[01-Jan-2010 05:57:48] <python600rr> hello
[01-Jan-2010 05:58:06] <python600rr> is anybody up
[01-Jan-2010 06:43:41] <Dieterbe> no
[01-Jan-2010 08:35:08] <python600rr> I can't install the ZenPacks.community.AIX-1.1-py2.4.egg
[01-Jan-2010 09:16:49] <python600rr> never mind
[01-Jan-2010 18:21:29] <ke4qqq> rmatte: any chance you are on
[01-Jan-2010 20:44:16] <python600rr> how easy is it to create a plugin(.egg)
[01-Jan-2010 20:45:17] <ke4qqq> python600rr: zenpacks are relatively easy
[01-Jan-2010 20:45:32] <ke4qqq> depends on how complicated you want it, but simple ones are simple.
[01-Jan-2010 20:45:45] <ke4qqq> you should read Jane Curry's 88-page missive on the subject
[01-Jan-2010 20:46:33] <python600rr> is it located on the the zenoss site or just goog it
[01-Jan-2010 20:48:01] <python600rr> lookin at the documentation it looks like if you can write scripts in python then it should not be a problem, correct?!?
[01-Jan-2010 20:50:45] <python600rr> ke4qqq: I found it! Thanks!
[01-Jan-2010 20:51:56] <ke4qqq> google works
[01-Jan-2010 20:52:08] <ke4qqq> python600rr: if it's really simple, no python really needed
[01-Jan-2010 20:52:15] <ke4qqq> if it's more complex, python helps
[01-Jan-2010 20:53:12] <python600rr> ok, I am in the process of learning python..few scripts here and there, and I think that's what I'll work on...Thanks for your help!
[01-Jan-2010 20:53:29] <ke4qqq> my pleasure
[01-Jan-2010 20:55:15] <ke4qqq> Jane's docs are so comprehensive, I think Zenoss should buy them from her and relabel them.
[01-Jan-2010 20:59:11] <python600rr> ke4qqq: I just downloaded it.
[01-Jan-2010 20:59:47] <RobertLaptop> WMI just seems to hate me.
[01-Jan-2010 21:00:06] <ke4qqq> Windows hates us all
[01-Jan-2010 21:00:55] <python600rr> RobertLaptop: what are you tryna do?
[01-Jan-2010 21:01:36] <RobertLaptop> That is the problem. I am tiring to copy what is working on 2 other winodow 2k3 machines with IIS and the 3rd box is giving me an erro
[01-Jan-2010 21:01:57] <python600rr> That's windows for ya...:-)
[01-Jan-2010 21:02:06] <RobertLaptop> Yea
[01-Jan-2010 21:02:56] <RobertLaptop> Have anyone seen "zenperfwmi|/Status/Wmi||4|Could not read the WMI value (list index out of range). "
[01-Jan-2010 21:03:20] <RobertLaptop> A search of the zenoss site shows nothing.
[01-Jan-2010 21:04:49] <RobertLaptop> And yea windows sucks.
[01-Jan-2010 21:04:59] <python600rr> I don't know if you can but try looking at the source for the wmi script and see where it's having a problem. I think there is a way to test it on the box
[01-Jan-2010 21:05:39] <python600rr> I'm sure someone on here could inform you of a way to see where your having the issue
[01-Jan-2010 21:06:34] <RobertLaptop> Yea. I not afraid to debug. Just not sure were to look. I am assuming it is a problem with 2k3.
[01-Jan-2010 21:08:06] <python600rr> try looking at the code and wscript.echo out where you think it could be returing an index
[01-Jan-2010 21:08:56] <RobertLaptop> I know zero python.
[01-Jan-2010 21:10:21] <python600rr> I thought it would have been written in WMI, instead of python.
[01-Jan-2010 21:11:50] <python600rr> WoW! I didn't know they had wmi mods for python (I just looked it up). never mind
[01-Jan-2010 21:14:29] <ke4qqq> RobertLaptop: is nthis with MSSQL?
[01-Jan-2010 21:14:53] <RobertLaptop> MSSQL shouldn't be.
[01-Jan-2010 21:15:06] <RobertLaptop> Wait MSSQL client is installed.
[01-Jan-2010 21:15:33] <ke4qqq> but not the MSSQL WMI zenpack
[01-Jan-2010 21:16:05] <RobertLaptop> No I have the MSSQL WMI zenpack. But I don't have it manually bind to the system.
[01-Jan-2010 21:16:36] <RobertLaptop> Are you saying that is the problem?
[01-Jan-2010 21:18:18] <ke4qqq> someone else reported the problem in mid december with the mssql wmi zenpack
[01-Jan-2010 21:18:21] <ke4qqq> but might not be related
[01-Jan-2010 21:20:06] <RobertLaptop> Let me remove the package and see if the issue goes away.
[01-Jan-2010 21:20:24] <ke4qqq> unlikely if it isn't bound
[01-Jan-2010 21:22:15] <RobertLaptop> I have found something bind though device.
[01-Jan-2010 21:24:33] <RobertLaptop> But that is the main diff between the 3 boxes. 1 has MSSQL Client the other 2 doesn't
[01-Jan-2010 21:25:01] <RobertLaptop> Nope.
[01-Jan-2010 21:28:23] <RobertLaptop> But I wonder if it has something to do with sqlclient being installed.
[01-Jan-2010 21:34:58] <RobertLaptop> Ops spook to soon. Restarted Zenoss. And now it seems to be working. Data is showing up.
[01-Jan-2010 21:39:16] <ke4qqq> ahhh restarting zenoss is a must for wmi zenpacks after installation
[01-Jan-2010 21:39:20] <ke4qqq> and a few other zenpacks as well
[01-Jan-2010 21:41:50] <RobertLaptop> Yea. I removed the WMI - MSSQL Zenpack. Restarted ZenOSS.
[01-Jan-2010 21:42:52] <RobertLaptop> Problem seems to have gone away. I am guessing the MSSQL Zenpack assumes SQL Client == SQL Server or something else like it.
[01-Jan-2010 21:44:11] <RobertLaptop> What I would like to find is a back that watches IIS5
[01-Jan-2010 21:50:15] <ke4qqq> good
[02-Jan-2010 08:36:25] nassy_ is now known as nassy
[02-Jan-2010 11:25:56] <sergeymasushko> hi guys. is there a way to disable modeling of some interfaces for a device? for example I do not want to monitor ngX and tunX interfaces on some device due to these interfaces are dynamically created and destroyed.
[02-Jan-2010 13:15:29] <python600rr> have you looked for it in that device to see if it's listed for monitoring
[02-Jan-2010 14:54:38] <chemist> blist
[02-Jan-2010 16:17:04] <RobertLaptop> Anyone around?
[02-Jan-2010 16:17:46] <RobertLaptop> I am having some issue with the ircd template. Getting an error about a script not being found.
[02-Jan-2010 16:18:29] <python600rr> is it looking for a zenpack that needs to be installed
[02-Jan-2010 16:19:10] <RobertLaptop> What Pack. It can as part of the Core ZenPack's?
[02-Jan-2010 16:19:40] <RobertLaptop> can=came
[02-Jan-2010 16:19:52] <python600rr> what is the error it's looking for
[02-Jan-2010 16:19:58] <python600rr> what is the error
[02-Jan-2010 16:20:38] <RobertLaptop> Cmd: /opt/zenoss/ZenPacks/ZenPacks.zenoss.IRCDMonitor-1.0.1-py2.4.egg/ZenPacks/zenoss/IRCDMonitor/libexec/check_ircd.py -H 192.168.100.57 -p 6667 -w 50 -c 100 - Code: 127 - Msg: Command not found
[02-Jan-2010 16:21:28] <python600rr> look at your zenpacks and see if you have a ZenPacks.zenoss.IRCDMonitor-1.0.1-py.2.4.egg
[02-Jan-2010 16:21:51] <python600rr> if not then you need to install it and then restart zenoss then try again
[02-Jan-2010 16:22:23] <RobertLaptop> /opt/zenoss/ZenPacks/ZenPacks.zenoss.IRCDMonitor-1.0.1-py2.4.egg/ folder exists and has stuff in it.
[02-Jan-2010 16:22:38] <RobertLaptop> I will restart ZenOSS and see if that helps.
[02-Jan-2010 16:24:35] <python600rr> if not make sure that when you run zenpack --install ZenPacks.zenoss.IRCDMonitor-1.0.1-py2.4.egg it installs
[02-Jan-2010 16:24:57] <RobertLaptop> It was installed as part of the .rpm
[02-Jan-2010 16:25:12] <RobertLaptop> I can install using the .egg file
[02-Jan-2010 16:25:32] <python600rr> so it installs when you run the command?
[02-Jan-2010 16:26:06] <python600rr> locate check_ircd.py is it where it should be?
[02-Jan-2010 16:27:23] <RobertLaptop> yep
[02-Jan-2010 16:27:33] <RobertLaptop> Yes
[02-Jan-2010 16:30:40] <RobertLaptop> When I tired to install the egg file and got the following error /opt/zenoss/Products/ZenPacks.zenoss.IRCDMonitor-1.0.1-py2.4.egg/skins
[02-Jan-2010 16:30:54] <python600rr> does that exist
[02-Jan-2010 16:31:02] <RobertLaptop> OSError: [Errno 20] Not a directory: '/opt/zenoss/Products/ZenPacks.zenoss.IRCDMonitor-1.0.1-py2.4.egg/skins'
[02-Jan-2010 16:31:06] <python600rr> and is there any thing in it
[02-Jan-2010 16:31:13] <RobertLaptop> checking now
[02-Jan-2010 16:33:45] <RobertLaptop> easier fix. Delete and reinstall the ZenPack
[02-Jan-2010 16:33:53] <python600rr> yep
[02-Jan-2010 16:45:04] <RobertLaptop> Didn't work. I don't care that much about the check. I will play with it later.
[02-Jan-2010 16:46:52] <python600rr> ok
[02-Jan-2010 16:47:51] <python600rr> I would uninstall zenoss and install the zenoss-stack for my platform
[02-Jan-2010 16:49:40] <RobertLaptop> I don't like zenoss-stack. If I did I would be running debian
[02-Jan-2010 16:52:51] <python600rr> your running rh right?!?
[02-Jan-2010 16:53:33] <RobertLaptop> CentOS.
[02-Jan-2010 16:53:35] <python600rr> it's bassically the same thing except it has all of the dependencies
[02-Jan-2010 16:53:58] <python600rr> same thing, Im using Fedora and RH5 and I dont have any issues
[02-Jan-2010 16:54:30] <RobertLaptop> I am guessing I have missed some python requirement for the IRCd
[02-Jan-2010 16:54:44] <python600rr> possibly
[02-Jan-2010 16:55:01] <python600rr> try running the script using python and see what you get
[02-Jan-2010 16:55:12] <RobertLaptop> But I am also guessing that IRC isn't a high priorty item.
[02-Jan-2010 16:56:47] <python600rr> probably not but I would check the site for an update on that
[02-Jan-2010 16:58:01] <RobertLaptop> I forgot to restart zenoss.
[02-Jan-2010 16:59:00] <python600rr> ok
[02-Jan-2010 17:01:29] <RobertLaptop> Still not working. I will deal with it next week.
[02-Jan-2010 17:03:37] <RobertLaptop> Not sure if I am doing it right but ran it command line and this is the output
[02-Jan-2010 17:03:40] <RobertLaptop> python /opt/zenoss/ZenPacks/ZenPacks.zenoss.IRCDMonitor-1.0.1-py2.4.egg/ZenPacks/zenoss/IRCDMonitor/libexec/check_ircd.py -H 192.168.100.57 -p 6667 -w 50 -c 100
[02-Jan-2010 17:03:40] <RobertLaptop> Traceback (most recent call last):
[02-Jan-2010 17:03:40] <RobertLaptop> File "/opt/zenoss/ZenPacks/ZenPacks.zenoss.IRCDMonitor-1.0.1-py2.4.egg/ZenPacks/zenoss/IRCDMonitor/libexec/check_ircd.py", line 15, in ?
[02-Jan-2010 17:03:40] <RobertLaptop> import Globals
[02-Jan-2010 17:03:41] <RobertLaptop> ImportError: No module named Globals
[02-Jan-2010 17:06:03] <python600rr> that's probably where you have the problem, Globals I would see if there is a bugtrac out for it and submit if it isn't
[02-Jan-2010 17:06:33] <RobertLaptop> ok
[02-Jan-2010 17:13:15] <RobertLaptop> I wonder if this could be the issue? http://dev.zenoss.org/trac/changeset/9520
[03-Jan-2010 04:16:58] <Apachez> I have setup the devices... and now what... how do I get the graphs from them?
[03-Jan-2010 04:26:15] <Apachez> ahh
[03-Jan-2010 04:26:29] <Apachez> change to snmp v1 in zproperties then select "model device" to fetch new data :D
[03-Jan-2010 04:38:11] <Apachez> hmpf
[03-Jan-2010 04:38:20] <Apachez> still no graphs under Perf for these devices
[03-Jan-2010 04:38:29] <Apachez> the interfaces shows up in OS now
[03-Jan-2010 04:40:32] <Apachez> but there are graphs when I go and click on each interface... but that sucks...
[03-Jan-2010 04:40:42] <Apachez> I guess im alllll alone in here? ;)
[03-Jan-2010 04:54:08] <Apachez> hmm
[03-Jan-2010 04:54:14] <Apachez> zenoss dashboard doesnt work with firefox
[03-Jan-2010 04:54:25] <Apachez> 3.5.6 that is
[03-Jan-2010 05:39:40] <Apachez> gah
[03-Jan-2010 05:39:40] <Apachez> ERROR: zenpack command failed. Reason: OSError: [Errno 20] Not a directory: '/usr/local/zenoss/zenoss/Products/ZenPacks.community.deviceAdvDetail-1.2.1-py2.4.egg/skins'
[03-Jan-2010 05:39:44] <Apachez> 2.5.1 is broken :(
[03-Jan-2010 05:50:09] <Apachez> * Try unzipping the zenpack so that it ends in .egg and uploading that.
[03-Jan-2010 05:50:09] <Apachez> * It looks like your browser changed the file from a .egg to a .zip. Rename it to <zenpack file>.egg and try to install it.
[03-Jan-2010 05:50:24] <Apachez> heh,,, zenoss should apply for auto zip/gz detection while installing zenpacks me thinks :P
[03-Jan-2010 07:17:12] <Diddi> Apachez: regarding the graph issue, i guess that's solved by now since it takes a while for zenoss to actually gather data to draw a graph(:
[03-Jan-2010 09:07:34] <Apachez> Diddi: well I have graphs when I go to devive -> OS -> click on one of the ip interfaces
[03-Jan-2010 09:16:37] <Apachez> and and I guess I have to setup a custom multigraph report to have them all in one page?
[03-Jan-2010 09:16:52] <Apachez> whats buggers me is that dashboard doesnt seem to work in firefo 3.5.6
[03-Jan-2010 09:16:56] <Apachez> anyone else noticed this?
[03-Jan-2010 09:17:02] <Apachez> running zenoss community 2.5.1
[03-Jan-2010 09:17:13] <Apachez> on ubuntu-server 9.10 installed through apt-get
[03-Jan-2010 09:29:22] <python600rr> im using the same version on fedora 11 and I have no problem, and Im using ff 3.5.6
[03-Jan-2010 09:29:50] <python600rr> what is the error saying
[03-Jan-2010 09:31:47] <python600rr> Apachez
[03-Jan-2010 09:42:27] <Apachez> python600rr: i just posted it on the forum http://community.zenoss.org/message/44121#44121
[03-Jan-2010 09:42:38] <Apachez> there is no error, the dashboard is just blank
[03-Jan-2010 09:42:49] <Apachez> or well theres a 1cm in height grey frame in top
[03-Jan-2010 09:42:51] <Apachez> but thats it
[03-Jan-2010 09:43:07] <python600rr> do you have flash
[03-Jan-2010 09:43:14] <Apachez> yup
[03-Jan-2010 09:43:20] <Apachez> youtube works at least :P
[03-Jan-2010 09:43:25] <python600rr> lol
[03-Jan-2010 09:43:49] <python600rr> you don't see anything but a grey frame?
[03-Jan-2010 09:44:00] <Apachez> well I have the menu etc to the left
[03-Jan-2010 09:44:14] <python600rr> do you have the add portlet link on the right
[03-Jan-2010 09:44:24] <Apachez> but the main frame has just a 1cm (approx) in height grey frame that is filling the width
[03-Jan-2010 09:44:26] <Apachez> and nothing more
[03-Jan-2010 09:44:50] <Apachez> hang on Im firing the system up :)
[03-Jan-2010 09:44:57] <python600rr> ok
[03-Jan-2010 09:45:10] <RobertLaptop> I use kubntu 9.10 without issue.
[03-Jan-2010 09:45:18] <Apachez> shockwave flash is enabled but not shockwave for director
[03-Jan-2010 09:45:22] <Apachez> in firefox that is
[03-Jan-2010 09:45:25] <Apachez> firefox is on windows
[03-Jan-2010 09:46:23] <Apachez> k its up
[03-Jan-2010 09:46:32] <Apachez> no portlet link
[03-Jan-2010 09:46:39] <Apachez> what I have to the upper right is:
[03-Jan-2010 09:47:15] <Apachez> device/ip search \n admin preferences logout help \n zenoss server time: 16:47:25
[03-Jan-2010 09:47:57] <Apachez> below the server time line (which has blueish background) there is the main frame with dark grey
[03-Jan-2010 09:48:29] <Apachez> within this huge dark greyed background frame theres a small frame approx 1cm in height which is light grey with 1px black border
[03-Jan-2010 09:48:44] <Apachez> it fills almost the full width of the dark grey frame
[03-Jan-2010 09:48:48] <Apachez> but is only 1cm in heigh
[03-Jan-2010 09:48:55] <Apachez> now if I go in using ie6...
[03-Jan-2010 09:49:35] <Apachez> I get a couple of error on page javascript errors but then 1 secon later the dashboad is loading
[03-Jan-2010 09:49:59] <Apachez> and the configure layout.. add portlet... stop refresh is visible
[03-Jan-2010 09:51:18] <python600rr> 32bit or 64
[03-Jan-2010 09:52:05] <Apachez> the server is 64
[03-Jan-2010 09:52:08] <Apachez> the windows klient is 32
[03-Jan-2010 09:52:37] <python600rr> ok, maybe it has something to do with the versions
[03-Jan-2010 09:53:03] <python600rr> Is the version of zenoss a 64-bit version
[03-Jan-2010 09:53:33] <python600rr> try using the 32 bit version of firefox
[03-Jan-2010 09:53:44] <python600rr> I am using 32 bit, and I don't have any issues
[03-Jan-2010 09:55:10] <Apachez> and opera 9.64 on the same windowsbox will display the dashboard but right after it has loaded it says "lost connection to the server" in the upper left in red text
[03-Jan-2010 09:55:27] <Apachez> server: 64 bit ubuntu, 64 bit zenoss
[03-Jan-2010 09:55:34] <Apachez> client: 32bit windows, 32 bit firefox
[03-Jan-2010 09:56:14] <python600rr> ok, can you access zenoss from the localhost
[03-Jan-2010 09:56:19] <python600rr> with firefox
[03-Jan-2010 09:56:40] <python600rr> http://localhost:8080
[03-Jan-2010 09:58:36] <Apachez> I dont have any gui on the server
[03-Jan-2010 10:00:49] <python600rr> see if you have lynx on your computer..text based browser, to see what it gives u
[03-Jan-2010 10:05:52] <Apachez> yeah im in with lynx
[03-Jan-2010 10:05:54] <Apachez> now what? :)
[03-Jan-2010 10:06:51] <Apachez> Loading...
[03-Jan-2010 10:06:51] <Apachez> [spinning_wheel_throbber.gif]
[03-Jan-2010 10:06:51] <Apachez> [close.gif]
[03-Jan-2010 10:06:51] <Apachez> * admin
[03-Jan-2010 10:06:51] <Apachez> * Preferences
[03-Jan-2010 10:08:49] <python600rr> I just tried and I thought you could see the add portlet link, but you cant
[03-Jan-2010 10:09:49] <Apachez> upgraded to opera 10.10
[03-Jan-2010 10:10:03] <Apachez> same thing with upper left
[03-Jan-2010 10:10:09] <Apachez> first you see in red text "last updated"
[03-Jan-2010 10:10:19] <Apachez> and then 1 secon later it says "lost connection "
[03-Jan-2010 10:14:06] <python600rr> hmm
[03-Jan-2010 10:14:31] <python600rr> that's weird that you can see it with IE but not with firefox or opera
[03-Jan-2010 10:15:22] <python600rr> see if firebug will help you figure out the issue
[03-Jan-2010 10:15:33] <python600rr> it will allow you to debug the output from firefox
[03-Jan-2010 10:15:36] <python600rr> or opera
[03-Jan-2010 10:16:34] <RobertLaptop> Question any idea why I would be randomly getting auth error with WMI?
[03-Jan-2010 10:17:41] <python600rr> ummmm it's microsoft..lol!
[03-Jan-2010 10:18:12] <python600rr> just kidding, but did you check in the log file for windows to see if there are failures
[03-Jan-2010 10:20:00] <RobertLaptop> No that would probably be a good idea :)
[03-Jan-2010 10:20:09] <python600rr> lol@
[03-Jan-2010 10:21:17] <RobertLaptop> Another question. Any way to get stats on iis 5? I use the ZenPack for IIS6 and it works fine. Anyway to pull the same info for 5?
[03-Jan-2010 10:22:20] <python600rr> Have you checked the site, Im pretty sure they have something or you could develop something
[03-Jan-2010 10:22:41] <python600rr> they have stuff for AIX 5.3
[03-Jan-2010 10:22:42] <python600rr> lol!
[03-Jan-2010 10:24:25] <RobertLaptop> I guess I just need to search for build yourself templates. BTW nothing in the event logs on the windows box.
[03-Jan-2010 10:25:31] <RobertLaptop> Speaking of templates is there a way to bond a template to a device group? Example /devices/servers/windows/ii6 always gets the IIS6 template. Right now I have to bond them manually?
[03-Jan-2010 10:26:33] <python600rr> I have not tried the templates yet?
[03-Jan-2010 10:26:36] <RobertLaptop> At some point I am going to give this mantaince over to an MCSE so I have to make it as simple as possible :)
[03-Jan-2010 10:26:37] <python600rr> .
[03-Jan-2010 10:29:14] <python600rr> there is a good doc about creating zenpacks from jane curry
[03-Jan-2010 10:32:40] <python600rr> I was looking at the documentation and it looks like you can setup a class organizer and then assign templates that way
[03-Jan-2010 10:34:36] <python600rr> so under Classes click on Devices and then which device you want it under "Server" would probably be the case, and then from the drop down select new organizer
[03-Jan-2010 10:35:36] <python600rr> or better yet Classes: devices/servers/windows then select from the dropdown arrow new organizer
[03-Jan-2010 10:35:48] <python600rr> then it will show up in your classes
[03-Jan-2010 10:35:57] <python600rr> then apply your template
[03-Jan-2010 10:36:05] <python600rr> I just created one for AIX
[03-Jan-2010 10:42:49] <RobertLaptop> I will have to try that. It is likely one of those hide commands you don't see until you actually use it that go that is what that is for.
[03-Jan-2010 10:52:57] <python600rr> probably
[03-Jan-2010 11:06:50] <RobertLaptop> python600rr I was creating the sub class just find. I added the device type to the class. What I didn't think to do was bind the device type to the class. I missed the binding step.
[03-Jan-2010 11:07:11] <RobertLaptop> Now that I did it. It makes since.
[03-Jan-2010 11:07:56] <python600rr> :-)
[03-Jan-2010 11:36:15] <Apachez> hahaha
[03-Jan-2010 11:36:19] <Apachez> I solved the browser problem
[03-Jan-2010 11:37:52] <Apachez> it was due to how the browser identifies itself
[03-Jan-2010 11:52:36] <python600rr> did you use firebug
[03-Jan-2010 11:55:28] <Apachez> even better
[03-Jan-2010 11:55:31] <Apachez> I used live header
[03-Jan-2010 11:55:32] <Apachez> http://community.zenoss.org/message/44141#44141
[03-Jan-2010 11:57:16] <python600rr> congratz!!
[03-Jan-2010 11:57:29] <python600rr> now you can enjoy
[03-Jan-2010 12:03:38] <Apachez> somewhat
[03-Jan-2010 12:03:52] <Apachez> I still need to figure out how to setup a "page" which lists all graphs per device
[03-Jan-2010 12:04:15] <python600rr> go to the alert settings
[03-Jan-2010 12:06:09] <python600rr> settings, users, click on the user who needs the alerts and then alerting rules
[03-Jan-2010 12:07:10] <python600rr> never mind I was distracted, I gave you "paging"
[03-Jan-2010 12:07:44] <Apachez> yeah :(
[03-Jan-2010 12:07:52] <Apachez> I can see graphs per interface
[03-Jan-2010 12:08:12] <Apachez> but then I need to go device -> [xxxx] -> OS -> interfaceX and now I see the traffic graph
[03-Jan-2010 12:08:27] <Apachez> but I want to device -> see all traffic graphs
[03-Jan-2010 12:08:39] <Apachez> like 3-4 clicks less and also see all graphs at the same time
[03-Jan-2010 12:08:44] <python600rr> It takes a while for the graph to start showing info, atleast for me it did
[03-Jan-2010 12:08:49] <Apachez> I guess Ill have to learn how to setup reports
[03-Jan-2010 12:09:01] <Apachez> well I have graphs
[03-Jan-2010 12:09:14] <Apachez> but I dont see all the graphs at one place
[03-Jan-2010 12:09:22] <Apachez> here is what I have to do currently
[03-Jan-2010 12:09:30] <Apachez> click device -> [xxxx] -> OS -> interfaceX
[03-Jan-2010 12:09:37] <Apachez> now I see traffic graph for interfaceX
[03-Jan-2010 12:09:41] <Apachez> click back
[03-Jan-2010 12:09:45] <Apachez> click interfaceX2
[03-Jan-2010 12:09:55] <Apachez> now I only see traffic graph for interfaceX2
[03-Jan-2010 12:10:09] <Apachez> there will be shitloads of clicks if I want to see all 48 ports that this switch contains
[03-Jan-2010 12:11:00] <python600rr> lol!
[03-Jan-2010 13:04:13] <RobertLaptop> Yea Apachez that would be a lot of clicks
[03-Jan-2010 14:25:53] <Apachez> RobertLaptop: any suggestion on how to lower them in an easy fashion? :)
[03-Jan-2010 14:26:15] <RobertLaptop> Create a custom Template
[03-Jan-2010 14:27:21] <RobertLaptop> I am working on a template for my ASG right now.
[03-Jan-2010 14:27:44] <RobertLaptop> The hardest part is getting the qid right
[03-Jan-2010 20:53:18] bzed__ is now known as bzed
[04-Jan-2010 05:58:57] <neuromancer> hi, anyone available to help?
[04-Jan-2010 06:01:00] <neuromancer> anyone?
[04-Jan-2010 06:09:49] <tehhobbit> I can try
[04-Jan-2010 06:13:20] <Apachez> I can try aswell
[04-Jan-2010 06:14:12] <tehhobbit> Apachez: are you on efnet by any chance ?
[04-Jan-2010 06:15:28] <neuromancer> hi, sorry for the delay tehhobbit
[04-Jan-2010 06:21:48] <Apachez> tehhobbit: why do you wonder?
[04-Jan-2010 06:22:42] <neuromancer> well i just installed the dell core monitoring zenpack and for some reason all of my devices have dissapeared
[04-Jan-2010 06:23:09] <neuromancer> i have since uninstalled it and restarted zen daemons but devices are still not on the list
[04-Jan-2010 06:23:42] <neuromancer> however, wierdly the alerts still work and they get sent to all users but in the admin section i can only see my user and the admin user
[04-Jan-2010 06:24:06] <neuromancer> im thinking a reinstall unless i can find a fix?
[04-Jan-2010 06:29:03] <tehhobbit> Apachez: just recognized the nickname from #sweden@efnet
[04-Jan-2010 09:33:42] <QubeZ> hello all.
[04-Jan-2010 09:34:03] <rmatte> hey
[04-Jan-2010 09:34:12] <QubeZ> i just installed Zenoss 2.5.1 fresh install on a fresh install of CentOS. When I go to Settings -> ZenPacks, I dont see anything list like on my older installs.
[04-Jan-2010 09:34:37] <QubeZ> even when I add a device and go to bind templates, there is only the "device" template there. Nothing else to choose from.
[04-Jan-2010 09:41:10] <rmatte> QubeZ: you say it's a fresh install...
[04-Jan-2010 09:41:17] <Apachez> what browser?
[04-Jan-2010 09:41:19] <rmatte> eugh
[04-Jan-2010 09:41:32] <rmatte> I wish people would wait more than 5 minutes after asking a question before leaving
[04-Jan-2010 09:41:32] <Apachez> did you modify your useragent string? ;)
[04-Jan-2010 09:41:36] <rmatte> he left
[04-Jan-2010 09:41:39] <Apachez> uhh already gone
[04-Jan-2010 09:41:41] <Apachez> too bad
[04-Jan-2010 09:41:59] <Apachez> im still happy of my finding regarding useragent bug in zenoss :)
[04-Jan-2010 09:42:04] <rmatte> oh well, if he wants an answer to his question he'll be back
[04-Jan-2010 09:42:11] <rmatte> useragent bug?
[04-Jan-2010 09:42:48] <rmatte> oh, where if you specify that your browser is firefox instead of Opera or whatever it'll display correctly?
[04-Jan-2010 09:42:59] <Apachez> well I have modified useragent which zenoss didnt like so my dashboard went ballistic
[04-Jan-2010 09:43:05] <rmatte> ah
[04-Jan-2010 09:43:05] <Apachez> yup
[04-Jan-2010 09:43:16] <Apachez> I have removed that im on windows, which windows, which patches, which version etc
[04-Jan-2010 09:43:21] <rmatte> I doubt that's QubeZ issue though...
[04-Jan-2010 09:43:43] <Apachez> you never know...
[04-Jan-2010 09:43:44] <rmatte> he said it's a fresh 2.5.1 install, I'm guessing he hasn't actually installed any ZenPacks yet :P
[04-Jan-2010 09:44:00] <rmatte> fresh meaning not an upgrade
[04-Jan-2010 09:44:08] <rmatte> ZenPacks don't just magically install themselves
[04-Jan-2010 09:44:14] <Apachez> sure but fresh as in never installed previously? :P
[04-Jan-2010 09:44:25] <rmatte> what difference would that make?
[04-Jan-2010 09:44:27] <Apachez> or fresh as in "same browser and everything but the zenoss and server is new" ? :P
[04-Jan-2010 09:44:38] <rmatte> If he installed previously then uninstalled then reinstalled he'd be starting from scratch
[04-Jan-2010 09:44:38] <Apachez> well opera 10.10 didnt work either with zenoss
[04-Jan-2010 09:44:53] <Apachez> so I guess there can be all sort of shit in the (I suppose) javascripts
[04-Jan-2010 09:45:06] <rmatte> He also mentioned that it worked fine previously, meaning he was using the same browser, meaning it's not a browser issue
[04-Jan-2010 09:45:18] <Apachez> perhaps another browser?
[04-Jan-2010 09:45:21] <Apachez> now we'll never know...
[04-Jan-2010 09:45:23] <rmatte> highly doubt it
[04-Jan-2010 09:45:39] <rmatte> I find that browser issues only make up about 5 percent of Zenoss issues, if not less
[04-Jan-2010 09:46:26] <Apachez> by the way
[04-Jan-2010 09:46:33] <Apachez> 300 secs is the default per scan in zenoss right ?
[04-Jan-2010 09:46:40] <rmatte> correct, 5 minutes
[04-Jan-2010 09:46:49] <rmatte> pings are done every 60 seconds
[04-Jan-2010 09:46:51] <Apachez> any downside with lowering to ehh 60 or 30 ?
[04-Jan-2010 09:47:05] <Apachez> im thinking if im hitting the 32 bit bug with snmp
[04-Jan-2010 09:47:25] <rmatte> well, you'd have to re-generate all of your RRDs, and it would be extremely intensive depending on the number of devices you are monitoring
[04-Jan-2010 09:47:42] <tehhobbit> Apachez: that makes 2 of us , espsially if you mix both linux, solaris and windows
[04-Jan-2010 09:47:53] <tehhobbit> none of the default snmp clients work the same
[04-Jan-2010 09:47:53] <Apachez> more than 14 316 557 bit/s during 5min will hit the roof
[04-Jan-2010 09:48:19] <Apachez> because right now I see a spike of 40Mbit/s on one of the graphs
[04-Jan-2010 09:48:21] <tehhobbit> Apachez: on linux you can always use the ifXtable and v3
[04-Jan-2010 09:48:27] <Apachez> but the packet/s doesnt follow it
[04-Jan-2010 09:48:50] <Apachez> its an allied telesyn switch
[04-Jan-2010 09:48:56] <Apachez> which supports snmp 2
[04-Jan-2010 09:49:03] <Apachez> but the interfaces only seem to work with v1
[04-Jan-2010 09:49:17] <QubeZ> which is the package to download the core zenpacks? every time i go to download page for my centos rpm install, i only see the actual application and no rpm for zenpacks
[04-Jan-2010 09:49:17] <tehhobbit> then the ifXtable is missing from the agent
[04-Jan-2010 09:49:54] <tehhobbit> Apachez: you would think that the vendors would have realized by now that ppl accually use full gbit :(
[04-Jan-2010 09:50:30] <tehhobbit> Apachez: or humm , have you changed to using the 64bit ethnet template ?
[04-Jan-2010 09:51:21] <OiScout> QubeZ: http://community.zenoss.org/docs/DOC-3480
[04-Jan-2010 09:51:29] <tehhobbit> QubeZ: http://sourceforge.net/projects/zenoss/files/zenpacks-2.5%20%28Latest%20Stable%29/zenpacks-2.5.1/zenoss-core-zenpacks-2.5.1.el4.x86_64.rpm/download
[04-Jan-2010 09:51:35] <Apachez> it choosed 64bit template automagically
[04-Jan-2010 09:51:44] <Apachez> hmm
[04-Jan-2010 09:51:44] <QubeZ> thanks
[04-Jan-2010 09:51:45] <tehhobbit> http://sourceforge.net/projects/zenoss/files/zenpacks-2.5%20%28Latest%20Stable%29/zenpacks-2.5.1/zenoss-core-zenpacks-2.5.1.el5.i386.rpm/download
[04-Jan-2010 09:51:48] <tehhobbit> sorry
[04-Jan-2010 09:51:48] <Apachez> prtg reports 40mbit aswell
[04-Jan-2010 09:52:05] <rmatte> QubeZ: so you figured out that you actually have to install the ZenPacks on a fresh install? lol
[04-Jan-2010 09:52:13] <tehhobbit> Apachez: then the switch agent is missing ifX
[04-Jan-2010 09:52:17] <rmatte> QubeZ: you left 1 second before I could answer your questions
[04-Jan-2010 09:52:20] <Apachez> ifx what ?
[04-Jan-2010 09:52:28] <QubeZ> rmatte lol, ya cuz i felt dumb ;)
[04-Jan-2010 09:52:33] <rmatte> ;)
[04-Jan-2010 09:52:38] <QubeZ> rmatte weird though, not sure why the zenpacks aren't on the same page as the isntaler rpm
[04-Jan-2010 09:52:43] <eidolon> okay, total wish for the day. - the device / ip search tab at the top of zenoss? should be able to do a substring search
[04-Jan-2010 09:52:44] * eidolon grumps.
[04-Jan-2010 09:52:51] <rmatte> QubeZ: because they are on the ZenPacks page
[04-Jan-2010 09:53:01] <rmatte> QubeZ: the ZenPacks page is there for a reason :P lol
[04-Jan-2010 09:53:12] <QubeZ> rmatte guess it makes sense =)
[04-Jan-2010 09:53:49] <rmatte> If you go to the ZenPack page and scroll down to the "Zenoss Core ZenPacks" section you'll see "ZenPack Bundle" (for future reference)
[04-Jan-2010 09:53:51] <tehhobbit> Apachez: ifXtable is the module + mib that contains 64bit counters for ethernet interfaces
[04-Jan-2010 09:54:03] <Apachez> can I use snmpwalk to somehow debug why this allied telesyn switch doesnt seem to give me 64 bit counters
[04-Jan-2010 09:54:12] <Apachez> or how can I verify if it supports 64bit counters ?
[04-Jan-2010 09:54:49] <rmatte> Apachez: well, the interface type in snmp should be ethernetCsmacd_64
[04-Jan-2010 09:54:54] <rmatte> if it's not, it doesn't support them
[04-Jan-2010 09:54:57] <tehhobbit> Apachez: got to Devices/Network/Switch/Template -> check the ethernet 64 template and try the oid manually
[04-Jan-2010 09:55:04] <tehhobbit> s/got/go/
[04-Jan-2010 09:55:06] <rmatte> since Zenoss is monitoring with 32bit I assume it's showing as ethernetCsmacd
[04-Jan-2010 09:55:46] <rmatte> generally that means that it doesn't support 64bit, though it might just be stupidity on the part of the manufacturer
[04-Jan-2010 09:57:12] <Apachez> bye th e way those 40mbit/s seems to have been a ddos
[04-Jan-2010 09:57:18] <Apachez> or dos
[04-Jan-2010 09:57:19] <tehhobbit> Apachez: try snmpwalk <bla bla password bla bla > .1.3.6.1.2.1.31.1.1.1.7
[04-Jan-2010 09:57:40] <tehhobbit> that should if the device supports it give you a 64bit counter if out octets
[04-Jan-2010 09:57:53] <rmatte> try with -v1 and -v2c
[04-Jan-2010 09:57:58] <rmatte> you might get different results
[04-Jan-2010 09:58:27] <tehhobbit> rmatte: v1 never supports 64 bit counters
[04-Jan-2010 09:58:41] <Apachez> im in the template now
[04-Jan-2010 09:58:51] <Apachez> trying to learn how to navgiate in there to do the debug through gui :)
[04-Jan-2010 09:59:28] <Apachez> ohh you meant to find out the oid
[04-Jan-2010 09:59:36] <Apachez> and then use it in snmpwalk through cli
[04-Jan-2010 09:59:48] <tehhobbit> yup
[04-Jan-2010 10:01:28] <rmatte> yeh, you need to see if you get a value from it
[04-Jan-2010 10:01:52] <rmatte> actually, tehhobbit, the problem is that I believe that may not be a full OID
[04-Jan-2010 10:02:08] <rmatte> I think the collector plugin adds some stuff to end of it based on the if index
[04-Jan-2010 10:03:01] <tehhobbit> rmatte: not its the full one, used it to debug this very problem less than a week ago
[04-Jan-2010 10:03:19] <rmatte> k
[04-Jan-2010 10:04:18] <Apachez> snmpwalk -Os -c community -v 1 x.x.x.x 1.3.6.1.2.1.31.1.1.1.7
[04-Jan-2010 10:04:18] <Apachez> snmpwalk -Os -c community -v 2c x.x.x.x 1.3.6.1.2.1.31.1.1.1.7
[04-Jan-2010 10:04:24] <Apachez> they gives me the same output
[04-Jan-2010 10:04:33] <Apachez> ifHCInUcastPkts.5 = Counter64: 13993860
[04-Jan-2010 10:04:47] <Apachez> but snmp v1 doesnt support 64bit counters ?
[04-Jan-2010 10:04:53] <tehhobbit> http://pastie.org/766134
[04-Jan-2010 10:04:56] <Apachez> then why am I getting Counter64 in the reply ?
[04-Jan-2010 10:06:17] <tehhobbit> because there quite a few different devices outthere with a retarded snmp aget ? :D
[04-Jan-2010 10:06:27] <tehhobbit> angent*
[04-Jan-2010 10:06:34] <rmatte> yup, like I said before, stupidity of the manufacturer
[04-Jan-2010 10:06:46] <twm1010> :D
[04-Jan-2010 10:07:01] <Apachez> but what does it mean?
[04-Jan-2010 10:07:07] <rmatte> had they labeled the interface type as ethernetCsmacd_64 instead of ethernetCsmacd you wouldn't be having the issue
[04-Jan-2010 10:07:11] <Apachez> that the manufact returns 64bit data for v1 requests ?
[04-Jan-2010 10:07:38] <twm1010> Can you set the device to only do v1?
[04-Jan-2010 10:07:41] <twm1010> perhaps that would do it?
[04-Jan-2010 10:08:26] <Apachez> devices -> network -> zproprties I changed to v2c
[04-Jan-2010 10:08:36] <twm1010> not in Zenoss, at the device itself
[04-Jan-2010 10:08:39] <Apachez> so I guess ill have to wait like 15 mins or so to see if it works ?
[04-Jan-2010 10:08:44] <Apachez> dunno
[04-Jan-2010 10:08:54] <Apachez> they are supposed to use 2c by default
[04-Jan-2010 10:09:06] <Apachez> and I just verified with snmpwalk that v2c works to get interface states
[04-Jan-2010 10:09:20] <rmatte> you might also have to create a local copy of the ethernetCsmacd template on the interfaces and modify it to look like the ethernetCsmacd_64 template
[04-Jan-2010 10:09:34] <rmatte> since Zenoss auto-assigns interface templates based on interface type
[04-Jan-2010 10:09:42] <twm1010> That sounds like a way around it...
[04-Jan-2010 10:10:11] <rmatte> you'd have to do it individually for each interface though
[04-Jan-2010 10:10:27] <rmatte> since Zenoss is kind of weak in it's local template system
[04-Jan-2010 10:13:32] <rmatte> well, I think I found a simple way to have Zenoss stop automatically disabling my thresholds
[04-Jan-2010 10:13:40] <rmatte> just need to comment out 1 line and then write a transform
[04-Jan-2010 10:13:52] <rmatte> testing it now...
[04-Jan-2010 10:14:15] <Apachez> I just verified with snmpwalk that all oid's in ethernetCsmacd_64 works with this particular unit
[04-Jan-2010 10:14:21] <Apachez> using v2c
[04-Jan-2010 10:14:33] <Apachez> will now try ethernetCsmacd and v1 just for the fun of it while Im at it
[04-Jan-2010 10:15:18] <rmatte> now if you could contact the manufacturer and tell them that they are id10ts...
[04-Jan-2010 10:15:20] <rmatte> :)
[04-Jan-2010 10:15:27] <Apachez> "truncting unsigned value to 32bit" counter32
[04-Jan-2010 10:15:46] <rmatte> "truncating" I assumed?
[04-Jan-2010 10:15:52] <rmatte> :)
[04-Jan-2010 10:15:55] <Apachez> hmm
[04-Jan-2010 10:16:01] <Apachez> operstatus is missing in _64 template
[04-Jan-2010 10:16:13] <Apachez> rmatte: my misspells are free of charge ;)
[04-Jan-2010 10:17:29] <rmatte> Apachez: yeh, you need to add that manually, it's not a default thing
[04-Jan-2010 10:17:51] <Apachez> wtf
[04-Jan-2010 10:17:56] <rmatte> ?
[04-Jan-2010 10:18:03] <Apachez> and you say that NOW? :P
[04-Jan-2010 10:18:19] <rmatte> May I direct you to: http://community.zenoss.org/docs/DOC-2494
[04-Jan-2010 10:18:30] <rmatte> which is probably where you got your other operstatus datapoint from in the first place
[04-Jan-2010 10:18:52] <rmatte> :P
[04-Jan-2010 10:19:10] <Apachez> nope there were included in the template
[04-Jan-2010 10:19:22] <rmatte> I've been trying to convince them to add operational status monitoring as a built in feature, and it sounds like it's going to eventually happen.
[04-Jan-2010 10:19:42] <Apachez> hmm a bug or maybe not
[04-Jan-2010 10:19:54] <rmatte> so on a fresh install it was included?
[04-Jan-2010 10:19:57] <Apachez> the ifoperstatus in the ethernetCsmacd template has a dot as first char in the oid
[04-Jan-2010 10:20:01] <Apachez> yup
[04-Jan-2010 10:20:04] <Apachez> 2.5.1
[04-Jan-2010 10:20:19] <rmatte> it always adds a dot to the front of it automatically
[04-Jan-2010 10:20:22] <rmatte> that doesn't mean anything
[04-Jan-2010 10:20:29] <Apachez> its not there for the other oids
[04-Jan-2010 10:20:35] <rmatte> hm
[04-Jan-2010 10:20:51] <rmatte> maybe someone added it in to the development version for testing and it didn't get removed before packaging for deployment was done
[04-Jan-2010 10:26:03] <Apachez> where is the datapoint menu ?
[04-Jan-2010 10:26:08] <Apachez> I only see add datapoint alias
[04-Jan-2010 10:28:18] <rmatte> alias?
[04-Jan-2010 10:28:22] <rmatte> that's for reports
[04-Jan-2010 10:28:27] <rmatte> are you actually in the template?
[04-Jan-2010 10:28:38] <rmatte> oh wait, nevermind
[04-Jan-2010 10:28:44] <rmatte> you added an snmp datasource right?
[04-Jan-2010 10:28:59] <rmatte> the datapoints are automatically created for snmp datasources since 2.4.x
[04-Jan-2010 10:29:08] <rmatte> (I know you use to have to manually create them in 2.3.x)
[04-Jan-2010 10:29:53] <Kristopher424> Happy New Year, all! Have a quick ?... When I drill into a device and choose "OS" I can see interface ge-0/1/0 as well as all other interfaces and their usage charts. When I go to Reports>Performance>Interface Utilization...I can see two interfaces from that device and not ge-0/1/0. What goes on here?
[04-Jan-2010 10:30:42] <rmatte> Kristopher424: what is the interface type listed for that interface when you click on it?
[04-Jan-2010 10:30:50] <rmatte> you'll see a bunch of info including type
[04-Jan-2010 10:31:22] <Kristopher424> ge-0/1/0 = ethernetCsmacd_64
[04-Jan-2010 10:31:39] <rmatte> ok, so that's fine... hmmmm
[04-Jan-2010 10:31:53] <Kristopher424> [It's a fiber port on a switch...ge-0/0/x is copper and ge-0/1/x
[04-Jan-2010 10:31:56] <rmatte> you are aware that the report will only show 40 interfaces by default?
[04-Jan-2010 10:31:56] <Kristopher424> is fiber
[04-Jan-2010 10:32:05] <rmatte> and that there's a show all button at the bottom that you have to click to show more?
[04-Jan-2010 10:32:18] <Kristopher424> Yep, I've tried the "Show All"
[04-Jan-2010 10:32:49] <rmatte> When you click on the device you actually see graphs for it?
[04-Jan-2010 10:32:56] <Kristopher424> Yessir.
[04-Jan-2010 10:32:58] <rmatte> erm, on the interface rather
[04-Jan-2010 10:33:07] <Kristopher424> That too, yes :)
[04-Jan-2010 10:33:14] <Apachez> rmatte: hmpf then that faq should be updated :P
[04-Jan-2010 10:33:22] <rmatte> Apachez: which?
[04-Jan-2010 10:33:31] <rmatte> oh, the one for datasources
[04-Jan-2010 10:33:36] <rmatte> talk to mrayzenoss about that ;)
[04-Jan-2010 10:34:01] <rmatte> Kristopher424: no idea then, those were the most likely issues
[04-Jan-2010 10:34:44] <Apachez> rmatte :)
[04-Jan-2010 10:35:29] <Apachez> by the way, did you try myisam as mysql storage compared to innodb?
[04-Jan-2010 10:35:41] <rmatte> nope
[04-Jan-2010 10:36:19] <Apachez> are all the db commands located in the same file ?
[04-Jan-2010 10:36:28] <Apachez> or are they spread out through all zenoss files ?
[04-Jan-2010 10:36:37] <Apachez> I was thinking if I should take a look at it
[04-Jan-2010 10:36:57] <rmatte> Pretty sure there are bits and pieces spread out
[04-Jan-2010 10:37:24] <rmatte> but I've never really dug in to the code for that
[04-Jan-2010 10:37:48] <Kristopher424> rmatte, what makes the interface show on that report? If it has traffic?
[04-Jan-2010 10:38:57] <rmatte> Kristopher424: Zenoss has to have collected performance data for it
[04-Jan-2010 10:39:18] <rmatte> Kristopher424: other than that, Zenoss just does a for loop through all of the interfaces on each device
[04-Jan-2010 10:39:26] <rmatte> Kristopher424: so I'm not sure how it could be missing it
[04-Jan-2010 10:39:59] <rmatte> Kristopher424: can you hit the select all button, then scroll all the way to the bottom and screenshot it then put it up at tinypic.com please?
[04-Jan-2010 10:40:17] <rmatte> also screenshot the section for that device
[04-Jan-2010 10:40:23] <rmatte> I just want to see a couple things
[04-Jan-2010 10:59:12] <eidolon> hrm. standard nagios return string format is digit|Variable=value - where 'digit' is a return code. 0 = all is well, 1 or more == severity level, yes?
[04-Jan-2010 10:59:31] * eidolon is writing a Cmd monitor
[04-Jan-2010 11:00:37] <rmatte> I just do OK|Whatever=
[04-Jan-2010 11:00:43] <rmatte> but whatever tickles your fancy
[04-Jan-2010 11:01:12] <rmatte> If the script fails to gather performance data I have it return "Unknown"
[04-Jan-2010 11:01:57] <rmatte> I'm not sure how Zenoss actually interprets the digit, but I know it's not required
[04-Jan-2010 11:02:10] <eidolon> oh interesitng.
[04-Jan-2010 11:02:11] <eidolon> okay
[04-Jan-2010 11:02:32] <eidolon> is that documented somewhere? and zenoss is 'smart' enought o know the Cmd it's running will return Nagios formatted results, yes?
[04-Jan-2010 11:02:45] <rmatte> it expects it to, yes
[04-Jan-2010 11:02:52] <cluther> eidolon: The exit code of your program maps to the Zenoss severity. 0=Clear, 1=Severity you set in command datasource, 2=1+severity you set in command data source.
[04-Jan-2010 11:02:52] <eidolon> ok, tx
[04-Jan-2010 11:02:54] <Kristopher424> rmatte, sorry I had to step away. I'll get those screenshots posted in just a moment
[04-Jan-2010 11:03:00] <eidolon> got it.
[04-Jan-2010 11:03:02] <eidolon> thanks cluther.
[04-Jan-2010 11:03:15] <rmatte> Kristopher424: np
[04-Jan-2010 11:03:16] <cluther> The text before the | will become the event's summary if required.
[04-Jan-2010 11:03:29] <rmatte> cluther: ah cool, didn't know that
[04-Jan-2010 11:03:34] <eidolon> awesome. thanks again.
[04-Jan-2010 11:03:39] <cluther> The name=value name=value â€¦ will be stored according to the data points on your data source.
[04-Jan-2010 11:03:50] <eidolon> is there a simple URL for that? i keep hitting nagios API / Monitor design docs that are awful to read.
[04-Jan-2010 11:03:54] <eidolon> yah, i know about the data points.
[04-Jan-2010 11:04:03] <mrayzenoss> eidolon: http://community.zenoss.org/docs/DOC-3909#d4e3252
[04-Jan-2010 11:04:10] * eidolon clix
[04-Jan-2010 11:04:31] <eidolon> ohh, awesome
[04-Jan-2010 11:04:37] <eidolon> right, and zentestcommand. egg-zellent
[04-Jan-2010 11:08:02] <Apachez> zproperties, are those per template type or are zproperties global ?
[04-Jan-2010 11:09:05] <rmatte> global
[04-Jan-2010 11:10:12] <Kristopher424> rmatte, This is the bottom of the interface perf report. You can see the two interfaces on that switch as well, but the fiber doesn't show. http://tinypic.com/view.php?pic=2njaoba&s=6
[04-Jan-2010 11:10:59] <Kristopher424> rmatte, This is in the Device>OS menu, you can see both the copper and the highlighted fiber ports. http://tinypic.com/view.php?pic=winiuq&s=6
[04-Jan-2010 11:12:06] <rmatte> what's the name of the fiber int again?
[04-Jan-2010 11:12:33] <rmatte> oh, ge-0/1/0
[04-Jan-2010 11:12:42] <rmatte> and yes, the .0 would be virtual
[04-Jan-2010 11:12:45] <Kristopher424> ge-0/1/0
[04-Jan-2010 11:13:02] <Kristopher424> Then we have a virtual loopback on it which is ge-0/1/0.0
[04-Jan-2010 11:13:07] <rmatte> yeh, but that's not the only one not showing up
[04-Jan-2010 11:13:17] <rmatte> 0/0/1 to 0/0/9 aren't showing either
[04-Jan-2010 11:13:32] <Kristopher424> Right, ge-0/0/1 and 0/0/9 don't have anything connected to them
[04-Jan-2010 11:14:01] <rmatte> I see
[04-Jan-2010 11:14:22] <rmatte> I see the report that says 1 of 156...
[04-Jan-2010 11:14:33] <rmatte> those are the exact last 2 interfaces in the report?
[04-Jan-2010 11:14:38] <Kristopher424> Yep.
[04-Jan-2010 11:14:44] <rmatte> you're absolutely positive that there aren't more after those?
[04-Jan-2010 11:15:04] <Kristopher424> If there are I can't get to them. There's no next page button or selection
[04-Jan-2010 11:15:08] <Kristopher424> it's grayed out.
[04-Jan-2010 11:15:14] <rmatte> right, true
[04-Jan-2010 11:15:19] <rmatte> hmmm
[04-Jan-2010 11:15:31] <Apachez> but you can set zproperties per device aswell ?
[04-Jan-2010 11:16:12] <rmatte> Apachez: the zProperties fields themselves are global, in the sense that they will always be displayed on every device
[04-Jan-2010 11:16:23] <rmatte> Apachez: but you can set local settings per device (if that's what you're asking)
[04-Jan-2010 11:16:40] <Kristopher424> rmatte, I ran a filter to only show interfaces on that report that are from that switch... Results: http://tinypic.com/r/2ev9jz5/6
[04-Jan-2010 11:17:24] <rmatte> Kristopher424: I'm honestly not sure what would be causing that, they should be displayed
[04-Jan-2010 11:17:42] <rmatte> Without actually having the Zenoss instance to monkey with I can't really help much more
[04-Jan-2010 11:18:32] <Kristopher424> Hmm.
[04-Jan-2010 11:19:19] <Apachez> hmm
[04-Jan-2010 11:19:32] <Apachez> when I choose to model device with 2c I get timed out: are your SNMP settings correct?
[04-Jan-2010 11:19:37] <Apachez> but it works to model with v1
[04-Jan-2010 11:19:53] <Apachez> but if I change to v2c after modeling with v1, then v2c will be used for the queries right ?
[04-Jan-2010 11:22:07] <rmatte> modelling defaults to v1 when available
[04-Jan-2010 11:22:20] <rmatte> monitoring is different
[04-Jan-2010 11:22:46] <rmatte> modelling should default to whatever you set when adding the device but it doesn't
[04-Jan-2010 11:22:54] <Apachez> because when I change in zproperties to v2c and then model devie it will bring me the above error
[04-Jan-2010 11:23:19] <theacolyte-> man, ever since I put in Zenoss I'm catching all sorts of things I didn't before.
[04-Jan-2010 11:23:32] <theacolyte-> This is crazy
[04-Jan-2010 11:23:37] <rmatte> theacolyte-: yup lol
[04-Jan-2010 11:29:12] <Apachez> hmm
[04-Jan-2010 11:29:21] <Apachez> tried the manual but what does "lock" mean in zenoss?
[04-Jan-2010 11:29:27] <Apachez> that the interface wont be monitored?
[04-Jan-2010 11:29:37] <Apachez> or is it locked from being deleted ?
[04-Jan-2010 11:29:53] <Apachez> err I mean if lock is like "paus" ?
[04-Jan-2010 11:45:47] <rmatte> Apachez: locked means that a remodeling won't affect that object
[04-Jan-2010 11:46:05] <rmatte> Apachez: it basically keeps Zenoss from automatically removing or modifying the object
[04-Jan-2010 11:54:45] <Apachez> ahh
[04-Jan-2010 11:54:46] <Apachez> tnx
[04-Jan-2010 11:55:18] <rmatte> np
[04-Jan-2010 11:57:01] <RobertLaptop> Ok anyone know anything about ZenPacks.community.LinuxMonitorAddOn-1.0-py2.4.egg? I enabled it and turned on the template and all data collection for that device stop?
[04-Jan-2010 11:58:12] <rmatte> RobertLaptop: read the description of the pack, I think all it does is add monitoring for interfaces and memory via SNMP
[04-Jan-2010 11:58:16] <rmatte> from what I recall
[04-Jan-2010 11:58:20] <rmatte> you may have to rebind the templates
[04-Jan-2010 11:58:38] <rmatte> it's specifically for SSH monitoring, so if you're monitoring via SNMP it doesn't apply to you
[04-Jan-2010 11:59:05] <RobertLaptop> It for a system montord using ssh
[04-Jan-2010 11:59:26] <rmatte> I have to head to a meeting, bbl
[04-Jan-2010 12:01:58] <RobertLaptop> When I added it to the machines template. The Standard ZenPacks.zenoss.LinuxMonitor stopped collecting.
[04-Jan-2010 12:18:00] <rmatte> back
[04-Jan-2010 12:20:53] <RobertLaptop> any idea why the standard ssh monitor would stop collecting when I turned on the template?
[04-Jan-2010 12:23:06] <rmatte> it shouldn't
[04-Jan-2010 12:23:28] <rmatte> check your collector plugins to make sure that they are still bound
[04-Jan-2010 12:23:33] <rmatte> it might have unbound some of them
[04-Jan-2010 12:23:45] <rmatte> all of the ssh plugins should be in the active plugins list
[04-Jan-2010 12:23:50] <rmatte> otherwise it won't collect the data
[04-Jan-2010 12:24:09] <eidolon> okay, help please.
[04-Jan-2010 12:24:21] <eidolon> i was working on a new datasource / cmd datasource
[04-Jan-2010 12:24:39] <eidolon> i had plugged in the commands in zenoss against... adevice or a template
[04-Jan-2010 12:24:42] <eidolon> but i can't find where i did it.
[04-Jan-2010 12:25:01] <eidolon> there's only one of these - it's calling a shell script - but i have no clue where it is in zenoss now. and it's not active, so it won't actually trigger.
[04-Jan-2010 12:25:15] <eidolon> can i grep / search for a string somehow? other than device / ip?
[04-Jan-2010 12:25:24] <rmatte> ok, hold up...
[04-Jan-2010 12:25:28] * eidolon holds up.
[04-Jan-2010 12:25:56] <rmatte> click on Devices in the left hand menu (Devices, not Device List)
[04-Jan-2010 12:26:12] <eidolon> done
[04-Jan-2010 12:26:18] <rmatte> From the dropdown select More -> All Templates
[04-Jan-2010 12:26:29] <rmatte> (I assume you remember what the template was called?)
[04-Jan-2010 12:26:42] <eidolon> maybe? :-/
[04-Jan-2010 12:26:52] <rmatte> was it applied directly to a device, or to a class?
[04-Jan-2010 12:27:02] <twm1010> he doesn't know :)
[04-Jan-2010 12:27:04] <eidolon> that i don't know :(
[04-Jan-2010 12:27:09] <eidolon> i don't remember where i was in the process.
[04-Jan-2010 12:27:10] <rmatte> how does that even happen!?
[04-Jan-2010 12:27:16] <rmatte> lol
[04-Jan-2010 12:27:18] <eidolon> there is no more->all templates
[04-Jan-2010 12:27:26] <rmatte> in the top dropdown
[04-Jan-2010 12:27:27] <eidolon> cuz i started this 2.5 weeks ago before i went on vaca.
[04-Jan-2010 12:27:28] <rmatte> not the bottom one
[04-Jan-2010 12:27:36] <eidolon> ah, got it
[04-Jan-2010 12:27:39] * eidolon browses
[04-Jan-2010 12:27:54] <eidolon> 1 of 131. :)
[04-Jan-2010 12:28:07] <rmatte> click the show all button at the bottom
[04-Jan-2010 12:28:13] <rmatte> and yes, there will be a lot
[04-Jan-2010 12:28:40] <cluther> eidolon: Your browser history may be of use.
[04-Jan-2010 12:28:40] <eidolon> i found it
[04-Jan-2010 12:28:41] <twm1010> that all templates menu eluded me for a whole day once :|
[04-Jan-2010 12:28:44] <eidolon> no, i got it
[04-Jan-2010 12:28:45] <rmatte> anyways, work your way through that list, that's your best bet
[04-Jan-2010 12:28:50] <eidolon> yep, it's a template.
[04-Jan-2010 12:28:51] <rmatte> cool
[04-Jan-2010 12:29:01] <eidolon> /Devices/Server/Templates/MySQL Slave
[04-Jan-2010 12:29:12] <eidolon> with a data source of MySQL-Slave Status
[04-Jan-2010 12:29:15] <rmatte> twm1010: I stumbled across it like 5 or 6 months after I started using Zenoss, it's fairly elusive
[04-Jan-2010 12:29:15] <eidolon> that's what i was working on
[04-Jan-2010 12:29:19] <eidolon> rmatte, you totally get a cookie.
[04-Jan-2010 12:29:25] <rmatte> w00t!
[04-Jan-2010 12:29:59] <rmatte> as long as it's a ginger_snap
[04-Jan-2010 12:30:50] <eidolon> meh. note to self that datasources should have spaces in their names.
[04-Jan-2010 12:31:09] <rmatte> I never put spaces in the names of the actual datasource
[04-Jan-2010 12:31:28] <rmatte> I change the name of the graphpoint when I'm creating the graph though
[04-Jan-2010 12:31:30] <rmatte> to pretty it up
[04-Jan-2010 12:31:36] <twm1010> i didn't understand the datasource / datapoint relationship until the light bulb went off with multi-graph reports
[04-Jan-2010 12:31:49] <rmatte> dataSourcesShouldBeNamedLikeThisInMyOpinion
[04-Jan-2010 12:32:04] <eidolon> howVeryCamelCase
[04-Jan-2010 12:32:10] <rmatte> hehe
[04-Jan-2010 12:32:18] * eidolon is a java programmer :)
[04-Jan-2010 12:32:27] <eidolon> (among other things)
[04-Jan-2010 12:32:37] <twm1010> once that light bulb went off, it made sense, and i really made some nice multi-graphs
[04-Jan-2010 12:32:44] <rmatte> yup
[04-Jan-2010 12:32:53] <eidolon> yeah, i did a couple of them collecting sales data via an http monitor
[04-Jan-2010 12:33:01] <eidolon> (scraping a source page for data)
[04-Jan-2010 12:35:20] <rmatte> cool
[04-Jan-2010 12:35:36] <rmatte> synthetic transaction essentially
[04-Jan-2010 12:46:31] <Kristopher424> rmatte, is there a way to convert those interface graphs to show MB instead of bits?
[04-Jan-2010 12:51:13] <twm1010> yep
[04-Jan-2010 12:51:32] <Kristopher424> RPN?
[04-Jan-2010 12:51:38] <twm1010> yep :)
[04-Jan-2010 12:52:38] <Kristopher424> I'm assuming it uses bytes, because current RPN is 8, *. add 1000000, /?
[04-Jan-2010 12:52:49] <twm1010> uh, lemme look at mine real quick
[04-Jan-2010 12:52:54] <Kristopher424> Cool, thanks.
[04-Jan-2010 12:54:20] <twm1010> i think i did... 8192,*
[04-Jan-2010 12:54:29] <twm1010> or... x 8 x 1024?
[04-Jan-2010 12:55:01] <cluther> 1024,/
[04-Jan-2010 12:55:22] <Kristopher424> 8, *, 1024, /? or is it just 1024/?
[04-Jan-2010 12:55:28] <twm1010> well, you want to go from bits to byes, and then from bytes to megabytes, right?
[04-Jan-2010 12:55:43] <cluther> It's already bytes.. so you just need to divide by 1024.
[04-Jan-2010 12:55:51] <Kristopher424> Ahh, it's already in bytes b/c 8, * is doing bytes to bits.
[04-Jan-2010 12:56:13] <twm1010> ahhh right, yeah im looking at a different one, sorry, mr brain hurts a bit today...
[04-Jan-2010 12:56:52] <twm1010> in my case i was going the other direction
[04-Jan-2010 12:57:04] <twm1010> KBytes to Mbits
[04-Jan-2010 12:57:06] <eidolon> OK|Seconds_behind=0
[04-Jan-2010 12:57:08] <eidolon> BOOYAH
[04-Jan-2010 12:57:16] <eidolon> that was from a zentestcommand called on an applied template.
[04-Jan-2010 12:57:20] <eidolon> i totally rock. *nod*
[04-Jan-2010 12:58:38] <Kristopher424> Uh... I broke it?
[04-Jan-2010 12:59:10] <Kristopher424> added 1024,/ ... now the values on the left show 1.0k, 2.0k and the far left shows bits/sec
[04-Jan-2010 12:59:50] <rmatte> Kristopher424: hunh, it grabs them in bits and the %s at the end of the format for the graphpoint automatically converts them to Kb, Mb, or Gb
[04-Jan-2010 13:00:04] <rmatte> They are automatically converted by default
[04-Jan-2010 13:01:05] <rmatte> For example, here are graphs from a ZenPack of mine: http://dmon.org/graphics/zenoss/windowsperf.png
[04-Jan-2010 13:01:05] <Kristopher424> Okay, so leave the RPN at 8,* and the %s at the end of "format" automatically chooses the correct value?
[04-Jan-2010 13:01:29] <rmatte> notice how the memory is shown as 33.55G, the actual value being collected is in bytes, Zenoss does the conversion by default
[04-Jan-2010 13:01:30] <Kristopher424> so the M on the values on the left stands for Meg?
[04-Jan-2010 13:01:35] <Kristopher424> not Million?
[04-Jan-2010 13:01:35] <rmatte> yes
[04-Jan-2010 13:01:44] <rmatte> well, a million technically is a meg
[04-Jan-2010 13:01:50] <rmatte> a meg is a million bytes
[04-Jan-2010 13:02:04] <rmatte> 1000000
[04-Jan-2010 13:02:32] <rmatte> but if it were to go up to 1000000000 it would be displayed as 1G
[04-Jan-2010 13:02:36] <rmatte> so yes
[04-Jan-2010 13:02:46] <Kristopher424> Gotchya. Thanks for explaining that!
[04-Jan-2010 13:02:52] <Kristopher424> (The RPN Part )
[04-Jan-2010 13:03:04] <rmatte> the format string for the graphpoint will be %5.2lf%s
[04-Jan-2010 13:03:15] <rmatte> the %s part ensures that Zenoss does the conversion before displaying the data
[04-Jan-2010 13:03:21] <Kristopher424> That's what I have. .. What does %5.2lf do?
[04-Jan-2010 13:03:41] <twm1010> its the formatting, hold on, the RRD website has good documentation on that
[04-Jan-2010 13:03:59] <rmatte> it's formatting, the f defines it as a float value, the .2 represents how many decimal places are shown after the number
[04-Jan-2010 13:04:21] <rmatte> so if you did %5.0lf you'd see no decimal points (it would round the value)
[04-Jan-2010 13:04:28] <Kristopher424> Sweet
[04-Jan-2010 13:05:52] <OiScout> Has anyone been able to get the WMI MSSQL zenpack working?
[04-Jan-2010 13:05:59] <twm1010> yeah, basically that's what the documentation says
[04-Jan-2010 13:06:15] <OiScout> Everytime I apply the template to a machine all WMI graphing stops
[04-Jan-2010 13:06:19] <twm1010> there is quite a bit more you can do with the graphs with custom commands, hide legends, change the anti-aliasing, etc
[04-Jan-2010 13:06:26] <rmatte> you don't require RRD for what you're doing
[04-Jan-2010 13:06:48] <rmatte> oh wait, I'm thinking RPN, nevermind
[04-Jan-2010 13:06:56] <rmatte> I meant you don't need to use RPN for the conversion
[04-Jan-2010 13:07:30] <rmatte> RPN is useful for converting from celsius to farenheit or whatever
[04-Jan-2010 13:07:47] <rmatte> or for adding graphpoints together
[04-Jan-2010 13:08:12] <twm1010> ack, i can't find the doc page on the RRD website that shows all the different things
[04-Jan-2010 13:09:01] <Kristopher424> RPN for C to F : /,10,1.8,*,32,+
[04-Jan-2010 13:09:07] <Kristopher424> rmatte helped me with that one :) haha
[04-Jan-2010 13:09:19] <rmatte> yup
[04-Jan-2010 13:09:25] <rmatte> why the 10 though?
[04-Jan-2010 13:09:53] <Kristopher424> Oh, my bad...The radios that I added that conversion to have a funky *10 before SNMP sends the data
[04-Jan-2010 13:10:00] <rmatte> hehe
[04-Jan-2010 13:10:25] <twm1010> here it is
[04-Jan-2010 13:10:25] <rmatte> it should just be ,1.8,*,32,+
[04-Jan-2010 13:10:25] <twm1010> http://oss.oetiker.ch/rrdtool/doc/rrdgraph_graph.en.html
[04-Jan-2010 13:10:28] <rmatte> I believe
[04-Jan-2010 13:10:53] <twm1010> for manipulating the format
[04-Jan-2010 13:14:39] <rmatte> twm1010: some of that stuff should be included as options in the UI though, such as hiding the legend
[04-Jan-2010 13:18:59] <eidolon> hrm... okay... so i have a custom template that's talking toa mysql server.
[04-Jan-2010 13:19:10] <eidolon> the problem is, the template needs the mysql port number to talk to to run the cmd
[04-Jan-2010 13:19:15] <eidolon> tht's not a zproperty
[04-Jan-2010 13:19:35] <eidolon> can i add a custom zproperty so that when this template is imported, it can have a customized setting?
[04-Jan-2010 13:25:04] <rmatte> you could use a custom property instead of a zProperty
[04-Jan-2010 13:26:02] <rmatte> If you go to a device, and select More -> Custom from the dropdown you'll see custom properties, the default being cDateTest
[04-Jan-2010 13:26:30] <rmatte> If you click on Devices on the left hand menu and select More -> Custom Schema you can create one
[04-Jan-2010 13:26:35] <rmatte> make sure the name of it starts with a c
[04-Jan-2010 13:26:56] <rmatte> you can then use it in your template by calling on: here.cWhateverYouNamedIt
[04-Jan-2010 13:28:35] <rmatte> you could also add a zproperty
[04-Jan-2010 13:28:41] <rmatte> with the following python code in zendmd
[04-Jan-2010 13:28:49] <rmatte> >>> ch = devices
[04-Jan-2010 13:28:49] <rmatte> >>> ch._setProperty('zTest', 'Test', type='string')
[04-Jan-2010 13:28:49] <rmatte> >>> commit()
[04-Jan-2010 13:29:14] <rmatte> you'd want to have that code (minus the commit) in your __init__.py for your ZenPack if you want it added automatically
[04-Jan-2010 13:49:48] <eidolon> a chart that says "cur: nan" - means it hasn't received any data yet, yes? (performance chart)
[04-Jan-2010 13:50:09] <theacolyte-> correct
[04-Jan-2010 13:50:12] <theacolyte-> give it 15 minutes or so
[04-Jan-2010 13:50:58] <eidolon> hm. k.
[04-Jan-2010 13:52:57] <twm1010> it means it doesn't have enough values to generate a graph
[04-Jan-2010 13:53:04] <twm1010> but it does mean its getting data
[04-Jan-2010 13:53:14] <theacolyte-> Right, I guess that's a better way of putting it
[04-Jan-2010 13:53:14] <twm1010> otherwise it would say RRD file not found or something like that
[04-Jan-2010 13:53:59] <theacolyte-> I have lost some steam on my WMI zenpacks
[04-Jan-2010 13:54:04] <theacolyte-> It's so much work
[04-Jan-2010 13:54:16] <theacolyte-> Tempted to just go back to up/down and let Event Logging do its trick
[04-Jan-2010 14:07:41] <rmatte> twm1010: you're fairly decently versed in RRD I'm assuming?
[04-Jan-2010 14:09:20] <rmatte> well, here's a general RRD question, hopefully someone will know the answer
[04-Jan-2010 14:09:29] <crow2> 42
[04-Jan-2010 14:09:37] crow2 is now known as coofamani
[04-Jan-2010 14:09:44] <rmatte> crow2: very funny :P
[04-Jan-2010 14:10:36] <rmatte> but yeh, when I view one of my graphs for CPU over a day, the max value is 97.1 percent, when I view it over a week (including that same day) the max value is 61.94... my understanding was that the max value should reflect the maximum value that was seen during that period, or am I missing something here?
[04-Jan-2010 14:10:47] <rmatte> it's more like it's doing some sort of funky averaging to come out with the max value
[04-Jan-2010 14:11:37] <theacolyte-> I've noticed that with anything using RRD
[04-Jan-2010 14:11:42] <theacolyte-> if you ever find a way to explain it let me know
[04-Jan-2010 14:11:46] <theacolyte-> cacti will do that, too
[04-Jan-2010 14:11:55] <rmatte> eugh, that's kind of stupid lol
[04-Jan-2010 14:12:30] <RobertLaptop> It likely has to do with timeframe's
[04-Jan-2010 14:13:29] <rmatte> RobertLaptop: If you take the max value during 1 day, then expand the query to include 6 more days but still include that one day, the max value should be the same or higher
[04-Jan-2010 14:13:42] <rmatte> RobertLaptop: doesn't make sense for it to be any lower
[04-Jan-2010 14:15:50] <RobertLaptop> If I understand the problem right. You are thinking it is looking at a signal data point. Instead if I understand what RDD is doing is the 1 day view is taking the range and breaking the data into X number of pieces. Each piece will be 5 min segments. When you look at a week. Each segment is say and hour. So it will give you the max that was held for 1 hour.
[04-Jan-2010 14:16:21] <RobertLaptop> vers 5 min.
[04-Jan-2010 14:18:10] <rmatte> let me show you a screenshot...
[04-Jan-2010 14:20:25] <rmatte> http://dmon.org/graphics/zenoss/rrdmax.png
[04-Jan-2010 14:20:48] <rmatte> see how the CPU utilization is clearly at 80ish percent over a quarter of the period selected
[04-Jan-2010 14:20:55] <rmatte> but the max value shows as 61.94
[04-Jan-2010 14:21:17] <rmatte> actually, hmmm
[04-Jan-2010 14:21:26] <rmatte> oh, I think I see what you mean
[04-Jan-2010 14:22:19] <rmatte> I still find it very strange that slight spikes are not taken in to account
[04-Jan-2010 14:22:44] <rmatte> so when it creates those 1 hour slices is it doing that based on an average then?
[04-Jan-2010 14:24:31] <rmatte> maybe it has something to do with it viewing the data as though it were consolidated
[04-Jan-2010 14:27:57] <cluther> http://dev.zenoss.org/trac/ticket/4914
[04-Jan-2010 14:28:38] <rmatte> ah
[04-Jan-2010 14:30:59] <rmatte> cluther: that ticket was closed...
[04-Jan-2010 14:31:04] <rmatte> and a patch was applied in 2.4.2?
[04-Jan-2010 14:32:27] <rmatte> so technically if I apply patch 13304 it should fix it?
[04-Jan-2010 14:38:02] <mrayzenoss1> well, 13304 wasn't targeted for the 2.4.x branch, so it may not apply cleanly (YMMV)
[04-Jan-2010 14:38:11] mrayzenoss1 is now known as mrayzenoss
[04-Jan-2010 14:38:56] <rmatte> it did apply cleanly
[04-Jan-2010 14:39:05] <rmatte> now to see if it actually works for the reports too and not just the graphs
[04-Jan-2010 14:40:32] <rmatte> apparently it doesn't work for the report
[04-Jan-2010 14:54:30] <rmatte> yeh ok, so it fixes the graphs but the CPU report still has the problem, hmmm
[04-Jan-2010 14:58:27] <RobertLaptop> Are you able to see what he is doing to clean up the graph. Find the issue is half the battle most of the time. The other half is generally getting the core team to accept a fix.
[04-Jan-2010 15:00:16] <rmatte> well, here are the changes that were made to fix it: http://dev.zenoss.org/trac/changeset/13304
[04-Jan-2010 15:01:13] <RobertLaptop> I can't read a lick of python.
[04-Jan-2010 15:02:07] <rmatte> basically he's just pulling in the values different, but the report is using the getRRDValue function, so I have no idea how you'd make that function work the same way
[04-Jan-2010 15:02:45] <RobertLaptop> Sorry no clue.
[04-Jan-2010 15:03:08] <RobertLaptop> Just though out ideas hoping it might click for you.
[04-Jan-2010 15:03:36] <rmatte> nah, it's not something simple
[04-Jan-2010 15:05:20] <rmatte> but oh well, that patch works well so I'll apply that on all my boxes tomorrow night during my change window
[04-Jan-2010 15:05:31] <theacolyte-> ooh
[04-Jan-2010 15:05:35] <theacolyte-> was there an answer to that RRD stuff?
[04-Jan-2010 15:05:49] <rmatte> yes
[04-Jan-2010 15:05:59] <rmatte> http://dev.zenoss.org/trac/ticket/4914
[04-Jan-2010 15:06:03] <rmatte> it's fixed in 2.5.x
[04-Jan-2010 15:06:22] <rmatte> I just tested patch 13304 against 2.4.5 and it worked fine
[04-Jan-2010 15:09:00] <rmatte> But it doesn't fix the values in the reports
[04-Jan-2010 15:09:10] <rmatte> My CPU report is still messed up
[04-Jan-2010 15:24:19] <twm1010> rmatte: around?
[04-Jan-2010 15:27:02] <rmatte> yeh
[04-Jan-2010 15:27:16] <rmatte> once of our engineers enabled syslogging on a couple firewalls for a new client and pointed it to zenoss
[04-Jan-2010 15:27:22] * rmatte slaps forehead
[04-Jan-2010 15:27:27] <twm1010> I'll throw this out there...
[04-Jan-2010 15:27:38] <rmatte> I noticed it after about 18 minutes since the CPU spiked, over 40000 events in 18 minutes
[04-Jan-2010 15:27:40] <twm1010> I'd like to use Zenoss to get an inventory of port usage on our core switches
[04-Jan-2010 15:27:41] <rmatte> shoot
[04-Jan-2010 15:27:53] <twm1010> i can easily go to each device, and look
[04-Jan-2010 15:28:01] <twm1010> but i'd like to export it so i can sort and such
[04-Jan-2010 15:28:16] <rmatte> so what's wrong with the interface utilization report?
[04-Jan-2010 15:28:43] * twm1010 discovers the interface utilization report
[04-Jan-2010 15:28:47] <rmatte> lol
[04-Jan-2010 15:29:35] <twm1010> its taking a sweet time to come up, but that fits the bill
[04-Jan-2010 15:29:48] <twm1010> now the MAC addresses shown on the interface page, is that the interface of the port, or the device using it
[04-Jan-2010 15:30:53] <rmatte> the MAC of the port
[04-Jan-2010 15:31:06] <twm1010> that's what i figured, they looked too similar.
[04-Jan-2010 15:31:16] <twm1010> the differences seem to go with the blades
[04-Jan-2010 15:31:20] <rmatte> and yeh, the interface report takes the longest to generate out of any report
[04-Jan-2010 15:31:32] <twm1010> once its up, can i narrow it to a device or class?
[04-Jan-2010 15:32:08] <theacolyte-> rmatte: Sounds like they need to adjust a few things
[04-Jan-2010 15:32:19] <theacolyte-> 40k messages in 18 minutes is something configured really wrong
[04-Jan-2010 15:32:20] <zgold> Hello! I'm trying to set a MinMaxThreshold where the Max Value is actually one of the datapoints in the template. Is this possible? I'm trying to read up on TALES and cannot seem to find any good answer
[04-Jan-2010 15:33:56] <rmatte> twm1010: http://pastebin.com/m5f9e3733
[04-Jan-2010 15:34:09] <rmatte> twm1010: that's a script that you can use to generate the report from the commandline
[04-Jan-2010 15:34:20] <rmatte> you can just do > whatever.csv to send it to a csv file
[04-Jan-2010 15:34:34] <rmatte> you might find it useful
[04-Jan-2010 15:34:53] <twm1010> sweet! thanks
[04-Jan-2010 15:35:57] <rmatte> np
[04-Jan-2010 15:36:18] <rmatte> I'm going to publish all of my report scripts in the near future
[04-Jan-2010 15:36:18] <twm1010> i really don't need the utilization though
[04-Jan-2010 15:36:25] <rmatte> I wrote them to be used for automated reporting that we do
[04-Jan-2010 15:36:26] <twm1010> just the name, administrative, and operational status
[04-Jan-2010 15:36:45] <rmatte> we basically have a php script that spits out a monthly report automatically
[04-Jan-2010 15:36:51] <rmatte> completely formatted with graphs and stuff
[04-Jan-2010 15:37:07] <rmatte> ah
[04-Jan-2010 15:37:17] <rmatte> you might have to write something a little more custom then
[04-Jan-2010 15:37:26] <rmatte> it wouldn't be overly difficult to do
[04-Jan-2010 15:37:40] <rmatte> probably like 4 or 5 lines of code
[04-Jan-2010 15:38:08] <rmatte> If I have a second I'll throw something together
[04-Jan-2010 15:38:20] <rmatte> (I don't have a second right now lol)
[04-Jan-2010 15:47:36] <twm1010> sure, maybe tomorrow or something... this will do for now
[04-Jan-2010 15:47:42] <rmatte> k
[04-Jan-2010 15:48:06] <rmatte> my workload is insane for the next couple of days, last week I had almost nothing to do and this week I'm getting things dropped on me left and right
[04-Jan-2010 15:48:10] <rmatte> (when it rains it pours)
[04-Jan-2010 15:48:29] <twm1010> im right there with ya man, don't i usually chat about servers? :)
[04-Jan-2010 15:50:37] <rmatte> hehe
[04-Jan-2010 15:53:18] <twm1010> I wish the reports would come up asking for criteria first, before running
[04-Jan-2010 15:53:30] <twm1010> instead of making me wait for all 1500 interfaces to calculate averages
[04-Jan-2010 15:53:38] <twm1010> just to get the 300 i need
[04-Jan-2010 15:54:08] <rmatte> the reports are quite limited at the moment
[04-Jan-2010 15:54:16] <theacolyte-> You know you are using Cisco stuff too option when you try to enable on your linux boxes
[04-Jan-2010 15:54:21] <rmatte> I actually opened a ticket for asking for criteria first
[04-Jan-2010 15:54:25] <twm1010> lol :)
[04-Jan-2010 15:54:26] <theacolyte-> administrator@zenoss:~$ en
[04-Jan-2010 15:54:26] <theacolyte-> -bash: en: command not found
[04-Jan-2010 15:54:29] <theacolyte-> :(
[04-Jan-2010 15:54:40] <rmatte> it's stupid, you have to wait for the report to load, then select criteria, wait for it to load, then click show all, wait for it to load
[04-Jan-2010 15:54:42] <twm1010> kind of like typing ls at a windows command prompt?
[04-Jan-2010 15:54:51] <theacolyte-> Yeah
[04-Jan-2010 15:54:55] <rmatte> so it essentially generates the report 3 times before you get a proper full view
[04-Jan-2010 15:54:56] <theacolyte-> I do that constantly, and I'm a windows admin
[04-Jan-2010 15:55:05] <rmatte> also, just a warning, compare the output CSV to the report
[04-Jan-2010 15:55:07] <twm1010> i might try my hand at it actually, i think i've tried once before
[04-Jan-2010 15:55:19] <rmatte> it used to output from the last 7 days rather than what you actually selected
[04-Jan-2010 15:55:24] <rmatte> I think they fixed that, but check anyways
[04-Jan-2010 15:55:29] <twm1010> rmatte: sure, it doesn't just export what it just calculated, it calculates it again doesn't it
[04-Jan-2010 15:55:40] <rmatte> I found like 5 or 6 bugs with the reports system the first 15 minutes that I used it :P
[04-Jan-2010 15:55:48] <twm1010> you'd think an export would be instant
[04-Jan-2010 15:56:02] <rmatte> yeh, the reports should cache after the first run
[04-Jan-2010 15:56:15] <rmatte> it's brutal that they don't
[04-Jan-2010 15:56:39] <rmatte> and it probably wouldn't be overcomplicated to do
[04-Jan-2010 15:56:42] <twm1010> where under zenhome are the report scripts?
[04-Jan-2010 15:56:53] <rmatte> $ZENHOME/Products/ZenReports
[04-Jan-2010 15:57:08] <rmatte> the formatting is done by .rpt files in the reports directory
[04-Jan-2010 15:57:20] <rmatte> the actual data collection is done by .py files in the plugins directory
[04-Jan-2010 15:57:25] <rmatte> plugins is what you'd want
[04-Jan-2010 15:58:37] <twm1010> right... i tried this once before I think
[04-Jan-2010 15:58:59] <rmatte> I've modified some of my reports slightly
[04-Jan-2010 15:59:08] <rmatte> To properly use datapoints in some of my custom templates
[04-Jan-2010 15:59:58] <rmatte> If you modify a reports plugin you need to restart Zenoss
[04-Jan-2010 16:00:08] <rmatte> (from what I recall)
[04-Jan-2010 16:00:19] <rmatte> since they are compiled at runtime
[04-Jan-2010 16:00:30] <twm1010> this doesn't look that hard, i can basically see what its doing but i don't know python
[04-Jan-2010 16:01:04] <twm1010> i see where its getting the RRD data, creating the total, etc.... so i can pretty much wipe half of this out
[04-Jan-2010 16:01:20] <rmatte> python is easy enough to read even for people who don't really know it
[04-Jan-2010 16:01:25] <rmatte> I picked it up pretty quickly
[04-Jan-2010 16:01:37] <rmatte> I still have quite a bit to learn but I'm pretty good
[04-Jan-2010 16:01:45] <RobertLaptop> Is there an easy way to move a template from one system to another?
[04-Jan-2010 16:02:08] <rmatte> RobertLaptop: you mean copy a local template?
[04-Jan-2010 16:02:10] <twm1010> RobertLaptop: hrmmm, a quick and dirty zenpack maybe?
[04-Jan-2010 16:02:22] <rmatte> nah, I think he means a local template
[04-Jan-2010 16:02:26] <RobertLaptop> rmatte yes.
[04-Jan-2010 16:02:28] <mrayzenoss> zgold: like this? http://www.thetechchef.com/zenoss-snmp-min-max-update/
[04-Jan-2010 16:02:29] <rmatte> you can't copy local templates to ZenPacks
[04-Jan-2010 16:02:43] <rmatte> I may have a solution for you, give me a sec
[04-Jan-2010 16:02:51] <twm1010> so where it says something like r = Utils.Record(device=d,interface=i,etc,etc) those variables are passed over to the report definition
[04-Jan-2010 16:02:57] <RobertLaptop> I tired creating a zenpack out of it. But the data source did copy.
[04-Jan-2010 16:03:28] <rmatte> ok, RobertLaptop: go to the templates page for that device
[04-Jan-2010 16:03:35] <RobertLaptop> ok
[04-Jan-2010 16:03:38] <RobertLaptop> There
[04-Jan-2010 16:03:44] <rmatte> then add /manage to the end of the URL
[04-Jan-2010 16:03:47] <rmatte> and hit enter
[04-Jan-2010 16:03:52] <rmatte> you'll be brought in to ZMI
[04-Jan-2010 16:04:05] <rmatte> you should see an object listed that matches the name of the template
[04-Jan-2010 16:04:20] <rmatte> correct?
[04-Jan-2010 16:04:50] <RobertLaptop> Doing now.
[04-Jan-2010 16:04:57] <zgold> mrayzenoss: ah, thats a bit more round-a-bout than I was expecting
[04-Jan-2010 16:05:43] <rmatte> RobertLaptop: actually hold on... I want to be absolutely certain that we're on the same page here...
[04-Jan-2010 16:06:04] <rmatte> when you say move a template from 1 system to another, are you talking about moving it between Zenoss instances or between devices on a Zenoss instance?
[04-Jan-2010 16:06:17] <RobertLaptop> Zenoss Instance.
[04-Jan-2010 16:06:21] <mrayzenoss> zgold: there's also predictive threshold http://community.zenoss.org/docs/DOC-3408
[04-Jan-2010 16:06:31] <rmatte> oh, then you'll want to make a ZenPack
[04-Jan-2010 16:06:40] <rmatte> I thought you were asking how to copy a local template from 1 device to another
[04-Jan-2010 16:06:44] <zgold> mrayzenoss: I was hoping to just be able to do ${curTemplate/template_name_datapoint_name} or something
[04-Jan-2010 16:07:10] <Apachez> how come zenoss uses pyhon 2.4 instead of 2.6 ?
[04-Jan-2010 16:07:24] <rmatte> Apachez: because the version of Zope that it uses relies on it
[04-Jan-2010 16:07:28] <RobertLaptop> rmatte how to I include datapoints into the templates inside the new zenpack?
[04-Jan-2010 16:07:32] <rmatte> Apachez: version 2.6 will have new Zope and new Python
[04-Jan-2010 16:07:52] <rmatte> RobertLaptop: you can include the whole template...
[04-Jan-2010 16:08:04] <rmatte> RobertLaptop: go to the templates page, put a checkmark next to the template
[04-Jan-2010 16:08:14] <rmatte> RobertLaptop: them from the dropdown select Add to ZenPack
[04-Jan-2010 16:08:17] <rmatte> then*
[04-Jan-2010 16:08:30] <RobertLaptop> So far what I did the 1st time.
[04-Jan-2010 16:08:45] <rmatte> so what's the issue then?
[04-Jan-2010 16:09:40] <RobertLaptop> It included the template but not the data points?
[04-Jan-2010 16:10:01] <rmatte> eugh, no
[04-Jan-2010 16:10:11] <rmatte> when you include the template it automatically knows to include everything in the template
[04-Jan-2010 16:10:14] <RobertLaptop> Let me retry.
[04-Jan-2010 16:10:32] <RobertLaptop> It isn't showing up in the new install for some reason :(
[04-Jan-2010 16:10:39] <rmatte> even if it doesn't look like the datapoints are there, export it anyways and install it
[04-Jan-2010 16:10:44] <rmatte> it'll work
[04-Jan-2010 16:10:56] <rmatte> oh?
[04-Jan-2010 16:10:56] <RobertLaptop> ok
[04-Jan-2010 16:11:08] <RobertLaptop> Retring
[04-Jan-2010 16:11:11] <rmatte> when you browse to the ZenPack you should see the stuff listed
[04-Jan-2010 16:11:32] <rmatte> under the ZenPack provides section
[04-Jan-2010 16:11:54] <RobertLaptop> ok. So I did it right. I will redue it.
[04-Jan-2010 16:20:47] <twm1010> what would be the zenDMD command to list the attributes of an interface
[04-Jan-2010 16:21:08] <twm1010> or is that a loaded question :)
[04-Jan-2010 16:22:41] <theacolyte-> aslkdjkasd
[04-Jan-2010 16:22:51] <theacolyte-> just noticed that I'm getting NAN's on my interfaces graphs on my ASA
[04-Jan-2010 16:22:55] <theacolyte-> that shouldn't be happening
[04-Jan-2010 16:23:28] <theacolyte-> it's grabbing ip/mac/link state/etc but giving me NAN's
[04-Jan-2010 16:24:42] <theacolyte-> and testing the template it does retrieve #'s
[04-Jan-2010 16:26:57] <rmatte> what's the interface type?
[04-Jan-2010 16:27:13] <rmatte> actually, I might know what's causing that...
[04-Jan-2010 16:27:40] <rmatte> go in to the zProperties for that device
[04-Jan-2010 16:27:51] <rmatte> you'll see a zProperty called zMaxOIDPerRequest
[04-Jan-2010 16:27:55] <rmatte> is it set to 40?
[04-Jan-2010 16:31:19] <rmatte> twm1010: there is no single command
[04-Jan-2010 16:32:05] <rmatte> twm1010: you would need to set a variable to equal the context of the interface (like evt equals the context of an event in a transform)
[04-Jan-2010 16:32:18] <rmatte> then you'd do: int. and hit tab to display the possible calls
[04-Jan-2010 16:32:45] <rmatte> twm1010: you're just wanting interface name and oper/admin status for monitored interfaces, correct?
[04-Jan-2010 16:32:45] <twm1010> yeah, i found a forum post here
[04-Jan-2010 16:33:11] <rmatte> I have some spare time now, let me hash something together
[04-Jan-2010 16:33:13] <twm1010> interface name, oper/admin status, link speed owuld be nice
[04-Jan-2010 16:33:27] <twm1010> hopefully i can step myself through and understand what you're doing
[04-Jan-2010 16:33:41] <twm1010> im looking at some zendmd examples to change monitoring
[04-Jan-2010 16:37:22] <theacolyte-> rmatte: looking now, sorry had to AFK
[04-Jan-2010 16:38:05] <theacolyte-> rmatte: it is 40
[04-Jan-2010 16:38:18] <rmatte> theacolyte-: change it to 10
[04-Jan-2010 16:38:40] <theacolyte-> Well.. that's not logical,, do I need to wait?
[04-Jan-2010 16:38:42] <rmatte> theacolyte-: then it'll work, Cisco ASAs are very particular about packet size, so 40 OIDs in 1 request is too much
[04-Jan-2010 16:39:00] <theacolyte-> ah.
[04-Jan-2010 16:39:02] <rmatte> theacolyte-: after you change it you need to wait 3 polling cycles for the graph to start getting values (15 minutes)
[04-Jan-2010 16:39:08] <theacolyte-> Man
[04-Jan-2010 16:39:13] <theacolyte-> Thanks
[04-Jan-2010 16:39:17] <theacolyte-> again
[04-Jan-2010 16:39:23] <rmatte> np
[04-Jan-2010 16:39:37] <theacolyte-> Now to figure out how to get interface graphs on the perf tab!
[04-Jan-2010 16:39:47] <theacolyte-> not sure if I'm that hot on having to click around
[04-Jan-2010 16:39:48] <twm1010> you won't
[04-Jan-2010 16:39:55] <twm1010> perf tab is for device performance, not interfaces
[04-Jan-2010 16:40:19] <theacolyte-> Well, interface throughput has every bit to do with performance, but yeah, I figured that's how it was designed
[04-Jan-2010 16:40:30] <rmatte> you won't be able to see them on the perf tab
[04-Jan-2010 16:40:36] <twm1010> interface graphs are on the OS tab under each interface or when you create graph reports or multi-graph reports
[04-Jan-2010 16:40:38] <rmatte> the graphs are generated based on the interface index
[04-Jan-2010 16:40:45] <rmatte> so they are categorized under each interface
[04-Jan-2010 16:40:49] <rmatte> there's no way around that
[04-Jan-2010 16:40:55] <theacolyte-> Yeah
[04-Jan-2010 16:41:05] <rmatte> you could manually create a graph to poll certain OIDs, but it's really not worth it
[04-Jan-2010 16:41:18] <theacolyte-> Yeah, that really isn't
[04-Jan-2010 16:41:25] <twm1010> and those oids are subject to change from time to time
[04-Jan-2010 16:41:31] <rmatte> ok, I've got the tricky parts of this script done... now I just need to format it correctly
[04-Jan-2010 16:41:34] <twm1010> if you add/remove devices potentially
[04-Jan-2010 16:41:53] <twm1010> rmatte: I guess since I don't know the schema of things, I just get lost at what the variables are.
[04-Jan-2010 16:42:06] <rmatte> here's what I have so far:
[04-Jan-2010 16:42:08] <rmatte> http://pastebin.com/m13e0eab
[04-Jan-2010 16:42:22] <rmatte> the operstatus and admin status show as 0, 1, or 2
[04-Jan-2010 16:42:29] <rmatte> I need to translate those to make sense when output
[04-Jan-2010 16:42:36] <rmatte> the rest is basically done
[04-Jan-2010 16:42:46] <rmatte> now if I can remember what each one stands for
[04-Jan-2010 16:43:07] <twm1010> lol, well, if we run the script, i can compare the output to what the gui shows
[04-Jan-2010 16:43:35] <twm1010> its going to be monitored, unmonitored, and... locked? or ignored maybe?
[04-Jan-2010 16:43:44] <rmatte> ah, 1 is up and 2 is down
[04-Jan-2010 16:43:51] <twm1010> k
[04-Jan-2010 16:44:05] <rmatte> ok, should have this done in a few minutes
[04-Jan-2010 16:44:36] <sergeymasushko> hi guys. can you help me to investigate one issue with process monitoring:
[04-Jan-2010 16:44:36] <sergeymasushko> zenoss reported me that 'vim' process is not running... I do not see that process under the 'OS' tab of the device. In the zenprocess.log I found only the following: http://pastebin.com/m45f9c684
[04-Jan-2010 16:44:36] <sergeymasushko> where from come that event?
[04-Jan-2010 16:46:01] <twm1010> Is the process running?
[04-Jan-2010 16:46:23] <theacolyte-> What a ridiculous question
[04-Jan-2010 16:46:28] <twm1010> :)
[04-Jan-2010 16:46:51] <sergeymasushko> dunno. I don't want to monitor it at all =)
[04-Jan-2010 16:46:52] <theacolyte-> hehe, sorry, I'm feeling slightly sarcastic today
[04-Jan-2010 16:47:38] <twm1010> i would guess perhaps you added vim to processes and set monitor to true?
[04-Jan-2010 16:47:42] <twm1010> and then removed it? or what?
[04-Jan-2010 16:48:54] <sergeymasushko> no. I do not monitor that process earlier... and delete nothing recently (before that event appears)
[04-Jan-2010 16:48:54] <rmatte> hmmm, this script is giving me grief
[04-Jan-2010 16:49:24] <rmatte> ah, got it
[04-Jan-2010 16:51:53] <rmatte> twm1010: your script: http://pastebin.com/m5786b8f0
[04-Jan-2010 16:52:12] <rmatte> save it to a file, run it, and redirect
[04-Jan-2010 16:52:21] <rmatte> so ./script.py > whatever.csv
[04-Jan-2010 16:52:30] <sergeymasushko> here is a list of all processes which I monitor http://pastebin.com/ma1505e1
[04-Jan-2010 16:52:33] <theacolyte-> rmatte: that worked
[04-Jan-2010 16:52:34] <rmatte> it'll only show monitored interfaces
[04-Jan-2010 16:52:44] <theacolyte-> KNow how to do a 95 percentile? :)
[04-Jan-2010 16:53:00] <twm1010> yeah i just ran what you pasted earlier by hand :)
[04-Jan-2010 16:53:01] <rmatte> theacolyte-: no, but I believe there was a ZenPack released recently for that
[04-Jan-2010 16:53:15] <theacolyte-> I'll take a look
[04-Jan-2010 16:53:44] <rmatte> hmmm, it wasn't released, but I could have swore something was in the works
[04-Jan-2010 16:53:53] * rmatte tries to remember where he saw that
[04-Jan-2010 16:54:03] <twm1010> that script will be my plugin, right?
[04-Jan-2010 16:54:28] <rmatte> twm1010: I wouldn't even try to put that in to the UI
[04-Jan-2010 16:54:39] <rmatte> twm1010: you'd need a custom rpt file written as well, just run the script by hand
[04-Jan-2010 16:54:56] <twm1010> sure, that can come later with some mucking... ok let me try
[04-Jan-2010 16:54:57] <rmatte> throw it on the box, save it as a .py, chmod +x it, run it as the zenoss user and redirect the output to a file
[04-Jan-2010 16:55:14] <rmatte> then grab the file and load it in to excel or whatever
[04-Jan-2010 16:55:54] <sergeymasushko> guys: any ideas?
[04-Jan-2010 16:56:25] <rmatte> theacolyte-: here's a thread on 95th percentile, it might be useful: http://community.zenoss.org/message/30048#30048
[04-Jan-2010 16:56:31] <twm1010> sergeymasushko: just delete the process?
[04-Jan-2010 16:56:38] <rmatte> theacolyte-: ask Matt whether or not there was a ZenPack in the works, thought there was
[04-Jan-2010 16:56:41] <rmatte> I could be mistaken though
[04-Jan-2010 16:57:08] <rmatte> sergeymasushko: sorry, what's the issue again?
[04-Jan-2010 16:57:13] <rmatte> oh wait nevermind
[04-Jan-2010 16:57:22] <RobertLaptop> rmatte I redid the same steps and I see the datasource in the installed pack not sure why it didn't work the 1st time.
[04-Jan-2010 16:57:27] <sergeymasushko> twm1010: thanks for the suggestion.... but as I said I _do not see_ that process under the 'OS' tab of the device!!!
[04-Jan-2010 16:57:36] <twm1010> rmatte: that works, much appreciated
[04-Jan-2010 16:57:37] <rmatte> sergeymasushko: when was the last time you restarted Zenoss?
[04-Jan-2010 16:57:54] <Diddi> howdy, iirc it is possible to visual bandwidth usage in the google map portlet? but I just can't recall any papers showing how it's done (:
[04-Jan-2010 16:57:58] <rmatte> sergeymasushko: before you added that device I presume?
[04-Jan-2010 16:57:59] <sergeymasushko> rmatte: it's hard to say
[04-Jan-2010 16:58:39] <twm1010> rmatte: hrmmm... need to add device name before interface as well
[04-Jan-2010 16:58:46] <rmatte> sergeymasushko: ok, one thing to know is that zenprocess is really really stupid (and needs to be rewritten), once it's configured it monitors fine, but after doing any changes I always find that I have to restart zenprocess to get it to monitor properly
[04-Jan-2010 16:59:17] <rmatte> sergeymasushko: what probably happened is that you added the device, someone happened to be running vim at the time, so it picked up on that for whatever reason, then you removed the process from the OS tab but zenprocess still thinks it's there
[04-Jan-2010 16:59:28] <rmatte> sergeymasushko: restart zenprocess and it should solve your problem
[04-Jan-2010 16:59:39] <rmatte> (you don't need to restart zenoss entirely, just zenprocess)
[04-Jan-2010 16:59:42] <twm1010> sergeymasushko: I think rmatte is right
[04-Jan-2010 16:59:51] <rmatte> twm1010: oh right, my bad... one sec
[04-Jan-2010 17:00:11] <sergeymasushko> ok. I'll try it
[04-Jan-2010 17:00:12] <sergeymasushko> thanks
[04-Jan-2010 17:01:04] <rmatte> twm1010: http://pastebin.com/m5b360181
[04-Jan-2010 17:01:19] <rmatte> sergeymasushko: no problem
[04-Jan-2010 17:03:56] <twm1010> rmatte: perfect
[04-Jan-2010 17:04:29] <sergeymasushko> btw. guys please look at my second pastebin... is it correct way to define processes to monitor?
[04-Jan-2010 17:05:16] <rmatte> I literally just define them as the name of the process
[04-Jan-2010 17:05:23] <rmatte> like, for named, I just put "named"
[04-Jan-2010 17:05:33] <twm1010> you can use reg. expressions too
[04-Jan-2010 17:05:34] <rmatte> Zenoss is smart enough to find anything matching that process name
[04-Jan-2010 17:05:47] <rmatte> yeh, I just find zenprocess blows at regex
[04-Jan-2010 17:05:49] <theacolyte-> Damn you rmatte
[04-Jan-2010 17:05:49] <twm1010> it actually creates the reg. expression for you, right?
[04-Jan-2010 17:05:52] <theacolyte-> Damn you and always being right
[04-Jan-2010 17:05:56] <theacolyte-> How dare you be such a good resource
[04-Jan-2010 17:06:00] <rmatte> theacolyte-: hehe
[04-Jan-2010 17:06:15] <rmatte> theacolyte-: I used to be the one in here asking the questions, I'm just spreading the wealth
[04-Jan-2010 17:06:22] <theacolyte-> hehe awesome
[04-Jan-2010 17:06:30] <theacolyte-> well, if you ever have aa windows question, shoot :p
[04-Jan-2010 17:06:37] <rmatte> will do
[04-Jan-2010 17:06:57] <rmatte> I figure if I help people out they are more likely to help me in return, so it works in my favour
[04-Jan-2010 17:07:03] <theacolyte-> hehe
[04-Jan-2010 17:07:30] <rmatte> There are also all sorts of little tricks and quirks that I've learned about Zenoss which really aren't documented anywhere, so I like passing those on
[04-Jan-2010 17:07:34] <sergeymasushko> ok. then can you suggest me the way how to define http process to except monitoring of tail process e.g. 'tail -f /bla/bla/bla/httpd-error.log ??
[04-Jan-2010 17:07:48] <twm1010> rmatte: I'll try my hand at adding the current interface description :)
[04-Jan-2010 17:08:15] <rmatte> sergeymasushko: yeh...
[04-Jan-2010 17:08:59] <rmatte> sergeymasushko: when you add the process go to the edit tab and change ignore paramaters to true
[04-Jan-2010 17:09:19] <rmatte> sergeymasushko: that will keep it from looking at options, it will only look at the process name itself
[04-Jan-2010 17:09:24] <rmatte> sergeymasushko: that's the simplest way
[04-Jan-2010 17:09:34] <rmatte> sergeymasushko: restart zenprocess after you make the change
[04-Jan-2010 17:09:41] <twm1010> would that be int.description(), ?
[04-Jan-2010 17:09:41] <sergeymasushko> is there another way?
[04-Jan-2010 17:09:59] <rmatte> sergeymasushko: the way that you're trying, but like I said, I don't find zenprocess does overly well with regex expressions
[04-Jan-2010 17:10:19] <rmatte> If you want to test your regex expressions use this: http://gskinner.com/RegExr/
[04-Jan-2010 17:10:25] <rmatte> awesome tool for testing regex
[04-Jan-2010 17:10:25] <sergeymasushko> that processes are appear and disappear frequently
[04-Jan-2010 17:10:52] <rmatte> what's the issue with having it ignore the parameter?
[04-Jan-2010 17:11:14] <rmatte> the only process you really care about is /usr/sbin/httpd (or whatever the path happens to be)
[04-Jan-2010 17:11:38] <rmatte> by telling to ignore parameters when looking for that process you're telling it to avoid stuff like tail -f bleh
[04-Jan-2010 17:11:45] <sergeymasushko> the paths can be different on different devices
[04-Jan-2010 17:11:49] <rmatte> it'll only find the actual httpd process itself
[04-Jan-2010 17:11:57] <rmatte> that's fine, it's smart enough to know that
[04-Jan-2010 17:12:03] <rmatte> just put "httpd" as the regex
[04-Jan-2010 17:12:13] <rmatte> and tell it to ignore parameters
[04-Jan-2010 17:12:17] <rmatte> I guarantee it'll work
[04-Jan-2010 17:12:44] <sergeymasushko> thanks for explanation
[04-Jan-2010 17:12:48] <rmatte> np
[04-Jan-2010 17:13:26] <rmatte> (don't forget to restart zenprocess after making the changes hehe)
[04-Jan-2010 17:14:02] * theacolyte- can almost retire his cacti install
[04-Jan-2010 17:14:11] <rmatte> hehe
[04-Jan-2010 17:14:19] <sergeymasushko> rmatte: sure :)
[04-Jan-2010 17:14:19] <rmatte> Zenoss does what Cacti does, but better
[04-Jan-2010 17:15:03] <twm1010> rmatte: I don't think if.description is quite what i'm after
[04-Jan-2010 17:15:58] <twm1010> hrmm, wait, sure is, seems to have bad output for some device, but on the whole i got what i need :)
[04-Jan-2010 17:16:05] <twm1010> thanks again, see you in the morning
[04-Jan-2010 17:16:21] <rmatte> it's not if.description, I'm calling on int.getInterfaceName()
[04-Jan-2010 17:16:23] <rmatte> ;)
[04-Jan-2010 17:16:49] <twm1010> roger that, im out for the night! take it easy
[04-Jan-2010 17:16:54] <rmatte> later
[04-Jan-2010 17:20:54] <theacolyte-> The real issue will be trying to get this graph from Cacti into Zenoss
[04-Jan-2010 17:20:57] <theacolyte-> it's pretty complicated
[04-Jan-2010 17:21:25] <rmatte> hehe
[04-Jan-2010 17:21:34] <rmatte> well, what's so complex about it?
[04-Jan-2010 17:24:26] <theacolyte-> Nothing really, other than I haven't played with SNMP
[04-Jan-2010 17:24:31] <theacolyte-> do you just put in OIDs?
[04-Jan-2010 17:26:16] <Fuzzy_T> theacolyte-: it is not anymore difficult than setting up graphs in cacti. Just different. :)
[04-Jan-2010 17:26:37] <Fuzzy_T> theacolyte-: you should be able to just put in the OID depending on what it is.
[04-Jan-2010 17:27:07] <theacolyte-> Excellent
[04-Jan-2010 17:27:18] <theacolyte-> seeing as I already have the graph, this may be easy
[04-Jan-2010 17:27:37] <Fuzzy_T> should not be hard try it first then ask 4 help when you get stuck.
[04-Jan-2010 17:27:49] <Fuzzy_T> the OID should already be in cacti.
[04-Jan-2010 17:27:56] <theacolyte-> yeah
[04-Jan-2010 17:27:58] <theacolyte-> it is
[04-Jan-2010 17:28:20] <theacolyte-> sorry, I didn't mean to cry for help on that hehe
[04-Jan-2010 17:28:22] <theacolyte-> I haven't tried it yet
[04-Jan-2010 17:29:11] <theacolyte-> Gotta graph my equallogic SAN (that's what I was referring to)
[04-Jan-2010 17:29:15] <Fuzzy_T> I did not see your cry for help yet so no prob.
[04-Jan-2010 17:29:19] <theacolyte-> hehe
[04-Jan-2010 17:29:29] <Fuzzy_T> what are you graphing?
[04-Jan-2010 17:29:32] <Fuzzy_T> on the SAN?
[04-Jan-2010 17:29:59] <theacolyte-> fan speed, latency, controller temps, total byes/sec, and ios
[04-Jan-2010 17:30:02] <theacolyte-> iops
[04-Jan-2010 17:30:25] <theacolyte-> Then it's onto doing my cat switches
[04-Jan-2010 17:31:26] <theacolyte-> yeah, this provides OID's
[04-Jan-2010 17:31:29] <theacolyte-> this should be easy
[04-Jan-2010 17:31:30] <Fuzzy_T> theacolyte-: you should look into rmatte's cisco pack but beware it will take a while to load all the SNMP MIB's
[04-Jan-2010 17:32:04] <Fuzzy_T> Nothing there sounds too difficult. Good thing is it is easy to set thresholds to alert on after you get the data / graphs setup.
[04-Jan-2010 17:32:12] <theacolyte-> yeah, no joke
[04-Jan-2010 17:34:31] <theacolyte-> bah
[04-Jan-2010 17:34:40] <theacolyte-> I have to add another NIC to my zenoss box!
[04-Jan-2010 17:34:47] <theacolyte-> (good thing it's virtualized!)
[04-Jan-2010 17:40:29] <rmatte> I'm out, going to see Avatar, see you guys tomorrow.
[04-Jan-2010 17:44:08] <theacolyte-> Avatar was amazing
[04-Jan-2010 17:44:10] <theacolyte-> have fun
[04-Jan-2010 17:46:06] <theacolyte-> Phew just got a huge error heh
[04-Jan-2010 17:46:15] <theacolyte-> database conflict error (oid 0x0df0, class Products.ZenModel.PerformanceConf.PerformanceConf)
[04-Jan-2010 17:47:18] <Fuzzy_T> hmm you got that when you put the OID in?
