-2009 16:16:48] <rmatte> you'll get the hang of it soon enough
[20-Aug-2009 16:16:58] <rmatte> after a while it just becomes second nature
[20-Aug-2009 16:19:14] <markeriv> nod I hope so
[20-Aug-2009 16:19:15] <markeriv> lol
[20-Aug-2009 16:19:15] <markeriv> well
[20-Aug-2009 16:19:19] <markeriv> more like i hope to get it configured
[20-Aug-2009 16:19:21] <markeriv> copy it once
[20-Aug-2009 16:19:24] <markeriv> and never touch it again
[20-Aug-2009 16:19:35] <markeriv> other than backups ect
[20-Aug-2009 16:19:45] <markeriv> hence all this dev work before we have the product
[20-Aug-2009 16:20:46] <markeriv> hmm now this is interesting
[20-Aug-2009 16:20:53] <markeriv> under the minimum point
[20-Aug-2009 16:21:04] <markeriv> the available RRD variables does not have the Minimum
[20-Aug-2009 16:21:30] <markeriv> what would happen if i deleted the fping_min file? Would it auto recreate it with data?
[20-Aug-2009 16:22:22] <mrayzenoss> yeah
[20-Aug-2009 16:22:25] <markeriv> i suppose arguably I do not need that particular variable as most customers dont care how small their jitter is, rather how large
[20-Aug-2009 16:22:29] <markeriv> alright
[20-Aug-2009 16:22:32] <markeriv> i will try doing that
[20-Aug-2009 16:22:34] <markeriv> see if that fixes it
[20-Aug-2009 16:22:39] <markeriv> or if it gives cant find RRD file issues
[20-Aug-2009 16:22:51] <rmatte> markeriv: knowing customers, the more info you can throw at them, the better
[20-Aug-2009 16:23:01] <markeriv> yea but sometimes less is more
[20-Aug-2009 16:23:05] <rmatte> you have to aim to impress
[20-Aug-2009 16:23:21] <markeriv> i cant count how many times I have said that a customer's utilization on their T1 is low only to have them yell at me
[20-Aug-2009 16:23:37] <markeriv> and tell me i better damn well fix it
[20-Aug-2009 16:23:38] <markeriv> lol
[20-Aug-2009 16:23:56] <rmatte> lol
[20-Aug-2009 16:24:11] <rmatte> I'd be like "sure thing" *fire up bittorrent*
[20-Aug-2009 16:24:16] <rmatte> I'm done for the day
[20-Aug-2009 16:24:20] <rmatte> talk to you guys tomorrow
[20-Aug-2009 16:24:20] <markeriv> LOL!
[20-Aug-2009 16:24:25] <markeriv> thanks a bunch rmatte
[20-Aug-2009 16:24:30] <rmatte> no problem
[20-Aug-2009 16:24:32] <markeriv> everyone has been a big help
[20-Aug-2009 16:29:00] <rhett> is everything working for you?
[20-Aug-2009 16:34:22] <mrayzenoss> later all
[20-Aug-2009 16:36:35] <rhett> good bye
[20-Aug-2009 18:04:02] <etank> anyone know if i can run zenoss on a CentOS 5 x86 box and have a distrobuted collector running on a Red Hat EL 4 machine?
[20-Aug-2009 18:04:17] <etank> s/distrobuted/distributed/
[20-Aug-2009 19:12:30] <rhett> i know you can run it on cent5 dont know about collector on cent4
[20-Aug-2009 19:12:38] <rhett> thats my easy answer!
[21-Aug-2009 00:00:46] [disconnected at Fri Aug 21 00:00:46 2009]
[21-Aug-2009 00:00:46] [connected at Fri Aug 21 00:00:46 2009]
[21-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[21-Aug-2009 06:52:08] <_mingau_> hello people
[21-Aug-2009 06:54:01] <ckrough> good morning
[21-Aug-2009 06:55:30] <_mingau_> morning
[21-Aug-2009 06:55:41] <_mingau_> does anybody use freebsd here?
[21-Aug-2009 07:02:05] <_mingau_> ?
[21-Aug-2009 07:05:56] <_mingau_> testing?
[21-Aug-2009 07:07:08] <ckrough> Not too many people do. I don't believe it's an
[21-Aug-2009 07:07:29] <ckrough> 'officially' supported OS for Zenoss. I think straterra uses it, he's the only one I've seen though.
[21-Aug-2009 07:09:48] <_mingau_> i use too
[21-Aug-2009 07:09:49] <_mingau_>
[21-Aug-2009 07:10:04] <ckrough> I assumed
[21-Aug-2009 07:22:13] <rmatte> scp2004 uses it, but he's running in to problems as well
[21-Aug-2009 07:22:38] <rmatte> FreeBSD + Zenoss == Problematic
[21-Aug-2009 07:23:12] <rmatte> seems like everyone who uses that combo runs in to problems, then when they fix one problem something else breaks
[21-Aug-2009 07:23:19] <rmatte> and it's a never-ending ordeal
[21-Aug-2009 07:25:46] <rmatte> doesn't running Zenoss on FreeBSD partially use FreeBSD's Linux emulation?
[21-Aug-2009 07:26:00] <rmatte> or does it run 100% native after compilation?
[21-Aug-2009 07:28:39] <rmatte> ah, you need to add the proc line to fstab, so yeh it does use a bit of emulation
[21-Aug-2009 07:30:43] <_mingau_> hey Ryan
[21-Aug-2009 07:30:58] <_mingau_> i know its problematic
[21-Aug-2009 07:30:59] <_mingau_>
[21-Aug-2009 07:31:02] <_mingau_> u know...
[21-Aug-2009 07:31:08] <_mingau_> its my karma
[21-Aug-2009 07:34:40] <_mingau_> i am trying to make my boss understand that i need any linux...
[21-Aug-2009 07:35:43] <rmatte> hehe
[21-Aug-2009 07:35:55] <rmatte> I thought they had agreed to install CentOS?
[21-Aug-2009 07:36:13] <rmatte> or maybe that was scp2004
[21-Aug-2009 07:37:09] <_mingau_> kkk
[21-Aug-2009 07:38:38] <rmatte> CentOS is just the free version of RedHat
[21-Aug-2009 08:11:03] <_mingau_> i know
[21-Aug-2009 08:48:41] <rmatte> good morning Matt
[21-Aug-2009 08:49:38] <mrayzenoss> morning
[21-Aug-2009 08:49:48] <mrayzenoss> today's the day
[21-Aug-2009 08:49:59] <mrayzenoss> I've got 12 entries for the ZenPack contest that I need to get through
[21-Aug-2009 08:50:04] <_mingau_> hello big matt
[21-Aug-2009 08:50:10] <jb> doesnt look like mine is going to make it
[21-Aug-2009 08:50:12] <_mingau_> hey...i have a question...
[21-Aug-2009 08:50:13] <jb> but it wouldn't win anyways
[21-Aug-2009 08:50:32] <rmatte> mrayzenoss: I might as well submit mine, but I'm going to write up an instruction blurb for the ZenPack page
[21-Aug-2009 08:50:57] <rmatte> since it's not as polished as I'd like it to be, but it works fine
[21-Aug-2009 08:51:28] <_mingau_> about autodiscover windows devices...
[21-Aug-2009 08:51:56] <_mingau_> the authentication...
[21-Aug-2009 08:52:13] <_mingau_> it shows that the user must be a local admin...
[21-Aug-2009 08:52:17] <_mingau_> so...
[21-Aug-2009 08:53:21] <_mingau_> will this process write something?
[21-Aug-2009 08:53:28] <_mingau_> or only read wmi?
[21-Aug-2009 08:53:35] <jb> it only reads
[21-Aug-2009 08:54:11] <jb> but non-privledged users do not have full read access to the WMI namespace
[21-Aug-2009 08:56:43] <_mingau_> hunnn
[21-Aug-2009 08:57:37] <ckrough> yay: http://webaj.com/how-setup-splunk-monitoring-zenoss-logs.htm
[21-Aug-2009 08:57:48] <ckrough> was just going to start writing recipes for that
[21-Aug-2009 08:58:21] <ckrough> meh, doesnt look like he did fancy recipes. I guess Im still doing it...
[21-Aug-2009 08:59:44] <mrayzenoss> ckrough: check the date on that article, don't get too attached to the details
[21-Aug-2009 09:00:01] <ckrough> There arent any details in that article...
[21-Aug-2009 09:00:06] <mrayzenoss> ugh, spammers responding to spammers
[21-Aug-2009 09:01:47] <ckrough> I havent looking into it much, but I assume I'll need to write some tranforms to clean up the data once splunk snarfs it
[21-Aug-2009 09:02:06] <mrayzenoss> there was someone from splunk on here the other day, I believe they use Zenoss
[21-Aug-2009 09:02:08] <ckrough> splunk transforms that is, not zenoss transforms
[21-Aug-2009 09:02:32] <ckrough> Im tired of browsing through all the text files on the collectors, time to consolidate that mess
[21-Aug-2009 09:03:51] <rmatte> mrayzenoss: I just sent an email to you with my ZenPack details
[21-Aug-2009 09:09:19] <ckrough> someone forgot to put a cabinet row into maintenance mode last night during a maint window. It took almost 4 hours for the SMS pages to get caught up
[21-Aug-2009 09:09:53] <jb> yikes
[21-Aug-2009 09:11:45] <jb> http://pastebin.com/m18889c9b
[21-Aug-2009 09:11:51] <jb> anyone have a clue on how to properly format this transform?
[21-Aug-2009 09:13:02] <ckrough> jb: this comes in via zentrap?
[21-Aug-2009 09:13:17] <jb> yeah
[21-Aug-2009 09:13:28] <jb> i just dont know how to format evt.message properly
[21-Aug-2009 09:13:43] <jb> i'd like each of those on a new line
[21-Aug-2009 09:15:37] <rmatte> you might be able to use \n linebreaks
[21-Aug-2009 09:15:53] <rmatte> like "firstline\nsecondline\nthirdline"
[21-Aug-2009 09:16:07] <rmatte> but I've never tried it in a transform
[21-Aug-2009 09:18:26] <rmatte> >>> var1 = 'hey\nthere'
[21-Aug-2009 09:18:26] <rmatte> >>> print var1
[21-Aug-2009 09:18:26] <rmatte> hey
[21-Aug-2009 09:18:26] <rmatte> there
[21-Aug-2009 09:18:27] <rmatte> >>>
[21-Aug-2009 09:18:33] <jb> yeah
[21-Aug-2009 09:18:43] <ckrough> is that how its handled internall though?
[21-Aug-2009 09:18:54] <rmatte> no idea
[21-Aug-2009 09:19:05] <rmatte> you'd have to try it out and see
[21-Aug-2009 09:19:35] <jb> any line breaks makes the transform invalid
[21-Aug-2009 09:20:38] <rmatte> ah
[21-Aug-2009 09:20:40] <ckrough> try \\n? maybe it needs to be escaped so their code doesnt choke
[21-Aug-2009 09:20:47] <jb> but im probably not doing the variables right
[21-Aug-2009 09:20:58] <jb> evt.message =
[21-Aug-2009 09:20:59] <jb> 'Machine Type / Model / Serial: %s' % (machineModelSerial)
[21-Aug-2009 09:21:02] <jb> even that makes it invalid
[21-Aug-2009 09:22:04] <rmatte> well, it needs to be right after the evt.message = probably
[21-Aug-2009 09:22:17] <rmatte> I don't think you can do it the way you're trying to
[21-Aug-2009 09:22:22] <jb> yeah, me either
[21-Aug-2009 09:23:46] <ckrough> You are not alone http://forums.zenoss.com/viewtopic.php?p=14871#14871
[21-Aug-2009 09:23:54] <rmatte> yeh, I just saw that post
[21-Aug-2009 09:24:43] <jb> hm yep
[21-Aug-2009 09:25:42] <jb> on a side note.. any of you guys pulling I/O from server HBAs?
[21-Aug-2009 09:25:57] <ckrough> yes
[21-Aug-2009 09:26:10] <jb> windows hosts?
[21-Aug-2009 09:26:17] <ckrough> no, RHEL
[21-Aug-2009 09:26:30] <ckrough> on dells with OpenManage, which dumps into snmp
[21-Aug-2009 09:26:37] <jb> ok
[21-Aug-2009 09:26:38] <jb> exactly what I need
[21-Aug-2009 09:26:46] <jb> i think the dellmon zenpack that im using breaks that
[21-Aug-2009 09:27:08] <jb> so openmanage provides I/O for everything? perc attached local disks as well as fiber attached storage?
[21-Aug-2009 09:27:09] <ckrough> I had to modify something, lets see if I can figure out what it was... been a while
[21-Aug-2009 09:27:25] <ckrough> Im not using FAS, so not sure on that.
[21-Aug-2009 09:32:24] <ckrough> Im using zenoss.snmp/UCDHardDiskMap to get it I think
[21-Aug-2009 09:33:04] <ckrough> I remember having to modify a Dell specifiz plugin at one point, but we abandoned those changes on the last upgrade. Now that I look, not all devices are getting IO
[21-Aug-2009 09:33:57] <jb> ah ok
[21-Aug-2009 09:34:14] <jb> im wondering if i can get these stats directly from the HBA
[21-Aug-2009 09:34:28] <ckrough> via?
[21-Aug-2009 09:34:43] <jb> not sure
[21-Aug-2009 09:34:48] <jb> maybe snmp?
[21-Aug-2009 09:34:57] <jb> probably not
[21-Aug-2009 09:35:50] <jb> probably easier to pull it from OM/windows
[21-Aug-2009 09:44:37] <jb> anyone here using egor's DellMon pack?
[21-Aug-2009 09:52:41] RoninX341 is now known as etank
[21-Aug-2009 09:53:15] <gwb2351> chudler_: the sentryCDU community zenpack you're developing is working great!
[21-Aug-2009 09:58:24] <chudler_> gwb2351: happy to hear it. It still has some rough edges... I will be improving those Some Day Soon
[21-Aug-2009 10:08:28] <gwb2351> the only thing I've modified are adding a "10,/" to the RDN of the graphs. the only things I'm having problems with is some "aggregate load" calculations are returning negative numbers (verified that snmpget of the load OID is providing valid input, but haven't gotten any farther than that).
[21-Aug-2009 10:09:28] <chudler_> I am tracking some bugs related to non-updating values, and the one you mentioned about the "centiamps". That other one you mentioned, I hope is related to the failure to update RRD values under certain conditions.
[21-Aug-2009 10:11:35] <gwb2351> ah... is the RRD failure to update this bug (http://dev.zenoss.org/trac/ticket/5306) ?
[21-Aug-2009 10:13:33] <gwb2351> i'll try the workaround on the few devices that aren't reporting properly and see if that fixes them
[21-Aug-2009 10:13:48] <eidolon> hey do classes such as Devices/Storage/Filer do an automatic remodel every 6 hours or so? we had a problem where alerts were going off, but the filesystem had been resized. the docs imply a remodel happens automatically every few hours.
[21-Aug-2009 10:13:55] <eidolon> but this one didn't - i remodelled by hand, and it cleared the condition.
[21-Aug-2009 10:14:10] <eidolon> (target machine is a netapp filer, obviously)
[21-Aug-2009 10:18:48] <mrayzenoss> eidolon: default remodeling cycle is 720 minutes, set on the collector (every 12 hours)
[21-Aug-2009 10:20:39] <eidolon> hyum. okay, mray. i need to dig and see if that's working
[21-Aug-2009 10:20:40] <eidolon> thanks
[21-Aug-2009 10:21:43] <TBKDan> I want to have two alerting periods: one during working hours (m-f 8-5) and one for the rest of the time. The 8-5 would be immediate alerting, and the rest of the time would be on a 5 minute delay. Right now I have the two rules set up, but if a machine goes down for more than 5 minutes, i get two alerts (one immediately and one in 5 minutes). Is there a way to just disable that 5 min rule during 8-5 or do I have to set up individual rules for weekdays and
[21-Aug-2009 10:24:32] <mrayzenoss> TBKDan: You should be able to configure your Alert Schedule that way
[21-Aug-2009 10:24:49] <mrayzenoss> so you already have one in place
[21-Aug-2009 10:25:13] <mrayzenoss> and you can edit the scedule details
[21-Aug-2009 10:25:45] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch02s06.html
[21-Aug-2009 10:29:31] <TBKDan> mrayzenoss: I guess to simplify the question a bit... can I tell a rule to be *inactive* for a time period and active for the rest? Right now the way I understand it is I can tell it to be active for a time period and inactive for the rest.
[21-Aug-2009 10:30:07] <mrayzenoss> TBKDan: yeah, that's what the schedule would give you
[21-Aug-2009 10:33:57] <TBKDan> mrayzenoss: So if I create an active period of M-F 8-5 and set enabled to false, does that mean it will not alert during that time but will the rest of the time?
[21-Aug-2009 10:34:55] <mrayzenoss> no, I think what you want to do is set the active period from 5pm-8am M-F
[21-Aug-2009 10:35:03] <mrayzenoss> (and the weekend)
[21-Aug-2009 10:35:18] <mrayzenoss> enabled is just if you want to turn it back on later
[21-Aug-2009 10:35:40] <TBKDan> Arg, ok.. that's what I was hoping I wouldn't have to do
[21-Aug-2009 10:37:08] <mrayzenoss> but if you set your "rest of the time" to be the default setting and then have the 8-5 setting to be scheduled, that may work
[21-Aug-2009 10:37:33] <mrayzenoss> so "rest of time" is 1 page and 8-5 is 2 pages?
[21-Aug-2009 10:38:10] <TBKDan> That's what I was trying to eliminate. Right now we get duplicate pages during business hours.
[21-Aug-2009 10:38:17] <mrayzenoss> ahh, hmmm.
[21-Aug-2009 10:38:37] <mrayzenoss> I guess our alerting could be smarter
[21-Aug-2009 10:38:50] <TBKDan> Additionally, in the "Repeat" section of the schedule, weekend is not an option
[21-Aug-2009 10:38:53] <mrayzenoss> the reason it's not is that we allow you to set alert schedules all over the place
[21-Aug-2009 10:39:27] <mrayzenoss> so some alerts go to different people at different times for different systems and groups, figuring out the overlap is non-trivial
[21-Aug-2009 10:40:22] <mrayzenoss> so I guess you have to do 4 alert schedules? 12am-8am, 8-5, 5-12pm and the weekend?
[21-Aug-2009 10:40:55] <jb> i do that..
[21-Aug-2009 10:41:00] <jb> after 5PM and weekends..
[21-Aug-2009 10:41:06] <jb> and then normal 6am-5PM
[21-Aug-2009 10:41:08] <jb> during the week
[21-Aug-2009 10:41:21] <TBKDan> For weekends you have two jobs, one for each day?
[21-Aug-2009 10:41:27] <jb> sec
[21-Aug-2009 10:41:28] <mrayzenoss> jb: does 5pm-8am weekdays work?
[21-Aug-2009 10:41:45] <mrayzenoss> jb: ie. does it pick up Monday at 12am?
[21-Aug-2009 10:42:19] <jb> oh..
[21-Aug-2009 10:42:19] <jb> hm
[21-Aug-2009 10:42:23] <jb> actually, i only do after 5PM
[21-Aug-2009 10:42:34] <jb> looks like I don't differentiate between weekdays/weekends
[21-Aug-2009 10:43:20] <TBKDan> Would adding "Weekend" repeats be difficult?
[21-Aug-2009 10:43:47] <TBKDan> I don't know if I missed something in the GUI or if it's somewhere deeper in the code
[21-Aug-2009 10:45:44] <TBKDan> brb
[21-Aug-2009 10:59:43] <jb> ok i think im going to buy this OIDView Professinal program
[21-Aug-2009 10:59:51] <jb> unless someone can reccomend a better free one?
[21-Aug-2009 11:00:18] <TBKDan> Back
[21-Aug-2009 11:03:11] <mrayzenoss> jb: which features do you really like about it?
[21-Aug-2009 11:03:19] <mrayzenoss> jb: I think I used that at a previous employer
[21-Aug-2009 11:03:34] <jb> well, it has multiple views which makes it really nice.. it has this LiveGrid thing
[21-Aug-2009 11:03:39] <jb> that allows you to view entire tables easily
[21-Aug-2009 11:03:47] <jb> and it auto-downloads/compiles any additional MIBs that it needs
[21-Aug-2009 11:04:04] <mrayzenoss> cool
[21-Aug-2009 11:04:10] <jb> its $299 fo a single user
[21-Aug-2009 11:04:34] <mrayzenoss> I don't have a good recommendation, I'm just looking for ideas to suggest for the mib-browser zenpack
[21-Aug-2009 11:04:45] <jb> ah
[21-Aug-2009 11:05:24] <mrayzenoss> Kells has been adding minor touch-ups here and there, but I've been pushing Brandon to adopt it into Core and beef it up
[21-Aug-2009 11:05:35] <mrayzenoss> that said, we need more developers
[21-Aug-2009 11:06:37] <mrayzenoss> I should probably rename the MIB browser the MIB Utils ZenPack
[21-Aug-2009 11:06:54] <mrayzenoss> since it does a little more than that
[21-Aug-2009 11:12:11] _mingau_ is now known as apocalipse
[21-Aug-2009 11:13:03] <apocalipse> hey people
[21-Aug-2009 11:14:41] <apocalipse> where can i find job postings list?
[21-Aug-2009 11:15:04] <mrayzenoss> http://www.zenoss.com/about/jobs
[21-Aug-2009 11:15:09] <mrayzenoss> that one?
[21-Aug-2009 11:15:50] <rmatte> hopefully he meant that one since this isn't a recruitment agency
[21-Aug-2009 11:16:01] <apocalipse>
[21-Aug-2009 11:20:45] <rmatte> mrayzenoss: I guess you have your work set out for you today with all the procrastinary last minute submissions hehe
[21-Aug-2009 11:21:15] <mrayzenoss> ugh
[21-Aug-2009 11:21:24] * rmatte is guilty
[21-Aug-2009 11:21:32] <mrayzenoss> so I just got an email from my boss saying he wanted the ZenPack contest wrapped up by Monday
[21-Aug-2009 11:21:40] <rmatte> ah
[21-Aug-2009 11:21:43] <mrayzenoss> I have 13 ZenPacks to process, which on average take 1 hour each
[21-Aug-2009 11:21:56] <chudler_> D'oh!
[21-Aug-2009 11:22:47] <chudler_> "um, I'm gonna need you to go ahead and come in on Saturday"
[21-Aug-2009 11:23:05] <rmatte> lol
[21-Aug-2009 11:23:56] <rmatte> ask if it can be extended to Tuesday or something
[21-Aug-2009 11:24:07] <rmatte> but make today the absolute final submission date
[21-Aug-2009 11:24:13] <rmatte> I doubt 1 day will hurt anything
[21-Aug-2009 11:25:17] <mrayzenoss> the final submission date was last Friday
[21-Aug-2009 11:26:24] <mrayzenoss> I've made a few exceptions for ZenPacks that would have come in if I'd been around to bug people, but I was on vacation
[21-Aug-2009 11:27:14] <rmatte> ah
[21-Aug-2009 11:42:41] sakura_ is now known as x-ip
[21-Aug-2009 11:50:38] <markeriv> hello everyone
[21-Aug-2009 11:51:24] <markeriv> anyone around? I am trying to decide if I should install zenoss on my new server today or if I should wait for 2.4.3, and what additional features/fixes might be in that version
[21-Aug-2009 11:52:42] <mrayzenoss> markeriv: here's the list of closed tickets for 2.4.3: http://dev.zenoss.org/trac/report/6
[21-Aug-2009 11:53:05] <mrayzenoss> once all the "verification" ones move to "closed" the release will be ready
[21-Aug-2009 11:53:34] <mrayzenoss> that said, our QA Manager crashed his bike this morning and is missing the day after having to get stiches
[21-Aug-2009 11:53:41] <markeriv> ouch
[21-Aug-2009 11:54:36] <mrayzenoss> so no tickets are being added at this point, unless they're insanely bad
[21-Aug-2009 11:54:52] <rmatte> yikes
[21-Aug-2009 11:55:00] <rmatte> motorbike?
[21-Aug-2009 11:55:10] <mrayzenoss> no, just a mtn. bike
[21-Aug-2009 11:55:14] <rmatte> ah
[21-Aug-2009 11:55:19] <rmatte> what did he crash it in to?
[21-Aug-2009 11:55:20] <rmatte> lol
[21-Aug-2009 11:56:07] <markeriv> sigh this new quadcore xeon 2.4 is purring on my desk with its 8 gigs of ram and making me drool over the fact that I have not installed anything on it yet
[21-Aug-2009 11:56:33] <mrayzenoss> a source build would probably be 2.4.3 unless QA rejects any of the tickets
[21-Aug-2009 11:56:55] <chudler_> if you just install SSH with some easy to guess passwords, you wont have to go through the trouble of installing anything. Someone will do it for you
[21-Aug-2009 11:57:01] <victor-> i'm not sure what I've changed, but it seems like zen.Device is unable to add devices via hostname anymore. I have a script that uses xmlrpc that is now failing, and i'm getting the same result in the web gui when adding a device by name. from the host CLI, I can ping the devices, so that's working properly. any ideas of what i should check?
[21-Aug-2009 11:57:25] <markeriv> lol chudler
[21-Aug-2009 11:57:58] <chudler_> that's what I call outsourcing
[21-Aug-2009 11:58:15] <mrayzenoss> I call that "cloud computing"
[21-Aug-2009 12:00:37] <victor-> zenoss's event.log shows an error like: INFO zen.Device device name 'hostname.new' for ip ''
[21-Aug-2009 12:01:01] <mrayzenoss> victor-: so DNS isn't finding the machine's IP?
[21-Aug-2009 12:04:41] <victor-> mrayzenoss: the linux system DNS resolution works fine - I can run 'nslookup hostname.new', etc, and it resolves. it's just zenoss that seems to be having trouble.
[21-Aug-2009 12:07:13] <mrayzenoss> victor-: weird, I assume if you add the device by IP address, it then finds the right name?
[21-Aug-2009 12:07:29] <mrayzenoss> but if you add it by hostname, it doesn't work...?
[21-Aug-2009 12:08:12] <geniusj> Hey all.. I have what hopefully is a quick question.. I have a cacti script that I'm trying to turn into a performance monitor in zenoss. I get the script added, it tests fine, etc. But it doesn't seem to generate any actual data. I'm wondering if it's a data points issue or something? I named the data points after the key in the key/value pairs that the cacti script spits out. e.g. num_gets: 5 - data point name = num_gets
[21-Aug-2009 12:09:01] <geniusj> I'm wondering if anyone has gotten cacti scripts successfully imported. I didn't see much online about it other than that it supports it
[21-Aug-2009 12:11:46] <ckrough> How can I get my collectors to reconnect to the master gracefully? The main box locked up and I had to reboot, now I'll need to convince the collectors to reconnect, but I have to stagger them so they dont overwhelm it
[21-Aug-2009 12:13:30] <mrayzenoss> geniusj: I haven't migrated Cactii plugins, but let's just assume your script is working and debug it like a Nagios plugin...
[21-Aug-2009 12:13:36] <mrayzenoss> geniusj: is the template bound?
[21-Aug-2009 12:13:49] <mrayzenoss> geniusj: do you get no graphs, or graphs that don't fill in with any data?
[21-Aug-2009 12:15:29] <mrayzenoss> ckrough: I'm guessing support will have better answers
[21-Aug-2009 12:15:37] <ckrough> thanks
[21-Aug-2009 12:15:53] <mrayzenoss> ckrough: I know there's a lot of work going into stability and recovery for distributed collectors in King Crab
[21-Aug-2009 12:16:43] <ckrough> I keep forgetting that distributed collectors is only ent
[21-Aug-2009 12:21:44] <victor-> ah, found my problem to be that zenoss/bin/zensocket wasn't owned by root:root.
[21-Aug-2009 12:22:01] <mrayzenoss> victor-: source build? gotta remember that step
[21-Aug-2009 12:22:24] <geniusj> mrayzenoss: I get graphs, they just dont' fill with data
[21-Aug-2009 12:23:02] <mrayzenoss> geniusj: so that indicates the template is bound and the graph defined, but something about the data source is off
[21-Aug-2009 12:23:27] <geniusj> mrayzenoss: Yea. What's the best way to debug that?
[21-Aug-2009 12:23:53] <mrayzenoss> based on a similar discussion yesterday, get someone in here to debug your output for you
[21-Aug-2009 12:23:57] <geniusj> mrayzenoss: most of the debug methods I've found just show you the output of the command, which doesn't help me.. I want to see what it lookks like after it's parsed
[21-Aug-2009 12:24:14] <mrayzenoss> geniusj: what does the output look like when you hit "Test"
[21-Aug-2009 12:24:28] <geniusj> let me pull it up
[21-Aug-2009 12:25:27] <geniusj> uptime:598850 cmd_get:8753723 curr_connections:5 connection_structures:297 rusage_user:389.422798 total_connections:52334 cmd_set:1575591 time:1250875514 bytes_written:21919121415 get_misses:8107930 rusage_system:461.534836 bytes_read:22351049321 get_hits:645793
[21-Aug-2009 12:26:42] <mrayzenoss> so are you graphing any of those?
[21-Aug-2009 12:27:02] <geniusj> yea.. cmd_get and cmd_set for now
[21-Aug-2009 12:27:27] <geniusj> they're type COUNTER (hopefully that's correct)?
[21-Aug-2009 12:27:29] <straterra> I showed my CIO Zenoss..and he wants me to configure it for his other company
[21-Aug-2009 12:27:33] <mrayzenoss> I meant, are they working?
[21-Aug-2009 12:27:42] <geniusj> oh.. no, not working
[21-Aug-2009 12:27:53] <mrayzenoss> geniusj: and the parser is set to Cactii or Auto?
[21-Aug-2009 12:27:57] <geniusj> yup
[21-Aug-2009 12:28:12] <mrayzenoss> straterra: excellent
[21-Aug-2009 12:28:19] <straterra> Just core..
[21-Aug-2009 12:28:29] <mrayzenoss> BTW, we'll have a blog post up today about free training at Ohio LinuxFest
[21-Aug-2009 12:28:35] <straterra> Oh, cool
[21-Aug-2009 12:28:40] <straterra> I'm in Indianapolis..so..
[21-Aug-2009 12:28:56] <mrayzenoss> straterra: then I'll see you there
[21-Aug-2009 12:29:01] <straterra> lol..maybe
[21-Aug-2009 12:29:22] <straterra> quick question..using snmp traps, I could like..have a printer notify Zenoss when it was running low on ink or something, right?
[21-Aug-2009 12:29:37] <mrayzenoss> straterra: if your printer is cool enough to send traps, yes
[21-Aug-2009 12:29:52] <straterra> It may not be
[21-Aug-2009 12:29:59] <straterra> I'm not even sure it does SNMP at all..but it may
[21-Aug-2009 12:30:02] <mrayzenoss> straterra: or check this out and add a threshold: http://www.zenoss.com/community/projects/zenpacks/printertoner
[21-Aug-2009 12:30:04] <geniusj> mrayzenoss, I'm going to try one other thing.. I'm wondering if there has to be a datapoint defined for every single key/value pair that's output by the command, because there currently isn't
[21-Aug-2009 12:30:43] <straterra> Oh, VERY nice
[21-Aug-2009 12:30:54] <straterra> We have primarily xerox and hp printers
[21-Aug-2009 12:32:09] <Belial_> 18:07:14 <@mrayzenos> victor-: weird, I assume if you add the device by IP address, it then finds the right name?
[21-Aug-2009 12:32:23] <Belial_> Apologies, Mispaste. Silly Putty\
[21-Aug-2009 12:39:45] <geniusj> hm.. no luck I think
[21-Aug-2009 12:43:30] <markeriv> lol
[21-Aug-2009 12:43:36] <markeriv> i actually like that feature
[21-Aug-2009 12:43:43] <markeriv> wish I knew how to turn it on for fedora
[21-Aug-2009 12:44:42] * mrayzenoss off to lunch
[21-Aug-2009 12:58:20] <twm1010> afternoon
[21-Aug-2009 13:17:29] <jb> it returns values
[21-Aug-2009 13:34:57] <straterra> Is there anyway to give a device name to something that isn't in DNS?
[21-Aug-2009 13:37:49] <straterra> mrayzenoss: Some..funky stuff going on after I install that zenpack. Zenoss just asked me to send an error report
[21-Aug-2009 13:38:13] <rmatte> straterra: just add the device by IP and then rename it
[21-Aug-2009 13:38:16] <rmatte> there's a rename function
[21-Aug-2009 13:38:59] <straterra> mrayzenoss: when I tried to remove community.snmp.PrinterTonerMap from the device heirarchy
[21-Aug-2009 13:39:03] <straterra> rmatte: alrighty
[21-Aug-2009 13:39:46] <straterra> oh..I have to restart zope
[21-Aug-2009 13:39:48] <straterra> That's cool
[21-Aug-2009 14:01:59] <geniusj> Looks like my ZenCommand ZenRRD task might not even attempt to run at all. Though I can test it, etc. Anything I need to verify or need to have to make sure that it actually runs? I'm fairly new to zenoss. The data source is enabled..
[21-Aug-2009 14:04:47] <rmatte> well, I think I finally have my Zenoss upgrade/standardization steps worked out, now I just need to test them on Monday for the deployment Tuesday
[21-Aug-2009 14:04:48] <mrayzenoss> check the path? make sure that whatever the command is works with the zenoss user
[21-Aug-2009 14:05:55] <mrayzenoss> there's a bug about the 'Test' button not using the same path as zencommand
[21-Aug-2009 14:06:08] <geniusj> I have zencommand running in debug mode
[21-Aug-2009 14:06:16] <geniusj> or with debug logging rather
[21-Aug-2009 14:06:18] <rmatte> yeh, which basically makes the "test" useless for the time being
[21-Aug-2009 14:06:18] <mrayzenoss> do you see the command get called?
[21-Aug-2009 14:06:22] <geniusj> but I don't even see it mentioned
[21-Aug-2009 14:06:33] <mrayzenoss> try using the full path in the data source
[21-Aug-2009 14:06:34] <geniusj> all I see is this over and over: 2009-08-21 12:05:41 DEBUG zen.zencommand: Queueing event {'manager': 'localhost', 'timeout': 180, 'device': 'localhost', 'eventClass': '/Heartbeat', 'component': 'zencommand', 'agent': 'zencommand'}
[21-Aug-2009 14:06:40] <geniusj> ok.. will do
[21-Aug-2009 14:07:29] <geniusj> do I have to restart zencommand or anything after I change the path?
[21-Aug-2009 14:07:49] <geniusj> and btw, what I did before was set zCommandPath I believe
[21-Aug-2009 14:09:55] <mrayzenoss> I dunno, I seem to remember Chet saying to blank it out, since it usually screws up
[21-Aug-2009 14:10:07] <mrayzenoss> and to just stick with absolute paths when possible
[21-Aug-2009 14:10:24] <rmatte> yeh, absolute paths work much better
[21-Aug-2009 14:10:31] <rmatte> at the moment anywyas
[21-Aug-2009 14:10:53] <geniusj> ok
[21-Aug-2009 14:13:33] <geniusj> would the component or event key fields prevent it from working? Right now they're set fairly arbitrarily
[21-Aug-2009 14:13:45] <geniusj> or event class for that matter?
[21-Aug-2009 14:14:47] <straterra> And I'm having another issue with that printertonermap zenpack
[21-Aug-2009 14:14:55] <straterra> User-supplied Python expression (here.maxToner * .05) for minimum value caused error: ['toner_toner']
[21-Aug-2009 14:15:18] <mrayzenoss> geniusj: those shouldn't cause any issues
[21-Aug-2009 14:15:58] <mrayzenoss> geniusj: what are you monitoring?
[21-Aug-2009 14:16:06] <geniusj> memcachedb
[21-Aug-2009 14:16:38] <mrayzenoss> cool, but unfortunately not something I have running that I could test
[21-Aug-2009 14:17:00] <mrayzenoss> I don't know if there's an issue with the Cacti parsers, nothing in Trac about it
[21-Aug-2009 14:17:10] <geniusj> just to trace my steps:
[21-Aug-2009 14:17:14] <mrayzenoss> but my guess is something is weird about the way the output is being parsed
[21-Aug-2009 14:18:14] <geniusj> I created a template in my memcachedb server heirarchy, I added a COMMAND data source using a script that produces cacti style output. Chose cacti for parser, enabled service. Added data points. Then I bound the template to that same heirarchy
[21-Aug-2009 14:18:46] <geniusj> that's about it.. so the area for the graph is there for each server, just nog raph itself
[21-Aug-2009 14:18:55] <geniusj> not sure if I missed any steps
[21-Aug-2009 14:19:09] <mrayzenoss> yeah, that sounds right
[21-Aug-2009 14:19:11] <straterra> mrayzenoss: Do I need to like..manually add the OID's for max?
[21-Aug-2009 14:19:58] <geniusj> oh.. I also added a custom field for memcachedbPort
[21-Aug-2009 14:20:04] <mrayzenoss> I don't think so, I've used the ZenPack myself, let me go see what it's doing
[21-Aug-2009 14:20:06] <geniusj> wel.. cMemcachedbPort I believe
[21-Aug-2009 14:20:38] <mrayzenoss> geniusj: maybe try creating a new template and modify the command to do just 1 data point and see if that works
[21-Aug-2009 14:20:54] <mrayzenoss> if 1 works, then try ramping it up?
[21-Aug-2009 14:20:56] <rmatte> straterra: do you have the collector plugin from that ZenPack bound to the class the device is under?
[21-Aug-2009 14:21:14] <rmatte> straterra: it looks like it's expecting to find a value in here.maxToner but it's not finding it at all
[21-Aug-2009 14:21:28] <straterra> rmatte: Yeah..
[21-Aug-2009 14:21:36] <straterra> The collector plugin is bound
[21-Aug-2009 14:21:42] <rmatte> hmmm
[21-Aug-2009 14:21:45] <straterra> Is there any way to edit/take a look at that plugin?
[21-Aug-2009 14:21:56] <rmatte> I've never used that ZenPack, so don't really know what to tell you
[21-Aug-2009 14:21:57] <mrayzenoss> the source is available
[21-Aug-2009 14:22:19] <mrayzenoss> trac-zenpacks/browser/zenpacks/ZenPacks.community.PrinterToner
[21-Aug-2009 14:22:28] <mrayzenoss> I know there's someone working on it in the zenoss-zenpacks forum
[21-Aug-2009 14:22:31] <straterra> mrayzenoss: You're trying to sabotage me
[21-Aug-2009 14:22:32] <rmatte> mrayzenoss: when an upgrade is performed, does it overwrite the existing rpt files?
[21-Aug-2009 14:22:40] <rmatte> and report plugins, for that matter
[21-Aug-2009 14:23:22] <straterra> I suppose I can get by for now by creating a local copy and just hard coding the number (not use percentages)
[21-Aug-2009 14:23:28] <mrayzenoss> rmatte: if it's not in a ZenPack, I'd be nervous about it
[21-Aug-2009 14:23:53] <mrayzenoss> I don't know where modified reports are stored
[21-Aug-2009 14:23:57] <rmatte> mrayzenoss: not really that easy to store reports in ZenPacks lol, I guess I'll have to look in to it
[21-Aug-2009 14:24:10] <mrayzenoss> there are 3 report ZenPacks
[21-Aug-2009 14:24:38] <rmatte> for custom reports it's simple, but for modified .rpt files it's not as simple as just "add to zenpack"
[21-Aug-2009 14:24:48] <mrayzenoss> ahh, yeah
[21-Aug-2009 14:25:02] <straterra> God..python makes no sense to me
[21-Aug-2009 14:25:03] <mrayzenoss> I'd say take one of the existing report ZenPacks and replace the .rpt file in there
[21-Aug-2009 14:25:11] <rmatte> I've modified the memory.py, and cpu.py files under report plugins
[21-Aug-2009 14:25:11] <straterra> trac-zenpacks/browser/zenpacks/ZenPacks.community.PrinterToner/ZenPacks/community/PrinterToner/PrinterToner.py
[21-Aug-2009 14:25:21] <rmatte> and I intend to modify the rpt files themselves at some point
[21-Aug-2009 14:25:48] <rmatte> yeh, I'll look in to doing something like that
[21-Aug-2009 14:26:35] <geniusj> Hmm.. When I try to add a Service, the port field doesn't show up
[21-Aug-2009 14:26:49] <rmatte> geniusj: an IP service?
[21-Aug-2009 14:27:23] <geniusj> rmatte, yea. under registered
[21-Aug-2009 14:27:39] <rmatte> geniusj: oh, you're trying to actually add one to the list, or to a device?
[21-Aug-2009 14:27:42] <geniusj> it does show up for existing registered services.. just not this new one
[21-Aug-2009 14:27:45] <geniusj> add one to the list
[21-Aug-2009 14:28:41] <markeriv> What is the best ping uptime monitoring zenpack?
[21-Aug-2009 14:28:47] <markeriv> iyo?
[21-Aug-2009 14:29:17] <straterra> rmatte: its funny..cause under PrinterToner it shows the max toner o.O
[21-Aug-2009 14:29:21] <geniusj> I guess I can just add it from /manage, but I figured I'm missing something
[21-Aug-2009 14:30:09] <rmatte> geniusj: hmmm, you're right
[21-Aug-2009 14:30:15] <rmatte> just tried it myself
[21-Aug-2009 14:30:20] <straterra> And the plugin seems to work without the template..and it shows a warning about low toner on this printer
[21-Aug-2009 14:30:37] <straterra> I'd just like to adjust when it triggers warning/error
[21-Aug-2009 14:30:58] <straterra> The only place I saw to do that is the template..but now the template isn't loaded for that device..but its getting those values from somewhere
[21-Aug-2009 14:34:15] <rmatte> hmmm, I might have to log a trac case for this, that's weird
[21-Aug-2009 14:34:33] <rmatte> when you add a new IP Service under Registered it doesn't allow you to set the port of the service
[21-Aug-2009 14:34:57] <rmatte> oh wait, I might know why
[21-Aug-2009 14:35:31] <geniusj> err
[21-Aug-2009 14:35:32] <geniusj> fixed
[21-Aug-2009 14:35:33] <geniusj> lol
[21-Aug-2009 14:35:37] <geniusj> Add IP Service
[21-Aug-2009 14:35:40] <geniusj> I clearly wasn't looking
[21-Aug-2009 14:35:42] <rmatte> yeh, you clicked the wrong link
[21-Aug-2009 14:35:42] <rmatte> lol
[21-Aug-2009 14:35:46] <rmatte> I just noticed that
[21-Aug-2009 14:38:09] <mrayzenoss> geniusj: does your command show up when you run zencommand run -v10?
[21-Aug-2009 14:38:17] <geniusj> mrayzenoss, no
[21-Aug-2009 14:38:45] <geniusj> I'm about to move it from device to service though.. and failing that, I"ll recreate it as you suggested, starting with 1 property
[21-Aug-2009 14:39:04] <geniusj> err.. data point
[21-Aug-2009 14:41:40] <mrayzenoss> geniusj: so if you run zencommand run -v10 -d yourmachine, there's nothing in the log?
[21-Aug-2009 14:43:28] <geniusj> http://pastebin.com/m573d7ff4
[21-Aug-2009 14:44:44] <mrayzenoss> yup, that's empty
[21-Aug-2009 14:51:51] <mrayzenoss> geniusj: does your script call $devname?
[21-Aug-2009 14:53:03] <cgibbons> hmm
[21-Aug-2009 15:02:37] <geniusj> the script itself?
[21-Aug-2009 15:02:52] <geniusj> or the command definition?
[21-Aug-2009 15:05:27] <geniusj> /opt/zenoss/common/libexec/memcachedb.py --port ${here/cMemcachedbPort} ${here/manageIp}
[21-Aug-2009 15:07:15] <markeriv> hey guys, i got a quick snmp question. I am getting this perf/snmp error
[21-Aug-2009 15:07:53] <markeriv> error reading value for Serial0.1 on IP (oid .1.3.6.1.2.1.2.2.1.20.4 is bad)
[21-Aug-2009 15:08:14] <markeriv> is this the whole needs a 0 at the end bug?
[21-Aug-2009 15:08:40] <markeriv> I have added like 6 1721 ciscos at this point and they are spamming me with this error every 5 mins
[21-Aug-2009 15:08:53] <mrayzenoss> that's the typical culprit
[21-Aug-2009 15:09:18] <markeriv> hmm, where would I go about editing that OID?
[21-Aug-2009 15:09:41] <rmatte> markeriv: that's because that's not a physical serial interface, it's a sub-interface
[21-Aug-2009 15:09:54] <rmatte> it's trying to read the errors but it can't since there are no errors, since it's a virtual interface
[21-Aug-2009 15:10:04] <markeriv> aha
[21-Aug-2009 15:10:17] <rmatte> you need to create a local copy of the interface template and delete the errors datapoint
[21-Aug-2009 15:10:31] <markeriv> sweet beans
[21-Aug-2009 15:10:41] <rmatte> you'll have to do that for every one of those interfaces if you have more
[21-Aug-2009 15:10:52] <rmatte> I had to do it for like 50 devices
[21-Aug-2009 15:10:54] <markeriv> *cries*
[21-Aug-2009 15:10:56] <rmatte> on one of my zenoss boxes
[21-Aug-2009 15:11:02] <markeriv> i got like
[21-Aug-2009 15:11:06] <markeriv> over 1k of these suckers
[21-Aug-2009 15:11:29] <markeriv> lol, back to the chain gang i suppose
[21-Aug-2009 15:11:31] <mrayzenoss> so some have the error and some don't?
[21-Aug-2009 15:13:03] <markeriv> No, i mean i have over a thousand of these routers deployed
[21-Aug-2009 15:13:11] <markeriv> and each one is having that virtual interface
[21-Aug-2009 15:13:23] <markeriv> maybe i can just not monitor that interface... hmmm
[21-Aug-2009 15:15:44] <rmatte> mrayzenoss: zenoss doesn't differentiate between physical serial interfaces and virtual
[21-Aug-2009 15:16:03] <rmatte> but you still want to collect error info for your physical
[21-Aug-2009 15:16:21] <rmatte> so you need to make local copies of the template for every single virtual interface and remove the errors datasource
[21-Aug-2009 15:16:25] <rmatte> it's very tedious
[21-Aug-2009 15:16:40] <rmatte> once it's setup it works fine, but it's pretty brutal to have to do
[21-Aug-2009 15:16:50] <mrayzenoss> seems like something you should be able to do in zendmd
[21-Aug-2009 15:17:12] <jb> hey rmatte
[21-Aug-2009 15:17:23] <jb> remember the other day when we were wondering about the RRD stuff?
[21-Aug-2009 15:17:27] <jb> storing values in RRD cache..
[21-Aug-2009 15:17:31] <rmatte> yeh
[21-Aug-2009 15:17:37] <markeriv> hmm, is there a way i can just choose to ignore that particular error?
[21-Aug-2009 15:17:39] <jb> well
[21-Aug-2009 15:17:46] <markeriv> so that it is not saved or logged globally?
[21-Aug-2009 15:18:01] <rmatte> mrayzenoss: maybe, I'll have to look in to it, it wouldn't be easy though
[21-Aug-2009 15:18:04] <jb> you do that when you want things in your modeler like "Status" or "Health" or something to be checked more often than 12 hours (default modeler time)
[21-Aug-2009 15:18:15] <rmatte> ah
[21-Aug-2009 15:18:15] <jb> since RRDs are generated every 5 minutes
[21-Aug-2009 15:18:21] <rmatte> gotcha
[21-Aug-2009 15:18:34] <jb> just fyi
[21-Aug-2009 15:24:09] <markeriv> hmm crup, there are interfaces on this router that i do need to monitor with that so i cannot just delete that graph
[21-Aug-2009 15:24:10] <markeriv> ponder
[21-Aug-2009 15:32:27] <markeriv> hmm well i figured out how to do what you suggested
[21-Aug-2009 15:32:30] <markeriv> at least now i can stop the flow
[21-Aug-2009 15:37:20] <jb> so if i add a parent device class to a zenpack
[21-Aug-2009 15:37:27] <jb> that will add all children classes too, right?
[21-Aug-2009 15:39:00] <markeriv> wootah, figured out a way to just not monitor that interface
[21-Aug-2009 15:39:27] <markeriv> same deal essentially, and more easily educatable so mr. script can do it himself
[21-Aug-2009 15:42:30] <rmatte> lovely, when I run my cpu report it looks fine but when I export it it somehow magically comes up with different values
[21-Aug-2009 15:43:35] <rmatte> omg, I think I know why...
[21-Aug-2009 15:43:47] <rmatte> it's like it's exporting between the default start and end time
[21-Aug-2009 15:43:59] <rmatte> even though a different start and end time were specified
[21-Aug-2009 15:48:05] <rmatte> it's like the export is completely ignoring the options
[21-Aug-2009 15:49:03] <markeriv> thats not comfy
[21-Aug-2009 15:50:29] <markeriv> is it happening for all reports?
[21-Aug-2009 15:50:32] <rmatte> no, it sucks, because now my boss expects to fix it
[21-Aug-2009 15:50:34] <markeriv> or just cpu?
[21-Aug-2009 15:50:36] <rmatte> good question...
[21-Aug-2009 15:52:07] <markeriv> and in 2.4.2 or 2.4.3?
[21-Aug-2009 15:52:34] <rmatte> this particular box is 2.3.3
[21-Aug-2009 15:52:36] <rmatte> so I have no idea
[21-Aug-2009 15:52:55] <rmatte> I haven't moved any of these custom reports to other boxes yet
[21-Aug-2009 15:53:11] <rmatte> well, modified reports
[21-Aug-2009 15:53:17] <rmatte> though my modifications wouldn't cause this
[21-Aug-2009 15:53:20] <markeriv> try creating a new custom report with the CPU field?
[21-Aug-2009 15:54:03] <rmatte> no, I just added more datapoints to the report plugin
[21-Aug-2009 15:54:13] <rmatte> I didn't even touch the report file itself
[21-Aug-2009 15:54:33] <markeriv> got a backup of the original you can try?
[21-Aug-2009 15:55:14] <rmatte> I have the original on other servers, but like I said, my modifications can't have caused this
[21-Aug-2009 15:55:19] <rmatte> it's literally impossible
[21-Aug-2009 15:55:59] <markeriv> hmm
[21-Aug-2009 15:56:48] <rmatte> yup, it does the exact same thing on a report I haven't even touched
[21-Aug-2009 15:56:56] <rmatte> that's really bad
[21-Aug-2009 15:56:57] <rmatte> lol
[21-Aug-2009 15:57:10] <rmatte> new trac ticket
[21-Aug-2009 15:57:37] <rmatte> does no one ever use reports or something?
[21-Aug-2009 15:57:43] <rmatte> I'm finding bug after bug after bug lol
[21-Aug-2009 15:58:03] <ckrough> Im out. have a good weekend
[21-Aug-2009 16:10:54] <mrayzenoss> free Zenoss training at Ohio LinuxFest: http://blog.zenoss.com/2009/08/21/zenoss-community-day-september-25th/
[21-Aug-2009 16:12:20] <straterra> Boss thought it was pretty cool I'm graphing one of the printers now
[21-Aug-2009 16:12:23] <straterra> I got that zenpack to work
[21-Aug-2009 16:12:28] <mrayzenoss> cool
[21-Aug-2009 16:12:32] <rmatte> I think I'm going to end up just writing scripts to generate reports
[21-Aug-2009 16:12:48] <jb> ibm3584 zenpack is nearly complete
[21-Aug-2009 16:12:51] <jb> due to MUCH help from egor
[21-Aug-2009 16:12:52] <straterra> I think I'll write a report on toner usage and such
[21-Aug-2009 16:13:40] <jb> well, its mostly inventory/modeled information
[21-Aug-2009 16:13:42] <jb> oops
[21-Aug-2009 16:15:19] <markeriv> what other bugs are you finding rmatte?
[21-Aug-2009 16:16:04] <geniusj> mrayzenoss: the target class for my memcachedb template is Products.ZenModel.Device - is that fine?
[21-Aug-2009 16:16:20] <rmatte> besides the bug with exporting you mean?
[21-Aug-2009 16:17:22] <mrayzenoss> geniusj: sounds fine to me, as long as it's bound to the device classes you want
[21-Aug-2009 16:19:44] <rmatte> markeriv: I'm prepping some trac tickets right now, I'll paste them once I have them logged
[21-Aug-2009 16:20:11] <jb> so zenoss finally got some street cred today at work..
[21-Aug-2009 16:20:18] <jb> helped pinpoint a few MSSQL problems
[21-Aug-2009 16:21:44] <jb> (which MOM didnt)
[21-Aug-2009 16:22:09] <mrayzenoss> yay us!
[21-Aug-2009 16:23:06] <markeriv> cool
[21-Aug-2009 16:23:17] <markeriv> woot go Z
[21-Aug-2009 16:27:07] <rmatte> mrayzenoss: http://dev.zenoss.org/trac/ticket/5438
[21-Aug-2009 16:28:09] * mrayzenoss sighs
[21-Aug-2009 16:28:16] <markeriv> ?
[21-Aug-2009 16:28:31] <mrayzenoss> I just know how swamped all the devs are
[21-Aug-2009 16:28:36] <mrayzenoss> and we only have 5
[21-Aug-2009 16:29:05] <rmatte> yeh, I understand that part
[21-Aug-2009 16:31:53] <jb> hmm
[21-Aug-2009 16:31:56] <jb> so i have one instance of zenoss that has like 30 transforms
[21-Aug-2009 16:32:09] <jb> i need to take these and put them onto another instance of zenoss and integrate them into another zenpack
[21-Aug-2009 16:32:31] <jb> what would you say is the best way to do that? create a temporary zenpack and put the transforms in it?
[21-Aug-2009 16:33:24] <rmatte> http://dev.zenoss.org/trac/ticket/5439
[21-Aug-2009 16:33:29] <rmatte> that's another one I noticed today
[21-Aug-2009 16:34:09] <mrayzenoss> I only opened this one today: http://dev.zenoss.org/trac/ticket/5437
[21-Aug-2009 16:34:39] <rmatte> ah, that's a good one
[21-Aug-2009 16:34:45] <rmatte> probably a relatively quick fix too
[21-Aug-2009 16:34:57] <rmatte> just need to add a check in to the code
[21-Aug-2009 16:35:13] <mrayzenoss> yeah, cgibbons is refactoring the daemons a bit to share more code
[21-Aug-2009 16:35:22] <mrayzenoss> so it may get fixed for King Crab
[21-Aug-2009 16:36:18] <rmatte> the report export problem is pretty brutal, my boss is pretty upset about that so I'll need to come up with some other solution
[21-Aug-2009 16:36:42] <rmatte> I think I'll just write scripts that take arguments and dump the data as csv files as a workaround for the time being
[21-Aug-2009 16:37:45] <rmatte> unless I can figure out how to fix it
[21-Aug-2009 16:38:38] <markeriv> when ya fix it wanna let me know? lol.
[21-Aug-2009 16:38:49] <markeriv> hmm wonder if it affects my version to
[21-Aug-2009 16:38:54] <rmatte> if I do fix it I'll be submitting a patch
[21-Aug-2009 16:39:05] <markeriv> cool
[21-Aug-2009 16:39:07] <rmatte> I tested on 2.4.1 and it's the same issue
[21-Aug-2009 16:39:13] <rmatte> so I'd imagine you're affected as well
[21-Aug-2009 16:39:14] <markeriv> bleh
[21-Aug-2009 16:39:16] <rmatte> test it out
[21-Aug-2009 16:39:36] <rmatte> generate an interface utilization report, it'll show default values (which is average over the last 7 days)
[21-Aug-2009 16:39:44] <rmatte> then change the start and end time until the values are different
[21-Aug-2009 16:39:56] <rmatte> then export, and you should see the same values as you saw by default in the exported report
[21-Aug-2009 16:40:04] <rmatte> instead of the values shown in your browser
[21-Aug-2009 16:40:24] <rmatte> you can do it with cpu report, memory report, whatever
[21-Aug-2009 16:40:32] <rmatte> as long as you can get it to display different values
[21-Aug-2009 16:41:20] <rmatte> my boss called me over and was like "uhhh, Ryan, I don't think I can trust these numbers."
[21-Aug-2009 16:41:30] <rmatte> lol
[21-Aug-2009 16:43:17] <markeriv> bleh
[21-Aug-2009 16:43:27] <markeriv> i screwed up my data collection by moving devices down a step
[21-Aug-2009 16:43:34] <markeriv> think i may have fixed it but still meh
[21-Aug-2009 16:43:44] <rmatte> "down a step"?
[21-Aug-2009 16:44:04] <markeriv> yea
[21-Aug-2009 16:44:07] <markeriv> basically i took
[21-Aug-2009 16:44:12] <rmatte> like to a different organizer?
[21-Aug-2009 16:44:16] <markeriv> Devices/Network/Router/Cisco
[21-Aug-2009 16:44:24] <markeriv> and added /On-Net Product
[21-Aug-2009 16:44:29] <rmatte> ah
[21-Aug-2009 16:44:29] <markeriv> and all the graphs stopped working
[21-Aug-2009 16:44:42] <markeriv> half tempted just to move it back and say screw it
[21-Aug-2009 16:44:43] <rmatte> well yeh, because the template wouldn't be applied to that organizer by default
[21-Aug-2009 16:44:46] <markeriv> yea
[21-Aug-2009 16:44:46] <rmatte> you need to manually do that
[21-Aug-2009 16:45:03] <markeriv> i copied the template into the class
[21-Aug-2009 16:45:21] <rmatte> instead of copying you really should just bind the template from higher up to the class
[21-Aug-2009 16:45:28] <rmatte> that way you only have to update the template in 1 place
[21-Aug-2009 16:45:35] <markeriv> hmm nod
[21-Aug-2009 16:46:25] <markeriv> hmm, i am in devices/network/cisco
[21-Aug-2009 16:46:34] <markeriv> when i select the device template and go to bind templates
[21-Aug-2009 16:46:39] <markeriv> it only has the current directory
[21-Aug-2009 16:46:52] <markeriv> however I can copy it into the new directory /On-Net
[21-Aug-2009 16:46:53] <rmatte> hmmm
[21-Aug-2009 16:47:02] <markeriv> and from there bind it into the new directory
[21-Aug-2009 16:47:56] <rmatte> shows up fine for me
[21-Aug-2009 16:48:26] <markeriv> hmmm
[21-Aug-2009 17:06:14] <geniusj> mrayzenoss, are you out of annapolis? or are you guys all over the place?
[21-Aug-2009 17:06:41] <geniusj> I used to work with some of the original people there when I was at USi
[21-Aug-2009 17:06:49] <rmatte> they have an office in Austin as well
[21-Aug-2009 17:06:58] <geniusj> ah ok
[21-Aug-2009 17:06:58] <rmatte> and employees from all over
[21-Aug-2009 17:07:04] <rmatte> who work remote
[21-Aug-2009 17:07:10] <mrayzenoss> I'm in Austin
[21-Aug-2009 17:07:35] <mrayzenoss> 3.5/5 of the developers are in Austin
[21-Aug-2009 17:07:36] <rmatte> the austin office looks really nice
[21-Aug-2009 17:07:39] <geniusj> they wrote USi's monitoring system and then left to start ZenOSS
[21-Aug-2009 17:07:40] <rmatte> I want to steal your office
[21-Aug-2009 17:07:52] <mrayzenoss> cgibbons is in Houston but he comes to Austin a lot
[21-Aug-2009 17:08:29] <mrayzenoss> Ian is in Canada
[21-Aug-2009 17:08:51] <geniusj> no annapolis respresented on irc?
[21-Aug-2009 17:08:52] <mrayzenoss> they kinda want to keep developers in-house if possible
[21-Aug-2009 17:09:19] <mrayzenoss> well, Chet is in Annapolis and he's in here occasionally
[21-Aug-2009 17:09:33] <rmatte> yeh, chet hasn't been around much lately, I assume because he's busy
[21-Aug-2009 17:09:46] <mrayzenoss> npmccallum is in Kentucky, he works remote
[21-Aug-2009 17:10:05] <mrayzenoss> yeah, not sure what Chet's been up to, he may be on vacation
[21-Aug-2009 17:10:09] <rmatte> mrchippy is in austin I assume
[21-Aug-2009 17:10:25] <mrayzenoss> Yeah, mrchippy, jplouis, bedwards and cgibbons
[21-Aug-2009 17:10:59] <mrayzenoss> and the QA Manager and our build engineer
[21-Aug-2009 17:11:18] <mrayzenoss> most support and client services is in Annapolis or remote
[21-Aug-2009 17:16:55] <geniusj> holy crap
[21-Aug-2009 17:17:00] <geniusj> it's working
[21-Aug-2009 17:17:19] <geniusj> 1 monitor down, a bunch to go.. lol
[21-Aug-2009 17:17:28] <mrayzenoss> what's working? The Cacti plugin?
[21-Aug-2009 17:17:46] <geniusj> yea
[21-Aug-2009 17:17:51] <geniusj> I restarted zenoss
[21-Aug-2009 17:17:56] <geniusj> :-\
[21-Aug-2009 17:19:50] <geniusj> ok.. not quite WORKING yet.. but I am getting feedback.. it's at least running it
[21-Aug-2009 17:19:55] <geniusj> so it should be working shortly
[21-Aug-2009 17:20:45] <rmatte> alright, I'm out for the night, have a nice weekend guys
[21-Aug-2009 17:22:20] <geniusj> zenmodeler heartbeat failure
[21-Aug-2009 17:22:20] <geniusj> that's new :-D
[21-Aug-2009 17:22:39] <geniusj> rmatte, see ya
[22-Aug-2009 00:00:46] [disconnected at Sat Aug 22 00:00:46 2009]
[22-Aug-2009 00:00:46] [connected at Sat Aug 22 00:00:46 2009]
[22-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[22-Aug-2009 05:45:14] <prefixthis_> hello
[22-Aug-2009 05:45:52] <prefixthis_> is it possible to have zenoss not send all the syslog output from devices directly to the event history?
[22-Aug-2009 10:52:49] <prefixthis> i have syslog output from a switch into zenoss. zenoss delays the link down syslog messages until the interface comes back up, and instead displays a linkdown trap, and everything gets moved to the history after the interfaces comes back up
[22-Aug-2009 10:54:27] <prefixthis> and the trap itself only displays "snmp trap snmp_linkDown", without the port number (i have to check out the details to get that far). Anyone can give me a hint?
[22-Aug-2009 11:12:36] <Belial_>
[23-Aug-2009 00:00:46] [disconnected at Sun Aug 23 00:00:46 2009]
[23-Aug-2009 00:00:46] [connected at Sun Aug 23 00:00:46 2009]
[23-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[23-Aug-2009 04:33:35] <diazepam> hi all - im new to Zenoss and have just finished reading the docs but am still having trouble getting snmp on my ubuntu server to communicate with my zenoss system
[23-Aug-2009 04:34:26] <diazepam> snmpwalk -v1 -c rocommunity localhost - command works fine from the ubuntu server but
[23-Aug-2009 04:34:48] <diazepam> snmpwalk -v1 -c rocommunity 10.73.75.100 - times out from the zenoss machine
[23-Aug-2009 05:30:44] <diazepam> okay sorted that one out
[23-Aug-2009 05:30:59] <diazepam> needed to add my bridge interface to the snmpd conf
[23-Aug-2009 16:36:44] <darkest_night> hi @ all
[23-Aug-2009 16:37:01] <darkest_night> i have a question... how do i come to the zenoss-monitor?
[23-Aug-2009 16:37:12] <darkest_night> do i need any special website to enter?
[24-Aug-2009 00:00:46] [disconnected at Mon Aug 24 00:00:46 2009]
[24-Aug-2009 00:00:46] [connected at Mon Aug 24 00:00:46 2009]
[24-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[24-Aug-2009 07:26:07] <Disconnect> aahgrhght. wtf does it take to get zenoss alerting rules to stay disabled? i go in, set it to disabled, hit save. back to the main list, its disabled. but by tomorrow morning it'll be on again..
[24-Aug-2009 08:54:52] <rmatte> Disconnect: that's strange, never heard of that behaviour before
[24-Aug-2009 08:55:05] <rmatte> good morning Matt
[24-Aug-2009 08:55:39] <jb> so you cant add an entire event class to zenpack?
[24-Aug-2009 08:55:40] <jb> hrm.
[24-Aug-2009 08:55:45] <Disconnect> disabled it 3 times last week, came in to find it on again this morning. tried removing the disabled ones so we'll see what happens (it'll be a while anyway before zenoss is cleared to send mail.. its paging on discovered devices ..)
[24-Aug-2009 08:56:25] <jb> thats sort of annoying
[24-Aug-2009 09:00:32] <rmatte> Disconnect: add a filter to the alert rule not to alert on devices in device class /Discovered
[24-Aug-2009 09:00:34] <rmatte> that's what we did
[24-Aug-2009 09:00:52] <rmatte> alert filters are your friend
[24-Aug-2009 09:04:51] <scp2004> Hi Guys !!!!
[24-Aug-2009 09:07:05] <scp2004> I want to monitore a Linux Server in Zenoss .. And Collect some snmp informations .. How should I proceed ?
[24-Aug-2009 09:07:33] <scp2004> I installed the Zenpacks HPMonitor and DellMonitor . Its right ?
[24-Aug-2009 09:07:50] <jb> yes.. but you need to have OpenManage and Insight installed on those servers as well
[24-Aug-2009 09:08:34] <scp2004> I think also have. Cause they was monitored buy another zenoss ..And was ok
[24-Aug-2009 09:09:26] <scp2004> I add the device , a FreeBsd Linux Server running on a HP server in: /Devices /Server /Linux
[24-Aug-2009 09:09:55] <Disconnect> yah. we've got other sorting issues right now (old hosts that aren't doing snmp etc are filtered into their proper slots, etc) but yah. and there -really- needs to be a "copy this alert template"
[24-Aug-2009 09:14:13] <scp2004> And In /Devices /Server /Linux now i can see in zCollectorPlugins : HPCpuMap, DellCpuMap .... Its right now ?
[24-Aug-2009 09:14:21] <rocket> morning matt
[24-Aug-2009 09:14:30] <jb> yawn
[24-Aug-2009 09:14:36] <rocket> Hey jb
[24-Aug-2009 09:14:42] <jb> so when you add /Devices/Storage/IBM to a zenpack
[24-Aug-2009 09:14:51] <jb> hey rocket
[24-Aug-2009 09:15:03] <jb> WARNING:zen.AddToPack:Unable to find context path /zport/dmd/Devices/Storage (line 4 ?) for IBM
[24-Aug-2009 09:15:03] <jb> WARNING:zen.AddToPack:Not committing any changes
[24-Aug-2009 09:15:03] <jb> ERROR: zenpack command failed. Reason: Unable to create object using the following attributes:
[24-Aug-2009 09:15:03] <jb> * id: /zport/dmd/Devices/Storage/IBM
[24-Aug-2009 09:15:12] <jb> it wont just create it if it doesn't exist?
[24-Aug-2009 09:15:29] <mrayzenoss> morning all
[24-Aug-2009 09:16:17] <jb> or am I not adding it correctly?
[24-Aug-2009 09:16:42] <mrayzenoss> jb: is there already a /Devices/Storage class?
[24-Aug-2009 09:16:58] <jb> oh.
[24-Aug-2009 09:17:02] <jb> core doesn't have that
[24-Aug-2009 09:17:13] <mrayzenoss> nope
[24-Aug-2009 09:17:26] <jb> so my goal is to add /Devices/Storage/IBM/Tape Library
[24-Aug-2009 09:17:28] <mrayzenoss> it gets added by one of the zenpacks along the way
[24-Aug-2009 09:17:36] <jb> it wont just recursively create those classes?
[24-Aug-2009 09:19:27] <rocket> jb: nope
[24-Aug-2009 09:19:32] <jb> wow, thats annoying
[24-Aug-2009 09:19:41] <jb> so you have to tell your users to create them ahead of time?
[24-Aug-2009 09:19:47] <rocket> jb: I would create a core storage zenpack first
[24-Aug-2009 09:20:13] <rocket> jb: or you can add code to the init function that checks for it .. and if it doesnt exist create it
[24-Aug-2009 09:20:24] <jb> is that what you did for AIX?
[24-Aug-2009 09:21:02] <mrayzenoss> jb: yeah, I just put that as a prerequisite
[24-Aug-2009 09:21:17] <jb> a manual prereq?
[24-Aug-2009 09:22:33] <jb> i bet my zPythonClass zprop wont get exported either
[24-Aug-2009 09:22:37] <jb> so i'll need to set that in init
[24-Aug-2009 09:23:19] <mrayzenoss> jb: yeah, some ZenPacks handle creating their own device classes, some don't
[24-Aug-2009 09:23:30] <mrayzenoss> usually if you're going 2 or 3 deep, you'll have to create something
[24-Aug-2009 09:23:43] <jb> ok, I think I'll just do it in __init__
[24-Aug-2009 09:23:49] <jb> seems like a cleaner way
[24-Aug-2009 09:23:51] <mrayzenoss> in this case, you'll require that they add Storage, so it doesn't get deleted when they delete the ZenPack
[24-Aug-2009 09:24:18] <jb> hm, good point
[24-Aug-2009 09:26:46] <jb> ok, got another question
[24-Aug-2009 09:27:19] <jb> i have 30 or so odd transforms on one install, and I need to move those to another zenoss instance where I am packaging this zenpack.. whats the easiest way to do that?
[24-Aug-2009 09:31:00] <mrayzenoss> export them to another ZenPack?
[24-Aug-2009 09:31:06] <jb> yeah thats what i was thinking
[24-Aug-2009 09:31:18] <mrayzenoss> the '?' is because I've never done that
[24-Aug-2009 09:31:26] <jb> i think thats the best way
[24-Aug-2009 09:31:30] <jb> create a "temp" zenpack
[24-Aug-2009 09:31:43] <jb> install that on the dev instance, and add them to the production zenpack
[24-Aug-2009 09:36:12] <jb> hmm
[24-Aug-2009 09:37:02] <jb> WARNING:zen.AddToPack:Unable to find context path /zport/dmd/Events/HW/Backup/Tape Library/instances (line 4 ?) for alertChangerMechHWCritical
[24-Aug-2009 09:38:18] <jb> /Events/HW/Backup/Tape Library/ does exist
[24-Aug-2009 09:38:35] <jb> <jb> /Events/HW/Backup/Tape Library/ does exist
[24-Aug-2009 09:38:37] <jb> oops
[24-Aug-2009 09:38:40] <jb> ERROR: zenpack command failed. Reason: Unable to create object using the following attributes:
[24-Aug-2009 09:38:40] <jb> * id: /zport/dmd/Events/HW/Backup/Tape Library/instances/alertChangerMechHWCritical
[24-Aug-2009 09:39:41] <malbon> hi folks
[24-Aug-2009 09:39:56] <jb> but im guessing "instances" does not?
[24-Aug-2009 09:40:15] <malbon> I would like to take zenprocess and use it with a non-standard mib, any one have some words of wisdom to help me do this?
[24-Aug-2009 09:45:14] <jb> i guess that instances gets created when you add the first event mapping?
[24-Aug-2009 09:52:32] <rmatte> malbon: zenprocess has nothing to do with Mibs
[24-Aug-2009 09:52:56] <rmatte> malbon: The only daemon that uses Mibs in zentrap, to translate snmp traps
[24-Aug-2009 09:53:18] <rmatte> is zentrap* rather
[24-Aug-2009 10:23:34] <jb> cgibbons: would you know how I can add a device class/organizer (not using the GUI)?
[24-Aug-2009 10:23:46] <jb> i was thinking app.zport.dmd.Devices.addOrganizer...
[24-Aug-2009 10:25:51] crow is now known as ckrough
[24-Aug-2009 10:32:40] <malbon> rmatte: not entirely true as zenprocess uses the HostResources HRSWRun MIB for it's collection points.
[24-Aug-2009 10:33:05] <malbon> rmatte: I want to use an alternative set of OIDs within zenprocess (or a lookalike)
[24-Aug-2009 10:33:38] <jb> anyone?
[24-Aug-2009 10:43:39] <nzle> How does one install a perl module such as Net::Telnet on a Zenoss appliance? conary doesnt find it, and cpan is unable to build it. since there are not dev tools on it and conary seems unable to install them as well.
[24-Aug-2009 10:46:38] <nzle> I need to write a special purpose nagios style plugin that connects on a port, does some dialog, and interpets the response.
[24-Aug-2009 10:48:07] <mrayzenoss> nzle: if you could post that on zenoss-dev, I'll get our build engineer to respond. He's busy pushing out 2.4.3 right now, but if anyone knows the answer, it would be him
[24-Aug-2009 10:48:31] <mrayzenoss> nzle: I searched around some for it, but if it's not in conary I'm not sure how to add the trove
[24-Aug-2009 10:48:39] <nzle> Ok, Ill try that...
[24-Aug-2009 11:05:24] <scp2004> [mrayzenoss] Hi ..To collect snmp from a server in linux i need to install Community Zenpacks or Core Zenpacks ? I need snmp from FreeBsd devices in HP Server. So I installed Core Zenpack HPMonitor . Its right ?
[24-Aug-2009 11:07:44] <mrayzenoss> nah, SNMP works just fine for Linux out of the box. Just put it under /Server/Linux
[24-Aug-2009 11:10:58] <scp2004> [mrayzenoss] But , i also put. in /Server/Linux . Now I have the event: SNMP agent down
[24-Aug-2009 11:11:21] <scp2004> [mrayzenoss] but i know that its not down. Cause i have another zenoss monitoring it
[24-Aug-2009 11:17:28] <mrayzenoss> scp2004: is the community string correct?
[24-Aug-2009 11:18:02] <scp2004> [mrayzenoss] i put public . in other zenoss are public and its working
[24-Aug-2009 11:18:22] <mrayzenoss> scp2004: try re-modeling the device
[24-Aug-2009 11:18:25] <rocket> scp2004: check the zenoss logs for errors
[24-Aug-2009 11:19:21] <scp2004> [rocket] re-modeling ? how ido this ?
[24-Aug-2009 11:20:49] <mrayzenoss> go to the status page for that device, and select the page menu->Manage->Model Device
[24-Aug-2009 11:21:01] <mrayzenoss> that will show you the modeling run and any errors
[24-Aug-2009 11:22:20] <jb> <object id='/zport/dmd/Events/HW/Backup/Tape Library/instances/alertChangerMechHWDiag' module='Products.ZenEvents.EventClassInst' clas
[24-Aug-2009 11:22:21] <jb> oops
[24-Aug-2009 11:23:32] <scp2004> [mrayzenoss] doing ...
[24-Aug-2009 11:56:26] <venturaville> anyone got an example of uploading an icon through a zenpack for a new device class (without using hard paths)
[24-Aug-2009 11:59:07] <jb> working on that 9509 pack?
[24-Aug-2009 12:00:30] <jb> btw, do you just look in the dev guide for this stuff?
[24-Aug-2009 12:05:15] <rmatte> for starters, but some stuff isn't in the dev guide
[24-Aug-2009 12:05:18] <rmatte> it's not 100% yet
[24-Aug-2009 12:08:12] <venturaville> Cisco 9509?
[24-Aug-2009 12:08:18] <jb> yah
[24-Aug-2009 12:09:27] <venturaville> got a link to what you are working on?
[24-Aug-2009 12:09:52] <rocket> jb: community zenpacks are a great place to look for things too
[24-Aug-2009 12:10:11] <rocket> jb: eg I have them checked out to a directory so I can look at all of them
[24-Aug-2009 12:10:13] <jb> yeah, i usually just grep against the entire svn tree
[24-Aug-2009 12:10:14] <jb>
[24-Aug-2009 12:23:04] <ckrough> Is there still an issue with upgrading/replacing zenpacks? Coworker remembers there being an issue with overwriting zenpacks with newer versions of the same zenpack.
[24-Aug-2009 12:23:42] <ckrough> ie, I create a zenpack, deploy it, then add some more functionality and want to deploy the updated one on servers that have the original. That shouldn't cause issues, right?
[24-Aug-2009 12:24:34] <venturaville> rmatte: I've got a bunch of 9509s if you need some help.........
[24-Aug-2009 12:26:21] <mrayzenoss> ckrough: there was more of an issue with Zip zenpacks
[24-Aug-2009 12:26:45] <ckrough> mrayzenoss: that's confirmed as resolved with eggs?
[24-Aug-2009 12:27:21] <mrayzenoss> yeah, I mean assuming the issue is related to the fact that .zip zenpacks didn't install to a zenpacks directory, they dumped junk into the Products folder
[24-Aug-2009 12:27:31] <ckrough> gotcha
[24-Aug-2009 12:27:32] <mrayzenoss> which didn't play well with lots of things
[24-Aug-2009 12:27:45] <mrayzenoss> zenoss upgrades would blow up zenpacks
[24-Aug-2009 12:27:47] <ckrough> thanks
[24-Aug-2009 12:29:32] <dranch> Rmatte, you in?
[24-Aug-2009 12:30:39] <dranch> I need to figure out why specific Zenoss COMMAND tests aren't running and why zencommand doesn't log both the working tests and the non-working tests!
[24-Aug-2009 12:31:14] <mrayzenoss> nzle: Glenn posted a response to your thread
[24-Aug-2009 12:54:50] <rmatte> dranch: I'm in but I'm pretty busy so I'm watching the channel on and off
[24-Aug-2009 12:55:28] <rmatte> dranch: tests as in, fill in device name and click "Test"?
[24-Aug-2009 13:03:23] <markeriv> holy cow
[24-Aug-2009 13:03:26] <markeriv> i am still in here O.o
[24-Aug-2009 13:03:57] <dranch> hey Rmatte: the manual tests have always worked
[24-Aug-2009 13:04:27] <rmatte> but they aren't now?
[24-Aug-2009 13:05:08] <rmatte> "<dranch> I need to figure out why specific Zenoss COMMAND tests aren't running"
[24-Aug-2009 13:05:17] <dranch> they don't seem to run automatically and the logging says it's not running ANY commands for hosts when infact it IS running them for some templates
[24-Aug-2009 13:05:54] <rmatte> quite honestly, beyond the troubleshooting we did last time around I don't really know what to tell you, I've never had issues like that with command based datapoints
[24-Aug-2009 13:05:59] <rmatte> and I use them all the time
[24-Aug-2009 13:06:27] <rmatte> obviously something is messed up with your zencommand, it's either a problem with the daemon itself, with your Zenoss settings, or with the datapoints
[24-Aug-2009 13:06:43] <dranch> yeah.. what's troublesome is the logging. It says it has ZERO commands to run yet in some templates, it shows the results of the tests
[24-Aug-2009 13:07:25] <rmatte> is there any difference in the output or configuration between those datapoints and the ones that aren't running?
[24-Aug-2009 13:08:09] <rmatte> also, how long do the commands take to run through? This might be an issue where zencommand is running out of cycle time.
[24-Aug-2009 13:08:40] <dranch> I don't know if you saw my other comments on IRC on friday but it seems datapoints aren't required to get tests to execute
[24-Aug-2009 13:08:52] <rmatte> datasources
[24-Aug-2009 13:08:55] <rmatte> is what I meant
[24-Aug-2009 13:09:10] <dranch> The zencommand logs say it's idle for like 46 seconds here and there.
[24-Aug-2009 13:09:19] <rmatte> also, I don't see how a datasource can work with no datapoints
[24-Aug-2009 13:09:36] <rmatte> even snmp based datasources have datapoints
[24-Aug-2009 13:09:44] <rmatte> they just get created automatically now and you can't see them
[24-Aug-2009 13:10:57] <dranch> ok.. I guess that would work
[24-Aug-2009 13:12:00] <rmatte> Where did these templates come from?
[24-Aug-2009 13:12:03] <dranch> I also have a screwy maintenance window somewhere that's putting our primary database servers in maintenance mode! It's scary!
[24-Aug-2009 13:12:07] <rmatte> did you make them, or are they from a ZenPack?
[24-Aug-2009 13:12:19] <dranch> I have no idea. I inherited this setup
[24-Aug-2009 13:12:27] <rmatte> dranch: but you can't see the maintenance window in the list?
[24-Aug-2009 13:12:35] <dranch> nope
[24-Aug-2009 13:12:49] <rmatte> sounds like your Zenoss install is quite fubar
[24-Aug-2009 13:12:53] <rmatte> hmmm
[24-Aug-2009 13:13:42] <dranch> it does work but the harder I dig into it.. the more I learn it's not functioning. What I don't understand is that I've added COMMANDs before and they work.. it's just this new one doesn't work on an automatic fashion
[24-Aug-2009 13:13:59] <rmatte> is the maintenance window specifically affecting a group of devices with a common organizer?
[24-Aug-2009 13:14:11] <dranch> yes
[24-Aug-2009 13:14:19] <rmatte> go to that organizer
[24-Aug-2009 13:14:45] <dranch> There are 8 machines in the group. 3 are production and 5 are maintenance mode. I'll come back some time later and all 8 machines are in maintenance mode
[24-Aug-2009 13:15:49] <jb> it would be nice if you could generate a quick PDF based report of perf graphs
[24-Aug-2009 13:15:57] <jb> i get tired of copying and pasting them into emails
[24-Aug-2009 13:17:03] <rmatte> dranch: go to the organizer that contains the services, then click More -> Administration
[24-Aug-2009 13:17:11] <rmatte> check for Maintenance Windows
[24-Aug-2009 13:17:13] <dranch> Organizer --> More --> Administration pull down --> No maintenance windows
[24-Aug-2009 13:17:24] <rmatte> ok, now check each device individually
[24-Aug-2009 13:17:30] <rmatte> check the admin section for each
[24-Aug-2009 13:18:10] <rmatte> someone might have added local maintenance windows to the devices themselves and forgot about them
[24-Aug-2009 13:18:15] <dranch> there IS one for the higher level organizer though the stop state is "original"
[24-Aug-2009 13:18:52] <mrayzenoss> "original" can be troublesome if that's not "Production"
[24-Aug-2009 13:18:59] <dranch> I'll set it to change it back to PRODUCTIOn
[24-Aug-2009 13:19:04] <rmatte> yeh
[24-Aug-2009 13:19:24] <dranch> ok.. back to the COMMAND stuff.
[24-Aug-2009 13:19:51] <dranch> If I run a manual test, shouldn't it be logged in zencommand.log when running with -v 10?
[24-Aug-2009 13:20:13] <rmatte> no, -v10 just makes the output of the command you're running more verbose
[24-Aug-2009 13:20:17] <rmatte> it doesn't guarantee logging
[24-Aug-2009 13:20:51] <dranch> Ok.. so is there a more verbose setting to see what's going on here?
[24-Aug-2009 13:21:01] <rmatte> 10 is the max to my knowledge
[24-Aug-2009 13:21:21] <rmatte> zencommand -v10 --cycle is the best you can do
[24-Aug-2009 13:21:56] <dranch> the manual run works fine but greping comes up empty: tail --lines=10000 zencommand.log | grep -i vip
[24-Aug-2009 13:23:00] <rmatte> zencommand run -v10 --cycle will run the daemon in the foreground
[24-Aug-2009 13:23:04] <rmatte> it should show everything
[24-Aug-2009 13:23:20] <rmatte> I'm running it right now on a box and it's providing plenty of info
[24-Aug-2009 13:24:10] <dranch> Can you see any results from manual runs?
[24-Aug-2009 13:25:03] <rmatte> I see the output of each command run, and how many events are queued as it goes
[24-Aug-2009 13:25:39] <rmatte> DEBUG:zen.zencommand:Total of 2 queued events
[24-Aug-2009 13:25:39] <rmatte> DEBUG:zen.zencommand:Next command in 281 seconds
[24-Aug-2009 13:26:14] <dranch> so if you manually run a given COMMAND and grep for the script name in the zencommand.log file, can you find it?
[24-Aug-2009 13:26:50] <rmatte> when you manually run it I don't think it logs at all
[24-Aug-2009 13:26:56] <rmatte> the logging is for the automatic stuff
[24-Aug-2009 13:27:26] <dranch> Ok.. that's unfortunate. So, is there a way to trace or log why zenoss doesn't think it needs to run this command?
[24-Aug-2009 13:27:34] <rmatte> and it only logs errors
[24-Aug-2009 13:27:44] <rmatte> like, my last log entry is from 4 days ago
[24-Aug-2009 13:27:50] <rmatte> 2009-08-20 17:12:04 WARNING zen.zencommand: Command timed out on device SPS: '$ZENHOME/Products/SNMPWindowsPerfMonitor/libexec/winmem.sh "192.168.10.6"
[24-Aug-2009 13:28:25] <rmatte> but your isn't apparently erroring out, it's just not running at all
[24-Aug-2009 13:28:34] <dranch> exactly
[24-Aug-2009 13:29:35] <rmatte> why don't you try installing Zenoss on another box and trying the exact same commands to see if it's still messed up
[24-Aug-2009 13:29:44] <rmatte> that way you'll know whether it's the commands or your install
[24-Aug-2009 13:30:26] <dranch> that's a pretty time intensive test... It might come down to that to get up to a new release but it's going to take some time to get things all running, etc.
[24-Aug-2009 13:30:55] <dranch> I would have hoped that the logging would have been better than this. To me, logging is #1 and everything else is second
[24-Aug-2009 13:31:04] <rmatte> well, I don't know what else to suggest, it's not automatically running the commands, I believe we've gone over almost every relevant setting, so I'm out of ideas
[24-Aug-2009 13:31:26] <rmatte> when you run with a -v10 you get even more output than the log file
[24-Aug-2009 13:31:34] <rmatte> so run wth -v10, cap it to a file, and see what you get
[24-Aug-2009 13:31:56] <rmatte> and I'd say run the whole thing, not just on one specific device
[24-Aug-2009 13:32:06] <rmatte> zencommand run -v10 --cycle
[24-Aug-2009 13:32:13] <dranch> that's what I'm doing. The script that the COMMAND runs is called check-vip.sh and I can't find the string "vip" in the logs
[24-Aug-2009 13:32:34] <rmatte> yes, but are you also searching the output of the actual command for it?
[24-Aug-2009 13:32:40] <rmatte> forget about the log files and focus on the output
[24-Aug-2009 13:32:45] <dranch> no.. just the script it should be running
[24-Aug-2009 13:32:51] <rmatte> the log files appear to only capture very specific info
[24-Aug-2009 13:33:13] <rmatte> do a zencommand run -v10 --cycle and look for anything odd in the output
[24-Aug-2009 13:33:21] <rmatte> it might show you where it's tripping up
[24-Aug-2009 13:33:37] <rmatte> capture it to a file and comb through it
[24-Aug-2009 13:33:45] <rmatte> the alternative is to do a fresh Zenoss install
[24-Aug-2009 13:33:56] <rmatte> (which you might end up doing anyways)
[24-Aug-2009 13:34:07] <rmatte> did you setup this install to begin with, or did you inherit it from someone else?
[24-Aug-2009 13:34:20] <dranch> inherit..
[24-Aug-2009 13:34:26] <rmatte> yeh, those are always the worst
[24-Aug-2009 13:34:35] <rmatte> since you have no idea what the last person might have messed around with
[24-Aug-2009 13:34:36] <dranch> I would hope that I could fix this..
[24-Aug-2009 13:34:58] <rmatte> they might have gone in and messed with code or something, not noticing that the changes caused other issues
[24-Aug-2009 13:35:16] <rmatte> it could be like finding a needle in a haystack
[24-Aug-2009 13:35:39] <rmatte> just out of curiosity, why are you not able to do an upgrade, to see if that resolves the issues?
[24-Aug-2009 13:36:00] <dranch> yeah.. agreed. With your help on friday, seeing that commands were being issued to other machines but zencommand logging that no commands need to be run was very, um, enlightening
[24-Aug-2009 13:36:18] <dranch> manpower
[24-Aug-2009 13:36:39] <rmatte> is this a single Zenoss instance or multiple?
[24-Aug-2009 13:36:45] <dranch> single
[24-Aug-2009 13:36:51] <rmatte> then why is manpower an issue?
[24-Aug-2009 13:37:04] <rmatte> I'm upgrading 5 Zenoss servers on Tuesday night by myself
[24-Aug-2009 13:37:20] <rmatte> (tomorrow night)
[24-Aug-2009 13:37:52] <rmatte> If an upgrade is done properly, none of your ZenPacks or settings will be affected
[24-Aug-2009 13:37:59] <rmatte> since it's using the same Zope database afterwards
[24-Aug-2009 13:38:15] <dranch> It's somewhat customized with nagios plugins, etc. Have to get the new zenoss, get the dependencies fixed (the RPMs never seem to have worked for the previous admin), identify and install the required nagios plugins, hopefully just restore the DB (I don't have high hopes), etc.
[24-Aug-2009 13:38:33] <rmatte> I see
[24-Aug-2009 13:38:42] <rmatte> yeh, I don't like the RPM installs
[24-Aug-2009 13:38:43] <dranch> this would also be on a different machine and NOT the same machine
[24-Aug-2009 13:38:48] <rmatte> I use stack installer, much nicer
[24-Aug-2009 13:39:01] <dranch> I personally like RPMs but it seems the RPMs for Zenoss aren't built properly
[24-Aug-2009 13:39:34] <dranch> brb
[24-Aug-2009 13:39:51] <rmatte> dranch: yeh, different machine is fairly simple, just do a fresh install of the same version on the new machine, tar up $ZENHOME on the old machine, move it to the new one, start it up and make sure it has all the previous settings and everything
[24-Aug-2009 13:39:53] <rmatte> then from there, just do the upgrade
[24-Aug-2009 13:40:06] <rmatte> (you'd also have to copy over any other files that it depends on which exist outside of $ZENHOME
[24-Aug-2009 13:41:14] <rmatte> like if you have any customs scripts stored in /opt or whatever that Zenoss uses)
[24-Aug-2009 13:41:21] <jb> i never have an issue with the RPM's..
[24-Aug-2009 13:41:29] <jb> not in 2.3+ at least
[24-Aug-2009 13:41:51] <rmatte> RPMs get annoying, whining about dependencies yet not being able to find them to download them half the time
[24-Aug-2009 13:41:58] <rmatte> so then you have to go googling for RPM packages
[24-Aug-2009 13:42:01] <jb> so frontend it with yum
[24-Aug-2009 13:42:12] <jb> or your package management solution of choice..
[24-Aug-2009 13:42:27] <dranch> rmatte: so you recommend to install the OLD version of zenoss on the new machine, move the DB over, and then upgrade it?!
[24-Aug-2009 13:42:45] <rmatte> dranch: that would be the best solution
[24-Aug-2009 13:43:01] <dranch> interesting. assuming upgraing from 2.2.4 to current is supported
[24-Aug-2009 13:43:06] <rmatte> dranch: unless you want to upgrade it on the old machine and then move it over, either way works, I just figure this way would end up being less hassle
[24-Aug-2009 13:43:19] <rmatte> dranch: it may not be, you'll have to follow the upgrade path
[24-Aug-2009 13:43:43] <dranch> Ah... is the path clearly documented somewhere?
[24-Aug-2009 13:43:47] <rmatte> dranch: plus, if the upgrade fails on the new machine at least the old one is still running that way
[24-Aug-2009 13:43:56] <rmatte> dranch: yeh it is, hold on...
[24-Aug-2009 13:44:09] <dranch> exactly.. I can't mess with the current one (even if it's not running everything it should be)
[24-Aug-2009 13:44:11] <mrayzenoss> dranch: http://www.zenoss.com/community/docs/installation-guide/2.4.2/ch08s01.html#d4e866
[24-Aug-2009 13:44:39] <rmatte> yeh, so you need to go to 2.3.3 and then go to 2.4
[24-Aug-2009 13:44:40] <dranch> thanks mray
[24-Aug-2009 13:45:14] <dranch> ok. Mray.. have you seen this thread? anything to add on why COMMANDS don't run automatically?
[24-Aug-2009 13:45:41] <magnachef> so a sad (yet good) day...looks like I'm going to be taking a programming job within my company and won't be doing the NMS stuff anymore
[24-Aug-2009 13:46:12] <mrayzenoss> dranch: I've been passively listening, I don't have much to add
[24-Aug-2009 13:46:12] <dranch> Ha.. you'll probably be doing both jobs!
[24-Aug-2009 13:46:15] <rmatte> magnachef: ah, congrats on the new job
[24-Aug-2009 13:46:23] <mrayzenoss> magnachef: congrats, you'll be missed
[24-Aug-2009 13:46:25] <dranch> mray: bummer.. ok.
[24-Aug-2009 13:46:36] <magnachef> haha...we'll see how well I can document!
[24-Aug-2009 13:46:38] <magnachef> thanks
[24-Aug-2009 13:46:58] <mrayzenoss> dranch: my suggestion would be to create a new device class and add a simple version of the command you're running
[24-Aug-2009 13:46:59] <magnachef> I have a feeling that I'll still be a lurker
[24-Aug-2009 13:47:04] <dranch> It's just bizzare to me that some tests run.. others don't
[24-Aug-2009 13:47:19] <mrayzenoss> dranch: then add a device to the device class and see what works
[24-Aug-2009 13:47:20] <dranch> I can try that
[24-Aug-2009 13:47:50] <mrayzenoss> of course, if you're running 2.2.4 you're getting long in the tooth and 2.4.x should be better
[24-Aug-2009 13:47:59] <dranch> does it need to be new from the very top level?
[24-Aug-2009 13:48:21] <dranch> yeah.. 2.4.x is on the map.. all the windows boxes have been throwing fits with zenoss
[24-Aug-2009 13:48:34] <mrayzenoss> dranch: no, but it will inherit the templates of devices higher in the hierarchy (but not bind them)
[24-Aug-2009 13:49:03] <mrayzenoss> dranch: wait, maybe it does bind the templates if they exist at creation time
[24-Aug-2009 13:49:19] <dranch> mray: ok, I didn't know about the inherited but non-binded "feature"
[24-Aug-2009 13:49:30] <dranch> I'll just create a new toplevel
[24-Aug-2009 13:49:32] <mrayzenoss> dranch: basically, if you're having trouble with /Devices/Server/Blah, don't test a 'clean' version with /Devices/Server/Blah/New
[24-Aug-2009 13:50:28] <rmatte> mrayzenoss: in reports, you can select a period of time (start/end) to collect data for, how would I manually specify that in a d.getRRDValue command?
[24-Aug-2009 13:52:56] <dranch> So I added database5 to the new /Devices/Test class yet it put it back in /Devices /Server /Linux /database /replication
[24-Aug-2009 13:53:06] <dranch> wtf
[24-Aug-2009 13:53:17] <rmatte> "put it back in"?
[24-Aug-2009 13:53:27] <straterra> That's what she said?
[24-Aug-2009 13:53:30] <straterra> Sorry..
[24-Aug-2009 13:53:36] <dranch> it didn't show up it /Devices/Test
[24-Aug-2009 13:53:37] <rmatte>
[24-Aug-2009 13:53:39] <dranch> nice..
[24-Aug-2009 13:53:53] <rmatte> dranch: sounds screwy
[24-Aug-2009 13:53:58] <rmatte> did you copy it over?
[24-Aug-2009 13:54:29] <dranch> adding it as a new device
[24-Aug-2009 13:54:40] <mrayzenoss> dranch: doesn't it already exist?
[24-Aug-2009 13:54:50] <dranch> yup
[24-Aug-2009 13:54:57] <rmatte> then why add it as a new device?
[24-Aug-2009 13:55:10] <rmatte> you just need to move it from it's old organizer to the new one
[24-Aug-2009 13:55:15] <dranch> to see if I can get these COMMANDs to run under a clean template
[24-Aug-2009 13:55:44] <rmatte> dranch: just moving it to the new organizer will give it a clean template
[24-Aug-2009 13:56:07] <rmatte> the template is bound to the organizer, not the device (unless you specifically created a local copy of the template)
[24-Aug-2009 13:56:08] <dranch> Hmmm.. let me add a NEW device and run this COMMAND on that one
[24-Aug-2009 13:56:19] <dranch> ok.. that's good to know
[24-Aug-2009 13:56:54] <rmatte> You should read the Zenoss Admin guide top to bottom if you haven't already
[24-Aug-2009 13:57:01] <rmatte> you'd get a better feel for the system
[24-Aug-2009 13:57:44] <dranch> I've read parts of it but I agree... though the docs are a little fragmented if you ask me
[24-Aug-2009 13:58:07] <rmatte> I didn't find them to be very fragmented
[24-Aug-2009 13:58:43] <dranch> ok.. new machine added. let me now try to run the COMMAND script there
[24-Aug-2009 13:59:56] <mrayzenoss> dranch: I believe 2.3 was when the Admin Guide got it's first big revamp
[24-Aug-2009 14:00:12] <mrayzenoss> and each release has had more clean ups and updats
[24-Aug-2009 14:00:14] <dranch> oh.. ok. That might be a reason
[24-Aug-2009 14:00:41] <mrayzenoss> (had to go look at the 2.2 docs to verify that)
[24-Aug-2009 14:00:55] <mrayzenoss> and King Crab will continue to have more upgrades
[24-Aug-2009 14:01:45] <dranch> Ok.. the new template runs a manual test via the webui fine
[24-Aug-2009 14:02:11] <dranch> the result is OK|status=1 and it's set to /cmd/fail
[24-Aug-2009 14:02:51] <rmatte> ah, return dict(start=startDate, end=endDate, function=how), so it'd be like d.getRRDValue('08/17/2009', '08/24/2009', 'MAX')
[24-Aug-2009 14:03:01] * rmatte test
[24-Aug-2009 14:03:03] <rmatte> s
[24-Aug-2009 14:04:41] <dranch> mray: Do you agree with rmatte that commands shouldn't execute unless a datapoint is defined?
[24-Aug-2009 14:05:00] <dranch> or a threshhold?
[24-Aug-2009 14:07:17] <mrayzenoss> you don't have to have a threshold
[24-Aug-2009 14:07:42] <dranch> but you do need to set a datapoint?
[24-Aug-2009 14:12:23] <dranch> Tests aren't running on the new machine.. I just added a GAUGE datapoint to see if that changes anything
[24-Aug-2009 14:13:53] <jb> venturaville: do you guys have a 3584?
[24-Aug-2009 14:13:56] <jb> or rocket?
[24-Aug-2009 14:26:46] <rmatte> eugh, I wish rocket or venturaville were around, I'm trying to figure out how to convert a time such as 07/17/2009 to something that getRRDValue will understand as a start or end time
[24-Aug-2009 14:26:51] <rmatte> so far the best example I've found is:
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> import time
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> mystart = time.time() - (3600 * 8)
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> args = dict(start=mystart)
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> f.getRRDValue('usedBlocks',**args)
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> 10021455.52
[24-Aug-2009 14:27:16] <rmatte> but, that doesn't explain how to convert a date such as that, it just kind of gives a general idea
[24-Aug-2009 14:29:25] <rmatte> hmmm, I might have figured out how, just need to code it up
[24-Aug-2009 14:36:57] <jb> hmm
[24-Aug-2009 14:37:07] <jb> i think my first official zenpack is finished
[24-Aug-2009 14:37:44] <mrayzenoss> ummm... yay? (My backlog is 16 ZenPacks deep)
[24-Aug-2009 14:37:54] <kgoedtel> pile it on!
[24-Aug-2009 14:38:28] <jb> hah matt
[24-Aug-2009 14:38:52] <mrayzenoss> it's OK, I did 10 over the weekend
[24-Aug-2009 14:39:06] <mrayzenoss> and should have them all published today
[24-Aug-2009 14:39:17] <jb> i just need to write the docs
[24-Aug-2009 14:39:39] <rmatte> hmmm, I can use the find() function in zendmd but I can't use it in a standalone script with the standard zendmd imports
[24-Aug-2009 14:39:48] <rmatte> lovely, now I need to figure out what I need to import to be able to use it
[24-Aug-2009 14:40:46] <mrayzenoss> #!/usr/bin/env python
[24-Aug-2009 14:40:46] <mrayzenoss> import Globals
[24-Aug-2009 14:40:46] <mrayzenoss> from Products.ZenUtils import Time
[24-Aug-2009 14:40:46] <mrayzenoss> from Products.ZenUtils.ZenScriptBase import ZenScriptBase
[24-Aug-2009 14:40:50] <mrayzenoss> like that?
[24-Aug-2009 14:41:03] <rmatte> #!/usr/bin/env python
[24-Aug-2009 14:41:03] <rmatte> import Globals
[24-Aug-2009 14:41:03] <rmatte> from Products.ZenUtils.ZenScriptBase import ZenScriptBase
[24-Aug-2009 14:41:03] <rmatte> from transaction import commit
[24-Aug-2009 14:41:03] <rmatte> dmd = ZenScriptBase(connect=True).dmd
[24-Aug-2009 14:41:04] <rmatte> from Products.ZenUtils import Time
[24-Aug-2009 14:41:06] <rmatte> is what I have
[24-Aug-2009 14:41:22] <jb> from string import strip, find, split, lower, atoi
[24-Aug-2009 14:41:43] <rmatte> thanks jb
[24-Aug-2009 14:42:04] <jb> np
[24-Aug-2009 14:42:18] <rmatte> weird
[24-Aug-2009 14:42:22] <rmatte> return s.find(*args)
[24-Aug-2009 14:42:23] <rmatte> TypeError: find/rfind/index/rindex() takes at least 1 argument (0 given)
[24-Aug-2009 14:42:32] <rmatte> d = find('XP03LYNv02')
[24-Aug-2009 14:42:35] <rmatte> that's the line
[24-Aug-2009 14:43:45] <rmatte> http://pastebin.com/m63e343c3
[24-Aug-2009 14:43:50] <rmatte> there's the whole script I'm testing
[24-Aug-2009 14:44:25] <jb> hmm wonder why my template bindings aren't being exported to a zenpack
[24-Aug-2009 14:44:26] <rmatte> I'm basically just trying to make a script where I can specify start, end, and method (average, maximum, etc...) and have getRRDValue retrieve info based on it
[24-Aug-2009 14:44:34] <rmatte> I'm basically trying to design scripts to dump reports
[24-Aug-2009 14:45:20] <rmatte> I want to at least get it working with 1 value for starters, and then replace the values with variables
[24-Aug-2009 14:46:29] <rmatte> ultimately it's going to be something like: script.py --start 08/17/2009 --end 08/24/2009 --function MAXIMUM
[24-Aug-2009 14:47:10] <rmatte> I'll probably build different functions in to it for different reports too, so that I can have something like --report cpu or --report memory
[24-Aug-2009 14:47:16] <rmatte> and have it dump the report as a csv
[24-Aug-2009 14:47:24] <rmatte> just doing this as a workaround until the reporting is fixed
[24-Aug-2009 14:48:12] <rmatte> it'll allow more flexibility in how it's output as well
[24-Aug-2009 14:50:27] <jb> cant help ya :/
[24-Aug-2009 14:50:47] <rmatte> weird how it's complaining that I'm not passing a value to find when I am
[24-Aug-2009 14:51:18] daMaestro|isBack is now known as daMaestro
[24-Aug-2009 15:24:58] <rmatte> It's not not, or isnot
[24-Aug-2009 15:25:05] <rmatte> woops
[24-Aug-2009 15:25:14] <dranch_> is snot
[24-Aug-2009 15:25:19] <dranch_> ;-)
[24-Aug-2009 18:17:18] daMaestro|isBack is now known as daMaestro
[25-Aug-2009 00:00:46] [disconnected at Tue Aug 25 00:00:46 2009]
[25-Aug-2009 00:00:46] [connected at Tue Aug 25 00:00:46 2009]
[25-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[25-Aug-2009 07:50:24] <mrayzenoss> Morning all
[25-Aug-2009 08:12:01] <ckrough> morning
[25-Aug-2009 08:36:26] <ckrough> Where does zenoss determine which processes it's going to monitor? I was under the impression it would only pick up on stuff defined in /Classes/Processes
[25-Aug-2009 08:36:38] <ckrough> But its picking up on some random stuff now
[25-Aug-2009 08:52:57] <cgibbons> bad regex?
[25-Aug-2009 09:13:33] <rmatte> good morning
[25-Aug-2009 10:03:34] <ckrough> cgibbons: sry, was at a meeting. No, I don't think it's bad regex since this is new behaviour (seemingly) but we havent modified Processes
[25-Aug-2009 10:04:04] <ckrough> cgibbons: Is my understanding correct in that it should only attempt to monitor processes that are defined under /Processes?
[25-Aug-2009 10:05:54] <cgibbons> as far as I understand (don't weigh that too heavily). I'd run zenmodeler run -v10 -d yourdevice real quick just to see what it shows when it discovers processes
[25-Aug-2009 10:06:09] <ckrough> Ill try that
[25-Aug-2009 11:36:38] <jb> mrayzenoss: just sent you a zenpack..
[25-Aug-2009 11:36:53] <mrayzenoss> just got it, thanks!
[25-Aug-2009 11:37:15] <jb>
[25-Aug-2009 11:37:16] <mrayzenoss> nice documentation
[25-Aug-2009 11:37:29] <jb> thanks
[25-Aug-2009 11:38:11] <rmatte> jb++ for docs
[25-Aug-2009 11:38:14] <rmatte>
[25-Aug-2009 11:38:20] <jb>)
[25-Aug-2009 11:38:21] <jb> http://thunder.jbdesign.net/docs/zenoss/zenoss-ibm3584mon.html
[25-Aug-2009 11:38:45] <rmatte> nice
[25-Aug-2009 12:07:38] <rmatte> eugh, our internet connection is being flaky right now
[25-Aug-2009 12:24:42] <dranch> Hey Rmatte?
[25-Aug-2009 12:24:48] <rmatte> yup?
[25-Aug-2009 12:24:51] <dranch> hey..
[25-Aug-2009 12:25:16] <dranch> so interetingly enough, that COMMAND stuff eventually started running under a new template. but it took HOURS
[25-Aug-2009 12:25:28] <rmatte> strange
[25-Aug-2009 12:25:42] <dranch> So obviously, something is hosed in the other template. Anyway, I question to you about binding
[25-Aug-2009 12:25:48] <dranch> A question that is
[25-Aug-2009 12:26:17] <dranch> I have a /devices /server /windows template. In that template is a MySQL test that I don't want run on one specific machine.
[25-Aug-2009 12:27:27] <dranch> What's strange is if I click on bind templates, only the top "devices" template is bound. the MySQL template is not highlighted BUT it is being run
[25-Aug-2009 12:27:52] <rmatte> yeh, so just go to that 1 specific device, go to More -> Templates, then go to Bind Templates
[25-Aug-2009 12:27:58] <rmatte> ctrl-click to unselect that template
[25-Aug-2009 12:28:00] <rmatte> then click ok
[25-Aug-2009 12:28:43] <dranch> That's what I'm saying, it's not highlight.. it's not bound
[25-Aug-2009 12:28:49] <rmatte> dranch: does the mysql template even show up in the list of templates?
[25-Aug-2009 12:28:54] <dranch> yes
[25-Aug-2009 12:29:09] <dranch> It's not highlighted yet it's still run
[25-Aug-2009 12:29:18] <rmatte> a quick and dirty way would be to make a local copy of the template, go in, and delete all of the datapoints, thresholds, and graphs
[25-Aug-2009 12:29:27] <rmatte> problem solved
[25-Aug-2009 12:29:32] <dranch> man.. I can't form sentences this morning
[25-Aug-2009 12:29:56] <rmatte> I'm like that some mornings
[25-Aug-2009 12:31:43] <dranch> Ok, with a local copy, the MySQL is still not bound. Let me try "push changes" to see if the behavior is any different
[25-Aug-2009 12:33:20] <rmatte> yeh, but if you create the local copy and remove everything from the template itself, then it won't be monitoring anything
[25-Aug-2009 12:33:29] <rmatte> you're effectively creating a blank template as a placeholder
[25-Aug-2009 12:33:48] <dranch> well, I need to monitor the server.. it's just the issue that HTTP isn't running on this specific one
[25-Aug-2009 12:34:24] <rmatte> eugh
[25-Aug-2009 12:34:37] <rmatte> you're trying to unbind the template, meaning you don't want to monitor anything in that template on that device
[25-Aug-2009 12:34:38] <rmatte> correct?
[25-Aug-2009 12:34:59] <rmatte> so I'm saying, create a local copy of the template, then edit the local template and remove everything in it
[25-Aug-2009 12:35:08] <dranch> but it's not bound (according to the GUI)
[25-Aug-2009 12:35:28] <dranch> Then re-create all the graphs, etc?
[25-Aug-2009 12:35:31] <rmatte> if you're seeing it in the perf tab, then it's somehow bound
[25-Aug-2009 12:35:40] <rmatte> you're using an old version, maybe there's a bug in the binding
[25-Aug-2009 12:35:53] <rmatte> why re-create?
[25-Aug-2009 12:36:14] <dranch> Yeah.. I found another but last night. For maintenance windows, the "restore former state" is busted
[25-Aug-2009 12:36:24] <rmatte> yeh
[25-Aug-2009 12:56:50] * pgega is away: I'm busy
[25-Aug-2009 13:00:03] * pgega is back (gone 00:03:15)
[25-Aug-2009 13:08:33] pgega is now known as pgega|away
[25-Aug-2009 13:08:33] <rmatte> hmmm, that Ex-Chat project (what npmccallum is using) seems pretty dead.
[25-Aug-2009 13:08:51] <mrayzenoss1> I think that's XChat2
[25-Aug-2009 13:09:06] <dranch> What does it do?
[25-Aug-2009 13:09:14] <mrayzenoss1> http://xchat.org/
[25-Aug-2009 13:09:18] <mrayzenoss1> just another IRC client
[25-Aug-2009 13:09:24] <rmatte> http://sourceforge.net/projects/ex-chat/
[25-Aug-2009 13:09:25] <mrayzenoss1> I used it on Linux and OSX
[25-Aug-2009 13:09:28] <rmatte> I thought it was that
[25-Aug-2009 13:09:32] <rmatte> it's an X-Chat clone
[25-Aug-2009 13:09:41] <dranch> wow.. how many does the world need?!
[25-Aug-2009 13:09:51] <mrayzenoss1> dunno, cgibbons wrote one
[25-Aug-2009 13:09:52] <rmatte> just two, iirc and bitchx
[25-Aug-2009 13:09:53] <rmatte> >:)
[25-Aug-2009 13:10:28] <mrayzenoss1> I'm using Adium for IRC now. 1 IM client for everything
[25-Aug-2009 13:10:44] <rmatte> yeh, I hear Adium is really good
[25-Aug-2009 13:10:59] <rmatte> I use bitchx for irc, and gaim (pidgin) for everything else
[25-Aug-2009 13:11:07] <mrayzenoss1> IRC, AIM, Yahoo, Gtalk, Facebook, Twitter
[25-Aug-2009 13:11:11] <mrayzenoss1> all in 1
[25-Aug-2009 13:12:09] <dranch> I'm using Adium now
[25-Aug-2009 13:12:26] <dranch> pidgin has an IRC client too
[25-Aug-2009 13:12:44] <mrayzenoss1> yeah, Pidgin and Adium use the same libraries (libpurple)
[25-Aug-2009 13:12:54] <rmatte> http://en.wikipedia.org/wiki/IrcII <--- Now there's an IRC client
[25-Aug-2009 13:12:55] <rmatte>
[25-Aug-2009 13:13:04] <dranch> yup.. and share the same vulerabilities too
[25-Aug-2009 13:14:26] <dranch> isn't bitchx based on ircII?
[25-Aug-2009 13:15:22] <rmatte> yeh, it was originally an ircII script
[25-Aug-2009 13:15:32] <rmatte> then it got converted to a standalone app by panasync
[25-Aug-2009 13:15:39] <rmatte> I'm using BX with Osiris script
[25-Aug-2009 13:15:44] <rmatte> osiri s-1c/ bitch x-1.1-final
[25-Aug-2009 13:16:38] <dranch> reading the wikipedia stuff on it.. didn't know it was merged into epic
[25-Aug-2009 13:17:20] <rmatte> yeh
[25-Aug-2009 13:17:39] mrayzenoss1 is now known as mrayzenoss
[25-Aug-2009 13:17:53] <jb> so
[25-Aug-2009 13:17:56] <jb> when i export to CSV
[25-Aug-2009 13:18:05] <jb> the times are always wrong
[25-Aug-2009 13:18:11] <jb> oh
[25-Aug-2009 13:18:14] <jb> its just excel.
[25-Aug-2009 13:18:22] <rmatte> jb
[25-Aug-2009 13:18:23] <rmatte> http://dev.zenoss.org/trac/ticket/5438
[25-Aug-2009 13:18:27] <rmatte> http://dev.zenoss.org/trac/ticket/5439
[25-Aug-2009 13:19:15] <rmatte> it's not just excel
[25-Aug-2009 13:19:19] <rmatte> the export function is fubar
[25-Aug-2009 13:20:40] <rmatte> I've just been copying and pasting right from the web page to get around it
[25-Aug-2009 13:20:56] <rmatte> but I'm working on a python script to temporarily replace some of the more important reports
[25-Aug-2009 13:20:59] <rmatte> until they can fix that
[25-Aug-2009 13:22:05] <rmatte> I probably won't have the script done until some time next week
[25-Aug-2009 13:22:08] <rmatte> too much on my plate right now
[25-Aug-2009 13:23:38] <rmatte> plus I'm developing my python skills as I go
[25-Aug-2009 13:23:42] <rmatte> so lots of googling/reading
[25-Aug-2009 13:23:59] <rmatte> once the script is done I'll release it
[25-Aug-2009 13:24:34] <jb> hmm
[25-Aug-2009 13:24:36] <jb> yeah the times are wrong
[25-Aug-2009 13:24:51] <rmatte> yeh, when you export it always exports the average over the last 7 days
[25-Aug-2009 13:24:55] <rmatte> it ignores what you selected
[25-Aug-2009 13:24:58] <jb> lame
[25-Aug-2009 13:24:59] <jb> :/
[25-Aug-2009 13:25:08] <jb> thats a big problem
[25-Aug-2009 13:25:23] <rmatte> these are pretty obvious bugs, which tells me that not many people bother to use the reports system apparently
[25-Aug-2009 13:25:36] <rmatte> though it's a very important component
[25-Aug-2009 13:25:37] <jb> i am just exporting events
[25-Aug-2009 13:25:40] <jb> not reports
[25-Aug-2009 13:25:46] <rmatte> oh
[25-Aug-2009 13:25:51] <rmatte> is it wrong for that as well?
[25-Aug-2009 13:26:05] * rmatte tests
[25-Aug-2009 13:26:08] <jb> oh
[25-Aug-2009 13:26:09] <jb> actually
[25-Aug-2009 13:26:13] <rmatte> I thought you were exporting reports
[25-Aug-2009 13:26:21] <jb> no
[25-Aug-2009 13:26:22] <jb> its correct
[25-Aug-2009 13:26:23] <jb> im stupid
[25-Aug-2009 13:26:37] <rmatte> events export fine
[25-Aug-2009 13:26:39] <rmatte>
[25-Aug-2009 13:32:06] <rmatte> well, I'm out, talk to you guys later
[25-Aug-2009 13:32:14] <straterra> later
[25-Aug-2009 14:35:27] <cgibbons> I wrote one of what, mrayzenoss?
[25-Aug-2009 14:39:38] <venturaville> hi
[25-Aug-2009 14:39:53] <venturaville> anyone interested in LSI storage arrays (e.g. IBM rebranded DS* models and such?)
[25-Aug-2009 14:49:38] <jb> yes
[25-Aug-2009 14:49:40] <jb> of course
[25-Aug-2009 14:49:58] <jb> we use TPC for the most part
[25-Aug-2009 14:53:30] <venturaville> jb: I mimicked what someone else did and I am using pywbem along with engenioprovider to be the CIMOM -> array link
[25-Aug-2009 14:53:40] <venturaville> seems to work
[25-Aug-2009 14:55:05] <venturaville> I was hoping someone else could get the same setup going so they could help me test
[25-Aug-2009 14:55:19] <jb> sure
[25-Aug-2009 14:55:30] <venturaville> media access statistics: (uRecoveredSeekOperations, None) (uTotalReadBlocksRequested, 2568448853L) (uTotalLargeWrites, 0L) (uTotalWriteBlocksRequested, 1097960875L) (uTotalBlocksRequested, 3666409728L) (uDescription, uMedia Access Statistics for Volume vol2) (uTotalLargeWriteBlocksRequested, 0L) (uUnrecoverableReadOperations, None) (uWriteOperations, 233603915L) (uTotalCacheReadChecks, 969917377L) (uRecoveredReadOperations, None) (uUnrecoverableSee
[25-Aug-2009 14:55:32] <jb> venturaville: btw do you guys use IBM 3584 tape libs?
[25-Aug-2009 14:56:01] <venturaville> jb: we avoid using tape libraries at all costs ... most of our backups and video tape libraries are handled in another group
[25-Aug-2009 14:56:09] <jb> oh ok
[25-Aug-2009 14:56:14] <jb> just finished a pack for 3584's
[25-Aug-2009 14:56:21] <venturaville> cool.
[25-Aug-2009 14:56:31] <jb> but yeah, i can help test
[25-Aug-2009 14:56:34] <venturaville> we looked at them a while back when doing video archival... ended up using storagetek instead
[25-Aug-2009 14:56:38] <jb> i have a lab DS4500
[25-Aug-2009 14:56:46] <jb> and several production DS4700's
[25-Aug-2009 14:56:52] <venturaville> I'm testing against a 4700
[25-Aug-2009 14:56:53] <jb> and an 8000 soon
[25-Aug-2009 14:56:58] <venturaville> shark?
[25-Aug-2009 14:57:03] <jb> yeah
[25-Aug-2009 14:57:10] <jb> unfortuantly
[25-Aug-2009 14:57:15] <jb> i hate these damn things
[25-Aug-2009 14:58:28] <venturaville> see if you can grab a copy of pywbem, and then try this: http://pastebin.com/d5a320607
[25-Aug-2009 14:58:48] <venturaville> insert your own username/password pair for the CIMOM server
[25-Aug-2009 14:58:52] <venturaville> as well as the cimom server
[25-Aug-2009 14:59:11] <venturaville> I'm probably going to have to do those as zproperties since there doesn't seem to be a way to do auto discovery
[25-Aug-2009 15:00:04] <jb> hmm
[25-Aug-2009 15:00:23] <jb> not even sure where the cimom server is
[25-Aug-2009 15:00:45] <jb> im not a storage guy.. i may need to get our storage dude to help
[25-Aug-2009 15:00:47] <jb> he'll be back tommorow
[25-Aug-2009 15:01:02] <venturaville> k... I'll check back tomorrow then
[25-Aug-2009 15:17:05] <mrayzenoss> Zenoss 2.4.3 is up on SourceForge
[25-Aug-2009 15:17:14] <chudler> niiiiice
[25-Aug-2009 15:17:26] <mrayzenoss> I'm about to get on a flight to Zenoss HQ, so the blog post will be delayed until later
[25-Aug-2009 15:17:40] <mrayzenoss> but the release notes are there and so are the bits
[25-Aug-2009 15:19:14] <mrayzenoss> BTW, iPhone tethering rocks
[25-Aug-2009 15:19:29] <chudler> yes. yes it does.
[25-Aug-2009 15:19:42] <chudler> If only I could make it work with linux
[25-Aug-2009 15:20:03] <mrayzenoss> heh, it barely works with OSX
[25-Aug-2009 15:20:45] <chudler> I have heard that bluetooth works fine everywhere, since it is a standard protocol. NO USB, though
[25-Aug-2009 15:21:00] <chudler> my netbook doesnt have USB :-( I got on the bandwagon too soon, I guess
[25-Aug-2009 15:21:02] <mrayzenoss> well, tethering isn't supported by AT&T
[25-Aug-2009 15:21:09] <chudler> I mean, it doesnt have bluetooth
[25-Aug-2009 15:21:13] <mrayzenoss> so to get it to work is a bit of hackery
[25-Aug-2009 15:21:17] <mrayzenoss> I'm actually using USB
[25-Aug-2009 15:21:17] <straterra> mine has bluetooth..wifi..usb..
[25-Aug-2009 15:25:06] <rocket> hey matt
[25-Aug-2009 15:25:10] <mrayzenoss> yo
[25-Aug-2009 15:25:22] <straterra> i use tethering with my verizon phone over wifi
[25-Aug-2009 15:26:39] <rocket> mrayzenoss: how goes the zenpack committing?
[25-Aug-2009 15:27:00] <mrayzenoss> I checked 9 into SVN
[25-Aug-2009 15:27:08] <mrayzenoss> I'm writing up webpages for them on the flight
[25-Aug-2009 15:27:14] <mrayzenoss> I've got all the screenshots uploaded
[25-Aug-2009 15:27:23] <mrayzenoss> so there should be a flurry of posts tonight
[25-Aug-2009 15:28:01] <chudler> amd the dramatic conclusion to the contest? ;-)
[25-Aug-2009 15:28:20] <mrayzenoss> ahh, the judges are still reviewing
[25-Aug-2009 15:28:29] <rocket> heh ..
[25-Aug-2009 15:28:35] <mrayzenoss> the list of 50 entries will be up tonight probably
[25-Aug-2009 15:28:47] <mrayzenoss> and I should have the judging done by Friday
[25-Aug-2009 15:28:59] <gwb235> if there's a community zenpack that you're using that's having problems, and you have support agreement with zenoss, does that cover helping debug why the community zenpack is having problems?
[25-Aug-2009 15:29:42] <mrayzenoss> gwb235: probably depends how bad it is
[25-Aug-2009 15:29:53] <mrayzenoss> I mean, officially they won't support it but they probably will
[25-Aug-2009 15:30:11] <mrayzenoss> Support debugs tons of customer issues with their own ZenPacks
[25-Aug-2009 15:30:15] <mrayzenoss> gwb235: which one?
[25-Aug-2009 15:30:58] <gwb235> chudler's SentryCDU bit. It works great about 60% of the time and I feel bad bothering him taking time away from his real job.
[25-Aug-2009 15:31:10] <mrayzenoss> heh, he's a customer too
[25-Aug-2009 15:31:21] <chudler> eheh, I should really fix that! ehe. I wouldnt pass the buck :-)
[25-Aug-2009 15:31:22] <mrayzenoss> gwb235: are you using his new .egg version?
[25-Aug-2009 15:31:28] <gwb235> yes
[25-Aug-2009 15:31:44] <gwb235> we've talked a bit, it's not a critical failure, just something I need to close the loop on
[25-Aug-2009 15:32:08] <chudler> Is it the thing where the page gives an error while viewing the "details"? That really is a zenoss fault, but everything else is me
[25-Aug-2009 15:32:40] <gwb235> it's the thing where i can snmpwalk the device and get the right output, but the aggregate load graphs are giving me negative numbers
[25-Aug-2009 15:33:16] <chudler> oh damn, yeah thats me with a bug somewhere, I suspect.
[25-Aug-2009 15:33:53] <gwb235> i tried using --showdevicereusults but am not experiences/skilled enough to follow the data in --> performance graph RRD values
[25-Aug-2009 15:36:07] <chudler> hmm. We should work together on this. I might have some free time in 1-2 weeks :-)
[25-Aug-2009 15:36:46] <gwb235> let me know, I'll be glad to help debug anytime. maybe a shared VNC session or something so you can see first hand
[25-Aug-2009 15:39:45] <gwb235> any help, much appreciated. I have some longhorn schwag if that helps sweeten the pot.
[25-Aug-2009 15:41:31] <chudler> I will let you know. I really want this to be fixed. Work priorities are not in line with it at the moment, but I am getting there.
[25-Aug-2009 15:41:40] <magnachef> what's the goal release date of King Crab?
[25-Aug-2009 15:42:23] <mrayzenoss> magnachef: this fall
[25-Aug-2009 15:42:33] <mrayzenoss> beta should start in the next few weeks
[25-Aug-2009 15:42:41] <cgibbons> they have me chained to my desk so i won't leave
[25-Aug-2009 15:42:52] <magnachef> haha...nice!
[25-Aug-2009 15:43:21] <magnachef> I'm writing up my documentation for our next NMS guy, so I'm leaving a note about this
[25-Aug-2009 15:43:48] <magnachef> do you guys have a link to enhancements, bugfixes, etc?
[25-Aug-2009 15:43:58] <chudler> what is this "documentation" of which you speak ;-)
[25-Aug-2009 15:44:23] <magnachef> documenting my current job, as I'm moving on to a programming job
[25-Aug-2009 15:44:24] <mrayzenoss> magnachef: for 2.4.3?
[25-Aug-2009 15:45:00] <magnachef> well, more for what King Crab will give us...I'm currently 2.3.3
[25-Aug-2009 15:45:18] <mrayzenoss> magnachef: http://www.zenoss.com/community/docs/
[25-Aug-2009 15:45:34] <mrayzenoss> ahh, the mysterious roadmap
[25-Aug-2009 15:45:34] <magnachef> ...basically leaving upgrades up to the new guy :-)
[25-Aug-2009 15:46:09] <magnachef> yeah, roadmap sounds like what I'm looking for...I'm sure it's ever changing...but at least having some big hitters
[25-Aug-2009 15:46:13] <sgsax> we've been using zenoss for a few months now for basic system monitoring and hve been pretty happy with it
[25-Aug-2009 15:46:50] <sgsax> but today, I've been trying to add a custom snmp monitor for some temp probes we just purchased and am getting stuck
[25-Aug-2009 15:47:00] <mrayzenoss> we don't have a public roadmap for King Crab, I'll hit up PM to publish something before we beta it
[25-Aug-2009 15:47:17] <sgsax> I don't *need* to install the MIBs for it, because I already know the full OID, but it'd be nice if I could
[25-Aug-2009 15:47:26] <sgsax> but I seem to be having path problems
[25-Aug-2009 15:47:47] <magnachef> mrayzenoss: thanks
[25-Aug-2009 15:48:27] <sgsax> anybody offer a hand?
[25-Aug-2009 15:48:31] <mrayzenoss> you can look up the closed tickets http://dev.zenoss.org/trac/report/5
[25-Aug-2009 15:48:31] <mrayzenoss> the other reports here show what's in the queue, what's been reviewed, etc. http://dev.zenoss.org/trac/report
[25-Aug-2009 15:48:31] <mrayzenoss> flight's boarding, gotta go
[25-Aug-2009 15:49:28] <jb> when is ent 2.4.3 dropping?
[25-Aug-2009 15:49:36] <mrayzenoss> sgsax: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch06s02.html double check the .0 on the oid
[25-Aug-2009 15:49:41] <mrayzenoss> 2.4.3 dropped
[25-Aug-2009 15:49:41] <mrayzenoss> and I'm out
[25-Aug-2009 15:49:59] <jb> not enterprise
[25-Aug-2009 15:50:29] <sgsax> I got the trailing .0
[25-Aug-2009 15:50:41] <jb> whats the problem?
[25-Aug-2009 15:50:44] <sgsax> I've read through that chapter, and think I have everything added correctly
[25-Aug-2009 15:50:45] <jb> are you getting OID is bad errors?
[25-Aug-2009 15:51:15] <sgsax> but I'm a bit confused about data points and data point aliases
[25-Aug-2009 15:51:23] <sgsax> and my graphs aren't coming out right
[25-Aug-2009 15:53:21] <jb> i really need to figure out a way to capture HBA I/O stats at the host level
[25-Aug-2009 15:55:47] <rocket> jb: for aix?
[25-Aug-2009 15:55:58] <jb> aix/windows/linux
[25-Aug-2009 15:56:11] <jb> we use IBM TPC for monitoring the san fabric and such
[25-Aug-2009 15:56:17] <jb> but i really want to be able to capture local I/O at the host
[25-Aug-2009 15:57:10] <rocket> ah I see
[25-Aug-2009 15:58:42] <jb> i can get windows i/o using the informant MIB
[25-Aug-2009 16:00:39] <chudler> I get tons of info from /proc/scsi/qla2xxx/*, but it is very specific to those cards
[25-Aug-2009 16:00:48] <jb> yeah
[25-Aug-2009 16:02:41] <jb> egor?
[25-Aug-2009 16:02:43] <jb> oops
[25-Aug-2009 16:04:28] <jb> for windows its all good and easy
[25-Aug-2009 16:04:40] <jb> sucks that it can't be common across all hosts though
[25-Aug-2009 16:16:06] <chudler> If its just I/O, that is abstracted by the linux disk subsystem, but it won't tell you the actual bytes the fiber is passing. I am mildly surprised that windows unifies that among the drivers
[25-Aug-2009 16:19:44] <jb> yeah its the same for windows
[25-Aug-2009 16:20:00] <jb> its not really coming from the fiber
[25-Aug-2009 17:19:13] <kgoedtel> haha
[25-Aug-2009 17:19:24] <kgoedtel> oops
[25-Aug-2009 18:10:33] <cgibbons> hurm
[25-Aug-2009 19:35:57] <ideopathic> do zenpacks need to be installed as the zenoss user?
[26-Aug-2009 00:00:46] [disconnected at Wed Aug 26 00:00:46 2009]
[26-Aug-2009 00:00:46] [connected at Wed Aug 26 00:00:46 2009]
[26-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[26-Aug-2009 06:37:47] edwardam1 is now known as edwardam
[26-Aug-2009 07:35:47] <ckrough> ah, coming in to gappy graphs. good morning to you too zenoss
[26-Aug-2009 08:28:15] <Apocalipse> hi people
[26-Aug-2009 08:28:28] <tripitaka> Hi, I have upgraded to 2.4.3 and I'm now getting these errors: http://www.pastebin.ca/1543144 - is it safe to downgrade by running an older stack installer?
[26-Aug-2009 08:29:28] <Apocalipse> maybe Master rmatte can help u
[26-Aug-2009 08:29:30] <Apocalipse>
[26-Aug-2009 08:38:51] <Apocalipse> does anybody knows how to change RRD graphs?
[26-Aug-2009 08:51:38] kevin7kal_ is now known as kevin7kal
[26-Aug-2009 09:28:04] <jb> venturaville: here?
[26-Aug-2009 09:30:47] <venturaville> jb: I'm here
[26-Aug-2009 09:31:50] <Apocalipse> hello people...
[26-Aug-2009 09:32:10] <Apocalipse> i would like to know if someone knows how to change RRD graphs
[26-Aug-2009 09:33:25] <Apocalipse> people?
[26-Aug-2009 09:33:43] <mrayzenoss> what do you want to change?
[26-Aug-2009 09:34:32] <Apocalipse> http://imagebin.org/61076
[26-Aug-2009 09:34:56] <Apocalipse> this graph is cpu percent usage...
[26-Aug-2009 09:35:06] <Apocalipse> i would like to change the units...
[26-Aug-2009 09:35:26] <Apocalipse> its showing 8.0 M , 6.0 M...etc
[26-Aug-2009 09:36:37] <mrayzenoss> back on the template, on the graph definition is where you put units
[26-Aug-2009 09:38:44] <Apocalipse> i went there...
[26-Aug-2009 09:38:47] <Apocalipse> put %
[26-Aug-2009 09:39:59] <mrayzenoss> I think you may have to remove the existing .RRD file to get the updated file format
[26-Aug-2009 09:40:38] <Apocalipse> hunnn
[26-Aug-2009 09:41:53] <Matir> Anyone ever seen Zenoss switch a process from unmonitored to monitored repeatedly? I keep getting alerts that process httpd is not running, even though I set it not to monitor that. (Additionally, httpd IS running)
[26-Aug-2009 09:42:05] <jb> ok, i cannot for the life of me figure out why this graph is showing all 0's.. i deleted the RRD, deleted the device, etc
[26-Aug-2009 09:42:13] <jb> verified that the OID's are good
[26-Aug-2009 09:43:38] <jb> venturaville: was it you that was talking about SVC?
[26-Aug-2009 09:43:43] <Apocalipse> matt...i got this in "graph point" format=%5.2lf%s
[26-Aug-2009 09:44:10] <venturaville> jb: I think multiple people have but yes I have been exploring using CIM for SVC and LSI Arrays.
[26-Aug-2009 09:44:32] <venturaville> unfrotunately CIM for SVC does not have performance data ... only structural and status data
[26-Aug-2009 09:45:03] <jb> have you looked at SNMP?
[26-Aug-2009 09:45:03] <venturaville> unless it is just somehow buried in there
[26-Aug-2009 09:45:12] <venturaville> jb: there isn't squat in SNMP
[26-Aug-2009 09:45:19] <jb> pfft
[26-Aug-2009 09:45:32] <jb> i was just looking at this "Inventory Email" that we get from it
[26-Aug-2009 09:45:35] <jb> and it has lots of data
[26-Aug-2009 09:46:11] <jb> mostly modeling data though
[26-Aug-2009 09:46:58] <venturaville> yep ... basically the same as what is in CIM
[26-Aug-2009 09:47:19] <venturaville> we have been wondering for a while if TPC actually gets its 'performance' data from the arrays to do its calculations for it
[26-Aug-2009 09:47:47] <venturaville> performance data is available in CIM for those
[26-Aug-2009 09:49:06] <jb> do you guys have TPC?
[26-Aug-2009 09:50:31] <venturaville> only parts of it
[26-Aug-2009 10:11:11] <rmatte> mrayzenoss: the upgrades to 2.4.3 that I did last night went very smoothly, and there is noticeable performance improvement in 2.4.3 compared to 2.4.1 and 2.3.3, which was nice to see.
[26-Aug-2009 10:14:33] <mrayzenoss> yay
[26-Aug-2009 10:32:20] <cgibbons> hmmm
[26-Aug-2009 10:49:36] <jb> when is 2.4.3 enterprise dropping?
[26-Aug-2009 10:52:27] <mrayzenoss> I think it's available through the support portal
[26-Aug-2009 10:56:08] <jb> oh
[26-Aug-2009 10:56:09] <jb> it is
[26-Aug-2009 10:56:12] <jb> it wasn't yesterday
[26-Aug-2009 11:02:54] <rmatte> yesterday they only released 2.4.3 around 6 or 7pm EST
[26-Aug-2009 11:25:19] <jb> hm
[26-Aug-2009 11:25:40] <jb> so, im trying to add Informant's I/O stats to the "Filesystem" box on the Software tab
[26-Aug-2009 11:25:44] <jb> instead of the hardware tab
[26-Aug-2009 11:25:54] <jb> the /Devices/Server/Filesystem template
[26-Aug-2009 11:26:05] <jb> not sure this is going to work though
[26-Aug-2009 11:30:42] <rmatte> you're just trying to do it by moving the template?
[26-Aug-2009 11:31:07] <rmatte> I think different types of templates have pretty concrete display paramaters
[26-Aug-2009 11:31:19] <jb> well, i took the IO datasources from the "HardDisk" template and created them on the "FileSystem" template
[26-Aug-2009 11:31:30] <rmatte> ah
[26-Aug-2009 11:31:35] <jb> im not sure the indexes are workign properly
[26-Aug-2009 11:31:38] <jb> or something
[26-Aug-2009 11:31:44] <rmatte> I think those datasources are collected by a collector plugin
[26-Aug-2009 11:34:40] <jb> which?
[26-Aug-2009 11:38:28] <rmatte> whichever has to do with filesystem
[26-Aug-2009 11:38:37] <rmatte> I don't know the full name off the top of my head
[26-Aug-2009 11:39:49] <jb> i think you need to somehow index the OID's
[26-Aug-2009 11:39:55] <jb> since 1.3.6.1.4.1.9600.1.1.1.1.18 contains stats for all disks
[26-Aug-2009 11:40:04] <jb> but the template doesn't know which is C:, D:, E:, etc
[26-Aug-2009 11:42:08] <jb> trying to figure out where that modeler is
[26-Aug-2009 11:42:28] <jb> cgibbons: you are the windows dude.. are you around by chance?
[26-Aug-2009 11:42:59] <rmatte> seems like moving it is more trouble than it's worth
[26-Aug-2009 11:43:04] <jb> well
[26-Aug-2009 11:43:07] <jb> the problem is..
[26-Aug-2009 11:43:21] <jb> egor's DellMon pack overrides the normal "Hard Disks" box in "Hardware"
[26-Aug-2009 11:43:39] <jb> so the I/O graphs no longer work
[26-Aug-2009 11:43:51] <jb> because its modeling that data directly from Dell's RAID controller
[26-Aug-2009 11:44:08] <jb> so i figured putting them in the Filesystems template would be good
[26-Aug-2009 11:44:58] <rmatte> ah
[26-Aug-2009 11:45:35] <rmatte> I just use the zenoss dell monitor zenpack instead
[26-Aug-2009 11:45:38] <rmatte> rather than egor's
[26-Aug-2009 11:45:50] <jb> egor's provides much more data
[26-Aug-2009 11:46:08] <rmatte> yeh, but I'm not generally interested in any of the other data it provides
[26-Aug-2009 11:46:22] <jb> a
[26-Aug-2009 11:46:24] <jb> h
[26-Aug-2009 12:56:47] <cgibbons> i am now, jb
[26-Aug-2009 13:01:54] <jb> sup cgibbons..
[26-Aug-2009 13:02:07] <cgibbons> you asked about winders earlier?
[26-Aug-2009 13:02:14] <jb> im trying to figure out how to move the I/O graphs from the InformantHardDisk to the Filesystems template
[26-Aug-2009 13:02:25] <jb> and im guessing there is some snmp indexing problems
[26-Aug-2009 13:10:47] <rmatte> woohoo, I just got a zenny zebra
[26-Aug-2009 13:11:01] <jb> hah
[26-Aug-2009 13:11:07] <jb> i have like 20 of those things floating around the office
[26-Aug-2009 13:11:12] <rmatte> lol
[26-Aug-2009 13:11:17] <rmatte> got a zebra, a pen, and a shirt
[26-Aug-2009 13:11:30] <rmatte> now I just need my jacket and I'll have the full ensemble
[26-Aug-2009 13:11:32] <jb> a NO NODE LEFT BEHIND shirt?
[26-Aug-2009 13:11:40] <gwb2351> if any of the zenoss guys in Austin want to come down to UT, I'll trade a longhorn for a zebra...
[26-Aug-2009 13:11:40] <rmatte> no, it says...
[26-Aug-2009 13:12:01] <rmatte> "Making enterprise management weird"
[26-Aug-2009 13:12:05] <jb> oh ih ave that one
[26-Aug-2009 13:12:28] <cgibbons> zebras.... http://yfrog.com/ccuvyj
[26-Aug-2009 13:12:28] <rmatte> It's a nice shirt
[26-Aug-2009 13:12:47] <cgibbons> i ride by there on my way home, gwb (live downtown)
[26-Aug-2009 13:12:57] <rmatte> cgibbons: haha, nice
[26-Aug-2009 13:13:09] <rmatte> I have 4 zebras here, but I can't keep them all
[26-Aug-2009 13:13:30] <rmatte> need to give some to other members of my team
[26-Aug-2009 13:13:50] <gwb2351> pick a day next week, bother rbriggs, and maybe we can do lunch + token schwag exchange
[26-Aug-2009 13:14:34] <jb> cgibbons: would you happen to have any idea on how to do that?
[26-Aug-2009 13:15:20] <cgibbons> well you would think it would just work it just doesn't collect after the datapoints have been moved?
[26-Aug-2009 13:15:28] <jb> nope
[26-Aug-2009 13:15:38] <jb> well.. how does it know how to index properly?
[26-Aug-2009 13:15:55] <jb> lDiskDiskWriteBytesPerSec = 1.3.6.1.4.1.9600.1.1.1.1.18
[26-Aug-2009 13:16:03] <jb> which contains all drives
[26-Aug-2009 13:16:03] <cgibbons> ah so it's instance based?
