[01-Jul-2010 02:07:45] biertie_laptop is now known as biertie
[01-Jul-2010 02:26:39] fOrsberg is now known as forsberg
[01-Jul-2010 06:01:51] * Simon4 bangs his head against net-snmp
[01-Jul-2010 06:07:05] Has anyone installed the ShowGraphPortlet ?
[01-Jul-2010 06:07:24] It doesn't seem to appear as an option for me….even after restarting zenoss
[01-Jul-2010 06:24:23] hmm, I have that installed I htink
[01-Jul-2010 06:24:41] I have that ZenPack installed OK
[01-Jul-2010 06:25:00] Jane_Curry what version of zenoss are you on
[01-Jul-2010 06:25:35] 2.5.2 on SuSE 10.3
[01-Jul-2010 06:25:44] * Simon4 is on 2.5.2 on RHEL
[01-Jul-2010 06:25:59] show graph portlet works here
[01-Jul-2010 06:26:03] The zenpack installed ok…no errors were tossed….but when I go to my dashboard and click Add Portlet …nothing shows
[01-Jul-2010 06:27:46] st3v3o: zenpack --list gives me this:
[01-Jul-2010 06:27:46] ZenPacks.SCC.ShowGraphPortlet (/opt/zenoss/ZenPacks/ZenPacks.SCC.ShowGraphPortlet-1.01-py2.4.egg)
[01-Jul-2010 06:29:46] running a list on my box now …should be pretty much what you run Simon4 …. I"m on Centos with zenoss 2.5.2
[01-Jul-2010 06:30:18] ZenPacks.SCC.ShowGraphPortlet (/vol/opt/zenoss/ZenPacks/ZenPacks.SCC.ShowGraphPortlet-1.01-py2.4.egg)
[01-Jul-2010 06:30:38] yeah same thing ...odd
[01-Jul-2010 06:34:57] Simon4 under Settings/Portlets…does it show up
[01-Jul-2010 07:34:22] I've finally had to upgrade to a new server as my old server died. I have 2.4.5 installed. We also upgraded the OS to a 64 bit from a 32 bit OS. We converted all the RRD files. I installed the 2.4.5 64 bit EL5 rpm and restored the databases and copied over the $ZENHOME directory. Now, zenhub won't stay up. I've set the log to debug but it really doesn't show much of interest - it loads the plugins and then some other notice info, and then Returned
[01-Jul-2010 07:34:22] a connection; Pool size: 1 just before it crashes ... any ideas?
[01-Jul-2010 07:42:17] st3v3o: sorry, ran away. Yes it shows up - I have it on my default dashboard
[01-Jul-2010 07:42:32] odd
[01-Jul-2010 08:03:37] How do you debug zenhub anyway? Especially when there's no traceback or anything in its log
[01-Jul-2010 08:06:22] jmp242: can you pastie or similar the output?
[01-Jul-2010 08:06:45] and also
[01-Jul-2010 08:06:54] what do you get when you run zeoctl status and zopectl status ?
[01-Jul-2010 08:07:28] yup - I'm also just looking at an strace of it run -v10
[01-Jul-2010 08:07:42] it looks like it's basically timing out doing select queries against MySQL
[01-Jul-2010 08:08:10] if I'm reading the output to console correctly lol
[01-Jul-2010 08:09:42] http://pastebin.com/UN8MePrP
[01-Jul-2010 08:09:55] that's the debug log from zenhub on one start request
[01-Jul-2010 08:10:14] ug
[01-Jul-2010 08:10:17] and it's not running after that?
[01-Jul-2010 08:10:22] nope
[01-Jul-2010 08:10:47] zeoctl status is program running
[01-Jul-2010 08:10:52] mysql is running?
[01-Jul-2010 08:11:02] same for zopectl
[01-Jul-2010 08:11:25] if you pull up the webui (you don't need zenhub for that) does it whinge about mysql being down or not able to be connected to?
[01-Jul-2010 08:11:28] mysqld dead but subsys locked
[01-Jul-2010 08:11:35] interesting on mysqld status
[01-Jul-2010 08:11:40] yeah, that won't help
[01-Jul-2010 08:11:43] does look like it's MySQL problem
[01-Jul-2010 08:11:50] as the strace was implying
[01-Jul-2010 08:11:57] ok - I'll go fix that lol
[01-Jul-2010 08:12:21] thanks bunches so far
[01-Jul-2010 08:12:26] np
[01-Jul-2010 08:17:52] Guest94988 is now known as ckrough
[01-Jul-2010 08:24:11] well, it all sounded so good, but I made a mistake and did the check for mysqld as the zenoss user and not as root
[01-Jul-2010 08:24:16] mysql is running
[01-Jul-2010 08:24:30] does each install of zenoss randomize the MySQL password it created?
[01-Jul-2010 08:25:12] i.e. would there be a permission issue conneecting to mysql as zenoss if I re-installed zenoss and restored the database?
[01-Jul-2010 08:29:53] so you can check what zenoss thinks the username and password are
[01-Jul-2010 08:32:59] in /opt/zenoss/etc/hubpasswd?
[01-Jul-2010 08:34:15] nope
[01-Jul-2010 08:34:19] * Simon4 grabs the command
[01-Jul-2010 08:34:37] dmd.ZenEventManager.password
[01-Jul-2010 08:34:52] nm, you are talking mysql, my bad
[01-Jul-2010 08:35:28] oh wait, that is it
[01-Jul-2010 08:35:31] that is is
[01-Jul-2010 08:35:38] echo "print dmd.ZenEventManager.password" | zendmd
[01-Jul-2010 08:35:40] just open zendmd as zenoss user and dmd.ZenEventManager.password should contain it
[01-Jul-2010 08:35:43] or that
[01-Jul-2010 08:35:44] as the zenoss user will get it
[01-Jul-2010 08:35:49] yeah dmd.ZenEventManager.password is right
[01-Jul-2010 08:36:11] and dmd.ZenEventManager.username ..just to verify
[01-Jul-2010 08:36:33] host also, if you want to check that it's localhost
[01-Jul-2010 08:37:03] Simon4 good point
[01-Jul-2010 08:38:23] looks good thanks
[01-Jul-2010 08:48:01] forsberg is now known as fOrsberg
[01-Jul-2010 08:48:27] hmm, manual use of the username + password lets us log into Mysql from the command line
[01-Jul-2010 08:48:36] no luck though with fixing zenhub
[01-Jul-2010 08:48:56] whats the problem?
[01-Jul-2010 08:49:29] zenhub doesn't run
[01-Jul-2010 08:49:37] http://pastebin.com/UN8MePrP
[01-Jul-2010 08:49:46] is the only log from zenhub even with debug logging on
[01-Jul-2010 08:50:03] it seems like it can't talk to mysql, but the uname/pass and host are all correct
[01-Jul-2010 08:50:35] are you connecting to a local instance of mysql ?
[01-Jul-2010 08:50:43] yes
[01-Jul-2010 08:51:00] and does the mysql user have the right permissions
[01-Jul-2010 08:51:16] I mean the user that zenoss uses, not 'mysql'
[01-Jul-2010 08:51:28] yes, the zenoss user can see the tables etc
[01-Jul-2010 08:51:36] also, even when it cant connect, zenhub still runs
[01-Jul-2010 08:52:01] at least it has for me. I've had the sql server completely down and zenoss keeps ticking, just doesnt do event related things
[01-Jul-2010 08:52:13] yea, zenhub isn't runnign
[01-Jul-2010 08:52:36] sucks when things crash but dont log
[01-Jul-2010 08:52:48] strace seems to imply that it's hanging on trying to do select( stuff)
[01-Jul-2010 08:52:59] select(10, [6 7 9], [], [], {0, 990552}) = 0 (Timeout)
[01-Jul-2010 08:53:02] is an example
[01-Jul-2010 08:53:20] does hundreds of those
[01-Jul-2010 08:57:20] I think "lsof -d '6,7,9'" will give you a list of the files being read there
[01-Jul-2010 08:57:45] it's not a mysql select, it's a file system rw, read in this case
[01-Jul-2010 08:58:44] its waiting for fds 6,7, and 9 to become available, may want to look at your iowait in top
[01-Jul-2010 08:58:56] interesting - it's data.fs
[01-Jul-2010 08:59:00] and some log files
[01-Jul-2010 08:59:02] and some temp files
[01-Jul-2010 08:59:10] probably normal, as long as the iowait isnt too high
[01-Jul-2010 08:59:11] this new server shouldn't have IO Wait issues
[01-Jul-2010 08:59:17] its a dual quad core xeon
[01-Jul-2010 08:59:23] and SAS 15k disks
[01-Jul-2010 08:59:33] whats the iowait %?
[01-Jul-2010 09:00:15] would it be under wa?
[01-Jul-2010 09:00:20] correct
[01-Jul-2010 09:00:31] 0.2%
[01-Jul-2010 09:00:58] if it was bad you would see it up near 80%+ for large periods of time
[01-Jul-2010 09:01:14] those selects() are probably normal
[01-Jul-2010 09:01:43] ok
[01-Jul-2010 09:01:57] hmm, I'll try and analyze the strace more
[01-Jul-2010 09:21:52] well, when I strace zenhub
[01-Jul-2010 09:22:22] it actually doesn't look like it's crashing then
[01-Jul-2010 09:22:31] just when running as a daemon from the US
[01-Jul-2010 09:22:33] err UI
[01-Jul-2010 09:22:59] productive morning eh?
[01-Jul-2010 09:23:12] not really lol
[01-Jul-2010 09:23:36] it did keep running ...
[01-Jul-2010 09:28:07] mrayzenoss is now known as mray
[01-Jul-2010 09:28:08] I'm not sure how to track this down more - I don't suppose trying to update might overwrite the broken parts or if it would make it worse
[01-Jul-2010 09:29:50] that sounds scary
[01-Jul-2010 09:30:09] so when you start it normally it dies, but when you run in debug it stays up?
[01-Jul-2010 09:30:20] oh on
[01-Jul-2010 09:30:30] when I run strace zenhub -v10
[01-Jul-2010 09:30:53] it seems to keep going as far as I can tell, but if I start the daemon - which is set at debug logging - it crashes
[01-Jul-2010 09:31:34] I'm going to try without strace
[01-Jul-2010 09:33:41] Ive seen this before
[01-Jul-2010 09:33:46] but I dont remember what it was
[01-Jul-2010 09:34:49] running in the foreground does seem to keep working oddly
[01-Jul-2010 09:34:53] which I do also recall
[01-Jul-2010 09:34:56] seeing in the forums
[01-Jul-2010 09:35:06] of course, with it running, now other daemons are crashing :(
[01-Jul-2010 09:36:13] What's the last message that you get from zenhub before it dies?
[01-Jul-2010 09:39:37] see my pastebin http://pastebin.com/UN8MePrP
[01-Jul-2010 09:39:44] that's all I get for each run
[01-Jul-2010 09:40:05] and the process actually exits?
[01-Jul-2010 09:40:20] as far as I can tell - I get the red icon in the UI
[01-Jul-2010 09:40:27] and I can't find it in a ps -aux
[01-Jul-2010 09:42:14] oddly enough, it does seem to keep running when run in the foreground even though the UI still shows red :?
[01-Jul-2010 09:42:40] though with it running in the foreground - other daemons keep crashing
[01-Jul-2010 09:43:43] fOrsberg is now known as forsberg
[01-Jul-2010 09:43:47] zenperfsnmp , zencommand, zenprocess
[01-Jul-2010 09:43:51] specifically
[01-Jul-2010 09:44:16] they are giving me
[01-Jul-2010 09:45:06] http://pastebin.com/zMZ1siVz
[01-Jul-2010 09:45:54] but I checked and corrected the permissions in the filesystem
[01-Jul-2010 09:47:24] Wow, the only thing that I can think of is something weird like that is selinux
[01-Jul-2010 09:47:47] mmm, this is a new install of EL5.4_64
[01-Jul-2010 09:47:51] I'll check
[01-Jul-2010 09:48:18] Yeah, there have been some 'updates', so a couple of new features have been added, IIRC
[01-Jul-2010 09:50:09] the old server was EL5.4 but 32 bit
[01-Jul-2010 09:55:22] I've got a uniform fleet of 25 systems monitored through Zenoss and 4 of the 25 are showing 900% utilization of the /boot file system. I've checked the systems and the /boot file systems are not utilized above 15%. I've also checked snmpwalk output for the systems and the numbers reported there are correct.
[01-Jul-2010 09:55:45] Seems like something is getting screwed up in Zenoss for these 4 systems and I'm not sure where to poke around. Anyone have any suggestions?
[01-Jul-2010 10:00:08] I will suggest to make a backup copy of the RRD files, then blow them away in the $ZENHOME/perf/Devices/ directory and see if the RRD files are a bit off
[01-Jul-2010 10:01:04] Welcome to another Zenoss Developer IRC Session, kells is on deck today and other Zenoss devs are in the channel
[01-Jul-2010 10:01:17] The model might also be messed up
[01-Jul-2010 10:01:25] Oh! SNMP, right
[01-Jul-2010 10:01:53] yes, snmp
[01-Jul-2010 10:02:15] I don't really care about the data that's been collected so far, so I can just delete it all without backing it up.
[01-Jul-2010 10:02:53] SNMP data collection works strictly off of index numbers, btw. So whatever the index was determined to be from the last model is what zenoss thinks is going on. So the model (ie what is the max size) can be different than what zenpersnmp collects (ie the entry and snmp index X is pointing to a different filesystem)
[01-Jul-2010 10:03:20] So hence you can get strange numbers occasionally from the SNMP agent if the model and the collection part are out of sync
[01-Jul-2010 10:03:31] Try running zenmodeler again first
[01-Jul-2010 10:03:39] I've re-modeled the system several times
[01-Jul-2010 10:04:41] Check the snmp indxe for the boot filesystem, and then compare it against the table order to make sure that what the device says is at /boot and then see what zenperfsnmp run -v10 -d device --showdeviceresults says is coming back at that OID
[01-Jul-2010 10:05:36] I'm making the somewhat safe assumption that /boot isn't insanely huge (like 3TB) and is an easily representable 32-bit number
[01-Jul-2010 10:06:22] Yes, that's correct, /boot is ~200MB
[01-Jul-2010 10:06:30] Let me try that last bit and get back to you
[01-Jul-2010 10:09:35] Simon4_ is now known as Simon4
[01-Jul-2010 10:09:52] hello
[01-Jul-2010 10:13:08] greetings
[01-Jul-2010 10:14:17] hello
[01-Jul-2010 10:15:53] hello
[01-Jul-2010 10:17:11] Hey mray did you ever grab that aggregate pdu zenpack?
[01-Jul-2010 10:17:33] nickanderson: I did, it's #1 on my list of zenpacks to publish today
[01-Jul-2010 10:17:44] my laptop died Wednesday of last week when I was at Velocity
[01-Jul-2010 10:17:57] cool, ive been busy so haven't had time to think about it
[01-Jul-2010 10:18:00] suck
[01-Jul-2010 10:18:04] so I was out of commission until yesterday
[01-Jul-2010 10:18:13] today is zenpacks, zenpacks, zenpacks
[01-Jul-2010 10:19:23] speaking of zenpacks, does anyone know of a zenpack for ssh collection for a free bsd host
[01-Jul-2010 10:19:24] Aggregate PDU, 64-bit MSSQL, HP Printer Trap Transforms, SNMP Location, MySQL SSH, SQL Data Source, Zeus Load Balancers, Open Gear MIBs, Open Gear Console Server and MSMQ
[01-Jul-2010 10:19:38] bah, thats a lot of zenpacks :)
[01-Jul-2010 10:19:49] * Simon4 won't add to that list today ;)
[01-Jul-2010 10:19:57] nickanderson: plus a few updates: memcached and Cfengine
[01-Jul-2010 10:20:03] Simon4: please send more
[01-Jul-2010 10:20:21] mray: I need to finish doco, probably get it done tongith after work
[01-Jul-2010 10:20:27] * nickanderson is ready for the ui refresh
[01-Jul-2010 10:20:33] jdross: my goal with my half-finished OS X ZenPack tutorial series was that someone would pick that up and do FreeBSD
[01-Jul-2010 10:20:49] jdross: docs/DOC-5857
[01-Jul-2010 10:20:59] jdross: docs/DOC-5865
[01-Jul-2010 10:22:28] question for the Devs re: zenpacks... is it possible for a custom command parser to update more than one component on a device?
[01-Jul-2010 10:22:56] Ill have to look through that, because I was considering writing it myself anyway since i couldnt find one
[01-Jul-2010 10:23:09] jdross: I'm happy to answer questions
[01-Jul-2010 10:23:33] currently from my limited experience it seems you need a command per component, which is a bit wasteful when the command in question returns the data for all components anyway
[01-Jul-2010 10:24:17] kells?
[01-Jul-2010 10:24:36] That's currently a limitation of the parser format atm :(
[01-Jul-2010 10:24:53] If we're talking about subcomponents, the story is a little different, of course
[01-Jul-2010 10:25:38] But subcomponents are, ahem, in flux
[01-Jul-2010 10:25:46] k
[01-Jul-2010 10:25:58] well at least I know I'm doing it as well as I can currently - cheers :)
[01-Jul-2010 10:34:26] folks, we have 2.4.2 and are trying to inject REST events, has this interface broken?
[01-Jul-2010 10:34:44] Oh dear.....
[01-Jul-2010 10:35:28] There's a few 'differences' between 2.3, 2.4.2, 2.4.5 and 2.5.x wrt REST, IIRC
[01-Jul-2010 10:36:03] What is the error message?
[01-Jul-2010 10:36:12] FWIW - I think I fixed my problem
[01-Jul-2010 10:36:14] As well as an example REST statement
[01-Jul-2010 10:36:23] Oh? What was it?
[01-Jul-2010 10:36:25] it turned out to be fileystem permissions after all
[01-Jul-2010 10:36:33] Doh!
[01-Jul-2010 10:36:34] our change from 32 bit to 64 bit
[01-Jul-2010 10:36:41] zensocket?
[01-Jul-2010 10:36:42] changed /opt/zenoss/perf to root root
[01-Jul-2010 10:36:48] DOH!
[01-Jul-2010 10:36:51] rahter than root zenoss lol
[01-Jul-2010 10:36:58] so no write access then lol
[01-Jul-2010 10:37:01] kells: being directed to the login page.
[01-Jul-2010 10:37:54] sorry my bad, we are on 2.4.5 until this coming weekend
[01-Jul-2010 10:37:57] The username/password in the REST are correct, right?
[01-Jul-2010 10:38:05] yeah I can log in using them.
[01-Jul-2010 10:39:36] In case you missed it, we beta #4 was released yesterday
[01-Jul-2010 10:39:56] and… we put out a press release at Cisco Live announcing 3.0 will be released in July
[01-Jul-2010 10:40:02] mray: what's new in 3.0 for the time impaired?
[01-Jul-2010 10:40:11] malbon: more unicorns
[01-Jul-2010 10:40:12] new UI
[01-Jul-2010 10:40:48] mray: ok, that sounds like it is worth having.
[01-Jul-2010 10:40:56] malbon: for enterprise: http://www.zenoss.com/about/news/Zenoss_Introduces_Dynamic_Service_Assurance_Vision_with_Release_of_Enterprise_3_0.html
[01-Jul-2010 10:41:13] for Core: new UI, gobs of bug fixes
[01-Jul-2010 10:41:51] Python 2.6, Zope 2.12
[01-Jul-2010 10:41:57] malborn: 2.4.5 had something weird about it
[01-Jul-2010 10:41:58] documented REST API :)
[01-Jul-2010 10:42:09] mray: funny I read that yesterday. I found it content-lite
[01-Jul-2010 10:42:30] malbon: let me throw up a screenshot of what they were talking about
[01-Jul-2010 10:43:34] mray: sounds good.
[01-Jul-2010 10:43:36] http://imagebin.org/103685
[01-Jul-2010 10:44:24] I'll open a case for my REST issue
[01-Jul-2010 10:44:43] mray: yes, ok, but that is only for the Cisco stack right?
[01-Jul-2010 10:44:48] malbon: nope
[01-Jul-2010 10:44:55] mray: no?
[01-Jul-2010 10:45:05] right now it's Cisco, VMware, NetApp, Brocade so far
[01-Jul-2010 10:45:22] if you're not using Cisco, you won't see that layer
[01-Jul-2010 10:45:23] mray: no IBM or HP?
[01-Jul-2010 10:45:30] malbon: slowly getting to them
[01-Jul-2010 10:45:46] * ckrough drool
[01-Jul-2010 10:45:54] mray: ok, fair enough.
[01-Jul-2010 10:46:07] malborn: if you're moving from 2.4.5 to 2.5.2, then you might want to wait on the REST issue
[01-Jul-2010 10:46:14] since you're on 2.4, in 2.5.2 they added the Data Center View: http://imagebin.org/103686 and http://imagebin.org/103687
[01-Jul-2010 10:46:27] kells: yeah quite possibly.
[01-Jul-2010 10:47:07] mray: actually I think that bit is quite neat.
[01-Jul-2010 10:47:26] mray: documents stuff nicely. now only if Zenoss could organise patching lists.
[01-Jul-2010 10:47:43] mray: Egads man! The VMware - NetApp link barely started working yesterday afternoon :)
[01-Jul-2010 10:47:57] kells: I'm just letting people know you're busy :p
[01-Jul-2010 10:48:12] AFAIK, Brocade won't have the link support
[01-Jul-2010 10:48:26] there are a ton of other fixes and enhancements across the board, for Core and Enterprise
[01-Jul-2010 10:48:41] kells: work harder.... :P
[01-Jul-2010 10:48:46] There's a slightly updated Brocade UI, however
[01-Jul-2010 10:49:18] Heh. We just need some FC stuff to test against and things would go a bit better.
[01-Jul-2010 10:49:47] kells: contact me by email. I can provide some test fibre stuff, and potentially snmp walks of our live.
[01-Jul-2010 10:50:01] Sweet! Thanks!
[01-Jul-2010 10:50:06] kells: via webex or equiv if required.
[01-Jul-2010 10:50:09] kells: I found the index for the /boot filesystem and ran that command you gave me, but I didn't get much output from it. Is that normal? I definitely didn't see any OID type information.
[01-Jul-2010 10:51:08] ad4m: Sometimes zenperfsnmp doesn't catch the configuration info before the cycle starts, so you need to run zenperfsnmp with -c and then kill it after it collects some perf data
[01-Jul-2010 10:51:17] kells: actually I'll check with our 24 hour guys if they mind setting up sessions because are you in EST or another US timezone?
[01-Jul-2010 10:52:12] EST or EDT (I can never remember :))
[01-Jul-2010 10:52:23] eastern seaboard?
[01-Jul-2010 10:53:05] Toronto
[01-Jul-2010 10:53:20] ok, cool. that is doable then, GMT+5
[01-Jul-2010 10:56:00] kells: I'm actually seeing a log line that might indicate a problem loading configuration for the device... WARNING zen.zenperfsnmp: Error loading config for devices ['web10']
[01-Jul-2010 10:56:02] good morning
[01-Jul-2010 10:56:15] hi guys, is a virtualbox appliance coming for beta 3.0?
[01-Jul-2010 10:56:26] I setup Zenoss last night because i was tired of wasting my time trying to get zabbix up and running :)
[01-Jul-2010 10:56:39] I could use a little direction if thats alright
[01-Jul-2010 10:56:47] I grabbed a WMI zenpack and got that installed
[01-Jul-2010 10:57:08] kells: Is that normal or something to work out?
[01-Jul-2010 10:57:13] ad4m: Is there a corresponding error in the zenhub.log file around the same time? Probably a traceback?
[01-Jul-2010 10:57:18] ali3n0: no, there will be the usual VMware appliance, but that's the only virtual appliance provided by Zenoss
[01-Jul-2010 10:57:25] i'm not really sure what to do now, is there a nice setup guide?
[01-Jul-2010 10:57:44] Parabola: community/documentation has the Getting Started Guide
[01-Jul-2010 10:57:46] mray, ah ok, tnx anyway
[01-Jul-2010 10:57:48] my end goal is to monitor windows services
[01-Jul-2010 10:57:51] can zenoss do taht?
[01-Jul-2010 10:57:56] i have not seen an agent
[01-Jul-2010 10:58:04] Parabola: yes, out of the box
[01-Jul-2010 10:58:12] oh okay thank you
[01-Jul-2010 10:58:14] mray, can I ask you why you chose vm instead of vbox?
[01-Jul-2010 10:58:15] Parabola: docs/DOC-2606 has the Getting Started Webinar video
[01-Jul-2010 10:58:33] if i have any stupdi questiosn may i ask here?
[01-Jul-2010 10:58:55] ad4m: I'm also assuming that web10 is the box in question, right?
[01-Jul-2010 10:58:55] ali3n0: not specifically my choice, but I'm pretty sure that's what rPath supported
[01-Jul-2010 10:59:10] ali3n0: and since VMware Player is free on Windows, that's what we targetted
[01-Jul-2010 10:59:17] kells: Yes, one of the 4 boxes
[01-Jul-2010 10:59:32] mray, I was just curious
[01-Jul-2010 10:59:56] ali3n0: we may be revisiting the vmware appliance after 3.0 is released
[01-Jul-2010 11:00:05] you know thats the second time in a week I have heard "Was setting up zabbix, got frustrated and installed Zenoss"
[01-Jul-2010 11:00:06] I'm on linux and I'd like to try zenoss out so it would be good for me
[01-Jul-2010 11:00:23] ali3n0: I believe there's a VMware Player for Linux
[01-Jul-2010 11:00:25] ali3n0: there is vmware player for linux as well
[01-Jul-2010 11:00:27] there is
[01-Jul-2010 11:00:33] uh, ok I'll try it out
[01-Jul-2010 11:00:35] i know what he means though
[01-Jul-2010 11:00:41] i like vbox better too
[01-Jul-2010 11:00:42] ali3n0: plus the stack installer is really simple
[01-Jul-2010 11:00:53] yeh i use vbox
[01-Jul-2010 11:00:55] so you could just install it on a VM on vbox
[01-Jul-2010 11:00:56] I use to prefer opensource stuff so I always tried vbox before
[01-Jul-2010 11:01:21] ok I'll try it out right now, thanks
[01-Jul-2010 11:01:23] ali3n0: i think you can convert the image to vbox as well
[01-Jul-2010 11:01:34] when you say 3.0 beta, are you speaking of a zenoss beta, if so should i use that instead of the stable?
[01-Jul-2010 11:01:40] Parabola: community/about/webinars is the Getting Started Webinar, next one is July 13
[01-Jul-2010 11:01:47] nickanderson, going to know in a sec ;)
[01-Jul-2010 11:02:14] Parabola: are you planning on going into production this month?
[01-Jul-2010 11:02:15] uh, July 13! cool I have to bookmark this
[01-Jul-2010 11:02:21] mray yep
[01-Jul-2010 11:02:27] this week
[01-Jul-2010 11:02:28] lol
[01-Jul-2010 11:02:35] Parabola: 3.0 is currently in beta, but will be officially released this month
[01-Jul-2010 11:02:35] today if i can get it done
[01-Jul-2010 11:02:58] kells: arranged with the 24 hour guys, early next week ok?
[01-Jul-2010 11:04:01] mray, july 13: will you show beta3?
[01-Jul-2010 11:04:24] ali3n0: hmmm… I typically do the webinars with the current stable release
[01-Jul-2010 11:04:48] ok. last one was the Baltimore one, right?
[01-Jul-2010 11:05:23] ali3n0: the webinars are every 2 weeks. The next free day of training is at Ohio LinuxFest
[01-Jul-2010 11:05:53] mray, ah, sorry, I was thinking about the free video tutorial I've found on the community site
[01-Jul-2010 11:06:10] malborn: Might have to bump out to a little later -- we have some scheduled meetings at the beginning (after the Zenoss Monday holiday)
[01-Jul-2010 11:06:11] yeah, that was from LISA in Baltimore last November
[01-Jul-2010 11:07:12] any chance to get Ohio LinuxFest videos?
[01-Jul-2010 11:07:15] m4adm: That might mean that it might be fastest to remove the device and re-add it (ie web10)
[01-Jul-2010 11:07:24] kells: when ever suits.
[01-Jul-2010 11:07:41] ali3n0: I'll look into it, we should have updated training materials and it'll be over the 3.0 UI by then (September)
[01-Jul-2010 11:07:50] Coolio. I have to check with one other guy here to line up some time. Thanks again!
[01-Jul-2010 11:08:00] mray, great. the LISA one is very very good stuff for a beginner
[01-Jul-2010 11:08:29] never told you thanks enough for the great community effort you do
[01-Jul-2010 11:08:52] thanks!
[01-Jul-2010 11:09:01] hope your project will be inspiring for other OSS projects too
[01-Jul-2010 11:09:20] I'm pretty sure it is already for a lot of coders out there
[01-Jul-2010 11:09:28] looking forward to LISA this year, back in San Jose, CA
[01-Jul-2010 11:09:33] odd
[01-Jul-2010 11:09:50] when i installed zenoss, the only method of data collection i have is SNMP
[01-Jul-2010 11:09:59] and i dont think thats working
[01-Jul-2010 11:10:05] its just using ping right now on the two hosts i added
[01-Jul-2010 11:10:20] Parabola: make sure your devices are in the right device class
[01-Jul-2010 11:10:20] kells: I'm not seeing anything in the zenhub.log for times when I ran zenperfsnmp. I've got to run now though, so thanks for your help. If you're around when I get back I'll catch up with you then.
[01-Jul-2010 11:10:31] mray: they are
[01-Jul-2010 11:10:33] one is windows
[01-Jul-2010 11:10:37] another is linux, which is accurate
[01-Jul-2010 11:10:48] Parabola: and you've got the right credentials in place and everything?
[01-Jul-2010 11:11:00] mray, talking about inspiration, I was looking for a new language to learn and I chose python, 99% because of zenoss is
[01-Jul-2010 11:11:03] creds? no
[01-Jul-2010 11:11:04] Parabola: service monitoring requires WMI creds
[01-Jul-2010 11:11:11] i figured as much
[01-Jul-2010 11:11:13] it never prompted
[01-Jul-2010 11:11:16] maybe not 99% but you get the point :)
[01-Jul-2010 11:12:13] this is a little much
[01-Jul-2010 11:12:29] what sectiosn do i need to read to get this working, i don thave time to read a 190 page PDF right now :(
[01-Jul-2010 11:12:40] i need to get this up and running, im on page like, 16 lol
[01-Jul-2010 11:14:35] im under the status of the windows box
[01-Jul-2010 11:14:39] i click edit, theres lots of shit
[01-Jul-2010 11:14:47] nothing about credentials
[01-Jul-2010 11:14:57] and most of this this product should be able to detect on its own anyway
[01-Jul-2010 11:15:30] Okay, I'm back to the mines! Later
[01-Jul-2010 11:18:40] lets try this from another angle, if i give you guys money, will i be able to get some of this resolve without me reading 200 pages
[01-Jul-2010 11:18:45] everyone likes money right
[01-Jul-2010 11:20:16] this is zgay
[01-Jul-2010 11:20:43] what is the point of putting a z in front of everything
[01-Jul-2010 11:21:13] Parabola: unfortunately the only monitoring system that requires you to do nothing is the one where you hire someone else to do it
[01-Jul-2010 11:21:21] honestly thats not true
[01-Jul-2010 11:21:30] ip monitor, now owned by solarwinds does
[01-Jul-2010 11:21:46] you scan ranges or a subnets
[01-Jul-2010 11:22:01] thats it?
[01-Jul-2010 11:22:03] nice
[01-Jul-2010 11:22:03] it detects, and will automatically add everything, or give you a list of all services..ect on the box
[01-Jul-2010 11:22:08] and you can select which to monitor
[01-Jul-2010 11:22:14] i highly doubt it though
[01-Jul-2010 11:22:15] yeah its only 2 grand too, im about to buy it
[01-Jul-2010 11:22:19] well i dont care what you doubt
[01-Jul-2010 11:22:20] i've used it
[01-Jul-2010 11:22:24] ok
[01-Jul-2010 11:22:27] had over 300 servers in it
[01-Jul-2010 11:22:32] was the easiest thing i've ever setup
[01-Jul-2010 11:23:10] only PITA with it was renaming the servers, since they showed up by IP
[01-Jul-2010 11:23:28] i'm hoping i missing somthing when i installed zenoss
[01-Jul-2010 11:23:38] because this isnt doing any of the shit its supposed too
[01-Jul-2010 11:23:56] minus all the stupidass z's
[01-Jul-2010 11:24:00] looks like IP Monitor requires you to enter your credentials as well
[01-Jul-2010 11:24:03] i see command protocol is set to SSH on a windows server
[01-Jul-2010 11:24:10] mray yeah it does, easy on a domain
[01-Jul-2010 11:24:27] zCommandUsername
[01-Jul-2010 11:24:36] what the fuck does that mean
[01-Jul-2010 11:24:43] zenablepassword isnt a yes or no
[01-Jul-2010 11:24:47] its an edit field, do i type yes?
[01-Jul-2010 11:25:39] i know this is probobly a very nice product
[01-Jul-2010 11:25:45] well, the last 30 minutes you could have read at least 10 pages of the manual but you preferred to write stuff here, I guess you're not such in a rush...
[01-Jul-2010 11:25:46] it looks well written, thats why i assume im missing somthing
[01-Jul-2010 11:26:01] zWinUser and zWinPassword for WMI monitoring
[01-Jul-2010 11:26:03] no i'm just going through the webGUI and trying to make it work
[01-Jul-2010 11:26:05] Parabola: check for zWinUser and zWinPassword
[01-Jul-2010 11:26:09] gah, beaten :)
[01-Jul-2010 11:26:12] yeah got it filled in
[01-Jul-2010 11:26:24] remodel your device and see what shows up
[01-Jul-2010 11:26:31] remodel? k
[01-Jul-2010 11:26:38] how do i uh do that
[01-Jul-2010 11:26:54] nvm
[01-Jul-2010 11:26:57] model device i assume
[01-Jul-2010 11:27:15] is there a way for this to not be done one by one?
[01-Jul-2010 11:27:22] ie, it knows its windows so it asks for credentials?
[01-Jul-2010 11:27:24] yup, just set it on the device class
[01-Jul-2010 11:27:33] dare i ask why its not already there
[01-Jul-2010 11:27:40] 1 ERROR zen.ZenModeler Timeout connecting to zenhub: is it running?
[01-Jul-2010 11:27:42] Huston
[01-Jul-2010 11:28:36] cool
[01-Jul-2010 11:28:38] somthing killed zenhub
[01-Jul-2010 11:29:15] what OS are you using and how did you install?
[01-Jul-2010 11:29:27] ubuntu 10.04 x64, installed from the DL link, a .bin file
[01-Jul-2010 11:30:06] its not seeing any interfaces
[01-Jul-2010 11:30:11] no processes. nothing :(
[01-Jul-2010 11:30:58] are there some requirements im missing
[01-Jul-2010 11:31:08] or does it just not support server 08
[01-Jul-2010 11:31:59] under device type, all i have is SNMP
[01-Jul-2010 11:32:04] when i go to add a server
[01-Jul-2010 11:33:36] Parabola: navigate to /Devices/Server/Windows, click on the "zProperties" tab and enter in the WMI username and password. Save. Then click the downarrow to the left of the "Classes" tab and select manage->add device. Type in the name of the server you're adding and hit okay
[01-Jul-2010 11:34:08] I know people are monitoring 2008 with no problems
[01-Jul-2010 11:34:21] I don't happen to monitor Windows myself
[01-Jul-2010 11:34:53] and what about domains?
[01-Jul-2010 11:34:56] There is a known WMI leak on Server 2008, fixed in SP1
[01-Jul-2010 11:34:57] * Simon4 doesn't either, but the instructions above make sense from what other people have said in here previously
[01-Jul-2010 11:35:00] its not a local login
[01-Jul-2010 11:35:20] lets see if it likes \
[01-Jul-2010 11:35:23] Enter your winuser as domainname\username
[01-Jul-2010 11:35:24] * Simon4 would guess the username should be DOMAINuser then
[01-Jul-2010 11:35:34] * mray is off to lunch
[01-Jul-2010 11:35:51] someone has to tell me why you put a z in front of everything
[01-Jul-2010 11:35:53] it doesnt look cool
[01-Jul-2010 11:35:56] and it makes it harder to read
[01-Jul-2010 11:36:15] i cant think of a single logical reason to do that in a UI
[01-Jul-2010 11:36:25] sorry, that question is above my pay grade
[01-Jul-2010 11:36:32] its zRetarded
[01-Jul-2010 11:37:52] Many of the 'z's have been dropped from the UI in 3.0
[01-Jul-2010 11:38:06] the fact that someone saw that and went "well gee this looks good"
[01-Jul-2010 11:38:10] is just brain boggling
[01-Jul-2010 11:38:27] oh ffs
[01-Jul-2010 11:38:30] snmp isnt on on this server
[01-Jul-2010 11:38:33] eargpkhjadsfagdsljh4fjdjnSlR
[01-Jul-2010 11:39:22] I am pretty new here, but I'm guessing that the initial 'z' was used to distinguish Zenoss-specific items in the Zope DB, and the field names bled out into the UI
[01-Jul-2010 11:42:13] zawesome
[01-Jul-2010 11:44:03] so i saved my wmi creds
[01-Jul-2010 11:44:05] i go back to that window
[01-Jul-2010 11:44:08] the fields are blank
[01-Jul-2010 11:44:10] that normal?
[01-Jul-2010 11:44:45] nvm its class specific
[01-Jul-2010 11:44:54] Did you define them at the device level, or ... oks
[01-Jul-2010 11:45:46] great
[01-Jul-2010 11:45:51] snmp setup and it sitll times out
[01-Jul-2010 11:47:47] And this is the /Server/Windows class, yes?
[01-Jul-2010 11:48:15] Any helpful events in the Event Console?
[01-Jul-2010 11:49:05] And of course, even though you set up Win credentials, SNMP still requires the community string too
[01-Jul-2010 11:51:26] snmp service installed and started
[01-Jul-2010 11:51:29] windows firewall off
[01-Jul-2010 11:51:38] still times out
[01-Jul-2010 11:53:10] Anything in the Zenoss Event Console regarding this device?
[01-Jul-2010 11:53:35] ah ha
[01-Jul-2010 11:53:40] * pmcguire is hoping there is more than just "connection timed out" message
[01-Jul-2010 11:53:43] after installing, the traps wont show up
[01-Jul-2010 11:53:45] :)
[01-Jul-2010 11:53:48] have to logout and back in lol
[01-Jul-2010 11:53:55] then the traps tab appears
[01-Jul-2010 11:56:08] Hmmm, I'm not too well-versed in the old UI, I just started recently so my main exposure has been just the new 3.0 UI in the development trunk - so I'm not sure what you mean by the Traps tab, or whether this particular refresh behavior is better in 3.0 or not, but I'm glad to see you put up a smiley, that sounds like *some* progress
[01-Jul-2010 11:56:36] nope, in windows
[01-Jul-2010 11:56:38] not zenoss
[01-Jul-2010 11:56:54] yeah i personally dont mind spending a month figuring somthing out
[01-Jul-2010 11:57:00] but the client needs it yesterweek
[01-Jul-2010 11:57:03] blah blah blah
[01-Jul-2010 11:57:25] boy, if it weren't for these pesky clients, we could really get some work done!
[01-Jul-2010 11:57:35] HW Model
[01-Jul-2010 11:57:35] .1.3.6.1.4.1.311.1.1.3.1.3
[01-Jul-2010 11:57:37] lmao
[01-Jul-2010 11:57:39] nice
[01-Jul-2010 11:57:56] alright good got it working now, somewhat
[01-Jul-2010 11:57:58] it cant see drive space
[01-Jul-2010 11:57:59] wtf
[01-Jul-2010 11:58:19] does zenoss not support esxi vms?
[01-Jul-2010 11:59:04] I believe there's a zenpack to talk to the esxi management interface
[01-Jul-2010 11:59:11] check on the site under "ZenPacks"
[01-Jul-2010 11:59:28] alright
[01-Jul-2010 11:59:43] oh man, this cant see the services running on the machine?
[01-Jul-2010 11:59:48] guys, I'm a bit confused by the beta 3.0 page (docs/DOC-5854). Available versions are 2.5.70-743 for every downloadable
[01-Jul-2010 12:00:05] ali3n0: that's v3 beta (the newest one)
[01-Jul-2010 12:00:13] go figure, etc, but it is :)
[01-Jul-2010 12:00:37] Simon4, ok cool thanks. I wasn't sure so...
[01-Jul-2010 12:01:02] if zenoss cant see what services are running on a windows box, this wont work :(
[01-Jul-2010 12:02:40] Parabola: I believe it can, then you set up the ones you care about under /Services/Winservice
[01-Jul-2010 12:03:20] oh good
[01-Jul-2010 12:04:05] well i see 197 services listed in there
[01-Jul-2010 12:04:12] however under the server its self i cant see services
[01-Jul-2010 12:04:13] see the "count" on the right hand side
[01-Jul-2010 12:04:25] Parabola: I'm off to lunch, I leave you in Simon4's capable hands
[01-Jul-2010 12:04:41] when you model the device you should see the count increase for any services that were on that server
[01-Jul-2010 12:04:52] yeah i see
[01-Jul-2010 12:04:54] i think i got it
[01-Jul-2010 12:05:09] if you then set that service's monitor value to "true" it'll appear under the OS tab for the device (on a good day, and if my guesses are correct based on how linux boxes work :D )
[01-Jul-2010 12:05:19] you may need to remodel it to pick it up
[01-Jul-2010 12:06:06] lol
[01-Jul-2010 12:06:21] basic principle = 1) set up services you're interested in, enable monitoring, 2) Zenmodeler runs twice a day (by default) and looks at what is on your servers, compares it to enabled checks and then monitors those services that are enabled.
[01-Jul-2010 12:06:36] yeah enable monitoring, how
[01-Jul-2010 12:06:45] i have like 6 services selected right now
[01-Jul-2010 12:06:50] i dont see anything in the context menu
[01-Jul-2010 12:07:11] ah, click on the service name, go to Edit, change Monitor to True
[01-Jul-2010 12:08:37] gotcha
[01-Jul-2010 12:11:59] I seem to be having an issue with zenwin... I deleted a device, but am still getting /Status/Winservice events in the console for the deleted device every 10 minutes or so.
[01-Jul-2010 12:12:18] any ideas how I can nuke it :o
[01-Jul-2010 12:15:30] restart the zenwin daemon
[01-Jul-2010 12:15:42] Tried that a few times
[01-Jul-2010 12:16:03] also tried clearing all heartbeats from the event manager
[01-Jul-2010 12:16:56] * Simon4 has to head home - it's dinner time
[01-Jul-2010 12:17:16] Parabola: good luck - others should be kicking around in here soon enough to help out etc
[01-Jul-2010 12:30:04] so
[01-Jul-2010 12:30:16] under easy add
[01-Jul-2010 12:30:22] my device type dropdown is now empty
[01-Jul-2010 12:30:28] it had 3 snmp entries in there 15min ago
[01-Jul-2010 12:30:33] all daemons are running
[01-Jul-2010 12:37:38] Is it possible to configure the default dashboard layout?
[01-Jul-2010 13:08:08] Trying again if anyone is awake:
[01-Jul-2010 13:08:10] I seem to be having an issue with zenwin... I deleted a device, but am still getting /Status/Winservice events in the console for the deleted device every 10 minutes or so.
[01-Jul-2010 13:22:32] Is anyone user the Websensor zenpack …and has created templates ?
[01-Jul-2010 13:23:03] or graphs in the templates rather
[01-Jul-2010 13:29:50] * gwb2351 is tilting at the "Manually adding an Ip Service to a device never showing matches or activating the OK button" windmill.
[01-Jul-2010 13:33:52] Is there a way to set a memory threshold via a percentage ?
[01-Jul-2010 13:40:53] st3c3o: i have tested WMIDataSource from SVN against Zenoss 2.5.2 and 3.0 both works for me.
[01-Jul-2010 13:41:39] dang…any suggestions on trying to hunt down the gremlins in my stuff ?
[01-Jul-2010 13:41:58] st3v3o: yes, I just don't know exactly what variables you need. The example in the admin guide (6.2.7) uses "here.speed * .85/8" for an interface threshold
[01-Jul-2010 13:42:18] so it seems you should be able to get the memory similarly
[01-Jul-2010 13:42:28] st3v3o: zenoss restart?
[01-Jul-2010 13:42:35] multiple times
[01-Jul-2010 13:42:43] even completely bounced the box
[01-Jul-2010 13:43:04] mray ..I'm just looking to set a threshold for memory at 90%
[01-Jul-2010 13:45:36] st3v3o: try to install WMIDataSource 2.5.73, i will post it on my zenoss blog soon.
[01-Jul-2010 13:45:53] ok will be waiting
[01-Jul-2010 13:46:04] thanks for the quick replys bigegor_
[01-Jul-2010 13:46:23] st3v3o: np
[01-Jul-2010 13:47:29] st3v3o: aren't you measuring memory by a percentage?
[01-Jul-2010 13:49:42] memAvailReal_memAvailReal….is the number in bytes if I'm not mistaken
[01-Jul-2010 13:53:51] st3v3o: people/bigegor/blog/2010/06/30/some-more-zenpacks-for-zenoss-30-beta
[01-Jul-2010 13:54:14] unstall the current zenpack 1st
[01-Jul-2010 13:54:17] or over right
[01-Jul-2010 13:54:21] over write
[01-Jul-2010 13:55:19] st3v3o: simply install it over
[01-Jul-2010 13:59:46] st3v3o: looks like dev.hw.totalMemory is the full memory amount
[01-Jul-2010 13:59:59] so I guess the threshold would be dev.hw.totalMemory * .9
[01-Jul-2010 14:00:18] off to a meeting
[01-Jul-2010 14:02:50] bigegor_ just installed but still can't edit/create WMI data sources ..going to restart zenoss and test again
[01-Jul-2010 14:05:50] WOOT!
[01-Jul-2010 14:06:00] bigegor_ I can now edit them
[01-Jul-2010 14:06:36] I suspect I can also add…thanks again for the help
[01-Jul-2010 14:12:40] st3v3o: np
[01-Jul-2010 14:28:55] mrayzenoss is now known as mray
[01-Jul-2010 14:40:58] taking mrays suggestions I looked at the current graph points from linux memory utilization and noticed that the RPN uses ${here/hw/totalMemory}. After re-reading the admin guide it looks like here.hw.totalMemory * .9 for a max threshold should give me 90%. If I change .9 to .1 I don't get any event …currently the system is using 22% memory..
[01-Jul-2010 14:53:24] st3v3o: do you have it formatted in rpn? ${here/hw/totalMemory},.1,*
[01-Jul-2010 14:54:51] it was already formated….I left that field default
[01-Jul-2010 14:55:45] 1024,*,${here/hw/totalMemory},/,1,-,-100,* that is what is in the memAvailReal RPN field
[01-Jul-2010 14:56:16] ooof
[01-Jul-2010 14:56:47] I created an additional threshold with a data point of memAvailReal_memAvilReal and a Max Value of here.hw.totalMemory * .1
[01-Jul-2010 14:57:27] can you just try it putting in an absolute numeric threshold, instead of a calculation?
[01-Jul-2010 14:57:38] and see if that works at least
[01-Jul-2010 14:58:08] it seems that I heard that there is some limitation on computations in thresholds
[01-Jul-2010 14:59:15] ok I entered a max value of 10
[01-Jul-2010 15:00:42] guess I have to wait for the next snmp poll for the threshold to get hit
[01-Jul-2010 15:00:51] yeah
[01-Jul-2010 15:02:04] I was just talking about that with one of the other devs yesterday - for testing he sets the frequency at 30 up at the /Devices class so he doesn't have to wait 5 minutes after making a change
[01-Jul-2010 15:03:18] that only works for the zenwinperf though, the other collectors do it on a per-collector basis.
[01-Jul-2010 15:04:21] alright an absolute value in the max threshold wored.
[01-Jul-2010 15:04:42] but that really sucks ..that I'll need to set thresholds per machine
[01-Jul-2010 15:04:48] there had to be a way to accomplish this
[01-Jul-2010 15:07:22] st3v3o: I've found rmatte to be fairly good with threshold issues
[01-Jul-2010 15:07:30] * Simon4_ drops him in it :)
[01-Jul-2010 15:31:16] * Simon4_ plays around with putting /opt/zenoss/var into a ramdisk
[01-Jul-2010 15:31:25] Simon4_ is now known as Simon4
[01-Jul-2010 17:34:44] Later all, off on vacation for the next 10 days
[01-Jul-2010 17:43:39] have fun!
[01-Jul-2010 17:45:01] your stomping grounds, Colorado
[01-Jul-2010 17:45:33] brrr, bring a sweater or 10 :)
[01-Jul-2010 18:32:24] hello all
[01-Jul-2010 18:32:28] rmatte: you around?
[01-Jul-2010 18:46:13] forsberg is now known as fOrsberg
[01-Jul-2010 19:01:17] I knew I was missing a channel
[01-Jul-2010 19:01:29] How'd the dev chat go earlier?
[01-Jul-2010 22:57:16] cd #eng
[01-Jul-2010 22:57:20] w/w
[01-Jul-2010 22:58:11] Can my config from 2.5 be imported into 3.0
[02-Jul-2010 00:00:30] [disconnected at Fri Jul 2 00:00:30 2010]
[02-Jul-2010 00:00:31] [connected at Fri Jul 2 00:00:31 2010]
[02-Jul-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]
[02-Jul-2010 01:30:47] pmcfadin_ is now known as pmcfadin
[02-Jul-2010 01:50:40] fOrsberg is now known as forsberg
[02-Jul-2010 06:16:20] hi
[02-Jul-2010 06:17:23] I am getting started with zenoss, and tried out the vmware appliance for 2.5.2
[02-Jul-2010 06:17:48] just to try the perf graphs out, I added "localhost" to the list of devices
[02-Jul-2010 06:17:57] seeing that snmpd is running in the appliance
[02-Jul-2010 06:18:30] however, I do not get perf graphs; instead, in event.log, I see "ERROR RenderServer Failed to generate a graph" all over
[02-Jul-2010 06:20:06] dv_: I'm just running to get the tube into the office, but could give you a hand in an hour or so if you're still about and no-one else in here is awake before them
[02-Jul-2010 06:20:08] then even
[02-Jul-2010 06:21:23] mmkay
[02-Jul-2010 06:21:46] i'm sure its just some stupid error, but my first contact with zenoss happened about two workdays ago :)
[02-Jul-2010 06:22:06] check zenperfsnmp.log to see if it's collecting from localhost all right
[02-Jul-2010 06:22:13] and the events on the localhost device to see fi there are any errors
[02-Jul-2010 06:22:42] * Simon4_ runs
[02-Jul-2010 06:38:10] hmm.
[02-Jul-2010 06:38:16] somehow, the error magically vanished.
[02-Jul-2010 06:40:52] tehhobbi1 is now known as tehhobbit
[02-Jul-2010 07:06:30] 38 minutes to work. not bad ;)
[02-Jul-2010 07:07:47] Simon4, it works now. no idea why though.
[02-Jul-2010 07:07:54] it just fixed itself :)
[02-Jul-2010 07:08:02] :)
[02-Jul-2010 07:08:13] dv_: possibly just needed some time to collect some data
[02-Jul-2010 07:08:40] * Simon4 peers at his email
[02-Jul-2010 07:08:49] 29 new reports generated in the last hour
[02-Jul-2010 07:08:55] we have crazy tenants
[02-Jul-2010 07:09:00] Simon4: are you simon lockhart? (just curious)
[02-Jul-2010 07:09:10] fragfutter: nope, Simon Helson
[02-Jul-2010 07:09:20] * Simon4 wonders why simon lockhart rings a bell
[02-Jul-2010 07:09:31] whois your ip address ;)
[02-Jul-2010 07:09:48] ahh :)
[02-Jul-2010 07:10:34] its absolutely too hot to prepare a usefull zenoss presentation
[02-Jul-2010 07:10:55] fragfutter: yes.
[02-Jul-2010 07:11:08] :P
[02-Jul-2010 07:11:12] * Simon4 has to enterprise up the dev server and move it to new hardware this afternoon, and motivation is seriously lacking
[02-Jul-2010 07:12:21] i think 10 slides are enough for the concepts, everything else has to happen with live demonstration and the whiteboard.
[02-Jul-2010 07:43:45] i need icecream http://www.youtube.com/watch?v=fvUQQF5S4Dg
[02-Jul-2010 07:58:41] Simon4: are you zenoss Simon?
[02-Jul-2010 07:59:14] ckrough: I'm simonh on the forums, don't work for Zenoss
[02-Jul-2010 07:59:19] roer
[02-Jul-2010 07:59:21] roger
[02-Jul-2010 07:59:33] * Simon4 asks too many silly questions to be working for zenoss :)
[02-Jul-2010 07:59:38] there is an engineer at Zenoss named Simon, was just curious
[02-Jul-2010 08:00:13] Simon4: you could be a fresh new employee who skipped courses :P
[02-Jul-2010 08:00:21] dv_: heh
[02-Jul-2010 08:03:17] heh that was me .. :p
[02-Jul-2010 08:03:21] so tell us simon, why did yo skip courses?
[02-Jul-2010 08:04:02] rocket: you are simon?
[02-Jul-2010 08:04:16] only course I've done is learning by reading the source code and swearing a lot :)
[02-Jul-2010 08:04:19] oh, you skipped courses
[02-Jul-2010 08:04:27] * ckrough is confused
[02-Jul-2010 08:04:47] no I am not .. I am Eric .. former community junkie that created many community zenpacks
[02-Jul-2010 08:05:10] predictive thresholds, enhanced the aix zenpack etc ..
[02-Jul-2010 09:05:02] all pirates on deck!
[02-Jul-2010 09:05:19] hahaha
[02-Jul-2010 09:05:20] arrr
[02-Jul-2010 09:06:09] yarr
[02-Jul-2010 09:11:47] so i had an issue the other day at a branch, and the cpu on one of our routers spiked to 95% for 15 minutes... i have a threshold set at 80%, but it didn't generate any events...
[02-Jul-2010 09:24:04] Morning
[02-Jul-2010 09:54:30] <|rt|> is there a way to manually set up a dependancy? I have a smart switch (no snmp) on my network so I can do a ping test to make sure it's up but I zenoss doesn't see it as a hop on the network so if it goes down a bunch of alerts are going to go off
[02-Jul-2010 09:55:13] imho zenoss only detects layer3 (routed) dependencies
[02-Jul-2010 09:55:40] <|rt|> fragfutter: yeah that's what I've been seeing when searching for this
[02-Jul-2010 09:56:14] having a flat network with 5k devices, this is a nightmare if one of the core switches fails ...
[02-Jul-2010 09:56:20] <|rt|> well this smartswitch does support vlans' so maybe i'll just set it up on a separate vlan/network
[02-Jul-2010 09:56:59] <|rt|> fragfutter: flat network with 5k devices sounds like a nightmare to begin with :)
[02-Jul-2010 09:57:16] the word is begin, it was already there...
[02-Jul-2010 09:59:36] <|rt|> then again if my smartswitch goes down then no alerts are going to get out b/c the network here would be almost completely down
[02-Jul-2010 09:59:52] <|rt|> the DMZ would be up but that would be about it
[02-Jul-2010 10:00:42] that is the bad thing about alerting by mail sms-gateways. You realy need to have a seperate alerting channel (local cell phone, second line).
[02-Jul-2010 10:00:51] <|rt|> yeah
[02-Jul-2010 10:01:42] <|rt|> I could hook this box directly to our gateway which has 2 internet connections but that gateway would still be a single point of failure
[02-Jul-2010 10:03:27] but if the gateway goes down, about a gazillion people notice and hammer your phone anyway
[02-Jul-2010 10:04:02] <|rt|> yeah that much is true...if it happens during work hours
[02-Jul-2010 10:24:43] rmatte, you alive?
[02-Jul-2010 10:25:16] I am now
[02-Jul-2010 10:28:18] so i'm not sure what's up, but i had a CPU spike on a router and didn't get any events.
[02-Jul-2010 10:28:31] did it actually register on the graph as having spiked?
[02-Jul-2010 10:28:31] i was hoping to setup email notifications if that happens again, but without an event i don't see how...
[02-Jul-2010 10:28:35] yep
[02-Jul-2010 10:28:39] 95.5% max
[02-Jul-2010 10:28:41] like, do you see a point where it exceeded the threshold line on the graph?
[02-Jul-2010 10:28:46] yes
[02-Jul-2010 10:28:54] and you've checked your event history to make sure that there's nothing in there?
[02-Jul-2010 10:28:56] but if i zoom in the graph it initially drops below the point
[02-Jul-2010 10:29:01] like the graph's scale doesn't keep up
[02-Jul-2010 10:29:12] but if i zoom in far enough, then it goes back to the spike being above the threshold.
[02-Jul-2010 10:29:15] well, the graph gets summarized the further back you go
[02-Jul-2010 10:29:20] yea, i looked at the history
[02-Jul-2010 10:29:26] no it's when i'm zooming in
[02-Jul-2010 10:29:31] not when i'm changing time
[02-Jul-2010 10:29:50] yeh, I mean if you zoom out far enough the scale gets summarized so you won't see it
[02-Jul-2010 10:29:54] you really have to zoom in
[02-Jul-2010 10:30:02] but yeh, I'm not sure why that didn't trigger
[02-Jul-2010 10:30:15] yea the default zoom shows the spike going over the threshold
[02-Jul-2010 10:30:23] and if i zoom in far enough i see that it spiked for about 15 minutes.
[02-Jul-2010 10:30:26] check your zenperfsnmp log is really the only suggestion I can make
[02-Jul-2010 10:30:33] hrm
[02-Jul-2010 10:30:33] ok
[02-Jul-2010 10:30:52] sometimes those cycles take a loooong time
[02-Jul-2010 10:31:03] hehe
[02-Jul-2010 10:31:07] i know it's because of the hardware and crap i have running on this system
[02-Jul-2010 10:44:36] it looks like it's making it thru the full cycle
[02-Jul-2010 10:44:41] last one took 209.08 seconds
[02-Jul-2010 10:44:51] doesn't have any left in queue or any failures
[02-Jul-2010 10:51:45] hi, i have a quick question. i've set up an alert rule to notify me on any events in a a certain class (/status/snmp). if i set up an event transform to drop rules with a given message summary, will those events pass to the alert rule before being dropped
[02-Jul-2010 10:52:01] or, will dropping the event simply ignore them and mean that there is no need to acknowledge
[02-Jul-2010 10:52:55] hey, does anyone know how to customize the device class a Zenpack is using? I used to be able to just copy / bind the templates, but that doesn't seem to be possible with the Libvert zenpack
[02-Jul-2010 10:55:21] Guest134324: events that are dropped are dropped silently
[02-Jul-2010 10:55:31] so you wouldn't get an alert on them
[02-Jul-2010 10:56:05] jmp242: thanks. so if i write a transform something to the effect of if evt.summary == "summary text here: evt._action = 'drop'
[02-Jul-2010 10:56:26] obviously with the closed quote before the : ... that should just silently drop it
[02-Jul-2010 10:56:34] works for me
[02-Jul-2010 10:57:16] yes Guest134324
[02-Jul-2010 11:00:09] sorry for one other silly question... the syntax for else if on event transforms is elif, correct
[02-Jul-2010 11:02:20] They are python code
[02-Jul-2010 11:02:41] so I believe that is correct, but check with the python docs for 2.4 I believe for the current version
[02-Jul-2010 11:08:48] hey guys
[02-Jul-2010 11:09:13] quick question, with the alert emails, the links in them do not work, its pointing at the box's hostname, zenoss
[02-Jul-2010 11:10:23] is there a way to reformat that URL in that email, to where i can point it at the externally accessible URL?
[02-Jul-2010 11:14:20] I believe you can just change the alert content in the GUI
[02-Jul-2010 11:14:55] Go into the alerting rule that you are worried about
[02-Jul-2010 11:15:04] say, prefs , alerting rules
[02-Jul-2010 11:15:38] message tab and you could edit it there... But to change the auto links
[02-Jul-2010 11:15:48] that ack for instance, you need to change the hostname I think
[02-Jul-2010 11:16:51] yeah thats the problem
[02-Jul-2010 11:16:53] its NATed
[02-Jul-2010 11:17:01] so the URL is a shared IP on a certain port for webGUI
[02-Jul-2010 11:17:38] message/22976?tstart=0
[02-Jul-2010 11:17:41] google is your friend
[02-Jul-2010 11:18:25] close :)
[02-Jul-2010 11:18:41] it isnt accisble from a DNS name
[02-Jul-2010 11:18:45] its only an IP :(
[02-Jul-2010 11:19:33] Try setting it to an IP them?
[02-Jul-2010 11:19:35] then
[02-Jul-2010 11:19:47] I don't see that they have it in the GUI yet so still the old way
[02-Jul-2010 11:20:03] ahh, so set the hostname to the IP
[02-Jul-2010 11:20:08] then it would work
[02-Jul-2010 11:20:08] Yea, try that
[02-Jul-2010 11:20:11] Lol nice
[02-Jul-2010 11:20:17] is there a way in the GUI to change hostname?
[02-Jul-2010 11:20:28] Not that I've found
[02-Jul-2010 11:20:57] sweet man
[02-Jul-2010 11:21:31] ty for the help
[02-Jul-2010 11:21:35] oh, one more Q
[02-Jul-2010 11:21:42] i saw 3.0 betas are out and i'm curious
[02-Jul-2010 11:22:05] if i use the webGUI to backup the stuff, would i be able to isntall 3.0 beta and import that data
[02-Jul-2010 11:22:37] You don't even need to do that, you can install over the existing version - though you don't really want to do that
[02-Jul-2010 11:22:48] but if you want to migrate to a dev server for instance
[02-Jul-2010 11:23:07] well i was stupid and used the ubuntu .bin file for install
[02-Jul-2010 11:23:07] then yes, you can take the backups and import them, but you also need the MySQL dumps (if they do them I don't recall)
[02-Jul-2010 11:23:14] and your $ZENHOME
[02-Jul-2010 11:23:30] i would like to switch to an apt repo
[02-Jul-2010 11:23:38] You then need to make sure that the zenoss user has full read/write permissions
[02-Jul-2010 11:23:47] on the $ZENHOME
[02-Jul-2010 11:23:50] kk
[02-Jul-2010 11:23:50] when you're done
[02-Jul-2010 11:24:00] Mmm, I only use RHEL5 for Zenoss
[02-Jul-2010 11:24:14] Well, EL5 anyway
[02-Jul-2010 11:25:04] centOS?
[02-Jul-2010 11:26:41] this is only day 2 on my zenoss install
[02-Jul-2010 11:26:44] switched from xabbix
[02-Jul-2010 11:26:52] its nice, i like it alot
[02-Jul-2010 11:27:04] I jush wish the webUI was a little more friendly, and that it could monitor ESXi hosts
[02-Jul-2010 11:27:56] i saw a zenpack for it, and it doesnt install right
[02-Jul-2010 11:28:24] bad directorys or somthing
[02-Jul-2010 11:30:13] Scientific Linux actually
[02-Jul-2010 11:30:34] I've not tried ESXi hosts - we use KVM which is built into EL5.4
[02-Jul-2010 11:30:57] I am fighting with that - but only because I don't want to use the device class /Devices/libverthost/kvm
[02-Jul-2010 11:32:03] hehe
[02-Jul-2010 11:39:27] hi everyone, I'm having a newbie problem :$
[02-Jul-2010 11:39:43] I'm having timeout in snmo
[02-Jul-2010 11:39:45] snmp*
[02-Jul-2010 11:41:19] in /etc/snmp/snmpd.conf I have: createuser IMNEWBIE MD5 THISISSECRET DES PRIVPASS
[02-Jul-2010 11:41:45] and: rwuser IMNEWBIE
[02-Jul-2010 11:42:52] I'm trying to probe with: snmpwalk -v 3 -a md5 -A THISISSECRET -u IMNEWBIE myserver.com
[02-Jul-2010 11:43:15] what do I miss?
[02-Jul-2010 11:43:29] oh, and the error is: snmpwalk: Timeout
[02-Jul-2010 11:43:40] chachan, can you snmpwalk the entire device?
[02-Jul-2010 11:44:59] arrrghhh, what do you mean?, do it in the snmp server?
[02-Jul-2010 11:45:40] no...
[02-Jul-2010 11:45:47] I can do it on localhost
[02-Jul-2010 11:45:56] from the zenoss server
[02-Jul-2010 11:46:41] I'm doing from zenoss server and it doesn't work, but localhost it's working
[02-Jul-2010 11:46:53] is*
[02-Jul-2010 11:47:03] snmpwalk the entire device from the zenoss server
[02-Jul-2010 11:47:54] so "snmpwalk -v1 -c system
[02-Jul-2010 11:48:02] although you probably don't need the "system" at the end
[02-Jul-2010 11:48:14] and replace the <> items
[02-Jul-2010 11:48:19] w/o the brackets
[02-Jul-2010 11:48:31] ok, let me see
[02-Jul-2010 11:50:09] Timeout: No Response from www.newstogram.com
[02-Jul-2010 11:50:19] firewall issue probably then
[02-Jul-2010 11:50:23] i'm assuming you can ping it
[02-Jul-2010 11:50:26] can you nmap it?
[02-Jul-2010 11:50:27] arrrghhh, I open 161 and 162 UDP
[02-Jul-2010 11:50:40] i think 161 needs to be open tcp... not positive tho.
[02-Jul-2010 11:50:56] mmm
[02-Jul-2010 11:51:18] I will open it, let me see
[02-Jul-2010 11:53:49] arrrghhh, I stopped the firewall and I still get timeout error :(
[02-Jul-2010 11:54:01] chachan, can you ping the device?
[02-Jul-2010 11:54:35] arrrghhh, yes I can, I'll try nmap-ing
[02-Jul-2010 11:54:44] just to see what ports it sees as open
[02-Jul-2010 11:54:57] assuming the ports are open, i would look back to the SNMP config on the server.
[02-Jul-2010 11:55:45] 161 and 162 TCP
[02-Jul-2010 11:56:01] I have 162 by UDP :s
[02-Jul-2010 11:56:17] k
[02-Jul-2010 11:56:25] check the SNMP config on the server then
[02-Jul-2010 11:56:26] but, now firewall is down, I don't understand why I still getting timeout error
[02-Jul-2010 11:56:49] you ran that nmap from the zenoss server right?
[02-Jul-2010 11:56:52] arrrghhh, what do I need to check=
[02-Jul-2010 11:57:07] chachan, depends on the server, kinda outside the scope of this help room...
[02-Jul-2010 11:57:07] arrrghhh, yes, I do that from Zenoss
[02-Jul-2010 11:57:25] arrrghhh, I understand
[02-Jul-2010 11:57:43] it's on centos
[02-Jul-2010 11:57:46] chachan, basically if you can't even snmpwalk the device, there's something blocking you or some bad config on the server.
[02-Jul-2010 11:58:16] even if I can snmpwalk the device locally?
[02-Jul-2010 11:58:20] did you set up the community strings correctly? and you're using the same string for the snmwalk command?
[02-Jul-2010 11:58:33] that's why i thought it would've been firewall.
[02-Jul-2010 11:58:49] hardware or software, but if nmap shows the port as open...
[02-Jul-2010 11:58:57] but firewall is down
[02-Jul-2010 11:59:04] i guess the other question would be, are you using the default port 161 for snmp?
[02-Jul-2010 11:59:14] arrrghhh, yes, I do
[02-Jul-2010 11:59:27] chachan, right but if you can't snmpwalk the device then there's something wrong... not with zenoss, with your network.
[02-Jul-2010 11:59:38] the community string is still by default
[02-Jul-2010 12:00:12] or with the snmp server
[02-Jul-2010 12:00:28] yes, which would be part of your network...
[02-Jul-2010 12:00:35] my point is, zenoss is not your problem.
[02-Jul-2010 12:01:02] arrrghhh, ok, then, thanks a lot for the help, I'll search about it ;)
[02-Jul-2010 12:01:07] np
[02-Jul-2010 12:29:22] arrrghhh, are you there?
[02-Jul-2010 12:29:26] yes
[02-Jul-2010 12:29:56] about to go back to the work network tho, so make it quick :P
[02-Jul-2010 12:30:03] when I nmap from zenoss with firewall down, I get "161/tcp filtered snmp"
[02-Jul-2010 12:30:13] ok
[02-Jul-2010 12:30:31] filtered?, why if the firewall is down?
[02-Jul-2010 12:30:44] hrm
[02-Jul-2010 12:30:45] not sure
[02-Jul-2010 12:30:45] telnet doesn't open :(
[02-Jul-2010 12:30:51] do you have any network equipment between those devices that could be blocking the connection?
[02-Jul-2010 12:31:23] rmatte, I don't think so
[02-Jul-2010 12:31:31] well, that's the only other possible explanation
[02-Jul-2010 12:31:37] your ISP may be filtering it...
[02-Jul-2010 12:31:39] unless you don't really have the firewall down
[02-Jul-2010 12:31:43] do you have private lines?
[02-Jul-2010 12:31:50] why don't you try explicitly allowing those ports in the firewall config?
[02-Jul-2010 12:31:52] rmatte, but zenoss is on VPS, could be datacenter details?
[02-Jul-2010 12:32:06] or there's another hardware firewall in the middle.
[02-Jul-2010 12:32:06] chachan: It's possible that they block traffic to those ports
[02-Jul-2010 12:32:34] rmatte, yes, I will ask them
[02-Jul-2010 12:32:41] chachan: contact your provider (but add the rules to the firewall first to test)
[02-Jul-2010 12:32:44] arrrghhh, rmatte, thanks for the help again
[02-Jul-2010 12:32:48] maybe you just think it's down but it's not really
[02-Jul-2010 12:32:49] np
[02-Jul-2010 12:32:52] np
[02-Jul-2010 12:33:02] rmatte, those rules are still added
[02-Jul-2010 12:33:24] I will ask
[02-Jul-2010 12:33:27] k
[02-Jul-2010 12:33:51] chachan: oh, one last thing to try...
[02-Jul-2010 12:33:55] traceroute -p 161
[02-Jul-2010 12:33:57] to the device
[02-Jul-2010 12:34:01] see where it stops
[02-Jul-2010 12:34:07] ah i forget you can trace to specific ports.
[02-Jul-2010 12:34:19] rmatte, that's a good one :) let me see
[02-Jul-2010 12:34:28] arrrghhh, yes, I forget it too
[02-Jul-2010 12:34:40] i shouldn't, considering my role int he company... lol.
[02-Jul-2010 12:34:45] :)
[02-Jul-2010 12:37:03] hmmm, you might actually want to do: traceroute -T -p 161
[02-Jul-2010 12:37:09] oh no wait
[02-Jul-2010 12:37:15] nevermind, snmp is UDP
[02-Jul-2010 12:37:22] so yeh, just -p should be fine
[02-Jul-2010 12:37:36] so snmp is udp
[02-Jul-2010 12:37:41] correct
[02-Jul-2010 12:37:42] then you can close that tcp port chachan
[02-Jul-2010 12:37:50] it can be done via TCP, but UDP is the default
[02-Jul-2010 12:37:57] makes sense, i don't know why you would want it to be TCP.
[02-Jul-2010 12:38:02] just more overhead
[02-Jul-2010 12:38:03] exactly
[02-Jul-2010 12:38:08] arrrghhh, but nmap says me that 161 is TCP
[02-Jul-2010 12:38:28] chachan, yea i had you open it tho right? i had you open 161 to both protocols just to be sure.
[02-Jul-2010 12:38:33] I didn't touch it, it's by default
[02-Jul-2010 12:38:36] oh
[02-Jul-2010 12:38:36] ok
[02-Jul-2010 12:38:48] i just didn't want to leave stuff open on your network if it doesn't need to be ;)
[02-Jul-2010 12:39:07] chachan: nmap shouldn't even seen your UDP port by default
[02-Jul-2010 12:39:12] unless you ran it with some flags
[02-Jul-2010 12:39:33] rmatte, mmm, ok
[02-Jul-2010 12:39:41] ah didn't even think about that.
[02-Jul-2010 12:39:56] i should probably stop trying to diagnose issues when i can't even solve my own haha
[02-Jul-2010 12:40:08] for example, here's an nmap of one of my Zenoss servers which has an SNMP agent on it:
[02-Jul-2010 12:40:09] PORT STATE SERVICE
[02-Jul-2010 12:40:09] 22/tcp open ssh
[02-Jul-2010 12:40:09] 111/tcp open rpcbind
[02-Jul-2010 12:40:09] 8081/tcp open blackice-icecap
[02-Jul-2010 12:40:09] 8180/tcp open unknown
[02-Jul-2010 12:40:09] 10000/tcp open snet-sensor-mgmt
[02-Jul-2010 12:40:09] 20000/tcp open unknown
[02-Jul-2010 12:40:40] that's just from doing nmap with no flags
[02-Jul-2010 12:41:29] I get it
[02-Jul-2010 12:46:38] forsberg is now known as fOrsberg
[02-Jul-2010 12:57:00] hey, does anyone know how to customize the device class a Zenpack is using? I used to be able to just copy / bind the templates, but that doesn't seem to be possible with the Libvert zenpack
[02-Jul-2010 12:57:34] jmp242: pretty sure that pack is hardcoded to use those locations
[02-Jul-2010 12:57:43] jmp242: you'd have to go in and change the code
[02-Jul-2010 12:58:04] when I say locations I mean device class
[02-Jul-2010 13:00:31] well, I'm up to 14 production Zenoss servers now, yay
[02-Jul-2010 13:00:40] damn
[02-Jul-2010 13:03:14] Ahh, ok.
[02-Jul-2010 13:03:30] Am I the only one who finds it a bit annoying for zenpacks to dictate the device classes?
[02-Jul-2010 13:03:39] or is there a new best practice there?
[02-Jul-2010 13:04:21] jmp242: it's an object based database, so in the xml file you need to specify a location for the class, then any scripts built need to center around that class for the more advanced packs
[02-Jul-2010 13:04:30] I'm sure there would be some way to do it dynamically
[02-Jul-2010 13:05:11] Yea, the reason it's come up is our VM hosts run Linux... I'd like to get all the linux monitoring that works for us under /Devices/Server/Linux
[02-Jul-2010 13:05:19] to also apply to the KVM hosts
[02-Jul-2010 13:06:09] I see
[02-Jul-2010 13:06:35] you could always just copy the Devices template from /Devices/Server/Linux to the one that the ZenPack creates
[02-Jul-2010 13:06:50] same end result, just probably not organizationally what you'd expect
[02-Jul-2010 13:07:12] yea, I would if it would let me.
[02-Jul-2010 13:07:21] I'm beginning to wonder if that Zenpack installed correctly
[02-Jul-2010 13:07:29] it's not a choice in the copy to dialog
[02-Jul-2010 13:07:53] what's not a choice?
[02-Jul-2010 13:08:38] err /libverthost
[02-Jul-2010 13:08:52] /Server/libvirtHost
[02-Jul-2010 13:08:55] is what it is
[02-Jul-2010 13:09:24] then there are a bunch of subclasses under that
[02-Jul-2010 13:09:35] not that either
[02-Jul-2010 13:09:47] for ESX, KVM, LXC, OpenNebula, OpenVZ, QEMU, UML, VirtualBox, and XEN
[02-Jul-2010 13:09:56] not all of them are functional/tested
[02-Jul-2010 13:10:07] Yea, I know that
[02-Jul-2010 13:10:11] for instance, ESX doesn't work properly
[02-Jul-2010 13:10:27] XEN works the best apparently
[02-Jul-2010 13:10:35] I must have broken it by trying to copy it from zope over
[02-Jul-2010 13:10:43] possibly
[02-Jul-2010 13:11:23] now, the question - would an zenpack uninstall break it more I wonder
[02-Jul-2010 13:11:34] doubt it
[02-Jul-2010 13:11:43] ok, I'll try uninstall / reinstall
[02-Jul-2010 13:15:31] mmm
[02-Jul-2010 13:15:38] well, it cleaned up my database I think
[02-Jul-2010 13:16:03] but I still can't copy anything (in the Zenoss UI) from /Devices/Server/Linux to /Devices/libVirtHost
[02-Jul-2010 13:19:44] hey rmatte you around ?
[02-Jul-2010 13:19:58] yup
[02-Jul-2010 13:20:13] lol it seems we all get unofficial support from rmatte
[02-Jul-2010 13:20:16] jmp242: you should be able to copy anything to anything
[02-Jul-2010 13:20:35] arrrghhh: that's what I'm here for lol (well, one of the reasons)
[02-Jul-2010 13:20:42] give and ye shall receive
[02-Jul-2010 13:20:46] haha well we appreciate it.
[02-Jul-2010 13:21:05] some said you might be able to help with the threshold questions
[02-Jul-2010 13:21:16] st3v3o: which questions?
[02-Jul-2010 13:21:17] That is indeed what I would expect
[02-Jul-2010 13:22:18] I'm trying to create a memory threshold for 90%
[02-Jul-2010 13:22:49] ok
[02-Jul-2010 13:23:22] my datapoint is memAvailReal_memAvailReal and i tired the following expression but no luck
[02-Jul-2010 13:23:22] here.hw.totalMemory * .9
[02-Jul-2010 13:23:30] for the max value
[02-Jul-2010 13:23:44] have you set somethign like this up
[02-Jul-2010 13:23:47] what snmp values do you have to work with?
[02-Jul-2010 13:24:53] I'm using the default linux device template..and relevant values seem to be memAvailReal and possibly memAvailSwap
[02-Jul-2010 13:26:00] I see, ok...
[02-Jul-2010 13:26:18] give me a couple minutes, I'll whip something up
[02-Jul-2010 13:27:24] I should actually do this on my servers anyways, I've done it for windows servers, cisco devices, and hp devices, just hadn't gotten around to doing it for Linux
[02-Jul-2010 13:29:18] memory thresholds are pretty complex to do properly, took me a long time to figure it out
[02-Jul-2010 13:30:18] yeah I'd love to know the exact logic behind it
[02-Jul-2010 13:30:45] yeah, I'll show you once I'm done getting the properties I need
[02-Jul-2010 13:30:56] sweet
[02-Jul-2010 13:32:31] ok, I found the 2 bits that I need... won't be long
[02-Jul-2010 13:33:05] yeh, this is going to be a lot easier than the others I did...
[02-Jul-2010 13:33:19] afk…bathroom break
[02-Jul-2010 13:34:16] k
[02-Jul-2010 13:35:36] ah actually it will be
[02-Jul-2010 13:35:41] no big deal though
[02-Jul-2010 13:36:58] afk, I'll be back in about 10
[02-Jul-2010 13:43:51] Hey again
[02-Jul-2010 13:44:14] So i'm adding some switches and dont want SNMP, I just want ping, is there a way to add these without it trying to use snmp?
[02-Jul-2010 13:44:54] TinyTim, you can set each device to just ignore SNMP...
[02-Jul-2010 13:45:10] well i have like 100 or more i need to add
[02-Jul-2010 13:45:23] and i'm _not_ going to edit each one seperately
[02-Jul-2010 13:45:24] how are you adding them?
[02-Jul-2010 13:45:32] well then...
[02-Jul-2010 13:45:33] Add each one to /Ping?
[02-Jul-2010 13:45:33] right now, easy add, and adding them
[02-Jul-2010 13:45:40] i guess that settles that then.
[02-Jul-2010 13:47:12] jmp242: i did add one and i still had SNMP alerts on the dashboard for it
[02-Jul-2010 13:47:13] lol
[02-Jul-2010 13:47:30] you'll have to disable them by hand, or use the zenbatchload to import them.
[02-Jul-2010 13:47:55] i have 500 devices, and i had to go thru each one and disable interfaces that i didn't want monitored. basically it sucks, but suck it up .
[02-Jul-2010 13:48:57] That's weird, the /Ping class should have zSNMPIgnore set to true
[02-Jul-2010 13:49:09] Did you add the device directly or move it afterwords
[02-Jul-2010 13:52:26] after i think
[02-Jul-2010 13:52:31] it was yesterday i dont remember
[02-Jul-2010 13:52:36] i see what you're saying im trying it now
[02-Jul-2010 13:52:50] they need to let you add a range of addresses
[02-Jul-2010 13:52:54] make my life so much easier
[02-Jul-2010 13:54:48] im trying to get a sonicwall snmp working
[02-Jul-2010 13:54:51] no port option
[02-Jul-2010 14:02:30] st3v3o: I'm back, should have something shortly, I only have total memory and available memory so I need to do it reverse of how I'm used to doing it, shouldn't be a big deal
[02-Jul-2010 14:02:51] ok
[02-Jul-2010 14:04:18] If you do it after, it will have the zProperties and config set from where you initially added it
[02-Jul-2010 14:04:34] You'll want to try adding one directly to /Ping
[02-Jul-2010 14:15:34] got it jmp242
[02-Jul-2010 14:15:36] you're the man
[02-Jul-2010 14:18:02] You probably could script something in ZenDMD to change the zProperties (or some other ways) for existing devices. I don't quite know how to do that though
[02-Jul-2010 14:19:10] dmd.Devices.zSnmpMonitorIgnore
[02-Jul-2010 14:19:18] probably set to true somehow
[02-Jul-2010 14:19:21] python stuff
[02-Jul-2010 14:20:01] though first you got to find the device
[02-Jul-2010 14:22:27] if dr.hasProperty('zWmiMonitorIgnore'):
[02-Jul-2010 14:22:27] dr.zWmiMonitorIgnore = True
[02-Jul-2010 14:22:27] else:
[02-Jul-2010 14:22:27] dr._setProperty('zWmiMonitorIgnore', True, type='boolean')
[02-Jul-2010 14:22:27] dr could be a device or a device class.
[02-Jul-2010 14:22:53] for zWmiMonitorIgnore - but could change to zSnmpMonitorIgnore
[02-Jul-2010 14:22:58] st3v3o: http://fpaste.org/fbTz/
[02-Jul-2010 14:23:27] so make dr = /Ping
[02-Jul-2010 14:23:32] etc. . .
[02-Jul-2010 14:23:42] looking now
[02-Jul-2010 14:23:48] At lease per message/7417?tstart=0
[02-Jul-2010 14:24:04] st3v3o: let me know if you run in to any problems with it, I haven't tested it since the Linux server that I have on my test box isn't reporting snmp info suddenly for some reason
[02-Jul-2010 14:24:12] st3v3o: but it should work
[02-Jul-2010 14:26:32] over right the current memory transform …or append under it
[02-Jul-2010 14:26:38] so how would one monitor windows machine with WMI behind a firewall
[02-Jul-2010 14:26:41] since this thing has no agent
[02-Jul-2010 14:26:55] oh, what memory transform do you already have?
[02-Jul-2010 14:27:10] is it from the Windows SNMP Performance ZenPack?
[02-Jul-2010 14:27:24] http://fpaste.org/sWpO/
[02-Jul-2010 14:27:54] interesting, wonder what that's from, ok one sec
[02-Jul-2010 14:28:23] ummm
[02-Jul-2010 14:28:29] that threshold that you have won't even technically work
[02-Jul-2010 14:28:36] it's missing a line lol
[02-Jul-2010 14:28:44] anyways, I'll fix it, one sec
[02-Jul-2010 14:28:44] jmp242: halp
[02-Jul-2010 14:28:52] wonder if one of the other guys dropped it in
[02-Jul-2010 14:29:12] yeah looks like it
[02-Jul-2010 14:29:17] i'll just over right with yours
[02-Jul-2010 14:29:23] it's identical to mine
[02-Jul-2010 14:29:27] which I wrote from scratch
[02-Jul-2010 14:29:36] so they must have took it from one of my packs or something
[02-Jul-2010 14:29:49] maybe but …copy and paste fail :P
[02-Jul-2010 14:29:57] so no ideas
[02-Jul-2010 14:30:01] here: http://fpaste.org/V0HQ/
[02-Jul-2010 14:30:11] yeh, they didn't copy and paste the dev = line
[02-Jul-2010 14:30:20] which means dev isn't set, so the transform can't function
[02-Jul-2010 14:31:06] ok i'll use this latest one
[02-Jul-2010 14:31:49] they basically just make it so that the threshold message displays the percentage of utilization
[02-Jul-2010 14:31:53] instead of the bytes
[02-Jul-2010 14:31:56] which would be useless
[02-Jul-2010 14:31:57] lol
[02-Jul-2010 14:32:02] fOrsberg is now known as forsberg
[02-Jul-2010 14:32:17] also go in to your memory graph and add the threshold to it
[02-Jul-2010 14:32:25] that way you'll see a red line on the graph representing the threshold
[02-Jul-2010 14:32:36] yeah I added it to the graph..but it doesn't show up
[02-Jul-2010 14:34:48] going to set the threshold to 20%
[02-Jul-2010 14:34:55] see if I can get an alert to trigger
[02-Jul-2010 14:36:28] TinyTim: well, you really can't unless you can pass WMI which doesn't make much sense. You could use a VPN if allowed I guess.
[02-Jul-2010 14:36:43] well this is kind of a show stopper :(
[02-Jul-2010 14:36:50] does the paid version have an agent?
[02-Jul-2010 14:36:57] i need to be able to get WMI info from clients PCs
[02-Jul-2010 14:37:44] No, Zenoss is agentless
[02-Jul-2010 14:37:48] sweet
[02-Jul-2010 14:38:04] Depending on what you mean, you could perhaps use a remote collector
[02-Jul-2010 14:38:07] jassinpain_ is now known as jassinpain
[02-Jul-2010 14:38:25] if you're talking about a network where you could install another computer on the network to do the scanning
[02-Jul-2010 14:38:25] rmatte…I'm not sure its correctly calculating the max value as I don't see a red bar when the threshold is added to the graph..but I do see it if I set an absolute value
[02-Jul-2010 14:38:39] and VPN from it to the central collector
[02-Jul-2010 14:39:00] but if there's a bunch of single computers you want to use WMI on, then you need to be able to talk to the WMI port on them
[02-Jul-2010 14:39:30] I suppose you might be able to hack something together with a program that gathers data and sends it as SNMP traps to Zenoss or something
[02-Jul-2010 14:39:47] but most community users are monitoring their internal networks
[02-Jul-2010 14:40:21] There was some though a while ago to agents, but I guess Zenoss Inc decided that they want to compete as the agentless solution so far and haven't written any agent(s).
[02-Jul-2010 14:41:37] yeah im looking into remote collectors now
[02-Jul-2010 14:41:42] that will work for the larger clients
[02-Jul-2010 14:42:28] Yea - The newer Collector tool docs/DOC-5861
[02-Jul-2010 14:42:45] makes it pretty easy to do with Core, and the paid product supports it right out of the box
[02-Jul-2010 14:43:46] i just need to enable SNMP on like 2032323024 windows xp machines :( lol
[02-Jul-2010 14:43:53] rmatte..I have to run the wife calls…if you have more time to take a look if you could shoot me an email
[02-Jul-2010 14:43:57] steve@arin.net
[02-Jul-2010 14:44:30] Yea, and you probably need to install the windows feature too
[02-Jul-2010 14:44:41] TinyTim: but what were you going to use SNMP for?
[02-Jul-2010 14:45:18] hello all
[02-Jul-2010 14:45:23] i have the NFS monitoring transform setup but I recall there is a way to get zenoss to run a remote ssh command if the event NFS disconnected is triggered. Can anyone refresh my memory?
[02-Jul-2010 14:46:05] set up an event command Quadro
[02-Jul-2010 14:46:08] err
[02-Jul-2010 14:46:13] QubeZ:
[02-Jul-2010 14:47:14] and it looks like you can't use the built in Use SSH like you can for monitoring
[02-Jul-2010 14:47:28] so you'd need something like the Nagios Run_by_SSH plugin
[02-Jul-2010 15:09:25] hi everyone again, I'm getting this error: http://pastebin.org/376368
[02-Jul-2010 15:16:40] chachan: on every device or just one?
[02-Jul-2010 15:17:21] rmatte, I'm trying with one, it's a fresh installation and I only have one device
[02-Jul-2010 15:17:40] I see
[02-Jul-2010 15:20:04] I get the same in zenoss.snmp.NewDeviceMap and zenoss.snmp.DeviceMap
[02-Jul-2010 15:21:09] could be the "ordering"?
[02-Jul-2010 15:21:26] of the plugin, of course...
[02-Jul-2010 15:28:56] do I need to create a SNMPv3 rw user?
[02-Jul-2010 15:29:34] rmatte, I mean, I created the user, but I didn't put the line "rwuser", do I need it?
[02-Jul-2010 15:35:06] rmatte: you recall helping me run a remote ssh command to remount nfs if the transform spit out an event on Zenoss awhile ago? I followed your wiki tutorial but for somereason, i have this shell script lying around on my servers but i can't recall how i called it
[02-Jul-2010 15:38:23] QubeZ: you probably called it via Event Manager -> Commands
[02-Jul-2010 15:39:02] chachan: not sure, I've never used snmp v3 before
[02-Jul-2010 15:39:15] chachan: are you sure you're modelling a device that is actually using snmpv3?
[02-Jul-2010 15:39:20] v3 isn't very widely used
[02-Jul-2010 15:39:53] rmatte, yes, I'm sure, how do you secure the snmp?, strong comunity word and firewall?
[02-Jul-2010 15:40:08] rmatte: thanks
[02-Jul-2010 15:40:16] QubeZ: np
[02-Jul-2010 15:40:40] chachan: just strong community word and configure the agents to only accept connections from a specific IP
[02-Jul-2010 15:41:03] firewall would technically work better
[02-Jul-2010 15:41:35] do you do it with com2sec?
[02-Jul-2010 15:42:14] nope
[02-Jul-2010 15:42:58] well, yes, on linux servers
[02-Jul-2010 15:43:04] but every agent has it's own way of doing it
[02-Jul-2010 15:43:40] I still don't understand why you're getting that error though
[02-Jul-2010 15:43:46] rmatte, I get, then, I will follow your steps and I'll stop using v3
[02-Jul-2010 15:43:47] and it's hard to help you if you only have 1 device to test with
[02-Jul-2010 15:43:59] it's entirely possible that Zenoss is choking on the SNMP output from that 1 specific device
[02-Jul-2010 15:44:10] well, I added a new device with a rwuser and I didn't get this error
[02-Jul-2010 15:44:11] determining if it's choking on the output from other devices would help
[02-Jul-2010 15:44:29] k
[02-Jul-2010 15:44:52] I'll try with non-v3 and I will let you know
[02-Jul-2010 15:45:15] k
[02-Jul-2010 15:45:26] rmatte: I forgot what class I need to set the command to with your transform method
[02-Jul-2010 15:45:38] basically, how do I make sure this ssh script only fires off when your event is written
[02-Jul-2010 15:45:52] QubeZ: check in the transform, whatever class the transform sets it to
[02-Jul-2010 15:46:01] rmatte: k
[02-Jul-2010 15:46:51] you could really just filter by the summary alone though
[02-Jul-2010 15:47:02] summary contains whatever
[02-Jul-2010 15:49:12] rmatte: i see your summary is NFS Share ... unavailable. I'll use that
[02-Jul-2010 15:49:30] then I can just use ssh ${manageIP}:/zenoss/remountnfs.sh ?
[02-Jul-2010 15:49:39] then I can just use ssh root@${manageIP}:/zenoss/remountnfs.sh ?
[02-Jul-2010 15:49:58] yeh, just do: if summary contains 'NFS Share'
[02-Jul-2010 15:50:02] or whatever
[02-Jul-2010 15:50:22] thanks
[02-Jul-2010 15:50:32] maybe even: if summary contains 'NFS Share' and summary contains 'unavailable'
[02-Jul-2010 15:53:11] zenoss.cmd.* are use only by ssh?, or can I use them by snmp?
[02-Jul-2010 15:53:25] only ssh
[02-Jul-2010 15:53:36] btw, rmatte, now I don't get the error (non-v3)
[02-Jul-2010 15:53:43] ah
[02-Jul-2010 15:53:57] cool
[02-Jul-2010 15:53:57] yep
[02-Jul-2010 15:53:57] :)
[02-Jul-2010 15:53:59] thanks
[02-Jul-2010 15:58:46] np
[02-Jul-2010 16:02:17] If some of you need cabling inspiration. The winner of the Sysadmin Neat Freak Award: http://people.cs.kuleuven.be/~bart.vanbrabant/sysadmin_lisa_10/img-8.html
[02-Jul-2010 16:02:33] You won't believe it
[02-Jul-2010 16:04:55] how do I get people with non-global permissions the ability to see Reports? (they show up blank/0 available)
[02-Jul-2010 16:05:05] pretty nice, but I've honestly seen nicer
[02-Jul-2010 16:05:14] i.e. they don't have global Operator, only operator on a set of Administered Objects
[02-Jul-2010 16:05:33] but would like to see some reports not directly related to their administered objects
[02-Jul-2010 16:05:54] gwb2351: I'm guessing you're talking enterprise?
[02-Jul-2010 16:06:00] yes
[02-Jul-2010 16:06:09] no idea
[02-Jul-2010 16:06:15] you could ask Zenoss support
[02-Jul-2010 16:06:15] ha ha :)
[02-Jul-2010 16:06:21] i will
[02-Jul-2010 16:06:27] this channel has more core users than anything
[02-Jul-2010 16:06:38] ;)
[02-Jul-2010 16:06:40] rmatte: We were all laughing here. This is someone who has obviously never moved any of that equipment
[02-Jul-2010 16:06:47] and a great depth of knowledge that is priceless sometimes
[02-Jul-2010 16:06:57] pmcfadin: yeh, really
[02-Jul-2010 16:07:25] we move one server and it seems you get a knot that would confuse a whole troop of boyscouts
[02-Jul-2010 16:08:01] lol
[02-Jul-2010 16:08:22] that's what vms and blade chassis are for :P
[02-Jul-2010 16:08:54] Ha. I was just thinking... the best reason to virtualize everything
[02-Jul-2010 16:08:59] physically moving servers is overrated
[02-Jul-2010 16:09:57] Cisco has a virtual switch so you can have a virtual cable mess. They think of everything
[02-Jul-2010 16:12:12] nice
[02-Jul-2010 16:36:48] where can I see why zenoss can't collete information from device?
[02-Jul-2010 16:39:18] well, make sure the device is snmpwalkable first
[02-Jul-2010 16:43:19] can I pass port parameter to snmpwalk?, I don't see it in the "--help"
[02-Jul-2010 16:44:42] -p
[02-Jul-2010 16:45:10] no :(
[02-Jul-2010 16:47:47] IPADDRESS:PORT
[02-Jul-2010 16:47:59] rmatte, I can snmpwalk from zenoss
[02-Jul-2010 16:48:12] are you using a non-standard port for SNMP?
[02-Jul-2010 16:48:18] yes
[02-Jul-2010 16:48:32] but, the trap is short
[02-Jul-2010 16:48:35] :s
[02-Jul-2010 16:48:55] right, but Zenoss is expecting the standard port by default, so that's probably why it's not collecting
[02-Jul-2010 16:49:08] it doesn't just magically know what port you're using if you're not using the standard one
[02-Jul-2010 16:49:10] rmatte, but I set it
[02-Jul-2010 16:49:26] you set zSnmpPort ?
[02-Jul-2010 16:49:30] yes
[02-Jul-2010 16:49:31] ah
[02-Jul-2010 16:49:39] so what happens when you model the device?
[02-Jul-2010 16:49:44] what's the output look like?
[02-Jul-2010 16:50:15] I don't have any error, but I can't see any colleted data in OS tab
[02-Jul-2010 16:50:17] and on that note... vacation #2
[02-Jul-2010 16:50:28] chachan: it's showing as unknown you mean?
[02-Jul-2010 16:50:36] chachan: or the OS tab is completely blank?
[02-Jul-2010 16:50:48] rmatte, that's right
[02-Jul-2010 16:50:53] as nothing happen
[02-Jul-2010 16:50:53] which is right?
[02-Jul-2010 16:50:57] unknown or blank?
[02-Jul-2010 16:51:06] it's blank, sorruy
[02-Jul-2010 16:51:14] then your snmp permissions aren't set correctly
[02-Jul-2010 16:51:20] it's only spitting out limited snmp info
[02-Jul-2010 16:51:23] linux server yes?
[02-Jul-2010 16:51:57] yes, it's CentOS 5.5
[02-Jul-2010 16:52:12] where can I see that permissions?
[02-Jul-2010 16:52:27] here's the snmpd.conf that I use
[02-Jul-2010 16:52:28] http://pastebin.com/4CWeFVWD
[02-Jul-2010 16:52:40] just take that, and modify it
[02-Jul-2010 16:52:51] only a few lines need to be changed
[02-Jul-2010 16:52:56] trapsink 192.168.0.1 your-key
[02-Jul-2010 16:53:03] trap2sink 192.168.0.1 your-key
[02-Jul-2010 16:53:08] rocommunity your-key
[02-Jul-2010 16:53:13] rwcommunity your-key
[02-Jul-2010 16:53:19] syslocation "Set this to the location"
[02-Jul-2010 16:53:25] syscontact whoever@contact.com
[02-Jul-2010 16:53:28] modify those lines
[02-Jul-2010 16:53:39] replace your-key with the snmp community string
[02-Jul-2010 16:53:50] and replace 192.168.0.1 with the IP of your Zenoss server
[02-Jul-2010 16:54:07] ok, I will
[02-Jul-2010 16:54:13] the rwcommunity is optional
[02-Jul-2010 16:54:34] you can comment it out if you want
[02-Jul-2010 17:04:02] OMG, rmatte, that .conf rocks!
[02-Jul-2010 18:00:07] I think I get a bug, when I'm trying to autodiscover with a different snmp port, it fails, I have to model again and set to port again
[02-Jul-2010 18:10:00] see you guys, have a great weekend
[02-Jul-2010 20:05:46] forsberg is now known as fOrsberg
[03-Jul-2010 00:00:30] [disconnected at Sat Jul 3 00:00:30 2010]
[03-Jul-2010 00:00:30] [connected at Sat Jul 3 00:00:30 2010]
[03-Jul-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]
[03-Jul-2010 01:38:35] fuzz is now known as Guest68193
[03-Jul-2010 05:48:27] fuzz is now known as Guest21289
[03-Jul-2010 07:45:40] fOrsberg is now known as forsberg
[03-Jul-2010 08:33:22] forsberg is now known as fOrsberg
[03-Jul-2010 08:56:52] fuzz is now known as Guest35754
[03-Jul-2010 10:12:08] fuzz is now known as Guest70392
[03-Jul-2010 10:51:14] fOrsberg is now known as forsberg
[03-Jul-2010 11:33:52] fuzz is now known as Guest65646
[03-Jul-2010 12:48:51] fuzz is now known as Guest76747
[03-Jul-2010 13:25:55] rmatte,
[03-Jul-2010 13:26:15] can you do me a favor and du -sh you /usr/local/zenoss dir please?
[03-Jul-2010 13:26:25] s/you/your
[03-Jul-2010 14:06:32] rhettardo_ is now known as rhettardo
[03-Jul-2010 14:50:45] fuzz is now known as Guest50054
[03-Jul-2010 18:47:55] fuzz is now known as Guest95030
[03-Jul-2010 19:58:57] forsberg is now known as fOrsberg
[03-Jul-2010 20:09:21] bzed is now known as bzed_
[03-Jul-2010 20:09:31] bzed_ is now known as bzed__
[03-Jul-2010 20:09:37] bzed__ is now known as bzed
[03-Jul-2010 22:58:02] can someone please du -sh /usr/local/zenoss for me and tell me how big theirs is?
[03-Jul-2010 23:03:32] du -sh /usr/local/zenoss
[03-Jul-2010 23:03:33] du: cannot access `/usr/local/zenoss': No such file or directory
[03-Jul-2010 23:03:36] :)
[03-Jul-2010 23:03:47] du -sh /opt/zenoss
[03-Jul-2010 23:03:47] 2.5G /opt/zenoss
[03-Jul-2010 23:04:49] rrd-data and mysql are elsewhere
[03-Jul-2010 23:05:23] ahhh ok thanks sytem you're a gentleman and a scholar :)
[03-Jul-2010 23:06:57] rrd takes about 100G, mysql 23G
[03-Jul-2010 23:07:52] about 700 monitored devices
[03-Jul-2010 23:08:21] * Bryanstein thinks wtf
[03-Jul-2010 23:08:42] sytem, what are the specs of the box running zenoss?
[03-Jul-2010 23:09:00] How much RAM does it consume at that level?
[03-Jul-2010 23:11:37] 8 cores, 16G memory, SAN-disks
[03-Jul-2010 23:15:51] Cached memery 10G, so zenoss itself wont use lot of memory at the moment
[03-Jul-2010 23:17:07] but i think it will use more when it has been running more than few hours what it is now
[03-Jul-2010 23:17:41] we must stop/start zenoss at least weekly, or it will fill /tmp with some unlinked logs
[04-Jul-2010 00:00:30] [disconnected at Sun Jul 4 00:00:30 2010]
[04-Jul-2010 00:00:31] [connected at Sun Jul 4 00:00:31 2010]
[04-Jul-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]
[04-Jul-2010 00:05:31] fuzz is now known as Guest70778
[04-Jul-2010 01:18:15] fuzz is now known as Guest71640
[04-Jul-2010 03:43:44] fOrsberg is now known as forsberg
[04-Jul-2010 04:13:57] forsberg is now known as fOrsberg
[04-Jul-2010 04:14:08] fuzz is now known as Guest41609
[04-Jul-2010 05:15:14] fuzz is now known as Guest32412
[04-Jul-2010 06:09:05] fuzz is now known as Guest93011
[04-Jul-2010 07:21:04] fuzz is now known as Guest95783
[04-Jul-2010 08:50:08] fuzz is now known as Guest30862
[04-Jul-2010 10:18:49] fOrsberg is now known as forsberg
[04-Jul-2010 10:32:46] forsberg is now known as fOrsberg
[04-Jul-2010 10:59:48] fOrsberg is now known as forsberg
[04-Jul-2010 13:09:34] frozty_s1 is now known as frozty_sa
[04-Jul-2010 14:44:56] fuzz is now known as Guest45818
[04-Jul-2010 15:40:45] forsberg is now known as fOrsberg
[04-Jul-2010 16:34:07] fuzz is now known as Guest96347
[05-Jul-2010 00:00:30] [disconnected at Mon Jul 5 00:00:30 2010]
[05-Jul-2010 00:00:31] [connected at Mon Jul 5 00:00:31 2010]
[05-Jul-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]
[05-Jul-2010 01:25:59] what daemons config time means?
[05-Jul-2010 01:26:25] on collectors daemon performance-page
[05-Jul-2010 01:26:54] zenperfsnmp is taking huge peaks sometime
[05-Jul-2010 01:49:43] fOrsberg is now known as forsberg
[05-Jul-2010 04:04:07] Jesper is now known as JesperS
[05-Jul-2010 04:07:10] JesperS is now known as Jesper-
[05-Jul-2010 04:10:03] Jesper- is now known as JesperS
[05-Jul-2010 06:35:55] Hello, is it possible to get Zenoss to moitor and alert on active process using snmp? I have my snmp setup on my targets, but it does seem that zenoss knows about the processes I want to monitor.
[05-Jul-2010 06:43:55] baffle_ is now known as baffle
[05-Jul-2010 07:21:22] Hello, is it possible to get Zenoss to moitor and alert on active process using snmp? I have my snmp setup on my targets, but it does seem that zenoss knows about the processes I want to monitor. Or about installed software which I can access usinmg snmpwalk
[05-Jul-2010 08:23:25] I made an addition to the SNMP collectors, how I do instruct Zenoss to poll again?
[05-Jul-2010 08:42:29] How can I list what plugins are currently available to me on my Zenoss server?
[05-Jul-2010 09:04:48] forsberg is now known as fOrsberg
[05-Jul-2010 09:08:51] what you mean with plugins?
[05-Jul-2010 09:10:56] zcollectorplugins
[05-Jul-2010 09:19:58] page where you can add them to some device or group?
[05-Jul-2010 09:31:36] My question is, how can I determine what zcollectorplugins are available to me? I can add them to the list, but I have to know what they are beforeand. Do they map to standard MIB names, somehow?
[05-Jul-2010 09:32:20] realitygeoff_: when you click to add zcollector plugins on the edit zProperties page, you should be given a list where you can drag the ones you want into the current list
[05-Jul-2010 09:32:24] that'll show you all the available ones
[05-Jul-2010 09:36:14] I don't see that.
[05-Jul-2010 09:36:27] Under zPropertEdit, right?
[05-Jul-2010 09:36:55] so if you navigate to a device class, click on the zProperties tab
[05-Jul-2010 09:37:09] then under zCollectorPlugins you can hit the "Edit" link
[05-Jul-2010 09:38:56] /zport/dmd/Devices/zSortableProperty?prop=zCollectorPlugins
[05-Jul-2010 09:39:59] and link "add fields" shows all available plugins
[05-Jul-2010 10:07:18] maybe i'm an idiot, but where is the "Commands tab of the event manager" ?
[05-Jul-2010 10:09:59] forget it, i *am* an idiot
[05-Jul-2010 10:15:32] fOrsberg is now known as forsberg
[05-Jul-2010 10:42:50] lol
[05-Jul-2010 11:00:08] forsberg is now known as fOrsberg
[05-Jul-2010 11:11:32] hmmm where can I get ZenPacks.zenoss.PreUpgrade30 from?
[05-Jul-2010 11:18:11] pretty sure that's an enterprise only ZenPack
[05-Jul-2010 11:18:34] for now anyways
[05-Jul-2010 11:18:44] If you're an enterprise user talk to Zenoss support
[05-Jul-2010 11:18:52] if not, wait until 3.0 is actually officially released
[05-Jul-2010 11:26:08] ah i see
[05-Jul-2010 11:44:16] hey guys, if i setup 3.0 beta will i be able to upgrade to 3.0 final
[05-Jul-2010 11:50:16] Parabol: in theory yes, but I wouldn't recommend doing that
[05-Jul-2010 11:50:41] Parabol: If you're setting up a prod system, use 2.5.2 for now
[05-Jul-2010 11:50:55] and upgrade to 3.0 or wait for bugfix releases such as 3.0.1 or 3.0.2 before upgrading
[05-Jul-2010 11:51:06] if you're just making a server to mess around on then by all means
[05-Jul-2010 11:51:14] no its for production
[05-Jul-2010 11:51:20] then don't use beta
[05-Jul-2010 11:51:23] that's just dumb lol
[05-Jul-2010 11:51:31] 3.0 seems more polished
[05-Jul-2010 11:51:43] it has a prettier interface yes
[05-Jul-2010 11:51:46] my employer saw a video on youtbue
[05-Jul-2010 11:51:51] dynamic somthing assurance
[05-Jul-2010 11:51:52] but there are underlying bugs
[05-Jul-2010 11:51:54] and hes all omg omg
[05-Jul-2010 11:52:04] there's a reason why it's a beta
[05-Jul-2010 11:52:06] he wants me to scrap the 2.5.2 and setup 3.0 lol
[05-Jul-2010 11:52:17] tell him to hold his horses
[05-Jul-2010 11:52:38] 3.0 will be out soon enough, no point in jumping the gun
[05-Jul-2010 11:54:55] well if we do all this work on 2.5.2
[05-Jul-2010 11:55:07] well, they've fixed some bugs since the last beta and introduced some new ones
[05-Jul-2010 11:55:11] time to make a list and log tickets
[05-Jul-2010 11:55:31] and then have to throw it all out
[05-Jul-2010 11:55:46] because we have to do all this work over again
[05-Jul-2010 11:56:50] why would you have to throw it all out?
[05-Jul-2010 11:56:53] it's called an upgrade
[05-Jul-2010 11:56:56] the work carries over
[05-Jul-2010 11:57:12] so you can upgrade to 3.0 from 2.5
[05-Jul-2010 11:57:13] all that changed is the interface, the underlying functionality is still very much the same from the old interface to the new one
[05-Jul-2010 11:57:20] of course
[05-Jul-2010 11:57:29] you can always upgrade from previous versions to newer ones
[05-Jul-2010 11:57:35] thats odd
[05-Jul-2010 11:57:41] how is that odd lol?
[05-Jul-2010 11:57:42] you can go from 2.5 to 3, but not from 3 beta to 3 final?
[05-Jul-2010 11:57:49] well who the hell makes somthing that does that
[05-Jul-2010 11:57:54] I'm not saying that you can't...
[05-Jul-2010 11:57:58] oh okay
[05-Jul-2010 11:58:05] so an upgrade from 2.5 to 3.0 is not recommended either
[05-Jul-2010 11:58:07] I'm saying that they don't really test that, and that it's inadvisable to do that
[05-Jul-2010 11:58:17] which means the answer is you cant upgrade
[05-Jul-2010 11:58:22] and im back to my original problem
[05-Jul-2010 11:58:27] an upgrade from 2.5 to 3.0 is advisable once 3.0 is actually officially released
[05-Jul-2010 11:58:29] it's not released
[05-Jul-2010 11:58:32] it's a beta
[05-Jul-2010 11:58:38] i live on betas :P
[05-Jul-2010 11:58:41] my phone runs nightly builds lol
[05-Jul-2010 11:58:48] that's horribly bad practice for production systems
[05-Jul-2010 11:58:56] but if that's how you do things, have at e'r
[05-Jul-2010 11:59:18] nono
[05-Jul-2010 11:59:20] not work, personal
[05-Jul-2010 11:59:24] i always go for betas at home :)
[05-Jul-2010 11:59:36] you WILL be able to upgrade cleanly from 2.5 to 3.0 once it's released
[05-Jul-2010 11:59:38] patience
[05-Jul-2010 11:59:39] lol
[05-Jul-2010 11:59:55] thats all i needed :) you're the man
[05-Jul-2010 12:00:01] thanks lol
[05-Jul-2010 12:02:50] 110% the man
[05-Jul-2010 12:03:07] neat product tho, once i get this all setup this hsould be a good free alternative to ipmonitor
[05-Jul-2010 12:03:43] yeh, it is a good product
[05-Jul-2010 12:06:06] solarwinds does have fancier display of stats and stuff though
[05-Jul-2010 12:06:46] Zenoss is extremely extendable though, you can do almost anything with it
[05-Jul-2010 12:08:07] its looking that way
[05-Jul-2010 12:08:14] i'm trying to wrap my head around it :)
[05-Jul-2010 12:08:20] this is only my 3rd day with the product
[05-Jul-2010 12:08:24] i've got WMI working, SNMP working
[05-Jul-2010 12:08:35] working on a zenpack that will mmonitor your SSL certs
[05-Jul-2010 12:08:37] for experation :D
[05-Jul-2010 12:08:44] talking to the developer now
[05-Jul-2010 12:08:55] had some problems getting it installed, now that its installed its not showing up
[05-Jul-2010 12:09:55] ah
[05-Jul-2010 12:10:20] that's one thing that'll be interesting for the 3.0 upgrades, migrating the ZenPacks properly
[05-Jul-2010 12:10:44] which is why I'm advising people to wait for the official release so that more experienced users can come up with an effective solution
[05-Jul-2010 12:11:23] I'm in the same boat as you, I'll have 14 Production Zenoss servers to upgrade from 2.5 to 3.0
[05-Jul-2010 12:11:47] so I'll be doing a lot of testing to get the absolute best method before performing the actual upgrades
[05-Jul-2010 12:12:11] Zenoss can be a bit testy
[05-Jul-2010 12:43:51] hmmm, the IpService monitoring seems to have greatly improved in 3.0
[05-Jul-2010 12:44:10] excellent
[05-Jul-2010 12:45:12] I have a question, been reading manuals/tutorials over the weekend with no luck :[
[05-Jul-2010 12:46:59] We have a script that returns the 1 / 0 with the threshold being 0 clear and 1 'somethings up'. We want to return a custom string to the event from the script though. IE: "please check apache.broken in /var/tmp/" instead of the default event message for thresholds reached "threshold of rsi.https-500-log.monitor exceeded: current value 1.00"
[05-Jul-2010 12:47:22] pretty sure that's whatever you put before the | in nagios format
[05-Jul-2010 12:47:34] THIS IS THE MESSAGE|
[05-Jul-2010 12:47:37] or whatever
[05-Jul-2010 12:47:55] the command datasource section of the admin guide should explain that
[05-Jul-2010 12:48:45] nice, so I can't find anywhere to configure alerts in the 3.0 interface
[05-Jul-2010 12:48:46] Thanks rmatte, I'll have a look
[05-Jul-2010 12:48:49] np
[05-Jul-2010 12:51:57] hmmm, no developers around today
[05-Jul-2010 12:51:58] shame
[05-Jul-2010 12:52:12] I wanted to run some stuff by them to make sure they are actually bugs before logging them
[05-Jul-2010 12:52:20] guess I'll just log them anyways
[05-Jul-2010 12:56:35] ah, I see why I couldn't see alerting rules
[05-Jul-2010 12:56:47] admin can't modify alerting rules for other users anymore, which is bad
[05-Jul-2010 12:57:02] woah
[05-Jul-2010 12:57:04] that's very bad
[05-Jul-2010 12:57:09] yup
[05-Jul-2010 12:57:19] I'll add it to my list of bugs to log
[05-Jul-2010 12:57:22] I already have about 10
[05-Jul-2010 12:57:37] next week is beta playing week for me
[05-Jul-2010 12:57:48] yeh, this week is it for me since I actually have time to
[05-Jul-2010 12:57:57] going to log a bunch of trac tickets
[05-Jul-2010 13:20:52] Hey your fave guy is back :)
[05-Jul-2010 13:21:07] so, under performance tab, CPU graph shows up, RAM does not
[05-Jul-2010 13:21:12] but under OS i can see free ram and such
[05-Jul-2010 13:22:35] for what type of device?
[05-Jul-2010 13:23:13] sorry
[05-Jul-2010 13:23:16] windows server 2008 R2 x64
[05-Jul-2010 13:23:23] ONLY THE LATEST AND GREATEST FOR ME
[05-Jul-2010 13:23:28] being monitored via WMI?
[05-Jul-2010 13:23:47] yup
[05-Jul-2010 13:23:51] under OS stats show up
[05-Jul-2010 13:23:52] check the wmi device template to see if it even supports monitoring memory utilization
[05-Jul-2010 13:23:55] that may not have been added
[05-Jul-2010 13:24:01] it does support it
[05-Jul-2010 13:24:03] it works
[05-Jul-2010 13:24:08] just under Perf the graph is broken
[05-Jul-2010 13:24:12] not as in all zeros
[05-Jul-2010 13:24:19] how is it broken?
[05-Jul-2010 13:24:20] as in , a little magnifying glass shows up lol
[05-Jul-2010 13:24:26] like its not there
[05-Jul-2010 13:24:29] but CPU one shows up just fine
[05-Jul-2010 13:24:38] when did you add that device?
[05-Jul-2010 13:24:42] 3 days ago
[05-Jul-2010 13:24:54] ok, you'll want to talk to bigegor about that when he's around
[05-Jul-2010 13:25:03] I don't do any WMI monitoring with that pack so I have no idea
[05-Jul-2010 13:25:19] it's probably something simple to fix
[05-Jul-2010 13:26:50] alright
[05-Jul-2010 13:27:09] you can also check the forum to see if anyone has had similar issues
[05-Jul-2010 13:27:15] and if not, post there and see what responses you get
[05-Jul-2010 13:27:20] that's the best I can suggest
[05-Jul-2010 13:41:58] you're the man, yet again
[05-Jul-2010 13:41:58] thank you rmatte
[05-Jul-2010 13:47:55] np
[05-Jul-2010 14:13:16] bigegor: around ?
[05-Jul-2010 14:13:37] zykes: hi
[05-Jul-2010 14:15:44] wondering, your mysql odbc thing, what permission does the user need ?
[05-Jul-2010 14:19:03] woot, my paintball gear just arrived
[05-Jul-2010 14:19:13] zykey:I think, the same like others mysql ZenPacks.
[05-Jul-2010 14:19:53] bigegor: hey man
[05-Jul-2010 14:19:59] i was told to talk to you :) do you a minute?
[05-Jul-2010 14:20:06] having a WMI issue with server 2008 R2
[05-Jul-2010 14:20:45] bigegor: that is ?
[05-Jul-2010 14:21:28] Parabol: sorry, i don't have 2008 in my environment.
[05-Jul-2010 14:24:05] bigegor: well it may not be an 08 issue
[05-Jul-2010 14:24:15] the issue is in the Perf tab, the RAM graph doesnt show up at all
[05-Jul-2010 14:24:42] my CPU one is broken too, which im working on that, it shows up but theres some RRD error
[05-Jul-2010 14:29:05] Parabol: first of all, in which Device Class you have put your server?
[05-Jul-2010 14:29:48] beigegor: servers/windows/Terminal Servers
[05-Jul-2010 14:29:49] why
[05-Jul-2010 14:30:09] zykes-: do you mean My SQL permissions?
[05-Jul-2010 14:31:10] Parabol: unbind Device template and bind WMIDevice template to your server
[05-Jul-2010 14:31:33] bigegor: correct
[05-Jul-2010 14:32:09] im sorry how do i do that, i've been using this product for a total of 3 days now :)
[05-Jul-2010 14:35:21] zykes-: SELECT must be granted
[05-Jul-2010 14:36:20] Parabol: go to the templates section for the class (the templates tab)
[05-Jul-2010 14:36:26] then select bind templates from the dropdown
[05-Jul-2010 14:36:35] then hold down ctrl and select the templates that you want
[05-Jul-2010 14:36:57] in this case you want WMIDevice
[05-Jul-2010 14:37:46] thanks, rmatte ;)
[05-Jul-2010 14:39:45] um WMI isnt in that bo
[05-Jul-2010 14:39:46] box*
[05-Jul-2010 14:40:06] theres only one entry, and its the classes "path"
[05-Jul-2010 14:40:12] so /servers/windows/terminal services/
[05-Jul-2010 14:40:19] and the rest is cut off, what should be in tehre
[05-Jul-2010 14:41:21] and in the list of availible templates I dont see WMI
[05-Jul-2010 14:41:50] did you not install the WMI Performance Template ZenPack?
[05-Jul-2010 14:42:01] performance, no i got the WMI data zenpack though
[05-Jul-2010 14:42:01] Just installing the WMI Data Source ZenPack alone is not enough
[05-Jul-2010 14:42:06] oooo
[05-Jul-2010 14:42:09] bigegor: i got that
[05-Jul-2010 14:42:10] fail!
[05-Jul-2010 14:42:10] i think
[05-Jul-2010 14:42:13] i'll test it
[05-Jul-2010 14:42:21] the WMI Data Source ZenPack just adds a WMI daemon and the WMI Data Source type
[05-Jul-2010 14:42:31] you still need to get the other pack to have the pre-built template
[05-Jul-2010 14:42:57] Parabol: Ok, now you have two options: 1. install SNMP Informant (http://www.snmp-informant.com/) or install WMIDataSource and WMIPerf_Windows ZenPacks
[05-Jul-2010 14:43:54] or use my Windows SNMP Performance ZenPacks
[05-Jul-2010 14:43:58] ;)
[05-Jul-2010 14:44:13] i just downloaded your performance zenpack egor
[05-Jul-2010 14:44:13] but he obviously wants to use WMI
[05-Jul-2010 14:44:15] will that work?
[05-Jul-2010 14:44:18] so he just needs the WMIPerf pack
[05-Jul-2010 14:45:34] Parabol: you have to install WMIDataSource ZenPack first
[05-Jul-2010 14:46:52] Error occured: [Failure instance: Traceback (failure with no frames): Products.ZenHub.PBDaemon.RemoteException: Remote exception: exceptions.AttributeError: 'ZenPack' object has no attribute '__of__'Traceback (most recent call last): File "/usr/local/zenoss/zenoss/Products/ZenHub/PBDaemon.py", line 75, in inner return callable(*args, **kw) File "/usr/local/zenoss/zenoss/Products/ZenHub/services/ThresholdMixin.py", line 24, in remote_ge
[05-Jul-2010 14:47:03] i get this after installing the perf zenpack
[05-Jul-2010 14:47:08] when i remodleed the device
[05-Jul-2010 14:47:27] Parabol: restart Zenoss
[05-Jul-2010 14:47:43] Parabol: Zenoss needs to be restarted after installing certain ZenPacks
[05-Jul-2010 14:47:49] the server its self?
[05-Jul-2010 14:47:54] no, just zenoss
[05-Jul-2010 14:47:59] not the whole server
[05-Jul-2010 14:48:12] just become the zenoss user and do zenoss restart from the commandline
[05-Jul-2010 14:48:22] then zenoss status afterwards to make sure everything is running
[05-Jul-2010 14:48:28] to become the zenoss user do: sudo su - zenoss
[05-Jul-2010 14:48:30] from the commandline
[05-Jul-2010 14:48:43] bah i dont have ssh enabled
[05-Jul-2010 14:49:12] back to vcenter
[05-Jul-2010 14:49:12] brb
[05-Jul-2010 14:49:12] enable it
[05-Jul-2010 14:49:15] k
[05-Jul-2010 14:53:15] http://www.explosm.net/db/files/Comics/Matt/On-behalf-of-men-everywhere,-sorry-ladies.-We-just-cant-help-it..png
[05-Jul-2010 14:54:39] bigegor: funny thing is that I've got user zenoss-monitor added to the db server but it's not updating graphs
[05-Jul-2010 14:55:11] and under "Software > databases" all databases has size == 0 gb
[05-Jul-2010 14:58:09] zykes-: create on zenoss server DSN for you MySQL server in /etc/odbc.ini, then run isql from commandline as zenoss user (isql YOURDSN zenoss-monitor pasword)
[05-Jul-2010 14:58:19] Apachez: nice, lol
[05-Jul-2010 14:59:09] dsn how ?
[05-Jul-2010 15:01:23] rmatte :(
[05-Jul-2010 15:01:35] i installed the perf zenpack, and i already had the datasource pack
[05-Jul-2010 15:01:38] i remodeled, no errors
[05-Jul-2010 15:01:44] but the CPU graphs are still broken, missing RRD
[05-Jul-2010 15:01:48] and the RAM one doesnt show up at all
[05-Jul-2010 15:01:57] its empty, just a little blue icon where the graph should be
[05-Jul-2010 15:01:59] you didn't bind the template like we explained earlier
[05-Jul-2010 15:02:14] go to the templates section of the class and select bind template from the dropdown menu
[05-Jul-2010 15:02:25] then select the WMIDevice template (the one you were saying didn't show up earlier)
[05-Jul-2010 15:02:27] and bind it
[05-Jul-2010 15:06:09] ooh yes
[05-Jul-2010 15:06:10] my fault :(
[05-Jul-2010 15:06:15] sorry im doing 4 things at once
[05-Jul-2010 15:06:27] lol
[05-Jul-2010 15:09:17] zykes-: have you file /home/zenoss/.odbc.ini on your zenoss server?
[05-Jul-2010 15:09:36] alright now they show up
[05-Jul-2010 15:09:53] good
[05-Jul-2010 15:09:53] but they are listed twice, and it looks like the original entry is still there at the bottom with the missing RRD message attached
[05-Jul-2010 15:10:06] that's because you bound both templates probably
[05-Jul-2010 15:10:29] only bind the WMIDevice template
[05-Jul-2010 15:10:35] LOL
[05-Jul-2010 15:10:36] smart man
[05-Jul-2010 15:10:49] I auto-discovered my devices, but how can i move them into the network\switches? =)
[05-Jul-2010 15:10:58] fOrsberg is now known as forsberg
[05-Jul-2010 15:11:01] Zfield: by hand?
[05-Jul-2010 15:11:16] you mean remove them then manually add them?
[05-Jul-2010 15:11:23] no
[05-Jul-2010 15:11:33] i looked for the option, and i can't seem to find it to change it
[05-Jul-2010 15:11:35] =(
[05-Jul-2010 15:11:45] I mean go to the device list, put a checkmark next to the ones you want to move and select the options from the dropdown to change class
[05-Jul-2010 15:11:48] it's not difficult
[05-Jul-2010 15:11:57] lol im sure it isn't
[05-Jul-2010 15:12:01] it's under the manage menu I believe
[05-Jul-2010 15:12:25] Manage -> Change Class
[05-Jul-2010 15:12:42] oh not even
[05-Jul-2010 15:12:47] it's just "Move to Class..."
[05-Jul-2010 15:12:50] :)
[05-Jul-2010 15:13:19] * Zfield is blind =(
[05-Jul-2010 15:13:30] * rmatte recommends lasik
[05-Jul-2010 15:13:32] :P
[05-Jul-2010 15:13:33] say i click on my switch
[05-Jul-2010 15:13:40] :)
[05-Jul-2010 15:13:46] lol
[05-Jul-2010 15:13:53] ok, so you click on your switch...
[05-Jul-2010 15:13:58] then where?
[05-Jul-2010 15:14:11] I'm explaining how to move many at once
[05-Jul-2010 15:14:20] click on Device List on the left hand menu
[05-Jul-2010 15:14:24] well.. that just makes it easier!
[05-Jul-2010 15:14:25] =)
[05-Jul-2010 15:14:27] put checkmarks next to the devices you want to move
[05-Jul-2010 15:14:41] then from the dropdown select "Move to Class..."
[05-Jul-2010 15:14:44] then select the class
[05-Jul-2010 15:15:11] ahhhhhhhhhhhhhhhhhhhhhhh
[05-Jul-2010 15:15:15] If you want to do it from an actual device page (by clicking on a device), it would be Manage -> Change Class...
[05-Jul-2010 15:15:19] i didn't notice the dropdown at all
[05-Jul-2010 15:15:25] you can do the same for groups, locations, etc...
[05-Jul-2010 15:15:43] yeh, the dropdown is hidden at first
[05-Jul-2010 15:15:53] once you know it's there though you're good
[05-Jul-2010 15:17:31] thanks
[05-Jul-2010 15:17:58] np
[05-Jul-2010 15:22:56] after the dsn stuff then bigegor ?
[05-Jul-2010 15:23:05] so
[05-Jul-2010 15:24:51] zykes-: can you connect to MySQL server with isql command?
[05-Jul-2010 15:26:11] yeah, to the zenoss one
[05-Jul-2010 15:26:35] i'll test with the other one
[05-Jul-2010 15:40:13] what the hell
[05-Jul-2010 15:40:19] i wiped system/data/cache
[05-Jul-2010 15:40:24] i installed p3's ROM
[05-Jul-2010 15:40:29] and i swear to god i had txt messages
[05-Jul-2010 15:40:33] it didnt have me activate
[05-Jul-2010 15:40:37] what happened?
[05-Jul-2010 15:41:36] wrong channel
[05-Jul-2010 15:41:37] fuck
[05-Jul-2010 15:45:58] http://www.cryptome.org/nsa-drive-by-176.jpg
[05-Jul-2010 16:25:48] note to those using community NetApp zenpack to update ethernetCsmacd template so that it graphs correct values: message/50279#50279
[05-Jul-2010 17:02:42] hi there , how can i remove the event logs checking for wiondows devices ?
[05-Jul-2010 17:02:47] the wmi one
[05-Jul-2010 17:06:04] orioni: zWinEventlog=False
[05-Jul-2010 17:07:36] yes but where its defined that ?
[05-Jul-2010 17:07:43] i cannot find that parameter
[05-Jul-2010 17:08:18] orioni: zProperty of Device or Device Class
[05-Jul-2010 17:08:42] thanx man
[05-Jul-2010 17:08:43] ok bigegor i've verified that it works.
[05-Jul-2010 17:08:44] just foudn that
[05-Jul-2010 17:08:52] the DSN thing works fine
[05-Jul-2010 17:09:02] but it still doesn't update
[05-Jul-2010 17:10:29] zykes-: try run as zenoss: zenperfodbc run -v10 -d yourserver
[05-Jul-2010 17:10:47] as zenoss user
[05-Jul-2010 17:12:53] and the last question
[05-Jul-2010 17:13:06] the map of network is not showing on the home page of zennos
[05-Jul-2010 17:13:16] any check / modification i have to do ?
[05-Jul-2010 17:13:49] http://pastebin.org/384037
[05-Jul-2010 17:16:21] zykes-: unbind MySqlDatabase template from device
[05-Jul-2010 17:17:00] ok ?
[05-Jul-2010 17:17:08] shouldn't that be there Å
[05-Jul-2010 17:19:11] zykes-: MySqlDatabase is components template for MySQL database component, MySQL - device template, you must bind only MySQL template to your server.
[05-Jul-2010 17:19:16] sizes still show 0 gb though
[05-Jul-2010 17:19:38] ah
[05-Jul-2010 17:19:40] wayt 15 min (3 collection cycles)
[05-Jul-2010 17:19:44] wait
[05-Jul-2010 17:19:49] any idea why the devices are not shown on the zenoss map , not google map
[05-Jul-2010 17:20:46] dpes it update the databases automatically ?
[05-Jul-2010 17:20:49] if new are added
[05-Jul-2010 17:23:03] zykes-: after device remodeling, if you add community.odbc.MySqlDatabaseMap plugin to zCollectorPlugins list
[06-Jul-2010 00:00:30] [disconnected at Tue Jul 6 00:00:30 2010]
[06-Jul-2010 00:00:30] [connected at Tue Jul 6 00:00:30 2010]
[06-Jul-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]
[07-Jul-2010 00:00:30] [disconnected at Wed Jul 7 00:00:30 2010]
[07-Jul-2010 00:00:31] [connected at Wed Jul 7 00:00:31 2010]
[07-Jul-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]
[07-Jul-2010 01:17:06] fOrsberg is now known as forsberg
[07-Jul-2010 08:21:44] ckrough is now known as Guest99383
[07-Jul-2010 08:23:09] I've tried 10 times but upgrading from 2.4.5 to 2.5.2 kills my Zenoss install (Virtual Appliance).
[07-Jul-2010 08:23:15] I get the following in my event.log: "ERROR ZPublisher BeforeTraverse: Error while invoking hook: "acl_users""
[07-Jul-2010 08:23:24] all services start fine. browsing to the site gives me a pretty generic "An error was encountered while publishing this resource. The requested resource does not exist."
[07-Jul-2010 09:19:19] Modeling servers in China takes a long long time -_-
[07-Jul-2010 09:20:23] cause they need to tunnel under the "great wall" :P
[07-Jul-2010 09:21:12] No..because its 700ms latency
[07-Jul-2010 09:22:21] ouch
[07-Jul-2010 09:23:08] you need thicker pipes, and higher pressure
[07-Jul-2010 09:24:09] how big is everyones events database?
[07-Jul-2010 09:24:46] <|rt|> jb that's a bit personal isn't
[07-Jul-2010 09:24:50] <|rt|> :P
[07-Jul-2010 09:24:54] :)
[07-Jul-2010 09:25:27] 200M (yeah, so many contacts on facebook ...)
[07-Jul-2010 09:27:15] Anyone have a solution to adding devices to Zenoss with a duplicate IP address?
[07-Jul-2010 09:27:46] ugh, I have one device that keeps insisting has dead processes, although the process indicators on the OS page are green.. deleting/readding the device did nothing
[07-Jul-2010 09:27:54] Process Down events just keep getting generated
[07-Jul-2010 09:28:14] Jane_Curry: i had a link how to do it through dmd, but never tried. Shall i search it?
[07-Jul-2010 09:28:14] Morning all, I've created a process (that checks for an ldap process name (slapd)) two of my ldap servers find/model it but the third refuses, is there any thing i can check for? Logs?
[07-Jul-2010 09:28:43] There was a comment back in Jan 08 - message/15469#15469 - but its a bit sparse
[07-Jul-2010 09:29:42] Jane_Curry: http://nettikconsulting.wordpress.com/2010/03/19/zenoss-more-than-one-device-with-a-single-ip/
[07-Jul-2010 09:44:04] Argh..Zenoss thinks the two new Windows servers I added are supposed to run http..but they arent
[07-Jul-2010 09:49:02] fragfutter: That's magic - many thanks
[07-Jul-2010 09:49:07] * Simon4 beams
[07-Jul-2010 09:49:13] that's my blog :D
[07-Jul-2010 09:49:49] Simon4: - you mean the one about duplicate IP addresses??
[07-Jul-2010 09:49:54] Jane_Curry: yeah
[07-Jul-2010 09:50:46] then many thanks to you too - exactly what I need to do -
[07-Jul-2010 09:51:09] looks like you posted it about 3 months back - any bad experiences since or is it safe?
[07-Jul-2010 09:51:24] it seems fine, we have had boxes added in that way running for more months than that
[07-Jul-2010 09:53:01] it came up a some time and i filed it under "nice, need to try if i need it" ;)
[07-Jul-2010 09:59:06] I removed a threshold from a template..how long does that take to go in to effect?
[07-Jul-2010 10:37:19] hey hey
[07-Jul-2010 10:40:59] |/Status/Wbem||4|Could not get WMI Instance (Received NT code 0x800700a4 from query: SELECT LoadPercentage FROM Win32_Processor).
[07-Jul-2010 10:41:06] any ideas on this , it keeps popping up
[07-Jul-2010 10:49:25] Parabola: 0x800700A4 No more threads can be created in the system. ERROR_MAX_THRDS_REACHED
[07-Jul-2010 10:52:23] egor, thank you mate
[07-Jul-2010 10:53:02] i'm googling now, its not really helping lol
[07-Jul-2010 10:53:10] is this a setting on the windows boxes i need to change
[07-Jul-2010 10:53:13] or am i just up shits creek?
[07-Jul-2010 10:55:27] i found this too, message/46701#46701
[07-Jul-2010 10:55:56] Parabola: have you another WMI monitoring tool? I heve seeng one MSSQL monitoring tool which sent WMI Request to server every minute.
[07-Jul-2010 10:56:11] hmm
[07-Jul-2010 10:56:23] he has an old zabbix box still
[07-Jul-2010 10:56:28] but he told me its only pinging devices
[07-Jul-2010 10:56:29] no wmi
[07-Jul-2010 10:56:30] or snmp
[07-Jul-2010 10:57:30] i can track that down, is that what causes it though?
[07-Jul-2010 10:57:41] if that ends up not being the case, what can i do to increase the threshold
[07-Jul-2010 10:58:42] Parabola: you can simply try to restart winmgmt service on target server
[07-Jul-2010 10:59:05] uh oh, i wonder if it could be the vmware tools
[07-Jul-2010 10:59:13] pulling data for the ESX host
[07-Jul-2010 11:03:48] is there a wiki page (or something) that people have dumped their event transforms into as examples others can use? this thread (thread/7829?start=0&tstart=0) referenced this page (http://www.zenoss.com/community/wiki/EventTransforms) but I can't find the "new" one?
[07-Jul-2010 11:10:36] There are some examples in my Zenoss Event Management paper - docs/DOC-3538
[07-Jul-2010 11:19:53] thanks!
[07-Jul-2010 11:27:03] so, what is the release milestone 'Avalon' denoting? 3.1?
[07-Jul-2010 11:43:44] <|rt|> I just bought Zenoss Core Network and System Monitoring...are their any other books out worth reading on Zenoss?
[07-Jul-2010 11:44:12] <|rt|> I'm thinking about getting a book on snmp just to help dive into that a bit more....but figured I'd start with the zenoss book
[07-Jul-2010 12:05:10] I have two /Status/WinService events (zenwin) for a device that I deleted 2 weeks ago still showing up every 10 mins. I have Cleared all Heartbeats and restarted zenoss, zenwin, the server, etc. Any ideas how I can make these go away fro good?
[07-Jul-2010 13:40:51] I have two /Status/WinService events (zenwin) for a device that I deleted 2 weeks ago still showing up every 10 mins. I have Cleared all Heartbeats and restarted zenoss, zenwin, the server, etc. Any ideas how I can make these go away for good?
[07-Jul-2010 13:41:10] yeh, thats very annoying.
[07-Jul-2010 13:41:26] you probably need to reindex
[07-Jul-2010 13:41:44] "dmd.Services.reIndex()" then "commit()" in zendmd.
[07-Jul-2010 13:47:54] thx will try that
[07-Jul-2010 13:54:45] jb: no dice, looks like they are still there (restarted zenoss after doing that reindex)
[07-Jul-2010 15:02:30] various 2.2 and 2.4 era zenpacks are giving me a primaryAq() missing error when doing (auto)modeling - e.g. the AP9730 power unit pack. Is this a well known compatibility issue with 2.5 v.s. older ?
[07-Jul-2010 15:30:25] rmatte: You around?
[07-Jul-2010 17:18:52] is there a way in zenoss core to run a command when a condition is met
[07-Jul-2010 17:18:57] ie, a service stops, so restart it
[07-Jul-2010 17:22:44] Parabola: not that I know of
[07-Jul-2010 17:22:50] Simon4_ is now known as Simon4
[07-Jul-2010 17:23:28] wow, really?
[07-Jul-2010 17:23:33] i've got a server where mydns keeps dying
[07-Jul-2010 17:23:43] it would be great if i could get emailed alerts on the services running on the linux box
[07-Jul-2010 17:23:48] as well as having it restart them :(
[07-Jul-2010 17:24:00] so you can be alerted to their failure, but not so much the restarting
[07-Jul-2010 17:24:31] well thats a start :)
[07-Jul-2010 17:24:43] add said process to /Processes
[07-Jul-2010 17:25:04] and configure appropriately, then remodel devices where the process is running
[07-Jul-2010 17:25:13] how would i name this to make it work
[07-Jul-2010 17:25:21] do i need the absolute path to the file running
[07-Jul-2010 17:25:31] you shouldn't, just hte proc name
[07-Jul-2010 17:28:15] zcommand
[07-Jul-2010 17:28:23] i assume that would be where i put the creds for SSH right?
[07-Jul-2010 17:38:58] Parabola: http://www.google.co.uk/url?sa=t&source=web&cd=3&ved=0CB8QFjAC&url=http%3A%2F%2Fwww.skills-1st.co.uk%2Fpapers%2Fjane%2Fprocess_monitoring.pdf&ei=8QE1TJOqCYqOjAfh1LjfBw&usg=AFQjCNGR-ynT6VeDj-l5ZmJQ7maansJdOw&sig2=E7YF2AJ2Rirg-vx4gIjO7w
[07-Jul-2010 17:39:08] oops
[07-Jul-2010 17:39:12] wrong url
[07-Jul-2010 17:39:29] www.skills-1st.co.uk/papers/jane/process_monitoring.pdf
[07-Jul-2010 17:39:36] that's the one
[07-Jul-2010 17:40:18] good explanation on using ssh to run commands on devices using ssh
[07-Jul-2010 17:41:51] you can tweak it so zenoss will run a script on a device when a condition is met
[07-Jul-2010 17:42:07] like a service not running
[07-Jul-2010 17:43:43] thanks man :)
[07-Jul-2010 17:43:46] im reading it now
[07-Jul-2010 17:44:50] www.skills-1st.co.uk/papers/jane
[07-Jul-2010 17:44:57] some good reads there
[07-Jul-2010 17:45:03] I have two /Status/WinService events (zenwin) for a device that I deleted 2 weeks ago still showing up every 10 mins. I have Cleared all Heartbeats and restarted zenoss, zenwin, the server, etc, also re-indexed services with zmdmd. Any ideas how I can make these go away for good?
[07-Jul-2010 17:46:16] DJDeFi: did you just delete the device?
[07-Jul-2010 17:47:48] It was a few weeks ago, i believe it was just deleted yes.... I had tried to re-add it but I no longer have the WMI credentials
[07-Jul-2010 17:48:15] search for it's ip address under Networks
[07-Jul-2010 17:48:21] and delete it from there
[07-Jul-2010 17:49:29] I found on a few occasions that just removing the device from the /Devices folder doesn't quite do it
[07-Jul-2010 17:50:42] hmm no luck, it still shows up in the event console as the device name rather than an IP
[07-Jul-2010 17:52:16] hold on
[07-Jul-2010 17:56:07] awesome PDF
[07-Jul-2010 17:57:40] DJDeFi: what happens if you search for the device?
[07-Jul-2010 17:58:18] top right corner of the ui
[07-Jul-2010 17:58:43] no results :/
[07-Jul-2010 17:59:33] hmm
[07-Jul-2010 18:00:13] are they events for services?
[07-Jul-2010 18:00:50] yes, BackupExec services stopped events
[07-Jul-2010 18:01:44] try searching for the services themselves in the list of winservices
[07-Jul-2010 18:02:14] they might be locked, so its reporting them because they are 'down'
[07-Jul-2010 18:02:18] they are their... says Count= 0 and both are set to Monitor = False
[07-Jul-2010 18:02:21] bbiab
[07-Jul-2010 18:11:32] wow
[07-Jul-2010 18:15:23] forsberg is now known as fOrsberg
[07-Jul-2010 18:24:30] chemist you here man
[07-Jul-2010 18:30:04] anyone here familiar with linux boxes / SSH monitoring?
[07-Jul-2010 18:33:59] a bit
[07-Jul-2010 18:40:32] hey man
[07-Jul-2010 18:40:45] so i'm reading through this pdf im in about 25 pages and its too much info to soak in for me at once
[07-Jul-2010 18:41:02] may i tell you what i'm looking to do, and if yo have any hints or can summerize a bit?
[07-Jul-2010 18:45:30] hello guys, Im having some weird issue. When I run a snmpwalk from zenoos, the command output starts throwing all the info but then I get a "Command timed out for.. IP" after a minute and the whole information does not get printed. When I try the same command over cli it works fine, any advice?
[07-Jul-2010 18:48:48] not sure robert,y sorry
[07-Jul-2010 18:48:51] i'm worthless with zenoss :D
[07-Jul-2010 18:49:25] robert45: weird
[07-Jul-2010 18:49:43] yeah
[07-Jul-2010 18:53:52] does anyone know how to troubleshoot this kind of stuff?
[07-Jul-2010 18:55:17] robert45: you might time the cmd line command to see how long it takes
[07-Jul-2010 18:55:32] robert45: then try in Zenoss and see if maybe zenoss is just giving up too soon
[07-Jul-2010 18:55:39] perhaps some value can be adjusted to make Zenoss wait longer
[07-Jul-2010 18:55:40] not sure
[07-Jul-2010 18:56:20] I was thinking at some zenoss timeout setting because the output differs everytime I run it, not sure if there is one
[07-Jul-2010 19:02:36] robert45: How is your Zenoss configured? Are you proxying to it from Apache or something?
[07-Jul-2010 19:02:48] aclark nope
[07-Jul-2010 19:03:33] ah
[07-Jul-2010 22:49:06] some one is alive ?
[07-Jul-2010 22:53:48] i have a question about zenoss and cisco routers
[07-Jul-2010 23:02:13] <|rt|> JuStIcIa_: just ask your question and someone may answer
[07-Jul-2010 23:02:43] <|rt|> not me though....not using zenoss to monitor nay cisco gear but I'm sure there are some in here
[07-Jul-2010 23:26:56] well
[07-Jul-2010 23:27:04] i have some problems with the oid
[07-Jul-2010 23:27:23] they present me the components status in grey
[07-Jul-2010 23:28:05] and when i go to the event console i see one error
[07-Jul-2010 23:28:23] i install all mibs of cisco
[07-Jul-2010 23:28:38] and i install the zenpack too
[07-Jul-2010 23:28:48] ?
[07-Jul-2010 23:28:58] and i dont know what to do about it
[07-Jul-2010 23:40:29] I installed a zenpack that added some collectors to a specific class, can i add thoes collectors to another class?
[08-Jul-2010 00:00:30] [disconnected at Thu Jul 8 00:00:30 2010]
[08-Jul-2010 00:00:30] [connected at Thu Jul 8 00:00:30 2010]
[08-Jul-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]
[08-Jul-2010 00:06:33] hmmmmm: Disk space: xxxxxxx.xxxx.fi /data00 505.9% > 90.0%
[08-Jul-2010 00:08:28] system was upgraded and free disk space shows unknown to zenoss
[08-Jul-2010 00:20:35] did you remodel?
[08-Jul-2010 00:24:04] yes, and that fixed
[08-Jul-2010 00:26:09] this was first time this happened instantly after upgrade, we had have these alerts about once a week
[08-Jul-2010 00:27:01] usually with older solaris-servers, where remodel wont help, but this was linux
[08-Jul-2010 00:28:36] yeah im still new to zenoss
[08-Jul-2010 00:28:39] had it less than a week
[08-Jul-2010 00:28:46] i'm still trying to get stuff via SSH :) hehe
[08-Jul-2010 00:29:08] i got WMI working alright, though it appears that Vmware tools also make WMI calls, because the vms keep getting alerts for max WMI threads
[08-Jul-2010 02:51:05] fOrsberg is now known as forsberg
[08-Jul-2010 03:15:17] Guest65862 is now known as bzed
[08-Jul-2010 03:15:47] bzed is now known as Guest76666
[08-Jul-2010 04:16:05] Guest76666 is now known as bzed
[08-Jul-2010 04:16:35] bzed is now known as Guest4503
[08-Jul-2010 05:16:42] Guest4503 is now known as bzed
[08-Jul-2010 05:17:12] bzed is now known as Guest11442
[08-Jul-2010 05:37:36] Anyone in that talk ZenPack creation details???
[08-Jul-2010 05:41:58] what talk ?
[08-Jul-2010 06:01:57] hi folks.
[08-Jul-2010 06:02:24] If I have an event mapping, how can I make sure that each instance of the evt.message is written to the log?
[08-Jul-2010 06:12:41] I have created a ZenPack that creates a new type of device called service with extra attributes
[08-Jul-2010 06:13:03] I have a modeler plugin that collects SNMP data
[08-Jul-2010 06:13:16] I can create a device of my new class - no problem
[08-Jul-2010 06:13:59] but running the modeler plugin always results in a KeyError: 'deviceStatus' message
[08-Jul-2010 06:17:30] Guest11442 is now known as bzed
[08-Jul-2010 06:18:00] bzed is now known as Guest75515
[08-Jul-2010 06:18:06] wow that sounds wierd.
[08-Jul-2010 06:18:57] Jane_Curry: you're extending Device? or have re-written it?
[08-Jul-2010 06:20:30] Jane_Curry: The only reference I can find for deviceStatus (in 2.5.2) is the factory_type_information structire
[08-Jul-2010 06:22:53] Not touched Device.py - I have a new object that inherits from Device
[08-Jul-2010 06:23:02] I found the same thing for deviceStatus - not sure why it's down there when I run a modeler plugin..
[08-Jul-2010 06:23:04] yeah
[08-Jul-2010 06:23:29] I guess you've restarted all the components before doing the model?
[08-Jul-2010 06:23:45] (I know it is a numpty question for which I apologise)
[08-Jul-2010 06:28:27] Yup - upside, downside, you name-it
[08-Jul-2010 06:29:46] the thing I don't understand is why the modeller is looking at the factory_type_information.
[08-Jul-2010 06:30:13] it is almost as though something is badly formed which should be detected on the start up of zenhub.
[08-Jul-2010 06:30:56] (sorry gotta go to a meeting, I'll be back in a little bit)
[08-Jul-2010 06:51:36] let me know where you want the bundle sent to....
[08-Jul-2010 07:18:19] Guest75515 is now known as bzed
[08-Jul-2010 07:18:50] bzed is now known as Guest55410
[08-Jul-2010 07:28:07] good morning
[08-Jul-2010 07:29:04] I have linux monitoring zenpack and the addon perf pack installed, i'm getting some data, but its missing a couple of things and i cant seemt o get hem to show up
[08-Jul-2010 07:29:20] the CPU graph on the perf tab, and the drive space stats under OS
[08-Jul-2010 07:30:46] no error messages on anything
[08-Jul-2010 07:31:44] i dont mind doing the leg work, but whats a good way to track this down with no error messages?
[08-Jul-2010 07:33:04] take a look at the $ZENHOME/log/event.log
[08-Jul-2010 07:34:44] thank you st3v3o
[08-Jul-2010 07:34:50] i feel bad always bugging you guys, i really do
[08-Jul-2010 07:35:03] I ask a ton of questions...
[08-Jul-2010 07:35:24] i'm just having a hard time with this product, i've had it installed less than a week
[08-Jul-2010 07:35:56] there is a pretty good learning curve.
[08-Jul-2010 07:40:37] ah okay
[08-Jul-2010 07:40:55] theres an error in the log over and over
[08-Jul-2010 07:41:54] however i cant tell if its for this device
[08-Jul-2010 07:42:53] "ERROR renderserver failed to generate a graph traceback (most recent call last): file /usr/local/zenoss/zenoss/proudcts/zenrrd/renderserver.py line 190
[08-Jul-2010 07:43:03] in$rrdtool.graph(*gopts)
[08-Jul-2010 08:17:01] alright this is pissing me off
[08-Jul-2010 08:17:10] it wont fucking snmp scan this host, and i know snmp works
[08-Jul-2010 08:17:22] no matter what class i add it under it says "snmp monitoring off for "
[08-Jul-2010 08:17:26] what is causing this shit
[08-Jul-2010 08:17:34] i changed the zproperty, and even used other classes
[08-Jul-2010 08:19:08] Guest55410 is now known as bzed
[08-Jul-2010 08:19:16] does anyone know why on certain devices, when I remodel them the serial number that I manually set on them get set to N/A?
[08-Jul-2010 08:19:38] bzed is now known as Guest19357
[08-Jul-2010 08:24:44] Parabola: just checking, the production state of the box is something other than "Decomissioned", right?
[08-Jul-2010 08:26:08] does by any chance onyone know if i can do shaded graphs in rrd? i need to draw a horizontal area that is green for normal values and gets redder the more we reach the limit.
[08-Jul-2010 08:32:35] fragfutter: that's not possible
[08-Jul-2010 08:32:44] rmatte: damn.
[08-Jul-2010 08:32:56] fragfutter: RRD can do transparency, but not gradient
[08-Jul-2010 08:33:22] transparency depending on the value?