[04-Jan-2010 17:47:30] <theacolyte-> well, that's weird, the device is added
[04-Jan-2010 17:47:35] <theacolyte-> no that's what happened when I tried to add it
[04-Jan-2010 17:48:39] <theacolyte-> Wonder if it has to do with 0 templates applied
[04-Jan-2010 17:49:46] <Fuzzy_T> not sure.
[04-Jan-2010 17:49:58] <theacolyte-> I will just ignore it
[04-Jan-2010 17:50:04] <theacolyte-> Just like this network down error!
[04-Jan-2010 18:22:44] <theacolyte-> Weird, my SAN is throwing this trap that has no details
[04-Jan-2010 18:23:04] <theacolyte-> Authentication failure apparently
[04-Jan-2010 18:23:08] <theacolyte-> Not sure where that is coming from
[04-Jan-2010 19:24:49] rhettardo_ is now known as rhettardo
[04-Jan-2010 19:52:00] Fuzzy_T_ is now known as Fuzzy_T
[05-Jan-2010 00:00:30] [disconnected at Tue Jan 5 00:00:30 2010]
[05-Jan-2010 00:00:31] [connected at Tue Jan 5 00:00:31 2010]
[05-Jan-2010 04:11:49] <moto-moi> What's the best way to use a drac/ilom/rsa2 in zenoss?
[05-Jan-2010 04:12:09] <moto-moi> I guess I need a way to combine the 'normal' ip of a server to the drac/ilom/rsa2?
[05-Jan-2010 04:15:52] <baffle> moto-moi: Actually, I've not found a good way.
[05-Jan-2010 04:16:09] <moto-moi> that doesn't sound good :+
[05-Jan-2010 04:16:20] <moto-moi> atm every server of mine has such an interface :P
[05-Jan-2010 04:16:35] <baffle> moto-moi: We have all drac/ilom/rsa/ipmi etc information in a separate inventory database. :)
[05-Jan-2010 04:17:50] <moto-moi> that is indeed an option, but it would be nice to combine those kind of data imo
[05-Jan-2010 04:25:10] <Raven737> Hi... so i just setup the core version just to try it out. We are currently using Nagios for monitoring and i head that zenoss is much better and can even use nagios plugins/clients directly and that is where i am stuck at right now. On most windows machines we have nsclient++ for nagios setup, can i use that with zenoss and if so how? Thanks!
[05-Jan-2010 04:29:54] <nassy> Raven737: i am using nagios too and am thinking of moving to zenoss also. i am now learning aboit zenoss. i think the way zenoss works is it uses a windows admin login to get info via wmi
[05-Jan-2010 04:33:13] <nassy> i dont know or sure but if your nagios plugin works in zenoss (not sure how to set it up yet) then it probably can query nsclient++. the wiki should have info. but can you get the win info from wmi?
[05-Jan-2010 04:34:45] <Raven737> nassy: yes, i tried that (gave admin credentials to the add device wizard) but it didn't seem to work for me. I installed the snmp service and after that i get some very basic information (machine name and windows version)... how do i get all the rest? Important things like cpu, mem and hdd stats?
[05-Jan-2010 04:36:20] <nassy> Raven737: i think there is a dll you have to download and install to get cpu etc from snmp
[05-Jan-2010 04:39:38] <nassy> i dont know where you can find the dll though. yeah, i had similar problems with WMI and ssh (on mac os x) also. still trying to figure everything out though.
[05-Jan-2010 04:39:46] <Raven737> nassy: oh, wmi seems to work, i just need to create/assign a template that collects more then just up time?
[05-Jan-2010 04:41:41] <nassy> i'm still learning. i am not sure. the built in classes don't work for you?
[05-Jan-2010 04:58:11] <Raven737> nassy: thanks for the suggestion, changed device class ro Server/Windows... now shoing some more things, but not quite working i think...
[05-Jan-2010 06:17:29] nassy_ is now known as nassy
[05-Jan-2010 07:43:49] crow is now known as coofamani
[05-Jan-2010 08:02:32] <OiScout> anyone have any ideas why almost every time I get the following error zenperfwmi stops running "ERROR zen.zenperfwmi: NT_STATUS_CONNECTION_REFUSED"
[05-Jan-2010 08:03:12] <twm1010> it means WMI the client is refusing the WMI connection from the DNS server
[05-Jan-2010 08:03:50] <OiScout> right, but why does zenperfwmi shutdown for all of zenoss just because one box is refusing a connection?
[05-Jan-2010 08:05:29] <twm1010> if you remove that one node, does it run fine?
[05-Jan-2010 08:06:44] <OiScout> no it eventually just happens on another node
[05-Jan-2010 08:07:22] <twm1010> thats pretty odd, did it ever work?
[05-Jan-2010 08:07:52] <OiScout> it works for the most part
[05-Jan-2010 08:08:20] <OiScout> just every once and a while, lets say 3 days it hits a server gets NT_STATUS.... and bombs out
[05-Jan-2010 08:09:38] <OiScout> WARNING zen.zenperfwmi: Closing watcher of <server> is the last thing I see in the logs
[05-Jan-2010 08:12:26] <twm1010> hrmm, not sure, i've never had the daemon crash on me like that
[05-Jan-2010 08:12:32] <twm1010> search the forum?
[05-Jan-2010 08:15:57] <OiScout> Thanks anyway, figure it was worth asking
[05-Jan-2010 08:16:13] <OiScout> I have never had deamons act like this either until I started using this zenperfwmi zenpack
[05-Jan-2010 08:19:58] <twm1010> that zenpack works flawless for me, im using the older version on zenoss 2.4.2 though
[05-Jan-2010 08:24:30] <twm1010> rmatte: around?
[05-Jan-2010 08:28:18] <OiScout> yeah I am still using 2.4.5 with wmidata 1.5
[05-Jan-2010 08:29:57] <RobertLaptop> I though 1.5 was only for 2.5.1?
[05-Jan-2010 08:31:01] <OiScout> I don't know but I had more problems before I upgraded that :)
[05-Jan-2010 08:32:17] <twm1010> yes, 1.5 is very much only for 2.5+
[05-Jan-2010 08:32:39] <twm1010> they changed a lot and he had to rewrite the daemon, i think you need to downgrade
[05-Jan-2010 08:33:17] <twm1010> or maybe it was 2.0 for zenoss 2.5
[05-Jan-2010 08:33:19] <twm1010> im checking mine
[05-Jan-2010 08:33:28] <NewBee> Is there a way to controld what devices check for a win services. say any system in a SQL group get checked for the sqlserver service?
[05-Jan-2010 08:33:56] <twm1010> NewBee: Not really no
[05-Jan-2010 08:33:59] <OiScout> 1.5 added compatibility for Zenoss 2.5 but doesn't say it is required
[05-Jan-2010 08:34:15] <twm1010> I'm pretty sure it is though, they changed the way the daemon hooks in.
[05-Jan-2010 08:34:33] <twm1010> mrayzenoss: ping
[05-Jan-2010 08:36:11] <OiScout> so should I upgrade to 2.0.1 and then upgrade zenoss to 2.5.1?
[05-Jan-2010 08:36:32] <NewBee> So you have to go to each system and add in the OS tabe that is should look for the sql services? and if you have say 100 servers?
[05-Jan-2010 08:37:06] <OiScout> brb, meeting.... Thanks for all the info
[05-Jan-2010 08:41:09] <rocket> Hello :)
[05-Jan-2010 08:42:02] <twm1010> i'm using wmi datasource 1.1
[05-Jan-2010 08:43:46] <twm1010> NewBee: No... there is an easier way, but you can't focus it down to just your SQL servers
[05-Jan-2010 08:44:09] <twm1010> so if you go to services on the left side, and go to win services, you will see every windows "service" that zenoss has picked up over WMI
[05-Jan-2010 08:44:37] <twm1010> if you change the zProperty of say... the MSSQL service, it will start monitoring that service on all boxes it has found it on
[05-Jan-2010 08:52:48] <NewBee> ok
[05-Jan-2010 08:52:59] <NewBee> thank you
[05-Jan-2010 08:53:40] <twm1010> for my SQL, i turned it on, and then also changed the severity of it being down to critical
[05-Jan-2010 08:56:32] <Fuzzy_T> mrayzenoss: is the Webcast being recorded so those of us that missed it can watch later?
[05-Jan-2010 08:59:46] <NewBee> Do you need to remodel the servers? or will the services just show up under the OS tabs
[05-Jan-2010 09:01:17] <twm1010> if WMI is working they'll start showing up
[05-Jan-2010 09:01:33] <NewBee> ok
[05-Jan-2010 09:01:36] <twm1010> if you go to the OS tab of a device, under services, there is a little check box called monitored that is a filter
[05-Jan-2010 09:01:42] <twm1010> uncheck that, you'll see every service on the box
[05-Jan-2010 09:11:25] <mrayzenoss> twm1010: pong
[05-Jan-2010 09:12:02] <mrayzenoss> Fuzzy_T: we forgot to record itâ€¦ our first time managing the webinar without our marketing manager :-$
[05-Jan-2010 09:17:11] <twm1010> i missed a webinar? crap
[05-Jan-2010 09:17:18] <Fuzzy_T> me 2
[05-Jan-2010 09:17:38] <twm1010> i almost always glean something useful just listening in the background if i can't give it my full attention
[05-Jan-2010 09:17:41] <Fuzzy_T> tweet sent out 5 hours ago but I did not see it until 8:45 CST
[05-Jan-2010 09:20:02] <twm1010> oh a getting started webinar... thats alright
[05-Jan-2010 09:24:37] <mrayzenoss> lots of content here: http://community.zenoss.org/docs/DOC-2606
[05-Jan-2010 09:30:04] <eidolon> hi folks - i'm tuning a WMI alert that keeps waking us up :) For CPU Utilization over 90, it's currently under Perf/CPU, Sevrity Warn, and Escalation Count 5 - what we really want is "if this is ove r90 for more than 5-10 minutes, then alert us." - the escalation count is... how many 'levels' it increments each time, right? so if I set that to '1', it'll take 4-5 alerts to raise this to 'critical' level, right?
[05-Jan-2010 09:52:19] <venturaville> is anyone working on a KVM monitoring module for zenoss?
[05-Jan-2010 09:53:21] <cluther> eidolon: It only ever escalates one severity.
[05-Jan-2010 09:53:36] <eidolon> so what is the count?
[05-Jan-2010 09:53:47] <eidolon> it'll escalate one severity every x counts?
[05-Jan-2010 09:53:50] <cluther> eidolon: So the escalation count is how many times the threshold has to be exceeded (simultaneously) before the escalation occurs.
[05-Jan-2010 09:53:54] <eidolon> ahhh.
[05-Jan-2010 09:54:07] <eidolon> mmkay. standard polling for wmi is,what, every 5 minutes?
[05-Jan-2010 09:54:15] <cluther> Yeah.
[05-Jan-2010 09:54:23] <eidolon> okay, cool. i can calc from there.
[05-Jan-2010 09:54:59] <mrayzenoss> venturaville: I've heard people mention we need a libvirt one
[05-Jan-2010 09:55:08] <twm1010> wouldn't your escalation count then be... 2? :)
[05-Jan-2010 09:55:22] <eidolon> twm1010: hmm? *lost*
[05-Jan-2010 09:55:40] <venturaville> mrayzenoss: libvirt would cover a bunch, and would probably be a good way to go ... I'm looking at doing that and a puppet monitoring one in my free time
[05-Jan-2010 09:55:59] <twm1010> well... you wanted above 90 for 10 minutes right?
[05-Jan-2010 09:56:09] <eidolon> ah, yes, right. i got it.
[05-Jan-2010 09:56:17] * eidolon is slwo this morning
[05-Jan-2010 09:56:19] <mrayzenoss> venturaville: you're aware of the Cloud Computing ZenPack contest right? $2000 for first place
[05-Jan-2010 09:56:20] <twm1010> problem is... the cpu count is not average cpu, but current cpu
[05-Jan-2010 09:56:36] <twm1010> at least, for Windows
[05-Jan-2010 09:56:42] <mrayzenoss> venturaville: libvirt, kvm and puppet would all count as entries
[05-Jan-2010 09:56:46] <venturaville> mrayzenoss: had not heard of it yet no. got a link?
[05-Jan-2010 09:56:48] <eidolon> well, the windows boxes handle cpu calculation differnt than the linux boxen. have i mentioned ihate windows? :)
[05-Jan-2010 09:57:07] <mrayzenoss> venturaville: http://community.zenoss.org/blogs/zenossblog/2009/10/26/zenoss-in-the-clouds-zenpack-contest
[05-Jan-2010 09:57:10] <twm1010> I was under the impression the Windows figure is not an average at all, but a current value.
[05-Jan-2010 09:58:38] <venturaville> mrayzenoss: thanks, fun timing :-)
[05-Jan-2010 09:58:51] <mrayzenoss> yeah, especially since there aren't any entries yet :)
[05-Jan-2010 09:58:56] <mrayzenoss> so odds are good
[05-Jan-2010 09:59:30] <mrayzenoss> venturaville: I posted this re: puppet http://community.zenoss.org/thread/11945?tstart=0
[05-Jan-2010 09:59:37] <mrayzenoss> as a suggestion
[05-Jan-2010 09:59:43] <mrayzenoss> switching networks, BRB
[05-Jan-2010 10:00:14] <venturaville> mrayzenoss: I have seen that before, though it doesn't jive that well with the way our setup
[05-Jan-2010 10:01:33] <eidolon> twm1010: that's likely right. so if you run one task that is churning on something for a few minutes, it sets off the alarms.
[05-Jan-2010 10:01:37] <eidolon> really irritating.
[05-Jan-2010 10:03:02] <twm1010> eidolon: lets call it our need to know, is windows cpu usage an average, or current value, and if its current, how can we get a 5 min average without polling like mad
[05-Jan-2010 10:04:03] <eidolon> right, i'd like to get that average. this is essentially what i'm doing with the "start low and escalate" thing
[05-Jan-2010 10:07:03] <mrayzenoss> venturaville: if you have any questions about the contest, just let me know
[05-Jan-2010 10:11:53] <OiScout> how can I make sure an old zenpack is removed
[05-Jan-2010 10:11:59] <OiScout> unable to start zenoss
[05-Jan-2010 10:12:34] <twm1010> OiScout: rutroh, removed WMI zenpack and now its broke? :O
[05-Jan-2010 10:12:54] <OiScout> yeah
[05-Jan-2010 10:12:56] <OiScout> pkg_resources.VersionConflict: (ZenPacks.community.WMIDataSource 1.1 (/opt/zenoss/ZenPacks/ZenPacks.community.WMIDataSource-1.1-py2.4.egg), Requirement.parse('ZenPacks.community.WMIDataSource>=1.5'))
[05-Jan-2010 10:13:23] <OiScout> last time I take your advice twm ;)
[05-Jan-2010 10:13:51] <OiScout> zenpack --remove says that 1.5 is not installed
[05-Jan-2010 10:23:28] <OiScout> twm1010: I just reinstalled 1.5 and am back up and running. I am going to try to dig further into this problem. Thanks for your help
[05-Jan-2010 10:40:05] <twm1010> OiScout: I think in your current state, your only option is to upgrade to Zenoss 2.5
[05-Jan-2010 11:13:26] <RobertLaptop> Question my one router vender has provided me with a text file with MiB in it. Is there a way to import that file into ZenOSS?
[05-Jan-2010 11:15:18] <RobertLaptop> Question my one router vendor has provided me with a text file with MiB in it. Is there a way to import that file into ZenOSS? It looks to be in some kinda of standardized format.
[05-Jan-2010 11:15:45] <rmatte> yes, there is
[05-Jan-2010 11:15:53] <rmatte> read up on zenmib in the admin guide
[05-Jan-2010 11:16:36] <rmatte> you basically just do: zenmib run <filename>

[05-Jan-2010 11:16:40] <rmatte> as the zenoss user
[05-Jan-2010 11:17:15] <theacolyte-> There's also something in the GUI
[05-Jan-2010 11:17:31] <rmatte> theacolyte-: no, only if you install the Mib Browser ZenPack
[05-Jan-2010 11:17:35] <eidolon> okay, my custom alert just returned this string:
[05-Jan-2010 11:17:36] <theacolyte-> ahh
[05-Jan-2010 11:17:43] <eidolon> NOSLAVE|Slave is not running!
[05-Jan-2010 11:17:53] <eidolon> but the alert sent out just says Messge: NOSLAVE
[05-Jan-2010 11:17:55] <theacolyte-> You should beat him then
[05-Jan-2010 11:18:02] <eidolon> how do i get the rest of the text message into the alert message?
[05-Jan-2010 11:18:03] <theacolyte-> Sorry, that was a horrible, horrible joke
[05-Jan-2010 11:18:23] <rmatte> eidolon: yeh, only the part before the | becomes the message
[05-Jan-2010 11:18:24] <eidolon> (it's a shell comand - it used return code to say this is a critical failure)
[05-Jan-2010 11:18:34] <rmatte> eidolon: the stuff after is interpreted as datasources
[05-Jan-2010 11:18:44] <rmatte> or datapoints rather
[05-Jan-2010 11:18:53] <eidolon> rmatte: can I provide more information? or should i just write it to standard out?
[05-Jan-2010 11:19:03] <eidolon> (with no pipes)
[05-Jan-2010 11:19:14] <rmatte> eidolon: just provide further info before the | instead of after
[05-Jan-2010 11:19:21] <rmatte> I'm sure it's not rocket science to change it to do that
[05-Jan-2010 11:19:42] <rmatte> It should look like NOSLAVE - Slave is not running!|
[05-Jan-2010 11:19:59] <rmatte> :)
[05-Jan-2010 11:20:28] <eidolon> yeah, i changed it :)
[05-Jan-2010 11:20:31] <rmatte> hehe
[05-Jan-2010 11:21:08] <eidolon> ooh, and it looks like restrting the slave cleared the condition.
[05-Jan-2010 11:21:10] <eidolon> Friggin Awesome.
[05-Jan-2010 11:22:00] <rmatte> hehe
[05-Jan-2010 11:22:15] <rmatte> yeh, when the command succeeds Zenoss knows to clear the alert
[05-Jan-2010 11:23:22] <eidolon> http://pastebin.stonekeep.com/6439
[05-Jan-2010 11:23:30] <eidolon> that's the monitor script.
[05-Jan-2010 11:23:48] <rmatte> nice
[05-Jan-2010 11:25:28] <rmatte> well, I need to get back to work (have 3 clients to commission, woohoo)
[05-Jan-2010 11:25:32] <rmatte> cheers for now
[05-Jan-2010 11:41:13] <jb> the throughput graphs
[05-Jan-2010 11:41:16] <jb> for network interfaces
[05-Jan-2010 11:41:25] <jb> that is Mbit right?
[05-Jan-2010 11:41:58] <cluther> jb: bits/sec
[05-Jan-2010 11:42:20] <jb> the value says "90M"
[05-Jan-2010 11:42:26] <jb> and the legend says "bits/sec"
[05-Jan-2010 11:42:31] <cluther> The values will be autoscaled using the SI prefixes though. http://en.wikipedia.org/wiki/SI_prefix
[05-Jan-2010 11:42:36] <jb> so thats 90,000,000 bits/sec?
[05-Jan-2010 11:42:47] <cluther> yup
[05-Jan-2010 11:43:31] <jb> ok.. so roughly 11 MB/sec
[05-Jan-2010 11:43:38] <jb> or 85 Mbit
[05-Jan-2010 11:44:49] <cluther> We assume 8bits per byte. So 90Mbits would be 11.25Mbytes.
[05-Jan-2010 11:45:16] <jb> if I want to add additional graphs for the "Network Interfaces," but only for Windows devices, should I copy the etherenetCsmacd template to /Devices/Server/Windows and modify it?
[05-Jan-2010 11:45:32] <cluther> yep.
[05-Jan-2010 11:45:38] <jb> thanks :)
[05-Jan-2010 11:52:20] <jb> HRM
[05-Jan-2010 11:52:29] <jb> where is cgibbons? :)
[05-Jan-2010 11:53:19] <jb> having a zenwinperf issue :(
[05-Jan-2010 11:58:07] <malbon> Hi folks, in re: ipv6 are you looking for sites which aready run some IPv6 to help with IPv6 devt in core?
[05-Jan-2010 12:03:13] <rmatte> malbon: you'd have to ask mrayzenoss
[05-Jan-2010 12:04:02] <malbon> rmatte: sure, thanks. I'll ping him a note
[05-Jan-2010 12:05:04] <rmatte> np
[05-Jan-2010 12:20:05] <eidolon> hey is there a way to tell zenoss to NOT throw an alert only on a specific moniotr for a period of time? we have a windows service that will shut down during a specific time window, but hte machine still needs be monitored during that time - just don't want to be alerted about this particular sercvice during that time.
[05-Jan-2010 12:24:50] <mrayzenoss> eidolon: take a look at maintenance windows
[05-Jan-2010 12:25:09] <mrayzenoss> eidolon: there are different types, some that turn off alerting but still monitor
[05-Jan-2010 12:26:03] <eidolon> yeah, the problem is maintenance windows are machine-wide
[05-Jan-2010 12:26:21] <eidolon> we have a service that should only be 'running' 12 hours a day.
[05-Jan-2010 12:26:37] <eidolon> (no, i didn't design this :)
[05-Jan-2010 12:28:15] <mrayzenoss> Alert Rules scheduling?
[05-Jan-2010 12:28:37] <eidolon> i'd have to make an alert rule on a per-alert basis... :-/
[05-Jan-2010 12:35:25] <rmatte> maintenance windows really should be able to be applied to any object, whether it be a whole device or a specific interface or service
[05-Jan-2010 12:35:35] <eidolon> i agree :)
[05-Jan-2010 12:35:37] <rmatte> but that's not currently the case
[05-Jan-2010 12:35:39] <eidolon> but i don't think they can.
[05-Jan-2010 12:35:41] <eidolon> yeah
[05-Jan-2010 12:35:42] <rmatte> so get crafty
[05-Jan-2010 12:35:43] <rmatte> ;)
[05-Jan-2010 12:36:28] <sergeymasushko> hi guys. is it possible to disable modeling of particular Ethernet interfaces for a device?
[05-Jan-2010 12:37:10] <rmatte> eidolon: you'd have to make 1 global alerting rule for everything except the rules that you want to handle differently, then make separate alerting rules for those specific alerts
[05-Jan-2010 12:37:36] <eidolon> yea
[05-Jan-2010 12:37:41] <rmatte> sergeymasushko: you can only disable based on type or name, so it's all or nothing
[05-Jan-2010 12:38:07] <rmatte> sergeymasushko: you can lock the interfaces if you don't want them being updated
[05-Jan-2010 12:38:11] <eidolon> i already have a global alerting rule in place - hmm. how is precedence handled? can i put a rule 'ahead' of the global? or do i have to sadd a 'not' to the global?
[05-Jan-2010 12:38:39] <rmatte> eidolon: you would add a filter to your global that basically says "unless the alert contains these properties, then don't process it using this alerting rule"
[05-Jan-2010 12:38:51] * eidolon nods.
[05-Jan-2010 12:38:53] <eidolon> hm.
[05-Jan-2010 12:39:00] <rmatte> then have the other alerting rule have filters that basically say "only if the alert matches these parameters"
[05-Jan-2010 12:39:02] <sergeymasushko> rmatte: ok. for example I do not want to monitor ngX and tunX interfaces on some device due to these interfaces are dynamically created and destroyed.
[05-Jan-2010 12:39:13] <eidolon> yah, i see.
[05-Jan-2010 12:39:13] <eidolon> ok.
[05-Jan-2010 12:39:16] <eidolon> let me tinker
[05-Jan-2010 12:39:42] <rmatte> sergeymasushko: right, so you'd set that in zProperties
[05-Jan-2010 12:40:13] <rmatte> zInterfaceMapIgnoreNames and zInterfaceMapIgnoreTypes
[05-Jan-2010 12:40:21] <sergeymasushko> rmatte: can you provide an example please
[05-Jan-2010 12:40:30] <sergeymasushko> ok
[05-Jan-2010 12:40:35] <rmatte> sergeymasushko: there is already one in there
[05-Jan-2010 12:41:23] <rmatte> you'd just change ^(Voice Encapsulation|Voice Over IP Peer|Foreign Exchange Station) to ^(Voice Encapsulation|Voice Over IP Peer|Foreign Exchange Station|ng|tun)
[05-Jan-2010 12:41:30] <rmatte> that should work fine
[05-Jan-2010 12:41:55] <rmatte> I'd only do it for specific devices though
[05-Jan-2010 12:41:58] <rmatte> I wouldn't do it globally
[05-Jan-2010 12:42:10] <rmatte> unless you don't think you have any devices that would have tunnels that you want to monitor
[05-Jan-2010 12:44:00] <sergeymasushko> rmatte: woiuld it be correct: zInterfaceMapIgnoreNames - ng*|tun*
[05-Jan-2010 12:44:06] <sergeymasushko> would*
[05-Jan-2010 12:44:14] <rmatte> no, that's what the ^() does
[05-Jan-2010 12:44:25] <rmatte> just do it the way I showed
[05-Jan-2010 12:44:42] <rmatte> ^(ng|tun|)
[05-Jan-2010 12:44:45] <rmatte> erm
[05-Jan-2010 12:44:46] <rmatte> ^(ng|tun)
[05-Jan-2010 12:44:55] <sergeymasushko> thanks
[05-Jan-2010 12:44:58] <rmatte> that means it'll ignore anything that starts with ng or tun
[05-Jan-2010 12:45:05] <rmatte> doesn't matter what it ends with
[05-Jan-2010 12:45:25] <rmatte> np
[05-Jan-2010 12:51:41] <rmatte> man I hate these small clients that we take on which I have to cram on the same Zenoss box
[05-Jan-2010 12:51:45] <rmatte> they all use 192.168.0.x
[05-Jan-2010 12:51:53] * rmatte cries quietly
[05-Jan-2010 12:54:26] <twm1010> lol :)
[05-Jan-2010 12:54:31] <twm1010> doesn't everyone?
[05-Jan-2010 12:54:39] <rmatte> eugh
[05-Jan-2010 12:54:46] <rmatte> I don't even use that at home
[05-Jan-2010 12:54:50] <rmatte> I use 10.0.0.0
[05-Jan-2010 12:54:51] <twm1010> we use 172.16.x.x
[05-Jan-2010 12:55:34] <rmatte> we partially use 172.30.30.x and we also use 30.30.0.0
[05-Jan-2010 12:56:09] <twm1010> the last place i worked at was quite hilarious
[05-Jan-2010 12:56:11] <rmatte> well, 30.30.1.0, 30.30.241.0, 30.30.254.0
[05-Jan-2010 12:56:28] <twm1010> 1100 machines, one physical segment
[05-Jan-2010 12:56:34] <rmatte> that's brutal
[05-Jan-2010 12:56:44] <twm1010> 1.x.x.x with a mask of 255.0.0.0
[05-Jan-2010 12:56:52] * rmatte slaps forehead
[05-Jan-2010 12:56:57] <rmatte> that is SUCH bad practice
[05-Jan-2010 12:57:14] <rmatte> good luck trying to secure that network
[05-Jan-2010 12:57:14] <twm1010> and... it gets better
[05-Jan-2010 12:57:22] <twm1010> the switches weren't hub and spoke
[05-Jan-2010 12:57:27] <twm1010> they were daisy chained
[05-Jan-2010 12:57:36] <rmatte> that's just asking for failure
[05-Jan-2010 12:57:40] <twm1010> and when you got to the last switch at one site, it connected to a fiber bridge to another site
[05-Jan-2010 12:57:46] <twm1010> :)
[05-Jan-2010 12:57:50] <rmatte> lol
[05-Jan-2010 12:58:03] <rmatte> so you could take down the whole site by unplugging 1 ethernet cord, awesome
[05-Jan-2010 12:58:09] <twm1010> 2.4 million packet collisions an hour
[05-Jan-2010 12:58:22] <rmatte> lmao
[05-Jan-2010 12:58:24] <rmatte> that's awesome
[05-Jan-2010 12:58:31] <rmatte> I wish we could be that cool with our network
[05-Jan-2010 12:58:39] <twm1010> In fact, one day in the helpdesk room, I found a really long cable, and plugged it into our switch, lo and behold, it lit up so I left it in there
[05-Jan-2010 12:58:56] <twm1010> Little did I know, that cable, was 200+ ft long, and went all the way around the room and back
[05-Jan-2010 12:58:59] <twm1010> so i created a loop
[05-Jan-2010 12:59:31] <rmatte> actually, my team used to work together at a different company in a NOC... we then all came to this company after that one got bought out/liquidated... this company really only had server engineers... the first time I loaded a bunch of our network equipment in to Zenoss I wanted to cry
[05-Jan-2010 12:59:47] <rmatte> tons of switches with absolutely no port labels, badly IPd, badly connected
[05-Jan-2010 12:59:52] <twm1010> well that loop brought down two whole sites :)
[05-Jan-2010 12:59:57] <rmatte> nice
[05-Jan-2010 13:00:27] <twm1010> watching 36 switches blink in unison every time any device broadcasted was kinda cool
[05-Jan-2010 13:00:38] <rmatte> hehe
[05-Jan-2010 13:00:53] <twm1010> The funny part was a few months before I left a new infrastructure manager came in and redid the whole thing correctly right?
[05-Jan-2010 13:01:11] <twm1010> And just shortly after we had the worst site to site outages and problems ever with the actual network
[05-Jan-2010 13:01:18] <twm1010> the perception was hard to deal with the management
[05-Jan-2010 13:01:31] <twm1010> seeing how the inefficient wrong way to do it had never failed.
[05-Jan-2010 13:02:58] <eidolon> twm1010: you totally should write that up and send it to TheDailyWTF.com
[05-Jan-2010 13:08:24] <eidolon> wait, is 30.30.x.x another unroutable subnet?
[05-Jan-2010 13:09:02] <jgillis> I don't believe it's RFC1918
[05-Jan-2010 13:10:38] <twm1010> haha, perhaps i should
[05-Jan-2010 13:11:07] <eidolon> yeah, i thought 10/8, 172.16/12, 192.168/16 - yeah, foncifmred. unless there's new rfc :)
[05-Jan-2010 13:15:27] <twm1010> eidolon: that is correct
[05-Jan-2010 13:17:21] <jgillis> So I'm having a problem with Zenoss discovering Networks on my network devices. After adding all my routers, zenoss discovered the networks in use, but found them as /24s, when most of them are /30 or /27s
[05-Jan-2010 13:17:30] <jgillis> I
[05-Jan-2010 13:18:04] <jgillis> I've gone into the Networks zProperties and added 23-32 network masks to zDefaultNetworkTree, and re-modeled, but it hasn't changed anything
[05-Jan-2010 13:19:25] <jgillis> This is (I think) causing my google maps portlet to draw connections between every node in my network, since they are all addressed loops out of the same /24
[05-Jan-2010 13:19:53] <jgillis> Anyone have any idea how I can get Zenoss to properly identify the IP Networks in use in my network?
[05-Jan-2010 13:20:09] <twm1010> did you assign those network devices to sites?
[05-Jan-2010 13:20:30] <jgillis> Yes
[05-Jan-2010 13:21:00] <jgillis> It's got the devices properly in the sites, but every site has a link to every other site, and that is not how my network is built
[05-Jan-2010 13:21:08] <jgillis> it's more of a physical ring topology
[05-Jan-2010 13:21:49] <jgillis> I'm wondering if I have to delete the Networks from zenoss and let it re-discover them after modifying the zDefaultNetworkTree?
[05-Jan-2010 13:23:40] <rmatte> eidolon: 30.30.x.x is allocated to NSA I believe
[05-Jan-2010 13:23:55] <rmatte> eidolon: we picked it since we're a NOC, and no one will be using it, so it avoids any possible conflicts
[05-Jan-2010 13:24:34] <eidolon> dbs@boomer:~$ whois 30.0.0.0
[05-Jan-2010 13:24:34] <eidolon> OrgName: DoD Network Information Center
[05-Jan-2010 13:24:40] <eidolon> ooo, scary!
[05-Jan-2010 13:24:40] <rmatte> sorry, DoD
[05-Jan-2010 13:24:41] <rmatte> yeh
[05-Jan-2010 13:24:51] <rmatte> I knew it was some government department
[05-Jan-2010 13:24:51] <rmatte> lol
[05-Jan-2010 13:28:49] <eidolon> ORLY
[05-Jan-2010 13:28:51] <eidolon> muahaha
[05-Jan-2010 13:29:00] <rmatte> ;)
[05-Jan-2010 13:29:01] <eidolon> "your product SUCKS SUCKS SUCKS!!!!!"
[05-Jan-2010 13:29:06] <rmatte> :P
[05-Jan-2010 13:29:07] <eidolon> "oh, and can you add this feature?"
[05-Jan-2010 13:33:06] <rmatte> lol
[05-Jan-2010 13:33:26] <rmatte> man zenprocess drives me nuts sometimes
[05-Jan-2010 13:33:49] <rmatte> It thinks that 5 processes are down across 4 servers when they really aren't
[05-Jan-2010 13:34:02] <rmatte> I think I fixed it, just need to wait for the alerts to clear
[05-Jan-2010 13:34:26] <eidolon> i've had problems with that a lot
[05-Jan-2010 13:34:41] <eidolon> the zenprocess stuff is sort of dynamic - i got an alert that told me 'more' was down.
[05-Jan-2010 13:34:44] <rmatte> it happened after I rebooted the zenoss server
[05-Jan-2010 13:34:48] <eidolon> O.-
[05-Jan-2010 13:34:58] <rmatte> nah, I have it configured properly not to do stuff like that lol
[05-Jan-2010 13:35:29] <eidolon> mostly i turned off httpd monitoring - cuz we were guessing zenprocess kept watching a specific child process in apache, and when that one got reaped, it yelled.
[05-Jan-2010 13:38:03] <eidolon> hmm. i don't have zenoss monitoring itself.
[05-Jan-2010 13:38:05] * eidolon adds.
[05-Jan-2010 13:38:10] <eidolon> (we're having some performance troubles)
[05-Jan-2010 13:39:12] <rmatte> I have 1 Zenoss instance that monitors all of the others
[05-Jan-2010 13:39:38] <eidolon> we have 3 right now. one for internal machines, one for external machines - the middle one is an 'old' install that ate itself - we're migrating off it
[05-Jan-2010 13:39:49] <rmatte> hehe
[05-Jan-2010 13:39:53] <rmatte> "ate itself"?
[05-Jan-2010 13:40:34] <RobertLaptop> I have lots of software that does that. Mostly windows based but every once in a while it happens under linux.
[05-Jan-2010 13:40:44] <eidolon> yeah, to the point where i had a couple zenoss enterprise folks webexed into it and they went "how? wha... how'd you do this? okay, this install is reallllly hosed."
[05-Jan-2010 13:40:54] <eidolon> like, they couldn't figure out where half my classes went, etc.
[05-Jan-2010 13:41:00] <eidolon> so we jdecided to nuke from orbit
[05-Jan-2010 13:41:02] <rmatte> lol
[05-Jan-2010 13:41:23] <eidolon> it really was the only way to be sure.
[05-Jan-2010 13:41:25] <rmatte> yeh, if you lose default event and device classes then you're pretty screwed
[05-Jan-2010 13:42:25] <eidolon> therebuild has gone really well. there were also tons of localized templates - someone went bonkers with "create local copy" "hack the crap out of it"
[05-Jan-2010 13:43:52] <rmatte> lol
[05-Jan-2010 13:44:06] <rmatte> I try to use that very sparingly
[05-Jan-2010 13:44:11] <eidolon> as you should :)
[05-Jan-2010 13:44:49] <rmatte> I wish you could set local properties in templates for a specific object, not having to copy the entire template
[05-Jan-2010 13:44:56] <rmatte> like how you can set local properties in zproperties
[05-Jan-2010 13:44:57] <eidolon> yea
[05-Jan-2010 13:45:03] <rmatte> that would make things sooooooooooooo much easier
[05-Jan-2010 13:45:06] <eidolon> i ahd that problem with this mysql monitor.
[05-Jan-2010 13:45:24] <rmatte> it's such a hassle not being able to do that
[05-Jan-2010 13:45:26] <eidolon> i have a template that checks the slave status, but i need to pass it the port number cuz we have some machines that are running more than one mysql instance.
[05-Jan-2010 13:45:43] <eidolon> and a zproperty won't do it - it's 4 mysql servers on one machine.
[05-Jan-2010 13:45:46] <RobertLaptop> Yea I have 1 mysql 4 install left had to create a local template for that
[05-Jan-2010 13:45:52] <rmatte> right, you'd need to use a custom property or a zproperty
[05-Jan-2010 13:45:59] <rmatte> I explained that to you the other day ;)
[05-Jan-2010 13:46:26] <eidolon> mmhmm :)
[05-Jan-2010 13:46:39] <rmatte> then for the port number you'd just do ${here/cWhateverYouCalledTheCustomProperty}
[05-Jan-2010 13:46:45] <eidolon> yeah, ugly though :(
[05-Jan-2010 13:46:49] <rmatte> not really
[05-Jan-2010 13:46:52] <eidolon> well, wait
[05-Jan-2010 13:47:00] <rmatte> a lot less ugly than a full local template
[05-Jan-2010 13:47:03] <eidolon> the zproperty is machine based.
[05-Jan-2010 13:47:08] <rmatte> correct
[05-Jan-2010 13:47:09] <eidolon> i need 4 zproperties.
[05-Jan-2010 13:47:11] <rmatte> same with custom properties
[05-Jan-2010 13:47:13] <eidolon> for all 4 mysql instances.
[05-Jan-2010 13:47:33] <rmatte> then you'd need 4 properties, whether they be zProperties or Custom Properties
[05-Jan-2010 13:47:39] <eidolon> i'd have to do something like ${here/MySQLPortForServer1} and 2} and 3} and 4}
[05-Jan-2010 13:47:57] <rmatte> yeh, but that's assuming that you actually have 4 instances on every server
[05-Jan-2010 13:48:01] <rmatte> which you probably don't
[05-Jan-2010 13:48:06] <eidolon> nope, don't.
[05-Jan-2010 13:48:08] <eidolon> anyway
[05-Jan-2010 13:48:21] <rmatte> yeh, then the local template is really the only way to go in that case unfortunately
[05-Jan-2010 13:48:24] <eidolon> i really need to figure outwhy i can't model this local machine. so the port number stuff needs to be shelved fro af ew :)
[05-Jan-2010 13:48:44] <rmatte> do you have snmp enabled on the local box?
[05-Jan-2010 13:49:15] <eidolon> well... probably not. no snmpd running. so i'm lookinng for how to start it (centos 5.3)
[05-Jan-2010 13:49:20] <eidolon> service start snmpd ?
[05-Jan-2010 13:49:52] <eidolon> snmpd is stopped
[05-Jan-2010 13:49:52] <eidolon> snmptrapd is stopped
[05-Jan-2010 13:49:52] <rmatte> you'll probably have to install snmpd
[05-Jan-2010 13:49:53] <eidolon> yeah
[05-Jan-2010 13:49:57] <rmatte> if it doesn't come standard
[05-Jan-2010 13:49:58] <eidolon> no, it's there, just not started.
[05-Jan-2010 13:50:01] <rmatte> k
[05-Jan-2010 13:50:06] <rmatte> you'll need to configure it properly too
[05-Jan-2010 13:50:14] <eidolon> bah.
[05-Jan-2010 13:50:17] <eidolon> :)
[05-Jan-2010 13:51:02] <rmatte> http://pastebin.com/m41fee192
[05-Jan-2010 13:51:03] <rmatte> there
[05-Jan-2010 13:51:12] <rmatte> just fill in the proper values
[05-Jan-2010 13:51:21] <eidolon> "yes, rmatte."
[05-Jan-2010 13:51:26] * eidolon zombie-like edits
[05-Jan-2010 13:51:39] <rmatte> change the ip and community string for trapsink lines, and for rocommunity and rwcommunity
[05-Jan-2010 13:51:54] <rmatte> also for syslocation
[05-Jan-2010 13:51:57] <rmatte> syscontact
[05-Jan-2010 13:52:05] <rmatte> that's about it
[05-Jan-2010 13:53:01] <rmatte> restart snmpd once the config is in place and you're in business
[05-Jan-2010 13:53:09] * eidolon is all modelled, thanks :)
[05-Jan-2010 13:53:12] <rmatte> np
[05-Jan-2010 13:54:13] ideopathic_ is now known as ideopathic
[05-Jan-2010 13:54:35] <eidolon> huh, and right off the bat, some of our performance problems may be identified.
[05-Jan-2010 13:54:37] <eidolon> threshold of low swap not met: current value 682464.00
[05-Jan-2010 13:54:42] <eidolon> we may be paging.
[05-Jan-2010 13:54:51] <rmatte> that threshold is so damn useless
[05-Jan-2010 13:54:55] <eidolon> hah
[05-Jan-2010 13:54:58] <rmatte> that kicks off on every single one of my linux boxes
[05-Jan-2010 13:55:04] <rmatte> I just outright disabled it
[05-Jan-2010 13:55:09] <eidolon> snrk
[05-Jan-2010 13:55:14] <eidolon> fine, take the wind out of my sails.
[05-Jan-2010 13:55:29] <rmatte> well, the problem with that threshold is that it's not dynamic at all
[05-Jan-2010 13:55:41] <rmatte> it's hardcoded at 1024000000 or whatever
[05-Jan-2010 13:55:50] <rmatte> so it always assumed you have 1 gig of swap
[05-Jan-2010 13:55:55] <rmatte> assumes*
[05-Jan-2010 13:56:04] <rmatte> you'd have to manually tune it for each server, too much hassle
[05-Jan-2010 13:56:07] <eidolon> SwapTotal: 2064376 kB
[05-Jan-2010 13:56:07] <eidolon> SwapFree: 682476 kB
[05-Jan-2010 13:56:29] <rmatte> it expects you to have 1024000 kB free by default
[05-Jan-2010 13:56:33] <rmatte> hence why it's kicking off
[05-Jan-2010 13:56:40] <eidolon> yeah i see
[05-Jan-2010 13:57:41] <rmatte> eugh, now to try and resolve these stupid IP address conflicts for this client that I'm commissioning...
[05-Jan-2010 13:57:49] <eidolon> have fun :)
[05-Jan-2010 13:58:12] <rmatte> oh I will
[05-Jan-2010 14:00:01] <RobertLaptop> rmatte question. I have been creating local templates to deal with the fact the an OiD is moving around on boxes due I think the how the CPU is identified. Will that trick work for them? http://community.zenoss.org/docs/DOC-3395
[05-Jan-2010 14:00:49] <rmatte> RobertLaptop: what type of device?
[05-Jan-2010 14:01:05] <RobertLaptop> Software router.
[05-Jan-2010 14:01:19] <rmatte> the way that I handle floating OIDs is by writing custom scripts
[05-Jan-2010 14:01:28] <rmatte> then using them as command based datasources
[05-Jan-2010 14:01:33] <rmatte> there really is no other way to do it
[05-Jan-2010 14:01:41] <rmatte> short of writing a whole collector plugin
[05-Jan-2010 14:01:52] <RobertLaptop> ok. I will stick with the working local template :)
[05-Jan-2010 14:02:12] <rmatte> If you want to be really efficient with Zenoss you really should get in to some sort of scripting
[05-Jan-2010 14:02:18] <rmatte> heck, even bash scripting works
[05-Jan-2010 14:02:36] <RobertLaptop> I know bash and php just have no clue how to do anything in python
[05-Jan-2010 14:03:05] <RobertLaptop> but I haven't read anything about using bash or php inside zenoss in any of the docs
[05-Jan-2010 14:04:02] <rmatte> you don't need to use python to do what I was describing
[05-Jan-2010 14:04:06] <rmatte> I'll show you an example...
[05-Jan-2010 14:04:37] <RobertLaptop> I love examples :)
[05-Jan-2010 14:05:44] <rmatte> http://pastebin.com/m3487c3f8
[05-Jan-2010 14:06:07] <rmatte> that's a script that I wrote to collect memory usage info from windows devices via SNMP (since the OID wanders based on the number of physical drives in the system)
[05-Jan-2010 14:06:39] <rmatte> I'll show you the one I did for CPU...
[05-Jan-2010 14:07:05] <rmatte> http://pastebin.com/m5539a40e
[05-Jan-2010 14:07:20] <rmatte> The CPU OIDs have a tendency to wander from one day to another
[05-Jan-2010 14:07:31] <rmatte> .1 becomes .2 and so on
[05-Jan-2010 14:08:27] <rmatte> you basically run the scripts like: ./script.sh <device> <community string>

[05-Jan-2010 14:08:36] <rmatte> then it outputs the data in nagios format
[05-Jan-2010 14:08:43] <rmatte> so you just need to integrate it with Zenoss, which is easu
[05-Jan-2010 14:08:45] <rmatte> easy*
[05-Jan-2010 14:09:31] <RobertLaptop> Looking at the pastbin's now
[05-Jan-2010 14:09:36] <rmatte> k
[05-Jan-2010 14:10:32] <rmatte> I basically do an snmpwalk for a specific string rather than OID, then when I find it I use cut to remove the last number from the OID, then use that to call on other OIDs and get the data that I need
[05-Jan-2010 14:10:37] <rmatte> you could do something similar
[05-Jan-2010 14:11:29] <RobertLaptop> ok. Then how to you call if from a datapoint?
[05-Jan-2010 14:12:01] <rmatte> this is where command based datasources come in...
[05-Jan-2010 14:12:31] <rmatte> you create a template, then create a command based datasource (not snmp based)
[05-Jan-2010 14:12:36] <RobertLaptop> I seems to miss the doc that describs that.
[05-Jan-2010 14:12:48] <rmatte> I assure you that it's in the admin guide
[05-Jan-2010 14:13:04] <coofamani> yup, do the same
[05-Jan-2010 14:13:09] <RobertLaptop> Do you know what section. I seem to be blind.
[05-Jan-2010 14:13:13] <rmatte> then you just do something like: /path/to/the/script.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"
[05-Jan-2010 14:13:16] <rmatte> as the command
[05-Jan-2010 14:13:32] <rmatte> and since mine actually collects data, it displays it like...
[05-Jan-2010 14:13:59] <rmatte> OK|CPU1=72 CPU2=63 CPU3=23 CPU4=90 Total=57
[05-Jan-2010 14:14:19] <rmatte> then you'd create datapoints under the datasource called CPU1 CPU2 CPU3 and Total (case sensitive)
[05-Jan-2010 14:14:26] <rmatte> then you can use those datapoints in graphs
[05-Jan-2010 14:15:10] <rmatte> Since it's outputting in nagios format, Zenoss knows to store the data it collects from CPU1=72 in the CPU1 datapoint RRD
[05-Jan-2010 14:15:40] <rmatte> I include my scripts in an actual ZenPack, so I do: ${here/ZenPackManager/packs/ZenPacks.Nova.Windows.SNMPPerfMonitor.Simple/path}/libexec/wincpu.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"
[05-Jan-2010 14:15:54] <rmatte> but if you just put the script somewhere on the server you can call directly on it
[05-Jan-2010 14:16:22] <rmatte> /usr/local/zenoss/zenoss/scripts/script.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"
[05-Jan-2010 14:16:25] <rmatte> or whatever
[05-Jan-2010 14:16:45] <RobertLaptop> I guess I need to reread that section of the admin doc.
[05-Jan-2010 14:17:00] <rmatte> yeh, that's honestly probably the most useful feature of Zenoss
[05-Jan-2010 14:17:11] <rmatte> custom scripts open up a huge world of monitoring
[05-Jan-2010 14:17:21] <rmatte> that isn't possible with the standard snmp datapoints
[05-Jan-2010 14:17:49] <RobertLaptop> Yea. I actually do a custom script for postfix. But all the customization is on the actually postfix server and net-snmp just sends back data.
[05-Jan-2010 14:19:08] <rmatte> I see
[05-Jan-2010 14:25:12] <rmatte> section 6.3
[05-Jan-2010 14:25:18] <rmatte> Monitoring Using ZenCommand
[05-Jan-2010 14:25:31] <RobertLaptop> Thanks I was rereading Section 5.
[05-Jan-2010 14:25:37] <rmatte> hehe
[05-Jan-2010 14:25:52] <rmatte> read the PDF btw, the HTML version is too hard to read in my opinion
[05-Jan-2010 14:27:12] <RobertLaptop> I generally prefer to read the HTML version Unless I print them out.
[05-Jan-2010 14:27:22] <rmatte> to each his own
[05-Jan-2010 14:28:33] <rmatte> they screwed up a part of that, it says to add a data point named "time" but it should be named "responseTime"
[05-Jan-2010 14:28:46] <rmatte> in the 6.3.3 example
[05-Jan-2010 14:28:49] <rmatte> just FYI
[05-Jan-2010 14:29:22] <RobertLaptop> ty
[05-Jan-2010 14:38:39] <RobertLaptop> Question on remote collectors. What is actually stored on them? I am guessing the master SQL server stores all the SQL data. Does the RDD files store on the master webserver or are they stored on the remote collector?
[05-Jan-2010 14:39:06] <rmatte> RRD is stored on the collector I believe
[05-Jan-2010 14:39:24] <rmatte> but I'm not 100% on that, I've never really done much work with remote collectors
[05-Jan-2010 14:40:02] <RobertLaptop> I have 4 large remote sites and 60 smaller ones I kinda want to run remote collectors on those 4 larger sites.
[05-Jan-2010 14:40:26] <RobertLaptop> But I need to know how much ram and storage I will need in the VM's for it.
[05-Jan-2010 14:40:56] <rmatte> depends on how many devices each collector will be monitoring
[05-Jan-2010 14:41:36] <coofamani> yes, RRDs are store in $ZENHOME/perf/Devices on the remote collector
[05-Jan-2010 14:42:53] <RobertLaptop> Well if data is stored on the remote collector that makes a big diff then if just needed for collecting and processing. Since coofamani confirmed they are stored that means I needs a lot more storage to store those files.
[05-Jan-2010 14:43:14] <rmatte> RRD files don't take up much space man
[05-Jan-2010 14:43:50] <rmatte> I'm monitoring 250 devices on a server with 20GB of disk space and total utilization is just under half that
[05-Jan-2010 14:44:07] <rmatte> and I've been monitoring for over half a year, also... RRD files don't grow after they are created
[05-Jan-2010 14:44:17] <rmatte> so the only thing that can add to the space usage is the actual events DB
[05-Jan-2010 14:44:45] <rmatte> let me check the total size of the RRD directory on that server..
[05-Jan-2010 14:45:35] <coofamani> I think my typical RRD file, with a looong consolidation and retention period (longer than default) is under 10MB each
[05-Jan-2010 14:45:50] <rmatte> yeh, they are pretty tiny
[05-Jan-2010 14:46:01] <RobertLaptop> I am wonder if I could use the ZenOSS VMware image for my remote collectors or if I was going to need a full blown install.
[05-Jan-2010 14:46:19] <rmatte> RobertLaptop: I'd highly recommend doing an install by hand
[05-Jan-2010 14:46:38] <rmatte> RobertLaptop: The VMware image is more of a demo type thing in my opinion, I wouldn't use it for production
[05-Jan-2010 14:46:51] <rmatte> :/usr/local/zenoss/zenoss/perf$ du -sh
[05-Jan-2010 14:46:51] <rmatte> 1.8G.
[05-Jan-2010 14:47:03] <rmatte> so yeh, all of the RRD data for 250 devices is 1.8GB
[05-Jan-2010 14:47:20] <rmatte> and that's including windows devices, cisco routers, switches, and few other things
[05-Jan-2010 14:47:23] <RobertLaptop> Do you do heavy montoring it light?
[05-Jan-2010 14:47:33] <rmatte> pardon?
[05-Jan-2010 14:47:39] <RobertLaptop> it=or
[05-Jan-2010 14:47:46] <rmatte> well, define "heavy"
[05-Jan-2010 14:48:05] <RobertLaptop> ok. Then a 10 gig CentOS image should be fine then.
[05-Jan-2010 14:48:15] <rmatte> probably
[05-Jan-2010 14:48:20] <coofamani> really going to depend on the number of datapoints you monitor.
[05-Jan-2010 14:48:24] <rmatte> you can always expand the filesystem if need be
[05-Jan-2010 14:48:29] <rmatte> that's the beauty of VMs
[05-Jan-2010 14:48:46] <RobertLaptop> Free ESXi doesn't allow you to
[05-Jan-2010 14:49:06] <RobertLaptop> Although I have gotten really good at moving the file system around
[05-Jan-2010 14:49:19] <rmatte> I monitor almost 17000 datapoints on the server with 250 devices
[05-Jan-2010 14:49:20] <RobertLaptop> If the causes I have needed it.
[05-Jan-2010 14:49:33] <RobertLaptop> In
[05-Jan-2010 14:49:37] <rmatte> so roughly 68 per device
[05-Jan-2010 14:49:40] <coofamani> I have one collector monitoring ~150 devices and it has 100G worth of RRDS, another monitoring 1422 using only 33G, so datapoints are everything
[05-Jan-2010 14:50:25] <rmatte> coofamani: yup
[05-Jan-2010 14:50:39] <rmatte> coofamani: hence why I tune the crap out of my Zenoss installs to make them as efficient as possible
[05-Jan-2010 14:50:49] <coofamani> yeah
[05-Jan-2010 14:51:00] <rmatte> my average snmp cycle times are 18.78 seconds
[05-Jan-2010 14:51:05] <RobertLaptop> I would est. the avg will be less then 100 data points a device. We will mostly care about the basics.
[05-Jan-2010 14:51:34] <rmatte> at least I think that's seconds
[05-Jan-2010 14:51:38] <coofamani> rmatte: what EXT3 mount options do you use?
[05-Jan-2010 14:51:41] <coofamani> yes its seconds
[05-Jan-2010 14:51:54] <rmatte> coofamani: whatever the ubuntu server standards are, nothing special
[05-Jan-2010 14:52:35] <rmatte> those 250 devices are also spread out all over the U.S, Mexico, and Europe
[05-Jan-2010 14:52:38] <coofamani> Are you referring to dumped extraneous alerts/monitoring when you say tune?
[05-Jan-2010 14:52:52] <coofamani> rmatte: check out 'noatime' and 'data=writeback', should make a big improvement
[05-Jan-2010 14:52:55] <rmatte> so to be able to collect 17000 datapoints in 18 seconds across 2 continents, that's pretty good
[05-Jan-2010 14:53:10] <rmatte> coofamani: cool, I'll have a look
[05-Jan-2010 14:53:55] <coofamani> I would only use data=writeback if you have battery backed RAID, etc... or a power outage could cause some issues
[05-Jan-2010 14:54:22] <rmatte> well, we have a pretty expensive SAN that we're using, so I would assume that it has that
[05-Jan-2010 14:54:34] <coofamani> heh, yeah, I would hope so
[05-Jan-2010 15:05:59] <rmatte> eugh, I think there's something wrong with the tunnel to this client, sometimes the devices respond to snmp normally and sometimes they time out
[05-Jan-2010 15:10:10] * eidolon is setting up a pair of monitors showing the main event log for our NOC monitoring
[05-Jan-2010 15:10:17] <eidolon> almost professional like.
[05-Jan-2010 15:10:24] <rmatte> hehe
[05-Jan-2010 15:10:49] <rmatte> we just have all of our zenoss servers feeding in to our ticketing system as a central location, works good
[05-Jan-2010 15:11:55] <eidolon> *nod* well, we have 2 installs - one for IT, one for production. it's in front of all us sysadmins - the idea is to ber able to see non-acknowledged alerts up on that.
[05-Jan-2010 15:13:17] <twm1010> professional? what's that?
[05-Jan-2010 15:13:35] <rmatte> ah
[05-Jan-2010 15:15:22] <rmatte> we have 1 install for lab (which I primarily use), one install to monitor our noc infrastructure, one install to monitor our internal infrastructure, then 10 client installs
[05-Jan-2010 15:16:22] <rmatte> we'll probably be up to like 20 client installs by end of year (hopefully)
[05-Jan-2010 15:16:52] <ke4qqq> or your client installs just collectors?
[05-Jan-2010 15:16:57] <ke4qqq> s/or/are/
[05-Jan-2010 15:17:01] <rmatte> all separate instances
[05-Jan-2010 15:17:18] <eidolon> twm1010: it's what we pretend to be :)
[05-Jan-2010 15:17:42] <eidolon> so what is a 'client install'? you set up zenoss for people to monitor their machines?
[05-Jan-2010 15:18:32] <rmatte> a company signs a contract with us to monitor their servers and network equipment, we setup a new Zenoss VM and load their stuff in to it, then configure it to generate tickets in the ticketing system when events occur and perform any necessary event tuning as part of the initial commissioning process
[05-Jan-2010 15:19:00] <rmatte> we also monitor some smaller companies which only have like 3 to 10 devices, we have 3 zenoss installs that are multi-client installs
[05-Jan-2010 15:19:10] <eidolon> this sounds a lot like one of our partners - inforonics
[05-Jan-2010 15:19:15] <rmatte> so we load as many clients as we can on to each instance without IP overlap
[05-Jan-2010 15:19:38] <rmatte> If we start getting IP overlap we fire up a new instance
[05-Jan-2010 15:19:49] <twm1010> ah, so some clever routing masks all the different re-usage of 192.168.?
[05-Jan-2010 15:19:53] <rmatte> we haven't even touched the third VM yet, we're only working off the first 2 so far
[05-Jan-2010 15:20:19] <RobertLaptop> I was about to say you must hate 192.168.0.x and 1.x
[05-Jan-2010 15:20:26] <rmatte> not really routing masks, some has 192.168.1.x some use 10.0.0.x, some use 192.168.0.x, we just make sure that we load them on in such a way that they don't overlap
[05-Jan-2010 15:20:30] <rmatte> and we setup the routing accordingly
[05-Jan-2010 15:20:47] <rmatte> and yes, I really do
[05-Jan-2010 15:26:31] <rmatte> but those are our lower priority clients anyways
[05-Jan-2010 15:40:18] <nickanderson> hello
[05-Jan-2010 16:01:25] <mrayzenoss> greetings
[05-Jan-2010 16:12:01] <eidolon> hey the event console
[05-Jan-2010 16:12:12] <eidolon> there's a 'Start' 'Stop' and a number field to the right of it.
[05-Jan-2010 16:12:20] <eidolon> if i change that number field to '5', it should refersh every 5 seconds, right?
[05-Jan-2010 16:12:29] <eidolon> do i have to hit 'load' or 'run' or 'enter' or something?
[05-Jan-2010 16:13:42] <rmatte> enter
[05-Jan-2010 16:14:07] <rmatte> and make sure it says "Stop"
[05-Jan-2010 16:14:17] <rmatte> if it says "Start" it means you clicked to stop the auto-refresh
[05-Jan-2010 16:16:13] <eidolon> yeah, so if i type 5, and hit enter, nothing happens. cursor stays there.
[05-Jan-2010 16:21:26] <rmatte> right, so leave it and it should refresh shortly
[05-Jan-2010 16:21:38] <rmatte> you're aware that those are seconds though right?
[05-Jan-2010 16:21:43] <rmatte> 5 seconds is a bit of overkill
[05-Jan-2010 16:22:30] <rmatte> hmmm, actually, it's not working for me either
[05-Jan-2010 16:22:35] <rmatte> hold on, let me figure this out...
[05-Jan-2010 16:22:54] <baffle> I have a lot of messages like this in my log: 2010-01-05 23:17:21,238 INFO zen.ZenHub: Worker reports 2010-01-05 23:17:21,237 WARNING zen.ZenStatus: device 'sig14-rs1.net.kq.no' network '10.178.1.0/24' not in topology
[05-Jan-2010 16:23:18] <rmatte> baffle: that's harmless, I have tons of those too
[05-Jan-2010 16:23:47] <rmatte> baffle: it's just complaining that it discovered a route on one of the devices to a network which it doesn't have devices in inventory for
[05-Jan-2010 16:23:54] <baffle> rmatte: Yes, I know they are harmless, and I know why they come. :) But I want to know if there is a good way to suppress them and/or manually configure routes so that ZenOss understands the network topology.
[05-Jan-2010 16:23:56] <rmatte> baffle: in most cases there's nothing you can do about that
[05-Jan-2010 16:24:34] <rmatte> you could just remove those routes from devices
[05-Jan-2010 16:24:41] <rmatte> and from the networks section
[05-Jan-2010 16:24:44] <rmatte> then it wouldn't even see them
[05-Jan-2010 16:25:03] <rmatte> not sure if it'd work
[05-Jan-2010 16:25:16] <rmatte> it's honestly wayyyy more trouble than it's worth
[05-Jan-2010 16:25:34] <baffle> rmatte: The reason /I/ have these, is that that (and heaps of other RFC1918) network is behind some linux routers that uses separate routing tables/policy routing for all the networks. So the routers themselves only have /32 addresses in those networks..
[05-Jan-2010 16:26:11] <rmatte> Zenoss sees /32 as /24, it's a bug
[05-Jan-2010 16:26:34] <baffle> rmatte: The netmask for the network is usually /24, and it polls devices behind the router that sees that network as a /24.
[05-Jan-2010 16:26:58] <rmatte> http://dev.zenoss.org/trac/ticket/5666
[05-Jan-2010 16:27:03] <rmatte> read that
[05-Jan-2010 16:27:32] <baffle> rmatte: So; It gets the correct network as a /24 from the devices behind the router, and it sees the subnet /32 from the router, but it does not see any router that has a route to the network; Because SNMP only gives the default routing table. :-/
[05-Jan-2010 16:28:21] <rmatte> ah, because it's linux routing rather than an actual router?
[05-Jan-2010 16:29:00] <rmatte> (I mean, Linux isn't exposing any of those other routes)
[05-Jan-2010 16:29:43] <baffle> rmatte: The SNMP daemon isn't exposing any of the *other* routing tables, no. The default routing table is basically just a default route. But there are separate routing tables for every network.
[05-Jan-2010 16:29:55] <baffle> rmatte: I don't even know if a Cisco router normally reports that?
[05-Jan-2010 16:29:56] <rmatte> This is all I can think of which might help: http://community.zenoss.org/docs/DOC-3231
[05-Jan-2010 16:30:04] <rmatte> other than that, no clue
[05-Jan-2010 16:30:22] <rmatte> oh, you're talking about the default ROUTE
[05-Jan-2010 16:30:24] <rmatte> not routing table
[05-Jan-2010 16:30:30] <rmatte> (don't confuse me)
[05-Jan-2010 16:30:31] <rmatte> :P
[05-Jan-2010 16:30:32] <baffle> rmatte: Upgraded the MySQL backend server yesterday btw. Magic improvement to performance. ;)
[05-Jan-2010 16:30:40] <rmatte> cool
[05-Jan-2010 16:31:06] <rmatte> maybe you should suggest that for future releases
[05-Jan-2010 16:31:18] <mrayzenoss> baffle: what did you upgrade it to?
[05-Jan-2010 16:32:08] <baffle> rmatte: Maybe I'm confusing myself? :) The linux snmp deamon only exposes routing table 0. I have 40-50 routing tables in /etc/iproute2/rt_tables..
[05-Jan-2010 16:32:19] <rmatte> I see
[05-Jan-2010 16:32:24] <baffle> mrayzenoss: What I had available.. A virtual machine. :-)
[05-Jan-2010 16:32:27] <rmatte> I'm sure there's some way to expose those too
[05-Jan-2010 16:32:42] <rmatte> you can expose basically anything with snmpd provided you do the research and configuration
[05-Jan-2010 16:32:43] <baffle> mrayzenoss: It has 16G RAM and 15K FC storage (3PAR F400 SAN).
[05-Jan-2010 16:32:59] <mrayzenoss> baffle: ahh, I thought you meant you'd upgraded MySQL :)
[05-Jan-2010 16:33:03] <rmatte> oh, thought you meant you actually changed the version of MySQL
[05-Jan-2010 16:33:06] <baffle> rmatte: Yes, ofcourse, and we have a custom SNMP agent that handles a lot of services allready. :)
[05-Jan-2010 16:33:06] <rmatte> not the hardware
[05-Jan-2010 16:33:27] <rmatte> There you go, confusing people again
[05-Jan-2010 16:33:28] <rmatte> tsk tsk
[05-Jan-2010 16:33:36] <baffle> rmatte: But I don't think it is supported by the standard MIBs.. So I'd have to make a magic zenoss extension/modeller as well. Argh. :)
[05-Jan-2010 16:33:58] <rmatte> probably
[05-Jan-2010 16:34:02] <baffle> mrayzenoss: Oh, I did both actually.
[05-Jan-2010 16:34:07] <rmatte> or you could just buy real routers :P
[05-Jan-2010 16:34:12] <rmatte> Cisco is your friend
[05-Jan-2010 16:34:27] <baffle> rmatte: I have Ciscos allready, these are only for monitoring/management traffic.
[05-Jan-2010 16:34:35] <rmatte> I see
[05-Jan-2010 16:34:53] <rmatte> couldn't you create routes in table 0 instead of the other tables for the routes that it's complaining about?
[05-Jan-2010 16:34:58] <rmatte> I mean, a routing table is a routing table
[05-Jan-2010 16:35:27] <rmatte> That way they'd be exposed
[05-Jan-2010 16:35:49] <baffle> rmatte: Using cisco boxes with "unlimited" VRFs are quite expensive.. And I also often have to do nasty magic like packet mangling and portforwarding to fix issues..
[05-Jan-2010 16:36:08] <rmatte> well yeh, if you need VRFs it costs a fortune
[05-Jan-2010 16:36:35] <baffle> rmatte: The point of the separate routing tables (aka VRFs) is that each of these networks go into the network of separate customers..
[05-Jan-2010 16:37:15] <rmatte> right
[05-Jan-2010 16:37:31] <baffle> rmatte: So it is important that they can't reach eachother. :)
[05-Jan-2010 16:37:59] <rmatte> We have a Cisco 2800 that we use to tunnel in to all of our client's networks
[05-Jan-2010 16:38:02] <rmatte> and it works fine
[05-Jan-2010 16:38:16] <rmatte> and we're not paying any crazy licensing fees since it's not an ASA or anything
[05-Jan-2010 16:38:34] <rmatte> C2800NM
[05-Jan-2010 16:38:46] <baffle> rmatte: I do magic hacks and packet mangling as well, not sure if a stock Cisco router can do tha.t
[05-Jan-2010 16:39:08] <rmatte> what's the point of the hacks/packet mangling?
[05-Jan-2010 16:39:50] <baffle> Not tired the 2800 yet.. We have 6509 for core switching and 3650G for access in one datacenter and 6513 for core+access in the other.
[05-Jan-2010 16:40:02] <baffle> s/tired/tried/
[05-Jan-2010 16:40:21] <baffle> rmatte: Sometimes you have to work around strange things. :)
[05-Jan-2010 16:41:17] <rmatte> hehe
[05-Jan-2010 16:41:50] <rmatte> workarounds can be good or bad
[05-Jan-2010 16:42:04] <baffle> rmatte: Like.. I had this SAN that wouldn't send traps outside it's directly connected network, even if everything else (access etc) worked from external networks.. So I made the router accept SNMP-traps, rewrite the packet, and send it to ZenOss.. Sure, that is not very special, but even still. :)
[05-Jan-2010 16:42:37] <rmatte> yeh, I had a RAS card like that in an IBM server (like an ILO board)
[05-Jan-2010 16:42:44] <rmatte> would only respond on local network
[05-Jan-2010 16:42:45] <baffle> (Eventually we found out why it did that and fixed it, but anyway)
[05-Jan-2010 16:43:09] <rmatte> lol
[05-Jan-2010 16:43:39] <baffle> rmatte: Does ZenOss see the routing tables of all your VRFs on the 2800?
[05-Jan-2010 16:43:54] <rmatte> I'll check what it sees...
[05-Jan-2010 16:43:55] <baffle> rmatte: Because afaik ZenOss basically doesn't support multiple routing tables..
[05-Jan-2010 16:44:37] <rmatte> hmmm, I specifically only have it collecting local routes
[05-Jan-2010 16:44:53] <rmatte> (setting in zProperties)
[05-Jan-2010 16:45:09] <rmatte> let me change that setting and see...
[05-Jan-2010 16:45:31] <rmatte> if the page ever decides to load
[05-Jan-2010 16:45:39] <baffle> rmatte: Wich kind of sucks, we have a few customers with overlapping RFC1918 networks. Actually, the worst is this one customer with overlapping RFC1918 *internally*. The use 10.0.0.0/24, 10.0.0.0/23 and 10.0.0.0/22.. Machines are sometimes on same, sometimes on different VLANs, sometimes on multiple VLANs and they have to talk to eachother. Yay legacy!
[05-Jan-2010 16:46:00] <baffle> rmatte: I hope you don't have a full BGP feed on it. :)
[05-Jan-2010 16:46:03] <rmatte> ah, actually I don't have that setting enabled
[05-Jan-2010 16:46:37] <rmatte> no no
[05-Jan-2010 16:46:47] <rmatte> that's exactly why I disabled that setting on some of our client boxes
[05-Jan-2010 16:46:59] <rmatte> routers in europe with 300000 BGP routes in them
[05-Jan-2010 16:47:14] <rmatte> ridiculous
[05-Jan-2010 16:47:41] <baffle> rmatte: Ridiculous? :)
[05-Jan-2010 16:48:04] <rmatte> it appears to only have routes for our local stuff
[05-Jan-2010 16:48:13] <rmatte> like from the router to the client Zenoss boxes
[05-Jan-2010 16:48:16] <rmatte> nothing beyond that
[05-Jan-2010 16:48:33] <baffle> rmatte: Yup. I think the standard MIB only has support for one (default) routing table.
[05-Jan-2010 16:49:01] <rmatte> oh well
[05-Jan-2010 16:49:31] <baffle> rmatte: So basically you never get network dependencies right if you use VRFs.
[05-Jan-2010 16:50:06] <rmatte> basically, but honestly, I don't see what the remote networks really have to do with network dependencies, all you really care about is from your router back when you're monitoring your internal stuff
[05-Jan-2010 16:50:19] <rmatte> and when you're monitoring your clients you only care about their router forward
[05-Jan-2010 16:50:27] <baffle> rmatte: True enough.
[05-Jan-2010 16:51:19] <rmatte> woot, ok, so finished 2 commissions today, now I get to go have a few beers before I start tonight's 2 and a half hour change
[05-Jan-2010 16:51:24] * rmatte sighs
[05-Jan-2010 16:51:41] <baffle> rmatte: So it is not like it is a *problem*, just annyoing. And it makes the cute flash network widget confused. :'-(
[05-Jan-2010 16:52:04] <rmatte> that flash network widget honestly isn't that great, I'd prefer something less flashy and more functional
[05-Jan-2010 16:52:24] <baffle> rmatte: Yes, I heard the worlds smallest violin playing while I wrote that.
[05-Jan-2010 16:52:29] <rmatte> The old old version of HP OpenView that I used in the past had an absolutely hideous looking network map, but it worked amazingly
[05-Jan-2010 16:53:05] <rmatte> bright reds and greens that would make your eyes bleed, but it was very accurate
[05-Jan-2010 16:53:06] <baffle> rmatte: So did our installation of WhatsUp Gold in the 90s. :)
[05-Jan-2010 16:53:12] <rmatte> lol
[05-Jan-2010 16:53:28] <rmatte> I never used WhatsUp Gold, but I've heard of it
[05-Jan-2010 16:54:03] <rmatte> anyways, I need to get out of here before I get sucked in to doing something else (get going while the going's good as they say)
[05-Jan-2010 16:54:05] <rmatte> cheers guys
[05-Jan-2010 16:54:13] <baffle> rmatte: Cheers.
[05-Jan-2010 17:14:30] Fuzzy_T_ is now known as Fuzzy_T
[05-Jan-2010 20:10:53] <Apachez> hmpf im still struggeling with setting up a page that shows all my device graphs in one page
[05-Jan-2010 20:28:34] <Apachez> holy f**k this will take time
[05-Jan-2010 20:28:47] <Apachez> isnt there a way to create those graph reports with cli or something ?
[06-Jan-2010 00:00:30] [disconnected at Wed Jan 6 00:00:30 2010]
[06-Jan-2010 00:00:30] [connected at Wed Jan 6 00:00:30 2010]
[06-Jan-2010 07:04:53] <twm1010> Apachez: You need to look at multi-graph reports
[06-Jan-2010 07:04:57] <twm1010> It will save you a ton of time.
[06-Jan-2010 09:09:44] <baffle> Does anyone know if there is a fix for 2.5.1 for bug #5795? I.e. breakage of Event Console due to unicode-characters in the database.
[06-Jan-2010 09:10:01] <baffle> I see there is a patch linked, but that seems to be for some future ZenOss version.
[06-Jan-2010 09:10:15] <baffle> I can't find the code/file in my installation.
[06-Jan-2010 09:13:39] <jb> argh, how I long for better AD authentication
[06-Jan-2010 09:13:43] <jb> :/
[06-Jan-2010 09:19:23] <jb> anyone here using AD/LDAP auth?
[06-Jan-2010 09:20:29] <RobertLaptop> jb it looks to be part of commerical zenoss so it will only happen if someone in the communty writes one.
[06-Jan-2010 09:21:19] <RobertLaptop> There is kinda LDAP auth support but it is only for creating ID's not deleting or password sync
[06-Jan-2010 09:24:31] <twm1010> jb: i tried it out on my test box, it was too picky
[06-Jan-2010 10:17:32] <theacolyte-> LDAP auth support shouldn't ever have anything to do with your password beyond passing it to LDAp t osee if it passes
[06-Jan-2010 11:14:28] <twm1010> any vmware gurus around?
[06-Jan-2010 11:15:46] <theacolyte-> twm1010: I'm a VCP
[06-Jan-2010 11:15:47] <theacolyte-> hehe
[06-Jan-2010 11:16:04] <twm1010> :) well, i have an easy question, we use standalone esx 3 servers with internal storage right now
[06-Jan-2010 11:16:28] <twm1010> if we want to use shared storage (fiber), does an esx standard license cut it, or do we need enterprise?
[06-Jan-2010 11:16:38] <theacolyte-> standard will work
[06-Jan-2010 11:16:57] <twm1010> so enterprise is just for using like vmotion and loadlbalancing etc
[06-Jan-2010 11:17:54] <theacolyte-> Enteprise only factors in with VI3
[06-Jan-2010 11:17:57] <theacolyte-> Not with ESX itself
[06-Jan-2010 11:18:01] <theacolyte-> but. yeah
[06-Jan-2010 11:18:27] <twm1010> ah, so if we run ESX servers in an island with FC storage, standard license is good
[06-Jan-2010 11:18:49] <twm1010> and if we want to cluster them, then we need VI/VS "enterprise". got it
[06-Jan-2010 11:19:07] <theacolyte-> I think you can still cluster them, but it won't give you HA or anything, no
[06-Jan-2010 11:19:12] <theacolyte-> I'm rusty on licnesing that far back
[06-Jan-2010 11:19:33] <willwh> hi guys :)
[06-Jan-2010 11:19:34] <twm1010> well, we're puttin in esx4 servers now, but the licensing is so confusing.
[06-Jan-2010 11:20:09] <willwh> quick question - the domain in the alert templates - where would I be able to change that?
[06-Jan-2010 11:20:21] <willwh> so people can update notices through email
[06-Jan-2010 11:20:33] <twm1010> like, the from address?
[06-Jan-2010 11:20:42] <willwh> i.e. the box my instance is hosted on is set up as "zenoss.local"
[06-Jan-2010 11:20:47] <willwh> no, not the from address
[06-Jan-2010 11:21:08] <willwh> but the links in the alerts are like; http://zenoss.local/*
[06-Jan-2010 11:21:45] <twm1010> oh, the URL... i seem to remember that being in a config file
[06-Jan-2010 11:21:52] <willwh> yes I would've though so
[06-Jan-2010 11:21:56] <willwh> trying to track it down :)
[06-Jan-2010 11:23:55] <twm1010> http://community.zenoss.org/message/12636#12636
[06-Jan-2010 11:24:46] <twm1010> that should set you straight
[06-Jan-2010 11:24:57] <willwh> awesome, thanks muchly twm1010
[06-Jan-2010 11:25:11] <willwh> had everything running for about 3 weeks now (this is the last little bit of config)
[06-Jan-2010 11:25:13] <twm1010> no problem! stick around, its not always so lively in here ;)
[06-Jan-2010 11:25:13] <willwh> absolutely love it
[06-Jan-2010 11:25:25] <willwh> yeah - I've spoken to mray a fair bit
[06-Jan-2010 11:25:31] <willwh> when I was initially setting up
[06-Jan-2010 11:25:45] <willwh> it's just saved me so much time in the last 2 weeks...
[06-Jan-2010 11:26:03] <willwh> and money.... I've canned 4 boxes which were doing almost nothing
[06-Jan-2010 11:26:12] <willwh> funny how you just loose/forget about 'em sometimes
[06-Jan-2010 11:26:13] <willwh> ahaha
[06-Jan-2010 11:26:37] <twm1010> yeah, i discovered a server i built out 4 months ago for an e-mail server migration that we never put into production
[06-Jan-2010 11:26:46] <willwh> haha :)
[06-Jan-2010 11:26:48] <twm1010> it's just sitting there with a terabyte of storage twiddling its thumbs
[06-Jan-2010 11:29:11] <willwh> haha
[06-Jan-2010 11:29:15] <willwh> much like my office at home
[06-Jan-2010 11:29:23] <willwh> I bought a couple of 1.5Tb disks in sales for my NAS
[06-Jan-2010 11:29:29] <willwh> they are not even unwrapped yet:(
[06-Jan-2010 11:29:41] <willwh> spend too much time at work :P
[06-Jan-2010 11:37:28] <gwb235> what are people doing to monitor the collector processes (i.e. collector_zencommand) are running? From time to time the collector will have a daemon die and I don't notice until I go looking for data and find out it hasn't been collecting (the daemon died and no notification)
[06-Jan-2010 11:38:23] <twm1010> I don't think the same processes are running all the time.
[06-Jan-2010 11:38:37] <twm1010> it kicks off python scripts/zencommands at will
[06-Jan-2010 11:39:33] <gwb235> i'm talking about the collector_* daemons (zencommand, zenperfsnmp, zentrap), which I believe do run all the time
[06-Jan-2010 11:39:48] <gwb235> they sleep in between cycles (but remain running)
[06-Jan-2010 11:40:08] <twm1010> well, wouldn't heartbeat catch those?
[06-Jan-2010 11:40:42] <gwb235> if it does, i'm missing where that's being reported in the event console/log...
[06-Jan-2010 11:41:18] <twm1010> if i stop any zenoss daemon, i get a critical event within seconds
[06-Jan-2010 11:42:23] <gwb235> ...checking manually...
[06-Jan-2010 11:45:17] <gwb235> no event in the event console about stopped collector_zencommand process...
[06-Jan-2010 11:47:26] * gwb235 away
[06-Jan-2010 12:23:16] <twm1010> so.... setup a process monitor?
[06-Jan-2010 12:45:39] <baffle> Does anyone know if there is a fix for 2.5.1 for bug #5795? I.e. breakage of Event Console due to unicode-characters in the database.
[06-Jan-2010 12:45:42] <baffle> I see there is a patch linked, but that seems to be for some future ZenOss version.
[06-Jan-2010 12:45:46] <baffle> I can't find the code/file in my installation.
[06-Jan-2010 13:12:49] <Apachez> baffle: got an url for that fix?
[06-Jan-2010 13:12:51] <Apachez> or bug?
[06-Jan-2010 13:14:09] <theacolyte-> time to walk some snmp... yay!
[06-Jan-2010 13:19:10] <theacolyte-> So I'm having an issue actually
[06-Jan-2010 13:19:18] <theacolyte-> I created a few datapoints, with the wrong name, and renamed them
[06-Jan-2010 13:19:29] <theacolyte-> however they are being referenced by the old name when setting up graph points
[06-Jan-2010 13:19:42] <theacolyte-> it's cosmetic, but I'd like to fix it before releasing this zenpack
[06-Jan-2010 13:20:43] <theacolyte-> I guess I can just recreate everything
[06-Jan-2010 13:21:16] <cluther> You really have to recreate the data sources.
[06-Jan-2010 13:21:40] <theacolyte-> yeah
[06-Jan-2010 13:21:42] <theacolyte-> figured
[06-Jan-2010 13:21:54] <cluther> I've had that happen when I created data sources and then renamed them. The data point that was created beneath them automatically doesn't get its name updated.
[06-Jan-2010 13:22:46] <theacolyte-> Doing an equallogic san zenpack
[06-Jan-2010 13:23:01] <twm1010> ooh, i like equallogic :)
[06-Jan-2010 13:23:13] <mrayzenoss> just hit me up when it's ready, my queue is empty
[06-Jan-2010 13:23:16] <theacolyte-> Only concern I have is I think the OID's change between models
[06-Jan-2010 13:23:31] <theacolyte-> So I'm not sure how to address that when i send it to you mray
[06-Jan-2010 13:23:38] <mrayzenoss> mray@zenoss.com
[06-Jan-2010 13:23:57] <theacolyte-> This one will be ready quick vs my windows ones because this is pretty cut and dry to do
[06-Jan-2010 13:24:03] <theacolyte-> So I should be able to send it today
[06-Jan-2010 13:24:37] <mrayzenoss> excellent, I'll be spending 10 hours on planes the next 2 days, which is when I usually do a lot of ZenPack work :)
[06-Jan-2010 13:24:42] <theacolyte-> hehe
[06-Jan-2010 13:24:45] <theacolyte-> going anywhere good?
[06-Jan-2010 13:25:13] <mrayzenoss> I'm going to Pasadena for the BCS Championship Game :)
[06-Jan-2010 13:25:19] <theacolyte-> damn you sir.
[06-Jan-2010 13:25:21] <mrayzenoss> so if you like college football... yes
[06-Jan-2010 13:25:34] <theacolyte-> Although I have a bad feeling about UT in that game
[06-Jan-2010 13:25:36] <venturaville> watched my team go down in flames last night :-/
[06-Jan-2010 13:25:41] <theacolyte-> bama is gonna kill us
[06-Jan-2010 13:25:45] <mrayzenoss> venturaville: GT?
[06-Jan-2010 13:25:48] <venturaville> yeah
[06-Jan-2010 13:26:01] <theacolyte-> I'm technically a Duck though and we got killed by ohio state :(
[06-Jan-2010 13:26:20] <mrayzenoss> I'm expecting a close game, but Texas has the defense to keep it close
[06-Jan-2010 13:26:37] <theacolyte-> Hey... when are you going to allow special characters in datapoint names?!
[06-Jan-2010 13:27:22] <mrayzenoss> open a ticket, we're working on reworking those pages for the next release
[06-Jan-2010 13:27:35] <theacolyte-> remind me how? I really should be doing this
[06-Jan-2010 13:27:52] <mrayzenoss> http://dev.zenoss.org
[06-Jan-2010 13:28:03] <mrayzenoss> you can use zenoss/zenoss or I can send you your own account if you want one
[06-Jan-2010 13:28:09] <theacolyte-> No one has submitted one for that? that is weird
[06-Jan-2010 13:28:23] <theacolyte-> Yeah, I wouldn't mind my own just so I can get emails on it
[06-Jan-2010 13:28:31] <theacolyte-> theacolyte.fn@gmail.com
[06-Jan-2010 13:28:39] <mrayzenoss> what do you want for your login?
[06-Jan-2010 13:28:45] <theacolyte-> theacolyte is fine
[06-Jan-2010 13:29:37] <theacolyte-> Reason special characters are nice is you can't do stuff like, Total bytes/sec
[06-Jan-2010 13:29:43] <theacolyte-> I end up doing Total bytes per sec
[06-Jan-2010 13:29:57] <theacolyte-> There are worse things in life to deal with, but that wouldn't be bad :)
[06-Jan-2010 13:31:15] <mrayzenoss> ok, account created
[06-Jan-2010 13:31:19] <mrayzenoss> check your email
[06-Jan-2010 13:31:43] <theacolyte-> haha, got it
[06-Jan-2010 13:31:49] <mrayzenoss> :)
[06-Jan-2010 13:31:57] <theacolyte-> and I agree with you
[06-Jan-2010 13:33:10] <mrayzenoss> we do defect reviews twice a week, Enhancement reviews every other week
[06-Jan-2010 13:33:26] <theacolyte-> looks like no one has submitted anything on this which is interesting
[06-Jan-2010 13:33:29] <theacolyte-> call me crazy I guess
[06-Jan-2010 13:33:35] <mrayzenoss> we did a big review of 2.5.2 tickets yesterday, so we're narrowing down those
[06-Jan-2010 13:35:32] <theacolyte-> I will set it as a bug!
[06-Jan-2010 13:35:35] <theacolyte-> kidding
[06-Jan-2010 13:35:48] <theacolyte-> What's 2.5.2 called? Stonecrab?
[06-Jan-2010 13:36:03] <mrayzenoss> no, 2.5.2 is just the next maintenance release
[06-Jan-2010 13:36:09] <mrayzenoss> Stone Crab is the next major release
[06-Jan-2010 13:36:18] <mrayzenoss> 2.5 was King Crab, 2.4 was Blue Crab
[06-Jan-2010 13:36:20] <theacolyte-> Ah, don't see how to select 2.5.2 I guess I can file it against 2.5.1
[06-Jan-2010 13:36:46] <mrayzenoss> if it's proposed for a maintenance release there's a field for that, but this is more of an enhancement
[06-Jan-2010 13:36:47] <mrayzenoss> http://dev.zenoss.org/trac/wiki/HowToAddTicket
[06-Jan-2010 13:37:20] <theacolyte-> Submitted!
[06-Jan-2010 13:37:37] <theacolyte-> Done yet?
[06-Jan-2010 13:37:40] <theacolyte-> Have you fixed it yet?
[06-Jan-2010 13:37:43] <theacolyte-> Come on guys!
[06-Jan-2010 13:38:41] <mrayzenoss> it could be awhile :)
[06-Jan-2010 13:38:46] <theacolyte-> hehe
[06-Jan-2010 13:38:57] <theacolyte-> No, I want instant gratification on a product I haven't paid for
[06-Jan-2010 13:39:59] <mrayzenoss> we do accept patches :)
[06-Jan-2010 13:40:06] <twm1010> LOL
[06-Jan-2010 13:40:17] <theacolyte-> haha
[06-Jan-2010 13:40:28] <theacolyte-> That's one thing you'll mnever see out of me
[06-Jan-2010 13:40:32] <theacolyte-> I am not a programmer, by choice
[06-Jan-2010 13:42:47] <venturaville> does anyone have a good suggestion for a graphical XML editor (in linux)
[06-Jan-2010 13:42:56] <venturaville> I hate editing objects.xml
[06-Jan-2010 13:43:03] <mrayzenoss> emacs :)
[06-Jan-2010 13:43:10] <theacolyte-> err this is weird
[06-Jan-2010 13:43:19] <venturaville> ()()()()()()()() is my response to that :-)
[06-Jan-2010 13:43:31] <theacolyte-> these OIDs work in cacti, but not zenoss, and snmpwalk shows nothing
[06-Jan-2010 13:44:09] <gwb235> are the priority 1 enhancements typically the ones that get the green light for the next release?
[06-Jan-2010 13:44:43] <theacolyte-> snmpwalk -czenoss -v1 10.0.100.10 1.3.6.1.4.1.12740.2.1.12.1.6.1.1681978106 against 10.0.100.10
[06-Jan-2010 13:44:47] <theacolyte-> DONE in 0 seconds
[06-Jan-2010 13:45:00] <theacolyte-> Oh, y'know this may just come through in v2
[06-Jan-2010 13:45:21] <theacolyte-> Although hell if I know how to change that
[06-Jan-2010 13:45:33] <theacolyte-> nm
[06-Jan-2010 13:46:05] <theacolyte-> Yep, that was it
[06-Jan-2010 13:46:18] <theacolyte-> Does anything actually still only support v1?
[06-Jan-2010 13:46:34] <gwb235> old netapp filers
[06-Jan-2010 13:46:43] <theacolyte-> ouch
[06-Jan-2010 13:48:52] <venturaville> conglomerate seems to work for xml editing but is slow as dirt
[06-Jan-2010 13:51:51] <baffle> Apachez: Sorry, had to get the kid into bed. URL to bug is http://dev.zenoss.org/trac/ticket/5795 and URL to patch is http://dev.zenoss.org/trac/changeset/16151 ... But I sure am confused. :)
[06-Jan-2010 13:52:42] <Apachez> btw zserver-threads 1 ftw :)
[06-Jan-2010 13:54:17] <Apachez> that "router.py" doesnt exist in 2.5.1 at least
[06-Jan-2010 13:54:20] <Apachez> dunno what they patched
[06-Jan-2010 13:54:26] <Apachez> also mysql uses utf8 as default
[06-Jan-2010 13:54:34] <Apachez> so that bug fix is really strange
[06-Jan-2010 13:54:58] <Apachez> or maybe they mean that they use iso towards mysql and should use iso towards json aswell ?
[06-Jan-2010 13:56:43] <Apachez> ./zenoss/Products/ZenUtils/Ext.py that file contains the same stuff as router.py in trunk
[06-Jan-2010 13:56:51] <gwb235> is there a way to search across all trac tickets for something (looking for solaris zone monitoring) ?
[06-Jan-2010 13:58:10] <Apachez> however many other things in Ext.py wont match router.py from trunk
[06-Jan-2010 13:59:07] <mrayzenoss> gwb235: Trac has search
[06-Jan-2010 14:00:50] <gwb235> i was searching within the individual enhancement, developer queue reports, couldn't find anything... wading through the larger search now.
[06-Jan-2010 14:01:53] <mrayzenoss> gwb235: since the Solaris ZenPack is in Enterprise, the defect would be on the Enterprise Trac, which isn't publicly available
[06-Jan-2010 14:02:04] <gwb235> ah
[06-Jan-2010 14:02:32] <mrayzenoss> gwb235: there is a ticket "Solaris Zones - kstat parser causes traceback" that is closed for 2.5.2
[06-Jan-2010 14:02:50] <gwb235> cool. that was one of mine!
[06-Jan-2010 14:03:02] <mrayzenoss> if you hit up support they can probably get you the patch
[06-Jan-2010 14:07:27] <theacolyte-> mrayzenoss: I'm concerned that if I release this zenpack that the OID's may change for people
[06-Jan-2010 14:07:33] <theacolyte-> I read somepalce they did
[06-Jan-2010 14:07:45] <theacolyte-> Is it ok to release a zenpack with caveats like that?
[06-Jan-2010 14:08:14] <mrayzenoss> yeah, that's fine
[06-Jan-2010 14:08:31] <mrayzenoss> lots of ZenPacks have notes like "tested on model 234, 456 may not work"
[06-Jan-2010 14:08:52] <mrayzenoss> users of other models can make local modifications or update your zenpack accordingly
[06-Jan-2010 14:09:00] <mrayzenoss> or make their own
[06-Jan-2010 14:09:11] <theacolyte-> excellente
[06-Jan-2010 14:09:28] <theacolyte-> just going to do some testing overnight, then I'll shoot it over your way tomorrow for my first of many zenpacks
[06-Jan-2010 14:09:52] <mrayzenoss> that's what I like to hear
[06-Jan-2010 14:11:54] <Apachez> can zenpacks be used to store graphs and history aswell ?
[06-Jan-2010 14:11:57] <gwb235> what if all we have is some useful templates developed, but nothing else?
[06-Jan-2010 14:12:01] <Apachez> like place devices in a zenpack for backup?
[06-Jan-2010 14:12:14] <theacolyte-> gwb235: That's what most of mine are
[06-Jan-2010 14:12:15] <Apachez> and then use that zenpack on another zenoss (with the same version) ?
[06-Jan-2010 14:12:19] <theacolyte-> You can make zenpacks out of that
[06-Jan-2010 14:13:20] <gwb235> i guess i should read how-to create zenpacks. anybody else in here use netscalers (load balancing devices) ?
[06-Jan-2010 14:13:49] <mrayzenoss> gwb235: templates can definitely be stored/exported in ZenPacks
[06-Jan-2010 14:14:53] <mrayzenoss> Apachez: no, ZenPacks do not store devices and their history
[06-Jan-2010 14:15:05] <mrayzenoss> zenbackup is for that
[06-Jan-2010 14:15:33] <mrayzenoss> gwb235: Jane Curry's ZenPack paper is excellent: http://www.skills-1st.co.uk/papers/jcurry.html
[06-Jan-2010 14:16:02] <gwb235> but is it evil to create a zenpack and then say "It Works For Me!(tm)"
[06-Jan-2010 14:16:15] <theacolyte-> Man this is killing me
[06-Jan-2010 14:16:22] <theacolyte-> I'm not sure why this keeps happening to me
[06-Jan-2010 14:16:26] <theacolyte-> MISSING RRD FILE
[06-Jan-2010 14:16:38] <theacolyte-> Always seems to happen when I'm recreating a graph/datapoints
[06-Jan-2010 14:16:59] <mrayzenoss> gwb235: yeah, I try to at least smoketest that they install cleanly and look for obvious errors
[06-Jan-2010 14:17:14] <mrayzenoss> gwb235: frequently I don't have access to the devices they're intended for
[06-Jan-2010 14:19:17] <twm1010> theacolyte-: Are you changing datapoint types by chance?
[06-Jan-2010 14:19:24] <theacolyte-> No
[06-Jan-2010 14:19:45] <theacolyte-> I seem to run into this fairly consistantly actually
[06-Jan-2010 14:20:01] <theacolyte-> In this particular occasion, I modified the OID for the datapoint
[06-Jan-2010 14:20:10] <twm1010> well, missing RRD file would hint that either the data file is missing, or its not running through rrdcreate for some reason
[06-Jan-2010 14:20:30] <twm1010> so either the OID is bad, the zencommand script is bad or can't write to the folder, or ... the collector hasn't picked up the change yet?
[06-Jan-2010 14:20:45] <theacolyte-> Oh, does hte collector need to pick up the change?
[06-Jan-2010 14:20:58] <theacolyte-> I just assumed it recrated the RRD file when you click save
[06-Jan-2010 14:20:59] <twm1010> yeah, its not always instantaneous, thats what push changes does
[06-Jan-2010 14:21:03] <theacolyte-> ahhh
[06-Jan-2010 14:21:11] <twm1010> no, the RRD is created the first time it gets the data
[06-Jan-2010 14:21:20] <twm1010> then you get NAN graphs until it has enough datapoints to draw the graph
[06-Jan-2010 14:21:34] <twm1010> <-- thats how i understand it ;)
[06-Jan-2010 14:21:55] <theacolyte-> makes sense
[06-Jan-2010 14:22:26] <theacolyte-> I ended up just deleting the device and readding it since I know that woorks
[06-Jan-2010 14:22:37] <theacolyte-> I'm also (obviously) not worried about losing history
[06-Jan-2010 14:25:21] <theacolyte-> Nope that didn't work this time, I'll leave it be and go to lunch
[06-Jan-2010 14:31:30] <Apachez> anyone else who have noticed that when loading graphs (either device->perf or report->graph reports) the graph elements loads twice?
[06-Jan-2010 14:31:37] <Apachez> at least it looks so in the browser
[06-Jan-2010 14:33:19] <Apachez> according to live header there is also two access for each element to renderserver
[06-Jan-2010 14:41:56] <theacolyte-> god
[06-Jan-2010 14:42:02] <theacolyte-> I can't fix this missing rrd file stuff
[06-Jan-2010 14:43:47] <baffle> Apachez: My zenoss_events database is latin1. I'm all for converting tables to utf8, but I think Zenoss expects it to be latin1. Especially when you consider the fix. To the file that doesn't exist in 2.5.1. :)
[06-Jan-2010 14:44:06] <theacolyte-> and yeah, confirmed, the RRD file does not exist, period
[06-Jan-2010 14:44:17] <baffle> Apachez: Yes, I saw there were some simmilarities..
[06-Jan-2010 14:45:27] <baffle> mrayzenoss: The best is to create a modeller for the ZenPack that supports X models, with ability to expand it with new models. Imho. Creating modellers is quite hard, tho'.
[06-Jan-2010 14:47:14] <mrayzenoss> baffle: agreed
[06-Jan-2010 14:53:33] <Apachez> http://community.zenoss.org/thread/12441?start=30&tstart=0
[06-Jan-2010 14:53:41] <Apachez> anyone else noticed that graphs are being loaded twice?
[06-Jan-2010 14:56:05] <baffle> Apachez: I haven't tought about it, but it might? :)
[06-Jan-2010 14:56:11] <mrayzenoss> I'm asking dev
[06-Jan-2010 14:58:36] <mrayzenoss> Apachez: please open a ticket for it, Ian says it was intentional 3 years ago to get around an issue that is no longer present
[06-Jan-2010 15:00:09] <willwh> hey guys, looking for a little assistance; I am using this guide - http://community.zenoss.org/docs/DOC-3909 - to write a zencommand, to check availability of a page, with regex
[06-Jan-2010 15:00:17] <baffle> mrayzenoss: Are you planning on releasing a patch for issue 5795 on 2.5.1 or will I have to wait for 2.5.2?
[06-Jan-2010 15:01:10] <willwh> I think it could be something to do with the eventclass?
[06-Jan-2010 15:01:23] <mrayzenoss> baffle: since it's Maintenance Target and Status are "2.5.2" and "Accepted", it will be fixed for 2.5.2. When the patch is available is another question
[06-Jan-2010 15:01:27] <willwh> Name or service not known - the error it is throwing
[06-Jan-2010 15:02:38] <mrayzenoss> baffle: as soon as you see a patch like "16XXX for 2.5.x", that's the proper patch
[06-Jan-2010 15:03:07] <mrayzenoss> of course, it's not considered safe until QA has moved the ticket to "Verified"
[06-Jan-2010 15:03:56] <willwh> actually - not using regex is just fine
[06-Jan-2010 15:04:00] <willwh> I am getting successes
[06-Jan-2010 15:04:05] <mrayzenoss> willwh: checking for the availability of a page? like HTTPMonitor?
[06-Jan-2010 15:04:06] <willwh> I guess that's enough for that...
[06-Jan-2010 15:04:14] <willwh> yes
[06-Jan-2010 15:04:23] <willwh> am I looking at old info then>

[06-Jan-2010 15:04:24] <willwh> ?
[06-Jan-2010 15:04:30] <willwh> I probably should have asked the best way to do that anyway
[06-Jan-2010 15:04:43] <mrayzenoss> nah, that page is just describing how the HTTPMonitor ZenPack works
[06-Jan-2010 15:04:53] <mrayzenoss> it's just a wrapper for the check_http script
[06-Jan-2010 15:04:58] <mrayzenoss> I believe
[06-Jan-2010 15:05:01] <willwh> ok
[06-Jan-2010 15:05:53] <willwh> so what is the best way to go about using that
[06-Jan-2010 15:06:25] <willwh> I don't want to poll the box over snmp - just check http avail
[06-Jan-2010 15:06:54] <willwh> the annoying thing is, I don't have shell access for the box running zenoss
[06-Jan-2010 15:06:58] <willwh> or this would be a bit easier
[06-Jan-2010 15:07:03] <gwb235> simplest is to use HttpMonitor template
[06-Jan-2010 15:07:18] <willwh> gwb235: yes, but, do I add a device?
[06-Jan-2010 15:07:26] <willwh> I am not sure quite how to start - if that makes sense?
[06-Jan-2010 15:07:37] <willwh> I am monitoring 22 boxes with it already
[06-Jan-2010 15:07:41] <gwb235> add a device (somewhere) and bind the HttpMonitor template
[06-Jan-2010 15:07:44] <gwb235> to the device
[06-Jan-2010 15:08:07] <gwb235> if you need something "special" in the regex match, create a local copy and modify Url/regular expression appropriate
[06-Jan-2010 15:08:32] <willwh> yeah, ok, so I've got that working without the regex
[06-Jan-2010 15:08:39] <mrayzenoss> willwh: http://community.zenoss.org/docs/DOC-4074 and http://community.zenoss.org/docs/DOC-3576
[06-Jan-2010 15:08:40] <Apachez> mrayzenoss: any news regarding that doubleloading of graphs bug? :)
[06-Jan-2010 15:08:53] <mrayzenoss> Apachez: you didn't see my comment?
[06-Jan-2010 15:08:58] <willwh> mrayzenoss: thank you!!!
[06-Jan-2010 15:09:04] <Apachez> nope
[06-Jan-2010 15:09:10] <Apachez> too many irc windows here :P
[06-Jan-2010 15:09:13] <mrayzenoss> Apachez: please open a ticket for it, Ian says it was intentional 3 years ago to get around an issue that is no longer present
[06-Jan-2010 15:09:23] <Apachez> ohh
[06-Jan-2010 15:09:44] <mrayzenoss> willwh: and this one's good too: http://community.zenoss.org/docs/DOC-3234
[06-Jan-2010 15:10:48] <Apachez> gah
[06-Jan-2010 15:10:56] <Apachez> where do I open tickets?
[06-Jan-2010 15:11:01] <Apachez> obviously not in "Community development"
[06-Jan-2010 15:11:08] <Apachez> or Im blind or so
[06-Jan-2010 15:11:17] <Apachez> ahh THERE
[06-Jan-2010 15:11:21] <Apachez> damnit its well hidden :)
[06-Jan-2010 15:11:28] <mrayzenoss> dev.zenoss.org zenoss/zenoss
[06-Jan-2010 15:11:44] <mrayzenoss> I can give you your own account if you want
[06-Jan-2010 15:13:33] <gwb235> there was a video you guys (zenoss) posted in the past month related to training/zenoss day at some conference... where did I see that?
[06-Jan-2010 15:13:50] <mrayzenoss> gwb235: it was on the blog and it's here: http://community.zenoss.org/docs/DOC-2606
[06-Jan-2010 15:13:52] <coofamani> http://community.zenoss.org/blogs/zenossblog/2009/12/21/zenoss-training-videos-from-community-day-baltimore
[06-Jan-2010 15:14:17] <baffle> Tried patching ZenUtils/json.py to force latin-1, but that didn't seem to work. Bah.
[06-Jan-2010 15:14:18] <gwb235> i dig this irc bot called mrayzenoss
[06-Jan-2010 15:14:21] <mrayzenoss> gwb235: http://barcampecm.com/
[06-Jan-2010 15:14:24] <mrayzenoss> heh
[06-Jan-2010 15:16:41] <gwb235> the community day was what i was looking for, thanks!
[06-Jan-2010 15:16:57] <mrayzenoss> yeah, just thought you'd like to know about an Austin event
[06-Jan-2010 15:19:59] <Apachez> mrayzenoss: ok created http://dev.zenoss.com/trac/ticket/6004 :)
[06-Jan-2010 15:20:50] <mrayzenoss> Apachez: looks good
[06-Jan-2010 15:21:09] <Apachez> tnx :)
[06-Jan-2010 15:21:16] <Apachez> so it will be fixed by lunch? ;)
[06-Jan-2010 15:21:36] <Apachez> by the way is there some quickfix I could try on my own to fix this until its officially fixed?
[06-Jan-2010 15:21:43] <Apachez> or is it more complicated than that?
[06-Jan-2010 15:21:48] <mrayzenoss> no idea
[06-Jan-2010 15:22:22] <mrayzenoss> we'll have a defect review tomorrow, it may get escalated or not.
[06-Jan-2010 15:22:29] <baffle> Apachez: The double loading? It's not mission critical..?
[06-Jan-2010 15:22:43] <Apachez> baffle: ?
[06-Jan-2010 15:22:47] <Apachez> well its a bug :)
[06-Jan-2010 15:23:05] <Apachez> takes longer time to load pages aswell as brings you larger logs and more load on the poor zope :)
[06-Jan-2010 15:24:09] <baffle> Apachez: My bug is worse, the Event Console doesn't work. :-)
[06-Jan-2010 15:24:40] <Apachez> pfff :P
[06-Jan-2010 15:24:46] <Apachez> it works here ;)
[06-Jan-2010 15:25:19] <coofamani> anyone seen this? "WARNING ZEO.zrpc: (21508) CW: error connecting to ('localhost', 8100): ECONNREFUSED"
[06-Jan-2010 15:25:33] <mrayzenoss> coofamani: are you using multiple collectors?
[06-Jan-2010 15:25:38] <coofamani> yeah
[06-Jan-2010 15:25:44] <coofamani> after an upgrade from 2.3.3 to 2.4.5
[06-Jan-2010 15:26:10] <coofamani> mrayzenoss: this is ckrough
[06-Jan-2010 15:26:19] <mrayzenoss> there was a security fix in 2.4.X for Enterprise that required a change to the zeo.conf
[06-Jan-2010 15:26:23] <coofamani> ah
[06-Jan-2010 15:26:24] <coofamani> thanks
[06-Jan-2010 15:27:51] <mrayzenoss> coofamani: you might want to respond to this: http://community.zenoss.org/docs/DOC-2496
[06-Jan-2010 15:28:23] <mrayzenoss> I know I've told multiple people that the zeo.conf needs to be updated, I just can't remember what the exact fix was
[06-Jan-2010 15:28:26] <Apachez> is that security fix propagated to community version aswell ?
[06-Jan-2010 15:28:54] <mrayzenoss> Apachez: ostensibly it's not present in Core, since distributed collectors weren't supported
[06-Jan-2010 15:29:12] <Apachez> k
[06-Jan-2010 15:29:21] <mrayzenoss> but basically distributed collectors could be used to pull files from each other
[06-Jan-2010 15:29:21] <Apachez> but otherwise its community -> enterprise ?
[06-Jan-2010 15:29:37] <Apachez> except for specific features but im thinking of lets say "my" doubleloading graph bug
[06-Jan-2010 15:30:36] <mrayzenoss> yeah, the features that are in the same in Core and Enterprise don't have fixes held back
[06-Jan-2010 15:30:38] <coofamani> the upgrade process doesnt handle daemons.txt well
[06-Jan-2010 15:30:58] <coofamani> especially if you've customized it on the collectors, it just steamrolls the collector daemons.txt with whatever is on the mastr
[06-Jan-2010 15:31:05] <mrayzenoss> yeah, there are a bunch of 2.5.2 tickets related to upgrading distributed collectors
[06-Jan-2010 15:31:37] <Apachez> whats the eta of 2.5.2 ?
[06-Jan-2010 15:31:41] <coofamani> having multiple hubs/collectors per box introduces a whole other set of fun
[06-Jan-2010 15:33:00] <mrayzenoss> this is the ticket that changed the zeo.conf: http://dev.zenoss.org/trac/ticket/5255
[06-Jan-2010 15:33:11] <mrayzenoss> Apachez: we did the first review of 2.5.2 tickets yesterday
[06-Jan-2010 15:33:46] <mrayzenoss> Apachez: http://dev.zenoss.org/trac/query?status=assigned&status=backlog&status=closed&status=new&status=verification&group=patch_state&patch=2.5.2&col=id&col=summary&col=status&col=owner&col=priority&col=milestone&col=component&col=changetime&report=6&order=status
[06-Jan-2010 15:34:07] <theacolyte-> Ok
[06-Jan-2010 15:34:10] <theacolyte-> So I'm at a loss
[06-Jan-2010 15:34:16] <theacolyte-> Why is it not c reating an RRD file?
[06-Jan-2010 15:34:25] <theacolyte-> The file physically doesn't exist
[06-Jan-2010 15:36:20] <coofamani> thanks mrayzenoss bot
[06-Jan-2010 15:36:21] <theacolyte-> the OID works, I tested it with the test function
[06-Jan-2010 15:36:36] <theacolyte-> I deleted the graph, recreated it, still nothing
[06-Jan-2010 15:36:42] <mrayzenoss> yeah, mrayzenoss bot should get back to work :)
[06-Jan-2010 15:36:44] <theacolyte-> I deleted the device, recreated it, nothing
[06-Jan-2010 15:37:13] <theacolyte-> forums list people with the same problem, but none of the fixes apply
[06-Jan-2010 15:37:28] <robo> hi: about an hour ago i installed zenoss and have it monitoring 3 devices. I can snmp walk to them but I don't get any performance graphs. Any suggestions?
[06-Jan-2010 15:37:45] <theacolyte-> are the graphs blank or?
[06-Jan-2010 15:38:11] <robo> i don't see any images at all
[06-Jan-2010 15:38:27] <theacolyte-> That usually means wait for awhile, but I've also found restarting zenoss fixes it
[06-Jan-2010 15:38:36] <baffle> mrayzenoss: Isn't multiple collectors working in core? Always tought it was. GUI sure makes me think it works. :)
[06-Jan-2010 15:40:26] <theacolyte-> Wait, wait I think I see the issue
[06-Jan-2010 15:40:48] <mrayzenoss> baffle: yeah, I've complained about the UI misleading people
[06-Jan-2010 15:41:17] <mrayzenoss> it sets up the option to send things to another collector, but doesn't configure the collector and set it up for you
[06-Jan-2010 15:41:51] <mrayzenoss> I expect it to be cleaned up in the new UI
[06-Jan-2010 15:42:13] <rocket> mrayzenoss: lies!!! :)
[06-Jan-2010 15:43:27] <robo> theacolyte-, i restarted zenoss using `zenoss-stack restart` -- how long do you think I need to wait for graphs?
[06-Jan-2010 15:43:33] <robo> or is there a good way i can verify it's collecting
[06-Jan-2010 15:44:18] <theacolyte-> 15 minutes
[06-Jan-2010 15:44:44] <robo> theacolyte-, it's been longer than 15 minutes. Also I went into the CPU utilization reports and there is 'N/A' for all the data
[06-Jan-2010 15:44:46] <theacolyte-> Does anyone know if snmpwalk that comes with zenoss would conflict with say, a locall installed snmpwalk
[06-Jan-2010 15:45:35] <robo> i'm guessing that zenoss would use the installed snmpwalk
[06-Jan-2010 15:45:39] <robo> at least that would be what would make sense
[06-Jan-2010 15:46:04] <theacolyte-> well zenoss comes with snmpwalk
[06-Jan-2010 15:46:23] <theacolyte-> but, I want to install snmpwalk on the base OS
[06-Jan-2010 15:47:26] <robo> are you sure you don't have it already?
[06-Jan-2010 15:49:59] <baffle> theacolyte-: No, there is no conflict.
[06-Jan-2010 15:50:13] <theacolyte-> excellent
[06-Jan-2010 15:50:20] <theacolyte-> beats having to open up SNMP even further
[06-Jan-2010 15:51:57] <robo> needs a device class for graphs
[06-Jan-2010 15:56:43] <Apachez> mrayzenoss: tnx
[06-Jan-2010 15:57:04] <Apachez> I assume default polling in zenoss for snmp is 300 secs between each poll
[06-Jan-2010 15:57:15] <Apachez> where do I change if I want to poll every 30 secs instead ?
[06-Jan-2010 15:57:19] <Apachez> or is that just bad?
[06-Jan-2010 16:00:30] <QubeZ> anyone use this Java monitoring zenpack? It's just a zip with an init.py file and some xml file within a folder. No instructions on the page abotu what to do with it exactly. I've followed the other directions so far: http://community.zenoss.org/docs/DOC-3385
[06-Jan-2010 16:03:01] <mrayzenoss> QubeZ: I don't know anything about that ZenPack, but I've wondered if it's a candidate for deprecation
[06-Jan-2010 16:03:19] <mrayzenoss> QubeZ: won't most Java shops use JMX?
[06-Jan-2010 16:03:36] <QubeZ> should i not be using it? i am trying to monitor our openfire (spark) service
[06-Jan-2010 16:04:50] <QubeZ> is this better: http://community.zenoss.org/docs/DOC-3396?