[26-Aug-2009 13:16:06] <jb> yeah
[26-Aug-2009 13:16:22] <cgibbons> the modeler plugins are what does the indexing usually so lesee
[26-Aug-2009 13:17:01] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.67.58 = STRING: "C:"
[26-Aug-2009 13:17:01] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.68.58 = STRING: "D:"
[26-Aug-2009 13:18:47] <cgibbons> okay, I think (danger) you'll need to make a new modeling plugin that collects from the right OIDs so that it can index the FileSystem components you're after. if you look @ HRFileSystemMap.py right now it does that, but for .1.3.6.1.2.1.25.2.3.1 in the host-resources-v2-mib
[26-Aug-2009 13:19:34] <jb> one sec
[26-Aug-2009 13:19:35] <cgibbons> you could, if you want to experiment, manually index using zendmd
[26-Aug-2009 13:19:38] <rmatte> well, I officially, finally, have all of my linux boxes with the same ZenPacks running at the same version
[26-Aug-2009 13:19:40] <rmatte> w00t
[26-Aug-2009 13:19:51] <cgibbons> as long as they are indexed, zenperfsnmp should collect the data (once told of the change)
[26-Aug-2009 13:23:08] <jb> hmmm
[26-Aug-2009 13:25:00] <cgibbons> of course the tricky thing is I think components can only have one snmpindex right now hrm
[26-Aug-2009 13:25:22] <jb> yeah '
[26-Aug-2009 13:25:33] <jb> basically im just trying to figure out how to display this I/O data
[26-Aug-2009 13:26:01] <cgibbons> does this need to be generic so when you model any new Windows device you'll see it in your template, or do you just need it for a couple of servers, etc?
[26-Aug-2009 13:26:02] <jb> because.. egor's DellMonPack override the default HardDisks on the "Hardware" tab with disks that are modeled directly from Dell's Raid controller
[26-Aug-2009 13:26:15] <cgibbons> yeah
[26-Aug-2009 13:26:22] <jb> so i need to display all logical windows drives and their corresponding I/O somewhere
[26-Aug-2009 13:26:32] <jb> every windows server
[26-Aug-2009 13:30:56] <cgibbons> talking to chet about it internally... he said no, can't index multiple objects on a single component. when he's done stuff like that before he's used a command plugin to collect the secondary data.
[26-Aug-2009 13:31:26] <jb> hm ok
[26-Aug-2009 13:31:27] <cgibbons> so that would be a combination of a new modeler to discover the indexes during modeling, and then a command plugin to monitor the data :
[26-Aug-2009 13:31:37] <cgibbons> not a lot of code but not just super easy either
[26-Aug-2009 13:31:39] <jb> yeah.. seems like too much trouble
[26-Aug-2009 13:32:03] <cgibbons> might be easier to change the host template to have your common 2 or 3 file systems in the main perf tab
[26-Aug-2009 13:32:20] <jb> any other suggestions on how to do it?.. I guess i could just create a new modeler and add a new box to the hardware tab
[26-Aug-2009 13:32:23] <jb> hm yeah
[26-Aug-2009 13:33:07] <cgibbons> if your servers are somewhat standardized you might catch 80% of what you need right off the bat
[26-Aug-2009 13:33:30] <jb> only thing is.. some servers have more logical drives than others, etc
[26-Aug-2009 13:33:34] <jb> c,d,e,f,g,h etc
[26-Aug-2009 13:33:44] <cgibbons> yeah
[26-Aug-2009 13:34:12] <cgibbons> that data should be available via WMI, no?
[26-Aug-2009 13:34:25] <jb> no
[26-Aug-2009 13:34:30] <jb> at least i don't think it is
[26-Aug-2009 13:34:32] <cgibbons> really? hrm lemme look
[26-Aug-2009 13:34:39] <jb> i think all thats available via WMI is combined IOPS?
[26-Aug-2009 13:34:46] <rmatte> you can get anything you can via snmp via wmi
[26-Aug-2009 13:34:54] <rmatte> you just need to figure out how to get it
[26-Aug-2009 13:35:00] <rmatte> it's not available by default in core
[26-Aug-2009 13:37:07] <cgibbons> and lDiskDiskWriteBytesPerSec is a logical disk counter not a physical disk counter, right?
[26-Aug-2009 13:39:02] <jb> yeah as far as I know
[26-Aug-2009 13:39:13] <cgibbons> have you looked @ Win32_PerfFormattedData_PerfDisk_LogicalDisk ?
[26-Aug-2009 13:39:41] <jb> you think this would be easier using WMI?
[26-Aug-2009 13:40:15] <cgibbons> 'select name,DiskWriteBytesPerSec from Win32_PerfFormattedData_PerfDisk_LogicalDisk' will do it, provided logical disk stats are enabled on the box
[26-Aug-2009 13:40:36] <cgibbons> dunno if it would be easier, you still have the whole component modeling/collection issue to deal with
[26-Aug-2009 13:41:20] <bigegor> how about wbem?
[26-Aug-2009 13:41:54] <cgibbons> wbem == wmi
[26-Aug-2009 13:42:22] <bigegor> wmi != wbem
[26-Aug-2009 13:43:34] <bigegor> wmi = CIM over DCOM but wbem = CIM over http[s]
[26-Aug-2009 13:44:03] <venturaville> http://www.serverwatch.com/tutorials/article.php/2178391/Understanding-WMI-WBEM-CIM-etc.htm
[26-Aug-2009 13:44:09] <cgibbons> it's still the same queries into the CIM namespace, getting the data out of it isn't what jb we're having issues with
[26-Aug-2009 13:45:02] <jb> bleh im thinking just writing a new modeler and displaying it separate of filesystems somewhere
[26-Aug-2009 13:45:17] <cgibbons> getting the data out of the windows box is easy, organizing it into the zenoss infrastructure in a natural way is what really the problem is here
[26-Aug-2009 13:45:25] <jb> exactly
[26-Aug-2009 13:46:18] <jb> theres just no easy/logical way to fit it in
[26-Aug-2009 13:47:00] <cgibbons> is there any relationship between the index in that other mib back to host-resources-v2?
[26-Aug-2009 13:47:12] <jb> i dont believe so
[26-Aug-2009 13:47:16] <jb> it has some funky indexing
[26-Aug-2009 13:47:19] <bigegor> no
[26-Aug-2009 13:47:37] <cgibbons> bummer
[26-Aug-2009 13:47:46] <jb> the index i like "2.67.58"
[26-Aug-2009 13:47:55] <jb> is
[26-Aug-2009 13:48:17] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.67.58 = STRING: "C:"
[26-Aug-2009 13:48:17] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.68.58 = STRING: "D:"
[26-Aug-2009 13:48:19] <jb> which is that
[26-Aug-2009 13:52:54] <cgibbons> yeah I dunno for you if SNMP or WMI based would be easier, but a plugin and a command data-type in your template would dit
[26-Aug-2009 13:52:57] <cgibbons> er do it
[26-Aug-2009 13:53:09] <jb> actually
[26-Aug-2009 13:53:14] <jb> i can get it from PhysicalDisk
[26-Aug-2009 13:53:40] <jb> at \PhysicalDisk\Disk Read Bytes/sec
[26-Aug-2009 13:54:23] <cgibbons> yeah and physical disk counters are always enabled by default iirc
[26-Aug-2009 13:54:23] <jb> i wouldnt know how to do it in WMI though, I don't think..
[26-Aug-2009 13:54:27] <jb> right
[26-Aug-2009 13:55:43] <jb> i would need to loop thru all instances and display them
[26-Aug-2009 13:58:53] <venturaville> wmi is a pain to loop through
[26-Aug-2009 13:59:01] <venturaville> I'm working with it right now for the lsi array stuff
[26-Aug-2009 13:59:14] <jb> yeah.. no idea where to even begin with that
[26-Aug-2009 13:59:43] <venturaville> try yawn
[26-Aug-2009 13:59:52] <venturaville> (though I don't know if it can talk directly to windows WMI)
[26-Aug-2009 14:01:31] <cgibbons> this at least provides data: wmic -U '.\me' //somewhere.local 'select name,DiskWriteBytesPerSec from Win32_PerfRawData_PerfDisk_PhysicalDisk'
[26-Aug-2009 14:02:02] <cgibbons> but then you've got that standard funky stuff with partitions vs. logical volumes
[26-Aug-2009 14:02:44] <jb> yeah.. ok since someone at zen already did it using snmp
[26-Aug-2009 14:03:02] <jb> can't i just take that modeler and stick it in a new "box" on the hardware tab?
[26-Aug-2009 14:03:17] <jb> the standard windows "Hard Disks" box is all I need..
[26-Aug-2009 14:06:08] <jb> egor.. that sound possible to you?
[26-Aug-2009 14:08:30] <rocket> jb ... impossible ... jk ...
[26-Aug-2009 14:08:35] <bigegor> i would like try to implement CIM based device model in zenoss
[26-Aug-2009 14:08:39] <jb>
[26-Aug-2009 14:08:52] <jb> well, i mean.. zenoss is already grabbing and indexing this data properly
[26-Aug-2009 14:08:58] <jb> i just need to make it display somewhere!
[26-Aug-2009 14:09:13] <rocket> jb get me up to speed
[26-Aug-2009 14:09:29] <rocket> you have data obtained and stored in the model?
[26-Aug-2009 14:09:32] <jb> ok
[26-Aug-2009 14:09:51] <cgibbons> there's already support in core for modeling via WMI, there's just no plugins provided in core that do it, nor are there any collectors in core that use it, but the framework is there
[26-Aug-2009 14:10:06] <jb> by default.. zenoss uses the windows informant MIB to display "Hard Disks" on the hardware tab.. this template (/Devices/Server/Windows/HardDisks) contains I/O graphs obtained from the informant MIB
[26-Aug-2009 14:10:42] <jb> but, when egor's DellMon Pack is used.. it overrides all things in the hardware tab and replaces hard disks with ones that are modeled from Dell's management agent
[26-Aug-2009 14:11:15] <jb> which a.) only displays local drives attached to dell's raid controller and b.) dell doesnt provide i/o counters
[26-Aug-2009 14:11:36] <jb> so, im basically trying to take this HardDisks data (IO) and display it somewhere else
[26-Aug-2009 14:11:47] <rocket> jb: so basically you want to take the dellmon pack and remove the hard disk stuff?
[26-Aug-2009 14:11:50] <jb> cgibbons: im enterprise..
[26-Aug-2009 14:12:02] <cgibbons> oh, detail...
[26-Aug-2009 14:12:05] <cgibbons> easy
[26-Aug-2009 14:12:10] <jb> rocket: well, no, because that data is useful too.
[26-Aug-2009 14:12:25] <cgibbons> you can monitor the PerfMon path directly in enterprise
[26-Aug-2009 14:12:34] <jb> but, i also need to display the I/O data for every disk
[26-Aug-2009 14:12:36] <cgibbons> and the component is probably already being modeled
[26-Aug-2009 14:13:47] <jb> hmm!
[26-Aug-2009 14:14:00] <cgibbons> you good with zendmd?
[26-Aug-2009 14:14:08] <jb> not really :/
[26-Aug-2009 14:15:09] <rocket> jb: ok I think I know your solution
[26-Aug-2009 14:15:21] <bigegor> jb: what happend if you remove DellHardDiskMap from zCollectorPlugins list and remodel your server?
[26-Aug-2009 14:15:29] <jb> sec
[26-Aug-2009 14:16:27] <rocket> bigegor: your zenpack overwrites the hard disk template correct?
[26-Aug-2009 14:17:27] <cgibbons> jb: try this, too, inside zendmd: find("yourdevicehere").os.filesystems.C.perfmonInstance
[26-Aug-2009 14:18:03] <bigegor> no, it add new DellHardDisk template
[26-Aug-2009 14:18:19] <jb> cgibbons: AttributeError: 'NoneType' object has no attribute 'os'
[26-Aug-2009 14:18:35] <cgibbons> you choose a valid device name inside the find?
[26-Aug-2009 14:18:58] <rocket> bigegor: it leaves the old harddisk template?
[26-Aug-2009 14:18:59] <jb> oh oops.. typo.
[26-Aug-2009 14:19:08] <rmatte> maybe he needs to do d = find("bleh") then d.os.filesystems.C.perfmonInstance?
[26-Aug-2009 14:19:10] <rocket> bigegor: how are you overridding the template?
[26-Aug-2009 14:19:20] <rmatte> type
[26-Aug-2009 14:19:23] <rmatte> typo rather
[26-Aug-2009 14:19:34] <jb> cgibbon: yah that didn't work.. same error
[26-Aug-2009 14:19:44] <bigegor> it doesn't touch HardDisk template
[26-Aug-2009 14:20:12] <jb> bigegor: so what makes it dissapear?
[26-Aug-2009 14:20:26] <cgibbons> k, that would be there depending upon which plugin(s) you have bound. but in any event, if you've got enterprise, we can get your components modeled with a perfmon instance, and then you can add any perfmon datasource to the template and magic happens, independent of the snmp stuff
[26-Aug-2009 14:21:21] <rocket> jb: oh yeah I bet he is overriding the page template
[26-Aug-2009 14:21:23] <jb> uh oh
[26-Aug-2009 14:21:28] <jb> now im erroring out everywhere.
[26-Aug-2009 14:21:49] <jb> yeah thats what hes doing
[26-Aug-2009 14:21:59] <rocket> jb: eg for what you are looking for you need to hack the page template to have both displayed
[26-Aug-2009 14:22:18] <jb> exacto.
[26-Aug-2009 14:22:34] <jb> ive somehow managed to botch this production instance
[26-Aug-2009 14:22:38] <jb> without doing anything
[26-Aug-2009 14:23:35] <bigegor> component with meta_type = DellHardDisk bind DellHardDisk tempate automatically
[26-Aug-2009 14:24:50] <rocket> ok I got it ...
[26-Aug-2009 14:25:03] <jb> well, now zen is completely broken
[26-Aug-2009 14:25:14] <cgibbons> yikes
[26-Aug-2009 14:25:21] <jb> and i have no idea how that happened!
[26-Aug-2009 14:25:23] <jb> sec
[26-Aug-2009 14:28:12] <jb> zope just dies
[26-Aug-2009 14:28:40] <cgibbons> uh
[26-Aug-2009 14:28:51] <cgibbons> let's blame rocket, he's more idle
[26-Aug-2009 14:29:03] <jb> well now its running
[26-Aug-2009 14:29:07] <jb> but not responding to anything via the web
[26-Aug-2009 14:30:07] <jb> hrm, nothing in logs
[26-Aug-2009 14:31:43] <jb> yeah, zope just keeps dying :/
[26-Aug-2009 14:40:11] <rocket> cgibbons: !!!!
[26-Aug-2009 14:40:39] <rocket> blame me not! It wasnt my zenpack that removed the old info .. ;p
[26-Aug-2009 14:40:56] <jb> hah
[26-Aug-2009 14:40:59] <cgibbons>
[26-Aug-2009 14:41:21] <rmatte> yikes, you fried your zope?
[26-Aug-2009 14:41:27] <jb> looks like it.
[26-Aug-2009 14:41:31] <rocket> bigegor: if I could make a recommendation .. I would update your page template
[26-Aug-2009 14:41:36] <jb> i have no idea how..
[26-Aug-2009 14:41:53] <rocket> bigegor: if the original modellor collects the hdisk data display the info if found
[26-Aug-2009 14:42:01] <rmatte> jb: what errors are you seeing in the zopectl log file?
[26-Aug-2009 14:42:06] <rocket> bigegor: else do not display it ..
[26-Aug-2009 14:42:14] <jb> http://pastebin.com/m2d80162d
[26-Aug-2009 14:42:35] <rocket> bigegor: my aix zenpack has a couple example areas that do that in the page templates around the printer and print queues
[26-Aug-2009 14:42:49] <rmatte> unknown status byte value, hmmm
[26-Aug-2009 14:44:12] <rmatte> For starters, here's the code that generates that error: http://pastebin.com/m1ba60cb3
[26-Aug-2009 14:44:36] <bigegor> rocket: what page template do you mean?
[26-Aug-2009 14:44:41] <cgibbons> okay, ian says that's just a bad cache. move that zeo1-1.zec file outside of $ZENHOME/var (shutdown zeo first and then restart it)
[26-Aug-2009 14:45:08] <rocket> bigegor: advanced details page template you wrote
[26-Aug-2009 14:45:39] <rocket> bigegor: you only reference the hp disk data and it is automatically pulling in the hp modelled information
[26-Aug-2009 14:46:30] <rocket> the web interface needs to be updated to display the other disk information if collected from the original modeller
[26-Aug-2009 14:46:51] <rocket> or your new modeller should be updated to have the data jb is looking for
[26-Aug-2009 14:47:00] <jb> cgibbons: woot.. and we're back.
[26-Aug-2009 14:47:01] <jb>
[26-Aug-2009 14:47:04] <jb> thanks
[26-Aug-2009 14:47:06] <rmatte> hehe
[26-Aug-2009 14:47:14] <jb> rmatte: gj tracking down that code too..
[26-Aug-2009 14:47:17] <bigegor> ok. i will try
[26-Aug-2009 14:47:19] <jb> it was cache related just as that snippet implied
[26-Aug-2009 14:49:02] <rmatte> yeh, I forgot about the .zec file though, I remember having to do that once over half a year ago
[26-Aug-2009 14:49:13] <jb> ok, so bascially we need to alter the advanced details hardware template to tell it to display the original info..
[26-Aug-2009 14:49:20] <rocket> jb: yes
[26-Aug-2009 14:49:31] <rocket> jb: thats exactly what is missing
[26-Aug-2009 14:49:55] <rocket> jb: the original page template needs to be merged with the new advanced details page template
[26-Aug-2009 14:50:19] <jb> right
[26-Aug-2009 14:50:26] <jb> just parts of it, at least
[26-Aug-2009 14:50:46] <rocket> yes the disk parts of it
[26-Aug-2009 14:53:57] <jb> and this would be deviceHardwareDetail.pt?
[26-Aug-2009 14:55:51] <rocket> yes
[26-Aug-2009 14:55:56] <rocket> I believe so
[26-Aug-2009 15:04:39] <bigegor> i have done some tests. its work. Without DellHardDisk and with UCDHardDiskMap (Linux HD performace). it shows all my disks
[26-Aug-2009 15:04:56] <bigegor> after modeling
[26-Aug-2009 15:12:35] <jb> hm
[26-Aug-2009 15:12:53] <jb> so it shows you the dell hard disk model and the normal one?
[26-Aug-2009 15:14:40] <bigegor> no dell disks but disks from OS
[26-Aug-2009 15:14:52] <jb> ah.. any way to display them both?
[26-Aug-2009 15:19:21] <rocket> jb: he might have to extend the device model to show both
[26-Aug-2009 15:21:32] <bigegor> thanks rocket. That's right. End another problem UCD and Informant is OS specific
[26-Aug-2009 15:21:42] <jb> yeah
[26-Aug-2009 15:21:45] <jb> its annoying
[26-Aug-2009 15:22:52] <jb> yeah informant stuff should only be bound to /Devices/Server/Windows
[26-Aug-2009 15:23:33] <jb> but since UCD (Linux) doesn't provide the hard disk information, it shouldn't be displayed at all on linux machines (only the Dell info).. right?
[26-Aug-2009 15:24:56] <bigegor> UCD provide hard disk information
[26-Aug-2009 15:25:14] <jb> hm...
[26-Aug-2009 15:25:50] <bigegor> UCD is part of net-snmp
[26-Aug-2009 15:25:54] <jb> right
[26-Aug-2009 15:26:38] <jb> so is it going to be possible to get both your Dell HD info
[26-Aug-2009 15:26:45] <jb> and the informant? (for IO)
[26-Aug-2009 15:27:24] <rocket> jb it might mess up reports but I can see that it would be a possible
[26-Aug-2009 15:27:52] <jb> hmm.. or maybe rename one?
[26-Aug-2009 15:27:57] <jb> so they both aren't named "Hard Disks"
[26-Aug-2009 15:29:33] <bigegor> no. informant used 'harddisks' relation and DellHardDisk used it to. by modeling last plugin owerwrite data collected by other one.
[26-Aug-2009 15:30:07] <jb> hrm
[26-Aug-2009 15:30:07] <jb> k
[26-Aug-2009 15:31:21] <bigegor> i have the same problem by HP modeling. Have a look on HPMon zenpack.
[26-Aug-2009 15:31:48] <rocket> bigegor: look at my aix zenpack it shows how to make new device components
[26-Aug-2009 15:31:52] <jb> could we rename DellHardDisks relation to "dellharddisk" or something?
[26-Aug-2009 15:32:10] <rocket> bigegor: eg the printqueue is a new device component and would be like jb's DellHardDisks
[26-Aug-2009 15:32:11] <jb> cause i think status of the dell harddisks is important too
[26-Aug-2009 15:32:56] <bigegor> rocket: i know this way to. You use zPythonClass
[26-Aug-2009 15:33:43] <rocket> bigegor: yes that is the case
[26-Aug-2009 15:36:22] <bigegor> rocket: using of zPythonClass had some limitation
[26-Aug-2009 15:38:47] <rocket> there is no other way to create a new device component that I know of then...
[26-Aug-2009 15:43:09] <bigegor> rocket: have a look on deviceAdvDetail and RDBMS zenpacks __init__.py
[26-Aug-2009 15:46:14] <rmatte> what's the function to pull availability for a device?
[26-Aug-2009 15:48:41] <rmatte> anyone?
[26-Aug-2009 15:52:17] <rmatte> I need to duplicate what the availability report is doing
[26-Aug-2009 16:04:08] <rmatte> ummm, wow, the memory utilization report must have changed between 2.4.1 and 2.4.3, I copied my custom report over and it doesn't work at all
[26-Aug-2009 16:04:42] <rmatte> actually crap, I didn't copy the report file itself over
[26-Aug-2009 16:12:34] <bigegor> jb: can you try this command? wbemcli ein -noverify 'https://username:password@your_dell_server:5989/root/cimv2:CIM_DiskDrive'
[26-Aug-2009 16:14:41] <jb> one sec
[26-Aug-2009 16:15:29] <jb> would this be my domain username/password?
[26-Aug-2009 16:15:57] <bigegor> yes
[26-Aug-2009 16:16:21] <jb> i have nothing on 5989
[26-Aug-2009 16:17:01] <bigegor> no wbem
[26-Aug-2009 16:17:14] <jb> yeah
[26-Aug-2009 16:17:18] <bigegor> but from IBM?
[26-Aug-2009 16:17:23] <jb> windows doesn't have a WBEM provider by default
[26-Aug-2009 16:17:32] <jb> i didnt think..
[26-Aug-2009 16:18:17] <jb> nothing on my IBM boxes either
[26-Aug-2009 16:18:18] <bigegor> but management agents rom HP and IBM
[26-Aug-2009 16:19:12] <jb> ah
[26-Aug-2009 16:19:13] <bigegor> have you IBM Director installed?
[26-Aug-2009 16:19:22] <jb> probably not on these
[26-Aug-2009 16:19:31] <jb> i uninstalled director on my dev boxes for the time being
[26-Aug-2009 16:42:53] <rmatte> how would I go about printing an interface name?
[26-Aug-2009 16:43:45] <rmatte> if I print "i" in d.os.interfaces
[26-Aug-2009 16:44:24] <rmatte> it prints out like: <IpInterface at Vlan50>
[26-Aug-2009 16:44:32] <rmatte> I need it to output normally, just the name of the interface
[26-Aug-2009 16:44:44] <rmatte> oh wait
[26-Aug-2009 16:45:07] <rmatte> got it
[26-Aug-2009 16:45:10] <rmatte> just came to me, i.id
[26-Aug-2009 16:51:17] <mrayzenoss> rmatte: http://blog.zenoss.com/2009/08/24/tip-of-the-month-zendmd-and-device-interfaces-in-zenoss/
[26-Aug-2009 17:02:41] <rmatte> mrayzenoss: nice
[26-Aug-2009 17:02:46] <rmatte> I'm outta here, talk to you tomorrow
[26-Aug-2009 17:05:24] ideopathic_ is now known as ideopathic
[26-Aug-2009 18:01:02] kisielk_ is now known as kisielk
[26-Aug-2009 18:05:38] <cgibbons> hurm
[27-Aug-2009 00:00:46] [disconnected at Thu Aug 27 00:00:46 2009]
[27-Aug-2009 00:00:46] [connected at Thu Aug 27 00:00:46 2009]
[27-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[27-Aug-2009 00:53:15] <ironpaw> hi
[27-Aug-2009 00:53:20] <ironpaw> anyone around?
[27-Aug-2009 00:55:34] <Appiah> 48 in the channel
[27-Aug-2009 00:55:37] <Appiah> others*
[27-Aug-2009 02:00:39] <ironpaw> oh
[27-Aug-2009 02:00:53] <ironpaw> ive just been here an there have been a heap of other ppl here but none respond
[27-Aug-2009 02:00:58] <ironpaw> much like i didnt just then
[27-Aug-2009 02:03:02] <ironpaw> i appear to have issues adding data sources to devices in 2.4.3
[27-Aug-2009 02:03:07] <ironpaw> is this a known issue?
[27-Aug-2009 06:55:04] <tripitaka> Hi, I'm seeing a lot of false positives, particularly for http service. The load on the zenoss server is not very high, and tcpdump shows what look like good http connections from the monitoring service. What are the likely causes of the false positives?
[27-Aug-2009 08:37:03] sgsax is now known as newt
[27-Aug-2009 08:37:42] newt is now known as sgsax
[27-Aug-2009 08:40:59] <sergeymasushko> hi 2all. Can anyone explain, what does '2009-08-27 02:50:22 WARNING zen.ZenPing: Configuration still loading. Started at Thu Aug 20 11:04:37 2009' mean in the 'zenping.log' ?
[27-Aug-2009 08:42:36] <sergeymasushko> I've experienced several fake UP/DOWN alerts from zenoss and I tried to find out what was the problem...
[27-Aug-2009 08:47:53] <rmatte> eugh, after our upgrade to 2.4.3 a lot of the CPU performance graphs are not populating data at all
[27-Aug-2009 08:47:59] <rmatte> and some are completely missing RRD files
[27-Aug-2009 08:48:59] <jb> morning
[27-Aug-2009 08:49:06] <rmatte> good morning
[27-Aug-2009 08:49:12] <rmatte> well, bad morning for me
[27-Aug-2009 08:49:12] <rmatte> lol
[27-Aug-2009 08:49:41] <sergeymasushko> morning guys
[27-Aug-2009 08:53:23] <jb> hehe
[27-Aug-2009 08:53:34] <jb> yeah, ive got to get these graphs fixed too
[27-Aug-2009 08:54:06] <rmatte> yeah, but what's causing them to go like that in the first place
[27-Aug-2009 08:54:17] <jb> i think my problem is a bit different
[27-Aug-2009 08:54:26] <jb> or are your saying "0" too?
[27-Aug-2009 08:55:24] <rmatte> a bunch are saying 0, some are saying nan, and some have no rrd files at all
[27-Aug-2009 08:55:32] <rmatte> but most are saying 0
[27-Aug-2009 08:55:57] <rmatte> it's horrible because I need to generate reports for my boss and now I'm missing a TON of report data
[27-Aug-2009 08:56:19] <rmatte> we may have to revert back to the older versions if I can't fix this
[27-Aug-2009 08:56:59] <jb> im going to delete my graph definition all together
[27-Aug-2009 08:57:01] <jb> and try again
[27-Aug-2009 08:57:05] <jb> although thats not really an option for you
[27-Aug-2009 08:58:52] <rmatte> nope
[27-Aug-2009 08:59:07] <rmatte> these templates have been working perfectly for months
[27-Aug-2009 08:59:14] <rmatte> I doubt it has anything to do with the actual template
[27-Aug-2009 09:00:36] <jb> hrm yeah didnt work for me
[27-Aug-2009 09:01:00] <rmatte> I just don't see why some graphs are working fine and others aren't
[27-Aug-2009 09:01:04] <jb> snmpwalk -cfcsnmp1ro -v1 172.26.101.66 1.3.6.1.4.1.8072.1.3.2.3.1.1.11.112.111.115.116.102.105.120.83.101.110.116
[27-Aug-2009 09:01:05] <jb> against fc-iftpcln01.corp.follett.com
[27-Aug-2009 09:01:05] <jb> NET-SNMP-EXTEND-MIB::nsExtendOutput1Line."postfixSent" = STRING: 228131
[27-Aug-2009 09:01:19] <rmatte> mine is a command based datapoint
[27-Aug-2009 09:02:53] <rmatte> well, my boss is going to kill me unless I can fix this
[27-Aug-2009 09:03:13] <rmatte> too bad chet's not around
[27-Aug-2009 09:13:13] <rmatte> well, I'm going to try deleting RRD files on one device for starters to see if that'll fix it
[27-Aug-2009 09:17:10] <rmatte> weird, I just noticed something
[27-Aug-2009 09:17:33] <jb> renaming by graphs/datapoints seemed to do it
[27-Aug-2009 09:17:59] <rmatte> omg
[27-Aug-2009 09:18:12] <rmatte> I included an old script in my snmp performance ZenPack, that's why it's all screwed
[27-Aug-2009 09:18:23] <rmatte> damnit
[27-Aug-2009 09:20:02] <rmatte> When I run the command against systems with 1 CPU it outputs like CPU1=0 Total=0
[27-Aug-2009 09:20:10] <rmatte> but it should be OK|CPU1=0 Total=0
[27-Aug-2009 09:20:15] <rmatte> I need to fix that
[27-Aug-2009 09:21:54] <rmatte> I'll need to release an updated version to Matt too, since it's posted on the ZenPack page
[27-Aug-2009 09:22:00] <rmatte> gah
[27-Aug-2009 09:25:43] <rmatte> hey Matt
[27-Aug-2009 09:26:03] <rmatte> my new Zenpack has an old version of the cpu collection script (I hadn't noticed)
[27-Aug-2009 09:26:05] <mrayzenoss> bad news
[27-Aug-2009 09:26:07] <mrayzenoss> http://dev.zenoss.org/trac/ticket/5466
[27-Aug-2009 09:26:21] <mrayzenoss> we're pushing 2.4.4 out
[27-Aug-2009 09:26:50] <jb> eek
[27-Aug-2009 09:26:56] <rmatte> pushing it out today?
[27-Aug-2009 09:27:25] <rmatte> anyways, I'm going to update my windows performance zenpack and send the updated version to you
[27-Aug-2009 09:27:28] <mrayzenoss> we're spinning up builds right now
[27-Aug-2009 09:27:36] <mrayzenoss> thanks
[27-Aug-2009 09:27:42] <rmatte> the old version of the script causes no cpu collection for anything under 1 Processor
[27-Aug-2009 09:27:44] <rmatte> lol
[27-Aug-2009 09:29:12] <jb> how can I reset values for all datasources?
[27-Aug-2009 09:29:16] <jb> removing the rrd did nothing
[27-Aug-2009 09:29:42] <rmatte> ok, I'm trying to understand this trac...
[27-Aug-2009 09:30:43] <rmatte> does that only affect zenoss email alerts?
[27-Aug-2009 09:30:49] <ckrough> ouch
[27-Aug-2009 09:31:11] <rmatte> "No CLEARs are sent out to agents, so the think devices that are up are down etc."
[27-Aug-2009 09:31:28] <rmatte> no clears are sent to collectors is what he means?
[27-Aug-2009 09:33:28] <rmatte> from looking at the event history on one of my boxes, clear events are working fine in the event console
[27-Aug-2009 09:33:34] <rmatte> so does that bug only affect notification?
[27-Aug-2009 09:34:08] <ckrough> clears dont get sent to collectors
[27-Aug-2009 09:35:41] <mrayzenoss> we're pulling 2.4.3 and 2.4.4 should be out tomorrow, possibly late today
[27-Aug-2009 09:36:03] <jb> ARGH
[27-Aug-2009 09:36:22] <ckrough> <nelson> ha ha </nelson>
[27-Aug-2009 09:36:30] <jb> now the graph is back to all 0's
[27-Aug-2009 09:36:31] <jb> :/
[27-Aug-2009 09:38:27] <rmatte> there, fixed the script
[27-Aug-2009 09:41:51] <dollarbang> Getting ready to install 2.4.3 of Zenoss, I'm getting this error backing up the etc directory "cp: cannot open `etc/zeo.zdsock' for reading: No such device or address" the file has these attributes "0 srwx------ 1 zenoss zenoss 0 Aug 13 11:03 zeo.zdsock".
[27-Aug-2009 09:42:26] <jb> dont install 2.4.3
[27-Aug-2009 09:42:41] <jb> wait for 2.4.4..
[27-Aug-2009 09:44:17] <dollarbang> jb: why, is there something wrong with 2.4.3? the error has to do with a file that was created by either the stacker during the previous installation, or Linux doesn't like the file.
[27-Aug-2009 09:44:24] <jb> scroll up..
[27-Aug-2009 09:45:09] <dollarbang> jb: oh, just saw this: mrayzenoss: we're pulling 2.4.3 and 2.4.4 should be out tomorrow, possibly late today
[27-Aug-2009 09:45:18] <jb> http://dev.zenoss.org/trac/ticket/5466
[27-Aug-2009 09:45:29] <rmatte> mrayzenoss: is a patch available yet?
[27-Aug-2009 09:45:46] <rmatte> I'd rather patch now rather than re-upgrade everything
[27-Aug-2009 09:46:09] <rmatte> I won't be able to do another full upgrade until next tuesday anyways
[27-Aug-2009 09:59:05] <rmatte> great, now I'm somehow screwed this ZenPack up even more, wtf
[27-Aug-2009 09:59:09] <rmatte> I've*
[27-Aug-2009 10:00:10] <rmatte> oh, hmmm
[27-Aug-2009 10:02:10] <rmatte> needed to restart Zenoss
[27-Aug-2009 10:06:43] <vinzpr> hello
[27-Aug-2009 10:07:19] <vinzpr> anybody here?
[27-Aug-2009 10:08:50] <jb> ?
[27-Aug-2009 10:09:02] <vinzpr> hi
[27-Aug-2009 10:10:17] <rmatte> yes, there are a few people here
[27-Aug-2009 10:10:21] <rmatte> if you have a question just ask
[27-Aug-2009 10:10:36] <jb> yes, don't wait on us to ask if you need help..
[27-Aug-2009 10:11:08] <vinzpr> ok..
[27-Aug-2009 10:12:21] <vinzpr> i upgrade from 2.4.2 to 2.4.3 when I check the port 8100 if open by doing telnet remotely it didnt response but doing it localhost it works
[27-Aug-2009 10:12:35] <rmatte> check your firewall
[27-Aug-2009 10:12:46] <vinzpr> i've already disable firewall
[27-Aug-2009 10:12:50] <jb> can you telnet from localhost?
[27-Aug-2009 10:12:55] <rmatte> and it should be 8080, unless you specifically changed it?
[27-Aug-2009 10:12:57] <vinzpr> yes it response
[27-Aug-2009 10:13:11] <jb> telnet localhost 8100 works?
[27-Aug-2009 10:13:11] <rmatte> 8100 is for something else, it's not the default web port
[27-Aug-2009 10:13:27] <vinzpr> but on the other machine doing telnet host.domain.com 8100 it didnt response
[27-Aug-2009 10:13:41] <vinzpr> is for zeodb
[27-Aug-2009 10:13:45] <jb> yeah its zeodb
[27-Aug-2009 10:14:15] <vinzpr> i tried to remove the package again then install fresh 2.4.3 its didnt work
[27-Aug-2009 10:14:19] <jb> its probably bound to localhost
[27-Aug-2009 10:14:21] <vinzpr> but on 2.4.2 it works
[27-Aug-2009 10:14:23] <jb> or you have a firewall issue
[27-Aug-2009 10:14:37] <vinzpr> i close the firewall already
[27-Aug-2009 10:14:43] <vinzpr> is open on all ports
[27-Aug-2009 10:14:51] <jb> whats the problem?
[27-Aug-2009 10:14:55] <jb> why do you need to connect to it remotely?
[27-Aug-2009 10:15:10] <vinzpr> im doing a distributed collector
[27-Aug-2009 10:15:29] <jb> you can try changing "localhost:8100" in zope.conf to IP:8100
[27-Aug-2009 10:15:30] <vinzpr> and changing the config on zope.conf and pointing to your main server
[27-Aug-2009 10:15:34] <jb> not sure
[27-Aug-2009 10:15:59] <vinzpr> in 2.4.2 it works well
[27-Aug-2009 10:16:11] <vinzpr> especially on dong zendmd on remote
[27-Aug-2009 10:16:39] <vinzpr> it response but now it didnt
[27-Aug-2009 10:17:19] <vinzpr> i have to upgrade into latest version because of the fix on keyerror
[27-Aug-2009 10:17:28] <jb> 2.4.3 has other issues
[27-Aug-2009 10:17:34] <jb> its being pulled..
[27-Aug-2009 10:17:42] <jb> 2.4.4 will be out late today or tommorow
[27-Aug-2009 10:17:58] <vinzpr> ok
[27-Aug-2009 10:18:05] <vinzpr> thats nice to hear
[27-Aug-2009 10:18:08] <mrayzenoss> as soon as the patch is available, I'll put it here and the forums
[27-Aug-2009 10:18:17] <vinzpr> hi matt
[27-Aug-2009 10:18:26] <vinzpr> i emailed you awhile ago
[27-Aug-2009 10:19:13] <vinzpr> i hope it works on the latest version
[27-Aug-2009 10:21:04] <vinzpr> so the problem of port 8100 will be fix on 2.4.4?
[27-Aug-2009 10:21:42] <mrayzenoss> the only fix going into 2.4.4 is the patch to http://dev.zenoss.org/trac/ticket/5466
[27-Aug-2009 10:21:48] <rmatte> mrayzenoss: awesome, thanks. I'll have that updated ZenPack to you in moments
[27-Aug-2009 10:21:58] <jb> this graph is really starting to annoy me
[27-Aug-2009 10:22:04] <jb> it makes no sense.
[27-Aug-2009 10:22:24] <rmatte> lol
[27-Aug-2009 10:22:37] <rmatte> neither does life
[27-Aug-2009 10:22:57] <rmatte> ...and we have cpu graphs!
[27-Aug-2009 10:23:00] * rmatte rejoices
[27-Aug-2009 10:23:13] <rmatte> although I think some other graphs are still malfunctioning, but at least this knocks down the majority
[27-Aug-2009 10:24:48] <chickenandbeans> Hi, anyone here good with zendmd? Need to know how to use the manage_addDataPointGraphPoints method
[27-Aug-2009 10:24:55] <chickenandbeans> anyone?
[27-Aug-2009 10:26:23] <vinzpr> hi matt...i saw the ticket..is the latest patch includes the problem on port 8100?
[27-Aug-2009 10:26:45] <mrayzenoss> I don't think so
[27-Aug-2009 10:26:49] <mrayzenoss> is there a ticket for that?
[27-Aug-2009 10:27:11] <rmatte> no, he's just noticing that he can't hit port 8100 remotely on one of his boxes
[27-Aug-2009 10:27:16] <vinzpr> no there is not ticket for that
[27-Aug-2009 10:27:19] <rmatte> but it's probably bound to localhost by default, no?
[27-Aug-2009 10:27:25] <jb> yes.. zope.conf
[27-Aug-2009 10:28:11] <vinzpr> zope.conf i point to the other server
[27-Aug-2009 10:29:49] <vinzpr> my main server is main.domain.com and on the other server i change the config of zope.conf localhost:8100 into main.domain.com:8100
[27-Aug-2009 10:29:54] <vinzpr> and it didnt work
[27-Aug-2009 10:30:44] <vinzpr> trying on my other server zendmd, it didnt response can connect to main.domain.com
[27-Aug-2009 10:32:04] <chickenandbeans> so can anyone here help?
[27-Aug-2009 10:38:07] <vinzpr> is this issue can be fixed on 2.4.4?
[27-Aug-2009 10:40:30] <mrayzenoss> vinzpr: no. There's no ticket and 2.4.4 is going to have 1 ticket
[27-Aug-2009 10:40:42] <mrayzenoss> vinzpr: so help me understand
[27-Aug-2009 10:41:22] <mrayzenoss> vinzpr: You're running a distributed collector and it's not respecting the zope.conf?
[27-Aug-2009 10:42:52] <vinzpr> im running distributed collector then i change localhost:8100 to main.domain.com:8100 of zope.conf to the other machine
[27-Aug-2009 10:43:43] <vinzpr> so my zeodb on the second machine will be the main.domain.com
[27-Aug-2009 10:45:20] <vinzpr> but on 2.4.2 it running well
[27-Aug-2009 10:45:29] <vinzpr> i have no issue on it
[27-Aug-2009 10:46:04] <vinzpr> i can even telnet main.domain.com 8100
[27-Aug-2009 10:46:09] <vinzpr> and it does response
[27-Aug-2009 10:48:15] <mrayzenoss> and all the conf files have been updated and you've restarted everything?
[27-Aug-2009 10:48:24] <vinzpr> yes
[27-Aug-2009 10:48:57] <vinzpr> except for zeodb
[27-Aug-2009 10:53:15] <vinzpr> any suggestion?
[27-Aug-2009 10:53:21] <vinzpr> how can i fix this?
[27-Aug-2009 10:55:39] <dberger> is there any reason why an event submitted via xmlrpc wouldn't appear in the event console?
[27-Aug-2009 10:59:53] <mrayzenoss> dberger: is the severity too low?
[27-Aug-2009 11:00:45] <venturaville> anyone here familiar with pywbem?
[27-Aug-2009 11:01:05] <mrayzenoss> venturaville: Egor and Kells
[27-Aug-2009 11:01:11] <vinzpr> hi matt..do you understand now my problem on port 8100?
[27-Aug-2009 11:01:13] <dberger> mrayzenoss, no
[27-Aug-2009 11:01:20] <dberger> severity is critical
[27-Aug-2009 11:01:24] <mrayzenoss> dberger: version?
[27-Aug-2009 11:01:27] <venturaville> mrayzenoss: thanks ... I'll keep an eye out for them
[27-Aug-2009 11:03:43] <dberger> mrayzenoss, remind me how to check that
[27-Aug-2009 11:08:06] <vinzpr> mrayzenoss, so when i excute zendmd on my remote server it response me like this WARNING:ZEO.zrpc:(11602) CW: error connecting to ('main.domain.com', 8100): ECONNREFUSED
[27-Aug-2009 11:09:04] <mrayzenoss> dberger: Settings->Versions
[27-Aug-2009 11:09:20] <dberger> mrayzenoss, 2.4.1
[27-Aug-2009 11:09:51] <dberger> mrayzenoss, sorry...i've had several versions installed and i got a little confused
[27-Aug-2009 11:10:28] <mrayzenoss> vinzpr: OK, I think the zeo.conf got changed
[27-Aug-2009 11:11:41] <vinzpr> mrayzenoss, so this is the problem on 2.4.3?
[27-Aug-2009 11:11:43] <mrayzenoss> I think this ticket may be the one that affected you
[27-Aug-2009 11:11:44] <mrayzenoss> http://dev.zenoss.org/trac/ticket/5255
[27-Aug-2009 11:14:40] <vinzpr> im sorry but on 2.4.2 its working fine
[27-Aug-2009 11:14:52] <vinzpr> and the ticket specified the target fix is on 2.4.3
[27-Aug-2009 11:15:09] <mrayzenoss> yeah, it was patched in 2.4.3
[27-Aug-2009 11:15:24] <mrayzenoss> so when you upgraded to 2.4.3 it broke your configuration
[27-Aug-2009 11:15:45] <mrayzenoss> distributed collectors aren't supported in Core, so this change wasn't tested in Core
[27-Aug-2009 11:19:17] <vinzpr> so do i have to manually execute the zeopack.py?
[27-Aug-2009 11:19:46] <daMaestro> i'm having issues mapping events to Ignore
[27-Aug-2009 11:20:01] <daMaestro> it doesn't seem like the regex is being applied to messages
[27-Aug-2009 11:21:00] <mrayzenoss> vinzpr: I'm sorry, I really don't know much about distributed collectors
[27-Aug-2009 11:21:26] <mrayzenoss> vinzpr: I don't mess with them myself, since they're in Enterprise
[27-Aug-2009 11:21:37] <mrayzenoss> anyone else using distributed collectors with Core? rmatte?
[27-Aug-2009 11:21:39] <daMaestro> Accepted publickey for username|pam_unix\(sshd:session\): session closed for user username
[27-Aug-2009 11:21:58] <daMaestro> that is my regex, ssh is the event key.. and it's part of /Ignore .. but it's not matching any events
[27-Aug-2009 11:32:06] <vinzpr> hi mrayzenoss, i got it and i was able to telnet on it
[27-Aug-2009 11:32:36] <mrayzenoss> I responded to your forum thread
[27-Aug-2009 11:32:37] <vinzpr> your right the problem is on the config file
[27-Aug-2009 11:32:47] <mrayzenoss> you'll probably want to update the community distributed collector docs
[27-Aug-2009 11:32:59] <jb> it was zope.conf?
[27-Aug-2009 11:33:04] <jb> binding to localhost?
[27-Aug-2009 11:33:15] <vinzpr> and then i notice in zeo.conf has a address localhost:8100
[27-Aug-2009 11:33:22] <vinzpr> it should be address 8100
[27-Aug-2009 11:33:25] <vinzpr> thats its
[27-Aug-2009 11:33:26] <jb> isnt that what I told you an hour ago?
[27-Aug-2009 11:33:36] <vinzpr> sure mray i will update the forum
[27-Aug-2009 11:33:40] <mrayzenoss> thanks
[27-Aug-2009 11:33:42] <vinzpr> and i will also contribute on it
[27-Aug-2009 11:34:16] <vinzpr> i will udpate the community distribute collectors docs soon with failover of zenoss
[27-Aug-2009 11:34:48] <vinzpr> thanks also jb
[27-Aug-2009 11:35:02] <jb> np
[27-Aug-2009 11:35:07] <jb> thanks for updating the docs
[27-Aug-2009 11:35:09] <vinzpr> maybe i will wait for the release of 2.4.4 then upgrade it again
[27-Aug-2009 11:35:12] <jb> many people have been looking for docs
[27-Aug-2009 11:36:05] <vinzpr> ok..thank you guys again let me do some testing now..
[27-Aug-2009 11:44:40] <rmatte> nah, we tried distributed collectors with core but it was too much of a hassle
[27-Aug-2009 11:44:49] <rmatte> so we're just using separate instances
[27-Aug-2009 12:04:22] <rmatte> hmmmm, events don't appear to automatically age in 2.4.3 either
[27-Aug-2009 12:26:18] <rmatte> mrayzenoss: events don't seem to be automatically aging at all in 2.4.3 either
[27-Aug-2009 12:26:31] <rmatte> not sure if that's it's own issue or if it's related to the clear events issue
[27-Aug-2009 12:26:32] <mrayzenoss> I'll pass that on
[27-Aug-2009 12:26:40] <mrayzenoss> probably related, I'll check with devs
[27-Aug-2009 12:27:04] <rmatte> I had events (that came in as traps) from yesterday still sitting there when I have it set to age out after an hour
[27-Aug-2009 12:35:04] <rmatte> mrayzenoss: you might want to change the channel topic
[27-Aug-2009 12:35:05] <rmatte>
[27-Aug-2009 12:38:30] <rmatte> mrayzenoss: the trac ticket has been set to fixed, resolution is patch 14744, wondering if we can go ahead with patching?
[27-Aug-2009 12:38:59] <mrayzenoss> QA hasn't finished verifying it
[27-Aug-2009 12:39:03] <rmatte> k
[27-Aug-2009 12:39:05] <mrayzenoss> but they're working on it
[27-Aug-2009 12:39:20] <mrayzenoss> you can try it if you like
[27-Aug-2009 12:39:27] <rmatte> no thanks, I'll wait
[27-Aug-2009 12:39:55] <chudler> take you're time, get it right :-) nobody wants to see a 2.4.5 tomorrow
[27-Aug-2009 12:39:57] <mrayzenoss> I'll be sending out announcements and whatnot once the patch is tested
[27-Aug-2009 12:39:57] <rmatte> I'm just anxious because we've probably had zero monitoring on certain devices for over 48 hours
[27-Aug-2009 12:40:19] <mrayzenoss> ugh
[27-Aug-2009 12:40:28] <rmatte> if something bad happens our clients will whoop us lol
[27-Aug-2009 12:40:33] <mrayzenoss> sounds like another test case we need
[27-Aug-2009 12:40:52] <rmatte> yeh
[27-Aug-2009 12:40:57] <mrayzenoss> feel free to jump in: http://blog.zenoss.com/2009/07/10/help-zenoss-expand-our-testing/
[27-Aug-2009 12:41:04] <rmatte> aging events would be a nice test to do too before release
[27-Aug-2009 12:41:27] <mrayzenoss> adding that to the ticket
[27-Aug-2009 12:45:21] <rmatte> k
[27-Aug-2009 12:46:09] <rmatte> I'm going to add reports to that, since obviously not much attention is paid to them
[27-Aug-2009 12:46:26] <rmatte> I guess it's not considered a "Critical" component
[27-Aug-2009 12:47:59] <rmatte> so did they say the aging is related?
[27-Aug-2009 12:48:06] <rmatte> or are they not sure yet?
[27-Aug-2009 12:48:33] <mrayzenoss> it should be related
[27-Aug-2009 12:48:43] <mrayzenoss> they're going to test it with a 1-hour age-out
[27-Aug-2009 12:48:44] <rmatte> k
[27-Aug-2009 12:49:03] <rmatte> I need to log a feature request to have the age-out specified in seconds
[27-Aug-2009 12:49:15] <rmatte> because we'd rather do it for a half hour, but "0.5" doesn't work
[27-Aug-2009 12:52:24] <mrayzenoss> yeah, that sounds valid
[27-Aug-2009 12:52:55] <mrayzenoss> throw in a patch and it might make it into King Crab
[27-Aug-2009 12:53:00] <jb> so is there any other way to troubleshoot graphs reporting 0?
[27-Aug-2009 12:53:53] <mrayzenoss> if you turn up the logging, you can see the values getting recorded
[27-Aug-2009 12:54:08] <jb> logging on?
[27-Aug-2009 12:54:12] <jb> zenperfsnmp?
[27-Aug-2009 12:54:25] <mrayzenoss> yeah
[27-Aug-2009 12:55:06] <jb> ok set to debug
[27-Aug-2009 13:04:39] <rmatte> timeout?
[27-Aug-2009 13:05:34] <mrayzenoss1> changed rooms
[27-Aug-2009 13:05:37] <mrayzenoss1> in a meeting
[27-Aug-2009 13:05:39] <rmatte> ah
[27-Aug-2009 13:07:58] <rmatte> the other thing is, after the patch is applied and Zenoss is restarted, the collector should start fresh and not think anything is down I assume?
[27-Aug-2009 13:08:11] <rmatte> or will it still think devices are down until a fake clear event is generated?
[27-Aug-2009 13:08:11] <dberger> so, I can see events in device list or in the device issues portlet, but not in the event console
[27-Aug-2009 13:08:22] <dberger> i'm submitting these events via xmlrpc
[27-Aug-2009 13:08:34] <mrayzenoss1> dberger: what maintenance state are they in?
[27-Aug-2009 13:08:40] <dberger> new
[27-Aug-2009 13:08:51] <rmatte> "new" isn't a maintenance state?
[27-Aug-2009 13:09:19] <rmatte> maintenance states are Production, Pre-Productions, Test, Maintenance, and Decommissioned
[27-Aug-2009 13:09:45] <mrayzenoss1> probably Production then
[27-Aug-2009 13:10:18] <mrayzenoss1> just that some maintenance states don't show on the dashboard, but you're seeing the opposite of that
[27-Aug-2009 13:10:23] <mrayzenoss1> on dash, no event console?
[27-Aug-2009 13:10:24] <dberger> I see prodState = Production(1000)
[27-Aug-2009 13:10:32] <rmatte> yeh, so production
[27-Aug-2009 13:11:15] <rmatte> I think he's seeing the count of the events in the device list, but not actually seeing the event in the event console
[27-Aug-2009 13:11:24] <dberger> right
[27-Aug-2009 13:11:33] <rmatte> hmmm
[27-Aug-2009 13:11:44] <rmatte> and you're using xmlrpc code to generate the events?
[27-Aug-2009 13:11:48] <dberger> and if i click on the devie in the device list, it gives me an event console specific to that device, and i see it there
[27-Aug-2009 13:11:53] <dberger> yes
[27-Aug-2009 13:11:58] <rmatte> can you pastebin the code?
[27-Aug-2009 13:12:56] <rmatte> (other non-xmlrpc generated events are displaying in the event console I assume?)
[27-Aug-2009 13:13:13] <dberger> yes
[27-Aug-2009 13:13:30] <rmatte> k, pastebin the code and we'll try to figure out what you're missing
[27-Aug-2009 13:16:12] <rmatte> mloven always reminds me of McLoving
[27-Aug-2009 13:16:16] <rmatte> McLovin*
[27-Aug-2009 13:16:17] <rmatte> rather
[27-Aug-2009 13:17:29] <dberger> http://pastebin.com/mc6dcd09
[27-Aug-2009 13:20:54] <rmatte> well, first thing I notice is that you're setting an event summary but not an event message
[27-Aug-2009 13:23:02] <rmatte> doesn't look like you're specifying an event class either
[27-Aug-2009 13:23:19] <rmatte> ah nevermind
[27-Aug-2009 13:23:22] <rmatte> testOrganizer
[27-Aug-2009 13:23:38] <dberger> oh right, i should include that
[27-Aug-2009 13:23:48] <rmatte> do you have /testOrganizer as an event class in Events?
[27-Aug-2009 13:24:06] <dberger> yes
[27-Aug-2009 13:24:14] <rmatte> k
[27-Aug-2009 13:24:45] <rmatte> there's no agent specified, not sure if that's going to cause issues or not
[27-Aug-2009 13:24:55] <dberger> agent?
[27-Aug-2009 13:25:09] <rmatte> example: agent zensyslog
[27-Aug-2009 13:26:22] <rmatte> but it may do that on it's own
[27-Aug-2009 13:26:40] <dberger> agent gets left blank
[27-Aug-2009 13:27:38] <rmatte> k
[27-Aug-2009 13:27:58] <rmatte> other than that, I can't really see any other issues with i
[27-Aug-2009 13:28:00] <rmatte> it*
[27-Aug-2009 13:28:15] <rmatte> what severity does it come in as?
[27-Aug-2009 13:28:20] <dberger> 5
[27-Aug-2009 13:28:29] <rmatte> where are you setting that?
[27-Aug-2009 13:28:35] <rmatte> oh nevermind
[27-Aug-2009 13:28:38] <rmatte> I see it now
[27-Aug-2009 13:29:26] <rmatte> yeh, no idea, it looks good from what I can see other than message missing
[27-Aug-2009 13:30:05] <dberger> which gets copied from the summary field before going into the db
[27-Aug-2009 13:32:17] <dberger> could it have something to do with the component?
[27-Aug-2009 13:32:36] <dberger> that's just something made up that hasn't been configured with zenoss
[27-Aug-2009 13:32:40] <rmatte> possibly, but I doubt it
[27-Aug-2009 13:32:46] <rmatte> you could try leaving component blank
[27-Aug-2009 13:32:59] <rmatte> there is no component configuration in Zenoss
[27-Aug-2009 13:33:04] <rmatte> component can be anything
[27-Aug-2009 13:33:26] <rmatte> it's just a place to list the affected component
[27-Aug-2009 13:33:28] <dberger> that's what i thought
[27-Aug-2009 13:33:28] <rmatte> so an interface name
[27-Aug-2009 13:33:31] <rmatte> or PRINTER
[27-Aug-2009 13:33:33] <rmatte> or whatever
[27-Aug-2009 13:34:47] <dberger> BINGO!
[27-Aug-2009 13:34:53] <rmatte> ?
[27-Aug-2009 13:35:04] <dberger> 'eventClass':'/testOrganizer'
[27-Aug-2009 13:35:06] <dberger> not
[27-Aug-2009 13:35:09] <rmatte> oh
[27-Aug-2009 13:35:12] <rmatte> you were missing the slash
[27-Aug-2009 13:35:14] <dberger> 'eventClass':'testOrganizer'
[27-Aug-2009 13:35:18] <rmatte> yeh, that's would definitely cause issues
[27-Aug-2009 13:35:24] <rmatte> lol
[27-Aug-2009 13:35:52] <rmatte> I just didn't know whether the xmlrpc code needed it or not
[27-Aug-2009 13:35:54] <dberger> funny that that ends up being inconsistent
[27-Aug-2009 13:35:56] <rmatte> but apparently it does
[27-Aug-2009 13:58:26] <rmatte> mrayzenoss1:
[27-Aug-2009 13:58:27] <rmatte> Changed 7 minutes ago by bbibeault ¶
[27-Aug-2009 13:58:27] <rmatte> * patch_state changed from Accepted to Patched
[27-Aug-2009 13:58:58] <rmatte> does that mean it passed QA?
[27-Aug-2009 14:00:45] <mrayzenoss1> I'm bugging them about it right now, making sure they tested auto-aging
[27-Aug-2009 14:01:13] <rmatte> k cool
[27-Aug-2009 14:01:18] <mrayzenoss1> but clears are definitely working again
[27-Aug-2009 14:01:23] mrayzenoss1 is now known as mrayzenoss
[27-Aug-2009 14:01:24] <rmatte> good stuff
[27-Aug-2009 14:01:41] <rmatte> the aging is pretty important to our setup though at the moment, so that's definitely a must for me
[27-Aug-2009 14:01:44] <mrayzenoss> and the bits are being uploaded to SourceForge
[27-Aug-2009 14:01:56] <rmatte> cool
[27-Aug-2009 14:02:04] <rmatte> I can just do this as a zenpatch though right?
[27-Aug-2009 14:02:22] <rmatte> I don't have time to do a full blown upgrade of each system until next Tuesday night
[27-Aug-2009 14:02:39] <rmatte> well, I don't have an official change window until then
[27-Aug-2009 14:02:46] <rmatte> but I can do emergency patching
[27-Aug-2009 14:04:10] <mrayzenoss> yeah 'zenpatch 14744' and 'zenactions restart'
[27-Aug-2009 14:04:27] <rmatte> nice, don't even have to restart all of zenoss
[27-Aug-2009 14:04:39] <rmatte> once I hear about the aging then I'll apply it
[27-Aug-2009 14:04:45] <mrayzenoss> yeah, it's zenactions blowing up and dying
[27-Aug-2009 14:04:53] <mrayzenoss> which sucks in many, many ways
[27-Aug-2009 14:04:54] <rmatte> lovely
[27-Aug-2009 14:05:45] <rmatte> actually, if the patch is officially final I'll patch
[27-Aug-2009 14:05:53] <rmatte> I assume it is since you said they were uploading the bits
[27-Aug-2009 14:05:56] <rmatte> ?
[27-Aug-2009 14:07:19] <rmatte> yeh, zenactions is responsible for aging
[27-Aug-2009 14:07:37] <rmatte> so I'll bet that's what was causing it
[27-Aug-2009 14:07:45] <mrayzenoss> yeah, we expect it works
[27-Aug-2009 14:07:46] <rmatte> so is it safe to patch or should I hold off?
[27-Aug-2009 14:08:01] <mrayzenoss> I think it's safe
[27-Aug-2009 14:08:15] <rmatte> k, I be patchin' then
[27-Aug-2009 14:09:13] <rmatte> I don't need to stop zenactions before patching do I?
[27-Aug-2009 14:09:18] <mrayzenoss> nope
[27-Aug-2009 14:09:20] <rmatte> k
[27-Aug-2009 14:09:36] <mrayzenoss> is zenactions still up? It's just in zombie mode if it is
[27-Aug-2009 14:09:51] <mrayzenoss> that stack trace just blows it up
[27-Aug-2009 14:10:46] <JKilgore> I'm trying to set the time on the zenoss appliance, but I can't get it to stick - anyone here that might be able to offer some help?
[27-Aug-2009 14:17:18] <rmatte> there, all our prod stuff is patched
[27-Aug-2009 14:17:21] <djack> hello everybody
[27-Aug-2009 14:17:31] <rmatte> I think it's still up yeh
[27-Aug-2009 14:17:38] <rmatte> let me check on the last box I have to patch
[27-Aug-2009 14:17:54] <djack> can somebody give me a hand ?
[27-Aug-2009 14:18:10] <djack> I have some problems with creation of rrd graphs ... and i don't know why
[27-Aug-2009 14:19:14] <rmatte> mrayzenoss: should I restart all of Zenoss, on each box, or will it come to it's senses on it's own?
[27-Aug-2009 14:19:17] <venturaville> jb: you awake?
[27-Aug-2009 14:19:24] <rmatte> like, if devices have already gone down during the issue
[27-Aug-2009 14:19:42] <rmatte> won't I need a full restart so that the collector finally realizes what's going on?
[27-Aug-2009 14:19:43] <mrayzenoss> rmatte: I think just zenactions
[27-Aug-2009 14:19:50] <rmatte> k
[27-Aug-2009 14:19:56] <djack> nobody?
[27-Aug-2009 14:20:10] <rmatte> djack: state the issue
[27-Aug-2009 14:21:06] <rmatte> mrayzenoss: yeh, it is running, so probably in zombie mode
[27-Aug-2009 14:21:30] <rmatte> probably why they didn't notice it during the QA for 2.3.3
[27-Aug-2009 14:21:32] <rmatte> erm
[27-Aug-2009 14:21:33] <rmatte> 2.4.3
[27-Aug-2009 14:21:48] <djack> well i added 6 server that are behind a firewall .... from the server command line i can ping them and do snmpwalks so i know the firewall does not blcok the snmp and icmp packages. however the perf data is not updated on the zenoss server ?
[27-Aug-2009 14:22:22] <mrayzenoss> djack: and they're in the right device class?
[27-Aug-2009 14:22:38] <djack> the same as the other servers
[27-Aug-2009 14:22:52] <djack> they are in /server/linux/
[27-Aug-2009 14:22:56] <rmatte> djack: when you snmpwalk them, does it seem like there are less lines than when you walk the others?
[27-Aug-2009 14:23:00] <rmatte> like, significantly less?
[27-Aug-2009 14:23:41] <rmatte> and are they using the same snmp config as the other linux boxes?
[27-Aug-2009 14:23:42] <djack> nope not at all and when i look at the device even all the installed rpm packages are in the list but no cpu/mem/disc and other very relevant data
[27-Aug-2009 14:23:54] <djack> rmatte yep
[27-Aug-2009 14:24:37] <rmatte> if you do something like: snmpwalk -v1 -c <community> <device> hrProcessorLoad
[27-Aug-2009 14:24:39] <rmatte> what do you see?
[27-Aug-2009 14:25:14] <venturaville> RecoveredSeekOperations=None TotalReadBlocksRequested=3097094828 TotalLargeWrites=0 TotalWriteBlocksRequested=1218624050 TotalBlocksRequested=20751582 Description="Media Access Statistics for Volume vol2" TotalLargeWriteBlocksRequested=0 UnrecoverableReadOperations=None WriteOperations=237045636 TotalCacheReadChecks=988275308 RecoveredReadOperations=None UnrecoverableSeekOperations=None ElementName="Volume vol2" AverageBlocksRequested=0 TotalCacheR
[27-Aug-2009 14:25:31] <venturaville> sweet ... finally getting volume stats out of LSI arrays....
[27-Aug-2009 14:25:38] <rmatte> nice
[27-Aug-2009 14:25:42] <rmatte> wrote a custom script I see
[27-Aug-2009 14:26:42] <venturaville> yep
[27-Aug-2009 14:27:03] <djack> nothing
[27-Aug-2009 14:28:12] <rmatte> mrayzenoss: I've got a couple of trap events that came in for reboots, so I'll leave them sit and see if they age. I'll let you know.
[27-Aug-2009 14:28:30] <djack> but neither for the servers that do enumerate the cpu memory and disc snmp data
[27-Aug-2009 14:28:59] <djack> the server are HP Proliant 365 and 380
[27-Aug-2009 14:29:10] <mrayzenoss> rmatte: yeah, me and QA are waiting on our time outs
[27-Aug-2009 14:29:11] <djack> with Suse Enterprise 11
[27-Aug-2009 14:29:20] <rmatte> hehe
[27-Aug-2009 14:30:10] <djack> rmate is ther info about the zenpatches and where to download them ?
[27-Aug-2009 14:30:30] <rmatte> djack: to fix 2.4.3 you mean?
[27-Aug-2009 14:30:34] <djack> yep
[27-Aug-2009 14:30:39] <rmatte> it's simple
[27-Aug-2009 14:30:42] <djack> i installed it ...
[27-Aug-2009 14:30:46] <rmatte> login to the zenoss server, become the zenoss user
[27-Aug-2009 14:30:51] <rmatte> then issue the following 2 commands:
[27-Aug-2009 14:30:53] <rmatte> zenpatch 14744
[27-Aug-2009 14:30:59] <rmatte> zenactions restart
[27-Aug-2009 14:31:01] <rmatte> that's it
[27-Aug-2009 14:31:10] <djack> well i don't have internet access from the monitoring machine
[27-Aug-2009 14:31:16] <rmatte> oh
[27-Aug-2009 14:31:24] <rmatte> one sec
[27-Aug-2009 14:32:07] <djack> but the fix will not fix the ping issues and my snmp problems i guess :-(
[27-Aug-2009 14:32:32] <rmatte> http://dev.zenoss.org/trac/changeset/14744
[27-Aug-2009 14:32:34] <rmatte> that's the patch
[27-Aug-2009 14:32:42] <rmatte> you can download it from the links on the bottom of the page
[27-Aug-2009 14:32:46] <rmatte> either as a diff or as a zip
[27-Aug-2009 14:33:14] <rmatte> and one second, I'll get the command to patch it
[27-Aug-2009 14:33:25] <rmatte> (download it as a diff and put it somewhere)
[27-Aug-2009 14:33:56] <djack> i have to download it upload it to a server scp it to an other server and then scp it to the zenoss server
[27-Aug-2009 14:34:04] <rmatte> well do that
[27-Aug-2009 14:34:12] <rmatte> once the diff is on the zenoss server you do...
[27-Aug-2009 14:34:41] <rmatte> patch /usr/local/zenoss/zenoss/Products/ZenEvents/zenactions.py < changeset_r14744.diff
[27-Aug-2009 14:34:47] <rmatte> then zenactions restart
[27-Aug-2009 14:35:20] <djack> oki
[27-Aug-2009 14:35:46] <rmatte> the patch might, if it's thinking those devices are down it might have something to do with it
[27-Aug-2009 14:35:51] <rmatte> apply the patch, wait a bit, and see
[27-Aug-2009 14:36:03] <rmatte> if it starts collecting then good, if not, then we'll have to figure out what's causing it
[27-Aug-2009 14:40:56] <djack> how can i see if the patch is working ?
[27-Aug-2009 14:41:12] <rmatte> did you patch and restart?
[27-Aug-2009 14:41:31] <djack> yep
[27-Aug-2009 14:41:47] <djack> I'm quick :-p
[27-Aug-2009 14:42:11] <rmatte> check $ZENHOME/log/zenactions.log
[27-Aug-2009 14:42:19] <rmatte> the latest entries should be good entries
[27-Aug-2009 14:42:28] <rmatte> whereas if you scroll up you'll see a bunch of python errors
[27-Aug-2009 14:43:14] <rmatte> You should see stuff like: 2009-08-27 15:38:53 INFO zen.ZenActions: Processed 1 commands in 0.002599
[27-Aug-2009 14:43:14] <rmatte> 2009-08-27 15:38:53 INFO zen.ZenActions: processed 0 rules in 0.01 secs
[27-Aug-2009 14:43:15] <rmatte> 2009-08-27 15:39:53 INFO zen.ZenActions: Running
[27-Aug-2009 14:43:56] <rmatte> egor, if you're running 2.4.3 you'll need to patch
[27-Aug-2009 14:44:33] <bigegor> i have 2.4.1 now.
[27-Aug-2009 14:44:41] <rmatte> k, then you're fine
[27-Aug-2009 14:44:55] <bigegor> i know
[27-Aug-2009 14:45:09] <rmatte> did you apply that security patch?
[27-Aug-2009 14:45:42] <rmatte> http://dev.zenoss.org/trac/ticket/5279
[27-Aug-2009 14:46:29] <djack> no python errors anymore in the zenactions.log
[27-Aug-2009 14:47:04] <bigegor> thanks for info
[27-Aug-2009 14:47:06] <venturaville> odd question: how do you deal with counters that only update every 10 minutes... is that something that should simply be changed for the device class?
[27-Aug-2009 14:47:20] <djack> but still no rrd and the status of the devies are still down
[27-Aug-2009 14:47:35] <rmatte> djack: so the devices are showing as down?
[27-Aug-2009 14:47:41] <djack> yep
[27-Aug-2009 14:47:54] <rmatte> djack: If they are showing as ping down then it's not even going to try to monitor them or generate RRDs
[27-Aug-2009 14:47:57] <rmatte> you need to resolve that
[27-Aug-2009 14:48:08] <rmatte> are they pingable?
[27-Aug-2009 14:48:45] <djack> yep even via the command ping tool in the zenoss interface :-(
[27-Aug-2009 14:49:11] <rmatte> djack: try restarting Zenoss and see if it still shows them as down
[27-Aug-2009 14:50:06] <djack> i read something about problems with zenping on the forums
[27-Aug-2009 14:50:24] <rmatte> I'm not having any issues with it
[27-Aug-2009 14:50:38] <rmatte> and I'm running 5 production Zenoss boxes
[27-Aug-2009 14:50:39] <djack> they stay down ?
[27-Aug-2009 14:50:48] <rmatte> stay down?
[27-Aug-2009 14:50:49] <rmatte> hunh?
[27-Aug-2009 14:50:58] <djack> yep they don't come up
[27-Aug-2009 14:51:08] <djack> grmbl
[27-Aug-2009 14:51:15] <jb> can you not ping them from the zenoss box?
[27-Aug-2009 14:51:19] <jb> using "ping"
[27-Aug-2009 14:51:22] <rmatte> jb: he can apparently
[27-Aug-2009 14:51:38] <rmatte> he can ping using the command in the Zenoss UI as well
[27-Aug-2009 14:51:53] <jb> zenping running?
[27-Aug-2009 14:52:11] <djack> zenoss@snlgwk01:~> zenping status
[27-Aug-2009 14:52:11] <djack> program running; pid=26301
[27-Aug-2009 14:52:35] <jb> hrm
[27-Aug-2009 14:53:00] <djack> strange problem he ....
[27-Aug-2009 14:53:13] <rmatte> try zenping run -v10
[27-Aug-2009 14:53:18] <rmatte> see if you get any strange errors
[27-Aug-2009 14:54:19] <venturaville> jb: I am now getting volume statistics out of the DS4xxx arrays we have :-)
[27-Aug-2009 14:54:23] <rmatte> hmmmm
[27-Aug-2009 14:54:33] <jb> venturaville: nice! what are you using?
[27-Aug-2009 14:54:38] <rmatte> when I do zenping run -v10 I'm seeing a lot of: DEBUG:zen.Ping:unexpected pkt
[27-Aug-2009 14:54:46] <rmatte> anyone else seeing that?
[27-Aug-2009 14:55:05] <rmatte> or is this another wonderful 2.4.3 nuance
[27-Aug-2009 14:55:05] <djack> DEBUG:zen.Ping:pj timeout for 10.253.2.3
[27-Aug-2009 14:55:06] <djack> DEBUG:zen.Ping:pj fail for 10.253.2.3
[27-Aug-2009 14:55:06] <djack> DEBUG:zen.ZenPing:Failed 10.253.2.3 267
[27-Aug-2009 14:55:06] <djack> WARNING:zen.ZenPing:ip 10.253.2.3 is down
[27-Aug-2009 14:55:06] <djack> DEBUG:zen.ZenPing:Queueing event {'severity': 5, 'component': '', 'agent': 'zenping', 'summary': 'ip 10.25 3.2.3 is down', 'manager': 'localhost', 'eventGroup': 'Ping', 'eventState': 0, 'device': 'frs-db01', 'even tClass': '/Status/Ping', 'ipAddress': '10.253.2.3'}
[27-Aug-2009 14:55:06] <rmatte> ?
[27-Aug-2009 14:55:08] <djack> DEBUG:zen.ZenPing:Total of 1 queued events
[27-Aug-2009 14:55:10] <djack> DEBUG:zen.Ping:pj timeout for 10.253.2.4
[27-Aug-2009 14:55:12] <venturaville> jb: I'm using pywbem to fetch the data through engenio provider (CIM -> SMI provider)
[27-Aug-2009 14:55:20] <jb> hmm
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.3'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.4'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.67'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.68'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.69'
[27-Aug-2009 14:55:30] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.70'
[27-Aug-2009 14:55:32] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.71'
[27-Aug-2009 14:55:34] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:36] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:37] <jb> whoa
[27-Aug-2009 14:55:38] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:39] <jb> use pastebin
[27-Aug-2009 14:55:39] <jb> heh
[27-Aug-2009 14:55:40] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:42] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:44] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:46] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:56:09] <djack> the 129 is a firewall ....
[27-Aug-2009 14:56:19] <rmatte> I'm wondering if those debug messages are normal or not
[27-Aug-2009 14:56:26] <rmatte> I've never done a zenping run, so I don't know
[27-Aug-2009 14:56:32] <djack> lol
[27-Aug-2009 14:56:35] <rmatte> and I don't have a zenoss box that isn't 2.4.3 at this point
[27-Aug-2009 14:56:57] <rmatte> venturaville: do you have any pre-2.4.3 boxes?
[27-Aug-2009 14:57:12] <djack> rmatte i also run it 2.4.3 version
[27-Aug-2009 14:57:21] <djack> DEBUG:zen.ZenPing:Queueing event {'severity': 5, 'component': '', 'agent': 'zenping', 'summary': 'ip 10.25 3.2.3 is down', 'manager': 'localhost', 'eventGroup': 'Ping', 'eventState': 0, 'device': 'frs-db01', 'even tClass': '/Status/Ping', 'ipAddress': '10.253.2.3'}
[27-Aug-2009 14:57:43] <rmatte> djack yours are actually showing fail lines
[27-Aug-2009 14:57:58] <djack> fail lines?
[27-Aug-2009 14:58:13] <rmatte> [03:48pm] <djack> DEBUG:zen.Ping:pj timeout for 10.253.2.3
[27-Aug-2009 14:58:13] <rmatte> [03:48pm] <djack> DEBUG:zen.Ping:pj fail for 10.253.2.3
[27-Aug-2009 14:58:13] <rmatte> [03:48pm] <djack> DEBUG:zen.ZenPing:Failed 10.253.2.3 267
[27-Aug-2009 14:58:32] <djack> yep idd
[27-Aug-2009 14:58:44] <djack> howcome is the question
[27-Aug-2009 14:59:24] <rmatte> show me what you get as ping replies from the device
[27-Aug-2009 14:59:26] <djack> why does a normal ping work but not the zenping ....
[27-Aug-2009 14:59:26] <rmatte> paste the output
[27-Aug-2009 14:59:51] <djack> zenoss@snlgwk01:~> ping 10.253.2.3
[27-Aug-2009 14:59:52] <djack> PING 10.253.2.3 (10.253.2.3) 56(84) bytes of data.
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=1 ttl=64 time=0.198 ms
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=2 ttl=64 time=0.213 ms
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=3 ttl=64 time=0.235 ms
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=4 ttl=64 time=0.136 ms
[27-Aug-2009 14:59:54] <djack> ^C
[27-Aug-2009 14:59:56] <djack> --- 10.253.2.3 ping statistics ---
[27-Aug-2009 14:59:58] <djack> 4 packets transmitted, 4 received, 0% packet loss, time 3000ms
[27-Aug-2009 15:00:00] <djack> rtt min/avg/max/mdev = 0.136/0.195/0.235/0.039 ms
[27-Aug-2009 15:00:38] <djack> my point of view the ping seems good
[27-Aug-2009 15:00:51] <rmatte> those debug messages don't seem to affect my ping monitoring at all
[27-Aug-2009 15:00:55] <rmatte> so I'd imagine they are harmless
[27-Aug-2009 15:01:04] <rmatte> (the unexpected pkt ones)
[27-Aug-2009 15:01:09] <djack> hmm
[27-Aug-2009 15:02:25] <venturaville> rmatte: we have 2.4.2
[27-Aug-2009 15:02:27] <djack> i can't figure it out
[27-Aug-2009 15:02:49] <djack> i need the machine for stress testing of our internet site
[27-Aug-2009 15:02:59] <venturaville> rmatte: we have not deployed to production yet, though, as we are still trying to work out a deal for zenoss enterprise licensing
[27-Aug-2009 15:03:37] <djack> well if i can prove that zenoss works my boss his visa card is mine
[27-Aug-2009 15:04:20] <djack> but then i need to some this problem
[27-Aug-2009 15:05:39] <rmatte> venturaville: ah
[27-Aug-2009 15:06:09] <rmatte> djack: those are the only devices which you are having this issue with though right?
[27-Aug-2009 15:07:41] <djack> yep
[27-Aug-2009 15:08:18] <djack> i have 7 devices that work but they are hanging before the firewall and not behind .....
[27-Aug-2009 15:09:01] <jb> "Basically, the ICMP ping packet generated by an OS ping or a Zenoss ping command contains a different load than the packet generated by the zenping daemon. I've not verified this myself, but you can try using a sniffer to see if that is the case. "
[27-Aug-2009 15:09:05] <jb> http://forums.zenoss.com/viewtopic.php?t=10785
[27-Aug-2009 15:09:10] <jb> did you see this thread?
[27-Aug-2009 15:09:27] <djack> yes how can i fix that ?
[27-Aug-2009 15:09:47] <jb> sounds like a firewall issue..
[27-Aug-2009 15:09:58] <jb> look at the actual packets to see if they are different
[27-Aug-2009 15:10:00] <jb> (ping / zenping)
[27-Aug-2009 15:10:13] <jb> i don't see how they could be
[27-Aug-2009 15:10:15] <jb> but who knows..
[27-Aug-2009 15:10:47] <djack> why don't just use the standard ping command in stead of the zenping?
[27-Aug-2009 15:12:15] <djack> it seems strang that zenoss changes the load of the ping package....
[27-Aug-2009 15:14:37] <djack> jb that was my post ;-)
[27-Aug-2009 15:16:46] <djack> but there is no resolution ...
[27-Aug-2009 15:18:56] <rmatte> djack: oh, I might have an idea
[27-Aug-2009 15:19:08] <rmatte> actually, hmmm, let me check on something...
[27-Aug-2009 15:19:09] <djack> rmatte please you can make my night
[27-Aug-2009 15:19:20] <djack> and i can score points :-)
[27-Aug-2009 15:20:56] <rmatte> ah damn, nah, I was thinking about zMaxOIDPerRequest, but that doesn't apply to ping obviously
[27-Aug-2009 15:21:21] <rmatte> are you able to view the firewall logs?
[27-Aug-2009 15:21:26] <djack> nope
[27-Aug-2009 15:21:35] <djack> it is managed externally
[27-Aug-2009 15:21:46] <rmatte> I'd imagine it has something to do with the firewall blocking the ping packets coming from Zenoss because it doesn't like something about them
[27-Aug-2009 15:21:50] <rmatte> there's really no other explanation
[27-Aug-2009 15:22:20] <djack> is there a way to replace the zenping with the zencommand ping ?
[27-Aug-2009 15:22:32] <djack> on a device level?
[27-Aug-2009 15:22:34] <rmatte> well, there is, but it'd tedious
[27-Aug-2009 15:22:38] <rmatte> it's*
[27-Aug-2009 15:22:57] <djack> i can try nothing to lose
[27-Aug-2009 15:23:23] <rmatte> well hold on, I want to read up on this a bit more
[27-Aug-2009 15:23:52] <djack> oki
[27-Aug-2009 15:25:31] <rmatte> afk for a bit
[27-Aug-2009 15:27:20] <jb> hey egor
[27-Aug-2009 15:27:28] <bigegor> hey
[27-Aug-2009 15:28:07] <jb> gotta question
[27-Aug-2009 15:28:16] <jb> from socket import gethostbyaddr
[27-Aug-2009 15:28:16] <jb> evt.summary = "LTM Service Down"
[27-Aug-2009 15:28:16] <jb> ip = getattr(evt, 'bigipNotifyObjNode', 'Unknown')
[27-Aug-2009 15:28:16] <jb> port = getattr(evt, 'bigipNotifyObjPort', 'Unknown')
[27-Aug-2009 15:28:16] <jb> host = gethostbyaddr(ip)
[27-Aug-2009 15:28:18] <jb> evt.component = 'Node: %s:%s' % (host,port)
[27-Aug-2009 15:28:19] <jb> oops
[27-Aug-2009 15:28:23] <jb> http://pastebin.com/m64e86e09
[27-Aug-2009 15:28:39] <jb> that returns something like:
[27-Aug-2009 15:28:40] <jb> Component: Node: ('wrxxfscsaas01.fsc.follett.com', [], ['172.31.223.30']):82
[27-Aug-2009 15:29:00] <jb> whats the trick to get it to return "Component: Node: wrxxfscsass01.fsc.follett.com:82"
[27-Aug-2009 15:31:36] <bigegor> evt.component = 'Node: %s:%s' % (host[0],port)
[27-Aug-2009 15:31:47] <jb> [0] thats right
[27-Aug-2009 15:31:48] <jb> thanks
[27-Aug-2009 15:32:55] <rmatte> ok I'm back
[27-Aug-2009 15:33:55] <djack> and and ?
[27-Aug-2009 15:34:01] <rmatte> djack: can you do zenping run -v10 again but pastebin the whole output?
[27-Aug-2009 15:34:04] <rmatte> I'd like to skim over it
[27-Aug-2009 15:34:21] <rmatte> because you only pasted a small part of it before but we might have missed some other lines
[27-Aug-2009 15:36:16] <q7joey> i've applied the 14744 patch, but i seem to have a phantom process that generates events, but isn't in the process list. is this related to the 2.4.3 sickness?
[27-Aug-2009 15:36:53] <rmatte> q7joey: phantom process?
[27-Aug-2009 15:36:55] <rmatte> example?
[27-Aug-2009 15:37:48] <djack> http://pastebin.com/d69cb8139
[27-Aug-2009 15:38:11] <q7joey> events for mysqldump not running, but mysqldump isn't being monitored. i had previously had a regex that match mysqld (now /usr/libexec/myqld) so that is probably where it came from, but i can't find anywhere to delete it. do you want screenshots?
[27-Aug-2009 15:39:42] <rmatte> please
[27-Aug-2009 15:43:34] <q7joey> http://www.spiretech.com/~joey/event.png and os.png
[27-Aug-2009 15:45:24] <rmatte> q7joey: first thing I'd do is go in to zendmd and do a reindex() and commit()
[27-Aug-2009 15:45:37] <rmatte> I remember some process monitoring got out of wack for me one time and that fixed it
[27-Aug-2009 15:46:07] <q7joey> ok, trying that now.
[27-Aug-2009 15:46:56] <q7joey> now i have to watch to see if the timestamp updates
[27-Aug-2009 15:47:05] <jb> just clear the event..
[27-Aug-2009 15:47:07] <jb> and maybe remodel the device
[27-Aug-2009 15:47:21] <rmatte> yeh, a remodel couldn't hurt
[27-Aug-2009 15:47:24] <q7joey> i did the clear/remodel already. didn't help.
[27-Aug-2009 15:47:32] <jb> after you reindexed?
[27-Aug-2009 15:47:51] <q7joey> not yet. waiting to see if the timestamp increases.
[27-Aug-2009 15:48:53] <q7joey> timestamp just was touched. i'll try clear/remodel again.
[27-Aug-2009 15:49:45] <q7joey> and just to make sure, clear==move to history?
[27-Aug-2009 15:51:05] <jb> yep
[27-Aug-2009 15:52:01] <q7joey> it's back...
[27-Aug-2009 15:56:20] <q7joey> i guess i'll live with it for now and see if 2.4.4 clears it up.
[27-Aug-2009 15:56:34] <jb> i have some windows services that do the same thing
[27-Aug-2009 15:56:57] <djack> rmatte any ideas how to solve the zenping problem?
[27-Aug-2009 15:59:08] <jb> argh
[27-Aug-2009 15:59:14] <jb> soemthing keeps corrupting my cache
[27-Aug-2009 15:59:21] <q7joey> another issue that has been on the mailing list, but no answers to is gaps in graphs that started after i upgraded to 2.4.2. http://www.spiretech.com/~joey/render.png is a good example.
[27-Aug-2009 15:59:39] <q7joey> from looking at logs, i suspect 32/64 bit confusion.
[27-Aug-2009 15:59:51] <q7joey> 2009-08-27 13:57:56 ERROR netsnmp: truncating unsigned value to 32 bits (2)
[27-Aug-2009 16:00:01] <jb> that almost looks like zenperfsnmp problems
[27-Aug-2009 16:00:05] <jb> is it all of your SNMP graphs?
[27-Aug-2009 16:00:33] <q7joey> only graphs that have counters that wrap quickly and aren't using 64 bit counters.
[27-Aug-2009 16:00:57] <rmatte> q7joey: how many devices do you monitor on that box?
[27-Aug-2009 16:01:01] <q7joey> is there an easy way to force zenoss to choose 64 bit counters? it seems to be random as to which ones it decides to.
[27-Aug-2009 16:01:18] <straterra> not every device supports it, thats why
[27-Aug-2009 16:01:44] <q7joey> on a device that does 64, it only chooses 64 for some interfaces.
[27-Aug-2009 16:02:47] <rmatte> q7joey: you mean the templates?
[27-Aug-2009 16:02:58] <q7joey> i'm not running out of cpu or i/o. here is a good graph from the same device as the bad one: http://www.spiretech.com/~joey/render2.png the blank there is from my upgrading to 2.4.3
[27-Aug-2009 16:03:19] <rmatte> they are chosen based on interface type. If you click on an interface and check the "type" value, that's what it's chosen based on
[27-Aug-2009 16:03:27] <q7joey> right, it only uses the 64 bit templates for some interfaces, even if other have hc counters.
[27-Aug-2009 16:04:03] <rmatte> so if the interface type is ethernetCsmacd it uses that template, if it's ethernetCsmacd_64 it uses that template
[27-Aug-2009 16:04:12] <rmatte> so maybe the interfaces aren't identified properly via snmp
[27-Aug-2009 16:05:06] <q7joey> ds3_64 doesn't seem to use the hc values. is it only ethernetcsmacd_64 that will do it?
[27-Aug-2009 16:05:35] <rmatte> q7joey: the templates do whatever they are configured to do
[27-Aug-2009 16:06:08] <q7joey> so if i clone ether_64 to ds3_64, is that the method?
[27-Aug-2009 16:06:57] <rmatte> q7joey: that would be your best bet
[27-Aug-2009 16:07:07] <rmatte> cloning a non-64bit template certainly wouldn't help
[27-Aug-2009 16:07:30] <q7joey> i guess what i'm asking is will it magically map a ds3_64 interface type to use a ds3_64 template?
[27-Aug-2009 16:07:37] <rmatte> correct
[27-Aug-2009 16:07:47] <rmatte> as long as the template is in /Devices
[27-Aug-2009 16:08:14] <rmatte> you need to copy templates and make new ones for each interface type
[27-Aug-2009 16:08:18] <rmatte> I've made like 20 of them so far
[27-Aug-2009 16:08:33] <q7joey> i'm going to experiment right now...
[27-Aug-2009 16:12:47] <q7joey> ok, no obvious clone option. what is the right way to copy a template in /devices?
[27-Aug-2009 16:13:01] <mrayzenoss> I'm gonna go catch a flight home, 2.4.4 should be out in the morning. The code has been tagged, the 1 patch is the only difference from 2.4.3
[27-Aug-2009 16:13:23] <mrayzenoss> q7joey: there's a copy option for templates
[27-Aug-2009 16:13:50] <q7joey> but that seems to copy to another path. can i copy to /devices and it will make a new name?
[27-Aug-2009 16:13:54] <rmatte> yeh, go to /Devices, click on the template tab, put a checkmark next to the template you want to copy
[27-Aug-2009 16:14:00] <rmatte> then from the dropdown just select copy template
[27-Aug-2009 16:14:02] <rmatte> and copy it to /
[27-Aug-2009 16:14:12] <rmatte> then you'll see copy_of_whatever
[27-Aug-2009 16:14:18] <rmatte> go in to that and rename it to whatever you want
[27-Aug-2009 16:14:25] <q7joey> copy to /devices, right?
[27-Aug-2009 16:15:18] <rmatte> yes
[27-Aug-2009 16:17:20] <q7joey> ah, i see. everything is relative to /devices...
[27-Aug-2009 16:18:13] <rmatte> yes
[27-Aug-2009 16:21:33] <djack> hmmm
[27-Aug-2009 16:24:43] <rmatte> for interface templates anyways
[27-Aug-2009 16:25:04] <rmatte> /Devices is the "root" though, if that's what you mean
[27-Aug-2009 16:28:13] <q7joey> i'm waiting to see if my 64 bit magic worked. is there any easy way to convert 32 bit rrd to 64 bit to keep my history? i'm guessing not.
[27-Aug-2009 16:29:50] <rmatte> I don't believe so
[27-Aug-2009 16:30:13] <rmatte> if you google around you might find some way though
[27-Aug-2009 16:31:39] <q7joey> hmm, i'm looking at the rrd files and there may not be any real difference internally. a rename may be all i need...
[27-Aug-2009 16:34:06] <rmatte> that's entirely possible
[27-Aug-2009 16:34:11] <rmatte> RRD is pretty standard
[27-Aug-2009 16:34:51] <q7joey> i was just thinking that it might have raw 32/64 data and therefore would be different. things look ok after my copy, but i'll wait for a few updates and see if it survives.
[27-Aug-2009 16:37:48] <rmatte> I think it just stores values
[27-Aug-2009 16:38:08] <rmatte> the amount of traffic passing through a 32 bit interface is technically the same as when it's passing through a 64
[27-Aug-2009 16:38:16] <rmatte> in terms of throughput and speed
[27-Aug-2009 16:38:54] <q7joey> but a 64 bit counter could potentially create more than 32 bits worth of data per sample (that's the whole reason to use 64 bit counters). but it looks like zenoss doesn't do anything different with rrd.
[27-Aug-2009 16:39:03] <rmatte> yeh, that's true
[27-Aug-2009 16:39:12] <rmatte> cool
[27-Aug-2009 16:39:33] <rmatte> I think the difference in processing is just done internally within Zenoss, the way it stores the data to the RRD is most likely the same though
[27-Aug-2009 16:40:04] <rmatte> man, it's so nice to have all these systems patched and know that they are working properly
[27-Aug-2009 16:40:18] <rmatte> I even patched the ZVersion.py file, so I'm effectively running 2.4.4 on everything
[27-Aug-2009 16:41:20] <rmatte> now tomorrow I can focus on fixing up some reports
[27-Aug-2009 16:43:28] <rmatte> let's just hope there isn't a version 2.4.5 in the works by tomorrow
[27-Aug-2009 16:43:30] <rmatte>
[27-Aug-2009 16:44:07] <djack> lol
[27-Aug-2009 16:44:22] <djack> well i hope it will fix the zenping payload issue :-p
[27-Aug-2009 16:44:56] <rmatte> did you ever pastebin that output?
[27-Aug-2009 16:45:12] <djack> yeah i did
[27-Aug-2009 16:45:25] <djack> http://pastebin.com/d69cb8139
[27-Aug-2009 16:45:28] <rmatte> thanks
[27-Aug-2009 16:45:34] <rmatte> I'm curious to see what's causing that
[27-Aug-2009 16:45:42] <djack> me to
[27-Aug-2009 16:47:34] <rmatte> http://forums.zenoss.com/viewtopic.php?p=35458
[27-Aug-2009 16:47:42] <rmatte> that issue has been around for a while it'd seem
[27-Aug-2009 16:47:49] <djack> idd
[27-Aug-2009 16:48:07] <djack> nobody seems to solve it ....
[27-Aug-2009 16:49:01] <rmatte> I have to leave in about 15 mins, but if you come around tomorrow I'll help you troubleshoot this more
[27-Aug-2009 16:49:16] <rmatte> I'd like to know the answer in case I ever run across it, or in case we can get a patch made for it
[27-Aug-2009 16:49:38] <rmatte> I was just too busy today to really be in here too much
[27-Aug-2009 16:49:42] <djack> rmatte how late tomorrow ?
[27-Aug-2009 16:49:47] <djack> or can we mail ?
[27-Aug-2009 16:49:52] <rmatte> well, what time is it there right now?
[27-Aug-2009 16:50:14] <djack> in belgium 8 minutes before 12
[27-Aug-2009 16:50:17] <rmatte> yikes
[27-Aug-2009 16:50:21] <djack> lol
[27-Aug-2009 16:50:26] <rmatte> 6 hours difference then
[27-Aug-2009 16:50:32] <djack> you ?
[27-Aug-2009 16:50:37] <djack> states i guess
[27-Aug-2009 16:50:40] <rmatte> which means I will be in here at about 4pm your time
[27-Aug-2009 16:50:44] <rmatte> nah, canada
[27-Aug-2009 16:50:52] <djack> lol
[27-Aug-2009 16:51:06] <djack> btw the ping it self is done in the ascping.py ?
[27-Aug-2009 16:51:18] <rmatte> most of the people in here are EST or CST, so you're at a disadvantage when it comes to time for support lol
[27-Aug-2009 16:51:44] <djack> well no problem i'm a night kind of person
[27-Aug-2009 16:52:34] <rmatte> $ZENHOME/Products/ZenStatus/AsyncPing.py
[27-Aug-2009 16:52:39] <rmatte> that's the file doing the ping
[27-Aug-2009 16:52:46] <rmatte> apparently
[27-Aug-2009 16:52:55] <djack> well i'm going to look in the construction of the payload...
[27-Aug-2009 16:53:03] <djack> i found python code that works :-p
[27-Aug-2009 16:53:06] <rmatte> hmmm, this is interesting
[27-Aug-2009 16:53:09] <rmatte> http://forums.zenoss.com/viewtopic.php?t=99&sid=e0fa6b22b54b139ca0a3e2145c01d13b
[27-Aug-2009 16:53:22] <rmatte> Quoting the forum:
[27-Aug-2009 16:53:23] <rmatte> http://forums.zenoss.com/viewtopic.php?t=99&sid=e0fa6b22b54b139ca0a3e2145c01d13b
[27-Aug-2009 16:53:24] <rmatte> oops
[27-Aug-2009 16:53:28] <rmatte> http://forums.zenoss.com/viewtopic.php?t=99&sid=e0fa6b22b54b139ca0a3e2145c01d13b
[27-Aug-2009 16:53:30] <rmatte> bah
[27-Aug-2009 16:53:33] <rmatte> why is it not copying
[27-Aug-2009 16:53:45] <djack> what?
[27-Aug-2009 16:53:55] <rmatte> If you change:
[27-Aug-2009 16:53:55] <rmatte> icmppkt = icmp.Packet(ipreply.data)
[27-Aug-2009 16:53:55] <rmatte> to:
[27-Aug-2009 16:53:55] <rmatte> icmppkt = icmp.Packet(ipreply.data, 0)
[27-Aug-2009 16:53:55] <rmatte> In $ZENHOME/Products/ZenStatus/AsyncPing.py, then it won't fail the
[27-Aug-2009 16:53:56] <rmatte> checksum and log the error, but you will still get "unexpected" log
[27-Aug-2009 16:53:58] <rmatte> messages.
[27-Aug-2009 16:54:00] <rmatte> that
[27-Aug-2009 16:54:02] <rmatte> lol
[27-Aug-2009 16:54:04] <rmatte> that's what the posting says
[27-Aug-2009 16:54:10] <rmatte> not sure if that's already changed or not
[27-Aug-2009 16:54:14] <rmatte> but if it's not, it's worth a shot
[27-Aug-2009 16:54:27] <rmatte> and if it works after changing it I'll bring that to Matt's attention tomorrow
[27-Aug-2009 16:54:28] <djack> let's check
[27-Aug-2009 16:55:28] <djack> restarting zenoss
[27-Aug-2009 16:55:28] <rmatte> ah, it looks like it's already there
[27-Aug-2009 16:55:57] <rmatte> plog.debug("checksum failure on packet %r", ipreply.data)
[27-Aug-2009 16:55:57] <rmatte> try:
[27-Aug-2009 16:55:57] <rmatte> icmppkt = icmp.disassemble(ipreply.data, 0)
[27-Aug-2009 16:56:07] <rmatte> oh wait
[27-Aug-2009 16:56:09] <rmatte> that's different
[27-Aug-2009 16:56:49] <rmatte> I don't see anything like either line in the file
[27-Aug-2009 16:57:31] <rmatte> I think that...
[27-Aug-2009 16:57:32] <rmatte> ipreply = ip.disassemble(data)
[27-Aug-2009 16:57:32] <rmatte> try:
[27-Aug-2009 16:57:32] <rmatte> icmppkt = icmp.disassemble(ipreply.data)
[27-Aug-2009 16:57:32] <rmatte> except ValueError:
[27-Aug-2009 16:57:33] <rmatte> plog.debug("checksum failure on packet %r", ipreply.data)
[27-Aug-2009 16:57:35] <rmatte> try:
[27-Aug-2009 16:57:37] <rmatte> icmppkt = icmp.disassemble(ipreply.data, 0)
[27-Aug-2009 16:57:39] <rmatte> does all that
[27-Aug-2009 16:57:43] <rmatte> so doesn't look like it'll make a difference
[27-Aug-2009 16:58:42] <djack> :-(
[27-Aug-2009 17:04:49] <djack> i tricked zenping to believe the host is up
[27-Aug-2009 17:05:51] <rmatte> that's only a temporary solution
[27-Aug-2009 17:05:52] <djack> but no graphing then also
[27-Aug-2009 17:05:58] <djack> idd
[27-Aug-2009 17:06:05] <rmatte> you know, you could just disable ping monitoring all together in zProperties for now
[27-Aug-2009 17:06:21] <rmatte> then use SNMP Agent Down events as indication that the device is down
[27-Aug-2009 17:06:32] <djack> do the snmp stuff works then
[27-Aug-2009 17:06:32] <rmatte> just for the time being until we can figure out what's causing that
[27-Aug-2009 17:06:41] <rmatte> it should work, yeh
[27-Aug-2009 17:06:50] <djack> where do i change it ?
[27-Aug-2009 17:07:01] <rmatte> you know where zProperties is right?
[27-Aug-2009 17:07:04] <rmatte> go in to that for each device
[27-Aug-2009 17:07:21] <rmatte> Change zPingMonitorIgnore from False to True
[27-Aug-2009 17:07:25] <rmatte> and it'll ignore ping monitoring
[27-Aug-2009 17:07:32] <rmatte> then it should hopefully start collecting SNMP data
[27-Aug-2009 17:08:01] <rmatte> (I would have suggested this earlier but I was trying to possibly fix the actual ping issue, since it's nice to have ping monitoring)
[27-Aug-2009 17:08:36] <djack> can the snmp be forced?
[27-Aug-2009 17:08:37] <rmatte> g'day Chad
[27-Aug-2009 17:08:46] <rmatte> forced?
[27-Aug-2009 17:08:59] <rmatte> When you disable the ping monitoring Zenoss will attempt to monitor via SNMP no matter what
[27-Aug-2009 17:09:24] <rmatte> the only reason it hasn't been is because when something is showing as down via ping Zenoss stops all monitoring for that device
[27-Aug-2009 17:09:33] <rmatte> because there's no point in monitoring an unreachable device
[27-Aug-2009 17:09:48] <rmatte> but if you tell it not to do the ping monitoring, then it'll monitor with snmp
[27-Aug-2009 17:10:04] <djack> lets'hope
[27-Aug-2009 17:10:24] <rmatte> but this is really more of a temporary solution, since the ping monitoring is pretty important to have
[27-Aug-2009 17:10:54] <djack> all graphs show nan
[27-Aug-2009 17:10:56] <djack> hmmm
[27-Aug-2009 17:11:32] <rmatte> they'll be nan for 3 polling cycles
[27-Aug-2009 17:11:37] <rmatte> (about 15 to 20 mins)
[27-Aug-2009 17:11:41] <rmatte> then they should start populating
[27-Aug-2009 17:11:43] <rmatte> that's normal
[27-Aug-2009 17:11:46] <djack> oki
[27-Aug-2009 17:12:49] <djack> the zpersnmp.logs shows that a success on 16 devices this is beter then 9
[27-Aug-2009 17:13:09] <rmatte> yup
[27-Aug-2009 17:13:13] <rmatte> I'm sure it's going to work
[27-Aug-2009 17:13:45] <rmatte> you just won't get "ip blah is down" when the device goes down, you'll see "SNMP agent down" instead
[27-Aug-2009 17:13:51] <djack> well tommorow or in my timezone today i will come back to maybe help to troubleshout so m more
[27-Aug-2009 17:13:59] <rmatte> cool
[27-Aug-2009 17:14:30] <djack> well i can learn from it also ;-)
[27-Aug-2009 17:14:36] <rmatte> yup
[27-Aug-2009 17:15:01] <rmatte> well, I'm going to head out of here
[27-Aug-2009 17:15:06] <rmatte> I'll talk to you later
[27-Aug-2009 17:15:20] <djack> oki have a nice evening and thanks
[27-Aug-2009 17:15:27] <rmatte> you too
[27-Aug-2009 17:15:29] <rmatte> no problem
[27-Aug-2009 17:16:06] <djack> oooh i have values ;-)
[27-Aug-2009 17:16:08] <djack> wiiiiiiiiiiiiiiiiiiiiii
[27-Aug-2009 17:16:11] <rmatte> good stuff
[27-Aug-2009 17:28:44] <ironpaw> hi
[27-Aug-2009 17:38:23] <djack> hi
[27-Aug-2009 17:42:43] <djack> exit
[27-Aug-2009 17:50:26] <ironpaw> aww
[27-Aug-2009 17:50:30] <ironpaw> no love
[27-Aug-2009 23:04:11] <davetoo> hmm
[27-Aug-2009 23:06:05] <davetoo> anybody awake here? I just discovered that deleting a Network breaks all the Interface objects that had IPs in that network.
[28-Aug-2009 00:00:46] [disconnected at Fri Aug 28 00:00:46 2009]
[28-Aug-2009 00:00:46] [connected at Fri Aug 28 00:00:46 2009]
[28-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[28-Aug-2009 04:55:25] <Dieterbe> anyone here using zenperfwmi? i get syntax errors in the log file (and events for it) but i don't know for which queries. i've set debug=True in its config and reloaded the daemon, but i still don't see the faulty queries in the logfile
[28-Aug-2009 08:55:12] <rmatte> good morning Matt
[28-Aug-2009 08:55:29] <mrayzenoss> morning
[28-Aug-2009 08:56:09] <rmatte> no other blockers were found overnight I hope
[28-Aug-2009 08:56:10] <rmatte> lol
[28-Aug-2009 09:15:07] <mrayzenoss> They're taking a hard look at http://dev.zenoss.com/trac/ticket/5469
[28-Aug-2009 09:44:38] <rmatte> yeh, I was looking at that
[28-Aug-2009 09:44:43] <rmatte> I saw Jane's post
[28-Aug-2009 09:45:02] <rmatte> that would be a nice one to fix as well hehe
[28-Aug-2009 09:45:16] <mrayzenoss> I think we're going to fix it too
[28-Aug-2009 09:45:22] <rmatte> cool
[28-Aug-2009 09:45:29] <rmatte> I look forward to the patch
[28-Aug-2009 09:45:31] <rmatte>
[28-Aug-2009 09:45:56] <rmatte> I don't have any command datapoints setup that way right now, so it's not as big of a deal for me as the other patch was
[28-Aug-2009 10:13:22] <dollarbang> is there a specific requirement for Zenoss to run Zenoss secure 'https'? Its on the intranet, so its relately secure anyway. Reason I ask is that I think that running Zenoss 'https' instead of 'http' causes everything to run more encrypted, could the encryption be affecting the maps? I've also noticed that every now and then I get an error "Error - Server connection error"... Can this be caused by the 'https' encryption?
[28-Aug-2009 10:21:26] <rmatte> dollarbang: how are you running it as https? Just via a proxy?
[28-Aug-2009 10:26:55] <dollarbang> rmatte: I knew if I asked a question, it would be over my head immediately. One of our Network OPS guys thinks that running Zenoss Core 2.4.2 is causing issues with Google Maps. He has his own Zenoss server 2.3.3, his entire device list was dumped and loaded onto another Zenoss Core 2.4.2 server (shutting down his server in the process to keep collisions down). Except for the different rev levels, the 2.4.2 version is using HTTPS for access to the Zenos
[28-Aug-2009 10:28:59] <rmatte> well, do you have the right API key for google maps?
[28-Aug-2009 10:29:10] <rmatte> when you sign up for the key you might have to include https:// as the prefix
[28-Aug-2009 10:29:32] <rmatte> Go to Settings, the API key input field is right at the bottom
[28-Aug-2009 10:29:40] <rmatte> with a link next to it for instructions on getting a key
[28-Aug-2009 10:32:58] <dollarbang> there is a key there, I'll check on the Google web site.
[28-Aug-2009 10:34:03] <rmatte> k
[28-Aug-2009 10:34:29] <rmatte> I'm thinking that they key might have been for http:// prefix, so when you switched to https:// it started not liking the key
[28-Aug-2009 10:35:02] <rmatte> also, the server connection errors just indicate that you've temporarily lost contact with the server, which could be for a number of reasons
[28-Aug-2009 10:35:15] <rmatte> If you refresh the page does it come back right away when that happens?
[28-Aug-2009 10:37:15] <dollarbang> rmatte: need to break for lunch, when I return, I'll gen another key.
[28-Aug-2009 10:38:04] <rmatte> k
[28-Aug-2009 10:49:27] <rmatte> mrayzenoss: have you seen http://forums.zenoss.com/viewtopic.php?t=10830 yet?
[28-Aug-2009 10:57:23] <jb> ok
[28-Aug-2009 10:57:31] <jb> time to fix these problems
[28-Aug-2009 10:57:47] <jb> cgibbons: more zendmd issues started popping up again :/
[28-Aug-2009 10:58:01] <jb> cgibbons: workign with kells in just a sec to get them fixed though
[28-Aug-2009 10:58:02] <cgibbons> oh?
[28-Aug-2009 10:58:16] <jb> yeah.. cache keeps getting corrupted, zendmd errors, etc
[28-Aug-2009 10:59:39] <mrayzenoss1> ugh, 2.4.5 Monday
[28-Aug-2009 11:00:02] <jb> really?
[28-Aug-2009 11:00:17] <mrayzenoss1> jb: don't worry, Enterprise people won't get it until Wednesday
[28-Aug-2009 11:00:24] <rmatte> lol
[28-Aug-2009 11:00:27] <jb> im still 2.4.2 anyways
[28-Aug-2009 11:00:55] <rmatte> 2.4.2 is fine as long as you trust all of your standard users logging in to Zenoss
[28-Aug-2009 11:00:58] <mrayzenoss1> http://dev.zenoss.com/trac/ticket/5469 is the ticket that caused 2.4.5
[28-Aug-2009 11:01:00] <jb> yeah
[28-Aug-2009 11:01:15] <rmatte> ah
[28-Aug-2009 11:01:24] <cgibbons> 2.4.5?
[28-Aug-2009 11:01:26] <cgibbons> not 2.4.4?
[28-Aug-2009 11:01:35] <jb> 2.4.4 fixed the event clear emails
[28-Aug-2009 11:01:52] <jb> 2.4.5 fixes the severity
[28-Aug-2009 11:01:57] <jb> (i think)
[28-Aug-2009 11:04:18] <mrayzenoss1> it's 2.4.5 because 2.4.4 was publicly available via apt, yum and appliance upgrades for probably 2 hours
[28-Aug-2009 11:04:27] <cgibbons> ahhh snap
[28-Aug-2009 11:08:08] <rocket> bah ..thats not good
[28-Aug-2009 11:14:08] mrayzenoss1 is now known as mrayzenoss
[28-Aug-2009 11:18:26] <etank> heh. we are doing our enterprise install today
[28-Aug-2009 11:18:56] <etank> guess we will do an upgrade on Wednesday then
[28-Aug-2009 11:31:52] <rmatte> http://dev.zenoss.org/trac/ticket/5473
[28-Aug-2009 11:32:12] <rmatte> can someone else who is running 2.4.3 test that to confirm please?
[28-Aug-2009 11:32:25] <rmatte> I'm seeing that behaviour across all 5 of my production Zenoss boxes
[28-Aug-2009 11:41:52] <schwartz> sya morning
[28-Aug-2009 12:52:18] <dollarbang> rmatte: I'm back, you asked about the API key. It was generated according to the Zenoss doc and with https:
[28-Aug-2009 12:57:50] <davetoo> uh oh
[28-Aug-2009 12:57:53] <davetoo> patch to the patch?
[28-Aug-2009 13:20:51] <rmatte> davetoo: no, just another patch
[28-Aug-2009 13:30:32] <djack> hellow world
[28-Aug-2009 13:31:53] <djack> small question is there a way to monitor the load per core via snmp ?
[28-Aug-2009 13:37:30] <cgibbons> yay community
[28-Aug-2009 13:38:45] <mrayzenoss> cgibbons: what's that?
[28-Aug-2009 13:39:00] <cgibbons> yellow hat from mrchippy
[28-Aug-2009 13:39:02] <mrayzenoss> djack: haven't seen per core, only per processor
[28-Aug-2009 13:39:12] <mrayzenoss> yay community!
[28-Aug-2009 13:39:35] <rocket>
[28-Aug-2009 13:40:23] <rmatte> djack: for what kind of server?
[28-Aug-2009 13:41:45] <rmatte> windows, linux?
[28-Aug-2009 13:45:48] * rocket starts grumbling ... how am I supposed to get work done if I am tired....
[28-Aug-2009 13:46:22] <rmatte> ...how am I supposed to get work done when my enthusiasm is sitting at the bar right now?...
[28-Aug-2009 13:46:39] <rmatte> ...and I'm tempted to follow it...
[28-Aug-2009 13:47:23] <cgibbons> you see the 15 minute caffeine powernap article, rocket?
[28-Aug-2009 13:48:08] <rocket> cgibbons: no ... tell me more ..
[28-Aug-2009 13:48:25] <cgibbons> yeah: http://lifehacker.com/software/sleep/reboot-your-brain-with-a-caffeine-nap-306029.php
[28-Aug-2009 13:48:35] <rocket> cgibbons: I was repairing a oracle/sap db server yesterday from 8am-3am ...
[28-Aug-2009 13:48:41] <cgibbons> ouf
[28-Aug-2009 13:51:01] <rocket> yeah .. sleep deprivation is making it hard to do the tedious hacking on zenoss ...
[28-Aug-2009 13:51:22] <rocket> eg the copyright stuff .. the comments etc ..
[28-Aug-2009 13:53:08] <cgibbons> i really miss the powerful IDEs the Java world had for that stuff, I rarely had to do the tedious broilerplate work... the python tools just aren't really that good (yet?)
[28-Aug-2009 13:53:32] * rocket is using vim and havent made the macros yet to automate ... :/
[28-Aug-2009 13:54:07] <cgibbons> yeah I used vim only for a long-time here but eventually switched to Eclipse + pydev once I got it all setup right.
[28-Aug-2009 13:54:14] <jb> i need to checkout eclipse
[28-Aug-2009 13:54:42] <rocket> cgibbons: sometime I will have to have you show me the setup
[28-Aug-2009 13:54:45] <jb> ive been using (eek, i know) nano+python template
[28-Aug-2009 13:54:56] <cgibbons> sure, i've got a partial write on how to do it, i'll finish tha tup
[28-Aug-2009 13:55:02] <jb> ah nice.
[28-Aug-2009 13:55:06] <cgibbons> mrchippy wrote up one for us when we first started here, but it's dated now
[28-Aug-2009 13:55:17] <jb> ugh, i still need to figure out this IO thing
[28-Aug-2009 13:55:21] <jb> i'll do it next week.
[28-Aug-2009 13:55:30] <rocket> cgibbons: I need to work on the installer too so I can have a native gentoo install ...
[28-Aug-2009 13:55:54] <rocket> right now I am having to do everything in a separate vm .. so I have to replicate config files etc across ...
[28-Aug-2009 13:56:14] <davetoo> jb?
[28-Aug-2009 13:56:16] <davetoo> jan?
[28-Aug-2009 13:56:22] <cgibbons> you just don't use a source build on your gentoo box?
[28-Aug-2009 13:56:34] <mrayzenoss> rocket: installer? Like a Gentoo build?
[28-Aug-2009 13:57:10] <davetoo> mrayzenoss: are you in the Annapolis office?
[28-Aug-2009 13:57:22] <mrayzenoss> I flew back to Austin last night
[28-Aug-2009 13:57:25] <rocket> mrayzenoss: yes
[28-Aug-2009 13:57:26] <davetoo> oh,
[28-Aug-2009 13:57:34] <davetoo> I was in Annapolis W-Fri last week
[28-Aug-2009 13:57:38] <davetoo> crazy weather
[28-Aug-2009 13:58:00] <cgibbons> man i haven't been to annapolis since the summer party we had there last july, crazy
[28-Aug-2009 13:58:07] <rocket> cgibbons: I cant remember my issue but the source installer borked on my gentoo machine
[28-Aug-2009 13:58:15] <cgibbons> huh
[28-Aug-2009 13:58:27] <jb> heh.. have you guys eaten at Lemongrass?
[28-Aug-2009 13:58:34] <jb> in annapolis, right down the street from zenoss..
[28-Aug-2009 13:58:39] <davetoo> I flew out Saturday morning, but one of the guys was supposed to leave Friday night.. and everthing from DC to Maine was closed for weather
[28-Aug-2009 13:58:51] <davetoo> I just "ate" at Fado'
[28-Aug-2009 13:58:53] <rocket> cgibbons: I believe it was due to library versions installed etc .. something wasnt compiling...
[28-Aug-2009 13:59:07] <mrayzenoss> davetoo: at at Carpaccio's, next to Fado
[28-Aug-2009 13:59:17] <davetoo> we got some slices there
[28-Aug-2009 13:59:27] <jb> carpaccios is good
[28-Aug-2009 13:59:30] <jb> so is fados
[28-Aug-2009 13:59:34] <jb> but lemongrass > *
[28-Aug-2009 13:59:38] <cgibbons> annapolis is a bit too cold for me, brrrr
[28-Aug-2009 13:59:50] <jb> best mojitos ever
[28-Aug-2009 14:03:06] <djack> rmatte linux servers
[28-Aug-2009 14:03:15] <rmatte> cgibbons: come to Ottawa, Canada, then you'll have a new appreciation for the weather in Annapolis
[28-Aug-2009 14:03:22] <jb> or chicago
[28-Aug-2009 14:03:26] <rmatte> djack: ah
[28-Aug-2009 14:03:29] <djack> nah belgium :-p
[28-Aug-2009 14:03:41] kevin7kal_ is now known as kevin7kal
[28-Aug-2009 14:03:43] <rmatte> djack: you could do it via ssh with a script
[28-Aug-2009 14:03:49] <cgibbons> oh, i know... i actually grew up in milwaukee, brr... and i got to spend quite a bit of time in Calgary in '07 @ a customer when i worked at my previous place. brr.
[28-Aug-2009 14:03:50] <rocket> heh Minnesota can be quite odd from a weather perspective
[28-Aug-2009 14:04:26] <rocket> 90's one day ... 60's the next
[28-Aug-2009 14:04:27] <djack> hmmm nagios has a script for that see if i can get it to work
[28-Aug-2009 14:05:21] <djack> rmatte already some ideas about the zenping issue ?
[28-Aug-2009 14:09:34] <rmatte> djack: I haven't had time to look at that at all today
[28-Aug-2009 14:09:37] <rmatte> way too busy
[28-Aug-2009 14:10:28] <rmatte> I'm getting sent love notes from some of our engineers about the latest Zenoss update
[28-Aug-2009 14:10:46] <rmatte> </sarcasm>
[28-Aug-2009 14:11:04] <rmatte> although 14744 fixed the major issue, so I'm happy enough
[28-Aug-2009 14:14:00] <djack> rmatte i don't follow the the sarcasm but no mather
[28-Aug-2009 14:15:00] <rmatte> the sarcasm was emphasized on the word "love"
[28-Aug-2009 14:15:16] <rmatte> lol
[28-Aug-2009 14:16:07] <djack> i don't know what kind or relation you have with uour engineers he ;-p
[28-Aug-2009 14:16:57] <davetoo> I probably shouldn't have been so quick to make a vm template from 2.4.3
[28-Aug-2009 14:17:29] <rmatte> djack: generally a good one, it's just been a rough day
[28-Aug-2009 14:17:42] <rmatte> davetoo: why?
[28-Aug-2009 14:18:01] <rmatte> davetoo: just apply the patches and your template will be fine
[28-Aug-2009 14:19:16] <davetoo> yeah, just have to convert out and back in
[28-Aug-2009 14:19:28] <davetoo> it's just for development anyway
[28-Aug-2009 14:19:34] <rmatte> yeh
[28-Aug-2009 14:20:02] <davetoo> I'm finding some interesting "features" in 2.4.x
[28-Aug-2009 14:20:14] <davetoo> doesn't anybody else use the ZMI to poke around/debug?
[28-Aug-2009 14:21:00] <mrayzenoss> be back soon
[28-Aug-2009 14:21:18] <rmatte> nah, my zenoss jacket is here but shipping department is closed
[28-Aug-2009 14:21:27] <rmatte> and it's actually chilly out today, I could use the jacket
[28-Aug-2009 14:21:28] <rmatte>
[28-Aug-2009 14:21:33] <rmatte> nah = bah
[28-Aug-2009 14:22:35] <davetoo> rmatte: where are you?
[28-Aug-2009 14:22:49] <djack> rmatte i had a nice day showing off the zenoss to the bosses they where impressed
[28-Aug-2009 14:22:53] <rocket> rmatte: jacket is nice ..
[28-Aug-2009 14:23:03] <rmatte> djack: good to hear
[28-Aug-2009 14:23:14] <rmatte> rocket: cool lol
[28-Aug-2009 14:23:24] <djack> rmatte can i submit a bug for the zenping ?
[28-Aug-2009 14:23:36] <rmatte> djack: do you have a trac account yet?
[28-Aug-2009 14:23:43] <djack> not yet
[28-Aug-2009 14:23:43] <davetoo> Cisco NX-OS is driving me bonkers
[28-Aug-2009 14:24:00] <rmatte> let me check if there's already a backlogged bug for it
[28-Aug-2009 14:24:16] <djack> rmatte if so kick it to the top :-)
[28-Aug-2009 14:24:32] <davetoo> rmatte: you zenoss crew?
[28-Aug-2009 14:24:46] <rmatte> http://dev.zenoss.com/trac/ticket/4886
[28-Aug-2009 14:24:48] <rmatte> there you go
[28-Aug-2009 14:25:07] <rmatte> sort of
[28-Aug-2009 14:25:09] <rmatte> similar issue
[28-Aug-2009 14:25:25] <rmatte> actually no it's not
[28-Aug-2009 14:25:30] <djack> lol
[28-Aug-2009 14:25:30] <rmatte> the title of the bug report just sucks
[28-Aug-2009 14:25:31] <rmatte> lol
[28-Aug-2009 14:26:56] <rmatte> djack: if you write up a detailed bug report I'll file it for you
[28-Aug-2009 14:27:05] <rmatte> send it to rmatte@novanetworks.com
[28-Aug-2009 14:27:28] <djack> well i see what i can provide :-)
[28-Aug-2009 14:27:29] <rmatte> also, you may want to ask mrayzenoss to make a Trac account for you so you can file bugs yourself
[28-Aug-2009 14:27:58] <davetoo> What is the purpose of the "Version" field? Is it supposed to be the version the bug is in? Or the version targeted for the fix?
[28-Aug-2009 14:28:11] <rmatte> but I'd do that in a couple weeks since he's ridiculously busy
[28-Aug-2009 14:28:22] <rmatte> davetoo: yeh
[28-Aug-2009 14:28:37] <rmatte> davetoo: if it's present in all versions then put ALL, if it appeared in 2.4.3 then put 2.4.3
[28-Aug-2009 14:28:55] <rmatte> and platform is generally "ALL"
[28-Aug-2009 14:29:01] <djack> rmatte what do you call detailed ?
[28-Aug-2009 14:29:12] <davetoo> 'k, 'cause there's freeform text in the body for the build/version
[28-Aug-2009 14:29:18] <rmatte> djack: a clear explanation of the issue you are experience with log lines included
[28-Aug-2009 14:29:22] <davetoo> where I'm generally putting my rpm package name
[28-Aug-2009 14:29:28] <rmatte> the issue your are experienCING
[28-Aug-2009 14:29:29] <rmatte> rather
[28-Aug-2009 14:30:00] <rmatte> davetoo: ah
[28-Aug-2009 14:30:11] <rmatte> davetoo: well, it's really up to you, I try to make the bug reports more generalize
[28-Aug-2009 14:30:14] <rmatte> generalized*
[28-Aug-2009 14:30:24] <davetoo> I've only done two with this new format
[28-Aug-2009 14:30:32] <davetoo> been a few months, before that.
[28-Aug-2009 14:30:42] <djack> rmatte i see if i can even add the firewall logs :-)
[28-Aug-2009 14:30:48] <davetoo> What's the python idiom/equivalent for the perl diamond operator?
[28-Aug-2009 14:31:22] <rmatte> djack: that'd help
[28-Aug-2009 14:31:39] <davetoo> stupid Cisco NX-OS doesn't have working snmp for the IP interfaces/routes yet.
[28-Aug-2009 14:31:56] <djack> btw i had something strange thing today... i added the apachemonitor template to one of the webservers when i went and look to the perfdata all the regular perf of the cpu and memory and so was f*cked
[28-Aug-2009 14:32:08] <rmatte> davetoo: http://xorlogic.blogspot.com/2007/12/perls-diamond-operator-in-python.html
[28-Aug-2009 14:32:57] <davetoo> thanks
[28-Aug-2009 14:33:01] <rmatte> np
[28-Aug-2009 14:33:11] <davetoo> I still don't know the right keywords to search yet
[28-Aug-2009 14:33:16] <rmatte> hehe
[28-Aug-2009 14:33:18] <davetoo> Know a lot of perl, not so much python yet
[28-Aug-2009 14:33:27] <davetoo> you say 'dict', I say 'hash'
[28-Aug-2009 14:33:42] <rmatte> hehe
[28-Aug-2009 14:34:01] <rmatte> I'm still learning python myself, and I don't even know perl that well
[28-Aug-2009 14:34:06] <rmatte> but I'm a god with bash scripting
[28-Aug-2009 14:34:08] <rmatte>
[28-Aug-2009 14:34:25] <davetoo> I hurt my brain on bash scripting
[28-Aug-2009 14:34:30] <davetoo> arrays and such
[28-Aug-2009 14:34:35] <rmatte> I've been using unix type operating systems since I was 14 or 15
[28-Aug-2009 14:34:38] <rmatte> so it's grown on me
[28-Aug-2009 14:34:54] <davetoo> when I started using unix, there was no bash
[28-Aug-2009 14:35:02] <davetoo> but I didn't keep it up.
[28-Aug-2009 14:35:03] <rmatte> yeh, but I'm only 24
[28-Aug-2009 14:35:07] <rmatte> so I never had the chance
[28-Aug-2009 14:35:08] <rmatte>
[28-Aug-2009 14:35:25] <davetoo> so the fileinput module seems to be what I need
[28-Aug-2009 14:35:26] <rmatte> I'm the only 24 year I know who can boast 8-9 years unix experience
[28-Aug-2009 14:35:32] <davetoo> heh
[28-Aug-2009 14:35:38] <davetoo> I know a few
[28-Aug-2009 14:35:46] <rmatte> it's fairly rare though
[28-Aug-2009 14:35:53] <davetoo> yes
[28-Aug-2009 14:35:59] <rmatte> I started out with FreeBSD for my first 5 years
[28-Aug-2009 14:36:02] <rmatte> then moved on to Linux
[28-Aug-2009 14:36:10] <rmatte> I think it's coming up on almost 5 years of Linux
[28-Aug-2009 14:36:18] <davetoo> I started with Slackware 1.14 I think....
[28-Aug-2009 14:36:26] <davetoo> I'm too old to remember
[28-Aug-2009 14:36:35] <davetoo> but I mostly use FreeBSD now
[28-Aug-2009 14:36:42] <rmatte> I started with FreeBSD, then moved to Gentoo Linux (before it even had an installer, did the entire installation by hand with commands)
[28-Aug-2009 14:36:50] <rmatte> now I'm using Ubuntu since I'm sick of the hassle
[28-Aug-2009 14:36:57] <rmatte> I still have Gentoo on my home server though
[28-Aug-2009 14:36:58] <djack> 11 years of linux :-) debian lover still am but because not commercial supported it's suse currently
[28-Aug-2009 14:37:16] <rmatte> I prefer Ubuntu over Suse
[28-Aug-2009 14:37:18] <davetoo> My first Linux install was done by purchasing a case of 100 1.4M floppies and filling those at the ISP where I worked,
[28-Aug-2009 14:37:23] <davetoo> then carting them home to do the install
[28-Aug-2009 14:37:45] <djack> rmatte we made a choice at the office and then there was not yet commercial supported ubuntu
[28-Aug-2009 14:37:47] <rmatte> brutality
[28-Aug-2009 14:37:52] <cwj> nostalgia mode engage
[28-Aug-2009 14:38:01] <rmatte> Ubuntu is commercially supported though
[28-Aug-2009 14:38:01] <davetoo> The other thing I miss about perl is the easy interactive debugger
[28-Aug-2009 14:38:07] <rmatte> there's a company that does Ubuntu support
[28-Aug-2009 14:38:08] <davetoo> or maybe I haven't read enough yet
[28-Aug-2009 14:38:09] <djack> i work for a larg retailer so i need to make calls to a servicedesk when needed
[28-Aug-2009 14:38:25] <rmatte> that's why Dell uses Ubuntu on their systems when people want Linux instead of Windows
[28-Aug-2009 14:38:28] <cwj> canonical ltd offers commerical ubuntu support
[28-Aug-2009 14:38:37] <rmatte> yeh, canonical
[28-Aug-2009 14:38:38] <djack> rmatte it is but 7 years ago ....
[28-Aug-2009 14:38:39] <rmatte> that's the one
[28-Aug-2009 14:38:44] <rmatte> ah
[28-Aug-2009 14:38:50] <djack> ubunto is debain based ;-)
[28-Aug-2009 14:39:15] <rmatte> davetoo: debugger: http://docs.python.org/library/pdb.html
[28-Aug-2009 14:39:25] <rmatte> djack: yeh, I know lol
[28-Aug-2009 14:39:28] <cwj> i am just starting to get traction on replacing all our SuSE VMs with Ubuntu VMs
[28-Aug-2009 14:39:35] <cwj> what a glorious day that will be
[28-Aug-2009 14:40:06] <rmatte> yeh
[28-Aug-2009 14:40:07] <davetoo> rmatte: I haven't figured out if/how to fire it up, load the module(s) without running them, and set breakpoints before running
[28-Aug-2009 14:40:22] <rmatte> we're using Ubuntu Server 8.04 on our VMs and it has been 100% smooth sailing
[28-Aug-2009 14:40:25] <rmatte> no issues at all
[28-Aug-2009 14:40:37] <djack> well we have to migrate our point of sales soon so i will have a choice again , when the cash application can run on linux
[28-Aug-2009 14:40:46] <rmatte> davetoo: that page sort of explains that
[28-Aug-2009 14:40:53] <davetoo> 'k
[28-Aug-2009 14:40:57] <cwj> i have been working on a script to deploy new Ubuntu 8.04 VMs using preseed and our VMware infrastructure
[28-Aug-2009 14:41:04] <cwj> almost works
[28-Aug-2009 14:41:40] <rmatte> b(reak) [[filename:]lineno | function[, condition]]
[28-Aug-2009 14:41:42] <rmatte> With a lineno argument, set a break there in the current file.....
[28-Aug-2009 14:41:53] <rmatte> that's what you're looking for
[28-Aug-2009 14:42:05] <davetoo> probably
[28-Aug-2009 14:43:54] <djack> btw i have problems (time-out) when loading the zenpack.hp.mibs and zenpack.cisco.mibs ?
[28-Aug-2009 14:44:13] <djack> i tried to do it via the web interface and from the command line it self
[28-Aug-2009 14:44:21] <rmatte> djack: you're trying to install from the web UI aren't you?
[28-Aug-2009 14:44:41] <rmatte> the cisco mibs will install fine from the commandline, but not from the UI
[28-Aug-2009 14:44:47] <rmatte> and they take about 5-6 hours to install
[28-Aug-2009 14:44:50] <davetoo> sheesh
[28-Aug-2009 14:44:55] <rmatte> so you're probably just not being patient enough
[28-Aug-2009 14:45:13] <rmatte> are you familliar with the 'screen' command?
[28-Aug-2009 14:45:14] <djack> rmatte what ....
[28-Aug-2009 14:45:20] <djack> yeap
[28-Aug-2009 14:45:33] <rmatte> run the install inside a screen, so if your ssh connection dies it'll still install
[28-Aug-2009 14:45:36] <davetoo> I just did an snmptranslate of all the cisco mibs that NX-OS uses. Over 50k oids
[28-Aug-2009 14:45:39] <rmatte> your ssh connection is probably timing out
[28-Aug-2009 14:45:49] <rmatte> I remember having to run it inside a screen for it to finish
[28-Aug-2009 14:46:04] <davetoo> I do almost nothing *outside* of a screen session
[28-Aug-2009 14:46:17] <davetoo> screen is my emacs
[28-Aug-2009 14:46:19] <djack> rmatte i will run it from a script
[28-Aug-2009 14:46:21] <rmatte> so the actual install inside a screen session is timing out?
[28-Aug-2009 14:46:25] <rmatte> I've never ever seen that before
[28-Aug-2009 14:46:54] <davetoo> rmatte: I think you confused my comment for a response from djack
[28-Aug-2009 14:46:56] <rmatte> djack: install "screen"
[28-Aug-2009 14:46:57] <djack> did forget to install screen and pfff access from my sofa to the server in Holland is difficult
[28-Aug-2009 14:47:04] <rmatte> and use a screen session to install them
[28-Aug-2009 14:47:21] <rmatte> and it'll take like 5-6 hours for the cisco pack
[28-Aug-2009 14:47:27] <rmatte> no idea how long for the HP pack
[28-Aug-2009 14:47:37] <rmatte> also, once they are installed use the zendmd commands on those zenpack pages
[28-Aug-2009 14:47:43] <rmatte> otherwise your UI performance will suck
[28-Aug-2009 14:47:52] <rmatte> took me forever to figure out that fix
[28-Aug-2009 14:48:30] <djack> rmatte it's a f*cking hp proliant 360 G6 with 2 dual quad core cpu's and 16 gig of memory ...
[28-Aug-2009 14:48:49] <djack> made a error ordering the server hihi
[28-Aug-2009 14:49:22] <djack> rmatte do you mean reindex() and commit() stuff
[28-Aug-2009 14:49:36] <davetoo> how is your storage configured?
[28-Aug-2009 14:49:59] <rmatte> lol
[28-Aug-2009 14:50:10] <rmatte> djack: no
[28-Aug-2009 14:50:14] <djack> 3 sas 15 k disc in Raid 5
[28-Aug-2009 14:50:19] <rmatte> djack: go to the actual pages where you download those zenpacks
[28-Aug-2009 14:50:21] <djack> rmatte please enlight me
[28-Aug-2009 14:50:23] <rmatte> there are zendmd commands listed
[28-Aug-2009 14:50:35] <rmatte> after the packs are installed you have to execute those
[28-Aug-2009 14:50:49] <rmatte> it basically removes the zenpack from the zenpack list after install
[28-Aug-2009 14:50:56] <rmatte> which eliminates the performance issues
[28-Aug-2009 14:52:11] <djack> i see
[28-Aug-2009 14:54:32] <jb> hrm
[28-Aug-2009 14:54:40] <jb> why isnt my evt.message transform appearing in emails?
[28-Aug-2009 14:54:41] <jb> :/
[28-Aug-2009 14:57:54] <jb> hey egor/rocket?
[28-Aug-2009 14:59:15] <jb> ah nevermind!
[28-Aug-2009 14:59:23] <jb> yo rmatte
[28-Aug-2009 14:59:30] <rocket> rmatte: but can you add aix, solaris and hpux to the mix?
[28-Aug-2009 14:59:31] <rocket>
[28-Aug-2009 15:00:21] <rocket> rmatte: heh and I am one of the gentoo devs who wrote genkernel and catalyst ..
[28-Aug-2009 15:01:09] <cwj> does drobbins still work on gentoo? i used to work with him on stampede a long long time ago
[28-Aug-2009 15:01:36] <cwj> i guess not
[28-Aug-2009 15:02:55] <rmatte> jb: sup?
[28-Aug-2009 15:03:09] <jb> do you remember that thread on the forums about the guy asking how to do multi-line transforms?
[28-Aug-2009 15:03:16] <rmatte> rocket: yeh, mrayzenoss told me
[28-Aug-2009 15:03:20] <rmatte> rocket: props for that
[28-Aug-2009 15:03:38] <rmatte> rocket: I actually remember seeing your email address floating around back in the day, just didn't know you were the same rocket
[28-Aug-2009 15:03:46] <rmatte> rocket@gentoo.org
[28-Aug-2009 15:04:00] <rmatte> jb: yeh, I do
[28-Aug-2009 15:04:01] <rocket> same one
[28-Aug-2009 15:04:14] <jb> remember what it was called?
[28-Aug-2009 15:04:18] <jb> or the URL?
[28-Aug-2009 15:04:26] <jb> finally figured out how to do it
[28-Aug-2009 15:05:08] <rmatte> hmmm, I'll see if I can find it
[28-Aug-2009 15:06:07] <jb> yeah i can't seem to locate it
[28-Aug-2009 15:06:48] <jb> http://pastebin.com/m3eb2c056
[28-Aug-2009 15:06:52] <jb> (an email alert)
[28-Aug-2009 15:08:27] <rmatte> jb: what's the code for the transform look like?
[28-Aug-2009 15:08:32] <jb> sec
[28-Aug-2009 15:08:48] <jb> http://pastebin.com/m79cc0e03
[28-Aug-2009 15:09:13] <rmatte> quite honestly, just post the code as a new forum post, and name it "Transform to create event message with multiple lines"
[28-Aug-2009 15:09:25] <rmatte> I'm sure other people will be interested in knowing
[28-Aug-2009 15:09:36] <rmatte> and whoever was originally looking for a solution will probably come across it eventually
[28-Aug-2009 15:09:59] <rmatte> ah, so it was newlines
[28-Aug-2009 15:10:05] <jb> yeah
[28-Aug-2009 15:10:09] <rmatte> I wasn't sure whether that would actually work or not
[28-Aug-2009 15:10:12] <jb> just had it formatted incorrectly
[28-Aug-2009 15:10:16] <rmatte> I remember testing that in zendmd and it worked fine
[28-Aug-2009 15:10:21] <jb> updating my transforms now
[28-Aug-2009 15:10:22] <rmatte> ah
[28-Aug-2009 15:10:42] <rmatte> yeh, you had it formatted on multiple lines
[28-Aug-2009 15:10:46] <rmatte> but that usually gets quite messy
[28-Aug-2009 15:11:07] <rmatte> good stuff though
[28-Aug-2009 15:11:30] <rmatte> by the way, are you running 2.4.3?
[28-Aug-2009 15:11:41] <jb> nope
[28-Aug-2009 15:11:43] <jb> 2.4.2
[28-Aug-2009 15:11:48] <rmatte> ah
[28-Aug-2009 15:12:05] <rmatte> I need someone to test a bug I filed in 2.4.3 to see if they see the same problem: http://dev.zenoss.org/trac/ticket/5473
[28-Aug-2009 15:12:11] <rmatte> I've patched my 2.4.3 obviously
[28-Aug-2009 15:12:17] <rmatte> it actually shows as 2.4.4 right now hehe
[28-Aug-2009 15:13:12] <rocket> rmatte: nooooooooooooooooo
[28-Aug-2009 15:13:18] <rocket> no patching ....
[28-Aug-2009 15:13:18] <rmatte> ?
[28-Aug-2009 15:13:22] <rocket>
[28-Aug-2009 15:13:23] <rmatte> why?
[28-Aug-2009 15:13:27] <rocket> dont do it man ...
[28-Aug-2009 15:13:33] <rocket> just pulling your leg ...
[28-Aug-2009 15:13:36] <rmatte> I figured
[28-Aug-2009 15:13:37] <rmatte>
[28-Aug-2009 15:13:40] <djack> lol
[28-Aug-2009 15:13:49] * rmatte fwaps rocket
[28-Aug-2009 15:14:11] * rocket bends over ... oops I dont think you wanted to fwap that ...
[28-Aug-2009 15:14:30] * rmatte turns away in disgust
[28-Aug-2009 15:14:34] <djack> lol
[28-Aug-2009 15:14:36] <rocket> heh ...
[28-Aug-2009 15:14:37] <rhett> what the hell did i just walk into
[28-Aug-2009 15:14:46] <djack> i brings back memories
[28-Aug-2009 15:14:50] <rmatte> rhett: still trying to figure that out for myself
[28-Aug-2009 15:15:21] <rocket> a discussion with someone who has had way to little sleep after fixing 9 oracle/sap databases yesterday
[28-Aug-2009 15:15:44] <rmatte> I spent all day yesterday patching and verifying our Zenoss boxes
[28-Aug-2009 15:16:09] <rmatte> not quite as fun as fixing databases
[28-Aug-2009 15:16:38] <rocket> well it was more like the volume groups after an upgrade by a junior admin were hosed
[28-Aug-2009 15:16:47] <rmatte> ah
[28-Aug-2009 15:17:11] <rocket> not fun to see a missing disk on the big volume groups where everything is striped across it ..
[28-Aug-2009 15:17:12] <rmatte> junior admin = PFY, rocket = BOFH
[28-Aug-2009 15:17:42] * rocket deletes rmattes homedir giving him more space ...
[28-Aug-2009 15:17:45] <rhett> your guys' days were much more eventful than mine
[28-Aug-2009 15:17:54] <rmatte> hehe
[28-Aug-2009 15:18:22] <rmatte> you guys have updated your linux kernels I hope?
[28-Aug-2009 15:18:28] <rhett> i just worked on some cacti templates and worked on my car
[28-Aug-2009 15:18:31] <rmatte> that newly discovered exploit is brutal
[28-Aug-2009 15:18:32] * rocket is starting to checkin his homedir into git ... slowly and surely .. figuring out all of its lovely pitfalls
[28-Aug-2009 15:19:01] <rocket> rmatte: hadnt paid attention lately... whats going on with the kernel now?
[28-Aug-2009 15:19:26] <rocket> then again everything I have is behind a firewall at home ... and we only have 2 linux boxen at work ..
[28-Aug-2009 15:19:40] <rmatte> rocket: http://blog.iweb.com/en/2009/08/critical-security-vulnerability-affecting-linux-servers/2923.html
[28-Aug-2009 15:19:42] <rocket> unless you count the hmc's which are up to IBM to send us fixes for
[28-Aug-2009 15:19:43] <rmatte> it's a good one
[28-Aug-2009 15:19:46] <rhett> just checked yum. no updates
[28-Aug-2009 15:19:51] <rmatte> affects all kernels back to 2001
[28-Aug-2009 15:20:26] <rmatte> http://archives.neohapsis.com/archives/fulldisclosure/2009-08/0174.html
[28-Aug-2009 15:20:31] <rmatte> that's a more detailed description
[28-Aug-2009 15:21:20] <rmatte> the exploit requires the user to have access to the box, but it's basically a really quick way for a standard user to gain root privileges
[28-Aug-2009 15:21:43] <rocket> well if there is a hack for ssh or openvpn I am screwed
[28-Aug-2009 15:22:09] <rocket> otherwise there is no way anyone gets close to my boxes ... those are the only two ports open externally at all
[28-Aug-2009 15:22:48] * rocket wonders if the suse enterprise boxes have a fix for them ...