[08-Jul-2010 08:33:28] I've seen people try to get transparency working in Zenoss and failed as well
[08-Jul-2010 08:33:35] so it doesn't even do that very well apparently
[08-Jul-2010 08:35:00] i wanted realy small graphs (like 10x80 pixels) of my network interfaces. because you can't read a curve this small, i though of using a color coded bar. basicly fill the whole graph green and represent usage by color
[08-Jul-2010 08:35:28] like a heatmap
[08-Jul-2010 08:37:22] yeh, I see what you're getting at
[08-Jul-2010 08:37:46] the only way to do that would be with multiple datapoints representing different colors
[08-Jul-2010 08:37:54] but it would be extremely tricky
[08-Jul-2010 08:38:03] and impractical
[08-Jul-2010 08:39:36] running python against the dmd, get the interface speed from it, get the last value via rrdvalue and plot it using imlib. i wonder how many cpu cycles this would burn
[08-Jul-2010 08:47:59] wtf
[08-Jul-2010 08:48:00] User-supplied Python expression ((here.speed or 1e9) / 8 * .75) for maximum value caused error: ['ifconfig_ifInOctets', 'ifconfig_ifOutOctets']
[08-Jul-2010 08:48:51] so all of a sudden its not finding even half the info it was
[08-Jul-2010 08:48:56] i am readding the same host
[08-Jul-2010 08:48:59] to the SAME class
[08-Jul-2010 08:49:03] and nothing is showing up
[08-Jul-2010 08:49:04] fuck
[08-Jul-2010 08:49:30] this product is very angering
[08-Jul-2010 08:59:26] rmatte your monitoring windows boxes with zenoss correct? Are you monitoring exchange at all ?
[08-Jul-2010 08:59:33] Parabola: you've tried remodelling the device?
[08-Jul-2010 08:59:58] st3v3o: only exchange processes, no performance data or anything
[08-Jul-2010 09:00:05] ah
[08-Jul-2010 09:00:13] exchange has integrated monitoring...
[08-Jul-2010 09:00:26] ... if it doesn't work users come running and complain
[08-Jul-2010 09:00:40] yeh, so true
[08-Jul-2010 09:00:41] lol
[08-Jul-2010 09:01:05] I'm surprised someone hasn
[08-Jul-2010 09:01:18] hasn't made an exchange template that works with Egor's WMI Data Source pack
[08-Jul-2010 09:04:07] rmatte yeah it made it worse lol
[08-Jul-2010 09:04:36] i got it, removed the class and readded it, then the zenpack
[08-Jul-2010 09:04:41] and it worked, i dunno
[08-Jul-2010 09:04:45] however, i still cant see cpu Perf graph
[08-Jul-2010 09:04:46] its empty
[08-Jul-2010 09:04:56] and, disk space (total) shows up, but not used
[08-Jul-2010 09:05:40] yeah I'm looking to see if I can use Egor's pack to pull performance data
[08-Jul-2010 09:05:55] it can :)
[08-Jul-2010 09:06:01] the wmi perf pack works very well
[08-Jul-2010 09:06:06] this is a linux box i'm dealing with
[08-Jul-2010 09:07:09] Does anyone know how to retrieve the service key from an event? I have automatic restart of windows services almost working, but in some cases the service name is slightly different than the Event component name. For instance "TSM Client Scheduler" is the component and often times the Windows Service name, but some systems have "TSM Scheduler" as the service name.
[08-Jul-2010 09:07:18] WOW
[08-Jul-2010 09:07:20] it just magically worked
[08-Jul-2010 09:07:24] drive space is there now
[08-Jul-2010 09:07:24] Parabola: then you haven't configured SNMP correctly on the server, it's not Zenoss' fault :P
[08-Jul-2010 09:07:27] and so is CPU graph
[08-Jul-2010 09:07:30] rmatte not using snmp
[08-Jul-2010 09:07:35] its all SSH and python scripts
[08-Jul-2010 09:07:43] shit its not working
[08-Jul-2010 09:07:47] the graph shows up now
[08-Jul-2010 09:07:51] I see
[08-Jul-2010 09:07:51] but they all say nan%
[08-Jul-2010 09:08:08] eugh, ok, rule of thumb...
[08-Jul-2010 09:08:14] it takes 3 polling cycles for data to start showing up
[08-Jul-2010 09:08:18] aka 15 minutes
[08-Jul-2010 09:08:18] you need to wait at least five minutes before they have enough data.
[08-Jul-2010 09:08:25] have some patience and it'll show up
[08-Jul-2010 09:08:25] oh okay
[08-Jul-2010 09:08:30] :)
[08-Jul-2010 09:08:33] (why someone prefers ssh over snmp, no idea)
[08-Jul-2010 09:08:39] * rmatte shrugs
[08-Jul-2010 09:08:44] because i couldnt get everything working with snmp
[08-Jul-2010 09:08:47] and its hard to get help sometimes
[08-Jul-2010 09:08:48] SNMP gives a heck of a lot more monitoring capability
[08-Jul-2010 09:08:51] but to each his own
[08-Jul-2010 09:08:55] and someone had zenpacks that did what i needed
[08-Jul-2010 09:09:02] http://www.xkcd.com/
[08-Jul-2010 09:09:14] Parabola: I gave you a working SNMP config the other day and told you what lines to change
[08-Jul-2010 09:09:25] i better do not ask
[08-Jul-2010 09:10:04] that's the best xkcd I've seen in a while
[08-Jul-2010 09:10:06] lol
[08-Jul-2010 09:10:25] lmfao
[08-Jul-2010 09:12:58] thats a pretty good one...
[08-Jul-2010 09:13:09] http://www.xkcd.com/753/ this is a good classic one
[08-Jul-2010 09:17:28] I have some things in Nagios I have NO idea how to move over to zenoss
[08-Jul-2010 09:18:21] I have a lot of hosts that only get pinged..I don't want/need for them to be graphed or anything like that
[08-Jul-2010 09:19:13] In Nagios, I'll have 3 'hosts' for an external firewall. One points to its public, external IP. Another points to its backup public, external IP. The other points to the other side of its VPN tunnel.
[08-Jul-2010 09:19:31] straterra: put them in the /Ping class
[08-Jul-2010 09:19:49] Should I create three hosts like before?
[08-Jul-2010 09:19:56] Guest19357 is now known as bzed
[08-Jul-2010 09:20:04] straterra: and when you're adding them, change "Discovery Protocol" from auto to none
[08-Jul-2010 09:20:08] so that it doesn't try to model them
[08-Jul-2010 09:20:19] yes, create 3 hosts if you want 3 IPs monitored
[08-Jul-2010 09:20:26] bzed is now known as Guest67
[08-Jul-2010 09:21:41] what happens if you clean up a memory leak in a system process? users start to use the memory, two eclipses, 1gb and 1.2gb of ram.
[08-Jul-2010 09:21:54] lol
[08-Jul-2010 10:03:56] I am writing a zenpack with a specialised type of device which has extra attributes
[08-Jul-2010 10:04:31] Does anyone have examples / suggestions for extending the Edit tab of a device to permit
[08-Jul-2010 10:04:51] a user to add values for the new attributes of this device?
[08-Jul-2010 10:07:00] has anyone experienced zenoss going balistic alerting with error: snmp_open messages ?
[08-Jul-2010 10:07:50] Jane_Curry: I don't have an example, but at a guess you would want to replace editDevice.pt in your zenpack
[08-Jul-2010 10:10:44] Simon4: could you expand a little.....
[08-Jul-2010 10:11:37] Do I copy editDevice.pt from (I guess) Products/Zenmodel/skins/.... stick it in to my ZenPack skins dir and then just modify it??
[08-Jul-2010 10:12:30] Jane_Curry: that's where I would start, yeah
[08-Jul-2010 10:12:50] you could also copy/paste it out of /manage
[08-Jul-2010 10:13:13] not quite sure how to tell your zenpack to use that template as opposed to the one in zenmodel however
[08-Jul-2010 10:13:25] ooh
[08-Jul-2010 10:13:27] maybe I do
[08-Jul-2010 10:13:58] { 'id' : 'edit'
[08-Jul-2010 10:13:58] , 'name' : 'Edit'
[08-Jul-2010 10:13:58] , 'action' : 'editDevice'
[08-Jul-2010 10:13:59] , 'permissions' : ("Change Device",)
[08-Jul-2010 10:13:59] },
[08-Jul-2010 10:14:23] in your class, just change the action to "customEditDevice" or similar, save your edited skin as customEditDevice.pt and see how you get on
[08-Jul-2010 10:14:58] so for the Blade Chassis zenpack that code is in BladeChassis.py, which is my extension of the Device
[08-Jul-2010 10:20:15] Anyone here use JMX monitoring much? I don't want to open a support case for what might be a simple thing - but I just tried bringing up tomcat6 so I can fix the two ports it uses for JMX
[08-Jul-2010 10:20:22] however the jmx template only asks me for a single port
[08-Jul-2010 10:20:45] Guest67 is now known as bzed
[08-Jul-2010 10:20:51] i imagine i need to change the zenjmx template to put in a 'raw service url'
[08-Jul-2010 10:20:56] but i'm not sure that's the right thing to do :)
[08-Jul-2010 10:21:15] bzed is now known as Guest84029
[08-Jul-2010 10:43:10] Help!: "undefined symbol: rl_line_buffer"
[08-Jul-2010 11:19:52] i need some help figuring out what the regex for some processes on a linux box would be, i need to monitor /etc/init.d/mydns
[08-Jul-2010 11:20:50] nvm i get it :)
[08-Jul-2010 11:21:17] :( the linux monitor zenpack doesnt grab processes
[08-Jul-2010 11:21:34] Guest84029 is now known as bzed
[08-Jul-2010 11:22:03] bzed is now known as Guest73662
[08-Jul-2010 11:22:37] what linux monitor zenpack ?
[08-Jul-2010 11:22:48] default basic zenoss installation has that feature
[08-Jul-2010 11:25:38] well it didnt work for shit for me
[08-Jul-2010 11:25:41] so i grabbed a zenpack
[08-Jul-2010 11:25:46] it monitors all of it over SSH
[08-Jul-2010 11:26:31] hmmm well it works fine, but ok I havent played much with the ssh
[08-Jul-2010 11:26:36] monitor pack
[08-Jul-2010 11:26:42] i couldnt get the snmp to work right with this box
[08-Jul-2010 11:26:45] and rather than bugging people about it
[08-Jul-2010 11:26:48] ahh
[08-Jul-2010 11:26:50] i just used what i found that worked
[08-Jul-2010 11:26:54] i have perf tab, os stats..ect
[08-Jul-2010 11:26:59] everything works but services/processes
[08-Jul-2010 11:27:01] ok well if snmp doesnt work, then process monitoring does not either
[08-Jul-2010 11:27:08] okok
[08-Jul-2010 11:27:53] so in $ ps aux
[08-Jul-2010 11:27:57] what does the process look like ?
[08-Jul-2010 11:29:48] im looking for it
[08-Jul-2010 11:29:49] i dont see it
[08-Jul-2010 11:30:19] Simon4: Hmmm - editDevice.pt doesn't actually have anything very useful in it - can't quite see where the "real" form comes from...
[08-Jul-2010 11:30:48] Jane_Curry: hmmm, bugger
[08-Jul-2010 11:31:21] in spades!
[08-Jul-2010 11:31:39] okay i really need to get this fixed
[08-Jul-2010 11:31:40] NFO zen.ZenModeler SNMP monitoring off for 174.46.66.9
[08-Jul-2010 11:31:43] its off, why
[08-Jul-2010 11:31:49] i have ignore set to false in zproperties
[08-Jul-2010 11:31:52] and i remodeled
[08-Jul-2010 11:32:02] Can you elaborate on your "you could also copy/paste it out of /manage"
[08-Jul-2010 11:32:21] mwahahah
[08-Jul-2010 11:32:27] even though it says that it looks like it worked
[08-Jul-2010 11:32:31] should i be concerened
[08-Jul-2010 11:32:54] Simon4: Can you elaborate on your "you could also copy/paste it out of /manage"
[08-Jul-2010 11:33:04] Jane_Curry: if you go to the edit tab on a device
[08-Jul-2010 11:33:09] then add /manage to the url in your url bar
[08-Jul-2010 11:33:14] it'll show you the template
[08-Jul-2010 11:34:25] Simon4: hmm - which gives me the same nice piece of code that doesn't really have the form in it
[08-Jul-2010 11:34:34] * Simon4 ponders
[08-Jul-2010 11:35:56] here/addDevice/macros/body
[08-Jul-2010 11:36:05] * Simon4 wonders how to find what that points to
[08-Jul-2010 11:37:18] When checking for processes, two of my hosts find the slapd in : "/usr/local/libexec/slapd -h ldap:/// " yet one does not.. any clues ?
[08-Jul-2010 11:39:27] Parabola ? you cannot find the process in the process list on the box and you wonder why zenoss cant find it either ?
[08-Jul-2010 11:39:31] or am i misunderstanding
[08-Jul-2010 11:39:44] you misunderstood or i mispoke
[08-Jul-2010 11:39:47] however i got it resolved
[08-Jul-2010 11:39:49] g0rdy did you try to remodel the device ?
[08-Jul-2010 11:40:31] yes, no luck.. also, i added it manually, remodelled it and it went away -
[08-Jul-2010 11:47:25] * Simon4 goes to ask someone
[08-Jul-2010 11:52:42] Jane_Curry: Products/ZenModel/skins/zenmodel/addDevice.pt is the file
[08-Jul-2010 11:53:09] we have an in-house zenpack here that the __init.py__ install process just replaces that with a custom one, but that makes the edit page modification global to zenoss
[08-Jul-2010 11:53:23] not sure how you make it local to a specific device type, but it should be doable
[08-Jul-2010 11:54:14] possibly making your own editDevice, then changing this line to point to your own template:
[08-Jul-2010 11:54:14]
[08-Jul-2010 11:54:29] Simon4: Yeh - just found the addDevice.pt - hacking now - in the dark - one hand tied behind my back....
[08-Jul-2010 11:54:36] the best way :)
[08-Jul-2010 11:55:16] right, hometime, beer in the sun I think
[08-Jul-2010 12:00:12] howdy folks
[08-Jul-2010 12:01:11] can someone explain how zenoss pulls process memory and cpu usage
[08-Jul-2010 12:02:01] depends on what OS
[08-Jul-2010 12:02:07] linux
[08-Jul-2010 12:02:30] snmp i assume?
[08-Jul-2010 12:02:55] I see that it uses an OSProcess Template
[08-Jul-2010 12:03:16] but how does it query with no OID
[08-Jul-2010 12:03:56] MIB
[08-Jul-2010 12:04:06] might be somthing like this :)
[08-Jul-2010 12:04:06] http://www.intermapper.com/custom-probes-a-snmp/177-snmp-oids-in-custom-probes
[08-Jul-2010 12:04:08] read the link
[08-Jul-2010 12:04:40] howdy
[08-Jul-2010 12:05:07] sup jane
[08-Jul-2010 12:20:18] st3v3o: collector plugins
[08-Jul-2010 12:20:49] st3v3o: and device templates
[08-Jul-2010 12:20:55] http://docs.huihoo.com/zenoss/admin-guide/2.1.1/ch11s03.html
[08-Jul-2010 12:21:06] trying to see if I can manually pull that data
[08-Jul-2010 12:21:10] that's an ANCIENT version of the admin guide
[08-Jul-2010 12:21:21] oh ok
[08-Jul-2010 12:21:27] pull what data?
[08-Jul-2010 12:21:36] is there a limit on how long an eventKey can be?
[08-Jul-2010 12:21:51] current CPU usage for a process via SNMP
[08-Jul-2010 12:21:51] Absenth: not to my knowledge
[08-Jul-2010 12:22:03] st3v3o: Zenoss already polls that
[08-Jul-2010 12:22:22] Guest73662 is now known as bzed
[08-Jul-2010 12:22:30] st3v3o: click on any process and you'll see it
[08-Jul-2010 12:22:44] Yeah I see that under the process…but one of the guys wants to use that data point on a different graph
[08-Jul-2010 12:22:52] bzed is now known as Guest75
[08-Jul-2010 12:22:52] rmatte: writing a few scripts to monitor the status of our zfs snapshots. I think I'll use the snapshot name, as the event id.
[08-Jul-2010 12:23:18] with luck, you'll be right and there won't be a limit, since the names are ~ 50 characters :)
[08-Jul-2010 12:23:20] and I can't figure out how to use that data point in a different template
[08-Jul-2010 12:23:37] st3v3o: well, you can either make a multi-graph report with the datapoints that you want, or you'll have to come up with something custom, which won't be easy since that value is part of an OID tree
[08-Jul-2010 12:24:00] Is there a multi graph report tutorial ?
[08-Jul-2010 12:24:08] the admin guide
[08-Jul-2010 12:24:23] community/documentation/official_documentation/zenoss-guide
[08-Jul-2010 12:24:26] grab the one for 2.5
[08-Jul-2010 12:24:51] thought honestly, creating multigraph reports is pretty easy
[08-Jul-2010 12:24:54] though*
[08-Jul-2010 12:25:13] docs/DOC-4743
[08-Jul-2010 12:25:30] that explains how
[08-Jul-2010 12:53:00] Is there any way to rename a multi graph report collection item
[08-Jul-2010 12:53:40] its pulling the name from the regex query which is to long to be read on the graph
[08-Jul-2010 12:54:48] * Simon4_ thinks
[08-Jul-2010 12:55:29] st3v3o: you could change the format of the legend to use another variable?
[08-Jul-2010 12:56:11] …. *scratches head*
[08-Jul-2010 12:56:28] hokay
[08-Jul-2010 12:56:36] * Simon4_ finds where he means
[08-Jul-2010 12:57:45] I'll give a little more info…the collection Item(s) are specific Device/Component a mysql process found on specific boxes
[08-Jul-2010 12:57:45] http://monitor2.arin.net:8080/zport/dmd/Devices/Server/Linux/Whois-Proto-RWS/devices/whoisrws1.proto.arin.net/os/processes/usr_sbin_mysqld%20e9fdb9fb0483b8c995ef3dd0b9567376
[08-Jul-2010 12:58:10] ah yeah, the good old "process name gets turned into unreadable shit" thing
[08-Jul-2010 12:58:15] * Simon4_ keeps meaning to open a ticket about that
[08-Jul-2010 12:58:25] yeah and it wont let me give it a friendly name
[08-Jul-2010 12:58:30] I want it to file under the process name you specify under /Processes, not the match hash
[08-Jul-2010 12:59:59] yeah that would be nice
[08-Jul-2010 13:00:12] so no fix/ work around :?
[08-Jul-2010 13:00:25] st3v3o: if it's just for one process, you can head for the graph definition
[08-Jul-2010 13:00:32] * Simon4_ waits for the page to load :)
[08-Jul-2010 13:01:18] its for one process …but on multiple boxes
[08-Jul-2010 13:01:38] and the legend is so long it doens't read what line is linked to what box
[08-Jul-2010 13:01:39] so in the graph definition you can set the format of the legend
[08-Jul-2010 13:02:03] how does I do that
[08-Jul-2010 13:02:17] just getting there, sorry this install is being horribly slow :/
[08-Jul-2010 13:02:34] no worries….I appreciate the help a LOT
[08-Jul-2010 13:02:59] so, when you edit hte multi-graph report you have collections, and presumably a graph definition + graph group, yeah?
[08-Jul-2010 13:03:54] yeah
[08-Jul-2010 13:04:01] if you click on the graph definition, then on the datapoint within that graph definition, you'll see a field called "Legend"
[08-Jul-2010 13:04:19] yes
[08-Jul-2010 13:04:37] you could set that to something like ${here/hostname} mysql
[08-Jul-2010 13:04:42] then see what the graph looks like
[08-Jul-2010 13:04:46] so could I use ${device/id}
[08-Jul-2010 13:05:05] ok let me try hostname
[08-Jul-2010 13:05:33] Simon4 you rock
[08-Jul-2010 13:05:48] :) no probs
[08-Jul-2010 13:05:53] Simon4_ is now known as Simon4
[08-Jul-2010 13:23:11] Guest75 is now known as bzed
[08-Jul-2010 13:23:40] bzed is now known as Guest54963
[08-Jul-2010 13:41:55] forsberg is now known as fOrsberg
[08-Jul-2010 13:57:17] is there a variable like ${here/process} for the process class
[08-Jul-2010 13:58:03] or a list of these somewhere
[08-Jul-2010 13:58:10] hmm, I get em out of zendmd
[08-Jul-2010 13:58:17] maybe I should blog about this sometime :)
[08-Jul-2010 13:58:27] here = the current device
[08-Jul-2010 13:58:35] so things below that are attributes of that device
[08-Jul-2010 13:58:53] is there a way to list attributes of the current device ?
[08-Jul-2010 14:00:11] so you can do this
[08-Jul-2010 14:00:12] >>> dev.os.processes.crond.id
[08-Jul-2010 14:00:12] 'crond'
[08-Jul-2010 14:00:24] have you messed with zendmd at all?
[08-Jul-2010 14:00:31] nope
[08-Jul-2010 14:02:02] docs/DOC-2528
[08-Jul-2010 14:02:18] is an interesting article, it talks about dev.os.interfaces, but you could explore processes similarilty
[08-Jul-2010 14:02:26] zendmd you run as the zenoss user on the cmd line
[08-Jul-2010 14:02:36] and it has tab-completion
[08-Jul-2010 14:02:55] so you can do such things as dev. to see what properties/methods are available
[08-Jul-2010 14:03:12] easiest way to find a device (from my experience) is
[08-Jul-2010 14:03:17] dev. doesn't auto complete
[08-Jul-2010 14:03:19] dev = dmd.Devices.findDevice('hostname')
[08-Jul-2010 14:03:27] devices works though
[08-Jul-2010 14:23:59] Guest54963 is now known as bzed
[08-Jul-2010 14:24:32] bzed is now known as Guest16878
[08-Jul-2010 14:37:19] Any pyton/zendmd scripting people here :)?
[08-Jul-2010 14:39:04] I want to write a python for loop script that executes a command each rotation you would normally type into the zendmd
[08-Jul-2010 14:39:47] each rotation would let the zendmd run this for example: zport.dmd.ZenUsers.admin.manage_addActionRule('VARIABLE)
[08-Jul-2010 15:09:22] In the "Define Commands" section (Settings>Commands>Add Command) I go to add a command, the ID/Label takes but the actual command itself doesn't save, any ideas ?
[08-Jul-2010 15:13:47] http://i31.tinypic.com/35aozs9.jpg
[08-Jul-2010 15:24:48] Guest16878 is now known as bzed
[08-Jul-2010 15:25:17] bzed is now known as Guest49952
[08-Jul-2010 15:28:19] gah.. Anyone know how this can be done, putting a variable within a zendmd command?
[08-Jul-2010 15:28:36] zport.dmd.ZenUsers.admin.ARRAY[1].delay=600
[08-Jul-2010 15:28:47] that sort of thing
[08-Jul-2010 15:34:46] hello, does zenoss support invoking remote commands/scripts upon detecting a failure? for ex. execute /foo/bar/custom_script if apache crashes
[08-Jul-2010 15:54:00] TheBeast: yes, set them up in the "Event Manager"
[08-Jul-2010 15:54:41] https://yourserver/zport/dmd/ZenEventManager/listEventCommands
[08-Jul-2010 16:25:36] Guest49952 is now known as bzed
[08-Jul-2010 16:26:06] bzed is now known as Guest47970
[08-Jul-2010 17:13:06] trashguy_ is now known as trashguy
[08-Jul-2010 17:26:25] Guest47970 is now known as bzed
[08-Jul-2010 17:26:54] bzed is now known as Guest19289
[08-Jul-2010 18:27:13] Guest19289 is now known as bzed
[08-Jul-2010 18:27:43] bzed is now known as Guest3764
[08-Jul-2010 19:28:02] Guest3764 is now known as bzed
[08-Jul-2010 19:28:32] bzed is now known as Guest72297
[08-Jul-2010 20:28:50] Guest72297 is now known as bzed
[08-Jul-2010 20:29:20] bzed is now known as Guest94878
[08-Jul-2010 21:29:39] Guest94878 is now known as bzed
[08-Jul-2010 21:30:08] bzed is now known as Guest96696
[08-Jul-2010 22:30:27] Guest96696 is now known as bzed
[08-Jul-2010 22:30:57] bzed is now known as Guest25901
[08-Jul-2010 23:03:06] so whats the best way to find out why a monitored process isnt actually being monitored
[08-Jul-2010 23:03:13] i stopped a service and its still green :(
[08-Jul-2010 23:05:28] ap
[08-Jul-2010 23:17:33] hm?
[08-Jul-2010 23:31:16] Guest25901 is now known as bzed
[08-Jul-2010 23:31:45] bzed is now known as Guest44071
[09-Jul-2010 00:00:30] [disconnected at Fri Jul 9 00:00:30 2010]
[09-Jul-2010 00:00:30] [connected at Fri Jul 9 00:00:30 2010]
[09-Jul-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]
[09-Jul-2010 00:31:54] Guest44071 is now known as bzed
[09-Jul-2010 00:32:23] bzed is now known as Guest28723
[09-Jul-2010 01:32:42] Guest28723 is now known as bzed
[09-Jul-2010 01:33:12] bzed is now known as Guest3522
[09-Jul-2010 01:48:47] Parabola: if you disable monitoring, both the icon/button in the "M" and in the "Status" column should be grey and not green.
[09-Jul-2010 02:33:31] Guest3522 is now known as bzed
[09-Jul-2010 02:34:07] bzed is now known as Guest92222
[09-Jul-2010 02:35:22] anyone here?
[09-Jul-2010 03:27:25] i have some switchports with user-pcs. During the night some of them turn there PC of. Now when during the night i get a model run, the ports are modelled with 10mbit. Next day, user turns his pc on, pulls his profile from the net and this is done with gigbit speed, so i get an event that port limit is reached. Any idea how to handle this
[09-Jul-2010 03:34:19] Guest92222 is now known as bzed
[09-Jul-2010 03:34:49] bzed is now known as Guest98810
[09-Jul-2010 03:37:12] fOrsberg is now known as forsberg
[09-Jul-2010 04:35:08] Guest98810 is now known as bzed
[09-Jul-2010 04:35:38] bzed is now known as Guest41080
[09-Jul-2010 05:35:56] Guest41080 is now known as bzed
[09-Jul-2010 05:36:26] bzed is now known as Guest81398
[09-Jul-2010 05:50:52] hello
[09-Jul-2010 05:51:16] zenoss collects data like the cpu load nicely in my system,
[09-Jul-2010 05:51:26] but is it possible to dump the measurements to a file?
[09-Jul-2010 05:51:33] say, a CSV or similar?
[09-Jul-2010 05:51:41] have a read about rrdtool
[09-Jul-2010 05:52:05] the cpu stats are just stored in a rrd database file, which you can dump out using command-line tools into a bunch of formats
[09-Jul-2010 05:52:56] hmm
[09-Jul-2010 05:53:01] and where is the rrd database?
[09-Jul-2010 05:53:13] $ZENHOME/perf/Devices/
[09-Jul-2010 05:53:23] alright, thanks
[09-Jul-2010 06:01:44] i have some switchports with user-pcs. During the night some of them turn there PC of. Now when during the night i get a model run, the ports are modelled with 10mbit. Next day, user turns his pc on, pulls his profile from the net and this is done with gigbit speed, so i get an event that port limit is reached. Any idea how to handle this?
[09-Jul-2010 06:18:14] fragfutter: Cisco switch?
[09-Jul-2010 06:18:28] malbon: various (cisco, hp, foundry)
[09-Jul-2010 06:19:07] fragfutter: you have 2 choices, 1, turn on WOL, 2 set the speed manually from the cli on the switch.
[09-Jul-2010 06:21:26] malbon: all switchports have to be fully dynamic. Turning WoL on will ease the problem but not erase it. I have ports that are unplugged at some time and then plugged back in. I'm looking in the direction to remodel a port the moment i receive a port up trap, would that be possible? Or model ports with there maximum physical speed not the negotiated one, but i can't find an OID reporting the physical speed.
[09-Jul-2010 06:23:30] no
[09-Jul-2010 06:24:04] not unless you override the template which you can do, but I wouldn't advise it.
[09-Jul-2010 06:24:40] malbon: so there is an oid with the max physical speed?
[09-Jul-2010 06:25:48] no there isn't sadly
[09-Jul-2010 06:26:23] why can't you just use the 'speed ' on the ciscos to override the reporting of the actual link speed?
[09-Jul-2010 06:28:19] malbon: because there are 10k ports and you can plug in anything, anywhere, anytime ;)
[09-Jul-2010 06:28:58] then increase your model frequency. ;)
[09-Jul-2010 06:29:29] you could for example have a command against a 'link up' trap from each of your switches. It would also enable you to track system moves.
[09-Jul-2010 06:30:19] so if you're running a normal office and not a conf centre, the moves and changes people should only be doing <100 per day which is quite easy to plan for in terms of zenoss load.
[09-Jul-2010 06:30:47] malbon: that was my workaround idea. I would prefer only to model the port, but i think it's the best short term solution.
[09-Jul-2010 06:31:17] malbon: more then 100 moves per day.
[09-Jul-2010 06:31:19] fragfutter: no it is a horrible kludge.
[09-Jul-2010 06:31:50] fragfutter: it does sound like a bit of an extreme requirement to me.
[09-Jul-2010 06:32:00] hmmmm my zeneventlog on a dedicated collector dies with
[09-Jul-2010 06:32:05] python[24029] general protection rip:2ba49a2003aa rsp:7fffe5d9d110 error:0
[09-Jul-2010 06:32:27] malbon: monitoring a fully dynamic network, with minimal control over the clients. yes bad think.
[09-Jul-2010 06:33:29] fragfutter: what sort of dynamic network? Conferencing suite, or Univeristy, big company ???
[09-Jul-2010 06:34:03] malbon: i could hack the modelling to have a lower limit of gigabit. the higher speeds are not dynamic and the number of slower ports could be neglected (ignored that they have false values).
[09-Jul-2010 06:34:24] yes that is another approach.
[09-Jul-2010 06:34:38] you can actually do that in the template, you don't need to hack anything.
[09-Jul-2010 06:34:39] msg malbon www.gsi.de -> science institute
[09-Jul-2010 06:34:45] *gg*
[09-Jul-2010 06:34:57] * fragfutter congrats himself, now its in the history
[09-Jul-2010 06:34:59] ;)
[09-Jul-2010 06:35:14] ah that makes sense, loads of freaky researchers with their insane requirements (no offense)
[09-Jul-2010 06:35:22] yes.
[09-Jul-2010 06:36:33] malbon: i can have if-then-else for a max value?
[09-Jul-2010 06:36:45] Guest81398 is now known as bzed
[09-Jul-2010 06:36:52] I have seen something somewhere in the web about that, I am trying to remember where.
[09-Jul-2010 06:36:56] fragfutter: why do you want to monitor ports with workstations ?
[09-Jul-2010 06:37:15] bzed is now known as Guest91001
[09-Jul-2010 06:37:40] tehhobbit: not necessary workstation, could also be an experiment server. and if they eat up all the available bandwith i want to know
[09-Jul-2010 06:38:17] fragfutter: I think this is what you want thread/7152;jsessionid=6E1B09EF826BED2246E1057ED52E7636.node0?decorator=print&displayFullThread=true
[09-Jul-2010 06:38:46] doesnt monitor of just up and downlinks solve that
[09-Jul-2010 06:39:12] fragfutter: because if its just traffic inside the switch it shouldnt be a problem
[09-Jul-2010 06:40:42] tehhobbit: it helps pinning the offending port faster. Monitoring the uplinks will helps only to identify the switch. also to saturate the uplink you need to have more then one gigbit.
[09-Jul-2010 06:41:07] would this work? max((here.speed or 1e9), 1e9) / 8 * .75
[09-Jul-2010 06:41:43] fragfutter: well you can still co9llect performance data, just dont need to have traps / tresholds on them
[09-Jul-2010 06:42:09] that way it would be enough with a quick look under the os tab on the switch
[09-Jul-2010 06:42:28] just tie the alarms to the uplink
[09-Jul-2010 06:42:43] you mean i disable thresholds by default and add them only to the uplink ports?
[09-Jul-2010 06:43:03] yes
[09-Jul-2010 06:43:58] bleh, that's horrible too. means you have loads of templates to manage. :(
[09-Jul-2010 06:44:00] this works. max((here.speed or 1e9), 1e9) / 8 * .75
[09-Jul-2010 06:44:28] * malbon is definitely in the one-size-fits-all school of thought
[09-Jul-2010 06:44:53] malbon: yes. define once in the tree, do not specify anything localy on the device
[09-Jul-2010 06:45:24] malbon: then use traps instead of thresholds , or always use same ports for uplinks =)
[09-Jul-2010 06:45:48] tehhobbit: or do capacity planning properly?
[09-Jul-2010 06:46:11] malbon: you mean give everyone his 10gig link as requested?
[09-Jul-2010 06:47:06] malbon: not an options in fragfutters case scientists will immediately come up with something that ruins any chance of sane planning =)
[09-Jul-2010 06:47:36] i think, the max solution is good enough. if the device is modelled wrong i don't get a warning. higher speed ports are not dynamic and modelled correctly. lower speed ports can saturate their ports, they won't disturb anything
[09-Jul-2010 06:48:33] if there is a problem and you are in doubt, remodel the switch before continuing investigation.
[09-Jul-2010 06:49:47] tehhobbit: I don't neccessarily agree. We work with very high volumes of data, and with a similar environment in some ways.
[09-Jul-2010 06:50:23] malbon: i don't even know which devices will be attached.
[09-Jul-2010 06:51:28] malbon: check the gsi.de site, they take different researchers from all around, hard to do planning when you have no how,when or what
[09-Jul-2010 06:51:29] fragfutter: have you deployed 802.1x?
[09-Jul-2010 06:52:05] malbon: nerfing the users is always an option, but not always a good one
[09-Jul-2010 06:52:05] malbon: yes.
[09-Jul-2010 06:58:32] bzed_ is now known as Guest52435
[09-Jul-2010 07:41:34] Guest99383 is now known as ckrough
[09-Jul-2010 07:42:01] * Simon4 looks at some of the mysql queries that zenoss does, when just navigating around
[09-Jul-2010 07:42:09] * Simon4 starts to understand why it's so damn slow at times
[09-Jul-2010 07:54:27] re
[09-Jul-2010 08:03:44] well, when you load a device page it has to calculate availability for that device over the last 7 days I believe
[09-Jul-2010 08:03:47] which is fairly intensive
[09-Jul-2010 08:03:57] mysqlwise
[09-Jul-2010 08:08:17] forsberg is now known as fOrsberg
[09-Jul-2010 08:16:12] Hopefully someone here can help me with a problem. It's not directly zenoss related, however i think Zenoss might be part of the solution.
[09-Jul-2010 08:17:30] We have two servers, running FreeBSD with a carp interface that's shared between them. Each server is running MySQL, with one being the master, and the other a slave in a replication pair. We are using ifstated to promote whichever server is the carp master to mysql master, and demote the carp backup to mysql slave.
[09-Jul-2010 08:18:45] this works well, and allows a server to power cycle, without dropping more than a couple packets. What it doesn't do is protect us against a service failure. Say for instance MySQL on the master fails. We are using the MySQL monitor zenpack on our Zenoss server which has been working great for us.
[09-Jul-2010 08:19:30] Do any of you think an appropriate solution to this might be to have zenoss log into the server with the failed service and down the carp0 interface? Or are there other methods you would suggest I've perhaps not thought of?
[09-Jul-2010 08:25:31] Hey there - is there a way to list all the devices a specific template is bound to ?
[09-Jul-2010 08:29:27] Weetos: used zendmd before?
[09-Jul-2010 08:29:35] >>> for dev in dmd.Devices.getSubDevices():
[09-Jul-2010 08:29:36] ... if "TemplateName" in dev.zDeviceTemplates:
[09-Jul-2010 08:29:36] ... print dev.id
[09-Jul-2010 08:29:36] ...
[09-Jul-2010 08:29:37] yep
[09-Jul-2010 08:29:56] woot ! that was fast ! thank you !
[09-Jul-2010 08:30:07] np
[09-Jul-2010 08:30:24] rmatte: yeah, I just hurled the query at the dba team here with the comment of "pls make fast now kthx"
[09-Jul-2010 08:30:41] there's got to be a way to speed that stuff up, the current query does multiple table scans
[09-Jul-2010 08:30:45] yawn
[09-Jul-2010 08:31:16] http://tech.hj60.com/iostat-monitoring-for-zenoss-and-such/
[09-Jul-2010 08:34:30] ok, I broke something in 2.5. hoping someone knows the answer. when I run zenping in the forground I get this error. RemoteException: Remote exception: zExceptions.NotFound: xxxxxxxxxx Traceback (most recent call last):
[09-Jul-2010 08:34:46] Absenth: Why wouldn't you just have a script cronned to run every minute that checks whether or not MySQL is failed and downs the interface?
[09-Jul-2010 08:34:48] where xxxxxxx is a device name that has been deleted.
[09-Jul-2010 08:34:56] Absenth: I don't see where Zenoss even needs to be involved in that
[09-Jul-2010 08:35:37] Simon4> That worked like a charm, thank you again
[09-Jul-2010 08:35:45] Weetos: sweet as
[09-Jul-2010 08:36:03] ptom_: When you say you broke it, what exactly was changed?
[09-Jul-2010 08:36:33] not really sure, I just assume that i deleted a device
[09-Jul-2010 08:37:20] i tried cleaning up the catalog and reindexing
[09-Jul-2010 08:38:13] is there a cache file or something that I can delete or is this somewhere in the zope db
[09-Jul-2010 08:45:10] you might have .zec files
[09-Jul-2010 08:45:15] I believe is what the extension is
[09-Jul-2010 08:45:21] in $ZENHOME/var
[09-Jul-2010 08:45:28] those would be cache files
[09-Jul-2010 08:45:29] i thought they did away with those
[09-Jul-2010 08:45:45] They may have, I know I don't have any
[09-Jul-2010 08:45:54] yeah, I don't have any either.
[09-Jul-2010 08:45:56] hmm
[09-Jul-2010 08:46:07] well, that's the only caching that zope would be doing really
[09-Jul-2010 08:46:17] in terms of flat file caching anyways
[09-Jul-2010 08:48:23] rmatte: is chet around today?
[09-Jul-2010 08:50:51] no idea
[09-Jul-2010 08:50:58] (I don't actually work for Zenoss lol)
[09-Jul-2010 08:51:02] I haven't seen him around
[09-Jul-2010 08:52:04] speaking of, have you seen Matt?
[09-Jul-2010 08:52:19] not in the last few days
[09-Jul-2010 08:52:26] I think he's at some conference
[09-Jul-2010 08:52:48] either that or on vacation
[09-Jul-2010 09:00:16] or at the pool, or melted from the heat
[09-Jul-2010 09:23:13] hi folks - here's a neat trick
[09-Jul-2010 09:23:25] we have an entry on the event console... that is for a machine that doesn't exist.
[09-Jul-2010 09:23:27] http://screencast.com/t/M2E0YWIxYjU
[09-Jul-2010 09:23:40] note that 'lwmmail01' isn't even a link.
[09-Jul-2010 09:40:09] bigegor_: found your hp proliant montior pack. very nice.
[09-Jul-2010 09:44:28] morning all
[09-Jul-2010 09:58:11] fOrsberg is now known as forsberg
[09-Jul-2010 09:59:04] wow, we have some extremely litterate people here
[09-Jul-2010 09:59:43] "high there ryan does this mean we want to remove this vlan from monitoring ?"
[09-Jul-2010 09:59:49] high there!? really!?
[09-Jul-2010 10:00:12] hahaha …smoking the reefer
[09-Jul-2010 10:00:14] literate*
[09-Jul-2010 10:00:19] lol, apparently
[09-Jul-2010 10:01:23] what skool it go too!
[09-Jul-2010 10:01:34] hehe
[09-Jul-2010 10:02:01] rmatte: fixed my problem from earlier, I just readded the device
[09-Jul-2010 10:02:08] ptom_: cool
[09-Jul-2010 10:02:15] something must have been funked up in zope
[09-Jul-2010 10:02:42] yeah, wierd.
[09-Jul-2010 10:03:15] has anybody a transform for Linux /Perf/Interface event class that adds the interface component to the "threshold of Utilizatoin XXX exceeded:" events?
[09-Jul-2010 10:04:24] gwb2352: not on hand, but you would basically insert it using regex or something
[09-Jul-2010 10:04:43] gwb2352: wouldn't be difficult, check some transform examples
[09-Jul-2010 10:04:55] looking at the event details, there's nothing there to grab to associate the event with an interface
[09-Jul-2010 10:05:52] I'm familiar with using transforms to extend the summaries using additional information in the event details, but not sure how to take the event and go back to the device to find what interface it's associated with.
[09-Jul-2010 10:05:56] good morning
[09-Jul-2010 10:06:06] gwb2352: it should do that automatically
[09-Jul-2010 10:06:12] gwb2352: it should set it as the component
[09-Jul-2010 10:06:17] i'm getting Unable to read processes on device ; Timeout on device.. count is 215
[09-Jul-2010 10:06:40] it does for cisco (i get components of FastEthernetx/x) but not for linux
[09-Jul-2010 10:06:44] however under OS, i see OS processes, the selected/added processes are listed as green, however when i kill one it isnt showing that the service is dead
[09-Jul-2010 10:07:03] it should work the same no matter what the interface
[09-Jul-2010 10:07:10] I don't personally have that issue
[09-Jul-2010 10:07:21] huh.
[09-Jul-2010 10:07:36] Parabol: SSH monitoring I assume?
[09-Jul-2010 10:08:07] rmatte: i did enable SNMP but i'm not sure if its working
[09-Jul-2010 10:08:13] i honestly dont know how to tell :(
[09-Jul-2010 10:08:28] gwb2352: I'm saying there's no reason why Zenoss would specifically distinguish a Linux interface from a Cisco interface
[09-Jul-2010 10:08:33] this is the summary for my linux devices "summary threshold of Utilization 90 perc exceeded: current value 1224793.41"
[09-Jul-2010 10:08:45] right, forget the summary
[09-Jul-2010 10:08:50] what about the component?
[09-Jul-2010 10:09:09] component is blank
[09-Jul-2010 10:09:32] Parabol: you probably enabled it but you don't have the config file done properly, SNMP has "paranoid" settings by default, which won't work with Zenoss
[09-Jul-2010 10:09:44] oh :(
[09-Jul-2010 10:09:52] I gave you my config to use the other day
[09-Jul-2010 10:09:56] I'll paste it again
[09-Jul-2010 10:09:58] <3
[09-Jul-2010 10:10:06] this is the snmpd config right?
[09-Jul-2010 10:10:11] yes
[09-Jul-2010 10:10:22] "eventKey intf|intf_ifInOctets|Utilization 90 perc"
[09-Jul-2010 10:10:39] http://pastebin.com/gWPzQkrk
[09-Jul-2010 10:10:44] no, not the eventKey
[09-Jul-2010 10:10:48] look for "component"
[09-Jul-2010 10:10:53] yes, component is blank
[09-Jul-2010 10:11:07] k, well, it shouldn't be, so not sure what changed to cause that
[09-Jul-2010 10:11:20] I've never seen it blank like that
[09-Jul-2010 10:11:27] wierd
[09-Jul-2010 10:11:43] i'll see what zenoss support says
[09-Jul-2010 10:11:45] danke
[09-Jul-2010 10:11:50] i thought it was "odd"
[09-Jul-2010 10:11:57] oh, you're an enterprise customer?
[09-Jul-2010 10:12:11] yeh, support will probably be able to work with you a bit more efficiently than I can
[09-Jul-2010 10:12:12] trust me, you guys are the best support in the world :)
[09-Jul-2010 10:12:12] ;)
[09-Jul-2010 10:12:26] I second that
[09-Jul-2010 10:12:40] well, I do what I can lol, but enterprise support is good as a backup plan if you have it
[09-Jul-2010 10:13:03] they did help me with my "unable to add IpService" problem
[09-Jul-2010 10:13:19] i'll report back in here after the zendmd script completes to see if it can help others
[09-Jul-2010 10:13:26] cool
[09-Jul-2010 10:13:40] yeh, if they do figure out what it is be sure to let us know
[09-Jul-2010 10:13:51] (I had >70,000 IpServices in the Services bit)
[09-Jul-2010 10:14:05] wow, that's a lot
[09-Jul-2010 10:14:14] so there's a limit?
[09-Jul-2010 10:14:31] not sure, but it takes a long time for the auto-match bits to scan against something that large
[09-Jul-2010 10:14:44] ah
[09-Jul-2010 10:14:45] you're supposed to have in the 3-4K range?
[09-Jul-2010 10:14:57] not sure, checking now lol
[09-Jul-2010 10:15:11] yeh, 3773 is the default from the looks of it
[09-Jul-2010 10:15:18] under IpService
[09-Jul-2010 10:15:28] first step for me is to get to there, then we'll see what happens
[09-Jul-2010 10:15:49] yup
[09-Jul-2010 10:16:11] hmmm, I'm debating if I want to do coding on my zenoss daemon, or continue with 3.0 beta testing today
[09-Jul-2010 10:16:19] leaning towards the testing to get it out of the way
[09-Jul-2010 10:16:41] oh no, that's what I wanted to try doing, get LDAP working with 3.0
[09-Jul-2010 10:16:49] what are your thoughts on 3.0 so far? I haven't touched it
[09-Jul-2010 10:17:01] interface is nice but there's room for improvement
[09-Jul-2010 10:17:24] rmatte: there is *always* room for improvement :)
[09-Jul-2010 10:17:25] http://pastebin.com/qiXJvPXJ
[09-Jul-2010 10:17:31] that's what I have so far
[09-Jul-2010 10:17:39] going to start logging trac tickets once the list is complete
[09-Jul-2010 10:17:43] which should be soon enough
[09-Jul-2010 10:17:44] huh
[09-Jul-2010 10:17:54] rmatte: we definitely want the feedback though
[09-Jul-2010 10:17:57] quite a list
[09-Jul-2010 10:17:58] npmccallum: yeh, I know, but there's some really obvious stuff that could be fixed right now
[09-Jul-2010 10:18:15] npmccallum: mostly usability more than anything, the interface seems quite solid
[09-Jul-2010 10:18:20] rmatte: please file any upgrade issues you see today if possible
[09-Jul-2010 10:18:36] they are going to be highest priority at this stage of the game
[09-Jul-2010 10:18:40] the only upgrade issue I saw so far was the issue with it resetting the admin password to default
[09-Jul-2010 10:18:45] oh, and ZenPack issues
[09-Jul-2010 10:18:52] not sure how you guys are going to be handling that
[09-Jul-2010 10:19:05] are you going to code some sort of live migration or something?
[09-Jul-2010 10:19:25] migration will require downtime
[09-Jul-2010 10:19:41] what zenpack issues are you seeing?
[09-Jul-2010 10:20:15] well, the 2 ZenPacks that I had installed that had daemons needed to be removed prior to the upgrade
[09-Jul-2010 10:20:23] otherwise they couldn't be installed or removed afterwards
[09-Jul-2010 10:20:34] not being able to remove them afterwards was the big pain
[09-Jul-2010 10:20:52] please file that issues ASAP
[09-Jul-2010 10:20:56] I ended up rolling back to a snapshot I had taken, removing them, then upgrading
[09-Jul-2010 10:20:57] k
[09-Jul-2010 10:20:58] s/issues/issue/
[09-Jul-2010 10:22:20] rmatte: there is a really cool change in 3.0 for authors of zenpacks with subcomponents
[09-Jul-2010 10:22:48] Hi, everyone
[09-Jul-2010 10:22:49] rmatte: you can now do something like: compname = "os/interfaces/eth0"
[09-Jul-2010 10:23:02] npmccallum: nice
[09-Jul-2010 10:24:05] then relname = "mynewrelation"
[09-Jul-2010 10:24:38] objects = []
[09-Jul-2010 10:24:38] for i in interfaces:
[09-Jul-2010 10:24:38] objects.append(makeRelMap(i))
[09-Jul-2010 10:24:38] for c in i.children():
[09-Jul-2010 10:24:38] objects.append(makeRelMap(c))
[09-Jul-2010 10:24:38] return objects
[09-Jul-2010 10:25:54] I was wondering why I couldn't login to Zenoss trac with my credentials just before realizing that I was trying to login to our trac system
[09-Jul-2010 10:25:57] facepalm
[09-Jul-2010 10:29:20] hi
[09-Jul-2010 10:29:28] I have a weird problem with syslog forwarding
[09-Jul-2010 10:29:51] some messages do not show up in the events panel
[09-Jul-2010 10:29:58] even though I use "*.debug" for forwarding
[09-Jul-2010 10:30:11] npmccallum: nice
[09-Jul-2010 10:30:19] is there a way to check whether or not zenoss received these messages?
[09-Jul-2010 10:30:39] not if they are being dropped
[09-Jul-2010 10:30:51] if they aren't being dropped then you can check in history
[09-Jul-2010 10:30:58] thanks for the snmp config rmatte :)
[09-Jul-2010 10:31:06] just got off the phone and found it :) hehe
[09-Jul-2010 10:31:44] hmm
[09-Jul-2010 10:31:47] they arent in the history
[09-Jul-2010 10:31:53] who drops messages?
[09-Jul-2010 10:31:53] Parabol: no problem, modify it a bit obviously
[09-Jul-2010 10:31:57] the receiving syslogd?
[09-Jul-2010 10:32:44] rmatte: actually that was the road we traveled down. I just updated the ifstated configuration to take action if mysqlping failed as well, adjusting the advskew on each server appropriately
[09-Jul-2010 10:33:05] rmatte: will need to do that for other critical services in the hosting stack as well going forward. :)
[09-Jul-2010 10:33:38] thanks. I was thinking along the lines of having zenoss down the interface on the device that failed. I think you're right, and having it closer to the machine is better.
[09-Jul-2010 10:34:20] for sure
[09-Jul-2010 10:34:27] no point in having a middle man if you don't need one
[09-Jul-2010 10:34:41] and with that in mind, death to contracting agencies
[09-Jul-2010 10:34:53] rmatte: agreed :)
[09-Jul-2010 10:39:06] also, with having the script run locally you could have it automatically up the interface once the issue is resolved
[09-Jul-2010 10:39:13] in case it's just a blip or something
[09-Jul-2010 10:40:08] is it possible zenoss drops messages with a debug priority?
[09-Jul-2010 10:44:11] npmccallum: http://dev.zenoss.org/trac/ticket/7020#comment:1
[09-Jul-2010 10:45:24] oops
[09-Jul-2010 10:45:31] rmatte: do you ahve the link to your 3.0 findings from the other day?
[09-Jul-2010 10:45:49] http://pastebin.com/qiXJvPXJ
[09-Jul-2010 10:45:53] that's what I have so far
[09-Jul-2010 10:45:56] thanks
[09-Jul-2010 10:45:58] np
[09-Jul-2010 10:46:17] I've only logged the first one as a trac ticket so far, but I'll try to have the rest in by end of day
[09-Jul-2010 10:46:23] *nod*
[09-Jul-2010 10:46:55] * Simon4 runs off to enjoy the sillyhot weather
[09-Jul-2010 10:46:57] have a great weekend all
[09-Jul-2010 10:46:58] realitygeoff_ is now known as realitygeoff
[09-Jul-2010 10:48:11] is there any difference in running reIndex() or dmd.Device.reIndex() from zendmd ?
[09-Jul-2010 10:48:37] dmd.Devices
[09-Jul-2010 10:49:56] ptom_: yes
[09-Jul-2010 10:50:15] dmd.Devices.reIndex() only reindexes the Devices portion of the database
[09-Jul-2010 10:50:22] reindex() reindexes everything
[09-Jul-2010 10:50:57] huh, I thought so but had to ask. TY
[09-Jul-2010 10:51:01] np
[09-Jul-2010 10:54:42] holy crap, they are up to LDAPMultiPlugins 1.11, thought they were still at 1.7, then noticed that they switched pages
[09-Jul-2010 10:56:23] ok gotta go - have a nice weekend
[09-Jul-2010 11:05:50] heh, every single one of my feature requests/bugs have been pushed off again
[09-Jul-2010 11:13:55] just got those emails as well
[09-Jul-2010 11:40:27] theacolyte: mine all got pushed off as well
[09-Jul-2010 11:40:35] you're not the only one lol
[09-Jul-2010 11:41:28] I hope you don't plan on having a "/" in a datapoint name any time soon
[09-Jul-2010 11:41:32] It's been pushed to "Future"
[09-Jul-2010 11:42:07] Back, to the future!
[09-Jul-2010 11:42:22] why would you need a / in a datapoint name?
[09-Jul-2010 11:42:24] lol
[09-Jul-2010 11:43:20] bytes/sec
[09-Jul-2010 11:44:03] oh and hope you don't want to just put a datapoint in the summary portion of an RRD
[09-Jul-2010 11:45:07] wow
[09-Jul-2010 11:45:17] they really changed the way the LDAP plugins work
[09-Jul-2010 11:45:24] ?
[09-Jul-2010 11:45:28] the new ones
[09-Jul-2010 11:45:37] How so?
[09-Jul-2010 11:45:39] I'm working on getting the latest LDAP plugins working in 3.0 beta #4
[09-Jul-2010 11:45:58] the configuration in the zope management interface is very different
[09-Jul-2010 11:45:59] I never actually bothered doing LDAP on y install actually
[09-Jul-2010 11:46:02] instead of adding "LDAP"
[09-Jul-2010 11:46:14] you add "Domain Authentication Plugin"
[09-Jul-2010 11:47:16] hmmm, or maybe the plugins just aren't installed correctly, though I'm pretty sure they are
[09-Jul-2010 11:47:46] Well
[09-Jul-2010 11:47:51] If it does NTLM that's be neat
[09-Jul-2010 11:48:31] hehe
[09-Jul-2010 11:49:01] I remember Jane was working on getting it working, I wonder if she made any progress
[09-Jul-2010 11:49:11] there's obviously going to need to be a new guide written up
[09-Jul-2010 12:01:44] Anyone fluent in python? Silly question!
[09-Jul-2010 12:02:07] never used it
[09-Jul-2010 12:02:11] Simon4_ is now known as Simon4
[09-Jul-2010 12:02:32] same here =\ str = 'hello-test-level1' : I want to cut 'test' out of that string,
[09-Jul-2010 12:03:20] sorry, return 'test', I wan to cut 'hello-' and '-level1'
[09-Jul-2010 12:03:40] str.split('-')[1]
[09-Jul-2010 12:03:44] for a quick filthy way to do it
[09-Jul-2010 12:04:09] you're the best hah
[09-Jul-2010 12:04:11] thanks
[09-Jul-2010 12:04:44] hehe
[09-Jul-2010 12:04:53] yeh, that was a simple one, no regex required on that one
[09-Jul-2010 12:05:24] yeah, I was feeling lazy
[09-Jul-2010 12:05:28] hehe
[09-Jul-2010 12:05:42] I've only used PHP / Javascript. It works perfectly for all 200 strings, thank you :)
[09-Jul-2010 12:05:50] np
[09-Jul-2010 12:16:47] It's finally raining here
[09-Jul-2010 12:16:58] we've had scorching humid weather this week
[09-Jul-2010 12:17:36] it's about 32 degrees outside in the shade, unheard of for London generally
[09-Jul-2010 12:18:24] It was 40+ all this week
[09-Jul-2010 12:18:30] oof
[09-Jul-2010 12:18:32] which is relatively unheard of for us
[09-Jul-2010 12:18:33] where are you based?
[09-Jul-2010 12:18:37] Ottawa, Canada
[09-Jul-2010 12:18:44] ah cool
[09-Jul-2010 12:18:47] I mean, we do get 30+ weather in summer
[09-Jul-2010 12:18:51] but almost never 40+
[09-Jul-2010 12:19:29] yeah, 40+ is well hot
[09-Jul-2010 12:19:57] yup
[09-Jul-2010 12:20:13] we had the queen over here recently, she toured RIM and was given a blackberry
[09-Jul-2010 12:20:16] lol
[09-Jul-2010 12:21:14] mahaaha
[09-Jul-2010 12:22:31] hopefully they gave her a roaming package
[09-Jul-2010 12:44:14] * Zfield wonders if there is anyone here
[09-Jul-2010 12:44:39] yes
[09-Jul-2010 12:45:17] quick question about snmpv3 - where do i set that up in zenoss?
[09-Jul-2010 12:47:50] i'm trying to get my devices to send to zenoss, rather than i have now where its going to the device
[09-Jul-2010 12:48:16] that made no sense lol
[09-Jul-2010 12:48:32] right now zenoss logs into my devices, rather then trapping messages
[09-Jul-2010 12:49:43] you can't send v3 traps to zenoss, it doesn't support em
[09-Jul-2010 12:50:10] you can send v2 or below though, and they'll turn up under /Unknown in the events view initially
[09-Jul-2010 12:50:37] so zenoss can only log into snmpv3?
[09-Jul-2010 12:50:49] yeah, it'll collect from v3
[09-Jul-2010 12:51:04] just zentrap which is the trap collection daemon won't talk v3
[09-Jul-2010 12:51:16] * Simon4 is hoping someone will correct him if he's wrong, but that's how I understand it
[09-Jul-2010 12:52:26] ahhh ok thanks, that saved me from looking for that section even more =)
[09-Jul-2010 12:53:52] is it a bad practise to have zenoss log into a switch, and also receive traps from the same device?
[09-Jul-2010 12:56:35] also, how would i enable zenoss to accept traps for the 'boobies' community string?
[09-Jul-2010 13:03:50] Simon4: is correct
[09-Jul-2010 13:04:09] Zfield: no, it's best practice to have it polling AND receiving traps
[09-Jul-2010 13:04:13] traps don't give you performance info
[09-Jul-2010 13:04:36] they are generally for hardware/software related errors, temperature alerts, etc...
[09-Jul-2010 13:04:45] you should be using both
[09-Jul-2010 13:05:13] everything I monitor is being polled and sending traps/syslogs back to Zenoss
[09-Jul-2010 13:10:41] ahh ok thanks
[09-Jul-2010 13:10:50] i wasn't sure if one would conflict with another
[09-Jul-2010 13:12:28] a week or so I was trying to get filesystem monitoring working for a linux box and shortly after I asked a question it started working so we assumed it was just a matter of waiting long enough for 3 collections to occur.
[09-Jul-2010 13:13:03] The new host I tried to setup has both net-snmp installed, and proper ssh credentials, but all I have is the filesystem sizes. Its not updating.
[09-Jul-2010 13:13:22] can someone list the basic steps need to get this going?
[09-Jul-2010 13:14:44] Zfield: they are designed to work together
[09-Jul-2010 13:15:24] nzle: everything else modeled correctly? you're not seeing any debug messages for that device?
[09-Jul-2010 13:15:41] nzle: Is it graphing CPU ok?
[09-Jul-2010 13:17:03] no, modeled ok, Can you say what the basic collectors needed are? the original configuration was change at some point and allmost all of the collectors were removed from the default config so I have been experimenting with adding them back in. Where are all the collectors described and the default ones listed?
[09-Jul-2010 13:18:20] lol
[09-Jul-2010 13:18:22] http://www.phdcomics.com/comics/archive.php?comicid=1335
[09-Jul-2010 13:18:50] nzle: oh, well, if it modeled the filesystems then you have the correct collector plugin for that
[09-Jul-2010 13:18:57] let me see what I have in my linux class just to be sure
[09-Jul-2010 13:19:06] cpu listed as GenuineIntel_ Intel(R) Xeon(R) CPU E5430 _ 2.66GHz, no graph..
[09-Jul-2010 13:19:20] no graph?
[09-Jul-2010 13:19:27] like it's not there at all?
[09-Jul-2010 13:19:31] or it's not generating?
[09-Jul-2010 13:19:45] no graphs showing on the perf tab.
[09-Jul-2010 13:19:57] then you need to bind the template
[09-Jul-2010 13:20:28] do you know how to do that?
[09-Jul-2010 13:20:46] I did it once recently, what template?
[09-Jul-2010 13:20:56] probably "Device"
[09-Jul-2010 13:21:10] yup, Device, and Devices
[09-Jul-2010 13:21:15] so that you get uptime stats
[09-Jul-2010 13:21:37] now, for plugins... I have....
[09-Jul-2010 13:21:45] zenoss.snmp.NewDeviceMap
[09-Jul-2010 13:21:51] zenoss.snmp.DeviceMap
[09-Jul-2010 13:21:56] I have device bound for the device in question.
[09-Jul-2010 13:22:05] zenoss.snmp.InterfaceMap
[09-Jul-2010 13:22:10] zenoss.snmp.RouteMap
[09-Jul-2010 13:22:16] zenoss.snmp.IpServiceMap
[09-Jul-2010 13:22:24] zenoss.snmp.HRFileSystemMap
[09-Jul-2010 13:22:32] zenoss.snmp.HRSWRunMap
[09-Jul-2010 13:22:38] zenoss.snmp.CpuMap
[09-Jul-2010 13:22:59] If you have it bound you should be seeing graphs
[09-Jul-2010 13:23:07] unless you somehow ruined that template
[09-Jul-2010 13:23:21] there should be 2 templates
[09-Jul-2010 13:23:23] or atleast the graph frames ;>
[09-Jul-2010 13:23:24] Device and Devices
[09-Jul-2010 13:23:27] you should bind both
[09-Jul-2010 13:23:39] yeh exactly, you should be seeing the graph frames
[09-Jul-2010 13:23:48] if you don't then it's either not bound or the template is screwed
[09-Jul-2010 13:24:20] If you messed up Zenoss you should start from scratch
[09-Jul-2010 13:24:42] people who just jump in and start digging without reading the entire admin guide beforehand sometimes end up deleting stuff they shouldn't have
[09-Jul-2010 13:24:47] another host is ok so I doubt the template is fubar.
[09-Jul-2010 13:24:49] and changing settings that shouldn't be changed
[09-Jul-2010 13:24:56] oh
[09-Jul-2010 13:25:13] are you sure you don't have custom bindings on that one host?
[09-Jul-2010 13:25:25] bindings can be done at the device level, and at the class level
[09-Jul-2010 13:25:30] where do I find the devices template to rebind it?
[09-Jul-2010 13:25:57] you need to navigate to /Server/Linux/Templates
[09-Jul-2010 13:26:02] then select bind templates from the dropdown
[09-Jul-2010 13:26:07] and select both Device and Devices
[09-Jul-2010 13:26:09] then click save
[09-Jul-2010 13:26:14] hold down ctrl to select both
[09-Jul-2010 13:26:35] (this is very basic stuff which is documented in the admin guide, I would suggest you read it)
[09-Jul-2010 13:28:50] I've read it, I will again this weekend. Between re-reading my HP San manuals and the HP Insight Manager manuals.
[09-Jul-2010 13:29:11] don't just skim over it, actually read it ;)
[09-Jul-2010 13:29:29] I have a bad habbit of skimming over tech manuals, and that's what I did the first time I read the admin guide
[09-Jul-2010 13:29:38] but then I finally read it in full and I was good to go after that
[09-Jul-2010 13:29:55] there's a lot of things that you'll read and go "oh, so that's how that's done"
[09-Jul-2010 13:33:54] yep, pulled away...
[09-Jul-2010 13:35:09] rmatte: I think everyone has that bad habit
[09-Jul-2010 13:36:53] hehe
[09-Jul-2010 13:37:57] i second that
[09-Jul-2010 13:49:03] got it I think, server wasn't in the right class.
[09-Jul-2010 14:01:34] yep, that was it.
[09-Jul-2010 14:02:51] do i need to install a zenpack before being able to receive traps?
[09-Jul-2010 14:03:16] Zfield: no
[09-Jul-2010 14:03:24] Zfield: full snmp capability is built in to Zenoss
[09-Jul-2010 14:03:37] Zfield: you will need to install MIBs for your devices though
[09-Jul-2010 14:03:48] min in black's?
[09-Jul-2010 14:03:50] *men
[09-Jul-2010 14:03:56] poor joke.. i know :-p
[09-Jul-2010 14:03:59] forsberg is now known as fOrsberg
[09-Jul-2010 14:03:59] :|
[09-Jul-2010 14:04:10] lol
[09-Jul-2010 14:04:16] management information base?
[09-Jul-2010 14:04:36] correct
[09-Jul-2010 14:04:43] linux mib is built-in?
[09-Jul-2010 14:04:43] you use zenmib to install them
[09-Jul-2010 14:04:51] linux mib is built in, yes
[09-Jul-2010 14:05:07] but for stuff like Cisco and HP you'll have to install them
[09-Jul-2010 14:05:20] I made a Cisco MIBs ZenPack that can be used for Cisco gear
[09-Jul-2010 14:05:38] i do have some cisco, hp, and 2 linksys poe switches
[09-Jul-2010 14:05:54] so im guessing for my linksys switch it isn't as easy as throwing in a community string and clikc ADD
[09-Jul-2010 14:06:02] no idea
[09-Jul-2010 14:06:07] I've never used a linksys switch
[09-Jul-2010 14:06:10] it might be, it might not be
[09-Jul-2010 14:06:26] you'd definitely need to make a custom template for it in Zenoss
[09-Jul-2010 14:06:29] if that's what you mean
[09-Jul-2010 14:06:43] provided they even support snmp
[09-Jul-2010 14:06:48] well i have a SNMP section, it asks for a Recipient IP, notification type: Traps, and notification version
[09-Jul-2010 14:07:08] oh and a community string and port
[09-Jul-2010 14:07:08] right, then specify the IP of the zenoss server, set version to 1 or 2
[09-Jul-2010 14:07:15] set the community string
[09-Jul-2010 14:07:19] ok, did all that
[09-Jul-2010 14:07:19] leave port as default
[09-Jul-2010 14:07:26] is there anything i need to do on the zenoss side?
[09-Jul-2010 14:07:31] you'll need to get linksys mibs and install them
[09-Jul-2010 14:07:40] find the mibs, the admin guide explains the rest
[09-Jul-2010 14:08:13] and if you want to monitor cpu/memory or whatever else you need to research what SNMP OIDs to use and create a device template
[09-Jul-2010 14:08:17] ohhhh ok i was confused for a second, i thought i was to install the MIB to the switch which made no sense at all to me
[09-Jul-2010 14:08:24] no
[09-Jul-2010 14:08:30] you install the Mibs in to Zenoss
[09-Jul-2010 14:08:32] using zenmib
[09-Jul-2010 14:08:48] so that it is able to properly translate the traps when they come in
[09-Jul-2010 14:08:53] so the mib is like a decoder ...
[09-Jul-2010 14:08:56] yea what you just said
[09-Jul-2010 14:08:58] gotcha
[09-Jul-2010 14:08:58] yes
[09-Jul-2010 14:09:06] you anwsered my question before i finished =)
[09-Jul-2010 14:09:12] lol
[09-Jul-2010 14:09:48] is windows support also built-in, or would i need a mib for that too?
[09-Jul-2010 14:09:57] anyone here use the esx zenpack in 2.5.x? I just upgraded to 3.x as my eval was over and I was bored and now it has all kinds of awesome esx reports
[09-Jul-2010 14:09:59] you need to install Mibs for that
[09-Jul-2010 14:10:03] I wasn't sure if those were in the old one and i just missed them
[09-Jul-2010 14:10:06] and make sure they work before you start doing transforms
[09-Jul-2010 14:10:26] our windows Mibs were installed incorrectly, and now we're up to like 100 transforms that I would have to redo if I installed the Mibs
[09-Jul-2010 14:10:35] so we're kind of screwed in that department at the moment
[09-Jul-2010 14:10:36] lol
[09-Jul-2010 14:10:42] lol
[09-Jul-2010 14:11:21] if you see the trap come in as enterprises. with a bunch of numbers after it, it's not setup properly
[09-Jul-2010 14:11:27] they should come in with meaninful names
[09-Jul-2010 14:11:41] like: snmp trap getOffYourArseAndFixMe
[09-Jul-2010 14:12:23] eugh, 1 more hour then home
[09-Jul-2010 14:12:24] woot
[09-Jul-2010 14:16:28] rmatte: pick your brain?
[09-Jul-2010 14:17:08] mmmm bbbrrraaaiinnnsssss
[09-Jul-2010 14:17:32] :)
[09-Jul-2010 14:18:30] hmm weird, in 3.0 it seems like the netapp modelling thing wants to ssh
[09-Jul-2010 14:18:47] ashp: core or enterprise?
[09-Jul-2010 14:19:14] enterprise
[09-Jul-2010 14:19:26] hrmm can't help you there, i use the community netapp zenpacks
[09-Jul-2010 14:19:52] what's confuzzling me right now is i have two nodes on the other side of a firewall, that i can clearly snmpwalk, yet i keep getting snmp agent down events from them
[09-Jul-2010 14:21:04] im looking at the manual, and i'm still clueless where to get the mibs from
[09-Jul-2010 14:21:15] where to get what mibs?
[09-Jul-2010 14:21:39] yeah
[09-Jul-2010 14:21:53] and if that makes no sense..... then you know how f***ed i am
[09-Jul-2010 14:22:24] oh wait, mib's arent zenoss specific.. duhhh
[09-Jul-2010 14:22:27] nvm i think im good now
[09-Jul-2010 14:22:58] lol, zenoss comes with a few
[09-Jul-2010 14:23:08] you provide the rest, SCP them onto the server, run zenmib to import
[09-Jul-2010 14:23:17] there is a zenpack which adds some gui dialogs to make it easier
[09-Jul-2010 14:23:52] oooh, where can i find those? im here community/zenpacks
[09-Jul-2010 14:24:02] but searched for gui and didn't get anything :-p
[09-Jul-2010 14:24:06] everytime i have to deal with mibs i want to stab myself in the face
[09-Jul-2010 14:24:07] if that helps :(
[09-Jul-2010 14:24:22] no, no it doesn't lol
[09-Jul-2010 14:24:43] i think i'm gonna go hang myself in the server room with some network cables
[09-Jul-2010 14:24:46] docs/DOC-3412
[09-Jul-2010 14:24:56] works fairly well for me in 2.5.2
[09-Jul-2010 14:25:08] i don't add mibs often, i mostly do performance, not so much alerting
[09-Jul-2010 14:25:13] im working up to that :)
[09-Jul-2010 14:26:21] thx for the link
[09-Jul-2010 14:28:09] hrmm, i increase my snmp timeout value for that device class, let's see how that goes
[09-Jul-2010 14:28:27] i don't get it tho, it seems like if i can snmpget, i should be good, theres nothing else to it
[09-Jul-2010 14:29:27] you download mibs from the manufacturer's website usually
[09-Jul-2010 14:29:41] yea, im on cisco's MIB site atm
[09-Jul-2010 14:29:51] and .. there is... a .. lot
[09-Jul-2010 14:29:59] ftp://ftp-sj.cisco.com/pub/mibs/v2/
[09-Jul-2010 14:30:52] there's a Cisco Mibs ZenPack
[09-Jul-2010 14:31:22] first ZenPack I made: docs/DOC-3404
[09-Jul-2010 14:31:36] it'll take a long time to install, so install it from the commandline
[09-Jul-2010 14:31:42] and follow the zendmd steps once it's installed
[09-Jul-2010 14:31:49] rmatte: you ever seen SNMPwalk work, yet zenoss still gets snmp agent down events?
[09-Jul-2010 14:31:53] i can model the device even
[09-Jul-2010 14:32:43] I honestly think the ASA between me and the node is doing something screwy, but you'd think i would get nothing, if that was the case
[09-Jul-2010 14:33:27] Is there a replacement for the alerting rules as a user? In nagios we had all kinds of groups defined etc etc and none of that seems doable.
[09-Jul-2010 14:33:41] ashp: sure you can
[09-Jul-2010 14:33:46] users can have their own rules
[09-Jul-2010 14:33:51] or they can be put into groups that have rules
[09-Jul-2010 14:34:45] twm1010: yeh, try changing zMaxOIDPerRequest from 40 to 20 or 10
[09-Jul-2010 14:34:56] twm1010: what kind of device is it?
[09-Jul-2010 14:35:19] its a linux machine, but you're right, its on the other side of an ASA, and i had to adjust maxOID for ASA's, maybe thats affecting these devices
[09-Jul-2010 14:35:21] ah yeh, ASAs are picky about the amount of SNMP requests sent at one time
[09-Jul-2010 14:35:27] so drop that value down
[09-Jul-2010 14:35:29] let me see what i set the ASA's to
[09-Jul-2010 14:35:30] and it should fix the problem
[09-Jul-2010 14:35:59] only drop it for the devices that you need to
[09-Jul-2010 14:36:02] don't do it globally
[09-Jul-2010 14:36:12] Oh, so I add each user (even if they never log in), then define those as groups, then do alerting rules per group?
[09-Jul-2010 14:36:30] Only thing that sucks about that is that it's confusing to get a good overview of what's going where, but I can work with it :)
[09-Jul-2010 14:36:34] ashp: correct
[09-Jul-2010 14:36:49] let's see if that does the trick, dropped it from 40 to 20 for that device class
[09-Jul-2010 14:36:58] I signed off on buying enterprise today so I figured I better like, work out the rest of the TODO list items :D
[09-Jul-2010 14:36:59] twm1010: I'm guessing it will
[09-Jul-2010 14:37:02] only 4 devices, very few OID's being queries, should be good
[09-Jul-2010 14:37:10] ashp: hehe, nice
[09-Jul-2010 14:37:24] ashp: i am going to try and make the case for enterprise here next year
[09-Jul-2010 14:37:40] My case for enterprise was "here's nagios.. here's what zenoss does with ESX monitoring"
[09-Jul-2010 14:37:42] i think i've demonstrated the value of core enought to quit dicking around and get the big daddy
[09-Jul-2010 14:37:45] i pretty much made my case with that alone
[09-Jul-2010 14:37:56] I'm going to try to make the case for enterprise here soon to replace the other POS software that they bought
[09-Jul-2010 14:38:01] tho, rmatte over there does some serious monitoring on core
[09-Jul-2010 14:38:13] yeh, I've got 14 core servers running
[09-Jul-2010 14:38:18] well, the only gripe i have with zenoss is i want variable polling
[09-Jul-2010 14:38:30] my largest install went up to 401 devices
[09-Jul-2010 14:38:31] wow, that's some monitoring
[09-Jul-2010 14:38:33] solarwinds lets me set polling on individual interfaces of devices
[09-Jul-2010 14:38:38] which isn't huge, but it's pretty big
[09-Jul-2010 14:38:39] I feel a bit guilty considering I'm only monitoring like, 50 devices
[09-Jul-2010 14:38:59] twm1010: solarwinds has the WORST web interface I've ever seen
[09-Jul-2010 14:39:19] twm1010: Changing a text field from "False" to "True" by typing it in is a nightmare
[09-Jul-2010 14:39:25] they apparently haven't heard of dropdowns
[09-Jul-2010 14:39:27] LOL WTF
[09-Jul-2010 14:39:29] u seroius?
[09-Jul-2010 14:39:33] very
[09-Jul-2010 14:39:41] I almost died when I saw that after we bought it
[09-Jul-2010 14:39:43] oh I agree, you also can't set maintenance windows without using SQL scripts
[09-Jul-2010 14:39:48] I don't manage it thank god
[09-Jul-2010 14:39:54] though they are trying to dump it on me
[09-Jul-2010 14:39:57] they might as well have made it case sensitive... true would throw an error, where True would be good lol
[09-Jul-2010 14:40:02] for deliberate, defined monitoring of network circuits, its good
[09-Jul-2010 14:40:08] good reporting engine too
[09-Jul-2010 14:40:28] rmatte: if you don't mind me asking, what on earth do you monitor with 14 boxes? :O
[09-Jul-2010 14:40:31] where in zenoss you filter out what you don't want, solarwinds makes you be explicit about what you do
[09-Jul-2010 14:40:45] ashp: we're an MSP, each box has a different client on it
[09-Jul-2010 14:40:50] ahhh
[09-Jul-2010 14:40:56] I used to do managed services stuff at ntt
[09-Jul-2010 14:41:04] some boxes have around 20 devices, the majority have 100+, and 1 has 400+
[09-Jul-2010 14:41:12] we sure as hell didn't have anything as nice as zenoss for monitoring, however
[09-Jul-2010 14:41:17] i have about 200 devices
[09-Jul-2010 14:41:18] hehe
[09-Jul-2010 14:41:22] i worked for some other uk managed services companies
[09-Jul-2010 14:41:28] and each one had an even worse monitoring system
[09-Jul-2010 14:41:32] I'm at over 1000 devices total, and we're honestly just getting started
[09-Jul-2010 14:41:36] with the very worst being tivoli
[09-Jul-2010 14:41:51] we've only been at this for a year, though most of our staff used to work at a big MSP before they got bought out and liquidated
[09-Jul-2010 14:41:55] so we moved over here
[09-Jul-2010 14:42:07] I used tivoli at my last job
[09-Jul-2010 14:42:08] especially considering our sales people kept telling customers that tivoli would 'predict failures in advance'
[09-Jul-2010 14:42:11] those fuckers :/
[09-Jul-2010 14:42:13] pain in the arse
[09-Jul-2010 14:42:15] wish i could do this full time
[09-Jul-2010 14:42:23] dear customers, tivoli can't see the future, sorry
[09-Jul-2010 14:42:28] well, their ticketing system, not their monitoring
[09-Jul-2010 14:42:34] but i'm also the VOIP admin, the exchange admin, the groupwise admin, the NetApp admin, you get the idea
[09-Jul-2010 14:42:40] nice
[09-Jul-2010 14:42:57] we used to get about 10,000 events for tivoli a day, all of which had to be cleared by hand
[09-Jul-2010 14:43:05] I'm the Zenoss admin, otrs admin, ziptie admin, sysadmin of almost anything to do with the monitoring infrastructure
[09-Jul-2010 14:43:07] you could clear them in batches by searching for key words but it was a total nightmare
[09-Jul-2010 14:43:25] plus I do development work
[09-Jul-2010 14:43:31] i'm at a law school so i'm in charge of anything that runs on linux, but we have a tiny baby environment here compared to what i was used to
[09-Jul-2010 14:43:39] hehe
[09-Jul-2010 14:43:50] is ziptie any good? we were talking just this morning about some kind of inventory system
[09-Jul-2010 14:43:50] we have each Zenoss instance feeding in to our ticketing system
[09-Jul-2010 14:43:59] so our techs works out of the ticketing system as their event console
[09-Jul-2010 14:44:02] rmatte: why wasn't I working for YOUR managed services company
[09-Jul-2010 14:44:05] and just go in to Zenoss for troubleshooting
[09-Jul-2010 14:44:10] are there any good open source ticketing systems?
[09-Jul-2010 14:44:16] ntt at least had an amazing inventory system, but the rest was mostly manual
[09-Jul-2010 14:44:20] ziptie is no longer supported/developed
[09-Jul-2010 14:44:27] but it's a good system
[09-Jul-2010 14:44:42] they got bought out by the same company that bought the MSP we were working at out
[09-Jul-2010 14:45:03] now it's called network authority and I think they only sell it, no more opensource
[09-Jul-2010 14:45:11] we still use ziptie though, it gets the job done
[09-Jul-2010 14:45:17] and it's free
[09-Jul-2010 14:45:30] it's funny, all the software that does the real heavy lifting for us is open source
[09-Jul-2010 14:45:31] we really do need to look into something for the whole change management/inventory purpose
[09-Jul-2010 14:45:37] rmatte: is the free version still available?
[09-Jul-2010 14:45:41] we have about 100 servers in the entire environment so it's hardly worth anything fancy
[09-Jul-2010 14:45:48] the crap products that they pay tons of money for do almost nothing except look pretty for clients (such as solarwinds)
[09-Jul-2010 14:45:56] i won't lie, i sure enjoy not doing managed services stuff anymore :)
[09-Jul-2010 14:46:02] twm1010: you can probably find it somewhere...
[09-Jul-2010 14:46:16] I can send it to you if I can't find a link...
[09-Jul-2010 14:46:39] cool, what do you use for ticketing?
[09-Jul-2010 14:46:44] didn't you integrate zenoss into it?
[09-Jul-2010 14:47:17] otrs
[09-Jul-2010 14:47:39] yeh, my manager wrote a perl script that actually creates the tickets, and I wrote a zenoss daemon that calls on the script
[09-Jul-2010 14:47:58] http://dmon.org/zenticket.html
[09-Jul-2010 14:48:02] that's the info on the daemon
[09-Jul-2010 14:49:26] soooooo, im trying to install a zenpack...
[09-Jul-2010 14:49:33] [root@zenoss zenoss]# zenpack --install ZenPacks.community.mib_utils-1.08-py2.4.egg
[09-Jul-2010 14:49:33] -bash: zenpack: command not found
[09-Jul-2010 14:49:33] [root@zenoss zenoss]#
[09-Jul-2010 14:50:06] * Zfield RTFM and is still sol
[09-Jul-2010 14:50:26] you need to be the zenoss user bud
[09-Jul-2010 14:50:32] sudo su - zenoss
[09-Jul-2010 14:50:40] never do anything zenoss related as root
[09-Jul-2010 14:51:07] wow, that looks pretty good
[09-Jul-2010 14:51:20] probably should run it on a dedicated box, a VM would probably do
[09-Jul-2010 14:51:22] twm1010: It's not bad, still room for improvement
[09-Jul-2010 14:51:33] it get's the job done tho, right?
[09-Jul-2010 14:51:40] twm1010: it does indeed
[09-Jul-2010 14:51:41] well wadda ya knwo, she works =)
[09-Jul-2010 14:51:48] thx
[09-Jul-2010 14:51:52] np
[09-Jul-2010 14:52:18] so realistically, i never need to use my root accout
[09-Jul-2010 14:52:20] *account
[09-Jul-2010 14:52:29] cool, good catching up, gotta jet
[09-Jul-2010 14:52:31] take it easy
[09-Jul-2010 15:00:57] man i wish they'd publish the updated docs for 3.0 so i can understand how to make this 'disk firmware' report work
[09-Jul-2010 15:11:11] bzed is now known as Guest41793
[09-Jul-2010 15:15:43] Guest41793 is now known as bzed_
[09-Jul-2010 15:17:57] bzed_ is now known as bzed
[09-Jul-2010 15:23:27] wow, versata has done a good job of wiping ziptie off the face of the internet
[09-Jul-2010 15:23:35] I think I'll setup a new page so that people can download it
[09-Jul-2010 15:23:43] since I have all the binaries archived
[09-Jul-2010 15:23:49] I don't have the source archived though
[09-Jul-2010 16:58:39] hi everyone, I wish to know how can I set up alerts for ping fails, so if a device is down after "n" ping fail zenoss send the alert, is it clear?
[09-Jul-2010 17:30:20] fOrsberg is now known as forsberg
[09-Jul-2010 20:22:47] Within the debian/ubuntu world, which is considered the best snmp package to be running on zenoss-monitored systems?
[09-Jul-2010 20:34:14] mota: snmpd?
[09-Jul-2010 20:35:07] thanks. I was interested in discussion of whether folks thought anythng better had come along.
[10-Jul-2010 00:00:30] [disconnected at Sat Jul 10 00:00:30 2010]
[11-Jul-2010 00:00:31] [connected at Sun Jul 11 00:00:31 2010]
[11-Jul-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]
[11-Jul-2010 02:40:43] ashp you still here ? anyways its pretty easy to test
[11-Jul-2010 02:40:48] i like that zenpack alot
[11-Jul-2010 02:49:26] i can test for you if you give me example site ;>
[11-Jul-2010 15:13:41] Simon4_ is now known as Simon4
[11-Jul-2010 20:05:15] some one speak on spanish ?
[11-Jul-2010 23:39:11] arafa
[12-Jul-2010 00:00:30] [disconnected at Mon Jul 12 00:00:30 2010]
[12-Jul-2010 00:00:30] [connected at Mon Jul 12 00:00:30 2010]
[12-Jul-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]
[12-Jul-2010 01:46:42] forsberg is now known as fOrsberg
[12-Jul-2010 02:36:19] fOrsberg is now known as forsberg
[12-Jul-2010 06:31:27] Hi! is it possible to define a threshold that uses 2 datapoints in the same datasource?
[12-Jul-2010 06:32:03] i have a gauge of events which have to be processed, and a counter which tracks the latest processed event
[12-Jul-2010 06:32:19] so on the RRD graph, i see amount of events needing to be processed, and the rate at which events are processed
[12-Jul-2010 06:32:39] now i want a way to say "alert me when the rate is 0 and pending events > 0'
[12-Jul-2010 06:32:40] so you only want the threshold to trip if both values go over
[12-Jul-2010 06:33:21] I just tested and you can select multiple data points when creating a threshold
[12-Jul-2010 06:34:04] I don't believe you can have different values for the data points though
[12-Jul-2010 06:34:15] each would need to trip on the same threshold ..
[12-Jul-2010 06:35:14] ugh
[12-Jul-2010 06:40:14] but I can see where that would be VERY useful
[12-Jul-2010 06:42:05] there are some references in google results. like http://copilotco.com/mail-archives/zenoss.2008/msg04940.html
[12-Jul-2010 06:56:42] meh, everyone just does workarounds
[12-Jul-2010 07:00:06] maybe it will be a feature built into 3.x
[12-Jul-2010 07:03:05] i doubt it
[12-Jul-2010 07:54:59] morning all
[12-Jul-2010 07:56:19] morning
[12-Jul-2010 08:32:05] With zenoss is it easy to monitor /etc/fstab and alert if filesystems aren't mounted?
[12-Jul-2010 08:32:18] We had a big datacenter outage and we had issues with filesystems not being remounted and no monitoring
[12-Jul-2010 08:56:44] Parabola_ is now known as Parabola
[12-Jul-2010 08:56:54] good morning
[12-Jul-2010 08:57:22] i've got a debian box thats being monitored via SNMP, i've got some OS processes being watched, apache2, mydns and a couple others
[12-Jul-2010 08:57:39] its showing the status as up, however when i stop mydns it doesnt seem to detect that its stopped
[12-Jul-2010 09:00:53] Parabola: when did you add the processes in to monitoring?
[12-Jul-2010 09:01:32] ashp: in my experience, zenoss just relies on snmp, which only includes filesystems which are mounted
[12-Jul-2010 09:01:42] ashp: for what you want, you should probably write your own check
[12-Jul-2010 09:02:13] rmatte: days ago
[12-Jul-2010 09:02:23] Parabola: zenoss process monitoring is very buggy
[12-Jul-2010 09:02:31] rmatte: i know over the weekend i had tested with another box, i dont remmeber which, stopping services and it noticed and emailed
[12-Jul-2010 09:02:44] alright, whats the fix
[12-Jul-2010 09:02:48] I need this to work :-\
[12-Jul-2010 09:02:50] Parabola: restart zenprocess
[12-Jul-2010 09:02:57] alright
[12-Jul-2010 09:03:12] Parabola: zenprocess is very finicky, I always restart it after making any kind of changes to process monitoring
[12-Jul-2010 09:03:25] Parabola: if that doesn't work, perform a reindex() and commit() in zendmd
[12-Jul-2010 09:03:34] i dont know what you mean by that
[12-Jul-2010 09:03:41] service restarted
[12-Jul-2010 09:03:48] rmatte: does it make a difference in flaky-ness if the process is being monitored by SNMP or by command?
[12-Jul-2010 09:03:51] you become the zenoss user via ssh, then type zendmd
[12-Jul-2010 09:04:00] and it'll bring you in to a special python shell for zenoss
[12-Jul-2010 09:04:05] then you type reindex()
[12-Jul-2010 09:04:07] and hit enter
[12-Jul-2010 09:04:14] then you type commit() once that finishes
[12-Jul-2010 09:04:14] i despise process monitoring
[12-Jul-2010 09:04:16] and hit enter
[12-Jul-2010 09:04:20] its so buggy.. :/
[12-Jul-2010 09:04:32] jb: yeh, it's buggy, but I've got it down to a fine science at this point lol
[12-Jul-2010 09:04:52] alright so i stopped the service
[12-Jul-2010 09:04:56] well, it seems like it works, and then all of a sudden it breaks
[12-Jul-2010 09:05:04] rmatte: what regexp do you normally use to monitor one process?
[12-Jul-2010 09:05:06] like "mysqld"
[12-Jul-2010 09:05:09] the problem is that it's way overengineered. zenoss developers/managers think they know better then sysadmins
[12-Jul-2010 09:05:36] yeah the shits not working
[12-Jul-2010 09:05:44] i stopped mysdns and waited about 60 seconds
[12-Jul-2010 09:05:46] refreshed, nothing
[12-Jul-2010 09:05:58] jb: I just use a regex that matches the name
[12-Jul-2010 09:06:05] all people i talk to all say "i want to say: class or device foo should have processes foo and bar". not "let zenoss automatically detect shit which is incorrect and generates incorrect and unreliable alerts"
[12-Jul-2010 09:06:06] "mysqld" in that case
[12-Jul-2010 09:06:34] Dieterbe: I don't find that it detects things that are incorrect
[12-Jul-2010 09:06:50] Dieterbe: it's never going to detect a process that isn't there are long as it's configured correctly
[12-Jul-2010 09:07:04] so this is odd
[12-Jul-2010 09:07:07] it often finds and starts monitoring which you don't care about
[12-Jul-2010 09:07:13] +processes
[12-Jul-2010 09:07:19] under classes > processes , i have them added with the service and the name as the same
[12-Jul-2010 09:07:31] Dieterbe: so you're saying you'd care about those processes on one device but not necessarily on another?
[12-Jul-2010 09:07:32] however when i go under the server, it changes the name to the full file path
[12-Jul-2010 09:07:45] rmatte: of course
[12-Jul-2010 09:07:54] Parabola: it's supposed to do that
[12-Jul-2010 09:08:07] so does that mean its seeing the services on the box correctly?
[12-Jul-2010 09:08:09] Parabola: the name set in processes is just a regex for locating the processes to monitor
[12-Jul-2010 09:08:13] if its renaming the name to the full path like that
[12-Jul-2010 09:08:17] yes, it does
[12-Jul-2010 09:08:25] okay well it doesnt detect when they stop still
[12-Jul-2010 09:08:31] reindex
[12-Jul-2010 09:08:32] i swear im going to stab somone
[12-Jul-2010 09:08:54] do a reindex as I explained then restart zenprocess again
[12-Jul-2010 09:09:00] and what you dont get is i cant just do that
[12-Jul-2010 09:09:04] i dont have SSH on the zenoss VM
[12-Jul-2010 09:09:05] and make sure you're actually restarting the same process that it detected
[12-Jul-2010 09:09:10] as its a VM i cant walk over to a KVM
[12-Jul-2010 09:09:15] and as im on a mac i dont have the vmware console
[12-Jul-2010 09:09:17] k, you really really need to enable ssh
[12-Jul-2010 09:09:26] commandline is invaluable when working with zenoss
[12-Jul-2010 09:09:27] no, they really need to write a working product
[12-Jul-2010 09:09:48] The product works fine, it just has some quirks
[12-Jul-2010 09:09:53] then it doesnt work fine
[12-Jul-2010 09:09:58] i'm not being rude, i like it dont get me wrong
[12-Jul-2010 09:09:59] I use it to successfully monitor tons of processes
[12-Jul-2010 09:10:03] on both windows and linux
[12-Jul-2010 09:10:06] but at the same time im getting yelled at over this
[12-Jul-2010 09:10:08] because it randomly works
[12-Jul-2010 09:10:11] this weekend, it was spot on
[12-Jul-2010 09:10:21] notified me right away, no remediation of course which is STUPID
[12-Jul-2010 09:10:23] but it at least told me
[12-Jul-2010 09:10:30] now it just doesnt work, no one changed anything
[12-Jul-2010 09:10:41] *shrugs*
[12-Jul-2010 09:10:51] have you tried remodelling that device?
[12-Jul-2010 09:11:04] yes i did
[12-Jul-2010 09:11:13] k, well reindexing is the only thing left
[12-Jul-2010 09:11:29] but if anyone thinks i'm going to remodel every 24 hours just in case monitoring doesnt work
[12-Jul-2010 09:11:32] they are crazy :)
[12-Jul-2010 09:11:40] there has to be a fix for this
[12-Jul-2010 09:11:44] that's not what I was suggesting
[12-Jul-2010 09:11:50] I only remodel by hand personally
[12-Jul-2010 09:11:55] and the process monitoring works just fine
[12-Jul-2010 09:12:04] in a way you are mate, 24 hours ago it was fine
[12-Jul-2010 09:12:09] now its broke with no user intervention
[12-Jul-2010 09:12:55] wow even better
[12-Jul-2010 09:12:56] why do you only remodel by hand?
[12-Jul-2010 09:13:39] doesn't the approach you take to operationally/administratively down interface monitoring muck with that? or do you simply remodel when you take an interface down intentionally?
[12-Jul-2010 09:14:05] twm1010: because our clients are monitored across tunnels, and some of them have really slow links, I don't want Zenoss remodelling everything at intervals
[12-Jul-2010 09:14:38] and no, we only monitor very specific interfaces, such as switch trunks, router ports, etc...
[12-Jul-2010 09:14:49] and they don't change often
[12-Jul-2010 09:15:00] if they do we get notified by the client of the change
[12-Jul-2010 09:15:03] and remodel the device
[12-Jul-2010 09:15:18] gotcha
[12-Jul-2010 09:15:21] it fits your model then
[12-Jul-2010 09:15:26] we change often
[12-Jul-2010 09:15:27] yeh, it works
[12-Jul-2010 09:15:46] if I were monitoring a LAN or WAN that I had direct access to, then of course I'd do automated modelling
[12-Jul-2010 09:16:16] for one of our clients, some devices take upwards of 30 minutes to remodel
[12-Jul-2010 09:16:16] lol
[12-Jul-2010 09:16:17] lol
[12-Jul-2010 09:16:25] so, this is neat
[12-Jul-2010 09:16:32] after about 5 times of starting and stopping
[12-Jul-2010 09:16:33] when you snmpwalk them you can comfortably read the output line by line
[12-Jul-2010 09:16:35] it shows up as stopped
[12-Jul-2010 09:16:45] however, it didnt show up as stopped until after i started it
[12-Jul-2010 09:16:53] Dieterbe: Yeah, I was hoping to use ssh and somehow compare fstab to what it gets from snmp every so often, something along those lines
[12-Jul-2010 09:16:55] since we're going through a tunnel, bouncing around in the U.S, and then going over to europe across painfully slow links to hit them
[12-Jul-2010 09:18:29] okay so i think my issue is moer the dns service than zenoss
[12-Jul-2010 09:18:34] i had two pids, i killed both
[12-Jul-2010 09:18:37] started the service
[12-Jul-2010 09:18:42] and its still showing up as dead on zenoss
[12-Jul-2010 09:18:44] ashp: i think it's pretty safe to say "what snmp returns me, is what is mounted", so you could simplify your check to "compare fstab to what is mounted"
[12-Jul-2010 09:19:05] Parabola: it's not going to detect that it's back up instantly :P
[12-Jul-2010 09:19:11] Parabola: 5 minute polling cycles
[12-Jul-2010 09:19:20] Dieterbe: Yeah, I've never written any custom zenoss stuff as I only just signed off on purchasing enterprise, but i think i can probably work it out
[12-Jul-2010 09:19:22] wow
[12-Jul-2010 09:19:23] it can't be too hard :)
[12-Jul-2010 09:19:26] is there a way to change that?
[12-Jul-2010 09:19:33] for certain devices**
[12-Jul-2010 09:19:43] Parabola: it's a global setting per collector
[12-Jul-2010 09:19:49] crap
[12-Jul-2010 09:20:03] if i setup a remote collector later for other offices, would i be able to change the value there?
[12-Jul-2010 09:20:09] yes
[12-Jul-2010 09:20:18] alright
[12-Jul-2010 09:20:23] where do i change that value then?
[12-Jul-2010 09:20:29] this is all on the same subnet right now anyways
[12-Jul-2010 09:20:36] click collectors
[12-Jul-2010 09:20:39] ashp: are you good at bash scripting? just check if the important filesystems from /etc/fstab are listed in /proc/mounts
[12-Jul-2010 09:20:47] click localhost
[12-Jul-2010 09:20:52] ah gotcha
[12-Jul-2010 09:20:57] i didnt even notice collectors down there :)
[12-Jul-2010 09:21:05] Process Cycle Interval (secs)
[12-Jul-2010 09:21:07] ashp: you should use a nagios-like script, it's the easiest and works fine. many examples online as well (monitoringexchange.org for instance)
[12-Jul-2010 09:21:09] is what you're looking for
[12-Jul-2010 09:21:09] how much bandwidth and CPU load would i expect to get from this if i set to 60 seconds?
[12-Jul-2010 09:21:22] you can also change SNMP Performance Cycle Interval (secs)
[12-Jul-2010 09:21:37] you could just change the value for process cycle interval globally
[12-Jul-2010 09:21:54] that would monitor all processes at that interval but all other snmp info will be collected at the regular 5 mins
[12-Jul-2010 09:22:08] oh that sounds like a better idea
[12-Jul-2010 09:22:14] is that also under the collector settings?
[12-Jul-2010 09:22:20] yeh
[12-Jul-2010 09:22:27] it's the third option down
[12-Jul-2010 09:22:30] yeah i see it :)
[12-Jul-2010 09:22:32] awesome
[12-Jul-2010 09:22:39] I'm not sure how that'll affect your graphing of process details
[12-Jul-2010 09:22:45] :-\
[12-Jul-2010 09:22:47] since the graphs are set for 5 mins by default
[12-Jul-2010 09:22:53] my concern here is going to be cpu load from this
[12-Jul-2010 09:23:02] if i change to 30 seconds, or 60 seconds will it blow my shit up?
[12-Jul-2010 09:23:09] well, how many processes are we talking here?
[12-Jul-2010 09:23:13] 10 maybe
[12-Jul-2010 09:23:27] theres 10 total added, and theres 3 or 4 boxes right now being watched
[12-Jul-2010 09:23:29] 60 seconds won't blow anything up with 10 processes
[12-Jul-2010 09:23:36] alright
[12-Jul-2010 09:23:46] that means it's basically only collecting 40 OIDs per minute
[12-Jul-2010 09:23:51] which is a cakewalk for Zenoss
[12-Jul-2010 09:24:04] what value would i need to start worrying?
[12-Jul-2010 09:24:25] lol, maybe like 10 seconds or something
[12-Jul-2010 09:24:29] lol
[12-Jul-2010 09:24:32] depends on a few things
[12-Jul-2010 09:24:37] one second it is then
[12-Jul-2010 09:24:38] I wouldn't set it that low though
[12-Jul-2010 09:24:42] hehe
[12-Jul-2010 09:24:57] 60 seconds is more than enough in my opinion
[12-Jul-2010 09:24:58] after i change that value do i need to restart any services, or should it just work
[12-Jul-2010 09:25:02] 60 seconds is fine yes
[12-Jul-2010 09:25:22] I think it just works, but I've never personally tried
[12-Jul-2010 09:25:27] it may require restarting Zenoss
[12-Jul-2010 09:26:16] the process graphs for cpu, memory, and process count will be affected by the change since the RRD is configured for 5 minute intervals, so they probably won't scale properly...
[12-Jul-2010 09:26:26] I'll check and see what needs to be changed for that
[12-Jul-2010 09:26:37] I know it's something in the template
[12-Jul-2010 09:26:58] i dont mind if the graphs for process are messed up, cpu and mem i do though
[12-Jul-2010 09:27:17] oh if you can find what i need to change to fix it, that would be incredible
[12-Jul-2010 09:28:28] the cpu and memory graphs per process will be messed up, not the ones on the actual perf tab for the device
[12-Jul-2010 09:28:51] oh the perf tab is where i view that data anyway
[12-Jul-2010 09:28:55] so thoes will still be fine?
[12-Jul-2010 09:29:10] yup
[12-Jul-2010 09:29:25] the graphs that will be messed up are the ones that you see when you actually click on the process name
[12-Jul-2010 09:29:46] if I can remember how to fix that I'll let you know
[12-Jul-2010 09:29:58] I know RRD options need to be changed
[12-Jul-2010 09:31:09] oh yeh, the default rrd create command... need to override that in the template, I just need to figure out the math to get it down to 60 seconds properly
[12-Jul-2010 09:34:03] how did you become so versed in this product rmatte?
[12-Jul-2010 09:34:08] just a long time running? hehe
[12-Jul-2010 09:35:05] Parabola: over a year, yeh
[12-Jul-2010 09:35:18] rmatte, i've been googling and i'm trying to find a way to do remediation
[12-Jul-2010 09:35:38] for linux and windows would be great, but right now linux is really the only place its needed
[12-Jul-2010 09:36:04] i assume since it supports SSH it cant be much of a stretch to do
[12-Jul-2010 09:36:39] oh, well, the only way to do that really is via Event Manager -> Commands
[12-Jul-2010 09:36:54] basically you add a command and setup filters that need to match to kick off the command
[12-Jul-2010 09:37:03] stuff like "if event summary contains whatever"
[12-Jul-2010 09:37:18] event manager, nice , i'll take a look now
[12-Jul-2010 09:37:42] under commands, i assume?
[12-Jul-2010 09:37:47] correct
[12-Jul-2010 09:38:09] you could code an expect script to login and restart a process, then set it to kick off on an event down for that process
[12-Jul-2010 09:38:30] you can pass stuff from zenoss to the command such as device name, ip, etc...
[12-Jul-2010 09:38:39] ip is obviously recommended
[12-Jul-2010 09:38:43] yeah
[12-Jul-2010 09:39:25] is there any documentation on this? my reason for asking is i assume i have to have the command, ssh in, accept cert, provide password, then run the actual needed command
[12-Jul-2010 09:40:00] here's an example of a command that we used to run...
[12-Jul-2010 09:40:03] Parabola: have a read up on ssh using keys, vs passwords
[12-Jul-2010 09:40:05] /mspfiles/scripts/create_ticket.pl -autonotify -customer 'nnc-00001' -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
[12-Jul-2010 09:40:36] so to pass the IP to the script you use: ${dev/manageIp}
[12-Jul-2010 09:41:12] wow thats quite a mouthful
[12-Jul-2010 09:41:19] Parabola: you're not familliar with expect scripts?
[12-Jul-2010 09:41:24] not at all, no sir
[12-Jul-2010 09:41:47] I see, it's not zenoss specific, it's a unix scripting language
[12-Jul-2010 09:41:48] i'm reading up now
[12-Jul-2010 09:41:52] yeah i see that
[12-Jul-2010 09:42:06] rather than wasting your time, i'll read up on this, if i cant get my script to work, may i ask for assistance?
[12-Jul-2010 09:42:09] basically you just want to make a script that triggers ssh, logs in automatically, runs the command to restart the process, then exits
[12-Jul-2010 09:42:18] and you want to be able to pass an IP to the script
[12-Jul-2010 09:42:35] sure, my expect is rusty but I'll do what I can
[12-Jul-2010 09:43:21] there are probably some examples out there that mostly match what you want
[12-Jul-2010 09:44:38] :)
[12-Jul-2010 09:47:54] Parabola: http://www.linuxjournal.com/article/3065
[12-Jul-2010 09:48:03] that's worth a read before you get started
[12-Jul-2010 09:48:40] nice thanks
[12-Jul-2010 09:48:47] it explains how to use autoexpect, which could be used to at least generate a base line script
[12-Jul-2010 09:49:04] then you just need to modify it to be able to pass the IP to it, which isn't overly difficult from what I recall
[12-Jul-2010 09:49:18] actually, I think I have an old script laying around that I pass an argument to...
[12-Jul-2010 09:50:32] yeh, here is it: http://fpaste.org/Tyf3/
[12-Jul-2010 09:50:46] so you just need the: set ipaddr [lrange $argv 0 0] line
[12-Jul-2010 09:50:46] can you add port 1 on rack switch 7 to vlan 8 for me please (Windows Corp Netowork)
[12-Jul-2010 09:50:59] then you can use $ipaddr in the script to represent the IP
[12-Jul-2010 09:51:00] sorry wrong windows
[12-Jul-2010 09:51:05] use autoexpect to generate the rest
[12-Jul-2010 09:51:05] st3v3o: sure thing :D
[12-Jul-2010 09:51:43] lol
[12-Jul-2010 09:52:08] Thanks Simon4 let me know when you've got that done :P
[12-Jul-2010 09:52:24] I'll just need the enable password ;)
[12-Jul-2010 09:53:56] eugh, just saw an email with someone here asking a client to provide us with serial numbers to new cisco equipment... it's called "show ver" dummy
[12-Jul-2010 09:54:11] :|
[12-Jul-2010 09:55:16] thanks for the help :)
[12-Jul-2010 09:55:21] ill be afkish while i read this
[12-Jul-2010 09:55:23] no problem
[12-Jul-2010 09:55:26] if you need me just say my name
[12-Jul-2010 09:55:32] ill be workin on this
[12-Jul-2010 09:56:18] k
[12-Jul-2010 09:59:37] hi all, last time i have posted some of my zenpacks with Zenoss 3.0 support, anyone tried them? I need feedback.
[12-Jul-2010 10:01:59] nope sorry mate
[12-Jul-2010 10:02:01] still on 2.5
[12-Jul-2010 10:04:41] Parabola: all my Zenpacks are backward compatible with Zenoss 2.5.2
[12-Jul-2010 10:05:01] egor yeah i'm using some of yours right now on 2.5.2 :) they work great, good work
[12-Jul-2010 10:06:32] bigegor: I will install the wmi datasource on my 3.0 dev box this afternoon and test
[12-Jul-2010 10:06:37] its the most important zenpack i use :)
[12-Jul-2010 10:06:52] need to update to 3.0beta4 first tho
[12-Jul-2010 10:08:25] mrayzenoss is now known as mray
[12-Jul-2010 10:09:10] twm1010: you can use egg packages from people/bigegor/blog/2010/06/30/some-more-zenpacks-for-zenoss-30-beta or build yourself from SVN
[12-Jul-2010 10:17:41] i wish someone would build a zenpack for the IBM XIV storage array :P
[12-Jul-2010 10:21:16] bigegor for your wmi data source zenpack … Is there a way I can use the deamon via command line for scripting purposes? or would you just recommand wmic
[12-Jul-2010 10:22:40] st3v3o: use wmic if you're doing it via commandline
[12-Jul-2010 10:22:44] Greetings everyone, back from vacation
[12-Jul-2010 10:22:51] mray: welcome back
[12-Jul-2010 10:22:57] st3v3o: you can use WMIClient.py
[12-Jul-2010 10:23:12] mray: hello
[12-Jul-2010 10:23:25] rmatte: thanks
[12-Jul-2010 10:23:34] bigegor: hello
[12-Jul-2010 10:23:44] anything happen around here while I was gone?
[12-Jul-2010 10:24:03] the channel fell in to a black hole but we recovered it
[12-Jul-2010 10:24:10] apart from that nothing much
[12-Jul-2010 10:24:15] :)
[12-Jul-2010 10:24:17] sounds good, I'm sure zenoss-logger caught it
[12-Jul-2010 10:24:19] lol
[12-Jul-2010 10:24:21] lol
[12-Jul-2010 10:24:21] hehe
[12-Jul-2010 10:24:24] i used autoexpect
[12-Jul-2010 10:24:25] it was fail
[12-Jul-2010 10:24:32] Parabola: really? hmmm
[12-Jul-2010 10:24:37] yup
[12-Jul-2010 10:24:47] Parabola: oh well, hunt down some examples
[12-Jul-2010 10:25:03] it was very messy
[12-Jul-2010 10:25:06] and took up a ton of lines
[12-Jul-2010 10:25:14] i'm going to try again with somthing more basic
[12-Jul-2010 10:26:41] k cool
[12-Jul-2010 10:26:48] yeah
[12-Jul-2010 10:26:56] once you have one written you can just recycle it
[12-Jul-2010 11:11:22] hmmm, if I want a threshold to trigger at 0, do I set the min value to 0 or 1?
[12-Jul-2010 11:11:26] does anyone know?
[12-Jul-2010 11:11:34] I'm thinking it's 0, but I could be wrong
[12-Jul-2010 11:15:40] do thresholds trigger right on the value set or when the value set is passed?
[12-Jul-2010 11:42:09] guess I'll just set it to 1 and hope for the best
[12-Jul-2010 11:42:23] not a big deal if it's 1 or 0 in this case, but was curious for the future
[12-Jul-2010 11:42:38] I'd guess it would be 1...
[12-Jul-2010 11:43:06] meaning "1 is the min value that things are still ok"
[12-Jul-2010 11:43:43] yeh, I actually just confirmed that by looking at chet's interface operational status threshold setup
[12-Jul-2010 11:43:50] he has max and min as 1 meaning 1 is fine
[12-Jul-2010 11:43:59] so that must be it
[12-Jul-2010 11:44:01] thanks :)
[12-Jul-2010 13:00:20] hey rmatte you here mate
[12-Jul-2010 13:00:27] the pre-upgrade zenpack for 3.0rc1
[12-Jul-2010 13:00:34] is that for previous betas as well?
[12-Jul-2010 13:00:45] i've used auto-expect script to generate a script, however im getting some weird errors
[12-Jul-2010 13:00:46] or just 2.5.x -> 3.0RC1
[12-Jul-2010 13:00:53] and was hoping you'd be able to help
[12-Jul-2010 13:47:04] 3.0RC1 starting :)
[12-Jul-2010 13:59:58] dibbz: ?
[12-Jul-2010 14:04:07] forsberg is now known as fOrsberg
[12-Jul-2010 14:31:54] http://github.com/nickanderson/ZenPacks.community.CheckDomain http://github.com/nickanderson/ZenPacks.community.NginxStatus
[12-Jul-2010 14:35:23] nickanderson: how's that different from docs/DOC-4551 ?
[12-Jul-2010 14:37:19] dosnt look like its very different, hadnt seen that one
[12-Jul-2010 14:38:00] I can publish that one too :)
[12-Jul-2010 14:39:15] i didnt see any nginx ones
[12-Jul-2010 14:39:21] so that one should be new
[12-Jul-2010 14:40:37] yeah, that's new to me
[12-Jul-2010 14:41:16] I've got 5 in my queue, including your nginx
[12-Jul-2010 14:41:37] sounds like your staying busy
[12-Jul-2010 14:41:49] I did 12 over my vacation :)
[12-Jul-2010 14:46:52] working on vacation, tsk tsk tsk
[12-Jul-2010 14:49:30] hey rmatte
[12-Jul-2010 14:49:35] so i've got my expect script done
[12-Jul-2010 14:50:10] cool
[12-Jul-2010 14:50:27] yeah took me a bit hehe, but i get it now
[12-Jul-2010 14:50:34] so im trying to setup my event
[12-Jul-2010 14:50:41] i have my device class and event class selected
[12-Jul-2010 14:50:55] but how do i specify, say if this particular OS process stops?
[12-Jul-2010 14:51:30] ok, you need to go back to Event Manager -> Commands
[12-Jul-2010 14:51:32] then add a command
[12-Jul-2010 14:51:41] and you'll see filter options at the bottom when you go in to it
[12-Jul-2010 14:51:43] right
[12-Jul-2010 14:51:51] which one of these can i pick a specific process? im not seeing it
[12-Jul-2010 14:51:53] you need to configure those filter options to match that particular event
[12-Jul-2010 14:52:13] so if event comes in for this device, and the summary contains this (process name), do this
[12-Jul-2010 14:52:34] you don't pick the specific process, you'd do something like: if summary contains mysqld
[12-Jul-2010 14:52:41] or whatever
[12-Jul-2010 14:52:52] you'll probably want to make it as specific as possible
[12-Jul-2010 14:53:36] gotcha
[12-Jul-2010 14:53:39] im testing now, i stopped the service
[12-Jul-2010 14:53:44] going to check out the actual event
[12-Jul-2010 14:58:45] cool
[12-Jul-2010 14:59:39] I assume you tested the expect script already?
[12-Jul-2010 15:19:30] yeah
[12-Jul-2010 15:19:33] it worked by hand
[12-Jul-2010 15:20:47] however the script i tested didnt start i service, i created a txt file instead
[12-Jul-2010 15:20:57] lemme try this
[12-Jul-2010 15:24:12] how odd
[12-Jul-2010 15:24:25] the snmp service on a large number of AD servers decided to simply stop responding
[12-Jul-2010 15:30:55] hi: i have servers in a linux class that when i jack the load to 100% cpu and a load of 7 it doesn't send any events. Do I need to enable something so it alerts when the server is under high load?
[12-Jul-2010 15:34:58] actually, looks like i have to set a threshold
[12-Jul-2010 15:38:58] robo: yes, set a threshold in the templatwe
[12-Jul-2010 15:39:00] template*
[12-Jul-2010 15:39:09] the default linux thresholds are garbage
[12-Jul-2010 15:39:21] Hi everyone, can a device belong to two or more class?
[12-Jul-2010 15:39:36] chachan: no, it can only belong to 1 class, but it can belong to multiple groups
[12-Jul-2010 15:41:20] yeah, it has a high load threshhold of 1200. I jumped that down to 600 and hit a load of 15 and still no alert
[12-Jul-2010 15:41:25] i wonder how low that needs to be...
[12-Jul-2010 15:42:08] rmatte, can I set my own events to groups?
[12-Jul-2010 15:42:22] chachan: what do you mean set events to groups?
[12-Jul-2010 15:43:55] rmatte, I want to create "event commands" to check web service content
[12-Jul-2010 15:44:27] so, I will need group some device to add them these "events commands"
[12-Jul-2010 15:44:36] need to*
[12-Jul-2010 15:44:44] chachan: explain "event commands"
[12-Jul-2010 15:44:51] I don't really see what you're getting at with that
[12-Jul-2010 15:45:30] explain your exact goal step by step so that I understand
[12-Jul-2010 15:46:33] rmatte, ok, I checking out the administration guide, and it says I can create "event commands" to be execute it repeatly, I want to create events to check web service content
[12-Jul-2010 15:46:50] did I make myself clear?
[12-Jul-2010 15:47:56] it says I can create them using TALES syntax
[12-Jul-2010 15:48:46] do you want to know where is this point in the guide?
[12-Jul-2010 15:49:06] well, I understand event commands, but I think you're missing the point of what an event command is
[12-Jul-2010 15:49:18] it's a command that triggers if an alert actually comes in and matches the rules
[12-Jul-2010 15:49:32] unless you're talking about command datasources?
[12-Jul-2010 15:49:39] oh, I understand
[12-Jul-2010 15:49:42] what are you exactly trying to check?
[12-Jul-2010 15:49:46] give me a scenario
[12-Jul-2010 15:51:30] rmatte, I'm trying to check web service, an example could be that I execute: "php -c http://foo.bar", and it returns me something like "1" for "ok" status or "0" for "fail" status
[12-Jul-2010 15:52:17] so you're trying to make sure that a website is functioning properly basically?
[12-Jul-2010 15:52:52] These zenpacks are what you want, specifically the second one, but the first may be useful too...
[12-Jul-2010 15:52:54] docs/DOC-3431
[12-Jul-2010 15:52:58] docs/DOC-3757
[12-Jul-2010 15:53:35] rmatte, ok, I'll check these out
[12-Jul-2010 15:53:39] rmatte, thanks
[12-Jul-2010 15:53:42] np
[12-Jul-2010 15:54:26] rmatte, could I group machines to differents configurations of this zenpack?
[12-Jul-2010 15:54:41] rmatte, or isn't needly?
[12-Jul-2010 15:54:53] you can create your own event classes and copy the template over then modify it however you like
[12-Jul-2010 15:55:57] rmatte, ok :)
[12-Jul-2010 15:56:01] but remember that a device can't be in multiple classes
[12-Jul-2010 15:56:46] :)
[12-Jul-2010 16:01:32] can someone point me in the correct direction for getting disk io graphs for linux hosts? i can't seem to get this working w/ this latest installation.
[12-Jul-2010 16:01:36] running 2.5.2
[12-Jul-2010 16:01:55] added the ucddisk collector to the linux device class
[12-Jul-2010 16:02:03] nonsenso they are working fine for me in 2.5.2, so it's definitely not Zenoss causing the issue
[12-Jul-2010 16:02:14] I assume you're referring to the IO graph on the perf tab?
[12-Jul-2010 16:02:38] yeah, i bet it's something i'm doing wrong (tm)
[12-Jul-2010 16:02:45] ucddisk?
[12-Jul-2010 16:02:58] no, i'm trying to add disk io graphs in the hw tabs for each disk
[12-Jul-2010 16:03:17] oh, I've never seen that done
[12-Jul-2010 16:03:30] Is there even an OID available via net-snmp to support that?
[12-Jul-2010 16:03:33] yeah
[12-Jul-2010 16:03:52] then it shouldn't be rocket science to do
[12-Jul-2010 16:04:12] i know! hehe. i've been smacking my head on this one probably for too long.
[12-Jul-2010 16:04:37] What's the OID?
[12-Jul-2010 16:04:45] when i model the device i get the error that the OID is bad despite the fact that i can walk it through zenoss and via cli w/o issue
[12-Jul-2010 16:05:00] well, here's the 4 i'm looking at:
[12-Jul-2010 16:05:01] 1.3.6.1.4.1.2021.13.15.1.1.3
[12-Jul-2010 16:05:04] 1.3.6.1.4.1.2021.13.15.1.1.4
[12-Jul-2010 16:05:05] right, but you have to understand that Zenoss is using a collector plugin to collect that data
[12-Jul-2010 16:05:08] 1.3.6.1.4.1.2021.13.15.1.1.5
[12-Jul-2010 16:05:10] 1.3.6.1.4.1.2021.13.15.1.1.6
[12-Jul-2010 16:05:11] so it tacks on the disk index automatically
[12-Jul-2010 16:05:26] so when you add the datasource you just want to add it as: 1.3.6.1.4.1.2021.13.15.1.1
[12-Jul-2010 16:05:49] er, how do i get points on a graph then?
[12-Jul-2010 16:05:59] i'm using those OIDs as datasources
[12-Jul-2010 16:06:08] create a datasource with that oid, create a graph, then add the datasource to the graph
[12-Jul-2010 16:06:10] in the graph i setup (that's not wrking)
[12-Jul-2010 16:06:25] you can't use those OIDs, you need to shave off the last number
[12-Jul-2010 16:06:30] since Zenoss will add that automatically
[12-Jul-2010 16:06:42] each of those OIDs represents a disk
[12-Jul-2010 16:06:42] aight. lemme try that. thx rmatte. :)
[12-Jul-2010 16:06:52] you need to edit the /Server/FileSystem template
[12-Jul-2010 16:06:59] and add a new data source as: 1.3.6.1.4.1.2021.13.15.1.1
[12-Jul-2010 16:07:02] then add a graph
[12-Jul-2010 16:07:06] and add a datasource to the graph
[12-Jul-2010 16:07:08] and that should do it
[12-Jul-2010 16:07:22] er, no those OIDs represent the written/wrote operations both in accesses and byes.
[12-Jul-2010 16:07:46] so the disk index gets appended to the end of those then?
[12-Jul-2010 16:07:49] or what?
[12-Jul-2010 16:07:52] if you append an index number to them then you get the disk
[12-Jul-2010 16:07:52] yeah
[12-Jul-2010 16:07:57] I see
[12-Jul-2010 16:08:08] well those should work then, if they're not I have no idea why
[12-Jul-2010 16:08:44] yeah, it's probably something boneheaded or somesuch.
[12-Jul-2010 16:08:47] the only templat they'll work in (if they do work) is the FileSystem template under /Server
[12-Jul-2010 16:09:50] though you'll probably want to make a copy of it in /Server/Linux before editing it
[12-Jul-2010 16:10:02] since those OIDs probably won't apply to other types of snmp enabled servers
[12-Jul-2010 16:10:08] yeah i don't think they do.
[12-Jul-2010 16:10:10] so you don't want to edit it for the whole /Server class
[12-Jul-2010 16:10:14] nod.
[12-Jul-2010 16:29:45] rmatte
[12-Jul-2010 16:30:04] two things, firstly i've got a working script, assuming it doesnt prompt for a yes/no before connecting
[12-Jul-2010 16:30:15] (which i need help on that if statement)
[12-Jul-2010 16:30:31] actually its just that, there is no other
[12-Jul-2010 16:35:43] that's easy
[12-Jul-2010 16:35:46] one sec
[12-Jul-2010 16:36:00] here i'll pastebin you the script
[12-Jul-2010 16:36:18] here...
[12-Jul-2010 16:36:19] http://fpaste.org/Tt3g/
[12-Jul-2010 16:36:26] that accounts for the yes/no
[12-Jul-2010 16:36:32] just add that in to your script
[12-Jul-2010 16:37:11] oh nice
[12-Jul-2010 16:37:16] hmmm, it still has my old password, oh well, it's been changed since lol
[12-Jul-2010 16:37:19] that set ipaddr line, what does that mean
[12-Jul-2010 16:37:29] that's to pass the IP to the script
[12-Jul-2010 16:37:41] so you'd execute the script like /script.xpt 192.168.0.1
[12-Jul-2010 16:37:53] and the script grabs the 192.168.0.1 and sets it to ipaddr
[12-Jul-2010 16:38:01] oo
[12-Jul-2010 16:38:01] which you can then call on in the script with $ipaddr
[12-Jul-2010 16:38:19] which is what you need if you want it to apply to multiple devices
[12-Jul-2010 16:38:30] you'll have to do 1 script per process though
[12-Jul-2010 16:40:32] what does interact do
[12-Jul-2010 16:41:35] that just hands it over to me so that I can type commands
[12-Jul-2010 16:41:41] it's just an auto-login script
[12-Jul-2010 16:41:56] you wouldn't normally use that in an expect script
[12-Jul-2010 16:46:37] damn
[12-Jul-2010 16:46:38] its not working
[12-Jul-2010 16:47:10] so it's not kicking off the script?
[12-Jul-2010 16:47:13] no clue
[12-Jul-2010 16:47:15] how would i tell
[12-Jul-2010 16:47:29] you should make a shell script that writes to a file then executes the script
[12-Jul-2010 16:47:36] so that you have some indication whether it ran or not
[12-Jul-2010 16:47:51] and then execute that shell script instead
[12-Jul-2010 16:48:18] then change it to execute the script directly once you have it working
[12-Jul-2010 16:48:38] man is there an easier way
[12-Jul-2010 16:48:42] i cant get ssh on the server right now
[12-Jul-2010 16:48:44] also, be aware that the script will only execute if the event count is 1
[12-Jul-2010 16:48:48] ooo
[12-Jul-2010 16:48:51] so if the event has been racking up event counts, clear it
[12-Jul-2010 16:48:58] then when the new event comes in it'll kick it off
[12-Jul-2010 16:49:32] if i delete the event will that clear it?
[12-Jul-2010 16:49:44] deleting == moving to history
[12-Jul-2010 16:49:50] you never actually delete events, they just get moved
[12-Jul-2010 16:49:52] but yes
[12-Jul-2010 16:49:54] that will do it
[12-Jul-2010 16:54:40] hmm
[12-Jul-2010 16:55:39] I have to take off for the day in a minute, I'll be around tomorrow
[12-Jul-2010 16:55:40] cheers
[12-Jul-2010 16:55:51] later mate
[12-Jul-2010 17:08:48] rmatte, HTTPMonitor only works with apche?, I need apache and nginx for check for specific content as result in a web service
[12-Jul-2010 17:12:02] chachan: I should be publishing nickanderson's nginx in the next couple of days: http://github.com/nickanderson/ZenPacks.community.NginxStatus/tree/1.0 is the source
[12-Jul-2010 17:13:25] cool, I'll be aware about it
[12-Jul-2010 17:57:10] dang it. still no love. i did double check my previous zenoss installations and this was working in older versions..
[12-Jul-2010 17:58:06] (re: using ucd disk mib to get bytes written and read io graphs per hdd/partition)
[12-Jul-2010 22:57:42] Hey guys can I use zenoss to track sw licenses? I would input the license info manually but want alerts from zenoss
[12-Jul-2010 23:01:58] Mray, great talk on floss weekly
[12-Jul-2010 23:05:37] has anyone been able to get zenoss to view the event logs on win boxes? i have the centos zenoss server joined to my AD domain but that doesnt seem to help
[13-Jul-2010 00:00:30] [disconnected at Tue Jul 13 00:00:30 2010]
[13-Jul-2010 00:00:30] [connected at Tue Jul 13 00:00:30 2010]
[13-Jul-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]
[13-Jul-2010 02:08:08] fOrsberg is now known as forsberg
[13-Jul-2010 02:58:03] Hello. I am trying to get a zenpack installed from the command line, but I am getting: ERROR: zenpack command failed. Reason: ZipImportError: not a Zip file: '/home/zenoss/ZenPacks.community.NginxStatus-1.0-py2.4.egg'
[13-Jul-2010 03:08:50] nevermind that. I am now getting: ERROR: zenpack command failed. Reason: KeyError: 'nickanderson-ZenPacks.community.NginxStatus-d1975e8'
[13-Jul-2010 03:37:51] re: the db.xvidoes.xnxx error, is this a case where running a normal repair tables command would be the thing to do?
[13-Jul-2010 03:38:49] wrong window... sorry
[13-Jul-2010 05:08:25] hmm
[13-Jul-2010 05:09:09] when SNMP traps come in, what's the first point I could safely run transforms that would apply to all incoming traps?
[13-Jul-2010 05:09:21] some of these end up in /Unknown
[13-Jul-2010 05:09:43] * Simon4 waits for someone to answer that so he can use the answer for his own evil purposes :)
[13-Jul-2010 05:10:52] haha
[13-Jul-2010 05:13:00] Simon4: blunt approach is to add the transform in /Events, but loadwise I dunno if I'd want to do that in general
[13-Jul-2010 05:13:07] frozty_sa: yeah, agreed
[13-Jul-2010 05:13:24] a simple match on whether the event is snmptrap should be fine to prevent unnecessary computing, but even so..
[13-Jul-2010 05:13:38] * frozty_sa digs around the daemons
[13-Jul-2010 05:20:18] Simon4: actually, looks like doing them at /Events might be the better option, based on servlet/JiveServlet/showImage/102-4766-20-2420/Event%2520Processing.png
[13-Jul-2010 05:28:06] forsberg is now known as fOrsberg
[13-Jul-2010 05:28:43] is there an easy way to see the traffic (performance graphs) of all interfaces on a switch?
[13-Jul-2010 05:30:37] fOrsberg is now known as forsberg
[13-Jul-2010 05:34:25] offhand, I think the quickest way would be a report
[13-Jul-2010 05:34:25] not sure if you could make it auto-iterate through all interfaces, checking now quickly
[13-Jul-2010 05:36:04] fragfutter: a graph report is the fastest way
[13-Jul-2010 05:36:15] I keep meaning to make a zenpack to add a tab for "all interfaces"
[13-Jul-2010 05:36:22] would be a cool feature :)
[13-Jul-2010 05:36:36] Simon4: indeed
[13-Jul-2010 05:37:06] I've sometimes felt that I want some of cacti's presentation-ability in zenoss, but it's not end-of-the-world stuff for us just yet
[13-Jul-2010 05:37:26] our device scale is small enough to run both concurrently, and trending is definitely something cacti does "better"
[13-Jul-2010 05:37:51] speaking of trending, I need to try and figure out why I can't get the holt-winters thresholds zenpack to work
[13-Jul-2010 05:38:11] fragfutter: /Reports /Performance Reports /Interface Utilization
[13-Jul-2010 06:04:42] Can I use zenoss to track sw licenses?
[13-Jul-2010 06:07:47] That is add sw license info of to machines detected by zenoss and get a notification when a license expires?
[13-Jul-2010 06:10:54] you could make a zencommand to do that, yes
[13-Jul-2010 06:11:38] have the zencommand run whatever the command is to check whether a license is expired or not, then work with that
[13-Jul-2010 06:17:22] frozty_sa: thanks but interface utilization is not realy what i need. i could create a multigraph report for every switch. But i'm not too happy about this (even if i do this automaticly with a script)
[13-Jul-2010 06:17:51] fragfutter: yeah, I know what you mean
[13-Jul-2010 06:18:13] like I said, it's a bit iffy still in zenoss when you're coming from a cacti background
[13-Jul-2010 06:18:24] correct.
[13-Jul-2010 06:19:52] my favorite missing features for zenoss are LLDP (layer2 topology) support and overview reports
[13-Jul-2010 06:21:24] the LLDP stuff is a regular point of discussion, and it's something on my potential todo list to see what I can get added to it
[13-Jul-2010 06:21:48] the current way zenoss handles network maps is routing-table based, which has its downsides
[13-Jul-2010 06:31:36] Frotz tks
[13-Jul-2010 06:59:40] what is zenoss using for snmp polling? (which python library or binding?)
[13-Jul-2010 07:05:13] twisted
[13-Jul-2010 08:23:16] woo
[13-Jul-2010 08:23:20] * Simon4 looks at the zopedb stats
[13-Jul-2010 08:23:42] 100 object reads /second on our Data.fs
[13-Jul-2010 08:31:59] how do you determine that?
[13-Jul-2010 08:33:59] jb: http://servername:8080/manage - needs the admin password
[13-Jul-2010 08:34:26] then you go to database under control panel and there's an activity tab
[13-Jul-2010 08:35:40] hm
[13-Jul-2010 08:35:42] 72 object loads
[13-Jul-2010 08:35:46] 582 connections
[13-Jul-2010 08:36:11] * Simon4 is getting 36k objects loads per 6 minutes
[13-Jul-2010 08:36:15] we have lots of users :)
[13-Jul-2010 08:36:37] good morning!
[13-Jul-2010 08:38:32] i've got a router that has a dynamic ip. i added the hostname and it picks it up, however when the IP changes, to "fix" the alerts i have to clear out the IP, then it will grab the new one
[13-Jul-2010 08:38:46] is there a fix for this?
[13-Jul-2010 08:39:38] Parabola: that's an interesting one
[13-Jul-2010 08:39:44] since zenoss keys devices by their IP
[13-Jul-2010 08:39:48] yeah
[13-Jul-2010 08:39:53] yeah, zabbix didnt do this, i'm being told
[13-Jul-2010 08:40:04] and i quote "zabbix didnt do this, it would update the IP"
[13-Jul-2010 08:40:23] so it's just the alerts not clearing that's the issue?
[13-Jul-2010 08:40:37] or is it that when the ip changes you lose data because zenoss doesn't pick it up until remodel time?
[13-Jul-2010 08:41:48] Parabola: It's always puzzled my why people use dynamic IPs on routers? Is there some special case scenario at play?
[13-Jul-2010 08:42:04] simon
[13-Jul-2010 08:42:12] it just blasts us wtih emails when the IP changes
[13-Jul-2010 08:42:19] and its not puzzling its called money
[13-Jul-2010 08:42:23] this company has 75 stores
[13-Jul-2010 08:42:38] they have regular cable connections at them
[13-Jul-2010 08:42:44] due to cost some of them dont have static
[13-Jul-2010 08:43:04] if i go into the device and clear it out it picks up the new IP
[13-Jul-2010 08:43:05] and all is well
[13-Jul-2010 08:43:22] I see
[13-Jul-2010 08:43:39] You couldn't have configured the sites to connect a tunnel back to the main site regardless of their IP?
[13-Jul-2010 08:45:18] honestly we arent going to reconfigure 75 sites just for the sake of monitoring
[13-Jul-2010 08:45:26] theres really no way to do this?
[13-Jul-2010 08:45:29] because thats 100% stupid
[13-Jul-2010 08:46:28] well what do you expect?
[13-Jul-2010 08:46:44] it to have at least the same functionality as the products its claiming to be better than
[13-Jul-2010 08:46:45] why don't you configure the sites with dyndns?
[13-Jul-2010 08:47:08] but even then you'd need some script in Zenoss to change the IP
[13-Jul-2010 08:47:39] it is using dyndns..
[13-Jul-2010 08:47:54] no-ip in particular but yeah, it is
[13-Jul-2010 08:48:01] ok, so your issue is that Zenoss isn't automatically grabbing the new IP?
[13-Jul-2010 08:48:06] yup :)
[13-Jul-2010 08:48:17] which for some stupid reason changes way more than i would like it too
[13-Jul-2010 08:48:19] like, weekly
[13-Jul-2010 08:48:45] I'm surprised, I had regular residential cable and my IP didn't change for over a year lol
[13-Jul-2010 08:49:38] well, the way I'd do it in your case is to write a script that constantly checks the hostnames of the devices to see if the IP has changed and updates Zenoss if it does, though it's a pretty hacky way of doing it
[13-Jul-2010 08:49:54] there really is no other way at the moment, that would be a good idea for a ZenPack though
[13-Jul-2010 08:50:40] which if i was able to create i would do that for everyone :)
[13-Jul-2010 08:50:51] so, back to expect for a script? lol
[13-Jul-2010 08:51:01] nah, it wouldn't be an expect script
[13-Jul-2010 08:51:10] where else can i put a script?
[13-Jul-2010 08:51:19] you mean a shells script and run via cron?
[13-Jul-2010 08:51:23] shell script*
[13-Jul-2010 08:51:24] yeh
[13-Jul-2010 08:51:45] I'm checking if there's a better solution
[13-Jul-2010 08:52:33] hmmm, I wish chet was around, he'd probably have a solution for that
[13-Jul-2010 08:53:05] let me test something...
[13-Jul-2010 08:54:14] unhunh, ok... this might be a better solution...
[13-Jul-2010 08:54:15] I'm wondering, I've upgraded my Zenoss install lots, and have the old 1.0.1 HTTPMonitor Zenpack
[13-Jul-2010 08:54:26] Can I just install the newer .egg and it will update
[13-Jul-2010 08:54:30] rmatte if i make a script that fixes this
[13-Jul-2010 08:54:32] or do I need to remove the old zenpack?
[13-Jul-2010 08:54:37] how would i create a zenpack for everyone?
[13-Jul-2010 08:54:41] are you able to configure a script on the devices that kicks off when they detect an IP change and update the no-ip entry?
[13-Jul-2010 08:55:09] I know that some dyndns clients support that sort of thing, I'm wondering if those devices do or not
[13-Jul-2010 08:55:52] If they do, then there's a really simple solution
[13-Jul-2010 08:58:02] well, it might be complex depending on the degree of scripting that they support (if any)
[13-Jul-2010 08:58:05] If I do remove the old HTTPMonitor zenpack, will it delete devices it's bound to if the devices are *not* in the Device Class it creates?
[13-Jul-2010 08:58:38] jmp242: no, only if they are in that class
[13-Jul-2010 08:59:03] jmp242: it removes the class it created and all objects under that class, but nothing else
[13-Jul-2010 08:59:16] there should be a warning message that explains that when removing ZenPacks in my opinion
[13-Jul-2010 08:59:32] It sucks when devices just magically dissapear
[13-Jul-2010 08:59:33] lol
[13-Jul-2010 09:00:35] man, there needs to be a zProperty that can be set to monitor via the hostname instead of the IP
[13-Jul-2010 09:01:10] There is a warning rmatte
[13-Jul-2010 09:01:14] hence my worry
[13-Jul-2010 09:01:29] but it wasn't specific enough - it said anything provided by it would be deleted
[13-Jul-2010 09:01:37] If that class shows up under the ZenPack's components list then you'll have problems
[13-Jul-2010 09:01:38] so I was wondering if I had to unbind the template everywhere
[13-Jul-2010 09:01:51] no, template bindings have no effect on devices
[13-Jul-2010 09:02:00] it won't delete the devices just because the template is bound to them
[13-Jul-2010 09:02:09] but if they are in the HTTP class that it created, then you'll lose them
[13-Jul-2010 09:02:11] ok, should be fine, I've got no devices in the deviec class it's going to kill
[13-Jul-2010 09:02:19] why are you removing it?
[13-Jul-2010 09:02:23] well
[13-Jul-2010 09:02:29] I can't create new data sources
[13-Jul-2010 09:02:33] as HTTPMon type
[13-Jul-2010 09:02:39] it lets me select it in the GUI
[13-Jul-2010 09:02:45] but then creates an SNMP dfata source
[13-Jul-2010 09:02:58] So I was going to upgrade to the .egg
[13-Jul-2010 09:03:05] but I don't know that I can do that in place or not
[13-Jul-2010 09:03:07] you restarted Zenoss after installing the ZenPack I assume?
[13-Jul-2010 09:03:28] I installed this zenpack on my Zenoss install back in version 2.0.2 if I recall correctly
[13-Jul-2010 09:03:34] oh
[13-Jul-2010 09:03:35] I've restarted a lot since then
[13-Jul-2010 09:03:37] that's ancient man
[13-Jul-2010 09:03:44] you should always upgrade core packs
[13-Jul-2010 09:03:44] especially to upgrade through 2.5.2 now
[13-Jul-2010 09:04:03] yeh
[13-Jul-2010 09:04:15] So can I just do an upgrade in place with the .egg?
[13-Jul-2010 09:04:20] Have you tried just installing the .egg without removing the .zip?
[13-Jul-2010 09:04:29] not yet
[13-Jul-2010 09:04:34] I didn't know if it would work
[13-Jul-2010 09:04:36] ZenPacks migrated from .zip to .egg are supposed to automatically handle that
[13-Jul-2010 09:04:47] just try installing the .egg and it should handle it
[13-Jul-2010 09:04:49] ok, I'll try just installing the .egg
[13-Jul-2010 09:04:53] cool
[13-Jul-2010 09:05:01] I usually don't change things until they don't work lol
[13-Jul-2010 09:05:04] lol
[13-Jul-2010 09:05:09] so I don't upgrade when I should always
[13-Jul-2010 09:05:12] kind of a bad habbit, but oh well
[13-Jul-2010 09:05:24] I find with Zenoss it's good to be current, but not too current
[13-Jul-2010 09:05:27] there's a happy balance
[13-Jul-2010 09:05:28] lol
[13-Jul-2010 09:05:29] Well, it's cyclical when I can work with Zenoss
[13-Jul-2010 09:05:37] I see
[13-Jul-2010 09:05:39] sometimes I just can't mess with it
[13-Jul-2010 09:05:42] and need it to be working
[13-Jul-2010 09:05:47] I work with it all the time, so I take it for granted
[13-Jul-2010 09:05:52] :)
[13-Jul-2010 09:06:12] the stability increase from 2.4.5 to 2.5.2 was excellent
[13-Jul-2010 09:06:35] I can actually leave my Zenoss servers running for over 2 months without worrying about them leaking memory all over the place
[13-Jul-2010 09:06:37] it's lovely
[13-Jul-2010 09:06:50] looks like it works
[13-Jul-2010 09:06:55] good stuff
[13-Jul-2010 09:06:57] the .zip is gone and the .egg is there
[13-Jul-2010 09:07:12] now to see if I can see inside the data source without a traceback lol
[13-Jul-2010 09:07:16] yeh, there's a special file in the .eggs of zenpacks that were migrated from .zips that tells them how to handle the upgrade
[13-Jul-2010 09:07:25] restart Zenoss before you try
[13-Jul-2010 09:09:16] Man, my new Zenoss server is a dream - much faster than the old one lol
[13-Jul-2010 09:09:37] restart done in 10 seconds instead of 5 mintues
[13-Jul-2010 09:15:53] nice
[13-Jul-2010 09:16:00] yeh, that is an improvement
[13-Jul-2010 09:16:23] I love having vms for my zenoss servers, can always go in and increase resources when needed
[13-Jul-2010 09:16:59] So in 2.5.2, why would it not let me add an IP Service on a device?
[13-Jul-2010 09:17:08] It just greys out the ok button
[13-Jul-2010 09:17:41] I assume it has something to do with that I can't do SNMP, but I should be able to see if port 80 is responding I would think...
[13-Jul-2010 09:17:42] you need to type the ip service in and wait for a dropdown, then select it from there
[13-Jul-2010 09:17:53] and then you'll see it in square brackets and you'll be able to click ok
[13-Jul-2010 09:18:01] sometimes it won't give you the dropdown, takes a few tries
[13-Jul-2010 09:18:07] so cancel out and open it again until it does
[13-Jul-2010 09:18:10] (it's buggy)
[13-Jul-2010 09:18:21] should be resolved in the new interface
[13-Jul-2010 09:18:51] I see
[13-Jul-2010 09:18:53] worked this time
[13-Jul-2010 09:18:58] it was a bit non-obvious
[13-Jul-2010 09:18:59] thanks
[13-Jul-2010 09:19:17] no problem, and yeh, it is unobvious
[13-Jul-2010 09:19:24] I just know from experience
[13-Jul-2010 09:19:25] lol
[13-Jul-2010 09:30:41] mrayzenoss is now known as mray
[13-Jul-2010 09:30:54] mray: good morning
[13-Jul-2010 09:31:53] rmatte: mornign
[13-Jul-2010 09:32:00] s/mornign/morning/
[13-Jul-2010 09:32:04] hehe
[13-Jul-2010 09:32:12] that actually works with Skype
[13-Jul-2010 09:32:26] seriously?
[13-Jul-2010 09:32:28] that's awesome
[13-Jul-2010 09:32:32] yup
[13-Jul-2010 09:35:42] son of a ...
[13-Jul-2010 09:35:52] another failed Time Machine recovery with OSX
[13-Jul-2010 09:36:02] that sucks
[13-Jul-2010 09:36:17] I thought time machine was supposed to be pretty foolproof lol
[13-Jul-2010 09:36:26] I'm a higher-grade of fool
[13-Jul-2010 09:38:38] so now I have to do a direct disk to disk copy, which means I have to be away from my computer for several hours…
[13-Jul-2010 09:56:17] bigegor: checked in the ZenPacks.community.SQLDatasource and ZenPacks.community.ZenODBC update last night
[13-Jul-2010 09:58:55] mray will they be posted at community/zenpacks
[13-Jul-2010 09:58:57] mray: thanks. both Zenpacks are up-to-date
[13-Jul-2010 10:04:40] st3v3o: yeah, soon
[13-Jul-2010 10:04:53] just need documentation :)
[13-Jul-2010 10:06:11] cool….SQLDataSource sounds promising
[13-Jul-2010 10:06:12] mray: i will do it
[13-Jul-2010 10:09:47] st3v3o: with SQLDataSource you can use any Pythons DB-API compatible modules to access database. And it has some additional features ported from WMIDataSource.
[13-Jul-2010 10:10:04] sweet
[13-Jul-2010 10:10:34] I really like the WMIDataSource...
[13-Jul-2010 10:10:53] :)
[13-Jul-2010 10:11:40] * rmatte wants to strangle whoever is in charge of decision concerning snmp at vmware
[13-Jul-2010 10:12:14] you can either collect vm info, or you can collector system info, but you can't collect both anymore
[13-Jul-2010 10:12:21] collect*
[13-Jul-2010 10:13:35] <|rt|> If I login as admin the google map shows all 3 locations but if I log in as my user I only see our main office on the map....I'm assuming it's a permission issue but have no idea where to look to correct it
[13-Jul-2010 10:14:56] |rt|: enterprise or core?
[13-Jul-2010 10:15:01] <|rt|> core
[13-Jul-2010 10:15:05] hmmm, that's odd then
[13-Jul-2010 10:15:26] are all of the locations at the same level on the map?
[13-Jul-2010 10:15:31] or do you have to drill down to some?
[13-Jul-2010 10:15:46] <|rt|> they are all at the same level
[13-Jul-2010 10:15:53] <|rt|> atleast as admin they are
[13-Jul-2010 10:16:05] have you tried clearing the geocache on your user?
[13-Jul-2010 10:16:35] Locations -> Dropdown Menu -> Clear Geocache Code
[13-Jul-2010 10:16:37] <|rt|> yeah....but do I need to go to each location to clear that or is that a global cache for all locations that are cleared
[13-Jul-2010 10:16:39] then view the map again
[13-Jul-2010 10:16:43] global
[13-Jul-2010 10:16:51] <|rt|> I've already tried that multiple times
[13-Jul-2010 10:17:01] have you tried modified the location addresses a bit?
[13-Jul-2010 10:17:05] to simplify them?
[13-Jul-2010 10:17:10] modifying*
[13-Jul-2010 10:17:20] <|rt|> and I can tell that it has indeed been cleared b/c you can see it reaquire the addresses
[13-Jul-2010 10:17:37] right, but after it re-acquires them, can you see those locations?
[13-Jul-2010 10:17:47] <|rt|> tried modifying it as well....the 2 addresses that don't show up in my user account are just city and state
[13-Jul-2010 10:17:55] I see
[13-Jul-2010 10:17:58] <|rt|> nope it's still just the one
[13-Jul-2010 10:18:19] <|rt|> I've tried putting actual street addresses for the other 2 locations to see if that would help as well to no avail
[13-Jul-2010 10:18:35] no idea then, I *highly* doubt it's a permissions issue since there aren't really any permissions in core which dictates who can or can't see specific locations
[13-Jul-2010 10:19:03] ACLs like that are only implemented in enterprise
[13-Jul-2010 10:19:29] <|rt|> hmm...let me try to create a new user and see what happens with it
[13-Jul-2010 10:19:36] I'm sure you could get at permissions like that via the zope management interface, but I doubt you've even been in there?
[13-Jul-2010 10:19:41] k
[13-Jul-2010 10:20:03] by the way, are there devices assigned to those locations right now?
[13-Jul-2010 10:20:10] or do they have 0 devices in them?
[13-Jul-2010 10:22:03] <|rt|> hmm new user can see all 3 locations
[13-Jul-2010 10:22:07] <|rt|> there are devices in each location
[13-Jul-2010 10:22:29] <|rt|> nope I haven't been into zope at all
[13-Jul-2010 10:22:58] <|rt|> let me see what happens if I change this new user to a ZenManager and then add them to the group my other user is in
[13-Jul-2010 10:23:08] <|rt|> to see if either of those changes replicate the issue
[13-Jul-2010 10:23:30] k
[13-Jul-2010 10:23:50] by the way, don't use ZenManager if you're trying to assign admin permissions
[13-Jul-2010 10:23:54] use Manager
[13-Jul-2010 10:24:52] grrr... i'm getting zenperfsnmp timeouts on devices, which is then causing zenpersnmp to issue a heartbeat failure event (just started last night)
[13-Jul-2010 10:24:58] <|rt|> yeah my main user has both zenmanager and manager permissions
[13-Jul-2010 10:25:09] just give Manager, it's redundant
[13-Jul-2010 10:25:16] I've set the zproperties to only try once, with a 1s timeout, but zenperfsnmp still waits 300s ?
[13-Jul-2010 10:25:25] <|rt|> k
[13-Jul-2010 10:25:40] <|rt|> I haven't gotten to read up on the roles yet
[13-Jul-2010 10:26:15] gwb2352: what zProperties value did you set exactly?
[13-Jul-2010 10:26:46] zSnmpTries, zSnmpTimeout
[13-Jul-2010 10:26:57] <|rt|> looks like it's just an issue with that account for some reason
[13-Jul-2010 10:27:01] gwb2352: the amount of time it waits is dependant on zCollectorClientTimeout I believe
[13-Jul-2010 10:27:13] <|rt|> I'll delete my original account and recreate it
[13-Jul-2010 10:27:14] gwb2352: the problem is, if you decrease that to 1 second, good luck modelling anything
[13-Jul-2010 10:27:19] it's something along the lines of "I want to figure this out, but don't want to get zenperfsnmp errors in the meantime"
[13-Jul-2010 10:27:37] gwb2352: what exactly are you trying to accomplish?
[13-Jul-2010 10:27:56] it's all local LAN, 5 of the 7 devices are still working, just 2 stopped magically last night
[13-Jul-2010 10:28:06] oh, you're getting timeouts
[13-Jul-2010 10:28:11] so I have 7 identical devices
[13-Jul-2010 10:28:14] 5 are still working
[13-Jul-2010 10:28:16] why would you DECREASE values if you're getting timeouts?
[13-Jul-2010 10:28:34] the timeout is because the snmp query never returns
[13-Jul-2010 10:28:58] which (my thinking) was causing the zenperfsnmp daemon to not complete in its cycle time, which (my thinking) would then cause the heartbeat email
[13-Jul-2010 10:29:01] then just toggle zSnmpMonitorIgnore to true for those devices until you fix SNMP on them
[13-Jul-2010 10:29:06] :)
[13-Jul-2010 10:29:06] don't mess around with timeout settings
[13-Jul-2010 10:29:17] i like that
[13-Jul-2010 10:29:31] <|rt|> rmatte: deleting and recreating the account worked....just in case anyone else runs across that issue
[13-Jul-2010 10:29:46] |rt|: cool, but bizarre lol
[13-Jul-2010 10:30:18] gwb2352: by the way, that won't help with the heartbeats
[13-Jul-2010 10:30:31] gwb2352: heartbeats are way too sensitive in 2.5.2, they fire off all the time
[13-Jul-2010 10:30:36] gwb2352: I just completely ignore them
[13-Jul-2010 10:30:47] gwb2352: as long as the Zenoss processes are all running you're good
[13-Jul-2010 10:31:29] add something to your alerting rule to filter them out
[13-Jul-2010 10:31:40] well, they did alert me to the fact that snmp stopped working randomly last night...
[13-Jul-2010 10:31:57] you sure it actually stopped working?
[13-Jul-2010 10:32:16] yeah - commandline zenperfsnmp -v 10 -d XXXXX no love either
[13-Jul-2010 10:32:27] did it error out, or what?
[13-Jul-2010 10:32:42] <|rt|> rmatte: it could have been the base location setting in that portlet
[13-Jul-2010 10:33:05] |rt|: that's exactly what I was thinking, which is why I asked you if the locations were multilevel
[13-Jul-2010 10:33:15] <|rt|> If I set that to one of the locations the others go away even if I zoom out for some reason
[13-Jul-2010 10:33:18] I end up with a "2010-07-13 10:32:54,549 DEBUG zen.zenperfsnmp: Total of 1 queued events"
[13-Jul-2010 10:33:22] and then an eventual timeout
[13-Jul-2010 10:33:30] |rt|: ah, I didn't know it would work even if it wasn't multilevel
[13-Jul-2010 10:33:34] <|rt|> I would have thought that setting would have just centered a location
[13-Jul-2010 10:33:51] |rt|: no, doesn't work that way
[13-Jul-2010 10:34:05] <|rt|> oh i see what you're saying so you'll see all the locations in that hierarchy
[13-Jul-2010 10:34:18] gwb2352: odd, well, disable snmp monitoring for those devices as I stated and then it should be fine
[13-Jul-2010 10:34:35] <|rt|> so if I set the base locations to just /Locations I see all the offices
[13-Jul-2010 10:34:56] ...and hopefully I get the time to figure out why it stopped working last night. 3AM is an odd time which makes me believe the device admin wasn't mucking with things.