[06-Jan-2010 16:05:41] <mrayzenoss> I would assume so, since most recent Java applications use JMX
[06-Jan-2010 16:07:42] <QubeZ> thanks, much easier installation =)
[06-Jan-2010 16:08:08] <mrayzenoss> yeah, that other ZenPack is for Java SNMP, a solution that never got much traction in the Java world
[06-Jan-2010 16:08:21] <mrayzenoss> especially once JMX got sorted out (JDK 1.4?)
[06-Jan-2010 16:09:27] <mrayzenoss> QubeZ: I found this: http://www.igniterealtime.org/community/message/198389#198389
[06-Jan-2010 16:10:40] <QubeZ> thanks, ill open up my jmx
[06-Jan-2010 16:11:05] <mrayzenoss> good luck
[06-Jan-2010 16:11:48] <QubeZ> mrayzenoss do i need to allow authentication or will it work just default?
[06-Jan-2010 16:12:01] <QubeZ> i see zJmxAuthenticate is set to false by default in zproperties
[06-Jan-2010 16:17:40] <mrayzenoss> QubeZ: probably depends on your configuration
[06-Jan-2010 17:05:31] <theacolyte-> heh, RRD is fixed, though I can't figure out what I did
[06-Jan-2010 17:05:42] <theacolyte-> actually, I think I know
[06-Jan-2010 17:05:51] <theacolyte-> What happens when you have a duplicate OID on the same graph? ;)
[06-Jan-2010 17:11:34] <theacolyte-> well, I'm officially done with cacti
[06-Jan-2010 17:15:45] <RobertLaptop> Well theacolyte I have a filling you will know.
[06-Jan-2010 17:16:05] <theacolyte-> hehe
[06-Jan-2010 17:16:22] <RobertLaptop> But as long as you use two diff data point names you should be ok.
[06-Jan-2010 17:16:32] <theacolyte-> It didn't seem to link it
[06-Jan-2010 17:16:36] <theacolyte-> Like, rather
[06-Jan-2010 17:16:41] <theacolyte-> Now that I fixed it, it's fine
[06-Jan-2010 17:16:54] <theacolyte-> It's funny actually, had I not had that RRD issue I wouldn't have caught my bad OID
[06-Jan-2010 17:17:10] <RobertLaptop> ok
[06-Jan-2010 17:17:41] <baffle> Anyone had these?
[06-Jan-2010 17:17:42] <baffle> 2010-01-07T00:15:15 INFO ZPublisher.Conflict ConflictError at /zport/RenderServer/render: database conflict error (oid 0x32, class Products.ZenUtils.PObjectCache.PObjectCache) (80 conflicts (2 unresolved) since startup at Wed Jan 6 23:40:34 2010)
[06-Jan-2010 17:17:50] <baffle> More and more coming.. :-/
[06-Jan-2010 17:21:34] <baffle> Maybe a reindex() fixes it.. Hmm..
[06-Jan-2010 17:25:52] <theacolyte-> Anyone have a resource for setting up rrd formatting in zenoss?
[06-Jan-2010 17:26:18] <mrayzenoss> theacolyte-: http://delicious.com/mray/rrd
[06-Jan-2010 17:27:29] <theacolyte-> I'll take a look, but I'm trying to figure out why this is showing me a much, much bigger number than cacti did
[06-Jan-2010 17:27:41] <theacolyte-> like instead of seeing 43, I'm seeing 4323788901238123
[06-Jan-2010 17:30:55] <RobertLaptop> sounds like a unit issue
[06-Jan-2010 17:31:22] <theacolyte-> It is, but I'm not sure how to fix it
[06-Jan-2010 17:33:10] <baffle> Strange, I've fixed sitecustomize.py to have "sys.setdefaultencoding('utf-8')", I've patched json.py with "return simplejson.dumps(value, encoding='iso-8859-1')" but my event view is still broken. :-/
[06-Jan-2010 17:33:52] <theacolyte-> oh i see the base setting
[06-Jan-2010 17:33:54] <theacolyte-> that's it i bet
[06-Jan-2010 17:33:55] <baffle> I.e. if I scroll a few pages down I come to *some* event that has *something* wrong with it.
[06-Jan-2010 17:34:34] <theacolyte-> no, that's not... damn
[06-Jan-2010 17:40:03] <Apachez> 4 323 788 901 238 123
[06-Jan-2010 17:40:18] <Apachez> perhaps divide with 100000000000000 ? ;)
[06-Jan-2010 17:40:32] <theacolyte-> hehe it's actually, let'se see
[06-Jan-2010 17:40:52] <Apachez> but its a strange number to divide with
[06-Jan-2010 17:40:52] <theacolyte-> 1525989247.29
[06-Jan-2010 17:41:09] <theacolyte-> It's like divide by 1,000,000
[06-Jan-2010 17:42:36] <theacolyte-> It should be 152.59 I believe
[06-Jan-2010 17:45:50] <theacolyte-> I wish I knew more about this stuff
[06-Jan-2010 17:46:24] <theacolyte-> 1024,/
[06-Jan-2010 17:46:42] <baffle> mrayzenoss: Have you seen anything like this before? http://paste2.org/p/597442
[06-Jan-2010 17:48:21] <theacolyte-> Why does this work in cacti though, that;s the part I don't understand
[06-Jan-2010 17:48:33] <theacolyte-> That # isn't varying enough to be right, either
[06-Jan-2010 17:49:05] <theacolyte-> same thing is happening with a couple of my other counters -they are staying completely static
[06-Jan-2010 17:49:54] <theacolyte-> and the outputted number from SNMP is just as wack, so
[06-Jan-2010 17:54:10] <mrayzenoss> baffle: http://community.zenoss.org/message/41346#41346
[06-Jan-2010 17:54:17] <mrayzenoss> that was my first hit when searching
[06-Jan-2010 17:54:19] <mrayzenoss> gotta run
[06-Jan-2010 17:54:52] <mrayzenoss> searching for "Products.ZenUtils.PObjectCache.PObjectCache" btw
[06-Jan-2010 18:03:59] <theacolyte-> heh, I can't figure this out and I don't even know what to search for
[06-Jan-2010 18:12:36] <baffle> Is it normal for zenmodeler/zenhubworker/zope to use 99-101? I know it is on my servers, but I am just wondering if everyone else is seeing this. It probably means that it is being effective.
[06-Jan-2010 18:24:13] <theacolyte-> It would be superb if there were a cacti/xml import utility!
[06-Jan-2010 19:20:07] <Apachez> what are the 8789 and 8081 ports being used for in zenoss?
[06-Jan-2010 19:20:19] <Apachez> and how can I change them to bind to 127.0.0.1 instead ?
[06-Jan-2010 20:13:22] <Apachez> hmpf
[06-Jan-2010 20:13:31] <Apachez> shouldnt this be needed for using proxy towards zenoss?
[06-Jan-2010 20:13:32] <Apachez> ProxyPass / http://127.0.0.1:8080/
[06-Jan-2010 20:13:32] <Apachez> ProxyPassReverse / http://127.0.0.1:8080/
[06-Jan-2010 20:13:49] <Apachez> or is modrewrite needed aswell?
[06-Jan-2010 21:07:32] nassy_ is now known as nassy
[07-Jan-2010 00:00:30] [disconnected at Thu Jan 7 00:00:30 2010]
[07-Jan-2010 00:00:30] [connected at Thu Jan 7 00:00:30 2010]
[07-Jan-2010 01:47:57] <Troubadix09> Hi all
[07-Jan-2010 04:14:53] <tehhobbit> he
[07-Jan-2010 04:14:56] <tehhobbit> hi*
[07-Jan-2010 04:23:32] <Kabu> hello :)
[07-Jan-2010 04:25:24] <Kabu> i'm new french user of zenoss and i'm looking for restart zenoss by web interface xD
[07-Jan-2010 04:25:43] <Kabu> how can i do??
[07-Jan-2010 04:26:05] <tehhobbit> hold let me check
[07-Jan-2010 04:26:28] <tehhobbit> Settings -> Daemons
[07-Jan-2010 04:27:20] <Kabu> oh yeah :)
[07-Jan-2010 04:27:21] <Kabu> thx
[07-Jan-2010 09:05:37] <rmatte> good morning folks
[07-Jan-2010 09:05:46] <twm1010> hey there
[07-Jan-2010 09:07:00] <Troubadix09> hi rmatte
[07-Jan-2010 09:07:01] <rmatte> Kabu: tu ne peux pas recommencer Zenoss completement a part du site
[07-Jan-2010 09:07:19] <rmatte> Kabu: ca doit etre fait du ligne the commande
[07-Jan-2010 09:10:12] <rmatte> hmmm, I typed the instead of de, my French is rusty lol
[07-Jan-2010 09:16:00] <Kabu> :)
[07-Jan-2010 09:34:52] <coofamani> morning
[07-Jan-2010 09:34:54] <coofamani> quiet in here
[07-Jan-2010 09:47:28] <rmatte> hmmm, need to figure out how to suppress all events for a certain host...
[07-Jan-2010 09:47:44] <rmatte> actually, shouldn't be too difficult
[07-Jan-2010 09:55:20] <twm1010> uhh.. put it in maintenance?
[07-Jan-2010 09:56:58] <Rolfs> is there an easy way to add multiple data sources of the same type template?
[07-Jan-2010 09:57:13] <Rolfs> I mean datapoints
[07-Jan-2010 09:58:20] <rmatte> Rolfs: can you explain a bit better what you're trying to do exactly?
[07-Jan-2010 09:59:27] <Rolfs> the template I've adding has several datapoints that the script prints as: something | key1=val1 key2=val2 key3=val3 key4=val4 .... etc
[07-Jan-2010 09:59:49] <mrchippy> yo, everybody. this is Chip. mattray is out of town cuz he's going to the bcs champ game tonight.
[07-Jan-2010 10:00:13] <rmatte> Rolfs: right, so add a command type datasource and then create the datapoints under it
[07-Jan-2010 10:00:28] <rmatte> Rolfs: make sure that they are named exactly after the keys
[07-Jan-2010 10:00:31] <rmatte> case sensitive
[07-Jan-2010 10:00:37] <Rolfs> yeah, that part I understand
[07-Jan-2010 10:00:39] <rmatte> if Key1 is CPU1 then make the datapoint called CPU1
[07-Jan-2010 10:00:44] <Rolfs> but I want to make it more easy
[07-Jan-2010 10:00:52] <smcingvale> Hi Chip, I want to make a multi-graph report for diskspace. I'd prefer to do it by partition. How does this work? If i set it up like my filesystem template (usinig used_bloakc) the multi-graph report shows up blank. Is this even possible?
[07-Jan-2010 10:01:14] <rmatte> Rolfs: If you're talking about having Zenoss dynamically create datapoints based on the number of keys, no, that's not possible
[07-Jan-2010 10:01:32] <Rolfs> i would like to not add the datapoint manualy for every value
[07-Jan-2010 10:01:56] <Rolfs> or key
[07-Jan-2010 10:02:22] <rmatte> well, how many keys do you have to add?
[07-Jan-2010 10:02:32] <rmatte> or datapoints rather
[07-Jan-2010 10:03:19] <willwh> hi guys :)
[07-Jan-2010 10:03:24] <twm1010> Rolfs: The whole point of a template is to only have to do it once...
[07-Jan-2010 10:03:37] <willwh> ipservice time outs.... where might those be? :)
[07-Jan-2010 10:04:04] <Rolfs> Well.. we have a lot of different and special-made munin scripts
[07-Jan-2010 10:04:11] <mrchippy> smcingvale: digesting and looking...
[07-Jan-2010 10:04:20] <smcingvale> cool, thanks.
[07-Jan-2010 10:04:37] <Rolfs> to say it like that. We have a lot of keys... some munin graphs generates a lot of keys dependent on the data.. is there a zendmd way to f.eks add them ?
[07-Jan-2010 10:05:28] <Rolfs> twm: the next problem would be that I cannot add the same template more than once on one server with different parameters.. afik..
[07-Jan-2010 10:05:53] <Rolfs> f.eks 4-5 different databases..
[07-Jan-2010 10:07:10] <Rolfs> how about having a button in zenoss saying.. Try to auto-fetch datapoints.. and then let you edit them after..
[07-Jan-2010 10:07:13] <twm1010> i see your point, some command datasources for monitoring a proprietary database will vary from server to server unless you're monitoring a cluster or mirrored pair
[07-Jan-2010 10:08:36] <Rolfs> I have 110 templates in the /Devices and using them and binding them to several servers/classes
[07-Jan-2010 10:10:32] <Rolfs> I think I found a script for copying templates with zendmd on the zenoss page.. but I havent looked into how to do it per datapoint.. any good suggestions there?
[07-Jan-2010 10:11:05] <Rolfs> <3 zendmd
[07-Jan-2010 10:12:23] <mrchippy> smcingvale: can you give me the steps you're following? which template are you using (the straight-up FileSystem)?
[07-Jan-2010 10:15:18] <smcingvale> I'm currently using the default FileSystem template that uses the HOST-RESOURCES mib. This works fine when I browse to the OS tab for each device. I can see each partition's usage, etc. I want to collect a bunch of devices filesystemm usage, by partition, into a multi-graph report. All I did was create a report using the usedBlocks datapoint (basically copied whats in the filesystem template). I get nothing in the report when I do this...I think th
[07-Jan-2010 10:17:44] <smcingvale> I'd ideally like to have each partition as a separate graph with about 10 devices on the graph. That way I can see which devices are deviating from the norm, etc. Not sure how to do this, thought I could do it like the filesystem template but not sure if there's something more involved or not.
[07-Jan-2010 10:17:47] <twm1010> you did reference the datasource, and the datapoint right?
[07-Jan-2010 10:19:00] <twm1010> I would think in your collections, you'll have to choose the components themselves, not the nodes, i could be wrong
[07-Jan-2010 10:20:01] <mrchippy> smcingvale: i think it's one or the other, unfortunately. a datapoint for each device in a bunch or a datapoint for each component on a device. i'll fool around some more here and see if i can cook something up.
[07-Jan-2010 10:20:43] <mrchippy> smcingvale: (and you're looking for a datapoint for a named component on each device in a group of devices, right?)
[07-Jan-2010 10:21:18] <smcingvale> mrchippy: correct.
[07-Jan-2010 10:22:10] <smcingvale> twm1010: Is there a way to referencec the components themselves (ie not the nodes) in a multi-graph report?
[07-Jan-2010 10:23:02] <twm1010> I would think so, just like you can reference interfaces specifically
[07-Jan-2010 10:24:13] <smcingvale> How do I set this up in the report? Is there a way to make a collection that references components (or interfaces)?
[07-Jan-2010 10:25:05] <smcingvale> twm1010: Think I just figured it out...didn't notice the Item Type of Device/Component...will give this a shot.
[07-Jan-2010 10:25:09] <twm1010> when you define the collection
[07-Jan-2010 10:25:14] <twm1010> there is n option for componenet
[07-Jan-2010 10:25:19] <twm1010> you can choose a disk
[07-Jan-2010 10:25:37] <twm1010> When I was creating a report mixing bandwidth and apache/iis stuff, i had to reference the interfaces directly
[07-Jan-2010 10:25:43] <twm1010> in a separate collection
[07-Jan-2010 10:25:59] <twm1010> Its pretty granular, but not as automatic as we'd like, but it gets it done.
[07-Jan-2010 10:26:06] <mrchippy> yeah, you can manually build the list.
[07-Jan-2010 10:26:26] <smcingvale> Awesome, thanks guys. I'll play around with this for a bit.
[07-Jan-2010 10:26:58] <fr500> hi
[07-Jan-2010 10:27:08] <fr500> i'm testing zenoss
[07-Jan-2010 10:27:20] <fr500> I added two devices on two locations with the addresses set
[07-Jan-2010 10:27:29] <fr500> they appear on the map fine
[07-Jan-2010 10:27:35] <fr500> but there is no line connecting them
[07-Jan-2010 10:27:50] <fr500> I already set the property to draw links to true
[07-Jan-2010 10:28:02] <cparlette> There are four conditions that must be met for a line to be drawn.
[07-Jan-2010 10:28:02] <cparlette> 1. DeviceX and DeviceY both have network interfaces on a shared IP subnet.
[07-Jan-2010 10:28:02] <cparlette> 2. The shared IP subnet has its zDrawMapLinks property set to true.
[07-Jan-2010 10:28:02] <cparlette> 3. DeviceX and DeviceY are in LocationA and LocationB respectively.
[07-Jan-2010 10:28:02] <cparlette> 4. LocationA and LocationB are shown on the current Google Map.
[07-Jan-2010 10:29:16] <Rolfs> fr500: have you put up the router in your zenoss?
[07-Jan-2010 10:29:21] <fr500> Location A and Location B inside a common organizer
[07-Jan-2010 10:29:37] <smcingvale> mychippy,twm1010: That worked beautifully. Thanks.
[07-Jan-2010 10:29:46] <fr500> Rolfs, yes, the devices are correctly modeled and set to the corresponding location
[07-Jan-2010 10:31:29] <smcingvale> My other question is: Is there anyway to import a bunch of OIDs into a template...either through the interfacce or programatically. I have a mib file containing some custom oids for custom apps...want to import about 200 of these into a template so we can use them to monitor. Don't want to have to go one by one. I've already imported the mibs...just need to use them to actively monitor...
[07-Jan-2010 10:32:16] <twm1010> yikes.... 200 oids?!?
[07-Jan-2010 10:32:33] <twm1010> i think you'd have to script it
[07-Jan-2010 10:33:12] <smcingvale> That's what I was thinking. Don't have a lot of experience and couldn't find a lot in the docs. Could you point me in the right direction on how to get started writing a script...or, are there any currently out there?
[07-Jan-2010 10:33:27] <Rolfs> Oh.. that would be almost like the same question as I have.. If you program it.. I would like to know how it's done
[07-Jan-2010 10:34:13] <smcingvale> If I get it working I'll definitely post it in the forums...just need to know where to start :). This is basically what zenpacks do...
[07-Jan-2010 10:35:08] <Rolfs> probably have to use zendmd, patch it so you have a tab compleation is my first tip
[07-Jan-2010 10:35:11] <twm1010> i dunno if you can design a template from zenDMD
[07-Jan-2010 10:35:14] <fr500> Rolfs, well one of the devices is down, maybe that's why the line isn't drawn?
[07-Jan-2010 10:35:28] <twm1010> but, it might be possible, but complex
[07-Jan-2010 10:35:34] <twm1010> not every datapoint is the same type
[07-Jan-2010 10:35:56] <twm1010> I think it would depend on just how many times you're going to have to do it.
[07-Jan-2010 10:36:10] <twm1010> if you only need one template with those oid's, I'd suck it up and start typing.
[07-Jan-2010 10:36:23] <twm1010> but in Rolfs case where each template is different, yikes
[07-Jan-2010 10:36:48] <twm1010> a slick windows user using AutoIT might pull it off by manipulating the gui, but zenoss isn't always so snappy :)
[07-Jan-2010 10:37:15] <twm1010> so, zenDMD it is, but i don't know if its even capable of creating a template,
[07-Jan-2010 10:37:35] <smcingvale> What about checking out some zenpack source? Don't a lot of the zenpacks simply create device template and stick a bunch of oids in there...I just want that capability with any generic oid list...
[07-Jan-2010 10:37:40] <Rolfs> mm I would just like to be able to have a button next to the 'test' button saying. Try to add datapoints from output of command
[07-Jan-2010 10:37:57] <smcingvale> OK, I can create the template myself...then use the script to import the oids. Think that would work?
[07-Jan-2010 10:38:15] <twm1010> Rolfs: thats a good suggestion, you should open up a trac ticket for a feature enhancement
[07-Jan-2010 10:38:54] <twm1010> especially for database monitoring on assimilar targets, it could be really useful
[07-Jan-2010 10:40:35] <Rolfs> seems to me that if you want to create a zenpack out to import the OIDs, you will have to create a objects.xml file .. which I have never tried to do..
[07-Jan-2010 10:44:43] <mrchippy> you could do either. you can create templates/ds/dp's in zendmd and export them for versioning in the zenpack (objects.xml), or you could create objects.xml directly. Either will work. But I definitely agree that a visual way to do this easily would be a kicka$$ feature.
[07-Jan-2010 10:44:44] <Rolfs> I think I whould have tried to copy a small ZenPack containing a small amount of OIDs and modified it and then imported it.. but after a backup of the zenoss..
[07-Jan-2010 10:44:48] <rmatte> eugh wtf... I'm getting utilization alerts for interfaces even though I have interface utilization thresholds disabled
[07-Jan-2010 10:45:36] <smcingvale> I'll give that a shot and see what I can come up with.
[07-Jan-2010 10:45:37] <mrchippy> rmatte: which daemon? snmp or command?
[07-Jan-2010 10:45:56] <rmatte> snmp
[07-Jan-2010 10:46:10] <rmatte> it just started happening
[07-Jan-2010 10:46:39] <mrchippy> were the thresholds just disabled recently?
[07-Jan-2010 10:46:46] <rmatte> nope
[07-Jan-2010 10:46:50] <rmatte> it's been over a month
[07-Jan-2010 10:47:09] <rmatte> and it's only happening on one specific device for one specific port
[07-Jan-2010 10:47:46] <Rolfs> ImportError: No module named ZenCollector.daemon - Is that a new module in zenoss 2.5 ?
[07-Jan-2010 10:47:56] <mrchippy> huh. can you run a "zenperfsnmp -d <deviceid> -v10" and post the log to the forums. i'll take a look
[07-Jan-2010 10:48:14] <mrchippy> Rolfs: yes it is
[07-Jan-2010 10:48:18] <mrchippy> where'd you see that?
[07-Jan-2010 10:48:20] <rmatte> mrchippy: k
[07-Jan-2010 10:48:37] <Rolfs> mrchippy: is it a way to fix it when using a plugin in a 2.4.1 version?
[07-Jan-2010 10:50:25] <rmatte> hmmm, there's really nothing in the zenperfsnmp run that would explain it
[07-Jan-2010 10:50:32] <rmatte> I don't even see the threshold being triggered
[07-Jan-2010 10:52:03] <mrchippy> rmatte: well, that stinx. hmmm...
[07-Jan-2010 10:52:22] <mrchippy> Rolfs: which plugin? where'd you get it?
[07-Jan-2010 10:52:58] <Rolfs> I get it when I restart zenoss. I have added the ZenPacks.community.ZenODBC-2.1.1-py2.4.egg plugin in my zenoss 2.4.1
[07-Jan-2010 10:53:16] <Rolfs> File "/opt/zenoss/ZenPacks/ZenPacks.community.ZenODBC-2.1.1-py2.4.egg/ZenPacks/community/ZenODBC/zenperfodbc.py", line 36, in ?
[07-Jan-2010 10:53:16] <Rolfs> from Products.ZenCollector.daemon import CollectorDaemon
[07-Jan-2010 10:53:24] <mrchippy> Rolfs: ok, i'll take a look at that bad boy
[07-Jan-2010 10:54:32] <rmatte> mrchippy: oh well, maybe I just need to restart Zenoss on that server, it was crazy slow earlier since we were gathering data for automated reports from it, plus it was getting hammered with monitoring
[07-Jan-2010 10:55:09] <Rolfs> is it a way to add the CollectorDaemon somehow, or would that mess up a lot?
[07-Jan-2010 10:55:41] <willwh> hey guys, sorry to ask this again, but, I am trying to find the timeout settings for ipservice/http
[07-Jan-2010 10:55:47] <mrchippy> rmatte: maybe the stress caused something to lose the configuration update? but just for one device?
[07-Jan-2010 10:56:08] <willwh> one of my boxes reported http being down; event was cleared 1 minute late.... I just wanted to tune it a little
[07-Jan-2010 10:56:10] <rmatte> mrchippy: maybe
[07-Jan-2010 10:56:58] <rmatte> seems to have settled down, I'll wait and see if it happens again
[07-Jan-2010 10:57:14] <mrchippy> Rolfs: no, the CollectorDaemon stuff comes from an overhaul of the daemons. we're moving all the daemons over to this framework. you'd wind up having to do an upgrade.
[07-Jan-2010 10:58:01] <Rolfs> willwh: try to look under the OS tab.. The Event Might tell you more about how the event was discovered.
[07-Jan-2010 10:58:24] <Rolfs> willwh: You could turn off monitoring of the http service under OS-tab
[07-Jan-2010 10:58:25] <mrchippy> Rolfs: the problem is that apparently the setup.py allowed that zenpack to be installed when it's not compatible. the version compatibility string probably didn't get updated :(
[07-Jan-2010 10:59:25] <Rolfs> mrchippy: I guess so..
[07-Jan-2010 10:59:46] <mrchippy> Rolfs: yikes, the COMPAT_ZENOSS_VERS is ">=2.5". even worse, maybe zenoss's compatibility check is not working :O
[07-Jan-2010 10:59:49] <willwh> Rolfs: great thank you
[07-Jan-2010 11:01:22] <theacolyte-> Finding SNMP to be even harder to work with than WMI
[07-Jan-2010 11:01:55] <eidolon> hey folks, what's the seperator for Nagios format checkers for multiple data sources? I have OK|Field=somevalue
[07-Jan-2010 11:01:59] <Rolfs> theacolyte-: why is that?
[07-Jan-2010 11:01:59] <eidolon> wht if ther'e more than one Field?
[07-Jan-2010 11:02:23] <willwh> snmp gives me everything I need
[07-Jan-2010 11:02:26] <Rolfs> asdasdasd |key=value key2=value2
[07-Jan-2010 11:02:27] <theacolyte-> Rolfs: I have a template that is working just fine in Cacti that I tried to move over - same OID's and the numbers are way off
[07-Jan-2010 11:02:32] <eidolon> space. okay, thanks
[07-Jan-2010 11:03:49] <theacolyte-> Unfortunatley I deleted the old template of of cacti to try a different one, and this one doesn't even translate over - they do some weird OID manipulation I haven't figured out yet
[07-Jan-2010 11:03:52] <rmatte> theacolyte-: you sure cacti isn't doing some sort of processing on the numbers?
[07-Jan-2010 11:04:05] <rmatte> theacolyte-: or maybe you need to change the format options for the graph in Zenoss
[07-Jan-2010 11:04:32] <theacolyte-> rmatte: I'm going to figure out how to get these graphs back together so I can troubleshoot and I'll ping ya
[07-Jan-2010 11:04:45] <rmatte> theacolyte-: it's technically impossible for the values to be noticeably different if you're using the exact same OID as cacti is
[07-Jan-2010 11:04:54] <rmatte> theacolyte-: have you tried snmpwalking the OIDs to see what you get?
[07-Jan-2010 11:04:57] <theacolyte-> That's what I'm saying.
[07-Jan-2010 11:05:09] <theacolyte-> I did
[07-Jan-2010 11:05:14] <rmatte> ...and?
[07-Jan-2010 11:05:15] <theacolyte-> Same values it Zenoss was reporting
[07-Jan-2010 11:05:24] <theacolyte-> s/it/that/
[07-Jan-2010 11:05:30] <rmatte> yeh, then Cacti is doing something else to the values obviously
[07-Jan-2010 11:05:35] <rmatte> if they aren't showing up as you'd expect
[07-Jan-2010 11:05:57] <theacolyte-> I don't even know how to figure it out, but at the moment I don't have it in place to troubleshoot
[07-Jan-2010 11:06:21] <rmatte> you'd need to know what it's doing to the values, once you know that it'd be simple from then on
[07-Jan-2010 11:07:35] <theacolyte-> yeah
[07-Jan-2010 11:07:54] <theacolyte-> I tried doing some RPN, but then I ran into an issue where the value actually wasn't changing
[07-Jan-2010 11:08:13] <theacolyte-> But, this new cacti template I'm trying to adapt to zenoss is weird and I can't figure out how they have the OIDs set so I'm walking
[07-Jan-2010 11:08:25] <rmatte> lol
[07-Jan-2010 11:08:49] <theacolyte-> The problem with equallogic SAN's is the OIDs change on each chassis (or, maybe, model)
[07-Jan-2010 11:09:12] <willwh> Rolfs: had a really good hunt around - I still don't see *why* this report http as being down - it doesn't explicitly say timeout... but I think so
[07-Jan-2010 11:09:45] <willwh> nothing in the event log on the box, so, I'd like to tune the timeout
[07-Jan-2010 11:09:49] <rmatte> willwh: telnet to port 80 on whatever device and type a bunch of garbage then hit enter and make sure you actually get some sort of return
[07-Jan-2010 11:09:55] <Rolfs> willwh: did you move it to history?
[07-Jan-2010 11:10:01] <rmatte> willwh: Zenoss does more than just checking to see if port 80 is open
[07-Jan-2010 11:10:07] <willwh> ah
[07-Jan-2010 11:10:25] <willwh> well it's not a web service that runs there.... but that is the 1 and only report of it being down
[07-Jan-2010 11:10:49] <willwh> we service streams on port 80 from most of our boxes, so, the latency to our boxes can be quite high
[07-Jan-2010 11:11:01] <Rolfs> them move the event to history if you have removed it from beeing monitored in the os-page
[07-Jan-2010 11:11:01] <rmatte> I see
[07-Jan-2010 11:11:09] <willwh> I've already tuned the ping timeout to 5seconds (yes I need that - at peak some boxes take aorund 4.5s to respond, although they are working quite happily)
[07-Jan-2010 11:12:00] <willwh> just serving video at 800mbps
[07-Jan-2010 11:12:05] <willwh> :)
[07-Jan-2010 11:12:43] <rmatte> Zenoss connects to port 80 and expects some sort of http return
[07-Jan-2010 11:12:56] <Rolfs> and you have debugged it with trying to fetch the data with the zencommand run -d device or somemthing like that?
[07-Jan-2010 11:12:57] <rmatte> so if it's not seeing that then it's going to think it's down
[07-Jan-2010 11:12:59] <willwh> which it will recieve
[07-Jan-2010 11:13:08] <willwh> (this is a windows media services box)
[07-Jan-2010 11:13:13] <willwh> so it does give an http response
[07-Jan-2010 11:13:17] <rmatte> k
[07-Jan-2010 11:13:28] <mrchippy> smcingvale: check out http://community.zenoss.org/thread/12475
[07-Jan-2010 11:13:29] <willwh> like I said - I've been monitoring this box for ~3days
[07-Jan-2010 11:13:31] <willwh> this is the only event
[07-Jan-2010 11:13:43] <willwh> it was a "hiccup" - looks like a timeout to me
[07-Jan-2010 11:13:49] <Apachez> are there any template or such for selfmonitoring on the zenoss ?
[07-Jan-2010 11:13:55] <Rolfs> could it be that all the daemons where busy serving.. and none where available to answer your zenoss check?
[07-Jan-2010 11:14:04] <Apachez> or do I have to install snmpd manually on the zenoss server first?
[07-Jan-2010 11:14:16] <willwh> Rolfs: nope
[07-Jan-2010 11:14:40] <willwh> 40mbps out / only serving aroun 1800 audio streams right now
[07-Jan-2010 11:14:50] <willwh> is there a timeout I can configure on those checks?
[07-Jan-2010 11:15:26] <theacolyte-> wow
[07-Jan-2010 11:15:33] <theacolyte-> someone that actually uses windows media services
[07-Jan-2010 11:15:38] <theacolyte-> which version?
[07-Jan-2010 11:15:39] <willwh> unfortunately
[07-Jan-2010 11:15:43] <willwh> we use a bunch of stuff
[07-Jan-2010 11:16:08] <theacolyte-> I ran it for awhile here, it wasn't tooooooo bad to deal with, but it felt like an awful, awful hack
[07-Jan-2010 11:16:19] <willwh> yeah
[07-Jan-2010 11:16:26] <willwh> using Wowza Media server to dist most stuff
[07-Jan-2010 11:16:32] <willwh> and icecast
[07-Jan-2010 11:16:36] <willwh> and darwin
[07-Jan-2010 11:16:43] <theacolyte-> What, if you don't mind, does your company use it for?
[07-Jan-2010 11:16:43] <QubeZ> man this asterisk monitoring zenpack is just killing my server :(
[07-Jan-2010 11:16:44] <willwh> some shoutcast setups too actually
[07-Jan-2010 11:16:59] <willwh> theacolyte-: we are a streaming media host :)
[07-Jan-2010 11:17:16] <theacolyte-> willwh: Powerstream? :P
[07-Jan-2010 11:17:22] <willwh> no :)
[07-Jan-2010 11:17:27] <willwh> we are a small outfit ;)
[07-Jan-2010 11:17:30] <theacolyte-> ah
[07-Jan-2010 11:17:44] <theacolyte-> I had looked into outsourcing it, but ended up doing it internally
[07-Jan-2010 11:17:49] <theacolyte-> powestream was a company I looked into
[07-Jan-2010 11:18:38] <Rolfs> QubeZ: why is it killing your server?
[07-Jan-2010 11:18:57] <theacolyte-> This could all be solved if Zenoss had a button: Import from Cacti XML template
[07-Jan-2010 11:20:28] <QubeZ> Rolfs sometimes python doing an AMI check of the asterisk statistics just hangs and CPU goes to 100%
[07-Jan-2010 11:20:42] <QubeZ> i have 6 asterisk servers monitored, just two python scripts running again two of my servers was at 100% cpu
[07-Jan-2010 11:20:46] <QubeZ> i killed the process
[07-Jan-2010 11:21:26] <theacolyte-> Just ask your users this:
[07-Jan-2010 11:21:50] <theacolyte-> "Look guys, would you rather we monitor for it to be up and down, and keep it down consistantly, or have it down ocassionally and not know?"
[07-Jan-2010 11:23:54] <rmatte> QubeZ: thought there was a brand new Asterisk ZenPack released recently that fixed some of those performance issues?
[07-Jan-2010 11:25:37] <Rolfs> i find it quite usefull to have the "zencommand run -v10 -d ${device/id}" as a command to run on the server from the webgui
[07-Jan-2010 11:26:13] <QubeZ> rmatte the ssh one? its limited in what it reports i believe
[07-Jan-2010 11:35:19] <Rolfs> was it suddently quiet here?
[07-Jan-2010 11:42:08] <Apachez> http://dev.zenoss.com/trac/ticket/6014 http://dev.zenoss.com/trac/ticket/6015 what do you think? feature requests regarding zenoss
[07-Jan-2010 11:44:25] <baffle> Has anyone had problems with going crazy with subnetting and networks? Example: http://paste2.org/p/598627
[07-Jan-2010 11:44:53] <rmatte> Apachez: ask Matt Ray for your own Trac account next time you see him.
[07-Jan-2010 11:44:55] <baffle> That's over 500 networks like 10.0.0.0/24 and 10.0.0.0/23 nested inside eachother..
[07-Jan-2010 11:45:41] <rmatte> baffle: you mean that's the way Zenoss is picking them up?
[07-Jan-2010 11:46:44] <twm1010> There is a way to get zenping to rebuild the topology, isn't there?
[07-Jan-2010 11:46:46] <Apachez> rmatte: will do
[07-Jan-2010 11:47:00] <Apachez> he offered me one yesterday but I didnt think at that time that I needed one :P
[07-Jan-2010 11:47:34] <rmatte> Apachez: ummm, Zenoss does support snmp traps
[07-Jan-2010 11:47:43] <rmatte> Apachez: the only snmp traps it doesn't support are snmpv3
[07-Jan-2010 11:47:59] <rmatte> Apachez: that's what the zentrap daemon handles
[07-Jan-2010 11:48:12] <twm1010> You know, I don't know how the "networks" are generated.
[07-Jan-2010 11:48:35] <twm1010> Is it solely from IP's assigned to Interfaces? or do the routing tables on the routers play a part?
[07-Jan-2010 11:49:15] <twm1010> I used to think the google map relied on routing tables but cluther corrected me and said its only when two interfaces at different sites are on the same subnet
[07-Jan-2010 11:49:35] <baffle> rmatte: Seems like it..
[07-Jan-2010 11:49:42] <rmatte> Apachez: and your request to have Zenoss monitor itself... it basically already does that, just not to the degree that you're specifying... If you go to collectors -> localhost -> perf you'll see performance stats on the daemons, it also alerts on commands that have timed out, and on missed heartbeats between daemons.
[07-Jan-2010 11:49:49] <Apachez> rmatte: then why doesnt my zenoss-stack 2.5.1 listen on udp 162 or whatever the default port is for snmptraps ?
[07-Jan-2010 11:50:00] <baffle> rmatte: Problem is there is a lot of duplicate allocations in the RFC1918 space for our various customers.
[07-Jan-2010 11:50:12] <rmatte> Apachez: you sure you don't have something already running on that port?
[07-Jan-2010 11:50:20] <baffle> rmatte: And they use IPs from that space for internal communication.
[07-Jan-2010 11:50:35] <baffle> rmatte: And it seems ZenOss is unable to understand subnets properly..
[07-Jan-2010 11:50:36] <rmatte> Apachez: I assure you that zenoss listens for traps by default with no configuration required from a standard stack install
[07-Jan-2010 11:50:59] <Apachez> ehm
[07-Jan-2010 11:51:02] <Apachez> no it doesnt :)
[07-Jan-2010 11:51:04] <rmatte> Apachez: I have 13 Zenoss servers running, all properly receiving traps and I didn't have to touch a thing
[07-Jan-2010 11:51:13] <Apachez> or maybe im blind or something
[07-Jan-2010 11:51:24] <Apachez> netstat -an | grep -i listen
[07-Jan-2010 11:51:27] <rmatte> no offence, but you must be lol
[07-Jan-2010 11:51:29] <rmatte> :P
[07-Jan-2010 11:51:33] <Apachez> tcp 0 0 0.0.0.0:8789 0.0.0.0:* LISTEN
[07-Jan-2010 11:51:33] <Apachez> tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN
[07-Jan-2010 11:51:33] <Apachez> tcp 0 0 127.0.0.1:8100 0.0.0.0:* LISTEN
[07-Jan-2010 11:51:33] <Apachez> tcp 0 0 127.0.0.1:3307 0.0.0.0:* LISTEN
[07-Jan-2010 11:51:33] <Apachez> tcp 0 0 0.0.0.0:80 0.0.0.0:* LISTEN
[07-Jan-2010 11:51:33] <Apachez> tcp 0 0 127.0.0.1:8080 0.0.0.0:* LISTEN
[07-Jan-2010 11:51:35] <Apachez> tcp 0 0 0.0.0.0:8081 0.0.0.0:* LISTEN
[07-Jan-2010 11:51:43] <Apachez> where in that list do you see udp 162? :P
[07-Jan-2010 11:51:43] <rmatte> check zentrap.log
[07-Jan-2010 11:52:14] <twm1010> I see it, right there...
[07-Jan-2010 11:52:19] <twm1010> you might need special glasses :)
[07-Jan-2010 11:52:47] <Apachez> what should I look for in the zentraps?
[07-Jan-2010 11:52:59] <rmatte> Apachez: no no no
[07-Jan-2010 11:53:04] <rmatte> Apachez: snmp traps are udp
[07-Jan-2010 11:53:07] <rmatte> not tcp
[07-Jan-2010 11:53:14] <Apachez> yes udp
[07-Jan-2010 11:53:23] <Apachez> and should be visible when you do netstat -an | grep -i listen
[07-Jan-2010 11:53:35] <rmatte> I just did that exact same command on one of my boxes that I am receiving traps on perfectly fine and it doesn't show 162 either
[07-Jan-2010 11:53:35] <Apachez> uhh
[07-Jan-2010 11:53:37] <Apachez> nevermind
[07-Jan-2010 11:53:44] <Apachez> listen doesnt exist in udp
[07-Jan-2010 11:53:46] <Apachez> gah
[07-Jan-2010 11:53:50] <rmatte> correct
[07-Jan-2010 11:53:52] * Apachez goes and hide behind a tree
[07-Jan-2010 11:53:59] <Apachez> well then 6015 can be closed :P
[07-Jan-2010 11:54:05] <rmatte> lol
[07-Jan-2010 11:54:43] <Apachez> and regarding 6014...
[07-Jan-2010 11:54:46] <Apachez> now I see it
[07-Jan-2010 11:55:00] <Apachez> perhaps I should change 6014 into "adding bandwidth graphs for collector as default"
[07-Jan-2010 11:55:26] <Apachez> but perhaps 6014 should remain
[07-Jan-2010 11:55:32] <Apachez> when you go to report -> graph report
[07-Jan-2010 11:55:46] <Apachez> the collector doesnt exist as device for add new graph
[07-Jan-2010 11:55:49] <rmatte> Apachez: the problem is that bandwidth graphs depend on what system it's running on... if it's running on FreeBSD for instance, you wouldn't be gathering the bandwidth info the same way as you would on Linux
[07-Jan-2010 11:56:11] <Apachez> rmatte: isnt there a snmpd for fbsd aswell?
[07-Jan-2010 11:56:15] <rmatte> Apachez: they've tried to make more Zenoss specific monitors rather than system monitors
[07-Jan-2010 11:56:18] <Apachez> I was thinking if that could be included in the stack
[07-Jan-2010 11:56:37] <Apachez> yeah thats what the request is about
[07-Jan-2010 11:56:41] <Apachez> system monitors of the zenoss server
[07-Jan-2010 11:56:50] <rmatte> Apachez: that assumes that the person is running snmpd on the server, which is not always the case
[07-Jan-2010 11:57:09] <rmatte> Apachez: If you need to configure it to monitor locally it's not hard to do it yourself
[07-Jan-2010 11:57:15] <Apachez> which is why it should be included in the stack
[07-Jan-2010 11:57:17] <Apachez> like mysql is
[07-Jan-2010 11:57:37] <Apachez> I know its not hard... it was more of an request for default settings
[07-Jan-2010 11:57:45] <rmatte> I disagree with that, because in the event that you're already running snmpd on the system then it would conflict
[07-Jan-2010 11:58:01] <Apachez> would it ?
[07-Jan-2010 11:58:02] <Apachez> why ?
[07-Jan-2010 11:58:13] <Apachez> the snmpd is not handling traps
[07-Jan-2010 11:58:18] <Apachez> I mean receiving stuff
[07-Jan-2010 11:58:20] <rmatte> well, you'd have 2 snmp daemons, whichever one grabs the ports first would be the active one
[07-Jan-2010 11:58:25] <Apachez> so it doesnt need to listen to any port
[07-Jan-2010 11:58:37] <Apachez> ohh you mean that way
[07-Jan-2010 11:58:43] <rmatte> yes
[07-Jan-2010 11:58:44] <Apachez> well... the same problem exists for tcp8080
[07-Jan-2010 11:58:59] <Apachez> if you already have something running on tcp8080 and tries to install zenoss... what will happen then?
[07-Jan-2010 11:59:02] <rmatte> which is why that's configurable when installing
[07-Jan-2010 11:59:06] <rmatte> we run on port 8180
[07-Jan-2010 11:59:14] <Apachez> the installation will hopefully notify the user
[07-Jan-2010 11:59:22] <rmatte> It does
[07-Jan-2010 11:59:35] <rmatte> If it detects something running on 8080 it prompts for a different port
[07-Jan-2010 11:59:39] <Apachez> then it could do the same with snmpd
[07-Jan-2010 11:59:53] <rmatte> I just don't personally see the value is actually including snmpd in the stack install
[07-Jan-2010 12:00:06] <Apachez> well different port wouldnt be an option perhaps but notify in event that local snmpd is already running therefor the zenoss-snmpd wont start
[07-Jan-2010 12:00:19] <Apachez> the same value as mysql is included? :P
[07-Jan-2010 12:00:39] <rmatte> mysql is part of the core functionality of Zenoss, i.e: it can't work without it
[07-Jan-2010 12:00:51] <rmatte> so that makes sense to have included
[07-Jan-2010 12:00:54] <Apachez> pfff
[07-Jan-2010 12:01:00] <theacolyte-> Man, this new cacti format is really weird
[07-Jan-2010 12:01:05] <Apachez> all im saying is that this was/is just a feature request :)
[07-Jan-2010 12:01:22] <rmatte> I doubt they are going to do it, but alright
[07-Jan-2010 12:01:46] <rmatte> It takes 5 minutes to install snmpd yourself, configure it, and add localhost in as a device
[07-Jan-2010 12:02:08] <Apachez> apt-get install snmpd
[07-Jan-2010 12:02:09] <Apachez> now what? ;)
[07-Jan-2010 12:02:27] <rmatte> take the example config, change 5 lines, and add localhost as a device
[07-Jan-2010 12:02:47] <theacolyte-> What do you guys use for snmp browsing (with MIBs)
[07-Jan-2010 12:03:08] <rmatte> I don't browse Mibs at all personally
[07-Jan-2010 12:03:19] <theacolyte-> So how do you know what OID's to monitor?
[07-Jan-2010 12:03:23] <rmatte> google
[07-Jan-2010 12:03:26] <rmatte> or snmpwalk
[07-Jan-2010 12:03:45] <theacolyte-> How do you know from snmpwalk what an OID does? :P
[07-Jan-2010 12:03:46] <twm1010> theacolyte-: there are a few out there that are free
[07-Jan-2010 12:03:55] <twm1010> they all seem written in java, and are buggy as hell :)
[07-Jan-2010 12:03:59] <rmatte> Apachez: here's a braindead snmpd.conf: http://pastebin.com/m5cd21a9
[07-Jan-2010 12:04:06] <rmatte> Apachez: fill in the blanks and you're rolling
[07-Jan-2010 12:04:13] <twm1010> serverscheck mib browser is the one i use the most
[07-Jan-2010 12:04:23] <twm1010> or i jump onto our solarwinds box and check the mib database
[07-Jan-2010 12:04:47] <twm1010> too bad i can't find some way to just dump that whole thing into zenoss, now that would be nice :)
[07-Jan-2010 12:05:06] <rmatte> theacolyte-: in some cases you can figure it out, some cases you can't, but a quick google generally yields the answer, unless it's some obscure device
[07-Jan-2010 12:05:16] <theacolyte-> I can't figure it out. hehe
[07-Jan-2010 12:05:17] <rmatte> theacolyte-: in which case I suppose then I'd dig in to the Mibs
[07-Jan-2010 12:05:42] <rmatte> theacolyte-: you could try the mib browser ZenPack for Zenoss
[07-Jan-2010 12:06:50] <rmatte> Apachez: you probably want to close out http://dev.zenoss.com/trac/ticket/6015
[07-Jan-2010 12:07:24] <theacolyte-> bah only does v1 tho
[07-Jan-2010 12:08:26] <rmatte> oh, didn't know that
[07-Jan-2010 12:08:31] <rmatte> thought it did v2 too
[07-Jan-2010 12:11:29] <theacolyte-> holy crap that serverscheck browser is buggy
[07-Jan-2010 12:13:02] <rmatte> lol
[07-Jan-2010 12:20:40] <twm1010> yeah, you have to click a lot of buttons twice
[07-Jan-2010 12:20:42] <twm1010> hehehe
[07-Jan-2010 12:20:53] <twm1010> but if you can get a mib loaded, its pretty nice because it will do walks, and check tables, etc
[07-Jan-2010 12:21:13] <twm1010> it seems like every free one out there uses that basic layout and underlying engine, and is just as buggy
[07-Jan-2010 12:21:13] <rmatte> cool
[07-Jan-2010 12:21:20] <rmatte> hehe
[07-Jan-2010 12:21:55] <twm1010> in fact, some kind of DLL or other app error maybe java runtime, i haven't cared to look, won't let me even run that on my own workstation now
[07-Jan-2010 12:23:58] <theacolyte-> Oh, I found something
[07-Jan-2010 12:24:07] <theacolyte-> ireasoning mib browser
[07-Jan-2010 12:24:21] <theacolyte-> awful to launch on windows like all java apps, but THIS is nice.
[07-Jan-2010 12:27:10] <twm1010> i think i tried that one too, lemme look
[07-Jan-2010 12:27:30] <twm1010> yeah, i have it
[07-Jan-2010 12:27:54] <Apachez> hmpf
[07-Jan-2010 12:27:58] <twm1010> im amazed my own workstation hasn't exploded, i'm usually anal retentive about not having too many apps installed, but my start menu is like 3 pages long
[07-Jan-2010 12:28:06] <Apachez> how do I configure snmpd to monitor my local interfaces?
[07-Jan-2010 12:28:14] * Apachez runs away to google
[07-Jan-2010 12:28:54] <Apachez> and wheres mray when you need him? :P
[07-Jan-2010 12:28:56] <willwh> twm1010: it happens too easily
[07-Jan-2010 12:29:06] <willwh> I just got a new box here; win7 64bit (I like it)
[07-Jan-2010 12:29:13] <willwh> already about 2 pages of installed crap
[07-Jan-2010 12:29:19] <rmatte> Apachez: it does that by default
[07-Jan-2010 12:29:20] <willwh> when I was younger I used to organize it all
[07-Jan-2010 12:29:25] <willwh> I can't be arsed any longer
[07-Jan-2010 12:29:34] <rmatte> Apachez: just add localhost as a device and model, you'll see the interfaces on the OS tab
[07-Jan-2010 12:29:58] <twm1010> I actually use an often ignored feature of the windows taskbar and created my own toolbar
[07-Jan-2010 12:30:11] <Apachez> rmatte: ill try that
[07-Jan-2010 12:30:22] <twm1010> hidden at the top of my screen, stored in my home directory, collection of shortcuts, for XP it gets the job done
[07-Jan-2010 12:30:38] <twm1010> i like win7's dock, very mac-like :)
[07-Jan-2010 12:30:46] <theacolyte-> This is weird
[07-Jan-2010 12:31:01] <theacolyte-> I'm getting the OID's from this browser for only one controller, but all of them are doing that
[07-Jan-2010 12:31:05] <theacolyte-> (this is a SAN with 2 controllers)
[07-Jan-2010 12:31:31] <Apachez> by the way
[07-Jan-2010 12:31:36] <Apachez> the different device groups
[07-Jan-2010 12:31:39] <Apachez> they are just groups right?
[07-Jan-2010 12:31:45] <twm1010> device groups?
[07-Jan-2010 12:31:56] <Apachez> I mean for a default install the group "Network" isnt anything more than just a groupname?
[07-Jan-2010 12:32:04] <Apachez> or are there special features connected to each group?
[07-Jan-2010 12:32:11] <twm1010> do you mean device classes?
[07-Jan-2010 12:32:18] <twm1010> like /server/windows ?
[07-Jan-2010 12:32:19] <Apachez> I mean will I brake anything if I delete all device classes?
[07-Jan-2010 12:32:21] <Apachez> yeah
[07-Jan-2010 12:32:28] <twm1010> yeah, absolutely, do not do that
[07-Jan-2010 12:32:40] <twm1010> the classes control the settings, what scripts/plugins are used to model devices
[07-Jan-2010 12:32:40] <Apachez> hmpf
[07-Jan-2010 12:32:42] <Apachez> too late :)
[07-Jan-2010 12:32:49] <twm1010> dood, reinstall
[07-Jan-2010 12:33:10] <twm1010> re-doing all of that by hand is... going to suck
[07-Jan-2010 12:33:17] <Apachez> gah damnit
[07-Jan-2010 12:33:29] <Apachez> well its a testsetup so well :)
[07-Jan-2010 12:33:36] <twm1010> If you need to logically group things, uses "systems" or "groups" organizers instead
[07-Jan-2010 12:33:56] <twm1010> then you can do all sorts of good stuff, without affecting what settings are used against the devices
[07-Jan-2010 12:34:12] <Apachez> well I was thinking of having the groups directly as sub devices
[07-Jan-2010 12:34:22] <Apachez> so I dont have to click gazillion of times
[07-Jan-2010 12:34:44] <Apachez> where can I read more about these different classes?
[07-Jan-2010 12:34:57] <Apachez> like is /device/network/switch different from /device/network/router ?
[07-Jan-2010 12:35:04] <Apachez> other then just the name of course
[07-Jan-2010 12:35:05] <twm1010> think of the device classes as a prebuilt filesystem hierarchy
[07-Jan-2010 12:35:30] <twm1010> with a ton of preloaded and configured and inheritable settings and templates
[07-Jan-2010 12:35:44] <twm1010> you can of course blow the whole thing away if you know it top to bottom
[07-Jan-2010 12:36:18] <twm1010> but you'll have to change every bloody setting and there are tons of them not to mention some of the templates were stored at lower levels of that hierarchy
[07-Jan-2010 12:36:21] <Apachez> well its organizers that I have added
[07-Jan-2010 12:36:35] <twm1010> So like.... we have 3 active directories right?
[07-Jan-2010 12:36:42] <Apachez> in my case?
[07-Jan-2010 12:36:44] <twm1010> So I didn't nuke /server/windows and created /windows
[07-Jan-2010 12:36:59] <twm1010> I created /Server/Windows/AD1/
[07-Jan-2010 12:37:04] <twm1010> and /Server/Windows/AD2/ etc
[07-Jan-2010 12:37:24] <Apachez> well I have SLA, ISP_<name> and <name>NET
[07-Jan-2010 12:37:27] <twm1010> so i can assign different zwinUsername and zwinPasswords to different windows servers
[07-Jan-2010 12:37:50] <Apachez> so what I would do is to create all those three in /network instead ?
[07-Jan-2010 12:38:16] <twm1010> well you use custom organizers under /Devices to choose what settings apply
[07-Jan-2010 12:38:36] <twm1010> not so much how your devices relate to each other thought they may already
[07-Jan-2010 12:38:54] <markeriv> hey guys, i am fiddling through the reporting section and trying to figure out how to schedule reports and auto email them... anyone care to point the blind man in the right way?
[07-Jan-2010 12:38:54] <theacolyte-> man this mib stuff is confusing, it's been so long since I've used it
[07-Jan-2010 12:39:16] <theacolyte-> Why is it that one of my OID's thats confirmed working in Zenoss (at least I think it's the right value) doesn't match the one this MIB is giving me?
[07-Jan-2010 12:39:19] <twm1010> markeriv: only known way is the cron and the utility reportmail
[07-Jan-2010 12:39:27] <Apachez> hmm
[07-Jan-2010 12:39:28] <twm1010> markeriv: search the forum for it, you'll find what you need quick enough
[07-Jan-2010 12:39:39] <Apachez> so whats the zenoss proper way of group things then?
[07-Jan-2010 12:39:50] <twm1010> Well, you're grouping them for what purpose
[07-Jan-2010 12:40:01] <Apachez> to know what the device is meant for
[07-Jan-2010 12:40:04] <twm1010> a common set of settings or templates to monitor with?
[07-Jan-2010 12:40:12] <Apachez> like if its a device that I only to ping checks against (SLA)
[07-Jan-2010 12:40:31] <Apachez> or if its a device within my isp that I do pingchecks against (ISP_<name>)
[07-Jan-2010 12:40:32] <twm1010> In that case there was a built-in class called /Ping I believe
[07-Jan-2010 12:40:50] <Apachez> or if its my own equipment which I do pingchecks + snmp + syslog/snmptraps (<name>NET)
[07-Jan-2010 12:41:00] <twm1010> those devices would also go under /Ping
[07-Jan-2010 12:41:14] <twm1010> your own equipment, would go in the device classes that make sense
[07-Jan-2010 12:41:22] <Apachez> what I did was to create my own organizers and place the devices in there
[07-Jan-2010 12:41:22] <twm1010> like... /Servers/Windows /Server/Linux
[07-Jan-2010 12:41:29] <Apachez> and then change the template to bind to b_fping
[07-Jan-2010 12:41:31] <Apachez> and voila
[07-Jan-2010 12:41:46] <rmatte> Apachez: yeh, you should leave the default device organizer and just add on to them... like /Network/Switch/Cisco or /Network/Switch/HP
[07-Jan-2010 12:41:53] <twm1010> By deleting the default classes though you deleted a ton of settings and templates
[07-Jan-2010 12:42:02] <rmatte> Apachez: you really shouldn't be deleting the default stuff, reinstall
[07-Jan-2010 12:42:28] <twm1010> look at zProperties of a class to get an idea what I'm saying
[07-Jan-2010 12:42:34] <rmatte> Like twm1010 stated, you just nuked a bunch of templates and settings by doing that
[07-Jan-2010 12:42:56] <theacolyte-> ahhhhh ha
[07-Jan-2010 12:43:03] <theacolyte-> interface issue re: my previous question, foudn it
[07-Jan-2010 12:44:11] <twm1010> Apachez: My recommendation to you... is to take a break, and sit and slowly read the admin guide's section on device classes
[07-Jan-2010 12:44:34] <rmatte> read the entire admin guide top to bottom honestly
[07-Jan-2010 12:44:46] <rmatte> it'll answer 90% of the questions that you're likely to ask
[07-Jan-2010 12:44:46] <theacolyte-> I really should do that
[07-Jan-2010 12:44:59] <twm1010> Chapter 9 is what I'm referring to
[07-Jan-2010 12:45:01] <theacolyte-> But before I do, riddle me this: Can you justm onitor a text string snmp value?
[07-Jan-2010 12:45:07] <theacolyte-> good vs bad, that kind of thing
[07-Jan-2010 12:45:29] <rmatte> theacolyte-: you would probably have to do that via a script and command datasource
[07-Jan-2010 12:45:36] <theacolyte-> ah
[07-Jan-2010 12:45:51] <theacolyte-> Would want to do something like monitor battery status on this raid controller
[07-Jan-2010 12:46:12] <rmatte> I'm surprised it doesn't give the status as a numeric value
[07-Jan-2010 12:46:14] <rmatte> or does it?
[07-Jan-2010 12:46:16] <twm1010> are you sure it doesn't return an integer?
[07-Jan-2010 12:46:19] <theacolyte-> it says "Good"
[07-Jan-2010 12:46:30] <twm1010> well yeah, but is it translating a 1 to = good ?
[07-Jan-2010 12:46:31] <theacolyte-> it is an integre though
[07-Jan-2010 12:46:32] <rmatte> that's really really bad practice on the part of the manufacturer lol
[07-Jan-2010 12:46:33] <theacolyte-> ah
[07-Jan-2010 12:46:43] <rmatte> oh, it is an integer
[07-Jan-2010 12:46:45] <twm1010> i mean is your mib browser doing that for you
[07-Jan-2010 12:46:47] <theacolyte-> weird that it would translate 1 to good
[07-Jan-2010 12:46:54] <twm1010> it will say it in the mib description sometimes
[07-Jan-2010 12:47:00] <rmatte> well, the Mib probably has the definition for that
[07-Jan-2010 12:47:10] <twm1010> do an snmpget from your zenoss server's commandline against the device, and see what you get back
[07-Jan-2010 12:47:13] <rmatte> you need to make Zenoss understand that 1 is good and 0 is bad via an event transform
[07-Jan-2010 12:47:15] <twm1010> if it says 1 or 5 or hatever
[07-Jan-2010 12:47:16] <theacolyte-> you're right
[07-Jan-2010 12:47:24] <theacolyte-> syntax
[07-Jan-2010 12:47:28] <rmatte> (if 0 actually is bad)
[07-Jan-2010 12:47:37] <theacolyte-> 0 is not present, 2 is bad
[07-Jan-2010 12:47:41] <theacolyte-> 3 is unknown
[07-Jan-2010 12:47:44] <theacolyte-> 1 guess as to what 1 is
[07-Jan-2010 12:48:16] <rmatte> right, is that value sent via a trap or polled?
[07-Jan-2010 12:48:25] <theacolyte-> Not sure how to even tell
[07-Jan-2010 12:48:38] <rmatte> well, it's available via an snmpwalk yes?
[07-Jan-2010 12:48:46] <theacolyte-> yeah
[07-Jan-2010 12:48:53] <rmatte> ok, then here's how you go about it...
[07-Jan-2010 12:49:02] <rmatte> add an snmp datasource for that OID
[07-Jan-2010 12:49:19] <rmatte> then create a threshold with a max value of 2 and a min value of 0
[07-Jan-2010 12:49:25] <twm1010> i might be done both, things get dicey here...
[07-Jan-2010 12:49:33] <rmatte> that way a threshold will kick off if it dips below 1 or goes above 1
[07-Jan-2010 12:49:34] <twm1010> i find it all depend on whether you trust traps or not
[07-Jan-2010 12:49:39] <theacolyte-> rmatte: oh, yeah of course
[07-Jan-2010 12:49:49] <rmatte> then create an event transform to translate the value to whatever it means when an event is generated
[07-Jan-2010 12:50:11] <twm1010> rmatte: I'm surprised there isn't a mechanism to generate an event transform from a MIB
[07-Jan-2010 12:50:29] <rmatte> twm1010: well, that would be pretty tricky
[07-Jan-2010 12:50:53] <rmatte> twm1010: the difference between what's needed from 1 to the other would vary greatly
[07-Jan-2010 12:50:55] <twm1010> sure, its so inconsistent, some manufacturers have good mibs and others simply say "its broken"
[07-Jan-2010 12:51:02] <rmatte> yup
[07-Jan-2010 12:51:18] <rmatte> it really does need to be manually done unfortunately
[07-Jan-2010 12:51:29] <twm1010> i did finally get your cisco zenpack installed, i haven't run the command to remove the entry yet
[07-Jan-2010 12:51:33] <rmatte> but a bit of dev time and you've got what you need
[07-Jan-2010 12:51:36] <twm1010> the gui is a little sluggish rigt now
[07-Jan-2010 12:51:50] <jb> erm, is it me, or can you not add an alert filter based on group class?
[07-Jan-2010 12:51:56] <rmatte> twm1010: your zProperties pages will take 5 to 15 minutes to load until you run that zendmd command
[07-Jan-2010 12:52:00] <twm1010> jb: let me look
[07-Jan-2010 12:52:16] <rmatte> twm1010: as well as the ZenPack tab in Settings, and some other pages (like when you're trying to add a datasource for example)
[07-Jan-2010 12:52:29] <twm1010> yeah, i've noticed
[07-Jan-2010 12:52:40] <rmatte> twm1010: It was a nightmare before I found that fix lol
[07-Jan-2010 12:52:54] <twm1010> jb: I see device groups as a filter
[07-Jan-2010 12:53:07] <rmatte> anyways, I need to get back to commissioning devices
[07-Jan-2010 12:53:26] <rmatte> jb: you can
[07-Jan-2010 12:53:37] <rmatte> jb: we had ours filtering based on like 4 different groups at one point
[07-Jan-2010 12:53:56] <Apachez> hmpf
[07-Jan-2010 12:54:00] <twm1010> rmatte: I think I'll do it now :)
[07-Jan-2010 12:54:06] <rmatte> twm1010: good plan
[07-Jan-2010 12:54:07] <Apachez> zenoss couldnt detect any interfaces on localhost when I did snmp
[07-Jan-2010 12:54:20] <Apachez> who in here claimed it would ? :P
[07-Jan-2010 12:54:22] <rmatte> Apachez: try snmpwalking localhost, see if it actually returns anything
[07-Jan-2010 12:54:40] <rmatte> Apachez: I know for a fact that it will, provided you configure snmpd properly
[07-Jan-2010 12:54:43] <twm1010> Apachez: If you deleted the standard /Device class structure then it probably doesn't have the interface modeling plugins assigned to the class :)
[07-Jan-2010 12:54:57] <rmatte> twm1010: actually, that's a good point
[07-Jan-2010 12:55:15] <rmatte> twm1010: though I thought that plugin was available at the root level?
[07-Jan-2010 12:55:20] <Apachez> twm1010: it worked to model my switches?
[07-Jan-2010 12:55:21] * rmatte checks
[07-Jan-2010 12:55:32] <twm1010> rmatte: do i have to restart zope afterwards?
[07-Jan-2010 12:55:34] <Apachez> so I guess its at root level
[07-Jan-2010 12:55:44] <rmatte> twm1010: nope, no restart required
[07-Jan-2010 12:55:59] <rmatte> twm1010: you should just see an instant improvement
[07-Jan-2010 12:56:33] <Apachez> walking localhost works just fine
[07-Jan-2010 12:56:42] <Apachez> but no interfaces are returned
[07-Jan-2010 12:56:51] <rmatte> Apachez: the other thing is that you blew away the /Server/Linux class which has beautiful pre-made templates for Linux servers
[07-Jan-2010 12:56:57] <Apachez> snmpwalk -v 2c -c public localhost
[07-Jan-2010 12:56:59] <rmatte> Apachez: but like I said, snmpwalk localhost
[07-Jan-2010 12:57:14] <Apachez> SNMPv2-MIB::sysDescr.0 = STRING: Linux monitor 2.6.31-16-generic #53-Ubuntu SMP Tue Dec 8 04:02:15 UTC 2009 x86_64
[07-Jan-2010 12:57:21] <Apachez> yaddayadda but nothing about interfaces
[07-Jan-2010 12:57:26] <rmatte> can you pastebin the entire output of snmpwalk?
[07-Jan-2010 12:57:48] <rmatte> I've seen partial snmpwalks due to incorrect configuration before, I want to see what it's giving you
[07-Jan-2010 12:58:05] <Apachez> sure hang on
[07-Jan-2010 12:58:19] <rmatte> k
[07-Jan-2010 12:59:01] <twm1010> does the remove command take a few minutes as well?
[07-Jan-2010 12:59:07] <rmatte> yes
[07-Jan-2010 12:59:11] <twm1010> ah, gotcha
[07-Jan-2010 12:59:12] <rmatte> like 15 mins or so
[07-Jan-2010 12:59:20] <rmatte> the commit will take about 7 or so
[07-Jan-2010 12:59:51] <Apachez> http://www.pastebin.ca/1741723
[07-Jan-2010 13:00:03] <Apachez> there the walk of localhost
[07-Jan-2010 13:00:13] <rmatte> that's the entire walk that it's giving you?
[07-Jan-2010 13:00:28] <rmatte> yeh, you definitely have a configuration issue
[07-Jan-2010 13:00:29] <Apachez> what I did is apt-get install snmpd then I changed the community (public in the pastebin but its named )
[07-Jan-2010 13:00:31] <Apachez> yup
[07-Jan-2010 13:00:46] <twm1010> Edited Zenpack info page to include the wait.
[07-Jan-2010 13:00:51] <rmatte> you modified the config that I pastebinned you?
[07-Jan-2010 13:00:59] <rmatte> twm1010: thanks
[07-Jan-2010 13:01:08] <Apachez> which config?
[07-Jan-2010 13:01:20] <Apachez> I think the group stuff is wrong
[07-Jan-2010 13:01:33] <Apachez> MyROGroup v2c need to be pointed to current secmodel
[07-Jan-2010 13:01:40] <rmatte> http://pastebin.com/m64e02c49
[07-Jan-2010 13:01:46] <rmatte> you need to edit the following lines...
[07-Jan-2010 13:01:52] <rmatte> trapsink 192.168.0.1 your-key
[07-Jan-2010 13:01:58] <rmatte> trap2sink 192.168.0.1 your-key
[07-Jan-2010 13:02:13] <rmatte> for traps, so set the IP as 127.0.0.1 and yourkey to the snmp string
[07-Jan-2010 13:02:17] <rmatte> and
[07-Jan-2010 13:02:18] <rmatte> rocommunity your-key
[07-Jan-2010 13:02:23] <rmatte> rwcommunity your-key
[07-Jan-2010 13:02:34] <rmatte> set whatever keys you want for read write and read only
[07-Jan-2010 13:02:40] <rmatte> then
[07-Jan-2010 13:02:41] <rmatte> syslocation "Set this to the location"
[07-Jan-2010 13:02:44] <rmatte> set the location info
[07-Jan-2010 13:02:48] <rmatte> and
[07-Jan-2010 13:02:49] <rmatte> syscontact whoever@contact.com
[07-Jan-2010 13:02:52] <rmatte> the contact info
[07-Jan-2010 13:02:56] <rmatte> then save it and restart snmpd
[07-Jan-2010 13:03:00] <rmatte> and snmpwalk again
[07-Jan-2010 13:03:11] <rmatte> don't touch the stuff at the bottom of the config at all
[07-Jan-2010 13:03:21] <Apachez> hmm
[07-Jan-2010 13:08:16] <theacolyte-> does zenoss figure out 64bit counters vs 32?
[07-Jan-2010 13:08:35] <rmatte> theacolyte-: if the interface is of type ethernetCsmacd_64 then yes
[07-Jan-2010 13:09:05] <twm1010> rmatte: mmm.. fast gui again :D
[07-Jan-2010 13:09:10] <rmatte> twm1010: cool
[07-Jan-2010 13:09:12] <Apachez> hmm
[07-Jan-2010 13:09:15] <Apachez> its related to the access
[07-Jan-2010 13:09:37] <theacolyte-> man,. this is great, remembering how to use snmp again
[07-Jan-2010 13:09:44] <theacolyte-> if you don't use it, you lose it
[07-Jan-2010 13:09:51] <rmatte> Apachez: if you take the config that I pastebinned and modify it the way I told you it'll work, it's the exact config that we're using to monitor all of our Linux boxes (including interfaces)
[07-Jan-2010 13:09:59] <theacolyte-> basic stuff is easy, but trying to find stuff when it isn't very well documented, is...
[07-Jan-2010 13:10:19] <Apachez> com2sec readonly default public
[07-Jan-2010 13:10:22] <rmatte> theacolyte-: that's when you come to me :)
[07-Jan-2010 13:10:24] <Apachez> and then walk "public" will work
[07-Jan-2010 13:10:35] <Apachez> but if I change that to some other community (same as paranoid) then it doesnt
[07-Jan-2010 13:10:56] <theacolyte-> jehehe
[07-Jan-2010 13:11:23] <rmatte> Apachez: which is why I said not to touch the stuff at the bottom, that's default stuff that shouldn't be changed
[07-Jan-2010 13:11:29] <rmatte> Apachez: the keys are set higher up in the config
[07-Jan-2010 13:11:42] <markeriv> anyone know scripting well? I keep getting this error when trying to follow the Scheduling reports instructions on the forums
[07-Jan-2010 13:11:55] <theacolyte-> OO
[07-Jan-2010 13:11:57] <theacolyte-> this is useful
[07-Jan-2010 13:11:59] <theacolyte-> http://oss.oetiker.ch/rrdtool/doc/rrdgraph_graph.en.html
[07-Jan-2010 13:12:21] <markeriv> bah
[07-Jan-2010 13:12:24] <markeriv> aint letting me link it
[07-Jan-2010 13:12:31] <rmatte> theacolyte-: yup, very
[07-Jan-2010 13:12:32] <markeriv> emailutilization2mark.sh: line 4: unexpected EOF while looking for matching `"'
[07-Jan-2010 13:12:51] <rmatte> markeriv: I would say you're missing a quotation mark somewhere
[07-Jan-2010 13:13:04] <markeriv> hmmm
[07-Jan-2010 13:14:12] <markeriv> found it, --address="joe@example.com \
[07-Jan-2010 13:14:17] <rmatte> ;)
[07-Jan-2010 13:14:18] <markeriv> missing in the help doc lol
[07-Jan-2010 13:14:52] <rmatte> generally whenever you see a message that says anything like "while looking for matching" you're just missing some closing bracket or quotation mark
[07-Jan-2010 13:14:52] <Apachez> rocommunity your-key
[07-Jan-2010 13:15:00] <rmatte> Apachez: yes, that line
[07-Jan-2010 13:15:15] <rmatte> change your-key to the string you want
[07-Jan-2010 13:15:15] <markeriv> nod, just not used to scripting
[07-Jan-2010 13:15:16] <Apachez> and change that into what ?
[07-Jan-2010 13:15:21] <markeriv> wish i knew more about it
[07-Jan-2010 13:15:26] <rmatte> rocommunity thisismykey
[07-Jan-2010 13:15:28] <rmatte> whatever you want
[07-Jan-2010 13:15:29] <markeriv> --address="joe@example.com" \
[07-Jan-2010 13:15:35] <Apachez> ohh the name of the community?
[07-Jan-2010 13:15:40] <Apachez> rocommunity public
[07-Jan-2010 13:15:41] <Apachez> like that ?
[07-Jan-2010 13:15:52] <rmatte> markeriv: you'll get the hang of it
[07-Jan-2010 13:15:58] <rmatte> yes
[07-Jan-2010 13:16:05] <rmatte> same for rwcommunity private
[07-Jan-2010 13:16:21] <markeriv> oie, now i am getting a report mail error... ReportMail.py: error: no such option: --password
[07-Jan-2010 13:16:31] <rmatte> the trapsink and trap2sink lines should be configured for 127.0.0.1 public
[07-Jan-2010 13:16:44] <rmatte> and then just configure the location and contact info, restart snmpd and you're in business
[07-Jan-2010 13:16:46] <Apachez> doh
[07-Jan-2010 13:16:49] <Apachez> man smpd.conf ftw :)
[07-Jan-2010 13:16:50] <Apachez> rocommunity COMMUNITY [SOURCE [OID | -V VIEW [CONTEXT]]]
[07-Jan-2010 13:16:52] <QubeZ> anyone familiar with the new asterisk zenpack (ssh)? I set it up properly. Even when I run the cmd it tries to execute remotely, I get data back but the graphs are not populating. Do I need to remodel the device?
[07-Jan-2010 13:17:08] <rmatte> markeriv: hmmmm, is that a supported option?
[07-Jan-2010 13:17:23] <Apachez> and if I dont configure rwcommunity then there will be no rw access right ?
[07-Jan-2010 13:17:28] <rmatte> QubeZ: did you wait 3 polling cycles?
[07-Jan-2010 13:17:31] <Apachez> or will it default to something terrible
[07-Jan-2010 13:17:41] <QubeZ> rmatte i've waited about an hour
[07-Jan-2010 13:17:45] <markeriv> oie
[07-Jan-2010 13:17:53] <rmatte> Apachez: there just won't be rw access, you can comment that out if you want
[07-Jan-2010 13:17:59] <markeriv> this preformmated script in the help doc is full of holes
[07-Jan-2010 13:18:05] <rmatte> QubeZ: hmmm
[07-Jan-2010 13:18:30] <QubeZ> rmatte hmm now im seeing this error: Cmd: /usr/bin/ssh root@10.253.8.11 /zenoss/show_core_channels.sh - Code: 255 - Msg: Exit status out of range, exit takes only integer arguments in the range 0-255
[07-Jan-2010 13:18:31] <rmatte> markeriv: seems like it
[07-Jan-2010 13:18:37] <rmatte> markeriv: repost it afterwards if you can
[07-Jan-2010 13:18:47] <Apachez> VICTORY!
[07-Jan-2010 13:18:53] <Apachez> rocommunity public 127.0.0.1
[07-Jan-2010 13:18:56] <Apachez> was the trick
[07-Jan-2010 13:19:11] <rmatte> QubeZ: for some reason it's trying to exit out of the exit value range
[07-Jan-2010 13:19:12] <Apachez> funny that this part wasnt in the default snmpd.conf in ubuntu 9.10
[07-Jan-2010 13:19:20] <markeriv> oie
[07-Jan-2010 13:19:31] <markeriv> so got past those bugs, and now it is giving me some crazy stuff
[07-Jan-2010 13:19:51] <theacolyte-> Ok so I actually figured out what my problem was, I just don't know how to fix it
[07-Jan-2010 13:20:26] <markeriv> i think it isnt liking the space in ="$REPORTS_URL/Reports/Graph Reports/Windsor"
[07-Jan-2010 13:20:33] <rmatte> Apachez: we used an snmpd config generator, I forget what the command was
[07-Jan-2010 13:20:34] <theacolyte-> I'm pullg RxData/Txdata, which are very large numbers because they appear to be aggregate totals - for isntance, this one is 136T
[07-Jan-2010 13:20:43] <theacolyte-> But, cacti can extract MB/sec from that
[07-Jan-2010 13:20:49] <theacolyte-> How can I do that with zenoss?
[07-Jan-2010 13:21:04] <rmatte> markeriv: if it's in quotes it shouldn't care about the space
[07-Jan-2010 13:21:23] <rmatte> markeriv: what's the error?
[07-Jan-2010 13:21:41] <markeriv> figured out that error, now on to the next one
[07-Jan-2010 13:21:51] <markeriv> didnt like $REPORTS_URL
[07-Jan-2010 13:21:52] <twm1010> theacolyte-: If it's a (TOTAL) style counter, you should change the datapoint type from gauge to derive
[07-Jan-2010 13:21:56] <markeriv> wanted me to put in the whole
[07-Jan-2010 13:22:03] <markeriv> public IP of the server
[07-Jan-2010 13:22:16] <twm1010> that way it derives the difference between the previous and current value, and divides it automatically by the polling interval
[07-Jan-2010 13:22:20] <Apachez> wohhooo ITS ALIVE
[07-Jan-2010 13:22:21] <Apachez> :)
[07-Jan-2010 13:22:25] <theacolyte-> twm1010: let me try that, sec
[07-Jan-2010 13:22:27] <twm1010> effectively giving you what you're after
[07-Jan-2010 13:22:40] <twm1010> you will have to delete your RRD files if you change your datapoint type
[07-Jan-2010 13:22:43] <markeriv> and now on to the next error
[07-Jan-2010 13:22:48] <markeriv> \sigh
[07-Jan-2010 13:22:51] <rmatte> theacolyte-: look at the way the default ethernetCsmacd template is configured, the datasources and datapoints are setup in a certain way
[07-Jan-2010 13:23:30] <markeriv> does anyone else consistently have dns issues with servers running zenoss?
[07-Jan-2010 13:24:11] <twm1010> markeriv: In my time with Zenoss I have found it to be good practice to have your DNS configuration on the host damn near perfect before installing it :)
[07-Jan-2010 13:24:18] <theacolyte-> I'm NAN'd at the moment, so I'll have to wait
[07-Jan-2010 13:24:31] <markeriv> i just keep having the server clear its dns server
[07-Jan-2010 13:24:33] <markeriv> like
[07-Jan-2010 13:24:37] <markeriv> it isnt looking to 4.2.2.1
[07-Jan-2010 13:24:46] <markeriv> and looks locally
[07-Jan-2010 13:25:16] <markeriv> i can manually force it back from the box, but i think part of the issue is how zenoss allows you to search by name
[07-Jan-2010 13:25:36] <markeriv> its like zenoss has its own built in dns server that it keeps defaulting to
[07-Jan-2010 13:25:39] <twm1010> if you do a search in zenoss, it doesn't search DNS, it searches known devicd IDs
[07-Jan-2010 13:26:10] <twm1010> when a device is discovered, it resolves the hostname via DNS, but thats about it, and it will probably do that using the OS's builting mechanisms
[07-Jan-2010 13:26:25] <rmatte> yeh, it searches zope, doesn't touch dns
[07-Jan-2010 13:26:26] <twm1010> so whatever the zenoss server's /etc/resolv.conf is set to
[07-Jan-2010 13:26:40] <rmatte> (when doing a search for a device)
[07-Jan-2010 13:26:48] <rmatte> afk for a few
[07-Jan-2010 13:28:06] <twm1010> take my leave for today, i have to focus on some things, take it easy
[07-Jan-2010 13:28:50] <theacolyte-> BAM
[07-Jan-2010 13:28:53] <theacolyte-> it was derive
[07-Jan-2010 13:28:55] <theacolyte-> thanks!
[07-Jan-2010 13:33:59] <theacolyte-> man, you guys are great
[07-Jan-2010 13:40:20] <rmatte> hehe
[07-Jan-2010 13:40:50] <rmatte> yeh, I have to take my leave too for a bit, have to finish loading devices in to a new Zenoss install
[07-Jan-2010 13:40:52] <rmatte> cheers for now
[07-Jan-2010 13:43:12] <rmatte> actually, I'm back, devices are taking a while to load so I have them lined up ready to go lol
[07-Jan-2010 13:43:14] <theacolyte-> I am goign to start getting really annoyed about RRD Files being missing
[07-Jan-2010 13:43:16] <theacolyte-> hehe
[07-Jan-2010 13:43:30] <rmatte> they shouldn't be missing unless it's not collecting properly
[07-Jan-2010 13:43:46] <theacolyte-> it seems like you are in a world of hurt if you ever remove/rename datapoints/graphpoints
[07-Jan-2010 13:44:09] <rmatte> the RRD files correspond to the datapoint names
[07-Jan-2010 13:44:14] <rmatte> not the graphpoint names though
[07-Jan-2010 13:44:17] <theacolyte-> It should collect fine, the OID test works
[07-Jan-2010 13:44:20] <rmatte> you can rename a graphpoint with no effect
[07-Jan-2010 13:44:56] <rmatte> right, but the OID test doesn't necessarily mean you have the right OID, since it uses snmpwalk
[07-Jan-2010 13:45:04] <rmatte> you might need to add a .0 to the end of the OID
[07-Jan-2010 13:45:14] <rmatte> you're probably having it try to gather an snmp tree
[07-Jan-2010 13:45:15] <theacolyte-> Removing rhe device and readding it always seems to fix it though
[07-Jan-2010 13:45:16] <rmatte> which it won't like
[07-Jan-2010 13:45:30] <theacolyte-> I'm testing the OID from zenoss tho
[07-Jan-2010 13:45:35] <theacolyte-> Which appears to work with snmpwalk
[07-Jan-2010 13:45:40] <rmatte> right, by clicking the test button right?
[07-Jan-2010 13:45:42] <theacolyte-> yeah
[07-Jan-2010 13:45:45] <rmatte> ok, let me explain this better
[07-Jan-2010 13:45:54] <theacolyte-> I've walked a tree with that before, I know what it looks like
[07-Jan-2010 13:45:59] <theacolyte-> hehe
[07-Jan-2010 13:45:59] <rmatte> if I do an snmpwalk on hrProcessorLoad on a machine with 1 processor, I'll get 1 line returned
[07-Jan-2010 13:46:10] <rmatte> but on a multi-cpu machine I'll get multiple lines returned from the same OID
[07-Jan-2010 13:46:16] <rmatte> That's an snmp tree
[07-Jan-2010 13:46:21] <theacolyte-> Right, I know what you're saying
[07-Jan-2010 13:46:27] <rmatte> Zenoss does not like those, though you won't be able to tell with snmpwalk
[07-Jan-2010 13:46:29] <theacolyte-> If you add aa 1 you get proc 1, 2 proc 2
[07-Jan-2010 13:46:35] <theacolyte-> hmmm didn't know that
[07-Jan-2010 13:46:42] <rmatte> so you need to add a .0 to the end of the OID for Zenoss to use it
[07-Jan-2010 13:46:54] <rmatte> Zenoss knows the difference between an actual OID and a tree
[07-Jan-2010 13:48:10] <theacolyte-> I read ya
[07-Jan-2010 13:48:26] <theacolyte-> This isn't that though ,believe me
[07-Jan-2010 13:48:43] <theacolyte-> There's actually a few forum posts about it
[07-Jan-2010 13:49:24] <rmatte> hmmm, I've never really had issues like what you're describing
[07-Jan-2010 13:49:56] <rmatte> other than having to add a .0 to the end, that's really the only scenario where I've waited a long time and not had an RRD generated
[07-Jan-2010 13:50:46] <theacolyte-> This behavior is that I'll create a datapoint and a graph, dislike what I've named it, deleted and created it the way I've wanted with the same OID and it doesn't work
[07-Jan-2010 13:50:50] <theacolyte-> until I delete/readd the device
[07-Jan-2010 13:50:57] <theacolyte-> (it was workng before I did the change)
[07-Jan-2010 13:51:22] <rmatte> hmmm, I've never had that issue
[07-Jan-2010 13:51:58] <rmatte> you do know that you can change the way a datapoint is displayed on a graph without renaming or removing the datasource itself right?
[07-Jan-2010 13:52:13] <theacolyte-> Yeah, I just want it clean
[07-Jan-2010 13:52:14] <rmatte> you just add it to the graph, click on the graphpoint, change the name, and click submit
[07-Jan-2010 13:52:16] <theacolyte-> because this will be a zenpack
[07-Jan-2010 13:52:18] <rmatte> ah, k
[07-Jan-2010 13:52:24] <theacolyte-> it's just cosmetic
[07-Jan-2010 13:52:25] <rmatte> fair enough
[07-Jan-2010 13:53:49] <theacolyte-> you are right to doubt me though hehe
[07-Jan-2010 13:54:02] <rmatte> I always start with the simple stuff and work down from there
[07-Jan-2010 13:54:07] <rmatte> no matter who I'm talking to
[07-Jan-2010 13:54:15] <theacolyte-> yeah absolutely
[07-Jan-2010 13:55:30] <theacolyte-> yup, deleted and now my graphs are showing without the RRD error
[07-Jan-2010 13:55:35] <theacolyte-> So it's a bug of some sort
[07-Jan-2010 13:55:36] <theacolyte-> I'll file one
[07-Jan-2010 13:55:44] <rmatte> k
[07-Jan-2010 13:56:05] <theacolyte-> I got my very own login
[07-Jan-2010 13:56:17] <rmatte> the most annoying thing lately was Zenoss automatically disable my thresholds
[07-Jan-2010 13:56:25] <rmatte> I had to find the code that was doing it and comment it out
[07-Jan-2010 13:56:32] <willwh> hi guys - if there are interfaces under the OS tab I don't want to track; I delete the interfaces I don't want; and then select any I do want to monitor and choose "lock from deletion and update"?
[07-Jan-2010 13:56:52] <rmatte> willwh: don't delete them, just set them to not be monitored
[07-Jan-2010 13:56:58] <willwh> ah
[07-Jan-2010 13:56:58] <willwh> ok
[07-Jan-2010 13:57:00] <rmatte> willwh: then lock them all down
[07-Jan-2010 13:57:02] <willwh> perfect
[07-Jan-2010 13:57:03] <willwh> :)
[07-Jan-2010 13:57:49] <willwh> you guys are fantastic :)
[07-Jan-2010 13:57:54] <rmatte> we try
[07-Jan-2010 13:58:01] <rmatte> ;)
[07-Jan-2010 13:59:05] <willwh> very much appreciate the assistance
[07-Jan-2010 13:59:22] <rmatte> no problem
[07-Jan-2010 13:59:40] <theacolyte-> http://dev.zenoss.org/trac/ticket/6019
[07-Jan-2010 14:00:33] <rmatte> theacolyte-: instead of deleting the device, did you try just clearing the contents of $ZENHOME/perf/Devices/<devicename>?
[07-Jan-2010 14:00:48] <theacolyte-> I didn't, no
[07-Jan-2010 14:00:55] <rmatte> that would probably work too
[07-Jan-2010 14:01:01] <theacolyte-> I didn't think it would make a difference because the file itself was missing
[07-Jan-2010 14:01:04] <rmatte> since that's basically what you're doing by deleting the device
[07-Jan-2010 14:01:07] <theacolyte-> All the other files were there
[07-Jan-2010 14:01:28] <rmatte> it might have, you'd have to see
[07-Jan-2010 14:02:51] <rmatte> oh I just thought of something else
[07-Jan-2010 14:02:59] <rmatte> you might have to go to Manage -> Push Changes
[07-Jan-2010 14:03:14] <rmatte> to push the changes to the collector, otherwise it only sees the old datapoint and not the new one
[07-Jan-2010 14:03:26] <rmatte> it only pushes changes globally every hour
[07-Jan-2010 14:03:30] <theacolyte-> interesting
[07-Jan-2010 14:03:55] <rmatte> you can do it from the device group where the template is
[07-Jan-2010 14:03:58] <rmatte> or from the device page
[07-Jan-2010 14:04:02] <theacolyte-> yeah I se it
[07-Jan-2010 14:04:09] <theacolyte-> I'll give that a try when I see it again (I will)
[07-Jan-2010 14:04:15] <rmatte> hehe
[07-Jan-2010 14:04:21] <theacolyte-> I'm fickle with naming
[07-Jan-2010 14:04:38] <theacolyte-> The last time I just ran into this it was because I didn't capitalize something
[07-Jan-2010 14:04:47] <rmatte> hehe
[07-Jan-2010 14:08:19] <rmatte> mista chippy
[07-Jan-2010 14:19:56] <rmatte> eugh, there are too many damn types of Cisco wireless gear
[07-Jan-2010 14:23:19] <rmatte> almost time to re-design the HP ProCurve template that I made back when I hardly knew anything about Zenoss lol
[07-Jan-2010 14:23:34] <rmatte> it'll be a cakewalk this time around
[07-Jan-2010 14:37:56] <markeriv> alright
[07-Jan-2010 14:37:59] <markeriv> so i fixed my dns issue
[07-Jan-2010 14:38:14] <markeriv> and I am still having problems with reportmail
[07-Jan-2010 14:38:23] <markeriv> now its emailing successfully... but sending a blank email
[07-Jan-2010 14:39:59] <cparlette> what's the full reportmail command you're running?
[07-Jan-2010 14:40:30] <markeriv> bash -lc '$ZENHOME/scripts/emailutilization2mark.sh'
[07-Jan-2010 14:41:01] <rmatte> well, I think he wants what the script is running lol
[07-Jan-2010 14:41:02] <markeriv> here is the part of the report that isnt making it
[07-Jan-2010 14:41:04] <markeriv> --subject="Linkwatch Utilization Report" \
[07-Jan-2010 14:41:05] <markeriv> --url="http://205.234.66.69:8080/zport/dmd/Reports/Graph Reports/Windsor"
[07-Jan-2010 14:41:05] <markeriv> EOF
[07-Jan-2010 14:41:13] <markeriv> will try to copy the entire script
[07-Jan-2010 14:41:22] <rmatte> pastebin.com is your friend
[07-Jan-2010 14:41:26] <markeriv> cat <<EOF > $ZENHOME/scripts/emailutilization2mark.sh
[07-Jan-2010 14:41:26] <rmatte> don't paste it here
[07-Jan-2010 14:41:26] <markeriv> #!/bin/sh
[07-Jan-2010 14:41:26] <markeriv> $ZENHOME/bin/reportmail \
[07-Jan-2010 14:41:26] <markeriv> --user=Mark \
[07-Jan-2010 14:41:26] <markeriv> --passwd=”1h0m3w0rld” \
[07-Jan-2010 14:41:27] <markeriv> --from="linkwatch@telnesbroadband.com" \
[07-Jan-2010 14:41:27] <rmatte> lol
[07-Jan-2010 14:41:29] <markeriv> --address="Markeriv@live.com" \
[07-Jan-2010 14:41:31] <markeriv> --subject="Linkwatch Utilization Report" \
[07-Jan-2010 14:41:31] <Apachez> dont
[07-Jan-2010 14:41:32] * rmatte cries
[07-Jan-2010 14:41:33] <markeriv> --url="http://205.234.66.69:8080/zport/dmd/Reports/Graph Reports/Windsor"
[07-Jan-2010 14:41:33] <Apachez> just dont
[07-Jan-2010 14:41:35] <markeriv> EOF
[07-Jan-2010 14:41:37] <markeriv> doh
[07-Jan-2010 14:41:43] <markeriv> least its a fubar password :P
[07-Jan-2010 14:41:46] * Apachez slaps markeriv using a wooden fork
[07-Jan-2010 14:41:50] <markeriv> ouch
[07-Jan-2010 14:41:59] <rmatte> pastebin.com
[07-Jan-2010 14:42:02] <rmatte> pweez
[07-Jan-2010 14:42:04] <rmatte> :)
[07-Jan-2010 14:42:31] <markeriv> http://pastebin.com/m5488d817
[07-Jan-2010 14:42:43] <markeriv> thats actually really neato :D
[07-Jan-2010 14:42:46] <cparlette> not sure if this would matter using a .sh script, but when putting a reportmail command directly into cron, you have to replace spaces in the URL with '\%20'
[07-Jan-2010 14:43:51] <rmatte> aloha mark
[07-Jan-2010 14:43:55] <markeriv> aloha
[07-Jan-2010 14:44:10] <markeriv> tried that, cparlette, same issue
[07-Jan-2010 14:44:15] <rmatte> I meant mrhinkle, but aloha to you too markeriv
[07-Jan-2010 14:44:19] <rmatte> hehe
[07-Jan-2010 14:44:21] <markeriv> :-P
[07-Jan-2010 14:46:22] <cparlette> does just running reportmail from the CLI with similar flags work?
[07-Jan-2010 14:46:25] <markeriv> question, how do i delete something off of pastebin?
[07-Jan-2010 14:46:44] <rmatte> markeriv: you can't really
[07-Jan-2010 14:46:56] <rmatte> markeriv: by default it deletes it after a month
[07-Jan-2010 14:47:07] <markeriv> joy
[07-Jan-2010 14:47:07] <rmatte> markeriv: you can set it to delete it after 1 day, or to never delete it
[07-Jan-2010 14:47:12] <markeriv> *goes off to change the pw*
[07-Jan-2010 14:47:33] <rmatte> markeriv: search engines don't see those pages though
[07-Jan-2010 14:47:51] <rmatte> markeriv: so the chances of someone actually coming across it are slim to none
[07-Jan-2010 14:47:57] <markeriv> nod
[07-Jan-2010 14:48:03] <markeriv> still, already changed em
[07-Jan-2010 14:48:09] <rmatte> good idea
[07-Jan-2010 14:48:15] <markeriv> anyhow
[07-Jan-2010 14:48:24] * rmatte cancels his hacking attempt
[07-Jan-2010 14:48:24] <markeriv> think it might be the -lc flag on bash?
[07-Jan-2010 14:48:41] <markeriv> sometime i am gonna hafta learn all these cutesy mirc commands :P
[07-Jan-2010 14:48:43] <rmatte> what's the -lc flag do anyways?
[07-Jan-2010 14:48:52] <markeriv> no idea
[07-Jan-2010 14:48:55] <markeriv> its whats in the help doc
[07-Jan-2010 14:48:56] <rmatte> markeriv: IRC commands, mIRC is just the client
[07-Jan-2010 14:49:02] <rmatte> ;)
[07-Jan-2010 14:49:05] <markeriv> :P
[07-Jan-2010 14:49:19] <markeriv> sigh, i dont even know how you put myname: before your response
[07-Jan-2010 14:49:20] <markeriv> lol
[07-Jan-2010 14:49:34] * rmatte like this?
[07-Jan-2010 14:49:40] <markeriv> that
[07-Jan-2010 14:49:41] <rmatte> it's /me
[07-Jan-2010 14:49:47] * markeriv slaps self.
[07-Jan-2010 14:49:50] <rmatte> hehe
[07-Jan-2010 14:49:52] <markeriv> wootah, 1 down
[07-Jan-2010 14:49:59] <markeriv> how about where u make my name show up
[07-Jan-2010 14:50:03] <markeriv> before your response
[07-Jan-2010 14:50:04] <mrhinkle> Hi Ryan
[07-Jan-2010 14:50:27] <rmatte> markeriv: that's built in to my irc client, I just hit tab, not sure if mIRC supports that or not, haven't used mIRC in years.
[07-Jan-2010 14:50:49] <rmatte> well, I start typing your name then hit tab rather and it auto-completes it
[07-Jan-2010 14:50:56] <markeriv> rmatte: ahhh
[07-Jan-2010 14:51:18] <rmatte> I'm using a commandline IRC client called BitchX
[07-Jan-2010 14:51:19] <markeriv> rmatte: I see how its done, fear me and my 1337 non-existant skillz
[07-Jan-2010 14:51:26] <rmatte> hehe
[07-Jan-2010 14:51:58] <rmatte> osiris-1c/bitchx-1.1-final
[07-Jan-2010 14:53:27] <rmatte> but yeh anyways
[07-Jan-2010 14:53:54] <rmatte> markeriv: http://www.ircbeginner.com/ircinfo/ircc-commands.html
[07-Jan-2010 14:54:59] <theacolyte-> Woot.
[07-Jan-2010 14:55:02] <theacolyte-> Graphs are workin!
[07-Jan-2010 14:55:07] <rmatte> theacolyte-: cool
[07-Jan-2010 14:55:14] <theacolyte-> Someone is going to have to write some back end code for this though, and it can't be me
[07-Jan-2010 14:55:18] <theacolyte-> The OID changes
[07-Jan-2010 14:55:27] <rmatte> hehe
[07-Jan-2010 14:55:38] <theacolyte-> for xample
[07-Jan-2010 14:55:43] <rmatte> is it just a single OID that changes?
[07-Jan-2010 14:55:45] <theacolyte-> bytes RX = 1.3.6.1.4.1.12740.2.1.12.1.8.1.168197810 for me
[07-Jan-2010 14:55:55] <theacolyte-> But it's going to be 1.3.6.1.4.1.12740.2.1.12.1.8.1.127398127389071233123 for someone else
[07-Jan-2010 14:56:04] <theacolyte-> Not sure why the did that actually
[07-Jan-2010 14:56:20] <rmatte> ok, but if you snmpwalk 1.3.6.1.4.1.12740.2.1.12.1.8.1 does it always just return 1 line?
[07-Jan-2010 14:56:23] <theacolyte-> You can find it by walking reeally easy, if you wrote a script that did a walk with some regex
[07-Jan-2010 14:56:37] <theacolyte-> let me walk it, sec
[07-Jan-2010 14:56:54] <rmatte> theacolyte-: just whip a bash script together that does it, do you know any bash scripting?
[07-Jan-2010 14:57:04] <theacolyte-> Name/OID: eqlMemberTxData.1.1681978106; Value (Counter64): 136632254231040
[07-Jan-2010 14:57:37] <rmatte> ok, so you're basically looking for: snmpwalk -v1 -c whatever device eqlMemberTxData ?
[07-Jan-2010 14:57:50] <theacolyte-> essentially
[07-Jan-2010 14:58:00] <theacolyte-> I'm not sure how it really works though
[07-Jan-2010 14:58:19] <rmatte> well, if you figure out how it works I could whip something together fairly quickly
[07-Jan-2010 14:58:21] <theacolyte-> It's weird that 1.3.6.1.4.1.12740.2.1.12.1.8.1 works and 1.3.6.1.4.1.12740.2.1.12.1.8.1.168197810 works
[07-Jan-2010 14:58:46] <rmatte> right, but what I need to know is whether or not 1.3.6.1.4.1.12740.2.1.12.1.8.1 can return multiple values on a system
[07-Jan-2010 14:58:50] <rmatte> or whether it's always just 1 line
[07-Jan-2010 14:59:16] <theacolyte-> ah let me actually walk it this time
[07-Jan-2010 14:59:21] <rmatte> k
[07-Jan-2010 14:59:50] <theacolyte-> Doesn't connect actually, sorry
[07-Jan-2010 14:59:54] <theacolyte-> Still learning this nterface for this tool
[07-Jan-2010 14:59:59] <rmatte> lol
[07-Jan-2010 15:00:13] <theacolyte-> walking 1.3.6.1.4.1.12740.2.1.12.1.8.1.168197810 works
[07-Jan-2010 15:00:14] <rmatte> you could just snmpwalk from commandline
[07-Jan-2010 15:00:17] <theacolyte-> walking 1.3.6.1.4.1.12740.2.1.12.1.8.1
[07-Jan-2010 15:00:19] <theacolyte-> does not
[07-Jan-2010 15:00:30] <rmatte> with just plain snmpwalk?
[07-Jan-2010 15:00:35] <theacolyte-> bah
[07-Jan-2010 15:00:39] <theacolyte-> fine, I'll use the real tool :P
[07-Jan-2010 15:00:43] <Fuzzy_T> snmpwalk should traverse the tool
[07-Jan-2010 15:00:48] <Fuzzy_T> ooops tree rather
[07-Jan-2010 15:00:49] <rmatte> what if you snmpwalk for eqlMemberTxData?
[07-Jan-2010 15:01:08] <rmatte> and yes, use the actual snmpwalk command
[07-Jan-2010 15:01:13] <rmatte> since that's what I'd be using for the script
[07-Jan-2010 15:01:33] <theacolyte-> snmpwalk -v 2c -c zenoss 10.0.100.10 eqlMemberTxData
[07-Jan-2010 15:01:33] <theacolyte-> eqlMemberTxData: Unknown Object Identifier (Sub-id not found: (top) -> eqlMemberTxData)
[07-Jan-2010 15:01:44] <rmatte> try 1.3.6.1.4.1.12740.2.1.12.1.8.1
[07-Jan-2010 15:01:54] <theacolyte-> let me check something
[07-Jan-2010 15:02:21] <theacolyte-> see the problem is the name isn't just TxData
[07-Jan-2010 15:02:33] <theacolyte-> its TxData + that OID # I was mentioning above
[07-Jan-2010 15:02:35] <theacolyte-> let me walk that though
[07-Jan-2010 15:02:49] <rmatte> don't walk the name plus the OID
[07-Jan-2010 15:02:52] <theacolyte-> snmpwalk -v 2c -c zenoss 10.0.100.10 1.3.6.1.4.1.12740.2.1.12.1.8.1
[07-Jan-2010 15:02:52] <theacolyte-> SNMPv2-SMI::enterprises.12740.2.1.12.1.8.1.1681978106 = Counter64: 136632287708160
[07-Jan-2010 15:02:56] <theacolyte-> I'm not
[07-Jan-2010 15:03:03] <rmatte> just walk 1.3.6.1.4.1.12740.2.1.12.1.8.1
[07-Jan-2010 15:03:09] <theacolyte-> I did
[07-Jan-2010 15:03:11] <rmatte> ok
[07-Jan-2010 15:03:16] <theacolyte-> that' what you see up there hehe
[07-Jan-2010 15:03:24] <rmatte> so it only returns 1 line, but what I'm wondering is if it always only returns 1 line
[07-Jan-2010 15:03:30] <rmatte> no matter what device you're doing it on
[07-Jan-2010 15:03:37] <rmatte> or if it can possibly generate more
[07-Jan-2010 15:03:43] <theacolyte-> well this should only have 1 line
[07-Jan-2010 15:03:48] <theacolyte-> not sure how it would have more
[07-Jan-2010 15:03:54] * markeriv continues to bash his head into the wall.
[07-Jan-2010 15:03:56] <theacolyte-> that's not the same for some of the other things I'm monitoring though
[07-Jan-2010 15:04:03] <rmatte> well, it represents rx and tx
[07-Jan-2010 15:04:10] <rmatte> what if there were multiple interfaces or something?
[07-Jan-2010 15:04:19] <theacolyte-> That's a total for the entire SAN I'm pulling
[07-Jan-2010 15:04:23] <rmatte> ah
[07-Jan-2010 15:04:43] <theacolyte-> Actually, now that I'm looking at what I'm trying to monitor they would all be st atic across models
[07-Jan-2010 15:04:53] <rmatte> but do you have more than 1 fiber card running to the SAN?
[07-Jan-2010 15:05:05] <rmatte> or more than 1 gig int (depending on how you have it hooked up)
[07-Jan-2010 15:05:06] <theacolyte-> It's iscsi, so it uses the controllers built-in
[07-Jan-2010 15:05:10] <theacolyte-> Yeah
[07-Jan-2010 15:05:11] <theacolyte-> there are 3
[07-Jan-2010 15:05:14] <theacolyte-> (per controller)
[07-Jan-2010 15:05:15] <rmatte> I see
[07-Jan-2010 15:05:22] <theacolyte-> But that does change between models
[07-Jan-2010 15:05:25] <rmatte> and that value is a combination of all 3?
[07-Jan-2010 15:05:25] <theacolyte-> Some have 2
[07-Jan-2010 15:05:27] <theacolyte-> Right
[07-Jan-2010 15:05:33] <rmatte> k
[07-Jan-2010 15:05:34] <theacolyte-> I'm 99.99% sure
[07-Jan-2010 15:05:49] <theacolyte-> because it's MemberTxData the member is the SAN itself
[07-Jan-2010 15:05:58] <rmatte> well, let me cook something up real quick...
[07-Jan-2010 15:06:50] <theacolyte-> rmatte: I still don't know why I use .1.3.6.1.4.1.12740.2.1.12.1.8.1.1681978106 and .1.3.6.1.4.1.12740.2.1.12.1.8.1.1 works as well
[07-Jan-2010 15:07:47] <theacolyte-> If I can just use .1.3.6.1.4.1.12740.2.1.12.1.8.1.1 then that should be constant for all models
[07-Jan-2010 15:07:59] <rmatte> try it, but it might not work
[07-Jan-2010 15:08:04] <theacolyte-> *however* everything I read has to calculate that, including this cacti one I was using
[07-Jan-2010 15:08:13] <theacolyte-> Well it seems to work for me, but I don't know...
[07-Jan-2010 15:08:21] <theacolyte-> Just doesn't make sense both works
[07-Jan-2010 15:10:13] <rmatte> This would work: http://pastebin.com/m1129804b
[07-Jan-2010 15:10:24] <rmatte> If you need it
[07-Jan-2010 15:10:59] <rmatte> you'd run it like: ./script.sh <device> <snmp string>

[07-Jan-2010 15:11:11] <theacolyte-> And does that return the 1681978106 #?
[07-Jan-2010 15:11:21] <rmatte> save it, run it, and see
[07-Jan-2010 15:11:23] <rmatte> it should
[07-Jan-2010 15:11:45] <rmatte> it won't actually return that number, it'll return the proper value for the snmp string under that OID
[07-Jan-2010 15:11:46] <theacolyte-> lesse
[07-Jan-2010 15:12:04] <rmatte> It'll output OK|value=whatever
[07-Jan-2010 15:12:23] <theacolyte-> snmp_build: unknown failuresnmpwalk: Error building ASN.1 representation (Can't build OID for variable)
[07-Jan-2010 15:13:07] <rmatte> hunh
[07-Jan-2010 15:13:27] <rmatte> oh crap
[07-Jan-2010 15:13:29] <rmatte> one second
[07-Jan-2010 15:13:52] <rmatte> http://pastebin.com/m36796b9f
[07-Jan-2010 15:13:58] <rmatte> I somehow pasted something extra before the OID
[07-Jan-2010 15:14:01] <rmatte> that should work
[07-Jan-2010 15:14:28] <theacolyte-> it finds the value yeah
[07-Jan-2010 15:14:33] <rmatte> yeh
[07-Jan-2010 15:14:51] <theacolyte-> I'll file that away for when I figure out how the hell to integrate it
[07-Jan-2010 15:15:07] <rmatte> hehe, read the part of the admin guide regarding zencommand
[07-Jan-2010 15:15:13] <theacolyte-> I really wish you could display a value on a graph summary but not the graph itself
[07-Jan-2010 15:15:14] <rmatte> I think it's section 6.3 if I recall correctly
[07-Jan-2010 15:15:25] <rmatte> that'll explain how to use it
[07-Jan-2010 15:15:39] <theacolyte-> I'm setting aside tomorrow to read the manual front to back
[07-Jan-2010 15:15:45] <rmatte> good idea
[07-Jan-2010 15:16:35] <theacolyte-> rmatte: By the way, this is essentially what I'm recreating, to check out what it's doing
[07-Jan-2010 15:16:38] <theacolyte-> http://docs.cacti.net/usertemplate:host:dellequallogicgroup
[07-Jan-2010 15:16:49] <theacolyte-> No manual editing of OID values is required. Just assign this host template to a Dell Equallogic device and it will find the members.
[07-Jan-2010 15:17:46] <rmatte> I see
[07-Jan-2010 15:18:46] <rmatte> well, you might not be able to get it exactly the same, but you should be able to do most of it
[07-Jan-2010 15:19:45] <theacolyte-> I may just throw this in the wild with some quick instructions how to tailor it
[07-Jan-2010 15:19:49] <theacolyte-> it's notthat hard, even I could do it
[07-Jan-2010 15:19:55] <theacolyte-> just snmpwalk your san, find the big #
[07-Jan-2010 15:20:00] <theacolyte-> replace mine with yours
[07-Jan-2010 15:20:15] <rmatte> you could do that, but the script would be easier hehe
[07-Jan-2010 15:20:34] <rmatte> If you want an example of how you'd integrate the script, grab my Windows SNMP Performance Monitor ZenPack
[07-Jan-2010 15:20:42] <rmatte> and see how I integrated the scripts in to that
[07-Jan-2010 15:21:02] <rmatte> it's honestly not hard
[07-Jan-2010 15:21:09] <theacolyte-> Did you release it?
[07-Jan-2010 15:21:20] <rmatte> yeh, there are 2 versions of it on the ZenPack page
[07-Jan-2010 15:21:24] <rmatte> just grab the simple version
[07-Jan-2010 15:22:12] <rmatte> the version up there isn't even the latest version, I'm polishing it up for another release
[07-Jan-2010 15:22:19] <rmatte> I re-did the graphs to make them look nicer
[07-Jan-2010 15:22:38] <rmatte> and I modified it so that the threshold line is properly shown on the graph
[07-Jan-2010 15:22:56] <rmatte> The graphs look like: http://dmon.org/graphics/zenoss/windowsperf.png
[07-Jan-2010 15:23:49] <theacolyte-> I wonder why I created mine then :P
[07-Jan-2010 15:23:52] <theacolyte-> Mine just uses WMI tho
[07-Jan-2010 15:23:58] <theacolyte-> I wonder if I should even release it
[07-Jan-2010 15:24:13] <rmatte> there's already a windows performance template that uses WMI
[07-Jan-2010 15:24:45] <rmatte> http://community.zenoss.org/docs/DOC-3500
[07-Jan-2010 15:24:46] <theacolyte-> It's.... lacking in detail
[07-Jan-2010 15:24:50] <theacolyte-> That's why I made mine
[07-Jan-2010 15:25:00] <rmatte> well, what did you add?
[07-Jan-2010 15:25:36] <QubeZ> when i test this asterisknew zenpack command against my server it says "host key verfication failed", however running it from command line runs fine: /usr/bin/ssh root@${dev/manageIp} /zenoss/show_core_channels.sh
[07-Jan-2010 15:25:50] <rmatte> mine is lacking in detail too, but it provides the really important stuff, it's pretty intensive to collect the data without snmp informant
[07-Jan-2010 15:25:55] <theacolyte-> http://imgur.com/ef1Hi.png
[07-Jan-2010 15:26:05] <rmatte> I want to convert my scripts to python but I need to figure out how to do the snmp stuff first
[07-Jan-2010 15:26:12] <QubeZ> i run the command line as root, are there permissions issues with the zenoss user?
[07-Jan-2010 15:26:57] <theacolyte-> rmatte: That's rough, I understand it, but units,formatting,etc aren't tweak3ed yet
[07-Jan-2010 15:27:03] <theacolyte-> I'm basically going in all directions right now
[07-Jan-2010 15:27:05] <rmatte> theacolyte-: oh right, you showed me this before
[07-Jan-2010 15:27:16] <theacolyte-> yeah
[07-Jan-2010 15:27:53] <rmatte> the problem with that one is the way you're collecting memory, you'll never be able to threshold on it
[07-Jan-2010 15:28:15] <theacolyte-> Yeah, I never got around to fixing that
[07-Jan-2010 15:28:20] <theacolyte-> IT'S ON MY LIST
[07-Jan-2010 15:28:21] <theacolyte-> oops
[07-Jan-2010 15:28:36] <rmatte> hehe
[07-Jan-2010 15:28:39] <Diddi> is there any documentation with what oids zenoss look at, after a basic installation?