[28-Aug-2009 15:23:00] <rocket> oh well .. they arent production yet ...
[28-Aug-2009 15:23:13] <rmatte> hehe
[28-Aug-2009 15:24:07] <rmatte> anyone running 2.4.3, can you test and confirm: http://dev.zenoss.org/trac/ticket/5473
[28-Aug-2009 15:24:37] <rmatte> rocket: that old ssh hack combined with this would be brutal
[28-Aug-2009 15:24:42] <rhett> 2.4.2 here
[28-Aug-2009 15:24:42] <rmatte> but ssh is pretty rock solid now
[28-Aug-2009 15:25:47] <djack> hmmm question
[28-Aug-2009 15:25:59] <jb> my zen boxes are not exposed to the Internet at all..
[28-Aug-2009 15:26:01] <djack> i deleted the localhost device .... can it be recoved?
[28-Aug-2009 15:27:11] <rmatte> djack: you mean you deleted the actual collector I assume?
[28-Aug-2009 15:27:19] <rmatte> since "localhost" doesn't show up in the device list
[28-Aug-2009 15:27:22] <rocket> rmatte: fine .. make me setup port knocking too ...
[28-Aug-2009 15:27:27] <rocket> rmatte: be that way ...
[28-Aug-2009 15:27:32] <rmatte> lol
[28-Aug-2009 15:27:43] <djack> rmatte it was in my list
[28-Aug-2009 15:27:51] <rmatte> djack: well it shouldn't have been
[28-Aug-2009 15:28:20] <djack> well then it is good that i deleted it.
[28-Aug-2009 15:28:21] <rmatte> djack: deleting localhost from your device list won't cause issues
[28-Aug-2009 15:28:34] <rmatte> deleting the actual collector from the collector list would
[28-Aug-2009 15:28:35] <djack> well i can remember why it was in the list
[28-Aug-2009 15:28:44] <rmatte> why?
[28-Aug-2009 15:28:57] <djack> to test the apachemonitor package
[28-Aug-2009 15:29:01] <rmatte> ah
[28-Aug-2009 15:29:10] <djack> I'm going to set it up again
[28-Aug-2009 15:29:11] <rmatte> well, removing it won't cause issues, so you're fine
[28-Aug-2009 15:29:56] <djack> oki then
[28-Aug-2009 15:31:22] <djack> hmmm my fping is not graphing ...
[28-Aug-2009 15:31:58] <rmatte> this started just recently or it's been not graphing for a while?
[28-Aug-2009 15:32:46] <djack> 20 minutes ago
[28-Aug-2009 15:33:10] <djack> i added the template to a machine see the stuff in the perf tab but no data
[28-Aug-2009 15:34:23] <rmatte> well it usually takes like 15-20 mins for it to start graphing data
[28-Aug-2009 15:34:30] <rmatte> if you do a zencommand run -d <device>
[28-Aug-2009 15:34:34] <rmatte> do you see it run?
[28-Aug-2009 15:34:37] <rmatte> actually...
[28-Aug-2009 15:34:45] <rmatte> zencommand run -v10 -d <device>
[28-Aug-2009 15:34:57] <jb> hrm.. how is it that you fix the links in email alerts again? mine all have http://localhost/
[28-Aug-2009 15:34:59] <djack> can you combine zencommand with snmp on a device ?
[28-Aug-2009 15:35:12] <rmatte> djack: well, what do you mean by combine?
[28-Aug-2009 15:35:28] <rmatte> you can have a template with as many different types of datasources as you want in it
[28-Aug-2009 15:35:58] <rmatte> you can collect via custom script, WMI, SNMP, and SSH all at the same time from the same template
[28-Aug-2009 15:36:18] <djack> oki
[28-Aug-2009 15:36:53] <djack> my first thing i going to create myself is a check of our sap servers
[28-Aug-2009 15:37:22] <djack> like zenoss more and more
[28-Aug-2009 15:41:58] <kgoedtel> jb: zenactions.conf add a line that says- zopeurl http://www.example.com:8080
[28-Aug-2009 15:48:33] <rmatte> djack: yeh, once you start learning how to bend it to your will it gets pretty neat
[28-Aug-2009 15:50:01] <djack> haha i know why graphs are not made
[28-Aug-2009 15:50:18] <djack> zenoss does not have enough right to access the fping command
[28-Aug-2009 15:51:54] <djack> ssst the big boss just walked in ;-0
[28-Aug-2009 15:52:12] <davetoo> he's just the PIO
[28-Aug-2009 15:53:38] <rmatte> he's the pilot-induced oscillation?
[28-Aug-2009 15:53:54] <rmatte> or the programmed input/output?
[28-Aug-2009 15:56:12] <rmatte>
[28-Aug-2009 15:58:59] <jb> %(eventUrl)s
[28-Aug-2009 15:59:03] <jb> anyone know where to update this?
[28-Aug-2009 15:59:18] <jb> its "localhost" right now which obviously doesn't allow the links in the email alerts to work..
[28-Aug-2009 16:00:03] <rmatte> You're looking for the "zopeurl" option to the zenactions daemon. Edit $ZENHOME/etc/zenactions.conf and add a line like the following. Then restart zenactions
[28-Aug-2009 16:00:03] <rmatte> zopeurl http://your.server:8080
[28-Aug-2009 16:00:09] <jb> also trying to figure out how to properly apply CSS to the email alerts
[28-Aug-2009 16:00:11] <rmatte> got that from a post
[28-Aug-2009 16:00:15] <jb> (so I can change the font)
[28-Aug-2009 16:01:29] <rmatte> can't help with the CSS, no idea
[28-Aug-2009 16:01:58] <jb> cool.. thanks
[28-Aug-2009 16:02:25] <rmatte> np
[28-Aug-2009 16:04:27] <kgoedtel> i already answered that question!
[28-Aug-2009 16:04:29] <kgoedtel> haha
[28-Aug-2009 16:04:39] <rmatte> ah, I just saw that
[28-Aug-2009 16:04:46] <rmatte> I was away from my computer for a bit so didn't notice
[28-Aug-2009 16:04:56] <jb> oh
[28-Aug-2009 16:04:58] <jb> crap, i missed it too
[28-Aug-2009 16:05:02] <kgoedtel>
[28-Aug-2009 16:05:03] <jb> kgoedtel: thanks!
[28-Aug-2009 16:05:17] <kgoedtel> np lol
[28-Aug-2009 16:05:25] <jb> i wonder if there is a main template stored somewhere in a flat file
[28-Aug-2009 16:05:41] <jb> i can't add CSS directly to the alert format in the GUI
[28-Aug-2009 16:10:30] <rmatte> you could always just create some sort of script that creates and sends a CSS formatted email
[28-Aug-2009 16:10:42] <rmatte> and add it under the commands tab in the Event Manager section
[28-Aug-2009 16:11:26] <rmatte> Here's an example of a command that we use to call on a script to create tickets in our ticketing system...
[28-Aug-2009 16:11:47] <rmatte> /mspfiles/scripts/create_ticket.pl -autonotify -customer 'gen-00025' -device '${dev/id}' -deviceIP '${dev/manageIp}' -collector `hostname -f` -first '${evt/firstTime}' -last '${evt/lastTime}' -count '${evt/count}' -summary '${evt/summary}' -noteTitle 'System Monitor Error' -note '${evt/message}' -severity '${evt/severity}' -group '${evt/DeviceGroups}' -impact '${evt/DevicePriority}' -component '${evt/component}' -queue 'Front Line' >> ~/APIOutput.txt
[28-Aug-2009 16:11:57] <rmatte> you could do something similar but have the script send an email
[28-Aug-2009 16:16:22] <djack> grmbl just lost on ebay
[28-Aug-2009 16:16:47] <rmatte> lost for what?
[28-Aug-2009 16:16:57] <djack> a nas for home network
[28-Aug-2009 16:17:20] <djack> went for 76 euro ex discs
[28-Aug-2009 16:20:32] <rmatte> ah
[28-Aug-2009 16:20:49] <rmatte> what brand?
[28-Aug-2009 16:21:25] <djack> something unknown
[28-Aug-2009 16:21:47] <djack> GIGABIT DUAL NAS RAID SYSTEM
[28-Aug-2009 16:21:56] <rmatte> ah
[28-Aug-2009 16:22:33] <djack> something small for at home could use 1 T space that is mirrored
[28-Aug-2009 16:23:52] <rmatte> I want the Synology DS-207+
[28-Aug-2009 16:24:28] <jb> i have a readynas pro
[28-Aug-2009 16:24:29] <jb> 6bay
[28-Aug-2009 16:24:29] <rmatte> it has LAMP support, Built in BitTorrent, NFS support
[28-Aug-2009 16:24:39] <jb> ^^ it has all of that..
[28-Aug-2009 16:24:44] <jb> and i can write at like 100
[28-Aug-2009 16:24:47] <jb> 100MB/sec to it
[28-Aug-2009 16:24:48] <rmatte> nice
[28-Aug-2009 16:24:49] <jb> its pricey though
[28-Aug-2009 16:24:57] <rmatte> I don't need a 6bay though
[28-Aug-2009 16:25:00] <rmatte> that's just too much
[28-Aug-2009 16:25:04] <rmatte> 2 bay would be perfect
[28-Aug-2009 16:25:08] <jb> yeah im only using 4 right now
[28-Aug-2009 16:25:17] <jb> but its easy to auto-expand the "raid5" containers..
[28-Aug-2009 16:25:17] <rocket> rmatte: the qnap looks nice ...
[28-Aug-2009 16:25:17] <rmatte> 2 1 terabyte drives mirrored
[28-Aug-2009 16:25:31] <jb> i have 4 x 1.5's right no
[28-Aug-2009 16:25:32] <jb> now
[28-Aug-2009 16:25:45] <davetoo> poo
[28-Aug-2009 16:25:50] <davetoo> another place ZMI is broken
[28-Aug-2009 16:25:52] <rmatte> http://ncix.com/products/?sku=39256&vpn=SS-439-Pro&manufacture=QNAP%20Systems%20Inc.
[28-Aug-2009 16:25:57] <rmatte> that one looks nice
[28-Aug-2009 16:26:00] <rocket> jb: how is the ready nas?
[28-Aug-2009 16:26:12] <rmatte> wow, qnap has an 8 bay lol
[28-Aug-2009 16:26:27] <rmatte> it's over a grand canadian though
[28-Aug-2009 16:26:37] <rocket> jb: ouch ready nas pro is expensive ...
[28-Aug-2009 16:26:57] <jb> yeah it is
[28-Aug-2009 16:27:07] <jb> its really overkill for home usage
[28-Aug-2009 16:27:13] <jb> unless you are an enthusiast
[28-Aug-2009 16:27:19] <davetoo> I need to build a fileserver for the house
[28-Aug-2009 16:27:20] <jb> i paid i think $1100 for the barebones
[28-Aug-2009 16:27:29] <rmatte> I still think the Synology NASes look good for the price
[28-Aug-2009 16:27:33] * rocket needs jb's job ...
[28-Aug-2009 16:27:37] <jb> haha
[28-Aug-2009 16:27:48] <davetoo> I wish I could take something home from work; I work for one of the *big* nas/san mfrs
[28-Aug-2009 16:27:50] <jb> it runs debian
[28-Aug-2009 16:27:57] <jb> its really just a linux box.. with hot swappable bays, etc
[28-Aug-2009 16:28:04] <rocket> the qnap runs some flavor of linux ...
[28-Aug-2009 16:28:14] <jb> has lots of features built in
[28-Aug-2009 16:28:53] <rmatte> I want a NetApp for home
[28-Aug-2009 16:28:55] <jb> similar to the qnaps
[28-Aug-2009 16:32:53] <djack> rmatte we have 2 netapps at work....
[28-Aug-2009 16:32:56] <djack> they suck
[28-Aug-2009 16:36:48] <cwj> sorry, does not compute
[28-Aug-2009 16:37:16] <cwj> only netapp ive used that sucked was 8+ years old
[28-Aug-2009 16:39:16] <davetoo> man...
[28-Aug-2009 16:39:33] <davetoo> something about the factory stuff, I think, is creating inconsistent objects
[28-Aug-2009 16:39:50] <davetoo> I can't access ipRouteEntry objects from the ZMI
[28-Aug-2009 16:40:00] <davetoo> nexthopip property is missing
[28-Aug-2009 16:40:04] <mrayzenoss> rmatte: mrchippy checked out 5473 and said it's not a regression, at least not since 2.4.2
[28-Aug-2009 16:45:28] <rmatte> ok...
[28-Aug-2009 16:45:43] <rmatte> but were you able to reproduce it in 2.4.3?
[28-Aug-2009 16:47:00] <rmatte> seems like an odd issue to suddenly spring up
[28-Aug-2009 16:47:34] <davetoo> feh
[28-Aug-2009 16:47:39] <davetoo> I left a pair of shoes in Annapolis
[28-Aug-2009 16:47:44] <rmatte> lol
[28-Aug-2009 16:48:00] <davetoo> and the fine people at Holiday In Expre$$ won't ship them to me on their dime
[28-Aug-2009 16:48:03] <davetoo> such service
[28-Aug-2009 16:48:07] <davetoo> Inn
[28-Aug-2009 16:48:28] <davetoo> the lame thing is... two of the guys in training got rooms at the Westin for less than what I paid for HI Express
[28-Aug-2009 16:48:30] <davetoo> !!!
[28-Aug-2009 16:49:00] <mrayzenoss> dang, the Westin isn't cheap
[28-Aug-2009 16:49:17] <davetoo> They said they paid less than the $107 we paid at HI Express
[28-Aug-2009 16:49:24] <mrayzenoss> rmatte: yeah, Chip said he tested on 2.4.3 and 2.4.2
[28-Aug-2009 16:49:25] <rmatte> brutal
[28-Aug-2009 16:49:37] <davetoo> Some special popped up on some website, I guess.
[28-Aug-2009 16:49:38] <rmatte> mrayzenoss: so it is reproduceable
[28-Aug-2009 16:49:41] <rmatte> ?
[28-Aug-2009 16:49:53] <mrayzenoss> davetoo: seriously? I paid $159 a night...
[28-Aug-2009 16:49:56] <mrayzenoss> nuts
[28-Aug-2009 16:50:08] <mrayzenoss> rmatte: definitely reproducible, just not a regression
[28-Aug-2009 16:50:19] <rmatte> mrayzenoss: ok, but when will it be fixed then?
[28-Aug-2009 16:50:35] <rmatte> It's causing major issues for us from an inventory perspective
[28-Aug-2009 16:50:40] <rmatte> and when generating report info
[28-Aug-2009 16:50:59] <djack> going to bed
[28-Aug-2009 16:51:04] <mrayzenoss> later
[28-Aug-2009 16:51:11] <mrayzenoss> rmatte: no idea
[28-Aug-2009 16:51:27] <mrayzenoss> hardening iterations are starting now, so features are mostly frozen for King Crab
[28-Aug-2009 16:51:37] <rmatte> eugh, crappy
[28-Aug-2009 16:51:43] <mrayzenoss> so defects should be getting all the attention
[28-Aug-2009 16:52:02] <mrayzenoss> rmatte: the good news is that's not a feature, it's a bug
[28-Aug-2009 16:52:02] <rmatte> so we could have to go over 2 months without that
[28-Aug-2009 16:52:11] <rmatte> yeh
[28-Aug-2009 16:52:16] <rmatte> I'd say so, it worked before
[28-Aug-2009 16:52:21] <rmatte> it's a lack of a feature if anything
[28-Aug-2009 16:52:22] <rmatte> lol
[28-Aug-2009 16:53:03] <rmatte> I don't suppose any report fixes will be making their way in to King Crab?
[28-Aug-2009 16:53:31] <rmatte> such as exports that actually work?
[28-Aug-2009 16:53:48] <mrayzenoss> no idea, I wasn't in the iteration planning meeting, I just know the next 2 weeks are for getting the beta ready
[28-Aug-2009 16:53:56] <rmatte> k
[28-Aug-2009 16:54:12] <mrayzenoss> and Development will be fixing bugs until KC is frozen
[28-Aug-2009 16:55:54] <rmatte> I'd love for them to fix the report bugs so that exports work, it's absolutely killing me right now because instead of doing stuff that I should be doing I'm having to write python scripts to dump report data
[28-Aug-2009 16:55:58] <mrayzenoss> here's the open queue: http://dev.zenoss.org/trac/report/3
[28-Aug-2009 16:56:10] <rmatte> and I just found out that my interface utilization script is broken when I thought it was working properly
[28-Aug-2009 16:56:17] <rmatte> so it's back to the drawing board on Monday for that one
[28-Aug-2009 16:56:30] <mrayzenoss> so 11 priority 1s for King Crab right now, those get done first
[28-Aug-2009 16:57:10] <mrayzenoss> 5453 looks like it touches reporting, maybe someone will be digging into those
[28-Aug-2009 16:58:16] <rmatte> the whole report system needs someone to go over it with a fine toothed comb
[28-Aug-2009 16:58:23] <rmatte> because it is definitely not Production grade at this point
[28-Aug-2009 16:58:30] <rmatte> whereas everything else in Zenoss is
[28-Aug-2009 16:59:05] <davetoo> mrayzenoss: doesn't anybody besides me use the ZMI for debugging/education/development?
[28-Aug-2009 16:59:09] <rmatte> I'd rather have working reports before a new UI
[28-Aug-2009 16:59:09] <mrayzenoss> yeah, I think it's currently high on the overhaul list for the next release, which isn't to say lots of the defects against it won't be fixed
[28-Aug-2009 16:59:27] <davetoo> sorry, I'll let you guys finish this thread
[28-Aug-2009 16:59:48] <rmatte> davetoo: the zope management interface?
[28-Aug-2009 16:59:52] <rmatte> I use it all the time, why?
[28-Aug-2009 17:00:04] <davetoo> try looking at, well,
[28-Aug-2009 17:00:13] <davetoo> try looking at the properties for /Devices, for instance
[28-Aug-2009 17:00:26] <davetoo> or properties for an IpRoute
[28-Aug-2009 17:00:36] <rmatte> what about it?
[28-Aug-2009 17:00:52] <davetoo> for me, it barfs,
[28-Aug-2009 17:01:14] <mrayzenoss> davetoo: hmm... me too
[28-Aug-2009 17:01:16] <rmatte> ok, well, by "look at properties" what are you doing exactly?
[28-Aug-2009 17:01:20] <rmatte> I'm just going to: http://lab01.novanoc.com:8180/zport/dmd/Devices/manage
[28-Aug-2009 17:01:31] <davetoo> in the zmi, select the "Properties" tab on the right
[28-Aug-2009 17:01:36] <rmatte> oh right
[28-Aug-2009 17:01:39] <davetoo> mrayzenoss: I have a case and a ticket open,
[28-Aug-2009 17:01:47] <rmatte> yeh, same here
[28-Aug-2009 17:01:58] <davetoo> for the missing devtypes properties
[28-Aug-2009 17:02:13] <rmatte> looks like there's a bug in the getProperty function in OFS.PropertyManager module
[28-Aug-2009 17:02:28] <davetoo> I'm not sure where the problem is,
[28-Aug-2009 17:02:29] <rmatte> yeh
[28-Aug-2009 17:02:37] <davetoo> but seems like the Factory stuff isn't working right
[28-Aug-2009 17:02:45] <davetoo> because something is creating inconsistent objects
[28-Aug-2009 17:02:59] <davetoo> I found it in zendmd (essentially)
[28-Aug-2009 17:03:17] <davetoo> because I have some scripts that depend upon propertyItems() to dump stuff to a zenpack
[28-Aug-2009 17:03:30] <rmatte> OHHHH
[28-Aug-2009 17:03:36] <rmatte> I just figured out the bug in my script
[28-Aug-2009 17:03:38] <rmatte> stupid me
[28-Aug-2009 17:03:49] <rmatte> I should have it setting percentUsed to 0
[28-Aug-2009 17:03:59] <rmatte> so that if it can't calculate it it doesn't take the value from the last interface
[28-Aug-2009 17:04:09] <rmatte> that's why the info I dumped is all screwed
[28-Aug-2009 17:04:11] <davetoo> and just now, I'm working on stuff to add routes to a devicen and some of the routes are missing the nexthopip property. They're not just empty/null, they are missing.
[28-Aug-2009 17:04:12] <rmatte> everything else is fine
[28-Aug-2009 17:04:40] <davetoo> so I guess I dont' understand what good the _properties stuff does
[28-Aug-2009 17:05:04] <davetoo> I dont' really understand a lot about Zope yet,
[28-Aug-2009 17:05:21] <davetoo> to know if that's from a standard Zope product or a Zenoss thing.
[28-Aug-2009 17:07:45] <rmatte> Zope can be pretty dirty
[28-Aug-2009 17:08:02] <davetoo> it's sure complex
[28-Aug-2009 17:08:09] <davetoo> or complicated
[28-Aug-2009 17:08:20] <davetoo> I forget which is the correct word to use
[28-Aug-2009 17:08:30] <rmatte> yeh, it's eye-bleed inducing
[28-Aug-2009 17:11:03] <mrayzenoss> my understanding is that Zope 2.11.2 in King Crab is some steps towards less ocular bleeding
[28-Aug-2009 17:16:58] <rmatte> I think I'm going to spend some time on Monday writing up a detailed review/report on the Zenoss reporting system which highlights improvements that could be made, or bugs that could be addressed. There honestly aren't that many, but they are biggies.
[28-Aug-2009 17:18:37] <mrayzenoss> that would be appreciated
[28-Aug-2009 17:19:10] <rmatte> I'll make it as professional as possible so that it's something you could bring to a meeting or something.
[28-Aug-2009 17:19:18] <rmatte> with screenshots and such
[28-Aug-2009 17:19:25] <rmatte> anyways, I'm out, have a good weekend guys
[28-Aug-2009 17:19:33] <mrayzenoss> you too, later
[28-Aug-2009 17:26:56] <davetoo> hmm
[28-Aug-2009 17:27:06] <davetoo> how can I debug page template calls?
[28-Aug-2009 17:36:20] <davetoo> uh oh
[28-Aug-2009 17:36:22] <davetoo> heh
[28-Aug-2009 17:37:20] <davetoo> I'm not sure that making ZenModel refreshable in zope is a good thing
[28-Aug-2009 17:48:23] <davetoo> ok, well,
[28-Aug-2009 17:48:30] <davetoo> I managed to figure out how to add ip routes with zendmd
[28-Aug-2009 17:48:40] <davetoo> I'm done with Zenoss for the day
[28-Aug-2009 17:48:42] <davetoo> brain hurts
[28-Aug-2009 17:49:20] <davetoo> I'll be having nightmares tonight about mibs and zope page templates and exceptions and...
[28-Aug-2009 18:33:06] <davetoo> what the heck
[28-Aug-2009 18:33:17] <davetoo> why isn't zendisc cycliing through all of my snmp strings?
[28-Aug-2009 18:33:51] <davetoo> n/mind
[28-Aug-2009 18:34:16] <davetoo> I guess it is, it just isn't running zenmodel
[28-Aug-2009 18:34:18] <davetoo> odd
[29-Aug-2009 00:00:46] [disconnected at Sat Aug 29 00:00:46 2009]
[29-Aug-2009 00:00:46] [connected at Sat Aug 29 00:00:46 2009]
[29-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[29-Aug-2009 02:05:51] <davetoo> uh oh
[29-Aug-2009 13:07:15] <stc_s> anyone knows about zenping problems? and to resolve it to more 300 devices or more?
[29-Aug-2009 13:07:29] <stc_s> I have a serious problem with zenoss
[29-Aug-2009 13:09:22] <stc_s> help
[29-Aug-2009 16:59:19] <etank> is this page still accurate? http://www.zenoss.com/community/docs/howtos/zenbatchload-howto
[29-Aug-2009 17:05:28] <etank> looks like the -i is no longer valid
[29-Aug-2009 17:05:34] <davetoo> hmm
[29-Aug-2009 17:05:56] <davetoo> rtfc
[29-Aug-2009 17:06:18] <davetoo> which version?
[29-Aug-2009 20:10:15] kevin7kal_ is now known as kevin7kal
[29-Aug-2009 23:14:50] <davetoo> damn cisco
[30-Aug-2009 00:00:46] [disconnected at Sun Aug 30 00:00:46 2009]
[30-Aug-2009 00:00:46] [connected at Sun Aug 30 00:00:46 2009]
[30-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[30-Aug-2009 16:15:40] <djack> hello world
[31-Aug-2009 00:00:46] [disconnected at Mon Aug 31 00:00:46 2009]
[31-Aug-2009 00:00:46] [connected at Mon Aug 31 00:00:46 2009]
[31-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[31-Aug-2009 09:10:56] <mrayzenoss> morning, anyone here zenpatch 2.4.3 on Friday?
[31-Aug-2009 09:11:07] <mrayzenoss> rmatte?
[31-Aug-2009 09:11:15] <mrayzenoss> looking for reasons not to push out 2.4.5
[31-Aug-2009 09:20:57] <rmatte> hey
[31-Aug-2009 09:21:06] <rmatte> well, what does 2.4.5 contain?
[31-Aug-2009 09:21:11] <rmatte> patchwise
[31-Aug-2009 09:21:44] <rmatte> just those 2 patches?
[31-Aug-2009 09:22:32] <mrayzenoss> 14744 and 14769
[31-Aug-2009 09:22:37] <rmatte> In my opinion: http://dev.zenoss.org/trac/ticket/5473 should be fixed before next official release since it's a very obvious bug
[31-Aug-2009 09:22:44] <mrayzenoss> I'm looking at the other 2 tickets assigned to 2.4.5
[31-Aug-2009 09:23:04] <rmatte> In addition to not being able to assign product details from the Edit tab, you can't properly add products via the edit tab either
[31-Aug-2009 09:23:05] <mrayzenoss> yeah, but we need to get 2.4.5 out
[31-Aug-2009 09:23:26] <rmatte> ok, well then I suppose, despite that obvious bug, everything else has been running smoothly
[31-Aug-2009 09:24:41] <jb> morning
[31-Aug-2009 09:25:28] <mrayzenoss> 2.4.5 is going to be just the 2 tickets then
[31-Aug-2009 09:25:34] <rmatte> k
[31-Aug-2009 09:26:21] <rmatte> You guys probably shouldn't wait longer than a week or two to get 2.4.6 out though
[31-Aug-2009 09:27:13] <mrayzenoss> well, the KC beta is dropping next week and half our developers are out, so it may be awhile
[31-Aug-2009 09:28:24] <rmatte> that sucks for me, but alright lol
[31-Aug-2009 09:28:25] <mrayzenoss> this week is VMworld and Red Hat Summit, so we're at both
[31-Aug-2009 09:29:48] <rmatte> I'm going to have engineers as well as my boss crawling all over me on that one
[31-Aug-2009 09:29:59] <rmatte> since we generate reports for billing that need to include hardware details
[31-Aug-2009 09:31:56] <jb> who is going to be at RHS?
[31-Aug-2009 09:32:31] <mrayzenoss> I know Mark and someone from marketing, I think one of the professional service engineers
[31-Aug-2009 09:32:47] <mrayzenoss> and sales and our director of alliances
[31-Aug-2009 09:33:13] <mrayzenoss> Mark = VP of Community, my boss
[31-Aug-2009 09:33:33] <jb> yeah
[31-Aug-2009 09:33:42] <jb> i'll be out of town, or I would go..
[31-Aug-2009 09:34:07] <mrayzenoss> it's nice that Red Hat and VMware have their big conferences at exactly the same time...
[31-Aug-2009 09:34:29] <mrayzenoss> so we've got people having to go to both
[31-Aug-2009 09:37:37] <rmatte> lol
[31-Aug-2009 10:10:59] <gwb2351> what's the easiest way to track down why a perf graph is not getting updated in zenoss?
[31-Aug-2009 10:11:10] <gwb2351> right now I 1) check the template for the data source command
[31-Aug-2009 10:11:19] <gwb2351> 2) validate that snmpwalk is returning a value
[31-Aug-2009 10:11:38] <gwb2351> how to 3) figure out why that value is not getting added/plotted?
[31-Aug-2009 10:16:10] <mrayzenoss> zensnmpperf run -v10 -d host
[31-Aug-2009 10:17:00] <mrayzenoss> verify that the template is bound
[31-Aug-2009 10:17:22] <rmatte> well, if the template isn't bound he shouldn't even see the graphs to begin with
[31-Aug-2009 10:17:24] <mrayzenoss> I assume 1) includes using the "Test" command
[31-Aug-2009 10:17:31] <mrayzenoss> agreed
[31-Aug-2009 10:17:40] <mrayzenoss> is that "empty graph" or "no graph"
[31-Aug-2009 10:17:46] <mrayzenoss> no graph means no template
[31-Aug-2009 10:17:58] <rmatte> yeh, like is the graph actually missing, or do you see that graph with "nan" values?
[31-Aug-2009 10:18:01] <gwb2351> well, i know the template is bound because the history of the data shows information, but then it just drops off (with nan) at odd times
[31-Aug-2009 10:18:01] <mrayzenoss> empty graph means something wonky in the data source, or it's not populated yet
[31-Aug-2009 10:18:21] <gwb2351> the graph is there; with some data; but from time to time (for long stretches, like hours) no data gets added
[31-Aug-2009 10:18:29] <rmatte> what is the template monitoring?
[31-Aug-2009 10:18:45] <gwb2351> in this instance, it's the powerware template (for a UPS)
[31-Aug-2009 10:18:56] <rmatte> and it uses command based datapoints?
[31-Aug-2009 10:19:12] <gwb2351> the data source is "/opt/zenoss/libexec/check_snmp -H ${here/manageIp} -C ${here/zSnmpCommunity} -r '1|3|4|5' -c 2 -o 1.3.6.1.4.1.534.1.2.5.0"
[31-Aug-2009 10:19:20] <rmatte> If you run a test of the command against a device, how long does it take for it to run through?
[31-Aug-2009 10:19:21] <gwb2351> which i can run from the command line, no problem, with an actual result
[31-Aug-2009 10:19:25] <gwb2351> <1s
[31-Aug-2009 10:20:15] <gwb2351> SNMP OK - 5 | iso.3.6.1.4.1.534.1.2.5.0=5 , real 0m0.005s
[31-Aug-2009 10:21:26] <gwb2351> the other SNMP values for the template are working/graphing OK, just not this data point
[31-Aug-2009 10:21:28] <mrayzenoss> why are you using check_snmp?
[31-Aug-2009 10:21:31] <rmatte> is this a premade template, from a ZenPack?
[31-Aug-2009 10:21:34] <rmatte> or did you make this?
[31-Aug-2009 10:21:41] <gwb2351> premade from community zenpack
[31-Aug-2009 10:22:47] <gwb2351> http://www.zenoss.com/community/projects/zenpacks/powerware-ups
[31-Aug-2009 10:23:46] <mrayzenoss> yeah, he must have just pulled a nagios script over rather than do it from scratch
[31-Aug-2009 10:24:08] <mrayzenoss> an snmp-based data source would probably be more performant
[31-Aug-2009 10:24:31] <rmatte> yeh, I don't get why they are even using a command based datasource for that
[31-Aug-2009 10:24:33] <mrayzenoss> maybe try adding a parallel snmp-based template that hits the same OIDs and see if the gaps go away
[31-Aug-2009 10:24:38] <rmatte> since they are just pulling straight SNMP values
[31-Aug-2009 10:24:43] <gwb2351> OK, I'll try that
[31-Aug-2009 10:40:28] <mrayzenoss> I gotta go give a demo, back later
[31-Aug-2009 10:40:40] <rmatte> have fun
[31-Aug-2009 10:40:49] <davetoo> heh
[31-Aug-2009 11:06:13] <mrhinkle> FYi...Our Zenoss community manager is presenting a live demo of Zenoss Core and Q&A via Webex you can sign up and ask questions for the next hour - http://forms.zenoss.com/getting-started
[31-Aug-2009 12:27:19] <gemineye> Does anyone know when the next release of Core is schedule for?
[31-Aug-2009 12:28:54] <etank> gemineye: i think i saw last week that it should be this week.
[31-Aug-2009 12:28:57] <etank> 2.4.4 i think
[31-Aug-2009 12:29:14] <jb> 2.4.5
[31-Aug-2009 12:29:17] <jb> today probably..
[31-Aug-2009 12:30:00] <gemineye> so just bug fixes.
[31-Aug-2009 12:30:25] <gemineye> We're at 2.4.2, they skipping some levels?
[31-Aug-2009 12:34:27] <jb> 2.4.3/2.4.4 had some bugs..
[31-Aug-2009 12:34:38] <jb> 2.4.4/2.4.5 just fix a couple of issues
[31-Aug-2009 12:35:17] <rmatte> 2.4.5 is just 2.4.3 with zenpatch 14744 and zenpatch 14769 applied to it
[31-Aug-2009 12:38:39] <etank> if i email support@zenoss.com should it show up as a case?
[31-Aug-2009 12:38:46] <etank> for enterprise that is
[31-Aug-2009 12:40:10] <rmatte> bah, just noticed that they changed the memory report slightly so my customized memory report plugin doesn't work with the .rpt file
[31-Aug-2009 12:40:46] <rmatte> and I didn't back up the .rpt file
[31-Aug-2009 12:40:59] <rmatte> oh well, yet another project for tomorrow
[31-Aug-2009 12:48:30] <davetoo> damn
[31-Aug-2009 12:48:39] <davetoo> the more I work on this the more work I have to do
[31-Aug-2009 13:00:07] <rmatte> davetoo: yup, it's exponential for a while
[31-Aug-2009 13:01:07] <davetoo> well, one o fthe things that left a bad taste in the mouths of the level-1 people, from our previous core installs, is alert floods
[31-Aug-2009 13:05:45] <rmatte> yeh
[31-Aug-2009 13:14:54] <nzle> I wrote a nagios plugin for an internal program we have here and its working fine being called by zencommand via the devices local template, but I only see it on the dashboard, or the device status when somthing is wrong. I'd like to see it when its OK as well.
[31-Aug-2009 13:16:21] <nzle> Same issue with a heartbeat. How can I get a visible indicator of the hearbeat test before it recieves a heartbeat.
[31-Aug-2009 13:24:05] <nzle> ?
[31-Aug-2009 13:31:25] <nzle> Did my question of a few minutes ago get sent since I was not properly identified to to the server yet or should I resend it?
[31-Aug-2009 15:14:57] <rmatte> nzle: ummm, that's kind of pointless no?
[31-Aug-2009 15:15:19] <rmatte> I mean, if you don't see an indicator that something is wrong, then it's obviously ok
[31-Aug-2009 15:16:21] <rmatte> I don't see why you need an indicator specifically telling you something is ok
[31-Aug-2009 15:16:33] <rmatte> you generally just assume everything is ok unless you get an alert indicating otherwise
[31-Aug-2009 15:17:08] <rmatte> or are you talking about having it send a clear command?
[31-Aug-2009 15:17:36] <rmatte> (like, to automatically clear the previous event)
[31-Aug-2009 15:23:30] <davetoo> hmm
[31-Aug-2009 15:23:46] <davetoo> zenpatch 14744
[31-Aug-2009 15:23:53] <davetoo> patching file ZenEvents/zenactions.py
[31-Aug-2009 15:23:53] <davetoo> Hunk #1 FAILED at 273.
[31-Aug-2009 15:23:53] <davetoo> Hunk #2 FAILED at 304.
[31-Aug-2009 15:23:53] <davetoo> Hunk #3 succeeded at 314 with fuzz 1 (offset -13 lines).
[31-Aug-2009 15:23:53] <davetoo> 2 out of 3 hunks FAILED -- saving rejects to file ZenEvents/zenactions.py.rej
[31-Aug-2009 15:24:16] <davetoo> D'oh
[31-Aug-2009 15:24:25] * davetoo reads more closely
[31-Aug-2009 15:24:36] <gwb2351> there is value in having an explicit "everything is OK" value; systems hose themselves hard enough that they're unable to send alerts quite often.
[31-Aug-2009 15:24:39] <rmatte> davetoo: what version of Zenoss are you running?
[31-Aug-2009 15:24:43] <davetoo> the wrong one
[31-Aug-2009 15:24:51] <rmatte> obviously
[31-Aug-2009 15:24:52] <rmatte> lol
[31-Aug-2009 15:25:01] <rmatte> you need to have 2.4.3 installed for those patches to work
[31-Aug-2009 15:25:15] <rmatte> If you don't then just wait for 2.4.5 to be released
[31-Aug-2009 15:25:28] <davetoo> yeah, this is just a development system
[31-Aug-2009 15:25:32] <davetoo> moving to a bigger box
[31-Aug-2009 15:28:03] <rmatte> cool
[31-Aug-2009 15:30:30] <davetoo> my first development box is too slow
[31-Aug-2009 15:30:37] <davetoo> only one disk
[31-Aug-2009 15:30:47] <rmatte> our lab box is pretty darn slow, but it's because we're running like 10 VMs on it
[31-Aug-2009 15:30:59] <rmatte> our production stuff is lightning fast
[31-Aug-2009 15:31:14] <davetoo> I'm using vms for zenpack testing
[31-Aug-2009 15:31:40] <rmatte> we use vms for everything
[31-Aug-2009 15:31:40] <davetoo> but I need a "big" development system to see the impact of a big load of systems
[31-Aug-2009 15:31:48] <davetoo> we're moving that way
[31-Aug-2009 15:31:56] <rmatte> but I develop zenpacks on my lab box and then deploy them out to the other zenoss vms
[31-Aug-2009 15:32:11] <davetoo> so you use zenoss on vms for production?
[31-Aug-2009 15:32:15] <rmatte> yup
[31-Aug-2009 15:32:16] <davetoo> What do you do for storage?
[31-Aug-2009 15:32:21] <rmatte> SAN
[31-Aug-2009 15:32:21] <davetoo> And how many devices?
[31-Aug-2009 15:32:37] <davetoo> Enterprise? or Core?
[31-Aug-2009 15:32:38] <rmatte> it varies, we support multiple clients, so each VM is for a different client
[31-Aug-2009 15:32:42] <rmatte> Core
[31-Aug-2009 15:32:58] <davetoo> ah
[31-Aug-2009 15:33:02] <rmatte> each VM has a VPN tunnel to the client's network
[31-Aug-2009 15:33:19] <rmatte> the largest network we're monitoring right now is only about 370 devices
[31-Aug-2009 15:33:21] <davetoo> Yeah, we'll still have a bunch of core systems to monitor stuff nobody wants to pay to license,
[31-Aug-2009 15:33:23] <rmatte> routers, switches, and windows servers
[31-Aug-2009 15:33:30] <davetoo> but we're bringing up our Enterprise system.
[31-Aug-2009 15:33:43] <rmatte> cool
[31-Aug-2009 15:33:49] <rmatte> we can't afford enterprise at the moment
[31-Aug-2009 15:33:56] <rmatte> and we're making due with core quite well
[31-Aug-2009 15:34:02] <rmatte> we're just having to be ingenious about it
[31-Aug-2009 15:34:03] <rmatte> lol
[31-Aug-2009 15:34:17] <davetoo> We can't afford not to have the support And not having a centralized interface.
[31-Aug-2009 15:34:26] <rmatte> ah
[31-Aug-2009 15:34:31] <davetoo> Finally got some $$ support for monitoring.
[31-Aug-2009 15:34:43] <etank> davetoo: im in the same boat
[31-Aug-2009 15:34:43] <rmatte> I'm pretty good at providing support for our stuff
[31-Aug-2009 15:34:51] <rmatte> but we're probably not as large scale as you guys yet
[31-Aug-2009 15:34:52] <etank> we just did our enterprise install last friday
[31-Aug-2009 15:35:02] <davetoo> 2.4.2 is doing much better than previous versions.
[31-Aug-2009 15:35:14] <rmatte> 2.4.2 has a security bug in it though
[31-Aug-2009 15:35:19] <davetoo> rmatte: well, we started out with 2500 nodes
[31-Aug-2009 15:35:35] <davetoo> but if we get the ESX monitoring done nicely,
[31-Aug-2009 15:35:47] <rmatte> ah
[31-Aug-2009 15:35:47] <davetoo> that will probably explode as other groups adopt it for VMware
[31-Aug-2009 15:35:57] <davetoo> etank: how many collectors?
[31-Aug-2009 15:36:18] <etank> davetoo: right now one. but we are going to work on getting some others out soon.
[31-Aug-2009 15:36:33] <etank> we have some devices that are only available through a vpn connections
[31-Aug-2009 15:36:50] <etank> so we are going to try putting collectors in those areas
[31-Aug-2009 15:36:53] <rmatte> we have 3 blade boxes right now, each with 15GHz processing and 32 gigs of RAM, then we have a SAN (after raiding it can store over 5TB I believe). We have an image of a Zenoss client base that we just clone, change the IPs and hostname on, and it's good to go.
[31-Aug-2009 15:37:05] <rmatte> we can have a new client up and monitored in less than a day
[31-Aug-2009 15:37:15] <rmatte> and all our Zenoss instances tie in to our ticketing system
[31-Aug-2009 15:37:20] <davetoo> oh
[31-Aug-2009 15:37:26] <rmatte> so we don't need a global event console since the ticketing system basically does that
[31-Aug-2009 15:37:32] <davetoo> then you must have done a lot of work on event filtering
[31-Aug-2009 15:37:36] <etank> if things go well then maybe the corporate IT guys will adopt Zenoss as their solution too
[31-Aug-2009 15:37:59] <etank> rmatte: what ticketing system do you use?
[31-Aug-2009 15:38:04] <rmatte> davetoo: I've done a lot of tuning so far, yeh
[31-Aug-2009 15:38:21] <rmatte> etank: a customized version of OTRS
[31-Aug-2009 15:38:42] <davetoo> that's one area we haven't done a good job at, previously,and why some of the groups have a bad taste in their mouth.
[31-Aug-2009 15:38:54] <davetoo> I have got to come up with some kind of summarization and flood protection.
[31-Aug-2009 15:39:22] <rmatte> I still have tuning on my list to do, I basically just do the tuning on my lab box, load it in to a ZenPack, then push the ZenPack upgrade across all servers
[31-Aug-2009 15:39:32] <rmatte> it's the easiest way to keep them all up to date
[31-Aug-2009 15:39:53] <rmatte> there are some items I haven't had time to tune yet this week that engineers are bugging me about
[31-Aug-2009 15:39:57] <davetoo> {how} do you manage dependencies? layer-2 and/or apps?
[31-Aug-2009 15:40:29] <rmatte> we don't do layer 2, and Zenoss handles app dependencies alright, it's a bit off sometimes
[31-Aug-2009 15:40:43] <rmatte> although that's why we have a select all button in our ticketing system hehe
[31-Aug-2009 15:40:46] <davetoo> We have some *big* layer 2 switches
[31-Aug-2009 15:40:59] <rmatte> if a flood of events comes in the front line guys just select a bunch of tickets and relate them back to 1
[31-Aug-2009 15:41:02] <rmatte> then work that 1
[31-Aug-2009 15:41:05] <rmatte> so it's not that bad
[31-Aug-2009 15:41:23] <davetoo> I do not like our ticketing system
[31-Aug-2009 15:41:37] <rmatte> we're doing smaller scale monitoring for the most part, I doubt we'll be monitoring anything over 1000 devices any time soon
[31-Aug-2009 15:42:38] <rmatte> anything over 1000 on one box I mean to say
[31-Aug-2009 15:43:26] <rmatte> davetoo: our ticketing system is a bit off right now too, but that's because we're still doing development work
[31-Aug-2009 15:43:50] <rmatte> we still have a customer portal and a cmdb system being developed. The CMDB is going to contain all contact info, circuit info, etc... for each device
[31-Aug-2009 15:47:39] <rmatte> by the way, if you're using 2.4.2 you probably want to do zenpatch 14181
[31-Aug-2009 15:47:43] <rmatte> if you haven't already
[31-Aug-2009 15:47:48] <rmatte> http://dev.zenoss.org/trac/ticket/5279
[31-Aug-2009 15:48:08] <davetoo> thanks
[31-Aug-2009 15:48:25] <rmatte> nasty security hole
[31-Aug-2009 15:50:37] <rmatte> which is why they are rushing out 2.4.5 even though it has bugs in it.
[31-Aug-2009 15:51:36] <straterra> Lemme see what I have
[31-Aug-2009 15:52:02] <rmatte> it affects any 2.4.x version before 2.4.3
[31-Aug-2009 15:52:04] <straterra> I'm on 2.4.2
[31-Aug-2009 15:52:13] <rmatte> yeh, then apply that patch
[31-Aug-2009 15:52:17] <rmatte> and restart Zenoss
[31-Aug-2009 15:52:21] <straterra> Is it easy to update zenoss?
[31-Aug-2009 15:52:40] <rmatte> depends on how you installed it
[31-Aug-2009 15:52:49] <straterra> rhel rpm
[31-Aug-2009 15:52:53] <rmatte> then yes
[31-Aug-2009 15:53:09] <davetoo> I used to do source installs. Very slow.
[31-Aug-2009 15:53:14] <davetoo> unreliable.
[31-Aug-2009 15:53:17] <rmatte> it's just like doing any other rpm update
[31-Aug-2009 15:53:38] <davetoo> but you have to READ THE INSTRUCTIONS
[31-Aug-2009 15:53:57] <davetoo> Because there are required upgrade paths if you've waited a long time.
[31-Aug-2009 15:54:16] <straterra> So..wheres the rpm I need?
[31-Aug-2009 15:54:18] <rmatte> I started learning unix with FreeBSD when I was 15
[31-Aug-2009 15:54:34] <rmatte> so I've done my share of source installs
[31-Aug-2009 15:54:37] <rmatte> lol
[31-Aug-2009 15:55:17] <rmatte> I think the first version of FreeBSD that I used was at the end of the 4.x series
[31-Aug-2009 15:55:29] <rmatte> or the very start of 5.x
[31-Aug-2009 15:56:20] <rmatte> straterra: there are none yet, 2.4.3 was pulled back, 2.4.2 is the latest version right now
[31-Aug-2009 15:56:34] <straterra> Ah
[31-Aug-2009 15:56:37] <straterra> I can wait then..
[31-Aug-2009 15:56:41] <rmatte> but if you run "zenpatch 14181" as the zenoss user, that'll patch that security hole for the time being
[31-Aug-2009 15:56:44] <straterra> That issue isn't a big deal for me
[31-Aug-2009 15:57:00] <mrayzenoss> as long as you trust your users
[31-Aug-2009 15:57:01] <straterra> our zenoss server is internal only
[31-Aug-2009 15:57:12] <straterra> None of the users know it even exists
[31-Aug-2009 15:57:14] <rmatte> yeh, hopefully you trust your users
[31-Aug-2009 15:57:23] <cgibbons> well, that only took all day
[31-Aug-2009 15:57:29] <straterra> And they can't get to it..that server is on its own subnet
[31-Aug-2009 15:59:06] <rmatte> we were more referring to other users on the zenoss box
[31-Aug-2009 15:59:11] <rmatte> rather than "users" in general
[31-Aug-2009 15:59:15] <rmatte> but yeh, you're fine then
[31-Aug-2009 15:59:42] <davetoo> I wish I understood LDAP more
[31-Aug-2009 15:59:55] <rmatte> yeh, we use LDAP for everything
[31-Aug-2009 16:00:12] <rmatte> ticketing system login, wiki login, zenoss login, etc...
[31-Aug-2009 16:00:17] <davetoo> I don't understand why it has to "search"
[31-Aug-2009 16:00:26] <davetoo> one system I have does'nt need a Manager user
[31-Aug-2009 16:00:38] <davetoo> it just simply tries to bind to the server with the user's credentials,
[31-Aug-2009 16:00:44] <davetoo> and if it succeeds, it's authorized
[31-Aug-2009 16:00:49] <straterra> rmatte: are you familiar with foxmarks/xmarks ?
[31-Aug-2009 16:00:58] <rmatte> davetoo: hunh?
[31-Aug-2009 16:01:02] <rmatte> straterra: nope
[31-Aug-2009 16:01:06] <davetoo> tocotox
[31-Aug-2009 16:01:20] <rmatte> davetoo: oh I see what you're saying
[31-Aug-2009 16:01:22] <straterra> It's a plugin for firefox that lets you sync your bookmarks/saved passwords (passwords are encrypted)
[31-Aug-2009 16:01:30] <rmatte> davetoo: then you have a badly configure LDAP server
[31-Aug-2009 16:01:32] <straterra> I run one here at work..a server, that is
[31-Aug-2009 16:01:35] <davetoo> I'm not using ldap for groups or anything
[31-Aug-2009 16:01:50] <rmatte> straterra: ah
[31-Aug-2009 16:01:51] <straterra> I wrote my own frontend for it and it ties in to AD LDAP
[31-Aug-2009 16:02:16] <rmatte> cool
[31-Aug-2009 16:02:23] <chudler> what? the xmarks ties into LDAP? neato, but why?
[31-Aug-2009 16:02:37] <straterra> So that users can use their login username/pw
[31-Aug-2009 16:02:42] <rmatte> davetoo: ours requires an administrative account to connect to LDAP, if it's not an LDAP admin it doesn't work, which is the way it should be.
[31-Aug-2009 16:02:51] <straterra> It's password protected to each user is only allowed access to their own bookmarks
[31-Aug-2009 16:02:52] <chudler> ah, I see, its for auth and possibly decryption?
[31-Aug-2009 16:03:16] <straterra> They can set a seperate pw for their encryption
[31-Aug-2009 16:03:42] <straterra> It's basically apache2 + mysql + php + AD LDAP
[31-Aug-2009 16:08:17] <nzle> rmatte, responding to somthing a while ago regarding my wanting an indicator of an OK service check. I dont think its 'pointless' how about the case where a script or somthing generating the event is borked. An indicator of a test in 'Unknown' status would tell me that. Assuming no news is good news would not.
[31-Aug-2009 16:09:11] <rmatte> nzle: well, if that's what you want then you're going to probably have to do it with a transform
[31-Aug-2009 16:10:06] <rmatte> if result is blah: evt.summary = "BAD"; if result is good: evt.summary = "OK"
[31-Aug-2009 16:10:08] <rmatte> or whatever
[31-Aug-2009 16:10:39] <nzle> doesnt a transform still need an incoming event? What I want is to configure an Expected event and have it show Unknown untill it gets someting.
[31-Aug-2009 16:11:10] <nzle> Ill look up transforms and read up on them... thanx for the direction..
[31-Aug-2009 16:11:26] <rmatte> nzle: ok, do you want an actual OK event generated in the event console?
[31-Aug-2009 16:11:49] <rmatte> actually, I might have a better solution for you
[31-Aug-2009 16:11:52] <rmatte> depending on what you want
[31-Aug-2009 16:11:59] <nzle> No, I want a null event. An indicator that I expect an event.
[31-Aug-2009 16:12:40] <rmatte> nzle: ok, you're not explaining this very clearly, what's your definition of a "null" event
[31-Aug-2009 16:12:44] <rmatte> one that clears the bad event?
[31-Aug-2009 16:13:08] <rmatte> give me an example
[31-Aug-2009 16:14:40] <chudler> sounds like a heartbeat to me..
[31-Aug-2009 16:14:41] <nzle> Take a heartbeat for example. As I understand it, you have an incoming event and you transform it to give it a timeout. If another doesnt come before the timeout, it goes critical. What if I didnt get a heartbeat at all following a failure that tood out zenoss too. I would have no indication of a lost heartbeat.
[31-Aug-2009 16:16:22] * rmatte scratches his head: "following a failure that tood out zenoss too"
[31-Aug-2009 16:16:37] <rmatte> (and I know that's "took")
[31-Aug-2009 16:16:39] <nzle> unless I had a indicator on the dashboard where I expect a heartbeat. Leave it grey or somthing to signal no status, then go green or red or yellow when you know somthing
[31-Aug-2009 16:17:18] <rmatte> you basically saying that you want something that is going to be an indication that you're actually actively monitoring that item?
[31-Aug-2009 16:18:00] <nzle> Yes.
[31-Aug-2009 16:18:45] <rmatte> ok, well, the way you're wanting to do it would involve actually scripting and changing stuff in Zenoss, which, since you seem to be starting out with it, is not an option... but I'll tell you how I'd do it...
[31-Aug-2009 16:19:22] <rmatte> I'd have the script output 0 if everything is ok and 1 if it's not, I would then creat a threshold and name it sometime appropriate, then set the Max value to 1
[31-Aug-2009 16:19:43] <rmatte> so when the script outputs 1, it triggers the threshold
[31-Aug-2009 16:19:49] <rmatte> when it goes back to 0, it clears the threshold
[31-Aug-2009 16:20:13] <rmatte> if the command times out while running, for whatever reason, Zenoss already has something built in which will generate a "Command timed out" event in the event console
[31-Aug-2009 16:20:21] <rmatte> so right away you'll know something's wrong
[31-Aug-2009 16:20:40] <rmatte> and if worst comes to worst and the whole zencommand daemon dies, resulting in no commands running, you'll receive a heartbeat alert for zencommand
[31-Aug-2009 16:21:08] <rmatte> if the entire Zenoss instance dies, at that point you're pretty much screwed no matter what, unless you have some script checking it every few minutes and alerting if it's down
[31-Aug-2009 16:21:58] <rmatte> (and when it clears the threshold, it sends an actual clear event which can be seen in the event history, so you know exactly what time the alert cleared)
[31-Aug-2009 16:24:33] <rmatte> you could even make a threshold called something else that thresholds on 0
[31-Aug-2009 16:24:38] <rmatte> and have it send clear events
[31-Aug-2009 16:24:41] <nzle> Ok, let me kick that around and see what I come up with. I was just assuming that I was missing somthing. Apparently Im not and I just need to adjust my point of view. In Nagios you configured hosts, and services, all the configured services for a host showed up unknown till they got checked and got an OK, Warning, or Critical.
[31-Aug-2009 16:25:13] <rmatte> yeh, forget about Nagios, Zenoss works differently
[31-Aug-2009 16:26:27] <rmatte> but yeh, read up on device templates, thresholds, transforms
[31-Aug-2009 16:26:39] <rmatte> I'd advise reading through the Zenoss Admin Guide in it's entirety
[31-Aug-2009 16:26:42] <rmatte> it's quite good
[31-Aug-2009 16:27:00] <rmatte> it'll be a day's read at least, but it's worth it
[31-Aug-2009 16:27:17] <nzle> I did a while back when I was first starting. Probably time to again since I have some background with it now.
[31-Aug-2009 16:27:28] <mrayzenoss> or even better, come to the free Zenoss training at Ohio LinuxFest
[31-Aug-2009 16:27:42] <rmatte> yeh, if you can make it out to Ohio
[31-Aug-2009 16:27:56] <nzle> Perhaps...Ill have to see if I can sell that idea to the boss.
[31-Aug-2009 16:28:03] <mrayzenoss> nzle: where are you at?
[31-Aug-2009 16:28:09] <nzle> Wisconsin.
[31-Aug-2009 16:28:17] <mrayzenoss> oh man, that's just a 10 hour drive
[31-Aug-2009 16:28:28] <rmatte> lol
[31-Aug-2009 16:28:29] <nzle> $$?
[31-Aug-2009 16:28:33] <mrayzenoss> the training is free
[31-Aug-2009 16:28:57] <mrayzenoss> it's the same material presented in the $$ Zenoss Enterprise training, minus the Enteprise bits and the hands-on lab
[31-Aug-2009 16:29:01] <mrayzenoss> so 1 day instead of 2
[31-Aug-2009 16:29:07] <nzle> When is it?
[31-Aug-2009 16:29:22] <nzle> Where is info online about it?
[31-Aug-2009 16:29:36] <mrayzenoss> just a sec, I'll get the URL
[31-Aug-2009 16:30:26] <rmatte> http://blog.zenoss.com/2009/08/21/zenoss-community-day-september-25th/
[31-Aug-2009 16:30:30] <nzle> found it in the newsletter
[31-Aug-2009 16:30:36] <mrayzenoss> just noticed the link is bad
[31-Aug-2009 16:30:43] <mrayzenoss> http://blog.zenoss.com/2009/08/21/zenoss-community-day-september-25th/
[31-Aug-2009 16:30:47] <mrayzenoss> at least on the blog it's bad
[31-Aug-2009 16:30:55] <rmatte> beatcha to it
[31-Aug-2009 16:31:00] <nzle> got it... thanx. Ill see if I can work it out.
[31-Aug-2009 16:31:10] <mrayzenoss> we did the training at SCALE, it went really well
[31-Aug-2009 16:31:19] <mrayzenoss> 50 people took the 50 spots, people standing in the back
[31-Aug-2009 16:31:33] <mrayzenoss> Ohio has 100 seats
[31-Aug-2009 16:31:37] <rmatte> it's fun stuff to learn hehe
[31-Aug-2009 16:31:40] <rmatte> it's not a boring product
[31-Aug-2009 16:31:52] <mrayzenoss> plus an open bar!
[31-Aug-2009 16:31:56] <nzle> Is there registration, or is it first come first served?
[31-Aug-2009 16:31:58] <rmatte> seriously?
[31-Aug-2009 16:32:03] <rmatte> I'd hit that training so hard
[31-Aug-2009 16:32:06] <rmatte> lol
[31-Aug-2009 16:32:10] <mrayzenoss> 8:00 - Ohio Linux pre-party at Barley’s sponsored by Zenoss
[31-Aug-2009 16:32:14] <nzle> <g>
[31-Aug-2009 16:32:26] <rmatte> "Does everyone have their template ready?" "I'm smashed, can I go home?"
[31-Aug-2009 16:32:42] <mrayzenoss> there's registration
[31-Aug-2009 16:32:55] <mrayzenoss> I'm sure there will be walk-ups, SCALE had them and they stood in the back
[31-Aug-2009 16:33:04] <rmatte> cool
[31-Aug-2009 16:34:52] <nzle> But its not all walkup, one can reserve a seat ahead of time?
[31-Aug-2009 16:35:01] <mrayzenoss> yeah, fixing the link right now
[31-Aug-2009 16:35:49] <mrayzenoss> the first "sign-up" link is good
[31-Aug-2009 16:37:33] <nzle> Cool. I see if I can sell the idea. RU one of the presenters?
[31-Aug-2009 16:37:39] <mrayzenoss> yeah
[31-Aug-2009 16:37:54] <mrayzenoss> npmccallum will handle the training, since he normally teaches it
[31-Aug-2009 16:38:02] <mrayzenoss> I'll be handling the Q&A sections and assisting
[31-Aug-2009 16:38:17] <mrayzenoss> and then on Saturday I'll be manning the booth and doing a demo
[31-Aug-2009 16:39:26] <rmatte> well, I'm out of here, talk to you guys tomorrow, have a good evening/night/morning
[31-Aug-2009 16:39:33] <mrayzenoss> later
[31-Aug-2009 16:39:42] <nzle> Thanx for the advise....
[31-Aug-2009 16:39:46] <rmatte> np
[31-Aug-2009 16:47:36] <nzle> mray, you still here??
[31-Aug-2009 16:48:12] <mrayzenoss> yeah
[31-Aug-2009 17:02:57] <davetoo> oh
[31-Aug-2009 17:03:01] <davetoo> mrayzenoss: you in SF?
[31-Aug-2009 17:03:06] <mrayzenoss> Austin
[31-Aug-2009 17:03:10] <davetoo> oh
[31-Aug-2009 17:03:17] <davetoo> who's at VMWorld?
[31-Aug-2009 17:03:21] <mrayzenoss> I didn't make the VMworld trip, that's mostly Enterprise
[31-Aug-2009 17:03:27] <davetoo> aye
[31-Aug-2009 17:03:34] <mrayzenoss> sales, marketing, director of alliances, mrchippy
[31-Aug-2009 17:03:47] <mrayzenoss> CEO, CMO
[31-Aug-2009 17:03:54] <mrayzenoss> CTO
[31-Aug-2009 17:04:06] <davetoo> I met Bill and Erik at training
[31-Aug-2009 17:04:14] <mrayzenoss> yeah, CEO and CTO are there
[31-Aug-2009 17:04:36] <davetoo> Too many others from my office are already up there, so I have to stay here
[31-Aug-2009 17:04:50] <davetoo> Or I would go ask Erik a couple of followup questions from Friday a week ago.
[31-Aug-2009 17:04:52] <mrayzenoss> well, I'd rather be at Red Hat Summit, which is also right now
[31-Aug-2009 17:05:06] <mrayzenoss> which we're also at and presenting
[31-Aug-2009 17:05:24] <mrayzenoss> so sales and marketing are spread thin
[31-Aug-2009 17:05:33] <mrayzenoss> and some of them are hitting both events
[31-Aug-2009 17:05:39] <davetoo> lots of air travel
[31-Aug-2009 17:05:55] <mrayzenoss> yeah, SF and Chicago
[31-Aug-2009 17:21:54] <davetoo> uh oh, zenpack --delete strikes again
[31-Aug-2009 17:27:21] <davetoo> this is interesting: it deleted the script but left the compiled .pyc
[31-Aug-2009 17:27:34] <davetoo> so it's still working
[31-Aug-2009 17:29:31] <gemineye> Hi mrayzenoss, long time no see.
[31-Aug-2009 17:29:44] <mrayzenoss> hey, how's it going?
[31-Aug-2009 17:30:04] <gemineye> Too bad you're not in SF, that's a 30 min drive from me.
[31-Aug-2009 17:30:23] * davetoo is in sunnyvale
[31-Aug-2009 17:30:29] <gemineye> ditto
[31-Aug-2009 17:30:55] <davetoo> thank god for air conditining and electricity, eh?
[31-Aug-2009 17:31:22] <gemineye> Yeah, but if the building gets too hot, I can always rely on our server rooms to be cool.
[31-Aug-2009 17:31:27] <gemineye> davetoo, you work at netapp?
[31-Aug-2009 17:31:33] <davetoo> oops
[31-Aug-2009 17:31:39] <davetoo> I need a proxy
[31-Aug-2009 17:31:40] <davetoo> yeah
[31-Aug-2009 17:31:43] <mrayzenoss> heh
[31-Aug-2009 17:32:08] <gemineye> I work at Blue Coat We're looking to refresh all our storage with new NetApp.
[31-Aug-2009 17:32:27] <mrayzenoss> maybe you could trade a Blue Coat proxy for a NetApp
[31-Aug-2009 17:32:42] <davetoo> I'm so far away from sales it's ... a good thing, actually
[31-Aug-2009 17:32:45] <gemineye> SURE!!! Not such a good deal for NetApp, but why not
[31-Aug-2009 17:33:11] <davetoo> no, I need a vm in a colo shop or something
[31-Aug-2009 17:33:21] <davetoo> or better yet, get in somebody's onion network
[31-Aug-2009 17:33:25] <gemineye> mrayzenoss, when's the next feature update going to happen to core?
[31-Aug-2009 17:34:09] <mrayzenoss> start of beta is penciled in for next week
[31-Aug-2009 17:34:16] <gemineye> I just have everything at a friends colo, I'm tunneling through SSH outside of Blue Coat since we don't allow irc.
[31-Aug-2009 17:34:33] <gemineye> When is 2.4.5 being released?
[31-Aug-2009 17:34:44] <mrayzenoss> looks like tomorrow morning
[31-Aug-2009 17:37:08] <gemineye> Hey, are there memory leaks in core? Every 4 days I have to stop core, kill off some python processes, and restart it. Every 4 days or so all 4GB of memory is used up and I've only got 181 clients being monitored, 48 of them are just ping tests.
[31-Aug-2009 17:37:39] <mrayzenoss> I know cgibbons is periodically finding them in Windows
[31-Aug-2009 17:37:49] <mrayzenoss> and in some of the wmi stuff
[31-Aug-2009 17:37:53] <gemineye> Mine is running on CentOS 5
[31-Aug-2009 17:38:31] <mrayzenoss> 2.4.2?
[31-Aug-2009 17:38:49] <gemineye> I noticed that zencommand, zenperfsnmp, and zenwin would die due to no memory being left, but today zeneventlog crashed with an error in the log.
[31-Aug-2009 17:38:53] <gemineye> yeah, 2.4.2
[31-Aug-2009 17:39:48] <gemineye> Where's the event log for zeneventlog?
[31-Aug-2009 17:39:58] <davetoo> /opt/zenoss/log
[31-Aug-2009 17:40:11] <gemineye> nope
[31-Aug-2009 17:40:13] <davetoo> hmm
[31-Aug-2009 17:40:28] * davetoo hasn't done any windows monitoring yet
[31-Aug-2009 17:41:05] <gemineye> /usr/local/zenoss/zenoss/log
[31-Aug-2009 17:41:09] <gemineye> stack install
[31-Aug-2009 17:41:25] <mrayzenoss> $ZENHOME/log/
[31-Aug-2009 17:42:02] <cgibbons> there's another memory leak opened in my queue, lemme see what the details are
[31-Aug-2009 17:42:12] <cgibbons> looks like for zenmodeler
[31-Aug-2009 17:42:28] <gemineye> Do these problems exist also in Enterprise?
[31-Aug-2009 17:43:03] <mrayzenoss> Enterprise uses the same zenmodeler, so yeah
[31-Aug-2009 17:43:16] <cgibbons> those wouldn't be different there, but that kind of problem is usually an error handling bug
[31-Aug-2009 17:50:56] <mrayzenoss> later all, heading home
[31-Aug-2009 22:21:18] bzed_ is now known as bzed
[01-Sep-2009 00:00:46] [disconnected at Tue Sep 1 00:00:46 2009]
[01-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[01-Aug-2009 00:48:57] <Diddi> hm, I have a switch under monitoring... is it possible to change stuff in it using zenoss, like interface description etc..?
[01-Aug-2009 00:49:08] <Diddi> link up/down,,
[01-Aug-2009 00:51:47] <Diddi> ah.. could always create my own commands for that
[01-Aug-2009 00:51:51] <Diddi> i guess
[01-Aug-2009 05:47:42] <Diddi> is it possible to draw map links only between devices that are actually connected to each other, instead of drawing links between all devices in the same network?
[01-Aug-2009 05:48:16] <Diddi> got a couple of switches in the same subnet... but they're not all connected to eachother (:
[02-Aug-2009 00:00:46] [disconnected at Sun Aug 2 00:00:46 2009]
[02-Aug-2009 00:00:46] [connected at Sun Aug 2 00:00:46 2009]
[02-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[02-Aug-2009 05:38:37] <Diddi> I'm using Paging to send SMS instead of Page (calling another program only).. Is it possible to add newlines in the Paging message?
[02-Aug-2009 05:49:00] <Diddi> I've tried \n and \r in different combinations without success.. they're just sent as-is
[02-Aug-2009 15:06:48] ke4qqq_ is now known as ke4qqq
[02-Aug-2009 21:30:12] <sergeymasushko> how can force remodeling of all devices?
[02-Aug-2009 22:13:52] edwardam1 is now known as edwardam
[03-Aug-2009 00:00:46] [disconnected at Mon Aug 3 00:00:46 2009]
[03-Aug-2009 00:00:46] [connected at Mon Aug 3 00:00:46 2009]
[03-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[03-Aug-2009 03:12:09] <Dieterbe> when i create a datasource for a template and put a nagios command in it. I can select the severity, but is this a "catch all" severity? because the nagios plugin will return different exit codes from warning/error/unknown states. it seems a bit weird that all these conditions map to only 1 severity in zenoss
[03-Aug-2009 08:52:28] edwardam1 is now known as edwardam
[03-Aug-2009 10:42:56] <eidolon> hi folks - i'm configuring up windows clients on our zenoss enterprise installation. i can use WMI or SNMP - we're mosty a linux shop, so SNMP is the preferred method. is there a reason to use one ove rthe other? do all zenpacks work with both?
[03-Aug-2009 10:44:17] <ckrough> I've heard WMI is more reliable for Windows nodes
[03-Aug-2009 10:46:17] <mrayzenoss> so officially we try to make them equally, but the SNMP from MS is somewhat limited
[03-Aug-2009 10:46:31] <mrayzenoss> but the WMI performance with Server 2008 and Vista reportedly sucks
[03-Aug-2009 10:46:41] <mrayzenoss> not us, MS's implementation
[03-Aug-2009 10:46:59] <mrayzenoss> but WMI uses TCP so it should be more network reliable
[03-Aug-2009 10:47:13] <mrayzenoss> but since it's MicroSoft, it also tends to have more memory leaks server-side
[03-Aug-2009 10:47:36] <mrayzenoss> WMI gives you event log and service monitoring, which aren't exposed via SNMP
[03-Aug-2009 10:49:19] <Disconnect> so in the end, the answer is.. don't use windows?
[03-Aug-2009 10:49:20] <mrayzenoss> so the short answer is "use WMI"
[03-Aug-2009 10:49:24] <mrayzenoss> heh
[03-Aug-2009 10:49:43] <mrayzenoss> Disconnect: you said it, not me
[03-Aug-2009 10:49:47] <Disconnect> heh
[03-Aug-2009 10:50:10] <Disconnect> (admittedly, *nix doesn't expose syslog via snmp either..)
[03-Aug-2009 10:50:46] <mrayzenoss> well, there is a tool for exporting Windows Event Log with syslog
[03-Aug-2009 10:51:42] <mrayzenoss> http://www.intersectalliance.com/projects/SnareWindows/
[03-Aug-2009 10:52:02] <mrayzenoss> again, short answer is "use WMI and stay up on your patches"
[03-Aug-2009 11:11:08] <eidolon> mray: that's a very good answer, thank you - oh crap, he left. but... my issue with WMI is this we also seem to be getting system event log traffic into zenoss. stuff we have no interest whatsoever in tracking
[03-Aug-2009 11:11:11] <eidolon> how can we turn that off?
[03-Aug-2009 11:19:21] <rmatte> you want to eliminate all syslog messages via WMI?
[03-Aug-2009 11:19:44] <rmatte> or just certain ones?
[03-Aug-2009 11:22:08] <rmatte> If you want to disable all of them, then you need to go in to zProperties for the Windows class and change zWinEventlog to false. If you want to keep monitoring it but only monitor higher level events then you need to change the value for zWinEventlogMinSeverity
[03-Aug-2009 11:22:29] <eidolon> ooooo
[03-Aug-2009 11:22:35] <eidolon> that sounds awesome rmatte. *makes notes*
[03-Aug-2009 11:22:52] <eidolon> i need to remodel some of these machines that were set up as SNMP, and make them WMI, then i'll set those properties.
[03-Aug-2009 11:23:49] <rmatte> if you want to block certain messages then you need to go to events on the left, go to a class (you can optionally make one called "Suppressed" or something, or you can use an existing class), then go to the dropdown click Add Mapping, name it whatever you want (WinEventlogSuppress1 or whatever), then click on it to access it's properties...
[03-Aug-2009 11:25:10] <eidolon> hum
[03-Aug-2009 11:25:13] <rmatte> Go to the Edit tab, set the event class key to "defaultmapping" and in the Regex box, type in part of the message that you're seeing, in the example box put the whole message you're seeing, then click save and go to the zProperties tab, set zEventAction to "drop" and click save
[03-Aug-2009 11:25:25] <eidolon> rmatte, how can i tell how polling is happening? because on my path, I have: /Devices /Server /Windows /WMI_Active Directory /BOS-ADS01
[03-Aug-2009 11:25:25] <rmatte> that's for suppressing individual messages
[03-Aug-2009 11:25:35] <eidolon> butunder Attributes, it's got all SNMP settings.
[03-Aug-2009 11:26:03] <rmatte> each class can have local attributes assigned to it, it all depends on how you setup the zProperties after creating that class
[03-Aug-2009 11:26:13] * eidolon didn't create this class
[03-Aug-2009 11:26:22] <eidolon> i'm tempted to delete it and remodel it.
[03-Aug-2009 11:26:27] <rmatte> you didn't create the WMI_Active Directory class?
[03-Aug-2009 11:26:37] <eidolon> no. i inherited this setup.
[03-Aug-2009 11:26:47] <rmatte> ah, well the person who created that class then
[03-Aug-2009 11:26:59] <rmatte> you need to go in to the zProperties and take a look at what's enabled/disabled
[03-Aug-2009 11:27:01] * eidolon sees zAxlPassword nad Username in the zproperties
[03-Aug-2009 11:27:15] <rmatte> Also, if you scroll to the bottom of zProperties, you'll see what fields are set locally
[03-Aug-2009 11:27:19] <rmatte> there's a dropdown with a list
[03-Aug-2009 11:27:20] <eidolon> zdevice templates - Device_WMI and ActiveDirectory
[03-Aug-2009 11:27:31] <eidolon> ok
[03-Aug-2009 11:27:45] <eidolon> local properties have... great.
[03-Aug-2009 11:27:46] <rmatte> mrayzenoss: FYI, I'm on vacation for the next 2 weeks, so I probably won't be around much, but I'll stop in a bit
[03-Aug-2009 11:28:05] <eidolon> zwinpassword, zwinusername, zsnmp port...
[03-Aug-2009 11:28:09] <eidolon> mixed signals
[03-Aug-2009 11:28:23] <rmatte> the fields are always going to be there, they are global
[03-Aug-2009 11:28:30] <rmatte> it all depends on what they are actually set to
[03-Aug-2009 11:28:46] <eidolon> ok. i see.
[03-Aug-2009 11:28:59] <eidolon> zwinsuer and zwinpassword are set. zwmimonitorignore is flase
[03-Aug-2009 11:29:03] <rmatte> and it is possible to do both WMI and SNMP monitoring on the same device
[03-Aug-2009 11:29:10] <eidolon> (sorry if i'm sounding stupid here)
[03-Aug-2009 11:29:31] <rmatte> then WMI monitoring is enabled for that device
[03-Aug-2009 11:29:37] <eidolon> ok
[03-Aug-2009 11:29:55] <rmatte> there's zWinEventlog which determines whether the event log is monitored or not
[03-Aug-2009 11:30:02] <rmatte> there's also the severity setting for it
[03-Aug-2009 11:30:14] <rmatte> and there will be zSnmpMonitorIgnore, and such
[03-Aug-2009 11:30:21] <eidolon> 'zwineventlog' 'trus' 'zwineventlogminseverity' '2'
[03-Aug-2009 11:30:24] <rmatte> which should be set to true if you're doing straight WMI
[03-Aug-2009 11:30:36] <rmatte> yeh, that min severity is the default
[03-Aug-2009 11:30:42] <eidolon> zsnmpmonitorignore is set to false
[03-Aug-2009 11:30:45] <eidolon> so we're doing both
[03-Aug-2009 11:30:58] <rmatte> is there an snmp community string specified?
[03-Aug-2009 11:31:02] <eidolon> yes
[03-Aug-2009 11:31:24] <eidolon> okay hang on.
[03-Aug-2009 11:31:28] * eidolon backs up a few steps.
[03-Aug-2009 11:31:43] <rmatte> Under zSnmpCommunity?
[03-Aug-2009 11:31:49] <eidolon> yeah, hang on a sec.
[03-Aug-2009 11:33:09] <straterra> My new netbook has arrived..Mwuahahaha
[03-Aug-2009 11:34:35] <rmatte> lol
[03-Aug-2009 11:35:44] <eidolon> okee. turned off wineventlog for the WMI and WMI_Active Directory classes...
[03-Aug-2009 11:35:49] * eidolon continues fiddling.
[03-Aug-2009 11:38:12] <eidolon> i love classes and organizers with no devices or sub-devices.
[03-Aug-2009 11:38:15] * eidolon cleans up.
[03-Aug-2009 11:39:17] <rmatte> lol
[03-Aug-2009 11:39:25] <rmatte> some are there by default
[03-Aug-2009 11:39:54] <rmatte> I wouldn't go too crazy deleting stuff, you might need some of it in the future
[03-Aug-2009 11:40:01] <Disconnect> yah i discovered that removing some of them breaks zenmigrate or whatever the update app is :/
[03-Aug-2009 11:40:02] <eidolon> so i already see a problem. i suspect thes guys are going to want us to monitor their apps and event logs for system errors.
[03-Aug-2009 11:40:09] <eidolon> but i dont know if they're using any consistent naming scheme.
[03-Aug-2009 11:40:29] <eidolon> like, i have 38,515 occurances of:
[03-Aug-2009 11:40:30] <eidolon> Windows Service 'SQLSERVERAGENT' is down
[03-Aug-2009 11:40:30] <rmatte> naming scheme for?
[03-Aug-2009 11:40:35] <eidolon> application errors.
[03-Aug-2009 11:40:44] <rmatte> I see
[03-Aug-2009 11:41:27] * eidolon has so much to learn.
[03-Aug-2009 11:42:33] <rmatte> Yeh, it takes a while
[03-Aug-2009 11:43:26] <eidolon> well, i keep wanting to go "See that error message? add an ignore rule for it."
[03-Aug-2009 11:43:29] <eidolon> and zenoss doesn't work that way.
[03-Aug-2009 11:43:45] <rmatte> I explained how to ignore errors
[03-Aug-2009 11:43:48] <eidolon> you did.
[03-Aug-2009 11:43:50] <eidolon> i saw
[03-Aug-2009 11:44:33] <rmatte> it's a bit more complex than just point and click, but it's fairly simple
[03-Aug-2009 11:44:57] <eidolon> yah, i see it.
[03-Aug-2009 11:45:08] <eidolon> well, right now what i've done is shut down event logging.
[03-Aug-2009 11:45:34] <eidolon> since i templated to the users "i will be monitoring system up, load, cpu, memory, and remote sockets. if youw ant more, tell me what you want."
[03-Aug-2009 11:45:37] <eidolon> and they never got back to me.
[03-Aug-2009 11:45:40] <eidolon> so. f'em.
[03-Aug-2009 11:45:49] * eidolon BOFH's.
[03-Aug-2009 11:46:00] <rmatte> hehe
[03-Aug-2009 11:46:25] <rmatte> with your sidekick, the PFY?
[03-Aug-2009 11:46:27] <eidolon> can i have no pre-production machines show upin the event list?
[03-Aug-2009 11:46:39] <eidolon> ala, i like having a very clean dashboard.
[03-Aug-2009 11:46:49] <eidolon> well, event console,sorry
[03-Aug-2009 11:47:04] <rmatte> eidolon: you effectively don't want to monitor them at all?
[03-Aug-2009 11:47:09] <rmatte> or just not have events show up?
[03-Aug-2009 11:47:13] <eidolon> let me think a moment.
[03-Aug-2009 11:47:53] <eidolon> i dont' want events to show up. i want to be able to... hmm. basically i want to be able to go to event console, and only see production machines. but if i look up a specific pre-production box, i want to see events / performance stats from it.
[03-Aug-2009 11:48:44] * eidolon may be asking for too much.
[03-Aug-2009 11:48:51] <rmatte> well, there's no real way to filter it like that, what you can do is make an event mapping much in the way that I described above, but do a transform that checks to see the production state of a device, and if it's "Pre-Production" sends the event to history immediately
[03-Aug-2009 11:49:12] <rmatte> then you could always go to the events tab of the pre-prod machines and check the event history to see what has come in
[03-Aug-2009 11:49:18] <mloven> event views?
[03-Aug-2009 11:49:55] <rmatte> mloven: hunh?
[03-Aug-2009 11:50:26] <mloven> wouldn't be exactly what he wanted, but he could create an event view that doesn't have any pre-production devices in it...
[03-Aug-2009 11:53:23] <eidolon> that sounds better.
[03-Aug-2009 11:53:46] <eidolon> not sure what that is though
[03-Aug-2009 11:54:06] <rmatte> mloven: you talking about a portlet basically?
[03-Aug-2009 11:54:11] <eidolon> (what i've been doing is setting up groups. a group of 'production' machines, and having a view of htem)
[03-Aug-2009 11:54:22] <rmatte> The transform would be:
[03-Aug-2009 11:54:22] <rmatte> if evt.prodState == 'Pre-Production':
[03-Aug-2009 11:54:23] <rmatte> evt._action = 'history'
[03-Aug-2009 11:54:25] <rmatte> for my method
[03-Aug-2009 11:54:53] <mloven> It doesn't completely solve the issue because the events from Pre-Production devices will still show up in the event console, but you could have a dedicated event view that only shows Production devices' events
[03-Aug-2009 11:55:09] <eidolon> how do you build event views?
[03-Aug-2009 11:55:46] <mloven> Settings>Users>yourusername>Event Views
[03-Aug-2009 11:57:44] <mloven> you would just specify Production State > Pre-Production
[03-Aug-2009 11:58:23] <rmatte> or Production-State = Production if you only care about Production alerts
[03-Aug-2009 11:58:35] <eidolon> hmmmmmmmmmmmmmm.
[03-Aug-2009 11:59:03] <rmatte> my method would resolve the issue for the main event console
[03-Aug-2009 11:59:06] <rmatte> up to you
[03-Aug-2009 12:01:40] <eidolon> actually i like hte custom views.
[03-Aug-2009 12:01:45] <mloven> yep. both work. my way is just like flipping a filter that doesn't show devices that aren't in Production. the events are still there....lurking.... rmatte's way actually gets rid of the events. Depends on what you want.
[03-Aug-2009 12:01:45] * eidolon is playing now.
[03-Aug-2009 12:02:59] <eidolon> awesome.
[03-Aug-2009 12:03:00] <eidolon> look at that
[03-Aug-2009 12:53:16] <konfoo> anyone here monitoring using the dell pack?
[03-Aug-2009 12:53:45] <konfoo> degrading the raid array raises an alert in the hardware properties page but the overall status is OK
[03-Aug-2009 14:14:14] <eidolon> hmm. how long should zenbackup take to run?
[03-Aug-2009 14:14:28] <eidolon> we have about 100 machines.
[03-Aug-2009 15:12:28] <rhett> eidolon that all depends on how fast the machine is that you're doing the backup on
[03-Aug-2009 15:13:02] <rhett> it tgz's everything into the backup archive
[03-Aug-2009 16:05:49] <mrayzenoss> New WMI Data Source ZenPack: http://www.zenoss.com/community/projects/zenpacks/wmidatasource
[03-Aug-2009 16:33:03] <malbon> excellent. A 3rd way of getting WMI data
[03-Aug-2009 16:44:21] <rmatte> actually, these new WMI packs that are coming out are going to be way nicer than the existing Dastrup ones
[03-Aug-2009 22:28:13] <ironpaw> anyone around?
[03-Aug-2009 22:28:23] <ironpaw> where can i find the conary migrate binary?
[04-Aug-2009 00:00:46] [disconnected at Tue Aug 4 00:00:46 2009]
[04-Aug-2009 00:00:46] [connected at Tue Aug 4 00:00:46 2009]
[04-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[04-Aug-2009 00:28:41] <raddy> Hello Everybody
[04-Aug-2009 00:28:48] <raddy> Anybody out there?
[04-Aug-2009 03:19:52] <raddy> Hello Everybody
[04-Aug-2009 03:20:14] <raddy> Can anybody suggest a regexp for monitoring postgres process?
[04-Aug-2009 03:20:38] <raddy> just using postgres shows 50 processes.
[04-Aug-2009 03:29:37] <raddy> Anybody there???
[04-Aug-2009 03:29:53] <raddy> How do i monitor postgres process?
[04-Aug-2009 03:35:17] <raddy> Anybody there??
[04-Aug-2009 03:41:55] <raddy> Anybody out there???
[04-Aug-2009 04:26:14] <raddy> Hello Everybody
[04-Aug-2009 04:26:19] <raddy> Anybody there?????
[04-Aug-2009 04:39:26] <raddy> Anybody there?
[04-Aug-2009 04:40:01] <raddy> Can anybody suggest a good regex for postgres
[04-Aug-2009 04:57:03] <raddy> Hi folke
[04-Aug-2009 04:57:21] <folke> Hi raddy.
[04-Aug-2009 05:02:06] <raddy> folke : Can you suggest a good regex for postgres.
[04-Aug-2009 05:02:39] <folke> raddy, regex testing?
[04-Aug-2009 05:02:55] <raddy> folke : Nope, real usage.
[04-Aug-2009 05:03:28] <raddy> folke : zenoss is showing too much of processes when i just use postgres
[04-Aug-2009 05:04:10] <folke> raddy: Ah, ok.. Sry I had the same problem before when I monitored a few linux nodes.
[04-Aug-2009 05:04:31] <folke> Do you see the right precesses atleas? Or every process on the server?
[04-Aug-2009 05:04:59] <raddy> folke : right process alias i think
[04-Aug-2009 05:05:44] <folke> Do you run the latest 2.4.2, I know that there has been trouble with processes and linux historically.
[04-Aug-2009 05:06:03] <raddy> folke : Yup, latest version.
[04-Aug-2009 05:07:25] <folke> Ok, I am for the moment trying to "upgrade" our old 2.1.3 env to 2.4.2
[04-Aug-2009 05:07:39] <folke> So I will try to mess with processes in nix again.
[04-Aug-2009 05:07:43] <raddy> folke : ok
[04-Aug-2009 05:07:56] <raddy> folke : Do you have solution?
[04-Aug-2009 05:08:06] <folke> raddy: Sry, not for the momen..
[04-Aug-2009 05:08:18] <raddy> folke : ok
[04-Aug-2009 05:47:48] <raddy> Anybody there?
[04-Aug-2009 05:52:52] baffle_ is now known as baffle
[04-Aug-2009 08:11:32] <klinstifen> morning all
[04-Aug-2009 08:11:49] <klinstifen> can someone confirm for me the size of their data.fs file in $ZENHOME/var
[04-Aug-2009 08:12:16] <klinstifen> mine is 14G
[04-Aug-2009 08:12:31] <klinstifen> that seems high to me especially because I only monitor 100 or so devices
[04-Aug-2009 08:18:59] <Dieterbe> klinstifen: just a sec..
[04-Aug-2009 08:19:46] <Dieterbe> 53M Aug 4 14:07 Data.fs
[04-Aug-2009 08:20:17] <Dieterbe> that's an installation that monitors about 12 devices, and it's up for .. 2 months or so (rough estimate)
[04-Aug-2009 08:20:34] <Dieterbe> 1-2 months
[04-Aug-2009 08:20:47] <klinstifen> hmmm
[04-Aug-2009 08:20:55] <klinstifen> i am monitoring 39 devices
[04-Aug-2009 08:20:56] <Dieterbe> how should one debug "non working" event transformations?
[04-Aug-2009 08:22:00] <klinstifen> and it's been up for just over a year
[04-Aug-2009 08:22:43] <klinstifen> what's not working with your transform?
[04-Aug-2009 08:24:13] <Dieterbe> well, to monitor my boxes i need to go through 2 vpn's. one endpoint is vpn0, the one after that is called firewall. i have this rule:
[04-Aug-2009 08:24:16] <Dieterbe> http://pastebin.ca/1517859
[04-Aug-2009 08:24:51] <Dieterbe> but it doesn't work, but at this point the firewall box is not reachable by ping anymore, however i'm getting notifications about all devices being down
[04-Aug-2009 08:26:24] <klinstifen> so what event is this transforming? the firewall down event?
[04-Aug-2009 08:27:04] <klinstifen> can you ping firewall box from the zenoss server cli?
[04-Aug-2009 08:27:30] <Dieterbe> no, it should work on all events for all other boxes who are "behind" firewall. i've put this rule in /Events
[04-Aug-2009 08:27:31] <raddy> Hello Everybody
[04-Aug-2009 08:27:50] <raddy> Can anybody suggest a good regex for postgres?
[04-Aug-2009 08:28:33] <Dieterbe> klinstifen: well, now it pings again, but a few minutes ago it didn't
[04-Aug-2009 08:30:20] <raddy> When i just use postgres it shows lot of processes
[04-Aug-2009 08:30:47] <raddy> Can anyone suggest?
[04-Aug-2009 08:31:00] <ckrough> What is the question?
[04-Aug-2009 08:31:52] <raddy> Can anybody suggest a good regex for postgres
[04-Aug-2009 08:32:18] <raddy> It shows lot of postgres processes when i use just "postgres" as regexp
[04-Aug-2009 08:32:40] <ckrough> oh, I see what your doing. Lemme see what I have in mine.
[04-Aug-2009 08:34:03] <ckrough> maybe match on an expression based around 'postmaster'?
[04-Aug-2009 08:34:20] <ckrough> Im not running postgres to check atm, but I think postmaster is the parent process
[04-Aug-2009 08:34:25] <raddy> ckrough : postmaster is useless
[04-Aug-2009 08:35:22] <cgibbons> hmmm
[04-Aug-2009 08:39:10] <Dieterbe> raddy: what about 'postgres: '
[04-Aug-2009 08:39:24] <Dieterbe> or /usr/bin/postgres for the 'main one' ?
[04-Aug-2009 08:39:59] <ckrough> good site for testing this stuff out: http://gskinner.com/RegExr/
[04-Aug-2009 08:40:11] <Dieterbe> so, how can i debug non-working event transforms?
[04-Aug-2009 08:40:24] <raddy> Dieterbe : /user/bin/postgres didn't work, but i didn't try postgres:
[04-Aug-2009 08:45:00] klinstifen_ is now known as klinstifen
[04-Aug-2009 08:52:44] <Dieterbe> so, klinstifen .. any idea?
[04-Aug-2009 08:52:56] <ckrough> DO you most of you run your ethernet templates with COUNTER or DERIVE datasources? Im dealing with spikes a little too often using COUNTER with a max value of one petabit
[04-Aug-2009 08:53:19] <klinstifen> i don't know if i am reading your transfer code correctly, but it looks like it will only report the status of either ping vpn0 or ping firewall
[04-Aug-2009 08:53:21] <klinstifen> is that correct?
[04-Aug-2009 08:53:22] <ckrough> actually, a max value of 141 terabits since IOS returns bytes...
[04-Aug-2009 08:54:58] <raddy> Dieterbe : postgres: shows all postgres processses
[04-Aug-2009 08:55:27] <ckrough> can you put your process list in a pastebin?
[04-Aug-2009 08:55:35] <Dieterbe> klinstifen: well i handle 2 different cases separately. first i check if it's a server behind vpn0 (that includes firewall). any such event depends on the ping state of vpn0. then i check for all devices behind firewall. these depend on the ping state of firewall
[04-Aug-2009 08:57:33] <klinstifen> well one way you could check it is to just change the code so that the text of the event changes
[04-Aug-2009 08:57:36] <klinstifen> know what i mean?
[04-Aug-2009 08:57:47] <klinstifen> you're not sure which part of the transform is breaking, right?
[04-Aug-2009 08:58:12] <klinstifen> so why not split it up and just have it to something innocuous like changing the text of the event.
[04-Aug-2009 08:58:33] <raddy> ckrough : Is there any other route?
[04-Aug-2009 08:58:36] <klinstifen> then then you drill down further and see if the problem is with the ping check or device name check etc
[04-Aug-2009 09:00:43] <klinstifen> and another thing...can you ping the devices behind firewall from your zenoss server...from the command line?
[04-Aug-2009 09:01:24] <raddy> basically using postgres shows all the idle postgres processes
[04-Aug-2009 09:02:39] <klinstifen> dieterbe: maybe it's just a route issue
[04-Aug-2009 09:08:05] <Dieterbe> klinstifen: ?? no it's just my event transforms that don't work.
[04-Aug-2009 09:08:36] <klinstifen> ah...well then yeah i would suggest breaking em down and testing it that way...
[04-Aug-2009 09:15:23] <mrayzenoss> New WMI Windows Performance ZenPack: http://www.zenoss.com/community/projects/zenpacks/wmiperf_windows
[04-Aug-2009 09:17:36] <Dieterbe> klinstifen: but then i have to bring down my vpn just to be able to test :/ unless i fire multiple events myself
[04-Aug-2009 09:18:40] <klinstifen> you could create a test event.. i think there are docs on that
[04-Aug-2009 09:19:11] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch05s02.html
[04-Aug-2009 09:19:20] <mrayzenoss> or just zensendevent
[04-Aug-2009 09:19:30] <klinstifen> or you could briefly change the ip within zenoss for one of the devices and then zenoss will think it is down
[04-Aug-2009 09:19:46] <klinstifen> even easier, thanks mray
[04-Aug-2009 09:19:58] <mrayzenoss> it's what I'm here for
[04-Aug-2009 09:25:52] <raddy> Hello Everybody
[04-Aug-2009 09:26:07] <raddy> Please suggest good regexp for postgres
[04-Aug-2009 09:38:04] <sto6ma9ch> Hello.
[04-Aug-2009 09:38:18] <sto6ma9ch> I have an issue with zenperfsnmp
[04-Aug-2009 09:38:45] <sto6ma9ch> The log shows "Error loading config for devices ['..." for every device
[04-Aug-2009 09:39:01] <sto6ma9ch> Then I receive a zenperfsnmp heartbeat failure
[04-Aug-2009 09:39:38] <cgibbons> that would first make me think zenhub was sick; have you checked its log file?
[04-Aug-2009 09:39:55] <sto6ma9ch> Not yet. Checking now.
[04-Aug-2009 09:41:06] <sto6ma9ch> Just restarted zenperfsnmp. Tailing zenhub.log
[04-Aug-2009 09:43:30] <sto6ma9ch> This is the only error thus far:
[04-Aug-2009 09:43:33] <sto6ma9ch> ERROR zen.MinMaxCheck: User-supplied Python expression ((here.totalBlocks * here.zFileSystemSizeOffset) * .9) for maximum value caused error: ['usedBlocks_usedBlocks']
[04-Aug-2009 09:44:58] <sto6ma9ch> Ahh, just found one:
[04-Aug-2009 09:45:01] <sto6ma9ch> CRITICAL zen.hub: ZenPing 'cml-zensvr.as.local' not found, ignoring network topology.
[04-Aug-2009 09:49:11] <sto6ma9ch> That's the ZenOSS server itself.
[04-Aug-2009 09:49:26] <sto6ma9ch> Just added it. Restarting zenperfsnmp and tailing zenhub.log
[04-Aug-2009 09:51:20] <ckrough> What type of datasource should I be using with ethernet templates? COUNTER or DERIVE? datasources? Im dealing with spikes a little too often using COUNTER
[04-Aug-2009 09:56:56] <sto6ma9ch> OK, got that error resolved. Now I see "INFO zen.hub: building pingtree from cml-zensvr.as.local"
[04-Aug-2009 09:58:03] <sto6ma9ch> After that, I receive these messages:
[04-Aug-2009 09:58:21] <sto6ma9ch> "'router1.as.local' network '10.90.16.0/20' not in topology"
[04-Aug-2009 10:00:06] <sto6ma9ch> Each of the networks that are "not in the topology" are listed in the NEtworks section of Zenoss
[04-Aug-2009 10:02:17] <sto6ma9ch> Not sure how to add these networks to the topology. Any advice?
[04-Aug-2009 10:15:41] <sto6ma9ch> I just received the zenperfsnmp heartbeat failure message again after restarting the service
[04-Aug-2009 10:26:49] <DesignsEdg1> Hello all !
[04-Aug-2009 10:30:19] <Dieterbe> hi
[04-Aug-2009 10:53:13] <mrayzenoss> gotta go do the Getting Started Webinar, bbiab
[04-Aug-2009 10:57:21] <Dieterbe> if there are syntax errors in event transforms, are these logged somewhere?
[04-Aug-2009 11:04:54] <jsm> I applied a set of Collector Plugins to /Servers/Linux, but they don't seem to get inherited by its sub-classes. why?
[04-Aug-2009 11:14:44] <al_x> how can i change the links in the emails zenoss sends out so they contain an actual ip, and not localhost:8080
[04-Aug-2009 11:16:10] <chudler> al_x: set zopeurl in zenactions.conf
[04-Aug-2009 11:21:52] <al_x> ever seen this before
[04-Aug-2009 11:21:53] <al_x> snmp_build: unknown failuresnmpwalk: Error building ASN.1 representation (Can't build OID for variable
[04-Aug-2009 11:22:50] <al_x> oh nm heh
[04-Aug-2009 11:28:00] <syntaxcollector> We're running Zenoss open source version at our company, we have two locations our office and colo. I want to deploy a second collector at the colo and be able to read the data off of our primary zenoss instance at the office.
[04-Aug-2009 11:28:12] <syntaxcollector> can we do this with the open source edition? the manual seems to say no
[04-Aug-2009 11:29:36] <al_x> chudler after i set zopeurl do i need to restart zenoss?
[04-Aug-2009 11:29:53] <chudler> probably, at least zenactions.
[04-Aug-2009 11:38:05] <al_x> is there a way to test-fire an email
[04-Aug-2009 11:38:15] <al_x> or is it easier just to block snmp at the firewall and have it trip 'naturally'
[04-Aug-2009 12:16:32] <mrayzenoss> jsm: I'm not sure collector plugins get applied post-creation of the subclass
[04-Aug-2009 12:17:33] <mrayzenoss> syntaxcollector: you can manually do it, there are assorted docs in the community for it: http://delicious.com/mray/collectors/
[04-Aug-2009 12:20:42] <syntaxcollector> mrayzenoss: thanks, I was looking at that first link earlier today
[04-Aug-2009 12:51:39] <jsm> mrayzenoss: yeah, they aren't, grumble.. i'm just deleting the devices and re-adding them
[04-Aug-2009 12:51:51] <jsm> you'd think it would be a feature
[04-Aug-2009 12:51:56] <jsm> sounds like a bug to me
[04-Aug-2009 12:55:56] <brice> What does an Authentication Trap Failure received on the snmptrap service mean?
[04-Aug-2009 12:56:20] <jsm> that your snmp trap failed to authenticate i'd guess
[04-Aug-2009 12:57:06] <brice> so mismatched community names between myd evice and trap service?
[04-Aug-2009 12:59:08] <jb> anybody here running IBM xseries boxes?
[04-Aug-2009 13:00:15] <brice> older ones
[04-Aug-2009 13:00:24] <jb> do you run director on them?
[04-Aug-2009 13:00:29] <brice> i dont
[04-Aug-2009 13:00:39] <jb> we are a dell shop (we run openmanage).. and i'm trying to figure out how director works
[04-Aug-2009 13:20:14] <brice> nm I figured it out
[04-Aug-2009 13:20:36] <brice> hrm was hoping the few cheap dell printers we have would send an snmp trap back that it completed a print job of x pages
[04-Aug-2009 13:20:53] <mrayzenoss> jsm: open a ticket for that, I can see arguments on both sides for it
[04-Aug-2009 13:20:54] <brice> to confirm against whose i'm pulling from WMI
[04-Aug-2009 14:20:21] <gwb2353> is this the command you would use to figure out why zenoss says SNMP isn't responding, but you can snmpwalk successfully from the command line of the collector?
[04-Aug-2009 14:20:35] <gwb2353> [zenoss@eye ~]$ zenperfsnmp run -v 10 --monitor cornea --device com10-chicken-z1r3
[04-Aug-2009 14:20:59] <gwb2353> (where eye is the hub and cornea is a collector)
[04-Aug-2009 14:21:38] <gwb2353> tcpdump is telling me it's not quite right (no traffic from collector to device when command is issued)
[04-Aug-2009 14:24:44] <gwb2353> note that when I model the device, everything works OK.
[04-Aug-2009 14:25:29] <gwb2353> just can't get the "/Status/SNMP SNMP agent down" even to quit showing up
[04-Aug-2009 14:28:41] <jb> twm?
[04-Aug-2009 14:28:47] <jb> anybody here know of a good windows based MIB browser?
[04-Aug-2009 14:39:58] <rhett> hmm. not really. there is one i used thats old but i dont remember what its called
[04-Aug-2009 14:45:05] <chudler> I hadn't noticed the "watchdog" and "watchdogPath" options in the daemon configs before. I can't find documentation for it, either. Anyone can provide a link?
[04-Aug-2009 14:45:14] <gwb2353> [zenoss@cornea etc]$ zenmodeler run -d com10-chicken-z1r3 -v 10 --hubhost=eye runs properly (no errors)
[04-Aug-2009 14:54:51] <gwb2353> getting further (I think) "[zenoss@cornea etc]$ zenperfsnmp run -d com10-chicken-z1r3 -v 10 --hubhost=eye --configfile=/opt/zenoss/etc/cornea_zenperfsnmp.conf"
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:Fetching config for ['com10-chicken-z1r3']
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:Getting device ping issues
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:unresponsive devices: [['com10-chicken-z2r2', 1, 3]]
[04-Aug-2009 14:55:16] <gwb2353> DEBUG:zen.zenperfsnmp:Sent a 'stop' event
[04-Aug-2009 14:56:46] <jb> can you snmpwalk it?
[04-Aug-2009 14:56:53] <gwb2353> (note: ping from the command line works, and snmpwalk is fine)
[04-Aug-2009 14:57:17] <jb> and you have the SNMP community set properly?
[04-Aug-2009 14:57:42] <gwb2353> yes - i can use "zenmodeler" and everything is good and double-checked zproperties
[04-Aug-2009 14:58:20] <jb> hmm
[04-Aug-2009 15:01:51] <gwb2353> works (5/5): zenmodeler run -d com10-chicken-z1r3 -v 10 --hubhost=eye --configfile=/opt/zenoss/etc/cornea_zenperfsnmp.conf
[04-Aug-2009 15:02:16] <gwb2353> fails (5/5): zenperfsnmp run -d com10-chicken-z1r3 -v 10 --hubhost=eye --configfile=/opt/zenoss/etc/cornea_zenperfsnmp.conf
[04-Aug-2009 15:02:39] <gwb2353> ...with the "device ping issues"
[04-Aug-2009 15:17:19] <twm1010> afternoon
[04-Aug-2009 15:17:29] <twm1010> anytone know how to test the apache monitor via command line?
[04-Aug-2009 15:36:47] <jb> hey matt
[04-Aug-2009 15:37:21] <mrayzenoss> howdy
[04-Aug-2009 15:37:28] <jb> not much.. working on a IBM zenpack
[04-Aug-2009 15:37:46] <jb> quick q.. when I edit a modeler plugin .py.. what daemons need to be restarted?
[04-Aug-2009 15:37:59] <mrayzenoss> Egor has an IBMMon he's working on as well
[04-Aug-2009 15:38:04] <jb> oh crap
[04-Aug-2009 15:38:04] <jb> hah
[04-Aug-2009 15:38:06] <mrayzenoss> not sure what it's for
[04-Aug-2009 15:38:27] <jb> is egor on IRC?
[04-Aug-2009 15:38:34] <mrayzenoss> yeah, he's bigegor_
[04-Aug-2009 15:38:48] <jb> wonder if thats the best way to get ahold of him
[04-Aug-2009 15:46:37] <twm1010> i'd think the _ means he is away
[04-Aug-2009 15:50:45] <twm1010> mrayzenoss: is there a built-in method to track latency?
[04-Aug-2009 15:51:09] <twm1010> I see the fping zenpack, that seems like overkill, more datapoints than i'm after
[04-Aug-2009 15:51:25] <mrayzenoss> twm1010: not that I'm aware of...
[04-Aug-2009 15:51:51] <gwb2353> i'm using the ZenPacks.Nagios.CheckPing; pretty simple to setup
[04-Aug-2009 15:51:58] <mrayzenoss> there was/is an fping collector associated with the Dev Guide that never got finished
[04-Aug-2009 15:52:17] <mrayzenoss> it's the example for writing a new collector
[04-Aug-2009 15:52:27] <mrayzenoss> but it never made it into the product
[04-Aug-2009 15:52:42] <mrayzenoss> it would be lighter weight than the command fping or the Nagios CheckPing
[04-Aug-2009 15:53:56] <twm1010> Hrmm.. i think the nagios one will cut it just fine.
[04-Aug-2009 16:05:31] <gwb2353> anybody use the SentryCDU zenpack (under development) here? https://webshare.uchicago.edu/users/chudler/zenoss/
[04-Aug-2009 16:05:46] <chudler> yes :-p
[04-Aug-2009 16:06:08] <gwb2353> doh! it's that annoying guy from UT that keeps on emailing you!
[04-Aug-2009 16:06:25] <mrayzenoss> chudler: just 10 more days until the end of the ZenPack contest
[04-Aug-2009 16:06:42] <jb> mrayzenoss: whats the current URL to the zenpack documentation?
[04-Aug-2009 16:06:47] <chudler> no worries. I am <-> this close to having something to release. I am overworked lately
[04-Aug-2009 16:06:49] <jb> notes on developing zenpacks, etc
[04-Aug-2009 16:07:13] <mrayzenoss> jb: http://www.zenoss.com/community/docs/ Dev Guide and Extended Monitoring Guide
[04-Aug-2009 16:07:31] <jb> thanks
[04-Aug-2009 16:07:37] <mrayzenoss> jb: more here: trac-zenpacks/
[04-Aug-2009 16:08:21] <gwb2353> if i install what you currently have posted, will that mess things up later? (just curious to see what I can get from all of our Sentry devices)
[04-Aug-2009 16:08:36] <gwb2353> (non-production monitoring system, under deployment right now)
[04-Aug-2009 16:09:08] <mrayzenoss> gwb2353: I can bump the version on the Sentry ZenPack when chudler sends it in, so you "upgrade" to whatever is the latest
[04-Aug-2009 16:09:33] <chudler> I suppose it would work. I have no idea.
[04-Aug-2009 16:10:16] <chudler> I dont know what actually happens when a zenpack is updated. If you install the next test one, it probably is better to install clean, otherwise its hard to know waht we are testing.
[04-Aug-2009 16:10:34] <chudler> But in the release version, do whatever matt says :-)
[04-Aug-2009 16:10:56] <gwb2353> no worries; i'm just itching to see what the power draw is in our DC
[04-Aug-2009 16:11:04] <gwb2353> (current draw)
[04-Aug-2009 16:14:39] <chudler> gwb2353: could you tell me the vendor of those?
[04-Aug-2009 16:16:54] <gwb2353> http://products.servertech.com/products/smartpower/
[04-Aug-2009 16:17:51] <chudler> thats good. some resellers puts the sentry3 mib under their oid, which the pack handles but it has to be added if I don't already know about it. I think this vendor is fine.
[04-Aug-2009 16:49:21] <jb> that was odd
[04-Aug-2009 16:49:21] <jb>
[04-Aug-2009 21:47:55] <duffyd> hi all, I'm getting this error when trying to run the stack installer: bash: ./zenoss-stack-2.4.2-linux.bin: No such file or directory
[04-Aug-2009 21:48:03] <duffyd> which is really strange given that the file is there and the perms are all fine
[04-Aug-2009 21:48:10] <duffyd> is there some particular pre-reqs I need
[04-Aug-2009 21:48:15] <duffyd> I'm running ubuntu 8.04
[04-Aug-2009 22:02:07] edwardam1 is now known as edwardam
[04-Aug-2009 22:10:37] <duffyd> what are the pre-reqs for the stack installer?
[05-Aug-2009 00:00:46] [disconnected at Wed Aug 5 00:00:46 2009]
[05-Aug-2009 00:00:46] [connected at Wed Aug 5 00:00:46 2009]
[05-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[05-Aug-2009 02:37:47] <duffyd> hi what would be the best way to setup a snmp service via an ssh tunnel?
[05-Aug-2009 02:39:56] <duffyd> i.e. zenoss uses the ssh tunnel to check the snmp services on a remote box
[05-Aug-2009 03:07:23] <duffyd> does zenoss use snmp for all its linux device communication?
[05-Aug-2009 03:24:24] <rhett> you can do monitoring over ssh
[05-Aug-2009 03:24:36] <rhett> i personally use ssh though for all my hosts
[05-Aug-2009 03:30:28] <duffyd> rhett: really
[05-Aug-2009 03:30:34] <duffyd> how do you set that up
[05-Aug-2009 03:30:44] <duffyd> all the stuff I've read indicates its a maintenance nightmare
[05-Aug-2009 03:34:37] <rhett> oh sorry. i mean to say i use snmp for all my hosts
[05-Aug-2009 03:35:04] <rhett> i've only used ssh monitoring once on a test vm and didnt really go much further with it
[05-Aug-2009 03:42:17] <duffyd> rhett: ok thanks for that
[05-Aug-2009 08:12:09] <twm1010> duffyd: still around?
[05-Aug-2009 08:12:17] <duffyd> aye
[05-Aug-2009 08:12:53] <duffyd> twm1010:
[05-Aug-2009 08:12:55] <twm1010> as far as pre-reqs for stack, not really any different than from source
[05-Aug-2009 08:13:24] <twm1010> the whole point of stack, is that it comes with everything ya need
[05-Aug-2009 08:13:46] <twm1010> now, hardware pre-reqs are something else, a zenoss box should have some decent CPU and ram, and fast disk
[05-Aug-2009 08:17:19] <Dieterbe> is it just me or don't event transformation rules get inherited from the root class if your class is /Unknown?
[05-Aug-2009 08:20:57] <duffyd> twm1010: yeah my issue was that I was using the 32-bit installer on a 64-bit machine
[05-Aug-2009 08:31:49] <klinstifen> morning all
[05-Aug-2009 08:32:15] <klinstifen> my events db has grown way too big
[05-Aug-2009 08:32:28] <klinstifen> the back up last night has events.sql at ~12GB
[05-Aug-2009 08:32:42] <twm1010> that'll do it every time
[05-Aug-2009 08:32:43] <klinstifen> does the "clear event cache" in the zenoss console flush the entire db?
[05-Aug-2009 08:32:51] <twm1010> no, i don't think so
[05-Aug-2009 08:32:54] <klinstifen> hrm
[05-Aug-2009 08:33:20] <twm1010> there is a line: Delete Historical Events Older Than (days)
[05-Aug-2009 08:33:41] <twm1010> what i don't know, is when it does that, and if it truncates the DB afterwards
[05-Aug-2009 08:33:47] <twm1010> probably not
[05-Aug-2009 08:36:25] <klinstifen> my previous backups were only 300MB
[05-Aug-2009 08:36:29] <klinstifen> but that's the zipped file
[05-Aug-2009 08:36:39] <klinstifen> i am restoring one now to see how big it is uncompressed
[05-Aug-2009 08:36:57] <klinstifen> i can't complete the backup job now because i run out of disk space
[05-Aug-2009 08:37:12] <klinstifen> and then zenoss ends up dying later because it can't write to the temp file : /
[05-Aug-2009 08:37:52] <twm1010> let me see how big mine is
[05-Aug-2009 08:38:03] <jb> klinstifen: you can specify the temp file..
[05-Aug-2009 08:38:05] <klinstifen> cool thx
[05-Aug-2009 08:38:07] <klinstifen> yeah
[05-Aug-2009 08:38:08] <jb> if it is filling /tmp up
[05-Aug-2009 08:38:16] <klinstifen> it is
[05-Aug-2009 08:38:22] <jb> but it still uses /tmp in my experience
[05-Aug-2009 08:38:26] <jb> i think it may be a bug in 2.4.2
[05-Aug-2009 08:38:47] <klinstifen> the zenbackup is writing to /tmp
[05-Aug-2009 08:39:02] <klinstifen> "/tmp/tmpmnP5Ux/zenbackup
[05-Aug-2009 08:39:14] <klinstifen> and in there the events.sql is ~12GB
[05-Aug-2009 08:39:27] <jb> are you specifying --temp-dir on zenbackup?
[05-Aug-2009 08:39:32] <klinstifen> let me check
[05-Aug-2009 08:39:53] <klinstifen> no i am not
[05-Aug-2009 08:39:58] <jb> give that a try
[05-Aug-2009 08:40:04] <klinstifen> just the standard "zenbackup --save-mysql-access"
[05-Aug-2009 08:40:08] <twm1010> hrmm... I installed stack to /apps/zenoss
[05-Aug-2009 08:40:12] <twm1010> where would i find my events.sql?
[05-Aug-2009 08:40:54] <twm1010> found this: -rw-rw---- 1 mysql mysql 1881145344 Aug 5 08:40 ibdata1
[05-Aug-2009 08:41:04] <jb> events.sql is just a dump
[05-Aug-2009 08:41:06] <jb> that zenbackup creates
[05-Aug-2009 08:41:22] <klinstifen> you could see how big the events db is by going into mysql
[05-Aug-2009 08:41:28] <klinstifen> assuming you are running mysql
[05-Aug-2009 08:41:39] <twm1010> well im running the mysql instance thats part of the stack
[05-Aug-2009 08:41:56] <klinstifen> yeah
[05-Aug-2009 08:42:04] <klinstifen> so am i
[05-Aug-2009 08:42:14] <twm1010> running a backup right now... let's see
[05-Aug-2009 08:42:52] <klinstifen> twm1010: mysql -u zenoss -p
[05-Aug-2009 08:42:58] <klinstifen> and then the password you created during the install
[05-Aug-2009 08:44:06] <klinstifen> http://forums.mysql.com/read.php?108,201578,201578
[05-Aug-2009 08:44:13] <klinstifen> just ran that command against my events db
[05-Aug-2009 08:44:21] <klinstifen> and it is only ~3GB
[05-Aug-2009 08:44:29] <klinstifen> yet the dumb is ~12GB?
[05-Aug-2009 08:45:03] <klinstifen> restore complete
[05-Aug-2009 08:45:15] <klinstifen> untar'ing
[05-Aug-2009 08:48:31] <mrayzenoss> Good morning
[05-Aug-2009 08:48:35] <ckrough> sup
[05-Aug-2009 08:49:02] <mrayzenoss> in case you missed it, another new WMI ZenPack last night: http://www.zenoss.com/community/projects/zenpacks/wmiperf_iis6
[05-Aug-2009 08:49:27] <twm1010> excellent, very nice.
[05-Aug-2009 08:49:39] <twm1010> I might have a chance to use that... I spent yesterday getting our Apache monitoring going
[05-Aug-2009 08:50:43] <twm1010> Matt: I've never seen WMI as a data source type?
[05-Aug-2009 08:50:49] <twm1010> will that zenpack work on core?
[05-Aug-2009 08:50:54] <jb> yes
[05-Aug-2009 08:51:00] <jb> egor created the WMI datasource type
[05-Aug-2009 08:51:07] <jb> it depends on his WMI pack
[05-Aug-2009 08:51:54] <twm1010> wait, so if i install his WMI zenpack I don't have to write all these damn perl scripts for getting WMI information?
[05-Aug-2009 08:52:00] <twm1010> *bangs head on desk*
[05-Aug-2009 08:52:12] <jb> http://www.zenoss.com/community/projects/zenpacks/wmidatasource
[05-Aug-2009 08:52:18] <jb> correct
[05-Aug-2009 08:52:20] <jb> just the WMI queries
[05-Aug-2009 08:52:23] <twm1010> <-- jackass
[05-Aug-2009 08:52:42] <twm1010> sigh... well, at least i learned something
[05-Aug-2009 08:52:46] <jb>
[05-Aug-2009 08:52:47] <ckrough> there ya go
[05-Aug-2009 08:53:04] <ckrough> ... if I could take make all of the things I've done the hard way...
[05-Aug-2009 08:54:31] <twm1010> klinstifen: how long did yuor backup take?
[05-Aug-2009 08:54:39] <twm1010> my timeout is set to 180seconds and it still times out
[05-Aug-2009 08:55:16] <klinstifen> let me check some time stamps
[05-Aug-2009 08:55:39] <klinstifen> let's see...it starts at 12:20AM
[05-Aug-2009 08:56:13] <klinstifen> haha i mean 8:00PM
[05-Aug-2009 08:56:24] <klinstifen> and the modify stamp says 9:03
[05-Aug-2009 08:56:27] <klinstifen> so...about an hour
[05-Aug-2009 08:57:22] <twm1010> wow
[05-Aug-2009 08:57:23] <twm1010> ok
[05-Aug-2009 08:57:32] <twm1010> WMI datasource zenpack fails install: ERROR: zenpack command failed. Reason: [Errno 20] Not a directory: '/apps/zenoss/zenoss/Products/COPYRIGHT.txt/skins'
[05-Aug-2009 08:59:00] <ckrough> anyone here running rancid+zenoss integration?
[05-Aug-2009 09:04:48] <klinstifen> mray: my events db has grown too big..if I use the "clear event cache" will it flush the entire db?
[05-Aug-2009 09:05:24] <klinstifen> jb: i would specify another path but there is only one disk on my zenoss box
[05-Aug-2009 09:05:50] <jb> oh.. /tmp is the same filesystem as everything else?
[05-Aug-2009 09:06:12] <klinstifen> yeah
[05-Aug-2009 09:06:15] <jb> gotcha
[05-Aug-2009 09:06:23] <jb> then you have a problem
[05-Aug-2009 09:06:39] <klinstifen> yeah
[05-Aug-2009 09:06:47] <twm1010> can you do an NFS mount to something else?
[05-Aug-2009 09:07:20] <klinstifen> i could
[05-Aug-2009 09:07:25] <klinstifen> and for the sake of the backups that might work
[05-Aug-2009 09:07:41] <klinstifen> i'm just surprised that the events.sql is so large
[05-Aug-2009 09:07:48] <klinstifen> when the events db itself doesnt' seem to be so big
[05-Aug-2009 09:07:50] <mrayzenoss> klinstifen: you can get into your events DB and clean it up
[05-Aug-2009 09:07:56] <mrayzenoss> http://www.zenoss.com/Members/netdata/how-to-clean-your-events-database/
[05-Aug-2009 09:08:27] <klinstifen> mray: ok that's what i was thinking
[05-Aug-2009 09:09:53] <klinstifen> should the events.sql be about the same size as the actual events db?
[05-Aug-2009 09:10:58] <mrayzenoss> well, the events DB has several tables (http://dev.zenoss.org/trac/browser/branches/zenoss-2.4.x/Products/ZenEvents/db/zenevents.sql)
[05-Aug-2009 09:11:13] <mrayzenoss> and there will probably be a lot more in history than in status
[05-Aug-2009 09:17:04] <Dieterbe> how should i check the memory consumption of processes matching a certain regex on a windows box? i can't use zenoss process monitoring because zenoss will add up the usages of all processes that match the regex together. and i want to use WMI
[05-Aug-2009 09:17:33] <klinstifen> mray: ok...i was using this to calculate size of entire events db
[05-Aug-2009 09:17:35] <klinstifen> http://forums.mysql.com/read.php?108,201578,201578#msg-201578
[05-Aug-2009 09:18:23] <klinstifen> maybe that isn't counting the right information?
[05-Aug-2009 09:19:24] <klinstifen> jb & twm1010: looks like i have to use an nfs mount (or find other things to delete from this box) because the restore of my backup from a week ago is pretty large as well
[05-Aug-2009 09:19:48] <klinstifen> the tar file is small (300MB) but untarred it's pretty big
[05-Aug-2009 09:23:13] <klinstifen> mray: woops, bad math on my part...the events db is ~3gb and the events.sql is ~1.2GB
[05-Aug-2009 09:23:31] <klinstifen> the big file is the .fs file
[05-Aug-2009 09:27:26] <mrayzenoss> the .fs is the config DB, zope's object database
[05-Aug-2009 09:32:18] <Dieterbe> anyone knows the answer to my question?
[05-Aug-2009 09:34:11] <mrayzenoss> Dieterbe: there is no process monitoring via WMI yet, only service monitoring
[05-Aug-2009 09:35:28] <Dieterbe> well, i'm okay with writing my own script / datasource command
[05-Aug-2009 09:35:55] <Dieterbe> as long as i can pull the values from the windows box, and use the data in zenoss for graphing and alerting
[05-Aug-2009 09:35:58] <mrayzenoss> then you'd probably want to check out the new WMI data source
[05-Aug-2009 09:36:31] <jb> http://www.zenoss.com/community/projects/zenpacks/wmidatasource
[05-Aug-2009 09:36:43] <klinstifen> or use wmic?
[05-Aug-2009 09:37:13] <mrayzenoss> klinstifen: the new WMI data source should replace the need for wmic
[05-Aug-2009 09:37:26] <mrayzenoss> I assume you can craft WQL queries to get the processes
[05-Aug-2009 09:37:43] <klinstifen> it's that so? hmm...must download
[05-Aug-2009 09:38:06] <mrayzenoss> yeah, it's supposedly faster and handles credentials better than wmic-based stuff
[05-Aug-2009 09:38:11] <mrayzenoss> same calls under the covers
[05-Aug-2009 09:38:13] <klinstifen> awesome
[05-Aug-2009 09:38:18] <klinstifen> yeah wmic was odd
[05-Aug-2009 09:38:26] <klinstifen> you had to flip the slash in domain/username
[05-Aug-2009 09:38:46] <klinstifen> even though zenoss used it the other way for grabbing sys logs
[05-Aug-2009 09:38:48] <klinstifen> cool
[05-Aug-2009 09:38:52] <klinstifen> screenshots look promising
[05-Aug-2009 09:39:01] <klinstifen> i'll have to try this out and convert some of my personal zenpacks
[05-Aug-2009 09:40:15] <Dieterbe> hmm i already use 2 wmi zenpacks ( http://blog.dastrup.com/?p=13 and http://www.zenoss.com/community/projects/zenpacks/windowswmidevicetemplate ) won't this new one conflict/break anyone in my current setup?
[05-Aug-2009 09:40:17] <mrayzenoss> gotta run, be back in a bit
[05-Aug-2009 09:40:48] <mrayzenoss> Dieterbe: I believe the new WMI ones from Egor Puzanov are meant to replace those
[05-Aug-2009 09:41:08] <mrayzenoss> so far there's a Windows Perf and an IIS Perf
[05-Aug-2009 09:41:30] <mrayzenoss> in addition to the WMI data source they both use
[05-Aug-2009 09:41:44] <mrayzenoss> so someone could do an Exchange, AD, whatever else you run on MS
[05-Aug-2009 09:42:31] <mrayzenoss> and unlike the 2 you mentioned, the new WMI-based ones should be faster and more secure (ie. no passwords in logs)
[05-Aug-2009 09:43:06] <mrayzenoss> gotta go
[05-Aug-2009 10:13:10] <Dieterbe> hmm odd, i'm googling with terms like 'wql memory usage', 'wql measure memory etc' but i don't have any good results :/ can anyone tell me where i should look?
[05-Aug-2009 10:13:20] <Dieterbe> (to measure memory usage of single processes with wmi/wql)
[05-Aug-2009 10:51:11] <venturaville> does anyone remember who was working on the IBM/LSI storage array module?
[05-Aug-2009 10:51:30] <ckrough> Egor?
[05-Aug-2009 10:51:54] <ckrough> jb as well
[05-Aug-2009 10:53:37] <venturaville> actually it was the IBM SVC we were discussing
[05-Aug-2009 10:53:43] <venturaville> and it was zenChild
[05-Aug-2009 10:53:48] <venturaville> doesn't look like he is on now though
[05-Aug-2009 10:56:33] <jb> ventura
[05-Aug-2009 10:56:39] <jb> im TRYING to work on IBM stuff right now
[05-Aug-2009 10:56:49] <jb> im about to scream though.. IBM director is driving me bonkers.
[05-Aug-2009 10:57:12] <jb> but i'd be interested in what you are tryign to do with SVC as well
[05-Aug-2009 10:58:24] <venturaville> zenChild is working on an IBM SVC zenpack: http://github.com/zenchild/ZenPacks.ndgov.IBMsvc/tree/master
[05-Aug-2009 10:58:49] <venturaville> I was hoping to get caught up with someone's efforts now that I have the time
[05-Aug-2009 10:58:51] <jb> nice!
[05-Aug-2009 10:59:05] <jb> well, egor and I are working on an IBMMon zenpack
[05-Aug-2009 10:59:11] <venturaville> I have a list of commands to extract data from SMcli
[05-Aug-2009 10:59:14] <jb> do you have any experience with IBM Director Agents?
[05-Aug-2009 10:59:44] <venturaville> zilch
[05-Aug-2009 10:59:50] <venturaville> we mostly use Storage Manager
[05-Aug-2009 11:00:07] <jb> yeah
[05-Aug-2009 11:00:10] <jb> we are a SVC shop here
[05-Aug-2009 11:00:13] <jb> that pack looks promising
[05-Aug-2009 11:00:17] <jb> it uses SNMP?
[05-Aug-2009 11:00:17] <venturaville> we have SVCs
[05-Aug-2009 11:00:21] <venturaville> sadly no
[05-Aug-2009 11:00:35] <venturaville> the SVCs and LSI based arrays are SNMP dumb
[05-Aug-2009 11:00:48] <jb> cimon?
[05-Aug-2009 11:00:56] <venturaville> and it looks as though going through the CIMON agents doesn't give me performance data
[05-Aug-2009 11:01:03] <venturaville> only structural and config data
[05-Aug-2009 11:01:15] <jb> how is zenchild modeling the existing data?
[05-Aug-2009 11:01:32] <venturaville> svcinfo commands from the looks of things
[05-Aug-2009 11:01:43] <jb> via ssh?
[05-Aug-2009 11:01:53] <venturaville> not sure
[05-Aug-2009 11:01:59] <venturaville> I'll download the update and take a look
[05-Aug-2009 11:02:03] <jb> k
[05-Aug-2009 11:02:11] <jb> i despise this IBM director bs
[05-Aug-2009 11:02:15] <venturaville> +1
[05-Aug-2009 11:02:22] <jb> why can't they have a simple solution like HP/Dell
[05-Aug-2009 11:02:23] <jb> GRR
[05-Aug-2009 11:02:33] <jb> anything IBM has to be so damn complicated..
[05-Aug-2009 11:02:39] <venturaville> I got some SMcli commands for the arrays themselves from the LSI folk
[05-Aug-2009 11:02:45] <venturaville> if you have access to storage manager
[05-Aug-2009 11:02:53] <jb> i actually was just upgrading my SM
[05-Aug-2009 11:02:56] <venturaville> as for the SVCs we might be stuck with direct ssh access for perf data
[05-Aug-2009 11:03:29] <venturaville> take a look at yawn if you have not used it before
[05-Aug-2009 11:03:36] <venturaville> I think zenChild was the one who pointed me at it
[05-Aug-2009 11:03:40] <jb> yawn?
[05-Aug-2009 11:03:51] <venturaville> http://sourceforge.net/apps/mediawiki/pywbem/index.php?title=YAWN
[05-Aug-2009 11:04:05] <jb> ahh
[05-Aug-2009 11:04:12] <venturaville> works pretty well
[05-Aug-2009 11:04:24] <venturaville> use root/lsissi
[05-Aug-2009 11:06:36] <venturaville> just need to point it at one of your CIMOM servers
[05-Aug-2009 11:09:25] <jb> k.. i'll try that in a bit
[05-Aug-2009 11:28:14] <jb> hi egor..
[05-Aug-2009 11:42:02] <otakup0pe> anyone good with the web services around
[05-Aug-2009 11:42:17] <otakup0pe> how does one easily query devices for os/status/hw info
[05-Aug-2009 11:53:56] <twm1010> anyone had issues installing the WMI datasource ZP on stack?
[05-Aug-2009 12:01:49] <klinstifen> mine went through without issue
[05-Aug-2009 12:02:02] <klinstifen> gonna test it out as soon
[05-Aug-2009 12:10:09] <twm1010> hrmm, i'll try it on my dev machine
[05-Aug-2009 12:11:31] <twm1010> same error on stack install on SLES as on RHEL
[05-Aug-2009 12:50:34] <venturaville> jb: which director version are you using?
[05-Aug-2009 12:50:46] <jb> trying to use 6.11
[05-Aug-2009 12:50:49] <jb> platform agent
[05-Aug-2009 12:50:59] <venturaville> SNMPv2-SMI::enterprises.2.6.159.200.3.1.1.1.1.1 = STRING: "International Business Machines Corporation"
[05-Aug-2009 12:50:59] <venturaville> SNMPv2-SMI::enterprises.2.6.159.200.3.1.1.1.2.1 = STRING: "IBM Director Server"
[05-Aug-2009 12:50:59] <venturaville> SNMPv2-SMI::enterprises.2.6.159.200.3.1.1.1.3.1 = STRING: "4.20"
[05-Aug-2009 12:51:03] <jb> its not registering alerts/subscription with the cimom automatically
[05-Aug-2009 12:51:34] <jb> and a bunch of other issues
[05-Aug-2009 12:56:09] <venturaville> apparently our CIMOM agent is running on a nonstandard port for SVC
[05-Aug-2009 13:02:09] <mrayzenoss> otakup0pe: I tried to do stuff with web services, but it's easier to just put your own Python scripts on Zope
[05-Aug-2009 13:02:30] <mrayzenoss> you can easily expose whatever you want as blocks of HTML or whatever
[05-Aug-2009 13:04:22] <mrayzenoss> examples are here: http://zenossdeki.svn.sourceforge.net/viewvc/zenossdeki/zenoss/
[05-Aug-2009 13:04:57] <Garn32> When ever I attempt to create a backup using the GUI it fails with raise child_exception
[05-Aug-2009 13:04:57] <Garn32> OSError
[05-Aug-2009 13:04:57] <Garn32> :
[05-Aug-2009 13:04:57] <Garn32> [Errno 2] No such file or directory
[05-Aug-2009 13:05:24] <Garn32> is the backup function broken in the GUI of what?
[05-Aug-2009 13:07:23] <mrayzenoss> Garn32: which version?
[05-Aug-2009 13:08:03] <Garn32> 2.4.2
[05-Aug-2009 13:08:26] <mrayzenoss> not that I'm aware of, nothing in trac for that
[05-Aug-2009 13:08:37] <mrayzenoss> which installer?
[05-Aug-2009 13:08:45] <Garn32> its from source
[05-Aug-2009 13:09:24] <mrayzenoss> ahh, let me try my source installs
[05-Aug-2009 13:12:02] <mrayzenoss> mine works, Debian 5 x86 from source: http://pastebin.com/d8d19eec
[05-Aug-2009 13:12:11] <mrayzenoss> 2.4.2 from SVN
[05-Aug-2009 13:12:20] <mrayzenoss> pastebin the whole stack
[05-Aug-2009 13:12:34] <Garn32> Output:
[05-Aug-2009 13:12:35] <Garn32> /usr/local/zenoss/bin/zenbackup -v10 --save-mysql-access
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:Use /tmp/tmpt1BGMx/zenbackup as a staging directory for the backup
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:Can we access ZODB through Zeo?
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:/usr/local/zenoss/bin/python /usr/local/zenoss/bin/zeoup.py -p 8100 -h localhost
[05-Aug-2009 13:12:35] <Garn32> DEBUG:zenbackup:Elapsed time: 0.50
[05-Aug-2009 13:12:37] <Garn32> INFO:zenbackup:Getting MySQL dbname, user, password from ZODB.
[05-Aug-2009 13:12:39] <Garn32> DEBUG:zenbackup:Writing MySQL credentials to /tmp/tmpt1BGMx/zenbackup/backup.settings
[05-Aug-2009 13:12:41] <Garn32> INFO:zenbackup:Backing up events database.
[05-Aug-2009 13:12:43] <Garn32> DEBUG:zenbackup:mysqldump -uzenoss ***************** --single-transaction --routines events --result-file=/tmp/tmpt1BGMx/zenbackup/events.sql
[05-Aug-2009 13:12:46] <Garn32> Traceback (most recent call last):
[05-Aug-2009 13:12:48] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 420, in ?
[05-Aug-2009 13:12:50] <Garn32> if zb.makeBackup():
[05-Aug-2009 13:12:54] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 388, in makeBackup
[05-Aug-2009 13:12:56] <Garn32> self.backupEventsDatabase()
[05-Aug-2009 13:12:58] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 280, in backupEventsDatabase
[05-Aug-2009 13:13:00] <Garn32> (output, warnings, returncode) = self.runCommand(cmd, obfuscated_cmd)
[05-Aug-2009 13:13:02] <Garn32> File "/usr/local/zenoss/Products/ZenUtils/ZenBackup.py", line 62, in runCommand
[05-Aug-2009 13:13:04] <Garn32> proc = Popen(cmd, stdout=PIPE, stderr=PIPE)
[05-Aug-2009 13:13:06] <Garn32> File "/usr/local/zenoss/lib/python2.4/subprocess.py", line 543, in __init__
[05-Aug-2009 13:13:08] <Garn32> errread, errwrite)
[05-Aug-2009 13:13:10] <Garn32> File "/usr/local/zenoss/lib/python2.4/subprocess.py", line 975, in _execute_child
[05-Aug-2009 13:13:12] <Garn32> raise child_exception
[05-Aug-2009 13:13:14] <Garn32> OSError
[05-Aug-2009 13:13:16] <Garn32> :
[05-Aug-2009 13:13:18] <Garn32> [Errno 2] No such file or directory
[05-Aug-2009 13:13:20] <Garn32> DONE
[05-Aug-2009 13:13:24] <Garn32> sorry
[05-Aug-2009 13:14:29] <Garn32> http://pastebin.com/m3b9abe6b
[05-Aug-2009 13:14:40] <Garn32> for the output given to me
[05-Aug-2009 13:19:46] <mrayzenoss> hmm... is there a $ZENHOME/backups directory?
[05-Aug-2009 13:19:57] <Garn32> yes that directory exists
[05-Aug-2009 13:20:06] <mrayzenoss> and it's owned by the 'zenoss' user?
[05-Aug-2009 13:20:23] <Garn32> yes
[05-Aug-2009 13:20:57] <Garn32> the other odd thing is that the command line zenbackup works just fine
[05-Aug-2009 13:21:04] <mrayzenoss> that is odd
[05-Aug-2009 13:21:37] <mrayzenoss> so if you use the same /usr/local/zenoss/bin/zenbackup -v10 --save-mysql-access call as the zenoss user, everything works?
[05-Aug-2009 13:21:58] <Garn32> let me check again real quick
[05-Aug-2009 13:22:40] <Garn32> yep ran just fine
[05-Aug-2009 13:23:15] <mrayzenoss> I'll run it by the devs, but maybe open a ticket.
[05-Aug-2009 13:23:47] <Garn32> ok I'll open a ticket
[05-Aug-2009 13:23:57] <mrayzenoss> put that stack trace in there as well
[05-Aug-2009 13:24:06] <mrayzenoss> you are running 2.4.2 right, not trunk?
[05-Aug-2009 13:24:27] <mrayzenoss> just have to check with source installs
[05-Aug-2009 13:25:05] <twm1010> mrayzenoss: is there a more verbose way to troubleshoot a zenpack install?
[05-Aug-2009 13:25:06] <Garn32> not trunk
[05-Aug-2009 13:25:22] <mrayzenoss> if you go to Settings->Versions you should see the SVN number after the Zenoss version
[05-Aug-2009 13:25:25] <mrayzenoss> put that in the ticket too
[05-Aug-2009 13:25:28] <Garn32> ok
[05-Aug-2009 13:26:39] <mrayzenoss> twm1010: zenpack -v10 ZenPacks.community.yourzenpack.egg
[05-Aug-2009 13:27:40] <twm1010> i mean to say, that won't install
[05-Aug-2009 13:27:52] <mrayzenoss> pastebin the error
[05-Aug-2009 13:28:25] <mrayzenoss> Garn32: SELinux?
[05-Aug-2009 13:28:36] <Garn32> its on a debain system
[05-Aug-2009 13:28:44] <mrayzenoss> yeah, me too
[05-Aug-2009 13:28:46] <twm1010> http://pastebin.com/d11f04ac9
[05-Aug-2009 13:29:08] <Garn32> no SELinux installed
[05-Aug-2009 13:29:32] <mrayzenoss> twm1010: your .egg got renamed as a .zip
[05-Aug-2009 13:32:30] <twm1010> let me test with the different extension
[05-Aug-2009 13:32:38] <twm1010> thats the file as it came down from the zenpack site
[05-Aug-2009 13:33:41] <mrayzenoss> Garn32: any chance the path changed between starting zenoss and running the zenbackup from the command line?
[05-Aug-2009 13:33:58] <mrayzenoss> twm1010: are you using IE? it sometimes rewrites the extension
[05-Aug-2009 13:34:34] <Garn32> none that I can think of
[05-Aug-2009 13:35:04] <mrayzenoss> just passing on suggestions from kells and chet
[05-Aug-2009 13:35:19] <mrayzenoss> they surmise that mysqldump isn't in the zenoss application's path
[05-Aug-2009 13:36:40] <Garn32> ok, how would I go about fixing that?
[05-Aug-2009 13:37:27] <Garn32> should I still put the ticket in?
[05-Aug-2009 13:37:38] <mrayzenoss> here's my path settings
[05-Aug-2009 13:37:47] <mrayzenoss> export ZENHOME=/usr/local/zenoss
[05-Aug-2009 13:37:47] <mrayzenoss> export PYTHONPATH=$ZENHOME/lib/python
[05-Aug-2009 13:37:47] <mrayzenoss> export PATH=$ZENHOME/bin:$PATH
[05-Aug-2009 13:37:56] <twm1010> jeebus
[05-Aug-2009 13:37:56] <mrayzenoss> for my zenoss account
[05-Aug-2009 13:38:11] <twm1010> you are right, renamed to .egg, i guess did somehow end up using IE to get that...
[05-Aug-2009 13:38:25] <twm1010> sigh... all that... for that.
[05-Aug-2009 13:38:51] <mrayzenoss> Garn32: I followed this a bit as well: http://dev.zenoss.org/trac/browser/trunk/inst/docs/INSTALL_DEB-ETCH.txt
[05-Aug-2009 13:38:58] <mrayzenoss> it's a bit dated
[05-Aug-2009 13:39:25] <Garn32> mysqldump is in the zenoss users path
[05-Aug-2009 13:39:26] <mrayzenoss> but steps 4, 5, 6, 9, 10 and 11
[05-Aug-2009 13:39:34] <mrayzenoss> are worth double-checking
[05-Aug-2009 13:42:47] <Garn32> checking the sudo bit
[05-Aug-2009 13:45:14] <Garn32> nope everything looks fine from those steps
[05-Aug-2009 13:50:53] <Garn32> just put the trouble ticket it #5395
[05-Aug-2009 13:52:23] <mrayzenoss> thanks
[05-Aug-2009 13:55:49] <venturaville> mrayzenoss: sounds like the zenoss guys are swinging by our office next week :-)
[05-Aug-2009 13:55:59] <otakup0pe> mrayzenoss: yeah i realize that
[05-Aug-2009 13:56:08] <otakup0pe> i'll probably end up mixing it between that and submitting patches to the zenoss cats
[05-Aug-2009 13:56:10] <mrayzenoss> venturaville: who's that?
[05-Aug-2009 13:56:23] <ckrough> demand swag!
[05-Aug-2009 13:56:44] <venturaville> mrayzenoss: not sure who all is coming out ... sounds as though they are doing an enterprise demo then
[05-Aug-2009 13:57:08] <mrayzenoss> well, good luck to us!
[05-Aug-2009 13:58:02] <mrayzenoss> otakup0pe: yeah, I just found it easier to cheat and have Zope serve up exactly what I wanted rather than make a real REST call, parse it on the other side and then format it
[05-Aug-2009 14:01:21] <otakup0pe> mrayzenoss: yeah i'm kionda of the same opinion
[05-Aug-2009 14:01:40] <otakup0pe> when i say REST i really just mean "machine friendly web services with semanticly correct http response codes"
[05-Aug-2009 14:39:29] <straterra> How do I stop paging for one machine entirely?
[05-Aug-2009 14:39:53] <mrayzenoss> put it in a maintenance window
[05-Aug-2009 14:40:13] <mrayzenoss> create a Group for it, put the device in there and put a "Testing" maintenance window on just that device
[05-Aug-2009 14:40:19] <mrayzenoss> that never ends
[05-Aug-2009 14:40:31] <mrayzenoss> it'll still get monitored, just no alerts
[05-Aug-2009 14:40:35] <straterra> Hmm
[05-Aug-2009 14:40:52] <straterra> So..new group, drop the machine in there..and put a maintenance window on the group?
[05-Aug-2009 14:40:57] <mrayzenoss> yeah
[05-Aug-2009 14:41:22] <mrayzenoss> there may be other ways, that just popped into my head first
[05-Aug-2009 14:41:31] <mrayzenoss> you still want monitoring right?
[05-Aug-2009 14:42:01] <mrayzenoss> do you want events from it?