[13-Jul-2010 10:35:02] gwb2352: when you're ready to switch them back scroll down to the bottom of zProperties and delete the local property so that they inherit settings from higher up
[13-Jul-2010 10:35:11] <|rt|> well I thought I tried changing that in my old account but since I've deleted and recreated it I can't go back to check to see if that was the problem or not
[13-Jul-2010 10:35:44] |rt|: that probably was it
[13-Jul-2010 10:37:30] <|rt|> yeah seem pretty likely
[13-Jul-2010 10:41:53] anybody else notice in 2.5.2 (if using LDAP authentication) you can't change /ZenUsers/uid information? (the "Enter your password to confirm changes" seems to require uid password, not Manager password)
[13-Jul-2010 10:41:58] hey guys
[13-Jul-2010 10:42:13] Parabola: fwiw, running your monitoring deployment over tunnels if the remote endpoints are a surefire way to lessen alerts and probably well worth doing
[13-Jul-2010 10:42:33] hey frozty
[13-Jul-2010 10:42:52] rmatte: the details around how a client IP address changes are left up to the ISP
[13-Jul-2010 10:43:04] (re your remark of an IP not changing in over a year)
[13-Jul-2010 10:43:32] frozty_sa: the ip changes so much
[13-Jul-2010 10:43:41] i have no clue why
[13-Jul-2010 10:43:46] Parabola: also, you gain the upside of being to monitor the point-to-point tunnel IP
[13-Jul-2010 10:43:58] we cannot create a tunnel from every single one of these sites
[13-Jul-2010 10:44:00] its not happening
[13-Jul-2010 10:44:07] its odd that zabbix does this just fine
[13-Jul-2010 10:44:07] and
[13-Jul-2010 10:44:10] yes you can, it's 3 hours worth of effort at most :P
[13-Jul-2010 10:44:12] zenoss wont
[13-Jul-2010 10:44:17] zabbix indexes hosts on hostnames
[13-Jul-2010 10:44:22] zenoss classifies them on IP
[13-Jul-2010 10:44:31] I've got extensive experience on both
[13-Jul-2010 10:44:46] while i understand what you're saying
[13-Jul-2010 10:44:47]
[13-Jul-2010 10:44:52] i'm kind of stuck finding a fix for this
[13-Jul-2010 10:44:56] and the tunnel isnt an option
[13-Jul-2010 10:45:07] i'm just the engine here, i'm not the guy pushing the pedal
[13-Jul-2010 10:45:08] I know about 35% of the zabbix codebase as well, and I can tell you that for anything even remotely complex, I'd never want to ever touch zabbix again
[13-Jul-2010 10:45:11] Parabola: okay well
[13-Jul-2010 10:45:20] yeah i like zenoss better myself
[13-Jul-2010 10:45:28] but these routers support no-ip out of the box
[13-Jul-2010 10:45:33] what you could do is have zenoss monitor a collection of private IPs
[13-Jul-2010 10:45:35] and thats what they're using and i cant change this setup
[13-Jul-2010 10:45:58] and then on the zenoss host itself have iptables DNAT rules that are dynamically updated with the endpoint IP addresses
[13-Jul-2010 10:46:15] write a small script to update the DNAT table every so often as the dyn IPs change
[13-Jul-2010 10:46:30] cunning :)
[13-Jul-2010 10:46:37] and then run it every however-long it's required to run
[13-Jul-2010 10:46:54] hmm
[13-Jul-2010 10:47:02] Simon4: I often joke to people that my job is to be a glue specialist ;)
[13-Jul-2010 10:47:16] :)
[13-Jul-2010 10:47:35] * Simon4 just works on the fact that he's generally one google search ahead of the customer
[13-Jul-2010 10:47:56] Parabola: how is the no-ip info being updated? directly from the routers?
[13-Jul-2010 10:48:08] Parabola: best of both worlds for you, in that you never have to hack zenoss and maintain any updates, and you gain everything that zenoss has while also not needing to plonk stuff down behind the customer CPEs
[13-Jul-2010 10:48:26] :)
[13-Jul-2010 10:48:30] rmatte: yup
[13-Jul-2010 10:49:27] Parabola: so no way to execute some sort of script when they do it?
[13-Jul-2010 10:49:43] * frozty_sa bets it's a dumb cheap&nasty consumer DSL router
[13-Jul-2010 10:49:51] probably
[13-Jul-2010 10:50:03] and my next bet will be that there's a mailserver in the head office as well ;)
[13-Jul-2010 10:50:34] I was going to suggest having the router call on Zenoss and update the IP when it updates the no-ip
[13-Jul-2010 10:50:42] but that's probably not possible with fisher price routers
[13-Jul-2010 10:50:50] Hmm
[13-Jul-2010 10:51:01] well, it could be if they run something like Tomato firmware
[13-Jul-2010 10:51:08] yeh
[13-Jul-2010 10:51:13] rmatte: some of them allow you to set a specific service where to update the IP
[13-Jul-2010 10:51:13] or even OpenWRT
[13-Jul-2010 10:51:43] I <3 Tomato firmware
[13-Jul-2010 10:51:47] since the DynDNS "protocol" is technically just either an HTTP POST or ordered GET
[13-Jul-2010 10:52:00] but not all give you that option
[13-Jul-2010 10:52:09] sooo much better than the linksys default - no more random lockups requiring a reboot of the router
[13-Jul-2010 10:52:34] back
[13-Jul-2010 10:52:36] they are sonic walls
[13-Jul-2010 10:52:48] OTOH, if it's an office, they ought to look into something like Vyatta or m00nwall or such at least
[13-Jul-2010 10:53:03] but theres a bunch of clients
[13-Jul-2010 10:53:07] frozty_sa: right, I'm aware, but I'm saying that it would need to trigger a script which determines what the new IP is, then access a URL such as http://whatever:8080/zport/dmd/Devices/Workstation/yourhostname/setManageIp?ip=192.168.0.1
[13-Jul-2010 10:53:10] so we have a million and 5 different devices out there
[13-Jul-2010 10:53:22] jmp242: pfffft. you overestimate the amount of bad management and practice out there :P
[13-Jul-2010 10:53:25] frozty_sa: and sets the new IP in Zenoss automatically
[13-Jul-2010 10:53:28] err, underestimate
[13-Jul-2010 10:53:29] that would eliminate the issue
[13-Jul-2010 10:53:35] rmatte: no sure, I know what you're saying
[13-Jul-2010 10:53:43] the downside is just that they're all shit :D
[13-Jul-2010 10:53:49] yeh
[13-Jul-2010 10:53:50] so i assume snmp cannot account for this?
[13-Jul-2010 10:54:06] Parabola: well, if you could send a trap to the zenoss server from the device, maybe
[13-Jul-2010 10:54:12] Parabola: not unless you can get the device to trap the new IP back to zenoss
[13-Jul-2010 10:54:24] Parabola: then have a transform that sets the new IP and moves the trap to history
[13-Jul-2010 10:54:25] ifOperStatus trap
[13-Jul-2010 10:54:32] and then what rmatte said
[13-Jul-2010 10:55:15] frozty_sa: not sure if ifOperStatus would do the trick though, and IP address can be re-assigned without the interface actually going down
[13-Jul-2010 10:55:16] * Simon4 wonders how you work out which device that trap came from
[13-Jul-2010 10:55:23] i guess the problem is there sint one fix for this
[13-Jul-2010 10:55:23] s/and/an
[13-Jul-2010 10:55:27] we have over 200 clients
[13-Jul-2010 10:55:28] since it's coming from a "new" ip..
[13-Jul-2010 10:55:33] some we put equpiment in , some they put it in
[13-Jul-2010 10:55:37] and what Simon4 said
[13-Jul-2010 10:56:12] Simon4: zenoss keeps those alerts
[13-Jul-2010 10:56:18] this isnt by any chance a feature in 3.0 is it
[13-Jul-2010 10:56:20] the only fix is to somehow have the sites themselves deliver the new IP to Zenoss and have it updated
[13-Jul-2010 10:56:23] to monitor host by dns isntead of ip
[13-Jul-2010 10:56:35] frozty_sa: sure, but what tags those alerts back to the original device, unless you can encode the hostname in the trap
[13-Jul-2010 10:56:36] Parabola: nope, not that I've seen
[13-Jul-2010 10:56:46] Parabola: 3.0 is 2.5.2 with a new UI skin basically
[13-Jul-2010 10:56:49] Simon4: orite, yeah
[13-Jul-2010 10:56:59] Parabola: nothing has really changed on the back end between the versions
[13-Jul-2010 10:57:13] Simon4: some of what we've got on our DSL infrastructure would help Parabola, but we're probably in a different country entirely :p
[13-Jul-2010 10:57:15] other than new zope and new python
[13-Jul-2010 10:57:26] heheh
[13-Jul-2010 10:57:32] yeah these are ATT DSL lines
[13-Jul-2010 10:57:36] leases dont last more than 48 hours
[13-Jul-2010 10:57:42] that's so lame
[13-Jul-2010 10:57:50] Umm, doesn't business connections come with static IP/
[13-Jul-2010 10:57:52] ?
[13-Jul-2010 10:57:57] did i say business?
[13-Jul-2010 10:58:02] he's probably working off residential
[13-Jul-2010 10:58:02] they are residental dsl lines
[13-Jul-2010 10:58:06] jmp242: not if management decides to go cheap on it ;)
[13-Jul-2010 10:58:17] you guys act like i sat down with a sheet of paper and planned all this shit
[13-Jul-2010 10:58:17] I had ATT residential DSL for years, static IP was free
[13-Jul-2010 10:58:17] we dynamically autobind people's username in a specific fashion to a DNS record, so people can check up their own IPs without any special stuff
[13-Jul-2010 10:58:19] You want to monitor remote residential DSL equipment?
[13-Jul-2010 10:58:20] I used to work for them too
[13-Jul-2010 10:58:20] Simon4: ^
[13-Jul-2010 10:58:24] its consulting, you dont always have a choice
[13-Jul-2010 10:58:29] I had business internet in my apartment for $90/month, are you guys really so cheap as to not be able to afford that sort of thing?
[13-Jul-2010 10:58:39] jmp242: jesus chrsit read up
[13-Jul-2010 10:58:49] rmatte: you arent fucking listening
[13-Jul-2010 10:58:53] i do not work for this company
[13-Jul-2010 10:58:56] they refuse to change it
[13-Jul-2010 10:59:01] I see
[13-Jul-2010 10:59:03] i'm a god damn consultant
[13-Jul-2010 10:59:05] I missed that part
[13-Jul-2010 10:59:06] yeah
[13-Jul-2010 10:59:11] we service about 200 clients
[13-Jul-2010 10:59:15] some we put service in for
[13-Jul-2010 10:59:17] others already had it
[13-Jul-2010 10:59:25] theres not much i can do in this case man
[13-Jul-2010 10:59:25] mmm, I have some doubts that Zenoss was really designed for this sort of network layout
[13-Jul-2010 10:59:38] yeah i'm getting yelled at because i picked zenoss
[13-Jul-2010 10:59:38] I've always had the feeling (with Core anyway)
[13-Jul-2010 10:59:44] yeh, I misunderstood, I thought these were 200 sites belonging to the same organization
[13-Jul-2010 10:59:45] my bad
[13-Jul-2010 10:59:49] its okay :)
[13-Jul-2010 10:59:54] sorry im just frustrated
[13-Jul-2010 11:00:02] i picked zenoss to move to from zabbix due to the agent bullshit
[13-Jul-2010 11:00:04] understandable
[13-Jul-2010 11:00:07] then he drops this in my lap
[13-Jul-2010 11:00:09] jmp242: for instance here in SA, you basically down a VPN-dialing router if you want a static IP. and then the provider who supplies the IP rips you about >=10x times the cost of other services, for a more contended service
[13-Jul-2010 11:00:11] that the idea was there was one internal network and you could talk to the devices directly
[13-Jul-2010 11:00:16] "oh yea monitor these DSL lines too"
[13-Jul-2010 11:00:25] Parabola: I think going with that script option is probably your best bet
[13-Jul-2010 11:00:32] and it'll stay stable as well
[13-Jul-2010 11:00:39] it's not like cron will go and fuck out any time soon
[13-Jul-2010 11:00:45] yeah
[13-Jul-2010 11:00:58] script writes out a ferm config file, ferm reloads it
[13-Jul-2010 11:01:07] frozty_sa: if you use a VPN dialing router
[13-Jul-2010 11:01:10] sounds above my head on a product i know little about
[13-Jul-2010 11:01:10] I wouldn't alert on ping events until the event count hits 3 though, that way you avoid false positives
[13-Jul-2010 11:01:18] since cron can only run once per minute
[13-Jul-2010 11:01:21] can't you use servers / endpoints in another country to save money?
[13-Jul-2010 11:01:22] drop monit in for good measure to make sure thing stay up
[13-Jul-2010 11:01:30] I'm thinking stuff like SwissVPN...
[13-Jul-2010 11:01:35] rmatte: doesnt matter, the shit changes every 2 days at least
[13-Jul-2010 11:01:40] jmp242: no, you essentially get a blackbox from them
[13-Jul-2010 11:01:42] at 2am i get woken up by shit loads of emails
[13-Jul-2010 11:01:48] Zenoss didn't care for our DMVPN approach at first, i had to keep deleting the networks and let the remodeling rebuild them
[13-Jul-2010 11:02:18] what would it take for me to get this feature added
[13-Jul-2010 11:02:22] Parabola: right, I'm suggesting a script that constantly resolves and watches the hostnames every minute and automatically updates the IP in Zenoss if it changes
[13-Jul-2010 11:02:23] Money
[13-Jul-2010 11:02:27] how mcuh
[13-Jul-2010 11:02:35] jmp242: in terms of numbers, DSL here is around $10 per 15GB or whatever. if you want that same package on a static IP, you're likely to pay $150
[13-Jul-2010 11:02:45] but Zenoss pings every minute, so to avoid false positives you'd only want to alert after 3 minutes of missed pings
[13-Jul-2010 11:02:47] rmatte: which sounds great, however i dont have the skillset to write such a script
[13-Jul-2010 11:02:49] that's really the best option
[13-Jul-2010 11:02:58] Parabola: It would take you buying the paid version I'm sure
[13-Jul-2010 11:02:58] Parabola: let me throw something together
[13-Jul-2010 11:03:13] theacolyte: the paid version DOES support devices by DNS not just ip?
[13-Jul-2010 11:03:22] I missed the whole thing
[13-Jul-2010 11:03:29] Parabola: nah it doesn't
[13-Jul-2010 11:03:29] Parabola: Are you asking if you can add a device by FQDN?
[13-Jul-2010 11:03:38] theacolyte: he wants to monitor by host
[13-Jul-2010 11:03:39] no, he wants to monitor it by FQDN
[13-Jul-2010 11:03:43] not add it
[13-Jul-2010 11:03:44] theacolyte: not have hardcoded IPs
[13-Jul-2010 11:03:54] theacolyte: nutshell: monitor by host, ATT home DSL lines in businesses, leases evey 48 hours
[13-Jul-2010 11:03:55] Because i add all my devices by their name and not by IP
[13-Jul-2010 11:03:56] theacolyte: since his sites have dynamically assigned IPs that change every 48 hours
[13-Jul-2010 11:03:59] jmp242: and it literally becomes a case of you can't even easily jump providers without plugging out the DSL link itself from the blackbox router they give you, and putting it into another one
[13-Jul-2010 11:04:00] ahhhhhhhhh
[13-Jul-2010 11:04:13] theacolyte: but after that 2 days i have to login to zenoss, clear out the IP and let it resolve it again
[13-Jul-2010 11:04:13] twm1010 already said dmvpn :(
[13-Jul-2010 11:04:19] theres 75 stores for this company
[13-Jul-2010 11:04:23] and its alot of work to do..
[13-Jul-2010 11:04:38] dmvpn did work, but during rollout it threw a big fit on us
[13-Jul-2010 11:05:07] and you've have to assume you've got similar or same equipment at each site
[13-Jul-2010 11:06:12] Do you have tunnels to the sites? or no?
[13-Jul-2010 11:06:44] You need something agent based that can phone home and update it
[13-Jul-2010 11:06:49] Zenoss may not be your product honestly
[13-Jul-2010 11:07:43] yeah so im finding out
[13-Jul-2010 11:07:46] kinda dissapointing, i like it
[13-Jul-2010 11:07:54] i've spent alot of time figuring stuff out with it and bothering you guys
[13-Jul-2010 11:08:10] That's why it's always a good idea to set requirements before you start evaluating
[13-Jul-2010 11:08:19] (That doesn't help you now obviously)
[13-Jul-2010 11:08:19] Well, its a model problem right, its meant to do something a certain way, ver yhard to make a product fit every mold
[13-Jul-2010 11:08:19] hehe
[13-Jul-2010 11:08:29] i agree
[13-Jul-2010 11:08:37] honestly i had no idea i was going to be asked to monitor this
[13-Jul-2010 11:08:42] But now you know!
[13-Jul-2010 11:08:42] it was for linux/windows servers
[13-Jul-2010 11:08:46] well... does it all have to be one system?
[13-Jul-2010 11:08:51] monitor the boxes, drive space, and ospreocess / servers
[13-Jul-2010 11:08:56] services*
[13-Jul-2010 11:08:58] what do you mean
[13-Jul-2010 11:09:04] i'm not running zabbix and this thats retarded
[13-Jul-2010 11:09:29] I use Zenoss for most of my monitoring and alerting, I use our leftover SolarWinds license for explicit monitoring of interfaces, since it has variable polling and a nice reporting engine
[13-Jul-2010 11:10:07] I use a combination of HP SIM and Zenoss
[13-Jul-2010 11:10:13] until Zenoss breaks from RRD to a database option, I won't get that nice variable polling without using multiple collectors, and even then, it's still per device
[13-Jul-2010 11:10:18] There's a fair amount of duplication there, but I'm fine with that
[13-Jul-2010 11:10:54] rmatte: are you around?
[13-Jul-2010 11:11:04] I've yet to use SIM, still meaning to attack the ILO2 and Insight agents configs for best practice for Zenoss
[13-Jul-2010 11:11:12] do you have anything to that effect already ?
[13-Jul-2010 11:11:33] You mean integrating the HP agents into Zenoss with zenpacks?
[13-Jul-2010 11:11:36] nah
[13-Jul-2010 11:11:54] the whole shbang really
[13-Jul-2010 11:12:05] Yeah, I haven't
[13-Jul-2010 11:12:14] trap forwarding from ILO to zenoss, etc, what is the way to do it right?
[13-Jul-2010 11:12:28] Ah, no idea, that's why I run SIM
[13-Jul-2010 11:12:33] when in rome...
[13-Jul-2010 11:12:51] kill romans
[13-Jul-2010 11:12:55] yeah, we're starting to go back to HP, but I've got Dell, IBM, HP, Mac, etc.
[13-Jul-2010 11:13:12] Scary enough, we even have some Compaq servers still running
[13-Jul-2010 11:13:29] so what product should i be using
[13-Jul-2010 11:13:37] does solar winds ip monitor support what i need?
[13-Jul-2010 11:13:42] That's a really bad thing to do
[13-Jul-2010 11:13:43] i've used it before but its been a couple years
[13-Jul-2010 11:13:44] Do your own research
[13-Jul-2010 11:13:48] None of us will know your requirements
[13-Jul-2010 11:13:59] Sit down, write out everything you have learned through the process and create a list of what you need
[13-Jul-2010 11:14:01] my requirements are aparently zenoss mixed with some zabbix
[13-Jul-2010 11:14:02] Then shop it around
[13-Jul-2010 11:14:06] That doesn't work
[13-Jul-2010 11:14:08] lol
[13-Jul-2010 11:14:10] shhh
[13-Jul-2010 11:14:10] Don't be lazy with it -
[13-Jul-2010 11:14:12] hehe
[13-Jul-2010 11:14:37] be thankful that you didn't waste anything but time/money and didn't buy a product that doesnt' do what you need it to do
[13-Jul-2010 11:15:01] Parabola: I don't think SW is the one for you either
[13-Jul-2010 11:15:18] alot of time
[13-Jul-2010 11:15:19] was just making the point that you may need more than one product to get done what you're after
[13-Jul-2010 11:15:19] If you do engage solarwinds, for the love of god don't give them your phone number
[13-Jul-2010 11:15:23] but i am billing for it
[13-Jul-2010 11:15:26] lol good to know
[13-Jul-2010 11:15:28] they like to call?
[13-Jul-2010 11:15:32] You can... say that
[13-Jul-2010 11:15:40] I wouldn't recommend SW
[13-Jul-2010 11:15:48] It's pretty hackish
[13-Jul-2010 11:15:52] it's very pretty but the backend and management interface blow chunks
[13-Jul-2010 11:15:55] Their network monitoring is really good
[13-Jul-2010 11:16:01] wtf else is there
[13-Jul-2010 11:16:04] But the application side, not so much
[13-Jul-2010 11:16:10] It depeneds on your #*()@*#)* requirements
[13-Jul-2010 11:16:23] I could just start spouting bullshit for you, but I don't think it would help you
[13-Jul-2010 11:16:34] yeh, if you can lay out a point form of everything you need it to be able to do
[13-Jul-2010 11:17:00] im about to stab you all
[13-Jul-2010 11:17:03] :P
[13-Jul-2010 11:17:16] * rmatte sighs
[13-Jul-2010 11:17:19] lol
[13-Jul-2010 11:17:20] The best we can do is let you know what's out there
[13-Jul-2010 11:17:22] adn there is a lot
[13-Jul-2010 11:17:27] i understand
[13-Jul-2010 11:17:31] i'm sticking with zenoss if i can
[13-Jul-2010 11:17:35] the other monitoring is quite nice
[13-Jul-2010 11:17:43] well, try the script then that rmatte suggested
[13-Jul-2010 11:17:56] well, I'm throwing something together right now for him...
[13-Jul-2010 11:17:58] he was nice enough to whip a quick one up for me he said
[13-Jul-2010 11:18:08] rmatte is too nice
[13-Jul-2010 11:18:10] I'm actually kinda surprised Zenoss can't monitor by hostname
[13-Jul-2010 11:18:11] I'm not done whipping yet lol
[13-Jul-2010 11:18:12] he knows this
[13-Jul-2010 11:18:18] this is a badass product
[13-Jul-2010 11:18:19] twm1010: so am I
[13-Jul-2010 11:18:21] i want to keep it
[13-Jul-2010 11:18:29] I guess I never noticed that behavior before
[13-Jul-2010 11:18:30] twm1010: there are quite a few small and stupid shortcomings still present
[13-Jul-2010 11:18:31] wonder if i paid the devs, if they would add it
[13-Jul-2010 11:18:36] I add by domain name, but it uses the IP anyway?
[13-Jul-2010 11:18:43] When you hit the Reset IP context drop down, there is text there that says you can put in a hostname
[13-Jul-2010 11:18:49] Parabola: they do do PS work for core, so you could ask
[13-Jul-2010 11:18:53] Parabola: I believe professional services tasks are available
[13-Jul-2010 11:18:54] Parabola: it would be expensive though
[13-Jul-2010 11:18:55] heh, snap
[13-Jul-2010 11:19:01] whats expensive?
[13-Jul-2010 11:19:09] more than a couple grand?
[13-Jul-2010 11:19:15] They charge like over $100/hour I believe
[13-Jul-2010 11:19:18] to be honest if i can get my employer to pay for this
[13-Jul-2010 11:19:26] would you guys like to have that feature?
[13-Jul-2010 11:19:31] contact them and get a quote
[13-Jul-2010 11:19:34] I wouldn't need it
[13-Jul-2010 11:19:34] i'd pay to get it added to Core
[13-Jul-2010 11:19:39] stfu theacolyte yes you do
[13-Jul-2010 11:19:42] Parabola: absolutely
[13-Jul-2010 11:19:46] "Are you sure you would like to reset the IP? Leave blank to use DNS." does that just lookup the IP via DNS and then store it?
[13-Jul-2010 11:19:55] i mean assuming they let me release it
[13-Jul-2010 11:19:55] I don't monitor... dynamic stuff
[13-Jul-2010 11:19:57] I wouldn't need it now, but it would most likely come in handy eventually
[13-Jul-2010 11:20:02] or they include the stuff i bought IN the next update to core
[13-Jul-2010 11:20:04] i would pay for it
[13-Jul-2010 11:20:09] if they say only i can have it, i wouldnt buy it
[13-Jul-2010 11:20:14] I remember some forum thread about this at one point?
[13-Jul-2010 11:20:32] I wonder if you could do this in a round-about sort of way through a performance template bound to a host
[13-Jul-2010 11:20:40] Parabola: nah, it's your decision if you want it rolled in or not as far as I understand.
[13-Jul-2010 11:20:41] using dynamic DNS
[13-Jul-2010 11:20:56] rmatte: then im going to contact them
[13-Jul-2010 11:21:11] since i cant help out with questions in here very often, the least i can do is put my wallet in there and help
[13-Jul-2010 11:21:19] im sure alot of people would like that feature
[13-Jul-2010 11:21:36] it would be handy indeed
[13-Jul-2010 11:22:13] So if I get this right, you're going to monitor devices on the other side of a router with a dynamic address?
[13-Jul-2010 11:22:31] he's monitoring sites by their internet address
[13-Jul-2010 11:22:34] no tunnels or anything
[13-Jul-2010 11:22:42] just a whole bunch of dsl sites
[13-Jul-2010 11:22:51] well, if those sites answer on port 80
[13-Jul-2010 11:22:51] with dynamic hosts configured for them
[13-Jul-2010 11:22:58] No
[13-Jul-2010 11:23:00] Not good enough!
[13-Jul-2010 11:23:01] yeh, he could use the httpmon zenpack
[13-Jul-2010 11:23:10] but that wouldn't give him full monitoring
[13-Jul-2010 11:23:21] true
[13-Jul-2010 11:23:25] Parabola: are you just monitoring for ping status?
[13-Jul-2010 11:23:32] Parabola: or are you doing full snmp at the sites?
[13-Jul-2010 11:23:43] if he wants more, he needs a VPN tunnel most likely
[13-Jul-2010 11:23:54] yeh, but vpn tunnels are out of the question apparently
[13-Jul-2010 11:24:01] just ping is what i "need"
[13-Jul-2010 11:24:07] right now its snmp
[13-Jul-2010 11:24:12] the one im testing with is a sonicwall
[13-Jul-2010 11:24:13] Parabola: do those routers respond on port 80 at each site?
[13-Jul-2010 11:24:18] some do, why
[13-Jul-2010 11:24:24] not the routers
[13-Jul-2010 11:24:27] but webservers behind them
[13-Jul-2010 11:24:36] because you could use the httpmonitor zenpack, which is capable of monitoring websites by just their hostnames
[13-Jul-2010 11:24:58] what you wouldnt' have is CPU/Memory/Disk, etc.
[13-Jul-2010 11:25:01] <|rt|> some opensource projects that allow for paid features charge a lower rate if the changes can be released to the community
[13-Jul-2010 11:25:03] yeh
[13-Jul-2010 11:25:08] <|rt|> I know that pfsense does that
[13-Jul-2010 11:25:27] |rt|: I don't think Zenoss does that, they have a fixed rate as far as I'm aware
[13-Jul-2010 11:25:36] but wouldn't hurt to ask
[13-Jul-2010 11:26:04] <|rt|> a bummer....i think it's a good incentive for companies to help give back
[13-Jul-2010 11:26:14] agreed
[13-Jul-2010 11:27:32] rmatte: OTRS has 1600 sysconfig options (my brain hurts)
[13-Jul-2010 11:27:49] twm1010: haha, did you just start using it?
[13-Jul-2010 11:27:50] but not a difficult install, seems ridiculously flexible
[13-Jul-2010 11:27:58] it's very very flexible
[13-Jul-2010 11:28:01] that's why we use it
[13-Jul-2010 11:28:07] though we obviously have custom coding done for it
[13-Jul-2010 11:28:37] Yeah, I like it personally, but I think our managers will want something more polished and pretty (they never seem to care about whether its functional)
[13-Jul-2010 11:28:45] Installing additional modules can be done right from the UI too, it's a piece of cake
[13-Jul-2010 11:28:50] look in to the ITSM modules
[13-Jul-2010 11:28:53] you'll love them
[13-Jul-2010 11:29:07] otrs 3.0 will have a completely new interface
[13-Jul-2010 11:29:10] yeah I installed damn near everything to play around with it, it's really slick, some of it not intuitive
[13-Jul-2010 11:29:11] much more polished and such
[13-Jul-2010 11:29:45] (not sure what it is about the 3.0 version number coinciding with interface redesign)
[13-Jul-2010 11:29:47] :)
[13-Jul-2010 11:30:11] alright im back
[13-Jul-2010 11:30:17] i emailed sales for a quote on the feature
[13-Jul-2010 11:30:25] twm1010: have you looked at any others yet?
[13-Jul-2010 11:30:28] not really
[13-Jul-2010 11:30:29] told them in the email i wanted the feature included in the CORE release
[13-Jul-2010 11:30:51] Parabola: cool, it'll be interesting to see if they'll agree or not
[13-Jul-2010 11:31:00] they better
[13-Jul-2010 11:31:04] or i start a'stabbin
[13-Jul-2010 11:31:08] lol
[13-Jul-2010 11:32:03] twm1010: if you want a more polished look: http://blog.bestpractical.com/2008/07/today-were-rele.html#screenshots
[13-Jul-2010 11:32:20] twm1010: RT isn't as powerful as otrs from what I've seen, but it certainly is pretty
[13-Jul-2010 11:32:30] rmatte: does modeling need to happen on my main collector or can a remote collector take care of it?
[13-Jul-2010 11:32:51] ptom: modelling takes place under whatever collector a device is assigned to
[13-Jul-2010 11:33:41] does your dist collector tool update zenmodeler.conf?
[13-Jul-2010 11:34:14] Parabola: lol
[13-Jul-2010 11:34:33] i'll check that out too
[13-Jul-2010 11:34:35] sup
[13-Jul-2010 11:34:46] mray: hop to it
[13-Jul-2010 11:34:48] hehe
[13-Jul-2010 11:37:01] hey guys , mainly rmatte
[13-Jul-2010 11:37:10] i have to head out to a clienst office to drop somthing off
[13-Jul-2010 11:37:23] rmatte: if you are still working on that script, can you message me the link? i wont see it in here
[13-Jul-2010 11:37:26] thanks man
[13-Jul-2010 11:37:39] i'll let you guys know when i hear back from zenoss on feature
[13-Jul-2010 11:38:18] anyone familiar with lldp? i can't figure out how to get the local endpoint (which local port of the chassis) corresponds to an entry in lldpRemEntry.
[13-Jul-2010 11:38:53] What are you using it for, though?
[13-Jul-2010 11:39:00] (in the context of zenoss)
[13-Jul-2010 11:39:31] theacolyte: currently doing layer2 discovery (outside of zenoss) so if i know that i can figure it out i could bring it into zenoss
[13-Jul-2010 11:40:04] Man, if you figure that out, let me know
[13-Jul-2010 11:40:09] (Although I'm doing DDP)
[13-Jul-2010 11:40:11] err CDP
[13-Jul-2010 11:40:31] theacolyte: ciscos should also support lldp
[13-Jul-2010 11:40:48] It does
[13-Jul-2010 11:42:16] Parabola: I have to tackle some other stuff today so I won't be able to finish the script... you might not need it though depending on what the Zenoss PS guys say.
[13-Jul-2010 11:42:32] If you do still need it, get in touch with me later on during the week
[13-Jul-2010 11:42:37] rmatte: i understand man, i'm here every day anyway lol
[13-Jul-2010 11:42:42] cool
[13-Jul-2010 11:42:55] thanks again :)
[13-Jul-2010 11:42:58] np
[13-Jul-2010 11:43:01] i owe you lunch
[13-Jul-2010 11:43:06] come to cols. ohio
[13-Jul-2010 11:43:07] well, not yet
[13-Jul-2010 11:43:08] lol
[13-Jul-2010 11:43:11] lol true
[13-Jul-2010 11:43:26] ohio is far for me to go for lunch :)
[13-Jul-2010 11:46:32] Parabola: I'll be there for Ohio LinuxFest
[13-Jul-2010 11:49:13] sweet
[13-Jul-2010 11:49:16] we should meet up
[13-Jul-2010 11:49:20] and make fun of rmatte
[13-Jul-2010 11:49:21] hmmmm, I don't get it, in the 3.0 beta there's a menu option to add an IpInterface, which only prompts you for an ID, then adds the interface with that id... what possible purpose could that serve? lol
[13-Jul-2010 11:49:24] we'll have a day of free training too
[13-Jul-2010 11:49:40] same for other components
[13-Jul-2010 11:50:42] hmmm, oh well
[13-Jul-2010 11:52:16] so apparently management wants us to move to WMI monitoring for every windows server that we are currently monitoring
[13-Jul-2010 11:52:18] FML
[13-Jul-2010 11:52:48] guess I'm going to get really familliar with Egor's ZenPacks
[13-Jul-2010 11:54:42] * rmatte despises WMI
[13-Jul-2010 11:55:12] * forsberg 2
[13-Jul-2010 11:55:18] rmatte: why?
[13-Jul-2010 11:55:26] not why you despise, but why switching?
[13-Jul-2010 11:55:32] mray: bulky, unstable, pain in the arse permissions
[13-Jul-2010 11:55:36] ah
[13-Jul-2010 11:55:41] because management said so
[13-Jul-2010 11:55:44] lol
[13-Jul-2010 11:56:02] they want to be able to do detailed monitoring of exchange or whatever, I'm going to recommend that we ONLY use WMI on servers that we need to
[13-Jul-2010 11:56:07] and leave everything else alone
[13-Jul-2010 11:56:35] but they probably won't listen
[13-Jul-2010 11:58:38] rmatte: say that I'm plan on discontinuing support for WMIDataSource ;)
[13-Jul-2010 11:58:55] bigegor: the thought crossed my mind
[13-Jul-2010 11:59:14] bigegor: but I doubt you actually are lol
[13-Jul-2010 11:59:30] yes, i am
[13-Jul-2010 11:59:45] bigegor: seriously? was this the last release?
[13-Jul-2010 11:59:46] what?
[13-Jul-2010 11:59:54] i mean lol
[13-Jul-2010 12:00:05] lol
[13-Jul-2010 12:00:09] haha
[13-Jul-2010 12:00:11] oh no3s
[13-Jul-2010 12:00:16] rmatte: http://twitter.com/mattray/status/18451379519
[13-Jul-2010 12:00:38] mray: nice
[13-Jul-2010 12:00:38] :)
[13-Jul-2010 12:01:03] bigegor: nononononononononononononononono
[13-Jul-2010 12:01:07] bigegor: so seriously, you are or you aren't?
[13-Jul-2010 12:02:01] don't worry, i will continue.
[13-Jul-2010 12:03:50] BTW, we'll be starting a new ZenPack contest with the 3.0 release
[13-Jul-2010 12:04:08] bigegor will probably win again :p
[13-Jul-2010 12:04:21] but I'm open for suggestions for prizes
[13-Jul-2010 12:04:42] zebra steaks
[13-Jul-2010 12:05:05] how about a core bugfix contest instead? no one ever works on the infrastructure
[13-Jul-2010 12:06:09] rmatte's got patches that you guys haven't taken yet
[13-Jul-2010 12:06:36] +1 for bugfix contest.
[13-Jul-2010 12:06:58] OK, we'll take patches as entries into the contest
[13-Jul-2010 12:07:04] I prefer a "All prizes go to theacolyte" contest
[13-Jul-2010 12:07:16] I'm shooting for 8 prizes
[13-Jul-2010 12:07:29] Excellent
[13-Jul-2010 12:07:33] I'll PM you my address
[13-Jul-2010 12:07:38] probably 2x$1000 for the grand prizes and 6x$400 or so for runner-ups
[13-Jul-2010 12:07:48] s/address/paypal/
[13-Jul-2010 12:08:07] so I'm open for prize ideas (besides cash)
[13-Jul-2010 12:08:08] i suggest a big golden B for the winner, we already know who will get it ;)
[13-Jul-2010 12:08:13] lol
[13-Jul-2010 12:08:18] hehe
[13-Jul-2010 12:09:10] last year we sent bigegor a QNAP something or other (a 419p?)
[13-Jul-2010 12:12:15] zenoss indexes interfaces by their description?
[13-Jul-2010 12:28:00] hey guys - I've got a zenoss install running in a Jumpbox instance
[13-Jul-2010 12:28:04] got a bit of a strange one
[13-Jul-2010 12:28:29] On a server I am monitoring zenoss is showing snmpd as down, yet I am able to poll the system with snmpwalk. As well zenoss is polling the server every five minutes for stats, but the check every sixty seconds for status is not coming through, any guesses?
[13-Jul-2010 12:28:35] when trying to install a zenpack (zenjmx) through the GUI - it errors out on me, saying, zenpack /bin/sh command not found
[13-Jul-2010 12:28:44] although logged in as zenoss in the shell I can run the command
[13-Jul-2010 12:28:53] the zenoss process is running as the zenoss user
[13-Jul-2010 12:29:05] so...... I tried just running the install for the zenpack from the shell
[13-Jul-2010 12:29:15] willwh: did you unzip it?
[13-Jul-2010 12:29:17] and ofc it errors out saying - easy_install not found
[13-Jul-2010 12:29:20] the egg? yeah
[13-Jul-2010 12:29:31] that's all i got.
[13-Jul-2010 12:29:32] but - I can run easy_install from the shell as zenoss user again
[13-Jul-2010 12:29:36] so doubly odd
[13-Jul-2010 12:30:05] if I just easy_install myzenpack.Egg - it seems to install fine
[13-Jul-2010 12:30:22] but ofc on restarting Zenoss, Daemon: zenjmx /etc/init.d/zenoss: line 105: /usr/local/zenoss/bin/zenjmx: No such file or directory
[13-Jul-2010 12:34:07] I'm not quite sure what's required in /usr/local/zenoss/bin/zenjmx - although I'm guessing my easy_install dropped it somewhere else
[13-Jul-2010 12:34:36] what were you installing with easy_install?
[13-Jul-2010 12:34:48] well - the egg - or trying to
[13-Jul-2010 12:35:37] http://pastebin.com/Cwv10snU
[13-Jul-2010 12:35:55] can you try zenpack --install zenpack.egg ?
[13-Jul-2010 12:35:59] if it's a zenpack, use "zenpack —install ZenPacks.zenoss.ZenJMX-3.3.2-py2.4.egg"
[13-Jul-2010 12:37:43] that is where things get funky
[13-Jul-2010 12:37:44] http://pastebin.com/K47Msk1k
[13-Jul-2010 12:38:00] I can run easy_install as the zenoss user at the shell - which is what I am running zenpack --install as btw
[13-Jul-2010 12:38:08] and it falls over like that
[13-Jul-2010 12:38:12] seriously odd
[13-Jul-2010 12:38:38] that is odd, can you try to download that zenpack again?
[13-Jul-2010 12:38:54] maybe it got corrupted somehow
[13-Jul-2010 12:38:57] I sure can
[13-Jul-2010 12:39:04] I'm going in to a meeting in a sec
[13-Jul-2010 12:39:05] back in 20
[13-Jul-2010 12:53:32] ptom: tried downloading and unzipping again
[13-Jul-2010 12:53:39] still throwing the same easy_install error
[13-Jul-2010 12:58:02] I've never used easy_install to install a zenpack, I have alwayes either used the ui or zenpack --install. Wondering if easy_install messed something up.
[13-Jul-2010 12:58:44] Yea, what is easy_install?
[13-Jul-2010 12:58:49] never heard of it
[13-Jul-2010 12:59:02] it's a python thing
[13-Jul-2010 12:59:07] it's run by zenpack I assume
[13-Jul-2010 12:59:41] as zenpack --install ***.egg -> errors to easy_install not found
[13-Jul-2010 12:59:52] which is just bizarre
[13-Jul-2010 13:00:53] can you go to the UI and see if that zenpack is listed in there?
[13-Jul-2010 13:01:08] if it is, I would delete it, restart zope and try again
[13-Jul-2010 13:01:16] it's not listed
[13-Jul-2010 13:01:42] And using the UI fails the same way?
[13-Jul-2010 13:02:00] nice, more good news, no I'm expected to somehow monitor call manager and unity in Zenoss Core
[13-Jul-2010 13:02:04] * rmatte slams head against desk
[13-Jul-2010 13:02:22] I guess that means that the POS product that they bought to do that monitoring is being scrapped
[13-Jul-2010 13:02:30] fun fun
[13-Jul-2010 13:03:01] rmatte: did you pick the wrong day to stop sniffing glue ?
[13-Jul-2010 13:03:55] jmp242: yup
[13-Jul-2010 13:04:22] ptom: apparently
[13-Jul-2010 13:04:25] willwh: I would restart zope and try to install it again.
[13-Jul-2010 13:04:27] hmm, does any zenpack install
[13-Jul-2010 13:04:33] or is it just this one failing
[13-Jul-2010 13:04:41] I haven't tried another
[13-Jul-2010 13:06:27] I will in a sec here
[13-Jul-2010 13:08:16] anyone have any problems with the nagios checkping zenpack? The command runs, the graph shows up, but I am getting the NaN entries.
[13-Jul-2010 13:08:21] any suggestions for something with no dependencies?
[13-Jul-2010 13:09:15] ptom: did you just add it?
[13-Jul-2010 13:09:29] ptom: it takes 3 polling cycles to start graphing (15 minutes)
[13-Jul-2010 13:09:35] so if you haven't waited that long yet, patience
[13-Jul-2010 13:09:35] lol
[13-Jul-2010 13:09:47] no, I have been waiting... days
[13-Jul-2010 13:11:17] ah, strange
[13-Jul-2010 13:11:19] I did change one part of the script though. ${dev/devname} to ${dev/manageIp}
[13-Jul-2010 13:11:29] that shouldn't make a difference
[13-Jul-2010 13:11:35] in fact that's how it should be
[13-Jul-2010 13:11:36] i didn't think so
[13-Jul-2010 13:11:39] i agree
[13-Jul-2010 13:11:41] doing it with devname is just stupid
[13-Jul-2010 13:11:42] lol
[13-Jul-2010 13:11:58] :) that's what I thought too
[13-Jul-2010 13:13:29] zencommand should look to see if a device is up before running.
[13-Jul-2010 13:14:06] it does
[13-Jul-2010 13:14:15] but only if a ping event came in
[13-Jul-2010 13:14:28] if it manages to run before Zenoss sees the device as down, then you get a command failed event
[13-Jul-2010 13:15:13] huh
[13-Jul-2010 13:15:39] if zenping hasn't seen the device as being down yet, the other daemons will hit it if it's during a polling cycle
[13-Jul-2010 13:16:00] you're right. my bad
[13-Jul-2010 13:16:07] if there's a ping down event for it, then all the other daemons ignore it
[13-Jul-2010 13:16:07] ;)
[13-Jul-2010 13:37:24] hi: I seem to be having this problem: thread/11690?start=15&tstart=0 -- apache processes are running but zenoss is saying that httpd is down when it's really up. The advise in that thread is to "change the regexp to "^.*\/httpd$". This regexp will not get the defunct daemon and everything will be ok." Can anyone give me a breadcrumb on how to do that?
[13-Jul-2010 13:38:01] I see under the OS Tab -> OS Processes there is /usr/sbin/httpd and httpd -- which is probably the problem
[13-Jul-2010 13:40:40] has anyone succeeded in monitoring win 7 home premium (<<
[13-Jul-2010 13:41:43] the beancounters tried to save some money and we ended up with a batch of these
[13-Jul-2010 13:42:10] robo: click services on the left
[13-Jul-2010 13:42:13] then ip services
[13-Jul-2010 13:42:14] find httpd
[13-Jul-2010 13:42:16] change the regex
[13-Jul-2010 13:42:43] sorry, missed a step, it'll be under privileged
[13-Jul-2010 13:43:05] search for http
[13-Jul-2010 13:43:35] or is it an actually process that it's complaining about?
[13-Jul-2010 13:44:46] ty rmatte, found it. I'm not exactly sure that's the process it's complaining about
[13-Jul-2010 13:45:06] well... it's complaining about httpd
[13-Jul-2010 13:45:21] "Process not running: httpd"
[13-Jul-2010 13:45:50] oi, I just noticed my install of Zenoss has no device classes (Besides the 4 I made); add device -> Device Class Path -> only the 4 I've made... I also notice there are no /Server/Linux templates?
[13-Jul-2010 13:45:56] Why Zenoss =(
[13-Jul-2010 13:46:52] that sounds like a munged install
[13-Jul-2010 13:50:04] varnii: that's bad
[13-Jul-2010 13:50:10] varnii: start from scratch
[13-Jul-2010 13:50:25] varnii: you need the default classes and templates
[13-Jul-2010 13:50:34] hmm rmatte, interestingly if i go into privileged -> http under "Server Instances" I don't see the server(s) in question. I see a bunch of solaris servers that work. I don't see the linux servers that have the problem
[13-Jul-2010 13:50:49] robo: then maybe it's httpd under processes?
[13-Jul-2010 13:51:00] what section is it in in the OS tab?
[13-Jul-2010 13:51:06] that'll be a dead giveaway
[13-Jul-2010 13:51:34] in the OS tab it's under OS Processes
[13-Jul-2010 13:51:46] rmatte: Yeah, the strange thing is I didn't need them from the start, I added our entire production line, added / created a script to add 81 custom alerting rules, and now I need to monitor disk usage haha :D
[13-Jul-2010 13:51:52] I'm fucked :}
[13-Jul-2010 13:51:54] jmp242: so I tested another couple of plugins
[13-Jul-2010 13:51:57] still erroring out the same way
[13-Jul-2010 13:52:07] robo: then it's in processes, not services
[13-Jul-2010 13:52:24] ahah! okay, that makes sense -- hehe
[13-Jul-2010 13:52:34] robo: do you see it in Processes?
[13-Jul-2010 13:53:09] hmm
[13-Jul-2010 13:53:18] rmatte are you matt ray
[13-Jul-2010 13:53:23] no
[13-Jul-2010 13:53:25] lies
[13-Jul-2010 13:53:25] lol
[13-Jul-2010 13:53:27] I'm Ryan Matte
[13-Jul-2010 13:53:28] that would be mray
[13-Jul-2010 13:53:28] so many of the same ltters
[13-Jul-2010 13:53:32] oh cool
[13-Jul-2010 13:53:57] i'm on the webinar right now, i asked about the monitoring by host :D
[13-Jul-2010 13:54:04] ah
[13-Jul-2010 13:54:05] yeah
[13-Jul-2010 13:54:11] thiis is a good webinar
[13-Jul-2010 13:54:16] "NOPE, moving along!"
[13-Jul-2010 13:54:18] lol
[13-Jul-2010 13:54:33] yeh, the webinars are good for getting a start with the product
[13-Jul-2010 13:54:51] rmatte, yup -- found it. I changed the regex to ^.*\/httpd
[13-Jul-2010 13:54:57] hopefully that works. ty for your help!
[13-Jul-2010 13:55:18] np
[13-Jul-2010 13:56:39] bah :)
[13-Jul-2010 13:56:48] rmatte: I don't suppose you have any ideas on my zenpack issue?
[13-Jul-2010 13:57:02] * rmatte is reading up
[13-Jul-2010 13:57:05] thank you
[13-Jul-2010 13:57:50] is that the only zenpack doing that to you?
[13-Jul-2010 13:58:10] hmm I'm asking this question without much research, we are in a bit of a mess though! Could anyone briefly explain how to use SNMP to mointor disk usage, I've got my snmpd.conf setup properly, ' disk /usr 15% '
[13-Jul-2010 13:58:24] not sure where to start on the zenoss side of things
[13-Jul-2010 13:58:37] rmatte: nope - every one I've tried
[13-Jul-2010 13:59:06] Zenoss 2.5.0
[13-Jul-2010 13:59:16] first off, upgrade to 2.5.2
[13-Jul-2010 13:59:21] 2.5.0 is buggier than all hell
[13-Jul-2010 13:59:34] oh
[13-Jul-2010 13:59:35] :)
[13-Jul-2010 13:59:56] after the upgrade, then try
[13-Jul-2010 14:00:01] are there docs for upgrading?
[13-Jul-2010 14:00:05] I can search >.>
[13-Jul-2010 14:00:12] what install method did you use?
[13-Jul-2010 14:00:26] and yes there are, it's called the install guide
[13-Jul-2010 14:00:30] read the latest one
[13-Jul-2010 14:00:37] it's available in pdf format on the site
[13-Jul-2010 14:00:50] perfect - I just grabbed a jumpbox
[13-Jul-2010 14:00:54] so it came pre-installed
[13-Jul-2010 14:00:55] community/documentation
[13-Jul-2010 14:00:59] I am not sure what method they would use
[13-Jul-2010 14:00:59] Zenoss Installation Guide
[13-Jul-2010 14:01:01] yeah - got it already
[13-Jul-2010 14:01:02] thank
[13-Jul-2010 14:01:03] you
[13-Jul-2010 14:01:09] np
[13-Jul-2010 14:01:58] nice presentation mray
[13-Jul-2010 14:03:09] ah wonderful; ii jumpbox-zenoss 2.5.0-2 JumpBox Package for Zenoss
[13-Jul-2010 14:03:10] Parabola: thanks
[13-Jul-2010 14:03:17] I really am having no luck today
[13-Jul-2010 14:03:19] :)
[13-Jul-2010 14:03:42] mray: do you know the average turn around time for response from sales?
[13-Jul-2010 14:03:54] like, will it be a week or two before i hear back
[13-Jul-2010 14:04:39] I would hope that sales would be on that like white on rice
[13-Jul-2010 14:04:40] lol
[13-Jul-2010 14:04:55] yeah me too, i want to get that feature since you're a LAZY SCRIPTER!!!
[13-Jul-2010 14:04:57] :P
[13-Jul-2010 14:05:04] lol
[13-Jul-2010 14:05:13] the feature would be better anyways, much cleaner
[13-Jul-2010 14:05:25] yeah i agree
[13-Jul-2010 14:07:41] Parabola: probably depends who you are and how big you are :)
[13-Jul-2010 14:07:50] :(
[13-Jul-2010 14:08:11] makes sense though, i'm not very big at all, 5'11 160lbs :P
[13-Jul-2010 14:08:17] lol
[13-Jul-2010 14:08:23] lol
[13-Jul-2010 14:09:05] common, it's the internet, you mean to say you're 6'5" 220lbs
[13-Jul-2010 14:10:15] hi rmatte, how are you?
[13-Jul-2010 14:12:45] good thanks, you?
[13-Jul-2010 14:16:52] I'm fine
[13-Jul-2010 14:17:06] needing some help with HttpMonitor :s
[13-Jul-2010 14:18:06] I'll explain my scenario so you can see my doubt
[13-Jul-2010 14:18:17] k
[13-Jul-2010 14:18:51] I have like 8 servers running web services, daemons and other stuff
[13-Jul-2010 14:19:26] ok
[13-Jul-2010 14:19:51] I have some scripts written on php that check for error in web service content
[13-Jul-2010 14:20:34] so, If I run a script, it can tell me weather a web service is fine or is not
[13-Jul-2010 14:21:48] I wish to use HttpMonitor to call these scripts frecuently and check for his answer to see weather is error, warning, etc
[13-Jul-2010 14:21:58] did I make mysql clear?
[13-Jul-2010 14:22:06] myself
[13-Jul-2010 14:22:08] hahaha
[13-Jul-2010 14:22:09] ROFL
[13-Jul-2010 14:22:30] you would be much better off using this chachan: docs/DOC-3757
[13-Jul-2010 14:22:49] I do the same thing I believe chachan
[13-Jul-2010 14:23:25] I have perl scripts that return errors / error messages, and you want zenoss command datasources to monitor them?
[13-Jul-2010 14:24:57] varnii, yes
[13-Jul-2010 14:25:46] I want to zenoss call them and check their own status code and submits alerts if is needly, etc
[13-Jul-2010 14:26:26] sorry for my english :s
[13-Jul-2010 14:28:37] It has a lot to do with the output of scripts I found. We ended up creating a basic default_check script on our zenoss instance that is run at set intervals to grab the return of other various scripts. SCRIPT_1 checks for 500 errors in apache.log lets say, it returns pass or fail, the default_check script checks the cgi for pass/fails, and outputs the error message if it's a fail, and also EXITS a numercial value that zenoss can read, 0 or 1
[13-Jul-2010 14:29:18] default_script_500log, the zencommand looks like /usr/local/zenoss/libexec/default_script -H HOST -u URL_TO_CGI
[13-Jul-2010 14:29:21] if that makes sense..
[13-Jul-2010 14:32:14] varnii, I understand
[13-Jul-2010 14:32:29] excellent, good luck =)
[13-Jul-2010 14:32:35] thanks!
[13-Jul-2010 14:33:26] Synthetic Web Transactions is excellent for that ;>
[13-Jul-2010 14:33:31] and yes as rmatte mentioned, read up on some things, especially any NAGIOS output related documents, very helpful
[13-Jul-2010 14:33:38] yeh, which is what I was saying
[13-Jul-2010 14:33:43] jups
[13-Jul-2010 14:33:53] (regarding was forsberg said)
[13-Jul-2010 14:34:03] ;)gotcha
[13-Jul-2010 14:34:06] twill is an awesome language for it
[13-Jul-2010 14:34:10] (i just confirmed and backed rmatte up)
[13-Jul-2010 14:34:10] :)
[13-Jul-2010 14:34:17] (yes he did)
[13-Jul-2010 14:34:51] (why are we suddenly talking like this?)
[13-Jul-2010 14:35:08] (dunno, its been a long day and im silly)
[13-Jul-2010 14:35:12] [break the mold]
[13-Jul-2010 14:35:18] (fair enough reason)
[13-Jul-2010 14:35:24] (oh no)
[13-Jul-2010 14:35:26] (he broke it)
[13-Jul-2010 14:36:26] and just when I was about to put .)(. in brackets
[13-Jul-2010 14:36:32] * rmatte huffs
[13-Jul-2010 14:37:02] :]
[13-Jul-2010 14:37:32] do you think once 3.0 core is released the upgrade will fix all of our missing stuff from 2.5.2 :o?
[13-Jul-2010 14:37:47] varnii: highly doubt it
[13-Jul-2010 14:38:02] varnii: that stuff is only created on install
[13-Jul-2010 14:39:17] long week ahead then ;|
[13-Jul-2010 14:40:04] or you could just use zenbackup to backup only your devices, and do a fresh install
[13-Jul-2010 14:40:06] * rmatte shrugs
[13-Jul-2010 14:40:26] I really don't recommend using the vmware appliance
[13-Jul-2010 14:40:42] fresh install of linux + fres install of Zenoss is optimum
[13-Jul-2010 14:40:48] fresh*
[13-Jul-2010 14:41:04] how many devices you have?
[13-Jul-2010 14:42:15] only 18, creating all our templates again is what takes awhile, I didn't make a script for it
[13-Jul-2010 14:42:32] oi
[13-Jul-2010 14:43:04] should be doable tho :) im still recovering from my 400+ devices reinstall :/
[13-Jul-2010 14:43:05] you should be able to create a zenpack and add all those templates to the zenpack
[13-Jul-2010 14:43:26] forsberg: ouch
[13-Jul-2010 14:43:28] I am still working on my 4000+ going from 2.2.4 to 2.5.2
[13-Jul-2010 14:43:29] yeah depending on how they are made, and ive gotten alot of rutine with zenoss in the process
[13-Jul-2010 14:43:32] ptom: very good point, I never thought of that
[13-Jul-2010 14:43:37] ^^ what ptom said
[13-Jul-2010 14:43:43] lol
[13-Jul-2010 14:43:47] lol
[13-Jul-2010 14:43:49] :)
[13-Jul-2010 14:43:59] i usually give terrible advice
[13-Jul-2010 14:44:08] you might have just saved the day
[13-Jul-2010 14:44:15] a few days at least
[13-Jul-2010 14:44:49] one thing about that tho, ive had problems with having "strange" zenpacks, because if you later need to use backup/restore etc, you need an install with the same zenpacks to restore to
[13-Jul-2010 14:46:23] yes, good point. You could copy/rename them after you import the zenpack and then delete the zenpack I suppose
[13-Jul-2010 14:47:03] theres my bit of bad advice.
[13-Jul-2010 14:47:18] hah
[13-Jul-2010 14:47:43] no
[13-Jul-2010 14:47:45] you cannot
[13-Jul-2010 14:47:48] or
[13-Jul-2010 14:47:54] i didnt find the right way to do that anyways
[13-Jul-2010 14:48:10] even if i copy it out and everything, it deletes with the zenpack
[13-Jul-2010 14:48:27] im not finnished exploring it, but should just be aware of it
[13-Jul-2010 14:48:52] huh, good to know
[13-Jul-2010 14:49:12] yeah, thanks, I think this installation is a little more fubard that I thought. Add zenpack = Report error page
[13-Jul-2010 14:49:56] good to know though
[13-Jul-2010 14:52:49] looks like I get to drive home through a monsoon. I can't wait.
[13-Jul-2010 14:53:39] we have a lot of rain here too today
[13-Jul-2010 14:54:57] it's good for my lawn. I guess.
[13-Jul-2010 15:01:26] mrayzenoss is now known as mray
[13-Jul-2010 15:07:29] guess who's back
[13-Jul-2010 15:07:35] back again
[13-Jul-2010 15:07:36] me
[13-Jul-2010 15:07:38] duh
[13-Jul-2010 15:07:40] gibbon's back
[13-Jul-2010 15:07:59] that's all I got
[13-Jul-2010 15:08:26] gibbons'*
[13-Jul-2010 15:09:37] my boss asked me to put our core links on 60 second collection intervals. I told him I needed a new server. He walked away.
[13-Jul-2010 15:09:46] lol
[13-Jul-2010 15:10:16] ptom: heh, I've had similar conversations :)
[13-Jul-2010 15:10:49] nice
[13-Jul-2010 15:10:59] I guess I understand where he is comign from.
[13-Jul-2010 15:11:12] it would saturate some of our links though
[13-Jul-2010 15:11:19] Everyone would love to sample every 30 or 60 seconds and keep the data forever.
[13-Jul-2010 15:11:31] yeh, ridiculous though
[13-Jul-2010 15:11:36] realistically, 5 minutes is fine
[13-Jul-2010 15:11:46] And then you draw the exponential chart of price and that usually solves the problem.
[13-Jul-2010 15:11:55] hehe
[13-Jul-2010 15:11:55] hahaah
[13-Jul-2010 15:13:14] even if i did change it from 300 to 60, wouldn't I have to re-generate the rrds?
[13-Jul-2010 15:13:20] and lose historical data
[13-Jul-2010 15:13:21] minor update to the Fortigate zenpack
[13-Jul-2010 15:13:23] yes, you would
[13-Jul-2010 15:13:44] you'd have to modify the RRD create settings for the collector too to match
[13-Jul-2010 15:14:14] i didn't think of that
[13-Jul-2010 15:14:44] would be interesting to see somone integrate collectd and zenoss
[13-Jul-2010 15:14:52] you could get your high resolution graphs that way
[13-Jul-2010 15:15:37] high resolution graphs == more load and more IO
[13-Jul-2010 15:16:08] you can get high resolution graphs in Zenoss, settings need to be tweaked
[13-Jul-2010 15:16:10] but it's intensive
[13-Jul-2010 15:16:23] you'd want to use an SSD or something
[13-Jul-2010 15:17:13] it's hard enough to get a standard sata drive around here.
[13-Jul-2010 15:19:13] lol
[13-Jul-2010 15:22:14] cgibbons: any idea what the intention of the options to add an IpInterface, IpRouteEntry, or whatever component is for in the new UI?
[13-Jul-2010 15:22:27] not really
[13-Jul-2010 15:22:31] I mean, it prompts for an ID, and adds an interface or whatever under that id which is editable
[13-Jul-2010 15:22:36] but it's essentially completely useless
[13-Jul-2010 15:22:41] I don't even know why it's there
[13-Jul-2010 15:29:11] guess I'll log a trac for it and see what is said
[13-Jul-2010 15:33:41] so guys - asked the guys over at jumpbox for some help on their forums
[13-Jul-2010 15:33:46] after looking in to things a little further
[13-Jul-2010 15:33:58] root@zenoss:~# dpkg -l | grep -i zenoss
[13-Jul-2010 15:33:58] ii jumpbox-zenoss 2.5.0-2 JumpBox Package for Zenoss
[13-Jul-2010 15:34:15] ofc - apt-get install jumpbox-zenoss says it's at the current version
[13-Jul-2010 15:34:32] not according to their zenoss page - it states current version as 2.5.2
[13-Jul-2010 15:34:36] that is the current package that they offer
[13-Jul-2010 15:35:13] http://www.jumpbox.com/app/zenoss
[13-Jul-2010 15:35:31] ofc nothing in their docs on upgrade etc / without launching a new jumpbox and copying over configs
[13-Jul-2010 15:35:36] which seems redundant
[13-Jul-2010 15:35:45] will see what happens
[13-Jul-2010 15:37:22] can you do an apt-get update ?
[13-Jul-2010 15:38:41] I did
[13-Jul-2010 15:38:48] no jumpbox-zenoss package for upgrade
[13-Jul-2010 15:39:00] I'm a long time ubuntu user, so well versed there, mostly
[13-Jul-2010 15:39:09] :) I had to ask
[13-Jul-2010 15:43:10] wow
[13-Jul-2010 15:43:13] jumpbox is a fucking ripoff
[13-Jul-2010 15:43:14] enoss requires dependencies like Zope, Python, NetSNMP, RRDtool, and MySQL to work effectively. You could spend days, even weeks configuring your hardware to support it
[13-Jul-2010 15:43:35] really? weeks? because its not like all of that doesnt come with the binaries.
[13-Jul-2010 15:44:30] it took me weeks and weeks to type the yum install line required to get zenoss working
[13-Jul-2010 15:44:33] each day i'd press one more key
[13-Jul-2010 15:44:36] yeah
[13-Jul-2010 15:44:48] i didnt even add teh repo i just used a .bin file
[13-Jul-2010 15:47:11] yeh, the .bin is the stack installer
[13-Jul-2010 15:47:17] that's my favourite method
[13-Jul-2010 15:47:21] anyways, I'm out for the day
[13-Jul-2010 15:47:22] cheers
[13-Jul-2010 15:47:26] yeah me too rmatte
[13-Jul-2010 15:47:27] o/
[13-Jul-2010 15:47:35] Parabola: yeah - to be fair it is a bit
[13-Jul-2010 15:47:41] I have zenoss running myself at home
[13-Jul-2010 15:47:49] but for work - CTO wanted a jumpbox
[13-Jul-2010 15:48:00] ...why
[13-Jul-2010 15:48:01] we have it running on an ec2 instance
[13-Jul-2010 15:48:03] sounds like a great way to always be behind
[13-Jul-2010 15:48:08] yeah
[13-Jul-2010 15:48:11] no joke, stupid CTO
[13-Jul-2010 15:48:13] I hear you guys :P
[13-Jul-2010 15:48:18] thats why he makes 5x what you do
[13-Jul-2010 15:49:32] hurrah, that's my zenoss server ordered
[13-Jul-2010 15:50:08] 24G of ram, some.. kind of xeon, 3x300G disks... probably my nicest server :(
[13-Jul-2010 15:50:30] ashp: nice!
[13-Jul-2010 15:51:16] the rest of my servers are virtual and have 30G of disk and as little memory as i can get away with
[13-Jul-2010 15:51:17] why all that
[13-Jul-2010 15:51:27] 24gb of ram for zenoss?
[13-Jul-2010 15:51:41] i told him i needed 8G, he ordered 24G
[13-Jul-2010 15:51:47] i think at this point ram is so cheap they don't bother doing less
[13-Jul-2010 15:52:00] why 8
[13-Jul-2010 15:52:06] how many devices are you monitoring
[13-Jul-2010 15:52:17] i picked the mid-level recommendations outta the install docs
[13-Jul-2010 15:52:24] i'm running it on a VM (esxi 4 host) with 2 or 4gb of ram
[13-Jul-2010 15:52:46] right now it's only about 50 devices but we expect to increase that in the future, and my boss doesn't want the monitoring virtual
[13-Jul-2010 15:52:49] in case esx dies
[13-Jul-2010 15:54:07] woot my first install of exchange 2010
[13-Jul-2010 15:54:13] on server 2008 r2, new stuff FTL
[13-Jul-2010 15:58:16] my condolences
[13-Jul-2010 15:58:58] * nickanderson turns back to his
[13-Jul-2010 16:00:21] hey all, about to migrate from alienvault to zenoss (OS), can the vm image be deployed within esxi? also, is a x64 bit image avail for esxi?
[13-Jul-2010 16:01:30] lol
[13-Jul-2010 16:01:48] sideone: dont fuck with prebuilt VM templates/ OVAs
[13-Jul-2010 16:01:52] they are out of date and have problems
[13-Jul-2010 16:01:57] just setup a linux VM and install zenoss
[13-Jul-2010 16:01:57] thanks!
[13-Jul-2010 16:02:06] ok cool, base ubuntu x64 ok?
[13-Jul-2010 16:02:09] yup
[13-Jul-2010 16:02:18] they have a repo to add to your apt repos
[13-Jul-2010 16:02:21] so the install is quick
[13-Jul-2010 16:02:29] thanks buddy.. any prereqs or just run the installer script i guess?
[13-Jul-2010 16:02:40] Parabola; hah even better
[13-Jul-2010 16:02:43] :) yup
[13-Jul-2010 16:02:48] no more makes, apt is great
[13-Jul-2010 16:02:54] if you have any problems let us know
[13-Jul-2010 16:02:58] apt > yum, rpm, etc..
[13-Jul-2010 16:02:58] i'm not very useful with zenoss
[13-Jul-2010 16:03:03] but i can help with debian/ubuntu
[13-Jul-2010 16:03:14] and exchange
[13-Jul-2010 16:03:17] ;)
[13-Jul-2010 16:03:19] lol
[13-Jul-2010 16:03:23] i'm having problems with that too
[13-Jul-2010 16:03:27] cool thanks, been on deb* since i switched from slak back in the 90s
[13-Jul-2010 16:03:29] the schema changes arent working
[13-Jul-2010 16:03:32] hosted exchange ftw
[13-Jul-2010 16:03:33] :P
[13-Jul-2010 16:03:34] sideone: :) nice
[13-Jul-2010 16:03:37] willwh: thats what im setting up
[13-Jul-2010 16:03:41] but still a *bsd guy
[13-Jul-2010 16:03:43] software as a managed service = the future
[13-Jul-2010 16:03:49] it is certainly
[13-Jul-2010 16:03:51] i like zimbra
[13-Jul-2010 16:03:57] i'm consulting for a company helping them move to it
[13-Jul-2010 16:04:04] P2Ving their servers over to vmware
[13-Jul-2010 16:04:04] :)
[13-Jul-2010 16:04:10] moving clients to terminal services
[13-Jul-2010 16:04:12] ive deployed nag* many times, how is zo? is it nag with better admin/config?
[13-Jul-2010 16:04:14] its good stuff
[13-Jul-2010 16:04:36] sideone: if oyu're a bsd guy, can you build me a hybrid install of FreeNAS and PFsense?
[13-Jul-2010 16:04:44] I have a box at home for each and its stupid
[13-Jul-2010 16:04:46] Parabola; why freenas?
[13-Jul-2010 16:04:53] 1 sec.
[13-Jul-2010 16:04:55] and due to controller card, i cant create a VM
[13-Jul-2010 16:04:59] doesnt have to be freenas
[13-Jul-2010 16:05:03] it just needs to have samba and ZFS
[13-Jul-2010 16:05:23] i would use openfiler
[13-Jul-2010 16:05:33] much better and iSCSI support if needed
[13-Jul-2010 16:05:33] why
[13-Jul-2010 16:05:42] i already have my ZFS pool setup
[13-Jul-2010 16:05:43] !google freenas vs openfiler
[13-Jul-2010 16:05:48] been using it for ever
[13-Jul-2010 16:05:48] [Google] #1 Openfiler vs FreeNAS: Tips for building your own NAS | blog ... ( http://blog.patyuen.com/lessons/technology/openfiler-vs-freenas-tips-for-building-your-own-nas/ )
[13-Jul-2010 16:05:50] [Google] Desc: If you run multiple computers in your home or office, centralized storage can be a daunting task. Ideally, you want your files to be available to all ...
[13-Jul-2010 16:06:16] Parabola, nothing wrong with it, i just think openfiler is a bit smoother
[13-Jul-2010 16:06:24] can even build it as an esxi base
[13-Jul-2010 16:07:02] whats openfiler built on?
[13-Jul-2010 16:07:36] rpath iirc
[13-Jul-2010 16:08:10] i believe rpath
[13-Jul-2010 16:08:11] http://www.openfiler.com/community/download/
[13-Jul-2010 16:08:19] they have vm's ready for deployment
[13-Jul-2010 16:08:29] you can prob tweak one with tables?
[13-Jul-2010 16:08:37] if you dont aboslutly need pf
[13-Jul-2010 16:09:05] no no the issue is my controller
[13-Jul-2010 16:09:09] i cant use a VM :(
[13-Jul-2010 16:09:28] i need SMB/CIFS and ZFS support on my pFsense box
[13-Jul-2010 16:09:29] then i'm set
[13-Jul-2010 16:09:33] i can throw one of the boxes out
[13-Jul-2010 16:09:56] ahh
[13-Jul-2010 16:10:08] Parabola; i am testing endian currently
[13-Jul-2010 16:10:17] i was using pfS at home
[13-Jul-2010 16:10:33] i like endian's setup and ntop support
[13-Jul-2010 16:12:30] i moved from untangle to pfsense last night
[13-Jul-2010 16:12:35] stupid performance issues
[13-Jul-2010 16:12:39] core2duo, 6GB of ram
[13-Jul-2010 16:13:00] 3 users, 25mbit/1mbit
[13-Jul-2010 16:18:38] i really didnt like untangle
[13-Jul-2010 16:19:18] but i used to develop routers for GECF worldwide on a freebsd*/open* platform, so i guess i dont like alot of *nix routers.
[13-Jul-2010 16:19:50] 3/users and 25MB? what porn are ya slanging?
[13-Jul-2010 16:21:44] n/a
[13-Jul-2010 16:23:46] hah
[13-Jul-2010 17:01:08] cool, I've got a preliminary list of "known incompatible Community ZenPacks with Zenoss 3.0"
[13-Jul-2010 17:02:11] <|rt|> mray: when is 3.0 coming out....last press release I saw mentioned july 10th
[13-Jul-2010 17:02:33] hmm… the releases I saw said "July"
[13-Jul-2010 17:02:44] I wouldn't hold your breath for the 10th ;)
[13-Jul-2010 17:03:31] <|rt|> hehe
[13-Jul-2010 17:04:07] <|rt|> is there any way to check if your zenpacks have any updates available?
[13-Jul-2010 17:04:29] sadly no
[13-Jul-2010 17:04:45] I'm contacting the authors of the ones I know are 3.0-incompatible
[13-Jul-2010 17:09:57] <|rt|> is there a page the talks about what new features are going to be in 3.0?
[13-Jul-2010 17:10:54] <|rt|> if I look for things tagged 3.0 on the website I just see posts about the rc and alpha releases but no release notes
[13-Jul-2010 17:14:06] Just upgraded to 3.0 RC1 - is it normal for zenactions to not be running?
[13-Jul-2010 17:14:26] (and where can I post bug reports?)
[13-Jul-2010 17:14:51] (tried to do a backup and it failed with a Python error. I got a screenshot I can submit)
[13-Jul-2010 17:34:07] NM - posting it in the forum...
[13-Jul-2010 18:09:38] forsberg is now known as fOrsberg
[13-Jul-2010 21:57:24] If anyone's around… 3.0-Incompatible Community ZenPacks: docs/DOC-5914
[13-Jul-2010 23:22:25] Hello. I am getting a "key error" when installing a zenpack. What does that mean?
[14-Jul-2010 00:00:30] [disconnected at Wed Jul 14 00:00:30 2010]
[14-Jul-2010 00:00:30] [connected at Wed Jul 14 00:00:30 2010]
[14-Jul-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]
[14-Jul-2010 00:31:09] realitygeoff: sounds like a python error. paste the full output at a pastebin
[14-Jul-2010 00:34:56] http://pastebin.com/fb3EFbdF
[14-Jul-2010 01:01:10] hi
[14-Jul-2010 01:16:11] fOrsberg is now known as forsberg
[14-Jul-2010 01:18:24] hi
[14-Jul-2010 01:35:13] Ok, I worked around the key file error by adding the egg file directly, but now my question is: how do I associate this template with my /Server/Linux class?
[14-Jul-2010 01:43:29] you want to make it on the class so all servers in that class will inherit it ?
[14-Jul-2010 01:43:36] in all future
[14-Jul-2010 01:45:19] if so you go devices in left bar menu - then browse to server/linux - click the templates menu @ top - then get the lower dropdown menu and chose bind templates
[14-Jul-2010 07:59:01] bzed_ is now known as Guest90611
[14-Jul-2010 08:06:28] hello
[14-Jul-2010 08:06:41] zenoss uses rrd files to record statistics like the cpu usage
[14-Jul-2010 08:06:54] however, does it prune the rrd files at some point by default?
[14-Jul-2010 08:07:06] or does it record everything until disk space runs out?
[14-Jul-2010 08:07:43] I actually want the latter, since I want to test a distributed system
[14-Jul-2010 08:08:09] and the cpu/io/ram usage/etc. measurements of each device are necessary to draw some conclusions
[14-Jul-2010 08:09:00] this implies that (a) zenoss must not throw away any portion of the rrd data (b) I somehow need to make a copy of the rrd files so that I can give said copy to other people who will draw conclusions from it
[14-Jul-2010 08:09:26] rrd files are self pruning by design
[14-Jul-2010 08:09:32] also, it would be useful to extract the event console history in some way
[14-Jul-2010 08:09:43] have a read about rrdtool
[14-Jul-2010 08:10:06] you can modify the create command to make it keep data for as long as you like basically
[14-Jul-2010 08:10:58] yes,thats what I mean
[14-Jul-2010 08:11:07] just set the time covered to a week for example
[14-Jul-2010 08:11:10] that would be enough
[14-Jul-2010 08:11:25] can this be done via the zenoss web frontend?
[14-Jul-2010 08:11:39] yup, under collectors->localhost
[14-Jul-2010 08:12:03] you can change the create command to create the rrd DS's that you want
[14-Jul-2010 08:18:06] hmm
[14-Jul-2010 08:18:08] "rows defines how many generations of data values are kept in an RRA."
[14-Jul-2010 08:18:14] what is a "generation of data values" ?
[14-Jul-2010 08:20:32] http://oss.oetiker.ch/rrdtool/tut/rrd-beginners.en.html
[14-Jul-2010 08:20:34] have a read about rrd. it aggregates data for an interval (user defined) and stores it in row.
[14-Jul-2010 08:20:59] oh. I missed that one. thanks.
[14-Jul-2010 08:29:30] is the event log cycle interval somehow related to the rrd tool step size (which is given in seconds as well) ?
[14-Jul-2010 08:33:05] the polling cycle interval is related
[14-Jul-2010 08:33:12] snmp polling cycle
[14-Jul-2010 08:33:28] hm. in what way? 1:1? I cant find a documentation about this
[14-Jul-2010 08:33:35] by default it polls every 5 mins, if you modify the cycle time you need to modify the RRD params to match
[14-Jul-2010 08:33:44] okay
[14-Jul-2010 08:33:44] I don't know off the top of my head
[14-Jul-2010 08:33:48] haven't changed it in a while
[14-Jul-2010 08:34:15] however, the only rrd params I can modify - at least as far as I can see it right now - are the RRA specs
[14-Jul-2010 08:34:22] you need to make sure that there is a least one completed snmp-cycle for the aggregation interval.
[14-Jul-2010 08:35:01] if you complete more cycles rrd will aggregate, if you have less you are missing a datapoint
[14-Jul-2010 08:35:15] okay, so if the snmp performance cycle interval is 300 secs, then the --step parameter will be 300 or more
[14-Jul-2010 08:35:22] someone really needs to write a Zenoss specific doc about this
[14-Jul-2010 08:35:22] lol
[14-Jul-2010 08:35:27] with actual examples
[14-Jul-2010 08:35:41] that way we don't have to half-ass explain it every time someone asks about it
[14-Jul-2010 08:35:48] :)
[14-Jul-2010 08:36:01] if you read about rrd (and understand it) it is rather obvious
[14-Jul-2010 08:36:03] pfff, training by committee is fun
[14-Jul-2010 08:36:52] isn't it trial in front of a comittee
[14-Jul-2010 08:37:05] well yes, if you do not have enough pdps to create a cdp, you miss a data point
[14-Jul-2010 08:38:16] nah, strike that. my bad.
[14-Jul-2010 08:40:21] so if each pdp is generated every 300 seconds, and 1 pdp = 1 cdp, and there are 600 rows, then the rra will hold data for 300 * 600 seconds -> 50 hours
[14-Jul-2010 08:56:28] forsberg is now known as fOrsberg
[14-Jul-2010 09:10:31] dv_: well, it'll hold high resolution data for that long
[14-Jul-2010 09:10:47] dv_: before it starts summarizing it
[14-Jul-2010 09:24:12] OK, second time this has happened, very weird
[14-Jul-2010 09:25:01] I've got 11 Windows03 servers, all in the same domain, that all randomly decided to stop responding to SNMP requests
[14-Jul-2010 09:25:44] twm1010: hmmm, firewall?
[14-Jul-2010 09:25:55] all have firewalls disabled
[14-Jul-2010 09:26:05] its not that zenoss can't hit them via SNMP, nothing can
[14-Jul-2010 09:26:12] right, but you're obviously going through network equipment to get to them
[14-Jul-2010 09:26:13] if i bounce the SNMP service, it works again
[14-Jul-2010 09:26:20] no firewall between, same LAN
[14-Jul-2010 09:26:20] if someone was screwing around with ACLs on a router or something
[14-Jul-2010 09:26:35] same physical switch, in the case of the zenoss server
[14-Jul-2010 09:26:42] odd
[14-Jul-2010 09:26:43] no idea
[14-Jul-2010 09:27:00] I've never had that sort of thing happen
[14-Jul-2010 09:34:57] and to confirm something: in case of the mysql monitor core zenpack, if somebody resets the DB that is being watched, and therefore drops the zenoss user, zenoss reports this to me
[14-Jul-2010 09:35:18] I do -not- have to restart something manually after the zenoss user has been recreated, right?
[14-Jul-2010 09:36:42] what do you mean after the zenoss user has been recreated?
[14-Jul-2010 09:36:48] why would you have to recreate it?
[14-Jul-2010 09:37:01] well because somebody dropped the zenoss mysql user by accident
[14-Jul-2010 09:37:15] oh, but not the database right?
[14-Jul-2010 09:37:18] no
[14-Jul-2010 09:37:24] then you're fine
[14-Jul-2010 09:37:32] no restart required
[14-Jul-2010 09:37:49] yes, the graph was there all the time, and mysql wasnt killed or something
[14-Jul-2010 09:37:55] so its just a little empty space in the graph, nothing more
[14-Jul-2010 09:38:01] hrmm, interesting
[14-Jul-2010 09:38:12] all 11 servers coincide with a group policy change i made yesterday
[14-Jul-2010 09:38:27] except i didn't touch the SNMP policy GPO, i created a different one
[14-Jul-2010 09:39:17] dv_ what graph are you referring to?
[14-Jul-2010 09:39:52] twm1010: sounds like a lead
[14-Jul-2010 09:39:55] what did you change?
[14-Jul-2010 09:40:22] I pushed a root CA cert for our Novell environment out to our AD systems
[14-Jul-2010 09:40:44] but why only these 11 servers? i have about 35
[14-Jul-2010 09:41:06] Is it possible to add custom schema objects to a zenpack?
[14-Jul-2010 09:41:47] nickanderson: read the zenoss developer's guide, it's possible to add anything to a ZenPack
[14-Jul-2010 09:45:18] oh i can never remember, what's the pastebin for images?
[14-Jul-2010 09:45:32] tinypic?
[14-Jul-2010 09:45:47] yup
[14-Jul-2010 09:45:52] that's the one I use
[14-Jul-2010 09:45:53] imagebin.org
[14-Jul-2010 09:46:02] that works too
[14-Jul-2010 09:46:07] there's a bunch of them
[14-Jul-2010 09:46:15] rmatte: the core mysql monitor zenpack produces extra graphs for mysql statistics
[14-Jul-2010 09:46:22] like, amount of select statements per second
[14-Jul-2010 09:46:25] dv_: ah, I see
[14-Jul-2010 09:46:42] I thought that was just for monitoring remote MySQL instances
[14-Jul-2010 09:47:09] look at my CPU graph
[14-Jul-2010 09:47:11] http://i30.tinypic.com/21mc2rm.jpg
[14-Jul-2010 09:47:20] how odd is that
[14-Jul-2010 09:47:56] twm1010: corrupt RRDs probably, delete them
[14-Jul-2010 09:48:34] well, that sucks, what causes corruption?
[14-Jul-2010 09:48:57] * rmatte shrugs
[14-Jul-2010 09:49:06] I've seen it happen a few times, but not often
[14-Jul-2010 09:49:17] I've had to delete the odd RRD to get Zenoss graphing again
[14-Jul-2010 09:49:20] rare though
[14-Jul-2010 09:50:09] If that is the reason and you can see old graph data you may be able to dump and restore the RRD file to preserve the history you have.
[14-Jul-2010 09:50:23] Go into $ZENHOME/perf/Devices/devicename and do something like..
[14-Jul-2010 09:50:33] rrdtool dump ssCpuUser.rrd > ssCpuUser.xml
[14-Jul-2010 09:50:55] rm ssCpuUser.rrd && rrdtool restore ssCpuUser.xml ssCpuUser.rrd
[14-Jul-2010 09:50:59] On each of the affected files.
[14-Jul-2010 09:51:11] i'll give it a shot, can't hurt, i can always delete it
[14-Jul-2010 09:51:23] these are for my Zenoss server itself, so nobody really looks at this but me
[14-Jul-2010 09:52:25] twm1010: if that actually works let me know since that would be cool
[14-Jul-2010 09:53:23] i'll let you know in a sec, i just cycled my server for the first time in a long time
[14-Jul-2010 09:53:34] cool
[14-Jul-2010 09:58:28] Well, I suppose I have to wait a few polling cycles to find out
[14-Jul-2010 09:59:04] how does the "load average" graph work?
[14-Jul-2010 09:59:25] it polls the snmp values
[14-Jul-2010 09:59:43] yes, but what is that supposed to tell me, I see the y axis is "processes"
[14-Jul-2010 09:59:55] is that supposed to tell me the number of processes currently crunching away?
[14-Jul-2010 09:59:59] it's a bad label.
[14-Jul-2010 10:00:34] login to the system, run top you will se the values in the top right corner (or run w)
[14-Jul-2010 10:00:56] it's the number of processes that wanted to run
[14-Jul-2010 10:01:10] I see
[14-Jul-2010 10:01:12] but it's not accurate
[14-Jul-2010 10:01:21] so is that a good gauge of load, or what?
[14-Jul-2010 10:01:37] if you can interpret it yes. otherwise you can use it to detect trends.
[14-Jul-2010 10:01:54] if a process is blocked by I/O it will still add to the load value
[14-Jul-2010 10:02:04] but not spike the CPU, makes sense
[14-Jul-2010 10:02:20] you can have a perfectly fine NFS server with a load of 100+
[14-Jul-2010 10:02:52] so... load over all system resources, where in Wintel land we typically just look at CPU
[14-Jul-2010 10:03:01] normaly i would say max-load should be number of cores
[14-Jul-2010 10:03:13] no.
[14-Jul-2010 10:04:10] found an article: http://www.linuxjournal.com/article/9001
[14-Jul-2010 10:04:48] it's the number of processes that would like to do something (CPU wise). But as also blocked processes contribute to the value it can misslead you.
[14-Jul-2010 10:05:04] But normaly on high load you need to investigate and check why.
[14-Jul-2010 10:05:12] fOrsberg is now known as forsberg
[14-Jul-2010 10:05:22] makes sense, my server is dual hyperthreaded xeons
[14-Jul-2010 10:05:24] hence 4.0 max
[14-Jul-2010 10:06:54] rmatte: odd, i only ran chet's suggestion on one of the RRD's, yet all are now working after the reboot
[14-Jul-2010 10:08:15] twm1010: if you posted the tinypic then your system would benefit from a faster disk-system. Most of the time it is waiting for I/O which is most likely your harddisk. (or try to use a different filesystem for your perf data with noatime and stuff like this)
[14-Jul-2010 10:08:33] lol weird
[14-Jul-2010 10:08:43] Yeah, I know I have an I/O bottleneck
[14-Jul-2010 10:08:50] guess it was something to do with zenperfsnmp
[14-Jul-2010 10:08:52] it was the first time they gave me a real server to run zenoss on
[14-Jul-2010 10:08:59] lol
[14-Jul-2010 10:09:11] hush, i run on wood!
[14-Jul-2010 10:09:13] i did a single RAID5
[14-Jul-2010 10:09:27] next time I'll do a mirror for the OS and a RAID10 for Zenoss
[14-Jul-2010 10:09:27] and there you named your problem.
[14-Jul-2010 10:10:16] It's not kiling me tho, and I did run through most of the suggestions for improving disk I/O, noatime, etc
[14-Jul-2010 10:10:57] the biggest performance boost came when I truncated the mysql database and set my retention at 7 days until I can get most of the events under control.
[14-Jul-2010 10:11:07] my windows machines are ridiculously chatty
[14-Jul-2010 10:11:46] * Simon4 is about to go to 7 day mysql also
[14-Jul-2010 10:11:55] some of our queries take >30 seconds, it's horrible
[14-Jul-2010 10:12:10] Simon4: I actually have it at 1 day right now.
[14-Jul-2010 10:12:35] Guest90611 is now known as bzed
[14-Jul-2010 10:12:37] cluther: do duplicate events actually get stored more than once, or is a counter in a record simply incremented?
[14-Jul-2010 10:13:25] twm1010: counter is incremented
[14-Jul-2010 10:13:48] and lastTime value is set
[14-Jul-2010 10:14:07] gotcha. what killed my mysql was a whole bunch of unique events that I was basically ignoring
[14-Jul-2010 10:14:10] bad OID type things
[14-Jul-2010 10:14:56] unclassified events, over and over
[14-Jul-2010 10:15:18] I think if I sat here for two hours classifying things, tomorrow, I could do it again
[14-Jul-2010 10:15:20] yeh, I always fix bad OIDs, debug messages, whatever
[14-Jul-2010 10:15:32] unclassified events need to be tuned to be dropped
[14-Jul-2010 10:15:36] if you don't care about them
[14-Jul-2010 10:15:43] don't do proper tuning and ye shall pay the price
[14-Jul-2010 10:15:44] :)
[14-Jul-2010 10:15:59] * Simon4 is currently working on dropping bad oid events, or at least pruning em after 60 min or something
[14-Jul-2010 10:16:00] well, hence the mysql truncate
[14-Jul-2010 10:16:27] I got a massive UI performance boost fixing most of the bad OID events with blank interface templates
[14-Jul-2010 10:16:39] well, bad oid events can be actually fixed by creating local copies of templates and removing those OIDs from them
[14-Jul-2010 10:16:40] the rest came when I dropped the mysql db, it actually runs well now
[14-Jul-2010 10:16:45] I assume they are probably interface related
[14-Jul-2010 10:16:57] yeah, most of them, or Linux devices that don't support the standard perf template
[14-Jul-2010 10:17:20] seems like Linux based appliances tend to use their own OID's, I guess host resources weren't good enough (yet the duplicate the same f'in stats)
[14-Jul-2010 10:17:36] if they don't "support" it it's because their snmp security settings are too locked down
[14-Jul-2010 10:17:42] you should have the server admins fix that
[14-Jul-2010 10:18:05] twm1010: best bet is to try and get MIBs from the vendors
[14-Jul-2010 10:18:06] well, i could see that, they use some proprietary locked down linux, and have to setup their own OID's for things that would normally be available
[14-Jul-2010 10:18:15] then load them into net-snmp as much as possible
[14-Jul-2010 10:18:35] lame
[14-Jul-2010 10:19:07] frozty_sa: It doesn't have anything to do with Mibs from the vendors, it's just the snmpd.conf file
[14-Jul-2010 10:19:13] they have paranoid settings enabled
[14-Jul-2010 10:19:20] hence why he can't gather any perf data
[14-Jul-2010 10:19:35] sure, but on most appliances, you can't access that file, so they provide some web gui or something
[14-Jul-2010 10:19:44] our borderware spam filtering appliances are like that
[14-Jul-2010 10:19:48] they're obviously linux
[14-Jul-2010 10:19:48] well, I'm jumping into the conversation here halfway, had assumed we were discussing proxied requests for some sort of app-server specific stuff
[14-Jul-2010 10:20:03] but the standard linux template doesn't work
[14-Jul-2010 10:20:07] for instance, our stuff all work that way
[14-Jul-2010 10:20:09] twm1010: erf
[14-Jul-2010 10:20:13] shooting offense!
[14-Jul-2010 10:20:21] they duplicated all the OID's for stuff like CPU, load, memory, etc.
[14-Jul-2010 10:20:28] under their own part of the tree
[14-Jul-2010 10:20:33] what the hell for, though?
[14-Jul-2010 10:20:34] ah
[14-Jul-2010 10:20:40] so... I suppose I can just copy the template and fix the OIDs
[14-Jul-2010 10:20:40] that's quite lame
[14-Jul-2010 10:20:46] yeh
[14-Jul-2010 10:20:50] but thats the kind of bad OID things I have, and havent' had time to go fix
[14-Jul-2010 10:28:46] forsberg is now known as fOrsberg
[14-Jul-2010 10:29:45] hi: if i go to community/zenpacks and click on the Microsoft Internet Information Services (IIS) link, I don't actually see the Zenpack. Is that specific IIS zenpack provided anywhere?
[14-Jul-2010 10:30:55] robo: that's listed under the enterprise zenpacks section
[14-Jul-2010 10:31:03] you can only get those when you actually buy zenoss enterprise
[14-Jul-2010 10:31:16] the only packs core users have access to are the community and core zenpacks
[14-Jul-2010 10:31:41] ah, gotcha rmatte
[14-Jul-2010 10:31:48] If you need IIS performance use this pack: docs/DOC-3400
[14-Jul-2010 10:31:56] Egor's WMI IIS6 Performance pack
[14-Jul-2010 10:32:05] cool; ty for the suggestion
[14-Jul-2010 10:32:08] np
[14-Jul-2010 10:34:28] fOrsberg is now known as forsberg
[14-Jul-2010 10:52:55] (you need the WMI Data Source) pack to be able to use the IIS6 pack
[14-Jul-2010 10:52:58] just a heads up
[14-Jul-2010 11:11:11] theacolyte: actually it's kind of fun.
[14-Jul-2010 11:11:33] theacolyte: we have an extension on our phone system that's "a room full of monkeys
[14-Jul-2010 11:12:00] theacolyte: actually it's kind of fun.
[14-Jul-2010 11:12:09] gah.
[14-Jul-2010 11:12:13] * Absenth is sorry :)
[14-Jul-2010 11:12:27] stay absent from absinth ;)
[14-Jul-2010 11:12:48] ^^ was in response to something many pagges up regarding "if you engage solarwinds, please god don't give them your phone number" or something to that extent.
[14-Jul-2010 11:13:03] * Absenth regularly sends vendors to "a room full of monkeys"
[14-Jul-2010 11:13:55] what will i hear there?
[14-Jul-2010 11:14:13] it sounds exactly like a small room, with about two dozen monkeys.
[14-Jul-2010 11:15:36] nice function. now if there would be an option so they couldn't hang up...
[14-Jul-2010 11:17:52] is there some other way to give username and password in http-get to zenoss than http://user:pass@host/path
[14-Jul-2010 11:19:10] i think my irc-bot (eggdrop) does not understand that format
[14-Jul-2010 11:26:07] Absenth: that is brilliant
[14-Jul-2010 11:26:20] what kind of phone system?
[14-Jul-2010 11:26:26] who, in his wisdom decided to exclude ext2 from a default kernel config? *gna*
[14-Jul-2010 11:26:48] :O isn't /boot usually ext2 ?
[14-Jul-2010 11:40:58] twm1010: yeh, it's supposed to be but doesn't have to be
[14-Jul-2010 11:42:25] the only advantage is that ext2 is ext3 without the journalling
[14-Jul-2010 11:42:41] and you only access /boot at boot time, so journalling is useless, waste of space
[14-Jul-2010 11:42:50] but it's not a huge deal, just best practice
[14-Jul-2010 11:43:14] and having an ext2 based ramdisk just blows up...
[14-Jul-2010 11:49:21] what i dont get is how fsck can take 12 hours on 2b ext3
[14-Jul-2010 11:49:29] i thought journaling was to prevent that
[14-Jul-2010 11:49:40] takes 2 seconds on ufs2 with journal on bsd
[14-Jul-2010 11:49:47] forsberg: it prevents it in case of a crash. not on a regular check
[14-Jul-2010 11:49:50] k
[14-Jul-2010 12:02:17] is that 2tb ext3?
[14-Jul-2010 12:02:25] 2b would take a millisecond :P
[14-Jul-2010 12:02:55] till the kernel makes an oops, because the journal did not fit inside the 2b ;)
[14-Jul-2010 12:04:39] yes sorry
[14-Jul-2010 12:04:40] 2tb :)
[14-Jul-2010 12:05:05] had 2 the other day, one took 4-5 hours and the other needed a second pass etc
[14-Jul-2010 12:05:08] took 12 hours
[14-Jul-2010 12:05:09] :(
[14-Jul-2010 12:05:36] perfect occasion to start my freebsd zfs rant, so it was atleast good for something
[14-Jul-2010 12:06:09] like having a license that makes zfs compatible with linux
[14-Jul-2010 12:12:26] twm1010: asterisk
[14-Jul-2010 12:12:35] twm1010: sorry for delay meetings, and lunch :)
[14-Jul-2010 12:16:26] hi everyone
[14-Jul-2010 12:18:07] that's too funny, I've been posting in the ubuntu forum and arrrghhh from in here responded
[14-Jul-2010 12:18:17] so I asked he was the same arrrghhh from #zenoss, turns out he is
[14-Jul-2010 12:18:43] I think I'm goin to have to file a bug with the Linux kernel/nfs developers
[14-Jul-2010 12:18:45] small world =P
[14-Jul-2010 12:18:53] since I've found an nfs bug that isn't distro specific
[14-Jul-2010 12:20:44] forsberg is now known as fOrsberg
[14-Jul-2010 14:03:11] quick question: can a check be run manually as you can in nagios?
[14-Jul-2010 14:03:31] ie, can i just run the script with the same format/arguments as shown in zenoss?
[14-Jul-2010 14:08:12] vlaurenz: it depends on what script you're talking about
[14-Jul-2010 14:10:30] openldap-repmon.py
[14-Jul-2010 14:10:48] don't know if it's custom or not
[14-Jul-2010 14:11:09] where are you seeing that script?
[14-Jul-2010 14:11:12] in some zenpack?
[14-Jul-2010 14:11:45] nm i think it's custom written here
[14-Jul-2010 14:11:52] it's in /opt/zenoss/scripts
[14-Jul-2010 14:12:19] rmatte: i'm _really_ new to zenoss -- nagios expert though
[14-Jul-2010 14:14:22] hmm
[14-Jul-2010 14:15:48] I've got a transform to do some conditional auto-acks on alerts, using dmd.ZenEventManager.manage_ackEvents(evid) in the transform
[14-Jul-2010 14:16:02] (ripped the format from how I was using it in the direct zendmd console)
[14-Jul-2010 14:16:27] but it doesn't appear to be working..is the dmd top-level object available in transforms?
[14-Jul-2010 14:16:44] frozty_sa: Starting in Zenoss 2.5 it is.
[14-Jul-2010 14:16:53] thanks rmatte
[14-Jul-2010 14:17:21] vlaurenz: usually if you've got something loaded as a zencommand, you can call it at any time you want to by testing it against a specific device
[14-Jul-2010 14:17:45] cluther: ah cool. so this is probably just some missing syntax or something then
[14-Jul-2010 14:19:27] Yeah, that syntax is correct. It's only a matter of arriving at the correct evid.
[14-Jul-2010 14:20:08] Wait. That param to ackEvents should be a list of evids. If you only have one, use [evid]
[14-Jul-2010 14:20:47] oh, wait..I think I found the problem
[14-Jul-2010 14:21:01] count vs evt.count
[14-Jul-2010 14:21:21] cluther: it accepts a singleton item
[14-Jul-2010 14:21:29] so forcing it as a list isn't necessary
[14-Jul-2010 14:21:34] frozty_sa: You're right. It wraps it for you.
[14-Jul-2010 14:21:43] or at least, it seemed to, when I was doing a quick check this afternoon
[14-Jul-2010 14:43:17] hi, i'm brand new to zenoss, and am creating a datacollector plugin that i would like to use, how do i go about registering it with zenoss so that it is used to pull info?
[14-Jul-2010 14:43:44] john---: if it's a standard shell command you're looking at creating a template with a command datasource
[14-Jul-2010 14:44:15] the admin or dev guide will show you how to do such things
[14-Jul-2010 14:44:45] i couldnt find what i'm trying to do maybe i'm not clued in enough yet
[14-Jul-2010 14:44:54] i have my .py written and it extends CommandPlugin
[14-Jul-2010 14:45:03] ah k
[14-Jul-2010 14:45:16] what docs do i need to read so that zenoss can load that and use it?
[14-Jul-2010 14:45:30] or do i just place it in a directory?
[14-Jul-2010 14:46:00] you've gone beyond my knowledge - but it should go in a zenpack as a custom datasource at a guess
[14-Jul-2010 14:46:13] where it gets registered, etc
[14-Jul-2010 14:46:20] ya
[14-Jul-2010 14:46:34] do you know what in the zenpacks i need to edit?
[14-Jul-2010 14:46:44] to load it
[14-Jul-2010 14:46:52] I don't, the zenpacks I've written don't go that far
[14-Jul-2010 14:47:02] (I don't work for zenoss)
[14-Jul-2010 14:47:12] ah =[
[14-Jul-2010 14:47:15] lol nor do i
[14-Jul-2010 14:47:34] you could check out zenpacks.zenoss.org and see if there's something you could copy
[14-Jul-2010 14:47:50] cool
[14-Jul-2010 14:47:50] will do
[14-Jul-2010 14:47:52] that or someone in here or on the forums should be able to help, I'm pretty sure it's been done before
[14-Jul-2010 14:47:59] * Simon4 is keen to learn that one also
[14-Jul-2010 14:48:41] john---: Put it under the modeler/plugins directory in your ZenPack somewhere.
[14-Jul-2010 14:48:49] is that all i need to do?
[14-Jul-2010 14:48:50] Make sure you have __init__.py files in all directories leading up to it and including it.
[14-Jul-2010 14:48:51] Yeah
[14-Jul-2010 14:48:55] cool
[14-Jul-2010 14:49:01] glad its that easy ^_^
[14-Jul-2010 14:49:02] ty
[14-Jul-2010 14:49:06] Then if you do a "zopectl restart" you'll find it in the list when you go to "Collector Plugins"
[14-Jul-2010 14:49:08] sweet, i knew someone would turn out :)
[14-Jul-2010 14:49:15] s/out/up
[14-Jul-2010 15:11:11] <|rt|> is there an easy way to surpress an alert on a threshhold that has an snmptrap alert associated with it?
[14-Jul-2010 15:14:13] |rt|: I don't quite get what you're getting at
[14-Jul-2010 15:14:22] you want to suppress a threshold IF a trap comes in?
[14-Jul-2010 15:14:33] <|rt|> yeah I may have my terminology off
[14-Jul-2010 15:14:49] can you maybe explain scenario-wise?
[14-Jul-2010 15:14:53] in more detail
[14-Jul-2010 15:15:33] <|rt|> what I'm trying to do say I'm monitoring a device via snmp polling and also using snmp traps
[14-Jul-2010 15:16:17] <|rt|> what's happening is the cpu is going beyond a threshhold so I'm getting an alert on both the snmp trap and from the performanc monitoring that zenoss is doing
[14-Jul-2010 15:17:24] so the device is configured to trap when cpu utilization rises to a certain level
[14-Jul-2010 15:17:25] ?
[14-Jul-2010 15:17:38] <|rt|> yes
[14-Jul-2010 15:17:50] why wouldn't you just outright disable the threshold in Zenoss then?
[14-Jul-2010 15:17:57] or disable the threshold trap on the device
[14-Jul-2010 15:18:12] you're overcomplicating things by having both and then trying to figure out how to auto-clear one of them
[14-Jul-2010 15:18:15] it just doesn't make sense
[14-Jul-2010 15:18:26] <|rt|> and then zenoss's performance monitoring throws the alert a few minutes later
[14-Jul-2010 15:18:43] not if you disable the threshold in Zenoss
[14-Jul-2010 15:18:55] <|rt|> how do you do that?
[14-Jul-2010 15:19:02] just go to the performance template
[14-Jul-2010 15:19:07] and set enabled to false on it
[14-Jul-2010 15:19:10] it's very easy to do
[14-Jul-2010 15:19:11] <|rt|> make a local template and change it?
[14-Jul-2010 15:19:24] correct
[14-Jul-2010 15:19:31] copy the template locally
[14-Jul-2010 15:19:35] then set it to disabled
[14-Jul-2010 15:19:45] problem solved
[14-Jul-2010 15:20:02] <|rt|> yeah now I just have to figure out if I want to do it that way or just disable that trap
[14-Jul-2010 15:20:13] I'd disable the trap personally
[14-Jul-2010 15:20:18] polling is much more effective
[14-Jul-2010 15:20:21] <|rt|> the nice thing with the perf template is that it clears itself when the situation corrects itself
[14-Jul-2010 15:20:29] exactly
[14-Jul-2010 15:20:32] <|rt|> which is normally in a few seconds
[14-Jul-2010 15:32:16] afternoon all
[14-Jul-2010 15:32:27] i have a quick python question
[14-Jul-2010 15:32:40] can someone look at the code in this script
[14-Jul-2010 15:32:42] docs/DOC-2484;jsessionid=FBFD65E63EA00451891256EAF290E31B.node0
[14-Jul-2010 15:33:48] it seems to return either 0 or 1, but not in the TAILS fashion like STATUS|DATAPOINT=VALUE
[14-Jul-2010 15:34:14] would i need to modify the return output to have the scripts output be attached to a datapoint?
[14-Jul-2010 15:34:55] err...i guess this is really a python/datapoint question
[14-Jul-2010 15:35:54] klinstifen: basically make the output match say a nagios format, then set the parser of the template to be nagios that way it would match the datapoints specified in the template
[14-Jul-2010 15:37:24] and this is the only way this will work correct?
[14-Jul-2010 15:37:33] there is no generic zenoss'esque datapoint name that i can use
[14-Jul-2010 15:37:40] that will just catch returns of 1 or 0?
[14-Jul-2010 15:38:08] klinstifen: you're either exiting with an exit code, or you're providing data to a datapoint
[14-Jul-2010 15:38:11] it's one or the other
[14-Jul-2010 15:38:40] exiting with an exit code instantly generates an alert
[14-Jul-2010 15:38:46] ah ok, i guess that was really the heart of my qeustion
[14-Jul-2010 15:39:05] If you need to provide data to a datapoint then you output in nagios format
[14-Jul-2010 15:39:11] right
[14-Jul-2010 15:39:33] so if i wanted to exit with a code AND get data then i would need two command datasources?
[14-Jul-2010 15:39:45] no
[14-Jul-2010 15:39:53] here's the just of it...
[14-Jul-2010 15:40:04] say you wanted to exit with a code AND provide a datasource
[14-Jul-2010 15:40:07] right
[14-Jul-2010 15:40:28] you would have the script exit with error code 1, which is a system thing, there's no actual output to stdout...
[14-Jul-2010 15:40:57] you would also have it output like: THIS IS MY ERROR MESSAGE|myDataPoint=1
[14-Jul-2010 15:41:15] the message displayed in the alert generated in Zenoss would be "THIS IS MY ERROR MESSAGE"
[14-Jul-2010 15:41:21] and the datapoint value would still be collected
[14-Jul-2010 15:41:50] a datapoint needs to be added with the exact same name (case sensitive) as what you're using in the script (in this case, myDataPoint) under the datasource for it to work
[14-Jul-2010 15:42:04] you can do that in the GUI
[14-Jul-2010 15:42:14] while editing the template
[14-Jul-2010 15:42:46] understand?
[14-Jul-2010 15:43:24] perfectly
[14-Jul-2010 15:43:36] i have created several command datasources that collected datapoints
[14-Jul-2010 15:43:44] but none that have also grabbed alert messages
[14-Jul-2010 15:43:51] k
[14-Jul-2010 15:44:09] and the script should always exit with error code 1, correct?
[14-Jul-2010 15:44:11] pretty sure the severity is based on the exit code too
[14-Jul-2010 15:44:16] I think 1 is warning, 2 is critical
[14-Jul-2010 15:44:23] but I've never actually done it myself
[14-Jul-2010 15:44:27] ok
[14-Jul-2010 15:44:32] i'll try it out
[14-Jul-2010 15:44:34] thanks for the clarifying
[14-Jul-2010 15:44:35] k
[14-Jul-2010 15:44:37] np
[14-Jul-2010 16:02:19] nooo im going to loose my internets!
[14-Jul-2010 16:02:44] oh well, worth it for a better router =)
[14-Jul-2010 16:09:36] I'm out for the day, cheers guys
[14-Jul-2010 17:00:47] fOrsberg is now known as forsberg
[14-Jul-2010 17:25:05] nickanderson: publishing the nginx zenpack right now
[14-Jul-2010 17:25:56] cool
[14-Jul-2010 17:38:33] nickanderson: docs/DOC-7441
[14-Jul-2010 17:38:44] feel free to update it however
[14-Jul-2010 17:43:24] lol
[14-Jul-2010 17:43:30] you put my weak screenshots in there
[14-Jul-2010 17:43:47] I'm not running nginx, better than the ones I'd come up with :)
[14-Jul-2010 17:43:59] Very cool.
[14-Jul-2010 17:44:30] nickanderson: I hope it's good, I added it to the list of Community ZenPacks we mentioned in the 3.0 press release
[14-Jul-2010 17:45:43] 3.0 is out?
[14-Jul-2010 17:46:34] i havent tested it in 3.0
[14-Jul-2010 17:46:59] nothing special but not anything special with the aggregate pdu one either and it was on the list for not working with 3.0
[14-Jul-2010 17:47:03] nickanderson: It'll work since it uses a COMMAND data source.
[14-Jul-2010 17:48:38] nickanderson: I didn't do the regression test myself, so hopefully I can find some cycles to recreate it
[14-Jul-2010 17:48:47] but 3.0 is about to drop
[14-Jul-2010 17:52:52] core?
[14-Jul-2010 17:58:24] yes
[14-Jul-2010 18:00:08] oo
[14-Jul-2010 18:00:25] how long approximately, I was just about to do a reinstall of 2.5.2 ;)
[14-Jul-2010 18:09:30] forsberg is now known as fOrsberg
[14-Jul-2010 18:38:58] mrayzenoss is now known as mray
[14-Jul-2010 18:39:33] nickanderson: did you update the fortigate zenpack?
[14-Jul-2010 18:40:43] i did
[14-Jul-2010 18:40:50] not to make it compatible with 3.0 though
[14-Jul-2010 18:41:18] did i do somthing wrong? i followed the tagging instructions
[14-Jul-2010 18:42:22] or thought i did
[14-Jul-2010 18:42:24] :/
[14-Jul-2010 18:42:26] that's fine, I'm just making a big list of zenpacks to list as updated
[14-Jul-2010 18:42:33] oh
[14-Jul-2010 18:42:38] egor updated about 15 for 3.0
[14-Jul-2010 18:42:44] so I was going through the tags
[14-Jul-2010 18:42:48] and saw the fortigate
[14-Jul-2010 18:42:56] yeah i just made some additions
[14-Jul-2010 18:43:08] I havent installed 3.0 yet
[14-Jul-2010 18:43:19] I'll try to do another regression sweep on the zenpacks soon
[14-Jul-2010 21:40:53] jassinpain_ is now known as jassinpain
[15-Jul-2010 00:00:30] [disconnected at Thu Jul 15 00:00:30 2010]
[15-Jul-2010 00:00:30] [connected at Thu Jul 15 00:00:30 2010]
[15-Jul-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]
[15-Jul-2010 01:20:54] fOrsberg is now known as forsberg
[15-Jul-2010 02:59:08] * frozty_sa makes an event mapping under /Cisco/WCCP called funnypacket
[15-Jul-2010 03:42:10] ah man, <3 the ability to do event transforms
[15-Jul-2010 03:42:11] so useful
[15-Jul-2010 06:25:01] except this thing is being odd. http://slexy.org/view/s21H6a1hgY is my transform. any ideas why it's not autoclosing the event?
[15-Jul-2010 06:52:08] frozty_sa: you sure your regexp matches
[15-Jul-2010 06:52:28] (and in line three you could leave out the test for None
[15-Jul-2010 06:55:30] frozty_sa: oh, and check the famous match vs search
[15-Jul-2010 06:55:56] i've had os processes on linux boxes that "line up" as far as regex goes
[15-Jul-2010 06:55:59] and it still didnt wrok
[15-Jul-2010 07:27:31] fragfutter: deadcertain. the eventclass gets changed and the severity is changed
[15-Jul-2010 07:27:42] just the close that doesn't happen
[15-Jul-2010 07:30:43] frozty_sa: could you close the event via zProperties?
[15-Jul-2010 07:33:37] frozty_sa: or evt._action = 'drop'
[15-Jul-2010 07:34:31] frozty_sa: i would think the event hasn't reacht the eventmanager yet, so you can't delete it.
[15-Jul-2010 07:34:32] what is the effect of evt._action = drop? just to close it?
[15-Jul-2010 07:34:41] mmm
[15-Jul-2010 07:34:57] frozty_sa: that would be history
[15-Jul-2010 07:35:10] okay, that could work
[15-Jul-2010 07:35:41] lemme see what it does
[15-Jul-2010 08:11:28] I'm using the MySQL ZenPack and under the Perf tab for a given host, all I get is broken images for all the graphs. Is there a log file some where I can check to see what's not working?
[15-Jul-2010 08:12:42] fragfutter: gah, that didn't work :/
[15-Jul-2010 08:12:47] _action, that is
[15-Jul-2010 08:37:04] good morning all
[15-Jul-2010 08:37:07] i know how to get LLDP Link information for a switch via SNMP. I would like to add this Information to the Interface Display of the switch. Where would i start?
[15-Jul-2010 08:37:10] morning
[15-Jul-2010 08:37:11] or afternoon/evening for some
[15-Jul-2010 08:37:19] afternoon ;)
[15-Jul-2010 08:41:29] well, fun day ahead, testing WMI pack, testing WMI pack with upgrade, cat scan at noon, then afternoon will be spent getting LDAP authentication working with 3.0
[15-Jul-2010 08:45:32] lol
[15-Jul-2010 08:45:37] horray catscan
[15-Jul-2010 08:47:06] lol
[15-Jul-2010 08:47:19] * Simon4 is auditing blades - joy
[15-Jul-2010 08:49:07] hehe
[15-Jul-2010 08:49:29] * rmatte is wondering whell the heck we're going to get our 4th blade
[15-Jul-2010 08:49:35] when*
[15-Jul-2010 08:50:00] * Simon4 looks guiltily at his half-finished bladechassis release document
[15-Jul-2010 08:50:06] I need more hours in my days :/
[15-Jul-2010 08:50:30] I need more days in my hours
[15-Jul-2010 08:51:16] In general is there a specific log file that captures the output of stuff run from ZenPacks?
[15-Jul-2010 08:51:43] ad4m: you mean the output that you see when installing a ZenPack?
[15-Jul-2010 08:52:46] ad4m: there's a logfile for each daemon. the log file to track actually day to day errors depends on what daemon the ZenPack uses
[15-Jul-2010 08:52:54] actual*
[15-Jul-2010 08:55:51] rmatte: do you have any idea what would cause zeneventlog to segfault on only the collector , the hub works fine, both running rpm install of 2.5.2 and same wh
[15-Jul-2010 08:55:54] python[24029] general protection rip:2ba49a2003aa rsp:7fffe5d9d110 error:0
[15-Jul-2010 08:56:24] tehhobbit: unfortunately I don't... I don't really use zeneventlog
[15-Jul-2010 08:56:37] tehhobbit: is there any traceback in the log file?
[15-Jul-2010 08:56:52] no
[15-Jul-2010 08:56:58] just segfaults
[15-Jul-2010 08:57:03] you're checking the log file on the collector itself right?
[15-Jul-2010 08:57:09] yes
[15-Jul-2010 08:58:41] seems to die when connecting to certain windows machines
[15-Jul-2010 08:59:05] hmmm
[15-Jul-2010 08:59:28] have you tried logging in to the collector and running the daemon in verbose?
[15-Jul-2010 08:59:41] zeneventlog -v10 run
[15-Jul-2010 08:59:52] or zeneventlog run -v10 --cycle
[15-Jul-2010 09:00:18] http://pastie.org/1045626 get some of these but not always at the same time as it dies
[15-Jul-2010 09:00:28] forsberg is now known as fOrsberg
[15-Jul-2010 09:00:58] (07/15 15:37:06) < rmatte> good morning all
[15-Jul-2010 09:01:00] liar :p
[15-Jul-2010 09:01:07] rmatte: I'm trying to find the output from a run of whatever is in the MySQL ZenPack. Something isn't work with it because all the perf graphs are broken, so I'm trying to figure out what's broken. Any ideas where to look for that sort of information?
[15-Jul-2010 09:01:28] frozty_sa: haha
[15-Jul-2010 09:02:20] ad4m: I've never seen that ZenPack, is it command datasource based, snmp datasource based, or something else entirely?
[15-Jul-2010 09:03:21] rmatte: So I'm very new to Zenoss and don't know all the terms yet, but I believe its a command datasource because it appears to run a Python script to obtain information from a given MySQL server.
[15-Jul-2010 09:03:36] tehhobbit: that's the smoking gun but I don't personally know what to make of it
[15-Jul-2010 09:03:58] tehhobbit: daemons definitely should not be crashing, log a trac ticket explaining the situation with that error message included
[15-Jul-2010 09:04:19] tehhobbit: probably just an oversight on the error handling
[15-Jul-2010 09:06:10] tehhobbit: also, have you tried removing that particular device over to the main collector to see if it ends up crashing on there as well?
[15-Jul-2010 09:06:18] removing == moving
[15-Jul-2010 09:06:19] rather
[15-Jul-2010 09:08:41] ad4m: then login to the server via ssh, become the Zenoss user, and do: zencommand run -v10 -d devicename
[15-Jul-2010 09:08:47] and check the output for errors
[15-Jul-2010 09:09:06] you may also want to try executing that script against the server by hand
[15-Jul-2010 09:09:27] or by clicking on the datasource in the template, adding a device name in to the test input field and clicking test
[15-Jul-2010 09:09:38] those are your main troubleshooting steps for that issue
[15-Jul-2010 09:10:02] to become zenoss user just do: sudo su - zenoss
[15-Jul-2010 09:11:58] ok, I'll give it a go, thanks
[15-Jul-2010 09:12:02] np
[15-Jul-2010 09:14:34] rmatte: no but gonna try now
[15-Jul-2010 09:14:51] k
[15-Jul-2010 09:20:57] rmatte: I'm running zencommand but it doesn't seem to be doing a whole lot. I tried it once with the device name in question, and then again with an invalid device name and both outputs were more or less the same.
[15-Jul-2010 09:21:23] Do you have to prefix the device name with anything? Like /Devices/Server/Linux?
[15-Jul-2010 09:21:25] ad4m: you used -v10?
[15-Jul-2010 09:21:33] because if not, yes they will be the same
[15-Jul-2010 09:21:51] and no, you don't prefix the device name with anything
[15-Jul-2010 09:21:55] Yes, I used -v10
[15-Jul-2010 09:22:02] try adding --cycle to the end
[15-Jul-2010 09:22:04] see if that works
[15-Jul-2010 09:22:28] ok
[15-Jul-2010 09:22:41] I don't see how the output can be the same with -v10
[15-Jul-2010 09:23:25] i want to add a new section to the OS tab of a device.
[15-Jul-2010 09:24:14] fragfutter: start reading the Zenoss developer's guide and have fun
[15-Jul-2010 09:24:58] fragfutter: there's no way to do it via the gui, there's coding involved.
[15-Jul-2010 09:25:16] so i create a zenpack, create a class (derived from OSComponent) that holds the information rows, add a relation to Products.ZenModel.OperatingSystem.
[15-Jul-2010 09:25:46] I don't know the exact steps since I've never done it.
[15-Jul-2010 09:25:56] fragfutter: Go to http://yourzenoss:8080/zport/portal_skins/zenmodel/deviceOsDetail/manage
[15-Jul-2010 09:26:12] Click customize, then you can edit the source of the OS tab.
[15-Jul-2010 09:26:37] Assuming you don't have Egor's advancedDeviceDetails pack installed. He uses a different OS tab.
[15-Jul-2010 09:26:39] cluther: if i want to have it in the zenpack, is there a metal-slot system?
[15-Jul-2010 09:27:17] check the Advanced Device Details pack as an example
[15-Jul-2010 09:27:30] Just drop a deviceOsDetail.pt into your ZenPack's skins directory if you want to override the core one. You may want to check out Egor's pack as his implements a slot based system that's much more pluggable.
[15-Jul-2010 09:27:37] Then you can just make your pack dependent upon his.
[15-Jul-2010 09:27:53] cluther: ok. i already have this pack. so i will look there
[15-Jul-2010 09:33:07] fragfutter: regarding that earlier thing, what's the way of doing it via zProperties?
[15-Jul-2010 09:33:54] actually, wait
[15-Jul-2010 09:34:14] in the event transform there is a zProperties tab where you can set zEventaction
[15-Jul-2010 09:34:19] morning
[15-Jul-2010 09:34:22] for example zport/dmd/Events/Perf/Interface/zPropertyEdit
[15-Jul-2010 09:34:47] sigh, silly mistakes again. 'evt._action = drop' vs 'evt._action = "drop"'...
[15-Jul-2010 09:34:52] %e_longweek
[15-Jul-2010 09:34:56] :P
[15-Jul-2010 09:34:59] one more day
[15-Jul-2010 09:34:59] lol
[15-Jul-2010 09:35:12] at least you didn't do evt.action
[15-Jul-2010 09:35:22] which is the most common mistake
[15-Jul-2010 09:35:37] does zenoss allow for some kind of role based management?
[15-Jul-2010 09:36:12] I am looking into zenoss to replace our current combination of nagios/cacti/smokeping/phpsyslog because those are just too many applications and everything is everywhere...
[15-Jul-2010 09:36:14] rmatte: earlier I was sitting staring at this one transform, going "why the fuuuuuuuuuuuuu aren't you working?! it runs perfectly in console dmd!!"
[15-Jul-2010 09:36:48] hehe
[15-Jul-2010 09:36:54] rmatte: up until the point I realized I'd taken my list-code from the console and pasted it into the transform. and in the console I refer to events as 'e' when I'm iterating through the list
[15-Jul-2010 09:37:06] ah
[15-Jul-2010 09:37:09] lol
[15-Jul-2010 09:37:49] I've been doing a lot of work earlier this week, so that it'd be okay when I'm effectively a zombie tomorrow
[15-Jul-2010 09:38:03] I feel like a zombie today
[15-Jul-2010 09:38:31] today is day 1 of quitting smoking
[15-Jul-2010 09:38:44] is tomorrow day 1 as well?
[15-Jul-2010 09:38:45] rmatte: good on ya :)
[15-Jul-2010 09:38:46] Good luck
[15-Jul-2010 09:38:58] frozty_sa: nope, tomorrow is day 2, smartass :P
[15-Jul-2010 09:38:59] I'm at almost 2 years
[15-Jul-2010 09:38:59] lol
[15-Jul-2010 09:39:07] rmatte: Patch/Gum?
[15-Jul-2010 09:39:08] * Simon4 stopped about 5 years ago, first week was horrible, it got much better very quickly after that
[15-Jul-2010 09:39:20] rmatte: there are some smokers at the office here, and some quitting smokers
[15-Jul-2010 09:39:29] and every so often the counter gets reset ;)
[15-Jul-2010 09:39:46] theacolyte: not yet, I've gotten myself down to about 2 to 3 smokes per day (extra light) over the past year and a half, going to try cold turkey, but if it gets too bad I'll go on the patch
[15-Jul-2010 09:39:52] ah
[15-Jul-2010 09:40:09] Yeah that shouldn't be that bad then... I was a pack of marlboros a day for 10+ years
[15-Jul-2010 09:40:23] jeez
[15-Jul-2010 09:40:29] I used chantix
[15-Jul-2010 09:40:30] I don't understand how you people can do that
[15-Jul-2010 09:40:32] And it really, really worked
[15-Jul-2010 09:40:40] "You people" as in a human being?
[15-Jul-2010 09:41:00] well, I've been smoking for 10 years now, and at one point I was smoking over a pack a day of export a green (aka the green death)
[15-Jul-2010 09:41:02] lol
[15-Jul-2010 09:41:10] export a
[15-Jul-2010 09:41:14] silly canadian ciggarettes
[15-Jul-2010 09:41:18] lol
[15-Jul-2010 09:41:25] friend of mine was on 1~2 packs of camel a day since the age of 15 or something..he finally had to start killing it last year, otherwise he'd be like an 85-year old man by the time he's 30
[15-Jul-2010 09:41:29] which would be 6 years from now
[15-Jul-2010 09:41:41] time to head into the orifice
[15-Jul-2010 09:41:43] le sigh
[15-Jul-2010 09:41:48] it's almost time to /leave/ here
[15-Jul-2010 09:42:19] hehe
[15-Jul-2010 09:44:09] * tehhobbit is thinking about heading home too
[15-Jul-2010 09:44:14] almost 5 pm
[15-Jul-2010 09:44:57] * Simon4 wonders where tehhobbit is
[15-Jul-2010 09:44:57] 15min 3s
[15-Jul-2010 09:45:25] Simon4: the not-the-slow-side of the world ;)
[15-Jul-2010 09:45:50] ah duh... europe
[15-Jul-2010 09:46:00] Sweden :P
[15-Jul-2010 09:46:04] for some reason my brain put one hour ahead of the UK somewhere in the atlantic
[15-Jul-2010 09:46:14] I'm in .za
[15-Jul-2010 09:47:15] .za == Zebra Alley?
[15-Jul-2010 09:47:21] * rmatte looks hopeful
[15-Jul-2010 09:47:29] soccer...
[15-Jul-2010 09:47:44] now try again
[15-Jul-2010 09:47:46] rmatte: lulz
[15-Jul-2010 09:48:05] *vuvvuvuvuvuvuvzuela*
[15-Jul-2010 09:48:26] We don't take kindly to your vuvuzela around these parts!
[15-Jul-2010 09:49:04] * frozty_sa owns vu.vuze.la
[15-Jul-2010 09:49:04] and sometime I'll do something useful with it
[15-Jul-2010 09:49:17] seriously?
[15-Jul-2010 09:49:29] that's okay, most of the world doesn't take kindly to america's "counter-terrorism" :P
[15-Jul-2010 09:49:35] rmatte: yip, seriously :)
[15-Jul-2010 09:49:36] rmatte: removing the device seemed to solve it, was a windowsmachine and tried wmi manually against it and was kinda funky then too
[15-Jul-2010 09:49:49] but will write a ticket to get correct errorhandling
[15-Jul-2010 09:49:51] tehhobbit: cool
[15-Jul-2010 09:49:53] frozty_sa: there is only one thing to do with that site
[15-Jul-2010 09:50:02] embedded vuvuzela sound on page load, very loud
[15-Jul-2010 09:50:19] ummm
[15-Jul-2010 09:50:20] Simon4: that's been done
[15-Jul-2010 09:50:28] did you see vuvuzela fm?
[15-Jul-2010 09:50:31] vuze.la is not registered
[15-Jul-2010 09:50:33] no
[15-Jul-2010 09:50:35] according to whois
[15-Jul-2010 09:50:49] * Simon4 could take a wild guess however
[15-Jul-2010 09:51:07] Created On:15-Jun-2010 15:38:07 UTC
[15-Jul-2010 09:51:36] fragfutter: yeah, that sounds about right
[15-Jul-2010 09:51:50] rmatte: some tld's aren't properly supported by whois
[15-Jul-2010 09:52:16] frozty_sa: vuvuze.la works fine in whois
[15-Jul-2010 09:52:27] *shrug*
[15-Jul-2010 09:52:33] so I doubt that's the issue hehe
[15-Jul-2010 09:52:34] whois vuze.la
[15-Jul-2010 09:52:42] works for me
[15-Jul-2010 09:52:52] I see info
[15-Jul-2010 09:52:54] Status:INACTIVE
[15-Jul-2010 09:52:57] meaning unregistered
[15-Jul-2010 09:53:04] rmatte: hmm, mine bitches about nameserver not found. which is fine, since I haven't added any DNS information
[15-Jul-2010 09:53:16] lol
[15-Jul-2010 09:53:26] I just didn't have time to do anything with it
[15-Jul-2010 09:53:26] oh maybe I'm reading this wrong
[15-Jul-2010 09:53:32] Expiration Date:15-Jun-2011 23:59:59 UTC
[15-Jul-2010 09:53:40] not sure what the status actually refers to then
[15-Jul-2010 09:53:53] meh
[15-Jul-2010 09:54:07] you would get a domain not found for unregistered domains.
[15-Jul-2010 09:54:25] I reckon I might use it as my consultant address or something
[15-Jul-2010 09:54:28] fragfutter: well, each whois server tends to have it's own funky way of showing that
[15-Jul-2010 09:54:33] so I wasn't sure
[15-Jul-2010 09:54:48] because I reckon it's not a word people can forget now :)
[15-Jul-2010 09:54:56] hehe
[15-Jul-2010 09:55:28] man, I should upload my pics off the final to smugmug still :/
[15-Jul-2010 09:55:33] my best domain is demonic.cc
[15-Jul-2010 09:55:36] * rmatte shrugs
[15-Jul-2010 09:55:51] if anyone here has any cloning machines, please let me know. I might volunteer for the beta program
[15-Jul-2010 09:55:58] rmatte: hey, I think I've seen that somewhere
[15-Jul-2010 09:55:59] lol
[15-Jul-2010 09:56:11] oh, the mibs thing?
[15-Jul-2010 09:56:15] frozty_sa: yeh, it's a tech blog
[15-Jul-2010 09:56:24] and yes, the Cisco Mibs guide was originally posted on there
[15-Jul-2010 09:56:27] yar, that's where I saw it
[15-Jul-2010 09:58:46] yar!
[15-Jul-2010 09:59:03] coming up on a year and a half since I started using Zenoss
[15-Jul-2010 09:59:26] doesn't seem that long
[15-Jul-2010 10:01:31] uhoh
[15-Jul-2010 10:01:53] lots of devs
[15-Jul-2010 10:02:01] what's the occasion?
[15-Jul-2010 10:02:24] we've decided to stop fixing bugs and just hang out here instead
[15-Jul-2010 10:02:35] mrayzenoss is now known as mray
[15-Jul-2010 10:02:37] Interesting approach
[15-Jul-2010 10:02:55] Yeah, it's a revolutionary new software development paradigm
[15-Jul-2010 10:03:02] better code through conversation
[15-Jul-2010 10:03:24] :)
[15-Jul-2010 10:04:30] oh, just noticed the RC1 is out for 3.0, time to have some fun
[15-Jul-2010 10:06:41] rmatte: a little birdy told me if you look in the right places you'll find a newer version
[15-Jul-2010 10:06:53] oh reallllly
[15-Jul-2010 10:07:31] * fragfutter marks the course "101 how to excite geeks" as completed
[15-Jul-2010 10:08:12] http://dev.zenoss.org/deb/
[15-Jul-2010 10:08:15] It's next to the lucky charms, at the end of the rainbow..
[15-Jul-2010 10:08:39] http://dev.zenoss.org/yum/
[15-Jul-2010 10:08:50] I use plain stack :)
[15-Jul-2010 10:09:23] it's in stable so it is final?
[15-Jul-2010 10:10:03] mray: where are the .bins located?
[15-Jul-2010 10:10:49] Hi everyone, I'm having this "problem": http://pastebin.org/397533
[15-Jul-2010 10:11:06] this url returns a json response
[15-Jul-2010 10:11:15] rmatte: on sourceforge, which hasn't gone public yet
[15-Jul-2010 10:11:20] patience :)
[15-Jul-2010 10:11:22] mray: ah
[15-Jul-2010 10:11:38] I don't know why it says me "Invalid HTTP response received from host: "
[15-Jul-2010 10:11:56] is it normal if the response is json?
[15-Jul-2010 10:12:01] hi, can anyone confirm this error? message/50562#50562
[15-Jul-2010 10:12:18] I mean, the content doesn't have html structure
[15-Jul-2010 10:12:53] bigegor: not yet, but I'm going to be testing the RC1 with the ZenPack today
[15-Jul-2010 10:12:55] chachan: sure you aren't getting a 500 or something? that'll come back with html
[15-Jul-2010 10:13:03] chachan: you miss the /
[15-Jul-2010 10:13:13] hmm - paul wasn't that error egor posted happening with the 1.3.10 version of the wmi externallib?
[15-Jul-2010 10:13:21] iancmcc, I'm sure
[15-Jul-2010 10:13:23] fragfutter, O.o
[15-Jul-2010 10:13:43] * pmcguire checking my wmi notes...
[15-Jul-2010 10:13:44] :)
[15-Jul-2010 10:13:56] that is it
[15-Jul-2010 10:14:07] fragfutter, iancmcc, thanks
[15-Jul-2010 10:14:47] caused by duplicate defn's of IEnumWbemClassObject in both Query.py and Wbem.py
[15-Jul-2010 10:15:11] yeah, wonder if RC1 still had that version of the library & hadn't picked up 1.3.11 yet
[15-Jul-2010 10:15:43] so i have a zenpack, that contains a class derived from OSComponent, it defines a relation to os. What do i need to do, so that zenoss picks this relation up?
[15-Jul-2010 10:16:14] i can update my ZenPack, but i can not test it (I have no Windows PC here)
[15-Jul-2010 10:16:21] i thought i would simply retart zope and it would show it under /zport/dmd/Devices/Server/Linux/devices/device/os
[15-Jul-2010 10:16:40] bigegor: I have to leave for a cat scan shortly, but when I get back I can test it for you
[15-Jul-2010 10:17:28] bigegor: we're apparently moving to WMI monitoring from SNMP soon, so I need to do some pretty heavy testing
[15-Jul-2010 10:18:51] I also still need to work on a guide for ldap authentication in 3.0, as soon as I actually figure out how to use the newer versions of the LDAP plugin
[15-Jul-2010 10:19:18] nightmare so far
[15-Jul-2010 10:19:32] isn't zenoss levering users from zope?
[15-Jul-2010 10:19:45] yeh, and the ldap plugin is a zope plugin
[15-Jul-2010 10:20:09] but the current guide doesn't work for the newer version that's required
[15-Jul-2010 10:20:17] hehe. then i had it a few years ago
[15-Jul-2010 10:20:24] so I need to monkey around with it until I figure it out
[15-Jul-2010 10:20:28] then write up a new guide
[15-Jul-2010 10:20:40] steal it from plone ;)
[15-Jul-2010 10:20:49] lol
[15-Jul-2010 10:20:56] that might be an idea
[15-Jul-2010 10:21:24] WMIDataSource 2.6.1 available from SVN (need testing against Zenoss 3.0 RC1 and Zenoss 2.5.2)
[15-Jul-2010 10:21:55] bigegor: http://dev.zenoss.com/trac/changeset?new=20861%40trunk%2Fwmi%2Fpysamba%2Fwbem&old=20779%40trunk%2Fwmi%2Fpysamba%2Fwbem
[15-Jul-2010 10:22:15] bigegor: this shows the changeset that fixes that particular problem.
[15-Jul-2010 10:27:40] pmcguire: i've simply moved defn of IEnumWbemClassObject right before wbem.py import, it must work for Zenoss 2.5.2 too.
[15-Jul-2010 10:29:41] bigegor: just be sure that there are *no* duplicates or redefinitions of that name (or com_context, either)
[15-Jul-2010 10:29:51] i have a zenpack class like this http://fpaste.org/9MEz/, what will i have to do that zenoss recognizes the relationship and shows it in the zope management interface?
[15-Jul-2010 10:33:37] pmcguire: http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.WMIDataSource/ZenPacks/community/WMIDataSource/Query.py
[15-Jul-2010 10:37:09] bigegor: Why are lines 18-19 necessary? IEnumWhatever and com_context are already defined in pysamba.wbem.wbem
[15-Jul-2010 10:38:00] why muck around with patches and just grab the newer build of the library?
[15-Jul-2010 10:38:19] pmcguire: but not for Zenoss 2.5.2
[15-Jul-2010 10:41:55] bigegor: Chad is right, I backported these to the 2.5.x branch
[15-Jul-2010 10:42:28] just make sure you're using the 1.3.11 version of the externalib not 1.3.10, 1.3.10 is bd
[15-Jul-2010 10:42:29] er bad
[15-Jul-2010 10:45:46] I blame the new guy...
[15-Jul-2010 10:46:08] (oh wait, *I'm* the new guy!)
[15-Jul-2010 10:46:49] fOrsberg is now known as forsberg
[15-Jul-2010 10:53:20] it looks like i have to define a relationship on both ends (in my case OperatingSystem and MyClass). How would i define the OperatingSystem end in a zenpack. I now edited the file, but this cant be the solution for a ZenPack
[15-Jul-2010 10:54:16] pmcguire, cgibbons: Zenoss 2.5.2 included pysamba 1.3.6, last WMIDataSource ZenPack is backward compatible with Zenoss 2.5.2
[15-Jul-2010 10:56:53] fragfutter: lines 80-82 in http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.RDBMS/ZenPacks/community/RDBMS/__init__.py
[15-Jul-2010 10:57:39] bigegor: thanks.
[15-Jul-2010 10:58:08] fragfutter: np
[15-Jul-2010 10:58:19] o
[15-Jul-2010 10:58:28] mducharme_ is now known as mducharme-laptop
[15-Jul-2010 11:03:36] bigegor: the 2.5.x SVN branch has the latest wmi tarball - http://dev.zenoss.com/trac/browser/branches/zenoss-2.5.x/inst/externallibs/wmi-1.3.11.tar.bz2
[15-Jul-2010 11:04:20] but you're right, the version shipped with 2.5.2 was a much earlier version (before my time)
[15-Jul-2010 11:04:43] wohoo. i have a new section in the OS tab. Now i need to write the modeller ;)
[15-Jul-2010 11:04:47] paul/egor: can we back up a minute? if we're talking about the forum post that egor posted earlier, that problem is just an issue with the library version in RC1, and will go away with a newer RC build or final. is there another issue we're talking about, too?
[15-Jul-2010 11:05:20] cgibbons: bigegor has his own flavor of Query.py in his custom zenpack
[15-Jul-2010 11:05:23] fragfutter: You do know the OS tab is gone in 3.0, right?
[15-Jul-2010 11:05:25] fwiw the 3.0.0 final is in the yum and apt repos
[15-Jul-2010 11:05:54] iancmcc: no.
[15-Jul-2010 11:05:57] iancmcc: *grmbl*
[15-Jul-2010 11:06:58] iancmcc: what will i get instead?
[15-Jul-2010 11:07:10] fragfutter: it's cool, you can reuse most of it, but you'll have to plug it in differently
[15-Jul-2010 11:08:12] iancmcc: ok. it has been complicated enough to plug it into the os tab, but i need to prove that (what i want to do) works anyway.
[15-Jul-2010 11:08:32] Looking at your changes, there is no harm in having the lines 18-19 as you posted, bigegor. Pre-1.3.11, you will define new names that are not redefined elsewhere; post 1.3.11 the names get defined, but will never be used, as pysamba.wbem.wbem will redefine them, and the wbem names will be the ones used. Just take care to NEVER PUT ANY REFS to those names before "from pysamba.wbem.wbem import *"
[15-Jul-2010 11:08:52] fragfutter: You'll likely make it a tab of its own. Unless it's a custom component, in which case it'll just show up on its own
[15-Jul-2010 11:09:20] iancmcc: actually it would be an extension of the IpInterface. But one step at a time
[15-Jul-2010 11:09:38] That's ok, if it has its own meta_type it'll show up separately. Yeah, when you get there...
[15-Jul-2010 11:10:21] fragfutter: see the components tree in the left pane here: http://public-demo.zenoss.com/zport/dmd/Devices/Server/Linux/devices/public-demo.zenoss.loc/devicedetail#deviceDetailNav:device_overview
[15-Jul-2010 11:10:31] auth is admin/zenoss
[15-Jul-2010 11:11:54] pmcguire: do you mean after?
[15-Jul-2010 11:12:48] bigegor: after is okay, because then you will be using the names as defined in pysamba.wbem.wbem
[15-Jul-2010 11:13:51] alternatively, you could follow the wbem import with a test for the existence of com_context and/or IEnumWbemClassObject, and only define them if they were not already imported from the wbem module
[15-Jul-2010 11:15:33] The thing is, if those names are defined in the wbem module, you *must* use the names as defined in that module
[15-Jul-2010 11:17:37] and not redefine them
[15-Jul-2010 11:18:15] pmcguire: i've reimplement Query.py because i need '_PATH' property of instance, is it posible to get this value in another way?
[15-Jul-2010 11:21:54] (looking at bigegor's Query.py again to see just he means...)
[15-Jul-2010 11:24:09] pmcguire: def wbemInstanceWithQualifiersToPython(obj)
[15-Jul-2010 11:26:40] pmcguire: __path returned in form root/cimv2:Win32_Processor.DeviceID='CPU1' if property DeviceID has Qualifier 'key' with value True.
[15-Jul-2010 11:31:47] bigegor: ok, I'm looking at the method that you added
[15-Jul-2010 11:36:40] I don't see a way around it - you can't do this in the caller, because by the time the data is returned, the underlying objects have been talloc_free'd, so the data is not there any more
[15-Jul-2010 11:38:54] i created a snmp modeller plugin, restartet zenoss, surfed to my device, added the plugin to the collector plugins, saved it. restartet zenoss. then run zenmodeler from CLI. it does not list my plugin and does not process it.
[15-Jul-2010 11:39:04] pmcguire: ok
[15-Jul-2010 11:40:22] if anybody know how to implement streaming output in ZenPack (like device modeling output window) with new Zenoss UI
[15-Jul-2010 11:47:32] fragfutter: check zenoss logs for errors (zenhub.log)
[15-Jul-2010 11:52:07] bigegor: thanks again ;)
[15-Jul-2010 11:52:31] how can i tell zenhub to reload the python script (there is a typo) without restarting everything?
[15-Jul-2010 11:54:08] never mind. restarting only zenhub is sufficient. fast enough for testing.
[15-Jul-2010 11:54:10] fragfutter: You can't. You need to restart zenhub and zenmodeler.
[15-Jul-2010 11:54:38] As long as you only make updates to the process method of your modeler plugin you only have to restart zenmodeler.
[15-Jul-2010 11:56:37] 3.0 is on the download pages
[15-Jul-2010 11:57:45] yippie
[15-Jul-2010 11:57:53] when is 3.0ent dropping?
[15-Jul-2010 11:58:37] jb: probably by the end of the month?
[15-Jul-2010 11:58:41] cool
[15-Jul-2010 12:02:20] bigegor: Your wmidatasource work with 3.0? :)
[15-Jul-2010 12:03:18] theacolyte: can you test SVN version (2.6.1)?
[15-Jul-2010 12:03:39] Not withotu a lot of work
[15-Jul-2010 12:10:44] LLDP found a port 0.9.1: {'sys': 'zj034', 'desc': 'GigabitEthernet16'} this is localPort {'desc': '9'}
[15-Jul-2010 12:10:47] yeah!
[15-Jul-2010 12:11:39] now how do i create the relation Objects with this information?
[15-Jul-2010 12:19:45] i'm real new to zenoss and trying to understand how it all fits together....how do data collector plugins map to things seen in the web ui? are there definitions for object map values to models stored anywhere?
[15-Jul-2010 12:24:29] nm i think i've figured out what i needed
[15-Jul-2010 12:54:33] bleh, I had to get an IV in my arm for the cat scan
[15-Jul-2010 12:54:36] * rmatte hates needles
[15-Jul-2010 12:54:46] glad it's over with
[15-Jul-2010 12:55:29] john---: it's pretty complex, but it's all built around Zope
[15-Jul-2010 13:14:13] anyone using the nagios checkping zenpack?
[15-Jul-2010 14:39:38] is there a way to check or log file to see if the MySQL zenpack is working properly ?
[15-Jul-2010 14:40:11] dec3pti01: no, it won't be in the log file
[15-Jul-2010 14:40:12] I still have no perf stats ... I'm able to login to the mysql box using the credentials I setup from the zenoss system
[15-Jul-2010 14:40:29] you need to test the command, and run the daemon by hand like zencommand run -v10
[15-Jul-2010 14:40:48] go to the datasource and run it as a test against a device from the GUI
[15-Jul-2010 14:41:10] try running the script against the device by hand from the commandline as the zenoss user
[15-Jul-2010 14:41:13] make sure it completes
[15-Jul-2010 14:41:23] it's basic troubleshooting
[15-Jul-2010 14:45:34] rmatte: forgot about using that ... checking it out now thanks
[15-Jul-2010 14:46:04] np
[15-Jul-2010 14:52:32] hmm I followed exactly what's on doc5634 for mysql monitoring and the template was added to this device but zentestcommand says
[15-Jul-2010 14:52:44] No COMMAND or SNMP datasource mysql applies to device C2-11_cdn
[15-Jul-2010 14:55:46] I ran zencommand as well but nothing about mysql on the output
[15-Jul-2010 14:58:55] mrayzenoss is now known as mray
[15-Jul-2010 14:59:30] the check_mysql_stats.py script under ZenPacks.zenoss.MySqlMonitor-2.1.1-py2.4.egg libexec works just fine
[15-Jul-2010 15:02:14] mray: bonjour
[15-Jul-2010 15:02:25] indeed
[15-Jul-2010 15:06:20] does it matter that the check_mysql_stats.py is not under the $ZENHOME/libexec directory ?
[15-Jul-2010 15:06:34] well, where is the template set to execute it from?
[15-Jul-2010 15:06:56] scripts are usually stored in the ZenPack's directory itself and dynamically referenced using a variable
[15-Jul-2010 15:07:16] the stuff in $ZENHOME/libexec is just what's included with Zenoss upon install
[15-Jul-2010 15:10:39] rmatte: "where is the template set to execute it from ?" where do I see that ? I don't see any zProperties option for the template itself
[15-Jul-2010 15:11:42] dec3pti01: you don't know how to get to the templates section of a class?
[15-Jul-2010 15:13:17] dec3pti01: can I suggest reading the Zenoss Admin manual please
[15-Jul-2010 15:13:59] rmatte: i have dude I just haven't messed with it in a while but I'm there now
[15-Jul-2010 15:14:05] k
[15-Jul-2010 15:14:12] find the MySQL template
[15-Jul-2010 15:14:15] go in to it
[15-Jul-2010 15:14:20] you'll see the datasource
[15-Jul-2010 15:14:30] what's the exact command it's executing?
[15-Jul-2010 15:17:42] it doesn't show .. all it has is ${here/manageIp} for the source and MySqlMonitor for the SourceType
[15-Jul-2010 15:18:21] unlike the one for example that I setup to monitor slave replication in which I created the template and the the sourcetype as command and the actuall command to run in the source
[15-Jul-2010 15:18:44] oh
[15-Jul-2010 15:18:50] so it's using a special datasource type
[15-Jul-2010 15:18:56] yep
[15-Jul-2010 15:19:02] did it install it's own daemon?
[15-Jul-2010 15:19:09] something like zenperfmysql or something?
[15-Jul-2010 15:20:25] nope
[15-Jul-2010 15:20:28] also, it may have just installed a collector plugin and you don't have it bound to the class
[15-Jul-2010 15:20:38] check the plugins list
[15-Jul-2010 15:22:20] can you point me to that ZenPack's zenpack page?
[15-Jul-2010 15:22:28] so that I have a better idea of what we're dealing with here
[15-Jul-2010 15:23:51] docs/DOC-5634 chap12 and docs/DOC-3501
[15-Jul-2010 15:25:22] maybe this zenpack install is broken ... I don't see any mysql plugins available fro the zCollectoPlugins
[15-Jul-2010 15:26:39] dec3pti01: it has no plugins
[15-Jul-2010 15:26:48] bigegor_ is now known as bigegor
[15-Jul-2010 15:31:26] did you restart zenoss after installing the pack?
[15-Jul-2010 15:31:32] if not, that's almost certainly your problem
[15-Jul-2010 15:32:08] dec3pti01: ZenPacks.zenoss.MySqlMonitor is Zenoss core zenpack. It provide MySqlMonitor data source with hardcoded command line and Data Points and MySQL performance template
[15-Jul-2010 15:33:28] yeh, he installed it and there's no data on the graphs
[15-Jul-2010 15:33:36] I'm thinking he installed it but didn't restart Zenoss
[15-Jul-2010 15:33:45] we'll find out as soon as he responds
[15-Jul-2010 15:33:47] :)
[15-Jul-2010 15:34:12] dec3pti01: you did follow the steps on the ZenPack page about adding the zenoss user as well I hope?
[15-Jul-2010 15:34:30] adding the zenoss user to the MySQL server I mean
[15-Jul-2010 15:35:00] the other thing that pops out at me is...
[15-Jul-2010 15:35:01] Note
[15-Jul-2010 15:35:01] Pay particular attention to the MySQL Version 5+ setting in the data source. If you are monitoring pre-v5 installations of MySQL, then you must set this value to False. If you are monitoring pre-v5 and v5+ installations, then create two templates: one for MySQL installations earlier than v5 and another for those after.
[15-Jul-2010 15:35:12] verify the version and make sure your setting is right
[15-Jul-2010 15:35:18] I've gotta go for the day
[15-Jul-2010 15:35:29] I'll be back tomorrow
[15-Jul-2010 15:35:33] dec3pti01: newest version of core MySqlMonitor zenpack is not testable (doesn't support Test of data source)
[15-Jul-2010 15:36:22] yes I have created it the mysql user and that works fine since running the command found under libexec for the zenpack worked just fine
[15-Jul-2010 15:36:40] I have nto restarted zenoss since the section on the doc did not mentioned so
[15-Jul-2010 15:36:48] yeh, restart it
[15-Jul-2010 15:36:51] that's necessary
[15-Jul-2010 15:39:13] bigegor_ is now known as bigegor
[15-Jul-2010 15:45:36] thanks everyone
[15-Jul-2010 15:45:47] sorry for the headache
[15-Jul-2010 15:56:59] Hmm, anyone here know of a decent web crawler/link checker (not really zenoss related, but still)
[15-Jul-2010 17:31:03] where the hell do you go in v3 to set class to a device
[15-Jul-2010 17:32:28] woot!
[15-Jul-2010 17:33:11] you mean in discovery or later when you want to reclassify it?
[15-Jul-2010 17:33:37] reclassify
[15-Jul-2010 17:33:53] i added a device and thought i set a class and i guess i didnt.
[15-Jul-2010 17:34:21] couple of ways at least. you can drag & drop it into the new class organizer with the left-side nav. of if you're on the device screen, glick the gear menu at the bottom and you'll see a change class in there.
[15-Jul-2010 17:34:46] oh crap
[15-Jul-2010 17:34:52] i forgot about the click and drag.
[15-Jul-2010 17:35:00] all that beta testing did me no good
[15-Jul-2010 17:37:58] thanks for the refresher
[15-Jul-2010 18:37:37] im trying to add a ping template to my v3 install but when I hadd the command datasource I cant add any datapoints. where do i do that at?
[15-Jul-2010 18:38:35] nevermind
[15-Jul-2010 19:14:10] Anyone still awake in here ?
[15-Jul-2010 21:12:02] is anyone here aware of any tricks for getting a little more perf/event info from an esxi host?
[15-Jul-2010 21:12:40] esxi4 that is
[15-Jul-2010 21:24:14] yyes
[15-Jul-2010 21:24:20] oh hi, this is #zenoss
[15-Jul-2010 21:24:38] vmware has their own snmp version, which is pretty awful
[15-Jul-2010 21:24:55] if you don't care about getting into zenoss you can use the rcli version esxtop
[15-Jul-2010 21:34:14] mota:
[15-Jul-2010 23:50:26] theacolyte: I am not at all sure what you're saying there
[15-Jul-2010 23:51:38] I did manage to enable really really basic snmp (it just says yeah, this is vmware) by editing /etc/vmware/snmp.xml and then doing /sbin/services.sh restart
[15-Jul-2010 23:53:14] but I'm still not getting the basics into zenoss. you know, basic perf numbers like cpu/ram/nic utilization, or how full the disks/stores are
[15-Jul-2010 23:54:35] so are you saying there's some secret rcli sauce that does this?
[16-Jul-2010 00:00:30] [disconnected at Fri Jul 16 00:00:30 2010]
[16-Jul-2010 00:00:31] [connected at Fri Jul 16 00:00:31 2010]
[16-Jul-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]
[16-Jul-2010 00:15:36] mota
[16-Jul-2010 00:15:54] you know that vmware makes snmp stuff for ESX/ESXi right?
[16-Jul-2010 00:17:02] I know that what I need is a no-cost way to get ESXi to give Zenoss more info to work with. Are you saying that's available?
[16-Jul-2010 00:17:11] one sec
[16-Jul-2010 00:17:15] lemme verify
[16-Jul-2010 00:18:16] MIB modules for use with ESX/ESXi and vCenter. These MIB modules support all ESX, ESXi and vCenter product releases through version 4.0 and 4.1.
[16-Jul-2010 00:20:17] k i downloaded them
[16-Jul-2010 00:20:21] i have a zip file full of MIBs
[16-Jul-2010 00:22:45] * Parabola shrugs
[16-Jul-2010 00:22:51] i was signed in when i got them tho
[16-Jul-2010 03:52:33] is it realy necessary to have a big registration form (and a small skip this link) in front of the download.