[07-Jan-2010 15:28:39] <theacolyte-> we should merge some of those counters with yours
[07-Jan-2010 15:28:42] <theacolyte-> unless you have them
[07-Jan-2010 15:28:45] <theacolyte-> since you're doing yours ebtter
[07-Jan-2010 15:28:51] <theacolyte-> altough some of those are WMI only
[07-Jan-2010 15:28:53] <rmatte> well, mine uses SNMP though
[07-Jan-2010 15:29:19] <rmatte> my aim for mine is really to make something that keeps people from having to use WMI
[07-Jan-2010 15:29:36] <rmatte> since SNMP is fully supported in Core but WMI isn't
[07-Jan-2010 15:29:44] <rmatte> so SNMP is the logical choice in most cases
[07-Jan-2010 15:29:51] <rmatte> WMI can be used in the odd instance to fill in gaps
[07-Jan-2010 15:30:40] <theacolyte-> Well
[07-Jan-2010 15:30:47] <theacolyte-> You absolutely need it for a lot of windows monitoring
[07-Jan-2010 15:30:50] <theacolyte-> like SQL for instance
[07-Jan-2010 15:30:57] <rmatte> right
[07-Jan-2010 15:30:59] <theacolyte-> And that's what I'm using, but I agree
[07-Jan-2010 15:31:03] <rmatte> which would be an odd instance
[07-Jan-2010 15:31:04] <rmatte> hehe
[07-Jan-2010 15:33:13] <rmatte> afk for a sec
[07-Jan-2010 15:36:57] <rmatte> back
[07-Jan-2010 15:40:34] <rmatte> Diddi: it depends on what group you drop the device it
[07-Jan-2010 15:40:35] <rmatte> in*
[07-Jan-2010 15:41:09] <rmatte> Diddi: It looks at a bunch of standard OIDs for interfaces, routes, etc... and the rest depends on what device templates are applied to the device organizer or to the device itself
[07-Jan-2010 15:42:45] <theacolyte-> just dropped a site :(
[07-Jan-2010 15:43:38] <Diddi> rmatte: ah, yes. but I was thinking if there was a paper, easier to look at than to look at all individual device template (:
[07-Jan-2010 15:43:52] <rmatte> nope
[07-Jan-2010 15:44:05] <Diddi> ah, ok (:
[07-Jan-2010 15:44:13] <rmatte> it's all in Zenoss, you just need to look around
[07-Jan-2010 15:44:22] <Diddi> will do that
[07-Jan-2010 15:44:25] <rmatte> what exactly are you trying to do?
[07-Jan-2010 15:44:34] <Diddi> nothing really, just curious
[07-Jan-2010 15:44:38] <rmatte> ah
[07-Jan-2010 15:52:20] <markeriv> well, slowly making progress
[07-Jan-2010 15:52:40] <markeriv> can get the report... but now the subject line is refusing to over write
[07-Jan-2010 15:53:01] <markeriv> mmm linux... when every step requires troubleshooting :P
[07-Jan-2010 15:54:05] * markeriv continues to bash his head into the wall.
[07-Jan-2010 15:54:29] * markeriv decides the wall is not hard enough and proceeds to move on to his firewall.
[07-Jan-2010 15:54:49] <QubeZ> i have my jmx template set to port 12345 and that is the port configured in my JVM options, however digging deeper into the settings of the heap collector graph, it shows port 11111.. what is this? do I need to change this to 12345?
[07-Jan-2010 16:08:46] <eidolon> hum. so i'm trying to add a solaris box to my zenoss 2.4 install - there's no pulldown for it. am i missing a zenpack?
[07-Jan-2010 16:08:56] <eidolon> (we're primarily linux with a smattering of windows boxen)
[07-Jan-2010 16:09:14] <theacolyte-> &hahahahahaha
[07-Jan-2010 16:09:20] <eidolon> uh oh
[07-Jan-2010 16:09:23] <eidolon> "This can't be good."
[07-Jan-2010 16:09:29] <theacolyte-> I figured out my mapping issue, I probably shouldn't have put the location in the description field.
[07-Jan-2010 16:09:46] <theacolyte-> eidolon: There's a solaris zenpack you have to install, but where are you trying to add the box?
[07-Jan-2010 16:09:58] <eidolon> add device -> easy add
[07-Jan-2010 16:10:41] <theacolyte-> I don't use that inerface honestly
[07-Jan-2010 16:10:47] <theacolyte-> But it's probably set up in your Devices
[07-Jan-2010 16:10:57] <theacolyte-> And that's something you need to do
[07-Jan-2010 16:11:00] <theacolyte-> It's not a zenpack thing
[07-Jan-2010 16:11:53] <eidolon> aha /device/server/solaris is there
[07-Jan-2010 16:12:13] <theacolyte-> I don't see how you can get it in the easy setup stuff
[07-Jan-2010 16:14:22] <eidolon> meh. solaris uses snmpdx?
[07-Jan-2010 16:14:33] <theacolyte-> No idea
[07-Jan-2010 16:16:27] <theacolyte-> Hey do any of you have issues setting base locations for maps?
[07-Jan-2010 16:40:58] <rmatte> theacolyte-: well, what sorts of issues?
[07-Jan-2010 16:41:48] <theacolyte-> Like when you add 2 sublocations to a main location with 2 different freaking states and it shows the whole map :P
[07-Jan-2010 16:41:52] <theacolyte-> (which is my fault)
[07-Jan-2010 16:41:53] <theacolyte-> (again)
[07-Jan-2010 16:41:59] <theacolyte-> I'm seriously my own worst enemy
[07-Jan-2010 16:42:13] <theacolyte-> I'm not qualified to touch much more than a piece of paper
[07-Jan-2010 16:42:42] <theacolyte-> I need to fix the default cisco template I think though, it's not reading subifs right
[07-Jan-2010 16:42:43] <rmatte> hehe
[07-Jan-2010 16:43:08] <rmatte> what do you mean by not reading them right?
[07-Jan-2010 16:43:27] <theacolyte-> well the OS tab interfaces
[07-Jan-2010 16:43:40] <theacolyte-> Error reading value for "FastEthernet0/0.3-802.1Q vLAN subif" on x.x.x.x (oid .1.3.6.1.2.1.2.2.1.20.10 is bad)
[07-Jan-2010 16:43:48] <theacolyte-> It's happening to all my vlan subifs
[07-Jan-2010 16:43:48] <rmatte> that has nothing to do with the actual cisco template
[07-Jan-2010 16:44:21] <rmatte> the reason that's happening is because the ethernetCsmacd template tries to collect error count via snmp, which sub interfaces don't support
[07-Jan-2010 16:44:36] <theacolyte-> ahh
[07-Jan-2010 16:44:38] <rmatte> so you'll have to create a local copy of the template on each interface and remove the error datasources and graph
[07-Jan-2010 16:44:50] <rmatte> (wish there was an easier way, but there isn't)
[07-Jan-2010 16:44:55] <theacolyte-> That sounds like a really big pain in the ass
[07-Jan-2010 16:45:00] <rmatte> yes, yes it is
[07-Jan-2010 16:45:07] <theacolyte-> So you just copy ethernetCsmacd template and make one for each router?
[07-Jan-2010 16:45:14] <theacolyte-> Man
[07-Jan-2010 16:45:16] <rmatte> no
[07-Jan-2010 16:45:19] <rmatte> one for each interface
[07-Jan-2010 16:45:29] <rmatte> that template is an interface template so it applies at the interface level
[07-Jan-2010 16:45:40] <rmatte> unless you don't care about errors on your other interfaces
[07-Jan-2010 16:45:42] <theacolyte-> woah...
[07-Jan-2010 16:45:45] <theacolyte-> I din't realize that
[07-Jan-2010 16:45:50] <rmatte> yup
[07-Jan-2010 16:45:51] <theacolyte-> That's new
[07-Jan-2010 16:46:03] <theacolyte-> There's even a convienient button that allows you to make a local copy
[07-Jan-2010 16:46:21] <theacolyte-> I guess it makes sense there wouldn't be errors, but zenoss should handle that better
[07-Jan-2010 16:47:43] <rmatte> yeh, it should
[07-Jan-2010 16:47:54] <rmatte> I've been meaning to open a Trac ticket about that
[07-Jan-2010 16:47:55] <theacolyte-> seems like a matter of a quick regexp
[07-Jan-2010 16:47:57] <theacolyte-> which is easy for me to say
[07-Jan-2010 16:48:09] <theacolyte-> I'll open one
[07-Jan-2010 16:48:09] <rmatte> nah, it's not
[07-Jan-2010 16:48:33] <rmatte> the interfaces are applied based on interface type, and there's no distinction between the type of the sub interface and the main interface
[07-Jan-2010 16:48:50] <theacolyte-> But couldn't they just look for net/ ?
[07-Jan-2010 16:48:54] <theacolyte-> In the IF name
[07-Jan-2010 16:49:06] <theacolyte-> that would get fastethernet/ethernet/tengigabitethernet
[07-Jan-2010 16:51:35] <theacolyte-> http://dev.zenoss.org/trac/ticket/6020
[07-Jan-2010 16:51:46] <theacolyte-> http://dev.zenoss.org/trac/ticket/6020
[07-Jan-2010 16:51:52] <theacolyte-> oops, was scrolled up
[07-Jan-2010 16:52:38] <rmatte> they could, but that's not the way Zenoss works
[07-Jan-2010 16:52:55] <rmatte> it doesn't look at the name of the interface at all when assigning interface templates
[07-Jan-2010 16:53:12] <theacolyte-> ah
[07-Jan-2010 16:53:25] <theacolyte-> I have 3 open bugs now!
[07-Jan-2010 16:53:36] <theacolyte-> Actuall y2 are enhancements, 1 is a bug
[07-Jan-2010 16:53:46] <Apachez> 3 ?
[07-Jan-2010 16:53:47] <Apachez> me too
[07-Jan-2010 16:53:57] <Apachez> but I need to get hold on mray to get my own trac login :P
[07-Jan-2010 16:54:06] <theacolyte-> I have one!
[07-Jan-2010 16:54:16] <theacolyte-> mray can set one up for you
[07-Jan-2010 16:54:23] <QubeZ> i am trying to load jmxmonitoring with the following flags but my openfire server does not start: ""-Dcom.sun.management.jmxremote.port=12345 -Dcom.sun.management.jmxremote.authenticate=false -Dcom.sun.management.jmxremote.ssl=false" any ideas?
[07-Jan-2010 16:54:33] <QubeZ> oops, wrong channel
[07-Jan-2010 16:57:41] <Apachez> theacolyte-: I know thats why I need to get in touch with mray
[07-Jan-2010 16:57:48] <Apachez> but he didnt show up in here toda :S
[07-Jan-2010 16:58:34] <rmatte> lol, even Matt needs a day off from time to time
[07-Jan-2010 16:58:44] <theacolyte-> lies.
[07-Jan-2010 16:58:55] <theacolyte-> he is a slave to this IRC cchannel!
[07-Jan-2010 16:59:13] <rmatte> lol
[07-Jan-2010 16:59:36] <rmatte> so am I, and I've had enough I tell you! no more cotton picking! freedom!
[07-Jan-2010 16:59:39] * rmatte pulls the plug
[07-Jan-2010 16:59:53] * rmatte looks around and sees that he's still here
[07-Jan-2010 16:59:55] <rmatte> damn!
[07-Jan-2010 16:59:56] <theacolyte-> hahahahaha
[07-Jan-2010 17:00:04] <theacolyte-> If you leave, I may as well throw in the towel!
[07-Jan-2010 17:00:09] <rmatte> lol
[07-Jan-2010 17:00:14] <rmatte> or read the admin guide
[07-Jan-2010 17:00:15] <rmatte> :P
[07-Jan-2010 17:00:20] <Apachez> and now for something completely different http://www.generalhilarity.com/wp-content/uploads/2009/12/belive-me.jpg
[07-Jan-2010 17:00:20] <theacolyte-> I could do that, but that's not fun
[07-Jan-2010 17:00:24] <rmatte> hehe
[07-Jan-2010 17:02:26] <theacolyte-> hah this is interesting
[07-Jan-2010 17:02:51] <theacolyte-> ah nevermind, my ISP did that on purpose
[07-Jan-2010 17:02:57] <rmatte> yeh, someone obviously got a hold of his login details
[07-Jan-2010 17:02:58] <theacolyte-> They set the Serial # of my devices to their name
[07-Jan-2010 17:03:19] <rmatte> I see
[07-Jan-2010 17:03:27] <rmatte> yeh, they generally do do that
[07-Jan-2010 17:03:38] <theacolyte-> chassis-id?
[07-Jan-2010 17:03:39] <theacolyte-> that's lame
[07-Jan-2010 17:03:55] <theacolyte-> I'm going to go through and do serial/asset tag#'s soon anyway
[07-Jan-2010 17:05:34] <theacolyte-> create local copy is handy
[07-Jan-2010 17:07:07] <rmatte> yup
[07-Jan-2010 17:08:06] <theacolyte-> Don't suppose you know how to zoom the map portlet out?
[07-Jan-2010 17:08:10] <theacolyte-> By default
[07-Jan-2010 17:08:31] <theacolyte-> I have 3 sites a few lbocks from eachother and on the map the dots are cut in half on the top and bottom (they're basically in a line)
[07-Jan-2010 17:09:49] <rmatte> nah, it defaults to a certain zoom
[07-Jan-2010 17:10:32] <rmatte> there's no way to override it
[07-Jan-2010 17:10:52] <theacolyte-> well then, that's a bummer!
[07-Jan-2010 17:11:16] <theacolyte-> ah well, not a game changer like the network map not doing essentially the same thing
[07-Jan-2010 17:11:26] <rmatte> hehe
[07-Jan-2010 17:11:51] <theacolyte-> (you can't view subnets, like 192.168.0.0 /14
[07-Jan-2010 17:11:55] <theacolyte-> errr /16
[07-Jan-2010 17:11:59] <rmatte> here's my best map: http://dmon.org/graphics/zenoss/gen2.png
[07-Jan-2010 17:12:20] <theacolyte-> bah maps are only interesting if you have a lot of sites!
[07-Jan-2010 17:12:21] <rmatte> theacolyte-: yeh, that's filed as a bug
[07-Jan-2010 17:12:23] <theacolyte-> I only have 5 (
[07-Jan-2010 17:12:34] <rmatte> hehe
[07-Jan-2010 17:12:36] <theacolyte-> Do you just work on monitoring all day?
[07-Jan-2010 17:12:43] <theacolyte-> seems like that's a full time job
[07-Jan-2010 17:12:47] <rmatte> our part of the company does monitoring, yeh
[07-Jan-2010 17:12:52] <rmatte> and I manage Zenoss all day
[07-Jan-2010 17:12:59] <rmatte> I do other stuff, but this is my primary job
[07-Jan-2010 17:13:02] <theacolyte-> And you haven't slit your wrists yet
[07-Jan-2010 17:13:03] <rmatte> and it keeps me quite busy
[07-Jan-2010 17:13:07] <rmatte> nah, I love it
[07-Jan-2010 17:13:12] <theacolyte-> hehe, so do I
[07-Jan-2010 17:13:21] <theacolyte-> It's miles better than all the other ones I've used
[07-Jan-2010 17:13:31] <rmatte> I'm going to be taking on managing solarwinds as well soon
[07-Jan-2010 17:13:38] <rmatte> (oh yay)
[07-Jan-2010 17:13:40] <theacolyte-> solarwinds is decent, but it's hackish
[07-Jan-2010 17:13:47] <theacolyte-> NPM or APM or both?
[07-Jan-2010 17:13:51] <rmatte> I just don't like that it has to run on windows
[07-Jan-2010 17:13:52] <theacolyte-> Actually ou need NMP for APM
[07-Jan-2010 17:14:20] <rmatte> poor decision on their part to not make it cross-platform
[07-Jan-2010 17:14:34] <theacolyte-> All their stuff is like that
[07-Jan-2010 17:14:42] <theacolyte-> once you get your hands on solarwinds though, you will see why
[07-Jan-2010 17:14:43] <theacolyte-> hehehe
[07-Jan-2010 17:14:49] <theacolyte-> I almost bought it a couple years back
[07-Jan-2010 17:15:42] <theacolyte-> Their network monitoring is nice
[07-Jan-2010 17:15:45] <theacolyte-> The application side isn't
[07-Jan-2010 17:15:50] <rmatte> well, we didn't really want it, but now we're stuck with it
[07-Jan-2010 17:15:54] <theacolyte-> (though they came out with a new version that's probably better)
[07-Jan-2010 17:15:58] <rmatte> we wanted EIQ
[07-Jan-2010 17:16:18] <theacolyte-> Haven't seen/used it
[07-Jan-2010 17:16:33] <rmatte> http://www.eiqnetworks.com/
[07-Jan-2010 17:16:34] <theacolyte-> I've used netcool, hyperic, WUG, and now zenoss
[07-Jan-2010 17:16:47] <RobertLaptop> I am at close to 30 sites My biggest issue is I have over 10 closely packet together in the northeast
[07-Jan-2010 17:16:48] <rmatte> how do you rank Zenoss in comparison to Hyperic?
[07-Jan-2010 17:16:55] <theacolyte-> hehe, honestly? :)
[07-Jan-2010 17:17:01] <rmatte> yeh
[07-Jan-2010 17:17:05] <rmatte> honestly
[07-Jan-2010 17:17:15] <theacolyte-> I like hyperic a lot for some things... it's great with windows, it has templates and much easier to work with WMI
[07-Jan-2010 17:17:18] <theacolyte-> MUCH easier.
[07-Jan-2010 17:17:29] <theacolyte-> However, extending it is much harder than zenoss - you can't but it's more difficult
[07-Jan-2010 17:17:37] <rmatte> ah
[07-Jan-2010 17:17:43] <theacolyte-> Like I have a problem monitoring my exchange environment
[07-Jan-2010 17:17:54] <theacolyte-> They provide the template, but I've had a forum posting for 2 years now
[07-Jan-2010 17:17:58] <theacolyte-> Nothing
[07-Jan-2010 17:18:07] <rmatte> ah
[07-Jan-2010 17:18:10] <theacolyte-> They did just get bought by springsource then vmware, so I bet you that's going to change
[07-Jan-2010 17:18:21] <rmatte> probably
[07-Jan-2010 17:18:22] <theacolyte-> Also the compan ywas basically unwilling to talk to you if you didn't have a contract
[07-Jan-2010 17:18:36] <rmatte> Yeh, Zenoss is so not like that
[07-Jan-2010 17:18:36] <theacolyte-> They are right down the street from me (literally)
[07-Jan-2010 17:18:44] <theacolyte-> right, that's actually why I've stuck with it
[07-Jan-2010 17:18:54] <rmatte> they are great with the community
[07-Jan-2010 17:18:55] <theacolyte-> bnetween you, mray, and the rest of the channel it's been great
[07-Jan-2010 17:19:11] <rmatte> hehe
[07-Jan-2010 17:19:32] <RobertLaptop> I have used JFFnms and opennms. Both jffnms is only really good for small sites. opennms seems really limited.
[07-Jan-2010 17:19:43] <theacolyte-> Oh, I've used opennms as well
[07-Jan-2010 17:19:51] <theacolyte-> I liked opennms
[07-Jan-2010 17:19:59] <rmatte> yeh, I looked at opennms a bit before
[07-Jan-2010 17:19:59] <theacolyte-> But I couldn't figure it out
[07-Jan-2010 17:20:11] <theacolyte-> Zenoss REALLY seems to shine on creating custom templates
[07-Jan-2010 17:20:16] <theacolyte-> It's actually really easy
[07-Jan-2010 17:20:23] <rmatte> yeh, that's it's major strength
[07-Jan-2010 17:20:23] <theacolyte-> Despite how I make it look LOL
[07-Jan-2010 17:20:26] <rmatte> which is why I love it
[07-Jan-2010 17:20:42] <theacolyte-> I have barely scratched the surface too
[07-Jan-2010 17:20:48] <rmatte> if a client comes to us and needs some weird obscure thing monitored to certain specifications we don't have to say "sorry, no can do", we say "give us 2 days"
[07-Jan-2010 17:20:48] <rmatte> lol
[07-Jan-2010 17:20:57] <theacolyte-> I still need to add a ton of devices, templates, web monitoring, ssh monitoring, etc
[07-Jan-2010 17:21:02] <theacolyte-> right, exactly
[07-Jan-2010 17:21:44] <rmatte> well, it's easy, and it's not, there's little tricks and stuff that you pick up as you go along which really aren't documented anywhere, or not documented well
[07-Jan-2010 17:21:53] <RobertLaptop> Yea it does. Opennms was much easier for basic setup but seems really hard to expend
[07-Jan-2010 17:22:02] <rmatte> but that's what I'm in here for, to pass the knowledge along
[07-Jan-2010 17:22:07] <rmatte> since I know all the little tricks now
[07-Jan-2010 17:22:07] <rmatte> lol
[07-Jan-2010 17:22:36] <theacolyte-> Yeah, I do the same in all my channels
[07-Jan-2010 17:22:45] <rmatte> yeh, we've done some much to extend zenoss including integration with our ticketing system and with the custom CMDB that we'll be rolling out at the end of next week.
[07-Jan-2010 17:22:49] <theacolyte-> a year ago I didn't know anything about vmware products beyond desktop stuff, now I'm certified
[07-Jan-2010 17:22:52] <rmatte> s/some/so
[07-Jan-2010 17:22:57] <theacolyte-> what do you use for ticketing again?
[07-Jan-2010 17:23:03] <rmatte> OTRS
[07-Jan-2010 17:23:06] <rmatte> but heavily modified
[07-Jan-2010 17:23:12] <theacolyte-> ah I hope so
[07-Jan-2010 17:23:15] <theacolyte-> Not a fan of OTRS
[07-Jan-2010 17:23:16] <RobertLaptop> Yea. Actually that reminds I have a question. I was rereading section 6 in the admin manually learning more about zencommands.
[07-Jan-2010 17:23:19] <theacolyte-> I use kayako personally
[07-Jan-2010 17:23:39] <rmatte> never seen kayako
[07-Jan-2010 17:23:47] <theacolyte-> It's ridiculously featured
[07-Jan-2010 17:23:50] <rmatte> oh, you have to pay for it?
[07-Jan-2010 17:23:51] <theacolyte-> www.kayako.com
[07-Jan-2010 17:23:55] <theacolyte-> yeah, but it's cheep
[07-Jan-2010 17:24:02] <theacolyte-> like doesn't register in cost compared to say trackit
[07-Jan-2010 17:24:03] <RobertLaptop> Cerb4 fan myself.
[07-Jan-2010 17:24:05] <theacolyte-> or remedy
[07-Jan-2010 17:24:11] <theacolyte-> $250
[07-Jan-2010 17:24:47] <theacolyte-> The new version of Kayako will have an API
[07-Jan-2010 17:24:55] <theacolyte-> http://www.kayako.com/
[07-Jan-2010 17:25:05] <rmatte> $250/year or one time?
[07-Jan-2010 17:25:08] <theacolyte-> check it out - it's freaking _amazing_ and I've done a fair amount of dilligence looking for better
[07-Jan-2010 17:25:11] <theacolyte-> perpetual
[07-Jan-2010 17:25:16] <rmatte> ah
[07-Jan-2010 17:25:18] <theacolyte-> support is a yearly fee for upgrades though, but it's nominal
[07-Jan-2010 17:25:36] <theacolyte-> Tons of companies use it, so the forums are decent
[07-Jan-2010 17:25:45] <theacolyte-> support is awful, though improving from how it used to be
[07-Jan-2010 17:25:52] <rmatte> yeh, but it seems to be geared more towards standard helpdesk than towards a NOC environment
[07-Jan-2010 17:25:56] <rmatte> we'd have to hack the hell out of it
[07-Jan-2010 17:26:00] <theacolyte-> It is yeah
[07-Jan-2010 17:26:09] <theacolyte-> It actually started for webhosts
[07-Jan-2010 17:26:16] <theacolyte-> (which is how I found it... 5 years ago now)
[07-Jan-2010 17:26:17] <rmatte> honestly, OTRS is very similar, and it's completely free
[07-Jan-2010 17:26:31] <RobertLaptop> Under sections 6.5 in the admin manual it talks about an ssh agent is that old data.
[07-Jan-2010 17:26:47] <theacolyte-> It has a ton more features, and isn't totally ugly and offputting to users though
[07-Jan-2010 17:26:49] <rmatte> RobertLaptop: yeh, you don't need the agent anymore
[07-Jan-2010 17:26:54] <RobertLaptop> ok.
[07-Jan-2010 17:27:10] <theacolyte-> My main beef with OTRS was I know jack about programming (as you know now)
[07-Jan-2010 17:27:15] <rmatte> theacolyte-: well, part of what we did was pretty OTRS up on the user end of things hehe
[07-Jan-2010 17:27:20] <theacolyte-> I couldn't customize myself out of a paper bag
[07-Jan-2010 17:27:30] <theacolyte-> Unless there's a pretty checkbox
[07-Jan-2010 17:27:36] <RobertLaptop> Well OTRS is in Perl and the iterface is ugly in my view.
[07-Jan-2010 17:27:55] <theacolyte-> s/ugly/heinous/g
[07-Jan-2010 17:27:59] <theacolyte-> hehe
[07-Jan-2010 17:28:02] <theacolyte-> sorry I couldn't resist it
[07-Jan-2010 17:28:04] <rmatte> let me grab a screenshot and show you how we changed it
[07-Jan-2010 17:28:23] <theacolyte-> I'd love to see it
[07-Jan-2010 17:28:28] <rmatte> just need to sensor out all the device names and stuff lol
[07-Jan-2010 17:29:09] <theacolyte-> I accept bank account numbers and social security #'s
[07-Jan-2010 17:29:26] <RobertLaptop> I personally use Cerb4.com but I have done a lot of hacking of the code to add stuff in.
[07-Jan-2010 17:31:50] <kisielk> hm, it seems my XML-RPC has broken :s
[07-Jan-2010 17:32:14] <kisielk> using zport/dmd/Devices/search now seems to give an error
[07-Jan-2010 17:32:15] <kisielk> d
[07-Jan-2010 17:32:23] <kisielk> id something change in 2.5 WRT to that functionality?
[07-Jan-2010 17:33:02] <rmatte> http://dmon.org/graphics/otrs.png
[07-Jan-2010 17:33:11] <rmatte> that's how we have it so far, but we're constantly improving it
[07-Jan-2010 17:33:17] <rmatte> that's just a view of the ticket queue
[07-Jan-2010 17:34:08] <rmatte> the colours of the cells represent severity
[07-Jan-2010 17:34:20] <rmatte> which is automatically sent when the tickets are generated from Zenoss
[07-Jan-2010 17:34:27] <rmatte> s/sent/set
[07-Jan-2010 17:34:43] <kisielk> rmatte: better start replacing those batteries!
[07-Jan-2010 17:34:51] <rmatte> hehe
[07-Jan-2010 17:35:03] <rmatte> that's for one of their clients, not us :P
[07-Jan-2010 17:35:18] <rmatte> and that's literally what the trap comes in as
[07-Jan-2010 17:35:32] <RobertLaptop> Once we cleanup and get zenoss running smothly we will be doing the same thing.
[07-Jan-2010 17:35:59] <RobertLaptop> For some reason dmog.org never comes up for me.
[07-Jan-2010 17:36:07] <kisielk> RobertLaptop: that's what I said when I installed Zenoss a year ago, still hasn't happened :/
[07-Jan-2010 17:36:09] <rmatte> dmon.org you mean?
[07-Jan-2010 17:36:29] <rmatte> do an nslookup on it, see if it resolves
[07-Jan-2010 17:36:42] <theacolyte-> Would you expect that in a zenpack or monitoring in Zenoss that you would monitor more than just counters? Stuff like RAID status for example?
[07-Jan-2010 17:36:52] <theacolyte-> which is an ok/bad/whatever
[07-Jan-2010 17:37:21] <rmatte> RobertLaptop: here - http://tinypic.com/view.php?pic=soobc6&s=6
[07-Jan-2010 17:37:46] <rmatte> theacolyte-: raid status is done via traps
[07-Jan-2010 17:37:58] <rmatte> theacolyte-: generally anyways
[07-Jan-2010 17:38:06] <theacolyte-> It's worth noting I do trust my SAN with traps for that
[07-Jan-2010 17:38:10] <theacolyte-> (I also get emails)
[07-Jan-2010 17:38:16] <theacolyte-> I guess I'm asking in a general POV for zenpacks
[07-Jan-2010 17:38:49] <rmatte> ZenPacks are whatever you want them to be, but generally I'm more interested in performance info from a ZenPack
[07-Jan-2010 17:38:55] <theacolyte-> Yeah
[07-Jan-2010 17:38:59] <theacolyte-> It's more of a general question I guess
[07-Jan-2010 17:39:03] <theacolyte-> But that's fair
[07-Jan-2010 17:40:25] <RobertLaptop> Yea. I need to figuire out how to prcess trap's they all come in as numbers.
[07-Jan-2010 17:40:54] <rmatte> you need to figure out what the numbers means and write an event transform for each trap
[07-Jan-2010 17:41:58] <theacolyte-> remind me the RPN for doing stuff like totals on a duplicate graph point?
[07-Jan-2010 17:42:11] <theacolyte-> I'll write it down this time :P
[07-Jan-2010 17:43:05] <RobertLaptop> I though I installed the MiB for my device but the emails still reference the numbers.
[07-Jan-2010 17:47:12] <rmatte> like adding 2 graphpoints?
[07-Jan-2010 17:47:23] <theacolyte-> yeah
[07-Jan-2010 17:47:42] <rmatte> well, the current graphpoint is automatically at the start
[07-Jan-2010 17:47:50] <rmatte> so you need: othergraphpoint,+
[07-Jan-2010 17:48:02] <rmatte> the end thing would look like currentgraphpoint,othergraphpoint,+
[07-Jan-2010 17:48:14] <rmatte> which is essentially currentgraphpoint + othergraphpoint
[07-Jan-2010 17:48:14] <theacolyte-> So you actually put in the names for the graph pointS?
[07-Jan-2010 17:48:17] <theacolyte-> yeah
[07-Jan-2010 17:48:27] <rmatte> you'll see a list of available graphpoints
[07-Jan-2010 17:48:35] <rmatte> some will say stuff like -raw after them
[07-Jan-2010 17:48:44] <rmatte> don't use those, just use the standard graphpoint name
[07-Jan-2010 17:49:02] <rmatte> make sure the graphpoint that you're doing the RPN on is lower in the sequence than the one that you're adding to it
[07-Jan-2010 17:49:32] * theacolyte- does a test
[07-Jan-2010 17:50:05] <theacolyte-> ah I see it
[07-Jan-2010 17:50:05] <rmatte> the idea behind RPN is that it eliminates brackets
[07-Jan-2010 17:50:32] <theacolyte-> borked image
[07-Jan-2010 17:50:39] <rmatte> so instead of doing (13 - 2) * 43 you'd do 13,2,-,43*
[07-Jan-2010 17:50:50] <rmatte> erm
[07-Jan-2010 17:50:53] <theacolyte-> it's polish
[07-Jan-2010 17:50:55] <theacolyte-> hehe
[07-Jan-2010 17:50:59] <rmatte> 13,2,-,43,*
[07-Jan-2010 17:51:00] <rmatte> yeh
[07-Jan-2010 17:51:05] <theacolyte-> I created a duplicate of my last graph point
[07-Jan-2010 17:51:10] <theacolyte-> IO_Read,IO_Write,+
[07-Jan-2010 17:51:17] <theacolyte-> in RPN
[07-Jan-2010 17:51:28] <rmatte> you basically put the numbers that you want to do the operation on, then the operation after it
[07-Jan-2010 17:51:58] <rmatte> then the next operation that you want to work with, which would be the value that you came out with in the first operation along with the next number
[07-Jan-2010 17:52:14] <rmatte> then the operation to be used on those
[07-Jan-2010 17:53:10] <theacolyte-> Yeah I did what I put above and the image is broken, alas
[07-Jan-2010 17:53:27] <rmatte> ((23 + 6) - 5) * 48 == 23,6,+,5,-,48,*
[07-Jan-2010 17:54:01] <rmatte> anyways, it's late and I need to get out of here
[07-Jan-2010 17:54:02] <rmatte> ttyl
[07-Jan-2010 17:54:07] <theacolyte-> alright man, take it easy
[07-Jan-2010 17:54:12] <rmatte> you too
[07-Jan-2010 18:29:32] <theacolyte-> nice, jsut finished my equallogic template OID's so now it's time to make them pretty
[07-Jan-2010 18:30:55] <RobertLaptop> That is cool. Want to do some templates for me :)
[07-Jan-2010 18:31:10] <theacolyte-> $150/hr, sure :P
[07-Jan-2010 19:18:31] <willwh> anyone hosting any hardware with rapidswitch?
[07-Jan-2010 19:18:35] <willwh> they just died it looks like
[07-Jan-2010 19:20:22] <theacolyte-> webhostingtalk is usually a great resource for outages I find
[08-Jan-2010 00:00:30] [disconnected at Fri Jan 8 00:00:30 2010]
[08-Jan-2010 00:00:30] [connected at Fri Jan 8 00:00:30 2010]
[08-Jan-2010 02:07:24] <tehhobbit> How do I go about igonring certain syslog messages ?
[08-Jan-2010 02:29:19] <tehhobbit> morning Apachez
[08-Jan-2010 03:14:58] <Troubadix09> Hi all
[08-Jan-2010 03:21:20] <Apachez> morrn
[08-Jan-2010 04:16:39] <Apachez> regarding tuning of zope: http://pypi.python.org/pypi/jarn.checkinterval/1.0
[08-Jan-2010 04:20:45] <moto-moi> Is there a way I can use two classes for a device?
[08-Jan-2010 05:00:15] <Apachez> is it only mray who can setup trac accounts?
[08-Jan-2010 08:05:48] <rmatte> moto-moi: you can't use 2 classes for a device, why would you want to?
[08-Jan-2010 08:07:19] <moto-moi> rmatte: Well, I have included several zenpacks, and some of them are put in /devices/ssh/linux and some in /devices/linux , and I wish to use both :P
[08-Jan-2010 08:09:09] <rmatte> then copy the templates and whatever else over to a third class and use that
[08-Jan-2010 08:09:24] <moto-moi> hm, ok
[08-Jan-2010 08:09:27] <moto-moi> I'll try that :)
[08-Jan-2010 08:09:29] <moto-moi> thanks :)
[08-Jan-2010 08:09:44] <rmatte> but you do understand that /server/linux is designed for snmp and /server/ssh/linux is designed for ssh right?
[08-Jan-2010 08:09:55] <rmatte> some of those plugins and templates are not going to jive well
[08-Jan-2010 08:12:33] <rmatte> tehhobbit: you were asking about ignoring syslog messages, you still around?
[08-Jan-2010 08:46:16] <Apachez> !seen mray
[08-Jan-2010 08:46:19] <Apachez> :S
[08-Jan-2010 08:47:02] <twm1010> morning gents
[08-Jan-2010 08:48:22] <rmatte> morning
[08-Jan-2010 08:50:16] <RobertLaptop> Morning
[08-Jan-2010 08:52:37] <RobertLaptop> The question man is back. Since section 6.5 is out of date in the admin manual does anyone have any info or adding commands on the remote side of an ssh device aka /device/ssh/linux ? I want to graphic some more things.
[08-Jan-2010 09:12:53] <tehhobbit> rmatte: yeah but I found it
[08-Jan-2010 09:12:58] <tehhobbit> mapped them to ignore
[08-Jan-2010 09:13:35] <rmatte> cool
[08-Jan-2010 09:33:19] <RobertLaptop> rmatte yes I do know the diff bettwen /device/linux (snmp) and /device/ssh/linux (ssh) why I listed that so there wouldn't be a question. :) I use the core ssh package and found that some of the other packages for ssh don't work so I wanted to create my own.
[08-Jan-2010 09:34:11] <rmatte> I see
[08-Jan-2010 09:35:35] <RobertLaptop> You so well with running local scripts that I was hoping you had some insite into ssh. :)
[08-Jan-2010 09:38:02] <RobertLaptop> rmatte ^^ You were so helpful with running local scripts...
[08-Jan-2010 09:38:50] <rmatte> I don't do ssh monitoring at all
[08-Jan-2010 09:38:53] <rmatte> all snmp and some WMI
[08-Jan-2010 09:39:53] <RobertLaptop> I use those mostly myself. Lower overhead. But this box is only ssh.
[08-Jan-2010 09:40:09] <rmatte> what are you looking to gather that Zenoss isn't already providing?
[08-Jan-2010 09:45:19] <RobertLaptop> The only preformance metric it is displaying is CPU
[08-Jan-2010 10:06:47] <rmatte> then you don't have the right collector plugins
[08-Jan-2010 10:06:57] <rmatte> it should do CPU, memory, hard disk, and interfaces
[08-Jan-2010 10:07:18] <rmatte> those are the things available via ssh, you can also have it check processes I believe
[08-Jan-2010 10:07:39] <rmatte> it also monitors swap I believe, but I'm not 100% sure
[08-Jan-2010 10:08:37] <RobertLaptop> I just uninstalled the adv ssh plugin I though turning the template off would fix the issue but didn't follow up but it seems to be still causing an issue.
[08-Jan-2010 10:12:22] <rmatte> there's a bug with the advanced ssh pack in 2.5.x
[08-Jan-2010 10:12:43] <rmatte> There's an issue where the plugins don't get properly bound and you can't bind them from the UI without making some modification
[08-Jan-2010 10:12:50] <rmatte> there was a forum posting about it, but I forget where
[08-Jan-2010 10:12:57] <rmatte> mrayzenoss knows about it, if you see him around
[08-Jan-2010 10:13:04] <gwb235> if you're not receiving an email alert (setup in alerting rules) what's the first process/logfile to look at to debug?
[08-Jan-2010 10:13:24] <rmatte> gwb235: the logs on your mailserver
[08-Jan-2010 10:13:31] <rmatte> to see if it's actually hitting it
[08-Jan-2010 10:13:48] <rmatte> in a lot of cases mailservers will receive the request to send the email but will block it
[08-Jan-2010 10:13:54] <rmatte> so you'll have to add a rule to allow it
[08-Jan-2010 10:14:30] <rmatte> you need to check the sending mailserver, and if that's good you need to check the receiving mailserver
[08-Jan-2010 10:14:58] <RobertLaptop> User-supplied Python expression (here.threshold) for minimum value caused error: ['rpm_rpm']?
[08-Jan-2010 10:15:38] <RobertLaptop> Do I just need to add a zproperty?
[08-Jan-2010 10:15:49] <rmatte> well, what is here.threshold?
[08-Jan-2010 10:15:59] <rmatte> is that just something you made up?
[08-Jan-2010 10:16:20] <coofamani> here.OMGPONIES!
[08-Jan-2010 10:16:24] <rmatte> lol
[08-Jan-2010 10:16:25] <rmatte> MOAR!
[08-Jan-2010 10:16:39] <RobertLaptop> It looks like it is part of the ZenPack Dell.
[08-Jan-2010 10:17:09] <rmatte> check the message for that event in the event details
[08-Jan-2010 10:17:21] <rmatte> it should say something like /blah/blah/blah has been disabled
[08-Jan-2010 10:17:29] <cparlette> gwb235: also check zenactions.log, thats where you'll see if an alerting email was sent (or failed)
[08-Jan-2010 10:17:33] <rmatte> (it'll automatically disable the threshold that it's having issues with)
[08-Jan-2010 10:18:45] <RobertLaptop> It looks like it is tring to read something in to decide a thresold that wasn't sent but the zenpack installer.
[08-Jan-2010 10:19:09] <rmatte> I see
[08-Jan-2010 10:19:33] <gwb2352> hmmm.... zenactions log hasn't been updated since 12/29. which daemon feeds that file?
[08-Jan-2010 10:20:00] <cparlette> restart zenactions (do a full restart, involving stopping it, check for running processes, then start it)
[08-Jan-2010 10:20:10] <cparlette> you'll probably get a flood of emails when it starts
[08-Jan-2010 10:20:23] <rmatte> hehe
[08-Jan-2010 10:20:48] <rmatte> zenactions stop
[08-Jan-2010 10:20:52] <rmatte> zenactions status
[08-Jan-2010 10:20:55] <rmatte> make sure it's stopped
[08-Jan-2010 10:20:58] <rmatte> zenactions start
[08-Jan-2010 10:22:00] <Apachez> which service is using zope.conf?
[08-Jan-2010 10:22:17] <Apachez> and by the way, is there a faq lists of which ports (and why) zenoss uses ?
[08-Jan-2010 10:22:23] <Apachez> like whats on 8079 and 8081 ?
[08-Jan-2010 10:24:08] <gwb2352> holy load average of 280 batman (upon zenactions restart)
[08-Jan-2010 10:24:42] <rmatte> you mean 2.80?
[08-Jan-2010 10:24:53] <gwb2352> load average: 300.02, 127.60, 48.19
[08-Jan-2010 10:24:58] <rmatte> holy crap
[08-Jan-2010 10:25:05] <rmatte> I haven't seen a load average that high in a long time
[08-Jan-2010 10:25:11] <theacolyte-> rmatte: what cmdb do you use?
[08-Jan-2010 10:25:13] <gwb2352> tons of deadlocked python processes (upon zenactions restart)
[08-Jan-2010 10:26:33] <rmatte> theacolyte-: fully custom, built from the ground up, we're actually rolling it out this upcoming week
[08-Jan-2010 10:26:45] <theacolyte-> I'll take a copy
[08-Jan-2010 10:26:46] <theacolyte-> hehe
[08-Jan-2010 10:26:50] <rmatte> hehe, no can do :P
[08-Jan-2010 10:26:52] <theacolyte-> I need to find something
[08-Jan-2010 10:27:02] <theacolyte-> I haven't looked at all though, just need to
[08-Jan-2010 10:27:09] <rmatte> we looked around but there really wasn't anything open source available that fit our needs
[08-Jan-2010 10:27:22] <rmatte> oh wait, I remember someone mentioning something that looked promising, let me see if I can dig that up
[08-Jan-2010 10:27:50] <RobertLaptop> why would a template have in min for a threshold " here.threshold "?
[08-Jan-2010 10:28:05] <rmatte> RobertLaptop: no idea, that's pretty dumb
[08-Jan-2010 10:28:23] <rmatte> unless they added some custom property or zproperty called "threshold"
[08-Jan-2010 10:28:28] <rmatte> which is horrible horrible practice
[08-Jan-2010 10:28:42] <RobertLaptop> They didn't
[08-Jan-2010 10:28:47] <rmatte> since it should be called cWhatever or zWhatever
[08-Jan-2010 10:28:51] <rmatte> ok, then I have no clue
[08-Jan-2010 10:29:02] <rmatte> unless it's somehow provided by the collector plugin
[08-Jan-2010 10:29:15] <rmatte> (if there is one included with the pack)
[08-Jan-2010 10:29:32] <RobertLaptop> Not sure.
[08-Jan-2010 10:29:47] <rmatte> even then, I would have hoped that they could pick a name that's more distinctive than just "threshold"
[08-Jan-2010 10:29:54] <rmatte> what if another pack decides to use the same thing?
[08-Jan-2010 10:30:23] <RobertLaptop> It looks like it simple extension the standard snmp plug-in but do understand collectors really well yet.
[08-Jan-2010 10:30:30] <gwb2352> the people in #lopsa would probably have some guidance irc.lopsa.org re: CMDB
[08-Jan-2010 10:31:10] <rmatte> theacolyte-: hmmm, there was this french developed open source cmdb that someone mentioned in here ages ago, trying to remember what it was called
[08-Jan-2010 10:31:10] <theacolyte-> ahhh lopsa
[08-Jan-2010 10:31:16] <theacolyte-> been awhile since I've heard that name
[08-Jan-2010 10:31:45] <gwb2352> they're pretty active bunch
[08-Jan-2010 10:32:02] <gwb2352> 149 people in the channel
[08-Jan-2010 10:33:10] <theacolyte-> I haven't even looked into any options yet, so I'll wait until I know more about the field before playing ball o nthat one
[08-Jan-2010 10:33:15] <theacolyte-> Thanks though, I'll write that down
[08-Jan-2010 10:33:30] <theacolyte-> I'm in the middle of trying to find nice colors for my graphs
[08-Jan-2010 10:33:31] <sciolist> theacolyte-: I think rmatte is referring to GLPI
[08-Jan-2010 10:34:26] <sciolist> and.or OCS 9which uses GLPI in the background IIRC)
[08-Jan-2010 10:34:38] <sciolist> s/9/\(/
[08-Jan-2010 10:35:11] <theacolyte-> I found glpi, what's OCS?
[08-Jan-2010 10:36:16] <theacolyte-> Ah there it is I think
[08-Jan-2010 10:36:23] <theacolyte-> OCS inventory
[08-Jan-2010 10:36:34] <sciolist> yep
[08-Jan-2010 10:36:37] <theacolyte-> Only thing I didn't like about OCS is it requires an agent
[08-Jan-2010 10:36:48] <sciolist> they both do, actually
[08-Jan-2010 10:36:53] <theacolyte-> ah, damn
[08-Jan-2010 10:37:04] <theacolyte-> All that stuff is available via WMI, not sure why they don't use it
[08-Jan-2010 10:37:14] <sciolist> yeah, my thinking was "What's wrong with just using SNMP?"
[08-Jan-2010 10:37:28] <sciolist> well, on UNIX-like thingies
[08-Jan-2010 10:37:41] <theacolyte-> Agent is just an issue because I have no deployment software (or undeployment if it sucks)
[08-Jan-2010 10:37:55] <theacolyte-> It's on our radar, we just haven't gotten there yet
[08-Jan-2010 10:38:01] <theacolyte-> 320 employees
[08-Jan-2010 10:40:25] <theacolyte-> rmatte: I'm using colourlovers.com for palette's for my graphs
[08-Jan-2010 10:40:26] <theacolyte-> hehe
[08-Jan-2010 10:41:15] <theacolyte-> http://imgur.com/n49mK.png
[08-Jan-2010 10:41:32] <rmatte> cool
[08-Jan-2010 10:42:30] <theacolyte-> I really, really wish you could display a value on a graph without graphing it
[08-Jan-2010 10:42:35] <theacolyte-> I'm putting in a feature request
[08-Jan-2010 10:42:42] <theacolyte-> A good example would be atotal
[08-Jan-2010 10:43:01] <rmatte> yeh, I agree
[08-Jan-2010 10:43:30] <rmatte> like when I graph a percentage on a graph that's displaying bytes, the percentage is obviously not going to scale, but I don't want a separate graph just to display it
[08-Jan-2010 10:43:38] <theacolyte-> yeah...
[08-Jan-2010 10:43:41] <theacolyte-> that's another really good one
[08-Jan-2010 10:43:49] <theacolyte-> (that I've also run into)
[08-Jan-2010 10:44:30] <rmatte> yeh
[08-Jan-2010 10:45:06] <rmatte> well... they have an option not to draw a line, which I assumed would do what you'd think it would do
[08-Jan-2010 10:45:12] <rmatte> but it actually removes the whole point from the legend
[08-Jan-2010 10:45:20] <theacolyte-> lol
[08-Jan-2010 10:45:21] <rmatte> I don't see what point there is to even having the option then
[08-Jan-2010 10:46:26] <rmatte> eugh, if I see one more command timed out message on this box I'm going to scream
[08-Jan-2010 10:47:20] <theacolyte-> http://dev.zenoss.org/trac/ticket/6023
[08-Jan-2010 10:52:55] <venturaville> does anyone know of any good examples of modeler plugin which does not use SNMP (i.e. I am using an alternate modeler method through a Python API)?
[08-Jan-2010 10:54:35] <gwb2352> having had to restart the hub due to high load average caused by restarting zenactions, all collectors are reporting hearbeat failures for zen* daemons. Do I / should I restart the collectors as well?
[08-Jan-2010 11:03:33] <theacolyte-> Man
[08-Jan-2010 11:03:39] <theacolyte-> What's the trick with graph colors
[08-Jan-2010 11:03:49] <theacolyte-> This transparency/alpha blend is killing me
[08-Jan-2010 11:04:30] <Apachez> when using rrd
[08-Jan-2010 11:04:36] <theacolyte-> Yeah
[08-Jan-2010 11:04:40] <theacolyte-> I'm doing area graphs
[08-Jan-2010 11:04:52] <Apachez> well I hit enter too soon :P
[08-Jan-2010 11:05:04] <theacolyte-> Like this graph I have 2 shades of green and 1 shade of red it ends up being brown
[08-Jan-2010 11:05:06] <Apachez> when using rrd isnt it possible to extract number of bytes that was transfered for the time period ?
[08-Jan-2010 11:05:16] <theacolyte-> yes
[08-Jan-2010 11:05:24] <theacolyte-> derive does that
[08-Jan-2010 11:05:36] <Apachez> I mean in total ?
[08-Jan-2010 11:05:41] <Apachez> where do I find this in zenoss?
[08-Jan-2010 11:05:46] <Apachez> bevcause I have the graph
[08-Jan-2010 11:05:47] <theacolyte-> Oh
[08-Jan-2010 11:05:48] <theacolyte-> No idea
[08-Jan-2010 11:05:56] <theacolyte-> You could probably translate your rrd file
[08-Jan-2010 11:06:04] <Apachez> and now I would like to find out "how many bytes were transfered 1th-2nd jan" for example
[08-Jan-2010 11:06:04] <theacolyte-> but don't quote me
[08-Jan-2010 11:06:14] <Apachez> like prtg does in a table format
[08-Jan-2010 11:06:19] <Apachez> I would be a nifty addon
[08-Jan-2010 11:06:32] <Apachez> like click on a graph and get a table of transfer per day along with coverage for that day
[08-Jan-2010 11:14:14] <Apachez> im in love with zenoss, if we just could make the gui faster it would be great :)
[08-Jan-2010 11:23:43] <rmatte> yeh, the gui can be a bit sluggish
[08-Jan-2010 11:23:57] <rmatte> but it's not horrible
[08-Jan-2010 11:24:01] <theacolyte-> I can't find colors that work!
[08-Jan-2010 11:24:06] <theacolyte-> I may just throw in the towel
[08-Jan-2010 11:24:09] <theacolyte-> Kill myself
[08-Jan-2010 11:24:20] <rmatte> lol
[08-Jan-2010 11:24:25] <rmatte> that's a bit drastic
[08-Jan-2010 11:24:37] <theacolyte-> hahahahah
[08-Jan-2010 11:25:21] <rmatte> but if you do decide on that route, remember, it's down the road, not across the street
[08-Jan-2010 11:25:40] <theacolyte-> up stream, not across?
[08-Jan-2010 11:25:57] <rmatte> lol
[08-Jan-2010 11:26:25] <theacolyte-> The problem is these graphs are alpha blending
[08-Jan-2010 11:26:32] <theacolyte-> if they were solid colors and didn't blend it would be better
[08-Jan-2010 11:34:51] <rmatte> theacolyte-: yeh
[08-Jan-2010 11:35:04] <rmatte> theacolyte-: I could have swore there was an option to disable that, but I could be wrong
[08-Jan-2010 11:56:42] <theacolyte-> rmatte: stacked!
[08-Jan-2010 11:56:45] <theacolyte-> that's how to fix it
[08-Jan-2010 12:01:33] <rmatte> hehe
[08-Jan-2010 12:01:39] <rmatte> well, not really
[08-Jan-2010 12:01:43] <rmatte> because that actually stacks them
[08-Jan-2010 12:01:53] <rmatte> that's for values that you want to basically add together
[08-Jan-2010 12:02:02] <rmatte> not for individual values
[08-Jan-2010 12:02:15] <rmatte> since there's no overlap with that
[08-Jan-2010 12:03:46] <theacolyte-> it seems like it fixed my issue
[08-Jan-2010 12:06:17] <theacolyte-> actually, you're right
[08-Jan-2010 12:06:24] <theacolyte-> it does fix the issue, but it also does what you're talkin bout
[08-Jan-2010 12:06:36] <theacolyte-> it works fine on graphs that are technically not stacked
[08-Jan-2010 12:08:13] <venturaville> are there docs anywhere for PythonPlugin?
[08-Jan-2010 12:14:44] <theacolyte-> I can't even find anything in the documentation for this, I found a bug that eluded to being able to set AA % in the RGB value
[08-Jan-2010 12:14:58] <theacolyte-> i.e FFFFFF00
[08-Jan-2010 12:19:35] <theacolyte-> There has to be a way to do it though because if you can't it would be total BS
[08-Jan-2010 12:20:59] <theacolyte-> If I put in aa % in the hex value it breaks the graph
[08-Jan-2010 12:35:48] <cryptographrix> y0
[08-Jan-2010 12:35:59] <cryptographrix> having a prob with corrupt tables
[08-Jan-2010 12:37:48] <cryptographrix> history table is corrupt
[08-Jan-2010 12:51:24] <Apachez> how corrupt ?
[08-Jan-2010 12:51:28] <Apachez> like databasecorrupt?
[08-Jan-2010 12:51:37] <Apachez> or just bad values for zenoss to recognize?
[08-Jan-2010 12:59:00] <cryptographrix> like database corrupt
[08-Jan-2010 12:59:01] <cryptographrix> :/
[08-Jan-2010 13:00:28] <twm1010> anyone want to here my painful lesson of the day?
[08-Jan-2010 13:00:44] <twm1010> *hear
[08-Jan-2010 13:00:51] <Apachez> http://www.softwareprojects.com/resources/programming/t-how-to-fix-mysql-database-myisam-innodb-1634.html
[08-Jan-2010 13:01:00] <Apachez> its a bit hard to fix innodb tables
[08-Jan-2010 13:01:06] <Apachez> much easier with myisam
[08-Jan-2010 13:05:26] <cryptographrix> I could totally drop the data and recreate - don't need the event history
[08-Jan-2010 13:24:57] <cryptographrix> is there a schema somewhere for zenoss that I could just import/run?
[08-Jan-2010 13:27:02] <nonsenso> can someone explain to me how zenoss's collector sets the OS Version?
[08-Jan-2010 13:27:18] <nonsenso> i'm completely stuck on this and it's driving me batty. :)
[08-Jan-2010 13:42:44] <cparlette> i think it's in the NewDeviceMap plugin
[08-Jan-2010 13:44:47] <nonsenso> do you know if both the DeviceMap and NewDeviceMap collector plugins are neccessary?
[08-Jan-2010 13:45:04] <cparlette> yes, they do different things
[08-Jan-2010 13:45:25] <cparlette> well, i should say, i think only NewDeviceMap is needed for OS Version
[08-Jan-2010 13:45:29] <cparlette> but you should have both applied
[08-Jan-2010 13:45:36] <cparlette> generally speaking
[08-Jan-2010 13:45:53] <nonsenso> aight.
[08-Jan-2010 13:46:43] <nonsenso> ty. :)
[08-Jan-2010 13:46:52] <cparlette> np
[08-Jan-2010 13:47:20] <nonsenso> what's your involvement with zenoss? user? developer?
[08-Jan-2010 13:49:10] <cparlette> enterprise support
[08-Jan-2010 13:49:28] <cparlette> dont tell anybody i'm here :)
[08-Jan-2010 13:49:37] <nonsenso> hehe. i don't know what you're talking about. :)
[08-Jan-2010 13:49:37] <chkraise> i know
[08-Jan-2010 13:49:38] <chkraise> lol
[08-Jan-2010 13:49:58] <chkraise> Enterprise customer here - Hi chris .. Hector here
[08-Jan-2010 13:50:05] <chkraise> Ragingwire
[08-Jan-2010 13:50:09] <cparlette> oh hey hector
[08-Jan-2010 13:50:31] <chkraise> I need to log back into the portal and resolv that wmi issue through nat we opened up last week
[08-Jan-2010 13:50:37] <nonsenso> i've done a handful of small core installs for clients. everyone seems to love it so far.
[08-Jan-2010 13:51:16] <chkraise> yeah its nice, i was reluctant at first because I love nagios
[08-Jan-2010 13:51:17] <cparlette> nonsenso: awesome, glad to hear it
[08-Jan-2010 13:51:30] <chkraise> used netsaint -> nagios for ever till now
[08-Jan-2010 13:51:36] <nonsenso> same here.
[08-Jan-2010 13:51:39] <chkraise> and i got to say zenoss is a nice upgrade
[08-Jan-2010 13:51:40] <cparlette> chkraise: yeah, i'll have to see what else i can find about getting around the NAT issues
[08-Jan-2010 13:51:49] <nonsenso> now ppl are like, "what you don't use cacti and nagios anymore?!"
[08-Jan-2010 13:52:25] <chkraise> its ok chris we found that if you modify the local hosts files, and upgrade to 2.5.1 it works
[08-Jan-2010 13:52:36] <chkraise> through the natted IP
[08-Jan-2010 13:52:59] <cparlette> oh interesting, just hardcode the NAT IP into the windows hosts?
[08-Jan-2010 13:53:03] <chkraise> yep
[08-Jan-2010 13:53:04] <cparlette> or into the zenoss hosts
[08-Jan-2010 13:53:05] <cparlette> ok
[08-Jan-2010 13:53:07] <chkraise> both
[08-Jan-2010 13:53:10] <chkraise> we did both sides
[08-Jan-2010 13:54:18] <chkraise> Lunch time..
[08-Jan-2010 13:54:28] <cparlette> that's definitely useful to know, we've had a couple others bring up issues with NAT'd addresses
[08-Jan-2010 14:06:50] <nonsenso> aight. so the oid that newdevicemap pulls returns just the linux kernel version. in my installation of zenoss, it's somehow mapping to a defined product. in my case "Red Hat Enterprise Linux". how does that mapping work?
[08-Jan-2010 14:08:05] <nonsenso> ideally, i'd like to see the version of redhat in the OS Version field.. that's muh goal.
[08-Jan-2010 14:34:36] <jsm> i'm noticing extremely poor performance when click on a device in the device list.. it's taking minutes for it to yield a page.. i'm guessing there's something wonky going on with the db, must be taking forever to return a query
[08-Jan-2010 14:40:00] <gwb2351> is there a way to get zenoss to use the OS-supplied ssh command instead of python built-in to data collection?
[08-Jan-2010 14:43:03] <rmatte> jsm: did you have an event flood or something recently?
[08-Jan-2010 14:43:28] <rmatte> jsm: when you click on a device to load a device page it has to go through the last 7 days woth of events to calculate the availability for the device
[08-Jan-2010 14:43:52] <jsm> rmatte: i will check that, thanks. also i'm going to tune my.cnf a bit.
[08-Jan-2010 14:43:52] <rmatte> jsm: so if you have a really slow db (old hardware), or a lot of events in the console from the past 7 days then yes, you're going to have issues
[08-Jan-2010 14:43:59] <rmatte> cool
[08-Jan-2010 14:44:05] <jsm> any my.cnf recommendations ?
[08-Jan-2010 14:44:25] <rmatte> one second
[08-Jan-2010 14:44:41] <rmatte> http://community.zenoss.org/docs/DOC-2521
[08-Jan-2010 14:44:47] <rmatte> there's some mysql tuning info there
[08-Jan-2010 14:44:52] <rmatte> as well as some other general tuning info
[08-Jan-2010 14:45:24] <jsm> awesome, thanks
[08-Jan-2010 14:46:03] <rmatte> np
[08-Jan-2010 14:55:03] <Apachez> jsm: the problem is mostly not mysql but the application server (zope) that zenoss uses
[08-Jan-2010 14:55:32] <jsm> Apachez: what's the best way to tune that?
[08-Jan-2010 14:55:37] <Apachez> zope?
[08-Jan-2010 14:55:40] <jsm> yes
[08-Jan-2010 14:55:49] <Apachez> not much you can do as far as I know
[08-Jan-2010 14:56:26] <Apachez> you could try to in zope.conf set zserver-threads 1 and python-check-interal 1000 (or whatever the spelling is for that, its just below zserver-threads in the conf file) and then restart zenoss-stack
[08-Jan-2010 14:56:27] <rmatte> there are some zope tuning suggestions in that same document I pasted
[08-Jan-2010 14:56:32] <rmatte> but I'm not sure how effective they are
[08-Jan-2010 14:56:38] <rmatte> they seemed to speed things up a bit when I did them
[08-Jan-2010 14:56:48] <Apachez> for my.cnf you can do several stuff but most of them are only minor and will have effect if you have a huge database
[08-Jan-2010 14:57:15] <Apachez> you could also try "OPTIMIZE TABLE xxx;" for every table in events
[08-Jan-2010 14:57:26] <Apachez> however since its innodb only the index will be sorted but that could be enough
[08-Jan-2010 14:57:43] <Apachez> I tried convering the tables into myisam but that didnt work well with 2.5.1
[08-Jan-2010 14:57:54] <Apachez> so im currently doing some query log digging right now
[08-Jan-2010 14:58:25] <Apachez> I can paste my current my.cnf settings (not yet optimal since I test different stuff but they are sure better then the nonexisting default settings :)
[08-Jan-2010 14:58:57] <jsm> wish zenoss would allow postgresql as the db
[08-Jan-2010 14:59:32] <theacolyte-> rmatte: I gave up on it and had to post on the forums
[08-Jan-2010 15:00:17] <Apachez> http://www.pastebin.ca/1743178
[08-Jan-2010 15:00:20] <Apachez> NOTE! NOTE! NOTE! NOTE! NOTE!
[08-Jan-2010 15:00:28] <Apachez> comment out innodb_file_per_table = 1
[08-Jan-2010 15:00:46] <Apachez> you will also need to do this after you use the above settings:
[08-Jan-2010 15:00:55] <Apachez> 1) /etc/init.d/zenoss-stack stop mysql
[08-Jan-2010 15:00:56] <jsm> Daemon: zopectl Error: 'pool-size' is not a known key name
[08-Jan-2010 15:01:01] <jsm> interesting
[08-Jan-2010 15:01:22] <jsm> funny since they list it here: http://community.zenoss.org/docs/DOC-2521
[08-Jan-2010 15:01:23] <Apachez> 2) rm /usr/local/zenoss/mysql/data/ib_logfile
[08-Jan-2010 15:01:29] <Apachez> 3) /etc/init.d/zenoss-stack start mysql
[08-Jan-2010 15:02:03] <jsm> ah, put it in the wrong place
[08-Jan-2010 15:02:18] <Apachez> 2) rm /usr/local/zenoss/mysql/data/ib_logfile*
[08-Jan-2010 15:02:32] <Apachez> since the settings use different logsize for innodb transactionlogs
[08-Jan-2010 15:03:00] <Apachez> the file_per_table thingy is to let each table have its own tablespace
[08-Jan-2010 15:03:13] <Apachez> to lower the diskusage when they all share the same space
[08-Jan-2010 15:03:23] <Apachez> the problem with sharing same space is when you get a huge history
[08-Jan-2010 15:03:25] <Apachez> and delete entries
[08-Jan-2010 15:03:37] <Apachez> the file will remain the previous size (thats how innodb works)
[08-Jan-2010 15:03:55] <Apachez> only way to shrink it is to recreate it (from backup.sql file)
[08-Jan-2010 15:04:30] <Apachez> so if using per file space then a simple "ALTER TABLE xxx ENGINE=MYISAM;" followed by "ALTER TABLE xxx ENGINE=INNODB"; would regain the lost space
[08-Jan-2010 15:04:42] <Apachez> also optimize table is nice
[08-Jan-2010 15:04:50] <Apachez> will sort index and regain space on myisam
[08-Jan-2010 15:04:55] <Apachez> but on innodb it will only sort index
[08-Jan-2010 15:05:56] <RobertLaptop> Thank you Apachez
[08-Jan-2010 15:11:01] <RobertLaptop> Apachez will the Alter Table tick work if you created the DB wiouth out innodb_file_per_table = 1 set?
[08-Jan-2010 15:12:44] <Apachez> RobertLaptop: no, you will need additional methods then
[08-Jan-2010 15:12:49] <Apachez> methods=steps
[08-Jan-2010 15:12:57] <Apachez> but basically
[08-Jan-2010 15:14:08] <Apachez> 1) alter table xxx engine=myisam; for every table in the events database 2) stop mysql service 3) delete ib* files in data dir 4) start mysql 5) alter table xxx engine=innodb;
[08-Jan-2010 15:14:23] <Apachez> and I always run optimzie table when I have the chance
[08-Jan-2010 15:14:31] <Apachez> O I forgot... shutdown zenoss first
[08-Jan-2010 15:14:40] <Apachez> so you wont get broken records during your maintence
[08-Jan-2010 15:15:05] <Apachez> the shutdown mysql and start everything again through zenoss-stack and then verify in the gui that you can reach devices and events
[08-Jan-2010 15:15:21] <Apachez> but if you do that you could add the per file thingy aswell
[08-Jan-2010 15:15:25] <RobertLaptop> ok.
[08-Jan-2010 15:15:40] <Apachez> so when you convert back to innodb each table will have its own tablespace
[08-Jan-2010 15:15:52] <Apachez> which will make it easier for you to keep track of tablesizes
[08-Jan-2010 15:15:53] <RobertLaptop> Yea. That is the plan.
[08-Jan-2010 15:16:20] <Apachez> most important is to have a filebackup of the datadir AFTER you have shutdown the mysql service
[08-Jan-2010 15:16:52] <Apachez> this way you can just shutdown mysql, clean datadir and copy back stuff form backup and then start mysql again
[08-Jan-2010 15:17:02] <Apachez> also not the chown / chmod so they are correct
[08-Jan-2010 15:17:20] <Apachez> note
[08-Jan-2010 15:17:34] <RobertLaptop> Yea backups are always good
[08-Jan-2010 15:29:41] <theacolyte-> Someone answer my forum post
[08-Jan-2010 15:29:45] <theacolyte-> http://community.zenoss.org/thread/12469
[08-Jan-2010 15:29:46] <theacolyte-> go!
[08-Jan-2010 15:42:10] <rmatte> theacolyte-: I saw this in the docs for rrdgraph: Optionally you may add another hexadecimal number specifying the transparency (FF is solid).
[08-Jan-2010 15:42:19] <theacolyte-> Yeah I tried that
[08-Jan-2010 15:42:21] <rmatte> so if you do FFFFFFFF
[08-Jan-2010 15:42:22] <rmatte> ?
[08-Jan-2010 15:42:22] <theacolyte-> Breaks the graphs
[08-Jan-2010 15:42:27] <rmatte> hmmm
[08-Jan-2010 15:42:27] <theacolyte-> broken image
[08-Jan-2010 15:42:44] <rmatte> no idea then
[08-Jan-2010 15:42:54] <theacolyte-> Yeah
[08-Jan-2010 15:42:59] <theacolyte-> I searched for bugs on it too
[08-Jan-2010 15:43:06] <theacolyte-> sad, sadface
[08-Jan-2010 15:43:13] <theacolyte-> totally borks my graphs prettiness
[08-Jan-2010 15:44:56] <Apachez> what about FFFFFF50 ?
[08-Jan-2010 15:45:17] <Apachez> if that breaks the graph aswell then you have your problem elsewhere
[08-Jan-2010 15:53:10] <rmatte> Apachez: not necessarily, Zenoss might not just like the additional 2 values
[08-Jan-2010 15:53:26] <theacolyte-> yeah it doesn't like that
[08-Jan-2010 15:53:30] <theacolyte-> which si funny
[08-Jan-2010 15:53:33] <theacolyte-> there's a bug for the same thing
[08-Jan-2010 15:53:38] <theacolyte-> almost looks like a regression
[08-Jan-2010 15:54:14] <theacolyte-> http://dev.zenoss.org/trac/ticket/1660
[08-Jan-2010 15:55:03] <Apachez> heh
[08-Jan-2010 15:55:13] <Apachez> not much knowledge in that one who created the tables
[08-Jan-2010 15:55:17] <Apachez> PRIMARY KEY (`evid`,`name`),
[08-Jan-2010 15:55:18] <Apachez> KEY `evididx` (`evid`)
[08-Jan-2010 15:55:23] <Apachez> redundant ehh? :)
[08-Jan-2010 15:58:02] <Apachez> table log is missing primary key?
[08-Jan-2010 15:58:10] <Apachez> but I dont know whats it used for either
[08-Jan-2010 15:58:14] <Apachez> the table that is
[08-Jan-2010 15:59:23] <Apachez> also the naming isnt consistent
[08-Jan-2010 15:59:36] <Apachez> who is in charge of the zenoss db structure?
[08-Jan-2010 15:59:48] <Apachez> perhaps I should have a small chat with that person :)
[08-Jan-2010 16:00:51] <rmatte> Apachez: it's passed through tons of different developers
[08-Jan-2010 16:01:02] <rmatte> Apachez: chances are whoever did that doesn't work for them anymore
[08-Jan-2010 16:13:09] <Apachez> order by lastTime DESC, lastTime DESC limit 0, 100
[08-Jan-2010 16:13:13] <Apachez> funny code :P
[08-Jan-2010 16:13:26] <Apachez> I really hope all zenoss db calls are in a single file? :)
[08-Jan-2010 16:16:13] <rmatte> holy crap do people ever need to learn how to update a wiki here
[08-Jan-2010 16:16:20] <rmatte> our wiki code is an absolute disaster
[08-Jan-2010 16:21:06] <theacolyte-> This is weird
[08-Jan-2010 16:21:19] <theacolyte-> Modelling one of my ESX servers and it's not showing interfaces
[08-Jan-2010 16:21:29] <theacolyte-> I can see them when I walk though
[08-Jan-2010 16:25:45] <Apachez> ALTER TABLE detail DROP KEY `evididx`;
[08-Jan-2010 16:25:50] <Apachez> ALTER TABLE heartbeat ADD KEY `lastTimeidx` (`lastTime`)
[08-Jan-2010 16:26:01] <Apachez> ALTER TABLE status ADD KEY `Ix_eventClass_severity_eventState` (`eventClass`, `severity`, `eventState`);
[08-Jan-2010 16:26:01] <Apachez> ALTER TABLE status ADD KEY `Ix_device_severity_eventState` (`device`, `severity`, `eventState`);
[08-Jan-2010 16:26:01] <Apachez> ALTER TABLE status ADD KEY `Ix_eventClass_eventState` (`eventClass`, `eventState`);
[08-Jan-2010 16:26:08] <Apachez> then run optimize table on all 6 tables
[08-Jan-2010 16:26:16] <Apachez> and then take a look in the gui if its any faster
[08-Jan-2010 16:29:03] <Apachez> jsm: still around ?
[08-Jan-2010 16:30:18] <jsm> yes
[08-Jan-2010 16:30:27] <jsm> the tips helped for sure
[08-Jan-2010 16:30:31] <jsm> but still not the performance i'd like to see
[08-Jan-2010 16:31:23] <Apachez> try this
[08-Jan-2010 16:31:39] <Apachez> ALTER TABLE status ADD KEY `Ix_device_severity_eventState` (`device`, `severity`, `eventState`);
[08-Jan-2010 16:31:50] <Apachez> and then run OPTIMIZE TABLE status;
[08-Jan-2010 16:31:56] <Apachez> and then go in to the devicelist
[08-Jan-2010 16:32:19] <Apachez> I will try to do some more querylogging to find out what queries are being runned towards the db
[08-Jan-2010 16:32:35] <Apachez> however it would be much more easier if all queries was collected in a single dbaccesss file
[08-Jan-2010 16:32:42] <Apachez> but I guess they arent :(
[08-Jan-2010 16:32:55] <Apachez> will try some grepping to see if I can locate some sort of dbfile
[08-Jan-2010 16:36:45] <theacolyte-> man, today is my brick wall with zenoss day
[08-Jan-2010 16:40:48] <Apachez> ./zenoss/Products/ZenEvents/EventManagerBase.py: select = "select count(*) from %s where %s" % (self.statusTable, where)
[08-Jan-2010 16:40:52] <Apachez> wtf this sucks
[08-Jan-2010 16:41:09] <Apachez> in terms of easy way to locate which queries are being runned
[08-Jan-2010 16:41:25] <Apachez> now I have to do it the hard way and enable query logging and then try to click everywehre in the gui
[08-Jan-2010 16:41:26] <Apachez> hmpf
[08-Jan-2010 16:51:38] <rmatte> theacolyte-: I've had tons of those
[08-Jan-2010 16:51:54] <theacolyte-> I actually am not sure hwat to do about either of these
[08-Jan-2010 16:52:02] <theacolyte-> I'm not sure why this isn't modelling my NIC's in this ESX server
[08-Jan-2010 16:52:05] <theacolyte-> I'm looking right at the OID
[08-Jan-2010 17:00:10] <theacolyte-> is ethernetCsmacd always applied to devices by default?
[08-Jan-2010 17:00:16] <theacolyte-> rmatte: ^
[08-Jan-2010 17:00:32] <rmatte> it's applied to any interface with and interface type of ethernetCsmacd
[08-Jan-2010 17:00:49] <rmatte> and by default it's applied to all interfaces unless there's another template that matches the interface type
[08-Jan-2010 17:01:00] <rmatte> the name of the template has to be an exact match (case sensitive)
[08-Jan-2010 17:01:08] <rmatte> click on any interface and you'll see type listed
[08-Jan-2010 17:01:15] <theacolyte-> There are 0 interfaces listed
[08-Jan-2010 17:01:21] <rmatte> hunh?
[08-Jan-2010 17:01:24] <theacolyte-> Yeah
[08-Jan-2010 17:01:29] <rmatte> you don't have any interfaces on your OS tab for any devices?
[08-Jan-2010 17:01:30] <theacolyte-> my ESX server with 6 NIC ports
[08-Jan-2010 17:01:35] <theacolyte-> has no interfaces
[08-Jan-2010 17:01:53] <rmatte> ok, now you added this server as a device in Zenoss?
[08-Jan-2010 17:02:00] <theacolyte-> yep
[08-Jan-2010 17:02:02] <theacolyte-> Name/OID: ifDescr.2; Value (OctetString): vmnic0
[08-Jan-2010 17:02:14] <rmatte> and you modeled the device?
[08-Jan-2010 17:02:17] <theacolyte-> yep
[08-Jan-2010 17:02:23] <theacolyte-> and remodeled to be sure
[08-Jan-2010 17:03:19] <rmatte> ok, well, are there any error while modelling?
[08-Jan-2010 17:03:29] <rmatte> did it say anything about not being able to monitor interfaces?
[08-Jan-2010 17:03:35] <rmatte> erm, model rather
[08-Jan-2010 17:04:06] <theacolyte-> Unable to get data for bladjsda from either ipAddrTable or ipNetToMediaTable -- skipping model
[08-Jan-2010 17:04:18] <theacolyte-> oh, I know why now
[08-Jan-2010 17:04:26] <theacolyte-> those NIC's don't actually have IP's
[08-Jan-2010 17:04:38] <rmatte> ok... I know a way you can get around that
[08-Jan-2010 17:04:53] <theacolyte-> that being said though, in one of the zenpacks for ESX that doesn't have networking available has those NICs listed
[08-Jan-2010 17:05:21] <rmatte> the zenpack for ESX uses different methods to gather the data
[08-Jan-2010 17:06:14] <rmatte> Ok, so I opened ticket http://dev.zenoss.org/trac/ticket/5691 for that problem a while ago
[08-Jan-2010 17:06:28] <rmatte> the fix is detailed in that ticket (well, it's a workaround, not a fix)
[08-Jan-2010 17:07:06] <rmatte> edit /usr/local/zenoss/zenoss/Products/DataCollector/plugins/zenoss/snmp/InterfaceMap.py
[08-Jan-2010 17:07:26] <rmatte> find the section shown in the ticket and do what's show
[08-Jan-2010 17:07:28] <rmatte> shown*
[08-Jan-2010 17:07:43] <rmatte> (comment out the line "return None" as shown)
[08-Jan-2010 17:07:45] <rmatte> then restart Zenoss
[08-Jan-2010 17:07:48] <rmatte> then remodel the device
[08-Jan-2010 17:08:01] <theacolyte-> wow that's weird....
[08-Jan-2010 17:08:20] <theacolyte-> let me give it a whirl
[08-Jan-2010 17:08:24] <rmatte> k
[08-Jan-2010 17:08:25] <theacolyte-> goddamn man you know everything about this product
[08-Jan-2010 17:08:43] <rmatte> Yeh, I've been using it for a long time, I don't know everything though lol
[08-Jan-2010 17:08:53] <rmatte> the man who seems to know everything about it is Chet Luther
[08-Jan-2010 17:08:56] <rmatte> (cluther)
[08-Jan-2010 17:09:02] <rmatte> but he works for Zenoss Inc. lol
[08-Jan-2010 17:10:51] <theacolyte-> restarting zenoss stack
[08-Jan-2010 17:11:06] <rmatte> cool
[08-Jan-2010 17:12:46] <theacolyte-> didn't do it
[08-Jan-2010 17:13:29] <rmatte> really?
[08-Jan-2010 17:13:33] <theacolyte-> yeah
[08-Jan-2010 17:13:36] <rmatte> that's quite weird then
[08-Jan-2010 17:13:40] <theacolyte-> same error
[08-Jan-2010 17:13:48] <rmatte> yeh, the error will still show
[08-Jan-2010 17:13:49] <theacolyte-> thing is those interfaces don't have IP's
[08-Jan-2010 17:13:50] <theacolyte-> sec
[08-Jan-2010 17:13:52] <rmatte> but the interface should be there
[08-Jan-2010 17:13:59] <rmatte> interfaces*
[08-Jan-2010 17:14:03] <rmatte> the change doesn't eliminate the error
[08-Jan-2010 17:14:08] <theacolyte-> http://community.zenoss.org/docs/DOC-3481
[08-Jan-2010 17:14:09] <rmatte> but it should make the modelling work anyways
[08-Jan-2010 17:14:13] <theacolyte-> oh
[08-Jan-2010 17:14:23] <theacolyte-> nope, nothing :(
[08-Jan-2010 17:14:30] <rmatte> ah, weird then
[08-Jan-2010 17:14:54] <rmatte> you're 100% sure you commented out the right line?
[08-Jan-2010 17:15:07] <theacolyte-> else:
[08-Jan-2010 17:15:07] <theacolyte-> log.error("Unable to get data for %s from either ipAddrTable or"
[08-Jan-2010 17:15:07] <theacolyte-> " ipNetToMediaTable -- skipping model" % device.id)
[08-Jan-2010 17:15:07] <theacolyte-> # return None
[08-Jan-2010 17:15:32] <theacolyte-> file is in /usr/local/zenoss/zenoss/Products/DataCollector/plugins/zenoss/snmp/InterfaceMap.py
[08-Jan-2010 17:16:49] <rmatte> ypu
[08-Jan-2010 17:16:51] <rmatte> yup*
[08-Jan-2010 17:23:56] <theacolyte-> I wish I knew something about how to monitor through SSH, I shoudl read about that
[08-Jan-2010 17:24:10] <theacolyte-> I could just run esxtop and parse that
[08-Jan-2010 17:31:57] <Apachez> jsm: tried the new index yet?
[08-Jan-2010 17:37:16] <jsm> Apachez: i will try those on monday
[08-Jan-2010 17:37:24] <jsm> but thank you!
[08-Jan-2010 17:40:26] <RobertLaptop> Has anyone gotten this to work under CentOS? http://linuxtrek1.blogspot.com/2009/02/zenoss-monitor-free-esxi-version.html
[08-Jan-2010 17:42:52] <Apachez> hmpf
[08-Jan-2010 17:42:56] <Apachez> 46 queries to analyze
[08-Jan-2010 17:48:47] <RobertLaptop> Question how do you tell python where to look for it library files?
[08-Jan-2010 17:49:28] <RobertLaptop> I need pywbem and the wbem module installs it as a library but the script I downloaded doesn't know were to find it?
[08-Jan-2010 17:50:45] <Apachez> 19 unique queries
[09-Jan-2010 00:00:30] [disconnected at Sat Jan 9 00:00:30 2010]
[09-Jan-2010 00:00:30] [connected at Sat Jan 9 00:00:30 2010]
[09-Jan-2010 14:10:44] <Apachez> did zenoss support v2c traps or only v1 ?
[09-Jan-2010 14:17:08] <Apachez> do
[09-Jan-2010 15:10:48] <chemist> Apachez: yes it does
[09-Jan-2010 15:17:48] <Apachez> I guess v2c traps are prefered over v1 ?
[09-Jan-2010 15:29:25] <theacolyte-> yeah
[09-Jan-2010 15:32:54] <chemist> v3 would be even better,but I dont think Zenoss supports v3 traps
[09-Jan-2010 17:41:24] <Apachez> hmpf mray is still missing
[10-Jan-2010 00:00:30] [disconnected at Sun Jan 10 00:00:30 2010]
[10-Jan-2010 00:00:30] [connected at Sun Jan 10 00:00:30 2010]
[10-Jan-2010 07:29:21] <skipzoid> hey all
[10-Jan-2010 07:30:14] <skipzoid> any gurus here ?
[10-Jan-2010 09:46:27] <Apachez> skipzoid: yeah?
[10-Jan-2010 12:24:17] <m4rzito> hi
[11-Jan-2010 00:00:30] [disconnected at Mon Jan 11 00:00:30 2010]
[11-Jan-2010 00:00:30] [connected at Mon Jan 11 00:00:30 2010]
[11-Jan-2010 08:31:40] <eidolon> mornin monitoring geeks.
[11-Jan-2010 08:31:51] <RobertLaptop> Morning
[11-Jan-2010 08:37:44] <eidolon> so how do folks solve the "i only want such and such a service monitored for this timeframe" problem? We have a process that should be monitored during a certain window, but is down other times.
[11-Jan-2010 08:37:53] <eidolon> we can't put in maintenance windows - that affects an entire host.
[11-Jan-2010 08:40:05] <RobertLaptop> Good question. One I hope you get an answer for as I have a box that has the same issue. :)
[11-Jan-2010 08:42:24] <eidolon> my understanding is that it's a known issue, and there's no known fix ;(
[11-Jan-2010 08:44:16] * eidolon also vaguely wishes i could say "See this templte? bind it to all hosts."
[11-Jan-2010 08:44:24] * eidolon has 40 hosts to add the ntpmonitor template to.
[11-Jan-2010 08:45:22] <RobertLaptop> eidolon are the 40 hosts in some kind of group?
[11-Jan-2010 08:45:32] <eidolon> yyyyessss..... they are.
[11-Jan-2010 08:45:36] * eidolon senses a path :)
[11-Jan-2010 08:45:52] <RobertLaptop> Pull up the group select template bind template.
[11-Jan-2010 08:46:28] <eidolon> not sure where that is, soryr?
[11-Jan-2010 08:46:44] <eidolon> (i have customn groups. like /Groups/Customer/[customername]...)
[11-Jan-2010 08:47:03] <eidolon> so if i could say 'eveyrone in /groups/customer, bind the NtpTemplate, i'd be good
[11-Jan-2010 08:47:46] <RobertLaptop> Pretty much
[11-Jan-2010 08:48:15] <eidolon> how? :)
[11-Jan-2010 08:49:06] <RobertLaptop> Checking now. I know I did something like that?
[11-Jan-2010 08:49:19] <eidolon> k
[11-Jan-2010 08:51:01] <RobertLaptop> Sorry I am bad. It only works with classes sorry for getting your hopes up.
[11-Jan-2010 08:51:19] * eidolon wails!
[11-Jan-2010 08:51:21] * eidolon was all set!
[11-Jan-2010 08:51:23] * eidolon sniffs
[11-Jan-2010 09:18:56] <rmatte> eidolon, personally I'd do it with a transform
[11-Jan-2010 09:19:33] <rmatte> eidolon: transforms are python scripts, therefore you have access to date and time info... you could write a script which auto-drops the event unless it's during a certain time window
[11-Jan-2010 09:19:41] <rmatte> eidolon: kind of a dirty solution, but it'd work
[11-Jan-2010 09:20:31] <RobertLaptop> If you know python that is :)
[11-Jan-2010 09:20:52] <rmatte> RobertLaptop: google is your friend, Python is honestly not that difficult to pick up if you put the effort in...
[11-Jan-2010 09:21:01] <rmatte> and if you're managing Zenoss it's basically a must
[11-Jan-2010 09:21:26] <rmatte> You have no idea how powerful event transforms can be until you learn Python
[11-Jan-2010 09:23:08] <rmatte> Lots of people who have looked at my code and don't know Python have said: "That's Python!? That's so easy to read and I don't even know Python."
[11-Jan-2010 09:27:59] <cparlette> i agree, a lot of times if you write pseudocode for what you want to do, it's likely pretty close to python already
[11-Jan-2010 09:28:42] <rmatte> yeh
[11-Jan-2010 09:30:12] <RobertLaptop> Since all my pheudocode look like c/php I wouldn't bet on that :)
[11-Jan-2010 09:31:26] <rmatte> then it's not really pseudocode lol
[11-Jan-2010 09:31:35] <rmatte> pseudocode should almost be plain english
[11-Jan-2010 09:31:53] <rmatte> with a general structure
[11-Jan-2010 09:33:03] <RobertLaptop> Haven't you heard the old saying. The keep making programing lang to look more like English until someone realized programmers don't know English
[11-Jan-2010 09:33:55] <rmatte> yeh, more of a saying than truth though
[11-Jan-2010 09:34:09] <rmatte> unless you don't speak english at all
[11-Jan-2010 09:34:22] <RobertLaptop> They fact is I learned programing in C so Plain lang to me writing out a decision tree will look very C like.
[11-Jan-2010 09:34:52] <rmatte> I used to write in C too, but it's quite primitive at this point
[11-Jan-2010 09:35:03] <rmatte> gotta teach the old dog new tricks
[11-Jan-2010 09:35:20] <cparlette> if you have knowledge of a programming language, and pull up a couple of transform examples to look off of, you'll be able to do quite a bit
[11-Jan-2010 09:35:35] <rmatte> yeh, transforms are braindead simple to code
[11-Jan-2010 09:41:38] <RobertLaptop> That I will give you. That is how I learned a little perl. I had scripts that I played with until they gave me what I wanted. A few example transforms would do the same. :)
[11-Jan-2010 09:42:24] <eidolon> rmatte: yeah, my python-fu is limited.
[11-Jan-2010 09:42:44] <rmatte> hehe
[11-Jan-2010 09:43:12] <eidolon> yeah, a nice tutorial on them would be handy - also, a situation where you can tinker with them without blowing up your entire production environonment would be good. "Here's how to put at test transform in place, and tinker around with it. You can set the alerting on this to just mail you, or just run it by hand."
[11-Jan-2010 09:43:59] <cparlette> there's an enhancement request in that asks for something to show the "path" an event took and where it was modified
[11-Jan-2010 09:44:10] <cparlette> which would help when troubleshooting your transforms
[11-Jan-2010 09:44:44] <RobertLaptop> A guide on testing stuff would be great. I am still tiring to get the esxi script to run. It is telling me that pxwbem isn't present but I find it when search my HD some it is a setting issue.
[11-Jan-2010 09:45:15] <RobertLaptop> some=so
[11-Jan-2010 09:54:24] <rmatte> RobertLaptop: probably a path issue
[11-Jan-2010 09:55:28] <RobertLaptop> rmatte I am sure just have no idea how to define paths with python.
[11-Jan-2010 09:57:19] <rmatte> well, depends on how the script executes the external app/script
[11-Jan-2010 09:59:28] <RobertLaptop> Command line
[11-Jan-2010 10:03:22] <RobertLaptop> I am tiring to get this guide to work. http://linuxtrek1.blogspot.com/2009/02/zenoss-monitor-free-esxi-version.html
[11-Jan-2010 10:07:10] <eidolon> ohh, esx monitoring
[11-Jan-2010 10:07:22] <eidolon> we have zenoss enterprise, so we have esx monitors arelady, i'm pretty sure.
[11-Jan-2010 10:08:17] <RobertLaptop> ESXi !- ESX. ESX is easy you enable snmp and you are done. ESXi doesn't support snmp at all you only option is WBEM.
[11-Jan-2010 10:08:24] <eidolon> i know :)
[11-Jan-2010 10:08:31] <eidolon> mostly i'm just being irritating :)
[11-Jan-2010 10:20:02] <theacolyte-> rmatte: No one has fixed my problems :(
[11-Jan-2010 10:27:37] <rmatte> theacolyte-: which?
[11-Jan-2010 10:27:48] <theacolyte-> rmatte: Transparency issue
[11-Jan-2010 10:28:40] <theacolyte-> http://community.zenoss.org/message/44363#44363
[11-Jan-2010 10:29:34] <rmatte> oh
[11-Jan-2010 10:29:38] <rmatte> yeh I remember
[11-Jan-2010 10:29:46] <rmatte> it's probably a bug
[11-Jan-2010 10:29:55] <rmatte> did you open a trac ticket for it?
[11-Jan-2010 10:29:58] <theacolyte-> Today I'm back to figuring out how the heck I can graph my ESX hosts
[11-Jan-2010 10:30:07] <theacolyte-> No, but I'm going to by EOD if I don't get an answer on the forums
[11-Jan-2010 10:30:12] <rmatte> cool
[11-Jan-2010 10:36:50] <theacolyte-> omg it's mray
[11-Jan-2010 10:36:52] <theacolyte-> time to harass
[11-Jan-2010 10:37:03] <theacolyte-> welcome back :)
[11-Jan-2010 10:37:17] <mrayzenoss> yup
[11-Jan-2010 10:42:56] <rmatte> yeh, welcome back
[11-Jan-2010 10:43:00] <rmatte> ;)
[11-Jan-2010 10:43:45] <theacolyte-> It's a shame about the game
[11-Jan-2010 10:43:53] <theacolyte-> it looked like mccoy barely got touched
[11-Jan-2010 10:44:11] <rmatte> mrayzenoss: this was downgraded to a 4 as it's "working as designed" but there's still a problem with manually setting software info: http://dev.zenoss.org/trac/ticket/5473
[11-Jan-2010 10:44:27] <rmatte> perhaps it could be put back to a 2 or a 3?
[11-Jan-2010 10:49:27] <rmatte> actually hmmm weird
[11-Jan-2010 10:49:36] <rmatte> there's 3 types of entries in products
[11-Jan-2010 10:49:44] <rmatte> Hardware, Software, and Operating Syste,
[11-Jan-2010 10:49:47] <rmatte> System*
[11-Jan-2010 10:50:00] <mrayzenoss> rmatte: yeah, I saw that, I'll look into it
[11-Jan-2010 10:50:01] <rmatte> after doing a reindex the Operating System entries show up
[11-Jan-2010 10:50:10] <rmatte> but the Software entries don't
[11-Jan-2010 10:54:33] <rmatte> anyways, I updated the ticket with that info and I'll leave it to you guys, it's not as critical as the hardware info was
[11-Jan-2010 11:03:22] <theacolyte-> database conflict error (oid 0x0df0, class Products.ZenModel.PerformanceConf.PerformanceConf)
[11-Jan-2010 11:04:39] <theacolyte-> Funny, it got added anyway
[11-Jan-2010 11:14:12] <theacolyte-> Anyone know if the mysql zenpack allows monitoring of mysql itself?
[11-Jan-2010 11:14:17] <theacolyte-> (besides the counters)
[11-Jan-2010 11:14:44] <rmatte> you mean status?
[11-Jan-2010 11:14:54] <theacolyte-> Yeah
[11-Jan-2010 11:14:57] <rmatte> you could just do that via WMI services monitoring
[11-Jan-2010 11:15:13] <theacolyte-> I *could*
[11-Jan-2010 11:15:14] <theacolyte-> hehe
[11-Jan-2010 11:15:16] <rmatte> The pack is really only for performance
[11-Jan-2010 11:15:42] <theacolyte-> got it
[11-Jan-2010 11:16:19] <rmatte> It could be converted to do status, but why reinvent the wheel
[11-Jan-2010 11:16:38] <rmatte> well, time to go grab some lunch, afk for a few
[11-Jan-2010 11:17:28] <theacolyte-> True
[11-Jan-2010 11:23:54] <theacolyte-> Man, none of these zenpacks ever work right
[11-Jan-2010 11:24:01] <theacolyte-> I'm like 1/5
[11-Jan-2010 11:24:09] <rmatte> lol
[11-Jan-2010 11:26:14] <rmatte> well, there are tons of ZenPacks but the quality isn't always there
[11-Jan-2010 11:26:40] <rmatte> like even my current windows performance ZenPacks aren't that great, but I'm working to improve them for a future release
[11-Jan-2010 11:26:55] <theacolyte-> I just end up rebuilding everything
[11-Jan-2010 11:26:58] <theacolyte-> This one is the IIS6 one
[11-Jan-2010 11:27:18] <theacolyte-> trying to rememebr my wmic comamnd line
[11-Jan-2010 11:27:29] <rmatte> If you end up rebuilding it you could always submit it as an update, provided it provides all original promised functionality
[11-Jan-2010 11:27:47] <rmatte> ummm, well hold up there
[11-Jan-2010 11:27:55] <rmatte> the old IIS6 one or the new one?
[11-Jan-2010 11:28:14] <rmatte> ...does it use the WMI Data Source pack?
[11-Jan-2010 11:28:18] <theacolyte-> It does
[11-Jan-2010 11:28:23] <rmatte> k cool
[11-Jan-2010 11:28:39] <theacolyte-> It's the one egor did
[11-Jan-2010 11:28:54] <rmatte> k
[11-Jan-2010 11:29:03] <rmatte> what wasn't working about it?
[11-Jan-2010 11:29:24] <theacolyte-> The path is wrong I think
[11-Jan-2010 11:29:26] <theacolyte-> (WMI path)
[11-Jan-2010 11:29:37] <theacolyte-> ERROR: Retrieve result data.
[11-Jan-2010 11:29:38] <theacolyte-> NTSTATUS: NT code 0x80041010 - NT code 0x80041010
[11-Jan-2010 11:29:50] <theacolyte-> That's from my experience a bad counter
[11-Jan-2010 11:31:28] <theacolyte-> It may just be my classes need to be rebuilit tho
[11-Jan-2010 11:31:41] <theacolyte-> In fact, that's what it looks like
[11-Jan-2010 11:31:52] <willwh> morning folks
[11-Jan-2010 11:34:18] <theacolyte-> A quick wmiadap /f fixed it
[11-Jan-2010 11:35:09] <rmatte> yeh, wmiadap /f is a nice command :)
[11-Jan-2010 11:35:28] <theacolyte-> I swear, I should just run it on all my systems
[11-Jan-2010 11:35:40] <rmatte> might as well
[11-Jan-2010 11:35:53] <rmatte> hola Nate
[11-Jan-2010 11:36:20] <theacolyte-> I really should also fix my windows memory chart so you can do thresholds
[11-Jan-2010 11:36:28] <rmatte> yup
[11-Jan-2010 11:36:37] <rmatte> you just need total added to it
[11-Jan-2010 11:37:00] <rmatte> then you can add a threshold against used and do something like: here.getRRDValue('Total') * 0.9
[11-Jan-2010 11:37:28] <rmatte> But then you may possibly run in to this: http://dev.zenoss.org/trac/ticket/5978
[11-Jan-2010 11:37:41] <Apachez> mrayzenoss: can I get a trac account please? :)
[11-Jan-2010 11:37:41] <rmatte> Which I've found a rough workaround for (detailed in the ticket)
[11-Jan-2010 11:38:09] <theacolyte-> I hope not, since it's WMI
[11-Jan-2010 11:38:14] <theacolyte-> But I'll let you know :)
[11-Jan-2010 11:38:22] <theacolyte-> Thanks for letting me cheat off your bug so I don't have to ask how to do it though
[11-Jan-2010 11:38:25] <theacolyte-> hehe
[11-Jan-2010 11:38:30] <rmatte> hehe
[11-Jan-2010 11:40:18] <willwh> quick quesiton guys - threshold warnings for throughput on windows devices?
[11-Jan-2010 11:40:22] <willwh> I am only using snmp
[11-Jan-2010 11:40:31] <willwh> ethernetCsmacd should be warning at 75% throughput?
[11-Jan-2010 11:41:28] <rmatte> by default it's 70 I believe
[11-Jan-2010 11:41:48] <rmatte> but it's based on the speed of the port
[11-Jan-2010 11:42:09] <rmatte> so if it's a 100mbit port it won't alert until you're using 75Mbit throughput on the interface
[11-Jan-2010 11:42:17] <rmatte> erm, 70Mbit by default rather
[11-Jan-2010 11:42:33] <RobertLaptop> What does wmiadap /f do?
[11-Jan-2010 11:42:52] <rmatte> RobertLaptop: rebuilds your WMI classes
[11-Jan-2010 11:43:10] <rmatte> RobertLaptop: sometimes you'll find that a class is missing, that'll fix it
[11-Jan-2010 11:43:41] <RobertLaptop> Does that fix those random errors about NT login / status errors?
[11-Jan-2010 11:43:59] <rmatte> depends on what's causing the errors
[11-Jan-2010 11:44:16] <theacolyte-> sometimes/most of the time
[11-Jan-2010 11:44:17] <theacolyte-> hehe
[11-Jan-2010 11:44:18] <rmatte> It's not going to fix WMI NT AUTH errors
[11-Jan-2010 11:44:32] <theacolyte-> but it does fix those NTERROR's I linked above
[11-Jan-2010 11:45:08] <RobertLaptop> OK. I was hoping I am assuming the WMI errors I am see are because the links can see some pretty latency sometimes.
[11-Jan-2010 11:45:25] <theacolyte-> rmatte: It's a shame that when I r edo this memory graph I'm going to lose my historical stuff :(
[11-Jan-2010 11:45:26] <rmatte> speaking of which, I have a windows 2003 server that I'm getting NT_STATUS_ACCESS_DENIED for in Zenoss, but it's using the exact same domain account which is working perfectly fine on it's sister server
[11-Jan-2010 11:45:45] <willwh> rmatte - interesting
[11-Jan-2010 11:45:46] <theacolyte-> DOMAIN\user?
[11-Jan-2010 11:45:50] <rmatte> and it was working previously on this one, just magically stopped working one day
[11-Jan-2010 11:45:53] <willwh> I had 95Mbit going through one 100mbit box
[11-Jan-2010 11:45:54] <rmatte> yeh, domain\user
[11-Jan-2010 11:45:58] <theacolyte-> That should work
[11-Jan-2010 11:46:03] <willwh> according to the Interface report
[11-Jan-2010 11:46:03] <rmatte> I know lol
[11-Jan-2010 11:46:06] <rmatte> that's why I'm stumped
[11-Jan-2010 11:46:08] <theacolyte-> I would try other RPC calls
[11-Jan-2010 11:46:22] <theacolyte-> Like try managing that server from another
[11-Jan-2010 11:46:26] <theacolyte-> services.msc or something
[11-Jan-2010 11:46:34] <rmatte> theacolyte-: yeh, good idea
[11-Jan-2010 11:47:27] <theacolyte-> Why isn't there just a template view under browse by hehe
[11-Jan-2010 11:49:07] <willwh> rmatte: any ideas on how I troubleshoot that a bit?
[11-Jan-2010 11:49:17] <theacolyte-> rmatte: Is your ticket backwards, or what? :)
[11-Jan-2010 11:49:21] <theacolyte-> http://dev.zenoss.org/trac/ticket/5978
[11-Jan-2010 11:49:30] <theacolyte-> Add an SNMP datasource for total amount of memory and name it Used.
[11-Jan-2010 11:49:31] <theacolyte-> Add an SNMP datasource for total used memory and name it Total.
[11-Jan-2010 11:49:33] <willwh> mm, hang on :)
[11-Jan-2010 11:50:48] <rmatte> theacolyte-: I'm able to log directly in to the server with that account for starters
[11-Jan-2010 11:52:07] <theacolyte-> rmatte: it seems to me you'd want to add a datasource for total amount of memory and name it total, and a total used memory and name it used
[11-Jan-2010 11:52:30] <rmatte> theacolyte-: read the first comment that I left for that ticket
[11-Jan-2010 11:52:52] <theacolyte-> LOL
[11-Jan-2010 11:53:02] <rmatte> ;)
[11-Jan-2010 11:53:05] <theacolyte-> It's not my fault, that part didn't show on my screen!
[11-Jan-2010 11:53:11] <theacolyte-> I had to scroll down!
[11-Jan-2010 11:53:50] <rmatte> hehe
[11-Jan-2010 11:54:23] <theacolyte-> I am my own worst enemy
[11-Jan-2010 11:54:25] <theacolyte-> you know that
[11-Jan-2010 11:55:53] <rmatte> ;)
[11-Jan-2010 11:55:59] <rmatte> man I hate WMI sometimes
[11-Jan-2010 11:56:07] <theacolyte-> I hear you.
[11-Jan-2010 11:56:45] <rmatte> I'm waiting to login to this sister server so that I can try doing remote management to see if it works
[11-Jan-2010 11:56:57] <rmatte> though remote management doesn't rely on WMI
[11-Jan-2010 11:57:07] <rmatte> and if it is a login issue, wouldn't cached credentials handle that?
[11-Jan-2010 11:58:02] <theacolyte-> rmatte: Not remote management, no
[11-Jan-2010 11:58:21] <theacolyte-> I would just test to see if you can remotely connect
[11-Jan-2010 11:58:31] <rmatte> yeh, working on it lol
[11-Jan-2010 11:58:36] <rmatte> if this other server ever lets me login
[11-Jan-2010 11:58:42] <rmatte> stuck at Loading your personal settings...
[11-Jan-2010 11:58:47] <theacolyte-> hehe
[11-Jan-2010 11:58:49] * rmatte huffs
[11-Jan-2010 11:59:07] <theacolyte-> Trying to find the right classes for memory
[11-Jan-2010 12:01:29] <theacolyte-> I thinkI really love trying to find classes
[11-Jan-2010 12:01:37] <theacolyte-> - I think
[11-Jan-2010 12:01:43] <rmatte> lol
[11-Jan-2010 12:02:10] <theacolyte-> There may have a ctually been a reason why I went with available memory
[11-Jan-2010 12:02:16] <theacolyte-> hehehe
[11-Jan-2010 12:03:08] <willwh> ah guys - think I've been stupid
[11-Jan-2010 12:03:12] <willwh> if I lock interfaces from updates
[11-Jan-2010 12:03:20] <willwh> they won't throw warnings when thresholds are hit
[11-Jan-2010 12:03:22] <willwh> ?
[11-Jan-2010 12:03:40] <theacolyte-> Are you just trying to remove the threshold?
[11-Jan-2010 12:03:41] <rmatte> theacolyte-: well, if you get total you can do some RPN magic and stuff to show used rather than available
[11-Jan-2010 12:04:02] <theacolyte-> I found used
[11-Jan-2010 12:04:14] <willwh> theacolyte-: no I want to make sure I am getting notifications of interfaces over 70/75% utilization
[11-Jan-2010 12:04:15] <rmatte> willwh: if you lock them from updates yes they will, as long as they are in monitored status
[11-Jan-2010 12:04:21] <willwh> huh, weird
[11-Jan-2010 12:04:26] <willwh> I am not sure what happened here then
[11-Jan-2010 12:04:37] <rmatte> when they say "updates" they are referring to modelling, not polling
[11-Jan-2010 12:04:40] <willwh> 100mbit box @ 95Mbps over the weekend
[11-Jan-2010 12:04:42] <willwh> ah ok
[11-Jan-2010 12:04:51] <willwh> and I didn't get any notifications
[11-Jan-2010 12:05:03] <rmatte> well, you keep saying box
[11-Jan-2010 12:05:07] <rmatte> you mean interface?
[11-Jan-2010 12:05:49] <theacolyte-> rmatte: haha, it looks like even egor's template does availablekbytes
[11-Jan-2010 12:06:00] <theacolyte-> that means this may be worse than I had even imagined
[11-Jan-2010 12:06:28] <rmatte> theacolyte-: perhaps, yet another reason why I'm not a fan of WMI monitoring, things that should be simple are not, then again Microsoft really screwed up their SNMP agent too
[11-Jan-2010 12:06:45] <willwh> rmatte: sorry I mean interface
[11-Jan-2010 12:06:59] <willwh> yes, a 100mbit interface on one of my servers hit 95Mbps
[11-Jan-2010 12:07:06] <willwh> it is a monitored interface
[11-Jan-2010 12:07:13] <willwh> argh phone, back asap
[11-Jan-2010 12:10:11] <theacolyte-> damn, it's 2 classes
[11-Jan-2010 12:17:18] <rmatte> willwh: weird that it didn't threshold then, you've checked the template on the interface to make sure that the threshold is enabled?
[11-Jan-2010 12:17:25] <theacolyte-> Win32_PhysicalMemory\Capacity
[11-Jan-2010 12:17:40] <theacolyte-> and CommittedBytes FROM Win32_PerfFormattedData_PerfOS_Memory
[11-Jan-2010 12:17:47] <rmatte> stupid that they didn't roll it in to the same class
[11-Jan-2010 12:17:59] <theacolyte-> god forbid, right?
[11-Jan-2010 12:18:03] <rmatte> but at least you know what to poll now
[11-Jan-2010 12:18:17] <Kristopher424> How can I set custom descriptions for OIDs? "snmp trap room-laert-11e-snmp-trap" to say "Commericial Power Off" for the certain OID.
[11-Jan-2010 12:18:25] <rmatte> yeh, Microsoft really had their heads up their ***es when they designed this stuff
[11-Jan-2010 12:18:47] <rmatte> Kristopher424: map the event to and event class, then create an event transform
[11-Jan-2010 12:18:56] <rmatte> Kristopher424: the transform would just be 1 line in that case...
[11-Jan-2010 12:19:06] <rmatte> evt.summary = 'Commercial Power Off'
[11-Jan-2010 12:19:22] <Kristopher424> Whoa, lost me already. How do I add the event to an event class?
[11-Jan-2010 12:19:32] <rmatte> ok, do you have the event sitting in your event console?
[11-Jan-2010 12:19:36] <Kristopher424> Yes.
[11-Jan-2010 12:19:43] <rmatte> ok, you're using 2.5.x I assume?
[11-Jan-2010 12:19:50] <Kristopher424> Yessir.
[11-Jan-2010 12:20:08] <rmatte> highlight the event, you'll see a button in the top right of the console that looks like an upside down candlestick
[11-Jan-2010 12:20:14] <rmatte> so highlight the event and click that
[11-Jan-2010 12:20:20] <rmatte> then select an event class to map it to
[11-Jan-2010 12:20:29] <rmatte> If there's no suitable class you can create on
[11-Jan-2010 12:20:32] <rmatte> one*
[11-Jan-2010 12:20:40] <rmatte> If you want to create on, click Events in the left hand menu
[11-Jan-2010 12:20:47] <rmatte> then create a new organizer wherever you want
[11-Jan-2010 12:21:05] <rmatte> for example, Cisco syslog traps would go in /Events/Cisco/Syslog or whatever
[11-Jan-2010 12:21:16] <rmatte> so create organizers that make sense for the trap
[11-Jan-2010 12:21:20] <rmatte> then drop the mapping in there
[11-Jan-2010 12:21:34] <rmatte> sorry, I meant a button in the top left
[11-Jan-2010 12:21:42] <rmatte> (doing this off the top of my head right now lol)
[11-Jan-2010 12:22:07] <Kristopher424> Ahh okay.
[11-Jan-2010 12:22:14] <rmatte> once you have the mapping, navigate to it, click on it
[11-Jan-2010 12:22:17] <rmatte> then click on the edit tab
[11-Jan-2010 12:22:21] <rmatte> you'll see a bunch of text fields
[11-Jan-2010 12:22:26] <rmatte> one of them will say transform
[11-Jan-2010 12:22:36] <rmatte> put the line that I mentioned before in to that and click save
[11-Jan-2010 12:22:42] <rmatte> then click on the zProperties tab
[11-Jan-2010 12:22:51] <rmatte> and from that tab, set the severity that you want it to come in as
[11-Jan-2010 12:22:54] <rmatte> and then you're done
[11-Jan-2010 12:23:49] <rmatte> it's simple once you get used to it ;)
[11-Jan-2010 12:24:08] <rmatte> transforms are really powerful too, you can do a lot more than that with them, that's just a simple one
[11-Jan-2010 12:25:25] <Kristopher424> Sweet, yo. Can I rename that Event to "Transform Comm Power Off / RmAlert" so it's easier to find?
[11-Jan-2010 12:27:42] <rmatte> rename the event mapping you mean?
[11-Jan-2010 12:28:52] <willwh> rmatte: found the issue - it's just that that threshold only throws a severity of "warning"
[11-Jan-2010 12:29:01] <willwh> I only have email notifications for 4+
[11-Jan-2010 12:29:01] <willwh> :)
[11-Jan-2010 12:29:37] <willwh> all sorted now; thank you!
[11-Jan-2010 12:30:14] <rmatte> np
[11-Jan-2010 12:30:47] <rmatte> Kristopher424: if you mean renaming the event mapping, yes, just go back to the edit tab where you set the transform and change the name at the top then save it
[11-Jan-2010 12:30:55] <rmatte> it might not allow you to use the "/" character though
[11-Jan-2010 12:33:47] * theacolyte- sits and waits impatiently for his graphs to draw
[11-Jan-2010 12:34:19] <theacolyte-> I still don't know how you can have non-whole numbers for stuff like queue lengths
[11-Jan-2010 12:34:28] <theacolyte-> How can I have a queue length of 1.21?!
[11-Jan-2010 12:34:36] <theacolyte-> Is it 1? is it 2?
[11-Jan-2010 12:46:41] <rmatte> just change your format string in the graph
[11-Jan-2010 12:46:51] <rmatte> format is %5.2lf%s by default
[11-Jan-2010 12:46:59] <rmatte> change it to %5.0lf
[11-Jan-2010 12:47:01] <rmatte> and problem solved
[11-Jan-2010 12:47:25] <rmatte> I don't know how Zenoss winds up with fractions though
[11-Jan-2010 12:47:49] <rmatte> but once the format string is modified it won't matter
[11-Jan-2010 12:48:06] <theacolyte-> Man this is really really annoying
[11-Jan-2010 12:48:10] <theacolyte-> This RRD stuff keeps happening
[11-Jan-2010 12:48:18] <rmatte> ?
[11-Jan-2010 12:48:19] <theacolyte-> all the old RRD's still exist for all my machines
[11-Jan-2010 12:48:23] <theacolyte-> None of the new RRDs are there
[11-Jan-2010 12:48:30] <theacolyte-> I deleted the graph, and recreated a new one
[11-Jan-2010 12:48:35] <theacolyte-> (new name this time)
[11-Jan-2010 12:49:03] <rmatte> I've honestly never had the problems that you're describing, but I'm not using 2.5.x, so it might be a new bug or something
[11-Jan-2010 12:49:28] <theacolyte-> This is seriously making it completely unusable
[11-Jan-2010 12:49:51] <theacolyte-> And I just noticed this server has no interfaces listed
[11-Jan-2010 12:50:44] <theacolyte-> The RRD one is the more serious issue
[11-Jan-2010 12:50:50] <theacolyte-> I can't keep deleting and readding devices
[11-Jan-2010 12:54:07] <theacolyte-> ah I see why
[11-Jan-2010 12:54:15] <theacolyte-> pythonThresholdException: User-supplied Python expression (here.getRRDValue('TotalMemory') * 0.9) for maximum value caused error: ['Memory_CommitedBytes']
[11-Jan-2010 12:54:42] <theacolyte-> Hehehehehe
[11-Jan-2010 12:57:05] <theacolyte-> If that's really what's causing the broken image, then there needs to be ab etter handler/indication for it :P
[11-Jan-2010 12:57:58] <theacolyte-> Not sure why it didn't delete the old RRD's though...
[11-Jan-2010 13:06:08] <theacolyte-> Well, I see the error, but no idea how to fix it since the error doesn't mean anything to me
[11-Jan-2010 13:06:11] <theacolyte-> ERROR RenderServer Failed to generate a graph
[11-Jan-2010 13:10:32] * theacolyte- throws up his hands
[11-Jan-2010 13:10:34] <theacolyte-> I can't fix this
[11-Jan-2010 13:16:37] <rmatte> lol
[11-Jan-2010 13:16:39] RobertLaptop_ is now known as RobertLaptop
[11-Jan-2010 13:16:46] <rmatte> that's the spirit!