[05-Aug-2009 14:42:24] <straterra> I want monitoring..just no emails/pages
[05-Aug-2009 14:42:27] <straterra> At all
[05-Aug-2009 14:44:18] <mrayzenoss> I believe "Pre-Production" is monitoring with events with no alerts
[05-Aug-2009 14:44:52] <straterra> whats 'maintenance' ?
[05-Aug-2009 14:45:05] <mrayzenoss> heh, I knew you'd ask
[05-Aug-2009 14:45:20] <mrayzenoss> the Admin Guide has 3 of the 5
[05-Aug-2009 14:45:26] <mrayzenoss> I know I've opened a ticket for that somewhere
[05-Aug-2009 14:45:26] <straterra> Ah, ok
[05-Aug-2009 14:46:00] <mrayzenoss> there we go: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch08s02.html
[05-Aug-2009 14:46:56] <straterra> so..maintenance is it showing on the dashboard, monitoring, but just no paging
[05-Aug-2009 14:47:08] <mrayzenoss> yes
[05-Aug-2009 14:47:19] <straterra> (That wording is a bit..weird on the page)
[05-Aug-2009 14:47:25] <mrayzenoss> agreed
[05-Aug-2009 14:48:14] <straterra> So..for a maintenance window that never expires, I just make it reoccure daily..and last exactly 1 day long, right?
[05-Aug-2009 14:49:29] <mrayzenoss> yeah, that works. Seems there used to be a "Stop Production State" pulldown where you could set it to go into the same state
[05-Aug-2009 14:49:34] <mrayzenoss> maybe that was causing issues
[05-Aug-2009 14:50:52] <straterra> Thanks a lot
[05-Aug-2009 14:51:08] <mrayzenoss> no problem.
[05-Aug-2009 14:51:18] <straterra> Oh..one last thing. We have ZenOSS in a VM right now. Is it pretty easy to migrate everything to a real machine?
[05-Aug-2009 14:51:40] <mrayzenoss> yeah
[05-Aug-2009 14:52:08] <mrayzenoss> we should add that to the Install Guide
[05-Aug-2009 14:52:19] <straterra> Restoring from a backup?
[05-Aug-2009 14:52:31] <mrayzenoss> no, migrating from VM to other installers
[05-Aug-2009 14:52:35] <mrayzenoss> restoring is covered
[05-Aug-2009 14:52:40] <straterra> Ah, ok
[05-Aug-2009 14:52:45] <straterra> I'm not using the appliance
[05-Aug-2009 14:52:59] <straterra> I'm using a CentOS 5.3 VM...and the real machine will be CentOS 5.3 too
[05-Aug-2009 14:54:55] <ckrough> zenhub is finicky
[05-Aug-2009 14:56:10] <mrayzenoss> straterra: so that should be similar
[05-Aug-2009 14:56:16] <mrayzenoss> and we don't have great docs for that use case, but we should
[05-Aug-2009 14:56:21] <mrayzenoss> opening a ticket
[05-Aug-2009 14:56:37] <straterra> One last thing..for real this time, heh..Are the installed zenpacks located somewhere that I can easily copy them off for reinstallation before I import my backup?
[05-Aug-2009 14:56:59] <mrayzenoss> $ZENHOME/ZenPacks
[05-Aug-2009 14:57:56] <straterra> Oh, nifty. I do notice one of my zenpacks isn't in there though
[05-Aug-2009 14:58:02] <straterra> Not that I care..I don't use that one
[05-Aug-2009 14:58:11] <mrayzenoss> is it a .zip ZenPack?
[05-Aug-2009 14:58:18] <straterra> It may have been
[05-Aug-2009 14:58:39] <mrayzenoss> yeah, it probably installed it into $ZENHOME/Products, which is one of the reasons we stopped using them
[05-Aug-2009 14:59:19] <straterra> gotcha
[05-Aug-2009 14:59:37] <straterra> Would it be bad if I just didn't reinstall that zenpack, then?
[05-Aug-2009 14:59:56] <mrayzenoss> only if you have stuff that uses it
[05-Aug-2009 15:00:17] <venturaville> jb: if you get a chance to play with yawn, let me know if you come across any actual performance data... as near as I can tell it is getting it all from the array CIMOMs and not from SVC (darned if I can find it in the director CIMOM anyways)
[05-Aug-2009 15:21:29] <mrayzenoss> straterra: http://dev.zenoss.org/trac/ticket/5396
[05-Aug-2009 15:22:24] <mrayzenoss> straterra: it's quite involved if you really want to migrate everything
[05-Aug-2009 15:23:19] <ckrough> How can I avoid spikes in graphs when an interface counter rolls over due to a power outage or other abnormal event?
[05-Aug-2009 15:23:29] <straterra> Hmm :/
[05-Aug-2009 15:23:57] <straterra> Its involved to migrate from rpm install to rpm install?
[05-Aug-2009 15:24:21] <mrayzenoss> no, the most pain is 32 to 64 bit conversions
[05-Aug-2009 15:24:37] <straterra> Ah..that won't be happening
[05-Aug-2009 15:24:37] <mrayzenoss> which is typically what happens when people go VMware -> RPM
[05-Aug-2009 15:25:01] <straterra> I'm going from a heavily used (read as slow) VM server to a lowend 1U server
[05-Aug-2009 15:25:02] <twm1010> finally got that WMI datasource going, but queries don't work against any box
[05-Aug-2009 15:26:08] <mrayzenoss> straterra: here are the notes from professional services
[05-Aug-2009 15:26:17] <mrayzenoss> Here is a brief overview of some of the tools used for data import/export:
[05-Aug-2009 15:26:17] <mrayzenoss> • zendevicedump / zendeviceload Creates a XML file of device configuration information from the ZODB. Requires the same device classes on the target system as on the source system.
[05-Aug-2009 15:26:17] <mrayzenoss> • zendump / zenload Creates a XML file of the entire ZODB, including menu items. Requires the same version of Zenoss as well as the same ZenPacks on the target system as on the source system. Some XSL (or editing) is required to modify some paths that point to on-disk files.
[05-Aug-2009 15:26:17] <mrayzenoss> • zenbackup / zenrestore Used to recover in the event of a system failure. Requires the same version of Zenoss as well as the same ZenPacks on the target system as on the source system. This is capable of being used to migrate between OS versions on the same architecture (ie 32-bit to 32-bit, 64-bit to 64-bit)
[05-Aug-2009 15:26:21] <mrayzenoss> • rrdtool dump / import Used to convert RRD data to XML and back again.
[05-Aug-2009 15:26:23] <mrayzenoss> • mysqldump / mysql Used to export MySQL data from one system to another.
[05-Aug-2009 15:31:43] <straterra> I think I'd just use zenbackup
[05-Aug-2009 15:32:55] <mrayzenoss> yeah, same OS, same architecture, same path and same ZenPacks installed and you should be fine
[05-Aug-2009 15:33:30] <straterra> Yup yup yup and yup!
[05-Aug-2009 15:46:09] <chudler> I am curious, has there been any zenpacks or solutions for people with web environments running virtual hosts? I remember seeing an example zenpack a while back... did anyone run with it?
[05-Aug-2009 15:47:55] <chudler> it used to come up on the forum sometimes. its not trivial to monitor a large virtual hosting environment, since the IP addresses and hostnames are all shared by a single box
[05-Aug-2009 15:48:29] <mrayzenoss> yeah, I actually had a discussion with the dev working on VMware stuff regarding that
[05-Aug-2009 15:49:01] <mrayzenoss> it's not something we do well, so there may be some changes in the next release to make non-device things fit the model better
[05-Aug-2009 15:49:57] <chudler> that sounds really nice. for other things also. I might have a chance in the next few months to expand monitoring to our 1000+ virtual hosts, so I will either build it or keep waiting ;-)
[05-Aug-2009 15:50:18] <mrayzenoss> well, this would be for the Stone Crab release next Spring
[05-Aug-2009 15:50:31] <mrayzenoss> King Crab will be in beta in about a month
[05-Aug-2009 15:51:07] <chudler> can't wait. You guys don't sit still, thats for sure!
[05-Aug-2009 15:51:26] <mrayzenoss> the good news is we now have funding for another developer
[05-Aug-2009 15:53:06] <straterra> My software has releaes names of STD's..but crabs work too
[05-Aug-2009 15:53:38] <ckrough> you guys talking about VirtualHosts (httpd) or Virtual Hosts (vm/xen)?
[05-Aug-2009 15:54:06] <chudler> httpd
[05-Aug-2009 15:54:23] <mrayzenoss> chudler was talking httpd, but we realize we need to have a common way of representing multiple things under a single IP
[05-Aug-2009 15:54:56] <mrayzenoss> whether it's httpd, a cluster of databases or virtual hosts under ESX
[05-Aug-2009 15:55:16] <ckrough> gotcha
[05-Aug-2009 15:55:22] <chudler> ideally the devices could have templates, alert rules, and the whole nine "as if it was a device". But I am not holding my breath on that much of an overhaul
[05-Aug-2009 15:55:35] <ckrough> mac addresses are always unique...
[05-Aug-2009 15:56:00] <ckrough> nm
[05-Aug-2009 15:56:01] <mrayzenoss> ckrough: yeah, I think we tried that with VMware
[05-Aug-2009 15:56:08] <mrayzenoss> ckrough: but they don't have to be
[05-Aug-2009 15:56:23] <ckrough> same problem. I dont NAT with my vms... if you nat still have the same issue
[05-Aug-2009 15:56:29] <ckrough> and it does nothing for VIrtualHosts
[05-Aug-2009 15:56:34] <ckrough> fingers faster than brain
[05-Aug-2009 15:58:22] <mrayzenoss> yeah, I think we fixed an issue with VMware where we now support VMs with the same name and same IP address and same MACs under different ESX servers
[05-Aug-2009 16:08:37] <victor-> i've got a server that uses VPN to connect, each time launching an interface called 'tun0'. the problem is that zenoss is creating a unique 'tun0' interface in the device's "OS" tab. is there some way to only have zenoss recognize one 'tun0' device?
[05-Aug-2009 16:11:23] edwardam is now known as edwardam_phone
[05-Aug-2009 16:21:41] <rhett> hm, dont know about that one
[05-Aug-2009 16:22:01] <mrayzenoss> victor-: you could lock it from getting remodeled
[05-Aug-2009 16:23:44] <victor-> Ok, I was hoping that'll work. Thanks.
[05-Aug-2009 16:42:24] edwardam_phone is now known as edwardam
[05-Aug-2009 17:19:22] <mrayzenoss> just hit a very nice milestone: http://blog.zenoss.com/2009/08/05/zenoss-community-posts-100th-zenpack/
[05-Aug-2009 17:20:18] <rhett> nice
[05-Aug-2009 17:25:08] <locohost> gratz
[05-Aug-2009 22:10:06] <otakup0pe> awesome i managed to add new rest methods :3
[05-Aug-2009 22:10:11] <otakup0pe> who wants to know the secret
[05-Aug-2009 22:29:59] <rhett> :O
[05-Aug-2009 22:51:49] <otakup0pe> hehe zenoss doesn't like having the ZenModel product refreshed
[05-Aug-2009 22:51:54] <otakup0pe> but it makes testing web services sooo easy
[05-Aug-2009 23:25:23] <otakup0pe> http://i30.tinypic.com/v6n792.png from left to right load average, cpu load and memory
[06-Aug-2009 00:00:46] [disconnected at Thu Aug 6 00:00:46 2009]
[06-Aug-2009 00:00:46] [connected at Thu Aug 6 00:00:46 2009]
[06-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[06-Aug-2009 00:46:31] <rhett> is this second life
[06-Aug-2009 00:54:11] <sergeymasushko> rhett: what do you mean? 0o
[06-Aug-2009 01:50:37] <rhett> sergeymasushko [09:24] <otakup0pe> http://i30.tinypic.com/v6n792.png from left to right load average, cpu load and memory this
[06-Aug-2009 07:01:47] <roadies> what plugin am I missing to get the network map "Select Device or Map" to display
[06-Aug-2009 08:08:25] <otakup0pe> rhett: opensim, but close enough
[06-Aug-2009 08:27:09] <TBKDan> I'm trying to monitor a winservice on a particular box. When I go to add it on the device's OS tab, it finds it just fine in the "Add WinService" box, but when I click ok, it says "Could not find a WinService named ..." The service is definitely there on the server, and in /Services/WinServices. Any idea why it's saying it can't find it?
[06-Aug-2009 08:37:02] <twm1010> you want to add it to just the one box, right?
[06-Aug-2009 08:37:22] <TBKDan> Correct
[06-Aug-2009 08:37:51] <twm1010> I've been wondering how to do that, because if you change the zprop on the service, it monitors every instance
[06-Aug-2009 08:38:05] <TBKDan> Yup.
[06-Aug-2009 08:38:32] <TBKDan> I don't understand what the purpose of "Add WinService" is if it doesn't let you add ones that it has in it's list...
[06-Aug-2009 08:38:41] <twm1010> yeah, seems odd, but try this
[06-Aug-2009 08:38:47] <twm1010> go to services/winservice, find the one you want
[06-Aug-2009 08:38:54] <twm1010> click on it so it shows all instances
[06-Aug-2009 08:39:33] <twm1010> find the server you want, but click the name of the service under the "name" column
[06-Aug-2009 08:39:47] <TBKDan> Ahhhh
[06-Aug-2009 08:40:07] <TBKDan> That works for me
[06-Aug-2009 08:40:20] <twm1010> seems a little odd, maybe ask matt ray when he drops in, i've been meaning to figure that out myself
[06-Aug-2009 08:40:34] <TBKDan> As long as I have a way to add it to one specific device, I'm happy (even if it's somewhat of a roundabout way
)[06-Aug-2009 08:40:35] <TBKDan> Thanks!
[06-Aug-2009 08:41:00] <twm1010> dude, i feel like a doof
[06-Aug-2009 08:41:05] <twm1010> check this, go to that add winservice dialog
[06-Aug-2009 08:41:10] <twm1010> and double click in the entry box
[06-Aug-2009 08:41:27] <TBKDan> Well, that's what I've been doing.
[06-Aug-2009 08:41:39] <TBKDan> I select the service from that list, click ok, then it says that it can't find it
[06-Aug-2009 08:41:48] <twm1010> huh, weird
[06-Aug-2009 08:42:14] <TBKDan> My thoughts
[06-Aug-2009 08:42:19] <twm1010> mine does the same thing
[06-Aug-2009 08:42:22] <twm1010> if it makes you feel better
[06-Aug-2009 08:42:32] <twm1010> probably has to do with characters in the name of the service or something
[06-Aug-2009 08:42:39] <TBKDan> Spaces? lol
[06-Aug-2009 08:43:07] <twm1010> perhaps, i think the other method is easier anyway with a tabbed browser you could set 5 servers pretty quick
[06-Aug-2009 08:44:04] <TBKDan> But if I'm already on that device setting it up, it's kind of annoying.
[06-Aug-2009 08:44:19] <TBKDan> I should just be able to type in right there.
[06-Aug-2009 08:44:30] <TBKDan> And to add several services is annoying
[06-Aug-2009 08:48:15] <twm1010> indeed...
[06-Aug-2009 08:49:09] <twm1010> i think it does have to do with the name of the service
[06-Aug-2009 08:49:13] <twm1010> some work, some don't
[06-Aug-2009 08:52:11] <TBKDan> If the service name has spaces, it spazzes.
[06-Aug-2009 08:52:22] <TBKDan> The description doesn't matter, the service name does.
[06-Aug-2009 08:52:33] <TBKDan> I was able to add IISADMIN service.
[06-Aug-2009 08:53:05] <twm1010> I wonder can you enclose it in ''
[06-Aug-2009 08:54:14] <twm1010> I would wager there is a tracking bug/ticket for it already
[06-Aug-2009 08:54:42] <twm1010> I would rather see an entirely separate tab for services, that showed all services on the box
[06-Aug-2009 08:55:20] <twm1010> hah, there already is
[06-Aug-2009 08:55:26] <TBKDan> Details? lol
[06-Aug-2009 08:55:42] <twm1010> in the OS tab, under win services, there is a checkbox called "monitored"
[06-Aug-2009 08:55:45] <twm1010> its jsut a filter
[06-Aug-2009 08:55:49] <twm1010> uncheck it, and watch
[06-Aug-2009 08:55:57] <TBKDan> Ahhh
[06-Aug-2009 08:56:14] <TBKDan> Doesn't let you know if they're up or down though
[06-Aug-2009 08:56:24] <twm1010> right, because they're not monitored
[06-Aug-2009 08:56:25] <TBKDan> Though it does gives you startmode which is nice
[06-Aug-2009 08:56:31] <jb> morning..
[06-Aug-2009 08:56:36] <twm1010> once you monitor it, the checkbox should be checked
[06-Aug-2009 08:56:38] <TBKDan> Goooooooood morning
[06-Aug-2009 08:56:41] <TBKDan> yup
[06-Aug-2009 08:56:55] <TBKDan> So you can do it that way then
[06-Aug-2009 08:57:07] <TBKDan> Show all unmonitored and then click that service from there
[06-Aug-2009 08:57:26] <TBKDan> Not as convenient as the "Add Service" dialog, but still better than hunting it down for each service
[06-Aug-2009 08:57:26] <TBKDan>
[06-Aug-2009 08:57:38] <twm1010> right, hit show all as well, put checkboxes next to the ones you want to monitor hit the drop down and choose monitor enable
[06-Aug-2009 08:57:42] <twm1010> you could do all of them at once
[06-Aug-2009 08:58:22] <TBKDan> Ahhhhh there we go
[06-Aug-2009 08:58:26] <TBKDan> Ok, that makes me a happy camper
[06-Aug-2009 08:58:27] <TBKDan> Danke! lol
[06-Aug-2009 08:58:53] <twm1010> i wonder, if that works for osprocesses too
[06-Aug-2009 08:59:36] <twm1010> no, it does not, it needs a regex for that, but it does work for ports
[06-Aug-2009 09:03:29] <jb> what do you guys transform snmp_coldstart/alertsystemup to?
[06-Aug-2009 09:03:32] <jb> just curious
[06-Aug-2009 09:07:33] <roadies> what plugin is required for Firefox to allow the network map to work, aka "Select Device or Network" display
[06-Aug-2009 09:09:20] <twm1010> I haven't done anything with it yet.
[06-Aug-2009 09:09:33] <twm1010> I'd rather just move it to history
[06-Aug-2009 09:09:41] <twm1010> roadies: probably flash?
[06-Aug-2009 09:16:11] <roadies> twm1010: Thanks, got that on, I guess it doesnt work with Fedora running on a PPC *sign*
[06-Aug-2009 09:16:35] <TBKDan> roadies: It doesn't work for me on Windows either. Can't get it to display any nodes =/
[06-Aug-2009 09:17:35] <roadies> TBKDan: double *sign*
[06-Aug-2009 09:18:17] <roadies> TBKDan: Fedora on Lenovo laptops works great however :-D
[06-Aug-2009 09:19:09] <TBKDan> heh, it works fine for 192.168.0.0....
[06-Aug-2009 09:19:16] <TBKDan> Any other subnet, it doesn't.
[06-Aug-2009 09:21:22] <roadies> it only works to first set the device/net within the subnet view of zenoss, then U can browse by clicking on the device/nodes to get that device/node view
[06-Aug-2009 09:22:01] <TBKDan> Yeah, I've got it seeing a few more now, but nowhere near all the subnets that I have configured.
[06-Aug-2009 09:26:26] <roadies> I assume you slide the "number of hops" bar over... this is a limit I wish they would fix
[06-Aug-2009 09:29:08] <TBKDan> I have a device that has all these vlans connected to it, yet it only shows a few
[06-Aug-2009 09:41:10] <unkwon> hi
[06-Aug-2009 09:42:15] <unkwon> i need help
[06-Aug-2009 09:42:36] <jb> um.. you are going to need to be a bit more specific
[06-Aug-2009 09:43:32] <unkwon> i have installed zenoss server and i need monitoring BackupExec server with snmp trap send by the backup server
[06-Aug-2009 09:43:49] <unkwon> that's work perfeclty but i have problem with Jobsuccess
[06-Aug-2009 09:44:25] <unkwon> when the server zenoss receive the message and i have selected severity 0 (clear) the message as deleted from the history and status
[06-Aug-2009 09:44:25] <jb> whats the problem? is zenoss receiving the trap?
[06-Aug-2009 09:44:47] <unkwon> how i can leave the message with green color in the history
[06-Aug-2009 09:45:51] <unkwon> i have tested diffent solution with transformation but the message as deleted finaly
[06-Aug-2009 09:45:59] <jb> a clear will move to Event History..
[06-Aug-2009 09:47:43] <unkwon> if i leave the jobsuccess in /unknow class but i have the ZProperties in /unknown i clear the message stay in the history with green color background
[06-Aug-2009 09:48:00] <unkwon> it's not possible change just the background color
[06-Aug-2009 09:48:12] <unkwon> i need green color for this message
[06-Aug-2009 09:49:14] <jb> it will appear green in the event history.. you may be able to get the clear message not to move to history with a transform, but I'm not quite sure.
[06-Aug-2009 09:50:02] <unkwon> my server delete the message with clear severity if i assign in class
[06-Aug-2009 09:50:25] <unkwon> i have tested with 2.3 2.4 of zenoss and it's same
[06-Aug-2009 09:55:06] <otakup0pe> oh i forgot about this lil' party
[06-Aug-2009 09:59:39] <otakup0pe> so who likes performance metrics
[06-Aug-2009 09:59:56] <mrayzenoss> oh yeah!
[06-Aug-2009 10:00:07] <mrayzenoss> JP and I will be hosting the session today, he should be here in a second
[06-Aug-2009 10:00:34] <otakup0pe> the only question is what's the patch submission process
[06-Aug-2009 10:00:46] <otakup0pe> i meant the only question -i have- is
[06-Aug-2009 10:00:48] <mrayzenoss> other Zenoss devs may come and go, like cgibbons and npmccallum
[06-Aug-2009 10:02:11] <npmccallum> If I'm on IRC, I'm usually idling in here...
[06-Aug-2009 10:02:12] <mrayzenoss> patch submission is easy. Open a ticket, attach the patch. Fill out the contributor agreement and we patch it
[06-Aug-2009 10:02:22] <npmccallum> can't say I'm always around though
[06-Aug-2009 10:02:27] <explosiv0SX> I have a question: shall I just fire it off?
[06-Aug-2009 10:02:32] <jb> sure.
[06-Aug-2009 10:02:45] <otakup0pe> mrayzenoss: word i'll check out the contrib agreement
[06-Aug-2009 10:02:55] <npmccallum> explosiv0SX: go ahead
[06-Aug-2009 10:02:59] <mrayzenoss> otakup0pe: http://www.zenoss.com/forms/contribute
[06-Aug-2009 10:03:08] <explosiv0SX> What's are zenoss' plans re: rest/api or web services going forward?
[06-Aug-2009 10:03:24] <explosiv0SX> I see some xml methods in the source and I see methods that return JSON.
[06-Aug-2009 10:03:30] <explosiv0SX> which is the way forward?
[06-Aug-2009 10:03:34] <mrayzenoss> explosiv0SX: so far it's been kinda ad-hoc, we add things as needed
[06-Aug-2009 10:03:52] <mrayzenoss> we have talked about making it more rigorous and formal, which would be more useful for external users
[06-Aug-2009 10:04:17] <mrayzenoss> not SOAP formal, just standardizing it
[06-Aug-2009 10:04:24] <mrayzenoss> and being consistent
[06-Aug-2009 10:04:26] <explosiv0SX> right -- SOAP is a little heavy-handed.
[06-Aug-2009 10:04:35] <otakup0pe> consistency++
[06-Aug-2009 10:04:38] <explosiv0SX> REST -> XML or JSON is more than fine.
[06-Aug-2009 10:05:07] <explosiv0SX> so no plans to deprecate any of the JSON or XML methods?
[06-Aug-2009 10:05:20] <mrayzenoss> no, everybody uses them
[06-Aug-2009 10:05:33] <mrayzenoss> well, I don't know if every single one is in use
[06-Aug-2009 10:05:39] <explosiv0SX> great.
[06-Aug-2009 10:05:40] <mrayzenoss> but I haven't heard anything about deprecating
[06-Aug-2009 10:05:44] <npmccallum> explosiv0SX: if we do deprecate them, they are marked as such before there removal...
[06-Aug-2009 10:05:56] <otakup0pe> yeah that makes more sense than just leaving them in for ever.
[06-Aug-2009 10:06:03] <mrayzenoss> so you'll have at least a release to get ready
[06-Aug-2009 10:06:07] <explosiv0SX> great.
[06-Aug-2009 10:06:09] <npmccallum> explosiv0SX: but I don't know that we've ever actually removed a method
[06-Aug-2009 10:06:10] <explosiv0SX> thank you.
[06-Aug-2009 10:06:17] <explosiv0SX> doesn't appear so ;-)
[06-Aug-2009 10:06:37] <npmccallum> explosiv0SX: we're a bit like GNOME in that regard
[06-Aug-2009 10:06:43] <explosiv0SX> right on.
[06-Aug-2009 10:06:56] <mrayzenoss> we are planning on deprecating RHEL 4 with the next release, and Zip-style ZenPacks
[06-Aug-2009 10:07:05] <mrayzenoss> not web services, but deprecations
[06-Aug-2009 10:07:57] <npmccallum> and I assume when that happens some code will probably die off at some point, but most of it isn't really external stuff
[06-Aug-2009 10:08:22] <npmccallum> the most important APIs that people care about are usually device and event management
[06-Aug-2009 10:08:46] <otakup0pe> oh here's a question. was there already rest methods for total memory/swap and i just couldn't find them ?
[06-Aug-2009 10:09:26] <npmccallum> otakup0pe: I'm not sure what you mean... are you trying to get the current memory usage on a device?
[06-Aug-2009 10:09:38] <otakup0pe> no the total memory/swap installed on a device
[06-Aug-2009 10:09:59] <otakup0pe> i couldn't find rest methods so i ended up adding them myself heh
[06-Aug-2009 10:10:27] <otakup0pe> i.e. totalSwapString from operatingsystem.py (iirc)
[06-Aug-2009 10:11:53] <explosiv0SX> otakup0pe: you can get it like: find('myserver').os.exportXml(sys.stdout)
[06-Aug-2009 10:12:09] <explosiv0SX> but not sure if that helps because of that pesky output stream parameter.
[06-Aug-2009 10:13:00] <explosiv0SX> btw -- that was written for zendmd.
[06-Aug-2009 10:13:20] explosiv0SX is now known as rak
[06-Aug-2009 10:13:45] <ckrough> What is the recommended datasource type for ethernet traffic? COUNTER or DERIVE? I'm trying to avoid spikes when interface counters roll over unexpectedly.
[06-Aug-2009 10:15:22] <otakup0pe> rak: that answer didn't make sense ? i meant a rest method.
[06-Aug-2009 10:15:38] <otakup0pe> let me just make a patch and it will make sense :3
[06-Aug-2009 10:17:46] <rak> well, this leads to another question:
[06-Aug-2009 10:18:03] <rak> I'd love to get device info with: http://zenoss.corp.goldlasso.com/zport/dmd/Devices/Server/Linux/devices/myserver/exportXml
[06-Aug-2009 10:18:10] <rak> the method is accessed.
[06-Aug-2009 10:18:29] <rak> but an error is thrown saying it's missing a param.
[06-Aug-2009 10:18:36] <rak> This parameter is the output stream.
[06-Aug-2009 10:18:58] <rak> is there anything I can pass to this method so say -- use the current response stream?
[06-Aug-2009 10:19:22] <otakup0pe> http://pastebin.ca/1520574 was all i added btw
[06-Aug-2009 10:20:40] <otakup0pe> hmm exportXml throws not found on some devices and throws an error on other devices for me. but i'm at work where i only have access to the (outdated) vmware appliance
[06-Aug-2009 10:20:53] <rak> cool. I probably could do the same to expose the exportXml function..
[06-Aug-2009 10:21:06] <otakup0pe> yeah the trick is the freakin @rtype tag
[06-Aug-2009 10:21:20] <otakup0pe> that seems to let zope know about the method
[06-Aug-2009 10:22:43] <jplouis> ckrough: I believe zenoss uses DERIVE with a RRD Min of 0 for interfaces
[06-Aug-2009 10:23:32] <ckrough> jplouis: actually, reading up on it just now, the recommended method is COUNTER with a max.
[06-Aug-2009 10:24:02] <ckrough> jplouis: device with a min will get a log of unknowns during normal counter rollovers apparently
[06-Aug-2009 10:24:26] <ckrough> jplouis: derive I mean, not device
[06-Aug-2009 10:27:19] <jplouis> from the RRD docs "so for high bandwidth interfaces and a 32bit counter, DERIVE with min=0 is probably preferable"
[06-Aug-2009 10:31:10] <daMaestro> does anyone know how to create a report that takes interface stats from specific devices and give a total number of bits transferred in a given time period
[06-Aug-2009 10:31:24] <daMaestro> so, it will actually total the rrd data
[06-Aug-2009 10:33:02] <npmccallum> otakup0pe, rak: there is a better way to get things like total memory
[06-Aug-2009 10:33:10] <otakup0pe> npmccallum: hooray !
[06-Aug-2009 10:33:28] <npmccallum> otakup0pe, rak: http://<hostname>:8080/zport/dmd/Devices/Server/Linux/devices/<device>/hw/getProperty?id=totalMemory
[06-Aug-2009 10:33:35] <otakup0pe> nice
[06-Aug-2009 10:33:41] <otakup0pe> getProperty is that a zope thing ?
[06-Aug-2009 10:33:42] <npmccallum> you can do that with any property
[06-Aug-2009 10:33:56] <npmccallum> I don't recall if it is a zope thing or not
[06-Aug-2009 10:34:16] <rak> ok, that'll do nicely.
[06-Aug-2009 10:34:22] <npmccallum> jplouis: do you know if getProperty comes from zope? or is it a zenoss thing?
[06-Aug-2009 10:34:37] <otakup0pe> looks like a zope-cmf thing ?
[06-Aug-2009 10:34:48] <otakup0pe> word thanks man if we are ever at the same conference i'll buy you a beer
[06-Aug-2009 10:34:54] <jplouis> I believe from Zope
[06-Aug-2009 10:35:08] <npmccallum> otakup0pe: let me know what conference to show up to and I'll take you up on it
[06-Aug-2009 10:37:26] <ckrough> jplouis: Im running 64bit counters, take a look at the line in the doc right after the one you pasted, do you take that to mean 64bit with DERIVE and a max, or 64bit with COUNTER and a max
[06-Aug-2009 10:37:57] <npmccallum> otakup0pe: I'll be at OLF
[06-Aug-2009 10:39:00] <jplouis> I read it as COUNTER with a max
[06-Aug-2009 10:39:15] <mrayzenoss> otakup0pe: me too
[06-Aug-2009 10:39:32] <jb> where is OLF?
[06-Aug-2009 10:39:55] <jb> orlando?
[06-Aug-2009 10:40:06] <mrayzenoss> Ohio
[06-Aug-2009 10:40:09] <jb> ah
[06-Aug-2009 10:40:11] <mrayzenoss> Ohio Linux Fest
[06-Aug-2009 10:40:13] <jb> you guys coming to chi anytime soon?
[06-Aug-2009 10:40:18] <mrayzenoss> actually yes
[06-Aug-2009 10:40:26] <mrayzenoss> our Product Manager is headed up there
[06-Aug-2009 10:40:37] <mrayzenoss> not sure who else is going, he's talking to some customers
[06-Aug-2009 10:40:46] <mrayzenoss> jb: are you on the list?
[06-Aug-2009 10:41:08] <jb> hm, I haven't heard anything about it..
[06-Aug-2009 10:41:15] <jb> i went to a campit a few months ago
[06-Aug-2009 10:41:38] <jb> i'd be interested though..
[06-Aug-2009 10:42:09] <otakup0pe> eh i'm in montreal and in the middle of a release cycle so i won't be going to con's any time soon
[06-Aug-2009 10:42:16] <otakup0pe> heh that should have been a sad emoticon
[06-Aug-2009 10:42:26] <mrayzenoss> jb: I'll hook you up, he was looking for customers and core users in Chicago
[06-Aug-2009 10:42:33] <otakup0pe> btw i should have some video of my third gen sl/zenoss mashups once i figure out fraps
[06-Aug-2009 10:42:40] <otakup0pe> and remember how2machinima
[06-Aug-2009 10:42:43] <mrayzenoss> otakup0pe: awesome
[06-Aug-2009 10:42:46] <jb> mrayzenoss: yeah, pass him my contact info if you don't mind..
[06-Aug-2009 10:42:49] <jb> is it brandon?
[06-Aug-2009 10:43:28] <mrayzenoss> yeah
[06-Aug-2009 10:43:44] <jb> argh, i think i actually accidentally deleted an email about vmware from him yesterday
[06-Aug-2009 10:44:01] <mrayzenoss> I'll ask him to resend that too
[06-Aug-2009 10:44:18] <jb> thanks
[06-Aug-2009 10:46:47] <raddy> Hello Everybody
[06-Aug-2009 10:47:15] <raddy> Can anybody suggest a good regexp for monitoring postgres process?
[06-Aug-2009 10:48:10] <raddy> When i just use postgres, it shows lot of processes.
[06-Aug-2009 10:49:21] <npmccallum> isn't postmaster the name of the master process? (I'm just guessing here, its been a while since I used postgres...)
[06-Aug-2009 10:49:45] <npmccallum> woah, I was right: http://www.postgresql.org/docs/8.1/static/app-postmaster.html
[06-Aug-2009 10:53:52] <raddy> Monitoring master process was useless
[06-Aug-2009 10:54:08] <raddy> As the memory usage n'all very 2-5 MB
[06-Aug-2009 10:54:25] <raddy> Monitoring postgres process is important,
[06-Aug-2009 10:54:29] <ckrough> what are you trying to do?
[06-Aug-2009 10:57:52] <jb> however, if postmaster dies, you are SOL
[06-Aug-2009 10:59:47] <raddy> ckrough : I want to monitor postgres memory usage.
[06-Aug-2009 11:00:07] <ckrough> so your not looking for availability, your looking for usage. gotcha
[06-Aug-2009 11:04:27] <mrayzenoss> Just a heads up, the ZenPack Contest only has 9 days left
[06-Aug-2009 11:04:42] <mrayzenoss> so if you were planning on entering, please send it them in soon
[06-Aug-2009 11:04:56] <chudler_> thanks, I have one or two to put in at the last minute.
[06-Aug-2009 11:05:27] chudler_ is now known as chudler
[06-Aug-2009 11:12:11] <raddy> ckrough : I can monitor availability seperately
[06-Aug-2009 11:12:26] <jb> mrayzenoss: is there any interest in zenpacks that use zenwinperf?
[06-Aug-2009 11:12:40] <raddy> ckrough : I want to monitor performance metrics for postgres process.
[06-Aug-2009 11:13:02] <mrayzenoss> jb: not sure what to do with Enterprise-dependendent zenpacks
[06-Aug-2009 11:13:09] <jb> yeah..
[06-Aug-2009 11:19:20] <otakup0pe> here's another questin actually. what specifically does it mean when getRRDValue returns 204. that the last snmp query failed ?
[06-Aug-2009 11:20:03] <otakup0pe> oh wait it does that for 0 responses nm >_<
[06-Aug-2009 11:25:07] <mrayzenoss> thanks to everyone who stopped by, I'll have a transcript up later on today
[06-Aug-2009 11:25:23] <mrayzenoss> I'll still be around, but the devs are off the hook
[06-Aug-2009 11:31:37] <Matir> I'm having a weird issue with my Zenoss dashboard... all of the groups added show an event status of "Green", even though some devices in those groups have events that are neither clear nor acknowledged
[06-Aug-2009 11:36:27] <rhett> do all the device shave teh same production state
[06-Aug-2009 11:38:09] <Matir> no, not all devices do
[06-Aug-2009 11:38:16] <Matir> does it only show events if they are production?
[06-Aug-2009 11:41:21] <mrayzenoss> Matir: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch08s02.html has which production states are on the dashboard
[06-Aug-2009 11:41:37] <Matir> ok, thanks
[06-Aug-2009 11:41:42] <Matir> I'll see if that addresses it
[06-Aug-2009 11:43:22] <Matir> ok, I get it... Production States are more zenoss-centric and don't necessarily align with the production states we use internally
[06-Aug-2009 11:43:27] <Matir> thanks!
[06-Aug-2009 12:09:23] <rhett> i just noticed i typed out my last line terribly
[06-Aug-2009 12:25:10] <Diddi> rhett: quite funny, though (:
[06-Aug-2009 13:41:46] <daMaestro> what is the recommended upgrade path from a 2.2.4 instance?
[06-Aug-2009 13:41:56] <daMaestro> is it to upgrade to 2.3 then 2.4?
[06-Aug-2009 13:42:16] <daMaestro> because i just tried the 2.2.4 -> latest.. and that fails
[06-Aug-2009 13:43:06] <magnachef> I think so...there are docs somewhere on the zenoss website
[06-Aug-2009 13:47:18] <chudler> anyone heard of using Zenoss with either modbus or bacnet?
[06-Aug-2009 13:51:19] <chudler> I've got PHB's asking about it. Here comes a time sink
[06-Aug-2009 13:55:11] <rhett> haha
[06-Aug-2009 13:56:55] <mrayzenoss> daMaestro: http://www.zenoss.com/community/docs/installation-guide/2.4.2/ch08s01.html#d4e869
[06-Aug-2009 15:31:05] edwardam is now known as edwardam_Fooz
[06-Aug-2009 15:31:30] edwardam_Fooz is now known as edwardamFoodz
[06-Aug-2009 15:39:49] daMaestro|isBack is now known as daMaestro
[06-Aug-2009 16:01:21] edwardamFoodz is now known as edwardam
[06-Aug-2009 16:04:43] edwardam is now known as edwardam_FoodRea
[06-Aug-2009 17:06:49] edwardam_FoodRea is now known as edwardam
[06-Aug-2009 18:22:40] daMaestro|isBack is now known as daMaestro
[06-Aug-2009 18:22:46] <daMaestro> mrayzenoss, ping
[06-Aug-2009 18:22:55] <daMaestro> mrayzenoss, i see you were the first to comment on my bug: http://dev.zenoss.com/trac/ticket/3740
[06-Aug-2009 18:22:58] <mrayzenoss> pong
[06-Aug-2009 18:23:12] <daMaestro> mrayzenoss, it got closed.. even though (so far i'm upgraded to 2.3.3) it's not fixed
[06-Aug-2009 18:23:33] <daMaestro> it might be in 2.4 ... just about to make that move, but the point of the bug was not addressed in the bug that was marked as related
[06-Aug-2009 18:23:52] <mrayzenoss> hmmm... chet verified it, but build 309 may have been 2.4
[06-Aug-2009 18:23:59] <daMaestro> mrayzenoss, i'll include an updated patch soon, but for example.. lighttpd requires you to send \r\n not just \n
[06-Aug-2009 18:24:18] <daMaestro> k, i'll wait to get up to date on the latest release.. just wanted to ping you 11 months later with a reply
[06-Aug-2009 18:24:19] <daMaestro> lol
[06-Aug-2009 18:24:46] <mrayzenoss> try pulling down a 2.4.2 VM and see if it has it
[06-Aug-2009 18:24:57] <mrayzenoss> that way you don't have to install anything
[06-Aug-2009 18:25:06] <daMaestro> i'm literally doing a backup of the 2.3.3 and will be going to 2.4 shortly
[06-Aug-2009 18:25:15] <mrayzenoss> cool
[06-Aug-2009 18:25:29] <daMaestro> finding and merging in all my local changes has been.... fun.
[06-Aug-2009 18:26:09] <mrayzenoss> yeah, whenever professional services makes a change for a customer, it gets pushed into the next maintenance release to avoid that problem
[06-Aug-2009 18:36:16] <sacpinball> mrayzenoss i just updated zenoss in my development system to 2.4.2
[06-Aug-2009 18:36:30] <sacpinball> and i am seeing a weird problem you might know the answer to
[06-Aug-2009 18:36:47] <sacpinball> a solaris 8 server being monitored via netsnmp
[06-Aug-2009 18:36:58] <sacpinball> i add to monitor the sshd daemon
[06-Aug-2009 18:37:02] <sacpinball> in the os process field
[06-Aug-2009 18:37:08] <mrayzenoss> sacpinball: I can't say I've heard good things about SNMP on Solaris
[06-Aug-2009 18:37:30] <sacpinball> a few hours later, zenoss removes that from being monitored
[06-Aug-2009 18:37:32] <sacpinball> any ideas
[06-Aug-2009 18:37:52] <sacpinball> removing object sshd from rel processes on device os
[06-Aug-2009 18:37:54] <mrayzenoss> take a look at Jane Curry's process monitoring paper: http://www.skills-1st.co.uk/papers/jcurry.html It's great
[06-Aug-2009 18:38:06] <mrayzenoss> I was just headed home
[06-Aug-2009 18:38:19] <sacpinball> no problem
[06-Aug-2009 18:38:20] <sacpinball> thanks
[06-Aug-2009 18:38:25] <sacpinball> i will look at that paper
[06-Aug-2009 18:38:28] <mrayzenoss> later
[06-Aug-2009 18:38:28] <locohost> doesnt solaris have its own non-net-snmp snmp daemon
[06-Aug-2009 18:38:40] <sacpinball> it sure does, but its horrible
[06-Aug-2009 18:39:08] <sacpinball> i looked around and found that a lot of people are using netsnmp to monitor their solaris hosts
[06-Aug-2009 18:43:27] <locohost> yeah, i guess thats pre-solaris 9
[06-Aug-2009 18:43:47] <sacpinball> yeah; i only have 6 of these solaris 8 dogs left
[06-Aug-2009 18:43:56] <sacpinball> so i might just go with ssh monitoring for them
[06-Aug-2009 18:44:13] <sacpinball> seems like net-snmp works, but i have seen two wierd things already
[06-Aug-2009 18:46:05] <locohost> sounds exciting
[06-Aug-2009 18:46:16] <sacpinball> lol
[06-Aug-2009 18:46:16] <sacpinball> yep
[06-Aug-2009 18:46:49] <locohost> you think that is bad, once i made a cisco catalyst 6509E reboot with an snmpwalk, cisco said it was an unpublished bug and suggested a firmware upgrade
[06-Aug-2009 18:47:03] <locohost> i mean, ios upgrade
[06-Aug-2009 18:47:04] <sacpinball> yikes
[06-Aug-2009 18:49:11] <daMaestro> @mray yeah... you guys implemented it not how it should be implemented
[06-Aug-2009 18:49:21] <daMaestro> @mray we need more then just \n
[06-Aug-2009 18:57:27] <moqq> hey, how can i adjust the parameters for alerts? e.g. i have a linux system that's getting memory swap alerts, but for that specific system i don't really care. how can i adjust the threshold or severity of that event?
[06-Aug-2009 19:48:47] bzed_ is now known as bzed
[07-Aug-2009 00:00:46] [disconnected at Fri Aug 7 00:00:46 2009]
[07-Aug-2009 00:00:46] [connected at Fri Aug 7 00:00:46 2009]
[07-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[07-Aug-2009 01:12:32] <KeithBilly> Hey, i'm trying to install 2.4.2 over 2.3 and it keeps flaking on the rrddeps build of glib, looking for iconv on freebsd 6.3
[07-Aug-2009 01:13:15] <KeithBilly> i'm following the posts by http://www.caseyhillman.com/?p=88 to deal with this but can't seem to get glib to see iconv or get iconv to build before glib
[07-Aug-2009 01:13:23] <KeithBilly> can anyone help me fix this problem
[07-Aug-2009 01:16:26] edwardam is now known as edwardamZzZz
[07-Aug-2009 02:45:39] <Dieterbe> is it just me or can't you do 'limit <int>' in WQL ?
[07-Aug-2009 08:52:59] <klinstifen> morning all
[07-Aug-2009 08:53:11] <klinstifen> so my zenoss install has grown to big for me to do a backup to my local disk
[07-Aug-2009 08:53:46] <klinstifen> i want to use an nfs share as the temp-dir for zenbackup
[07-Aug-2009 08:53:54] <klinstifen> has anyone done this successfully?
[07-Aug-2009 08:54:05] <klinstifen> i am getting permission errors
[07-Aug-2009 08:54:19] <klinstifen> no matter how i mount the share (mount, smbmount, sudo mount, etc
[07-Aug-2009 08:56:39] <npmccallum> klinstifen: I think this is the wrong channel to ask your question (since it has nothing to do with Zenoss)
[07-Aug-2009 08:59:54] <LW> Question, I am trying to add an ip service to be monitored, which happens to be in the 18000 range. It was never autodetected when I modeled, so I manually added it manually, yet it still does not find it. I noticed that it was bound via a ipv6 port, and most other apps were binding with ipv4 that were showing up, so I added the ipv6 snmp map, and it found ssh on the box, but not the other service, the other thing I noticed that it's b
[07-Aug-2009 10:02:19] <djjay> hi. I've seen some references to zenoss honoring ./.ssh/config but so far it doesn't seem to be true. anyone know?
[07-Aug-2009 10:37:44] edwardam1 is now known as edwardam
[07-Aug-2009 10:38:05] edwardam is now known as edwardamExer
[07-Aug-2009 10:44:57] <pat2man> Quick question, I want to run the nagios check_ping command against a specific IP address for one of our routers. Can I apply an interface template that will run check_ping? The interface template section of the documentation is rather sparse.
[07-Aug-2009 10:48:08] <mrayzenoss> pat2man: if you create a template that utilizes the nagios command, you can bind it to whichever device classes you need
[07-Aug-2009 10:48:19] <mrayzenoss> I've got it attached on /Devices
[07-Aug-2009 10:48:34] <pat2man> right i can bind it to devices but I want to bind it to a specific interface
[07-Aug-2009 10:48:46] <mrayzenoss> hmmm...
[07-Aug-2009 10:49:00] <pat2man> or can I specify a specific interface in the command section of the template?
[07-Aug-2009 10:49:16] <mrayzenoss> I see what you're asking, I'm looking
[07-Aug-2009 10:53:22] <mrayzenoss> pat2man: the interface that will be tested is the one from the zenoss server to that machine
[07-Aug-2009 10:53:42] <mrayzenoss> is the machine reachable from multiple interfaces from the zenoss server?
[07-Aug-2009 10:53:53] <pat2man> yes
[07-Aug-2009 10:53:56] <mrayzenoss> ugh
[07-Aug-2009 10:55:10] <pat2man> it seems like I should be able to apply a template to the specific interface that I want to monitor
[07-Aug-2009 10:55:26] <pat2man> and just have performance metrics for that interface instead of for the host
[07-Aug-2009 10:56:28] <mrayzenoss> yeah, there's nothing in the man page for check_ping that makes me think it would take any sort of guidance
[07-Aug-2009 10:56:48] <pat2man> well check_ping just runs against an IP
[07-Aug-2009 10:56:49] <mrayzenoss> I mean really, we just call the shell and execute the script, passing the IP address
[07-Aug-2009 10:57:00] <pat2man> in my case I run this: check_ping -H ${dev/id} -w100,80% -c500,100% -p1 | sed 's/, RTA = / RTA=/;s/ loss = / |loss=/'
[07-Aug-2009 10:57:08] <mrayzenoss> ahh, sorry, brainfart
[07-Aug-2009 10:57:20] <pat2man> but ${dev/id} is always the primary IP of the device
[07-Aug-2009 10:57:24] <mrayzenoss> change -H ${dev/id} to -H ${here/manageIp}
[07-Aug-2009 10:58:02] <pat2man> would that be for a device template or an interface template?
[07-Aug-2009 10:58:05] <mrayzenoss> hmm... I wonder if there's a ${here/interfaces[0] or something
[07-Aug-2009 10:58:11] <mrayzenoss> device template
[07-Aug-2009 10:58:34] <pat2man> well I mean I could always create a local copy of my template for each device and manually put in the IP too
[07-Aug-2009 10:58:46] <pat2man> but I would rather have an interface template that I can just bind to any interfaces I need it on
[07-Aug-2009 11:00:38] <pat2man> if its not obvious I can just submit a new feature request
[07-Aug-2009 11:01:49] <mrayzenoss> I'm just not sure where it would get bound
[07-Aug-2009 11:03:12] <venturaville> jb: I still cannot find performance data for the SVCs in CIM.....
[07-Aug-2009 11:03:16] <pat2man> well the admin guide has this to say:
[07-Aug-2009 11:03:17] <pat2man> 2.1.3. Interface Templates
[07-Aug-2009 11:03:17] <pat2man> Interface templates are applied to network interfaces by using a special type of binding. Instead of using the name of the underlying class, Zenoss looks for a template with the same name as the interface type. You can find this type in the details information for any network interface (from the OS tab of its containing device).
[07-Aug-2009 11:03:17] <pat2man> If Zenoss cannot locate a template that matches the interface type, then it uses the ethernetCsmacd template.
[07-Aug-2009 11:05:32] <mrayzenoss> ahh, so if you were to name the check_ping template "eth1" and bind it to the interfaces, it would work on just that interface
[07-Aug-2009 11:05:48] <mrayzenoss> at least, that's what it sounds like to me, I haven't messed with that
[07-Aug-2009 11:07:12] edwardamExer is now known as edwardam
[07-Aug-2009 11:08:18] <pat2man> mrayzenoss: well I think they are talking about the interface type not the name
[07-Aug-2009 11:08:22] <pat2man> type is ethernetCsmacd_64
[07-Aug-2009 11:08:29] <pat2man> name is ethernet0/0
[07-Aug-2009 11:09:21] <mrayzenoss> ahh, correct
[07-Aug-2009 11:09:38] <pat2man> well I will just make a local copy for now and manually enter the IP
[07-Aug-2009 11:09:44] <pat2man> and put in a feature request
[07-Aug-2009 11:10:20] <mrayzenoss> yeah, that would probably be useful
[07-Aug-2009 11:37:08] <jb> venturaville:
[07-Aug-2009 11:54:42] <daMaestro> anyone know what needs to be done to have zencommand timeouts send clears if on the next poll the command correctly executes?
[07-Aug-2009 11:54:45] <daMaestro> it used to do that with 2.2.x
[07-Aug-2009 11:56:23] <daMaestro> basically, we have having zencommand events being raised, but they never clear without manual intervention
[07-Aug-2009 11:56:34] <daMaestro> (now running 2.4.2)
[07-Aug-2009 12:19:19] <mrayzenoss> daMaestro: http://dev.zenoss.org/trac/ticket/5359
[07-Aug-2009 12:19:41] <mrayzenoss> awaiting verification for the 2.4.3 release
[07-Aug-2009 12:19:46] <mrayzenoss> but there's a patch
[07-Aug-2009 12:20:09] <mrayzenoss> hmm... this one too: http://dev.zenoss.org/trac/ticket/5100
[07-Aug-2009 12:21:41] <mrayzenoss> here's the tickets currently slated for 2.4.3: http://dev.zenoss.org/trac/report/6
[07-Aug-2009 12:26:59] <daMaestro> mrayzenoss, cool; i updated my bug with a new patch for 2.4.2
[07-Aug-2009 12:28:17] <daMaestro> http://dev.zenoss.com/trac/ticket/3740
[07-Aug-2009 12:29:01] <daMaestro> mrayzenoss, that patch makes zenoss act like most of the hardware based load balancers/healthchecks i've worked with
[07-Aug-2009 12:29:43] <daMaestro> i think most people would expect that behavior, but would recommend a release note point if it's implemented that if you need to send literal \n, etc it would need to be escaped
[07-Aug-2009 12:32:29] <daMaestro> actually ... on a brief ipython session it doesn't look like a simple \\n prints \n so yeah.
[07-Aug-2009 12:32:45] <daMaestro> i'd still recommend it, we use it and it allows you to do advanced testing of tcp sessions ;-)
[07-Aug-2009 12:33:02] <daMaestro> mrayzenoss, thanks for the link to the patches, applying now
[07-Aug-2009 12:33:31] <mrayzenoss> yeah, 2.4.3 is a few weeks away
[07-Aug-2009 12:33:48] <mrayzenoss> should be last 2.4.x release before King Crab
[07-Aug-2009 12:35:48] <daMaestro> odd... that changeset doesn't apply against 2.4.2 cleanly
[07-Aug-2009 12:36:14] <mrayzenoss> perhaps other patches are interfering?
[07-Aug-2009 12:36:18] <mrayzenoss> or missing
[07-Aug-2009 12:36:23] <daMaestro> http://dev.zenoss.org/trac/changeset/14424#file0
[07-Aug-2009 12:36:36] <daMaestro> that doesn't apply to 2.4.2 cleanly
[07-Aug-2009 12:37:29] <mrayzenoss> there are quite a few patches to zencommand.py
[07-Aug-2009 12:46:47] <daMaestro> http://fpaste.org/ZtKX/ is what i'll give a shot
[07-Aug-2009 13:02:47] <folke> I'm trying to get an snmp perf template working.. But fails miserably on the first counter...
[07-Aug-2009 13:03:17] <folke> Everything looks good in the datasource. I'm getting the right values.
[07-Aug-2009 13:03:39] <folke> And a good ol snmpget gets the right stuff SNMPv2-SMI::enterprises.3711.7.1.1.1.1.2.0 = INTEGER: 358
[07-Aug-2009 13:03:52] <folke> But nuthin shows up in the graphs..
[07-Aug-2009 13:04:43] <daMaestro> you have to add a data point
[07-Aug-2009 13:05:09] <daMaestro> so for custom OIDs, you have to setup a datasource for each OID and then assign a data point to each OID
[07-Aug-2009 13:05:21] <daMaestro> at least that is how it was in 2.2, not sure about 2.4... have not looked yet
[07-Aug-2009 13:06:10] <folke> daMaestro, I am so lost. I can't find that anywhare in 2.4.2
[07-Aug-2009 13:06:13] <folke>
[07-Aug-2009 13:07:15] <folke> In the datasource I can only add an datapoint alias..
[07-Aug-2009 13:08:40] <folke> I had it working in our 2.1.x enviroment.. But no cigar now...
[07-Aug-2009 13:11:07] <folke> I think that they are called Graph Points nowdays.. But still nuthin.
[07-Aug-2009 13:13:49] <folke> when I do a zenperfsnmp run -v 10 -d jadda.jadda.net there do not look like it asks my perftemplate..
[07-Aug-2009 13:13:54] <daMaestro> ahhh
[07-Aug-2009 13:14:05] <daMaestro> no, it's still data source
[07-Aug-2009 13:14:13] <daMaestro> but you specify the name and the oid in the same object now
[07-Aug-2009 13:14:18] <daMaestro> rather then multiple representations
[07-Aug-2009 13:15:11] <folke> daMaestro, Humm, I'll take a look at it..
[07-Aug-2009 13:16:59] <folke> I have the name and the oid, there.. Enabled and everything. And when I test it against a device it returns the right value..
[07-Aug-2009 13:18:32] <daMaestro> k, so add those data points to a graph
[07-Aug-2009 13:18:38] <daMaestro> and then bind the template to a device
[07-Aug-2009 13:19:43] <folke> Aye, done. THe graph shows up with the right legend. But it gets no data.
[07-Aug-2009 13:20:27] <folke> Looks like it auint grabbing it when I run zenperfsmp.
[07-Aug-2009 13:20:49] <mrayzenoss> folke: that'll take a few minutes to get some data going
[07-Aug-2009 13:21:29] <folke> I started it a few hours ago at work.. And now when I got home I checked it out..
[07-Aug-2009 13:21:55] <folke> I have just missed something trivial.. Just need to find it
[07-Aug-2009 13:22:24] <folke> It's my first shot at 2.4.x
[07-Aug-2009 13:26:47] <folke> here is the output from a zenperfsnmp run: http://pastebin.com/d228fe20d
[07-Aug-2009 13:27:20] <kisielk> I'm really starting to hate zenprocess
[07-Aug-2009 13:28:27] <folke> I think that INFO:zen.zenperfsnmp:Sent 0 OID requests
[07-Aug-2009 13:28:27] <folke> INFO:zen.zenperfsnmp:Queried 0 devices is not optimal if you want to hace the graph filled
[07-Aug-2009 13:45:15] <folke> Humm, starting to think that I have broken sumfink in my lab vmware guest enviroment.. The snmp is really working flaky.. Evern with the ordinary linux snmp counters...
[07-Aug-2009 13:45:45] <mrayzenoss> folke: I'm testing adding my own snmp datasource to see if I can figure what's blocking you
[07-Aug-2009 13:47:50] <folke> mrayzenoss, Perhaps it's only pebkac this time
I will take the vmware guest back a snap and see if I can get stability in the linux prefs..[07-Aug-2009 13:51:02] <mrayzenoss> well, I'm trying to add a simple dhcp client count and I'm getting a blank graph. I figure once I get that working I'll throw up a quick howto
[07-Aug-2009 13:57:19] <folke> That sounds great. I will continue over here in sweden for a few hours more
[07-Aug-2009 14:40:12] <mrayzenoss> took awhile, but my perf graphs just started showing up
[07-Aug-2009 14:40:50] <mrayzenoss> I had to add a .0 to the end of the OID I was collecting (as noted in the Admin Guide)
[07-Aug-2009 14:47:59] <folke> mrayzenoss, when I add .0 the test in the Data Source tab fails. But I try to run it like that a few minutes.
[07-Aug-2009 14:48:18] <mrayzenoss> yeah, I'm testing different variations of the template now
[07-Aug-2009 14:48:22] <mrayzenoss> to see what works
[07-Aug-2009 14:50:30] <folke> Just to test I made a local copy of Device snmp template and inserted my oids in it.. To see what happens with it.
[07-Aug-2009 14:52:07] <mrayzenoss> it does appear that once you add a custom template to a device, it seems to not bind any other changes made to the device class templates
[07-Aug-2009 14:52:43] <mrayzenoss> so far, only the templates with a .0 added seem to be working
[07-Aug-2009 14:56:40] <folke> The local device modded template is starting to view graphs now.
[07-Aug-2009 14:57:34] <folke> But my "home made" one with the added zero is empty..
[07-Aug-2009 14:58:25] <folke> I think that I already have the end value in thati oid, and get the right intger from it.
[07-Aug-2009 15:00:49] <folke> It just seems that it wont "parse" the template even if looks like it is binded in the template tab..
[07-Aug-2009 15:26:48] edwardam is now known as edwardamAFK
[07-Aug-2009 15:58:45] edwardamAFK is now known as edwardam
[07-Aug-2009 15:59:31] <mrayzenoss> I'll be out the next week on vacation, if anyone's looking for me
[07-Aug-2009 16:12:39] <folke> Now I'll getting somewhere.. Both graphs getting data now.
[07-Aug-2009 16:17:28] <folke> After I deleted the rrd file and restarted zenoss things starting to light up..
[07-Aug-2009 17:37:36] <mrayzenoss> New WBEM-based ZenPacks
[07-Aug-2009 17:38:04] <mrayzenoss> WBEM DataSource http://www.zenoss.com/community/projects/zenpacks/wbemdatasource
[07-Aug-2009 17:38:17] <mrayzenoss> WBEM LinuxMonitor http://www.zenoss.com/community/projects/zenpacks/wbem_linux
[07-Aug-2009 17:43:29] <mrayzenoss> Last Week for the Summer of ZenPacks Contest: http://tinyurl.com/optb2d
[07-Aug-2009 17:46:11] <mrayzenoss> later all!
[07-Aug-2009 18:17:10] daMaestro|isBack is now known as daMaestro
[07-Aug-2009 19:44:17] <malibu> Hi there, I just installed zenoss for the first time from the CentOS 5 stack installer.. But the admin:zenoss ID isn't working
[07-Aug-2009 19:44:29] <malibu> Anyone know what the default ID/pass might be?
[07-Aug-2009 23:34:44] [disconnected at Fri Aug 7 23:34:44 2009]
[08-Aug-2009 00:00:46] [connected at Sat Aug 8 00:00:46 2009]
[08-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[08-Aug-2009 06:24:05] <raddy> Hello Everybody
[08-Aug-2009 06:24:12] <raddy> Anybod out there????
[08-Aug-2009 06:28:20] <raddy> Anybody there???
[08-Aug-2009 08:59:14] <straterra> Indeed
[08-Aug-2009 08:59:22] <straterra> Patience..
[08-Aug-2009 09:37:24] <brice> Did zenoss write their own WMI module or is there one available out there?
[08-Aug-2009 09:38:51] <brice> only WMI module i've found for python is one that needs to be ran on a windows box
[08-Aug-2009 11:20:58] <brice> or does it use wbemcli
[09-Aug-2009 00:00:46] [disconnected at Sun Aug 9 00:00:46 2009]
[09-Aug-2009 00:00:46] [connected at Sun Aug 9 00:00:46 2009]
[09-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[09-Aug-2009 20:13:57] <Ford> greetings people
[09-Aug-2009 20:17:23] <Ford> what's happening ?
[10-Aug-2009 00:00:46] [disconnected at Mon Aug 10 00:00:46 2009]
[10-Aug-2009 00:00:46] [connected at Mon Aug 10 00:00:46 2009]
[10-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[10-Aug-2009 10:05:12] <dollarbang> question for the zenoss community, we are using ldap security for signon verification. I have added someone, but for some reason, the person is unable to logon. What log file on the server can I read to find the error.
[10-Aug-2009 10:06:55] <venturaville1> that WBEM module is nice, but it is probably something that needs to make it into base zenoss at some point
[10-Aug-2009 10:09:44] <jb> yeah, i havent had a chance to try it yet
[10-Aug-2009 10:16:18] <venturaville1> jb: once the dev server is up I will see if it can be used against our arrays, though it obviously won't be able to get the structure from them the way it is built
[10-Aug-2009 10:16:51] <venturaville1> i.e. the volumes on it
[10-Aug-2009 10:17:42] <venturaville1> looks like he is just using pywbem so it should be possible to replicate the same work
[10-Aug-2009 13:57:59] willwh1 is now known as willwh
[10-Aug-2009 15:38:28] <twm1010> bigegor: I'm having trouble getting any WMI queries to work with the datasource ZenPack
[11-Aug-2009 00:00:46] [disconnected at Tue Aug 11 00:00:46 2009]
[11-Aug-2009 00:00:46] [connected at Tue Aug 11 00:00:46 2009]
[11-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[11-Aug-2009 05:14:26] <Dieterbe> If you make a threshold for a performance template and you select multiple datapoints. will it trigger if any of them hits the specified value, or when the sum of them hits the value?
[11-Aug-2009 07:05:50] <Dieterbe> twm1010: what's your problem? i use it and it works for me
[11-Aug-2009 09:49:15] <raddy> Is it possible to monitor Tomcat via ZenOSS without installing the enterprice edition?
[11-Aug-2009 09:58:51] <twm1010> hrmm.. probably
[11-Aug-2009 09:59:17] <raddy> twm1010 : How?
[11-Aug-2009 09:59:34] zenNoob is now known as chickenandbeans
[11-Aug-2009 10:05:06] <chickenandbeans> hi, need help with a zenoss problem, anyone got a minute or two to spare?
[11-Aug-2009 10:06:26] <jb> anybody here transform savy?
[11-Aug-2009 10:07:04] <chickenandbeans> Just want to know if this has been solved: http://forums.zenoss.com/viewtopic.php?t=6706&highlight=
[11-Aug-2009 10:07:04] <jb> http://pastebin.ca/1525153
[11-Aug-2009 10:07:10] <jb> trying to figure out how I can fix this..
[11-Aug-2009 10:14:19] <chickenandbeans> Any takers on my problem?
[11-Aug-2009 10:17:45] <chickenandbeans> It's the sudden lack of graphing issue: http://forums.zenoss.com/viewtopic.php?t=6706&highlight=
[11-Aug-2009 10:51:57] <TreeStump> hey all
[11-Aug-2009 10:53:23] <TreeStump> im just wondering if someone can provide some assitance
[11-Aug-2009 10:53:46] <TreeStump> i have a zenoss install
[11-Aug-2009 10:53:50] <TreeStump> and for some reason
[11-Aug-2009 10:55:02] <TreeStump> my graphs for bandwidth and performance are wierd
[11-Aug-2009 10:55:24] <TreeStump> isee they drop out for hours at a time, and show absolutely no data
[11-Aug-2009 10:56:36] <TreeStump> when an mrtg graph of the same connections for example are viewed
[11-Aug-2009 10:56:46] <TreeStump> it shows data throughout the day
[11-Aug-2009 10:57:36] <TreeStump> any ideas?
[11-Aug-2009 11:18:42] <twm1010> TreeStump: I would say that if the bandwidth monitoring of interfaces AND the performance graphs for CPU/memory are showing gaps in the same spots
[11-Aug-2009 11:18:53] <twm1010> that your zenoss box is having difficulty with something regarding polling
[11-Aug-2009 11:19:43] <twm1010> or i should say that something is affecting its ability to poll
[11-Aug-2009 11:21:50] <TreeStump> maybe because its behind NAT?
[11-Aug-2009 11:23:00] <twm1010> mmm... perhaps, but is the MRTG box also behind NAT?
[11-Aug-2009 11:23:53] <twm1010> if there are gaps in the data, it means A) the collect did not get to the SNMP agents of the device to retrieve it
[11-Aug-2009 11:24:16] <twm1010> or B) the collector did not attempt to
[11-Aug-2009 11:25:01] <twm1010> how long are the gaps?
[11-Aug-2009 11:39:08] <TreeStump> maybe 6-8 hours
[11-Aug-2009 11:44:01] <twm1010> thats a long time
[11-Aug-2009 11:44:39] <twm1010> is it simultaneous across all devices, just those outside the NAT?
[11-Aug-2009 11:45:37] <TreeStump> well i only have one device inside the NAT at the moment being monitored - i just added another but it wont have data yet
[11-Aug-2009 11:45:47] <TreeStump> that one devise is a cisco switch
[11-Aug-2009 11:46:21] <TreeStump> I havent received data on it since 12:00AM yesterday morning
[11-Aug-2009 11:46:32] <TreeStump> just over 24 hours ago
[11-Aug-2009 11:46:37] <TreeStump> hang on
[11-Aug-2009 11:46:52] <TreeStump> the clock on the machine keeps resetting
[11-Aug-2009 11:47:01] <TreeStump> could that have anything to do with it haha
[11-Aug-2009 12:04:21] richard is now known as Guest39185
[11-Aug-2009 12:04:28] <Guest39185> hey all, started using zenoss today and im wondering how i get it to display a list of software which is installed on a linux box?
[11-Aug-2009 12:24:02] <TreeStump> too far outta my depth mate
[11-Aug-2009 14:09:37] <TreeStump> so after some investigating
[11-Aug-2009 14:10:22] <TreeStump> i noticed, that on each device it reports data as being collected last houurs ago
[11-Aug-2009 14:10:36] <TreeStump> under management i did edit collectors
[11-Aug-2009 14:10:41] <TreeStump> changed some settings
[11-Aug-2009 14:10:47] <TreeStump> was hoping it would check more often
[11-Aug-2009 14:11:10] <TreeStump> i noticed even graphs are missing on the localhost collector
[11-Aug-2009 14:11:25] <TreeStump> is there a way i can reset the settings of the collector?
[11-Aug-2009 14:20:20] <twm1010> TreeStump: the collector picks up changes automatically after some time, but it seems like you're having more "general" problems then settings issues
[11-Aug-2009 14:23:20] <twm1010> Is zenperfsnmp running?
[11-Aug-2009 15:38:30] <TreeStump> twm1010: how do I know if its running?
[11-Aug-2009 15:42:51] <twm1010> go settings look at daemons
[11-Aug-2009 16:00:32] <TreeStump> yeah it says its running
[11-Aug-2009 16:00:48] <TreeStump> zentrap and zenjmx are not though
[11-Aug-2009 18:04:35] <jsm> what format is this supposed to produce? http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch08s05.html? it's seems to be a string of a python list... which i'm not quite sure how to treat as a real list.. I would guess it should be JSON format, but i can't seem to use the JSON module inside of that function.
[12-Aug-2009 00:00:46] [disconnected at Wed Aug 12 00:00:46 2009]
[12-Aug-2009 00:00:46] [connected at Wed Aug 12 00:00:46 2009]
[12-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[12-Aug-2009 08:06:33] <jb> anybody have a reccomendation for a python book?
[12-Aug-2009 08:31:21] <brice> http://diveintopython.org/
[12-Aug-2009 08:57:22] <jb> anybody here have any modeler experience?
[12-Aug-2009 09:22:55] <jb> hrm, can't figure out why my modelers are not gettign compiled.
[12-Aug-2009 09:24:57] <jb> relm -l
[12-Aug-2009 09:38:44] <jb> ah.. no __init__.py's
[12-Aug-2009 11:22:29] <x-ip> hi all
[12-Aug-2009 11:22:52] <x-ip> i'm trying and learning zenoss right now .... its ... awsome :,)
[12-Aug-2009 11:39:51] <jsm> what wire format does zenoss use when it returns data as described here? http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch08s05.html
[12-Aug-2009 11:39:55] <jsm> it is not JSON format.
[12-Aug-2009 12:43:08] burlyscudd__ is now known as burlyscudd
[12-Aug-2009 13:04:33] burlyscudd_ is now known as burlyscudd
[12-Aug-2009 19:41:58] <tpdd> Sorry for the newbie question, but can someone point me to a document that can explain how to use zenoss to monitor a linux dhcp server?
[13-Aug-2009 00:00:46] [disconnected at Thu Aug 13 00:00:46 2009]
[13-Aug-2009 00:00:46] [connected at Thu Aug 13 00:00:46 2009]
[13-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[13-Aug-2009 06:42:11] <etank> anyone had expierence with getting zenoss to monitor a windows cluster?
[13-Aug-2009 06:42:24] <etank> I have 4 that I need to monitor
[13-Aug-2009 06:43:32] <etank> the problem is that i can add each node but when i add the cluster node it says that the IP is already assigned to another node
[13-Aug-2009 07:09:00] <etank> setting zInterfaceMapIgnoreTypes to ethernetCsmacd removes all nics but the loopback :/
[13-Aug-2009 07:15:08] <etank> hmm .. i can still ping it though. just not get stats on the nic
[13-Aug-2009 08:02:28] <jb> morning..
[13-Aug-2009 08:23:45] sakura_ is now known as x-ip
[13-Aug-2009 08:33:24] <x-ip> morning all =)
[13-Aug-2009 09:31:54] <jb> yawn, I wish I could figure out why these .py's aren't compiling.
[13-Aug-2009 09:53:46] <x-ip> anybody knows where is zTransportPreference ?
[13-Aug-2009 09:54:03] <x-ip> in vs 2.4.2 seems isnt at zPreferences
[13-Aug-2009 09:54:04] <x-ip> =)
[13-Aug-2009 09:54:24] <x-ip> also, i want to monitor a remote machine that runs ssh in a different port
[13-Aug-2009 09:55:02] <x-ip> than default 22, i set zCommandPort to the correct port number and seems that zenoss isnt using that value
[13-Aug-2009 10:06:16] <x-ip> hmmm surely i'm wrong ... i'm looking at zenoss/Products/DataCollector/zenmodeler.py, line 268, method collectDevice and the self.cmdCollect taks three args , device, ip and timeout but seems to not take a port number
[13-Aug-2009 10:06:57] <chudler> it probably inherits it from some its parent class, or even higher up the chain
[13-Aug-2009 10:07:12] <x-ip> maybe the use from another port different from 22 (ssh's default) isnt still coded ?
[13-Aug-2009 10:07:59] <x-ip> surely charlieS ... only trying to know why zCommandPort isnt working for me
[13-Aug-2009 10:14:57] <Jane_Curry> <jb> - not sure what .py files you are struggling with but I have just put a paper on "creating Zenoss ZenPacks" on the forum and our website
[13-Aug-2009 10:15:17] <Jane_Curry> It includes some debugging hints and tips
[13-Aug-2009 10:15:34] <Jane_Curry> Get the paper either from http://www.zenoss.com/Members/jcurry/creating-zenoss-zenpacks/
[13-Aug-2009 10:15:54] <Jane_Curry> or from http://www.skills-1st.co.uk/papers/jane/zenpacks/
[13-Aug-2009 10:22:47] <jb> ah one sec
[13-Aug-2009 10:23:01] <jb> i didn't know you had that paper..
[13-Aug-2009 10:23:08] <jb> basically, I have made my new modelers
[13-Aug-2009 10:23:28] <jb> but I'm trying to now define a new class (I guess)
[13-Aug-2009 10:24:22] <jb> the modelers are compiling fine
[13-Aug-2009 10:24:51] <jb> but my classes arent.. and im not sure im doing it correctly.. hopefully this paper will enlighten me
)[13-Aug-2009 10:26:29] <Jane_Curry> <jb> I've only finished the draft of the paper today so it only went public 3 hours back - I'm looking for reviews and comments
[13-Aug-2009 10:26:42] <jb> awesome.. I'm reading it now!
[13-Aug-2009 10:27:06] * x-ip reading it too
[13-Aug-2009 10:43:05] <Jane_Curry> Anyone know of a Zenoss feature or ZenPack that collec6ts "Serial Number" from a Windows box?
[13-Aug-2009 10:43:41] <Jane_Curry> (Don't speak "Doze" very well so I'm not sure whether this is information that is available)
[13-Aug-2009 10:46:30] <jb> certain systems management agents (Dell OpenManage, HP Insight, IBM Director, etc) offer this in their MIB's
[13-Aug-2009 11:06:17] <Jane_Curry> Anyone got an easy method for creating pdf copies of reports?
[13-Aug-2009 11:07:20] <Jane_Curry> There seemed to be a forum thread doing serious discussion of possibilities last October but I don't see any obvious ZenPacks
[13-Aug-2009 11:07:53] <Jane_Curry> I also see a note from Matt Ray saying pdf reports is NOT currently in plan for King Crab - unless someone knows better?
[13-Aug-2009 11:15:19] <ali3> hallow
[13-Aug-2009 11:17:36] <ali3> please help
[13-Aug-2009 11:42:42] <jb> jane?
[13-Aug-2009 11:52:07] <tpdd> can anyone explain how to monitor a dhcpd service (running on a linux box with the command tools). When I model the server it shows up but it won't let me turn on monitoring
[13-Aug-2009 12:34:54] edwardam1 is now known as edwardam
[13-Aug-2009 12:44:00] <twm1010> tpdd: it shows up on the OS tab under services?
[13-Aug-2009 12:46:43] <korozion> Hi there, I have many towers, each with some access points on them. Each AP has many customers. Each tower also has a router and/or a switch and some other snmp equipment. Currently I have an old version of What's Up Gold as seen here ... http://www.linuxgeneration.org/pics/WUG1.jpg. I require the map and the ability to browse to the web page of my APs and customer devices from it. Is this something ZenOSS can help me with?
[13-Aug-2009 12:48:11] <rhett> yes
[13-Aug-2009 12:49:03] <rhett> i wish i would have used zenoss when i was in the wisp/wireless world
[13-Aug-2009 12:49:13] <rhett> much easier to manage by geographical location
[13-Aug-2009 12:52:10] <korozion> hmm
[13-Aug-2009 12:52:16] <korozion> nice. Are the maps easy to edit?
[13-Aug-2009 12:53:30] <rhett> you add locations in zenoss then tag devices to that location. the lines between locations are drawn based on the routing tables
[13-Aug-2009 12:53:43] <rhett> and its all quite easy to do
[13-Aug-2009 12:54:33] <rmatte> korozion: when you create a location, you need to click on that location, you'll see a field where you can click an "edit" link to specify the address of the location, that needs to be filled out properly for the location to show up on the map
[13-Aug-2009 12:59:57] <_mingau_> hello people
[13-Aug-2009 13:00:02] <rmatte> hi
[13-Aug-2009 13:00:05] <_mingau_> i need a little help
[13-Aug-2009 13:00:18] <korozion> rmatte: very nice, thanks
[13-Aug-2009 13:00:48] <twm1010> _mingau_: go ahead
[13-Aug-2009 13:00:58] <_mingau_> i am trying to install default mibs...
[13-Aug-2009 13:01:06] <_mingau_> but i got an error
[13-Aug-2009 13:02:09] <rmatte> default Mibs?
[13-Aug-2009 13:02:45] <rmatte> for what type of OS/Device?
[13-Aug-2009 13:02:54] <_mingau_> yes
[13-Aug-2009 13:02:57] <_mingau_> freebsd 7.2
[13-Aug-2009 13:03:26] <_mingau_> its a fresh install of zenoss core...
[13-Aug-2009 13:03:55] <_mingau_> so i was tryinf to load mibs from share/mibs/iana/
[13-Aug-2009 13:04:15] <rmatte> I never had to install Mibs for Linux, I think that Zenoss already comes with all of the Mibs that you would need to translate traps
[13-Aug-2009 13:04:22] <rmatte> I assume you're using net-snmp?
[13-Aug-2009 13:04:56] <_mingau_> yes
[13-Aug-2009 13:05:03] <_mingau_> hey
[13-Aug-2009 13:05:07] <_mingau_> now i got it...
[13-Aug-2009 13:05:12] <_mingau_> hehehhee
[13-Aug-2009 13:05:27] <_mingau_> SMIPATH was wrong
[13-Aug-2009 13:06:13] <rmatte> ummmm, you aren't installing them as the root user are you?
[13-Aug-2009 13:06:23] <_mingau_> nonono
[13-Aug-2009 13:06:32] <_mingau_> i wrote wrong the SMIPATH
[13-Aug-2009 13:06:38] <_mingau_> God
[13-Aug-2009 13:06:42] <twm1010> rmatte: I'm still moving those mibs, a little at a time...
[13-Aug-2009 13:06:42] <twm1010> haha
[13-Aug-2009 13:06:55] <twm1010> a zenDMD line to just move them all and let it crank, would be nice
[13-Aug-2009 13:07:05] <rmatte> twm1010: brutal
[13-Aug-2009 13:07:25] <rmatte> twm1010: I'm sure that'd be possible somehow, I just don't know how unfortunately hehe
[13-Aug-2009 13:07:43] <_mingau_>
[13-Aug-2009 13:07:46] <twm1010> Im thinking the slowness is filesystem performance on my zenoss server
[13-Aug-2009 13:07:52] <twm1010> disk throughput really
[13-Aug-2009 13:08:07] <twm1010> though CPU on that guy is pretty high alot of the time as well
[13-Aug-2009 13:08:12] <rmatte> twm1010: When I was moving them it was on a 15ghz server, and I had all of the processing power to the one box at the time so it went fairly quickly
[13-Aug-2009 13:09:24] <twm1010> well, no big deal really, its a convenience thing, only really a pain when i add new ones and gotta skim for them
[13-Aug-2009 13:12:43] <rmatte> yeh
[13-Aug-2009 13:13:46] <_mingau_> thx people
[13-Aug-2009 13:20:08] <TreeStump> hheya guys
[13-Aug-2009 13:20:26] <TreeStump> i came in here a few days back wondeing if anyone could help
[13-Aug-2009 13:20:47] <TreeStump> i think now, if i can't find a solution i'll just reformat haha so sick of this damn thign
[13-Aug-2009 13:21:30] <TreeStump> i started having issues where snmp was only collecting data for part of the day, there would be massive gaps in my performance/bandwidth graphs
[13-Aug-2009 13:21:49] <TreeStump> now no data is collected at allby the looks of it
[13-Aug-2009 13:22:05] <TreeStump> any ideas anyone what the prob could be?
[13-Aug-2009 13:23:17] <twm1010> It honestly sounds like the SNMP collector (zenperfsnmp) is not running
[13-Aug-2009 13:29:22] <TreeStump> hmm it id
[13-Aug-2009 13:29:24] <TreeStump> is*
[13-Aug-2009 13:29:30] <rmatte> TreeStump: become the zenoss user and issue the command: zenoss status
[13-Aug-2009 13:29:36] <TreeStump> funny thing though when I checked it out today
[13-Aug-2009 13:29:55] <TreeStump> it said last collection was an hour ago(first time it said that in ages)
[13-Aug-2009 13:30:00] <TreeStump> bu still no data
[13-Aug-2009 13:30:32] <TreeStump> rmatte: how do i become the zenoss user? give it a passwd, and login through SSH?
[13-Aug-2009 13:30:51] <rmatte> TreeStump: nah, login, become root, then "su - zenoss"
[13-Aug-2009 13:35:49] <twm1010> either way will do.
[13-Aug-2009 13:36:13] <twm1010> I think you need to do some basic troubleshooting here. 1. can zenoss ping the devices
[13-Aug-2009 13:36:19] <twm1010> 2. can zenoss SNMPwalk the device
[13-Aug-2009 13:36:26] <rmatte> yeh, I just find it best not to allow direct logins to the Zenoss account, one less security issue to worry about
[13-Aug-2009 13:36:37] <twm1010> 3. are there events being logged that might indicate what your problem is
[13-Aug-2009 13:37:28] <rmatte> there's a bunch of things that could be wrong, you might need to remodel devices, reindex the database, make sure that zenperfsnmp isn't erroring out somehow, stuff like that.
[13-Aug-2009 13:37:31] <jb> rmatte: have you by chance developed any new object classes for your zenpacks?
[13-Aug-2009 13:37:50] <TreeStump> hmm as much as I hate to say it, i think i have much larger problems to worry about, now that I look at it
[13-Aug-2009 13:37:53] <rmatte> jb: no, I haven't
[13-Aug-2009 13:38:12] <tpdd> twn1010: it does show up but not as monitored
[13-Aug-2009 13:38:18] <TreeStump> seems like some major major network conjestion
[13-Aug-2009 13:38:21] <rmatte> TreeStump: what are you seeing?
[13-Aug-2009 13:38:29] <rmatte> ah
[13-Aug-2009 13:38:53] <TreeStump> actually no not network conjestion
[13-Aug-2009 13:38:58] <rmatte> that'd do it, I'd recommend upping your snmp timeout value in zproperties, and upping your collector cycle times in properties for the collector
[13-Aug-2009 13:39:19] <TreeStump> jsut some really wierd network issues
[13-Aug-2009 13:39:25] <TreeStump> hmm ok
[13-Aug-2009 13:39:28] <TreeStump> thanks for the tips
[13-Aug-2009 13:39:31] <rmatte> np
[13-Aug-2009 13:39:49] <TreeStump> ill ook into this network issue and then come back to you if i don't get anywhere after that
[13-Aug-2009 13:39:51] <TreeStump> thanks gain
[13-Aug-2009 13:39:54] <TreeStump> again*
[13-Aug-2009 13:44:07] <jb> any devs around?
[13-Aug-2009 13:46:35] <rmatte> Have there been any announcements as to when 2.4.3 is going to be released?
[13-Aug-2009 13:47:38] <twm1010> haven't seen anything
[13-Aug-2009 14:15:59] <TreeStump> ok so network issues were nothing
[13-Aug-2009 14:16:13] <TreeStump> still cant seem to get anything
[13-Aug-2009 14:16:18] <TreeStump> i can model devides
[13-Aug-2009 14:16:21] <TreeStump> devices*
[13-Aug-2009 14:16:34] <TreeStump> and ping them
[13-Aug-2009 14:17:43] <tpdd> so if you have to monitor upd based services is the prefered way to do that by using nagios plugins?
[13-Aug-2009 14:22:14] <TreeStump> zenperfsnmp is running
[13-Aug-2009 14:22:27] <rmatte> check $ZENHOME/logs/zenperfsnmp.log
[13-Aug-2009 14:22:33] <rmatte> see if there are any errors
[13-Aug-2009 14:22:46] <rmatte> you might also want to try doing "zenperfsnmp start -v10"
[13-Aug-2009 14:22:58] <rmatte> and see what it spits out
[13-Aug-2009 14:24:38] <jb> argh.. what daemon actually does something with object classes?
[13-Aug-2009 14:24:45] <jb> i cannot figure out why these won't compile.
[13-Aug-2009 14:24:57] <rmatte> jb: what issue are you experiencing relating to them?
[13-Aug-2009 14:25:06] <TreeStump> zenperfsnmp is running
[13-Aug-2009 14:25:07] <jb> they aren't compiling
[13-Aug-2009 14:25:08] <jb>
[13-Aug-2009 14:25:13] <TreeStump> is what i reports
[13-Aug-2009 14:25:33] <rmatte> jb: when starting Zenoss, or what?
[13-Aug-2009 14:25:35] <jb> yes
[13-Aug-2009 14:25:39] <jb> or zenhub..
[13-Aug-2009 14:25:44] <rmatte> jb: hmmm, what's the exact error?
[13-Aug-2009 14:25:48] <jb> there is none..
[13-Aug-2009 14:25:56] <rmatte> zenhub run -v10
[13-Aug-2009 14:26:00] <jb> yeah, nothing..
[13-Aug-2009 14:26:04] <TreeStump> the log reports that the cycle completes with success... but i see there are rrdtool errors
[13-Aug-2009 14:26:06] <rmatte> (how do you know it's not compiling them?)
[13-Aug-2009 14:26:13] <jb> because its not creating the .pyc's
[13-Aug-2009 14:26:18] <rmatte> ah
[13-Aug-2009 14:26:20] <rmatte> gotcha
[13-Aug-2009 14:26:34] <rmatte> hmmm
[13-Aug-2009 14:26:39] <rmatte> when did this start happening?
[13-Aug-2009 14:26:46] <jb> this is a custom zenpack im working on..
[13-Aug-2009 14:26:54] <rmatte> oh
[13-Aug-2009 14:26:56] <jb> im not entirely sure I have the relationships and classes setup properly
[13-Aug-2009 14:27:00] <jb> which could be part of the problem
[13-Aug-2009 14:27:06] <rmatte> probably is lol
[13-Aug-2009 14:27:28] <jb> this stuff is really confusing
[13-Aug-2009 14:27:31] <rmatte> I don't have any real experience with that, I'm just starting to learn the complexities of ZenPack editing
[13-Aug-2009 14:27:37] <jb> even after reading jane's paper and looking at egor's packs..
[13-Aug-2009 14:27:46] <jb> yeah, me too
[13-Aug-2009 14:27:50] <kgoedtel> does anyone know if there is a way or if someone has attempted to make a threshold/alert based on multiple conditions being true?
[13-Aug-2009 14:27:55] <jb> i have my modelers built!
[13-Aug-2009 14:28:16] <rmatte> kgoedtel: you
[13-Aug-2009 14:28:18] <rmatte> oops
[13-Aug-2009 14:28:34] <rmatte> kgoedtel: you'd have to make the threshold and then apply a transform to it
[13-Aug-2009 14:29:01] <kgoedtel> yeah i was hoping i could somehow do it all within a template
[13-Aug-2009 14:29:33] <rmatte> kgoedtel: nope, Zenoss isn't setup for that at the moment
[13-Aug-2009 14:29:47] <kgoedtel> damn, logical operators for that would be nice :]
[13-Aug-2009 14:29:53] <rmatte> Here's an example of a threshold transform: http://www.zenoss.com/Members/netdata/show-the-usage-in-a-filesytem-threshold/
[13-Aug-2009 14:30:25] <kgoedtel> thanks
[13-Aug-2009 14:30:31] <rmatte> np
[13-Aug-2009 14:31:57] <rmatte> well, I'm on vacation and I've been lounging around all day so far, time to shower & get dressed lol
[13-Aug-2009 14:32:02] <rmatte> afk
[13-Aug-2009 15:15:04] <rmatte> rmatte test
[13-Aug-2009 15:15:08] <rmatte> oops
[13-Aug-2009 16:23:04] <cwj> i have an instance of zenoss-stack core which i have configured for monitoring a bunch of devices via SNMP and Nagios/NRPE checks. the monitoring was working fine for like two weeks, and now i am seeing a ton of failures for both SNMP agents and NRPE checks, even though the same checks succeed every time I run them from the command line on the server where this instance of zenoss is running.
[13-Aug-2009 16:23:18] <cwj> i've been reviewing the logs but still can't make sense of why things are failing
[13-Aug-2009 16:23:36] <cwj> the NRPE checks in zenoss are returning exit code 2
[13-Aug-2009 16:23:47] <cwj> but they return exit code 0 when run from the command line
[13-Aug-2009 16:24:07] <cwj> can anyone suggest something i might have missed?
[13-Aug-2009 21:07:58] <alley> hello
[13-Aug-2009 21:17:47] <alley> ello
[13-Aug-2009 21:17:48] <alley> allow
[13-Aug-2009 22:26:10] <alley> alow
[13-Aug-2009 22:42:50] <alley> halow all
[13-Aug-2009 22:44:07] <korozion> you ok?
[14-Aug-2009 00:00:46] [disconnected at Fri Aug 14 00:00:46 2009]
[14-Aug-2009 00:00:46] [connected at Fri Aug 14 00:00:46 2009]
[14-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[14-Aug-2009 05:18:59] [2]folke is now known as folke
[14-Aug-2009 07:24:05] <baffle> I'm experiencing huge gaps in everyting that is polled over SNMP; What are the common cause for this?
[14-Aug-2009 07:32:37] <folke> baffle, is the server under heavy load?
[14-Aug-2009 07:40:00] <baffle> folke: Not really.
[14-Aug-2009 07:41:19] <baffle> folke: Or, well, it sometimes has some processes that uses a lot of CPU, but it is a quad core 2.33 with 22G ram. 4G is used by ZenOss, rest is fscache.
[14-Aug-2009 07:41:36] <baffle> folke: 5 minute load is 0.18.
[14-Aug-2009 08:12:23] <folke> baffle, then it should not be a problem..
[14-Aug-2009 08:13:37] <baffle> folke: But it is. :-)
[14-Aug-2009 08:14:39] <baffle> folke: Hmm, I have a lot of backtraces in the perfsnmp log, due to some devices reporting the wrong values. I.e. something that usd to be an int is now a string.. Does a crash like that impact all other devices that happend to be polled by the same thread?
[14-Aug-2009 08:38:57] <folke> baffle, Are there some device that is always having trouble..
[14-Aug-2009 08:39:52] <folke> I had one that messed up the zenperfsnmp deamon.. But that was in 2.0.x i think.
[14-Aug-2009 08:44:32] <baffle> folke: Yes, there is a few devices that bakes it barf.
[14-Aug-2009 08:45:51] <folke> if you disable snmp for a moment on those efvil ones, and see if the gaps stops on the working ones?
[14-Aug-2009 16:38:08] <otakup0pe> hmm how do i get zenoss to track network utilization
[14-Aug-2009 16:38:18] <otakup0pe> like total down/upload per interface per "time period"
[14-Aug-2009 16:39:03] <otakup0pe> rather than bw/sec i want to know N GB per day
[14-Aug-2009 17:46:56] <otakup0pe> oh come on there has to be someone in here who is still sober
[14-Aug-2009 17:46:57] <otakup0pe>
[14-Aug-2009 18:03:37] <rhett> haha
[14-Aug-2009 18:04:26] <rhett> i may be able to be of some help
[14-Aug-2009 18:47:19] <otakup0pe> :3
[14-Aug-2009 18:47:27] <otakup0pe> did my question make sense ?
[14-Aug-2009 18:58:35] <rhett> yeah it makes perfect sense. i monitor all that in cacti. just busy and dont know if anyone has already done that or not
[14-Aug-2009 18:58:48] <rhett> for cacti that is
[14-Aug-2009 18:58:59] <rhett> errrr zenoss
[14-Aug-2009 19:15:54] <otakup0pe> what snmp oid's do you use
[14-Aug-2009 19:19:04] <rhett> you use the in/out octets oid's and then have a comment print the ammount of bytes
[14-Aug-2009 19:19:48] <rhett> in cacti for example i use COMMENT:Total Out: |sum:auto:current:2:auto|
[14-Aug-2009 19:19:56] <otakup0pe> rather than do any fancy math on it to get the kbps
[14-Aug-2009 19:19:56] <otakup0pe> ok
[14-Aug-2009 19:19:59] <otakup0pe> i'll poke at this
[15-Aug-2009 00:00:46] [disconnected at Sat Aug 15 00:00:46 2009]
[15-Aug-2009 00:00:46] [connected at Sat Aug 15 00:00:46 2009]
[15-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[15-Aug-2009 16:50:54] <dcoderhed> Howdy
[16-Aug-2009 00:00:46] [disconnected at Sun Aug 16 00:00:46 2009]
[16-Aug-2009 00:00:46] [connected at Sun Aug 16 00:00:46 2009]
[16-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[16-Aug-2009 04:24:38] <philjray> Hi All
[16-Aug-2009 04:25:00] <philjray> I am new to Zenoss and have a question regarding MIBs
[16-Aug-2009 04:25:46] <philjray> Having looked at the documentation and wiki etc. more info is still required - by me anyway!
[16-Aug-2009 04:39:47] <philjray> Hi All
[16-Aug-2009 04:39:54] <philjray> I am new to Zenoss and have a question regarding MIBs
[16-Aug-2009 04:40:02] <philjray> Having looked at the documentation and wiki etc. more info is still required - by me anyway!
[16-Aug-2009 04:41:04] <philjray> I have the 3rd party MIB installed on Zenoss via the web interface
[16-Aug-2009 04:41:31] <philjray> What I need to know is how to "apply" that MIB to the hardware that I want to monitor
[16-Aug-2009 20:08:39] <ironpaw> hi
[16-Aug-2009 20:08:40] <ironpaw> anyone here?
[16-Aug-2009 20:08:50] <ironpaw> why wont my snmpignore=true work?
[16-Aug-2009 20:08:53] <ironpaw> even when pushing changes?
[16-Aug-2009 20:08:59] <ironpaw> as per this thread:
[16-Aug-2009 20:09:01] <ironpaw> http://forums.zenoss.com/viewtopic.php?p=28410
[16-Aug-2009 20:09:05] <ironpaw> this is 2.4.2
[17-Aug-2009 00:00:46] [disconnected at Mon Aug 17 00:00:46 2009]
[17-Aug-2009 00:00:46] [connected at Mon Aug 17 00:00:46 2009]
[17-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[17-Aug-2009 07:32:41] <_mingau_> hi people
[17-Aug-2009 07:35:08] <_mingau_> does anybody has info about wmi monitoring in zenoss?
[17-Aug-2009 07:43:49] <_mingau_> hello?
[17-Aug-2009 09:10:37] * rmatte is back from vacation
[17-Aug-2009 09:11:54] <straterra> how was it?
[17-Aug-2009 09:32:14] <twm1010> rmatte: you around?
[17-Aug-2009 09:34:56] <rmatte> yeh
[17-Aug-2009 09:35:03] <rmatte> fairly busy today, but what's up?
[17-Aug-2009 09:35:16] <rmatte> straterra: quite relaxing
[17-Aug-2009 09:37:16] <twm1010> you done any transformation of poller results?
[17-Aug-2009 09:37:36] <twm1010> take a value from an OID, pump it through a formula to change, like celsius to fahrenheit, etc
[17-Aug-2009 09:47:53] <rmatte> no, I haven't
[17-Aug-2009 09:55:11] <rmatte> The way I'd do it is to write a script to collect the data and convert it, then just use that as a command based datapoint
[17-Aug-2009 09:55:51] <rmatte> You can convert the data in the graph using RPN, but that won't actually change the values stored in the RRD itself
[17-Aug-2009 09:59:09] <twm1010> that's fine, actually
[17-Aug-2009 10:52:08] <fuzz> i recently upgraded to zenoss-stack 2.4.2, and it seems that gaps started to appear in a lot of graphs after this. changing workers from 2 to 4 in zenhub seems to have helped somewhat, but many hosts still have ~ 1-5 hours gaps between every 0.1-1 hours of data.. could this be related to a known issue in 2.4.2?
[17-Aug-2009 10:52:36] <fuzz> it also seems that this only applies to graphs generated by zenperfsnmp
[17-Aug-2009 10:52:40] <fuzz> ping/process graphs seems fine
[17-Aug-2009 10:54:49] <baffle> And also graphs done by executing commands.
[17-Aug-2009 10:56:30] <rmatte> fuzz: network latency?
[17-Aug-2009 10:59:07] <fuzz> rmatte: i doubt it.. one of the affected boxes is on the same subnet as the zenoss box
[17-Aug-2009 11:00:27] <rmatte> hmmm
[17-Aug-2009 11:00:48] <rmatte> how many devices are you monitoring, and what kind of hardware is on the Zenoss server?
[17-Aug-2009 11:02:10] <fuzz> about 500 devices
[17-Aug-2009 11:02:13] <fuzz> let me check the hw
[17-Aug-2009 11:02:15] <baffle> HW is IBM x3650, Quad Core 2.33, 20G ram.
[17-Aug-2009 11:02:19] <fuzz> right
[17-Aug-2009 11:03:23] <baffle> RAID 10 15k disks if memory serves me right.
[17-Aug-2009 11:03:37] <baffle> fuzz: How many hosts are there again?
[17-Aug-2009 11:04:07] <fuzz> baffle: there are 425 dirs under perf/Devices so i guess the same number of hosts
[17-Aug-2009 11:05:17] <rmatte> hmmm, definitely not a hardware issue
[17-Aug-2009 11:05:52] <rmatte> Have you checked the zenperfsnmp log file?
[17-Aug-2009 11:06:10] <fuzz> yes.. i've grepped for a host i know has gaps in the graphs, and it is not mentioned
[17-Aug-2009 11:06:32] <fuzz> i also tried to log each rrd update and it does not seem to try to save rrd data for that host
[17-Aug-2009 11:06:40] <fuzz> don't know if it tries to poll it yet
[17-Aug-2009 11:07:36] <rmatte> have you played around with any collector settings?
[17-Aug-2009 11:07:40] <rmatte> such as cycle time?
[17-Aug-2009 11:09:28] <baffle> collector settings: http://paste2.org/p/383519
[17-Aug-2009 11:09:38] <baffle> SNMP is 300.
[17-Aug-2009 11:09:56] <baffle> I think it is the default.
[17-Aug-2009 11:10:01] <rmatte> yeh it is
[17-Aug-2009 11:10:13] <rmatte> are you monitoring a lot of servers, or mostly network?
[17-Aug-2009 11:10:15] <fuzz> btw.. the host i mentioned turned out to not be on the subnet after all. but i'm still pretty convinced it's not a network issue. i can snmpwalk the box fine from the zenoss box..
[17-Aug-2009 11:10:35] <baffle> rmatte: Most is servers, but there are some switches/routers as well.
[17-Aug-2009 11:10:45] <rmatte> estimate of how many?
[17-Aug-2009 11:11:03] <baffle> rmatte: Like.. 4 big switcher/routers, 10-20 switches.
[17-Aug-2009 11:11:11] <rmatte> servers I meant
[17-Aug-2009 11:11:12] <rmatte> sorry
[17-Aug-2009 11:11:25] <baffle> Around 400 I guess.
[17-Aug-2009 11:11:27] <rmatte> (servers are the most intensive to collect data for)
[17-Aug-2009 11:11:31] <rmatte> ah, quite a few
[17-Aug-2009 11:11:54] <rmatte> you upgraded to 2.4.2 from which version?
[17-Aug-2009 11:12:01] <baffle> 2.4.0 or 2.4.1.
[17-Aug-2009 11:12:19] <rmatte> hmmm
[17-Aug-2009 11:12:46] <baffle> rmatte: Actually, 4 big switches are Cisco 6513 wich loads of 48 port switchmodules, so.. I think those are the worst. :-)
[17-Aug-2009 11:13:22] <rmatte> well, possibly, are you monitoring port operational status via a threshold?
[17-Aug-2009 11:13:50] <rmatte> such as: http://www.zenoss.com/Members/cluther/polling-interface-status/
[17-Aug-2009 11:13:50] <baffle> Graphs for the switches are OK btw.
[17-Aug-2009 11:14:29] <rmatte> the servers that aren't graphing, if you load up a list of servers, is it the servers near the end of the list that aren't graphing
[17-Aug-2009 11:14:33] <rmatte> or does it seem random?
[17-Aug-2009 11:14:53] <baffle> rmatte: Hmm, no, I don't think we do.
[17-Aug-2009 11:15:08] <baffle> rmatte: What is the best way to get the list of servers?
[17-Aug-2009 11:16:12] <rmatte> hmmm, I guess you don't group them the same way we do lol...
[17-Aug-2009 11:16:38] <rmatte> we have organizers under Groups for Network, Security, and Server
[17-Aug-2009 11:16:53] <rmatte> not sure, really depends on how you organize them
[17-Aug-2009 11:17:00] <rmatte> are they all under the same devices organizer?
[17-Aug-2009 11:17:33] <baffle> rmatte: Hmm, we have /Network /Server etc, and under /Server we have OS and Vendor etc etc.
[17-Aug-2009 11:18:10] <baffle> But if servers are polled alphabetically, then there is no diffence from servers starting with a or z.
[17-Aug-2009 11:18:53] <rmatte> yeh, they are polled alphabetically in my experience
[17-Aug-2009 11:20:58] <fuzz> it may be worth mentioning that we had an issue (around the same time as the update) with modeling some devices that had a network interface with ip in 10.0.0.0/24. some python function recursed infinitely and eventually hit the max recursion limit and died, apparently because the 10.0.0.0 network object was a child of itself (looking at the object in the zope db showed 10.0.0.0/10.0.0.0/10.0.0.0/10.0.0.0/etc..). i had to delete the object from the zope db..
[17-Aug-2009 11:21:08] <fuzz> i wouldn't think it should be related, but who knows..
[17-Aug-2009 11:25:34] <rmatte> hmmmm
[17-Aug-2009 11:25:58] <rmatte> so increasing the hub workers did help a bit?
[17-Aug-2009 11:26:00] <fuzz> btw.. there are devices with gaps that does not have a network interface in 10.0.0.0
[17-Aug-2009 11:26:06] <baffle> rmatte: Not, not really.
[17-Aug-2009 11:26:31] <rmatte> fuzz: gaps?
[17-Aug-2009 11:26:41] <baffle> Actually, most servers have public IPs.
[17-Aug-2009 11:26:46] <fuzz> rmatte: gaps in the graphs.. missing data in the rrds..
[17-Aug-2009 11:27:02] <rmatte> ah
[17-Aug-2009 11:27:21] <rmatte> well, if the gaps are in 10.0.0.0, it may be related to that change
[17-Aug-2009 11:27:30] <baffle> No, graphs are everywhere.
[17-Aug-2009 11:27:54] <baffle> We have a few /16 and /17 nets, so servers are spread all over.
[17-Aug-2009 11:28:00] <rmatte> all I can think of is that snmp is timing out while polling
[17-Aug-2009 11:28:21] <rmatte> try increasing the cycle and snmp values for the collector
[17-Aug-2009 11:28:31] <rmatte> increase the snmp timeout value in zPropeties under /Devices as well
[17-Aug-2009 11:29:06] <fuzz> rmatte: would such a timeout be visible in the log? is there somewhere in the code i can add a debug statement to check it?
[17-Aug-2009 11:29:52] <rmatte> it would be visible I believe
[17-Aug-2009 11:30:10] <rmatte> but I don't have any logs as reference to check since I've got my stuff pretty well tuned
[17-Aug-2009 11:30:21] <rmatte> try this...
[17-Aug-2009 11:30:30] <rmatte> as zenoss user: zenperfsnmp run -v10
[17-Aug-2009 11:30:35] <rmatte> and watch it run
[17-Aug-2009 11:30:46] <rmatte> see if there are any errors
[17-Aug-2009 11:33:19] <fuzz> seems fine
[17-Aug-2009 11:33:20] <fuzz> http://pastebin.com/d105833dc
[17-Aug-2009 11:33:35] <fuzz> it completed suspiciously quickly
[17-Aug-2009 11:34:08] <fuzz> success:0 fail:0 pending:0 todo:0
[17-Aug-2009 11:34:09] <fuzz> hm?
[17-Aug-2009 11:36:44] <rmatte> yeh, it does the same for me, hmmm
[17-Aug-2009 11:37:22] <baffle> Maybe you should stop zenperfsnmp first.
[17-Aug-2009 11:37:26] <baffle> I.e. the daemon.
[17-Aug-2009 11:37:33] <rmatte> zenperfsnmp run -v 10 --cycle
[17-Aug-2009 11:37:36] <rmatte> try that
[17-Aug-2009 11:37:46] <rmatte> yeh, you might have to stop the daemon that's running
[17-Aug-2009 11:38:27] <rmatte> yeh --cycle works
[17-Aug-2009 11:38:58] <rmatte> or seems to anyways
[17-Aug-2009 11:39:17] <fuzz> yes.. that's better. running now.. no errors so far.. lots of "Updating threshold..."
[17-Aug-2009 11:39:25] <rmatte> yeh, that's normal
[17-Aug-2009 11:45:20] <fuzz> still running..
[17-Aug-2009 11:45:50] <rmatte> yeh, could take a while, you have a lot of servers lol
[17-Aug-2009 11:45:52] <fuzz> some warnings about '38 degrees C' cannot be parsed as float etc..
[17-Aug-2009 11:46:04] <fuzz> but not for (all) the relevant hosts
[17-Aug-2009 11:46:08] <rmatte> that wouldn't cause this
[17-Aug-2009 11:46:22] <rmatte> that's just complaing that a string can't be parsed as a numeric
[17-Aug-2009 11:46:31] <fuzz> right..
[17-Aug-2009 11:46:45] <fuzz> other warnings about hosts down and bad oids, but still not for (all) the relevant hosts
[17-Aug-2009 11:47:29] <rmatte> yeh, you're really just looking for failures, errors, or timeouts
[17-Aug-2009 11:50:31] <fuzz> mm
[17-Aug-2009 11:50:37] <fuzz> i'll get back to you when it's done :-)
[17-Aug-2009 11:50:58] <TreeStump> heya guys
[17-Aug-2009 11:51:11] <twm1010> Hello
[17-Aug-2009 11:51:49] <TreeStump> me again haha
[17-Aug-2009 11:52:02] <twm1010> had any luck?
[17-Aug-2009 11:52:17] <TreeStump> i reinstalled the os and zenoss
[17-Aug-2009 12:02:19] <TreeStump> so far so good
[17-Aug-2009 12:02:19] <TreeStump> im wodering though
[17-Aug-2009 12:02:20] <TreeStump> wondering*
[17-Aug-2009 12:02:26] <TreeStump> is it possible to add a zenuser who can only view the devices they are assigned to administer?
[17-Aug-2009 12:02:40] <rmatte> not in Core
[17-Aug-2009 12:02:53] <rmatte> device ACLs are a feature of enterprise
[17-Aug-2009 12:03:02] <TreeStump> nope?
[17-Aug-2009 12:03:06] <TreeStump> only in the enterprise version?
[17-Aug-2009 12:03:10] <rmatte> correct
[17-Aug-2009 12:03:10] <TreeStump> tim to see how much that costs then
[17-Aug-2009 12:03:28] <TreeStump> time*
[17-Aug-2009 12:03:30] <TreeStump> thanks
[17-Aug-2009 12:03:43] <rmatte> http://www.zenoss.com/product/pricing
[17-Aug-2009 12:03:45] <rmatte> np
[17-Aug-2009 12:07:09] <TreeStump> i thought it was cheap
[17-Aug-2009 12:07:16] <TreeStump> then I just realised
[17-Aug-2009 12:07:25] <TreeStump> its per resource
[17-Aug-2009 12:07:45] <TreeStump> haha
[17-Aug-2009 12:07:48] <TreeStump> oh man
[17-Aug-2009 12:09:16] <rmatte> yeh, it's quite pricy
[17-Aug-2009 12:10:31] <rmatte> You can assign admin access to specific devices per user in Core
[17-Aug-2009 12:10:38] <rmatte> but they'll still be able to view everything
[17-Aug-2009 12:11:18] <rmatte> If you go to Settings -> Users -> <click on a user> -> Administered Objects
[17-Aug-2009 12:13:02] <rmatte> http://www.zenoss.com/community/docs/zenoss-guide/2.2.4/ch06s02.html
[17-Aug-2009 12:13:39] <straterra> I'm about to try a zenoss migration
[17-Aug-2009 12:17:44] <TreeStump> matte: thanks, I already tried that because I assumed that it would only allow them access to those devices... thanks anyways...
[17-Aug-2009 12:18:41] <TreeStump> actually for the functionality I need, its probably cheaper for me to buy heaps of VPS's and just create multiple installs of zenoss
[17-Aug-2009 12:19:45] <fuzz> rmatte: the --cycle flag makes zenperfsnmp run over and over again? i have multiple "*** Cycle completed ***, Sent xxx oid requests, Queried 297 devices" entries in the output now..
[17-Aug-2009 12:21:52] <rmatte> fuzz: I believe it simulates running as a daemon, but in the foreground
[17-Aug-2009 12:22:02] <rmatte> ctrl-c out of it and the background daemon will take over
[17-Aug-2009 12:22:03] <fuzz> rmatte: grepped for CRITICAL in the output now, and i have see no relevant lines, only misc (other) oids that could not be polled/saved
[17-Aug-2009 12:22:13] <rmatte> grep for fail:
[17-Aug-2009 12:22:15] <rmatte> and error
[17-Aug-2009 12:23:57] <fuzz> well.. there are quite a few matches
[17-Aug-2009 12:24:05] <fuzz> but not much relevant it seems
[17-Aug-2009 12:24:11] <fuzz> INFO:zen.zenperfsnmp:success:289 fail:2 pending:6 todo:0
[17-Aug-2009 12:25:14] <fuzz> nothing for |grep -i fail|grep <troubledevice> or |grep -i error|grep <troubledevice>
[17-Aug-2009 12:25:21] <fuzz> however..
[17-Aug-2009 12:25:35] <fuzz> grepping for trouble devices now.. it suddenly seems to have updated some rrds
[17-Aug-2009 12:27:34] <TreeStump> do you think it would be possible to write a zenpack(plugin) which would allow the extra ACL functionality?
[17-Aug-2009 12:28:30] <chudler> yes, it is possible. It is usually how the enterprise features work
[17-Aug-2009 12:29:44] <TreeStump> because there is no way im gonna have enough devices to even qualify for buying enterprise
[17-Aug-2009 12:32:01] <baffle> TreeStump: We have a custom portal for our customers, where they get access to their graphs via a magic proxy.
[17-Aug-2009 12:33:47] <malbon> yo folks.
[17-Aug-2009 12:34:39] <malbon> has anyone used sha or md5 authentication digests with the web front end? I would like to use rest calls with some of our apps but don't want to give the developers and users the real password.
[17-Aug-2009 12:35:48] <rmatte> fuzz: well, you said the problem is intermittent, so checking logs during the times that the data wasn't being collected would be your best bet, you'd have to make sure that the logs haven't rolled over since then as well
[17-Aug-2009 12:40:41] <fuzz> rmatte: i grepped for a few trouble devices in the log earlier (the log should not have rolled over by then), and there were no matches..
[17-Aug-2009 12:40:42] <baffle> malbon: I.e. set up anonymous auth and allow people in with apache auth? Might work.
[17-Aug-2009 12:40:53] <baffle> fuzz: But did the RRDs update?
[17-Aug-2009 12:42:04] <fuzz> baffle: for appdev (a trouble device) they did now yes..
[17-Aug-2009 12:47:16] <rmatte> fuzz: I'm out of ideas then
[17-Aug-2009 12:47:22] <rmatte> maybe check zenhub log?
[17-Aug-2009 12:49:45] <fuzz> actually.. it seems it has updated almost all the rrds now, after the zenperf.. -cycle.. had run for a while
[17-Aug-2009 12:50:11] <rmatte> sounds like a timeout issue then
[17-Aug-2009 12:50:16] <fuzz> baffle: /srv/zenoss/check-dead-rrds.py [maxage] - shows rrds without data since maxage
[17-Aug-2009 12:50:27] <rmatte> increase your timeout values and cycle time by 100 or so
[17-Aug-2009 12:50:43] <fuzz> i guess.. strange that there was no entries in the log though..
[17-Aug-2009 12:50:44] <fuzz> will do
[17-Aug-2009 12:51:59] <malbon> baffle: yes, I was thinking about that to kerberise it, but I'd prefer not to go down that route at this time.
[17-Aug-2009 12:52:35] <straterra> When I migrate to a new zenoss install..do I have to apply templates to their groups before I do the import..or will the import that care of that?
[17-Aug-2009 12:52:59] <baffle> fuzz: Maybe we should update zenperfsnmp threads?
[17-Aug-2009 12:53:35] <rmatte> straterra: importing devices?
[17-Aug-2009 12:53:48] <straterra> No
[17-Aug-2009 12:53:58] <straterra> I'm doing a full backup of one zenoss and importing in a fresh install of another
[17-Aug-2009 12:54:14] <straterra> Basically from a VM to a real machine
[17-Aug-2009 12:54:17] <rmatte> I see, I'd apply them beforehand just to be safe
[17-Aug-2009 12:54:23] <rmatte> you'll have to apply them either way
[17-Aug-2009 12:55:24] <straterra> by hand again?
[17-Aug-2009 12:55:43] <rmatte> I don't know what setup you had before
[17-Aug-2009 12:55:49] <rmatte> explain?
[17-Aug-2009 12:56:00] <straterra> I have a Centos 5.3 vm with everything configured how I want it
[17-Aug-2009 12:56:15] <straterra> I want to move it to a real machine..so..I have a machine installed with centos 5.3
[17-Aug-2009 12:56:26] <rmatte> why not just copy the whole $ZENHOME directory over?
[17-Aug-2009 12:56:29] <straterra> I was going to do a backup in zenoss for the vm and import it
[17-Aug-2009 12:56:47] <straterra> I have to dump the db and such too, right?
[17-Aug-2009 12:57:07] <rmatte> you have to backup anything in the mysql DB, but not in Zope
[17-Aug-2009 12:57:19] <rmatte> if you copy $ZENHOME the zope db will be copied along with it
[17-Aug-2009 12:57:21] <rmatte> it's just a flat file
[17-Aug-2009 12:57:32] <straterra> If I rsync $ZENHOME over, and then do a db dump/import..I should be fine?
[17-Aug-2009 12:57:43] <rmatte> yes
[17-Aug-2009 12:57:56] <rmatte> If it doesn't work I'll be really surprised
[17-Aug-2009 12:58:03] <rmatte> did you use stack installer to install though?
[17-Aug-2009 12:58:06] <straterra> Might as well try it
[17-Aug-2009 12:58:09] <straterra> I used the RHEL RPM's
[17-Aug-2009 12:58:13] <rmatte> hmmmm
[17-Aug-2009 12:58:21] <straterra> Same OS..same RPM's
[17-Aug-2009 12:58:29] <rmatte> you might want to do a clean install of Zenoss, then copy $ZENHOME over, then dump/import
[17-Aug-2009 12:58:39] <straterra> I have a clean install of zenoss
[17-Aug-2009 12:58:43] <straterra> Just installed it
[17-Aug-2009 12:58:50] <rmatte> same version?
[17-Aug-2009 12:58:54] <straterra> Yes
[17-Aug-2009 12:58:58] <straterra> Everything is the same
[17-Aug-2009 12:59:05] <rmatte> then yes, copy $ZENHOME over and dump/import should work fine
[17-Aug-2009 12:59:18] <straterra> Alrighty..waiting on this slow VM to shut down zenoss
[17-Aug-2009 12:59:23] <rmatte> you should essentially have the exact same settings that you had with the old install
[17-Aug-2009 12:59:30] <rmatte> lol
[17-Aug-2009 12:59:40] <straterra> Thats why we're moving to a dedicated machine
[17-Aug-2009 12:59:43] <straterra> The VM is ungodly slow
[17-Aug-2009 12:59:54] <rmatte> we have VMs, but they are on fast servers
[17-Aug-2009 13:00:15] <straterra> Our servers are old..and are supposed to be upgraded extremely soon
[17-Aug-2009 13:00:28] <rmatte> hehe
[17-Aug-2009 13:00:40] <rmatte> some companies don't understand the words "end of life cycle"
[17-Aug-2009 13:01:32] <straterra> God, I know
[17-Aug-2009 13:01:50] <straterra> zenoss still isn't shut down
[17-Aug-2009 13:02:01] <venturaville> ditto (said while staring at a long list of 10 year old boxes in inventory)
[17-Aug-2009 13:02:03] <rmatte> that's brutal
[17-Aug-2009 13:02:11] <rmatte> I'm surprised it ran at all on that thing
[17-Aug-2009 13:02:24] <scp2004> Hi guys ..Im a Zenoss user from Brazil . And monitoring 475 hosts .. I need some help with a problem . Somebody ?
[17-Aug-2009 13:02:25] <straterra> It wasn't pretty
[17-Aug-2009 13:02:36] <straterra> scp2004: sure..just state your issue
[17-Aug-2009 13:03:16] <rmatte> scp2004: just ask and we'll help you as best we can
[17-Aug-2009 13:03:49] <scp2004> Thanks .. I m using zenoss 2.4.1 on Freebsd 7.2 following the casey tutorial on Zenoss Forum ... But the daemon zentrap never come up !
[17-Aug-2009 13:04:11] <straterra> Thank jesus for slow zopectl
[17-Aug-2009 13:04:26] <scp2004> I tried all .. Install on Bsd 7.0 , 7.1 .. Installing zenoss 2.4.0 , 2.4.1 but still same
[17-Aug-2009 13:05:09] <scp2004> Matt Ray told me to speak with you
[17-Aug-2009 13:07:16] <scp2004> straterra Thanks .. I m using zenoss 2.4.1 on Freebsd 7.2 following the casey tutorial on Zenoss Forum ... But the daemon zentrap never come up !
[17-Aug-2009 13:07:39] <scp2004> [straterra] I tried all .. Install on Bsd 7.0 , 7.1 .. Installing zenoss 2.4.0 , 2.4.1 but still same
[17-Aug-2009 13:12:02] <scp2004> Somebody help ?
[17-Aug-2009 13:13:38] <straterra> rmatte: do you know what db zenoss uses by default
[17-Aug-2009 13:13:40] <straterra> ?^
[17-Aug-2009 13:14:11] <scp2004> events
[17-Aug-2009 13:14:12] <baffle> scp2004: You have ofcourse checked the logfile?
[17-Aug-2009 13:14:30] <scp2004> yes . but in this case , dont make any log
[17-Aug-2009 13:14:52] <scp2004> when i try start zentrap manually with the command ./zentrap start its the same
[17-Aug-2009 13:15:01] <scp2004> dont come up ..and dont make any log
[17-Aug-2009 13:15:24] <baffle> scp2004: Did you try the debug flags?
[17-Aug-2009 13:15:40] <scp2004> its not problem with same port cause i just checked
[17-Aug-2009 13:16:24] <scp2004> <baffle> are you talking about ./zentrap run or ./zentrap --v10 ..something like this ?
[17-Aug-2009 13:17:31] <scp2004> scp2004 Are you ?
[17-Aug-2009 13:17:37] <scp2004> [scp2004] Are you ?
[17-Aug-2009 13:18:42] <rmatte> straterra: MySQL for events, Zope for everything else
[17-Aug-2009 13:19:05] <straterra> Right..but just the events db?
[17-Aug-2009 13:19:11] <rmatte> correct
[17-Aug-2009 13:19:22] <rmatte> everything else is in Zope
[17-Aug-2009 13:19:34] <scp2004> the name of db is events man
[17-Aug-2009 13:19:55] <scp2004> Somebody knows why my zentrap dont come up in freeBsd ?
[17-Aug-2009 13:24:33] <rmatte> zentrap doesn't start?
[17-Aug-2009 13:28:37] <rmatte> try: zentrap run -v10
[17-Aug-2009 13:28:39] <rmatte> check for errors
[17-Aug-2009 13:28:55] <rmatte> Has anything been announced as to when 2.4.3 is going to be released?
[17-Aug-2009 13:43:36] <scp2004> [rmatte] here the error : Segmentation fault: 11 (core dumped)
[17-Aug-2009 13:47:53] <straterra> Wow..the new monitoring server for zenoss is now...200 megs in to swap
[17-Aug-2009 13:53:40] <scp2004> when i try ./zentrap run --v10 here the erro : Segmentation fault: 11 (core dumped) .. Any idea ?
[17-Aug-2009 14:04:02] <baffle> scp2004: zentrap run -v10 or something, yes.
[17-Aug-2009 14:04:29] <baffle> scp2004: Strange, it's made in Python..
[17-Aug-2009 14:04:43] <baffle> scp2004: It should produce a coredump, then.
[17-Aug-2009 14:12:45] <scp2004> how i debug it in Pyton to test ?
[17-Aug-2009 14:12:58] <scp2004> [baffle] how i debug it in Pyton to test ?
[17-Aug-2009 14:16:27] <baffle> scp2004: To be honest, I have no idea why it crashes.. Maybe there is a problem with one of the bundled modules? Maybe you can use truss or something to see what happens?
[17-Aug-2009 14:22:53] <rmatte> scp2004: more importantly, what did you change recently?
[17-Aug-2009 14:23:03] <rmatte> something must have changed to cause that, things don't generally break on their own
[17-Aug-2009 14:23:50] <scp2004> i dont change anything man .. I just install following the casey tutorial in zenoss forum and it doesnt work !
[17-Aug-2009 14:24:29] <rmatte> ah, you're a BSD user?
[17-Aug-2009 14:25:01] <scp2004> [rmatte] i tried install it again more than 10 times and its the same .. Only zentrap doesnt run
[17-Aug-2009 14:25:01] <scp2004> Yes man .. I m using freeBsd 7.2
[17-Aug-2009 14:25:15] <scp2004> any idea ?
[17-Aug-2009 14:26:08] <rmatte> nope, Zenoss with FreeBSD is quite shaky at the moment
[17-Aug-2009 14:26:15] <rmatte> very hard to support
[17-Aug-2009 14:26:34] <rmatte> just looking at this guide makes my eyes water
[17-Aug-2009 14:26:37] <rmatte> so many tweaks
[17-Aug-2009 14:27:18] <rmatte> you're literally editing code to get it to compile, any one of the changes could cause segfaults
[17-Aug-2009 14:28:27] <rmatte> I'd go through and double/triple check your edits
[17-Aug-2009 14:30:04] <rmatte> also, the guide was written for 2.4.1, it's possible that it won't work properly with 2.4.2 due to changes made between versions
[17-Aug-2009 14:30:06] <rmatte> but no idea
[17-Aug-2009 14:30:52] <scp2004> Im using 2.4.1 ..Ok ..thank you ill try ask to casey
[17-Aug-2009 14:30:53] <scp2004> ty
[17-Aug-2009 14:31:24] <rmatte> Yeh, Casey would be the one to speak with, I'll have to setup a virtual BSD server and try this out sometime so that I'm more helpful with FreeBSD related problems
[17-Aug-2009 14:32:07] <rmatte> I used FreeBSD for 5 years, it's nice and stable but installing apps can be a huge chore
[17-Aug-2009 14:32:13] <rmatte> which is ultimately why I switched to Linux
[17-Aug-2009 14:32:51] <scp2004> [rmatte] Will help me a lot if u do this .. Cause i have no idea that other user that have sucess in run zenoss on FreeBsd 7.2 ..
[17-Aug-2009 14:33:11] <rmatte> there aren't many, email Casey and see if they can help
[17-Aug-2009 14:33:16] <scp2004> If you can .. Try install it following the casey steps .. To verify if we have the same problem
[17-Aug-2009 14:33:23] <rmatte> k
[17-Aug-2009 14:33:32] <scp2004> ty ...bye
[17-Aug-2009 14:37:14] <rmatte> or he could just use Linux for his Zenoss server lol
[17-Aug-2009 14:38:14] <cwj> and violate the purity of his GNU-free network? why the very thought..
[17-Aug-2009 14:38:28] <rmatte> lol
[17-Aug-2009 14:38:40] <rmatte> they use RedHat there as well apparently
[17-Aug-2009 14:38:45] <rmatte> so it's not exactly GNU free
[17-Aug-2009 14:39:05] <cwj> well im not really one to pass up an opportunity for poking fun
[17-Aug-2009 14:39:20] <rmatte> hehe
[17-Aug-2009 15:03:48] <sacpinball> Error collecting performance data: NT_STATUS_NET_WRITE_FAULT
[17-Aug-2009 15:04:02] <sacpinball> WMI on a windows box
[17-Aug-2009 15:13:40] <rmatte> sacpinball: http://forums.zenoss.com/viewtopic.php?p=37591
[17-Aug-2009 15:13:43] <rmatte> that thread may help you
[17-Aug-2009 15:13:52] <sacpinball> thx.. will give it a look
[17-Aug-2009 15:21:46] <rmatte> oh great, I was playing with some report plugins on a server and now it's messed up
[17-Aug-2009 15:21:51] <rmatte> grrr
[17-Aug-2009 15:22:04] <sacpinball> just restore from backup
[17-Aug-2009 15:22:04] <sacpinball> lol
[17-Aug-2009 15:22:40] <rmatte> nah, I need to fix this, it's a production box
[17-Aug-2009 15:24:16] <rmatte> finally
[17-Aug-2009 15:38:09] <rmatte> man, these default reports are horrible
[17-Aug-2009 15:38:23] <rmatte> the CPU report divides values by 100 even though they are already collected as a percentage
[17-Aug-2009 15:39:00] <rmatte> so they end up being like 0.1 instead of 10 percent
[17-Aug-2009 16:10:37] <rmatte> nice, I just found a new feature that I didn't know about lol
[17-Aug-2009 16:11:00] <rmatte> if you go to a device class, then go to More -> All Templates it lists all the local templates
[17-Aug-2009 16:11:15] <rmatte> I never used to be able to keep track of all the local ones I had created
[17-Aug-2009 16:11:56] <cwj> someone in this channel pointed that out to me recently
[17-Aug-2009 16:12:00] <cwj> very very helpful
[17-Aug-2009 16:12:05] <rmatte> yeh, very
[17-Aug-2009 16:12:21] <cwj> now if only i could copy templates up the device tree without using the zope manager interface
[17-Aug-2009 16:12:24] <cwj> maybe someday
[17-Aug-2009 16:12:58] <rmatte> yeh
[17-Aug-2009 16:13:10] <rmatte> zope manager interface can be super handy for copying templates between classes
[17-Aug-2009 16:13:24] <cwj> yep
[17-Aug-2009 16:13:38] <cwj> thats actually the only reason ive needed to use it thus far
[17-Aug-2009 16:13:43] <rmatte> actually, thanks for reminding me about that, I need to use that in a few hehe
[17-Aug-2009 16:34:50] <rmatte> anyone who's good at python...
[17-Aug-2009 16:34:51] <rmatte> memFree = d.getRRDValue('mem5minFree', **summary)
[17-Aug-2009 16:34:52] <rmatte> memUsed = d.getRRDValue('mem5minUsed', **summary)
[17-Aug-2009 16:34:52] <rmatte> percentUsed = memFree / memUsed * 100
[17-Aug-2009 16:35:07] <rmatte> that last line, it's not right since I'm getting an error
[17-Aug-2009 16:35:18] <rmatte> I need memFree divided by memUsed then multiplied by 100
[17-Aug-2009 16:35:33] <rmatte> I also need to cut the number down so that it doesn't show a bunch of decimals
[17-Aug-2009 16:35:36] <rmatte> ideas?
[17-Aug-2009 16:53:06] <rmatte> hmmm, after testing it seems that dividing a smaller number by a larger number is python results in "0"
[17-Aug-2009 16:53:11] <rmatte> any way around this?
[17-Aug-2009 16:58:01] <rmatte> think I got it
[17-Aug-2009 16:58:02] <rmatte> round((1.0*memUsed/memFree)*100)
[17-Aug-2009 17:02:04] <rmatte> this code should be working but it's not
[17-Aug-2009 17:02:04] <rmatte> if percentUsed is None:
[17-Aug-2009 17:02:04] <rmatte> memFree = round(d.getRRDValue('mem5minFree', **summary))
[17-Aug-2009 17:02:04] <rmatte> memUsed = round(d.getRRDValue('mem5minUsed', **summary))
[17-Aug-2009 17:02:05] <rmatte> percentUsed = round((1.0*memUsed/memFree)*100)
[17-Aug-2009 17:02:08] <rmatte> no idea why
[17-Aug-2009 17:09:09] <rmatte> anyways, I'm out of here
[18-Aug-2009 00:00:46] [disconnected at Tue Aug 18 00:00:46 2009]
[18-Aug-2009 00:00:46] [connected at Tue Aug 18 00:00:46 2009]
[18-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[18-Aug-2009 03:12:06] [disconnected at Tue Aug 18 03:12:06 2009]
[19-Aug-2009 00:00:46] [connected at Wed Aug 19 00:00:46 2009]
[19-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[19-Aug-2009 06:28:25] <varbouk> Hello all
[19-Aug-2009 08:18:13] <jb> morning!
[19-Aug-2009 08:23:53] <jb> cgibbons: around?
[19-Aug-2009 08:25:22] <DogWater> can zenoss make systems fire off yum updates, etc?
[19-Aug-2009 08:26:08] <jb> not out of the box..
[19-Aug-2009 08:35:22] <cgibbons> yeah
[19-Aug-2009 08:37:13] <jb> cgibbons: i'm trying to link to a new product name in a skin using productLink like I have seen in several other zenpacks..
[19-Aug-2009 08:37:29] <jb> but it is not working (no data is displayed)
[19-Aug-2009 08:37:35] <jb> can you think of anything that I would be missing?
[19-Aug-2009 08:38:08] <jb> <td class="tablevalues" tal:content="structure frame/getProductLink"/>
[19-Aug-2009 08:38:12] <jb> is what I'm trying to use..
[19-Aug-2009 08:38:21] <jb> is it because maybe the model is not in the database?
[19-Aug-2009 08:39:39] <cgibbons> no clue really
but lemme poke around a bit[19-Aug-2009 08:39:49] <jb> thanks!
[19-Aug-2009 08:47:00] <jb> sup venturaville..
[19-Aug-2009 08:47:07] <venturaville> jb: howdy :-)
[19-Aug-2009 08:47:36] <ckrough> jb: what does productLink do?
[19-Aug-2009 08:48:17] <jb> ckrough: it looks like it links you to related products in the zopedb.. ie: http://fc-zenoss01:8080/zport/dmd/Manufacturers/Unknown/products/DELL%20%20%20%20%20ST373455SS
[19-Aug-2009 08:48:40] <jb> basically, I have a table that displays information about a product, one field being the "Model"
[19-Aug-2009 08:48:51] <jb> i want to be able to click on that model, and show all other products with the same model
[19-Aug-2009 08:50:35] <jb> if that makes any since..
[19-Aug-2009 08:50:37] <jb> sense
[19-Aug-2009 08:50:58] <ckrough> yup
[19-Aug-2009 09:01:19] <Mituc> hi guys!
[19-Aug-2009 09:01:37] <Mituc> for some reason the zencommand process crashes every few days
[19-Aug-2009 09:01:45] <Mituc> I can't find anything into the logs
[19-Aug-2009 09:02:08] <Mituc> and I can't find a list with the configuration directive it supports appart of what it already has set by default (monitor localhost)
[19-Aug-2009 09:02:29] <mrayzenoss> go to Settings->Daemons->Edit Config and you can see the options
[19-Aug-2009 09:02:37] <Mituc> is there any way of adding a debug directive in the configuration file instead of launching it manually with -v10?
[19-Aug-2009 09:02:50] <Mituc> mrayzenoss: that's what I'm looking at..
[19-Aug-2009 09:02:50] <mrayzenoss> what I just said will work
[19-Aug-2009 09:03:09] <Mituc> but I can only see the already existing options there
[19-Aug-2009 09:03:22] <mrayzenoss> ahh, are you using pre-2.4?
[19-Aug-2009 09:03:23] <Mituc> not all the configuration directives that it supports
[19-Aug-2009 09:03:26] <Mituc> yep
[19-Aug-2009 09:03:29] <Mituc> pre 2.4
[19-Aug-2009 09:03:56] <Mituc> 2.3.0 to be more precise
[19-Aug-2009 09:04:07] <Mituc> I'm using the RPM version
[19-Aug-2009 09:04:27] <Mituc> and I'm not sure if upgrading the RPM to a higher version will also do all the migration stuff that it's supposed to
[19-Aug-2009 09:04:58] <Mituc> not that I'm lazy to check, but I didn't have time to do that since I didn't need to touch the current setup
[19-Aug-2009 09:06:10] <Mituc> ok
[19-Aug-2009 09:06:20] <Mituc> so I'll probably have to look into upgrading it
[19-Aug-2009 09:14:26] <rmatte> good morning folks
[19-Aug-2009 09:15:10] <jb> sup
[19-Aug-2009 09:15:15] <mrayzenoss> morning
[19-Aug-2009 09:16:52] <rmatte> I get to setup 2 new Zenoss boxes today, yay
[19-Aug-2009 09:17:30] * ckrough is setting up irc services... longest. config. evar.
[19-Aug-2009 09:17:43] <rmatte> hehe, I haven't setup IRC services in years
[19-Aug-2009 09:17:45] <jb> using what ircd?
[19-Aug-2009 09:17:53] <ckrough> inspircd and anope
[19-Aug-2009 09:17:57] <rmatte> I've done it for ratbox, bahamut, and unrealircd
[19-Aug-2009 09:18:02] <jb> ratbox ftw
[19-Aug-2009 09:18:08] * jb = efnet
[19-Aug-2009 09:18:08] <rmatte> yeh, ratbox is awesome
[19-Aug-2009 09:18:13] <jb> i know the ratbox devs..
[19-Aug-2009 09:18:15] <ckrough> is ratbox good? was going to try it, but inspircd has been working well
[19-Aug-2009 09:18:30] <rmatte> ckrough: ratbox is excellent
[19-Aug-2009 09:18:33] <ckrough> hmm
[19-Aug-2009 09:18:37] <jb> yeah, ratbox is very stable..
[19-Aug-2009 09:19:04] <rmatte> yeh, we've been running it on our IRC network for like 3 or 4 years now, and haven't had any problems with it
[19-Aug-2009 09:19:47] <jb> def needsCleaningString(self):
[19-Aug-2009 09:19:47] <jb> clstatusmap = {1: 'Yes', 2: 'No'}
[19-Aug-2009 09:19:47] <jb> clstatus = self.cacheRRDValue('needsCleaning', 'Unknown')
[19-Aug-2009 09:19:48] <jb> return clstatusmap.get(clstatus, 'Unknown')
[19-Aug-2009 09:19:58] <jb> argh, why is this always returning 'Unknown'
[19-Aug-2009 09:20:05] <jb> even though needsCleaning != 0
[19-Aug-2009 09:20:51] <jb> oh.
[19-Aug-2009 09:20:53] <jb> stupid typo
[19-Aug-2009 09:21:13] <rmatte> ?
[19-Aug-2009 09:21:26] <rmatte> Values or something?
[19-Aug-2009 09:21:30] <jb> my modeler was using "needCleaning," not "needsCleaning"
[19-Aug-2009 09:21:34] <rmatte> ah
[19-Aug-2009 09:21:35] <rmatte> lol
[19-Aug-2009 09:21:41] <jb> hopefully thats it
[19-Aug-2009 09:21:47] <rmatte> I'd imagine it is
[19-Aug-2009 09:23:00] <ckrough> does this just return 1,0,or unknown? self.cacheRRDValue('needsCleaning', 'Unknown')
[19-Aug-2009 09:23:16] <jb> 1, 2, unknown
[19-Aug-2009 09:23:19] <ckrough> 1,2,Unknown I mean
[19-Aug-2009 09:23:21] <jb> yeah
[19-Aug-2009 09:23:31] <jb> im a bit confused as to what the cacheRRDValue is for..
[19-Aug-2009 09:23:36] <jb> but im trying to follow an example
[19-Aug-2009 09:23:54] <rmatte> it probably pulls from cached data
[19-Aug-2009 09:23:59] <rmatte> rather than from the RRD itself
[19-Aug-2009 09:24:02] <rmatte> which speeds it up
[19-Aug-2009 09:24:13] <rmatte> just a guess
[19-Aug-2009 09:24:38] <ckrough> "read an RRDValue with and cache it"
[19-Aug-2009 09:24:44] <ckrough> no sure that helps
[19-Aug-2009 09:24:54] <rmatte> yeh, just saw that hehe
[19-Aug-2009 09:24:59] <jb> i think it has something to do with a threshold
[19-Aug-2009 09:25:00] <rmatte> on huihoo.com
[19-Aug-2009 09:25:29] <rmatte> it reads the RRD value then caches it
[19-Aug-2009 09:25:52] <jb> to be checked against a threshold? .. i think
[19-Aug-2009 09:26:11] <ckrough> it looks like it tries to read value from _cache as well?
[19-Aug-2009 09:26:15] <jb> like, if it needs cleaned (ie, needsCleaningString = 1) trigger a threshold
[19-Aug-2009 09:26:18] <ckrough> and it its not there it creates teh cache
[19-Aug-2009 09:27:00] <rmatte> jb: I highly doubt that function directly relates to thresholds
[19-Aug-2009 09:27:13] <jb> i can't figure out why rrdCache was used then
[19-Aug-2009 09:27:14] <jb> for th is
[19-Aug-2009 09:27:16] <jb> this..
[19-Aug-2009 09:27:21] <rmatte> jb: It reads an RRD value and then commits it to cache if it's not already there, that's all the function does
[19-Aug-2009 09:27:23] <jb> other values/vars do not use it
[19-Aug-2009 09:27:28] <ckrough> it looks like it reads/caches rrd values
[19-Aug-2009 09:27:32] <rmatte> if you want it to apply to a threshold, you need to script it
[19-Aug-2009 09:27:43] <ckrough> what rmatte said.
[19-Aug-2009 09:27:47] <jb> yeah
[19-Aug-2009 09:29:29] <ckrough> the four stages of zenoss admins: "MYEYES!", "Why would they do it that way??", "Detective", "Chet"
[19-Aug-2009 09:29:39] <rmatte> lol
[19-Aug-2009 09:29:54] <rmatte> yeh, Chet does generally have the answers for sure
[19-Aug-2009 09:31:20] <Disconnect> ckrough: you skipped "omg what did i ever do to you?" and "where did my charts go??" (not necessarily in that order)
[19-Aug-2009 09:31:28] <ckrough> lol
[19-Aug-2009 09:31:29] <jb> argh wtf
[19-Aug-2009 09:31:41] <ckrough> "wait, that interface cant push 2.5 petabits per second"
[19-Aug-2009 09:31:57] <mrayzenoss> I'm pretty sure 'Chet' is an unattainable goal for mortals
[19-Aug-2009 09:32:03] <Disconnect> and step one is usually "hey, that looks neat!" otherwise you'd never get to step 2.
[19-Aug-2009 09:32:44] <jb> still returning unknown..
[19-Aug-2009 09:33:29] <rmatte> mrayzenoss: Attaining Chethood seems like a valid challenge, I shall attempt it
[19-Aug-2009 09:34:17] <jb> def getRRDNames(self):
[19-Aug-2009 09:34:17] <jb> """
[19-Aug-2009 09:34:17] <jb> Return the datapoint name of this access device
[19-Aug-2009 09:34:17] <jb> 'needsCleaning_needsCleaning' and 'status_status'
[19-Aug-2009 09:34:17] <jb> """
[19-Aug-2009 09:34:18] <jb> return ['needsCleaning_needsCleaning', 'status_status'
[19-Aug-2009 09:34:22] <jb> hm
[19-Aug-2009 09:34:25] <rmatte> (Apparently the secret is to sell your soul to a penguin, eat an apple, and then feed a mouse to a python)
[19-Aug-2009 09:34:42] <rmatte> but shhhhh
[19-Aug-2009 09:35:14] <rmatte> ummm, do you have to specify the datapoint name and the data source?
[19-Aug-2009 09:35:36] <rmatte> datapoint names would technically just be needsCleaning and status
[19-Aug-2009 09:36:02] <rmatte> but I may be wrong in this case
[19-Aug-2009 09:36:20] <rmatte> are you pulling that format straight for an example?
[19-Aug-2009 09:36:25] <rmatte> from*
[19-Aug-2009 09:36:26] <jb> yeah
[19-Aug-2009 09:36:29] <rmatte> ah, ok
[19-Aug-2009 09:36:33] <rmatte> probably right then hehe
[19-Aug-2009 09:36:41] <jb> still trying to figure out why needsCleaningString is returning unknown though
[19-Aug-2009 09:42:36] <rmatte> did you restart Zenoss after modifying your collector?