[16-Jul-2010 03:53:11] gotta harvest email addresses somehow ;)
[16-Jul-2010 03:53:23] *bah*
[16-Jul-2010 03:53:46] yeah, it frustrates me also
[16-Jul-2010 03:55:01] make it nice, give a real incentive to register and you get addresses that are realy interrested. Otherwise you end up with bullshit, or users that will be angry if you mail them.
[16-Jul-2010 03:55:08] marketing should learn...
[16-Jul-2010 04:07:10] buuuuuug
[16-Jul-2010 04:07:35] i can't select multiple events in the console to ack them at once.
[16-Jul-2010 04:08:23] * Simon4 really should do something about the 12,000 debug events :/
[16-Jul-2010 04:14:16] :( the handy search box that was always visible top right is gone.
[16-Jul-2010 04:14:46] on v3?
[16-Jul-2010 04:14:52] yes
[16-Jul-2010 04:15:42] hmm, that's going to suck
[16-Jul-2010 04:15:47] now it's click infrastructure and the search box there seems to be buggy
[16-Jul-2010 04:15:48] * Simon4 uses that _all_ the time
[16-Jul-2010 04:16:13] ajax everywhere :(
[16-Jul-2010 04:20:15] it's nice that they allow you to skip signup; a lot of sites do not
[16-Jul-2010 04:46:13] fragfutter: ctrl+slect doesn't work?
[16-Jul-2010 04:46:16] select*
[16-Jul-2010 04:55:56] not initialy, after zenoss restart it works. strange
[16-Jul-2010 05:17:07] in zenoss3, how do i specify the columns that are displayed for my custom meta_type ?
[16-Jul-2010 05:17:23] strange that a restart fixed the problem ? ;>
[16-Jul-2010 05:17:30]
[16-Jul-2010 05:18:02] forsberg: there's an embedded windows vm in there somewhere :>
[16-Jul-2010 05:18:12] forsberg: maybee i was to fast and the upgrade was still running. Zope comes up but migration scripts are still running.
[16-Jul-2010 06:27:49] question, can one setup a alerting rule for "ON DUTY"
[16-Jul-2010 06:28:14] that's only active from 16 > 8 next day + from friday 16 to sunday ?
[16-Jul-2010 06:28:27] i got it working in the week days but not the weekends
[16-Jul-2010 06:32:34] zykes-: zenos 3 or 2?
[16-Jul-2010 06:33:20] create a rule that is active every day from 16:00-8:00
[16-Jul-2010 06:33:39] create another rule that is active on Saturday 8:00-16:00
[16-Jul-2010 06:33:44] and a third for sunday
[16-Jul-2010 06:50:46] the ajax pages are missing a "loading" indicator.
[16-Jul-2010 07:05:50] anyone awake ?
[16-Jul-2010 07:13:42] SNMPv2-SMI::enterprises.2.3.51.2.2.1.1.2.0 = STRING: "38.00 Centigrade"
[16-Jul-2010 07:15:28] so uhm, what's the difference between evt._action = "drop" and evt._action = "history"?
[16-Jul-2010 07:15:40] I remember seeing it somewhere, but I can't find the bloody page today
[16-Jul-2010 07:16:23] my actual question is "does drop delete the event entirely? and does history just close it?"
[16-Jul-2010 07:19:05] frozty_sa: at a wild guess, drop = totally gone, history = just send straight to the history without turning up as a current event, so yeah, closed
[16-Jul-2010 07:22:08] yeah, I think history is what I actually want here
[16-Jul-2010 07:28:13] is there a way to setup a global default dashboard in the core eddition…that will still allow users to customize if desired ?
[16-Jul-2010 07:32:53] anyone have experience with building graph from a reply like this ?
[16-Jul-2010 07:32:53] SNMPv2-SMI::enterprises.2.3.51.2.2.10.1.1.1.8.1 = STRING: "390W"
[16-Jul-2010 07:34:15] im only used to doing from straight numbers, and its not working as i expected :)
[16-Jul-2010 07:48:14] how can custom components control their display in the new zenoss v3 UI?
[16-Jul-2010 07:48:33] with their own custom componentpanel javascript
[16-Jul-2010 07:49:25] uh? not something with IComponent?
[16-Jul-2010 07:50:19] It should be doc'ed in the ZenPack conversion guide, but in short you'd make a new Info object and interface to represent the component, and then a new extension of ComponentGridPanel in JavaScript that is registered for that component.
[16-Jul-2010 07:51:53] on a related note, we just added something in trunk (didnt make it in for 3.0) that lets you register friendly names for the component meta type, so you can control how it is presented in the navigation grid on the left. some of the meta names are a bit... ugly
[16-Jul-2010 07:52:23] where can i find the conversion guide?
[16-Jul-2010 07:52:34] hurm now that is a good question, moment
[16-Jul-2010 07:52:55] i've done a bunch of component UI customization recently for an enterprise zenpack, so at least I can help with that one :)
[16-Jul-2010 07:53:28] cgibbons: so don't leave ;) i want to have a basic info display of my data today!
[16-Jul-2010 07:53:34] ah here we go message/50142
[16-Jul-2010 07:53:51] section 1.7
[16-Jul-2010 07:58:27] cgibbons: you don't happen to have a small zenpack with an example? (info.py, interface.py, configure.zcml=?
[16-Jul-2010 07:59:19] no, although I could swizzled one up. but it's probably just as fast to do one interactively with you. bigegor has converted his though (not sure which ones) to use that format.
[16-Jul-2010 08:00:20] forsberg is now known as fOrsberg
[16-Jul-2010 08:00:33] * Simon4 starts logging this channel
[16-Jul-2010 08:00:48] for one of my smaller components, my interface extends IComponentInfo and just adds serialNumber, product, and manufacturer properties
[16-Jul-2010 08:01:25] let me mock the interface.py and info.py up for one
[16-Jul-2010 08:01:39] would be nice.
[16-Jul-2010 08:04:38] hi: what does this mean in terms of load? "threshold of high utilization exceeded: current value 965453.51"
[16-Jul-2010 08:04:58] robo: sounds like an IP Interface
[16-Jul-2010 08:05:08] hmm
[16-Jul-2010 08:05:17] ah, good call!
[16-Jul-2010 08:05:45] so that would be in bytes?
[16-Jul-2010 08:05:49] I guess I should read up on that
[16-Jul-2010 08:05:59] i think its simply bits/sec
[16-Jul-2010 08:06:03] so 1Mbit
[16-Jul-2010 08:06:50] so here's an example info interface: http://pastebin.com/mk9mPE3m
[16-Jul-2010 08:07:50] cgibbons: i will try how far i can get with it.
[16-Jul-2010 08:09:41] and the matching info object: http://pastebin.com/Aba6vnp8
[16-Jul-2010 08:09:56] small enough to be very useful but still lots of magic within
[16-Jul-2010 08:10:06] your DummyComponent has a Property serialNumber?
[16-Jul-2010 08:10:32] and the interface class defines that this should be displayer and the column title is "Serial"
[16-Jul-2010 08:11:41] You'll still need to make the actual component grid javascript to pick & choose what things are displayed in the component grid. Imagine the info interface/object as an API for the component. The title you see there is used in the 'Details' section when you click on the component.
[16-Jul-2010 08:12:24] cgibbons: do you have time to hold my hand? then i would like to do it step by step.
[16-Jul-2010 08:12:43] i'll be here for a while. but i'll be sure to disappear at the most inopportune time ;)
[16-Jul-2010 08:13:32] start small and just pick a couple of simple columns you want to display to make it easy
[16-Jul-2010 08:14:06] cgibbons: this is my OSComponent class: http://pastebin.com/fEVqUX6y
[16-Jul-2010 08:14:46] ok, 4 text properties easy
[16-Jul-2010 08:14:46] * fragfutter glances at his ip-phone, looks like it is down. the magic of ip.
[16-Jul-2010 08:15:37] you could just make each one like: remSysName = schema.Text(title=u'Remote System Name', readonly=True) in the info interface
[16-Jul-2010 08:15:46] http://pastebin.com/aRB2zqPu
[16-Jul-2010 08:15:52] and then each one can use ProxyProperty in the info object, since it's just a direct representation of the property in your component
[16-Jul-2010 08:16:06] yep
[16-Jul-2010 08:16:24] that is inferfaces.py or info.py?
[16-Jul-2010 08:16:31] thats' interfaces.py; don't forget the imports!
[16-Jul-2010 08:16:52] yes. should all be there
[16-Jul-2010 08:17:03] so far. whats next?
[16-Jul-2010 08:17:57] now the info.py with the implementation of the interface
[16-Jul-2010 08:19:25] what are property and info decorators doing? and what is a ProxyProperty?
[16-Jul-2010 08:19:59] ProxyProperty is just a short-cut that will take the given property name from the actual component and use it as the info object's property
[16-Jul-2010 08:20:16] so if you say foobar = ProxyProperty('eat_my_shorts') it will take self._object.eat_my_shorts and use it for the value of the foobar property
[16-Jul-2010 08:20:28] so this would be enough for me.
[16-Jul-2010 08:20:52] an info object _adapts_ (zope component architecture terminology) another object, and that's given in the constructor and assigned to self._object
[16-Jul-2010 08:21:29] please hold, i will write it. :)
[16-Jul-2010 08:21:33] the @info decorator would only be used if you were doing something that needed to return another component and wanted it adapted with its own info object... like the manufacturer or product entity, but otherwise you will rarely need it
[16-Jul-2010 08:23:51] http://pastebin.com/peX8HCqx
[16-Jul-2010 08:28:38] looks right
[16-Jul-2010 08:28:45] do you already have a configure.zcml ?
[16-Jul-2010 08:29:14] http://pastebin.com/iEHnWMqq
[16-Jul-2010 08:29:44] cool. i'd say restart zope and if it comes up happily all that is good.
[16-Jul-2010 08:29:58] AttributeError: 'module' object has no attribute 'ILLDLink'
[16-Jul-2010 08:31:03] ah. missing a P
[16-Jul-2010 08:31:15] yeah, better to get all that squared away before writing any javascript
[16-Jul-2010 08:34:40] fragfutter, so the eth0 number I got earlier.. I don't think it's bits/sec. I checked all the eth0 threshold emails and it seems the highest was 1355962.25. I checked another bandwidth graph and it shows that it peaked at 700Mbps around the same time zenoss alerted 1355962.25. If it was bytes then it would only have been only around 7megabit.
[16-Jul-2010 08:35:37] robo: what is zenoss thinking that the max speed of the interface is?
[16-Jul-2010 08:37:28] cgibbons: i will drop a device and remodel it, to make sure all objects are created with this pack
[16-Jul-2010 08:37:53] you won't need to worry about that, this is all done at the UI-layer so it doesn't need any of the normal zope magic
[16-Jul-2010 08:38:06] i.e. if the component existed at all previously the info stuff will get applied
[16-Jul-2010 08:38:29] cgibbons: but zope complained when i wanted to inspect the properties of the object.
[16-Jul-2010 08:38:49] I'll take one of the small edge switches ;)
[16-Jul-2010 08:39:06] good call fragfutter.. it thinks it's 10.000Mbps
[16-Jul-2010 08:39:15] this is a gigabit interface
[16-Jul-2010 08:39:35] robo: nothing was connected (or was turned off) last night?
[16-Jul-2010 08:40:45] fragfutter, not that i know of
[16-Jul-2010 08:41:06] zenoss was right in that the server was using a ton of bandwidth. It's just that the numbers mis-match
[16-Jul-2010 08:41:20] well, either was right or got lucky
[16-Jul-2010 08:42:14] it's a vm guest. vmware shows the server at 700mb of bandwidth being used. Zenoss alerts at 1355962.25 whatevers. Maybe that number is wrong because it only thinks it's a 10mb link
[16-Jul-2010 08:42:58] cgibbons: ok. device is discovered and remodelled. Up next on the menu?
[16-Jul-2010 08:43:46] robo: no the link max speed only annoys with the alert threshold it does not mess up the numbers
[16-Jul-2010 08:44:02] okay, next we need to setup a javascript. i'd just create a resources directory in your zenpack (or put it in the root if you prefer), add an empty .js file of your naming, and hten add the browser:viewlet section to your configure.zcml
[16-Jul-2010 08:48:34] http://pastebin.com/NGFnHtDD
[16-Jul-2010 08:48:51] looks right, restart zope to make sure the syntax is all kosher
[16-Jul-2010 08:50:16] ok
[16-Jul-2010 08:55:16] you can use this for the template of your javascript: http://pastebin.com/GXNvXVPw
[16-Jul-2010 08:56:04] remove foobar and add in the 4 properties you want in the fields list and also add columns for them
[16-Jul-2010 08:56:45] you need uid, usesMonitorAttribute, monitor, monitored in the fields list even if you don't have columns for them (the base componentgridpanel and related code use them). those are populated in your info object automatically by the base classes.
[16-Jul-2010 09:01:19] done
[16-Jul-2010 09:01:38] k, you won't need to restart zope, juust go to the component in your device screen and refresh brwoser if necessary
[16-Jul-2010 09:02:20] the grid did not change
[16-Jul-2010 09:02:51] ok, clear cache and retry - also useful to have firebug or something similar installed when doing this so you can see if the browser is infact loading that resource and if there are any errors in the JS file
[16-Jul-2010 09:05:00] (i need to upgrade firefox and firebug, in the mean time...)
[16-Jul-2010 09:05:02] i have a question if my OComponent got the right data.
[16-Jul-2010 09:05:26] so i go into dmd, find my device, get my component
[16-Jul-2010 09:05:52] LLDPLink at /zport/dmd/Devices/Network/Switch/devices/.../os/lldplink
[16-Jul-2010 09:06:14] but l.locPortDesc
[16-Jul-2010 09:06:27] gives an attributeError
[16-Jul-2010 09:07:43] huh
[16-Jul-2010 09:07:54] whatcha get if you look at l._properties and l.__dict__ ?
[16-Jul-2010 09:08:11] >>> l._properties
[16-Jul-2010 09:08:12] ({'type': 'string', 'id': 'snmpindex', 'mode': 'w'}, {'type': 'boolean', 'id': 'monitor', 'mode': 'w'}, {'type': 'string', 'id': 'locPortDesc', 'mode': ''}, {'type': 'string', 'id': 'locIndex', 'mode': ''}, {'type': 'string', 'id': 'remPortDesc', 'mode': ''}, {'type': 'string', 'id': 'remSysName', 'mode': ''})
[16-Jul-2010 09:08:14] fOrsberg is now known as forsberg
[16-Jul-2010 09:08:52] oh did your component define a default value for them?
[16-Jul-2010 09:08:57] didn't see it in your snippet
[16-Jul-2010 09:09:11] no.
[16-Jul-2010 09:11:02] the complete LLDPLink.py http://pastebin.com/z5HeAKPy
[16-Jul-2010 09:11:59] good morning
[16-Jul-2010 09:12:40] does anyone know if there is a zenoss certification?
[16-Jul-2010 09:12:43] try adding locPortDesc = "", locIndex = "", remPortDesc = "", remSysName = "" to the class (newlines instead of commas) and see if that works
[16-Jul-2010 09:16:44] cgibbons: ok. added. zenoss restarted.
[16-Jul-2010 09:19:06] (and there is a javascript error in my LLDPLink.js)
[16-Jul-2010 09:25:32] does my modeller make sense?
[16-Jul-2010 09:25:33] http://pastebin.com/nHXGxa8w
[16-Jul-2010 09:34:33] got the javascript error.
[16-Jul-2010 09:34:48] but somehow my model seems to be broken.
[16-Jul-2010 09:35:07] and there is no change in the grid
[16-Jul-2010 09:36:35] ah, now the model works. it was cached.
[16-Jul-2010 09:36:41] but the grid is still the same.
[16-Jul-2010 09:37:01] what's the js look like?
[16-Jul-2010 09:37:32] http://pastebin.com/KwZr0srC
[16-Jul-2010 09:37:36] anyone know if zenoss plays happily with mysql 5.1?
[16-Jul-2010 09:37:56] that's all I use it with, simon
[16-Jul-2010 09:38:16] cgibbons: that's a perfect answer, cheers :)
[16-Jul-2010 09:44:04] cgibbons: http://pastebin.com/0PGejc77 guess this is the correct version, but still no different grid
[16-Jul-2010 09:44:56] browser cache all cleared, etc?
[16-Jul-2010 09:45:17] checked in firefox, got the corrected script
[16-Jul-2010 09:45:30] oh i see
[16-Jul-2010 09:45:49] Ext.reg('LLDPLinkPanel', ZC.LLDPLinkPanel);
[16-Jul-2010 09:46:27] . o O (and i have no clue what i'm doing)
[16-Jul-2010 09:46:56] * rmatte digs his teeth in to the 3.0 release
[16-Jul-2010 09:46:57] it needs a documented API rather just an example, hopefully one will be written up :)
[16-Jul-2010 09:49:04] i get a javascript exception, but in another script?
[16-Jul-2010 09:49:27] this.config[a] is undefined (in ext-all.js)
[16-Jul-2010 09:52:15] oh, i left a foobar.
[16-Jul-2010 09:52:28] it works.
[16-Jul-2010 09:52:30] *woot*
[16-Jul-2010 09:53:55] yay!
[16-Jul-2010 09:54:12] last question for the day. How do i modify the modeller plugins starting at one point in the device class hierachy in zenoss 3?
[16-Jul-2010 09:55:21] answer, click the blue arrow "Details"
[16-Jul-2010 09:55:29] :)
[16-Jul-2010 09:55:39] totaly intuitive interface :(
[16-Jul-2010 09:55:49] it'll be cool when each release doesn't break everything
[16-Jul-2010 09:55:53] is any of the UI developers here?
[16-Jul-2010 09:56:14] fragfutter: you'd want to talk to ian when you see him
[16-Jul-2010 09:56:14] the complete interface lacks "loading" indications for its ajax requests.
[16-Jul-2010 09:56:20] mrayzenoss is now known as mray
[16-Jul-2010 09:56:30] fragfutter: yup, I was going to log that as a trac
[16-Jul-2010 09:56:35] it does indeed lack that, drives me crazy
[16-Jul-2010 09:56:35] it's on my list
[16-Jul-2010 09:57:44] one should not build an interface with ajax. build a normal interface and then add functionality using it. *brrr*
[16-Jul-2010 09:57:46] summoning spell worked again!
[16-Jul-2010 09:57:57] iancmcc: the complete interface lacks "loading" indications for its ajax requests.
[16-Jul-2010 09:58:17] Yes, in many places.
[16-Jul-2010 09:58:19] cgibbons: your sure the pentagram is uninterrupted and he can't escape?
[16-Jul-2010 09:59:29] i have a stupid question. i detect LLDP Links and create objects for them. How will the modeller detect if a link goes away?
[16-Jul-2010 09:59:41] We forgot some. The things that take a long time, meaning the grid loading, has the indicator.
[16-Jul-2010 10:02:11] s/has/have. I'm bad with English today, apparently.
[16-Jul-2010 10:02:23] at least you didn't use "havening"
[16-Jul-2010 10:02:42] I'm sure you'll find several minor things.
[16-Jul-2010 10:03:31] iancmcc: loading the infrastructure section takes a while
[16-Jul-2010 10:03:38] iancmcc: there was no loader there last I checked
[16-Jul-2010 10:03:42] unless that changed recently
[16-Jul-2010 10:03:51] hmm, it ought to be
[16-Jul-2010 10:03:55] yeah, that load time is getting annoying
[16-Jul-2010 10:04:02] how many devices?
[16-Jul-2010 10:04:11] doesn't matter how many
[16-Jul-2010 10:04:12] It is almost certainly a mysql issue
[16-Jul-2010 10:04:15] oh, that's odd
[16-Jul-2010 10:04:17] does it with 3 devices
[16-Jul-2010 10:04:24] I can get it with 2 :)
[16-Jul-2010 10:04:27] I added 3 test devices, and it took like 30 seconds at times
[16-Jul-2010 10:04:36] although in an earlier beta it didn't
[16-Jul-2010 10:04:41] so not sure what changed to cause that
[16-Jul-2010 10:04:59] what the hell...it's working fine here, so I'm guessing it isn't a code issue. How many events in your status table?
[16-Jul-2010 10:05:47] there should be less than 1000, it was created from our customer template image, which is basically bare bone
[16-Jul-2010 10:05:57] check http://public-demo.zenoss.com/zport/dmd/itinfrastructure
[16-Jul-2010 10:06:09] it loads in just under two seconds for me
[16-Jul-2010 10:06:18] yeh, that's much better
[16-Jul-2010 10:06:23] hmm ok that's good
[16-Jul-2010 10:06:24] let me install the actual 3.0 release and try
[16-Jul-2010 10:06:26] 3.0, w00t
[16-Jul-2010 10:06:28] well, good is the wrong word
[16-Jul-2010 10:06:29] the zope managment interface barfs if i want to look at the root node (host:8080/manage)
[16-Jul-2010 10:06:29] last time I tried was 2 releases ago
[16-Jul-2010 10:06:36] but at least it's a data point
[16-Jul-2010 10:07:08] fragfutter: with 3.0?
[16-Jul-2010 10:07:12] yes
[16-Jul-2010 10:07:42] fragfutter: I believe you're restricted to /zport/dmd/manage
[16-Jul-2010 10:07:57] you can no longer slap /manage all over the place
[16-Jul-2010 10:08:16] hey, it's still zope ;)
[16-Jul-2010 10:08:43] mine doesn't barf. Define "barf", please
[16-Jul-2010 10:08:47] mray: that's something I need to test once I get LDAP auth working... the password auth for changing settings
[16-Jul-2010 10:08:58] I've been putting manage on everything
[16-Jul-2010 10:09:07] that is, the ZMI should be totally unaffected.
[16-Jul-2010 10:09:39] iancmcc: http://pastebin.com/ZWRA7SfD
[16-Jul-2010 10:10:08] well, you aren't logged in of course
[16-Jul-2010 10:10:14] 'user': ,
[16-Jul-2010 10:10:28] iancmcc: but it doesn't ask me
[16-Jul-2010 10:10:35] yeah
[16-Jul-2010 10:10:37] that's a bug.
[16-Jul-2010 10:10:42] I had good intentions.
[16-Jul-2010 10:10:46] :)
[16-Jul-2010 10:10:50] But apparently didn't account for all the possibilities
[16-Jul-2010 10:11:00] iancmcc!
[16-Jul-2010 10:11:01] log in through the app first, then do /manage
[16-Jul-2010 10:11:04] Are you a grown up now? :-)
[16-Jul-2010 10:11:09] and i'm logged into zenoss itself
[16-Jul-2010 10:11:19] If by that you mean "a US resident," then yes.
[16-Jul-2010 10:11:27] heh
[16-Jul-2010 10:11:28] fragfutter: as admin?
[16-Jul-2010 10:11:30] and a Texas no less
[16-Jul-2010 10:11:33] no i mean working in an office ;-)
[16-Jul-2010 10:11:47] OH yeah, that too
[16-Jul-2010 10:11:47] yeah time to purchase fire arms
[16-Jul-2010 10:12:27] and boots and a hat
[16-Jul-2010 10:12:32] iancmcc: role manager, is that enough?
[16-Jul-2010 10:13:12] I think that root always needs admin
[16-Jul-2010 10:13:40] what does the pre-upgrade pack actually do?
[16-Jul-2010 10:13:49] Prepare the global catalog
[16-Jul-2010 10:13:55] it indexes basically everything in your syste
[16-Jul-2010 10:13:58] I see
[16-Jul-2010 10:14:04] which could take you down for a while if you have a ton of stuff
[16-Jul-2010 10:14:09] iancmcc: ok it does. The bug is, that it does not ask for a username
[16-Jul-2010 10:14:21] yes, and that's the fault of the error_message thing.
[16-Jul-2010 10:14:28] I'll open a ticket.
[16-Jul-2010 10:14:33] thanks
[16-Jul-2010 10:15:52] hmm, it's just zenoss 3.0 core right, not enterprise?
[16-Jul-2010 10:15:58] fragfutter just successfully implemented a custom component panel this morning, ian
[16-Jul-2010 10:16:08] i'm stuck waiting a week for my server to arrive so by the time i eventually get it, i guess 3.0 enterprise will have landed :/
[16-Jul-2010 10:16:30] i added a new OSComponent. how can i explain an existing device, that it now has a new relationship? I could drop the device and add it, then it has it. But is there a different possibility
[16-Jul-2010 10:16:46] @ashp - Enterprise has not been released yet.
[16-Jul-2010 10:17:04] "by the time i eventually get it, i guess 3.0 enterprise will have landed"
[16-Jul-2010 10:17:08] he's citing the future
[16-Jul-2010 10:17:15] ;)
[16-Jul-2010 10:17:40] cgibbons: that is awesome. High marks, fragfutter.
[16-Jul-2010 10:17:51] and cgibbons holded my hands all the time
[16-Jul-2010 10:18:01] i only got half of the stuff i did ;)
[16-Jul-2010 10:18:12] that's okay the other half I don't understand either
[16-Jul-2010 10:18:16] cgibbons: custom component panel?
[16-Jul-2010 10:18:35] yeah, for customizing the grid display for a component in the new UI
[16-Jul-2010 10:18:49] ah, nice
[16-Jul-2010 10:19:02] It's rough for the first release, hopefully we'll be able to make the process easier.
[16-Jul-2010 10:19:05] that doesn't involve being able to select what columns are displayed?
[16-Jul-2010 10:19:19] I'll die if I can't see interface descriptions
[16-Jul-2010 10:19:32] I could care less about mac addresses and IPs
[16-Jul-2010 10:20:06] rmatte: It is easy to change, but very different from before
[16-Jul-2010 10:20:21] But much easier to put in a zenpack
[16-Jul-2010 10:20:41] k cool
[16-Jul-2010 10:20:46] and zenpack modifications don't stomp on each other. That is, you can have several packs that each modify component information without overriding each other
[16-Jul-2010 10:21:03] very nice, so it's more pluggable basically
[16-Jul-2010 10:21:04] used to be of course that the last pack loaded wins, and its os tab controlled
[16-Jul-2010 10:21:05] yeah
[16-Jul-2010 10:21:15] yeah, that's sweet
[16-Jul-2010 10:21:33] it's still non-obvious though
[16-Jul-2010 10:21:43] i have two open questions for this week. How to tell all my devices that they now have a new relationship without deleting and adding them (like upgrading the pickeled Zope object)? Second: how does my modellere tell zenoss that the previously detected device is no longer there (LLDP Links).
[16-Jul-2010 10:21:50] oh well, it will be eventually
[16-Jul-2010 10:21:51] We need a tutoria
[16-Jul-2010 10:21:52] l
[16-Jul-2010 10:22:02] +s
[16-Jul-2010 10:22:08] fragfutter: dev.buildRelations()
[16-Jul-2010 10:22:16] for the second I don't know
[16-Jul-2010 10:23:35] iancmcc: no. that doesn't work.
[16-Jul-2010 10:23:49] um.
[16-Jul-2010 10:24:09] http://dev.zenoss.org/trac/ticket/7051
[16-Jul-2010 10:24:22] ok so...maybe I'm misunderstanding what you want
[16-Jul-2010 10:24:43] you've plugged in something to Device._relations or something?
[16-Jul-2010 10:25:13] i have a new OSComponent. It has a relationship. so for new devices i have device.os.lldplinks. But devices that got added before i installed the zenpack know nothing about it.
[16-Jul-2010 10:25:26] OH sorry
[16-Jul-2010 10:25:32] device.os.buildRelations(), then
[16-Jul-2010 10:26:07] ah....
[16-Jul-2010 10:26:11] yes. thats it.
[16-Jul-2010 10:26:46] hi rmatte, do you know if there is a zenoss certifitation?
[16-Jul-2010 10:26:53] iancmcc: in zencatalog something that needs to be running permanently thereafter?
[16-Jul-2010 10:26:57] or is it just used pre-upgrade?
[16-Jul-2010 10:27:04] the latter
[16-Jul-2010 10:27:16] (I selectively run daemons, so I'll need to add it to the list then)
[16-Jul-2010 10:27:28] it's only a daemon in the pre-upgrade
[16-Jul-2010 10:27:38] otherwise it's just a script that makes it easier to reindex things
[16-Jul-2010 10:27:41] oh
[16-Jul-2010 10:27:45] so the upgrade removes it?
[16-Jul-2010 10:27:49] and for the second question. i walk the lldp-mib of a switch. For every link i find, i create a link object. so what happens on the next modelling run if a link no longer exists. I won't detect it missing. it will never vanish from zenoss
[16-Jul-2010 10:27:54] rmatte: yes
[16-Jul-2010 10:27:59] ah ok, cool
[16-Jul-2010 10:28:06] I'll just start it manually then when doing the upgrades
[16-Jul-2010 10:28:09] thanks
[16-Jul-2010 10:28:13] well, no. The upgrade removes the pack
[16-Jul-2010 10:28:18] right
[16-Jul-2010 10:28:19] there's a /new/ zencatalog in 3.0
[16-Jul-2010 10:28:20] i think
[16-Jul-2010 10:28:24] oh
[16-Jul-2010 10:28:24] i don't think we removed it
[16-Jul-2010 10:28:28] but it's just a script
[16-Jul-2010 10:28:33] in case you feel like reindexing things
[16-Jul-2010 10:28:34] so it doesn't need to be running but it's recommended that it is?
[16-Jul-2010 10:28:38] not at all
[16-Jul-2010 10:28:44] ok, cool
[16-Jul-2010 10:28:50] I prefer the old fashioned way of reindexing
[16-Jul-2010 10:28:51] everything is indexed in the normal course of running things
[16-Jul-2010 10:29:01] but just in case you want to, you can.
[16-Jul-2010 10:29:04] it's just as easy to do it from zendmd
[16-Jul-2010 10:29:09] :)
[16-Jul-2010 10:29:12] it is, but some people don't like zendmd
[16-Jul-2010 10:29:21] true
[16-Jul-2010 10:29:24] I love it
[16-Jul-2010 10:29:26] so meh
[16-Jul-2010 10:30:22] this reindexing is a bit more intense than the standard
[16-Jul-2010 10:30:26] taking quite a bit longer
[16-Jul-2010 10:31:26] yes, it will take a long time
[16-Jul-2010 10:31:32] but it means we can make a much, much better UI
[16-Jul-2010 10:31:46] I'm cringing seeing how long it's taking on a server that's essentially a blank slate
[16-Jul-2010 10:31:52] when I have one with over 400 devices on it
[16-Jul-2010 10:31:58] :)
[16-Jul-2010 10:31:59] it takes about ten minutes on a fresh 2.5 install
[16-Jul-2010 10:32:04] ah
[16-Jul-2010 10:32:05] depending on hardware of course
[16-Jul-2010 10:32:10] yeh
[16-Jul-2010 10:32:28] that's why we did it pre-upgrade instead of post
[16-Jul-2010 10:32:34] makes sense
[16-Jul-2010 10:32:46] otherwise your system would be down while it indexed
[16-Jul-2010 10:32:49] or your UI would be blank
[16-Jul-2010 10:32:50] or both
[16-Jul-2010 10:33:01] as long as it doesn't actually take zenoss down it's fine
[16-Jul-2010 10:33:30] this is all good information. says the man sitting silently in the corner
[16-Jul-2010 10:33:34] this one's at 12% so far
[16-Jul-2010 10:33:52] rmatte: you're upgrading dev servers or your prod boxes?
[16-Jul-2010 10:33:56] it's a hassle, but only for this upgrade. after this no big deal.
[16-Jul-2010 10:34:06] twm1010: it's a test VM
[16-Jul-2010 10:34:07] i could swear you just recently did 2.5.2, I did it right after you
[16-Jul-2010 10:34:21] twm1010: I took our client image, cloned it, and now I'm upgrading it from 2.5.2 to 3.0
[16-Jul-2010 10:34:39] it's snapshotted at 2.5 too, so I can keep rolling back to test stuff
[16-Jul-2010 10:34:48] Do Egor's zenpacks need to be pre-upgrade to 3.0 compatible versions?
[16-Jul-2010 10:34:56] yessir
[16-Jul-2010 10:34:58] we had that problem with 2.4-2.5
[16-Jul-2010 10:35:07] upgrade to his latest releases
[16-Jul-2010 10:35:15] they are cross-compatible
[16-Jul-2010 10:35:20] twm1010: Egor's upgraded all of his zenpacks
[16-Jul-2010 10:35:29] please upgrade them before going to 3.0
[16-Jul-2010 10:35:35] I'm using WMI datasource 1.5 on 2.5.2
[16-Jul-2010 10:35:45] mray: the upgrade notice at the top of the WMI ZenPack page needs changing
[16-Jul-2010 10:35:52] it's still the old message from the last time this happenede
[16-Jul-2010 10:35:56] -e
[16-Jul-2010 10:35:57] :)
[16-Jul-2010 10:36:02] rmatte: yeah, I'm going to be removing that when I refresh this list
[16-Jul-2010 10:36:03] docs/DOC-5914
[16-Jul-2010 10:36:12] ah, niceness
[16-Jul-2010 10:36:13] is this dynamic service assurance stuff on the homepage part of core too?
[16-Jul-2010 10:36:13] iancmcc: want to hear other UI bugs?
[16-Jul-2010 10:36:13] Egor's sent me 10+ ZenPacks to update
[16-Jul-2010 10:36:26] fragfutter: make tickets
[16-Jul-2010 10:36:37] twm1010: it's marketing talk, but it's mostly about Enterprise visualization stuff
[16-Jul-2010 10:36:46] ah that blows, a bunch of rocket's packs broke
[16-Jul-2010 10:36:47] iancmcc: then i need a trac account
[16-Jul-2010 10:36:56] fragfutter: email me and I'll hook you up
[16-Jul-2010 10:36:58] mray@zenoss.com
[16-Jul-2010 10:37:18] synthetic web transactions failed, bah
[16-Jul-2010 10:37:42] rmatte: doh!
[16-Jul-2010 10:37:46] rmatte: I've emailed all the authors
[16-Jul-2010 10:37:56] Andrea Consadori and Egor already fixed a few on that list
[16-Jul-2010 10:37:59] well, if the synthetic pack doesn't get fixed by the author I'll fix it
[16-Jul-2010 10:38:01] The changes required to update them should be pretty minimal
[16-Jul-2010 10:38:03] that's a darn good pack
[16-Jul-2010 10:38:28] rmatte: can you pipe the synthetic checks through a proxy server?
[16-Jul-2010 10:38:36] It's mostly UI, but the python 2.6 upgrade definitely obsoleted soe stuff
[16-Jul-2010 10:38:40] twm1010: why wouldn't you be able to?
[16-Jul-2010 10:38:55] twm1010: a proxy is a proxy, it makes no difference to twill
[16-Jul-2010 10:39:17] I think if I used that it would be a requirement, so that our checks come in the front door
[16-Jul-2010 10:39:34] twm1010: I'm 99% sure it'd work
[16-Jul-2010 10:41:02] I wonder if anyone is monitoring content switches and loadbalancers with Zenoss
[16-Jul-2010 10:41:04] mray: is rocket allowed to fix the broken packs if he wanted to, or are those orphaned?
[16-Jul-2010 10:41:14] twm1010: there's a Zeus zenpack
[16-Jul-2010 10:41:16] So I may just be bold enough to move to 3.0
[16-Jul-2010 10:41:32] mray: did egor send you a new wmidatasource? it shows it works with 3, but I hate to find out the hardway
[16-Jul-2010 10:41:36] I would like to see someone take a look at WebInject as a replacement for the Twill ZenPack
[16-Jul-2010 10:41:37] http://www.webinject.org/
[16-Jul-2010 10:41:43] theacolyte: verify that any packs on which you depend are compatible first
[16-Jul-2010 10:41:47] theacolyte: yeah, Egor updated his stuff
[16-Jul-2010 10:41:53] theacolyte: I believe 2.7 is the latest
[16-Jul-2010 10:41:54] otherwise hopefully you're golden
[16-Jul-2010 10:41:59] 2.7 is the one you want
[16-Jul-2010 10:42:01] Got it
[16-Jul-2010 10:42:26] I guess it's what snapshots are for in the worst case... i'll let you know how it goes
[16-Jul-2010 10:42:31] I'll be testing it out to make sure it still actually works as expected
[16-Jul-2010 10:42:35] as soon as this install is done
[16-Jul-2010 10:42:43] is there a page with a feature comparison of 2.5.2 vs 3.0?
[16-Jul-2010 10:42:53] theacolyte: I'm beginning to wish my Zenoss server was virtual :|
[16-Jul-2010 10:42:55] damn your silly epub format btw!
[16-Jul-2010 10:43:03] but stack pretty much lets me make a quick full backup anyway, so...
[16-Jul-2010 10:43:05] What about us poor kindle users!
[16-Jul-2010 10:43:08] phoenix20: not sure if there's a really formal one, but the press release should contain info on that
[16-Jul-2010 10:43:13] phoenix20: check the release notes too
[16-Jul-2010 10:43:28] @theacolyte - Put in a feature request for whatever format you want the docs in, and we can consider it.
[16-Jul-2010 10:43:40] phoenix20: the release notes are available from the download page
[16-Jul-2010 10:43:46] bbibeault: I probably shoud, though really - if you know what you're doing you can convert it
[16-Jul-2010 10:43:53] rmatte: nice. thanks
[16-Jul-2010 10:44:03] phoenix20: no problem
[16-Jul-2010 10:44:25] As I understand it we decided to try an alternate format to see what folks thought. That is good feedback.
[16-Jul-2010 10:44:48] The sad part is it's such a moving target
[16-Jul-2010 10:45:17] all the more reason to have target practice :)
[16-Jul-2010 10:45:27] although epub is where everything is headed, so you were right - if I were to pick a single format that'd be it
[16-Jul-2010 10:45:37] (y)
[16-Jul-2010 10:45:44] How about this, I'll put it into trac tonight, wtih a converted version
[16-Jul-2010 10:46:16] mray: are there any packs for A10 load balancers?
[16-Jul-2010 10:46:21] even in enterprise?
[16-Jul-2010 10:46:37] That is fine, thanks.
[16-Jul-2010 10:47:26] twm1010: I haven't seen any
[16-Jul-2010 10:50:10] cgibbons: result of your help: http://i26.tinypic.com/f2l9om.png
[16-Jul-2010 10:50:27] fragfutter: nice
[16-Jul-2010 10:51:27] next step, generate links for the localport, the remote systems (requires going over ip address and searching them in the DB) and remote ports.
[16-Jul-2010 10:51:40] fragfutter: you have my attention, what is that?
[16-Jul-2010 10:51:52] correlating switch port assignment to host?
[16-Jul-2010 10:52:17] doesn't the Bridge MIB ZenPack do that?
[16-Jul-2010 10:52:18] twm1010: layer2 links between LLDP capable devices (switches normaly)
[16-Jul-2010 10:53:17] that's pretty cool
[16-Jul-2010 10:53:17] mray: LLDP is easier. Discovering Links with bridge mib requires a complete list of mac addresses. one port can have more then one mac address entry in the bridge table
[16-Jul-2010 10:54:32] mray: there is an algorithm to do it. But it get's slow on huge nets. With LLDP you can get the link for every device without knowing anything about it's neighbours
[16-Jul-2010 10:54:50] that is links between devices that both speak lldp.
[16-Jul-2010 10:55:08] cool
[16-Jul-2010 10:56:01] and as some of the switches are not capable to hold the complete arp cache in memory, guess why i prefer lldp ;)
[16-Jul-2010 10:56:23] fragfutter: You're going to make this a ZenPack?
[16-Jul-2010 10:57:01] twm1010: it will currently not remove links that are unplugged. And i need to ask management about the legal stuff of releasing it.
[16-Jul-2010 10:59:18] releasing under gpl would be a non issue, but signing over to zenoss will be.
[16-Jul-2010 10:59:46] so, this happens during modeling basically?
[16-Jul-2010 10:59:53] twm1010: yes
[16-Jul-2010 11:00:24] well, if you're releasing it under the gpl you're not really "signing it over" to anyone in particular
[16-Jul-2010 11:00:29] seems like a moot point
[16-Jul-2010 11:00:55] rmatte: if i want to add it to zenoss subversion i need to sign it over.
[16-Jul-2010 11:01:06] I see
[16-Jul-2010 11:01:30] yeah, that way we could eventually move it into the product
[16-Jul-2010 11:01:40] well, you could always just keep it in your svn but allow it to be downloaded for the community zenpacks page
[16-Jul-2010 11:01:43] not a huge deal
[16-Jul-2010 11:01:59] mray: totaly clear. as you have a dual license you need to have all the rights.
[16-Jul-2010 11:02:13] but first it needs cleanup.
[16-Jul-2010 11:02:23] Fighting likewise open before I upgrade
[16-Jul-2010 11:02:24] software legalities make my head spin
[16-Jul-2010 11:02:31] yeah, but the Contribution Form is a mutual license, you keep your rights
[16-Jul-2010 11:02:59] anyway, hope to have it someday :)
[16-Jul-2010 11:03:07] right, so it's not really signing it over so much as granting rights to it
[16-Jul-2010 11:03:18] we are looking to start moving more Community ZenPacks to Core
[16-Jul-2010 11:03:19] mray: but i hand over something i did during working hours and you might decide to close zenoss core.
[16-Jul-2010 11:03:39] i will ask mgmt and we'll see.
[16-Jul-2010 11:03:50] fragfutter: while that's exceedingly unlikely, I understand the reservations
[16-Jul-2010 11:03:51] I can't see core getting closed
[16-Jul-2010 11:03:57] the whole business model is built up around it
[16-Jul-2010 11:04:05] and I'd be encouraging a Community fork :)
[16-Jul-2010 11:04:10] for sure
[16-Jul-2010 11:04:52] overall quality would drop if that happened, less bug testers
[16-Jul-2010 11:05:21] again, exceedingly unlikely since the Zenoss business model is built around Open Source
[16-Jul-2010 11:05:26] yup
[16-Jul-2010 11:05:34] it seems to work well too
[16-Jul-2010 11:08:48] I would still really like to see more community involvement in the core infrastructure itself instead of just zenpacks. not that this has been our focus, and it's difficult given the legacy code architecture
[16-Jul-2010 11:08:53] woohoo, zencatalog finally finished
[16-Jul-2010 11:08:56] time to install 3.0
[16-Jul-2010 11:09:05] test system?
[16-Jul-2010 11:09:15] yeh
[16-Jul-2010 11:09:22] how many devs?
[16-Jul-2010 11:09:24] 3.0 in production is a ways off for me
[16-Jul-2010 11:09:29] I haven't added any yet
[16-Jul-2010 11:10:00] is there a way to change the notification for a given problem? for example, i monitor ajava process thats our erp
[16-Jul-2010 11:10:12] i want the page to say erp is down, not that java is down
[16-Jul-2010 11:10:21] straterra: do it with a transform
[16-Jul-2010 11:10:21] I finally got likewise working again, so time to install
[16-Jul-2010 11:10:41] alrighty
[16-Jul-2010 11:12:54] Actually though I htink I will go up to 10.04 on this box first
[16-Jul-2010 11:13:33] zenoss 3? i just upgraded last week
[16-Jul-2010 11:14:23] . o O (did you know, that you can spend minutes just closing all the windows i had on my workspaces)
[16-Jul-2010 11:16:58] cgibbons: I'd love to contribute,I'm not a coder, I'm lucky to write a shell script :|
[16-Jul-2010 11:18:47] weeeeekend
[16-Jul-2010 11:22:02] Anyone have a good screenshot of mySQL monitoring graphs?
[16-Jul-2010 11:23:17] twm1010: I can give you some for my SSH-variant of it
[16-Jul-2010 11:23:27] same graphs
[16-Jul-2010 11:25:14] http://imagebin.org/105587
[16-Jul-2010 11:32:27] http://hesten.foppa.dk/files/Capture.JPG
[16-Jul-2010 11:32:33] ups same yeah :)
[16-Jul-2010 11:33:08] cool, some tomcat screenshots would be even better :)
[16-Jul-2010 11:33:32] * forsberg did not follow
[16-Jul-2010 11:33:36] damn my mysql is busy
[16-Jul-2010 11:33:37] ;>
[16-Jul-2010 11:36:46] as is mine
[16-Jul-2010 11:36:47] Queries per second avg: 184.949
[16-Jul-2010 11:36:54] ok, so, I followed the proper steps for the 3.0 upgrade and after the upgrade...
[16-Jul-2010 11:36:55] Daemon: zenperfwmi /usr/local/zenoss/zenoss/bin/zenoss: line 106: /usr/local/zenoss/zenoss/bin/zenperfwmi: No such file or directory
[16-Jul-2010 11:36:55] Daemon: zenticket /usr/local/zenoss/zenoss/bin/zenoss: line 106: /usr/local/zenoss/zenoss/bin/zenticket: No such file or directory
[16-Jul-2010 11:37:07] and that's with the latest wmi pack installed beforehand
[16-Jul-2010 11:37:23] hmm, what i need to do when i want every event having some text in summary to generate alert
[16-Jul-2010 11:37:48] now i have alert rule maching that event, but only first event will trigger it
[16-Jul-2010 11:38:43] next event will only show in counter, and when event is moved to history, next event works ok
[16-Jul-2010 11:39:08] I'll try installing the WMI pack overtop after the upgrade
[16-Jul-2010 11:39:50] sytem: add more event states than 'new"
[16-Jul-2010 11:40:01] might make it keep sending em
[16-Jul-2010 11:40:05] or just remove that filter
[16-Jul-2010 11:40:14] removing it is probably the way, actually
[16-Jul-2010 11:40:38] ok, so installing the pack after the upgrade fixes it
[16-Jul-2010 11:41:15] filters in alerting rule are "summary contains test-event" and "device contatains xxx.yyy.zz"
[16-Jul-2010 11:41:25] hmm, crap
[16-Jul-2010 11:41:55] installing the other pack does too
[16-Jul-2010 11:43:09] alright, time to login and check it out
[16-Jul-2010 11:43:34] ok creating catalog
[16-Jul-2010 11:43:47] suprisingly going from 8.04LTS > 10.04LTS worked this time
[16-Jul-2010 11:44:07] yeh, the infrastructure page is still taking way too long to load
[16-Jul-2010 11:44:11] even with 0 devices
[16-Jul-2010 11:53:31] the time it takes to update catalogs pre-upgrade may be an issue for some
[16-Jul-2010 11:53:43] you may want to put a warning on that
[16-Jul-2010 11:53:49] WARNING: THIS CAN TAKE A LONG TIME TO COMPLETE
[16-Jul-2010 11:56:22] here's a fun one I just found...
[16-Jul-2010 11:56:24] Navigate to Advanced -> ZenPacks, then click on Daemons. Notice that there is no longer any indication of which section you are in in the top menu. This happens when clicking from ZenPacks to Versions as well.
[16-Jul-2010 11:56:39] I'm just downloading 3.0 now
[16-Jul-2010 11:57:02] my list of bugs is growing, though most of them are small bugs
[16-Jul-2010 11:57:04] like that one
[16-Jul-2010 11:57:34] installing now
[16-Jul-2010 11:59:26] nice, they fixed some link colors that were driving me nuts
[16-Jul-2010 12:00:40] hehe
[16-Jul-2010 12:00:44] been at 100% installation for awhile
[16-Jul-2010 12:00:59] it's technically at 99%
[16-Jul-2010 12:01:00] :P
[16-Jul-2010 12:01:04] haha
[16-Jul-2010 12:01:10] Ah, the Microsoft progress bar
[16-Jul-2010 12:01:20] the cursor needs to move another step for it to complete
[16-Jul-2010 12:01:22] :P
[16-Jul-2010 12:01:34] True.. there's still the %
[16-Jul-2010 12:01:39] 0% ______________ 50% ______________ 100%
[16-Jul-2010 12:01:39] ########################################
[16-Jul-2010 12:01:45] damn you %
[16-Jul-2010 12:01:50] hehe
[16-Jul-2010 12:02:01] most people misinterpret that
[16-Jul-2010 12:02:14] I've seen people kill the install early and then complain that zenoss doesn't work properly
[16-Jul-2010 12:02:34] There's a direct relationship between % and elapsed/time to go
[16-Jul-2010 12:02:57] sweet
[16-Jul-2010 12:02:58] If it zoomes through from 1-99% and the last 1% takes 300x as long...
[16-Jul-2010 12:03:02] hehe
[16-Jul-2010 12:03:02] I get this when adding a device: http://fpaste.org/19Kh/
[16-Jul-2010 12:03:21] :(
[16-Jul-2010 12:03:40] the device did get added
[16-Jul-2010 12:03:46] but not modeled properly
[16-Jul-2010 12:04:03] just got a zenoss alarm: : 'NoneType' object has no attribute 'clone'
[16-Jul-2010 12:04:11] nice
[16-Jul-2010 12:04:46] ah looks like it's finished!
[16-Jul-2010 12:05:03] yeh great, so I can't add or model devices
[16-Jul-2010 12:05:07] strike 1
[16-Jul-2010 12:05:18] the last beta I tested worked fine for doing this...
[16-Jul-2010 12:05:28] though egor mentioned that it might have something to do with his pack
[16-Jul-2010 12:05:33] I'll remove the pack and then try
[16-Jul-2010 12:05:37] Which one? Wmi?
[16-Jul-2010 12:06:07] yeh
[16-Jul-2010 12:06:19] oh, woot
[16-Jul-2010 12:06:23] my map is properly zoomed out now
[16-Jul-2010 12:06:42] oh my god
[16-Jul-2010 12:06:47] you can't shift select events?
[16-Jul-2010 12:07:04] you should be able to
[16-Jul-2010 12:07:20] haha, I cna't
[16-Jul-2010 12:07:24] not in chrome at least
[16-Jul-2010 12:07:33] though in the event console, you can hit the selecvt button
[16-Jul-2010 12:07:38] sweet, I can't remove the ZenPack
[16-Jul-2010 12:07:40] AttributeError: 'NoneType' object has no attribute 'clone'
[16-Jul-2010 12:07:47] this is already becoming a pain in the ass lol
[16-Jul-2010 12:07:48] ah, I got that one
[16-Jul-2010 12:08:06] nope
[16-Jul-2010 12:08:11] you can't select more than one event with shift
[16-Jul-2010 12:08:18] http://fpaste.org/6ANF/
[16-Jul-2010 12:08:21] hey guys :)
[16-Jul-2010 12:08:35] rmatte: so after my interesting issues with jumpbox/ec2 & my 2.5.0
[16-Jul-2010 12:08:42] well time to roll back and test without the WMI packs
[16-Jul-2010 12:08:47] do you think 3.0.0 beta is stable enough to run in prod
[16-Jul-2010 12:08:55] Good news, 3.0 has released
[16-Jul-2010 12:08:58] willwh: 3.0 was released today, look at the topic
[16-Jul-2010 12:09:02] ;)
[16-Jul-2010 12:09:04] zomg !
[16-Jul-2010 12:09:08] * willwh dances
[16-Jul-2010 12:09:14] As to if it's stable enough... not sure
[16-Jul-2010 12:09:19] good luck with it though, it's making me cry
[16-Jul-2010 12:09:25] oh really?
[16-Jul-2010 12:09:25] :/
[16-Jul-2010 12:09:27] but I'm upgrading from 2.5.2 to 3.0 to test it
[16-Jul-2010 12:09:31] ah
[16-Jul-2010 12:09:34] a clean install should be alright
[16-Jul-2010 12:09:36] I don't really need to upgrade I don't think
[16-Jul-2010 12:09:54] the interface is way better
[16-Jul-2010 12:09:57] you guys have a debian repo right?
[16-Jul-2010 12:10:03] or am I being crazy?
[16-Jul-2010 12:10:08] nice... I can't start zenperfwmi
[16-Jul-2010 12:10:25] theacolyte: yeh, I had the same problem, just install it overtop of what's already there
[16-Jul-2010 12:10:28] and you'll be able to start it
[16-Jul-2010 12:10:30] got it
[16-Jul-2010 12:10:35] yeah the last time I had the issue I had to do that
[16-Jul-2010 12:10:35] but you might not be able to model anything afterwards
[16-Jul-2010 12:10:43] that was from 2.5.1 > 2.5.2
[16-Jul-2010 12:10:47] let me try starting it manually first
[16-Jul-2010 12:10:50] ah
[16-Jul-2010 12:10:54] k
[16-Jul-2010 12:11:54] uh oh
[16-Jul-2010 12:11:55] it's not in bin
[16-Jul-2010 12:12:02] what?
[16-Jul-2010 12:12:07] zenperfwmi
[16-Jul-2010 12:12:13] is usually in zenoss/bin right?
[16-Jul-2010 12:12:41] $ZENHOME/bin
[16-Jul-2010 12:12:42] yes
[16-Jul-2010 12:12:52] if it's not there, install the pack overtop of what's there
[16-Jul-2010 12:12:55] and it'll fix it
[16-Jul-2010 12:12:55] looks like it's not yeah
[16-Jul-2010 12:12:57] so yeah
[16-Jul-2010 12:13:04] ( was going to try manually starting it)
[16-Jul-2010 12:13:06] that's no good
[16-Jul-2010 12:13:27] so yeah the upgrade does remove zenperfwmi
[16-Jul-2010 12:13:28] lol
[16-Jul-2010 12:13:34] just made sure
[16-Jul-2010 12:13:37] yup
[16-Jul-2010 12:13:46] it removed my custom daemon from /bin too
[16-Jul-2010 12:13:55] it doesn't technically remove the daemon itself, just the symlink to it
[16-Jul-2010 12:13:58] rmatte do you use a proxy via apache to supply SSL for zenoss….if so do you happen to have a wiki article on that
[16-Jul-2010 12:13:59] which is a bug in my opinion
[16-Jul-2010 12:14:04] it shouldn't be doing that
[16-Jul-2010 12:14:29] st3v3o: we do, but no article, and I'm not the one who set it up
[16-Jul-2010 12:14:47] st3v3o: we use sun web proxy
[16-Jul-2010 12:14:54] it's braindead simple to configure
[16-Jul-2010 12:15:02] This not being able to select more than one event is already pissing me off
[16-Jul-2010 12:15:06] looks like the proxy is working ..but it wont allow me to log on
[16-Jul-2010 12:15:17] st3v3o: weird
[16-Jul-2010 12:15:34] http://pastebin.com/LzRErvb5
[16-Jul-2010 12:16:00] I love the new networks tab so far though
[16-Jul-2010 12:16:05] * rmatte shrugs
[16-Jul-2010 12:16:06] no idea
[16-Jul-2010 12:16:29] theacolyte: yeh, the layout is nice, there's just annoyances that need to be cleaned up
[16-Jul-2010 12:18:17] it's sort of hard to see if something is loaded or laoding
[16-Jul-2010 12:18:46] yup
[16-Jul-2010 12:19:37] Al my uptime went to unknown
[16-Jul-2010 12:19:38] haha
[16-Jul-2010 12:19:53] Also what does this gray snoflake icon mean
[16-Jul-2010 12:20:02] it'll take it some time to cache the uptime info again
[16-Jul-2010 12:20:05] you didn't actually lose it
[16-Jul-2010 12:20:26] ah
[16-Jul-2010 12:20:40] Man, the interface is much more usable
[16-Jul-2010 12:20:52] bigegor: so I installed the new versions of your pack, upgraded, and got the following when adding or modelling devices: http://fpaste.org/19Kh/
[16-Jul-2010 12:21:05] re: multiple events clearing - you have to do it in the console with the select all button
[16-Jul-2010 12:21:11] I'm gonna call that one a bug and send it up
[16-Jul-2010 12:21:17] going to try removing the packs prior to upgrading to see if I can model
[16-Jul-2010 12:22:19] rmatte: same thing with event console cache do you think?
[16-Jul-2010 12:22:43] theacolyte: event console cache?
[16-Jul-2010 12:23:12] the event console shouldn't take time to populate as long as alerts are coming in
[16-Jul-2010 12:23:26] theacolyte: you've made sure that mysql is running right?
[16-Jul-2010 12:23:36] the upgrade didn't start mysql for me afterwards
[16-Jul-2010 12:23:40] I had to start it by hand
[16-Jul-2010 12:24:08] bigegor: I also got this when trying to remove the WMI packs: http://fpaste.org/6ANF/
[16-Jul-2010 12:24:11] rmatte: Job completed at 2010-07-16 13:02:24. Result: success. All's well that ends well ;)
[16-Jul-2010 12:24:20] wait... it will run without mysql?
[16-Jul-2010 12:24:24] bigegor: nope
[16-Jul-2010 12:24:30] bigegor: it didn't model anything lol
[16-Jul-2010 12:24:41] mysql is running
[16-Jul-2010 12:24:48] theacolyte: of course it'll run without mysql
[16-Jul-2010 12:24:52] as long as zope is running
[16-Jul-2010 12:24:57] Looks like my event history is so big it times out
[16-Jul-2010 12:24:57] you'll get errors though
[16-Jul-2010 12:25:04] I see
[16-Jul-2010 12:25:13] They are all WMI issues
[16-Jul-2010 12:25:14] it shouldn't be timing out, they need to fix that
[16-Jul-2010 12:25:17] * theacolyte st ares at bigegor
[16-Jul-2010 12:25:29] I kid I need to fix it myself
[16-Jul-2010 12:26:04] Zenoss shouldn't choke because of a large event database
[16-Jul-2010 12:26:10] that's just dumb
[16-Jul-2010 12:27:16] ah there it goes
[16-Jul-2010 12:27:20] I think it did need to do some sort of cache
[16-Jul-2010 12:27:23] Now it's loading quickly
[16-Jul-2010 12:27:27] cool
[16-Jul-2010 12:27:36] first load was like 10 minutes, and showed "disconnected" error
[16-Jul-2010 12:27:38] rmatte: seriously, it doesn't looks like wmi problem
[16-Jul-2010 12:28:07] bigegor: well, I guess we'll see, I'm rolling back to a snapshot, I'm going to remove the WMI packs and do the same upgrade on the same server
[16-Jul-2010 12:28:19] if I'm suddenly able to model devices properly then it'll obviously have something to do with it
[16-Jul-2010 12:33:57] ah well
[16-Jul-2010 12:34:01] seems like the upgrad went pretty well
[16-Jul-2010 12:34:05] so on to other things
[16-Jul-2010 12:34:22] Loading. Please wait....
[16-Jul-2010 12:34:28] lol
[16-Jul-2010 12:34:45] wtf
[16-Jul-2010 12:35:10] my device list (that has maybe 50 things on it btw) just sits at loading. please wait...
[16-Jul-2010 12:35:33] you get a loading dialog!?
[16-Jul-2010 12:35:39] yeah.
[16-Jul-2010 12:35:41] you're lucky
[16-Jul-2010 12:35:42] lol
[16-Jul-2010 12:35:53] I can't actually load my device list anymore
[16-Jul-2010 12:35:57] it was working at one point
[16-Jul-2010 12:35:59] give it time
[16-Jul-2010 12:36:07] I'm at 2 minutes for 50 devices
[16-Jul-2010 12:36:08] it takes like 30 seconds to load with even 0 devices for me
[16-Jul-2010 12:36:13] they screwed it up somehow
[16-Jul-2010 12:36:23] well
[16-Jul-2010 12:36:27] at least it looks good as it loads.
[16-Jul-2010 12:36:45] which is why I'm just treating 3.0 as a beta, and will get around to upgrading once a couple of bugfix releases are done
[16-Jul-2010 12:36:47] oh you know what
[16-Jul-2010 12:36:54] i think the issue may be with chrome
[16-Jul-2010 12:36:58] possibly
[16-Jul-2010 12:37:01] which i suppose isn't on the supported browser list
[16-Jul-2010 12:37:04] so i suppose that's my fault
[16-Jul-2010 12:37:05] chrome is not an officially supported browser
[16-Jul-2010 12:37:15] (FWIW Chrome worked fine with 2.5.2)
[16-Jul-2010 12:37:17] it's "partially supported"
[16-Jul-2010 12:37:21] firefox is the way to go
[16-Jul-2010 12:37:27] ug
[16-Jul-2010 12:37:36] don't want to go back to that memory leaking, cpu hungry browser :/
[16-Jul-2010 12:37:46] it's really not that bad
[16-Jul-2010 12:37:59] 3.0 is a lot less system intensive
[16-Jul-2010 12:38:01] I find it funny that I now say that about Firefox, I'm sure IE was the bad browser one time :)
[16-Jul-2010 12:38:04] what's bad is I have to have 3 browsers
[16-Jul-2010 12:38:04] and at least it renders pages properly
[16-Jul-2010 12:38:23] IE is garbage, what are you talking about?
[16-Jul-2010 12:38:28] it's just as slow as Firefox, if not slower
[16-Jul-2010 12:38:33] and it can't render a page to save it's life
[16-Jul-2010 12:38:41] rmatte: I'm just surprised I ended up saying it about Firefox
[16-Jul-2010 12:38:46] I totally agree IE is garbage
[16-Jul-2010 12:38:47] lol
[16-Jul-2010 12:39:30] despite firefox being slower than browsers like chrome and opera, it renders pages better than any browser out there
[16-Jul-2010 12:40:19] it also has the best plugins
[16-Jul-2010 12:40:20] :)
[16-Jul-2010 12:40:24] that it does
[16-Jul-2010 12:40:38] greasemonkey and firebug are awesome
[16-Jul-2010 12:40:45] and given I've stopped using facebook, I don't have to take the 10% cpu hit at idle any more
[16-Jul-2010 12:40:52] lol
[16-Jul-2010 12:41:02] well, facebook is ajax central
[16-Jul-2010 12:41:03] that site chewed cpu on javascript stuff, even when you weren't using it
[16-Jul-2010 12:41:06] yah
[16-Jul-2010 12:41:19] yeh, that's because of the facebook messenger
[16-Jul-2010 12:41:41] yslow
[16-Jul-2010 12:41:44] if you disable that it's not too bad
[16-Jul-2010 12:41:44] don't forget yslow
[16-Jul-2010 12:41:51] it's less than a third of the cpu in chrome
[16-Jul-2010 12:41:58] chrome = superfast js
[16-Jul-2010 12:42:15] so it would be really nice if zenoss supported it, especially in 3
[16-Jul-2010 12:42:22] chrome is really fast in general, now if they can get the rendering up to par with firefox I'd be game to switch
[16-Jul-2010 12:42:31] well, they'd also need something like greasemonkey
[16-Jul-2010 12:43:05] I use this greasemonkey script all the time: http://userscripts.org/scripts/show/73700
[16-Jul-2010 12:43:15] it's awesome
[16-Jul-2010 12:43:30] ooh fancy
[16-Jul-2010 12:43:35] you click one checkbox, hold down the mouse button and drag over others and it checks them all
[16-Jul-2010 12:49:06] Loading. Please wait....
[16-Jul-2010 12:49:19] That should be the motto for this release
[16-Jul-2010 12:49:38] Zenoss Community Version "Loading. Please wait..." 3.0
[16-Jul-2010 12:50:11] lol
[16-Jul-2010 12:50:22] too funny
[16-Jul-2010 12:50:41] * Simon4 will try the enterprise version when it comes out and see whether it's faster
[16-Jul-2010 12:50:51] then we can grep for the "sleep(10)" lines in the code ;)
[16-Jul-2010 12:50:55] haha
[16-Jul-2010 12:51:17] if environment == theacolyte then sleep(30) at random between 10 and 30 seconds
[16-Jul-2010 12:51:37] what's your hardware env?
[16-Jul-2010 12:52:02] cgibbons: it's a vm with 1 vcpu (2.5ghz 54xx xeon), 2GB of RAM on a SAN
[16-Jul-2010 12:53:41] Not a lot of devices though
[16-Jul-2010 12:53:45] I may have 10k data points
[16-Jul-2010 12:54:23] SAN is fiber attached I would assume?
[16-Jul-2010 12:54:57] forsberg is now known as fOrsberg
[16-Jul-2010 12:55:20] * rmatte wants to sneak up on our Microsoft CRM server and slight it on fire
[16-Jul-2010 12:55:23] rmatte: iscsi
[16-Jul-2010 12:55:46] ah
[16-Jul-2010 12:55:59] slight = light
[16-Jul-2010 12:56:16] worst software evar!
[16-Jul-2010 13:03:00] bigegor: just about to add a device without your packs installed, moment of truth...
[16-Jul-2010 13:04:58] it's still taking forever to load the infrastructure page with 0 devices
[16-Jul-2010 13:06:19] bigegor: it works perfectly without the WMI packs installed
[16-Jul-2010 13:06:28] so it is definitely a problem with one of those packs
[16-Jul-2010 13:06:42] rmatte: what does mysql "show processlist" say when you try and load that infrastructure page?
[16-Jul-2010 13:06:44] just out of interest
[16-Jul-2010 13:07:29] where do I do that?
[16-Jul-2010 13:07:36] just at the mysql console?
[16-Jul-2010 13:07:38] yup
[16-Jul-2010 13:07:48] it'll show you any active queries and how long they're taking
[16-Jul-2010 13:07:57] you'll need a ; on the end
[16-Jul-2010 13:08:07] all I get is another prompt line
[16-Jul-2010 13:08:15] ah, the semicolon :)
[16-Jul-2010 13:08:16] ah, that's why
[16-Jul-2010 13:08:19] sorry
[16-Jul-2010 13:09:01] it loaded faster this last time
[16-Jul-2010 13:09:05] so not sure if that's really helpful
[16-Jul-2010 13:09:39] I was just interested, a lot of the UI slowness I see in 2.5.2 can be traced to an actual slow mysql query
[16-Jul-2010 13:09:45] yeh, there's nothing interesting in there
[16-Jul-2010 13:10:02] ah well, cheers
[16-Jul-2010 13:10:04] there's always the same amount of rows listed
[16-Jul-2010 13:10:21] and they are all the same processes
[16-Jul-2010 13:10:30] right
[16-Jul-2010 13:10:40] if you catch a slow query it'll show the query and the time it's taken
[16-Jul-2010 13:10:48] a well behaved db you'll almost never see anything in there
[16-Jul-2010 13:10:50] yeh, I didn't catch any
[16-Jul-2010 13:10:53] I'll try again later
[16-Jul-2010 13:12:06] rmatte: have you another ZenPacks installed?
[16-Jul-2010 13:12:20] bigegor: lots
[16-Jul-2010 13:13:16] here's the list: http://fpaste.org/9OvA/
[16-Jul-2010 13:14:08] not having any issues with any of those after the upgrade
[16-Jul-2010 13:14:13] so far anyways
[16-Jul-2010 13:14:57] here are the errors I was seeing when adding with the WMI pack again, in case you need them: http://fpaste.org/19Kh/
[16-Jul-2010 13:15:15] and this is the error I got when I tried to remove one of the WMI packs: http://fpaste.org/6ANF/
[16-Jul-2010 13:15:30] which is a good idea for another test, see if I can successfully remove a ZenPack after the migration...
[16-Jul-2010 13:15:45] I need to log a bunch of time in our CRM system first though
[16-Jul-2010 13:16:00] * rmatte yawns
[16-Jul-2010 13:22:59] bigegor_ is now known as bigegor
[16-Jul-2010 13:23:46] rmatte: hmm, you doesn't have any wmi related modeler plugins. How this error can be related with WMIDataSource ZenPack?
[16-Jul-2010 13:27:03] hello
[16-Jul-2010 13:27:14] Simon4: I found most of my slow GUI problems to be caused by the calculation of availability against MySQL stored evnets
[16-Jul-2010 13:27:29] twm1010: yeah, me also
[16-Jul-2010 13:27:45] one for the knowleadable: can device.id be the fqdn or the IP?
[16-Jul-2010 13:27:49] that one query is something unmentionable :)
[16-Jul-2010 13:27:57] It's nice to have dynamic availability, but it would be really nice if that was also stored in RRD, so we can show history of availability over time
[16-Jul-2010 13:28:16] rather than have to calculate it off evnets
[16-Jul-2010 13:29:02] agreed
[16-Jul-2010 13:29:13] I found the code, this could be doable ;)
[16-Jul-2010 13:36:31] Simon4: keep me posted? I'm asked for this alot
[16-Jul-2010 13:37:00] twm1010: hokay
[16-Jul-2010 13:37:20] Kinda interesting if you think about it
[16-Jul-2010 13:37:26] you'd almost want a custom RRD for this
[16-Jul-2010 13:37:34] and a way to have reports that can break it down
[16-Jul-2010 13:37:39] and average it out in a few ways
[16-Jul-2010 13:38:07] so keep a few values stored
[16-Jul-2010 13:38:21] daily availability, weekly, monthly, yearly
[16-Jul-2010 13:38:24] you could make something like a zenavailability daemon
[16-Jul-2010 13:38:27] and then the ability to report on that in histority
[16-Jul-2010 13:38:36] *historically
[16-Jul-2010 13:38:37] which would keep said rrd updated based on nicely indexed queries on mysql
[16-Jul-2010 13:38:40] by organizer
[16-Jul-2010 13:38:52] anyone?
[16-Jul-2010 13:38:54] then you could have multiple data series in teh rrd like you're saying
[16-Jul-2010 13:39:01] so there are community and enterprise organizer availability reports
[16-Jul-2010 13:39:16] chemist: not sure sorry
[16-Jul-2010 13:39:27] np, thanx
[16-Jul-2010 13:39:45] but rather than calculate, have it just lookup
[16-Jul-2010 13:39:51] yeah
[16-Jul-2010 13:40:04] I doubt most people need availability to be calculated more than once per day
[16-Jul-2010 13:40:37] you can actually tweak the caching of those availability queries, but given they can run to 120-150 seconds a query, someone is still the sucker who hits the terrible page load time
[16-Jul-2010 13:40:49] could a single RRD have a daily that rolls into a weekly, that rolls into a monthly, that rolls into a yearly?
[16-Jul-2010 13:40:56] yup
[16-Jul-2010 13:41:23] given you could keep it as a gauge, it would just average it off into the RRA's
[16-Jul-2010 13:41:40] so you get monthly calculated as the average of the last month's dailies
[16-Jul-2010 13:42:15] sure, if you only query once a day, it should stay pretty tidy
[16-Jul-2010 13:42:33] yeah, and you could just have a background task to populate the rrd
[16-Jul-2010 13:42:37] so users never hit that query
[16-Jul-2010 13:43:50] I'm not a coder, but it sounds like this would be fairly simple
[16-Jul-2010 13:44:02] it wouldn't be horribly hard, that's for sure
[16-Jul-2010 13:44:09] * Simon4 takes a copy of the log for later
[16-Jul-2010 13:44:12] I'm keen to give it a go
[16-Jul-2010 13:44:15] im trying to think of what kind of reports could come from this
[16-Jul-2010 13:44:24] by organizer would be nice
[16-Jul-2010 13:44:41] yeah
[16-Jul-2010 13:44:56] or by system or group folder, etc
[16-Jul-2010 13:45:04] a monthly availability trend graph, by say location, or system, would be very useful
[16-Jul-2010 13:45:21] Great, my 3.0 core binary installation is hung at 100%, been 15 minutes and it hasn't installed the uninstall script yet :|
[16-Jul-2010 13:45:26] eff
[16-Jul-2010 13:45:34] varnii: go get a beer :)
[16-Jul-2010 13:46:34] -__- long day of file hunting haha
[16-Jul-2010 13:47:32] server fart, wow 30 min install but it just finished lol
[16-Jul-2010 13:48:20] bigegor: you're not understanding...
[16-Jul-2010 13:48:34] what I did was, I had the WMI packs installed (all the latest ones) and I upgraded to 3.0
[16-Jul-2010 13:48:37] then I got that error
[16-Jul-2010 13:48:46] so since it's a VM, I had a snapshot of it when it was 2.5.2
[16-Jul-2010 13:48:49] so I rolled it back
[16-Jul-2010 13:48:59] then removed all the WMI packs (since it wouldn't let me uninstall them in 3.0)
[16-Jul-2010 13:49:04] then I did the same upgrade
[16-Jul-2010 13:49:07] and the error was gone
[16-Jul-2010 13:49:13] so it clearly has something to do with the WMI packs
[16-Jul-2010 13:49:40] rmatte: or with upgrade...
[16-Jul-2010 13:51:00] rmatte: which WMIDataSource version has you on zenoss 2.5.2
[16-Jul-2010 13:52:40] bigegor: I upgraded to the absolute latest
[16-Jul-2010 13:52:42] 2.7
[16-Jul-2010 13:52:46] before upgrading to 3.0
[16-Jul-2010 13:53:08] maybe if I remove the WMI packs, and install them after the upgrade they'll work properly
[16-Jul-2010 13:53:19] maybe it's the upgrade itself while having the packs installed that's messing something up
[16-Jul-2010 13:53:28] rmatte: try this
[16-Jul-2010 14:01:21] I'm all ears
[16-Jul-2010 14:02:04] rmatte: reinstall wmi zenpacks after upgrade
[16-Jul-2010 14:02:12] k
[16-Jul-2010 14:02:35] hey quick question.... the 'zenoss core network and system monitoring book' that was published in 2008
[16-Jul-2010 14:02:39] is it still reasonably up to date ?
[16-Jul-2010 14:02:46] in terms of the device / event classes ?
[16-Jul-2010 14:03:25] if not, can anyone recommend another zenoss book ?
[16-Jul-2010 14:03:29] otakup0pe: yes, the interface has obviously drastically changed in 3.0
[16-Jul-2010 14:03:38] but everything else is pretty much the same
[16-Jul-2010 14:03:40] oh 3.0 just came out too didn't it hmm
[16-Jul-2010 14:03:58] and there is no other Zenoss book yet, one is in the works, but they are waiting for 3.0 to mature enough
[16-Jul-2010 14:04:31] ok so i'm recommending it to the sysadmin at work since it seems like the concepts are all the same, just the gooey changed
[16-Jul-2010 14:04:34] thx rmatte :)
[16-Jul-2010 14:06:28] np
[16-Jul-2010 14:06:44] anyone running zenoess on EC2?
[16-Jul-2010 14:06:55] not to monitor ec2, but running on an instance
[16-Jul-2010 14:06:57] zenoss*
[16-Jul-2010 14:07:10] that how my jumpbox is conf'd (but it runs a stripped ubuntu)
[16-Jul-2010 14:07:19] I'm guessing I can just upgrade via the stack installer for ubuntu
[16-Jul-2010 14:07:28] although might just go with a fresh install
[16-Jul-2010 14:07:35] I only have about 40 devices to model
[16-Jul-2010 14:07:51] willwh_ is now known as willwh
[16-Jul-2010 14:08:04] bigegor: no error if I install it after the upgrade
[16-Jul-2010 14:08:43] I did have the exchange and IIS packs installed too though
[16-Jul-2010 14:08:46] haven't installed those yet
[16-Jul-2010 14:08:49] going to do it right now
[16-Jul-2010 14:09:08] rmatte: as you can see UPGRADE is a problem
[16-Jul-2010 14:09:15] right
[16-Jul-2010 14:09:35] which is not such a good thing
[16-Jul-2010 14:10:37] agreed
[16-Jul-2010 14:12:13] eugh, this upgrade is going to be a nightmare if there are problems with other ZenPacks
[16-Jul-2010 14:16:55] ;\
[16-Jul-2010 14:18:32] ok, so note to self, remove WMI packs prior to upgrade, and reinstall them afterwards
[16-Jul-2010 14:20:02] hello I'm trying to create a CommandPLugin to model disk data, and I need to pull some zProperties -- what is the best way to go about doing this?
[16-Jul-2010 14:20:16] collectorplugin i guess i should call it
[16-Jul-2010 14:20:28] d = dmd.Devices.findDevice('devicename')
[16-Jul-2010 14:20:31] d.zWinUser
[16-Jul-2010 14:20:33] or whatever
[16-Jul-2010 14:20:36] it's pretty simple
[16-Jul-2010 14:20:37] o nice
[16-Jul-2010 14:20:40] awesome
[16-Jul-2010 14:20:56] whats the proper way to import dmd
[16-Jul-2010 14:21:14] sry just started hacking away at zenoss yesterday
[16-Jul-2010 14:21:23] ring: docs/DOC-3217;jsessionid=80D27122402B508FC8040FBA903347A4.node0
[16-Jul-2010 14:24:12] ty
[16-Jul-2010 14:25:09] hmm
[16-Jul-2010 14:25:13] i'm passed the device attribute though
[16-Jul-2010 14:25:13] already
[16-Jul-2010 14:25:14] so it looks like
[16-Jul-2010 14:25:14] err device object
[16-Jul-2010 14:25:42] but the zattribute i am looking for doesnt appear to be there
[16-Jul-2010 14:25:57] what do you mean by "passed" it?
[16-Jul-2010 14:26:20] which zProperty?
[16-Jul-2010 14:26:28] zProperty i'm looking for is
[16-Jul-2010 14:26:38] 'zFileSystemMapIgnoreNames', 'zFileSystemMapIgnoreTypes'
[16-Jul-2010 14:26:45] i do dir(device) and it doesnt show
[16-Jul-2010 14:26:56] and doing getattr does nothing either
[16-Jul-2010 14:27:02] so it looks like for w/e reason its not being passed
[16-Jul-2010 14:27:54] >>> d = dmd.Devices.findDevice('gen01.novanoc.com')
[16-Jul-2010 14:27:54] >>> d.zFileSystemMapIgnoreNames
[16-Jul-2010 14:27:54] ''
[16-Jul-2010 14:27:54] >>> d.zFileSystemMapIgnoreTypes
[16-Jul-2010 14:27:54] ['other', 'ram', 'virtualMemory', 'removableDisk', 'floppyDisk', 'compactDisk', 'ramDisk', 'flashMemory']
[16-Jul-2010 14:27:55] >>>
[16-Jul-2010 14:28:09] meh
[16-Jul-2010 14:28:10] >>> d.zFileSystemMapIgnoreTypes[1]
[16-Jul-2010 14:28:10] 'ram'
[16-Jul-2010 14:28:10] >>> d.zFileSystemMapIgnoreTypes[2]
[16-Jul-2010 14:28:10] 'virtualMemory'
[16-Jul-2010 14:28:10] >>> d.zFileSystemMapIgnoreTypes[3]
[16-Jul-2010 14:28:10] 'removableDisk'
[16-Jul-2010 14:28:34] you could do something like: for t in d.zFileSystemMapIgnoreTypes:
[16-Jul-2010 14:28:43] you're not seeing any values listed for those?
[16-Jul-2010 14:28:49] it's possible that they are blank in Zenoss for you
[16-Jul-2010 14:29:03] though they zFileSystemMapIgnoreTypes should not be by default
[16-Jul-2010 14:30:12] i get
[16-Jul-2010 14:30:13] DeviceProxy instance has no attribute 'zFileSystemMapIgnoreTypes'
[16-Jul-2010 14:30:20] ring: i have better solution: add to your modeller class deviceProperties = CommandPlugin.deviceProperties + ( 'zFileSystemMapIgnoreNames', 'zFileSystemMapIgnoreTypes')
[16-Jul-2010 14:31:05] dunno what the difference is between the dmd device and DeviceProxy is
[16-Jul-2010 14:31:09] i assume thats probably my issue though
[16-Jul-2010 14:31:28] yeh, DeviceProxy is probably completely different
[16-Jul-2010 14:31:36] try bigegor's suggestion
[16-Jul-2010 14:31:41] he knows more about this than I do
[16-Jul-2010 14:31:41] ring: and in process method you can use it with getattr(device, 'zFileSystemMapIgnoreTypes', None)
[16-Jul-2010 14:39:11] bigegor: so the WMI Data Source pack can't monitor filesystems?
[16-Jul-2010 14:40:14] rmatte: WMI Data Source doesn't provide any templates
[16-Jul-2010 14:40:30] right, but I thought it provided full modelling?
[16-Jul-2010 14:40:58] templates have nothing to do with monitoring filesystems, interfaces, processes, etc...
[16-Jul-2010 14:41:00] rmatte: yes
[16-Jul-2010 14:41:03] that's all done via collector plugin
[16-Jul-2010 14:41:31] I guess the FileSystems template and the Interface templates would though
[16-Jul-2010 14:41:36] hmm
[16-Jul-2010 14:41:37] rmatte: WMI Data Source doesn't provide any collector plugins too
[16-Jul-2010 14:41:59] I see, so it theoretically has the capability to fully model a device, but it doesn't actually do it?
[16-Jul-2010 14:44:28] rmatte: use WMIPerf_Windows Zenpacj if you need this
[16-Jul-2010 14:44:35] Zenpack
[16-Jul-2010 14:45:13] rmatte: docs/DOC-3500
[16-Jul-2010 14:46:11] right, I already have that pack installed
[16-Jul-2010 14:46:15] k cool
[16-Jul-2010 14:47:32] WMI modeling infrastructure is in straight core. you can write a new WMI plugin easily without any extra zenpacks.
[16-Jul-2010 14:48:08] I see
[16-Jul-2010 14:48:16] only for root/cimv2 namespace ;)
[16-Jul-2010 14:49:12] bigegor: which modeler do i add that to?
[16-Jul-2010 14:51:06] ring: which modeler?
[16-Jul-2010 14:51:14] hmm
[16-Jul-2010 14:51:16] so i added
[16-Jul-2010 14:51:23] egor, we should see if we can get the features you need for Query, etc. put into core for 3.1
[16-Jul-2010 14:51:36] deviceProperties = CommandPlugin.deviceProperties... to the CommandPlugin class
[16-Jul-2010 14:51:47] i now have attributes put they are not poppulated from the zProperties
[16-Jul-2010 14:52:05] i am sorry if my quetions arent real clear i'm kind of new to zenoss
[16-Jul-2010 14:52:40] cgibbons: np. do it.
[16-Jul-2010 14:54:03] cgibbons: wait... you can do WMI natively now?
[16-Jul-2010 14:54:18] root/cimv2 is fine with me
[16-Jul-2010 14:55:25] ring: if you want create new CommandPlugin, you will create new class yourCmdPlugin(CommandPlugin) right?
[16-Jul-2010 14:55:44] yes
[16-Jul-2010 14:55:57] ishould have been clearer i put the lines in the LinuxCommandPlugin class
[16-Jul-2010 14:57:41] ring: you must redefine property 'deviceProperties' in your class
[16-Jul-2010 14:57:55] ok
[16-Jul-2010 14:58:19] ring simply add this line to you class: deviceProperties = CommandPlugin.deviceProperties + ( 'zFileSystemMapIgnoreNames', 'zFileSystemMapIgnoreTypes')
[16-Jul-2010 14:59:21] ring: it takes deviceProperties from 'CommandPlugin' plus your properties
[16-Jul-2010 14:59:31] ah
[16-Jul-2010 15:00:08] ring: deviceProperties type is tuple
[16-Jul-2010 15:01:15] let me pastebin my code for you
[16-Jul-2010 15:01:21] man, that not being able to shift click to select in the event console is really really annoying
[16-Jul-2010 15:02:02] bigegor: this is what i have http://pastebin.com/7MbbgE1R
[16-Jul-2010 15:02:23] skipfsnames and skipfstypes appear to be empty
[16-Jul-2010 15:03:01] o woops
[16-Jul-2010 15:03:02] i'm an idiot
[16-Jul-2010 15:03:04] pasted wrong thing
[16-Jul-2010 15:04:09] why? looks good for me ;)
[16-Jul-2010 15:04:23] it errors
[16-Jul-2010 15:05:02] ok
[16-Jul-2010 15:05:03] so ihave
[16-Jul-2010 15:05:13] http://pastebin.com/tU5igrQJ in /usr/local/zenoss/zenoss/Products/DataCollector/plugins/zenoss/cmd/linux
[16-Jul-2010 15:05:19] when i model i get this error
[16-Jul-2010 15:06:24] rmatte: yes, it really, really is
[16-Jul-2010 15:06:27] no values are returned when i log zFileSystemMapIgnoreNAmes
[16-Jul-2010 15:06:33] rmatte: it's actually driving me slightly insane
[16-Jul-2010 15:06:40] rmatte: In fact, if I could roll back, I may even do that
[16-Jul-2010 15:07:44] woohoo
[16-Jul-2010 15:07:53] just did my first fully successful modelling via WMI
[16-Jul-2010 15:08:19] grats
[16-Jul-2010 15:08:34] rmatte: I thought I was imagining things with the shift click
[16-Jul-2010 15:08:35] bigegor: nice work
[16-Jul-2010 15:08:49] twm1010: nah, that bug has been mentioned to ian before
[16-Jul-2010 15:08:50] rmatte: thanks ;)
[16-Jul-2010 15:08:54] I'm surprised it didn't get fixed
[16-Jul-2010 15:08:57] ctrl click still works tho
[16-Jul-2010 15:09:06] yeh, but ctrl-click is tedious
[16-Jul-2010 15:09:14] You may as well do it one by one if you do that
[16-Jul-2010 15:10:18] ring: skipfsnames = getattr(device, 'zFileSystemMapIgnoreNames', None)
[16-Jul-2010 15:11:57] k
[16-Jul-2010 15:12:28] ok i did that
[16-Jul-2010 15:12:34] the values are None
[16-Jul-2010 15:12:47] it doesnt seem to be reading it from the zattributes
[16-Jul-2010 15:17:13] ill keep digging
[16-Jul-2010 15:17:18] thank you for all your help egor
[16-Jul-2010 15:17:36] ring: np
[16-Jul-2010 15:25:28] bigegor: tip for the next release... in the interface errors graph, you have "Recive" and "Send"
[16-Jul-2010 15:25:36] that should be "Received" "Sent"
[16-Jul-2010 15:25:50] or just "Input" "Output"
[16-Jul-2010 15:26:11] :)
[16-Jul-2010 15:26:26] or "Inbound" "Outbound"
[16-Jul-2010 15:26:29] to match the others
[16-Jul-2010 15:26:29] rmatte: thanks. can you check other things?
[16-Jul-2010 15:26:31] would probably be best
[16-Jul-2010 15:26:37] I wonder how hard it would be to fix that shift issue
[16-Jul-2010 15:26:42] yeh, I'll let you know as I come across anything
[16-Jul-2010 15:26:47] can someone look at this? http://pastebin.ca/1902003
[16-Jul-2010 15:26:48] I thought shift selecting was inherant
[16-Jul-2010 15:26:58] so far the other stuff looks good
[16-Jul-2010 15:27:06] the transform is take straight from the zenoss blog
[16-Jul-2010 15:27:18] I'm liking it so far, but rmatte - I'm getting htat loading issue from time to itme in firefox
[16-Jul-2010 15:27:51] theacolyte: which one?
[16-Jul-2010 15:28:09] the devices Loading. Please wait...
[16-Jul-2010 15:28:43] from here to be precise:
[16-Jul-2010 15:29:19] blogs/zenossblog/2009/05/28/tip-of-the-month-event-suppression
[16-Jul-2010 15:31:32] chemist: checking...
[16-Jul-2010 15:31:40] theacolyte: yeh, they really need to fix that
[16-Jul-2010 15:32:10] I think if that shift thing was fixed and that I wouldn't have any issues
[16-Jul-2010 15:32:20] therre's also a bit of a curve learning the new stuff, but so far it's not horrendus
[16-Jul-2010 15:32:23] chemist: ummm, that's a bit weird
[16-Jul-2010 15:32:24] Just need to make sure you have time
[16-Jul-2010 15:32:33] Definately a major version release
[16-Jul-2010 15:32:50] chemist: you're not setting device to something?
[16-Jul-2010 15:33:32] hmmmm
[16-Jul-2010 15:33:39] chemist: I think those transforms are ancient
[16-Jul-2010 15:33:45] chemist: what exactly are you trying to do?
[16-Jul-2010 15:34:35] suppress all events from a locationn when device x (router1.bellevue etc in this case) is down
[16-Jul-2010 15:35:01] and where did you put the transform?
[16-Jul-2010 15:35:07] did you put it at the root level as explained?
[16-Jul-2010 15:35:28] no, /Events/Status/Ping
[16-Jul-2010 15:35:40] it says right in that document to put it at the root level
[16-Jul-2010 15:35:53] so the odd functions he's using probably only work at that level
[16-Jul-2010 15:35:54] the root level one is for a different transform
[16-Jul-2010 15:36:07] oh
[16-Jul-2010 15:36:19] rmatte: done (SVN ver. 2.2.1)
[16-Jul-2010 15:36:36] that's so weird how he's done that
[16-Jul-2010 15:36:42] bigegor: cool
[16-Jul-2010 15:37:03] chemist: let me rewrite that for you real quick
[16-Jul-2010 15:37:09] I am trying to achieve= if device x at such location is ping down, suppress all other events for that location
[16-Jul-2010 15:38:23] ok, so router1.bellevue.datatote.co.uk is the head end router for the Belle Vue site I'd imagine
[16-Jul-2010 15:38:40] correct
[16-Jul-2010 15:38:44] I see what you're getting at, the code just isn't right, I don't see how this could have ever worked lol, I'll fix it up
[16-Jul-2010 15:39:02] you're a diamond geezer :)
[16-Jul-2010 15:39:22] when are you getting your zenoss master title?
[16-Jul-2010 15:39:29] I've had it for a while
[16-Jul-2010 15:39:30] ;)
[16-Jul-2010 15:39:40] oops :)
[16-Jul-2010 15:39:53] I'm just not in the list that shows up when you search for it
[16-Jul-2010 15:41:31] * rmatte yawns while he waits for zendmd to load
[16-Jul-2010 15:43:11] k wtf, zendmd is taking a ridiculous amount of time to load on this one server, time to switch to a different one
[16-Jul-2010 15:43:18] ah, it finally loaded
[16-Jul-2010 15:44:26] chemist: http://pastebin.ca/1902018
[16-Jul-2010 15:44:29] that should work
[16-Jul-2010 15:45:13] instead of evt.eventState = 2...
[16-Jul-2010 15:45:22] I'd do: evt._action = 'drop'
[16-Jul-2010 15:45:24] or
[16-Jul-2010 15:45:31] evt._action = 'history'
[16-Jul-2010 15:45:47] your choice
[16-Jul-2010 15:46:16] bigegor: i restarted zenoss and everything works now, thank you very much
[16-Jul-2010 15:46:39] actually
[16-Jul-2010 15:46:42] ring:np
[16-Jul-2010 15:46:44] it could be simplified even further
[16-Jul-2010 15:48:58] one sec
[16-Jul-2010 15:49:02] just verifying some stuff
[16-Jul-2010 15:51:04] chemist: http://pastebin.ca/1902025
[16-Jul-2010 15:51:05] voila
[16-Jul-2010 15:51:20] much simpler and will actually work
[16-Jul-2010 15:53:19] * chemist is checking
[16-Jul-2010 15:54:39] hmmm, evt.eventState = 2 might actually be better
[16-Jul-2010 15:54:45] since that acknowledges the events
[16-Jul-2010 15:54:49] rather than moving them to history
[16-Jul-2010 15:55:04] or you could outright drop them, but that'll affect your availability stats
[16-Jul-2010 15:57:23] drop sounds good to me
[16-Jul-2010 15:57:29] and it works :)
[16-Jul-2010 15:57:42] what do you drink?
[16-Jul-2010 15:57:49] lol
[16-Jul-2010 15:57:56] keith's
[16-Jul-2010 15:58:00] ??
[16-Jul-2010 15:58:06] it's a beer
[16-Jul-2010 15:58:11] ah, ok
[16-Jul-2010 15:58:13] lol
[16-Jul-2010 15:58:34] http://www.keiths.ca/
[16-Jul-2010 15:58:53] don't know how, but I will get you a case somehow
[16-Jul-2010 15:58:58] lol
[16-Jul-2010 15:59:07] all good, not necessary :)
[16-Jul-2010 15:59:29] thank you very much :)
[16-Jul-2010 15:59:37] if you want to drop them just change 'history' to 'drop'
[16-Jul-2010 15:59:39] you're welcome
[16-Jul-2010 16:00:12] I have a week off starting tuesday, no work, kids, or wife, in a sleepy little spanish town
[16-Jul-2010 16:00:25] nice
[16-Jul-2010 16:00:31] you will not believe how many python pdf's I have
[16-Jul-2010 16:00:36] haha
[16-Jul-2010 16:00:59] I just have the O'Reilly book titled Learning Python
[16-Jul-2010 16:01:02] it's good for reference
[16-Jul-2010 16:01:22] I use google a lot too
[16-Jul-2010 16:01:29] and I chat in #python on here the odd time
[16-Jul-2010 16:01:33] though they are rude in there
[16-Jul-2010 16:01:34] lol
[16-Jul-2010 16:01:45] I had noticed that
[16-Jul-2010 16:01:56] of th python channel lol
[16-Jul-2010 16:02:09] "Oh, you don't know advanced python? WELL, we're CLEARLY better than you!"
[16-Jul-2010 16:02:16] or at least that's how it feels every time I go in there
[16-Jul-2010 16:02:34] but they do help
[16-Jul-2010 16:02:50] it's like getting advice from a bunch of starbucks lackeys
[16-Jul-2010 16:03:06] stuckup assistance
[16-Jul-2010 16:03:07] :P
[16-Jul-2010 16:03:32] let's hope they are not eavesdropping in the channel :)
[16-Jul-2010 16:03:37] meh
[16-Jul-2010 16:03:50] I tell it how it is
[16-Jul-2010 16:03:52] lol
[16-Jul-2010 16:05:03] eugh, can't wait until work is over
[16-Jul-2010 16:05:06] so in that transform dc1_router = etc etc, is just defining the variable dc1_router?
[16-Jul-2010 16:05:09] concert tonight then paintball tomorrow
[16-Jul-2010 16:05:09] woot
[16-Jul-2010 16:05:17] correct
[16-Jul-2010 16:05:30] it's defining dc1_router as the "path" to that device
[16-Jul-2010 16:05:37] which allows you to query it for properties
[16-Jul-2010 16:05:45] which is why you're able to get the ping status
[16-Jul-2010 16:05:52] so it can be any arbitrary name i want, instead of dc1_router
[16-Jul-2010 16:05:53] via dc1_router.getPingStatus()
[16-Jul-2010 16:05:57] yessir
[16-Jul-2010 16:05:59] ok
[16-Jul-2010 16:06:04] make sure you change it in both spots
[16-Jul-2010 16:06:11] yep
[16-Jul-2010 16:07:57] could I use an IP address instead of a device name?
[16-Jul-2010 16:12:12] in the find function you mean?
[16-Jul-2010 16:12:28] yes
[16-Jul-2010 16:12:32] yes you can
[16-Jul-2010 16:12:35] IP or hostname
[16-Jul-2010 16:18:45] lmao
[16-Jul-2010 16:18:46] http://pastebin.com/GSCEcd76
[16-Jul-2010 16:18:52] chat transcript from #python
[16-Jul-2010 16:20:36] sounds like a very serious channel......lol
[16-Jul-2010 16:20:53] Ah, python geeks
[16-Jul-2010 16:20:54] cute
[16-Jul-2010 16:24:59] Damn you shift button
[16-Jul-2010 16:25:00] hahahaha
[16-Jul-2010 16:25:04] This is almost unusable
[16-Jul-2010 16:25:10] I may just let events stack up and pray for a patch
[16-Jul-2010 16:25:31] rmatte: you're a programmer - fix it!
[16-Jul-2010 16:26:40] rmatte: can I use a regex in the device name?
[16-Jul-2010 16:26:49] theacolyte: I might actually take a stab at it
[16-Jul-2010 16:27:05] chemist: in which part of the transform?
[16-Jul-2010 16:27:17] the find function?
[16-Jul-2010 16:27:38] or for the matching?
[16-Jul-2010 16:27:50] fro both
[16-Jul-2010 16:29:28] not for the find
[16-Jul-2010 16:29:48] well, the problem arises with, what if it finds more than 1 device?
[16-Jul-2010 16:29:49] we use router1.blablabla forthe vpn endpoint and router1ext.blablabla for the external ip
[16-Jul-2010 16:30:00] oh, I see
[16-Jul-2010 16:30:05] so you want to dynamically extract
[16-Jul-2010 16:30:11] I have to go for the day...
[16-Jul-2010 16:30:17] but I can help you with that on Monday
[16-Jul-2010 16:30:34] ok, thanks for yor efforts
[16-Jul-2010 16:30:50] enjoy your concert, and paitballing
[16-Jul-2010 16:31:45] I shall
[16-Jul-2010 16:31:47] :)
[16-Jul-2010 16:31:58] I get to shoot coworkers with my new paintball gun, woot
[16-Jul-2010 16:32:17] may the force be with you lol
[16-Jul-2010 16:32:40] lol
[16-Jul-2010 16:32:47] oh god I need to go
[16-Jul-2010 16:32:51] I need to paintball
[16-Jul-2010 16:34:01] rmatte: How hard was LDAP in 3?
[16-Jul-2010 16:34:06] Allegeldy it's much easier
[16-Jul-2010 16:35:33] theacolyte: if it is easier I would like to meet the people who make that claim
[16-Jul-2010 16:35:37] since I've yet to figure it out
[16-Jul-2010 16:35:49] lol
[16-Jul-2010 16:35:54] thread/13980?tstart=0
[16-Jul-2010 16:36:56] I'll send him a msg on monday and get the steps from him so that I can write up a proper guide
[16-Jul-2010 16:37:06] I'd like to do it this time
[16-Jul-2010 16:43:30] go for it, as long as the guide gets written
[16-Jul-2010 16:45:13] no ... the intention is to follow YOUR guide
[16-Jul-2010 16:45:14] ;)
[16-Jul-2010 16:45:42] hey rmatte, quick question
[16-Jul-2010 16:45:59] so now thinking, I can't easily do a fresh install - so thinking about upgrading
[16-Jul-2010 16:46:08] running 2.5.0 installed from a jumpbox-zenoss.deb
[16-Jul-2010 16:46:16] for which their doesn't seem to be an updated version
[16-Jul-2010 16:46:38] kinda jammed on which way to go now :)
[16-Jul-2010 16:46:52] do I install 2.5.0 native (from stack installer) - restore my backup - then start upgrading?
[16-Jul-2010 16:46:57] that sounds like a bit of a nightmare
[16-Jul-2010 16:56:50] Anyone know why 'Test Data Source' in the 3.0 GUI is just a black screen, and nothing happens?
[16-Jul-2010 16:59:32] it should work
[16-Jul-2010 17:00:22] yeah, same template commands that worked in 2.5.2, I'll keep digging
[16-Jul-2010 17:02:53] Hi everyone, I'm getting this: ""threshold of high utilization exceeded: current value 1365288.38" in two interfaces
[16-Jul-2010 17:03:13] of course, value is differente in the other interface
[16-Jul-2010 17:03:59] does anyone know what could be?
[16-Jul-2010 17:05:14] oh, and in eventKey I got this: ifOutOctets_ifOutOctets|high utilization
[16-Jul-2010 17:05:41] mrayzenoss: yeah, the commands run perfectly from the command line, there's no text at all in the testdatasource box
[16-Jul-2010 17:05:58] anyone else seen/fixed this issue?
[16-Jul-2010 17:06:31] varnii: which browser?
[16-Jul-2010 17:06:45] reportedly webkit ones have a few new issues
[16-Jul-2010 17:06:46] chrome, ah right I should have checked that
[16-Jul-2010 17:08:15] mrayzenoss: yeah, thanks
[16-Jul-2010 17:08:17] chrome is the issue
[16-Jul-2010 17:08:25] works in firefox
[16-Jul-2010 17:08:47] chrome has issues
[16-Jul-2010 17:08:48] yes
[16-Jul-2010 17:09:16] :) <3
[16-Jul-2010 17:12:26] haha
[16-Jul-2010 17:12:34] I just had amoment where I'm looking at my graphs and saw a gap for today
[16-Jul-2010 17:12:39] then realized that's because I was updating to 3.0
[16-Jul-2010 20:49:19] great work guys, upgrading to 3.0 now :)
[16-Jul-2010 20:56:12] uh oh
[16-Jul-2010 21:13:27] question: it didnt say in the install guide, what happens if after 5min or so theres nothing in you zencatalog log?
[16-Jul-2010 21:14:37] i've only got like 15 devices lol
[16-Jul-2010 21:15:10] nevermind, was user error
[16-Jul-2010 22:35:45] wow
[16-Jul-2010 22:35:51] the new webgui is much nicer, great work guys
[16-Jul-2010 23:09:39] Parabola: Dammit you are making me want to play w/Zenoss ;-)
[17-Jul-2010 00:00:30] [disconnected at Sat Jul 17 00:00:30 2010]
[17-Jul-2010 00:00:30] [connected at Sat Jul 17 00:00:30 2010]
[17-Jul-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]
[17-Jul-2010 00:00:43] [freenode-info] please register your nickname...don't forget to auto-identify! http://freenode.net/faq.shtml#nicksetup
[17-Jul-2010 00:25:15] One question and one question only about Z3.0 - when I select one or more events in Event Console. Can I send them to history by pressing the DEL key?
[17-Jul-2010 00:25:57] (I fall into that trap *every* time. I press DEL. Then I remember, 'oh yeah, gotta move mouse and click the button, bummer')
[17-Jul-2010 01:41:08] fOrsberg is now known as forsberg
[17-Jul-2010 05:26:06] <`nstuff> hi can someone explain how i change the device class in v3?
[17-Jul-2010 05:26:28] <`nstuff> in the docs it states you just drag them, but no joy :/
[17-Jul-2010 08:50:47] forsberg is now known as fOrsberg
[17-Jul-2010 11:08:01] hi there
[17-Jul-2010 11:08:32] guys.. i'm trying to use winexe to run a command on a vbox machine without a domain
[17-Jul-2010 11:09:02] the problem is I keep getting ERROR: Failed to open connection - NT_STATUS_LOGON_FAILURE
[17-Jul-2010 11:09:24] i'm running winexe -W . -U 10.1.1.103/suportedi //10.1.1.103 "cmd.exe"
[17-Jul-2010 11:11:07] any idea on what I'm doing wrong?!
[17-Jul-2010 12:34:21] fOrsberg is now known as forsberg
[17-Jul-2010 16:34:27] [disconnected at Sat Jul 17 16:34:27 2010]
[18-Jul-2010 00:00:31] [connected at Sun Jul 18 00:00:31 2010]
[18-Jul-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]
[18-Jul-2010 15:06:27] OK, so I am getting "SNMP agent down" events against Windows 2003R2 and 2008R2 servers. This is true even when the servers have firewall disabled and SNMP agent configured properly. snmpwalk from the zenoss server fetches data from these servers. othe W2003R2, w2008R2 servers do not experience snmp down.
[18-Jul-2010 15:06:49] Anyone seen this and know of a solution? Kind of infuriating
[18-Jul-2010 15:26:02] mota: check zenperfsnmp.log (in $ZENHOME/logs) and see if perhaps the snmp collection is failing sometimes - maybe the timeout is a little short?
[18-Jul-2010 15:35:33] timeout is at 10 seconds
[18-Jul-2010 15:35:59] and zsnmptires=3
[18-Jul-2010 15:36:07] *zsnmptries
[18-Jul-2010 15:36:14] checking that log, hangon
[18-Jul-2010 15:40:57] aha. 2010-07-18 13:37:35,944 WARNING zen.zenperfsnmp: Failed to collect on myhost.mydomain.com (exceptions.ValueError: invalid literal for int(): C)
[18-Jul-2010 15:42:39] apparently the script interprets these failures to parse driveletters as snmp down
[18-Jul-2010 15:43:39] what's odd though, is that it is reporting the same failure for another server which it does NOT do the snmp agent down alert on. Hmm.
[18-Jul-2010 15:43:50] I wonder if this should be filed as a bug?
[18-Jul-2010 15:45:03] the other thing to check is whether you're getting multiple "agent down" events, since they won't auto-clear
[18-Jul-2010 15:45:19] so a single event won't actually go away, even if snmp then comes back up or the error is resolved et
[18-Jul-2010 15:45:21] etc
[18-Jul-2010 15:45:45] yes, multiple agent down events
[18-Jul-2010 15:48:29] it's probably worth asking people in here tomorrow - there are a few regulars in here on weekdays who run snmp against a lot of windows boxes
[18-Jul-2010 15:48:35] not so many people about on a Sunday
[18-Jul-2010 15:50:34] yeah, will do; at this rate I don't see a solution appearing today
[18-Jul-2010 16:16:01] Thanks for the suggestions, Simon4
[18-Jul-2010 16:16:41] mota: no probs, sorry I couldn't help more
[18-Jul-2010 19:05:05] hmm, how can I run zenperfsnmp manually for testing?
[18-Jul-2010 20:23:20] mota: it should have some flags if you run it from the command line
[18-Jul-2010 20:23:42] mota: make sure you are running it as zenoss (sudo su - zenoss)
[18-Jul-2010 20:23:49] And I have no clue what they are (zenperfsnmp)
[18-Jul-2010 20:23:58] What do you mean?
[18-Jul-2010 20:24:08] Oh - it's basically a python script
[18-Jul-2010 20:24:17] yes, and?
[18-Jul-2010 20:24:40] I'm not sure what you are asking then.
[18-Jul-2010 20:24:58] what args should I use with zenperfsnmp to make it run manually?
[18-Jul-2010 20:25:15] That I'm not sure - that's the it should have some flags part
[18-Jul-2010 20:25:28] I've never had cause to do it with zenperfsnmp
[18-Jul-2010 20:25:32] google reveals nothing but things that do not work
[18-Jul-2010 20:25:35] I have with zenperfwmi and I know that one has something
[18-Jul-2010 20:26:27] ahhh. ok, sudo to zenoss user worked, it will not run as root
[18-Jul-2010 20:26:31] right
[18-Jul-2010 20:26:33] at least, not without error
[18-Jul-2010 20:27:14] this speeds up troubleshooting anyway
[18-Jul-2010 22:46:14] So I just couldn't resolve it. I ended up turning off snmp monitoring for /devices/windows entirely; wmi does a fine job I think
[18-Jul-2010 22:47:35] err /devices/server/windows
[18-Jul-2010 22:48:06] I'll noodle with it again when try zenoss 3
[19-Jul-2010 00:00:30] [disconnected at Mon Jul 19 00:00:30 2010]
[19-Jul-2010 00:00:30] [connected at Mon Jul 19 00:00:30 2010]
[19-Jul-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]
[19-Jul-2010 04:18:04] hello, anyone available for a simple (i hope) support request?
[19-Jul-2010 04:20:52] i'm trying to upgrade from 3.0beta2 to 3.0final (RHEL5 rpm), but the preupgrade zenpack won't install, and the final rpm complains about missing preupgrade
[19-Jul-2010 04:23:37] I'm pretty sure there isn't an upgrade path from the betas to the final
[19-Jul-2010 04:23:41] only from 2.5.2
[19-Jul-2010 04:24:46] so you're saying i should completely remove the beta and install the final? do you think i will be able to retain my config?
[19-Jul-2010 04:28:13] I'm not sure, I just remember seeing warnings saying not to run poduction on the beta because it wasn't upgradeable.
[19-Jul-2010 04:28:42] ok, thanks
[19-Jul-2010 04:29:10] I would post on the forums and ask the question, maybe there is a migration path that you could use
[19-Jul-2010 04:29:23] worst case you could certainly keep your rrd files anyway (which drive all the graphing)
[19-Jul-2010 04:33:22] that would be good. but it's not critical, mostly i wanted to avoid reconfiguring every device, if there is a chance to migrate the config to the new version.
[19-Jul-2010 04:45:04] i posted the question on zenoss-testing, will wait for answers there. thanks again.
[19-Jul-2010 04:46:01] damn, i mis-posted on zenoss-users actually :(
[19-Jul-2010 04:46:48] moved
[19-Jul-2010 06:21:30] why does zenoss still try snmp queries against a device thats down and how do I make it stop ?
[19-Jul-2010 06:28:15] tehhobbit: blogs/zenossblog/2009/05/28/tip-of-the-month-event-suppression
[19-Jul-2010 06:28:32] I think that is what you are looking for
[19-Jul-2010 06:42:31] tnx but not really, what happens is that perfsnmp takes to long to complete due to waiting for timeout on devs it knows are down / have snmp agent down
[19-Jul-2010 06:42:53] but it was a nice link will fix that one too straight away
[19-Jul-2010 06:55:16] Has anyone had issues with with modifying a data source type from gauge to counter and it not taking effect if the template was already bound to a device. I couldn't get the points to graph correctly until the device (and performance data) was removed, the the device was re-added and template rebound.
[19-Jul-2010 06:57:13] did you try to model the device ?
[19-Jul-2010 06:57:39] yes, I also removed the template ..modeled the device. Rebound the template and modeled the device again
[19-Jul-2010 07:05:06] is there a way to delete performance data for a specific template ?
[19-Jul-2010 07:24:14] st3v3o: yeah, the rrd files are in $ZENHOME/perf/Devices/devicename
[19-Jul-2010 07:24:25] the names will correspond to the datasource names in the template
[19-Jul-2010 07:25:42] i think you should stop zenoss before deleting
[19-Jul-2010 07:26:43] fragfutter … why stop zenoss ?
[19-Jul-2010 07:27:00] just curious
[19-Jul-2010 07:27:53] st3v3o: i didn't look into the code, but i would assume that zenoss does not check for file existence during _every_ write. At least i would only check on startup ;)
[19-Jul-2010 07:28:48] cool I'm gonna test and see if unbinding the template before deletion is sufficient
[19-Jul-2010 07:30:22] looks like unbinding the template works find
[19-Jul-2010 07:30:23] * Simon4 hasn't had issues with just removing the files, they seem to reappear each collection cycle just fine
[19-Jul-2010 07:30:26] fine
[19-Jul-2010 07:31:11] Is there a way to list all devices that have template X bound ?
[19-Jul-2010 07:32:35] should be possible from zendmd
[19-Jul-2010 07:33:56] yeah
[19-Jul-2010 07:33:58] 2 secs
[19-Jul-2010 07:35:08] >>> for dev in dmd.Devices.getSubDevices():
[19-Jul-2010 07:35:08] ... if "TemplateName" in dev.zDeviceTemplates:
[19-Jul-2010 07:35:09] ... print dev.id
[19-Jul-2010 07:35:09] ...
[19-Jul-2010 07:35:41] dammit need moar ram
[19-Jul-2010 07:35:52] everyone needs more ram :)
[19-Jul-2010 07:35:55] my main collector just ran out of swap
[19-Jul-2010 07:36:08] the moment using zope or java
[19-Jul-2010 07:36:57] accually it was zenmodeler that used most of the ram
[19-Jul-2010 07:40:42] how can i delete a relation from zendmd?
[19-Jul-2010 07:42:13] fragfutter: http://nettikconsulting.wordpress.com/2010/06/02/cleaning-up-the-zopedb-removing-interfaces-and-ip-routes-from-devices/
[19-Jul-2010 07:42:25] the _remove(rel) method can be your friend
[19-Jul-2010 07:46:46] Simon4: not sure. i also need to clean out the object itself, so like deleting an interface like device/os/interfaces/eth0
[19-Jul-2010 07:47:24] will i have a loose object eth0 inside the zopedb then?
[19-Jul-2010 07:47:45] fragfutter: I don't believe so, but I could be wrong
[19-Jul-2010 07:47:49] * Simon4 wonders where rmatte is :)
[19-Jul-2010 07:58:55] Completed 30396 out of an estimated 86419, 35.17 percent done bah!
[19-Jul-2010 07:59:10] wont finnish before its time to go home
[19-Jul-2010 08:07:35] tehhobbit: so you can leave, cause it won't finish before anyway...
[19-Jul-2010 08:08:04] hehe
[19-Jul-2010 08:08:06] true
[19-Jul-2010 08:08:31] but would be nice to have the upgrade done so that I can claim that I have done something today
[19-Jul-2010 08:30:22] nice. zenoss is cleaning up during modelling itself.
[19-Jul-2010 08:38:53] anyone have the src rpm of the zenpack bundle ?
[19-Jul-2010 08:41:56] iancmcc: morning, will you be in irc all the time now?
[19-Jul-2010 08:42:14] When I remember to open Colloquy, yeah.
[19-Jul-2010 08:42:22] coolio
[19-Jul-2010 08:45:17] iancmcc: yeah working on a 15" takes getting used to. i routinely switch between 24" LED and 15" MBP depending on mood and what needs to be done
[19-Jul-2010 08:46:20] aclark: I really don't mind it, because the proportions are somehow just right on the 15", but it is a drag to have to cmd-tab to Skype and suchlike.
[19-Jul-2010 08:47:32] yah
[19-Jul-2010 08:54:03] hey all, about to install v3 :)
[19-Jul-2010 08:54:14] hmmmm dammit think I might just gotten a snafu
[19-Jul-2010 08:54:24] sideone: so start converting and come back tomorrow ;)
[19-Jul-2010 08:54:50] hahah
[19-Jul-2010 08:54:51] :)
[19-Jul-2010 08:55:00] just laying down a new OS now
[19-Jul-2010 08:55:13] about to create a production env, been beta under 252
[19-Jul-2010 08:55:15] w/vm
[19-Jul-2010 08:55:25] I have 2 collectors (1 main) but when upgrading to 3.0 (rpm) it doesnt seem to read the config file to see where the zopedb is so claiming I'm not ready for upgrade ?
[19-Jul-2010 08:55:31] i want x64, so im building
[19-Jul-2010 08:55:31] any suggestions
[19-Jul-2010 08:55:48] any guides on how to tweek a linux env for zo 3?
[19-Jul-2010 08:57:01] sideone: buy more ram
[19-Jul-2010 08:57:12] hahha
[19-Jul-2010 08:57:21] 4GB or more for a 50dev network?
[19-Jul-2010 08:57:41] sideone: sounds enough.
[19-Jul-2010 08:57:50] sideone: next, buy faster harddisks
[19-Jul-2010 08:57:54] creating my vm on esxi, how much space should i allocate for 50devices? i was thinking 35gb?
[19-Jul-2010 08:58:14] fragfutter, im striping for the os
[19-Jul-2010 08:59:04] sideone: we've been testing around 200 devices on a 4GB system, you should be totally fine there
[19-Jul-2010 08:59:13] sideone: you are what?
[19-Jul-2010 08:59:21] thanks, how much space did you allocate to the os?
[19-Jul-2010 08:59:35] oh sorry, my raid for the o/s is striped.
[19-Jul-2010 09:00:00] dunno space for the OS. Space for Zenoss depends on how much perf data you want to keep and how many events you expect.
[19-Jul-2010 09:00:56] im really unsure as i havent rolled out zen before. i was hoping 35GB of partition could give me a 1-2 months of history
[19-Jul-2010 09:00:59] public-demo.zenoss.com only has 3.6GB allocated for the entire filesystem including the OS, but it isn't doing any collection.
[19-Jul-2010 09:01:05] 35GB is plenty.
[19-Jul-2010 09:01:14] thx
[19-Jul-2010 09:01:24] sideone: keep the performance data on a seperate disk.
[19-Jul-2010 09:01:42] yeah, i/o will be your bottleneck depending on how frequently you collect
[19-Jul-2010 09:01:54] sideone: what do you mean by striping? raid 1?
[19-Jul-2010 09:02:18] but a 50-device network isn't really taxing at all so you're probably fine without putting it on a separate disk.
[19-Jul-2010 09:02:54] . o O (until you add more devices and need to reconfigure, because you thought, it would be fine)
[19-Jul-2010 09:03:14] i can create a seperate data store for the disc, which is on raid6, but the o/s is running on raid 0 (non redundant, just quick), i have my vm snapshots in case of erro.
[19-Jul-2010 09:03:16] sure, but moving your perf data after the fact is pretty easy.
[19-Jul-2010 09:03:16] s/error.
[19-Jul-2010 09:04:05] sideone: you gain nothing with raid0 for the os. And i would not use raid6 for the perf data.
[19-Jul-2010 09:05:06] sideone: how many disks do you have on which controller/enclosure?
[19-Jul-2010 09:09:38] for dev in dmd.Devices.getSubDevices():
[19-Jul-2010 09:09:38] print dev.id
[19-Jul-2010 09:09:47] That should list all devices no ?
[19-Jul-2010 09:10:16] am I missing some sort of commit or execute ?
[19-Jul-2010 09:10:42] st3v3o: sounds good.
[19-Jul-2010 09:10:54] >>> for dev in dmd.Devices.getSubDevices():
[19-Jul-2010 09:10:54] ... print dev.id
[19-Jul-2010 09:10:54] ...
[19-Jul-2010 09:10:57] I don't get any output
[19-Jul-2010 09:11:04] backspace, backspace, enter
[19-Jul-2010 09:11:25] how could I script that
[19-Jul-2010 09:12:00] you want a python script that lists all devices?
[19-Jul-2010 09:12:12] yes
[19-Jul-2010 09:12:46] fragfutter, i have all of my o/s's running on this esx box with r0, but the data usually sits on a r6
[19-Jul-2010 09:12:59] 10 disks total
[19-Jul-2010 09:13:03] 2TB/each
[19-Jul-2010 09:13:30] r6 instead of r5 due to the 10^14 read error
[19-Jul-2010 09:14:04] sideone: raid6 (and raid5) s*** in write preformance. But with 50devices it should be fine.
[19-Jul-2010 09:14:41] ahh.. if the i/o's fail, is it tough to repoint the data in zenoss?
[19-Jul-2010 09:15:14] no. just mount disk directly on the perf folder or symlink it.
[19-Jul-2010 09:15:31] cool thanks
[19-Jul-2010 09:22:22] st3v3o: http://pastebin.com/u63mcTbp
[19-Jul-2010 09:23:04] cool thanks fragfutter…I found this just a sec ago also docs/DOC-3217;jsessionid=80D27122402B508FC8040FBA903347A4.node0
[19-Jul-2010 09:23:13] but looks like only the zenoss user can run these scripts
[19-Jul-2010 09:23:30] st3v3o: you need the python version bundled with zenoss
[19-Jul-2010 09:23:40] ah
[19-Jul-2010 09:24:00] change first line to /opt/zenoss/bin/python
[19-Jul-2010 09:24:55] I was just gonna ask that
[19-Jul-2010 09:47:23] so I'm trying to use a variable with zendmd but it doesn't seem to be working
[19-Jul-2010 09:47:24] http://pastebin.com/ptFHqEFd
[19-Jul-2010 09:47:36] its does read the variable template
[19-Jul-2010 09:47:40] any suggestions
[19-Jul-2010 09:49:49] what are the quotes for?
[19-Jul-2010 09:52:50] if "Name of Template" in dev.zDeviceTemplates:
[19-Jul-2010 09:53:07] I just appended the quotes to the variable
[19-Jul-2010 09:53:13] why?
[19-Jul-2010 09:53:23] zDeviceTempaltes is a list of strings
[19-Jul-2010 09:53:33] and those strings don't have the additional Quotes
[19-Jul-2010 09:53:49] as your variable template (after you read it) is already a string
[19-Jul-2010 09:53:56] you don't need to add quotes
[19-Jul-2010 09:54:03] …yeah the code from pastebin doesn't add additional quotes
[19-Jul-2010 09:54:34] st3v3o: id does.
[19-Jul-2010 09:54:40] st3v3o: it does.
[19-Jul-2010 09:54:45] just drop line 11
[19-Jul-2010 09:55:37] then the print outputs asdf …not "asdf"
[19-Jul-2010 09:56:01] but regardless of that …the template variable in line 16 is not being read
[19-Jul-2010 09:56:25] it is.
[19-Jul-2010 09:56:35] I stand corrected
[19-Jul-2010 09:57:12] I assumed that if there was a space in the template name that it would need quotes
[19-Jul-2010 09:57:46] st3v3o: why? you are using a programming language not a bash.
[19-Jul-2010 09:59:33] true.
[19-Jul-2010 09:59:34] erm
[19-Jul-2010 09:59:36] template = raw_input ("Enter Template Name: ")
[19-Jul-2010 09:59:39] what the heck is that?
[19-Jul-2010 09:59:45] if it's a function, there shouldn't be a space
[19-Jul-2010 10:00:35] hmmm, it does work with the space
[19-Jul-2010 10:00:40] didn't think it would
[19-Jul-2010 10:00:43] yeah it does
[19-Jul-2010 10:01:14] my python scripting foo is weak though…need to fix that pronto
[19-Jul-2010 10:01:32] yeh, that won't work...
[19-Jul-2010 10:01:41] rmatte: i think it is even syntactical correct. but yes it should be cleaned. as should be line 11. Either a , behind the template or leave out the brackets.
[19-Jul-2010 10:01:44] since if whatever in checks for a perfect match
[19-Jul-2010 10:01:54] it's like doing a re.match
[19-Jul-2010 10:02:40] http://pastebin.com/TWkSfiQj
[19-Jul-2010 10:02:43] this is what I have now and it works
[19-Jul-2010 10:03:02] ah, apparently it is a search
[19-Jul-2010 10:03:13] I generally use re so I wasn't aware
[19-Jul-2010 10:03:27] >>> var = 'The big black sheep'
[19-Jul-2010 10:03:27] >>> if 'sheep' in var:
[19-Jul-2010 10:03:27] ... print "woot"
[19-Jul-2010 10:03:27] ...
[19-Jul-2010 10:03:28] woot
[19-Jul-2010 10:03:30] lol
[19-Jul-2010 10:03:48] cool
[19-Jul-2010 10:03:48] rmatte: either re.search or variable.startswith
[19-Jul-2010 10:04:00] Product documentation in EPUB e-book format. You can read EPUB books from your iPhone, iPad, <-- petition for zenoss to throw in a free ipad with the docs with all new subscriptions
[19-Jul-2010 10:04:04] (especially my subscription)
[19-Jul-2010 10:04:11] fragfutter: wouldn't startswith be limited to the start of the string?
[19-Jul-2010 10:04:34] fragfutter: I'm more partial to re.match and re.search myself
[19-Jul-2010 10:04:47] rmatte: yes. you started with re.match, which is also rooted at the beginning, unless you add a '.*'
[19-Jul-2010 10:06:16] and although you have a nice re library, it doesn't hur to use strings ;)
[19-Jul-2010 10:07:38] lol
[19-Jul-2010 10:09:42] for a stack installer, should i install mysql prior to zenoss installation?
[19-Jul-2010 10:14:39] sideone: no, MySQL is packaged with it
[19-Jul-2010 10:14:48] sideone: just run the stack installer, all dependencies are included
[19-Jul-2010 10:15:04] sideone: it includes it's own copy of Python as well
[19-Jul-2010 10:16:58] sweet. thx
[19-Jul-2010 10:59:57] bigegor_ is now known as bigegor
[19-Jul-2010 11:00:16] eugh, this is so dumb
[19-Jul-2010 11:00:38] I'm getting "Unable to get data for schleich01.novanoc.com from either ipAddrTable or ipNetToMediaTable" for 2 of my Linux servers when I model them, yet I can hit those OIDs via snmpwalk
[19-Jul-2010 11:00:43] same for all the CPU and Memory info
[19-Jul-2010 11:00:50] not sure what's tripping zenoss up yet
[19-Jul-2010 11:04:50] yeh, this is majorly messed up
[19-Jul-2010 11:05:03] Zenoss is treating the OIDs as though they don't exist, yet they can be snmpwalked perfectly fine
[19-Jul-2010 11:05:05] grrrr
[19-Jul-2010 11:05:20] and there's no indication of what's causing it when I run zenperfsnmp against the device with -v10
[19-Jul-2010 11:07:27] poor bastard
[19-Jul-2010 11:08:08] I have to spend the whole morning clearing errors one by one
[19-Jul-2010 11:08:57] why?
[19-Jul-2010 11:21:43] iptables on the box?
[19-Jul-2010 11:22:00] er, linux or win
[19-Jul-2010 11:22:52] nvm!
[19-Jul-2010 11:23:45] linux
[19-Jul-2010 11:23:50] anyone out there monitoring Tomcat?
[19-Jul-2010 11:23:56] and it can't be iptables or snmpwalk would be affected as well
[19-Jul-2010 11:24:24] it's something that Zenoss is choking on
[19-Jul-2010 11:24:42] the snmpd config on the box is identical to others that are working fine
[19-Jul-2010 11:25:22] it's obviously some bug in Zenoss, just need to dig through the code and figure out what's causing it
[19-Jul-2010 11:25:23] strange, have you tried removing the machine zendmd commit(), and re-adding?
[19-Jul-2010 11:25:27] which I shall be doing shortly
[19-Jul-2010 11:25:41] varnii: the machine was just added, that won't help
[19-Jul-2010 11:25:58] and there's another one that's doing the same thing that's been in there for a while, and was working fine at some point
[19-Jul-2010 11:26:23] I'll figure it out, it's just going to be a pain in the ass
[19-Jul-2010 11:26:41] yeah, symptoms are odd. Good luck, coffee +1
[19-Jul-2010 11:26:49] tnx
[19-Jul-2010 11:27:01] rmatte: Because I need to clear everything in between another error
[19-Jul-2010 11:27:04] It's a sad sad story
[19-Jul-2010 11:27:17] theacolyte: why not just use zendmd to do it?
[19-Jul-2010 11:27:23] Thankfully I only have 30 or so
[19-Jul-2010 11:27:34] rmatte: kind of against the point of it isn't it ?:P
[19-Jul-2010 11:27:39] how so?
[19-Jul-2010 11:27:57] you're wanting to selectively move certain errors to history but leave the rest, correct?
[19-Jul-2010 11:28:03] (or at least, that's how you've explained it)
[19-Jul-2010 11:28:33] yes.
[19-Jul-2010 11:28:46] then as I asked before, why not use zendmd? wouldn't be hard
[19-Jul-2010 11:28:48] I think the point is, I'd prefer to do it via the gui rather than learn how to do it via zenmd
[19-Jul-2010 11:29:04] what do the events that you're clearing look like?
[19-Jul-2010 11:29:12] I can have a zendmd script whipped up in 10 seconds
[19-Jul-2010 11:29:23] They are pretty much random
[19-Jul-2010 11:29:27] Different sources
[19-Jul-2010 11:29:36] different devices
[19-Jul-2010 11:29:37] right, but similar summary?
[19-Jul-2010 11:30:00] is there any way to match them, or do you just have to "know"?
[19-Jul-2010 11:30:43] just have to know
[19-Jul-2010 11:30:52] k
[19-Jul-2010 11:33:49] this is roughly how you'd do it:
[19-Jul-2010 11:33:51] http://pastebin.com/8KbKT8sm
[19-Jul-2010 11:33:54] Hey, does anyone know what the python code is for 'contains' operator in zenoss, instead of ' = / is '
[19-Jul-2010 11:34:04] only you'd do it based on summary rather than device and component
[19-Jul-2010 11:34:23] varnii: "in"
[19-Jul-2010 11:34:30] if "string" in variable:
[19-Jul-2010 11:34:41] rmatte: wicked, thanks
[19-Jul-2010 11:34:43] np
[19-Jul-2010 11:45:24] so I upgraded snmpd and that didn't help, hmmm
[19-Jul-2010 11:45:51] zenoss 3.0?
[19-Jul-2010 11:46:03] no
[19-Jul-2010 11:46:05] 2.5.2
[19-Jul-2010 11:46:16] I'm not pushing 3.0 in to production for quite a while
[19-Jul-2010 11:53:21] * rmatte calls upon the mystical powers of Chet
[19-Jul-2010 11:53:28] * rmatte glances around
[19-Jul-2010 11:53:31] darn
[19-Jul-2010 12:00:06] heh
[19-Jul-2010 12:00:10] you're a funny guy ryan
[19-Jul-2010 12:00:37] :)
[19-Jul-2010 12:00:49] yeh, this issue is really weird
[19-Jul-2010 12:01:02] so for some reason... this: iptable = tabledata.get("ipAddrTable")
[19-Jul-2010 12:01:06] is not working
[19-Jul-2010 12:01:15] in the plugin
[19-Jul-2010 12:01:58] this as well: iptable = tabledata.get("ipNetToMediaTable")
[19-Jul-2010 12:02:03] both are yielding blank variables
[19-Jul-2010 12:02:08] it'll be fun to figure out why
[19-Jul-2010 12:04:00] trying to find some tomcat or java monitoring screenshots
[19-Jul-2010 12:06:34] by java you mean jmx?
[19-Jul-2010 12:06:55] yeah
[19-Jul-2010 12:07:07] servlet/JiveServlet/showImage/102-2616-13-1162/ZenJMXGraphs.png
[19-Jul-2010 12:07:31] can't find anything for tomcat
[19-Jul-2010 12:09:02] "There is a Tomcat ZenPack available in Enterprise, but the Zen JMX Collector ZenPack provides the JMX data sources needed for monitoring Tomcat"
[19-Jul-2010 12:10:13] they probably have some screenshots of it kicking around, you'd have to ask Matt
[19-Jul-2010 12:32:33] Is there a zendmd tutorial ?
[19-Jul-2010 12:42:58] i have added a new windows server device, but find the error, "Could not read the Windows event log (NT code 0xc002001b). Check your username/password settings and verify network connectivity. ". Where do i set the user/pass for this device?
[19-Jul-2010 12:44:08] st3v3o: read the developer guide and community posts, lots of helpful stuff
[19-Jul-2010 12:44:15] zProperties
[19-Jul-2010 12:44:47] community/documentation/official_documentation/zenoss-dev-guide
[19-Jul-2010 12:44:52] click on the "arrow" on the tab menu ..more..zProperties
[19-Jul-2010 12:47:55] varnii thanks
[19-Jul-2010 12:48:24] word, np. Another helpful thing I've found is digging through the source code, and viewing page sources for attribute names etc
[19-Jul-2010 13:01:41] the tab completion is useful too
[19-Jul-2010 13:03:40] sideone: you're looking for zWinUser and zWinPassword under zProperties
[19-Jul-2010 13:03:56] sideone: only set them on the actual windows server class that you have the devices in, don't set it at the root level
[19-Jul-2010 13:15:31] rmatte; i had set them with the correct user/pass, but recieve a different message, "Could not read the Windows event log (NT_STATUS_IO_TIMEOUT). Check your username/password settings and verify network connectivity. " Do i need to integrate LDAP/AD or setup a wmi type of process on the server?
[19-Jul-2010 13:16:08] is it a local user account?
[19-Jul-2010 13:16:32] and yes, WMI does need to be properly configured on the server
[19-Jul-2010 13:16:55] if it's a local user account put ./ in front of the username in zWinUser
[19-Jul-2010 13:16:57] rmatte; nope its a domain admin account. should i specify DOMAIN\user ? also, is there a wmi guide for zenoss? im very newb with wmi
[19-Jul-2010 13:17:08] yeh, you need to specify domain\user
[19-Jul-2010 13:17:13] otherwise it has no idea what it's doing
[19-Jul-2010 13:17:15] gotcha. brb thanks :)
[19-Jul-2010 13:17:17] k
[19-Jul-2010 13:17:43] btw, 3.0 is very fluid, i really like the AI
[19-Jul-2010 13:18:56] I assume you mean the UI?
[19-Jul-2010 13:19:13] yeh, it's not bad, there are some annoying bugs that still need to be fixed
[19-Jul-2010 13:20:17] nice, so when Zenoss polls for the uptime on this particular device it gets...
[19-Jul-2010 13:20:35] 14:07:49.860551 IP (tos 0x0, ttl 64, id 0, offset 0, flags [DF], proto UDP (17), length 286) sensored.novanoc.com.snmp > 30.30.241.10.47608: { SNMPv2c { GetResponse(241) R=1803379265 25.1.1.0=[noSuchObject] [|snmp] } }
[19-Jul-2010 13:20:47] but if I do it via snmpwalk...
[19-Jul-2010 13:21:06] 14:14:24.732634 IP (tos 0x0, ttl 64, id 0, offset 0, flags [DF], proto UDP (17), length 81) sensored.novanoc.com.snmp > 30.30.241.10.41851: { SNMPv2c C=nova-public { GetResponse(33) R=839631284 25.1.1.0=182198147 } }
[19-Jul-2010 13:21:09] exact same OID
[19-Jul-2010 13:21:28] for some reason when Zenoss polls, it returns "noSuchObject" but snmpwalk polling the exact same things return normally
[19-Jul-2010 13:21:35] quite annoying and bizarre
[19-Jul-2010 13:22:29] wish a dev was around so that I could ask about it
[19-Jul-2010 13:25:15] hmmm, it seems like Zenoss isn't passing a community string for the device
[19-Jul-2010 13:25:44] my god, if it's as simple as the zProperties being screwy on this box I'm going to scream
[19-Jul-2010 13:27:29] weird
[19-Jul-2010 13:27:44] it automatically added "public" as the snmp string locally when I added the devices
[19-Jul-2010 13:27:48] why the heck is it doing that
[19-Jul-2010 13:28:06] well, that fixes it
[19-Jul-2010 13:28:08] bah
[19-Jul-2010 13:28:15] now I need to figure out why it's doing that
[19-Jul-2010 13:30:08] glad I finally got those devices modeled
[19-Jul-2010 13:30:54] haha
[19-Jul-2010 13:30:56] nice!
[19-Jul-2010 13:31:08] btw, i think the errors have stopped.
[19-Jul-2010 13:34:00] i need to do a bunch of configuration i believe. just reading thourhg the manual now. I am using this system for a couple lin*/unix* boxen and a cisco router/switches via snmp. The bulk of what i would like to monitor are windows servers+workstations. I have 10 server (phys/vm), and 50+ workstations (DHCP). What would be the ideal way to monitor these boxes? wmi, snmp, etc? Should i discover devices by hostname for the dhcp based workstations? is ther
[19-Jul-2010 13:35:16] sideone: I don't know too many people monitoring workstations with Zenoss
[19-Jul-2010 13:36:15] monitoring DHCP devices with Zenoss isn't really possible since it's IP address oriented
[19-Jul-2010 13:36:23] so you'd constantly be changing IPs for devices
[19-Jul-2010 13:36:31] ahh..
[19-Jul-2010 13:36:37] well, it's possible, just not very simple to do
[19-Jul-2010 13:36:57] no sure why you'd want to monitor workstations anyways?
[19-Jul-2010 13:37:04] it's just a huge pain in the arse
[19-Jul-2010 13:37:06] generally users monitor workstations fairly well
[19-Jul-2010 13:37:07] and has no real value
[19-Jul-2010 13:37:17] they certainly make enough noise if they don't work ;)
[19-Jul-2010 13:37:21] rmatte, true enough
[19-Jul-2010 13:37:21] exactly
[19-Jul-2010 13:37:31] rmatte: I think if you don't have zSnmpCommunity set at all, anywhere, it tries the first entry that works on discovery
[19-Jul-2010 13:37:41] i was actually just looking for a few polls, low disk space, mem over consumed, 100% proc, etc..
[19-Jul-2010 13:37:43] in the list of zsnmpcommunities
[19-Jul-2010 13:38:03] sideone: yeh, but once again, they are workstations, let the users handle it themselves
[19-Jul-2010 13:38:20] sideone: they should be intelligent enough to realize that when the computer is warning them about disk space, it's time to clean stuff up
[19-Jul-2010 13:38:25] bah, panzy, give'em all static IP's :)
[19-Jul-2010 13:38:38] yeh, or give them all static IPs and have at er'
[19-Jul-2010 13:39:28] rmatte hahah
[19-Jul-2010 13:39:33] rmatte: curious, you made a point to do all SNMP based monitoring for Windows, right?
[19-Jul-2010 13:39:34] Why?
[19-Jul-2010 13:39:46] Why not WMI datasource + WMI device perf zenpack
[19-Jul-2010 13:39:52] twm1010: I just said I personally prefer it since it's simpler to setup and more stable
[19-Jul-2010 13:39:57] a lot less overhead too
[19-Jul-2010 13:40:10] but using Egor's pack to do full WMI monitoring is an alternate solution
[19-Jul-2010 13:40:23] twm1010 is wmi the 'proper' way to monitor win* servers with zo?
[19-Jul-2010 13:40:35] sideone: there is no "proper way"
[19-Jul-2010 13:40:44] rmatte; haha
[19-Jul-2010 13:40:53] rmatte; what is the most popular way? :)
[19-Jul-2010 13:40:56] well, WMI datasource uses WMI, but I think zenoss enterprise uses RPC
[19-Jul-2010 13:40:58] Zenoss Core has full SNMP support built in...
[19-Jul-2010 13:41:08] for full WMI support, you need the WMI Data Source ZenPack
[19-Jul-2010 13:41:15] and the WMI Performance Template ZenPack as well
[19-Jul-2010 13:41:18] both were made by Egor
[19-Jul-2010 13:41:30] then you can add devices to /Devices/CIM/WMI and model them using WMI
[19-Jul-2010 13:41:54] Well I'm curious if I switch over to it, will it use existing RRDs assuming he kept datasources and graph points the same
[19-Jul-2010 13:42:05] or, you can monitor via SNMP, but you need my Windows SNMP Performance Monitor packs to monitor CPU and Memory without requiring snmp-informant
[19-Jul-2010 13:42:11] either way, you need a ZenPack
[19-Jul-2010 13:42:24] twm1010: no, the RRDs will be different
[19-Jul-2010 13:43:06] hrmmm... so I'd have to do a cleanup then
[19-Jul-2010 13:43:20] I wish there was a way to cleanup RRDs that haven't been written to in a while, I suppose a simply shell script would do it
[19-Jul-2010 13:43:26] ok time to look into zenpacks :)
[19-Jul-2010 13:43:40] twm1010: yeh
[19-Jul-2010 13:43:44] but I'm curious to look into what kind of performance hit I might take, does the WMI datasource create that much overhead?
[19-Jul-2010 13:44:17] twm1010: Zenoss used to do it automatically but there was a bug where it ended up deleting RRD files that were being actively written to but not properly timestamped by the system
[19-Jul-2010 13:44:40] which was a linux related bug
[19-Jul-2010 13:44:43] is there a util to run to force it?
[19-Jul-2010 13:44:55] twm1010: none that I know of off the top of my head
[19-Jul-2010 13:45:17] twm1010: why wouldn't you just clear out the device folders for those devices in $ZENHONE/Perf?
[19-Jul-2010 13:45:22] and let them regenerate
[19-Jul-2010 13:45:34] Well. I only want it to nuke ones that aren't being used, right?
[19-Jul-2010 13:45:38] not all of them
[19-Jul-2010 13:45:53] well, I thought you were talking about if you were switching to WMI instead of SNMP
[19-Jul-2010 13:46:02] well sure, but not all my devices are windows
[19-Jul-2010 13:46:06] I doubt you have many stale ones
[19-Jul-2010 13:46:14] yeh, I'm aware of that
[19-Jul-2010 13:46:21] so make a list of all of the devices that are
[19-Jul-2010 13:46:32] then have a batch script that only deletes the folders matching those device names
[19-Jul-2010 13:46:32] you'd be surprised, I've got a lot of servers that for some reason snmp informant doesn't work
[19-Jul-2010 13:46:36] or causes the agent to crash
[19-Jul-2010 13:46:55] that's why I may re-org
[19-Jul-2010 13:46:56] I don't use snmp-informant
[19-Jul-2010 13:47:09] some of these systems have a mix tho, standard zenoss/informant template, and then some extra data collected via WMI datasource
[19-Jul-2010 13:47:11] snmp-informant == pain in the arse
[19-Jul-2010 13:47:24] so like for my exchange box, I've got lots of good RRD data for e-mail stats
[19-Jul-2010 13:47:32] I see
[19-Jul-2010 13:47:38] well then handle that one manually
[19-Jul-2010 13:47:43] you obviously have to be intelligent about it
[19-Jul-2010 13:47:53] automate it as best you can
[19-Jul-2010 13:49:32] God damn... 'm having an issue with 3.0: Prior to this version, I was adding SERVER1 (4 IPs), the primary IP would monitor SNMP/disk/load/etc, while the remaining IPs where all sites and to be adding separately with no SNMP to monitor individual services for those sites.
[19-Jul-2010 13:50:07] 3.0, when I add anything after the device, it errors out, and says the IP is already on 'SERVER1' for example
[19-Jul-2010 13:50:21] am I going about this wrong?
[19-Jul-2010 13:51:04] excuse my poor English skills, trying to resolve this quickly hah
[19-Jul-2010 13:55:10] rmatte: I think I found my informant issue. Some over-permissioned helpdesk gremlin decided to uninstall informant :O
[19-Jul-2010 13:56:29] That, or maybe installing 2003SP2 set snmp plugins back to defaults
[19-Jul-2010 14:07:07] twm1010: could be
[19-Jul-2010 14:14:01] Q: Why not WMI datasource + WMI device perf zenpack?
[19-Jul-2010 14:14:01] A: Because, WMI we can use for monitoring of M$ Exchnge, M$ SQL, M$ IIS or some other M$ stuff, and not only CPU load and memory utilization of Windows OS. I'm preffer separate zenpack for every product, as alternative to single SUPER-MEGA-ALL-IN-ONE Zenpack. Small zenpack for one product is easy to manage and update, and more importend, it can be provided and supported by another person too.
[19-Jul-2010 14:22:33] bigegor_ is now known as bigegor
[19-Jul-2010 14:24:27] hey bigegor
[19-Jul-2010 14:24:36] hi
[19-Jul-2010 14:24:43] i just started exploring zenpacks, and found wmi data source :)
[19-Jul-2010 14:24:53] just swithcing over to zen from nag*
[19-Jul-2010 14:26:15] i would like to use zen* to monitor mostly windows servers, errors, possibly event logs?, and performance. On top of that, i would like to hit applications and layer 3 information. I believe the best way to get started is to use your wmi datasource pack
[19-Jul-2010 14:26:47] I use it alot
[19-Jul-2010 14:26:54] sorry for such a newb question, but should i couple it with other packs? (i did see your q/a above :)
[19-Jul-2010 14:26:55] Exchange, Terminal Services stats
[19-Jul-2010 14:26:55] sideone: yeh, use that ZenPack with the WMI Performance Template pack as well
[19-Jul-2010 14:26:56] not realy
[19-Jul-2010 14:27:07] sideone: event log monitoring is already built in to Zenoss core
[19-Jul-2010 14:27:17] rmatte; really?
[19-Jul-2010 14:27:29] event and windows service monitoring
[19-Jul-2010 14:27:31] time to read, i thought events tab != win* events
[19-Jul-2010 14:27:53] event tab == ALL events
[19-Jul-2010 14:27:58] regardless of what they are
[19-Jul-2010 14:28:04] ahh, very very very nice!
[19-Jul-2010 14:28:09] does it tail?
[19-Jul-2010 14:28:15] or do you have to refresh?
[19-Jul-2010 14:28:26] it auto-refreshes at the set refresh time
[19-Jul-2010 14:28:32] wicked.
[19-Jul-2010 14:28:58] ok so i am going to start with wmi, by using egors' wmi perf temp pack and the wmi datasource
[19-Jul-2010 14:29:06] k
[19-Jul-2010 14:29:07] thanks guys.
[19-Jul-2010 14:29:12] put the devices in /Devices/CIM/WMI
[19-Jul-2010 14:29:28] ok thanks, brb 15, learning eggs :)
[19-Jul-2010 14:29:37] never been a python gu
[19-Jul-2010 14:29:39] s/guy
[19-Jul-2010 14:29:46] and make sure you set the WMI username/pass in zProperties, (zWinUser, zWinPassword)
[19-Jul-2010 14:29:59] when you download the eggs they'll be .egg.zip
[19-Jul-2010 14:30:03] unzip them before installing
[19-Jul-2010 14:30:10] it needs to be just a .egg before installing properly
[19-Jul-2010 14:30:16] will do, i just used wmi explorer to make sure my wmi is able to be queried.
[19-Jul-2010 14:30:20] ok cool thanks rm
[19-Jul-2010 14:30:21] the Zenoss community site automatically zips stuff when uploaded unfortunately
[19-Jul-2010 14:30:24] np
[19-Jul-2010 14:30:39] rmatte; prob better for security/health of data for a zip
[19-Jul-2010 14:30:44] instead of egg.
[19-Jul-2010 14:30:48] well, a .egg is zipped
[19-Jul-2010 14:31:04] so it's basically just taking something that is already zipped and zipping it again
[19-Jul-2010 14:31:27] ahh, an egg is a compression for python? guess it already has crc checking.
[19-Jul-2010 14:31:43] well, it uses .zip format for compression as far as I can tell
[19-Jul-2010 14:31:48] it's just called a .egg
[19-Jul-2010 14:31:58] ahh. ok going to get wmi stuff.
[19-Jul-2010 14:32:02] it has a specific file/directory structure in it
[19-Jul-2010 14:32:03] k
[19-Jul-2010 14:40:35] roaksoax__ is now known as andreserl
[19-Jul-2010 14:55:48] brb.
[19-Jul-2010 15:32:12] hey all.
[19-Jul-2010 15:33:24] yo
[19-Jul-2010 15:33:38] rmatte, sorry got hung up with gpo issues.
[19-Jul-2010 15:33:46] back to zo for 30 mins :)
[19-Jul-2010 15:37:45] cool
[19-Jul-2010 15:48:15] grabbed the wmi datasource and perf, trying to integrate before 10 mins :)
[19-Jul-2010 15:54:14] restart zenoss after you install those packs
[19-Jul-2010 15:54:19] yep.
[19-Jul-2010 15:54:35] time for server work. ill jump on this again tomorrow.
[19-Jul-2010 15:54:39] thanks for the help today!
[19-Jul-2010 15:54:40] k
[19-Jul-2010 15:54:41] np
[19-Jul-2010 16:56:31] I've got a script that I use to add Location organizers. I'm using the technique described here to connect: docs/DOC-2483 . I'm basically just doing dmd.Locations.manage_addOrganizer('/Locations/Foo') followed by a commit(). The new organizers can be referenced through zendmd, but they don't show up in the web interface (and I get an error that they already exist if I try to create them through there)
[19-Jul-2010 16:57:09] the odd part is that if I do the same command directly to zendmd, it works fine. This technique worked for me in 2.x releases (I'm using 3.0 rc, should have specified that). Any ideas?
[19-Jul-2010 17:13:07] wayne___: you must be missing something but I'm not sure what... you should check what Zenoss does in the sourcecode when an organizer is created via the UI for clues on what you're missing
[19-Jul-2010 17:13:37] actually, let me check a script I have
[19-Jul-2010 17:14:33] hmmm nope
[19-Jul-2010 17:14:38] looks like that's the way to do it
[19-Jul-2010 17:14:39] dmd.Locations.manage_addOrganizer(name)
[19-Jul-2010 17:15:07] http://fpaste.org/iiOd/
[19-Jul-2010 17:15:15] there's a batch location script
[19-Jul-2010 17:15:26] ugh
[19-Jul-2010 17:15:45] cgibbons might have insight
[19-Jul-2010 17:16:30] actually hmmm, shouldn't he just be doing: dmd.Locations.manage_addOrganizer('Foo')?
[19-Jul-2010 17:16:44] or dmd.Locations.manage_addOrganizer('/Foo')?
[19-Jul-2010 17:17:20] since he's already referencing locations in the dmd path
[19-Jul-2010 17:17:27] I think I started adding /Locations after seeing it in an example somewhere else
[19-Jul-2010 17:18:02] previously I was just using dmd.Locations.createOrganizer('/Foo'), with the same results
[19-Jul-2010 17:19:15] if I change the script to print the command instead of running it, I can then paste it into zendmd and it works exactly the way I'd expect it to
[19-Jul-2010 17:20:38] dmd.Locations.createOrganizer('/Ugh') also works through zendmd, but if I do it from my script the organizer won't show up in the web interface
[19-Jul-2010 17:22:58] on a separate issue, my device list keeps going back to only showing devices with a production state of 'production' every time I go to somewhere other than the infrastructure "tab" - is this normal or should I be looking for an issue in my setup?
[19-Jul-2010 17:23:35] can you pastebin your who script?
[19-Jul-2010 17:23:39] I want to see what you have in it
[19-Jul-2010 17:23:48] your whole script*
[19-Jul-2010 17:23:49] rather
[19-Jul-2010 17:24:23] wayne___: well, do you have devices that aren't in a Production state?
[19-Jul-2010 17:24:29] wayne___: also, is this in 2.5.2 or 3.0?
[19-Jul-2010 17:24:55] If it's 3.0 these could all be bugs with the new interface for all I know
[19-Jul-2010 17:25:07] rmatte: 3.0, I have some devices with testing as their production state
[19-Jul-2010 17:27:17] afternoon
[19-Jul-2010 17:27:56] http://www.fpaste.org/Ms3p/
[19-Jul-2010 17:28:08] is there some standard way of hooking in code to run when the CMDB gets updated with data?
[19-Jul-2010 17:28:26] wayne__: shoot, and that doesn't work? one sec...
[19-Jul-2010 17:28:32] doh, forgot to take out a password
[19-Jul-2010 17:28:36] guess I'll be changing that!
[19-Jul-2010 17:28:43] wayne___: oh, I see why
[19-Jul-2010 17:28:47] * iancmcc will tactfully ignore that
[19-Jul-2010 17:29:03] wayne___: have you not noticed the dropdown in the device list for the Production State column?
[19-Jul-2010 17:29:11] click that and notice that Production is the only state selected
[19-Jul-2010 17:29:17] select the other states that you want to see
[19-Jul-2010 17:30:15] wayne___: and I'm pretty sure the default is production, you have to select what you want to see if you want to view anything else
[19-Jul-2010 17:30:19] rmatte: yup, I've got that part. As soon as I navigate away from the infrastructure tab (say, to the dashboard) and back, it goes back to only having production selected
[19-Jul-2010 17:30:28] yeh, that's by design I suspect
[19-Jul-2010 17:30:31] wayne__: Apparently there's an issue with the zcml loading.
[19-Jul-2010 17:30:40] nite all
[19-Jul-2010 17:30:48] rmatte: ok, just wanted to make sure it wasn't something weird on my end, thanks
[19-Jul-2010 17:30:52] iancmcc: the device list is not remember it's filters, is that normal?
[19-Jul-2010 17:30:57] wayne__: Tho I don't see why as it's right there on line 36...
[19-Jul-2010 17:31:01] remembering*
[19-Jul-2010 17:31:20] rmatte: Yeah, we didn't do the state remembering on the device list. That's either a design decision or an oversight, I can't remember which.
[19-Jul-2010 17:31:33] I'll take a note and check it out
[19-Jul-2010 17:31:43] thanks
[19-Jul-2010 17:35:17] honestly, for the device list and the event console it would be nice to remember the state/severity settings but to reset everything else on reload of the page, including setting the data back to the default of 24 hours back
[19-Jul-2010 17:35:49] I hate having to remove the last summary search that I did, and then change the date since it's picking up tickets from a month ago when I last happened to check the history on the given server
[19-Jul-2010 17:35:52] lots of things would be nice, including world peace. send me beer.