[11-Jan-2010 13:17:10] <theacolyte-> Man, I literally can't figure htis out
[11-Jan-2010 13:17:18] <theacolyte-> it's a combination of a bunch of issues
[11-Jan-2010 13:17:39] <theacolyte-> RRD files aren't being created, I'm getting that error which is causing the RRD's to not be generated
[11-Jan-2010 13:17:56] <theacolyte-> At the same time, the old graph that I deleted - the RRD files still exist.
[11-Jan-2010 13:18:15] <rmatte> RRD files are not based on graphs
[11-Jan-2010 13:18:19] <rmatte> they are based on datasources
[11-Jan-2010 13:18:28] <theacolyte-> 2010-01-11T11:12:29 ERROR RenderServer Failed to generate a graph
[11-Jan-2010 13:18:28] <theacolyte-> Traceback (most recent call last):
[11-Jan-2010 13:18:28] <theacolyte-> File "/usr/local/zenoss/zenoss/Products/ZenRRD/RenderServer.py", line 190, in render
[11-Jan-2010 13:18:32] <theacolyte-> rrdtool.graph(*gopts)
[11-Jan-2010 13:18:32] <rmatte> the actual graph itself has nothing to do with it
[11-Jan-2010 13:19:56] <theacolyte-> The symptom is the graphs don't display.
[11-Jan-2010 13:20:40] <rmatte> and you're sure the datasources are good?
[11-Jan-2010 13:20:48] <theacolyte-> 100%.
[11-Jan-2010 13:20:59] <rmatte> then open a trac case, sounds buggy to me
[11-Jan-2010 13:21:12] <theacolyte-> Tested via both Zenoss and from windows itself
[11-Jan-2010 13:21:13] <rmatte> I've never had an issue like that, although I'm wondering if it's WMI Data Source specific
[11-Jan-2010 13:21:15] <theacolyte-> it's for all of my machines
[11-Jan-2010 13:21:20] <theacolyte-> that have this template applied
[11-Jan-2010 13:21:27] <rmatte> it's happening with WMI Data Sources...
[11-Jan-2010 13:21:32] <theacolyte-> yes
[11-Jan-2010 13:21:38] <rmatte> can you reproduce the same issue with SNMP Data Sources?
[11-Jan-2010 13:21:48] <theacolyte-> I don't even know if you can get this info from snmp
[11-Jan-2010 13:21:54] <rmatte> It could be a bug in Egor's pack rather than in Zenoss
[11-Jan-2010 13:22:06] <rmatte> no, I mean just grab any info, no matter what
[11-Jan-2010 13:22:16] <theacolyte-> this WMI stuff is a freaki nightmare
[11-Jan-2010 13:22:16] <rmatte> and try to get the same RRD behaviour
[11-Jan-2010 13:22:22] <theacolyte-> I think I can, since it works fine for interfaces
[11-Jan-2010 13:22:34] <rmatte> ok
[11-Jan-2010 13:22:47] <rmatte> then you'll want to speak with Egor next time he's around and ask him if he might know what's causing that
[11-Jan-2010 13:23:07] <theacolyte-> Any particular time he's around? I haven't seen him say anything since I joined last month
[11-Jan-2010 13:23:08] <rmatte> the zenperfwmi daemon probably isn't fully handling the RRDs properly
[11-Jan-2010 13:23:38] <rmatte> he pops in once in a while, best way to get in touch with him though is to send him a message on the community site
[11-Jan-2010 13:26:44] <theacolyte-> I may just scrap WMI entirely
[11-Jan-2010 13:26:49] <theacolyte-> This is much too annoying to deal with
[11-Jan-2010 13:26:55] <theacolyte-> I know why you don't do it now
[11-Jan-2010 13:27:06] <theacolyte-> Although to be honest, without WMI zenoss is of limited use
[11-Jan-2010 13:27:15] <theacolyte-> (for me, since I'm 100% windows)
[11-Jan-2010 13:27:47] <RobertLaptop> 100% windows man your days must suck.
[11-Jan-2010 13:28:01] <theacolyte-> I'm sorry?
[11-Jan-2010 13:29:10] <rmatte> hehe
[11-Jan-2010 13:29:34] <rmatte> well, SNMP is fine for basic monitoring
[11-Jan-2010 13:29:48] <rmatte> but when you get in to monitoring stuff like exchange then you need WMI
[11-Jan-2010 13:30:08] <rmatte> WMI is just too unreliable in my opinion, they need to fix the damn thing
[11-Jan-2010 13:30:24] <theacolyte-> The problem now becomes, I'm so invested timewise into this product that I can
[11-Jan-2010 13:30:29] <theacolyte-> can't back out
[11-Jan-2010 13:30:30] <rmatte> It'll just magically stop working out of the blue, which rarely happens with SNMP
[11-Jan-2010 13:31:42] <theacolyte-> Hmm
[11-Jan-2010 13:31:48] <theacolyte-> I tried the graph with a working datapoint
[11-Jan-2010 13:31:54] <rmatte> theacolyte-: just tell Egor about the problem you're having, maybe he can fix up the ZenPack and release a new version
[11-Jan-2010 13:32:01] <rmatte> he's usually really good about fixing bugs in his packs
[11-Jan-2010 13:32:07] <rmatte> and it sounds like it's probably a bug with his pack
[11-Jan-2010 13:33:19] <rmatte> That WMI Data Source pack is one of if not the most advanced ZenPack out there, so it's bound to have bugs that pop up here and there.
[11-Jan-2010 13:34:23] <mrayzenoss> if it's any consolation, the Enterprise Windows stuff is always getting updated too, Windows is a moving target
[11-Jan-2010 13:35:01] <RobertLaptop> Yea most of my problems seem to be on windows.
[11-Jan-2010 13:35:02] <theacolyte-> I just don't know what to do
[11-Jan-2010 13:35:07] <theacolyte-> I always seem to run into issues with WMI
[11-Jan-2010 13:35:11] <theacolyte-> SNMP ... never
[11-Jan-2010 13:35:24] <theacolyte-> But, the problem is WMI is a requirement
[11-Jan-2010 13:35:37] <theacolyte-> I sent the PM we'll see what happens
[11-Jan-2010 13:36:18] <rmatte> yeh, WMI's reliability is brutal
[11-Jan-2010 13:36:30] <rmatte> I've had many instanced where WMI just quits working out of the blue
[11-Jan-2010 13:36:41] <rmatte> sometimes I'd have to perform a full reboot of the server to get it working again
[11-Jan-2010 13:36:42] <RobertLaptop> I seems to just get random WMI error's
[11-Jan-2010 13:36:44] <rmatte> huge pain
[11-Jan-2010 13:37:24] <rmatte> luckily our clients are generally happy with just CPU and Memory monitoring
[11-Jan-2010 13:38:20] <theacolyte-> well
[11-Jan-2010 13:38:21] <rmatte> mrayzenoss: I'm going to take a stab at writing a collector plugin for my ZenPacks rather than doing stand-alone python scripts
[11-Jan-2010 13:38:25] <theacolyte-> I may not have to deal with it much longer
[11-Jan-2010 13:38:30] <theacolyte-> Looks like I may be getting a new job
[11-Jan-2010 13:38:34] <rmatte> mrayzenoss: lord knows how long it'll take me
[11-Jan-2010 13:38:57] <rmatte> theacolyte-: well, that's convenient lol
[11-Jan-2010 13:39:02] <mrayzenoss> theacolyte-: congratulations?
[11-Jan-2010 13:39:02] <theacolyte-> hahahaha
[11-Jan-2010 13:39:07] <theacolyte-> Nah, not yet, I'm interviewing
[11-Jan-2010 13:39:11] <theacolyte-> That was more of a joke than anything
[11-Jan-2010 13:39:15] <rmatte> hehe
[11-Jan-2010 13:39:25] <theacolyte-> May be easier to find a job than to fix this though
[11-Jan-2010 13:39:29] <mrayzenoss> be sure to list your old job on zenoss-jobs :p
[11-Jan-2010 13:39:31] <theacolyte-> I wish K new python
[11-Jan-2010 13:39:38] <theacolyte-> I knew python rather
[11-Jan-2010 13:39:52] <rmatte> theacolyte-: I didn't know any at all until I started using Zenoss, just learn it
[11-Jan-2010 13:39:52] <theacolyte-> I wouldn't suggest this position to my worst enemey
[11-Jan-2010 13:39:58] <rmatte> theacolyte-: it's honestly not that hard
[11-Jan-2010 13:40:10] <theacolyte-> Problem is my mind is not programming oriented
[11-Jan-2010 13:40:23] <rmatte> theacolyte-: so you've never done any kind of scripting?
[11-Jan-2010 13:40:25] <theacolyte-> I guess I just had a really negative experience tryign to learn early java in college
[11-Jan-2010 13:40:31] <theacolyte-> just powershell
[11-Jan-2010 13:40:35] <rmatte> ah
[11-Jan-2010 13:40:52] <rmatte> I had lots of bash scripting experience going in to learning Python, so I guess I had a leg up
[11-Jan-2010 13:41:06] <theacolyte-> The only real scripting in Windows until powershell was vbscript
[11-Jan-2010 13:41:09] <rmatte> I've also coded in like 10 other languages in the past
[11-Jan-2010 13:41:13] <theacolyte-> But I stayed away because it was essentially programming
[11-Jan-2010 13:41:18] <theacolyte-> Well, you're cheating then!
[11-Jan-2010 13:41:20] <rmatte> vbscript is very powerful
[11-Jan-2010 13:41:29] <theacolyte-> It is, but I may as well of learned C#
[11-Jan-2010 13:41:31] <theacolyte-> or C++
[11-Jan-2010 13:41:38] <rmatte> when I say coded in 10 other languages, I don't mean particularily well lol
[11-Jan-2010 13:41:52] <rmatte> bash was the first language which I can say that I truly mastered
[11-Jan-2010 13:42:01] <theacolyte-> I decided to focus more on the server stuff, ask me about esx/vsphere/exchange/mssql, etc
[11-Jan-2010 13:42:05] <rmatte> my Python is improving but I have tons to learn still
[11-Jan-2010 13:42:12] <theacolyte-> I can configure an ASA firewall, a router, switch, etc
[11-Jan-2010 13:42:26] <theacolyte-> In otherwords, too much to learn to focus on scripting hehe
[11-Jan-2010 13:42:28] <rmatte> yeh, I've done all that stuff too
[11-Jan-2010 13:42:46] <theacolyte-> I may just start crying
[11-Jan-2010 13:42:46] <rmatte> I went to school for network/server administration
[11-Jan-2010 13:42:52] <rmatte> I just happen to have a nack for scripting
[11-Jan-2010 13:43:01] <theacolyte-> My boss will expect this working when he comes back tomorrow, if it's not, I'm in for a fun conversation hehe
[11-Jan-2010 13:43:10] <theacolyte-> ah well
[11-Jan-2010 13:43:34] <theacolyte-> Ever seen it when no interfaces are detected ... on a windows host?
[11-Jan-2010 13:43:35] <rmatte> well, do what you can at least and if there are a few things missing you can add on later
[11-Jan-2010 13:43:41] <theacolyte-> yeah exactly
[11-Jan-2010 13:43:46] <rmatte> theacolyte-: only when not using SNMP
[11-Jan-2010 13:43:47] <theacolyte-> focusing on up/down until I can figure this out
[11-Jan-2010 13:43:56] <rmatte> theacolyte-: you can't detect interfaces via WMI in Zenoss core
[11-Jan-2010 13:44:41] <rmatte> Interfaces and disks need to be done via SNMP
[11-Jan-2010 13:44:52] <theacolyte-> Weird, it's detecting it on one of my servers, but not all
[11-Jan-2010 13:45:03] <rmatte> make sure your SNMP permissions are correct
[11-Jan-2010 13:45:27] <theacolyte-> are interfaces in device [devices/server/windows0?
[11-Jan-2010 13:45:28] <rmatte> eugh, I have a meeting to go to in 10 mins (yay)
[11-Jan-2010 13:45:38] <rmatte> hunh?
[11-Jan-2010 13:45:44] <rmatte> the interfaces would show up on the OS tab
[11-Jan-2010 13:45:46] <theacolyte-> I thought they were in ethernetCsmacd, which is alwasys enabled
[11-Jan-2010 13:45:47] <theacolyte-> right
[11-Jan-2010 13:45:59] <rmatte> it depends on the interface type
[11-Jan-2010 13:46:07] <rmatte> when you click on an interface you'll see: Type: whatever
[11-Jan-2010 13:46:20] <rmatte> if it's ethernetCsmacd then it'll use that template
[11-Jan-2010 13:46:36] <rmatte> if it's something else it'll use that template by default until another template which matches the name is created
[11-Jan-2010 13:46:37] <theacolyte-> Yeah, it's a few of my servers now with no interfaces listed heh
[11-Jan-2010 13:46:42] <rmatte> I have like 20 interface templates
[11-Jan-2010 13:46:56] <rmatte> well, that has to do with modelling not templates
[11-Jan-2010 13:47:02] <rmatte> if there are no interfaces listed remodel the device
[11-Jan-2010 13:47:10] <rmatte> if that doesn't work, snmpwalk it and make sure you get output
[11-Jan-2010 13:47:31] <rmatte> and you should get more than just 30 lines, if you only get a few lines something isn't configured right
[11-Jan-2010 13:47:32] <theacolyte-> hmm yeah I see zenoss.snmp.InterfaceMap
[11-Jan-2010 13:47:46] <theacolyte-> seems like it's stopped at changes in configuration applied
[11-Jan-2010 13:47:58] <rmatte> that's where it always stops
[11-Jan-2010 13:48:08] <rmatte> that means it's done modelling
[11-Jan-2010 13:48:10] <theacolyte-> it's still going though... or at least the page is still loading
[11-Jan-2010 13:48:20] <rmatte> it shouldn't be
[11-Jan-2010 13:48:35] <theacolyte-> lovely, I love being that person where it shouldn't be doing something
[11-Jan-2010 13:48:46] <rmatte> haha
[11-Jan-2010 13:48:54] <rmatte> I've gotta head to a meeting, I'll be back after
[11-Jan-2010 13:48:57] <theacolyte-> alright man
[11-Jan-2010 13:48:57] <rmatte> ttyl
[11-Jan-2010 13:48:59] <theacolyte-> take it easy
[11-Jan-2010 13:49:01] <rmatte> you too
[11-Jan-2010 13:49:03] <rmatte> good luck
[11-Jan-2010 13:49:06] <theacolyte-> hehe
[11-Jan-2010 13:49:09] <rmatte> keep screwing around with it, you'll get it
[11-Jan-2010 13:49:12] <theacolyte-> yeah
[11-Jan-2010 13:49:14] <rmatte> but seriously, snmpwalk the devices
[11-Jan-2010 13:49:18] <theacolyte-> just a question of how much lost hair
[11-Jan-2010 13:49:20] <rmatte> make sure you get adequate output
[11-Jan-2010 14:26:35] <daveborg> Does anyone have much experience with the ShowGraphPortlet zenpack? I am on Zenoss 2.5.1 and trying to use version 1.01 of the zenpack. When I try to install it, I get and error saying [Errno 20] Not a directory: '/home/zenoss/Products/ZenPacks.SCC.ShowGraphPortlet-1.01-py2.4.egg/skins'. Any ideas?
[11-Jan-2010 14:27:59] <Apachez> mrayzenoss: tnx for the account :)
[11-Jan-2010 14:28:26] <Apachez> mrayzenoss: is it hard to update and set those tickets I created as "zenoss" to be set to have been created by my account instead ?
[11-Jan-2010 14:28:31] <Apachez> I can bring you the ip
[11-Jan-2010 14:36:12] <rmatte> I'm back
[11-Jan-2010 14:37:48] <rmatte> daveborg: is the file a .egg.zip?
[11-Jan-2010 14:38:18] <daveborg> rmatte: it is.
[11-Jan-2010 14:38:45] <rmatte> daveborg: yeh, jive, the software that Zenoss uses for the community site auto-zips files when they are uploaded... unzip it then install it
[11-Jan-2010 14:39:01] <rmatte> some old ZenPacks are zips but they will just be .zip, any new ZenPacks are always .egg
[11-Jan-2010 14:39:10] <rmatte> so If you see .egg.zip unzip the pack before installing
[11-Jan-2010 14:39:18] <daveborg> I did not know that. Good to know for the future.
[11-Jan-2010 14:39:24] <daveborg> Thanks
[11-Jan-2010 14:39:28] <rmatte> np
[11-Jan-2010 14:39:49] <theacolyte-> I am reasonably sure I fixed it
[11-Jan-2010 14:39:53] <theacolyte-> But it's weird that it's the problem
[11-Jan-2010 14:40:00] <rmatte> It says it at the top of the ZenPack page in big bold letters, maybe they need to make it flash or something lol
[11-Jan-2010 14:40:11] <theacolyte-> (fixed the interfaces one)
[11-Jan-2010 14:40:18] <rmatte> theacolyte-: what was the issue?
[11-Jan-2010 14:40:25] <theacolyte-> v1 vs v2c
[11-Jan-2010 14:40:31] <theacolyte-> need to run v2c
[11-Jan-2010 14:40:31] <rmatte> oh
[11-Jan-2010 14:40:42] <rmatte> that's weird, I've never had to do that with windows snmp
[11-Jan-2010 14:40:42] <theacolyte-> don't say you knew that
[11-Jan-2010 14:40:43] <theacolyte-> haha
[11-Jan-2010 14:40:54] <rmatte> it just works
[11-Jan-2010 14:41:09] <theacolyte-> testing it on another one of my servers to see if it was a fluke first
[11-Jan-2010 14:41:20] <rmatte> whatever works though
[11-Jan-2010 14:41:35] <theacolyte-> man... that's weird
[11-Jan-2010 14:41:36] <theacolyte-> that's what it was
[11-Jan-2010 14:41:46] <theacolyte-> I wonder
[11-Jan-2010 14:41:51] <theacolyte-> Oh you know I bet you I know what it is
[11-Jan-2010 14:42:05] <theacolyte-> these have the HP agents installe,d I bet you dollars to donuts it changes it to v2
[11-Jan-2010 14:42:50] <theacolyte-> But, as you said, if it fixes it, who cares
[11-Jan-2010 14:43:01] <rmatte> that may be what's causing it
[11-Jan-2010 14:44:02] <rmatte> Well, our custom CMDB got rolled out today, seems to be working well enough, still lots of kinks to work out of it
[11-Jan-2010 14:44:48] <theacolyte-> I am clawing my way up the zenoss ladder
[11-Jan-2010 14:44:53] <rmatte> hehe
[11-Jan-2010 14:44:57] <rmatte> it's the only way to do it
[11-Jan-2010 14:45:54] <theacolyte-> I'm really worried about that WMI thing though
[11-Jan-2010 14:45:57] <theacolyte-> That could be a gamechanger
[11-Jan-2010 14:46:01] <theacolyte-> Although it's up to my boss really
[11-Jan-2010 14:46:11] <theacolyte-> I'm so glad I stepped down from management
[11-Jan-2010 14:46:23] <theacolyte-> best move I ever made
[11-Jan-2010 14:46:40] <rmatte> hehe
[11-Jan-2010 14:47:18] <theacolyte-> I have servers that don't have snmpinformant that are showing free space on my drives, but they are vmware guests - I bet you vmware tools makes those OID's available
[11-Jan-2010 14:48:14] <rmatte> you don't need snmp informant for disk stats
[11-Jan-2010 14:48:20] <theacolyte-> I thought you did?
[11-Jan-2010 14:48:26] <rmatte> nope
[11-Jan-2010 14:48:37] <rmatte> just CPU and Memory, which my ZenPack handles without having to use informant
[11-Jan-2010 14:48:46] <theacolyte-> Sad part is
[11-Jan-2010 14:48:50] <theacolyte-> I'm going to end up using your zenpack
[11-Jan-2010 14:48:54] <theacolyte-> Instead of mine
[11-Jan-2010 14:49:11] <theacolyte-> My EqualLogic zenpack is solid though - if I can figure out that transparency stuff I'll release it
[11-Jan-2010 14:49:33] <rmatte> well, the current one that's publicly available isn't as good as my updated version, but I'm not releasing a new one until I've improved the performance
[11-Jan-2010 14:49:42] <theacolyte-> I keep hearing that... release it already!
[11-Jan-2010 14:49:42] <rmatte> but it'll do basically what you need it to for the time being
[11-Jan-2010 14:50:05] <theacolyte-> I haven't played with this yet, so don't judgement yet
[11-Jan-2010 14:50:06] <theacolyte-> hehe
[11-Jan-2010 14:50:06] <rmatte> :P
[11-Jan-2010 14:50:10] <theacolyte-> But...
[11-Jan-2010 14:50:20] <rmatte> can't release it yet, have lots of work to do on it
[11-Jan-2010 14:50:21] <theacolyte-> I have a graph that's supposed to be 2.10T that's showing up as 2.10M
[11-Jan-2010 14:50:27] <rmatte> how many servers are you going to be monitoring with it?
[11-Jan-2010 14:50:46] <Kristopher424> rmatte, On our old NMS, the NMS got the description and information from the trap automatically... I'm wondering why Zenoss is showing "snmp-room-monitor-snmp-11e" instead of the same description our old NMS had and I'm having to add transformers for them.
[11-Jan-2010 14:51:16] <theacolyte-> rmatte: 30 or so
[11-Jan-2010 14:51:27] <rmatte> Kristopher424: because your old NMS probably had pre-made scripts that did that
[11-Jan-2010 14:51:44] <rmatte> Kristopher424: Zenoss comes with limited pre-made stuff, it's more of a framework to develop your own customizations around
[11-Jan-2010 14:51:55] <theacolyte-> I wish you could add a column for location in device list
[11-Jan-2010 14:52:02] <theacolyte-> I'm going to put in an enhancement request
[11-Jan-2010 14:52:13] <Kristopher424> Oh, okay. I thought traps pretty much worked the same way... if the trap had a descrip. in it, zenoss would show it.
[11-Jan-2010 14:52:38] <rmatte> Kristopher424: no, traps are not even close to as simple as you're making them out to be lol
[11-Jan-2010 14:52:50] <rmatte> Kristopher424: different traps provide different properties with different names
[11-Jan-2010 14:53:11] <rmatte> Kristopher424: then you have to create an event mapping and transform to grab those properties and display them the way you would live when the event comes in
[11-Jan-2010 14:53:40] <rmatte> Kristopher424: some traps will come in with a property which defines the severity of the event, so you can have your transform set the actual severity of the event based on that when it comes in
[11-Jan-2010 14:54:15] <rmatte> Kristopher424: or it might come in with just a number which represents a certain type of event, like 3 could mean cable has been disconnected, 5 could mean power supply fault, and you need to write a transform to translate those
[11-Jan-2010 14:54:18] <rmatte> stuff like that
[11-Jan-2010 14:55:03] <rmatte> If you double click on the trap event to open the details window and expand the details list you'll see the trap properties at the bottom of the list
[11-Jan-2010 14:55:16] <Kristopher424> Gotchya. I know you showed me how to map an event to a tranformer earlier, can I make a transformer based off of a manually typed in OID? That way I don't have to trigger every alarm in order to transform them?
[11-Jan-2010 14:56:04] <rmatte> well... that's where it gets tricky
[11-Jan-2010 14:56:26] <rmatte> some traps will come in with a fully translated name because you have the Mibs loaded
[11-Jan-2010 14:56:43] <rmatte> some will come in with a partially translated name like: enterprise.1.0.23.40.10
[11-Jan-2010 14:57:05] <rmatte> very rarely you'll get one come in as snmp trap <oid>

[11-Jan-2010 14:57:20] <rmatte> depending on what the traps look like when they come in, that's what the mapping has to be
[11-Jan-2010 14:57:37] <rmatte> when you create a mapping there's a name for the mapping as well as an EventClassKey
[11-Jan-2010 14:57:46] <rmatte> the key needs to match the eventClassKey of the event when it comes in
[11-Jan-2010 14:57:52] <rmatte> that's how Zenoss associates it
[11-Jan-2010 14:58:09] <Kristopher424> Hmm.
[11-Jan-2010 14:58:15] <Kristopher424> This got all too complicated.
[11-Jan-2010 14:58:28] <rmatte> Zenoss definitely has a learning curve, it takes time
[11-Jan-2010 14:58:46] <Kristopher424> So, even though when you double click the trap it shows the OID, you can't have Zenoss call on the OID to transform it?
[11-Jan-2010 14:58:46] <rmatte> It's very easy to do basic stuff with, but once you get in to the development side of it, well, pull of your sleeves
[11-Jan-2010 14:58:47] <rmatte> lol
[11-Jan-2010 14:59:02] <rmatte> Kristopher424: not if it's translating it using Mibs
[11-Jan-2010 14:59:24] <rmatte> Kristopher424: If you have the Mibs loaded, find the OID in the Mibs and find out what the Mibs are going to make the trap look like when it comes in
[11-Jan-2010 14:59:43] <rmatte> such as snmp-room-monitor-snmp-11e for example
[11-Jan-2010 15:00:33] <Kristopher424> Ok. What is the dedupid?
[11-Jan-2010 15:01:07] <rmatte> that's just some id that Zenoss uses, you don't have to worry about that
[11-Jan-2010 15:01:12] <Kristopher424> ah ok.
[11-Jan-2010 15:01:21] <rmatte> the actual trap values will be near the bottom
[11-Jan-2010 15:01:43] <rmatte> (I asked the developers to split them off in to their own section like it used to be, since it's confusing)
[11-Jan-2010 15:02:04] <Kristopher424> Hmm, too bad I can't just choose what part of the trap I want to show in the description...the dumb thing is sitting right there in alarmmessage1.0
[11-Jan-2010 15:02:04] <rmatte> "monitor" should be the last event value before the trap values
[11-Jan-2010 15:02:12] <rmatte> yuo can
[11-Jan-2010 15:02:16] <rmatte> you can* rather
[11-Jan-2010 15:02:29] <rmatte> instead of evt.summary = 'whatever' you'd do...
[11-Jan-2010 15:03:24] <rmatte> msg = getattr(evt, 'alarmmessage1.0', 'Unknown')
[11-Jan-2010 15:03:24] <rmatte> evt.summary = '%s' % (msg)
[11-Jan-2010 15:03:29] <rmatte> that would be your transform
[11-Jan-2010 15:03:45] <Kristopher424> what is the 'Unknown'?
[11-Jan-2010 15:04:09] <rmatte> If alarmmessage1.0 happens to not exist (which is unlikely) it'll set msg to 'Unknown'
[11-Jan-2010 15:04:20] <rmatte> so your summary will become 'Unknown' instead of being blank
[11-Jan-2010 15:04:35] <rmatte> it's just an alternate value for the message
[11-Jan-2010 15:04:41] <rmatte> you can set it to something else if you want
[11-Jan-2010 15:05:19] <Kristopher424> Gotchya.
[11-Jan-2010 15:05:45] <Kristopher424> So, for some reason that transformer set all the traps to the same event. It says the count is 21.
[11-Jan-2010 15:05:53] <Kristopher424> Can I show all 21 of these individually?
[11-Jan-2010 15:06:14] <rmatte> Ok, I'll explain how Zenoss does things...
[11-Jan-2010 15:06:57] <rmatte> If an event matches a previous event, instead of creating a completely separate event in the database it increments the count on the existing event and increases the lastTime value to the date/time that the last event came in, it's to save space in the database
[11-Jan-2010 15:07:09] <rmatte> so instead of having 1000 entries in the database for the same recurring event you have 1
[11-Jan-2010 15:07:45] <Apachez> rmatte: are the queries used collected in a single file or such?
[11-Jan-2010 15:07:51] <rmatte> so you know when the first event came in and when the last event came in (effectively the span) and you know how many times it occured, but it doesn't take up extra space in the DB
[11-Jan-2010 15:08:10] <Kristopher424> Hmm, okay... I guess that would work.
[11-Jan-2010 15:09:10] <rmatte> Apachez: which queries are you referring to?
[11-Jan-2010 15:10:13] <rmatte> Apachez: If you're talking about the database queries, all I know is that they are done through python functions that are coded in to Zenoss. I use those functions when I'm doing anything to do with events, I don't generally use actual MySQL DB commands
[11-Jan-2010 15:10:34] <rmatte> Apachez: I highly doubt they are all in one spot, they are probably spread out all over the place
[11-Jan-2010 15:12:34] <rmatte> for example, to move an event to history:
[11-Jan-2010 15:12:36] <rmatte> dmd.ZenEventManager.manage_deleteEvents(evt.evid)
[11-Jan-2010 15:12:43] <rmatte> where evt.evid is the event id
[11-Jan-2010 15:13:08] <rmatte> If you wanted to comb through active events:
[11-Jan-2010 15:13:09] <rmatte> for e in dmd.ZenEventManager.getEventList([], "", "lastTime ASC, firstTime ASC"):
[11-Jan-2010 15:13:53] <Kristopher424> Thanks for the help, rmatte
[11-Jan-2010 15:13:57] <rmatte> np
[11-Jan-2010 15:19:43] <Kristopher424> How does the page feature work in Zenoss?
[11-Jan-2010 15:21:57] <rmatte> Kristopher424: all the stuff you've been asking is covered in detail in the admin guide, you really should read it
[11-Jan-2010 15:22:06] <rmatte> http://sourceforge.net/projects/zenoss/files/Documentation/zenoss-2.5.x-docs/Zenoss_Administration_06-102009-2.5-v01.pdf/download
[11-Jan-2010 15:22:29] <rmatte> It'll answer about 90% of the questions that you are liable to ask in here
[11-Jan-2010 15:30:29] <Kristopher424> haha, thanks.
[11-Jan-2010 15:31:40] <rmatte> eugh, finally figured out why utilization threshold alerts are coming in, they are actually error threshold alerts but the transform isn't specific enough
[11-Jan-2010 15:44:47] <Apachez> I tried the other day to create some good indexes for the mysql in zenoss
[11-Jan-2010 15:44:58] <Apachez> holy crap what many kind of queries there were
[11-Jan-2010 15:45:10] <Apachez> in short, hard to setup any good indexes :(
[11-Jan-2010 15:45:29] <rmatte> yeh
[11-Jan-2010 15:45:31] <mrayzenoss> yeah, it sucks. I'm looking at MySQL 5.4 for the next release
[11-Jan-2010 15:45:43] <mrayzenoss> apparently that's the "performance" release
[11-Jan-2010 15:46:06] <mrayzenoss> but there may be too many moving parts right now for that to go in
[11-Jan-2010 15:46:57] <QubeZ> hi all
[11-Jan-2010 15:47:34] <rmatte> well, I have 2 projects planned... coding collector plugins for my windows snmp perf zenpack and taking a stab at coding a zenpack to make setting up remote collectors with zenoss core easier
[11-Jan-2010 15:47:48] <QubeZ> i just tested ssh commands for AsteriskNew zenpack via my shell and it works but when I run the command against my asterisk server via the zenoss webui, it says host key verification failed. Is the zenpack running as zenoss user.. is this the issue?
[11-Jan-2010 15:48:12] <rmatte> QubeZ: it would run it as zenoss user, yes
[11-Jan-2010 15:48:30] <Apachez> any of you who have a large db I can do test on?
[11-Jan-2010 15:48:34] <QubeZ> gotcha, ok thats my issue
[11-Jan-2010 15:48:38] <Apachez> I mean like sending me a copy of backup.sql
[11-Jan-2010 15:49:05] <Apachez> because when I tested with my altered indexes the content is to small to be able to do any real measurements
[11-Jan-2010 15:49:08] <mrayzenoss> QubeZ: looking at the ZenPack, it looks like it's hard-coded to root
[11-Jan-2010 15:49:09] <rmatte> Apachez: I can't, our databases are confidential
[11-Jan-2010 15:49:28] <Apachez> of course im not asking for any confidential dbs :)
[11-Jan-2010 15:49:39] <rmatte> Apachez: those are the only type I have lol
[11-Jan-2010 15:49:54] <Apachez> but there are 57 other people in here ;)
[11-Jan-2010 15:50:06] <rmatte> who probably also have confidential dbs
[11-Jan-2010 15:50:12] <Apachez> pfff
[11-Jan-2010 15:50:14] <rmatte> it's an enterprise product
[11-Jan-2010 15:50:15] <rmatte> lol
[11-Jan-2010 15:50:30] <Apachez> "enterprise" with a dreadfully slow gui :P
[11-Jan-2010 15:50:47] <Apachez> mrayzenoss: why isnt any of my requests proposed for 2.5.2? :P
[11-Jan-2010 15:53:01] <rmatte> Apachez: well, the gui wouldn't be slow if you were running a remote collector and using the 1 box just for the GUI
[11-Jan-2010 15:53:16] <rmatte> Apachez: when the box is doing a polling cycle and using 100% CPU of course the UI is going to be slow
[11-Jan-2010 15:53:44] <rmatte> Apachez: anything else would be running slow on that box during that time as well
[11-Jan-2010 15:57:40] <gwb2352> do other linux ssh based collections show specific traffic for virtual (eth0:1, eth0:2) traffic, or just for eth0 ?
[11-Jan-2010 16:03:41] <mrayzenoss> Apachez: examples?
[11-Jan-2010 16:03:47] <theacolyte-> mrayzenoss: fix all my problems!
[11-Jan-2010 16:03:59] <theacolyte-> hehe
[11-Jan-2010 16:04:11] <mrayzenoss> Apachez: if you need something from a pristine Zenoss install, just grab a VMware image
[11-Jan-2010 16:10:39] <theacolyte-> rmatte: fyi
[11-Jan-2010 16:11:01] <gwb2352> never mind, i see that virtual interfaces don't have traffic counters explicitly associated with them :(
[11-Jan-2010 16:11:23] <theacolyte-> It looks like when you model it doesn't actually do interfaces until after the configuration prompt I was talking about earlier
[11-Jan-2010 16:11:50] <theacolyte-> Do you guys
[11-Jan-2010 16:11:58] <theacolyte-> monitor your zenoss servers, from within zenoss? :)
[11-Jan-2010 16:13:36] <mrayzenoss> rmatte: I've got someone looking at 5473 again
[11-Jan-2010 16:16:58] <kisielk> my zenoss server is soooo slowww... I think I need newer hardware for it :/
[11-Jan-2010 16:18:29] <kisielk> what's the recommended specs for a Zenoss server anyway?
[11-Jan-2010 16:18:37] <theacolyte-> The one thing I will say a bout zenoss is watch for latency
[11-Jan-2010 16:19:21] <rmatte> mrayzenoss: cool, thanks
[11-Jan-2010 16:20:03] <theacolyte-> Nothing like the internet at your monitoring point going down for a second... just got 12380123 emails
[11-Jan-2010 16:20:08] <mrayzenoss> kisielk: there are a couple of popular threads in the forums, but after you have plenty of RAM you'll have to worry about disk I/O
[11-Jan-2010 16:20:36] <kisielk> yeah
[11-Jan-2010 16:20:48] <kisielk> so I'm guessing something fast disks and 8GB of RAM or thereabouts should be good
[11-Jan-2010 16:21:22] <theacolyte-> kisielk: How many devices?
[11-Jan-2010 16:21:36] <kisielk> 300
[11-Jan-2010 16:21:37] <rmatte> theacolyte-: that's a lot of emails lol
[11-Jan-2010 16:21:42] <theacolyte-> rmatte: hehe
[11-Jan-2010 16:21:49] <rmatte> theacolyte-: how many devices are you monitoring?
[11-Jan-2010 16:21:56] <theacolyte-> Just got off on that phone interview, it went great
[11-Jan-2010 16:22:00] <theacolyte-> Right now, or eventually?
[11-Jan-2010 16:22:12] <kisielk> currently running on a dual-proc dual-core Athlon box, with regular SATA disks and 4GB of RAM
[11-Jan-2010 16:22:12] <rmatte> theacolyte-: well, you said you got a ridiculous amount of emails...
[11-Jan-2010 16:22:18] <kisielk> and it is like... really damn slow
[11-Jan-2010 16:22:23] <rmatte> theacolyte-: Zenoss should have noticed everything ping down and stopped monitoring them
[11-Jan-2010 16:22:24] <theacolyte-> rmatte: Ah, I wasn't really being serious... I got something like 20
[11-Jan-2010 16:22:36] <rmatte> hence you should have only seen as many alerts as you have devices
[11-Jan-2010 16:22:40] <rmatte> ah ok
[11-Jan-2010 16:22:44] <rmatte> lol
[11-Jan-2010 16:22:53] <theacolyte-> And it wasn't actually that bad, i wouldn't want to turn it off
[11-Jan-2010 16:22:57] <theacolyte-> Blow me up if that connection goes down
[11-Jan-2010 16:23:00] <rmatte> yeh
[11-Jan-2010 16:23:02] <rmatte> for sure
[11-Jan-2010 16:23:02] <theacolyte-> I want an alarm clock that queues against it
[11-Jan-2010 16:23:05] <theacolyte-> Or an airraid siren
[11-Jan-2010 16:23:19] <rmatte> that's the way we look at it, if a tunnel goes down to one of our clients we'll never miss it
[11-Jan-2010 16:24:50] <rmatte> but we don't get emails, we just get a flood of tickets lol
[11-Jan-2010 16:25:25] <theacolyte-> hehe
[11-Jan-2010 16:25:37] <theacolyte-> I'd be hesitant to have that connection in fear of inundating my poor helpdesk guys
[11-Jan-2010 16:26:06] <theacolyte-> That's actually a hurdle I'm going to have to figure out soon - what to give them access to
[11-Jan-2010 16:27:40] <rmatte> well, our front line guys are NOC staff, so that's all they do all day, they are used to it
[11-Jan-2010 16:27:51] <theacolyte-> That makes sense
[11-Jan-2010 16:28:02] <rmatte> all they do is pick 1 ticket, relate the rest back to it then put whatever update they need in the ticket then close it when the link comes back up
[11-Jan-2010 16:29:51] <theacolyte-> Ah, so your staff does your event correlation? :L)
[11-Jan-2010 16:30:46] <rmatte> for the time being, yeh
[11-Jan-2010 16:30:53] <rmatte> eventually we'll code a layer to do that
[11-Jan-2010 16:31:03] <theacolyte-> I wonder if you can output zenoss events to a syslog server?
[11-Jan-2010 16:31:23] <theacolyte-> output/forward
[11-Jan-2010 16:31:30] <rmatte> why would you want to?
[11-Jan-2010 16:31:36] <theacolyte-> Splunk
[11-Jan-2010 16:31:40] <rmatte> ah
[11-Jan-2010 16:31:51] <theacolyte-> Then you could do real correlation
[11-Jan-2010 16:31:57] <rmatte> you could probably rig up something to do it, but it'd take some dev work
[11-Jan-2010 16:32:52] <kisielk> theacolyte-: do you use splunk?
[11-Jan-2010 16:32:58] <theacolyte-> I bet you it wouldn't even be that difficult
[11-Jan-2010 16:33:02] <theacolyte-> kisielk: yeah
[11-Jan-2010 16:33:09] <kisielk> how's it working for you?
[11-Jan-2010 16:33:27] <theacolyte-> Actually, I should correct that, not at this moment, but I have here (and other locations)
[11-Jan-2010 16:33:31] <rmatte> we really prefer to have the systems suppress whatever they need to then have a pair of human eyes actually analyze the events that make it through
[11-Jan-2010 16:33:32] <theacolyte-> have been using it since it came out off and on
[11-Jan-2010 16:33:35] <theacolyte-> splunk is amazing
[11-Jan-2010 16:33:57] <kisielk> I took a brief look at it, but I was more confused than anything
[11-Jan-2010 16:34:07] <rmatte> lol
[11-Jan-2010 16:34:13] <kisielk> seems like it would take a fair bit of work to get rid of the noise
[11-Jan-2010 16:34:18] <mrayzenoss> speaking of syslog, our QA is looking for suggestions for testing: http://community.zenoss.org/message/44319#44319
[11-Jan-2010 16:34:40] <theacolyte-> kisielk: It's not, the whole point of it is filtering
[11-Jan-2010 16:35:06] <theacolyte-> I may play with zenoss syslogging htough
[11-Jan-2010 16:35:11] <kisielk> well, I only spent like 5 minutes with it
[11-Jan-2010 16:35:13] <rmatte> mrayzenoss: if they can fix this in terms of syslog I would be very happy: http://dev.zenoss.org/trac/ticket/5289
[11-Jan-2010 16:35:25] <theacolyte-> kisielk: the filters can be *VERY* specific
[11-Jan-2010 16:35:45] <theacolyte-> and it's designed to be lcicked around and it's also very fast
[11-Jan-2010 16:36:22] <theacolyte-> Hmm, I wonder if I can monitor my pgp server
[11-Jan-2010 16:37:25] <rmatte> hmmm, my workstation has better uptime than some of our servers :P
[11-Jan-2010 16:37:31] <theacolyte-> lol
[11-Jan-2010 16:37:39] <rmatte> been doing a lot of changes lately though which required reboots
[11-Jan-2010 16:37:47] <rmatte> My workstation is at 69 days right now
[11-Jan-2010 16:38:05] <theacolyte-> wow my pgp server actually has a link on it's interface
[11-Jan-2010 16:38:11] <theacolyte-> Download MIBs...
[11-Jan-2010 16:39:22] <rmatte> lol
[11-Jan-2010 16:39:36] <theacolyte-> Apparently though, Zenoss odesn'
[11-Jan-2010 16:39:42] <theacolyte-> seem to have an upload mibs button (
[11-Jan-2010 16:40:00] <rmatte> well, if you install the mib browser zenpack it does
[11-Jan-2010 16:40:09] <rmatte> but you can also just install them from commandline
[11-Jan-2010 16:40:11] <theacolyte-> only works with v1 (!!!)
[11-Jan-2010 16:40:16] <theacolyte-> yeah reading how to do that now
[11-Jan-2010 16:40:26] <mrayzenoss> rmatte: I'll look into getting that targetted for 2.5.2, in the meantime: http://dev.zenoss.org/trac/ticket/5898
[11-Jan-2010 16:40:28] <rmatte> zenmib run <mib filename>

[11-Jan-2010 16:41:01] <rmatte> mrayzenoss: that's a superb idea
[11-Jan-2010 16:41:18] <rmatte> mrayzenoss: zenhub is already busy enough as it is
[11-Jan-2010 16:41:18] <theacolyte-> Oh hey mrayzenoss
[11-Jan-2010 16:41:21] <mrayzenoss> there's another ticket for eventlog
[11-Jan-2010 16:41:27] <rmatte> sweet
[11-Jan-2010 16:41:29] <mrayzenoss> but I think it's Stone Crab
[11-Jan-2010 16:41:47] <theacolyte-> Don't suppose you know anything about transparancies/alpha channels and area graphs?
[11-Jan-2010 16:42:02] <theacolyte-> Before I file a bug
[11-Jan-2010 16:42:21] <rmatte> wonder if that would apply cleanly to 2.4.5 >:)
[11-Jan-2010 16:42:33] <rmatte> I'll test tomorrow
[11-Jan-2010 16:42:47] <rmatte> my 2.4.5 boxes are so patched up lol
[11-Jan-2010 16:43:00] <theacolyte-> patched enough where you can't ever upgrade? :P
[11-Jan-2010 16:43:07] <mrayzenoss> beauty of open source
[11-Jan-2010 16:44:24] <rmatte> theacolyte-: no, upgrading would work fine
[11-Jan-2010 16:44:38] <theacolyte-> can you run zenmib as root?
[11-Jan-2010 16:44:43] <theacolyte-> or do you have to run it as zenoss?
[11-Jan-2010 16:44:45] <rmatte> theacolyte-: run it as the zenoss user
[11-Jan-2010 16:44:51] <rmatte> theacolyte-: don't run any zenoss stuff as root
[11-Jan-2010 16:45:40] <rmatte> theacolyte-: the first time I used zenmib I ran it as root but it was a pain in the ass since I had to do all sorts of imports and setting env variables and stuff which I later found out could be avoided by using the zenoss user
[11-Jan-2010 16:45:42] <rmatte> ;)
[11-Jan-2010 16:45:49] <theacolyte-> lol
[11-Jan-2010 16:45:50] <theacolyte-> word.
[11-Jan-2010 16:47:01] <rmatte> hopefully 2.5.2 is going to be a big step above 2.5.1 then I can consider upgrading
[11-Jan-2010 16:47:08] <theacolyte-> oh, I see OIDs to monitor!
[11-Jan-2010 16:47:17] <rmatte> lol
[11-Jan-2010 16:47:24] <rmatte> in the Mib?
[11-Jan-2010 16:47:31] <theacolyte-> yeah
[11-Jan-2010 16:47:37] <rmatte> cool
[11-Jan-2010 16:48:16] <kisielk> I hope it will fix all the crap I'm dealing with
[11-Jan-2010 16:48:38] <theacolyte-> Only an act of god can fix this WMI issue :(
[11-Jan-2010 16:49:34] <theacolyte-> mrayzenoss: http://community.zenoss.org/thread/12469?tstart=0 this is blocking my zenpack release - do you know if that's a bug or not?
[11-Jan-2010 16:54:59] <rmatte> This is blocking my next zenpack releases: http://dev.zenoss.org/trac/ticket/5978
[11-Jan-2010 16:55:18] <theacolyte-> Yeah, that one is hatin' it
[11-Jan-2010 16:55:28] <theacolyte-> I have a suspicion that's why I'm in the hole I am now
[11-Jan-2010 16:56:37] <rmatte> qI'm out, later
[11-Jan-2010 16:56:41] <theacolyte-> see ya man
[11-Jan-2010 17:02:36] <mrayzenoss> theacolyte-: open a bug for it will you? I think Alpha was added in 2.4, but we'll check it out
[11-Jan-2010 17:02:43] <theacolyte-> Yeah, will do
[11-Jan-2010 17:02:51] <mrayzenoss> defect review is tomorrow
[11-Jan-2010 17:02:58] <theacolyte-> sorry to bother you directly like this
[11-Jan-2010 17:03:11] <mrayzenoss> no worries
[11-Jan-2010 17:11:33] <theacolyte-> http://dev.zenoss.org/trac/ticket/6037
[11-Jan-2010 17:11:40] <theacolyte-> not that you need a blow by blow, but FYI in case you need it
[11-Jan-2010 17:11:49] <theacolyte-> I will follow up in 15 minutes to see if it's fixed
[11-Jan-2010 17:41:03] <mrayzenoss> rmatte: any chance that http://dev.zenoss.org/trac/ticket/5796 fixes http://dev.zenoss.org/trac/ticket/5289 ?
[11-Jan-2010 23:16:51] <rhettardo> i've forgotten how to do a simple task. for a IpService i want to make it only fire off alerts if the service comes up as failed X amount of times. can someone refresh my memory on how to do that?
[12-Jan-2010 00:00:30] [disconnected at Tue Jan 12 00:00:30 2010]
[12-Jan-2010 00:00:30] [connected at Tue Jan 12 00:00:30 2010]
[12-Jan-2010 04:03:14] <moto-moi> User-supplied Python expression (here.totalBlocks * .9) for maximum value caused error: ['disk_usedBlocks']
[12-Jan-2010 04:03:29] <moto-moi> Could it be some plugins don't like 2TB+ disks?
[12-Jan-2010 05:37:53] <moto-moi> hm, is it me, or is the ircd zenpack broken :( I get http://community.zenoss.org/message/42563#42563 and when I search in trac there is a (autogenerated) report, but no sollution :(
[12-Jan-2010 06:58:57] <chickenandbeans> hi, any zendmd gurus here?
[12-Jan-2010 07:00:04] <chickenandbeans> I've written a few scripts using dmd = ZenScriptBase(connect=True).dmd
[12-Jan-2010 07:00:37] <chickenandbeans> i''m trying to pipe the output of script 1 into script 2. This results in a db conflict error
[12-Jan-2010 07:01:22] <chickenandbeans> I reckon it's a locked connection issue?
[12-Jan-2010 07:01:25] <chickenandbeans> any takers?
[12-Jan-2010 08:08:58] <chickenandbeans> hi anyone about yet?
[12-Jan-2010 08:12:14] <chickenandbeans> :'(
[12-Jan-2010 08:12:31] <chickenandbeans> really need a hand here
[12-Jan-2010 08:12:50] <chickenandbeans> end product will be some really nice zenoss scripts..
[12-Jan-2010 08:12:56] <chickenandbeans> zendmd woes
[12-Jan-2010 08:15:07] <mrayzenoss> I assume you've seen the wiki zendmd stuff? http://community.zenoss.org/community/documentation/wiki/zendmd
[12-Jan-2010 08:15:24] <chickenandbeans> yeah, it's a problem not on there
[12-Jan-2010 08:15:36] <chickenandbeans> I've written a few scripts using dmd = ZenScriptBase(connect=True).dmd
[12-Jan-2010 08:15:41] <chickenandbeans> i''m trying to pipe the output of script 1 into script 2. This results in a db conflict error
[12-Jan-2010 08:15:54] <chickenandbeans> i'm thinking it's a locking issue?
[12-Jan-2010 08:16:30] <chickenandbeans> script1.py | script2.py = db confilict error
[12-Jan-2010 08:17:00] <chickenandbeans> script1.py > script1.out && cat script1.out | script2.py works!
[12-Jan-2010 08:17:22] <chickenandbeans> hope I explained that correctly
[12-Jan-2010 08:18:00] <mrayzenoss> yeah, I follow but I don't know about having zendmd scripts push info to each otherâ€¦ why not make a single script?
[12-Jan-2010 08:18:29] <chickenandbeans> because, script1 is useful on it's own as it's script2
[12-Jan-2010 08:18:44] <chickenandbeans> i'm creating a suite of mini applications
[12-Jan-2010 08:18:59] <chickenandbeans> it'll be cool... once it's sorted :)
[12-Jan-2010 08:25:59] <chickenandbeans> yup, it's a toughie..hence my :'(
[12-Jan-2010 08:34:49] <chickenandbeans> Stack trace is on the forum.. message 44420
[12-Jan-2010 08:50:18] <chickenandbeans> mmm, i've posted my problem on the dev forum... not sure how active it is though
[12-Jan-2010 08:50:41] <chickenandbeans> I reckon it'll be an easy one for a zendmd ninja
[12-Jan-2010 08:56:40] <RobertLaptop> Is there more info on event class mapping other then the 2-3 pages inside the admin manual?
[12-Jan-2010 09:05:03] <mrayzenoss> RobertLaptop: http://www.skills-1st.co.uk/papers/jcurry.html check out the Zenoss Event Management paper
[12-Jan-2010 09:06:43] <RobertLaptop> mrayzenoss will read things.
[12-Jan-2010 09:06:46] <RobertLaptop> thanks
[12-Jan-2010 09:09:01] <eidolon> hey folks, can i, from the command line, disable a monitor on a host?
[12-Jan-2010 09:09:27] <eidolon> i'm thinking of 'timed' maintenance windows - i need to disable alerting on a service during a time window. (not a host, a service). if i can cron a 'zenalertshaddup' command or something, that would work great.
[12-Jan-2010 09:09:54] <chickenandbeans> it's on the wiki under zendmd
[12-Jan-2010 09:10:15] <eidolon> which wiki?
[12-Jan-2010 09:10:51] <cparlette> do you just care about email alerts? if so, i'd just set up an alerting schedule for that alerting rule
[12-Jan-2010 09:11:08] <chickenandbeans> the zenoss wiki. There are a bunch of zendmd snippets you can customise
[12-Jan-2010 09:11:31] <eidolon> well, the problem is we have hundreds of machines and thousands of possible alerts. i'd ahve to set filters saying "this alert, this schedule, this machine, this service, don't alert during this window" - and that gets ugly really fast.
[12-Jan-2010 09:12:09] <eidolon> okay, what zenoss wiki?
[12-Jan-2010 09:12:19] <eidolon> googling 'zenoss wiki' puts to community.zenoss.org ? that one?
[12-Jan-2010 09:12:44] <chickenandbeans> search for it on the zenoss wiki
[12-Jan-2010 09:13:32] <eidolon> din't make me hurt you :)
[12-Jan-2010 09:13:57] <chickenandbeans> once you see how to do it, it's all about deciphering the api i'm afraid.. thank the gods for tab complete
[12-Jan-2010 09:14:15] <eidolon> a url would help.
[12-Jan-2010 09:15:18] <cparlette> http://community.zenoss.org/community/documentation/wiki/zendmd
[12-Jan-2010 09:15:33] <mrayzenoss> eidolon: http://community.zenoss.org/blogs/zenossblog/2009/12/23/tip-of-the-month-dynamically-creating-maintenance-windows
[12-Jan-2010 09:15:43] <eidolon> THANK you.
[12-Jan-2010 09:15:59] <chickenandbeans> ah beat me to it
[12-Jan-2010 09:16:01] <mrayzenoss> command line maintenance windows right?
[12-Jan-2010 09:16:34] <eidolon> yessir.
[12-Jan-2010 09:16:47] <mrayzenoss> then that's what you need
[12-Jan-2010 09:16:52] <eidolon> not really though
[12-Jan-2010 09:16:53] * eidolon sighs.
[12-Jan-2010 09:17:06] <eidolon> i need to stop an alert from happening on a device - for a specific service.
[12-Jan-2010 09:17:45] <eidolon> here's teh situation. we have mysql servers that are set up to replicate from one to the other. for a 4 hour period between like 2am and 6am, the replication is stopped for backups. but our slave monitor alerts go off for that device during that window.
[12-Jan-2010 09:18:05] <eidolon> i want to be able to say "stop monitorign" at 2am, and say "start monitoring" at 6am. but only for that one monitor. not for the entire host.
[12-Jan-2010 09:19:31] <chickenandbeans> device/host... i'm lost here. Are you saying you are monitoring several mysql instances on one device
[12-Jan-2010 09:19:42] <chickenandbeans> and that you want to stop monitoring one of those instances?
[12-Jan-2010 09:19:51] <cparlette> you should be able to through zendmd, make a script that will set monitor on that specific process to false
[12-Jan-2010 09:19:54] <cparlette> then another to set it to true
[12-Jan-2010 09:20:10] <cparlette> i dont know it off the top of my head so i dont know if that helps, but it's possible
[12-Jan-2010 09:20:27] <eidolon> chickenandbeans: no.
[12-Jan-2010 09:20:41] <eidolon> chickenandbeans: my understanding of maintenance windows is - you can set a maintenance window on a HOST.
[12-Jan-2010 09:20:47] <eidolon> but a single host can have dozens of monitors on it.
[12-Jan-2010 09:20:55] <chickenandbeans> That's good, then what cparlette applies
[12-Jan-2010 09:21:01] <eidolon> i need to stop one of the monitors from alerting for that maintenance window, but leave all the other monitors active.
[12-Jan-2010 09:21:25] <eidolon> yeah, i need to learn zendmd and api stuff. which i'm sure will generate a raft of other questions. since my python is weak at best :)
[12-Jan-2010 09:21:58] <chickenandbeans> dev = find('mydevice')
[12-Jan-2010 09:22:12] <chickenandbeans> dev.monitor=False
[12-Jan-2010 09:22:27] <eidolon> chickenandbeans: do you really think before you answer questions?
[12-Jan-2010 09:22:34] <eidolon> since that will shut down monitoring on the entire device.
[12-Jan-2010 09:22:44] <eidolon> which, as i've said now 4 times, is what i don't want.
[12-Jan-2010 09:23:21] <chickenandbeans> I did ask that, but you said no
[12-Jan-2010 09:23:38] <chickenandbeans> a bit of miscommunication there, sorry.
[12-Jan-2010 09:24:28] <chickenandbeans> I've spent the last week going through the zendmd (creating templates, graphs, adding devices, monitoring stuff etc)
[12-Jan-2010 09:25:10] <cparlette> http://pastebin.com/m15bd64bf
[12-Jan-2010 09:25:14] <cparlette> i think that should work
[12-Jan-2010 09:25:31] <eidolon> that's good.
[12-Jan-2010 09:25:35] <cparlette> pasted into zendmd
[12-Jan-2010 09:25:50] <eidolon> i still have to understand zendmd. but... that's a specific process monitor... and isn't quite what i'm doing.
[12-Jan-2010 09:25:57] <eidolon> sorry this is sounding complex.
[12-Jan-2010 09:26:50] <eidolon> i have a Cmd monitor on a specific host - part of a template. the command queries the mysql server and asks for it's slave (replication) status, nd throws an error if it fails.
[12-Jan-2010 09:26:56] <eidolon> let me screenshot the monitor.
[12-Jan-2010 09:27:07] <cparlette> ah, didnt know it was a command data source
[12-Jan-2010 09:27:12] <mrayzenoss> you want a maintenance window for a template, or for a command data source
[12-Jan-2010 09:27:14] <eidolon> 'sok.
[12-Jan-2010 09:27:40] <eidolon> mrayzenoss: i'm... not... sure. the template includes a command datasource.
[12-Jan-2010 09:28:28] <chickenandbeans> then what on earth do you want
[12-Jan-2010 09:28:33] <eidolon> i want it all!
[12-Jan-2010 09:28:35] <eidolon> hang on a sec. :)
[12-Jan-2010 09:28:50] <chickenandbeans> i'm out.
[12-Jan-2010 09:28:54] <mrayzenoss> sorry, gotta run to a meeting
[12-Jan-2010 09:28:54] <chickenandbeans> :)
[12-Jan-2010 09:29:12] <eidolon> awesome. :)
[12-Jan-2010 09:29:41] <chickenandbeans> a bit rude there on the 'do you think' part too ...
[12-Jan-2010 09:30:39] <eidolon> well from my side, i had spelled out the problem several times, and it wasn't getting through. i had also asked for the zenoss wiki you referred to, and you wouldn't answer what it was... part of irc help is being clear with answers, and it was very frustrating that i felt i was asking a clear question, and was getting very oblique answers.
[12-Jan-2010 09:31:41] <eidolon> http://www.homeport.org/~dbs/screenshots/slavemonitors.png <-- this is the alert in the event console. Those slave alerts are what i'm trying to 'mask out' for a maintenance window.
[12-Jan-2010 09:32:55] <chickenandbeans> Riiiight.... anyone got any further clues on: http://community.zenoss.org/message/44420#44420
[12-Jan-2010 09:33:28] <chickenandbeans> I've checked out the api (ZenScriptBase).. clear as mud :)
[12-Jan-2010 09:33:40] <eidolon> http://www.homeport.org/~dbs/screenshots/mysqlslavetemplate.png <-- and this is the template definition applied to that host.
[12-Jan-2010 09:34:13] <eidolon> goal: how do i set a maintenance window on -just- that template on -just- that host?
[12-Jan-2010 09:34:33] <eidolon> (i think... that might have been clear / direct enough? trying here. :)
[12-Jan-2010 09:36:26] <cparlette> eidolon: if that template isn't a local copy on the device, you won't be able to disable the data source for a single device
[12-Jan-2010 09:36:38] <eidolon> it isn't a local copy. :(
[12-Jan-2010 09:36:52] <cparlette> if you make a local copy, or if that template is only bound to that one device, then you can change it
[12-Jan-2010 09:36:56] <eidolon> i was wondering if i could put a transform or somethin gin the event itself... that would be checking the device name.
[12-Jan-2010 09:37:14] <cparlette> in a transform you could, yes
[12-Jan-2010 09:37:40] <cparlette> i haven't worked with times in transforms, but i'm sure it would work
[12-Jan-2010 09:37:59] <cparlette> to either drop or move to history if the lastTime is within a certain range
[12-Jan-2010 09:38:13] <eidolon> *nods*
[12-Jan-2010 09:47:25] <chickenandbeans> mmm "self.poollock = Lock()" curiouser and curiouser..
[12-Jan-2010 10:09:28] <chickenandbeans> ah, just found the wiki page with eidolon's query on.. well I think i did.. his explaination changed a bit... never mind eh
[12-Jan-2010 10:21:19] <chickenandbeans> does anyonee know when the next dev zenoss irc session is?
[12-Jan-2010 10:31:09] <theacolyte-> phew busy day of zenoss alerts
[12-Jan-2010 10:31:22] <Grizmawe> good afternoon - I am getting an error that zenoss cannot save RRD data for a disk on one of my servers: Unable to save data for OID .1.3.6.1.4.1.232.3.2.5.1.1.51.0.129 in RRD Devices/l<hostname>/hw/harddisks/HardDisk0 - can anybody translate, couldnt find anything by googling. The traceback in the error reports invalid literal fo. Is this the server not returning the data in a correct format or is it an issue with the RRD?
[12-Jan-2010 10:32:25] bzed_ is now known as bzed
[12-Jan-2010 11:20:40] <rmatte> Grizmawe: try deleting the RRD file
[12-Jan-2010 11:20:48] <rmatte> Grizmawe: also, remodel the device
[12-Jan-2010 11:23:05] <Grizmawe> rmatte, thanks, will try
[12-Jan-2010 11:23:36] <rmatte> k
[12-Jan-2010 11:37:23] <rmatte> eugh, man it pisses me off when Zenoss automatically puts the checkmark back in the HttpMon checkbox for SSL
[12-Jan-2010 11:38:06] <rmatte> aha, managed to uncheck it in the management interface
[12-Jan-2010 11:38:06] <rmatte> woot
[12-Jan-2010 11:45:05] <chickenandbeans> :) That interface has saved much time..
[12-Jan-2010 11:46:05] <rmatte> yup
[12-Jan-2010 11:49:10] <chickenandbeans> bye all, i'll try and become a regular on here.
[12-Jan-2010 11:54:21] <nonsenso> morning. :)
[12-Jan-2010 12:03:28] <rmatte> morning
[12-Jan-2010 12:04:18] <rmatte> heya Egor, what's new an exciting?
[12-Jan-2010 12:04:27] <rmatte> s/an/and
[12-Jan-2010 12:07:06] <bigegor> hi,
[12-Jan-2010 12:07:20] <bigegor> i've update WMIDataSourec and WBEMDataSourec on SVN
[12-Jan-2010 12:08:13] <bigegor> Changelog:
[12-Jan-2010 12:08:25] <bigegor> * Query syntax unification with WBEMDataSource (WQL, InstanceName)
[12-Jan-2010 12:08:25] <bigegor> * Queries sorting (join multiple queries in one query)
[12-Jan-2010 12:08:25] <bigegor> * CIMDataSource support
[12-Jan-2010 12:08:25] <bigegor> * support for DataPoint Aliases (evaluate befor write RRD)
[12-Jan-2010 12:08:25] <bigegor> * WMIClient can be used standalon, based on modified Query.py from pysamba
[12-Jan-2010 12:08:27] <bigegor> * WMIPlugin CollectorPlugin added (any namespaces, tables format like in WBEMPlugin)
[12-Jan-2010 12:08:29] <bigegor> * added __path attribute for Wbem Insatnce
[12-Jan-2010 12:08:31] <bigegor> * support for avg, count, sum, min, max function for datapoints with multiline result
[12-Jan-2010 12:08:34] <bigegor> * WmiProxy support
[12-Jan-2010 12:09:50] <bigegor> beta quality, need more testing
[12-Jan-2010 12:13:14] <rmatte> cool
[12-Jan-2010 12:14:45] <bigegor> I will try to implement WMI/WBEM only monitoring without SNMP dependencies
[12-Jan-2010 12:22:19] <rmatte> bigegor: that would be awesome
[12-Jan-2010 12:24:27] <bigegor> rmatte: but i hope it possible now with new WMIDataSource
[12-Jan-2010 12:25:00] <rmatte> hopefully
[12-Jan-2010 12:25:01] <skipzoid> hello all
[12-Jan-2010 12:25:25] <rmatte> skipzoid: hey
[12-Jan-2010 12:26:09] <rmatte> bigegor: I'm going to start a project to create a ZenPack which makes using multiple collectors with Zenoss Core easier. No idea how long it'll take.
[12-Jan-2010 12:26:13] <skipzoid> quick q. on the graphs where cur/avg/max are - how easy is it to add min too ?
[12-Jan-2010 12:26:59] <rmatte> skipzoid: you'd have to add it in to the RRD options in the template
[12-Jan-2010 12:27:05] <rmatte> for each template
[12-Jan-2010 12:27:18] <skipzoid> ahhhhh - book time again ?
[12-Jan-2010 12:27:18] <rmatte> not sure what the actual option would be, you'll have to hunt around for it
[12-Jan-2010 12:27:52] <rmatte> why do you need min out of curiosity?
[12-Jan-2010 12:28:44] <skipzoid> i'm graphing satellite downlink signal levels, dB is always degative figures
[12-Jan-2010 12:29:10] <skipzoid> negative*
[12-Jan-2010 12:30:59] <bigegor> rmatte: what do you mean by "multiple collectors"? distributed collectors?
[12-Jan-2010 12:31:31] <rmatte> skipzoid: ah, I tried graphing negative values about a year back and it never worked
[12-Jan-2010 12:31:43] <rmatte> bigegor: yeh, distributed
[12-Jan-2010 12:31:51] <bigegor> super!
[12-Jan-2010 12:32:02] <rmatte> bigegor: there are instructions on how to do it, but I want to make a pack out of it to make it braindead simple to do
[12-Jan-2010 12:32:44] <rmatte> bigegor: the configuration will be done via commandline commands since I don't really want to waste time on a UI component of it right away
[12-Jan-2010 12:32:56] <rmatte> bigegor: but I'll make it as simple as possible
[12-Jan-2010 12:33:41] <skipzoid> they are graphing ok, all beit downwards instead of upwards if you know what I mean and the thresholds are working and triggering sending text messages etc fine, its just sometimes we need to look back and display the levels and what the min/max was for a period, so we can blame the weather and not equipment control
[12-Jan-2010 12:34:09] <rmatte> hehe
[12-Jan-2010 12:34:56] <bigegor> rmatte: i've also thinking about such as ZenPack
[12-Jan-2010 12:35:06] <theacolyte-> bigegor: hello there
[12-Jan-2010 12:35:13] <skipzoid> another q. the graphing is displayed over 12 hours, I'd like to be able to select from the drop down 2 hours ?
[12-Jan-2010 12:35:17] <bigegor> hey
[12-Jan-2010 12:35:32] <rmatte> bigegor: cool, I've been thinking about it just recently since our company is going to have a need for distributed collectors
[12-Jan-2010 12:35:36] <theacolyte-> bigegor: sorry to bug you in here, but did you happen to see my PM? :)
[12-Jan-2010 12:36:06] <rmatte> bigegor: I also want to take a crack at writing a collector plugin for my existing windows performance zenpacks
[12-Jan-2010 12:36:34] <bigegor> theacolyte-:yes, but without any idea :(
[12-Jan-2010 12:37:01] <theacolyte-> bigegor: ouch
[12-Jan-2010 12:37:43] <theacolyte-> That's a real bummer
[12-Jan-2010 12:38:34] <bigegor> theacolyte-: you can try new WMIDataSource from SVN it write data in RRD only as float.
[12-Jan-2010 12:39:06] <theacolyte-> It's weird
[12-Jan-2010 12:39:11] <theacolyte-> there's no reason for it to not work
[12-Jan-2010 12:39:21] <theacolyte-> the datasource works, but the damn graphs never get generated
[12-Jan-2010 12:41:48] <bigegor> theacolyte-: i your PM, i can see only RenderServer Error. Sorry
[12-Jan-2010 12:41:58] <theacolyte-> What else do you need?
[12-Jan-2010 12:43:05] <rmatte> RenderServer Error is extremely general by the way theacolyte-, that errors shows up every time any graph doesn't render for whatever reason
[12-Jan-2010 12:43:17] <theacolyte-> Well, tell me what you need and I can provide it
[12-Jan-2010 12:43:21] <theacolyte-> I don't know any of this stuff
[12-Jan-2010 12:43:26] <rmatte> It's equivalent to Zenoss saying "Sorry, the graph didn't show up"
[12-Jan-2010 12:43:27] <rmatte> lol
[12-Jan-2010 12:44:12] <theacolyte-> the problem is if I can't get this to work there are some pretty good odds I need to revaluate a monitoring system, and I've already invested a ton of time and energy in it
[12-Jan-2010 12:45:36] <theacolyte-> Unfortunately my boss may not want to continue heh
[12-Jan-2010 12:45:55] <rmatte> so it's generating the RRDs but not the graphs?
[12-Jan-2010 12:46:00] <theacolyte-> No RRDs
[12-Jan-2010 12:46:07] <rmatte> hmmmm
[12-Jan-2010 12:46:10] <bigegor> theacolyte-:try to start zenperfwmi daemon with debug option: zenperfwmi run -v10 -d yourserver
[12-Jan-2010 12:46:15] <rmatte> did it ever generate them properly?
[12-Jan-2010 12:46:24] <theacolyte-> it never generated them period
[12-Jan-2010 12:46:29] <willwh> hi guys - anyone monitoring stuff on the EC2 cloud? (if so, how?) - their disk images don't seem to have snmp on them
[12-Jan-2010 12:46:30] <theacolyte-> and the RRD's from my old graph still exist
[12-Jan-2010 12:46:33] <rmatte> yeh, do what egor just said, and look for any detailed errors
[12-Jan-2010 12:46:36] <theacolyte-> (never got cleared - this may be working as intended)
[12-Jan-2010 12:46:40] <theacolyte-> bigegor: trying right now
[12-Jan-2010 12:49:02] <theacolyte-> or I would if I knew linux better
[12-Jan-2010 12:49:08] <theacolyte-> sudo su - zenoss zenperfwmi run -v10 -d localhost
[12-Jan-2010 12:49:08] <theacolyte-> su: invalid option -- v
[12-Jan-2010 12:49:52] <theacolyte-> bigegor: can I change that config from the zenoss gui?
[12-Jan-2010 12:49:56] <theacolyte-> is that just setting logging to debug?
[12-Jan-2010 12:50:59] <bigegor> theacolyte-: yes and yes
[12-Jan-2010 12:51:05] <theacolyte-> got it, set it to debug
[12-Jan-2010 12:51:18] <theacolyte-> bigegor: now, how to reproduce the error, heh
[12-Jan-2010 12:51:35] <bigegor> wait 5 min ;)
[12-Jan-2010 12:51:54] <theacolyte-> ah I removed the datasource
[12-Jan-2010 12:51:58] <theacolyte-> let me readd it if I can remember
[12-Jan-2010 12:52:47] <theacolyte-> looks like I may have freaking deleted the datapoints without writing them down.
[12-Jan-2010 12:53:01] <willwh> go ahead and answer my own question :) http://community.zenoss.org/message/41179#41179
[12-Jan-2010 12:53:29] <bigegor> later
[12-Jan-2010 12:53:38] <theacolyte-> great
[12-Jan-2010 13:06:35] <QubeZ> im running 2.5.1 and when I try to add a datasource to one of my templates (AsteriskNew), zenoss hangs and I have to restart my services
[12-Jan-2010 13:06:38] <QubeZ> bug?
[12-Jan-2010 13:08:19] <theacolyte-> Missing RRD files, but the graphs show up now.
[12-Jan-2010 13:08:24] <theacolyte-> I'm done with WMI and Zenoss though
[12-Jan-2010 13:08:30] <theacolyte-> I can't take the frustration and I can't fix it
[12-Jan-2010 13:08:41] <theacolyte-> Zenoss is wonderful for SNMP
[12-Jan-2010 13:09:02] <theacolyte-> That's a real shame
[12-Jan-2010 13:09:04] <cparlette> if graphs just started showing up, then zenoss is collecting data
[12-Jan-2010 13:09:14] <theacolyte-> No, this isn't the normal delay
[12-Jan-2010 13:10:00] <theacolyte-> RRD files never get generated.
[12-Jan-2010 13:12:26] <theacolyte-> Is there a command I can use to delete unused RRD files, or do I just need to wipe everything out?
[12-Jan-2010 13:12:41] <theacolyte-> I'm removing the WMI zenpack and it seems that the RRD files remain
[12-Jan-2010 13:14:02] <cparlette> nothing besides bash commands
[12-Jan-2010 13:14:07] <theacolyte-> heh, k
[12-Jan-2010 13:22:13] <theacolyte-> Heh, probably with getting rid of WMI and just cutting back to SNMP is I lose event logs
[12-Jan-2010 13:22:17] <theacolyte-> I guess this is it then
[12-Jan-2010 13:29:06] <cparlette> zeneventlog can work independant of your template data sources
[12-Jan-2010 13:29:20] <cparlette> so you can leave wmi enabled on the devices, and leave zwinuser/zwinpassword, and zeneventlog will still work
[12-Jan-2010 13:29:30] <cparlette> even if you're using snmp to collect cpu/memory/etc
[12-Jan-2010 13:41:02] <Apachez> theacolyte-: the shame is that zope is so f**king slow
[12-Jan-2010 13:42:16] <theacolyte-> It's not slow for me, I just can't get this to work
[12-Jan-2010 13:42:20] <theacolyte-> And I'm done bashing my head against it
[12-Jan-2010 13:42:22] <theacolyte-> It's too frustrating
[12-Jan-2010 13:42:30] <theacolyte-> What's even more frustrating is the time investment
[12-Jan-2010 13:42:44] <theacolyte-> Anyways, I will save you guys my rant, it's unprofessional
[12-Jan-2010 13:42:45] <theacolyte-> Sorry
[12-Jan-2010 13:50:32] <Apachez> how can it NOT be slow?
[12-Jan-2010 13:59:24] <rmatte> theacolyte-: you have to do sudo su - zenoss as it's own command
[12-Jan-2010 13:59:34] <theacolyte-> Heh yeah figured that out
[12-Jan-2010 13:59:34] <rmatte> theacolyte-: you can't tack a command on to the end of it
[12-Jan-2010 13:59:37] <rmatte> lol
[12-Jan-2010 13:59:47] <theacolyte-> I did figure out my WMI issue and the graph stuff
[12-Jan-2010 13:59:55] <theacolyte-> But I ran into the same bug I always run into
[12-Jan-2010 13:59:57] <theacolyte-> so...
[12-Jan-2010 14:00:11] <theacolyte-> It's like a constant uphill battle, it's killing me
[12-Jan-2010 14:00:19] <theacolyte-> I really hope they answer http://dev.zenoss.org/trac/ticket/6019#comment:2
[12-Jan-2010 14:00:32] <theacolyte-> What actually creates the RRD file itself?
[12-Jan-2010 14:00:49] <rmatte> technically what that command does is the following, it launches sudo which gives you root privileges, then sudo executes su which is the switch user program, since the command is su - zenoss it knows to switch to the zenoss user, and the - just passes extra environment info over once you switch users... sudo su zenoss works too
[12-Jan-2010 14:01:03] <theacolyte-> ahh
[12-Jan-2010 14:01:10] <theacolyte-> I suppose that makes sense
[12-Jan-2010 14:01:20] <rmatte> just so that you understand the command ;)
[12-Jan-2010 14:01:29] <theacolyte-> hehe
[12-Jan-2010 14:02:18] <rmatte> they use rrdtools to generate the RRD files
[12-Jan-2010 14:02:42] <theacolyte-> And what calls RRDtool?
[12-Jan-2010 14:02:54] <theacolyte-> I'm trying to understand the workflow, maybe I can hack at it
[12-Jan-2010 14:03:36] <rmatte> have you tried just remodelling the device after changing the template?
[12-Jan-2010 14:04:30] <theacolyte-> believe I did, one sec
[12-Jan-2010 14:04:51] <theacolyte-> no effect
[12-Jan-2010 14:05:03] <rmatte> have you tried just deleting the whole RRD directory?
[12-Jan-2010 14:05:10] <rmatte> it should automatically recreate it
[12-Jan-2010 14:05:31] <theacolyte-> That fixes it
[12-Jan-2010 14:05:38] <theacolyte-> However, that also deletes all my historical info
[12-Jan-2010 14:05:47] <theacolyte-> Right?
[12-Jan-2010 14:06:07] <rmatte> ok, you have to understand that RRD files are based on the name of the datapoint
[12-Jan-2010 14:06:10] <rmatte> not the OID itself
[12-Jan-2010 14:06:35] <rmatte> so unless you manually rename your old RRD files to match the new name of the datapoint, you won't be keeping your historical data
[12-Jan-2010 14:07:20] <rmatte> how much historical data do you have? you're in initial setup phase, I doubt you'll be doing much template editing when you're done
[12-Jan-2010 14:07:38] <theacolyte-> Nothing too much, but this isn't sustainable
[12-Jan-2010 14:07:54] <rmatte> I've never seen the issue that you're describing
[12-Jan-2010 14:08:00] <rmatte> have you tried just restarting Zenoss?
[12-Jan-2010 14:08:06] <theacolyte-> A google search shows that it's not all that uncommon
[12-Jan-2010 14:08:09] <theacolyte-> yeah, but let me try again
[12-Jan-2010 14:08:18] <rmatte> It should always generate new RRD files if there isn't currently a matching one
[12-Jan-2010 14:08:37] <theacolyte-> You would think so
[12-Jan-2010 14:08:51] <theacolyte-> What I know works is removing the device and readding it
[12-Jan-2010 14:08:55] <rmatte> it does for me, so I have no idea what's so different about your install
[12-Jan-2010 14:09:06] <theacolyte-> Man, I don't know, it doesn't make any sense
[12-Jan-2010 14:09:14] <rmatte> other than the fact that you're using Zenoss 2.5 and I'm using 2.4.5
[12-Jan-2010 14:09:23] <theacolyte-> Maybe that's my fatal flaw
[12-Jan-2010 14:09:39] <rmatte> 2.5 might have a bug that's causing that behaviour
[12-Jan-2010 14:10:38] <rmatte> did you see any errors when running zenperfwmi with -v10?
[12-Jan-2010 14:11:13] <theacolyte-> none other than the ones I already knew about
[12-Jan-2010 14:11:48] <rmatte> so it won't generate the RRD even if you wait over an hour?
[12-Jan-2010 14:12:13] <theacolyte-> well
[12-Jan-2010 14:12:17] <theacolyte-> I can try waiting an hour
[12-Jan-2010 14:12:22] <theacolyte-> I never really leave it untouched
[12-Jan-2010 14:12:27] <theacolyte-> I thought the hour was the remodel though
[12-Jan-2010 14:12:32] <rmatte> here's what I think may be happening...
[12-Jan-2010 14:12:44] <theacolyte-> The files literally don't exist
[12-Jan-2010 14:13:17] <rmatte> The collector is not having enough time to become aware of the new state of the template (the collector only gets updates every hour by default), therefore it's collecting based on what the template used to look like rather than what it currently looks like, hence no RRD files
[12-Jan-2010 14:13:32] <theacolyte-> I thought that's what push changes did?
[12-Jan-2010 14:13:41] <rmatte> it is, so you've tried that?
[12-Jan-2010 14:13:44] <theacolyte-> Right
[12-Jan-2010 14:13:45] <theacolyte-> Yes
[12-Jan-2010 14:14:00] <rmatte> and you waited like 15 minutes and never had an RRD generated?
[12-Jan-2010 14:14:09] <theacolyte-> It's been 15 minutes, yeah
[12-Jan-2010 14:14:29] <rmatte> ...and this is an SNMP datapoint?
[12-Jan-2010 14:14:44] <theacolyte-> WMI
[12-Jan-2010 14:14:52] <rmatte> you mentioned an "OID" in your trac ticket
[12-Jan-2010 14:15:00] <rmatte> OID is a SNMP term
[12-Jan-2010 14:15:24] <theacolyte-> heh yeah
[12-Jan-2010 14:15:26] <theacolyte-> you're right
[12-Jan-2010 14:15:32] <theacolyte-> I just added a comment since I can't edit the original
[12-Jan-2010 14:15:36] <bigegor> hey
[12-Jan-2010 14:15:38] <rmatte> cool
[12-Jan-2010 14:15:41] <rmatte> hey
[12-Jan-2010 14:15:48] <theacolyte-> Although I think that may have happened to me with SNMP as well yesterday
[12-Jan-2010 14:16:32] <rmatte> theacolyte-: I've never seen it happen with SNMP, or with command based datasources. Maybe you should install 2.4.5 and try that to see if you get the same behaviour
[12-Jan-2010 14:16:42] <rmatte> If not, then use 2.4.5 until it gets fixed in 2.5
[12-Jan-2010 14:16:56] <theacolyte-> No joke
[12-Jan-2010 14:17:02] <theacolyte-> I wonder how I can migrate all my stuff over though
[12-Jan-2010 14:17:09] <theacolyte-> ah well, I can always spin up a new VM
[12-Jan-2010 14:17:24] <theacolyte-> of course, depending on how well this phone interview goes, it may be a non issue LOL
[12-Jan-2010 14:17:50] <rmatte> If 2.4.5 does work you'll want to install these patches: http://community.zenoss.org/docs/DOC-4434$
[12-Jan-2010 14:17:54] <rmatte> erm: http://community.zenoss.org/docs/DOC-4434
[12-Jan-2010 14:18:18] <rmatte> I'd say spin up a new VM and see if it works for starters
[12-Jan-2010 14:18:27] <theacolyte-> I favorited it
[12-Jan-2010 14:18:32] <rmatte> k
[12-Jan-2010 14:18:34] <theacolyte-> Well, the one thign that's going for me is I can reproduce this problem
[12-Jan-2010 14:18:36] <theacolyte-> Every single time
[12-Jan-2010 14:18:36] <bigegor> theacolyte-:have you found the problem?
[12-Jan-2010 14:18:45] <theacolyte-> Create graph, change graph, expect RRD files, not get them
[12-Jan-2010 14:18:49] <theacolyte-> bigegor: With the WMI part, yes
[12-Jan-2010 14:19:04] <rmatte> what was the problem with WMI?
[12-Jan-2010 14:19:04] <theacolyte-> I'm not going to tell you what it was, because rmatte may actually kill me
[12-Jan-2010 14:19:17] <rmatte> lol
[12-Jan-2010 14:19:23] <bigegor> :)
[12-Jan-2010 14:19:27] <theacolyte-> Ok
[12-Jan-2010 14:19:27] <rmatte> Let me guess, it was your query
[12-Jan-2010 14:19:30] <theacolyte-> Yep.
[12-Jan-2010 14:19:32] <theacolyte-> #&*()@*)(#)@(#()@#*)@#
[12-Jan-2010 14:19:39] <theacolyte-> Apparently committed has 2 M's and 2 T's
[12-Jan-2010 14:19:46] <rmatte> I asked you like 5 times whether you were 100% sure that the datasources were correct lol
[12-Jan-2010 14:19:48] <rmatte> :P
[12-Jan-2010 14:19:54] <theacolyte-> I swear I had it right at some point
[12-Jan-2010 14:19:57] <theacolyte-> It's all confused now
[12-Jan-2010 14:20:11] <rmatte> theacolyte-: you know that when you create a new datapoint you have to remove the old graphpoint and add the new one right?
[12-Jan-2010 14:20:30] <theacolyte-> Yeah
[12-Jan-2010 14:20:33] <theacolyte-> I delete the graph
[12-Jan-2010 14:20:35] <theacolyte-> Delete the datapoints
[12-Jan-2010 14:20:38] <theacolyte-> and recreate it all
[12-Jan-2010 14:21:04] <rmatte> What if you just outright delete the template and recreate it from scratch?
[12-Jan-2010 14:21:09] <rmatte> since that's basically what you're doing anyways
[12-Jan-2010 14:21:28] <theacolyte-> I could do that
[12-Jan-2010 14:21:31] <theacolyte-> I may actually try that
[12-Jan-2010 14:21:36] <theacolyte-> it's easier than the alternatives
[12-Jan-2010 14:21:40] <rmatte> yup
[12-Jan-2010 14:22:00] <theacolyte-> brb, undocking and going to conference room with blinds, interview in 10
[12-Jan-2010 14:22:26] <rmatte> cool
[12-Jan-2010 14:23:51] <theacolyte-> Either way
[12-Jan-2010 14:24:00] <theacolyte-> Whatever is happening, shouldn't be happening
[12-Jan-2010 14:24:21] <rmatte> yup
[12-Jan-2010 14:24:40] <rmatte> well, it's not happening in 2.4.5 from what I can tell, so you might take a stab at trying that
[12-Jan-2010 14:25:00] <theacolyte-> And regardless of how ridiculously stupid I am about having stuff named right, it can't be that, and this consistantly
[12-Jan-2010 14:25:05] <theacolyte-> Even I'm not that dumb
[12-Jan-2010 14:25:38] <theacolyte-> Maybe I should make an interface suggestion on that datapoint
[12-Jan-2010 14:25:51] <theacolyte-> something as simple as validating it against the class wouldn't be a bad idea,
[12-Jan-2010 14:26:01] <theacolyte-> Or a real error instead of a broken image
[12-Jan-2010 14:26:45] <rmatte> well, usually you'll see "RRD File not found" in the graph
[12-Jan-2010 14:26:50] <rmatte> rather than a completely broken image
[12-Jan-2010 14:27:08] <theacolyte-> from my experienc,e and I should document this
[12-Jan-2010 14:27:12] <rmatte> Completely broken images are generally from something else
[12-Jan-2010 14:27:16] <theacolyte-> a broken image = bad OID/Class
[12-Jan-2010 14:27:32] <rmatte> did the zenoss trac system suddenly die?
[12-Jan-2010 14:27:32] <theacolyte-> Missing RRD file = myg bug
[12-Jan-2010 14:27:36] <rmatte> I suddenly can't get to it
[12-Jan-2010 14:28:09] <mrayzenoss> we're having connectivity issues
[12-Jan-2010 14:28:13] <mrayzenoss> at our colo
[12-Jan-2010 14:28:19] <rmatte> ah ok
[12-Jan-2010 14:28:23] <theacolyte-> Quick, call the rack monkies!
[12-Jan-2010 14:28:27] <rmatte> good timing for it, just when I need to use trac lol
[12-Jan-2010 14:28:33] <rmatte> curse the gods!
[12-Jan-2010 14:28:42] <theacolyte-> hehee
[12-Jan-2010 14:29:10] <theacolyte-> I think something is wrong with my wireless
[12-Jan-2010 14:29:19] <theacolyte-> Reply from 74.125.19.103: bytes=32 time=1200ms TTL=56
[12-Jan-2010 14:29:28] <rmatte> I'll say
[12-Jan-2010 14:29:40] <theacolyte-> AS fun as SSH with 1.2 seconds worth of lagv is
[12-Jan-2010 14:29:48] <rmatte> haha
[12-Jan-2010 14:29:59] <rmatte> type a paragraph and watch it slowly trickle in
[12-Jan-2010 14:30:07] <theacolyte-> And kind of stuttered , too
[12-Jan-2010 14:30:13] <rmatte> yeh
[12-Jan-2010 14:30:21] <theacolyte-> w w w w w w w w w w h h h a a tT?
[12-Jan-2010 14:31:28] <rmatte> finally, trac is back
[12-Jan-2010 14:40:38] <mrayzenoss> ugh
[12-Jan-2010 15:01:00] <theacolyte-> rmatte: Well my interview went really well
[12-Jan-2010 15:01:08] <theacolyte-> Unfortunately though, my RRD files are still missing
[12-Jan-2010 15:01:12] <theacolyte-> I need to grab a quick bite to eat though
[12-Jan-2010 15:03:19] <Apachez> I hope that interview wasnt about recovering RRD ;)
[12-Jan-2010 15:05:02] <nonsenso> hehe.
[12-Jan-2010 15:12:32] <rmatte> lol
[12-Jan-2010 15:19:42] mrayzenoss1 is now known as mrayzenoss
[12-Jan-2010 15:23:04] <theacolyte-> hahahahaha
[12-Jan-2010 15:23:07] <theacolyte-> that was a good one
[12-Jan-2010 15:23:19] <theacolyte-> no it definately wasn't - and they also did ask me to spell committed
[12-Jan-2010 15:26:58] <nonsenso> theacolyte-: what city were you interviewing in?
[12-Jan-2010 15:27:25] <theacolyte-> This job is in Berkeley
[12-Jan-2010 15:27:35] <nonsenso> hey! i live in berkeley.
[12-Jan-2010 15:27:38] <nonsenso> :P
[12-Jan-2010 15:27:39] <theacolyte-> No shit?
[12-Jan-2010 15:27:42] <nonsenso> srsly.
[12-Jan-2010 15:27:50] <theacolyte-> Berkeley, CA?
[12-Jan-2010 15:27:53] <nonsenso> aye aye.
[12-Jan-2010 15:27:56] <theacolyte-> Woah...
[12-Jan-2010 15:28:22] <Apachez> o'rly?
[12-Jan-2010 15:38:58] <rmatte> they asked you to spell committed?
[12-Jan-2010 15:38:59] <rmatte> why?
[12-Jan-2010 15:39:29] <theacolyte-> LOL
[12-Jan-2010 15:39:34] <theacolyte-> no that was the counter I messed up on
[12-Jan-2010 15:39:39] <theacolyte-> I thought it was commited
[12-Jan-2010 15:39:43] <theacolyte-> For no really good reason
[12-Jan-2010 16:50:41] <nonsenso> anyone around? :)
[12-Jan-2010 16:56:07] <mrayzenoss> kinda quiet, figured I was offline like our colo was
[12-Jan-2010 16:58:45] <nonsenso> so i'm kinda curious about template binding best practices. from everything i've read and done it sounds like best practice is to bind the template to the device class.
[12-Jan-2010 16:59:23] <nonsenso> but say i have apache and mysql systems. some systems run both and some just one. in this case the only way to cope is to bind local to each device.
[12-Jan-2010 16:59:44] <nonsenso> so should it really be, bind to the device class, except in the above conditions?
[12-Jan-2010 17:06:20] <theacolyte-> I'm sitting here crying in my soup
[12-Jan-2010 17:06:42] <theacolyte-> I'm creating device classes for them personally
[12-Jan-2010 17:06:51] <theacolyte-> But you can obviously bind them locally
[12-Jan-2010 17:10:51] <nonsenso> yeah but then you have all these local bindings. the idea of local bindings like that for 400+ systems makes me wanna cry.
[12-Jan-2010 17:14:31] <Fuzzy_T> wow major earth quake in Haiti
[12-Jan-2010 17:15:43] <theacolyte-> eesh
[12-Jan-2010 17:15:47] <theacolyte-> we've had a bunch out here as well
[12-Jan-2010 17:15:53] <theacolyte-> inclduing a 6.5 way north,
[12-Jan-2010 17:17:37] <Fuzzy_T> yeah CA has been a hot spot lately
[12-Jan-2010 17:17:45] * Fuzzy_T is glad I live in TX :)
[12-Jan-2010 17:17:57] <Fuzzy_T> s/I/he
[12-Jan-2010 17:18:19] <nonsenso> hehe. man, i dunno. tx has some weather tho.
[12-Jan-2010 17:18:28] <nonsenso> tornadoes. all those "season" things.
[12-Jan-2010 17:19:26] <theacolyte-> I'm in TX starting Friday
[12-Jan-2010 17:19:36] <nonsenso> where in tx?
[12-Jan-2010 17:19:39] <theacolyte-> Austin
[12-Jan-2010 17:19:56] <nonsenso> i have fam in tx and lived there at some points. heh.
[12-Jan-2010 17:20:08] <nonsenso> going to austin was always a treat.
[12-Jan-2010 17:20:57] <theacolyte-> It's a great city
[12-Jan-2010 17:21:08] <theacolyte-> The only other place in the US that I'd live
[12-Jan-2010 17:31:06] <mrayzenoss> Anyone in Austin feel free to ping me for lunch :)
[12-Jan-2010 21:56:12] <theacolyte-> hmmm just realized zenoss doesn't do up/down for interfaces
[12-Jan-2010 21:57:26] <theacolyte-> Don't worry, I'm reading a long post on how to do it
[12-Jan-2010 21:57:28] <theacolyte-> :)
[12-Jan-2010 22:12:52] <rmatte> there's actually a relatively short post on how to do it :P
[12-Jan-2010 22:13:50] <rmatte> http://community.zenoss.org/docs/DOC-2494
[12-Jan-2010 22:14:07] <rmatte> put the transform/mapping in /Status/IpInterface though
[12-Jan-2010 22:14:11] <rmatte> create it if it doesn't exist
[12-Jan-2010 22:14:26] <rmatte> that'll make it so that the status icon for the interface gets updated on the OS tab
[13-Jan-2010 00:00:30] [disconnected at Wed Jan 13 00:00:30 2010]
[13-Jan-2010 00:00:30] [connected at Wed Jan 13 00:00:30 2010]
[13-Jan-2010 05:17:28] <chickenandbeans> hi
[13-Jan-2010 05:18:07] <chickenandbeans> dmd.Devices.rrdTemplates.mytemplate.datasources.mydataSource.datapoints.mydataSource.rrdtype = "COUNTER"
[13-Jan-2010 05:18:30] <chickenandbeans> The above used to work pre 2.5.1... now it's kind of funky
[13-Jan-2010 05:19:17] <chickenandbeans> One can set the value on via zendmd and print it out ok, but on the GUI, the value remains unchanged
[13-Jan-2010 05:20:04] <chickenandbeans> anyone encountered this?
[13-Jan-2010 05:20:23] <chickenandbeans> Or is it my command?
[13-Jan-2010 08:06:35] <RobertLaptop> Anyone in yet this morning who can help me with the ZenPack competent IRCD?
[13-Jan-2010 08:14:14] <Fuzzy_T> there is an IRCD zenpack?
[13-Jan-2010 08:14:23] <tehhobbit> I get this error from once of our collectors (http://pastie.org/776429) but not the main server zenpack is copied to the collector
[13-Jan-2010 08:17:06] <Fuzzy_T> hmm looks like something went wrong with the DF collector portion or something I have not seeen that but I don't have a large install either.
[13-Jan-2010 08:17:23] <Fuzzy_T> I have no distributed servers only a single vm instance in my lab
[13-Jan-2010 08:19:48] <tehhobbit> the thing is that you cant install zenpacks on the collector thru the webui (it only installs it self on the main server running the webpart) and I suspect that this pack needed something more than just the files
[13-Jan-2010 08:19:54] <tehhobbit> cant figure out what thou
[13-Jan-2010 08:20:29] <tehhobbit> and doc's for distributed monitoring with the core version is to say the least, limited
[13-Jan-2010 08:20:45] <tehhobbit> (since you get that feature with the enterprice version)
[13-Jan-2010 08:24:49] <RobertLaptop> Fuzzy_T there is a ZenPack for IRCD
[13-Jan-2010 08:25:36] <RobertLaptop> Fuzzy_T http://community.zenoss.org/docs/DOC-3490
[13-Jan-2010 08:28:31] <RobertLaptop> I guess it is still a little early yet.
[13-Jan-2010 09:30:09] <rmatte> good morning
[13-Jan-2010 09:30:44] <Apachez> morrn
[13-Jan-2010 09:30:50] <Apachez> but its more of evening over here
[13-Jan-2010 09:30:55] <Apachez> 16:32'ish
[13-Jan-2010 09:31:07] <rmatte> lol
[13-Jan-2010 09:31:15] <rmatte> it's 10:32ish am here
[13-Jan-2010 09:34:08] <willwh> morning :)
[13-Jan-2010 09:34:12] <willwh> it's 7:34am :P
[13-Jan-2010 09:34:20] * willwh yawns
[13-Jan-2010 09:38:55] <markeriv> hey guys, got a question for you all that hopefully wont be too much of a pain
[13-Jan-2010 09:39:31] <markeriv> under reports/performance reports/graph reports i have created a single report, and configured it to auto send out via reportmail daily
[13-Jan-2010 09:39:56] <markeriv> what I would like to do is change this to a monthly setup, but need to figure out how to make the graph show a full month instead of the last 48 hrs
[13-Jan-2010 09:41:21] <markeriv> I can manually toggle it to weekly or monthly mode, however... that doesnt change the automated report
[13-Jan-2010 09:42:23] <chickenandbeans> wow, didn't know you could do that. Great
[13-Jan-2010 09:49:54] <markeriv> well, right now it just shows 48 hrs
[13-Jan-2010 09:50:02] <markeriv> and of course that doesnt satisfy this particular customer
[13-Jan-2010 09:53:06] <willwh> ah - I don't have shell access to the zenoss install I'm running
[13-Jan-2010 09:53:09] <willwh> :x
[13-Jan-2010 09:53:10] <chickenandbeans> :) as always. Don't you love customers? Sorry I can't help man
[13-Jan-2010 09:53:28] <willwh> markeriv: have you looked at the report template in zope?
[13-Jan-2010 09:55:35] <willwh> http://zenhost/zport/dmd/Reports/manage
[13-Jan-2010 09:55:50] <willwh> I don't think there is a link in the gui to manage reports
[13-Jan-2010 09:56:00] <willwh> type it :)
[13-Jan-2010 10:12:36] <rmatte> markeriv: The reports system is a bit weak at the moment, I wouldn't be surprised if there isn't a way to do that
[13-Jan-2010 10:13:05] <willwh> ah
[13-Jan-2010 10:20:44] <markeriv> doh
[13-Jan-2010 10:21:31] <markeriv> zport/dmd/Reports/Manage gives me a dead link
[13-Jan-2010 10:21:46] <markeriv> silly case sensitives
[13-Jan-2010 10:23:04] <markeriv> yea looking through here isnt getting me anywhere
[13-Jan-2010 10:36:21] <rmatte> do manage not Manage
[13-Jan-2010 10:36:23] <rmatte> case sensitive
[13-Jan-2010 10:36:44] <theacolyte-> So, something happened last night, not sure what
[13-Jan-2010 10:36:49] <theacolyte-> I got 150 emails though
[13-Jan-2010 10:37:01] <rmatte> emails for what types of alerts?
[13-Jan-2010 10:37:54] <theacolyte-> Everything being down
[13-Jan-2010 10:38:00] <theacolyte-> Which it definately wasn't
[13-Jan-2010 10:38:07] <rmatte> right, but you dodn't have 150 devices I'm assuming?
[13-Jan-2010 10:38:12] <theacolyte-> No
[13-Jan-2010 10:38:18] <theacolyte-> We're talking like HTTP, SNMP, WMI, everything being down
[13-Jan-2010 10:38:20] <theacolyte-> and it wasn't.
[13-Jan-2010 10:38:21] <rmatte> I'm guessing you had some missed pings
[13-Jan-2010 10:38:24] <theacolyte-> it was all of my sites
[13-Jan-2010 10:38:47] <rmatte> something to do with your management link
[13-Jan-2010 10:39:05] <rmatte> might have had errors across it, or some other type of carrier related issues
[13-Jan-2010 10:39:33] <theacolyte-> tryign to look at that now
[13-Jan-2010 10:39:35] <theacolyte-> seems logical
[13-Jan-2010 10:39:49] <rmatte> that's all it really could have been
[13-Jan-2010 10:40:13] <theacolyte-> The problem with going from no monitoring to having it is you don't know what to expect
[13-Jan-2010 10:40:40] <rmatte> yup
[13-Jan-2010 10:40:51] <theacolyte-> I'm finding all sorts of stuff
[13-Jan-2010 10:40:52] <theacolyte-> it's crazy
[13-Jan-2010 10:41:01] <theacolyte-> Thankfully I won't have to deal with it if I get this new job
[13-Jan-2010 10:41:01] <theacolyte-> LOL
[13-Jan-2010 10:41:05] <rmatte> What kind of line are you using as your management link?
[13-Jan-2010 10:41:16] <theacolyte-> It's a 100mbit colo connection
[13-Jan-2010 10:41:21] <theacolyte-> and not cheap cogent, either
[13-Jan-2010 10:41:22] <rmatte> ah
[13-Jan-2010 10:41:23] <theacolyte-> it's internap
[13-Jan-2010 10:41:40] <rmatte> well, if the problem starts happening more often be sure to get on their asses about it lol
[13-Jan-2010 10:41:50] <theacolyte-> As far as I know, it has never happened
[13-Jan-2010 10:41:57] <theacolyte-> I've never had a single second of downtime in 2.5 years
[13-Jan-2010 10:42:09] <theacolyte-> until last night?
[13-Jan-2010 10:42:11] <rmatte> It's possible that there was some issue in the cloud
[13-Jan-2010 10:42:14] <theacolyte-> it's possible that they lost one of their routers
[13-Jan-2010 10:42:17] <theacolyte-> routes
[13-Jan-2010 10:42:18] <theacolyte-> yeah
[13-Jan-2010 10:42:33] <theacolyte-> All I know is I was woken 3 times last night with a flood of emails
[13-Jan-2010 10:42:40] <theacolyte-> But I was so tired I didn't get up to look LOL
[13-Jan-2010 10:42:58] <rmatte> I remember seeing a good chunk of the UK go down one time when some guys broke in and managed to steal a bunch of fiber cards
[13-Jan-2010 10:43:03] <rmatte> ;)
[13-Jan-2010 10:43:06] <theacolyte-> hehehhee
[13-Jan-2010 10:43:27] <rmatte> BT were scrambling since they didn't have enough fiber cards in the country to replace them
[13-Jan-2010 10:45:25] <theacolyte-> That reminds me of the cut we had out here
[13-Jan-2010 10:45:40] <theacolyte-> Someone deliberatly cut some fiber here in the bay area
[13-Jan-2010 10:45:43] <theacolyte-> 4 differet points
[13-Jan-2010 10:45:45] <theacolyte-> over 1k strands
[13-Jan-2010 10:45:57] <theacolyte-> even cell phones were down in some areas
[13-Jan-2010 10:46:00] <theacolyte-> it made CNN
[13-Jan-2010 10:47:54] <theacolyte-> took them something like 36 hours to splice it back together
[13-Jan-2010 10:48:03] <theacolyte-> s/splice/fuse/g
[13-Jan-2010 10:48:15] <rmatte> wow
[13-Jan-2010 10:48:26] <rmatte> yeh, splicing takes a long time
[13-Jan-2010 10:48:50] <theacolyte-> And the other problem is that once you do it, you will signal loss forever
[13-Jan-2010 10:48:54] <theacolyte-> you can't ever get back to 100%
[13-Jan-2010 10:49:02] <rmatte> yup
[13-Jan-2010 10:49:15] <rmatte> well, depends, fiber doesn't really suffer from "signal loss"
[13-Jan-2010 10:49:30] <rmatte> as long as the fiber is spliced together perfectly you can get back to close to 100% if not 100%
[13-Jan-2010 10:49:32] <theacolyte-> snr
[13-Jan-2010 10:49:37] <rmatte> copper is a different story
[13-Jan-2010 10:49:39] <theacolyte-> close, but never the same though
[13-Jan-2010 10:50:52] <theacolyte-> it was a pretty big disaster though
[13-Jan-2010 10:50:56] <twm1010> any netapp certified people around?
[13-Jan-2010 10:51:12] <theacolyte-> twm1010: The only channel I know of that has netapp people is in #vmware
[13-Jan-2010 10:51:16] <theacolyte-> not sure if anyone is certified though
[13-Jan-2010 10:51:27] <theacolyte-> I'm on equallogic personally, too cheap for netapp
[13-Jan-2010 10:54:35] <rmatte> One of my friends knows a lot about NetApp but he's not actually certified... he manages NetApp and other solutions for NBC Universal
[13-Jan-2010 10:54:46] <theacolyte-> lol
[13-Jan-2010 10:54:56] <theacolyte-> I know a guy that works for NBC universal that does the same thing
[13-Jan-2010 10:54:59] <twm1010> ah, well my ? is actually about how they do their certs... no biggy
[13-Jan-2010 10:55:00] <theacolyte-> What's his first name?
[13-Jan-2010 10:55:06] <rmatte> Derrick
[13-Jan-2010 10:55:11] <theacolyte-> Ah, guy I know is Steve
[13-Jan-2010 10:55:15] <theacolyte-> Out in CA though
[13-Jan-2010 10:55:18] <rmatte> Steve works with him
[13-Jan-2010 10:55:21] <theacolyte-> LOL
[13-Jan-2010 10:55:29] <theacolyte-> Is Steve's last name 4 letters/
[13-Jan-2010 10:55:35] <rmatte> yeh, Derrick was supposed to move to CA but he couldn't get across the border
[13-Jan-2010 10:55:44] <theacolyte-> jesus it's a small world
[13-Jan-2010 10:55:45] <rmatte> Steve is in L.A right?
[13-Jan-2010 10:55:48] <theacolyte-> yeah
[13-Jan-2010 10:55:50] <rmatte> yeah
[13-Jan-2010 10:55:53] <theacolyte-> consultant now though
[13-Jan-2010 10:55:59] <rmatte> yeh, same with Derrick
[13-Jan-2010 10:56:04] <theacolyte-> that's freaking hillarious
[13-Jan-2010 10:56:06] <rmatte> they don't work directly for NBC
[13-Jan-2010 10:56:10] <theacolyte-> yeah
[13-Jan-2010 10:56:22] <chickenandbeans> hi: this is Steve
[13-Jan-2010 10:56:28] <theacolyte-> ?
[13-Jan-2010 10:56:30] <twm1010> LOL
[13-Jan-2010 10:56:30] <chickenandbeans> joke
[13-Jan-2010 10:56:33] <theacolyte-> haha
[13-Jan-2010 10:56:35] <rmatte> Derrick moved to Calgary recently but he's coming back to Ottawa in a little under a year, he just works from home so he can pretty much go wherever he wants
[13-Jan-2010 10:56:55] <rmatte> but yeh, that's funny that you know Steve
[13-Jan-2010 10:57:02] <theacolyte-> I met Steve because he came up here to SF for the vmware class
[13-Jan-2010 10:57:09] <theacolyte-> a vmware class, I suppose
[13-Jan-2010 10:58:48] <rmatte> cool
[13-Jan-2010 10:58:58] <theacolyte-> Steve is EMC certified iirc
[13-Jan-2010 10:59:41] <theacolyte-> Just goes to show, watch out who you piss off on IRC
[13-Jan-2010 10:59:55] <rmatte> yeh, I wasn't too sure about his actual certifications... I've never spoken with him or anything, Derrick just mentions him a lot
[13-Jan-2010 11:00:07] <rmatte> hehe
[13-Jan-2010 11:15:32] <theacolyte-> yeah looks like something did happen last night
[13-Jan-2010 11:16:08] <theacolyte-> I wish I could figure out how to handle up/down monitoring, I followed some doc last night that didn't work
[13-Jan-2010 11:16:18] <theacolyte-> made a copy of the ethernet template and added one
[13-Jan-2010 11:16:24] <theacolyte-> one/threshold
[13-Jan-2010 11:19:14] <twm1010> up/down monitoring?
[13-Jan-2010 11:19:31] <twm1010> rmatte: didn't chet right an article on that?
[13-Jan-2010 11:19:33] <twm1010> using event transforms?
[13-Jan-2010 11:21:17] <rmatte> well, you mean up/down in terms of interface operational status?
[13-Jan-2010 11:21:33] <rmatte> http://community.zenoss.org/docs/DOC-2494
[13-Jan-2010 11:21:48] <rmatte> I use that for all of the interfaces which I monitor
[13-Jan-2010 11:21:57] <theacolyte-> That's the one
[13-Jan-2010 11:22:10] <twm1010> I still wish interfaces worked like winservices
[13-Jan-2010 11:22:27] <theacolyte-> To be honest, I didn't play with it _too_ much
[13-Jan-2010 11:22:28] <twm1010> to where you could flip a zproperty, and then be explicit about what you want to monitor, rather than filter out what you don't want
[13-Jan-2010 11:22:29] <theacolyte-> I was so tired
[13-Jan-2010 11:22:32] <theacolyte-> I may give it another try
[13-Jan-2010 11:22:52] <theacolyte-> I had tried it on a serial interface on one of my routers, that I set admin down, so that may be why
[13-Jan-2010 11:22:56] <theacolyte-> But that does change ifoperstatus
[13-Jan-2010 11:26:15] <theacolyte-> Wow actually something major happened
[13-Jan-2010 11:27:21] <rmatte> I would assume so
[13-Jan-2010 11:27:22] <rmatte> lol
[13-Jan-2010 11:28:40] <theacolyte-> Can't figure out what it is though lol
[13-Jan-2010 11:29:38] <theacolyte-> looks like my vcenter server restarted itself
[13-Jan-2010 11:29:43] <rmatte> yikes
[13-Jan-2010 11:29:57] <theacolyte-> hahahahdjhasd
[13-Jan-2010 11:29:57] <rmatte> we've had that happen once, but then we patched everything and it stopped doing that
[13-Jan-2010 11:30:00] <theacolyte-> I know what it is
[13-Jan-2010 11:30:08] <rmatte> ?
[13-Jan-2010 11:30:14] <theacolyte-> automatic updates
[13-Jan-2010 11:30:18] <rmatte> ohhhh
[13-Jan-2010 11:30:24] <rmatte> badly configured evidently
[13-Jan-2010 11:30:44] <theacolyte-> it's patch tuesday
[13-Jan-2010 11:30:46] <theacolyte-> that's what happened
[13-Jan-2010 11:31:14] <rmatte> but yeh, they don't called it black tuesday for nothing
[13-Jan-2010 11:31:25] <rmatte> s/called/call
[13-Jan-2010 11:31:43] <theacolyte-> I suppose that's what it looks like
[13-Jan-2010 11:34:25] <theacolyte-> That's great
[13-Jan-2010 11:35:29] <theacolyte-> I do have a couple servers that auto reboot by design though
[13-Jan-2010 11:35:31] <theacolyte-> Just 2
[13-Jan-2010 11:42:14] <rmatte> ah
[13-Jan-2010 11:42:21] <rmatte> might want to set maintenance windows then
[13-Jan-2010 11:50:24] <RobertLaptop> Anyone who can help me with the core ZenPack for IRCD http://community.zenoss.org/docs/DOC-3490 ?
[13-Jan-2010 11:54:30] <rmatte> RobertLaptop: well, what are you having problems with?
[13-Jan-2010 11:55:19] <rmatte> RobertLaptop: http://community.zenoss.org/docs/DOC-4018
[13-Jan-2010 11:56:04] <RobertLaptop> Getting the following events in my log.
[13-Jan-2010 11:56:28] <RobertLaptop> Code: 127 - Msg: Command not found
[13-Jan-2010 11:57:12] <RobertLaptop> I have checked. The file exists. I have ran the script manually without issue. I uninstalled and reinstalled the ZenPack.
[13-Jan-2010 11:57:29] <RobertLaptop> Just keep getting the same error over and over again.
[13-Jan-2010 11:57:41] <rmatte> The file may exist, but it could be a permissions issue?
[13-Jan-2010 11:57:58] <rmatte> you were able to run the script as the Zenoss user?
[13-Jan-2010 11:58:39] <RobertLaptop> Yes. It runs without error when ran under the Zenoss User.
[13-Jan-2010 11:58:55] <rmatte> check in zProperties for the device
[13-Jan-2010 11:58:56] <rmatte> zCommandPath
[13-Jan-2010 11:59:01] <rmatte> make sure it's the correct path
[13-Jan-2010 12:00:18] <RobertLaptop> Not correct.
[13-Jan-2010 12:00:34] <rmatte> then that's your issue
[13-Jan-2010 12:00:35] <RobertLaptop> I am fixing now. I am suprissed nothing else brook
[13-Jan-2010 12:00:52] <rmatte> well, zCommandPath isn't used overly often
[13-Jan-2010 12:01:06] <rmatte> hence why it's usually overlooked
[13-Jan-2010 12:01:24] <rmatte> well, time for me to go grab some grub, bbiab
[13-Jan-2010 12:02:05] <RobertLaptop> thanks rmatte
[13-Jan-2010 12:23:47] <RobertLaptop> rmatte didn't fix the issue.
[13-Jan-2010 12:24:23] <RobertLaptop> I can pastebin the full error but the command it says it is running runs without error when ran on the command line.
[13-Jan-2010 12:24:35] <chemist> blist
[13-Jan-2010 13:18:43] <skipzoid> hello
[13-Jan-2010 13:19:28] <skipzoid> everyone snowed in ?