[19-Aug-2009 09:43:51] <jb> yeah..
[19-Aug-2009 09:47:13] <rmatte> no idea then, it should be able to pull the value if it's being set properly
[19-Aug-2009 09:48:00] <twm1010> morning
[19-Aug-2009 09:48:26] <rmatte> good morning
[19-Aug-2009 09:51:05] <twm1010> ever get an answer to when 2.4.3 is expected?
[19-Aug-2009 09:52:29] <straterra> rmatte: is adding mibs for a new type of device hard?
[19-Aug-2009 09:52:47] <straterra> I want to create a new type of device for our UPS..and graph its information
[19-Aug-2009 09:54:50] <rmatte> you don't need Mibs to graph information
[19-Aug-2009 09:54:53] <rmatte> just for traps
[19-Aug-2009 09:54:57] <rmatte> will the UPS be sending traps?
[19-Aug-2009 09:55:26] <straterra> No
[19-Aug-2009 09:55:27] <mrayzenoss> twm1010: should be early next week
[19-Aug-2009 09:55:41] <straterra> Why don't I need mibs?
[19-Aug-2009 09:55:42] <twm1010> thanks matt
[19-Aug-2009 09:55:47] <rmatte> straterra: then you don't need Mibs, you just need to find what OIDs correspond to the information that you want to gather
[19-Aug-2009 09:55:54] <straterra> Oh..thats what I man
[19-Aug-2009 09:55:56] <straterra> mean^
[19-Aug-2009 09:55:58] <rmatte> straterra: then create a device template and apply it to that class
[19-Aug-2009 09:55:58] <twm1010> straterra: MIBs are for classifying traps, not gathering peformance data
[19-Aug-2009 09:56:00] <straterra> I have the OID's
[19-Aug-2009 09:56:06] <twm1010> what brand UPS curiously?
[19-Aug-2009 09:56:12] <rmatte> straterra: ok, then just create a new Device template
[19-Aug-2009 09:56:14] <straterra> liebert
[19-Aug-2009 09:56:27] <straterra> I went through and mapped damn near every OID heh
[19-Aug-2009 09:56:32] <twm1010> does it use the standard UPS mib or a custom one?
[19-Aug-2009 09:56:37] <straterra> No idea
[19-Aug-2009 09:57:00] <rmatte> mapped every OID?
[19-Aug-2009 09:57:03] <straterra> yeha
[19-Aug-2009 09:57:05] <straterra> yeah
[19-Aug-2009 09:57:09] <rmatte> like, to a device template?
[19-Aug-2009 09:57:20] <straterra> Did an snmp walk..and matched up the OID information to information about the UPS
[19-Aug-2009 09:57:28] <rmatte> ah, gotcha
[19-Aug-2009 09:57:36] <rmatte> have you not created a device template before?
[19-Aug-2009 09:57:38] <straterra> like charge percentage, time remaining, input lines, output lines, frequency, etc etc
[19-Aug-2009 09:57:41] <straterra> Nope
[19-Aug-2009 09:57:47] <twm1010> there is a standard, RFC1628 mib that shows you the OID's you want
[19-Aug-2009 09:57:51] <twm1010> is is three-phase
[19-Aug-2009 09:57:54] <twm1010> ?
[19-Aug-2009 09:57:58] <straterra> Er..no idea o.O
[19-Aug-2009 09:58:06] <twm1010> how many kVa ?
[19-Aug-2009 09:58:11] <rmatte> straterra: it's very straightforward...
[19-Aug-2009 09:58:20] <straterra> twm1010: I dunno
I'll have to look it up[19-Aug-2009 09:58:22] <rmatte> create a new device organizer, or use an existing one
[19-Aug-2009 09:58:23] <straterra> We just got it
[19-Aug-2009 09:58:25] <rmatte> go in to the organizer
[19-Aug-2009 09:58:27] <twm1010> k
[19-Aug-2009 09:58:44] <twm1010> i am monitoring tripp-lites, APC's, and chlorites
[19-Aug-2009 09:58:46] <rmatte> /Power/UPS will probably be what you want
[19-Aug-2009 09:58:59] <twm1010> the tripp-lites diverted from the RFC standard OID's, the others were pretty close
[19-Aug-2009 09:59:02] <rmatte> then maybe add a sub-organizer for liebert
[19-Aug-2009 09:59:14] <rmatte> then go to the organizer where you plan to put the devices
[19-Aug-2009 09:59:19] <rmatte> and click on the "Templates" tab
[19-Aug-2009 09:59:33] <twm1010> i'm monitoring input/output voltage, output load, and some environment sensors
[19-Aug-2009 10:00:15] <rmatte> you can either copy the "Device" template down from /Devices or create a new template called whatever
[19-Aug-2009 10:00:34] <straterra> Then I just go about adding the OIDs, right?
[19-Aug-2009 10:00:36] <rmatte> then click on the dropdown arrow and select Bind Templates
[19-Aug-2009 10:00:42] <rmatte> highlight the template you created and click ok
[19-Aug-2009 10:00:46] <rmatte> then click on the template name
[19-Aug-2009 10:00:59] <rmatte> add snmp type data sources
[19-Aug-2009 10:01:05] <rmatte> it's pretty much self explanatory from that point
[19-Aug-2009 10:01:30] <rmatte> you can then create thresholds and graphs from the data points
[19-Aug-2009 10:01:37] <straterra> So like..if a line from snmpwalk is SNMPv2-SMI::mib-2.33.1.1.1.0 = STRING: "Blablabla", 2.33.1.1.1.0 is the full OID, right?
[19-Aug-2009 10:01:43] <rmatte> no
[19-Aug-2009 10:01:48] <rmatte> that's not the full OID
[19-Aug-2009 10:01:50] <twm1010> that's the RFC standard MIB.
[19-Aug-2009 10:01:51] <straterra> no?
[19-Aug-2009 10:02:06] <rmatte> the "SNMPv2-SMI::mib-" part is masking part of it
[19-Aug-2009 10:02:15] <rmatte> since your standard RFC Mibs are translating that part
[19-Aug-2009 10:02:28] <rmatte> there's some flag for snmpwalk to keep it from translating the OIDs
[19-Aug-2009 10:02:56] <twm1010> 1.3.6.1.2.1.33.1.3.3.1.3.1 is input voltage
[19-Aug-2009 10:03:19] <straterra> Yeah
[19-Aug-2009 10:03:20] <twm1010> (for single phase)
[19-Aug-2009 10:03:30] <straterra> Thats close to what I have written down
[19-Aug-2009 10:04:08] <straterra> I have mib-2.33.1.3.3.1.3.1 for volts input
[19-Aug-2009 10:05:12] <rmatte> you need to add -0n to your snmpwalk options
[19-Aug-2009 10:05:18] <rmatte> that'll display OIDs numerically
[19-Aug-2009 10:05:29] <straterra> aha
[19-Aug-2009 10:05:57] <straterra> twm1010: So..you're telling me that these OID's are standard?
[19-Aug-2009 10:06:31] <rmatte> no, just the first part of them
[19-Aug-2009 10:06:54] <straterra> well..when I did that -On switch, I get this for the input volts..
[19-Aug-2009 10:07:01] <rmatte> the first part specifies the version of SNMP, and some other standard stuff
[19-Aug-2009 10:07:06] <rmatte> the second part is the non-standard
[19-Aug-2009 10:07:16] <straterra> .1.3.6.1.2.1.33.1.3.3.1.3.1 = INTEGER: 246
[19-Aug-2009 10:07:29] <twm1010> that would be 246volts
[19-Aug-2009 10:07:32] <straterra> Right
[19-Aug-2009 10:07:33] <rmatte> yeh, that's the raw data for the input volts
[19-Aug-2009 10:07:50] <straterra> But my question is..if there OID are standard for UPS's..is there an existing template/zenpack?
[19-Aug-2009 10:08:13] <rmatte> I doubt they use the same OIDs between different companies
[19-Aug-2009 10:08:20] <twm1010> No, I don't think so, some manufacturers diverge from these, some don't, also depends on if it's single-phase or three-phase
[19-Aug-2009 10:08:28] <straterra> Ok
[19-Aug-2009 10:08:33] <rmatte> you're better off to taylor it to your needs
[19-Aug-2009 10:09:02] <twm1010> For our ginormous UPS at HQ I use manufacturer specific stuff
[19-Aug-2009 10:09:11] <rmatte> Zenoss is great, but be prepared to spend some development time hehe
[19-Aug-2009 10:09:35] <straterra> That's fine
[19-Aug-2009 10:10:01] <twm1010> rmatte: Yes it is, if only I could get some enthusiasm from my team. You'd think a group of Linux geeks/scripters would be gung-ho, but they're not, they like their Hobbit monitor..
[19-Aug-2009 10:10:03] <straterra> So..I go to add a data source..say I want to add the input volts..
[19-Aug-2009 10:10:33] <straterra> I'd put in .1.3.6.1.2.1.33.1.3.3.1.3.1 and select snmp as the type?
[19-Aug-2009 10:10:45] <twm1010> no, give it a logical name, like input volts
[19-Aug-2009 10:10:53] <straterra> Oh..doh
[19-Aug-2009 10:11:00] <twm1010> i do it ala brand.... so tl_inputvolts for tripp-lite input volts
[19-Aug-2009 10:11:09] <straterra> then .1.3.6.1.2.1.33.1.3.3.1.3.1 as the OID
[19-Aug-2009 10:11:18] <twm1010> then on the next screen you put in the OID.
[19-Aug-2009 10:11:22] <rmatte> I'm pretty sure Zenoss can easily trump that hobbit system
[19-Aug-2009 10:11:54] <rmatte> try to keep spaces out of the datapoint names for good measure
[19-Aug-2009 10:12:02] <rmatte> call it inputVolts or something
[19-Aug-2009 10:12:03] <twm1010> rmatte: oh yes, it can woop it three ways from sunday, but they hate change... it took me a WEEK to get their apache mod_status URL out of them just so i could demonstrate that the Apache Zenpack was every bit if not more powerful.
[19-Aug-2009 10:12:20] <straterra> Gauge type will be ok for this kind of stuff, right?
[19-Aug-2009 10:12:21] <rmatte> twm1010: haha, nice
[19-Aug-2009 10:12:34] <twm1010> Right... you're wanting the values, not rate of change or anything like that
[19-Aug-2009 10:12:40] <rmatte> yes, gauge is usually what you'll be using
[19-Aug-2009 10:12:47] <rmatte> the other types are for very special circumstances
[19-Aug-2009 10:12:51] <rmatte> I've yet to use any of them
[19-Aug-2009 10:14:12] <rmatte> eugh, I'm waiting for an 11gig transfer to finish between our lab nfs share and our SAN in a remote data center
[19-Aug-2009 10:14:17] <rmatte> it's taking forevar
[19-Aug-2009 10:15:20] <straterra> So like.. there is a string for the name and model and stuff too..where do I put that?
[19-Aug-2009 10:15:38] <rmatte> where are you seeing this?
[19-Aug-2009 10:15:47] <rmatte> oh, an snmp string you mean?
[19-Aug-2009 10:15:48] <twm1010> well, it sounds like you did a walk of the whole device
[19-Aug-2009 10:15:59] <rmatte> that would go under "Products" section, not in the template
[19-Aug-2009 10:16:08] <straterra> Yeah
[19-Aug-2009 10:16:10] <straterra> twm1010: I did
[19-Aug-2009 10:16:13] <rmatte> I wouldn't load every single OID from the walk in to the template
[19-Aug-2009 10:16:18] <straterra> Right
[19-Aug-2009 10:16:19] <rmatte> only load stuff you know you'll want to monitor
[19-Aug-2009 10:17:08] <straterra> lunch time..ill inquire about the products section when I get back
[19-Aug-2009 10:17:36] <twm1010> im leaving shortly myself.
[19-Aug-2009 10:17:42] <rmatte> the products section is pretty straightforward
[19-Aug-2009 10:18:47] <DogWater> Where do you specify the username/password Zenoss uses to get the event log on a windows host?
[19-Aug-2009 10:19:01] <rmatte> zProperties, zWinUser zWinPass
[19-Aug-2009 10:19:16] <DogWater> righto
[19-Aug-2009 10:19:20] <rmatte>
[19-Aug-2009 10:20:00] <rmatte> If it's a local admin account you may have to put ./ in front of the username
[19-Aug-2009 10:20:14] <rmatte> if it's a domain account you need to put the domain then a slash either way, then the username
[19-Aug-2009 10:20:19] <twm1010> DogWater: if you have a common account, do it as the zProperties of the class, and not the device
[19-Aug-2009 10:20:28] <rmatte> yeh
[19-Aug-2009 10:20:32] <rmatte> good point
[19-Aug-2009 10:20:42] <DogWater> Is it a good idea to make a specific domain user for Zenoss?
[19-Aug-2009 10:20:51] <DogWater> or do you need a domain admin account?
[19-Aug-2009 10:21:08] <rmatte> the account needs to be full admin, so if it's a domain account it has to be domain admin
[19-Aug-2009 10:21:20] * ckrough shudders
[19-Aug-2009 10:21:23] <rmatte> well, it needs to have admin on all boxes
[19-Aug-2009 10:21:31] <rmatte> not necessarily full admin access to AD
[19-Aug-2009 10:21:52] <rmatte> so it doesn't have to be a "true" domain admin account
[19-Aug-2009 10:22:26] <DogWater> So if I create a user called Zenoss and make it a local admin on the windows boxes that would work?
[19-Aug-2009 10:22:37] <rmatte> just make sure that you limit ssh access to the Zenoss box since anyone with root access will be able to dump that username and password if they know how
[19-Aug-2009 10:22:45] <rmatte> DogWater: correct
[19-Aug-2009 10:22:57] <DogWater> it doesn't store it in any sort of hash/encryption?
[19-Aug-2009 10:23:10] <rmatte> nope, I've complained about that before
[19-Aug-2009 10:23:15] <DogWater> neat-o
[19-Aug-2009 10:23:16] <DogWater> ...
[19-Aug-2009 10:24:11] <twm1010> Yes, and if one was to gain access to zenoss and create a simple command using zWinUser and zWinPassword and monitor the output, there they are.
[19-Aug-2009 10:24:16] <rmatte> dumping the password is as simple as becoming Zenoss user, going in to zendmd, and doing something like...
[19-Aug-2009 10:24:23] <rmatte> >>> mydev = find("172.30.30.8")
[19-Aug-2009 10:24:23] <rmatte> >>> mydev.zWinPassword
[19-Aug-2009 10:24:27] <rmatte> and it'll spit it out
[19-Aug-2009 10:24:28] <twm1010> or does it *** them now?
[19-Aug-2009 10:26:01] <rmatte> You can also open $ZENHOME/var/Data.fs and search for "Password" and you'll find it
[19-Aug-2009 10:26:18] <rmatte> takes a while to open the file since it's huge
[19-Aug-2009 10:27:14] <rmatte> but either way, just keep the server as secure as possible and you should be fine
[19-Aug-2009 10:27:23] <rmatte> they can't really do anything via the web interface
[19-Aug-2009 10:31:59] <cgibbons> You can configure a user that doesn't have admin rights for remote monitoring, but it's rather involved.
[19-Aug-2009 10:32:17] <cgibbons> If you're running Server 2008 Microsoft has created a new local group that has the right access rights to make it easier, too.
[19-Aug-2009 10:32:39] x-ip is now known as xip|comiendo
[19-Aug-2009 10:32:58] <rmatte> cgibbons: We've done that before and still ran in to problems
[19-Aug-2009 10:33:16] <cgibbons> it's easy to miss stuff, very non-trivial
[19-Aug-2009 10:33:45] <rmatte> well, we had an MCSE try it, and he couldn't get it to work 100%
[19-Aug-2009 10:33:52] <rmatte> lol
[19-Aug-2009 10:35:10] <cgibbons> we've got it documented in the knowledge base but there's still a lot of edge cases, especially if you're trying to monitor a domain controller isself
[19-Aug-2009 10:35:53] <cgibbons> I never went through setting it up via domain group policy, either, which is really the right solution
[19-Aug-2009 10:36:05] <rmatte> yeh
[19-Aug-2009 10:40:06] <rmatte> hmmm, I'm getting pretty frequent heartbeat failures on one of our zenoss boxes, but it doesn't seem to affect collection
[19-Aug-2009 10:45:23] <DogWater> is there some manual way to test from the Zenoss box to see if the domain credentials provided should work? Its kind of odd, I assigned a winuser/password to the class, and 3 out of 4 of the machines work but one doesnt.
[19-Aug-2009 10:47:03] <rmatte> I used to use a perl script that came with one of James Dastrup's zenpacks for testing, but those ZenPacks are obsolete now
[19-Aug-2009 10:47:29] <rmatte> If it's not working and has the exact same credentials as the other servers, then the problem is on the server side
[19-Aug-2009 10:49:24] <rmatte> cgibbons: Any news on when 2.4.3 is due for release?
[19-Aug-2009 10:49:33] <cgibbons> real soon now
[19-Aug-2009 10:49:39] <jb> yawn
[19-Aug-2009 10:49:40] <mrayzenoss> rmatte: tentatively scheduled for next week
[19-Aug-2009 10:50:12] <rmatte> ok, I have an upgrade change on Tuesday, so if it's out by then I'll be ecstatic, if not, then I'll hold off on the update until the week after
[19-Aug-2009 10:50:28] <cgibbons> dogwater: what I do to test connectivity for Windows is this: wmic -U 'domain\account' //host 'select name from win32_computersystem'
[19-Aug-2009 10:51:08] xip|comiendo is now known as x-ip
[19-Aug-2009 10:51:37] <DogWater> hm, that gives invalid global switch ;D
[19-Aug-2009 10:53:06] <cgibbons> from the zenoss server? got the right case & quotes and everything?
[19-Aug-2009 10:53:47] <DogWater> Zenoss is running on linux in my environment, i'm trying to do that from the box itself that isn't working just to see what it says.
[19-Aug-2009 10:53:55] <jb> can you guys reccomend any python beginner books?
[19-Aug-2009 10:54:22] <rmatte> jb: you might have to do something like $ZENHOME/bin/wmic
[19-Aug-2009 10:54:33] <jb>
[19-Aug-2009 10:55:06] <rmatte> sorry...
[19-Aug-2009 10:55:27] <rmatte> /usr/local/zenoss/common/bin/wmic
[19-Aug-2009 10:55:32] <rmatte> that's where mine is
[19-Aug-2009 10:56:23] <DogWater> is it supposed to return a prompt for a password or do I need to specify that on the cmmd line?
[19-Aug-2009 10:56:58] <rmatte> specify it as cgibbons stated
[19-Aug-2009 10:57:07] <rmatte> wmic -U 'domain\account' //host 'select name from win32_computersystem'
[19-Aug-2009 10:57:25] <rmatte> I think you need to put a : after account and put the password
[19-Aug-2009 10:57:31] <cgibbons> it'll prompt you for the password
[19-Aug-2009 10:57:38] <rmatte> ah
[19-Aug-2009 10:57:40] <DogWater> ah, it just says: NTSTATUS: NT_STATUS_ACCESS_DENIED - Access denied
[19-Aug-2009 10:57:42] <cgibbons> if you want to include it in the command line it would be 'domain\user%password'
[19-Aug-2009 10:57:56] <cgibbons> then there you go, bad username, bad password, or account locked out
[19-Aug-2009 10:58:07] <jb> TypeError: 'int' object is not callable
[19-Aug-2009 10:58:07] <DogWater> well i'm using the domain administrator to test..
[19-Aug-2009 10:58:07] <jb> hm
[19-Aug-2009 10:58:11] <DogWater> ;-)
[19-Aug-2009 10:58:28] <rmatte> then something is wrong on that server's config
[19-Aug-2009 10:58:35] <mrayzenoss> jb: I like http://oreilly.com/catalog/9780596513986/
[19-Aug-2009 10:58:49] <jb> thats the one I had my eye on..
[19-Aug-2009 10:59:06] <rmatte> jb: that's the book sitting right next to me right now, it's quite good
[19-Aug-2009 10:59:12] <jb> ok nice
[19-Aug-2009 10:59:18] <jb> i'm going to pick it up..
[19-Aug-2009 10:59:36] <mrayzenoss> next version comes out Oct. 2
[19-Aug-2009 10:59:53] <jb> hmm
[19-Aug-2009 11:00:07] <rmatte> just buy the third edition jb, I doubt there'll be a huge difference
[19-Aug-2009 11:00:08] <rmatte> hehe
[19-Aug-2009 11:00:20] <jb> probably not
[19-Aug-2009 11:00:22] <jb> python 3 maybe?
[19-Aug-2009 11:00:30] <mrayzenoss> Zenoss is supposed to move to Python 2.6 in the April release, which will still be the version covered by the 3rd edition
[19-Aug-2009 11:00:39] <jb> right
[19-Aug-2009 11:01:38] <DogWater> yeah, something goofy with this machine.
[19-Aug-2009 11:04:04] <straterra> Ok..back from lunch
[19-Aug-2009 11:04:45] <straterra> rmatte: So..where would I add the OID for like..device name and such?
[19-Aug-2009 11:06:41] <DogWater> hrmmm.
[19-Aug-2009 11:07:26] <cgibbons> can you remote desktop into that machine with those credentials, dogwater?
[19-Aug-2009 11:09:16] <rmatte> ok, so if I'm getting single heartbeat failures every 15 minutes or so, any thoughts on what I could adjust to make this stop?
[19-Aug-2009 11:09:25] <rmatte> I already have 3 zenhub workers on that box
[19-Aug-2009 11:10:05] <rmatte> oh wait, I only have 2, I'll try bumping that up during our next change window
[19-Aug-2009 11:11:37] <straterra> I want to like..fill out the OS information for this new template/device
[19-Aug-2009 11:12:44] <rmatte> If you want it populated automatically when modelled you need to add entries in the Products section
[19-Aug-2009 11:13:00] <rmatte> If you want to manually select it each time, you still need to add entries in to the Products section
[19-Aug-2009 11:13:04] <straterra> Erhm..wheres that?
[19-Aug-2009 11:13:15] <rmatte> right in the left hand meny
[19-Aug-2009 11:13:16] <rmatte> menu*
[19-Aug-2009 11:13:24] <straterra> Oh..duh
[19-Aug-2009 11:13:28] <rmatte>
[19-Aug-2009 11:13:42] <rmatte> lol
[19-Aug-2009 11:14:25] <straterra> So, I add liebert, then an actual product
[19-Aug-2009 11:15:21] <straterra> So, what do I add to make it pick that up upon modeling?
[19-Aug-2009 11:16:14] <DogWater> cgibbons: yep
[19-Aug-2009 11:16:27] <DogWater> cgibbons: I also checked the dcom settings
[19-Aug-2009 11:20:29] <DogWater> cgibbons: I can also connect to the remote host with compmgmt.msc via my vista desktop
[19-Aug-2009 11:20:36] <DogWater> using my account
[19-Aug-2009 11:20:37] <cgibbons> interesting
[19-Aug-2009 11:20:59] <cgibbons> can you add -d99 argument to the start of that wmic command-line ? that'll dump the protocol communications and we can see where it's actually failing and it might provide more clue
[19-Aug-2009 11:22:07] <DogWater> rpc fault: DCERPC_FAULT_ACCESS_DENIED
[19-Aug-2009 11:22:31] <cgibbons> can you paste the whole output to pastebin or e-mail for me?
[19-Aug-2009 11:22:40] <straterra> I'm looking at the product stuff for IBM..and thats very confusing
[19-Aug-2009 11:22:56] <DogWater> sure im me your email
[19-Aug-2009 11:22:59] <straterra> I put the OID of the manufacturer as the hadward name..and the product key?
[19-Aug-2009 11:25:37] <straterra> Is there a guide on this?
[19-Aug-2009 11:26:11] <rmatte> straterra: you need to put the appropriate OIDs for that model
[19-Aug-2009 11:26:13] <cgibbons> that's the earliest place in the protocol communication that you can get an error, so we're likely looking at DCOM settings. you said you checked that though?
[19-Aug-2009 11:26:35] <rmatte> straterra: there are guides to that and other things in the Zenoss Admin guide
[19-Aug-2009 11:26:40] <straterra> kk
[19-Aug-2009 11:26:43] <rmatte> I'd really recommend that you read through it once
[19-Aug-2009 11:27:03] <rmatte> You'll learn a lot from it
[19-Aug-2009 11:27:24] <straterra> I've glanced over it..but not a full read
[19-Aug-2009 11:27:41] <straterra> Erhm..I'm not finding the section I'm looking for..
[19-Aug-2009 11:27:47] * straterra goes to take another look
[19-Aug-2009 11:28:24] <DogWater> cgibbons: I just made sure the dcom settings on the broken one matched a working one.
[19-Aug-2009 11:29:01] <cgibbons> another thing you can try is to use wbemtest from another Windows box in that domain to the one that's not working; i'd expect that to fail, too.
[19-Aug-2009 11:29:29] <cgibbons> what kind of box is the one that isn't working?
[19-Aug-2009 11:29:40] <straterra> rmatte: do you know what section, per chance?
[19-Aug-2009 11:29:45] <DogWater> 2k3 x64 domain controller
[19-Aug-2009 11:30:41] <rmatte> straterra: not sure, ctrl-f
[19-Aug-2009 11:30:46] <straterra> I did that
[19-Aug-2009 11:30:59] <cgibbons> ooh, it's a domain controller, are any of the other working boxes DCs, too?
[19-Aug-2009 11:31:01] <straterra> I see nothing about it in the guide
[19-Aug-2009 11:31:12] <DogWater> yeah, one of them is, its 2k8 though
[19-Aug-2009 11:31:41] <cgibbons> okay, DCs throw some more complexity into the mix (more rules). the account is a domain admin too?
[19-Aug-2009 11:31:51] <DogWater> yeah, domain\administrator
[19-Aug-2009 11:31:54] <rmatte> straterra: just go to the Products section and look at some of the products already there
[19-Aug-2009 11:32:00] <straterra> I did that
[19-Aug-2009 11:32:00] <rmatte> straterra: check the Cisco one
[19-Aug-2009 11:32:07] <rmatte> straterra: it's not overcomplicated
[19-Aug-2009 11:32:35] <rmatte> you said you had a product key that corresponds to the model
[19-Aug-2009 11:32:40] <straterra> Ok..so..the Product Key is the OID that holds the string, right?
[19-Aug-2009 11:32:52] <rmatte> so create an organize for lietberg and then add the model number under that container with the associated product OID
[19-Aug-2009 11:32:55] <rmatte> and then you're done
[19-Aug-2009 11:33:16] <rmatte> I'm not sure what the OID would be, you'll have to hunt it down
[19-Aug-2009 11:33:23] <cgibbons> A domain admin should 'just work' unless the server has been locked down before. Hrm. Can you double-check these settings, too? http://technet.microsoft.com/en-us/library/cc787533(WS.10).aspx
[19-Aug-2009 11:33:24] <rmatte> google for it if need be
[19-Aug-2009 11:33:28] <straterra> I have the OID
[19-Aug-2009 11:33:45] <rmatte> then just add it in with the corresponding model name and you're done
[19-Aug-2009 11:33:50] <straterra> .1.3.6.1.2.1.33.1.1.1.0 = STRING: "Liebert Corporation", .1.3.6.1.2.1.33.1.1.2.0 = STRING: "Nfinity Power System", .1.3.6.1.2.1.33.1.1.3.0 = STRING: "V5007", .1.3.6.1.2.1.33.1.1.4.0 = STRING: "3.000.0"
[19-Aug-2009 11:33:50] <DogWater> i'm thinking i may have run that security wizard thing on the server in the past, its an old box
[19-Aug-2009 11:34:02] <rmatte> v5007 is what you're looking for
[19-Aug-2009 11:34:07] <cgibbons> probably
[19-Aug-2009 11:34:32] <straterra> So.. I put in that OID and the string V5007
[19-Aug-2009 11:34:38] <rmatte> add a lietbert container, then under that container add V5007 with .1.3.6.1.2.1.33.1.1.3.0
[19-Aug-2009 11:34:51] <straterra> Right
[19-Aug-2009 11:35:00] <rmatte> you name the object V5007 and specify that OID for it
[19-Aug-2009 11:35:03] <rmatte> and then you're done
[19-Aug-2009 11:35:26] <rmatte> when you model and Zenoss finds that OID it should automatically assign the make as Lietbert and the model as V5007
[19-Aug-2009 11:35:38] <rmatte> If you have other models you can add the OIDs for those too
[19-Aug-2009 11:35:51] <rmatte> if it's doesn't automatically assign the model info, you can just do it manually from the edit tab for the device
[19-Aug-2009 11:36:00] <straterra> Alright..let me give it a shot
[19-Aug-2009 11:36:01] <straterra> Thanks
[19-Aug-2009 11:36:14] <rmatte> no prob
[19-Aug-2009 11:38:13] <DogWater> cgibbons: the 'root' namespace has the same security configuration as the other 3 machines
[19-Aug-2009 11:38:17] <straterra> Didn't model it correctly..but oh well I guess
[19-Aug-2009 11:38:34] <straterra> it shows the hw make as unknown and hw model as
[19-Aug-2009 11:38:34] <straterra> it shows the hw make as unknown and hw model as .1.3.6.1.4.1.476.1.42
[19-Aug-2009 11:38:41] <straterra> Stupid irssi..sorry
[19-Aug-2009 11:39:14] <rmatte> I guess .1.3.6.1.4.1.476.1.42 is the string it's looking for then
[19-Aug-2009 11:39:19] <rmatte> add that in to products instead
[19-Aug-2009 11:39:31] <rmatte> also, remove the .1.3.6.1.4.1.476.1.42 product that it will have automatically added
[19-Aug-2009 11:40:04] <rmatte> it'll be under Unknown
[19-Aug-2009 11:40:35] <straterra> It got that value from .1.3.6.1.2.1.1.2.0 for some reason
[19-Aug-2009 11:40:37] <straterra> Oh well
[19-Aug-2009 11:40:41] <rmatte> and if that doesn't work then just manually assign it
[19-Aug-2009 11:40:53] <rmatte> what version of Zenoss are you using?
[19-Aug-2009 11:41:06] <jb> cgibbons: were you ever able to hunt anything down about getProductLink?
[19-Aug-2009 11:41:19] <straterra> Let me check
[19-Aug-2009 11:41:32] <straterra> 2.4.2
[19-Aug-2009 11:41:44] <rmatte> k, it should be able to auto-assign make/model then
[19-Aug-2009 11:41:54] <rmatte> there was a bug where it wouldn't do it properly, but that was in older versions
[19-Aug-2009 11:42:57] <straterra> Now to wait and see if my input voltage gets graphed properly
[19-Aug-2009 11:43:11] <rmatte> it should be graphing by now
[19-Aug-2009 11:43:14] <cgibbons> nope sorry
[19-Aug-2009 11:43:17] <rmatte> it's been well over 3 collection cycles
[19-Aug-2009 11:43:24] <jb> k
[19-Aug-2009 11:43:34] <straterra> I have nan in the graph
[19-Aug-2009 11:43:50] <rmatte> when did you finish the template?
[19-Aug-2009 11:44:03] <straterra> before I left for lunch
[19-Aug-2009 11:44:10] <straterra> I just added the device a couple minutes ago
[19-Aug-2009 11:44:13] <rmatte> then if it's not graphing by now there's a problem
[19-Aug-2009 11:44:21] <rmatte> click on the data source and run a test on a UPS device
[19-Aug-2009 11:44:29] <rmatte> input the UPS device name in and click test
[19-Aug-2009 11:44:34] <rmatte> see what you get
[19-Aug-2009 11:44:58] <otakup0p1> is it possible to manually edit the software for a device
[19-Aug-2009 11:45:05] <otakup0p1> or is that only only only for windows wmi voodoo
[19-Aug-2009 11:45:25] <rmatte> otakup0p1: the software list in Zenoss you mean?
[19-Aug-2009 11:45:30] <otakup0p1> i think
[19-Aug-2009 11:45:30] <otakup0p1> ack
[19-Aug-2009 11:45:33] otakup0p1 is now known as otakup0pe
[19-Aug-2009 11:45:38] <twm1010> i thought the software list came from SNMP actually
[19-Aug-2009 11:45:39] <straterra> SNMPv2-SMI::mib-2.33.1.3.3.1.3.1 = INTEGER: 245
[19-Aug-2009 11:45:44] <otakup0pe> twm1010: oh !
[19-Aug-2009 11:45:51] <straterra> Thats the result of the test..so..it works
[19-Aug-2009 11:45:56] <rmatte> straterra: hmmm, give it a bit more time then, but it really should be working by now
[19-Aug-2009 11:46:44] <straterra> Any way to force collection?
[19-Aug-2009 11:47:04] <rmatte> from commandline as zenoss user: zenperfsnmp run --cycle
[19-Aug-2009 11:47:46] <rmatte> actually
[19-Aug-2009 11:47:48] <straterra> So..if I run that a few times real fast..
[19-Aug-2009 11:47:51] <rmatte> from commandline as zenoss user: zenperfsnmp run -d <devicename>
[19-Aug-2009 11:47:53] <rmatte> try that
[19-Aug-2009 11:48:02] <straterra> i already ran the other one lol
[19-Aug-2009 11:48:09] <straterra> I'll wait for it to finish
[19-Aug-2009 11:48:19] <rmatte> well ctrl-c out of it, the cycle one just keeps cycling
[19-Aug-2009 11:48:22] <straterra> graph is populated now \O/
[19-Aug-2009 11:48:28] <rmatte> good
[19-Aug-2009 11:48:35] <straterra> Kickass..thanks
[19-Aug-2009 11:48:36] <rmatte> then ctrl-c out of it and leave it be
[19-Aug-2009 11:48:45] <straterra> yup
[19-Aug-2009 11:48:46] <rmatte> do a zenoss status to make sure all your Zenoss daemons are running
[19-Aug-2009 11:48:52] <rmatte> "zenoss status"
[19-Aug-2009 11:49:02] <straterra> yup
[19-Aug-2009 11:49:22] <rmatte> good
[19-Aug-2009 11:49:36] <rmatte> congrats, you've just created your first device template
[19-Aug-2009 11:50:33] <rmatte> it usually takes 3 polling cycles, so about 15-20 minutes for a graph to populate with data
[19-Aug-2009 11:50:41] <straterra> Gotcha
[19-Aug-2009 11:50:52] <straterra> Now to add all of these data points and graphs
[19-Aug-2009 11:51:42] <rmatte> hehe
[19-Aug-2009 11:51:50] <straterra> and maybe save it as a zenpack when I'm done
[19-Aug-2009 11:51:57] <rmatte> yeh, you could do that
[19-Aug-2009 11:52:14] <rmatte> hmmm, time for me to grab something to eat, afk for a few
[19-Aug-2009 11:52:18] <straterra> Is there anyway to do some simple math on the return stuff?
[19-Aug-2009 11:52:21] <straterra> Enjoy
[19-Aug-2009 11:52:38] <straterra> I need to divide/multiply the output by 10
[19-Aug-2009 11:56:39] <jb> you guys know if there is a routine to set "OS Make" like there is for OS Versino (setOSProductKey)?
[19-Aug-2009 11:59:27] <rmatte> jb, I believe the Make gets set to the container name that the version info is in in Products
[19-Aug-2009 11:59:35] <rmatte> same as product make/model
[19-Aug-2009 12:00:11] <rmatte> so you'd make a container called Microsoft for instance and then put a version item in it with the correct OID and name of the version
[19-Aug-2009 12:00:44] <jb> ah ha
[19-Aug-2009 12:00:54] <jb> thanks
[19-Aug-2009 12:00:57] <rmatte> np
[19-Aug-2009 12:02:46] <scp2004> Hi Guys ! .. Can someone help me with a question ?
[19-Aug-2009 12:02:57] <rmatte> scp2004: just ask and we'll help as best we can
[19-Aug-2009 12:03:13] <rmatte> (asking to ask is kind of counter-productive)
[19-Aug-2009 12:03:46] <scp2004> [rmatte] Thanks.
.. I need a report in zenoss that give me the Name of device and Ip ...[19-Aug-2009 12:03:57] <rmatte> that's it?
[19-Aug-2009 12:04:01] <rmatte> just name and IP?
[19-Aug-2009 12:04:09] <scp2004> yep
[19-Aug-2009 12:04:14] <rmatte> Easy...
[19-Aug-2009 12:04:16] <rmatte> click on Reports
[19-Aug-2009 12:04:21] <rmatte> then click on Custom Device Reports
[19-Aug-2009 12:04:35] <rmatte> go to the dropdown and select Add Device Report
[19-Aug-2009 12:04:43] <rmatte> name it something like "List of Devices" or whatever
[19-Aug-2009 12:04:45] <rmatte> click on it
[19-Aug-2009 12:04:56] <rmatte> click the Edit tab
[19-Aug-2009 12:05:06] <scp2004> done
[19-Aug-2009 12:05:10] <scp2004> and..
[19-Aug-2009 12:05:20] <rmatte> Under Sort Column put: getId
[19-Aug-2009 12:05:26] <rmatte> Under Columns put:
[19-Aug-2009 12:05:27] <rmatte> getId
[19-Aug-2009 12:05:27] <rmatte> getManageIp
[19-Aug-2009 12:05:40] <rmatte> Under Column Names put:
[19-Aug-2009 12:05:41] <rmatte> Device
[19-Aug-2009 12:05:41] <rmatte> IP Address
[19-Aug-2009 12:05:49] <rmatte> Sort Sense should be: asc
[19-Aug-2009 12:05:55] <rmatte> Path should be /
[19-Aug-2009 12:05:58] <rmatte> then click Save
[19-Aug-2009 12:06:00] <rmatte> and you're done
[19-Aug-2009 12:06:57] <rmatte> then just click the Report tab and it will generate the report
[19-Aug-2009 12:07:07] <scp2004> doing ..
[19-Aug-2009 12:07:36] <scp2004> [rmatte] Ooooow ..Very Cool ...Thank you man ..
[19-Aug-2009 12:07:40] <rmatte> no problem
[19-Aug-2009 12:08:13] <rmatte> there are other things you can add to columns too, like:
[19-Aug-2009 12:08:14] <rmatte> getHWManufacturerName
[19-Aug-2009 12:08:14] <rmatte> getHWProductName
[19-Aug-2009 12:08:14] <rmatte> getHWSerialNumber
[19-Aug-2009 12:08:14] <rmatte> getPingStatusString
[19-Aug-2009 12:08:14] <rmatte> getSnmpStatusString
[19-Aug-2009 12:08:18] <rmatte> getPriorityString
[19-Aug-2009 12:08:26] <twm1010> rmatte: are those documented?
[19-Aug-2009 12:08:37] <rmatte> yeh, somewhere, lol
[19-Aug-2009 12:08:58] <rmatte> I have this bookmarked: http://www.zenoss.com/community/docs/zenoss-api-docs/2.1/ZenModel.Device.Device-class.html
[19-Aug-2009 12:09:02] <rmatte> some of them are listed in there
[19-Aug-2009 12:09:35] <scp2004> [rmatte] Maybe you can help me with this ...
[19-Aug-2009 12:09:35] <rmatte> There are also:
[19-Aug-2009 12:09:36] <rmatte> getLocationName
[19-Aug-2009 12:09:36] <rmatte> getCreatedTimeString
[19-Aug-2009 12:09:36] <rmatte> getProdState
[19-Aug-2009 12:09:45] <rmatte> save the list I just pasted to a text file somewhere for future use
[19-Aug-2009 12:10:00] <twm1010> can you do multiple sorts?
[19-Aug-2009 12:10:25] <scp2004> [rmatte] Look .. I need this report . Cause my zentrap doesn't run on FreeBsd , so I m using Centos now
[19-Aug-2009 12:10:53] <rmatte> twm1010: you can sort by any column, but I don't know about actual multiple sorts
[19-Aug-2009 12:10:59] <twm1010> Are you going to use the report to import the devices to a new install?
[19-Aug-2009 12:10:59] <rmatte> twm1010: the sorting can be a bit flaky
[19-Aug-2009 12:11:14] <rmatte> scp2004: oh, there are better ways to back up devices
[19-Aug-2009 12:11:24] <twm1010> rmatte: was just curious, my solarwinds reports have multiple sorts, and filtering, which is really nice
[19-Aug-2009 12:11:29] <scp2004> [rmatte] But , i have 485 Devices listed . And when i used zenbackup / zenrestore to Bsd from Centos. Doesnt work ...
[19-Aug-2009 12:12:07] <rmatte> twm1010: yeh, we're going to probably be using solarwinds for reporting since Zenoss' reporting features are a bit weak at the moment
[19-Aug-2009 12:12:14] <rmatte> I'm still designing a bunch of custom reports for us
[19-Aug-2009 12:12:39] <rmatte> scp2004: one sec, I have to check some docs on what the proper procedure is
[19-Aug-2009 12:12:40] <scp2004> [rmatte] Cause the zenrestore , restores all confs . And its very different confs in Bsd to Centos ..
[19-Aug-2009 12:13:11] <scp2004> [rmatte] So .. I need a command to restore only devices .. Not configs ... You know ?
[19-Aug-2009 12:13:39] <rmatte> scp2004: yeh, working on it...
[19-Aug-2009 12:14:06] <twm1010> scp2004: there may be a command to do just that
[19-Aug-2009 12:14:28] <rmatte> twm1010: yeh, there's some option for zenbackup to just do devices
[19-Aug-2009 12:14:31] <rmatte> I'm seaching for it
[19-Aug-2009 12:14:34] <rmatte> searching*
[19-Aug-2009 12:15:50] <scp2004> [rmatte] Will help me a lot ..Thanks ...
[19-Aug-2009 12:16:49] <straterra> All UPS values added \O/
[19-Aug-2009 12:16:54] <straterra> and graphed..sweeeeet
[19-Aug-2009 12:17:05] <rmatte> hmmm, I thought there was but looks like maybe not
[19-Aug-2009 12:17:39] <rmatte> I'll keep looking
[19-Aug-2009 12:18:17] <straterra> Weird
[19-Aug-2009 12:18:34] <straterra> I have datapoints labeled like.. Input_Volts
[19-Aug-2009 12:18:49] <straterra> Yet..in the graph its called volts_inputvolts
[19-Aug-2009 12:19:14] <straterra> Oh..I can just rename it
[19-Aug-2009 12:19:17] <rmatte> you probably made a data source called volts, then made a datapoint below it called inputvolts?
[19-Aug-2009 12:19:25] <rmatte> so it combines them
[19-Aug-2009 12:19:31] <rmatte> datasource_datapoint
[19-Aug-2009 12:19:54] <straterra> Gotcha
[19-Aug-2009 12:20:11] <rmatte> also, I wouldn't make it a habbit of using underscore for datapoint or data source names
[19-Aug-2009 12:20:19] <rmatte> try to just do stuff like inputVolts
[19-Aug-2009 12:20:21] <rmatte> or whatever
[19-Aug-2009 12:20:40] <rmatte> quoting: "[01:12pm] <straterra> I have datapoints labeled like.. Input_Volts"
[19-Aug-2009 12:21:17] <straterra> Gotcha
[19-Aug-2009 12:22:56] <rmatte> scp2004: yeh, looks like you might have to just resort to using that report
[19-Aug-2009 12:23:02] <rmatte> but at least you'll only have to do it once
[19-Aug-2009 12:33:20] <scp2004> [rmatte] i ll have to go for now ..but if you find .PLease post here .. I ll see ..Thank you man ..
[19-Aug-2009 12:33:48] <scp2004> [rmatte] About this command to restore only devices ..
[19-Aug-2009 12:47:09] <DogWater> cgibbons: yeah so I have no clue about this, lol
[19-Aug-2009 12:59:25] <jb> hey
[19-Aug-2009 12:59:27] <jb> you core users
[19-Aug-2009 12:59:36] <jb> can core monitor windows services, etc out of the box?
[19-Aug-2009 12:59:50] <twm1010> Yes, indeed.
[19-Aug-2009 12:59:50] <jb> been a while since I have used core..
[19-Aug-2009 13:03:05] <rmatte> jb: it can monitor services via WMI, you can monitor processes via SNMP, you can monitor event log via WMI or via SNMP traps
[19-Aug-2009 13:03:06] <jb> without an agent on the windows box?
[19-Aug-2009 13:03:18] <rmatte> jb: Performance data can be collected via either, but WMI is easier to do it with
[19-Aug-2009 13:03:27] <rmatte> jb: Zenoss is 100% agentless
[19-Aug-2009 13:03:36] <jb> well, core used to use a winagent
[19-Aug-2009 13:03:40] <jb> before your time
[19-Aug-2009 13:03:40] <rmatte> Unless you consider SNMP or WMI services an agent
[19-Aug-2009 13:03:41] <jb>
[19-Aug-2009 13:03:44] <rmatte> which they technically are not
[19-Aug-2009 13:03:55] <rmatte> probably
[19-Aug-2009 13:04:04] <rmatte> I've only been using Zenoss for like 8 or 9 months now
[19-Aug-2009 13:04:09] <jb> yeah
[19-Aug-2009 13:04:20] <rmatte> I've learned it pretty inside and out though
[19-Aug-2009 13:33:28] <dranch> Anyone have some time to help with a command script that won't even throw errors in the zencommand.log file?
[19-Aug-2009 13:34:03] <rmatte> this is something in Event Manager -> Commands?
[19-Aug-2009 13:34:34] <dranch> This thread covers what I see but even putting the script in /usr/local/zenoss/libexec doesn't resolve: http://forums.zenoss.com/viewtopic.php?t=5467&postdays=0&postorder=asc&start=30
[19-Aug-2009 13:35:07] <dranch> If I do the manual test via the webui, it works as expected BUT the command won't run automatically
[19-Aug-2009 13:35:32] <rmatte> what's the command?
[19-Aug-2009 13:36:09] <dranch> It's a little bash script I wrote that will return a 1 if there is a failiure
[19-Aug-2009 13:36:16] <mrayzenoss> dranch: I ran into that awhile ago http://dev.zenoss.org/trac/ticket/5175
[19-Aug-2009 13:36:17] <rmatte> I see
[19-Aug-2009 13:37:14] <dranch> ok.. let me give it an absolute path
[19-Aug-2009 13:37:17] <dranch> hold
[19-Aug-2009 13:37:34] <rmatte> well, how do you have it right now?
[19-Aug-2009 13:37:54] <rmatte> the proper way to specify zenhome in a command based datasource is $$ZENHOME
[19-Aug-2009 13:37:57] <rmatte> not just $ZENHOME
[19-Aug-2009 13:38:13] <twm1010> rmatte: why?
[19-Aug-2009 13:38:16] <rmatte> for example, I use the following on many different zenoss boxes
[19-Aug-2009 13:38:17] <rmatte> $$ZENHOME/Products/SNMPWindowsPerfMonitor/libexec/wincpu.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"
[19-Aug-2009 13:38:20] <rmatte> and it works great
[19-Aug-2009 13:38:43] <rmatte> twm1010: I think the extra $ tells Zenoss to look for an environment variable, instead of a variable within Zope
[19-Aug-2009 13:38:50] <dranch> no path.. I just specify use SSH: no; event class /cmd/fail; severity: critical; cycle time: 60; command template: check-vip.sh database-fail
[19-Aug-2009 13:39:30] <rmatte> yeh, command template should be a full path
[19-Aug-2009 13:39:37] <rmatte> you can use $$ZENHOME to shorten it
[19-Aug-2009 13:39:52] <rmatte> when you're doing the test it's running it as though you were running it from a shell I believe, so it works
[19-Aug-2009 13:39:55] <dranch> I have other commands that work find though they DO use the SSH transport
[19-Aug-2009 13:40:23] <twm1010> rmatte: that makes sense, that would be why some of my perl scripts mess up getting to $ZENHOME.
[19-Aug-2009 13:40:32] <dranch> So I should use: $$ZENHOME/check-vip.sh database-fail ?
[19-Aug-2009 13:40:33] <rmatte> twm1010: probably
[19-Aug-2009 13:40:43] <rmatte> well, where is the check-vip.sh file stored?
[19-Aug-2009 13:40:54] <rmatte> what's the absolute path to it?
[19-Aug-2009 13:41:09] <dranch> usr/local/zenoss/libexec
[19-Aug-2009 13:41:50] <rmatte> ok, and if you do echo "$ZENHOME"
[19-Aug-2009 13:41:52] <rmatte> what do you get?
[19-Aug-2009 13:42:11] <dranch> I'm configuring this from the WebUI. let me try from the CMD line
[19-Aug-2009 13:42:15] <rmatte> k
[19-Aug-2009 13:42:33] <rmatte> zenhome paths are different depending on what installation method you used
[19-Aug-2009 13:42:36] <rmatte> that's why I'm checking
[19-Aug-2009 13:42:47] <dranch> $ echo $ZENHOME
[19-Aug-2009 13:42:49] <dranch> /usr/local/zenoss
[19-Aug-2009 13:42:55] <rmatte> ok...
[19-Aug-2009 13:43:08] <rmatte> so you'd do $$ZENHOME/libexec/check-vip.sh
[19-Aug-2009 13:43:13] <dranch> So I guess the proper setting is $$ZENHOME/libexec/check-vip.sh
[19-Aug-2009 13:43:18] <rmatte> correct
[19-Aug-2009 13:43:25] <scp2004> [rmatte] Hi man ..Do you found ? The command ? To restore only devices ?
[19-Aug-2009 13:43:39] <dranch> What I don't understand is why I wouldn't see errors in the zencommand.log file with a -v 10 setting
[19-Aug-2009 13:44:53] <mrayzenoss> that's why I opened the ticket
[19-Aug-2009 13:45:06] <mrayzenoss> it needs to have logging and the same path as the test button
[19-Aug-2009 13:45:40] <dranch> That previous URL I have continues on into the weeds but it seems it's intermittant for some people
[19-Aug-2009 13:46:27] <scp2004> [mrayzenoss] Hi man .. rmatte are trying help me .. But lets see if you know too .. I need a command like zenrestore. But to restore only devices , without configs ..Do you know ?
[19-Aug-2009 13:47:22] <mrayzenoss> zendevicedump?
[19-Aug-2009 13:48:00] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch04s02.html#d4e1761
[19-Aug-2009 13:48:33] <scp2004> i ll see ..thank you !
[19-Aug-2009 13:49:18] <dranch> Nope.. no change on my side
[19-Aug-2009 13:49:38] <mrayzenoss> dranch: try restarting zenhub and zencommand
[19-Aug-2009 13:50:03] <rmatte> scp2004: no, I don't think it's possible, unless Matt knows of a way...
[19-Aug-2009 13:50:16] <rmatte> Matt, is there some way to only export/import devices between Zenoss instances
[19-Aug-2009 13:50:22] <mrayzenoss> rmatte: that's the link http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch04s02.html#d4e1761
[19-Aug-2009 13:50:23] <rmatte> without exporting/importing anything else?
[19-Aug-2009 13:50:37] <rmatte> ah sorry, I was away from my desk for a few minutes, boss called me away
[19-Aug-2009 13:50:43] <dranch> I've restarted zencommand now with -v 10 and I'm greping for check-vip".. I'll try zenhub in a moment
[19-Aug-2009 13:50:48] <scp2004> [rmatte] I m reading this now ... Its true .. Only devices ..
[19-Aug-2009 13:50:58] <rmatte> ah right, zendevicedump
[19-Aug-2009 13:51:03] <rmatte> I totally forgot about that
[19-Aug-2009 13:51:04] <scp2004> yep ..
[19-Aug-2009 13:52:08] <dranch> Hmmmm.. still nothing showing up in the log
[19-Aug-2009 13:54:24] <dranch> As usual, the manual test works via the webui when I specify the zenoss machine as the "test against device"
[19-Aug-2009 13:54:50] <dranch> The only other thing to note is that I'm running this via a template and not a specific machine
[19-Aug-2009 13:56:10] <dranch> brb
[19-Aug-2009 13:59:14] <scp2004> [mrayzenoss] when i use zendevicedump : WARNING:ZEO.zrpc:(32152) CW: error connecting to ('localhost', 8100): EINVAL
[19-Aug-2009 14:00:20] <mrayzenoss> weird, haven't seen that
[19-Aug-2009 14:02:49] <rmatte> dranch: you do know the command needs to output in a certain format right?
[19-Aug-2009 14:03:08] <rmatte> scp2004: is Zenoss running when you're using the command?
[19-Aug-2009 14:10:39] <rmatte> man, King crab is only like 2-3 months away, can't wait
[19-Aug-2009 14:11:35] <mrayzenoss> I think the beta is supposed to start soon
[19-Aug-2009 14:11:51] <DogWater> crap, no matter what I do i can't get wmic to work on this remote host, although i've gotten a bit further.
[19-Aug-2009 14:12:06] <rmatte> I'll most likely be setting up a lab box with the beta, but I really can't wait until we can get the final version on to our Prod boxes
[19-Aug-2009 14:12:16] <rmatte> especially since it'll accurately reflect interface status on the OS tab
[19-Aug-2009 14:12:26] <rmatte> so no more confusion
[19-Aug-2009 14:18:54] <rmatte> haha
[19-Aug-2009 14:19:11] <rmatte> I went through and added a bunch of comments on the new event console demo site
[19-Aug-2009 14:19:16] <rmatte> I'm reading the responses
[19-Aug-2009 14:21:58] <mrayzenoss> supposedly Ian has been putting patches into the demo
[19-Aug-2009 14:22:06] <mrayzenoss> he's integrating it in trunk right now
[19-Aug-2009 14:22:16] <rmatte> not much really seems to have changed
[19-Aug-2009 14:22:26] <mrayzenoss> some IE bugs
[19-Aug-2009 14:22:44] <rmatte> it's still slow opening the details portion in firefox, when you click Show more details... it just closes the details window lol
[19-Aug-2009 14:23:23] <rmatte> and even with those problems it's still nicer than the current event console
[19-Aug-2009 14:23:24] <rmatte> lol
[19-Aug-2009 14:27:37] <rmatte> does the UI consultant you guys hired actually code, or just assists to plan out the layout?
[19-Aug-2009 14:27:48] <mrayzenoss> she does not code
[19-Aug-2009 14:27:53] <rmatte> ah
[19-Aug-2009 14:28:14] <rmatte> so it's more about the aesthetic aspect of it
[19-Aug-2009 14:28:17] <mrayzenoss> she's spent a lot of time on workflow analysis
[19-Aug-2009 14:28:25] <rmatte> ah
[19-Aug-2009 14:28:33] <mrayzenoss> so things won't be buried 7 clicks away
[19-Aug-2009 14:28:37] <rmatte> so it's more about efficiency
[19-Aug-2009 14:28:39] <rmatte> yeh
[19-Aug-2009 14:28:39] <jb> i havent seen updated screeshots in quite some time
[19-Aug-2009 14:28:45] <jb> wonder if alot has changed..
[19-Aug-2009 14:28:55] <rmatte> jb: http://public-demo.zenoss.com/evconsole
[19-Aug-2009 14:28:56] <mrayzenoss> jb: http://public-demo.zenoss.com/evconsole
[19-Aug-2009 14:28:57] <rmatte> seen that?
[19-Aug-2009 14:29:01] <jb> yeah i have seen that
[19-Aug-2009 14:29:07] <mrayzenoss> that's all that's public so far
[19-Aug-2009 14:29:08] <jb> im talking others
[19-Aug-2009 14:29:08] <rmatte> that's basically it
[19-Aug-2009 14:29:18] <jb>
[19-Aug-2009 14:29:27] <rmatte>
[19-Aug-2009 14:30:29] <mrayzenoss> so the Add Device wizard in 2.4 was the first of the new screens, then the Event Console in King Crab and then everything else in Stone Crab
[19-Aug-2009 14:30:32] <mrayzenoss> is the plan at least
[19-Aug-2009 14:30:41] <mrayzenoss> which is why we're hiring another developer
[19-Aug-2009 14:31:02] <jb> obviously some things will need to change in blue crab (the navigation, etc)
[19-Aug-2009 14:32:14] <mrayzenoss> yeah, Ian's putting the Event Console into Subversion trunk right now, so if you're a glutton for punishment...
[19-Aug-2009 14:32:53] <mrayzenoss> or you can wait for the beta
[19-Aug-2009 14:33:01] <mrayzenoss> which will have installers and VMs
[19-Aug-2009 14:40:25] <eidolon> hey folks - i'm enabling Microsoft SQL server monitoring via a template - but i'm not getting any actual data (it shows up int he performance details). do i need to specify a login/password in zproperties for sql server to get performance stats from it?
[19-Aug-2009 14:41:26] <jb> ok cool.. I'll figure it out
[19-Aug-2009 14:42:53] <mrayzenoss> eidolon: depends which ZenPack you're using to monitor
[19-Aug-2009 14:42:59] <magnachef> hey mray, I have a Zenoss scalability question
[19-Aug-2009 14:43:04] <eidolon> let me check.
[19-Aug-2009 14:43:40] <eidolon> actiually i'm not sure how to check
[19-Aug-2009 14:44:00] <eidolon> i see:
[19-Aug-2009 14:44:01] <eidolon> ${here/ZenPackManager/packs/ZenPacks.community.Perfmon/path}/libexec/perfmon.pl 1 "${dev/manageIp}" "${dev/zWinUser}" "${dev/zWinPassword}" '\${dev/zSQLInstance}:Access Methods\Full Scans/sec' "SQLServer_FullTableScansSec"
[19-Aug-2009 14:44:10] <magnachef> let's say that I have around 1000 devices, and I want to be able to assign a variable polling time to certain groups of servers. Some polling every 15 sec, some 30 sec, some 1 min, and some 5 min. Does this sound reasonable to do in Zenoss?
[19-Aug-2009 14:45:31] <mrayzenoss> eidolon: I think that's this one: http://www.zenoss.com/community/projects/zenpacks/sql-server-performance
[19-Aug-2009 14:45:58] <mrayzenoss> magnachef: right now polling is a function of which protocol you're using
[19-Aug-2009 14:46:17] <mrayzenoss> magnachef: so SNMP is managed at the collector level and WMI and SSH are at the device class
[19-Aug-2009 14:46:35] <eidolon> looks like it's using the zWinUser to log in via WMI, but not actually collecting any data. i'mwondering if the 'zenoss' login i'm using needs to be granted something internally on the machine
[19-Aug-2009 14:47:06] <mrayzenoss> magnachef: ping is also set on the collector
[19-Aug-2009 14:47:16] <mrayzenoss> magnachef: hmm... WMI is actually on the collector too
[19-Aug-2009 14:47:54] <mrayzenoss> eidolon: you may need to look at the Perl script and see what it's calling, then try those by hand to check your WMI privileges
[19-Aug-2009 14:48:13] <mrayzenoss> magnachef: I believe in King Crab all collector intervals will move to the device class
[19-Aug-2009 14:50:23] <chudler_> that.... WOULD BE NIIIICE!!!
[19-Aug-2009 14:50:55] <chudler_> not that it matters, if what one wants more resolution on graphs.
[19-Aug-2009 14:51:23] <magnachef> yeah, that's nice
[19-Aug-2009 14:52:02] <dranch> rmatte, you still around?
[19-Aug-2009 14:52:13] <dranch> You mentioned before: dranch: you do know the command needs to output in a certain format right?
[19-Aug-2009 14:52:20] <magnachef> I'm assuming that all the rrd magic is handled properly as well? or is that something the user will need to tweak?
[19-Aug-2009 14:53:03] <twm1010> mrayzenoss: collector intervals are set at the collector now?
[19-Aug-2009 14:53:53] <mrayzenoss> twm1010: yeah Collectors->localhost->Edit
[19-Aug-2009 14:54:48] <twm1010> mrayzenoss: so this is moving to a zproperty?
[19-Aug-2009 14:55:04] <mrayzenoss> unsure where it's going
[19-Aug-2009 14:55:39] <twm1010> Off topic: the collected RRD data, is it ever summarized?
[19-Aug-2009 14:56:06] <mrayzenoss> twm1010: yeah, the RRD Create command has the summarization schedules
[19-Aug-2009 14:56:46] <mrayzenoss> twm1010: older data gets summarized the farther out it gets
[19-Aug-2009 14:57:30] <twm1010> I see entries like: RRA:AVERAGE:0.5:1:600
[19-Aug-2009 14:57:39] <mrayzenoss> those are them
[19-Aug-2009 14:57:54] <mrayzenoss> http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[19-Aug-2009 14:57:57] <magnachef> that is unless you have enough steps created to hold enough single poll values
[19-Aug-2009 14:58:36] <magnachef> you'd need to bump up that 600 value
[19-Aug-2009 14:59:16] <magnachef> mray: have you guys been following reconnoiter?
[19-Aug-2009 14:59:59] <twm1010> So is summarization done on a routine? Or does RRD do it as new data is written?
[19-Aug-2009 15:00:13] <magnachef> as new data is written
[19-Aug-2009 15:00:38] <twm1010> ah, very good, but with a lot of datapoints, thats a lot disk constant disk i/o i'd imagine
[19-Aug-2009 15:00:51] <magnachef> so in the above example, after 600, the last value is gone from that RRA
[19-Aug-2009 15:01:17] <magnachef> s/600/600 data points/
[19-Aug-2009 15:01:27] <twm1010> so if you collect every 5 minutes
[19-Aug-2009 15:01:56] <magnachef> after you 601st collection, you will lose your first collected value
[19-Aug-2009 15:02:15] <twm1010> right, that value is summarized into an archive
[19-Aug-2009 15:02:40] <magnachef> yeah, but is probably already in the other RRAs
[19-Aug-2009 15:02:47] <mrayzenoss> right
[19-Aug-2009 15:02:52] <magnachef> if I remember correctly, it'll update all RRAs at the same time
[19-Aug-2009 15:02:58] <mrayzenoss> so there are multiple RRAs being used
[19-Aug-2009 15:03:15] <magnachef> but some RRAs will require multiple data points before making a datapoint for that RRA
[19-Aug-2009 15:03:43] <twm1010> ideally i only have to fool with it if i need to retain accuracy for a longer or shorter period
[19-Aug-2009 15:04:18] <magnachef> yeah, I try to keep 5 minute averages for about 2 years
[19-Aug-2009 15:04:24] <mrayzenoss> magnachef: I looked into Reconnoiter a little
[19-Aug-2009 15:05:12] <magnachef> I just watched a demo from oscon
[19-Aug-2009 15:05:35] <magnachef> the aggregate graph building is something I'd definately like to see in Zenoss
[19-Aug-2009 15:05:51] <mrayzenoss> I'll watch the demo, I read about it before OSCON
[19-Aug-2009 15:06:34] <magnachef> yeah, my company has done some work with the OmniTi guys, and I've talked briefly with Theo about it
[19-Aug-2009 15:07:31] <mrayzenoss> it looks a lot like BMC Patrol Express
[19-Aug-2009 15:08:07] <magnachef> I've never seen BMC Patrol before ($$$$)
[19-Aug-2009 15:08:21] <mrayzenoss> a few of us at Zenoss were on that team
[19-Aug-2009 15:08:27] <magnachef> ahh
[19-Aug-2009 15:08:33] <jb> oh wow.. i can't believe I accidentally just /quit
[19-Aug-2009 15:08:34] <jb> ... Total uptime : 552d 4h 39m 34s
[19-Aug-2009 15:08:35] <jb>
[19-Aug-2009 15:09:04] <mrayzenoss> at least architecturally
[19-Aug-2009 15:09:27] <jb> are the devs here tommorow?
[19-Aug-2009 15:09:37] <mrayzenoss> yup, mrchippy is on call
[19-Aug-2009 15:09:39] <magnachef> a big issue I'm having is that now that we're gathering a lot of data through Zenoss and Cacti, management is wanting more and more reports, which can be quite tedious in both systems (IMHO)
[19-Aug-2009 15:09:42] <jb> k
[19-Aug-2009 15:09:52] <mrayzenoss> yeah, our reports suck
[19-Aug-2009 15:10:04] <twm1010> mrayzenoss: Yeah, they do
[19-Aug-2009 15:10:09] <magnachef> yeah, so that's a HUGE thing with my company now
[19-Aug-2009 15:10:12] <mrayzenoss> some guy from BIRT said he was going to do an extractor, but I haven't heard back from him
[19-Aug-2009 15:10:38] <magnachef> ...the other issue is the variable polling stuff and allowing it to scale
[19-Aug-2009 15:11:27] <magnachef> so is there a big initative to fix the polling suckiness?
[19-Aug-2009 15:11:46] <mrayzenoss> yeah
[19-Aug-2009 15:11:50] <mrayzenoss> so the scheduling is now asynchronous in trunk, which will provide scaling improvements.
[19-Aug-2009 15:12:25] <mrayzenoss> but the work to move the configuration hasn't been scheduled yet
[19-Aug-2009 15:12:30] <mrayzenoss> we need more devs
[19-Aug-2009 15:12:35] <twm1010> mrayzenoss: to tell you the truth, the reporting is what stopped us from getting enterprise.
[19-Aug-2009 15:12:53] <magnachef> how much are you hiring devs for? :-)
[19-Aug-2009 15:13:05] <mrayzenoss> pay is good
[19-Aug-2009 15:13:26] <magnachef> can't find enough quality guys?
[19-Aug-2009 15:13:36] <mrayzenoss> just opened up the spot recently
[19-Aug-2009 15:13:54] <magnachef> I basically program in Perl...I'd need to tighten up my Python-Fu!
[19-Aug-2009 15:14:03] <mrayzenoss> http://www.zenoss.com/about/jobs#position_senior-web-developer
[19-Aug-2009 15:14:13] <twm1010> it might be easier to put some money into research to clone cluther
[19-Aug-2009 15:14:25] <magnachef> heh
[19-Aug-2009 15:16:49] <mrayzenoss> if you know anyone who would be a good fit, send them our way
[19-Aug-2009 15:17:07] <magnachef> hmm...where is the position out of
[19-Aug-2009 15:17:15] <mrayzenoss> Austin or Annapolis
[19-Aug-2009 15:18:08] <mrayzenoss> but if you're badass enough like cgibbons they'll let you work remote
[19-Aug-2009 15:18:30] <cgibbons> ugh
[19-Aug-2009 15:18:34] <mrayzenoss> heh
[19-Aug-2009 15:19:30] <magnachef> so will you guys be attending Ohio Linux Fest?
[19-Aug-2009 15:19:35] <mrayzenoss> I'll be there
[19-Aug-2009 15:19:42] <mrayzenoss> we're gonna have a free training day
[19-Aug-2009 15:19:53] <twm1010> when is that?
[19-Aug-2009 15:19:55] <magnachef> that would be awesome...I'm about 3 hrs away and was planning on going
[19-Aug-2009 15:20:02] <mrayzenoss> training will be on Friday
[19-Aug-2009 15:20:03] <magnachef> ...that would seal the deal
[19-Aug-2009 15:20:21] <mrayzenoss> condensed version of the 2-day training we offer for $
[19-Aug-2009 15:20:32] <magnachef> yeah, I'd like to get me some of that
[19-Aug-2009 15:21:08] <twm1010> hrmm, i MIGHT swing that...
[19-Aug-2009 15:22:03] <magnachef> so who do I need to pressure to make sure that happens? @mrhinkle?
[19-Aug-2009 15:22:07] <magnachef> :-)
[19-Aug-2009 15:22:10] <mrayzenoss> it's done
[19-Aug-2009 15:22:20] <mrayzenoss> I'm making my plane reservations today
[19-Aug-2009 15:22:24] <mrayzenoss> to get there on Thursday
[19-Aug-2009 15:22:25] <magnachef> oh...well that was easy!
[19-Aug-2009 15:23:35] <magnachef> so class on Friday?
[19-Aug-2009 15:23:41] <mrayzenoss> yeah
[19-Aug-2009 15:24:11] <magnachef> alright, cool. I'll keep my eyes open for the updates
[19-Aug-2009 15:24:12] <dranch> mrayzenoss: what would be the reason for data source "commands" not being logged at all?
[19-Aug-2009 15:24:50] <mrayzenoss> dranch: I'm not sure, that's why it took me so long to discover that bug
[19-Aug-2009 15:25:09] <mrayzenoss> so you're using zencommand run -v10 -d yourdevice ?
[19-Aug-2009 15:25:11] <dranch> So I'm at a loss of how to get this working. I can't see it fail and it doesn't work
[19-Aug-2009 15:25:37] <dranch> the manual webui test works fine.. nothing happens otherwise
[19-Aug-2009 15:26:33] <dranch> If I run that command, I see
[19-Aug-2009 15:26:38] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 15:26:39] <dranch> DEBUG:zen.zencommand:Finished config fetch
[19-Aug-2009 15:26:41] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 15:26:42] <dranch> INFO:zen.zencommand:Daemon zencommand shutting down
[19-Aug-2009 15:26:44] <dranch> DEBUG:zen.zencommand:removing service EventService
[19-Aug-2009 15:26:45] <dranch> DEBUG:zen.zencommand:removing service CommandConfig
[19-Aug-2009 15:26:47] <dranch> INFO:zen.zencommand:zencommand shutting down
[19-Aug-2009 15:26:50] <dranch> Seems it thinks there aren't any commands to run
[19-Aug-2009 15:26:59] <mrayzenoss> try 'zencommand stop'
[19-Aug-2009 15:27:05] <mrayzenoss> which will stop the zencommand daemon
[19-Aug-2009 15:27:11] <dranch> done
[19-Aug-2009 15:27:51] <dranch> and then what.. ./zencommand start -v 10 ?
[19-Aug-2009 15:27:56] <jb> so.. I have a object class that is not compiling. does anyone know where I can look for any debugging? zenhub.log is not showing any pertienent information
[19-Aug-2009 15:28:19] <mrayzenoss> no, just the zencommand run -v 10
[19-Aug-2009 15:28:34] <dranch> ok.. done
[19-Aug-2009 15:28:44] <mrayzenoss> anything relevant in the output?
[19-Aug-2009 15:29:58] <dranch> nope..nothing for the expected hostname or it's resolved IP address
[19-Aug-2009 15:31:28] <mrayzenoss> crap, my home machine crashed
[19-Aug-2009 15:31:44] <mrayzenoss> I was going to log in there and see what my zencommands are doing
[19-Aug-2009 15:32:02] <magnachef> hopefully not crashing your machine ;-)
[19-Aug-2009 15:32:17] <dranch> It doesn't seem to be logging any commands for this template or machine
[19-Aug-2009 15:32:17] <rmatte> too bad you don't have some UPS that you can login to and reboot power
[19-Aug-2009 15:32:50] <mrayzenoss> dranch: do you have an empty graph?
[19-Aug-2009 15:32:53] <mrayzenoss> or nothing?
[19-Aug-2009 15:33:07] <dranch> I'm not looking to graph, only send alerts
[19-Aug-2009 15:33:16] <mrayzenoss> is the template bound?
[19-Aug-2009 15:33:25] <dranch> how do I confrim that?
[19-Aug-2009 15:33:45] <mrayzenoss> go to the device class with the template
[19-Aug-2009 15:33:53] <mrayzenoss> click "Templates"
[19-Aug-2009 15:33:54] <dranch> ok
[19-Aug-2009 15:34:06] <dranch> ok
[19-Aug-2009 15:34:07] <mrayzenoss> go to the page menu and select "Bind Templates"
[19-Aug-2009 15:34:26] <dranch> ok..
[19-Aug-2009 15:34:41] <mrayzenoss> is your custom template highlighted like Device?
[19-Aug-2009 15:34:51] <dranch> I see device, ethernetcdmad*, filesystem, ipservice, mysql, osprocesses, and winservice
[19-Aug-2009 15:35:01] <dranch> device is highlighed
[19-Aug-2009 15:35:12] <dranch> that's where I created my data source
[19-Aug-2009 15:36:01] <mrayzenoss> no, on the Templates page, there's a menu next to "Available Performance Templates"
[19-Aug-2009 15:36:28] <mrayzenoss> and Bind Templates is a choice there
[19-Aug-2009 15:36:45] <dranch> right.. under the down arrow right?
[19-Aug-2009 15:36:52] <mrayzenoss> yeah
[19-Aug-2009 15:37:12] <dranch> right.. and that pops up a window with the list of templates I listed above
[19-Aug-2009 15:37:18] <mrayzenoss> ahh, ok
[19-Aug-2009 15:37:27] <mrayzenoss> and your custom template is bound?
[19-Aug-2009 15:37:36] <dranch> Device [ /devices/server/linux/database/replication] is the top entry and it's highlighed
[19-Aug-2009 15:38:01] <mrayzenoss> so you modified the Device template itself, instead of a custom template?
[19-Aug-2009 15:38:03] <dranch> Should I click on "ok" to re-bind it?
[19-Aug-2009 15:38:09] <dranch> yes
[19-Aug-2009 15:38:27] <dranch> should I place it somewhere else?
[19-Aug-2009 15:38:27] <mrayzenoss> that should be ok
[19-Aug-2009 15:39:01] <dranch> I have three database servers as part of this template and I'd like to run this command on all three machines
[19-Aug-2009 15:40:18] <mrayzenoss> I assume you've pushed your changes and restarted everything for grins?
[19-Aug-2009 15:41:19] <dranch> This is a single machine and I have restarted a few times (zencommand and zenhub). Should I do the whole thing?
[19-Aug-2009 15:42:02] <dranch> I don't understand why zenoss things there aren't any commands to run. There are a total of three commands on that page and it seems NONE of them are being run
[19-Aug-2009 15:42:43] <mrayzenoss> dranch: I'm with you, it sounds like everything should be in the right places
[19-Aug-2009 15:43:07] <mrayzenoss> perhaps remove the custom device template and try adding the commands manually
[19-Aug-2009 15:43:18] <mrayzenoss> I mean, a separate template for each
[19-Aug-2009 15:43:19] <dranch> This has been enlightening in the fact that now I know that Zenoss isn't running anything that's been configured
[19-Aug-2009 15:43:30] <dranch> Uh... ok, let me try
[19-Aug-2009 15:43:30] <mrayzenoss> just to narrow it down
[19-Aug-2009 15:43:46] <dranch> There are also a ton of SNMP values
[19-Aug-2009 15:44:13] <mrayzenoss> ok, just copy and paste the command stuff into a new template
[19-Aug-2009 15:46:06] <mrayzenoss> man, I was working on a bug related to this right before I left on vacation 2 weeks ago
[19-Aug-2009 15:46:13] <mrayzenoss> on my home box that crashed
[19-Aug-2009 15:46:20] <dranch> figures huh?
[19-Aug-2009 15:46:29] <dranch> (for me at least)
[19-Aug-2009 15:46:39] <mrayzenoss> I noticed it had crashed this morning and I was going to take a screenshot so I could debug it
[19-Aug-2009 15:46:51] <mrayzenoss> some weird kernel issue on my Dell mini
[19-Aug-2009 15:46:59] <mrayzenoss> it
[19-Aug-2009 15:47:06] <mrayzenoss> it's been crashing after 2 weeks
[19-Aug-2009 15:47:11] <mrayzenoss> fairly consistently
[19-Aug-2009 15:47:12] <dranch> Ok.. so I added the new template
[19-Aug-2009 15:47:19] <jb> how can I verify relationships in zope?
[19-Aug-2009 15:47:21] <dranch> do I need to bind it?
[19-Aug-2009 15:47:24] <jb> somehow thru zport?
[19-Aug-2009 15:47:33] <mrayzenoss> dranch: yes
[19-Aug-2009 15:47:37] <scp2004> [mrayzenoss] sorry man ..I was afk .. About the error using zendevicedump .. I must use it with zenoss running ?
[19-Aug-2009 15:47:46] <mrayzenoss> scp2004: yes
[19-Aug-2009 15:47:54] <scp2004> oh ..ok ..i ll try
[19-Aug-2009 15:48:25] <dranch> In the bind performance templates, do I only need to select this new "DBmonitor" entry and select ok or do multiple things need to be selected?
[19-Aug-2009 15:50:23] <rmatte> I can't wait until the day when http://dev.zenoss.org/trac/ticket/5289 gets fixed. It's usually not a big deal but it's absolutely driving me up the wall right now
[19-Aug-2009 15:51:01] <jb> hmm
[19-Aug-2009 15:51:01] <jb> you know
[19-Aug-2009 15:51:05] <jb> i think i need to manually rebuild my relations.
[19-Aug-2009 15:51:48] <scp2004> [mrayzenoss] Waf ! ...
.. File "/usr/local/zenoss/Products/ZenRelations/ExportDevices.py", line 229, in ?[19-Aug-2009 15:51:48] <scp2004> ex.export()
[19-Aug-2009 15:51:53] <dranch> Ray? thoughts?
[19-Aug-2009 15:52:16] <mrayzenoss> dranch: yeah, select the new one and the previous ones
[19-Aug-2009 15:52:21] <rmatte> dranch: depends on what templates you want assigned to it
[19-Aug-2009 15:52:26] <dranch> So ALL should be selected?
[19-Aug-2009 15:52:34] <mrayzenoss> all that you want to use
[19-Aug-2009 15:52:35] <rmatte> I wouldn't select ALL
[19-Aug-2009 15:52:38] <rmatte> just the relevant ones
[19-Aug-2009 15:53:03] <mrayzenoss> templates are available hierarchally, so there may be some defined farther up the hierarchy that you don't care for
[19-Aug-2009 15:53:19] <dranch> How can I tell?
[19-Aug-2009 15:53:53] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 15:54:15] <rmatte> dranch: you just kind of have to know
[19-Aug-2009 15:55:54] <dranch> restarted zencommand and zenhub, still schedule has 0 commands
[19-Aug-2009 15:57:06] <scp2004> [mrayzenoss] any idea ?
.. File "/usr/local/zenoss/Products/ZenRelations/ExportDevices.py", line 229, in ?[19-Aug-2009 15:57:33] <mrayzenoss> scp2004: use pastebin.com and post the whole stack trace
[19-Aug-2009 15:57:51] <mrayzenoss> dranch: try "Push Changes"
[19-Aug-2009 15:58:00] <dranch> Where's that?
[19-Aug-2009 15:58:04] <mrayzenoss> from the device in question, go to the page menu
[19-Aug-2009 15:58:43] <mrayzenoss> on the status page, below "Manage->Model Device"
[19-Aug-2009 15:59:37] <dranch> ok.. done
[19-Aug-2009 16:00:12] <rmatte> yeh, Push Changes is your friend
[19-Aug-2009 16:00:20] <mrayzenoss> did that work?
[19-Aug-2009 16:00:26] <rmatte> since the collector takes upwards of an hour to grab updated info from zenhub
[19-Aug-2009 16:00:44] <mrayzenoss> lazy zenhub
[19-Aug-2009 16:00:54] <rmatte> yeh lol
[19-Aug-2009 16:01:06] <mrayzenoss> it needs more workers
[19-Aug-2009 16:01:17] <dranch> :zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:01:27] <rmatte> is the command enabled?
[19-Aug-2009 16:01:54] <dranch> enabled: true
[19-Aug-2009 16:02:01] <rmatte> ok
[19-Aug-2009 16:02:03] <rmatte> hmmm
[19-Aug-2009 16:02:36] <rmatte> but other command based datapoints are working?
[19-Aug-2009 16:03:14] <dranch> name: DB VIP monitor; source type: command; enabled true; use ssh: false; component: DB VIP heartbeat; event class: /Cmd/Fail; Event key: DB heartbeat; severity: critical; cycle time: 60l command template: $$ZENHOME/libexec/check-vip.sh database-fail
[19-Aug-2009 16:03:45] <dranch> No, there are two other commands that I "THOUGHT" were running but it seems they aren't being issued either
[19-Aug-2009 16:04:01] <rmatte> cycle time: 60l?
[19-Aug-2009 16:04:14] <dranch> sorry: 60
[19-Aug-2009 16:04:18] <rmatte> k
[19-Aug-2009 16:04:53] <rmatte> what's the output of that script look like?
[19-Aug-2009 16:05:11] <dranch> In the device template, there is a SSH commend to run mysql, and another command to run /opt/zenoss/libexec/check_mysql_stat -u zenoss -p xxx -H ${device/manageIp} -S
[19-Aug-2009 16:05:21] <dranch> seems those aren't running now either.
[19-Aug-2009 16:05:42] <rmatte> but they were working fine previously I assume?
[19-Aug-2009 16:06:07] <dranch> test against device: database5 ---> Executing command against database5.admin.zvents.com
[19-Aug-2009 16:06:08] <dranch> 1
[19-Aug-2009 16:06:10] <dranch> DONE in 1 seconds
[19-Aug-2009 16:06:24] <rmatte> ummmm, that's not really how a command based datapoint works
[19-Aug-2009 16:06:26] <dranch> I've inherited this setup.. I don't have much history with it
[19-Aug-2009 16:06:45] <rmatte> you need the output to be something like OK|status=1 or OK|status=0 or Unknown
[19-Aug-2009 16:06:58] <rmatte> then you need to make a datapoint named "status" under the data source
[19-Aug-2009 16:07:09] <dranch> hmmm.. ok. let me change the output
[19-Aug-2009 16:07:12] <rmatte> otherwise Zenoss doesn't know how to interpret the data
[19-Aug-2009 16:07:44] <rmatte> you can even just do |status=1 or |status=2 or Unknown
[19-Aug-2009 16:07:52] <rmatte> Nagios format output
[19-Aug-2009 16:07:56] <dranch> So if the test is /Cmd/fail, and it's a fail, the result should be a OK|status=0?
[19-Aug-2009 16:08:09] <rmatte> yes
[19-Aug-2009 16:08:27] <rmatte> if 0 is what indicated a failure
[19-Aug-2009 16:08:31] <rmatte> indicates*
[19-Aug-2009 16:08:50] <rmatte> and if it succeeds it should be OK|status=1
[19-Aug-2009 16:08:59] <rmatte> the OK| just represents that it was able to gather the data
[19-Aug-2009 16:09:14] <rmatte> if you whatever reason it can't check (which is unlikely in this case) it should output Unknown
[19-Aug-2009 16:09:19] <rmatte> but I wouldn't worry about that part
[19-Aug-2009 16:09:37] <dranch> But it from the above zencommand logs, the commands don't seem to be running in the first place
[19-Aug-2009 16:09:41] <dranch> or is that a logging issue?
[19-Aug-2009 16:09:48] <mrayzenoss> dranch: the instructions with screenshots for http://www.zenoss.com/community/projects/zenpacks/nagios-check-ping work well
[19-Aug-2009 16:09:58] <rmatte> well, if you haven't created any datapoints under the command then it's not going to try to bother running
[19-Aug-2009 16:10:28] <rmatte> the datapoint name needs to match what's in the output
[19-Aug-2009 16:10:46] <rmatte> so if you had OK|whatever=1 your datapoint name would be whatever
[19-Aug-2009 16:11:20] <dranch> Ok, for a /Cmd/Fail setup, and and an test FAILURE, it's now returning: OK|status=1
[19-Aug-2009 16:11:24] <rmatte> that way you can have scripts outputting multiple info like OK|CPU1=7 CPU2=10 CPU3=20 CPU4=80
[19-Aug-2009 16:11:42] <rmatte> and then create the datapoints CPU1, CPU2, CPU3, and CPU4 under the command
[19-Aug-2009 16:11:52] <rmatte> and zenoss will assign the corresponding value to each datapoint
[19-Aug-2009 16:12:09] <rmatte> ok, so 1 indicates a failure I assume?
[19-Aug-2009 16:12:13] <dranch> Again, I'm just looking for a failure alert...
[19-Aug-2009 16:12:15] <dranch> correct
[19-Aug-2009 16:12:20] <dranch> I don't need to graph, etc
[19-Aug-2009 16:12:29] <rmatte> ok, so now, under the command options, you'll see a DataPoints section
[19-Aug-2009 16:12:41] <dranch> yup..
[19-Aug-2009 16:12:44] <rmatte> you need to add a datapoint that's called status
[19-Aug-2009 16:12:51] <rmatte> you can just use gauge for the type
[19-Aug-2009 16:12:58] <dranch> ok
[19-Aug-2009 16:13:07] <rmatte> then you can add your threshold against that datapoint and set it to 1, so it triggers when it hits 1
[19-Aug-2009 16:13:44] <dranch> there is name: status; type: gauge; RRD min <null>;l RRD max <null> and what do I send in create cmd: ?
[19-Aug-2009 16:14:12] <rmatte> just leave the datapoint options at default
[19-Aug-2009 16:14:20] <rmatte> you don't need to change anything, the datapoint just has to exist
[19-Aug-2009 16:14:58] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:15:03] <rmatte> once it's there just create a threshold against it, and hopefully Zenoss will run it
[19-Aug-2009 16:15:35] <dranch> ok, and that's not done in the template? if not, where?
[19-Aug-2009 16:15:44] <rmatte> the threshold is done right in the template
[19-Aug-2009 16:15:52] <rmatte> in the thresholds section
[19-Aug-2009 16:16:11] <rmatte> add a min-max threshold, name it whatever
[19-Aug-2009 16:16:20] <rmatte> select that datapoint
[19-Aug-2009 16:16:26] <rmatte> it'll be whatever_status
[19-Aug-2009 16:16:37] <rmatte> set the max value to 1, and set the event class
[19-Aug-2009 16:16:43] <rmatte> severity as well if you wish
[19-Aug-2009 16:16:45] <rmatte> then click save
[19-Aug-2009 16:16:47] <dranch> I'm back at the templates tab, under the down arrow, I don't see any option for threshold
[19-Aug-2009 16:17:01] <rmatte> no, go in to the template itself
[19-Aug-2009 16:17:08] <rmatte> and there's a threshold section
[19-Aug-2009 16:17:12] <dranch> Ah.. I see it. sorry
[19-Aug-2009 16:17:16] <rmatte> no prob
[19-Aug-2009 16:17:21] <rmatte> follow my above instructions
[19-Aug-2009 16:17:47] <rmatte> make sure the threshold is enabled as well
[19-Aug-2009 16:18:02] <rmatte> then at least you're done for the config end of things
[19-Aug-2009 16:18:43] <rmatte> then all we need to do is figure out why the heck the command isn't running
[19-Aug-2009 16:18:47] <dranch> ok.. done but do I have to restart zencommand, etc?
[19-Aug-2009 16:19:07] <rmatte> At this point i'd just restart Zenoss for the heck of it to see if that clears it up
[19-Aug-2009 16:19:33] <dranch> restarting
[19-Aug-2009 16:19:40] <rmatte> k
[19-Aug-2009 16:20:13] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:20:20] <rmatte> gah
[19-Aug-2009 16:20:40] <rmatte> what options are you using with zencommand?
[19-Aug-2009 16:21:03] <dranch> cd /usr/local/zenoss/bin; ./zencommand run -v10 -d database5.admin
[19-Aug-2009 16:21:14] <rmatte> try: zencommand run -v10 --cycle
[19-Aug-2009 16:21:21] <rmatte> see if any commands run through at all
[19-Aug-2009 16:21:36] <dranch> without the cycle, I do see:
[19-Aug-2009 16:21:40] <dranch> DEBUG:zen.zencommand:run
[19-Aug-2009 16:21:41] <dranch> DEBUG:zen.zencommand:Connecting to localhost
[19-Aug-2009 16:21:43] <dranch> DEBUG:zen.zencommand:Logging in as admin
[19-Aug-2009 16:21:44] <dranch> WARNING:zen.zencommand:Reconnected to ZenHub
[19-Aug-2009 16:21:46] <dranch> DEBUG:zen.zencommand:setting up services EventService, CommandConfig
[19-Aug-2009 16:21:47] <dranch> DEBUG:zen.zencommand:chaining getInitialServices with d2
[19-Aug-2009 16:21:49] <dranch> DEBUG:zen.zencommand:callback after getting service EventService
[19-Aug-2009 16:21:50] <dranch> DEBUG:zen.zencommand:callback after getting service CommandConfig
[19-Aug-2009 16:21:52] <dranch> DEBUG:zen.zencommand:Calling connected.
[19-Aug-2009 16:21:53] <dranch> DEBUG:zen.zencommand:connected
[19-Aug-2009 16:21:55] <dranch> DEBUG:zen.zencommand:Sending event {'severity': 0, 'component': 'zencommand', 'agent': 'zencommand', 'summary': 'started', 'manager': 'localhost', 'device': 'admin2.admin.zvents.com', 'eventClass': '/App/Start'}
[19-Aug-2009 16:21:56] <dranch> DEBUG:zen.zencommand:Fetching config
[19-Aug-2009 16:21:58] <dranch> DEBUG:zen.zencommand:Updated configCycleInterval config to 360
[19-Aug-2009 16:21:59] <dranch> DEBUG:zen.zencommand:Loading classes ['Products.ZenModel.MinMaxThreshold']
[19-Aug-2009 16:22:01] <dranch> DEBUG:zen.thresholds:Updating threshold ('zeneventlog cycle time', ('localhost', ''))
[19-Aug-2009 16:22:02] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenmodeler cycle time', ('localhost', ''))
[19-Aug-2009 16:22:04] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenperfsnmp cycle time', ('localhost', ''))
[19-Aug-2009 16:22:05] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenping cycle time', ('localhost', ''))
[19-Aug-2009 16:22:07] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenprocess cycle time', ('localhost', ''))
[19-Aug-2009 16:22:09] <dranch> DEBUG:zen.thresholds:Updating threshold ('zenwin cycle time', ('localhost', ''))
[19-Aug-2009 16:22:11] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:22:13] <dranch> DEBUG:zen.zencommand:Finished config fetch
[19-Aug-2009 16:22:15] <dranch> INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:22:17] <dranch> INFO:zen.zencommand:Daemon zencommand shutting down
[19-Aug-2009 16:22:19] <dranch> DEBUG:zen.zencommand:removing service EventService
[19-Aug-2009 16:22:21] <dranch> DEBUG:zen.zencommand:removing service CommandConfig
[19-Aug-2009 16:22:23] <dranch> INFO:zen.zencommand:zencommand shutting down
[19-Aug-2009 16:22:25] <dranch> --cycle didn't change anything
[19-Aug-2009 16:22:32] <rmatte> strange
[19-Aug-2009 16:22:49] <rmatte> so it's running the zenoss internal commands but not a single template command
[19-Aug-2009 16:23:14] <dranch> seems like it
[19-Aug-2009 16:23:34] <rmatte> ah
[19-Aug-2009 16:23:36] <rmatte> http://forums.zenoss.com/viewtopic.php?p=36653
[19-Aug-2009 16:23:43] <rmatte> the upgrade caused a "null-value" for zSshConcurrentSessions. zencommand doesnt
[19-Aug-2009 16:23:43] <rmatte> run at all without any value. Setting a value will fix it. I set it to 10.
[19-Aug-2009 16:23:58] <dranch> could be that the others are misconfigured though. There are only 4 thresholds. CPU utilizatio, raw user 400, raw user 500, and this new DB VIP change
[19-Aug-2009 16:23:58] <rmatte> that's apparently the solution
[19-Aug-2009 16:24:09] <rmatte> go in to zProperties and set that value
[19-Aug-2009 16:24:48] <dranch> It's not run via SSH
[19-Aug-2009 16:25:03] <rmatte> doesn't matter, that seems to cause general issues with zencommand
[19-Aug-2009 16:25:05] <rmatte> regardless
[19-Aug-2009 16:25:13] <rmatte> just check to make sure a value is set
[19-Aug-2009 16:25:18] <mrayzenoss> is that error in the zenhub log?
[19-Aug-2009 16:25:34] <jb> yawn
[19-Aug-2009 16:26:43] <dranch> I don't have that option available in this zProperties
[19-Aug-2009 16:26:44] <rmatte> just check in zProperties for /Devices, if it's already set then great, if not, then that's your problem
[19-Aug-2009 16:26:47] <rmatte> default value is 10
[19-Aug-2009 16:26:53] <rmatte> what version of Zenoss?
[19-Aug-2009 16:27:13] <dranch> Zenoss 2.2.4
[19-Aug-2009 16:27:18] <rmatte> I have a 2.3.3 box that doesn't have that option
[19-Aug-2009 16:27:21] <rmatte> ah, you're super oldschool
[19-Aug-2009 16:27:25] <dranch> Upgrading isn't an option
[19-Aug-2009 16:27:25] <rmatte> ok, that changes things
[19-Aug-2009 16:28:14] <rmatte> well, i'm pretty much out of ideas, if we don't have any output or logs to go on there's not much that can be done
[19-Aug-2009 16:28:23] <rmatte> need to have some idea of what may be wrong
[19-Aug-2009 16:28:31] <dranch> What do you need?
[19-Aug-2009 16:28:52] <rmatte> you've checked $ZENHOME/logs to see if any of the files contain errors?
[19-Aug-2009 16:29:03] <rmatte> (errors that may be related to the cause of this?)
[19-Aug-2009 16:29:14] <dranch> let me see if anything new is in there
[19-Aug-2009 16:29:26] <dranch> nothing in zencommand.log
[19-Aug-2009 16:29:26] <robos> hi: for the disk i/o performance graph, i have a solaris box that shows the graph climbing.. which is inaccurate. However the linux box works as expected. Any suggestions on what to check?
[19-Aug-2009 16:30:29] <mrayzenoss> robos: SNMP or SSH?
[19-Aug-2009 16:30:30] <rmatte> robos: just climbing infinitely?
[19-Aug-2009 16:30:42] <robos> snmp. And it climbs infinitely... just keeps going up
[19-Aug-2009 16:30:46] <robos> never goes down
[19-Aug-2009 16:31:21] <mrayzenoss> sounds like it's probably got the wrong OID
[19-Aug-2009 16:31:30] <robos> wrong OID?
[19-Aug-2009 16:31:32] <rmatte> yeh, does sound that way
[19-Aug-2009 16:31:44] <rmatte> sounds like it's got an OID for uptime or something
[19-Aug-2009 16:32:00] <robos> hmm
[19-Aug-2009 16:32:01] <mrayzenoss> robos: so SNMP uses individual datapoints called OIDs that expose a metric for collection
[19-Aug-2009 16:32:02] <rmatte> (or something else that increments)
[19-Aug-2009 16:32:15] <robos> ah
[19-Aug-2009 16:32:26] <robos> gotcha
[19-Aug-2009 16:32:29] <mrayzenoss> which version of Solaris?
[19-Aug-2009 16:33:06] <rmatte> mrayzenoss: is it maybe because of the datapoint type?
[19-Aug-2009 16:33:09] <rmatte> might not be a gauge?
[19-Aug-2009 16:33:21] <robos> solaris 10
[19-Aug-2009 16:33:28] <robos> i have a bunch added and they all do that
[19-Aug-2009 16:34:37] <rmatte> actually yeh, the type wouldn't affect it
[19-Aug-2009 16:34:56] <dranch> rmatte: so looking in the zencommand.log and there ARE other zencommands being run for other database servers that are under a different template. Btw, those tests DON'T have any datapoints or threasholds configured
[19-Aug-2009 16:35:32] <rmatte> dranch: hmmm, so their output is just 1 or whatever?
[19-Aug-2009 16:36:41] <rmatte> I'm leaving for the day, gotta go to catch my ride
[19-Aug-2009 16:37:03] <dranch> <sigh>.. well, thanks for the help you've given me.. though I'm still foobared
[19-Aug-2009 16:39:35] <mrayzenoss> robos: this is all I could find: http://dev.zenoss.org/trac/ticket/5237
[19-Aug-2009 16:39:47] <mrayzenoss> and I asked the dev about I/O and he said he hadn't seen anything like that
[19-Aug-2009 16:40:27] <robos> hmm, so what does this mean mrayzenoss?
[19-Aug-2009 16:40:35] <robos> i'm looking at that ticket..
[19-Aug-2009 16:41:14] <mrayzenoss> it means we fixed some other stuff for Solaris 10, but not your bug
[19-Aug-2009 16:41:20] <robos> ah
[19-Aug-2009 16:41:31] <robos> then i wonder why nobody else has this problem
[19-Aug-2009 16:41:35] <mrayzenoss> but I know we've got Solaris users
[19-Aug-2009 16:41:47] <robos> this is solaris 10 on SPARC... a T1000 i believe
[19-Aug-2009 16:43:03] <robos> *shrugs, guess I'll look into this tomorrow
[19-Aug-2009 16:43:08] <robos> cool; ty for your time guys
[19-Aug-2009 16:45:10] <dranch> myray: when do you get in tomorrow?
[19-Aug-2009 16:45:42] <dranch> Since some templates run commands and others don't, I think something uglier is afoot
[19-Aug-2009 16:46:40] <mrayzenoss> I'll be in for the IRC session tomorrow
[19-Aug-2009 16:46:43] <mrayzenoss> and afterwards
[19-Aug-2009 16:46:57] <dranch> sure.. but what hours PST?
[19-Aug-2009 16:47:07] * mrayzenoss does the math
[19-Aug-2009 16:47:12] <mrayzenoss> 8am
[19-Aug-2009 16:47:47] <mrayzenoss> from the blog: Zenoss Core Reaches 1 Million Downloads http://blog.zenoss.com/2009/08/19/network-monitoring/
[19-Aug-2009 16:47:52] <dranch> Ok.. I'll keep hacking around to see if I can get anything to run. By seeing other commands in other templates running WITHOUTH thresholds, etc. this is something else
[19-Aug-2009 16:48:09] <mrayzenoss> dranch: yeah, I suspect something is hosed in the binding
[19-Aug-2009 16:48:17] <mrayzenoss> just hard to find that
[19-Aug-2009 16:48:29] <dranch> yeah..
[19-Aug-2009 16:48:39] <mrayzenoss> try this... go to the status page for the machine being monitored
[19-Aug-2009 16:48:54] <dranch> ok
[19-Aug-2009 16:49:10] <mrayzenoss> select More->Templates
[19-Aug-2009 16:49:20] <dranch> uhuh
[19-Aug-2009 16:49:30] <mrayzenoss> is your template bound there?
[19-Aug-2009 16:49:59] <dranch> yup.. both the modified Device template and the new "DBmonitor" one. Neither seem to run
[19-Aug-2009 16:50:04] <mrayzenoss> crap
[19-Aug-2009 16:50:22] <mrayzenoss> I saw templates bound at the device class not get bound to the device before
[19-Aug-2009 16:50:26] <dranch> I even made the DBMonitor one to be a "local" copy. No change
[19-Aug-2009 16:50:41] <dranch> The path for that template is: /Devices /Server /Linux /database /replication /database5.admin.zvents.com
[19-Aug-2009 16:50:57] <mrayzenoss> my home box will be back up tomorrow, I'll have a better testbed
[19-Aug-2009 16:50:58] <dranch> This one DOES work: /Devices /Server /Linux /database /non-replication /load-database2.admin.zvents.com
[19-Aug-2009 16:51:24] <dranch> It shows only one template: device
[19-Aug-2009 16:52:06] <dranch> ha.. it also shows: INFO:zen.zencommand:---------- - schedule has 0 commands
[19-Aug-2009 16:52:13] <dranch> I don't trust the logging now
[19-Aug-2009 16:52:36] <mrayzenoss> well, the schedule is things that are the queue
[19-Aug-2009 16:52:41] <mrayzenoss> so the queue isn't backed up
[19-Aug-2009 16:52:55] <dranch> that's not an atomic test?
[19-Aug-2009 16:53:05] <mrayzenoss> I'm not sure
[19-Aug-2009 16:53:15] <dranch> Sure seems like it is
[19-Aug-2009 16:53:16] * mrayzenoss wishes he could get to his test box
[19-Aug-2009 16:53:32] <mrayzenoss> I've got a bunch of zencommands on my home box
[19-Aug-2009 16:54:37] <dranch> Hmmm.. seems that this load-database2 WAS working.. but stopped around 2:18.. I broke something
[19-Aug-2009 16:54:57] <dranch> 2009-08-19 14:18:49 DEBUG zen.zencommand: The result of "/opt/zenoss/libexec/check_mysql_stat -u zenoss -p xxx -H 10.0.30.174" was "Uptime: 30365675 Threads: 1
[19-Aug-2009 16:56:37] <dranch> ack.. false alarm.. zencommand was missing the -v10 option.. it's still running
[19-Aug-2009 16:57:27] <dranch> even still, the manual zencommand shows INFO:zen.zencommand:---------- - schedule has 0 commands :: which is wrong
[19-Aug-2009 18:10:46] kisielk_ is now known as kisielk
[19-Aug-2009 18:58:15] <straterra> rmatte: you around?
[20-Aug-2009 00:00:46] [disconnected at Thu Aug 20 00:00:46 2009]
[20-Aug-2009 00:00:46] [connected at Thu Aug 20 00:00:46 2009]
[20-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[20-Aug-2009 05:15:05] cwj_ is now known as cwj
[20-Aug-2009 09:16:35] <ckrough> "INFO ZODB.Conflict database read conflict error" when I try to run a /viewGraphReport , and I get no graph report.
[20-Aug-2009 09:16:39] <rmatte> has anyone else noticed that they've removed the --debug option from zenmib in latest Zenoss versions?
[20-Aug-2009 09:16:56] <ckrough> nah, kinda load mib once and never look back for me
[20-Aug-2009 09:17:09] <rmatte> that option was pure gold, I can't believe that
[20-Aug-2009 09:17:21] <rmatte> I would have never been able to get the Cisco Mibs loaded without it
[20-Aug-2009 09:17:44] <ckrough> which version?
[20-Aug-2009 09:17:57] <rmatte> and now it appears to be gone, since the script isn't accepting the option, and there's absolutely no reference to it in zenmib.py
[20-Aug-2009 09:18:02] <rmatte> 2.4.1
[20-Aug-2009 09:18:07] <rmatte> and later versions
[20-Aug-2009 09:19:36] <jb> morning
[20-Aug-2009 09:19:41] <rmatte> good morning
[20-Aug-2009 09:19:49] <ckrough> morning
[20-Aug-2009 09:22:12] <rmatte> yeh, the option is available on my 2.3.3 boxes
[20-Aug-2009 09:22:16] <rmatte> but not on my 2.4.1 boxes
[20-Aug-2009 09:22:24] <rmatte> no idea why they'd remove that option
[20-Aug-2009 09:23:04] <rmatte> I'll ask Matt about it when he signs on
[20-Aug-2009 09:24:57] <rmatte> hopefully there's some alternative
[20-Aug-2009 09:25:09] <ckrough> does it provide more than V10
[20-Aug-2009 09:25:27] <rmatte> ckrough: hmmm, that's actually a good question
[20-Aug-2009 09:25:34] <ckrough> Maybe it was redundna
[20-Aug-2009 09:25:37] <ckrough> redundant
[20-Aug-2009 09:25:54] <rmatte> yeh, I never thought of that (old habbits die hard)
[20-Aug-2009 09:25:58] <rmatte> testing now...
[20-Aug-2009 09:29:05] <rmatte> seems to
[20-Aug-2009 09:29:13] <rmatte> I guess it was redundant
[20-Aug-2009 09:29:33] <rmatte> I was just used to using it, I'll have to update my article on loading cisco Mibs to include that instead
[20-Aug-2009 09:30:02] <ckrough> cool
[20-Aug-2009 09:30:21] <ckrough> so, your writeup for the mibs is strictly to get all the traps recognized, right?
[20-Aug-2009 09:30:46] <rmatte> yeh, Mibs translate traps
[20-Aug-2009 09:30:51] <rmatte> Zenoss doesn't use them for anything else
[20-Aug-2009 09:32:07] <twm1010> a MIB walker to help design performance templates would be fantastic.
[20-Aug-2009 09:32:26] <ckrough> What do you mean
[20-Aug-2009 09:33:21] <twm1010> well... for instance in the SolarWinds universal device polar, you can browse the MIB database built-in, find the vendor, device, etc, and test the value you want to monitor, right there in the window, and then turn that into an assignable poller.
[20-Aug-2009 09:33:39] <twm1010> Zenoss could do the same thing, except you could just say "create a datasource" and choose which template to drop it into.
[20-Aug-2009 09:33:48] <rmatte> twm1010: install the Mib Browser ZenPack, it's sort of like that
[20-Aug-2009 09:33:49] <ckrough> that would be nice
[20-Aug-2009 09:33:55] <twm1010> i have it, can it do that?
[20-Aug-2009 09:34:00] <rmatte> well, in terms of browsing the Mib tree
[20-Aug-2009 09:34:07] <rmatte> you still need to add the OIDs to templates by hand
[20-Aug-2009 09:34:11] <rmatte> but that takes seconds to do anyways
[20-Aug-2009 09:34:24] <rmatte> the Mib browser window opens in a new window too, which is nice
[20-Aug-2009 09:34:37] <rmatte> so you can browse the Mibs tree while editing the template
[20-Aug-2009 09:35:04] <rmatte> I'm sure it's not as flashy as solarwinds' system though
[20-Aug-2009 09:35:22] <rmatte> but it runs on Linux, so it gets points for that
[20-Aug-2009 09:35:48] <twm1010> and it doens't cost you 20 grand a year either.
[20-Aug-2009 09:35:54] <jb> yes it does
[20-Aug-2009 09:35:55] <jb>
[20-Aug-2009 09:36:01] <ckrough> heh
[20-Aug-2009 09:36:02] <twm1010> err.. well SW isn't that much, its 20 initial, 3 a year
[20-Aug-2009 09:36:13] <rmatte> if you buy enterprise
[20-Aug-2009 09:36:13] <ckrough> Enterprise...
[20-Aug-2009 09:36:16] <jb> zenoss is the same amount every year.. its subscription based.
[20-Aug-2009 09:36:19] <jb> right.
[20-Aug-2009 09:36:28] <jb> the only thing that solarwinds excels at is reporting and network device monitoring..
[20-Aug-2009 09:36:35] <ckrough> and performance with large installs
[20-Aug-2009 09:36:40] <jb> yeah
[20-Aug-2009 09:36:46] <rmatte> Zenoss excels at network device monitoring as well in my opinion
[20-Aug-2009 09:36:50] <ckrough> we've outscaled RRD... just throwing hardware at it now
[20-Aug-2009 09:36:56] <rmatte> network device monitoring isn't crazy complex
[20-Aug-2009 09:36:59] <jb> rmatte: with customization, yeah
[20-Aug-2009 09:37:07] <twm1010> ckrough: are you using distributed collectors?
[20-Aug-2009 09:37:18] <jb> im looking to do a core installation to monitor 900 windows boxes
[20-Aug-2009 09:37:19] <rmatte> jb: well yeh, it's not one of these "ready to go out of the box" type things, but I actually prefer to customize it myself
[20-Aug-2009 09:37:23] <ckrough> yeah, last count I was at 26
[20-Aug-2009 09:37:33] <rmatte> more flexibility, so provided you spend the development time you can monitor literally anything
[20-Aug-2009 09:37:37] <twm1010> wow, 26 collectors? how many nodes?
[20-Aug-2009 09:37:40] <rmatte> even some obscure POS that no one has heard of before
[20-Aug-2009 09:37:40] <jb> you have 26 collectors?
[20-Aug-2009 09:38:11] <ckrough> thats total over 5 regions, so 5 installs.. largest is 10k devices, total is around 18k devices
[20-Aug-2009 09:38:25] <ckrough> just network gear
[20-Aug-2009 09:38:30] <ckrough> dont do much systems monitoring
[20-Aug-2009 09:38:30] <jb> nice
[20-Aug-2009 09:38:36] <twm1010> 5 regions 5 installs, 5 collectors at each?
[20-Aug-2009 09:38:57] <ckrough> 1 master at each, then sets of collectors. largest has 10 collectors
[20-Aug-2009 09:39:21] <ckrough> lots and lots of OIDs since Im grabbing info for each interface on each switch
[20-Aug-2009 09:39:29] <twm1010> i'd imagine so.
[20-Aug-2009 09:40:03] <twm1010> you on core or enterprise?
[20-Aug-2009 09:40:06] <ckrough> ent
[20-Aug-2009 09:40:40] <jb> ckrough: are you pleased with the ent distributed collectors?
[20-Aug-2009 09:40:44] <jb> i'll be throwing a couple up soon
[20-Aug-2009 09:40:54] <jb> do you have to maintain things such as zenpacks, etc on every collector
[20-Aug-2009 09:41:08] <ckrough> not in the later versions. the master pushes out just about everything
[20-Aug-2009 09:41:13] <jb> awesome
[20-Aug-2009 09:41:16] <ckrough> very
[20-Aug-2009 09:41:17] <jb> i was concerned about that
[20-Aug-2009 09:41:28] <jb> so whats your upgrade process like?
[20-Aug-2009 09:41:30] <ckrough> Used to do it that way. ClusterSSH ftw
[20-Aug-2009 09:41:36] <rmatte> yeh, that'd be nice, I have multiple core instances that I have to keep up to date by hand
[20-Aug-2009 09:41:36] <jb> yeah
[20-Aug-2009 09:41:39] <rmatte> it's fun fun
[20-Aug-2009 09:41:43] <rmatte> though I have scripts that help me
[20-Aug-2009 09:41:44] <rmatte> hehe
[20-Aug-2009 09:41:49] <ckrough> upgrade process involves Chet
[20-Aug-2009 09:41:56] <jb> thats what i thought
[20-Aug-2009 09:42:10] <twm1010> I am unforunately stuck with only my production server right now :|
[20-Aug-2009 09:42:17] <jb> oh, are you RS?
[20-Aug-2009 09:42:23] <ckrough> yeah, it's got to. Quite a few customisations rolled in now that get goofy on upgrades
[20-Aug-2009 09:42:30] <ckrough> yes
[20-Aug-2009 09:42:33] <jb> ok
[20-Aug-2009 09:42:52] <jb> where are you based out of if you dont mind me asking?
[20-Aug-2009 09:43:01] <ckrough> company or me? Im in San Antonio
[20-Aug-2009 09:43:05] <jb> ah ok
[20-Aug-2009 09:43:07] <rmatte> oh, rackspace, we monitor some servers that are hosted by you guys
[20-Aug-2009 09:43:13] <ckrough> oh yeah? nice
[20-Aug-2009 09:43:18] <rmatte> yup
[20-Aug-2009 09:43:30] <ckrough> rmatte: what region?
[20-Aug-2009 09:43:46] <rmatte> and some cisco gear looks like
[20-Aug-2009 09:43:52] <rmatte> looks like they are in europe
[20-Aug-2009 09:44:27] <rmatte> well, they are named rackspace
[20-Aug-2009 09:44:49] <ckrough> just availability I assume
[20-Aug-2009 09:44:56] <ckrough> /I hope/
[20-Aug-2009 09:44:59] <rmatte> I think this company we're monitoring has stuff hosted by you guys in the U.S too
[20-Aug-2009 09:45:08] <rmatte> probably
[20-Aug-2009 09:46:25] <jb> "{'firmware': '85P1'}"
[20-Aug-2009 09:46:41] <jb> whats the best way to just get "85P1" out of that?
[20-Aug-2009 09:46:52] <ckrough> as a string or as a tuple?
[20-Aug-2009 09:46:56] <jb> string
[20-Aug-2009 09:47:08] <jb> ... markeriv [n=markeriv@67.103.221.82] has joined #zenoss
[20-Aug-2009 09:47:10] <jb> oops
[20-Aug-2009 09:47:11] <ckrough> http://gskinner.com/RegExr/
[20-Aug-2009 09:47:14] <rmatte> ckrough: ah, it looks like they are actually hosted by one of your partners or something for you guys
[20-Aug-2009 09:47:14] <jb> om.firmware = firmware.get(om.snmpindex)
[20-Aug-2009 09:47:19] <jb> thats om.firmware
[20-Aug-2009 09:47:20] <rmatte> IXEUROPE
[20-Aug-2009 09:47:22] <markeriv> Greetings fellow Zenoss Junkies
[20-Aug-2009 09:47:49] <jb> ckrough: oh nice, ive never seen that
[20-Aug-2009 09:47:57] <twm1010> IX?
[20-Aug-2009 09:48:03] <ckrough> I love that site. I should send him a donation at this point
[20-Aug-2009 09:48:09] <twm1010> markeriv: welcome.
[20-Aug-2009 09:48:11] <_mingau_> hello all people
[20-Aug-2009 09:49:06] <jb> hmmm
[20-Aug-2009 09:49:36] * jb tries to figure it out
[20-Aug-2009 09:49:40] <rmatte> yeh, RegExr is awesome
[20-Aug-2009 09:50:16] <markeriv> Quick question, has anyone else had problems in 2.4.2 with rrd files not being generated?
[20-Aug-2009 09:50:31] <markeriv> I am having to manually go create em O.o
[20-Aug-2009 09:50:52] <chudler_> that sucks. It hasn't happened to me.
[20-Aug-2009 09:50:53] <ckrough> jb: match group 1 in : ^{'[\w]*':\s'([\w]*)'}$
[20-Aug-2009 09:51:11] <twm1010> markeriv: any trouble I had with that, was related to the datasource setup
[20-Aug-2009 09:51:18] <ckrough> jb: my regex foo is ugly thought, it could probably be more elegant, but that should be accurate
[20-Aug-2009 09:51:25] <markeriv> I would typically think it would be a permissions issue, but all but 1 of the rrd files was correctly populating on device install
[20-Aug-2009 09:51:26] <twm1010> for instance I was using incomplete OID's forgetting to put a .1 on the end
[20-Aug-2009 09:51:31] <jb> ckrough: how did you get that?
[20-Aug-2009 09:51:34] <jb> using the site
[20-Aug-2009 09:51:43] <_mingau_> does anybody use freebsd here?
[20-Aug-2009 09:51:52] <ckrough> jb: entered your string in the main window, then toyed with the match line up top
[20-Aug-2009 09:52:13] <rmatte> haha
[20-Aug-2009 09:52:17] <ckrough> jb: enter it then mouse over the regex and it will explain each portion, then mouse over the highlight match it will show the match groups
[20-Aug-2009 09:52:20] <rmatte> he tried to paste something to be a flooded off
[20-Aug-2009 09:52:26] <rmatte> _mingau_: pastebin.com
[20-Aug-2009 09:52:30] <rmatte> then paste the URL
[20-Aug-2009 09:52:31] <rmatte> much easier
[20-Aug-2009 09:52:36] <_mingau_> kkkkkkkkkk
[20-Aug-2009 09:52:49] <markeriv> hmm, i am using the zenpack ZenPacks.BlakeDrager.fping
[20-Aug-2009 09:53:15] <rmatte> wow, my fingers seems to be going faster than my brain this morning, that should have read: "he tried to paste something to me and flooded off"
[20-Aug-2009 09:53:18] <markeriv> so I have not edited any OIDs
[20-Aug-2009 09:53:26] <twm1010> markeriv: i would think then you need to look at debug events or try and execute the command used to collect the fping data manually, to see if it succeeds
[20-Aug-2009 09:53:56] <jb> hmm
[20-Aug-2009 09:54:07] <twm1010> generally if the RRD files is not created it's 1) a permissions issue or 2) the data is not being collected correctly
[20-Aug-2009 09:54:17] <markeriv> hmm
[20-Aug-2009 09:54:26] <markeriv> i got that folder set to -R 777
[20-Aug-2009 09:54:39] <jb> group 1: 85P1
[20-Aug-2009 09:54:49] <twm1010> what im saying, it is more likely to be a collection issue, and not a permissions issue
[20-Aug-2009 09:54:58] <markeriv> hmm ok
[20-Aug-2009 09:55:29] <markeriv> so check the OID of the fping_min.rrd
[20-Aug-2009 09:55:41] <twm1010> fping probably doesn't have an OID, its not an SNMP based datasource
[20-Aug-2009 09:55:42] <markeriv> thats the one that i cannot get to auto populate on install
[20-Aug-2009 09:55:48] <markeriv> doh
[20-Aug-2009 09:55:53] <markeriv> thats the one thats breaking lol
[20-Aug-2009 09:56:11] <markeriv> not having any similar issues with SNMP devices... just fping
[20-Aug-2009 09:56:12] <twm1010> well, did you assign the fping template to any devices?
[20-Aug-2009 09:56:34] <markeriv> aye, 3 at this point. 1 linux server (fedora) and 2 cisco 1721s
[20-Aug-2009 09:56:44] <twm1010> gotcha...
[20-Aug-2009 09:56:53] <twm1010> well.... look at that template, and see what command its trying to run
[20-Aug-2009 09:56:56] <markeriv> all utilization graphs and charts work correctly by default
[20-Aug-2009 09:56:59] <twm1010> login as the zenoss user, and try and run it yourself
[20-Aug-2009 09:56:59] <markeriv> nod
[20-Aug-2009 09:58:11] <markeriv> kk
[20-Aug-2009 10:01:01] <twm1010> let me start up my dev box, i think i installed the fping zenpack there
[20-Aug-2009 10:01:26] <mrayzenoss> mrchippy and I will be taking questions today, he should be here in a second
[20-Aug-2009 10:01:36] <twm1010> mrayzenoss: welcome.
[20-Aug-2009 10:01:42] <straterra> I added a process to monitor.. It's a Windows machine and the executable is JAMES.exe. I added that as the regex, and its running..but Zenoss says it isnt
[20-Aug-2009 10:01:45] <rmatte> oh right, developer day
[20-Aug-2009 10:01:50] <straterra> Oh, cool
[20-Aug-2009 10:01:57] <twm1010> straterra: how much time did you give it to find the process?
[20-Aug-2009 10:02:05] <straterra> About..8 hours
[20-Aug-2009 10:02:16] <twm1010> hrmm, that should be plenty, and SNMP is working fine to the server?
[20-Aug-2009 10:02:29] <straterra> yup
[20-Aug-2009 10:02:43] <markeriv> <twm1010> thanks a bunch for the help btw
[20-Aug-2009 10:02:48] <straterra> I'm sitting next to the machine..I see the process running
[20-Aug-2009 10:02:51] <twm1010> try changing your regex to.... ^JAMES and then remodel the device and see what happens
[20-Aug-2009 10:02:56] <straterra> no .exe?
[20-Aug-2009 10:03:08] <twm1010> give it a shot... see how it goes.
[20-Aug-2009 10:03:11] <straterra> ok
[20-Aug-2009 10:03:14] <_mingau_> people...
[20-Aug-2009 10:03:27] <_mingau_> does anybody use freebsd here?
[20-Aug-2009 10:03:58] <straterra> Indeed
[20-Aug-2009 10:06:05] <twm1010> markeriv: i had a thought, is fping installed? as well as it's corresponding zenpack?
[20-Aug-2009 10:06:13] <markeriv> yup
[20-Aug-2009 10:06:52] <markeriv> after I manually created the file the graphs increment correctly
[20-Aug-2009 10:07:33] <markeriv> manually created it by coping cp -b fping_max.rrd fping_min.rrd
[20-Aug-2009 10:08:40] <twm1010> i've never had to manually create an RRD file
[20-Aug-2009 10:08:53] <markeriv> *jealous*
[20-Aug-2009 10:09:05] <twm1010> well, you shouldn't have to.
[20-Aug-2009 10:09:09] <markeriv> ahh that was worth a good laugh
[20-Aug-2009 10:09:30] <twm1010> the collector has to put the data somewhere, so if its not putting it anywhere, its not collecting it
[20-Aug-2009 10:09:56] <markeriv> hmm, that makes sense...
[20-Aug-2009 10:10:26] <ckrough> I've never had to manually create on either.
[20-Aug-2009 10:11:50] <markeriv> hmm now that is very curious
[20-Aug-2009 10:12:13] <markeriv> when clicking on /Devices/rrdTemplates/b_fping/datasources/fping/datapoints/min
[20-Aug-2009 10:12:17] <markeriv> under the Zenpack
[20-Aug-2009 10:12:35] <markeriv> i get An error was encountered while publishing this resource. The requested resource does not exist.
[20-Aug-2009 10:12:46] <markeriv> so its creation file is missing
[20-Aug-2009 10:13:02] <markeriv> just adding the rrd file did not "fix the graph so to speak"
[20-Aug-2009 10:13:43] <twm1010> don't go under the zenpack
[20-Aug-2009 10:13:47] <twm1010> go to the template itself
[20-Aug-2009 10:14:12] <twm1010> see what command is being run to get the datasource min
[20-Aug-2009 10:14:43] <twm1010> it will probably something like scriptnamehere.py $zmanageIP or something like that
[20-Aug-2009 10:15:30] <markeriv> /usr/sbin/fping -c 3 -sq ${dev/manageIp} 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /' ?
[20-Aug-2009 10:16:41] <markeriv> hmm
[20-Aug-2009 10:16:52] <markeriv> when I use the Test against device feature i get good output
[20-Aug-2009 10:16:58] <rmatte> hmmm, there's no "RDP" IpService by default?
[20-Aug-2009 10:16:59] <rmatte> weird
[20-Aug-2009 10:18:51] <rmatte> hehe, I just found logs of the day I was building the Cisco Mibs and running in to problems
[20-Aug-2009 10:18:51] <rmatte> back in January
[20-Aug-2009 10:20:42] <rmatte> ah, it's MS WBT Server
[20-Aug-2009 10:22:13] <twm1010> rmatte: why would they monitor the port when they can monitor the service itself over WMI?
[20-Aug-2009 10:22:31] <twm1010> perhaps to see if a firewall is stopping things from listening.
[20-Aug-2009 10:22:31] <rmatte> twm1010: because we're not monitoring windows servers using WMI
[20-Aug-2009 10:22:44] <rmatte> and also, we're more interested in connectivity rather than the service
[20-Aug-2009 10:22:47] <twm1010> markeriv: that's a hell of a command
[20-Aug-2009 10:22:56] <rmatte> connectivity could be interrupted, but the service might show as good
[20-Aug-2009 10:23:09] <twm1010> login as zenoss user and execute /usr/sbin/fpin -c 3 -sq TESTIPHERE
[20-Aug-2009 10:23:15] <twm1010> and see if the data comes back
[20-Aug-2009 10:23:26] <twm1010> it looks like the script depends on both fping, and sed to work properly
[20-Aug-2009 10:24:03] <twm1010> well, thats not technically a script its just a massive command
[20-Aug-2009 10:24:22] <jb> 2009-08-20T10:19:13 ERROR RenderServer Failed to generate a graph
[20-Aug-2009 10:24:25] <jb> File "/opt/zenoss/Products/ZenRRD/RenderServer.py", line 190, in render
[20-Aug-2009 10:24:25] <jb> rrdtool.graph(*gopts)
[20-Aug-2009 10:24:26] <jb> hm
[20-Aug-2009 10:24:48] <twm1010> <-- out to lunch, back in an hour
[20-Aug-2009 10:25:03] <markeriv> good output to 4.2.2.1
[20-Aug-2009 10:25:55] <mrchippy> jb, is that the extent of the traceback, or is there more?
[20-Aug-2009 10:26:00] <jb> nothing more
[20-Aug-2009 10:26:04] <markeriv> using that command
[20-Aug-2009 10:26:25] <jb> yes, thats a traceback
[20-Aug-2009 10:26:34] <jb> any ideas?
[20-Aug-2009 10:27:07] <jb> oh.
[20-Aug-2009 10:27:54] <jb> debug saying the OIDs are bad..
[20-Aug-2009 10:33:41] <mrayzenoss> kinda a quiet day
[20-Aug-2009 10:34:07] <ckrough> yup
[20-Aug-2009 10:34:17] <ckrough> I guess all the issues are resolved. Time for beers/
[20-Aug-2009 10:34:22] <jb> erm, wonder why zen is telling me these OID's are bad..
[20-Aug-2009 10:34:22] <jb> .1.3.6.1.4.1.2.6.182.3.7.1.0 = INTEGER: 709
[20-Aug-2009 10:34:38] <rmatte> yeh it is
[20-Aug-2009 10:37:01] <jb> ok, i have a question..
[20-Aug-2009 10:37:19] <jb> I have events (SNMP) traps that come in..
[20-Aug-2009 10:37:42] <jb> with a summary of "Dell OpenManage: Memory Device Failure"
[20-Aug-2009 10:37:52] <jb> or "Dell OpenManage: Cooling device Failure"
[20-Aug-2009 10:38:12] <jb> but when they get emailed, the text gets mangled..
[20-Aug-2009 10:38:15] <jb> example: "Dell OpenManage: Cooling deviceFailure"
[20-Aug-2009 10:38:23] <jb> (no spaces)
[20-Aug-2009 10:38:29] <jb> Dell OpenManage: SystemRedundancy has been lost
[20-Aug-2009 10:38:36] <jb> anybody notice this?
[20-Aug-2009 10:42:09] <mrayzenoss> jb: is your OID message like this? Error reading value for "dhcp0" on planet-express (oid .1.3.6.1.4.1.63.501.3.3.1 is bad)
[20-Aug-2009 10:42:27] <jb> yeah, sorry, I fixed that.
[20-Aug-2009 10:42:31] <jb> it needed a .0
[20-Aug-2009 10:42:44] <mrayzenoss> yeah, that's what I was about to recommend
[20-Aug-2009 10:42:49] <rmatte> yeh, without the .0 you'll still get the output but Zenoss won't like it
[20-Aug-2009 10:43:07] <jb> yeah, i forgot about that
[20-Aug-2009 10:44:03] x-ip is now known as x-ip|launch
[20-Aug-2009 10:44:14] <jb> ah, another graph related question though..
[20-Aug-2009 10:45:50] <jb> i have one graph that is showing "0" for every value
[20-Aug-2009 10:45:57] <_mingau_> testing
[20-Aug-2009 10:46:05] <rmatte> jb: then that's what it's getting from snmp it would seem
[20-Aug-2009 10:46:11] <_mingau_> hello matt
[20-Aug-2009 10:46:13] <rmatte> jb: you sure you are polling the right OIDs?
[20-Aug-2009 10:46:17] <jb> yes, one would seem
[20-Aug-2009 10:46:21] <jb> it just randomly stopped working though
[20-Aug-2009 10:46:26] <mrayzenoss> _mingau_: greetings
[20-Aug-2009 10:46:28] <jb> one would think rather
[20-Aug-2009 10:46:36] <rmatte> jb: might have to restart snmp or something
[20-Aug-2009 10:46:39] <jb> sec.. i had an event generated from it
[20-Aug-2009 10:46:46] <jb> its not all graphs for that host, and the OID's are fine
[20-Aug-2009 10:47:21] <rmatte> jb: I've had it where CPU monitoring just died, but all other monitoring worked fine on a windows host before, so we had to restart snmp on it and then everything was fine
[20-Aug-2009 10:47:33] <magnachef> I have a question on something that's been a pain for a while now. If I have a device with multiple interfaces, that have multiple IPs that are bound to apache, lighttpd, and mysql, there are times when I can discover and add the device with our admin interface and IP, and it'll see the services and their associated IPs, but the IP Services check will fail unless I change it to an IP that's shared between the services
[20-Aug-2009 10:47:38] <jb> k, i'll try that
[20-Aug-2009 10:48:10] <magnachef> sometimes with lighttpd, http, and mysql, this doesn't work, and I need to drop monitoring of the mysql IP Service
[20-Aug-2009 10:48:10] <jb> g$ snmpwalk -v2c -c fcsnmp1ro fc-iftpcln01 1.3.6.1.4.1.8072.1.3.2.3.1.1.14.112.111.115.116.102.105.120.66
[20-Aug-2009 10:48:11] <jb> NET-SNMP-EXTEND-MIB::nsExtendOutput1Line."postfixBounced" = STRING: 253
[20-Aug-2009 10:48:18] <jb> just odd, cause i don't have any "bad OID" events
[20-Aug-2009 10:49:04] <magnachef> is there a way that I can add the device with my admin IP, have it find the various other IPs and their services, then have them test the services via those IPs and not the admin IP?
[20-Aug-2009 10:49:19] <rmatte> jb: you might have to kill your RRD files and let them regenerate
[20-Aug-2009 10:49:33] <jb> yeah, thats what I was thinking
[20-Aug-2009 10:49:40] <rmatte> it's worth a try
[20-Aug-2009 10:50:13] <jb> ok, did that.. we'll just wait a bit now
[20-Aug-2009 10:50:26] <rmatte> yup, 15-20 mins and we should know
[20-Aug-2009 11:06:00] <jb> so can any devs give me tips on how to troubleshoot object classes that wont compile?
[20-Aug-2009 11:11:22] <mrayzenoss> from mrchippy: magnachef: you can click on the service on the device, choose the ip for the service to checked on, then go back to the os tab, check the service, and lock the service.
[20-Aug-2009 11:13:25] <mrayzenoss> mrchippy's battery died, he should be back soon
[20-Aug-2009 11:16:16] <magnachef> let me find one with multiple IPs. So it'll only give me an option for choosing an IP if there's more than one IP?
[20-Aug-2009 11:16:54] <magnachef> hmmm...looks like it doesn't...it just lists the IPs
[20-Aug-2009 11:17:05] <magnachef> I'm still running 2.3.3 if that matters
[20-Aug-2009 11:20:32] <rmatte> does anyone know of a way to install the SNMP service on a windows 2000 box without a windows 2000 cd?
[20-Aug-2009 11:20:54] <jb> make the media available via the network?
[20-Aug-2009 11:21:33] <rmatte> I'm more looking for some download link to a package that I could use instead or something (we're trying to dig up a 2000 cd but haven't been able to find one so far, and I just want to get this done and over with lol)
[20-Aug-2009 11:29:22] <mrchippy> sorry, magnachef. disconnected for a bit there
[20-Aug-2009 11:36:23] <rmatte> mrchippy: recap...
[20-Aug-2009 11:36:24] <rmatte> [12:10pm] <magnachef> let me find one with multiple IPs. So it'll only give me an option for choosing an IP if there's more than one IP?
[20-Aug-2009 11:36:24] <rmatte> [12:10pm] <magnachef> hmmm...looks like it doesn't...it just lists the IPs
[20-Aug-2009 11:36:24] <rmatte> [12:11pm] <magnachef> I'm still running 2.3.3 if that matters
[20-Aug-2009 11:38:17] <mrchippy> thanks, rmatte. yeah, it's not even in 2.4. it has been added and will be available in King Crab.
[20-Aug-2009 11:39:11] <rmatte> yeh, King crab seems to have a lot of great enhancements, I can't wait
[20-Aug-2009 11:39:36] <rmatte> (I'm most looking forward to port status actually reflecting the current status of the port)
[20-Aug-2009 11:40:04] <jb> so can any devs give me tips on how to troubleshoot object classes that wont compile?
[20-Aug-2009 11:40:06] <jb>
[20-Aug-2009 11:41:06] <rmatte> jb: are you using something like eclipse to author your python?
[20-Aug-2009 11:41:32] <jb> no
[20-Aug-2009 11:41:35] <jb> im using vi
[20-Aug-2009 11:41:40] <jb>
[20-Aug-2009 11:41:44] <rmatte> you might try running it through an IDE to clean up any spacing problems first
[20-Aug-2009 11:41:45] <jb> eclipse.. hmm
[20-Aug-2009 11:41:52] <rmatte> python is very very particular about spacing
[20-Aug-2009 11:41:55] <jb> yeah
[20-Aug-2009 11:42:22] <rmatte> that reminds me, I need to install IDLE on here
[20-Aug-2009 11:42:31] <jb> i saw one reccomended on the forums
[20-Aug-2009 11:42:39] <jb> preferrably, i'd like one that i can forward over X11
[20-Aug-2009 11:42:42] <rmatte> idle is the simplest
[20-Aug-2009 11:42:48] <rmatte> and idle can run on X11
[20-Aug-2009 11:43:21] <jb> ah it's packaged with python
[20-Aug-2009 11:43:31] <rmatte> generally
[20-Aug-2009 11:44:30] <rmatte> eclipse is a bit more robust, but idle is good for beginners
[20-Aug-2009 11:44:48] <rmatte> and it's very lightweight
[20-Aug-2009 11:46:15] <rmatte> general rule, indents in python should be 4 spaces
[20-Aug-2009 11:46:21] <rmatte> no tabs
[20-Aug-2009 11:46:27] <jb> yeah
[20-Aug-2009 11:46:35] <rmatte> unless you're using something like zendmd, then it's single spaces
[20-Aug-2009 11:47:07] <jb> ok i'll take a look at idle
[20-Aug-2009 11:47:11] <rmatte> cool
[20-Aug-2009 11:47:53] <straterra> rmatte: I have server 2k disk
[20-Aug-2009 11:48:34] <rmatte> straterra: it's alright, I'll let the admin responsible for the server dig one up
[20-Aug-2009 11:48:41] <straterra> Alright
[20-Aug-2009 11:48:44] <rmatte> we're a microsoft partner so I'm sure we can dig one up somehow
[20-Aug-2009 11:48:50] <rmatte> we'll download it if need be
[20-Aug-2009 11:48:52] <straterra> I have pretty much all Windows disks..so..if you need anything..lemme know
[20-Aug-2009 11:49:00] <rmatte> cool
[20-Aug-2009 11:49:02] <rmatte> thanks
[20-Aug-2009 11:49:04] <straterra> Yup
[20-Aug-2009 11:49:19] <straterra> My work needs an 'admin nap time'
[20-Aug-2009 11:49:56] <rmatte> hehe
[20-Aug-2009 11:53:06] <straterra> I didnt get to sleep until after 6AM..and I was back in here at 9:30 AM
[20-Aug-2009 11:56:39] <rmatte> brutal
[20-Aug-2009 11:57:02] <rmatte> I've pulled a 25 hour shift here when we were setting up a Zenoss box for a large client
[20-Aug-2009 11:57:04] <rmatte> that was fun
[20-Aug-2009 11:57:48] <rmatte> It was our first production Zenoss box ever, 350 devices spread across 3 continents
[20-Aug-2009 11:58:05] <rmatte> and not much went right, or we would have been done in half the time
[20-Aug-2009 11:58:06] <rmatte> lol
[20-Aug-2009 12:01:09] <straterra> I was moving a VM across storage arrays
[20-Aug-2009 12:02:34] <rmatte> ah
[20-Aug-2009 12:02:58] <straterra> It's still going
[20-Aug-2009 12:03:08] <rmatte> brutal
[20-Aug-2009 12:03:14] <rmatte> big vm?
[20-Aug-2009 12:03:19] <straterra> Yeah
[20-Aug-2009 12:03:26] <straterra> I wish we had 10ge :/
[20-Aug-2009 12:03:33] <rmatte> hehe
[20-Aug-2009 12:03:38] <rmatte> yeh, I'd bet that'd help
[20-Aug-2009 12:03:57] <rmatte> too bad the storage arrays aren't linked with fiber or something
[20-Aug-2009 12:04:22] <straterra> I don't think it'd help..the problem is vmware
[20-Aug-2009 12:04:32] <rmatte> ah
[20-Aug-2009 12:04:33] <straterra> It's not even close to maxing out gigabit
[20-Aug-2009 12:04:41] <straterra> ESX is just..transfering very slowly
[20-Aug-2009 12:04:46] <rmatte> crappy
[20-Aug-2009 12:05:45] <straterra> Good news is, my new work phone came in
[20-Aug-2009 12:09:50] <markeriv> anyone here know much about sed?
[20-Aug-2009 12:11:00] <chudler_> I just wish the spam on the zenoss forum could be dealt with once and for all :-(
[20-Aug-2009 12:11:16] <markeriv> yea
i saw that phone offer come through the other day lol[20-Aug-2009 12:11:30] <mrayzenoss> I've been deleting and banning them
[20-Aug-2009 12:11:49] <mrayzenoss> we're switching to a new platform soon, so please be patient
[20-Aug-2009 12:12:06] <chudler_> thats good. They are persistent little twerps
[20-Aug-2009 12:13:05] <markeriv> thx mray
[20-Aug-2009 12:13:16] <mrayzenoss> I wonder if I could outsource some of this, I'll ask the sysadmins if we can crowdsource the ban and delete
[20-Aug-2009 12:13:27] <straterra> Whats the best way to monitor an SSL website on port 8443?
[20-Aug-2009 12:13:29] <mrayzenoss> I figure some of you guys are trustworthy
[20-Aug-2009 12:15:25] <straterra> A new ip service?
[20-Aug-2009 12:17:43] <mrayzenoss> so anytime you see spam, just PM me and I'll get it
[20-Aug-2009 12:18:24] <straterra> Erhm..I created a new IP service..but can't find where to set the port o.O
[20-Aug-2009 12:18:57] <straterra> oh..nevermind heh
[20-Aug-2009 12:24:39] <erin> Is it cool for a a newbie zenoss to ask a question about a zenpack here? and yea I've searched the forums, RTFM & googled it all...
[20-Aug-2009 12:24:48] <erin> I'm a DBA & not the sys admin
[20-Aug-2009 12:25:00] <erin> thought I speak unix not win
[20-Aug-2009 12:25:53] <mrayzenoss> yup, this is a place of Q&A
[20-Aug-2009 12:27:44] <erin> Ok the zenpack was installed by my sys admin. I find it in zenoss. I bind it to my device (mysql server). I configure it with mysql user/password. I even test the user/pwd connection from the zenoss server.
[20-Aug-2009 12:28:03] <erin> I see the MySQL - Select Statistics title & a broken image ...
[20-Aug-2009 12:28:23] <erin> I search error logs & there's no error about a bad MySQL connection ...
[20-Aug-2009 12:28:34] <erin> I might be looking in the wrong error log tho....
[20-Aug-2009 12:28:48] <erin> how to get a real graph?
[20-Aug-2009 12:30:31] <mrayzenoss> how long has it been monitoring?
[20-Aug-2009 12:30:44] <mrayzenoss> which MySQL ZenPack is installed?
[20-Aug-2009 12:31:03] <scp2004> Hi Guys.. I need some help ...
[20-Aug-2009 12:31:15] <erin> I had the default MySQLMonitoring ZenPack installed.
[20-Aug-2009 12:31:24] <erin> It's been monitoring for about a week ...
[20-Aug-2009 12:31:36] <erin> long enough for there to be data ...
[20-Aug-2009 12:31:38] <mrayzenoss> and no graphs for anything on it?
[20-Aug-2009 12:31:54] <erin> well none on the MySQL zenpack.
[20-Aug-2009 12:31:55] <scp2004> I need to import a Mikrotik Mib to Zenoss 2.4.2 . How ? I m using CentOS 5
[20-Aug-2009 12:32:07] <scp2004> [mrayzenoss] I need to import a Mikrotik Mib to Zenoss 2.4.2 . How ? I m using CentOS 5
[20-Aug-2009 12:32:30] <erin> I've got graphs on IO, Mem, CPU, Load, NumTasksGraph
[20-Aug-2009 12:32:43] <erin> they were all there before I had the mysql zenpack installed.
[20-Aug-2009 12:32:54] <mrayzenoss> erin: double check the mysql settings: http://www.zenoss.com/community/docs/zenoss-extended-monitoring/2.4.2/ch11s03.html
[20-Aug-2009 12:33:42] <erin> ok. I'll be back in a second
[20-Aug-2009 12:33:47] <erin> ok more like 5 min
[20-Aug-2009 12:34:19] <mrayzenoss> scp2004: why don't you install the mib-browser ZenPack, it makes it easy to load MIBs from the UI
[20-Aug-2009 12:34:28] <mrayzenoss> http://www.zenoss.com/community/projects/zenpacks/mib-browser
[20-Aug-2009 12:36:15] <scp2004> [mrayzenoss] thanks .. But i need to import just one mib .. I think its more easy i do this for a command like zenmibload, not ?
[20-Aug-2009 12:36:16] <erin> ok this is a stupid question but the UI sux big time here. WTF is the "page menu"??? where I'm gonna select more--> templates?
[20-Aug-2009 12:36:58] <mrayzenoss> scp2004: sure, the zenpack just makes it simple
[20-Aug-2009 12:37:12] <mrayzenoss> scp2004: look at the command zenmib
[20-Aug-2009 12:37:38] <mrayzenoss> erin: yeah, it's non-intuitive at times
[20-Aug-2009 12:37:56] <erin> at times? gee I wish zenoss would hire a UX person.