[19-Jul-2010 17:36:28] :P
[19-Jul-2010 17:36:28] beer I could probably arrange for... though as far as the shipping company would be concerned it would be "olive oil"
[19-Jul-2010 17:36:36] hehe
[19-Jul-2010 17:37:03] well, time to go get my hair cut then head home, bbtmrw
[19-Jul-2010 17:37:26] thanks for the help
[19-Jul-2010 17:37:46] I'll bug iancmcc tomorrow to find out whether the loading issue is an RC bug
[19-Jul-2010 18:08:28] is there a way to schedule maintenance on applications rather than the entire device?
[19-Jul-2010 20:46:27] hello guys!
[19-Jul-2010 20:47:39] i use zenoss open source 2.5.x, and i want to upgrade to 3.0. Im monitoring just 10 hosts, but the release notes say that the server must have 4GB of ram. Mine has 2. Will it work, even slow? Or it wont allow me to install it?
[19-Jul-2010 20:48:40] ?
[19-Jul-2010 20:55:47] logyati: you'll probably be fine. My test box has 2GB and monitors more than 10 hosts
[19-Jul-2010 20:58:03] wayne___, ok thanks :)
[19-Jul-2010 21:32:58] well just upgraded
[19-Jul-2010 21:33:23] when i open infrastructure > processes i get an error "NameError global name 'ancestorValue' is not defined"
[19-Jul-2010 21:33:31] I can confirm you can monitor more than that with 2gb of memory
[19-Jul-2010 21:33:35] I'm doing like 40
[19-Jul-2010 22:07:14] i've got at least 40 devices on a VM with 2gb of ram
[19-Jul-2010 22:07:21] works fine, although the 3.0 webgui is def slower
[19-Jul-2010 22:07:27] not sure if theacolyte has run into that
[19-Jul-2010 22:07:43] 3.0 seems faster, but hat may be my recent reboot talking
[19-Jul-2010 22:07:50] box deffo uses 100% RAM
[19-Jul-2010 22:24:12] how do i change device's class in 3.0?
[19-Jul-2010 22:24:46] couldnt find
[19-Jul-2010 23:46:29] Any Zenoss Tools you would recommend for a local ISP?
[20-Jul-2010 00:00:30] [disconnected at Tue Jul 20 00:00:30 2010]
[20-Jul-2010 00:00:30] [connected at Tue Jul 20 00:00:30 2010]
[20-Jul-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]
[20-Jul-2010 01:08:22] is there some way to create alarm rule where "device class begins with /Something/" and "device class is not /something/foo/"
[20-Jul-2010 02:18:31] hmmmmm after upgrade I get a whole bunch of funky problems, main one is the device view isnt showing any devices
[20-Jul-2010 03:57:16] tehhobbit: what version did you upgrade from? I'm running 2.5.2 and am considering upgrading, but if there are too many funky issues I guess I'll wait for 3.0.1 or something. :-)
[20-Jul-2010 03:57:56] JesperS: 2.52. => 3.0 , but 90% of the issues seem to be with old zen packs
[20-Jul-2010 03:58:08] k
[20-Jul-2010 03:58:19] because they changed to python 2.6
[20-Jul-2010 03:58:40] oh ok
[20-Jul-2010 03:59:21] http://pastie.org/1051723 and this from zenhub
[20-Jul-2010 04:41:55] how can I get the total number of devices ?
[20-Jul-2010 07:00:28] tehhobbit: len(dmd.Devices.getSubDevices()) but i have no idea if there is a more efficient way
[20-Jul-2010 07:15:57] navigate to "Device List" and have a look at the number at the bottom of the table ;)
[20-Jul-2010 07:16:10] * Simon4 provides the Windows(tm) answer
[20-Jul-2010 07:28:41] anyone that can shed a light on the following http://pastie.org/1051898 from zenhub log after upgrade 2.5.2 => 3.0
[20-Jul-2010 07:29:06] err zenmodeller
[20-Jul-2010 07:29:15] get it for almost all devices
[20-Jul-2010 07:47:55] tehhobbit: did you run zencatalog
[20-Jul-2010 07:48:02] yes
[20-Jul-2010 07:48:13] but hmmmmm
[20-Jul-2010 07:48:19] I didnt on the collector
[20-Jul-2010 07:48:34] does it do something with the rrd files too ?
[20-Jul-2010 07:49:56] tehhobbit: you error indicates a missing attribute, i assume that some indexes in the zopedb are not correct
[20-Jul-2010 07:51:56] fragfutter: I have 1 main monitor, that I ran zencatalog on and one collector that I didnt but since they use the same zope db I shouldnt need to run it on both should I ?
[20-Jul-2010 07:57:38] tehhobbit: no idea.
[20-Jul-2010 08:13:08] man, some of the foreign support guys at one of the companies that we support have awesome english
[20-Jul-2010 08:13:20] apparently we're performing an upgradation of IOS on a Cisco device
[20-Jul-2010 08:13:35] and the upgradation has been preponed to a different time
[20-Jul-2010 08:14:34] epic upgradation
[20-Jul-2010 08:19:44] indeed
[20-Jul-2010 08:19:55] so, new community site issue...
[20-Jul-2010 08:19:57] docs/DOC-3404#cf
[20-Jul-2010 08:20:01] out of office messages
[20-Jul-2010 08:20:09] quite annoying
[20-Jul-2010 08:31:40] rmatte: time for some procmail filters maybe
[20-Jul-2010 08:43:05] hey all
[20-Jul-2010 08:53:43] wayne___ is now known as wtucker
[20-Jul-2010 09:58:36] Migrated to Zenoss 3.0 successfully :)
[20-Jul-2010 09:59:58] chachan: silver trophy zenoss3.0 awarded
[20-Jul-2010 10:00:12] chachan: come migrate mine, kthx
[20-Jul-2010 10:00:27] O.o
[20-Jul-2010 10:02:31] does any one know if HttpMonitor has a new version or if I need to do anything?
[20-Jul-2010 10:02:40] anyone*
[20-Jul-2010 10:50:28] tornado is now known as Guest25659
[20-Jul-2010 10:56:43] crickets
[20-Jul-2010 10:57:28] *zirp**zirp*
[20-Jul-2010 11:13:48] morning
[20-Jul-2010 11:14:17] working on adding support for a wireless access point into zenoss, and I was just wondering what organizer they should be placed in
[20-Jul-2010 11:14:25] the instructions say to create an organizer
[20-Jul-2010 11:14:55] but I would rather it goes into some place that is reasonably standard
[20-Jul-2010 11:38:53] hmm is anyone having problems with the zenoss.snmp.HRSWRunMap plugin with 3.0? It keeps on failing
[20-Jul-2010 11:40:32] http://pastebin.com/S8c5QNQw
[20-Jul-2010 11:40:34] I get that error
[20-Jul-2010 12:02:33] mducharme-laptop: if it were me I'd probably do Network/Bridge/Wireless AP/
[20-Jul-2010 12:03:45] wtucker - these wireless APs are not standard wifi.. they have 128 "slots" for SU's to get linked to them, almost like a switch with 128 interfaces
[20-Jul-2010 12:04:07] that made me unsure of using "bridge"
[20-Jul-2010 12:04:19] mducharme-laptop: SU?
[20-Jul-2010 12:06:54] mducharme-laptop: subscriber unit?
[20-Jul-2010 12:10:56] mducharme-laptop: technically speaking I'd say that bridge is more accurate since the wireless side is a single collision domain. From a logical perspective it sounds like it may present it self more like a switch, though, so /Network/Switch/ probably makes more sense from that perspective
[20-Jul-2010 12:32:51] so how many of you are using the 3.0 release so far?
[20-Jul-2010 12:33:45] cgibbons: I am, but not in production yet
[20-Jul-2010 12:34:58] wtucker yeah that's exactly it
[20-Jul-2010 13:29:10] hi folks :)
[20-Jul-2010 13:52:13] Alright guys. I need some serious help with zenoss... I have a voice circuit that's having issues and i'd like zenoss tell me when this ckt goes down.
[20-Jul-2010 13:52:33] Now I assume it's possible, I just don't have it setup correctly
[20-Jul-2010 13:52:52] arrrghhh: does it talk snmp?
[20-Jul-2010 13:52:58] arrrghhh: what does the voice circuit connect to?
[20-Jul-2010 13:53:02] The T1's operational status is currently DOWN but I have no events for the site
[20-Jul-2010 13:53:11] yes the router talks snmp
[20-Jul-2010 13:53:28] The voice ckt goes into a cisco 2851 router.
[20-Jul-2010 13:53:36] which version of zenoss?
[20-Jul-2010 13:53:40] 2.5.2
[20-Jul-2010 14:00:18] hello where can I file a bug report for 3.0?
[20-Jul-2010 14:05:07] grrrrr
[20-Jul-2010 14:05:38] someone put their out of office notification on, every time they receive an email notification from the Zenoss community site the out of office notification kicks in and it ends up posting it as a comment
[20-Jul-2010 14:05:44] so the Cisco Mibs page is full of them
[20-Jul-2010 14:05:48] * rmatte shakes fist
[20-Jul-2010 14:06:00] HAHAHAA that is awesome
[20-Jul-2010 14:06:16] Matt's not around today?
[20-Jul-2010 14:14:18] rmatte: that's why i never turn on my out of office, no matter how much my boss bitches
[20-Jul-2010 14:16:01] ashp: I never turn mine on either
[20-Jul-2010 14:17:44] Sorry guys havin issues w/irc @ work
[20-Jul-2010 14:17:58] Tryin to use my phone, so far not workin out too we'll...
[20-Jul-2010 14:18:05] Well rather
[20-Jul-2010 14:20:52] hrm
[20-Jul-2010 14:20:58] Looks like 3.0 is final... have you upgraded rmatte?
[20-Jul-2010 14:21:30] I just added a modeler.. the instructions say: To test your new modeler plugin, add it to the list of modeler plugins.
[20-Jul-2010 14:21:52] arrrghhh: heck no, I'm going to be waiting at least until the next bugfix release
[20-Jul-2010 14:21:55] I have made the .py file for the modeler but it is not appearing in the list of modeler plugins in the gui.. how do I get it to appear there?
[20-Jul-2010 14:22:14] arrrghhh: 2.5.0 was a disaster
[20-Jul-2010 14:22:21] Rmatte, haha makes sense. So can you help with my problem?
[20-Jul-2010 14:22:29] which problems?
[20-Jul-2010 14:22:48] oh, I see
[20-Jul-2010 14:22:53] Lol not plural. Unless you mean my personal problems.
[20-Jul-2010 14:23:22] just having the .py file in the directory is obviously not enough
[20-Jul-2010 14:23:37] you have 2 options
[20-Jul-2010 14:23:49] mducharme: in the zproperties for the organizer you add the modeller plugin into "collector plugins"
[20-Jul-2010 14:23:51] either have the device trap back to Zenoss when the voice circuit goes down (you need to configure the device to do that)
[20-Jul-2010 14:23:54] or, this: docs/DOC-2494
[20-Jul-2010 14:23:58] which is the way I do it
[20-Jul-2010 14:24:10] K
[20-Jul-2010 14:24:24] I'll look at the doc when I get back to my pc, thx.
[20-Jul-2010 14:25:32] np
[20-Jul-2010 14:25:55] simon4_ that's where I was, but it's not there
[20-Jul-2010 14:26:27] I clicked "add fields" and up comes available fields, but the modeler that I made is not there
[20-Jul-2010 14:26:42] not everyone thought it was, rmatte, you just break things :)
[20-Jul-2010 14:26:54] mducharme-laptop: I'm presuming you've installed your zenpack"
[20-Jul-2010 14:26:55] ?
[20-Jul-2010 14:27:34] cgibbons: I seem to recall a big rush to get 2.5.1 out
[20-Jul-2010 14:27:45] cgibbons: there were some pretty nasty bugs
[20-Jul-2010 14:28:02] the x.x.1 release is always quick because we couple it with the enterprise release
[20-Jul-2010 14:28:17] no simon4_ I do not have a zenpack yet
[20-Jul-2010 14:28:19] 3.0.1 will be out real soon now for the same reason
[20-Jul-2010 14:28:22] I am following the instructions here: docs/DOC-3820
[20-Jul-2010 14:28:25] yeh, but there was a bug in 2.5.0 where a regular user could escalate themselves to an admin
[20-Jul-2010 14:28:37] which was pretty rough
[20-Jul-2010 14:28:47] for some indeed
[20-Jul-2010 14:29:05] I guess it depends on how heavily people actually use Zenoss
[20-Jul-2010 14:29:11] people who don't use it heavily won't notice the issues
[20-Jul-2010 14:29:19] whereas I fully test it before deciding on upgrades
[20-Jul-2010 14:29:21] Oh I hope my company ponies up for the enterprise ed....
[20-Jul-2010 14:29:27] to make sure there aren't any surprises
[20-Jul-2010 14:29:32] and there usually still end up being surprises
[20-Jul-2010 14:29:50] simon4_ so how do I get my modeller into the list of available plugins
[20-Jul-2010 14:30:02] Yea, you can't possibly test every scenario
[20-Jul-2010 14:30:52] mducharme-laptop: ah I see, you've added the file into $ZENHOME/Products/DataCollector/plugins/zenoss/snmp ?
[20-Jul-2010 14:31:01] simon4_ yes
[20-Jul-2010 14:31:03] you might need to restart zenoss to get the new file picked up, not sure
[20-Jul-2010 14:32:40] simon4_ the instructions said that is the place to add it, I still do not know how this will go in a zenpack afterwards
[20-Jul-2010 14:32:56] mducharme-laptop: those instructions are fairly old
[20-Jul-2010 14:32:59] is there a better location for me to be placing this?
[20-Jul-2010 14:33:16] * Simon4_ tries to find the "how to make a zenpack" instructions
[20-Jul-2010 14:34:03] Simon4_: there's Jane Curry's paper on that and there's also the developer's guide
[20-Jul-2010 14:34:21] Simon4_: there's an example developer's pack kicking around too
[20-Jul-2010 14:34:23] rmatte: yeah, there was somewhere that covered initial creation, then installing via --link
[20-Jul-2010 14:34:32] which seems to be the easiest way to get it all started
[20-Jul-2010 14:34:37] why would you bother to create it by hand?
[20-Jul-2010 14:34:41] just create it in the UI
[20-Jul-2010 14:34:50] yeah, this is the instructions I'm trying to find :)
[20-Jul-2010 14:34:52] then copy it out to wherever, remove it from Zenoss
[20-Jul-2010 14:34:54] and then link it
[20-Jul-2010 14:34:57] takes 3 seconds
[20-Jul-2010 14:35:06] mducharme-laptop: what rmatte said :)
[20-Jul-2010 14:35:48] the plugin would go in the ZenPack's plugins directory after it's created
[20-Jul-2010 14:35:51] mducharme-laptop: have a read of this: docs/DOC-3783
[20-Jul-2010 14:35:58] that walks you through it
[20-Jul-2010 14:36:01] i think it's different tan that, there's a lot of uses who use it a lot more heavily than you & didn't have issues. they just ahve different use cases.
[20-Jul-2010 14:36:08] (for rmatte)
[20-Jul-2010 14:36:23] cgibbons: I couldn't even set product model info
[20-Jul-2010 14:36:41] cgibbons: so my device inventory reports were garbage
[20-Jul-2010 14:36:44] that was the major issue
[20-Jul-2010 14:36:56] and there was other fun stuff
[20-Jul-2010 14:37:00] 2.5.2 was a vast improvement
[20-Jul-2010 14:37:07] indeed
[20-Jul-2010 14:37:37] can create it all in the UI?
[20-Jul-2010 14:37:46] mducharme-laptop: initially, yeah
[20-Jul-2010 14:37:57] which sets up all the files/dirs for the zenpack
[20-Jul-2010 14:38:02] ok then
[20-Jul-2010 14:38:10] I should just remove the thing I created in zenoss/snmp?
[20-Jul-2010 14:38:12] you can't copy the file to it in the UI, but generating the ZenPack and exporting it is done in the UI
[20-Jul-2010 14:38:19] I didn't do very much with it at all
[20-Jul-2010 14:39:38] I added mibs already for the devices, should those go into the zenpack instead somehow?
[20-Jul-2010 14:40:18] mducharme-laptop: yeah you can ship mibs with your zenpack
[20-Jul-2010 14:40:47] go read docs above, they should answer most questions
[20-Jul-2010 14:42:50] grrrr, I had 2 small chunks taken from my leg earlier then the wounds were cauterized... the freezing is starting to wear off, painnnn
[20-Jul-2010 14:43:02] (biopsy)
[20-Jul-2010 14:43:07] codiene :)
[20-Jul-2010 14:43:13] yeh, wish I had some
[20-Jul-2010 14:43:26] they were just like "take tylenol or advil"
[20-Jul-2010 14:43:36] is there a doc i can read about adding a new component type for a device?
[20-Jul-2010 14:44:02] yes, the admin guide I assume
[20-Jul-2010 14:44:06] erm
[20-Jul-2010 14:44:09] developer's guide
[20-Jul-2010 14:44:11] meant to say
[20-Jul-2010 14:45:54] lmao, my buddy just got a phonecall and the person said: "hello i was wondering if anyone in the household speaks creol"
[20-Jul-2010 14:46:02] just wow
[20-Jul-2010 14:46:27] creole*
[20-Jul-2010 14:48:03] that is wow
[20-Jul-2010 14:48:08] quote of the week tbh :P
[20-Jul-2010 14:48:13] and I've heard some doozies this week
[20-Jul-2010 14:48:15] :D
[20-Jul-2010 14:48:23] lol
[20-Jul-2010 14:50:01] how could I change the variables in email message format?
[20-Jul-2010 14:50:04] database?
[20-Jul-2010 14:50:15] with event transforms, if I drop an event, I take it it doesn't hit _any_ mysql db tables?
[20-Jul-2010 14:50:18] what do you mean by change them?
[20-Jul-2010 14:50:30] Simon4_: correcto
[20-Jul-2010 14:50:40] I get "%(eventUrl)s" as "localhost", obviously, a broken link
[20-Jul-2010 14:50:42] Simon4_: transforms are applied prior to events touching the database
[20-Jul-2010 14:50:47] rmatte: awesome :)
[20-Jul-2010 14:50:53] * Simon4_ prunes his messy db :)
[20-Jul-2010 14:51:12] chachan: then I suggest you update the hosts file
[20-Jul-2010 14:51:24] cachan: with a proper hostname
[20-Jul-2010 14:51:30] chachan*
[20-Jul-2010 14:51:46] chachan: zenactions.conf in $ZENHOME/etc will let you set the url to zope, which is then used in the url's in your alerts
[20-Jul-2010 14:52:05] Simon4_: thought it was set based on the hosts file by default
[20-Jul-2010 14:52:21] rmatte: yeah, that's useless if your zenoss install is behind a load balancer/etc/etc
[20-Jul-2010 14:52:28] true enough
[20-Jul-2010 14:52:29] so zenactions lets you set it to anything
[20-Jul-2010 14:52:45] good to know
[20-Jul-2010 14:52:53] I don't use email alerting much
[20-Jul-2010 14:53:15] we need less of it, it's a curse
[20-Jul-2010 14:53:16] Hey rmatte - that doc, it'll enable up/down info on ALL monitored interfaces correct?
[20-Jul-2010 14:53:43] uhm, I'm on VPS, I think that hostname is managed by the VPS host and not for guess OS
[20-Jul-2010 14:53:49] arrrghhh: well, only if that datapoint is in the template for that interface type
[20-Jul-2010 14:53:52] so, I can't see my hostname O.o
[20-Jul-2010 14:53:55] arrrghhh: but yes
[20-Jul-2010 14:54:09] chachan: you can also go to "settings/daemons then click "edit config" on zenactions"
[20-Jul-2010 14:54:21] Cool, thanks again rmatte.
[20-Jul-2010 14:54:21] and set the zopeurl property
[20-Jul-2010 14:54:27] arrrghhh: np
[20-Jul-2010 14:54:28] nice gui way to do it
[20-Jul-2010 15:05:40] <|rt|> I'm upgrading to 3.0 now....hopefully it will go smoothly...but figured since Zenoss isn't in production here yet it's a pretty safe upgrade
[20-Jul-2010 15:08:33] so simon4_ is there some graphical way of doing a modeler now?
[20-Jul-2010 15:09:02] mducharme-laptop: no, you still have to write it, but as per the doc I linked, you create your skeleton zenpack using the GUI, then you can put your modeller into it
[20-Jul-2010 15:09:07] mducharme-laptop: modeler plugins are all code bud
[20-Jul-2010 15:09:32] ok so like before I copy DeviceMap.py into the zenpack
[20-Jul-2010 15:09:34] ?
[20-Jul-2010 15:10:07] well obviously rename it if it's a duplicate of a plugin
[20-Jul-2010 15:10:13] you can't have 2 plugins named the same thing
[20-Jul-2010 15:10:17] well yes, I know
[20-Jul-2010 15:10:29] it will be called TrangoMap.py because I am modeling trango devices
[20-Jul-2010 15:10:32] and yes, copy it in to the plugins directory in the ZenPack
[20-Jul-2010 15:10:45] and make sure there's an __init__.py in that directory
[20-Jul-2010 15:15:21] ok
[20-Jul-2010 15:15:38] do I need to make an snmp directory inside plugins like there is in the main zenoss directory
[20-Jul-2010 15:15:40] or no
[20-Jul-2010 15:16:01] nope
[20-Jul-2010 15:17:43] ok there we go, I have my trangomap.py
[20-Jul-2010 15:22:37] what order should my plugin be in compared to the default ones
[20-Jul-2010 15:23:00] my collector plugin
[20-Jul-2010 15:23:07] it doesn't really matter, I tend to put mine at the bottom just out of habit
[20-Jul-2010 15:28:22] hey all
[20-Jul-2010 15:29:02] i have been trying to query wmi from zenoss, but have been failing. i can query wmi from wmiexplorer without issue, any ideas?
[20-Jul-2010 15:30:34] sideone: well, what are you using in Zenoss to query it?
[20-Jul-2010 15:31:34] rmatte, i have installed 2 packs, (that we spoke about yesterday), wmidatasource and wmiperf_windows
[20-Jul-2010 15:31:50] however, even prior to that install, i had issues with auth from zWin user and pass
[20-Jul-2010 15:31:53] i would imagine
[20-Jul-2010 15:33:07] did you make sure those are configured correctly?
[20-Jul-2010 15:33:16] also, where did you put the device?
[20-Jul-2010 15:33:20] under config properties, i have zWinuser as domain\user (user==domain admin), and zWinPassword as the correct password
[20-Jul-2010 15:33:45] i have tried putting it in server/win and another server under /cim/wmi
[20-Jul-2010 15:33:52] both seem to have the same issue.
[20-Jul-2010 15:33:58] /cim/wmi is where it needs to be
[20-Jul-2010 15:34:11] so don't put it anywhere other than there
[20-Jul-2010 15:34:27] can i move the server to another place?
[20-Jul-2010 15:34:33] become zenoss user at the commandline and run: zenperfwmi run -v10
[20-Jul-2010 15:34:35] see what you get
[20-Jul-2010 15:34:36] from /server/win to cim/wmi
[20-Jul-2010 15:34:42] pl
[20-Jul-2010 15:34:43] ij
[20-Jul-2010 15:34:44] yes, move it
[20-Jul-2010 15:34:44] s/ok
[20-Jul-2010 15:35:34] hahah, sorry, is that a monitoring template? how do i move the server?
[20-Jul-2010 15:35:50] just drag the device in to the class
[20-Jul-2010 15:35:57] I assume you're using 3.0?
[20-Jul-2010 15:36:10] yes
[20-Jul-2010 15:36:17] yeh, just drag it
[20-Jul-2010 15:36:19] but im sorry, i dont know which screen allows for the drag
[20-Jul-2010 15:36:26] INFRASTRUCTURE
[20-Jul-2010 15:36:44] expand on the left until you see that organizer
[20-Jul-2010 15:36:49] sideone: is you overwrite template in /cim/wmi its bad
[20-Jul-2010 15:36:51] then drag the device from the list to that organizer
[20-Jul-2010 15:37:04] bigegor: I doubt he's over-written it
[20-Jul-2010 15:38:14] rmatte; i dragged, but under dev class, it still shows /server/windows after a refresh. does zen support firefox?
[20-Jul-2010 15:38:26] yes, it does
[20-Jul-2010 15:38:36] nm, im dum.
[20-Jul-2010 15:38:44] selected the row, and moved,
[20-Jul-2010 15:38:44] lol
[20-Jul-2010 15:38:48] i have too many plugins :)
[20-Jul-2010 15:38:49] k
[20-Jul-2010 15:39:21] ok this is weird
[20-Jul-2010 15:39:31] I wonder how I override this behavior
[20-Jul-2010 15:40:01] mducharme-laptop: order doesn't really matter much
[20-Jul-2010 15:40:05] HW Model .1.3.6.1.4.1.5454.1.30
[20-Jul-2010 15:40:15] mducharme: click on it
[20-Jul-2010 15:40:21] and change the name
[20-Jul-2010 15:40:30] oh
[20-Jul-2010 15:40:32] that simple?
[20-Jul-2010 15:40:42] Zenoss doesn't have every product in it's products database
[20-Jul-2010 15:40:48] so you need to manually define them sometimes
[20-Jul-2010 15:41:14] Is that device a netscreen?
[20-Jul-2010 15:41:22] oh no, trango
[20-Jul-2010 15:42:07] ok they are in the correct class now. ill check events.i also ran that command above, zenperfwmi run -v10, here is my result: http://www.pastebin.org/409344
[20-Jul-2010 15:42:51] #
[20-Jul-2010 15:42:52] 2010-07-20 16:40:20,402 DEBUG zen.WMIClient: Exception collecting query: Connection error NT_STATUS_IO_TIMEOUT
[20-Jul-2010 15:42:52] #
[20-Jul-2010 15:42:52] 2010-07-20 16:40:20,403 ERROR zen.zenperfwmi: Device fax2007: Connection error NT_STATUS_IO_TIMEOUT
[20-Jul-2010 15:43:02] getting a timeout for some reason
[20-Jul-2010 15:43:11] ok, that's better now
[20-Jul-2010 15:43:25] ok, next item on the list is getting it to see the interfaces.. it sees no interfaces right now
[20-Jul-2010 15:43:45] thats odd. did you see the error on mainsvr?
[20-Jul-2010 15:43:59] same issue?
[20-Jul-2010 15:44:20] 2010-07-20 16:40:20,401 ERROR zen.zenperfwmi: Device mainsvr: Connection error NT_STATUS_IO_TIMEOUT
[20-Jul-2010 15:44:46] yeh, same issue
[20-Jul-2010 15:44:46] i wonder why one product can query wmi and this cant?
[20-Jul-2010 15:45:00] sideone: is your server in LAN?
[20-Jul-2010 15:45:06] I have it querrying WMI perfectly fine on one of my servers
[20-Jul-2010 15:45:08] should I also make a copy of InterfaceMap.py
[20-Jul-2010 15:45:09] bigegor; yes.
[20-Jul-2010 15:45:15] you sure that WMI is configured correctly for full monitoring on that thing?
[20-Jul-2010 15:45:26] mducharme-laptop: yeah
[20-Jul-2010 15:45:26] rmatte; no i am not sure at all.
[20-Jul-2010 15:45:34] rmatte; i am very newb to wmi
[20-Jul-2010 15:45:39] neeeeewb
[20-Jul-2010 15:45:39] and then I customize that for the device.. gotcha
[20-Jul-2010 15:45:45] <----me
[20-Jul-2010 15:45:47] sideone: newwwwwwby!
[20-Jul-2010 15:45:51] * sideone is so newby :)
[20-Jul-2010 15:45:55] everyone, quick, point and laugh
[20-Jul-2010 15:46:03] * sideone pees allover himself.
[20-Jul-2010 15:46:22] so i can query from wmiexplorer, wouldnt that mean that wmi is functional on the server?
[20-Jul-2010 15:46:22] notice the dead silence after that last comment?
[20-Jul-2010 15:46:28] lol
[20-Jul-2010 15:46:30] rmatte; hahah
[20-Jul-2010 15:46:37] rmatte; nobody go near the smelly kid
[20-Jul-2010 15:46:45] sideone: yeh, it means it's partially functional
[20-Jul-2010 15:46:57] but, it doesn't mean that every WMI query that Zenoss needs is available
[20-Jul-2010 15:47:11] ah, is there a faq/howto on what is required for correct wmi functionality on a win* server?
[20-Jul-2010 15:47:23] bigegor: anything you know of?
[20-Jul-2010 15:47:31] no
[20-Jul-2010 15:47:38] rmatte; sorry wish i knew more about it.
[20-Jul-2010 15:47:38] sideone: google around
[20-Jul-2010 15:47:42] find a good WMI guide
[20-Jul-2010 15:47:48] the only way to know more is to learn
[20-Jul-2010 15:48:33] did i need to bind the script?
[20-Jul-2010 15:48:38] s/script/pack?
[20-Jul-2010 15:48:51] you mean the template?
[20-Jul-2010 15:49:10] aye
[20-Jul-2010 15:49:11] sideone: you can try to restart winmgmt service on the target server
[20-Jul-2010 15:49:16] if you don't see anything on the perf tab then it needs binding, but pretty sure it binds itself
[20-Jul-2010 15:49:20] ok. just reading now.
[20-Jul-2010 15:49:26] messing with wmic for testing./
[20-Jul-2010 15:49:34] k
[20-Jul-2010 15:49:57] thanks for the help again guys, just really need to see some data prior to getting approval on product
[20-Jul-2010 15:50:16] np
[20-Jul-2010 15:50:23] np
[20-Jul-2010 15:51:13] hmm, wmic seems to work properly uerying for Win32_computerSystem.
[20-Jul-2010 15:51:14] brb
[20-Jul-2010 15:53:13] sideone: I found syntax is very important to the WMI datasource zenpack
[20-Jul-2010 15:53:23] http://www.pastebin.org/409372
[20-Jul-2010 15:53:42] twm1010: well, he's just using the stock pack and it's not working apparently
[20-Jul-2010 15:53:52] this looks a bit better i think, changed my slashes for domain\user to domain/user .. been in windows for too long, forgot my smbclient days :)
[20-Jul-2010 15:53:54] he's getting NT_STATUS_IO_TIMEOUT errors
[20-Jul-2010 15:53:58] brb 5 mins, tqla shot time!
[20-Jul-2010 15:54:14] sideone: I have the slash backwards in my Zenoss and it works fine
[20-Jul-2010 15:54:22] monitoring and tequila, best mix ever
[20-Jul-2010 15:54:26] hahah
[20-Jul-2010 15:54:26] lol
[20-Jul-2010 15:54:34] I can't stand tequila
[20-Jul-2010 15:54:36] * sideone idles..and drinks.. and idles..
[20-Jul-2010 15:54:39] * Simon4_ tends to lean towards red wine personally
[20-Jul-2010 15:54:53] rmatte; use pineapple juice as a chaser and it rox!
[20-Jul-2010 15:55:11] I had some Johnnie Walker Blue Label last night, pure sex
[20-Jul-2010 15:55:38] $20 for an ounce
[20-Jul-2010 15:55:43] slurp
[20-Jul-2010 15:55:51] blue label is tasty
[20-Jul-2010 15:56:50] yup
[20-Jul-2010 15:57:20] NO
[20-Jul-2010 15:57:23] < scotsman
[20-Jul-2010 15:57:31] lagavulin ftw.
[20-Jul-2010 15:57:34] sideone: you can try send WMI query with WMIClient.py (look at end of file)
[20-Jul-2010 16:05:00] k
[20-Jul-2010 16:05:03] thanks egor.
[20-Jul-2010 16:06:20] sorry i dont know python, does the py file take any variables?
[20-Jul-2010 16:06:46] http://www.pastebin.org/409414
[20-Jul-2010 16:08:28] well, I'm out, later folks
[20-Jul-2010 16:08:33] later rmatte
[20-Jul-2010 16:08:37] thanks for the help bud
[20-Jul-2010 16:08:44] np
[20-Jul-2010 16:09:40] egor
[20-Jul-2010 16:09:49] does the py file take any variables?
[20-Jul-2010 16:09:53] http://www.pastebin.org/409414
[20-Jul-2010 16:10:05] sideone: yes
[20-Jul-2010 16:13:25] hey there.
[20-Jul-2010 16:13:35] what arguements is it looking for?
[20-Jul-2010 16:13:48] sideone: you don't need to know python to change one string with your values
[20-Jul-2010 16:14:18] bigegor; sorry, i just dont know what values/arguements it is expecting
[20-Jul-2010 16:14:45] this modeler is not picking up the interfaces
[20-Jul-2010 16:14:56] or maybe it is but the fields are just not mapped correctly to get the data
[20-Jul-2010 16:15:15] like the webui might be trying to display a field that isn't mapped
[20-Jul-2010 16:15:32] sideone: replace string url = "//username:password@127.0.0.1/root/cimv2" with your values
[20-Jul-2010 16:15:38] thx
[20-Jul-2010 16:21:47] how can I see whether this thing is pulling any data on the interfaces
[20-Jul-2010 16:22:08] sideone: command lines parameters for WMIClient.py: WMIClient.py -c "domain\username:password" -q "Win32_OperatingSystem" -f "Name Version" -a "Name Version"
[20-Jul-2010 16:22:21] mducharme-laptop: you can run zenmodeler from the commandline like zenmodeler run -d devicename -v 10
[20-Jul-2010 16:22:25] and you'll see much more output
[20-Jul-2010 16:23:23] you can then put things like log.debug("SNMP results: %r", results) in your modeler to see what's goingon
[20-Jul-2010 16:23:56] sideone: c - credentials, q - query, f - properties, a - aliases
[20-Jul-2010 16:24:27] thx
[20-Jul-2010 16:24:36] np
[20-Jul-2010 16:24:42] bigegor_ is now known as bigegor
[20-Jul-2010 16:26:17] is there a section for target?
[20-Jul-2010 16:27:14] http://www.pastebin.org/409457
[20-Jul-2010 16:33:34] sideone: something wrong with you zenoss installation. Python looking for modules in /var/mail.
[20-Jul-2010 16:33:52] hmm. i just did the stack install. i can reinstall if needed
[20-Jul-2010 16:34:06] can i set a variable to have it look somewhere else?
[20-Jul-2010 16:34:56] if not, can i reinstall ontop of itself or remove and install?
[20-Jul-2010 16:35:04] or should i create another linux os?
[20-Jul-2010 16:36:12] sideone: wait a moment
[20-Jul-2010 16:36:20] k
[20-Jul-2010 16:40:14] sideone: wrong WMIClient.py ;)
[20-Jul-2010 16:40:19] :)
[20-Jul-2010 16:40:21] ok.. it doesn't even seem to be running the process function
[20-Jul-2010 16:40:28] wtf
[20-Jul-2010 16:41:11] sideone: use from ZenPacks/ZenPacks.community.WMIDataSource/... directory
[20-Jul-2010 16:42:17] from the egg?
[20-Jul-2010 16:44:02] hah i see the url string :)
[20-Jul-2010 16:44:23] in egg? ;)
[20-Jul-2010 16:44:41] could only find / on that file in the egg.. just chmod to x
[20-Jul-2010 16:45:35] still errors.
[20-Jul-2010 16:46:36] look at /usr/local/zenoss/ZenPacks/ZenPacks.community.WMIDataSource/ZenPacks/community/WMIDataSource
[20-Jul-2010 16:46:58] try echo $ZENHOME
[20-Jul-2010 16:47:06] its odd, i have (by default stack install) /usr/local/zenoss/zenoss/ZenPacks..
[20-Jul-2010 16:47:40] http://www.pastebin.org/409498
[20-Jul-2010 16:53:21] sideone: it is directory, and you can open it.
[20-Jul-2010 16:53:35] anyone here knows what needs to be enabled in a hosts.allow besides snmpd and mysqld in order for mysqlMonitor to work ?
[20-Jul-2010 16:54:03] if I removed my settings on those hosts.allow|deny then it works fine
[20-Jul-2010 16:54:18] but whn I put it back in place it doesn't
[20-Jul-2010 16:56:03] i have tried with both the URL string and the WMIClient.py -c "domain\username:password" -q "Win32_OperatingSystem" -f "Name Version" -a "Name Version" runline, but get the following: http://www.pastebin.org/409528
[20-Jul-2010 16:57:35] maybe i should just scratch the install and start again?
[20-Jul-2010 16:59:33] well ill be back tomorrow, bigegor, thanks for your help, just closing shop for today.
[20-Jul-2010 16:59:37] night all.
[20-Jul-2010 17:07:10] anyone upgraded from 2.5 to 3 without any issues on debian ?
[20-Jul-2010 17:10:29] <|rt|> dec3pti01: I just hit issues with Ubuntu...but it's 10.04 which isn't listed in the "supported OS's"
[20-Jul-2010 17:11:15] <|rt|> I was using the deb stack-installer
[20-Jul-2010 17:11:27] <|rt|> going to try the .bin stack-installer next
[20-Jul-2010 17:21:15] Any one availble?
[20-Jul-2010 17:23:25] Well when somone reads this, Im wodering how to intagrate a Floorplan or a server rack picture, that would be interactive that you could click on, maybe have the status if its up or down, etc. Im useing Zenoss Core.
[20-Jul-2010 17:26:34] polk330: I have heard of that being in enterprise - and there's always weathermap that someone wrote about on the wiki
[20-Jul-2010 17:27:16] Yeah i saw that, But will that work with what im wanting to do, ke4qqq ?
[20-Jul-2010 17:28:28] polk330: no idea - depends on what you are trying to model I guess. There's also openlayers - which is OSM - so theoretically that would be more realistic perhaps for up/down status
[20-Jul-2010 17:29:05] hmmm.. Like a server set up where i could tell weather they are up or down and like the subdevices etc
[20-Jul-2010 17:29:40] if it's a critical must have thing - I'd talk to someone from zenoss - they were talking about the rack/floorplan level stuff last february, and allow you to see - say all wiring closets, zoom in on one rack in a wiring closet, and down to a single switch .
[20-Jul-2010 17:29:48] * Simon4_ discovers diagram view in enterprise
[20-Jul-2010 17:29:49] wooo
[20-Jul-2010 17:29:50] being a core-only user I have no experience with it.
[20-Jul-2010 17:30:29] yeah I think it is called diagram view
[20-Jul-2010 17:30:30] yeah.
[20-Jul-2010 17:31:29] * Simon4_ needs to work out how to make that actually useful
[20-Jul-2010 17:31:35] vs just a bunch of icons vs a list of servers
[20-Jul-2010 17:38:16] bigegor_ is now known as bigegor
[21-Jul-2010 00:00:30] [disconnected at Wed Jul 21 00:00:30 2010]
[21-Jul-2010 00:00:31] [connected at Wed Jul 21 00:00:31 2010]
[21-Jul-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]
[21-Jul-2010 00:55:16] hello, anyone home?
[21-Jul-2010 00:59:55] I am running Ubuntu 10.04 LTS and have a fresh install of zenoss v.3 via apt-get and for some reason I am stuck at the http://x.x.x.x:8080/zport/dmd/qs-step1 <-- "when I push submit" nothing. I can bypass this by changing step1 to 2 and then bypass again with the link to get to the dashboard, however my hope is that someone can point me in a direction to make this work properly? Any muse is appreciated.
[21-Jul-2010 02:35:54] I am test driving Zenoss 3.0. If I want the SNMP based filesystem monitor template to ignore /proc (for example) how do I do that?
[21-Jul-2010 02:51:29] add it to zFileSystemMapIgnoreNames in zProperties
[21-Jul-2010 02:51:37] morning all
[21-Jul-2010 02:51:51] ello Simon4
[21-Jul-2010 02:54:41] time for some test 3.0 enterprise action, I think
[21-Jul-2010 02:55:07] ah... thanks
[21-Jul-2010 02:55:17] Simon4: liar, it's *obviously* time for funky event transforms!
[21-Jul-2010 02:55:25] hah, that was last night's fun ;)
[21-Jul-2010 02:55:37] "lets see what happens to the load on the db if I drop all debug events"
[21-Jul-2010 02:55:43] haha
[21-Jul-2010 02:56:12] what sort of environment are you monitoring with zenoss, btw?
[21-Jul-2010 02:56:15] we're currently doing 150 update queries/second, and it's hurting
[21-Jul-2010 02:56:21] servers / network kit?
[21-Jul-2010 02:56:39] about 400 linux servers + associated network kit + blade chassis + env stuff
[21-Jul-2010 02:57:07] about to triple at least in size over the next 6 months, so we're going through a tech refresh stage of monitoring hardware atm
[21-Jul-2010 02:57:34] nice
[21-Jul-2010 02:58:32] I've got more network kit per monitoring location than I do servers
[21-Jul-2010 03:04:02] Hmm... ok. I cannot anything that looks quite right with zProperties. Is it that I have to add zFileSystemMapIgnoreNames as a new mapping in order to ignore certain filesystems?
[21-Jul-2010 03:04:19] cannot find anything that looks right, that is
[21-Jul-2010 03:06:48] realitygeoff: proc is normaly a ramdisk, as such it would already be ignored by zFileSystemMapIgnoreTypes
[21-Jul-2010 03:07:13] scratch that, its a pseudo filesystem. let me check the type ;)
[21-Jul-2010 03:08:54] filesystem type of proc is proc ;)
[21-Jul-2010 03:09:09] Simon4: but proc is unknown for mibs
[21-Jul-2010 03:10:26] realitygeoff: ok, it should come up as other filesystem and will be ignored by default.
[21-Jul-2010 03:10:33] I have "other, ram, virtualMemory" and proc is ignored
[21-Jul-2010 03:11:06] and net-snmp will ignore it itself and not even export it in the hrStorageTable
[21-Jul-2010 03:11:28] I am in "configuration properties" of "events > perf > Filesystem" area and I see no such related settings. Am I just in the wrong place?
[21-Jul-2010 03:11:54] Or is that SNMP itself is misconfigured?
[21-Jul-2010 03:12:17] realitygeoff: zenoss 3 or 2?
[21-Jul-2010 03:12:23] Zenoss 3.0
[21-Jul-2010 03:12:36] realitygeoff: ignore for one device or for a subtree?
[21-Jul-2010 03:12:55] All of /server/linux class
[21-Jul-2010 03:13:05] click infrastructure
[21-Jul-2010 03:13:14] left side click server/linux
[21-Jul-2010 03:13:41] ok.. I'm with you.
[21-Jul-2010 03:13:43] click details blue arrow
[21-Jul-2010 03:13:59] configuration properties
[21-Jul-2010 03:14:24] long list of zProperties on the right side
[21-Jul-2010 03:14:34] check zFileSystemMapIgnoreTypes
[21-Jul-2010 03:14:57] ahh... ok. I see it now. And I also have "other, ram, virtualMemory"
[21-Jul-2010 03:15:08] so you should not see proc
[21-Jul-2010 03:15:29] yes... relatively long list, but no proc
[21-Jul-2010 03:15:42] so what was the originial problem
[21-Jul-2010 03:15:46] hehe.
[21-Jul-2010 03:16:11] I am getting 100% utilization alerts for /proc and /dev on some boxes.
[21-Jul-2010 03:16:26] they should not show up
[21-Jul-2010 03:16:42] dev is ramdisk (on modern unixes) and proc is other
[21-Jul-2010 03:17:07] Ok. Possibly SNMP issue?
[21-Jul-2010 03:17:20] realitygeoff: walk it with snmp and check what is reported.
[21-Jul-2010 03:17:35] snmpwalk -v 2c -c public hostname.example.com 1.3.6.1.2.1.25.2.3.1
[21-Jul-2010 03:21:06] HOST-RESOURCES-MIB::hrStorageDescr.33 = STRING: /proc
[21-Jul-2010 03:21:21] I assume that is usually not present?
[21-Jul-2010 03:21:22] HOST-RESOURCES-MIB::hrStorageType.33
[21-Jul-2010 03:22:03] (what is the value of this)
[21-Jul-2010 03:22:21] HOST-RESOURCES-MIB::hrStorageType.33 = OID: HOST-RESOURCES-TYPES::hrStorageFixedDisk
[21-Jul-2010 03:22:28] *wrong*
[21-Jul-2010 03:22:39] how old is the snmp server?
[21-Jul-2010 03:23:26] quite old. circa 2005
[21-Jul-2010 03:24:24] So we can fix the SNMP layer, or is fairly simple to ignore /proc in the mappings settings on Zenoss?
[21-Jul-2010 03:24:44] is it net-snmp? can you modify the configuration? Then i would add ifnoreDisk there (it should not be reported and annoy all snmp monitoring systems)
[21-Jul-2010 03:24:59] if you prefer you can also fix it in zenoss.
[21-Jul-2010 03:25:30] zFileSystemMapIgnoreNames takes a regular expression, so ^/(proc|dev) should work
[21-Jul-2010 03:25:36] I'd like both options. Igore in Zenoss and fix the snmp during regular maintenance windows.
[21-Jul-2010 03:25:49] Ah cool. I can try that.
[21-Jul-2010 03:26:09] That should do me for now. Thanks for taking the time help me out with that.
[21-Jul-2010 03:28:24] run zenmodell in verbose mode from commandline and it will tell you what it ignores
[21-Jul-2010 03:28:27] -l
[21-Jul-2010 03:29:09] thanks
[21-Jul-2010 03:44:24] i know, it's early in the morning, but i'll ask so if anyone wakes up and sees it later...
[21-Jul-2010 03:44:40] It is the middle of the day for some us.
[21-Jul-2010 03:45:09] But I don't know much of anything.
[21-Jul-2010 03:45:32] i have a custom os-component. LLDP Links. It knows the IP of a remote System. I would like to make a link of the IP, best, directly to the device owning it. How and when would i do that? During Model or Rendering?
[21-Jul-2010 03:45:59] realitygeoff: most devs are in the channel around 20:00-24:00 UTC
[21-Jul-2010 03:46:28] fragfutter: you would link that during modelling I think
[21-Jul-2010 03:47:25] * Simon4 has an application that needs to do similar (linking load-balancer node definitions to the actual hosts)
[21-Jul-2010 03:48:18] Simon4: so i need: access to dmd (to search for the device). That would slow down modelling and on first discovery it won't find the device if it hasn't been in the DB before (less of an issue, lldp tends to be fairly static)
[21-Jul-2010 03:48:39] fragfutter: yeah, that's about where I got to
[21-Jul-2010 03:48:49] Simon4: and how would i store this and explain it to zenoss to render a link?
[21-Jul-2010 03:49:21] so you're just creating a relation that points to a Device object, that much I know
[21-Jul-2010 03:49:41] then somehow in the template you use getUrlLink for the relation (I think that's the function call)
[21-Jul-2010 03:49:59] I haven't done this yet, so don't take me as gospel
[21-Jul-2010 03:54:23] I need a devs in channel day and some time to document the problem better so I can ask
[21-Jul-2010 03:55:11] zenoss needs structured documentation for developers.
[21-Jul-2010 03:55:49] yeah, the dev guide is okay, but there's a huge amount that's left out for people to try and discover themselves
[21-Jul-2010 03:56:05] I end up grepping through the source or the community zenpacks a lot
[21-Jul-2010 03:58:25] ah man that's a pain
[21-Jul-2010 03:58:29] * Simon4 gets v3 installed
[21-Jul-2010 03:58:37] navigate to "infrastructure" then to a device
[21-Jul-2010 03:59:01] click on the events to view the events, and you get taken all the way out of infrastructure into events, with no easy way to navigate back to where you were
[21-Jul-2010 03:59:24] hit back ;)
[21-Jul-2010 03:59:38] right, without using the generally unreliable browser controls ;)
[21-Jul-2010 03:59:51] and what if I want to nav quickly to another device in the infrastructure tree?
[21-Jul-2010 03:59:52] but yes, navigation got worse.
[21-Jul-2010 04:00:04] type it in the top right corner search box...
[21-Jul-2010 04:00:09] ... that used to be there
[21-Jul-2010 04:00:14] * Simon4 laughs :D
[21-Jul-2010 04:00:38] * fragfutter is tempted to write a small script and integrate it in the quick search of the browser
[21-Jul-2010 04:00:50] time to open me some tickets in the enterprise ticketing system I think ;)
[21-Jul-2010 04:01:33] http://dev.zenoss.org/trac/ticket/7061
[21-Jul-2010 04:01:55] sweet
[21-Jul-2010 04:02:19] maybe our enterprise license can add some weight to these things
[21-Jul-2010 04:02:22] if you have enterprise, you should still open a ticket (more power)
[21-Jul-2010 04:02:26] yeah, will do
[21-Jul-2010 04:02:28] and will link that
[21-Jul-2010 04:02:29] :)
[21-Jul-2010 04:03:57] thanks.
[21-Jul-2010 04:05:17] and the search box on the left doesn't work as i expect. type in a devicename there. nothing is found, it searches for device classes. remove your search and it won't update to all device classes again.
[21-Jul-2010 04:05:50] sometimes ajax is not the best thing to do...
[21-Jul-2010 04:15:29] I am running Ubuntu 10.04 LTS and have a fresh install of zenoss v.3 via apt-get and for some reason I am stuck at the http://x.x.x.x:8080/zport/dmd/qs-step1 <-- "when I push submit" nothing. I can bypass this by changing step1 to 2 and then bypass again with the link to get to the dashboard, however my hope is that someone can point me in a direction to make this work properly? Any muse is appreciated.
[21-Jul-2010 04:15:36] Hello again. Just a simple question. If updating mappings or other properties in zProperties, is a Zenoss restart needed?
[21-Jul-2010 04:16:13] realitygeoff: IMHO no. zProperties are in the zodb and should be picked up live. They will be effective during the next model run
[21-Jul-2010 04:16:25] suntzu: cookies? adblock? noscript?
[21-Jul-2010 04:17:11] i have the server setup in ie as trusted and have "everything" enables "except the kitchen sink"
[21-Jul-2010 04:17:50] nothing on the local system in ie is blocking it. It just seems to hang.
[21-Jul-2010 04:18:41] the dashboard works and seems fine, with the exception of having to bypass qs-step1 each time I log in...
[21-Jul-2010 04:18:45] any thoughts?
[21-Jul-2010 04:18:51] no
[21-Jul-2010 04:19:02] ok, thanks.
[21-Jul-2010 04:19:48] do you know which file in zenoss forces the qs-step1 process on login?
[21-Jul-2010 04:20:10] as in: "where I can manually edit it out"?
[21-Jul-2010 04:38:19] 'lo
[21-Jul-2010 04:38:38] does zenoss allow you to select and combine/compare graphs?
[21-Jul-2010 04:38:58] for example you have a few servers and want to combine the graphs of their network interfaces
[21-Jul-2010 04:55:13] kokey: you want "Multi-Graph Reports"
[21-Jul-2010 04:55:21] there are some good vids on the zenoss site for how to make them
[21-Jul-2010 04:57:28] kokey: heya
[21-Jul-2010 04:57:39] kokey: probably not easily gonna do what I think you want it to do
[21-Jul-2010 04:58:22] but knowing you, you might just as well beat it into submission too.. :P
[21-Jul-2010 04:58:45] .wub 3
[21-Jul-2010 04:58:46] sldwgrtkrwnwpih4
[21-Jul-2010 05:03:36] i should make a video of my graphing stuff
[21-Jul-2010 05:04:16] multi graph reports, sounds promising then
[21-Jul-2010 05:04:29] it would be quite impressive to see it one day, based on what you told us about last time :)
[21-Jul-2010 05:04:31] just as i logged out of the demo
[21-Jul-2010 05:12:54] one day I'll do the effort of figuring out why the eventloop on the jabber bot keeps letting the XMPP session die
[21-Jul-2010 05:14:27] why does the layer3_catalog not contain an index on the ip address?
[21-Jul-2010 05:15:04] so what is the preferred way to search an IP through dmd?
[21-Jul-2010 05:53:29] hmm, zenoss 3 seems to want a _lot_ of vertical screen res
[21-Jul-2010 05:53:39] my 15" mac barely shows anything useful
[21-Jul-2010 07:19:16] hmm, anyone know where to configure how long the heartbeat timeouts are?
[21-Jul-2010 07:19:27] we get a lot of flapping, and it would be nice to just extend the timeout a little
[21-Jul-2010 07:46:08] zendaemon configs, I think
[21-Jul-2010 07:47:24] Simon4: "when in doubt, [ack-]grep"
[21-Jul-2010 07:47:29] :)
[21-Jul-2010 08:06:04] morning all
[21-Jul-2010 08:17:33] in a ZComponentGridPanel, how can i display the output of a function instead of a property?
[21-Jul-2010 08:34:45] bigegor_: do you know how to use ZComponentGridPanel?
[21-Jul-2010 08:38:02] cgibbons: do you know how to use ZComponentGridPanel?
[21-Jul-2010 08:38:12] yes
[21-Jul-2010 08:38:32] in a ZComponentGridPanel, how can i display the output of a function instead of a property?
[21-Jul-2010 08:38:58] you use the renderer: attribute in your column definition and set the equal to a function that does the work
[21-Jul-2010 08:39:33] cgibbons: not a javascript function. the object i want to render has a function that generates the output i want to display
[21-Jul-2010 08:40:52] ok, it's really the same as a property then in python, so you're info object would just do the same thing. i'm not sure if ProxyProperty in the info object would work (but I think it would) or if you'd need to write a function in your info object that just did self._object.myfunc() but either way that should be it
[21-Jul-2010 08:41:09] with the stack installer, the defaults show, /usr/local/zenoss for the install folder. last time i built from the stack it created /usr/local/zenoss/zenoss, should i adjust the path to /usr/local ?
[21-Jul-2010 08:43:01] cgibbons: so i don't need to add it to my object._properties? i just define a object.getMyFunction and in the jsvascript of the zgrid i add it?
[21-Jul-2010 08:43:14] yep
[21-Jul-2010 08:45:36] cgibbons: doesn't work.
[21-Jul-2010 08:47:19] there's a lot of little steps: add a field to the fields list in your component grid; add a column in your colmns list of the component grid for that field; make sure the field as a matching attribute name in your info object & interface; make sure the info object gets that data from somewhere (your function)
[21-Jul-2010 08:47:19] hello bigegor
[21-Jul-2010 08:47:41] cgibbons: ah, the interface and info object.
[21-Jul-2010 08:49:48] cgibbons: yeah. putting it through all interface and info objects it worked. ProxyProperty also works.
[21-Jul-2010 08:55:30] hi
[21-Jul-2010 08:55:56] afternoon
[21-Jul-2010 08:56:09] * Simon4 puppets up his zenoss boxen
[21-Jul-2010 08:57:04] cgibbons: i want to return a link to a device (if i can find the device in zenoss) or a textual string otherwise. Can you give me a hint how this is done?
[21-Jul-2010 08:57:37] cgibbons: i could return the device object and use a renderer, but how would i then return only a string if i can't find it?
[21-Jul-2010 08:58:07] cgibbons: or i could return a string with the full link, but this sounds wrong.
[21-Jul-2010 08:58:27] what's the column you're trying to display?
[21-Jul-2010 09:00:03] cgibbons: custom os component (containing LLDP links). i know the name and ip lldp tells me. Normaly i search dmd.Networks for the ip, get the device, return this link. But if the remote end of the lldp is not managed by zenoss i would come up empty. then i want to return the text i found during modelling.
[21-Jul-2010 09:14:08] Anyone have any idea why I'd get an event like: Unable to save data for process-monitor RRD Devices/path/to/a/device/and/os/monitor ...
[21-Jul-2010 09:14:38] jmp242: filesystem permissions?
[21-Jul-2010 09:17:40] That's what I thought, but the permissions seem OK to me -
[21-Jul-2010 09:17:42] -rwxrwxr-x 1 zenoss zenoss 35432 Jul 21 10:16 count_count.rrd
[21-Jul-2010 09:17:49] for instance
[21-Jul-2010 09:18:27] jmp242: maybe the data it's trying to save is incompatible with the rrd type?
[21-Jul-2010 09:18:31] i.e. trying to put a float into a counter
[21-Jul-2010 09:20:02] weird. I wonder why it'd be trying to do that.
[21-Jul-2010 09:20:22] This is a process monitor
[21-Jul-2010 09:20:31] on a Linux computer and using zenprocess to do the monitoring
[21-Jul-2010 09:20:49] As far as I can tell, nothing odd, though we had recently upgraded from 2.4.5 to 2.5.2
[21-Jul-2010 09:21:14] The event shows Data was value= 16, type=GAUGE, min=U, max=U
[21-Jul-2010 09:32:49] which function mangles "Gigabit10/2" to Gigbit10_2?
[21-Jul-2010 09:34:17] prepId
[21-Jul-2010 09:35:49] Simon4: does it make sure that when i have "Gigabit10/2" and "Gigbit10-2" as interface descriptions that it doesn't produce the same ids?
[21-Jul-2010 09:36:45] pirates on deck!
[21-Jul-2010 09:37:11] fragfutter: hmmmm, I haven't tried, I'm guessing they'll end up identical though
[21-Jul-2010 09:37:39] Simon4: which would kill during the model process, as this id is used as node-id in the zodb
[21-Jul-2010 09:39:34] Yar
[21-Jul-2010 09:40:10] I can't remember what problem I was having. Damnit.
[21-Jul-2010 09:40:31] no problem grog couldn't solve, obviously
[21-Jul-2010 09:41:12] Oh that was it... so I changed the default port on zenoss, and in the emails generated from zenoss, the links have the default port instead of the modified one
[21-Jul-2010 09:41:28] I prefer rum. :D
[21-Jul-2010 09:41:33] arrrghhh: have some more grog.
[21-Jul-2010 09:42:28] Lol
[21-Jul-2010 09:42:48] and now, i have no idea.
[21-Jul-2010 09:42:52] -w
[21-Jul-2010 09:50:43] can i tell findObjectsById to do an exact match?
[21-Jul-2010 10:05:45] dec3pti01 is now known as dec3pti0n
[21-Jul-2010 10:08:21] rmatte, you around dude?
[21-Jul-2010 10:19:27] morning
[21-Jul-2010 10:20:38] I have a question.. what determines the fields stored for the device, or fields for OS/Interfaces
[21-Jul-2010 10:21:08] I have certain types of devices where those fields do not make sense
[21-Jul-2010 10:21:35] what do you mean?
[21-Jul-2010 10:21:37] for instance, with this brand of wireless radio that we use, there is no single "speed" setting
[21-Jul-2010 10:21:51] there is an MIR and a CIR, for up and down directions
[21-Jul-2010 10:22:19] mducharme-laptop: the way stuff is gathered on the OS tab is hardcoded
[21-Jul-2010 10:22:22] I get the subscriber units to appear in zenoss as being interfaces of the AP
[21-Jul-2010 10:22:40] ok
[21-Jul-2010 10:22:50] I'm just trying to determine I'm doing this in the best possible way
[21-Jul-2010 10:22:52] mducharme-laptop: it's designed for the most common scenario
[21-Jul-2010 10:23:21] mducharme-laptop: I usually just monitor wireless controllers via traps, and then just do ping monitoring on the actual base stations if possible
[21-Jul-2010 10:23:26] I assume that's what you're dealing with?
[21-Jul-2010 10:23:45] we do traffic monitoring of individual customers
[21-Jul-2010 10:24:04] i.e. individual "interfaces" on the AP
[21-Jul-2010 10:24:21] I see
[21-Jul-2010 10:24:40] well, you'll have a tough time doing that if the interfaces don't actually provide an overall speed setting via SNMP
[21-Jul-2010 10:25:13] the counters are there for packets per second etc
[21-Jul-2010 10:25:15] since Zenoss relies on that to threshold on utilization
[21-Jul-2010 10:25:20] we collect the data right now with cacti
[21-Jul-2010 10:25:34] so cacti can do this but zenos can't?
[21-Jul-2010 10:25:34] when it says speed it's looking for the maximum capacity of the interface
[21-Jul-2010 10:25:53] oh, well I can just hard-code that in the collector
[21-Jul-2010 10:25:56] 1000000000 for example
[21-Jul-2010 10:26:08] ok, I can hard code that
[21-Jul-2010 10:26:12] if that's the issue
[21-Jul-2010 10:26:16] yup
[21-Jul-2010 10:26:25] ok, that's not a problem
[21-Jul-2010 10:27:42] I was also trying to figure out what to do with the overall traffic monitor for the AP
[21-Jul-2010 10:27:53] the AP has like a total traffic snmp graph
[21-Jul-2010 10:28:16] you can do the same thing in Zenoss
[21-Jul-2010 10:28:24] you just need to modify the interface template
[21-Jul-2010 10:28:27] add a new graph
[21-Jul-2010 10:28:28] but the eth0 interface on the AP is at totally different OIDs than the SU interfaces
[21-Jul-2010 10:28:34] and have it graph the sum of in/out
[21-Jul-2010 10:29:05] yeh, that's just stupidity on the vendor's part
[21-Jul-2010 10:29:09] I think the total traffic is basically the traffic on eth0 but eth0 does not appear as an interface because it's at a diferent spot
[21-Jul-2010 10:29:12] they should have treated them like any other interfaces
[21-Jul-2010 10:29:20] yes, I know
[21-Jul-2010 10:29:29] I will have to find some clever way in python of integrating it, I think
[21-Jul-2010 10:29:35] yup
[21-Jul-2010 10:29:50] adding a fake first record which will be eth0
[21-Jul-2010 10:40:31] cgibbons: now that i return a formatted link (which is fine for the grid), the details page of my object will barf. MyObject.myfunction is not JSON serializable
[21-Jul-2010 10:47:29] I still have to find some way of pulling the zenoss CMDB into OTRS
[21-Jul-2010 10:48:14] some kind of synchronization solution
[21-Jul-2010 10:48:19] it could end up being a little ugly
[21-Jul-2010 10:50:08] I just use otrs for ticketing, not as a cmdb
[21-Jul-2010 10:50:16] I find it's cmdb system isn't that great
[21-Jul-2010 10:50:28] rmatte yes but we are moving to itil
[21-Jul-2010 10:50:38] so we need to log tickets against cmdb items
[21-Jul-2010 10:50:51] I see
[21-Jul-2010 10:50:52] I don't actually need all the information, I just need all the devices/locations across
[21-Jul-2010 10:51:07] it makes senes to have some kind of data sync
[21-Jul-2010 10:51:13] yeh, we coded our own cmdb from scratch
[21-Jul-2010 10:51:19] it still needs a lot of work though
[21-Jul-2010 10:51:22] but it's coming along
[21-Jul-2010 10:51:24] having zenoss push down cmdb updates to OTRS
[21-Jul-2010 10:52:08] have fun with that hehe
[21-Jul-2010 10:52:44] We have zenoss/otrs ticket integration coded, but we heavily customized otrs to make it work
[21-Jul-2010 10:52:56] when an issue comes in a ticket is automatically generated in otrs from Zenoss
[21-Jul-2010 10:54:09] I was trying to find some kind of hook in the CMDB so that when the configuration is changed I can trigger the sync code
[21-Jul-2010 10:54:11] that way our analysts only have to watch otrs
[21-Jul-2010 10:54:21] since we have 14 different Zenoss core servers
[21-Jul-2010 10:54:40] yeh, not sure about hook code in cmdb
[21-Jul-2010 10:55:53] wow. zenoss is calling for ajax updates on every scroll of a componentGrid?
[21-Jul-2010 10:56:21] it calls for ajax updates on _everything_
[21-Jul-2010 10:56:22] fragfutter: hunh?
[21-Jul-2010 10:56:26] Crap. I can only scroll back like 100 lines.
[21-Jul-2010 10:56:41] fragfutter: the whole interface is heavily ajax based
[21-Jul-2010 10:57:56] rmatte: and i loaded a ComponentGrid. One of those fields is provided by my Objects, and they log every time this field is accessed (debugging). so i scroll through the grid and get a log entry for each row. ok. scroll up again and ... more log entries
[21-Jul-2010 10:58:04] Rmatte, did you see my relatively minor issue? If you have any suggestions, it's probably a setting I missed.
[21-Jul-2010 10:58:50] is it possible to upgrade the zenoss-stack over the gui or just with apt-get ?
[21-Jul-2010 10:59:19] i have already installed the PreUpgrade30 ZenPack .egg
[21-Jul-2010 11:00:25] There is an upgrade section, not sure if it'll actually upgrade or just notify you of upgrades however...
[21-Jul-2010 11:00:39] yeah that only seems to notify
[21-Jul-2010 11:02:02] dec3pti0n: assuming you used the zenoss-stack deb, download the .deb for the new version to the box and dpkg -i from the command line
[21-Jul-2010 11:02:46] yep i did
[21-Jul-2010 11:31:49] wow very nice interface :)
[21-Jul-2010 11:32:07] I do miss the server time on the right top corner though
[21-Jul-2010 11:32:11] rmatte: did you fix the shift thing?
[21-Jul-2010 11:37:51] dec3pti0n you can do it with apt get yes
[21-Jul-2010 11:38:13] dec3pti0n or you can get the .bin file, and install it, it will detect the previous install, just make sure you bakcup your data just incase
[21-Jul-2010 11:53:32] Hey all, sorry if someone already answered this but I changed the port on Zenoss, and in the emails Zenoss sends out it's using the default port of 8080. How can I update it?
[21-Jul-2010 11:58:55] arrrghhh: Settings -> Daemons -> zenactions -> Edit config -> zopeurl
[21-Jul-2010 12:01:56] Sweet! I figured it was simple.
[21-Jul-2010 12:06:30] dec3pti0n: yeh, the server time missing sucks
[21-Jul-2010 12:06:46] shhh, there he is
[21-Jul-2010 12:07:33] i just reinstalled linux, and reinstalled the zen stack for ubuntu x64.. i have added 2 wmi (source/perf) and am watching the events. I do not see the new server that i have installed (with zwinuser/zwinpass) on the dashboard under dev/cmi. Is there a place to find the status of the server if it is not listed?
[21-Jul-2010 12:08:02] sideone: if it's not listed then the status is good
[21-Jul-2010 12:08:05] that's the whole point
[21-Jul-2010 12:08:11] ahh!
[21-Jul-2010 12:08:15] so it only shows problems!
[21-Jul-2010 12:08:17] i see
[21-Jul-2010 12:08:22] indeed
[21-Jul-2010 12:08:26] thanks rm!
[21-Jul-2010 12:08:31] np
[21-Jul-2010 12:08:43] which portlet are you referring to out of curiosity?
[21-Jul-2010 12:08:47] Device Issues?
[21-Jul-2010 12:08:55] so, this install worked alot better. i did change the default path for the ubuntu stack to /usr/local instead of /usr/local/zenoss
[21-Jul-2010 12:09:05] as the default created .../zenoss/zenoss
[21-Jul-2010 12:09:39] rmatte; i am unsure of the dashboard, so i have added all portlets avail
[21-Jul-2010 12:09:42] yeh, it's supposed to be /usr/local/zenoss/zenoss
[21-Jul-2010 12:09:46] that's the proper way to have it
[21-Jul-2010 12:09:50] don't go messing around with it lol
[21-Jul-2010 12:09:50] ahh. crap
[21-Jul-2010 12:10:03] ok i think ill take one more shot at this thing :)
[21-Jul-2010 12:10:08] lol
[21-Jul-2010 12:10:14] i thought that was failed.
[21-Jul-2010 12:10:17] what portlet were you saying you weren't seeing that device in?
[21-Jul-2010 12:10:28] I want to make sure what I told you wasn't misinformation, pretty sure I'
[21-Jul-2010 12:10:32] pretty sure I'm right
[21-Jul-2010 12:10:33] on the dash, i dont see the server anywehre
[21-Jul-2010 12:10:37] but I don't use the dahboard much
[21-Jul-2010 12:10:41] dashboard*
[21-Jul-2010 12:10:45] yeh, but in which portlet?
[21-Jul-2010 12:10:52] the dashboard has many different portlets available
[21-Jul-2010 12:10:59] it's fully configurable
[21-Jul-2010 12:11:11] i haev all up, root org, oject watch, dev issues, prod states, messages, and zenoss issues
[21-Jul-2010 12:11:22] fragfutter: did you figure out your component grid thinger? i was gone for a while...
[21-Jul-2010 12:11:24] yeh
[21-Jul-2010 12:11:33] it'll only show up in device issues if there is actually an issue
[21-Jul-2010 12:11:34] so to check status on servers, i will need to check under infrastructure
[21-Jul-2010 12:11:55] if you don't see anything in device issues then there are no issues with that device
[21-Jul-2010 12:12:02] I always check under infrastructure personally anyways
[21-Jul-2010 12:12:04] i guess, i just want to see the servers stats, event logs, and other items
[21-Jul-2010 12:12:08] I hardly ever look at the dashboard
[21-Jul-2010 12:12:35] Isn't infrastructure only in 3.0+?
[21-Jul-2010 12:12:37] yeah, i will have email notifications as well, so i guess i wont be looking at dash either
[21-Jul-2010 12:12:43] arrrghhh: yes, it is
[21-Jul-2010 12:12:51] arrrghhh: I'm just using that as an example
[21-Jul-2010 12:12:56] it's "Device List" in 2.5.2
[21-Jul-2010 12:12:59] Fair enough :P
[21-Jul-2010 12:13:14] rmatte; can i pull in event logs from the server into zenoss?
[21-Jul-2010 12:13:15] everyone seems to have jumped at upgrading to 3.0 in here
[21-Jul-2010 12:13:20] so I have to use the 3.0 lingo
[21-Jul-2010 12:13:21] :P
[21-Jul-2010 12:13:23] heh
[21-Jul-2010 12:13:25] i like 3.0
[21-Jul-2010 12:13:34] meh, there are some annoying interface problems
[21-Jul-2010 12:13:42] started with 2.5 a month ago, then 3.0 shot out :)
[21-Jul-2010 12:13:43] some of which will be fixed in 3.0.1 hopefully
[21-Jul-2010 12:14:40] rmatte; so i have wmi with 3 sensors loaded, how do i see the values?
[21-Jul-2010 12:14:56] im in infrastructure, and am not sure what im looking for :)
[21-Jul-2010 12:14:57] sideone: what do you mean by 3 sensors?
[21-Jul-2010 12:15:10] click on the device obviously
[21-Jul-2010 12:15:12] oh it has wmi for PerfOSMemory, Processor, and sysUptime
[21-Jul-2010 12:15:23] i am looking in device
[21-Jul-2010 12:15:24] click on the device you want to view
[21-Jul-2010 12:15:29] then select whatever on the left hand side
[21-Jul-2010 12:15:35] interfaces, disks, whatever
[21-Jul-2010 12:15:40] Graphs will show performance graphs
[21-Jul-2010 12:15:44] such as CPU, Memory, etc...
[21-Jul-2010 12:15:55] thats the thing, there is no data
[21-Jul-2010 12:16:05] do you see graphs at all?
[21-Jul-2010 12:16:13] sorry, i see the load under graphs
[21-Jul-2010 12:16:21] execpt for memory util
[21-Jul-2010 12:16:24] the load?
[21-Jul-2010 12:16:28] CPU you mean?
[21-Jul-2010 12:16:36] I'd like to go to 3.0, but I'm going to wait.
[21-Jul-2010 12:16:37] yeah, i appologize i do see data
[21-Jul-2010 12:16:45] k
[21-Jul-2010 12:16:48] just not under memory
[21-Jul-2010 12:16:50] but you're not seeing memory data?
[21-Jul-2010 12:16:54] nope.
[21-Jul-2010 12:16:55] what does the memory graph look like?
[21-Jul-2010 12:16:59] Especially since this isn't in production, there's no rush.
[21-Jul-2010 12:17:02] does it says "No RRD" or something?
[21-Jul-2010 12:17:07] There was a problem rendering this graph. Either the file does not exist or an error has occurred. Initial graph creation can take up to 5 minutes. If the graph still does not appear, look in the Zope log file $ZENHOME/log/event.log for errors.
[21-Jul-2010 12:17:08] or do you see "nan" for the values?
[21-Jul-2010 12:17:29] ok, and when did you finish adding the WMI stuff in?
[21-Jul-2010 12:17:35] it takes up to 15 minutes for graphing to start
[21-Jul-2010 12:17:38] ahh
[21-Jul-2010 12:17:46] thats gotta be it, i knew it was longer than 5 mins
[21-Jul-2010 12:17:47] :)
[21-Jul-2010 12:17:48] it needs 3 polling cycles
[21-Jul-2010 12:17:51] gotcha
[21-Jul-2010 12:18:02] now i need to figure out what i can monitor :)
[21-Jul-2010 12:18:07] though it's odd that you're seeing cpu data
[21-Jul-2010 12:18:15] since the memory data should have been gathered at the same time
[21-Jul-2010 12:18:15] can i view ms event logs through this system?
[21-Jul-2010 12:18:33] you can't actually view the logs
[21-Jul-2010 12:18:39] i would like to have the ms server send all critical events to zo.
[21-Jul-2010 12:18:43] you can have it alert when event log events come in
[21-Jul-2010 12:19:00] rmatte; so wmi gathers event logs, but does not display?
[21-Jul-2010 12:19:02] that part is clearly explained in the Zenoss admin guide
[21-Jul-2010 12:19:06] please give it a read
[21-Jul-2010 12:19:11] cool,. i will.
[21-Jul-2010 12:19:13] thanks again rm.
[21-Jul-2010 12:19:29] not a ton of defects will be in core for 3.0.1: http://bit.ly/bXpEDT
[21-Jul-2010 12:19:33] I don't know if it actually uses WMI for the event logs, I think it uses something different
[21-Jul-2010 12:19:39] got through about 60% of guide now, but still have more to do!
[21-Jul-2010 12:19:48] you just set it up to monitor them in zProperties, and set the severity you want to see
[21-Jul-2010 12:19:58] and as they come in they'll be grabbed by Zenoss and events will be created for them
[21-Jul-2010 12:20:08] it can be quite noisy so you'll have to probably do some tuning
[21-Jul-2010 12:20:35] sweet! i saw the property, but was unsure of its effect
[21-Jul-2010 12:20:46] well, that's the effect lol
[21-Jul-2010 12:20:53] (it, being zeneventlog, does use WMI to collect from the MS Event Log on a Windows device)
[21-Jul-2010 12:20:55] hahah ! thanks buddy
[21-Jul-2010 12:20:56] but seriously, download the zenoss admin guide in pdf and read it
[21-Jul-2010 12:21:02] the UI has changed but the concepts have not
[21-Jul-2010 12:21:19] it'll answer 90% of what you're likely to ask in here
[21-Jul-2010 12:21:32] will do.
[21-Jul-2010 12:21:33] cgibbons: k, cool
[21-Jul-2010 12:21:43] thanks, good lunch reading.. be back in a bit.
[21-Jul-2010 12:21:48] k
[21-Jul-2010 12:23:04] btw, i was testing alienvault w/ snarecore, i reused the ip after taking out alienvault, and snare is reporting to zen.. zen seems to recognize it, is there already integration with snare?
[21-Jul-2010 12:27:17] <17SAA39QW> what are people setting their SMTP send/regex strings to (if changed from default) in Ip Service bit?
[21-Jul-2010 12:29:17] 17SAA39QW is now known as gwb235
[21-Jul-2010 12:29:21] wierd
[21-Jul-2010 12:55:46] ok, now I just have to figure out how this GetTableMap works
[21-Jul-2010 13:16:51] hrm
[21-Jul-2010 13:17:01] I need to know what class tabledata.get() returns
[21-Jul-2010 13:17:08] does anybody know?
[21-Jul-2010 13:32:44] ok this is wierd I can install zenpacks through cli just fine but over the gui on this zenoss install (which I did not do) It keeps saying that it can't find zenpack command
[21-Jul-2010 13:33:33] I see that the ctlscript.sh script calls the setenv.sh one so not sure why it keeps happening
[21-Jul-2010 14:06:35] hi
[21-Jul-2010 14:06:41] can i ask a question?
[21-Jul-2010 14:06:48] always
[21-Jul-2010 14:06:54] thanks
[21-Jul-2010 14:07:08] i've recently installed 3.0 version
[21-Jul-2010 14:07:23] and i traying to use Active Directory as auth a role
[21-Jul-2010 14:07:40] i've followed this http://nathaniel.themccallums.org/2008/03/13/ldapactivedirectory-authorizationauthentication-and-zenoss/
[21-Jul-2010 14:07:54] do you guys know if this is working for 3.0 version?
[21-Jul-2010 14:08:53] I don't personally, but people were talking about getting it working in here recently - not sure where they got to though sorry
[21-Jul-2010 14:09:29] hi: so for the linux low swap threshold the min value is 1024000. Does anyone know off hand how to interpret that number?
[21-Jul-2010 14:42:21] anyone seen this before ? Code: 1 - Msg: General error
[21-Jul-2010 14:49:45] hello all
[21-Jul-2010 14:49:54] how do I monitor tmpfs filesystems? its ramdisk but not really
[21-Jul-2010 14:50:25] i removed ramDisk from the ignored file systems but it still doesn't show up. It use to before I used tmpfs by using a block device that I formatted /dev/ram1 but im using the better tmpfs now
[21-Jul-2010 14:52:30] is there any zenoss plugin to make isc dhcp leases appear in networks?
[21-Jul-2010 15:02:56] QubeZ: did you model the device after making that change?
[21-Jul-2010 15:03:29] wtucker: yup
[21-Jul-2010 15:06:08] QubeZ: what does it show for the type in SNMP? (snmpwalk hrStorageTable)
[21-Jul-2010 15:07:04] wtucker: one sec
[21-Jul-2010 15:07:35] snmpwalk: No securityName specified (Sub-id not found: (top) -> hrStorageTable)
[21-Jul-2010 15:08:46] hmm... what OS/distribution is running on the box you want to monitor?
[21-Jul-2010 15:08:58] wtucker: CentOS 5
[21-Jul-2010 15:09:52] does it show anything if you walk mib-2.system instead of hrStorageTable?
[21-Jul-2010 15:10:13] still gives the security error
[21-Jul-2010 15:10:46] k, try snmpwalk -v2c -c mib-2.system
[21-Jul-2010 15:11:51] New q: will there be a fix for zenactions crashing/hanging from old Maint Windows in 3.x?
[21-Jul-2010 15:12:00] wtucker: k one sec
[21-Jul-2010 15:12:08] QubeZ: k. I won't need the output, just need to know if it works
[21-Jul-2010 15:13:02] wtucker: http://pastebin.com/cB7cbHNa
[21-Jul-2010 15:13:09] wtucker: oops
[21-Jul-2010 15:14:14] QubeZ: is your tmpfs in that list?
[21-Jul-2010 15:14:31] wtucker: nope
[21-Jul-2010 15:14:43] i've remodeled several times
[21-Jul-2010 15:15:12] my nfs is showing up since networkDisk is enabled and so is ramDisk (enabled) but i dont see it
[21-Jul-2010 15:15:20] maybe Zenoss doesn't see tmpfs as ramdisk device
[21-Jul-2010 15:16:33] snmpwalk is part of NET-SNMP, not part of Zenoss, so this tells us that the agent isn't even showing the tmpfs
[21-Jul-2010 15:17:29] one thing I'm not sure of is whether snmpd needs to be restarted to see new devices. I don't think it would, but...
[21-Jul-2010 15:18:31] i restarted it just to check but no change in result
[21-Jul-2010 15:31:00] Hi, is it possible for Zenoss to store and use logical dependency data - like this app uses active directory for authentication, etc?
[21-Jul-2010 15:34:07] q <-- any help?
[21-Jul-2010 15:40:11] QubeZ: hmm... sounds like the snmp daemon is the culprit... not usre if there's a configuration option you can change or if it'd be a source change to make it show that tmpfs
[21-Jul-2010 15:41:04] gschwar: my understanding is that this is a list of bugs that are slated to be worked on for 3.0.1: http://bit.ly/bXpEDT
[21-Jul-2010 15:41:50] wtucker: will look into it
[21-Jul-2010 15:42:22] mr-potter: you could store that sort of thing as custom properties, though I'm not sure what it would take to use them (e.g. if you're wanting to not get an alert for an app being unavailable if AD is unavailable)
[21-Jul-2010 15:42:55] wtucker: thanks...doesn't seem to have anything about zenactions...is there a workaround for maint. windows crashing zenactions without removing old data?
[21-Jul-2010 15:43:16] wtucker: thanks for your help
[21-Jul-2010 15:43:25] gschwar: not sure, I haven't encountered that one
[21-Jul-2010 15:43:45] gschwar: what kind of data need to be removed?
[21-Jul-2010 15:44:21] wtucker: old maintenance windows, and decommissioned servers or else zenactions refuses to process new windows or will crash
[21-Jul-2010 15:44:46] wtucker: wimilar to this: thread/5247
[21-Jul-2010 15:46:01] wtucker: current version is 2.5.2, looking at 3.x but the code from SVN is unchanged for zenactions
[21-Jul-2010 15:47:37] gschwar: probably not, then. I get the impression from that list of tickets that the focus for 3.0.1 is to fix problems that popped up in 3.0.0
[21-Jul-2010 15:47:48] QubeZ: np
[21-Jul-2010 15:48:08] wtucker: ic, i'll submit a bug report then
[21-Jul-2010 15:48:12] thx
[21-Jul-2010 15:49:32] wtucker, thanks
[21-Jul-2010 16:00:27] where do I set the default class for a device
[21-Jul-2010 16:00:37] i.e. the default container for certain make and model of device to go into
[21-Jul-2010 16:04:35] I'm not sure that you can
[21-Jul-2010 16:07:25] strange
[21-Jul-2010 16:10:25] it should be fairly easy to make a script that moves them around after the fact, but there might be a chicken-and-egg problem involved in doing it beforehand - does Zenoss know what make/model it is before it creates the device object?
[21-Jul-2010 16:29:23] mducharme-laptop: there is no out of the box way to do that
[21-Jul-2010 16:34:10] ok
[21-Jul-2010 16:34:18] not a huge deal
[21-Jul-2010 16:37:43] wierd, I can't find the custom tabs from my zenpack in v3
[21-Jul-2010 16:38:36] time to read more doco
[21-Jul-2010 16:50:07] ok this is weird
[21-Jul-2010 16:50:24] I'm trying to do a performance template for interfaces
[21-Jul-2010 16:50:36] it doesn't appear in the list of performance templates
[21-Jul-2010 16:56:58] how the heck do I make this work
[21-Jul-2010 17:03:16] ok this is the weirdest thing
[21-Jul-2010 17:06:28] mducharme-laptop: which version?
[21-Jul-2010 17:06:35] 2.5.2
[21-Jul-2010 17:06:49] the only performance template that shows up under the interfaces is ethernetCmsacd
[21-Jul-2010 17:07:01] which is not correct for this type of device
[21-Jul-2010 17:08:08] I made a performance template that was a copy of that
[21-Jul-2010 17:08:16] and customized it for what I need
[21-Jul-2010 17:08:40] but the interfaces are still using ethernetCmsacd
[21-Jul-2010 17:09:12] look at $ZENHOME/Products/DataCollector/plugins/zenoss/snmpInterfaceMap.py
[21-Jul-2010 17:12:13] correction snmp/InterfaceMap.py
[21-Jul-2010 17:12:30] ahh
[21-Jul-2010 17:12:33] I have a customized interfacemap file
[21-Jul-2010 17:15:40] I might need to add my own interface types
[21-Jul-2010 17:19:33] found a howto
[21-Jul-2010 17:21:28] mducharme-laptop: the copy of the template has to match the type of interface exactly
[21-Jul-2010 17:21:35] check the interface properties for the type
[21-Jul-2010 17:21:40] and name the template exactly that
[21-Jul-2010 17:21:42] case sensitive
[21-Jul-2010 17:21:48] then Zenoss will use that template insteadf
[21-Jul-2010 17:21:51] instead*
[21-Jul-2010 17:22:08] ahh ok
[21-Jul-2010 17:22:28] I think I will have to make my own type of interface
[21-Jul-2010 17:23:03] what do you mean by that?
[21-Jul-2010 17:23:37] the "type" is determined by what the type is defined as in snmp
[21-Jul-2010 17:23:44] it's a standard
[21-Jul-2010 17:24:16] if you look in InterfaceMap.py, there's a type index in the code
[21-Jul-2010 18:24:41] hi, anyone online? :)
[21-Jul-2010 18:25:38] I have few questions... I'm trying to integrate ZenOSS with Apache authentication... are there any docs available for that?
[21-Jul-2010 18:26:05] I found this for Zope: http://www.zope.org/Documentation/Misc/WEBSERVER.txt/view
[21-Jul-2010 18:26:28] It says I only need to set up a Rewrite rule, like this
[21-Jul-2010 18:26:37] RewriteRule ^/Zope/(.*) /usr/local/apache/cgi-bin/Zope.cgi/$1 [e=HTTP_CGI_AUTHORIZATION:%1,t=application/x-httpd-cgi,l]
[21-Jul-2010 18:26:54] But, what should I put instead of /usr/local/apache/cgi-bin/Zope.cgi ?!?!
[21-Jul-2010 18:28:18] jsosic: that's way old
[21-Jul-2010 18:28:41] jsosic: what do you mean "integrate with apache auth" ?
[21-Jul-2010 18:28:59] jsosic: usually you let Zenoss do it's own auth and just proxy to it from apache
[21-Jul-2010 18:29:13] I want to use mod_auth_cas, as we use that as SSO
[21-Jul-2010 18:29:21] and be able to use mod_auth_xradius
[21-Jul-2010 18:29:52] and it seems to me much simpler to do it through Apache rather than mangle with zope
[21-Jul-2010 18:31:11] jsosic: it would be, if it wasn't already done for you ;-)
[21-Jul-2010 18:31:33] mhm... but how? :)
[21-Jul-2010 18:31:47] do I need ZenOSS 3.0 for that?
[21-Jul-2010 18:32:17] jsosic: what i mean is, i have no idea how to use those auth mods… but i do know that Zenoss ships with an auth system out of the box
[21-Jul-2010 18:32:40] so it's not "mangling with zope" because Zenoss already did all the mangling ;-)
[21-Jul-2010 18:33:08] jsosic: i'd read up on alternative auth methods in the zenoss docs, maybe they support it
[21-Jul-2010 18:33:57] OK, I'll check that. in the meantime, why is the upper mentioned method old? it cannot be set up anymore or?
[21-Jul-2010 18:34:21] jsosic: we implemented certificate based auth with apache-ssl as the frontend, we created a new AuthHelper for zope and set it as the active, and had apache pass headers to our custom auth helper
[21-Jul-2010 18:34:41] it's non-trivial and I don't have all the details (I didn't write the helper) btu that might give you some clues as to one way to skin that cat :)
[21-Jul-2010 18:35:11] you can to to http://zenoss.server:8080/manage and use the admin login to explore the acl_users pluggable auth service config
[21-Jul-2010 18:35:46] nice, I'll check it asap
[21-Jul-2010 18:36:43] basically with a bit of python coding you should be able to translate any apache auth headers into something that zenoss considers a valid login - the guys at work even covered off dynamically adding a zenoss user if it didn't exist etc
[21-Jul-2010 18:37:51] I'm a perl monk, so I guess I'll have to dig into python too now...
[21-Jul-2010 18:38:01] you shouldn't find the move too hard
[21-Jul-2010 18:38:26] * Simon4 heads off to bed, good luck :)
[21-Jul-2010 18:38:32] thank you!
[21-Jul-2010 18:46:11] ugh
[21-Jul-2010 18:58:26] hm, I have trouble finding the appropriate ZenOSS documentation page for alternative authentication methods...
[21-Jul-2010 18:58:37] should I look into zope documents or zenoss docs?
[21-Jul-2010 19:14:02] Parabol is now known as Parabola
[21-Jul-2010 20:15:51] yes!
[21-Jul-2010 21:28:37] what linux distribution is the vmware zenoss based off?
[21-Jul-2010 22:34:46] looks like redhat (ie. /etc/sysconfig, chkconfig, etc)
[22-Jul-2010 00:00:30] [disconnected at Thu Jul 22 00:00:30 2010]
[22-Jul-2010 00:00:30] [connected at Thu Jul 22 00:00:30 2010]
[22-Jul-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]
[22-Jul-2010 03:52:42] hmmmm what have I done wrong in 3.0 when I want to scale the graphs the browser wants to dl a file called render
[22-Jul-2010 06:24:53] fragfutter: you there ?
[22-Jul-2010 06:25:05] tehhobbit: for about 90seconds ;)
[22-Jul-2010 06:25:21] how do I remove a dev with dmd ?
[22-Jul-2010 06:26:06] * Simon4 knows the answer if fragfutter needs to run
[22-Jul-2010 06:26:15] would be sweet
[22-Jul-2010 06:26:16] device.manage_delete or something
[22-Jul-2010 06:26:42] running into some problems with reindex from, 2.5.2 => 3.0
[22-Jul-2010 06:27:31] tehhobbit: dev = find('devicename') then dev.deleteDevice()
[22-Jul-2010 06:29:22] http://mrdoob.com ... fap
[22-Jul-2010 06:30:33] Simon4: can I do a wildcard seach ?
[22-Jul-2010 06:30:56] tehhobbit: not with taht function, btu you could walk the device tree with something like:
[22-Jul-2010 06:31:14] for dev in dmd.Devices.getSubDevices():
[22-Jul-2010 06:31:20] if "fred" in dev:
[22-Jul-2010 06:31:25] dev.deleteDevice()
[22-Jul-2010 06:31:26] etc
[22-Jul-2010 06:31:30] gah
[22-Jul-2010 06:31:37] second line = if "fred" in dev.id
[22-Jul-2010 06:31:40] http://pastie.org/1055143
[22-Jul-2010 06:31:50] seem to have fucked up things pretty badly
[22-Jul-2010 06:32:01] hey that's fairly cool
[22-Jul-2010 06:32:52] any ideas what to do `
[22-Jul-2010 06:33:32] try:
[22-Jul-2010 06:33:41] for dev in dmd.Devices.getSubDevices():
[22-Jul-2010 06:33:55] if "ibm3584.mgmt.slba.se" in dev.id:
[22-Jul-2010 06:34:08] print dev.id
[22-Jul-2010 06:34:20] and if that works, you can try deleting it from there (presuming that thta's what you want to do)
[22-Jul-2010 06:34:50] 2010-07-22 13:34:27 WARNING ZenModel.DeviceOrganizer Stale deviceSearch record: /zport/dmd/Devices/Storage/IBM/TapeLibrary/devices/193.10.249.110
[22-Jul-2010 06:35:17] a ha
[22-Jul-2010 06:35:25] catalog awesome
[22-Jul-2010 06:35:35] its one of the zenpacks that broke in 3.0
[22-Jul-2010 06:35:47] so zenpack is removed
[22-Jul-2010 06:35:58] and you've run reindex() ?
[22-Jul-2010 06:36:23] fails but sec and I'll give you stacktrace
[22-Jul-2010 06:39:51] http://pastie.org/1055156
[22-Jul-2010 06:40:19] thats what triggered me to clean out zenpacks in the first place
[22-Jul-2010 06:42:38] okay
[22-Jul-2010 06:42:40] http://pastie.org/1055159
[22-Jul-2010 06:42:47] stick that in a file, then run zendmd
[22-Jul-2010 06:42:57] see if you get anything coming back as ****BAD***
[22-Jul-2010 06:43:05] don't uncomment the two commented lines quite yet ;)
[22-Jul-2010 06:43:34] * Simon4 isn't sure if this will work on 3 or not, so better to be safe first run
[22-Jul-2010 06:45:23] cant get that to fail but it doesnt show all devs either
[22-Jul-2010 06:45:41] does it print any lines with ****BAD**** in front of them?
[22-Jul-2010 06:47:21] can I rescue the data from reindex (print info of the object it was working with
[22-Jul-2010 06:47:27] something like
[22-Jul-2010 06:47:30] try:
[22-Jul-2010 06:47:36] reindex()
[22-Jul-2010 06:47:42] except:
[22-Jul-2010 06:47:48] print dev
[22-Jul-2010 06:48:06] no, you would need to go find the reindex code itself I think
[22-Jul-2010 06:48:29] o dear
[22-Jul-2010 06:49:01] what is the device 193.10.249.110 ?
[22-Jul-2010 06:49:40] a taperobot
[22-Jul-2010 06:50:04] is it related to ibm3584.mgmt.slba.se ?
[22-Jul-2010 06:50:11] yes
[22-Jul-2010 06:50:18] hmm, okay
[22-Jul-2010 06:50:20] both are the same type of devs
[22-Jul-2010 06:50:20] * Simon4 thnks
[22-Jul-2010 06:50:32] ah right, so same type, but different physical devs
[22-Jul-2010 06:51:18] yes
[22-Jul-2010 06:57:38] what I really want to find out is at what dev the reindex fails
[22-Jul-2010 06:57:47] yeah
[22-Jul-2010 06:57:51] I'm just playing with the code a lite
[22-Jul-2010 06:57:53] little
[22-Jul-2010 06:58:20] hmm, this kinda works
[22-Jul-2010 06:58:37] in /opt/zenoss/Products/Zuul/catalog/global_catalog.py
[22-Jul-2010 06:59:09] there's a function called searchExcerpt
[22-Jul-2010 06:59:13] line 219ish
[22-Jul-2010 06:59:58] I hurled this in:
[22-Jul-2010 06:59:59] def searchExcerpt(self):
[22-Jul-2010 06:59:59] o = self._context
[22-Jul-2010 06:59:59] print o
[22-Jul-2010 07:00:09] then ran up zendmd and ran a reindex()
[22-Jul-2010 07:00:18] would be interesting to see what the last line of output is
[22-Jul-2010 07:00:31] (usual disclaimer about backing files up, etc :) )
[22-Jul-2010 07:04:50]
[22-Jul-2010 07:05:03] is the last item that gets indexed
[22-Jul-2010 07:05:53] * Simon4 wonders what one of those is
[22-Jul-2010 07:06:08] and I take it it then stack traces right below that?
[22-Jul-2010 07:06:22] yes
[22-Jul-2010 07:06:35] its a part of the hpmon zenpack
[22-Jul-2010 07:06:41] for insight manager
[22-Jul-2010 07:07:05] hmm
[22-Jul-2010 07:07:07] hokay
[22-Jul-2010 07:07:15] remove the print line
[22-Jul-2010 07:07:19] time for another fun code snippet to run
[22-Jul-2010 07:07:21] * Simon4 pastie's
[22-Jul-2010 07:07:52] http://pastie.org/1055188
[22-Jul-2010 07:08:02] similar to the last, but does components, which is wht that looks like
[22-Jul-2010 07:10:40] I was correct its the ibm tape libraries that f*** things up
[22-Jul-2010 07:11:38] so can I get direct access to the zope db and just remove them since they seem to be in some sort of catch 22
[22-Jul-2010 07:11:59] was that after running the latest script I pasted?
[22-Jul-2010 07:12:22] get alot of bad from them
[22-Jul-2010 07:16:59] hokay
[22-Jul-2010 07:17:03] so, you can clean up with that script
[22-Jul-2010 07:17:11] but I would recommend taking a zenbackup first
[22-Jul-2010 07:17:21] seemed to work with catalog.uncatalog_object(o.getPath())
[22-Jul-2010 07:17:28] once you have a backup, just uncomment out the two commented lines
[22-Jul-2010 07:17:33] trying reindex() again
[22-Jul-2010 07:17:37] and run the script a couple of times
[22-Jul-2010 07:18:14] heh, you're a step ahead of me :)
[22-Jul-2010 07:18:47] dammit then next thing will be to try and figure out how to get the zenpack 3.0 compatible
[22-Jul-2010 07:19:17] those to pieces of hardware are kind important to us
[22-Jul-2010 07:20:26] for sure
[22-Jul-2010 07:20:41] nopp reindex still fails
[22-Jul-2010 07:21:14] crap, same place?
[22-Jul-2010 07:23:34] oooh remeber the line we played with changed it to print o.getParentDeviceName
[22-Jul-2010 07:23:56] so hopefully I'll soon find what dev the reindex fails at
[22-Jul-2010 07:24:47] ah cool
[22-Jul-2010 07:25:14] or hmm maybe the device before since it fails :(
[22-Jul-2010 07:25:28] yeah, I was going from the "str is not callable" when I tried to print o
[22-Jul-2010 07:25:35] thinking that maybe it was a string, vs an object
[22-Jul-2010 07:28:07] wrapped it in try / except and printing self to see what that might give
[22-Jul-2010 07:30:48] was all I got
[22-Jul-2010 07:31:45] well might atleast be on the right track
[22-Jul-2010 07:35:32] ok now I have the object name that fails, can I somehow grab zenpacks and devices that has that object ?
[22-Jul-2010 07:39:48] by the looks of it its a component in the hp mon zenpack, which kinda sucks
[22-Jul-2010 07:40:36] cpqSsChassis is an abstraction of a HP Storage System Chassis.
[22-Jul-2010 07:40:38] you can try running that component cleaner snippet more times, and see if it gets picked up
[22-Jul-2010 07:40:54] that or you can get a little trickier
[22-Jul-2010 07:41:32] something like:
[22-Jul-2010 07:41:43] cat = dmd.Devices.componentSearch()
[22-Jul-2010 07:41:50] for o in catalog:
[22-Jul-2010 07:42:06] if "cpqSsChassis" in o.getPath:
[22-Jul-2010 07:42:13] well there is a svn tag right above the comment that says that it hasnt been updated since 1.1 althou they bumped the zenpack to 2.0 in order for it to be 3.0 compatible
[22-Jul-2010 07:42:23] catalog.uncatalog_object(o.getPath())
[22-Jul-2010 07:42:32] which may well just remove the lot from the catalog
[22-Jul-2010 07:42:50] I would be backing up before playing these games though, but Iv'e had success doing similar for interfaces and other htings
[22-Jul-2010 07:43:06] then a reindex rebuilds the catalogues from what is really there
[22-Jul-2010 07:43:14] well I have my 2.5.2 backup
[22-Jul-2010 07:43:37] you can run zenbackup --no-eventsdb --no-perfdata and it's really quick
[22-Jul-2010 07:44:01] that picks up the zopedb and zenpack config, which is enough to get you back if you accidentally remove too much from the db :)
[22-Jul-2010 07:44:40] http://pastie.org/1055224
[22-Jul-2010 07:45:09] this is what I did to get everything save for the broken components indexed
[22-Jul-2010 07:45:14] a ha
[22-Jul-2010 07:45:26] what did the "print o" come back with, was that the cpqSsChassis thing?
[22-Jul-2010 07:45:32] yes
[22-Jul-2010 07:47:11] just onemore thing how can I force remove a device
[22-Jul-2010 07:47:26] my 2 robots are now orphaned after removing the zenpack
[22-Jul-2010 07:47:42] so need to force remove them (gui cant do that
[22-Jul-2010 07:48:11] I'm not too sure, I have a device I need to remove also, I would think that possibly you could try removing the device relation from the organiser object
[22-Jul-2010 07:48:30] ok
[22-Jul-2010 07:50:26] how do I do that ?
[22-Jul-2010 07:51:11] heh, that bit I don't know
[22-Jul-2010 07:59:50] but how hard can it be to remove a device
[22-Jul-2010 08:05:25] re
[22-Jul-2010 08:05:26] wish they had used sql so I just could have removed it from the db
[22-Jul-2010 08:05:55] * tehhobbit wants a --force of some sort
[22-Jul-2010 08:06:45] fragfutter: you wouldnt happen to have any idea what to do
[22-Jul-2010 08:07:46] tehhobbit: i would need to read the irc log. can i have a summary?
[22-Jul-2010 08:08:19] fragfutter: I want to forcefully remove 2 orphaned devices
[22-Jul-2010 08:08:40] (zenpack and the organizer they were under are gone but they arent)
[22-Jul-2010 08:09:02] you can find them in zendmd?
[22-Jul-2010 08:09:26] yes but get a fail due to the organizer missing
[22-Jul-2010 08:09:38] when finding them you get a fail?
[22-Jul-2010 08:10:11] and the gui fails to show the device part so cant use that either
[22-Jul-2010 08:11:50] ok, you had a device, now the path/classification is gone. How do you still find the device?
[22-Jul-2010 08:11:59] * Simon4 has an idea
[22-Jul-2010 08:12:02] using zendmd, you type what?
[22-Jul-2010 08:12:03] just trying something at this end
[22-Jul-2010 08:12:14] you are using find("dasdasd") ?
[22-Jul-2010 08:12:22] yes
[22-Jul-2010 08:12:29] and what does it do/say?
[22-Jul-2010 08:12:34] sec
[22-Jul-2010 08:13:02] http://pastie.org/1055143
[22-Jul-2010 08:13:18] find is doing a catalog query, i assume your catalog is out of sync with the real objects.
[22-Jul-2010 08:14:43] fragfutter: that device and some other stuff is stopping the zenmigrate so stuck in the middle yes
[22-Jul-2010 08:18:49] lets work directly on the catalog
[22-Jul-2010 08:19:06] 'ibm3584.mgmt.slba.se' in [x.id for x in dmd.Devices.deviceSearch()]
[22-Jul-2010 08:19:22] checks if the device can still be found in the catalog
[22-Jul-2010 08:20:20] no it isnt since the indexing fails when it comes to that dev
[22-Jul-2010 08:20:33] can I from dmd add an organicer ?
[22-Jul-2010 08:20:38] tehhobbit: did you run this code and what did you get?
[22-Jul-2010 08:20:47] >>> 'ibm3584.mgmt.slba.se' in [x.id for x in dmd.Devices.deviceSearch()]
[22-Jul-2010 08:20:48] False
[22-Jul-2010 08:21:22] but your find('...') found it
[22-Jul-2010 08:21:27] yes
[22-Jul-2010 08:21:37] fOrsberg is now known as forsberg
[22-Jul-2010 08:21:44] uuh now it doesnt
[22-Jul-2010 08:21:51] and you reconnected your dmd after updating the index?
[22-Jul-2010 08:22:24] fragfutter: I cant update the index since it fails halfway when it tries to index the dev
[22-Jul-2010 08:23:52] tehhobbit: find('') is using the catalog dmd.Devices._getCatalog() which is /zport/dmd/Devices/deviceSearch. If find is finding the Device, but a direct Query of the device is not finding it, i'm mystified
[22-Jul-2010 08:24:06] (direct Query of the catalog)
[22-Jul-2010 08:24:33] but I just want the device gone, out , deleted etc
[22-Jul-2010 08:24:47] tehhobbit: the device IS gone. there are fragments in the catalog
[22-Jul-2010 08:25:25] http://pastie.org/1055283
[22-Jul-2010 08:25:32] this is where reindex fails
[22-Jul-2010 08:26:10] the catalog isnt the same as the db is it ?
[22-Jul-2010 08:26:24] the catalog is an object in the zodb
[22-Jul-2010 08:27:16] well I cant reindex till its completly gone
[22-Jul-2010 08:27:34] ok. do you know the zope management interface?
[22-Jul-2010 08:27:46] zendmd ?
[22-Jul-2010 08:27:50] no.
[22-Jul-2010 08:28:02] http://example.com:8080/zport/manage
[22-Jul-2010 08:28:39] aaaah
[22-Jul-2010 08:28:56] we will first make sure your device is no longer in the zodb
[22-Jul-2010 08:28:57] and now you want me to go in under the dmd/devices ?
[22-Jul-2010 08:29:14] which thows and exception
[22-Jul-2010 08:29:19] horizontal menu right end, click tab find
[22-Jul-2010 08:29:52] find objects of type "All types"; with ids "ibm...something". Click Find button. wait. it won't find your device
[22-Jul-2010 08:34:04] cant I use readd the organizer from zendmd ?
[22-Jul-2010 08:34:36] tehhobbit: first i need to understand the problem.
[22-Jul-2010 08:34:51] tehhobbit: my assumption is still, a f** catalog.
[22-Jul-2010 08:34:59] the device is under a organizer that is no long there
[22-Jul-2010 08:35:23] tehhobbit: yes. we had that. But i assume there is still an entry in the catalog.
[22-Jul-2010 08:35:29] tehhobbit: i want to find this entry
[22-Jul-2010 08:35:49] tehhobbit: so first we make sure that there is no object with the ibm id you named
[22-Jul-2010 08:36:01] tehhobbit: then we go to the catalog and see if there is an entry
[22-Jul-2010 08:36:11] fragfutter: no since it made reindex fail it was never in the catalog ( I tried removing the zenpack to see if that helped reindex)
[22-Jul-2010 08:36:29] tehhobbit: can we please check?
[22-Jul-2010 08:36:49] tehhobbit: did you execute the find in the zope management interface, and what did it return?
[22-Jul-2010 08:38:07] just hangs
[22-Jul-2010 08:38:26] oh done and no result
[22-Jul-2010 08:38:32] good.
[22-Jul-2010 08:38:59] now we go to the index. http://example.com:8080/zport/dmd/Devices/deviceSearch/manage_catalogView
[22-Jul-2010 08:40:38] in the path filter you can type the prefix of your ibm device like /zport/dmd/Devices/Tapes/something
[22-Jul-2010 08:40:44] then set-filter
[22-Jul-2010 08:42:02] can you now find your device in the list?
[22-Jul-2010 08:43:49] yes
[22-Jul-2010 08:43:57] so it is still in the catalog...
[22-Jul-2010 08:44:21] click checkbox to the left of it, click remove
[22-Jul-2010 08:44:38] fails
[22-Jul-2010 08:44:43] get a stacktrace
[22-Jul-2010 08:44:58] *hmpf*
[22-Jul-2010 08:45:15] see why I'm getting frustrated ?
[22-Jul-2010 08:45:49] but why not just put the organizer back
[22-Jul-2010 08:46:17] can i see the stacktrace?
[22-Jul-2010 08:47:20] Type:
[22-Jul-2010 08:47:21] Value: _index_parents
[22-Jul-2010 08:47:27] is all I got this time
[22-Jul-2010 08:48:13] i assume you have a backup?
[22-Jul-2010 08:48:14] 8080/zport/dmd/Devices/deviceSearch/manage_catalogAdvanced Update-Catalog
[22-Jul-2010 08:49:46] http://pastie.org/1055319 from event log
[22-Jul-2010 08:49:56] and dev is still there after update
[22-Jul-2010 08:50:25] yes. the catalog index is starting a transaction and doing rollback on error
[22-Jul-2010 08:51:25] I need a smoke before I kill somethinh
[22-Jul-2010 08:51:29] brb
[22-Jul-2010 08:52:11] does 8080/zport/dmd/Devices/deviceSearch/manage_catalogIndexes have some additional index?
[22-Jul-2010 08:58:57] dont know which ones are suppose to be there etc
[22-Jul-2010 09:00:10] http://pastie.org/1055339
[22-Jul-2010 09:00:35] forsberg is now known as fOrsberg
[22-Jul-2010 09:00:39] but I still dont get why I cant just readd the organizer
[22-Jul-2010 09:00:59] (from dmd)
[22-Jul-2010 09:01:10] it should solve the problem
[22-Jul-2010 09:01:18] tehhobbit: you can try.
[22-Jul-2010 09:01:30] how do I do that ?
[22-Jul-2010 09:02:04] do it through zope
[22-Jul-2010 09:02:12] 8080/zport/dmd/Devices
[22-Jul-2010 09:02:24] /manage_main
[22-Jul-2010 09:02:38] dropdownbox on the right, DeviceClass, Add
[22-Jul-2010 09:04:37] problem is its a sub organizer to storage and if I click the storage folder it fails
[22-Jul-2010 09:05:25] so I need some way to force it
[22-Jul-2010 09:05:44] so the storage folder is also f***
[22-Jul-2010 09:06:00] yes and I cant remove that either
[22-Jul-2010 09:07:06] sounds bad. what about restoring a point before you tried to migrate?
[22-Jul-2010 09:09:02] yeah well suppose that could work
[22-Jul-2010 09:09:10] but a hell lot of trouble
[22-Jul-2010 09:09:39] is there no way to access the raw db and just remove the stuff of to create the folder from dmd
[22-Jul-2010 09:10:10] tehhobbit: there is no rawdb. zodb is an object database. it does not contain tables and rows.
[22-Jul-2010 09:10:58] and if I try forcing creating deleting the folder
[22-Jul-2010 09:10:59] ?
[22-Jul-2010 09:12:48] are there any hacks to get the status of Nagios plugins to show up in the Device Status list? ie. have one of those handy green dots show up when check_http says a webserver is functional
[22-Jul-2010 09:12:57] the developer manual is somewhat vague about this
[22-Jul-2010 09:13:10] fragfutter: just cant see how it would be impossible to remove data
[22-Jul-2010 09:13:56] tehhobbit: same as in a relational db. if you have a trigger and that one fails, it won't let you remove the data
[22-Jul-2010 09:14:06] I know the dev has a key called IBM
[22-Jul-2010 09:14:24] dammit
[22-Jul-2010 09:14:50] this is gonna be alot of extra work
[22-Jul-2010 09:14:52] tehhobbit: so in zope you take the object, wake it up and tell him "delete yourself" and the object says "wait, i'm important, don't delete me"
[22-Jul-2010 09:15:33] --force-I-dont-care switch ftw
[22-Jul-2010 09:16:09] tehhobbit: and run around and cry if you get lot's of other obsucre errors in a week.
[22-Jul-2010 09:16:14] but how can the device not be deleteabel
[22-Jul-2010 09:16:40] yeah well
[22-Jul-2010 09:16:53] uninstall , back to 2.5 restore *sight*
[22-Jul-2010 09:18:45] well gives me some time to fix other stuff too
[22-Jul-2010 09:18:53] pretty shade install to start with
[22-Jul-2010 09:19:10] gone from testing to "hey lets make it prod" abit to fast
[22-Jul-2010 09:19:19] better to do it right this time
[22-Jul-2010 09:19:24] tehhobbit: i would assume a zenpack broke it.
[22-Jul-2010 09:19:27] so not all bad
[22-Jul-2010 09:19:55] fragfutter: been playing around installing / uninstalling lots of stuff past month
[22-Jul-2010 09:20:16] tehhobbit: and some packs did not uninstall cleanly
[22-Jul-2010 09:20:34] the ibm one is broken with 3.0
[22-Jul-2010 09:21:59] gonna reinstall and fix some other stuff before readding the devices I think
[22-Jul-2010 09:23:02] btw might as while I'm at it , can I set a default collector for a certain network ?
[22-Jul-2010 09:24:17] running single collector, no idea
[22-Jul-2010 09:24:49] tehhobbit: for when you discover devices? or just when you add them manually?
[22-Jul-2010 09:25:09] discover
[22-Jul-2010 09:25:33] so you could run zendisc on the command line via cron or manually
[22-Jul-2010 09:25:33] fragfutter: been thinking about running a single one too
[22-Jul-2010 09:25:48] and if you then split the commands you can set the collector on the commandline for a specific network
[22-Jul-2010 09:26:16] because dual collectors with core version is alot of work
[22-Jul-2010 09:30:50] tehhobbit: I hadn't even though of that, setting up 1 collector is easy, but setting up 2 remote wouldn't be very simple to do
[22-Jul-2010 09:30:55] hmmm, actually
[22-Jul-2010 09:31:00] nah, it should be the same
[22-Jul-2010 09:31:02] now that I think of it
[22-Jul-2010 09:31:38] it's no harder to set up >1
[22-Jul-2010 09:31:40] oh I just had 1 + main one
[22-Jul-2010 09:32:04] but cause me alot of grief anyways
[22-Jul-2010 09:34:28] yeh well, I don't think the guide for doing that in core is exactly perfect yet
[22-Jul-2010 09:34:29] everything from rpm packages wanting to run a ton of scripts , zenpacks being outof sync etc
[22-Jul-2010 09:34:51] I made my collector tool ZenPack based on the community guide, but I haven't had a chance to extensively test it to improve it
[22-Jul-2010 09:34:52] rmatte: no I know and you get it for enterprice
[22-Jul-2010 09:35:02] for one, I think the URL put in to the collectors section is incorrect
[22-Jul-2010 09:35:21] which egor pointed out to me at one point
[22-Jul-2010 09:35:26] haven't had a chance to test though
[22-Jul-2010 09:35:40] the most troublesome was that the standalone collector can see nets that the main one cant and when adding devs its done from the main server
[22-Jul-2010 09:36:17] tehhobbit: it shouldn't be done from the main server if you specify the alternate collector
[22-Jul-2010 09:36:20] if it is, then that's a bug
[22-Jul-2010 09:36:45] you need to specifically tell it to add the device using the other collector, and not just move it over to it after it's added
[22-Jul-2010 09:40:40] I think I did that but still the zendisc ran on the main one
[22-Jul-2010 09:40:58] * Simon4 had that problem also
[22-Jul-2010 09:41:16] I ended up just shelling into the collector and running zenmodeler manually on devices added on that collector
[22-Jul-2010 09:41:19] that worked
[22-Jul-2010 09:41:38] yeah same way I did it
[22-Jul-2010 09:41:54] * Simon4 hasn't tried that in enterprise yet actually
[22-Jul-2010 09:41:55] * Simon4 tries
[22-Jul-2010 09:42:10] would be a tad to expensive for us
[22-Jul-2010 09:42:44] my zenoss hardware arrived, hurrah!
[22-Jul-2010 09:42:51] woohoo
[22-Jul-2010 09:42:55] i wish 3.0 enterprise was out, because i don't want to install 2.5 and then have to upgrade immediately :)
[22-Jul-2010 09:43:19] think it will be some time until it comes out
[22-Jul-2010 09:43:41] I tried upgrading to the rc from the 2.5 test install i had and it didn't seem to work that great
[22-Jul-2010 09:43:41] * Simon4 is hoping that 3.0 enterprise comes out as 3.0.1 after all the community niggles are sorted
[22-Jul-2010 09:43:49] modelling servers didn't really work that well after
[22-Jul-2010 09:43:59] so maybe it's for the best that i'll use something older and established :)
[22-Jul-2010 09:44:54] Simon4: 3.0.1 will be Enterprise
[22-Jul-2010 09:45:51] ashp: the upgrade does have it's issues, but nothing that can't be overcome
[22-Jul-2010 09:46:14] rmatte: Sure, I just had no idea what I was doing, I just followed the instructions as best I could..
[22-Jul-2010 09:46:21] I'm sure by 3.0.1 it'll handle the upgrade just fine :)
[22-Jul-2010 09:46:27] if not i'll just start again
[22-Jul-2010 09:46:29] yeh, there's stuff outside the instructions
[22-Jul-2010 09:46:36] but that comes with experience
[22-Jul-2010 09:46:56] nah, I doubt 3.0.1 will handle the upgrade any better
[22-Jul-2010 09:46:59] the problem is ZenPacks
[22-Jul-2010 09:47:09] 3.0 uses Python 2.6
[22-Jul-2010 09:47:16] and all the ZenPacks are rigged to work with 2.4
[22-Jul-2010 09:47:25] some work fine after the upgrade, others need to be reinstalled
[22-Jul-2010 09:47:46] and when reinstalling them, the -py2.4 needs to be removed from the ZenPack filename
[22-Jul-2010 09:48:33] In my opinion, a migration script should have been created to smooth the ZenPack migration from version to version
[22-Jul-2010 09:48:36] but that was never done
[22-Jul-2010 09:49:32] I'm waiting on 3.0.1 to come out before I start creating a final upgrade guide for my servers
[22-Jul-2010 09:49:50] actually, I have trac tickets to log today, been putting it off
[22-Jul-2010 09:50:35] so basically it's going to be a nightmare to upgrade :)
[22-Jul-2010 09:50:44] ah well, it's never that easy
[22-Jul-2010 09:50:53] well, not a "nightmare", but complex for me at least
[22-Jul-2010 09:51:04] I have 14 Zenoss servers with 26 ZenPacks installed on each of them
[22-Jul-2010 09:51:18] Upgrades are always a lot of fun
[22-Jul-2010 09:51:18] yeah, i think you're in a much harder situation
[22-Jul-2010 09:51:35] the upgrade from 2.4 to 2.5 was no picnic either
[22-Jul-2010 09:52:10] i was hoping the server delay would be enough that we'd just dive right into 3.0 to save me the effort of tackling an upgrade early on :)
[22-Jul-2010 09:52:24] I'll end up just using the 2.5 install as more learning
[22-Jul-2010 09:53:23] it's up to you, I just don't personally see 3.0 as sufficient for my needs quite yet
[22-Jul-2010 09:53:33] I ordered a failover box in the hardware this time around, so upgrading will involve doing it alongside live, then failing people over to the "new" box I think
[22-Jul-2010 09:53:39] in-place sounds like it'll be tricky
[22-Jul-2010 09:53:41] especially since interface descriptions are no longer listed in the interface section
[22-Jul-2010 09:53:46] which is a major problem for me
[22-Jul-2010 09:54:15] but I guess that's what happens when you decide to have a side menu take up a quarter of the screen
[22-Jul-2010 09:54:35] no room for anything
[22-Jul-2010 09:55:06] yeah, the designers must all have desktops and 24-30" monitors
[22-Jul-2010 09:55:13] the columns should be configurable, and then there should be rollover popups which display detailed info about a given interface
[22-Jul-2010 09:55:15] on a laptop screen the new UI is incredibly cramped
[22-Jul-2010 09:55:22] that would be ideal for the new interface
[22-Jul-2010 09:55:46] I don't personally care to see IP and Mac Address info as columns, those would be fine as rollovers
[22-Jul-2010 09:55:55] but I absolutely need to see interface description as a column
[22-Jul-2010 09:56:22] * Simon4 nods
[22-Jul-2010 09:56:51] I'll be logging a trac about that shortly
[22-Jul-2010 09:56:56] since it's a regression in my opinion
[22-Jul-2010 10:02:49] from earlier - the first 3.x enterprise release will be 3.0.1 and might be as early as next week
[22-Jul-2010 10:03:10] cool
[22-Jul-2010 10:03:52] cgibbons: i did not manage to get a clean ComponentGrid with url links. If i have links in the grid, the details page fails.
[22-Jul-2010 10:05:18] it can be tricky
[22-Jul-2010 10:07:35] cgibbons: i may just stall in that case and wait
[22-Jul-2010 10:07:58] whatever you do, read the release notes before you upgrade
[22-Jul-2010 10:08:13] that apparently has been the source of a great number of issues in the 3.0 core version so far
[22-Jul-2010 10:09:29] cgibbons: http://pastie.org/1055470
[22-Jul-2010 10:09:46] cgibbons: would render it correct in the grid, but gives exception in the details page
[22-Jul-2010 10:10:59] you didn't make it a schema.Entity because it might there might not be an actual component found if it's an external, non-zenoss managed, thing on the other side of the link?
[22-Jul-2010 10:11:26] cgibbons: yes.
[22-Jul-2010 10:11:54] cgibbons: and i wouldn't know how to deal with a schema.Entity anyway ;)
[22-Jul-2010 10:12:21] When I've had that issue I've just had a blank column, but in your case you want to display something... Ian might have a suggestion for that. Ian: short-summary is that a property of this info object would normally be an entity, but sometimes it can just be a text value that needs to be displayed.
[22-Jul-2010 10:13:16] You can't just send it through? I thought stuff was in there to just return the string
[22-Jul-2010 10:13:22] schema.Entity would be easy; you'd just return the component in the property, use the @info descriptor around the property function to get the Info object created automatically, and use Zenoss.render.default_uid_renderer in the grid column. but that doesn't handle (i think) the scenario if it's not there.
[22-Jul-2010 10:13:25] a moment, checking the code
[22-Jul-2010 10:13:57] it might well be- so the .info object that finds the other entity would just return "foobar" instead of the other object and then the right things would happen? hmm that sounds right
[22-Jul-2010 10:14:10] most of that rendering code seems to be pretty smart about making sure it's got a uid or not
[22-Jul-2010 10:14:14] nope, you're right, it doesn't do the check. It should test for pathiness.
[22-Jul-2010 10:14:24] Shoot, that's by far the easiest solution.
[22-Jul-2010 10:14:49] Monkeypatch it in your pack
[22-Jul-2010 10:15:33] ashp: i think support has a document out for pre-release information for 3.0.x enterprise upgrades.
[22-Jul-2010 10:16:33] iancmcc: i'm lost.
[22-Jul-2010 10:17:37] wait. wt...
[22-Jul-2010 10:17:43] it's complicated unfortunately. ian & i have a lot of insider knowledge of this code so it's hard to verbalize it all :)
[22-Jul-2010 10:17:45] why is this working now.
[22-Jul-2010 10:17:58] magic
[22-Jul-2010 10:18:03] oh, cool, that'd be nice
[22-Jul-2010 10:18:12] fragfutter: one sec, I'm writing the modified method
[22-Jul-2010 10:20:16] fragfutter: You've got some custom javascript in your pack, I presume?
[22-Jul-2010 10:20:22] if so, throw this at the top of it: http://pastebin.ca/1906050
[22-Jul-2010 10:20:53] and by that I meant http://pastebin.ca/1906051
[22-Jul-2010 10:24:40] iancmcc: it works without it. if i use an info object with decorated functions it works. if i use ProxyProperty it doesn't
[22-Jul-2010 10:25:19] fragfutter: Cool, because monkeypatching is a dark path that has no turning
[22-Jul-2010 10:25:33] yesterday i used ProxyProperty and that failed
[22-Jul-2010 10:27:56] iancmcc: if i want to specify that the component has no graphs, but instead should display details by default?
[22-Jul-2010 10:28:23] the details thing is a bit tricky. Don't know we've done that yet.
[22-Jul-2010 10:28:38] Graphs are easy, though.
[22-Jul-2010 10:29:03] iancmcc: specifying that it has no graphs, so it would by default display events then?
[22-Jul-2010 10:29:17] iancmcc: like IpRouteEntry
[22-Jul-2010 10:29:56] presumably this is a monitored component?
[22-Jul-2010 10:30:07] IpRouteEntry isn't, so that's what's being used to check
[22-Jul-2010 10:30:11] iancmcc: it's actualy only something picked up during model.
[22-Jul-2010 10:30:38] if you set monitor=False on the info object graphs won't show up
[22-Jul-2010 10:30:45] IF that's something you can do.
[22-Jul-2010 10:30:50] iancmcc: yes i can
[22-Jul-2010 10:31:48] also make sure that 'monitor' is a field you're asking for in the ComponentGridPanel config, even if it isn't displayed in a column
[22-Jul-2010 10:32:11] iancmcc: yes. no idea why, but cgibbons told me so.
[22-Jul-2010 10:32:32] actually no, you don't have to. IpRouteEntry doesn't, which is presumably why it works
[22-Jul-2010 10:32:46] I bet if you take that out the graphs will disappear and you won't have to modify the info object at all
[22-Jul-2010 10:33:08] iancmcc: modified the info, restarted zenoss and it now works. it is only rather slow.
[22-Jul-2010 10:33:30] iancmcc: so if you have time, i would ask you some more questions.
[22-Jul-2010 10:36:22] a little bit
[22-Jul-2010 10:37:31] iancmcc: i model lldp links. during modelling i pull data with snmp. i know the local interface description, the remote management ip and the remote interface description. Works find, can be displayed
[22-Jul-2010 10:38:14] iancmcc: next step i wanted to make links of them. this is done during display.
[22-Jul-2010 10:38:39] you want to link to the interface/ip address?
[22-Jul-2010 10:39:09] for the interface to the interface and for the ip to the device owning it.
[22-Jul-2010 10:39:17] i have something that works
[22-Jul-2010 10:39:26] i simply want to ask it makes sense ;)
[22-Jul-2010 10:39:30] bring it
[22-Jul-2010 10:40:24] iancmcc: i take the remote management ip, search it ip=dmd.Networks.findIp("1.2.3.4"). If i find it, i use ip.getDeviceLink()
[22-Jul-2010 10:40:36] else display the string i found.
[22-Jul-2010 10:41:33] the interface is trickier.
[22-Jul-2010 10:41:45] ip.getDeviceLink() is old but will do for now
[22-Jul-2010 10:42:02] iancmcc: what should i use instead? ip.device().urlLink()?
[22-Jul-2010 10:42:26] no, just send the device's uid and use schema.Entity
[22-Jul-2010 10:43:02] wherever possible don't allow html to be built server-side. It's terrible and we're weeding it out when we can
[22-Jul-2010 10:43:22] nonetheless yours is functional, but I can't guarantee that api being supported
[22-Jul-2010 10:43:38] iancmcc: ok. so i either send a string (if i don't find it) or i send a uid.
[22-Jul-2010 10:44:05] yeah
[22-Jul-2010 10:45:10] iancmcc: i didn't understand the schema.Entity thing. so how would i do that?
[22-Jul-2010 10:45:24] device has no attribute uid
[22-Jul-2010 10:45:29] actually you don't have to use schema.Entity.
[22-Jul-2010 10:45:36] right, make it an info object
[22-Jul-2010 10:45:44] IInfo(device).uid
[22-Jul-2010 10:45:53] from Products.Zuul.interfaces import IInfo
[22-Jul-2010 10:46:57] so, sorry, you want this to display in your grid?
[22-Jul-2010 10:46:59] or in the details?
[22-Jul-2010 10:47:06] iancmcc: both
[22-Jul-2010 10:47:12] ah
[22-Jul-2010 10:47:21] so yeah, use schema.Entity in the interface
[22-Jul-2010 10:47:46] just like the previous field we were discussing. If you're returning a uid, it'll make it a link.
[22-Jul-2010 10:48:21] iancmcc: i'll write some pastes.
[22-Jul-2010 10:48:45] Gotta go to a meeting; I'll check it out after.
[22-Jul-2010 10:48:52] http://pastie.org/1055470
[22-Jul-2010 10:58:44] <|rt|> when you install the stack installer what's the default password to access the database...I need to restore a backup and can't figure out what the username and password to use to restore it
[22-Jul-2010 11:02:18] morning
[22-Jul-2010 11:02:39] question for you guys - if I install the mib for the device, do I still need to use the OIDs everywhere?
[22-Jul-2010 11:02:50] or can I just use the MIB names that the OIDs define
[22-Jul-2010 11:03:01] er the names that the mib defines
[22-Jul-2010 11:03:16] as in, I don't understand the purpose of the mib when it seems like I'm having to enter OIDs manually everywhere
[22-Jul-2010 11:04:08] mducharme-laptop: you get nicer texts in the events ;)
[22-Jul-2010 11:04:18] that's it?
[22-Jul-2010 11:04:32] that's all the mib does?
[22-Jul-2010 11:04:50] mducharme-laptop: i'm so used to oids, because most of the time you don't have the right mib around, that i didn't care and try ;)
[22-Jul-2010 11:11:25] mducharme-laptop: Mibs translate traps, that's it that's all
[22-Jul-2010 11:11:35] oh ok
[22-Jul-2010 11:14:32] iancmcc: a cleaner pastie http://pastie.org/1055587 (that hopefilly is understandable)
[22-Jul-2010 11:14:51] iancmcc: i'm out for the day, but if you msg me it would be very nice
[22-Jul-2010 11:30:34] hrm ok one other question
[22-Jul-2010 11:30:38] there are certain data items that I wish to collect
[22-Jul-2010 11:30:44] but I don't need them graphed - I just want to see a number
[22-Jul-2010 11:31:34] is there a way to just have numbers for things appear on the page where the graphs are?
[22-Jul-2010 11:31:57] things like firmware version etc
[22-Jul-2010 11:41:49] fOrsberg is now known as forsberg
[22-Jul-2010 11:43:46] mducharme-laptop: not out of the box, but hopefully eventually since it's been a common request
[22-Jul-2010 11:44:56] it just makes sense
[22-Jul-2010 11:44:59] yeah
[22-Jul-2010 11:51:26] that probably won't get added for quite a while though since there are higher priority issues that need to be worked first
[22-Jul-2010 11:51:37] unless a community ZenPack gets released that does that
[22-Jul-2010 12:12:08] <|rt|> rmatte: where does zenoss store the mysql username and password?
[22-Jul-2010 12:12:49] <|rt|> the stack installer never has you set the root mysql password so it either has a default password or generates and stores a random one somewhere
[22-Jul-2010 12:15:09] |rt|: you using 2.5 or 3.0?
[22-Jul-2010 12:15:32] just so I know what interface to use when walking you through this
[22-Jul-2010 12:17:08] <|rt|> 2.5
[22-Jul-2010 12:17:31] <|rt|> my update didn't work and killed my 2.5 install so I can't install the migration module
[22-Jul-2010 12:17:48] <|rt|> so I need to restore from a backup but i need the database username and password to do that
[22-Jul-2010 12:18:02] can you even get in to Zenoss though?
[22-Jul-2010 12:18:13] the password is stored in Zope
[22-Jul-2010 12:18:29] so you either need to get in to Zenoss, or open your Data.fs in a text editor and have fun combing through it
[22-Jul-2010 12:18:40] the default password should be "zenoss"
[22-Jul-2010 12:18:43] if you didn't set anything
[22-Jul-2010 12:19:26] <|rt|> hmm I think i tried zenoss...i'm in the middle fo jsut installing a clean 3.0 now....may just say screw it...my 2.5 install was only a couple of weeks old and wasn't in production yet
[22-Jul-2010 12:20:05] <|rt|> so I didn't lose anything that I can't redo in an hour other than some of the rrd data....but I have the rrd files so I can always copy those over to the new install once I get the devices set back up
[22-Jul-2010 12:20:05] I see
[22-Jul-2010 12:21:18] <|rt|> but the 3.0 stack .deb installer seems to break the 2.5 install if it doesn't find the 3.0 migration zenpack installed
[22-Jul-2010 12:21:35] <|rt|> I saw someone on the forum with the same issue
[22-Jul-2010 12:22:28] * rmatte just took his first shot at creating a quick collector plugin, moment of truth...
[22-Jul-2010 12:22:50] annnnnd it didn't work
[22-Jul-2010 12:22:52] hmmm
[22-Jul-2010 12:22:58] <|rt|> doh
[22-Jul-2010 12:24:36] I'm trying to copy the InterfaceAliasMap.py plugin and make one that pulls down CatOS descriptions
[22-Jul-2010 12:24:41] instead of the standard descriptions
[22-Jul-2010 12:26:30] hi there, is there anyone, who can help me with zenoss install from source?
[22-Jul-2010 12:27:30] drtic_cz: hardly anyone does source installs of Zenoss, I've never even done one and I've been using Zenoss for a long time
[22-Jul-2010 12:27:45] drtic_cz: your best bet is to follow the steps in the install guide
[22-Jul-2010 12:28:10] drtic_cz: just out of curiosity, why are you doing a source install in the first place?
[22-Jul-2010 12:28:41] rmatte: yes, I read the install guide - this isn't the problem probably..
[22-Jul-2010 12:29:06] because I'm using arch linux- there is no binaries for Arch :(
[22-Jul-2010 12:29:32] drtic_cz: The stack will probably work
[22-Jul-2010 12:29:50] drtic_cz: the stack installer will work on any version of linux
[22-Jul-2010 12:29:58] drtic_cz: it's the .bin file
[22-Jul-2010 12:30:05] I tried stack binary installer.. it freeze, without any signs of reason
[22-Jul-2010 12:30:15] no no, common misconception
[22-Jul-2010 12:30:19] it stops for a while at 99%
[22-Jul-2010 12:30:23] but it is actually still installing
[22-Jul-2010 12:30:31] people just have zero patience and usually end up killing it at the end
[22-Jul-2010 12:30:36] really?
[22-Jul-2010 12:30:37] if you run it and wait for it to finish it will work
[22-Jul-2010 12:30:39] yes, really
[22-Jul-2010 12:30:52] i'm waited for 15 minutes ant still nothing
[22-Jul-2010 12:30:59] leave it for more
[22-Jul-2010 12:31:11] it could take quite a while depending on the hardware you're installing it on
[22-Jul-2010 12:31:14] so, I will leave it over the night - thi should be enough
[22-Jul-2010 12:31:22] if it's still sitting at 99% after 30 minutes then I'd be concerned
[22-Jul-2010 12:31:29] thank you very much rmatte!
[22-Jul-2010 12:31:32] you shouldn't have to leave it overnight
[22-Jul-2010 12:31:33] np
[22-Jul-2010 12:31:52] No promises, but installation should be a tad faster in 3.1
[22-Jul-2010 12:31:59] one of the things being investigated now
[22-Jul-2010 12:32:07] and by "a tad" I mean "a lot"
[22-Jul-2010 12:32:19] iancmcc: the stack installer should really give some sort of notice about it sitting at 99% forever near the end of the install
[22-Jul-2010 12:32:30] tons of new users tend to kill the install thinking it's frozen
[22-Jul-2010 12:32:40] Yes. There's a ticket, and I don't know why we haven't done it yet.
[22-Jul-2010 12:32:48] either that, or have the % actually reflect the progress a bit better
[22-Jul-2010 12:32:49] lol
[22-Jul-2010 12:33:13] There's a limit to the extent we can see inside what's happening there
[22-Jul-2010 12:33:25] yeh, I know it's not you guys who build the installer
[22-Jul-2010 12:33:25] it's part of bitrock's thing
[22-Jul-2010 12:33:30] yeh
[22-Jul-2010 12:33:45] just adding a notice would be sufficient
[22-Jul-2010 12:34:10] "NOTE: The installer may appear to freeze at 99% near the end of the install, please be patient and allow the install to complete."
[22-Jul-2010 12:34:14] or something along those lines
[22-Jul-2010 12:34:46] iancmcc: how much do you know about collector plugins?
[22-Jul-2010 12:35:03] little, last time I was in there was a couple years ago
[22-Jul-2010 12:35:20] I'm trying to copy the InterfaceAliasMap.py plugin and modify it so that it will pull down CatOS interface descriptions
[22-Jul-2010 12:35:51] I thought it would be as simple as copying it to a different name and changing the OID, the datapoint, and the name in the plugin
[22-Jul-2010 12:35:53] but that didn't work
[22-Jul-2010 12:36:03] it seems to just block out InterfaceAliasMap instead of adding on
[22-Jul-2010 12:36:40] you renamed both the class and the module, yeah?
[22-Jul-2010 12:36:45] correct
[22-Jul-2010 12:36:55] let me show you what I have...
[22-Jul-2010 12:37:24] hurm
[22-Jul-2010 12:37:27] http://fpaste.org/IRug/
[22-Jul-2010 12:37:30] that's what I have
[22-Jul-2010 12:37:46] now I may be completely missing the point of how the InterfaceAliasMap plugin works
[22-Jul-2010 12:38:00] Maybe there's some hook in InterfaceMap that it's not jiving with
[22-Jul-2010 12:38:33] ignore the comment on lines 36 and 37, I haven't changed it yet
[22-Jul-2010 12:38:58] how often does zenoss remodel devices?
[22-Jul-2010 12:39:10] mducharme-laptop: it's a setting
[22-Jul-2010 12:39:13] rmatte: No clue. It looks right to me.
[22-Jul-2010 12:40:02] mducharme-laptop: one sec, finding where they stuck it in the new UI
[22-Jul-2010 12:40:14] <|rt|> i think in 2.x it was every 6 hours
[22-Jul-2010 12:40:29] |rt| by default, yeh, or every 12 or something
[22-Jul-2010 12:40:40] I believe 12
[22-Jul-2010 12:41:19] aha
[22-Jul-2010 12:41:37] I wonder why there are no nms systems with integrated ticketing.. it makes a certain amount of sense having it all in one application
[22-Jul-2010 12:41:47] mducharme-laptop: Advanced -> Collectors -> Edit -> Modeler Cycle Interval (mins)
[22-Jul-2010 12:42:02] cool rmatte, thanks
[22-Jul-2010 12:42:08] default is 720 mins which is 12 hours
[22-Jul-2010 12:42:26] can it be set individually for different devices?
[22-Jul-2010 12:42:38] or is it only available as a collector wide setting
[22-Jul-2010 12:42:46] no it's a collector thing
[22-Jul-2010 12:42:54] you could put devices on different collectors
[22-Jul-2010 12:42:57] but that's probably overkill
[22-Jul-2010 12:42:58] <|rt|> you can have more than one collector setup with different settings
[22-Jul-2010 12:43:15] You can also, if you really want to, just make a cronjob that runs zenmodeler
[22-Jul-2010 12:43:46] yeh, you could even have a cronjob that runs zenmodeler against a specific device
[22-Jul-2010 12:43:50] or specific group of devices
[22-Jul-2010 12:44:52] we have many devices over slow satellite links
[22-Jul-2010 12:45:09] we want to minimize traffic over those links
[22-Jul-2010 12:45:11] iancmcc: this collector plugin even seems to run fine, no errors when modelling, but it's obviously not working
[22-Jul-2010 12:45:41] iancmcc: I'm trying to run it in conjunction with InterfaceAliasMap, so maybe running the 2 at once is tripping it up, going to test without it in a second
[22-Jul-2010 12:51:48] nope, that's not it
[22-Jul-2010 12:52:44] yeah, that shouldn't be it
[22-Jul-2010 12:52:48] so it actually runs?
[22-Jul-2010 12:52:52] yup
[22-Jul-2010 12:53:06] it just doesn't do anything other than completely block out InterfaceAliasMap
[22-Jul-2010 12:53:21] I'm wondering if it has something to do with the fact that I'm not actually accessing the ifalias table
[22-Jul-2010 12:53:27] I switched the OID for something else
[22-Jul-2010 12:53:40] though it technically shouldn't from what I can see
[22-Jul-2010 12:53:42] here's something
[22-Jul-2010 12:53:51] so you're just schlocking this onto 'id', yeah?
[22-Jul-2010 12:53:56] line 48?
[22-Jul-2010 12:54:04] 43 I mean
[22-Jul-2010 12:54:09] was looking at InterfaceAliasMap
[22-Jul-2010 12:54:21] well, I'm going exactly was InterfaceAliasMap does
[22-Jul-2010 12:54:30] I just switched the oid in the table map
[22-Jul-2010 12:54:35] and the index of description
[22-Jul-2010 12:54:42] to .4
[22-Jul-2010 12:54:54] going == doing
[22-Jul-2010 12:55:01] was == what
[22-Jul-2010 12:55:05] man I can't type this morning
[22-Jul-2010 12:55:09] erm, afternoon
[22-Jul-2010 12:55:11] lol
[22-Jul-2010 12:56:09] yeah, that's possible (OID switching thing). I'm out of my sphere here, I'm just grasping
[22-Jul-2010 12:56:28] yeh, I know you're more UI
[22-Jul-2010 12:56:38] naah, just not this
[22-Jul-2010 12:56:42] ah
[22-Jul-2010 12:56:57] cgibbons knows a decent amount about collector plugins does he not?
[22-Jul-2010 12:57:08] or is he in the same boat?
[22-Jul-2010 12:57:10] some
[22-Jul-2010 12:58:58] hmmm, I kind of see where I might be going wrong here
[22-Jul-2010 12:59:36] not sure, need to dig a bit more
[22-Jul-2010 13:00:39] ohhhhh
[22-Jul-2010 13:00:40] I see
[22-Jul-2010 13:00:44] * rmatte slaps forehead
[22-Jul-2010 13:01:29] line 43 should actually be: results[1]['ifalias'][i_idx]['description'] = alias['description']
[22-Jul-2010 13:01:33] I believe
[22-Jul-2010 13:01:49] that is indeed what I meant before
[22-Jul-2010 13:01:53] yeh
[22-Jul-2010 13:01:57] blind guessing is my forte
[22-Jul-2010 13:01:57] let me test with that and see if it works
[22-Jul-2010 13:02:38] I didn't notice it was referencing the iftable table mapping and swapping out the value for id in there
[22-Jul-2010 13:02:44] just thought id was a variable or something
[22-Jul-2010 13:03:04] yeah you probably don't need to do anything, actually
[22-Jul-2010 13:03:10] that entire process method might be useless
[22-Jul-2010 13:03:22] because it's only there in InterfaceAliasMap so it'll use ifName for id
[22-Jul-2010 13:03:27] you don't care about that at all
[22-Jul-2010 13:03:53] well, I sort of do
[22-Jul-2010 13:04:04] ok.
[22-Jul-2010 13:04:04] but for description in ifalias instead
[22-Jul-2010 13:04:19] is there a better way to reload collector plugins than restarting Zenoss?
[22-Jul-2010 13:04:32] restart zenhub
[22-Jul-2010 13:04:41] just zenhub?
[22-Jul-2010 13:04:43] k cool
[22-Jul-2010 13:04:56] i believe so
[22-Jul-2010 13:05:03] k
[22-Jul-2010 13:05:08] well, time to test the plugin again
[22-Jul-2010 13:05:12] hopefully it just works this time
[22-Jul-2010 13:05:16] but chances are it won't
[22-Jul-2010 13:07:01] oh well, I have another 4 hours to figure this out lol
[22-Jul-2010 13:07:06] I'll just keep plugging away at it
[22-Jul-2010 13:07:27] going to release it as a ZenPack if I figure it out
[22-Jul-2010 13:08:07] oh by the way ian... there's no way to view interface descriptions in 3.0, that's pretty killer lol
[22-Jul-2010 13:08:21] I was going to log a trac ticket today, unless there's one already open somewhere
[22-Jul-2010 13:08:31] there is
[22-Jul-2010 13:08:34] somewhere
[22-Jul-2010 13:09:10] k
[22-Jul-2010 13:09:54] I can't switch to 3.0 until I'm able to have interface descriptions
[22-Jul-2010 13:10:00] since we use them all the time
[22-Jul-2010 13:10:50] http://dev.zenoss.org/trac/ticket/7076
[22-Jul-2010 13:10:56] annnnd the modeler plugin still isn't working, hmmm
[22-Jul-2010 13:11:24] cool
[22-Jul-2010 13:12:21] rmatte: You have a dev instance or something?
[22-Jul-2010 13:12:24] i have a patch for you to try
[22-Jul-2010 13:13:02] http://pastebin.ca/1906131
[22-Jul-2010 13:13:05] should take care of it
[22-Jul-2010 13:13:18] I really don't know if it'll get into 3.0.1, that's pretty locked down right now
[22-Jul-2010 13:13:24] but 3.0.2 certainly, and you can zenpatch easily
[22-Jul-2010 13:14:18] I missed it yesterday
[22-Jul-2010 13:14:22] Did they figure out how to fix shift?
[22-Jul-2010 13:14:30] And I didn't notice ifdesc, but if that's true, that's gonna suck
[22-Jul-2010 13:14:44] theacolyte: you mean shift on event console?
[22-Jul-2010 13:14:55] Yeah
[22-Jul-2010 13:14:58] that isn't a bug, it's intentional. I know it's annoying.
[22-Jul-2010 13:15:06] are you serious?
[22-Jul-2010 13:15:15] It was necessary for this reason
[22-Jul-2010 13:15:29] You can have millions of events in that table
[22-Jul-2010 13:15:31] How did that pass UAT?
[22-Jul-2010 13:15:38] iancmcc: won't that make it a bit crowded?
[22-Jul-2010 13:15:45] cut off the end of the descriptions for instance?
[22-Jul-2010 13:15:48] rmatte: Yeah, what do you want to get rid of?
[22-Jul-2010 13:15:59] What about having millions of events?
[22-Jul-2010 13:16:00] make IP address and Mac Address rollovers
[22-Jul-2010 13:16:04] You mean the way it loads events?
[22-Jul-2010 13:16:12] * rmatte shrugs
[22-Jul-2010 13:16:14] theacolyte: Because of millions of events, it has to load them in batches
[22-Jul-2010 13:16:24] Why can't I shift select in those batches?
[22-Jul-2010 13:16:25] at least get rid of Mac Address
[22-Jul-2010 13:16:28] That is a pretty major issue
[22-Jul-2010 13:16:39] I can think of tons and tons of use cases for shift
[22-Jul-2010 13:16:41] which means there's really no way to describe events that aren't loaded.
[22-Jul-2010 13:16:42] Mac Address isn't very useful and takes up a lot of space
[22-Jul-2010 13:16:45] Yes, there are tons of use cases
[22-Jul-2010 13:16:52] So why get rid of it?
[22-Jul-2010 13:16:52] but it breaks
[22-Jul-2010 13:16:58] because it introduced a bug
[22-Jul-2010 13:16:58] You could still use shift in the batch that's loaded
[22-Jul-2010 13:17:09] if you shift-select and an event has come in in the meantime
[22-Jul-2010 13:17:11] For me, the lack of shift is almost a deal breaker for me
[22-Jul-2010 13:17:12] in between the range
[22-Jul-2010 13:17:20] then you'll act on that event without ever seeing it
[22-Jul-2010 13:17:24] it's a terrible bug
[22-Jul-2010 13:17:34] I'm not a programmer,. so I can't speak to work arounds, but....
[22-Jul-2010 13:17:38] It is seriously obnoxious now
[22-Jul-2010 13:17:42] Like game changing obnoxious
[22-Jul-2010 13:17:48] yeh, it's pretty brutal
[22-Jul-2010 13:18:05] What happens when I say, want to clear everyting up to a point, and everything after that point?
[22-Jul-2010 13:18:13] filter, select all
[22-Jul-2010 13:18:14] ctrl-click click click click click click click click etc....
[22-Jul-2010 13:18:14] lol
[22-Jul-2010 13:18:16] Oh, I guess I get to clear 3000 events manually now
[22-Jul-2010 13:18:22] Right, select all won't work
[22-Jul-2010 13:18:26] Because that one event I want to keep
[22-Jul-2010 13:18:39] select all, then unselect the event.
[22-Jul-2010 13:18:43] well, do the select all, then hold down ctrl and click that 1 event
[22-Jul-2010 13:18:45] to unselect it
[22-Jul-2010 13:18:47] that's not that bad
[22-Jul-2010 13:18:49] ah you can do that still?
[22-Jul-2010 13:18:53] I should have tried that
[22-Jul-2010 13:18:58] yes. You can ctrl-click, just not shift.
[22-Jul-2010 13:18:59] lol
[22-Jul-2010 13:19:03] I haven't c leared any events since I went to 3.0
[22-Jul-2010 13:19:10] I recognize that it is a hassle but we have yet to come up with a solution
[22-Jul-2010 13:19:18] That's fair, but please - work on it
[22-Jul-2010 13:19:23] It's a serious step backwards
[22-Jul-2010 13:19:29] good that there's a workaround
[22-Jul-2010 13:19:31] * theacolyte clears his events
[22-Jul-2010 13:19:34] and the UI annoyance is less annoying than accidentally clearing events you never saw
[22-Jul-2010 13:19:41] this is true
[22-Jul-2010 13:19:48] iancmcc: why does it even act on events that aren't there?
[22-Jul-2010 13:19:58] because the UI can only know about ranges of events
[22-Jul-2010 13:20:09] rmatte: remind me how i can use zendmd to clear events with no email
[22-Jul-2010 13:20:09] either that or load up the id of every event first, which could be millions.
[22-Jul-2010 13:20:14] ok another question for you guys
[22-Jul-2010 13:20:25] so the selection is stored as, say, [0,100]
[22-Jul-2010 13:20:26] theacolyte: what do you mean by with no email?
[22-Jul-2010 13:20:30] (sorry I've been having so many and thanks for your help)
[22-Jul-2010 13:20:37] rmatte: no cleared emails
[22-Jul-2010 13:20:53] how does it know what performance template to use for a device
[22-Jul-2010 13:20:57] in the perf tab
[22-Jul-2010 13:21:03] theacolyte: what does emails have to do with it? are you talking acknowledged events?
[22-Jul-2010 13:21:09] yes
[22-Jul-2010 13:21:11] if an event comes in between 0 and 100 /after/ the page loaded but /before/ the button is pressed, the front end has no way of recognizing that.
[22-Jul-2010 13:21:14] I just need to clear my events
[22-Jul-2010 13:21:21] theacolyte: speak in terms of the event console, emails are irrelevant and confusing to me lol
[22-Jul-2010 13:21:24] the performance template used by the interfaces is set by the interface type name
[22-Jul-2010 13:21:31] theacolyte: all of them without exception?
[22-Jul-2010 13:21:44] yes
[22-Jul-2010 13:21:46] just need to close them all
[22-Jul-2010 13:21:49] how do you specify which performance template is used for the device in the Perf tab
[22-Jul-2010 13:22:05] theacolyte: http://fpaste.org/DbOe/
[22-Jul-2010 13:22:06] rmatte: I don't want to get 3000 emails sayign they are each cleared when I do it through the console
[22-Jul-2010 13:22:19] mducharme-laptop: It's the template named Device that's bound to the device or a parent
[22-Jul-2010 13:22:22] cool
[22-Jul-2010 13:22:24] thanks man
[22-Jul-2010 13:22:30] I know it says "deleting" but it's just moving them to history
[22-Jul-2010 13:22:34] it's just the way I worded it
[22-Jul-2010 13:22:35] lol
[22-Jul-2010 13:22:35] yeah
[22-Jul-2010 13:22:40] ahh iancmcc ok
[22-Jul-2010 13:23:31] rmatte: so you want to ditch macaddress?
[22-Jul-2010 13:23:41] that'll take a little more discussion than just me doing it
[22-Jul-2010 13:23:45] but I'll put it in the ticket
[22-Jul-2010 13:26:06] rmatte: weird that doesn't seem to work for some reason and what little I know it should
[22-Jul-2010 13:26:22] oh wait
[22-Jul-2010 13:26:25] i take that back, sorry, my bad
[22-Jul-2010 13:26:32] the enter key is the killer :)
[22-Jul-2010 13:27:03] oh god
[22-Jul-2010 13:27:06] it's gonna send me emails
[22-Jul-2010 13:27:11] hahahahahah
[22-Jul-2010 13:27:30] I hope this reboot happens before I get them all
[22-Jul-2010 13:28:00] lol
[22-Jul-2010 13:28:20] that email thing must be something built in to the backend
[22-Jul-2010 13:28:33] I've never had it send me any emails for cleared events
[22-Jul-2010 13:28:35] only got 275 of them
[22-Jul-2010 13:28:45] I hope it doesn't save them for when it comes back up
[22-Jul-2010 13:28:45] haha
[22-Jul-2010 13:28:50] it might
[22-Jul-2010 13:28:52] lol
[22-Jul-2010 13:28:54] but I doubt it
[22-Jul-2010 13:28:57] oh, it did.