[13-Jan-2010 13:20:51] <theacolyte-> Not much snow in California
[13-Jan-2010 13:21:02] <skipzoid> we've got loads here in the UK
[13-Jan-2010 13:21:10] <theacolyte-> Well, I should say, in the Bay Area
[13-Jan-2010 13:21:13] <theacolyte-> Tahoe got pelnty
[13-Jan-2010 13:25:01] <skipzoid> i'm playing with a tempreture device that is not pollable but can be programmed to telnet to another device and empart the new temp, i'd like to be able to graph this but need a few pointers
[13-Jan-2010 13:44:37] <rmatte> We got dumped on today too
[13-Jan-2010 13:45:24] <rmatte> skipzoid: you'd have to have it telnet to the Zenoss box and store the temp somewhere then have Zenoss grab it from there
[13-Jan-2010 13:47:36] <skipzoid> MMmm, yeah - that would work...
[13-Jan-2010 13:48:22] <skipzoid> i'm still not able to get the satellite graph to show min as well as max - I get an error somethign about RRD file missing etc
[13-Jan-2010 13:56:00] <rmatte> If the RRD file is missing it's because your datapoint is not correct
[13-Jan-2010 13:59:37] <hermes_> hi
[13-Jan-2010 14:00:22] <hermes_> has anyone installed zenoss on FC 10?
[13-Jan-2010 14:02:03] <rmatte> nope, but I'd imagine it's not much different as installing it on RHEL
[13-Jan-2010 14:02:44] <hermes_> the issue that i'm having is with zope.
[13-Jan-2010 14:03:04] <hermes_> it tried to build the database and it just hangs there
[13-Jan-2010 14:05:10] <rmatte> you're installing via RPM?
[13-Jan-2010 14:05:39] <hermes_> yes i did
[13-Jan-2010 14:06:10] <rmatte> I'm used to the stack installer, I've never seen it "build" the zope database
[13-Jan-2010 14:06:35] <rmatte> I just know that the stack installer hangs near the end for a while
[13-Jan-2010 14:06:45] <rmatte> and you have to leave it hang until it finishes
[13-Jan-2010 14:07:58] <hermes_> i'm dl it right now
[13-Jan-2010 14:08:18] <rmatte> the stack installer is probably the easiest way to install it
[13-Jan-2010 14:08:36] <hermes_> is there another app that will do this sort of monitoring?
[13-Jan-2010 14:08:46] <hermes_> i've used Opennms, NMIS, Nagios
[13-Jan-2010 14:08:48] <twm1010> hundreds :)
[13-Jan-2010 14:08:52] <rmatte> tons of them, but Zenoss is the best free open source one which I've found
[13-Jan-2010 14:08:56] <twm1010> if you got the $$ for it ;)
[13-Jan-2010 14:09:49] <rmatte> It's the easiest to customize from what I've found
[13-Jan-2010 14:10:42] <hermes_> cool
[13-Jan-2010 14:12:10] <rmatte> It's got a learning curve like any of them though, so be prepared to read
[13-Jan-2010 14:12:39] <rmatte> As long as you're willing to read the Zenoss Admin Guide from start to finish, I'll answer any other questions you have (I'm in here often)
[13-Jan-2010 14:12:50] <rmatte> there's just a lot of stuff which is commonly asked and fully covered in the admin guide
[13-Jan-2010 14:15:29] <hermes_> Thanks rmatte , I truly appreciate the guidance. i will see how this stack installer goes and continue to read the guide.
[13-Jan-2010 14:18:35] <rmatte> cool
[13-Jan-2010 14:31:17] <theacolyte-> rmatte: fix all of my issues plz
[13-Jan-2010 14:32:06] <theacolyte-> I need more money, that's issue #1
[13-Jan-2010 14:32:33] <twm1010> LO
[13-Jan-2010 14:32:36] <twm1010> L?!
[13-Jan-2010 14:32:42] <twm1010> my brain hurts
[13-Jan-2010 14:32:59] <twm1010> I'm using way too much chat-speak lately.
[13-Jan-2010 14:37:09] <rmatte> lol
[13-Jan-2010 14:37:26] <rmatte> theacolyte-: so do I, I'm still making exactly what I was when I was a senior network analyst
[13-Jan-2010 14:37:40] <rmatte> apparently the leap from that to managing our entire monitoring infrastructure doesn't merrit one
[13-Jan-2010 14:47:25] <theacolyte-> So this is weird
[13-Jan-2010 14:47:33] <theacolyte-> I'm getting an SNMP poll error
[13-Jan-2010 14:47:43] <theacolyte-> But for an OID I'm not monitoring
[13-Jan-2010 14:47:52] <theacolyte-> Is there a way to find out where it's coming from?
[13-Jan-2010 14:48:03] <twm1010> what's the OID?
[13-Jan-2010 14:48:11] <theacolyte-> Error reading value for "Internal RAS Server interface for dial in clients" on dcistsfo55-01.intranet.xxx.com (oid .1.3.6.1.2.1.2.2.1.20.65538 is bad)
[13-Jan-2010 14:48:12] <twm1010> are you sure an interface template isn't hitting it?
[13-Jan-2010 14:48:45] <twm1010> I would guess it's trying to use ethernetCSmacd to try and monitor an interface that is incompatible with it
[13-Jan-2010 14:48:52] <theacolyte-> Ah, it is running RRAS
[13-Jan-2010 14:49:03] <theacolyte-> Wonder how to fix that, since it's not showing up in interfaces
[13-Jan-2010 14:49:06] <twm1010> go to the interface, see what the "type" is, and create a blank template named after that "type" in your root
[13-Jan-2010 14:49:15] <twm1010> uhh... if its not showing up, i'm at a loss :)
[13-Jan-2010 14:49:19] <theacolyte-> yeah
[13-Jan-2010 14:49:22] <theacolyte-> if it was showing up, no problem
[13-Jan-2010 14:50:20] <theacolyte-> Its' wird though
[13-Jan-2010 14:50:25] <theacolyte-> The loopback interface is called
[13-Jan-2010 14:50:29] <twm1010> so you go to the device, and that interface isn't there?
[13-Jan-2010 14:50:30] <theacolyte-> Internal loopback interface for 127.0.0 network
[13-Jan-2010 14:50:47] <theacolyte-> Let me try remodeling
[13-Jan-2010 14:51:03] <theacolyte-> yeah, nothing... odd
[13-Jan-2010 14:51:50] <theacolyte-> MISSING RRD FILE:
[13-Jan-2010 14:53:20] <theacolyte-> I dunno... I'll search around
[13-Jan-2010 14:53:28] <theacolyte-> It is running RRAS, and RRAs does add an interface
[13-Jan-2010 14:54:03] <theacolyte-> But not under network connections, so it could trip it up
[13-Jan-2010 14:54:30] <twm1010> walk the device's network OID tree, and see if it shows up as an interface
[13-Jan-2010 14:54:37] <twm1010> don't ask me what part of the tree that is :)
[13-Jan-2010 14:54:55] <twm1010> it will either be enumerated like any other nic, or it won't
[13-Jan-2010 14:54:57] <theacolyte-> I was just about to ask
[13-Jan-2010 14:54:59] <theacolyte-> :P
[13-Jan-2010 14:55:07] <theacolyte-> it's near the top
[13-Jan-2010 14:59:54] <theacolyte-> well
[13-Jan-2010 14:59:54] <Buklov> Hello
[13-Jan-2010 14:59:56] <theacolyte-> I found the issue
[13-Jan-2010 15:00:04] <theacolyte-> that OID doesn't exist... but why is it trying to poll for it
[13-Jan-2010 15:03:17] <Buklov> I have not get to work ZenPacks.community.WMIPerf_Windows
[13-Jan-2010 15:03:18] <Buklov> ZenPacks.community.WMIPerf_Windows and ZenPacks.community.WMIDataSource usanvilis, but the graphics are not built
[13-Jan-2010 15:03:53] <Buklov> anyone can help me?
[13-Jan-2010 15:04:34] <Buklov> missing rrd file
[13-Jan-2010 15:04:50] <theacolyte-> I would help you, but I've actually had the exact same issue.
[13-Jan-2010 15:05:05] <theacolyte-> http://dev.zenoss.org/trac/ticket/6019
[13-Jan-2010 15:05:08] <theacolyte-> apparently I'm not crazy.
[13-Jan-2010 15:05:12] <theacolyte-> rmatte: ^^^^^
[13-Jan-2010 15:06:20] <Buklov> :)
[13-Jan-2010 15:06:46] <theacolyte-> Buklov: Are these templates you are creating?
[13-Jan-2010 15:06:51] <theacolyte-> Datapoints/graphs that YOU are creating?
[13-Jan-2010 15:07:54] <Buklov> if i press test, Executing command wmic -U 'tcdon\buklov'%'*****' //10.0.0.241 --namespace='root\cimv2' "SELECT AvailableKBytes, PagesPerSec FROM Win32_PerfRawData_PerfOS_Memory" against ts00.tcdon.ru
[13-Jan-2010 15:07:55] <Buklov> CLASS: Win32_PerfRawData_PerfOS_Memory
[13-Jan-2010 15:07:55] <Buklov> AvailableKBytes|PagesPersec
[13-Jan-2010 15:07:55] <Buklov> 14555720|17106824
[13-Jan-2010 15:07:55] <Buklov> DONE in 0 seconds
[13-Jan-2010 15:07:59] <Buklov> all ok
[13-Jan-2010 15:08:02] <theacolyte-> yup
[13-Jan-2010 15:08:10] <theacolyte-> that's tbhe bug.
[13-Jan-2010 15:08:12] <Buklov> bat no graph
[13-Jan-2010 15:08:17] <theacolyte-> no that it helps you
[13-Jan-2010 15:08:28] <theacolyte-> No one in here seems to be able to resolve it
[13-Jan-2010 15:08:31] <theacolyte-> See my link above
[13-Jan-2010 15:09:32] <Buklov> :(
[13-Jan-2010 15:11:33] <Buklov> I thought that I did something wrong
[13-Jan-2010 15:14:21] <theacolyte-> Sorry about that Buklov, but I don't think you have
[13-Jan-2010 15:14:27] <theacolyte-> Buklov: I have s een one thing work
[13-Jan-2010 15:14:34] <bigegor> theacolyte-:what name of you datasource
[13-Jan-2010 15:14:35] <theacolyte-> Buklov: If you delete the device and readd it, it should fix it
[13-Jan-2010 15:14:42] <theacolyte-> bigegor: I made my own in this case
[13-Jan-2010 15:14:57] <theacolyte-> Oh you mean the actual class?
[13-Jan-2010 15:15:02] <theacolyte-> bigegor: http://dev.zenoss.org/trac/ticket/6019
[13-Jan-2010 15:15:04] <bigegor> how it looks like?
[13-Jan-2010 15:15:25] <theacolyte-> bigegor: for what it's worth I don't think this is your pack
[13-Jan-2010 15:15:28] <theacolyte-> I think it's zenoss
[13-Jan-2010 15:15:32] <theacolyte-> let me post up a screenshot
[13-Jan-2010 15:15:48] <Buklov> bigegor i can give you web access
[13-Jan-2010 15:15:52] <rmatte> It may be Zenoss, but we might be able to identify the root cause with enough details
[13-Jan-2010 15:16:08] <bigegor> Buklov:no
[13-Jan-2010 15:16:47] <Buklov> bigegor ïî ðóññêè ïîíèìàåòå?
[13-Jan-2010 15:16:54] <theacolyte-> screenshots incoming
[13-Jan-2010 15:17:17] <bigegor> i have strange things if name of datasource contain '_' symbols
[13-Jan-2010 15:17:22] <theacolyte-> http://dev.zenoss.org/trac/attachment/ticket/6019/missing%20RRD.PNG
[13-Jan-2010 15:17:24] <theacolyte-> That's the missing RRD
[13-Jan-2010 15:17:52] <theacolyte-> http://dev.zenoss.org/trac/attachment/ticket/6019/missing%20RRD%202.PNG
[13-Jan-2010 15:17:55] <theacolyte-> That's my template
[13-Jan-2010 15:18:38] <Buklov> bigegor Do you speak Russian?
[13-Jan-2010 15:18:49] <bigegor> yes
[13-Jan-2010 15:19:24] <theacolyte-> bigegor: The 2nd file there shows you my queries/datasources
[13-Jan-2010 15:19:42] <theacolyte-> Those queries work (tested)
[13-Jan-2010 15:20:22] <Buklov> bigegor as there may talk in Russian? I understand you do not see the message if I write them in Russian
[13-Jan-2010 15:20:45] <Buklov> if i possible
[13-Jan-2010 15:20:50] <Buklov> *it
[13-Jan-2010 15:21:12] <bigegor> Buklov:yes
[13-Jan-2010 15:21:53] <bigegor> Buklov: you can write me PM in Zenoss Forum.
[13-Jan-2010 15:22:39] <theacolyte-> rmatte: Are you doing apache proxy for SSL?
[13-Jan-2010 15:22:47] <Buklov> ok
[13-Jan-2010 15:22:52] <Buklov> tanks
[13-Jan-2010 15:22:58] <Buklov> *thanks
[13-Jan-2010 15:23:30] <rmatte> theacolyte-: squid proxy I believe
[13-Jan-2010 15:23:36] <theacolyte-> ah
[13-Jan-2010 15:23:52] <theacolyte-> By the way did you look at my 2nd screenshot there? Maybe you can see something overtly wrong?
[13-Jan-2010 15:24:08] <rmatte> checking now
[13-Jan-2010 15:25:34] <rmatte> not as far as I can tell
[13-Jan-2010 15:25:50] <theacolyte-> yeah
[13-Jan-2010 15:25:55] <theacolyte-> I really think it has to be a bug
[13-Jan-2010 15:26:02] <theacolyte-> I'm glad (sad?) to see someone else running into it
[13-Jan-2010 15:26:44] <rmatte> theacolyte-: haven't had a chance to spin up another vm and try it with 2.4.5?
[13-Jan-2010 15:27:00] <rmatte> it would really help, because if you don't see the same thing with 2.4.5 that means it's something which was changed recently
[13-Jan-2010 15:27:04] <rmatte> which would narrow it down
[13-Jan-2010 15:27:25] <theacolyte-> Nah, I won't til I come back from vacation, I'm done at 5:30 today
[13-Jan-2010 15:27:29] <Buklov> I just installed zenoss wmi plugins added, installed zWinUser, zWinPassword and added a host. Maybe I have something not done?
[13-Jan-2010 15:27:30] <rmatte> k
[13-Jan-2010 15:27:33] <theacolyte-> Just finished a quick knowledge transfer
[13-Jan-2010 15:27:46] <theacolyte-> Buklov: You'd know if it was a password issue, you wouldn't be getting missing RRD files
[13-Jan-2010 15:27:57] <twm1010> Buklov: WMI plugins?
[13-Jan-2010 15:27:57] <theacolyte-> Buklov: Have you verified that the RRD files are, in fact missing?
[13-Jan-2010 15:27:59] <rmatte> yeh, you'd see an event access denies
[13-Jan-2010 15:28:02] <rmatte> denied*
[13-Jan-2010 15:28:08] <Buklov> twm1010 yes
[13-Jan-2010 15:28:22] <theacolyte-> twm1010: The problem isn't with WMI itself, it's something else
[13-Jan-2010 15:28:31] <theacolyte-> The queries work fine, the RRD files are not being created
[13-Jan-2010 15:29:08] <twm1010> i've never had to install a wmi plugin
[13-Jan-2010 15:29:10] <twm1010> ?
[13-Jan-2010 15:29:28] <cparlette> did you customize the RRD create command?
[13-Jan-2010 15:29:36] <theacolyte-> I don't even know how, so in my case, no
[13-Jan-2010 15:29:52] <cparlette> i think there's a problem with custom rrd create commands on data sources if the rrd file doesnt exist yet
[13-Jan-2010 15:29:59] <theacolyte-> for those of you that haven't seen it: http://dev.zenoss.org/trac/ticket/6019
[13-Jan-2010 15:30:17] <cparlette> if this issue is specific to that community pack, opening a trac ticket won't help
[13-Jan-2010 15:30:30] <twm1010> Buklov: I'm trying to make sure you're not following old documentation that tells you to setup a host to proxy to device over WMI.
[13-Jan-2010 15:31:06] <theacolyte-> cparlette: I don't see how it could be an issue with WMI itself
[13-Jan-2010 15:31:06] <bigegor> theacolyte-:what can you see, if you start:zenperfwmi run -v10 -d yourserver
[13-Jan-2010 15:31:12] <theacolyte-> bigegor: no problems
[13-Jan-2010 15:31:16] <theacolyte-> ran that for awhile yesterday
[13-Jan-2010 15:37:23] <Buklov> twm1010 i simle install zenos core 2.5.1, install wmi plugins, set user and pass, and add host, nothing more.
[13-Jan-2010 15:48:28] <bigegor> Buklov: you need to bind WMIDevice Prformance template to device or deviceclass
[13-Jan-2010 15:49:29] <theacolyte-> bigegor: I don't think that's the issue, he wouldn't see that MISSING RRD FILE thing unless you bound it
[13-Jan-2010 15:51:37] <bigegor> theacolyte-: but i want to see how looks values before sending to RRD.
[13-Jan-2010 15:52:00] <Buklov> bigegor How? Can you explain the steps?
[13-Jan-2010 15:52:03] <theacolyte-> ah, he said it above
[13-Jan-2010 15:52:17] <theacolyte-> 12:53 < Buklov> CLASS: Win32_PerfRawData_PerfOS_Memory
[13-Jan-2010 15:52:17] <theacolyte-> 12:53 < Buklov> AvailableKBytes|PagesPersec
[13-Jan-2010 15:52:17] <theacolyte-> 12:53 < Buklov> 14555720|17106824
[13-Jan-2010 15:52:17] <theacolyte-> 12:53 < Buklov> DONE in 0 seconds
[13-Jan-2010 15:53:40] <Buklov> bigegor If I understood correctly, it automatically added?
[13-Jan-2010 15:59:19] <bigegor> Buklov:what templates you can seen in Templates page of your server?
[13-Jan-2010 15:59:26] <Buklov> add ts00 http://pastebin.freeswitch.org/11782
[13-Jan-2010 16:00:08] <Buklov> Device /Devices Basic template that only collects sysUpTime.
[13-Jan-2010 16:00:08] <Buklov> IpService /Devices Place holder for future use.
[13-Jan-2010 16:00:08] <Buklov> OSProcess /Devices Monitors for OSProcess object
[13-Jan-2010 16:00:08] <Buklov> WinService /Devices Place holder for future use.
[13-Jan-2010 16:00:08] <Buklov> ethernetCsmacd /Devices Standard ethernet interface template with 75% utilization threshold
[13-Jan-2010 16:00:09] <Buklov> ethernetCsmacd_64 /Devices Template for 64-bit interface counters. Must use SNMP v2c for it to work.
[13-Jan-2010 16:00:53] <bigegor> Buklov: unbind Devices and bind WMIDevices
[13-Jan-2010 16:09:06] <rmatte> nah, he's pasting that right from /Devices not the device page itself
[13-Jan-2010 16:09:15] <rmatte> the device templates page would never show the ethernet templates
[13-Jan-2010 16:09:35] <Buklov> bigegor result the same
[13-Jan-2010 16:13:23] <bigegor> Buklov: after binding of Template, you can see result after 2 scan cycle (10min)
[13-Jan-2010 16:15:22] <Buklov> now i see graphics!
[13-Jan-2010 16:15:30] <Buklov> thanks
[13-Jan-2010 16:15:46] <rmatte> ah ok, so he just didn't have the right template bound?
[13-Jan-2010 16:18:14] <bigegor> template was ok.
[13-Jan-2010 16:20:47] <Buklov> but where is given by the frequency of the survey?
[13-Jan-2010 16:21:05] <bigegor> can somebody play a bit with new WMIDataSource ZenPack? I realy need feedback.
[13-Jan-2010 16:21:56] <rmatte> I don't really do any WMI monitoring right now, so I don't have any devices to test actual functionality with, or I would
[13-Jan-2010 16:22:07] <Buklov> I want to get value for a given interval?
[13-Jan-2010 16:22:23] <rmatte> Buklov: it does it every 5 minutes
[13-Jan-2010 16:24:22] <Buklov> i want 30 sec
[13-Jan-2010 16:24:47] <rmatte> Buklov: ummm, that's pretty intensive
[13-Jan-2010 16:28:39] <rmatte> What's the issue with only collecting every 5 minutes?
[13-Jan-2010 16:29:21] <Apachez> http://www.youtube.com/watch?v=4gPJHKTl2e4 BBC Reporter breaks 'unbreakable' mobile phone at CES
[13-Jan-2010 16:31:10] <rmatte> yeh, but it was the screen, what do they expect
[13-Jan-2010 16:31:21] <rmatte> it'd have to be made of bullet proof material
[13-Jan-2010 16:32:45] <theacolyte-> son of a bitch
[13-Jan-2010 16:32:50] <theacolyte-> I guess that's not my problem then
[13-Jan-2010 17:25:58] rhettardo_ is now known as rhettardo
[13-Jan-2010 17:51:49] <sjobeck> hey hi all. hope al are well. quick Q. does the community edition allow there to be proxy servers in remote offices & relay status to main server?
[13-Jan-2010 18:14:38] Fuzzy_T_ is now known as Fuzzy_T
[13-Jan-2010 18:15:04] <sjobeck1> can the community edition do "proxy" monitor servers?
[14-Jan-2010 00:00:30] [disconnected at Thu Jan 14 00:00:30 2010]
[14-Jan-2010 00:00:31] [connected at Thu Jan 14 00:00:31 2010]
[14-Jan-2010 06:10:24] <ShaneKu> anyone available to attempt to help me through some snmpv3 issues between my 2.4.5 zenoss appliance and freebsd 7/8 machines? snmpwalk and zenmodeler are working fine, but zenperfsnmp is timing out, tcpdumps aren't giving me much info...wondering where else to look for clues.
[14-Jan-2010 06:11:12] <tehhobbit> how do I create agregated reports, started with a multigraph report added all ports on a switch , choosed stacked, but the max value still isnt higher than the the max for a single port ( doesnt seem to stack )
[14-Jan-2010 06:41:51] <ShaneKu> anyone know where I can get a 2.5.2 vmware appliance?
[14-Jan-2010 07:20:55] <ShaneKu> to add on to my question above it is only affecting snmpv3 I just enabled v2c on one of my servers and zenperfsnmp can successfully connect. Does zenperfsnmp not support snmpv3 on Zenoss 2.4.5?
[14-Jan-2010 07:34:33] <Dieterbe_> we are doing maintainance work on some servers. zenoss gives critical events. i created maintainance windows for all the servers spanning the entire day (and yes, i set 'enabled' flag to true), but the events didn't go away
[14-Jan-2010 07:34:49] <Dieterbe_> i tried acking them and moving them to the history, but new events keep on being generated..??
[14-Jan-2010 08:17:13] <Dieterbe_> mrayzenoss: maybe you know this
[14-Jan-2010 08:17:18] <Dieterbe_> we are doing maintainance work on some servers. zenoss gives critical events. i created maintainance windows for all the servers spanning the entire day (and yes, i set 'enabled' flag to true), but the events didn't go away
[14-Jan-2010 08:17:23] <Dieterbe_> i tried acking them and moving them to the history, but new events keep on being generated..??
[14-Jan-2010 08:18:02] <mrayzenoss> if you go to those devices, which production state are they listed in?
[14-Jan-2010 08:18:21] <mrayzenoss> I assume they're supposed to be in "Maintenance" or something besides "Production"?
[14-Jan-2010 08:19:16] <Dieterbe_> State Maintenance
[14-Jan-2010 08:19:58] <Dieterbe_> yes they are supposed to be in state maintenance
[14-Jan-2010 08:20:11] <mrayzenoss> Dieterbe_: according to the admin guide, they should be eventing, just not alerting
[14-Jan-2010 08:20:11] <Dieterbe_> but we are still getting events
[14-Jan-2010 08:20:23] <mrayzenoss> "Maintenance - you want monitoring and collection to occur, and maybe or maybe not the device on the dashboard, just not alerting occurring"
[14-Jan-2010 08:20:51] <Dieterbe_> oh maybe that should be clarified then. i don't see the word 'event' in that string
[14-Jan-2010 08:20:56] <mrayzenoss> yeah
[14-Jan-2010 08:21:12] <mrayzenoss> the 3 things are monitoring, alerting and dashboard
[14-Jan-2010 08:21:15] <mrayzenoss> monitoring = events
[14-Jan-2010 08:21:55] <Dieterbe_> so when you have a NOC with a display that shows the event console, you are supposed to set devices to 'decomissioned' ?
[14-Jan-2010 08:21:56] <mrayzenoss> Pre-Production would be monitoring, but not on dashboards or alerting
[14-Jan-2010 08:22:18] <mrayzenoss> if you don't want them monitored
[14-Jan-2010 08:23:06] <Dieterbe_> the nomenclature feels a bit wrong, but yeah
[14-Jan-2010 08:23:24] <mrayzenoss> I'd agree that we need to say that monitoring means events
[14-Jan-2010 08:25:45] <Dieterbe_> would be useful also to have the ability to define custom maintenance states
[14-Jan-2010 08:25:52] <Dieterbe_> err production states
[14-Jan-2010 08:25:53] <mrayzenoss> I believe you do
[14-Jan-2010 08:26:36] <mrayzenoss> just a sec, let me find it
[14-Jan-2010 08:27:47] <mrayzenoss> on the Settings tab
[14-Jan-2010 08:27:51] <mrayzenoss> State Conversions
[14-Jan-2010 08:28:26] <mrayzenoss> hmmâ€¦ need to find more documentation
[14-Jan-2010 08:29:28] <Dieterbe_> i meant more like a page where you could define a lot of properties for each state (events enabled y/n, alerting enabled y/n, affect uptime counter y/n, etc)
[14-Jan-2010 08:29:59] <mrayzenoss> I believe that is planned, just not sure if it's in the next release or not
[14-Jan-2010 08:30:05] <Dieterbe_> k np
[14-Jan-2010 08:30:05] <mrayzenoss> a lot of UI changes are coming
[14-Jan-2010 08:30:49] <Dieterbe_> so i just set the servers to state decomissioned, moved a lot of events manually to history. but now new events are coming in again
[14-Jan-2010 08:30:54] <Dieterbe_> the whole event log is red again :(
[14-Jan-2010 08:31:00] <Dieterbe_> event console*
[14-Jan-2010 08:41:30] <Dieterbe_> any idea, mrayzenoss ?
[14-Jan-2010 08:46:01] <mrayzenoss> Dieterbe_: what version?
[14-Jan-2010 08:46:17] <Dieterbe_> 2.4.5
[14-Jan-2010 08:48:52] <mrayzenoss> are you sure the devices are in the "Decommissioned" production state?
[14-Jan-2010 08:55:10] <Dieterbe_> i'll double check hold on
[14-Jan-2010 08:55:46] <Dieterbe_> oh this is odd. so i updated all maintenance states to say decommisioned instead of maintenance
[14-Jan-2010 08:56:00] <Dieterbe_> but when checking device->status, they are still in maintenance
[14-Jan-2010 09:00:44] <Dieterbe_> hmm and setting the mainteance period to disabled and then enabled again doesn't help :(
[14-Jan-2010 09:02:40] <chickenandbeans> Afternoon
[14-Jan-2010 09:03:32] <Dieterbe_> i'll have one portion please
[14-Jan-2010 09:05:20] <mrayzenoss> Dieterbe_: are they in maintenance windows or did you set the state manually?
[14-Jan-2010 09:06:15] <Dieterbe_> maintenance windows
[14-Jan-2010 09:07:50] <mrayzenoss> overlapping maintenance windows can be troublesome
[14-Jan-2010 09:08:26] <Dieterbe_> nono, all servers only have 1 maintenance window. from today 00:00 , with a length of 2 days
[14-Jan-2010 09:11:50] <Dieterbe_> first i set it to maintenance, but that was not i wanted, so now i changed it to decomissioned
[14-Jan-2010 09:12:06] <Dieterbe_> ..in the production state.
[14-Jan-2010 09:12:16] <Dieterbe_> but the servers are still "stuck" in the maintenance state
[14-Jan-2010 09:34:36] <Dieterbe_> mrayzenoss: i just removed all the maintenance windows and recreated them, making sure i set it to decommisioned
[14-Jan-2010 09:34:54] <Dieterbe_> now at least the device state is correct
[14-Jan-2010 09:35:06] <Dieterbe_> lets hope the events don't come back now after i moved them to the history..
[14-Jan-2010 09:37:04] <Dieterbe_> i think it works \o/ i'll report a bug for that issue though
[14-Jan-2010 09:42:50] <mrayzenoss> Dieterbe_: http://dev.zenoss.org/trac/ticket/6048
[14-Jan-2010 09:43:57] <Dieterbe_> heh. isn't it easier for you to just update the text?
[14-Jan-2010 09:44:23] <mrayzenoss> nah, our doc writer keeps tight control
[14-Jan-2010 09:44:39] <mrayzenoss> since there are multiple branches of the documentation
[14-Jan-2010 09:45:32] <willwh> mornin' lads :)
[14-Jan-2010 09:45:54] <Dieterbe_> git makes that stuff really easy. eg you'd commit the chance and then the doc writer can pull branches or even individual commits (!) and merge them into his code base
[14-Jan-2010 09:46:00] <Dieterbe_> change*
[14-Jan-2010 09:48:26] <mrayzenoss> yeah, we keep talking about switching to git
[14-Jan-2010 09:52:12] <rmatte> yeh, I have really bad luck with maintenance windows to the point where I just move devices in and out of maintenance status by hand
[14-Jan-2010 09:54:43] <Dieterbe_> mrayzenoss: oopsie. i clicked submit a bit too soon. can you make the title something more meaningfull? http://dev.zenoss.org/trac/ticket/6049
[14-Jan-2010 09:54:58] <mrayzenoss> what do you want it to be?
[14-Jan-2010 09:55:06] <Dieterbe_> ie "changing start production state of maintenance window after it's enabled doesn't change state" or something
[14-Jan-2010 09:55:48] <mrayzenoss> ohâ€¦ looking at the ticket that's gonna get closed
[14-Jan-2010 09:55:59] <mrayzenoss> maintenance windows can't be updated once they've started
[14-Jan-2010 09:56:13] <mrayzenoss> they're really just triggers
[14-Jan-2010 09:56:22] <mrayzenoss> which is why overlapping windows can be confusing
[14-Jan-2010 09:57:21] <Dieterbe_> sounds like the implementation of the maintenance windows should be fixed then, instead of closing valid tickets because "it doesn't work with how maintenance windows are implemented"
[14-Jan-2010 09:58:09] <mrayzenoss> I could see that view
[14-Jan-2010 09:59:46] <mrayzenoss> I've updated the title, I imagine it will either get closed or changed to an enhancement, since it's changing the way the windows work
[14-Jan-2010 10:00:04] <mrayzenoss> rather than update an existing window, you could just delete it and add another
[14-Jan-2010 10:00:20] <mrayzenoss> but then it would lose your previous state...
[14-Jan-2010 10:00:27] <mrayzenoss> if it wasn't Production
[14-Jan-2010 10:02:46] <tehhobbit> mrayzenoss: do you know how to spec witch devs should be included in aggregated reports ( defaults to .*)
[14-Jan-2010 10:07:45] <rmatte> mrayzenoss: the reason they need to be editable on the fly is because sometimes a change can drag on, so you'd want to extend the window, or perhaps it finished ahead of schedule and you want monitoring back up asap
[14-Jan-2010 10:08:17] <mrayzenoss> I understand the need, I'm just saying you can add new windows which supersede them
[14-Jan-2010 10:08:40] <rmatte> Is that documented anywhere?
[14-Jan-2010 10:08:46] <rmatte> I don't recall reading that anywhere
[14-Jan-2010 10:08:59] <mrayzenoss> I'll lobby to get it marked as an enhancement
[14-Jan-2010 10:09:07] <rmatte> cool
[14-Jan-2010 10:09:10] <rocket> mrayzenoss: I would argue that way isnt very intuitive to an admin who is used to being edit windows in other products ..
[14-Jan-2010 10:09:46] <rmatte> yup
[14-Jan-2010 10:11:11] <mrayzenoss> yeah, but if you've past the Start Production State, is it intuitive that you should be able to change it after the fact?
[14-Jan-2010 10:11:33] <mrayzenoss> s/past/passed/
[14-Jan-2010 10:12:09] <rmatte> If you're past the start date you should have the ability to shorten or extend the end date/time
[14-Jan-2010 10:12:21] <rmatte> It would be extremely intuitive
[14-Jan-2010 10:13:03] <rmatte> For example, change management in many ticketing systems allows you to extend the window of a change if necessary
[14-Jan-2010 10:13:11] <mrayzenoss> I agree that changing the End should be allowed
[14-Jan-2010 10:13:13] <rmatte> so having the ability to do the same with maintenance windows would be lovely
[14-Jan-2010 10:13:30] <rmatte> It shouldn't allow the start to be edited once the window has "kicked off"
[14-Jan-2010 10:13:43] <rmatte> it's really the end date that people would be interested in
[14-Jan-2010 10:14:08] <mrayzenoss> well that's what he was asking for, editing the start production state
[14-Jan-2010 10:14:25] <mrayzenoss> after it's "kicked off"
[14-Jan-2010 10:14:34] <rmatte> oh, I don't see the point in that
[14-Jan-2010 10:14:45] <rmatte> Dieterbe_: explain what the benefit would be in your eyes
[14-Jan-2010 10:15:00] <mrayzenoss> rmatte: http://dev.zenoss.org/trac/ticket/6049
[14-Jan-2010 10:15:19] <rocket> rmatte: the only benefit I can see with that is if its a repeating maintenance window and the admin is trying to change the start time for following windows ..
[14-Jan-2010 10:15:28] <rmatte> mrayzenoss: another thing that I would absolutely LOVE to see for production windows would be a link that you could click to immediately end the maintenance mode and revert everything back to previous state
[14-Jan-2010 10:15:49] <rmatte> I'm having to do it all by hand right now]
[14-Jan-2010 10:16:17] <mrayzenoss> rmatte: that might have a enhancement already
[14-Jan-2010 10:16:37] <rmatte> oh, no
[14-Jan-2010 10:16:43] <rmatte> he's not even talking about start date
[14-Jan-2010 10:16:58] <rmatte> he's talking about being able to edit the state while the window is in effect
[14-Jan-2010 10:17:07] <mrayzenoss> right
[14-Jan-2010 10:17:14] <rmatte> which is stupid, because you should be intelligent enough to set it to the right state in the first place when you create the window
[14-Jan-2010 10:17:38] <rmatte> plus, what's the difference between Decommissioned and Maintenance really?
[14-Jan-2010 10:17:43] <rmatte> they both do the same thing
[14-Jan-2010 10:17:47] <mrayzenoss> Maintenance still monitors
[14-Jan-2010 10:17:57] <rmatte> really? thought it didn't
[14-Jan-2010 10:17:58] <mrayzenoss> and generates events, which started this whole discussion
[14-Jan-2010 10:18:23] <rmatte> I moved stuff in to maintenance on Tuesday night since we took every one of our customer tunnels down...
[14-Jan-2010 10:18:26] <rmatte> I didn't see a single event
[14-Jan-2010 10:18:34] <mrayzenoss> there are 3 things available in Production Statesâ€¦ appearing on the dashboard, generating alerts, and monitoring
[14-Jan-2010 10:18:48] <mrayzenoss> decommissioned turns everything off
[14-Jan-2010 10:18:52] <rmatte> I see
[14-Jan-2010 10:18:54] <mrayzenoss> maintenance still monitors, but no alerts
[14-Jan-2010 10:18:58] <mrayzenoss> or dashboard
[14-Jan-2010 10:19:33] <rmatte> well either way, I don't agree with being able to change state during a window, I was lobbying for being able to change the end time and having a button/link to instantly kill the maintenance window and revert to original state
[14-Jan-2010 10:19:39] <rmatte> I think those 2 things would actually be useful
[14-Jan-2010 10:20:04] <mrayzenoss> why don't you open an enhancement ticket? that sounds useful
[14-Jan-2010 10:20:55] <rmatte> I shall
[14-Jan-2010 10:21:05] <mrayzenoss> hmmâ€¦ I just edited a maintenance window in progress, shortened the end time
[14-Jan-2010 10:21:10] <mrayzenoss> will know if that works soon
[14-Jan-2010 10:21:22] <rmatte> It didn't the few times that I tried it
[14-Jan-2010 10:21:29] <rmatte> let me know if it does
[14-Jan-2010 10:21:32] <mrayzenoss> 2.5.1 :)
[14-Jan-2010 10:21:36] <rmatte> yeh, I know
[14-Jan-2010 10:21:38] <rmatte> that's why lol
[14-Jan-2010 10:21:51] <mrayzenoss> but I do like the "Quick Cancel"
[14-Jan-2010 10:22:10] <rmatte> Yeh, the Quick Cancel would be super useful
[14-Jan-2010 10:22:15] <rmatte> I'd use it all the time
[14-Jan-2010 10:22:59] <rocket> rmatte: can you change a window before its kicked off?
[14-Jan-2010 10:23:13] <mrayzenoss> maintenance window start/stops generate events as of 2.5 I believe
[14-Jan-2010 10:23:13] <rmatte> rocket: I'm not even sure of that, never tried it
[14-Jan-2010 10:23:17] <rmatte> rocket: I would hope so
[14-Jan-2010 10:23:44] <rocket> unfortunately our maintenance window changes every weekend .. its not static .. :/
[14-Jan-2010 10:24:18] <mrayzenoss> you can definitely edit them before they start, I'm testing if you can edit them once they've started
[14-Jan-2010 10:25:07] <mrayzenoss> and after that I guess I'll test what happens when you delete one in progress
[14-Jan-2010 10:25:11] <rmatte> someone just walked in to my office with a wireless router connected to an ethernet chord stretching all the way back to his office across the room, placed it above my desk and stated "there, you are now an access point"
[14-Jan-2010 10:25:17] * rmatte thinks wtf
[14-Jan-2010 10:25:48] <ckrough> whats the code name for 2.5? Is stone crab 2.5 or 2.6?
[14-Jan-2010 10:25:54] <rmatte> King Crab is 2.5
[14-Jan-2010 10:25:57] <rmatte> 2.6 is Stone Crab
[14-Jan-2010 10:26:00] <mrayzenoss> 2.5 is King Crab, Stone Crab is next
[14-Jan-2010 10:26:04] <ckrough> thanks
[14-Jan-2010 10:26:21] <theacolyte-> mrayzenoss: How did the whole defect thing go?
[14-Jan-2010 10:26:39] <theacolyte-> I'd love to know about my area graph question - but more importantly that missing rrd file one I filed
[14-Jan-2010 10:26:56] <rmatte> mrayzenoss: actually, if deleting one in progress reverted everything back instantly, that'd be nice too
[14-Jan-2010 10:27:09] <Dieterbe_> rmatte> Dieterbe_: explain what the benefit would be in your eyes <-- fixing mistakes
[14-Jan-2010 10:27:26] <mrayzenoss> rmatte: yeah
[14-Jan-2010 10:27:35] <mrayzenoss> theacolyte-: it's on the list for today's defect review: http://dev.zenoss.org/trac/report/1
[14-Jan-2010 10:28:15] <mrayzenoss> ok, I started a 1 hour maintenance window and after it started I updated the end time to 10 minutes.. Worked fine.
[14-Jan-2010 10:28:25] <mrayzenoss> everything just went back to Production
[14-Jan-2010 10:28:33] <theacolyte-> Cool
[14-Jan-2010 10:28:40] <theacolyte-> my RRD one didn't make it, but I an wait
[14-Jan-2010 10:28:48] <mrayzenoss> theacolyte-: it's on the list
[14-Jan-2010 10:28:49] <theacolyte-> I'm on vacation actually, starting yesterday until the 15th
[14-Jan-2010 10:28:50] <theacolyte-> hehe
[14-Jan-2010 10:28:56] <rmatte> mrayzenoss: cool
[14-Jan-2010 10:29:08] <theacolyte-> The transparency one made it
[14-Jan-2010 10:29:11] <rmatte> mrayzenoss: If you start one at 5 minutes and extend it to 10, does it work as well?
[14-Jan-2010 10:29:23] <mrayzenoss> rmatte: good question, I guess I can test that too
[14-Jan-2010 10:29:31] <theacolyte-> http://dev.zenoss.org/trac/ticket/6019
[14-Jan-2010 10:30:26] <rmatte> does anyone have sort of a collector plugin template or some good easy examples laying around?
[14-Jan-2010 10:30:51] <theacolyte-> brb, gotta get ready for interview
[14-Jan-2010 10:31:45] <rmatte> I'm determined to get my SNMP Performance ZenPack using a proper collector plugin in the interest of performance
[14-Jan-2010 10:32:11] <rmatte> launching 100 bash scripts every 5 minutes isn't very resource friendly
[14-Jan-2010 10:32:58] <mrayzenoss> deleted a maintenance window in progress and everything went back to Production
[14-Jan-2010 10:33:06] <ShaneKu> Does anyone know if the "Server Connection Error" bug in the Event Viewer of 2.5.1 on the vmware appliance is still open? Not having much luck finding it in the bug tracker...
[14-Jan-2010 10:33:10] <rmatte> sweet, didn't know it'd do that
[14-Jan-2010 10:33:19] <rmatte> well, that's effectively the quick kill button
[14-Jan-2010 10:33:37] <rocket> rmatte: just do a doc enhancement about it :p
[14-Jan-2010 10:33:44] <tehhobbit> was contemplating to run local plugins that wrote to a db instead, and then a collector to fetch the data
[14-Jan-2010 10:33:50] <mrayzenoss> ShaneKu: there are a bunch of Event Console tickets fixed http://dev.zenoss.org/trac/report/6
[14-Jan-2010 10:33:58] <mrayzenoss> for 2.5.2 when it comes out
[14-Jan-2010 10:34:08] <rmatte> rocket: yup I shall
[14-Jan-2010 10:34:10] <rocket> rmatte: deleting a in progress maintenance window reverts state .. :p
[14-Jan-2010 10:34:20] <ShaneKu> ..any virtual appliance of 2.5.2 floating around yet?? :)
[14-Jan-2010 10:34:36] <rmatte> rocket: yup
[14-Jan-2010 10:34:37] <mrayzenoss> we may have builds available next week
[14-Jan-2010 10:34:50] <tehhobbit> http://dev.zenoss.org/trac/ticket/6031 would love for this to be fixed
[14-Jan-2010 10:34:53] <mrayzenoss> ShaneKu: next week's QA Test Day will be around testing 2.5.2 candidates
[14-Jan-2010 10:36:33] <ShaneKu> excellent, having issues with snmpv3 and zenperfsnmp between the appliance and my freebsd boxes on 2.4.5. Would like to try 2.5.x with its newer version of net-snmp
[14-Jan-2010 10:50:04] <rmatte> I still need to open an enhancement request for "MOAR PONIES!"
[14-Jan-2010 10:56:09] <daveborg> I have searched for a good reference for creating MultiGraph reports and cannot locate one. Does anyone have a decent reference for this? I am trying to graph the throughput of the interfaces on a 48 port switch and segment the graphs into 4 sets of 12.
[14-Jan-2010 10:59:24] <chickenandbeans> Is there another dev irc meeting thing happening anytime soon?
[14-Jan-2010 11:00:06] <mrayzenoss> next Thursday
[14-Jan-2010 11:00:59] <chickenandbeans> and is it ok for none zenoss people to ask questions?
[14-Jan-2010 11:01:04] <mrayzenoss> every 2 weeks there's an hour or so dev meeting and we're ramping up doing a QA day all day
[14-Jan-2010 11:01:09] <mrayzenoss> of course
[14-Jan-2010 11:01:31] <chickenandbeans> cool
[14-Jan-2010 11:02:35] <mrayzenoss> it's called a Developer IRC session just because someone who should know more than me is guaranteed to be here
[14-Jan-2010 11:05:37] <chickenandbeans> :) Thurs 24th is the mozilla Calendar
[14-Jan-2010 11:07:25] <rmatte> lol
[14-Jan-2010 11:07:56] <rocket> mrayzenoss: no one knows more than you :p
[14-Jan-2010 11:08:06] <mrayzenoss> bah
[14-Jan-2010 11:08:27] <rmatte> you do always seem to have the "inside scoop" as it were
[14-Jan-2010 11:11:35] <chickenandbeans> :) So... what time is it?
[14-Jan-2010 11:11:47] <rmatte> Beer time!
[14-Jan-2010 11:12:05] * chickenandbeans dances
[14-Jan-2010 11:22:46] <forsberg> good evening :)
[14-Jan-2010 11:52:48] <rmatte> eugh, it's annoying how when items are added to ZenPacks it adds whatever links exist to devices as well
[14-Jan-2010 12:11:54] <forsberg> i have a desperate zenoss problem, any zenmasters able to help in here? :)
[14-Jan-2010 12:14:19] <rmatte> what's the issue?
[14-Jan-2010 12:14:22] <tehhobbit> forsberg: give it a try
[14-Jan-2010 12:17:17] <forsberg> in short, since i upgraded to 2.5.1 the MySqlMonitor have been broken, and I am not able to install or remove it. been searching forums etc, and last place i ended was running zendmd
[14-Jan-2010 12:17:55] <forsberg> when i try to run zendmd as zenoss user, it fails with some tracebacks and bring me back to prompt
[14-Jan-2010 12:18:25] <forsberg> im becomming rather desperate since I have so many devices that the task of moving them all to a fresh zenoss is....giving me a headache
[14-Jan-2010 12:19:38] <rmatte> ok, I'll need to see some errors
[14-Jan-2010 12:19:43] <rmatte> pastebin.com them please
[14-Jan-2010 12:19:57] <forsberg> from zendmd ?
[14-Jan-2010 12:20:20] <forsberg> http://p.tyk.nu/386165
[14-Jan-2010 12:20:23] <forsberg> this is from zendmd
[14-Jan-2010 12:21:02] <forsberg> tbh i have no idea if its related, but i guess i should be able to run zendmd anyways
[14-Jan-2010 12:21:26] <rmatte> well, it complaining about a lock file
[14-Jan-2010 12:21:53] <rmatte> you're not trying to run zendmd while zenoss isn't running are you?
[14-Jan-2010 12:22:03] <forsberg> zenoss is running
[14-Jan-2010 12:22:55] <forsberg> all daemons are running, zenoss is working, i have not registrered any other problems so far
[14-Jan-2010 12:23:06] <rmatte> go in to /opt/zenoss/var
[14-Jan-2010 12:23:10] <rmatte> stop zenoss
[14-Jan-2010 12:23:15] <rmatte> delete any .zec files you see
[14-Jan-2010 12:23:20] <rmatte> and any .lock files
[14-Jan-2010 12:23:23] <rmatte> then start zenoss again
[14-Jan-2010 12:23:26] <rmatte> and try zendmd again
[14-Jan-2010 12:26:38] <forsberg> excactly same thing, with same output
[14-Jan-2010 12:27:40] <rmatte> go to $ZENHOME/Products
[14-Jan-2010 12:27:52] <rmatte> do you see Hotfix_20070320 listed there?
[14-Jan-2010 12:27:58] <rmatte> and ZGadflyDA
[14-Jan-2010 12:28:04] <rmatte> and ZenTestRunner
[14-Jan-2010 12:28:21] <rmatte> and ZopeTutorial
[14-Jan-2010 12:28:28] <rmatte> ?
[14-Jan-2010 12:28:45] <forsberg> http://p.tyk.nu/386210
[14-Jan-2010 12:28:49] <forsberg> some of them
[14-Jan-2010 12:29:49] <forsberg> the ones who are there, seems to be from the day i upgraded, when things apparently went wrong :(
[14-Jan-2010 12:30:05] <rmatte> hmmm, the others seem to be in $ZENHOME/lib/python/Products
[14-Jan-2010 12:30:50] <forsberg> indeed
[14-Jan-2010 12:31:51] <rmatte> This seems to be similar to your issue: http://community.zenoss.org/message/41793;jsessionid=E4DA4AD5D13E330AA154927A584B5AD2.node0
[14-Jan-2010 12:32:46] <rmatte> http://dev.zenoss.org/trac/ticket/5798
[14-Jan-2010 12:33:00] <forsberg> ahh ok, so its related to the old wmi when upgrading maybe
[14-Jan-2010 12:33:01] <rmatte> that's the bug ticket that's open for that issue
[14-Jan-2010 12:33:02] <forsberg> god damn
[14-Jan-2010 12:33:13] <rmatte> well, if you had the old WMI, yes
[14-Jan-2010 12:33:16] <rmatte> did you?
[14-Jan-2010 12:34:28] <forsberg> i must admit im not sure, but i may very well have, just didnt think it was related, since its only the mysqlmonitor wich is not working
[14-Jan-2010 12:34:44] <forsberg> (and zendmd)
[14-Jan-2010 12:36:28] <ShaneKu> I'm trying to write a suppression tool using zensendevent based on one of the tips of the month, however I want a way to control the length. Anyone done this before? I'm pretty close with it but could use some very basic python help if anyone is free
[14-Jan-2010 12:37:15] <rmatte> forsberg: the fact that zendmd isn't loading is more important lol
[14-Jan-2010 12:37:29] <forsberg> hehe yeah
[14-Jan-2010 12:37:43] <rmatte> ShaneKu: control the length of?
[14-Jan-2010 12:38:05] <forsberg> to me it just wasnt as important, as everything is "working" except the mysqlmonitor
[14-Jan-2010 12:38:20] <rmatte> forsberg: right
[14-Jan-2010 12:38:39] <rmatte> forsberg: quickest solution will probably be to migrate to a new install.
[14-Jan-2010 12:38:47] <ShaneKu> the maintenece window, using a transport, it all works, but the manage_editMaintenanceWindow method expects time in days, hours, mins, and I want to be able to feed the time in minutes only, just wondering how to do that conversion with python
[14-Jan-2010 12:38:49] <rmatte> forsberg: you could try using zenmigrate to speed things up
[14-Jan-2010 12:40:52] <rmatte> ShaneKu: here's how to do it in seconds: http://stackoverflow.com/questions/572049/convert-seconds-to-days-minutes-and-hours-in-obj-c
[14-Jan-2010 12:40:57] <rmatte> minutes wouldn't be much different
[14-Jan-2010 12:41:30] <rmatte> oh, that's objective c though
[14-Jan-2010 12:41:31] <rmatte> grrr
[14-Jan-2010 12:41:32] <rmatte> one second
[14-Jan-2010 12:42:31] <ShaneKu> should be pretty easy, I'm just a python newb haha, fine with php, perl, shell...
[14-Jan-2010 12:42:48] <rmatte> well, python is pretty braindead simple for doing calculations
[14-Jan-2010 12:43:00] <rmatte> just figure out the mathematical calculation that you need then do it
[14-Jan-2010 12:43:08] <rmatte> time = 1890832
[14-Jan-2010 12:43:11] <rmatte> time * whatever
[14-Jan-2010 12:43:24] <rmatte> you might try #python, they are pretty good in there
[14-Jan-2010 12:43:35] <ShaneKu> will do, thanks
[14-Jan-2010 12:43:43] <rmatte> np
[14-Jan-2010 12:45:29] <forsberg> rmatte zenmigrate ?
[14-Jan-2010 12:45:40] <rmatte> forsberg: yeh, check the admin guide
[14-Jan-2010 12:46:03] <forsberg> ok will do, 1000 thanks for your time
[14-Jan-2010 12:46:24] <forsberg> ive reached the same conclusion a couple of times, but i REALLY wanted to fix the old install, since my work day is interrupted 117 times
[14-Jan-2010 12:48:19] <rmatte> lol
[14-Jan-2010 12:48:47] <rmatte> eugh, apparently the admin guide doesn't explain zenmigrate, but I'm pretty sure the install guide does
[14-Jan-2010 12:49:41] <forsberg> no i was looking, but you mean it can help move devices to new install or?
[14-Jan-2010 12:49:46] <rmatte> yes
[14-Jan-2010 12:49:52] <rmatte> that's what it's designed for
[14-Jan-2010 12:49:54] <forsberg> uhh that would be über
[14-Jan-2010 12:50:16] <rmatte> you chatting from an iphone?
[14-Jan-2010 12:50:21] <forsberg> nopes
[14-Jan-2010 12:50:34] <rmatte> ah, my iphone automatically adds the dots to the u on uber
[14-Jan-2010 12:50:34] <forsberg> i just used german ü
[14-Jan-2010 12:50:38] <rmatte> hence why I was asking
[14-Jan-2010 12:50:39] <rmatte> ah
[14-Jan-2010 12:50:41] <forsberg> to stress how über
[14-Jan-2010 12:50:41] <forsberg> ;>

[14-Jan-2010 12:50:45] <rmatte> lol
[14-Jan-2010 12:51:39] <rmatte> or sorry
[14-Jan-2010 12:51:43] <rmatte> you want zenbackup and zenrestore
[14-Jan-2010 12:51:45] <rmatte> not zenmigrate
[14-Jan-2010 12:53:25] <rmatte> I haven't used them myself so I can't help you with the actual syntax
[14-Jan-2010 12:53:32] <rmatte> I'm sure I'll end up using them eventually
[14-Jan-2010 12:54:00] <forsberg> thats ok, someone else must have done it, i wasnt aware that you could actually use it to move all devices...
[14-Jan-2010 12:54:06] <rmatte> yup
[14-Jan-2010 12:54:42] <forsberg> never gave it another thought, after i failed to find the restore option in the gui :)
[14-Jan-2010 12:56:13] <forsberg> http://docs.huihoo.com/zenoss/admin-guide/2.4.2/ch14s04.html
[14-Jan-2010 12:56:49] <forsberg> ill put up some iron tomorrow and try it out :) thanks again
[14-Jan-2010 12:56:51] <rmatte> yup, that's the section
[14-Jan-2010 13:00:57] <rmatte> np
[14-Jan-2010 13:23:38] <ShaneKu> I got that working, if anyone is interested here is the transform I put on the Suppress event class mapping: http://pastebin.com/m2ea9de49
[14-Jan-2010 13:24:15] <rmatte> cool, your python honestly isn't that bad lol
[14-Jan-2010 13:24:16] <ShaneKu> simple shell script to start and stop (remove) the maintenence window based on time in minutes - http://pastebin.com/m2bcd7ef
[14-Jan-2010 13:24:39] <ShaneKu> well I borrowed most of that from a tip of the month ;)
[14-Jan-2010 13:24:44] <rmatte> hehe
[14-Jan-2010 13:25:56] <mrayzenoss> feel free to post those to the wiki: http://community.zenoss.org/community/documentation/wiki
[14-Jan-2010 13:26:16] <ShaneKu> will do
[14-Jan-2010 13:29:31] <mrayzenoss> meeting time
[14-Jan-2010 13:47:14] <hermes> Hello - I just ran stack installer for zenoss on my FC 10 box. The out put is " [Errno 13] Permission denied: '/home/zenoss' The Python egg cache directory is currently set to: /home/zenoss/.python-eggs Perhaps your account does not have write access to this directory? You can change the cache directory by setting the PYTHON_EGG_CACHE environment variable to point to an accessible directory."
[14-Jan-2010 13:47:56] <rmatte> did you not run it as root?
[14-Jan-2010 13:48:03] <hermes> Yes i did
[14-Jan-2010 13:48:31] <hermes> changed permission on that folder
[14-Jan-2010 13:48:35] <rmatte> does /home/zenoss/.python_eggs exist?
[14-Jan-2010 13:48:43] <rmatte> erm .python-eggs rather
[14-Jan-2010 13:49:15] <hermes> no, the folder isn' there.
[14-Jan-2010 13:51:44] <hermes> the installer creates the zenoss folder in /usr/local/zenoss - i've ran this a couple of times. even created the folder zenoss under home.
[14-Jan-2010 13:51:54] <rmatte> that directory don't even exist for me either
[14-Jan-2010 13:52:03] <rmatte> so I doubt that the fact that it's not there is an issue
[14-Jan-2010 13:52:21] <rmatte> it should create all of that for you though
[14-Jan-2010 13:53:28] <rmatte> it seems to be complaining that you don't have write access to /home/zenoss as root, which is bizarre
[14-Jan-2010 13:54:35] <hermes> don't know what else to look for.
[14-Jan-2010 13:55:11] <rmatte> no idea, I've never had the stack installer fail on me
[14-Jan-2010 13:55:17] <rmatte> is that a perfectly clean FC 10 install?
[14-Jan-2010 13:55:23] <hermes> Yes
[14-Jan-2010 13:55:24] <rmatte> or do you have a bunch of other stuff running on it?
[14-Jan-2010 13:55:46] <rmatte> Are you guys a fedora shop? Like, is there any particular reason why you have to use FC 10?
[14-Jan-2010 13:55:59] <rmatte> Zenoss is better tested on RHEL, CentOS, Ubuntu Server
[14-Jan-2010 13:56:19] <hermes> yea, pretty much everything is FC
[14-Jan-2010 13:56:22] <rmatte> ah
[14-Jan-2010 13:56:36] <rmatte> is that install some custom build of FC?
[14-Jan-2010 13:56:46] <hermes> not at all.
[14-Jan-2010 13:56:51] <rmatte> k
[14-Jan-2010 13:57:15] <hermes> my other FC box is running opennms along with nmis.
[14-Jan-2010 13:57:37] <rmatte> ah
[14-Jan-2010 13:57:51] <rmatte> so you've tried RPM and Stack and they've both failed?
[14-Jan-2010 13:57:53] <hermes> this one, that i'm working on was supposed to be for Zenoss. Not sure if the boss will like a box on the network that isn't FC.
[14-Jan-2010 13:57:57] <hermes> Yea man. nada
[14-Jan-2010 13:58:05] <forsberg> can you make the dir yourself? :)
[14-Jan-2010 13:58:06] <rmatte> that's pretty weird
[14-Jan-2010 13:58:26] <hermes> tried that also.
[14-Jan-2010 13:58:48] <rmatte> What was the failure with RPM again?
[14-Jan-2010 13:59:39] <hermes> with the rpm, it just hangs there. trying to build a database
[14-Jan-2010 13:59:51] <rmatte> Hmmm
[14-Jan-2010 13:59:56] <rmatte> found this post by Matt: http://www.mail-archive.com/zenoss-users@zenoss.org/msg24641.html
[14-Jan-2010 14:00:12] <rmatte> apparently you're not the only one with stack installer issues on Fedora, though that's from like a whole year ago
[14-Jan-2010 14:00:33] <rmatte> Someone stated that FC 8 would work better than FC 10
[14-Jan-2010 14:00:56] <rmatte> oh, here's another comment
[14-Jan-2010 14:00:57] <rmatte> "for whatever reason when I tried the stack install from runlevel 3 the install
[14-Jan-2010 14:00:57] <rmatte> worked.
[14-Jan-2010 14:01:30] <hermes> let me give that a try
[14-Jan-2010 14:01:35] <rmatte> k
[14-Jan-2010 14:17:50] <hermes> no dice.
[14-Jan-2010 14:33:47] <rmatte> then I'd suggest you try FC 8
[14-Jan-2010 14:34:15] <rmatte> mrayzenoss: hermes is having issues installing Zenoss in Fedora Core 10, he's tried the RPM and the stack installer
[14-Jan-2010 14:34:29] <rmatte> any tips/workarounds that you're aware of?
[14-Jan-2010 14:36:00] <theacolyte-> awesome, my trac ticket was closed because it was a duplicate of a ticket submitted... 13 months ago with a last change of 3 months
[14-Jan-2010 14:37:01] <rmatte> theacolyte-: yup, that happens a lot
[14-Jan-2010 14:37:14] <theacolyte-> I guess I'll just make them line graphs
[14-Jan-2010 14:37:16] <rmatte> theacolyte-: the zenoss backlog is huge
[14-Jan-2010 14:37:19] <theacolyte-> Although it looks like I probably got the job
[14-Jan-2010 14:37:23] <theacolyte-> Interview went really well
[14-Jan-2010 14:37:26] <rmatte> cool
[14-Jan-2010 14:37:31] <theacolyte-> It may be a non-issue
[14-Jan-2010 14:37:38] <theacolyte-> Though I did mention I liked zenoss
[14-Jan-2010 14:37:39] <theacolyte-> hehe
[14-Jan-2010 14:37:44] <rmatte> hehe
[14-Jan-2010 14:37:49] <hermes> thanks rmatte.
[14-Jan-2010 14:37:59] <rmatte> line graphs aren't that bad, they just aren't as flashy
[14-Jan-2010 14:38:06] <theacolyte-> yeah
[14-Jan-2010 14:38:12] <rmatte> hermes: np, hopefully Matt wakes up and sees that :)
[14-Jan-2010 14:38:21] <rmatte> yikes
[14-Jan-2010 14:38:22] <hermes> i will keep on playing around with zenoss. looks great. don't want to give up yet
[14-Jan-2010 14:38:36] <rmatte> hermes: yeh, try FC 8
[14-Jan-2010 14:38:49] <hermes> k
[14-Jan-2010 14:42:06] <rmatte> mrayzenoss: hermes is having issues installing Zenoss in Fedora Core 10, he's tried the RPM and the stack installer
[14-Jan-2010 14:42:13] <rmatte> any tips/workarounds that you're aware of?
[14-Jan-2010 14:42:27] <mrayzenoss> pretty sure RPM won't work
[14-Jan-2010 14:42:34] <rmatte> well that helps :)
[14-Jan-2010 14:42:40] <rmatte> stack installer doesn't work either
[14-Jan-2010 14:43:04] <rmatte> he gets...
[14-Jan-2010 14:43:28] <rmatte> /home/zenoss/.python-eggs Perhaps your account does not have write access to this directory? You can change the cache directory by setting the PYTHON_EGG_CACHE environment variable to point to an accessible directory.
[14-Jan-2010 14:43:58] <mrayzenoss> is he installing zenpacks as root?
[14-Jan-2010 14:44:18] <rmatte> no, he's just running the stack installer as root on a fresh FC 10 install
[14-Jan-2010 14:44:46] <rmatte> seems a bit odd
[14-Jan-2010 14:45:36] <rmatte> I wonder if the failed RPM install left fragments behind which the stack installer doesn't like
[14-Jan-2010 14:46:07] <mrayzenoss> that sounds feasible
[14-Jan-2010 14:46:16] <mrayzenoss> especially if it wasn't uninstalled
[14-Jan-2010 14:46:33] <rmatte> hermes: is it possible that stuff got left behind by the RPM install?
[14-Jan-2010 14:47:04] <hermes> i uninstalled it.
[14-Jan-2010 14:47:14] <rmatte> ah, k
[14-Jan-2010 14:47:30] <rmatte> wonder if it completely uninstalled cleanly though
[14-Jan-2010 14:47:51] <hermes> i think it did. but not 100%
[14-Jan-2010 14:48:02] <mrayzenoss> ke4qqq: do you run Zenoss on Fedora?
[14-Jan-2010 14:48:10] <rmatte> you could always try another clean FC 10 install, then run the stack installer
[14-Jan-2010 14:48:40] * rmatte fwaps ke4qqq
[14-Jan-2010 14:48:42] <rmatte> wake up
[14-Jan-2010 14:51:19] <hermes> i could try that again. it's a test box
[14-Jan-2010 15:04:25] <hermes> rmatte - i will get on this again tomorrow and let you know how it turns out. thanks again man!
[14-Jan-2010 15:04:44] <hermes> i will also install it on my ubuntu laptop.
[14-Jan-2010 15:18:49] <ke4qqq> mrayzenoss: no, RHEL currently, CentOS prior to that
[14-Jan-2010 15:19:14] * ke4qqq reads up
[14-Jan-2010 15:19:53] <ke4qqq> whats sestatus return?
[14-Jan-2010 15:20:06] <ke4qqq> hermes: ^^
[14-Jan-2010 15:38:53] <hermes> SELinux status: disabled
[14-Jan-2010 15:48:11] ideopathic_ is now known as ideopathic
[14-Jan-2010 15:50:55] <hermes> rmatte, mrayzenoss and ke4qqq - thanks for the help/advise. i will jump on this ch. tomorrow and give you guys an update.
[14-Jan-2010 15:51:01] <hermes> nite
[14-Jan-2010 16:22:45] <rmatte> woohoo, got a ton of stuff tuned today
[14-Jan-2010 16:26:45] <rmatte> looks like tomorrow is a development day
[14-Jan-2010 17:02:41] <rmatte> Lol, found the oldest pic I've ever seen of Zenoss: http://image.moongift.jp/intro3/Zenoss_BDC1/zenoss043.png
[14-Jan-2010 17:03:16] <rmatte> and found a screenshot of the new UI: http://designingwebinterfaces.com/blog/wp-content/uploads/2009/07/drag_drop_reorg_zenoss.png
[14-Jan-2010 17:03:20] <rmatte> looks exceptionally awesome
[14-Jan-2010 17:03:36] <nonsenso> uhm. that looks awesome.
[14-Jan-2010 17:03:39] <nonsenso> can i have it now?
[14-Jan-2010 17:03:42] <rmatte> hehe
[14-Jan-2010 17:03:50] <rmatte> nope, wait 6 or 7 months
[14-Jan-2010 17:04:05] <nonsenso> dang.
[14-Jan-2010 17:04:12] <rmatte> actually, technically 4 months, but not sure if they're going to hit that
[14-Jan-2010 17:04:35] <zenethian> Heh
[14-Jan-2010 17:05:31] <rmatte> zenethian: it'd be cool if you guys released some screenshots of what's done so far with the new UI, I'm quite curious to see how it's coming along
[14-Jan-2010 17:05:31] * zenethian is on the new UI team.
[14-Jan-2010 17:05:47] <nonsenso> yay!
[14-Jan-2010 17:06:00] <zenethian> I'll ask if I can do that. Maybe give an exclusive IRC sneak peek.
[14-Jan-2010 17:06:06] <rmatte> That's the only screenshot that I've managed to find of it
[14-Jan-2010 17:06:11] <nonsenso> zenethian: are you a community developer or employee?
[14-Jan-2010 17:06:11] <rmatte> and it looks even better than I expected
[14-Jan-2010 17:06:18] <zenethian> Employee
[14-Jan-2010 17:06:36] <zenethian> screenshot inbound, give me a minute.
[14-Jan-2010 17:07:18] <nonsenso> rmatte: do you think it would be in bad taste to share the sneak peak preview you just posted around?
[14-Jan-2010 17:07:28] <nonsenso> presumably that's on someone's blog..
[14-Jan-2010 17:07:35] <rmatte> probably
[14-Jan-2010 17:08:42] <nonsenso> hrm. it's public.. http://designingwebinterfaces.com/2009/07
[14-Jan-2010 17:10:03] <zenethian> http://zen.sebistar.net/images/zenoss_newui_services.png
[14-Jan-2010 17:10:06] <zenethian> Enjoy
[14-Jan-2010 17:10:10] ideopathic_ is now known as ideopathic
[14-Jan-2010 17:10:18] <nonsenso> hotdamn. :)
[14-Jan-2010 17:10:55] <zenethian> I warn: that is in development and subject to change
[14-Jan-2010 17:11:57] <zenethian> It's in trunk though. For everyone who loves source builds.
[14-Jan-2010 17:12:12] <nonsenso> some of those design elements are in 2.5.1 i noticed. not sure when they were introduced.
[14-Jan-2010 17:13:14] <zenethian> Yeah, we started doing some work in Ext (the UI toolkit we're using now) to get an idea of how it will work.
[14-Jan-2010 17:14:09] * nonsenso is install remedy. ugh.
[14-Jan-2010 17:14:23] <nonsenso> the installer is so. freaking. horrible.
[14-Jan-2010 17:15:12] <zenethian> Any particular part of the install that you don't like?
[14-Jan-2010 17:15:24] <nonsenso> the part where it's java. and slow. and requires X.
[14-Jan-2010 17:15:28] <nonsenso> :)
[14-Jan-2010 17:15:35] <zenethian> Oh, THAT installer.
[14-Jan-2010 17:15:39] <zenethian> I thought you meant the install wizard.
[14-Jan-2010 17:15:56] <zenethian> I guess that's a "setup" wizard.
[14-Jan-2010 17:16:03] <nonsenso> and this is like my 3rd install because the first two had severe issues.
[14-Jan-2010 17:16:19] <nonsenso> presumptions about how apache is installed because they want you to use THEIR mod_jk, etc.
[14-Jan-2010 17:16:43] <nonsenso> like, "give us the install path for apache (we're presuming you installed it by source!"
[14-Jan-2010 17:17:21] <nonsenso> yeah, i used the rpm because there's well over 400+ boxes here. i'm not doing local src installs..
[14-Jan-2010 17:18:03] <nonsenso> </rant>

[14-Jan-2010 17:18:16] <zenethian> Hrm, which installer? I didn't think it was java, I thought the Linux installer was GTK based and the GUI could be disabled with a flag.
[14-Jan-2010 17:18:17] <zenethian> Uh we're talking about Zenoss still right? heh
[14-Jan-2010 17:18:26] <nonsenso> hehe. nono, remedy.
[14-Jan-2010 17:18:31] <zenethian> OH
[14-Jan-2010 17:18:33] <nonsenso> big scary ticketing system from BMC.
[14-Jan-2010 17:18:38] <zenethian> ha
[14-Jan-2010 17:18:40] <zenethian> My mistake.
[14-Jan-2010 17:18:42] <nonsenso> (:
[14-Jan-2010 17:19:17] <nonsenso> afk meeting time! nice meeting you sir!
[14-Jan-2010 17:19:20] <zenethian> I was going to say that we're doing something mighty wrong if our installer needs java and X.
[14-Jan-2010 17:19:23] <zenethian> have fun!
[14-Jan-2010 17:25:32] <mrayzenoss1> later all
[14-Jan-2010 17:25:39] <zenethian> seeya
[14-Jan-2010 18:03:40] <sjobeck> hey, hi, all, hope all are well.
[14-Jan-2010 18:03:40] <sjobeck> can anyone point me at the file where the DB (mySQL) connection string is in? (silly but i can not grep it from any where in the entire /opt/zenoss/ any where)
[14-Jan-2010 18:03:40] <sjobeck> thx !
[14-Jan-2010 19:10:31] <theacolyte-> No luck
[14-Jan-2010 19:10:34] <theacolyte-> You guys are stuck with me
[14-Jan-2010 19:14:30] <aclark> heh
[14-Jan-2010 21:00:01] ideopathic_ is now known as ideopathic
[14-Jan-2010 21:18:39] ideopathic_ is now known as ideopathic
[15-Jan-2010 00:00:30] [disconnected at Fri Jan 15 00:00:30 2010]
[15-Jan-2010 00:00:30] [connected at Fri Jan 15 00:00:30 2010]
[15-Jan-2010 02:00:40] <dalfz> hi, I've been running 2.4.5, and suddenly data in the performance graph is lost. i.e my temperature sensor "reset itself" a few days ago, so previous data is not shown in graph. any hints?
[15-Jan-2010 02:02:23] <dalfz> oh, and I just did a debian zenoss-stack upgrade to 2.5.1 to see if the historic data would be available again, but it didn't help. (upgrade went just fine)
[15-Jan-2010 03:34:47] jabular_ is now known as jabular
[15-Jan-2010 04:28:12] <tehhobbit> anyone that can point me in the direction of some docs to modify the ethernet collector to handle fc ports ?
[15-Jan-2010 04:57:24] <tehhobbit> anyone around that can explain a few things about the os->interface part to me
[15-Jan-2010 04:58:31] <tehhobbit> I want to change so that interfaces that comes back as fibre channel uses another template than the ones that comes back as ethernet*
[15-Jan-2010 05:46:48] <rayk3n> hi, how can i remove /perf/cpu eventclass from monitoring
[15-Jan-2010 06:02:49] <tehhobbit> rayk3n: this isnt really an answer to your question, but I removed the thresholds from the template instead
[15-Jan-2010 07:25:04] <twm1010> morning people.
[15-Jan-2010 07:26:01] nassy_ is now known as nassy
[15-Jan-2010 08:18:54] <Absenth> what would I have to change to prevent zenoss from quering what IP addresses are assigned to a server?
[15-Jan-2010 08:19:47] <Absenth> I run a FreeBSD Jails server. The Server has to have an IP address that's ultimately assigned to each jail. If I want to monitor the jails individually, I currently can't monitor the Jails server as Zenoss says the IP of the Jail is already in use by the host.
[15-Jan-2010 08:20:19] <twm1010> You probably want to remove the interface mapping plugins
[15-Jan-2010 08:20:22] <twm1010> of the device class
[15-Jan-2010 09:18:08] <mrayzenoss> Hmmmâ€¦ anyone want to upload their screenshots here? http://community.zenoss.org/docs/DOC-2616
[15-Jan-2010 09:30:16] <twm1010> hrmmm....
[15-Jan-2010 09:30:18] <twm1010> i can do that
[15-Jan-2010 09:45:55] <twm1010> how about a sweet looking multi-graph report?
[15-Jan-2010 09:52:40] <mrayzenoss> sounds good :)
[15-Jan-2010 09:56:25] <twm1010> @mrayzenoss I don't see the edit dialog?
[15-Jan-2010 09:59:15] <mrayzenoss> hmmâ€¦ I'll see if I can fix that
[15-Jan-2010 09:59:46] <mrayzenoss> it says anyone may edit
[15-Jan-2010 09:59:54] <mrayzenoss> are you logged in?
[15-Jan-2010 10:00:15] <mrayzenoss> off to a demo
[15-Jan-2010 10:04:39] <rmatte> good morning all
[15-Jan-2010 10:05:53] <forsberg> good morning :)
[15-Jan-2010 10:08:47] <forsberg> rmatte > small update: i did a backup/restore to a fresh install and it seems to have fixed some of the problems, there is still one left tho, are you here a little later?
[15-Jan-2010 10:09:02] <rmatte> I'll be around for a while, yeh
[15-Jan-2010 10:09:25] <rmatte> I'll be doing coding today though so I won't be in here consitently
[15-Jan-2010 10:09:29] <rmatte> consistently*
[15-Jan-2010 10:09:39] <forsberg> schweet, ill try to catch you later then maybe :P after dinner and baby is asleep
[15-Jan-2010 10:10:13] <forsberg> all the wierd product errors are gone, it still complains about the lock file tho.
[15-Jan-2010 10:10:45] <forsberg> if i stop zenoss and delete lock file and start, same thing happens, but if i delete the lock file while zenoss is running, and do zendmd twice, i enter zendmd with no errors
[15-Jan-2010 10:11:05] <rmatte> hmmm]
[15-Jan-2010 10:11:13] <rmatte> that's a really weird bug
[15-Jan-2010 10:11:32] <forsberg> yes, but i might have did something wrong with some zenpacks when restoring etc, so im gonna make those things right first
[15-Jan-2010 10:11:56] <forsberg> make sure i have same packs installed etc
[15-Jan-2010 10:12:42] <rmatte> well, I don't see how you can do something "wrong" with a ZenPack unless you're actually editing it
[15-Jan-2010 10:12:59] <rmatte> I don't see a ZenPack causing a lock file issue though
[15-Jan-2010 10:13:02] <rmatte> it must be something else
[15-Jan-2010 10:13:27] <forsberg> thats true, just want to make sure i have the same zenpacks installed when i backup and when i restore
[15-Jan-2010 10:14:10] <forsberg> and its rather strange that i can delete the lock file when its running and then first zendmd spams another error, and if i do it again i enter zendmd with no errors
[15-Jan-2010 10:14:17] <forsberg> and it keeps working untill i restart zenoss
[15-Jan-2010 10:14:36] <rmatte> right, that's the bug
[15-Jan-2010 10:14:47] <rmatte> I pasted a link to the bug ticket that's open for that yesterday
[15-Jan-2010 10:15:41] <forsberg> http://dev.zenoss.org/trac/ticket/5798
[15-Jan-2010 10:16:19] <rmatte> yup
[15-Jan-2010 10:16:37] <forsberg> but atleast this way i can get the mysqlmonitor to work, and my boss gets happy happy ;)
[15-Jan-2010 10:16:54] <rmatte> mrayzenoss: any idea about a fix for http://dev.zenoss.org/trac/ticket/5798 ?
[15-Jan-2010 10:17:06] <forsberg> [17:00:15] <@mrayzenoss> off to a demo
[15-Jan-2010 10:17:06] <rmatte> maybe you're aware of a guide or something on the forums that I'm not
[15-Jan-2010 10:17:10] <rmatte> ah
[15-Jan-2010 10:17:11] <forsberg> thats 12 minutes ago
[15-Jan-2010 10:17:23] <forsberg> 17 sorry
[15-Jan-2010 10:17:27] <rmatte> well, mrayzenoss1 joined
[15-Jan-2010 10:17:36] <rmatte> so that's probably him on his laptop at the demo
[15-Jan-2010 10:17:44] <rmatte> ;)
[15-Jan-2010 10:17:59] <forsberg> ahh affirmative
[15-Jan-2010 10:20:30] mrayzenoss1 is now known as mrayzenoss
[15-Jan-2010 10:21:56] <forsberg> the ticket has alot about daemons dying tho, i have none of that
[15-Jan-2010 10:22:02] <forsberg> but seems very similar :)
[15-Jan-2010 10:23:26] <rmatte> so you're having the problem on a completely clean install?
[15-Jan-2010 10:25:07] <rmatte> wow, someone is on crack
[15-Jan-2010 10:25:21] <rmatte> http://dev.zenoss.org/trac/ticket/5547 was closed as a duplicate of http://dev.zenoss.org/trac/ticket/2071
[15-Jan-2010 10:25:32] <rmatte> but 2071 has absolutely nothing to do with 5547
[15-Jan-2010 10:26:22] <forsberg> no first get the problem when i restore the data from the "sick" install
[15-Jan-2010 10:26:41] <forsberg> i have plenty other of install with no problems :P
[15-Jan-2010 10:27:08] <forsberg> installs
[15-Jan-2010 10:28:53] <forsberg> im using centos like the guy in the bug ticket
[15-Jan-2010 10:31:30] <mrayzenoss> rmatte: that's 2071 internal, should have noted that
[15-Jan-2010 10:31:31] <rmatte> hmmm, he might be talking about 2071 in their internal ticketing system.
[15-Jan-2010 10:31:39] <rmatte> yeh, I just just realized that
[15-Jan-2010 10:31:58] <rmatte> He did note that, I just didn't make the connection until now
[15-Jan-2010 10:32:00] <mrayzenoss> 2071 Create release note about officially abandoning zip zenpacks
[15-Jan-2010 10:32:10] <mrayzenoss> I'll get that moved external where it should be
[15-Jan-2010 10:32:12] <rmatte> ahhhh
[15-Jan-2010 10:32:40] <rmatte> thanks
[15-Jan-2010 10:33:32] <rmatte> forsberg: oh, I see
[15-Jan-2010 10:33:49] <rmatte> forsberg: you might be stuck moving the devices by hand to a new install then, if the problem seems to tag along in the import
[15-Jan-2010 10:34:00] <rmatte> actually, there's a way to only import the devices themselves
[15-Jan-2010 10:34:05] <rmatte> you may want to dig that up and try that
[15-Jan-2010 10:34:14] <rmatte> then the issue might not tag along to the new install
[15-Jan-2010 10:49:21] <forsberg> yeah but then i guess i need to do all the custom device classes and everything also
[15-Jan-2010 10:50:50] <rmatte> yessir
[15-Jan-2010 10:51:31] <rmatte> that's odd that the bug follows the export though
[15-Jan-2010 10:51:36] <rmatte> I wonder what could be causing that
[15-Jan-2010 10:54:00] <rmatte> I'll add that to the trac ticket (that it followed the export)
[15-Jan-2010 10:54:07] <rmatte> you just did a zenbackup and zenrestore?
[15-Jan-2010 10:57:02] iouns is now known as Iouns
[15-Jan-2010 11:06:18] <rmatte> forsberg: how did you move the stuff over? zenbackup/restore?
[15-Jan-2010 11:08:40] <forsberg> yes
[15-Jan-2010 11:09:01] <forsberg> zenbackup --save-mysql-access --file=/home/zenoss/BACKUPGULD
[15-Jan-2010 11:09:29] <forsberg> scp file over and restore with zenrestore --file=BACKUPGULD
[15-Jan-2010 11:19:56] <forsberg> i stop zenoss when i restore it, should i then delete zec and lock files in /opt/zenoss/var ?
[15-Jan-2010 11:20:03] <forsberg> before i start it again
[15-Jan-2010 11:22:05] <cgibbons> hmmm
[15-Jan-2010 11:26:21] <rmatte> well, the zec file shouldn't be an issue unless it's corrupt
[15-Jan-2010 11:26:26] <rmatte> and the .lock should be removed on it's own
[15-Jan-2010 11:29:06] <rmatte> I just noted that in the trac ticket
[15-Jan-2010 11:29:12] <rmatte> hopefully that'll help narrow it down
[15-Jan-2010 11:29:35] <forsberg> super, thanks
[15-Jan-2010 11:29:47] <forsberg> will try again in an hour or so
[15-Jan-2010 11:30:03] <rmatte> cool
[15-Jan-2010 11:30:10] <forsberg> delete all unused/wierd zenpakcs from the original install
[15-Jan-2010 11:30:12] <rmatte> there's a way to just zenbackup the devices only
[15-Jan-2010 11:30:20] <rmatte> I'd say try that, that'll hopefully do it
[15-Jan-2010 11:30:34] <forsberg> make sure the remaining zenpacks are also on the fresh install and try again
[15-Jan-2010 11:30:34] <forsberg> yes
[15-Jan-2010 11:30:43] <forsberg> that i will try last if, but i need mmuch more time for that
[15-Jan-2010 11:30:54] <forsberg> since its our main survailance
[15-Jan-2010 11:31:02] <forsberg> i cannot leave it in ruins ;>

[15-Jan-2010 11:31:03] <rmatte> yeh
[15-Jan-2010 11:31:12] <rmatte> hehe
[15-Jan-2010 11:32:06] <forsberg> thing is i need to change ip's on the iron also, since much of the devices is only open for the original ip, well im sure you know all thats involved, to properly test the new install
[15-Jan-2010 11:32:22] <rmatte> yup
[15-Jan-2010 11:32:30] <rmatte> I go through it daily lol
[15-Jan-2010 11:32:34] <forsberg> hehe ;:)
[15-Jan-2010 11:33:23] <forsberg> i have the hmm 24/7 this week(end) anyways so its a good time to try!
[15-Jan-2010 11:39:35] <rmatte> yup
[15-Jan-2010 11:55:17] nassy_ is now known as nassy
[15-Jan-2010 12:29:47] <rmatte> bah
[15-Jan-2010 12:29:59] <rmatte> mrayzenoss: http://www.zenoss.com/Members/hackman238/Errors-Discards-Duplex.zip
[15-Jan-2010 12:30:06] <rmatte> any idea where member's files were moved to?
[15-Jan-2010 12:30:10] <rmatte> I really need that file
[15-Jan-2010 12:56:19] <mrayzenoss> what is/was it?
[15-Jan-2010 12:57:03] <mrayzenoss> hackman238 is still very active, you could PM him
[15-Jan-2010 12:59:02] <forsberg> ack :/ totally no go, and i was wrong, the mysqlmonitor is still broken on the "fresh" install
[15-Jan-2010 12:59:13] <forsberg> fuk :[
[15-Jan-2010 13:06:05] <rmatte> I did PM him
[15-Jan-2010 13:06:15] <rmatte> but until he gets back to me my whole development day is blown without that
[15-Jan-2010 13:06:33] <rmatte> it's an error report that he wrote, I'll modify it a bit, but I don't want to duplicate work that's already done
[15-Jan-2010 13:10:57] <forsberg> when http://dev.zenoss.org/trac/ticket/5798 status says closed
[15-Jan-2010 13:11:07] <forsberg> does it mean its not looked at anymore?
[15-Jan-2010 13:11:58] <mrayzenoss> correct. If this work-around isn't working, please re-open the case and add a note about what you're seeing
[15-Jan-2010 13:12:48] <forsberg> rmatte updated it 100minutes ago
[15-Jan-2010 13:13:04] <forsberg> but i guess it doenst make much difference if its closed
[15-Jan-2010 13:13:06] <mrayzenoss> yeah, but it's still marked as closed
[15-Jan-2010 13:15:19] <gwb235> garrrr! what is the best thread for configuring WMI in Zenoss for systems with windows firewalls?
[15-Jan-2010 13:16:56] <mrayzenoss> gwb235: http://community.zenoss.org/docs/DOC-2520 might help
[15-Jan-2010 13:17:28] <cgibbons> and http://msdn.microsoft.com/en-us/library/aa389286(VS.85).aspx
[15-Jan-2010 13:28:14] <mrayzenoss> more meetings
[15-Jan-2010 13:47:32] <rmatte> mrayzenoss: the problem with the workaround for http://dev.zenoss.org/trac/ticket/5798 is that if someone accidentally upgrades without removing the zec files, then has the issue, there's no way for them to salvage their config. Even if they zenbackup and zenrestore to a clean install the issue follows them.
[15-Jan-2010 13:47:58] <rmatte> So something that's being backed up by zenbackup is causing the issue
[15-Jan-2010 13:48:27] <forsberg> and i discovered the issue like 2 months to late :/
[15-Jan-2010 13:48:35] <rmatte> Also, the deleting of zec files should arguably be done automatically during upgrade rather than having it listed as an upgrade step
[15-Jan-2010 13:49:19] <rmatte> If my Zenoss installs gots trashed because I didn't remove .zec files by hand before upgrading I wouldn't be overly impressed.
[15-Jan-2010 13:52:15] <mrayzenoss> reopen the ticket and mention that the zec files should be cleaned up by the installer, I assume the RPM install can do that
[15-Jan-2010 13:53:57] <rmatte> k
[15-Jan-2010 13:57:39] <rmatte> done
[15-Jan-2010 13:58:34] <forsberg> o//
[15-Jan-2010 14:17:39] <gwb235> using tcpdump, it looks like zenoss wmi needs these ports: 135, 139, 445, and a dynamically-assigned port > 1024. Does that jive with what others know?
[15-Jan-2010 14:18:03] <rmatte> you really need to leave all ports open to do effective WMI monitoring
[15-Jan-2010 14:18:21] <rmatte> since it's not always using the same ports
[15-Jan-2010 14:18:25] <gwb235> yeah... alas, in a .edu environment not feasible :(
[15-Jan-2010 14:18:33] <rmatte> then don't use WMI
[15-Jan-2010 14:18:35] <rmatte> use SNMP
[15-Jan-2010 14:20:01] <gwb235> I am quickly getting to that point myself, but deploying SNMP on windows boxes would face a extremely uphill and difficult challenge here.
[15-Jan-2010 14:20:10] <rmatte> why's that?
[15-Jan-2010 14:20:49] <gwb235> the windows box admins temperment is heavily weighted towards "don't install additional software"
[15-Jan-2010 14:21:06] <rmatte> the snmp service is not really "additional software"
[15-Jan-2010 14:21:11] <gwb235> even though yes technically having to enable GPO/firewall holes is just as much work
[15-Jan-2010 14:21:13] <rmatte> it's the same as installing the WMI service
[15-Jan-2010 14:21:16] <forsberg> its a windows option
[15-Jan-2010 14:22:24] <rmatte> WMI is unreliable by the way, you'd constantly get servers that just magically stop responding to WMI and require a restart of the service, or in some cases a reboot
[15-Jan-2010 14:22:48] <rmatte> One server that we were monitoring via WMI just suddenly stopped responding one day and we've tried absolutely everything to fix it and can't get it working again
[15-Jan-2010 14:22:55] <rmatte> so we just switched it to SNMP monitoring
[15-Jan-2010 14:23:42] <rmatte> we've had it happen quite a few times on other servers as well
[15-Jan-2010 14:23:45] <cparlette> snmp informant is additional software
[15-Jan-2010 14:23:56] <rmatte> cparlette: you don't need to use informant
[15-Jan-2010 14:24:50] <cparlette> for cpu/memory?
[15-Jan-2010 14:25:21] <rmatte> http://community.zenoss.org/docs/DOC-3386
[15-Jan-2010 14:25:24] <rmatte> http://community.zenoss.org/docs/DOC-3570
[15-Jan-2010 14:25:44] <rmatte> and I'm going to be working on writing an actual collector plugin for those packs soon enough too, so they'll be even more effective once that's done
[15-Jan-2010 14:26:10] <rmatte> I've already improved the graphs so that they look much better too, but I'm not releasing a new version until the collector plugin is done
[15-Jan-2010 14:26:23] <rmatte> the version that's up there right now is good enough to get by for now though
[15-Jan-2010 14:26:49] <gwb235> looking now... does the snmp query dump windows service(s) status ?
[15-Jan-2010 14:27:26] <rmatte> gwb235: no, only WMI does that, but what you can do is find the .exe names of all the processes for those services, then click on Processes in the left menu and add them in there
[15-Jan-2010 14:27:42] <rmatte> make sure that you change Ignore parameters to True for each one while adding them
[15-Jan-2010 14:27:49] <forsberg> oi nice, <3 to get rid of that wmi
[15-Jan-2010 14:27:50] <rmatte> then remodel the devices and it'll pick up the processes automatically
[15-Jan-2010 14:28:08] <rmatte> I do full SNMP monitoring across quite a few servers
[15-Jan-2010 14:28:21] <rmatte> one of my Zenoss boxes is monitoring 80 Windows servers with full SNMP
[15-Jan-2010 14:29:21] <rmatte> Not sure how many you guys have to monitor?
[15-Jan-2010 14:29:56] <gwb235> not too many, just trying to get it to work (and need to have windows service monitoring -- higher requirement that CPU/memory)
[15-Jan-2010 14:30:32] <rmatte> gwb235: well, like I said, you can monitor the processes via snmp, you just need to add the names of the processes (.exes) to the Processes section
[15-Jan-2010 14:54:33] <forsberg> yrk enoughs enough, im gonna turn to my private problems, as getting the remote working on asrock with xbmc live
[15-Jan-2010 14:54:35] <forsberg> good weekend all :)
[15-Jan-2010 15:12:11] cgibbons_ is now known as cgibbons
[15-Jan-2010 15:13:44] <robo> hi: i'm getting this error when trying to import a MIB: 2010-01-15 16:11:09,762 WARNING zen.zenmib: Unable to find a file providing the OID F5-BIGIP-COMMON-MIB
[15-Jan-2010 15:13:57] <robo> does this mean I need to some how ge the F5-BIGIP-COMMON-MIB file from somewhere?
[15-Jan-2010 15:20:18] <robo> well, i found that MIB but i'm still getting the same error
[15-Jan-2010 15:20:20] <rhettardo> yes
[15-Jan-2010 15:21:01] <rhettardo> append the path to F5-BIGIP-COMMON-MIB to the command you are running
[15-Jan-2010 15:21:08] <gwb235> robo: try doing a zenmib run * in a directory with all the F5 MIBs, that's what I had to do (instead of one at a time)
[15-Jan-2010 15:21:09] <rhettardo> well path/F5-BIGIP-COMMON-MIB
[15-Jan-2010 15:22:30] <robo> here is the command I'm using -- zenmib run /usr/local/zenoss/common/share/mibs/site/F5-BIGIP-COMMON-MIB.txt
[15-Jan-2010 15:22:59] <cgibbons> gwb235: you can actually limit WMI/DCOM to specific dynamic port ranges if you need to control him.
[15-Jan-2010 15:23:11] <robo> ah, zenmib run *
[15-Jan-2010 15:23:11] <cgibbons> er them
[15-Jan-2010 15:23:14] <robo> sounsd like a good idea
[15-Jan-2010 15:23:52] <robo> hmm, that gave me a different error: 2010-01-15 16:23:22,432 ERROR zen.zenmib: -- then said no mibs were loaded
[15-Jan-2010 15:25:57] <cparlette> try "zenmib run -v10" and youll get a lot more output
[15-Jan-2010 15:26:35] <rmatte> yeh, it'll tell you what Mibs it needs to load that one MIB
[15-Jan-2010 15:26:38] <rmatte> then you can do...
[15-Jan-2010 15:26:58] <rmatte> zenmib run ./FIRSTMIB.whatever ./SECONDMIB.whatever ./THIRDMIB.whatever
[15-Jan-2010 15:27:05] <rmatte> to load them in the correct order
[15-Jan-2010 15:27:31] <robo> http://pastie.org/780176
[15-Jan-2010 15:27:34] <robo> those are my errors
[15-Jan-2010 15:27:34] <rmatte> Building the Cisco Mibs was a frigging nightmare
[15-Jan-2010 15:27:51] <robo> i'll try switching the order
[15-Jan-2010 15:28:09] <rmatte> hmmmm
[15-Jan-2010 15:28:17] <rmatte> do you have libsmi2 and libsmi2-common installed?
[15-Jan-2010 15:28:54] <robo> let me see
[15-Jan-2010 15:28:59] <rmatte> some Mibs are dependant on libsmi2
[15-Jan-2010 15:29:07] <rmatte> so make sure it's installed on the system
[15-Jan-2010 15:29:29] <robo> i don't see it installed
[15-Jan-2010 15:29:35] <robo> so ty for the tip, let me try that
[15-Jan-2010 15:29:36] <rmatte> then install it
[15-Jan-2010 15:29:40] <rmatte> np
[15-Jan-2010 15:35:13] <robo> rmatte, i installed libsmi -- same deal
[15-Jan-2010 15:35:19] <robo> but it wasn't installed
[15-Jan-2010 15:36:27] <robo> also installed libmsi-common
[15-Jan-2010 15:36:37] <robo> though it's not libmsi2 -- think that matters?
[15-Jan-2010 15:47:04] <robo> is there a mibs troubleshooting guide anywhere?
[15-Jan-2010 15:53:35] <robo> rmatte, i wasn't doing it as the zenoss user!
[15-Jan-2010 15:53:40] <robo> so that's the problem :-)
[15-Jan-2010 15:57:29] <robo> okay, so the mib imported. How would I tell the F5 device to use the F5 mib?
[15-Jan-2010 15:58:10] <cparlette> the only things that use mibs are snmp traps
[15-Jan-2010 15:58:27] <cparlette> so if you have the F5 sending traps to zenoss, it will look more human readable
[15-Jan-2010 15:58:34] <cparlette> but if youre just doing monitoring, mibs dont matter
[15-Jan-2010 15:59:06] <rmatte> yup
[15-Jan-2010 15:59:10] <rmatte> what he said
[15-Jan-2010 16:01:04] <robo> oh, i'm just need to graph CPU
[15-Jan-2010 16:01:35] <robo> the problem is the cpu utilization graph says missing RRD file
[15-Jan-2010 16:02:12] <robo> and errors in the event log say "Error reading value for "ssCpuRawWait" on..."
[15-Jan-2010 16:02:47] <robo> snmpwalks fine and it reports fine for memory
[15-Jan-2010 16:02:52] <robo> it's just the cpu graph
[15-Jan-2010 16:04:23] <robo> so the F5 doesn't report ssCpuRawWait so I thought adding a MIB will get it to use whatever the F5 uses to report the CPU
[15-Jan-2010 16:06:07] <mrayzenoss> no http://community.zenoss.org/docs/DOC-2445#WhyisZenossnotmonitoringmydevicestuffIveloadedtheMIBs.
[15-Jan-2010 16:06:47] <robo> ah
[15-Jan-2010 16:06:55] <robo> so i need a zenpack for this
[15-Jan-2010 16:07:13] <mrayzenoss> or a custom SNMP Performance template
[15-Jan-2010 16:08:33] <robo> yeah
[15-Jan-2010 16:08:48] <robo> i did a snmpwalk on the f5 and i didn't see any cpu related MIBs
[15-Jan-2010 16:08:57] <robo> or OIDs -- whatever they're called
[15-Jan-2010 16:15:17] <rmatte> yeh, OIDs
[15-Jan-2010 18:51:14] <rrichardsr3> is zenoss a good replacement for nagios? (here come the bullets)
[15-Jan-2010 19:04:56] <cwj> i think it is
[15-Jan-2010 19:05:06] <cwj> since i can use my existing nrpe checks with it
[15-Jan-2010 19:17:38] <rrichardsr3> can you write plugins similar to how you can do with nagios?
[15-Jan-2010 19:17:49] <rrichardsr3> thanks for the response btw ;)
[15-Jan-2010 19:18:42] <cwj> yeah i think you can write plugins in python, or execute scripts remotely via ssh
[15-Jan-2010 19:19:39] <cwj> honestly haven't spent a lot of time with zenoss since 2.4, someone else does monitoring now
[15-Jan-2010 19:21:17] <rrichardsr3> ok cool thanks
[15-Jan-2010 21:25:51] <rayk3n> hi, how can i remove /perf/cpu eventclass from monitoring
[15-Jan-2010 21:54:03] nassy_ is now known as nassy
[16-Jan-2010 00:00:30] [disconnected at Sat Jan 16 00:00:30 2010]
[16-Jan-2010 00:00:30] [connected at Sat Jan 16 00:00:30 2010]
[16-Jan-2010 02:03:13] <rayk3n> hi, how can i remove /perf/cpu eventclass from monitoring
[16-Jan-2010 03:02:48] <rayk3n> asdasdd
[16-Jan-2010 06:15:45] <m4rzito> hi
[16-Jan-2010 06:15:46] <m4rzito> someone?
[16-Jan-2010 06:24:45] <nassy> hi
[16-Jan-2010 06:25:13] <m4rzito> hi nassy
[16-Jan-2010 06:25:35] <m4rzito> I try to monitor JMX with the mbean Runtime, I get StartTime its a epoch miliseconds
[16-Jan-2010 06:25:40] <m4rzito> and I want to print with rrdtool
[16-Jan-2010 06:25:45] <m4rzito> how i setup CDEF?
[16-Jan-2010 06:27:36] <nassy> i dont know but perhaps someone else does
[16-Jan-2010 06:29:12] <m4rzito> :(
[16-Jan-2010 06:48:25] <forsberg> anyone who have succesfully done some zenoss with qstat ? im making som graphs for plaeyrs on servers etc, but im not very happy with how i am presenting the data
[16-Jan-2010 06:50:16] <m4rzito> I try to monitor JMX with the mbean Runtime, I get StartTime its a epoch miliseconds
[16-Jan-2010 06:50:20] <m4rzito> how I CDEF it?
[16-Jan-2010 14:04:36] <sjobeck> hey, hi all. any one point me in the correct direction for moving my SQL db to another server? it is running but complaining it cant contact the DB.
[16-Jan-2010 14:04:39] <sjobeck> thx so much !
[16-Jan-2010 14:05:18] <sjobeck> it seems there is sill one reference to the old DB name some where. I just cant find it. I jsut cant figure out where it is. I have grep'd the entire directory looking for that one last reference to change.
[16-Jan-2010 14:05:47] <Apachez> it only store events in the db
[16-Jan-2010 14:06:42] <sjobeck> ok, thx for clarification, then it cant find the events DB
[16-Jan-2010 14:07:28] <sjobeck> i am not concerned with old data, by the way, fine to point at new empty DB on the dedicated SQL server & go from there
[16-Jan-2010 14:11:56] <sjobeck> in trying to edit the DB reference, I came across this:
[16-Jan-2010 14:11:57] <sjobeck> how about this onebash: /opt/zenoss/Products/ZenModel/zendmd.py: Permission denied
[16-Jan-2010 14:12:09] <sjobeck> bash: /opt/zenoss/Products/ZenModel/zendmd.py: Permission denied
[16-Jan-2010 14:12:20] <sjobeck> after doing a su zenoss -
[16-Jan-2010 14:16:36] <sjobeck> any one know how to execute the command referred to here in the context of moving the DB?
[16-Jan-2010 14:16:37] <sjobeck> http://www.mail-archive.com/zenoss-users@zenoss.org/msg23307.html
[16-Jan-2010 14:22:02] <sjobeck> any one out there ?
[16-Jan-2010 14:38:39] <rdeshpande> hey guys
[16-Jan-2010 14:50:45] RoninX341 is now known as etank
[16-Jan-2010 14:54:34] <rdeshpande> The ruby example in the 'Calling Methods with REST' does not seem to work. I get this error now: /opt/local/lib/ruby/1.8/xmlrpc/client.rb:352:in `new2': Wrong URI as parameter! (RuntimeError) .. any ideas?
[16-Jan-2010 23:20:18] tang_ is now known as tang
[17-Jan-2010 00:00:30] [disconnected at Sun Jan 17 00:00:30 2010]
[17-Jan-2010 00:00:30] [connected at Sun Jan 17 00:00:30 2010]
[17-Jan-2010 00:46:23] rhettardo_ is now known as rhettardo
[18-Jan-2010 00:00:30] [disconnected at Mon Jan 18 00:00:30 2010]
[18-Jan-2010 00:00:31] [connected at Mon Jan 18 00:00:31 2010]
[18-Jan-2010 00:49:13] rhettardo_ is now known as rhettardo
[18-Jan-2010 02:45:42] <cy_`> hello
[18-Jan-2010 02:46:57] <cy_`> hello, again ;)
[18-Jan-2010 02:52:31] <cy_`> hello forsberg
[18-Jan-2010 02:52:36] <forsberg> hello
[18-Jan-2010 02:52:37] <forsberg> :)
[18-Jan-2010 02:59:30] <cy_`> are you danish forsberg? ;)
[18-Jan-2010 03:00:50] <forsberg> yessir :)
[18-Jan-2010 03:01:13] <cy_`> nice
[18-Jan-2010 03:02:21] <cy_`> jeg ogsÃ¥
[18-Jan-2010 03:02:32] <forsberg> :]
[18-Jan-2010 03:02:33] <cy_`> well, half ;)
[18-Jan-2010 03:02:41] <forsberg> good enogh ;)
[18-Jan-2010 03:02:49] <cy_`> i've got the danish passport but i'm actually swiss..
[18-Jan-2010 03:02:59] <cy_`> my mums danish
[18-Jan-2010 03:03:20] <forsberg> ah :)
[18-Jan-2010 03:04:57] <cy_`> yeah
[18-Jan-2010 03:05:18] <cy_`> just as i saw that comendo page..
[18-Jan-2010 03:05:25] <cy_`> you work there? looks like fun
[18-Jan-2010 03:21:13] <forsberg> indeed i do
[18-Jan-2010 03:21:32] <cy_`> cewl ;)
[18-Jan-2010 03:21:57] <forsberg> system admin, with a desperate zenoss problem
[18-Jan-2010 03:22:05] <forsberg> resulting in having to move 400 devices now...
[18-Jan-2010 03:22:31] <cy_`> sounds like not so much fun
[18-Jan-2010 03:34:54] <cy_`> me i'm trying to get another zope prefix for my zenoss than /zport/dmd but i run into the same issues as http://community.zenoss.org/message/44516 does :(
[18-Jan-2010 03:38:56] <forsberg> oi
[18-Jan-2010 03:50:54] <salax> hi, is there any where in informing about alerts other than emailing n pager?
[18-Jan-2010 03:51:11] <salax> can zenoss alerts can be send thru irc?
[18-Jan-2010 03:54:46] <forsberg> its possible sure, just make a mail to irc script ;>

[18-Jan-2010 03:55:22] <forsberg> we have a bot wich checks mail and post on irc
[18-Jan-2010 04:13:29] <salax> forsberg, can u give me clear picture on this?
[18-Jan-2010 04:36:34] <forsberg> well its not combined with zenoss, but have an irc bot check mailbox and parse and paste result on irc
[18-Jan-2010 04:36:57] <forsberg> we parse mails to gms modem for sms alerts
[18-Jan-2010 05:18:36] <Rolfs> BadRequest: The property cRSA does not exist
[18-Jan-2010 05:18:51] <Rolfs> I have added cRSA as a custom property
[18-Jan-2010 05:19:21] <Rolfs> and have changed it via zendmd: d=find("nameofdevice") and d.cRSA="something"
[18-Jan-2010 05:20:07] <Rolfs> Seeing the value in the webinterface goes OK, but chaning the string value from the webinterface after using zendmd does not work.
[18-Jan-2010 05:20:13] <Rolfs> any ideas?
[18-Jan-2010 07:04:41] <chickenandbeans> Afternoon
[18-Jan-2010 07:42:22] MattD is now known as mdereus
[18-Jan-2010 08:16:59] <forsberg> wow the create zenpack and export really works well :>

[18-Jan-2010 08:17:54] <forsberg> when making templates, is there no reason to not keep it base under /Devices
[18-Jan-2010 08:18:06] <forsberg> instead of somewhere in /devices/servers/linux/mtablahbalh
[18-Jan-2010 08:18:28] <forsberg> s/no/any
[18-Jan-2010 08:51:32] <Dieterbe_> not that i know
[18-Jan-2010 09:01:37] <cdd> chip chip wake up
[18-Jan-2010 09:02:09] <cdd> so a pm has been assigned to my monitoring project and with it action items...ugh
[18-Jan-2010 09:02:23] <cdd> Here is what is is asking for:
[18-Jan-2010 09:02:30] <cdd> First step...... is to review the current Distributed Collector Model that was presented by me and "Sanity Test" it
[18-Jan-2010 09:02:44] <cdd> Next action is to bring that configuration to Zenoss engineering support for their validation that we have a proposed solution that will provide the efficiency and reliability we need. We are requesting the "minimum requirements", but also a potential for growth.
[18-Jan-2010 09:02:54] <cdd> Keep in mind, that in this initial stage of consolidation we are looking to deploy Zenoss as overlay tool to centralize our monitoring platform, but as we move forward we will look to "roll up" as much functionality from existing monitoring/management tools as possible. The potential may not exist today, but as Zenoss evolves we may have the opportunity to move towards a more unified monitoring/management platform.
[18-Jan-2010 09:03:03] <cdd> Is this a billable engagment Zenoss would explore?
[18-Jan-2010 09:03:41] <cdd> LOL min reqs but "room to grow"
[18-Jan-2010 09:04:11] <ckrough> cdd: Sounds like you want to talk with the Zenoss Enterprise people
[18-Jan-2010 09:04:38] <cdd> yeah I know going to but I know what the answer probably will be
[18-Jan-2010 09:04:52] <cdd> buy enterprise, receive the eng/support you need
[18-Jan-2010 09:05:25] <cdd> but mgmt isn't buying ent right now they just want zenoss to assist in machine specs
[18-Jan-2010 09:06:05] <aclark> cdd: so what are you asking then?
[18-Jan-2010 09:06:08] * aclark is confused ;-)
[18-Jan-2010 09:06:42] <cdd> they want to buy some billable hours for zenoss to bless the distrubted collector enviroment and to help size the servers appropriately.
[18-Jan-2010 09:06:49] <ckrough> You can pay an hourly consulting fee to them for assitance
[18-Jan-2010 09:07:14] <cdd> i didn't think they were doing that for core
[18-Jan-2010 09:07:24] <cdd> or so I recall last time I looked
[18-Jan-2010 09:07:27] <ckrough> http://www.zenoss.com/support#consulting
[18-Jan-2010 09:07:29] <ckrough> dunno
[18-Jan-2010 09:07:46] <ckrough> Im sure they would work something out if you could turn into an enterprise customer :)
[18-Jan-2010 09:08:17] <cdd> if my signture attority was 10k higher it would be a done deal
[18-Jan-2010 09:09:24] <cparlette> send an email to sales@zenoss.com with what you're looking for
[18-Jan-2010 09:09:31] <cdd> i'm gonna
[18-Jan-2010 09:09:37] <cdd> thx all
[18-Jan-2010 10:03:41] <willwh> howdy folks
[18-Jan-2010 10:03:55] <willwh> mrayzenoss: quick question for you
[18-Jan-2010 10:04:22] <willwh> the IPservices stuff... I want to change the timeout settings and cannot for the life of me find them
[18-Jan-2010 10:08:21] <mrayzenoss> hmmâ€¦ is that the Status Cycle Interval?
[18-Jan-2010 10:08:49] <mrayzenoss> I'm pretty glad we're redoing that stuff for the next release
[18-Jan-2010 10:08:50] <willwh> well, no, I am looking for the timeout settings I guess
[18-Jan-2010 10:08:57] <willwh> I am seeing httpservice down
[18-Jan-2010 10:09:02] <willwh> and then back up again a minute later
[18-Jan-2010 10:10:49] <willwh> so clearly they aren't down, but the checks are timing out
[18-Jan-2010 10:10:58] <mrayzenoss> yeah, I'm looking
[18-Jan-2010 10:11:04] <willwh> :)
[18-Jan-2010 10:11:05] <willwh> ty
[18-Jan-2010 10:11:07] <willwh> core btw
[18-Jan-2010 10:41:04] <rmatte> willwh: I'll see if I can find it
[18-Jan-2010 10:41:42] <mrayzenoss> no one responded to my query
[18-Jan-2010 10:41:45] <rmatte> zStatusConnectTimeout in zProperties I believe
[18-Jan-2010 10:42:07] <rmatte> that's the only property which makes sense
[18-Jan-2010 10:42:33] <mrayzenoss> jeez, how did I overlook that?
[18-Jan-2010 10:42:38] <mrayzenoss> I totally missed that
[18-Jan-2010 10:42:41] <rmatte> ;)
[18-Jan-2010 10:42:49] <rmatte> ctrl-f: "timeout"
[18-Jan-2010 10:42:55] <rmatte> next next
[18-Jan-2010 10:42:56] <mrayzenoss> I checked zProperties in Services, Devices and the configuration of the Collector and the daemons
[18-Jan-2010 10:43:12] <rmatte> ah
[18-Jan-2010 10:43:32] <rmatte> Timeout related properties tend to be under the Device zProperties
[18-Jan-2010 10:43:43] <rmatte> Seems to be the way they've laid it out
[18-Jan-2010 10:44:01] <rmatte> mrayzenoss: your dev guys really need to start releasing screenshots of the new UI :)
[18-Jan-2010 10:44:09] <mrayzenoss> coming soon
[18-Jan-2010 10:44:19] <rmatte> I'm anxious to see what's done so far
[18-Jan-2010 10:44:31] <mrayzenoss> yeah, me too :p
[18-Jan-2010 10:44:38] <rmatte> The screenshot of the wizard page that I saw looked incredible, easily nicest looking web app I've seen
[18-Jan-2010 10:45:23] <chickenandbeans> new UI?
[18-Jan-2010 10:48:28] <rmatte> chickenandbeans: the whole UI is being redesigned
[18-Jan-2010 10:48:31] <mrayzenoss> I don't think we've put anything official out yet, I'll dig up the screenshot from zenethian the other day and the mock-up from our UI consultant
[18-Jan-2010 10:48:38] <rmatte> cool
[18-Jan-2010 10:49:05] <willwh> :)
[18-Jan-2010 10:49:42] <willwh> rmatte: you sure?