[20-Aug-2009 12:38:17] <scp2004> [mrayzenoss] im reading about this . Just dont know where put the mib file to import ... The guide says /$ZENHOME/Share.. Dont have this folder in zenoss Centos Installation
[20-Aug-2009 12:38:22] <erin> I'm lost wheres the page menu?
[20-Aug-2009 12:38:25] <mrayzenoss> erin: we have and the UI revamp is in-progress
[20-Aug-2009 12:38:35] <erin> yea!
[20-Aug-2009 12:38:36] <cgibbons> it's that little drop-down triangle you see on the upper left of a table
[20-Aug-2009 12:38:37] <mrayzenoss> erin: it's the down-arrow to the left of "Status"
[20-Aug-2009 12:38:48] <cgibbons> also known as Zenoss New User Question #2
[20-Aug-2009 12:39:03] <cgibbons> horrible horrible menu
[20-Aug-2009 12:39:11] <mrayzenoss> yeah, you should fix it
[20-Aug-2009 12:39:17] <ckrough> zing
[20-Aug-2009 12:39:32] <erin> uh what place? dashboard?
[20-Aug-2009 12:39:41] <erin> at the page for the device?
[20-Aug-2009 12:40:07] <mrayzenoss> erin: which template are you trying to change? The one for the device class or the actual device?
[20-Aug-2009 12:40:29] <erin> I'm following the instructions on getting MySQLMonitor to work
[20-Aug-2009 12:40:49] <erin> item #2 under 3.2 Zenoss: From the page menu, select More → zProperties.
[20-Aug-2009 12:41:10] <erin> shite found it.
[20-Aug-2009 12:41:16] <erin> damn you guys this ui sux
[20-Aug-2009 12:41:56] <mrayzenoss> erin: may I suggest http://downloads.sourceforge.net/zenoss/Getting_Started_with_Zenoss.pdf
[20-Aug-2009 12:42:34] <erin> I already did. ANd I've already followed the instructions but my graphs are broken images.
[20-Aug-2009 12:43:07] <rhett> there are plenty of worse ui's
[20-Aug-2009 12:43:30] <rhett> its not bad. especially after you've used it
[20-Aug-2009 12:44:19] <erin> yea I know but my frustration is with this product.
[20-Aug-2009 12:44:38] <erin> so I see MySQL in the templates that have been "bind to device".
[20-Aug-2009 12:44:52] <erin> and the headers for these are showing up with broken graphs.
[20-Aug-2009 12:45:24] <scp2004> [mrayzenoss] look : Download and copy these files into /opt/zenoss/share/mibs/site/ .. dont have this folder in my zenoss centOS installation
[20-Aug-2009 12:45:41] <erin> I really can follow directions. I was very excited to see the title & area for the graphs when I first did this last thurs.
[20-Aug-2009 12:46:34] <erin> but alas I still have broken graphs. suggestion of error log to dig into?
[20-Aug-2009 12:46:42] <rhett> well if there are broken graphs it maybe the datasources are not there or do not have data
[20-Aug-2009 12:46:43] <erin> could be how the zenpack was installed?
[20-Aug-2009 12:46:54] <rhett> which could be a permissions error
[20-Aug-2009 12:47:09] <erin> nope -- db has data & is there. I can connect from the zenoss server to the db using the zenoss user/pwd
[20-Aug-2009 12:47:15] <erin> that part was a walk in the park!
[20-Aug-2009 12:47:40] <cgibbons> erin: if you want to get a taste of what the new UI will look like, http://public-demo.zenoss.com/
[20-Aug-2009 12:47:57] <rhett> no events under the host then i assume
[20-Aug-2009 12:48:23] <erin> which log?
[20-Aug-2009 12:48:55] <erin> OH! event.log!
[20-Aug-2009 12:49:06] <rhett> the log on the box you're trying to get the stats from
[20-Aug-2009 12:49:11] <erin> I'm getting this error (not sure if it's my graph but prolly is!)
[20-Aug-2009 12:49:22] <erin> I'm on zenoss server & in the log dir.
[20-Aug-2009 12:49:26] <rhett> ok
[20-Aug-2009 12:49:33] <erin> here's the error:
[20-Aug-2009 12:49:34] <erin> 2009-08-20T10:50:28 ERROR RenderServer Failed to generate a graph
[20-Aug-2009 12:49:34] <erin> Traceback (most recent call last):
[20-Aug-2009 12:49:34] <erin> File "/usr/local/zenoss/zenoss/Products/ZenRRD/RenderServer.py", line 190, in render
[20-Aug-2009 12:49:35] <erin> rrdtool.graph(*gopts)
[20-Aug-2009 12:50:09] <rhett> ok well thats for sure a graph error
[20-Aug-2009 12:50:25] <erin> how to fix?
[20-Aug-2009 12:50:39] <rhett> are you running zenoss or zenoss-stack?
[20-Aug-2009 12:50:43] <erin> I wonder why I'm getting other graphs?? is it the zenpack for MySQLMonitor
[20-Aug-2009 12:50:44] <rhett> i assume zenoss
[20-Aug-2009 12:50:54] <rhett> hm i dont know
[20-Aug-2009 12:51:06] <rhett> ive never run across this personally
[20-Aug-2009 12:51:28] <erin> running ubuntu so we've got teh zenoss-stack
[20-Aug-2009 12:51:43] <erin> I hate ubuntu nothing goes smoothly!
[20-Aug-2009 12:51:50] <erin>
[20-Aug-2009 12:51:51] <rhett> do other graphs work for the same host?
[20-Aug-2009 12:51:55] <erin> yup
[20-Aug-2009 12:52:05] <erin> the default cpu, mem, blah all do
[20-Aug-2009 12:52:08] <rhett> ok so its for sure only the mysql stuff
[20-Aug-2009 12:52:18] <erin> yup
[20-Aug-2009 12:52:31] <erin> and no I didn't install the mysql stuff. I'm the "end user".
[20-Aug-2009 12:53:46] <erin> should I have the zenpack de-installed & re-installed?
[20-Aug-2009 12:53:54] <erin> or is there an ubuntu zenpack for MySQL?
[20-Aug-2009 12:54:03] <scp2004> [mrayzenoss] I dont have perf graphs ! I need it .. Its cause i dont have the Mikrotik Mib installed ?
[20-Aug-2009 12:54:22] <rhett> i was curious if that had been done or if they tried to remove the zenpack from the host and recreate the graphs
[20-Aug-2009 12:54:34] <rhett> that would probably be the first step erin
[20-Aug-2009 12:54:40] <rhett> do you know if any other mysql graphs work?
[20-Aug-2009 12:54:49] <scp2004> [rmatte] I dont have perf graphs ! I need it .. Its cause i dont have the Mikrotik Mib installed ?
[20-Aug-2009 12:55:09] <scp2004> [straterra] I dont have perf graphs ! I need it .. Its cause i dont have the Mikrotik Mib installed ?
[20-Aug-2009 12:55:23] <erin> HAHA! on the re-install. They find the ui & setup as daunting. It's my test!
[20-Aug-2009 12:55:33] <erin> I'll get it re-installed & check back
[20-Aug-2009 12:55:37] <rhett> cool
[20-Aug-2009 12:56:06] <rhett> scp2004 install the mib isnt too difficult. As stated earlier look at the zenmib command
[20-Aug-2009 12:56:28] <markeriv> hey rhett
[20-Aug-2009 12:56:34] <markeriv> what type of perf graphs you looking for?
[20-Aug-2009 12:56:38] <markeriv> and are you monitoring via SNMP?
[20-Aug-2009 12:56:42] <rhett> i.e. zenmib run ~/mikrotik.mib
[20-Aug-2009 12:56:43] <mrayzenoss> scp2004: trac-zenpacks/wiki/CommunityFAQGeneral#HowdoesZenossuseMIBs
[20-Aug-2009 12:56:48] <markeriv> err scp2004 i mean
[20-Aug-2009 12:56:53] <scp2004> [rhett] Yes ..I just installed the Mikrotik mib few minutes ago .. Its ok now .. But still without graphs .. I need to wait next traps ?
[20-Aug-2009 12:57:04] <mrayzenoss> scp2004: you won't use the MIB to monitor the device
[20-Aug-2009 12:57:13] <rhett> you need to create templates for that mib
[20-Aug-2009 12:57:25] <mrayzenoss> or really just from the OIDs of interest
[20-Aug-2009 12:57:25] <scp2004> [mrayzenoss] so ..why i dont have perf graphs ?
[20-Aug-2009 12:58:26] <mrayzenoss> is your Mikrotik device a router? Do you have it under /Devices/Network/Router?
[20-Aug-2009 12:58:38] <scp2004> yep
[20-Aug-2009 12:59:13] <scp2004> [mrayzenoss] I created the class /Devices/Network/Router/Mikrotik ..
[20-Aug-2009 12:59:27] <scp2004> [mrayzenoss] And put it them
[20-Aug-2009 13:00:32] <mrayzenoss> scp2004: and you have SNMP enabled and monitoring?
[20-Aug-2009 13:00:45] <scp2004> [mrayzenoss] yes .. snmp its ok ..
[20-Aug-2009 13:01:03] <scp2004> [mrayzenoss] are collecting ok .. Only dont have graphs
[20-Aug-2009 13:01:15] <mrayzenoss> broken graphs or nothing?
[20-Aug-2009 13:01:29] <Belial_> Hi there, Not sure if distributed collectors on core are supported, but have been following a few guides without much sucess. I was wondering how to point things to the zope server (In this paste, pthouzenoss is a CNAME for pthouap004 and is the 2nd collector & ptabzzenoss is the one with zope etc running) http://pastebin.com/d2a4d9345 I though changing the configuration files as pasted would point it, along with the zope.conf change, but no sucess
[20-Aug-2009 13:02:03] <mrayzenoss> scp2004: just checked, there are no perf graphs defined in /Devices/Network/Router
[20-Aug-2009 13:02:13] <Belial_> If anyone could point me in a direction it would be most appriciated!
[20-Aug-2009 13:02:41] <mrayzenoss> scp2004: at least not under the "Perf" tab
[20-Aug-2009 13:02:54] <mrayzenoss> scp2004: there are some if you go to the interfaces
[20-Aug-2009 13:03:19] <mrayzenoss> Belial_: distributed collectors aren't supported in Core, but there are a few community docs for implementing them
[20-Aug-2009 13:03:33] <mrayzenoss> Belial_: http://delicious.com/mray/collectors/
[20-Aug-2009 13:03:33] <scp2004> [mrayzenoss] so i need to habilite this ? where ? in /Devices /Network /Router /Mikrotik zproperties ?
[20-Aug-2009 13:03:53] <mrayzenoss> scp2004: on the OS tab, do you have interfaces listed?
[20-Aug-2009 13:04:02] <scp2004> [mrayzenoss] yes
[20-Aug-2009 13:04:26] <mrayzenoss> click on one, like "wlan0" or whatever
[20-Aug-2009 13:04:47] <Belial_> Thanks mrayzenoss, I've been over those links, but guess a 13th read wont hurt
Grats on the 1mil DL's and thanks for a great product![20-Aug-2009 13:05:10] <mrayzenoss> scp2004: you should see Throughput, Packets and Errors
[20-Aug-2009 13:05:34] <scp2004> [mrayzenoss] oh ..i can see the graphs ..
[20-Aug-2009 13:06:14] <mrayzenoss> so that's what you get with the vanilla Router device class
[20-Aug-2009 13:06:20] <rmatte> oh you guys actually hit mil?
[20-Aug-2009 13:06:27] <mrayzenoss> rmatte: yup
[20-Aug-2009 13:06:40] <scp2004> [mrayzenoss] but when i click just in /device perf i cant see anything
[20-Aug-2009 13:06:43] <mrayzenoss> rmatte: and put your name in the press release
[20-Aug-2009 13:06:46] <rmatte> nice, congrats
[20-Aug-2009 13:06:51] <rmatte> haha, thanks
[20-Aug-2009 13:07:01] <rmatte> is the press release on blog?
[20-Aug-2009 13:07:18] <mrayzenoss> rmatte: http://www.zenoss.com/news/archive/zenosspressrelease.2009-08-20.6091752351
[20-Aug-2009 13:07:55] <rmatte> thanks
[20-Aug-2009 13:07:56] <mrayzenoss> scp2004: there won't be anything on the perf tab unless you define some custom data sources for them
[20-Aug-2009 13:08:46] <scp2004> [mrayzenoss] please .how i do this ?
[20-Aug-2009 13:12:11] <erin> Ok during the remove of the zenpack we got errors! it was installed as ROOT & not zenoss. we reinstalled as zenoss.
[20-Aug-2009 13:12:36] <erin> and now that I've got the binding, setup all memorized it's ready to go.
[20-Aug-2009 13:14:10] <mrayzenoss> oops, something crashed the matrix
[20-Aug-2009 13:14:51] <mrayzenoss> scp2004 seems to have disappeared
[20-Aug-2009 13:15:03] <mrayzenoss> didn't want my link to the Admin Guide I guess
[20-Aug-2009 13:15:08] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch06s02.html
[20-Aug-2009 13:15:32] <rmatte> hehe
[20-Aug-2009 13:15:50] <rhett> he'll be back
[20-Aug-2009 13:16:02] <mrayzenoss> if there's anything that's gotten substantially better in the last year, it's the docs
[20-Aug-2009 13:19:51] <rmatte> yeh, the admin guide is definitely a worthwhile read for anyone starting out with Zenoss
[20-Aug-2009 13:34:05] <Belial_> Its a well thumbed through printout on my desk
[20-Aug-2009 13:34:20] <erin> Ok event.log says this:
[20-Aug-2009 13:34:31] <erin> 2009-08-20T11:10:56 ERROR zen.Plugins Could not load plugins from ZenPacks. One of the ZenPacks is missing or broken.
[20-Aug-2009 13:34:43] <erin> The zenpack was installed as zenoss & not root.
[20-Aug-2009 13:34:55] <rmatte> erin: which pack?
[20-Aug-2009 13:34:58] <erin> ZenPacks.zenoss.MySqlMonitor-2.0.0-py2.4.egg
[20-Aug-2009 13:35:08] <rmatte> hmmm
[20-Aug-2009 13:35:09] <erin> is this the correct version?
[20-Aug-2009 13:35:13] <erin> weird huh?
[20-Aug-2009 13:35:18] <rmatte> if it's the latest version then it's the correct one
[20-Aug-2009 13:35:21] <mrayzenoss> did it install?
[20-Aug-2009 13:35:24] <erin> yea
[20-Aug-2009 13:35:33] <rmatte> oh wait, did you restart Zenoss after the install?
[20-Aug-2009 13:35:40] <erin> oh no?
[20-Aug-2009 13:35:43] <rmatte> do that
[20-Aug-2009 13:35:49] <rmatte> some packs don't play nice until after a restart
[20-Aug-2009 13:35:53] <erin> how do I restart Zenoss & need to do it as zenoss?
[20-Aug-2009 13:36:05] <rmatte> as zenoss user just do "zenoss restart"
[20-Aug-2009 13:36:18] <rmatte> then do "zenoss status" to make sure all the daemons are running after the restart
[20-Aug-2009 13:36:23] <rmatte> if not then do zenoss start
[20-Aug-2009 13:36:46] <rmatte> I find that doing zenoss stop, waiting a few seconds, then doing zenoss start tends to work better
[20-Aug-2009 13:36:59] <erin> well now you tell me!
[20-Aug-2009 13:37:01] <erin>
[20-Aug-2009 13:37:04] <rmatte> hehe
[20-Aug-2009 13:38:23] <erin> yea yr right zenjobs not running & zenactions not running did not start
[20-Aug-2009 13:38:36] <ckrough> I do the same, I also add in the step of killing python processes that wont die after the stop
[20-Aug-2009 13:38:45] <erin> mysql used to be that way so I always do a stop/start
[20-Aug-2009 13:39:22] x-ip|launch is now known as x-ip
[20-Aug-2009 13:39:26] <rmatte> erin: yeh, I might just log a trac ticket to suggest that some sort of delay be put in to the restart command
[20-Aug-2009 13:39:36] <rmatte> so that it waits like 3 seconds before starting it back up
[20-Aug-2009 13:40:07] <erin> got this warning:
[20-Aug-2009 13:40:09] <erin> 2009-08-20T11:40:17 WARNING OFS.Application Duplicate Product name
[20-Aug-2009 13:40:09] <erin> After loading Product 'Five' from '/usr/local/zenoss/zenoss/Products',
[20-Aug-2009 13:40:10] <erin> I skipped the one in '/usr/local/zenoss/zenoss/lib/python/Products'.
[20-Aug-2009 13:40:15] <rmatte> that's fine
[20-Aug-2009 13:40:24] <rmatte> I get that too after upgrade
[20-Aug-2009 13:40:31] <rmatte> doesn't hurt anything
[20-Aug-2009 13:40:34] <erin> ok so now that I've done the restart do I have to go thru the rebind process again?
[20-Aug-2009 13:40:47] <rmatte> nah, hopefully that should fix it
[20-Aug-2009 13:40:56] <erin> yea!
[20-Aug-2009 13:41:07] <rmatte> some zenpacks show as broken until a restart
[20-Aug-2009 13:41:26] <erin> oh what does that mean then? gotta wait the 15 min or?
[20-Aug-2009 13:41:35] <rmatte> yeh
[20-Aug-2009 13:41:38] <rmatte> do you see graphs?
[20-Aug-2009 13:42:00] <erin> ah HA! it works!!!
[20-Aug-2009 13:42:02] <erin> yippee.
[20-Aug-2009 13:42:04] <rmatte> perfect
[20-Aug-2009 13:42:31] <erin> now I wonder if I can get it to monitor my infobright server? hmmm I'll let you know!
[20-Aug-2009 13:42:49] <erin> infobright is a datawarehouse built on mysql
[20-Aug-2009 13:42:49] <rmatte> cool
[20-Aug-2009 13:43:07] <rmatte> ah, opensource too
[20-Aug-2009 13:43:08] <rmatte> nice
[20-Aug-2009 13:43:38] <erin> oh wait? whats this error mean?
[20-Aug-2009 13:43:39] <erin> 2009-08-20T11:43:31 INFO ZODB.Conflict database conflict error (oid 0x2c, class Products.ZenUtils.PObjectCache.PObjectCache) at /zport/RenderServer/render (5 conflicts, of which 0 were unresolved, since startup at Thu Aug 20 11:40:21 2009)
[20-Aug-2009 13:43:52] <rmatte> lovely
[20-Aug-2009 13:43:59] <rmatte> zope database conflicts, those can get messy
[20-Aug-2009 13:44:06] <erin> whats that mean?
[20-Aug-2009 13:44:12] <rhett> ewe. thats gross
[20-Aug-2009 13:44:17] <rmatte> first off stop zenoss and tar up the $ZENHOME directory to back it up
[20-Aug-2009 13:44:22] <rmatte> and throw it somewhere else on the drive
[20-Aug-2009 13:44:24] <ckrough> Those can be safely ignored, so I've been told
[20-Aug-2009 13:44:32] <rmatte> oh?
[20-Aug-2009 13:44:37] <ckrough> "0 were unresolved", meaning all good
[20-Aug-2009 13:44:40] <rmatte> I thought you had to attempt a reindex to fix them?
[20-Aug-2009 13:44:43] <rmatte> oh right
[20-Aug-2009 13:44:49] <rmatte> it must have auto-corrected them
[20-Aug-2009 13:44:54] <rmatte> good stuff then
[20-Aug-2009 13:44:58] <erin> ok it was labeled info & not error. whew
[20-Aug-2009 13:45:02] <ckrough> I was told by someone at zenoss (chet?) that they are more informational than actual errors
[20-Aug-2009 13:45:14] <rmatte> yeh, seems that way
[20-Aug-2009 13:45:36] <erin> I'll go back to lurking until my mysql test is over & then flip to infobright & see if zenoss monitors it.
[20-Aug-2009 13:45:52] <erin> then I'll let you all know (as well as infobright!)
[20-Aug-2009 13:53:10] <chudler_> :Q
[20-Aug-2009 13:53:16] <chudler_> wrong window -- sorry!
[20-Aug-2009 13:53:28] <chudler_> plus a typo! bad day
[20-Aug-2009 13:53:40] <ckrough> yay vi
[20-Aug-2009 13:53:53] <ckrough> and the bad caps...
[20-Aug-2009 13:55:52] <erin> well getting data in the mysql graphs! yippee so its really working.
[20-Aug-2009 13:55:57] <erin> still have to test infobright
[20-Aug-2009 13:56:34] <rhett> cool
[20-Aug-2009 13:57:05] <rhett> i have a tendency to check the datasources for data even if i dont have graphs appearing
[20-Aug-2009 13:57:30] <rhett> just to try and troubleshoot where the problem may actually be
[20-Aug-2009 14:15:21] <markeriv> hey guys
[20-Aug-2009 14:15:29] <markeriv> i am still poking around a problem that I mentioned earlier
[20-Aug-2009 14:15:40] <markeriv> and was hoping now that a few different heads are in here that we might make some headway
[20-Aug-2009 14:15:46] <markeriv> i am running ver 2.4.2
[20-Aug-2009 14:15:57] <markeriv> and having problems with Zenpacks
[20-Aug-2009 14:16:14] <markeriv> I have tried multiple zenpacks and seem to have the issue with several types
[20-Aug-2009 14:16:24] <markeriv> where data is not collecting and I have missing rrd files
[20-Aug-2009 14:16:38] <markeriv> I have gone in and manually added the files, but the data still is not collecting.
[20-Aug-2009 14:16:56] <markeriv> when I go into the zenpack template and run the test it appears to give good #s.
[20-Aug-2009 14:16:58] <erin> markeriv -- you installed as zenoss & did a zenoss stop, zenoss start?
[20-Aug-2009 14:17:22] <erin> I had a zenpack installed as root so I got no graphs
[20-Aug-2009 14:17:38] <erin> for that pack only.
[20-Aug-2009 14:17:41] <markeriv> the zenpack? I believe so, I have checked and all zenpack folders and items are owned by zenoss
[20-Aug-2009 14:17:48] <erin> the zenpacks installed as zenoss worked.
[20-Aug-2009 14:17:51] <erin> ok
[20-Aug-2009 14:17:59] <erin> then you stopped & started the zenoss?
[20-Aug-2009 14:18:05] <markeriv> will give that a go here
[20-Aug-2009 14:18:17] <markeriv> have rebooted the machine a few times tho
[20-Aug-2009 14:18:19] <markeriv> would that not do it?
[20-Aug-2009 14:18:27] <erin> dunno I'm a newbie!
[20-Aug-2009 14:18:37] <erin> thats just what I did & it fixed my problem
[20-Aug-2009 14:18:53] <markeriv>
[20-Aug-2009 14:19:00] <markeriv> i am a newb too
[20-Aug-2009 14:20:59] <markeriv> still getting nan's
[20-Aug-2009 14:21:13] <erin> check the event.log what's it complaining about?
[20-Aug-2009 14:21:14] <rmatte> markeriv: it takes 3 polling cycles for the nans to dissapear
[20-Aug-2009 14:21:20] <rmatte> so like 15 mins
[20-Aug-2009 14:21:25] <markeriv> ahh
[20-Aug-2009 14:21:27] <markeriv> kewl
[20-Aug-2009 14:21:32] <markeriv> so now its time to chill and dance
[20-Aug-2009 14:21:35] <rmatte> so even if it is working, you won't know for a bit
[20-Aug-2009 14:21:39] <erin> yup
[20-Aug-2009 14:21:54] <rmatte> seeing "nan" is actually a good thing because it means that the RRD files have been created
[20-Aug-2009 14:22:04] <rhett> another reason why i fetch the datasources
[20-Aug-2009 14:22:09] <markeriv> well, i went and manually made an rdd file
[20-Aug-2009 14:22:13] <markeriv> with the proper name
[20-Aug-2009 14:22:16] <markeriv> and that problem went away
[20-Aug-2009 14:22:16] <rmatte> oh
[20-Aug-2009 14:22:26] <rmatte> I've never had to manually make one
[20-Aug-2009 14:22:37] <markeriv> yea, from what I have heard mine's a unique experience
[20-Aug-2009 14:22:43] <rmatte> apparently
[20-Aug-2009 14:22:48] <erin> haha! don't you just love those!
[20-Aug-2009 14:22:53] <markeriv> for some reason I always end up doing things the hard way
[20-Aug-2009 14:23:01] <markeriv> you shoulda seen me when I was installing gentoo last year
[20-Aug-2009 14:23:06] <erin> not me if I can help it.
[20-Aug-2009 14:23:08] <rmatte> haha
[20-Aug-2009 14:23:14] <rmatte> gentoo is always a fun learning experience
[20-Aug-2009 14:23:16] <erin> I like centos/redhat myself
[20-Aug-2009 14:23:21] <erin> suse is the worst!
[20-Aug-2009 14:23:25] <rmatte> hey, at least last year the gentoo installer was actually around
[20-Aug-2009 14:23:31] <markeriv> my mboard had a chip that was specifically not supported by the (at the time, latest linux version)
[20-Aug-2009 14:23:31] <rmatte> the first time I installed it there was no installer
[20-Aug-2009 14:23:36] <rmatte> you did the whole installation by hand
[20-Aug-2009 14:23:42] <rmatte> now THAT was fun
[20-Aug-2009 14:23:43] <rmatte>
[20-Aug-2009 14:23:44] <erin> ubuntu is a pain for mysql dbs
[20-Aug-2009 14:23:45] <markeriv> but I found a patch for a different linux version
[20-Aug-2009 14:23:58] <markeriv> so i took the patch apart
[20-Aug-2009 14:24:05] <markeriv> figured out which part of the kernel it was patching
[20-Aug-2009 14:24:09] <markeriv> and went in and did it manually
[20-Aug-2009 14:24:11] <markeriv> recompiled
[20-Aug-2009 14:24:16] <markeriv> and was able to listen to sound
[20-Aug-2009 14:24:17] <rmatte> you probably had an nforce chip?
[20-Aug-2009 14:24:19] <markeriv> aye
[20-Aug-2009 14:24:36] <markeriv> and then last weekend, i plugged in fedora 11...
[20-Aug-2009 14:24:43] <markeriv> boom everything worked
[20-Aug-2009 14:24:44] <rmatte> (they don't play nice with linux, or never used to anyways, the latest kernel versions have been great since it looks like they finally added support for them)
[20-Aug-2009 14:24:46] <erin> sounds like I'm surrounded by sys admin geeks!
[20-Aug-2009 14:24:55] <rmatte> I upgraded to ubuntu 9.04 and my system stopped whining
[20-Aug-2009 14:25:01] <markeriv> meh, i am network engineer who likes to tinker
[20-Aug-2009 14:25:15] <markeriv> yea ubuntu is where I jacked the patch for when i fixed the sound for gentoo
[20-Aug-2009 14:25:15] <rmatte> hehe
[20-Aug-2009 14:25:20] <erin> aah I'm a dba who did a side trip as a sys admin in the 90s.
[20-Aug-2009 14:25:22] <rmatte> yeh
[20-Aug-2009 14:25:29] <rmatte> I had that sound problem too
[20-Aug-2009 14:25:40] <jb> gentoo?
[20-Aug-2009 14:25:40] <jb> yuck.
[20-Aug-2009 14:25:50] <rmatte> I'm running gentoo on my home server
[20-Aug-2009 14:25:53] <markeriv> its actually fairly nice once you get used to it
[20-Aug-2009 14:25:57] <markeriv> i still have it
[20-Aug-2009 14:26:05] <rhett> i do everything
[20-Aug-2009 14:26:29] <rmatte> markeriv: was your mobo a M3N-HT Deluxe/Mempipe per chance?
[20-Aug-2009 14:26:35] <rhett> buntu variants or centos for me
[20-Aug-2009 14:26:42] <rmatte> (Asus)
[20-Aug-2009 14:27:08] <jb> im a rhel/cent guy
[20-Aug-2009 14:27:47] <rmatte> I have the M3N-HT and it did not play nice with Linux at all until Ubuntu 9.04
[20-Aug-2009 14:28:07] <rmatte> sound corruption issues (I actually went out and bought a soundcard because of it)
[20-Aug-2009 14:28:32] <erin> Why not use a mac for yr desktop & centos for dev?
[20-Aug-2009 14:28:33] <rmatte> I couldn't boot the system in multicore mode since it wouldn't boot and even if it did it would continuously beep from the PC speaker
[20-Aug-2009 14:28:53] <rmatte> fun times
[20-Aug-2009 14:29:15] <rmatte> plus the bios version that came with the board initially would only allow the system to boot right after the CMOS was reset
[20-Aug-2009 14:29:23] <rmatte> so I finally upgraded the BIOS to fix that problem
[20-Aug-2009 14:29:27] <rmatte> lol
[20-Aug-2009 14:29:34] <chudler_> I had a system like that once. Was an ABIT board
[20-Aug-2009 14:30:15] <chudler_> BIOS self-corrupted right after boot, so it also required a reset on each boot.
[20-Aug-2009 14:30:48] <rmatte> lovely
[20-Aug-2009 14:31:21] <erin> zenoss question? does the reset button refresh the graphs?
[20-Aug-2009 14:31:53] <scp2004> Someone to help me how i can see the perf graphs on devices ?
[20-Aug-2009 14:32:03] <magnachef> resets time frames, zooming, etc
[20-Aug-2009 14:32:06] <rmatte> erin: it refreshes your view of the graphs
[20-Aug-2009 14:32:13] <erin> cool!
[20-Aug-2009 14:32:14] <jb> rmatte: hm, looks like restarting snmp and deleting the RRD's did nothing :/
[20-Aug-2009 14:32:17] <scp2004> Snmp are running ok ...bur i cant see the graphs in perf
[20-Aug-2009 14:32:34] <mrayzenoss> everybody be nice to Gentoo, rocket worked on the installer
[20-Aug-2009 14:32:38] <erin> ok zenoss's MySQL Monitor is monitoring my Infobright instance!!!! yippeee!
[20-Aug-2009 14:32:38] <rmatte> jb: I'm stumped then, you're saying that when you walk the OIDs it is giving you the correct data right?
[20-Aug-2009 14:32:50] <jb> rmatte: yep
[20-Aug-2009 14:32:58] <mrayzenoss> scp2004: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch06s02.html
[20-Aug-2009 14:33:11] <rmatte> mrayzenoss: really? I'll have to give him kudos for that next time I see him, it's a nice installer
[20-Aug-2009 14:33:35] <rmatte> mrayzenoss: and having previously installed gentoo completely by hand, I'm quite appreciative of it
[20-Aug-2009 14:33:36] <scp2004> [mrayzenoss] One more time thank you vm man ..sr ...
..[20-Aug-2009 14:34:28] <mrayzenoss> yeah, he was rocket@gentoo.org for awhile (I think he's no longer active)
[20-Aug-2009 14:37:19] <markeriv> rmatte no it wasnt, was an ASUS nforc
[20-Aug-2009 14:37:20] <markeriv> nforce
[20-Aug-2009 14:37:28] <rmatte> nforce is the chipset
[20-Aug-2009 14:37:32] <rmatte> I meant the model of the board
[20-Aug-2009 14:37:36] <markeriv> ahh
[20-Aug-2009 14:37:39] <markeriv> hmm
[20-Aug-2009 14:37:47] <markeriv> how can i check my mboard ver in linux?
[20-Aug-2009 14:37:50] <markeriv> its the same one
[20-Aug-2009 14:37:58] <rmatte> does it looks like: http://computersrus.co.za/catalog/images/M3N-HT%20DeluxeMempipe.JPG
[20-Aug-2009 14:37:59] <rmatte> ?
[20-Aug-2009 14:38:13] <markeriv> no
[20-Aug-2009 14:38:20] <rmatte> ok, it's some other nforce board then
[20-Aug-2009 14:38:23] <markeriv> aye
[20-Aug-2009 14:38:25] <markeriv> yours looks newer
[20-Aug-2009 14:38:42] <rmatte> yeh, it was the newest one available when I bought it, now I'm sure there's newer out there
[20-Aug-2009 14:38:43] <markeriv> hey scp is back
[20-Aug-2009 14:38:47] <markeriv> hey SCP, i had a thought
[20-Aug-2009 14:38:55] <markeriv> under the device, under OS
[20-Aug-2009 14:38:59] <markeriv> do you see the interfaces?
[20-Aug-2009 14:39:03] <markeriv> if so click the interface name
[20-Aug-2009 14:39:15] <rmatte> mrayzenoss: that explains his awesome coding skills
[20-Aug-2009 14:39:15] <markeriv> one that has an IP address assigned to it
[20-Aug-2009 14:39:32] <markeriv> anyhow
[20-Aug-2009 14:40:10] <markeriv> hmm has it been 15 mins?
[20-Aug-2009 14:40:34] <rmatte> just about
[20-Aug-2009 14:40:38] <markeriv> hope not, still getting nans
[20-Aug-2009 14:40:50] <rmatte> you can always cheat and do zenperfsnmp run -d <devicename>
[20-Aug-2009 14:40:52] <markeriv> should I have removed the device first?
[20-Aug-2009 14:40:58] <rmatte> or zenperfsnmp run --cycle
[20-Aug-2009 14:41:00] <markeriv> am getting SNMP data just fine
[20-Aug-2009 14:41:07] <markeriv> and utilization ect
[20-Aug-2009 14:41:14] <markeriv> the only graphs i cant seem to get working are ping related
[20-Aug-2009 14:41:16] <rmatte> no, no need to remove the device
[20-Aug-2009 14:41:34] <markeriv> the company i work for monitors around 1000 network devices
[20-Aug-2009 14:41:35] <rmatte> ping related?
[20-Aug-2009 14:41:38] <rmatte> command based?
[20-Aug-2009 14:41:42] <markeriv> err fping
[20-Aug-2009 14:41:52] <rmatte> yeh, so command based
[20-Aug-2009 14:41:52] <markeriv> i need basic uptime over time data
[20-Aug-2009 14:42:04] <markeriv> like that device, 6 months ago was it up and online
[20-Aug-2009 14:42:13] <markeriv> i have to be able to prove or disprove SLAs
[20-Aug-2009 14:42:19] <rmatte> I see
[20-Aug-2009 14:42:31] <markeriv> so latency and jitter info is critical for for that
[20-Aug-2009 14:42:34] <markeriv> hence fping
[20-Aug-2009 14:42:36] <rmatte> why not just make an uptime graph?
[20-Aug-2009 14:42:47] <markeriv> would love that too, however I am as they say, a true newb
[20-Aug-2009 14:42:54] <rmatte> lol
[20-Aug-2009 14:42:59] <markeriv> so far parts of this zenpack are working
[20-Aug-2009 14:43:08] <markeriv> but the parts that dont make it look ugly
[20-Aug-2009 14:43:14] <rmatte> right
[20-Aug-2009 14:43:25] <markeriv> eventually this is gonna get put into a portal
[20-Aug-2009 14:43:35] <markeriv> where our customers should be able to view it dynamically
[20-Aug-2009 14:43:50] <rmatte> yeh, we're doing something similar with our graphed data eventually
[20-Aug-2009 14:44:02] <rmatte> you guys are an MSP?
[20-Aug-2009 14:44:05] <markeriv> anyhow, cause of that the graphs have to look pretty
[20-Aug-2009 14:44:10] <markeriv> and an ISP
[20-Aug-2009 14:44:15] <rmatte> ah, cool
[20-Aug-2009 14:44:26] <markeriv> looking to replace Silverback
[20-Aug-2009 14:44:43] <markeriv> microsofts former 10,000$ an update gig
[20-Aug-2009 14:44:50] <rmatte> if you haven't already, you should download the zenoss admin guide pdf and read it front to back
[20-Aug-2009 14:44:50] <markeriv> although i think they sold it to IBM
[20-Aug-2009 14:45:00] <rmatte> that's what I did when I was starting out, and it really opens your eyes
[20-Aug-2009 14:45:02] <markeriv> have, still some parts are confusing
[20-Aug-2009 14:45:03] <mrayzenoss> I think Dell bought Silverback
[20-Aug-2009 14:45:08] <rmatte> you'll have plenty of "wow, I can do that?" moments
[20-Aug-2009 14:45:11] <markeriv> might have been them, it got sold
[20-Aug-2009 14:45:19] <markeriv> hehe
[20-Aug-2009 14:45:32] <markeriv> i am half debating running 2.3.3
[20-Aug-2009 14:45:36] <mrayzenoss> http://www.silverbacktech.com/
[20-Aug-2009 14:45:45] <markeriv> right now I just have zenoss running on my pc
[20-Aug-2009 14:45:45] <mrayzenoss> well, 2.4.3 is scheduled for next week
[20-Aug-2009 14:45:50] <markeriv> O.o
[20-Aug-2009 14:45:59] <rmatte> I have 2.3.3 running on some servers here, waiting for 2.4.3 release to upgrade
[20-Aug-2009 14:46:02] <markeriv> but we have a server that should be arriving this week
[20-Aug-2009 14:46:20] <markeriv> is it easy to upgrade or is it better to do a fresh install
[20-Aug-2009 14:46:27] <markeriv> since I am going to be installing like next week
[20-Aug-2009 14:46:40] <mrayzenoss> fresh install is probably easiest
[20-Aug-2009 14:46:50] <mrayzenoss> discovery is usually pretty fast
[20-Aug-2009 14:46:56] <mrayzenoss> unless you have a lot of customization
[20-Aug-2009 14:47:14] <markeriv> well, unfortunately the network aint so clean
[20-Aug-2009 14:47:31] <markeriv> best i could hope for is some way to add devices off of a spreadsheet
[20-Aug-2009 14:47:32] <rmatte> yeh, we can't really do discovery ever with our clients
[20-Aug-2009 14:47:39] <rmatte> unless we want like a thousand workstations in inventory
[20-Aug-2009 14:47:46] <jb> question about disovery
[20-Aug-2009 14:47:48] <markeriv> cause there is a few thousand devices
[20-Aug-2009 14:47:54] <jb> i have an entire 10.0.0.0/8 network
[20-Aug-2009 14:47:56] <markeriv> and they are not easily identifable as a block
[20-Aug-2009 14:48:02] <jb> all the hosts I want to monitor end in a .15
[20-Aug-2009 14:48:18] <jb> is it possible to tell zenoss just to scan/add the .15's?
[20-Aug-2009 14:48:28] <jb> (there is about 900 servers)
[20-Aug-2009 14:48:33] <rmatte> jb: I believe you can do something like that in the add device wizard
[20-Aug-2009 14:48:35] <markeriv> nice network
[20-Aug-2009 14:48:39] <jb> haha
[20-Aug-2009 14:48:43] <jb> its a LEGACY network.
[20-Aug-2009 14:48:43] <rmatte> I know it allows you to do specific ranges and stuff
[20-Aug-2009 14:48:52] <jb> for remote locations
[20-Aug-2009 14:48:58] <mrayzenoss> yeah, but that's not a range
[20-Aug-2009 14:49:03] <jb> yeah its not
[20-Aug-2009 14:49:07] <rmatte> mrayzenoss: yeh, true
[20-Aug-2009 14:49:17] <rmatte> jb: you might have to write a script to do that
[20-Aug-2009 14:49:23] <mrayzenoss> I wonder if the code supports something like 10.0.0-255.15
[20-Aug-2009 14:49:43] <mrayzenoss> I doubt it works that way
[20-Aug-2009 14:49:56] <rmatte> mrayzenoss: I doubt it, that's a pretty specific case, I've never heard of anyone wanting to add all .15's or something
[20-Aug-2009 14:49:57] <mrayzenoss> probably easiest to do a bulk load
[20-Aug-2009 14:50:33] <rmatte> I'd just do an auto-discovery, then select all, and uncheck all the .15s, then delete the rest
[20-Aug-2009 14:50:53] <mrayzenoss> and then turn off auto-discovery
[20-Aug-2009 14:50:55] <rmatte> or write a script to do it
[20-Aug-2009 14:51:06] <jb> i have a feeling that auto discovery on an entire /8 would blow zenoss up
[20-Aug-2009 14:51:17] <rmatte> jb: it is pretty intensive
[20-Aug-2009 14:51:26] <jb> well..
[20-Aug-2009 14:51:30] <rmatte> jb: you should just write a bash script to do it or something
[20-Aug-2009 14:51:32] <jb> yeah
[20-Aug-2009 14:51:36] <jb> that would be the easiest
[20-Aug-2009 14:51:38] <mrayzenoss> yeah, you can tweak how many discoveries to run in parallel
[20-Aug-2009 14:51:51] <mrayzenoss> I know Chet messes with that
[20-Aug-2009 14:51:53] <jb> rmatte: you said you were working on CPU stats without snmp informant?
[20-Aug-2009 14:52:09] <rmatte> jb: yeh, I have it working, it's just not very clean
[20-Aug-2009 14:52:19] <rmatte> like clean in terms of the script being bash
[20-Aug-2009 14:52:24] <rmatte> when it really should be python
[20-Aug-2009 14:52:24] <rmatte> lol
[20-Aug-2009 14:52:35] <rmatte> I can send you the pack anyways if you'd like
[20-Aug-2009 14:52:38] <jb> out of the box with core.. i should be able to monitor things like windows services, os processes, filesystems, routing table, etc?
[20-Aug-2009 14:52:48] <mrayzenoss> yeah
[20-Aug-2009 14:52:48] <rmatte> jb: correct
[20-Aug-2009 14:52:55] <jb> and i should be able to monitor 900 windows devices?
[20-Aug-2009 14:52:57] <jb> on a beefy box?
[20-Aug-2009 14:53:00] <rmatte> jb: but not performance data, since Microsoft's snmp agent completely blows
[20-Aug-2009 14:53:02] <mrayzenoss> probably
[20-Aug-2009 14:53:09] <jb> (quad core at least, 32GB ram, raid10, etc)
[20-Aug-2009 14:53:16] <rmatte> my scripts allow performance data to be monitored
[20-Aug-2009 14:55:29] <markeriv> does microsoft have a product that doesnt blow?
[20-Aug-2009 14:55:40] <mrayzenoss> their mice are nice
[20-Aug-2009 14:55:42] <markeriv> you know... I am really starting to like this xchat program...
[20-Aug-2009 14:55:50] <rmatte> markeriv: there are a team of experts trying to find one, but so far, no luck
[20-Aug-2009 14:55:52] <markeriv> it says the word "microsoft" is mispelled
[20-Aug-2009 14:56:02] <markeriv> hahaha
[20-Aug-2009 14:56:05] <markeriv> nice rmatte
[20-Aug-2009 14:56:06] <mrayzenoss> I like my XBox
[20-Aug-2009 14:56:13] <markeriv> yea xboxs are ok
[20-Aug-2009 14:56:15] <markeriv> so are 360s
[20-Aug-2009 14:56:27] <markeriv> but i am trying to push our whole office Open source
[20-Aug-2009 14:56:30] <mrayzenoss> I'm just saying, they'd be a great company if they stuck to hardware and game consoles
[20-Aug-2009 14:56:36] <rmatte> I'm a ps3 guy myself
[20-Aug-2009 14:56:38] <markeriv> hard to get people to give up outlook
[20-Aug-2009 14:56:43] <ckrough> our office went to OpenOffice from MS Office
[20-Aug-2009 14:56:50] <markeriv> yea, only problem with that
[20-Aug-2009 14:56:55] <markeriv> is converged calendar and email
[20-Aug-2009 14:57:01] <markeriv> outlook has it and does it great
[20-Aug-2009 14:57:05] <ckrough> yeah
[20-Aug-2009 14:57:09] <ckrough> outlook is fairly pimp
[20-Aug-2009 14:57:11] <markeriv> so far every linux comparison does it poorly
[20-Aug-2009 14:57:15] <markeriv> its sad
[20-Aug-2009 14:57:39] <markeriv> any linux mail program that tries to compete seriously on business markets needs at least the ability to accept a outlook invitation
[20-Aug-2009 14:57:48] <markeriv> for a conference
[20-Aug-2009 14:57:51] <chudler_> I am not rabidly anti-MS, but exchange has failed to impress me. meh
[20-Aug-2009 14:58:01] <ckrough> client based email is dead. long live SaaS
[20-Aug-2009 14:58:08] <rmatte> actually evolution is quite good
[20-Aug-2009 14:58:16] <rmatte> it's almost as good as outlook
[20-Aug-2009 14:58:18] <mrayzenoss> yeah, I used that for a long time
[20-Aug-2009 14:58:51] <markeriv> i personally am liking gmail
[20-Aug-2009 14:58:58] <markeriv> i also work for mechaps.com
[20-Aug-2009 14:59:07] <markeriv> and they use google as a host for their domain
[20-Aug-2009 14:59:10] <markeriv> and it works very well
[20-Aug-2009 14:59:12] <rmatte> I do use Outlook here, but only because we're a microsoft shop for internal stuff
[20-Aug-2009 14:59:20] <rmatte> I use it inside of virtualbox
[20-Aug-2009 14:59:25] <markeriv> hehe
[20-Aug-2009 14:59:25] <mrayzenoss> Zenoss uses Zimbra
[20-Aug-2009 14:59:28] <markeriv> now that i need to get
[20-Aug-2009 14:59:35] <markeriv> a virtual box for outlook
[20-Aug-2009 14:59:38] <rmatte> with windows vista which is magically eating up 11 gigs of hard drive space, and I can't figure out how
[20-Aug-2009 14:59:39] <markeriv> if i could set that up
[20-Aug-2009 14:59:43] <rmatte> since the files on disk are less than that
[20-Aug-2009 14:59:55] <markeriv> i could theoretically convert my whole office
[20-Aug-2009 15:00:11] <markeriv> and have no one notice... well... not too much
[20-Aug-2009 15:00:23] <rmatte> markeriv: virtualbox is just like vmware, but free
[20-Aug-2009 15:00:28] <rmatte> very good
[20-Aug-2009 15:00:30] <markeriv> bah still getting nans
[20-Aug-2009 15:00:36] <rmatte> crappy
[20-Aug-2009 15:00:54] <rmatte> I'm out of ideas for you, are you sure you had to create the RRDs by hand?
[20-Aug-2009 15:00:58] <rmatte> that might be throwing it off
[20-Aug-2009 15:01:43] <markeriv> yea was no file
[20-Aug-2009 15:01:48] <markeriv> i can try adding a new site
[20-Aug-2009 15:03:29] <rhett> i've been thinking of setting up zimbra but i really only care about getting caldav working on my iphone
[20-Aug-2009 15:03:53] <rhett> cosmo (chandler server) and bedework both have issues w/ ical on the damn iphone from my experience
[20-Aug-2009 15:04:03] <rhett> stupid iphone
[20-Aug-2009 15:04:04] <mrayzenoss> markeriv: look at zencommand logging, maybe there's something in there?
[20-Aug-2009 15:04:39] <mrayzenoss> rhett: I use 14 calendars on my iPhone
[20-Aug-2009 15:05:00] <mrayzenoss> 5 cal dav from Zimbra, 1 ActiveSync from Zimbra
[20-Aug-2009 15:05:12] <mrayzenoss> and a bunch from Gooogle
[20-Aug-2009 15:05:58] <rhett> ah cool. the zimbra stuff work well over caldav?
[20-Aug-2009 15:06:29] <mrayzenoss> it's gotten a lot better
[20-Aug-2009 15:07:04] <mrayzenoss> we have occasional hiccups
[20-Aug-2009 15:07:17] <jb> ..
[20-Aug-2009 15:07:21] <rmatte> If anyone is interested in my Windows SNMP performance ZenPack, here it is: http://dmon.org/downloads/zenoss/ZenPacks.Nova.Windows.SNMPPerfMonitor-1.0-py2.4.egg
[20-Aug-2009 15:07:34] <rmatte> you need to have bc and snmpwalk installed for it to work
[20-Aug-2009 15:08:02] <rmatte> it also creates folders under /Servers/Windows
[20-Aug-2009 15:08:23] <rmatte> so you'll have like /Servers/Windows/1 Processor, 2 Processors, 4 Processors, 8 Processors, and 16 Processors
[20-Aug-2009 15:08:43] <rmatte> you need to do: snmpwalk -v1 -c CommunityString <DeviceName> hrProcessorLoad
[20-Aug-2009 15:08:51] <rmatte> then count how many lines, that tell you which group it should go in
[20-Aug-2009 15:09:02] <rmatte> so if you see 4 lines, drop it in to the 4 processors group
[20-Aug-2009 15:09:44] <rmatte> like I said, it's really rough right now, but it does work
[20-Aug-2009 15:10:01] <rmatte> I'm using it to monitor over 70 servers on one Zenoss instance
[20-Aug-2009 15:11:02] <rhett> thats good to hear though mrayzenoss. cosmo seems to have some odd auth issue w/ the long hashed url's and bedework requires a trailing slash on the url which the iphone removes
[20-Aug-2009 15:12:06] <markeriv> trying adding a new device on the off chance that my adding the rdd file manually screwed it up
[20-Aug-2009 15:14:27] <markeriv> yea by default it is still missing
[20-Aug-2009 15:14:28] <markeriv> grr
[20-Aug-2009 15:14:41] <mrayzenoss> is the template bound?
[20-Aug-2009 15:15:33] <markeriv> the only line that references it in zencommand is this one
[20-Aug-2009 15:15:34] <markeriv> 2009-08-18 09:54:27 INFO zen.zencommand: Deleting command Emser Tile Oakbrook from check_ping -H Emser Tile Oakbrook -w 180,100% -c 300,100% | sed -e 's# - Packet loss#|LOSS#;s#%,##;s#ms##;s# = #=#g'
[20-Aug-2009 15:15:44] <markeriv> aye the graphs are physically there
[20-Aug-2009 15:15:52] <markeriv> i just get nans and missing rrd files
[20-Aug-2009 15:15:59] <markeriv> i bound the template to devices/routers/cisco
[20-Aug-2009 15:16:20] <markeriv> and that aye is for the template question mray, they are bound
[20-Aug-2009 15:16:58] <mrayzenoss> I thought you'd said you had manually added the RRD files, which is odd
[20-Aug-2009 15:17:51] <rmatte> he manually created them I believe he said
[20-Aug-2009 15:17:59] <rmatte> not sure what method was used?
[20-Aug-2009 15:18:14] <rmatte> did you just do something like "touch RRD-FILE-NAME.rrd"
[20-Aug-2009 15:18:15] <rmatte> ?
[20-Aug-2009 15:18:30] <markeriv> i copied the max file
[20-Aug-2009 15:18:41] <rmatte> oh
[20-Aug-2009 15:18:47] <markeriv> and renamed it min
[20-Aug-2009 15:19:10] <rmatte> hmmm
[20-Aug-2009 15:19:19] <markeriv> the file that isnt populating is fping_min
[20-Aug-2009 15:19:33] <rmatte> quite honestly, if one of my Zenoss instances got that messed up I'd be reinstalling lol
[20-Aug-2009 15:19:47] <rmatte> you obviously shouldn't have to create the file by hand
[20-Aug-2009 15:19:56] <rmatte> and if it's not doing it automatically, that's the real issue
[20-Aug-2009 15:19:57] <mrayzenoss> I guess I wasn't paying close enough attention, but what else is hosed?
[20-Aug-2009 15:20:13] <markeriv> just ping graphs
[20-Aug-2009 15:20:14] <markeriv> oie
[20-Aug-2009 15:20:19] <markeriv> 2nd reinstall already
[20-Aug-2009 15:20:23] <rmatte> ah
[20-Aug-2009 15:20:26] <markeriv> here is the confusing bit
[20-Aug-2009 15:20:31] <rmatte> hmmm, maybe it's the pack then
[20-Aug-2009 15:20:34] <markeriv> Executing command
[20-Aug-2009 15:20:34] <markeriv> /usr/sbin/fping -c 3 -sq 69.3.230.36 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /'
[20-Aug-2009 15:20:34] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:20:34] <markeriv> PING OK|xmt=3 rcv=3 loss=0, min/avg/max =min= 115 avg=115 max=115
[20-Aug-2009 15:20:35] <markeriv> DONE in 2 seconds
[20-Aug-2009 15:20:47] <markeriv> it appears to work when I run the command from the template
[20-Aug-2009 15:20:54] <markeriv> and it even pulls data
[20-Aug-2009 15:20:54] <rmatte> are those datapoints present under the datasource?
[20-Aug-2009 15:21:08] <markeriv> avg GAUGE
[20-Aug-2009 15:21:08] <markeriv> loss GAUGE
[20-Aug-2009 15:21:08] <markeriv> max GAUGE
[20-Aug-2009 15:21:08] <markeriv> min GAUGE
[20-Aug-2009 15:21:08] <markeriv> rcv GAUGE
[20-Aug-2009 15:21:09] <markeriv> xmt GAUGE
[20-Aug-2009 15:21:12] <markeriv> those are my data points
[20-Aug-2009 15:21:23] <rmatte> "min/avg/max =min= 115"
[20-Aug-2009 15:21:29] <rmatte> odd
[20-Aug-2009 15:21:38] <markeriv> 115ms for 3 pings
[20-Aug-2009 15:21:40] <rhett> yeah
[20-Aug-2009 15:21:49] <rhett> that seems a bit odd
[20-Aug-2009 15:21:59] <rhett> but possible
[20-Aug-2009 15:22:00] <rmatte> yeh, I just don't know why it's outputting it like that
[20-Aug-2009 15:22:01] <rmatte> bizarre
[20-Aug-2009 15:22:12] <markeriv> hmm i am noticing a little bit more jitter than that is suggesting
[20-Aug-2009 15:22:18] <rmatte> 115ms for 3 pings is plausible
[20-Aug-2009 15:22:23] <markeriv> enough that i am disbelieving 3 115s
[20-Aug-2009 15:22:40] <rhett> bump it up to 10 15 or 20 pings and see what happens
[20-Aug-2009 15:22:44] <markeriv> Executing command
[20-Aug-2009 15:22:44] <markeriv> /usr/sbin/fping -c 3 -sq 69.3.230.36 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /'
[20-Aug-2009 15:22:44] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:22:44] <markeriv> PING OK|xmt=3 rcv=3 loss=0, min/avg/max =min= 114 avg=117 max=120
[20-Aug-2009 15:22:47] <markeriv> DONE in 2 seconds
[20-Aug-2009 15:22:49] <markeriv> meh
[20-Aug-2009 15:22:51] <markeriv> that time was more regular
[20-Aug-2009 15:23:00] <rmatte> yeh
[20-Aug-2009 15:23:21] <rmatte> it's just this: min/avg/max =min= 114
[20-Aug-2009 15:23:24] <rmatte> that I'm wondering about
[20-Aug-2009 15:23:46] <rmatte> the output should be more like: PING OK|xmt=3 rcv=3 loss=0 min=114 avg=117 max=120
[20-Aug-2009 15:23:48] <rmatte> from my experience
[20-Aug-2009 15:24:19] <mrayzenoss> yeah, that's what I'm thinking
[20-Aug-2009 15:24:54] <markeriv> here is the command itself
[20-Aug-2009 15:24:59] <markeriv> should I delete the = at the end?
[20-Aug-2009 15:24:59] <markeriv> /usr/sbin/fping -c 3 -sq ${dev/manageIp} 2>&1 | /bin/sed 's/\([0-9]*\)\/\([0-9]*\)\/\([0-9]\)%/xmt=\1 rcv=\2 loss=\3/;s/\([0-9]*.[0-9]*\)\/\([0-9].*\)\/\([0-9].*\)/min=\1 avg=\2 max=\3/' | /bin/sed -r '/targets/,/real time/d;s/[0-9]*.[0-9]*.[0-9]*.[0-9]* : xmt\/rcv\/%loss = /PING OK|/;s/, min\/avg\/max = / /'
[20-Aug-2009 15:25:14] * mrayzenoss eyes glazed over immediately
[20-Aug-2009 15:25:19] <markeriv> uhh oh
[20-Aug-2009 15:25:22] <rmatte> that's quite the command
[20-Aug-2009 15:25:25] <markeriv> yea
[20-Aug-2009 15:25:27] <markeriv> i didnt build it
[20-Aug-2009 15:25:51] <rmatte> I'd get rid of "min\/avg\/max ="
[20-Aug-2009 15:25:53] <rmatte> and then try it
[20-Aug-2009 15:25:56] <markeriv> blame BlakeDrager
[20-Aug-2009 15:26:01] <rmatte> it seems to be inserting that as plain text
[20-Aug-2009 15:26:26] <markeriv> min\/avg\/max =
[20-Aug-2009 15:26:30] <markeriv> pasted to preserve
[20-Aug-2009 15:26:35] <rmatte> yup
[20-Aug-2009 15:26:46] <rmatte> well, you have the whole command pasted to preserve
[20-Aug-2009 15:26:57] <markeriv> PING OK|xmt=3 rcv=3 loss=0 min/avg/max =min= 116 avg=119 max=122
[20-Aug-2009 15:27:05] <markeriv> with that line changed
[20-Aug-2009 15:27:08] <rmatte> so it didn't change anything
[20-Aug-2009 15:27:08] <markeriv> hmm
[20-Aug-2009 15:27:10] <markeriv> ya
[20-Aug-2009 15:27:38] <rmatte> that sed makes my eyes bleed
[20-Aug-2009 15:27:51] <markeriv> lol
[20-Aug-2009 15:28:00] <markeriv> think we could rewrite it, or would that be paid time?
[20-Aug-2009 15:28:09] <rmatte> but yeh, you need to mess around with it until it looks like: PING OK|xmt=3 rcv=3 loss=0 min=114 avg=117 max=120
[20-Aug-2009 15:28:29] <rmatte> since zenoss is probably tripping up on it because of the extra junk
[20-Aug-2009 15:28:40] <rhett> against Cypress
[20-Aug-2009 15:28:40] <rhett> PING OK|loss=0 min=0.141 avg=0.181 max=0.215 mdev=0.027
[20-Aug-2009 15:28:40] <rhett> DONE in 4 seconds
[20-Aug-2009 15:28:50] <rhett> well that didnt work
[20-Aug-2009 15:28:55] <rmatte> lol
[20-Aug-2009 15:28:58] <rhett> /bin/ping -i .5 -n -q -c 10 Cypress | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:29:00] <rhett> there
[20-Aug-2009 15:29:22] <rhett> thats what I use
[20-Aug-2009 15:29:34] <markeriv> should that be fping or just ping?
[20-Aug-2009 15:30:11] <rmatte> markeriv: doesn't matter...
[20-Aug-2009 15:30:20] <rmatte> fping allows you to ping multiple hosts in parallel
[20-Aug-2009 15:30:30] <rmatte> but you're not doing that, you're just pinging a single host per command
[20-Aug-2009 15:30:38] <markeriv> ahh but i am going to be
[20-Aug-2009 15:30:46] <markeriv> got 1000 or so device to track >.>
[20-Aug-2009 15:30:53] <markeriv> meh
[20-Aug-2009 15:30:55] <rmatte> yes, but Zenoss is going to do that for you
[20-Aug-2009 15:30:56] <markeriv> that gave me zenoss error
[20-Aug-2009 15:30:58] <markeriv> ahh nod
[20-Aug-2009 15:31:19] <rmatte> fping just lets you be like ping <host1> <host2> <host3> or whatever
[20-Aug-2009 15:31:24] <rmatte> fping rather
[20-Aug-2009 15:31:30] <rmatte> but you don't need that
[20-Aug-2009 15:31:33] <markeriv> ahh nod
[20-Aug-2009 15:31:48] <rmatte> the command rhett pasted need different datapoints
[20-Aug-2009 15:31:53] <rmatte> they need to match the output
[20-Aug-2009 15:31:53] <rhett> yes
[20-Aug-2009 15:32:03] <markeriv> yikes
[20-Aug-2009 15:32:04] <markeriv> ok
[20-Aug-2009 15:32:10] <rmatte> so loss min avg max and mdev
[20-Aug-2009 15:32:24] <rmatte> you also need to adjust your graph to add the new datapoints
[20-Aug-2009 15:32:26] <rmatte> then test it
[20-Aug-2009 15:32:36] <markeriv> type is gauge?
[20-Aug-2009 15:32:40] <rhett> avg GAUGE
[20-Aug-2009 15:32:40] <rhett> loss GAUGE
[20-Aug-2009 15:32:40] <rhett> max GAUGE
[20-Aug-2009 15:32:40] <rhett> mdev GAUGE
[20-Aug-2009 15:32:40] <rhett> min GAUGE
[20-Aug-2009 15:32:42] <rhett> yes
[20-Aug-2009 15:32:44] <rmatte> correct
[20-Aug-2009 15:33:13] <markeriv> hmm
[20-Aug-2009 15:33:16] <markeriv> added that one data point
[20-Aug-2009 15:33:21] <markeriv> still getting an error when i test
[20-Aug-2009 15:33:49] <markeriv> /bin/ping -i .5 -n -q -c 10 ${dev/manageIp} | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:33:53] <markeriv> thats what I am running
[20-Aug-2009 15:33:57] <rmatte> lol, my friend's IM status is "The artist formerly known as #!/bin/sh"
[20-Aug-2009 15:34:03] <markeriv> lol
[20-Aug-2009 15:34:05] <rhett> /bin/ping -i .5 -n -q -c 10 $devname | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$$/ {;N;s/\n//}'
[20-Aug-2009 15:34:11] <rhett> thats mine
[20-Aug-2009 15:34:36] <rmatte> same thing
[20-Aug-2009 15:34:47] <markeriv> well
[20-Aug-2009 15:34:51] <markeriv> aint pumping out the same error
[20-Aug-2009 15:34:52] <kisielk> hm, is there a faster way to access custom event views?
[20-Aug-2009 15:34:55] <markeriv> now its this one
[20-Aug-2009 15:34:58] <kisielk> than going through Settings/Users/ etc
[20-Aug-2009 15:35:04] <markeriv> Executing command
[20-Aug-2009 15:35:04] <markeriv> /bin/ping -i .5 -n -q -c 10 Emser Tile Louisville | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:35:04] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:35:04] <markeriv> ping: unknown host Emser
[20-Aug-2009 15:35:05] <markeriv> DONE in 0 seconds
[20-Aug-2009 15:35:07] <kisielk> ideally, is there a way to add one to the dashboard?
[20-Aug-2009 15:35:08] <markeriv> least thats better
[20-Aug-2009 15:35:15] <markeriv> now its just not using its own internal dns check
[20-Aug-2009 15:35:17] <rmatte> kisielk: unfortunately, no
[20-Aug-2009 15:35:19] <markeriv> will try by IP
[20-Aug-2009 15:35:30] <rmatte> kisielk: bookmark it
[20-Aug-2009 15:35:30] <kisielk> damn, that sucks
[20-Aug-2009 15:35:39] <kisielk> how hard would it be for them to be dashboardable
[20-Aug-2009 15:35:43] <rmatte> map it to a hotkey sequence or something
[20-Aug-2009 15:35:47] <kisielk> it would make the feature infinitely more useful
[20-Aug-2009 15:35:49] <markeriv> nope, puts the name back in there
[20-Aug-2009 15:36:26] <mrayzenoss> kisielk: http://public-demo.zenoss.com/evconsole
[20-Aug-2009 15:36:29] <markeriv> looks like it is not converting the device name back to its IP address
[20-Aug-2009 15:36:42] <rmatte> kisielk: the new event console has filters for everything
[20-Aug-2009 15:36:48] <mrayzenoss> and saved views
[20-Aug-2009 15:36:48] <rmatte> so it's going to effectively replace those anyways
[20-Aug-2009 15:36:54] <kisielk> right
[20-Aug-2009 15:37:16] <kisielk> how far are w from seeing that?
[20-Aug-2009 15:37:17] <rhett> markeriv yeah. from the error you posted thats the problem
[20-Aug-2009 15:37:24] <mrayzenoss> beta is probably 3 weeks out
[20-Aug-2009 15:37:38] <rmatte> kisielk: final release is due around October
[20-Aug-2009 15:37:41] <rmatte> so it's not that far off
[20-Aug-2009 15:37:45] <kisielk> nice
[20-Aug-2009 15:37:50] <rmatte> King Crab is going to be awesome
[20-Aug-2009 15:38:01] <mrayzenoss> markeriv: in the datasource, what's the command line?
[20-Aug-2009 15:38:03] * rmatte rubs his hands together
[20-Aug-2009 15:38:22] <rhett> you cant go much further w/ the crab names after king crab
[20-Aug-2009 15:38:46] <markeriv> you mean the command you just had me enter?
[20-Aug-2009 15:38:46] <mrayzenoss> rhett: http://en.wikipedia.org/wiki/Crab
[20-Aug-2009 15:38:50] <markeriv> /bin/ping -i .5 -n -q -c 10 $devname | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$$/ {;N;s/\n//}'
[20-Aug-2009 15:38:52] <markeriv> or elsewhere?
[20-Aug-2009 15:39:17] <mrayzenoss> replace $devname with ${here/manageIp}
[20-Aug-2009 15:39:57] <markeriv> Executing command
[20-Aug-2009 15:39:57] <markeriv> /bin/ping -i .5 -n -q -c 10 69.3.230.36 | sed -n -e '/loss/p' -e '/rtt/p' | sed -e 's/.*\([0-9].*.% packet loss\).*/PING OK|loss=\1/;s/[0-9] received//;s/%//;s/, //;s/packet loss//' -e 's/rtt min\/avg\/max\/mdev = /min=/;s/ ms//;s/\// avg=/;s/\// max=/;s/\// mdev=/' | sed '/$/ {;N;s/\n//}'
[20-Aug-2009 15:39:57] <markeriv> against Emser Tile Louisville
[20-Aug-2009 15:39:57] <markeriv> PING OK|loss=0 min=116.007 avg=118.862 max=125.374 mdev=2.845
[20-Aug-2009 15:39:58] <markeriv> DONE in 4 seconds
[20-Aug-2009 15:39:58] <rmatte> mrayzenoss: hehe
[20-Aug-2009 15:40:00] <markeriv> wootah
[20-Aug-2009 15:40:06] <rhett> nice
[20-Aug-2009 15:40:06] <rmatte> rhett: next one will be stone crab
[20-Aug-2009 15:40:20] <markeriv> stop talking about crabs, you are making me hungry
[20-Aug-2009 15:40:26] <rmatte> omg, please name the next one after that "Sally Lightfoot Crab"
[20-Aug-2009 15:40:38] <rmatte> Arrow Crab would be a good one actually
[20-Aug-2009 15:40:38] <rhett> ha. i just mean because you know its king crab and you dont get much further up the hiearchy than the king
[20-Aug-2009 15:40:44] <mrayzenoss> ahh, true
[20-Aug-2009 15:40:46] <rmatte> Ghost Crab, oooh
[20-Aug-2009 15:40:53] <markeriv> emperor crab?
[20-Aug-2009 15:40:57] <markeriv> darth crab?
[20-Aug-2009 15:40:58] <mrayzenoss> Blue Crab, King Crab and Stone Crab taken so far
[20-Aug-2009 15:41:02] <markeriv> lord crab?
[20-Aug-2009 15:41:17] <rhett> coconut crab
[20-Aug-2009 15:41:25] <markeriv> does anyone have any lemon?
[20-Aug-2009 15:41:28] <rhett> ha
[20-Aug-2009 15:41:33] <mrayzenoss> basically if we release twice a year, we have a good 847 more names according to Wikipedia
[20-Aug-2009 15:41:40] <rmatte> yup
[20-Aug-2009 15:41:55] <markeriv> ok so now i have good points
[20-Aug-2009 15:41:55] <mrayzenoss> so just 423 years of Zenoss releases
[20-Aug-2009 15:41:58] <markeriv> now to edit the graph
[20-Aug-2009 15:42:10] <rmatte> Although "Liocarcinus vernalis Crab" doesn't really have much of a ring to it
[20-Aug-2009 15:42:13] <rmatte> hard to pronounce too
[20-Aug-2009 15:42:21] <mrayzenoss> trust me, in 400 years you won't care
[20-Aug-2009 15:42:22] <rmatte>
[20-Aug-2009 15:42:28] <rmatte> haha, true
[20-Aug-2009 15:42:50] <rmatte> well I might, what is going to monitor my status pod status?
[20-Aug-2009 15:42:55] <markeriv> just call it LVcrab
[20-Aug-2009 15:43:10] <rmatte> erm, stasis
[20-Aug-2009 15:43:10] <markeriv> your status pod? What about my cryo tube
[20-Aug-2009 15:43:26] <markeriv> i want something making sure i get let out on time
[20-Aug-2009 15:43:38] <chudler_> "Open the pod bay door please, Zen. I'm sorry Matt, I cannot do that."
[20-Aug-2009 15:43:48] <rmatte> lol
[20-Aug-2009 15:43:56] <markeriv> so erm
[20-Aug-2009 15:43:58] <markeriv> on your graph
[20-Aug-2009 15:44:04] <markeriv> what becomes pings sent/pings recieved
[20-Aug-2009 15:44:12] <markeriv> need to replace those
[20-Aug-2009 15:44:18] <rhett> http://www.allthepages.org/images/blog/coconut_crab.jpg such a lovely picture
[20-Aug-2009 15:44:19] <markeriv> we have avg, loss, max, mdev, and min
[20-Aug-2009 15:44:29] <markeriv> so do i just delete that chart?
[20-Aug-2009 15:44:53] <rmatte> that's a CRAB!?
[20-Aug-2009 15:44:56] <rmatte> crazy
[20-Aug-2009 15:45:09] <rhett> coconut crab
[20-Aug-2009 15:45:18] <rhett> supposed to be super delicious
[20-Aug-2009 15:45:21] <rmatte> that's a frigging monster
[20-Aug-2009 15:46:29] <rmatte> "some coconut crabs are rumored to steal shiny items such as pots and silverware from houses and tents."
[20-Aug-2009 15:46:35] <rmatte> how the heck would they even get in?
[20-Aug-2009 15:46:42] <rmatte> it's not like they can slip under the door
[20-Aug-2009 15:46:56] <chudler_> THEY HAVE THEIR WAYS. XMAS ISLAND. http://www.jurgenfreund.com/stories/xmascrabs/content/bin/images/large/1001785.jpg
[20-Aug-2009 15:47:19] <rmatte> hehe
[20-Aug-2009 15:48:16] <markeriv> so what does the point mdev do?
[20-Aug-2009 15:48:19] <markeriv> or how should I graph it
[20-Aug-2009 15:49:18] <rhett> it graphs deviation
[20-Aug-2009 15:49:19] <rhett> jitter
[20-Aug-2009 15:49:41] <rmatte> IPSLA style stuff
[20-Aug-2009 15:49:51] <markeriv> so latency and loss then
[20-Aug-2009 15:50:50] <rmatte> jitter is the variation in the time between packets arriving, caused by network congestion, timing drift, or route changes. A jitter buffer can be used to handle jitter.
[20-Aug-2009 15:50:57] <markeriv> nod
[20-Aug-2009 15:51:49] <rhett> i still like smokeping or the advanced ping templates for cacti more
[20-Aug-2009 15:51:52] <rhett> but it gives a good idea
[20-Aug-2009 15:53:02] <rmatte> I like actual IPSLA implemented on a core Cisco router
[20-Aug-2009 15:53:02] <markeriv> oie
[20-Aug-2009 15:53:05] <rmatte> that's the ultimate
[20-Aug-2009 15:53:08] <markeriv> i have really managed to screw this graph up lol
[20-Aug-2009 15:53:25] <rhett> too rich for my blood rmatte
[20-Aug-2009 15:53:45] <rmatte> hehe
[20-Aug-2009 15:54:03] <rmatte> we have it implemented for the links to our clients
[20-Aug-2009 15:54:16] <rmatte> so that we can provide them with stats on it in their monthly report
[20-Aug-2009 15:54:46] <rhett> cool
[20-Aug-2009 15:55:17] <rmatte> I wrote a bash script that collects the data from the router via SNMP and feeds it in to Zenoss
[20-Aug-2009 15:55:40] <rhett> cool
[20-Aug-2009 15:57:23] <rmatte> now I just need to figure out how to write proper snmp scripts in python
[20-Aug-2009 15:57:44] <rmatte> If I can modify my SNMP performance monitor zenpack to python it'll be perfect
[20-Aug-2009 16:00:05] <rmatte> I wonder if there's something similar to the cut command available in python
[20-Aug-2009 16:05:17] <rmatte> markeriv: but is it actually displaying data now?
[20-Aug-2009 16:06:13] <rmatte> afk
[20-Aug-2009 16:10:08] <markeriv> lemmie refresh
[20-Aug-2009 16:10:55] <markeriv> mostly
[20-Aug-2009 16:11:03] <markeriv> a couple fields still are not populating
[20-Aug-2009 16:11:05] <markeriv> but i can remove those
[20-Aug-2009 16:11:09] <markeriv> to get a clean looking graph
[20-Aug-2009 16:11:18] <markeriv> round trip minimum in particular
[20-Aug-2009 16:13:50] <rmatte> back
[20-Aug-2009 16:14:11] <rmatte> weird that min is not populating
[20-Aug-2009 16:14:17] <rmatte> the output is correct now with that command
[20-Aug-2009 16:14:19] <markeriv> at this point i just want to delete the round trip minimum and min difference in the graphs i have
[20-Aug-2009 16:14:44] <markeriv> well i have round trip maximum and round trip average
[20-Aug-2009 16:14:54] <markeriv> but they are listed twice in the graph description field
[20-Aug-2009 16:15:09] <markeriv> and the minimum still isnt populating
[20-Aug-2009 16:15:27] <rmatte> did you adjust your graph datapoints?
[20-Aug-2009 16:15:38] <rmatte> they shouldn't be showing up twice
[20-Aug-2009 16:15:48] <rmatte> you need to remove any duplicate datapoints in the graph
[20-Aug-2009 16:15:50] <markeriv> i believe so, but again one claims newbacy
[20-Aug-2009 16:15:51] <rmatte> start fresh
[20-Aug-2009 16:16:29] <markeriv> there we go
[20-Aug-2009 16:16:31] <markeriv> that did it
[20-Aug-2009 16:16:38] <markeriv> dude had some custom stuff in there
[20-Aug-2009 16:16:39] <rmatte> good
[20-Aug-2009 16:16:42] <markeriv> and i found it and nixed it
[20-Aug-2009 16:16:45] <markeriv> now it looks much cleaner
[20-Aug-2009 16:16:48] <rmatte> you'll get the hang of it soon enough
[20-Aug-2009 16:16:58] <rmatte> after a while it just becomes second nature
[20-Aug-2009 16:19:14] <markeriv> nod I hope so
[20-Aug-2009 16:19:15] <markeriv> lol
[20-Aug-2009 16:19:15] <markeriv> well
[20-Aug-2009 16:19:19] <markeriv> more like i hope to get it configured
[20-Aug-2009 16:19:21] <markeriv> copy it once
[20-Aug-2009 16:19:24] <markeriv> and never touch it again
[20-Aug-2009 16:19:35] <markeriv> other than backups ect
[20-Aug-2009 16:19:45] <markeriv> hence all this dev work before we have the product
[20-Aug-2009 16:20:46] <markeriv> hmm now this is interesting
[20-Aug-2009 16:20:53] <markeriv> under the minimum point
[20-Aug-2009 16:21:04] <markeriv> the available RRD variables does not have the Minimum
[20-Aug-2009 16:21:30] <markeriv> what would happen if i deleted the fping_min file? Would it auto recreate it with data?
[20-Aug-2009 16:22:22] <mrayzenoss> yeah
[20-Aug-2009 16:22:25] <markeriv> i suppose arguably I do not need that particular variable as most customers dont care how small their jitter is, rather how large
[20-Aug-2009 16:22:29] <markeriv> alright
[20-Aug-2009 16:22:32] <markeriv> i will try doing that
[20-Aug-2009 16:22:34] <markeriv> see if that fixes it
[20-Aug-2009 16:22:39] <markeriv> or if it gives cant find RRD file issues
[20-Aug-2009 16:22:51] <rmatte> markeriv: knowing customers, the more info you can throw at them, the better
[20-Aug-2009 16:23:01] <markeriv> yea but sometimes less is more
[20-Aug-2009 16:23:05] <rmatte> you have to aim to impress
[20-Aug-2009 16:23:21] <markeriv> i cant count how many times I have said that a customer's utilization on their T1 is low only to have them yell at me
[20-Aug-2009 16:23:37] <markeriv> and tell me i better damn well fix it
[20-Aug-2009 16:23:38] <markeriv> lol
[20-Aug-2009 16:23:56] <rmatte> lol
[20-Aug-2009 16:24:11] <rmatte> I'd be like "sure thing" *fire up bittorrent*
[20-Aug-2009 16:24:16] <rmatte> I'm done for the day
[20-Aug-2009 16:24:20] <rmatte> talk to you guys tomorrow
[20-Aug-2009 16:24:20] <markeriv> LOL!
[20-Aug-2009 16:24:25] <markeriv> thanks a bunch rmatte
[20-Aug-2009 16:24:30] <rmatte> no problem
[20-Aug-2009 16:24:32] <markeriv> everyone has been a big help
[20-Aug-2009 16:29:00] <rhett> is everything working for you?
[20-Aug-2009 16:34:22] <mrayzenoss> later all
[20-Aug-2009 16:36:35] <rhett> good bye
[20-Aug-2009 18:04:02] <etank> anyone know if i can run zenoss on a CentOS 5 x86 box and have a distrobuted collector running on a Red Hat EL 4 machine?
[20-Aug-2009 18:04:17] <etank> s/distrobuted/distributed/
[20-Aug-2009 19:12:30] <rhett> i know you can run it on cent5 dont know about collector on cent4
[20-Aug-2009 19:12:38] <rhett> thats my easy answer!
[21-Aug-2009 00:00:46] [disconnected at Fri Aug 21 00:00:46 2009]
[21-Aug-2009 00:00:46] [connected at Fri Aug 21 00:00:46 2009]
[21-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[21-Aug-2009 06:52:08] <_mingau_> hello people
[21-Aug-2009 06:54:01] <ckrough> good morning
[21-Aug-2009 06:55:30] <_mingau_> morning
[21-Aug-2009 06:55:41] <_mingau_> does anybody use freebsd here?
[21-Aug-2009 07:02:05] <_mingau_> ?
[21-Aug-2009 07:05:56] <_mingau_> testing?
[21-Aug-2009 07:07:08] <ckrough> Not too many people do. I don't believe it's an
[21-Aug-2009 07:07:29] <ckrough> 'officially' supported OS for Zenoss. I think straterra uses it, he's the only one I've seen though.
[21-Aug-2009 07:09:48] <_mingau_> i use too
[21-Aug-2009 07:09:49] <_mingau_>
[21-Aug-2009 07:10:04] <ckrough> I assumed
[21-Aug-2009 07:22:13] <rmatte> scp2004 uses it, but he's running in to problems as well
[21-Aug-2009 07:22:38] <rmatte> FreeBSD + Zenoss == Problematic
[21-Aug-2009 07:23:12] <rmatte> seems like everyone who uses that combo runs in to problems, then when they fix one problem something else breaks
[21-Aug-2009 07:23:19] <rmatte> and it's a never-ending ordeal
[21-Aug-2009 07:25:46] <rmatte> doesn't running Zenoss on FreeBSD partially use FreeBSD's Linux emulation?
[21-Aug-2009 07:26:00] <rmatte> or does it run 100% native after compilation?
[21-Aug-2009 07:28:39] <rmatte> ah, you need to add the proc line to fstab, so yeh it does use a bit of emulation
[21-Aug-2009 07:30:43] <_mingau_> hey Ryan
[21-Aug-2009 07:30:58] <_mingau_> i know its problematic
[21-Aug-2009 07:30:59] <_mingau_>
[21-Aug-2009 07:31:02] <_mingau_> u know...
[21-Aug-2009 07:31:08] <_mingau_> its my karma
[21-Aug-2009 07:34:40] <_mingau_> i am trying to make my boss understand that i need any linux...
[21-Aug-2009 07:35:43] <rmatte> hehe
[21-Aug-2009 07:35:55] <rmatte> I thought they had agreed to install CentOS?
[21-Aug-2009 07:36:13] <rmatte> or maybe that was scp2004
[21-Aug-2009 07:37:09] <_mingau_> kkk
[21-Aug-2009 07:38:38] <rmatte> CentOS is just the free version of RedHat
[21-Aug-2009 08:11:03] <_mingau_> i know
[21-Aug-2009 08:48:41] <rmatte> good morning Matt
[21-Aug-2009 08:49:38] <mrayzenoss> morning
[21-Aug-2009 08:49:48] <mrayzenoss> today's the day
[21-Aug-2009 08:49:59] <mrayzenoss> I've got 12 entries for the ZenPack contest that I need to get through
[21-Aug-2009 08:50:04] <_mingau_> hello big matt
[21-Aug-2009 08:50:10] <jb> doesnt look like mine is going to make it
[21-Aug-2009 08:50:12] <_mingau_> hey...i have a question...
[21-Aug-2009 08:50:13] <jb> but it wouldn't win anyways
[21-Aug-2009 08:50:32] <rmatte> mrayzenoss: I might as well submit mine, but I'm going to write up an instruction blurb for the ZenPack page
[21-Aug-2009 08:50:57] <rmatte> since it's not as polished as I'd like it to be, but it works fine
[21-Aug-2009 08:51:28] <_mingau_> about autodiscover windows devices...
[21-Aug-2009 08:51:56] <_mingau_> the authentication...
[21-Aug-2009 08:52:13] <_mingau_> it shows that the user must be a local admin...
[21-Aug-2009 08:52:17] <_mingau_> so...
[21-Aug-2009 08:53:21] <_mingau_> will this process write something?
[21-Aug-2009 08:53:28] <_mingau_> or only read wmi?
[21-Aug-2009 08:53:35] <jb> it only reads
[21-Aug-2009 08:54:11] <jb> but non-privledged users do not have full read access to the WMI namespace
[21-Aug-2009 08:56:43] <_mingau_> hunnn
[21-Aug-2009 08:57:37] <ckrough> yay: http://webaj.com/how-setup-splunk-monitoring-zenoss-logs.htm
[21-Aug-2009 08:57:48] <ckrough> was just going to start writing recipes for that
[21-Aug-2009 08:58:21] <ckrough> meh, doesnt look like he did fancy recipes. I guess Im still doing it...
[21-Aug-2009 08:59:44] <mrayzenoss> ckrough: check the date on that article, don't get too attached to the details
[21-Aug-2009 09:00:01] <ckrough> There arent any details in that article...
[21-Aug-2009 09:00:06] <mrayzenoss> ugh, spammers responding to spammers
[21-Aug-2009 09:01:47] <ckrough> I havent looking into it much, but I assume I'll need to write some tranforms to clean up the data once splunk snarfs it
[21-Aug-2009 09:02:06] <mrayzenoss> there was someone from splunk on here the other day, I believe they use Zenoss
[21-Aug-2009 09:02:08] <ckrough> splunk transforms that is, not zenoss transforms
[21-Aug-2009 09:02:32] <ckrough> Im tired of browsing through all the text files on the collectors, time to consolidate that mess
[21-Aug-2009 09:03:51] <rmatte> mrayzenoss: I just sent an email to you with my ZenPack details
[21-Aug-2009 09:09:19] <ckrough> someone forgot to put a cabinet row into maintenance mode last night during a maint window. It took almost 4 hours for the SMS pages to get caught up
[21-Aug-2009 09:09:53] <jb> yikes
[21-Aug-2009 09:11:45] <jb> http://pastebin.com/m18889c9b
[21-Aug-2009 09:11:51] <jb> anyone have a clue on how to properly format this transform?
[21-Aug-2009 09:13:02] <ckrough> jb: this comes in via zentrap?
[21-Aug-2009 09:13:17] <jb> yeah
[21-Aug-2009 09:13:28] <jb> i just dont know how to format evt.message properly
[21-Aug-2009 09:13:43] <jb> i'd like each of those on a new line
[21-Aug-2009 09:15:37] <rmatte> you might be able to use \n linebreaks
[21-Aug-2009 09:15:53] <rmatte> like "firstline\nsecondline\nthirdline"
[21-Aug-2009 09:16:07] <rmatte> but I've never tried it in a transform
[21-Aug-2009 09:18:26] <rmatte> >>> var1 = 'hey\nthere'
[21-Aug-2009 09:18:26] <rmatte> >>> print var1
[21-Aug-2009 09:18:26] <rmatte> hey
[21-Aug-2009 09:18:26] <rmatte> there
[21-Aug-2009 09:18:27] <rmatte> >>>
[21-Aug-2009 09:18:33] <jb> yeah
[21-Aug-2009 09:18:43] <ckrough> is that how its handled internall though?
[21-Aug-2009 09:18:54] <rmatte> no idea
[21-Aug-2009 09:19:05] <rmatte> you'd have to try it out and see
[21-Aug-2009 09:19:35] <jb> any line breaks makes the transform invalid
[21-Aug-2009 09:20:38] <rmatte> ah
[21-Aug-2009 09:20:40] <ckrough> try \\n? maybe it needs to be escaped so their code doesnt choke
[21-Aug-2009 09:20:47] <jb> but im probably not doing the variables right
[21-Aug-2009 09:20:58] <jb> evt.message =
[21-Aug-2009 09:20:59] <jb> 'Machine Type / Model / Serial: %s' % (machineModelSerial)
[21-Aug-2009 09:21:02] <jb> even that makes it invalid
[21-Aug-2009 09:22:04] <rmatte> well, it needs to be right after the evt.message = probably
[21-Aug-2009 09:22:17] <rmatte> I don't think you can do it the way you're trying to
[21-Aug-2009 09:22:22] <jb> yeah, me either
[21-Aug-2009 09:23:46] <ckrough> You are not alone http://forums.zenoss.com/viewtopic.php?p=14871#14871
[21-Aug-2009 09:23:54] <rmatte> yeh, I just saw that post
[21-Aug-2009 09:24:43] <jb> hm yep
[21-Aug-2009 09:25:42] <jb> on a side note.. any of you guys pulling I/O from server HBAs?
[21-Aug-2009 09:25:57] <ckrough> yes
[21-Aug-2009 09:26:10] <jb> windows hosts?
[21-Aug-2009 09:26:17] <ckrough> no, RHEL
[21-Aug-2009 09:26:30] <ckrough> on dells with OpenManage, which dumps into snmp
[21-Aug-2009 09:26:37] <jb> ok
[21-Aug-2009 09:26:38] <jb> exactly what I need
[21-Aug-2009 09:26:46] <jb> i think the dellmon zenpack that im using breaks that
[21-Aug-2009 09:27:08] <jb> so openmanage provides I/O for everything? perc attached local disks as well as fiber attached storage?
[21-Aug-2009 09:27:09] <ckrough> I had to modify something, lets see if I can figure out what it was... been a while
[21-Aug-2009 09:27:25] <ckrough> Im not using FAS, so not sure on that.
[21-Aug-2009 09:32:24] <ckrough> Im using zenoss.snmp/UCDHardDiskMap to get it I think
[21-Aug-2009 09:33:04] <ckrough> I remember having to modify a Dell specifiz plugin at one point, but we abandoned those changes on the last upgrade. Now that I look, not all devices are getting IO
[21-Aug-2009 09:33:57] <jb> ah ok
[21-Aug-2009 09:34:14] <jb> im wondering if i can get these stats directly from the HBA
[21-Aug-2009 09:34:28] <ckrough> via?
[21-Aug-2009 09:34:43] <jb> not sure
[21-Aug-2009 09:34:48] <jb> maybe snmp?
[21-Aug-2009 09:34:57] <jb> probably not
[21-Aug-2009 09:35:50] <jb> probably easier to pull it from OM/windows
[21-Aug-2009 09:44:37] <jb> anyone here using egor's DellMon pack?
[21-Aug-2009 09:52:41] RoninX341 is now known as etank
[21-Aug-2009 09:53:15] <gwb2351> chudler_: the sentryCDU community zenpack you're developing is working great!
[21-Aug-2009 09:58:24] <chudler_> gwb2351: happy to hear it. It still has some rough edges... I will be improving those Some Day Soon
[21-Aug-2009 10:08:28] <gwb2351> the only thing I've modified are adding a "10,/" to the RDN of the graphs. the only things I'm having problems with is some "aggregate load" calculations are returning negative numbers (verified that snmpget of the load OID is providing valid input, but haven't gotten any farther than that).
[21-Aug-2009 10:09:28] <chudler_> I am tracking some bugs related to non-updating values, and the one you mentioned about the "centiamps". That other one you mentioned, I hope is related to the failure to update RRD values under certain conditions.
[21-Aug-2009 10:11:35] <gwb2351> ah... is the RRD failure to update this bug (http://dev.zenoss.org/trac/ticket/5306) ?
[21-Aug-2009 10:13:33] <gwb2351> i'll try the workaround on the few devices that aren't reporting properly and see if that fixes them
[21-Aug-2009 10:13:48] <eidolon> hey do classes such as Devices/Storage/Filer do an automatic remodel every 6 hours or so? we had a problem where alerts were going off, but the filesystem had been resized. the docs imply a remodel happens automatically every few hours.
[21-Aug-2009 10:13:55] <eidolon> but this one didn't - i remodelled by hand, and it cleared the condition.
[21-Aug-2009 10:14:10] <eidolon> (target machine is a netapp filer, obviously)
[21-Aug-2009 10:18:48] <mrayzenoss> eidolon: default remodeling cycle is 720 minutes, set on the collector (every 12 hours)
[21-Aug-2009 10:20:39] <eidolon> hyum. okay, mray. i need to dig and see if that's working
[21-Aug-2009 10:20:40] <eidolon> thanks
[21-Aug-2009 10:21:43] <TBKDan> I want to have two alerting periods: one during working hours (m-f 8-5) and one for the rest of the time. The 8-5 would be immediate alerting, and the rest of the time would be on a 5 minute delay. Right now I have the two rules set up, but if a machine goes down for more than 5 minutes, i get two alerts (one immediately and one in 5 minutes). Is there a way to just disable that 5 min rule during 8-5 or do I have to set up individual rules for weekdays and
[21-Aug-2009 10:24:32] <mrayzenoss> TBKDan: You should be able to configure your Alert Schedule that way
[21-Aug-2009 10:24:49] <mrayzenoss> so you already have one in place
[21-Aug-2009 10:25:13] <mrayzenoss> and you can edit the scedule details
[21-Aug-2009 10:25:45] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch02s06.html
[21-Aug-2009 10:29:31] <TBKDan> mrayzenoss: I guess to simplify the question a bit... can I tell a rule to be *inactive* for a time period and active for the rest? Right now the way I understand it is I can tell it to be active for a time period and inactive for the rest.
[21-Aug-2009 10:30:07] <mrayzenoss> TBKDan: yeah, that's what the schedule would give you
[21-Aug-2009 10:33:57] <TBKDan> mrayzenoss: So if I create an active period of M-F 8-5 and set enabled to false, does that mean it will not alert during that time but will the rest of the time?
[21-Aug-2009 10:34:55] <mrayzenoss> no, I think what you want to do is set the active period from 5pm-8am M-F
[21-Aug-2009 10:35:03] <mrayzenoss> (and the weekend)
[21-Aug-2009 10:35:18] <mrayzenoss> enabled is just if you want to turn it back on later
[21-Aug-2009 10:35:40] <TBKDan> Arg, ok.. that's what I was hoping I wouldn't have to do
[21-Aug-2009 10:37:08] <mrayzenoss> but if you set your "rest of the time" to be the default setting and then have the 8-5 setting to be scheduled, that may work
[21-Aug-2009 10:37:33] <mrayzenoss> so "rest of time" is 1 page and 8-5 is 2 pages?
[21-Aug-2009 10:38:10] <TBKDan> That's what I was trying to eliminate. Right now we get duplicate pages during business hours.
[21-Aug-2009 10:38:17] <mrayzenoss> ahh, hmmm.
[21-Aug-2009 10:38:37] <mrayzenoss> I guess our alerting could be smarter
[21-Aug-2009 10:38:50] <TBKDan> Additionally, in the "Repeat" section of the schedule, weekend is not an option
[21-Aug-2009 10:38:53] <mrayzenoss> the reason it's not is that we allow you to set alert schedules all over the place
[21-Aug-2009 10:39:27] <mrayzenoss> so some alerts go to different people at different times for different systems and groups, figuring out the overlap is non-trivial
[21-Aug-2009 10:40:22] <mrayzenoss> so I guess you have to do 4 alert schedules? 12am-8am, 8-5, 5-12pm and the weekend?
[21-Aug-2009 10:40:55] <jb> i do that..
[21-Aug-2009 10:41:00] <jb> after 5PM and weekends..
[21-Aug-2009 10:41:06] <jb> and then normal 6am-5PM
[21-Aug-2009 10:41:08] <jb> during the week
[21-Aug-2009 10:41:21] <TBKDan> For weekends you have two jobs, one for each day?
[21-Aug-2009 10:41:27] <jb> sec
[21-Aug-2009 10:41:28] <mrayzenoss> jb: does 5pm-8am weekdays work?
[21-Aug-2009 10:41:45] <mrayzenoss> jb: ie. does it pick up Monday at 12am?
[21-Aug-2009 10:42:19] <jb> oh..
[21-Aug-2009 10:42:19] <jb> hm
[21-Aug-2009 10:42:23] <jb> actually, i only do after 5PM
[21-Aug-2009 10:42:34] <jb> looks like I don't differentiate between weekdays/weekends
[21-Aug-2009 10:43:20] <TBKDan> Would adding "Weekend" repeats be difficult?
[21-Aug-2009 10:43:47] <TBKDan> I don't know if I missed something in the GUI or if it's somewhere deeper in the code
[21-Aug-2009 10:45:44] <TBKDan> brb
[21-Aug-2009 10:59:43] <jb> ok i think im going to buy this OIDView Professinal program
[21-Aug-2009 10:59:51] <jb> unless someone can reccomend a better free one?
[21-Aug-2009 11:00:18] <TBKDan> Back
[21-Aug-2009 11:03:11] <mrayzenoss> jb: which features do you really like about it?
[21-Aug-2009 11:03:19] <mrayzenoss> jb: I think I used that at a previous employer
[21-Aug-2009 11:03:34] <jb> well, it has multiple views which makes it really nice.. it has this LiveGrid thing
[21-Aug-2009 11:03:39] <jb> that allows you to view entire tables easily
[21-Aug-2009 11:03:47] <jb> and it auto-downloads/compiles any additional MIBs that it needs
[21-Aug-2009 11:04:04] <mrayzenoss> cool
[21-Aug-2009 11:04:10] <jb> its $299 fo a single user
[21-Aug-2009 11:04:34] <mrayzenoss> I don't have a good recommendation, I'm just looking for ideas to suggest for the mib-browser zenpack
[21-Aug-2009 11:04:45] <jb> ah
[21-Aug-2009 11:05:24] <mrayzenoss> Kells has been adding minor touch-ups here and there, but I've been pushing Brandon to adopt it into Core and beef it up
[21-Aug-2009 11:05:35] <mrayzenoss> that said, we need more developers
[21-Aug-2009 11:06:37] <mrayzenoss> I should probably rename the MIB browser the MIB Utils ZenPack
[21-Aug-2009 11:06:54] <mrayzenoss> since it does a little more than that
[21-Aug-2009 11:12:11] _mingau_ is now known as apocalipse
[21-Aug-2009 11:13:03] <apocalipse> hey people
[21-Aug-2009 11:14:41] <apocalipse> where can i find job postings list?
[21-Aug-2009 11:15:04] <mrayzenoss> http://www.zenoss.com/about/jobs
[21-Aug-2009 11:15:09] <mrayzenoss> that one?
[21-Aug-2009 11:15:50] <rmatte> hopefully he meant that one since this isn't a recruitment agency
[21-Aug-2009 11:16:01] <apocalipse>
[21-Aug-2009 11:20:45] <rmatte> mrayzenoss: I guess you have your work set out for you today with all the procrastinary last minute submissions hehe
[21-Aug-2009 11:21:15] <mrayzenoss> ugh
[21-Aug-2009 11:21:24] * rmatte is guilty
[21-Aug-2009 11:21:32] <mrayzenoss> so I just got an email from my boss saying he wanted the ZenPack contest wrapped up by Monday
[21-Aug-2009 11:21:40] <rmatte> ah
[21-Aug-2009 11:21:43] <mrayzenoss> I have 13 ZenPacks to process, which on average take 1 hour each
[21-Aug-2009 11:21:56] <chudler_> D'oh!
[21-Aug-2009 11:22:47] <chudler_> "um, I'm gonna need you to go ahead and come in on Saturday"
[21-Aug-2009 11:23:05] <rmatte> lol
[21-Aug-2009 11:23:56] <rmatte> ask if it can be extended to Tuesday or something
[21-Aug-2009 11:24:07] <rmatte> but make today the absolute final submission date
[21-Aug-2009 11:24:13] <rmatte> I doubt 1 day will hurt anything
[21-Aug-2009 11:25:17] <mrayzenoss> the final submission date was last Friday
[21-Aug-2009 11:26:24] <mrayzenoss> I've made a few exceptions for ZenPacks that would have come in if I'd been around to bug people, but I was on vacation
[21-Aug-2009 11:27:14] <rmatte> ah
[21-Aug-2009 11:42:41] sakura_ is now known as x-ip
[21-Aug-2009 11:50:38] <markeriv> hello everyone
[21-Aug-2009 11:51:24] <markeriv> anyone around? I am trying to decide if I should install zenoss on my new server today or if I should wait for 2.4.3, and what additional features/fixes might be in that version
[21-Aug-2009 11:52:42] <mrayzenoss> markeriv: here's the list of closed tickets for 2.4.3: http://dev.zenoss.org/trac/report/6
[21-Aug-2009 11:53:05] <mrayzenoss> once all the "verification" ones move to "closed" the release will be ready
[21-Aug-2009 11:53:34] <mrayzenoss> that said, our QA Manager crashed his bike this morning and is missing the day after having to get stiches
[21-Aug-2009 11:53:41] <markeriv> ouch
[21-Aug-2009 11:54:36] <mrayzenoss> so no tickets are being added at this point, unless they're insanely bad
[21-Aug-2009 11:54:52] <rmatte> yikes
[21-Aug-2009 11:55:00] <rmatte> motorbike?
[21-Aug-2009 11:55:10] <mrayzenoss> no, just a mtn. bike
[21-Aug-2009 11:55:14] <rmatte> ah
[21-Aug-2009 11:55:19] <rmatte> what did he crash it in to?
[21-Aug-2009 11:55:20] <rmatte> lol
[21-Aug-2009 11:56:07] <markeriv> sigh this new quadcore xeon 2.4 is purring on my desk with its 8 gigs of ram and making me drool over the fact that I have not installed anything on it yet
[21-Aug-2009 11:56:33] <mrayzenoss> a source build would probably be 2.4.3 unless QA rejects any of the tickets
[21-Aug-2009 11:56:55] <chudler_> if you just install SSH with some easy to guess passwords, you wont have to go through the trouble of installing anything. Someone will do it for you
[21-Aug-2009 11:57:01] <victor-> i'm not sure what I've changed, but it seems like zen.Device is unable to add devices via hostname anymore. I have a script that uses xmlrpc that is now failing, and i'm getting the same result in the web gui when adding a device by name. from the host CLI, I can ping the devices, so that's working properly. any ideas of what i should check?
[21-Aug-2009 11:57:25] <markeriv> lol chudler
[21-Aug-2009 11:57:58] <chudler_> that's what I call outsourcing
[21-Aug-2009 11:58:15] <mrayzenoss> I call that "cloud computing"
[21-Aug-2009 12:00:37] <victor-> zenoss's event.log shows an error like: INFO zen.Device device name 'hostname.new' for ip ''
[21-Aug-2009 12:01:01] <mrayzenoss> victor-: so DNS isn't finding the machine's IP?
[21-Aug-2009 12:04:41] <victor-> mrayzenoss: the linux system DNS resolution works fine - I can run 'nslookup hostname.new', etc, and it resolves. it's just zenoss that seems to be having trouble.
[21-Aug-2009 12:07:13] <mrayzenoss> victor-: weird, I assume if you add the device by IP address, it then finds the right name?
[21-Aug-2009 12:07:29] <mrayzenoss> but if you add it by hostname, it doesn't work...?
[21-Aug-2009 12:08:12] <geniusj> Hey all.. I have what hopefully is a quick question.. I have a cacti script that I'm trying to turn into a performance monitor in zenoss. I get the script added, it tests fine, etc. But it doesn't seem to generate any actual data. I'm wondering if it's a data points issue or something? I named the data points after the key in the key/value pairs that the cacti script spits out. e.g. num_gets: 5 - data point name = num_gets
[21-Aug-2009 12:09:01] <geniusj> I'm wondering if anyone has gotten cacti scripts successfully imported. I didn't see much online about it other than that it supports it
[21-Aug-2009 12:11:46] <ckrough> How can I get my collectors to reconnect to the master gracefully? The main box locked up and I had to reboot, now I'll need to convince the collectors to reconnect, but I have to stagger them so they dont overwhelm it
[21-Aug-2009 12:13:30] <mrayzenoss> geniusj: I haven't migrated Cactii plugins, but let's just assume your script is working and debug it like a Nagios plugin...
[21-Aug-2009 12:13:36] <mrayzenoss> geniusj: is the template bound?
[21-Aug-2009 12:13:49] <mrayzenoss> geniusj: do you get no graphs, or graphs that don't fill in with any data?
[21-Aug-2009 12:15:29] <mrayzenoss> ckrough: I'm guessing support will have better answers
[21-Aug-2009 12:15:37] <ckrough> thanks
[21-Aug-2009 12:15:53] <mrayzenoss> ckrough: I know there's a lot of work going into stability and recovery for distributed collectors in King Crab
[21-Aug-2009 12:16:43] <ckrough> I keep forgetting that distributed collectors is only ent
[21-Aug-2009 12:21:44] <victor-> ah, found my problem to be that zenoss/bin/zensocket wasn't owned by root:root.
[21-Aug-2009 12:22:01] <mrayzenoss> victor-: source build? gotta remember that step
[21-Aug-2009 12:22:24] <geniusj> mrayzenoss: I get graphs, they just dont' fill with data
[21-Aug-2009 12:23:02] <mrayzenoss> geniusj: so that indicates the template is bound and the graph defined, but something about the data source is off
[21-Aug-2009 12:23:27] <geniusj> mrayzenoss: Yea. What's the best way to debug that?
[21-Aug-2009 12:23:53] <mrayzenoss> based on a similar discussion yesterday, get someone in here to debug your output for you
[21-Aug-2009 12:23:57] <geniusj> mrayzenoss: most of the debug methods I've found just show you the output of the command, which doesn't help me.. I want to see what it lookks like after it's parsed
[21-Aug-2009 12:24:14] <mrayzenoss> geniusj: what does the output look like when you hit "Test"
[21-Aug-2009 12:24:28] <geniusj> let me pull it up
[21-Aug-2009 12:25:27] <geniusj> uptime:598850 cmd_get:8753723 curr_connections:5 connection_structures:297 rusage_user:389.422798 total_connections:52334 cmd_set:1575591 time:1250875514 bytes_written:21919121415 get_misses:8107930 rusage_system:461.534836 bytes_read:22351049321 get_hits:645793
[21-Aug-2009 12:26:42] <mrayzenoss> so are you graphing any of those?
[21-Aug-2009 12:27:02] <geniusj> yea.. cmd_get and cmd_set for now
[21-Aug-2009 12:27:27] <geniusj> they're type COUNTER (hopefully that's correct)?
[21-Aug-2009 12:27:29] <straterra> I showed my CIO Zenoss..and he wants me to configure it for his other company
[21-Aug-2009 12:27:33] <mrayzenoss> I meant, are they working?
[21-Aug-2009 12:27:42] <geniusj> oh.. no, not working
[21-Aug-2009 12:27:53] <mrayzenoss> geniusj: and the parser is set to Cactii or Auto?
[21-Aug-2009 12:27:57] <geniusj> yup
[21-Aug-2009 12:28:12] <mrayzenoss> straterra: excellent
[21-Aug-2009 12:28:19] <straterra> Just core..
[21-Aug-2009 12:28:29] <mrayzenoss> BTW, we'll have a blog post up today about free training at Ohio LinuxFest
[21-Aug-2009 12:28:35] <straterra> Oh, cool
[21-Aug-2009 12:28:40] <straterra> I'm in Indianapolis..so..
[21-Aug-2009 12:28:56] <mrayzenoss> straterra: then I'll see you there
[21-Aug-2009 12:29:01] <straterra> lol..maybe
[21-Aug-2009 12:29:22] <straterra> quick question..using snmp traps, I could like..have a printer notify Zenoss when it was running low on ink or something, right?
[21-Aug-2009 12:29:37] <mrayzenoss> straterra: if your printer is cool enough to send traps, yes
[21-Aug-2009 12:29:52] <straterra> It may not be
[21-Aug-2009 12:29:59] <straterra> I'm not even sure it does SNMP at all..but it may
[21-Aug-2009 12:30:02] <mrayzenoss> straterra: or check this out and add a threshold: http://www.zenoss.com/community/projects/zenpacks/printertoner
[21-Aug-2009 12:30:04] <geniusj> mrayzenoss, I'm going to try one other thing.. I'm wondering if there has to be a datapoint defined for every single key/value pair that's output by the command, because there currently isn't
[21-Aug-2009 12:30:43] <straterra> Oh, VERY nice
[21-Aug-2009 12:30:54] <straterra> We have primarily xerox and hp printers
[21-Aug-2009 12:32:09] <Belial_> 18:07:14 <@mrayzenos> victor-: weird, I assume if you add the device by IP address, it then finds the right name?
[21-Aug-2009 12:32:23] <Belial_> Apologies, Mispaste. Silly Putty
\[21-Aug-2009 12:39:45] <geniusj> hm.. no luck I think
[21-Aug-2009 12:43:30] <markeriv> lol
[21-Aug-2009 12:43:36] <markeriv> i actually like that feature
[21-Aug-2009 12:43:43] <markeriv> wish I knew how to turn it on for fedora
[21-Aug-2009 12:44:42] * mrayzenoss off to lunch
[21-Aug-2009 12:58:20] <twm1010> afternoon
[21-Aug-2009 13:17:29] <jb> it returns values
[21-Aug-2009 13:34:57] <straterra> Is there anyway to give a device name to something that isn't in DNS?
[21-Aug-2009 13:37:49] <straterra> mrayzenoss: Some..funky stuff going on after I install that zenpack. Zenoss just asked me to send an error report
[21-Aug-2009 13:38:13] <rmatte> straterra: just add the device by IP and then rename it
[21-Aug-2009 13:38:16] <rmatte> there's a rename function
[21-Aug-2009 13:38:59] <straterra> mrayzenoss: when I tried to remove community.snmp.PrinterTonerMap from the device heirarchy
[21-Aug-2009 13:39:03] <straterra> rmatte: alrighty
[21-Aug-2009 13:39:46] <straterra> oh..I have to restart zope
[21-Aug-2009 13:39:48] <straterra> That's cool
[21-Aug-2009 14:01:59] <geniusj> Looks like my ZenCommand ZenRRD task might not even attempt to run at all. Though I can test it, etc. Anything I need to verify or need to have to make sure that it actually runs? I'm fairly new to zenoss. The data source is enabled..
[21-Aug-2009 14:04:47] <rmatte> well, I think I finally have my Zenoss upgrade/standardization steps worked out, now I just need to test them on Monday for the deployment Tuesday
[21-Aug-2009 14:04:48] <mrayzenoss> check the path? make sure that whatever the command is works with the zenoss user
[21-Aug-2009 14:05:55] <mrayzenoss> there's a bug about the 'Test' button not using the same path as zencommand
[21-Aug-2009 14:06:08] <geniusj> I have zencommand running in debug mode
[21-Aug-2009 14:06:16] <geniusj> or with debug logging rather
[21-Aug-2009 14:06:18] <rmatte> yeh, which basically makes the "test" useless for the time being
[21-Aug-2009 14:06:18] <mrayzenoss> do you see the command get called?
[21-Aug-2009 14:06:22] <geniusj> but I don't even see it mentioned
[21-Aug-2009 14:06:33] <mrayzenoss> try using the full path in the data source
[21-Aug-2009 14:06:34] <geniusj> all I see is this over and over: 2009-08-21 12:05:41 DEBUG zen.zencommand: Queueing event {'manager': 'localhost', 'timeout': 180, 'device': 'localhost', 'eventClass': '/Heartbeat', 'component': 'zencommand', 'agent': 'zencommand'}
[21-Aug-2009 14:06:40] <geniusj> ok.. will do
[21-Aug-2009 14:07:29] <geniusj> do I have to restart zencommand or anything after I change the path?
[21-Aug-2009 14:07:49] <geniusj> and btw, what I did before was set zCommandPath I believe
[21-Aug-2009 14:09:55] <mrayzenoss> I dunno, I seem to remember Chet saying to blank it out, since it usually screws up
[21-Aug-2009 14:10:07] <mrayzenoss> and to just stick with absolute paths when possible
[21-Aug-2009 14:10:24] <rmatte> yeh, absolute paths work much better
[21-Aug-2009 14:10:31] <rmatte> at the moment anywyas
[21-Aug-2009 14:10:53] <geniusj> ok
[21-Aug-2009 14:13:33] <geniusj> would the component or event key fields prevent it from working? Right now they're set fairly arbitrarily
[21-Aug-2009 14:13:45] <geniusj> or event class for that matter?
[21-Aug-2009 14:14:47] <straterra> And I'm having another issue with that printertonermap zenpack
[21-Aug-2009 14:14:55] <straterra> User-supplied Python expression (here.maxToner * .05) for minimum value caused error: ['toner_toner']
[21-Aug-2009 14:15:18] <mrayzenoss> geniusj: those shouldn't cause any issues
[21-Aug-2009 14:15:58] <mrayzenoss> geniusj: what are you monitoring?
[21-Aug-2009 14:16:06] <geniusj> memcachedb
[21-Aug-2009 14:16:38] <mrayzenoss> cool, but unfortunately not something I have running that I could test
[21-Aug-2009 14:17:00] <mrayzenoss> I don't know if there's an issue with the Cacti parsers, nothing in Trac about it
[21-Aug-2009 14:17:10] <geniusj> just to trace my steps:
[21-Aug-2009 14:17:14] <mrayzenoss> but my guess is something is weird about the way the output is being parsed
[21-Aug-2009 14:18:14] <geniusj> I created a template in my memcachedb server heirarchy, I added a COMMAND data source using a script that produces cacti style output. Chose cacti for parser, enabled service. Added data points. Then I bound the template to that same heirarchy
[21-Aug-2009 14:18:46] <geniusj> that's about it.. so the area for the graph is there for each server, just nog raph itself
[21-Aug-2009 14:18:55] <geniusj> not sure if I missed any steps
[21-Aug-2009 14:19:09] <mrayzenoss> yeah, that sounds right
[21-Aug-2009 14:19:11] <straterra> mrayzenoss: Do I need to like..manually add the OID's for max?
[21-Aug-2009 14:19:58] <geniusj> oh.. I also added a custom field for memcachedbPort
[21-Aug-2009 14:20:04] <mrayzenoss> I don't think so, I've used the ZenPack myself, let me go see what it's doing
[21-Aug-2009 14:20:06] <geniusj> wel.. cMemcachedbPort I believe
[21-Aug-2009 14:20:38] <mrayzenoss> geniusj: maybe try creating a new template and modify the command to do just 1 data point and see if that works
[21-Aug-2009 14:20:54] <mrayzenoss> if 1 works, then try ramping it up?
[21-Aug-2009 14:20:56] <rmatte> straterra: do you have the collector plugin from that ZenPack bound to the class the device is under?
[21-Aug-2009 14:21:14] <rmatte> straterra: it looks like it's expecting to find a value in here.maxToner but it's not finding it at all
[21-Aug-2009 14:21:28] <straterra> rmatte: Yeah..
[21-Aug-2009 14:21:36] <straterra> The collector plugin is bound
[21-Aug-2009 14:21:42] <rmatte> hmmm
[21-Aug-2009 14:21:45] <straterra> Is there any way to edit/take a look at that plugin?
[21-Aug-2009 14:21:56] <rmatte> I've never used that ZenPack, so don't really know what to tell you
[21-Aug-2009 14:21:57] <mrayzenoss> the source is available
[21-Aug-2009 14:22:19] <mrayzenoss> trac-zenpacks/browser/zenpacks/ZenPacks.community.PrinterToner
[21-Aug-2009 14:22:28] <mrayzenoss> I know there's someone working on it in the zenoss-zenpacks forum
[21-Aug-2009 14:22:31] <straterra> mrayzenoss: You're trying to sabotage me
[21-Aug-2009 14:22:32] <rmatte> mrayzenoss: when an upgrade is performed, does it overwrite the existing rpt files?
[21-Aug-2009 14:22:40] <rmatte> and report plugins, for that matter
[21-Aug-2009 14:23:22] <straterra> I suppose I can get by for now by creating a local copy and just hard coding the number (not use percentages)
[21-Aug-2009 14:23:28] <mrayzenoss> rmatte: if it's not in a ZenPack, I'd be nervous about it
[21-Aug-2009 14:23:53] <mrayzenoss> I don't know where modified reports are stored
[21-Aug-2009 14:23:57] <rmatte> mrayzenoss: not really that easy to store reports in ZenPacks lol, I guess I'll have to look in to it
[21-Aug-2009 14:24:10] <mrayzenoss> there are 3 report ZenPacks
[21-Aug-2009 14:24:38] <rmatte> for custom reports it's simple, but for modified .rpt files it's not as simple as just "add to zenpack"
[21-Aug-2009 14:24:48] <mrayzenoss> ahh, yeah
[21-Aug-2009 14:25:02] <straterra> God..python makes no sense to me
[21-Aug-2009 14:25:03] <mrayzenoss> I'd say take one of the existing report ZenPacks and replace the .rpt file in there
[21-Aug-2009 14:25:11] <rmatte> I've modified the memory.py, and cpu.py files under report plugins
[21-Aug-2009 14:25:11] <straterra> trac-zenpacks/browser/zenpacks/ZenPacks.community.PrinterToner/ZenPacks/community/PrinterToner/PrinterToner.py
[21-Aug-2009 14:25:21] <rmatte> and I intend to modify the rpt files themselves at some point
[21-Aug-2009 14:25:48] <rmatte> yeh, I'll look in to doing something like that
[21-Aug-2009 14:26:35] <geniusj> Hmm.. When I try to add a Service, the port field doesn't show up
[21-Aug-2009 14:26:49] <rmatte> geniusj: an IP service?
[21-Aug-2009 14:27:23] <geniusj> rmatte, yea. under registered
[21-Aug-2009 14:27:39] <rmatte> geniusj: oh, you're trying to actually add one to the list, or to a device?
[21-Aug-2009 14:27:42] <geniusj> it does show up for existing registered services.. just not this new one
[21-Aug-2009 14:27:45] <geniusj> add one to the list
[21-Aug-2009 14:28:41] <markeriv> What is the best ping uptime monitoring zenpack?
[21-Aug-2009 14:28:47] <markeriv> iyo?
[21-Aug-2009 14:29:17] <straterra> rmatte: its funny..cause under PrinterToner it shows the max toner o.O
[21-Aug-2009 14:29:21] <geniusj> I guess I can just add it from /manage, but I figured I'm missing something
[21-Aug-2009 14:30:09] <rmatte> geniusj: hmmm, you're right
[21-Aug-2009 14:30:15] <rmatte> just tried it myself
[21-Aug-2009 14:30:20] <straterra> And the plugin seems to work without the template..and it shows a warning about low toner on this printer
[21-Aug-2009 14:30:37] <straterra> I'd just like to adjust when it triggers warning/error
[21-Aug-2009 14:30:58] <straterra> The only place I saw to do that is the template..but now the template isn't loaded for that device..but its getting those values from somewhere
[21-Aug-2009 14:34:15] <rmatte> hmmm, I might have to log a trac case for this, that's weird
[21-Aug-2009 14:34:33] <rmatte> when you add a new IP Service under Registered it doesn't allow you to set the port of the service
[21-Aug-2009 14:34:57] <rmatte> oh wait, I might know why
[21-Aug-2009 14:35:31] <geniusj> err
[21-Aug-2009 14:35:32] <geniusj> fixed
[21-Aug-2009 14:35:33] <geniusj> lol
[21-Aug-2009 14:35:37] <geniusj> Add IP Service
[21-Aug-2009 14:35:40] <geniusj> I clearly wasn't looking
[21-Aug-2009 14:35:42] <rmatte> yeh, you clicked the wrong link
[21-Aug-2009 14:35:42] <rmatte> lol
[21-Aug-2009 14:35:46] <rmatte> I just noticed that
[21-Aug-2009 14:38:09] <mrayzenoss> geniusj: does your command show up when you run zencommand run -v10?
[21-Aug-2009 14:38:17] <geniusj> mrayzenoss, no
[21-Aug-2009 14:38:45] <geniusj> I'm about to move it from device to service though.. and failing that, I"ll recreate it as you suggested, starting with 1 property
[21-Aug-2009 14:39:04] <geniusj> err.. data point
[21-Aug-2009 14:41:40] <mrayzenoss> geniusj: so if you run zencommand run -v10 -d yourmachine, there's nothing in the log?
[21-Aug-2009 14:43:28] <geniusj> http://pastebin.com/m573d7ff4
[21-Aug-2009 14:44:44] <mrayzenoss> yup, that's empty
[21-Aug-2009 14:51:51] <mrayzenoss> geniusj: does your script call $devname?
[21-Aug-2009 14:53:03] <cgibbons> hmm
[21-Aug-2009 15:02:37] <geniusj> the script itself?
[21-Aug-2009 15:02:52] <geniusj> or the command definition?
[21-Aug-2009 15:05:27] <geniusj> /opt/zenoss/common/libexec/memcachedb.py --port ${here/cMemcachedbPort} ${here/manageIp}
[21-Aug-2009 15:07:15] <markeriv> hey guys, i got a quick snmp question. I am getting this perf/snmp error
[21-Aug-2009 15:07:53] <markeriv> error reading value for Serial0.1 on IP (oid .1.3.6.1.2.1.2.2.1.20.4 is bad)
[21-Aug-2009 15:08:14] <markeriv> is this the whole needs a 0 at the end bug?
[21-Aug-2009 15:08:40] <markeriv> I have added like 6 1721 ciscos at this point and they are spamming me with this error every 5 mins
[21-Aug-2009 15:08:53] <mrayzenoss> that's the typical culprit
[21-Aug-2009 15:09:18] <markeriv> hmm, where would I go about editing that OID?
[21-Aug-2009 15:09:41] <rmatte> markeriv: that's because that's not a physical serial interface, it's a sub-interface
[21-Aug-2009 15:09:54] <rmatte> it's trying to read the errors but it can't since there are no errors, since it's a virtual interface
[21-Aug-2009 15:10:04] <markeriv> aha
[21-Aug-2009 15:10:17] <rmatte> you need to create a local copy of the interface template and delete the errors datapoint
[21-Aug-2009 15:10:31] <markeriv> sweet beans
[21-Aug-2009 15:10:41] <rmatte> you'll have to do that for every one of those interfaces if you have more
[21-Aug-2009 15:10:52] <rmatte> I had to do it for like 50 devices
[21-Aug-2009 15:10:54] <markeriv> *cries*
[21-Aug-2009 15:10:56] <rmatte> on one of my zenoss boxes
[21-Aug-2009 15:11:02] <markeriv> i got like
[21-Aug-2009 15:11:06] <markeriv> over 1k of these suckers
[21-Aug-2009 15:11:29] <markeriv> lol, back to the chain gang i suppose
[21-Aug-2009 15:11:31] <mrayzenoss> so some have the error and some don't?
[21-Aug-2009 15:13:03] <markeriv> No, i mean i have over a thousand of these routers deployed
[21-Aug-2009 15:13:11] <markeriv> and each one is having that virtual interface
[21-Aug-2009 15:13:23] <markeriv> maybe i can just not monitor that interface... hmmm
[21-Aug-2009 15:15:44] <rmatte> mrayzenoss: zenoss doesn't differentiate between physical serial interfaces and virtual
[21-Aug-2009 15:16:03] <rmatte> but you still want to collect error info for your physical
[21-Aug-2009 15:16:21] <rmatte> so you need to make local copies of the template for every single virtual interface and remove the errors datasource
[21-Aug-2009 15:16:25] <rmatte> it's very tedious
[21-Aug-2009 15:16:40] <rmatte> once it's setup it works fine, but it's pretty brutal to have to do
[21-Aug-2009 15:16:50] <mrayzenoss> seems like something you should be able to do in zendmd
[21-Aug-2009 15:17:12] <jb> hey rmatte
[21-Aug-2009 15:17:23] <jb> remember the other day when we were wondering about the RRD stuff?
[21-Aug-2009 15:17:27] <jb> storing values in RRD cache..
[21-Aug-2009 15:17:31] <rmatte> yeh
[21-Aug-2009 15:17:37] <markeriv> hmm, is there a way i can just choose to ignore that particular error?
[21-Aug-2009 15:17:39] <jb> well
[21-Aug-2009 15:17:46] <markeriv> so that it is not saved or logged globally?
[21-Aug-2009 15:18:01] <rmatte> mrayzenoss: maybe, I'll have to look in to it, it wouldn't be easy though
[21-Aug-2009 15:18:04] <jb> you do that when you want things in your modeler like "Status" or "Health" or something to be checked more often than 12 hours (default modeler time)
[21-Aug-2009 15:18:15] <rmatte> ah
[21-Aug-2009 15:18:15] <jb> since RRDs are generated every 5 minutes
[21-Aug-2009 15:18:21] <rmatte> gotcha
[21-Aug-2009 15:18:34] <jb> just fyi
[21-Aug-2009 15:24:09] <markeriv> hmm crup, there are interfaces on this router that i do need to monitor with that so i cannot just delete that graph
[21-Aug-2009 15:24:10] <markeriv> ponder
[21-Aug-2009 15:32:27] <markeriv> hmm well i figured out how to do what you suggested
[21-Aug-2009 15:32:30] <markeriv> at least now i can stop the flow
[21-Aug-2009 15:37:20] <jb> so if i add a parent device class to a zenpack
[21-Aug-2009 15:37:27] <jb> that will add all children classes too, right?
[21-Aug-2009 15:39:00] <markeriv> wootah, figured out a way to just not monitor that interface
[21-Aug-2009 15:39:27] <markeriv> same deal essentially, and more easily educatable so mr. script can do it himself
[21-Aug-2009 15:42:30] <rmatte> lovely, when I run my cpu report it looks fine but when I export it it somehow magically comes up with different values
[21-Aug-2009 15:43:35] <rmatte> omg, I think I know why...
[21-Aug-2009 15:43:47] <rmatte> it's like it's exporting between the default start and end time
[21-Aug-2009 15:43:59] <rmatte> even though a different start and end time were specified
[21-Aug-2009 15:48:05] <rmatte> it's like the export is completely ignoring the options
[21-Aug-2009 15:49:03] <markeriv> thats not comfy
[21-Aug-2009 15:50:29] <markeriv> is it happening for all reports?
[21-Aug-2009 15:50:32] <rmatte> no, it sucks, because now my boss expects to fix it
[21-Aug-2009 15:50:34] <markeriv> or just cpu?
[21-Aug-2009 15:50:36] <rmatte> good question...
[21-Aug-2009 15:52:07] <markeriv> and in 2.4.2 or 2.4.3?
[21-Aug-2009 15:52:34] <rmatte> this particular box is 2.3.3
[21-Aug-2009 15:52:36] <rmatte> so I have no idea
[21-Aug-2009 15:52:55] <rmatte> I haven't moved any of these custom reports to other boxes yet
[21-Aug-2009 15:53:11] <rmatte> well, modified reports
[21-Aug-2009 15:53:17] <rmatte> though my modifications wouldn't cause this
[21-Aug-2009 15:53:20] <markeriv> try creating a new custom report with the CPU field?
[21-Aug-2009 15:54:03] <rmatte> no, I just added more datapoints to the report plugin
[21-Aug-2009 15:54:13] <rmatte> I didn't even touch the report file itself
[21-Aug-2009 15:54:33] <markeriv> got a backup of the original you can try?
[21-Aug-2009 15:55:14] <rmatte> I have the original on other servers, but like I said, my modifications can't have caused this
[21-Aug-2009 15:55:19] <rmatte> it's literally impossible
[21-Aug-2009 15:55:59] <markeriv> hmm
[21-Aug-2009 15:56:48] <rmatte> yup, it does the exact same thing on a report I haven't even touched
[21-Aug-2009 15:56:56] <rmatte> that's really bad
[21-Aug-2009 15:56:57] <rmatte> lol
[21-Aug-2009 15:57:10] <rmatte> new trac ticket
[21-Aug-2009 15:57:37] <rmatte> does no one ever use reports or something?
[21-Aug-2009 15:57:43] <rmatte> I'm finding bug after bug after bug lol
[21-Aug-2009 15:58:03] <ckrough> Im out. have a good weekend
[21-Aug-2009 16:10:54] <mrayzenoss> free Zenoss training at Ohio LinuxFest: http://blog.zenoss.com/2009/08/21/zenoss-community-day-september-25th/
[21-Aug-2009 16:12:20] <straterra> Boss thought it was pretty cool I'm graphing one of the printers now
[21-Aug-2009 16:12:23] <straterra> I got that zenpack to work
[21-Aug-2009 16:12:28] <mrayzenoss> cool
[21-Aug-2009 16:12:32] <rmatte> I think I'm going to end up just writing scripts to generate reports
[21-Aug-2009 16:12:48] <jb> ibm3584 zenpack is nearly complete
[21-Aug-2009 16:12:51] <jb> due to MUCH help from egor
[21-Aug-2009 16:12:52] <straterra> I think I'll write a report on toner usage and such
[21-Aug-2009 16:13:40] <jb> well, its mostly inventory/modeled information
[21-Aug-2009 16:13:42] <jb> oops
[21-Aug-2009 16:15:19] <markeriv> what other bugs are you finding rmatte?
[21-Aug-2009 16:16:04] <geniusj> mrayzenoss: the target class for my memcachedb template is Products.ZenModel.Device - is that fine?
[21-Aug-2009 16:16:20] <rmatte> besides the bug with exporting you mean?
[21-Aug-2009 16:17:22] <mrayzenoss> geniusj: sounds fine to me, as long as it's bound to the device classes you want
[21-Aug-2009 16:19:44] <rmatte> markeriv: I'm prepping some trac tickets right now, I'll paste them once I have them logged
[21-Aug-2009 16:20:11] <jb> so zenoss finally got some street cred today at work..
[21-Aug-2009 16:20:18] <jb> helped pinpoint a few MSSQL problems
[21-Aug-2009 16:21:44] <jb> (which MOM didnt)
[21-Aug-2009 16:22:09] <mrayzenoss> yay us!
[21-Aug-2009 16:23:06] <markeriv> cool
[21-Aug-2009 16:23:17] <markeriv> woot go Z
[21-Aug-2009 16:27:07] <rmatte> mrayzenoss: http://dev.zenoss.org/trac/ticket/5438
[21-Aug-2009 16:28:09] * mrayzenoss sighs
[21-Aug-2009 16:28:16] <markeriv> ?
[21-Aug-2009 16:28:31] <mrayzenoss> I just know how swamped all the devs are
[21-Aug-2009 16:28:36] <mrayzenoss> and we only have 5
[21-Aug-2009 16:29:05] <rmatte> yeh, I understand that part
[21-Aug-2009 16:31:53] <jb> hmm
[21-Aug-2009 16:31:56] <jb> so i have one instance of zenoss that has like 30 transforms
[21-Aug-2009 16:32:09] <jb> i need to take these and put them onto another instance of zenoss and integrate them into another zenpack
[21-Aug-2009 16:32:31] <jb> what would you say is the best way to do that? create a temporary zenpack and put the transforms in it?
[21-Aug-2009 16:33:24] <rmatte> http://dev.zenoss.org/trac/ticket/5439
[21-Aug-2009 16:33:29] <rmatte> that's another one I noticed today
[21-Aug-2009 16:34:09] <mrayzenoss> I only opened this one today: http://dev.zenoss.org/trac/ticket/5437
[21-Aug-2009 16:34:39] <rmatte> ah, that's a good one
[21-Aug-2009 16:34:45] <rmatte> probably a relatively quick fix too
[21-Aug-2009 16:34:57] <rmatte> just need to add a check in to the code
[21-Aug-2009 16:35:13] <mrayzenoss> yeah, cgibbons is refactoring the daemons a bit to share more code
[21-Aug-2009 16:35:22] <mrayzenoss> so it may get fixed for King Crab
[21-Aug-2009 16:36:18] <rmatte> the report export problem is pretty brutal, my boss is pretty upset about that so I'll need to come up with some other solution
[21-Aug-2009 16:36:42] <rmatte> I think I'll just write scripts that take arguments and dump the data as csv files as a workaround for the time being
[21-Aug-2009 16:37:45] <rmatte> unless I can figure out how to fix it
[21-Aug-2009 16:38:38] <markeriv> when ya fix it wanna let me know? lol.
[21-Aug-2009 16:38:49] <markeriv> hmm wonder if it affects my version to
[21-Aug-2009 16:38:54] <rmatte> if I do fix it I'll be submitting a patch
[21-Aug-2009 16:39:05] <markeriv> cool
[21-Aug-2009 16:39:07] <rmatte> I tested on 2.4.1 and it's the same issue
[21-Aug-2009 16:39:13] <rmatte> so I'd imagine you're affected as well
[21-Aug-2009 16:39:14] <markeriv> bleh
[21-Aug-2009 16:39:16] <rmatte> test it out
[21-Aug-2009 16:39:36] <rmatte> generate an interface utilization report, it'll show default values (which is average over the last 7 days)
[21-Aug-2009 16:39:44] <rmatte> then change the start and end time until the values are different
[21-Aug-2009 16:39:56] <rmatte> then export, and you should see the same values as you saw by default in the exported report
[21-Aug-2009 16:40:04] <rmatte> instead of the values shown in your browser
[21-Aug-2009 16:40:24] <rmatte> you can do it with cpu report, memory report, whatever
[21-Aug-2009 16:40:32] <rmatte> as long as you can get it to display different values
[21-Aug-2009 16:41:20] <rmatte> my boss called me over and was like "uhhh, Ryan, I don't think I can trust these numbers."
[21-Aug-2009 16:41:30] <rmatte> lol
[21-Aug-2009 16:43:17] <markeriv> bleh
[21-Aug-2009 16:43:27] <markeriv> i screwed up my data collection by moving devices down a step
[21-Aug-2009 16:43:34] <markeriv> think i may have fixed it but still meh
[21-Aug-2009 16:43:44] <rmatte> "down a step"?
[21-Aug-2009 16:44:04] <markeriv> yea
[21-Aug-2009 16:44:07] <markeriv> basically i took
[21-Aug-2009 16:44:12] <rmatte> like to a different organizer?
[21-Aug-2009 16:44:16] <markeriv> Devices/Network/Router/Cisco
[21-Aug-2009 16:44:24] <markeriv> and added /On-Net Product
[21-Aug-2009 16:44:29] <rmatte> ah
[21-Aug-2009 16:44:29] <markeriv> and all the graphs stopped working
[21-Aug-2009 16:44:42] <markeriv> half tempted just to move it back and say screw it
[21-Aug-2009 16:44:43] <rmatte> well yeh, because the template wouldn't be applied to that organizer by default
[21-Aug-2009 16:44:46] <markeriv> yea
[21-Aug-2009 16:44:46] <rmatte> you need to manually do that
[21-Aug-2009 16:45:03] <markeriv> i copied the template into the class
[21-Aug-2009 16:45:21] <rmatte> instead of copying you really should just bind the template from higher up to the class
[21-Aug-2009 16:45:28] <rmatte> that way you only have to update the template in 1 place
[21-Aug-2009 16:45:35] <markeriv> hmm nod
[21-Aug-2009 16:46:25] <markeriv> hmm, i am in devices/network/cisco
[21-Aug-2009 16:46:34] <markeriv> when i select the device template and go to bind templates
[21-Aug-2009 16:46:39] <markeriv> it only has the current directory
[21-Aug-2009 16:46:52] <markeriv> however I can copy it into the new directory /On-Net
[21-Aug-2009 16:46:53] <rmatte> hmmm
[21-Aug-2009 16:47:02] <markeriv> and from there bind it into the new directory
[21-Aug-2009 16:47:56] <rmatte> shows up fine for me
[21-Aug-2009 16:48:26] <markeriv> hmmm
[21-Aug-2009 17:06:14] <geniusj> mrayzenoss, are you out of annapolis? or are you guys all over the place?
[21-Aug-2009 17:06:41] <geniusj> I used to work with some of the original people there when I was at USi
[21-Aug-2009 17:06:49] <rmatte> they have an office in Austin as well
[21-Aug-2009 17:06:58] <geniusj> ah ok
[21-Aug-2009 17:06:58] <rmatte> and employees from all over
[21-Aug-2009 17:07:04] <rmatte> who work remote
[21-Aug-2009 17:07:10] <mrayzenoss> I'm in Austin
[21-Aug-2009 17:07:35] <mrayzenoss> 3.5/5 of the developers are in Austin
[21-Aug-2009 17:07:36] <rmatte> the austin office looks really nice
[21-Aug-2009 17:07:39] <geniusj> they wrote USi's monitoring system and then left to start ZenOSS
[21-Aug-2009 17:07:40] <rmatte> I want to steal your office
[21-Aug-2009 17:07:52] <mrayzenoss> cgibbons is in Houston but he comes to Austin a lot
[21-Aug-2009 17:08:29] <mrayzenoss> Ian is in Canada
[21-Aug-2009 17:08:51] <geniusj> no annapolis respresented on irc?
[21-Aug-2009 17:08:52] <mrayzenoss> they kinda want to keep developers in-house if possible
[21-Aug-2009 17:09:19] <mrayzenoss> well, Chet is in Annapolis and he's in here occasionally
[21-Aug-2009 17:09:33] <rmatte> yeh, chet hasn't been around much lately, I assume because he's busy
[21-Aug-2009 17:09:46] <mrayzenoss> npmccallum is in Kentucky, he works remote
[21-Aug-2009 17:10:05] <mrayzenoss> yeah, not sure what Chet's been up to, he may be on vacation
[21-Aug-2009 17:10:09] <rmatte> mrchippy is in austin I assume
[21-Aug-2009 17:10:25] <mrayzenoss> Yeah, mrchippy, jplouis, bedwards and cgibbons
[21-Aug-2009 17:10:59] <mrayzenoss> and the QA Manager and our build engineer
[21-Aug-2009 17:11:18] <mrayzenoss> most support and client services is in Annapolis or remote
[21-Aug-2009 17:16:55] <geniusj> holy crap
[21-Aug-2009 17:17:00] <geniusj> it's working
[21-Aug-2009 17:17:19] <geniusj> 1 monitor down, a bunch to go.. lol
[21-Aug-2009 17:17:28] <mrayzenoss> what's working? The Cacti plugin?
[21-Aug-2009 17:17:46] <geniusj> yea
[21-Aug-2009 17:17:51] <geniusj> I restarted zenoss
[21-Aug-2009 17:17:56] <geniusj> :-\
[21-Aug-2009 17:19:50] <geniusj> ok.. not quite WORKING yet.. but I am getting feedback.. it's at least running it
[21-Aug-2009 17:19:55] <geniusj> so it should be working shortly
[21-Aug-2009 17:20:45] <rmatte> alright, I'm out for the night, have a nice weekend guys
[21-Aug-2009 17:22:20] <geniusj> zenmodeler heartbeat failure
[21-Aug-2009 17:22:20] <geniusj> that's new :-D
[21-Aug-2009 17:22:39] <geniusj> rmatte, see ya
[22-Aug-2009 00:00:46] [disconnected at Sat Aug 22 00:00:46 2009]
[22-Aug-2009 00:00:46] [connected at Sat Aug 22 00:00:46 2009]
[22-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[22-Aug-2009 05:45:14] <prefixthis_> hello
[22-Aug-2009 05:45:52] <prefixthis_> is it possible to have zenoss not send all the syslog output from devices directly to the event history?
[22-Aug-2009 10:52:49] <prefixthis> i have syslog output from a switch into zenoss. zenoss delays the link down syslog messages until the interface comes back up, and instead displays a linkdown trap, and everything gets moved to the history after the interfaces comes back up
[22-Aug-2009 10:54:27] <prefixthis> and the trap itself only displays "snmp trap snmp_linkDown", without the port number (i have to check out the details to get that far). Anyone can give me a hint?
[22-Aug-2009 11:12:36] <Belial_>
[23-Aug-2009 00:00:46] [disconnected at Sun Aug 23 00:00:46 2009]
[23-Aug-2009 00:00:46] [connected at Sun Aug 23 00:00:46 2009]
[23-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[23-Aug-2009 04:33:35] <diazepam> hi all - im new to Zenoss and have just finished reading the docs but am still having trouble getting snmp on my ubuntu server to communicate with my zenoss system
[23-Aug-2009 04:34:26] <diazepam> snmpwalk -v1 -c rocommunity localhost - command works fine from the ubuntu server but
[23-Aug-2009 04:34:48] <diazepam> snmpwalk -v1 -c rocommunity 10.73.75.100 - times out from the zenoss machine
[23-Aug-2009 05:30:44] <diazepam> okay sorted that one out
[23-Aug-2009 05:30:59] <diazepam> needed to add my bridge interface to the snmpd conf
[23-Aug-2009 16:36:44] <darkest_night> hi @ all
[23-Aug-2009 16:37:01] <darkest_night> i have a question... how do i come to the zenoss-monitor?
[23-Aug-2009 16:37:12] <darkest_night> do i need any special website to enter?
[24-Aug-2009 00:00:46] [disconnected at Mon Aug 24 00:00:46 2009]
[24-Aug-2009 00:00:46] [connected at Mon Aug 24 00:00:46 2009]
[24-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[24-Aug-2009 07:26:07] <Disconnect> aahgrhght. wtf does it take to get zenoss alerting rules to stay disabled? i go in, set it to disabled, hit save. back to the main list, its disabled. but by tomorrow morning it'll be on again..
[24-Aug-2009 08:54:52] <rmatte> Disconnect: that's strange, never heard of that behaviour before
[24-Aug-2009 08:55:05] <rmatte> good morning Matt
[24-Aug-2009 08:55:39] <jb> so you cant add an entire event class to zenpack?
[24-Aug-2009 08:55:40] <jb> hrm.
[24-Aug-2009 08:55:45] <Disconnect> disabled it 3 times last week, came in to find it on again this morning. tried removing the disabled ones so we'll see what happens (it'll be a while anyway before zenoss is cleared to send mail.. its paging on discovered devices
..)[24-Aug-2009 08:56:25] <jb> thats sort of annoying
[24-Aug-2009 09:00:32] <rmatte> Disconnect: add a filter to the alert rule not to alert on devices in device class /Discovered
[24-Aug-2009 09:00:34] <rmatte> that's what we did
[24-Aug-2009 09:00:52] <rmatte> alert filters are your friend
[24-Aug-2009 09:04:51] <scp2004> Hi Guys !!!!
[24-Aug-2009 09:07:05] <scp2004> I want to monitore a Linux Server in Zenoss .. And Collect some snmp informations .. How should I proceed ?
[24-Aug-2009 09:07:33] <scp2004> I installed the Zenpacks HPMonitor and DellMonitor . Its right ?
[24-Aug-2009 09:07:50] <jb> yes.. but you need to have OpenManage and Insight installed on those servers as well
[24-Aug-2009 09:08:34] <scp2004> I think also have. Cause they was monitored buy another zenoss ..And was ok
[24-Aug-2009 09:09:26] <scp2004> I add the device , a FreeBsd Linux Server running on a HP server in: /Devices /Server /Linux
[24-Aug-2009 09:09:55] <Disconnect> yah. we've got other sorting issues right now (old hosts that aren't doing snmp etc are filtered into their proper slots, etc) but yah. and there -really- needs to be a "copy this alert template"
[24-Aug-2009 09:14:13] <scp2004> And In /Devices /Server /Linux now i can see in zCollectorPlugins : HPCpuMap, DellCpuMap .... Its right now ?