[22-Jul-2010 13:29:00] haha
[22-Jul-2010 13:29:01] haha
[22-Jul-2010 13:29:03] bombs away!
[22-Jul-2010 13:29:13] I think I had something like 4000 events
[22-Jul-2010 13:29:23] your mail server is huffing and puffing right now
[22-Jul-2010 13:29:26] If I was smart, I would have just removed my email address
[22-Jul-2010 13:29:32] lol
[22-Jul-2010 13:30:06] haha
[22-Jul-2010 13:30:10] there goes all the heartbeat messages
[22-Jul-2010 13:31:07] hahaha the server is completely unresponsive
[22-Jul-2010 13:31:17] so would anyone here like to share their perf data, if they've got a whole lot of devices, to the greater good of scale testing for the reporting project we're working on?
[22-Jul-2010 13:31:37] Sure, I'd be OK with that, but you would need to define the metrics
[22-Jul-2010 13:31:43] and testing methodologies
[22-Jul-2010 13:31:50] cgibbons: How do I get this thing to stop emailing me
[22-Jul-2010 13:32:00] lmao
[22-Jul-2010 13:32:03] flooooood
[22-Jul-2010 13:32:13] Only about 200 more to go
[22-Jul-2010 13:32:17] sorry, 2000
[22-Jul-2010 13:32:18] no idea
[22-Jul-2010 13:32:27] kill zenactions
[22-Jul-2010 13:32:28] you should have removed your email address from the system before clearing the events :P
[22-Jul-2010 13:32:36] then remove your email address
[22-Jul-2010 13:32:37] I can't SSH to the box
[22-Jul-2010 13:32:38] hahahaha
[22-Jul-2010 13:32:39] then restart zenactions
[22-Jul-2010 13:32:40] oh
[22-Jul-2010 13:32:45] then you're screwed
[22-Jul-2010 13:32:46] it's not responding
[22-Jul-2010 13:32:47] <|rt|> pull the plug
[22-Jul-2010 13:32:49] why can't you ssh?
[22-Jul-2010 13:32:51] oh
[22-Jul-2010 13:32:57] it's not even pingable hahaha
[22-Jul-2010 13:32:59] hmm, that's weird
[22-Jul-2010 13:33:00] lol @ |rt|
[22-Jul-2010 13:33:29] well
[22-Jul-2010 13:33:33] I will just go to lunch
[22-Jul-2010 13:33:35] they've always said that curiosity killed the cat, but apparently it died because it was trying to send 4000 emails
[22-Jul-2010 13:33:43] <|rt|> is there any way to tell if a zenpack is compatible with 3.x?
[22-Jul-2010 13:33:43] by the way, one notice - this is very, very, very much more slower than 2.0 was
[22-Jul-2010 13:33:54] 2.5 would do this in about 3 minutes, no unresponsiveness
[22-Jul-2010 13:34:13] lunch time
[22-Jul-2010 13:34:31] |rt|: if it doesn't have a -py2.4 in the filename
[22-Jul-2010 13:34:43] |rt|: and even if it does, that can be removed and it'll install anyways in most cases
[22-Jul-2010 13:34:51] easiest just to ask the author
[22-Jul-2010 13:34:54] most are compatible
[22-Jul-2010 13:35:19] I believe mray has a list going, but he's not here
[22-Jul-2010 13:35:33] iancmcc: so after an upgrade a lot of ZenPack folders still have the -py2.4 at the end of the name, but the ZenPacks appear to function, is that going to cause a problem in the future?
[22-Jul-2010 13:35:38] should they all ultimately be reinstalled?
[22-Jul-2010 13:35:52] it will not cause a problem
[22-Jul-2010 13:35:55] k
[22-Jul-2010 13:36:24] at least, not for that reason. But nothing's guaranteeing it's been tested against 2.6 in that case
[22-Jul-2010 13:36:46] nonetheless it'll probably be fine. There are a couple tiny little things that break them, and only a couple packs use those things
[22-Jul-2010 13:36:51] simplejson is the biggie
[22-Jul-2010 13:38:55] <|rt|> rmatte: ok I'll try that with teh foritgate zenpack
[22-Jul-2010 13:39:19] k
[22-Jul-2010 13:40:51] <|rt|> cool that worked....should I put in support ticket against that zenpack for them to rename it>?
[22-Jul-2010 13:43:17] <|rt|> hmm getting this error now on the device page when I click on the monitoring templates "DistributionNotFound" "ZenPacks.community.Fortigate"
[22-Jul-2010 13:43:36] hello all
[22-Jul-2010 13:43:51] |rt| did you restart Zenoss after installing it?
[22-Jul-2010 13:44:00] Zenoss does not show my tmpfs when I have remodeled the device and removed the ramDisk from ignored filesystems. Any ideas?
[22-Jul-2010 13:44:12] <|rt|> nope....just went to the zenpacks page and see the error there about restarting zope :)
[22-Jul-2010 13:44:26] QubeZ: did you remove the ramDisk from ignored filesystems PRIOR to remodeling?
[22-Jul-2010 13:44:44] rmatte: yup, remodeled many times since then and still doesn't show up. My NFS mount does however.
[22-Jul-2010 13:44:47] |rt| yeh, restart Zenoss
[22-Jul-2010 13:44:53] i even ran the snmp walk but it doesnt show up for the device
[22-Jul-2010 13:44:59] QubeZ: perhaps tmpfs is not defined as a ramDisk type?
[22-Jul-2010 13:45:02] <|rt|> rmatte: do you need to restart all of zenoss or just one of the deamons?
[22-Jul-2010 13:45:06] rmatte: how do I do that?
[22-Jul-2010 13:45:10] QubeZ: unless something has changed since yesterday, the snmp daemon (which is not part of Zenoss) is not providing data for that filesystem
[22-Jul-2010 13:45:16] QubeZ: you have to check with snmpwalk
[22-Jul-2010 13:45:27] QubeZ: if it's not showing up in snmpwalk then you're out of luck
[22-Jul-2010 13:45:34] wtucker: yup, so how do I fix that? I haven't been able to find many resources on the intenret.
[22-Jul-2010 13:45:56] thats so odd because tmpfs is very much used these days... surprised snmp doesn't report it
[22-Jul-2010 13:46:45] QubeZ: http://markmail.org/thread/nsceuwb225wm2cps
[22-Jul-2010 13:48:17] thanks, i dont see the attached patch though
[22-Jul-2010 13:48:35] it's not
[22-Jul-2010 13:48:43] I'm just making you aware that it's a known bug with net-snmp
[22-Jul-2010 13:48:46] nevermind, i see teh link
[22-Jul-2010 13:49:25] ah apparently there is a patch
[22-Jul-2010 13:49:37] at the sourceforge link
[22-Jul-2010 13:50:13] yup, i'll have to find that version of net-snmp and compile it
[22-Jul-2010 13:50:39] QubeZ: it's a pretty simple change, just find the source for the version you're currently running
[22-Jul-2010 13:51:01] i installed via yum, ya need to find src
[22-Jul-2010 13:51:10] and remove tmpfs from HRFS_ignores[] in agent/mibgroup/host/hr_filesys.c
[22-Jul-2010 13:52:31] there should be an srpm avilable from your distribution
[22-Jul-2010 13:54:07] I'm unable to create copies of the ethernetCsmacd or ethernetCsmacd_64 templates in 3.0.0. I get an exception: BadRequest: The property decription does not exist
[22-Jul-2010 13:54:58] other templates copy just fine. Before I try to track it down in parts of the source that I don't know, does the solution jump out to anyone?
[22-Jul-2010 13:55:57] <|rt|> how do you change the device class in 3.x?
[22-Jul-2010 13:56:38] |rt|: drag and drop. When you click for the drag also make sure you click on a part of the line that isn't text
[22-Jul-2010 13:56:55] (took me a while to figure that one out, too)
[22-Jul-2010 13:57:22] eugh, I just can't wrap my head around this damn code
[22-Jul-2010 13:57:30] <|rt|> thanks
[22-Jul-2010 13:57:41] or, Change Class from the gear menu when on the device page
[22-Jul-2010 13:58:45] cgibbons: I'm trying to copy InterfaceAliasMap.py and modify it to pull down CatOS interface descriptions... I know that something is screwed with it. It doesn't error out at all, it just doesn't work. I'm trying to wrap my head around the process function....
[22-Jul-2010 13:59:15] cgibbons: http://fpaste.org/wXYS/
[22-Jul-2010 13:59:19] the process part of it is down inside zenhub, are you running him in debug?
[22-Jul-2010 13:59:49] I want it to do what InterfaceAliasMap does, but instead of grabbing ifName and replacing ifDescr (id) with it, I want it to replace the description with
[22-Jul-2010 14:00:02] nope, not running him in debug, but I'm sure the code is screwed to begin with
[22-Jul-2010 14:00:15] it seems to be trying to match some sort of index value
[22-Jul-2010 14:00:17] it's at least ugly
[22-Jul-2010 14:00:46] yup
[22-Jul-2010 14:00:50] i'd at least just throw a pdb in there, run zenhub and then step through it
[22-Jul-2010 14:01:13] I guess I'll have to
[22-Jul-2010 14:01:40] this seems like it should be pretty simple
[22-Jul-2010 14:01:55] pulling down description from an alternate OID and overriding description with it
[22-Jul-2010 14:02:02] iancmcc: do you know off hand what your "phone home" code looks like? I.e. the stuff in zenoss that sends back stats (and also tells you of new releases, IIRC) ?
[22-Jul-2010 14:02:22] it looks like python code
[22-Jul-2010 14:02:26] lol
[22-Jul-2010 14:02:28] naw, what do you mean?
[22-Jul-2010 14:02:44] it's pretty old, I don't think we've been in there in a while
[22-Jul-2010 14:02:51] contrary to popular belief, iancmcc is not an svn system :)
[22-Jul-2010 14:03:00] <|rt|> how do you create groups, systems, and locations in 3.x?
[22-Jul-2010 14:03:10] <|rt|> this new UI is going to take a bit to get used to heh
[22-Jul-2010 14:03:11] |rt| from the menu on the bottom bar
[22-Jul-2010 14:03:17] |rt|: + button
[22-Jul-2010 14:03:19] bottom left
[22-Jul-2010 14:03:22] bottom left, as rmatte says
[22-Jul-2010 14:03:25] <|rt|> ah I didn't even see that
[22-Jul-2010 14:03:31] :)
[22-Jul-2010 14:04:03] the new UI will be painful for most existing users.. new users will find it easier
[22-Jul-2010 14:04:24] <|rt|> yeah it's not bad...just different....I'll figure it out :)
[22-Jul-2010 14:04:30] hopefully existing users will find it easier too. You don't have to click down into organizers anymore, that's a big win.
[22-Jul-2010 14:04:44] they will eventually, once they stop fighting it
[22-Jul-2010 14:05:11] and we fix some of the big bugs we don't know about yet
[22-Jul-2010 14:05:55] aclark: http://dev.zenoss.org/trac/browser/trunk/Products/ZenEvents/UpdateCheck.py
[22-Jul-2010 14:05:59] oh i found a minor one yesterday, Ian. just a wording. you don't move a device into an organizer (group, system). the little dialog that pops up when you do the drag & drop into them.
[22-Jul-2010 14:06:11] oh
[22-Jul-2010 14:06:16] iancmcc: neat, thx
[22-Jul-2010 14:06:25] pretty simple
[22-Jul-2010 14:06:59] hmmmm: "uc.getUpdate = lambda *unused: parseVersion('0.24.0')"
[22-Jul-2010 14:08:41] beats me
[22-Jul-2010 14:08:58] * aclark considering adding a feature like this to Plone
[22-Jul-2010 14:09:01] iancmcc: not having to click in to organizer would be a big win if the infrastructure page loaded quicker
[22-Jul-2010 14:09:14] rmatte: Yeah, I really can't reproduce that one
[22-Jul-2010 14:09:27] I know others can, but it's tough to narrow down. Mine loads in no time.
[22-Jul-2010 14:09:35] you using IE?
[22-Jul-2010 14:09:41] have you tried with Firefox?
[22-Jul-2010 14:09:46] I am offended you have even asked such a question.
[22-Jul-2010 14:09:52] * rmatte shrugs
[22-Jul-2010 14:09:58] you're doing UI design for both
[22-Jul-2010 14:09:58] I've tried it in all browsers, but anyway it's a server issue
[22-Jul-2010 14:10:01] true
[22-Jul-2010 14:10:05] I see
[22-Jul-2010 14:10:14] you tried public-demo, right?
[22-Jul-2010 14:10:35] I'll try it again right now
[22-Jul-2010 14:10:50] The server reported the following error:
[22-Jul-2010 14:10:50] Unable to connect to the server.
[22-Jul-2010 14:10:52] lol
[22-Jul-2010 14:10:53] hang on because public-demo is dying
[22-Jul-2010 14:10:56] heh
[22-Jul-2010 14:10:58] looks like public-demo is dead
[22-Jul-2010 14:11:00] * rmatte cries
[22-Jul-2010 14:11:07] i'm restarting mysql. It probably filled up the partition again.
[22-Jul-2010 14:11:10] what specs did you have on your 3.0 test instance, rmatte? that's been an issue for a few users.
[22-Jul-2010 14:11:11] <|rt|> seems fast here on my zenoss install but I don't have that many devices and I'm using Chrome
[22-Jul-2010 14:11:25] it's nice that the server exception window's tint isn't backwards there
[22-Jul-2010 14:11:26] (the min reqs. went up)
[22-Jul-2010 14:11:42] I've seen it pop up faded and then become dark when you drag it
[22-Jul-2010 14:11:45] lol
[22-Jul-2010 14:12:00] rmatte: that's because the server is dead and it can't load the background image
[22-Jul-2010 14:12:08] cgibbons: what are the minimum specs now?
[22-Jul-2010 14:12:40] my test instance has 2GB of RAM and upwards of 15GHz processing
[22-Jul-2010 14:12:47] though it averages 3 to 4GHz
[22-Jul-2010 14:12:49] ahh it's now 4gb :) you read the release notes right? docs/DOC-7442
[22-Jul-2010 14:13:15] rmatte: mysql stuff is what's screwing you over, I believe.
[22-Jul-2010 14:13:23] that's a bit much lol
[22-Jul-2010 14:13:26] 4GB minimum
[22-Jul-2010 14:13:47] Have you messed with mysql config on your test box?
[22-Jul-2010 14:13:51] nope
[22-Jul-2010 14:13:57] I'm fairly sure that's why
[22-Jul-2010 14:14:01] I see
[22-Jul-2010 14:14:04] the OTB config is suited for a blog, that's basically it
[22-Jul-2010 14:14:11] it's only using 1.5GB of RAM right now
[22-Jul-2010 14:14:17] yeh
[22-Jul-2010 14:14:26] I had to modify it on a few of my servers
[22-Jul-2010 14:14:29] but left it alone on some
[22-Jul-2010 14:14:45] it would be nice if you guys actually had the out of the box config make sense for the product
[22-Jul-2010 14:14:54] we don't ship mysql tho
[22-Jul-2010 14:14:58] oh i guess we do in the stack
[22-Jul-2010 14:15:00] yeh
[22-Jul-2010 14:15:03] exactly
[22-Jul-2010 14:15:10] the stack should come configured appropriate to Zenoss
[22-Jul-2010 14:15:13] lol
[22-Jul-2010 14:15:19] but also it depends totally on hardware
[22-Jul-2010 14:15:38] My VMs run on some beefy blades and have plenty of resources allocated to them
[22-Jul-2010 14:15:44] they are also running off a SAN
[22-Jul-2010 14:15:48] so no real IO issues
[22-Jul-2010 14:16:10] I'll try changing the MySQL config a bit later
[22-Jul-2010 14:16:36] either way, a loading dialog would be lovely
[22-Jul-2010 14:16:37] :)
[22-Jul-2010 14:16:57] "Ok, so it's sitting at a white screen... is it frozen?"
[22-Jul-2010 14:17:55] ?
[22-Jul-2010 14:17:58] whoops
[22-Jul-2010 14:18:01] lol
[22-Jul-2010 14:18:18] would it be useful for him to run with firebug net enabled so we can see where the time is being spent on the remote calls?
[22-Jul-2010 14:20:40] <|rt|> I think zenjobs must have crashed
[22-Jul-2010 14:20:57] <|rt|> just had to restart it...was wondering why it wasn't modeling the devices as I added them
[22-Jul-2010 14:25:53] I just tuned mysql, the speed seems to have improved
[22-Jul-2010 14:26:13] it still takes about 2 or 3 seconds
[22-Jul-2010 14:26:15] which isn't bad
[22-Jul-2010 14:26:23] A mere 3000 unread messages
[22-Jul-2010 14:26:24] could still use a quick loading dialog
[22-Jul-2010 14:26:31] theacolyte: nice
[22-Jul-2010 14:26:34] what was it before?
[22-Jul-2010 14:26:46] cgibbons: like 5 to 10 seconds before
[22-Jul-2010 14:26:52] annoying yes
[22-Jul-2010 14:26:56] cgibbons: though I've seen it take 25 to 30
[22-Jul-2010 14:26:57] more than 2 seconds = progress needed
[22-Jul-2010 14:27:22] yeh'
[22-Jul-2010 14:27:58] actually that unresponsiveness may actually be something other than zenoss that just broke
[22-Jul-2010 14:28:37] wow, yeah, it's not passing traffic (the VM)
[22-Jul-2010 14:29:01] hmm this is interesting
[22-Jul-2010 14:31:39] rmatte, you know anything about Matts work on a chef recipe ?
[22-Jul-2010 14:33:12] jassinpain: nope, haven't seen any of it
[22-Jul-2010 14:33:27] jassinpain: I'm aware that he's working on it, that's about it
[22-Jul-2010 14:33:33] wow this is really weird
[22-Jul-2010 14:33:45] zenoss is sending me emails, but I can't ping it or access it except from console
[22-Jul-2010 14:33:54] ;)
[22-Jul-2010 14:35:11] * theacolyte checks his CAM table
[22-Jul-2010 14:38:35] the emails might just be saturating the link
[22-Jul-2010 14:38:39] hence the loss of connectivity
[22-Jul-2010 14:38:51] no, that's not it
[22-Jul-2010 14:39:15] I can actually access it internally from the network it is on
[22-Jul-2010 14:39:21] I see
[22-Jul-2010 14:39:27] But not over VPN
[22-Jul-2010 14:39:33] and also vmware tools isn't running it says
[22-Jul-2010 14:39:39] perhaps you're going through different virtual interfaces?
[22-Jul-2010 14:39:49] depending on where you're connecting to it from
[22-Jul-2010 14:40:13] vmware tools not running could be a reason
[22-Jul-2010 14:40:28] you know what
[22-Jul-2010 14:40:33] I bet you that's actually the problem, but why!
[22-Jul-2010 14:40:34] Why!
[22-Jul-2010 14:41:06] if you can connect via console just start it back up
[22-Jul-2010 14:41:13] ok this is weird
[22-Jul-2010 14:41:18] yeap
[22-Jul-2010 14:41:25] this interface has operStatus and adminStatus both as down
[22-Jul-2010 14:41:33] but, the lights are both green in the OS interfaces page
[22-Jul-2010 14:41:41] the lights are green regardless of whether it is up or down
[22-Jul-2010 14:41:56] what is the point of that?
[22-Jul-2010 14:42:08] mducharme-laptop: They will be unless you have a setup like this: docs/DOC-2494
[22-Jul-2010 14:42:13] ehghhh I see what it is
[22-Jul-2010 14:42:17] also, if you remodel with it admin down, they should turn grey
[22-Jul-2010 14:42:21] instead of green
[22-Jul-2010 14:43:05] ahh I see
[22-Jul-2010 14:43:13] yeah we;re not gonna talk about this one :)
[22-Jul-2010 14:43:17] mducharme-laptop: also, the 2 lights don't represent oper status and admin status
[22-Jul-2010 14:43:20] I really shouldn't have done a dist-upgrade on ubuntu
[22-Jul-2010 14:43:22] hmm now i got it to read the /dev/shm tmpfs via snmp but not my custom on from /etc/fstab
[22-Jul-2010 14:43:25] oh
[22-Jul-2010 14:43:34] when I read the documentation it seemed to say that that is what those lights meant
[22-Jul-2010 14:43:34] they represent overall status (basically oper status), and whether or not that interface is set to be monitored or not
[22-Jul-2010 14:43:40] but it looked like the documentation was for an older version
[22-Jul-2010 14:43:41] ok
[22-Jul-2010 14:43:42] the documentation is outdated
[22-Jul-2010 14:43:45] it used to be that way
[22-Jul-2010 14:43:47] it was changed
[22-Jul-2010 14:43:49] ok
[22-Jul-2010 14:43:57] there's Status and the M column which means monitored
[22-Jul-2010 14:44:20] and if you do that interface polling setup it will change the status icon if an interface goes down
[22-Jul-2010 14:44:30] that's the way I monitor interfaces
[22-Jul-2010 14:45:28] gahhhhh, I can not see why this collector plugin isn't working, I figured out the process function and fixed it
[22-Jul-2010 14:45:36] it should be working now but it's not
[22-Jul-2010 14:45:37] garrrr
[22-Jul-2010 14:56:01] aurgh
[22-Jul-2010 14:56:13] I'm trying to figure out how to pull the oper status of these interfaces
[22-Jul-2010 14:56:21] unfortunately it's not so simple as just pulling it from snmp
[22-Jul-2010 14:56:35] the snmp value is either 0, 2, or 3 if it is down, and 1 if it is up
[22-Jul-2010 14:57:11] it will incorrectly read 3 as being up when it is down
[22-Jul-2010 14:57:33] what are you talking about? that guide I pasted monitor oper status
[22-Jul-2010 14:57:39] monitors*
[22-Jul-2010 14:57:43] just follow the guide
[22-Jul-2010 14:58:25] docs/DOC-2494
[22-Jul-2010 14:58:46] yes I followed the guide
[22-Jul-2010 14:58:51] so what's the issue?
[22-Jul-2010 14:59:06] if you set the threshold to a max of 1 and a min of 1 it will alert if the value is anything other than 1
[22-Jul-2010 14:59:09] this is not cisco operstatus
[22-Jul-2010 14:59:21] operstatus is a standard bud
[22-Jul-2010 14:59:47] it's not just cisco gear that follows it
[22-Jul-2010 15:00:17] If it shows up in the interfaces list in Zenoss, then you can monitor it using that guide
[22-Jul-2010 15:00:32] hahahahahaha
[22-Jul-2010 15:00:39] ahh I see
[22-Jul-2010 15:00:42] I am never doing another version upgrade on ubuntu again
[22-Jul-2010 15:00:48] I need to install tools
[22-Jul-2010 15:00:54] I don't have linux headers installed
[22-Jul-2010 15:00:59] theacolyte: nice
[22-Jul-2010 15:01:01] I can't install them... because no tools
[22-Jul-2010 15:01:11] apt-get install linux-server-headers I think it is
[22-Jul-2010 15:01:19] what do you mean no tools?
[22-Jul-2010 15:01:21] I know how to install them, but I can't talk to any repos
[22-Jul-2010 15:01:29] oh, right
[22-Jul-2010 15:01:39] chicken and the egg scenario lol
[22-Jul-2010 15:01:42] I actually don't know what to do
[22-Jul-2010 15:01:50] grab the .deb
[22-Jul-2010 15:01:55] how do I create the status ipinterface event class
[22-Jul-2010 15:01:56] create an iso with it
[22-Jul-2010 15:01:56] there is no add button
[22-Jul-2010 15:01:57] I suppose I could download the package, put it on a virtual iso
[22-Jul-2010 15:01:58] and mount it
[22-Jul-2010 15:02:01] yeah
[22-Jul-2010 15:02:03] haha
[22-Jul-2010 15:02:11] that's what I'd do, it's really your only choice
[22-Jul-2010 15:02:40] I will say there's a reason I'm a windows admin :)
[22-Jul-2010 15:02:45] lol
[22-Jul-2010 15:02:52] I'm a Linux admin all the way ;)
[22-Jul-2010 15:02:59] it's fun fun fun
[22-Jul-2010 15:03:23] you have a weird notion of fun
[22-Jul-2010 15:03:27] "Set the event class to /Status/IpInterface (if this event class does not exist, create it)."
[22-Jul-2010 15:03:28] eugh... I'm absolutely positive that this collector plugin code is good now
[22-Jul-2010 15:03:32] but it's still refusing to work properly
[22-Jul-2010 15:03:35] there is no option to create that event class
[22-Jul-2010 15:03:54] mducharme-laptop: yeh there is, hold on...
[22-Jul-2010 15:04:02] ok
[22-Jul-2010 15:04:33] rmatte: how do I even find the deb
[22-Jul-2010 15:04:52] Events -> Event Classes -> Status -> Click on the dropdown beside "SubClasses" -> Add New Organizer
[22-Jul-2010 15:05:06] mducharme-laptop: not sure how you could have missed that :P
[22-Jul-2010 15:05:25] theacolyte: do a uname -a for me please
[22-Jul-2010 15:05:25] ah there we go it timed out and gave me the url
[22-Jul-2010 15:05:31] ah ok
[22-Jul-2010 15:05:33] that works too
[22-Jul-2010 15:05:35] thankfully
[22-Jul-2010 15:05:35] :)
[22-Jul-2010 15:05:43] I was on us.archive.ubuntu.com but it's hard to find stuff hehe
[22-Jul-2010 15:05:58] well, if I had gotten your kernel version it would have been easy to find via google
[22-Jul-2010 15:06:05] but you're good now
[22-Jul-2010 15:06:35] hopefully it doesn't need dependencies to install
[22-Jul-2010 15:06:44] guess you'll just have to wait and see
[22-Jul-2010 15:06:52] it might need kernel-source or something along those lines
[22-Jul-2010 15:07:28] it'll tell you if it does
[22-Jul-2010 15:08:04] http://us.archive.ubuntu.com/ubuntu/pool/main/l/linux/
[22-Jul-2010 15:08:08] it doesn't
[22-Jul-2010 15:08:15] thankfully apt-get actually gave me all the paths and filenames
[22-Jul-2010 15:08:24] man
[22-Jul-2010 15:08:29] cool
[22-Jul-2010 15:08:32] this is exzactly why doing inplace upgrades is always a bad idea
[22-Jul-2010 15:08:37] it proves itself time and time again
[22-Jul-2010 15:08:44] there's always some hidden gotcha
[22-Jul-2010 15:09:22] mducharme-laptop: at the part that it says "More -> Transform" it'll be the menu in the bottom left hand corner
[22-Jul-2010 15:13:46] rmatte: heh, installed them, but it didn't do anything in /usr/src/linux/include
[22-Jul-2010 15:13:49] I wonder where it just put it
[22-Jul-2010 15:16:02] let me check on one of the boxes that I have it installed on
[22-Jul-2010 15:16:17] by the way, what version of ubuntu did you upgrade to?
[22-Jul-2010 15:17:34] theacolyte?
[22-Jul-2010 15:17:52] * rmatte yawns
[22-Jul-2010 15:18:02] 10.04
[22-Jul-2010 15:18:03] sorry
[22-Jul-2010 15:18:08] I'm about to kill this thing
[22-Jul-2010 15:18:18] I can' get this damn stuff installed
[22-Jul-2010 15:18:19] k, vmware tools doesn't play nice with that at all, but I have a patching method for it
[22-Jul-2010 15:18:23] hold on
[22-Jul-2010 15:18:41] yeah i' not getting that far
[22-Jul-2010 15:18:45] i'm installing headers
[22-Jul-2010 15:18:48] I need to modify it so that I can send it to you
[22-Jul-2010 15:18:52] and it's not actually putting the files anywhere
[22-Jul-2010 15:19:14] Hey all, I have setup monitoring of my T1's, and of course a ton of notifications came up on T1 cards that are unused. I don't really want to turn off monitoring on those endpoints... I put them as admin down (shut the interface...) and it's still throwing an event.
[22-Jul-2010 15:19:20] they should be somewhere like: /usr/src/linux-headers-2.6.32-23
[22-Jul-2010 15:19:31] you sure they aren't there?
[22-Jul-2010 15:20:36] that should have a symlink to /usr/src/linux/include
[22-Jul-2010 15:20:38] maybe that's the issue
[22-Jul-2010 15:20:43] good call let me check if this fails again
[22-Jul-2010 15:21:07] nah, you don't need a symlink to that
[22-Jul-2010 15:21:23] you do if vmware tools is looking for headers in that area
[22-Jul-2010 15:21:29] oh wait,
[22-Jul-2010 15:21:33] no no
[22-Jul-2010 15:21:46] I figured it out
[22-Jul-2010 15:21:49] you need to run /usr/bin/vmware-config-tools.pl again
[22-Jul-2010 15:21:52] I didn't actually know
[22-Jul-2010 15:21:52] yeah
[22-Jul-2010 15:21:54] and it'll actually find the headers
[22-Jul-2010 15:21:56] I had put it to sleep/fg
[22-Jul-2010 15:21:56] BUT
[22-Jul-2010 15:22:03] you need to patch one of the modules before building
[22-Jul-2010 15:22:06] or it won't build
[22-Jul-2010 15:22:09] yeah i don't care about hgfs and the other one
[22-Jul-2010 15:22:14] no, vmci
[22-Jul-2010 15:22:15] err vmci
[22-Jul-2010 15:22:22] k, up to you
[22-Jul-2010 15:22:22] lol
[22-Jul-2010 15:22:24] yeah I don't care
[22-Jul-2010 15:22:26] I personally patched it
[22-Jul-2010 15:22:27] hehe
[22-Jul-2010 15:22:37] it's being phased out ayway
[22-Jul-2010 15:22:48] meh, might as well have it running properly in the meantime
[22-Jul-2010 15:22:53] nein
[22-Jul-2010 15:22:53] * rmatte shrugs
[22-Jul-2010 15:22:54] hehe
[22-Jul-2010 15:22:59] It doesn't actually even try to run it
[22-Jul-2010 15:23:06] mine does
[22-Jul-2010 15:23:10] oh wait, it does...
[22-Jul-2010 15:23:19] yup
[22-Jul-2010 15:23:32] ok, ok, where's your patch :0
[22-Jul-2010 15:23:35] :P
[22-Jul-2010 15:23:37] Hopefully I can type it
[22-Jul-2010 15:23:39] hahaha
[22-Jul-2010 15:23:41] now I have to prep it for you, hold on
[22-Jul-2010 15:23:49] yeh, one sec
[22-Jul-2010 15:25:08] arrrghhh: did you model the device after shutting the interfaces?
[22-Jul-2010 15:25:17] yeah you have to do that
[22-Jul-2010 15:25:35] I have it all fully automated so I need to strip out the stuff you'll want to do it by hand
[22-Jul-2010 15:26:06] I have SSH baby
[22-Jul-2010 15:26:08] I have SSH
[22-Jul-2010 15:26:19] life is so much better with SSH
[22-Jul-2010 15:26:41] wtucker, no... didn't realize I needed to! How frequently does Zenoss remodel by default?
[22-Jul-2010 15:26:50] it doesn't
[22-Jul-2010 15:27:26] Oh, I thought it remodeled devices on its own every night or week...
[22-Jul-2010 15:27:31] I don't think so
[22-Jul-2010 15:27:38] Soemone else may be able to speak to that, but that's my understanding
[22-Jul-2010 15:27:46] theacolyte: http://fpaste.org/B1Ch/
[22-Jul-2010 15:27:58] obviously don't copy the part with *************
[22-Jul-2010 15:28:02] just everything below that
[22-Jul-2010 15:28:12] save it as pgtbl.h.patch
[22-Jul-2010 15:28:17] then issue the commands above
[22-Jul-2010 15:28:39] erm
[22-Jul-2010 15:28:45] ignore that echo line
[22-Jul-2010 15:28:48] forgot to remove it lol
[22-Jul-2010 15:28:51] it's part of my script
[22-Jul-2010 15:29:23] ok that makes most of my interfaces work
[22-Jul-2010 15:29:56] mducharme-laptop: if you have multiple interface templates that datapoint and threshold need to be added to each one
[22-Jul-2010 15:30:28] Der, I'm retarded. The events just needed to be cleared...
[22-Jul-2010 15:30:36] rmatte: so do I put "http://fpaste.org/B1Ch/" into pgtbl.h.patch?
[22-Jul-2010 15:30:54] ok cool
[22-Jul-2010 15:30:58] not the whole thing
[22-Jul-2010 15:30:59] yes, I just added it to another
[22-Jul-2010 15:31:06] just from "--- vmci-only.old/include/pgtbl.h 2009-10-21 03:43:27.000000000 +0200" down
[22-Jul-2010 15:31:09] one interface is 0 if it is operationally up and 255 if it is operationally down
[22-Jul-2010 15:31:16] the rest are commands that you need to issue to patch the stuff
[22-Jul-2010 15:31:18] comprende?
[22-Jul-2010 15:31:23] I had to change the threshold settings for that one
[22-Jul-2010 15:31:24] hahahahahaha
[22-Jul-2010 15:31:25] I'm kidding
[22-Jul-2010 15:31:28] :P
[22-Jul-2010 15:31:54] mducharme: seriously? I've never ever seen operational statuses like that
[22-Jul-2010 15:32:21] mducharme: in that case you'd set the max to 0
[22-Jul-2010 15:32:28] so when it exceeds 0 it triggers
[22-Jul-2010 15:33:34] rmatte these are weird devices
[22-Jul-2010 15:33:35] trango radio
[22-Jul-2010 15:33:44] not the same oid as the cisco operstatus
[22-Jul-2010 15:34:11] rmatte: worked perfectly
[22-Jul-2010 15:34:23] The vmci module loads perfectly into the running kernel.
[22-Jul-2010 15:36:13] ok I got everything working now
[22-Jul-2010 15:36:22] except, what to do for this one interface
[22-Jul-2010 15:36:30] the ethernet interface has no operational status
[22-Jul-2010 15:36:37] rmatte: Have I mentioned how much of a stud you are recently?
[22-Jul-2010 15:36:53] however, it is what the connectivity to the device runs through, so the device will be unreachable if the ethernet interface is down
[22-Jul-2010 15:37:22] so I'm not sure whether I really need to monitor operstatus with that
[22-Jul-2010 15:38:52] theacolyte: I try
[22-Jul-2010 15:39:39] mducharme-laptop: you're absolutely positive that it doesn't have operstatus?
[22-Jul-2010 15:39:47] I've never seen a device that doesn't
[22-Jul-2010 15:39:56] and I monitor a LOT of different types of interfaces
[22-Jul-2010 15:41:45] yes rmatte
[22-Jul-2010 15:41:51] I'm absolutely positive it does not have operstatus
[22-Jul-2010 15:42:19] no idea then, you might have to configure that device to trap to Zenoss on status changes then instead
[22-Jul-2010 15:42:21] * rmatte shrugs
[22-Jul-2010 15:42:32] it's stupid that it doesn't show the operstatus via snmp
[22-Jul-2010 15:42:37] right, but how do I detect if the device goes down
[22-Jul-2010 15:42:41] I will have to do a ping test
[22-Jul-2010 15:42:41] epic fail on the part of the vendor
[22-Jul-2010 15:42:57] all devices are pinged by default anyways
[22-Jul-2010 15:43:04] via zenping
[22-Jul-2010 15:43:13] I had to do a lot of programming in the modeler to get this device to work
[22-Jul-2010 15:43:19] so if that's the interface that you're actually hitting the device through it's a null issue
[22-Jul-2010 15:43:22] the su's are all in a nice table
[22-Jul-2010 15:43:34] but two interfaces are separate with totally different oids
[22-Jul-2010 15:43:42] right, standard OIDs
[22-Jul-2010 15:43:46] lol
[22-Jul-2010 15:43:59] and not even the same options for each one
[22-Jul-2010 15:44:06] they should
[22-Jul-2010 15:44:12] though that device sounds wonky
[22-Jul-2010 15:44:16] * rmatte shrugs
[22-Jul-2010 15:44:27] I'm used to Cisco, HP, Juniper type gear
[22-Jul-2010 15:44:34] which is generally no non-sense
[22-Jul-2010 15:44:56] if you want to have a look, there is a spreadsheet in http://www.trangobroadband.com/infobase/downloads/M900S_ap_1p1_MIB.zip
[22-Jul-2010 15:45:06] which contains a list of all oids
[22-Jul-2010 15:45:21] I don't really have time to look too deeply, need to get back to this modeler plugin that I'm trying to code
[22-Jul-2010 15:45:26] ok
[22-Jul-2010 15:45:27] which isn't working even though the code is fine
[22-Jul-2010 15:46:09] paste in a pastebin?
[22-Jul-2010 15:46:20] I've just done so much modeler coding I could probably figure it out lol
[22-Jul-2010 15:47:34] k one sec
[22-Jul-2010 15:47:54] now I just have to figure out how not to make the entire city go orange/yellow
[22-Jul-2010 15:47:58] http://fpaste.org/7Bze/
[22-Jul-2010 15:48:01] ignore the commented lines
[22-Jul-2010 15:48:05] just commented for testing
[22-Jul-2010 15:48:10] one customer is down, and the entire city is orange as a result
[22-Jul-2010 15:48:19] nice
[22-Jul-2010 15:48:30] you mean on the map?
[22-Jul-2010 15:48:57] the color of the dots on the map represent the highest severity event at each location
[22-Jul-2010 15:49:16] so if you have a bunch of customers all grouped under the same city organizer that's what you'll see
[22-Jul-2010 15:49:24] that's why I don't use sub-location organizers
[22-Jul-2010 15:49:30] I put all of the locations at the very top level
[22-Jul-2010 15:49:37] with the organizer name as the address of the location
[22-Jul-2010 15:49:52] I just added the log.info line recently
[22-Jul-2010 15:50:00] going to see if it gives me clues as to what's going on
[22-Jul-2010 15:50:31] is anything getting logged?
[22-Jul-2010 15:50:41] nope
[22-Jul-2010 15:50:42] no errors or anything
[22-Jul-2010 15:50:52] I'm going to add log lines in to help with checking things out
[22-Jul-2010 15:51:00] just getting around to trying that
[22-Jul-2010 15:51:10] I would first add something in at the beginning of the process function
[22-Jul-2010 15:51:19] to make sure that it's even running it
[22-Jul-2010 15:51:22] k, so that log.info line never kicked off
[22-Jul-2010 15:51:45] good idea
[22-Jul-2010 15:52:58] this is based off InterfaceAliasMap.py by the way
[22-Jul-2010 15:53:00] I just modified it
[22-Jul-2010 15:53:11] and that plugin works fine
[22-Jul-2010 15:56:21] oops, I didn't even have the plugin bound since I forgot that I was testing something else for a minute
[22-Jul-2010 15:56:23] I'll rebind it
[22-Jul-2010 15:59:31] you can use gettablemap to get only one oid?
[22-Jul-2010 16:00:12] ok...
[22-Jul-2010 16:00:12] so
[22-Jul-2010 16:00:18] it is running the process portion
[22-Jul-2010 16:00:27] it's just not making it through the ifs
[22-Jul-2010 16:00:33] now to figure out where it's dying off at
[22-Jul-2010 16:04:12] any guide for upgrading on a zenoss 5.5 system from zenoss 2.5.2 to 3.0 ?
[22-Jul-2010 16:04:26] will all the zenpacks work on the new version ?
[22-Jul-2010 16:04:33] it's making it to the second if but not the third, hmmm
[22-Jul-2010 16:04:43] zykes-: not all of them, but most
[22-Jul-2010 16:04:53] meaning ?
[22-Jul-2010 16:05:05] rmatte
[22-Jul-2010 16:05:08] well, some flat out won't work (some of the more complex ones)
[22-Jul-2010 16:05:12] put the if statement back in
[22-Jul-2010 16:05:29] k
[22-Jul-2010 16:05:36] if 'ifcatos' in results[1] and 'iftable' in results[1]:
[22-Jul-2010 16:05:41] you had an error in the if statement before
[22-Jul-2010 16:05:43] some, you'll have to remove the -py2.4 from the filename to get them to install
[22-Jul-2010 16:05:46] what type of installaer should i use ?
[22-Jul-2010 16:06:05] after the if put in a log command
[22-Jul-2010 16:06:19] b/c if the data is not getting into results the loop will not occur at all
[22-Jul-2010 16:06:24] mducharme-laptop: nah that wasn't an error, I just heavily edited the rest but haven't gotten around to changing that if statement yet
[22-Jul-2010 16:06:31] ok
[22-Jul-2010 16:06:40] mducharme-laptop that if statement isn't an issue, it would error out if it was
[22-Jul-2010 16:06:48] I realize that
[22-Jul-2010 16:06:57] so it makes it up to: for i_idx, iface in results[1]['iftable'].items():
[22-Jul-2010 16:07:05] but never seems to get past if a_idx == i_idx:
[22-Jul-2010 16:08:00] basically that should be comparing interface indexes to make sure that it's applying the right description to the right interface
[22-Jul-2010 16:08:09] not sure how it's not matching up
[22-Jul-2010 16:08:17] hmmm, unless....
[22-Jul-2010 16:10:40] crap, I don't think the indexes correspond
[22-Jul-2010 16:13:57] that would explain it :)
[22-Jul-2010 16:17:44] <|rt|> the command window in 3.x doesn't work with chrome
[22-Jul-2010 16:18:26] hmmm, how the heck am I going to get this to work with indexes that don't correspond, garrr
[22-Jul-2010 16:18:30] there has to be some way to link the indexes though
[22-Jul-2010 16:18:34] yeh
[22-Jul-2010 16:18:47] I'll have to look in to it before continuing
[22-Jul-2010 16:18:59] look at what's in the ifcatos table
[22-Jul-2010 16:19:07] maybe the interface number is one of the parameters in there
[22-Jul-2010 16:19:20] nah it's not, I already checked
[22-Jul-2010 16:19:34] actually hmmm
[22-Jul-2010 16:19:36] check this out: http://www.google.ca/
[22-Jul-2010 16:19:39] bah
[22-Jul-2010 16:19:44] message/12705
[22-Jul-2010 16:19:56] that's sort of what I'm loosely working from
[22-Jul-2010 16:20:08] oh wait
[22-Jul-2010 16:20:09] so it is
[22-Jul-2010 16:20:13] .1 is the index
[22-Jul-2010 16:20:32] hmmm
[22-Jul-2010 16:20:34] or not
[22-Jul-2010 16:21:00] oh no
[22-Jul-2010 16:21:02] it's .11
[22-Jul-2010 16:21:16] bingo
[22-Jul-2010 16:26:23] now I still need to figure out how the heck to get the process function to work properly
[22-Jul-2010 16:39:14] well, it's definitely pulling down the info ok...
[22-Jul-2010 16:41:25] I think I'm starting to get somewhere with this
[22-Jul-2010 17:08:13] hrm rmatte
[22-Jul-2010 17:08:21] what is happening is I have a few interfaces that are administratively down
[22-Jul-2010 17:08:38] those show as operationally down and put the device in error stage
[22-Jul-2010 17:08:53] then remodel so that Zenoss unmonitors them
[22-Jul-2010 17:09:02] it shouldn't even be monitoring those if they are admin down
[22-Jul-2010 17:09:09] but you need it to remodel to pick up on that
[22-Jul-2010 17:09:56] I remodeled, same thing
[22-Jul-2010 17:11:10] do I have to manually change monitoring to disabled when adminstatus is down?
[22-Jul-2010 17:11:40] are you absolutely sure that they are actually admin down and not just operationally down?
[22-Jul-2010 17:11:53] the interfaces are completely disabled on the device?
[22-Jul-2010 17:12:07] they are admin down
[22-Jul-2010 17:12:17] then yeh, you might have to change the monitoring manually
[22-Jul-2010 17:12:18] Operational Status Down Administrative Status Down
[22-Jul-2010 17:12:22] though I don't see why you'd have to
[22-Jul-2010 17:12:33] it usually works fine for me
[22-Jul-2010 17:12:39] anyways, just disable them by hand
[22-Jul-2010 17:12:41] and that'll be that
[22-Jul-2010 17:14:01] hrm
[22-Jul-2010 17:14:11] I finally got the modeler to set the descriptions, but it's overwriting whatever the alias modeler does so I just need to come up with a solution for that and then I'm good to go
[22-Jul-2010 17:14:11] this will be annoying when we are going in adding all of our devices
[22-Jul-2010 17:14:33] when you're adding your devices it should automatically pick up on what's admin down
[22-Jul-2010 17:14:36] it always does for me
[22-Jul-2010 17:14:37] to have to manually disable monitoring for the ports that are administratively disabled
[22-Jul-2010 17:14:45] if it doesn't pick up on it then it's not seeing it as truly being admin down
[22-Jul-2010 17:14:48] should I try maybe deleting the device and readding it to see?
[22-Jul-2010 17:14:57] worth a shot
[22-Jul-2010 17:15:09] though it should have done it the first time
[22-Jul-2010 17:17:28] anyways, I'm done for the day
[22-Jul-2010 17:17:32] I'll be back tomorrow
[22-Jul-2010 17:18:32] ok
[22-Jul-2010 17:24:41] ok, I'm going to have to figure this out tomorrow I guess
[22-Jul-2010 17:25:02] I just modeled a device, and two interfaces where the status is down are showing with green lights in the list
[22-Jul-2010 17:26:35] check with snmpwalk to make sure they are actually operationally down integerwise
[22-Jul-2010 17:26:46] that's what Zenoss checks for
[22-Jul-2010 17:26:53] if snmp is reporting the wrong thing then that's the problem
[22-Jul-2010 17:28:31] it is a non standard oid for the operational status
[22-Jul-2010 17:28:42] I took care of it through the actual modeler script
[22-Jul-2010 17:28:49] then that's the problem
[22-Jul-2010 17:28:59] zenmodeler is looking for the standard OID
[22-Jul-2010 17:29:25] k
[22-Jul-2010 17:29:32] I have the modeler taking the correct one
[22-Jul-2010 17:29:37] it isn't the modeler, it's something else
[22-Jul-2010 17:29:42] well if you "took care of it" and it's still not working, then it wasn't done correctly
[22-Jul-2010 17:29:45] the modeler is correctly putting the interface as down
[22-Jul-2010 17:29:56] there might be some other code that's specifically polling the status
[22-Jul-2010 17:30:04] yes.. that's what I'm thinking
[22-Jul-2010 17:30:07] you'd have to ask one of the devs
[22-Jul-2010 17:33:51] nice, now Zenoss suddenly just stopped even trying to use my plugin
[22-Jul-2010 17:34:35] oh I see why
[22-Jul-2010 17:34:41] because I didn't close a bracket :P
[22-Jul-2010 17:36:13] http://pastebin.com/kwFUj4yE
[22-Jul-2010 17:36:26] that's what I spent a few days doing
[22-Jul-2010 17:36:54] lines 157-166 are what set the adminstatus on those SU's
[22-Jul-2010 17:42:47] cool
[22-Jul-2010 17:44:03] took me forever to do
[22-Jul-2010 17:44:18] I had to manually set certain things
[22-Jul-2010 17:44:24] like the names of the eth0 and rf0 interfaces
[22-Jul-2010 17:45:04] and anything that wasn't alredy named, used the suid prefix with the id following it.. because no interface ids are available through snmp
[22-Jul-2010 17:45:53] I'm just working on getting it matching up the indexes properly right now
[22-Jul-2010 17:46:16] the loop of loops should cause it to loop through all possible combinations
[22-Jul-2010 17:46:27] yup
[22-Jul-2010 17:46:40] the problem is the way the plugin was originally designed doesn't quite jive
[22-Jul-2010 17:46:43] but I've almost got it fixed
[22-Jul-2010 17:47:28] yeah there has to be something else still pulling the adminstatus
[22-Jul-2010 17:48:27] it's all green.. yet those interfaces are administratively down.. and it's trying to graph the interface
[22-Jul-2010 17:50:39] I think I might finally have the indexes working properly
[22-Jul-2010 17:50:45] waiting for the model to finish to verify
[22-Jul-2010 17:51:04] then I just have to deal with the fact that it's overwriting what the InterfaceAliasMap plugin is down
[22-Jul-2010 17:51:06] doing*
[22-Jul-2010 17:51:19] so I either need to make it play nice with that plugin or just copy what that plugin does
[22-Jul-2010 17:51:31] I'll probably just make it copy it
[22-Jul-2010 17:51:46] won't be hard to do
[22-Jul-2010 17:53:15] bingo
[22-Jul-2010 17:53:23] got it working
[22-Jul-2010 17:53:48] I'll have to continue this tomorrow
[22-Jul-2010 17:53:49] cheers
[22-Jul-2010 17:54:34] ttyl, thanks!
[22-Jul-2010 19:57:55] forsberg is now known as fOrsberg
[23-Jul-2010 00:00:30] [disconnected at Fri Jul 23 00:00:30 2010]
[23-Jul-2010 00:00:30] [connected at Fri Jul 23 00:00:30 2010]
[23-Jul-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]
[23-Jul-2010 01:15:27] orange is now known as Guest68966
[23-Jul-2010 01:48:13] fOrsberg is now known as forsberg
[23-Jul-2010 03:44:19] forsberg is now known as fOrsberg
[23-Jul-2010 05:48:28] hello
[23-Jul-2010 05:48:46] in a machine hosting multiple VMs, would you put zenoss on the physical host, or on a separate VM?
[23-Jul-2010 06:03:29] depends on how much power you need
[23-Jul-2010 06:03:43] i run mine in a VM
[23-Jul-2010 06:10:54] well, what are the pros/cons?
[23-Jul-2010 06:11:09] to me, running it on the host sounds intuitive
[23-Jul-2010 06:11:16] but perhaps I am missing something
[23-Jul-2010 06:17:03] the same pro's/cons as running any production box in a vitural container
[23-Jul-2010 06:17:28] i have multiple monitoring boxes though so if any one goes down i can usually tell
[23-Jul-2010 06:17:53] but to be honest, i've had an issue w/ one of my VM hosts only once and that was a HDD firmware issue
[23-Jul-2010 08:50:19] hey all
[23-Jul-2010 08:51:57] hey
[23-Jul-2010 08:56:46] Do you guys know where I could find a list of the available methods for dmd/ZenUsers/ via a REST call ?
[23-Jul-2010 08:59:44] for the log, ZenUI3/browser/resources/js/deploy/zenoss-compiled-debug.js is broken. the function link returns nothing if passed link(null, "/something", null)
[23-Jul-2010 09:06:40] npmccallum: you got any hands in ZenUI3 and the ComponentGrid?
[23-Jul-2010 09:07:47] ah, just the man I was looking for
[23-Jul-2010 09:19:07] fOrsberg is now known as forsberg
[23-Jul-2010 09:56:16] can someone help me with this error TypeError: is not JSON serializable
[23-Jul-2010 10:08:25] morning
[23-Jul-2010 10:09:19] fragfutter: are you trying to stuff an object into a json document, vs maybe the name of the object or similar?
[23-Jul-2010 10:10:10] Simon4: i have an info object that needs to be transported by json. In the component grid it works, if i click the events page i get the error.
[23-Jul-2010 10:10:29] hmm
[23-Jul-2010 10:10:45] no idea then
[23-Jul-2010 10:13:43] now, only thing left to do
[23-Jul-2010 10:13:50] how do I get interfaces with adminStatus down to turn grey?
[23-Jul-2010 10:14:08] without having to manually shut off monitoring
[23-Jul-2010 10:14:22] mducharme-laptop: remodel?
[23-Jul-2010 10:14:35] this is a freshly modeled device
[23-Jul-2010 10:14:42] the adminstatus correctly shows as down in the interface
[23-Jul-2010 10:14:50] but the lights are both green in the interface list
[23-Jul-2010 10:19:18] obviously adminStatus is getting determined two ways
[23-Jul-2010 10:19:40] and the light is not turning grey because adminStatus on this device is a non standard OID
[23-Jul-2010 10:21:04] so my question is
[23-Jul-2010 10:21:06] how does it know what to check?
[23-Jul-2010 10:21:17] the modeler is correctly programmed to check the non-standard OID
[23-Jul-2010 10:21:39] because I did the modeler script for the device to properly detect the status
[23-Jul-2010 10:21:50] mducharme-laptop: ugh. what kind of device?
[23-Jul-2010 10:21:59] trango access point
[23-Jul-2010 10:22:05] a wireless access point
[23-Jul-2010 10:22:52] my modeler script works fine, and the interface status is always correct when you click on it to view the interface itself
[23-Jul-2010 10:23:11] but the interfaces list shows all interfaces with adminStatus down as green, as though they are all enabled
[23-Jul-2010 10:23:57] my guess is that there'd be something you'd have to do in the modeling script to tell it to not monitor the interface
[23-Jul-2010 10:27:47] good idea, maybe there's something in the main interfacemap script that handles that
[23-Jul-2010 10:27:57] b/c I had to make my own interfacemap script for this device
[23-Jul-2010 10:27:59] cgibbons: the details page of a component, can i tell it how to render?
[23-Jul-2010 10:28:12] tell it? through code?
[23-Jul-2010 10:29:12] cgibbons: my custom component now renders correctly in the componentGrid. I return schema.Entity objects and have a renderer in javascript that handles them. If i click on the details, i get empty fields for all fields that use a renderer in the grid
[23-Jul-2010 10:29:24] ok, there is nothing in InterfaceMap.py that handles that
[23-Jul-2010 10:29:36] so, what could the problem be
[23-Jul-2010 10:30:02] if I knew how zenoss internally determined whether to set those lights green or grey, it would be helpful
[23-Jul-2010 10:31:16] this is the script that I wrote to map the interfaces:
[23-Jul-2010 10:31:16] http://pastebin.com/kwFUj4yE
[23-Jul-2010 10:31:23] cgibbons: http://pastie.org/1057118 works for the grid, but has empty fields in the details page
[23-Jul-2010 10:31:28] it's a heavily modified version of the default interfacemap script
[23-Jul-2010 10:32:17] lines 157-166 handle setting the adminStatus and operStatus
[23-Jul-2010 10:32:55] mducharme-laptop: so the green/grey dot is filled by a function call that returns an image url based on the numeric value
[23-Jul-2010 10:33:13] the numeric value of what?
[23-Jul-2010 10:33:26] it would be worth checking that the value of operStatus and adminStatus are what you believe they should be by looking in zendmd
[23-Jul-2010 10:33:41] how do I look in zendmd?
[23-Jul-2010 10:33:55] lemme pastebin something, 2 secs
[23-Jul-2010 10:36:39] http://pastie.org/1057125
[23-Jul-2010 10:36:47] you run zendmd just off the commandline as the zenoss user
[23-Jul-2010 10:39:57] ok it's run
[23-Jul-2010 10:40:07] the output shows adminStatus for those two interfaces as 0
[23-Jul-2010 10:40:14] those two interfaces which are disabled
[23-Jul-2010 10:40:14] * Simon4 hunts down the function call, 2 secs
[23-Jul-2010 10:41:59] thanks for your help simon4
[23-Jul-2010 10:42:09] I would be lost if I had to look through the zenoss source
[23-Jul-2010 10:43:03] a ha
[23-Jul-2010 10:43:04] http://pastie.org/1057131
[23-Jul-2010 10:43:20] that's the function that the UI calls, you need to make your status's -1 to get grey :)
[23-Jul-2010 10:44:10] and there's nothing wrong with a bit of grep -R in the zenoss source ;)
[23-Jul-2010 10:45:13] hrm simon4
[23-Jul-2010 10:45:18] I don't quite buy that
[23-Jul-2010 10:45:29] what is status getting passed from?
[23-Jul-2010 10:47:32] mducharme-laptop: it's in the template for the os page, just finding the appropriate code
[23-Jul-2010 10:47:43] ok I just compared the values with my switch
[23-Jul-2010 10:47:47] which properly turns grey
[23-Jul-2010 10:47:55] the values in the switch are 2 for adminStatus and operStatus when it is down
[23-Jul-2010 10:48:35] * Simon4 checks something
[23-Jul-2010 10:49:21] I wish I had a document or something that said what the values of adminStatus and operStatus meant
[23-Jul-2010 10:49:42] if you go to the os detail of a device, then add /manage to the url
[23-Jul-2010 10:49:48] you can see the source of the page template
[23-Jul-2010 10:50:18] . o O (and the moment you upgrade to zenoss3, you are out of luck, javascript all over the house)
[23-Jul-2010 10:50:21] for interfaces
[23-Jul-2010 10:50:21] http://pastie.org/1057141
[23-Jul-2010 10:50:24] http://tools.cisco.com/Support/SNMP/do/BrowseOID.do?local=en&translate=Translate&objectInput=ifAdminStatus
[23-Jul-2010 10:51:04] fragfutter: having fun then?
[23-Jul-2010 10:51:07] so 2 is down..
[23-Jul-2010 10:51:13] that means I programmed this script wrong
[23-Jul-2010 10:51:17] i want tal, metal and all the nice stuff back
[23-Jul-2010 10:53:03] who can i ask about zenoss-compiled.js?
[23-Jul-2010 10:53:21] fragfutter: so both the zenpacks I have written don't work in 3, I'm not looking forward to getting my head around fixing them
[23-Jul-2010 10:56:08] yes!
[23-Jul-2010 10:56:12] setting those things to 2, it works now
[23-Jul-2010 10:56:13] mrayzenoss is now known as mray
[23-Jul-2010 10:56:16] this is excellent
[23-Jul-2010 10:56:20] ok now
[23-Jul-2010 10:56:23] there is one remaining problem
[23-Jul-2010 10:56:26] and this one is sorta weird
[23-Jul-2010 10:57:19] mray: can i ask you about zenoss-compiled.js?
[23-Jul-2010 10:58:28] Do you want to not use it so you can debug the js easier, ff?
[23-Jul-2010 10:58:47] cgibbons: no, i have an issue with the implemented link function
[23-Jul-2010 10:58:48] ahahhh
[23-Jul-2010 10:58:52] I'm not smoking crack
[23-Jul-2010 10:59:02] Yes you are
[23-Jul-2010 10:59:05] http://pastie.org/1057149
[23-Jul-2010 10:59:08] theacolyte: I wish ;)
[23-Jul-2010 10:59:12] hehe
[23-Jul-2010 10:59:20] shre the wealth, puff puff pass and all that
[23-Jul-2010 10:59:38] mducharme-laptop: so it is -1 when it hits the "what colour" function, there's just another function call in the way that translates the 2 from the operStatus
[23-Jul-2010 11:00:12] yeah simon that's what I figured
[23-Jul-2010 11:00:24] that's exactly what I figured was happening :)
[23-Jul-2010 11:00:34] fragfutter: I probably can't help you, cgibbons undoubtedly knows more
[23-Jul-2010 11:00:36] so this is good now, everything works, except one thing
[23-Jul-2010 11:01:00] and this one is weird
[23-Jul-2010 11:01:09] I developed the modeler in the zenpack directory
[23-Jul-2010 11:01:13] made changes every so often etc
[23-Jul-2010 11:01:22] debugging it until I got the final version of the modeler
[23-Jul-2010 11:01:51] now, here's the weird thing
[23-Jul-2010 11:01:57] whoops
[23-Jul-2010 11:02:07] I run the modeler and it runs the final one
[23-Jul-2010 11:02:22] the correct, finalized one
[23-Jul-2010 11:02:44] but when the modeler runs automatically on scheduled basis, however often that is, it's an older version of my modeler script
[23-Jul-2010 11:02:59] have you restarted zenmodeler?
[23-Jul-2010 11:03:02] so zenoss is caching the old modeler script somewhere
[23-Jul-2010 11:03:32] no, I didn't know I needed to, I will do that now
[23-Jul-2010 11:03:56] I would def try that, since that's what's running the scheduled modelling
[23-Jul-2010 11:04:44] fragfutter: ithink remoteDevice method must return not ip.getDeviceLink() but simply 'ip' (Device instance)
[23-Jul-2010 11:05:03] bigegor: yes. that part is done.
[23-Jul-2010 11:05:42] mducharme: you need to restart zenhub
[23-Jul-2010 11:05:47] bigegor: oh, forgot to fix the pastie.
[23-Jul-2010 11:05:51] zenhub is caching the modeler
[23-Jul-2010 11:05:56] bigegor: line 20.
[23-Jul-2010 11:06:02] http://pastie.org/1057155
[23-Jul-2010 11:06:34] fragfutter: line 20 is '#'
[23-Jul-2010 11:06:47] bigegor: late.... line 25
[23-Jul-2010 11:07:35] bigegor: the problem i have is, that the Info object returned has no name attribute. This is something i catched in my resource.js (line 30-37)
[23-Jul-2010 11:08:14] bigegor: but it doesn't help for the details page. So i need to learn why the info object has no name attribute.
[23-Jul-2010 11:08:15] fragfutter: you can redefine it
[23-Jul-2010 11:08:53] bigegor: redefine what?
[23-Jul-2010 11:09:00] all right, one other thing I noticed
[23-Jul-2010 11:09:07] how do I get the network map to display something?
[23-Jul-2010 11:09:32] fragfutter: name method in info.py like here: http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.HPEVAMon/ZenPacks/community/HPEVAMon/info.py
[23-Jul-2010 11:09:36] mducharme: make sure you've got it set to the right network
[23-Jul-2010 11:09:59] ie. "10.0.10.0" or whatever instead of the default
[23-Jul-2010 11:09:59] it says selected device or network
[23-Jul-2010 11:10:05] so I put in a device
[23-Jul-2010 11:10:16] I put in the IP of our core router
[23-Jul-2010 11:10:42] should that not display at the very least the core router?
[23-Jul-2010 11:11:03] fragfutter: method 'name' in class HPEVADiskDriveInfo
[23-Jul-2010 11:11:04] bigegor: but remoteDevice is a Zenoss Device object. it already has a name function
[23-Jul-2010 11:11:17] bigegor: i wonder why it is missing.
[23-Jul-2010 11:11:39] s/function/attribute/
[23-Jul-2010 11:11:56] I just tried specifying the network that my core router is on
[23-Jul-2010 11:12:12] instead of specifying the core router itself
[23-Jul-2010 11:12:14] still get just a white screen
[23-Jul-2010 11:12:55] empty map
[23-Jul-2010 11:12:58] gah, this is frustrating
[23-Jul-2010 11:13:30] I have the physical map appearing correctly (the google maps thing) but the network map shows nothing
[23-Jul-2010 11:13:50] bigegor: InstanceOfmyObject.remoteDevice().name() returns the correct name
[23-Jul-2010 11:14:23] I would normally think something is broken in zenoss here but
[23-Jul-2010 11:14:29] this is the vmware install
[23-Jul-2010 11:14:33] of 2.5.2
[23-Jul-2010 11:14:52] I would not think the vmware install would have network map broken
[23-Jul-2010 11:15:31] bigegor: also IInfo(InstanceOfMyObject).remoteDevice.name returns the correct name
[23-Jul-2010 11:18:02] fragfutter: try use Zenoss.render.default_uid_renderer as renderer
[23-Jul-2010 11:18:38] bigegor: in the grid it works fine with my renderer. Only the details page has no entry.
[23-Jul-2010 11:18:57] bigegor: because the details page ignores the renderer i specified
[23-Jul-2010 11:21:29] any ideas?
[23-Jul-2010 11:23:14] I just tried ie because someone said it was a flash player problem.. same thing happens though
[23-Jul-2010 11:24:40] fragfutter: i know schema.Entity used object ID as links name
[23-Jul-2010 11:26:45] ok well I'm upgrading my flash player to see if that makes any difference
[23-Jul-2010 11:27:19] the Details page is all done using this auto-form code in $ZENHOME/ZenUI3/browser/resource/js/zenoss/form - maybe LinkField.js would be what's actually being rendered
[23-Jul-2010 11:27:24] but yeah it doesn't use your component grid renderer
[23-Jul-2010 11:28:12] oh wait a second
[23-Jul-2010 11:28:13] there we go
[23-Jul-2010 11:35:24] cgibbons: as i can't specify my own renderer, i need to find out why a Device Info object i return has no name attribute.
[23-Jul-2010 11:44:42] is there no quick start guide for 3.0?
[23-Jul-2010 11:45:02] AwoL: it's coming soon
[23-Jul-2010 11:45:13] mray: oh okay, thanks!
[23-Jul-2010 11:47:34] is apache a prereq? it wasn't mentioned in the guide I followed
[23-Jul-2010 11:47:41] no
[23-Jul-2010 11:47:50] it's in the stack installer
[23-Jul-2010 11:47:54] everything you need is in it
[23-Jul-2010 11:48:17] ah cool - I'm assuming it's not running on port 80 then?
[23-Jul-2010 11:49:34] nope, 8080
[23-Jul-2010 11:49:46] great, thanks!
[23-Jul-2010 12:00:38] mducharme-laptop: select a network?
[23-Jul-2010 12:00:42] ups.
[23-Jul-2010 12:08:43] fragfutter I got it going
[23-Jul-2010 12:08:57] welp
[23-Jul-2010 12:09:06] there was a router that was not modeled in between my zenoss box and the core router
[23-Jul-2010 12:09:29] zenoss cannot generate a map unless all hops between it and the device specified are mapped
[23-Jul-2010 12:10:36] so
[23-Jul-2010 12:10:45] the change passsword feature is broken as far as I can tell in 3.0
[23-Jul-2010 12:10:50] is there a way to do it via zendmd?
[23-Jul-2010 12:11:12] zenpass?
[23-Jul-2010 12:12:01] ah didn't know there was a command
[23-Jul-2010 12:12:03] let me check that
[23-Jul-2010 12:13:06] that just seems to change "admin"
[23-Jul-2010 12:13:34] it's not working for other users?
[23-Jul-2010 12:13:39] what's the arg
[23-Jul-2010 12:13:54] I tried zenpass username and it says "Password for the Zenoss "admin" user:"
[23-Jul-2010 12:13:56] no arg, I just meant to clarify your problem
[23-Jul-2010 12:14:09] within the UI, you can't change passwords?
[23-Jul-2010 12:14:12] it does change admin
[23-Jul-2010 12:14:13] no
[23-Jul-2010 12:14:15] it does not work
[23-Jul-2010 12:14:24] hrm I'm getting errors on my zenoss box in monitoring
[23-Jul-2010 12:14:26] 127.0.0.1 sendto error Host localhost and zenoss.gobcn.ca are both using ip 127.0.0.1
[23-Jul-2010 12:14:50] I enter the new password, confirm it, then at the bottom enter it again
[23-Jul-2010 12:14:59] oh wait
[23-Jul-2010 12:15:06] * theacolyte tries something
[23-Jul-2010 12:15:30] ah, that worked - you know what it was - outlook and trailing spaces for copy & paste
[23-Jul-2010 12:15:34] I want to kill whoever did that
[23-Jul-2010 12:16:11] hrm what entry should I have in my hosts file on this zenoss server
[23-Jul-2010 12:16:14] to make that go away
[23-Jul-2010 12:16:21] right now I only have the one entry
[23-Jul-2010 12:16:41] 127.0.0.1 zenoss.gobcn.ca localhost.localdomain localhost zenoss
[23-Jul-2010 12:17:03] is it safe for me to remove localhost and localdomain from there?
[23-Jul-2010 12:17:17] that's my /etc/hosts file
[23-Jul-2010 12:26:45] gah I'm stupid
[23-Jul-2010 13:42:28] well, I had this collector plugin partially working, then I tried to add something else to it and it's not working as expected
[23-Jul-2010 13:42:44] it'll either do one thing or the other, not both
[23-Jul-2010 13:45:44] what's it doing rmatte?
[23-Jul-2010 13:45:56] is it that same one you were working on yesterday?
[23-Jul-2010 13:45:58] hi fellows
[23-Jul-2010 13:46:05] one sec, I'm just testing a couple of things
[23-Jul-2010 13:46:13] JonnyV: hey
[23-Jul-2010 13:47:22] mducharme-laptop: yeh same one, just getting back to working on it now...
[23-Jul-2010 13:47:36] I had it gathering the descriptions but not the proper interface names as well
[23-Jul-2010 13:47:45] so I'm working on getting it to successfully grab and apply both
[23-Jul-2010 13:48:53] Man I really do like the new interface a lot more
[23-Jul-2010 13:49:06] Now that I'm up to speed
[23-Jul-2010 13:49:09] it's also quite a bit quicker
[23-Jul-2010 13:51:35] theacolyte: yeh, it's not bad, I'm waiting for a few improvements that will be in 3.0.1
[23-Jul-2010 13:53:30] mducharme-laptop: http://fpaste.org/cXOF/
[23-Jul-2010 13:54:21] mducharme-laptop: I know for a fact that both of those sets of if statements are running through to completion
[23-Jul-2010 13:54:32] and I know for a fact that both of them work perfectly fine on their own
[23-Jul-2010 13:54:45] but as soon as I added the second GetTableMap it stopped working properly
[23-Jul-2010 13:54:59] it will only collect interface names, doesn't apply the descriptions for some reason
[23-Jul-2010 13:55:05] and I can't understand why it's doing that
[23-Jul-2010 13:55:32] maybe it's something in InterfaceMap.py
[23-Jul-2010 13:56:02] any thoughts?
[23-Jul-2010 13:57:24] any thoughts from anyone, for that matter?
[23-Jul-2010 14:02:39] sounds like something to debug.
[23-Jul-2010 14:02:57] that wasn't very useful so i was keeping it inside :)
[23-Jul-2010 14:03:02] :P
[23-Jul-2010 14:03:40] oh, hmmm
[23-Jul-2010 14:03:46] I think I see what it might be
[23-Jul-2010 14:05:48] anyone want to throw in their suggestions for prizes in the Summer of ZenPacks contest?
[23-Jul-2010 14:05:59] a private jet!
[23-Jul-2010 14:06:13] previous prizes were Dell minis, Chumbies and a QNAP NAS
[23-Jul-2010 14:06:21] netbook maybe?
[23-Jul-2010 14:07:19] I was going to say ipad, but I don't even like those things lol
[23-Jul-2010 14:07:33] GuruPlug? http://www.globalscaletechnologies.com/c-4-guruplugs.aspx
[23-Jul-2010 14:07:39] second gettablemap?
[23-Jul-2010 14:07:40] guruplugs are awesome
[23-Jul-2010 14:07:49] my friend has been ranting and raving about those for months
[23-Jul-2010 14:07:50] lol
[23-Jul-2010 14:07:53] what second GetTableMap?
[23-Jul-2010 14:07:54] not enough RAM for running Zenoss
[23-Jul-2010 14:08:09] mducharme-laptop: yeh, what's wrong with that?
[23-Jul-2010 14:08:22] oops missed your paste link
[23-Jul-2010 14:08:32] I was at the bank
[23-Jul-2010 14:08:34] lol
[23-Jul-2010 14:08:58] I made a change, testing to see if it fixed it
[23-Jul-2010 14:09:05] waiting on model to finish
[23-Jul-2010 14:10:01] so you're saying this works if you have just one or the other
[23-Jul-2010 14:10:05] but if you have both it doesn't work
[23-Jul-2010 14:10:36] correcto
[23-Jul-2010 14:10:48] and it always only works for the first GetTableMap
[23-Jul-2010 14:12:14] well, what I tried didn't work
[23-Jul-2010 14:12:15] hmmm
[23-Jul-2010 14:12:54] I think it has something to do with InterfaceMap.py itself
[23-Jul-2010 14:12:58] going to dig through it
[23-Jul-2010 14:13:38] why are you still pulling description from ifalias when you aren't using it?
[23-Jul-2010 14:14:03] I'm not pulling description from ifalias
[23-Jul-2010 14:14:11] I'm pulling ifName from ifalias
[23-Jul-2010 14:14:13] er
[23-Jul-2010 14:14:20] sorry, pulling description TO ifalias
[23-Jul-2010 14:14:36] I'm not
[23-Jul-2010 14:14:36] you're pulling the description field from the device into ifalias table
[23-Jul-2010 14:14:46] I've tested that description code, it works flawlessly
[23-Jul-2010 14:14:50] it does exactly what it's supposed to
[23-Jul-2010 14:15:22] the description gets pulled from the device in to ifcatos and then pushed in to the iftable table
[23-Jul-2010 14:15:26] it doesn't touch ifalias
[23-Jul-2010 14:15:29] ok, I just thought maybe it was just some glitch involving using "description" in two tables
[23-Jul-2010 14:15:40] I was looking at line 36 of your pastebin
[23-Jul-2010 14:15:41] no, it has to be that way to work
[23-Jul-2010 14:15:48] ok
[23-Jul-2010 14:15:55] oh wait, hmmm
[23-Jul-2010 14:16:03] sorry, I totally missed the point of what you were saying lol
[23-Jul-2010 14:16:07] one sec, I'll remove that
[23-Jul-2010 14:16:10] * rmatte facepalms
[23-Jul-2010 14:16:17] not sure if that'll make a difference though, doubt it
[23-Jul-2010 14:16:45] mray: Samsung i9000 (Android 2.1)
[23-Jul-2010 14:17:53] it will rmatte, I think
[23-Jul-2010 14:18:09] mducharme-laptop: k, guess we'll see lol
[23-Jul-2010 14:18:09] the reason is, if the table ifalias has description, InterfaceMap pulls description from ifalias automatically
[23-Jul-2010 14:18:22] I don't know the exact line number from interfacemap
[23-Jul-2010 14:18:25] but I'm looking at it right now
[23-Jul-2010 14:18:29] k
[23-Jul-2010 14:18:41] I'm just waiting for the model to finish to find out
[23-Jul-2010 14:18:43] makes sense
[23-Jul-2010 14:18:59] nope, that didn't help
[23-Jul-2010 14:19:05] still just set the names with no descriptions
[23-Jul-2010 14:19:30] ok, you'll have to cange InterfaceMap.py then
[23-Jul-2010 14:19:33] I will tell you which line
[23-Jul-2010 14:19:52] I don't want to change it though lol
[23-Jul-2010 14:20:02] iftable[ifidx]['description'] = data.get('description', '')
[23-Jul-2010 14:20:10] hang on a minute
[23-Jul-2010 14:20:14] I have a better idea
[23-Jul-2010 14:20:17] k
[23-Jul-2010 14:20:20] of how to work around this
[23-Jul-2010 14:20:25] cool
[23-Jul-2010 14:20:33] I don't want to touch any of the actual Zenoss plugins
[23-Jul-2010 14:20:38] you won't have to
[23-Jul-2010 14:20:41] k cool
[23-Jul-2010 14:20:44] store the value in the ifalias table
[23-Jul-2010 14:21:06] sounds like a plan, didn't know you could do that though
[23-Jul-2010 14:21:42] # results[1]['ifalias'][i_idx]['description'] = catos['description']
[23-Jul-2010 14:21:48] ah
[23-Jul-2010 14:21:51] I'll give it a whirl
[23-Jul-2010 14:21:52] if the indexes match you can just do that
[23-Jul-2010 14:22:01] they do match
[23-Jul-2010 14:22:03] otherwise you will need to change your loops to loop through the ifalias table instead
[23-Jul-2010 14:22:03] ok
[23-Jul-2010 14:22:06] that's all been tested
[23-Jul-2010 14:22:10] ok
[23-Jul-2010 14:22:45] k, testing...
[23-Jul-2010 14:23:41] wham
[23-Jul-2010 14:23:43] that worked
[23-Jul-2010 14:23:46] thanks
[23-Jul-2010 14:23:51] cool
[23-Jul-2010 14:24:00] nice to return some help after all the help I've gotten
[23-Jul-2010 14:24:03] lol
[23-Jul-2010 14:24:27] I think I'll rename it to InterfaceCatOsMap, the current name is bit long lol
[23-Jul-2010 14:26:09] you might also want to add a comment that explains why you are pulling description from catos into ifalias :)
[23-Jul-2010 14:27:52] hmmm, I renamed the class, the file, but it's still only showing the old name in the UI
[23-Jul-2010 14:28:09] restarting zenoss is probably in order
[23-Jul-2010 14:30:34] there we go, that did it
[23-Jul-2010 14:30:35] sweet
[23-Jul-2010 14:31:11] so glad it's working, thanks again
[23-Jul-2010 14:33:30] I wish zenoss core had built in ldap support
[23-Jul-2010 14:33:50] well, I still haven't figured out how to set it up in 3.0
[23-Jul-2010 14:33:54] but once I do I'll write a guide
[23-Jul-2010 14:34:10] it wasn't hard to setup in 2.5
[23-Jul-2010 14:34:22] we are still using 2.5
[23-Jul-2010 14:34:26] just need to figure out how to do it with the newer versions of the plugins
[23-Jul-2010 14:34:32] not sure whether we should move to 3.0 yet, worried about glitches
[23-Jul-2010 14:34:49] what distro are you running on?
[23-Jul-2010 14:34:58] running the vmware appliance
[23-Jul-2010 14:35:01] ah
[23-Jul-2010 14:35:03] which runs on rpath
[23-Jul-2010 14:35:12] yeh, I'm not a big fan of the appliance
[23-Jul-2010 14:35:20] is it possible to edit a datasource for a local template on 3.0 within the device ?
[23-Jul-2010 14:35:23] it's great for trying it out, but I wouldn't use it for production
[23-Jul-2010 14:35:35] the problem with the appliance is it's not 64-bit
[23-Jul-2010 14:35:50] at some point we might want to move zenoss to a 64-bit vm
[23-Jul-2010 14:35:54] all I get when clicking on "view and edit details" is a page to test the datasource !
[23-Jul-2010 14:35:55] dec3pti0n: everything that was possible in 2.5 should still be possible, so in short, yes
[23-Jul-2010 14:36:08] dec3pti0n: really? that's odd
[23-Jul-2010 14:36:39] mducharme-laptop: I've never seen ldap setup on an appliance, so I'm not sure how involved it would be
[23-Jul-2010 14:36:47] yeah it's wierd cause I also get nothing on the black screen when I model a device . it's just stuck doing nothing
[23-Jul-2010 14:37:17] mducharme-laptop: this is the best LDAP guide for 2.5: docs/DOC-2510
[23-Jul-2010 14:37:26] yeah I know I found that already
[23-Jul-2010 14:37:29] k
[23-Jul-2010 14:37:36] but it just looks to me like a hack
[23-Jul-2010 14:37:41] all my 2.5 boxes use ldap for auth
[23-Jul-2010 14:37:45] I'd be worried about upgrades breaking that
[23-Jul-2010 14:37:45] it's not a hack
[23-Jul-2010 14:37:52] it's a zope plugin
[23-Jul-2010 14:38:03] upgrades can break it, but it can always be fixed
[23-Jul-2010 14:38:10] just something you need to test before upgrading in production
[23-Jul-2010 14:38:44] that plugin is maintained alongside zope
[23-Jul-2010 14:39:05] I upgraded from 2.4 to 2.5 and it didn't break
[23-Jul-2010 14:39:10] but I know the 3.0 upgrade breaks it
[23-Jul-2010 14:39:23] newer version of zope, so newer version of plugin needed
[23-Jul-2010 14:39:31] but it looks like the plugin has changed substancially
[23-Jul-2010 14:42:00] ldap isn't a huge deal for us right now anyway
[23-Jul-2010 14:42:07] we only have a few users using it
[23-Jul-2010 14:42:09] eugh, my brother is practicing swing dancing upstairs, sounds like the house is going to collapse
[23-Jul-2010 14:42:18] * rmatte shakes fist
[23-Jul-2010 14:42:31] I really wish zenoss had integrated ticketing
[23-Jul-2010 14:42:33] ah
[23-Jul-2010 14:42:37] that would solve so many issues for me
[23-Jul-2010 14:42:40] lol
[23-Jul-2010 14:43:01] we developed ticket system integration for otrs, but we can't release it
[23-Jul-2010 14:43:04] why are things like BMC the only integrated monitoring/helpdesk system
[23-Jul-2010 14:43:11] * rmatte shrugs
[23-Jul-2010 14:43:28] generally because people know what ticketing system they want to use and do the integration themselves
[23-Jul-2010 14:43:31] your ticket system integration just uses email I imagine
[23-Jul-2010 14:43:36] nope
[23-Jul-2010 14:43:37] API
[23-Jul-2010 14:43:44] ahh so you're using the soap api
[23-Jul-2010 14:43:48] the tickets are generated via SOAP
[23-Jul-2010 14:44:02] yes, i was looking at coding that
[23-Jul-2010 14:44:12] there's a perl script that my manager wrote which generates the tickets
[23-Jul-2010 14:44:15] and I wrote this...
[23-Jul-2010 14:44:25] http://dmon.org/zenticket.html
[23-Jul-2010 14:45:07] ahh!!
[23-Jul-2010 14:45:14] that'e excellent
[23-Jul-2010 14:45:33] it behaves like any other zenoss daemon, can start/stop it from the UI and everything
[23-Jul-2010 14:46:12] right
[23-Jul-2010 14:46:22] why did your manager write a perl script to generate the tickets
[23-Jul-2010 14:46:31] you can use the soap api from python directly
[23-Jul-2010 14:47:00] yeh I know, we were originally just using the commands section of the event manager in Zenoss to kick off the perl script (he's more comfortable coding in perl)
[23-Jul-2010 14:47:12] right
[23-Jul-2010 14:47:15] but it only kicked off when an event count was 1, and was extremely tedious to configure
[23-Jul-2010 14:47:26] so I decided to write the daemon and use the existing script
[23-Jul-2010 14:47:32] can I use your script there and add on python based integration?
[23-Jul-2010 14:47:36] the ticket create script is really complex, I didn't want to have to recreate it
[23-Jul-2010 14:47:53] yeh, I can send you the actual ZenPack tomorrow
[23-Jul-2010 14:47:56] ermn
[23-Jul-2010 14:47:59] on Monday
[23-Jul-2010 14:47:59] rather
[23-Jul-2010 14:48:42] we heavily customized otrs though
[23-Jul-2010 14:48:42] ok cool
[23-Jul-2010 14:48:50] added fields to it, redesigned the queue view
[23-Jul-2010 14:49:08] do you use the ITIL version?
[23-Jul-2010 14:49:10] our ticket create script has all sorts of auto-correlation and stuff built in to it
[23-Jul-2010 14:49:13] or rather the ITSM version
[23-Jul-2010 14:49:20] yeh, we use ITSM modules
[23-Jul-2010 14:49:33] I actually just finished porting all our customizations from 2.3 to 2.4
[23-Jul-2010 14:49:38] so we'll be upgrading soon enough
[23-Jul-2010 14:49:43] yeah I was going to say, that would minimize the customization that you would have to do
[23-Jul-2010 14:49:49] yeh
[23-Jul-2010 14:49:54] the ITSM modules are godly
[23-Jul-2010 14:49:59] the new change system module is awesome
[23-Jul-2010 14:50:01] because ITSM has a lot more stuff in it
[23-Jul-2010 14:51:16] yeah
[23-Jul-2010 14:51:30] that's what i want to integrate somewhat with the zenoss cmdb
[23-Jul-2010 14:53:05] I do not need a lot of information
[23-Jul-2010 14:53:27] what I ideally want is each device from zenoss replicated into the OTRS DB
[23-Jul-2010 14:53:31] but not all the details
[23-Jul-2010 14:54:27] probably just name, ip address and a web url link
[23-Jul-2010 14:54:48] mducharme-laptop: http://i27.tinypic.com/bfhxth.jpg
[23-Jul-2010 14:54:57] an example of what we've done
[23-Jul-2010 14:56:08] rmatte, yes, that is exactly what i am doing
[23-Jul-2010 14:56:17] cool
[23-Jul-2010 14:56:27] otrs already has the capability of associating devices with tickets
[23-Jul-2010 14:56:32] that's why I want to sync the cmdb's
[23-Jul-2010 14:56:38] yeh
[23-Jul-2010 14:56:45] we didn't do it that way unfortunately lol
[23-Jul-2010 14:56:49] I would imagine that is what your customizations do
[23-Jul-2010 14:56:53] we have a seperate cmdb
[23-Jul-2010 14:57:08] what do you store for the devices themselves
[23-Jul-2010 14:57:16] in otrs
[23-Jul-2010 14:57:18] not very much I assume
[23-Jul-2010 14:57:19] nothing
[23-Jul-2010 14:57:27] we have links that go off to other things
[23-Jul-2010 14:57:38] like if you click on the device name in the queue it brings you right to the device in zenoss
[23-Jul-2010 14:57:41] stuff like that
[23-Jul-2010 14:57:52] there's links to go straight to relevant documentation for a given customer
[23-Jul-2010 14:58:20] there's an iframe if you open a ticket that gives you info on the device taken from the cmdb
[23-Jul-2010 14:59:52] rmatte yeah, that's pretty much what I'm planning on setting up, only using the cmdb
[23-Jul-2010 15:00:02] cool
[23-Jul-2010 15:00:11] what I'm going to do is prefix all the cmdb classes with ZENOSS::
[23-Jul-2010 15:00:14] for the imported ones
[23-Jul-2010 15:00:19] cool
[23-Jul-2010 15:00:27] that way I won't run into naming issues with zenoss -> otrs
[23-Jul-2010 15:00:31] yeh
[23-Jul-2010 15:00:56] we have a script that compares the names in Zenoss to the ones in OTRS and send an email report with any issues it finds
[23-Jul-2010 15:01:01] there's one thing I was curious about however
[23-Jul-2010 15:01:03] erm
[23-Jul-2010 15:01:06] to the ones in cmdb rather
[23-Jul-2010 15:01:29] ?
[23-Jul-2010 15:02:00] because your script calls a perl script to create the tickets, it has to store the config (which is likely I assume the soap url, username and password for otrs)
[23-Jul-2010 15:02:12] the config I see is in a config file
[23-Jul-2010 15:02:23] as it would have to be to be accessible by perl
[23-Jul-2010 15:02:37] is there a facility where zProperties could be added into a zenpack
[23-Jul-2010 15:02:56] no, everything is passed to the perl script via commandline options
[23-Jul-2010 15:02:57] so that the OTRS soap url, username and password could be set in the gui
[23-Jul-2010 15:03:09] perl doesn't actually see the config script of the daemon
[23-Jul-2010 15:03:14] ok
[23-Jul-2010 15:03:15] config file rather
[23-Jul-2010 15:03:31] so, is it possible to have parameters in a zenpack?
[23-Jul-2010 15:03:39] instead of bundling a config file
[23-Jul-2010 15:03:45] or manually creating one in etc
[23-Jul-2010 15:03:52] the config file isn't really "bundled"
[23-Jul-2010 15:04:05] a generic config file is generated the first time the daemon is run
[23-Jul-2010 15:04:16] or if it finds that a config file doesn't exist
[23-Jul-2010 15:04:31] right, but where does it pull the soap url, username and password from
[23-Jul-2010 15:04:42] the config data has to be stored somewhere
[23-Jul-2010 15:04:49] that's all harcoded in to the perl script that it's calling
[23-Jul-2010 15:04:53] it doesn't need any of that
[23-Jul-2010 15:05:00] right
[23-Jul-2010 15:05:09] and I'm asking if instead of hardcoding it, it's possible to have zproperties storing those
[23-Jul-2010 15:05:22] you could do that, yes
[23-Jul-2010 15:05:32] you'd have to modify the daemon to pull down those properties
[23-Jul-2010 15:05:35] which would be simple
[23-Jul-2010 15:05:46] how would you add the properties
[23-Jul-2010 15:05:54] can a zenpack have custom zproperties?
[23-Jul-2010 15:06:01] docs/DOC-2347
[23-Jul-2010 15:06:19] you'd add lines like that to the zenpack's __init__.py file
[23-Jul-2010 15:06:25] so they'd get added when the ZenPack is installed
[23-Jul-2010 15:06:42] ahh gotcha
[23-Jul-2010 15:06:50] ok, that helps a lot
[23-Jul-2010 15:07:10] only thing that doesn't include - the zproperties would not be for a device or event, they would be global
[23-Jul-2010 15:07:21] there are some ZenPacks out there that provide zProperties, you could just grab one and jack their code
[23-Jul-2010 15:07:26] the soap url and username and pasword would be the same for the ticketing system regardless of device
[23-Jul-2010 15:07:30] I can't think of any off-hand
[23-Jul-2010 15:07:33] mray might know
[23-Jul-2010 15:07:33] yeah I was trying to find a zenpack that did that
[23-Jul-2010 15:07:47] b/c I already had that thought
[23-Jul-2010 15:10:26] yeh
[23-Jul-2010 15:10:43] I'm trying to think of a pack that has zProperties
[23-Jul-2010 15:10:45] hmmm
[23-Jul-2010 15:18:15] yeh can't think of any, I know there are some though
[23-Jul-2010 15:18:22] ask mray next time you see him he'd probably know
[23-Jul-2010 15:18:36] seeing as he processes all zenpacks that are submitted
[23-Jul-2010 15:58:36] guys a question regarding alerting...
[23-Jul-2010 15:59:14] is it possible for an event to alert to several email lists like escalating an issue?
[23-Jul-2010 15:59:33] lets say a host is down and i want to alert windows admins
[23-Jul-2010 15:59:35] but
[23-Jul-2010 15:59:46] it's definitely possible, you'd need to setup multiple alerting rules
[23-Jul-2010 15:59:51] and setup different filters on them
[23-Jul-2010 16:00:03] if the host remains down for an hour alert another one
[23-Jul-2010 16:00:07] ohh..cool
[23-Jul-2010 16:00:37] i will try it then this weekend.. im moving from a nagios setup to this one...
[23-Jul-2010 16:00:42] cool
[23-Jul-2010 16:00:47] yeh, Zenoss is quite nice
[23-Jul-2010 16:00:49] its really great this tool
[23-Jul-2010 16:00:55] yup
[23-Jul-2010 16:00:56] very
[23-Jul-2010 16:00:58] yeps.. lotsss of features
[23-Jul-2010 16:01:01] hehe
[23-Jul-2010 16:02:01] :)
[23-Jul-2010 16:02:24] the best thing is to see people come here over and over again, just like i did mysel, with the same reaction
[23-Jul-2010 16:02:31] viva zenoss! \o/
[23-Jul-2010 16:02:37] hahah
[23-Jul-2010 16:02:44] yepss.. im really excitedd.. hehe
[23-Jul-2010 16:02:52] not best thing maybe, but really unique :P
[23-Jul-2010 16:03:05] like a kid with a new toy
[23-Jul-2010 16:03:12] hehe
[23-Jul-2010 16:03:12] yeahs :P
[23-Jul-2010 16:04:22] im reading quite a few links from zenoss site... im starting with docs/DOC-5974
[23-Jul-2010 16:04:34] do u guys have a recommendation for new users to read?
[23-Jul-2010 16:04:59] also reading the faq
[23-Jul-2010 16:05:01] I'd recommend reading the Zenoss Admin Guide
[23-Jul-2010 16:05:06] top to bottom
[23-Jul-2010 16:05:41] http://downloads.sourceforge.net/zenoss/Zenoss_Administration_06-072010-3.0-v01.pdf
[23-Jul-2010 16:06:24] ok... thanks.. will start with that.. thanks
[23-Jul-2010 16:06:28] np
[23-Jul-2010 16:11:15] cool.. just what i asked.. heeh.. 2.6.3. Escalation of Alerting
[23-Jul-2010 16:11:33] yup
[23-Jul-2010 16:11:39] that covers pretty much everything
[23-Jul-2010 16:11:43] well, all the basics anyways
[23-Jul-2010 16:19:40] great
[23-Jul-2010 16:34:44] sideone is now known as Ragenix
[23-Jul-2010 16:34:51] Ragenix is now known as sideone
[23-Jul-2010 16:36:43] hurm
[23-Jul-2010 16:36:50] indeed
[23-Jul-2010 16:42:15] hi, I am trying to set up a jmx datasource. Do I need to put anything in the "Attribute and Operation Configuration" Section in the datasource definition?
[23-Jul-2010 16:43:56] I've never used the jmx pack...
[23-Jul-2010 16:43:58] docs/DOC-5528#comment-2127
[23-Jul-2010 16:44:02] perhaps that helps?
[23-Jul-2010 16:46:31] rmatte, that is what I am working from but not giving me enough info to help me figure out if what I put in is correct and why I am getting an error
[23-Jul-2010 17:10:53] hello all
[23-Jul-2010 17:11:25] i can't find any help regarding SNMP elsewhere so I come here, if no one can help thats fine.. just hear it out :)
[23-Jul-2010 17:11:36] I patched my net-snmp source so that tmpfs will show up in snmp. However, when I run an snmpwalk, it only shows the default /dev/shm and not my custom mount point which is also a tmpfs. Any ideas where to start troubleshooting?
[23-Jul-2010 17:42:14] QubeZ: you'd be better off trying in #linux or something
[23-Jul-2010 17:42:18] I honestly have no idea
[23-Jul-2010 19:57:55] forsberg is now known as fOrsberg
[23-Jul-2010 20:29:16] anyone know how i go about adding new component types in 3.0?
[24-Jul-2010 00:00:30] [disconnected at Sat Jul 24 00:00:30 2010]
[24-Jul-2010 00:00:31] [connected at Sat Jul 24 00:00:31 2010]
[24-Jul-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]
[24-Jul-2010 02:23:35] jassinpain_ is now known as jassinpain
[24-Jul-2010 03:33:46] fOrsberg is now known as forsberg
[24-Jul-2010 03:36:09] realitygeoff_ is now known as realitygeoff
[24-Jul-2010 11:55:16] forsberg is now known as fOrsberg
[24-Jul-2010 12:50:24] hi guys
[24-Jul-2010 13:50:08] zProdStateThreshold says Production state threshold at which Zenoss will begin to monitor a device.
[24-Jul-2010 13:50:19] i dont understand that
[24-Jul-2010 13:50:33] i think i will have to read production states info
[24-Jul-2010 14:35:05] what do u guys recommend for modelling win servers? snmp or wmi? im using core 3.0
[24-Jul-2010 14:52:49] wmi
[24-Jul-2010 14:56:04] thanks theacolyte
[24-Jul-2010 14:56:47] will look for the doc to implement it now..
[24-Jul-2010 16:59:09] fOrsberg is now known as forsberg
[24-Jul-2010 20:30:00] forsberg is now known as fOrsberg
[24-Jul-2010 21:35:47] Guest68966 is now known as orange
[25-Jul-2010 00:00:30] [disconnected at Sun Jul 25 00:00:30 2010]
[25-Jul-2010 00:00:31] [connected at Sun Jul 25 00:00:31 2010]
[25-Jul-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]
[25-Jul-2010 00:40:26] So the time on my Zenoss (2.5) server is 22:39. But when I log a comment on an event, it's timestamp is 15:39
[25-Jul-2010 00:40:49] Can anyone comment on this disparity, and ways to correct?
[25-Jul-2010 00:46:02] (note, I and the server are in PST timezone; 15:39 is afaik a time that occurs in the past - literally before the event I am commenting on!
[25-Jul-2010 00:46:34] eg, there is no place in the world where it's 15:39 on Saturday 7/24
[25-Jul-2010 02:11:55] hello all
[25-Jul-2010 03:03:22] fOrsberg is now known as forsberg
[25-Jul-2010 04:07:27] forsberg is now known as fOrsberg
[25-Jul-2010 18:29:24] Hey I've got some, I guess you could say "freedom" concerns... I understand Zenoss is OSS but these days that means very little... Would you say Zenoss gives as many features to users of the community edition as it does the enterprise version?
[25-Jul-2010 18:30:31] well, we just decided to go with the enterprise version but really the only difference are some nice premade modules
[25-Jul-2010 18:30:42] you could, i believe, build exactly the same stuff yourself on the community version
[25-Jul-2010 18:31:12] which modules?
[25-Jul-2010 18:31:42] they have a really nice vmware module, which is one of the major attractions to me
[25-Jul-2010 18:31:53] that has all kinds of bundled reports and so forth, there's a netapp one, some others
[25-Jul-2010 18:32:13] so access to a module repository is granted in the enterprise edition
[25-Jul-2010 18:32:16] but all modules are OSS ?
[25-Jul-2010 18:32:46] i would think the enterprise modules are not OSS, but the community ones are.. I've only just finished doing a brief eval of zenoss so I'm no expert
[25-Jul-2010 18:33:21] i'll be honest, i've been doing linux stuff for over ten years now and I just don't care about OSS vs other licences in any way, so I can't answer most of this
[25-Jul-2010 18:34:18] I care about companies that create a community edition and make parts of their code OSS...simply to gain new business and not to offer any real support for the community at all
[25-Jul-2010 18:34:34] basically the new "Freeware" ... but the community helps them write the code for them
[25-Jul-2010 18:35:13] I guess I'd say, "hey, some OSS code is better than no OSS code and nobody makes the community write or do anything unless it benefits them too"
[25-Jul-2010 18:35:20] i'm in favour of everyone being content :)
[25-Jul-2010 18:36:54] hi guys
[25-Jul-2010 18:37:16] im having trouble while trying to create a bakcup from the GUI
[25-Jul-2010 18:37:22] running 3.0 core
[25-Jul-2010 18:38:04] An error was encountered while publishing this resource.
[25-Jul-2010 18:38:04] (, 'error_message')Traceback (innermost last): Module ZPublisher.Publish, line 231, in publish_module_standard Module ZP
[25-Jul-2010 18:38:13] i can pastebin if u want the whole error
[25-Jul-2010 18:39:05] i really was thinking maybe a user permission error.. im manager in the roles settings
[25-Jul-2010 18:44:24] ashp, Do you reckon that the enterprise zenpacks... are partly enterprise because they include proprietary bits on how to monitor the things they do that can't be disclosed?
[25-Jul-2010 18:44:35] thus a good reason for not making them open source
[25-Jul-2010 18:47:24] Lanta: No idea, considering they are written in python I doubt there's anything proprietary
[25-Jul-2010 18:47:33] the netapp module still grabs info via snmp for example
[25-Jul-2010 18:47:48] i think they are just modules with extra effort put into them as a nice bonus for enterprise users :)
[25-Jul-2010 19:26:55] it was just perms..
[25-Jul-2010 19:27:01] as admin i can do it alright
[26-Jul-2010 00:00:30] [disconnected at Mon Jul 26 00:00:30 2010]
[26-Jul-2010 00:00:30] [connected at Mon Jul 26 00:00:30 2010]
[26-Jul-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]
[26-Jul-2010 05:12:35] Lanta is now known as Lantizia
[26-Jul-2010 06:05:20] hmmmm after upgrade I only get a blank ajaxy window but doesnt model, what log should I look for debug (cant find anything in the modeler log)
[26-Jul-2010 06:23:49] hello
[26-Jul-2010 06:24:09] zenoss automatically does an snmpwalk when the device is modeled,
[26-Jul-2010 06:24:26] however some statistics arent discovered, even though I can manually see them using snmpwalk
[26-Jul-2010 06:24:46] is it possible to add specific OIDs to zenoss and make a graph out of it?
[26-Jul-2010 06:25:22] dv_: yes, you want to make a template to describe the OIDs, and also define the graphs
[26-Jul-2010 06:25:32] I see.
[26-Jul-2010 06:25:36] then you bind that template to the device class containing the devices you care about
[26-Jul-2010 06:25:40] I wonder if there is a template for this already
[26-Jul-2010 06:25:45] it is haproxy data
[26-Jul-2010 06:25:50] the admin guide has details for doing this I believe
[26-Jul-2010 06:26:18] the other option is have a look in the zenpacks section of the site, and see if someone has made a zenpack for the data you want to monitor
[26-Jul-2010 07:13:19] Simon4_ is now known as Simon4
[26-Jul-2010 08:06:42] Simon4: do you use 3.0 ?
[26-Jul-2010 08:07:00] tehhobbit: no, still hanging in on 2.5.2
[26-Jul-2010 08:08:40] * tehhobbit wishes he had done that
[26-Jul-2010 08:09:20] the hpmon zenpack doesnt seem to be able to find a certain key
[26-Jul-2010 08:09:21] KeyError: 'setHWProductKey'
[26-Jul-2010 08:09:22] KeyError: 'setHWProductKey'
[26-Jul-2010 08:09:43] can I somehow check if its valid / changed name
[26-Jul-2010 08:10:06] or is it just 3.0 that doesnt want to set it to none ?
[26-Jul-2010 08:11:26] hmm
[26-Jul-2010 08:11:29] on a 3.0 test box
[26-Jul-2010 08:11:30] >>> dev.setHWProductKey('moo')
[26-Jul-2010 08:11:31] works
[26-Jul-2010 08:11:37] dev.setHWProductKey(None)
[26-Jul-2010 08:11:38] also works
[26-Jul-2010 08:11:52] (after filling dev up with a device object)
[26-Jul-2010 08:15:57] its only on devs that dont always answer propperly (old dl380 g3)
[26-Jul-2010 08:37:43] dear zenoss, feel free to delay zenoss enterprise because until it's released I have no real work to do :)
[26-Jul-2010 08:37:49] and that means i can just listen to music and relax
[26-Jul-2010 08:39:54] you could try 3.0 on a labserver
[26-Jul-2010 08:43:01] that would imply work!
[26-Jul-2010 08:52:09] * Simon4 is playing "find the debug events and related templates"
[26-Jul-2010 08:52:12] zendmd scripting fun
[26-Jul-2010 09:27:51] Simon4: and related? hrm..
[26-Jul-2010 09:28:52] frozty_sa: so I'm walking the debug events on a device, then trying to trace em to a template, and then making a decision based on what software is on the device as to whether the template should really be bound
[26-Jul-2010 09:32:15] hehe
[26-Jul-2010 09:32:33] Simon4: well, at least yours is based only on "should this template be bound?" fun..
[26-Jul-2010 09:32:43] * frozty_sa has to play syslog matchup the whole time
[26-Jul-2010 09:32:53] and write so many many event transforms...
[26-Jul-2010 09:33:03] but damn, <3 event transforms. really really <3 'em
[26-Jul-2010 09:33:05] yeah, we haven't tried syslogging to zenoss yet
[26-Jul-2010 09:33:16] I'm pretty sure our install would explode in a cloud of mysql inserts
[26-Jul-2010 09:33:58] Simon4: depends on what you send, I syslog ~240 devs doesnt cause that much load
[26-Jul-2010 09:34:15] Simon4: it's a bit iffy
[26-Jul-2010 09:34:20] we have some applications that have been known to abuse syslog before
[26-Jul-2010 09:34:24] mostly because syslogs are shit
[26-Jul-2010 09:34:54] I need to go patch the syslog parsers though, so that I can relay messages at them from syslog-ng
[26-Jul-2010 09:35:04] because otherwise relaying stuff makes it all show up as localhost
[26-Jul-2010 09:35:20] (and sinking syslog straight into zensyslog has the chance of blackholing stuff)
[26-Jul-2010 09:35:26] *nod*
[26-Jul-2010 09:37:15] I could probably actually contribute that as a patch to zenoss, but I'll have to find out..
[26-Jul-2010 09:39:32] frozty_sa: uuuh I find the general unix syslog messages quite reliable and clear
[26-Jul-2010 09:40:18] tehhobbit: yeah, I don't just have unix syslogs to deal with ;)
[26-Jul-2010 09:40:55] I've got on average 7 different vendors worth of network gear pumping info to, and info being pumped from into, my zenoss instances
[26-Jul-2010 09:41:07] frozty_sa: aah tried sorting out the mess that is windows eventlog that wasnt fun at all
[26-Jul-2010 09:41:20] thankfully, I don't even really have to care about that
[26-Jul-2010 09:54:48] * frozty_sa O_o's at not seeming to find anything that makes use of $ZENHOME/Products/ZenEvents/zensendsyslog.py
[26-Jul-2010 10:15:45] good morning all
[26-Jul-2010 10:16:12] good morning
[26-Jul-2010 10:16:12] frozty_sa: that could just be there for testing purposes
[26-Jul-2010 10:25:43] rmatte: *shrug*
[26-Jul-2010 10:26:18] that does seem like a plausible reason, I would say
[26-Jul-2010 10:28:29] hi everyone, I'm getting networking issues, so my ping fail often. I would like to check a log where I can look at the timeout so I can tune on my zenping. Where could I search for this?
[26-Jul-2010 10:32:01] zenping.log, I found it, thanks :D
[26-Jul-2010 10:32:10] anytime ;)
[26-Jul-2010 10:33:07] looking over all the code for zensyslog, I'm left thinking it's probably safer to add a second mode of operation. the first is the normal zensyslog sink mode, the other is to take messages in from a relaying syslog host
[26-Jul-2010 10:33:57] it's a relatively trivial code change in terms of how to parse each of those, but the tricky part comes in if you need to decide if a message is one or the other
[26-Jul-2010 10:34:22] reckon I should maybe ask this question on zenoss-dev?
[26-Jul-2010 10:34:54] zenoss-dev is usually empty except on thursdays
[26-Jul-2010 10:35:20] #zenoss-dev, or zenoss-dev?
[26-Jul-2010 10:35:26] oh, the channel I meant
[26-Jul-2010 10:35:29] (I was thinking of the mailing list/forum)
[26-Jul-2010 10:35:30] you can ask in the forums section
[26-Jul-2010 10:45:32] I only get this: WARNING zen.ZenPing: ip 1.2.3.4 is down
[26-Jul-2010 10:46:03] but I don't get timeout or any information else, could I extend this log?
[26-Jul-2010 10:46:14] how*
[26-Jul-2010 11:00:22] chachan: it's not actually doing a "ping" like you'd do from the commandline
[26-Jul-2010 11:00:30] so you're not going to get TTL info and such
[26-Jul-2010 11:00:43] it's simply telling you that the device is unreachable via ping
[26-Jul-2010 11:00:55] If you need more detailed info use the fping ZenPack
[26-Jul-2010 11:01:21] I don't personally see the point though
[26-Jul-2010 11:04:31] If you need debug info just run it from the commandline like so: zenping run -v10
[26-Jul-2010 11:04:50] (as the zenoss user)
[26-Jul-2010 11:13:25] ok now
[26-Jul-2010 11:15:25] hey rmatte
[26-Jul-2010 11:15:32] you said you could send me a zenpack for the zenticket daemon?
[26-Jul-2010 11:15:56] also who was that guy you said I should ask for which zenpacks have zproperties in them
[26-Jul-2010 11:16:07] there was a guy who all zenpacks go through
[26-Jul-2010 11:18:59] Matt Ray
[26-Jul-2010 11:19:01] if you see him
[26-Jul-2010 11:19:06] he's the Zenoss community manager
[26-Jul-2010 11:19:15] I'm heading to lunch in a minute, but I'll send it to you after
[26-Jul-2010 11:19:41] on the condition that you send me your completed pack when you're done the modifications, if that's cool?
[26-Jul-2010 11:20:07] I need to get something back to justify sending it to you lol
[26-Jul-2010 11:20:29] (Just so I have something to tell management if they were to find out)
[26-Jul-2010 11:21:00] anyways, I'm off to lunch, bbiab
[26-Jul-2010 11:34:32] ok, yes of course I would sen dyou the completed pack when I'm done
[26-Jul-2010 11:34:42] what I want to do is get the pack with parameters and instructions
[26-Jul-2010 11:34:48] so that it can be posted up on the community zenpacks site
[26-Jul-2010 11:36:58] mrayzenoss is now known as mray
[26-Jul-2010 11:40:13] hey mray
[26-Jul-2010 11:40:30] greetings
[26-Jul-2010 11:40:32] I had a question for you
[26-Jul-2010 11:40:55] I want to write a zenpack that includes zproperties in it, to make it easier to set certain parameters for the zenpack
[26-Jul-2010 11:41:05] rmatte said that you might know of an existing zenpack that includes zproperties
[26-Jul-2010 11:41:12] so that I will have an example for how it is done
[26-Jul-2010 11:41:24] sure thing, it's simple
[26-Jul-2010 11:41:34] relatively :)
[26-Jul-2010 11:41:43] the zproperties I would have would be system-wide
[26-Jul-2010 11:41:51] meaning the would not be device zproperties or whatever
[26-Jul-2010 11:42:03] yeah, zProperties are defined system-wide
[26-Jul-2010 11:42:08] ok
[26-Jul-2010 11:42:13] that's what I want then
[26-Jul-2010 11:42:23] http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.ChefClient/ZenPacks/community/ChefClient/__init__.py
[26-Jul-2010 11:42:32] the zChefRole is defined here
[26-Jul-2010 11:42:44] you can use strings, lists, ints
[26-Jul-2010 11:42:45] need a quick answer on hardware requirements
[26-Jul-2010 11:43:14] I for a collector monitoring 1000 hosts
[26-Jul-2010 11:43:28] how do I download the entire zenpack?
[26-Jul-2010 11:43:35] I'd like to see how it is used
[26-Jul-2010 11:44:00] because this shows me how to set it, but not how to access it
[26-Jul-2010 11:44:24] mducharme: ok, I'm not using that zProperty yet with the Chef ZenPack
[26-Jul-2010 11:44:24] er it shows me how to create/set it but not how to access it
[26-Jul-2010 11:44:34] let me find another example
[26-Jul-2010 11:44:34] ok
[26-Jul-2010 11:44:37] ok...
[26-Jul-2010 11:44:40] thanks!
[26-Jul-2010 11:45:22] mducharme: http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.Cfengine/ZenPacks/community/Cfengine/CfengineClient.py
[26-Jul-2010 11:45:32] that shows setting zDeviceTemplates, which is a property
[26-Jul-2010 11:45:41] setZenProperty
[26-Jul-2010 11:46:15] http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.Cfengine/ZenPacks/community/Cfengine/__init__.py
[26-Jul-2010 11:46:23] sets zCfengineComplianceFile
[26-Jul-2010 11:46:50] which is used in http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.Cfengine/ZenPacks/community/Cfengine/modeler/plugins/community/cfenginemodeler.py
[26-Jul-2010 11:46:54] hi guys
[26-Jul-2010 11:47:03] ok cool, thanks mray
[26-Jul-2010 11:47:08] one other question, what is that middle parameter used for
[26-Jul-2010 11:47:09] '/tmp/cfengine-clients.txt'
[26-Jul-2010 11:47:16] in your chefclient one it is blank
[26-Jul-2010 11:47:21] jdross: the new Installation Guide for 3.0 has some hardware recommendations in Chatper 1
[26-Jul-2010 11:47:28] mducharme: that's the default
[26-Jul-2010 11:47:52] ahh! ok
[26-Jul-2010 11:48:01] gotcha, thanks!
[26-Jul-2010 11:48:09] mray: yeah, for single server installations, im going to be doing a distributed install
[26-Jul-2010 11:48:12] mducharme: more docs on buliding/using zenpacks here: community/developers/zenpack_development
[26-Jul-2010 11:48:25] jdross: the distributed servers will have similar specs
[26-Jul-2010 11:48:41] mray yeah I've already made a zenpack for the trango radios that we use
[26-Jul-2010 11:48:42] To show locations on map they must be under locations and cant be under sub locations?
[26-Jul-2010 11:49:05] JonnyV: the map shows the top-level, so you can drill-down to see other levels
[26-Jul-2010 11:49:14] JonnyV: ie. top-level == USA
[26-Jul-2010 11:49:21] sub-level == Texas, California
[26-Jul-2010 11:49:27] from the top-level all you see is USA
[26-Jul-2010 11:49:36] exactly.. that was going to be my example
[26-Jul-2010 11:49:37] click on it and you see the sub-organizers
[26-Jul-2010 11:49:41] ok.. thanks mray
[26-Jul-2010 11:49:54] and you have to have an address at each level
[26-Jul-2010 11:50:02] I will be posting the trango zenpack on the community site as soon as I make sure certain things are working
[26-Jul-2010 11:50:11] yes.. they worked when i put them as top level
[26-Jul-2010 11:50:57] mducharme: email me when you're looking to post it and I'll help you out mray@zenoss.com
[26-Jul-2010 11:52:38] i think i will put all locations as top level so they all show in the map..
[26-Jul-2010 11:52:51] at the same time
[26-Jul-2010 11:53:48] rmatte, thanks :)
[26-Jul-2010 11:56:57] mray, thanks!
[26-Jul-2010 12:17:45] ok, this may be a stupid question but
[26-Jul-2010 12:17:53] how do I get zenoss to talk to a device using snmp v2 instead of v1?
[26-Jul-2010 12:22:43] nevermind, figured it out
[26-Jul-2010 12:22:50] hehe
[26-Jul-2010 12:23:03] <|rt|> has anyone had any luck with the printertoner zenpack with 3.x?
[26-Jul-2010 12:23:34] <|rt|> zenhub is throwing an error here in the user supplied python expression for the min value on the thresholds
[26-Jul-2010 12:56:46] mducharme-laptop: privmsg me your email
[26-Jul-2010 13:18:22] hurm
[26-Jul-2010 13:38:09] guys, is there a doc where i can lear a bit about ldap and roels and groups in zenoss?
[26-Jul-2010 13:38:25] i read that only enterprise is quite granular on permissions
[26-Jul-2010 13:38:38] yeh, core is not very granular
[26-Jul-2010 13:38:44] and there isn't really any good document covering that
[26-Jul-2010 13:38:59] you're really stuck with just users or admins
[26-Jul-2010 13:39:06] yep.. im reading and not getting it.. hehe
[26-Jul-2010 13:39:10] ok
[26-Jul-2010 13:39:36] so all ldap groups that it imports wont really be useful
[26-Jul-2010 13:39:57] such as?
[26-Jul-2010 13:40:15] i imported from my AD.. sorry
[26-Jul-2010 13:41:21] and i get a bunch of groups like devs, wiki, and more.. most dont have to interact with zenoss but was just wondering
[26-Jul-2010 13:42:55] When I get an IP down event, if the device comes back up that down event should clear... right?
[26-Jul-2010 13:43:00] JonnyV: oh, it probably pulled those down from your LDAP server
[26-Jul-2010 13:43:06] they were probably already on there
[26-Jul-2010 13:43:14] what you do is assign an LDAP group to a role
[26-Jul-2010 13:43:20] so that anyone in that group gets that role
[26-Jul-2010 13:44:02] ok.. so the autorization is based on only the 3 roles zenoss provides
[26-Jul-2010 13:44:18] unless you want to create a new role, but good luck with that
[26-Jul-2010 13:44:24] there's like 100 checkboxes
[26-Jul-2010 13:44:27] haha
[26-Jul-2010 13:45:24] i dont think so.. too newbie with zenoss.. first im trying to figure out how to map/organize my infrastructure
[26-Jul-2010 13:45:51] pity on the roles... all other features are quite awesome
[26-Jul-2010 13:45:53] also, it's really only 2 roles
[26-Jul-2010 13:45:57] do not user ZenManager
[26-Jul-2010 13:46:02] just use ZenUser and Manager
[26-Jul-2010 13:46:34] well, LDAP isn't even officially supported in Core
[26-Jul-2010 13:46:44] by the way, do you have LDAP working in 3.0?
[26-Jul-2010 13:47:04] or 2.5.2?
[26-Jul-2010 13:52:28] on 3.0 rmatte
[26-Jul-2010 13:55:27] Sorry damn phone took a digger. Basically I came in this morning and Zenoss said 280 some odd devices were down, when I could ping them from within Zenoss. I just restarted the zenoss-stack, is that normal?
[26-Jul-2010 14:02:06] arrrghhh: that's obviously not normal
[26-Jul-2010 14:04:32] eugh, what the heck am I missing here
[26-Jul-2010 14:04:48] I have these Cisco UCOS devices that I'm trying to monitor processes on
[26-Jul-2010 14:05:03] they provide a full hrSWRun table just like a Linux server would
[26-Jul-2010 14:05:10] but Zenoss isn't detecting the processes
[26-Jul-2010 14:11:49] Rmatte, we had some network hardware fail over during the weekend
[26-Jul-2010 14:13:13] I'm sure that's the cause of all these devices listed as down, I just want to know why they were still listed as down this morning...
[26-Jul-2010 14:13:40] because Zenoss wasn't able to ping them
[26-Jul-2010 14:13:49] that's the only reason why they would have still been listed as down
[26-Jul-2010 14:14:09] zenping wasn't able to ping them
[26-Jul-2010 14:14:11] for whatever reason
[26-Jul-2010 14:14:20] I've never had a problem like what you're describing
[26-Jul-2010 14:15:48] I could ping them
[26-Jul-2010 14:16:04] But zenping could not... that doesn't make sense to me.
[26-Jul-2010 14:17:27] how do I install a zenpack in develoment mode
[26-Jul-2010 14:17:50] er development mode
[26-Jul-2010 14:21:41] figured it out
[26-Jul-2010 14:21:46] So why could the zenoss gui ping the device and zenping could not?
[26-Jul-2010 14:24:40] zenping is not actually using the ping command
[26-Jul-2010 14:24:50] when you're doing it from the gui it's just using the ping command
[26-Jul-2010 14:25:01] Zenoss uses twisted to perform it's network tests including pings
[26-Jul-2010 14:25:13] I've seen cases where I could ping a device using the ping command but zenping couldn't
[26-Jul-2010 14:25:20] something to do with packets
[26-Jul-2010 14:25:26] and the way they are handled
[26-Jul-2010 14:25:55] Well this happened last time this core network device failed over
[26-Jul-2010 14:26:08] I had to restart zenoss, it was full of red!
[26-Jul-2010 14:26:33] Simon4_ is now known as Simon4
[26-Jul-2010 14:26:49] I guess I should've just left it to see if you guys could find a problem with it. All the red was freakin me out tho lol
[26-Jul-2010 14:27:14] lol
[26-Jul-2010 14:27:30] oh wait, how long were the devices actually down for?
[26-Jul-2010 14:28:10] was it for more than a day?
[26-Jul-2010 14:28:19] I think so
[26-Jul-2010 14:28:24] Let me look
[26-Jul-2010 14:28:26] then I know what your problem was
[26-Jul-2010 14:28:41] Collectors -> localhost -> Maximum Ping Failures
[26-Jul-2010 14:28:48] by default it's set to 24 hours I believe
[26-Jul-2010 14:28:55] Oh
[26-Jul-2010 14:29:00] meaning if ping failures reaches that count it stops monitoring the device entirely
[26-Jul-2010 14:29:06] K 1 sec
[26-Jul-2010 14:29:08] I've already complained saying that was a stupid feature
[26-Jul-2010 14:29:13] but they decided to leave it in
[26-Jul-2010 14:29:29] I just set it to "99999999999" myself
[26-Jul-2010 14:29:35] which effectively disables it
[26-Jul-2010 14:29:57] also, once it stops monitoring it there is no indication in the UI that it has done so
[26-Jul-2010 14:30:09] It was set to 1440
[26-Jul-2010 14:30:09] and the only way to get it to monitor it again is to either remodel the device or restart Zenoss
[26-Jul-2010 14:30:16] right, 1440 is the default
[26-Jul-2010 14:30:16] Damn
[26-Jul-2010 14:30:26] That sounds like what happened...
[26-Jul-2010 14:30:31] it's really stupid, I found that out about 7 months ago and I flipped
[26-Jul-2010 14:30:37] we missed an outage because of it
[26-Jul-2010 14:30:45] since I thought it was monitoring a device when it actually wasn't
[26-Jul-2010 14:30:47] I'm really freakin glad this isn't in production
[26-Jul-2010 14:30:52] lol
[26-Jul-2010 14:30:55] Yea, that sucks!
[26-Jul-2010 14:30:58] that's the dumbest thing in Zenoss
[26-Jul-2010 14:31:08] they should default it to 0 and make that disable it
[26-Jul-2010 14:31:11] but it doesn't
[26-Jul-2010 14:31:12] Ah well. At least it's easily fixed
[26-Jul-2010 14:31:20] Agreed...
[26-Jul-2010 14:31:21] you need to set it to: 99999999999 or whatever
[26-Jul-2010 14:31:40] I opened a trac ticket for it but I think they might have just closed it out at some point
[26-Jul-2010 14:31:43] I'll check
[26-Jul-2010 14:31:43] I filled the box with 9's basically haha
[26-Jul-2010 14:31:56] actually I think Matt opened a ticket for it after I mentioned it
[26-Jul-2010 14:32:01] I don't think it ever went anywhere
[26-Jul-2010 14:32:11] yeh, that's what I did too
[26-Jul-2010 14:33:09] Lol my data point graph looks really good for that last 12 hr period haha
[26-Jul-2010 14:33:22] I'll bet
[26-Jul-2010 14:33:23] lol
[26-Jul-2010 14:34:08] I know most of Zenoss' little quirks like that
[26-Jul-2010 14:34:10] Ugh I definitely need more powerful hardware
[26-Jul-2010 14:34:15] that's the worst for new users who don't know about it
[26-Jul-2010 14:34:35] I should make a "10 things you should know when starting to use Zenoss" style document for the wiki
[26-Jul-2010 14:34:39] and include that on there
[26-Jul-2010 14:34:58] or they could just do what they should have done and set it to 0 by default to disable it :P
[26-Jul-2010 14:35:42] the justification that one of the devs gave me was "If I have a server that's down for over a day, and I know that server is down, I don't want Zenoss to keep alerting me about it"
[26-Jul-2010 14:35:51] and I was like "That's what production states are for, durrrr"
[26-Jul-2010 14:36:10] lol
[26-Jul-2010 14:37:10] 1440 / 60 == 24
[26-Jul-2010 14:37:17] so yeh, it is 24 hours by default
[26-Jul-2010 14:37:35] Ah
[26-Jul-2010 14:37:50] So there is no way to truly disable it then?
[26-Jul-2010 14:38:01] other than setting it to 999999999999, nope, not currently
[26-Jul-2010 14:38:07] I would think 0 would completely disable that "feature"
[26-Jul-2010 14:38:11] Huh
[26-Jul-2010 14:38:18] hopefully the trac ticket that was opened for it is still open
[26-Jul-2010 14:38:23] yeh, you'd think so, but it doesn't
[26-Jul-2010 14:38:25] we tested it
[26-Jul-2010 14:38:36] How long will it stop monitoring that device for?
[26-Jul-2010 14:38:45] until it's remodeled or Zenoss is restarted
[26-Jul-2010 14:38:50] Oy
[26-Jul-2010 14:38:55] and there's absolutely no indication that it's no longer monitoring it
[26-Jul-2010 14:38:59] which is the really scary part
[26-Jul-2010 14:39:08] Does zenoss remodel automatically at all?
[26-Jul-2010 14:39:12] you'd expect at least a notification or a special icon on the device or something
[26-Jul-2010 14:39:26] yes, by default it remodels automatically every 12 hours
[26-Jul-2010 14:39:32] I personally disable the automated remodelling
[26-Jul-2010 14:39:33] Ok
[26-Jul-2010 14:39:53] Collectors -> localhost -> Modeler Cycle Interval
[26-Jul-2010 14:40:02] that's where you can adjust how often it auto-remodels
[26-Jul-2010 14:40:40] I never did get the automatic device adding to work... i'd like that to run once a week or a month, do you do anything like that?
[26-Jul-2010 14:41:05] oh crap, I was supposed to log my time in our CRM system by today, whoops
[26-Jul-2010 14:41:08] time to get cracking on that
[26-Jul-2010 14:41:30] arrrghhh: no, automatic device adding is a nightmare
[26-Jul-2010 14:41:46] arrrghhh: I add everything by add, and if it's a really big add I use batch load scripts
[26-Jul-2010 14:41:55] by hand*
[26-Jul-2010 14:42:13] Bummer. I'm so lazy, and people don't always tell me when new crap is added.
[26-Jul-2010 14:42:24] Ever since I upgraded to 3.0, I get an attribute error for 'zProdStateThreshold' whenever I try to view multi-graph reports or create new multi-graph reports. Anyone have that problem or something similar? Or know how to fix this?
[26-Jul-2010 14:42:55] ad4m: unfortunately I haven't come across that
[26-Jul-2010 14:43:15] ad4m: If you don't find a solution, log it as a bug in trac
[26-Jul-2010 14:43:25] I wouldn't be surprised if it is a bug
[26-Jul-2010 14:43:26] and send beer!
[26-Jul-2010 14:43:32] indeed
[26-Jul-2010 14:43:41] and wenches
[26-Jul-2010 14:43:46] you guys work at zenoss?
[26-Jul-2010 14:43:52] not me, cgibbons does
[26-Jul-2010 14:43:59] I hope to eventually
[26-Jul-2010 14:45:06] i think a few of us here are going to annapolis on friday to meet with some zenoss folks
[26-Jul-2010 14:45:21] the annapolis zenoss people are boring and don't know how to party
[26-Jul-2010 14:45:32] cgibbons: if you can help me fix this problem i'll definitely buy you multiple beers if you're in annapolis
[26-Jul-2010 14:45:44] ah
[26-Jul-2010 14:46:10] cgibbons: lol
[26-Jul-2010 14:48:45] cgibbons: what code performs modelling of OS Processes?
[26-Jul-2010 14:49:08] I have a device which provides hrSWRun SNMP info, but Zenoss doesn't model/monitor the processes on it for some reason
[26-Jul-2010 14:49:17] I want to take a look at the code and see if I can figure out why
[26-Jul-2010 14:49:40] I saw OSProcess.py, but I was expecting some sort of modeller plugin
[26-Jul-2010 14:49:43] or does one not exist?
[26-Jul-2010 14:50:18] i would expect HRSWRunMap.py in Products/DataCollector/plugins/snmp
[26-Jul-2010 14:50:25] ahhhhh
[26-Jul-2010 14:50:31] * rmatte facepalms
[26-Jul-2010 14:50:33] thanks
[26-Jul-2010 14:51:27] ad4m: I'm not in Annapolis very often. been over a year actually. Most of the devs are in Austin.
[26-Jul-2010 14:51:37] I have a stupid SNMP question...I have a device that pushes traps to Zenoss, and one of the OIDs provides a link to click on - Zenoss seems to truncate this, is there any way to poll it manually or change how zenoss displays it?
[26-Jul-2010 14:52:36] cgibbons: Annapolis = business types?
[26-Jul-2010 14:54:32] rmatte: you have os processes defined for the processes you are hoping to monitor?
[26-Jul-2010 14:55:17] it's the corporate HQ, so most of the execs are there, business, sales mgmt. Austin office was founded in April '08 and now has dev, marketing and is growing rapidly.
[26-Jul-2010 14:56:41] support, prof services guys are in annapolis too
[26-Jul-2010 14:57:26] some good prof services guys in austin that we've used too
[26-Jul-2010 15:02:31] I really should have dropped by the austin office when I was out there the last time
[26-Jul-2010 15:02:40] matt ray said he'd buy me lunch!
[26-Jul-2010 15:11:57] lol
[26-Jul-2010 15:12:02] generous of him
[26-Jul-2010 15:12:20] I'd end up buying him some beers if I was out there
[26-Jul-2010 15:12:29] :)
[26-Jul-2010 15:13:22] Lol yea if anything I think I would owe Matt lunch!
[26-Jul-2010 15:14:01] hehe
[26-Jul-2010 15:14:30] we'd just buy each other beers probably since we tend to help each other out with stuff
[26-Jul-2010 15:14:50] woot, I finally got processes monitored on these UCOS devices
[26-Jul-2010 15:15:02] Yea, I can't say I've "helped" matt w/much of anything.
[26-Jul-2010 15:15:09] lol
[26-Jul-2010 15:15:25] I tried to help you in the ubuntuforums! Not much help tho ha.
[26-Jul-2010 15:15:33] I've done some ZenPack migrations, bug testing, etc...
[26-Jul-2010 15:15:41] yeh, I noticed lol
[26-Jul-2010 15:15:44] I'm still seeing that issue
[26-Jul-2010 15:15:50] guaranteed that's a kernel bug
[26-Jul-2010 15:15:51] Weird
[26-Jul-2010 15:16:00] Yea, I would say so.
[26-Jul-2010 15:16:18] You said it doesn't happen with 8.04 right?
[26-Jul-2010 15:16:19] ldap has been sucking ass in 10.04 too
[26-Jul-2010 15:16:24] it tends to drop off every now and then
[26-Jul-2010 15:16:35] doesn't happen with 8.04, correct
[26-Jul-2010 15:16:41] ldap is super stable in 8.04 too
[26-Jul-2010 15:16:45] Weird indeed. Yea, I'd stick with 8.04.
[26-Jul-2010 15:16:52] they seem to have really messed those 2 things up in 10.04 unfortunately
[26-Jul-2010 15:17:04] mrayzenoss is now known as mray
[26-Jul-2010 15:17:12] I'm thinking I may go back to it. 10.04 had some stupid changes done to it.
[26-Jul-2010 15:17:22] well, 10.04 is otherwise working fine
[26-Jul-2010 15:17:29] it's just ldap and nfs that are killers right now
[26-Jul-2010 15:17:33] Hrm
[26-Jul-2010 15:17:41] I haven't had any other problems with it
[26-Jul-2010 15:17:54] I just don't like landscape being included w/the server distro
[26-Jul-2010 15:17:59] well, they did remove python 2.4 and 2.5 from the repos, which was stupid in my opinion
[26-Jul-2010 15:18:01] well if you're ever in Austin add gwb2353 to the list of "grab a beer with"
[26-Jul-2010 15:18:04] Or the new phone-home motd system.
[26-Jul-2010 15:18:15] gwb2353: will do lol
[26-Jul-2010 15:18:28] phone-home motd system?
[26-Jul-2010 15:18:40] Yea... you didn't notice that?
[26-Jul-2010 15:18:53] well which part?
[26-Jul-2010 15:18:59] you know the motd system is customizable right?
[26-Jul-2010 15:19:21] Everytime you login (maybe only ssh) but it phones home with stats on your server
[26-Jul-2010 15:19:34] Yea, there's a thread on how to rip it out
[26-Jul-2010 15:20:11] Nothing official from Canocial on how-to do it tho, and there's debate on how complete it is.
[26-Jul-2010 15:21:09] But back to zenoss :D. Do you know how I can solve my snmp trap truncation problem?
[26-Jul-2010 15:21:24] arrrghhh: you just go in to /etc/update-motd.d
[26-Jul-2010 15:21:30] and you can modify/remove/add scripts
[26-Jul-2010 15:21:55] I didn't see anything that sends stats in there though, so that might be elsewhere
[26-Jul-2010 15:22:04] Yea, there's some debate on whether or not that removes all of it. I'll find the thread when I get back to my desk.
[26-Jul-2010 15:22:15] truncation problem?
[26-Jul-2010 15:22:45] Yea, I have a trap that one of the OIDs has a url, and it's never complete
[26-Jul-2010 15:22:54] odd
[26-Jul-2010 15:23:16] I've had urls in traps before but never had them truncated
[26-Jul-2010 15:23:25] you sure the device itself isn't causing the truncation?
[26-Jul-2010 15:23:27] The url is freakin enormous, for whatever reason. Geniuses at netscout.
[26-Jul-2010 15:23:31] ah
[26-Jul-2010 15:23:39] hmmm
[26-Jul-2010 15:23:59] Hrm... I'm having trouble polling the device manually
[26-Jul-2010 15:24:25] It's one of those that is pushed instead of pulled (if I understand it correctly that is...)
[26-Jul-2010 15:24:35] yes, that's what a trap is
[26-Jul-2010 15:25:00] snmp form of a syslog basically
[26-Jul-2010 15:25:48] I can't see any zentrap setting pertaining to truncation/size
[26-Jul-2010 15:27:29] Well the info sent in that one OID is ridiculous
[26-Jul-2010 15:27:49] It has a bunch of text and then the URL
[26-Jul-2010 15:28:32] Maybe I have to make a change on the netscout box rather than the zenoss side. Is there an outright limit on the length of an OID within SNMP?
[26-Jul-2010 15:28:38] yeh, it might be getting truncated before it even hits Zenoss
[26-Jul-2010 15:29:11] there might be a limit on the length of the value of an OID property
[26-Jul-2010 15:29:12] not sure
[26-Jul-2010 15:29:16] Is there anyway I can test?
[26-Jul-2010 15:29:38] I've never really done much testing, you'd need to find some linux utility that'll allow you to capture them as they come in
[26-Jul-2010 15:29:49] hmmmm
[26-Jul-2010 15:29:57] you could try running tcpdump for port 162
[26-Jul-2010 15:30:01] and see what you get
[26-Jul-2010 15:31:01] http://ubuntuforums.org/showthread.php?t=1470011&highlight=motd
[26-Jul-2010 15:31:16] That's the motd thread I was talkin about.
[26-Jul-2010 15:31:35] Hrm. Ok.
[26-Jul-2010 15:32:24] I removed that canonical motd part
[26-Jul-2010 15:32:27] I'll probably talk to the netscout folks since we should be paying for their support services ;)
[26-Jul-2010 15:32:28] and replaced it with my own
[26-Jul-2010 15:32:32] Yea
[26-Jul-2010 15:32:45] Still stupid that it phones home by default
[26-Jul-2010 15:32:48] but I left the parts that show when updates are available or when a reboot is required
[26-Jul-2010 15:32:58] since that's handy for upgrade tracking purposes
[26-Jul-2010 15:33:12] I guess they're trying to drum up enterprise $$$
[26-Jul-2010 15:33:21] Indeed
[26-Jul-2010 15:33:38] I have a relatively basic modelling question and haven't been able to find any more general place to ask it :)
[26-Jul-2010 15:34:17] Well quit stealing my username for starters! :P
[26-Jul-2010 15:34:31] Go ahead and ask!
[26-Jul-2010 15:34:43] lol
[26-Jul-2010 15:35:18] heh, I've been Arrgh for quite a few years... Anyway, I was going to ask where something as prosaic as a *website* fits into the CIM but now it looks like Zenoss has its own homegrown object model
[26-Jul-2010 15:35:42] I've been arrrghhh (just like that) since about '96...
[26-Jul-2010 15:36:12] pirate duel!
[26-Jul-2010 15:36:14] Arrgh`: CIM has always been a bit of overkill
[26-Jul-2010 15:36:16] And I don't know what you mean...
[26-Jul-2010 15:36:20] Interesting, probably about the same for me.
[26-Jul-2010 15:36:37] Lol!
[26-Jul-2010 15:37:13] No question that CIM is massive overkill for any real-world use case, but it does seem to be the de facto interchange format
[26-Jul-2010 15:38:09] Arrgh`: essentially we use a Device, which belongs to a Device Class which defines how things are monitored
[26-Jul-2010 15:38:41] I guess Zenoss is much more concerned with what I would consider infrastructure, rather than application modelling?
[26-Jul-2010 15:38:42] Arrgh`: the device has Status, OS, Hardware, Software and Performance Data
[26-Jul-2010 15:39:04] Arrgh`: not necessarily, there's just a bit of shoe-horning of devices into the model
[26-Jul-2010 15:39:11] ie. Devices have applications
[26-Jul-2010 15:39:38] but some applications become devices (VMware, Cfengine, etc.)
[26-Jul-2010 15:39:48] Device == IP address for the most part
[26-Jul-2010 15:40:13] to sum it up, classes have devices, devices have components
[26-Jul-2010 15:42:26] I'm doing the Getting Started Webinar tomorrow: community/about/webinars
[26-Jul-2010 15:42:36] first one with the 3.0 UI
[26-Jul-2010 15:42:46] So if you have a machine with a monolithic LAMP stack hosting a dozen unique VirtualHosted sites/domains that are all delivered through the same IP/port, would these typically be exposed separately in the model?
[26-Jul-2010 15:42:46] cool
[26-Jul-2010 15:43:12] mray: $5 says you use the line "Where the heck did we put that again?" at least once
[26-Jul-2010 15:43:20] rmatte: true
[26-Jul-2010 15:43:27] :)
[26-Jul-2010 15:43:35] Arrgh`: typically you'd pick a management IP for the machine, get the basics there
[26-Jul-2010 15:44:15] Arrgh`: then add the virtual hosts to a device class that doesn't concern itself with modeling the machine and add templates that monitor the LAMP stack
[26-Jul-2010 15:44:35] so really you watch the 'L' on 1 device and the AMPs on other devices
[26-Jul-2010 15:44:58] Interesting
[26-Jul-2010 15:45:08] yeah, kinda unweildy I'll admit
[26-Jul-2010 15:45:22] Is there any other IRC channel out there that discusses this kind of thing? :)
[26-Jul-2010 15:46:00] well, for Zenoss there's just #zenoss and #zenoss-testing
[26-Jul-2010 15:46:15] and #zenoss-testing is really quiet when we're not in alpha/beta
[26-Jul-2010 15:48:14] Signed up for tomorrow
[26-Jul-2010 15:48:27] try not to heckle me too much :)
[26-Jul-2010 15:48:27] Hopefully doesn't get canceled :P
[26-Jul-2010 15:48:39] Lol not like last time!
[26-Jul-2010 15:49:14] Oy, I was way too green. I'm still not very comfortable with the system
[26-Jul-2010 15:49:20] Getting better tho
[26-Jul-2010 15:51:25] :)
[26-Jul-2010 15:53:53] ok this is really weird
[26-Jul-2010 15:53:57] I'm trying to uninstall this zenpack
[26-Jul-2010 15:54:41] and it's failing to uninstall because: Value: The property zOTRSTicketArticleType does not exist
[26-Jul-2010 15:54:57] the thing is, that property exists, it's in the zProperties list
[26-Jul-2010 15:58:35] not sure
[26-Jul-2010 16:02:06] ahh I just reinstalled the zenpack on top of itself
[26-Jul-2010 16:02:18] nice
[26-Jul-2010 16:02:21] ok good, I now have all the zenproperties that I will need, at least for the basic OTRS stuff
[26-Jul-2010 16:02:40] the fact that the pack can't be uninstalled is a bit concerning though lol
[26-Jul-2010 16:02:43] I have like 25 zProperties
[26-Jul-2010 16:03:21] yeah I don't know what's up with that
[26-Jul-2010 16:07:52] ValueError: The property zOTRSTicketArticleType does not exist
[26-Jul-2010 16:08:10] that has to be a bug
[26-Jul-2010 16:08:17] because it's right there in the zproperties list
[26-Jul-2010 16:08:19] could be
[26-Jul-2010 16:09:52] it might have something to do with the huge number of zproperties I have in this zenpack
[26-Jul-2010 16:10:03] doubt it has to do with number
[26-Jul-2010 16:10:29] just some yummy buggy code somewhere
[26-Jul-2010 16:11:01] I've had quite a few problems while messing with ZenPacks in 3.0
[26-Jul-2010 16:11:11] Module OFS.PropertyManager, line 216, in _delProperty
[26-Jul-2010 16:11:16] I've installed some then been completely unable to remove them
[26-Jul-2010 16:11:51] that's where the error is being generated
[26-Jul-2010 16:12:19] right, so it's trying to perform a _delProperty
[26-Jul-2010 16:12:21] but on what
[26-Jul-2010 16:12:52] it's trying to delete the property zOTRSTicketArticleType
[26-Jul-2010 16:12:59] for instance, to remove a property from a device you'd do something like...
[26-Jul-2010 16:13:02] a=find("devicename")
[26-Jul-2010 16:13:03] a.__delattr__("zWinUser")
[26-Jul-2010 16:13:17] I guess it's trying to remove it from the root class
[26-Jul-2010 16:13:45] yeah it is
[26-Jul-2010 16:14:03] it is only specified at the root class at the moment
[26-Jul-2010 16:15:05] weird that it's failing
[26-Jul-2010 16:15:11] http://dev.zenoss.com/trac
[26-Jul-2010 16:15:14] login with zenoss/zenoss
[26-Jul-2010 16:15:16] log it as a bug
[26-Jul-2010 16:15:25] it's probably a bug with that zenpack
[26-Jul-2010 16:15:32] what's the full stack trace?
[26-Jul-2010 16:15:53] <|rt|> in Zenoss is there any concept of dependancies between services?
[26-Jul-2010 16:16:32] cgibbons: http://pastebin.com/nzrPqxac
[26-Jul-2010 16:17:18] Hmm I wonder if that zenpack has a custom __init__.py cleanup
[26-Jul-2010 16:17:33] no, it doesn't
[26-Jul-2010 16:17:42] hmmm
[26-Jul-2010 16:19:02] maybe if I can manually clean up the properties and then just remove the zenpack directory with zenoss shut down
[26-Jul-2010 16:19:24] and reinstall it
[26-Jul-2010 16:19:32] but I do not know how to manually delete zproperties
[26-Jul-2010 16:22:25] I will open a bug on this tomorrow
[26-Jul-2010 16:23:00] cgibbons: I made that pack, it's pretty plain
[26-Jul-2010 16:23:05] stock __init__.pys
[26-Jul-2010 16:23:18] good deal
[26-Jul-2010 16:23:22] the only thing I have done is added zproperties to rmatte's pack
[26-Jul-2010 16:23:29] in the __init__.py
[26-Jul-2010 16:23:35] let's see that code
[26-Jul-2010 16:24:21] http://pastebin.com/2jLCmbSN
[26-Jul-2010 16:26:24] certainly looks harmless
[26-Jul-2010 16:26:59] yes, and this is why I am so confused
[26-Jul-2010 16:27:12] it should not be having an issue deleting that property, because it is visible when I go into the zproperties tab
[26-Jul-2010 16:27:17] the technical term is "confuzzled"
[26-Jul-2010 16:29:08] it won't prevent me from getting going on coding this
[26-Jul-2010 16:29:12] at least
[26-Jul-2010 16:29:18] yeh
[26-Jul-2010 16:29:20] but it will eventually need to get fixed heh
[26-Jul-2010 16:29:30] yup
[26-Jul-2010 16:29:59] If I were coding and ran into the problem, I'd zenwipe and modify that init to do one zproperty at a time until I figured out what is up
[26-Jul-2010 16:31:55] will I lose my devices if I zenwipe?
[26-Jul-2010 16:32:00] yes
[26-Jul-2010 16:32:06] it restores it to a virgin state
[26-Jul-2010 16:32:17] don't want to do that then
[26-Jul-2010 16:32:38] cgibbons works with production instances so it's not a big deal in that case
[26-Jul-2010 16:32:45] erm
[26-Jul-2010 16:32:49] development instances
[26-Jul-2010 16:32:50] rather
[26-Jul-2010 16:32:59] * rmatte 's brain is on backwards apparently
[26-Jul-2010 16:33:03] yeah I was not expecting to need a development instance
[26-Jul-2010 16:33:04] heh
[26-Jul-2010 16:33:32] there has to be some manual way of cleaning up the zproperties
[26-Jul-2010 16:33:32] sweet, finally going to be upgrading otrs in 2 weeks
[26-Jul-2010 16:33:47] OTRS 3.0 is coming out in a few weeks
[26-Jul-2010 16:33:54] mducharme-laptop: you could try cleaning them up in zendmd
[26-Jul-2010 16:34:17] mducharme-laptop: yeh, but 3.0 has a completely new interface, so I'd have to re-port all my customizations over
[26-Jul-2010 16:34:27] better to just go to 2.4 for now since it's much much nicer than 2.3
[26-Jul-2010 16:34:35] and then I can get started on 3.0
[26-Jul-2010 16:35:56] our custom 2.4 has been ready to go for months now, I've just been waiting on the go-ahead
[26-Jul-2010 16:36:15] I fixed 2 bugs in otrs during the course of customizing it too
[26-Jul-2010 16:36:16] lol
[26-Jul-2010 16:36:26] submitted them to the bug list and they got patched
[26-Jul-2010 16:38:35] I want to get mine set up as to avoid needing customization
[26-Jul-2010 16:38:52] well, the base otrs kind of sucks though
[26-Jul-2010 16:39:01] the queueview is not exactly suited for NOC
[26-Jul-2010 16:40:25] Hi All!
[26-Jul-2010 16:40:52] hi
[26-Jul-2010 16:42:00] I have an alert that continously kicks off an alert stating the the "SNMP agent is down". A few seconds or minutes later a "Clear" alert is sent. Is there a way for me to modify the sensitivity of the monitor? Change the threshold of "failed" before an alert is emailed to the group?
[26-Jul-2010 16:42:04] Hi Rmatte!
[26-Jul-2010 16:42:37] newmon: for what type of device?
[26-Jul-2010 16:43:05] how could I reload the IP address for a device?, when I remodel this, zenoss still using an old IP address and don't resolve it :(
[26-Jul-2010 16:43:22] I have it happen on two different devices. HP printer and also Windows 2003 Storage Server
[26-Jul-2010 16:43:24] chachan: very easy, go to the page for that device
[26-Jul-2010 16:43:48] chachan: I assume you're using 3.0 at this point?
[26-Jul-2010 16:43:59] rmatte, you're right
[26-Jul-2010 16:43:59] yes i am
[26-Jul-2010 16:44:07] k hold on, I have to fire up that UI
[26-Jul-2010 16:44:21] hehe
[26-Jul-2010 16:44:50] ok, so go to the page for the device
[26-Jul-2010 16:44:56] then in the bottom left you'll see menus
[26-Jul-2010 16:45:04] click on the settings menu (the one with the gear)
[26-Jul-2010 16:45:15] then select Reset/Change IP Address
[26-Jul-2010 16:45:18] yes, I saw it
[26-Jul-2010 16:45:20] :)
[26-Jul-2010 16:45:20] pretty self explanatory after that
[26-Jul-2010 16:45:23] thanks!
[26-Jul-2010 16:45:44] newmon: hmmm
[26-Jul-2010 16:45:55] FYI-- im actually still using 2.5
[26-Jul-2010 16:45:55] newmon: do you have any packet loss to those devices from the Zenoss server?
[26-Jul-2010 16:46:01] k
[26-Jul-2010 16:46:06] let me run a mtr
[26-Jul-2010 16:46:42] check for packet loss... and another thing you can try is going in to the zProperties for those devices and changing zMaxOIDPerRequest from 40 to 20 or 10
[26-Jul-2010 16:47:02] you can also try increasing zSnmpTimeout
[26-Jul-2010 16:47:03] no. The zenoss server and the server in question are on the same lan
[26-Jul-2010 16:47:11] that's the best I can suggest
[26-Jul-2010 16:47:26] cool! i will change those settings one at a time
[26-Jul-2010 16:47:29] thanks
[26-Jul-2010 16:47:29] k
[26-Jul-2010 16:47:31] np
[26-Jul-2010 16:48:00] <|rt|> I think I just found a small bug in 3.0...when you lock or unlock components for a host the message that displays at the top says "locked %s components"
[26-Jul-2010 16:48:17] <|rt|> I'm assuming that %s is supposed to get filled in with the components name
[26-Jul-2010 16:48:31] does it take some minutes?, the reset seems not working
[26-Jul-2010 16:48:40] <|rt|> I should also try that under firefox to make sure it's not an issue with chrome
[26-Jul-2010 16:48:54] yea, its weird. Anytime i run a snmpwalk it works perfectly. It has not failed.
[26-Jul-2010 16:49:32] in chrome, when I model a device, it doesn't show any result, is that normal?
[26-Jul-2010 16:49:38] chachan: what do you mean it's not working?
[26-Jul-2010 16:49:49] <|rt|> yeah that's an issue with that widget updating
[26-Jul-2010 16:49:55] rmatte, the IP address is unchanged
[26-Jul-2010 16:50:14] <|rt|> chachan: anything like snmpwalk, ping, etc that opens up that overlay doesn't work in chrome
[26-Jul-2010 16:50:23] uhm
[26-Jul-2010 16:50:23] newmon: that's why I said to change zMaxOIDPerRequest, some devices don't like the amount of SNMP polls Zenoss tries to do at once
[26-Jul-2010 16:50:27] but snmpwalk always works
[26-Jul-2010 16:50:47] chachan: YOU need to set the IP by hand
[26-Jul-2010 16:50:48] <|rt|> chachan: the command is still executed...it just doesn't display the results to you
[26-Jul-2010 16:50:55] chachan: instead of hoping that Zenoss will pick up the new IP
[26-Jul-2010 16:51:18] chachan: or clear it and remodelling is supposed to work apparently too, but I've never personally tried it
[26-Jul-2010 16:51:27] as long as the device name is a valid hostname
[26-Jul-2010 16:51:44] got it. thanks. ive made the change and will observe. If i still have problems ill try your other suggestion.
[26-Jul-2010 16:51:51] k
[26-Jul-2010 16:51:52] it is, it just change the IP
[26-Jul-2010 16:51:57] <|rt|> that %s message shows up in firefox as well....should I file a bug?
[26-Jul-2010 16:51:59] then what's the problem?
[26-Jul-2010 16:52:14] |rt|: yes, that's obviously a coding issue
[26-Jul-2010 16:52:23] they messed up the syntax somewhere
[26-Jul-2010 16:53:09] <|rt|> rmatte: yeah works find if you're locking a device...just seems to be a problem if your dealing with a component
[26-Jul-2010 16:53:21] ah
[26-Jul-2010 16:53:32] yeh, file a bug, I'm sure that's an easy fix for them
[26-Jul-2010 16:53:45] <|rt|> well the lock works either way...but the message is wrong....you know what I mean
[26-Jul-2010 16:53:50] yup
[26-Jul-2010 16:53:53] cosmetic issue
[26-Jul-2010 16:53:56] <|rt|> yeah
[26-Jul-2010 16:54:14] <|rt|> where do you file bugs for zenoss?
[26-Jul-2010 16:54:22] http://dev.zenoss.com/trac
[26-Jul-2010 16:54:25] login with zenoss/zenoss
[26-Jul-2010 16:54:31] or ask Matt Ray for your own account
[26-Jul-2010 16:59:47] that's mray@zenoss.com if you're looking
[26-Jul-2010 16:59:59] once i added manually a device when it is modeled?
[26-Jul-2010 17:00:15] i assigned it to the class i wanted to
[26-Jul-2010 17:01:03] <|rt|> Matt for now I'll just file it using the zenoss/zenoss account....this is such a minor bug no reason to create an account
[26-Jul-2010 17:01:16] <|rt|> probably taking me longer to fill in the bug report than it will to fix
[26-Jul-2010 17:02:01] JonnyV: you can manually model it or it'll get modeled within 12 hours
[26-Jul-2010 17:02:38] yes.. i read the 12 hour period. How to do it manually?
[26-Jul-2010 17:03:02] JonnyV: what, model a device manually?
[26-Jul-2010 17:03:21] JonnyV: go to the device page for that device and check the menus, you'll find the remodel option
[26-Jul-2010 17:03:24] it's easy to find
[26-Jul-2010 17:03:46] ok... im looking.. thanks rmatte
[26-Jul-2010 17:03:52] np
[26-Jul-2010 17:03:54] <|rt|> mray: ok bug filed http://dev.zenoss.com/trac/ticket/7106
[26-Jul-2010 17:04:24] JonnyV: If you're using 3.0... hint hint, bottom left
[26-Jul-2010 17:04:34] middle meny
[26-Jul-2010 17:04:36] menu*
[26-Jul-2010 17:04:38] Model Device
[26-Jul-2010 17:05:03] yepss.. i already did that.. but returned with error
[26-Jul-2010 17:05:16] it is trying to use nmap
[26-Jul-2010 17:05:21] instead of snmp
[26-Jul-2010 17:05:37] then you assigned the nmap collector plugin to that class
[26-Jul-2010 17:05:41] remove the nmap collector plugin
[26-Jul-2010 17:05:51] <|rt|> with that bug filed it's time to head home
[26-Jul-2010 17:05:55] I need to log my time so I can't help out anymore for now, I'll be back in a bit
[26-Jul-2010 17:06:08] ok :)
[26-Jul-2010 17:07:14] what's a zenping heartbeat?