[18-Jan-2010 10:49:42] <mrayzenoss> http://designingwebinterfaces.com/blog/wp-content/uploads/2009/07/drag_drop_reorg_zenoss.png
[18-Jan-2010 10:49:50] <willwh> it's set to 15.0 currently
[18-Jan-2010 10:49:59] <willwh> I'm looking at /devices/windows
[18-Jan-2010 10:50:09] <rmatte> willwh: right, 15 seconds, I've seen web requests take up to 40 seconds depending on latency (no joke)
[18-Jan-2010 10:50:16] <willwh> ah
[18-Jan-2010 10:50:24] <mrayzenoss> that was a mock-up, this was a screenshot: http://zen.sebistar.net/images/zenoss_newui_services.png
[18-Jan-2010 10:50:29] <rmatte> bump it up to 30 and see if it solves your problem
[18-Jan-2010 10:50:35] <willwh> the couple of boxes that I am seeing issues with, I guess that is entirely possible
[18-Jan-2010 10:50:42] <willwh> I'll increase to 20 and see how I go
[18-Jan-2010 10:50:47] <willwh> thanks for the assistance, as per :)
[18-Jan-2010 10:50:55] <rmatte> mrayzenoss: looks nice
[18-Jan-2010 10:51:03] <rmatte> np
[18-Jan-2010 10:51:09] <mrayzenoss> we'll have alpha builds with it soon
[18-Jan-2010 10:51:26] <mrayzenoss> this week we'll have a QA day around beta-2.5.2 builds
[18-Jan-2010 10:51:34] <rmatte> nice
[18-Jan-2010 10:51:41] <rmatte> 2.5.2 is due out when?
[18-Jan-2010 10:51:45] <mrayzenoss> no date yet
[18-Jan-2010 10:52:10] <rmatte> I was waiting for you to be like "It's already out, sheesh Ryan, get with the program!"
[18-Jan-2010 10:52:15] <rmatte> since I hadn't checked in a while
[18-Jan-2010 10:52:24] <rmatte> :)
[18-Jan-2010 10:52:36] <mrayzenoss> we don't want to do another maintenance release, trying to dedicate as much resources as possible to Stone Crab, so we want to put out a really solid 2.5.2 and not return to the 2.5.x series
[18-Jan-2010 10:53:24] <rmatte> I agree with that, I would love to see a solid 2.5.2 release
[18-Jan-2010 10:53:35] <rmatte> If it takes longer, then so be it
[18-Jan-2010 10:54:01] <chickenandbeans> oooh pwetty
[18-Jan-2010 10:54:08] <rmatte> I'm going to have to orchestrate an update to 2.5.x at some point, just waiting for a few more things to be fixed
[18-Jan-2010 10:54:10] <mrayzenoss> here's the tickets so far: http://dev.zenoss.org/trac/report/6
[18-Jan-2010 10:54:34] <rmatte> I have 13 Zenoss boxes to upgrade, so failure isn't really an option ;)
[18-Jan-2010 10:54:50] <mrayzenoss> well, we'll have beta 2.5.2 releases to kick around
[18-Jan-2010 10:55:11] <rmatte> those are the tickets for 2.5.2 or for Stone Crab?
[18-Jan-2010 10:55:15] <mrayzenoss> 2.5.2
[18-Jan-2010 10:55:18] <rmatte> cool
[18-Jan-2010 10:55:54] <rmatte> ooooooh
[18-Jan-2010 10:55:54] <rmatte> http://dev.zenoss.org/trac/ticket/5907
[18-Jan-2010 10:55:57] <rmatte> I like that one
[18-Jan-2010 10:56:21] <mrayzenoss> yeah, Kells has been at a customer with a lot of syslog :)
[18-Jan-2010 10:57:06] <rmatte> This one is going to be great for us: http://dev.zenoss.org/trac/ticket/5926
[18-Jan-2010 10:57:26] <rmatte> We've been wanting our front line guys to only have read-only access to the console for a long time
[18-Jan-2010 10:58:04] <mrayzenoss> so you like that behavior? we weren't sure
[18-Jan-2010 10:58:18] <mrayzenoss> zenuser didn't have console access before
[18-Jan-2010 10:58:22] <rmatte> well, in our case we do, but I assume it can be toggled?
[18-Jan-2010 10:58:32] <mrayzenoss> in Zope...
[18-Jan-2010 10:58:46] <rmatte> That's fine then
[18-Jan-2010 10:59:06] <rmatte> our front line guys work out of the ticketing system and only really login to Zenoss to view stuff
[18-Jan-2010 10:59:18] <mrayzenoss> yeah, zenuser is supposed to be read-only
[18-Jan-2010 10:59:19] <rmatte> eventually we're going to have our ticketing system automatically moving events to history when tickets are closed
[18-Jan-2010 10:59:33] <mrayzenoss> but before it didn't read the console
[18-Jan-2010 10:59:39] <mrayzenoss> we thought that seemed kinda limiting
[18-Jan-2010 10:59:43] <rmatte> oh, they need to be able to see the console
[18-Jan-2010 10:59:46] <rmatte> just not touch anything
[18-Jan-2010 10:59:50] <mrayzenoss> agreed
[18-Jan-2010 11:00:43] <rmatte> actually...
[18-Jan-2010 11:00:52] <rmatte> is that what the purpose of the ZenManager group is?
[18-Jan-2010 11:01:07] <rmatte> I could never figure out the purpose of that group
[18-Jan-2010 11:01:18] <rmatte> It doesn't actually give manager level permissions
[18-Jan-2010 11:01:41] <rmatte> Wondering if it was originally meant just to give read/write access to the event console
[18-Jan-2010 11:02:49] <rmatte> well, if all of those tickets make it in to the 2.5.2 list I should be a relatively happy camper
[18-Jan-2010 11:03:25] <rmatte> Although the big one for me is: http://dev.zenoss.org/trac/ticket/5978
[18-Jan-2010 11:03:35] <rmatte> since that totally screws over a bunch of my templates
[18-Jan-2010 11:03:52] <rmatte> I've figured out how to do a dirty fix for it by hand, but it'd be nice to actually see it fixed
[18-Jan-2010 11:07:37] <mrayzenoss> I believe ZenManager can manage events, but not change settings?
[18-Jan-2010 11:08:28] <rmatte> I speculate that's the use for it, but I've no idea, since as it currently stands ZenUser and ZenManager don't do anything differently
[18-Jan-2010 11:08:33] <rmatte> guess we'll see in the next version
[18-Jan-2010 11:25:16] <rmatte> hmmm, hackman hasn't responded to my PM yet
[18-Jan-2010 11:25:32] <rmatte> :(
[18-Jan-2010 11:25:47] <rmatte> Looks like I'll have to make the error report from scratch
[18-Jan-2010 11:26:04] <rmatte> I just hate reinventing the wheel
[18-Jan-2010 11:26:46] <rmatte> Did they just completely delete the http://www.zenoss.com/Members directory?
[18-Jan-2010 11:26:58] <rmatte> or was it moved/backed up somewhere?
[18-Jan-2010 11:28:34] <mrayzenoss> rmatte: http://community.zenoss.org/people/ ?
[18-Jan-2010 11:29:22] <rmatte> Well, yeh, but I can't get to http://community.zenoss.org/people/hackman238/Errors-Discards-Duplex.zip
[18-Jan-2010 11:29:24] <rmatte> for example
[18-Jan-2010 11:29:32] <rmatte> so I guess a bunch of files were lost in the migration
[18-Jan-2010 11:31:00] <rmatte> aha, I might be able to just clean up the code that's here: http://community.zenoss.org/thread/9297
[18-Jan-2010 11:31:29] <rmatte> ah, he has a link to the file right in that post too
[18-Jan-2010 11:31:38] <rmatte> sweet
[18-Jan-2010 11:33:08] <rmatte> ah, the files he linked to are for version 2.5 only
[18-Jan-2010 11:34:20] <rmatte> hmmm, it'd take forever to clean up the white spacing in these
[18-Jan-2010 11:34:55] <rmatte> omg, I found the old cleaned up version that I did kicking around
[18-Jan-2010 11:34:57] <rmatte> yay
[18-Jan-2010 11:42:47] <rmatte> now to test it out
[18-Jan-2010 11:47:40] <rmatte> An Interface Errors report should be available by default
[18-Jan-2010 11:47:44] <rmatte> it's a pretty standard report
[18-Jan-2010 12:07:09] <rmatte> man, I had this report working at some point in the past, now it's erroring out
[18-Jan-2010 12:07:22] <kobalt> hows everyone doing
[18-Jan-2010 12:09:54] <rmatte> good, yourself?
[18-Jan-2010 12:10:21] <kobalt> good had a great new year got married =)
[18-Jan-2010 12:10:30] <rmatte> cool, congrats
[18-Jan-2010 12:11:13] <kobalt> yeah its nice to be back to my admin duties, we had a spot in our noc that needed to be covered so I was working that for a bit
[18-Jan-2010 12:12:26] <mrayzenoss> congrats
[18-Jan-2010 12:12:36] <kobalt> Thanks
[18-Jan-2010 12:13:09] <rmatte> lol, I'd hate to go back to doing noc work
[18-Jan-2010 12:13:35] <rmatte> used to be fun back when it wasn't below my skill level
[18-Jan-2010 12:14:01] <kobalt> heh
[18-Jan-2010 12:14:15] <kobalt> but I got to use Zenoss first hand so it helped a little
[18-Jan-2010 12:16:42] <kobalt> any idea's how I would do a threshold on percent of change? so if the data drops by 15% create an alert? Im looking online to try to figure it out hehe not getting far yet
[18-Jan-2010 12:17:44] <mrayzenoss> kobalt: if you're using COUNTER for your data point, that's rate of change
[18-Jan-2010 12:17:53] <mrayzenoss> threshold set accordingly
[18-Jan-2010 12:18:03] <rmatte> I think I finally got this report to work, woot
[18-Jan-2010 12:18:31] <kobalt> mrayzenoss: interface counters are using derive
[18-Jan-2010 12:18:37] <rmatte> kobalt: well, what are you thresholding on?
[18-Jan-2010 12:19:01] <kobalt> rmatte: interface bandwidth
[18-Jan-2010 12:19:14] <rmatte> actually, hmmm, my idea would mean that you'd need to be able to fetch the previous value... which I'm not sure how to do
[18-Jan-2010 12:20:00] <kobalt> yeah thats what I was trying to figure out
[18-Jan-2010 12:20:00] <mrayzenoss> DERIVED is same as COUNTER, just allows negative values
[18-Jan-2010 12:20:04] <rmatte> I was thinking that you could set the max and min thresholds to something like here.getRRDValue('whatever') * 0.15, or whatever
[18-Jan-2010 12:20:27] <kobalt> I do have min and max thresholds set
[18-Jan-2010 12:20:30] <rmatte> but you'd need to have here.getRRDValues fetch the last value before the current one somehow
[18-Jan-2010 12:20:30] <mrayzenoss> hmmâ€¦ I've seen this somewhere before
[18-Jan-2010 12:20:52] <rmatte> I actually needed to research this for specialized hard drive templates that I was working on
[18-Jan-2010 12:21:09] <kobalt> but what we are looking at is rate of change so if a customer is sitting at 40mb on a 100 mb interface and drops to 20 then that would be a indication of a possible issue
[18-Jan-2010 12:21:09] <rmatte> I wanted to make it so that I could set it based on quick percentages of change rather than total usage
[18-Jan-2010 12:21:39] <rmatte> kobalt: right, so do that you'd have to figure out how to grab the previous value from RRD
[18-Jan-2010 12:21:44] <rmatte> then it'd be a piece of cake
[18-Jan-2010 12:22:01] <kobalt> yeah
[18-Jan-2010 12:22:08] <rmatte> Zenoss dynamically works out where the thresholds are based on the previous value (15% above and 15% below previous value)
[18-Jan-2010 12:22:20] <rmatte> then if the current value exceeds either of those thresholds it generates an alert
[18-Jan-2010 12:22:27] <rmatte> the tricky part is, how do you fetch the previous value
[18-Jan-2010 12:22:40] <rmatte> there must be some way to do it with getRRDValue function
[18-Jan-2010 13:29:08] <kobalt> threw a post up to see if I get a response, thanks for your help rmatte
[18-Jan-2010 13:37:52] <rmatte> np
[18-Jan-2010 13:40:35] <Egyptian[laptop]> evening all
[18-Jan-2010 13:40:40] <Egyptian[laptop]> is the zenoss repo working?
[18-Jan-2010 13:44:29] <rmatte> Egyptian[laptop]: I assume it's not working for you?
[18-Jan-2010 13:45:24] <Egyptian[laptop]> actually .. i misread it .. its working but cant find the exact file
[18-Jan-2010 13:45:35] <Egyptian[laptop]> zenoss-2.5.1-440.el5.x86_64: failure: zenoss-2.5.1.el5.x86_64.rpm from zenoss: [Errno 256] No more mirrors to try.
[18-Jan-2010 13:45:49] <mrayzenoss> weird, that's sourceforge
[18-Jan-2010 13:45:50] <Egyptian[laptop]> and browsing the repo by hand shows that version 440 does not exist
[18-Jan-2010 13:46:35] <Egyptian[laptop]> notice the rpm name?
[18-Jan-2010 13:46:53] <mrayzenoss> Egyptian[laptop]: here's the link from SF directly: https://sourceforge.net/projects/zenoss/files/zenoss-2.5%20%28Latest%20Stable%29/zenoss-2.5.1/zenoss-2.5.1.el5.x86_64.rpm/download
[18-Jan-2010 13:47:19] <Egyptian[laptop]> i am already downloading it by hand but i wanted to do it via yum
[18-Jan-2010 13:47:39] <Egyptian[laptop]> but it is looking for a different package name
[18-Jan-2010 13:47:44] <mrayzenoss> http://dev.zenoss.org/yum/stable/
[18-Jan-2010 13:48:09] <Egyptian[laptop]> mrayzenoss: yah that one
[18-Jan-2010 13:48:22] <mrayzenoss> the RPM downloaded for me no problems
[18-Jan-2010 13:49:32] <mrayzenoss> if you can recreate your steps to get the issue in a forum post and send me the link, I'll run it by our build engineer. I wouldn't be surprised if something's wonky with SourceForge
[18-Jan-2010 13:51:58] <Egyptian[laptop]> i am using the yum repo and this is the message i get:
[18-Jan-2010 13:52:01] <Egyptian[laptop]> Error Downloading Packages:
[18-Jan-2010 13:52:03] <Egyptian[laptop]> zenoss-2.5.1-440.el5.i386: failure: zenoss-2.5.1.el5.i386.rpm from zenoss: [Errno 256] No more mirrors to try.
[18-Jan-2010 13:52:03] <Egyptian[laptop]> zenoss-2.5.1-440.el5.x86_64: failure: zenoss-2.5.1.el5.x86_64.rpm from zenoss: [Errno 256] No more mirrors to try.
[18-Jan-2010 13:52:53] <mrayzenoss> hmmâ€¦ Glenn appears to be out right now. I'll get him to look at it when he's back, I just need a forum post or something to link him to
[18-Jan-2010 13:54:40] <Egyptian[laptop]> sorry .. not registered yet
[18-Jan-2010 13:55:01] <rmatte> then register lol
[18-Jan-2010 13:55:04] <rmatte> it only takes a second
[18-Jan-2010 13:57:46] <Egyptian[laptop]> rmatte: ok whats the link?
[18-Jan-2010 13:58:02] <Egyptian[laptop]> and isnt there an issue queue this should go into?
[18-Jan-2010 13:58:07] <Egyptian[laptop]> a bugzilla? or mantis? or trac?
[18-Jan-2010 13:58:28] <rmatte> trac
[18-Jan-2010 13:58:33] <mrayzenoss> go ahead and put it on trac
[18-Jan-2010 13:58:45] <mrayzenoss> http://dev.zenoss.org/trac/ user/pass zenoss/zenoss
[18-Jan-2010 13:58:46] <rmatte> http://community.zenoss.org is the community site with the forum (immensely useful)
[18-Jan-2010 13:59:00] <rmatte> and trac is obviously self explanatory
[18-Jan-2010 13:59:50] <rmatte> generally I don't post to trac until I've determined that it's not an isolated issue (something that's wrong on my server but which can't be reproduced elsewhere)
[18-Jan-2010 14:00:53] <Egyptian[laptop]> hmm..would there be any general documentation on data center monitoring? to include for eg. environmental sensors? i am thinking fo documentation where zenoss plays a part of a bigger picture
[18-Jan-2010 14:01:11] <Egyptian[laptop]> rmatte: hokay if you say so
[18-Jan-2010 14:02:53] <Egyptian[laptop]> oh .. and is it possible to _disable_ network autodiscovery?
[18-Jan-2010 14:05:54] * Egyptian[laptop] thinks i broke the spell
[18-Jan-2010 14:07:33] <Egyptian[laptop]> figured out the problem .. the package name is wrong
[18-Jan-2010 14:08:15] <Egyptian[laptop]> rpm -qpi zenoss-2.5.1.el5.x86_64.rpm shows its release is 440.el5 but this is not reflected in the package name
[18-Jan-2010 14:13:43] <Egyptian[laptop]> umm.. where in the forum do i post this?
[18-Jan-2010 14:16:19] <mrayzenoss> zenoss-users would be fine
[18-Jan-2010 14:16:39] <mrayzenoss> I'll get someone to look at it
[18-Jan-2010 14:17:18] * Egyptian[laptop] is watching the story of cat stevens
[18-Jan-2010 14:30:49] <Egyptian[laptop]> mrayzenoss: done ..thanks
[18-Jan-2010 14:35:51] <rmatte> hmmmm, trying to figure out how to attach a file to a wiki article
[18-Jan-2010 14:35:57] <rmatte> in Jive I mean
[18-Jan-2010 14:36:00] <rmatte> I assume it's possible?
[18-Jan-2010 14:37:34] <mrayzenoss> yeah, at the bottom of the edit page there's "Attach Files"
[18-Jan-2010 14:39:38] <rmatte> ah k
[18-Jan-2010 14:40:28] <rmatte> there's a maximum of 5?
[18-Jan-2010 14:40:30] <rmatte> eugh
[18-Jan-2010 14:40:50] <rmatte> Figures, I have 7 files to attach
[18-Jan-2010 14:41:30] <mrayzenoss> I'll see if I can bump the limitâ€¦ what are you attaching?
[18-Jan-2010 14:41:46] <mrayzenoss> images and video are in-line
[18-Jan-2010 14:41:50] <mrayzenoss> and not attachements
[18-Jan-2010 14:45:16] <rmatte> python files
[18-Jan-2010 14:45:31] <rmatte> I'm releasing my stand-alone reports scripts
[18-Jan-2010 14:45:40] <rmatte> since people will undoubtedly find them useful
[18-Jan-2010 14:47:32] <mrayzenoss> perhaps zip them up in 1 zip?
[18-Jan-2010 14:47:39] <mrayzenoss> or tarball
[18-Jan-2010 14:48:14] <rmatte> well, I sort of wanted people to be able to pick and choose which they wanted to grab, but yeh, guess I'll have to do that
[18-Jan-2010 14:51:23] <mrayzenoss> rmatte: how many attachments are you going to add?
[18-Jan-2010 14:51:28] <mrayzenoss> I found the setting
[18-Jan-2010 14:52:53] <foppa> 7
[18-Jan-2010 14:52:55] <foppa> :)
[18-Jan-2010 14:53:02] <foppa> [21:40:51] <rmatte> Figures, I have 7 files to attach
[18-Jan-2010 14:55:50] <mrayzenoss> aight, bumped it to 10
[18-Jan-2010 15:08:25] <rmatte> mrayzenoss: http://community.zenoss.org/docs/DOC-4664
[18-Jan-2010 15:08:39] <rmatte> I just ended up dropping them on my server
[18-Jan-2010 15:13:03] <ckrough> rmatte: '
[18-Jan-2010 15:13:03] <ckrough> Stand-alone reports scripts
[18-Jan-2010 15:13:05] <ckrough> nice!
[18-Jan-2010 15:17:23] <rmatte> yup
[18-Jan-2010 15:17:27] <rmatte> ;)
[18-Jan-2010 15:17:52] <rmatte> we're using them to automate our monthly customer reports
[18-Jan-2010 15:17:59] <rmatte> (along with a lot of php magic)
[18-Jan-2010 15:18:01] <rmatte> lol
[18-Jan-2010 15:18:39] <mrayzenoss> awww, I bumped up the attachment limit so you could add them individually
[18-Jan-2010 15:19:01] <rmatte> lol, I appreciate that but it was too late lol
[18-Jan-2010 15:19:34] <mrayzenoss> I just hope dmon.org has good SLAs, I might make it a tip of the month :p
[18-Jan-2010 15:19:49] <rmatte> dmon.org is running in my home office lol
[18-Jan-2010 15:20:06] <rmatte> but I doubt that a few downloads of python files is really going to hurt it
[18-Jan-2010 15:20:12] <rmatte> I have business grade cable internet
[18-Jan-2010 15:20:21] <foppa> plol :)
[18-Jan-2010 15:26:21] <rmatte> I'll link them from jive, just going to take a minute or two
[18-Jan-2010 15:27:56] <rmatte> eugh, of course it zipped them all
[18-Jan-2010 15:28:03] <rmatte> bleh.
[18-Jan-2010 15:29:15] <rmatte> meh, I'll just leave it as is
[18-Jan-2010 15:29:30] <rmatte> having to unzip .py files individually will be annoying for people
[18-Jan-2010 15:32:17] <rmatte> yeh, combined the python files are 40KB
[18-Jan-2010 15:32:33] <rmatte> so I doubt my connection is going to have issues
[18-Jan-2010 15:43:05] <mrayzenoss> Zenoss QA Test Day January 21 âˆ’ 2.5.2 Beta: http://bit.ly/7wT0Fw
[18-Jan-2010 16:29:57] <mrayzenoss> New Example Techniques ZenPack from Chet: http://community.zenoss.org/docs/DOC-4671
[18-Jan-2010 16:48:23] <rmatte> nice
[18-Jan-2010 16:48:29] <rmatte> I'll load that up on my lab box
[18-Jan-2010 16:49:21] <rmatte> definitely going to check out the custom graph report stuff
[18-Jan-2010 16:51:08] <rmatte> woot, our custom cmdb just got a bunch of major updates today which will actually make it usable for once lol
[18-Jan-2010 17:59:19] <MattWoo> Anyone have problems with a memory leaK? Is it appropriate to ask this here?
[18-Jan-2010 18:07:53] <rmatte> MattWoo, well, what is making you think you have one?
[18-Jan-2010 18:19:16] <MattWoo> rmatte, Free Swap gradually falls from 2 GB to 500 MB over a period of 2 weeks after booting the system. Somewhere between 100 MB and 500 MB, the system starts behaving erratically. Alarms aren't send, data collection for graphs stops.
[18-Jan-2010 18:19:43] <rmatte> well how much physical memory is in the box?
[18-Jan-2010 18:20:26] <MattWoo> 4 GB
[18-Jan-2010 18:20:35] <rmatte> and how many devices are you monitoring?
[18-Jan-2010 18:20:57] <MattWoo> 32-bit CentOS 5.3 with 1 processor. Zenoss 2.4.5
[18-Jan-2010 18:21:52] <Apachez> sounds bad
[18-Jan-2010 18:21:58] <Apachez> what other stuff do you have running?
[18-Jan-2010 18:22:06] <Apachez> you could lower swapiness
[18-Jan-2010 18:22:24] <rmatte> I have a VM with 5 gigs of RAM running Ubuntu Server and Zenoss 2.4.5 monitoring 352 devices, and it's only using 3.5GB of RAM right now.
[18-Jan-2010 18:22:33] <MattWoo> The system is definitely over-utilized. Load is 2-4 all day. 598 devices monitored including 75 Win servers. SNMP monitoring on about 70% of devices. Lots of Cisco routers and switches.
[18-Jan-2010 18:22:52] <rmatte> are you running anything other than Zenoss on it?
[18-Jan-2010 18:23:01] <MattWoo> No, nothing besides Zenoss
[18-Jan-2010 18:23:07] <rmatte> yeh, 598 devices will do that man
[18-Jan-2010 18:23:15] <rmatte> you need to stick more RAM in that thing
[18-Jan-2010 18:23:19] <MattWoo> But a memory leak?
[18-Jan-2010 18:23:24] <rmatte> no
[18-Jan-2010 18:23:27] <rmatte> it's not a memory leak
[18-Jan-2010 18:23:35] <rmatte> monitoring that many devices via SNMP is extremely intensive
[18-Jan-2010 18:23:41] <MattWoo> I am not a very knowlegeable Linux guy
[18-Jan-2010 18:23:52] <rmatte> Zenoss does a lot of caching of data, the more devices you have the more caching happens
[18-Jan-2010 18:23:56] <rmatte> hence the more RAM usage
[18-Jan-2010 18:24:05] <rmatte> and ultimately swap usage if you don't have enough RAM
[18-Jan-2010 18:24:20] <rmatte> honestly, for monitoring that many devices you'd need 6 to 8 gigs of RAM
[18-Jan-2010 18:24:34] <MattWoo> So, steadily falling free swap in a Linux system is a sure sign of the system not having enough RAM?
[18-Jan-2010 18:24:55] <rmatte> correct, what's happening is that as Zenoss runs it's caching more data...
[18-Jan-2010 18:25:09] <rmatte> it'll eventually top out and stop taking up more space
[18-Jan-2010 18:25:17] <rmatte> but you'd need enough physical memory for it to use
[18-Jan-2010 18:25:47] <kobalt> 1600 devices here 90% switches, and I am using 32gb of ram and 136k of swap =)
[18-Jan-2010 18:25:50] <MattWoo> So, following that logic, if I had 8 GB of RAM, the system would start behaving weird in 4 weeks instead of 2. Is that right?
[18-Jan-2010 18:26:10] <MattWoo> Not trying to be snarky
[18-Jan-2010 18:26:21] <rmatte> I'm monitoring a little over half the devices that you are on one of my Zenoss boxes and I have 5.7GB of RAM in that Zenoss VM
[18-Jan-2010 18:26:30] <rmatte> no
[18-Jan-2010 18:26:38] <rmatte> I literally just said it would eventually top out
[18-Jan-2010 18:26:45] <rmatte> it wouldn't just continuously consume memory
[18-Jan-2010 18:27:03] <rmatte> Zenoss always starts with not much memory usage and then the usage increases
[18-Jan-2010 18:27:11] <MattWoo> I see
[18-Jan-2010 18:27:13] <rmatte> when I start Zenoss on that box it takes up about 1.5GB of Memory
[18-Jan-2010 18:27:14] <kobalt> I found with 6 gigs I could run about 500 devices
[18-Jan-2010 18:27:20] <rmatte> which increases until it hits about 3.5 to 4
[18-Jan-2010 18:27:24] <rmatte> then it settles at that point
[18-Jan-2010 18:27:28] <rmatte> you have twice the devices that I have
[18-Jan-2010 18:27:34] <MattWoo> So, I guess i need 8 GB to be fairly safe.
[18-Jan-2010 18:27:38] <rmatte> so I'd guess yours would settle around 5.5
[18-Jan-2010 18:27:52] <rmatte> Yeh, if you can get more RAM in there, I'd definitely go for 8GB
[18-Jan-2010 18:28:05] <rmatte> Then you won't have the swap space usage issues anymore
[18-Jan-2010 18:28:09] <MattWoo> My system load is 2-4 or 5 most of the day. Advice?
[18-Jan-2010 18:28:23] <MattWoo> Could that be the result of not enough RAM too?
[18-Jan-2010 18:28:25] <rmatte> with that many devices I'm not surprised
[18-Jan-2010 18:28:35] <rmatte> my box with 352 devices averages about 2 to 2.5
[18-Jan-2010 18:28:50] <rmatte> so it's not unusual for a box with almost twice as many devices to be at 2 to 4
[18-Jan-2010 18:28:52] <MattWoo> Or do I just need more CPU
[18-Jan-2010 18:29:10] <rmatte> What kind of speed CPU do you have now?
[18-Jan-2010 18:29:41] <MattWoo> 2.2 GHz I think - could be 2.4. Any way I can find that from the command line?
[18-Jan-2010 18:29:58] <MattWoo> CentOS
[18-Jan-2010 18:30:09] <rmatte> You're probably polling around 40000 datapoints every polling cycles, which means that system has 5 minutes each polling cycle to collect all that data
[18-Jan-2010 18:30:21] <rmatte> maybe 30000 but even still, very intensive
[18-Jan-2010 18:30:35] <rmatte> My box with 352 devices is polling about 18000
[18-Jan-2010 18:30:57] <rmatte> yeh, 2.2GHz is low
[18-Jan-2010 18:31:02] <MattWoo> Data Points graph hits 20K
[18-Jan-2010 18:31:03] <rmatte> I'd say minimum of 3GHz
[18-Jan-2010 18:31:16] RoninX341 is now known as elake
[18-Jan-2010 18:31:17] <Apachez> 30k doesnt sound much
[18-Jan-2010 18:31:25] <Apachez> unless the core of zenoss is really bad
[18-Jan-2010 18:31:29] <rmatte> k, so your datapoints aren't much higher than mine, but then again my server can use up to 10GHz if it needs to
[18-Jan-2010 18:31:32] <MattWoo> Do I need really fast disk I/O?
[18-Jan-2010 18:31:36] <kobalt> MattWoo: cat /proc/cpuinfo
[18-Jan-2010 18:31:39] <Apachez> 100 per second is nothing
[18-Jan-2010 18:31:43] <rmatte> disk I/O is really the main bottle neck
[18-Jan-2010 18:31:59] <rmatte> I have my Zenoss VMs running off a large SAN so access speeds are really high and it does help a lot
[18-Jan-2010 18:32:07] <MattWoo> model name : AMD Opteron(tm) Processor 148
[18-Jan-2010 18:32:07] <MattWoo> stepping : 1
[18-Jan-2010 18:32:07] <MattWoo> cpu MHz : 2210.217
[18-Jan-2010 18:32:11] <rmatte> I noticed an instant improvement when we moved to the SAN from local disks
[18-Jan-2010 18:32:11] <MattWoo> Sorry
[18-Jan-2010 18:32:21] <kobalt> so your proc is 2.2
[18-Jan-2010 18:32:21] <rmatte> if you could get a RAID 5 array or something at least, then it'd help
[18-Jan-2010 18:32:29] <MattWoo> We do have a SAN available
[18-Jan-2010 18:32:43] <Apachez> have you done any tuning?
[18-Jan-2010 18:32:49] <rmatte> kobalt: 2.2GHz is pretty weak for monitoring almost 600 devices on a single collector lol
[18-Jan-2010 18:33:17] <Apachez> demanding 2.2 ghz to monitor just 600 devs are just lolo
[18-Jan-2010 18:33:18] <MattWoo> I am not sure about tuning. I have done some basic MySQL tunin.
[18-Jan-2010 18:33:42] <MattWoo> I am also not terribly smart about MySQL.
[18-Jan-2010 18:33:53] <Apachez> the mysql is only used for events
[18-Jan-2010 18:33:59] <Apachez> basically syslog/snmptraps things
[18-Jan-2010 18:34:03] <rmatte> MattWoo: besides hardware upgrades you can try some of this: http://community.zenoss.org/docs/DOC-2521
[18-Jan-2010 18:34:21] <MattWoo> I have figured out how to optimize tables and I have a script that gets rid of a lot of debug events. My tables are no longer growing forever.
[18-Jan-2010 18:34:23] <rmatte> MattWoo: but I really recommend CPU and Memory upgrades for a start (memory especially)
[18-Jan-2010 18:35:11] <rmatte> Then maybe look in to using that SAN instead of local disks
[18-Jan-2010 18:35:33] <kobalt> on my system with the 6 gigs of ram I have a dual xeon DC procs at 1.8
[18-Jan-2010 18:36:10] <rmatte> we have between 15 and 30GHz in each of our blades, but that's split up between a ton of VMs
[18-Jan-2010 18:36:27] <MattWoo> Well, you all have been generous with your advice. Thanks. I think I will start with a RAM upgrade and move to a faster system with faster CPU if necessary. I might try the tuning doc you pointed out after teh RAM upgrade.
[18-Jan-2010 18:36:30] <kobalt> hehe server boast
[18-Jan-2010 18:36:39] <rmatte> The new blade we'll be getting will have 40GHz total and about 60GB of RAM, can't wait
[18-Jan-2010 18:36:47] <rmatte> ;)
[18-Jan-2010 18:36:55] <MattWoo> Gotta go home
[18-Jan-2010 18:37:00] <rmatte> cool
[18-Jan-2010 18:37:14] <rmatte> me too, getting late, time to head home
[18-Jan-2010 18:37:16] <rmatte> cheers guys
[18-Jan-2010 18:37:24] <kobalt> Later rmatte
[18-Jan-2010 18:43:17] etank is now known as elake
[18-Jan-2010 18:45:37] <ironpaw> anyone around?
[18-Jan-2010 18:45:53] <ironpaw> why can't i get alerts when interfaces reach the threshold ?
[18-Jan-2010 18:46:02] <ironpaw> even when a template is bound?
[18-Jan-2010 18:48:12] <rmatte> how do you even know that it's hitting the threshold?
[18-Jan-2010 18:48:38] <rmatte> and are you saying that you're literally not seeing an event for it in the event console
[18-Jan-2010 18:48:46] <rmatte> or your configured alerts aren't firing off for those events?
[18-Jan-2010 19:11:13] elake is now known as etank
[18-Jan-2010 21:52:51] theacoly1e- is now known as theacolyte
[19-Jan-2010 00:00:30] [disconnected at Tue Jan 19 00:00:30 2010]
[19-Jan-2010 00:00:30] [connected at Tue Jan 19 00:00:30 2010]
[19-Jan-2010 03:25:20] <Rolfs> How do I change custom properties from zendmd ?
[19-Jan-2010 03:25:44] <Rolfs> d=find("server-name")
[19-Jan-2010 03:26:01] <Rolfs> d.customproperty = "new string" ???
[19-Jan-2010 09:11:20] <mrayzenoss> anyone on here using the WMI MSSQL ZenPack? People keep giving me suggestions for how to fix it and I can't test it
[19-Jan-2010 09:13:03] <mrayzenoss> or if anyone wants to test the WMI queries against a SQL Server
[19-Jan-2010 09:26:27] <rmatte> someone in here was going to try fixing that pack as I recall...
[19-Jan-2010 09:26:55] <rmatte> theacolyte I believe
[19-Jan-2010 09:27:02] <rmatte> He could probably assist you with that
[19-Jan-2010 09:27:24] <chickenandbeans> Afternoon
[19-Jan-2010 09:27:34] <rmatte> g'day
[19-Jan-2010 09:28:27] <mrayzenoss> there are a bunch of suggestions on http://community.zenoss.org/docs/DOC-4641 and I don't know what's right
[19-Jan-2010 09:28:33] <chickenandbeans> written a nice python script that echo's a device template dmd commands :)
[19-Jan-2010 09:28:51] <chickenandbeans> *templates
[19-Jan-2010 09:29:03] <rmatte> chickenandbeans: you mean the commands that the template could support
[19-Jan-2010 09:29:05] <rmatte> ?
[19-Jan-2010 09:30:03] <chickenandbeans> no, let's say you want to export the dmd commands for a template on deviceX
[19-Jan-2010 09:30:25] <chickenandbeans> my script does it. Kind of an export via the command line
[19-Jan-2010 09:30:47] <rmatte> I see
[19-Jan-2010 09:30:50] <rmatte> example of output?
[19-Jan-2010 09:32:56] <chickenandbeans> well I say 'script', my python-fu is quite poor
[19-Jan-2010 09:32:56] <chickenandbeans> wow my wireless is terrible
[19-Jan-2010 09:33:31] <chickenandbeans> well let's take the standard device template..
[19-Jan-2010 09:33:45] <chickenandbeans> cattemplate -t mydevice Device would echo
[19-Jan-2010 09:33:59] <rmatte> mrayzenoss: I'm working on interface templates, I already have a pretty nice collection of them but once I'm done with what I have (I'll have 20 or more templates), I'll fire them over to you as a ZenPack and maybe they can be included in a future release (it would save people time from having to make their own)
[19-Jan-2010 09:34:21] <mrayzenoss> that sounds very handy
[19-Jan-2010 09:34:26] <chickenandbeans> dmd.manage_addRRDTemplate("Device") etc
[19-Jan-2010 09:34:48] <rmatte> I see
[19-Jan-2010 09:34:53] <mrayzenoss> chickenandbeans: so it's the stuff that's available when you tab-complete in zendmd on that template?
[19-Jan-2010 09:35:01] <chickenandbeans> pretty much
[19-Jan-2010 09:35:13] <chickenandbeans> but i did it before tab complete... :'( the pain
[19-Jan-2010 09:35:15] <chickenandbeans> :)
[19-Jan-2010 09:35:18] <rmatte> lol
[19-Jan-2010 09:35:35] <rmatte> I just patched my 2.4.5 installs so that they have tab completion
[19-Jan-2010 09:35:47] <mrayzenoss> sounds good for helping people see what's available. I'll put it into the queue for a tip of the month candidate at some point
[19-Jan-2010 09:35:53] <chickenandbeans> I am stuck on a problem though, which i hope will be solved on thursday in the irc dev chat
[19-Jan-2010 09:35:56] <mrayzenoss> since it may be handy for documentation
[19-Jan-2010 09:36:16] <rmatte> chickenandbeans: which problem?
[19-Jan-2010 09:37:38] <chickenandbeans> http://community.zenoss.org/thread/12497?tstart=0
[19-Jan-2010 09:37:40] <chickenandbeans> :)
[19-Jan-2010 09:37:58] <chickenandbeans> it's a tough one, but for a dev it should be really easy
[19-Jan-2010 09:38:35] <chickenandbeans> once done, i'll write my tools, test them in anger on prod then release 'em somehow
[19-Jan-2010 09:42:58] * chickenandbeans tries convert 12pm EST into GMT
[19-Jan-2010 09:43:38] <rmatte> chickenandbeans: you shouldn't really be trying to pipe anything between python scripts
[19-Jan-2010 09:43:42] <rmatte> you're in a bash mindset
[19-Jan-2010 09:43:53] <chickenandbeans> yes, but why not?
[19-Jan-2010 09:43:59] <chickenandbeans> it works with cat
[19-Jan-2010 09:44:06] <chickenandbeans> it's all just text
[19-Jan-2010 09:44:09] <rmatte> well, I'm sure there's a much more elegant way to do what you're trying to do
[19-Jan-2010 09:44:15] <rocket> rmatte: you can if you handle stdin within the python script .. :p
[19-Jan-2010 09:44:44] <chickenandbeans> yeah, it's a simple readline thingy
[19-Jan-2010 09:45:00] <rmatte> rocket: right, but I don't see why you wouldn't just incorporate the scripts together, or have the one script import the functions from the other script
[19-Jan-2010 09:45:03] <rmatte> but that's just me
[19-Jan-2010 09:45:05] <chickenandbeans> but it doesn't work, i know it's not the python because, i can do it other ways
[19-Jan-2010 09:45:14] <rocket> rmatte: I didnt see what he was doing though :p
[19-Jan-2010 09:45:26] <rmatte> chickenandbeans: you've coded the second script to accept the stdout from the first script?
[19-Jan-2010 09:45:36] <rmatte> rocket: nor did I
[19-Jan-2010 09:45:50] <chickenandbeans> it's more I have coded two useful scripts.
[19-Jan-2010 09:46:10] <chickenandbeans> they should work on their own or via piped input
[19-Jan-2010 09:46:18] <chickenandbeans> again, this isn't a python thing
[19-Jan-2010 09:46:24] <rocket> all python? or a mix of python and bash?
[19-Jan-2010 09:46:31] <rocket> or all bash?
[19-Jan-2010 09:46:33] <chickenandbeans> bash? lol
[19-Jan-2010 09:46:37] <chickenandbeans> no just python
[19-Jan-2010 09:46:40] <chickenandbeans> :)
[19-Jan-2010 09:46:59] <rmatte> maybe if you actually pastebin the scripts we could spot the issue
[19-Jan-2010 09:47:05] <rmatte> rocket is amazing at python
[19-Jan-2010 09:47:12] <rocket> chickenandbeans: I usally create a class or module out of one of the scripts and use the class as a library in my main script
[19-Jan-2010 09:48:40] <chickenandbeans> yes, sounds cool.
[19-Jan-2010 09:49:09] <chickenandbeans> But I really would love to know how to have multiple connections. It would be *so* useful in future scripting
[19-Jan-2010 09:49:35] <chickenandbeans> I haven't seen one example of it, I want to ask the devs whether it's possible
[19-Jan-2010 09:50:05] <rocket> multiple connections?
[19-Jan-2010 09:50:09] <chickenandbeans> I mean, one can do it when connecting to mysql
[19-Jan-2010 09:50:22] <chickenandbeans> yah, the stack trace indicates some sort of locking issue
[19-Jan-2010 09:51:02] <chickenandbeans> I reckon it'll be a really easy one for the devs
[19-Jan-2010 09:52:36] <rmatte> rocket: http://community.zenoss.org/thread/12497?tstart=0
[19-Jan-2010 09:52:43] <rmatte> that's the error he's referring to
[19-Jan-2010 09:52:48] <rmatte> you joined right after he pasted the link
[19-Jan-2010 09:52:58] <chickenandbeans> ah sorry
[19-Jan-2010 09:54:03] <chickenandbeans> It's weird because I am not performing any writes to zope. Really just reading values
[19-Jan-2010 09:55:44] <rocket> hrmmm I see what your getting at .. trying to figure out if there is a bug in zenscriptbase that basically disallows multiple connections
[19-Jan-2010 09:56:10] <rocket> or two scripts at the same time calling that code ...
[19-Jan-2010 09:58:25] <chickenandbeans> yeah, that's it
[19-Jan-2010 09:58:41] <chickenandbeans> a few other devs here have noticed the same thing
[19-Jan-2010 09:58:43] <mrayzenoss> off to a meetingâ€¦ enhancement reviews
[19-Jan-2010 09:58:53] <chickenandbeans> sound fun ;)
[19-Jan-2010 09:59:08] <mrayzenoss> just letting you know we do look at them ;)
[19-Jan-2010 10:01:07] <rocket> I would guess its a zope bug if anything
[19-Jan-2010 10:01:43] <rocket> maybe not .. looking at Cmdbase...
[19-Jan-2010 10:02:07] <chickenandbeans> yeah, I reckon it's in zenscriptbase
[19-Jan-2010 10:02:26] <chickenandbeans> could be a reason for a few gui, time out issues too
[19-Jan-2010 10:03:28] <kobalt> mrayzenoss, I have installed the WMIMSSQL but have not tested it yet, is there things that need to be fixed?
[19-Jan-2010 10:04:48] <mrayzenoss1> kobalt: look at the thread under the ZenPack, I just wanted someone to sanity check the various WMI queries
[19-Jan-2010 10:08:58] <rocket> chickenandbeans: which version of zope are you running or version of zenoss?
[19-Jan-2010 10:11:24] <chickenandbeans> 2.5.1
[19-Jan-2010 10:12:14] <chickenandbeans> so that's zop 2.11.2
[19-Jan-2010 10:12:18] <chickenandbeans> *zope even
[19-Jan-2010 10:12:26] <chickenandbeans> I love these eee's
[19-Jan-2010 10:14:59] <rocket> based on the traceback its an interaction with zope .. something about that interaction isnt happy
[19-Jan-2010 10:15:13] <rocket> I dont have my virtual machine available at the moment to debug ..
[19-Jan-2010 10:15:15] <chickenandbeans> yeah, my money's on the transaction.py.. but then again
[19-Jan-2010 10:15:25] <chickenandbeans> i'm just guessing
[19-Jan-2010 10:17:15] <rocket> ok want to try something for me ..
[19-Jan-2010 10:17:52] <rocket> comment out the commit() in ZenScriptBase and see if it works .. though it my blow up horribly
[19-Jan-2010 10:18:24] <rocket> eg http://dev.zenoss.com/trac/ticket/5128 seems related
[19-Jan-2010 10:18:37] <chickenandbeans> mmm i'll check it out
[19-Jan-2010 10:22:53] <chickenandbeans> by jove!
[19-Jan-2010 10:24:48] <chickenandbeans> yeah, that seemed to work
[19-Jan-2010 10:25:10] <chickenandbeans> can't use it till I understand it though :)
[19-Jan-2010 10:25:40] <rocket> I think commit ends up getting called twice .. at least thats what the ticket I posted is mentioning
[19-Jan-2010 10:26:25] <chickenandbeans> mmm
[19-Jan-2010 10:26:44] <rocket> eg the transactions that need to commit will later in the code
[19-Jan-2010 10:27:11] <chickenandbeans> i'll update my forum post with bit of info (i'll give you a shoutout :) ). I'd like to know the effect of this change on other things....
[19-Jan-2010 10:27:29] <rocket> you can also remove
[19-Jan-2010 10:27:50] <rocket> from transaction import commit
[19-Jan-2010 10:28:02] <rocket> as that method isnt used anymore
[19-Jan-2010 10:28:21] <chickenandbeans> yeah, are you a he or a she...
[19-Jan-2010 10:28:26] <chickenandbeans> writing the forum post :)
[19-Jan-2010 10:28:33] <rocket> he
[19-Jan-2010 10:31:03] <eidolon> *sigh* does anyone else have problems with zenoss getting slow to respond on web hits? sometimes i pause 10-15 seconds to get a page update.
[19-Jan-2010 10:31:21] <eidolon> (we have 123 devices monitored)
[19-Jan-2010 10:33:58] <rmatte> eidolon: yes, it will slow down during polling cycles
[19-Jan-2010 10:34:07] <eidolon> meh!
[19-Jan-2010 10:34:47] <rmatte> eidolon: when the box is using all of it's processing power to try and collect the data within the 5 minute window it slows the UI down
[19-Jan-2010 10:35:06] <eidolon> anything to be done about it? splitting services, anything like that?
[19-Jan-2010 10:36:05] <rmatte> eidolon: you can try tweaking some daemon settings to see if you can improve the performance
[19-Jan-2010 10:36:13] <eidolon> i worry because we're just going to add more and more machines to this host.
[19-Jan-2010 10:36:15] <rmatte> eidolon: using any command based datasources?
[19-Jan-2010 10:36:22] <eidolon> a few. not many.
[19-Jan-2010 10:36:30] <rmatte> lots of servers?
[19-Jan-2010 10:36:32] <eidolon> mostly basic performance templates.
[19-Jan-2010 10:36:32] <rmatte> or switches?
[19-Jan-2010 10:36:45] <eidolon> 123 devices. 95% servers.
[19-Jan-2010 10:36:54] <rmatte> yeh, monitoring servers is pretty intensive
[19-Jan-2010 10:37:01] <rmatte> you doing it via SNMP, WMI, or both?
[19-Jan-2010 10:41:20] <chickenandbeans> Ok, updated the forum post :)
[19-Jan-2010 10:42:02] <chickenandbeans> will keep it open and see if we can get to the bottom of this
[19-Jan-2010 10:45:45] <otakup0pe> hey i have a zenoss question ! how do i get a report showing total network transfers for an interface for a given time period
[19-Jan-2010 10:46:02] <otakup0pe> like the interface utilization report just totals instead of max/average
[19-Jan-2010 10:48:20] <rmatte> otakup0pe: honestly, if you take the average and multiply it by the period you'd get pretty close
[19-Jan-2010 10:48:34] <rmatte> but I don't know of any quick way to see the actual total
[19-Jan-2010 10:50:24] <eidolon> meh. i have /devices/server/linux on most of my devices. i want to add ntp checking on all servers. if i look up the class /device/server/linux, i can 'bind templates...' and att ntp monitor, but that doesn't seem right. what i really want to do is modify the tenmplate to include ntpmonitor. c'est possible?
[19-Jan-2010 10:51:05] <rmatte> why wouldn't that "seem" right?
[19-Jan-2010 10:51:12] <rmatte> you can have multiple templates bound
[19-Jan-2010 10:51:22] <rmatte> so if you bind Device and that other template they'll work together
[19-Jan-2010 10:52:17] <rmatte> go to bind templates, hold down the ctrl key, select both templates then click OK
[19-Jan-2010 10:52:27] <rmatte> you can bind both to the whole class or individually per device, it's up to you
[19-Jan-2010 10:53:17] <chickenandbeans> otakup0pe: I've seen some zendmd scripts knocking about that allow for this on the zenoss wiki, but I don't know how to do it via the gui. Sorry mate
[19-Jan-2010 10:56:24] <chickenandbeans> http://community.zenoss.org/docs/DOC-2528 - should be too hard to write a script based on tis to produce a template and a graph.
[19-Jan-2010 10:58:33] <eidolon> rmatte: yeah, but i'm binding on the class, not on the host itself
[19-Jan-2010 10:58:36] <eidolon> that's what's puzzling
[19-Jan-2010 10:58:45] <eidolon> i know i can bind multiple templates to a host.
[19-Jan-2010 10:58:57] <rmatte> you can bind multiple templates to a class as well
[19-Jan-2010 10:59:31] <eidolon> ahh. okay. so i'm in /zport/dms/Devices/Server/Linux/perfConfig
[19-Jan-2010 10:59:41] <eidolon> and i do bind templates, select both the linux template and ntpmonitor
[19-Jan-2010 10:59:51] <eidolon> i should see the ntp 'offset' perf template in all my hosts now, yes?
[19-Jan-2010 10:59:58] <rmatte> yes
[19-Jan-2010 10:59:58] <eidolon> (all that use that template)
[19-Jan-2010 11:00:10] <rmatte> well, if you're binding to the class you'll see it on every one
[19-Jan-2010 11:00:21] <rmatte> though if Zenoss can't gather the data you'll just see a blank graph
[19-Jan-2010 11:00:26] <rmatte> hence why it's better to be a bit more selective
[19-Jan-2010 11:00:49] <rmatte> if you have some that support it and some that don't I'd say it would be best to make a new organizer, something like /Server/Linux/NTP
[19-Jan-2010 11:00:56] <rmatte> then throw the hosts that support it in there
[19-Jan-2010 11:01:24] <rmatte> or, you can bind it individually to each server which supports it
[19-Jan-2010 11:01:31] <eidolon> well, okay, so the host i'm looking at is showing Device : /DEvices/Server/Linux
[19-Jan-2010 11:01:39] <chickenandbeans> Also binding a template to a group (or whatever) and a device underneath it causes badness.
[19-Jan-2010 11:02:01] <rmatte> chickenandbeans: I've never noticed that
[19-Jan-2010 11:02:10] <chickenandbeans> we have :)
[19-Jan-2010 11:02:16] <rmatte> define badness
[19-Jan-2010 11:02:16] <eidolon> and if i click on that template, I see 'Available Performance Tempaltes' - only /Devices/Server/Linux has a checkbox next to it, even though under Bind Performance Templates I see 'ntp monitor' is highlighted.
[19-Jan-2010 11:02:40] <chickenandbeans> basically after a restart, monitoring stops for all devices under that grouping
[19-Jan-2010 11:02:40] <rmatte> eidolon: yeh, that checkbox means nothing man, that just means that that template is present at that level
[19-Jan-2010 11:02:50] <rmatte> whereas the ntp template is obviously present at a higher level
[19-Jan-2010 11:02:54] <eidolon> okay. but i still don't see the ntp monitor chart.
[19-Jan-2010 11:02:56] <rmatte> you can still bind both of them
[19-Jan-2010 11:03:00] <eidolon> i did bind them :)
[19-Jan-2010 11:03:21] <rmatte> you're literally not seeing a graph at all?
[19-Jan-2010 11:03:30] <rmatte> or the graph just isn't generating?
[19-Jan-2010 11:03:53] <eidolon> literally not seeing the graph. it's not a broken image, not a no data. it's just not there.
[19-Jan-2010 11:04:09] <rmatte> have you checked more than one device?
[19-Jan-2010 11:04:37] <rmatte> ...and have you checked the template to make sure that it's ok (not missing graph or anything)?
[19-Jan-2010 11:05:10] <ckrough> rmatte: how do you usually pull a specific OID when you script it? 1) walk the descriptions to get the index number, then do an snmpget for the OID with that index?
[19-Jan-2010 11:05:24] <ckrough> sorry. 1) is all there is. no 2).
[19-Jan-2010 11:05:25] <eidolon> yes, checked 2-3 other devices.
[19-Jan-2010 11:05:38] <eidolon> one had ntpmonitor explicitely bound, and that worked. but i don't want to have to do that to all hosts.
[19-Jan-2010 11:05:42] <rmatte> ckrough: you mean such as in my Windows performance zenpack?
[19-Jan-2010 11:06:27] <rmatte> eidolon: remove all the explicit bindings
[19-Jan-2010 11:06:35] <rmatte> from what chickenandbeans is saying, those can cause issues
[19-Jan-2010 11:06:40] <eidolon> ung
[19-Jan-2010 11:06:45] <ckrough> rmatte: nah, pulling OIDs from IOS specifically
[19-Jan-2010 11:06:47] <eidolon> that means i need to go through all my hosts anyway.
[19-Jan-2010 11:06:50] <eidolon> fuck that.
[19-Jan-2010 11:07:10] <rmatte> afk for 5 mins
[19-Jan-2010 11:11:10] <chickenandbeans> You could try writing a script using zendmd
[19-Jan-2010 11:11:49] <chickenandbeans> to try and reset the bindings
[19-Jan-2010 11:12:29] <eidolon> yea, i already puit in time with my boss to work on learning enough dmd API stuff to be able to do magic.
[19-Jan-2010 11:12:33] <eidolon> haven't allocated it yet :)
[19-Jan-2010 11:24:31] <rmatte> yeh, dmd magic is very useful
[19-Jan-2010 11:35:59] <chickenandbeans> right home time
[19-Jan-2010 11:36:09] <chickenandbeans> rmatte, rocket... thanks for your help guys
[19-Jan-2010 11:38:57] <rmatte> np
[19-Jan-2010 11:39:15] <rmatte> eidolon: what are you pulling from IOS?
[19-Jan-2010 11:39:20] <rocket> np
[19-Jan-2010 11:39:25] <eidolon> wait wha huh?
[19-Jan-2010 11:39:30] * eidolon isn't doing anythign from ios
[19-Jan-2010 11:39:41] <rmatte> oh nevermind
[19-Jan-2010 11:39:49] <rmatte> ckrough: what are you pulling from IOS?
[19-Jan-2010 11:39:58] <eidolon> :)
[19-Jan-2010 11:40:04] <rmatte> lol
[19-Jan-2010 12:19:39] <ckrough> rmatte: CPU utilization, sometimes temp
[19-Jan-2010 12:29:45] <rmatte> ckrough: then why would you need to pull the index?
[19-Jan-2010 12:29:55] <rmatte> the OIDs are always the same on IOS
[19-Jan-2010 12:30:12] <ckrough> I dont think that's correcy
[19-Jan-2010 12:30:17] <ckrough> it depends on what slot the card is in
[19-Jan-2010 12:30:44] <ckrough> and it seems to vary from version to version
[19-Jan-2010 12:31:19] <ckrough> the question was more about, is there a more scalable way to do it
[19-Jan-2010 12:31:36] <ckrough> I cant think of one
[19-Jan-2010 12:52:51] <rmatte> oh, well for temperature maybe, the OID for CPU shouldn't change
[19-Jan-2010 12:53:25] <rmatte> and there really isn't any easy scalable way to do it, you'd have to code a collector plugin, then add a tab to the device page and have the info dynamically displayed there
[19-Jan-2010 12:53:29] <rmatte> really the only way to do it
[19-Jan-2010 12:53:58] <rmatte> I created devices groups for each number of CPUs when I made my windows performance packs
[19-Jan-2010 12:54:05] <rmatte> but slots on cisco devices are much more dynamic
[19-Jan-2010 12:55:02] <rmatte> I'd just figure out what the max amount of slots are then write a plugin which collects the data, and if the OID for a slot doesn't exist then the script outputs 0 for that datapoint
[19-Jan-2010 12:55:10] <rmatte> then you just always have it graph the datapoints
[19-Jan-2010 12:55:15] <rmatte> the ones that are always 0 don't exist
[19-Jan-2010 12:55:20] <rmatte> that's the simplest way of doing it
[19-Jan-2010 12:55:37] <rmatte> well, not a plugin, but a script, and use it as a command based datasource
[19-Jan-2010 12:56:14] <ckrough> the way I've been doing it is to walk the description table and seach for 'CPU for Switching', then snag the index off of that and poll it
[19-Jan-2010 12:56:31] <ckrough> yeah, via command based ds
[19-Jan-2010 12:56:46] <rmatte> 'CPU for switching'?
[19-Jan-2010 12:57:00] <rmatte> are you getting individual CPU info for stacked devices or something?
[19-Jan-2010 12:57:05] <rmatte> or for some sort of blade setup?
[19-Jan-2010 12:57:36] <ckrough> the supervisor card has it
[19-Jan-2010 12:57:41] <ckrough> for 650x chassis
[19-Jan-2010 12:57:56] <rmatte> I see, I generally only collect CPU info for the supervisor
[19-Jan-2010 12:58:37] <ckrough> so, for example:
[19-Jan-2010 12:58:38] <ckrough> SNMPv2-SMI::mib-2.47.1.1.1.1.2.3001 = STRING: "CPU of Switching Processor"
[19-Jan-2010 12:58:44] <ckrough> SNMPv2-SMI::mib-2.47.1.1.1.1.2.3017 = STRING: "CPU of Routing Processor
[19-Jan-2010 12:59:18] <rmatte> I just collect the cpu5min value
[19-Jan-2010 12:59:34] <rmatte> but I guess you require more detailed reporting
[19-Jan-2010 13:00:35] <ckrough> what OID do you use for that? THis is a 6509, I dont think they return that value
[19-Jan-2010 13:00:50] <ckrough> 1.3.6.1.4.1.9.2.1.58.0
[19-Jan-2010 13:01:14] <ckrough> hmm, this one does
[19-Jan-2010 13:01:17] <ckrough> arg
[19-Jan-2010 13:02:35] <ckrough> sorry, yeah... more detailed. we need the SP CPU as well
[19-Jan-2010 13:21:16] <rmatte> yeh
[19-Jan-2010 13:21:30] <rmatte> you'll have to do it with a custom script
[19-Jan-2010 13:43:33] <rmatte> eugh, I love how one interface type doesn't support the same OIDs as other interface types
[19-Jan-2010 13:43:54] <rmatte> like, the standard ethernet template does work for certain other types like frame-relay
[19-Jan-2010 13:43:59] <rmatte> but it doesn't work for any of these...
[19-Jan-2010 13:44:04] <rmatte> ATM Sub Interface
[19-Jan-2010 13:44:04] <rmatte> ATT DS0 Point (64 Kbps)
[19-Jan-2010 13:44:04] <rmatte> Asymetric Digitial Subscriber Loop (ADSL)
[19-Jan-2010 13:44:04] <rmatte> ISDN
[19-Jan-2010 13:44:04] <rmatte> ISDN S_T Interface
[19-Jan-2010 13:44:05] <rmatte> Link Access Protocol D (LAPD)
[19-Jan-2010 13:44:07] <rmatte> Voice Encapsulation
[19-Jan-2010 13:44:09] <rmatte> Voice Over IP Encapulation
[19-Jan-2010 13:44:11] <rmatte> atm
[19-Jan-2010 13:44:13] <rmatte> ds1
[19-Jan-2010 13:44:15] <rmatte> ds3
[19-Jan-2010 13:44:17] <rmatte> iso88025TokenRing
[19-Jan-2010 13:44:19] <rmatte> propPointToPointSerial
[19-Jan-2010 13:44:21] <rmatte> propVirtual
[19-Jan-2010 13:44:23] <rmatte> sonet
[19-Jan-2010 13:44:27] <rmatte> so I have to figure out what OIDs I should be using instead for each one
[19-Jan-2010 13:46:40] <kobalt> the propvirtual works for the standard ethernet oid except errors
[19-Jan-2010 13:47:03] <kobalt> I have sonet set to collect CV ES SES for my sonet interfaces
[19-Jan-2010 13:47:07] <rmatte> hmmmm, it wasn't working for me
[19-Jan-2010 13:47:53] <rmatte> I'll try it out again
[19-Jan-2010 13:47:55] <kobalt> what Im using it on are Unrouted Vlan interfaces
[19-Jan-2010 13:48:14] <rmatte> I'm using it on actual serial interfaces
[19-Jan-2010 13:48:33] <kobalt> not sure on that I know the Vlan interfaces work
[19-Jan-2010 13:48:51] <kobalt> and they show as propvirtual and propvirtual_64
[19-Jan-2010 13:48:53] <rmatte> well, going to check snmp and figure it out, maybe it does work and I just overlooked it
[19-Jan-2010 13:49:02] <rmatte> the other templates are going to be a pain though
[19-Jan-2010 13:49:17] <rmatte> since I know that nothing standard works on some of them
[19-Jan-2010 13:49:24] <kobalt> yeah the sonet ones are fun lol
[19-Jan-2010 13:49:37] <rmatte> what's CV, ES, and SES?
[19-Jan-2010 13:50:13] <kobalt> CV code viloation ES errored Seconds and SES Severly Errored Seconds
[19-Jan-2010 13:50:19] <rmatte> ah
[19-Jan-2010 13:50:26] <rmatte> nothing to do with utilization though?
[19-Jan-2010 13:50:27] <kobalt> basicly any errors on the sonet
[19-Jan-2010 13:50:43] <kobalt> not at this point
[19-Jan-2010 13:51:13] <rmatte> k
[19-Jan-2010 13:51:22] <rmatte> if I figure it out I'll pass the info along
[19-Jan-2010 13:51:55] <kobalt> I can do some reasearch on it I have 60 sonet nodes (cisco 15454 nodes)
[19-Jan-2010 13:52:02] <rmatte> cool
[19-Jan-2010 13:52:18] <rmatte> it'd help since that would be one less template that I have to make
[19-Jan-2010 13:52:27] <rmatte> once I have these sorted out I intend to release them as a pack
[19-Jan-2010 13:52:52] <kobalt> nice
[19-Jan-2010 13:53:39] <rmatte> or hopefully get them included in Zenoss by default
[19-Jan-2010 13:54:01] <rmatte> interface templates are a pain in the rear to develop lol
[19-Jan-2010 13:54:33] <rmatte> I wish Zenoss could tell the difference between physical and virtual frame relay interfaces
[19-Jan-2010 13:54:57] <rmatte> I had to create local copies of the frame-relay template on each of the virtual frame relay interfaces and disable the stuff for errors
[19-Jan-2010 13:55:10] <rmatte> since virtual interfaces don't support error info collection, only the physical interface does
[19-Jan-2010 13:55:37] <kobalt> yeah I had to do the same with the Vlan interfaces
[19-Jan-2010 13:55:46] <rmatte> yeah, it's a huge pain
[19-Jan-2010 13:55:52] <rmatte> takes up tons of time and very hard to maintain
[19-Jan-2010 13:56:44] <rmatte> eugh, this snmpwalk is taking forever
[19-Jan-2010 13:57:09] <rmatte> I should have picked a device with better throughput
[19-Jan-2010 13:57:17] <rmatte> this is some router in L.A, can't believe how slow it is
[19-Jan-2010 13:58:42] <kobalt> lol thankfully all my stuff is in the same state
[19-Jan-2010 14:00:32] <rmatte> lol
[19-Jan-2010 14:00:40] <rmatte> well, we have all sorts of different clients with stuff all over the place
[19-Jan-2010 14:00:52] <rmatte> this particular clients has stuff all over the U.S, Europe, and Mexico
[19-Jan-2010 14:01:19] <rmatte> We have lots of stuff that we monitor in Canada as well
[19-Jan-2010 14:01:54] <kobalt> seattle to spokane and everywhere between almost hehe for us
[19-Jan-2010 14:02:00] <rmatte> cool
[19-Jan-2010 14:36:05] <etank> Error
[19-Jan-2010 14:36:16] <etank> Confirmation password is empty or invalid. Please confirm your password for security reasons. <-- im getting this error
[19-Jan-2010 14:36:29] <etank> when i try to edit an email for a user
[19-Jan-2010 14:36:53] <etank> is anyone else seeing this?
[19-Jan-2010 14:37:15] <rmatte> do you have Manager privileges?
[19-Jan-2010 14:37:24] <mrayzenoss1> are you entering your credentials when you try to change the email address?
[19-Jan-2010 14:37:56] <rmatte> If you're a manager you don't need to enter a password when changing it
[19-Jan-2010 14:38:02] <mrayzenoss1> in 2.5 you do
[19-Jan-2010 14:38:03] <etank> i am manager
[19-Jan-2010 14:38:04] <rmatte> just modify the email address and click save
[19-Jan-2010 14:38:12] <etank> this is 2.5 ent
[19-Jan-2010 14:38:14] mrayzenoss1 is now known as mrayzenoss
[19-Jan-2010 14:38:18] <rmatte> mrayzenoss: oh?
[19-Jan-2010 14:38:24] <etank> we are using ldap authentication though
[19-Jan-2010 14:38:35] <mrayzenoss> yeah, it was added because of security concerns
[19-Jan-2010 14:38:36] <etank> i enter my password and that is the error i get
[19-Jan-2010 14:38:46] <mrayzenoss> ohâ€¦ that's not good
[19-Jan-2010 14:38:56] <rmatte> mrayzenoss: I'm thinking it was added but only works for the user which you're trying to change the setting for
[19-Jan-2010 14:38:57] <etank> mrayzenoss: no kidding :)
[19-Jan-2010 14:39:02] <rmatte> lol
[19-Jan-2010 14:39:19] <mrayzenoss> etank: waitâ€¦ so as admin you can't change other users' email addresses?
[19-Jan-2010 14:39:31] <etank> mrayzenoss: mine or others actually
[19-Jan-2010 14:39:59] <etank> i am just trying to enter the email since it is blank currently
[19-Jan-2010 14:40:08] <rmatte> mrayzenoss: I'm guessing that the feature was never tested with LDAP
[19-Jan-2010 14:40:28] <rmatte> It was probably tested with admin and regular local accounts
[19-Jan-2010 14:40:32] <mrayzenoss> no, we test with LDAP
[19-Jan-2010 14:40:39] <rmatte> hmmm
[19-Jan-2010 14:40:45] <etank> im wondering if there is something that i need to set up in ldap to make it work
[19-Jan-2010 14:40:57] <etank> ldap config in zope that is
[19-Jan-2010 14:40:57] <rmatte> etank: doubtful
[19-Jan-2010 14:41:14] <mrayzenoss> well, it works in 2.5.1 Core w/out LDAPâ€¦ I'd ping Support on this
[19-Jan-2010 14:41:34] <etank> i have authentication and user_enumeration checked in /zport/acl_users
[19-Jan-2010 14:42:24] <etank> hard to get notification with no email :/
[19-Jan-2010 14:43:31] <rmatte> you could probably just bypass Zenoss and set it directly in zope management interface
[19-Jan-2010 14:43:34] <rmatte> until it's fixed
[19-Jan-2010 14:43:50] <etank> or drop ldap for the moment
[19-Jan-2010 14:43:51] <rmatte> just navigate to that user's settings then put /manage after the url, then click on the properties tab in the top right
[19-Jan-2010 14:44:02] <rmatte> then throw the email address in and save the settings
[19-Jan-2010 14:44:04] <etank> yeah i was just in /manage
[19-Jan-2010 14:45:15] <etank> the error makes it look like im trying to change the password since it is asking for the confirmation password
[19-Jan-2010 14:47:00] <etank> well it is a work around at least
[19-Jan-2010 14:47:06] <rmatte> yup
[19-Jan-2010 14:47:29] <rmatte> mrayzenoss: I don't honestly even see the need for that since anyone who knows zope in the slightest could easily work around it
[19-Jan-2010 14:47:46] <rmatte> mrayzenoss: It's more of a nuissance than anything
[19-Jan-2010 14:47:50] <etank> rmatte: but dont they have to have manage ability?
[19-Jan-2010 14:48:19] <rmatte> etank: yes, but why should someone who has Manager privileges be prompted for their credentials when changing another user's info?
[19-Jan-2010 14:48:22] <rmatte> doesn't make sense
[19-Jan-2010 14:48:52] <rmatte> you're already logged in, and if it's meant as a secondary layer of security, the fact that you can bypass it with ZMI makes it useless
[19-Jan-2010 14:49:05] <mrayzenoss> if you walk away from your computer and leave it unattended, someone could come over and change settings. It's a security requirement
[19-Jan-2010 14:49:18] <etank> the only time i can see it being useful is if i leave my desktop unlocked
[19-Jan-2010 14:49:59] <rmatte> mrayzenoss: honestly, a more effective method would be having a spot where you could set a master email address where Zenoss sends alerts to in regards to any changes to the system
[19-Jan-2010 14:50:06] <rmatte> that way you'd know about it and be able to just change it back
[19-Jan-2010 14:50:28] <mrayzenoss> yeah, there's an enhancement ticket to add logging of all changes
[19-Jan-2010 14:50:41] <rmatte> but like I said, you can work around that with ZMI
[19-Jan-2010 14:51:00] <rmatte> I suppose it would be good for preventing changes by users who haven't touched Zenoss before
[19-Jan-2010 14:51:08] <rmatte> but anyone with a bit of experience would be able to figure that out
[19-Jan-2010 14:51:21] <mrayzenoss> yeah, we're working towards locking down everything eventually
[19-Jan-2010 14:51:26] <rmatte> cool
[19-Jan-2010 14:51:43] <rmatte> hopefully not to the point where it's prompting for a username and password for every change made
[19-Jan-2010 14:51:44] <etank> did someone say to open a support ticket for this?
[19-Jan-2010 14:51:53] <rmatte> mrayzenoss did
[19-Jan-2010 14:52:00] <etank> ok. will do.
[19-Jan-2010 14:52:01] <mrayzenoss> I did. If the LDAP stuff isn't working that way, that needs to be addressed
[19-Jan-2010 14:52:08] <etank> agreed
[19-Jan-2010 14:52:33] <mrayzenoss> might be a good 2.5.2 candidate especially
[19-Jan-2010 14:52:37] <rmatte> I mean, I understand the need for security but I also hate apps that take it too far and make it a nuissance to use the app
[19-Jan-2010 14:53:18] <mrayzenoss> you probably don't touch user settings too frequently, so the hassle should be minimal
[19-Jan-2010 14:53:35] <mrayzenoss> it's not like we're requiring user/pass for the event console :p
[19-Jan-2010 14:53:45] <rocket> I would say thats a relatively safe place to say few changes occur .. :p
[19-Jan-2010 14:54:30] <rmatte> mrayzenoss: no, I was thinking more along the lines of needing it every time a template is edited or something, if it slows down development work then it becomes annoying
[19-Jan-2010 14:54:47] <rmatte> rocket: lol
[19-Jan-2010 14:55:26] <mrayzenoss> rmatte: agreed, we're only adding it where we have to
[19-Jan-2010 14:55:46] * rocket needs some good humor ... /me is frustrated with work ... grr sometimes my job doesnt match my skills ... :p
[19-Jan-2010 14:55:47] <rmatte> I lock my workstation religiously so it's not an issue with me, but I see lots of other people around here who don't
[19-Jan-2010 14:56:02] <rmatte> and the odd time an email gets sents from their workstation professing their love for the same sex
[19-Jan-2010 14:56:38] <rmatte> rocket: doesn't match your skills meaning too simple or too complex?
[19-Jan-2010 14:57:07] <mrayzenoss> etank: just heard that's a known and fixed Enterprise bug
[19-Jan-2010 14:57:21] <etank> mrayzenoss: what was the fix?
[19-Jan-2010 14:57:28] <etank> or is it a 2.5.2 thing
[19-Jan-2010 14:57:30] <mrayzenoss> dunno, we're in the defect review right now
[19-Jan-2010 14:57:45] <mrayzenoss> Support should have the patch for you, but it'll be rolled up into 2.5.2
[19-Jan-2010 14:57:53] <etank> ok cool.
[19-Jan-2010 14:59:12] <rocket> gotta run to a meeting
[19-Jan-2010 14:59:28] <rmatte> have fun
[19-Jan-2010 15:24:31] <rmatte> mhmmm, I have a Cisco ASA which I can collect performance data on fine but I can't collect interface utilization info
[19-Jan-2010 15:24:38] <rmatte> the interface types are ethernetCsmacd
[19-Jan-2010 15:24:43] <rmatte> but there are no RRD files
[19-Jan-2010 15:29:35] <ckrough> the dreaded 'Code: 2 - Msg: Misuse of shell builtins'... anyone know how to get more detail from it?
[19-Jan-2010 15:30:26] <rmatte> check your log files
[19-Jan-2010 15:34:20] * ckrough installs Net::SNMP and facepalms
[19-Jan-2010 15:34:34] <rmatte> hmmmm
[19-Jan-2010 15:34:41] <ckrough> perl based command
[19-Jan-2010 15:34:43] <ckrough> new install
[19-Jan-2010 15:35:06] <rmatte> propPointToPointSerial looks like it does use the same OIDs as ethernet, wonder why it wasn't working previously
[19-Jan-2010 15:35:11] <rmatte> I'll have to test it out
[19-Jan-2010 15:35:52] <kobalt> rmatte: everything I have been looking at with SONET its a layer 1 protocal, so no way from the SONET side to actually caluclate bandwidth, other then by capacity
[19-Jan-2010 15:36:59] <rmatte> ah
[19-Jan-2010 15:39:30] <Kristopher424> I think I've messed up my temperature graphs. :/ They just don't load. I've got the RPN as /,10,1.8,*,32,+ Limit -1, type Line, Width 1, Format %5.2lf, Legend ${graphPoint/id} under the graph point.
[19-Jan-2010 15:40:10] <Kristopher424> The Data Source is enabled, with correct OID, type: Gauge.
[19-Jan-2010 15:40:55] <kobalt> wouldnt it be 10,/,1.8,*,32,+ thats how I have mine
[19-Jan-2010 15:41:21] <Kristopher424> kobalt, I'll give it a shot.
[19-Jan-2010 15:42:08] <Kristopher424> kobalt, worked! Thanks... The RPN that I gave works fine on our Trango radios and yours works on the Dragonwave radios. Wierd.
[19-Jan-2010 15:42:50] <kobalt> Kristopher424: well it works backwords so divide by 10 is 10,/
[19-Jan-2010 15:43:59] <Kristopher424> Cool. Thanks for the help.
[19-Jan-2010 15:44:12] <rmatte> well, what's the actual mathematical calculation again?
[19-Jan-2010 15:44:23] <rmatte> Kristopher424: you're going from celsius to farenheit?
[19-Jan-2010 15:44:38] <Kristopher424> yes.
[19-Jan-2010 15:45:20] <rmatte> Farenheit=1.8(Celsius)+32
[19-Jan-2010 15:45:44] <rmatte> so it would be: ,1.8,*,32,+
[19-Jan-2010 15:45:59] <Kristopher424> Right, but our radios (for some dumb reason) have a built-in *10 to any value queried by SNMP
[19-Jan-2010 15:46:15] <rmatte> oh right
[19-Jan-2010 15:46:33] <rmatte> so 30 farenheit would show up as 300?
[19-Jan-2010 15:46:37] <rmatte> or how does that work?
[19-Jan-2010 15:47:15] <Kristopher424> Well, it's celsius, so 17C would show up as 170C
[19-Jan-2010 15:47:18] <kobalt> the one I had to do for my DC plants was 100,/,1.8,*,32,+
[19-Jan-2010 15:47:24] <rmatte> ok
[19-Jan-2010 15:47:27] <rmatte> so you'd want to do...
[19-Jan-2010 15:47:45] <rocket> rmatte: quit messing with RPN ... :p
[19-Jan-2010 15:47:48] <rmatte> 10,/,1.8,*,32,+
[19-Jan-2010 15:48:04] <Kristopher424> Right, that's the one kobalt gave me and it worked.
[19-Jan-2010 15:48:05] foppa is now known as forsberg
[19-Jan-2010 15:48:09] <rmatte> rocket: I refuse
[19-Jan-2010 15:48:25] <rmatte> Kristopher424: ok, so what's the issue then?
[19-Jan-2010 15:48:25] <rmatte> lol
[19-Jan-2010 15:48:27] <rocket> rmatte: reminds me of my engineering days with my trusty hp calc .. ;p
[19-Jan-2010 15:48:44] <rmatte> rocket: I never used an hp calc, I specifically learned RPN for Zenoss
[19-Jan-2010 15:48:48] <rmatte> and now I understand it quite well
[19-Jan-2010 15:48:55] <kobalt> rmatte: you were just too slow =)
[19-Jan-2010 15:49:07] <rmatte> kobalt: bleh
[19-Jan-2010 15:49:37] <kobalt> lol I only knew cause I had to figure it out myself for our DC plants
[19-Jan-2010 15:50:01] <Kristopher424> rmatte, I had /,10,1.8,*,32,+ instead.
[19-Jan-2010 15:50:05] <rmatte> kobalt: I helped Kristopher424 with the original RPN formula and relearned RPN in the process, I thought it worked completely differently
[19-Jan-2010 15:50:25] <rmatte> Kristopher424: yeh, it's always the value(s) then the operation
[19-Jan-2010 15:50:28] <Kristopher424> Oh no, ... The latest update of FireFox doesn't load the dashboard properly.
[19-Jan-2010 15:50:44] <rmatte> Kristopher424: clear the cache
[19-Jan-2010 15:51:54] <rmatte> Kristopher424: effectively you were doing Celsius,/,10,1.8,*,32,+ which would translate to Celsius * 10 * 1.8 + 32
[19-Jan-2010 15:52:37] <dec3pti0n> the options to install zenoss is this stack option or source ?
[19-Jan-2010 15:52:42] <dec3pti0n> for debian
[19-Jan-2010 15:52:42] <rmatte> when what you want is Celsius,10,/,1.8,*,32,+ which is Celsius / 10 * 1.8 + 32
[19-Jan-2010 15:52:52] <rmatte> dec3pti0n: stack
[19-Jan-2010 15:53:02] <rmatte> dec3pti0n: there's also the stack installer rolled in to a .deb
[19-Jan-2010 15:53:48] <dec3pti0n> rmatte: the one that is available on the debian repo that is mentioned in the docs, is that the stack instaleer ?
[19-Jan-2010 15:54:08] <rmatte> dec3pti0n: any debian package of Zenoss is the stack installer
[19-Jan-2010 15:54:23] <rmatte> dec3pti0n: they don't have a truly native debian install yet
[19-Jan-2010 15:54:39] <rmatte> the stack installer works great though
[19-Jan-2010 15:55:19] <dec3pti0n> cool, it would also be nice if it allowed one to choose its db backend :) like Pg
[19-Jan-2010 15:55:42] <rmatte> dec3pti0n: it only works with MySQL
[19-Jan-2010 15:56:12] <rmatte> but MySQL is only used for storing the actual events
[19-Jan-2010 15:56:18] <rmatte> everything else uses the Zope database
[19-Jan-2010 15:56:27] <dec3pti0n> yeah I saw that .. that worries me :)
[19-Jan-2010 15:56:28] <forsberg> holy shit, its much more annoying to move from old install to new one than i thought. Because the standard /device templates have changed all my historic data, or atleast alot of it, is useless
[19-Jan-2010 15:56:31] <dec3pti0n> ah
[19-Jan-2010 15:56:34] <rmatte> dec3pti0n: nah, it works fine
[19-Jan-2010 15:57:05] <rmatte> dec3pti0n: it's a relational database and it works quite well for what Zenoss needs
[19-Jan-2010 15:57:58] <rmatte> forsberg: changed it how?
[19-Jan-2010 15:58:19] <rmatte> forsberg: and why didn't you make sure that you had duplicated all of your old templates and stuff prior to doing that?
[19-Jan-2010 15:58:29] <rmatte> forsberg: also, someone posted a fix for your issue today
[19-Jan-2010 15:59:52] <forsberg> oh ?
[19-Jan-2010 16:00:12] <forsberg> well tbh i liked the new templates better, so i initially thought it was a good thing :)
[19-Jan-2010 16:00:40] <rmatte> I'm trying to find the email that I got with the fix
[19-Jan-2010 16:00:44] <rmatte> you just need to modify your zope.conf
[19-Jan-2010 16:01:02] <forsberg> http://dev.zenoss.org/trac/ticket/5798
[19-Jan-2010 16:01:05] <forsberg> oi yes :P~
[19-Jan-2010 16:01:38] <rmatte> yeh, that's it
[19-Jan-2010 16:01:55] <forsberg> thanks maten, gonna try hmmmm tomorrow morning
[19-Jan-2010 16:03:00] <rmatte> np
[19-Jan-2010 16:04:14] <forsberg> i got so keen on a brand new install tho, that hmm :)
[19-Jan-2010 16:04:21] <forsberg> im not sure im gonna tell my boss its fixed hehe
[19-Jan-2010 16:04:52] <forsberg> i know 10x more zenoss now, and i already had hmm put my mind to all the positive things with new install
[19-Jan-2010 16:04:56] <forsberg> get the structure etc right
[19-Jan-2010 19:48:33] <salawank> hi, how do i send zenoss alerts from email to irc? have anyone done that?
[20-Jan-2010 00:00:30] [disconnected at Wed Jan 20 00:00:30 2010]
[20-Jan-2010 00:00:30] [connected at Wed Jan 20 00:00:30 2010]
[20-Jan-2010 03:18:07] nassy_ is now known as nassy
[20-Jan-2010 03:22:36] <chui2002> hi just joine channel
[20-Jan-2010 03:23:00] <forsberg> welcome
[20-Jan-2010 03:23:01] <forsberg> ^^
[20-Jan-2010 03:23:16] <chui2002> haven't even installed Zenoss yet
[20-Jan-2010 03:24:41] <chui2002> Platform: Clear OS;any heads up for this platform?
[20-Jan-2010 07:57:42] <chickenandbeans> Afternoon all
[20-Jan-2010 09:33:34] <ckrough> How do you all build your class tree? Is it strictly by function/role, brand, device type? Ours is currently a mixture.
[20-Jan-2010 10:25:14] <andyman53> Hey all
[20-Jan-2010 10:25:34] <andyman53> besides for its own internal monitoring of services does zenoss have any accommodation for heartbeats?
[20-Jan-2010 10:26:12] <andyman53> I have a VM that I am pinging but I need to send snmp traps from it for services.
[20-Jan-2010 10:26:27] <andyman53> I want a heartbeat log and audit to see when and if my monitoring script stops running.
[20-Jan-2010 10:26:41] <andyman53> Running powershell v1 so I can't monitor what I need remotely.
[20-Jan-2010 10:28:32] <ckrough> can you add the monitoring script as a monitored process for the device?
[20-Jan-2010 10:29:04] <andyman53> i don't think so
[20-Jan-2010 10:29:18] <andyman53> its just a powershell script thats set up to run in task scheduler
[20-Jan-2010 10:41:06] <mistich> goodday any one really good in min/max thresholds
[20-Jan-2010 10:41:45] <mistich> what I want to do it add free_memory and used_memory then set the min value for 50%
[20-Jan-2010 10:42:26] <mistich> so want to check all I do in min/max I select free_memory and used_memory in the datapoints
[20-Jan-2010 10:43:03] <mistich> then in Min Value I just put free_memory + used_memory * .5
[20-Jan-2010 10:43:07] <mistich> is this correct?
[20-Jan-2010 10:44:51] <andyman53> is anyone in here?
[20-Jan-2010 10:45:04] <gwb2352> what are people (if there are any) using to check email _delivery_ i.e. 1. send email 2. receive email in < 60s or alert ?
[20-Jan-2010 10:57:50] <gwb2352> i'm looking at the docs for ZenMailTx, but the POP requirement is troublesome (IMAP is OK)
[20-Jan-2010 10:58:31] <gwb2352> http://www.zenoss.co.uk/mailtx.htm was a good reference
[20-Jan-2010 11:02:11] <dec3pti0n> perhaps I have not gotten to this part in the docs yet but how many servers/devices can one zenoss box monitor ?
[20-Jan-2010 11:02:48] <ckrough> dec3pti0n: it varies, it's really the number of datapoints that impacts it
[20-Jan-2010 11:03:12] <ckrough> dec3pti0n: and since you could be monitoring 1 datapoint on one device and 10000 on the next, it's hard to say a number of target nodes
[20-Jan-2010 11:03:23] <ckrough> dec3pti0n: obviously your hardware has a major impact as well
[20-Jan-2010 11:03:35] <ckrough> dec3pti0n: for most people disk IO is the first limiting factor
[20-Jan-2010 11:03:44] <ckrough> dec3pti0n: zenoss is very disk intensive
[20-Jan-2010 11:05:42] <dec3pti0n> ckrough: thanks, I'm still on chp 2 so I guess I will read about this "datapoint" soon :)
[20-Jan-2010 11:07:30] * dec3pti0n the glosarry helped :)
[20-Jan-2010 11:09:11] <ckrough> dec3pti0n: sorry, datapoint/datasource... something being watched, like temperature or packets inbound, etc... disk spac
[20-Jan-2010 11:13:04] <mrayzenoss> beta 2.5.2 artifacts are now up on http://alpha.zenoss.com/files/2.5.2-beta/
[20-Jan-2010 11:13:16] <mrayzenoss> for QA Test Day tomorrow
[20-Jan-2010 11:16:27] <chickenandbeans> quick question: You know how you can put tales expressions in fields in zenoss, can it take regEx too?
[20-Jan-2010 11:17:11] <chickenandbeans> got some apache monitoring thing to set up one alias is "server-status" the other "status"... awesome huh
[20-Jan-2010 11:17:20] <chickenandbeans> would rather not change the apache config
[20-Jan-2010 11:19:48] <forsberg> dec3pti0n: at around 200+ devices (servers and network units) we started to get problems
[20-Jan-2010 11:25:49] <dec3pti0n> forsberg: really ? that's it
[20-Jan-2010 11:25:55] <ckrough> for me it's 1500
[20-Jan-2010 11:25:59] <ckrough> that number is irrelevant
[20-Jan-2010 11:26:59] <dec3pti0n> I'm still a newbie on zenoss but, it allows you to have a central box for viewing information and then several collectors , correct ?
[20-Jan-2010 11:27:40] <ckrough> yes. the enterprise version does that fairly easily. It's a manual process in CORE (unsupported?)
[20-Jan-2010 11:28:01] <dec3pti0n> you mean the free stack version ?
[20-Jan-2010 11:28:07] <ckrough> correct
[20-Jan-2010 11:28:19] <dec3pti0n> I see
[20-Jan-2010 11:28:38] <ckrough> is your environment distributed and/or very large?
[20-Jan-2010 11:28:58] <forsberg> but yeah its irrelevant
[20-Jan-2010 11:29:12] <dec3pti0n> I guess another option would be to have a zenoss box brought up to collect data from new devices when the time comes
[20-Jan-2010 11:29:35] <forsberg> we just split it up when it began to be to slow (ie more than 5 minutes in collecting all data)
[20-Jan-2010 11:29:50] <dec3pti0n> ckrough: starting around 1000 and it could go up pretty quick
[20-Jan-2010 11:30:05] <forsberg> 1500 devices?
[20-Jan-2010 11:30:13] <ckrough> forsberg: me? yes
[20-Jan-2010 11:30:18] <dec3pti0n> 1000 - 7500
[20-Jan-2010 11:30:52] <forsberg> hmm on a single server?, can you share some specs on that server?
[20-Jan-2010 11:31:12] <ckrough> dec3pti0n: you could probably get a good feel for it with CORE and a single big server, but I would plan for multiple collectors and the Enterprise version if you are monitoring a lot of datapoints
[20-Jan-2010 11:31:25] <forsberg> either im not good enough to remove unimportant stuff from polling, or my server is made of wood ;)
[20-Jan-2010 11:32:08] <ckrough> forsberg: 32g, dual quads, RAID10 with 15k drives, lots and lots of datapoints
[20-Jan-2010 11:32:22] <gwb2352> we're at 551 devices, no problems, no sweat (4 cores, 16GB)
[20-Jan-2010 11:33:16] <forsberg> are you doing tuning or just out of the box core?
[20-Jan-2010 11:34:09] <ckrough> mine is enterprise. definitely tuning of the OS and filesystem, but not much as far as discarding information in zenoss. we pretty much need to see everything it monitors/recieves.
[20-Jan-2010 11:34:41] <ckrough> but the difference between enterprise and CORE is mostly features and convenience, not performance
[20-Jan-2010 11:34:45] <forsberg> ok interesting
[20-Jan-2010 11:34:50] <ckrough> actually the same exact install, plus some zenpacks
[20-Jan-2010 11:34:59] <forsberg> yeah
[20-Jan-2010 11:37:16] <ckrough> and support, delicious delicious support.
[20-Jan-2010 11:37:33] <forsberg> this is quadIntel(R) Xeon(R) CPU E5335 @ 2.00GHz, on raid 10 15k, and only 4gb ram
[20-Jan-2010 11:37:34] <mrayzenoss> lol
[20-Jan-2010 11:37:39] <forsberg> but never seemed like that the ram was the problem
[20-Jan-2010 11:37:58] <forsberg> dunno wtf, there is a long way from 2-300 to 1500 devices
[20-Jan-2010 11:38:37] <ckrough> I think the RAM comes into play just as a compensation for being disk IO bound (caching)
[20-Jan-2010 11:39:02] <ckrough> forsberg: watch top for IOWAIT percentages, and iostat
[20-Jan-2010 11:39:14] <forsberg> yeah but if it never runs empty and starts to swap etc...
[20-Jan-2010 11:39:21] <chickenandbeans> im sure it's been said, but make sure your db isn't on a separate disk :)
[20-Jan-2010 11:39:24] <forsberg> i will, im in the middle of making a brand new install
[20-Jan-2010 11:39:24] <ckrough> if your iowait percentages are higher than say... 20% for most of the time, your disks are choking you
[20-Jan-2010 11:39:44] <forsberg> so i will surely check up on it, and throw some more ram in etc
[20-Jan-2010 11:40:13] <ckrough> sometimes zenhub.py can get to 100% on a single core and slow the whole show down too
[20-Jan-2010 11:40:42] <mistich> +100 on that one :)
[20-Jan-2010 11:47:29] <forsberg> but enterprise version for 1500+ devices... phew :P
[20-Jan-2010 11:47:57] <forsberg> if you can afford that, i guess you can afford a nicer server to install it on also :)
[20-Jan-2010 11:56:03] <rmatte> forsberg: you'll have a hard time even doing 300 devices with that hardware
[20-Jan-2010 11:57:48] <forsberg> yap, thats what im saying ;)
[20-Jan-2010 11:58:00] <mistich> rmatte: all depends on the number of datapoints
[20-Jan-2010 11:58:23] <mistich> and how much data you retain
[20-Jan-2010 11:58:39] <forsberg> ofcourse, it does, but i do think dec3pti0n was looking for an excact number either, but an estimate, 20 / 300 / 1500 servers
[20-Jan-2010 11:58:45] <forsberg> and now he got his answer
[20-Jan-2010 11:58:45] <forsberg> ;)
[20-Jan-2010 11:58:53] <forsberg> do NOT think
[20-Jan-2010 11:59:06] <mistich> yeah I wish there was a magic number
[20-Jan-2010 11:59:25] <forsberg> no its cool, after all it snmpwalks itself ;)
[20-Jan-2010 11:59:46] <mistich> tried to do a calculator once too many variables
[20-Jan-2010 12:00:12] <forsberg> 2-300 devices on a crappy cheap server :P 1500 on ninja hardware
[20-Jan-2010 12:00:53] <forsberg> ive got a brilliant argument for my boss, to get me some shiny new iron ^^
[20-Jan-2010 12:00:54] <mistich> thats about right
[20-Jan-2010 12:02:51] <mistich> want to push it I think we got 4500 with ssd drives
[20-Jan-2010 12:03:10] <mistich> that 2 ssd card
[20-Jan-2010 12:03:19] <forsberg> :P~
[20-Jan-2010 12:04:09] <mistich> might be more we did it in or test environment going to move it into production soon
[20-Jan-2010 12:04:30] <mistich> ckrough has more detail if your intrested
[20-Jan-2010 12:05:30] <forsberg> well not really, the company i work for wont get so much anytime too soon
[20-Jan-2010 12:06:37] <mistich> yeah it took me 2 years so know where your at
[20-Jan-2010 12:09:56] <rmatte> mistich: I was assuming that he'd be doing full monitoring on most of those 300 devices
[20-Jan-2010 12:10:04] <rmatte> mistich: obviously ping only monitoring doesn't count lol
[20-Jan-2010 12:10:34] <forsberg> ^^
[20-Jan-2010 12:10:56] <forsberg> its full monitoring yes, 200ish+ /device/server/linux + some custom shit
[20-Jan-2010 12:11:13] <forsberg> and 20 windwows servers and some switches/routers/firewalls
[20-Jan-2010 12:11:17] <rmatte> linux isn't that bad to monitor compared to windows
[20-Jan-2010 12:11:30] <rmatte> since linux actually has snmp done right
[20-Jan-2010 12:11:37] <rmatte> the windows snmp agent is a frigging nightmare
[20-Jan-2010 12:11:40] <forsberg> yes
[20-Jan-2010 12:11:55] <rmatte> I'm going to be taking the time to really amp up my windows performance zenpacks
[20-Jan-2010 12:12:03] <rmatte> going to script an actual collector for them
[20-Jan-2010 12:12:11] <forsberg> our new exchange server, wouldnt answer to snmp, so i removed the service, and it wanted to reboot
[20-Jan-2010 12:12:14] <forsberg> before i could install again
[20-Jan-2010 12:12:15] <forsberg> haha
[20-Jan-2010 12:12:23] <rmatte> wouldn't answer to snmp?
[20-Jan-2010 12:12:25] <rmatte> that's weird
[20-Jan-2010 12:12:30] <forsberg> well it wasnt firewall rules
[20-Jan-2010 12:12:35] <forsberg> and nothing happened when i walked it
[20-Jan-2010 12:12:37] <rmatte> In my experience snmp is more reliable than WMI
[20-Jan-2010 12:12:47] <rmatte> you might have had it configured incorrectly
[20-Jan-2010 12:12:53] <mistich> just change it out for a linux server problem solved :)
[20-Jan-2010 12:12:56] <rmatte> maybe it wasn't set to accept SNMP from that particular host?
[20-Jan-2010 12:12:59] <forsberg> nono, add an ip and a community
[20-Jan-2010 12:13:00] <rmatte> haha
[20-Jan-2010 12:13:07] <forsberg> hehe well wish i could :P
[20-Jan-2010 12:13:10] <rmatte> I love that mentality
[20-Jan-2010 12:13:26] <rmatte> "Eugh, windows is acting up again, screw it, get a linux server in here stat!"
[20-Jan-2010 12:13:35] <forsberg> :P
[20-Jan-2010 12:13:39] <rmatte> lol
[20-Jan-2010 12:13:47] <rmatte> I'm a big Linux guy, so I can appreciate that
[20-Jan-2010 12:13:57] <mistich> have to keep myself employed some how
[20-Jan-2010 12:14:06] <rmatte> (Though I've done my time with Windows too unfortunately, enough to know it inside and out)
[20-Jan-2010 12:14:15] <forsberg> would really love to, but I think that train is gone.....with the exchange/office thing
[20-Jan-2010 12:14:37] <rmatte> forsberg: well, exchange can be emulated, and openoffice can replace office
[20-Jan-2010 12:14:38] <rmatte> ;)
[20-Jan-2010 12:14:40] <mistich> more linux people use. more my skills go up in value
[20-Jan-2010 12:14:44] <forsberg> + i dont have to deal with it at all
[20-Jan-2010 12:14:57] <forsberg> we have a double team of MICROSOFT CERTIFIED PERSONEL
[20-Jan-2010 12:15:05] <mistich> just giving you a hard time
[20-Jan-2010 12:15:08] <rmatte> mistich: honestly, I like having Linux as a specialty thing
[20-Jan-2010 12:15:18] <rmatte> mistich: keeps the skills much more valuable than microsoft skills
[20-Jan-2010 12:15:33] <rmatte> I like being "The Linux guy"
[20-Jan-2010 12:15:34] <forsberg> funny thing is, the replaced all computers on the office with imacs
[20-Jan-2010 12:15:37] <mistich> no didn't waste my time on windows is all gui wouldn't take me long to pick it up
[20-Jan-2010 12:15:43] <rmatte> though around here we have like 5 other "Linux guys" but oh well
[20-Jan-2010 12:15:44] <forsberg> so all technicies etc dont even run outlook anymore
[20-Jan-2010 12:15:49] <forsberg> technicians
[20-Jan-2010 12:15:56] <rmatte> mistich: yeh, windows takes no time at all to learn
[20-Jan-2010 12:16:22] <mistich> like most of us we started with windows then found the light
[20-Jan-2010 12:16:24] <rmatte> forsberg: well, outlook will run on a Mac, provided you have the apple version of office
[20-Jan-2010 12:16:34] <forsberg> yeah, i know but we dont use that
[20-Jan-2010 12:16:58] <forsberg> only sales department, and direction
[20-Jan-2010 12:16:58] <forsberg> ;)
[20-Jan-2010 12:17:03] <mistich> yeah be nice if they came out with a linux version of Microsoft office office
[20-Jan-2010 12:17:52] <mistich> viso is the best app they have have yet to find a replacement
[20-Jan-2010 12:17:59] <forsberg> :P
[20-Jan-2010 12:18:59] <rmatte> It would be nice if they came out with an adobe master suite for Linux too, but the chances of that are very slim.
[20-Jan-2010 12:19:04] <forsberg> hehe
[20-Jan-2010 12:19:17] <rmatte> I'm sick of running photoshop in wine and virtualbox
[20-Jan-2010 12:19:19] <forsberg> i dont use all that shit.....ignorance is bliss
[20-Jan-2010 12:19:25] <rmatte> hehe
[20-Jan-2010 12:19:35] <rmatte> I used to do web design, so I love that stuff
[20-Jan-2010 12:19:42] <forsberg> use my roundcube mail and some text editors
[20-Jan-2010 12:19:55] <rmatte> I can make stuff like: http://dmon.org/graphics/nexusbg.jpg
[20-Jan-2010 12:20:03] <rmatte> I use roundcube mail on my home server
[20-Jan-2010 12:20:07] <rmatte> best webmail I've seen
[20-Jan-2010 12:20:11] <rmatte> I used horde before that
[20-Jan-2010 12:20:11] <forsberg> jupski
[20-Jan-2010 12:20:18] <forsberg> squirrel
[20-Jan-2010 12:20:19] <forsberg> ;)
[20-Jan-2010 12:20:25] <rmatte> I've never seen squirrel
[20-Jan-2010 12:20:34] <rmatte> I used pine for a while ;)
[20-Jan-2010 12:20:39] <rmatte> that was true hardcore email
[20-Jan-2010 12:21:45] <forsberg> gonna put on a movie so wife falls asleep, so i can get on with moving more devices
[20-Jan-2010 12:21:49] <forsberg> later :P
[20-Jan-2010 12:22:22] <rmatte> lol
[20-Jan-2010 12:22:25] <rmatte> later
[20-Jan-2010 12:22:33] <rmatte> (enjoy the show)
[20-Jan-2010 12:22:34] <rmatte> lol
[20-Jan-2010 12:37:20] <mistich> horde that brings back memories
[20-Jan-2010 12:37:23] <rmatte> hehe
[20-Jan-2010 12:37:53] <ckrough> when we retired pine at my last job a bunch of the old faculty printed "Save the trees" t-shirts in protest
[20-Jan-2010 12:37:56] <rmatte> horde was great until I saw roundcube
[20-Jan-2010 12:38:04] <rmatte> haha
[20-Jan-2010 12:38:08] <rmatte> that's awesome
[20-Jan-2010 12:38:21] <rmatte> I honestly miss pine
[20-Jan-2010 12:38:56] <rmatte> It made sorting through email so quick
[20-Jan-2010 12:40:07] <mistich> pine was wounderful
[20-Jan-2010 12:41:06] <rmatte> it can't do html formatting though (for obvious reasons), it's great for viewing email on secure mailservers though
[20-Jan-2010 12:41:17] <rmatte> just make sure to run pine on the server itself
[20-Jan-2010 12:43:41] * rmatte waits half an hour for an snmpwalk to finish
[20-Jan-2010 12:44:29] <rmatte> wow, squirrel looks overly simple
[20-Jan-2010 12:46:29] <dec3pti0n> does zenoss provide built-in templates at all ? like if I'm adding linux servers that would be part of a template that already has all services needed to be monitored ?
[20-Jan-2010 12:47:06] <rmatte> dec3pti0n: /Server/Linux
[20-Jan-2010 12:47:24] <ckrough> it picks up on most of the common services
[20-Jan-2010 12:47:24] <rmatte> provided you have snmp properly enabled on them and you have the snmp string added in to Zenoss that'll give you all you want
[20-Jan-2010 12:47:31] <ckrough> it's really easy to add others though
[20-Jan-2010 12:47:40] <rmatte> you're going to want to disable a couple of the thresholds in the Linux device template
[20-Jan-2010 12:47:54] <rmatte> there's an annoying low swap threshold which is set at 1024 (it's not dynamic)
[20-Jan-2010 12:47:58] <rmatte> I usually just delete that
[20-Jan-2010 12:48:00] <dec3pti0n> yeah I used that while adding the device but didn't pick up any running services, processes runnign on the system to be monitored
[20-Jan-2010 12:48:09] <rmatte> there's also a low idle threshold which I just disable
[20-Jan-2010 12:48:31] <rmatte> dec3pti0n: processes that you actually want monitored need to be added in to the processes section (click processes on the left)
[20-Jan-2010 12:48:49] <rmatte> since you obviously don't want it monitoring all processes
[20-Jan-2010 12:49:08] <rmatte> once they are added there just remodel the device (Manage -> Model Device from the dropdown menu)
[20-Jan-2010 12:49:08] <dec3pti0n> cool
[20-Jan-2010 13:11:24] <rmatte> also, if there are processes where you want Zenoss to ignore the actual flags when monitoring them (because it watches them by default) set ignore parameters to false for that particular process
[20-Jan-2010 13:11:48] <rmatte> for example you may have some program running which looks like /bin/program --something --blah --blah
[20-Jan-2010 13:12:05] <rmatte> but the next time it runs it looks like /bin/program --somethingelse --blah --blah
[20-Jan-2010 13:12:38] <rmatte> settings ignore parameters to true (I said false before, but I meant to say true) causes Zenoss to only watch for /bin/program
[20-Jan-2010 13:12:44] <rmatte> and disregard the options/flags
[20-Jan-2010 13:23:37] cwj is now known as FUDGE
[20-Jan-2010 13:24:09] FUDGE is now known as cwj
[20-Jan-2010 14:05:00] <rmatte> wow, it's pretty dead in here today lol
[20-Jan-2010 14:06:24] <venturaville> very quiet
[20-Jan-2010 14:06:45] <venturaville> incidentally, has anyone got a good example of a zenpack that has a library included in it...?
[20-Jan-2010 14:06:48] <rmatte> venturaville: hey, haven't seen you in a while
[20-Jan-2010 14:07:02] <rmatte> by "library" you mean?
[20-Jan-2010 14:07:20] <venturaville> in this case libvirt-python
[20-Jan-2010 14:07:29] <rmatte> oh
[20-Jan-2010 14:07:37] <rmatte> is that available via easy_install?
[20-Jan-2010 14:07:40] <venturaville> which has a straight up python and a companion so file
[20-Jan-2010 14:07:47] <venturaville> easy_install?
[20-Jan-2010 14:08:24] <rmatte> the problem with including the library is that the pack would be very system-specific
[20-Jan-2010 14:08:33] <venturaville> I'm trying to get around that
[20-Jan-2010 14:08:35] <rmatte> since it would either be 32 or 64bit
[20-Jan-2010 14:08:58] <rmatte> http://wiki.vpslink.com/Python_easy_install
[20-Jan-2010 14:09:19] <rmatte> If that package is available via that, you could just have the __init__.py file kick off the install of it automatically
[20-Jan-2010 14:09:30] <venturaville> that is exactly what I am looking for
[20-Jan-2010 14:09:44] <rmatte> see if it's available via easy_install then
[20-Jan-2010 14:10:24] <mrayzenoss> easy_install isn't on every platform: http://dev.zenoss.org/trac/ticket/5657
[20-Jan-2010 14:11:57] <rmatte> oh
[20-Jan-2010 14:11:59] <rmatte> didn't realize that
[20-Jan-2010 14:12:21] <patzer> hello
[20-Jan-2010 14:12:35] <rmatte> venturaville: honestly, just better to list it as a dependency and let people install it themselves
[20-Jan-2010 14:12:41] <rmatte> patzer: hello
[20-Jan-2010 14:12:50] <venturaville> I really need it to be tied to the zenoss python version though
[20-Jan-2010 14:13:06] <rmatte> well, that's going to change from 2.4 to 2.6 soon
[20-Jan-2010 14:13:14] <rmatte> so bad timing to do that
[20-Jan-2010 14:13:15] <rmatte> lol
[20-Jan-2010 14:13:17] <venturaville> yep, but if the zenpack does the setup......
[20-Jan-2010 14:13:45] <rmatte> basically it's just a matter of figuring out how to install it perfectly every time and then having the __init__.py do that
[20-Jan-2010 14:13:58] <venturaville> in theory I could convert my pythonplugin over to commandplugin and only use the check_libvirt script, but that has a performance penalty.....
[20-Jan-2010 14:14:19] <venturaville> any hints on how to get easy_install to use zenoss python?
[20-Jan-2010 14:14:29] <rmatte> venturaville: actually, technically what you're trying to do is impossible
[20-Jan-2010 14:14:38] <rmatte> ZenPacks install with zenoss user privileges
[20-Jan-2010 14:14:44] <rmatte> installing a library takes root privs
[20-Jan-2010 14:14:55] <venturaville> not if you are doing it into zenpack/lib ......
[20-Jan-2010 14:14:56] <rmatte> unless you're dropping the library in to the zenoss directory structure
[20-Jan-2010 14:15:27] <rmatte> well, if you're looking to do that you'll need to include a 32bit and 64bit version of the library plus 32 and 64bit versions for FreeBSD as well
[20-Jan-2010 14:15:33] <venturaville> right now the way I have it working is to symlink the libvirtmod.so to the system one (and make sure the versions match)
[20-Jan-2010 14:15:37] <rmatte> then have __init__.py do the copying
[20-Jan-2010 14:17:43] <venturaville> ah ... I just realized easy_install is in the zenoss stack...
[20-Jan-2010 14:17:44] <venturaville> cool
[20-Jan-2010 14:17:58] <rmatte> it is, but apparently it's not on every platform
[20-Jan-2010 14:18:08] <rmatte> so it's not exactly an ideal solution
[20-Jan-2010 14:18:11] <venturaville> in this case, I mainly only care about linux platforms...
[20-Jan-2010 14:18:18] <rmatte> I see
[20-Jan-2010 14:18:38] <rmatte> well, then you need to include a compiled 32 and 64bit version
[20-Jan-2010 14:18:42] <venturaville> nope
[20-Jan-2010 14:18:46] <rmatte> then have the script figure out what the system is and copy it
[20-Jan-2010 14:18:50] <venturaville> just need to include the python-libvirt source
[20-Jan-2010 14:19:00] <venturaville> library is LGPL v2 though..
[20-Jan-2010 14:19:04] <venturaville> not sure of the implications of that
[20-Jan-2010 14:19:16] <rmatte> LGPL is pretty rabid
[20-Jan-2010 14:19:36] <rmatte> As long as the pack is open source, which it will be, I doubt there'd be an issue
[20-Jan-2010 14:19:59] <rmatte> the whole point of GPL and LGPL is to ensure that code doesn't get ripped off and user in commercial applications which are sold for profit
[20-Jan-2010 14:20:02] <venturaville> it will be
[20-Jan-2010 14:20:16] <rmatte> s/user/used
[20-Jan-2010 14:20:49] <rmatte> but you're releasing it as a community zenpack for an open source project, so I doubt there'll be a problem
[20-Jan-2010 14:26:15] <rmatte> mrayzenoss: I don't see any improvement in the event console start/end time thing in 2.5.2
[20-Jan-2010 14:26:41] <rmatte> actually nevermind
[20-Jan-2010 14:26:48] <rmatte> just realized it is actually working now
[20-Jan-2010 14:27:09] <venturaville> libvirt isn't on the easy_install list as of yet
[20-Jan-2010 14:27:24] <rmatte> ah
[20-Jan-2010 14:27:56] <rmatte> well, you're going to have to actually include it with the pack then and write some code which figures our whether the system is 32 or 64bit and copies the proper library file over
[20-Jan-2010 14:29:16] <rmatte> hmmm, never noticed that before
[20-Jan-2010 14:29:33] <rmatte> there's an uptime for zope if you go to Settings -> Versions
[20-Jan-2010 14:29:35] <rmatte> good to know
[20-Jan-2010 14:47:56] <patzer> I am having some trouble refactoring my zenpack
[20-Jan-2010 14:48:09] <rmatte> "refactoring"?
[20-Jan-2010 14:48:17] <patzer> I have all the monitoring and such working well, but I made my own device type
[20-Jan-2010 14:48:33] <patzer> I just added some relations to the global OSComponent
[20-Jan-2010 14:48:51] <patzer> so now I am trying to reimplement using a new device type
[20-Jan-2010 14:49:12] <patzer> only I seem to be missing something, when I add new devices they are still of kind Device, not MyDevice
[20-Jan-2010 14:49:28] <rmatte> hmmm
[20-Jan-2010 14:49:31] <rmatte> not sure about that one
[20-Jan-2010 14:49:32] <patzer> and the MyDevice.py file is never compiled to a pyc, but I can't find anything in the logs to indicate a problem
[20-Jan-2010 14:50:09] <rmatte> how is Zenoss aware of MyDevice.py's existence?
[20-Jan-2010 14:50:31] <patzer> it's in a zenpack and the __init__.py for the zenpack IS compiled and executed when I restart
[20-Jan-2010 14:52:35] <rmatte> hmmm
[20-Jan-2010 14:52:57] <cgibbons> does your new device class set the zPythonClass property?
[20-Jan-2010 14:53:05] <patzer> no
[20-Jan-2010 14:53:10] <cgibbons> that's why
[20-Jan-2010 14:53:26] <rmatte> cgibbons to the rescue
[20-Jan-2010 14:53:37] <cgibbons> once in a while i know the answer, but not very often :)
[20-Jan-2010 14:53:39] <patzer> in that case I believe that step is missing from the examples
[20-Jan-2010 14:53:42] <rmatte> hehe
[20-Jan-2010 14:54:01] <cgibbons> quite likely that it is... in any event, that zProperty, when set, tells Zenoss to use the specificed python class instead of ZenModel/Device.py
[20-Jan-2010 14:55:04] <patzer> where do you set that, in the __init__.py?
[20-Jan-2010 14:55:27] <patzer> or in the gui for the organizer class
[20-Jan-2010 14:55:34] <cgibbons> whichever you prefer
[20-Jan-2010 14:55:59] <cgibbons> if you're not explicitly creating the device class in your __init__.py, then it is probably more consistent to set it in UI and have it saved off to your objects.xml
[20-Jan-2010 14:56:30] <cgibbons> if you want really rich examples of custom device classes, be sure to reference the PrinterToner and AIX community zenpacks
[20-Jan-2010 14:56:37] <cgibbons> both do custom subcomponents, etc
[20-Jan-2010 14:57:27] <patzer> how would you explicitly create it in the __init__.py? those two zenpacks would provide examples of that as well?
[20-Jan-2010 14:59:45] <cgibbons> luckily easily: deviceClass = dmd.Devices.createOrganizer("/Foobar")
[20-Jan-2010 15:00:00] <cgibbons> deviceClass.zPythonClass = "ZenPacks.community.Foobar.FoobarDevice"
[20-Jan-2010 15:00:07] <patzer> ah ok, I see
[20-Jan-2010 15:01:22] <patzer> ok, great this works much better :)
[20-Jan-2010 15:01:26] <patzer> thanks
[20-Jan-2010 15:04:32] <QubeZ> hello all
[20-Jan-2010 15:05:16] <QubeZ> how can I monitor an NFS mount on a server and have it alert me when that mount no longer exists? also, is there a way to remount it when it sees that the mount no longer exists?