[24-Aug-2009 09:14:21] <rocket> morning matt
[24-Aug-2009 09:14:30] <jb> yawn
[24-Aug-2009 09:14:36] <rocket> Hey jb
[24-Aug-2009 09:14:42] <jb> so when you add /Devices/Storage/IBM to a zenpack
[24-Aug-2009 09:14:51] <jb> hey rocket
[24-Aug-2009 09:15:03] <jb> WARNING:zen.AddToPack:Unable to find context path /zport/dmd/Devices/Storage (line 4 ?) for IBM
[24-Aug-2009 09:15:03] <jb> WARNING:zen.AddToPack:Not committing any changes
[24-Aug-2009 09:15:03] <jb> ERROR: zenpack command failed. Reason: Unable to create object using the following attributes:
[24-Aug-2009 09:15:03] <jb> * id: /zport/dmd/Devices/Storage/IBM
[24-Aug-2009 09:15:12] <jb> it wont just create it if it doesn't exist?
[24-Aug-2009 09:15:29] <mrayzenoss> morning all
[24-Aug-2009 09:16:17] <jb> or am I not adding it correctly?
[24-Aug-2009 09:16:42] <mrayzenoss> jb: is there already a /Devices/Storage class?
[24-Aug-2009 09:16:58] <jb> oh.
[24-Aug-2009 09:17:02] <jb> core doesn't have that
[24-Aug-2009 09:17:13] <mrayzenoss> nope
[24-Aug-2009 09:17:26] <jb> so my goal is to add /Devices/Storage/IBM/Tape Library
[24-Aug-2009 09:17:28] <mrayzenoss> it gets added by one of the zenpacks along the way
[24-Aug-2009 09:17:36] <jb> it wont just recursively create those classes?
[24-Aug-2009 09:19:27] <rocket> jb: nope
[24-Aug-2009 09:19:32] <jb> wow, thats annoying
[24-Aug-2009 09:19:41] <jb> so you have to tell your users to create them ahead of time?
[24-Aug-2009 09:19:47] <rocket> jb: I would create a core storage zenpack first
[24-Aug-2009 09:20:13] <rocket> jb: or you can add code to the init function that checks for it .. and if it doesnt exist create it
[24-Aug-2009 09:20:24] <jb> is that what you did for AIX?
[24-Aug-2009 09:21:02] <mrayzenoss> jb: yeah, I just put that as a prerequisite
[24-Aug-2009 09:21:17] <jb> a manual prereq?
[24-Aug-2009 09:22:33] <jb> i bet my zPythonClass zprop wont get exported either
[24-Aug-2009 09:22:37] <jb> so i'll need to set that in init
[24-Aug-2009 09:23:19] <mrayzenoss> jb: yeah, some ZenPacks handle creating their own device classes, some don't
[24-Aug-2009 09:23:30] <mrayzenoss> usually if you're going 2 or 3 deep, you'll have to create something
[24-Aug-2009 09:23:43] <jb> ok, I think I'll just do it in __init__
[24-Aug-2009 09:23:49] <jb> seems like a cleaner way
[24-Aug-2009 09:23:51] <mrayzenoss> in this case, you'll require that they add Storage, so it doesn't get deleted when they delete the ZenPack
[24-Aug-2009 09:24:18] <jb> hm, good point
[24-Aug-2009 09:26:46] <jb> ok, got another question
[24-Aug-2009 09:27:19] <jb> i have 30 or so odd transforms on one install, and I need to move those to another zenoss instance where I am packaging this zenpack.. whats the easiest way to do that?
[24-Aug-2009 09:31:00] <mrayzenoss> export them to another ZenPack?
[24-Aug-2009 09:31:06] <jb> yeah thats what i was thinking
[24-Aug-2009 09:31:18] <mrayzenoss> the '?' is because I've never done that
[24-Aug-2009 09:31:26] <jb> i think thats the best way
[24-Aug-2009 09:31:30] <jb> create a "temp" zenpack
[24-Aug-2009 09:31:43] <jb> install that on the dev instance, and add them to the production zenpack
[24-Aug-2009 09:36:12] <jb> hmm
[24-Aug-2009 09:37:02] <jb> WARNING:zen.AddToPack:Unable to find context path /zport/dmd/Events/HW/Backup/Tape Library/instances (line 4 ?) for alertChangerMechHWCritical
[24-Aug-2009 09:38:18] <jb> /Events/HW/Backup/Tape Library/ does exist
[24-Aug-2009 09:38:35] <jb> <jb> /Events/HW/Backup/Tape Library/ does exist
[24-Aug-2009 09:38:37] <jb> oops
[24-Aug-2009 09:38:40] <jb> ERROR: zenpack command failed. Reason: Unable to create object using the following attributes:
[24-Aug-2009 09:38:40] <jb> * id: /zport/dmd/Events/HW/Backup/Tape Library/instances/alertChangerMechHWCritical
[24-Aug-2009 09:39:41] <malbon> hi folks
[24-Aug-2009 09:39:56] <jb> but im guessing "instances" does not?
[24-Aug-2009 09:40:15] <malbon> I would like to take zenprocess and use it with a non-standard mib, any one have some words of wisdom to help me do this?
[24-Aug-2009 09:45:14] <jb> i guess that instances gets created when you add the first event mapping?
[24-Aug-2009 09:52:32] <rmatte> malbon: zenprocess has nothing to do with Mibs
[24-Aug-2009 09:52:56] <rmatte> malbon: The only daemon that uses Mibs in zentrap, to translate snmp traps
[24-Aug-2009 09:53:18] <rmatte> is zentrap* rather
[24-Aug-2009 10:23:34] <jb> cgibbons: would you know how I can add a device class/organizer (not using the GUI)?
[24-Aug-2009 10:23:46] <jb> i was thinking app.zport.dmd.Devices.addOrganizer...
[24-Aug-2009 10:25:51] crow is now known as ckrough
[24-Aug-2009 10:32:40] <malbon> rmatte: not entirely true as zenprocess uses the HostResources HRSWRun MIB for it's collection points.
[24-Aug-2009 10:33:05] <malbon> rmatte: I want to use an alternative set of OIDs within zenprocess (or a lookalike)
[24-Aug-2009 10:33:38] <jb> anyone?
[24-Aug-2009 10:43:39] <nzle> How does one install a perl module such as Net::Telnet on a Zenoss appliance? conary doesnt find it, and cpan is unable to build it. since there are not dev tools on it and conary seems unable to install them as well.
[24-Aug-2009 10:46:38] <nzle> I need to write a special purpose nagios style plugin that connects on a port, does some dialog, and interpets the response.
[24-Aug-2009 10:48:07] <mrayzenoss> nzle: if you could post that on zenoss-dev, I'll get our build engineer to respond. He's busy pushing out 2.4.3 right now, but if anyone knows the answer, it would be him
[24-Aug-2009 10:48:31] <mrayzenoss> nzle: I searched around some for it, but if it's not in conary I'm not sure how to add the trove
[24-Aug-2009 10:48:39] <nzle> Ok, Ill try that...
[24-Aug-2009 11:05:24] <scp2004> [mrayzenoss] Hi ..To collect snmp from a server in linux i need to install Community Zenpacks or Core Zenpacks ? I need snmp from FreeBsd devices in HP Server. So I installed Core Zenpack HPMonitor . Its right ?
[24-Aug-2009 11:07:44] <mrayzenoss> nah, SNMP works just fine for Linux out of the box. Just put it under /Server/Linux
[24-Aug-2009 11:10:58] <scp2004> [mrayzenoss] But , i also put. in /Server/Linux . Now I have the event: SNMP agent down
[24-Aug-2009 11:11:21] <scp2004> [mrayzenoss] but i know that its not down. Cause i have another zenoss monitoring it
[24-Aug-2009 11:17:28] <mrayzenoss> scp2004: is the community string correct?
[24-Aug-2009 11:18:02] <scp2004> [mrayzenoss] i put public . in other zenoss are public and its working
[24-Aug-2009 11:18:22] <mrayzenoss> scp2004: try re-modeling the device
[24-Aug-2009 11:18:25] <rocket> scp2004: check the zenoss logs for errors
[24-Aug-2009 11:19:21] <scp2004> [rocket] re-modeling ? how ido this ?
[24-Aug-2009 11:20:49] <mrayzenoss> go to the status page for that device, and select the page menu->Manage->Model Device
[24-Aug-2009 11:21:01] <mrayzenoss> that will show you the modeling run and any errors
[24-Aug-2009 11:22:20] <jb> <object id='/zport/dmd/Events/HW/Backup/Tape Library/instances/alertChangerMechHWDiag' module='Products.ZenEvents.EventClassInst' clas
[24-Aug-2009 11:22:21] <jb> oops
[24-Aug-2009 11:23:32] <scp2004> [mrayzenoss] doing ...
[24-Aug-2009 11:56:26] <venturaville> anyone got an example of uploading an icon through a zenpack for a new device class (without using hard paths)
[24-Aug-2009 11:59:07] <jb> working on that 9509 pack?
[24-Aug-2009 12:00:30] <jb> btw, do you just look in the dev guide for this stuff?
[24-Aug-2009 12:05:15] <rmatte> for starters, but some stuff isn't in the dev guide
[24-Aug-2009 12:05:18] <rmatte> it's not 100% yet
[24-Aug-2009 12:08:12] <venturaville> Cisco 9509?
[24-Aug-2009 12:08:18] <jb> yah
[24-Aug-2009 12:09:27] <venturaville> got a link to what you are working on?
[24-Aug-2009 12:09:52] <rocket> jb: community zenpacks are a great place to look for things too
[24-Aug-2009 12:10:11] <rocket> jb: eg I have them checked out to a directory so I can look at all of them
[24-Aug-2009 12:10:13] <jb> yeah, i usually just grep against the entire svn tree
[24-Aug-2009 12:10:14] <jb>
[24-Aug-2009 12:23:04] <ckrough> Is there still an issue with upgrading/replacing zenpacks? Coworker remembers there being an issue with overwriting zenpacks with newer versions of the same zenpack.
[24-Aug-2009 12:23:42] <ckrough> ie, I create a zenpack, deploy it, then add some more functionality and want to deploy the updated one on servers that have the original. That shouldn't cause issues, right?
[24-Aug-2009 12:24:34] <venturaville> rmatte: I've got a bunch of 9509s if you need some help.........
[24-Aug-2009 12:26:21] <mrayzenoss> ckrough: there was more of an issue with Zip zenpacks
[24-Aug-2009 12:26:45] <ckrough> mrayzenoss: that's confirmed as resolved with eggs?
[24-Aug-2009 12:27:21] <mrayzenoss> yeah, I mean assuming the issue is related to the fact that .zip zenpacks didn't install to a zenpacks directory, they dumped junk into the Products folder
[24-Aug-2009 12:27:31] <ckrough> gotcha
[24-Aug-2009 12:27:32] <mrayzenoss> which didn't play well with lots of things
[24-Aug-2009 12:27:45] <mrayzenoss> zenoss upgrades would blow up zenpacks
[24-Aug-2009 12:27:47] <ckrough> thanks
[24-Aug-2009 12:29:32] <dranch> Rmatte, you in?
[24-Aug-2009 12:30:39] <dranch> I need to figure out why specific Zenoss COMMAND tests aren't running and why zencommand doesn't log both the working tests and the non-working tests!
[24-Aug-2009 12:31:14] <mrayzenoss> nzle: Glenn posted a response to your thread
[24-Aug-2009 12:54:50] <rmatte> dranch: I'm in but I'm pretty busy so I'm watching the channel on and off
[24-Aug-2009 12:55:28] <rmatte> dranch: tests as in, fill in device name and click "Test"?
[24-Aug-2009 13:03:23] <markeriv> holy cow
[24-Aug-2009 13:03:26] <markeriv> i am still in here O.o
[24-Aug-2009 13:03:57] <dranch> hey Rmatte: the manual tests have always worked
[24-Aug-2009 13:04:27] <rmatte> but they aren't now?
[24-Aug-2009 13:05:08] <rmatte> "<dranch> I need to figure out why specific Zenoss COMMAND tests aren't running"
[24-Aug-2009 13:05:17] <dranch> they don't seem to run automatically and the logging says it's not running ANY commands for hosts when infact it IS running them for some templates
[24-Aug-2009 13:05:54] <rmatte> quite honestly, beyond the troubleshooting we did last time around I don't really know what to tell you, I've never had issues like that with command based datapoints
[24-Aug-2009 13:05:59] <rmatte> and I use them all the time
[24-Aug-2009 13:06:27] <rmatte> obviously something is messed up with your zencommand, it's either a problem with the daemon itself, with your Zenoss settings, or with the datapoints
[24-Aug-2009 13:06:43] <dranch> yeah.. what's troublesome is the logging. It says it has ZERO commands to run yet in some templates, it shows the results of the tests
[24-Aug-2009 13:07:25] <rmatte> is there any difference in the output or configuration between those datapoints and the ones that aren't running?
[24-Aug-2009 13:08:09] <rmatte> also, how long do the commands take to run through? This might be an issue where zencommand is running out of cycle time.
[24-Aug-2009 13:08:40] <dranch> I don't know if you saw my other comments on IRC on friday but it seems datapoints aren't required to get tests to execute
[24-Aug-2009 13:08:52] <rmatte> datasources
[24-Aug-2009 13:08:55] <rmatte> is what I meant
[24-Aug-2009 13:09:10] <dranch> The zencommand logs say it's idle for like 46 seconds here and there.
[24-Aug-2009 13:09:19] <rmatte> also, I don't see how a datasource can work with no datapoints
[24-Aug-2009 13:09:36] <rmatte> even snmp based datasources have datapoints
[24-Aug-2009 13:09:44] <rmatte> they just get created automatically now and you can't see them
[24-Aug-2009 13:10:57] <dranch> ok.. I guess that would work
[24-Aug-2009 13:12:00] <rmatte> Where did these templates come from?
[24-Aug-2009 13:12:03] <dranch> I also have a screwy maintenance window somewhere that's putting our primary database servers in maintenance mode! It's scary!
[24-Aug-2009 13:12:07] <rmatte> did you make them, or are they from a ZenPack?
[24-Aug-2009 13:12:19] <dranch> I have no idea. I inherited this setup
[24-Aug-2009 13:12:27] <rmatte> dranch: but you can't see the maintenance window in the list?
[24-Aug-2009 13:12:35] <dranch> nope
[24-Aug-2009 13:12:49] <rmatte> sounds like your Zenoss install is quite fubar
[24-Aug-2009 13:12:53] <rmatte> hmmm
[24-Aug-2009 13:13:42] <dranch> it does work but the harder I dig into it.. the more I learn it's not functioning. What I don't understand is that I've added COMMANDs before and they work.. it's just this new one doesn't work on an automatic fashion
[24-Aug-2009 13:13:59] <rmatte> is the maintenance window specifically affecting a group of devices with a common organizer?
[24-Aug-2009 13:14:11] <dranch> yes
[24-Aug-2009 13:14:19] <rmatte> go to that organizer
[24-Aug-2009 13:14:45] <dranch> There are 8 machines in the group. 3 are production and 5 are maintenance mode. I'll come back some time later and all 8 machines are in maintenance mode
[24-Aug-2009 13:15:49] <jb> it would be nice if you could generate a quick PDF based report of perf graphs
[24-Aug-2009 13:15:57] <jb> i get tired of copying and pasting them into emails
[24-Aug-2009 13:17:03] <rmatte> dranch: go to the organizer that contains the services, then click More -> Administration
[24-Aug-2009 13:17:11] <rmatte> check for Maintenance Windows
[24-Aug-2009 13:17:13] <dranch> Organizer --> More --> Administration pull down --> No maintenance windows
[24-Aug-2009 13:17:24] <rmatte> ok, now check each device individually
[24-Aug-2009 13:17:30] <rmatte> check the admin section for each
[24-Aug-2009 13:18:10] <rmatte> someone might have added local maintenance windows to the devices themselves and forgot about them
[24-Aug-2009 13:18:15] <dranch> there IS one for the higher level organizer though the stop state is "original"
[24-Aug-2009 13:18:52] <mrayzenoss> "original" can be troublesome if that's not "Production"
[24-Aug-2009 13:18:59] <dranch> I'll set it to change it back to PRODUCTIOn
[24-Aug-2009 13:19:04] <rmatte> yeh
[24-Aug-2009 13:19:24] <dranch> ok.. back to the COMMAND stuff.
[24-Aug-2009 13:19:51] <dranch> If I run a manual test, shouldn't it be logged in zencommand.log when running with -v 10?
[24-Aug-2009 13:20:13] <rmatte> no, -v10 just makes the output of the command you're running more verbose
[24-Aug-2009 13:20:17] <rmatte> it doesn't guarantee logging
[24-Aug-2009 13:20:51] <dranch> Ok.. so is there a more verbose setting to see what's going on here?
[24-Aug-2009 13:21:01] <rmatte> 10 is the max to my knowledge
[24-Aug-2009 13:21:21] <rmatte> zencommand -v10 --cycle is the best you can do
[24-Aug-2009 13:21:56] <dranch> the manual run works fine but greping comes up empty: tail --lines=10000 zencommand.log | grep -i vip
[24-Aug-2009 13:23:00] <rmatte> zencommand run -v10 --cycle will run the daemon in the foreground
[24-Aug-2009 13:23:04] <rmatte> it should show everything
[24-Aug-2009 13:23:20] <rmatte> I'm running it right now on a box and it's providing plenty of info
[24-Aug-2009 13:24:10] <dranch> Can you see any results from manual runs?
[24-Aug-2009 13:25:03] <rmatte> I see the output of each command run, and how many events are queued as it goes
[24-Aug-2009 13:25:39] <rmatte> DEBUG:zen.zencommand:Total of 2 queued events
[24-Aug-2009 13:25:39] <rmatte> DEBUG:zen.zencommand:Next command in 281 seconds
[24-Aug-2009 13:26:14] <dranch> so if you manually run a given COMMAND and grep for the script name in the zencommand.log file, can you find it?
[24-Aug-2009 13:26:50] <rmatte> when you manually run it I don't think it logs at all
[24-Aug-2009 13:26:56] <rmatte> the logging is for the automatic stuff
[24-Aug-2009 13:27:26] <dranch> Ok.. that's unfortunate. So, is there a way to trace or log why zenoss doesn't think it needs to run this command?
[24-Aug-2009 13:27:34] <rmatte> and it only logs errors
[24-Aug-2009 13:27:44] <rmatte> like, my last log entry is from 4 days ago
[24-Aug-2009 13:27:50] <rmatte> 2009-08-20 17:12:04 WARNING zen.zencommand: Command timed out on device SPS: '$ZENHOME/Products/SNMPWindowsPerfMonitor/libexec/winmem.sh "192.168.10.6"
[24-Aug-2009 13:28:25] <rmatte> but your isn't apparently erroring out, it's just not running at all
[24-Aug-2009 13:28:34] <dranch> exactly
[24-Aug-2009 13:29:35] <rmatte> why don't you try installing Zenoss on another box and trying the exact same commands to see if it's still messed up
[24-Aug-2009 13:29:44] <rmatte> that way you'll know whether it's the commands or your install
[24-Aug-2009 13:30:26] <dranch> that's a pretty time intensive test... It might come down to that to get up to a new release but it's going to take some time to get things all running, etc.
[24-Aug-2009 13:30:55] <dranch> I would have hoped that the logging would have been better than this. To me, logging is #1 and everything else is second
[24-Aug-2009 13:31:04] <rmatte> well, I don't know what else to suggest, it's not automatically running the commands, I believe we've gone over almost every relevant setting, so I'm out of ideas
[24-Aug-2009 13:31:26] <rmatte> when you run with a -v10 you get even more output than the log file
[24-Aug-2009 13:31:34] <rmatte> so run wth -v10, cap it to a file, and see what you get
[24-Aug-2009 13:31:56] <rmatte> and I'd say run the whole thing, not just on one specific device
[24-Aug-2009 13:32:06] <rmatte> zencommand run -v10 --cycle
[24-Aug-2009 13:32:13] <dranch> that's what I'm doing. The script that the COMMAND runs is called check-vip.sh and I can't find the string "vip" in the logs
[24-Aug-2009 13:32:34] <rmatte> yes, but are you also searching the output of the actual command for it?
[24-Aug-2009 13:32:40] <rmatte> forget about the log files and focus on the output
[24-Aug-2009 13:32:45] <dranch> no.. just the script it should be running
[24-Aug-2009 13:32:51] <rmatte> the log files appear to only capture very specific info
[24-Aug-2009 13:33:13] <rmatte> do a zencommand run -v10 --cycle and look for anything odd in the output
[24-Aug-2009 13:33:21] <rmatte> it might show you where it's tripping up
[24-Aug-2009 13:33:37] <rmatte> capture it to a file and comb through it
[24-Aug-2009 13:33:45] <rmatte> the alternative is to do a fresh Zenoss install
[24-Aug-2009 13:33:56] <rmatte> (which you might end up doing anyways)
[24-Aug-2009 13:34:07] <rmatte> did you setup this install to begin with, or did you inherit it from someone else?
[24-Aug-2009 13:34:20] <dranch> inherit..
[24-Aug-2009 13:34:26] <rmatte> yeh, those are always the worst
[24-Aug-2009 13:34:35] <rmatte> since you have no idea what the last person might have messed around with
[24-Aug-2009 13:34:36] <dranch> I would hope that I could fix this..
[24-Aug-2009 13:34:58] <rmatte> they might have gone in and messed with code or something, not noticing that the changes caused other issues
[24-Aug-2009 13:35:16] <rmatte> it could be like finding a needle in a haystack
[24-Aug-2009 13:35:39] <rmatte> just out of curiosity, why are you not able to do an upgrade, to see if that resolves the issues?
[24-Aug-2009 13:36:00] <dranch> yeah.. agreed. With your help on friday, seeing that commands were being issued to other machines but zencommand logging that no commands need to be run was very, um, enlightening
[24-Aug-2009 13:36:18] <dranch> manpower
[24-Aug-2009 13:36:39] <rmatte> is this a single Zenoss instance or multiple?
[24-Aug-2009 13:36:45] <dranch> single
[24-Aug-2009 13:36:51] <rmatte> then why is manpower an issue?
[24-Aug-2009 13:37:04] <rmatte> I'm upgrading 5 Zenoss servers on Tuesday night by myself
[24-Aug-2009 13:37:20] <rmatte> (tomorrow night)
[24-Aug-2009 13:37:52] <rmatte> If an upgrade is done properly, none of your ZenPacks or settings will be affected
[24-Aug-2009 13:37:59] <rmatte> since it's using the same Zope database afterwards
[24-Aug-2009 13:38:15] <dranch> It's somewhat customized with nagios plugins, etc. Have to get the new zenoss, get the dependencies fixed (the RPMs never seem to have worked for the previous admin), identify and install the required nagios plugins, hopefully just restore the DB (I don't have high hopes), etc.
[24-Aug-2009 13:38:33] <rmatte> I see
[24-Aug-2009 13:38:42] <rmatte> yeh, I don't like the RPM installs
[24-Aug-2009 13:38:43] <dranch> this would also be on a different machine and NOT the same machine
[24-Aug-2009 13:38:48] <rmatte> I use stack installer, much nicer
[24-Aug-2009 13:39:01] <dranch> I personally like RPMs but it seems the RPMs for Zenoss aren't built properly
[24-Aug-2009 13:39:34] <dranch> brb
[24-Aug-2009 13:39:51] <rmatte> dranch: yeh, different machine is fairly simple, just do a fresh install of the same version on the new machine, tar up $ZENHOME on the old machine, move it to the new one, start it up and make sure it has all the previous settings and everything
[24-Aug-2009 13:39:53] <rmatte> then from there, just do the upgrade
[24-Aug-2009 13:40:06] <rmatte> (you'd also have to copy over any other files that it depends on which exist outside of $ZENHOME
[24-Aug-2009 13:41:14] <rmatte> like if you have any customs scripts stored in /opt or whatever that Zenoss uses)
[24-Aug-2009 13:41:21] <jb> i never have an issue with the RPM's..
[24-Aug-2009 13:41:29] <jb> not in 2.3+ at least
[24-Aug-2009 13:41:51] <rmatte> RPMs get annoying, whining about dependencies yet not being able to find them to download them half the time
[24-Aug-2009 13:41:58] <rmatte> so then you have to go googling for RPM packages
[24-Aug-2009 13:42:01] <jb> so frontend it with yum
[24-Aug-2009 13:42:12] <jb> or your package management solution of choice..
[24-Aug-2009 13:42:27] <dranch> rmatte: so you recommend to install the OLD version of zenoss on the new machine, move the DB over, and then upgrade it?!
[24-Aug-2009 13:42:45] <rmatte> dranch: that would be the best solution
[24-Aug-2009 13:43:01] <dranch> interesting. assuming upgraing from 2.2.4 to current is supported
[24-Aug-2009 13:43:06] <rmatte> dranch: unless you want to upgrade it on the old machine and then move it over, either way works, I just figure this way would end up being less hassle
[24-Aug-2009 13:43:19] <rmatte> dranch: it may not be, you'll have to follow the upgrade path
[24-Aug-2009 13:43:43] <dranch> Ah... is the path clearly documented somewhere?
[24-Aug-2009 13:43:47] <rmatte> dranch: plus, if the upgrade fails on the new machine at least the old one is still running that way
[24-Aug-2009 13:43:56] <rmatte> dranch: yeh it is, hold on...
[24-Aug-2009 13:44:09] <dranch> exactly.. I can't mess with the current one (even if it's not running everything it should be)
[24-Aug-2009 13:44:11] <mrayzenoss> dranch: http://www.zenoss.com/community/docs/installation-guide/2.4.2/ch08s01.html#d4e866
[24-Aug-2009 13:44:39] <rmatte> yeh, so you need to go to 2.3.3 and then go to 2.4
[24-Aug-2009 13:44:40] <dranch> thanks mray
[24-Aug-2009 13:45:14] <dranch> ok. Mray.. have you seen this thread? anything to add on why COMMANDS don't run automatically?
[24-Aug-2009 13:45:41] <magnachef> so a sad (yet good) day...looks like I'm going to be taking a programming job within my company and won't be doing the NMS stuff anymore
[24-Aug-2009 13:46:12] <mrayzenoss> dranch: I've been passively listening, I don't have much to add
[24-Aug-2009 13:46:12] <dranch> Ha.. you'll probably be doing both jobs!
[24-Aug-2009 13:46:15] <rmatte> magnachef: ah, congrats on the new job
[24-Aug-2009 13:46:23] <mrayzenoss> magnachef: congrats, you'll be missed
[24-Aug-2009 13:46:25] <dranch> mray: bummer.. ok.
[24-Aug-2009 13:46:36] <magnachef> haha...we'll see how well I can document!
[24-Aug-2009 13:46:38] <magnachef> thanks
[24-Aug-2009 13:46:58] <mrayzenoss> dranch: my suggestion would be to create a new device class and add a simple version of the command you're running
[24-Aug-2009 13:46:59] <magnachef> I have a feeling that I'll still be a lurker
[24-Aug-2009 13:47:04] <dranch> It's just bizzare to me that some tests run.. others don't
[24-Aug-2009 13:47:19] <mrayzenoss> dranch: then add a device to the device class and see what works
[24-Aug-2009 13:47:20] <dranch> I can try that
[24-Aug-2009 13:47:50] <mrayzenoss> of course, if you're running 2.2.4 you're getting long in the tooth and 2.4.x should be better
[24-Aug-2009 13:47:59] <dranch> does it need to be new from the very top level?
[24-Aug-2009 13:48:21] <dranch> yeah.. 2.4.x is on the map.. all the windows boxes have been throwing fits with zenoss
[24-Aug-2009 13:48:34] <mrayzenoss> dranch: no, but it will inherit the templates of devices higher in the hierarchy (but not bind them)
[24-Aug-2009 13:49:03] <mrayzenoss> dranch: wait, maybe it does bind the templates if they exist at creation time
[24-Aug-2009 13:49:19] <dranch> mray: ok, I didn't know about the inherited but non-binded "feature"
[24-Aug-2009 13:49:30] <dranch> I'll just create a new toplevel
[24-Aug-2009 13:49:32] <mrayzenoss> dranch: basically, if you're having trouble with /Devices/Server/Blah, don't test a 'clean' version with /Devices/Server/Blah/New
[24-Aug-2009 13:50:28] <rmatte> mrayzenoss: in reports, you can select a period of time (start/end) to collect data for, how would I manually specify that in a d.getRRDValue command?
[24-Aug-2009 13:52:56] <dranch> So I added database5 to the new /Devices/Test class yet it put it back in /Devices /Server /Linux /database /replication
[24-Aug-2009 13:53:06] <dranch> wtf
[24-Aug-2009 13:53:17] <rmatte> "put it back in"?
[24-Aug-2009 13:53:27] <straterra> That's what she said?
[24-Aug-2009 13:53:30] <straterra> Sorry..
[24-Aug-2009 13:53:36] <dranch> it didn't show up it /Devices/Test
[24-Aug-2009 13:53:37] <rmatte>
[24-Aug-2009 13:53:39] <dranch> nice..
[24-Aug-2009 13:53:53] <rmatte> dranch: sounds screwy
[24-Aug-2009 13:53:58] <rmatte> did you copy it over?
[24-Aug-2009 13:54:29] <dranch> adding it as a new device
[24-Aug-2009 13:54:40] <mrayzenoss> dranch: doesn't it already exist?
[24-Aug-2009 13:54:50] <dranch> yup
[24-Aug-2009 13:54:57] <rmatte> then why add it as a new device?
[24-Aug-2009 13:55:10] <rmatte> you just need to move it from it's old organizer to the new one
[24-Aug-2009 13:55:15] <dranch> to see if I can get these COMMANDs to run under a clean template
[24-Aug-2009 13:55:44] <rmatte> dranch: just moving it to the new organizer will give it a clean template
[24-Aug-2009 13:56:07] <rmatte> the template is bound to the organizer, not the device (unless you specifically created a local copy of the template)
[24-Aug-2009 13:56:08] <dranch> Hmmm.. let me add a NEW device and run this COMMAND on that one
[24-Aug-2009 13:56:19] <dranch> ok.. that's good to know
[24-Aug-2009 13:56:54] <rmatte> You should read the Zenoss Admin guide top to bottom if you haven't already
[24-Aug-2009 13:57:01] <rmatte> you'd get a better feel for the system
[24-Aug-2009 13:57:44] <dranch> I've read parts of it but I agree... though the docs are a little fragmented if you ask me
[24-Aug-2009 13:58:07] <rmatte> I didn't find them to be very fragmented
[24-Aug-2009 13:58:43] <dranch> ok.. new machine added. let me now try to run the COMMAND script there
[24-Aug-2009 13:59:56] <mrayzenoss> dranch: I believe 2.3 was when the Admin Guide got it's first big revamp
[24-Aug-2009 14:00:12] <mrayzenoss> and each release has had more clean ups and updats
[24-Aug-2009 14:00:14] <dranch> oh.. ok. That might be a reason
[24-Aug-2009 14:00:41] <mrayzenoss> (had to go look at the 2.2 docs to verify that)
[24-Aug-2009 14:00:55] <mrayzenoss> and King Crab will continue to have more upgrades
[24-Aug-2009 14:01:45] <dranch> Ok.. the new template runs a manual test via the webui fine
[24-Aug-2009 14:02:11] <dranch> the result is OK|status=1 and it's set to /cmd/fail
[24-Aug-2009 14:02:51] <rmatte> ah, return dict(start=startDate, end=endDate, function=how), so it'd be like d.getRRDValue('08/17/2009', '08/24/2009', 'MAX')
[24-Aug-2009 14:03:01] * rmatte test
[24-Aug-2009 14:03:03] <rmatte> s
[24-Aug-2009 14:04:41] <dranch> mray: Do you agree with rmatte that commands shouldn't execute unless a datapoint is defined?
[24-Aug-2009 14:05:00] <dranch> or a threshhold?
[24-Aug-2009 14:07:17] <mrayzenoss> you don't have to have a threshold
[24-Aug-2009 14:07:42] <dranch> but you do need to set a datapoint?
[24-Aug-2009 14:12:23] <dranch> Tests aren't running on the new machine.. I just added a GAUGE datapoint to see if that changes anything
[24-Aug-2009 14:13:53] <jb> venturaville: do you guys have a 3584?
[24-Aug-2009 14:13:56] <jb> or rocket?
[24-Aug-2009 14:26:46] <rmatte> eugh, I wish rocket or venturaville were around, I'm trying to figure out how to convert a time such as 07/17/2009 to something that getRRDValue will understand as a start or end time
[24-Aug-2009 14:26:51] <rmatte> so far the best example I've found is:
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> import time
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> mystart = time.time() - (3600 * 8)
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> args = dict(start=mystart)
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> >>> f.getRRDValue('usedBlocks',**args)
[24-Aug-2009 14:26:53] <rmatte> [17-Jun-2009 12:43:54] <venturaville> 10021455.52
[24-Aug-2009 14:27:16] <rmatte> but, that doesn't explain how to convert a date such as that, it just kind of gives a general idea
[24-Aug-2009 14:29:25] <rmatte> hmmm, I might have figured out how, just need to code it up
[24-Aug-2009 14:36:57] <jb> hmm
[24-Aug-2009 14:37:07] <jb> i think my first official zenpack is finished
[24-Aug-2009 14:37:44] <mrayzenoss> ummm... yay? (My backlog is 16 ZenPacks deep)
[24-Aug-2009 14:37:54] <kgoedtel> pile it on!
[24-Aug-2009 14:38:28] <jb> hah matt
[24-Aug-2009 14:38:52] <mrayzenoss> it's OK, I did 10 over the weekend
[24-Aug-2009 14:39:06] <mrayzenoss> and should have them all published today
[24-Aug-2009 14:39:17] <jb> i just need to write the docs
[24-Aug-2009 14:39:39] <rmatte> hmmm, I can use the find() function in zendmd but I can't use it in a standalone script with the standard zendmd imports
[24-Aug-2009 14:39:48] <rmatte> lovely, now I need to figure out what I need to import to be able to use it
[24-Aug-2009 14:40:46] <mrayzenoss> #!/usr/bin/env python
[24-Aug-2009 14:40:46] <mrayzenoss> import Globals
[24-Aug-2009 14:40:46] <mrayzenoss> from Products.ZenUtils import Time
[24-Aug-2009 14:40:46] <mrayzenoss> from Products.ZenUtils.ZenScriptBase import ZenScriptBase
[24-Aug-2009 14:40:50] <mrayzenoss> like that?
[24-Aug-2009 14:41:03] <rmatte> #!/usr/bin/env python
[24-Aug-2009 14:41:03] <rmatte> import Globals
[24-Aug-2009 14:41:03] <rmatte> from Products.ZenUtils.ZenScriptBase import ZenScriptBase
[24-Aug-2009 14:41:03] <rmatte> from transaction import commit
[24-Aug-2009 14:41:03] <rmatte> dmd = ZenScriptBase(connect=True).dmd
[24-Aug-2009 14:41:04] <rmatte> from Products.ZenUtils import Time
[24-Aug-2009 14:41:06] <rmatte> is what I have
[24-Aug-2009 14:41:22] <jb> from string import strip, find, split, lower, atoi
[24-Aug-2009 14:41:43] <rmatte> thanks jb
[24-Aug-2009 14:42:04] <jb> np
[24-Aug-2009 14:42:18] <rmatte> weird
[24-Aug-2009 14:42:22] <rmatte> return s.find(*args)
[24-Aug-2009 14:42:23] <rmatte> TypeError: find/rfind/index/rindex() takes at least 1 argument (0 given)
[24-Aug-2009 14:42:32] <rmatte> d = find('XP03LYNv02')
[24-Aug-2009 14:42:35] <rmatte> that's the line
[24-Aug-2009 14:43:45] <rmatte> http://pastebin.com/m63e343c3
[24-Aug-2009 14:43:50] <rmatte> there's the whole script I'm testing
[24-Aug-2009 14:44:25] <jb> hmm wonder why my template bindings aren't being exported to a zenpack
[24-Aug-2009 14:44:26] <rmatte> I'm basically just trying to make a script where I can specify start, end, and method (average, maximum, etc...) and have getRRDValue retrieve info based on it
[24-Aug-2009 14:44:34] <rmatte> I'm basically trying to design scripts to dump reports
[24-Aug-2009 14:45:20] <rmatte> I want to at least get it working with 1 value for starters, and then replace the values with variables
[24-Aug-2009 14:46:29] <rmatte> ultimately it's going to be something like: script.py --start 08/17/2009 --end 08/24/2009 --function MAXIMUM
[24-Aug-2009 14:47:10] <rmatte> I'll probably build different functions in to it for different reports too, so that I can have something like --report cpu or --report memory
[24-Aug-2009 14:47:16] <rmatte> and have it dump the report as a csv
[24-Aug-2009 14:47:24] <rmatte> just doing this as a workaround until the reporting is fixed
[24-Aug-2009 14:48:12] <rmatte> it'll allow more flexibility in how it's output as well
[24-Aug-2009 14:50:27] <jb> cant help ya :/
[24-Aug-2009 14:50:47] <rmatte> weird how it's complaining that I'm not passing a value to find when I am
[24-Aug-2009 14:51:18] daMaestro|isBack is now known as daMaestro
[24-Aug-2009 15:24:58] <rmatte> It's not not, or isnot
[24-Aug-2009 15:25:05] <rmatte> woops
[24-Aug-2009 15:25:14] <dranch_> is snot
[24-Aug-2009 15:25:19] <dranch_> ;-)
[24-Aug-2009 18:17:18] daMaestro|isBack is now known as daMaestro
[25-Aug-2009 00:00:46] [disconnected at Tue Aug 25 00:00:46 2009]
[25-Aug-2009 00:00:46] [connected at Tue Aug 25 00:00:46 2009]
[25-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[25-Aug-2009 07:50:24] <mrayzenoss> Morning all
[25-Aug-2009 08:12:01] <ckrough> morning
[25-Aug-2009 08:36:26] <ckrough> Where does zenoss determine which processes it's going to monitor? I was under the impression it would only pick up on stuff defined in /Classes/Processes
[25-Aug-2009 08:36:38] <ckrough> But its picking up on some random stuff now
[25-Aug-2009 08:52:57] <cgibbons> bad regex?
[25-Aug-2009 09:13:33] <rmatte> good morning
[25-Aug-2009 10:03:34] <ckrough> cgibbons: sry, was at a meeting. No, I don't think it's bad regex since this is new behaviour (seemingly) but we havent modified Processes
[25-Aug-2009 10:04:04] <ckrough> cgibbons: Is my understanding correct in that it should only attempt to monitor processes that are defined under /Processes?
[25-Aug-2009 10:05:54] <cgibbons> as far as I understand (don't weigh that too heavily). I'd run zenmodeler run -v10 -d yourdevice real quick just to see what it shows when it discovers processes
[25-Aug-2009 10:06:09] <ckrough> Ill try that
[25-Aug-2009 11:36:38] <jb> mrayzenoss: just sent you a zenpack..
[25-Aug-2009 11:36:53] <mrayzenoss> just got it, thanks!
[25-Aug-2009 11:37:15] <jb>
[25-Aug-2009 11:37:16] <mrayzenoss> nice documentation
[25-Aug-2009 11:37:29] <jb> thanks
[25-Aug-2009 11:38:11] <rmatte> jb++ for docs
[25-Aug-2009 11:38:14] <rmatte>
[25-Aug-2009 11:38:20] <jb>
)[25-Aug-2009 11:38:21] <jb> http://thunder.jbdesign.net/docs/zenoss/zenoss-ibm3584mon.html
[25-Aug-2009 11:38:45] <rmatte> nice
[25-Aug-2009 12:07:38] <rmatte> eugh, our internet connection is being flaky right now
[25-Aug-2009 12:24:42] <dranch> Hey Rmatte?
[25-Aug-2009 12:24:48] <rmatte> yup?
[25-Aug-2009 12:24:51] <dranch> hey..
[25-Aug-2009 12:25:16] <dranch> so interetingly enough, that COMMAND stuff eventually started running under a new template. but it took HOURS
[25-Aug-2009 12:25:28] <rmatte> strange
[25-Aug-2009 12:25:42] <dranch> So obviously, something is hosed in the other template. Anyway, I question to you about binding
[25-Aug-2009 12:25:48] <dranch> A question that is
[25-Aug-2009 12:26:17] <dranch> I have a /devices /server /windows template. In that template is a MySQL test that I don't want run on one specific machine.
[25-Aug-2009 12:27:27] <dranch> What's strange is if I click on bind templates, only the top "devices" template is bound. the MySQL template is not highlighted BUT it is being run
[25-Aug-2009 12:27:52] <rmatte> yeh, so just go to that 1 specific device, go to More -> Templates, then go to Bind Templates
[25-Aug-2009 12:27:58] <rmatte> ctrl-click to unselect that template
[25-Aug-2009 12:28:00] <rmatte> then click ok
[25-Aug-2009 12:28:43] <dranch> That's what I'm saying, it's not highlight.. it's not bound
[25-Aug-2009 12:28:49] <rmatte> dranch: does the mysql template even show up in the list of templates?
[25-Aug-2009 12:28:54] <dranch> yes
[25-Aug-2009 12:29:09] <dranch> It's not highlighted yet it's still run
[25-Aug-2009 12:29:18] <rmatte> a quick and dirty way would be to make a local copy of the template, go in, and delete all of the datapoints, thresholds, and graphs
[25-Aug-2009 12:29:27] <rmatte> problem solved
[25-Aug-2009 12:29:32] <dranch> man.. I can't form sentences this morning
[25-Aug-2009 12:29:56] <rmatte> I'm like that some mornings
[25-Aug-2009 12:31:43] <dranch> Ok, with a local copy, the MySQL is still not bound. Let me try "push changes" to see if the behavior is any different
[25-Aug-2009 12:33:20] <rmatte> yeh, but if you create the local copy and remove everything from the template itself, then it won't be monitoring anything
[25-Aug-2009 12:33:29] <rmatte> you're effectively creating a blank template as a placeholder
[25-Aug-2009 12:33:48] <dranch> well, I need to monitor the server.. it's just the issue that HTTP isn't running on this specific one
[25-Aug-2009 12:34:24] <rmatte> eugh
[25-Aug-2009 12:34:37] <rmatte> you're trying to unbind the template, meaning you don't want to monitor anything in that template on that device
[25-Aug-2009 12:34:38] <rmatte> correct?
[25-Aug-2009 12:34:59] <rmatte> so I'm saying, create a local copy of the template, then edit the local template and remove everything in it
[25-Aug-2009 12:35:08] <dranch> but it's not bound (according to the GUI)
[25-Aug-2009 12:35:28] <dranch> Then re-create all the graphs, etc?
[25-Aug-2009 12:35:31] <rmatte> if you're seeing it in the perf tab, then it's somehow bound
[25-Aug-2009 12:35:40] <rmatte> you're using an old version, maybe there's a bug in the binding
[25-Aug-2009 12:35:53] <rmatte> why re-create?
[25-Aug-2009 12:36:14] <dranch> Yeah.. I found another but last night. For maintenance windows, the "restore former state" is busted
[25-Aug-2009 12:36:24] <rmatte> yeh
[25-Aug-2009 12:56:50] * pgega is away: I'm busy
[25-Aug-2009 13:00:03] * pgega is back (gone 00:03:15)
[25-Aug-2009 13:08:33] pgega is now known as pgega|away
[25-Aug-2009 13:08:33] <rmatte> hmmm, that Ex-Chat project (what npmccallum is using) seems pretty dead.
[25-Aug-2009 13:08:51] <mrayzenoss1> I think that's XChat2
[25-Aug-2009 13:09:06] <dranch> What does it do?
[25-Aug-2009 13:09:14] <mrayzenoss1> http://xchat.org/
[25-Aug-2009 13:09:18] <mrayzenoss1> just another IRC client
[25-Aug-2009 13:09:24] <rmatte> http://sourceforge.net/projects/ex-chat/
[25-Aug-2009 13:09:25] <mrayzenoss1> I used it on Linux and OSX
[25-Aug-2009 13:09:28] <rmatte> I thought it was that
[25-Aug-2009 13:09:32] <rmatte> it's an X-Chat clone
[25-Aug-2009 13:09:41] <dranch> wow.. how many does the world need?!
[25-Aug-2009 13:09:51] <mrayzenoss1> dunno, cgibbons wrote one
[25-Aug-2009 13:09:52] <rmatte> just two, iirc and bitchx
[25-Aug-2009 13:09:53] <rmatte> >:)
[25-Aug-2009 13:10:28] <mrayzenoss1> I'm using Adium for IRC now. 1 IM client for everything
[25-Aug-2009 13:10:44] <rmatte> yeh, I hear Adium is really good
[25-Aug-2009 13:10:59] <rmatte> I use bitchx for irc, and gaim (pidgin) for everything else
[25-Aug-2009 13:11:07] <mrayzenoss1> IRC, AIM, Yahoo, Gtalk, Facebook, Twitter
[25-Aug-2009 13:11:11] <mrayzenoss1> all in 1
[25-Aug-2009 13:12:09] <dranch> I'm using Adium now
[25-Aug-2009 13:12:26] <dranch> pidgin has an IRC client too
[25-Aug-2009 13:12:44] <mrayzenoss1> yeah, Pidgin and Adium use the same libraries (libpurple)
[25-Aug-2009 13:12:54] <rmatte> http://en.wikipedia.org/wiki/IrcII <--- Now there's an IRC client
[25-Aug-2009 13:12:55] <rmatte>
[25-Aug-2009 13:13:04] <dranch> yup.. and share the same vulerabilities too
[25-Aug-2009 13:14:26] <dranch> isn't bitchx based on ircII?
[25-Aug-2009 13:15:22] <rmatte> yeh, it was originally an ircII script
[25-Aug-2009 13:15:32] <rmatte> then it got converted to a standalone app by panasync
[25-Aug-2009 13:15:39] <rmatte> I'm using BX with Osiris script
[25-Aug-2009 13:15:44] <rmatte> osiri s-1c/ bitch x-1.1-final
[25-Aug-2009 13:16:38] <dranch> reading the wikipedia stuff on it.. didn't know it was merged into epic
[25-Aug-2009 13:17:20] <rmatte> yeh
[25-Aug-2009 13:17:39] mrayzenoss1 is now known as mrayzenoss
[25-Aug-2009 13:17:53] <jb> so
[25-Aug-2009 13:17:56] <jb> when i export to CSV
[25-Aug-2009 13:18:05] <jb> the times are always wrong
[25-Aug-2009 13:18:11] <jb> oh
[25-Aug-2009 13:18:14] <jb> its just excel.
[25-Aug-2009 13:18:22] <rmatte> jb
[25-Aug-2009 13:18:23] <rmatte> http://dev.zenoss.org/trac/ticket/5438
[25-Aug-2009 13:18:27] <rmatte> http://dev.zenoss.org/trac/ticket/5439
[25-Aug-2009 13:19:15] <rmatte> it's not just excel
[25-Aug-2009 13:19:19] <rmatte> the export function is fubar
[25-Aug-2009 13:20:40] <rmatte> I've just been copying and pasting right from the web page to get around it
[25-Aug-2009 13:20:56] <rmatte> but I'm working on a python script to temporarily replace some of the more important reports
[25-Aug-2009 13:20:59] <rmatte> until they can fix that
[25-Aug-2009 13:22:05] <rmatte> I probably won't have the script done until some time next week
[25-Aug-2009 13:22:08] <rmatte> too much on my plate right now
[25-Aug-2009 13:23:38] <rmatte> plus I'm developing my python skills as I go
[25-Aug-2009 13:23:42] <rmatte> so lots of googling/reading
[25-Aug-2009 13:23:59] <rmatte> once the script is done I'll release it
[25-Aug-2009 13:24:34] <jb> hmm
[25-Aug-2009 13:24:36] <jb> yeah the times are wrong
[25-Aug-2009 13:24:51] <rmatte> yeh, when you export it always exports the average over the last 7 days
[25-Aug-2009 13:24:55] <rmatte> it ignores what you selected
[25-Aug-2009 13:24:58] <jb> lame
[25-Aug-2009 13:24:59] <jb> :/
[25-Aug-2009 13:25:08] <jb> thats a big problem
[25-Aug-2009 13:25:23] <rmatte> these are pretty obvious bugs, which tells me that not many people bother to use the reports system apparently
[25-Aug-2009 13:25:36] <rmatte> though it's a very important component
[25-Aug-2009 13:25:37] <jb> i am just exporting events
[25-Aug-2009 13:25:40] <jb> not reports
[25-Aug-2009 13:25:46] <rmatte> oh
[25-Aug-2009 13:25:51] <rmatte> is it wrong for that as well?
[25-Aug-2009 13:26:05] * rmatte tests
[25-Aug-2009 13:26:08] <jb> oh
[25-Aug-2009 13:26:09] <jb> actually
[25-Aug-2009 13:26:13] <rmatte> I thought you were exporting reports
[25-Aug-2009 13:26:21] <jb> no
[25-Aug-2009 13:26:22] <jb> its correct
[25-Aug-2009 13:26:23] <jb> im stupid
[25-Aug-2009 13:26:37] <rmatte> events export fine
[25-Aug-2009 13:26:39] <rmatte>
[25-Aug-2009 13:32:06] <rmatte> well, I'm out, talk to you guys later
[25-Aug-2009 13:32:14] <straterra> later
[25-Aug-2009 14:35:27] <cgibbons> I wrote one of what, mrayzenoss?
[25-Aug-2009 14:39:38] <venturaville> hi
[25-Aug-2009 14:39:53] <venturaville> anyone interested in LSI storage arrays (e.g. IBM rebranded DS* models and such?)
[25-Aug-2009 14:49:38] <jb> yes
[25-Aug-2009 14:49:40] <jb> of course
[25-Aug-2009 14:49:58] <jb> we use TPC for the most part
[25-Aug-2009 14:53:30] <venturaville> jb: I mimicked what someone else did and I am using pywbem along with engenioprovider to be the CIMOM -> array link
[25-Aug-2009 14:53:40] <venturaville> seems to work
[25-Aug-2009 14:55:05] <venturaville> I was hoping someone else could get the same setup going so they could help me test
[25-Aug-2009 14:55:19] <jb> sure
[25-Aug-2009 14:55:30] <venturaville> media access statistics: (uRecoveredSeekOperations, None) (uTotalReadBlocksRequested, 2568448853L) (uTotalLargeWrites, 0L) (uTotalWriteBlocksRequested, 1097960875L) (uTotalBlocksRequested, 3666409728L) (uDescription, uMedia Access Statistics for Volume vol2) (uTotalLargeWriteBlocksRequested, 0L) (uUnrecoverableReadOperations, None) (uWriteOperations, 233603915L) (uTotalCacheReadChecks, 969917377L) (uRecoveredReadOperations, None) (uUnrecoverableSee
[25-Aug-2009 14:55:32] <jb> venturaville: btw do you guys use IBM 3584 tape libs?
[25-Aug-2009 14:56:01] <venturaville> jb: we avoid using tape libraries at all costs ... most of our backups and video tape libraries are handled in another group
[25-Aug-2009 14:56:09] <jb> oh ok
[25-Aug-2009 14:56:14] <jb> just finished a pack for 3584's
[25-Aug-2009 14:56:21] <venturaville> cool.
[25-Aug-2009 14:56:31] <jb> but yeah, i can help test
[25-Aug-2009 14:56:34] <venturaville> we looked at them a while back when doing video archival... ended up using storagetek instead
[25-Aug-2009 14:56:38] <jb> i have a lab DS4500
[25-Aug-2009 14:56:46] <jb> and several production DS4700's
[25-Aug-2009 14:56:52] <venturaville> I'm testing against a 4700
[25-Aug-2009 14:56:53] <jb> and an 8000 soon
[25-Aug-2009 14:56:58] <venturaville> shark?
[25-Aug-2009 14:57:03] <jb> yeah
[25-Aug-2009 14:57:10] <jb> unfortuantly
[25-Aug-2009 14:57:15] <jb> i hate these damn things
[25-Aug-2009 14:58:28] <venturaville> see if you can grab a copy of pywbem, and then try this: http://pastebin.com/d5a320607
[25-Aug-2009 14:58:48] <venturaville> insert your own username/password pair for the CIMOM server
[25-Aug-2009 14:58:52] <venturaville> as well as the cimom server
[25-Aug-2009 14:59:11] <venturaville> I'm probably going to have to do those as zproperties since there doesn't seem to be a way to do auto discovery
[25-Aug-2009 15:00:04] <jb> hmm
[25-Aug-2009 15:00:23] <jb> not even sure where the cimom server is
[25-Aug-2009 15:00:45] <jb> im not a storage guy.. i may need to get our storage dude to help
[25-Aug-2009 15:00:47] <jb> he'll be back tommorow
[25-Aug-2009 15:01:02] <venturaville> k... I'll check back tomorrow then
[25-Aug-2009 15:17:05] <mrayzenoss> Zenoss 2.4.3 is up on SourceForge
[25-Aug-2009 15:17:14] <chudler> niiiiice
[25-Aug-2009 15:17:26] <mrayzenoss> I'm about to get on a flight to Zenoss HQ, so the blog post will be delayed until later
[25-Aug-2009 15:17:40] <mrayzenoss> but the release notes are there and so are the bits
[25-Aug-2009 15:19:14] <mrayzenoss> BTW, iPhone tethering rocks
[25-Aug-2009 15:19:29] <chudler> yes. yes it does.
[25-Aug-2009 15:19:42] <chudler> If only I could make it work with linux
[25-Aug-2009 15:20:03] <mrayzenoss> heh, it barely works with OSX
[25-Aug-2009 15:20:45] <chudler> I have heard that bluetooth works fine everywhere, since it is a standard protocol. NO USB, though
[25-Aug-2009 15:21:00] <chudler> my netbook doesnt have USB :-( I got on the bandwagon too soon, I guess
[25-Aug-2009 15:21:02] <mrayzenoss> well, tethering isn't supported by AT&T
[25-Aug-2009 15:21:09] <chudler> I mean, it doesnt have bluetooth
[25-Aug-2009 15:21:13] <mrayzenoss> so to get it to work is a bit of hackery
[25-Aug-2009 15:21:17] <mrayzenoss> I'm actually using USB
[25-Aug-2009 15:21:17] <straterra> mine has bluetooth..wifi..usb..
[25-Aug-2009 15:25:06] <rocket> hey matt
[25-Aug-2009 15:25:10] <mrayzenoss> yo
[25-Aug-2009 15:25:22] <straterra> i use tethering with my verizon phone over wifi
[25-Aug-2009 15:26:39] <rocket> mrayzenoss: how goes the zenpack committing?
[25-Aug-2009 15:27:00] <mrayzenoss> I checked 9 into SVN
[25-Aug-2009 15:27:08] <mrayzenoss> I'm writing up webpages for them on the flight
[25-Aug-2009 15:27:14] <mrayzenoss> I've got all the screenshots uploaded
[25-Aug-2009 15:27:23] <mrayzenoss> so there should be a flurry of posts tonight
[25-Aug-2009 15:28:01] <chudler> amd the dramatic conclusion to the contest? ;-)
[25-Aug-2009 15:28:20] <mrayzenoss> ahh, the judges are still reviewing
[25-Aug-2009 15:28:29] <rocket> heh ..
[25-Aug-2009 15:28:35] <mrayzenoss> the list of 50 entries will be up tonight probably
[25-Aug-2009 15:28:47] <mrayzenoss> and I should have the judging done by Friday
[25-Aug-2009 15:28:59] <gwb235> if there's a community zenpack that you're using that's having problems, and you have support agreement with zenoss, does that cover helping debug why the community zenpack is having problems?
[25-Aug-2009 15:29:42] <mrayzenoss> gwb235: probably depends how bad it is
[25-Aug-2009 15:29:53] <mrayzenoss> I mean, officially they won't support it but they probably will
[25-Aug-2009 15:30:11] <mrayzenoss> Support debugs tons of customer issues with their own ZenPacks
[25-Aug-2009 15:30:15] <mrayzenoss> gwb235: which one?
[25-Aug-2009 15:30:58] <gwb235> chudler's SentryCDU bit. It works great about 60% of the time and I feel bad bothering him taking time away from his real job.
[25-Aug-2009 15:31:10] <mrayzenoss> heh, he's a customer too
[25-Aug-2009 15:31:21] <chudler> eheh, I should really fix that! ehe. I wouldnt pass the buck :-)
[25-Aug-2009 15:31:22] <mrayzenoss> gwb235: are you using his new .egg version?
[25-Aug-2009 15:31:28] <gwb235> yes
[25-Aug-2009 15:31:44] <gwb235> we've talked a bit, it's not a critical failure, just something I need to close the loop on
[25-Aug-2009 15:32:08] <chudler> Is it the thing where the page gives an error while viewing the "details"? That really is a zenoss fault, but everything else is me
[25-Aug-2009 15:32:40] <gwb235> it's the thing where i can snmpwalk the device and get the right output, but the aggregate load graphs are giving me negative numbers
[25-Aug-2009 15:33:16] <chudler> oh damn, yeah thats me with a bug somewhere, I suspect.
[25-Aug-2009 15:33:53] <gwb235> i tried using --showdevicereusults but am not experiences/skilled enough to follow the data in --> performance graph RRD values
[25-Aug-2009 15:36:07] <chudler> hmm. We should work together on this. I might have some free time in 1-2 weeks :-)
[25-Aug-2009 15:36:46] <gwb235> let me know, I'll be glad to help debug anytime. maybe a shared VNC session or something so you can see first hand
[25-Aug-2009 15:39:45] <gwb235> any help, much appreciated. I have some longhorn schwag if that helps sweeten the pot.
[25-Aug-2009 15:41:31] <chudler> I will let you know. I really want this to be fixed. Work priorities are not in line with it at the moment, but I am getting there.
[25-Aug-2009 15:41:40] <magnachef> what's the goal release date of King Crab?
[25-Aug-2009 15:42:23] <mrayzenoss> magnachef: this fall
[25-Aug-2009 15:42:33] <mrayzenoss> beta should start in the next few weeks
[25-Aug-2009 15:42:41] <cgibbons> they have me chained to my desk so i won't leave
[25-Aug-2009 15:42:52] <magnachef> haha...nice!
[25-Aug-2009 15:43:21] <magnachef> I'm writing up my documentation for our next NMS guy, so I'm leaving a note about this
[25-Aug-2009 15:43:48] <magnachef> do you guys have a link to enhancements, bugfixes, etc?
[25-Aug-2009 15:43:58] <chudler> what is this "documentation" of which you speak ;-)
[25-Aug-2009 15:44:23] <magnachef> documenting my current job, as I'm moving on to a programming job
[25-Aug-2009 15:44:24] <mrayzenoss> magnachef: for 2.4.3?
[25-Aug-2009 15:45:00] <magnachef> well, more for what King Crab will give us...I'm currently 2.3.3
[25-Aug-2009 15:45:18] <mrayzenoss> magnachef: http://www.zenoss.com/community/docs/
[25-Aug-2009 15:45:34] <mrayzenoss> ahh, the mysterious roadmap
[25-Aug-2009 15:45:34] <magnachef> ...basically leaving upgrades up to the new guy :-)
[25-Aug-2009 15:46:09] <magnachef> yeah, roadmap sounds like what I'm looking for...I'm sure it's ever changing...but at least having some big hitters
[25-Aug-2009 15:46:13] <sgsax> we've been using zenoss for a few months now for basic system monitoring and hve been pretty happy with it
[25-Aug-2009 15:46:50] <sgsax> but today, I've been trying to add a custom snmp monitor for some temp probes we just purchased and am getting stuck
[25-Aug-2009 15:47:00] <mrayzenoss> we don't have a public roadmap for King Crab, I'll hit up PM to publish something before we beta it
[25-Aug-2009 15:47:17] <sgsax> I don't *need* to install the MIBs for it, because I already know the full OID, but it'd be nice if I could
[25-Aug-2009 15:47:26] <sgsax> but I seem to be having path problems
[25-Aug-2009 15:47:47] <magnachef> mrayzenoss: thanks
[25-Aug-2009 15:48:27] <sgsax> anybody offer a hand?
[25-Aug-2009 15:48:31] <mrayzenoss> you can look up the closed tickets http://dev.zenoss.org/trac/report/5
[25-Aug-2009 15:48:31] <mrayzenoss> the other reports here show what's in the queue, what's been reviewed, etc. http://dev.zenoss.org/trac/report
[25-Aug-2009 15:48:31] <mrayzenoss> flight's boarding, gotta go
[25-Aug-2009 15:49:28] <jb> when is ent 2.4.3 dropping?
[25-Aug-2009 15:49:36] <mrayzenoss> sgsax: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch06s02.html double check the .0 on the oid
[25-Aug-2009 15:49:41] <mrayzenoss> 2.4.3 dropped
[25-Aug-2009 15:49:41] <mrayzenoss> and I'm out
[25-Aug-2009 15:49:59] <jb> not enterprise
[25-Aug-2009 15:50:29] <sgsax> I got the trailing .0
[25-Aug-2009 15:50:41] <jb> whats the problem?
[25-Aug-2009 15:50:44] <sgsax> I've read through that chapter, and think I have everything added correctly
[25-Aug-2009 15:50:45] <jb> are you getting OID is bad errors?
[25-Aug-2009 15:51:15] <sgsax> but I'm a bit confused about data points and data point aliases
[25-Aug-2009 15:51:23] <sgsax> and my graphs aren't coming out right
[25-Aug-2009 15:53:21] <jb> i really need to figure out a way to capture HBA I/O stats at the host level
[25-Aug-2009 15:55:47] <rocket> jb: for aix?
[25-Aug-2009 15:55:58] <jb> aix/windows/linux
[25-Aug-2009 15:56:11] <jb> we use IBM TPC for monitoring the san fabric and such
[25-Aug-2009 15:56:17] <jb> but i really want to be able to capture local I/O at the host
[25-Aug-2009 15:57:10] <rocket> ah I see
[25-Aug-2009 15:58:42] <jb> i can get windows i/o using the informant MIB
[25-Aug-2009 16:00:39] <chudler> I get tons of info from /proc/scsi/qla2xxx/*, but it is very specific to those cards
[25-Aug-2009 16:00:48] <jb> yeah
[25-Aug-2009 16:02:41] <jb> egor?
[25-Aug-2009 16:02:43] <jb> oops
[25-Aug-2009 16:04:28] <jb> for windows its all good and easy
[25-Aug-2009 16:04:40] <jb> sucks that it can't be common across all hosts though
[25-Aug-2009 16:16:06] <chudler> If its just I/O, that is abstracted by the linux disk subsystem, but it won't tell you the actual bytes the fiber is passing. I am mildly surprised that windows unifies that among the drivers
[25-Aug-2009 16:19:44] <jb> yeah its the same for windows
[25-Aug-2009 16:20:00] <jb> its not really coming from the fiber
[25-Aug-2009 17:19:13] <kgoedtel> haha
[25-Aug-2009 17:19:24] <kgoedtel> oops
[25-Aug-2009 18:10:33] <cgibbons> hurm
[25-Aug-2009 19:35:57] <ideopathic> do zenpacks need to be installed as the zenoss user?
[26-Aug-2009 00:00:46] [disconnected at Wed Aug 26 00:00:46 2009]
[26-Aug-2009 00:00:46] [connected at Wed Aug 26 00:00:46 2009]
[26-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[26-Aug-2009 06:37:47] edwardam1 is now known as edwardam
[26-Aug-2009 07:35:47] <ckrough> ah, coming in to gappy graphs. good morning to you too zenoss
[26-Aug-2009 08:28:15] <Apocalipse> hi people
[26-Aug-2009 08:28:28] <tripitaka> Hi, I have upgraded to 2.4.3 and I'm now getting these errors: http://www.pastebin.ca/1543144 - is it safe to downgrade by running an older stack installer?
[26-Aug-2009 08:29:28] <Apocalipse> maybe Master rmatte can help u
[26-Aug-2009 08:29:30] <Apocalipse>
[26-Aug-2009 08:38:51] <Apocalipse> does anybody knows how to change RRD graphs?
[26-Aug-2009 08:51:38] kevin7kal_ is now known as kevin7kal
[26-Aug-2009 09:28:04] <jb> venturaville: here?
[26-Aug-2009 09:30:47] <venturaville> jb: I'm here
[26-Aug-2009 09:31:50] <Apocalipse> hello people...
[26-Aug-2009 09:32:10] <Apocalipse> i would like to know if someone knows how to change RRD graphs
[26-Aug-2009 09:33:25] <Apocalipse> people?
[26-Aug-2009 09:33:43] <mrayzenoss> what do you want to change?
[26-Aug-2009 09:34:32] <Apocalipse> http://imagebin.org/61076
[26-Aug-2009 09:34:56] <Apocalipse> this graph is cpu percent usage...
[26-Aug-2009 09:35:06] <Apocalipse> i would like to change the units...
[26-Aug-2009 09:35:26] <Apocalipse> its showing 8.0 M , 6.0 M...etc
[26-Aug-2009 09:36:37] <mrayzenoss> back on the template, on the graph definition is where you put units
[26-Aug-2009 09:38:44] <Apocalipse> i went there...
[26-Aug-2009 09:38:47] <Apocalipse> put %
[26-Aug-2009 09:39:59] <mrayzenoss> I think you may have to remove the existing .RRD file to get the updated file format
[26-Aug-2009 09:40:38] <Apocalipse> hunnn
[26-Aug-2009 09:41:53] <Matir> Anyone ever seen Zenoss switch a process from unmonitored to monitored repeatedly? I keep getting alerts that process httpd is not running, even though I set it not to monitor that. (Additionally, httpd IS running)
[26-Aug-2009 09:42:05] <jb> ok, i cannot for the life of me figure out why this graph is showing all 0's.. i deleted the RRD, deleted the device, etc
[26-Aug-2009 09:42:13] <jb> verified that the OID's are good
[26-Aug-2009 09:43:38] <jb> venturaville: was it you that was talking about SVC?
[26-Aug-2009 09:43:43] <Apocalipse> matt...i got this in "graph point" format=%5.2lf%s
[26-Aug-2009 09:44:10] <venturaville> jb: I think multiple people have but yes I have been exploring using CIM for SVC and LSI Arrays.
[26-Aug-2009 09:44:32] <venturaville> unfrotunately CIM for SVC does not have performance data ... only structural and status data
[26-Aug-2009 09:45:03] <jb> have you looked at SNMP?
[26-Aug-2009 09:45:03] <venturaville> unless it is just somehow buried in there
[26-Aug-2009 09:45:12] <venturaville> jb: there isn't squat in SNMP
[26-Aug-2009 09:45:19] <jb> pfft
[26-Aug-2009 09:45:32] <jb> i was just looking at this "Inventory Email" that we get from it
[26-Aug-2009 09:45:35] <jb> and it has lots of data
[26-Aug-2009 09:46:11] <jb> mostly modeling data though
[26-Aug-2009 09:46:58] <venturaville> yep ... basically the same as what is in CIM
[26-Aug-2009 09:47:19] <venturaville> we have been wondering for a while if TPC actually gets its 'performance' data from the arrays to do its calculations for it
[26-Aug-2009 09:47:47] <venturaville> performance data is available in CIM for those
[26-Aug-2009 09:49:06] <jb> do you guys have TPC?
[26-Aug-2009 09:50:31] <venturaville> only parts of it
[26-Aug-2009 10:11:11] <rmatte> mrayzenoss: the upgrades to 2.4.3 that I did last night went very smoothly, and there is noticeable performance improvement in 2.4.3 compared to 2.4.1 and 2.3.3, which was nice to see.
[26-Aug-2009 10:14:33] <mrayzenoss> yay
[26-Aug-2009 10:32:20] <cgibbons> hmmm
[26-Aug-2009 10:49:36] <jb> when is 2.4.3 enterprise dropping?
[26-Aug-2009 10:52:27] <mrayzenoss> I think it's available through the support portal
[26-Aug-2009 10:56:08] <jb> oh
[26-Aug-2009 10:56:09] <jb> it is
[26-Aug-2009 10:56:12] <jb> it wasn't yesterday
[26-Aug-2009 11:02:54] <rmatte> yesterday they only released 2.4.3 around 6 or 7pm EST
[26-Aug-2009 11:25:19] <jb> hm
[26-Aug-2009 11:25:40] <jb> so, im trying to add Informant's I/O stats to the "Filesystem" box on the Software tab
[26-Aug-2009 11:25:44] <jb> instead of the hardware tab
[26-Aug-2009 11:25:54] <jb> the /Devices/Server/Filesystem template
[26-Aug-2009 11:26:05] <jb> not sure this is going to work though
[26-Aug-2009 11:30:42] <rmatte> you're just trying to do it by moving the template?
[26-Aug-2009 11:31:07] <rmatte> I think different types of templates have pretty concrete display paramaters
[26-Aug-2009 11:31:19] <jb> well, i took the IO datasources from the "HardDisk" template and created them on the "FileSystem" template
[26-Aug-2009 11:31:30] <rmatte> ah
[26-Aug-2009 11:31:35] <jb> im not sure the indexes are workign properly
[26-Aug-2009 11:31:38] <jb> or something
[26-Aug-2009 11:31:44] <rmatte> I think those datasources are collected by a collector plugin
[26-Aug-2009 11:34:40] <jb> which?
[26-Aug-2009 11:38:28] <rmatte> whichever has to do with filesystem
[26-Aug-2009 11:38:37] <rmatte> I don't know the full name off the top of my head
[26-Aug-2009 11:39:49] <jb> i think you need to somehow index the OID's
[26-Aug-2009 11:39:55] <jb> since 1.3.6.1.4.1.9600.1.1.1.1.18 contains stats for all disks
[26-Aug-2009 11:40:04] <jb> but the template doesn't know which is C:, D:, E:, etc
[26-Aug-2009 11:42:08] <jb> trying to figure out where that modeler is
[26-Aug-2009 11:42:28] <jb> cgibbons: you are the windows dude.. are you around by chance?
[26-Aug-2009 11:42:59] <rmatte> seems like moving it is more trouble than it's worth
[26-Aug-2009 11:43:04] <jb> well
[26-Aug-2009 11:43:07] <jb> the problem is..
[26-Aug-2009 11:43:21] <jb> egor's DellMon pack overrides the normal "Hard Disks" box in "Hardware"
[26-Aug-2009 11:43:39] <jb> so the I/O graphs no longer work
[26-Aug-2009 11:43:51] <jb> because its modeling that data directly from Dell's RAID controller
[26-Aug-2009 11:44:08] <jb> so i figured putting them in the Filesystems template would be good
[26-Aug-2009 11:44:58] <rmatte> ah
[26-Aug-2009 11:45:35] <rmatte> I just use the zenoss dell monitor zenpack instead
[26-Aug-2009 11:45:38] <rmatte> rather than egor's
[26-Aug-2009 11:45:50] <jb> egor's provides much more data
[26-Aug-2009 11:46:08] <rmatte> yeh, but I'm not generally interested in any of the other data it provides
[26-Aug-2009 11:46:22] <jb> a
[26-Aug-2009 11:46:24] <jb> h
[26-Aug-2009 12:56:47] <cgibbons> i am now, jb
[26-Aug-2009 13:01:54] <jb> sup cgibbons..
[26-Aug-2009 13:02:07] <cgibbons> you asked about winders earlier?
[26-Aug-2009 13:02:14] <jb> im trying to figure out how to move the I/O graphs from the InformantHardDisk to the Filesystems template
[26-Aug-2009 13:02:25] <jb> and im guessing there is some snmp indexing problems
[26-Aug-2009 13:10:47] <rmatte> woohoo, I just got a zenny zebra
[26-Aug-2009 13:11:01] <jb> hah
[26-Aug-2009 13:11:07] <jb> i have like 20 of those things floating around the office
[26-Aug-2009 13:11:12] <rmatte> lol
[26-Aug-2009 13:11:17] <rmatte> got a zebra, a pen, and a shirt
[26-Aug-2009 13:11:30] <rmatte> now I just need my jacket and I'll have the full ensemble
[26-Aug-2009 13:11:32] <jb> a NO NODE LEFT BEHIND shirt?
[26-Aug-2009 13:11:40] <gwb2351> if any of the zenoss guys in Austin want to come down to UT, I'll trade a longhorn for a zebra...
[26-Aug-2009 13:11:40] <rmatte> no, it says...
[26-Aug-2009 13:12:01] <rmatte> "Making enterprise management weird"
[26-Aug-2009 13:12:05] <jb> oh ih ave that one
[26-Aug-2009 13:12:28] <cgibbons> zebras.... http://yfrog.com/ccuvyj
[26-Aug-2009 13:12:28] <rmatte> It's a nice shirt
[26-Aug-2009 13:12:47] <cgibbons> i ride by there on my way home, gwb (live downtown)
[26-Aug-2009 13:12:57] <rmatte> cgibbons: haha, nice
[26-Aug-2009 13:13:09] <rmatte> I have 4 zebras here, but I can't keep them all
[26-Aug-2009 13:13:30] <rmatte> need to give some to other members of my team
[26-Aug-2009 13:13:50] <gwb2351> pick a day next week, bother rbriggs, and maybe we can do lunch + token schwag exchange
[26-Aug-2009 13:14:34] <jb> cgibbons: would you happen to have any idea on how to do that?
[26-Aug-2009 13:15:20] <cgibbons> well you would think it would just work
it just doesn't collect after the datapoints have been moved?[26-Aug-2009 13:15:28] <jb> nope
[26-Aug-2009 13:15:38] <jb> well.. how does it know how to index properly?
[26-Aug-2009 13:15:55] <jb> lDiskDiskWriteBytesPerSec = 1.3.6.1.4.1.9600.1.1.1.1.18
[26-Aug-2009 13:16:03] <jb> which contains all drives
[26-Aug-2009 13:16:03] <cgibbons> ah so it's instance based?
[26-Aug-2009 13:16:06] <jb> yeah
[26-Aug-2009 13:16:22] <cgibbons> the modeler plugins are what does the indexing usually so lesee
[26-Aug-2009 13:17:01] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.67.58 = STRING: "C:"
[26-Aug-2009 13:17:01] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.68.58 = STRING: "D:"
[26-Aug-2009 13:18:47] <cgibbons> okay, I think (danger) you'll need to make a new modeling plugin that collects from the right OIDs so that it can index the FileSystem components you're after. if you look @ HRFileSystemMap.py right now it does that, but for .1.3.6.1.2.1.25.2.3.1 in the host-resources-v2-mib
[26-Aug-2009 13:19:34] <jb> one sec
[26-Aug-2009 13:19:35] <cgibbons> you could, if you want to experiment, manually index using zendmd
[26-Aug-2009 13:19:38] <rmatte> well, I officially, finally, have all of my linux boxes with the same ZenPacks running at the same version
[26-Aug-2009 13:19:40] <rmatte> w00t
[26-Aug-2009 13:19:51] <cgibbons> as long as they are indexed, zenperfsnmp should collect the data (once told of the change)
[26-Aug-2009 13:23:08] <jb> hmmm
[26-Aug-2009 13:25:00] <cgibbons> of course the tricky thing is I think components can only have one snmpindex right now hrm
[26-Aug-2009 13:25:22] <jb> yeah '
[26-Aug-2009 13:25:33] <jb> basically im just trying to figure out how to display this I/O data
[26-Aug-2009 13:26:01] <cgibbons> does this need to be generic so when you model any new Windows device you'll see it in your template, or do you just need it for a couple of servers, etc?
[26-Aug-2009 13:26:02] <jb> because.. egor's DellMonPack override the default HardDisks on the "Hardware" tab with disks that are modeled directly from Dell's Raid controller
[26-Aug-2009 13:26:15] <cgibbons> yeah
[26-Aug-2009 13:26:22] <jb> so i need to display all logical windows drives and their corresponding I/O somewhere
[26-Aug-2009 13:26:32] <jb> every windows server
[26-Aug-2009 13:30:56] <cgibbons> talking to chet about it internally... he said no, can't index multiple objects on a single component. when he's done stuff like that before he's used a command plugin to collect the secondary data.
[26-Aug-2009 13:31:26] <jb> hm ok
[26-Aug-2009 13:31:27] <cgibbons> so that would be a combination of a new modeler to discover the indexes during modeling, and then a command plugin to monitor the data :
[26-Aug-2009 13:31:37] <cgibbons> not a lot of code but not just super easy either
[26-Aug-2009 13:31:39] <jb> yeah.. seems like too much trouble
[26-Aug-2009 13:32:03] <cgibbons> might be easier to change the host template to have your common 2 or 3 file systems in the main perf tab
[26-Aug-2009 13:32:20] <jb> any other suggestions on how to do it?.. I guess i could just create a new modeler and add a new box to the hardware tab
[26-Aug-2009 13:32:23] <jb> hm yeah
[26-Aug-2009 13:33:07] <cgibbons> if your servers are somewhat standardized you might catch 80% of what you need right off the bat
[26-Aug-2009 13:33:30] <jb> only thing is.. some servers have more logical drives than others, etc
[26-Aug-2009 13:33:34] <jb> c,d,e,f,g,h etc
[26-Aug-2009 13:33:44] <cgibbons> yeah
[26-Aug-2009 13:34:12] <cgibbons> that data should be available via WMI, no?
[26-Aug-2009 13:34:25] <jb> no
[26-Aug-2009 13:34:30] <jb> at least i don't think it is
[26-Aug-2009 13:34:32] <cgibbons> really? hrm lemme look
[26-Aug-2009 13:34:39] <jb> i think all thats available via WMI is combined IOPS?
[26-Aug-2009 13:34:46] <rmatte> you can get anything you can via snmp via wmi
[26-Aug-2009 13:34:54] <rmatte> you just need to figure out how to get it
[26-Aug-2009 13:35:00] <rmatte> it's not available by default in core
[26-Aug-2009 13:37:07] <cgibbons> and lDiskDiskWriteBytesPerSec is a logical disk counter not a physical disk counter, right?
[26-Aug-2009 13:39:02] <jb> yeah as far as I know
[26-Aug-2009 13:39:13] <cgibbons> have you looked @ Win32_PerfFormattedData_PerfDisk_LogicalDisk ?
[26-Aug-2009 13:39:41] <jb> you think this would be easier using WMI?
[26-Aug-2009 13:40:15] <cgibbons> 'select name,DiskWriteBytesPerSec from Win32_PerfFormattedData_PerfDisk_LogicalDisk' will do it, provided logical disk stats are enabled on the box
[26-Aug-2009 13:40:36] <cgibbons> dunno if it would be easier, you still have the whole component modeling/collection issue to deal with
[26-Aug-2009 13:41:20] <bigegor> how about wbem?
[26-Aug-2009 13:41:54] <cgibbons> wbem == wmi
[26-Aug-2009 13:42:22] <bigegor> wmi != wbem
[26-Aug-2009 13:43:34] <bigegor> wmi = CIM over DCOM but wbem = CIM over http[s]
[26-Aug-2009 13:44:03] <venturaville> http://www.serverwatch.com/tutorials/article.php/2178391/Understanding-WMI-WBEM-CIM-etc.htm
[26-Aug-2009 13:44:09] <cgibbons> it's still the same queries into the CIM namespace, getting the data out of it isn't what jb we're having issues with
[26-Aug-2009 13:45:02] <jb> bleh im thinking just writing a new modeler and displaying it separate of filesystems somewhere
[26-Aug-2009 13:45:17] <cgibbons> getting the data out of the windows box is easy, organizing it into the zenoss infrastructure in a natural way is what really the problem is here
[26-Aug-2009 13:45:25] <jb> exactly
[26-Aug-2009 13:46:18] <jb> theres just no easy/logical way to fit it in
[26-Aug-2009 13:47:00] <cgibbons> is there any relationship between the index in that other mib back to host-resources-v2?
[26-Aug-2009 13:47:12] <jb> i dont believe so
[26-Aug-2009 13:47:16] <jb> it has some funky indexing
[26-Aug-2009 13:47:19] <bigegor> no
[26-Aug-2009 13:47:37] <cgibbons> bummer
[26-Aug-2009 13:47:46] <jb> the index i like "2.67.58"
[26-Aug-2009 13:47:55] <jb> is
[26-Aug-2009 13:48:17] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.67.58 = STRING: "C:"
[26-Aug-2009 13:48:17] <jb> SNMPv2-SMI::enterprises.9600.1.1.1.1.1.2.68.58 = STRING: "D:"
[26-Aug-2009 13:48:19] <jb> which is that
[26-Aug-2009 13:52:54] <cgibbons> yeah I dunno for you if SNMP or WMI based would be easier, but a plugin and a command data-type in your template would dit
[26-Aug-2009 13:52:57] <cgibbons> er do it
[26-Aug-2009 13:53:09] <jb> actually
[26-Aug-2009 13:53:14] <jb> i can get it from PhysicalDisk
[26-Aug-2009 13:53:40] <jb> at \PhysicalDisk\Disk Read Bytes/sec
[26-Aug-2009 13:54:23] <cgibbons> yeah and physical disk counters are always enabled by default iirc
[26-Aug-2009 13:54:23] <jb> i wouldnt know how to do it in WMI though, I don't think..
[26-Aug-2009 13:54:27] <jb> right
[26-Aug-2009 13:55:43] <jb> i would need to loop thru all instances and display them
[26-Aug-2009 13:58:53] <venturaville> wmi is a pain to loop through
[26-Aug-2009 13:59:01] <venturaville> I'm working with it right now for the lsi array stuff
[26-Aug-2009 13:59:14] <jb> yeah.. no idea where to even begin with that
[26-Aug-2009 13:59:43] <venturaville> try yawn
[26-Aug-2009 13:59:52] <venturaville> (though I don't know if it can talk directly to windows WMI)
[26-Aug-2009 14:01:31] <cgibbons> this at least provides data: wmic -U '.\me' //somewhere.local 'select name,DiskWriteBytesPerSec from Win32_PerfRawData_PerfDisk_PhysicalDisk'
[26-Aug-2009 14:02:02] <cgibbons> but then you've got that standard funky stuff with partitions vs. logical volumes
[26-Aug-2009 14:02:44] <jb> yeah.. ok since someone at zen already did it using snmp
[26-Aug-2009 14:03:02] <jb> can't i just take that modeler and stick it in a new "box" on the hardware tab?
[26-Aug-2009 14:03:17] <jb> the standard windows "Hard Disks" box is all I need..
[26-Aug-2009 14:06:08] <jb> egor.. that sound possible to you?
[26-Aug-2009 14:08:30] <rocket> jb ... impossible ... jk ...
[26-Aug-2009 14:08:35] <bigegor> i would like try to implement CIM based device model in zenoss
[26-Aug-2009 14:08:39] <jb>
[26-Aug-2009 14:08:52] <jb> well, i mean.. zenoss is already grabbing and indexing this data properly
[26-Aug-2009 14:08:58] <jb> i just need to make it display somewhere!
[26-Aug-2009 14:09:13] <rocket> jb get me up to speed
[26-Aug-2009 14:09:29] <rocket> you have data obtained and stored in the model?
[26-Aug-2009 14:09:32] <jb> ok
[26-Aug-2009 14:09:51] <cgibbons> there's already support in core for modeling via WMI, there's just no plugins provided in core that do it, nor are there any collectors in core that use it, but the framework is there
[26-Aug-2009 14:10:06] <jb> by default.. zenoss uses the windows informant MIB to display "Hard Disks" on the hardware tab.. this template (/Devices/Server/Windows/HardDisks) contains I/O graphs obtained from the informant MIB
[26-Aug-2009 14:10:42] <jb> but, when egor's DellMon Pack is used.. it overrides all things in the hardware tab and replaces hard disks with ones that are modeled from Dell's management agent
[26-Aug-2009 14:11:15] <jb> which a.) only displays local drives attached to dell's raid controller and b.) dell doesnt provide i/o counters
[26-Aug-2009 14:11:36] <jb> so, im basically trying to take this HardDisks data (IO) and display it somewhere else
[26-Aug-2009 14:11:47] <rocket> jb: so basically you want to take the dellmon pack and remove the hard disk stuff?
[26-Aug-2009 14:11:50] <jb> cgibbons: im enterprise..
[26-Aug-2009 14:12:02] <cgibbons> oh, detail...
[26-Aug-2009 14:12:05] <cgibbons> easy
[26-Aug-2009 14:12:10] <jb> rocket: well, no, because that data is useful too.
[26-Aug-2009 14:12:25] <cgibbons> you can monitor the PerfMon path directly in enterprise
[26-Aug-2009 14:12:34] <jb> but, i also need to display the I/O data for every disk
[26-Aug-2009 14:12:36] <cgibbons> and the component is probably already being modeled
[26-Aug-2009 14:13:47] <jb> hmm!
[26-Aug-2009 14:14:00] <cgibbons> you good with zendmd?
[26-Aug-2009 14:14:08] <jb> not really :/
[26-Aug-2009 14:15:09] <rocket> jb: ok I think I know your solution
[26-Aug-2009 14:15:21] <bigegor> jb: what happend if you remove DellHardDiskMap from zCollectorPlugins list and remodel your server?
[26-Aug-2009 14:15:29] <jb> sec
[26-Aug-2009 14:16:27] <rocket> bigegor: your zenpack overwrites the hard disk template correct?
[26-Aug-2009 14:17:27] <cgibbons> jb: try this, too, inside zendmd: find("yourdevicehere").os.filesystems.C.perfmonInstance
[26-Aug-2009 14:18:03] <bigegor> no, it add new DellHardDisk template
[26-Aug-2009 14:18:19] <jb> cgibbons: AttributeError: 'NoneType' object has no attribute 'os'
[26-Aug-2009 14:18:35] <cgibbons> you choose a valid device name inside the find?
[26-Aug-2009 14:18:58] <rocket> bigegor: it leaves the old harddisk template?
[26-Aug-2009 14:18:59] <jb> oh oops.. typo.
[26-Aug-2009 14:19:08] <rmatte> maybe he needs to do d = find("bleh") then d.os.filesystems.C.perfmonInstance?
[26-Aug-2009 14:19:10] <rocket> bigegor: how are you overridding the template?
[26-Aug-2009 14:19:20] <rmatte> type
[26-Aug-2009 14:19:23] <rmatte> typo rather
[26-Aug-2009 14:19:34] <jb> cgibbon: yah that didn't work.. same error
[26-Aug-2009 14:19:44] <bigegor> it doesn't touch HardDisk template
[26-Aug-2009 14:20:12] <jb> bigegor: so what makes it dissapear?
[26-Aug-2009 14:20:26] <cgibbons> k, that would be there depending upon which plugin(s) you have bound. but in any event, if you've got enterprise, we can get your components modeled with a perfmon instance, and then you can add any perfmon datasource to the template and magic happens, independent of the snmp stuff
[26-Aug-2009 14:21:21] <rocket> jb: oh yeah I bet he is overriding the page template
[26-Aug-2009 14:21:23] <jb> uh oh
[26-Aug-2009 14:21:28] <jb> now im erroring out everywhere.
[26-Aug-2009 14:21:49] <jb> yeah thats what hes doing
[26-Aug-2009 14:21:59] <rocket> jb: eg for what you are looking for you need to hack the page template to have both displayed
[26-Aug-2009 14:22:18] <jb> exacto.
[26-Aug-2009 14:22:34] <jb> ive somehow managed to botch this production instance
[26-Aug-2009 14:22:38] <jb> without doing anything
[26-Aug-2009 14:23:35] <bigegor> component with meta_type = DellHardDisk bind DellHardDisk tempate automatically
[26-Aug-2009 14:24:50] <rocket> ok I got it ...
[26-Aug-2009 14:25:03] <jb> well, now zen is completely broken
[26-Aug-2009 14:25:14] <cgibbons> yikes
[26-Aug-2009 14:25:21] <jb> and i have no idea how that happened!
[26-Aug-2009 14:25:23] <jb> sec
[26-Aug-2009 14:28:12] <jb> zope just dies
[26-Aug-2009 14:28:40] <cgibbons> uh
[26-Aug-2009 14:28:51] <cgibbons> let's blame rocket, he's more idle
[26-Aug-2009 14:29:03] <jb> well now its running
[26-Aug-2009 14:29:07] <jb> but not responding to anything via the web
[26-Aug-2009 14:30:07] <jb> hrm, nothing in logs
[26-Aug-2009 14:31:43] <jb> yeah, zope just keeps dying :/
[26-Aug-2009 14:40:11] <rocket> cgibbons: !!!!
[26-Aug-2009 14:40:39] <rocket> blame me not! It wasnt my zenpack that removed the old info .. ;p
[26-Aug-2009 14:40:56] <jb> hah
[26-Aug-2009 14:40:59] <cgibbons>
[26-Aug-2009 14:41:21] <rmatte> yikes, you fried your zope?
[26-Aug-2009 14:41:27] <jb> looks like it.
[26-Aug-2009 14:41:31] <rocket> bigegor: if I could make a recommendation .. I would update your page template
[26-Aug-2009 14:41:36] <jb> i have no idea how..
[26-Aug-2009 14:41:53] <rocket> bigegor: if the original modellor collects the hdisk data display the info if found
[26-Aug-2009 14:42:01] <rmatte> jb: what errors are you seeing in the zopectl log file?
[26-Aug-2009 14:42:06] <rocket> bigegor: else do not display it ..
[26-Aug-2009 14:42:14] <jb> http://pastebin.com/m2d80162d
[26-Aug-2009 14:42:35] <rocket> bigegor: my aix zenpack has a couple example areas that do that in the page templates around the printer and print queues
[26-Aug-2009 14:42:49] <rmatte> unknown status byte value, hmmm
[26-Aug-2009 14:44:12] <rmatte> For starters, here's the code that generates that error: http://pastebin.com/m1ba60cb3
[26-Aug-2009 14:44:36] <bigegor> rocket: what page template do you mean?
[26-Aug-2009 14:44:41] <cgibbons> okay, ian says that's just a bad cache. move that zeo1-1.zec file outside of $ZENHOME/var (shutdown zeo first and then restart it)
[26-Aug-2009 14:45:08] <rocket> bigegor: advanced details page template you wrote
[26-Aug-2009 14:45:39] <rocket> bigegor: you only reference the hp disk data and it is automatically pulling in the hp modelled information
[26-Aug-2009 14:46:30] <rocket> the web interface needs to be updated to display the other disk information if collected from the original modeller
[26-Aug-2009 14:46:51] <rocket> or your new modeller should be updated to have the data jb is looking for
[26-Aug-2009 14:47:00] <jb> cgibbons: woot.. and we're back.
[26-Aug-2009 14:47:01] <jb>
[26-Aug-2009 14:47:04] <jb> thanks
[26-Aug-2009 14:47:06] <rmatte> hehe
[26-Aug-2009 14:47:14] <jb> rmatte: gj tracking down that code too..
[26-Aug-2009 14:47:17] <bigegor> ok. i will try
[26-Aug-2009 14:47:19] <jb> it was cache related just as that snippet implied
[26-Aug-2009 14:49:02] <rmatte> yeh, I forgot about the .zec file though, I remember having to do that once over half a year ago
[26-Aug-2009 14:49:13] <jb> ok, so bascially we need to alter the advanced details hardware template to tell it to display the original info..
[26-Aug-2009 14:49:20] <rocket> jb: yes
[26-Aug-2009 14:49:31] <rocket> jb: thats exactly what is missing
[26-Aug-2009 14:49:55] <rocket> jb: the original page template needs to be merged with the new advanced details page template
[26-Aug-2009 14:50:19] <jb> right
[26-Aug-2009 14:50:26] <jb> just parts of it, at least
[26-Aug-2009 14:50:46] <rocket> yes the disk parts of it
[26-Aug-2009 14:53:57] <jb> and this would be deviceHardwareDetail.pt?
[26-Aug-2009 14:55:51] <rocket> yes
[26-Aug-2009 14:55:56] <rocket> I believe so
[26-Aug-2009 15:04:39] <bigegor> i have done some tests. its work. Without DellHardDisk and with UCDHardDiskMap (Linux HD performace). it shows all my disks
[26-Aug-2009 15:04:56] <bigegor> after modeling
[26-Aug-2009 15:12:35] <jb> hm
[26-Aug-2009 15:12:53] <jb> so it shows you the dell hard disk model and the normal one?
[26-Aug-2009 15:14:40] <bigegor> no dell disks but disks from OS
[26-Aug-2009 15:14:52] <jb> ah.. any way to display them both?
[26-Aug-2009 15:19:21] <rocket> jb: he might have to extend the device model to show both
[26-Aug-2009 15:21:32] <bigegor> thanks rocket. That's right. End another problem UCD and Informant is OS specific
[26-Aug-2009 15:21:42] <jb> yeah
[26-Aug-2009 15:21:45] <jb> its annoying
[26-Aug-2009 15:22:52] <jb> yeah informant stuff should only be bound to /Devices/Server/Windows
[26-Aug-2009 15:23:33] <jb> but since UCD (Linux) doesn't provide the hard disk information, it shouldn't be displayed at all on linux machines (only the Dell info).. right?
[26-Aug-2009 15:24:56] <bigegor> UCD provide hard disk information
[26-Aug-2009 15:25:14] <jb> hm...
[26-Aug-2009 15:25:50] <bigegor> UCD is part of net-snmp
[26-Aug-2009 15:25:54] <jb> right
[26-Aug-2009 15:26:38] <jb> so is it going to be possible to get both your Dell HD info
[26-Aug-2009 15:26:45] <jb> and the informant? (for IO)
[26-Aug-2009 15:27:24] <rocket> jb it might mess up reports but I can see that it would be a possible
[26-Aug-2009 15:27:52] <jb> hmm.. or maybe rename one?
[26-Aug-2009 15:27:57] <jb> so they both aren't named "Hard Disks"
[26-Aug-2009 15:29:33] <bigegor> no. informant used 'harddisks' relation and DellHardDisk used it to. by modeling last plugin owerwrite data collected by other one.
[26-Aug-2009 15:30:07] <jb> hrm
[26-Aug-2009 15:30:07] <jb> k
[26-Aug-2009 15:31:21] <bigegor> i have the same problem by HP modeling. Have a look on HPMon zenpack.
[26-Aug-2009 15:31:48] <rocket> bigegor: look at my aix zenpack it shows how to make new device components
[26-Aug-2009 15:31:52] <jb> could we rename DellHardDisks relation to "dellharddisk" or something?
[26-Aug-2009 15:32:10] <rocket> bigegor: eg the printqueue is a new device component and would be like jb's DellHardDisks
[26-Aug-2009 15:32:11] <jb> cause i think status of the dell harddisks is important too
[26-Aug-2009 15:32:56] <bigegor> rocket: i know this way to. You use zPythonClass
[26-Aug-2009 15:33:43] <rocket> bigegor: yes that is the case
[26-Aug-2009 15:36:22] <bigegor> rocket: using of zPythonClass had some limitation
[26-Aug-2009 15:38:47] <rocket> there is no other way to create a new device component that I know of then...
[26-Aug-2009 15:43:09] <bigegor> rocket: have a look on deviceAdvDetail and RDBMS zenpacks __init__.py
[26-Aug-2009 15:46:14] <rmatte> what's the function to pull availability for a device?
[26-Aug-2009 15:48:41] <rmatte> anyone?
[26-Aug-2009 15:52:17] <rmatte> I need to duplicate what the availability report is doing
[26-Aug-2009 16:04:08] <rmatte> ummm, wow, the memory utilization report must have changed between 2.4.1 and 2.4.3, I copied my custom report over and it doesn't work at all
[26-Aug-2009 16:04:42] <rmatte> actually crap, I didn't copy the report file itself over
[26-Aug-2009 16:12:34] <bigegor> jb: can you try this command? wbemcli ein -noverify 'https://username:password@your_dell_server:5989/root/cimv2:CIM_DiskDrive'
[26-Aug-2009 16:14:41] <jb> one sec
[26-Aug-2009 16:15:29] <jb> would this be my domain username/password?
[26-Aug-2009 16:15:57] <bigegor> yes
[26-Aug-2009 16:16:21] <jb> i have nothing on 5989
[26-Aug-2009 16:17:01] <bigegor> no wbem
[26-Aug-2009 16:17:14] <jb> yeah
[26-Aug-2009 16:17:18] <bigegor> but from IBM?
[26-Aug-2009 16:17:23] <jb> windows doesn't have a WBEM provider by default
[26-Aug-2009 16:17:32] <jb> i didnt think..
[26-Aug-2009 16:18:17] <jb> nothing on my IBM boxes either
[26-Aug-2009 16:18:18] <bigegor> but management agents rom HP and IBM
[26-Aug-2009 16:19:12] <jb> ah
[26-Aug-2009 16:19:13] <bigegor> have you IBM Director installed?
[26-Aug-2009 16:19:22] <jb> probably not on these
[26-Aug-2009 16:19:31] <jb> i uninstalled director on my dev boxes for the time being
[26-Aug-2009 16:42:53] <rmatte> how would I go about printing an interface name?
[26-Aug-2009 16:43:45] <rmatte> if I print "i" in d.os.interfaces
[26-Aug-2009 16:44:24] <rmatte> it prints out like: <IpInterface at Vlan50>
[26-Aug-2009 16:44:32] <rmatte> I need it to output normally, just the name of the interface
[26-Aug-2009 16:44:44] <rmatte> oh wait
[26-Aug-2009 16:45:07] <rmatte> got it
[26-Aug-2009 16:45:10] <rmatte> just came to me, i.id
[26-Aug-2009 16:51:17] <mrayzenoss> rmatte: http://blog.zenoss.com/2009/08/24/tip-of-the-month-zendmd-and-device-interfaces-in-zenoss/
[26-Aug-2009 17:02:41] <rmatte> mrayzenoss: nice
[26-Aug-2009 17:02:46] <rmatte> I'm outta here, talk to you tomorrow
[26-Aug-2009 17:05:24] ideopathic_ is now known as ideopathic
[26-Aug-2009 18:01:02] kisielk_ is now known as kisielk
[26-Aug-2009 18:05:38] <cgibbons> hurm
[27-Aug-2009 00:00:46] [disconnected at Thu Aug 27 00:00:46 2009]
[27-Aug-2009 00:00:46] [connected at Thu Aug 27 00:00:46 2009]
[27-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[27-Aug-2009 00:53:15] <ironpaw> hi
[27-Aug-2009 00:53:20] <ironpaw> anyone around?
[27-Aug-2009 00:55:34] <Appiah> 48 in the channel
[27-Aug-2009 00:55:37] <Appiah> others*
[27-Aug-2009 02:00:39] <ironpaw> oh
[27-Aug-2009 02:00:53] <ironpaw> ive just been here an there have been a heap of other ppl here but none respond
[27-Aug-2009 02:00:58] <ironpaw> much like i didnt just then
[27-Aug-2009 02:03:02] <ironpaw> i appear to have issues adding data sources to devices in 2.4.3
[27-Aug-2009 02:03:07] <ironpaw> is this a known issue?
[27-Aug-2009 06:55:04] <tripitaka> Hi, I'm seeing a lot of false positives, particularly for http service. The load on the zenoss server is not very high, and tcpdump shows what look like good http connections from the monitoring service. What are the likely causes of the false positives?
[27-Aug-2009 08:37:03] sgsax is now known as newt
[27-Aug-2009 08:37:42] newt is now known as sgsax
[27-Aug-2009 08:40:59] <sergeymasushko> hi 2all. Can anyone explain, what does '2009-08-27 02:50:22 WARNING zen.ZenPing: Configuration still loading. Started at Thu Aug 20 11:04:37 2009' mean in the 'zenping.log' ?
[27-Aug-2009 08:42:36] <sergeymasushko> I've experienced several fake UP/DOWN alerts from zenoss and I tried to find out what was the problem...
[27-Aug-2009 08:47:53] <rmatte> eugh, after our upgrade to 2.4.3 a lot of the CPU performance graphs are not populating data at all
[27-Aug-2009 08:47:59] <rmatte> and some are completely missing RRD files
[27-Aug-2009 08:48:59] <jb> morning
[27-Aug-2009 08:49:06] <rmatte> good morning
[27-Aug-2009 08:49:12] <rmatte> well, bad morning for me
[27-Aug-2009 08:49:12] <rmatte> lol
[27-Aug-2009 08:49:41] <sergeymasushko> morning guys
[27-Aug-2009 08:53:23] <jb> hehe
[27-Aug-2009 08:53:34] <jb> yeah, ive got to get these graphs fixed too
[27-Aug-2009 08:54:06] <rmatte> yeah, but what's causing them to go like that in the first place
[27-Aug-2009 08:54:17] <jb> i think my problem is a bit different
[27-Aug-2009 08:54:26] <jb> or are your saying "0" too?
[27-Aug-2009 08:55:24] <rmatte> a bunch are saying 0, some are saying nan, and some have no rrd files at all
[27-Aug-2009 08:55:32] <rmatte> but most are saying 0
[27-Aug-2009 08:55:57] <rmatte> it's horrible because I need to generate reports for my boss and now I'm missing a TON of report data
[27-Aug-2009 08:56:19] <rmatte> we may have to revert back to the older versions if I can't fix this
[27-Aug-2009 08:56:59] <jb> im going to delete my graph definition all together
[27-Aug-2009 08:57:01] <jb> and try again
[27-Aug-2009 08:57:05] <jb> although thats not really an option for you
[27-Aug-2009 08:58:52] <rmatte> nope
[27-Aug-2009 08:59:07] <rmatte> these templates have been working perfectly for months
[27-Aug-2009 08:59:14] <rmatte> I doubt it has anything to do with the actual template
[27-Aug-2009 09:00:36] <jb> hrm yeah didnt work for me
[27-Aug-2009 09:01:00] <rmatte> I just don't see why some graphs are working fine and others aren't
[27-Aug-2009 09:01:04] <jb> snmpwalk -cfcsnmp1ro -v1 172.26.101.66 1.3.6.1.4.1.8072.1.3.2.3.1.1.11.112.111.115.116.102.105.120.83.101.110.116
[27-Aug-2009 09:01:05] <jb> against fc-iftpcln01.corp.follett.com
[27-Aug-2009 09:01:05] <jb> NET-SNMP-EXTEND-MIB::nsExtendOutput1Line."postfixSent" = STRING: 228131
[27-Aug-2009 09:01:19] <rmatte> mine is a command based datapoint
[27-Aug-2009 09:02:53] <rmatte> well, my boss is going to kill me unless I can fix this
[27-Aug-2009 09:03:13] <rmatte> too bad chet's not around
[27-Aug-2009 09:13:13] <rmatte> well, I'm going to try deleting RRD files on one device for starters to see if that'll fix it
[27-Aug-2009 09:17:10] <rmatte> weird, I just noticed something
[27-Aug-2009 09:17:33] <jb> renaming by graphs/datapoints seemed to do it
[27-Aug-2009 09:17:59] <rmatte> omg
[27-Aug-2009 09:18:12] <rmatte> I included an old script in my snmp performance ZenPack, that's why it's all screwed
[27-Aug-2009 09:18:23] <rmatte> damnit
[27-Aug-2009 09:20:02] <rmatte> When I run the command against systems with 1 CPU it outputs like CPU1=0 Total=0
[27-Aug-2009 09:20:10] <rmatte> but it should be OK|CPU1=0 Total=0
[27-Aug-2009 09:20:15] <rmatte> I need to fix that
[27-Aug-2009 09:21:54] <rmatte> I'll need to release an updated version to Matt too, since it's posted on the ZenPack page
[27-Aug-2009 09:22:00] <rmatte> gah
[27-Aug-2009 09:25:43] <rmatte> hey Matt
[27-Aug-2009 09:26:03] <rmatte> my new Zenpack has an old version of the cpu collection script (I hadn't noticed)
[27-Aug-2009 09:26:05] <mrayzenoss> bad news
[27-Aug-2009 09:26:07] <mrayzenoss> http://dev.zenoss.org/trac/ticket/5466
[27-Aug-2009 09:26:21] <mrayzenoss> we're pushing 2.4.4 out
[27-Aug-2009 09:26:50] <jb> eek
[27-Aug-2009 09:26:56] <rmatte> pushing it out today?
[27-Aug-2009 09:27:25] <rmatte> anyways, I'm going to update my windows performance zenpack and send the updated version to you
[27-Aug-2009 09:27:28] <mrayzenoss> we're spinning up builds right now
[27-Aug-2009 09:27:36] <mrayzenoss> thanks
[27-Aug-2009 09:27:42] <rmatte> the old version of the script causes no cpu collection for anything under 1 Processor
[27-Aug-2009 09:27:44] <rmatte> lol
[27-Aug-2009 09:29:12] <jb> how can I reset values for all datasources?
[27-Aug-2009 09:29:16] <jb> removing the rrd did nothing
[27-Aug-2009 09:29:42] <rmatte> ok, I'm trying to understand this trac...
[27-Aug-2009 09:30:43] <rmatte> does that only affect zenoss email alerts?
[27-Aug-2009 09:30:49] <ckrough> ouch
[27-Aug-2009 09:31:11] <rmatte> "No CLEARs are sent out to agents, so the think devices that are up are down etc."
[27-Aug-2009 09:31:28] <rmatte> no clears are sent to collectors is what he means?
[27-Aug-2009 09:33:28] <rmatte> from looking at the event history on one of my boxes, clear events are working fine in the event console
[27-Aug-2009 09:33:34] <rmatte> so does that bug only affect notification?
[27-Aug-2009 09:34:08] <ckrough> clears dont get sent to collectors
[27-Aug-2009 09:35:41] <mrayzenoss> we're pulling 2.4.3 and 2.4.4 should be out tomorrow, possibly late today
[27-Aug-2009 09:36:03] <jb> ARGH
[27-Aug-2009 09:36:22] <ckrough> <nelson> ha ha </nelson>
[27-Aug-2009 09:36:30] <jb> now the graph is back to all 0's
[27-Aug-2009 09:36:31] <jb> :/
[27-Aug-2009 09:38:27] <rmatte> there, fixed the script
[27-Aug-2009 09:41:51] <dollarbang> Getting ready to install 2.4.3 of Zenoss, I'm getting this error backing up the etc directory "cp: cannot open `etc/zeo.zdsock' for reading: No such device or address" the file has these attributes "0 srwx------ 1 zenoss zenoss 0 Aug 13 11:03 zeo.zdsock".
[27-Aug-2009 09:42:26] <jb> dont install 2.4.3
[27-Aug-2009 09:42:41] <jb> wait for 2.4.4..
[27-Aug-2009 09:44:17] <dollarbang> jb: why, is there something wrong with 2.4.3? the error has to do with a file that was created by either the stacker during the previous installation, or Linux doesn't like the file.
[27-Aug-2009 09:44:24] <jb> scroll up..
[27-Aug-2009 09:45:09] <dollarbang> jb: oh, just saw this: mrayzenoss: we're pulling 2.4.3 and 2.4.4 should be out tomorrow, possibly late today
[27-Aug-2009 09:45:18] <jb> http://dev.zenoss.org/trac/ticket/5466
[27-Aug-2009 09:45:29] <rmatte> mrayzenoss: is a patch available yet?
[27-Aug-2009 09:45:46] <rmatte> I'd rather patch now rather than re-upgrade everything
[27-Aug-2009 09:46:09] <rmatte> I won't be able to do another full upgrade until next tuesday anyways
[27-Aug-2009 09:59:05] <rmatte> great, now I'm somehow screwed this ZenPack up even more, wtf
[27-Aug-2009 09:59:09] <rmatte> I've*
[27-Aug-2009 10:00:10] <rmatte> oh, hmmm
[27-Aug-2009 10:02:10] <rmatte> needed to restart Zenoss
[27-Aug-2009 10:06:43] <vinzpr> hello
[27-Aug-2009 10:07:19] <vinzpr> anybody here?
[27-Aug-2009 10:08:50] <jb> ?
[27-Aug-2009 10:09:02] <vinzpr> hi
[27-Aug-2009 10:10:17] <rmatte> yes, there are a few people here
[27-Aug-2009 10:10:21] <rmatte> if you have a question just ask
[27-Aug-2009 10:10:36] <jb> yes, don't wait on us to ask if you need help..
[27-Aug-2009 10:11:08] <vinzpr> ok..
[27-Aug-2009 10:12:21] <vinzpr> i upgrade from 2.4.2 to 2.4.3 when I check the port 8100 if open by doing telnet remotely it didnt response but doing it localhost it works
[27-Aug-2009 10:12:35] <rmatte> check your firewall
[27-Aug-2009 10:12:46] <vinzpr> i've already disable firewall
[27-Aug-2009 10:12:50] <jb> can you telnet from localhost?
[27-Aug-2009 10:12:55] <rmatte> and it should be 8080, unless you specifically changed it?
[27-Aug-2009 10:12:57] <vinzpr> yes it response
[27-Aug-2009 10:13:11] <jb> telnet localhost 8100 works?
[27-Aug-2009 10:13:11] <rmatte> 8100 is for something else, it's not the default web port
[27-Aug-2009 10:13:27] <vinzpr> but on the other machine doing telnet host.domain.com 8100 it didnt response
[27-Aug-2009 10:13:41] <vinzpr> is for zeodb
[27-Aug-2009 10:13:45] <jb> yeah its zeodb
[27-Aug-2009 10:14:15] <vinzpr> i tried to remove the package again then install fresh 2.4.3 its didnt work
[27-Aug-2009 10:14:19] <jb> its probably bound to localhost
[27-Aug-2009 10:14:21] <vinzpr> but on 2.4.2 it works
[27-Aug-2009 10:14:23] <jb> or you have a firewall issue
[27-Aug-2009 10:14:37] <vinzpr> i close the firewall already
[27-Aug-2009 10:14:43] <vinzpr> is open on all ports
[27-Aug-2009 10:14:51] <jb> whats the problem?
[27-Aug-2009 10:14:55] <jb> why do you need to connect to it remotely?
[27-Aug-2009 10:15:10] <vinzpr> im doing a distributed collector
[27-Aug-2009 10:15:29] <jb> you can try changing "localhost:8100" in zope.conf to IP:8100
[27-Aug-2009 10:15:30] <vinzpr> and changing the config on zope.conf and pointing to your main server
[27-Aug-2009 10:15:34] <jb> not sure
[27-Aug-2009 10:15:59] <vinzpr> in 2.4.2 it works well
[27-Aug-2009 10:16:11] <vinzpr> especially on dong zendmd on remote
[27-Aug-2009 10:16:39] <vinzpr> it response but now it didnt
[27-Aug-2009 10:17:19] <vinzpr> i have to upgrade into latest version because of the fix on keyerror
[27-Aug-2009 10:17:28] <jb> 2.4.3 has other issues
[27-Aug-2009 10:17:34] <jb> its being pulled..
[27-Aug-2009 10:17:42] <jb> 2.4.4 will be out late today or tommorow
[27-Aug-2009 10:17:58] <vinzpr> ok
[27-Aug-2009 10:18:05] <vinzpr> thats nice to hear
[27-Aug-2009 10:18:08] <mrayzenoss> as soon as the patch is available, I'll put it here and the forums
[27-Aug-2009 10:18:17] <vinzpr> hi matt
[27-Aug-2009 10:18:26] <vinzpr> i emailed you awhile ago
[27-Aug-2009 10:19:13] <vinzpr> i hope it works on the latest version
[27-Aug-2009 10:21:04] <vinzpr> so the problem of port 8100 will be fix on 2.4.4?
[27-Aug-2009 10:21:42] <mrayzenoss> the only fix going into 2.4.4 is the patch to http://dev.zenoss.org/trac/ticket/5466
[27-Aug-2009 10:21:48] <rmatte> mrayzenoss: awesome, thanks. I'll have that updated ZenPack to you in moments
[27-Aug-2009 10:21:58] <jb> this graph is really starting to annoy me
[27-Aug-2009 10:22:04] <jb> it makes no sense.
[27-Aug-2009 10:22:24] <rmatte> lol
[27-Aug-2009 10:22:37] <rmatte> neither does life
[27-Aug-2009 10:22:57] <rmatte> ...and we have cpu graphs!
[27-Aug-2009 10:23:00] * rmatte rejoices
[27-Aug-2009 10:23:13] <rmatte> although I think some other graphs are still malfunctioning, but at least this knocks down the majority
[27-Aug-2009 10:24:48] <chickenandbeans> Hi, anyone here good with zendmd? Need to know how to use the manage_addDataPointGraphPoints method
[27-Aug-2009 10:24:55] <chickenandbeans> anyone?
[27-Aug-2009 10:26:23] <vinzpr> hi matt...i saw the ticket..is the latest patch includes the problem on port 8100?
[27-Aug-2009 10:26:45] <mrayzenoss> I don't think so
[27-Aug-2009 10:26:49] <mrayzenoss> is there a ticket for that?
[27-Aug-2009 10:27:11] <rmatte> no, he's just noticing that he can't hit port 8100 remotely on one of his boxes
[27-Aug-2009 10:27:16] <vinzpr> no there is not ticket for that
[27-Aug-2009 10:27:19] <rmatte> but it's probably bound to localhost by default, no?
[27-Aug-2009 10:27:25] <jb> yes.. zope.conf
[27-Aug-2009 10:28:11] <vinzpr> zope.conf i point to the other server
[27-Aug-2009 10:29:49] <vinzpr> my main server is main.domain.com and on the other server i change the config of zope.conf localhost:8100 into main.domain.com:8100
[27-Aug-2009 10:29:54] <vinzpr> and it didnt work
[27-Aug-2009 10:30:44] <vinzpr> trying on my other server zendmd, it didnt response can connect to main.domain.com
[27-Aug-2009 10:32:04] <chickenandbeans> so can anyone here help?
[27-Aug-2009 10:38:07] <vinzpr> is this issue can be fixed on 2.4.4?
[27-Aug-2009 10:40:30] <mrayzenoss> vinzpr: no. There's no ticket and 2.4.4 is going to have 1 ticket
[27-Aug-2009 10:40:42] <mrayzenoss> vinzpr: so help me understand
[27-Aug-2009 10:41:22] <mrayzenoss> vinzpr: You're running a distributed collector and it's not respecting the zope.conf?
[27-Aug-2009 10:42:52] <vinzpr> im running distributed collector then i change localhost:8100 to main.domain.com:8100 of zope.conf to the other machine
[27-Aug-2009 10:43:43] <vinzpr> so my zeodb on the second machine will be the main.domain.com
[27-Aug-2009 10:45:20] <vinzpr> but on 2.4.2 it running well
[27-Aug-2009 10:45:29] <vinzpr> i have no issue on it
[27-Aug-2009 10:46:04] <vinzpr> i can even telnet main.domain.com 8100
[27-Aug-2009 10:46:09] <vinzpr> and it does response
[27-Aug-2009 10:48:15] <mrayzenoss> and all the conf files have been updated and you've restarted everything?
[27-Aug-2009 10:48:24] <vinzpr> yes
[27-Aug-2009 10:48:57] <vinzpr> except for zeodb
[27-Aug-2009 10:53:15] <vinzpr> any suggestion?
[27-Aug-2009 10:53:21] <vinzpr> how can i fix this?
[27-Aug-2009 10:55:39] <dberger> is there any reason why an event submitted via xmlrpc wouldn't appear in the event console?
[27-Aug-2009 10:59:53] <mrayzenoss> dberger: is the severity too low?
[27-Aug-2009 11:00:45] <venturaville> anyone here familiar with pywbem?
[27-Aug-2009 11:01:05] <mrayzenoss> venturaville: Egor and Kells
[27-Aug-2009 11:01:11] <vinzpr> hi matt..do you understand now my problem on port 8100?
[27-Aug-2009 11:01:13] <dberger> mrayzenoss, no
[27-Aug-2009 11:01:20] <dberger> severity is critical
[27-Aug-2009 11:01:24] <mrayzenoss> dberger: version?
[27-Aug-2009 11:01:27] <venturaville> mrayzenoss: thanks ... I'll keep an eye out for them
[27-Aug-2009 11:03:43] <dberger> mrayzenoss, remind me how to check that
[27-Aug-2009 11:08:06] <vinzpr> mrayzenoss, so when i excute zendmd on my remote server it response me like this WARNING:ZEO.zrpc:(11602) CW: error connecting to ('main.domain.com', 8100): ECONNREFUSED
[27-Aug-2009 11:09:04] <mrayzenoss> dberger: Settings->Versions
[27-Aug-2009 11:09:20] <dberger> mrayzenoss, 2.4.1
[27-Aug-2009 11:09:51] <dberger> mrayzenoss, sorry...i've had several versions installed and i got a little confused
[27-Aug-2009 11:10:28] <mrayzenoss> vinzpr: OK, I think the zeo.conf got changed
[27-Aug-2009 11:11:41] <vinzpr> mrayzenoss, so this is the problem on 2.4.3?
[27-Aug-2009 11:11:43] <mrayzenoss> I think this ticket may be the one that affected you
[27-Aug-2009 11:11:44] <mrayzenoss> http://dev.zenoss.org/trac/ticket/5255
[27-Aug-2009 11:14:40] <vinzpr> im sorry but on 2.4.2 its working fine
[27-Aug-2009 11:14:52] <vinzpr> and the ticket specified the target fix is on 2.4.3
[27-Aug-2009 11:15:09] <mrayzenoss> yeah, it was patched in 2.4.3
[27-Aug-2009 11:15:24] <mrayzenoss> so when you upgraded to 2.4.3 it broke your configuration
[27-Aug-2009 11:15:45] <mrayzenoss> distributed collectors aren't supported in Core, so this change wasn't tested in Core
[27-Aug-2009 11:19:17] <vinzpr> so do i have to manually execute the zeopack.py?
[27-Aug-2009 11:19:46] <daMaestro> i'm having issues mapping events to Ignore
[27-Aug-2009 11:20:01] <daMaestro> it doesn't seem like the regex is being applied to messages
[27-Aug-2009 11:21:00] <mrayzenoss> vinzpr: I'm sorry, I really don't know much about distributed collectors
[27-Aug-2009 11:21:26] <mrayzenoss> vinzpr: I don't mess with them myself, since they're in Enterprise
[27-Aug-2009 11:21:37] <mrayzenoss> anyone else using distributed collectors with Core? rmatte?
[27-Aug-2009 11:21:39] <daMaestro> Accepted publickey for username|pam_unix\(sshd:session\): session closed for user username
[27-Aug-2009 11:21:58] <daMaestro> that is my regex, ssh is the event key.. and it's part of /Ignore .. but it's not matching any events
[27-Aug-2009 11:32:06] <vinzpr> hi mrayzenoss, i got it and i was able to telnet on it
[27-Aug-2009 11:32:36] <mrayzenoss> I responded to your forum thread
[27-Aug-2009 11:32:37] <vinzpr> your right the problem is on the config file
[27-Aug-2009 11:32:47] <mrayzenoss> you'll probably want to update the community distributed collector docs
[27-Aug-2009 11:32:59] <jb> it was zope.conf?
[27-Aug-2009 11:33:04] <jb> binding to localhost?
[27-Aug-2009 11:33:15] <vinzpr> and then i notice in zeo.conf has a address localhost:8100
[27-Aug-2009 11:33:22] <vinzpr> it should be address 8100
[27-Aug-2009 11:33:25] <vinzpr> thats its
[27-Aug-2009 11:33:26] <jb> isnt that what I told you an hour ago?
[27-Aug-2009 11:33:36] <vinzpr> sure mray i will update the forum
[27-Aug-2009 11:33:40] <mrayzenoss> thanks
[27-Aug-2009 11:33:42] <vinzpr> and i will also contribute on it
[27-Aug-2009 11:34:16] <vinzpr> i will udpate the community distribute collectors docs soon with failover of zenoss
[27-Aug-2009 11:34:48] <vinzpr> thanks also jb
[27-Aug-2009 11:35:02] <jb> np
[27-Aug-2009 11:35:07] <jb> thanks for updating the docs
[27-Aug-2009 11:35:09] <vinzpr> maybe i will wait for the release of 2.4.4 then upgrade it again
[27-Aug-2009 11:35:12] <jb> many people have been looking for docs
[27-Aug-2009 11:36:05] <vinzpr> ok..thank you guys again let me do some testing now..
[27-Aug-2009 11:44:40] <rmatte> nah, we tried distributed collectors with core but it was too much of a hassle
[27-Aug-2009 11:44:49] <rmatte> so we're just using separate instances
[27-Aug-2009 12:04:22] <rmatte> hmmmm, events don't appear to automatically age in 2.4.3 either
[27-Aug-2009 12:26:18] <rmatte> mrayzenoss: events don't seem to be automatically aging at all in 2.4.3 either
[27-Aug-2009 12:26:31] <rmatte> not sure if that's it's own issue or if it's related to the clear events issue
[27-Aug-2009 12:26:32] <mrayzenoss> I'll pass that on
[27-Aug-2009 12:26:40] <mrayzenoss> probably related, I'll check with devs
[27-Aug-2009 12:27:04] <rmatte> I had events (that came in as traps) from yesterday still sitting there when I have it set to age out after an hour
[27-Aug-2009 12:35:04] <rmatte> mrayzenoss: you might want to change the channel topic
[27-Aug-2009 12:35:05] <rmatte>
[27-Aug-2009 12:38:30] <rmatte> mrayzenoss: the trac ticket has been set to fixed, resolution is patch 14744, wondering if we can go ahead with patching?
[27-Aug-2009 12:38:59] <mrayzenoss> QA hasn't finished verifying it
[27-Aug-2009 12:39:03] <rmatte> k
[27-Aug-2009 12:39:05] <mrayzenoss> but they're working on it
[27-Aug-2009 12:39:20] <mrayzenoss> you can try it if you like
[27-Aug-2009 12:39:27] <rmatte> no thanks, I'll wait
[27-Aug-2009 12:39:55] <chudler> take you're time, get it right :-) nobody wants to see a 2.4.5 tomorrow
[27-Aug-2009 12:39:57] <mrayzenoss> I'll be sending out announcements and whatnot once the patch is tested
[27-Aug-2009 12:39:57] <rmatte> I'm just anxious because we've probably had zero monitoring on certain devices for over 48 hours
[27-Aug-2009 12:40:19] <mrayzenoss> ugh
[27-Aug-2009 12:40:28] <rmatte> if something bad happens our clients will whoop us lol
[27-Aug-2009 12:40:33] <mrayzenoss> sounds like another test case we need
[27-Aug-2009 12:40:52] <rmatte> yeh
[27-Aug-2009 12:40:57] <mrayzenoss> feel free to jump in: http://blog.zenoss.com/2009/07/10/help-zenoss-expand-our-testing/
[27-Aug-2009 12:41:04] <rmatte> aging events would be a nice test to do too before release
[27-Aug-2009 12:41:27] <mrayzenoss> adding that to the ticket
[27-Aug-2009 12:45:21] <rmatte> k
[27-Aug-2009 12:46:09] <rmatte> I'm going to add reports to that, since obviously not much attention is paid to them
[27-Aug-2009 12:46:26] <rmatte> I guess it's not considered a "Critical" component
[27-Aug-2009 12:47:59] <rmatte> so did they say the aging is related?
[27-Aug-2009 12:48:06] <rmatte> or are they not sure yet?
[27-Aug-2009 12:48:33] <mrayzenoss> it should be related
[27-Aug-2009 12:48:43] <mrayzenoss> they're going to test it with a 1-hour age-out
[27-Aug-2009 12:48:44] <rmatte> k
[27-Aug-2009 12:49:03] <rmatte> I need to log a feature request to have the age-out specified in seconds
[27-Aug-2009 12:49:15] <rmatte> because we'd rather do it for a half hour, but "0.5" doesn't work
[27-Aug-2009 12:52:24] <mrayzenoss> yeah, that sounds valid
[27-Aug-2009 12:52:55] <mrayzenoss> throw in a patch and it might make it into King Crab
[27-Aug-2009 12:53:00] <jb> so is there any other way to troubleshoot graphs reporting 0?
[27-Aug-2009 12:53:53] <mrayzenoss> if you turn up the logging, you can see the values getting recorded
[27-Aug-2009 12:54:08] <jb> logging on?
[27-Aug-2009 12:54:12] <jb> zenperfsnmp?
[27-Aug-2009 12:54:25] <mrayzenoss> yeah
[27-Aug-2009 12:55:06] <jb> ok set to debug
[27-Aug-2009 13:04:39] <rmatte> timeout?
[27-Aug-2009 13:05:34] <mrayzenoss1> changed rooms
[27-Aug-2009 13:05:37] <mrayzenoss1> in a meeting
[27-Aug-2009 13:05:39] <rmatte> ah
[27-Aug-2009 13:07:58] <rmatte> the other thing is, after the patch is applied and Zenoss is restarted, the collector should start fresh and not think anything is down I assume?
[27-Aug-2009 13:08:11] <rmatte> or will it still think devices are down until a fake clear event is generated?
[27-Aug-2009 13:08:11] <dberger> so, I can see events in device list or in the device issues portlet, but not in the event console
[27-Aug-2009 13:08:22] <dberger> i'm submitting these events via xmlrpc
[27-Aug-2009 13:08:34] <mrayzenoss1> dberger: what maintenance state are they in?
[27-Aug-2009 13:08:40] <dberger> new
[27-Aug-2009 13:08:51] <rmatte> "new" isn't a maintenance state?
[27-Aug-2009 13:09:19] <rmatte> maintenance states are Production, Pre-Productions, Test, Maintenance, and Decommissioned
[27-Aug-2009 13:09:45] <mrayzenoss1> probably Production then
[27-Aug-2009 13:10:18] <mrayzenoss1> just that some maintenance states don't show on the dashboard, but you're seeing the opposite of that
[27-Aug-2009 13:10:23] <mrayzenoss1> on dash, no event console?
[27-Aug-2009 13:10:24] <dberger> I see prodState = Production(1000)
[27-Aug-2009 13:10:32] <rmatte> yeh, so production
[27-Aug-2009 13:11:15] <rmatte> I think he's seeing the count of the events in the device list, but not actually seeing the event in the event console
[27-Aug-2009 13:11:24] <dberger> right
[27-Aug-2009 13:11:33] <rmatte> hmmm
[27-Aug-2009 13:11:44] <rmatte> and you're using xmlrpc code to generate the events?
[27-Aug-2009 13:11:48] <dberger> and if i click on the devie in the device list, it gives me an event console specific to that device, and i see it there
[27-Aug-2009 13:11:53] <dberger> yes
[27-Aug-2009 13:11:58] <rmatte> can you pastebin the code?
[27-Aug-2009 13:12:56] <rmatte> (other non-xmlrpc generated events are displaying in the event console I assume?)
[27-Aug-2009 13:13:13] <dberger> yes
[27-Aug-2009 13:13:30] <rmatte> k, pastebin the code and we'll try to figure out what you're missing
[27-Aug-2009 13:16:12] <rmatte> mloven always reminds me of McLoving
[27-Aug-2009 13:16:16] <rmatte> McLovin*
[27-Aug-2009 13:16:17] <rmatte> rather
[27-Aug-2009 13:17:29] <dberger> http://pastebin.com/mc6dcd09
[27-Aug-2009 13:20:54] <rmatte> well, first thing I notice is that you're setting an event summary but not an event message
[27-Aug-2009 13:23:02] <rmatte> doesn't look like you're specifying an event class either
[27-Aug-2009 13:23:19] <rmatte> ah nevermind
[27-Aug-2009 13:23:22] <rmatte> testOrganizer
[27-Aug-2009 13:23:38] <dberger> oh right, i should include that
[27-Aug-2009 13:23:48] <rmatte> do you have /testOrganizer as an event class in Events?
[27-Aug-2009 13:24:06] <dberger> yes
[27-Aug-2009 13:24:14] <rmatte> k
[27-Aug-2009 13:24:45] <rmatte> there's no agent specified, not sure if that's going to cause issues or not
[27-Aug-2009 13:24:55] <dberger> agent?
[27-Aug-2009 13:25:09] <rmatte> example: agent zensyslog
[27-Aug-2009 13:26:22] <rmatte> but it may do that on it's own
[27-Aug-2009 13:26:40] <dberger> agent gets left blank
[27-Aug-2009 13:27:38] <rmatte> k
[27-Aug-2009 13:27:58] <rmatte> other than that, I can't really see any other issues with i
[27-Aug-2009 13:28:00] <rmatte> it*
[27-Aug-2009 13:28:15] <rmatte> what severity does it come in as?
[27-Aug-2009 13:28:20] <dberger> 5
[27-Aug-2009 13:28:29] <rmatte> where are you setting that?
[27-Aug-2009 13:28:35] <rmatte> oh nevermind
[27-Aug-2009 13:28:38] <rmatte> I see it now
[27-Aug-2009 13:29:26] <rmatte> yeh, no idea, it looks good from what I can see other than message missing
[27-Aug-2009 13:30:05] <dberger> which gets copied from the summary field before going into the db
[27-Aug-2009 13:32:17] <dberger> could it have something to do with the component?
[27-Aug-2009 13:32:36] <dberger> that's just something made up that hasn't been configured with zenoss
[27-Aug-2009 13:32:40] <rmatte> possibly, but I doubt it
[27-Aug-2009 13:32:46] <rmatte> you could try leaving component blank
[27-Aug-2009 13:32:59] <rmatte> there is no component configuration in Zenoss
[27-Aug-2009 13:33:04] <rmatte> component can be anything
[27-Aug-2009 13:33:26] <rmatte> it's just a place to list the affected component
[27-Aug-2009 13:33:28] <dberger> that's what i thought
[27-Aug-2009 13:33:28] <rmatte> so an interface name
[27-Aug-2009 13:33:31] <rmatte> or PRINTER
[27-Aug-2009 13:33:33] <rmatte> or whatever
[27-Aug-2009 13:34:47] <dberger> BINGO!
[27-Aug-2009 13:34:53] <rmatte> ?
[27-Aug-2009 13:35:04] <dberger> 'eventClass':'/testOrganizer'
[27-Aug-2009 13:35:06] <dberger> not
[27-Aug-2009 13:35:09] <rmatte> oh
[27-Aug-2009 13:35:12] <rmatte> you were missing the slash
[27-Aug-2009 13:35:14] <dberger> 'eventClass':'testOrganizer'
[27-Aug-2009 13:35:18] <rmatte> yeh, that's would definitely cause issues
[27-Aug-2009 13:35:24] <rmatte> lol
[27-Aug-2009 13:35:52] <rmatte> I just didn't know whether the xmlrpc code needed it or not
[27-Aug-2009 13:35:54] <dberger> funny that that ends up being inconsistent
[27-Aug-2009 13:35:56] <rmatte> but apparently it does
[27-Aug-2009 13:58:26] <rmatte> mrayzenoss1:
[27-Aug-2009 13:58:27] <rmatte> Changed 7 minutes ago by bbibeault ¶
[27-Aug-2009 13:58:27] <rmatte> * patch_state changed from Accepted to Patched
[27-Aug-2009 13:58:58] <rmatte> does that mean it passed QA?
[27-Aug-2009 14:00:45] <mrayzenoss1> I'm bugging them about it right now, making sure they tested auto-aging
[27-Aug-2009 14:01:13] <rmatte> k cool
[27-Aug-2009 14:01:18] <mrayzenoss1> but clears are definitely working again
[27-Aug-2009 14:01:23] mrayzenoss1 is now known as mrayzenoss
[27-Aug-2009 14:01:24] <rmatte> good stuff
[27-Aug-2009 14:01:41] <rmatte> the aging is pretty important to our setup though at the moment, so that's definitely a must for me
[27-Aug-2009 14:01:44] <mrayzenoss> and the bits are being uploaded to SourceForge
[27-Aug-2009 14:01:56] <rmatte> cool
[27-Aug-2009 14:02:04] <rmatte> I can just do this as a zenpatch though right?
[27-Aug-2009 14:02:22] <rmatte> I don't have time to do a full blown upgrade of each system until next Tuesday night
[27-Aug-2009 14:02:39] <rmatte> well, I don't have an official change window until then
[27-Aug-2009 14:02:46] <rmatte> but I can do emergency patching
[27-Aug-2009 14:04:10] <mrayzenoss> yeah 'zenpatch 14744' and 'zenactions restart'
[27-Aug-2009 14:04:27] <rmatte> nice, don't even have to restart all of zenoss
[27-Aug-2009 14:04:39] <rmatte> once I hear about the aging then I'll apply it
[27-Aug-2009 14:04:45] <mrayzenoss> yeah, it's zenactions blowing up and dying
[27-Aug-2009 14:04:53] <mrayzenoss> which sucks in many, many ways
[27-Aug-2009 14:04:54] <rmatte> lovely
[27-Aug-2009 14:05:45] <rmatte> actually, if the patch is officially final I'll patch
[27-Aug-2009 14:05:53] <rmatte> I assume it is since you said they were uploading the bits
[27-Aug-2009 14:05:56] <rmatte> ?
[27-Aug-2009 14:07:19] <rmatte> yeh, zenactions is responsible for aging
[27-Aug-2009 14:07:37] <rmatte> so I'll bet that's what was causing it
[27-Aug-2009 14:07:45] <mrayzenoss> yeah, we expect it works
[27-Aug-2009 14:07:46] <rmatte> so is it safe to patch or should I hold off?
[27-Aug-2009 14:08:01] <mrayzenoss> I think it's safe
[27-Aug-2009 14:08:15] <rmatte> k, I be patchin' then
[27-Aug-2009 14:09:13] <rmatte> I don't need to stop zenactions before patching do I?
[27-Aug-2009 14:09:18] <mrayzenoss> nope
[27-Aug-2009 14:09:20] <rmatte> k
[27-Aug-2009 14:09:36] <mrayzenoss> is zenactions still up? It's just in zombie mode if it is
[27-Aug-2009 14:09:51] <mrayzenoss> that stack trace just blows it up
[27-Aug-2009 14:10:46] <JKilgore> I'm trying to set the time on the zenoss appliance, but I can't get it to stick - anyone here that might be able to offer some help?
[27-Aug-2009 14:17:18] <rmatte> there, all our prod stuff is patched
[27-Aug-2009 14:17:21] <djack> hello everybody
[27-Aug-2009 14:17:31] <rmatte> I think it's still up yeh
[27-Aug-2009 14:17:38] <rmatte> let me check on the last box I have to patch
[27-Aug-2009 14:17:54] <djack> can somebody give me a hand ?
[27-Aug-2009 14:18:10] <djack> I have some problems with creation of rrd graphs ... and i don't know why
[27-Aug-2009 14:19:14] <rmatte> mrayzenoss: should I restart all of Zenoss, on each box, or will it come to it's senses on it's own?
[27-Aug-2009 14:19:17] <venturaville> jb: you awake?
[27-Aug-2009 14:19:24] <rmatte> like, if devices have already gone down during the issue
[27-Aug-2009 14:19:42] <rmatte> won't I need a full restart so that the collector finally realizes what's going on?
[27-Aug-2009 14:19:43] <mrayzenoss> rmatte: I think just zenactions
[27-Aug-2009 14:19:50] <rmatte> k
[27-Aug-2009 14:19:56] <djack> nobody?
[27-Aug-2009 14:20:10] <rmatte> djack: state the issue
[27-Aug-2009 14:21:06] <rmatte> mrayzenoss: yeh, it is running, so probably in zombie mode
[27-Aug-2009 14:21:30] <rmatte> probably why they didn't notice it during the QA for 2.3.3
[27-Aug-2009 14:21:32] <rmatte> erm
[27-Aug-2009 14:21:33] <rmatte> 2.4.3
[27-Aug-2009 14:21:48] <djack> well i added 6 server that are behind a firewall .... from the server command line i can ping them and do snmpwalks so i know the firewall does not blcok the snmp and icmp packages. however the perf data is not updated on the zenoss server ?
[27-Aug-2009 14:22:22] <mrayzenoss> djack: and they're in the right device class?
[27-Aug-2009 14:22:38] <djack> the same as the other servers
[27-Aug-2009 14:22:52] <djack> they are in /server/linux/
[27-Aug-2009 14:22:56] <rmatte> djack: when you snmpwalk them, does it seem like there are less lines than when you walk the others?
[27-Aug-2009 14:23:00] <rmatte> like, significantly less?
[27-Aug-2009 14:23:41] <rmatte> and are they using the same snmp config as the other linux boxes?
[27-Aug-2009 14:23:42] <djack> nope not at all and when i look at the device even all the installed rpm packages are in the list but no cpu/mem/disc and other very relevant data
[27-Aug-2009 14:23:54] <djack> rmatte yep
[27-Aug-2009 14:24:37] <rmatte> if you do something like: snmpwalk -v1 -c <community> <device> hrProcessorLoad
[27-Aug-2009 14:24:39] <rmatte> what do you see?
[27-Aug-2009 14:25:14] <venturaville> RecoveredSeekOperations=None TotalReadBlocksRequested=3097094828 TotalLargeWrites=0 TotalWriteBlocksRequested=1218624050 TotalBlocksRequested=20751582 Description="Media Access Statistics for Volume vol2" TotalLargeWriteBlocksRequested=0 UnrecoverableReadOperations=None WriteOperations=237045636 TotalCacheReadChecks=988275308 RecoveredReadOperations=None UnrecoverableSeekOperations=None ElementName="Volume vol2" AverageBlocksRequested=0 TotalCacheR
[27-Aug-2009 14:25:31] <venturaville> sweet ... finally getting volume stats out of LSI arrays....
[27-Aug-2009 14:25:38] <rmatte> nice
[27-Aug-2009 14:25:42] <rmatte> wrote a custom script I see
[27-Aug-2009 14:26:42] <venturaville> yep
[27-Aug-2009 14:27:03] <djack> nothing
[27-Aug-2009 14:28:12] <rmatte> mrayzenoss: I've got a couple of trap events that came in for reboots, so I'll leave them sit and see if they age. I'll let you know.
[27-Aug-2009 14:28:30] <djack> but neither for the servers that do enumerate the cpu memory and disc snmp data
[27-Aug-2009 14:28:59] <djack> the server are HP Proliant 365 and 380
[27-Aug-2009 14:29:10] <mrayzenoss> rmatte: yeah, me and QA are waiting on our time outs
[27-Aug-2009 14:29:11] <djack> with Suse Enterprise 11
[27-Aug-2009 14:29:20] <rmatte> hehe
[27-Aug-2009 14:30:10] <djack> rmate is ther info about the zenpatches and where to download them ?
[27-Aug-2009 14:30:30] <rmatte> djack: to fix 2.4.3 you mean?
[27-Aug-2009 14:30:34] <djack> yep
[27-Aug-2009 14:30:39] <rmatte> it's simple
[27-Aug-2009 14:30:42] <djack> i installed it ...
[27-Aug-2009 14:30:46] <rmatte> login to the zenoss server, become the zenoss user
[27-Aug-2009 14:30:51] <rmatte> then issue the following 2 commands:
[27-Aug-2009 14:30:53] <rmatte> zenpatch 14744
[27-Aug-2009 14:30:59] <rmatte> zenactions restart
[27-Aug-2009 14:31:01] <rmatte> that's it
[27-Aug-2009 14:31:10] <djack> well i don't have internet access from the monitoring machine
[27-Aug-2009 14:31:16] <rmatte> oh
[27-Aug-2009 14:31:24] <rmatte> one sec
[27-Aug-2009 14:32:07] <djack> but the fix will not fix the ping issues and my snmp problems i guess :-(
[27-Aug-2009 14:32:32] <rmatte> http://dev.zenoss.org/trac/changeset/14744
[27-Aug-2009 14:32:34] <rmatte> that's the patch
[27-Aug-2009 14:32:42] <rmatte> you can download it from the links on the bottom of the page
[27-Aug-2009 14:32:46] <rmatte> either as a diff or as a zip
[27-Aug-2009 14:33:14] <rmatte> and one second, I'll get the command to patch it
[27-Aug-2009 14:33:25] <rmatte> (download it as a diff and put it somewhere)
[27-Aug-2009 14:33:56] <djack> i have to download it upload it to a server scp it to an other server and then scp it to the zenoss server
[27-Aug-2009 14:34:04] <rmatte> well do that
[27-Aug-2009 14:34:12] <rmatte> once the diff is on the zenoss server you do...