[26-Jul-2010 17:07:20] probably sthg to do with this. No command plugins found for 10.10.**
[26-Jul-2010 17:07:36] also No manage IP for 10.10.**
[26-Jul-2010 17:07:38] is something like a threshold?
[26-Jul-2010 17:09:28] chachan: heartbeat failures are when a daemon doesn't respond back in the expected amount of time
[26-Jul-2010 17:09:41] chachan: they are wayyyyyy too sensitive though, I just completely disregard them
[26-Jul-2010 17:10:09] uhm
[26-Jul-2010 17:10:28] I think I could restart it, yeap?
[26-Jul-2010 17:10:34] hunh?
[26-Jul-2010 17:10:44] no no, it doesn't mean that that daemon is unresponsive
[26-Jul-2010 17:10:49] mmm
[26-Jul-2010 17:11:01] I'm saying the heartbeat alerts are WAY too sensitive since 2.5
[26-Jul-2010 17:11:06] rendering them basically useless
[26-Jul-2010 17:11:12] I just completely disregard them
[26-Jul-2010 17:11:18] as long as the daemons are all running you're fine
[26-Jul-2010 17:11:25] mmm, ok
[26-Jul-2010 17:11:51] If they'd fix them to be less sensitive and not generate wackloads of false positives then it would actually be useful
[26-Jul-2010 17:12:45] rmatte, I got it, could I clean this event?
[26-Jul-2010 17:13:21] chachan: it'll auto-clear eventually, just leave it
[26-Jul-2010 17:13:28] mmm
[26-Jul-2010 17:13:32] great
[26-Jul-2010 17:13:33] set a filter to not alert on them
[26-Jul-2010 17:13:36] thanks again
[26-Jul-2010 17:13:36] and just leave them
[26-Jul-2010 17:13:38] np
[26-Jul-2010 17:15:10] i think the problem is that the device aint showing its ip rmatte. In the list i see the device name, which is the IP actually, but nothing on IP address field
[26-Jul-2010 17:15:49] the class is the same as 30 other devices that are being modeled correctly. they were discovered
[26-Jul-2010 17:16:14] see you guys
[26-Jul-2010 17:16:42] JonnyV: If you actually go to the device's page you can manually set the IP of the device
[26-Jul-2010 17:16:50] JonnyV: which is what I've been trying to tell you
[26-Jul-2010 17:16:55] you literally type the IP in and set it
[26-Jul-2010 17:17:19] don't confuse setting the name with setting the IP
[26-Jul-2010 17:17:22] 2 different things
[26-Jul-2010 17:17:30] exactly
[26-Jul-2010 17:17:37] that was confusing me
[26-Jul-2010 17:17:45] right, so go set the IP
[26-Jul-2010 17:17:46] as when i added it only shows set name
[26-Jul-2010 17:18:07] probably because the hostname that you added it as doesn't resolve
[26-Jul-2010 17:18:21] i put the ip as the name
[26-Jul-2010 17:18:37] so i can track it while testing.. hehe
[26-Jul-2010 17:18:49] that's fine, but make sure the actual IP is set as well
[26-Jul-2010 17:19:14] ok.. im in the overview of the device
[26-Jul-2010 17:20:12] you are using the old interface or the new?
[26-Jul-2010 17:20:35] default for 3.0
[26-Jul-2010 17:20:41] k, the new
[26-Jul-2010 17:20:43] bottom left
[26-Jul-2010 17:20:44] menus
[26-Jul-2010 17:20:52] the menu options are very self explanatory
[26-Jul-2010 17:20:57] find the one for setting the IP address
[26-Jul-2010 17:21:12] theree
[26-Jul-2010 17:21:15] pfff
[26-Jul-2010 17:21:23] mray: I thought this new UI was supposed to be easier for new users :P
[26-Jul-2010 17:21:24] thanksss.. i would have never found it alone
[26-Jul-2010 17:21:36] np
[26-Jul-2010 17:22:23] anyways, I need to get back to logging my time
[26-Jul-2010 17:22:58] rmatte: I need to set up a feedback session ;)
[26-Jul-2010 17:24:37] no, you need a guider tour of the new UI
[26-Jul-2010 17:24:41] guided*
[26-Jul-2010 17:24:43] posted as a video
[26-Jul-2010 17:25:01] or some sort of flash slideshow
[26-Jul-2010 17:26:03] yeah, the Getting Started Guide still needs to be updated and more new material will continue to be published
[26-Jul-2010 17:27:48] i think i found why it was failing. from logs, Found IP '10.10.2.254' on device '172.18.10.254'; skipping discovery
[26-Jul-2010 17:28:00] it gives a failure on the result
[26-Jul-2010 17:28:17] its an interface from another already discovered device
[26-Jul-2010 17:28:47] yes, it will do that
[26-Jul-2010 17:29:01] you can't have an interface on one device matching the IP of another device
[26-Jul-2010 17:30:31] the only way i am seeing it ok is in the networks section
[26-Jul-2010 17:31:34] where it points to the actual device
[26-Jul-2010 17:32:20] and comments the interface... pretty cool
[26-Jul-2010 17:32:53] yup
[26-Jul-2010 17:34:19] it already did the throughput graphs... great
[26-Jul-2010 17:40:01] i read about the Production state value and i understand it and its quite useful. What about the priority Value for each device? i suppose it has to do with the alerting system
[26-Jul-2010 18:06:39] I don't think it's used much
[26-Jul-2010 18:15:55] hehe
[26-Jul-2010 18:31:03] I use the priority value but it's used as part of our algorithm for determining ticket severity when creating tickets in our ticketing system
[26-Jul-2010 18:31:51] example, if a lower severity event comes in for a critical (highest) level device the ticket becomes critical severity
[26-Jul-2010 18:32:00] even if the issue itself wouldn't normally be considered critical
[26-Jul-2010 18:32:03] that sort of thing
[26-Jul-2010 18:32:24] anyways, I'm out for the night, ta
[26-Jul-2010 18:44:05] thanks rmatte
[26-Jul-2010 19:14:15] so an update: someone from zenoss (eric) got back with me
[26-Jul-2010 19:14:29] They will NOT let me pay them to develop the monitor by hostname feature
[26-Jul-2010 19:14:48] UNLESS, i buy the enterprise version, and they will NOT give the feature out in the CORE version even if i do all of this
[26-Jul-2010 19:14:58] so, way to run a company that doesnt want more features for the users, or more money
[26-Jul-2010 19:15:05] seems like a fucking brilliant strategy
[26-Jul-2010 20:42:09] does that stop you from just making it a community zenpack, Parabola?
[26-Jul-2010 20:48:10] although I'm prolly misunderstanding what you're after
[26-Jul-2010 21:32:42] yeah :)
[26-Jul-2010 21:32:52] what i'm looking for is not a feature, i want to monitor a host by hostname not just IP
[26-Jul-2010 21:33:05] like zabbix can do i, i emailed your sales asking if i could pay for the feature
[26-Jul-2010 21:33:19] however, if i paid for it to be written, i wanted it to be included in the CORE release so everyone could have it
[26-Jul-2010 21:33:28] since this would be thousands of dollars
[26-Jul-2010 21:33:45] and the response i got back wasnt very um, good?
[26-Jul-2010 21:34:01] sorry for the harsh outburst, it just bothers me when open souce projects are run stupidliy like that
[26-Jul-2010 21:34:23] i mean i'm offering to pay 100 bucks an hour or w/e the rate is for this feature and he seemed 100% uninterested
[26-Jul-2010 21:50:43] mrayzenoss is now known as mray
[26-Jul-2010 22:46:18] mrayzenoss is now known as mray
[27-Jul-2010 00:00:30] [disconnected at Tue Jul 27 00:00:30 2010]
[27-Jul-2010 00:00:30] [connected at Tue Jul 27 00:00:30 2010]
[27-Jul-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]
[27-Jul-2010 02:26:55] mrayzenoss is now known as mray
[27-Jul-2010 02:38:56] *sight* damm upgrade causes me nothing but problems
[27-Jul-2010 03:57:18] *hmpf*
[27-Jul-2010 03:57:46] surfing to infrastructure, looking at the tree, i have an indicator, that i have an information event on one of the device classes.
[27-Jul-2010 03:58:30] clicking the device class, shows all those devices, but the events in this overview have only indicators for warning and above.
[27-Jul-2010 03:59:56] either don't indicate informational events in the infrastructe tree, or list them also in the events columns.
[27-Jul-2010 06:00:55] morning all
[27-Jul-2010 07:37:01] morning ya'll
[27-Jul-2010 07:46:06] morning
[27-Jul-2010 09:11:41] mrayzenoss is now known as mray
[27-Jul-2010 09:19:48] mrayzenoss is now known as mray
[27-Jul-2010 09:45:21] <_rain_man_> ok i need help
[27-Jul-2010 09:46:05] <_rain_man_> has anyone got zenoss core to work on ubuntu?
[27-Jul-2010 09:47:01] * _rain_man_ slaps cgibbons around a bit with a large trout
[27-Jul-2010 09:47:03] * _rain_man_ slaps mray around a bit with a large trout
[27-Jul-2010 09:47:06] * _rain_man_ slaps npmccallum around a bit with a large trout
[27-Jul-2010 09:47:11] <_rain_man_> WAKE UP!
[27-Jul-2010 09:47:13] rainbow?
[27-Jul-2010 09:47:31] _rain_man_: I run my test boxes on 9.10 and 10.04
[27-Jul-2010 09:47:50] setting up for the Getting Started Webinar in 13 minutes on 10.04 right now
[27-Jul-2010 09:48:13] <_rain_man_> i just get a load of errors when installing on 10.04
[27-Jul-2010 09:48:43] <_rain_man_> i've followed the documentation
[27-Jul-2010 09:48:52] <_rain_man_> any tips?
[27-Jul-2010 09:49:32] I'm using the stack installer .deb
[27-Jul-2010 09:49:49] <_rain_man_> and did you install mysql first?
[27-Jul-2010 09:49:57] no, that's included in the stack installer
[27-Jul-2010 09:50:31] <_rain_man_> so you installed ubuntu server 10.04 then just installed the stack installer?
[27-Jul-2010 09:51:12] yeah, lemme find my apt-get line
[27-Jul-2010 09:51:30] sudo apt-get install fping emacs22-nox ttf-linux-libertine ttf-liberation curl
[27-Jul-2010 09:51:42] fping was there for the fping ZenPack
[27-Jul-2010 09:51:51] emacs22-nox is there because emacs rocks
[27-Jul-2010 09:51:55] weirdo
[27-Jul-2010 09:52:08] ttf-liberation fixes an Ubuntu font issue
[27-Jul-2010 09:53:12] <_rain_man_> ok thanks, and then you run /etc/init.d/zenoss-stack start ?
[27-Jul-2010 09:53:15] emacs rocks? hahah!
[27-Jul-2010 09:53:24] _rain_man_: yeah
[27-Jul-2010 09:53:31] _rain_man_: the stack installer creates the 'zenoss' user
[27-Jul-2010 09:54:27] <_rain_man_> ok thanks alot for you help mray!
[27-Jul-2010 09:55:40] theacolyte: no, he is right, emacs moves like a rock
[27-Jul-2010 09:55:52] oh, right - right
[27-Jul-2010 09:55:54] Getting Started Webinar in 5 minutes
[27-Jul-2010 09:56:10] mray: quick, install vim before you start ;)
[27-Jul-2010 09:56:19] <_rain_man_> does that webinar show you how to install zenoss?
[27-Jul-2010 09:56:44] heh, the stack installer is trivial
[27-Jul-2010 09:56:50] very
[27-Jul-2010 09:57:06] <_rain_man_> well all i get is errors evertime i've tried to run it :(
[27-Jul-2010 09:57:15] <_rain_man_> everytime*
[27-Jul-2010 09:57:32] wget http://blah.blah.blah/stack-installer-file.bin && chmod +x stack-installer-file.bin && ./stack-installer-file.bin
[27-Jul-2010 09:57:44] what kinds of errors?
[27-Jul-2010 09:58:34] <_rain_man_> I will try again (Y) cheers for your help
[27-Jul-2010 09:59:08] . o O (maximum strange)
[27-Jul-2010 09:59:26] pretty sure he didn't know to chmod +x the file
[27-Jul-2010 09:59:34] which would explain why he was getting errors running it :P
[27-Jul-2010 09:59:44] *smash*
[27-Jul-2010 09:59:48] lol
[27-Jul-2010 10:00:37] ok, last time i was sitting in front of a w2k8 server i was maybe feeling the same. Where did this go? where is that?
[27-Jul-2010 10:01:10] lol
[27-Jul-2010 10:01:23] that's like the control panel between XP and Vista
[27-Jul-2010 10:01:37] I was thinking to myself "Why the hell did they have to rename Add/Remove Programs!?"
[27-Jul-2010 10:01:52] *that* is the only thing that really messes me up
[27-Jul-2010 10:02:39] cgibbons: a startmenu with a searchbox, doesn't?
[27-Jul-2010 10:02:47] nope
[27-Jul-2010 10:03:14] oh, and the annoying elevated command-prompt stuff
[27-Jul-2010 10:03:30] yeh
[27-Jul-2010 10:03:33] that too
[27-Jul-2010 10:03:56] yeh, and the stupid search box instead of a run prompt
[27-Jul-2010 10:04:15] it does the same thing basically, it's just slower
[27-Jul-2010 10:04:18] hi guys
[27-Jul-2010 10:04:22] hello
[27-Jul-2010 10:04:38] rmatte: you didn't quote zenos3 User Interface?
[27-Jul-2010 10:04:58] fragfutter: I'm not finding it too bad in terms of navigating honestly
[27-Jul-2010 10:05:05] I've found where pretty much everything is
[27-Jul-2010 10:05:15] it has bugs obviously
[27-Jul-2010 10:05:22] I hope to see most of them fixed by 3.0.1
[27-Jul-2010 10:05:34] and I'll still probably hold out to 3.0.2 before upgrading
[27-Jul-2010 10:05:42] i'm guessing the ones you want won't be, rmatte :) you know where hte bug fix list is right?
[27-Jul-2010 10:05:44] rmatte: besides relying to heavy on ajax, i think it has some UI breaks that shouldn't happen
[27-Jul-2010 10:05:56] the biggest problem I have with the new interface is the lack of the ability to customize the component views
[27-Jul-2010 10:06:10] for instance, the interface description column is gone, but there's no way to add it
[27-Jul-2010 10:06:16] or to remove other columns to make room for it
[27-Jul-2010 10:06:49] cgibbons: I would assume in trac under reports
[27-Jul-2010 10:06:58] yeah, lemme get you the right one
[27-Jul-2010 10:07:02] k
[27-Jul-2010 10:07:30] http://bit.ly/bh2JgP
[27-Jul-2010 10:07:34] cgibbons: I'm a patient person, I can put up with 2.5.2 for a while since it works properly
[27-Jul-2010 10:07:36] i filed a new UI today http://dev.zenoss.org/trac/ticket/7111
[27-Jul-2010 10:07:49] thanks for the link
[27-Jul-2010 10:08:14] cgibbons: this was one that I wanted fixed: http://dev.zenoss.org/trac/ticket/6694
[27-Jul-2010 10:08:17] glad it will be
[27-Jul-2010 10:08:33] yeah that one was ugly. just took a while to get to in the rush to get 3.0 out.
[27-Jul-2010 10:08:41] yeh, I figured
[27-Jul-2010 10:08:47] you should have seen how bad it was with the Cisco UCS enterprise ZenPack. It has a dozen or so very ugly internal component meta names.
[27-Jul-2010 10:09:05] yeh, Matt was mentioning that at some point
[27-Jul-2010 10:09:28] cgibbons: your Call Manager pack doesn't also monitor Cisco Unity does it?
[27-Jul-2010 10:09:34] no idea
[27-Jul-2010 10:09:35] (the enterprise one)
[27-Jul-2010 10:09:36] bigegor has (or had) internal names for a lot of components
[27-Jul-2010 10:09:37] k
[27-Jul-2010 10:09:49] Egor's names are pretty standard
[27-Jul-2010 10:09:57] he didn't actually add any
[27-Jul-2010 10:10:03] he just used the existing ones
[27-Jul-2010 10:10:10] maybe he added 1 or 2
[27-Jul-2010 10:10:14] so you just want the description column in the IP Interface grid, rmatte?
[27-Jul-2010 10:10:33] cgibbons: yeh, but I don't really get how that can be done without making it look cluttered as all hell
[27-Jul-2010 10:10:38] rmatte: yes but they are "raw" snmp field names ;)
[27-Jul-2010 10:10:38] leseee
[27-Jul-2010 10:10:59] fragfutter: yeh, but that can easily be fixed with the changes in that trac ticket
[27-Jul-2010 10:11:08] fragfutter: they added a pretty name map for them
[27-Jul-2010 10:11:15] nice.
[27-Jul-2010 10:11:21] It would take Egor like 2 seconds to fix
[27-Jul-2010 10:11:50] cgibbons: that side menu is the main issue, takes up 1/4 of the screen
[27-Jul-2010 10:12:15] cgibbons: can we have a configurable grid in 3.1? Like i for component xy i want columns with the property a, b, d and g?
[27-Jul-2010 10:12:16] cgibbons: too bad there isn't some quick hide button for the menu which would also trigger a more detailed view
[27-Jul-2010 10:12:35] yeah the little hide/show arrows most of those navs have. It was in the specs, I think time ran out.
[27-Jul-2010 10:12:36] show/hide columns depending on whether the menu is hidden or not
[27-Jul-2010 10:12:47] would be sufficient
[27-Jul-2010 10:13:57] configurable columns would be nice, but I understand how complex it would be to code
[27-Jul-2010 10:14:25] sortable columns is something I'd really like to see at some point
[27-Jul-2010 10:14:35] but also not really necessary right now
[27-Jul-2010 10:15:00] ah yes indeed. they are in fact sortable, but a lot of the ones that would be useful sortable are not yet enabled because of some other bug, etc...
[27-Jul-2010 10:15:09] ah
[27-Jul-2010 10:15:15] gotcha
[27-Jul-2010 10:16:11] lol, this one's a biggy: http://dev.zenoss.org/trac/ticket/7101
[27-Jul-2010 10:16:12] hmm is IP interface description something that the device admin usually keys in and Zenoss will just pick up via SNMP?
[27-Jul-2010 10:16:24] cgibbons: that's exactly right
[27-Jul-2010 10:16:30] it's keyed in on the device itself
[27-Jul-2010 10:16:38] ok, that may be why it got ignored. most of our test devices have it as empty.
[27-Jul-2010 10:16:44] ah
[27-Jul-2010 10:16:49] yeh, it's a pretty essential thing
[27-Jul-2010 10:17:13] if you're managing a switch with over 300 ports on it, you kind of want to know what each of them is
[27-Jul-2010 10:17:27] indeed
[27-Jul-2010 10:17:47] I got my modeler plugin working for mapping the interface names and descriptions for catalyst switches
[27-Jul-2010 10:17:52] rmatte: is Gigbit9/34, not sufficient :)
[27-Jul-2010 10:17:56] going to be making a Cisco Catalyst ZenPack
[27-Jul-2010 10:18:09] fragfutter: no, it's not :P
[27-Jul-2010 10:18:33] rmatte: come on, with a little bit of training, you can remember all of them.
[27-Jul-2010 10:18:38] haha
[27-Jul-2010 10:18:42] I'll give you a quick patch for doing that.
[27-Jul-2010 10:19:08] I've seen the quick patch before...
[27-Jul-2010 10:19:20] I'm just waiting for it to formally make it in before I upgrade
[27-Jul-2010 10:19:48] but with Mac and IP in there, and no way to quickly hide the menu, it's going to be really crowded
[27-Jul-2010 10:19:55] was it opened as a defect/enhancement already then?
[27-Jul-2010 10:20:18] I think it was, if not I'd be happy to open a ticket for it
[27-Jul-2010 10:21:45] http://dev.zenoss.org/trac/ticket/7076
[27-Jul-2010 10:21:52] there it is
[27-Jul-2010 10:22:45] how can i find which version im zenoss i'm running?
[27-Jul-2010 10:22:51] cool - you can add this as a patch in the ticket and it'll get a lot more attention at the defect review: http://pastie.org/1062217
[27-Jul-2010 10:23:09] Ian kind of already added it
[27-Jul-2010 10:23:31] not checked in anyway :)
[27-Jul-2010 10:23:31] I'll add a comment to the ticket though about the menu hiding
[27-Jul-2010 10:23:43] i can't seem to find it looking though the menus
[27-Jul-2010 10:24:00] I'd open another enhancement request too about choosing which columns to display. That should be there as a feature.
[27-Jul-2010 10:24:31] Zfield, at the login screen for one
[27-Jul-2010 10:24:48] ahhh thnx
[27-Jul-2010 10:25:06] i have auto-login enabled, so my first screen is really the dashboard lol
[27-Jul-2010 10:25:13] or, more specifically, settings - versions
[27-Jul-2010 10:25:40] Is that a setting in Zenoss? Autologin?
[27-Jul-2010 10:25:58] is it worth upgrading to zenoss 3 or doing a fresh install - i only have 5 devices being monitored atm
[27-Jul-2010 10:26:02] but i will be adding everythign shortly
[27-Jul-2010 10:26:37] Zfield, I'm not going to v3 yet...
[27-Jul-2010 10:27:01] so you would recommend a begginer to just wait for all the bugs to be fixed/
[27-Jul-2010 10:27:18] If you don't have anything in the system, might as well do a fresh v3 install
[27-Jul-2010 10:27:21] I am
[27-Jul-2010 10:27:29] Depends on what you're using it for
[27-Jul-2010 10:27:44] Just playin around with it, no prob
[27-Jul-2010 10:27:54] In production? May want to wait.
[27-Jul-2010 10:27:55] having server 2k3/8 reporting to it
[27-Jul-2010 10:28:14] production
[27-Jul-2010 10:28:15] That doesn't really make a difference
[27-Jul-2010 10:29:13] Up to you really. Definitely want to play with it before rolling it out.
[27-Jul-2010 10:29:31] yeh, people always seem to have a production instance with no lab instance
[27-Jul-2010 10:29:40] A lab instance is absolutely essential in my opinion
[27-Jul-2010 10:30:02] heck, I have 3 lab instances
[27-Jul-2010 10:30:08] rmatte: I bring up a lab as needed, and try not to be on the bleeding edge
[27-Jul-2010 10:30:32] speaking of bleeding edge, i hate shitty cases that make you bleed =\
[27-Jul-2010 10:30:33] Ha, we just set ourselves up a lab for router & phone configs
[27-Jul-2010 10:30:35] twm1010: yeh, I usually just have 1 lab for zenpack development, but I have 3 right now because I'm doing various upgrade testing
[27-Jul-2010 10:30:46] Zenoss really isn't used yet haha
[27-Jul-2010 10:30:51] 2 of them will be killed when I'm done with them
[27-Jul-2010 10:30:59] rmatte: do you use WMI datasource? that's really the only thing I need to make sure is working before i go to 3.0
[27-Jul-2010 10:31:09] we're taking a look at enterprise tomorrow tho
[27-Jul-2010 10:31:21] twm1010: I actually just just started testing that out, since there were hints that we may be going that direction for our windows monitoring
[27-Jul-2010 10:31:28] I tested it in my 3.0 lab and it worked fine
[27-Jul-2010 10:31:32] I was able to model a server
[27-Jul-2010 10:31:59] I had to reinstall the pack after the upgrade to get it to work properly
[27-Jul-2010 10:32:05] other than that, fairly smooth
[27-Jul-2010 10:32:25] you should install version 2.7 of the pack PRIOR to an upgrade, then upgrade, and then install it again after the upgrade
[27-Jul-2010 10:32:30] then it'll work
[27-Jul-2010 10:32:31] ok
[27-Jul-2010 10:32:33] good to know
[27-Jul-2010 10:32:39] i think i use v1.5 right now
[27-Jul-2010 10:32:44] ah, that's ancient
[27-Jul-2010 10:32:56] v1.5 didn't even support full modelling via WMI as I recall
[27-Jul-2010 10:32:57] i knew there were newer, but i'm very much if it ain't broke don't fix it
[27-Jul-2010 10:33:13] yeah, i use a mix config, snmp with informant for most stats
[27-Jul-2010 10:33:23] with v 2.7 you just throw the device in to /Devices/CIM/WMI and model it
[27-Jul-2010 10:33:25] wmi datasource for things like SQL, exchange, terminal services
[27-Jul-2010 10:33:27] and it gets fully modelled with WMI
[27-Jul-2010 10:34:11] i'll have to test that out then. under /Server/Windows tho, i have a ton of subclasses
[27-Jul-2010 10:34:30] the modelling works in /CIM/WMI by default
[27-Jul-2010 10:34:39] I don't think it works outside of that class
[27-Jul-2010 10:34:42] with the way Egor coded it
[27-Jul-2010 10:34:47] you'd have to move the sub classes over
[27-Jul-2010 10:34:49] yeah, i would have to recreate or copy some templates over from there
[27-Jul-2010 10:34:51] and hope that they work properly
[27-Jul-2010 10:35:00] yup
[27-Jul-2010 10:35:18] anybody on 2.5 that uses LDAP authentication?
[27-Jul-2010 10:35:25] gwb2351: I am
[27-Jul-2010 10:35:47] whoops, forgot we talked about this before (unable to change ZenUser information)
[27-Jul-2010 10:36:00] what do you mean by unable to change it?
[27-Jul-2010 10:36:07] lay out what exactly you're trying to do
[27-Jul-2010 10:36:20] you didn't really go in to specifics yesterday
[27-Jul-2010 10:36:39] looking at /ZenUsers/username
[27-Jul-2010 10:36:46] trying to remove global role
[27-Jul-2010 10:36:54] (zenuser is currently selected)
[27-Jul-2010 10:36:59] click on it (to unselect)
[27-Jul-2010 10:37:20] any password I type in the "Enter your password to confirm changes" no changes are saved
[27-Jul-2010 10:38:12] wait, what!?!
[27-Jul-2010 10:38:17] it works now...
[27-Jul-2010 10:38:41] rmatte fixed it with his mind...
[27-Jul-2010 10:38:49] haha
[27-Jul-2010 10:40:28] * rmatte has super ESP Zenoss powers
[27-Jul-2010 10:46:19] They'll definitely hire you now bro ;)
[27-Jul-2010 10:46:30] we shall see
[27-Jul-2010 10:48:29] What do you want to do for them? Code or do QA?
[27-Jul-2010 10:48:48] I'd want to do engineering work
[27-Jul-2010 10:48:55] customer support basically
[27-Jul-2010 10:49:10] (which is what I do in here daily for free basically anyways :P)
[27-Jul-2010 10:49:18] :)
[27-Jul-2010 10:49:39] I like configuring/customizing Zenoss, it's fun
[27-Jul-2010 10:49:52] Nice
[27-Jul-2010 10:50:14] it's neat seeing what you can squeeze out of it
[27-Jul-2010 10:50:20] Yea, I hope we pony up for the enterprise version for the support.
[27-Jul-2010 10:50:29] twm1010: it's relatively easy to useWMI for any device class. Simply set the zCollectorPlugins property like /CIM/WMI, copy all performance templates from /CIM/WMI to your device class, set zSnmpMonitorIgnore to True and zWmiMonitorIgnore to False, don't foget zWinUser and zWinPasword
[27-Jul-2010 10:50:33] for instance, one of our clients wanted us to only alert on ports that contained a certain prefix in the port descriptions
[27-Jul-2010 10:50:35] Granted, you have given me ridiculous support...
[27-Jul-2010 10:50:48] so I wrote up a quick transform, and voila
[27-Jul-2010 10:50:55] Nice
[27-Jul-2010 10:51:15] I try, I figure what goes around comes around, I usually end up getting support back here and there when I need it
[27-Jul-2010 10:51:25] so it works out
[27-Jul-2010 10:51:26] Yea this software is very powerful. With great power also comes great complexity :P. I'm learning...
[27-Jul-2010 10:51:33] For sure
[27-Jul-2010 10:51:38] yeh, it is quite complex
[27-Jul-2010 10:51:53] in terms of configuration anyways
[27-Jul-2010 10:51:58] I started out very green with the ubuntuforums, and now I help more then I ask :D
[27-Jul-2010 10:52:08] the thing that really makes Zenoss incredibly powerful is the transforms system
[27-Jul-2010 10:52:22] the sky is the limit with it
[27-Jul-2010 10:52:28] Yea, I haven't even played with any of that.
[27-Jul-2010 10:52:28] I added the patch at http://dev.zenoss.org/trac/ticket/7076 (and separately tried http://pastie.org/1062217), but with either of them applied I still don't get a description column. In both cases I restarted zope (and even tried restarting the whole zenoss stack). Is there something else I need to restart/delete/etc.?
[27-Jul-2010 10:52:39] you know any python?
[27-Jul-2010 10:52:52] if not, be prepared to learn
[27-Jul-2010 10:52:58] Lol
[27-Jul-2010 10:53:03] it's honestly not that hard to learn the basics to write transforms
[27-Jul-2010 10:53:03] I know a little
[27-Jul-2010 10:53:11] I didn't know any python when I started out with Zenoss
[27-Jul-2010 10:53:12] But I don't use it that frequently.
[27-Jul-2010 10:53:20] rmatte: ZenPacks are making Zenoss incredibly powerful ;)
[27-Jul-2010 10:53:27] bigegor: that too
[27-Jul-2010 10:53:50] bigegor: I was just saying that transforms offer flexibility that I haven't seen in any other monitoring product
[27-Jul-2010 10:53:51] Same with perl. Great language, endless possibilty.... but I just don't use it that often anymore.
[27-Jul-2010 10:54:08] I know enough perl to be dangerous
[27-Jul-2010 10:54:10] lol
[27-Jul-2010 10:54:21] I much prefer python
[27-Jul-2010 10:54:24] arrrghhh: that word shouldn't be used in polite company ;)
[27-Jul-2010 10:54:40] +1 for python
[27-Jul-2010 10:54:44] wtucker: lol
[27-Jul-2010 10:54:44] Ha, what... perl?
[27-Jul-2010 10:54:52] Not a fan eh?
[27-Jul-2010 10:54:55] rmatte: http://geekandpoke.typepad.com/geekandpoke/perl/
[27-Jul-2010 10:55:05] * wtucker gets the soap
[27-Jul-2010 10:55:07] I love the xkcd comic where god admits that they hacked most of the universe together with perl
[27-Jul-2010 10:55:23] rmatte: except flying, that is done with python
[27-Jul-2010 10:55:27] http://xkcd.com/224/
[27-Jul-2010 10:55:58] rmatte: http://xkcd.com/353/
[27-Jul-2010 10:56:12] arrrghhh: used to love it, but I don't like the way it handles data types - and objects are really ugly in perl
[27-Jul-2010 10:56:19] fragfutter: awesome
[27-Jul-2010 10:56:39] lol
[27-Jul-2010 10:56:44] morning all
[27-Jul-2010 10:57:13] morning
[27-Jul-2010 10:57:52] rmatte: did ian's patch on #7076 work for you?
[27-Jul-2010 10:58:39] is there a python equivalent of the "touch" command?
[27-Jul-2010 10:59:02] wtucker: haven't tried it, I would imagine that it does work
[27-Jul-2010 10:59:03] mducharme-laptop: create a file or update a timestamp?
[27-Jul-2010 10:59:09] mducharme-laptop: to create a file or to change its date?
[27-Jul-2010 10:59:16] fragfutter: jinx
[27-Jul-2010 10:59:24] !
[27-Jul-2010 10:59:25] to create a file
[27-Jul-2010 10:59:46] open(file, "w").close()
[27-Jul-2010 10:59:53] mducharme-laptop: just use file() (or open()) and discard the handle
[27-Jul-2010 11:00:28] that's of course assuming that the file does not already exist
[27-Jul-2010 11:00:48] are you trying to do a lockfile?
[27-Jul-2010 11:01:47] rmatte wrote this script that writes to a log file which is fine, but if the log file does not exist the script does not run, so the log file has to be created manually
[27-Jul-2010 11:01:48] rmatte: k. Neither it nor the one that cgibbons pasted a link to earlier work for me... but it's quite possible that I missed something
[27-Jul-2010 11:02:02] I would rather change the script so that the log file gets auto created if it does not exist
[27-Jul-2010 11:02:29] mducharme-laptop: is the script in python?
[27-Jul-2010 11:02:30] mducharme-laptop: find the point where it is opened and open it with "a"
[27-Jul-2010 11:03:00] mducharme-laptop: the logfile is created automatically via the bash script
[27-Jul-2010 11:03:17] though I guess you're looking to create it via the daemon itself?
[27-Jul-2010 11:03:24] yes
[27-Jul-2010 11:03:41] I see
[27-Jul-2010 11:03:52] just to make the setup a bit simpler
[27-Jul-2010 11:04:03] * rmatte shrugs
[27-Jul-2010 11:04:08] I thought it was fairly simple lol
[27-Jul-2010 11:05:37] hehe
[27-Jul-2010 11:05:45] well yeah but I'm thinking again in terms of public consumption
[27-Jul-2010 11:05:56] I could manually go to a command prompt and go into the log directory
[27-Jul-2010 11:06:09] and touch the file
[27-Jul-2010 11:06:18] that's not what I mean
[27-Jul-2010 11:06:55] cat $ZENHOME/bin/zenticket
[27-Jul-2010 11:07:07] that's part of the zenpack
[27-Jul-2010 11:07:15] and it creates the config file the first time you start the daemon
[27-Jul-2010 11:07:26] or if the config file goes missing
[27-Jul-2010 11:10:00] no such file or directory
[27-Jul-2010 11:10:14] you did actually install the pack correctly right?
[27-Jul-2010 11:10:20] yup
[27-Jul-2010 11:10:22] that link should have been automatically created upon installation
[27-Jul-2010 11:10:41] wait...
[27-Jul-2010 11:10:45] are you the zenoss user right now
[27-Jul-2010 11:10:48] or some other user?
[27-Jul-2010 11:10:59] you need to be the zenoss user for $ZENHOME to resolve to a directory
[27-Jul-2010 11:11:00] zenoss user
[27-Jul-2010 11:11:02] I know
[27-Jul-2010 11:11:14] k, then I don't see how you could have installed the pack properly if that link is missing
[27-Jul-2010 11:11:26] i will release next version of WMIPerf_Windows ZenPack, can somebody test actual SVN version of this ZenPack
[27-Jul-2010 11:11:29] after installing the pack you should be able to do: zenticket start
[27-Jul-2010 11:11:33] and have it start the daemon
[27-Jul-2010 11:11:45] if you can't do that, then it's not installed properly
[27-Jul-2010 11:12:12] I can't do that
[27-Jul-2010 11:12:28] anyways...
[27-Jul-2010 11:12:34] in the ZenPack's zenpack directory
[27-Jul-2010 11:12:37] check out...
[27-Jul-2010 11:12:41] ZenPacks.Nova.Ticket.Daemon-2.1-py2.4.egg/ZenPacks/Nova/Ticket/Daemon/daemons/zenticket
[27-Jul-2010 11:12:52] that's the script that normally starts the daemon and creates the config file
[27-Jul-2010 11:13:02] you don't need to have the daemon itself create the config file
[27-Jul-2010 11:13:04] yes, I see that file there
[27-Jul-2010 11:13:05] there's no point
[27-Jul-2010 11:13:37] and that file gets linked to $ZENHOME/bin/zenticket automatically upon successful installation of the pack
[27-Jul-2010 11:13:45] so if it didn't get linked, then it's not installed properly
[27-Jul-2010 11:13:56] ok
[27-Jul-2010 11:14:05] the proper way to edit the pack is to do the following...
[27-Jul-2010 11:14:09] install the pack normally
[27-Jul-2010 11:14:12] for starters
[27-Jul-2010 11:14:21] no linking or any of that nonsense
[27-Jul-2010 11:14:34] I'm installing it through the web interface
[27-Jul-2010 11:14:35] using install zenpack
[27-Jul-2010 11:14:38] that's not the normal way?
[27-Jul-2010 11:15:05] that should work
[27-Jul-2010 11:15:14] how did you convert the pack to developer mode?
[27-Jul-2010 11:15:27] there was a cp command on the website
[27-Jul-2010 11:15:42] copying files from the zentemplate directory or something to the zenpack dir
[27-Jul-2010 11:15:47] ok, I don't understand how the link didn't get created in $ZENHOME/bin
[27-Jul-2010 11:16:01] what exactly is supposed to create the link?
[27-Jul-2010 11:16:04] I've tested the pack in 2.5 and 3.0, the link always gets created
[27-Jul-2010 11:16:08] zenoss does it itself
[27-Jul-2010 11:16:20] there's an automated routine to create a link to whatever is in the daemons directory
[27-Jul-2010 11:16:37] maybe it's because I'm now overriding class zenpack
[27-Jul-2010 11:16:44] probably
[27-Jul-2010 11:16:48] to set zproperties
[27-Jul-2010 11:17:05] you'll probably have to code in the linking yourself then
[27-Jul-2010 11:17:18] let me get you an example of what to put in __init__.py...
[27-Jul-2010 11:17:22] ok
[27-Jul-2010 11:17:22] bigegor: i have an issue with the hp-proliant pack. Sometimes i get flukes. That is it tells me all fans failed (status other). Next check the error is gone.
[27-Jul-2010 11:18:12] mducharme-laptop: http://fpaste.org/O5fV/
[27-Jul-2010 11:18:30] that's what I use to link my zencollectortool script in to $ZENHOME/bin
[27-Jul-2010 11:18:57] you'll obviously want to change self.path('bin') to 'daemon'
[27-Jul-2010 11:19:00] ok
[27-Jul-2010 11:19:01] right
[27-Jul-2010 11:19:11] and zencollectortool to zenticket
[27-Jul-2010 11:19:19] I'm surprised it overrides the linking
[27-Jul-2010 11:19:42] oh wait
[27-Jul-2010 11:19:47] crap, you'll need even more than that
[27-Jul-2010 11:19:57] because zenoss also somehow adds it to a daemons list somewhere
[27-Jul-2010 11:20:04] so that it knows to start it when you do a zenoss start
[27-Jul-2010 11:20:12] not sure where the list that it adds it to is
[27-Jul-2010 11:20:13] fragfutter: hmm. can it be SIM reported that status?
[27-Jul-2010 11:20:28] I think overriding the class isn't necessarily the way to go
[27-Jul-2010 11:21:28] bigegor: system insight manager? should have nothing to do with it.
[27-Jul-2010 11:21:58] bigegor: i have no traps regarding it in SIM.
[27-Jul-2010 11:23:21] fragfutter: i mean SIM Agent on server. HP Proliant ZenPack recived components status from SIM agent
[27-Jul-2010 11:24:10] bigegor: could be. but i wouldn't know why.
[27-Jul-2010 11:24:21] mducharme-laptop: there has to be a way to create those zProperties without screwing up the linking that Zenoss does
[27-Jul-2010 11:25:12] rmatte I only know the way that mray gave me
[27-Jul-2010 11:25:32] bigegor: if the snmp-agents fluke i expect them to also show up in SIM.
[27-Jul-2010 11:25:40] but it does appear to add it to the daemons list
[27-Jul-2010 11:25:45] meaning if I go into my daemons screen, it is there
[27-Jul-2010 11:27:00] the daemon starts now
[27-Jul-2010 11:27:01] ah
[27-Jul-2010 11:27:05] then you just need to make the link
[27-Jul-2010 11:27:19] yes, I did the updates based on the paste you sent me
[27-Jul-2010 11:27:29] now it does it correctly and starts
[27-Jul-2010 11:27:54] fragfutter: you can try to increase 'Escalate Count' for status threshold
[27-Jul-2010 11:29:25] general zenoss question - in the google map, why does the mouseover no longer work when you change the zoom?
[27-Jul-2010 11:29:52] mducharme-laptop: the mouseover has always been very finicky, no idea
[27-Jul-2010 11:30:02] bigegor: and now i need to find that in zenoss3
[27-Jul-2010 11:30:58] I have 3 places so cloes to each other on my google map they are almost overlapping
[27-Jul-2010 11:31:07] when I zoom in to see them better the mouseover no longer works
[27-Jul-2010 11:31:14] it's slightly irritating
[27-Jul-2010 11:31:40] log it as a bug
[27-Jul-2010 11:32:26] fragfutter: check HPFan and HPsdFan performance templates
[27-Jul-2010 11:32:34] k
[27-Jul-2010 11:33:23] bigegor: they should be at Advanced->MonitoringTemplates
[27-Jul-2010 11:33:37] bigegor: but they seem to be broken with zenoss3. AttributeError
[27-Jul-2010 11:34:55] fragfutter: hmm, let me look.
[27-Jul-2010 11:36:33] I'm setting mine up like your setup at first rmatte
[27-Jul-2010 11:36:46] that way I can make sure the calls are working before I start changing things
[27-Jul-2010 11:40:07] only thing I'm going to change is where you call the script, I'm just going to log that
[27-Jul-2010 11:41:27] how do i customize/overwrite a monitoring template in zenoss3?
[27-Jul-2010 11:42:22] fragfutter: I'm opening a P1 ticket about that craptacular workflow right now
[27-Jul-2010 11:42:36] wanna do for a device or a device class?
[27-Jul-2010 11:42:43] device class
[27-Jul-2010 11:42:49] (actualy both)
[27-Jul-2010 11:43:43] If you're on a device you can click on teh template you want to override and then bring up the gear menu and you'll see Override Template Here.
[27-Jul-2010 11:44:13] for device classes they'd be in Advanced -> Monitoring Templates - you're getting an error there?
[27-Jul-2010 11:44:56] cgibbons: the error is only for bigegor tempaltes ;) i want to customize network graph (zenos default)
[27-Jul-2010 11:45:21] ethernetCsmacd
[27-Jul-2010 11:47:11] ok yeah, go there, select an instance of the template (prolly the one in /Devices) and then you can choose a target device class for it to be overridden at.
[27-Jul-2010 11:47:52] ah, gear at the bottom, override template, choose target
[27-Jul-2010 11:47:53] I have 2 issues with that workflow that I've brought up, but I'm not sure how many people will have htem in the real world. One is that you'd have to already know where you were gonna put it and pre-create the device class. Seems l ike the device class override would make sense the same way the device override would, too (although having one place to do it all is nice, too).
[27-Jul-2010 11:52:25] cgibbons: issue one) in the infrastructure view i surf to some device-class. I see "oh, all these need to have blue curves for the graphs". Now i need to go to the Advanced tab, find the component, click customize, remember where i was in the infrastructure tab, place it. Zenoss2 i would have clicked the folder, selected the template i inherited and clicked customize
[27-Jul-2010 11:52:55] cgibbons: issue two) i can' easily see which templates are active in which folder, or which custom templates are attached to a folder
[27-Jul-2010 11:53:17] yeah cuz you h ave to go back to the device class/folder to see it. i agree.
[27-Jul-2010 11:53:43] i found another UI bug: http://dev.zenoss.org/trac/ticket/7120
[27-Jul-2010 11:54:15] getting nested frames when i check the overriding objects
[27-Jul-2010 11:54:24] ugh
[27-Jul-2010 11:54:55] a great deal of our bug fixing in the beta time frame work bugs like that. intersection between old backwards-compatiblity pages that weren't rewritten yet and the new stuff.
[27-Jul-2010 11:55:40] btw it is inconsisten to have a template at "/Devices" and a customized at "/Server/Cmd". It should be "/" and "/Server/cmd" or "/Devices" and "/Devices/Server/cmd"
[27-Jul-2010 11:56:11] yep
[27-Jul-2010 11:56:21] ugh
[27-Jul-2010 11:56:22] http://dev.zenoss.org/trac/attachment/ticket/7121/
[27-Jul-2010 11:57:19] subscribed
[27-Jul-2010 11:59:24] if i open a bug "give me a plain html 1.0 frameset without ajax but server rendered tal/metal templates" can i have it?
[27-Jul-2010 12:01:45] not gonna happen
[27-Jul-2010 12:02:03] cgibbons: damn.
[27-Jul-2010 12:04:26] and another: http://dev.zenoss.org/trac/ticket/7122
[27-Jul-2010 12:05:51] enough bugs for the day
[27-Jul-2010 12:05:55] * fragfutter is heading home
[27-Jul-2010 12:09:09] eugh, another stupid issue to figure out
[27-Jul-2010 12:09:19] zenping isn't able to ping a device but I can ping the device fine using the ping command
[27-Jul-2010 12:09:33] figures
[27-Jul-2010 12:10:08] going to grab lunch first, afk
[27-Jul-2010 12:14:26] back
[27-Jul-2010 12:17:54] 3.0 hit release? When did that happen? :)
[27-Jul-2010 12:18:09] like over a week ago
[27-Jul-2010 12:18:17] last Thursday
[27-Jul-2010 12:20:02] that's awesome
[27-Jul-2010 12:20:10] that means my procrastinating by a week and a half
[27-Jul-2010 12:20:10] the easiest way is the %s method
[27-Jul-2010 12:20:12] oops
[27-Jul-2010 12:20:16] SAVED me a ton of work.
[27-Jul-2010 12:22:20] 3.0.1 should be RSN now, too
[27-Jul-2010 12:22:32] er extra words whatever, same idea
[27-Jul-2010 12:28:05] <|rt|> how do you keep zenoss from deduping an event? My IPS sends a trap to Zenoss...but if I haven't cleared the previous event then it dedups the alerts b/c it thinks it's the same alert
[27-Jul-2010 12:28:31] <|rt|> I guess I can use the source IP address to differentiate the alerts
[27-Jul-2010 12:29:52] * mray off to lunch
[27-Jul-2010 12:32:29] Is it fairly easy to talk to NRPE from ZenOSS?
[27-Jul-2010 12:43:06] |rt|: you should write a transform that makes the trap more meaningful when it comes in
[27-Jul-2010 12:43:22] |rt|: as long as there's different data in the summary it'll treat it as a separate event
[27-Jul-2010 12:47:31] my god is frustrating to use the UI in infra-> Networks
[27-Jul-2010 12:48:08] it continuously changes focus to Networks when navigating the IPs on the Display tab
[27-Jul-2010 12:50:33] JonnyV: I'm in there now and I don't see what you mean
[27-Jul-2010 12:50:52] i have a few nets and subnets on the left
[27-Jul-2010 12:51:03] i choose one that has over 100 ips
[27-Jul-2010 12:51:04] right, so do I
[27-Jul-2010 12:51:17] they show at the bottom on the right
[27-Jul-2010 12:51:22] right
[27-Jul-2010 12:51:27] ok..
[27-Jul-2010 12:51:34] try to navigate through the ips
[27-Jul-2010 12:51:41] <|rt|> rmatte yeah I read that in the Zenoss Core Network and System Monitoring book....how do I access some of the other bits of information using TALES?
[27-Jul-2010 12:51:51] what do you mean by navigate through them?
[27-Jul-2010 12:52:01] use the arrows to go down
[27-Jul-2010 12:52:12] or pgdown
[27-Jul-2010 12:52:16] |rt|: what info are you referring to?
[27-Jul-2010 12:52:19] <|rt|> The 2 bits of information that I'd like to include in teh summary is the fgIPsTrapSigMsg and fgIpsTrapSrcIP
[27-Jul-2010 12:52:28] ok...
[27-Jul-2010 12:52:36] |rt|: 2 secs, I'll cook up an example
[27-Jul-2010 12:52:37] <|rt|> these comes with the snmp trap
[27-Jul-2010 12:52:45] <|rt|> ok thanks
[27-Jul-2010 12:52:53] the IP i was looking was a couple of pgdowns
[27-Jul-2010 12:53:33] and once i got it i did a few up and the focus went to the networks and change the subnet i was looking
[27-Jul-2010 12:53:39] so had to start all over again
[27-Jul-2010 12:53:42] hehe
[27-Jul-2010 12:54:34] |rt|: http://fpaste.org/dCOH/
[27-Jul-2010 12:55:09] JonnyV: ohhhhh, I see what you mean, the focus remains on the left
[27-Jul-2010 12:55:22] yeaa
[27-Jul-2010 12:55:23] hehe
[27-Jul-2010 12:55:44] u have to continuously click on the IPs windows
[27-Jul-2010 12:56:05] so the focus remains there
[27-Jul-2010 12:56:20] glad u understood me.. hehe..
[27-Jul-2010 12:56:37] nice
[27-Jul-2010 12:56:42] well, you could just scroll with your mouse lol
[27-Jul-2010 12:56:44] but that is weird
[27-Jul-2010 12:57:21] haha
[27-Jul-2010 12:57:57] definitely open it that as a bug. and be sure to mention 508 somewhere.
[27-Jul-2010 12:57:59] rmatte: where would I go to add an arbitrary command to be checked?
[27-Jul-2010 12:58:04] |rt|: the %s get replaced with the contents of the variables, understand?
[27-Jul-2010 12:58:11] I have scripts on the ZenOSS box that queries remote NRPE
[27-Jul-2010 12:58:31] straterra: you need to create a command datasource
[27-Jul-2010 12:58:39] straterra: that's detailed in the admin guide
[27-Jul-2010 12:58:52] kk
[27-Jul-2010 12:59:40] <|rt|> let me have a look
[27-Jul-2010 12:59:59] What term should I look for in the document? command datasource doesn't show anything
[27-Jul-2010 13:00:26] <|rt|> rmatte: what's the 3rd parameter in that getattr function call?
[27-Jul-2010 13:00:47] <|rt|> that says 'Unknown' do I need to replace that with the event class?
[27-Jul-2010 13:04:39] |rt|: that just means that if it's not able to actually grab a value it will set the variable to "Unknown" instead of just empty
[27-Jul-2010 13:04:43] it's optional
[27-Jul-2010 13:04:58] <|rt|> ah ok...now is that python or TALES?
[27-Jul-2010 13:05:03] python
[27-Jul-2010 13:05:08] TALES is not used in transforms
[27-Jul-2010 13:05:14] transforms are pure python
[27-Jul-2010 13:05:14] <|rt|> ah ok
[27-Jul-2010 13:05:25] <|rt|> thanks again
[27-Jul-2010 13:05:28] TALES is for things like reports
[27-Jul-2010 13:05:29] np
[27-Jul-2010 13:06:41] <|rt|> there's not a way to test a transform with a test event is there? b/c those variables won't have a value....well I guess it will just say unknown at that point
[27-Jul-2010 13:07:05] I don't think transforms get applied to test events
[27-Jul-2010 13:07:07] in my experience
[27-Jul-2010 13:07:12] only real way to test is with a real event
[27-Jul-2010 13:07:27] especially for things like traps
[27-Jul-2010 13:07:33] <|rt|> I guess we sit and wait :)
[27-Jul-2010 13:07:41] show me your end code
[27-Jul-2010 13:07:48] I can tell you whether it'll work or not
[27-Jul-2010 13:07:54] this is a really simple transform
[27-Jul-2010 13:08:10] <|rt|> yeah I only edited the text in the summary string from what you sent
[27-Jul-2010 13:08:11] also, when you save a transform, if the code is bad the text will turn red
[27-Jul-2010 13:08:17] if that happens, you have some sort of error in the code
[27-Jul-2010 13:08:19] just a tip
[27-Jul-2010 13:08:24] <|rt|> ok good to know
[27-Jul-2010 13:08:29] you left the %s right?
[27-Jul-2010 13:08:32] <|rt|> yeah
[27-Jul-2010 13:08:38] the first %s is the first variable, second is the second
[27-Jul-2010 13:08:48] and you could add a third or whatever if you had a third variable
[27-Jul-2010 13:08:56] pretty straightforward
[27-Jul-2010 13:09:16] <|rt|> yeah....I just didn't know how to access the variables from the event
[27-Jul-2010 13:09:37] so rmatte should i open a bug or something for that issue with the UI ? i really dont mind, but if it helps i have no problem with that
[27-Jul-2010 13:11:16] rmatte: I'm adding a datasource..but don't really know what Component means..or event key..or anye of that
[27-Jul-2010 13:11:28] Do you know where in the documentation it is?
[27-Jul-2010 13:12:04] straterra: you can just leave those blank
[27-Jul-2010 13:12:55] straterra: components would be shown in the component column in the event queue
[27-Jul-2010 13:13:06] eventKey is just an event property
[27-Jul-2010 13:13:09] both are optional
[27-Jul-2010 13:13:16] What about event class?
[27-Jul-2010 13:13:30] event class should be set to whatever event class you feel is relevant
[27-Jul-2010 13:13:43] And command template?
[27-Jul-2010 13:13:45] If it's a CPU OID you'd set it to /Perf/CPU for instance
[27-Jul-2010 13:14:03] /cmd/fail works, I guess
[27-Jul-2010 13:14:11] where do you see command template?
[27-Jul-2010 13:14:17] are you using the new UI or the old one?
[27-Jul-2010 13:14:22] In the data source
[27-Jul-2010 13:14:24] I've yet to make a template in the new UI
[27-Jul-2010 13:14:35] I'm on 2.5..so..
[27-Jul-2010 13:14:37] ah
[27-Jul-2010 13:14:39] old UI
[27-Jul-2010 13:14:41] Yeah
[27-Jul-2010 13:14:47] let me look for myself so i actually see what you see
[27-Jul-2010 13:14:50] kk
[27-Jul-2010 13:14:55] trying to do this off the top of my head is half-assed lol
[27-Jul-2010 13:15:35] command template is the actual command you want to execute
[27-Jul-2010 13:15:41] the script
[27-Jul-2010 13:15:52] and you would pass info to it
[27-Jul-2010 13:15:55] an example...
[27-Jul-2010 13:16:07] /libexec/wincpu.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"
[27-Jul-2010 13:16:33] that would execute wincpu.sh from the libexec directory and pass the IP of the device the SNMP string of the device as arguments to the script
[27-Jul-2010 13:16:41] Right
[27-Jul-2010 13:17:04] So..how do I define good or bad output?
[27-Jul-2010 13:17:18] well, that's where the script itself comes in
[27-Jul-2010 13:17:24] first off, what does the script do exactly?
[27-Jul-2010 13:17:36] Checks two servers postgres databases to make sure slony is working properly
[27-Jul-2010 13:17:53] When both databases match, it returns an ok status..when they dont, it tells how many records one is lagging by
[27-Jul-2010 13:17:54] ok, and if it's not it throws some sort of error?
[27-Jul-2010 13:18:02] I see
[27-Jul-2010 13:18:10] so what you'd want the script to do is this...
[27-Jul-2010 13:18:13] <|rt|> rmatte: I edited the example you gave me online to reflect my changes I dont' think there's any reason for it not to work. http://fpaste.org/PO3N/
[27-Jul-2010 13:18:19] if everything is ok, have it output: OK|
[27-Jul-2010 13:18:22] and exit normally
[27-Jul-2010 13:18:37] if something is wrong have it output: WHATEVER ERROR YOU WANT|
[27-Jul-2010 13:18:43] and have it exit with an exit code of 1 or 2
[27-Jul-2010 13:18:50] The script gives different exit codes, I'm told
[27-Jul-2010 13:19:03] It's an old nagios script, heh
[27-Jul-2010 13:19:03] well then just make sure the output is nagios format
[27-Jul-2010 13:19:06] STATUS|
[27-Jul-2010 13:19:09] ah
[27-Jul-2010 13:19:12] then it'll work fine
[27-Jul-2010 13:19:28] Zenoss handles Nagios scripts
[27-Jul-2010 13:19:42] so just have the command datasource execute it properly with any arguments that it needs (if any)
[27-Jul-2010 13:19:48] and let er' rip
[27-Jul-2010 13:23:02] cgibbons: navigate to Infrastructure -> Networks -> then click on the options menu in the bottom left... very funny
[27-Jul-2010 13:23:08] so many choices
[27-Jul-2010 13:23:33] actually hmmm, it only does it when you're on one of the networks
[27-Jul-2010 13:23:37] click on a network folder, then check it
[27-Jul-2010 13:23:51] hmmm, it's not doing it anymore
[27-Jul-2010 13:23:59] it was saying "Discover Devices" 3 times
[27-Jul-2010 13:24:03] * rmatte attempts to recreate
[27-Jul-2010 13:24:27] hmmm, guess it was a fluke, odd
[27-Jul-2010 13:25:13] Works
[27-Jul-2010 13:25:23] Can I change the Definition path once I've created it?
[27-Jul-2010 13:25:35] which definition path?
[27-Jul-2010 13:25:46] For the new template I made with the datasource
[27-Jul-2010 13:26:05] It's under /Devices right now and I need it for /Devices/Server/Linux/ERP
[27-Jul-2010 13:26:14] yes, you can move it freely
[27-Jul-2010 13:26:19] it won't affect anything
[27-Jul-2010 13:26:43] although I'd say just leave it at /Devices and bind it at the level that you need it at
[27-Jul-2010 13:26:45] works just the same
[27-Jul-2010 13:26:50] oh, ok
[27-Jul-2010 13:26:57] either that or copy it to the level you want it at and delete it from /Devices
[27-Jul-2010 13:26:58] up to you
[27-Jul-2010 13:29:50] So..where in the device can I check the status of that datasource?
[27-Jul-2010 13:30:07] you can't, it'll just generate an alert if something goes wrong
[27-Jul-2010 13:30:11] ah, ok
[27-Jul-2010 13:30:12] if it doesn't, then all is well
[27-Jul-2010 13:30:43] well, what does the actual output of that script look like?
[27-Jul-2010 13:31:15] you might be able to graph some data, depending on what the output looks like
[27-Jul-2010 13:31:21] but they may not have scripted it to provide any
[27-Jul-2010 13:35:33] It's text
[27-Jul-2010 13:35:45] Later I'll make it output just numbers to have it graph
[27-Jul-2010 13:35:52] Thanks for the help...you're a life saver
[27-Jul-2010 13:37:12] Hi friends, I was reading the developer's guide, looking for how to create events using python, so in an example, it says that I can import the library using Products.ZenEvents.ZenEventClasses
[27-Jul-2010 13:37:53] but, I found PBDaemon in Products.ZenHub.PBDaemon, is that the same?
[27-Jul-2010 13:39:48] look at $ZENHOME/bin/zensendevent
[27-Jul-2010 13:40:29] chachan: don't forget about zensendevent if you just need something quick
[27-Jul-2010 13:43:01] mray, well, I don't need something quick, I need to connect to a database, parse some web services, connect via ssh to another server, etc
[27-Jul-2010 13:45:23] so, I found it in a different location from what the guide show me
[27-Jul-2010 13:48:34] straterra: np
[27-Jul-2010 13:49:28] chachan: they are saying to actually look at the code for zensendevent
[27-Jul-2010 13:49:32] since it creates events
[27-Jul-2010 13:49:42] ;)
[27-Jul-2010 13:56:21] guys can u point me an easy, "long" explained alerts creation and capabilities guide? im reading the admin guide now
[27-Jul-2010 13:59:09] JonnyV, you can create alerts to users and groups through "Advanced" > "Users", choose an user or group and go to "Alerting Rules"
[27-Jul-2010 13:59:49] so, you can create alerts through it
[27-Jul-2010 13:59:58] yeps.. im there..
[27-Jul-2010 14:00:23] cool.. the groups one is interesting
[27-Jul-2010 14:00:47] so i dont tie alerts to users that can be deleted right
[27-Jul-2010 14:01:41] but im looking for more detailed info. will investigate the events info on the admin guide.
[27-Jul-2010 14:04:34] JonnyV: well, don't delete users unless necessary and you won't have issues with the alerts lol\
[27-Jul-2010 14:04:44] mrayzenoss is now known as mray
[27-Jul-2010 14:05:34] hehe.. yeah.. iknow.. but maybe when i leave the company they delete my account.. hehe
[27-Jul-2010 14:05:50] then tough luck for them
[27-Jul-2010 14:05:51] lol
[27-Jul-2010 14:05:59] hahaha
[27-Jul-2010 14:06:00] yeaa
[27-Jul-2010 14:06:12] i'm thinking right now that i dont care that much
[27-Jul-2010 14:06:13] that falls under the category of something you'd want to explain before leaving
[27-Jul-2010 14:06:14] haha
[27-Jul-2010 14:06:35] will depend on how i leave... hehe
[27-Jul-2010 14:06:59] i will create groups then
[27-Jul-2010 14:07:20] hopefully not carried out in chains
[27-Jul-2010 14:07:21] but who knows
[27-Jul-2010 14:07:35] hahaha
[27-Jul-2010 14:07:46] u never know
[27-Jul-2010 14:11:37] what should i put on the Alert ID? number or something descriptive? really dunno where it will impact
[27-Jul-2010 14:11:49] on the guide it says a name.
[27-Jul-2010 14:12:31] doesn't matter
[27-Jul-2010 14:12:35] whatever you want to call it
[27-Jul-2010 14:12:47] it can be "super galactic flying sheep" if that's what you fancy
[27-Jul-2010 14:13:01] hahaha
[27-Jul-2010 14:13:21] ok. i kinda liked that name
[27-Jul-2010 14:13:30] :)
[27-Jul-2010 14:16:50] hell oall
[27-Jul-2010 14:17:19] anyone have issues with net-snmp recognizing on /dev mounted tmpfs devices rather than any custom ones? I know this isn't #net-snmp but no one ever answers in there
[27-Jul-2010 14:18:26] i run the snmpwalk and it picks up my /dev/shm but no my custom mounted one and thats the one i need to monitor
[27-Jul-2010 14:18:46] QubeZ: still at it eh?
[27-Jul-2010 14:18:59] QubeZ: like I said, you'd be better to try in a linux or linux server channel
[27-Jul-2010 14:19:27] rmatte: ya man, its driving me nuts because at times my ramdisk fills up and I need to find out exactly when so I can run a script to clean up the dir before it becomes a major issue like it was today
[27-Jul-2010 14:19:55] QubeZ: Why don't you just check it via an SSH datasource, threshold on that, and call it a day?
[27-Jul-2010 14:20:18] rmatte: how do I do that?
[27-Jul-2010 14:20:33] QubeZ: the admin guide explains command ssh datasources
[27-Jul-2010 14:20:41] rmatte: ok thanks
[27-Jul-2010 14:20:46] np
[27-Jul-2010 14:25:05] rmatte: what is the table of contents descp? im looking at the ssh and/or datasource but see nothing in there that would be what i need. I see a modeling via ssh but i dont want to do that.
[27-Jul-2010 14:33:59] for your NFS mounted filesystems on a linux device in 2.5.2, do used, free, and % util show up as "unknown" ?
[27-Jul-2010 14:34:30] gwb235: no, they show up
[27-Jul-2010 14:34:37] hrm
[27-Jul-2010 14:34:43] you need to wait 15 minutes for them to populate
[27-Jul-2010 14:34:50] if you just added it
[27-Jul-2010 14:34:56] yeah, this device has been around a while
[27-Jul-2010 14:35:08] yeh, but did you just remodel it to see that share?
[27-Jul-2010 14:35:13] nope
[27-Jul-2010 14:35:22] k, no idea then
[27-Jul-2010 14:35:26] looks fine for me
[27-Jul-2010 14:35:31] try deleting the RRDs
[27-Jul-2010 14:37:58] what interval does a collector normally pick up changes?
[27-Jul-2010 14:38:10] i've gotten in the habit of pushing... but what's the interval?
[27-Jul-2010 14:51:24] tricky
[27-Jul-2010 14:54:42] what do the green lines on the google map between locations represent?
[27-Jul-2010 14:54:58] it just means that there are devices at both locations that are in the same subnet
[27-Jul-2010 14:55:01] that's all it means
[27-Jul-2010 14:55:23] so Zenoss assumes that they are in the same subnet because the sites are linked
[27-Jul-2010 14:55:50] in the ideal solution, when you're subnetting, you create tiny subnets with only 2 addresses to use for links between sites
[27-Jul-2010 14:56:11] that way you're not wasting IP space
[27-Jul-2010 14:56:46] though the mickey mouse way of doing it is having a bunch of sites that are 192.168.0.1/24 and linking them together
[27-Jul-2010 14:57:23] but that's uber sloppy
[27-Jul-2010 14:59:48] i use the 172.16 for links between network devices
[27-Jul-2010 14:59:55] 10. for giving out to gorups
[27-Jul-2010 15:00:12] cool
[27-Jul-2010 15:00:33] and 192.168 is for home (which long time agos ran into VPN conflicts and couldn't proxy subnets between home via VPN gateway)
[27-Jul-2010 15:00:36] we just use 30.30.x.x for everything, since it won't conflict with any IP spaces that our clients are using
[27-Jul-2010 15:00:50] It's the DoD IP space
[27-Jul-2010 15:00:53] no one really uses it
[27-Jul-2010 15:11:17] how do you access the device class through an event
[27-Jul-2010 15:11:21] I want to call up a zproperty value for a device
[27-Jul-2010 15:11:26] mducharme-laptop: like find out what class the device is in?
[27-Jul-2010 15:11:33] oh
[27-Jul-2010 15:11:35] well here's what I'm trying to do
[27-Jul-2010 15:11:39] you have that evt which is the event
[27-Jul-2010 15:11:45] and you call evt.device for the device name
[27-Jul-2010 15:12:01] normally to call up the zproperty for a device you would use device.zPropertyName
[27-Jul-2010 15:12:15] when I try evt.device.zPropertyName it doesn't work
[27-Jul-2010 15:12:24] yeh, that's easy...
[27-Jul-2010 15:12:29] you do something like...
[27-Jul-2010 15:12:30] it seems like evt.device is just a string instead of being a pointer at a device object
[27-Jul-2010 15:12:45] d = dmd.Devices.findDevice(evt.device)
[27-Jul-2010 15:12:54] d.zPropertyName
[27-Jul-2010 15:12:57] ahh!
[27-Jul-2010 15:12:58] thanks
[27-Jul-2010 15:13:01] np
[27-Jul-2010 15:14:00] LarsN is now known as LarsN_Away
[27-Jul-2010 15:16:17] mrayzenoss is now known as mray
[27-Jul-2010 15:17:59] yes!!!! it works!!!!
[27-Jul-2010 15:18:05] it is pulling the zproperties now
[27-Jul-2010 15:18:06] sweet
[27-Jul-2010 15:20:17] if I want to check if a boolean is true can I just say if variablename: or do I need to say if variablename == True:
[27-Jul-2010 15:20:32] if variablename
[27-Jul-2010 15:27:16] LarsN_Away is now known as LarsN
[27-Jul-2010 15:29:25] * LarsN is so glad to see Zenoss 3 documentation on the site :)
[27-Jul-2010 15:29:38] I'm lost in the new interface.
[27-Jul-2010 15:33:12] seems to have settled
[27-Jul-2010 15:33:29] pfff
[27-Jul-2010 15:33:41] lol
[27-Jul-2010 15:33:47] theacoly1e is now known as theacolyte
[27-Jul-2010 15:34:17] theacolyte is now known as Guest72896
[27-Jul-2010 15:35:53] Guest72896 is now known as theacolyte
[27-Jul-2010 15:38:12] rmatte: where in the admin guide does it show that? i looked at the datasource and ssh sections
[27-Jul-2010 15:38:25] is it related to setting up a zenplugin on the remote server?
[27-Jul-2010 15:38:44] no
[27-Jul-2010 15:39:02] let me find it
[27-Jul-2010 15:39:11] thanks man
[27-Jul-2010 15:40:25] hmmm, they actually have less info on it than I would have thought, hold on
[27-Jul-2010 15:40:46] k
[27-Jul-2010 15:41:21] QubeZ: eugh, they butchered some part of this admin guide
[27-Jul-2010 15:41:27] :(
[27-Jul-2010 15:41:43] well, the 3.0 version anyways
[27-Jul-2010 15:41:52] part of it say to refer to sections that no longer exist
[27-Jul-2010 15:41:52] im running 2.5
[27-Jul-2010 15:42:06] docs/DOC-4728
[27-Jul-2010 15:42:17] read from " About COMMAND Data Sources" down
[27-Jul-2010 15:42:31] k cool, that's the 2.5 guide that I just linked you to
[27-Jul-2010 15:42:39] thanks
[27-Jul-2010 15:42:42] np
[27-Jul-2010 15:43:16] ah wait
[27-Jul-2010 15:43:18] I can find better
[27-Jul-2010 15:43:31] docs/DOC-4770
[27-Jul-2010 15:43:32] voila
[27-Jul-2010 15:43:38] that's even better
[27-Jul-2010 15:44:48] * LarsN reads the documentation under 5.2.2
[27-Jul-2010 15:44:53] yay Templates!
[27-Jul-2010 15:45:16] yay!
[27-Jul-2010 15:45:20] what? :-)
[27-Jul-2010 15:45:48] has anyone imported Products.ZenHub.PBDaemon?
[27-Jul-2010 15:45:58] aclark: building a shiny new Zenoss 3.0 machine. On actual hardware rather than a Xen DomU this time.
[27-Jul-2010 15:46:21] aclark: and the Zenoss 3.0 interface is so different i'm actually having to read the documentation to get it to do what I want. :)
[27-Jul-2010 15:46:27] LarsN: nice
[27-Jul-2010 15:47:24] the whole "create local copy" of a bound template function is escaping me at the moment.
[27-Jul-2010 15:47:48] where can I find documentation on dmd functions?
[27-Jul-2010 15:47:54] LarsN: the concept of it, or the purpose of it?
[27-Jul-2010 15:48:12] twm1010: more like, where the functionality to create a local copy went. :)
[27-Jul-2010 15:48:24] oh, haha, uh... no idea :)
[27-Jul-2010 15:48:25] used to be a button next to the template
[27-Jul-2010 15:48:38] That's a good question for mray
[27-Jul-2010 15:48:43] it doesn't appear to be in the API docs
[27-Jul-2010 15:49:13] I have the api docs...
[27-Jul-2010 15:49:58] hrmm, don't templates have a whole new interface in v3?
[27-Jul-2010 15:50:05] yep
[27-Jul-2010 15:50:13] twm1010: exactly why I'm trying to read through the documentation.
[27-Jul-2010 15:50:20] with a few bugs
[27-Jul-2010 15:50:34] cgibbons: any ideas on where create local template went?
[27-Jul-2010 15:50:50] if you're trying to do it on a device, go to the device, click the template you want, and then click the gear menu and choose override template here
[27-Jul-2010 15:51:19] I was thinking /server/freebsd vs the /server for the Device template actually
[27-Jul-2010 15:51:30] cgibbons: http://dev.zenoss.org/trac/ticket/7121
[27-Jul-2010 15:51:33] for device classes go to advanced -> monitoring templates
[27-Jul-2010 15:51:38] and then see mray for huge list of bugs about it :)
[27-Jul-2010 15:51:42] although override template is functionally the same as copying local? thanks.
[27-Jul-2010 15:52:17] the metaphor in the UI is now 'override' because copying in the UI just exposed to the end-user what was going on behind the scenes, not what the workflow was really about.
[27-Jul-2010 15:52:17] might actually blow up this install and downgrade to 2.5 :)
[27-Jul-2010 15:52:52] cgibbons: hrmm... but in my brain that changes funtionality a little, it would be nice if the templates could act more like zproperties a little with inheritance
[27-Jul-2010 15:53:02] let 3.x get a few minor or dot releases :)
[27-Jul-2010 15:53:03] so if i could only override one setting in a template, that would be slick
[27-Jul-2010 15:53:54] alot of times i override something like... a threshold on a template right? but now its really a copy, and if make changes to the original, i then have to go back and adjust all the little iterations floating about
[27-Jul-2010 15:54:40] twm1010: We do this, although usually by OS type, rather than at the specific machine level.
[27-Jul-2010 15:55:08] which is why I want to copy the device template, override it.... into /server/FreeBSD
[27-Jul-2010 15:55:10] really not all that rampant, but it adds up, like say i adjusted created a local filesystem template for a single volume on a server and forget to go back and remove it to restore the inheritance from the parent once i've finally addressed a utilization isue
[27-Jul-2010 16:03:04] LarsN is now known as LarsN_Away
[27-Jul-2010 16:09:06] how could I import SendEvent.py?
[27-Jul-2010 16:10:30] hrm... this doesn't seem to do anything
[27-Jul-2010 16:10:44] for dg in dmd.DeviceGroup.getSubOrganizers():
[27-Jul-2010 16:10:56] that returns: AttributeError: DeviceGroup
[27-Jul-2010 16:14:53] how the heck do I get a list of all device groups
[27-Jul-2010 16:15:00] there has to be a call to do that
[27-Jul-2010 16:18:37] mducharme: you can use the getOrganizer method if you know what you're looking for
[27-Jul-2010 16:19:10] devdmd =Â self.getDmdRoot("Devices")
[27-Jul-2010 16:19:14] dclass =Â devdmd.getOrganizer("/Discovered")
[27-Jul-2010 16:20:26] dmd.Groups.getSubOrganizers() did the trick
[27-Jul-2010 16:21:46] except aurgh
[27-Jul-2010 16:21:51] now I don't know how to pull the time
[27-Jul-2010 16:21:52] er name
[27-Jul-2010 16:21:57] path rather
[27-Jul-2010 16:22:09] for dg in dmd.Groups.getSubOrganizers():
[27-Jul-2010 16:22:18] that works
[27-Jul-2010 16:22:32] but how do I get a string of the device group path
[27-Jul-2010 16:36:24] ok this is weird
[27-Jul-2010 16:37:04] when I use dmd.Devices.manage_addOrganizer('/Devices/whatever') from the command prompt it adds the organizer
[27-Jul-2010 16:37:11] but it isn't adding the organizer in this script
[27-Jul-2010 16:42:25] it's passing the string correctly, I know because I logged the output
[27-Jul-2010 16:42:41] the addOrganizer command just isn't doing anything
[27-Jul-2010 16:46:36] anyone who knows how to import Products.ZenHub.PBDaemon please :'(
[27-Jul-2010 16:47:17] I don't know if I have wrong path of if I need change it to get to this .py
[27-Jul-2010 16:47:36] I get ImportError: No module named ZenHub.PBDaemon
[27-Jul-2010 16:58:40] isn't it to suppossed that I need to have ZOPEHOME in my bashrc?
[27-Jul-2010 16:59:42] i had an unsuccessful install of the preupgrade egg for 3.0, what do I do?
[27-Jul-2010 16:59:59] error?
[27-Jul-2010 17:02:31] this is such a simple thing.. wish I knew why it wasn't working
[27-Jul-2010 17:02:54] more chicken blood
[27-Jul-2010 17:03:14] our system is not yet ready for upgrade. Please install
[27-Jul-2010 17:03:14] ZenPacks.zenoss.PreUpgrade30 and let it run to completion before
[27-Jul-2010 17:03:32] i re-ran the python egg and it did go through, this is what im trying to install zenoss 3.0 rpm
[27-Jul-2010 17:03:42] i haven't installed the core packs yet so i didn't add the --nodeps
[27-Jul-2010 17:05:52] i dont even see a link to download the core zenpacks
[27-Jul-2010 17:09:44] dmd.Devices.manage_addOrganizer('/Devices/DocTest')
[27-Jul-2010 17:10:01] that should be adding an organizer.. it does from the command prompt, but is not from the script
[27-Jul-2010 17:13:23] this preupgarde egg looks like it installs but when i try to install rpm, it errors out stating the system is not prepared for upgrade
[27-Jul-2010 17:13:27] im following the upgrade doc line by line
[27-Jul-2010 17:28:55] ahh I didn't realize the commit() function worked just then and there
[27-Jul-2010 18:27:21] guys, where to change the eventUrl, or ackUrl that appears on the message strings as variables? i was reading about zopeurl in zenactions
[27-Jul-2010 18:27:30] but i see it empty that file
[27-Jul-2010 18:27:37] 3.0 im running
[27-Jul-2010 18:54:25] I'm having some trouble with the "Systems" organizer in 3.0: it shows there are 28 systems, but the list is blank
[27-Jul-2010 20:17:26] mrayzenoss is now known as mray
[27-Jul-2010 20:34:24] <|rt|> rmatte: I got a hit on that event you helped me with the transform earlier today....the transform didn't work....it returned Unknown for both variables
[27-Jul-2010 20:49:22] guys, where to change the eventUrl, or ackUrl that appears on the message strings as variables in Alerting? i was reading about zopeurl in zenactions
[27-Jul-2010 20:49:31] but i see it empty that file
[27-Jul-2010 20:49:49] im using 3.0
[28-Jul-2010 00:00:30] [disconnected at Wed Jul 28 00:00:30 2010]
[28-Jul-2010 00:00:31] [connected at Wed Jul 28 00:00:31 2010]
[28-Jul-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]
[28-Jul-2010 03:49:20] hi
[28-Jul-2010 03:49:26] frozty_s1 is now known as frozty_sa
[28-Jul-2010 03:53:12] morning
[28-Jul-2010 03:53:38] eyeing message/51074...I'm tending to say I might be willing to try and implement this
[28-Jul-2010 03:53:51] it's something we need as well, and I see people again and again asking for it
[28-Jul-2010 03:54:50] my personal concern is mainly that I'm not entirely sure where to plug all the different parts of it together, so I'd want to speak to someone on the dev team, or who otherwise knows zenoss well enough, now and then when I'd need it
[28-Jul-2010 03:56:16] * Simon4 has a think
[28-Jul-2010 03:57:40] single parent device for any given device, tree style I guess, but you would need to walk the tree
[28-Jul-2010 03:58:21] so you would want to patch zenEventManager somewhere and maybe have a new relation for the device object which is "parent"
[28-Jul-2010 03:58:48] * Simon4 wonders if he's over simplifying this
[28-Jul-2010 03:59:02] damnit, I want to try this now, and I have a full day's work ahead :/
[28-Jul-2010 04:00:08] (we could make use of it, also)
[28-Jul-2010 04:03:00] Simon4: well, principly, you're on the right track
[28-Jul-2010 04:03:02] actually, if you had the relation, and a custom function as part of the device class spec, you could then just use an event transform to suppress the events, since you have access to the device object in the transform
[28-Jul-2010 04:03:31] frozty_sa: would any given device only ever have one "parent" device? i.e. one to many type relationship?
[28-Jul-2010 04:03:42] so say, bunch of hosts have a router as their "parent", etc
[28-Jul-2010 04:05:56] I reckon the logic would have to get implemented in the event manager
[28-Jul-2010 04:06:06] lemme just sort this other thing out quickly
[28-Jul-2010 04:10:25] Simon4: the case where I can think that would fail is device that have a failover configuration
[28-Jul-2010 04:10:35] such as hosts with full crossconnects
[28-Jul-2010 04:11:19] I'm not sure if there's a general solution to that at L3 (seeing as it's normally implemented with floating IPs), but L2 is doable
[28-Jul-2010 04:12:11] * Simon4 nods
[28-Jul-2010 04:12:38] so you would only want to suppress the alerts when _all_ parent devices were down/unreachable/etc
[28-Jul-2010 04:17:09] well, I first want to sort out the basic version, in terms of design
[28-Jul-2010 04:18:16] if one gets a down alert, the event processing should check if evt.device.getParent() != None, for instance
[28-Jul-2010 04:18:51] if it has a parent, get the status on the parent?
[28-Jul-2010 04:19:36] this would mean that one has to select events for the parent device..not sure if there's a device.state or something? (I haven't looked into how devices are presented, I'll probably have to poke at that)
[28-Jul-2010 04:25:02] %e_TodoListOverflow
[28-Jul-2010 04:36:41] hehe
[28-Jul-2010 04:37:03] frozty_sa: yeah, I'm going to have a play with this after work
[28-Jul-2010 04:38:41] I'll probably also have to discuss lots of it tonight, because more zenoss upperheads will be online then
[28-Jul-2010 04:38:50] yah
[28-Jul-2010 04:39:12] * frozty_sa rages and fumes
[28-Jul-2010 04:39:23] * frozty_sa looks if rancid is done yet
[28-Jul-2010 06:37:37] LarsN_Away is now known as LarsN
[28-Jul-2010 06:39:06] what's the process to move a device from /Discovered to /somewhere/else
[28-Jul-2010 06:39:13] on zenoss 3.0
[28-Jul-2010 06:40:44] drag and drop I believe
[28-Jul-2010 06:46:18] I'll share my initial opinions. The new UI..... Kinda like office 2007.
[28-Jul-2010 06:46:41] which is to say, i'm sure all the functionality I'm used to is there, and it might even work better. But i can't find anything.
[28-Jul-2010 06:47:25] LarsN and Simon4! :-)
[28-Jul-2010 06:48:20] morning aclark
[28-Jul-2010 06:50:30] LarsN: Morning, is it *that* bad? The new UI
[28-Jul-2010 06:51:02] aclark: well, there's a trac ticket on how "horribly broken" the bind template workflow is.
[28-Jul-2010 06:51:48] and it looks like all the developers have adopted apple's ui practices, putting add/remove/edit at the bottom left of the window.
[28-Jul-2010 06:51:53] ouch
[28-Jul-2010 06:52:03] it might be better. but I'm not seeing it.
[28-Jul-2010 06:52:46] pretty sure I'm going to downgrade to 2.5.x
[28-Jul-2010 06:53:05] and let at least one or two dot releases happen on 3.x first
[28-Jul-2010 06:53:54] right
[28-Jul-2010 06:54:15] where one or two, most likely means several months :)
[28-Jul-2010 06:54:27] LarsN: i remember being a bit shocked running the beta…
[28-Jul-2010 07:09:40] aclark: consequences will never be the same, cyber police have been informed of your activity.
[28-Jul-2010 07:10:26] Apachez: ?
[28-Jul-2010 07:26:48] morning aclark
[28-Jul-2010 07:26:53] or, afternoon as it now is :)
[28-Jul-2010 07:31:12] * LarsN waits for his coffee maker to finish brewing the 40 cups he put in it.
[28-Jul-2010 07:37:07] aclark: if you were looking at a device in zenoss. Let's call it plone01. And you saw an orange error, count 1, at the top of the dashboard for that device. Wouldn't you expect when clicking on that alert, that it would take you to the page specifically about that event?
[28-Jul-2010 07:37:47] LarsN: probably, sounds reasonable
[28-Jul-2010 07:38:12] LarsN: but I don't recall 2.x doing that… then again I'm not a heavy user
[28-Jul-2010 07:38:59] instead, to find out about that specific event I need to choose events from the menu on the left. If I click the big orange 1, it takes me to the overall events page. which would be fine, if it also added a filter for the specific device I was looking at, but of course, that's apparently expecting too much.
[28-Jul-2010 07:39:05] sorry /rant
[28-Jul-2010 07:41:35] LarsN: well as Apachez says, the cyber police have been informed of your activity :-)
[28-Jul-2010 07:46:45] aclark: heh, someone at Zenoss saw my tweet, and told me to e-mail him with a list of complaints :)
[28-Jul-2010 07:47:08] LarsN: of course
[28-Jul-2010 07:47:21] LarsN: those kids are on their game :-)
[28-Jul-2010 07:47:52] I give Zenoss Credit from a community support standpoint. :)
[28-Jul-2010 07:48:25] I'm sure most of my objections are because I'm used to doing something one way, and now it's been changed.
[28-Jul-2010 07:49:00] that's why I used Office 2007 as my example. It's technially easier than previous versions. But because everything has been moved, long time users HATED the new UI.
[28-Jul-2010 08:29:08] hey folks, so I just took the plunge and upgraded to 3.0 via the centos RPM, and I'm getting an error now when zeo tries to start
[28-Jul-2010 08:29:17] [zenoss@american ~]$ zeoctl start
[28-Jul-2010 08:29:17] Error: no program specified; use -p or -C
[28-Jul-2010 08:29:17] For help, use /opt/zenoss/bin/zeoctl -h
[28-Jul-2010 08:29:46] you probably want to run service zenoss start as root first time
[28-Jul-2010 08:29:56] ah ok, I will try that
[28-Jul-2010 08:30:50] <-- alot more happening now, thanks, I just tried zenoss start as zenoss
[28-Jul-2010 08:44:23] if I stop the stack installer, and blow up the /usr/local/zenoss folder
[28-Jul-2010 08:44:42] will that open the doors for re-installing fresh?
[28-Jul-2010 08:44:50] sorry, stop the stack processes
[28-Jul-2010 09:06:44] I have a devicer I cannot delete - any takers????
[28-Jul-2010 09:07:24] I get
[28-Jul-2010 09:07:29] ObjectNotFound: object was not found on
[28-Jul-2010 09:08:13] .. and indeed print d.deviceClass() from zendmd returns with None
[28-Jul-2010 09:08:59] * Simon4 runs screaming from that one
[28-Jul-2010 09:09:02] tried running fix_broken_devicesearch_componentsearch.py and fix_layer3_catalog.py to no avail
[28-Jul-2010 09:09:21] I've been screaming at it for 3 hours now!
[28-Jul-2010 09:09:44] bigger hammer?
[28-Jul-2010 09:10:10] Jane_Curry: what happs if you bin/instance debug and try to delete in there?
[28-Jul-2010 09:10:14] My Device List menu is also bust - responds with "unable to communicate with the server"
[28-Jul-2010 09:10:42] Jane_Curry: errr, zendmd (I think it's called in Zenoss land)
[28-Jul-2010 09:10:55] but all daemons are running and I can see devices by navigating down Devices/...
[28-Jul-2010 09:11:42] sorry - don't understand your bin/instance debug comment...
[28-Jul-2010 09:15:37] nothing in zenhub.log...
[28-Jul-2010 09:16:38] * ashp pokes people, still looking good for 3.0 enterprise this week?
[28-Jul-2010 09:28:36] mrayzenoss is now known as mray
[28-Jul-2010 09:28:49] mray - is zenfixit something core users can grab?
[28-Jul-2010 09:29:21] it's been floating around in the forums
[28-Jul-2010 09:29:41] since it's a tool that support is always tweaking, we didn't release it
[28-Jul-2010 09:30:10] yeah
[28-Jul-2010 09:30:27] jane has a buggy device she can't delete, wonder if that would solve the issue (broken reference to a device class)
[28-Jul-2010 09:31:21] cgibbons: many times when we have issues like that we can fix it by going into zendmd, running reindex() then commit() then restart the zenoss services.
[28-Jul-2010 09:31:24] I don't know where the canonical version is, but it shouldn't hurt
[28-Jul-2010 09:31:34] zenChild: yeah, that's one of the things it does
[28-Jul-2010 09:31:42] yeah, not sure that'll fix a device class reference since it's not really an index thing
[28-Jul-2010 09:31:51] it just combines all the things that possibly fix zodb issues and consolidates them
[28-Jul-2010 09:31:52] but harmless to try
[28-Jul-2010 09:35:09] tried the reindex and commit trick many times - it still wont go
[28-Jul-2010 09:35:22] Good morning everyone
[28-Jul-2010 09:36:30] I had been trying to import a Products.ZenHub.PBDaemon and I couldn't
[28-Jul-2010 09:37:15] Jane_Curry: check your email
[28-Jul-2010 09:47:52] LarsN is now known as LarsN_Away
[28-Jul-2010 10:39:40] hmmm too quiet today
[28-Jul-2010 10:39:45] indeed
[28-Jul-2010 10:44:56] ok.....
[28-Jul-2010 10:45:28] I am writing a zenpack that expands the attributes and relationships of a device
[28-Jul-2010 10:46:18] I have created a new Edit tab that shows new attributes and lets you edit them
[28-Jul-2010 10:47:02] I want to add a new entry to the relationship part of the edit tab that lists all instances of a device of a certain device class
[28-Jul-2010 10:48:19] .. rather like the Systems and Groups scrolling list
[28-Jul-2010 10:49:26] I have a line in the skins file that sez
[28-Jul-2010 10:50:28] pkgs python:here.dmd.Devices.Service.System.Application.Package.getOrganizerNames();
[28-Jul-2010 10:51:43] that gets me a single-element list of /Service/System/Application/Package - which is the class path..
[28-Jul-2010 10:52:12] what I want are the instances of those devices
[28-Jul-2010 10:52:59] what do I need to put on the end of my skins line?????
[28-Jul-2010 10:56:09] morning
[28-Jul-2010 10:57:02] you around rmatte?
[28-Jul-2010 10:57:13] . o O (and he is bored)
[28-Jul-2010 11:00:15] mducharme-laptop: yeh
[28-Jul-2010 11:01:21] not sure what you mean, Jane. are you asking about the rest of the syntax to finish the expression specifically, or?
[28-Jul-2010 11:02:38] mray: hiya
[28-Jul-2010 11:03:16] mray: dunno if you were around with the discussion Simon4 and I had earlier? I've got joins and parts filtered at the client I usually view IRC on..
[28-Jul-2010 11:06:24] cgibbons: yeh - I can't find the right combination to list the actual devices, rather than the device class path
[28-Jul-2010 11:07:47] So I have decided that I like 3
[28-Jul-2010 11:08:00] That shift stuff kills it a bit,b ut nice job
[28-Jul-2010 11:08:06] I really like the interface polish
[28-Jul-2010 11:08:51] frozty_sa: I missed it completely
[28-Jul-2010 11:09:08] i dislike it, but i'm still reading my emails on the command line...
[28-Jul-2010 11:09:18] wouldn't you want python:here.dmd.Devices.Service.System.Application.Package.getOrganizerNames(SubDevices(); instead?
[28-Jul-2010 11:09:27] :P
[28-Jul-2010 11:09:49] er
[28-Jul-2010 11:09:53] wow i totally hosed that paste
[28-Jul-2010 11:09:59] python:here.dmd.Devices.Service.System.Application.Package.getOrganizerNames(SubDevices(); :)
[28-Jul-2010 11:10:03] hmm
[28-Jul-2010 11:10:08] bah!
[28-Jul-2010 11:10:13] lmao
[28-Jul-2010 11:10:15] replace getOrganizerNames with getSubDevices
[28-Jul-2010 11:10:20] the message got accross ...
[28-Jul-2010 11:11:27] ok I have one that y'all might know - what's the difference between ifInOctets and ifHCInOctets SNMP values?
[28-Jul-2010 11:11:54] last one should be 64bit counter
[28-Jul-2010 11:12:18] ok, but representing the same data?
[28-Jul-2010 11:12:19] yeh, 64bit
[28-Jul-2010 11:12:21] cgibbons: but then in my list I get a great long string...
[28-Jul-2010 11:12:24] yes
[28-Jul-2010 11:12:25] I think HC stands for high count
[28-Jul-2010 11:13:33] ah I see, so getOrganizerNames gives you a list of strings, but getSubDevices gets you a list of device objects
[28-Jul-2010 11:14:24] you could use a list comprehension to pull out the device id's instead: [d.id for d in dmd.Devices.Server.getSubDevices()]
[28-Jul-2010 11:14:36] although, I wonder at that point if a catalog would be better hrm
[28-Jul-2010 11:14:38] but try that
[28-Jul-2010 11:14:58] cgibbons: I've been pondering it a bit the last few days as well
[28-Jul-2010 11:15:09] the counters/
[28-Jul-2010 11:15:09] ?
[28-Jul-2010 11:15:23] I don't know the internal zenoss datastructure very well yet, though, which is why I wanted to discuss things with people who do know it
[28-Jul-2010 11:15:32] oh, no, the dependency stuff
[28-Jul-2010 11:15:35] (or lack thereof)
[28-Jul-2010 11:15:59]
[28-Jul-2010 11:16:06] huh
[28-Jul-2010 11:16:15] Jane_Curry: don't print the webpage? ;)
[28-Jul-2010 11:16:41] cgibbons: it's possible that I've jumped onto the wrong train from the wrong platform here, too ;)
[28-Jul-2010 11:16:56] * Simon4 heads home, via the apple store, then some coding I think
[28-Jul-2010 11:17:00] frozty_sa: I shalt play :)
[28-Jul-2010 11:17:19] Simon4: I'm still at the office, but ping you when I get home as well?
[28-Jul-2010 11:17:34] frozty_sa: for sure, Il'l be an hour or two, need to get laptop + gym
[28-Jul-2010 11:17:53] * Simon4 has been given permission to upgrade the macbook :D
[28-Jul-2010 11:18:05] nice
[28-Jul-2010 11:18:13] i want one of the new pros. i miss the one i foolish sold off a few months ago :P
[28-Jul-2010 11:18:23] cgibbons: I have a nice mid-2009 15" for sale ;)
[28-Jul-2010 11:18:28] one careful owner :P
[28-Jul-2010 11:18:33] i mean the desktop one
[28-Jul-2010 11:18:40] ahhh
[28-Jul-2010 11:18:42] yeha, they're well nice
[28-Jul-2010 11:18:49] i had one of the original 2006 pros and it was still faster than even the current iMac i'm using for work
[28-Jul-2010 11:20:53] I had been trying to import a Products.ZenHub.PBDaemon but I cann't. It's on CentOS
[28-Jul-2010 11:21:43] anyone who had been working with it?
[28-Jul-2010 11:22:00] frozty_sa: I don't WANT the web page - I just want a list of devices of a certain class
[28-Jul-2010 11:22:39] With zendmd, I can do
[28-Jul-2010 11:22:44] >>> devpkg.dmd.Devices.Service.System.Application.Package.getSubDevices()
[28-Jul-2010 11:22:46] []
[28-Jul-2010 11:26:14] which gets me to the object I want .... what's printing the web page?????
[28-Jul-2010 11:27:42] i'd probably have to see the whole .pt to understand it more, Jane
[28-Jul-2010 11:44:32] is there a zenoss dev channel?
[28-Jul-2010 11:53:37] this is it
[28-Jul-2010 11:58:23] huh? did you see my .pt ??
[28-Jul-2010 11:58:36] sorry in & out the past few minutes, i will look shortly
[28-Jul-2010 11:58:40] the other comment was for chachan
[28-Jul-2010 11:58:46] many thanks......
[28-Jul-2010 12:00:45] cgibbons, ok, I had hours trying to import Products classes and I can't :(
[28-Jul-2010 12:00:54] from where?
[28-Jul-2010 12:01:09] from $ZENHOME/libexec
[28-Jul-2010 12:01:25] so from your own standlone python script? are you importing Globals first?
[28-Jul-2010 12:01:48] yes and no
[28-Jul-2010 12:01:54] global?
[28-Jul-2010 12:02:07] import Globals
[28-Jul-2010 12:02:58] uhm, no, I was trying with: import Products.ZenHub.PBDaemon
[28-Jul-2010 12:03:16] and it can't find it
[28-Jul-2010 12:04:08] ok, I can now import Globals and PBDaemon, I'll try to run my script
[28-Jul-2010 12:05:15] Globals is a zope thing that fixes up all the internal Product modules etc
[28-Jul-2010 12:11:59] cgibbons, ok, I already can import Globals, but I wish to send an event through my script
[28-Jul-2010 12:13:08] I'm reading the dev guide and says me that I need to import Products.ZenEvents.ZenClasses
[28-Jul-2010 12:13:42] so, how could I call sendEvent method?
[28-Jul-2010 12:15:51] I think that Products.ZenEvents.ZenClasses is not about python classes
[28-Jul-2010 12:18:31] does the example in the dev guide not work for you?
[28-Jul-2010 12:20:01] cgibbons, no
[28-Jul-2010 12:20:28] cgibbons, well, first, I didn't see the Globals import
[28-Jul-2010 12:20:54] dgibbons is a very bad man
[28-Jul-2010 12:20:57] cgibbons, ah, And I don't want to create a new class like guide says
[28-Jul-2010 12:22:17] well then wheat about the xmlprc example jsut below it instead?
[28-Jul-2010 12:23:06] cgibbons, ok, I will try it in a few minutes
[28-Jul-2010 12:23:22] cgibbons, I'll let you know if it works for me
[28-Jul-2010 12:24:14] admin:zenoss is meaning user:pass right?
[28-Jul-2010 12:24:39] the first example requires a PBDaemon instance because it needs to connect to zenhub & setup everything involved with that. If you arent doing a new daemon, then the xmlrpc way is more correct. Thats why we pointed you at zensendevent yesterday since it does that exactly. You can either use it as is or do it like the dev guide example is.
[28-Jul-2010 12:25:32] yeah that's the user pass
[28-Jul-2010 12:26:44] cgibbons, thanks, you pointed me yesterday to zensendevent, but, is it a python class, or what is it?
[28-Jul-2010 12:30:52] very impressed with v3.0, good work guys
[28-Jul-2010 12:36:24] it's a python script in $ZENHOME/bin that you can look at as an example or use it directly.
[28-Jul-2010 12:36:32] it you can send events from a shell script with it.
[28-Jul-2010 12:41:29] does zenoss 3 support device dependencies now?
[28-Jul-2010 12:42:17] mducharme-laptop: no
[28-Jul-2010 12:42:26] ok
[28-Jul-2010 12:42:29] it does layer 3 dependencies, but that doesn't really work very well
[28-Jul-2010 12:42:48] right
[28-Jul-2010 12:42:59] they need to implement later 2 dependencies and manual dependencies at some point to make it truly effective
[28-Jul-2010 12:43:04] layer*
[28-Jul-2010 12:43:05] because that's one thing I will need to sort out with OTRS
[28-Jul-2010 12:43:09] and the CMDB
[28-Jul-2010 12:43:19] OTRS allows you to define dependencies between devices in its CMDB
[28-Jul-2010 12:43:25] I see
[28-Jul-2010 12:43:32] it would be ideal to sync that down from zenoss to otrs
[28-Jul-2010 12:43:49] yeh
[28-Jul-2010 12:47:37] cgibbons, ok :)
[28-Jul-2010 12:47:40] cgibbons: is there some way to convert evt.firstTime and evt.lastTime to an integer for processing?
[28-Jul-2010 12:47:59] cgibbons: I'm basically trying to compare the last time of an event to the current time to determine if the event is an hour or more old
[28-Jul-2010 13:02:17] sideone: got wave?
[28-Jul-2010 13:02:48] rmatte: import datetime, timedelta
[28-Jul-2010 13:02:56] (timedelta in datetime)
[28-Jul-2010 13:03:03] sideone: err soz, not you
[28-Jul-2010 13:03:31] today needs another 4 hours or so, even though I don't have the energy for them :/
[28-Jul-2010 13:10:19] does anybody know how to trigger a delete of zproperties?
[28-Jul-2010 13:10:32] like a system wide delete
[28-Jul-2010 13:11:40] I just sent you a link that has an example of doing just that
[28-Jul-2010 13:12:10] looks quite clean too
[28-Jul-2010 13:13:56] ahh thanks
[28-Jul-2010 13:14:59] I need to figure out how to convert evt.lastTime to an integer
[28-Jul-2010 13:15:02] fun fun
[28-Jul-2010 13:15:25] I might have to split it up and store it as a list to be able to convert it
[28-Jul-2010 13:18:10] yup, I think that's the way to go about it
[28-Jul-2010 13:26:52] hmmm, that won't work either apparently
[28-Jul-2010 13:27:42] ah sweet
[28-Jul-2010 13:27:44] found out how
[28-Jul-2010 13:28:32] hi everyone
[28-Jul-2010 13:36:30] message/51133#51133
[28-Jul-2010 13:37:58] that's an awesome daisy-chain of function calls :)
[28-Jul-2010 13:38:03] yup
[28-Jul-2010 13:38:05] it works
[28-Jul-2010 13:39:16] rmatte: why would you do that?
[28-Jul-2010 13:39:40] twm1010: I'm going to add a routine to my daemon that auto-moves events older than an hour to history
[28-Jul-2010 13:39:52] since the Zenoss routine for doing it is flawed (which I already have a trac ticket open for)
[28-Jul-2010 13:40:00] flawed in all versions?
[28-Jul-2010 13:40:05] i wished it was more flexible
[28-Jul-2010 13:40:05] correct
[28-Jul-2010 13:40:13] it will never clear Critical events
[28-Jul-2010 13:40:35] have different send to history for the diff severities
[28-Jul-2010 13:41:02] the other possible use for this, which is actually incredible... is to have a transform do something based on the time between first time and last time
[28-Jul-2010 13:41:24] you could basically tell it something like, if the event has been occuring for over an hour... up the severity
[28-Jul-2010 13:41:26] that sort of thing
[28-Jul-2010 13:41:35] i do that on the alerting side
[28-Jul-2010 13:41:59] same kinda effect, on the assumption someone should be watching the dashboard
[28-Jul-2010 13:42:20] what is a database conflict error
[28-Jul-2010 13:42:22] one rule lets techs know about something after 15 minutes, and if nobody acknowledges it within an hour, a different rule hits a different group to escalate it
[28-Jul-2010 13:42:41] but, changing the severity would be nice too
[28-Jul-2010 13:42:48] don't some thresholds do that?
[28-Jul-2010 13:42:57] lets X occurrences escalate?
[28-Jul-2010 13:43:15] how can I track down the cause of a database conflict error
[28-Jul-2010 13:43:17] that does it based on counts
[28-Jul-2010 13:43:18] but yes
[28-Jul-2010 13:43:31] there are situations where doing this could apply though
[28-Jul-2010 13:44:39] mducharme-laptop: 99% of the time it's an indexing issue
[28-Jul-2010 13:44:44] go in to zendmd and do the following...
[28-Jul-2010 13:44:46] reindex()
[28-Jul-2010 13:44:49] commit()
[28-Jul-2010 13:44:51] ctrl-d to exit
[28-Jul-2010 13:45:04] that should resolve the conflict
[28-Jul-2010 13:45:13] is there a core equivalent to predictive thresholding?
[28-Jul-2010 13:45:28] twm1010: yes, the pack that rocket made
[28-Jul-2010 13:45:35] which is arguably better than the enterprise
[28-Jul-2010 13:45:38] I've determined that that's why I could not uninstall the pack, rmatte
[28-Jul-2010 13:46:04] twm1010: docs/DOC-3408
[28-Jul-2010 13:46:22] not sure if it'll work properly in 3.0
[28-Jul-2010 13:46:38] it should
[28-Jul-2010 13:46:43] since it's just a datasource type
[28-Jul-2010 13:46:47] i might have to try that
[28-Jul-2010 13:47:02] it'd be nice to do that on the filesystem template to trend growth
[28-Jul-2010 13:47:15] yeh
[28-Jul-2010 13:47:17] and then create a report that aggregates that
[28-Jul-2010 13:47:20] that would be useful
[28-Jul-2010 13:47:24] show how fast we're consuming disk across all servers
[28-Jul-2010 13:47:28] in case the filesystem usage suddenly shoots up or something
[28-Jul-2010 13:47:33] I might actually implement that myself
[28-Jul-2010 13:48:01] BBL
[28-Jul-2010 13:55:47] * Simon4 plays around with setting up a dependency tree for devices
[28-Jul-2010 13:58:18] mducharme-laptop: cool, it uninstalls after the reindex?
[28-Jul-2010 14:00:29] test
[28-Jul-2010 14:15:49] is there a zendmd (or other way) to rename a sub-system defined for a /Systems hierarchy?
[28-Jul-2010 14:17:03] gwb2351: you mean you want to rename the organizer itself?
[28-Jul-2010 14:17:12] pretty much
[28-Jul-2010 14:17:24] you can do that via the zope management interface, not sure what issues it may cause though
[28-Jul-2010 14:17:25] How long does it actually take to install zenoss using the debian package, it's utilizing 100% cpu for about 10-15 minutes now on 2.6ghz 4gb ram machine
[28-Jul-2010 14:17:53] Xeross: it'll sit at 99% for a while
[28-Jul-2010 14:18:06] i have an old organizer name (XXX) that is now know as YYY. There's 362 devices in the old XXX hierarchy, and wondering if I can rename XXX to YYY instead of creating YYY and then moving devices from XXX to YYY
[28-Jul-2010 14:20:07] i've done this in the past via:
[28-Jul-2010 14:20:21] [zenoss@eye zenbatchload]$ awk '{print $1}' oracle.txt | { while read line; do echo "* $line"; \ curl -L -u admin:XXXXX http://eye.its.utexas.edu:8080/zport/dmd/Devices/Server/Passive/devices/$line\ /manage_editDevice?performanceMonitor=cornea\&systemPaths=/Systems/Database/oracle 2>&1>/dev/null; done; }
[28-Jul-2010 14:20:37] but that's done per device. just wondering if there's a shortcut
[28-Jul-2010 14:21:55] rmatte: If I understand correctly zenoss is a SNMP frontend and more ?
[28-Jul-2010 14:22:04] way more
[28-Jul-2010 14:23:20] gwb2351: As long as it allows monitoring of my servers, and looks half decent I'm happy
[28-Jul-2010 14:27:06] If I already have an existing mysql database and I'm installing zenoss, does anyone know which file I edit to put in our personal details?
[28-Jul-2010 14:27:28] My coworker is working on automating the install but doesn't know what file to change so zenoss can connect and slap in the blank databases
[28-Jul-2010 14:30:52] hmm, this new laptop is superfast, my zenoss dev vm's fly
[28-Jul-2010 14:31:45] ashp: normally the auth details are kept in the zope db, I'm trying to think how you could inject the auth details into the first run process since that proces creates the zope db then populates the mysql etc
[28-Jul-2010 14:36:27] oh, argh, maybe we didn't do it that way, this is trickier then
[28-Jul-2010 14:36:40] I can adjust our mysql module to not set a password if it's being installed on the zenoss host
[28-Jul-2010 14:36:44] but it's a bit of a hack
[28-Jul-2010 14:37:05] I guess I had a false memory of tweaking something prior to first run so it could connect with our standard details
[28-Jul-2010 14:39:19] so the first run a) populates a zopedb, then b) sets up the mysql db using "zenoss:zenoss" as the user/pass
[28-Jul-2010 14:39:42] you could have a look at the first run script and see if your automation couldn't patch in a diff user/pass somewhere, or insert some dmd script to adjust it along the way
[28-Jul-2010 14:41:45] Xeross: yes, you'll be able to monitor your servers with it
[28-Jul-2010 14:41:51] Xeross: what kinds of servers?
[28-Jul-2010 14:42:16] Xeross: also, Zenoss support SNMP, WMI, SSH, and custom methods to grab data
[28-Jul-2010 14:42:20] supports*
[28-Jul-2010 14:46:58] Simon4: I'll have a look and try to find this first run script
[28-Jul-2010 14:47:09] I left it to him to handle but I think I overwhelmed him, I'll have a look :)
[28-Jul-2010 14:47:10] ashp: it's referenced in the init script
[28-Jul-2010 14:47:27] checks for a .firstrun or .freshinstall file or similar, then goes off and runs the first run stuff
[28-Jul-2010 14:57:47] cgibbons, I'm trying in this way: ./zensendevent --device server.com --eventclass "/Cmd/Fail" --auth admin:zenoss -v Testing
[28-Jul-2010 14:58:01] but I got this xmlrpclib.Fault:
[28-Jul-2010 14:58:21] user and password are right
[28-Jul-2010 14:58:44] do I need some rpc service running or something like that?
[28-Jul-2010 15:00:23] and then I tried using python console and I got this: xmlrpclib.ProtocolError:
[28-Jul-2010 15:01:00] I don't know why I only get errors, no matter the way :(
[28-Jul-2010 15:03:46] is localhost:8080 accessible?
[28-Jul-2010 15:03:55] like, are you doing this for a different server?
[28-Jul-2010 15:03:59] from*
[28-Jul-2010 15:06:55] rmatte, yes, it's the zenoss server and telnet request is right
[28-Jul-2010 15:07:05] I use the FQDN too
[28-Jul-2010 15:07:12] getting the same response
[28-Jul-2010 15:10:58] hello all
[28-Jul-2010 15:11:09] hi
[28-Jul-2010 15:11:18] is there a way to get Zenoss to send info messages to one address, warning to another, error to another without creating users for each and alert users per user?
[28-Jul-2010 15:12:04] QubeZ: you can create a dummy "alerts" user, and then in each alerting rule you can specify a different email address
[28-Jul-2010 15:12:24] thats what i've done
[28-Jul-2010 15:12:35] * Simon4 does that also
[28-Jul-2010 15:12:40] yeh, create 3 different alerting rules with different filters
[28-Jul-2010 15:12:45] then set the "Address (optional)" field
[28-Jul-2010 15:12:51] and you'll get exactly what you're asking for
[28-Jul-2010 15:13:14] i just set the email when i create the user, dont need to set optional field then right?
[28-Jul-2010 15:13:33] YES, BUT, you're asking to have emails sent to 3 different addresses
[28-Jul-2010 15:13:38] meaning don't set the email for the user
[28-Jul-2010 15:13:43] set it per alerting rule instead
[28-Jul-2010 15:13:45] understand?
[28-Jul-2010 15:13:54] weird that in firefox 3, i can't get the little gear icon to drop down options but in safari it works
[28-Jul-2010 15:14:09] QubeZ: works fine for me in Firefox 3
[28-Jul-2010 15:14:18] clear your browser cache
[28-Jul-2010 15:14:21] you might have old data
[28-Jul-2010 15:14:31] rmatte: nope, im asking to have info messages sent to one user info_notification@domain.com, and on our mail server, we've aliased that for three users so we can subscribe to the email list
[28-Jul-2010 15:14:43] yes, I understand that
[28-Jul-2010 15:14:48] * rmatte tears hair out...
[28-Jul-2010 15:14:53] ok, here's what you do...
[28-Jul-2010 15:14:55] sorry dude
[28-Jul-2010 15:15:10] you create 1 alerting rule called "info", and you set the filters on it to only trigger on info events
[28-Jul-2010 15:15:17] done
[28-Jul-2010 15:15:21] you then set the email address for that alerting rule to info_notification@domain.com
[28-Jul-2010 15:15:29] you then create a second alerting rule called "warning"
[28-Jul-2010 15:15:40] and do the same for that, but set the warning email address in that alerting rule
[28-Jul-2010 15:15:43] oh i see what you mean, got it
[28-Jul-2010 15:15:45] and repeat for the third alerting rules as well
[28-Jul-2010 15:15:49] rather then 3 users, one user with 3 emails
[28-Jul-2010 15:15:53] yes
[28-Jul-2010 15:16:08] 1 user with 3 alerting rules with a different address for each and different filters
[28-Jul-2010 15:16:18] *nod*
[28-Jul-2010 15:16:55] k, good
[28-Jul-2010 15:16:56] lol
[28-Jul-2010 15:30:40] * Simon4 laughs
[28-Jul-2010 15:30:53] so I'm sitting here, waiting for a device to be marked down by zenoss
[28-Jul-2010 15:30:57] wondering why it isn't, etc
[28-Jul-2010 15:31:07] turns out my dev box has zenping disabled :/
[28-Jul-2010 15:32:06] LOL
[28-Jul-2010 15:32:25] thats okay Simon, i've done worse
[28-Jul-2010 15:32:30] :)
[28-Jul-2010 15:33:02] like when i tried to do the whole operational interface status thing through the performance template and my voice gateways generated 4000 e-mails in about 90 seconds
[28-Jul-2010 15:33:04] to my whole team
[28-Jul-2010 15:33:09] that was awesome
[28-Jul-2010 15:33:20] :D
[28-Jul-2010 15:33:41] for a while our test zenoss box was set up to not send email just by having postfix turned off with chkconfig
[28-Jul-2010 15:33:49] a new admin came along, saw it was stopped it
[28-Jul-2010 15:34:02] postfix promptly tried to clear 5 months of zenoss alert mail queue
[28-Jul-2010 15:34:10] we don't stop it sending emails that way any more ;)
[28-Jul-2010 15:34:26] oops, saw it was stopped and started it, even
[28-Jul-2010 15:35:30] hurm
[28-Jul-2010 15:36:10] I've been trying to revisit my interface status notifications just haven't had time
[28-Jul-2010 15:36:31] I came to the conclusion that for just voice gateways, i would filter out serial interfaces and just watch the controllers via a zproperty ona device class
[28-Jul-2010 15:36:39] I'm having some sortf of inconsistency with zproperties
[28-Jul-2010 15:36:47] the same zproperty appeared twice in the list
[28-Jul-2010 15:36:49] problem is, at some remote sites I have voice gateways handling both data and voice circuits
[28-Jul-2010 15:36:56] I deleted it, they both disappeared
[28-Jul-2010 15:37:08] i suppose in those instances i'd override the inteface template for the data circuits and let it inherit the others
[28-Jul-2010 15:37:16] still, i'd prefer some better method
[28-Jul-2010 15:37:17] I can no longer delete it because it says it does not exist, but I can no longer re-create it because it says it exists
[28-Jul-2010 15:37:34] mducharme-laptop: try clearing your browser cache?
[28-Jul-2010 15:37:55] the browser cache would not impact the zendmd command line
[28-Jul-2010 15:38:21] heh, ok, you got me there ;)
[28-Jul-2010 15:39:17] dmd.Devices._setProperty('zpropertyname', 'hello', 'string') fails because it exists
[28-Jul-2010 15:39:31] dmd.devices._delProperty('zpropertyname') fails because it does not exist
[28-Jul-2010 15:39:42] er Devices
[28-Jul-2010 15:41:46] the specific error when it tries to create it with setproperty is
[28-Jul-2010 15:41:59] BadRequest: Id zOTRSSOAPLogin is invalid or duplicate
[28-Jul-2010 15:45:45] hrmm... does it work fine through the GUI?
[28-Jul-2010 15:45:55] if so, i'd think you've just got a simple syntax issue
[28-Jul-2010 15:49:28] no.. ti's not a syntax issue, the property is not there
[28-Jul-2010 15:50:17] but it is there
[28-Jul-2010 15:50:18] but it's not there
[28-Jul-2010 15:50:30] I think I have an inconsistent database :(
[28-Jul-2010 15:50:56] I may have to dig up phpmyadmin and try to find the offending record
[28-Jul-2010 15:50:58] and manually delete it
[28-Jul-2010 15:51:03] is there anyway to close an event and not have it send an email? For example, I have a bunch of fa interfaces on a switch that are showing down but are unused, so I've remove monitoring. I dont want to clear out the events the normal way because an email will go out for each one.
[28-Jul-2010 15:51:24] mducharme-laptop: I have news for you, and it's all bad
[28-Jul-2010 15:51:33] the offending record is in zope db, not mysql
[28-Jul-2010 15:51:38] so no phpmyadmin
[28-Jul-2010 15:51:53] what?
[28-Jul-2010 15:51:58] oh
[28-Jul-2010 15:52:02] how do I find it then
[28-Jul-2010 15:52:24] generally using the methods like you have been, or by finding other methods that do what you want
[28-Jul-2010 15:53:48] * Simon4 needs to get himself a proper zope book one of these days
[28-Jul-2010 15:54:21] does one exist?
[28-Jul-2010 15:54:33] probably not
[28-Jul-2010 15:54:41] I was being hopeful :)
[28-Jul-2010 15:56:47] valid_property_id is what it uses
[28-Jul-2010 16:08:22] I have some kind of inconsistency
[28-Jul-2010 16:08:25] and I have no idea why
[28-Jul-2010 16:08:31] nobody's perfect
[28-Jul-2010 16:08:43] * Simon4 chortles
[28-Jul-2010 16:12:22] I'm just trying to get these zproperties deleted.....
[28-Jul-2010 16:12:37] so that they are gone
[28-Jul-2010 16:12:40] so I can recreate them
[28-Jul-2010 16:12:46] I suppose I could make new ones with new names
[28-Jul-2010 16:12:52] that works
[28-Jul-2010 16:12:57] but if I use the old name, it does not
[28-Jul-2010 16:13:10] still that would make me nervous as far as upgrades go
[28-Jul-2010 16:13:29] if there is inconsistency like that, it would not be good if it caused upgrade issues
[28-Jul-2010 16:14:16] mducharme-laptop: would certainly be worth trying it on a fresh db etc, check if it's something that happens every time or just on the off chance
[28-Jul-2010 16:14:52] I am simply unable to create a zproperty with that name
[28-Jul-2010 16:15:03] zOTRSSOAPLogin or zOTRSSOAPPassword
[28-Jul-2010 16:15:06] it claims they are unavailable
[28-Jul-2010 16:15:24] yet I had properties with those names before
[28-Jul-2010 16:15:27] the showed up twice in the list
[28-Jul-2010 16:15:32] I saw duplicates of them....
[28-Jul-2010 16:15:53] in the zProperties list in the web browser the property names and values appeared twice
[28-Jul-2010 16:16:27] that's why I think this is inconsistent now
[28-Jul-2010 16:22:46] what precisely should I try with a fresh db, simon4?
[28-Jul-2010 16:23:31] and how do I temporarily get a fresh db
[28-Jul-2010 16:24:46] mducharme-laptop: I was thinking just replicating what you did in terms of creating and trying to remove/recreate these zproperties
[28-Jul-2010 16:26:12] what purpose would replicating what I did serve?
[28-Jul-2010 16:26:30] and I don't know if it will be possible to replicate
[28-Jul-2010 16:26:40] you were saying it made you nervous in terms of upgrades, I was just suggesting trying to see if it would replicate
[28-Jul-2010 16:26:48] I may have gotten the wrong end of the stick
[28-Jul-2010 16:27:00] well yes
[28-Jul-2010 16:27:09] because somehow I have zProperties that are half-there
[28-Jul-2010 16:27:16] there in one database table, not the in another
[28-Jul-2010 16:27:23] so they can neither be created nor deleted
[28-Jul-2010 16:27:46] that makes me nervous
[28-Jul-2010 16:27:47] citrus is now known as citrus2
[28-Jul-2010 16:28:17] again, I can work around this by slightly alterning the name
[28-Jul-2010 16:28:22] or changing the capitalization
[28-Jul-2010 16:28:25] in Zenoss 3.0 how do i change the WMI user/pass for all servers in a certain location
[28-Jul-2010 16:28:28] I can create properties with any slightly different name
[28-Jul-2010 16:28:32] but those names do not work
[28-Jul-2010 16:28:50] I have been doing a lot of creating and deleting of properties lately, it has all been fine, except this
[28-Jul-2010 16:37:04] I can create and delete the property at the dmd root, but not at the Devices level
[28-Jul-2010 16:50:52] is it possible to have conditional alerts in 3.0? such as if alert for this machine only if switch is up?
[28-Jul-2010 16:51:27] I know I bother you a lot with my question. But I can't really create an event at anyway
[28-Jul-2010 16:51:58] * Simon4 is working on conditional alerts in a zenpack right this second
[28-Jul-2010 16:52:03] not working quite yet though :/
[28-Jul-2010 16:52:52] Simon4, could you help me a little with my issue?
[28-Jul-2010 16:53:06] I can always try
[28-Jul-2010 16:54:16] thanks, I tried in many ways to create an event through python
[28-Jul-2010 16:55:08] chachan: you can use the "zensendevent" command-line utility, just call it from python
[28-Jul-2010 16:55:36] Simon4, I tried too
[28-Jul-2010 16:55:58] I'll let you know the response, please wait
[28-Jul-2010 16:57:52] Simon4, this is it: http://pastebin.org/426453
[28-Jul-2010 16:58:44] hrm, only thing I can think of is that you've changed the admin password to something else
[28-Jul-2010 16:58:50] otherwise I'm stumped sorry :/
[28-Jul-2010 16:59:16] Simon4, yes, it isn't the real password
[28-Jul-2010 16:59:29] but, the password hasn't special characters
[28-Jul-2010 16:59:49] ;'(
[28-Jul-2010 17:00:19] So Zenoss is running, snmpwalk works for both my systems, yet I get SNMP Agent Down errors
[28-Jul-2010 17:00:23] any idea why it's happening ?
[28-Jul-2010 17:00:38] chachan: sorry, I have no idea :/
[28-Jul-2010 17:00:44] I remember going somewhere in zenoss where I could browse arond the zope database
[28-Jul-2010 17:00:57] mducharme-laptop: stick /manage on the end of a url
[28-Jul-2010 17:07:23] is the properties tab under devices/manage supposed to throw an error?
[28-Jul-2010 17:08:04] like if you go to http://zenoss/zport/dmd/Devices/manage into the properties tab, does it throw an error?
[28-Jul-2010 17:08:29] So seems that zenoss wrongly configered the snmp community
[28-Jul-2010 17:10:23] the properties tab seems to work everywhere else except in Devices itself
[28-Jul-2010 17:10:30] in Devices itself it throws an error
[28-Jul-2010 17:12:21] How can I force ZenOss to update system info
[28-Jul-2010 17:12:30] it shows Memory/Swap: Unknown/Unknown
[28-Jul-2010 17:14:25] Xeross: that's done via a collector plugin...
[28-Jul-2010 17:14:35] what type of device are you modelling?
[28-Jul-2010 17:16:05] rmatte: 2 linux servers running SNMP, it probably failed fetching the info because it used a different SNMP community then I specified
[28-Jul-2010 17:20:38] rmatte, if I would like to use Products.ZenHub.PBDaemon through python script, what do I need?, don't see great help about it in the dev guide
[28-Jul-2010 17:21:14] I don't see*
[28-Jul-2010 17:24:56] Xeross: where did you specify the snmp community?
[28-Jul-2010 17:25:35] chachan: no idea, I've never used PBDaemon
[28-Jul-2010 17:26:07] chachan: perhaps this is what you're looking for?: docs/DOC-3217;jsessionid=80D27122402B508FC8040FBA903347A4.node0
[28-Jul-2010 17:27:47] rmatte, how do you create events?, do you have an example?
[28-Jul-2010 17:33:14] what do you mean?
[28-Jul-2010 17:33:33] you want to actually generate an event?
[28-Jul-2010 17:34:18] <|rt|> rmatte: I think I figured out why that transform you helped me with yesterday wasn't working
[28-Jul-2010 17:34:53] <|rt|> the field names are case sensitive and in the code we had some case issues
[28-Jul-2010 17:35:10] <|rt|> I gotta wait for another event to see if my tweak fixed it
[28-Jul-2010 17:35:15] yes, they are case sensitive
[28-Jul-2010 17:35:22] what did you change exactly?
[28-Jul-2010 17:35:43] rmatte, yes, create it through a command
[28-Jul-2010 17:35:49] I didn't recall seeing any case issues
[28-Jul-2010 17:36:05] chachan: I don't have an example on hand.
[28-Jul-2010 17:36:22] I've never had need to create an event using anything other than the UI or zensendevent
[28-Jul-2010 17:36:30] and I've only used zensendevent for testing
[28-Jul-2010 17:36:50] <|rt|> fgIPsTrapSigMsg was correct in this bit: iptrapsigmsg = getattr(evt, 'fgIpsTrapSigMsg', 'Unknown')
[28-Jul-2010 17:37:06] oh I see
[28-Jul-2010 17:37:20] I assumed you had cut and paste them to me
[28-Jul-2010 17:37:21] :P
[28-Jul-2010 17:37:30] ok, I tried but I get this: http://pastebin.org/426555
[28-Jul-2010 17:37:35] did I miss anything?
[28-Jul-2010 17:37:50] <|rt|> there was a similar issue with fgIpsTrapSrcIP...the P at the end of that should be p
[28-Jul-2010 17:38:10] getting a right answer from this would be great for me!
[28-Jul-2010 17:38:16] I get the same... let me figure this out
[28-Jul-2010 17:38:27] What would cause the memory utilization graph to say nan all over the place while the graph is displaying correctly
[28-Jul-2010 17:39:17] Xeross: you haven't waited 3 polling cycles (15 minutes)
[28-Jul-2010 17:39:32] just be patient and it'll start graphing eventually
[28-Jul-2010 17:41:19] rmatte: no it is already graphing, just the values beneath the graph show nan
[28-Jul-2010 17:44:53] chachan: hmmm, when I try passing the username/password to zenoss in the http line it doesn't login
[28-Jul-2010 17:45:01] chachan: I'm guessing that's what the issue is
[28-Jul-2010 17:45:45] chachan: looks like a bug
[28-Jul-2010 17:46:30] O.o
[28-Jul-2010 17:47:55] well, that doesn't sound great :( My boss is gonna kill me hehehe
[28-Jul-2010 17:48:31] I tried using wget and it response me the login form
[28-Jul-2010 17:48:37] Is my snmp properly secured if it's restricted to access of only 1 IP address, and it has a custom community ?
[28-Jul-2010 17:49:16] Xeross, I do it in that way
[28-Jul-2010 17:49:37] see you guys, I will stay logged
[28-Jul-2010 17:51:55] yeh, it's an authentication issue, and I can't seem to find the actual python code for creating an event
[28-Jul-2010 17:54:01] Seems my SNMP is configured to not return info on the memory size, etc. any pointers on how to correct this
[28-Jul-2010 17:55:04] by default net-snmp has paranoid settings in it's snmpd.conf
[28-Jul-2010 17:55:56] here's a working config: http://pastebin.com/yZ8Gw1uE
[28-Jul-2010 17:56:18] just modify it with your trap and community string settings (also location and contact)
[28-Jul-2010 17:56:43] either that or modify your existing config to remove the paranoid settings
[28-Jul-2010 17:59:23] rmatte: I commented a comomunities, and made my own readwrite one
[28-Jul-2010 17:59:41] rmatte: But with snmpwalk I don't get any memory info on the system view
[28-Jul-2010 18:02:11] chachan: I think I found how to do it via zendmd
[28-Jul-2010 18:03:07] is there a way to move multiple devices from one device class to another?
[28-Jul-2010 18:03:13] in 3.0?
[28-Jul-2010 18:03:17] rmatte: Or is the system view just for testing and irrelevant ?
[28-Jul-2010 18:03:33] bzed_ is now known as Guest94406
[28-Jul-2010 18:07:21] i have asked 3 questions in here.. not one can be answered. \
[28-Jul-2010 18:45:11] Xeross: you should see memory info there
[28-Jul-2010 18:45:28] are your performance graphs even showing yet?
[28-Jul-2010 18:46:48] if citrus2 had been a bit more patient he might have gotten his answers answered
[28-Jul-2010 18:48:44] his questions answered rather
[28-Jul-2010 19:09:47] chachan: I figured out how to create an event via zendmd
[28-Jul-2010 19:09:52] dmd.ZenEventManager.manage_addEvent(dict(device="testdevice", component="Whatever", summary="This is a test message!", eclass="/Cmd/Fail", severity="5"))
[28-Jul-2010 19:15:20] well, I'm well overdue to go home
[28-Jul-2010 19:15:25] later folks
[28-Jul-2010 19:16:07] Xeross: use the config that I pasted... you may think that you fixed your config but you probably didn't
[28-Jul-2010 19:16:37] also, make sure the device is in the /Server/Linux class
[28-Jul-2010 19:16:41] and not some other random class
[28-Jul-2010 19:16:50] since that class has the appropriate plugins for collecting that info
[28-Jul-2010 19:16:57] I'm out, I'll be back around tomorrow
[28-Jul-2010 19:16:58] cheers.
[28-Jul-2010 19:49:44] <|rt|> can you put html in an event summary with a transform and say have a clickable link in the zenoss UI?
[28-Jul-2010 19:53:03] <|rt|> looks like you can....atleast it works by manually putting html in the summary of a test event
[28-Jul-2010 20:27:53] rhettardo_ is now known as rhettardo
[28-Jul-2010 20:36:46] mrayzenoss is now known as mray
[28-Jul-2010 22:09:42] mrayzenoss is now known as mray
[28-Jul-2010 22:24:51] mrayzenoss is now known as mray
[28-Jul-2010 22:44:24] mrayzenoss is now known as mray
[29-Jul-2010 00:00:30] [disconnected at Thu Jul 29 00:00:30 2010]
[29-Jul-2010 00:00:31] [connected at Thu Jul 29 00:00:31 2010]
[29-Jul-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]
[29-Jul-2010 02:30:32] Hi there. I am trying to find the way to change the FS threshold using the default SNMP FS monitor, but strangely I cannot seem to find it in Zenoss 3.0. I can load and CPU thresholds, but not the FS threshold though I can see it in the graphs. Am I just missing something?
[29-Jul-2010 02:31:01] by load I actually mean find. I can find all of the others, just not the FS threshold settings.
[29-Jul-2010 02:32:02] rcngeoff: Advanced -> Monitoring Templates -> Filesystems -> /Server -> Thresholds
[29-Jul-2010 02:32:49] click locations, top, top, left, left, right
[29-Jul-2010 02:33:43] doh! Now it is obvious. Thanks.
[29-Jul-2010 02:34:23] it is unintuitive if you used zenoss2 before.
[29-Jul-2010 04:02:32] morning all
[29-Jul-2010 04:02:45] personal greetings ;)
[29-Jul-2010 04:03:21] mellow greetings to you ;)
[29-Jul-2010 04:56:57] frozty_sa: round one of me vs zenoss and dependencies = zenoss won, however I know what I was doing wrong, so need to play a bit more tongiht
[29-Jul-2010 04:57:00] tonight even
[29-Jul-2010 04:57:50] Simon4: ah k
[29-Jul-2010 04:57:59] I was so completely gone last night :/
[29-Jul-2010 04:58:07] * frozty_sa is getting/has gotten sick
[29-Jul-2010 04:58:13] crappy :/
[29-Jul-2010 04:58:17] aye
[29-Jul-2010 04:58:51] Simon4: how are you approaching it at the moment?
[29-Jul-2010 05:00:20] Simon4: Are you the guy who wrote the TechRant about adding more than one device with the same IP????
[29-Jul-2010 05:00:28] frozty_sa: so I tried monkeypatching the Device object to have a list of "parents" which was a list of device objects, then wrote a function that was monkeypatched in to basically run getStatus() on the parent deivces, working it's way up the tree
[29-Jul-2010 05:00:40] however a list of devices doesn't work, it needs to be a proper relation to the device
[29-Jul-2010 05:00:49] I think the plan is still sound though
[29-Jul-2010 05:00:51] Jane_Curry: I am
[29-Jul-2010 05:01:05] Simon4: I was pondering some things yesterday, though
[29-Jul-2010 05:01:33] I have been using your script for several weeks now but I suddenly find that some devices are throwing an error event with
[29-Jul-2010 05:02:06] 192.168.10.129 sendto error Host raddlefredpackage.class.example.org and zenossfredsystem.class.example.org are both using ip 192.168.10.129
[29-Jul-2010 05:02:13] Any ideas???
[29-Jul-2010 05:02:31] hmm, what daemon is that error in?
[29-Jul-2010 05:02:45] let's say we have a 3-tier structure. top-level we have one core router (CR), then two core switches below that (CS01/CS02), and then 15 hosts connected to each CS
[29-Jul-2010 05:03:05] I actually have 6 "devices" all with CNAME entries that point to one canonical DNS entry - only 2 of the 6 are winging
[29-Jul-2010 05:04:15] Error is coming from zenping
[29-Jul-2010 05:04:25] * Simon4 has a look at his config
[29-Jul-2010 05:04:56] if CS01 dies, each of H1-(1..15) will die, but if we process those hosts' events on (ie, we're going /up/ the tree), we'll have to do some sort of assert to suppress all of them if evt.device.parentDevice().getStatus != down?
[29-Jul-2010 05:05:04] Jane_Curry: cached DNS values, perhaps?
[29-Jul-2010 05:05:25] oh, wait, nevermind. I'm on a wrong page :P
[29-Jul-2010 05:06:48] Jane_Curry: I have ping supressed for my "clone" devices, they sit in a device class with pretty much everything turned off
[29-Jul-2010 05:07:01] since there's only really point in pinging it once
[29-Jul-2010 05:08:23] strange though that some of my CNAMEs are wingeing and others are not???
[29-Jul-2010 05:08:27] yeah, very
[29-Jul-2010 05:08:38] unless it happens that they're being grouped in some random way by zenping
[29-Jul-2010 05:08:49] in terms of the order in which it's pinging things
[29-Jul-2010 05:10:29] frozty_sa: initially I'm working on providing a method to the Device object which returns "is anything above me in the defined dependency chain down", which can then be used in an event transform to surpress events as needed
[29-Jul-2010 05:10:38] so you could basically say:
[29-Jul-2010 05:10:46] if dev.getParentStatus():
[29-Jul-2010 05:10:55] evt._action = "drop"
[29-Jul-2010 05:10:57] or whatnot
[29-Jul-2010 05:11:25] well, _action = "history" in my case, but yes
[29-Jul-2010 05:11:32] yeah
[29-Jul-2010 05:11:47] but at least for first cut provide enough functionality to let you event transform based on this
[29-Jul-2010 05:12:05] if one takes it out of event transforms, so that it becomes native, where does it go? in the eventmanager path somewhere?
[29-Jul-2010 05:12:31] and how do you assign/manage the host dependencies at the moment?
[29-Jul-2010 05:12:34] that would be my thought, and yeah, that would be round two I think
[29-Jul-2010 05:12:57] currently I set em with zendmd, but you could add it as a zProperty list
[29-Jul-2010 05:13:06] which would mean that it was inherited into device classes etc
[29-Jul-2010 05:13:08] mmm
[29-Jul-2010 05:13:23] how does the network map do it's thing?
[29-Jul-2010 05:13:26] its*
[29-Jul-2010 05:13:48] I'm wondering if things could attempt to seed the dependency stuff from that..
[29-Jul-2010 05:13:50] no idea, I haven't done very much with that side of things as it's always been buggy
[29-Jul-2010 05:14:10] you certainly could attempt to do that
[29-Jul-2010 05:19:11] I just added the DigMonitor zenpack to one of my devices, double clicking the plugin should make me able to configure it according to the manual
[29-Jul-2010 05:19:19] but the said configuration properties aren't there
[29-Jul-2010 05:19:31] http://pastie.org/1065229 get this when modelling hp procurve switches
[29-Jul-2010 05:19:44] doesnt seem that good
[29-Jul-2010 05:24:43] What kind of permissions should the zenoss ssh user have on a system
[29-Jul-2010 05:25:28] depends on what plugins you plan on using
[29-Jul-2010 05:27:01] tehhobbit: So give it limited access but access to the binaries it needs basically /
[29-Jul-2010 05:32:12] Guest94406 is now known as bzed
[29-Jul-2010 05:35:55] yes
[29-Jul-2010 08:14:27] <_rain_man_> is there any zenoss experts here at all?
[29-Jul-2010 08:16:33] <_rain_man_> :(
[29-Jul-2010 08:18:24] <_rain_man_> ok well if any of you could help with the following error it would be appreciated
[29-Jul-2010 08:19:20] <_rain_man_> Error running su zenoss -c "/usr/local/zenoss/zenoss/bin/zenoss" start : su: incorrect password
[29-Jul-2010 08:19:33] <_rain_man_> I get this error on both ubuntu and SUSE
[29-Jul-2010 08:20:13] su - zenoss
[29-Jul-2010 08:20:16] and set a password
[29-Jul-2010 08:20:36] well sudo su - ( to root 1st)
[29-Jul-2010 08:21:20] <_rain_man_> ok st3v3o thanks, but I understood that the zenoss installer would create the user and password
[29-Jul-2010 08:23:22] I havn't used the stack installed ..so I can't speak on that …I use the native RPMs
[29-Jul-2010 08:23:49] it creates the zenoss user but does not set a password (not even a blank one) explicitly.
[29-Jul-2010 08:24:05] <_rain_man_> ok thanks st3v3o
[29-Jul-2010 08:24:34] <_rain_man_> so cgibbons i have to give the zenoss user a password then install again?
[29-Jul-2010 08:25:02] no, you can set a password like st3v3o said, or perhaps even better configure sudo to let you send commands to zenoss as needed.
[29-Jul-2010 08:25:24] as zenoss, rather
[29-Jul-2010 08:25:43] <_rain_man_> oh ok I see
[29-Jul-2010 08:26:22] <_rain_man_> so the instruction manual is wrong :(
[29-Jul-2010 08:26:24] and if you're trying to just start zenoss, there's a script called /etc/init.d/zenoss-stack that you can use to start or stop the whole stack from root. sudo /etc/init.d/zenoss-stack start, for example.
[29-Jul-2010 08:26:43] hmm, which one you using? i'll look & tell our doc writer if iso
[29-Jul-2010 08:26:57] <_rain_man_> no I'm trying to install it and that's when I get the error
[29-Jul-2010 08:27:26] <_rain_man_> I'm using zenoss 3.0.0. on ubuntu and SUSE and see the same errors on both
[29-Jul-2010 08:29:49] That command you tried would work, if you were doing it from the root user. Presumably you had to run the stack installer as root, either in a root shell, or via sudo, to get it to install.
[29-Jul-2010 08:31:14] <_rain_man_> that is correct yes
[29-Jul-2010 08:33:01] <_rain_man_> It's frustrating when you copy exactly what the manual tells you to do, but it doesn't work!
[29-Jul-2010 08:35:24] again, which manual are you using?
[29-Jul-2010 08:35:38] I'm looking @ this one: docs/DOC-7398 and don't see that step
[29-Jul-2010 08:36:01] assuming you used the stack installer
[29-Jul-2010 08:37:31] A lot of times something like this is caused by an assumption in the documentation that an experienced admin wouldn't notice, but it isn't clear enough for newer users.
[29-Jul-2010 08:44:25] <_rain_man_> yes that is the manual i am using
[29-Jul-2010 08:45:02] <_rain_man_> i'm using the stack installer on both ubuntu and SUSE
[29-Jul-2010 08:59:56] Good morning everyone
[29-Jul-2010 09:00:08] rmatte, ok, I'll check it out
[29-Jul-2010 09:00:51] <_rain_man_> afternoon
[29-Jul-2010 09:01:13] <_rain_man_> chachan, you know anything about installing zenoss core?
[29-Jul-2010 09:02:20] _rain_man_, you mean, just zenoss?
[29-Jul-2010 09:02:44] <_rain_man_> yes
[29-Jul-2010 09:03:07] ok, I did it on
[29-Jul-2010 09:03:14] on CentOS and ubuntu
[29-Jul-2010 09:03:36] are you having any problem?
[29-Jul-2010 09:03:44] <_rain_man_> do you have any notes for when you done it on ubuntu at all?
[29-Jul-2010 09:05:36] <|rt|> ubuntu is pretty easy...just add the repository and apt-get install zenoss-stack
[29-Jul-2010 09:06:14] <_rain_man_> yes I tried that, but get errors
[29-Jul-2010 09:06:17] <|rt|> I did have to comment out a few lines in the /home/zenoss/.bashrc so that the cron.daily/zenoss would run
[29-Jul-2010 09:06:44] I did it using .deb and I read the official installation guide. But |rt| says that zenoss is in repo, so, you can also try this way :D
[29-Jul-2010 09:07:02] <|rt|> no zenoss project provides their own repo
[29-Jul-2010 09:07:14] |rt|, uhm, better :)
[29-Jul-2010 09:07:18] <|rt|> it's not in the ubuntu repo
[29-Jul-2010 09:07:21] <_rain_man_> I've tried both ways and get the same error to do with some user. So I tried on Suse and got the same :(
[29-Jul-2010 09:07:38] _rain_man_, can you tell more about this error?
[29-Jul-2010 09:07:49] <|rt|> you'll have to be more specific about the error if you want help
[29-Jul-2010 09:07:54] <_rain_man_> Error running su zenoss -c "/usr/local/zenoss/zenoss/bin/zenoss" start : su: incorrect password
[29-Jul-2010 09:08:01] pebkac
[29-Jul-2010 09:08:17] <_rain_man_> this is the error from the SUSE installation
[29-Jul-2010 09:08:26] sudo su
[29-Jul-2010 09:08:31] try it
[29-Jul-2010 09:08:56] <|rt|> when do you get that error?
[29-Jul-2010 09:09:06] <_rain_man_> right at the end of the installation
[29-Jul-2010 09:09:26] <|rt|> probably when it's trying to start zenoss
[29-Jul-2010 09:09:40] I think you have to "sudo" that command
[29-Jul-2010 09:09:43] <|rt|> try sudo /etc/init.d/zenoss-stack start
[29-Jul-2010 09:09:58] <|rt|> chachan or run it as root
[29-Jul-2010 09:10:10] <_rain_man_> the error is during installation
[29-Jul-2010 09:10:47] <_rain_man_> I think it's strange that it happens on 2 different operating systems :S
[29-Jul-2010 09:10:48] <|rt|> _rain_man_: I would say that installation is done at that point and it's trying to start the service for the first time
[29-Jul-2010 09:11:03] run installation also as root...
[29-Jul-2010 09:11:06] <_rain_man_> ah ok that makes sense
[29-Jul-2010 09:11:14] _rain_man: one confusion point... are typing that command in, or is the installation running that command & then failing?
[29-Jul-2010 09:11:19] <_rain_man_> yes the installation is run as root :)
[29-Jul-2010 09:11:22] are you, rather
[29-Jul-2010 09:11:31] wow, it's busy in here this morning
[29-Jul-2010 09:11:46] <_rain_man_> no I'm not typing it in, it's part of the installation
[29-Jul-2010 09:11:53] :s
[29-Jul-2010 09:12:02] ahh okay, that was not clear to me earlier
[29-Jul-2010 09:12:04] rmatte: its the guessing club, this morning
[29-Jul-2010 09:12:17] ok, try deleting this user before reinstall
[29-Jul-2010 09:12:27] <_rain_man_> ok right lets make this a bit easier
[29-Jul-2010 09:12:27] do you know how delete it?
[29-Jul-2010 09:12:39] <_rain_man_> how the hell did you all manage to install it???
[29-Jul-2010 09:12:42] <|rt|> hey rmatte last night I discovered that I could put html in the summary of an event with a transform...so I tweaked that transform a bit to add a link back to Fortinet's IPS definition database to the page on the attack sig id
[29-Jul-2010 09:13:01] fragfutter: nice
[29-Jul-2010 09:13:39] |rt|: nice
[29-Jul-2010 09:14:05] <_rain_man_> have any of you actually installed zenoss then?
[29-Jul-2010 09:14:06] |rt|: pretty spiffy, sounds like a tip of the month candidate to me... you should make a Wiki entry about that
[29-Jul-2010 09:14:18] _rain_man_: of course, I've been using it for over a year and a half :P
[29-Jul-2010 09:14:24] all of have, successfully, many many times
[29-Jul-2010 09:14:38] in fact those installers you're trying _rain_man get installed a few hundered times by QA before they get released
[29-Jul-2010 09:14:50] <|rt|> about the fact that you can put html in a summary?
[29-Jul-2010 09:14:58] If you're running the install as root there's no reason why su should even need a password
[29-Jul-2010 09:15:08] your system is misconfigured, it's not the installer's fault
[29-Jul-2010 09:15:27] <_rain_man_> lol I've tried different operating systems and get the same error
[29-Jul-2010 09:15:30] either way, that doesn't mean that the install didn't complete properly
[29-Jul-2010 09:15:32] <|rt|> I'm thinking about doign nice event classes for all the snmp traps that the fortigates send and update the zenpack for the summer of zenpack contest
[29-Jul-2010 09:15:37] become root and do /etc/init.d/zenoss-stack start
[29-Jul-2010 09:15:39] and you're done
[29-Jul-2010 09:15:50] <|rt|> I need to figure out how to do some reports and add that to the zenpack as well
[29-Jul-2010 09:15:58] <_rain_man_> ok rmatte 1 sec
[29-Jul-2010 09:16:14] _rain_man_: when you say "different operating systems" what distro did you try it on?
[29-Jul-2010 09:16:22] also, you are actually running the installer as root right?
[29-Jul-2010 09:16:32] you're not trying to run it from a regular user account?
[29-Jul-2010 09:16:34] <|rt|> rmatte: he's tried it on Ubuntu and Suse...but not sure what versions of either
[29-Jul-2010 09:16:42] <_rain_man_> ubuntu 8.04, ubuntu 10.04 and SUSE
[29-Jul-2010 09:16:59] the only time su prompts for a password is when you're not a root user
[29-Jul-2010 09:16:59] <_rain_man_> yes all have been run as root
[29-Jul-2010 09:17:05] very odd
[29-Jul-2010 09:17:32] <_rain_man_> I've tried the gui way on SUSE KDE and the terminal way both as root
[29-Jul-2010 09:18:21] <|rt|> no issues here with Ubuntu 10.04...other than the cron job not running due to some issues with the .bashrc file that zenoss puts in the zenoss home directory
[29-Jul-2010 09:18:21] explain step by step how you're installing it the terminal way
[29-Jul-2010 09:18:41] |rt|: yeh, I've tested the 3.0 installer on both 8.04 and 10.04, worked fine in both cases
[29-Jul-2010 09:18:44] <_rain_man_> the exact way that the document tells you how to do it
[29-Jul-2010 09:18:52] _rain_man_: humour me
[29-Jul-2010 09:19:00] what commands are you issuing?
[29-Jul-2010 09:19:14] <_rain_man_> you want me to copy and paste from the document?
[29-Jul-2010 09:19:31] sure
[29-Jul-2010 09:19:35] <_rain_man_> lol
[29-Jul-2010 09:19:50] _rain_man_, can you pastebin the installation error?
[29-Jul-2010 09:21:42] <_rain_man_> it would be easier if you told me how you got it to install lol
[29-Jul-2010 09:21:49] eugh
[29-Jul-2010 09:21:51] very simple
[29-Jul-2010 09:21:55] sudo su
[29-Jul-2010 09:22:01] wget
[29-Jul-2010 09:22:04] <_rain_man_> Install OS, then follow instructions in zenoss manual ?
[29-Jul-2010 09:22:05] chmod +x
[29-Jul-2010 09:22:08] ./
[29-Jul-2010 09:22:17] all there is to it
[29-Jul-2010 09:22:41] and make sure you're grabbing the appropriate file for your system
[29-Jul-2010 09:22:46] the stack installer should end in .bin
[29-Jul-2010 09:22:56] <_rain_man_> that is correct
[29-Jul-2010 09:22:58] and you should grab 32bit or 64bit version depending on what your server is
[29-Jul-2010 09:24:01] <_rain_man_> I think i'll just use the virtual app
[29-Jul-2010 09:24:08] * rmatte shrugs
[29-Jul-2010 09:24:18] up to you, I don't personally recommend the appliance for production use
[29-Jul-2010 09:24:22] <_rain_man_> until they redo the manual to actually work
[29-Jul-2010 09:24:23] it's fine for just messing around
[29-Jul-2010 09:24:43] _rain_man_: I just gave you the steps
[29-Jul-2010 09:24:44] <_rain_man_> yeah i know, which is why i wanted to install it :(
[29-Jul-2010 09:24:52] The stack installer is braindead simple
[29-Jul-2010 09:24:58] and did you even try what I said?
[29-Jul-2010 09:25:00] sudo su
[29-Jul-2010 09:25:05] then /etc/init.d/zenoss-stack
[29-Jul-2010 09:25:15] <_rain_man_> I know that rmatte, that's exactly what I've done about 10 times now
[29-Jul-2010 09:25:18] <|rt|> rmatte I think it's better to add the repository and let apt-get install the stack
[29-Jul-2010 09:25:19] the error you got just means that the installer was finished and was trying to start up zenoss
[29-Jul-2010 09:25:22] so the install is done
[29-Jul-2010 09:25:26] you just need to start zenoss
[29-Jul-2010 09:25:27] <_rain_man_> but I get an error towards the end of the installation
[29-Jul-2010 09:25:28] as root
[29-Jul-2010 09:25:32] <|rt|> makes apt-get update easy to stay up to date
[29-Jul-2010 09:25:38] with the command /etc/init.d/zenoss-stack
[29-Jul-2010 09:25:52] that error is irrelevant
[29-Jul-2010 09:26:14] like I said, that was just the installer trying to start up Zenoss and failing because for some reason su is prompting for a password on your systems when it shouldn't be
[29-Jul-2010 09:26:18] no idea why that is
[29-Jul-2010 09:26:19] become root
[29-Jul-2010 09:26:24] and do /etc/init.d/zenoss-stack
[29-Jul-2010 09:26:27] and you'll be in business
[29-Jul-2010 09:27:03] <_rain_man_> start zenoss-stack
[29-Jul-2010 09:27:05] <_rain_man_> ?
[29-Jul-2010 09:27:09] yes
[29-Jul-2010 09:27:16] that's the startup script for zenoss
[29-Jul-2010 09:27:19] start that as root
[29-Jul-2010 09:27:25] then zenoss will be running
[29-Jul-2010 09:27:44] <_rain_man_> it says it's already running
[29-Jul-2010 09:27:45] the installer was just complaining that it couldn't use the "su" command to properly start up Zenoss
[29-Jul-2010 09:27:56] rhettardo: su - zenoss
[29-Jul-2010 09:27:58] as root
[29-Jul-2010 09:28:03] then do: zenoss status
[29-Jul-2010 09:28:33] ignore that rhettardo: thing, ever time I type do: my client auto-completes to his nick
[29-Jul-2010 09:28:35] forgot about that
[29-Jul-2010 09:28:57] type: su - zenoss
[29-Jul-2010 09:29:04] <_rain_man_> lol ok
[29-Jul-2010 09:29:05] then: zenoss status
[29-Jul-2010 09:29:25] if it all shows as running then Zenoss is good to go
[29-Jul-2010 09:29:37] If not, just do a zenoss start to start whatever isn't running
[29-Jul-2010 09:29:50] <_rain_man_> ok i'll give it a go :)
[29-Jul-2010 09:30:14] k
[29-Jul-2010 09:30:40] <_rain_man_> ok i've got 2 that aren't running
[29-Jul-2010 09:30:44] <_rain_man_> sorry make that 3
[29-Jul-2010 09:31:39] <_rain_man_> trying to start them now
[29-Jul-2010 09:32:13] k
[29-Jul-2010 09:33:56] <_rain_man_> i tried to restart them but they haven't started :(
[29-Jul-2010 09:34:47] <_rain_man_> zenhub, zenjobs & zenactions are the culprits
[29-Jul-2010 09:35:13] hmmm
[29-Jul-2010 09:35:23] you're still the zenoss user?
[29-Jul-2010 09:35:36] <_rain_man_> yes i am
[29-Jul-2010 09:35:47] type: tail -500 $ZENHOME/log/zenhub.log
[29-Jul-2010 09:36:00] see any errors?
[29-Jul-2010 09:36:18] <_rain_man_> no
[29-Jul-2010 09:36:23] out of curiosity, do you already have other stuff running on that server?