[27-Aug-2009 14:34:41] <rmatte> patch /usr/local/zenoss/zenoss/Products/ZenEvents/zenactions.py < changeset_r14744.diff
[27-Aug-2009 14:34:47] <rmatte> then zenactions restart
[27-Aug-2009 14:35:20] <djack> oki
[27-Aug-2009 14:35:46] <rmatte> the patch might, if it's thinking those devices are down it might have something to do with it
[27-Aug-2009 14:35:51] <rmatte> apply the patch, wait a bit, and see
[27-Aug-2009 14:36:03] <rmatte> if it starts collecting then good, if not, then we'll have to figure out what's causing it
[27-Aug-2009 14:40:56] <djack> how can i see if the patch is working ?
[27-Aug-2009 14:41:12] <rmatte> did you patch and restart?
[27-Aug-2009 14:41:31] <djack> yep
[27-Aug-2009 14:41:47] <djack> I'm quick :-p
[27-Aug-2009 14:42:11] <rmatte> check $ZENHOME/log/zenactions.log
[27-Aug-2009 14:42:19] <rmatte> the latest entries should be good entries
[27-Aug-2009 14:42:28] <rmatte> whereas if you scroll up you'll see a bunch of python errors
[27-Aug-2009 14:43:14] <rmatte> You should see stuff like: 2009-08-27 15:38:53 INFO zen.ZenActions: Processed 1 commands in 0.002599
[27-Aug-2009 14:43:14] <rmatte> 2009-08-27 15:38:53 INFO zen.ZenActions: processed 0 rules in 0.01 secs
[27-Aug-2009 14:43:15] <rmatte> 2009-08-27 15:39:53 INFO zen.ZenActions: Running
[27-Aug-2009 14:43:56] <rmatte> egor, if you're running 2.4.3 you'll need to patch
[27-Aug-2009 14:44:33] <bigegor> i have 2.4.1 now.
[27-Aug-2009 14:44:41] <rmatte> k, then you're fine
[27-Aug-2009 14:44:55] <bigegor> i know
[27-Aug-2009 14:45:09] <rmatte> did you apply that security patch?
[27-Aug-2009 14:45:42] <rmatte> http://dev.zenoss.org/trac/ticket/5279
[27-Aug-2009 14:46:29] <djack> no python errors anymore in the zenactions.log
[27-Aug-2009 14:47:04] <bigegor> thanks for info
[27-Aug-2009 14:47:06] <venturaville> odd question: how do you deal with counters that only update every 10 minutes... is that something that should simply be changed for the device class?
[27-Aug-2009 14:47:20] <djack> but still no rrd and the status of the devies are still down
[27-Aug-2009 14:47:35] <rmatte> djack: so the devices are showing as down?
[27-Aug-2009 14:47:41] <djack> yep
[27-Aug-2009 14:47:54] <rmatte> djack: If they are showing as ping down then it's not even going to try to monitor them or generate RRDs
[27-Aug-2009 14:47:57] <rmatte> you need to resolve that
[27-Aug-2009 14:48:08] <rmatte> are they pingable?
[27-Aug-2009 14:48:45] <djack> yep even via the command ping tool in the zenoss interface :-(
[27-Aug-2009 14:49:11] <rmatte> djack: try restarting Zenoss and see if it still shows them as down
[27-Aug-2009 14:50:06] <djack> i read something about problems with zenping on the forums
[27-Aug-2009 14:50:24] <rmatte> I'm not having any issues with it
[27-Aug-2009 14:50:38] <rmatte> and I'm running 5 production Zenoss boxes
[27-Aug-2009 14:50:39] <djack> they stay down ?
[27-Aug-2009 14:50:48] <rmatte> stay down?
[27-Aug-2009 14:50:49] <rmatte> hunh?
[27-Aug-2009 14:50:58] <djack> yep they don't come up
[27-Aug-2009 14:51:08] <djack> grmbl
[27-Aug-2009 14:51:15] <jb> can you not ping them from the zenoss box?
[27-Aug-2009 14:51:19] <jb> using "ping"
[27-Aug-2009 14:51:22] <rmatte> jb: he can apparently
[27-Aug-2009 14:51:38] <rmatte> he can ping using the command in the Zenoss UI as well
[27-Aug-2009 14:51:53] <jb> zenping running?
[27-Aug-2009 14:52:11] <djack> zenoss@snlgwk01:~> zenping status
[27-Aug-2009 14:52:11] <djack> program running; pid=26301
[27-Aug-2009 14:52:35] <jb> hrm
[27-Aug-2009 14:53:00] <djack> strange problem he ....
[27-Aug-2009 14:53:13] <rmatte> try zenping run -v10
[27-Aug-2009 14:53:18] <rmatte> see if you get any strange errors
[27-Aug-2009 14:54:19] <venturaville> jb: I am now getting volume statistics out of the DS4xxx arrays we have :-)
[27-Aug-2009 14:54:23] <rmatte> hmmmm
[27-Aug-2009 14:54:33] <jb> venturaville: nice! what are you using?
[27-Aug-2009 14:54:38] <rmatte> when I do zenping run -v10 I'm seeing a lot of: DEBUG:zen.Ping:unexpected pkt
[27-Aug-2009 14:54:46] <rmatte> anyone else seeing that?
[27-Aug-2009 14:55:05] <rmatte> or is this another wonderful 2.4.3 nuance
[27-Aug-2009 14:55:05] <djack> DEBUG:zen.Ping:pj timeout for 10.253.2.3
[27-Aug-2009 14:55:06] <djack> DEBUG:zen.Ping:pj fail for 10.253.2.3
[27-Aug-2009 14:55:06] <djack> DEBUG:zen.ZenPing:Failed 10.253.2.3 267
[27-Aug-2009 14:55:06] <djack> WARNING:zen.ZenPing:ip 10.253.2.3 is down
[27-Aug-2009 14:55:06] <djack> DEBUG:zen.ZenPing:Queueing event {'severity': 5, 'component': '', 'agent': 'zenping', 'summary': 'ip 10.25 3.2.3 is down', 'manager': 'localhost', 'eventGroup': 'Ping', 'eventState': 0, 'device': 'frs-db01', 'even tClass': '/Status/Ping', 'ipAddress': '10.253.2.3'}
[27-Aug-2009 14:55:06] <rmatte> ?
[27-Aug-2009 14:55:08] <djack> DEBUG:zen.ZenPing:Total of 1 queued events
[27-Aug-2009 14:55:10] <djack> DEBUG:zen.Ping:pj timeout for 10.253.2.4
[27-Aug-2009 14:55:12] <venturaville> jb: I'm using pywbem to fetch the data through engenio provider (CIM -> SMI provider)
[27-Aug-2009 14:55:20] <jb> hmm
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.3'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.4'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.67'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.68'
[27-Aug-2009 14:55:29] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.69'
[27-Aug-2009 14:55:30] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.70'
[27-Aug-2009 14:55:32] <djack> DEBUG:zen.Ping:send icmp to '10.253.2.71'
[27-Aug-2009 14:55:34] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:36] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:37] <jb> whoa
[27-Aug-2009 14:55:38] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:39] <jb> use pastebin
[27-Aug-2009 14:55:39] <jb> heh
[27-Aug-2009 14:55:40] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:42] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:44] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:55:46] <djack> DEBUG:zen.Ping:unexpected pkt 10.253.2.129 <ICMP packet 0 0>
[27-Aug-2009 14:56:09] <djack> the 129 is a firewall ....
[27-Aug-2009 14:56:19] <rmatte> I'm wondering if those debug messages are normal or not
[27-Aug-2009 14:56:26] <rmatte> I've never done a zenping run, so I don't know
[27-Aug-2009 14:56:32] <djack> lol
[27-Aug-2009 14:56:35] <rmatte> and I don't have a zenoss box that isn't 2.4.3 at this point
[27-Aug-2009 14:56:57] <rmatte> venturaville: do you have any pre-2.4.3 boxes?
[27-Aug-2009 14:57:12] <djack> rmatte i also run it 2.4.3 version
[27-Aug-2009 14:57:21] <djack> DEBUG:zen.ZenPing:Queueing event {'severity': 5, 'component': '', 'agent': 'zenping', 'summary': 'ip 10.25 3.2.3 is down', 'manager': 'localhost', 'eventGroup': 'Ping', 'eventState': 0, 'device': 'frs-db01', 'even tClass': '/Status/Ping', 'ipAddress': '10.253.2.3'}
[27-Aug-2009 14:57:43] <rmatte> djack yours are actually showing fail lines
[27-Aug-2009 14:57:58] <djack> fail lines?
[27-Aug-2009 14:58:13] <rmatte> [03:48pm] <djack> DEBUG:zen.Ping:pj timeout for 10.253.2.3
[27-Aug-2009 14:58:13] <rmatte> [03:48pm] <djack> DEBUG:zen.Ping:pj fail for 10.253.2.3
[27-Aug-2009 14:58:13] <rmatte> [03:48pm] <djack> DEBUG:zen.ZenPing:Failed 10.253.2.3 267
[27-Aug-2009 14:58:32] <djack> yep idd
[27-Aug-2009 14:58:44] <djack> howcome is the question
[27-Aug-2009 14:59:24] <rmatte> show me what you get as ping replies from the device
[27-Aug-2009 14:59:26] <djack> why does a normal ping work but not the zenping ....
[27-Aug-2009 14:59:26] <rmatte> paste the output
[27-Aug-2009 14:59:51] <djack> zenoss@snlgwk01:~> ping 10.253.2.3
[27-Aug-2009 14:59:52] <djack> PING 10.253.2.3 (10.253.2.3) 56(84) bytes of data.
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=1 ttl=64 time=0.198 ms
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=2 ttl=64 time=0.213 ms
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=3 ttl=64 time=0.235 ms
[27-Aug-2009 14:59:52] <djack> 64 bytes from 10.253.2.129: icmp_seq=4 ttl=64 time=0.136 ms
[27-Aug-2009 14:59:54] <djack> ^C
[27-Aug-2009 14:59:56] <djack> --- 10.253.2.3 ping statistics ---
[27-Aug-2009 14:59:58] <djack> 4 packets transmitted, 4 received, 0% packet loss, time 3000ms
[27-Aug-2009 15:00:00] <djack> rtt min/avg/max/mdev = 0.136/0.195/0.235/0.039 ms
[27-Aug-2009 15:00:38] <djack> my point of view the ping seems good
[27-Aug-2009 15:00:51] <rmatte> those debug messages don't seem to affect my ping monitoring at all
[27-Aug-2009 15:00:55] <rmatte> so I'd imagine they are harmless
[27-Aug-2009 15:01:04] <rmatte> (the unexpected pkt ones)
[27-Aug-2009 15:01:09] <djack> hmm
[27-Aug-2009 15:02:25] <venturaville> rmatte: we have 2.4.2
[27-Aug-2009 15:02:27] <djack> i can't figure it out
[27-Aug-2009 15:02:49] <djack> i need the machine for stress testing of our internet site
[27-Aug-2009 15:02:59] <venturaville> rmatte: we have not deployed to production yet, though, as we are still trying to work out a deal for zenoss enterprise licensing
[27-Aug-2009 15:03:37] <djack> well if i can prove that zenoss works my boss his visa card is mine
[27-Aug-2009 15:04:20] <djack> but then i need to some this problem
[27-Aug-2009 15:05:39] <rmatte> venturaville: ah
[27-Aug-2009 15:06:09] <rmatte> djack: those are the only devices which you are having this issue with though right?
[27-Aug-2009 15:07:41] <djack> yep
[27-Aug-2009 15:08:18] <djack> i have 7 devices that work but they are hanging before the firewall and not behind .....
[27-Aug-2009 15:09:01] <jb> "Basically, the ICMP ping packet generated by an OS ping or a Zenoss ping command contains a different load than the packet generated by the zenping daemon. I've not verified this myself, but you can try using a sniffer to see if that is the case. "
[27-Aug-2009 15:09:05] <jb> http://forums.zenoss.com/viewtopic.php?t=10785
[27-Aug-2009 15:09:10] <jb> did you see this thread?
[27-Aug-2009 15:09:27] <djack> yes how can i fix that ?
[27-Aug-2009 15:09:47] <jb> sounds like a firewall issue..
[27-Aug-2009 15:09:58] <jb> look at the actual packets to see if they are different
[27-Aug-2009 15:10:00] <jb> (ping / zenping)
[27-Aug-2009 15:10:13] <jb> i don't see how they could be
[27-Aug-2009 15:10:15] <jb> but who knows..
[27-Aug-2009 15:10:47] <djack> why don't just use the standard ping command in stead of the zenping?
[27-Aug-2009 15:12:15] <djack> it seems strang that zenoss changes the load of the ping package....
[27-Aug-2009 15:14:37] <djack> jb that was my post ;-)
[27-Aug-2009 15:16:46] <djack> but there is no resolution ...
[27-Aug-2009 15:18:56] <rmatte> djack: oh, I might have an idea
[27-Aug-2009 15:19:08] <rmatte> actually, hmmm, let me check on something...
[27-Aug-2009 15:19:09] <djack> rmatte please you can make my night
[27-Aug-2009 15:19:20] <djack> and i can score points :-)
[27-Aug-2009 15:20:56] <rmatte> ah damn, nah, I was thinking about zMaxOIDPerRequest, but that doesn't apply to ping obviously
[27-Aug-2009 15:21:21] <rmatte> are you able to view the firewall logs?
[27-Aug-2009 15:21:26] <djack> nope
[27-Aug-2009 15:21:35] <djack> it is managed externally
[27-Aug-2009 15:21:46] <rmatte> I'd imagine it has something to do with the firewall blocking the ping packets coming from Zenoss because it doesn't like something about them
[27-Aug-2009 15:21:50] <rmatte> there's really no other explanation
[27-Aug-2009 15:22:20] <djack> is there a way to replace the zenping with the zencommand ping ?
[27-Aug-2009 15:22:32] <djack> on a device level?
[27-Aug-2009 15:22:34] <rmatte> well, there is, but it'd tedious
[27-Aug-2009 15:22:38] <rmatte> it's*
[27-Aug-2009 15:22:57] <djack> i can try nothing to lose
[27-Aug-2009 15:23:23] <rmatte> well hold on, I want to read up on this a bit more
[27-Aug-2009 15:23:52] <djack> oki
[27-Aug-2009 15:25:31] <rmatte> afk for a bit
[27-Aug-2009 15:27:20] <jb> hey egor
[27-Aug-2009 15:27:28] <bigegor> hey
[27-Aug-2009 15:28:07] <jb> gotta question
[27-Aug-2009 15:28:16] <jb> from socket import gethostbyaddr
[27-Aug-2009 15:28:16] <jb> evt.summary = "LTM Service Down"
[27-Aug-2009 15:28:16] <jb> ip = getattr(evt, 'bigipNotifyObjNode', 'Unknown')
[27-Aug-2009 15:28:16] <jb> port = getattr(evt, 'bigipNotifyObjPort', 'Unknown')
[27-Aug-2009 15:28:16] <jb> host = gethostbyaddr(ip)
[27-Aug-2009 15:28:18] <jb> evt.component = 'Node: %s:%s' % (host,port)
[27-Aug-2009 15:28:19] <jb> oops
[27-Aug-2009 15:28:23] <jb> http://pastebin.com/m64e86e09
[27-Aug-2009 15:28:39] <jb> that returns something like:
[27-Aug-2009 15:28:40] <jb> Component: Node: ('wrxxfscsaas01.fsc.follett.com', [], ['172.31.223.30']):82
[27-Aug-2009 15:29:00] <jb> whats the trick to get it to return "Component: Node: wrxxfscsass01.fsc.follett.com:82"
[27-Aug-2009 15:31:36] <bigegor> evt.component = 'Node: %s:%s' % (host[0],port)
[27-Aug-2009 15:31:47] <jb> [0] thats right
[27-Aug-2009 15:31:48] <jb> thanks
[27-Aug-2009 15:32:55] <rmatte> ok I'm back
[27-Aug-2009 15:33:55] <djack> and and ?
[27-Aug-2009 15:34:01] <rmatte> djack: can you do zenping run -v10 again but pastebin the whole output?
[27-Aug-2009 15:34:04] <rmatte> I'd like to skim over it
[27-Aug-2009 15:34:21] <rmatte> because you only pasted a small part of it before but we might have missed some other lines
[27-Aug-2009 15:36:16] <q7joey> i've applied the 14744 patch, but i seem to have a phantom process that generates events, but isn't in the process list. is this related to the 2.4.3 sickness?
[27-Aug-2009 15:36:53] <rmatte> q7joey: phantom process?
[27-Aug-2009 15:36:55] <rmatte> example?
[27-Aug-2009 15:37:48] <djack> http://pastebin.com/d69cb8139
[27-Aug-2009 15:38:11] <q7joey> events for mysqldump not running, but mysqldump isn't being monitored. i had previously had a regex that match mysqld (now /usr/libexec/myqld) so that is probably where it came from, but i can't find anywhere to delete it. do you want screenshots?
[27-Aug-2009 15:39:42] <rmatte> please
[27-Aug-2009 15:43:34] <q7joey> http://www.spiretech.com/~joey/event.png and os.png
[27-Aug-2009 15:45:24] <rmatte> q7joey: first thing I'd do is go in to zendmd and do a reindex() and commit()
[27-Aug-2009 15:45:37] <rmatte> I remember some process monitoring got out of wack for me one time and that fixed it
[27-Aug-2009 15:46:07] <q7joey> ok, trying that now.
[27-Aug-2009 15:46:56] <q7joey> now i have to watch to see if the timestamp updates
[27-Aug-2009 15:47:05] <jb> just clear the event..
[27-Aug-2009 15:47:07] <jb> and maybe remodel the device
[27-Aug-2009 15:47:21] <rmatte> yeh, a remodel couldn't hurt
[27-Aug-2009 15:47:24] <q7joey> i did the clear/remodel already. didn't help.
[27-Aug-2009 15:47:32] <jb> after you reindexed?
[27-Aug-2009 15:47:51] <q7joey> not yet. waiting to see if the timestamp increases.
[27-Aug-2009 15:48:53] <q7joey> timestamp just was touched. i'll try clear/remodel again.
[27-Aug-2009 15:49:45] <q7joey> and just to make sure, clear==move to history?
[27-Aug-2009 15:51:05] <jb> yep
[27-Aug-2009 15:52:01] <q7joey> it's back...
[27-Aug-2009 15:56:20] <q7joey> i guess i'll live with it for now and see if 2.4.4 clears it up.
[27-Aug-2009 15:56:34] <jb> i have some windows services that do the same thing
[27-Aug-2009 15:56:57] <djack> rmatte any ideas how to solve the zenping problem?
[27-Aug-2009 15:59:08] <jb> argh
[27-Aug-2009 15:59:14] <jb> soemthing keeps corrupting my cache
[27-Aug-2009 15:59:21] <q7joey> another issue that has been on the mailing list, but no answers to is gaps in graphs that started after i upgraded to 2.4.2. http://www.spiretech.com/~joey/render.png is a good example.
[27-Aug-2009 15:59:39] <q7joey> from looking at logs, i suspect 32/64 bit confusion.
[27-Aug-2009 15:59:51] <q7joey> 2009-08-27 13:57:56 ERROR netsnmp: truncating unsigned value to 32 bits (2)
[27-Aug-2009 16:00:01] <jb> that almost looks like zenperfsnmp problems
[27-Aug-2009 16:00:05] <jb> is it all of your SNMP graphs?
[27-Aug-2009 16:00:33] <q7joey> only graphs that have counters that wrap quickly and aren't using 64 bit counters.
[27-Aug-2009 16:00:57] <rmatte> q7joey: how many devices do you monitor on that box?
[27-Aug-2009 16:01:01] <q7joey> is there an easy way to force zenoss to choose 64 bit counters? it seems to be random as to which ones it decides to.
[27-Aug-2009 16:01:18] <straterra> not every device supports it, thats why
[27-Aug-2009 16:01:44] <q7joey> on a device that does 64, it only chooses 64 for some interfaces.
[27-Aug-2009 16:02:47] <rmatte> q7joey: you mean the templates?
[27-Aug-2009 16:02:58] <q7joey> i'm not running out of cpu or i/o. here is a good graph from the same device as the bad one: http://www.spiretech.com/~joey/render2.png the blank there is from my upgrading to 2.4.3
[27-Aug-2009 16:03:19] <rmatte> they are chosen based on interface type. If you click on an interface and check the "type" value, that's what it's chosen based on
[27-Aug-2009 16:03:27] <q7joey> right, it only uses the 64 bit templates for some interfaces, even if other have hc counters.
[27-Aug-2009 16:04:03] <rmatte> so if the interface type is ethernetCsmacd it uses that template, if it's ethernetCsmacd_64 it uses that template
[27-Aug-2009 16:04:12] <rmatte> so maybe the interfaces aren't identified properly via snmp
[27-Aug-2009 16:05:06] <q7joey> ds3_64 doesn't seem to use the hc values. is it only ethernetcsmacd_64 that will do it?
[27-Aug-2009 16:05:35] <rmatte> q7joey: the templates do whatever they are configured to do
[27-Aug-2009 16:06:08] <q7joey> so if i clone ether_64 to ds3_64, is that the method?
[27-Aug-2009 16:06:57] <rmatte> q7joey: that would be your best bet
[27-Aug-2009 16:07:07] <rmatte> cloning a non-64bit template certainly wouldn't help
[27-Aug-2009 16:07:30] <q7joey> i guess what i'm asking is will it magically map a ds3_64 interface type to use a ds3_64 template?
[27-Aug-2009 16:07:37] <rmatte> correct
[27-Aug-2009 16:07:47] <rmatte> as long as the template is in /Devices
[27-Aug-2009 16:08:14] <rmatte> you need to copy templates and make new ones for each interface type
[27-Aug-2009 16:08:18] <rmatte> I've made like 20 of them so far
[27-Aug-2009 16:08:33] <q7joey> i'm going to experiment right now...
[27-Aug-2009 16:12:47] <q7joey> ok, no obvious clone option. what is the right way to copy a template in /devices?
[27-Aug-2009 16:13:01] <mrayzenoss> I'm gonna go catch a flight home, 2.4.4 should be out in the morning. The code has been tagged, the 1 patch is the only difference from 2.4.3
[27-Aug-2009 16:13:23] <mrayzenoss> q7joey: there's a copy option for templates
[27-Aug-2009 16:13:50] <q7joey> but that seems to copy to another path. can i copy to /devices and it will make a new name?
[27-Aug-2009 16:13:54] <rmatte> yeh, go to /Devices, click on the template tab, put a checkmark next to the template you want to copy
[27-Aug-2009 16:14:00] <rmatte> then from the dropdown just select copy template
[27-Aug-2009 16:14:02] <rmatte> and copy it to /
[27-Aug-2009 16:14:12] <rmatte> then you'll see copy_of_whatever
[27-Aug-2009 16:14:18] <rmatte> go in to that and rename it to whatever you want
[27-Aug-2009 16:14:25] <q7joey> copy to /devices, right?
[27-Aug-2009 16:15:18] <rmatte> yes
[27-Aug-2009 16:17:20] <q7joey> ah, i see. everything is relative to /devices...
[27-Aug-2009 16:18:13] <rmatte> yes
[27-Aug-2009 16:21:33] <djack> hmmm
[27-Aug-2009 16:24:43] <rmatte> for interface templates anyways
[27-Aug-2009 16:25:04] <rmatte> /Devices is the "root" though, if that's what you mean
[27-Aug-2009 16:28:13] <q7joey> i'm waiting to see if my 64 bit magic worked. is there any easy way to convert 32 bit rrd to 64 bit to keep my history? i'm guessing not.
[27-Aug-2009 16:29:50] <rmatte> I don't believe so
[27-Aug-2009 16:30:13] <rmatte> if you google around you might find some way though
[27-Aug-2009 16:31:39] <q7joey> hmm, i'm looking at the rrd files and there may not be any real difference internally. a rename may be all i need...
[27-Aug-2009 16:34:06] <rmatte> that's entirely possible
[27-Aug-2009 16:34:11] <rmatte> RRD is pretty standard
[27-Aug-2009 16:34:51] <q7joey> i was just thinking that it might have raw 32/64 data and therefore would be different. things look ok after my copy, but i'll wait for a few updates and see if it survives.
[27-Aug-2009 16:37:48] <rmatte> I think it just stores values
[27-Aug-2009 16:38:08] <rmatte> the amount of traffic passing through a 32 bit interface is technically the same as when it's passing through a 64
[27-Aug-2009 16:38:16] <rmatte> in terms of throughput and speed
[27-Aug-2009 16:38:54] <q7joey> but a 64 bit counter could potentially create more than 32 bits worth of data per sample (that's the whole reason to use 64 bit counters). but it looks like zenoss doesn't do anything different with rrd.
[27-Aug-2009 16:39:03] <rmatte> yeh, that's true
[27-Aug-2009 16:39:12] <rmatte> cool
[27-Aug-2009 16:39:33] <rmatte> I think the difference in processing is just done internally within Zenoss, the way it stores the data to the RRD is most likely the same though
[27-Aug-2009 16:40:04] <rmatte> man, it's so nice to have all these systems patched and know that they are working properly
[27-Aug-2009 16:40:18] <rmatte> I even patched the ZVersion.py file, so I'm effectively running 2.4.4 on everything
[27-Aug-2009 16:41:20] <rmatte> now tomorrow I can focus on fixing up some reports
[27-Aug-2009 16:43:28] <rmatte> let's just hope there isn't a version 2.4.5 in the works by tomorrow
[27-Aug-2009 16:43:30] <rmatte>
[27-Aug-2009 16:44:07] <djack> lol
[27-Aug-2009 16:44:22] <djack> well i hope it will fix the zenping payload issue :-p
[27-Aug-2009 16:44:56] <rmatte> did you ever pastebin that output?
[27-Aug-2009 16:45:12] <djack> yeah i did
[27-Aug-2009 16:45:25] <djack> http://pastebin.com/d69cb8139
[27-Aug-2009 16:45:28] <rmatte> thanks
[27-Aug-2009 16:45:34] <rmatte> I'm curious to see what's causing that
[27-Aug-2009 16:45:42] <djack> me to
[27-Aug-2009 16:47:34] <rmatte> http://forums.zenoss.com/viewtopic.php?p=35458
[27-Aug-2009 16:47:42] <rmatte> that issue has been around for a while it'd seem
[27-Aug-2009 16:47:49] <djack> idd
[27-Aug-2009 16:48:07] <djack> nobody seems to solve it ....
[27-Aug-2009 16:49:01] <rmatte> I have to leave in about 15 mins, but if you come around tomorrow I'll help you troubleshoot this more
[27-Aug-2009 16:49:16] <rmatte> I'd like to know the answer in case I ever run across it, or in case we can get a patch made for it
[27-Aug-2009 16:49:38] <rmatte> I was just too busy today to really be in here too much
[27-Aug-2009 16:49:42] <djack> rmatte how late tomorrow ?
[27-Aug-2009 16:49:47] <djack> or can we mail ?
[27-Aug-2009 16:49:52] <rmatte> well, what time is it there right now?
[27-Aug-2009 16:50:14] <djack> in belgium 8 minutes before 12
[27-Aug-2009 16:50:17] <rmatte> yikes
[27-Aug-2009 16:50:21] <djack> lol
[27-Aug-2009 16:50:26] <rmatte> 6 hours difference then
[27-Aug-2009 16:50:32] <djack> you ?
[27-Aug-2009 16:50:37] <djack> states i guess
[27-Aug-2009 16:50:40] <rmatte> which means I will be in here at about 4pm your time
[27-Aug-2009 16:50:44] <rmatte> nah, canada
[27-Aug-2009 16:50:52] <djack> lol
[27-Aug-2009 16:51:06] <djack> btw the ping it self is done in the ascping.py ?
[27-Aug-2009 16:51:18] <rmatte> most of the people in here are EST or CST, so you're at a disadvantage when it comes to time for support lol
[27-Aug-2009 16:51:44] <djack> well no problem i'm a night kind of person
[27-Aug-2009 16:52:34] <rmatte> $ZENHOME/Products/ZenStatus/AsyncPing.py
[27-Aug-2009 16:52:39] <rmatte> that's the file doing the ping
[27-Aug-2009 16:52:46] <rmatte> apparently
[27-Aug-2009 16:52:55] <djack> well i'm going to look in the construction of the payload...
[27-Aug-2009 16:53:03] <djack> i found python code that works :-p
[27-Aug-2009 16:53:06] <rmatte> hmmm, this is interesting
[27-Aug-2009 16:53:09] <rmatte> http://forums.zenoss.com/viewtopic.php?t=99&sid=e0fa6b22b54b139ca0a3e2145c01d13b
[27-Aug-2009 16:53:22] <rmatte> Quoting the forum:
[27-Aug-2009 16:53:23] <rmatte> http://forums.zenoss.com/viewtopic.php?t=99&sid=e0fa6b22b54b139ca0a3e2145c01d13b
[27-Aug-2009 16:53:24] <rmatte> oops
[27-Aug-2009 16:53:28] <rmatte> http://forums.zenoss.com/viewtopic.php?t=99&sid=e0fa6b22b54b139ca0a3e2145c01d13b
[27-Aug-2009 16:53:30] <rmatte> bah
[27-Aug-2009 16:53:33] <rmatte> why is it not copying
[27-Aug-2009 16:53:45] <djack> what?
[27-Aug-2009 16:53:55] <rmatte> If you change:
[27-Aug-2009 16:53:55] <rmatte> icmppkt = icmp.Packet(ipreply.data)
[27-Aug-2009 16:53:55] <rmatte> to:
[27-Aug-2009 16:53:55] <rmatte> icmppkt = icmp.Packet(ipreply.data, 0)
[27-Aug-2009 16:53:55] <rmatte> In $ZENHOME/Products/ZenStatus/AsyncPing.py, then it won't fail the
[27-Aug-2009 16:53:56] <rmatte> checksum and log the error, but you will still get "unexpected" log
[27-Aug-2009 16:53:58] <rmatte> messages.
[27-Aug-2009 16:54:00] <rmatte> that
[27-Aug-2009 16:54:02] <rmatte> lol
[27-Aug-2009 16:54:04] <rmatte> that's what the posting says
[27-Aug-2009 16:54:10] <rmatte> not sure if that's already changed or not
[27-Aug-2009 16:54:14] <rmatte> but if it's not, it's worth a shot
[27-Aug-2009 16:54:27] <rmatte> and if it works after changing it I'll bring that to Matt's attention tomorrow
[27-Aug-2009 16:54:28] <djack> let's check
[27-Aug-2009 16:55:28] <djack> restarting zenoss
[27-Aug-2009 16:55:28] <rmatte> ah, it looks like it's already there
[27-Aug-2009 16:55:57] <rmatte> plog.debug("checksum failure on packet %r", ipreply.data)
[27-Aug-2009 16:55:57] <rmatte> try:
[27-Aug-2009 16:55:57] <rmatte> icmppkt = icmp.disassemble(ipreply.data, 0)
[27-Aug-2009 16:56:07] <rmatte> oh wait
[27-Aug-2009 16:56:09] <rmatte> that's different
[27-Aug-2009 16:56:49] <rmatte> I don't see anything like either line in the file
[27-Aug-2009 16:57:31] <rmatte> I think that...
[27-Aug-2009 16:57:32] <rmatte> ipreply = ip.disassemble(data)
[27-Aug-2009 16:57:32] <rmatte> try:
[27-Aug-2009 16:57:32] <rmatte> icmppkt = icmp.disassemble(ipreply.data)
[27-Aug-2009 16:57:32] <rmatte> except ValueError:
[27-Aug-2009 16:57:33] <rmatte> plog.debug("checksum failure on packet %r", ipreply.data)
[27-Aug-2009 16:57:35] <rmatte> try:
[27-Aug-2009 16:57:37] <rmatte> icmppkt = icmp.disassemble(ipreply.data, 0)
[27-Aug-2009 16:57:39] <rmatte> does all that
[27-Aug-2009 16:57:43] <rmatte> so doesn't look like it'll make a difference
[27-Aug-2009 16:58:42] <djack> :-(
[27-Aug-2009 17:04:49] <djack> i tricked zenping to believe the host is up
[27-Aug-2009 17:05:51] <rmatte> that's only a temporary solution
[27-Aug-2009 17:05:52] <djack> but no graphing then also
[27-Aug-2009 17:05:58] <djack> idd
[27-Aug-2009 17:06:05] <rmatte> you know, you could just disable ping monitoring all together in zProperties for now
[27-Aug-2009 17:06:21] <rmatte> then use SNMP Agent Down events as indication that the device is down
[27-Aug-2009 17:06:32] <djack> do the snmp stuff works then
[27-Aug-2009 17:06:32] <rmatte> just for the time being until we can figure out what's causing that
[27-Aug-2009 17:06:41] <rmatte> it should work, yeh
[27-Aug-2009 17:06:50] <djack> where do i change it ?
[27-Aug-2009 17:07:01] <rmatte> you know where zProperties is right?
[27-Aug-2009 17:07:04] <rmatte> go in to that for each device
[27-Aug-2009 17:07:21] <rmatte> Change zPingMonitorIgnore from False to True
[27-Aug-2009 17:07:25] <rmatte> and it'll ignore ping monitoring
[27-Aug-2009 17:07:32] <rmatte> then it should hopefully start collecting SNMP data
[27-Aug-2009 17:08:01] <rmatte> (I would have suggested this earlier but I was trying to possibly fix the actual ping issue, since it's nice to have ping monitoring)
[27-Aug-2009 17:08:36] <djack> can the snmp be forced?
[27-Aug-2009 17:08:37] <rmatte> g'day Chad
[27-Aug-2009 17:08:46] <rmatte> forced?
[27-Aug-2009 17:08:59] <rmatte> When you disable the ping monitoring Zenoss will attempt to monitor via SNMP no matter what
[27-Aug-2009 17:09:24] <rmatte> the only reason it hasn't been is because when something is showing as down via ping Zenoss stops all monitoring for that device
[27-Aug-2009 17:09:33] <rmatte> because there's no point in monitoring an unreachable device
[27-Aug-2009 17:09:48] <rmatte> but if you tell it not to do the ping monitoring, then it'll monitor with snmp
[27-Aug-2009 17:10:04] <djack> lets'hope
[27-Aug-2009 17:10:24] <rmatte> but this is really more of a temporary solution, since the ping monitoring is pretty important to have
[27-Aug-2009 17:10:54] <djack> all graphs show nan
[27-Aug-2009 17:10:56] <djack> hmmm
[27-Aug-2009 17:11:32] <rmatte> they'll be nan for 3 polling cycles
[27-Aug-2009 17:11:37] <rmatte> (about 15 to 20 mins)
[27-Aug-2009 17:11:41] <rmatte> then they should start populating
[27-Aug-2009 17:11:43] <rmatte> that's normal
[27-Aug-2009 17:11:46] <djack> oki
[27-Aug-2009 17:12:49] <djack> the zpersnmp.logs shows that a success on 16 devices this is beter then 9
[27-Aug-2009 17:13:09] <rmatte> yup
[27-Aug-2009 17:13:13] <rmatte> I'm sure it's going to work
[27-Aug-2009 17:13:45] <rmatte> you just won't get "ip blah is down" when the device goes down, you'll see "SNMP agent down" instead
[27-Aug-2009 17:13:51] <djack> well tommorow or in my timezone today i will come back to maybe help to troubleshout so m more
[27-Aug-2009 17:13:59] <rmatte> cool
[27-Aug-2009 17:14:30] <djack> well i can learn from it also ;-)
[27-Aug-2009 17:14:36] <rmatte> yup
[27-Aug-2009 17:15:01] <rmatte> well, I'm going to head out of here
[27-Aug-2009 17:15:06] <rmatte> I'll talk to you later
[27-Aug-2009 17:15:20] <djack> oki have a nice evening and thanks
[27-Aug-2009 17:15:27] <rmatte> you too
[27-Aug-2009 17:15:29] <rmatte> no problem
[27-Aug-2009 17:16:06] <djack> oooh i have values ;-)
[27-Aug-2009 17:16:08] <djack> wiiiiiiiiiiiiiiiiiiiiii
[27-Aug-2009 17:16:11] <rmatte> good stuff
[27-Aug-2009 17:28:44] <ironpaw> hi
[27-Aug-2009 17:38:23] <djack> hi
[27-Aug-2009 17:42:43] <djack> exit
[27-Aug-2009 17:50:26] <ironpaw> aww
[27-Aug-2009 17:50:30] <ironpaw> no love
[27-Aug-2009 23:04:11] <davetoo> hmm
[27-Aug-2009 23:06:05] <davetoo> anybody awake here? I just discovered that deleting a Network breaks all the Interface objects that had IPs in that network.
[28-Aug-2009 00:00:46] [disconnected at Fri Aug 28 00:00:46 2009]
[28-Aug-2009 00:00:46] [connected at Fri Aug 28 00:00:46 2009]
[28-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[28-Aug-2009 04:55:25] <Dieterbe> anyone here using zenperfwmi? i get syntax errors in the log file (and events for it) but i don't know for which queries. i've set debug=True in its config and reloaded the daemon, but i still don't see the faulty queries in the logfile
[28-Aug-2009 08:55:12] <rmatte> good morning Matt
[28-Aug-2009 08:55:29] <mrayzenoss> morning
[28-Aug-2009 08:56:09] <rmatte> no other blockers were found overnight I hope
[28-Aug-2009 08:56:10] <rmatte> lol
[28-Aug-2009 09:15:07] <mrayzenoss> They're taking a hard look at http://dev.zenoss.com/trac/ticket/5469
[28-Aug-2009 09:44:38] <rmatte> yeh, I was looking at that
[28-Aug-2009 09:44:43] <rmatte> I saw Jane's post
[28-Aug-2009 09:45:02] <rmatte> that would be a nice one to fix as well hehe
[28-Aug-2009 09:45:16] <mrayzenoss> I think we're going to fix it too
[28-Aug-2009 09:45:22] <rmatte> cool
[28-Aug-2009 09:45:29] <rmatte> I look forward to the patch
[28-Aug-2009 09:45:31] <rmatte>
[28-Aug-2009 09:45:56] <rmatte> I don't have any command datapoints setup that way right now, so it's not as big of a deal for me as the other patch was
[28-Aug-2009 10:13:22] <dollarbang> is there a specific requirement for Zenoss to run Zenoss secure 'https'? Its on the intranet, so its relately secure anyway. Reason I ask is that I think that running Zenoss 'https' instead of 'http' causes everything to run more encrypted, could the encryption be affecting the maps? I've also noticed that every now and then I get an error "Error - Server connection error"... Can this be caused by the 'https' encryption?
[28-Aug-2009 10:21:26] <rmatte> dollarbang: how are you running it as https? Just via a proxy?
[28-Aug-2009 10:26:55] <dollarbang> rmatte: I knew if I asked a question, it would be over my head immediately. One of our Network OPS guys thinks that running Zenoss Core 2.4.2 is causing issues with Google Maps. He has his own Zenoss server 2.3.3, his entire device list was dumped and loaded onto another Zenoss Core 2.4.2 server (shutting down his server in the process to keep collisions down). Except for the different rev levels, the 2.4.2 version is using HTTPS for access to the Zenos
[28-Aug-2009 10:28:59] <rmatte> well, do you have the right API key for google maps?
[28-Aug-2009 10:29:10] <rmatte> when you sign up for the key you might have to include https:// as the prefix
[28-Aug-2009 10:29:32] <rmatte> Go to Settings, the API key input field is right at the bottom
[28-Aug-2009 10:29:40] <rmatte> with a link next to it for instructions on getting a key
[28-Aug-2009 10:32:58] <dollarbang> there is a key there, I'll check on the Google web site.
[28-Aug-2009 10:34:03] <rmatte> k
[28-Aug-2009 10:34:29] <rmatte> I'm thinking that they key might have been for http:// prefix, so when you switched to https:// it started not liking the key
[28-Aug-2009 10:35:02] <rmatte> also, the server connection errors just indicate that you've temporarily lost contact with the server, which could be for a number of reasons
[28-Aug-2009 10:35:15] <rmatte> If you refresh the page does it come back right away when that happens?
[28-Aug-2009 10:37:15] <dollarbang> rmatte: need to break for lunch, when I return, I'll gen another key.
[28-Aug-2009 10:38:04] <rmatte> k
[28-Aug-2009 10:49:27] <rmatte> mrayzenoss: have you seen http://forums.zenoss.com/viewtopic.php?t=10830 yet?
[28-Aug-2009 10:57:23] <jb> ok
[28-Aug-2009 10:57:31] <jb> time to fix these problems
[28-Aug-2009 10:57:47] <jb> cgibbons: more zendmd issues started popping up again :/
[28-Aug-2009 10:58:01] <jb> cgibbons: workign with kells in just a sec to get them fixed though
[28-Aug-2009 10:58:02] <cgibbons> oh?
[28-Aug-2009 10:58:16] <jb> yeah.. cache keeps getting corrupted, zendmd errors, etc
[28-Aug-2009 10:59:39] <mrayzenoss1> ugh, 2.4.5 Monday
[28-Aug-2009 11:00:02] <jb> really?
[28-Aug-2009 11:00:17] <mrayzenoss1> jb: don't worry, Enterprise people won't get it until Wednesday
[28-Aug-2009 11:00:24] <rmatte> lol
[28-Aug-2009 11:00:27] <jb> im still 2.4.2 anyways
[28-Aug-2009 11:00:55] <rmatte> 2.4.2 is fine as long as you trust all of your standard users logging in to Zenoss
[28-Aug-2009 11:00:58] <mrayzenoss1> http://dev.zenoss.com/trac/ticket/5469 is the ticket that caused 2.4.5
[28-Aug-2009 11:01:00] <jb> yeah
[28-Aug-2009 11:01:15] <rmatte> ah
[28-Aug-2009 11:01:24] <cgibbons> 2.4.5?
[28-Aug-2009 11:01:26] <cgibbons> not 2.4.4?
[28-Aug-2009 11:01:35] <jb> 2.4.4 fixed the event clear emails
[28-Aug-2009 11:01:52] <jb> 2.4.5 fixes the severity
[28-Aug-2009 11:01:57] <jb> (i think)
[28-Aug-2009 11:04:18] <mrayzenoss1> it's 2.4.5 because 2.4.4 was publicly available via apt, yum and appliance upgrades for probably 2 hours
[28-Aug-2009 11:04:27] <cgibbons> ahhh snap
[28-Aug-2009 11:08:08] <rocket> bah ..thats not good
[28-Aug-2009 11:14:08] mrayzenoss1 is now known as mrayzenoss
[28-Aug-2009 11:18:26] <etank> heh. we are doing our enterprise install today
[28-Aug-2009 11:18:56] <etank> guess we will do an upgrade on Wednesday then
[28-Aug-2009 11:31:52] <rmatte> http://dev.zenoss.org/trac/ticket/5473
[28-Aug-2009 11:32:12] <rmatte> can someone else who is running 2.4.3 test that to confirm please?
[28-Aug-2009 11:32:25] <rmatte> I'm seeing that behaviour across all 5 of my production Zenoss boxes
[28-Aug-2009 11:41:52] <schwartz> sya morning
[28-Aug-2009 12:52:18] <dollarbang> rmatte: I'm back, you asked about the API key. It was generated according to the Zenoss doc and with https:
[28-Aug-2009 12:57:50] <davetoo> uh oh
[28-Aug-2009 12:57:53] <davetoo> patch to the patch?
[28-Aug-2009 13:20:51] <rmatte> davetoo: no, just another patch
[28-Aug-2009 13:30:32] <djack> hellow world
[28-Aug-2009 13:31:53] <djack> small question is there a way to monitor the load per core via snmp ?
[28-Aug-2009 13:37:30] <cgibbons> yay community
[28-Aug-2009 13:38:45] <mrayzenoss> cgibbons: what's that?
[28-Aug-2009 13:39:00] <cgibbons> yellow hat from mrchippy
[28-Aug-2009 13:39:02] <mrayzenoss> djack: haven't seen per core, only per processor
[28-Aug-2009 13:39:12] <mrayzenoss> yay community!
[28-Aug-2009 13:39:35] <rocket>
[28-Aug-2009 13:40:23] <rmatte> djack: for what kind of server?
[28-Aug-2009 13:41:45] <rmatte> windows, linux?
[28-Aug-2009 13:45:48] * rocket starts grumbling ... how am I supposed to get work done if I am tired....
[28-Aug-2009 13:46:22] <rmatte> ...how am I supposed to get work done when my enthusiasm is sitting at the bar right now?...
[28-Aug-2009 13:46:39] <rmatte> ...and I'm tempted to follow it...
[28-Aug-2009 13:47:23] <cgibbons> you see the 15 minute caffeine powernap article, rocket?
[28-Aug-2009 13:48:08] <rocket> cgibbons: no ... tell me more ..
[28-Aug-2009 13:48:25] <cgibbons> yeah: http://lifehacker.com/software/sleep/reboot-your-brain-with-a-caffeine-nap-306029.php
[28-Aug-2009 13:48:35] <rocket> cgibbons: I was repairing a oracle/sap db server yesterday from 8am-3am ...
[28-Aug-2009 13:48:41] <cgibbons> ouf
[28-Aug-2009 13:51:01] <rocket> yeah .. sleep deprivation is making it hard to do the tedious hacking on zenoss ...
[28-Aug-2009 13:51:22] <rocket> eg the copyright stuff .. the comments etc ..
[28-Aug-2009 13:53:08] <cgibbons> i really miss the powerful IDEs the Java world had for that stuff, I rarely had to do the tedious broilerplate work... the python tools just aren't really that good (yet?)
[28-Aug-2009 13:53:32] * rocket is using vim and havent made the macros yet to automate ... :/
[28-Aug-2009 13:54:07] <cgibbons> yeah I used vim only for a long-time here but eventually switched to Eclipse + pydev once I got it all setup right.
[28-Aug-2009 13:54:14] <jb> i need to checkout eclipse
[28-Aug-2009 13:54:42] <rocket> cgibbons: sometime I will have to have you show me the setup
[28-Aug-2009 13:54:45] <jb> ive been using (eek, i know) nano+python template
[28-Aug-2009 13:54:56] <cgibbons> sure, i've got a partial write on how to do it, i'll finish tha tup
[28-Aug-2009 13:55:02] <jb> ah nice.
[28-Aug-2009 13:55:06] <cgibbons> mrchippy wrote up one for us when we first started here, but it's dated now
[28-Aug-2009 13:55:17] <jb> ugh, i still need to figure out this IO thing
[28-Aug-2009 13:55:21] <jb> i'll do it next week.
[28-Aug-2009 13:55:30] <rocket> cgibbons: I need to work on the installer too so I can have a native gentoo install ...
[28-Aug-2009 13:55:54] <rocket> right now I am having to do everything in a separate vm .. so I have to replicate config files etc across ...
[28-Aug-2009 13:56:14] <davetoo> jb?
[28-Aug-2009 13:56:16] <davetoo> jan?
[28-Aug-2009 13:56:22] <cgibbons> you just don't use a source build on your gentoo box?
[28-Aug-2009 13:56:34] <mrayzenoss> rocket: installer? Like a Gentoo build?
[28-Aug-2009 13:57:10] <davetoo> mrayzenoss: are you in the Annapolis office?
[28-Aug-2009 13:57:22] <mrayzenoss> I flew back to Austin last night
[28-Aug-2009 13:57:25] <rocket> mrayzenoss: yes
[28-Aug-2009 13:57:26] <davetoo> oh,
[28-Aug-2009 13:57:34] <davetoo> I was in Annapolis W-Fri last week
[28-Aug-2009 13:57:38] <davetoo> crazy weather
[28-Aug-2009 13:58:00] <cgibbons> man i haven't been to annapolis since the summer party we had there last july, crazy
[28-Aug-2009 13:58:07] <rocket> cgibbons: I cant remember my issue but the source installer borked on my gentoo machine
[28-Aug-2009 13:58:15] <cgibbons> huh
[28-Aug-2009 13:58:27] <jb> heh.. have you guys eaten at Lemongrass?
[28-Aug-2009 13:58:34] <jb> in annapolis, right down the street from zenoss..
[28-Aug-2009 13:58:39] <davetoo> I flew out Saturday morning, but one of the guys was supposed to leave Friday night.. and everthing from DC to Maine was closed for weather
[28-Aug-2009 13:58:51] <davetoo> I just "ate" at Fado'
[28-Aug-2009 13:58:53] <rocket> cgibbons: I believe it was due to library versions installed etc .. something wasnt compiling...
[28-Aug-2009 13:59:07] <mrayzenoss> davetoo: at at Carpaccio's, next to Fado
[28-Aug-2009 13:59:17] <davetoo> we got some slices there
[28-Aug-2009 13:59:27] <jb> carpaccios is good
[28-Aug-2009 13:59:30] <jb> so is fados
[28-Aug-2009 13:59:34] <jb> but lemongrass > *
[28-Aug-2009 13:59:38] <cgibbons> annapolis is a bit too cold for me, brrrr
[28-Aug-2009 13:59:50] <jb> best mojitos ever
[28-Aug-2009 14:03:06] <djack> rmatte linux servers
[28-Aug-2009 14:03:15] <rmatte> cgibbons: come to Ottawa, Canada, then you'll have a new appreciation for the weather in Annapolis
[28-Aug-2009 14:03:22] <jb> or chicago
[28-Aug-2009 14:03:26] <rmatte> djack: ah
[28-Aug-2009 14:03:29] <djack> nah belgium :-p
[28-Aug-2009 14:03:41] kevin7kal_ is now known as kevin7kal
[28-Aug-2009 14:03:43] <rmatte> djack: you could do it via ssh with a script
[28-Aug-2009 14:03:49] <cgibbons> oh, i know... i actually grew up in milwaukee, brr... and i got to spend quite a bit of time in Calgary in '07 @ a customer when i worked at my previous place. brr.
[28-Aug-2009 14:03:50] <rocket> heh Minnesota can be quite odd from a weather perspective
[28-Aug-2009 14:04:26] <rocket> 90's one day ... 60's the next
[28-Aug-2009 14:04:27] <djack> hmmm nagios has a script for that see if i can get it to work
[28-Aug-2009 14:05:21] <djack> rmatte already some ideas about the zenping issue ?
[28-Aug-2009 14:09:34] <rmatte> djack: I haven't had time to look at that at all today
[28-Aug-2009 14:09:37] <rmatte> way too busy
[28-Aug-2009 14:10:28] <rmatte> I'm getting sent love notes from some of our engineers about the latest Zenoss update
[28-Aug-2009 14:10:46] <rmatte> </sarcasm>
[28-Aug-2009 14:11:04] <rmatte> although 14744 fixed the major issue, so I'm happy enough
[28-Aug-2009 14:14:00] <djack> rmatte i don't follow the the sarcasm but no mather
[28-Aug-2009 14:15:00] <rmatte> the sarcasm was emphasized on the word "love"
[28-Aug-2009 14:15:16] <rmatte> lol
[28-Aug-2009 14:16:07] <djack> i don't know what kind or relation you have with uour engineers he ;-p
[28-Aug-2009 14:16:57] <davetoo> I probably shouldn't have been so quick to make a vm template from 2.4.3
[28-Aug-2009 14:17:29] <rmatte> djack: generally a good one, it's just been a rough day
[28-Aug-2009 14:17:42] <rmatte> davetoo: why?
[28-Aug-2009 14:18:01] <rmatte> davetoo: just apply the patches and your template will be fine
[28-Aug-2009 14:19:16] <davetoo> yeah, just have to convert out and back in
[28-Aug-2009 14:19:28] <davetoo> it's just for development anyway
[28-Aug-2009 14:19:34] <rmatte> yeh
[28-Aug-2009 14:20:02] <davetoo> I'm finding some interesting "features" in 2.4.x
[28-Aug-2009 14:20:14] <davetoo> doesn't anybody else use the ZMI to poke around/debug?
[28-Aug-2009 14:21:00] <mrayzenoss> be back soon
[28-Aug-2009 14:21:18] <rmatte> nah, my zenoss jacket is here but shipping department is closed
[28-Aug-2009 14:21:27] <rmatte> and it's actually chilly out today, I could use the jacket
[28-Aug-2009 14:21:28] <rmatte>
[28-Aug-2009 14:21:33] <rmatte> nah = bah
[28-Aug-2009 14:22:35] <davetoo> rmatte: where are you?
[28-Aug-2009 14:22:49] <djack> rmatte i had a nice day showing off the zenoss to the bosses they where impressed
[28-Aug-2009 14:22:53] <rocket> rmatte: jacket is nice ..
[28-Aug-2009 14:23:03] <rmatte> djack: good to hear
[28-Aug-2009 14:23:14] <rmatte> rocket: cool lol
[28-Aug-2009 14:23:24] <djack> rmatte can i submit a bug for the zenping ?
[28-Aug-2009 14:23:36] <rmatte> djack: do you have a trac account yet?
[28-Aug-2009 14:23:43] <djack> not yet
[28-Aug-2009 14:23:43] <davetoo> Cisco NX-OS is driving me bonkers
[28-Aug-2009 14:24:00] <rmatte> let me check if there's already a backlogged bug for it
[28-Aug-2009 14:24:16] <djack> rmatte if so kick it to the top :-)
[28-Aug-2009 14:24:32] <davetoo> rmatte: you zenoss crew?
[28-Aug-2009 14:24:46] <rmatte> http://dev.zenoss.com/trac/ticket/4886
[28-Aug-2009 14:24:48] <rmatte> there you go
[28-Aug-2009 14:25:07] <rmatte> sort of
[28-Aug-2009 14:25:09] <rmatte> similar issue
[28-Aug-2009 14:25:25] <rmatte> actually no it's not
[28-Aug-2009 14:25:30] <djack> lol
[28-Aug-2009 14:25:30] <rmatte> the title of the bug report just sucks
[28-Aug-2009 14:25:31] <rmatte> lol
[28-Aug-2009 14:26:56] <rmatte> djack: if you write up a detailed bug report I'll file it for you
[28-Aug-2009 14:27:05] <rmatte> send it to rmatte@novanetworks.com
[28-Aug-2009 14:27:28] <djack> well i see what i can provide :-)
[28-Aug-2009 14:27:29] <rmatte> also, you may want to ask mrayzenoss to make a Trac account for you so you can file bugs yourself
[28-Aug-2009 14:27:58] <davetoo> What is the purpose of the "Version" field? Is it supposed to be the version the bug is in? Or the version targeted for the fix?
[28-Aug-2009 14:28:11] <rmatte> but I'd do that in a couple weeks since he's ridiculously busy
[28-Aug-2009 14:28:22] <rmatte> davetoo: yeh
[28-Aug-2009 14:28:37] <rmatte> davetoo: if it's present in all versions then put ALL, if it appeared in 2.4.3 then put 2.4.3
[28-Aug-2009 14:28:55] <rmatte> and platform is generally "ALL"
[28-Aug-2009 14:29:01] <djack> rmatte what do you call detailed ?
[28-Aug-2009 14:29:12] <davetoo> 'k, 'cause there's freeform text in the body for the build/version
[28-Aug-2009 14:29:18] <rmatte> djack: a clear explanation of the issue you are experience with log lines included
[28-Aug-2009 14:29:22] <davetoo> where I'm generally putting my rpm package name
[28-Aug-2009 14:29:28] <rmatte> the issue your are experienCING
[28-Aug-2009 14:29:29] <rmatte> rather
[28-Aug-2009 14:30:00] <rmatte> davetoo: ah
[28-Aug-2009 14:30:11] <rmatte> davetoo: well, it's really up to you, I try to make the bug reports more generalize
[28-Aug-2009 14:30:14] <rmatte> generalized*
[28-Aug-2009 14:30:24] <davetoo> I've only done two with this new format
[28-Aug-2009 14:30:32] <davetoo> been a few months, before that.
[28-Aug-2009 14:30:42] <djack> rmatte i see if i can even add the firewall logs :-)
[28-Aug-2009 14:30:48] <davetoo> What's the python idiom/equivalent for the perl diamond operator?
[28-Aug-2009 14:31:22] <rmatte> djack: that'd help
[28-Aug-2009 14:31:39] <davetoo> stupid Cisco NX-OS doesn't have working snmp for the IP interfaces/routes yet.
[28-Aug-2009 14:31:56] <djack> btw i had something strange thing today... i added the apachemonitor template to one of the webservers when i went and look to the perfdata all the regular perf of the cpu and memory and so was f*cked
[28-Aug-2009 14:32:08] <rmatte> davetoo: http://xorlogic.blogspot.com/2007/12/perls-diamond-operator-in-python.html
[28-Aug-2009 14:32:57] <davetoo> thanks
[28-Aug-2009 14:33:01] <rmatte> np
[28-Aug-2009 14:33:11] <davetoo> I still don't know the right keywords to search yet
[28-Aug-2009 14:33:16] <rmatte> hehe
[28-Aug-2009 14:33:18] <davetoo> Know a lot of perl, not so much python yet
[28-Aug-2009 14:33:27] <davetoo> you say 'dict', I say 'hash'
[28-Aug-2009 14:33:42] <rmatte> hehe
[28-Aug-2009 14:34:01] <rmatte> I'm still learning python myself, and I don't even know perl that well
[28-Aug-2009 14:34:06] <rmatte> but I'm a god with bash scripting
[28-Aug-2009 14:34:08] <rmatte>
[28-Aug-2009 14:34:25] <davetoo> I hurt my brain on bash scripting
[28-Aug-2009 14:34:30] <davetoo> arrays and such
[28-Aug-2009 14:34:35] <rmatte> I've been using unix type operating systems since I was 14 or 15
[28-Aug-2009 14:34:38] <rmatte> so it's grown on me
[28-Aug-2009 14:34:54] <davetoo> when I started using unix, there was no bash
[28-Aug-2009 14:35:02] <davetoo> but I didn't keep it up.
[28-Aug-2009 14:35:03] <rmatte> yeh, but I'm only 24
[28-Aug-2009 14:35:07] <rmatte> so I never had the chance
[28-Aug-2009 14:35:08] <rmatte>
[28-Aug-2009 14:35:25] <davetoo> so the fileinput module seems to be what I need
[28-Aug-2009 14:35:26] <rmatte> I'm the only 24 year I know who can boast 8-9 years unix experience
[28-Aug-2009 14:35:32] <davetoo> heh
[28-Aug-2009 14:35:38] <davetoo> I know a few
[28-Aug-2009 14:35:46] <rmatte> it's fairly rare though
[28-Aug-2009 14:35:53] <davetoo> yes
[28-Aug-2009 14:35:59] <rmatte> I started out with FreeBSD for my first 5 years
[28-Aug-2009 14:36:02] <rmatte> then moved on to Linux
[28-Aug-2009 14:36:10] <rmatte> I think it's coming up on almost 5 years of Linux
[28-Aug-2009 14:36:18] <davetoo> I started with Slackware 1.14 I think....
[28-Aug-2009 14:36:26] <davetoo> I'm too old to remember
[28-Aug-2009 14:36:35] <davetoo> but I mostly use FreeBSD now
[28-Aug-2009 14:36:42] <rmatte> I started with FreeBSD, then moved to Gentoo Linux (before it even had an installer, did the entire installation by hand with commands)
[28-Aug-2009 14:36:50] <rmatte> now I'm using Ubuntu since I'm sick of the hassle
[28-Aug-2009 14:36:57] <rmatte> I still have Gentoo on my home server though
[28-Aug-2009 14:36:58] <djack> 11 years of linux :-) debian lover still am but because not commercial supported it's suse currently
[28-Aug-2009 14:37:16] <rmatte> I prefer Ubuntu over Suse
[28-Aug-2009 14:37:18] <davetoo> My first Linux install was done by purchasing a case of 100 1.4M floppies and filling those at the ISP where I worked,
[28-Aug-2009 14:37:23] <davetoo> then carting them home to do the install
[28-Aug-2009 14:37:45] <djack> rmatte we made a choice at the office and then there was not yet commercial supported ubuntu
[28-Aug-2009 14:37:47] <rmatte> brutality
[28-Aug-2009 14:37:52] <cwj> nostalgia mode engage
[28-Aug-2009 14:38:01] <rmatte> Ubuntu is commercially supported though
[28-Aug-2009 14:38:01] <davetoo> The other thing I miss about perl is the easy interactive debugger
[28-Aug-2009 14:38:07] <rmatte> there's a company that does Ubuntu support
[28-Aug-2009 14:38:08] <davetoo> or maybe I haven't read enough yet
[28-Aug-2009 14:38:09] <djack> i work for a larg retailer so i need to make calls to a servicedesk when needed
[28-Aug-2009 14:38:25] <rmatte> that's why Dell uses Ubuntu on their systems when people want Linux instead of Windows
[28-Aug-2009 14:38:28] <cwj> canonical ltd offers commerical ubuntu support
[28-Aug-2009 14:38:37] <rmatte> yeh, canonical
[28-Aug-2009 14:38:38] <djack> rmatte it is but 7 years ago ....
[28-Aug-2009 14:38:39] <rmatte> that's the one
[28-Aug-2009 14:38:44] <rmatte> ah
[28-Aug-2009 14:38:50] <djack> ubunto is debain based ;-)
[28-Aug-2009 14:39:15] <rmatte> davetoo: debugger: http://docs.python.org/library/pdb.html
[28-Aug-2009 14:39:25] <rmatte> djack: yeh, I know lol
[28-Aug-2009 14:39:28] <cwj> i am just starting to get traction on replacing all our SuSE VMs with Ubuntu VMs
[28-Aug-2009 14:39:35] <cwj> what a glorious day that will be
[28-Aug-2009 14:40:06] <rmatte> yeh
[28-Aug-2009 14:40:07] <davetoo> rmatte: I haven't figured out if/how to fire it up, load the module(s) without running them, and set breakpoints before running
[28-Aug-2009 14:40:22] <rmatte> we're using Ubuntu Server 8.04 on our VMs and it has been 100% smooth sailing
[28-Aug-2009 14:40:25] <rmatte> no issues at all
[28-Aug-2009 14:40:37] <djack> well we have to migrate our point of sales soon so i will have a choice again , when the cash application can run on linux
[28-Aug-2009 14:40:46] <rmatte> davetoo: that page sort of explains that
[28-Aug-2009 14:40:53] <davetoo> 'k
[28-Aug-2009 14:40:57] <cwj> i have been working on a script to deploy new Ubuntu 8.04 VMs using preseed and our VMware infrastructure
[28-Aug-2009 14:41:04] <cwj> almost works
[28-Aug-2009 14:41:40] <rmatte> b(reak) [[filename:]lineno | function[, condition]]
[28-Aug-2009 14:41:42] <rmatte> With a lineno argument, set a break there in the current file.....
[28-Aug-2009 14:41:53] <rmatte> that's what you're looking for
[28-Aug-2009 14:42:05] <davetoo> probably
[28-Aug-2009 14:43:54] <djack> btw i have problems (time-out) when loading the zenpack.hp.mibs and zenpack.cisco.mibs ?
[28-Aug-2009 14:44:13] <djack> i tried to do it via the web interface and from the command line it self
[28-Aug-2009 14:44:21] <rmatte> djack: you're trying to install from the web UI aren't you?
[28-Aug-2009 14:44:41] <rmatte> the cisco mibs will install fine from the commandline, but not from the UI
[28-Aug-2009 14:44:47] <rmatte> and they take about 5-6 hours to install
[28-Aug-2009 14:44:50] <davetoo> sheesh
[28-Aug-2009 14:44:55] <rmatte> so you're probably just not being patient enough
[28-Aug-2009 14:45:13] <rmatte> are you familliar with the 'screen' command?
[28-Aug-2009 14:45:14] <djack> rmatte what ....
[28-Aug-2009 14:45:20] <djack> yeap
[28-Aug-2009 14:45:33] <rmatte> run the install inside a screen, so if your ssh connection dies it'll still install
[28-Aug-2009 14:45:36] <davetoo> I just did an snmptranslate of all the cisco mibs that NX-OS uses. Over 50k oids
[28-Aug-2009 14:45:39] <rmatte> your ssh connection is probably timing out
[28-Aug-2009 14:45:49] <rmatte> I remember having to run it inside a screen for it to finish
[28-Aug-2009 14:46:04] <davetoo> I do almost nothing *outside* of a screen session
[28-Aug-2009 14:46:17] <davetoo> screen is my emacs
[28-Aug-2009 14:46:19] <djack> rmatte i will run it from a script
[28-Aug-2009 14:46:21] <rmatte> so the actual install inside a screen session is timing out?
[28-Aug-2009 14:46:25] <rmatte> I've never ever seen that before
[28-Aug-2009 14:46:54] <davetoo> rmatte: I think you confused my comment for a response from djack
[28-Aug-2009 14:46:56] <rmatte> djack: install "screen"
[28-Aug-2009 14:46:57] <djack> did forget to install screen and pfff access from my sofa to the server in Holland is difficult
[28-Aug-2009 14:47:04] <rmatte> and use a screen session to install them
[28-Aug-2009 14:47:21] <rmatte> and it'll take like 5-6 hours for the cisco pack
[28-Aug-2009 14:47:27] <rmatte> no idea how long for the HP pack
[28-Aug-2009 14:47:37] <rmatte> also, once they are installed use the zendmd commands on those zenpack pages
[28-Aug-2009 14:47:43] <rmatte> otherwise your UI performance will suck
[28-Aug-2009 14:47:52] <rmatte> took me forever to figure out that fix
[28-Aug-2009 14:48:30] <djack> rmatte it's a f*cking hp proliant 360 G6 with 2 dual quad core cpu's and 16 gig of memory ...
[28-Aug-2009 14:48:49] <djack> made a error ordering the server hihi
[28-Aug-2009 14:49:22] <djack> rmatte do you mean reindex() and commit() stuff
[28-Aug-2009 14:49:36] <davetoo> how is your storage configured?
[28-Aug-2009 14:49:59] <rmatte> lol
[28-Aug-2009 14:50:10] <rmatte> djack: no
[28-Aug-2009 14:50:14] <djack> 3 sas 15 k disc in Raid 5
[28-Aug-2009 14:50:19] <rmatte> djack: go to the actual pages where you download those zenpacks
[28-Aug-2009 14:50:21] <djack> rmatte please enlight me
[28-Aug-2009 14:50:23] <rmatte> there are zendmd commands listed
[28-Aug-2009 14:50:35] <rmatte> after the packs are installed you have to execute those
[28-Aug-2009 14:50:49] <rmatte> it basically removes the zenpack from the zenpack list after install
[28-Aug-2009 14:50:56] <rmatte> which eliminates the performance issues
[28-Aug-2009 14:52:11] <djack> i see
[28-Aug-2009 14:54:32] <jb> hrm
[28-Aug-2009 14:54:40] <jb> why isnt my evt.message transform appearing in emails?
[28-Aug-2009 14:54:41] <jb> :/
[28-Aug-2009 14:57:54] <jb> hey egor/rocket?
[28-Aug-2009 14:59:15] <jb> ah nevermind!
[28-Aug-2009 14:59:23] <jb> yo rmatte
[28-Aug-2009 14:59:30] <rocket> rmatte: but can you add aix, solaris and hpux to the mix?
[28-Aug-2009 14:59:31] <rocket>
[28-Aug-2009 15:00:21] <rocket> rmatte: heh and I am one of the gentoo devs who wrote genkernel and catalyst ..
[28-Aug-2009 15:01:09] <cwj> does drobbins still work on gentoo? i used to work with him on stampede a long long time ago
[28-Aug-2009 15:01:36] <cwj> i guess not
[28-Aug-2009 15:02:55] <rmatte> jb: sup?
[28-Aug-2009 15:03:09] <jb> do you remember that thread on the forums about the guy asking how to do multi-line transforms?
[28-Aug-2009 15:03:16] <rmatte> rocket: yeh, mrayzenoss told me
[28-Aug-2009 15:03:20] <rmatte> rocket: props for that
[28-Aug-2009 15:03:38] <rmatte> rocket: I actually remember seeing your email address floating around back in the day, just didn't know you were the same rocket
[28-Aug-2009 15:03:46] <rmatte> rocket@gentoo.org
[28-Aug-2009 15:04:00] <rmatte> jb: yeh, I do
[28-Aug-2009 15:04:01] <rocket> same one
[28-Aug-2009 15:04:14] <jb> remember what it was called?
[28-Aug-2009 15:04:18] <jb> or the URL?
[28-Aug-2009 15:04:26] <jb> finally figured out how to do it
[28-Aug-2009 15:05:08] <rmatte> hmmm, I'll see if I can find it
[28-Aug-2009 15:06:07] <jb> yeah i can't seem to locate it
[28-Aug-2009 15:06:48] <jb> http://pastebin.com/m3eb2c056
[28-Aug-2009 15:06:52] <jb> (an email alert)
[28-Aug-2009 15:08:27] <rmatte> jb: what's the code for the transform look like?
[28-Aug-2009 15:08:32] <jb> sec
[28-Aug-2009 15:08:48] <jb> http://pastebin.com/m79cc0e03
[28-Aug-2009 15:09:13] <rmatte> quite honestly, just post the code as a new forum post, and name it "Transform to create event message with multiple lines"
[28-Aug-2009 15:09:25] <rmatte> I'm sure other people will be interested in knowing
[28-Aug-2009 15:09:36] <rmatte> and whoever was originally looking for a solution will probably come across it eventually
[28-Aug-2009 15:09:59] <rmatte> ah, so it was newlines
[28-Aug-2009 15:10:05] <jb> yeah
[28-Aug-2009 15:10:09] <rmatte> I wasn't sure whether that would actually work or not
[28-Aug-2009 15:10:12] <jb> just had it formatted incorrectly
[28-Aug-2009 15:10:16] <rmatte> I remember testing that in zendmd and it worked fine
[28-Aug-2009 15:10:21] <jb> updating my transforms now
[28-Aug-2009 15:10:22] <rmatte> ah
[28-Aug-2009 15:10:42] <rmatte> yeh, you had it formatted on multiple lines
[28-Aug-2009 15:10:46] <rmatte> but that usually gets quite messy
[28-Aug-2009 15:11:07] <rmatte> good stuff though
[28-Aug-2009 15:11:30] <rmatte> by the way, are you running 2.4.3?
[28-Aug-2009 15:11:41] <jb> nope
[28-Aug-2009 15:11:43] <jb> 2.4.2
[28-Aug-2009 15:11:48] <rmatte> ah
[28-Aug-2009 15:12:05] <rmatte> I need someone to test a bug I filed in 2.4.3 to see if they see the same problem: http://dev.zenoss.org/trac/ticket/5473
[28-Aug-2009 15:12:11] <rmatte> I've patched my 2.4.3 obviously
[28-Aug-2009 15:12:17] <rmatte> it actually shows as 2.4.4 right now hehe
[28-Aug-2009 15:13:12] <rocket> rmatte: nooooooooooooooooo
[28-Aug-2009 15:13:18] <rocket> no patching ....
[28-Aug-2009 15:13:18] <rmatte> ?
[28-Aug-2009 15:13:22] <rocket>
[28-Aug-2009 15:13:23] <rmatte> why?
[28-Aug-2009 15:13:27] <rocket> dont do it man ...
[28-Aug-2009 15:13:33] <rocket> just pulling your leg ...
[28-Aug-2009 15:13:36] <rmatte> I figured
[28-Aug-2009 15:13:37] <rmatte>
[28-Aug-2009 15:13:40] <djack> lol
[28-Aug-2009 15:13:49] * rmatte fwaps rocket
[28-Aug-2009 15:14:11] * rocket bends over ... oops I dont think you wanted to fwap that ...
[28-Aug-2009 15:14:30] * rmatte turns away in disgust
[28-Aug-2009 15:14:34] <djack> lol
[28-Aug-2009 15:14:36] <rocket> heh ...
[28-Aug-2009 15:14:37] <rhett> what the hell did i just walk into
[28-Aug-2009 15:14:46] <djack> i brings back memories
[28-Aug-2009 15:14:50] <rmatte> rhett: still trying to figure that out for myself
[28-Aug-2009 15:15:21] <rocket> a discussion with someone who has had way to little sleep after fixing 9 oracle/sap databases yesterday
[28-Aug-2009 15:15:44] <rmatte> I spent all day yesterday patching and verifying our Zenoss boxes
[28-Aug-2009 15:16:09] <rmatte> not quite as fun as fixing databases
[28-Aug-2009 15:16:38] <rocket> well it was more like the volume groups after an upgrade by a junior admin were hosed
[28-Aug-2009 15:16:47] <rmatte> ah
[28-Aug-2009 15:17:11] <rocket> not fun to see a missing disk on the big volume groups where everything is striped across it ..
[28-Aug-2009 15:17:12] <rmatte> junior admin = PFY, rocket = BOFH
[28-Aug-2009 15:17:42] * rocket deletes rmattes homedir giving him more space ...
[28-Aug-2009 15:17:45] <rhett> your guys' days were much more eventful than mine
[28-Aug-2009 15:17:54] <rmatte> hehe
[28-Aug-2009 15:18:22] <rmatte> you guys have updated your linux kernels I hope?
[28-Aug-2009 15:18:28] <rhett> i just worked on some cacti templates and worked on my car
[28-Aug-2009 15:18:31] <rmatte> that newly discovered exploit is brutal
[28-Aug-2009 15:18:32] * rocket is starting to checkin his homedir into git ... slowly and surely .. figuring out all of its lovely pitfalls
[28-Aug-2009 15:19:01] <rocket> rmatte: hadnt paid attention lately... whats going on with the kernel now?
[28-Aug-2009 15:19:26] <rocket> then again everything I have is behind a firewall at home ... and we only have 2 linux boxen at work ..
[28-Aug-2009 15:19:40] <rmatte> rocket: http://blog.iweb.com/en/2009/08/critical-security-vulnerability-affecting-linux-servers/2923.html
[28-Aug-2009 15:19:42] <rocket> unless you count the hmc's which are up to IBM to send us fixes for
[28-Aug-2009 15:19:43] <rmatte> it's a good one
[28-Aug-2009 15:19:46] <rhett> just checked yum. no updates
[28-Aug-2009 15:19:51] <rmatte> affects all kernels back to 2001
[28-Aug-2009 15:20:26] <rmatte> http://archives.neohapsis.com/archives/fulldisclosure/2009-08/0174.html
[28-Aug-2009 15:20:31] <rmatte> that's a more detailed description
[28-Aug-2009 15:21:20] <rmatte> the exploit requires the user to have access to the box, but it's basically a really quick way for a standard user to gain root privileges
[28-Aug-2009 15:21:43] <rocket> well if there is a hack for ssh or openvpn I am screwed
[28-Aug-2009 15:22:09] <rocket> otherwise there is no way anyone gets close to my boxes ... those are the only two ports open externally at all
[28-Aug-2009 15:22:48] * rocket wonders if the suse enterprise boxes have a fix for them ...
[28-Aug-2009 15:23:00] <rocket> oh well .. they arent production yet ...
[28-Aug-2009 15:23:13] <rmatte> hehe
[28-Aug-2009 15:24:07] <rmatte> anyone running 2.4.3, can you test and confirm: http://dev.zenoss.org/trac/ticket/5473
[28-Aug-2009 15:24:37] <rmatte> rocket: that old ssh hack combined with this would be brutal
[28-Aug-2009 15:24:42] <rhett> 2.4.2 here
[28-Aug-2009 15:24:42] <rmatte> but ssh is pretty rock solid now
[28-Aug-2009 15:25:47] <djack> hmmm question
[28-Aug-2009 15:25:59] <jb> my zen boxes are not exposed to the Internet at all..
[28-Aug-2009 15:26:01] <djack> i deleted the localhost device .... can it be recoved?
[28-Aug-2009 15:27:11] <rmatte> djack: you mean you deleted the actual collector I assume?
[28-Aug-2009 15:27:19] <rmatte> since "localhost" doesn't show up in the device list
[28-Aug-2009 15:27:22] <rocket> rmatte: fine .. make me setup port knocking too ...
[28-Aug-2009 15:27:27] <rocket> rmatte: be that way ...
[28-Aug-2009 15:27:32] <rmatte> lol
[28-Aug-2009 15:27:43] <djack> rmatte it was in my list
[28-Aug-2009 15:27:51] <rmatte> djack: well it shouldn't have been
[28-Aug-2009 15:28:20] <djack> well then it is good that i deleted it.
[28-Aug-2009 15:28:21] <rmatte> djack: deleting localhost from your device list won't cause issues
[28-Aug-2009 15:28:34] <rmatte> deleting the actual collector from the collector list would
[28-Aug-2009 15:28:35] <djack> well i can remember why it was in the list
[28-Aug-2009 15:28:44] <rmatte> why?
[28-Aug-2009 15:28:57] <djack> to test the apachemonitor package
[28-Aug-2009 15:29:01] <rmatte> ah
[28-Aug-2009 15:29:10] <djack> I'm going to set it up again
[28-Aug-2009 15:29:11] <rmatte> well, removing it won't cause issues, so you're fine
[28-Aug-2009 15:29:56] <djack> oki then
[28-Aug-2009 15:31:22] <djack> hmmm my fping is not graphing ...
[28-Aug-2009 15:31:58] <rmatte> this started just recently or it's been not graphing for a while?
[28-Aug-2009 15:32:46] <djack> 20 minutes ago
[28-Aug-2009 15:33:10] <djack> i added the template to a machine see the stuff in the perf tab but no data
[28-Aug-2009 15:34:23] <rmatte> well it usually takes like 15-20 mins for it to start graphing data
[28-Aug-2009 15:34:30] <rmatte> if you do a zencommand run -d <device>
[28-Aug-2009 15:34:34] <rmatte> do you see it run?
[28-Aug-2009 15:34:37] <rmatte> actually...
[28-Aug-2009 15:34:45] <rmatte> zencommand run -v10 -d <device>
[28-Aug-2009 15:34:57] <jb> hrm.. how is it that you fix the links in email alerts again? mine all have http://localhost/
[28-Aug-2009 15:34:59] <djack> can you combine zencommand with snmp on a device ?
[28-Aug-2009 15:35:12] <rmatte> djack: well, what do you mean by combine?
[28-Aug-2009 15:35:28] <rmatte> you can have a template with as many different types of datasources as you want in it
[28-Aug-2009 15:35:58] <rmatte> you can collect via custom script, WMI, SNMP, and SSH all at the same time from the same template
[28-Aug-2009 15:36:18] <djack> oki
[28-Aug-2009 15:36:53] <djack> my first thing i going to create myself is a check of our sap servers
[28-Aug-2009 15:37:22] <djack> like zenoss more and more
[28-Aug-2009 15:41:58] <kgoedtel> jb: zenactions.conf add a line that says- zopeurl http://www.example.com:8080
[28-Aug-2009 15:48:33] <rmatte> djack: yeh, once you start learning how to bend it to your will it gets pretty neat
[28-Aug-2009 15:50:01] <djack> haha i know why graphs are not made
[28-Aug-2009 15:50:18] <djack> zenoss does not have enough right to access the fping command
[28-Aug-2009 15:51:54] <djack> ssst the big boss just walked in ;-0
[28-Aug-2009 15:52:12] <davetoo> he's just the PIO
[28-Aug-2009 15:53:38] <rmatte> he's the pilot-induced oscillation?
[28-Aug-2009 15:53:54] <rmatte> or the programmed input/output?
[28-Aug-2009 15:56:12] <rmatte>
[28-Aug-2009 15:58:59] <jb> %(eventUrl)s
[28-Aug-2009 15:59:03] <jb> anyone know where to update this?
[28-Aug-2009 15:59:18] <jb> its "localhost" right now which obviously doesn't allow the links in the email alerts to work..
[28-Aug-2009 16:00:03] <rmatte> You're looking for the "zopeurl" option to the zenactions daemon. Edit $ZENHOME/etc/zenactions.conf and add a line like the following. Then restart zenactions
[28-Aug-2009 16:00:03] <rmatte> zopeurl http://your.server:8080
[28-Aug-2009 16:00:09] <jb> also trying to figure out how to properly apply CSS to the email alerts
[28-Aug-2009 16:00:11] <rmatte> got that from a post
[28-Aug-2009 16:00:15] <jb> (so I can change the font)
[28-Aug-2009 16:01:29] <rmatte> can't help with the CSS, no idea
[28-Aug-2009 16:01:58] <jb> cool.. thanks
[28-Aug-2009 16:02:25] <rmatte> np
[28-Aug-2009 16:04:27] <kgoedtel> i already answered that question!
[28-Aug-2009 16:04:29] <kgoedtel> haha
[28-Aug-2009 16:04:39] <rmatte> ah, I just saw that
[28-Aug-2009 16:04:46] <rmatte> I was away from my computer for a bit so didn't notice
[28-Aug-2009 16:04:56] <jb> oh
[28-Aug-2009 16:04:58] <jb> crap, i missed it too
[28-Aug-2009 16:05:02] <kgoedtel>
[28-Aug-2009 16:05:03] <jb> kgoedtel: thanks!
[28-Aug-2009 16:05:17] <kgoedtel> np lol
[28-Aug-2009 16:05:25] <jb> i wonder if there is a main template stored somewhere in a flat file
[28-Aug-2009 16:05:41] <jb> i can't add CSS directly to the alert format in the GUI
[28-Aug-2009 16:10:30] <rmatte> you could always just create some sort of script that creates and sends a CSS formatted email
[28-Aug-2009 16:10:42] <rmatte> and add it under the commands tab in the Event Manager section
[28-Aug-2009 16:11:26] <rmatte> Here's an example of a command that we use to call on a script to create tickets in our ticketing system...
[28-Aug-2009 16:11:47] <rmatte> /mspfiles/scripts/create_ticket.pl -autonotify -customer 'gen-00025' -device '${dev/id}' -deviceIP '${dev/manageIp}' -collector `hostname -f` -first '${evt/firstTime}' -last '${evt/lastTime}' -count '${evt/count}' -summary '${evt/summary}' -noteTitle 'System Monitor Error' -note '${evt/message}' -severity '${evt/severity}' -group '${evt/DeviceGroups}' -impact '${evt/DevicePriority}' -component '${evt/component}' -queue 'Front Line' >> ~/APIOutput.txt
[28-Aug-2009 16:11:57] <rmatte> you could do something similar but have the script send an email
[28-Aug-2009 16:16:22] <djack> grmbl just lost on ebay
[28-Aug-2009 16:16:47] <rmatte> lost for what?
[28-Aug-2009 16:16:57] <djack> a nas for home network
[28-Aug-2009 16:17:20] <djack> went for 76 euro ex discs
[28-Aug-2009 16:20:32] <rmatte> ah
[28-Aug-2009 16:20:49] <rmatte> what brand?
[28-Aug-2009 16:21:25] <djack> something unknown
[28-Aug-2009 16:21:47] <djack> GIGABIT DUAL NAS RAID SYSTEM
[28-Aug-2009 16:21:56] <rmatte> ah
[28-Aug-2009 16:22:33] <djack> something small for at home could use 1 T space that is mirrored
[28-Aug-2009 16:23:52] <rmatte> I want the Synology DS-207+
[28-Aug-2009 16:24:28] <jb> i have a readynas pro
[28-Aug-2009 16:24:29] <jb> 6bay
[28-Aug-2009 16:24:29] <rmatte> it has LAMP support, Built in BitTorrent, NFS support
[28-Aug-2009 16:24:39] <jb> ^^ it has all of that..
[28-Aug-2009 16:24:44] <jb> and i can write at like 100
[28-Aug-2009 16:24:47] <jb> 100MB/sec to it
[28-Aug-2009 16:24:48] <rmatte> nice
[28-Aug-2009 16:24:49] <jb> its pricey though
[28-Aug-2009 16:24:57] <rmatte> I don't need a 6bay though
[28-Aug-2009 16:25:00] <rmatte> that's just too much
[28-Aug-2009 16:25:04] <rmatte> 2 bay would be perfect
[28-Aug-2009 16:25:08] <jb> yeah im only using 4 right now
[28-Aug-2009 16:25:17] <jb> but its easy to auto-expand the "raid5" containers..
[28-Aug-2009 16:25:17] <rocket> rmatte: the qnap looks nice ...
[28-Aug-2009 16:25:17] <rmatte> 2 1 terabyte drives mirrored
[28-Aug-2009 16:25:31] <jb> i have 4 x 1.5's right no
[28-Aug-2009 16:25:32] <jb> now
[28-Aug-2009 16:25:45] <davetoo> poo
[28-Aug-2009 16:25:50] <davetoo> another place ZMI is broken
[28-Aug-2009 16:25:52] <rmatte> http://ncix.com/products/?sku=39256&vpn=SS-439-Pro&manufacture=QNAP%20Systems%20Inc.
[28-Aug-2009 16:25:57] <rmatte> that one looks nice
[28-Aug-2009 16:26:00] <rocket> jb: how is the ready nas?
[28-Aug-2009 16:26:12] <rmatte> wow, qnap has an 8 bay lol
[28-Aug-2009 16:26:27] <rmatte> it's over a grand canadian though
[28-Aug-2009 16:26:37] <rocket> jb: ouch ready nas pro is expensive ...
[28-Aug-2009 16:26:57] <jb> yeah it is
[28-Aug-2009 16:27:07] <jb> its really overkill for home usage
[28-Aug-2009 16:27:13] <jb> unless you are an enthusiast
[28-Aug-2009 16:27:19] <davetoo> I need to build a fileserver for the house
[28-Aug-2009 16:27:20] <jb> i paid i think $1100 for the barebones
[28-Aug-2009 16:27:29] <rmatte> I still think the Synology NASes look good for the price
[28-Aug-2009 16:27:33] * rocket needs jb's job ...
[28-Aug-2009 16:27:37] <jb> haha
[28-Aug-2009 16:27:48] <davetoo> I wish I could take something home from work; I work for one of the *big* nas/san mfrs
[28-Aug-2009 16:27:50] <jb> it runs debian
[28-Aug-2009 16:27:57] <jb> its really just a linux box.. with hot swappable bays, etc
[28-Aug-2009 16:28:04] <rocket> the qnap runs some flavor of linux ...
[28-Aug-2009 16:28:14] <jb> has lots of features built in
[28-Aug-2009 16:28:53] <rmatte> I want a NetApp for home
[28-Aug-2009 16:28:55] <jb> similar to the qnaps
[28-Aug-2009 16:32:53] <djack> rmatte we have 2 netapps at work....
[28-Aug-2009 16:32:56] <djack> they suck
[28-Aug-2009 16:36:48] <cwj> sorry, does not compute
[28-Aug-2009 16:37:16] <cwj> only netapp ive used that sucked was 8+ years old
[28-Aug-2009 16:39:16] <davetoo> man...
[28-Aug-2009 16:39:33] <davetoo> something about the factory stuff, I think, is creating inconsistent objects
[28-Aug-2009 16:39:50] <davetoo> I can't access ipRouteEntry objects from the ZMI
[28-Aug-2009 16:40:00] <davetoo> nexthopip property is missing
[28-Aug-2009 16:40:04] <mrayzenoss> rmatte: mrchippy checked out 5473 and said it's not a regression, at least not since 2.4.2
[28-Aug-2009 16:45:28] <rmatte> ok...
[28-Aug-2009 16:45:43] <rmatte> but were you able to reproduce it in 2.4.3?
[28-Aug-2009 16:47:00] <rmatte> seems like an odd issue to suddenly spring up
[28-Aug-2009 16:47:34] <davetoo> feh
[28-Aug-2009 16:47:39] <davetoo> I left a pair of shoes in Annapolis
[28-Aug-2009 16:47:44] <rmatte> lol
[28-Aug-2009 16:48:00] <davetoo> and the fine people at Holiday In Expre$$ won't ship them to me on their dime
[28-Aug-2009 16:48:03] <davetoo> such service
[28-Aug-2009 16:48:07] <davetoo> Inn
[28-Aug-2009 16:48:28] <davetoo> the lame thing is... two of the guys in training got rooms at the Westin for less than what I paid for HI Express
[28-Aug-2009 16:48:30] <davetoo> !!!
[28-Aug-2009 16:49:00] <mrayzenoss> dang, the Westin isn't cheap
[28-Aug-2009 16:49:17] <davetoo> They said they paid less than the $107 we paid at HI Express
[28-Aug-2009 16:49:24] <mrayzenoss> rmatte: yeah, Chip said he tested on 2.4.3 and 2.4.2
[28-Aug-2009 16:49:25] <rmatte> brutal
[28-Aug-2009 16:49:37] <davetoo> Some special popped up on some website, I guess.
[28-Aug-2009 16:49:38] <rmatte> mrayzenoss: so it is reproduceable
[28-Aug-2009 16:49:41] <rmatte> ?
[28-Aug-2009 16:49:53] <mrayzenoss> davetoo: seriously? I paid $159 a night...
[28-Aug-2009 16:49:56] <mrayzenoss> nuts
[28-Aug-2009 16:50:08] <mrayzenoss> rmatte: definitely reproducible, just not a regression
[28-Aug-2009 16:50:19] <rmatte> mrayzenoss: ok, but when will it be fixed then?
[28-Aug-2009 16:50:35] <rmatte> It's causing major issues for us from an inventory perspective
[28-Aug-2009 16:50:40] <rmatte> and when generating report info
[28-Aug-2009 16:50:59] <djack> going to bed
[28-Aug-2009 16:51:04] <mrayzenoss> later
[28-Aug-2009 16:51:11] <mrayzenoss> rmatte: no idea
[28-Aug-2009 16:51:27] <mrayzenoss> hardening iterations are starting now, so features are mostly frozen for King Crab
[28-Aug-2009 16:51:37] <rmatte> eugh, crappy
[28-Aug-2009 16:51:43] <mrayzenoss> so defects should be getting all the attention
[28-Aug-2009 16:52:02] <mrayzenoss> rmatte: the good news is that's not a feature, it's a bug
[28-Aug-2009 16:52:02] <rmatte> so we could have to go over 2 months without that
[28-Aug-2009 16:52:11] <rmatte> yeh
[28-Aug-2009 16:52:16] <rmatte> I'd say so, it worked before
[28-Aug-2009 16:52:21] <rmatte> it's a lack of a feature if anything
[28-Aug-2009 16:52:22] <rmatte> lol
[28-Aug-2009 16:53:03] <rmatte> I don't suppose any report fixes will be making their way in to King Crab?
[28-Aug-2009 16:53:31] <rmatte> such as exports that actually work?
[28-Aug-2009 16:53:48] <mrayzenoss> no idea, I wasn't in the iteration planning meeting, I just know the next 2 weeks are for getting the beta ready
[28-Aug-2009 16:53:56] <rmatte> k
[28-Aug-2009 16:54:12] <mrayzenoss> and Development will be fixing bugs until KC is frozen
[28-Aug-2009 16:55:54] <rmatte> I'd love for them to fix the report bugs so that exports work, it's absolutely killing me right now because instead of doing stuff that I should be doing I'm having to write python scripts to dump report data
[28-Aug-2009 16:55:58] <mrayzenoss> here's the open queue: http://dev.zenoss.org/trac/report/3
[28-Aug-2009 16:56:10] <rmatte> and I just found out that my interface utilization script is broken when I thought it was working properly
[28-Aug-2009 16:56:17] <rmatte> so it's back to the drawing board on Monday for that one
[28-Aug-2009 16:56:30] <mrayzenoss> so 11 priority 1s for King Crab right now, those get done first
[28-Aug-2009 16:57:10] <mrayzenoss> 5453 looks like it touches reporting, maybe someone will be digging into those
[28-Aug-2009 16:58:16] <rmatte> the whole report system needs someone to go over it with a fine toothed comb
[28-Aug-2009 16:58:23] <rmatte> because it is definitely not Production grade at this point
[28-Aug-2009 16:58:30] <rmatte> whereas everything else in Zenoss is
[28-Aug-2009 16:59:05] <davetoo> mrayzenoss: doesn't anybody besides me use the ZMI for debugging/education/development?
[28-Aug-2009 16:59:09] <rmatte> I'd rather have working reports before a new UI
[28-Aug-2009 16:59:09] <mrayzenoss> yeah, I think it's currently high on the overhaul list for the next release, which isn't to say lots of the defects against it won't be fixed
[28-Aug-2009 16:59:27] <davetoo> sorry, I'll let you guys finish this thread
[28-Aug-2009 16:59:48] <rmatte> davetoo: the zope management interface?
[28-Aug-2009 16:59:52] <rmatte> I use it all the time, why?
[28-Aug-2009 17:00:04] <davetoo> try looking at, well,
[28-Aug-2009 17:00:13] <davetoo> try looking at the properties for /Devices, for instance
[28-Aug-2009 17:00:26] <davetoo> or properties for an IpRoute
[28-Aug-2009 17:00:36] <rmatte> what about it?
[28-Aug-2009 17:00:52] <davetoo> for me, it barfs,
[28-Aug-2009 17:01:14] <mrayzenoss> davetoo: hmm... me too
[28-Aug-2009 17:01:16] <rmatte> ok, well, by "look at properties" what are you doing exactly?
[28-Aug-2009 17:01:20] <rmatte> I'm just going to: http://lab01.novanoc.com:8180/zport/dmd/Devices/manage
[28-Aug-2009 17:01:31] <davetoo> in the zmi, select the "Properties" tab on the right
[28-Aug-2009 17:01:36] <rmatte> oh right
[28-Aug-2009 17:01:39] <davetoo> mrayzenoss: I have a case and a ticket open,
[28-Aug-2009 17:01:47] <rmatte> yeh, same here
[28-Aug-2009 17:01:58] <davetoo> for the missing devtypes properties
[28-Aug-2009 17:02:13] <rmatte> looks like there's a bug in the getProperty function in OFS.PropertyManager module
[28-Aug-2009 17:02:28] <davetoo> I'm not sure where the problem is,
[28-Aug-2009 17:02:29] <rmatte> yeh
[28-Aug-2009 17:02:37] <davetoo> but seems like the Factory stuff isn't working right
[28-Aug-2009 17:02:45] <davetoo> because something is creating inconsistent objects
[28-Aug-2009 17:02:59] <davetoo> I found it in zendmd (essentially)
[28-Aug-2009 17:03:17] <davetoo> because I have some scripts that depend upon propertyItems() to dump stuff to a zenpack
[28-Aug-2009 17:03:30] <rmatte> OHHHH
[28-Aug-2009 17:03:36] <rmatte> I just figured out the bug in my script
[28-Aug-2009 17:03:38] <rmatte> stupid me
[28-Aug-2009 17:03:49] <rmatte> I should have it setting percentUsed to 0
[28-Aug-2009 17:03:59] <rmatte> so that if it can't calculate it it doesn't take the value from the last interface
[28-Aug-2009 17:04:09] <rmatte> that's why the info I dumped is all screwed
[28-Aug-2009 17:04:11] <davetoo> and just now, I'm working on stuff to add routes to a devicen and some of the routes are missing the nexthopip property. They're not just empty/null, they are missing.
[28-Aug-2009 17:04:12] <rmatte> everything else is fine
[28-Aug-2009 17:04:40] <davetoo> so I guess I dont' understand what good the _properties stuff does
[28-Aug-2009 17:05:04] <davetoo> I dont' really understand a lot about Zope yet,
[28-Aug-2009 17:05:21] <davetoo> to know if that's from a standard Zope product or a Zenoss thing.
[28-Aug-2009 17:07:45] <rmatte> Zope can be pretty dirty
[28-Aug-2009 17:08:02] <davetoo> it's sure complex
[28-Aug-2009 17:08:09] <davetoo> or complicated
[28-Aug-2009 17:08:20] <davetoo> I forget which is the correct word to use
[28-Aug-2009 17:08:30] <rmatte> yeh, it's eye-bleed inducing
[28-Aug-2009 17:11:03] <mrayzenoss> my understanding is that Zope 2.11.2 in King Crab is some steps towards less ocular bleeding
[28-Aug-2009 17:16:58] <rmatte> I think I'm going to spend some time on Monday writing up a detailed review/report on the Zenoss reporting system which highlights improvements that could be made, or bugs that could be addressed. There honestly aren't that many, but they are biggies.
[28-Aug-2009 17:18:37] <mrayzenoss> that would be appreciated
[28-Aug-2009 17:19:10] <rmatte> I'll make it as professional as possible so that it's something you could bring to a meeting or something.
[28-Aug-2009 17:19:18] <rmatte> with screenshots and such
[28-Aug-2009 17:19:25] <rmatte> anyways, I'm out, have a good weekend guys
[28-Aug-2009 17:19:33] <mrayzenoss> you too, later
[28-Aug-2009 17:26:56] <davetoo> hmm
[28-Aug-2009 17:27:06] <davetoo> how can I debug page template calls?
[28-Aug-2009 17:36:20] <davetoo> uh oh
[28-Aug-2009 17:36:22] <davetoo> heh
[28-Aug-2009 17:37:20] <davetoo> I'm not sure that making ZenModel refreshable in zope is a good thing
[28-Aug-2009 17:48:23] <davetoo> ok, well,
[28-Aug-2009 17:48:30] <davetoo> I managed to figure out how to add ip routes with zendmd
[28-Aug-2009 17:48:40] <davetoo> I'm done with Zenoss for the day
[28-Aug-2009 17:48:42] <davetoo> brain hurts
[28-Aug-2009 17:49:20] <davetoo> I'll be having nightmares tonight about mibs and zope page templates and exceptions and...
[28-Aug-2009 18:33:06] <davetoo> what the heck
[28-Aug-2009 18:33:17] <davetoo> why isn't zendisc cycliing through all of my snmp strings?
[28-Aug-2009 18:33:51] <davetoo> n/mind
[28-Aug-2009 18:34:16] <davetoo> I guess it is, it just isn't running zenmodel
[28-Aug-2009 18:34:18] <davetoo> odd
[29-Aug-2009 00:00:46] [disconnected at Sat Aug 29 00:00:46 2009]
[29-Aug-2009 00:00:46] [connected at Sat Aug 29 00:00:46 2009]
[29-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[29-Aug-2009 02:05:51] <davetoo> uh oh
[29-Aug-2009 13:07:15] <stc_s> anyone knows about zenping problems? and to resolve it to more 300 devices or more?
[29-Aug-2009 13:07:29] <stc_s> I have a serious problem with zenoss
[29-Aug-2009 13:09:22] <stc_s> help
[29-Aug-2009 16:59:19] <etank> is this page still accurate? http://www.zenoss.com/community/docs/howtos/zenbatchload-howto
[29-Aug-2009 17:05:28] <etank> looks like the -i is no longer valid
[29-Aug-2009 17:05:34] <davetoo> hmm
[29-Aug-2009 17:05:56] <davetoo> rtfc
[29-Aug-2009 17:06:18] <davetoo> which version?
[29-Aug-2009 20:10:15] kevin7kal_ is now known as kevin7kal
[29-Aug-2009 23:14:50] <davetoo> damn cisco
[30-Aug-2009 00:00:46] [disconnected at Sun Aug 30 00:00:46 2009]
[30-Aug-2009 00:00:46] [connected at Sun Aug 30 00:00:46 2009]
[30-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[30-Aug-2009 16:15:40] <djack> hello world
[31-Aug-2009 00:00:46] [disconnected at Mon Aug 31 00:00:46 2009]
[31-Aug-2009 00:00:46] [connected at Mon Aug 31 00:00:46 2009]
[31-Aug-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[31-Aug-2009 09:10:56] <mrayzenoss> morning, anyone here zenpatch 2.4.3 on Friday?
[31-Aug-2009 09:11:07] <mrayzenoss> rmatte?
[31-Aug-2009 09:11:15] <mrayzenoss> looking for reasons not to push out 2.4.5
[31-Aug-2009 09:20:57] <rmatte> hey
[31-Aug-2009 09:21:06] <rmatte> well, what does 2.4.5 contain?
[31-Aug-2009 09:21:11] <rmatte> patchwise
[31-Aug-2009 09:21:44] <rmatte> just those 2 patches?
[31-Aug-2009 09:22:32] <mrayzenoss> 14744 and 14769
[31-Aug-2009 09:22:37] <rmatte> In my opinion: http://dev.zenoss.org/trac/ticket/5473 should be fixed before next official release since it's a very obvious bug
[31-Aug-2009 09:22:44] <mrayzenoss> I'm looking at the other 2 tickets assigned to 2.4.5
[31-Aug-2009 09:23:04] <rmatte> In addition to not being able to assign product details from the Edit tab, you can't properly add products via the edit tab either
[31-Aug-2009 09:23:05] <mrayzenoss> yeah, but we need to get 2.4.5 out
[31-Aug-2009 09:23:26] <rmatte> ok, well then I suppose, despite that obvious bug, everything else has been running smoothly
[31-Aug-2009 09:24:41] <jb> morning
[31-Aug-2009 09:25:28] <mrayzenoss> 2.4.5 is going to be just the 2 tickets then
[31-Aug-2009 09:25:34] <rmatte> k
[31-Aug-2009 09:26:21] <rmatte> You guys probably shouldn't wait longer than a week or two to get 2.4.6 out though
[31-Aug-2009 09:27:13] <mrayzenoss> well, the KC beta is dropping next week and half our developers are out, so it may be awhile
[31-Aug-2009 09:28:24] <rmatte> that sucks for me, but alright lol
[31-Aug-2009 09:28:25] <mrayzenoss> this week is VMworld and Red Hat Summit, so we're at both
[31-Aug-2009 09:29:48] <rmatte> I'm going to have engineers as well as my boss crawling all over me on that one
[31-Aug-2009 09:29:59] <rmatte> since we generate reports for billing that need to include hardware details
[31-Aug-2009 09:31:56] <jb> who is going to be at RHS?
[31-Aug-2009 09:32:31] <mrayzenoss> I know Mark and someone from marketing, I think one of the professional service engineers
[31-Aug-2009 09:32:47] <mrayzenoss> and sales and our director of alliances
[31-Aug-2009 09:33:13] <mrayzenoss> Mark = VP of Community, my boss
[31-Aug-2009 09:33:33] <jb> yeah
[31-Aug-2009 09:33:42] <jb> i'll be out of town, or I would go..
[31-Aug-2009 09:34:07] <mrayzenoss> it's nice that Red Hat and VMware have their big conferences at exactly the same time...
[31-Aug-2009 09:34:29] <mrayzenoss> so we've got people having to go to both
[31-Aug-2009 09:37:37] <rmatte> lol
[31-Aug-2009 10:10:59] <gwb2351> what's the easiest way to track down why a perf graph is not getting updated in zenoss?
[31-Aug-2009 10:11:10] <gwb2351> right now I 1) check the template for the data source command
[31-Aug-2009 10:11:19] <gwb2351> 2) validate that snmpwalk is returning a value
[31-Aug-2009 10:11:38] <gwb2351> how to 3) figure out why that value is not getting added/plotted?
[31-Aug-2009 10:16:10] <mrayzenoss> zensnmpperf run -v10 -d host
[31-Aug-2009 10:17:00] <mrayzenoss> verify that the template is bound
[31-Aug-2009 10:17:22] <rmatte> well, if the template isn't bound he shouldn't even see the graphs to begin with
[31-Aug-2009 10:17:24] <mrayzenoss> I assume 1) includes using the "Test" command
[31-Aug-2009 10:17:31] <mrayzenoss> agreed
[31-Aug-2009 10:17:40] <mrayzenoss> is that "empty graph" or "no graph"
[31-Aug-2009 10:17:46] <mrayzenoss> no graph means no template
[31-Aug-2009 10:17:58] <rmatte> yeh, like is the graph actually missing, or do you see that graph with "nan" values?
[31-Aug-2009 10:18:01] <gwb2351> well, i know the template is bound because the history of the data shows information, but then it just drops off (with nan) at odd times
[31-Aug-2009 10:18:01] <mrayzenoss> empty graph means something wonky in the data source, or it's not populated yet
[31-Aug-2009 10:18:21] <gwb2351> the graph is there; with some data; but from time to time (for long stretches, like hours) no data gets added
[31-Aug-2009 10:18:29] <rmatte> what is the template monitoring?
[31-Aug-2009 10:18:45] <gwb2351> in this instance, it's the powerware template (for a UPS)
[31-Aug-2009 10:18:56] <rmatte> and it uses command based datapoints?
[31-Aug-2009 10:19:12] <gwb2351> the data source is "/opt/zenoss/libexec/check_snmp -H ${here/manageIp} -C ${here/zSnmpCommunity} -r '1|3|4|5' -c 2 -o 1.3.6.1.4.1.534.1.2.5.0"
[31-Aug-2009 10:19:20] <rmatte> If you run a test of the command against a device, how long does it take for it to run through?
[31-Aug-2009 10:19:21] <gwb2351> which i can run from the command line, no problem, with an actual result
[31-Aug-2009 10:19:25] <gwb2351> <1s
[31-Aug-2009 10:20:15] <gwb2351> SNMP OK - 5 | iso.3.6.1.4.1.534.1.2.5.0=5 , real 0m0.005s
[31-Aug-2009 10:21:26] <gwb2351> the other SNMP values for the template are working/graphing OK, just not this data point
[31-Aug-2009 10:21:28] <mrayzenoss> why are you using check_snmp?
[31-Aug-2009 10:21:31] <rmatte> is this a premade template, from a ZenPack?
[31-Aug-2009 10:21:34] <rmatte> or did you make this?
[31-Aug-2009 10:21:41] <gwb2351> premade from community zenpack
[31-Aug-2009 10:22:47] <gwb2351> http://www.zenoss.com/community/projects/zenpacks/powerware-ups
[31-Aug-2009 10:23:46] <mrayzenoss> yeah, he must have just pulled a nagios script over rather than do it from scratch
[31-Aug-2009 10:24:08] <mrayzenoss> an snmp-based data source would probably be more performant
[31-Aug-2009 10:24:31] <rmatte> yeh, I don't get why they are even using a command based datasource for that
[31-Aug-2009 10:24:33] <mrayzenoss> maybe try adding a parallel snmp-based template that hits the same OIDs and see if the gaps go away
[31-Aug-2009 10:24:38] <rmatte> since they are just pulling straight SNMP values
[31-Aug-2009 10:24:43] <gwb2351> OK, I'll try that
[31-Aug-2009 10:40:28] <mrayzenoss> I gotta go give a demo, back later
[31-Aug-2009 10:40:40] <rmatte> have fun
[31-Aug-2009 10:40:49] <davetoo> heh
[31-Aug-2009 11:06:13] <mrhinkle> FYi...Our Zenoss community manager is presenting a live demo of Zenoss Core and Q&A via Webex you can sign up and ask questions for the next hour - http://forms.zenoss.com/getting-started
[31-Aug-2009 12:27:19] <gemineye> Does anyone know when the next release of Core is schedule for?
[31-Aug-2009 12:28:54] <etank> gemineye: i think i saw last week that it should be this week.
[31-Aug-2009 12:28:57] <etank> 2.4.4 i think
[31-Aug-2009 12:29:14] <jb> 2.4.5
[31-Aug-2009 12:29:17] <jb> today probably..
[31-Aug-2009 12:30:00] <gemineye> so just bug fixes.
[31-Aug-2009 12:30:25] <gemineye> We're at 2.4.2, they skipping some levels?
[31-Aug-2009 12:34:27] <jb> 2.4.3/2.4.4 had some bugs..
[31-Aug-2009 12:34:38] <jb> 2.4.4/2.4.5 just fix a couple of issues
[31-Aug-2009 12:35:17] <rmatte> 2.4.5 is just 2.4.3 with zenpatch 14744 and zenpatch 14769 applied to it
[31-Aug-2009 12:38:39] <etank> if i email support@zenoss.com should it show up as a case?
[31-Aug-2009 12:38:46] <etank> for enterprise that is
[31-Aug-2009 12:40:10] <rmatte> bah, just noticed that they changed the memory report slightly so my customized memory report plugin doesn't work with the .rpt file
[31-Aug-2009 12:40:46] <rmatte> and I didn't back up the .rpt file
[31-Aug-2009 12:40:59] <rmatte> oh well, yet another project for tomorrow
[31-Aug-2009 12:48:30] <davetoo> damn
[31-Aug-2009 12:48:39] <davetoo> the more I work on this the more work I have to do
[31-Aug-2009 13:00:07] <rmatte> davetoo: yup, it's exponential for a while
[31-Aug-2009 13:01:07] <davetoo> well, one o fthe things that left a bad taste in the mouths of the level-1 people, from our previous core installs, is alert floods
[31-Aug-2009 13:05:45] <rmatte> yeh
[31-Aug-2009 13:14:54] <nzle> I wrote a nagios plugin for an internal program we have here and its working fine being called by zencommand via the devices local template, but I only see it on the dashboard, or the device status when somthing is wrong. I'd like to see it when its OK as well.
[31-Aug-2009 13:16:21] <nzle> Same issue with a heartbeat. How can I get a visible indicator of the hearbeat test before it recieves a heartbeat.
[31-Aug-2009 13:24:05] <nzle> ?
[31-Aug-2009 13:31:25] <nzle> Did my question of a few minutes ago get sent since I was not properly identified to to the server yet or should I resend it?
[31-Aug-2009 15:14:57] <rmatte> nzle: ummm, that's kind of pointless no?
[31-Aug-2009 15:15:19] <rmatte> I mean, if you don't see an indicator that something is wrong, then it's obviously ok
[31-Aug-2009 15:16:21] <rmatte> I don't see why you need an indicator specifically telling you something is ok
[31-Aug-2009 15:16:33] <rmatte> you generally just assume everything is ok unless you get an alert indicating otherwise
[31-Aug-2009 15:17:08] <rmatte> or are you talking about having it send a clear command?
[31-Aug-2009 15:17:36] <rmatte> (like, to automatically clear the previous event)
[31-Aug-2009 15:23:30] <davetoo> hmm
[31-Aug-2009 15:23:46] <davetoo> zenpatch 14744
[31-Aug-2009 15:23:53] <davetoo> patching file ZenEvents/zenactions.py
[31-Aug-2009 15:23:53] <davetoo> Hunk #1 FAILED at 273.
[31-Aug-2009 15:23:53] <davetoo> Hunk #2 FAILED at 304.
[31-Aug-2009 15:23:53] <davetoo> Hunk #3 succeeded at 314 with fuzz 1 (offset -13 lines).
[31-Aug-2009 15:23:53] <davetoo> 2 out of 3 hunks FAILED -- saving rejects to file ZenEvents/zenactions.py.rej
[31-Aug-2009 15:24:16] <davetoo> D'oh
[31-Aug-2009 15:24:25] * davetoo reads more closely
[31-Aug-2009 15:24:36] <gwb2351> there is value in having an explicit "everything is OK" value; systems hose themselves hard enough that they're unable to send alerts quite often.
[31-Aug-2009 15:24:39] <rmatte> davetoo: what version of Zenoss are you running?
[31-Aug-2009 15:24:43] <davetoo> the wrong one
[31-Aug-2009 15:24:51] <rmatte> obviously
[31-Aug-2009 15:24:52] <rmatte> lol
[31-Aug-2009 15:25:01] <rmatte> you need to have 2.4.3 installed for those patches to work
[31-Aug-2009 15:25:15] <rmatte> If you don't then just wait for 2.4.5 to be released
[31-Aug-2009 15:25:28] <davetoo> yeah, this is just a development system
[31-Aug-2009 15:25:32] <davetoo> moving to a bigger box
[31-Aug-2009 15:28:03] <rmatte> cool
[31-Aug-2009 15:30:30] <davetoo> my first development box is too slow
[31-Aug-2009 15:30:37] <davetoo> only one disk
[31-Aug-2009 15:30:47] <rmatte> our lab box is pretty darn slow, but it's because we're running like 10 VMs on it
[31-Aug-2009 15:30:59] <rmatte> our production stuff is lightning fast
[31-Aug-2009 15:31:14] <davetoo> I'm using vms for zenpack testing
[31-Aug-2009 15:31:40] <rmatte> we use vms for everything
[31-Aug-2009 15:31:40] <davetoo> but I need a "big" development system to see the impact of a big load of systems
[31-Aug-2009 15:31:48] <davetoo> we're moving that way
[31-Aug-2009 15:31:56] <rmatte> but I develop zenpacks on my lab box and then deploy them out to the other zenoss vms
[31-Aug-2009 15:32:11] <davetoo> so you use zenoss on vms for production?
[31-Aug-2009 15:32:15] <rmatte> yup
[31-Aug-2009 15:32:16] <davetoo> What do you do for storage?
[31-Aug-2009 15:32:21] <rmatte> SAN
[31-Aug-2009 15:32:21] <davetoo> And how many devices?
[31-Aug-2009 15:32:37] <davetoo> Enterprise? or Core?
[31-Aug-2009 15:32:38] <rmatte> it varies, we support multiple clients, so each VM is for a different client
[31-Aug-2009 15:32:42] <rmatte> Core
[31-Aug-2009 15:32:58] <davetoo> ah
[31-Aug-2009 15:33:02] <rmatte> each VM has a VPN tunnel to the client's network
[31-Aug-2009 15:33:19] <rmatte> the largest network we're monitoring right now is only about 370 devices
[31-Aug-2009 15:33:21] <davetoo> Yeah, we'll still have a bunch of core systems to monitor stuff nobody wants to pay to license,
[31-Aug-2009 15:33:23] <rmatte> routers, switches, and windows servers
[31-Aug-2009 15:33:30] <davetoo> but we're bringing up our Enterprise system.
[31-Aug-2009 15:33:43] <rmatte> cool
[31-Aug-2009 15:33:49] <rmatte> we can't afford enterprise at the moment
[31-Aug-2009 15:33:56] <rmatte> and we're making due with core quite well
[31-Aug-2009 15:34:02] <rmatte> we're just having to be ingenious about it
[31-Aug-2009 15:34:03] <rmatte> lol
[31-Aug-2009 15:34:17] <davetoo> We can't afford not to have the support
And not having a centralized interface.[31-Aug-2009 15:34:26] <rmatte> ah
[31-Aug-2009 15:34:31] <davetoo> Finally got some $$ support for monitoring.
[31-Aug-2009 15:34:43] <etank> davetoo: im in the same boat
[31-Aug-2009 15:34:43] <rmatte> I'm pretty good at providing support for our stuff
[31-Aug-2009 15:34:51] <rmatte> but we're probably not as large scale as you guys yet
[31-Aug-2009 15:34:52] <etank> we just did our enterprise install last friday
[31-Aug-2009 15:35:02] <davetoo> 2.4.2 is doing much better than previous versions.
[31-Aug-2009 15:35:14] <rmatte> 2.4.2 has a security bug in it though
[31-Aug-2009 15:35:19] <davetoo> rmatte: well, we started out with 2500 nodes
[31-Aug-2009 15:35:35] <davetoo> but if we get the ESX monitoring done nicely,
[31-Aug-2009 15:35:47] <rmatte> ah
[31-Aug-2009 15:35:47] <davetoo> that will probably explode as other groups adopt it for VMware
[31-Aug-2009 15:35:57] <davetoo> etank: how many collectors?
[31-Aug-2009 15:36:18] <etank> davetoo: right now one. but we are going to work on getting some others out soon.
[31-Aug-2009 15:36:33] <etank> we have some devices that are only available through a vpn connections
[31-Aug-2009 15:36:50] <etank> so we are going to try putting collectors in those areas
[31-Aug-2009 15:36:53] <rmatte> we have 3 blade boxes right now, each with 15GHz processing and 32 gigs of RAM, then we have a SAN (after raiding it can store over 5TB I believe). We have an image of a Zenoss client base that we just clone, change the IPs and hostname on, and it's good to go.
[31-Aug-2009 15:37:05] <rmatte> we can have a new client up and monitored in less than a day
[31-Aug-2009 15:37:15] <rmatte> and all our Zenoss instances tie in to our ticketing system
[31-Aug-2009 15:37:20] <davetoo> oh
[31-Aug-2009 15:37:26] <rmatte> so we don't need a global event console since the ticketing system basically does that
[31-Aug-2009 15:37:32] <davetoo> then you must have done a lot of work on event filtering
[31-Aug-2009 15:37:36] <etank> if things go well then maybe the corporate IT guys will adopt Zenoss as their solution too
[31-Aug-2009 15:37:59] <etank> rmatte: what ticketing system do you use?
[31-Aug-2009 15:38:04] <rmatte> davetoo: I've done a lot of tuning so far, yeh
[31-Aug-2009 15:38:21] <rmatte> etank: a customized version of OTRS
[31-Aug-2009 15:38:42] <davetoo> that's one area we haven't done a good job at, previously,and why some of the groups have a bad taste in their mouth.
[31-Aug-2009 15:38:54] <davetoo> I have got to come up with some kind of summarization and flood protection.
[31-Aug-2009 15:39:22] <rmatte> I still have tuning on my list to do, I basically just do the tuning on my lab box, load it in to a ZenPack, then push the ZenPack upgrade across all servers
[31-Aug-2009 15:39:32] <rmatte> it's the easiest way to keep them all up to date
[31-Aug-2009 15:39:53] <rmatte> there are some items I haven't had time to tune yet this week that engineers are bugging me about
[31-Aug-2009 15:39:57] <davetoo> {how} do you manage dependencies? layer-2 and/or apps?
[31-Aug-2009 15:40:29] <rmatte> we don't do layer 2, and Zenoss handles app dependencies alright, it's a bit off sometimes
[31-Aug-2009 15:40:43] <rmatte> although that's why we have a select all button in our ticketing system hehe
[31-Aug-2009 15:40:46] <davetoo> We have some *big* layer 2 switches
[31-Aug-2009 15:40:59] <rmatte> if a flood of events comes in the front line guys just select a bunch of tickets and relate them back to 1
[31-Aug-2009 15:41:02] <rmatte> then work that 1
[31-Aug-2009 15:41:05] <rmatte> so it's not that bad
[31-Aug-2009 15:41:23] <davetoo> I do not like our ticketing system
[31-Aug-2009 15:41:37] <rmatte> we're doing smaller scale monitoring for the most part, I doubt we'll be monitoring anything over 1000 devices any time soon
[31-Aug-2009 15:42:38] <rmatte> anything over 1000 on one box I mean to say
[31-Aug-2009 15:43:26] <rmatte> davetoo: our ticketing system is a bit off right now too, but that's because we're still doing development work
[31-Aug-2009 15:43:50] <rmatte> we still have a customer portal and a cmdb system being developed. The CMDB is going to contain all contact info, circuit info, etc... for each device
[31-Aug-2009 15:47:39] <rmatte> by the way, if you're using 2.4.2 you probably want to do zenpatch 14181
[31-Aug-2009 15:47:43] <rmatte> if you haven't already
[31-Aug-2009 15:47:48] <rmatte> http://dev.zenoss.org/trac/ticket/5279
[31-Aug-2009 15:48:08] <davetoo> thanks
[31-Aug-2009 15:48:25] <rmatte> nasty security hole
[31-Aug-2009 15:50:37] <rmatte> which is why they are rushing out 2.4.5 even though it has bugs in it.
[31-Aug-2009 15:51:36] <straterra> Lemme see what I have
[31-Aug-2009 15:52:02] <rmatte> it affects any 2.4.x version before 2.4.3
[31-Aug-2009 15:52:04] <straterra> I'm on 2.4.2
[31-Aug-2009 15:52:13] <rmatte> yeh, then apply that patch
[31-Aug-2009 15:52:17] <rmatte> and restart Zenoss
[31-Aug-2009 15:52:21] <straterra> Is it easy to update zenoss?
[31-Aug-2009 15:52:40] <rmatte> depends on how you installed it
[31-Aug-2009 15:52:49] <straterra> rhel rpm
[31-Aug-2009 15:52:53] <rmatte> then yes
[31-Aug-2009 15:53:09] <davetoo> I used to do source installs. Very slow.
[31-Aug-2009 15:53:14] <davetoo> unreliable.
[31-Aug-2009 15:53:17] <rmatte> it's just like doing any other rpm update
[31-Aug-2009 15:53:38] <davetoo> but you have to READ THE INSTRUCTIONS
[31-Aug-2009 15:53:57] <davetoo> Because there are required upgrade paths if you've waited a long time.
[31-Aug-2009 15:54:16] <straterra> So..wheres the rpm I need?
[31-Aug-2009 15:54:18] <rmatte> I started learning unix with FreeBSD when I was 15
[31-Aug-2009 15:54:34] <rmatte> so I've done my share of source installs
[31-Aug-2009 15:54:37] <rmatte> lol
[31-Aug-2009 15:55:17] <rmatte> I think the first version of FreeBSD that I used was at the end of the 4.x series
[31-Aug-2009 15:55:29] <rmatte> or the very start of 5.x
[31-Aug-2009 15:56:20] <rmatte> straterra: there are none yet, 2.4.3 was pulled back, 2.4.2 is the latest version right now
[31-Aug-2009 15:56:34] <straterra> Ah
[31-Aug-2009 15:56:37] <straterra> I can wait then..
[31-Aug-2009 15:56:41] <rmatte> but if you run "zenpatch 14181" as the zenoss user, that'll patch that security hole for the time being
[31-Aug-2009 15:56:44] <straterra> That issue isn't a big deal for me
[31-Aug-2009 15:57:00] <mrayzenoss> as long as you trust your users
[31-Aug-2009 15:57:01] <straterra> our zenoss server is internal only
[31-Aug-2009 15:57:12] <straterra> None of the users know it even exists
[31-Aug-2009 15:57:14] <rmatte> yeh, hopefully you trust your users
[31-Aug-2009 15:57:23] <cgibbons> well, that only took all day
[31-Aug-2009 15:57:29] <straterra> And they can't get to it..that server is on its own subnet
[31-Aug-2009 15:59:06] <rmatte> we were more referring to other users on the zenoss box
[31-Aug-2009 15:59:11] <rmatte> rather than "users" in general
[31-Aug-2009 15:59:15] <rmatte> but yeh, you're fine then
[31-Aug-2009 15:59:42] <davetoo> I wish I understood LDAP more
[31-Aug-2009 15:59:55] <rmatte> yeh, we use LDAP for everything
[31-Aug-2009 16:00:12] <rmatte> ticketing system login, wiki login, zenoss login, etc...
[31-Aug-2009 16:00:17] <davetoo> I don't understand why it has to "search"
[31-Aug-2009 16:00:26] <davetoo> one system I have does'nt need a Manager user
[31-Aug-2009 16:00:38] <davetoo> it just simply tries to bind to the server with the user's credentials,
[31-Aug-2009 16:00:44] <davetoo> and if it succeeds, it's authorized
[31-Aug-2009 16:00:49] <straterra> rmatte: are you familiar with foxmarks/xmarks ?
[31-Aug-2009 16:00:58] <rmatte> davetoo: hunh?
[31-Aug-2009 16:01:02] <rmatte> straterra: nope
[31-Aug-2009 16:01:06] <davetoo> tocotox
[31-Aug-2009 16:01:20] <rmatte> davetoo: oh I see what you're saying
[31-Aug-2009 16:01:22] <straterra> It's a plugin for firefox that lets you sync your bookmarks/saved passwords (passwords are encrypted)
[31-Aug-2009 16:01:30] <rmatte> davetoo: then you have a badly configure LDAP server
[31-Aug-2009 16:01:32] <straterra> I run one here at work..a server, that is
[31-Aug-2009 16:01:35] <davetoo> I'm not using ldap for groups or anything
[31-Aug-2009 16:01:50] <rmatte> straterra: ah
[31-Aug-2009 16:01:51] <straterra> I wrote my own frontend for it and it ties in to AD LDAP
[31-Aug-2009 16:02:16] <rmatte> cool
[31-Aug-2009 16:02:23] <chudler> what? the xmarks ties into LDAP? neato, but why?
[31-Aug-2009 16:02:37] <straterra> So that users can use their login username/pw
[31-Aug-2009 16:02:42] <rmatte> davetoo: ours requires an administrative account to connect to LDAP, if it's not an LDAP admin it doesn't work, which is the way it should be.
[31-Aug-2009 16:02:51] <straterra> It's password protected to each user is only allowed access to their own bookmarks
[31-Aug-2009 16:02:52] <chudler> ah, I see, its for auth and possibly decryption?
[31-Aug-2009 16:03:16] <straterra> They can set a seperate pw for their encryption
[31-Aug-2009 16:03:42] <straterra> It's basically apache2 + mysql + php + AD LDAP
[31-Aug-2009 16:08:17] <nzle> rmatte, responding to somthing a while ago regarding my wanting an indicator of an OK service check. I dont think its 'pointless' how about the case where a script or somthing generating the event is borked. An indicator of a test in 'Unknown' status would tell me that. Assuming no news is good news would not.
[31-Aug-2009 16:09:11] <rmatte> nzle: well, if that's what you want then you're going to probably have to do it with a transform
[31-Aug-2009 16:10:06] <rmatte> if result is blah: evt.summary = "BAD"; if result is good: evt.summary = "OK"
[31-Aug-2009 16:10:08] <rmatte> or whatever
[31-Aug-2009 16:10:39] <nzle> doesnt a transform still need an incoming event? What I want is to configure an Expected event and have it show Unknown untill it gets someting.
[31-Aug-2009 16:11:10] <nzle> Ill look up transforms and read up on them... thanx for the direction..
[31-Aug-2009 16:11:26] <rmatte> nzle: ok, do you want an actual OK event generated in the event console?
[31-Aug-2009 16:11:49] <rmatte> actually, I might have a better solution for you
[31-Aug-2009 16:11:52] <rmatte> depending on what you want
[31-Aug-2009 16:11:59] <nzle> No, I want a null event. An indicator that I expect an event.
[31-Aug-2009 16:12:40] <rmatte> nzle: ok, you're not explaining this very clearly, what's your definition of a "null" event
[31-Aug-2009 16:12:44] <rmatte> one that clears the bad event?
[31-Aug-2009 16:13:08] <rmatte> give me an example
[31-Aug-2009 16:14:40] <chudler> sounds like a heartbeat to me..
[31-Aug-2009 16:14:41] <nzle> Take a heartbeat for example. As I understand it, you have an incoming event and you transform it to give it a timeout. If another doesnt come before the timeout, it goes critical. What if I didnt get a heartbeat at all following a failure that tood out zenoss too. I would have no indication of a lost heartbeat.
[31-Aug-2009 16:16:22] * rmatte scratches his head: "following a failure that tood out zenoss too"
[31-Aug-2009 16:16:37] <rmatte> (and I know that's "took")
[31-Aug-2009 16:16:39] <nzle> unless I had a indicator on the dashboard where I expect a heartbeat. Leave it grey or somthing to signal no status, then go green or red or yellow when you know somthing
[31-Aug-2009 16:17:18] <rmatte> you basically saying that you want something that is going to be an indication that you're actually actively monitoring that item?
[31-Aug-2009 16:18:00] <nzle> Yes.
[31-Aug-2009 16:18:45] <rmatte> ok, well, the way you're wanting to do it would involve actually scripting and changing stuff in Zenoss, which, since you seem to be starting out with it, is not an option... but I'll tell you how I'd do it...
[31-Aug-2009 16:19:22] <rmatte> I'd have the script output 0 if everything is ok and 1 if it's not, I would then creat a threshold and name it sometime appropriate, then set the Max value to 1
[31-Aug-2009 16:19:43] <rmatte> so when the script outputs 1, it triggers the threshold
[31-Aug-2009 16:19:49] <rmatte> when it goes back to 0, it clears the threshold
[31-Aug-2009 16:20:13] <rmatte> if the command times out while running, for whatever reason, Zenoss already has something built in which will generate a "Command timed out" event in the event console
[31-Aug-2009 16:20:21] <rmatte> so right away you'll know something's wrong
[31-Aug-2009 16:20:40] <rmatte> and if worst comes to worst and the whole zencommand daemon dies, resulting in no commands running, you'll receive a heartbeat alert for zencommand
[31-Aug-2009 16:21:08] <rmatte> if the entire Zenoss instance dies, at that point you're pretty much screwed no matter what, unless you have some script checking it every few minutes and alerting if it's down
[31-Aug-2009 16:21:58] <rmatte> (and when it clears the threshold, it sends an actual clear event which can be seen in the event history, so you know exactly what time the alert cleared)
[31-Aug-2009 16:24:33] <rmatte> you could even make a threshold called something else that thresholds on 0
[31-Aug-2009 16:24:38] <rmatte> and have it send clear events
[31-Aug-2009 16:24:41] <nzle> Ok, let me kick that around and see what I come up with. I was just assuming that I was missing somthing. Apparently Im not and I just need to adjust my point of view. In Nagios you configured hosts, and services, all the configured services for a host showed up unknown till they got checked and got an OK, Warning, or Critical.
[31-Aug-2009 16:25:13] <rmatte> yeh, forget about Nagios, Zenoss works differently
[31-Aug-2009 16:26:27] <rmatte> but yeh, read up on device templates, thresholds, transforms
[31-Aug-2009 16:26:39] <rmatte> I'd advise reading through the Zenoss Admin Guide in it's entirety
[31-Aug-2009 16:26:42] <rmatte> it's quite good
[31-Aug-2009 16:27:00] <rmatte> it'll be a day's read at least, but it's worth it
[31-Aug-2009 16:27:17] <nzle> I did a while back when I was first starting. Probably time to again since I have some background with it now.
[31-Aug-2009 16:27:28] <mrayzenoss> or even better, come to the free Zenoss training at Ohio LinuxFest
[31-Aug-2009 16:27:42] <rmatte> yeh, if you can make it out to Ohio
[31-Aug-2009 16:27:56] <nzle> Perhaps...Ill have to see if I can sell that idea to the boss.
[31-Aug-2009 16:28:03] <mrayzenoss> nzle: where are you at?
[31-Aug-2009 16:28:09] <nzle> Wisconsin.
[31-Aug-2009 16:28:17] <mrayzenoss> oh man, that's just a 10 hour drive
[31-Aug-2009 16:28:28] <rmatte> lol
[31-Aug-2009 16:28:29] <nzle> $$?
[31-Aug-2009 16:28:33] <mrayzenoss> the training is free
[31-Aug-2009 16:28:57] <mrayzenoss> it's the same material presented in the $$ Zenoss Enterprise training, minus the Enteprise bits and the hands-on lab
[31-Aug-2009 16:29:01] <mrayzenoss> so 1 day instead of 2
[31-Aug-2009 16:29:07] <nzle> When is it?
[31-Aug-2009 16:29:22] <nzle> Where is info online about it?
[31-Aug-2009 16:29:36] <mrayzenoss> just a sec, I'll get the URL
[31-Aug-2009 16:30:26] <rmatte> http://blog.zenoss.com/2009/08/21/zenoss-community-day-september-25th/
[31-Aug-2009 16:30:30] <nzle> found it in the newsletter
[31-Aug-2009 16:30:36] <mrayzenoss> just noticed the link is bad
[31-Aug-2009 16:30:43] <mrayzenoss> http://blog.zenoss.com/2009/08/21/zenoss-community-day-september-25th/
[31-Aug-2009 16:30:47] <mrayzenoss> at least on the blog it's bad
[31-Aug-2009 16:30:55] <rmatte> beatcha to it
[31-Aug-2009 16:31:00] <nzle> got it... thanx. Ill see if I can work it out.
[31-Aug-2009 16:31:10] <mrayzenoss> we did the training at SCALE, it went really well
[31-Aug-2009 16:31:19] <mrayzenoss> 50 people took the 50 spots, people standing in the back
[31-Aug-2009 16:31:33] <mrayzenoss> Ohio has 100 seats
[31-Aug-2009 16:31:37] <rmatte> it's fun stuff to learn hehe
[31-Aug-2009 16:31:40] <rmatte> it's not a boring product
[31-Aug-2009 16:31:52] <mrayzenoss> plus an open bar!
[31-Aug-2009 16:31:56] <nzle> Is there registration, or is it first come first served?
[31-Aug-2009 16:31:58] <rmatte> seriously?
[31-Aug-2009 16:32:03] <rmatte> I'd hit that training so hard
[31-Aug-2009 16:32:06] <rmatte> lol
[31-Aug-2009 16:32:10] <mrayzenoss> 8:00 - Ohio Linux pre-party at Barley’s sponsored by Zenoss
[31-Aug-2009 16:32:14] <nzle> <g>
[31-Aug-2009 16:32:26] <rmatte> "Does everyone have their template ready?" "I'm smashed, can I go home?"
[31-Aug-2009 16:32:42] <mrayzenoss> there's registration
[31-Aug-2009 16:32:55] <mrayzenoss> I'm sure there will be walk-ups, SCALE had them and they stood in the back
[31-Aug-2009 16:33:04] <rmatte> cool
[31-Aug-2009 16:34:52] <nzle> But its not all walkup, one can reserve a seat ahead of time?
[31-Aug-2009 16:35:01] <mrayzenoss> yeah, fixing the link right now
[31-Aug-2009 16:35:49] <mrayzenoss> the first "sign-up" link is good
[31-Aug-2009 16:37:33] <nzle> Cool. I see if I can sell the idea. RU one of the presenters?
[31-Aug-2009 16:37:39] <mrayzenoss> yeah
[31-Aug-2009 16:37:54] <mrayzenoss> npmccallum will handle the training, since he normally teaches it
[31-Aug-2009 16:38:02] <mrayzenoss> I'll be handling the Q&A sections and assisting
[31-Aug-2009 16:38:17] <mrayzenoss> and then on Saturday I'll be manning the booth and doing a demo
[31-Aug-2009 16:39:26] <rmatte> well, I'm out of here, talk to you guys tomorrow, have a good evening/night/morning
[31-Aug-2009 16:39:33] <mrayzenoss> later
[31-Aug-2009 16:39:42] <nzle> Thanx for the advise....
[31-Aug-2009 16:39:46] <rmatte> np
[31-Aug-2009 16:47:36] <nzle> mray, you still here??
[31-Aug-2009 16:48:12] <mrayzenoss> yeah
[31-Aug-2009 17:02:57] <davetoo> oh
[31-Aug-2009 17:03:01] <davetoo> mrayzenoss: you in SF?
[31-Aug-2009 17:03:06] <mrayzenoss> Austin
[31-Aug-2009 17:03:10] <davetoo> oh
[31-Aug-2009 17:03:17] <davetoo> who's at VMWorld?
[31-Aug-2009 17:03:21] <mrayzenoss> I didn't make the VMworld trip, that's mostly Enterprise
[31-Aug-2009 17:03:27] <davetoo> aye
[31-Aug-2009 17:03:34] <mrayzenoss> sales, marketing, director of alliances, mrchippy
[31-Aug-2009 17:03:47] <mrayzenoss> CEO, CMO
[31-Aug-2009 17:03:54] <mrayzenoss> CTO
[31-Aug-2009 17:04:06] <davetoo> I met Bill and Erik at training
[31-Aug-2009 17:04:14] <mrayzenoss> yeah, CEO and CTO are there
[31-Aug-2009 17:04:36] <davetoo> Too many others from my office are already up there, so I have to stay here
[31-Aug-2009 17:04:50] <davetoo> Or I would go ask Erik a couple of followup questions from Friday a week ago.
[31-Aug-2009 17:04:52] <mrayzenoss> well, I'd rather be at Red Hat Summit, which is also right now
[31-Aug-2009 17:05:06] <mrayzenoss> which we're also at and presenting
[31-Aug-2009 17:05:24] <mrayzenoss> so sales and marketing are spread thin
[31-Aug-2009 17:05:33] <mrayzenoss> and some of them are hitting both events
[31-Aug-2009 17:05:39] <davetoo> lots of air travel
[31-Aug-2009 17:05:55] <mrayzenoss> yeah, SF and Chicago
[31-Aug-2009 17:21:54] <davetoo> uh oh, zenpack --delete strikes again
[31-Aug-2009 17:27:21] <davetoo> this is interesting: it deleted the script but left the compiled .pyc
[31-Aug-2009 17:27:34] <davetoo> so it's still working
[31-Aug-2009 17:29:31] <gemineye> Hi mrayzenoss, long time no see.
[31-Aug-2009 17:29:44] <mrayzenoss> hey, how's it going?
[31-Aug-2009 17:30:04] <gemineye> Too bad you're not in SF, that's a 30 min drive from me.
[31-Aug-2009 17:30:23] * davetoo is in sunnyvale
[31-Aug-2009 17:30:29] <gemineye> ditto
[31-Aug-2009 17:30:55] <davetoo> thank god for air conditining and electricity, eh?
[31-Aug-2009 17:31:22] <gemineye> Yeah, but if the building gets too hot, I can always rely on our server rooms to be cool.
[31-Aug-2009 17:31:27] <gemineye> davetoo, you work at netapp?
[31-Aug-2009 17:31:33] <davetoo> oops
[31-Aug-2009 17:31:39] <davetoo> I need a proxy
[31-Aug-2009 17:31:40] <davetoo> yeah
[31-Aug-2009 17:31:43] <mrayzenoss> heh
[31-Aug-2009 17:32:08] <gemineye> I work at Blue Coat
We're looking to refresh all our storage with new NetApp.[31-Aug-2009 17:32:27] <mrayzenoss> maybe you could trade a Blue Coat proxy for a NetApp
[31-Aug-2009 17:32:42] <davetoo> I'm so far away from sales it's ... a good thing, actually
[31-Aug-2009 17:32:45] <gemineye> SURE!!! Not such a good deal for NetApp, but why not
[31-Aug-2009 17:33:11] <davetoo> no, I need a vm in a colo shop or something
[31-Aug-2009 17:33:21] <davetoo> or better yet, get in somebody's onion network
[31-Aug-2009 17:33:25] <gemineye> mrayzenoss, when's the next feature update going to happen to core?
[31-Aug-2009 17:34:09] <mrayzenoss> start of beta is penciled in for next week
[31-Aug-2009 17:34:16] <gemineye> I just have everything at a friends colo, I'm tunneling through SSH outside of Blue Coat since we don't allow irc.
[31-Aug-2009 17:34:33] <gemineye> When is 2.4.5 being released?
[31-Aug-2009 17:34:44] <mrayzenoss> looks like tomorrow morning
[31-Aug-2009 17:37:08] <gemineye> Hey, are there memory leaks in core? Every 4 days I have to stop core, kill off some python processes, and restart it. Every 4 days or so all 4GB of memory is used up and I've only got 181 clients being monitored, 48 of them are just ping tests.
[31-Aug-2009 17:37:39] <mrayzenoss> I know cgibbons is periodically finding them in Windows
[31-Aug-2009 17:37:49] <mrayzenoss> and in some of the wmi stuff
[31-Aug-2009 17:37:53] <gemineye> Mine is running on CentOS 5
[31-Aug-2009 17:38:31] <mrayzenoss> 2.4.2?
[31-Aug-2009 17:38:49] <gemineye> I noticed that zencommand, zenperfsnmp, and zenwin would die due to no memory being left, but today zeneventlog crashed with an error in the log.
[31-Aug-2009 17:38:53] <gemineye> yeah, 2.4.2
[31-Aug-2009 17:39:48] <gemineye> Where's the event log for zeneventlog?
[31-Aug-2009 17:39:58] <davetoo> /opt/zenoss/log
[31-Aug-2009 17:40:11] <gemineye> nope
[31-Aug-2009 17:40:13] <davetoo> hmm
[31-Aug-2009 17:40:28] * davetoo hasn't done any windows monitoring yet
[31-Aug-2009 17:41:05] <gemineye> /usr/local/zenoss/zenoss/log
[31-Aug-2009 17:41:09] <gemineye> stack install
[31-Aug-2009 17:41:25] <mrayzenoss> $ZENHOME/log/
[31-Aug-2009 17:42:02] <cgibbons> there's another memory leak opened in my queue, lemme see what the details are
[31-Aug-2009 17:42:12] <cgibbons> looks like for zenmodeler
[31-Aug-2009 17:42:28] <gemineye> Do these problems exist also in Enterprise?
[31-Aug-2009 17:43:03] <mrayzenoss> Enterprise uses the same zenmodeler, so yeah
[31-Aug-2009 17:43:16] <cgibbons> those wouldn't be different there, but that kind of problem is usually an error handling bug
[31-Aug-2009 17:50:56] <mrayzenoss> later all, heading home
[31-Aug-2009 22:21:18] bzed_ is now known as bzed
[01-Sep-2009 00:00:46] [disconnected at Tue Sep 1 00:00:46 2009]