[29-Jul-2010 09:36:31] <_rain_man_> no it's a new installation
[29-Jul-2010 09:38:43] ok
[29-Jul-2010 09:38:48] firewall enabled?
[29-Jul-2010 09:39:16] <_rain_man_> no that is disabled
[29-Jul-2010 09:39:43] selinux?
[29-Jul-2010 09:39:52] hello all
[29-Jul-2010 09:39:56] hola
[29-Jul-2010 09:40:04] <_rain_man_> yes cgibbons
[29-Jul-2010 09:40:17] in v3.0, i go to ipInterface of my router and click on one of the fa interfaces and I could swear last night the graph showed below but now it doesn't even have that option.
[29-Jul-2010 09:40:25] only shows events, details and template as options
[29-Jul-2010 09:41:08] however, when i click on graphs - cpu util and mem is being graphed and up to date
[29-Jul-2010 09:41:36] guys, where to change the eventUrl, or ackUrl that appears on the message strings as variables in Alerting? i was reading about zopeurl in zenactions
[29-Jul-2010 09:41:59] but that file is empty in 3.0
[29-Jul-2010 09:42:22] <_rain_man_> Thank you all for your help
[29-Jul-2010 09:42:37] <_rain_man_> I'm going to try again for the 15th time (Y)
[29-Jul-2010 09:42:40] JonnyV, I want to change it to, but I don't see what to change
[29-Jul-2010 09:42:53] too*
[29-Jul-2010 09:43:17] i cant find it also.. :(
[29-Jul-2010 09:43:21] well, that wasn't very productive
[29-Jul-2010 09:43:28] rmatte: you skilled patience +10
[29-Jul-2010 09:45:51] is there a way to tell Zenoss to only throw out warning for threshold exceeded limits *if* it is sustained for x number of minutes? For example, I dont want to be notified if a server's cpu went to 95% for a few seconds during that one poll. Or better yet, if Zenoss sees that percentage in 3+ polls, then alert
[29-Jul-2010 09:46:15] QubeZ: you want "count = " in your alerting rule
[29-Jul-2010 09:47:05] Simon4: so I'd have to set a filter on the alert for that component? No way to just change it globally on the template?
[29-Jul-2010 09:47:20] QubeZ: no, it's only done at the alerting end
[29-Jul-2010 09:47:24] QubeZ: set it in your global alert rule
[29-Jul-2010 09:47:26] you can't set it per template
[29-Jul-2010 09:47:33] QubeZ: or do you have multiple rules?
[29-Jul-2010 09:48:03] Simon4: got it, thanks guys
[29-Jul-2010 09:48:21] didnt u find something chachan where to change it? you using 3.0 ?
[29-Jul-2010 09:48:23] also another issue im having is that my switches/routers have ipInterface greyed out so no ports are being graphed
[29-Jul-2010 09:48:54] No, I didn't find it. And yes, I'm on 3.0
[29-Jul-2010 09:49:41] yep.. weird..
[29-Jul-2010 09:49:47] conf files are empty
[29-Jul-2010 09:50:38] what conf files?
[29-Jul-2010 09:50:56] in zenoss home etc
[29-Jul-2010 09:52:17] fragfutter: what do you mean multiple rules? i haven't set any rules yet but I'd like to have separate rules for say 1) threshold of ethernet exceeded vs 2) cpu util high
[29-Jul-2010 09:53:10] QubeZ: multiple alert rules. Otherwise you don't get any mails.
[29-Jul-2010 09:53:41] fragfutter: i will be setting them now based on what Simon4 said which is to set alerts with count=2 if i want it to only email me when it hits that high
[29-Jul-2010 09:54:13] QubeZ: and why multiple?
[29-Jul-2010 09:55:17] fragfutter: because im under the assumption that i'll need one alert for cpu util and one for interface util
[29-Jul-2010 09:55:20] not true?
[29-Jul-2010 09:55:46] QubeZ: not true. unless you want different conditions when to send mail
[29-Jul-2010 09:56:44] fragfutter: i guess im not clear on how to construct the alert then because cpu and ethernet util are separate components that i want to alert on
[29-Jul-2010 09:56:50] QubeZ: a rule like event_state=new, production_state=production, severity>=warning, count>2
[29-Jul-2010 09:57:13] QubeZ: and then you get mail for any component that fails (and is in production)
[29-Jul-2010 09:57:21] rmatte, dude!, it works!!!
[29-Jul-2010 09:57:34] I'm very happy :D
[29-Jul-2010 09:57:46] rmatte, thanks a lot!
[29-Jul-2010 09:58:16] fragfutter: oh i see but lets say a warning is set (count=1) for threshold exceed, then it clears... count is set back to 0 right? so if I have something bouncing alot between polling cycles, i'd never know?
[29-Jul-2010 09:58:29] mrayzenoss is now known as mray
[29-Jul-2010 09:58:40] QubeZ: yes
[29-Jul-2010 09:59:06] QubeZ: or you remove the count rule and get a mail everytime...
[29-Jul-2010 09:59:07] thats scary but less likely i'd assume
[29-Jul-2010 09:59:52] JonnyV, I'll try and I'll let you know about it
[29-Jul-2010 10:00:47] great chachan .. i will too.. cheers
[29-Jul-2010 10:00:48] QubeZ: you can also remove the count rule. Instead turn on a delay for your alerting rule. So if something just shows up and goes away it won't mail you. But one-time events that are still there after delay will mail. Or it shows up and the working operator acknowledges it, if he is faster then delay, again no mail, etc.
[29-Jul-2010 10:01:47] fragfutter: i like that better
[29-Jul-2010 10:02:41] thanks man
[29-Jul-2010 10:03:13] also the issue where graphs were not showing up, i remodled the device and now they're back
[29-Jul-2010 10:04:49] nevermind, after i leave that device and come back to it... graphs are unselectable again
[29-Jul-2010 10:05:32] that sounds sub-optimal.
[29-Jul-2010 10:07:49] strange thing, my IpInterface(30) shows grey because I have a few interfaces unmonitored but the ones that are up and monitored are not showing graphs. If I "push changes" or remodel the device, they show up after a page refresh. But if I navigate away from the device then come back, I cannot select graphs from the dropdown.
[29-Jul-2010 10:08:54] The logic that chooses if Graphs should be on the drop-down or not is based on if the component is currently 'monitored'.
[29-Jul-2010 10:09:01] It's probably not quite right.
[29-Jul-2010 10:09:22] cgibbons: it was working just fine yesterday night then this morning I cannot select graphs. Do I need to delete the device and readd?
[29-Jul-2010 10:09:56] If you choose the Details drop-down, is the checkbox indicating the component is monitored or not checked?
[29-Jul-2010 10:10:15] its apparently all my network devices (not servers, just the fw/swt/rtr)... even my PIX doesn't show any interfaces (no components menu on left)
[29-Jul-2010 10:10:33] one sec, going back to that switch
[29-Jul-2010 10:11:48] cgibbons: doesn't show up anything in the details... just shows operational status, duplex mode etc..
[29-Jul-2010 10:11:57] status is up
[29-Jul-2010 10:12:39] confirming QubeZ bug. if i deactivate monitoring on one of the interfaces i can't see graphs, also for interfaces that are still monitored.
[29-Jul-2010 10:13:40] so the bug being that you flip the monitoring switch on just one interface, but then it goes off and hides the graphs for everything else on the device then?
[29-Jul-2010 10:14:12] cgibbons: no better ,)
[29-Jul-2010 10:14:28] cgibbons: it simply does not update the dropdown box if you select a different interface
[29-Jul-2010 10:14:51] Ah, but if you refreshed the page but choose one of the monitored components first, you'd see the drop-down?
[29-Jul-2010 10:14:59] cgibbons: if the first interface that was selected when the pages loaded has no monitoring, then you can't see graphs.
[29-Jul-2010 10:15:19] nasty
[29-Jul-2010 10:15:22] cgibbons: correct
[29-Jul-2010 10:15:37] too much ajax ;)
[29-Jul-2010 10:16:26] cgibbons: shall i open it?
[29-Jul-2010 10:16:38] sure if you don't mind, I'll get the gang to look at it right away
[29-Jul-2010 10:16:55] wish we had noticed that last week. would have been nice fix for 3.0.1
[29-Jul-2010 10:20:27] if I would like to clear a event, what class do I need to send it?
[29-Jul-2010 10:20:34] an event*
[29-Jul-2010 10:22:06] chachan: you mean for it to not send emails/alerts, just silently clear the event?
[29-Jul-2010 10:22:16] cgibbons: http://dev.zenoss.org/trac/ticket/7132
[29-Jul-2010 10:22:46] QubeZ, yes
[29-Jul-2010 10:24:43] QubeZ, I got it: "Clear"
[29-Jul-2010 10:24:58] k
[29-Jul-2010 10:26:41] chachan: you did "map event to class" and selected clear?
[29-Jul-2010 10:27:24] chachan: i ask becauase i've been wondering how to do this too. I need to clear out a bunch of old events and I dont want notifications to go out.
[29-Jul-2010 10:29:35] QubeZ, well, I'm creating these event through python script, so, what about clear it?, that's what I meaning, are we talking about the same?
[29-Jul-2010 10:31:21] chachan: oh, i thought you mean from the event console. I'm looking for a way in the web ui to remove events without triggering emails.
[29-Jul-2010 10:31:59] turn the rule off, temoprarily? :)
[29-Jul-2010 10:32:49] twm1010: ya thats what i've been doing, was hoping for a better way :)
[29-Jul-2010 10:39:08] rmatte, this dmd object has any api or doc?
[29-Jul-2010 10:39:47] anybody used this zenpack? docs/DOC-5817
[29-Jul-2010 10:39:51] its for monitoring amqp
[29-Jul-2010 10:40:13] but it doesnt work, I get the same error as the poster on the page: "Unable to connect to amqp queue celery on : 'generator' object has no attribute 'send'"
[29-Jul-2010 10:40:20] and unfortunately, the author doesnt respond
[29-Jul-2010 10:50:41] <|rt|> can you import modules and define functions in the script for a transform?
[29-Jul-2010 10:51:20] hm?
[29-Jul-2010 10:51:29] I just use "Install ZenPack" in the gui
[29-Jul-2010 10:51:40] dv_: rt didn't ask you ;)
[29-Jul-2010 10:51:44] oh :]
[29-Jul-2010 10:52:03] also if anyone can confirm this issue when there are old events in the event console and you disable/re-enable alerts, you get a host of emails again for all the listings in the event console.
[29-Jul-2010 10:52:04] |rt|: afaik at least imports.
[29-Jul-2010 10:52:47] <|rt|> ok I'll give it a try....I want to use the socket module to rev dns lookup an ip address for the summary
[29-Jul-2010 10:55:50] i even disabled alerts for a few minutes and after re-enabling alerts, i get a host of emails for old stuff
[29-Jul-2010 10:57:18] <|rt|> fragfutter: well teh transform box didn't go red so I guess it's good
[29-Jul-2010 10:57:42] <|rt|> pretty soon here I'll have a transform that's a couple hundres lines of python heh
[29-Jul-2010 10:58:53] <|rt|> I could use a geoIP service to provide a link to show you where the attack came from on google maps :)
[29-Jul-2010 11:00:01] <|rt|> hmm if I did that I wonder if I could do a report that would give you a link using google's api to map out all the attacks between two dates
[29-Jul-2010 11:24:53] cgibbons: any chances of http://dev.zenoss.org/trac/ticket/7061 making it into 3.0.2?
[29-Jul-2010 11:25:04] (global search box)
[29-Jul-2010 11:26:03] heading to a meeting, but here's something new for 3.0.1
[29-Jul-2010 11:26:03] docs/DOC-7453
[29-Jul-2010 11:26:55] mray: great!
[29-Jul-2010 11:27:57] mray: nice
[29-Jul-2010 11:28:03] that'll be really handy
[29-Jul-2010 11:28:27] major annoyance reliefed
[29-Jul-2010 11:29:58] i can't download it with svn?
[29-Jul-2010 11:35:21] I could really use some assistance with deleting this property that is half there
[29-Jul-2010 11:36:11] BadRequest: Id zOTRSSOAPLogin is invalid or duplicate
[29-Jul-2010 11:36:59] I wonder if the #zope channel would help
[29-Jul-2010 11:37:13] well hold up, do you know what level it's complaining at?
[29-Jul-2010 11:37:15] device level?
[29-Jul-2010 11:37:34] yes
[29-Jul-2010 11:37:41] a=find("devicename")
[29-Jul-2010 11:37:53] a.__delattr__("zOTRSSOAPLogin")
[29-Jul-2010 11:37:56] commit()
[29-Jul-2010 11:37:58] in zendmd
[29-Jul-2010 11:38:26] I meant the device organizer level
[29-Jul-2010 11:38:42] hang on let me pastabin this
[29-Jul-2010 11:38:44] er pastebin
[29-Jul-2010 11:39:01] http://pastebin.com/hCUvfgRB
[29-Jul-2010 11:39:39] oh
[29-Jul-2010 11:39:42] then do...
[29-Jul-2010 11:39:52] dmd.Devices.__delattr__("zOTRSSOAPLogin")
[29-Jul-2010 11:40:46] I think that did it!!!
[29-Jul-2010 11:41:06] cool
[29-Jul-2010 11:41:28] that usually cleans up BadRequest issues
[29-Jul-2010 11:41:30] yes, it's fixed
[29-Jul-2010 11:41:34] good stuff
[29-Jul-2010 11:41:40] thank you!
[29-Jul-2010 11:41:47] np
[29-Jul-2010 11:41:51] how do you know all this?
[29-Jul-2010 11:41:52] lol
[29-Jul-2010 11:41:56] experience
[29-Jul-2010 11:42:15] python ;)
[29-Jul-2010 11:42:23] There aren't many issues with Zenoss that I don't know how to fix at this point
[29-Jul-2010 11:42:34] now I can get back to coding this thing
[29-Jul-2010 11:43:08] It'll be a while before I can send you the updated zenticket pack... there's more stuff I want to add in, but when I send it to you you can just diff the files to see what I've added
[29-Jul-2010 11:43:26] It won't be complex to implement in to your version
[29-Jul-2010 11:44:08] I'm working the rest of the day on that, so hopefully I'll have the end product finished and tested by tomorrow
[29-Jul-2010 11:44:15] ok
[29-Jul-2010 11:44:38] mray: hero of the day!
[29-Jul-2010 11:44:58] who is? Matt?
[29-Jul-2010 11:45:10] mray is.
[29-Jul-2010 11:45:13] ah
[29-Jul-2010 11:45:15] what did he do?
[29-Jul-2010 11:45:20] simple search
[29-Jul-2010 11:45:23] oh right
[29-Jul-2010 11:45:28] yeh, that does look nice
[29-Jul-2010 11:45:41] link?
[29-Jul-2010 11:45:54] Simon4: two pages up ;)
[29-Jul-2010 11:46:04] Simon4: docs/DOC-7453
[29-Jul-2010 11:46:12] ah, I'm blind :)
[29-Jul-2010 11:46:19] hehe
[29-Jul-2010 11:46:22] i'm so much faster typing the names of devices then clicking through interfaces
[29-Jul-2010 11:46:31] oh good, doesn't work in enterprise ;)
[29-Jul-2010 11:46:53] * Simon4 will find a large stick and beat a product manager with it :)
[29-Jul-2010 11:47:02] haha
[29-Jul-2010 11:47:22] I don't see why it wouldn't work in enterprise
[29-Jul-2010 11:47:26] * rmatte shrugs
[29-Jul-2010 11:47:36] not tested, not approved
[29-Jul-2010 11:47:43] ^^^^^
[29-Jul-2010 11:47:55] my guess is it'll come in enterprise 3.0.1 anyway
[29-Jul-2010 11:48:21] i have an issue http://fpaste.org/9IfA/
[29-Jul-2010 11:48:41] if i view a monitoring template
[29-Jul-2010 11:49:17] sounds like it's trying to get the severity level of a threshold, and it's not being provided
[29-Jul-2010 11:49:27] where is that template from
[29-Jul-2010 11:49:29] and what is it?
[29-Jul-2010 11:49:33] rmatte: genius ;) now how to fix it.
[29-Jul-2010 11:49:55] bigegors compaq/hp monitoring stuff
[29-Jul-2010 11:50:18] 2.5 or 3.0?
[29-Jul-2010 11:50:21] 3
[29-Jul-2010 11:50:33] hmmm, I wonder if that pack has even been tested in 3 yet
[29-Jul-2010 11:50:57] so was the pack installer prior to an upgrade?
[29-Jul-2010 11:51:02] yes
[29-Jul-2010 11:51:13] try reinstalling the pack
[29-Jul-2010 11:51:28] I find that some packs stop working after the upgrade and the only way to get them working again is to install them again
[29-Jul-2010 11:51:36] that's what I had to do with Egor's WMI packs
[29-Jul-2010 11:52:27] you'll have to remove the "-py2.4" from the ZenPack filename for it to install in 3.0
[29-Jul-2010 11:53:05] actually there is a newer version, so upgrade....
[29-Jul-2010 11:53:35] yup
[29-Jul-2010 11:54:10] nice talking, could have checked that myself...
[29-Jul-2010 11:54:52] lol
[29-Jul-2010 11:56:31] anything wrong with "zenpack --remove ZenPacks.community.HPMon"? hangs sind four minutes, does nothing
[29-Jul-2010 11:58:34] oh boy, corrupted zodb.
[29-Jul-2010 11:58:56] reindex
[29-Jul-2010 11:58:58] lol
[29-Jul-2010 11:59:18] i know, playing with zope since years.
[29-Jul-2010 11:59:34] :)
[29-Jul-2010 11:59:47] the reason i learned python, zope.
[29-Jul-2010 12:00:13] and later i discovered that programming python outside of zope was actually fun.
[29-Jul-2010 12:02:14] ! damn nick auto complete
[29-Jul-2010 12:03:42] hmm.. anyone in here a pro with event transforms and class mapping?
[29-Jul-2010 12:05:28] bigegor: two days ago i complained about not being able to view monitoring templates for hpmon zenpack. upgrading resolved it.
[29-Jul-2010 12:07:04] crm_mon sends all it's traps under a single oid, so the eventClassKey is the same, so I'm having trouble mapping clear traps to clear error traps :/
[29-Jul-2010 12:07:07] Simon4: I consider myself pretty good at tuning
[29-Jul-2010 12:07:13] including transforms and mappings
[29-Jul-2010 12:07:22] Simon4: ask your question
[29-Jul-2010 12:08:56] rmatte: so crm_mon sends traps when bad and good things happen
[29-Jul-2010 12:09:07] every single trap comes in under the same oid
[29-Jul-2010 12:09:17] ok
[29-Jul-2010 12:09:23] I map one of the traps to my event class, they all appear in the event class
[29-Jul-2010 12:09:25] so far so good
[29-Jul-2010 12:09:40] ok...
[29-Jul-2010 12:09:56] however some traps are "monitor xxx fails" as the payload, some are "monitor xxx succeeds" as the payload, etc
[29-Jul-2010 12:10:04] ok
[29-Jul-2010 12:10:05] I want the monitor xxx succeeeds to clear the fails events
[29-Jul-2010 12:10:17] normally to do that I have two class mappings, one with the zproperty set to clear
[29-Jul-2010 12:10:31] I have something very similar configured to what you're asking...
[29-Jul-2010 12:10:33] hold on...
[29-Jul-2010 12:11:19] http://fpaste.org/yLJj/
[29-Jul-2010 12:11:38] that kicks in when an UP event comes in, and checks the queue for the corresponding down event, and moves it to history
[29-Jul-2010 12:11:39] ahhh, cunning!
[29-Jul-2010 12:12:05] cheers, I'll have a play, but that looks like it'll do nicely
[29-Jul-2010 12:12:13] * rmatte nods
[29-Jul-2010 12:14:05] fragfutter: cool
[29-Jul-2010 12:21:02] hmmm, nice, I'd never seen this page before: docs/DOC-2445
[29-Jul-2010 12:21:07] answers a lot of common questions
[29-Jul-2010 12:33:11] lets see
[29-Jul-2010 12:33:49] yeahh.. already seen it.. its nice
[29-Jul-2010 12:46:30] I was off in an All Hands meeting, sorry I forgot to attach the .EGG to the Simple Search ZenPack page
[29-Jul-2010 12:46:34] it's added now
[29-Jul-2010 12:46:59] Enterprise has an Advanced Search feature, the version in Core conflicts with it
[29-Jul-2010 12:47:30] alright so one of the other guys was messing with building a ZenPack and I think he just removed the directory…so now half the daemons wont start and we get an error like this ERROR: zenpack command failed. Reason: DistributionNotFound: ZenPacks.michaelv.MySQLStat
[29-Jul-2010 12:47:43] is there a way to make zenoss requery what zenpacks it has installed
[29-Jul-2010 12:47:52] or make it think this zenpack is removed ?
[29-Jul-2010 12:48:21] st3v3o: you can re-install that Zenpack if you have it… or use "zenpack —remove=ZenPacks.michaelv.MySQLStat"
[29-Jul-2010 12:48:37] if the files are gone, it might not work
[29-Jul-2010 12:48:51] yeah files are gone …he rm -rf them :/
[29-Jul-2010 12:49:03] st3v3o: worst case. cd $ZENHOME/ZenPacks/ ; edit easy-install.pth and remove the zenpack
[29-Jul-2010 12:49:10] eugh, going back and commenting code after it's been written is so boring
[29-Jul-2010 12:49:16] rocket: that works?
[29-Jul-2010 12:49:16] st3v3o: that will not remove the items from the object database
[29-Jul-2010 12:49:22] mray: yes
[29-Jul-2010 12:49:27] ok cool thanks mray
[29-Jul-2010 12:49:38] I would recommending then adding the zenpack in and removing it normally
[29-Jul-2010 12:49:46] I run into that sometimes when I use shared drives that get unmounted
[29-Jul-2010 12:49:48] * rmatte rubs his eyes
[29-Jul-2010 12:49:58] * rmatte points and says "Look! It's a rocket!"
[29-Jul-2010 12:49:59] mray: that keeps the code for the zenpack from loading ..
[29-Jul-2010 12:50:30] rmatte: heh :)
[29-Jul-2010 12:50:40] * rmatte switches to his crocodile hunter voice: "They are very rare to see in the wild like this!"
[29-Jul-2010 12:50:54] rmatte: do much with snmp process monitoring?
[29-Jul-2010 12:51:10] rocket: I do tons of snmp process monitoring
[29-Jul-2010 12:51:18] http://dev.zenoss.org/trac/ticket/7134 :)
[29-Jul-2010 12:51:27] guys why if i click to view the conf file for zenactions daemon ( i suppose it happens to all rest of daemons) it shows empty; and if i click the edit config it shows all the variables and values?
[29-Jul-2010 12:51:50] rmatte: a little patch I created to make events more useful ..
[29-Jul-2010 12:52:01] rocket: nice, handy
[29-Jul-2010 12:52:33] hrm
[29-Jul-2010 12:52:45] what class in zenoss should multi-layer switches go into in zenoss
[29-Jul-2010 12:52:50] cisco switches
[29-Jul-2010 12:52:54] rmatte: notice its noting the pid changes etc .. and if its totally down the last model time in the event ..
[29-Jul-2010 12:52:56] OK, gotta go back to my meeting
[29-Jul-2010 12:52:57] because they are somewhat between switches and routers
[29-Jul-2010 12:54:24] So, how come my ZenPack import of .egg files fails? (Zenoss 3.0)
[29-Jul-2010 12:54:26] what would you guys suggest
[29-Jul-2010 12:54:48] because there is /Devices/Network/Routers/Cisco or /Devices/Network/Switches
[29-Jul-2010 12:54:50] Orv the zenpack may not be compatible with 3.0
[29-Jul-2010 12:55:00] rocket: I thought it was; I'll recheck.
[29-Jul-2010 12:55:40] rocket: Yeah, it says it is...
[29-Jul-2010 12:56:02] It's the WMI Data Source ZenPack.
[29-Jul-2010 12:56:17] traceback clues? tried from command line?
[29-Jul-2010 12:56:29] other dependant zenpack?
[29-Jul-2010 12:56:43] good point; I'll try from command line.
[29-Jul-2010 12:56:44] try a restart zenoss
[29-Jul-2010 12:56:58] may be a browser bug too ..
[29-Jul-2010 12:57:08] could always try a different browser
[29-Jul-2010 12:58:47] well, from the command line, here's the first error (which probably leads to all the other errors):
[29-Jul-2010 12:58:52] (apologies for the paste)
[29-Jul-2010 12:59:10] zenpack --install ZenPacks.community.WMIPerf_Windows-2.2.egg
[29-Jul-2010 12:59:11] Link to http://pypi.python.org/simple/ZenPacks.community.WMIDataSource/ ***BLOCKED*** by --allow-hosts
[29-Jul-2010 13:00:52] * Orv checks dependencies...
[29-Jul-2010 13:00:52] I am guessing you have a dependancy issue
[29-Jul-2010 13:01:11] you need to install WMI Data Source pack beforehand
[29-Jul-2010 13:01:14] or it won't install
[29-Jul-2010 13:01:45] rmatte: heh I havent done much with community zenpacks lately ...
[29-Jul-2010 13:01:57] :)
[29-Jul-2010 13:02:07] you have a few broken ones that need fixin' apparently
[29-Jul-2010 13:02:10] mostly core fixes etc .. ;p
[29-Jul-2010 13:02:21] I hear that ..
[29-Jul-2010 13:02:33] not sure what effort that will take ..
[29-Jul-2010 13:02:45] no idea
[29-Jul-2010 13:02:53] you'd have to ask Egor
[29-Jul-2010 13:03:00] since he's done work to fix his packs
[29-Jul-2010 13:03:06] not much free time at the moment ... currently rewriting some core daemons etc .. for fun .. :p
[29-Jul-2010 13:03:12] ah
[29-Jul-2010 13:03:24] rocket: Tried installing another Zenpack; it worked fine. I'll chew on this a bit more - thanks.
[29-Jul-2010 13:03:41] but have a little one due in less than 10 weeks so there goes my time .. :p
[29-Jul-2010 13:06:25] Orv: install the WMI Data Source pack
[29-Jul-2010 13:06:33] it's a dependency of the WMIPerf_Windows pack
[29-Jul-2010 13:17:44] ok.. how do I get it to not show these lines between locations
[29-Jul-2010 13:18:35] on the google map
[29-Jul-2010 13:19:17] you're using 3.0 yes?
[29-Jul-2010 13:19:23] 2.5.2
[29-Jul-2010 13:19:33] click Networks on the left
[29-Jul-2010 13:19:46] zProperties
[29-Jul-2010 13:20:01] zDrawMapLinks
[29-Jul-2010 13:20:07] rmatte: That worked, thanks.
[29-Jul-2010 13:20:12] set it to False
[29-Jul-2010 13:20:15] Orv: no problem
[29-Jul-2010 13:20:15] ahh!
[29-Jul-2010 13:20:22] I can just turn that off for that network that is causing the problem
[29-Jul-2010 13:20:34] mducharme-laptop: yessir
[29-Jul-2010 13:20:49] ok, that's better
[29-Jul-2010 13:20:59] see we have some devices whose interfaces default to private IPs
[29-Jul-2010 13:21:11] zenoss is incorrectly thinking those are the same network
[29-Jul-2010 13:21:16] ah
[29-Jul-2010 13:21:41] by the way, I'm finding myself changing quite a bit of stuff in the daemon... and I'm commenting everything
[29-Jul-2010 13:21:56] I just changed the may that it matched the groups since it was flawed
[29-Jul-2010 13:22:02] using == instead of re.match
[29-Jul-2010 13:22:25] I was pretty new to python when I wrote this, I assumed re.match would only be true on an exact match, which I now know to be incorrect
[29-Jul-2010 13:24:24] ok
[29-Jul-2010 13:26:48] it'll eliminate some overhead too
[29-Jul-2010 13:31:16] rmatte: Well, that didn't work so well: " broken This pack was just installed, or is missing or broken. Please restart Zope. " (even after restarting Zenoss)
[29-Jul-2010 13:33:16] Orv: odd, you're sure zenoss restarted properly
[29-Jul-2010 13:33:23] don't use "zenoss restart"
[29-Jul-2010 13:33:26] use zenoss stop
[29-Jul-2010 13:33:31] rmatte, how did you get to that dmd method?, I wish to know more about it
[29-Jul-2010 13:33:34] then zenoss status to make sure everything actually did stop
[29-Jul-2010 13:33:40] once everything is stopped, then do a zenoss start
[29-Jul-2010 13:33:42] rmatte: service zenoss restart won't work?
[29-Jul-2010 13:33:57] chachan: what do you mean how did I get to it?
[29-Jul-2010 13:33:59] I'll give it a shot..
[29-Jul-2010 13:34:21] Orv: service zenoss restart is equivalent to zenoss restart, it's not very reliable I find
[29-Jul-2010 13:34:29] Orv: obviously do what I was describing as the Zenoss user
[29-Jul-2010 13:34:45] rmatte, how you know about this method, I want to know more about dmd
[29-Jul-2010 13:35:04] rmatte, but I don't see much information about it in the guide
[29-Jul-2010 13:35:25] chachan: I just use the tab completion in zendmd to find the functions I need
[29-Jul-2010 13:35:41] I actually discovered that particular function in a script that one of the zenoss devs wrote
[29-Jul-2010 13:35:53] rmatte: Nope, that didn't fix it.
[29-Jul-2010 13:36:02] I assume there's a command line 'uninstall'..
[29-Jul-2010 13:36:10] mmm
[29-Jul-2010 13:36:13] Orv: well hold up, you're using Zenoss 3.0?
[29-Jul-2010 13:36:20] really really really strange problem here
[29-Jul-2010 13:36:26] netmask = 255.255.255.128
[29-Jul-2010 13:36:53] rmatte, I'm search (grep -r ) in $ZENHOME to see what I can get. I'll you know
[29-Jul-2010 13:37:02] i'll let you know*
[29-Jul-2010 13:38:28] str(self.maskToBits(om.netmask.strip())) is 24
[29-Jul-2010 13:38:32] rmatte: Yes, Zenoss 3.0
[29-Jul-2010 13:38:40] chachan: well, which function in particular are you wondering about?
[29-Jul-2010 13:38:40] but er is 25
[29-Jul-2010 13:38:44] zenpack --list shows them all as installed..
[29-Jul-2010 13:38:53] but for whatever reason it's having the ip as a /24
[29-Jul-2010 13:38:58] Has Zenoss 3.0 gotten a fixed winexe yet (per in this thread: message/48714#48714)
[29-Jul-2010 13:39:18] mducharme-laptop: pretty sure there's been a trac ticket open for that issue for ages
[29-Jul-2010 13:39:19] or is there any issue with manually replacing the Zenoss RPM provided winexe with one that's fixed?
[29-Jul-2010 13:39:40] rmatte, I want to know if I can send some event ID or something like that to make an event "Clear"
[29-Jul-2010 13:39:50] Orv: was your WMI Data Source pack installed prior to an upgrade?
[29-Jul-2010 13:39:58] Orv: or was it a clean install of 3.0?
[29-Jul-2010 13:40:06] rmatte, with you help, I already can create event, but I don't know how to clear it
[29-Jul-2010 13:40:17] Orv: also, what version of the WMI Data Source pack did you install?
[29-Jul-2010 13:40:38] chachan: that blurb I pasted you shows how to clear it
[29-Jul-2010 13:41:12] rmatte: It was a clean install of 3.0 on Centos 5.5
[29-Jul-2010 13:41:28] chachan: the line that actually moves the event to history is...
[29-Jul-2010 13:41:29] dmd.ZenEventManager.manage_deleteEvents(cevt.evid)
[29-Jul-2010 13:41:31] It shows version 2.7 of WMI Data Source
[29-Jul-2010 13:41:51] Orv: strange, I have version 2.7 of the WMI Data Source pack working perfectly in 3.0
[29-Jul-2010 13:41:58] Orv: try this...
[29-Jul-2010 13:42:00] rmatte, that's what I need :D
[29-Jul-2010 13:42:01] as the zenoss user...
[29-Jul-2010 13:42:03] zendmd
[29-Jul-2010 13:42:06] reindex()
[29-Jul-2010 13:42:08] commit()
[29-Jul-2010 13:42:12] ctrl-d to exit
[29-Jul-2010 13:42:15] zenoss stop
[29-Jul-2010 13:42:18] zenoss status
[29-Jul-2010 13:42:22] make sure everything is stopped
[29-Jul-2010 13:42:24] zenoss start
[29-Jul-2010 13:42:27] that might fix it
[29-Jul-2010 13:42:43] Umm... where do I find dmd?
[29-Jul-2010 13:42:51] just become the zenoss user
[29-Jul-2010 13:42:53] oh
[29-Jul-2010 13:42:54] and type zendmd
[29-Jul-2010 13:43:34] working on it..
[29-Jul-2010 13:44:36] WARNING OFS.Uninstalled Could not import class 'ZenPack' from module 'ZenPacks.community.WMIPerf_Windows'
[29-Jul-2010 13:45:22] where are you seeing that?
[29-Jul-2010 13:45:42] while doing the reindex()
[29-Jul-2010 13:45:52] does it actually kill the reindex?
[29-Jul-2010 13:46:01] rmatte: No. It completed with no other warnings.
[29-Jul-2010 13:46:05] restarting zenoss now
[29-Jul-2010 13:46:11] did you commit()?
[29-Jul-2010 13:46:46] Yep
[29-Jul-2010 13:46:49] k
[29-Jul-2010 13:47:03] I'd say remove the ZenPacks.community.WMIPerf_Windows ZenPack if you can
[29-Jul-2010 13:47:06] then install it again
[29-Jul-2010 13:49:08] Hrm: broken This pack was just installed, or is missing or broken. Please restart Zope.
[29-Jul-2010 13:49:12] oops wrong paste
[29-Jul-2010 13:49:23] ZenPackException: ZenPack ZenPacks.community.WMIDataSource-2.7.egg is not installed.
[29-Jul-2010 13:49:48] rmatte: I have to go run and do some things; I'll chase this later. Thanks for your help!
[29-Jul-2010 13:50:04] np
[29-Jul-2010 13:50:05] later
[29-Jul-2010 13:50:41] rmatte: will ZenOSS do paging on route changes, per chance?
[29-Jul-2010 13:51:16] straterra: not out of the box... it only picks up on routes when devices are remodeled
[29-Jul-2010 13:51:24] Oh
[29-Jul-2010 13:51:34] straterra: you talking about BGP, EIGRP, OSPF, stuff like that?
[29-Jul-2010 13:51:38] OSPF, yes
[29-Jul-2010 13:51:48] Each plant has two internet connections and we use OSPF for fail over
[29-Jul-2010 13:51:54] the proper way to alert on those types of routing protocols is to have the device trap back to Zenoss
[29-Jul-2010 13:51:59] that's how we do it
[29-Jul-2010 13:52:11] You can have quagga do that?
[29-Jul-2010 13:52:26] not sure about quagga, I deal with Cisco gear myself
[29-Jul-2010 13:52:41] Hmm..
[29-Jul-2010 13:53:05] Quagga is linux based?
[29-Jul-2010 13:53:08] quagga uses a lot of the same features, so maybe
[29-Jul-2010 13:53:24] Linux and unix/unix based, yes
[29-Jul-2010 13:54:08] hmmm, if it can't trap back, you could generate an event in Zenoss via zensendevent, but unfortunately it appears to be broken
[29-Jul-2010 13:54:10] so that's no good
[29-Jul-2010 13:54:24] actually, that's what I wanted to talk to Matt about, I keep forgetting
[29-Jul-2010 13:54:28] I know a work around
[29-Jul-2010 13:54:42] I know how to do it via zendmd if needed
[29-Jul-2010 13:54:47] Write a nagios plugin real quick that runs on the router..and returns an error status if the primary route is the second interface
[29-Jul-2010 13:54:49] that's just about the only workaround that I know
[29-Jul-2010 13:55:00] yeh, you could do that as well
[29-Jul-2010 13:55:31] using an expect script or something
[29-Jul-2010 14:01:36] mray …I looked in the $ZENHOME\ZenPacks\easy-install.pth file and I don't see the missing zenpack listed that is causing me issues.
[29-Jul-2010 14:02:14] Is there another location that might be listing this missing zenpack…I can't reinstall because it was a zenpack that was made from the gui and was deleted
[29-Jul-2010 14:06:40] whoa, the severity row colors in the event console need to be lightened a bit :)
[29-Jul-2010 14:08:01] JonnyV, check this out message/45261
[29-Jul-2010 14:08:26] JonnyV, you could create a cron job
[29-Jul-2010 14:10:24] woohoo, I'm finally done commenting this script
[29-Jul-2010 14:12:28] nice
[29-Jul-2010 14:14:06] FYI, it's now called "Global Device Search"
[29-Jul-2010 14:14:13] docs/DOC-7453
[29-Jul-2010 14:15:18] chachan, the zopeurl appears in zenactions; edit from GUI...
[29-Jul-2010 14:15:47] from shell, file is empty. and clicking view config also shows it empty
[29-Jul-2010 14:16:04] click on edit and at the bottom you can change the http address
[29-Jul-2010 14:19:15] alright now even running zenpack —list give me a funky error
[29-Jul-2010 14:19:15] http://pastebin.com/BQGBAs8h
[29-Jul-2010 14:26:12] looks like something is screwed with that ZenPack
[29-Jul-2010 14:28:41] yeah ..it got deleted via rm -rf
[29-Jul-2010 14:28:49] with out doing a zenpack —remove
[29-Jul-2010 14:28:59] oh
[29-Jul-2010 14:29:08] well reinstall it
[29-Jul-2010 14:29:08] trying to "fix" it
[29-Jul-2010 14:29:14] its gone gone.
[29-Jul-2010 14:29:24] like you don't even have an egg file for it?
[29-Jul-2010 14:29:31] it was a zenpack a guy created from the web interface and the deleted
[29-Jul-2010 14:29:41] I see...
[29-Jul-2010 14:29:47] nope he deleted ALL the file… *face palm*
[29-Jul-2010 14:29:53] I know how to get rid of it
[29-Jul-2010 14:30:01] :)
[29-Jul-2010 14:30:07] go in to zendmd
[29-Jul-2010 14:30:45] alright heading there now
[29-Jul-2010 14:30:52] dmd.ZenPackManager.packs._delObject('ZenPacks.MichaelVitale.MySQLStats')
[29-Jul-2010 14:30:55] commit()
[29-Jul-2010 14:31:11] that should do it
[29-Jul-2010 14:31:11] 2010-07-29 15:31:01 WARNING zen.ToManyRelationshipBase Tried to delete object id 'ZenPacks.MichaelVitale.MySQLStats' but didn't find it on /zport/dmd/ZenPackManager/packs
[29-Jul-2010 14:31:11] >>> 3:30
[29-Jul-2010 14:31:11] File "", line 1
[29-Jul-2010 14:31:11] 3:30
[29-Jul-2010 14:31:11] ^
[29-Jul-2010 14:31:12] SyntaxError: invalid syntax
[29-Jul-2010 14:31:49] mhmmm
[29-Jul-2010 14:31:57] yeah I can't figure out where its stuck at
[29-Jul-2010 14:32:02] do you have a second instance of Zenoss running somewhere?
[29-Jul-2010 14:32:39] (by the way, tell whoever did that that they are a moron
[29-Jul-2010 14:32:40] lol
[29-Jul-2010 14:32:46] oh I did :P
[29-Jul-2010 14:33:00] I think I have one ….what are you thinking
[29-Jul-2010 14:33:01] you don't just go rm -rfing stuff when you don't know what you're doing
[29-Jul-2010 14:33:12] I'm thinking create a pack from the UI on the other instance
[29-Jul-2010 14:33:13] FNG
[29-Jul-2010 14:33:15] with the same name
[29-Jul-2010 14:33:18] export it
[29-Jul-2010 14:33:20] remove it
[29-Jul-2010 14:33:25] rmatte, is it possible to disable zenping or change it for a custom script?
[29-Jul-2010 14:33:27] then install the exported pack on that instance
[29-Jul-2010 14:33:32] that should restore it to how it should be
[29-Jul-2010 14:33:35] then remove it properly
[29-Jul-2010 14:33:49] you don't think it needs the same object…just same name >?
[29-Jul-2010 14:33:51] chachan: you want to disable it on a specific device?
[29-Jul-2010 14:33:58] st3v3o: same name, correct
[29-Jul-2010 14:34:04] cool let me try that
[29-Jul-2010 14:34:04] JonnyV, ok, that's great
[29-Jul-2010 14:34:07] k
[29-Jul-2010 14:34:32] rmatte, No, I want disable or change it for all
[29-Jul-2010 14:34:41] that's simple
[29-Jul-2010 14:34:47] :D
[29-Jul-2010 14:34:49] navigate to the /Devices class
[29-Jul-2010 14:34:53] then zProperties for that class
[29-Jul-2010 14:34:59] change zPingMonitorIgnore to true
[29-Jul-2010 14:35:00] done
[29-Jul-2010 14:35:24] that will make zenping ignore everything
[29-Jul-2010 14:35:34] you can do it for a specific class as well if you want
[29-Jul-2010 14:35:43] or specific devices
[29-Jul-2010 14:36:00] yep
[29-Jul-2010 14:37:14] can you still install older 2.4 zenpacks if they're not yet available for 3.0?
[29-Jul-2010 14:37:21] want to test out this SSL expiration zenpack
[29-Jul-2010 14:39:40] rmatte: If you come up with a proper solution for the route thing, lemme know
[29-Jul-2010 14:40:53] anyone have any issues posting to the forum today via e-mail?
[29-Jul-2010 14:41:45] QubeZ: you can, you need to remove the -py2.4 from the filename
[29-Jul-2010 14:41:56] QubeZ: they may or may not work depending on what's in them
[29-Jul-2010 14:42:31] jmp242: I hope so, I was getting sick of seeing endless out of office messages as posts
[29-Jul-2010 14:42:41] If they disabled that I'm personally happy
[29-Jul-2010 14:42:51] it takes 2 extra seconds to actually go to the site and login to post a reply
[29-Jul-2010 14:49:30] rmatte, done
[29-Jul-2010 14:50:07] did it work?
[29-Jul-2010 14:50:32] actually sorry, I mixed up your issue with st3v3o's
[29-Jul-2010 14:50:36] I'm sure it worked for you
[29-Jul-2010 14:50:36] lol
[29-Jul-2010 14:50:59] having issues installing the zenapck
[29-Jul-2010 14:51:15] what's the error?
[29-Jul-2010 14:52:07] rmatte, yep
[29-Jul-2010 14:52:20] hjehehe ok
[29-Jul-2010 14:52:24] Now its saying zenpack exists…but I can't do a zenapck —remove ZenPacks.MichaelVitale.MySQLStats
[29-Jul-2010 14:52:54] fpaste the output when you try to install it
[29-Jul-2010 14:53:18] if I restore from backup what happens…will I loose my rrd data ect.
[29-Jul-2010 14:54:42] st3v3o: not if you back it up, but I really recommend not doing that, I'm positive we can fix this
[29-Jul-2010 14:54:54] paste me the output when you try to install the pack please
[29-Jul-2010 14:54:58] I like the sound of that
[29-Jul-2010 14:55:15] sure I'll PM you
[29-Jul-2010 14:55:19] no
[29-Jul-2010 14:55:21] fpaste.org
[29-Jul-2010 14:55:27] pastebin ok
[29-Jul-2010 14:55:33] yes
[29-Jul-2010 14:55:39] JonnyV, done
[29-Jul-2010 14:55:43] I prefer fpaste, but whatever you prefer
[29-Jul-2010 14:56:00] cool.. :)
[29-Jul-2010 14:56:02] http://fpaste.org/AHEv/
[29-Jul-2010 14:56:14] when I look in $ZENHOME\Zenpacks…now I see the following
[29-Jul-2010 14:56:40] [zenoss@zen ZenPacks]$ ls | grep Vitale
[29-Jul-2010 14:56:40] ZenPacks.MichaelVitale.MySQLStats-1.0.0.egg
[29-Jul-2010 14:56:40] ZenPacks.MichaelVitale.MySQLStats-1.0-py2.4.egg
[29-Jul-2010 14:56:40] ZenPacks.MichaelVitale.MySQLStats-3.0.0.egg
[29-Jul-2010 14:56:40] ZenPacks.MichaelVitale.MySQLStats.egg
[29-Jul-2010 14:56:43] JonnyV, was useful fro you what I sent you?
[29-Jul-2010 14:56:45] there is a bunch of them
[29-Jul-2010 14:56:50] see an easy-install.pth file?
[29-Jul-2010 14:57:04] for*
[29-Jul-2010 14:57:31] do you know what directory he actually did the rm -rf on?
[29-Jul-2010 14:57:42] No.
[29-Jul-2010 14:57:47] yea... nice
[29-Jul-2010 14:57:50] actually, listen to rocket, he probably has some tricks up my sleeve that I don't
[29-Jul-2010 14:57:57] up HIS sleeve
[29-Jul-2010 14:58:00] rather
[29-Jul-2010 14:58:36] http://fpaste.org/8x7U/
[29-Jul-2010 14:58:41] here is after the zenpack remove
[29-Jul-2010 14:59:11] st3v3o: best recommendation .. restore directory from tape if it was accidentally removed
[29-Jul-2010 14:59:29] backup hasn't run on the box yet
[29-Jul-2010 14:59:34] he did this all today
[29-Jul-2010 14:59:47] second best .. look inside a $ZENHOME/ZenPacks/easy-install.pth file for the lines ZenPacks.Michael...
[29-Jul-2010 15:00:09] rocket I looked there also
[29-Jul-2010 15:00:10] the easy-install.pth file is what zenoss uses to actually load the code into memory
[29-Jul-2010 15:00:17] don't see it listed
[29-Jul-2010 15:00:50] dont see it at all? .. and what happens if you restart zenoss in this case?
[29-Jul-2010 15:01:10] yeh, that's a good question
[29-Jul-2010 15:01:29] half of the daemons have errors in teh logs and then quit
[29-Jul-2010 15:02:07] I'll bounce it again
[29-Jul-2010 15:02:20] st3v3o: can you reapply the zenpack on top of itself?
[29-Jul-2010 15:03:37] no because the dude deleted it
[29-Jul-2010 15:03:42] and rm -rf the directory
[29-Jul-2010 15:04:00] you may be somewhat hosed .. this isnt a simple fix
[29-Jul-2010 15:04:11] yeah thats what my googling tells me :P
[29-Jul-2010 15:04:35] this is why I setup a "staging" zenoss to play on ....
[29-Jul-2010 15:04:58] may want to look at removing zenpack objects via zendmd
[29-Jul-2010 15:05:39] we tried that
[29-Jul-2010 15:05:45] yearh rmatte suggested that …and when we did that it sayd it doesn't exist
[29-Jul-2010 15:06:08] I don't get why this is so difficult to figure out lol
[29-Jul-2010 15:06:26] because your not supposed to randomly delete stuff :P
[29-Jul-2010 15:06:32] ok stop zenoss... start zeo and zenhub
[29-Jul-2010 15:06:33] true
[29-Jul-2010 15:06:38] zemdmd
[29-Jul-2010 15:06:41] but I didn't envisage it being this difficult to fix
[29-Jul-2010 15:07:19] dmd.ZenPackManager.packs()
[29-Jul-2010 15:07:25] what does that show?
[29-Jul-2010 15:07:28] after a restart the following services don't come back up …zenping, zenstatus, zenperfsnmp, zencommnd ,
[29-Jul-2010 15:07:39] I dont want any other services up
[29-Jul-2010 15:07:46] I want them all down but zenhub and zeo
[29-Jul-2010 15:08:26] http://fpaste.org/ghJV/
[29-Jul-2010 15:09:02] how do I start a single zenoss daemon ?
[29-Jul-2010 15:09:20] ZenPacks.michaelv.MySQLStats
[29-Jul-2010 15:09:34] zendmd
[29-Jul-2010 15:09:58] dmd.ZenPackManager.packs._delObject('ZenPacks.michaelv.MySQLStats')
[29-Jul-2010 15:10:01] commit()
[29-Jul-2010 15:10:18] look for that in the easy-install.pth file as well
[29-Jul-2010 15:10:31] *runningn steps now*
[29-Jul-2010 15:11:25] not in easy-install.pth
[29-Jul-2010 15:11:50] >>> dmd.ZenPackManager.packs._delObject('ZenPacks.michaelv.MySQLStats')
[29-Jul-2010 15:11:50] >>> 4:10
[29-Jul-2010 15:11:50] File "", line 1
[29-Jul-2010 15:11:50] 4:10
[29-Jul-2010 15:11:50] ^
[29-Jul-2010 15:11:50] SyntaxError: invalid syntax
[29-Jul-2010 15:11:51] >>> commit()
[29-Jul-2010 15:12:24] did you enter 4:10 on the cmdline?
[29-Jul-2010 15:12:30] nope
[29-Jul-2010 15:12:38] I cut and paste my input
[29-Jul-2010 15:12:40] copy and paste from irc?
[29-Jul-2010 15:12:44] yeah
[29-Jul-2010 15:12:50] you got an irc timestamp in there
[29-Jul-2010 15:12:56] go ahead and commit()
[29-Jul-2010 15:13:06] mother..!@#$!@#$
[29-Jul-2010 15:13:21] ?
[29-Jul-2010 15:14:17] I think that did it
[29-Jul-2010 15:14:28] I was just pissed I copied the IRC time stamp
[29-Jul-2010 15:14:36] new guy move
[29-Jul-2010 15:14:37] ok ...
[29-Jul-2010 15:15:01] thanks rocket you is the man
[29-Jul-2010 15:15:05] and rmatte
[29-Jul-2010 15:15:10] Thank you for using zenoss support ... Please push forward the karma and consider a software license :)
[29-Jul-2010 15:15:12] each time I try to add a datasource to a template, I get: Unable to connect to the server.
[29-Jul-2010 15:16:36] check your $ZENHOME/logs you most likely have a daemon down
[29-Jul-2010 15:19:12] doesn't appear to be the case
[29-Jul-2010 15:19:34] 3.0? 2.5.2?
[29-Jul-2010 15:19:54] 3.0
[29-Jul-2010 15:19:55] may be a bug in 3.0
[29-Jul-2010 15:20:26] ya
[29-Jul-2010 15:20:29] file a ticket at http://dev.zenoss.com/trac ..
[29-Jul-2010 15:20:53] reviewed twice a week .. if you can replicate it and put down the steps it will get resolved faster
[29-Jul-2010 15:21:02] k
[29-Jul-2010 15:21:25] follow the ticket and if it is a bug there will be a patch you can apply once its fixed
[29-Jul-2010 15:23:04] damn, i really need to monitor my tmpfs ramdisk over the weekend too
[29-Jul-2010 15:23:32] for some damn reason, net-snmp (patched for tmpfs) only shows me /dev/shm and not my custom mount point
[29-Jul-2010 15:23:43] so Zenoss will not monitor it since net-snmp doesn't report it
[29-Jul-2010 15:27:13] 3.0 is pretty sweet much better than 2.5 but there does seem to be some bugs/inconsistencies. is there a planned bugfix release comeing soon?
[29-Jul-2010 15:29:50] citrus: there will be a few
[29-Jul-2010 15:29:56] 3.0.1 will be out next week I believe
[29-Jul-2010 15:30:07] I'd expect it to go to at least 3.0.3
[29-Jul-2010 15:30:37] 2.5.0 was pretty bad, same with 2.5.1, 2.5.2 is when they finally got the majority of the bugs fixed
[29-Jul-2010 15:30:45] so it'll follow a similar pattern with 3.0 I'd imagine
[29-Jul-2010 15:31:09] That's why I'm not upgrading my production boxes yet
[29-Jul-2010 15:31:18] I'm waiting for bugs to get fixed
[29-Jul-2010 15:31:43] i wish i hadn't upgraded
[29-Jul-2010 15:31:49] but nonetheless, will wait til next week
[29-Jul-2010 15:32:01] its a sweet interface, lots of nice stuff to get accustomed to
[29-Jul-2010 15:32:34] i am pretty much relearning it. the interface is almost completly different
[29-Jul-2010 15:32:38] but it makes a lot more sense
[29-Jul-2010 15:32:46] however it only runs well in chrome
[29-Jul-2010 15:32:51] firefox its slow and IE it freezes
[29-Jul-2010 15:34:23] nah, the organization of the interface is different but the elements are the same
[29-Jul-2010 15:34:29] I didn't find it difficult to find anything
[29-Jul-2010 15:34:56] but yeh, there are performance issues with it
[29-Jul-2010 15:34:58] it's quite slow
[29-Jul-2010 15:35:18] there's a heck of a lot more ajax, so that's probably what's causing the performance issues
[29-Jul-2010 15:35:18] <|rt|> not everything works in chrome though
[29-Jul-2010 15:35:29] yeh, chrome is only partially supported
[29-Jul-2010 15:35:31] its too bad there isn't more beta time for major releases like 3.0 you'd think there would have been tons of betas and rcs
[29-Jul-2010 15:35:40] there were.
[29-Jul-2010 15:35:47] months and months worth of public betas.
[29-Jul-2010 15:36:17] oh.. well you'd think more of the performance issues would have been ironed out then
[29-Jul-2010 15:36:17] cgibbons: the releases do tend to be "rushed" though I find
[29-Jul-2010 15:36:43] cgibbons: people are still actively finding multiple bugs per day, yet it gets released anyways
[29-Jul-2010 15:36:49] when is that not true?
[29-Jul-2010 15:36:52] i'm just saying a big . release like three.oh should be pretty darn stable
[29-Jul-2010 15:37:07] cgibbons: there are some projects that don't have glaring bugs when they are released
[29-Jul-2010 15:38:19] then again, releasing it increases your testing group, since a lot of people don't bother with beta testing the beta versions
[29-Jul-2010 15:38:28] I wish more people would
[29-Jul-2010 15:38:39] true
[29-Jul-2010 15:39:17] when was 3.0 release?
[29-Jul-2010 15:39:31] It amazes me how people wait until the first release and then complain about bugs when they made no effort to contribute to testing the betas and RCs
[29-Jul-2010 15:39:45] 3.0 was released 2 thursdays ago I believe
[29-Jul-2010 15:39:49] so 2 weeks ago
[29-Jul-2010 15:39:51] i was oblivious to the betas. i totally would have
[29-Jul-2010 15:40:09] citrus: beta links are always posted in here, so it's worth dropping by once in a while to check
[29-Jul-2010 15:40:41] i love running zenoss in a virtual machine.. i can easily roll back to a snapshot with a click of mouse button if its too buggy
[29-Jul-2010 15:40:52] lots of people don't read the release notes, either :)
[29-Jul-2010 15:41:37] i mean, i certainly never do
[29-Jul-2010 15:42:09] * Simon4 almost never does
[29-Jul-2010 15:42:16] at least, not until something breaks ;)
[29-Jul-2010 15:42:22] cgibbons: I usually skim over them
[29-Jul-2010 15:42:42] one thing they i am missing.. idk if otheres are having this issue, but my search is gone from the upper right part of the page. the admin guide shows it.. but i don't see it at all
[29-Jul-2010 15:42:47] I always beta test before putting it in to production though
[29-Jul-2010 15:42:51] search was a cluster
[29-Jul-2010 15:42:52] it's coming back
[29-Jul-2010 15:42:54] it's 'moved'
[29-Jul-2010 15:42:55] helps prevent those "holy shit" momentw
[29-Jul-2010 15:42:59] moments*
[29-Jul-2010 15:43:17] cgibbons: that search zenpack you mean?
[29-Jul-2010 15:43:32] i dunno the whole details
[29-Jul-2010 15:44:52] <|rt|> I could use a dependancy zenpack more than search :)
[29-Jul-2010 15:44:52] rmatte, I see the definitions of events methods here: $ZENHOME/Products/ZenEvents/EventManagerBase.py
[29-Jul-2010 15:45:10] look a nice place to find something else more :D
[29-Jul-2010 15:45:24] looks like*
[29-Jul-2010 15:45:58] guys, im creating alert rules and wanted to divide devices in groups or whatever so that for instance events from x group alerts only this person; events from another device group alert other person..
[29-Jul-2010 15:46:08] what would be the way to create something like that?
[29-Jul-2010 15:46:30] chachan: ah, I know the ones that I need already :)
[29-Jul-2010 15:46:46] JonnyV: put the devices into a group under /groups and use that as a filter on the alerting rules?
[29-Jul-2010 15:46:47] good :D
[29-Jul-2010 15:46:52] JonnyV: there are literally device groups you can create
[29-Jul-2010 15:47:04] JonnyV: (not to be confused with device classes)
[29-Jul-2010 15:47:09] i was aha
[29-Jul-2010 15:47:12] sorry
[29-Jul-2010 15:47:13] aha
[29-Jul-2010 15:47:15] JonnyV: then you can set filters in the events based on those groups
[29-Jul-2010 15:47:22] erm, in the alerting rules
[29-Jul-2010 15:47:23] rather
[29-Jul-2010 15:47:51] that was the way i was thinking of.. the easiest way that came to my mind.. Cool
[29-Jul-2010 15:48:01] yup
[29-Jul-2010 15:48:10] devices can be members of more than one group too
[29-Jul-2010 15:48:13] something to keep in mind
[29-Jul-2010 15:48:47] i will play with that now
[29-Jul-2010 15:48:52] cool
[29-Jul-2010 15:48:53] thanks for the tip guys
[29-Jul-2010 15:49:02] np
[29-Jul-2010 15:49:17] you can send the beer to 1700 woodward drive, Ottawa, Ontario, Canada
[29-Jul-2010 15:49:24] :)
[29-Jul-2010 15:49:29] hehehe
[29-Jul-2010 15:49:38] * Simon4 looks around the lounge for something random to send
[29-Jul-2010 15:49:50] <|rt|> he likes Papts Blue Ribbon
[29-Jul-2010 15:51:57] no, I like Keith's
[29-Jul-2010 15:51:59] :P
[29-Jul-2010 15:52:23] wow.. look at this well designed webpage. http://havenworks.com/
[29-Jul-2010 15:52:33] <|rt|> heh....I just brew my own :)
[29-Jul-2010 15:52:52] citrus: that is horrific
[29-Jul-2010 15:53:01] citrus: and extremely vulgar
[29-Jul-2010 15:53:05] <|rt|> citrus: atleast they didn't use the tag
[29-Jul-2010 15:53:20] but it has news that's fair and balanced. whats horrific about that
[29-Jul-2010 15:53:22] |rt|: yeh, it definitely needs some or some marquis
[29-Jul-2010 15:54:38] it looks as tho someone had some content and some html tags and threw them both in a blender and that webpage was a result of what happened
[29-Jul-2010 15:54:46] citrus: it would be balanced if it weren't for the margin rape
[29-Jul-2010 15:55:06] half the page is off the screen
[29-Jul-2010 15:55:20] haha i didn't see that till now
[29-Jul-2010 15:55:31] <|rt|> if you had 1000 monkeys with type writers you wouldn't end up with shakespear you'd end up with that website
[29-Jul-2010 15:56:07] indeed
[29-Jul-2010 15:56:14] <|rt|> maybe someone erased a css by accident
[29-Jul-2010 15:56:39] is there a way to mass remodel devices?
[29-Jul-2010 15:56:44] I know it happens automatically on schedule
[29-Jul-2010 15:56:47] although you'd have to substitute monkeys for rhinos with hemorrhoids
[29-Jul-2010 15:56:55] but is there a way to select a bunch of devices and say "model device"
[29-Jul-2010 15:57:11] mducharme-laptop: zenmodeler run
[29-Jul-2010 15:57:36] that models everything though right?
[29-Jul-2010 15:57:36] heh
[29-Jul-2010 15:57:47] oh, you want to model specific devices
[29-Jul-2010 15:57:51] yeah
[29-Jul-2010 15:57:57] ummm, let me check...
[29-Jul-2010 15:57:58] not everything, just 25 devices
[29-Jul-2010 15:58:10] zenmodeler run -p "/Device/Class/Here" will do a single device class
[29-Jul-2010 15:58:17] ahh that's what i want
[29-Jul-2010 15:58:31] * Simon4 uses that lots
[29-Jul-2010 15:58:34] nope, you can run multiple commands, just not multiple models
[29-Jul-2010 15:58:48] I personally wrote a script that I can just feed a list and it models the devices
[29-Jul-2010 15:58:53] want me to pass it to you?
[29-Jul-2010 15:59:08] that's ok, simon's command will be fine
[29-Jul-2010 15:59:15] it happens to be all the routers I want to remodel
[29-Jul-2010 15:59:22] ah, they are all in the same device class?
[29-Jul-2010 15:59:24] ah, ok
[29-Jul-2010 15:59:25] yup
[29-Jul-2010 15:59:26] ooooh here is a very nicely done flash website
[29-Jul-2010 15:59:28] http://www.miauk.com/mayaaspect/
[29-Jul-2010 16:00:04] here's the script
[29-Jul-2010 16:00:05] http://fpaste.org/6nwQ/
[29-Jul-2010 16:00:22] you save it as a .sh, chmod +x it, and throw it in /home/zenoss or something
[29-Jul-2010 16:00:30] then you create a text file which can be named anything
[29-Jul-2010 16:00:38] put the devices in it line by line
[29-Jul-2010 16:00:44] then ./model.sh devices.txt
[29-Jul-2010 16:00:45] or whatever
[29-Jul-2010 16:00:47] and off it goes
[29-Jul-2010 16:01:19] I should have used $ZENHOME instead of the full path, I wrote that script ages ago
[29-Jul-2010 16:09:08] whats the system organizer for? how can it be used?
[29-Jul-2010 16:09:18] same as groups
[29-Jul-2010 16:09:24] just a different way to organize devices
[29-Jul-2010 16:09:29] you can use it however you wish
[29-Jul-2010 16:09:56] cool
[29-Jul-2010 16:10:17] will leave it empty for now ..
[29-Jul-2010 16:10:27] cool
[29-Jul-2010 16:10:40] I don't even use systems
[29-Jul-2010 16:10:45] just groups and locations
[29-Jul-2010 16:11:21] :)
[29-Jul-2010 16:13:40] has anyone in here used zenos to check content on a website?
[29-Jul-2010 16:13:53] zenoss rather
[29-Jul-2010 16:13:56] when did you start using zenoss, rmatte? what ver?
[29-Jul-2010 16:14:14] citrus: yes, you'd be looking for the HTTP Monitor ZenPack and the Synthetic Transactions ZenPack
[29-Jul-2010 16:14:26] cgibbons: 2.3.2
[29-Jul-2010 16:14:39] actually I think I was even at 2.3.1 at one point
[29-Jul-2010 16:16:27] citrus, I do
[29-Jul-2010 16:16:51] ok so spring 2008
[29-Jul-2010 16:17:07] or fall or whenever we got those out
[29-Jul-2010 16:17:12] guess spring was a 2.2
[29-Jul-2010 16:17:56] it was a little over a year and a half ago I think
[29-Jul-2010 16:18:24] why do you ask?
[29-Jul-2010 16:18:45] I was wondering if you ever used zenoss before it got relatively stable :)
[29-Jul-2010 16:18:50] ah I see
[29-Jul-2010 16:18:56] nah, 2.3.3 was pretty stable
[29-Jul-2010 16:19:04] Prior to about 2.2.4 there wasn't any QA or real testing done.
[29-Jul-2010 16:19:05] although it was still unstable compared to 2.5
[29-Jul-2010 16:19:11] because of the memory leak in RRDTool
[29-Jul-2010 16:19:18] which was always just a real treat
[29-Jul-2010 16:19:25] i bet
[29-Jul-2010 16:19:28] I don't miss having to restart Zenoss on all of my servers every week
[29-Jul-2010 16:19:29] lol
[29-Jul-2010 16:19:48] when reading alert rules, zenoss works on a first match order from top to bottom? or chooses the most specific filter?
[29-Jul-2010 16:20:01] JonnyV: it uses all filters
[29-Jul-2010 16:20:19] think of it as filter and filter and filter and filter and
[29-Jul-2010 16:20:25] it has to meet all of the filters to proceeed
[29-Jul-2010 16:20:27] proceed*
[29-Jul-2010 16:20:47] ok... great to know that
[29-Jul-2010 16:25:57] if I would like to create an event which is already created, what happens?, does the count grows up?
[29-Jul-2010 16:26:04] yep
[29-Jul-2010 16:26:18] cool
[29-Jul-2010 16:27:26] yeh, I like the count setup, saves database space
[29-Jul-2010 16:33:03] at least with 2.5.1 i had to restart zenoss once a week too hopefully 3.0 is a bit better in that regard
[29-Jul-2010 16:39:17] citrus: there is a #zenoss-testing channel that gets pretty active in the alpha & beta testing period. our QA manager likes to engage the community to help out for that when it's time.
[29-Jul-2010 16:39:50] Anyone ever monitored a Dec terminal server with Zenoss?
[29-Jul-2010 16:40:38] wow i haven't seen one of those since the mid 90s
[29-Jul-2010 16:47:45] i guess the last model they introduced was in 2003
[29-Jul-2010 16:48:15] i was going to say thats freaking old. but i still have a couple servers here running from 2000
[29-Jul-2010 16:49:16] I remember writing DECnet -> LAT and TCP/IP -> LAT protocol translators for those when I was a young man. Ahh back in the old days when 500 lines of code could do something amazing.
[29-Jul-2010 16:52:36] bouncy
[29-Jul-2010 16:52:42] lol i remember righting my first line of programing. ahhh qbasic
[29-Jul-2010 17:01:57] holy moly
[29-Jul-2010 17:02:02] I just added lots of routers and switches
[29-Jul-2010 17:02:10] now my network map has like so many lines, it's insane
[29-Jul-2010 17:02:17] there are lines everywhere
[29-Jul-2010 17:06:30] i just dragged a device to a new device class 3 times before it worked
[29-Jul-2010 17:06:33] same way each time
[29-Jul-2010 17:06:35] hehe
[29-Jul-2010 17:12:19] how do i change the default device class "configuration properties" settings ?
[29-Jul-2010 17:17:14] it's simply the zproperties for the class
[29-Jul-2010 17:17:19] I assume you mean
[29-Jul-2010 17:26:13] where do you edit the zproperties for a device class in 3.0?
[29-Jul-2010 17:26:48] citrus: Configuration Properties
[29-Jul-2010 17:27:26] on the Device Class hit Details
[29-Jul-2010 17:27:31] and then you'll see it
[29-Jul-2010 17:27:35] as an option
[29-Jul-2010 17:27:37] ahhhhhhh
[29-Jul-2010 17:27:40] thank you
[29-Jul-2010 17:27:46] haven't hit the details button yet
[29-Jul-2010 17:27:53] part of our effort to make it easier for new users, and harder for existing users :)
[29-Jul-2010 17:28:14] old users don't work hard enough
[29-Jul-2010 17:28:22] get to work rmatte!
[29-Jul-2010 17:28:46] hehe... the details button is great
[29-Jul-2010 17:38:19] the zenoss.cmd.df is the plugin which get disk space information?
[29-Jul-2010 17:39:07] that's the modeler
[29-Jul-2010 17:39:19] for SSH-based collection
[29-Jul-2010 17:39:27] umm
[29-Jul-2010 17:40:02] i got it, I'm looking for change thresholds on disk space warning
[29-Jul-2010 17:40:11] for a class
[29-Jul-2010 17:40:47] but I don't find the class that contains these thresholds
[29-Jul-2010 17:51:23] yeah, the UI for that sucks
[29-Jul-2010 17:51:39] go Advanced->Monitoring Templates
[29-Jul-2010 17:51:49] FileSystem
[29-Jul-2010 17:51:56] and you can copy and override it there
[29-Jul-2010 17:52:20] hopefully it'll be fixed in 3.0.2
[29-Jul-2010 18:07:25] Anyone know what the ETA is for releasing 3.0.x?
[29-Jul-2010 18:07:32] Have they mentioned it yet?
[29-Jul-2010 18:11:13] I'm using Zenoss Core 2.5.2, anyone know why when I try to test a data source through ssh against a specific device, it ALWAYS runs against the zenoss server?
[29-Jul-2010 18:30:20] Can anyone see what I am doing wrong here? docs/DOC-3431#comment-2341
[29-Jul-2010 18:37:04] the above is me trying to get the HTTP MONITOR zenpack trying to work
[29-Jul-2010 19:15:10] How do you run a zencommand over ssh?
[29-Jul-2010 19:27:35] I just want to send a command over ssh to one of my devices and process the results. I've done this before but for the life of me, I can't get it to work again on a fresh install.
[29-Jul-2010 20:29:07] mrayzenoss is now known as mray
[29-Jul-2010 21:28:58] mrayzenoss is now known as mray
[29-Jul-2010 22:48:37] mrayzenoss is now known as mray
[29-Jul-2010 23:21:16] hey, mray, thanks for your answers wrt http monitor
[29-Jul-2010 23:28:37] I am stepping afk but will be back in a bit, if you car to be more interactive on that question.
[29-Jul-2010 23:28:43] s/car/care
[30-Jul-2010 00:00:30] [disconnected at Fri Jul 30 00:00:30 2010]
[30-Jul-2010 00:00:30] [connected at Fri Jul 30 00:00:30 2010]
[30-Jul-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]
[30-Jul-2010 04:42:37] wow, zenoss in chrome is significant faster then in firefox
[30-Jul-2010 04:43:46] chrome = speedy++ for javascript
[30-Jul-2010 04:43:52] pity it's not fully supported :(
[30-Jul-2010 05:01:49] can i hand over a list of devices to zendisc?
[30-Jul-2010 05:08:17] I wrap it in a script
[30-Jul-2010 05:08:33] for i in `cat devicelist`
[30-Jul-2010 05:08:33] do
[30-Jul-2010 05:08:42] zendisc --options $i
[30-Jul-2010 05:08:43] done
[30-Jul-2010 05:08:43] etc
[30-Jul-2010 05:08:45] Simon4: i just didn't want to call zendisc in a loop. spawning new ones all over the place
[30-Jul-2010 05:08:49] heh
[30-Jul-2010 05:08:51] but it's fun :)
[30-Jul-2010 05:09:17] Simon4: i started to care about the amount of processes i spawn.
[30-Jul-2010 05:09:31] fragfutter: well looping it like that you're only spawning one at a time
[30-Jul-2010 05:09:48] and throwing it out, and spawning again, and ...
[30-Jul-2010 05:10:01] how often are you doing this? presumably it's a once-off
[30-Jul-2010 05:10:16] but it's not an embedded system... heck "no" resource constrains.
[30-Jul-2010 05:10:41] not very often, just thinking if there is something better.
[30-Jul-2010 05:12:26] yeah fair enough
[30-Jul-2010 05:18:16] how do i add a device to a group?
[30-Jul-2010 05:23:50] ohhhhh, shiny drag and drop, nothing to click.
[30-Jul-2010 05:29:52] i think zenoss needs a UI Tester/Designer for the next release ;)
[30-Jul-2010 06:56:12] fragfutter: there are some screens in a compatibility mode yet that have just been reskinned they are to be replaced shortly
[30-Jul-2010 06:56:41] fragfutter: as to what shortly means thats up for debate :)
[30-Jul-2010 06:57:35] http://dev.zenoss.org/trac/ticket/7136
[30-Jul-2010 06:58:37] heh .. thats minor .. :p
[30-Jul-2010 06:59:08] sure it is, but i find stuff like this all over the place.
[30-Jul-2010 06:59:30] developer-mode: that's just spit and polish
[30-Jul-2010 06:59:43] user-mode: unfinished product
[30-Jul-2010 07:00:27] fragfutter: There are more serious things I would like to see addressed before something like that .. though that should be a quick to fix issue
[30-Jul-2010 07:01:03] yes, the bug is only there so it isn't lost. lowest priority affixed to it.
[30-Jul-2010 07:01:20] btw thanks for filing a bug on it :)
[30-Jul-2010 07:01:35] it may not be addressed without a bug :)
[30-Jul-2010 07:21:48] is there anything in zenoss like nagos flapdetection ?
[30-Jul-2010 07:21:54] nagios*
[30-Jul-2010 08:35:24] fOrsberg is now known as forsberg
[30-Jul-2010 08:46:22] I've been wondering too, because I've wanted to start using it
[30-Jul-2010 08:46:29] (flap detection)
[30-Jul-2010 08:48:14] tehhobbit: not currently
[30-Jul-2010 08:48:29] what exactly does the "flap detection" script do?
[30-Jul-2010 08:48:46] (I know what a flap is, I'm just asking for specifics)
[30-Jul-2010 08:48:55] if a service goes repeated down and up it will be flagged as flapping
[30-Jul-2010 08:49:10] for flapping services you only get _one_ notification
[30-Jul-2010 08:49:14] I see, yeh, there's nothing like that
[30-Jul-2010 08:50:04] typical notification is, "service down", "service up", "service down", "service flapping", .... wait ...., "service no longer flapping", "service up"
[30-Jul-2010 08:50:25] so its more then one mail, but at least not a hundred ;)
[30-Jul-2010 08:50:28] by service you mean for an interface?
[30-Jul-2010 08:50:39] anything nagios can check
[30-Jul-2010 08:50:49] I see
[30-Jul-2010 08:50:52] process running, interfaces, cpu load, ...
[30-Jul-2010 08:51:09] it only checks every 5 minutes anyways, pretty hard to determine something is flapping in that amount of time
[30-Jul-2010 08:51:35] rmatte: nagios explains its flap detection in detail (default also checks only every five minutes)
[30-Jul-2010 08:51:39] also, I'll point out that all of those events auto-clear, so you can always just set a repeating time on your alerting rules
[30-Jul-2010 08:51:56] so if the alert is still around in 15 minutes, send another email
[30-Jul-2010 08:51:58] sort of thing
[30-Jul-2010 08:52:05] rmatte: thats not the goal.
[30-Jul-2010 08:52:24] thats repeating, same feature in nagios.
[30-Jul-2010 08:52:38] I don't see why you'd be getting a hundred emails in the first place is what I'm saying
[30-Jul-2010 08:52:45] Zenoss increases the event count and groups events
[30-Jul-2010 08:52:51] so I don't see how that's even an issue
[30-Jul-2010 08:53:05] say something goes up for five minues, then is down for five minutes, then is up....
[30-Jul-2010 08:53:21] good morning everyone
[30-Jul-2010 08:53:31] zenoss would send "its down", "clear", "its down", "clear", "its down", "clear", ...
[30-Jul-2010 08:53:50] if i place a delay on the alert rule i would get nothing
[30-Jul-2010 08:54:30] fragfutter: I didn't mean delay, I meant repeat time
[30-Jul-2010 08:54:57] rmatte: that is if a service stays down it tells me every hour (or so) that it is down, still down, still down.
[30-Jul-2010 08:55:04] right
[30-Jul-2010 08:55:05] rmatte: different issue
[30-Jul-2010 08:55:14] so you just don't want to get an email every 5 minutes for each state change
[30-Jul-2010 08:55:30] rmatte: reset, start discussion over.
[30-Jul-2010 08:55:44] I understand what you're saying
[30-Jul-2010 08:55:53] rmatte: nagions approaches things as if it's still the same dedupid being treated
[30-Jul-2010 08:55:53] no, there's nothing like that built in...
[30-Jul-2010 08:56:16] where in zenoss' case it would be a new dedupid, thus hitting a new alert
[30-Jul-2010 08:56:34] but it's definitely something I've looked into writing after the dep code, if it didn't exist
[30-Jul-2010 08:56:41] with the altenating up-down service i will describe what nagios does and then what i want to have in nagios.
[30-Jul-2010 08:56:59] fragfutter: I understand, you don't need to re-explain
[30-Jul-2010 08:57:18] you want it to identify back to back ups/downs as "flaps" and create a "flap" event instead
[30-Jul-2010 08:57:37] nagios sends: service down, service flapping, clear. And nagios displays a service as flapping in the webinterface even if it is up while i look at the interface
[30-Jul-2010 08:58:01] so you need to do two things, the event handling and a third state of a service (flapping)
[30-Jul-2010 08:58:09] it's probably possible to hack it into zenoss by creating a generic event transform under event manager, but yeah, haven't poked at it..
[30-Jul-2010 08:58:30] well, doing it with an event transform won't add the gui component that he's talking about
[30-Jul-2010 08:58:55] rmatte: true enough
[30-Jul-2010 08:59:15] but I'm going to try steer clear of anything zope and web-component related for as long as possible ;)
[30-Jul-2010 08:59:15] or you need to create a new event type flapping that get's generated if a service is changing state more then N times during the last X checks and a clear event that fires if Service is up for at least Y checks
[30-Jul-2010 09:01:43] fragfutter: what do you have that flaps that badly?
[30-Jul-2010 09:02:47] currently nothing ;) it's a feature of nagios that i relied on monitoring a heavy loaded webserver farm. Flapping machines got kicked from the balancer pool
[30-Jul-2010 09:03:15] failovers were too expensive to have them often.
[30-Jul-2010 09:03:28] I see
[30-Jul-2010 09:04:16] it's more a "if zenoss can do everything nagios can do and more, there would be no more reason for nagios to exist". and world domination is the goal for zenoss, isn't it?
[30-Jul-2010 09:05:49] zenoss is already better than nagios in my opinion :P
[30-Jul-2010 09:06:49] rmatte: i can configure hardcore detailed nagios checks faster and easier then in zenoss.
[30-Jul-2010 09:07:07] and for me nagios is totaly deterministic in the behaviour.
[30-Jul-2010 09:07:31] who was asking about ramdisks with wmi again?
[30-Jul-2010 09:07:48] no ramdisks with snmp was it i think
[30-Jul-2010 09:08:02] maybe ..
[30-Jul-2010 09:08:37] rmatte: have zenoss remodel something at night and zenoss might stop monitoring something if i forgot to lock it. But the autodiscovery and modelling is the zenoss feature i want
[30-Jul-2010 09:10:01] fragfutter: turn off automatic modelling .. you can run it manually from cron
[30-Jul-2010 09:10:07] fragfutter: that's why I don't do automatic modelling :P
[30-Jul-2010 09:10:39] and it was ramdisks with snmp
[30-Jul-2010 09:10:48] or manually if you know a change ..
[30-Jul-2010 09:10:53] QubeZ was asking about it
[30-Jul-2010 09:10:59] but it was a net-snmp problem
[30-Jul-2010 09:11:02] not a Zenoss problem
[30-Jul-2010 09:11:24] yeh, our customers tell us when they change stuff
[30-Jul-2010 09:11:29] so we only remodel when necessary
[30-Jul-2010 09:11:59] rmatte: still there is more magic in zenoss (the reason i evaluate it). nagios is just less code and simpler and i went through it in detail so i understand it completely.
[30-Jul-2010 09:12:16] rmatte: well, in my case, I've got 7 remote networks to care about to make sure their edge devices are available and routable etc
[30-Jul-2010 09:12:44] fragfutter: I don't see how you can say it's harder to customize nagios than Zenoss... Zenoss supports Nagios plugins :P
[30-Jul-2010 09:13:00] erm, I meant harder to customize Zenoss than nagios
[30-Jul-2010 09:13:04] (in terms of flap detection)
[30-Jul-2010 09:13:18] yeh, but that's a very specific usage case
[30-Jul-2010 09:13:30] perhaps
[30-Jul-2010 09:13:31] I'm sure if you filed a feature request it would eventually be implemented
[30-Jul-2010 09:14:04] but that's something that a lot of networks people would like to have and they'd instantly swap to zenoss from nagios (well, quickly, not instantly necessarily) if they could
[30-Jul-2010 09:14:07] I personally like actually seeing the times when interfaces go up and down
[30-Jul-2010 09:14:10] even if they are flapping
[30-Jul-2010 09:14:16] makes it easier to troubleshoot the issue itself
[30-Jul-2010 09:14:18] rmatte: and it probably could be managed via a transform etc ..
[30-Jul-2010 09:14:31] rocket: no doubt, but he wants a UI component
[30-Jul-2010 09:14:32] rmatte: well, you'd still see the things either way
[30-Jul-2010 09:15:02] but having things indicated as "it's flapping" can tell you the difference of shouting at your provider, or looking for hardware faults
[30-Jul-2010 09:15:03] frozty_sa: I doubt they'd swap until dependencies are resolved
[30-Jul-2010 09:15:32] yeah well, Simon4 and I are discussing and fiddling at that ;)
[30-Jul-2010 09:16:05] I really want to see dependencies implemented
[30-Jul-2010 09:16:12] manual dependencies at the very least
[30-Jul-2010 09:16:20] me also
[30-Jul-2010 09:16:27] It would make a whole world of difference
[30-Jul-2010 09:16:28] * Simon4 will be playing again in the weekend/tonight
[30-Jul-2010 09:16:34] cool
[30-Jul-2010 09:17:54] rmatte: i have/had a perl script that accepts three parameters balancer pool, node, instance, i need to add all permutations of this to check (111, 112, 121, 211, 113, 131, ... in total i had 72 checks), i can whip it up in nagios in 60seconds by editing a config file.
[30-Jul-2010 09:19:37] rmatte: but on the other hand i needed to edit a config file vor every new server i added, not just stick it in a class and if it has a webserver, the server get's monitored, if not, this is ignored
[30-Jul-2010 09:26:53] mrayzenoss is now known as mray
[30-Jul-2010 09:34:00] good morning everyone
[30-Jul-2010 09:35:26] :o
[30-Jul-2010 09:35:42] good morning
[30-Jul-2010 09:36:26] i have a question about zenoss, i am a new user of the product so forgive me for my stupid question ahead of time.
[30-Jul-2010 09:37:18] I am using Zenoss to monitor network bandwidth, most devices have no issues but some arent displaying any info and giving me a 2010-07-30 10:51:54 WARNING zen.ZenDisc Client TESTROUTER4 timeout when adding
[30-Jul-2010 09:38:34] then you don't have them configured correctly for snmp
[30-Jul-2010 09:39:43] when we add these devices it does accept them but it just doesnt grab the interface information.
[30-Jul-2010 09:43:05] MIHAJLO: maybe the default community can read basics, but is not allowed to access the interfaces?
[30-Jul-2010 09:43:28] MIHAJLO: yeh, you either have the wrong SNMP string set in Zenoss or you don't have the devices configured for SNMP
[30-Jul-2010 09:43:31] it's one or the other
[30-Jul-2010 09:43:43] like with public you can get the administrator of the device and the location but nothing else
[30-Jul-2010 09:44:29] i would check running an snmpwalk manually
[30-Jul-2010 09:46:12] but if i add the device/int manually it comes up
[30-Jul-2010 09:50:37] MIHAJLO: you should never have to add them manually
[30-Jul-2010 09:50:47] MIHAJLO: perform a remodel on the device and pastebin the output
[30-Jul-2010 09:50:53] I want to see what kinds of errors you're getting
[30-Jul-2010 09:51:01] ok
[30-Jul-2010 09:51:06] MIHAJLO: also, it doesn't "come up
[30-Jul-2010 09:51:08] "
[30-Jul-2010 09:51:12] Hi Simon4.....
[30-Jul-2010 09:51:19] Zenoss doesn't do interface monitoring via polls unless you specifically configure it
[30-Jul-2010 09:51:40] MIHAJLO: docs/DOC-2494
[30-Jul-2010 09:52:00] MIHAJLO: That's what you need to monitor interface status and actually have the status change on the OS tab
[30-Jul-2010 09:52:14] MIHAJLO: otherwise it won't monitor actual operational status
[30-Jul-2010 10:09:23] mrayzenoss is now known as mray
[30-Jul-2010 10:11:33] mrayzenoss is now known as mray
[30-Jul-2010 10:12:44] Simon4_ is now known as Simon4
[30-Jul-2010 10:28:23] Simon4: wrt your script to create devices with duplicate IP addresses........
[30-Jul-2010 10:28:38] yah?
[30-Jul-2010 10:28:57] .. why DOESN'T it check for duplicate IP addresses, using this script??
[30-Jul-2010 10:29:14] because it's bypassing zendisc, basically
[30-Jul-2010 10:29:28] whilst if you use the UI, it runs zendisc on the hostname you have asked to add
[30-Jul-2010 10:29:31] I don't see anything that changes the checkDeviceExists defaults......
[30-Jul-2010 10:29:37] which then checks for duplicate
[30-Jul-2010 10:30:57] I would have to go dig into the function calls to explain further, I just know it works
[30-Jul-2010 10:36:06] Simon4: you have hacked multiple devices with the same managed ip?
[30-Jul-2010 10:36:40] rocket: yeah
[30-Jul-2010 10:36:55] are you sure thats not breaking something else where?
[30-Jul-2010 10:37:23] rocket: it hasn't for the last 5 months, and the "clone" device has 99% of hte collectors turned off on it
[30-Jul-2010 10:37:33] * rocket has got a feeling indexes somewhere aren't right ..
[30-Jul-2010 10:39:28] rocket: http://nettikconsulting.wordpress.com/2010/03/19/zenoss-more-than-one-device-with-a-single-ip/
[30-Jul-2010 10:39:31] being how I do it
[30-Jul-2010 10:40:11] if it is breaking something in the background I would be keen to know, and will edit said blog post accordingly :)
[30-Jul-2010 11:16:47] morning
[30-Jul-2010 11:22:10] forsberg is now known as fOrsberg
[30-Jul-2010 11:23:48] afternoon
[30-Jul-2010 11:33:47] scons -j 8
[30-Jul-2010 12:15:59] hello all
[30-Jul-2010 12:16:32] is the default for messages sent out for load average supposed to be in the thousands? I get warning messages stating current value = 1350.00, im assuming it means 13.50
[30-Jul-2010 12:18:45] ok looks like it returns an integer like 1350, but the data point is set to 100,/ so that should spit out the right number in the email but isn't
[30-Jul-2010 12:35:46] must be lunch time
[30-Jul-2010 12:38:04] hey there
[30-Jul-2010 12:38:15] where do I download the zenplugin.py package?
[30-Jul-2010 12:38:51] MrWGW: chances are good you don't need the zenplugin, it's been deprecated since 2.5
[30-Jul-2010 12:39:02] alright, here's the problem I'm haing
[30-Jul-2010 12:39:19] the ubuntu serers I'm monitoring ia ssh, I'm getting very little useful information from them relative to what I've gotten in the past
[30-Jul-2010 12:39:34] for example
[30-Jul-2010 12:39:40] I get the size of each disk, ut not the used space
[30-Jul-2010 12:39:47] and I get no info about OS processes, IP services, et cetera
[30-Jul-2010 12:39:55] I am ssh'ing them as a lmiited duser
[30-Jul-2010 12:40:55] I figured installing the plugin would be what was needed
[30-Jul-2010 12:41:06] I'm on Zenoss 2.5.1 installed on the Zenoss virtual appliance
[30-Jul-2010 12:41:29] you'll want to get the LinuxMonitor ZenPack
[30-Jul-2010 12:41:40] and put your Ubuntu boxes in /Server/SSH/Linux
[30-Jul-2010 12:41:50] all the servers are in /Server/SSH/Linux
[30-Jul-2010 12:41:56] all the Linux ones rather
[30-Jul-2010 12:41:59] was the zenpack already installed?
[30-Jul-2010 12:42:17] no zenpacks were installed
[30-Jul-2010 12:42:24] I was under the impression that ssh/linux worked out of the box
[30-Jul-2010 12:42:32] and in the past it had oddly enough on a different zenoss install that I did
[30-Jul-2010 12:42:55] it's included in the ZenPacks RPM for RPM installs, but most people use the stack or VM and need to add it
[30-Jul-2010 12:43:05] ok
[30-Jul-2010 12:47:40] should I remodel all the devices after installing it?
[30-Jul-2010 12:48:02] yeah
[30-Jul-2010 12:49:02] I'm getting much more info, but presumably since the ssh user I'm using is non-root I'm not getting a list of running processes
[30-Jul-2010 12:49:09] I'm also not getting info on used disk space oddly enough
[30-Jul-2010 12:49:18] not sure process monitoring works
[30-Jul-2010 12:49:43] or actually, I haven't tried it myself
[30-Jul-2010 12:49:56] MrWGW: I can't think of any reason you'd need root to monitor processes
[30-Jul-2010 12:51:30] the weirder thing is wh yI'm not getting statistics on filesystem usage
[30-Jul-2010 12:56:29] it takes 15 minutes to start seeing that info
[30-Jul-2010 12:56:42] also, process monitoring most definitely does not require root
[30-Jul-2010 12:56:48] I think the process monitoring just doesn't work
[30-Jul-2010 12:56:52] someone dropped the ball on it
[30-Jul-2010 13:00:23] I'm also not getting cpu statistics
[30-Jul-2010 13:00:27] I would assume it needs like 15 minutes?
[30-Jul-2010 13:04:40] yes
[30-Jul-2010 13:04:47] 3 polling cycles for everything
[30-Jul-2010 13:04:53] 5 minutes per polling cycle
[30-Jul-2010 13:04:55] 15 minutes
[30-Jul-2010 13:10:23] is there a lot of editin to the snmp file for linux to get good info from it or should it just be as easy as enabling the right community. btw most of the linux servers are debian
[30-Jul-2010 13:11:20] the default snmp config file has paranoid security settings
[30-Jul-2010 13:11:28] you need to edit it to allow the read only string to get all info
[30-Jul-2010 13:11:31] rather than limited info
[30-Jul-2010 13:12:18] Here's a working config you can use: http://pastebin.com/RWTkFmSW
[30-Jul-2010 13:12:26] obviously modify the settings
[30-Jul-2010 13:13:02] cool thank you that should help
[30-Jul-2010 13:16:57] whats the different between paranoid and readonly
[30-Jul-2010 13:17:29] oh paranoid is only allowed access to the system and readonly has access to all?
[30-Jul-2010 13:19:35] its been around 15 minutes but I'm still not seeing any perormance data
[30-Jul-2010 13:19:37] sorry i am still getting used to the snmp stuff. does system provide disk/cpu?
[30-Jul-2010 13:19:57] the system OID i guess
[30-Jul-2010 13:28:08] no
[30-Jul-2010 13:28:14] the system OID provides almost nothing
[30-Jul-2010 13:28:23] just very very basic system info
[30-Jul-2010 13:28:36] should cisco multilayer switches go into /Network/Switches or /Network/Routers/Cisco
[30-Jul-2010 13:28:49] because they do routing as well
[30-Jul-2010 13:28:55] mducharme-laptop: I made my own class of /Network/Switches/Cisco
[30-Jul-2010 13:29:02] and copied the template from /Network/Routers/Cisco
[30-Jul-2010 13:29:23] mducharme-laptop: you'll have to do lots of customization, the base templates aren't all that great
[30-Jul-2010 13:29:39] mducharme-laptop: Zenoss is more of a framework/blank slate more than anything
[30-Jul-2010 13:30:05] it already does more than we had before
[30-Jul-2010 13:31:11] it does do a lot, but be prepared to learn how to design performance templates
[30-Jul-2010 13:31:17] the guy who runs the network is pulled in too many different directions.. he was managing nagios/cacti/smokeping/phpsyslog/apstats
[30-Jul-2010 13:31:34] a smokeping pack for Zenoss would be nice
[30-Jul-2010 13:31:40] he would do updates on boxes but not have time to fully do the research on whether the upgrades were compatible
[30-Jul-2010 13:31:57] so he would upgrade php on the cacti box and cacti would be broken
[30-Jul-2010 13:32:02] nice
[30-Jul-2010 13:32:02] and we would have no graphs for weeks
[30-Jul-2010 13:32:32] he had too much stuff to manage and not enough time to do the research before upgrading things
[30-Jul-2010 13:33:14] with zenoss there should be no excuse, because he is only upgrading one system
[30-Jul-2010 13:33:31] yeh, it still requires a lot of testing beforehand though
[30-Jul-2010 13:34:30] yes.. but doing system updates on one box is simpler than doing system updates on 10 different boxes.. we have zenoss running in a vm too so can just snapshot or backup the vm before upgrading
[30-Jul-2010 13:34:52] yeh
[30-Jul-2010 13:34:53] something goes wrong, we revert
[30-Jul-2010 13:35:00] well, I have Zenoss on 14 different boxes
[30-Jul-2010 13:35:01] :)
[30-Jul-2010 13:35:10] that's different heh
[30-Jul-2010 13:35:41] it's quite the setup
[30-Jul-2010 13:36:14] i put my cisco pix in /network/router/cisco and enabled snmp on it. using snmpwalk i can get info from it but zenoss is persistant about the snmp agent being down in alerts.
[30-Jul-2010 13:36:36] the other good thing about zenoss here is that we can get more of our tech staf using it
[30-Jul-2010 13:36:47] citrus2: easy fix
[30-Jul-2010 13:36:54] at the moment our service desk staff have access to the network monitoring tools in nagios/cacti etc but they don't know how to use any of it
[30-Jul-2010 13:37:09] citrus2: go to zProperties for the device, change zMaxOIDPerRequest from 40 to 20
[30-Jul-2010 13:37:15] network documentation is in text files in subversion
[30-Jul-2010 13:37:42] rmatte, thank you.. can you explain that?
[30-Jul-2010 13:37:47] what does it do
[30-Jul-2010 13:37:59] citrus2: Cisco PIX and ASA are picky as hell about the size of the SNMP packets being sent to them
[30-Jul-2010 13:38:02] is it possible to get zenoss to store actual full cisco configs
[30-Jul-2010 13:38:11] and Zenoss tried to cram more in than snmpwalk does
[30-Jul-2010 13:38:15] ahh, sounds like cisco
[30-Jul-2010 13:38:15] tries*
[30-Jul-2010 13:38:42] mducharme-laptop: no, you'd want to use something like rancid for that
[30-Jul-2010 13:38:52] mducharme-laptop: we use ziptie, but it's a dead project now
[30-Jul-2010 13:39:17] we have something called ccm
[30-Jul-2010 13:39:21] was developed in houes
[30-Jul-2010 13:39:23] er in house
[30-Jul-2010 13:39:26] but it broke 3 months ago
[30-Jul-2010 13:39:30] and nobody has time to fix it
[30-Jul-2010 13:39:35] so we hav enothing now
[30-Jul-2010 13:39:42] man you guys need to hire some IT
[30-Jul-2010 13:39:56] mducharme-laptop: check out rancid
[30-Jul-2010 13:40:06] ziptie is good, but it's been pretty much wiped off the face of the earth
[30-Jul-2010 13:40:15] I plan to setup a site to host what remains of it
[30-Jul-2010 13:40:22] but I haven't gotten around to it yet
[30-Jul-2010 13:40:28] citrus2 the problem is the guy who custom coded this is the guy who runs the network
[30-Jul-2010 13:40:34] I was quite disappointed at ziptie going away
[30-Jul-2010 13:40:46] since he coded it he's the only one who knows how it works, and he's the busiest of all of us, so things get tied up with him
[30-Jul-2010 13:40:47] jmp242: I'm going to try to dig up the source from somewhere too
[30-Jul-2010 13:40:53] jmp242: and host that to encourage a fork
[30-Jul-2010 13:40:55] he's the only one who knows how to configure nagios/cacti/etc
[30-Jul-2010 13:41:06] we weren't monitoring all hosts because he was too busy to add some in there
[30-Jul-2010 13:41:07] jmp242: I have every binary that was available
[30-Jul-2010 13:41:15] so I'm still not getting any performance graphs at all and its been running for like 30 minutes
[30-Jul-2010 13:41:18] now we are monitoring everything with zenoss, and it's much simpler
[30-Jul-2010 13:41:31] MrWGW: then you have a genuine problem
[30-Jul-2010 13:41:35] also I'm not getting any inormation about disk space usage other than the size of each filesystem
[30-Jul-2010 13:41:43] Well, I never did end up using ziptie as it didn't work with our Procurves
[30-Jul-2010 13:41:45] MrWGW: this is with SSH?
[30-Jul-2010 13:41:48] yes
[30-Jul-2010 13:41:51] we seemed to have too low level ones :(
[30-Jul-2010 13:41:55] ssh monitoring of ubuntu 8.04 servers
[30-Jul-2010 13:42:08] MrWGW: they ae in the /Server/Linux/SSH class, right?
[30-Jul-2010 13:42:13] are*
[30-Jul-2010 13:42:49] yes
[30-Jul-2010 13:42:58] check the list of collector plugins for that class
[30-Jul-2010 13:43:06] make sure all the required SSH plugins are bound to that class
[30-Jul-2010 13:43:23] also check zencommand.log for errors
[30-Jul-2010 13:44:31] where is zencommand.log?
[30-Jul-2010 13:44:45] Settings -> Daemons -> zencommand -> log
[30-Jul-2010 13:45:00] ah cool I was getting ready to ssh on in
[30-Jul-2010 13:45:05] or SSH to the server, become the zenoss user, then it's in $ZENHOME/log/zencommand.log
[30-Jul-2010 13:45:18] to become zenoss user you do: sudo su - zenoss
[30-Jul-2010 13:45:32] interesting, when I view it via the web interface, that log file is blank
[30-Jul-2010 13:45:37] I almost get the impression btw
[30-Jul-2010 13:45:46] that the system isn't actually rescanning the hosts after it models them
[30-Jul-2010 13:46:14] are all the daemons showing as running on the daemons page?
[30-Jul-2010 13:46:19] do they all have green lights?
[30-Jul-2010 13:47:21] yes
[30-Jul-2010 13:47:26] k
[30-Jul-2010 13:47:51] docs/DOC-5967#ModelingSSHCOMMAND
[30-Jul-2010 13:47:59] read that and double check all that stuff
[30-Jul-2010 13:48:04] zencommand.log is blank
[30-Jul-2010 13:48:22] yeh, that just means it hasn't done anything, though it's odd that it's completely blank
[30-Jul-2010 13:48:25] but whatever
[30-Jul-2010 13:48:30] read what I just pasted
[30-Jul-2010 13:48:38] (and read the entire Zenoss admin guide when you have time)
[30-Jul-2010 13:49:30] for the plugins portion...
[30-Jul-2010 13:49:41] make sure all of the zenoss.cmd plugins are present in the left pane
[30-Jul-2010 13:49:48] if some aren't, drag them over
[30-Jul-2010 13:50:22] when you edit the config properties for a device class does it override any preconfigured "config properties" for specific hosts that are inside ?
[30-Jul-2010 13:50:39] yes, unless you have specific local configs on those
[30-Jul-2010 13:51:52] so. if you changed one setting for a specific host.. do all of the settings become unaffected when you edit the config properties for that class? or just that one setting is static?
[30-Jul-2010 13:52:19] If you just edit the 1 setting for a specific devices that's the only setting that becomes local
[30-Jul-2010 13:52:27] got it
[30-Jul-2010 13:52:29] thanks
[30-Jul-2010 13:52:29] all other settings are still inherited from higher up
[30-Jul-2010 13:52:38] scroll to the bottom of zProperties
[30-Jul-2010 13:52:45] there's a dropdown that shows the locally applied properties
[30-Jul-2010 13:52:49] and allows you to remove them
[30-Jul-2010 13:52:56] is there a way to see what settings are local easily/
[30-Jul-2010 13:52:57] ?
[30-Jul-2010 13:53:00] got it
[30-Jul-2010 13:53:02] thanks
[30-Jul-2010 13:53:04] np
[30-Jul-2010 13:54:15] are you like the IRC goto guy for Zenoss? do they pay you to sit on irc and be awesome?
[30-Jul-2010 13:54:26] nope, they pay Matt Ray to do that
[30-Jul-2010 13:54:32] I just do it for the heck of it
[30-Jul-2010 13:54:41] (I don't work for Zenoss)
[30-Jul-2010 13:54:55] I just have a ton of experience with it
[30-Jul-2010 13:55:13] I wish I got paid to do this lol
[30-Jul-2010 13:55:16] lol they SHOULD pay you
[30-Jul-2010 13:55:17] I just need to make rmatte gets more swag :)
[30-Jul-2010 13:55:28] :)
[30-Jul-2010 13:55:30] s/make/make sure/
[30-Jul-2010 13:56:01] I'l take 10 zennies, 3 king crab t-shirts, 3 stone crab t-shirts, 4 pens, and a life sized Zebra
[30-Jul-2010 13:56:23] how about just a LIVE zebra
[30-Jul-2010 13:56:29] that works too
[30-Jul-2010 13:56:39] that would be a cool pet to say you have..
[30-Jul-2010 13:56:39] Zenoss actually adopted some real Zebras
[30-Jul-2010 13:56:53] crazy for why? marketing?
[30-Jul-2010 13:56:55] hmmm rmatte all o that checks out
[30-Jul-2010 13:56:58] marketing
[30-Jul-2010 13:57:28] MrWGW, I don't know, it's really hard for me to troubleshoot something like this remotely when you're just starting out... perhaps your should do it with SNMP instead?
[30-Jul-2010 13:58:06] i want also 2 zennies ;)
[30-Jul-2010 13:58:09] that is a possibility
[30-Jul-2010 13:58:13] wow.. the more i use 3.0 the more little buggies i see i should start documenting them
[30-Jul-2010 13:58:33] citrus2: http://dev.zenoss.com/trac login with zenoss/zenoss to open bug tickets
[30-Jul-2010 13:58:42] though quite a few have already been reported
[30-Jul-2010 13:58:48] I still have a bunch left to log
[30-Jul-2010 13:59:04] bigegor_ is now known as bigegor
[30-Jul-2010 13:59:15] cool thanks
[30-Jul-2010 13:59:39] the weird thing about this is that I have setup ssh monitoring before
[30-Jul-2010 13:59:43] and indeed had no problems at all
[30-Jul-2010 13:59:59] Well, new releases always seem to end up a little buggy
[30-Jul-2010 14:00:00] I'm going to test add a server running a different OS that I'm successfully monitoring using a diferent zenoss VM
[30-Jul-2010 14:00:04] and see if that makes a diference
[30-Jul-2010 14:00:13] MrWGW: good idea
[30-Jul-2010 14:01:09] * rmatte wants to shoot whoever thought it was a good idea to only give our core sql server a 40GB vdisk
[30-Jul-2010 14:01:38] I load up a test database, run a test upgrade on it, and voila, only 2.4GB of disk space left
[30-Jul-2010 14:01:51] now I'm working on clearing the database records
[30-Jul-2010 14:03:10] rmatte, can't you expand the disk?
[30-Jul-2010 14:03:37] rmatte: what hypervisor?
[30-Jul-2010 14:03:46] jmp242: yeh, but it has production databases on it
[30-Jul-2010 14:03:56] jmp242: so that can only be done next week during the change window
[30-Jul-2010 14:04:14] Oh fun.
[30-Jul-2010 14:04:18] yup
[30-Jul-2010 14:05:04] MrWGW: ESX
[30-Jul-2010 14:05:31] Lucky you
[30-Jul-2010 14:05:38] well you can at least enlarge the .vmdk
[30-Jul-2010 14:05:40] and then grow the FS
[30-Jul-2010 14:05:45] yes
[30-Jul-2010 14:05:54] I'm currently fighting with KVM trying to get it to boot an XP guest :(
[30-Jul-2010 14:05:56] it won't be much work
[30-Jul-2010 14:06:06] jmp242: fun
[30-Jul-2010 14:06:17] also fire whoever gave it the undersized disk
[30-Jul-2010 14:06:27] That's probably my manager who did that
[30-Jul-2010 14:06:28] :P
[30-Jul-2010 14:06:28] oh fun fact: ESX 4.0 and better for some OSes supports memory and CPU hot add
[30-Jul-2010 14:06:32] oh well then
[30-Jul-2010 14:06:36] guess you won't be firing him
[30-Jul-2010 14:06:37] Yea - apparently there's a bug for 2k/XP NTFS disk geometry
[30-Jul-2010 14:06:51] MrWGW: Yes, we upgraded all the VM versions to support the hot add
[30-Jul-2010 14:06:57] so I have to hex edit the disk (somehow) and change the hex at a specific offset
[30-Jul-2010 14:07:14] ok well
[30-Jul-2010 14:07:19] I don't know how anyone missed this in a release - as it affects the Server line too :(
[30-Jul-2010 14:07:22] the system didn't collect the correct information in this case
[30-Jul-2010 14:07:28] jmp242: lame
[30-Jul-2010 14:07:32] where I tested it against the server that the other box was successfully polling
[30-Jul-2010 14:07:44] so that means there has to be something fucked up on this zenoss server
[30-Jul-2010 14:07:48] MrWGW: configuration issue somewhere then
[30-Jul-2010 14:07:51] pardon my anglo saxon
[30-Jul-2010 14:08:08] so it's modelling the disks initially?
[30-Jul-2010 14:08:14] but not actually monitoring them afterwards?
[30-Jul-2010 14:08:14] the weird thing is that when I intiially configured the two servers I did it in the same manner
[30-Jul-2010 14:08:23] that's what it looks like
[30-Jul-2010 14:08:28] and it never gets the used bytes
[30-Jul-2010 14:08:29] strange
[30-Jul-2010 14:08:48] I'm rather ill today, I'm very tempted to pay someone to just go in there and fix it and tell me where I screwed up
[30-Jul-2010 14:08:52] yeh, it sounds like the modeller plugins are working fine but not the collector plugins
[30-Jul-2010 14:09:04] should I perhaps restart zenoss?
[30-Jul-2010 14:09:05] lol
[30-Jul-2010 14:09:10] worth a shot
[30-Jul-2010 14:09:13] oh wait
[30-Jul-2010 14:09:21] you installed that ZenPack but didn't restart Zenoss afterwards?
[30-Jul-2010 14:09:25] no
[30-Jul-2010 14:09:34] that could be my moment of epic n00b fail
[30-Jul-2010 14:09:38] I installed it via the web GUI
[30-Jul-2010 14:09:46] yeh, restart zenoss for sure man
[30-Jul-2010 14:09:55] oh man
[30-Jul-2010 14:10:01] for very basic packs with just performance templates you don't usually need a restart
[30-Jul-2010 14:10:01] it all comes down to a reboot
[30-Jul-2010 14:10:12] but for things like collector plugins, daemons, and whatever else, you do
[30-Jul-2010 14:10:27] ok
[30-Jul-2010 14:10:30] so that was me being a n00b
[30-Jul-2010 14:10:35] indeed
[30-Jul-2010 14:10:40] fortunately I didn't have to pay anyone to clean up my mess in this instance
[30-Jul-2010 14:10:55] that would have hurt to pay someone to reboot it
[30-Jul-2010 14:10:58] If you'd go to the ZenPacks section in the UI you'd see a message complaining that the pack is broken
[30-Jul-2010 14:11:04] meaning you need to restart Zenoss
[30-Jul-2010 14:11:06] well it would've taught me a lesson
[30-Jul-2010 14:11:08] note by the way
[30-Jul-2010 14:11:20] I am an independent businessman and I do managed hosting
[30-Jul-2010 14:11:24] so yes, I'm one of those morons
[30-Jul-2010 14:11:35] but basically when I get stuck on something, if I don't have time to fix it myself I do hire people to do it
[30-Jul-2010 14:11:48] however usually I fix stuf myself
[30-Jul-2010 14:11:56] :)
[30-Jul-2010 14:11:56] I am also really good at fixing servers that are down
[30-Jul-2010 14:12:08] and I know how to do BGP routing and I have my own IP addresses and ASNs
[30-Jul-2010 14:12:11] well if you need help rebooting anything you can hire me :)
[30-Jul-2010 14:12:27] actually right now I'm looking for a relief network engineer
[30-Jul-2010 14:12:41] I have to travel to Europe in a few months and I'm looking for a guy to monitor the networks while I'm in the air
[30-Jul-2010 14:13:09] primarily a Cisco guy
[30-Jul-2010 14:13:21] and preferrably someone in the Los Angeles area
[30-Jul-2010 14:13:24] yeah i only ever got my ccna
[30-Jul-2010 14:13:29] not very far in cisco stuff
[30-Jul-2010 14:13:37] i'm in santa barbara!
[30-Jul-2010 14:13:38] hehe
[30-Jul-2010 14:13:47] I myself have no certs but I estimate I have CCNP-level knowledge in some areas
[30-Jul-2010 14:13:55] I know a lot about BGP but less about the ibgps
[30-Jul-2010 14:14:04] other than OSPF, which I use internally
[30-Jul-2010 14:14:12] my main area of weakness in Cisco though is QoS
[30-Jul-2010 14:14:17] I shoudl not have told zenoss to email me all events debug and above lol
[30-Jul-2010 14:14:30] mducharme-laptop: ah you must've been pounded there haha
[30-Jul-2010 14:14:45] I just thought I would do that for a few seconds to see if it worked
[30-Jul-2010 14:14:51] big mistake
[30-Jul-2010 14:14:52] lol
[30-Jul-2010 14:15:00] yup
[30-Jul-2010 14:15:10] in those few seconds it sent me like 8,000 emails or something
[30-Jul-2010 14:15:24] holy balls
[30-Jul-2010 14:15:43] there's a lot of debug stuff when you first add devices and they get modeled
[30-Jul-2010 14:16:20] alright I restarted zenoss
[30-Jul-2010 14:16:24] should I now remodel the deices?
[30-Jul-2010 14:16:43] zen modeler should run after a restart shouldn't it?
[30-Jul-2010 14:16:48] note by the way that I did actually get increased information after I installed the zenpack
[30-Jul-2010 14:17:10] I am now getting performance graphs
[30-Jul-2010 14:17:19] but I'm still not getting filesystem usage information
[30-Jul-2010 14:17:20] woot!
[30-Jul-2010 14:17:28] hrm
[30-Jul-2010 14:17:36] is it possible to show the interface description somehow in the event email
[30-Jul-2010 14:17:37] citrus2: blogs/zenossblog/2009/07/30/join-the-zenoss-herd
[30-Jul-2010 14:17:58] wow, that's weird
[30-Jul-2010 14:18:05] I just noticed that was posted exactly 1 year ago
[30-Jul-2010 14:18:06] lol
[30-Jul-2010 14:18:09] perhaps the output of the df -h command on Ubuntu is formatted diferently?
[30-Jul-2010 14:18:17] seems like it wasn't that long
[30-Jul-2010 14:18:27] b/c otherwise I cannot tell how important this is
[30-Jul-2010 14:18:31] MrWGW: highly doubt it
[30-Jul-2010 14:18:37] interface 0/22 on some switch is down
[30-Jul-2010 14:18:56] mducharme-laptop: you'd have to write a transform to do that
[30-Jul-2010 14:19:00] but I don't know what the description on that interface is to know how important that is.. there must be some way of pulling the description into the email
[30-Jul-2010 14:19:00] ok
[30-Jul-2010 14:19:14] is the default for messages sent out for load average supposed to be in the thousands? I get warning messages stating current value = 1350.00, im assuming it means 13.50. I changed the datapoint to 10000,/ because it was simply set to 100,/ but my messages are still coming through with the number in thousands rather than tens.
[30-Jul-2010 14:19:28] mducharme-laptop: let me show you a transform that I use to do monitoring based on interface description
[30-Jul-2010 14:19:38] ok
[30-Jul-2010 14:20:41] mducharme-laptop: this http://fpaste.org/uw0Q/ is designed to be used with this: docs/DOC-2494
[30-Jul-2010 14:20:41] I am getting more information right now
[30-Jul-2010 14:20:47] in fact I literally have everything except disk space usage
[30-Jul-2010 14:21:14] mducharme-laptop: if the interface description starts with SVxx, UVxx, etc... it won't generate an alert
[30-Jul-2010 14:21:23] it'll just drop it
[30-Jul-2010 14:21:25] ok
[30-Jul-2010 14:21:35] I just want it to tell me the description
[30-Jul-2010 14:21:53] in the alert email
[30-Jul-2010 14:21:56] right, the code that you'll need is in there
[30-Jul-2010 14:22:03] actually I checked the test system I added
[30-Jul-2010 14:22:04] but hold up
[30-Jul-2010 14:22:08] and it is also not getting filesystem usage information
[30-Jul-2010 14:22:11] how are you monitoring the actual interfaces?
[30-Jul-2010 14:22:15] whereas it is in the other zenoss install
[30-Jul-2010 14:22:37] are you doing this docs/DOC-2494 or something else?
[30-Jul-2010 14:23:04] yeah I'm doing that rmatte
[30-Jul-2010 14:23:16] MrWGW: not sure, there's obviously something different, compare zProperties, compare command plugins, etc... between the 2 instances
[30-Jul-2010 14:23:43] mducharme-laptop: ok, so getting the description will be easy...
[30-Jul-2010 14:23:57] Hi, I'm getting a weird behavior when I do: dmd = ZenScriptBase(connect=True).dmd
[30-Jul-2010 14:24:35] my "getopt.getopt" trigger an exception
[30-Jul-2010 14:24:40] i get weird behavior when i tell me gf shes fat
[30-Jul-2010 14:25:00] LOL
[30-Jul-2010 14:25:06] mducharme-laptop: something like: http://fpaste.org/BWgG/
[30-Jul-2010 14:25:25] erm oops
[30-Jul-2010 14:25:36] http://fpaste.org/1Bwd/
[30-Jul-2010 14:25:38] that rather
[30-Jul-2010 14:25:45] then the message will be set to the description
[30-Jul-2010 14:25:53] you can add other stuff to the message too if you want
[30-Jul-2010 14:26:03] then make sure the message is included in the alert
[30-Jul-2010 14:26:58] hrm
[30-Jul-2010 14:27:25] the zproprierties, command plugins etc at the device level are identical
[30-Jul-2010 14:27:27] you know python, so I assume you can figure it out
[30-Jul-2010 14:27:31] I would ideally have it in the summary
[30-Jul-2010 14:27:39] rmatte, did you have that error anytime when you use dmd?
[30-Jul-2010 14:27:44] MrWGW: There has to be some difference there somewhere
[30-Jul-2010 14:27:55] chachan: which error?
[30-Jul-2010 14:28:14] chachan: pastebin the actual error
[30-Jul-2010 14:28:18] the two systems were both installed from the same common VMware virtual appliance
[30-Jul-2010 14:28:42] the only difference is that the former runs on ESXi and the latter (malfunctioning) one on VMware Server
[30-Jul-2010 14:28:49] but in both cases it was deployed via VMware Cnverter
[30-Jul-2010 14:28:51] MrWGW: I'm just stating the obvious here, one works, one doesn't, something is different
[30-Jul-2010 14:28:54] and that shouldn't amount to squat
[30-Jul-2010 14:28:55] right
[30-Jul-2010 14:28:56] whether they came from the same machine or not
[30-Jul-2010 14:28:57] I get that
[30-Jul-2010 14:29:02] rmatte will that error out if the interface has no description?
[30-Jul-2010 14:29:03] let me check one other thing
[30-Jul-2010 14:29:14] let me try and add the screwed up server to the old system that is working
[30-Jul-2010 14:29:16] mducharme-laptop: no, it'll just set the message to ''
[30-Jul-2010 14:29:26] becasue if that works then that will be illucidating if nothing else
[30-Jul-2010 14:29:48] mducharme-laptop: add an: if interface.description: if you want
[30-Jul-2010 14:29:54] to be sure
[30-Jul-2010 14:30:49] chachan: I also need context, where are you doing that?
[30-Jul-2010 14:30:56] in a script?
[30-Jul-2010 14:31:03] if so, what user is executing the script?
[30-Jul-2010 14:31:26] yes, I'm trying to make it a little short
[30-Jul-2010 14:31:34] zenoss user
[30-Jul-2010 14:31:42] a little short?
[30-Jul-2010 14:31:56] yes, to not show you the whole script
[30-Jul-2010 14:32:12] docs/DOC-3217;jsessionid=80D27122402B508FC8040FBA903347A4.node0
[30-Jul-2010 14:32:13] small?
[30-Jul-2010 14:32:17] you know you need all of that right?
[30-Jul-2010 14:32:28] there's no way you'll get it to work otherwise
[30-Jul-2010 14:32:36] you can't just have: dmd = ZenScriptBase(connect=True).dmd
[30-Jul-2010 14:32:44] because it needs ZenScriptBase
[30-Jul-2010 14:33:00] you also need to import commit
[30-Jul-2010 14:33:05] and Globals
[30-Jul-2010 14:33:14] rmatte, yes, I do all those imports
[30-Jul-2010 14:33:15] because it makes use of those
[30-Jul-2010 14:33:35] show me the whole error you're seeing
[30-Jul-2010 14:33:40] fpaste.org it
[30-Jul-2010 14:33:50] also fpaste.org the script
[30-Jul-2010 14:33:59] yes, I'm on it
[30-Jul-2010 14:34:21] just to double check, you're running the script as the zenoss user, and zenoss is actually running at the time, correct?
[30-Jul-2010 14:34:27] rmatte
[30-Jul-2010 14:34:29] http://fpaste.org/CKaP/
[30-Jul-2010 14:34:37] will that work like I think it will?
[30-Jul-2010 14:35:40] mducharme-laptop: looks fine
[30-Jul-2010 14:35:56] well
[30-Jul-2010 14:36:02] you might want to add some spaces...
[30-Jul-2010 14:36:08] "interface" + interfaceinfo + "operationally up"
[30-Jul-2010 14:36:20] won't that turn out like interfaceblah blah blahoperationally up
[30-Jul-2010 14:36:21] ?
[30-Jul-2010 14:36:55] rmatte: when editing the datapoints where 100,/ is set, if I simply change to 10000,/ <-- this is sufficient to get it to produce correct numbers for load avg right? RIght now my 5min loadavg is showing 1350.00 when it should be 13.50
[30-Jul-2010 14:37:16] I've edited the data points to reflect this but its not showing up in the emails with the new number
[30-Jul-2010 14:38:04] QubeZ: yeh, that should work
[30-Jul-2010 14:38:15] QubeZ: actually sorry
[30-Jul-2010 14:38:24] rmatte, http://fpaste.org/bOM9/
[30-Jul-2010 14:38:27] QubeZ: that will only change the value displayed on the graph itself
[30-Jul-2010 14:38:36] QubeZ: to change it in the alerts you need to write a transform
[30-Jul-2010 14:38:43] Hello. I added a user. It sends out 3 emails with a password but none of them work. I am using zenoss-3.0.0.tar.gz on Gentoo. What can I do?
[30-Jul-2010 14:38:51] oops rmatte, good point
[30-Jul-2010 14:38:58] QubeZ: RPN is just for converting on the fly for graphs
[30-Jul-2010 14:38:58] :)
[30-Jul-2010 14:39:08] zaoul, check if you server can send mails first
[30-Jul-2010 14:39:54] Well, I'm receiving the mails
[30-Jul-2010 14:40:12] chachan: I see what your issue is...
[30-Jul-2010 14:40:16] So it seems the server can send them.
[30-Jul-2010 14:40:20] really???????
[30-Jul-2010 14:40:26] chachan: you're doing the actual #
[30-Jul-2010 14:40:28] dmd = ZenScriptBase(connect=True).dmd
[30-Jul-2010 14:40:34] ok
[30-Jul-2010 14:40:38] right in the classes/functions
[30-Jul-2010 14:40:46] but those classes don't even know what ZenScriptBase is
[30-Jul-2010 14:41:01] I assume that's why it's complaining
[30-Jul-2010 14:41:14] try doing #
[30-Jul-2010 14:41:15] dmd = ZenScriptBase(connect=True).dmd
[30-Jul-2010 14:41:19] right at the top of the script
[30-Jul-2010 14:41:22] under the imports
[30-Jul-2010 14:41:30] and forget about setting it in the functions themselves
[30-Jul-2010 14:41:52] rmatte, mmm
[30-Jul-2010 14:41:54] let me see
[30-Jul-2010 14:41:55] you only have to do it once
[30-Jul-2010 14:42:00] it doesn't need to be done multiple times
[30-Jul-2010 14:42:15] as far as I recall
[30-Jul-2010 14:42:20] rmatte: but the graphs are right, it shows on the left processes (12) and each time my alert goes off its because my loadavg went over 12 but the emails shows 1200
[30-Jul-2010 14:42:35] i can see why because snmp reports it as an integer 1200 rather than 12.00
[30-Jul-2010 14:42:36] More specifically when I create the user, no password is sent out at all. but when I reset password as admin. It sends 3 emails
[30-Jul-2010 14:42:36] QubeZ: yes, I just explained why that is
[30-Jul-2010 14:42:46] hrm rmatte
[30-Jul-2010 14:42:48] that isn't working
[30-Jul-2010 14:42:51] QubeZ: SNMP is reporting 1200, the RPN value is changing that on the fly when graphing
[30-Jul-2010 14:42:57] now it is printing threshold of operational satus exceeded
[30-Jul-2010 14:43:01] but when the threshold triggers it displays the raw value in the alert
[30-Jul-2010 14:43:04] like the transform isn't running at all
[30-Jul-2010 14:43:11] which is why you need to write a transform to convert that value
[30-Jul-2010 14:43:14] rmatte: ok thanks
[30-Jul-2010 14:43:30] mducharme-laptop: did the code turn red when you saved it?
[30-Jul-2010 14:43:37] yes
[30-Jul-2010 14:43:45] mducharme-laptop: then you have a syntax error somewhere
[30-Jul-2010 14:43:49] look carefully and fix it
[30-Jul-2010 14:44:03] it could be as simple as incorrect white space
[30-Jul-2010 14:44:11] mducharme-laptop: pastebin it to me again
[30-Jul-2010 14:44:13] oh yeah
[30-Jul-2010 14:44:15] found it
[30-Jul-2010 14:44:15] I'll have a look
[30-Jul-2010 14:44:17] k
[30-Jul-2010 14:44:20] I didn't indent after the for loop
[30-Jul-2010 14:44:25] ;)
[30-Jul-2010 14:44:41] if the code turns red, that's bad
[30-Jul-2010 14:44:45] remember that
[30-Jul-2010 14:44:46] :)
[30-Jul-2010 14:46:05] zaoul: ok, now for your issue...
[30-Jul-2010 14:46:12] cool that works!
[30-Jul-2010 14:46:19] I haven't quite wrapped my head around it...
[30-Jul-2010 14:46:27] interface suid77 (benice sanderson) is operationally down
[30-Jul-2010 14:46:28] you created a user, now what's this talk about emails with passwords?
[30-Jul-2010 14:46:28] rmatte, nothing :(
[30-Jul-2010 14:46:43] woohoo
[30-Jul-2010 14:46:52] chachan: you didn't pastebin the error
[30-Jul-2010 14:46:56] or I missed it if you did
[30-Jul-2010 14:47:10] I need to see the exact error
[30-Jul-2010 14:47:12] the full trace
[30-Jul-2010 14:47:21] rmatte, http://fpaste.org/RWNV/
[30-Jul-2010 14:47:25] mducharme-laptop: nice
[30-Jul-2010 14:47:28] we're reviewing potential defects to be included in 3.0.2; open up your favorites if you haven't yet :)
[30-Jul-2010 14:47:32] that's all what it returns me
[30-Jul-2010 14:47:36] that's weird
[30-Jul-2010 14:48:08] chachan: ummm, that's just bad coding on your part from the looks of it :P
[30-Jul-2010 14:48:36] oh
[30-Jul-2010 14:48:42] now all that I have to do is make some interfaces less important than others
[30-Jul-2010 14:48:46] #
[30-Jul-2010 14:48:46] event = {'device':d['--device'].replace('-service', '', 1), 'component':url,
[30-Jul-2010 14:48:46] #
[30-Jul-2010 14:48:46] 'summary':"["+ d['--device'].replace('-service', '', 1) +"]" + " [Web Service] " + message,
[30-Jul-2010 14:48:46] #
[30-Jul-2010 14:48:46] 'eventClass':'/Cmd/Fail', 'severity':"3"}
[30-Jul-2010 14:48:51] because I do not care whether bernice sanderson is down or not
[30-Jul-2010 14:48:52] lol
[30-Jul-2010 14:48:53] perhaps -device instead of --device?
[30-Jul-2010 14:49:19] the zenoss blog is the only way in which i was thanked for sysadmin day :(
[30-Jul-2010 14:49:25] hey guys
[30-Jul-2010 14:49:30] actually I'm getting filesystem usage information now
[30-Jul-2010 14:49:33] it just took it awhile
[30-Jul-2010 14:49:45] chachan: hmmm nope, looks like it's hitting the end of your script
[30-Jul-2010 14:49:52] so the only thing I'm not getting right now that I need is monitoring of memory
[30-Jul-2010 14:49:53] it's weird, because it works without dmd, I'll try anyway
[30-Jul-2010 14:49:55] chachan: check your code, something is funky for sure
[30-Jul-2010 14:49:56] I'm getting CPU performance graphs
[30-Jul-2010 14:50:03] and I'm getting filesystem usage information, ut I don't see anything about memory
[30-Jul-2010 14:50:08] other than under the hardware tab, it says how much I have
[30-Jul-2010 14:50:32] Ok, I will fix the arguments by hand
[30-Jul-2010 14:50:41] zenoss.cmd.linux.memory is in the collector plugins list however
[30-Jul-2010 14:50:46] MrWGW: It sounds like you didn't push changes to the collector so it's taking it a while to figure out what it should be monitoring
[30-Jul-2010 14:50:52] go to the device page for the device
[30-Jul-2010 14:50:53] avoiding "getopt.getopt"
[30-Jul-2010 14:51:00] and do: manage -> push changes
[30-Jul-2010 14:51:34] and then presumably wait like 15 minutes for it to poll everything?
[30-Jul-2010 14:51:48] I guess the push changes would be needed after remodelling the device following installing the ssh monitoring plugin?
[30-Jul-2010 14:53:10] well, this is why you push changes...
[30-Jul-2010 14:53:23] the collector only comes back and asks zenhub to tell it what it should be monitoring every hours
[30-Jul-2010 14:53:25] hour*
[30-Jul-2010 14:53:30] so you either wait the hour for it to do that
[30-Jul-2010 14:53:31] how do I not get it to say "clear" in the email when it comes back up again
[30-Jul-2010 14:53:35] or you force it to push the changes
[30-Jul-2010 14:53:50] because it's somewhat redundant to say CLEAR: device interface (description) is operationally UP
[30-Jul-2010 14:53:50] so remember to push changes when you change the way a device is monitored or add a new device
[30-Jul-2010 14:53:58] ok good stuff
[30-Jul-2010 14:54:07] now I'm not yet seeing any performance graph on memory utilization
[30-Jul-2010 14:54:13] mducharme-laptop: that's easy, just modify the email template in the alerting rule
[30-Jul-2010 14:54:15] the only perormance graph I have is on cpu utilization
[30-Jul-2010 14:54:21] I take it I sho uld just wait like 15 minutes or so?
[30-Jul-2010 14:55:13] mducharme-laptop: go to the alerting rule itself, then go to the message tab
[30-Jul-2010 14:55:22] ahh cool, thanks
[30-Jul-2010 14:55:27] np
[30-Jul-2010 14:55:35] you can change the formatting of the messages there too
[30-Jul-2010 14:55:47] rmatte, I created a user... I receive no password by email... As admin I reset password. I receive 3 emails from zenoss with passwords but none of them work.
[30-Jul-2010 14:55:50] %(whatever)s can be added to show any event properties
[30-Jul-2010 14:56:01] cool.. I think what is there is good enough
[30-Jul-2010 14:56:03] rmatte, 3 passwords/ 3 emails
[30-Jul-2010 14:56:06] for now anyhow
[30-Jul-2010 14:56:07] 3 emails with one click
[30-Jul-2010 14:56:18] zaoul: why don't you just explicitly set a password for the user rather than "resetting" it?
[30-Jul-2010 14:56:59] go to the user, type in the new password, confirm password, then fill out the admin password in the password confirmation at the bottom and click save
[30-Jul-2010 14:59:30] zaoul: basically, you're clicking the wrong button
[30-Jul-2010 14:59:36] you should be clicking save, not reset password
[30-Jul-2010 14:59:53] now I just need to figure out the best way to determine the importance of interfaces
[30-Jul-2010 15:00:00] Ah, Yes that works!.. more so I was confused by it wanted the admin password
[30-Jul-2010 15:00:01] to specify their importance
[30-Jul-2010 15:00:07] some interfaces it doesn't matter if they are up or down
[30-Jul-2010 15:00:12] others it does
[30-Jul-2010 15:00:18] mducharme-laptop: yeh, that's one thing they need to add as a feature, the ability to "tag" objects
[30-Jul-2010 15:00:19] rmatte, but still It seems there is a problem with it does not send a password out when the user is created
[30-Jul-2010 15:00:27] which could be used for severity or type
[30-Jul-2010 15:00:27] rmatte, and when resetting the password it sends three emails
[30-Jul-2010 15:00:40] b/c we are an isp and we have home users and important enterprise customers
[30-Jul-2010 15:00:50] if health canada is down we want to know right away
[30-Jul-2010 15:00:50] zaoul: it's not supposed to send the password out, do that yourself
[30-Jul-2010 15:00:57] but I do not care whether bernice sanderson is down
[30-Jul-2010 15:01:05] if she does she will phone us
[30-Jul-2010 15:01:06] lol
[30-Jul-2010 15:01:11] zaoul: you're the one setting the password so you obviously know what it is
[30-Jul-2010 15:01:37] mducharme-laptop: yeh, that's why we have interface description prefixes
[30-Jul-2010 15:01:39] actually on the subject of memory
[30-Jul-2010 15:01:44] even on my system that was otesnily working normally
[30-Jul-2010 15:01:45] mducharme-laptop: for instance, UVxx is a user port
[30-Jul-2010 15:01:48] there are no memory performance graphs
[30-Jul-2010 15:01:52] whereas IOxx is an infrastructure port
[30-Jul-2010 15:02:00] rmatte yeah the dificulty here is the interface descriptions can only be set in the SU's
[30-Jul-2010 15:02:07] andunder status, in the component type box
[30-Jul-2010 15:02:10] there's no entry for memory
[30-Jul-2010 15:02:11] mducharme-laptop: ah
[30-Jul-2010 15:02:16] so I think I need to tell it to monitor memory more closely
[30-Jul-2010 15:02:21] is that possible with ssh monitoring?
[30-Jul-2010 15:02:28] or would I need to go snmp for that?
[30-Jul-2010 15:02:35] there is no ability to set the description at the AP side, you can only read the description
[30-Jul-2010 15:02:47] so we would need to manually telnet into several hundred ip's to set descriptions
[30-Jul-2010 15:03:02] MrWGW: do not worry about that component box
[30-Jul-2010 15:03:11] MrWGW: the values you see in there are completely hardcoded
[30-Jul-2010 15:03:24] it doesn't mean something isn't monitored just because it doesn't show up as green in there
[30-Jul-2010 15:03:26] so what about adding a perormance graph for memory?
[30-Jul-2010 15:03:46] I think memory isn't supported via SSH, we ran in to this same stupidity the other day with someone else
[30-Jul-2010 15:03:59] it's supposed to support it, but it doesn't
[30-Jul-2010 15:04:00] really man, SNMPD is much much more flexible
[30-Jul-2010 15:04:06] SNMP*
[30-Jul-2010 15:04:06] oh this is bitterly ironic, I finally got it working and it can't monitor memory :-p
[30-Jul-2010 15:04:20] I monitor all my linux boxes via SNMP and it works fine
[30-Jul-2010 15:04:30] well I can do that in this case
[30-Jul-2010 15:05:22] the only question is, could I perhaps add a performance graph for memory by telling it to poll the "free" command?
[30-Jul-2010 15:05:26] because that's all it needs to do really
[30-Jul-2010 15:05:33] and I was under the impression that zenoss was somewhat extensible
[30-Jul-2010 15:05:47] It's definitely possible
[30-Jul-2010 15:05:54] read up on configuring your own SSH commands
[30-Jul-2010 15:05:59] and building performance templates
[30-Jul-2010 15:06:06] the question is, would it be easier to do that vs. just switching to snmp?
[30-Jul-2010 15:06:11] it's a pretty basic skill to have when using Zenoss, so you'll need to learn it sooner or later
[30-Jul-2010 15:06:22] easier to switch to snmp
[30-Jul-2010 15:06:34] plus snmp can do process monitoring properly as well
[30-Jul-2010 15:06:42] and it monitors other stuff like disk IO
[30-Jul-2010 15:06:54] can I just push out a standard snmpd.conf to all my servers, change the device class, remodel and push changes?
[30-Jul-2010 15:07:00] and of course set the snmp community string
[30-Jul-2010 15:07:01] yes
[30-Jul-2010 15:07:09] I pasted you an snmpd.conf file to use
[30-Jul-2010 15:07:11] twice
[30-Jul-2010 15:07:14] oh did you
[30-Jul-2010 15:07:19] yes
[30-Jul-2010 15:07:20] much to my chagrin I missed that
[30-Jul-2010 15:07:30] I have a nasty cold this morning
[30-Jul-2010 15:07:34] so let me sroll up
[30-Jul-2010 15:08:23] I'm scrolling up trying to find it
[30-Jul-2010 15:08:33] rmatte: if you could repaste it I would be almost infintiely obliged
[30-Jul-2010 15:08:54] I would also like to monitor mysql btw
[30-Jul-2010 15:09:00] I found a zenpak called mysql ssh monitor
[30-Jul-2010 15:10:30] I really have to get these dependencies working somehow
[30-Jul-2010 15:10:49] I just got 3 emails, that mooselake router, mooselake switch, and mooselake ap are down.. mooselake router is the only important one
[30-Jul-2010 15:11:03] you can still use the MySQL monitor in conjunction with SNMP monitoring
[30-Jul-2010 15:11:22] the mysql monitoring would require ssh in either case presumably?
[30-Jul-2010 15:11:26] MrWGW: eugh, I'll just repaste the config, I can't find it for the life of me
[30-Jul-2010 15:11:40] rmatte: I'm almost darn certain you didn't paste it, becasue I've been scrolling for it also
[30-Jul-2010 15:11:55] if its any consolation, I once imagined an entire cab of servers existing in a datacenter
[30-Jul-2010 15:11:58] that actually didn't
[30-Jul-2010 15:13:51] I pasted it way back but it's beyond by scrollback buffer
[30-Jul-2010 15:13:54] I know I pasted it twice
[30-Jul-2010 15:14:52] here
[30-Jul-2010 15:14:54] http://pastebin.com/kqwaTy50
[30-Jul-2010 15:16:46] I would assume I don 'tneed to mess with snmptrapd on the servers being monitored?
[30-Jul-2010 15:17:57] you need to configure the trap lines in the snmpd.conf file
[30-Jul-2010 15:18:04] if you want traps
[30-Jul-2010 15:18:18] here are the lines that need to be changed...
[30-Jul-2010 15:18:40] trapsink 192.168.0.1 your-key
[30-Jul-2010 15:18:41] on the rwcommunity line, shouldn't that be commented out? SNMP read write is widely considered to be dangerous IIRC
[30-Jul-2010 15:18:43] trap2sink 192.168.0.1 your-key
[30-Jul-2010 15:18:49] rocommunity your-key
[30-Jul-2010 15:18:52] rwcommunity your-key
[30-Jul-2010 15:18:55] syslocation "Set this to the location"
[30-Jul-2010 15:18:59] syscontact whoever@contact.com
[30-Jul-2010 15:19:01] that's it
[30-Jul-2010 15:19:20] you can comment out the rw line if you'd like
[30-Jul-2010 15:19:22] it's up to you
[30-Jul-2010 15:19:23] ttyl guys
[30-Jul-2010 15:19:54] rw is dangerous unless you're actually using it for something valid
[30-Jul-2010 15:20:00] and you have a strong community string
[30-Jul-2010 15:20:09] it's the idiots who leave it as "private" that get nailed
[30-Jul-2010 15:20:36] given it's not encrypted it's a dumb idea to run rw without v3 encryption tbh
[30-Jul-2010 15:20:36] rmatte: around?
[30-Jul-2010 15:20:47] twm1010: yessir
[30-Jul-2010 15:20:54] rmatte: can i pick your brain on something?
[30-Jul-2010 15:21:05] Simon4: yeh, though on an secure internal network it's not that bad
[30-Jul-2010 15:21:17] Simon4: it would obviously be stupid to run it over the internet unencrypted
[30-Jul-2010 15:21:20] rmatte: in my world there's no such thing :)
[30-Jul-2010 15:21:27] I'd like to create a multi-graph report, that would show, for each site, bandwidth utilization, one graph, per site, with every circuit at the site on the graph
[30-Jul-2010 15:21:41] except, not every interface at a site, uses the same interface template right?
[30-Jul-2010 15:21:46] now on syscontact
[30-Jul-2010 15:21:46] Simon4: :P
[30-Jul-2010 15:21:56] does that have any bearing on how zenoss does alerting?
[30-Jul-2010 15:22:02] twm1010: correct
[30-Jul-2010 15:22:12] can i simply set the graph definition for both 64-bit and 32-bit and it will display only the relevant ones?
[30-Jul-2010 15:22:31] twm1010: I'm honestly not sure, I've never tried what you're attempting
[30-Jul-2010 15:22:51] MrWGW: none
[30-Jul-2010 15:23:03] MrWGW: that will just show up in the contact field on the device status page
[30-Jul-2010 15:23:11] you see why that would be useful right? a nice consolidated report by site, showing what the WAN links are doing
[30-Jul-2010 15:23:13] MrWGW: it doesn't affect monitoring/alerting
[30-Jul-2010 15:23:17] its the mix that makes it complicated in my brain
[30-Jul-2010 15:23:30] twm1010: oh definitely, I just don't know if it's possible without some development work
[30-Jul-2010 15:23:36] twm1010: you'll get "missing rrd file" for the ones that are missing... but then what network devices do you have which don't support 64 bit counters?
[30-Jul-2010 15:23:53] rmatte: I figured as much
[30-Jul-2010 15:24:01] even our cheapest ciscos support em
[30-Jul-2010 15:24:06] but I wanted to make sure that zenoss didn't do anything cute like automatically send alerts to the device contact
[30-Jul-2010 15:24:10] Simon4: well, the interface type decides whether to use which template ot use
[30-Jul-2010 15:24:19] Simon4: Wouldn't the question be "What network interfaces do you have that don't support 64bit counters"?
[30-Jul-2010 15:24:33] and the answer would be, a lot
[30-Jul-2010 15:24:42] rmatte: hmm
[30-Jul-2010 15:24:48] MrWGW: nope, nothing cute
[30-Jul-2010 15:24:55] a single point to point T1/Serial interface uses ethernetCSmacD
[30-Jul-2010 15:24:58] all our linux/bsd/cisco/juniper ones do
[30-Jul-2010 15:25:13] twm1010: ahh :)
[30-Jul-2010 15:25:20] but a multilink comprised of two will be a PPP_64 and default to ethernetCSmacD64
[30-Jul-2010 15:25:28] twm1010: you could cheat
[30-Jul-2010 15:25:35] cheating works :)
[30-Jul-2010 15:25:50] and make both templates have the same DS names, which means in a multi-graph report you could refer to the same datapoints
[30-Jul-2010 15:26:04] You know, I thought about that
[30-Jul-2010 15:26:06] vs inOctects and InHCOctects, which is what yo uget by default
[30-Jul-2010 15:26:18] under trapsink and trap2sink I presumably change the IP address to that of the zenoss box?
[30-Jul-2010 15:26:20] right, and then just go rename the existing RRD's and I'm gravy
[30-Jul-2010 15:26:27] I wonder why they chose to separate?
[30-Jul-2010 15:26:27] twm1010: yeah
[30-Jul-2010 15:26:40] it's "clearer" in a pure sense
[30-Jul-2010 15:26:46] but less useful in your world
[30-Jul-2010 15:27:12] maybe the RRD create for the two is a little different?
[30-Jul-2010 15:27:20] it's not
[30-Jul-2010 15:27:41] but it's certainly more technically correct to have the two different names
[30-Jul-2010 15:27:42] mray: any thoughts on that?
[30-Jul-2010 15:28:09] Simon4: maybe, but why? all you care about is what's the utilization?
[30-Jul-2010 15:28:20] they are separate because they use different datapoints
[30-Jul-2010 15:28:32] twm1010: I dunno - I'm just brainstorming reasons why it would be different :)
[30-Jul-2010 15:28:32] they use different sources, but store the same value
[30-Jul-2010 15:28:32] and it's just cleaner
[30-Jul-2010 15:28:44] though I admit the interface template system could be improved quite a bit
[30-Jul-2010 15:28:45] right, two different OID's but you're asking the same question
[30-Jul-2010 15:28:50] it's kind of stitched together right now
[30-Jul-2010 15:28:58] it should be more intelligent/dynamic
[30-Jul-2010 15:28:59] it's just the way it is.
[30-Jul-2010 15:29:09] but oh well, maybe down the road
[30-Jul-2010 15:29:53] hrmm...
[30-Jul-2010 15:29:58] twm1010: sorry, was working on something else. What's up?
[30-Jul-2010 15:30:21] * rmatte is afk for a few
[30-Jul-2010 15:30:21] was curious, why ethernetCSmacD and ethernetCSmacD uses two different datasource names?
[30-Jul-2010 15:30:23] twm1010: basically the datapoint name follows the mib description
[30-Jul-2010 15:30:37] err, ethernetCSmacD64
[30-Jul-2010 15:31:24] Ah... well perhaps I will adjust it then, I see why it was done then to match the MIB, but I really don't care what the MIB says. I want bandwidth to be a common datasource name across any interface so I can combine the graphs of anything together
[30-Jul-2010 15:32:00] twm1010: yeh, that'll work
[30-Jul-2010 15:32:09] and yeh, it is just a name
[30-Jul-2010 15:32:16] perhaps you should look at the data source alias feature.
[30-Jul-2010 15:32:32] cgibbons: does that work for multi-graph reports?
[30-Jul-2010 15:32:39] I thought that only worked for performance reports?
[30-Jul-2010 15:32:40] i think it was specifically created for such.
[30-Jul-2010 15:32:48] think, no first-hand knowledge
[30-Jul-2010 15:32:54] I'll take a look
[30-Jul-2010 15:32:57] * Simon4 has never met that feature
[30-Jul-2010 15:33:02] time for some more reading :)
[30-Jul-2010 15:33:26] well, I don't see anything about them in the admin guide
[30-Jul-2010 15:33:57] maybe it's just burried
[30-Jul-2010 15:34:33] where is the alias added tho?
[30-Jul-2010 15:34:38] looks like that was new as of 2.5.x
[30-Jul-2010 15:34:49] I see in a datasource where an alias is added, but not where to manipulate those
[30-Jul-2010 15:34:57] it was present in 2.4, but only worked with the CPU report
[30-Jul-2010 15:35:09] it was supposed to be supported in other performance reports by 2.5
[30-Jul-2010 15:35:38] I do see it here already tho
[30-Jul-2010 15:36:05] hold on, I'm finding the documentation for you
[30-Jul-2010 15:36:06] just wait
[30-Jul-2010 15:36:07] oh I see
[30-Jul-2010 15:36:22] the aliases for the 64 bit template is set to "inputOctets__bytes"
[30-Jul-2010 15:36:24] I've not used it, I just know it's intent is all about normalization :/
[30-Jul-2010 15:36:29] * Simon4 wonders if it's set for the 32 bit one
[30-Jul-2010 15:36:32] it is
[30-Jul-2010 15:36:34] i see it here
[30-Jul-2010 15:36:43] that might address my immediate concern, let's find out :)
[30-Jul-2010 15:36:44] let us know if htat works in a report
[30-Jul-2010 15:39:08] mhmmm
[30-Jul-2010 15:39:10] This report uses data point aliases. (For more information about data point aliases, see "Data Point Aliases" in the chapter titled Core Monitoring.)
[30-Jul-2010 15:39:22] yet there is no "Core Monitoring" chapter anywhere
[30-Jul-2010 15:39:24] It won't let me add the datasource
[30-Jul-2010 15:39:32] heh
[30-Jul-2010 15:39:43] it doesn't like the underscores
[30-Jul-2010 15:39:56] it's expecting datasource_datapoint
[30-Jul-2010 15:40:01] and the alias has two underscores
[30-Jul-2010 15:40:04] I found a similar stupidity the other day, I can't remember what it was, it referred to some non-existant section
[30-Jul-2010 15:40:14] someone really needs to go through and proofread the documentation
[30-Jul-2010 15:41:19] rmatte: we always take documentation tickets :)
[30-Jul-2010 15:41:27] i guess it's expecting there to be one underscore acting as a delimiter
[30-Jul-2010 15:41:50] mray: I don't have time to personally read through it end to end
[30-Jul-2010 15:42:04] no, I just meant when you notice things please point them out
[30-Jul-2010 15:42:20] mray: yeh, I forgot the last one, I'll remember to point this one out
[30-Jul-2010 15:42:24] our doc writer is pretty diligent about updating the contents, she's just not a subject matter expert
[30-Jul-2010 15:42:38] I suppose these aliases are used by the aggregate performance or bandwidth reports
[30-Jul-2010 15:43:08] twm1010: yeah, the aliases are used so you can compare apples and oranges, ie. WMI data with SSH data with SNMP data
[30-Jul-2010 15:44:36] aliases are in this doc: docs/DOC-3908
[30-Jul-2010 15:47:18] oh cool, and you cna put the RPN in there too, to manipulate it ahead of time before the graph definition
[30-Jul-2010 15:47:21] that will save me time too
[30-Jul-2010 15:51:15] Ah, but I'd still have to rename the datasource to get what I want
[30-Jul-2010 15:51:20] because the alias is on the datapoint
[30-Jul-2010 15:51:23] not the datasource
[30-Jul-2010 15:51:48] since the two templates have two different datasource names
[30-Jul-2010 15:51:50] blast!
[30-Jul-2010 15:52:56] in my data sources (WMI, WBEM, SQL) i have extend alias concept little bit. Value can be evaluated befor writing to RRD
[30-Jul-2010 15:54:38] bigegor: the stored value is fine, it's the naming
[30-Jul-2010 15:54:57] it sounds like the only way to get the multi-graph I want is to rename the datasource and datapoint in one template or the other
[30-Jul-2010 15:55:03] and then rename all my RRDs as well
[30-Jul-2010 15:56:07] That's alright, its not a must-have. I'd rather not deviate the standard template like that if I don't have to. It would have been nice though to consolidate some things onto a single graph per site
[30-Jul-2010 15:56:50] maybe a custom graph definition would do it
[30-Jul-2010 15:57:38] in fact, it'd be very manual, but that probably would do the trick
[30-Jul-2010 15:58:54] weekend time! take it easy everyone
[30-Jul-2010 15:59:00] seu
[30-Jul-2010 15:59:02] meh
[30-Jul-2010 15:59:08] agreed
[30-Jul-2010 16:10:19] rmatte: the only weird thing I'm experiencing is that snmpd is not answering external connection attempts; even though it isn't firewalled
[30-Jul-2010 16:11:10] MrWGW: are you on ubuntu/debian?
[30-Jul-2010 16:11:32] Simon4: in this case yes
[30-Jul-2010 16:11:37] snmpd is running btw
[30-Jul-2010 16:11:39] check /etc/default/snmpd
[30-Jul-2010 16:11:43] normally I use centos
[30-Jul-2010 16:11:52] I found yesterday that it bound to 127.0.0.1 only
[30-Jul-2010 16:12:03] or whatever is similar, etc
[30-Jul-2010 16:12:07] Simon4: don't you know, that's the problem
[30-Jul-2010 16:12:23] SNMPDOPTS='-Lsd -Lf /dev/null -u snmp -I -smux -p /var/run/snmpd.pid 127.0.0.1'
[30-Jul-2010 16:12:29] :)
[30-Jul-2010 16:12:54] I would assume I should delete the localhost address from that?
[30-Jul-2010 16:13:07] yeah, just remove the 127.0.0.1
[30-Jul-2010 16:13:11] I could alternately rtfm but since you already know...
[30-Jul-2010 16:13:36] it just happens I hit exactly the same issue last night
[30-Jul-2010 16:13:42] indeed
[30-Jul-2010 16:23:36] anyone have issues where you disable alerts, do some maintenance, clear out the events and re-enable alerts yet you still get emails?
[30-Jul-2010 20:37:49] mrayzenoss is now known as mray
[30-Jul-2010 21:45:43] I click on add -> Monitoring -> enable -> ok -> submit ... nothing happens, window pop in ins still displayed
[31-Jul-2010 00:00:30] [disconnected at Sat Jul 31 00:00:30 2010]
[31-Jul-2010 00:00:30] [connected at Sat Jul 31 00:00:30 2010]
[31-Jul-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]
[31-Jul-2010 02:59:13] fOrsberg is now known as forsberg
[31-Jul-2010 03:02:14] garr
[31-Jul-2010 03:02:43] had some arb issue now. zencommand works perfectly if testing it from the interface, fails on automatic runs
[31-Jul-2010 03:03:00] up the log level on zencommand to debug, and suddenly it starts working
[31-Jul-2010 03:03:03] pfoey :p
[31-Jul-2010 11:35:33] http://cache.gawker.com/assets/images/comment/9/2010/07/4bc9e697f0b95d372b0cc989a6481264/original.png
[31-Jul-2010 13:34:03] guys where can i get a good(easy for beginner) MIB documentation? i really couldnt find something for beginners that extends a little
[31-Jul-2010 14:15:40] JonnyV: i'd buy this http://oreilly.com/catalog/9780596000202
[31-Jul-2010 14:16:12] checking...
[31-Jul-2010 14:17:33] nice..
[31-Jul-2010 14:17:38] thanks rhettardo
[31-Jul-2010 14:18:05] if you dont mind reading that much, its a good book if you're getting started w/ SNMP
[31-Jul-2010 14:18:48] i dont personally know of any sites that explain is well for a beginner though
[31-Jul-2010 14:20:29] yep... its quite long but o'reilly books always are.. hehe...
[31-Jul-2010 14:21:13] and i been looking for a few days and didnt find anything... had to read quite a lot of sites to understand a little
[31-Jul-2010 14:51:16] forsberg is now known as fOrsberg
[31-Jul-2010 17:00:56] the web like alert icon means its an unknown class perhaps?
[31-Jul-2010 19:21:07] fOrsberg is now known as forsberg
[31-Jul-2010 21:09:24] forsberg is now known as fOrsberg
[31-Jul-2010 21:39:20] i has talk?
[31-Jul-2010 21:39:32] wrong window
[31-Jul-2010 21:40:11] <|rt|> heh
[31-Jul-2010 21:40:20] <|rt|> you has wrong window
[31-Jul-2010 21:41:54] |rt|: my mirc client freaked out, and i litearlly couldnt talk
[31-Jul-2010 21:42:04] i thought i was muted on freenode, but that wasnt the case
[31-Jul-2010 21:42:39] <|rt|> seems not