[20-Jan-2010 15:05:31] <QubeZ> i enabled the RPC server template but i dont see where to access the stats
[20-Jan-2010 15:06:06] <rmatte> QubeZ: for that (and especially for the automatic remouting) you'd have to make a custom script
[20-Jan-2010 15:06:24] <QubeZ> so even for monitoring nfs i need a custom script?
[20-Jan-2010 15:06:46] <rmatte> QubeZ: well, no, an nfs share is seen just like any other filesystem would be
[20-Jan-2010 15:06:51] <rmatte> but for the automatic remount you would
[20-Jan-2010 15:07:08] <QubeZ> rmatte so i just need to remove the "networkDisk" from ignored file types?
[20-Jan-2010 15:07:43] <rmatte> I didn't even have to do that, we have an nfs share that's mounted on basically every one of our servers and Zenoss picked it up fine without me changing anything
[20-Jan-2010 15:07:57] <QubeZ> hmm mine did not, it only shows the ramdisk no nfs
[20-Jan-2010 15:13:08] <rmatte> that's odd
[20-Jan-2010 15:13:18] <rmatte> it shouldn't even show the ramdisk
[20-Jan-2010 15:13:34] <QubeZ> ya, but my ramdisk is showing up for some reason but the nfs isn't
[20-Jan-2010 15:14:27] <rmatte> My zFileSystemMapIgnoreTypes is:
[20-Jan-2010 15:14:28] <rmatte> other
[20-Jan-2010 15:14:28] <rmatte> ram
[20-Jan-2010 15:14:28] <rmatte> virtualMemory
[20-Jan-2010 15:14:28] <rmatte> removableDisk
[20-Jan-2010 15:14:28] <rmatte> floppyDisk
[20-Jan-2010 15:14:30] <rmatte> compactDisk
[20-Jan-2010 15:14:32] <rmatte> ramDisk
[20-Jan-2010 15:14:34] <rmatte> flashMemory
[20-Jan-2010 15:14:42] <rmatte> just remove networkDisk from there if it's there
[20-Jan-2010 15:14:52] <rmatte> and put ramDisk in (since you don't want to be monitoring it like that)
[20-Jan-2010 15:15:06] <QubeZ> i do want to monitor ramdisk as well
[20-Jan-2010 15:15:13] <rmatte> ok, well then leave it out
[20-Jan-2010 15:15:15] <rmatte> up to you
[20-Jan-2010 15:15:42] <rmatte> swap space is already monitored on the perf tab
[20-Jan-2010 15:15:55] <QubeZ> i have a custom ramdisk created of 1G for my asterisk recordings
[20-Jan-2010 15:16:03] <rmatte> oh, I see
[20-Jan-2010 15:16:06] <QubeZ> :)
[20-Jan-2010 15:16:08] <rmatte> makes sense
[20-Jan-2010 15:18:23] <QubeZ> weird, now it showed up after removing ramdisk and networkdisk and remodling
[20-Jan-2010 15:18:32] <forsberg> :)
[20-Jan-2010 15:18:39] <forsberg> how is that wierd?
[20-Jan-2010 15:18:45] <forsberg> wasnt that the intention?
[20-Jan-2010 15:19:06] <QubeZ> as far as networkdisk is yes but ramdisk was in my list of ignoredfiletypes but it was always available.
[20-Jan-2010 15:19:19] <QubeZ> so i guess my "weird" comment should have been directed at the ramdisk
[20-Jan-2010 15:19:35] <rmatte> you sure that's the type that it's displayed as though?
[20-Jan-2010 15:19:41] <forsberg> ahh sorry, but since its not there and you modelled, it should be there
[20-Jan-2010 15:20:01] <rmatte> by ramdisk, you're not referring to swap (which is what ramdisk type refers to)
[20-Jan-2010 15:20:08] <rmatte> you're referring to a solid state drive, right?
[20-Jan-2010 15:21:51] <QubeZ> no, filesystem ramdisk
[20-Jan-2010 15:22:18] <rmatte> If you go to the OS tab and click on it...
[20-Jan-2010 15:23:18] <rmatte> does it actually show ramdisk as type?
[20-Jan-2010 15:25:44] <QubeZ> i dont see any type heading... i just have mount total bytes used bytes ,free bytes, %util
[20-Jan-2010 15:28:38] <rmatte> leave it to microsoft to come up with a drive naming convention with a limit of 26
[20-Jan-2010 15:30:17] <rmatte> QubeZ: since there's no type heading, that's why it picks it up
[20-Jan-2010 15:30:27] <rmatte> because it's not explicitly defined as a ramdisk via snmp
[20-Jan-2010 15:36:38] <eidolon> hi folks - i have a CMD monitor in one of my hosts - but it stopped reporting about a week ago. See here: http://skitch.com/dbelfershevett/ntmcb/zenoss-httpmonitor
[20-Jan-2010 15:37:04] <eidolon> the problem is that i'm not sure how to debug this. I thought this would log into ZenCommand - so i turned th debug logging up on that, restarted it, and tail -f 'ed it, but i never see those commands run.
[20-Jan-2010 15:37:28] <mrayzenoss> zentestcommand?
[20-Jan-2010 15:37:41] <eidolon> that works.
[20-Jan-2010 15:37:42] <rmatte> as the zenoss user, do: zencommand run -v10 --cycle
[20-Jan-2010 15:38:45] <rmatte> just ctrl-c out of it when you're done with it
[20-Jan-2010 15:39:14] <eidolon> [zenoss@lwmzen03 bin]$ ./zentestcommand -d BellVHosts --datasource="CMDSPurchases"
[20-Jan-2010 15:39:14] <eidolon> |Purchases=38
[20-Jan-2010 15:40:19] <rmatte> I assume that your actual data point is called "Purchases"
[20-Jan-2010 15:40:20] <rmatte> ?
[20-Jan-2010 15:41:08] <patzer> hmm, one final question!
[20-Jan-2010 15:41:17] <eidolon> yep, see the screenshot?
[20-Jan-2010 15:41:20] <patzer> it seems my skins directory isn't being registered right
[20-Jan-2010 15:41:31] <rmatte> eidolon: I don't see your datapoint in the screenshot
[20-Jan-2010 15:41:35] <rmatte> just your datasource
[20-Jan-2010 15:41:39] <rmatte> they are 2 different things
[20-Jan-2010 15:41:57] <rmatte> the datapoint would be listed under the actual command
[20-Jan-2010 15:42:04] <patzer> /usr/local/zenoss/zenoss/Products/CMFCore/DirectoryView.py:431: UserWarning: DirectoryView ZenPacks.MyOrg.MyPack refers to a non-existing path ...
[20-Jan-2010 15:42:09] <patzer> but the path ... does exist
[20-Jan-2010 15:42:16] <eidolon> yes 'Purchases'
[20-Jan-2010 15:42:31] <rmatte> k
[20-Jan-2010 15:42:53] <rmatte> patzer: Is that after installing the pack or while installing the pack?
[20-Jan-2010 15:42:53] <eidolon> the thing is this -was- working
[20-Jan-2010 15:42:57] <eidolon> but stopped.
[20-Jan-2010 15:43:04] <eidolon> "what changed" "nothing! i swear!"
[20-Jan-2010 15:43:04] <patzer> rmatte: when I run zendmd it is a warning
[20-Jan-2010 15:43:09] <rmatte> eidolon: have you tried just restarting zenoss entirely?
[20-Jan-2010 15:43:17] <eidolon> yep i have.
[20-Jan-2010 15:43:18] <eidolon> no dice.
[20-Jan-2010 15:43:22] <rmatte> eidolon: also, perhaps the command is taking longer than usual?
[20-Jan-2010 15:43:29] <rmatte> perhaps you should increase the timeout for the command
[20-Jan-2010 15:43:46] <rmatte> you don't see the commands run at all when you do: zencommand run -v10 --cycle?
[20-Jan-2010 15:43:53] <eidolon> i haven't tried that, lemme do that.
[20-Jan-2010 15:43:58] <patzer> and I get a "requested resource" does not exist when trying to go to the page in the gui
[20-Jan-2010 15:44:20] <rmatte> patzer: did you install that pack without restarting Zenoss?
[20-Jan-2010 15:44:37] <eidolon> AIEEE
[20-Jan-2010 15:44:39] * eidolon is flooded!
[20-Jan-2010 15:44:40] <rmatte> patzer: if so, restart Zenoss and that should go away
[20-Jan-2010 15:44:48] <rmatte> eidolon: by?
[20-Jan-2010 15:44:53] <eidolon> i got MASSES of output.
[20-Jan-2010 15:44:59] <eidolon> i have zencommand in debug mode
[20-Jan-2010 15:45:04] <rmatte> ah
[20-Jan-2010 15:45:25] <rmatte> I don't generally actually put daemons in to debug mode, I just do <daemon> run -v10 --cycle
[20-Jan-2010 15:45:28] <rmatte> but to each his own
[20-Jan-2010 15:45:30] <rmatte> lol
[20-Jan-2010 15:45:36] <eidolon> whaddapal :)
[20-Jan-2010 15:45:51] <rmatte> lol
[20-Jan-2010 15:46:02] <rmatte> eidolon: so is it working again then or still not working?
[20-Jan-2010 15:46:12] <patzer> this is the new zenpack I am developing though
[20-Jan-2010 15:46:24] <patzer> I shouldn't have to reinstall it while developing it?
[20-Jan-2010 15:46:28] <eidolon> rrd charts are still showing 'nan'
[20-Jan-2010 15:46:32] <rmatte> patzer: right, when more complex zenpacks are installed Zenoss needs to be restarted after they are installed
[20-Jan-2010 15:46:35] <rmatte> otherwise you get odd behaviour
[20-Jan-2010 15:47:02] <rmatte> eidolon: they'd show nan for 3 polling cycles
[20-Jan-2010 15:47:14] <patzer> I have restarted several times already though
[20-Jan-2010 15:47:18] <eidolon> so here's a question
[20-Jan-2010 15:47:22] <eidolon> if that command ran, and timed out.
[20-Jan-2010 15:47:33] <rmatte> eidolon: another thing you could try is deleting the RRD files. I've had RRD files go sour on me before. It's very rare but it does happen.
[20-Jan-2010 15:47:37] <eidolon> i'm occasionally getting timeouts. so i'm wondering... does it need 3 successes in a row to chart?
[20-Jan-2010 15:47:43] <rmatte> yes
[20-Jan-2010 15:47:46] <rmatte> it needs 3 in a row
[20-Jan-2010 15:47:49] <eidolon> ahhhh. that may be a problem.
[20-Jan-2010 15:47:56] <patzer> you'd see the last value on say the graph though
[20-Jan-2010 15:48:01] <rmatte> Ideally you should have no timeouts at all
[20-Jan-2010 15:48:43] <eidolon> hmmmm
[20-Jan-2010 15:48:57] <rmatte> also, increase zCommandCycleTime in zProperties
[20-Jan-2010 15:49:01] <rmatte> make it 60 instead of 15
[20-Jan-2010 15:49:07] <eidolon> i'm getting this a lot
[20-Jan-2010 15:49:07] <rmatte> that'll probably eliminate your timeouts
[20-Jan-2010 15:49:09] <eidolon> [zenoss@lwmzen03 bin]$ ./zentestcommand -d BellVHosts --datasource="CMDSPurchases"
[20-Jan-2010 15:49:09] <eidolon> Command timed out
[20-Jan-2010 15:49:40] <rmatte> by default it's set at 15 or something
[20-Jan-2010 15:49:50] <patzer> that's probably your problem then :)
[20-Jan-2010 15:50:18] <rmatte> zCollectorClientTimeout could be increased as well
[20-Jan-2010 15:50:24] * eidolon zpropertyedits
[20-Jan-2010 15:50:37] <eidolon> zCommandCommandTimeout
[20-Jan-2010 15:50:38] <eidolon> that one?
[20-Jan-2010 15:51:08] <rmatte> yeh
[20-Jan-2010 15:51:09] <eidolon> currently 15
[20-Jan-2010 15:51:11] <rmatte> I'd set that to 60
[20-Jan-2010 15:51:22] <eidolon> done
[20-Jan-2010 15:51:25] <rmatte> (that's what I have mine set to since I use a lot of command based datasources)
[20-Jan-2010 15:51:25] * eidolon reruns testcommand
[20-Jan-2010 15:52:22] <eidolon> okay, seems better. i'll need to wait a few 5 minute cycels though
[20-Jan-2010 15:52:45] <rmatte> yup
[20-Jan-2010 15:54:46] <rmatte> patzer: does the zenoss user have access to the path of the pack?
[20-Jan-2010 15:55:00] <rmatte> (read/write)
[20-Jan-2010 15:57:21] <patzer> yes, it is in the default ZenPack subdirectory
[20-Jan-2010 16:00:57] <rmatte> patzer: if you have any .zec files in $ZENHOME/var try stopping Zenoss, deleting them, then starting Zenoss again
[20-Jan-2010 16:01:03] <rmatte> that may resolve your issue
[20-Jan-2010 16:02:22] <eidolon> okay, we'll see if that fixes it. i need to pack up and move to a different spot
[20-Jan-2010 16:02:25] <eidolon> rmatte, thank you as always :)
[20-Jan-2010 16:02:29] <rmatte> np
[20-Jan-2010 16:03:25] <patzer> ok, I'll try that
[20-Jan-2010 16:05:12] <patzer> I don't think I have any zec files
[20-Jan-2010 16:05:25] <patzer> but there is a symlink difference
[20-Jan-2010 16:05:55] <patzer> I refer to /usr/local/zenoss/ but that is a symlink to the real location at /local/zenoss-stack/
[20-Jan-2010 16:10:23] <rmatte> really?
[20-Jan-2010 16:10:26] <rmatte> I never noticed that
[20-Jan-2010 16:10:56] <rmatte> mine isn't a symlink
[20-Jan-2010 16:11:02] <rmatte> it's literally in /usr/local/zenoss
[20-Jan-2010 16:11:17] <patzer> it's just how we setup due to disk space likely
[20-Jan-2010 16:15:04] <rmatte> probably
[20-Jan-2010 16:17:21] <TBKDan1> Is there any way to export the entire zenoss configuration including devices, services monitored on each device, etc?
[20-Jan-2010 16:17:41] <patzer> rmatte: ok, I did a zenpack --install on it and that seems to have fixed it as suggested in some old forum post from 2007 :|
[20-Jan-2010 16:19:28] <patzer> thanks, I'm out of here for today
[20-Jan-2010 16:26:09] <rmatte> TBKDan1: read up on zenbackup/zenrestore
[20-Jan-2010 16:26:31] <rmatte> patzer: cool, later
[20-Jan-2010 16:27:48] <TBKDan1> rmatte: But that grabs the data.fs files, doesn't it? My data.fs is incredibly funky so I just want to start from scratch, importing the plain config
[20-Jan-2010 16:32:41] <dec3pti0n> ok I'm kind of confused here, I just added a device to zenoss and I have configured snmpd on the device
[20-Jan-2010 16:32:51] <QubeZ> rmatte so having the networkdisk enable, i notice it only monitors disk space to make sure it doesn't hit 90% util. However, in the event this nfs mount is dismounted or unreachable.. Zenoss will alert correct?
[20-Jan-2010 16:32:59] <QubeZ> and then i can write a custom script to invoke the remount?
[20-Jan-2010 16:33:27] <dec3pti0n> still on the OS tab for the device zenoss only finds the total bytes on the root file system !
[20-Jan-2010 16:34:24] <dec3pti0n> I see with snmpwalk under the UCD-SNMP-MIB that it shows quite a bit of info
[20-Jan-2010 16:39:42] <dec3pti0n> and it didn't pickup any memory !
[20-Jan-2010 16:43:00] <rmatte> QubeZ: I don't believe it would alert if it was unmounted
[20-Jan-2010 16:43:06] <rmatte> the only alert is the threshold which is based on usage
[20-Jan-2010 16:43:40] <rmatte> dec3pti0n: check the Performance tab, that's where memory is shown
[20-Jan-2010 16:44:04] <rmatte> the OS tab is for processes, interfaces, disks, and services
[20-Jan-2010 16:46:29] <QubeZ> is there a default way to monitor nfs mounts? i'd assume that was some type of de facto way to do it and implemented by Zenoss.
[20-Jan-2010 16:46:35] <QubeZ> been reading up on it but nothing so far
[20-Jan-2010 16:47:09] <rmatte> QubeZ: there's no built in way to do it
[20-Jan-2010 16:47:15] <rmatte> you'd have to put something together
[20-Jan-2010 16:47:30] <QubeZ> k
[20-Jan-2010 17:01:55] <forsberg> cant you test if it will alert you if you unmount it ?
[20-Jan-2010 17:03:08] <forsberg> or make a script to check if its mounted and return 0 or 1
[20-Jan-2010 17:03:26] <forsberg> or make a graph wich draws 0 and 1 and use threshold to alert
[20-Jan-2010 17:03:26] <forsberg> :)
[20-Jan-2010 17:03:41] <rmatte> forsberg: if it were unmounted I doubt you'd get anything more than a debug message
[20-Jan-2010 17:04:04] <rmatte> I could be wrong
[20-Jan-2010 17:04:05] <forsberg> i have no idea, but i would just test it, if he cannot unmount it, then make another test nfs moutnt
[20-Jan-2010 17:04:21] <rmatte> I'll test it right now...
[20-Jan-2010 17:05:07] <forsberg> but it gets under file systems?
[20-Jan-2010 17:05:24] <rmatte> yes, it does
[20-Jan-2010 17:05:46] <forsberg> i would hope it would alert if a whole part of the monitored filesystem disappears
[20-Jan-2010 17:05:46] <forsberg> :)
[20-Jan-2010 17:05:55] <rmatte> I just unmounted one that I'm monitoring
[20-Jan-2010 17:05:58] <rmatte> now to wait and see...
[20-Jan-2010 17:05:59] <QubeZ> forsberg i dont want to umount it because there are critical recordings going on... not really able to use it as a lab at the moment
[20-Jan-2010 17:06:09] <forsberg> i understand QubeZ
[20-Jan-2010 17:06:16] <forsberg> but then maybe make another test mount
[20-Jan-2010 17:06:22] <rmatte> QubeZ: I'm testing it right now
[20-Jan-2010 17:06:23] <forsberg> anyways seems like rmatte is doing it for you :)
[20-Jan-2010 17:06:24] <QubeZ> ya ill have to do that
[20-Jan-2010 17:06:33] <QubeZ> i feel like such a lazy loser
[20-Jan-2010 17:06:38] <rmatte> lol
[20-Jan-2010 17:06:39] <forsberg> :P
[20-Jan-2010 17:07:08] <rmatte> well, you're not a loser
[20-Jan-2010 17:07:09] <rmatte> :P
[20-Jan-2010 17:07:12] <forsberg> ^^
[20-Jan-2010 17:07:36] <cgibbons> hmmm
[20-Jan-2010 17:08:45] <rmatte> well, so far Zenoss has seen a bunch of processes shutdown on the server (they use the share so I had to shut them down)
[20-Jan-2010 17:08:48] <rmatte> but nothing else
[20-Jan-2010 17:09:04] <QubeZ> hmm
[20-Jan-2010 17:09:27] <forsberg> doenst make sense
[20-Jan-2010 17:09:32] <rmatte> Yeh, that's what I though, there's a debug message
[20-Jan-2010 17:09:37] <rmatte> but I actually just had a good idea
[20-Jan-2010 17:09:47] <rmatte> Error reading value for "/mspfiles" on lab01.novanoc.com (oid .1.3.6.1.2.1.25.2.3.1.6.32 is bad)
[20-Jan-2010 17:09:51] <rmatte> that's the message I got
[20-Jan-2010 17:10:50] <rmatte> what you could do is apply a transform on the /Perf/Snmp class and have it check debug messages, if a debug message comes in with your share name as part of it, you could have it up the severity of the event and change the message and summary
[20-Jan-2010 17:11:03] <rmatte> that'd work
[20-Jan-2010 17:11:30] <rmatte> something like...
[20-Jan-2010 17:11:59] <QubeZ> im going to have to read on transforms
[20-Jan-2010 17:12:57] <rmatte> import re
[20-Jan-2010 17:12:57] <rmatte> if re.search('/mspfiles', evt.summary) and evt.severity = 1:
[20-Jan-2010 17:12:57] <rmatte> evt.summary = 'The /mspfiles NTFS share is down!'
[20-Jan-2010 17:12:57] <rmatte> evt.message = 'The /mspfiles NTFS share is down!'
[20-Jan-2010 17:12:57] <rmatte> evt.severity = 5
[20-Jan-2010 17:14:31] <QubeZ> neat, thanks rmatte. ill be looking into that.
[20-Jan-2010 17:14:53] <forsberg> where is that to be typed in?
[20-Jan-2010 17:16:03] <rmatte> Navigate to /Events/Perf/Snmp
[20-Jan-2010 17:16:18] <rmatte> Then from the dropdown: More -> Transform
[20-Jan-2010 17:16:46] <forsberg> ah
[20-Jan-2010 17:17:03] <QubeZ> rmatte so using this will not override anyting else? just addition?
[20-Jan-2010 17:17:06] <rmatte> I screwed up a line of my script
[20-Jan-2010 17:17:16] <forsberg> :=
[20-Jan-2010 17:17:17] <rmatte> that should have been: and evt.severity == 1:
[20-Jan-2010 17:17:18] <forsberg> :)
[20-Jan-2010 17:17:19] <rmatte> not = 1
[20-Jan-2010 17:17:40] <rmatte> the rest should be fine
[20-Jan-2010 17:17:50] <QubeZ> rmatte also in your msg if should be NFS not NTFS :)
[20-Jan-2010 17:17:51] <rmatte> QubeZ: just addition
[20-Jan-2010 17:17:54] <QubeZ> appreciate the help dude
[20-Jan-2010 17:18:06] <rmatte> QubeZ: yeh, that'd my bad
[20-Jan-2010 17:18:07] <forsberg> hehehe
[20-Jan-2010 17:18:10] <rmatte> it's just an example damnit
[20-Jan-2010 17:18:11] <rmatte> :P
[20-Jan-2010 17:18:13] <rmatte> lol
[20-Jan-2010 17:18:13] <forsberg> :P
[20-Jan-2010 17:18:20] <QubeZ> so in my cashe i need to add thef ull path? /var/spool/asterisk/monitor/repo ?
[20-Jan-2010 17:18:21] <rmatte> :)
[20-Jan-2010 17:18:30] <QubeZ> lol
[20-Jan-2010 17:18:34] <rmatte> or part of the path, whatever you want
[20-Jan-2010 17:18:47] <rmatte> re.search searches the event summary to see if whatever you put in matches a part of the string
[20-Jan-2010 17:19:00] <QubeZ> evt.sevity = 5 is warning or error?
[20-Jan-2010 17:19:20] <rmatte> It's Critical
[20-Jan-2010 17:20:01] <rmatte> 4 is error
[20-Jan-2010 17:20:04] <rmatte> 3 is warning
[20-Jan-2010 17:20:31] <QubeZ> so i can add more code below that for other transforms?
[20-Jan-2010 17:20:46] <rmatte> correct
[20-Jan-2010 17:21:05] <rmatte> you'd just leave out the import re
[20-Jan-2010 17:21:18] <rmatte> but you'd just copy the rest, if you had other shares to do
[20-Jan-2010 17:21:35] <rmatte> It could be made to do it dynamically for all shares, but I don't have the time for that right now
[20-Jan-2010 17:21:44] <forsberg> geeze, tons of stuff i havent looked at yet it seems :)
[20-Jan-2010 17:21:50] <rmatte> I'll try putting something together tomorrow, it'll make a good tip of the month for Matt
[20-Jan-2010 17:22:14] <mrayzenoss> always happy for tips
[20-Jan-2010 17:22:32] <rmatte> I think this would make a good one since it's simple to implement
[20-Jan-2010 17:22:45] <QubeZ> rmatte but lets say that /var/spool/asterisk/monitor/repo exists, in this case it has to because i am mounting my nfs share to that directory. If NFS get umounted for some reason, then /var/spool/asterisk/monitor/repo still exists. How will this know the difference?
[20-Jan-2010 17:22:47] <rmatte> I just need to write up a fancier transform tomorrow which works for all disks
[20-Jan-2010 17:23:53] <rmatte> QubeZ: because Zenoss is not looking to see if the directory exists, the info that it grabs is similar to the info that you see when you do a "df"
[20-Jan-2010 17:24:08] <rmatte> So when it checks and sees that it can't gather usage information for the drive it generates a debug even
[20-Jan-2010 17:24:09] <rmatte> event*
[20-Jan-2010 17:24:30] <QubeZ> ahh and you're capturing that event and parsing it via regex
[20-Jan-2010 17:24:33] <QubeZ> got it!
[20-Jan-2010 17:24:40] <rmatte> correcto
[20-Jan-2010 17:24:54] <rmatte> I'll have to do the transform based on the OID I believe (hopefully part of the OID never changes)
[20-Jan-2010 17:25:08] <rmatte> then have it pluck the name of the drive out and adjust the severity/message
[20-Jan-2010 17:25:19] <rmatte> shouldn't be overly complex to do
[20-Jan-2010 17:25:24] <forsberg> but a nice thing
[20-Jan-2010 17:26:06] <rmatte> yup
[20-Jan-2010 17:26:34] <rmatte> I'll definitely write something better up tomorrow and post it on Wiki. I'll relay the link to you guys when I'm done.
[20-Jan-2010 17:26:52] <QubeZ> rmatte sweet, ill be idling here all night. About to head home soon.
[20-Jan-2010 17:27:03] <rmatte> yeh, I'm out soon too
[20-Jan-2010 17:27:51] <QubeZ> so now my focus would be now that the event is triggered, capture is somehow and invoke a shell script to remount the nfs
[20-Jan-2010 17:28:31] <rmatte> well, that's simple
[20-Jan-2010 17:28:40] <rmatte> Click on Event Manager on the left
[20-Jan-2010 17:28:44] <rmatte> then go to the Commands tab
[20-Jan-2010 17:28:58] <rmatte> you can add a command to run and set filters on it (like you would for an alert)
[20-Jan-2010 17:29:12] <rmatte> so now all you need is the script that actually logs in and does the automated remounts
[20-Jan-2010 17:29:41] <rmatte> which could be done pretty quickly in expect script
[20-Jan-2010 17:29:49] <QubeZ> sweet
[20-Jan-2010 17:30:35] <rmatte> I'm going to rig it up so that the component of the event gets set to the share name
[20-Jan-2010 17:31:04] <rmatte> so you could pass the device IP to the command via ${dev/manageIp}
[20-Jan-2010 17:31:17] <rmatte> and the share name via ${evt/component}
[20-Jan-2010 17:31:27] <rmatte> so you could make the command dynamic for any share.
[20-Jan-2010 17:31:42] <rmatte> and any device
[20-Jan-2010 17:31:50] <QubeZ> ya just gonna do ssh ${dev/manageIP} /root/zenoss_scripts/remountNFS.sh
[20-Jan-2010 17:32:00] <QubeZ> since i already shared the rsa keys between the servers, it wont ask for password
[20-Jan-2010 17:32:02] <rmatte> that'd work
[20-Jan-2010 17:32:27] <rmatte> if you had more than one share you could do: ssh ${dev/manageIP} /root/zenoss_scripts/remountNFS.sh ${evt/component}
[20-Jan-2010 17:32:37] <rmatte> provided the component is set to the name of the share
[20-Jan-2010 17:32:45] <QubeZ> not sure what filter to add tho based on your event transform i just used
[20-Jan-2010 17:32:49] <rmatte> or honestly, why not just do...
[20-Jan-2010 17:32:59] <rmatte> if you had more than one share you could do: ssh ${dev/manageIP} mount ${evt/component}
[20-Jan-2010 17:33:27] <rmatte> filter for Message or summary
[20-Jan-2010 17:33:42] <rmatte> and set it to is
[20-Jan-2010 17:33:48] <rmatte> and then put the message in
[20-Jan-2010 17:34:02] <QubeZ> i just set it to contains and set it to "repo"
[20-Jan-2010 17:34:07] <rmatte> that works
[20-Jan-2010 17:34:08] <QubeZ> because its the only NFS share we use
[20-Jan-2010 17:34:13] <QubeZ> whats the clear command?
[20-Jan-2010 17:34:22] <rmatte> hm?
[20-Jan-2010 17:34:33] <QubeZ> there are two boxes in the edit.. command and clear command
[20-Jan-2010 17:34:51] <rmatte> clear is to automatically clear another event of a certain class when that event comes in
[20-Jan-2010 17:35:04] <QubeZ> ahh ill leaev that alone
[20-Jan-2010 17:35:06] <rmatte> it's the reason why threshold events clear themselves
[20-Jan-2010 17:35:18] <rmatte> yeh, don't bother with that unless you absolutely need to
[20-Jan-2010 17:36:06] <rmatte> well, that'll make a fun little project for me tomorrow
[20-Jan-2010 17:36:32] <rmatte> should only take like 15 or 20 minutes to code and test hopefully
[20-Jan-2010 17:36:37] <QubeZ> this is all documented in the administartive guide?
[20-Jan-2010 17:36:45] <rmatte> what is?
[20-Jan-2010 17:36:58] <QubeZ> these snippets of help you've provided me
[20-Jan-2010 17:37:07] <QubeZ> i want to learn more abotu transforms and zenoss scripting
[20-Jan-2010 17:37:20] <rmatte> well, coding transforms is explained, but I don't think they go in to crazy detail
[20-Jan-2010 17:37:27] <rmatte> and the clear thing is definitely documented
[20-Jan-2010 17:37:51] <rmatte> as for the rest of it, I just came up with the whole using the debug message to monitor the share today lol
[20-Jan-2010 17:37:56] <rmatte> so I doubt that's documented anywhere ;)
[20-Jan-2010 17:38:14] <rmatte> but it'll be documented on Wiki tomorrow
[20-Jan-2010 17:38:57] <rmatte> make sure you changed...
[20-Jan-2010 17:38:59] <rmatte> if re.search('/mspfiles', evt.summary) and evt.severity = 1:
[20-Jan-2010 17:39:00] <rmatte> to
[20-Jan-2010 17:39:03] <rmatte> if re.search('/mspfiles', evt.summary) and evt.severity == 1:
[20-Jan-2010 17:39:07] <rmatte> or the transform won't work
[20-Jan-2010 17:39:47] <rmatte> (typo that I made)
[20-Jan-2010 17:40:16] <QubeZ> yup, i figured it was not assignment but rather compare so i changed it to double ='s
[20-Jan-2010 17:40:38] <QubeZ> so you saw the debug message where? on the zenoss console or gui?
[20-Jan-2010 17:40:54] <rmatte> they show up like any other event
[20-Jan-2010 17:41:00] <rmatte> they just aren't displayed by default
[20-Jan-2010 17:41:32] <rmatte> click on the dropdown under the severity column
[20-Jan-2010 17:41:38] <rmatte> and put a checkmark next to debug
[20-Jan-2010 17:42:35] <rmatte> the OID in the message is always the same (1.3.6.1.2.1.25.2.3.1.6) which is going to make writing the transform easy
[20-Jan-2010 17:42:50] <rmatte> well, part of the OID is always the same rather
[20-Jan-2010 17:42:52] <QubeZ> nice, ill tinker with it
[20-Jan-2010 17:42:58] <rmatte> cool
[20-Jan-2010 17:48:19] <QubeZ> gonna have to test umounting this share over the weekend when things are slow at this call center
[20-Jan-2010 17:48:40] <QubeZ> hopefully it will 1) trigger the event 2) email me its down 3) run the script 4) email me a "clear" msg
[20-Jan-2010 17:49:38] <QubeZ> later rmatte, thanks a bunch!
[20-Jan-2010 17:53:40] <rmatte> no problem, later
[20-Jan-2010 19:47:27] <baffle_> I've had a problem since my first ZenOss installation; I get huge gaps in *most* data polled via perfsnmp.. Is it just me? I've had the problem since 1.x, and now we run 2.5.1..
[20-Jan-2010 19:47:39] baffle_ is now known as baffle
[20-Jan-2010 19:48:34] <baffle> It comes and goes, but it is quite consistent, and the gaps seems to occur on most devices at the same time..
[20-Jan-2010 19:49:09] <baffle> MySQL is separate, and a beefy server. Zenoss is on a beefy server. No iowait to speak of (RAID10, 6 or 8 15k disks)
[20-Jan-2010 19:49:20] <baffle> 20G memory..
[20-Jan-2010 19:50:00] <baffle> Graps ceated from zenping is always 100%. Also if data is polled via local or remote commands.
[20-Jan-2010 19:51:16] <baffle> Sometimes I get gaps in all graphs from the Device template, but not in any other custom templates that also pull from SNMP; I.e. we have MySQL, Apache, Memcache etc SNMP templates that may or might not have gaps at the same time on the same device. :)
[20-Jan-2010 20:04:40] <baffle> Example where all graphs er broken: http://i.imgur.com/mDzJI.png
[20-Jan-2010 20:04:57] <baffle> Example where some graphs are broken: http://i.imgur.com/mDzJI.png
[20-Jan-2010 20:05:13] <baffle> Example where some graphs are broken: http://imgur.com/XYJdZ.png
[20-Jan-2010 20:05:30] <baffle> If anyone has any ideas, I sure could need some help. :)
[20-Jan-2010 20:11:17] * rmatte checks
[20-Jan-2010 20:13:31] <rmatte> If you go to your collector performance stats, what's the number of datapoints showing as?
[20-Jan-2010 20:16:45] <rmatte> baffle: I'm not going to be in here much longer, you alive?
[20-Jan-2010 21:54:36] <zgold_> Greetings! I am having an issue where zencommand gets return code 1 and no output from any python command run locally, however it works fine when I run the same command myself in bash
[20-Jan-2010 23:27:22] <sjobeck> hey, hi, all
[20-Jan-2010 23:27:45] <sjobeck> any one know why our zenoss shows pings failing to hosts that are easily ping-able from the same server when i try it? it is trying to ping as the "zenoss" user & failing? or something like that?
[20-Jan-2010 23:27:56] <sjobeck> i have effective-user set to zenoss inside zope.conf
[20-Jan-2010 23:28:29] <sjobeck> also, if i "su zenoss -" I can ping these hosts
[20-Jan-2010 23:28:55] <sjobeck> as well as fping
[20-Jan-2010 23:43:24] <sjobeck> hello
[20-Jan-2010 23:43:25] <sjobeck> any one ?
[20-Jan-2010 23:43:32] <sjobeck> bueller?
[20-Jan-2010 23:43:32] <sjobeck> bueller?
[20-Jan-2010 23:43:33] <sjobeck> bueller?
[20-Jan-2010 23:43:33] <sjobeck> bueller?
[20-Jan-2010 23:43:33] <sjobeck> bueller?
[21-Jan-2010 00:00:30] [disconnected at Thu Jan 21 00:00:30 2010]
[21-Jan-2010 00:00:30] [connected at Thu Jan 21 00:00:30 2010]
[21-Jan-2010 03:12:40] <baffle> rmatte: Hi, I know you're away but anyway. Number of datapoints is (max) 407.54 zencommand datapoints, 21530 SNMP datapoints, 2510 zenprocess datapoints.
[21-Jan-2010 05:55:11] <chickenandbeans> afternoon
[21-Jan-2010 08:58:38] <mrayzenoss> Get it while it's hot - http:://alpha.http://zenoss.com/2.5.2-beta/zenoss-2.5.2-533-x86-devel-beta.vmware.tar.gz
[21-Jan-2010 08:59:29] <eidolon> morningeeks.
[21-Jan-2010 08:59:36] <mrayzenoss> doh, double "::"
[21-Jan-2010 08:59:49] <mrayzenoss> http://alpha.zenoss.com/files/2.5.2-beta/
[21-Jan-2010 09:01:21] <eidolon> so rmatte / mrayzenoss - the suggestion yesterday on the timeout switch for my COMMAND monitors didn't work. i'm still getting no data in the chart, and no indication that these scripts are even being run.
[21-Jan-2010 09:07:46] patzer_ is now known as patzer
[21-Jan-2010 09:13:53] <eidolon> what's worse is there's an http monitor here... and i'm not even sure if it's doing anything.
[21-Jan-2010 09:22:56] <ckrough> NANOG is in Austin this year
[21-Jan-2010 09:31:38] <Jane_Curry> test
[21-Jan-2010 09:31:49] <mrayzenoss> ckrough: interesting
[21-Jan-2010 09:32:21] <rmatte> mrayzenoss: it's funny how the zenoss screenshot on the main community site page is of zenoss core :P
[21-Jan-2010 09:32:22] <rmatte> erm
[21-Jan-2010 09:32:26] <rmatte> zenoss enterprise rather
[21-Jan-2010 09:32:31] <rmatte> (should be of core)
[21-Jan-2010 09:33:19] <rmatte> eidolon: If it's not the timeout then delete the RRD files
[21-Jan-2010 09:33:43] <eidolon> i think it's something more... idon't think the host is being queried at all
[21-Jan-2010 09:33:51] <eidolon> i just opened a case in support.zenoss.com with screenshots.
[21-Jan-2010 09:33:52] <TBKDan1> rmatte: I'm redoing my zenoss core install and currently trying to install the cisco mib zenpack. It's been sitting at the zenpack install for easily over an hour. The CPU is pegged. Think it's actually still doing stuff or should I kill it?
[21-Jan-2010 09:38:46] <mrayzenoss> rmatte: fixing that
[21-Jan-2010 09:39:40] <eidolon> anyone know of a simple monitor to throw an alert if a filesystem is read only?
[21-Jan-2010 09:40:07] <eidolon> we're having problems on one of our fiber channel array with iscsi. the filesystem just goes read-only when connections get wonky.
[21-Jan-2010 09:40:45] <cwj> eidolon: if you find one i would surely love to know about it
[21-Jan-2010 09:40:47] <mrayzenoss> eidolon: make a command data source using "touch"?
[21-Jan-2010 09:40:56] <eidolon> that's what i was thinking.
[21-Jan-2010 09:41:06] <mrayzenoss> the error code should let you know if it works
[21-Jan-2010 09:41:22] <eidolon> it'll throw an error if the touch fails. i'd need to chew out the filesystem path.. hmm. also, it'd have to be set up to ssh into it as root.
[21-Jan-2010 09:42:35] <mrayzenoss> why root?
[21-Jan-2010 09:43:31] <rmatte> TBKDan1: that Cisco Mibs pack install take 5 to 7 hours
[21-Jan-2010 09:43:34] <rmatte> takes*
[21-Jan-2010 09:43:50] <TBKDan1> ....................................
[21-Jan-2010 09:43:55] <rmatte> also, make sure that you follow the zendmd steps listed on the ZenPack page when it's done
[21-Jan-2010 09:44:27] <rmatte> Hey, you can either let the pack install for 5 to 7 hours or take a shot at loading the Mibs yourself. I guarantee you that it's easier to just let the pack install
[21-Jan-2010 09:44:30] <TBKDan1> yeah, I saw the steps saying about 15 minutes to complete
[21-Jan-2010 09:44:39] <rmatte> 15 minutes to complete the zendmd
[21-Jan-2010 09:44:44] <TBKDan1> Yeah, I see that now :-)
[21-Jan-2010 09:45:10] <rmatte> that pack loads hundreds of thousands of objects in to Zope
[21-Jan-2010 09:45:33] <TBKDan1> Pssh, thousands shmousands. :-)
[21-Jan-2010 09:45:40] <rmatte> It says 4 or 5 hours to process on the ZenPack page, 5 is probably the most average
[21-Jan-2010 09:45:57] <rmatte> I've seen it take up to 7 or 8 but that was on a really slow box
[21-Jan-2010 09:45:59] <TBKDan1> Ok, thanks for the info
[21-Jan-2010 09:46:02] <rmatte> np
[21-Jan-2010 09:46:16] <TBKDan1> Hopefully I'll see a prompt after lunch lol
[21-Jan-2010 09:46:57] <eidolon> re: why root. well, what i'd like to do is apply a template to a filesystem on a monitored host. i don't think that's possible, so i'd need to pass the list of filesystems to the command script somehow. Each of those fs's are root mounted - I'd need to be root to write to, say / - so ssh to the target as root and touch.
[21-Jan-2010 09:46:59] <rmatte> hehe
[21-Jan-2010 09:47:56] <rmatte> oh, you just reminded me, I need to write up that tip on monitoring nfs share status
[21-Jan-2010 09:48:08] <rmatte> I'll have something up on Wiki within an hour or two
[21-Jan-2010 09:49:43] <rmatte> eidolon: I'm sure there's a better way to check if the filesystem is read-only without using touch
[21-Jan-2010 09:49:54] * eidolon is open to suggestions :)
[21-Jan-2010 09:53:40] <mrayzenoss> parse output of /proc/mounts?
[21-Jan-2010 09:53:46] <mrayzenoss> seems painful
[21-Jan-2010 09:54:11] <rmatte> nah, that'd be easy
[21-Jan-2010 09:54:17] <rmatte> eidolon: you're only interested in the root?
[21-Jan-2010 09:54:19] <rmatte> "/"
[21-Jan-2010 09:54:19] <rmatte> ?
[21-Jan-2010 09:54:38] <eidolon> nope. these are iscsi mounted filesystems.
[21-Jan-2010 09:54:43] <eidolon> a few machines have dozensof the.m
[21-Jan-2010 09:55:12] <eidolon> (we're a music / media delivery company. the content volumes are on sunstores, netapps, and a home grown FC array. about 200 terabytes.
[21-Jan-2010 09:59:19] <rmatte> I see
[21-Jan-2010 09:59:28] <rmatte> well then, that's still possible to do...
[21-Jan-2010 10:00:30] <rmatte> afk for 5, then I'll try to work out a one liner for you that does what you want.
[21-Jan-2010 10:08:36] <chickenandbeans> afternoon
[21-Jan-2010 10:08:47] <chickenandbeans> is the dev chat still happening today?
[21-Jan-2010 10:09:04] <mrayzenoss> yeah, it's starting an hour later than usual because of schedule conflicts
[21-Jan-2010 10:09:19] <mrayzenoss> so cgibbons will be here "officially" in 50 minutes
[21-Jan-2010 10:09:37] <chickenandbeans> thanking you
[21-Jan-2010 10:10:19] <rmatte> eidolon: so you want it to check all of the volumes and when they are read-only automatically remount them as read write?
[21-Jan-2010 10:10:40] <rmatte> eidolon: or simply alert on the fact that one is mounted as read-write?
[21-Jan-2010 10:10:42] <eidolon> auto-healing is really the next step. i just want to know when they go read only.
[21-Jan-2010 10:10:44] <venturaville> rmatte: I found a work around for my libvirt issue btw.....
[21-Jan-2010 10:10:57] <rmatte> eidolon: ok, I'm half way there, give me a few more minutes
[21-Jan-2010 10:11:02] <rmatte> venturaville: oh, what was it?
[21-Jan-2010 10:11:09] <venturaville> its devious ... you'll like it
[21-Jan-2010 10:11:15] <rmatte> hehe
[21-Jan-2010 10:11:38] <eidolon> heh. okay :)
[21-Jan-2010 10:11:39] <venturaville> I put an option on my check script to output pickled structure of what the host looks like when passed an option from the pythonplugin based modeler.....
[21-Jan-2010 10:19:27] <forsberg> o/
[21-Jan-2010 10:21:28] <rmatte> eidolon: try this one liner: http://pastebin.com/m36282a2
[21-Jan-2010 10:21:33] <rmatte> see if that gives you what you want
[21-Jan-2010 10:21:54] <eidolon> huh :)
[21-Jan-2010 10:21:57] <eidolon> sec :)
[21-Jan-2010 10:22:15] <rmatte> when I run it I see
[21-Jan-2010 10:22:15] <rmatte> rw
[21-Jan-2010 10:22:17] <eidolon> yah, i see it
[21-Jan-2010 10:22:18] <rmatte> erm
[21-Jan-2010 10:22:22] <rmatte> ok, so it works...
[21-Jan-2010 10:22:23] <eidolon> interesting.
[21-Jan-2010 10:22:26] <rmatte> now, what do you need changed in that?
[21-Jan-2010 10:22:27] <eidolon> yeah, it works on my centos box here.
[21-Jan-2010 10:22:43] <eidolon> i suppose i could pattern-match look for ' ro'
[21-Jan-2010 10:22:48] <rmatte> we obviously can't just leave it like that
[21-Jan-2010 10:22:54] <rmatte> yeh, that'd be easy to do
[21-Jan-2010 10:23:02] <eidolon> okay, thanks. i hadn't thought of going that way
[21-Jan-2010 10:23:06] <eidolon> i'll make up a template for this.
[21-Jan-2010 10:23:10] <rmatte> hehe cool
[21-Jan-2010 10:23:18] <rmatte> what do you need the output to be?
[21-Jan-2010 10:23:37] <rmatte> probably like: "/ ro|"
[21-Jan-2010 10:23:38] <rmatte> ?
[21-Jan-2010 10:24:45] <eidolon> well, i can do a pattern match
[21-Jan-2010 10:24:52] <rmatte> ok
[21-Jan-2010 10:24:57] <rmatte> I'll leave the code with you
[21-Jan-2010 10:25:01] <rmatte> I did the hard part ;)
[21-Jan-2010 10:25:05] <eidolon> seriously though, i can definitely run wth this. i men, i did a bunch of curl grep/sed/awk stuff early to do the same thing :)
[21-Jan-2010 10:25:08] <eidolon> you did, thanks :)
[21-Jan-2010 10:25:19] <rmatte> yeh, this is much much easier than doing that
[21-Jan-2010 10:25:43] <rmatte> people always use awk but I find that I can do almost anything just as well with cut
[21-Jan-2010 10:25:56] <rmatte> cut can replace awk/sed in many cases
[21-Jan-2010 10:25:56] <eidolon> i use sed to chop out stuff in a line a lot
[21-Jan-2010 10:26:12] <rmatte> yeh, it's good for that sort of thing
[21-Jan-2010 10:26:22] <eidolon> particularly when i'm trying to pull out a single value from a string. r ecently i had to scrape a webpage that had "There have been 400 purchases in the last 5 minutes."
[21-Jan-2010 10:26:30] <eidolon> getting hte '400' out and charting it was the target.
[21-Jan-2010 10:26:33] <eidolon> that was sed :)
[21-Jan-2010 10:27:18] <rmatte> hehe
[21-Jan-2010 10:29:01] <rmatte> I'm lazy, I'd just do: cat <output> | grep purchases | cut -d ' ' -f 4 | cut -f 1
[21-Jan-2010 10:29:09] <rmatte> actually
[21-Jan-2010 10:29:21] <rmatte> cat <output> | grep purchases | cut -d ' ' -f 4 | cut -d ' ' -f 1
[21-Jan-2010 10:29:22] <rmatte> rather
[21-Jan-2010 10:30:06] <rmatte> (there would be more to it depending on what the line looked like, obviously there would be some html code or something in there too
[21-Jan-2010 10:30:20] <eidolon> hisssssssssss..
[21-Jan-2010 10:30:30] <eidolon> gratuitous use of 'cat'! 10 yd penalty!
[21-Jan-2010 10:30:37] <rmatte> meh
[21-Jan-2010 10:30:50] <rmatte> grep purchases <output>

[21-Jan-2010 10:30:57] <rmatte> I just have a habbit of using cat
[21-Jan-2010 10:32:24] <rmatte> you call it "gratuitous", I call it "stylish"
[21-Jan-2010 10:32:32] <cgibbons> hurm
[21-Jan-2010 10:32:33] <rmatte> stylish as a cat!
[21-Jan-2010 10:33:00] <rmatte> ok, so time to get this nfs monitoring guide done
[21-Jan-2010 10:33:28] * rmatte fires up the ol' lab box
[21-Jan-2010 10:36:57] <robo> hi: i'm trying to find a zenpack for bigip f5. Does anyone know of a good place that I can start looking to see if someone already made one? I checked here http://community.zenoss.org/community/zenpacks -- but nothing for an f5. Safe to assume there isn't one for it?
[21-Jan-2010 10:38:01] <rmatte> robo: if it's not listed there chances are it doesn't exist
[21-Jan-2010 10:38:24] <robo> :-/
[21-Jan-2010 10:38:39] <mrayzenoss> robo: correct. Everything that has been though any vetting is there, my backlog is here: http://community.zenoss.org/docs/DOC-4624
[21-Jan-2010 10:38:56] <Apachez> robo: you can use splunk with the f5 app :)
[21-Jan-2010 10:39:10] <Apachez> but thats more for loganalysis
[21-Jan-2010 10:39:18] <robo> splunk, hmm
[21-Jan-2010 10:39:45] <robo> oh wait, just saw this on the zenpacks page -- "The following ZenPacks are provided by Zenoss Inc. to users of Zenoss Enterprise." -- Big-IP (the F5) is listed in there
[21-Jan-2010 10:39:55] <robo> so i guess they provide it for enterprise only
[21-Jan-2010 10:40:01] <forsberg> :|
[21-Jan-2010 10:40:05] <mrayzenoss> oh yeah, forgot about that one
[21-Jan-2010 10:40:16] <Apachez> are there no zenpack warez then on piratebay or something? :P
[21-Jan-2010 10:41:02] <robo> I wonder how much enterprise is
[21-Jan-2010 10:41:19] <forsberg> to much
[21-Jan-2010 10:41:20] <forsberg> :)
[21-Jan-2010 10:41:37] <robo> i don't see a price
[21-Jan-2010 10:41:42] <rmatte> man I'm bad at regex
[21-Jan-2010 10:41:51] <rmatte> ok, so to extract a value from between quotation marks
[21-Jan-2010 10:41:58] <rmatte> right now I have ' $(.*?)$ '
[21-Jan-2010 10:42:01] <rmatte> erm sorry
[21-Jan-2010 10:42:02] <forsberg> like 200 $ pr device ? i never forget the phonecall i had with zenoss sales person lol
[21-Jan-2010 10:42:12] <rmatte> ".*?"
[21-Jan-2010 10:42:20] <rmatte> I need to extract it but ommit the quotation marks
[21-Jan-2010 10:42:40] <forsberg> in my cocknewbie/naivity i totally misunderstood the pricing system, and the price he was talking was.....very different that my idea of what it would cost :)
[21-Jan-2010 10:42:48] <robo> $200 / device -- wow
[21-Jan-2010 10:42:53] <Muti> ".*?"
[21-Jan-2010 10:43:14] <Muti> "(.*?)" to capture
[21-Jan-2010 10:43:17] <robo> rmatte, something like /'(.*?)'/
[21-Jan-2010 10:43:19] <Apachez> forsberg: so he said 200 dollah per device and you thought 200 dollah per monitor server?
[21-Jan-2010 10:43:29] <forsberg> haha something like that
[21-Jan-2010 10:43:34] <forsberg> i thought a couple of thousands or something
[21-Jan-2010 10:43:46] <forsberg> and he was like half a mill danish kroners
[21-Jan-2010 10:43:59] <Apachez> heh
[21-Jan-2010 10:44:12] <Apachez> Im about to setup munin + splunk
[21-Jan-2010 10:44:12] <rmatte> robo: it's not liking that
[21-Jan-2010 10:44:22] <forsberg> i guess thats what enterprice grade stuff cost
[21-Jan-2010 10:44:25] <forsberg> but i was/am newbie ;)
[21-Jan-2010 10:44:39] <rmatte> oh well whatever, I'll just leave the quotes
[21-Jan-2010 10:44:44] <ckrough> rmatte: my monthly pimp of this dudes site http://www.gskinner.com/RegExr/
[21-Jan-2010 10:44:53] <rmatte> ckrough: using that right now
[21-Jan-2010 10:45:11] <robo> rmatte, what regex engine?
[21-Jan-2010 10:45:18] <robo> the one sed uses or one that php uses?
[21-Jan-2010 10:45:25] <robo> or better, what language?
[21-Jan-2010 10:45:39] <rmatte> robo: well the end goal is python
[21-Jan-2010 10:45:46] <forsberg> robo its http://www.zenoss.com/ for enterprise
[21-Jan-2010 10:45:49] <robo> oh, i don't know python :-/
[21-Jan-2010 10:45:52] <rmatte> but I'm guessing the one I'm testing with uses php's engine
[21-Jan-2010 10:46:04] <rmatte> robo: regex expressions are pretty universal
[21-Jan-2010 10:46:06] <robo> that should work for php
[21-Jan-2010 10:46:11] <forsberg> http://www.zenoss.com/product/pricing
[21-Jan-2010 10:46:29] <ckrough> rmatte: \'(.*)\'
[21-Jan-2010 10:46:34] <rmatte> ckrough: the only thing that would make the site better is if I could just highlight what I want and have it generate the expression for me lol
[21-Jan-2010 10:47:16] <Apachez> but what I dont get is if the enterprise version is the same install as the opensource version?
[21-Jan-2010 10:47:19] <rmatte> ckrough: it's double quotes
[21-Jan-2010 10:47:32] <Apachez> since the opensource verison have crippled mysql settings (or rather no settings at all) etc...
[21-Jan-2010 10:47:35] <rmatte> and that still matches the quotation marks as part of it
[21-Jan-2010 10:47:39] <Apachez> or is the enterprise version "optimized"?
[21-Jan-2010 10:47:46] <rmatte> here's the string...
[21-Jan-2010 10:47:47] <rmatte> Error reading value for "/mspfiles" on lab01.novanoc.com (oid .1.3.6.1.2.1.25.2.3.1.6.32 is bad)
[21-Jan-2010 10:47:47] <Apachez> and if so why isnt this commited back to community edition?
[21-Jan-2010 10:47:57] <rmatte> I'm looking to extract /mspfiles from the quotes
[21-Jan-2010 10:48:05] <rmatte> but it doesn't matter, I'll just keep the quotes
[21-Jan-2010 10:48:08] <rmatte> it's not a big deal
[21-Jan-2010 10:48:10] <ckrough> \"(.*)\" does it
[21-Jan-2010 10:48:23] <rmatte> I have that in there right now and it's not doing it
[21-Jan-2010 10:48:27] <ckrough> capture group 1 is /mspfiles
[21-Jan-2010 10:48:30] <Muti> zenoss@zenoss:~$ python
[21-Jan-2010 10:48:34] <Muti> >>> import re
[21-Jan-2010 10:48:37] <Muti> >>> m = re.search('"(.*?)"', 'Error reading value for "/mspfiles" on lab01.novanoc.com (oid .1.3.6.1.2.1.25.2.3.1.6.32 is bad)')
[21-Jan-2010 10:48:38] <ckrough> dont look at the highlight, look at the capture group (mouse over)
[21-Jan-2010 10:48:41] <Muti> >>> print m.group(1)
[21-Jan-2010 10:48:43] <Muti> /mspfiles
[21-Jan-2010 10:48:52] <ckrough> ^^
[21-Jan-2010 10:49:10] <rmatte> ohhhhhhhhh
[21-Jan-2010 10:49:11] <rmatte> I see
[21-Jan-2010 10:49:16] <ckrough> you were doing group 0?
[21-Jan-2010 10:49:24] <rmatte> it shows it as group 1
[21-Jan-2010 10:49:40] * ckrough python takes another victim
[21-Jan-2010 10:50:18] <robo> so if zenoss only provides the big-ip zenpack for enterprises customers, couldn't i just build my own zenpack for zenoss?
[21-Jan-2010 10:52:26] <rmatte> ok, got working code, now to finish this up...
[21-Jan-2010 10:52:29] <forsberg> have no idea how they difference it robo :) i havent really had any need for more than core...
[21-Jan-2010 10:52:56] <robo> gotcha
[21-Jan-2010 10:52:59] <robo> okay, ty :-)
[21-Jan-2010 10:54:05] <ckrough> robo: most of the stuff you could probably replicate in CORE. Distributed collectors and support are the two major differentiators.
[21-Jan-2010 10:54:17] <ckrough> robo: and you can do distributed in CORE but it's not nearly as easy as ENT
[21-Jan-2010 10:55:47] <mrayzenoss> robo: yes, if you build an F5 ZenPack I will post it on the site
[21-Jan-2010 10:56:07] <mrayzenoss> just because an Enterprise ZenPack exists doesn't mean the Community can't make one too
[21-Jan-2010 10:56:30] <Apachez> what about if someone just steals the enterprise zenpack and renames a few strings?
[21-Jan-2010 10:56:45] <mrayzenoss> Apachez: I might notice...
[21-Jan-2010 10:56:49] <Apachez> the downside is if there will exist two zenpacks for same thing
[21-Jan-2010 10:57:14] <rmatte> hmmmm, how the heck do I display filesystems for a device, I figured it would just be...
[21-Jan-2010 10:57:19] <rmatte> for f in d.os.filesystems:
[21-Jan-2010 10:57:23] <rmatte> print f.id
[21-Jan-2010 10:57:27] <rmatte> but apparently not
[21-Jan-2010 10:57:30] <robo> okay cool, looks like i'll learn how to build a zenpack. Ty for the help
[21-Jan-2010 10:59:13] <rmatte> should be f.mount but that's not working either
[21-Jan-2010 10:59:36] <ckrough> rmatte: pprint(dir(f)) may give some hints
[21-Jan-2010 10:59:54] <chickenandbeans> We good to go?
[21-Jan-2010 10:59:58] <venturaville> mrayzenoss: ... is the zenpack path available inside of the modeler plugin?
[21-Jan-2010 11:00:09] <rmatte> ckrough: AttributeError: __getitem__
[21-Jan-2010 11:00:23] <rmatte> it's doing that for everything I try
[21-Jan-2010 11:00:41] <ckrough> what that will do is list all of the available methods if there are any
[21-Jan-2010 11:01:00] <ckrough> in this case I guess its not applicable, but it's really helpful in a lot of situations
[21-Jan-2010 11:01:20] <chickenandbeans> I could do with a hand on: http://community.zenoss.org/message/44457
[21-Jan-2010 11:01:24] <rmatte> ckrough: well, the problem is that I know that this device has a disk listed on the OS tab, and if I go to the management interface I can see that the property I'd need to grab is f.mount
[21-Jan-2010 11:01:29] <rmatte> but it's not letting me
[21-Jan-2010 11:01:41] <rmatte> actually, hmmm, I may know why
[21-Jan-2010 11:01:42] <chickenandbeans> could be a bug, could be me :)
[21-Jan-2010 11:02:41] <rmatte> bleh, nope, thought it might have been because the device was in decommed state
[21-Jan-2010 11:03:12] <rmatte> ohhhhhhhhh
[21-Jan-2010 11:03:14] <rmatte> bleh
[21-Jan-2010 11:03:22] <rmatte> forgot the () after d.os.filesystems
[21-Jan-2010 11:03:28] * rmatte slaps forehead
[21-Jan-2010 11:03:31] <mrchippy> rmatte: doh! you beat me to it!
[21-Jan-2010 11:03:43] <rmatte> lol
[21-Jan-2010 11:03:54] <cgibbons> shoot the clock has chimed
[21-Jan-2010 11:04:08] <chickenandbeans> so.... umm yeah... is it a bug?
[21-Jan-2010 11:04:14] <chickenandbeans> I could do with a hand on: http://community.zenoss.org/message/44457 :)
[21-Jan-2010 11:04:44] <chickenandbeans> been playing around with it since... I haven't seen any side effects
[21-Jan-2010 11:06:40] <mrchippy> chickenandbeans: looking now
[21-Jan-2010 11:06:49] <chickenandbeans> cool :)
[21-Jan-2010 11:07:08] <mrayzenoss> venturaville: it's available in the objects.xml, ${here/ZenPackManager/packs/ZenPacks.community.VMwareEsx/path}/
[21-Jan-2010 11:08:45] <Simon4> question from me... the code at http://pastie.org/788441 works, however it doesn't set the DS type to COUNTER as expected... can anyone point me in the direction of how to make it fly?
[21-Jan-2010 11:08:54] <chickenandbeans> it's a weird one, 'rocket' did a good job on finding that line
[21-Jan-2010 11:10:31] <rmatte> well, time to test this transform out
[21-Jan-2010 11:11:16] <cgibbons> i suspect that commit is there for a very good reason
[21-Jan-2010 11:11:50] <chickenandbeans> I would agree
[21-Jan-2010 11:12:02] <chickenandbeans> but it does seem to be causing the error
[21-Jan-2010 11:12:20] <cgibbons> is your 2nd script working with any of the same database objects or is it just the act of running the 2nd dmd that causes the contention?
[21-Jan-2010 11:12:56] <chickenandbeans> it's just the 2nd dmd
[21-Jan-2010 11:13:14] <chickenandbeans> scripts just print out vars, nothing crazy
[21-Jan-2010 11:13:32] <cgibbons> but vars from zeo objects?
[21-Jan-2010 11:13:54] <cgibbons> hold on, i'll try something silly
[21-Jan-2010 11:15:06] <chickenandbeans> ummm it's just: print dmd.Devices.rrdTemplates.....
[21-Jan-2010 11:15:32] <chickenandbeans> All the paths exist, etc
[21-Jan-2010 11:15:42] <cgibbons> and the first one?
[21-Jan-2010 11:15:57] <rmatte> woot, my transform works
[21-Jan-2010 11:16:06] <chickenandbeans> just outputs a text var
[21-Jan-2010 11:16:17] <chickenandbeans> dmd.blah.blah.viewName()
[21-Jan-2010 11:19:15] <Muti> Question for devs: After looking through the bug-fixes I didn't see anythink talking about whether the SQL query from the deviceStatus page that reports Up/Down has been optimized for the case of a large history table
[21-Jan-2010 11:20:50] <chickenandbeans> Perhaps it's in transaction.py?
[21-Jan-2010 11:20:57] <cgibbons> checking one thing on that commit
[21-Jan-2010 11:21:06] <chickenandbeans> ah, wild guess there, ignore me
[21-Jan-2010 11:21:08] <chickenandbeans> ah cool
[21-Jan-2010 11:21:13] <cgibbons> that was added in 2.5 when we added support for encrypting properties
[21-Jan-2010 11:22:06] <mrayzenoss> Muti: was there a ticket for that somewhere?
[21-Jan-2010 11:22:28] <cgibbons> the dev that did that code is out to lunch, but I pinged him and added it to my task list for the day so I'll follow up on your forum post once I chat with him. okay?
[21-Jan-2010 11:22:49] <chickenandbeans> cool
[21-Jan-2010 11:23:13] <chickenandbeans> another problem that may be related was the setting of the rrdtype
[21-Jan-2010 11:23:31] <Muti> mrayzenoss: I'm an enterprise customer and had a case open for it, but I didn't see a Trac ticket for the issue.. I figured I'd just hop in here today and check up on it
[21-Jan-2010 11:24:07] <mrayzenoss> Muti: ahh, send me the case number and I'll poke around
[21-Jan-2010 11:24:09] <Simon4> question from me... the code at http://pastie.org/788441 works, however it doesn't set the DS type to COUNTER as expected... can anyone point me in the direction of how to make it fly?
[21-Jan-2010 11:24:22] <Simon4> (this is related to chickenandbeans :) )
[21-Jan-2010 11:24:53] <cgibbons> looking
[21-Jan-2010 11:25:17] <Simon4> thanks
[21-Jan-2010 11:26:20] <mrchippy> chickenandbeans: i replied in the forums. give me a shout if you need more info or i wasn't clear
[21-Jan-2010 11:26:41] * chickenandbeans reading :)
[21-Jan-2010 11:27:34] <baffle> I've had a problem since my first ZenOss installation; I get huge gaps in *most* data polled via perfsnmp.. Is it just me? I've had the problem since 1.x, and now we run 2.5.1. It comes and goes, but it is quite consistent, and the gaps seems to occur on most devices at the same time. MySQL is separate, and a beefy server. Zenoss is on a beefy server. No iowait to speak of (RAID10, 6 or 8 15k disks) and 20G memory.
[21-Jan-2010 11:27:40] <baffle> Graps ceated from zenping i always 100%. Also if data is polled via local or remote commands. Sometimes I geet gaps in all graphs from the Device template, but not in any other custom templates that also pull from SNMP; I.e. we have MySQL, Apache, Memcache etc SNMP templates that may or might not have gaps at the same time on the same device. :)
[21-Jan-2010 11:27:42] <mrchippy> Simon4: it doesn't set type to COUNTER on the object or in the RRD file?
[21-Jan-2010 11:27:46] <baffle> Example where all graphs are broken: http://i.imgur.com/mDzJI.png - Example where some graphs are broken: http://imgur.com/XYJdZ.png
[21-Jan-2010 11:27:49] <baffle> rmatte: I pinged out mentally yesterday it seems. :-)
[21-Jan-2010 11:28:02] <Simon4> mrchippy: no.. it doesn't throw errors, but it stays as GAUGE in the template
[21-Jan-2010 11:28:10] <Simon4> has me well stumped
[21-Jan-2010 11:28:15] <rmatte> baffle: lol
[21-Jan-2010 11:28:18] <cgibbons> hrm
[21-Jan-2010 11:28:49] <mrchippy> Simon4: It stays GAUGE via the UI or when you look in zendmd? or both?
[21-Jan-2010 11:28:59] <rmatte> baffle: I'm doing some dev work right now so can't try to help until I'm done
[21-Jan-2010 11:29:26] <baffle> Actually, I have to modify my earlier statement; I have 20% iowait on the server it seems.
[21-Jan-2010 11:29:30] <Simon4> it says GAUGE in the UI... when I was playing in comandline it would say COUNTER in zendmd, but only for that zendmd session - i.e. exit zendmd and re-enter it, and ti would be back to gauge
[21-Jan-2010 11:29:38] <Simon4> it's like the commit() doesn't take or something
[21-Jan-2010 11:29:52] <chickenandbeans> mrchippy: yeah, thanks for that, but there are no commits.
[21-Jan-2010 11:29:54] <baffle> Or, hm, not now.
[21-Jan-2010 11:30:40] <cgibbons> isn't it rrdtype not rrdType ?
[21-Jan-2010 11:31:05] <cgibbons> yeah I get an attribute exception if I do ds.datapoints()[0].rrdType but not if I do .rrdtype
[21-Jan-2010 11:31:14] * Simon4 tries quickly
[21-Jan-2010 11:31:15] <rmatte> cgibbons: yeh, you're right
[21-Jan-2010 11:31:25] <baffle> Hmm, actually, it seems as if some devices are quite happy now. Maybe my problems have changed..
[21-Jan-2010 11:31:27] <chickenandbeans> i've found the same problem occurs thought using the lowercase
[21-Jan-2010 11:31:46] <rmatte> baffle: honestly, it sounds like it has to do with your network link more than with your hardware
[21-Jan-2010 11:31:49] <mrayzenoss> baffle: I assume your machine isn't overloaded with I/O?
[21-Jan-2010 11:31:56] <rmatte> baffle: how many datapoints are you polling per cycle?
[21-Jan-2010 11:32:02] <rmatte> (check the collector performance info)
[21-Jan-2010 11:32:36] <cgibbons> it saved for me when I did it with lowercase
[21-Jan-2010 11:32:36] <baffle> mrayzenoss: I sometimes see bursts of IOwait, typically 20% in top. Now it is 0%.
[21-Jan-2010 11:33:18] <baffle> rmatte: zenperfsnmp avg: 21.46k
[21-Jan-2010 11:33:30] <rmatte> baffle: yeh, that's a pretty decent amount
[21-Jan-2010 11:33:31] <baffle> rmatte: zencommand avg: 353.52 .. zenprocess 2.07k.
[21-Jan-2010 11:33:40] <mrchippy> chickenandbeans: which means the commits (as you pointed out) are in the zenoss code. i can give evil foo for your script if you would like--
[21-Jan-2010 11:33:43] <rmatte> baffle: what's your average cycles times?
[21-Jan-2010 11:33:47] <rmatte> cycle*
[21-Jan-2010 11:34:20] <chickenandbeans> mrchippy: :) what does that mean?
[21-Jan-2010 11:34:23] <baffle> rmatte: zenperfsnmp avg: 25.87, zenping avg: 789.88m, zenstatus 3.51u zenmodeler 22.54k.
[21-Jan-2010 11:34:35] <baffle> (Uhh, does the notation actually mean something? :))
[21-Jan-2010 11:34:36] <Simon4> mrchippy: works a treat - nice spotting :) thanks
[21-Jan-2010 11:34:41] <rmatte> chickenandbeans: it's like tofu, but eviler
[21-Jan-2010 11:35:23] <baffle> The graph for Data Points is spotty for zenperfsnmp too.
[21-Jan-2010 11:35:50] <rmatte> baffle: those cycles times are decent, hmmm
[21-Jan-2010 11:36:59] <rmatte> baffle: is all the stuff that you're monitoring on the same LAN?
[21-Jan-2010 11:37:05] <rmatte> or are you monitoring a bunch of remote sites?
[21-Jan-2010 11:37:06] <baffle> rmatte: Noooo, hardly anything.
[21-Jan-2010 11:37:17] <baffle> rmatte: Not remote.
[21-Jan-2010 11:37:27] <baffle> rmatte: As in; They're in the same datacenter.
[21-Jan-2010 11:37:34] <baffle> rmatte: But 100 different VLANs.
[21-Jan-2010 11:37:37] <rmatte> I'm basically wondering if latency is a factor, but apparently it's not
[21-Jan-2010 11:37:50] <baffle> rmatte: And a huge amount of firewalls and routers galore. :)
[21-Jan-2010 11:38:12] <rmatte> you sure you're not just saturating the interface that you're monitoring through?
[21-Jan-2010 11:38:35] <rmatte> that's all I can think of in terms of causes
[21-Jan-2010 11:38:42] <Simon4> your firewalls aren't munching large UDP SNMP packets on the way through?
[21-Jan-2010 11:38:58] <baffle> Simon4: No, this happens on links without firewalls as well.
[21-Jan-2010 11:39:20] <baffle> Simon4: And ZenOss is not behind a firewall, only local IP-table rules.
[21-Jan-2010 11:39:32] <rmatte> baffle, go to zProperties for all devices and try changing zMaxOIDPerRequest from 40 to 20
[21-Jan-2010 11:40:50] <rmatte> you may also want to increase all timeout values in zProperties
[21-Jan-2010 11:41:03] <rmatte> just as a starting point
[21-Jan-2010 11:41:17] <rmatte> if you notice an improvement after doing that, you'll know it's most likely network related
[21-Jan-2010 11:41:39] <rmatte> the fact that it comes and goes as you describe is why I'm thinking network related
[21-Jan-2010 11:41:49] <baffle> rmatte: It has me stumped as well.
[21-Jan-2010 11:42:07] <rmatte> but make the changes that I just specified and see what happens
[21-Jan-2010 11:42:45] <rmatte> You can also try some of this: http://community.zenoss.org/docs/DOC-2521
[21-Jan-2010 11:42:48] <baffle> rmatte: Changed, guess I'll take a new look after and hour or so.
[21-Jan-2010 11:42:56] <rmatte> you have a beefy server, but your Zenoss may not be taking full potential of it
[21-Jan-2010 11:59:48] <etank> is there a way to set up zenoss so that when people log in they are anonymous (no rights to see anything) except for the Welcome portal?
[21-Jan-2010 12:00:27] <etank> or in other words where do i set the default portal layout for all users?
[21-Jan-2010 12:00:53] <rmatte> you can't really set a default portal layout for all users as far as I'm aware
[21-Jan-2010 12:00:57] <robo> $25,000 minimum for zenoss enterprise
[21-Jan-2010 12:01:00] <mrayzenoss> not yet
[21-Jan-2010 12:01:01] <robo> that's crazy
[21-Jan-2010 12:01:02] <rmatte> as far as the permissions go, that could be configured from ZMI
[21-Jan-2010 12:01:19] <rmatte> robo: you can talk them down lol
[21-Jan-2010 12:01:27] <etank> rmatte: ZMI?
[21-Jan-2010 12:01:34] <etank> zope
[21-Jan-2010 12:01:34] <rmatte> etank: Zope Management Interface
[21-Jan-2010 12:01:38] <etank> gotcha
[21-Jan-2010 12:01:54] <etank> i think i have the permissions right now
[21-Jan-2010 12:02:17] <etank> people logging in through LDAP auth can not see anything
[21-Jan-2010 12:02:37] <etank> but i would like to set it so that they can see the Welcome in the dashboard
[21-Jan-2010 12:02:39] <rmatte> well, I'll give you a little hint in terms of the portal...
[21-Jan-2010 12:03:01] <rocket> robo: there is the free version if you are willing to do your own coding of zenpacks and support yourself ...
[21-Jan-2010 12:03:02] <rmatte> navigate to settings, users
[21-Jan-2010 12:03:05] <rmatte> click on a user
[21-Jan-2010 12:03:16] <rmatte> put /manage after the URL
[21-Jan-2010 12:03:27] <rmatte> Click on the Properties tab
[21-Jan-2010 12:03:44] <etank> you mean the dashboardstate
[21-Jan-2010 12:03:44] <rmatte> you'll see a dashboardState property with something like...
[21-Jan-2010 12:03:46] <rmatte> {"layout":"2colbs", "columns":[[{"id":"googlemaps1246911720079", "title":"Locations", "datasource":{"baseLoc":"/Locations/Ottawa", "__class__":"YAHOO.zenoss.portlet.GoogleMapsDatasource"}, "bodyHeight":400, "refreshTime":"60", "__class__":"YAHOO.zenoss.portlet.GoogleMapsPortlet"}, {"id":"toplevelorgs1246559840062", "title":"Systems", "datasource":{"url":"/zport/getRootOrganizerInfo", "queryArguments":{"dataRoot":"Systems", "_dc":"1246912163112"}, "postContent"
[21-Jan-2010 12:03:57] <etank> right
[21-Jan-2010 12:04:04] <rmatte> so you need to somehow figure out where Zenoss grabs that from when creating a new user and modify it
[21-Jan-2010 12:04:41] <rmatte> It would be nice if they'd just add an option in the UI to do this sort of thing
[21-Jan-2010 12:04:46] <rmatte> but I'm sure it'll come eventually
[21-Jan-2010 12:04:56] <etank> but can not set defaults at this point
[21-Jan-2010 12:05:06] <etank> good enough i guess
[21-Jan-2010 12:06:55] * rmatte wonders why it's taking forever for Zenoss to realize that an nfs share has been remounted and monitor it again
[21-Jan-2010 12:07:06] <rmatte> wonder if the snmp index changed
[21-Jan-2010 12:07:31] * rmatte remodels
[21-Jan-2010 12:07:57] <rmatte> nope, it didn't
[21-Jan-2010 12:07:59] <rmatte> hmmm
[21-Jan-2010 12:08:48] <Simon4> rmatte: zenperfsnmp has cached the config that it's not there, and has a long config cycle interval set?
[21-Jan-2010 12:09:07] <rmatte> Simon4: that's possible, just thought of that, pushes changes to the device
[21-Jan-2010 12:09:11] <rmatte> pushed*
[21-Jan-2010 12:09:14] <rocket> robo: I dont know of any monitoring system that is cheap. There is no single api available to gather anything .. bugs in snmp .. all of that takes time and resources to support ..
[21-Jan-2010 12:09:26] * Simon4 runs into that far too often :)
[21-Jan-2010 12:10:08] <robo> rocket, yeah. We're spending about that much now for foglight
[21-Jan-2010 12:10:24] <rocket> robo: BMC is $1000 a node .. IBM's pricing isnt much better .. and neither of them really have a free version that you can use
[21-Jan-2010 12:10:30] <robo> i could get away with getting management to fork over $5k or so.. but $25k -- no way :-)
[21-Jan-2010 12:10:44] <rmatte> Solarwind's pricing is off the wall too, it's like $15,000 per collector
[21-Jan-2010 12:10:50] <robo> wow
[21-Jan-2010 12:11:16] <rocket> robo: at least you could start with the free version and the zenpacks you build would be compatible as you move up if you need to
[21-Jan-2010 12:11:35] <rmatte> yeh, we're using the core version quite successfully
[21-Jan-2010 12:11:35] <mrayzenoss> there's discounted pricing for .edus and Sales is looking for ways to lower their minimums
[21-Jan-2010 12:11:39] <rmatte> we just have a ton of core boxes
[21-Jan-2010 12:12:04] <etank> rmatte: we just switched from foglight to zenoss
[21-Jan-2010 12:12:20] <rmatte> I've never seen foglight
[21-Jan-2010 12:12:27] <rocket> robo: I am sure if you can think of a more reasonable way to do licensing with a decent business for zenoss they might consider it
[21-Jan-2010 12:12:31] <rmatte> the only other ones I've seen are Solarwinds and OpenView
[21-Jan-2010 12:12:39] <rmatte> an ancient version of OpenView at that
[21-Jan-2010 12:12:44] <rocket> robo: they are quite decent guys trying to make a living like all of us .. :p
[21-Jan-2010 12:12:51] <rmatte> yup
[21-Jan-2010 12:12:58] <rmatte> development costs money
[21-Jan-2010 12:13:00] <QubeZ> rmatte did you get that wiki up?
[21-Jan-2010 12:13:01] <robo> rocket, yeah -- I totally get that. I put in a call to sales to see if they can work with me
[21-Jan-2010 12:13:08] <etank> the sales guy at Zenoss that i worked with was great
[21-Jan-2010 12:13:09] <rmatte> QubeZ: still testing the code
[21-Jan-2010 12:13:10] <robo> why i had to leave a voice mail with someone from sales is beyond me :-)
[21-Jan-2010 12:13:16] <QubeZ> rmatte sweet :)
[21-Jan-2010 12:13:27] <rmatte> QubeZ: code is done, but there appear to be a couple of kinks, just need to iron them out
[21-Jan-2010 12:13:28] <mrayzenoss> robo: they're in the Sales kick-off meetings all week
[21-Jan-2010 12:13:35] <robo> ah
[21-Jan-2010 12:13:39] <mrayzenoss> 3 days of all day meetings
[21-Jan-2010 12:14:24] <mrayzenoss> I can poke them with a virtual stick if you don't hear back soon enough :)
[21-Jan-2010 12:14:40] <rmatte> man, for what we're spending no solarwinds collectors we could have unlimited Zenoss Enterprise
[21-Jan-2010 12:14:43] <rmatte> on*
[21-Jan-2010 12:14:54] <cgibbons> and don't forget there's always the balance that small customers are more expensive to service than big ones, so there's another reasonf or minimums :)
[21-Jan-2010 12:14:57] <rmatte> the only reason that our management wants to buy solarwinds over Zenoss is because of the reporting
[21-Jan-2010 12:15:07] <rmatte> Zenoss' reporting really really needs to improve
[21-Jan-2010 12:15:36] <mrayzenoss> at least our dev staff is finally ramping up
[21-Jan-2010 12:15:48] <forsberg> http://www.youtube.com/watch?v=CjOQ9r35uiU
[21-Jan-2010 12:15:49] <mrayzenoss> up to 8 right now and still hiring
[21-Jan-2010 12:15:53] <forsberg> ups sorry wrong chan
[21-Jan-2010 12:16:03] <rmatte> yeh, I'm glad to see that, maybe in about a year Zenoss will be at the point where I can push to have use actually buy the MSP Enterprise package
[21-Jan-2010 12:16:09] <rocket> mrayzenoss: I heard .. :p
[21-Jan-2010 12:16:13] <rmatte> s/use/us
[21-Jan-2010 12:17:14] <mrayzenoss> speaking of making Zenoss betterâ€¦ go grab a 2.5.2 beta and kick the tiresâ€¦ we want this thing to be solid and we need more testing http://alpha.zenoss.com/files/2.5.2-beta/
[21-Jan-2010 12:17:37] <rmatte> I've already installed the beta
[21-Jan-2010 12:17:42] <Simon4> cgibbons: question... if I run zenperfsnmp stop, and then zenperfsnmp start on our core install... it fails to load config for some devices... if I then reindex the db (via reindex() in zendmd), zenperfsnmp starts cleanly and things tick along as before
[21-Jan-2010 12:17:54] <Simon4> is this something other people run into?
[21-Jan-2010 12:17:54] <rmatte> ...and I really like the improvements that I've seen so far
[21-Jan-2010 12:18:21] <rmatte> Unfortunately I can't even consider upgrading until this is fixed: http://dev.zenoss.org/trac/ticket/5978
[21-Jan-2010 12:18:31] <rmatte> because it ruins half of my performance templates
[21-Jan-2010 12:18:34] <cgibbons> hmmm i've never seen anything like that just from starting/starting a daemon, unless there's more stuff going on (like device issues, etc.). but zenperfsnmp batch loads config, iirc.
[21-Jan-2010 12:18:46] <Simon4> yeah, it does, at the start
[21-Jan-2010 12:19:00] <rmatte> I also won't be able to release any of the new ZenPacks that I've been working on until that is fixed either
[21-Jan-2010 12:19:07] <mrayzenoss> rmatte: why no 2.5.2 proposed?
[21-Jan-2010 12:19:20] * Simon4 will document better and post on the forums - was just on the off-chance that it had been seen before as it's fairly irritating
[21-Jan-2010 12:19:25] <rmatte> no idea
[21-Jan-2010 12:19:36] <rmatte> bbibeault backlogged it and that was it
[21-Jan-2010 12:19:56] <mrayzenoss> we must have been in the defect review and reviewed it
[21-Jan-2010 12:20:13] <rmatte> I think it's a pretty major one
[21-Jan-2010 12:20:22] <rmatte> I had to actually go in and comment out that line on all of my servers
[21-Jan-2010 12:20:27] <rmatte> and they've been running fine without it
[21-Jan-2010 12:22:09] <cgibbons> chickenandbeans, you still around?
[21-Jan-2010 12:23:03] * Simon4 heads away, thanks all
[21-Jan-2010 12:23:52] <chickenandbeans> still here :)
[21-Jan-2010 12:24:26] * chickenandbeans is reading up
[21-Jan-2010 12:27:32] <chickenandbeans> mrchippy was extremely helpful... so much thanks :)
[21-Jan-2010 12:27:44] <cgibbons> think we've got a fix for you on that conflict error thinger
[21-Jan-2010 12:28:28] <chickenandbeans> yeah, tell me tell me
[21-Jan-2010 12:28:33] * chickenandbeans jumps up and down
[21-Jan-2010 12:31:53] <cgibbons> ok edit ZenScriptBase.py and change line 55 to
[21-Jan-2010 12:32:13] <cgibbons> if getattr(self.dmd, 'propertyTransfomers', None) is None:
[21-Jan-2010 12:32:21] <cgibbons> save & try your scripts again
[21-Jan-2010 12:33:40] <chickenandbeans> ok, cool... i'll comment out the work around and give it a go
[21-Jan-2010 12:33:42] <chickenandbeans> two secs
[21-Jan-2010 12:35:05] <cgibbons> grr UPS took a header
[21-Jan-2010 12:35:21] <chickenandbeans> mm no dice... the fix didn't work
[21-Jan-2010 12:35:28] <cgibbons> hrm
[21-Jan-2010 12:35:48] <cgibbons> still a conflict on DataRoot ?
[21-Jan-2010 12:36:36] <chickenandbeans> yeah, same error as in the thread
[21-Jan-2010 12:37:00] <chickenandbeans> ah two secs
[21-Jan-2010 12:37:05] <chickenandbeans> i'm an idiot
[21-Jan-2010 12:37:19] <rmatte> lol
[21-Jan-2010 12:37:32] <chickenandbeans> shush you
[21-Jan-2010 12:37:36] <chickenandbeans> ;)
[21-Jan-2010 12:37:58] <cgibbons> typo?
[21-Jan-2010 12:38:47] <chickenandbeans> indeed
[21-Jan-2010 12:38:50] <cgibbons> good :)
[21-Jan-2010 12:39:06] <chickenandbeans> ah ok that works!
[21-Jan-2010 12:39:11] <cgibbons> shew!
[21-Jan-2010 12:39:14] <chickenandbeans> tell me what we fixed
[21-Jan-2010 12:39:19] <chickenandbeans> by 'we' i mean you
[21-Jan-2010 12:39:20] <chickenandbeans> :)
[21-Jan-2010 12:39:31] <cgibbons> not me, I didn't notice the bug, but the author did immediately
[21-Jan-2010 12:39:49] <cgibbons> that code should have only ever set propertyTransformers once and only once, but it was doing it every time
[21-Jan-2010 12:40:02] <cgibbons> lemme get you a defect link in a moment
[21-Jan-2010 12:40:10] <cgibbons> i'll see if i can talk them into putting it into 2.5.2 too
[21-Jan-2010 12:40:39] <rmatte> cgibbons: maybe you can help me out with this...
[21-Jan-2010 12:41:01] <rmatte> if I unmount an nfs share, Zenoss generates a debug message and stops monitoring the share...
[21-Jan-2010 12:41:06] <chickenandbeans> what effect would that have?
[21-Jan-2010 12:41:14] <rmatte> when I remount the share I need to push changes to get Zenoss to start monitoring it again
[21-Jan-2010 12:41:34] <rmatte> is there some property that gets set on that object which tells Zenoss to stop checking it?
[21-Jan-2010 12:42:06] <rmatte> Basically, I'm working on a way to monitor NFS share status, I have a large portion of it done, this is the last thing that I need to figure out
[21-Jan-2010 12:42:19] <cgibbons> how is it being monitored?
[21-Jan-2010 12:42:33] <rocket> cgibbons: woohoo way to go .. :)
[21-Jan-2010 12:42:36] <cgibbons> chickenandbeans: benign except in your case only
[21-Jan-2010 12:42:57] <rmatte> cgibbons: ok, here's the transform: http://pastebin.com/m65b12350
[21-Jan-2010 12:43:07] <rmatte> I apply that to the /Perf/Snmp class
[21-Jan-2010 12:43:26] <rmatte> if it sees a debug message come in for filesystem OID it checks to make sure the drive is of type "networkDisk"
[21-Jan-2010 12:43:42] <rmatte> it then modifies the debug event to become an alert event for NFS being unavailable
[21-Jan-2010 12:44:09] <rmatte> but if I remount the nfs share right afterwards Zenoss completely ignores monitoring it for about an hour unless I manually push changes
[21-Jan-2010 12:44:17] <chickenandbeans> ah cool
[21-Jan-2010 12:44:30] <chickenandbeans> is there a ticket/defect link somewhere
[21-Jan-2010 12:44:31] <rmatte> I'm wondering if there's some bit that gets set on that object to tell Zenoss to ignore monitoring of it until next config cycle
[21-Jan-2010 12:44:54] <rmatte> If there is, I could have the transform unset the bit to have it continuously alert each cycle until the share is available again
[21-Jan-2010 12:45:57] <rmatte> NFS status monitoring is often requested so I decided to come up with a solution
[21-Jan-2010 12:46:31] <cgibbons> what about the monitored attribute of the component?
[21-Jan-2010 12:46:40] <rmatte> that doesn't seem to change
[21-Jan-2010 12:48:08] <rmatte> I'm looking at the properties for the root and the nfs share on the same server
[21-Jan-2010 12:48:14] <rmatte> and there's nothing obviously different
[21-Jan-2010 12:50:18] <rmatte> do you know of a way via dmd/python that I could force Zenoss to push changes to the collector for that particular object?
[21-Jan-2010 12:50:51] <rmatte> oh, I can do: d.collectDevice()?
[21-Jan-2010 12:50:57] <dec3pti0n> from 1 to 10 , how hard is it to setup collectors using the core free version ?
[21-Jan-2010 12:51:09] <rmatte> oh no, that's to remodel
[21-Jan-2010 12:51:10] <rmatte> not push
[21-Jan-2010 12:51:12] <rmatte> hmmmm
[21-Jan-2010 12:51:12] <rocket> dec3pti0n: from 1 to 10 .. :p
[21-Jan-2010 12:51:17] <dec3pti0n> :)
[21-Jan-2010 12:51:23] <cgibbons> if zenhub sees that the device/components changes, it's supposed to do an asynchronous update to the collector... but lesee
[21-Jan-2010 12:51:32] <cgibbons> there's that UI push config command so it must do something
[21-Jan-2010 12:51:39] <rocket> dec3pti0n: different people find it hard .. others find it easy .. its hard to say ..
[21-Jan-2010 12:51:45] <cgibbons> hang on lemme finish with this defect real quick
[21-Jan-2010 12:52:01] <dec3pti0n> rocket: is it explained how to do somewhere ?
[21-Jan-2010 12:52:09] <rmatte> cgibbons: thanks, no rush
[21-Jan-2010 12:53:13] <rocket> dec3pti0n: found this via google almost right away http://community.zenoss.org/docs/DOC-2496
[21-Jan-2010 12:53:31] <rmatte> aha d.pushConfig() looks like
[21-Jan-2010 12:53:34] <rmatte> going to try it
[21-Jan-2010 12:54:01] <dec3pti0n> rocket: cool thanks
[21-Jan-2010 12:55:01] <cgibbons> chickenandbeans: http://dev.zenoss.org/trac/ticket/6081
[21-Jan-2010 12:55:13] <cgibbons> we've got a defect review in an hour so it'll come up there
[21-Jan-2010 12:55:26] <cgibbons> you can zenpatch 16698 until it gets shipped
[21-Jan-2010 12:57:06] <chickenandbeans> editing the thread
[21-Jan-2010 12:58:44] <rmatte> if this d.pushConfig() works this'll be awesome
[21-Jan-2010 12:58:57] <cgibbons> either that or it'll eat your database
[21-Jan-2010 12:59:12] <rmatte> what will?
[21-Jan-2010 12:59:23] * rmatte looks around worriedly
[21-Jan-2010 13:00:51] <rmatte> according to a post by chet there's nothing wrong with d.pushConfig()
[21-Jan-2010 13:00:59] <cgibbons> i was joking
[21-Jan-2010 13:01:04] <rmatte> don't do that
[21-Jan-2010 13:01:04] <cgibbons> poorly
[21-Jan-2010 13:01:05] <rmatte> :P
[21-Jan-2010 13:01:25] <rmatte> when a Zenoss developer says something like that I take it seriously :P
[21-Jan-2010 13:01:25] <rocket> lol
[21-Jan-2010 13:02:29] <cgibbons> i have this strange way of causing things to break that i interact with, so it might be safe, but it might not be just because ;)
[21-Jan-2010 13:02:30] <cgibbons> ask rocket
[21-Jan-2010 13:02:39] <rmatte> lol
[21-Jan-2010 13:02:57] <rocket> cgibbons: I have a strange ability to notice breaking things .. :p
[21-Jan-2010 13:03:00] <rmatte> I did notice that there are often netsplits shortly after you join the channel
[21-Jan-2010 13:03:19] <rocket> cgibbons: skype wont work for me at all today .. :/
[21-Jan-2010 13:03:44] <cgibbons> so far mine has been great, so don't jinx me!
[21-Jan-2010 13:03:52] <rmatte> lol
[21-Jan-2010 13:03:55] <rocket> thats cuz I am offline .. ;p
[21-Jan-2010 13:05:31] <rmatte> am I going to have to do a txnCommit() after this d.pushConfig()?
[21-Jan-2010 13:05:38] <rmatte> or should it just work?
[21-Jan-2010 13:05:48] <chickenandbeans> updating forum thread
[21-Jan-2010 13:07:19] <chickenandbeans> right im outta here :)
[21-Jan-2010 13:07:50] <rmatte> later
[21-Jan-2010 13:07:54] <chickenandbeans> thanks guys for all your help
[21-Jan-2010 13:09:50] <rmatte> Once they get me a proper lab box here I'm going to take a shot at developing a zenpack for core which makes it simple to setup remote collectors
[21-Jan-2010 13:11:03] <forsberg> :P
[21-Jan-2010 13:11:08] * forsberg cheers for rmatte
[21-Jan-2010 13:11:28] <rmatte> hehe
[21-Jan-2010 13:16:03] <rmatte> eugh, I might actually have to remodel the device then push changes
[21-Jan-2010 13:22:19] <rmatte> it'd be nice if there were a way to tell Zenoss just to remodel for a specific object
[21-Jan-2010 13:22:35] <rmatte> actually, I wonder if I could do that
[21-Jan-2010 13:22:56] <rmatte> instead of d.collectDevice, think f.collectDevice() would work?
[21-Jan-2010 13:23:07] <rmatte> ah probably not
[21-Jan-2010 13:23:08] <rmatte> oh well
[21-Jan-2010 13:27:40] <rmatte> ah crap, that's not going to work, hmmmm
[21-Jan-2010 13:34:18] <baffle> rmatte: Hmm, I think you might be right, it might be network related.
[21-Jan-2010 13:35:12] <baffle> rmatte: Looking forward to seeing your project, wonder how popular it will be. :)
[21-Jan-2010 13:35:27] <baffle> rmatte: We have out own customer-facing portal for exposing graphs to customers.
[21-Jan-2010 13:35:28] <rmatte> aha
[21-Jan-2010 13:35:34] <rmatte> I'll need to add a: f.lockFromDeletion() in there
[21-Jan-2010 13:37:01] <rmatte> baffle: cool
[21-Jan-2010 13:40:50] <baffle> rmatte: Portal is on a separate server; it was a 1 week job I think.. It fetches all devices related to $usernames $company, and checks if they are tagged as "In Zenoss". Produces a list of all servers.. You click on servers to expose all Performance-graphs (from all tabs) that is available.. We also cache all the images in a local memcached-instance, in case customers are pressing refresh a lot. :)
[21-Jan-2010 13:41:48] <baffle> rmatte: We also have an SLA util, that collects availability reports for either Ping or SNMP (or, well, whatever) and generates nice reports.. But that one interacts with our ticketing systems SOAP-interface as well, to pull out SLAs on response times on tickets etc..
[21-Jan-2010 13:41:48] <rmatte> cool
[21-Jan-2010 13:42:09] <rmatte> yeh, we're doing something similar
[21-Jan-2010 13:42:23] <rmatte> Zenoss interacts with our ticketing system, then we have our custom CMDB tied in to Zenoss
[21-Jan-2010 13:42:40] <rmatte> and the CMDB will eventually be our portal and will control everything including auto-notifications
[21-Jan-2010 13:42:54] <baffle> rmatte: I think the new version is collecting availability reports every night, and storing them in a DB, so that generating the reports are faster. :)
[21-Jan-2010 13:43:09] <rmatte> for now it contains location/device contacts, circuit info, login details, etc...
[21-Jan-2010 13:43:19] <rmatte> cool
[21-Jan-2010 13:43:20] <baffle> rmatte: Yeah, we have a homebrew cmdb/inventory/invoice/crm system as well.
[21-Jan-2010 13:43:35] <baffle> (It is all connected)
[21-Jan-2010 13:43:40] <rmatte> we have something that automatically generates and formats a report for each client
[21-Jan-2010 13:44:19] <rmatte> not sure if you saw this: http://community.zenoss.org/docs/DOC-4664
[21-Jan-2010 13:44:20] <rmatte> ?
[21-Jan-2010 13:44:28] <rmatte> those are the scripts that we trigger to generate the report
[21-Jan-2010 13:44:37] <rmatte> then we have a php script that handles the formatting
[21-Jan-2010 13:45:29] <baffle> We have a web based interface for the reports; They need to be made manually because we have to filter out errors. :)
[21-Jan-2010 13:46:38] <rmatte> ah
[21-Jan-2010 13:46:53] <rmatte> well, we do have some human interaction
[21-Jan-2010 13:46:58] <rmatte> but for the most part they are automated
[21-Jan-2010 13:47:33] <rmatte> used to take us like 2 days to prepare each report by hand, now it takes like an hour from start to finish
[21-Jan-2010 13:47:46] <rmatte> most of which is spent running the reports
[21-Jan-2010 13:50:08] <QubeZ> hello all
[21-Jan-2010 13:50:11] <QubeZ> rmatte you around?
[21-Jan-2010 13:50:14] <rmatte> yup
[21-Jan-2010 13:50:18] <QubeZ> question...
[21-Jan-2010 13:51:07] <QubeZ> when i setup the command to run my remountNFS.sh when the message had "repo" in it as in "NFS share /repo is down!". What happens when a CLEAR event occurs from Zenoss? Wont it see that the word "repo" is in that message also and then i'll be caught in a cycle?
[21-Jan-2010 13:51:34] <rmatte> that's why you should be adding another filter for the event severity
[21-Jan-2010 13:51:57] <rmatte> make sure that the severity of the event is greater than 1 and you'll be fine
[21-Jan-2010 13:52:21] <QubeZ> i set it to Severity >= Critical
[21-Jan-2010 13:52:35] <rmatte> well, you don't really need the > lol
[21-Jan-2010 13:52:39] <rmatte> there's nothing higher than Critical
[21-Jan-2010 13:52:42] <rmatte> but that'll work
[21-Jan-2010 13:52:43] <QubeZ> true
[21-Jan-2010 13:53:14] <rmatte> I'm still working on this, it's a bit trickier than I thought yesterday
[21-Jan-2010 13:53:39] <QubeZ> so that transform code sets the severity to 5 which is critical. evt.severity = 5 for when this event occurs
[21-Jan-2010 13:53:48] <rmatte> correct
[21-Jan-2010 13:53:58] <baffle> rmatte: Yeah, it takes less than an hour now I think.
[21-Jan-2010 13:54:08] <rmatte> but honestly, you'll want to use this new transform code when it's done
[21-Jan-2010 13:54:14] <rmatte> the transform I gave you yesterday is quite rough
[21-Jan-2010 13:54:20] <QubeZ> this way, my emails will have status: critical and it will spit the message out, run the command (only when severity = 5) which is the case here
[21-Jan-2010 13:54:30] <rmatte> baffle: cool
[21-Jan-2010 13:54:36] <QubeZ> rmatte i will but before i use it, ill want to learn what you did exactly
[21-Jan-2010 13:54:46] <rmatte> I'll explain it on the wiki page
[21-Jan-2010 13:54:48] <rmatte> lol
[21-Jan-2010 13:54:50] <QubeZ> k
[21-Jan-2010 13:54:56] <baffle> rmatte: I want to open source as much as we can, but we never get around to doing that. :)
[21-Jan-2010 13:55:15] <QubeZ> so when severity is set to 5 and the command is run to remount, Zenoss sees this and sets the severity to what exactly?
[21-Jan-2010 13:55:20] <rmatte> baffle: I try to do that as much as possible, but same thing, it's a matter of time
[21-Jan-2010 13:55:28] <baffle> Like, our snmp agent is available for all if you know the URL..
[21-Jan-2010 13:55:39] <rmatte> hehe
[21-Jan-2010 13:56:16] <rmatte> QubeZ: it won't change the severity just because the drive was remounted
[21-Jan-2010 13:56:54] <rmatte> QubeZ: I'm not even sure that I'll be able to get it to auto-clear the event, I'm going to keep it simple and just focus on the initial alert
[21-Jan-2010 13:57:00] <QubeZ> so when Zenoss is clearing the event, how does it know to not run my sh script again, even with the condition severity >= critical
[21-Jan-2010 13:57:16] <rmatte> QubeZ: people can acknowledge it and move it to history manually anyways, it's not something that likely to happen often
[21-Jan-2010 13:57:29] <rmatte> QubeZ: oh, I see what you're asking
[21-Jan-2010 13:57:34] <rmatte> clear events always come in as sev 1
[21-Jan-2010 13:57:36] <baffle> rmatte: Hey, I think I might have found the problem.
[21-Jan-2010 13:57:37] <baffle> RX packets:2152223328 errors:0 dropped:40821589 overruns:0 frame:18097
[21-Jan-2010 13:57:48] <QubeZ> rmatte ahh that works for me then :)
[21-Jan-2010 13:57:49] <rmatte> so since your command requires sev 5 to kick off, it won't run it again
[21-Jan-2010 13:58:06] <rmatte> baffle: yeh, I thought so
[21-Jan-2010 13:58:23] <rmatte> It really really sounded like a network issue
[21-Jan-2010 13:58:39] <rmatte> although that's not a huge percentage of drops
[21-Jan-2010 13:59:09] <rmatte> that's 1.8%
[21-Jan-2010 13:59:26] <rmatte> and 2% is deemed as "Acceptable"
[21-Jan-2010 13:59:42] <rmatte> so I doubt it'd be causing much of an issue unless all of those drops were very recent
[21-Jan-2010 14:00:08] <QubeZ> zenoss is showing my 4 T1 circuit multilink at 3.088. I've remodelled and canot get it to read 6.176 and so the threshold events are running rampant.
[21-Jan-2010 14:00:11] <rmatte> oh my freaking god, I think I finally got this transform to work
[21-Jan-2010 14:00:34] * rmatte lets it run for a bit longer
[21-Jan-2010 14:00:53] <baffle> rmatte: Well, I think *most* traffic go thru. And I don't think those stats have been cleared in a while. :)
[21-Jan-2010 14:01:07] <rmatte> baffle: probably not
[21-Jan-2010 14:02:12] <baffle> rmatte: It's not on the ZenOss servers, it's on one Linux server on the same IP subnet that has lots of rfc1918 networks behind.. I.e. it has heaps of separate routing tables and stuff.. No idea why it is dropping packets, tho'. Switch reports no errors.. So it must be the OS. :-/
[21-Jan-2010 14:03:36] <rmatte> I see
[21-Jan-2010 14:04:03] <rmatte> QubeZ: then the speed that'd being revealed via snmp is wrong for that interface
[21-Jan-2010 14:04:12] <rmatte> QubeZ: it could be wrong in the actual config on the device
[21-Jan-2010 14:04:14] <QubeZ> rmatte will check that, thanks
[21-Jan-2010 14:04:55] <rmatte> ok, now to remount this drive and see if monitoring kicks back in
[21-Jan-2010 14:05:29] <rmatte> if this works then I'm done and I'll throw it up on wiki
[21-Jan-2010 14:05:35] <QubeZ> rmatte does it make sense to run a umount /repo just in case before the remount to make certain its unmounted?
[21-Jan-2010 14:06:01] <rmatte> QubeZ: nah, I wouldn't bother
[21-Jan-2010 14:06:34] <rmatte> QubeZ: if it weren't properly unmounted it would still be visible via snmp, the odds of that happening are slim to none
[21-Jan-2010 14:12:48] <rmatte> I just need to wait a while longer to make sure that Zenoss resumes monitoring of the drive in a timely fashion now after the changes that I made
[21-Jan-2010 14:13:28] <rmatte> it was taking like an hour to resume monitoring previously.
[21-Jan-2010 14:13:49] <rmatte> It should only take 15 minutes this time
[21-Jan-2010 14:15:39] <QubeZ> so the script executed for ya?
[21-Jan-2010 14:16:23] <rmatte> I didn't test with an actual script
[21-Jan-2010 14:16:30] <rmatte> I'm just focussing on the transform itself
[21-Jan-2010 14:16:49] <rmatte> what people choose to do with the event is up to them
[21-Jan-2010 14:18:15] <rmatte> I just want to make sure that the transform is working the way it's supposed to be
[21-Jan-2010 14:18:20] <QubeZ> k
[21-Jan-2010 14:24:06] <baffle> rmatte: Actually, it seems it is not those resets; I think they might be old. So, that's not it.
[21-Jan-2010 14:24:13] <rmatte> ah
[21-Jan-2010 14:24:22] <rmatte> I still think it's something network related
[21-Jan-2010 14:24:31] <rmatte> maybe somewhere down the line
[21-Jan-2010 14:24:33] <baffle> rmatte: Yeah, I sure haven't ruled it out. :)
[21-Jan-2010 14:24:44] <rmatte> I've never seen gaps that bad in graphs
[21-Jan-2010 14:24:59] <rmatte> the only other explanation is that the server isn't gathering the data quickly enough during the polling cycle
[21-Jan-2010 14:25:02] <baffle> rmatte: It's all L2 from behind that box, so I am quite confused. :)
[21-Jan-2010 14:25:05] <rmatte> so like half your data collection times up
[21-Jan-2010 14:25:32] <baffle> rmatte: Checking out all those performance tips you posted now.
[21-Jan-2010 14:30:59] <rmatte> sweet, this transform is working perfectly
[21-Jan-2010 14:31:04] <rmatte> time to publish
[21-Jan-2010 14:31:36] <rmatte> on a box that powerful those performance tips with drastically increase performance
[21-Jan-2010 14:31:55] <rmatte> they doubled and almost trippled performance on one of my boxes
[21-Jan-2010 14:32:10] <rmatte> It's monitoring 18000 datapoints and I rarely ever get any drops
[21-Jan-2010 14:32:25] <rmatte> and that's 18000 datapoints gathered across the world, not restricted to one location
[21-Jan-2010 14:34:50] <baffle> rmatte: Some of those datapoints are remote, but all in Norway. I think..
[21-Jan-2010 14:35:00] <rmatte> ah
[21-Jan-2010 14:35:11] <baffle> rmatte: I've made some modifications, but not all..
[21-Jan-2010 14:35:16] <rmatte> on that particular box I've got them all over europe, Mexico, and the U.S
[21-Jan-2010 14:35:25] <baffle> rmatte: Most are in our datacenters, tho'.
[21-Jan-2010 14:35:41] <rmatte> baffle: check this out: http://dmon.org/graphics/zenoss/gen2.png
[21-Jan-2010 14:36:26] <baffle> rmatte: Sweeeet.
[21-Jan-2010 14:36:30] <rmatte> hehe
[21-Jan-2010 14:36:58] <rmatte> ok, I need to author this wiki document, afk for a few
[21-Jan-2010 14:36:59] <ckrough> rmatte: are these all one install or are they each their own?
[21-Jan-2010 14:37:02] <baffle> rmatte: Wich reminds me that I kind of have to put up proper locations for all devices so I get cool graphics like that. ;) I just recently added proper locations for some of our DCs.
[21-Jan-2010 14:37:05] <eidolon> rmatte: whoah. :)
[21-Jan-2010 14:37:47] <rmatte> there's also a couple of devices in australia but I didn't bother showing that in the screenshot
[21-Jan-2010 14:38:14] <baffle> rmatte: But, hmm, do you have transparent tunnels (I.e. MPLS or something) between countries or have you hacked something to get the links up?
[21-Jan-2010 14:38:53] <baffle> rmatte: But this must be a lucky screenshot; Eveything is green...
[21-Jan-2010 14:39:00] <rmatte> I didn't do any work on their network
[21-Jan-2010 14:39:08] <rmatte> nor any modifications to Zenoss
[21-Jan-2010 14:39:30] <rmatte> From how it was explained to me, as long as both points have IPs in the same subnet Zenoss draws a line
[21-Jan-2010 14:39:35] <rmatte> it has nothing to do with routes or anything
[21-Jan-2010 14:40:06] <baffle> rmatte: Yeah, as long as they are directly connected, yeah. (I.e. in same subnet)
[21-Jan-2010 14:40:14] <rmatte> yup
[21-Jan-2010 14:40:24] <baffle> rmatte: So.. Hmm.. Maybe it's lots of VPN tunnels in every direction. :)
[21-Jan-2010 14:40:32] <rmatte> probably
[21-Jan-2010 14:40:42] <rmatte> this client's network is insane
[21-Jan-2010 14:40:45] <rmatte> lol
[21-Jan-2010 14:41:50] <baffle> Hmm, it is kinda nice to see all the nice lines. I really should get around to pasting in all locations for remote sites. :)
[21-Jan-2010 14:42:27] <baffle> rmatte: Do you do any flow accounting?
[21-Jan-2010 14:44:08] <baffle> Meh, it isn't that linuxbox, I see the same problem with another link.
[21-Jan-2010 14:46:48] <mrayzenoss> Anyone have a good FreeBSD SNMP template?
[21-Jan-2010 14:52:47] <rmatte> oh my freaking god
[21-Jan-2010 14:53:04] <rmatte> I just had the entire wiki page typed up and when I went to save it I got an error from Jive and lost everything
[21-Jan-2010 14:53:08] <rmatte> man that drives me nuts
[21-Jan-2010 14:53:26] <rmatte> An unexpected error has occurred
[21-Jan-2010 14:53:26] <rmatte> * The form has already been processed or no token was supplied, please try again.
[21-Jan-2010 14:55:09] <rmatte> god I hate Jive sometimes
[21-Jan-2010 14:56:38] <rmatte> man, I had that typed up so well too, this is going to take forever to retype
[21-Jan-2010 14:59:42] <baffle> mrayzenoss: Doesn't that use net-snmp? Or does it have its own bundled snmpd now?
[21-Jan-2010 15:02:53] <baffle> rmatte: Install this before you do that again: http://userscripts.org/scripts/show/7671
[21-Jan-2010 15:03:17] <baffle> rmatte: Or, well, there are other plugins as well that might be better..
[21-Jan-2010 15:03:24] <baffle> rmatte: I think I had an addon before.
[21-Jan-2010 15:03:28] <rmatte> I'm just resorting to gedit
[21-Jan-2010 15:05:03] <baffle> rmatte: Good move.
[21-Jan-2010 15:05:59] <mrayzenoss> baffle: http://dev.zenoss.org/trac/ticket/6080
[21-Jan-2010 15:11:56] <rmatte> Hot off the press: http://community.zenoss.org/docs/DOC-4673
[21-Jan-2010 15:15:13] <rmatte> woohoo
[21-Jan-2010 15:16:27] <rmatte> I'd love to see a FreeBSD template added
[21-Jan-2010 15:16:39] <rmatte> it's fairly common
[21-Jan-2010 15:22:45] <QubeZ> rmatte how come you put the entire OID string in your code? I recall you mentioning yesterday that only part of it stays the same in the debug message.
[21-Jan-2010 15:23:35] <mrayzenoss1> Zenoss is hiring: http://community.zenoss.org/message/44804
[21-Jan-2010 15:23:40] <rmatte> I didn't
[21-Jan-2010 15:24:05] <QubeZ> ahh nevermind, just finished editing a csv file so my eyes are loopy
[21-Jan-2010 15:24:10] <rmatte> lol
[21-Jan-2010 15:24:54] <rmatte> I put the whole OID that actually shows up in the filesystem template
[21-Jan-2010 15:25:04] <rmatte> but the collector plugin dynamically adds on to that OID
[21-Jan-2010 15:25:11] <rmatte> so that's not the whole OID
[21-Jan-2010 15:26:14] <rmatte> that transform actually works suprisingly well
[21-Jan-2010 15:26:24] <rmatte> I'm going to implement it for our NFS stuff
[21-Jan-2010 15:28:04] <dec3pti0n> on the zenoss docs, under the templates section, it mentions that common device templates includes for Mysql, apache, Active directory ... and others
[21-Jan-2010 15:28:18] <dec3pti0n> but I don't see those available on my zenoss core install
[21-Jan-2010 15:28:52] <dec3pti0n> I'm under /Devices All Performance templates, is that the wrong place ?
[21-Jan-2010 15:29:00] <mrayzenoss1> dec3pti0n: you just need to install more Core ZenPacks
[21-Jan-2010 15:29:06] <dec3pti0n> ah
[21-Jan-2010 15:29:14] <rmatte> dec3pti0n: the zenoss docs also cover some enterprise stuff too
[21-Jan-2010 15:29:17] <mrayzenoss1> http://community.zenoss.org/community/zenpacks
[21-Jan-2010 15:29:22] <dec3pti0n> I guess I need to get to that reading :)
[21-Jan-2010 15:29:26] <rmatte> hehe
[21-Jan-2010 15:29:52] <mrayzenoss1> well, they're not all installed by default since lots of them are kinda useless, but they probably should be
[21-Jan-2010 15:29:53] <dec3pti0n> I have not read anything about zenpacks on the docs yet
[21-Jan-2010 15:30:16] <rmatte> I wouldn't personally want them all installed by default
[21-Jan-2010 15:30:19] <mrayzenoss1> if you did an RPM install you can install all of them
[21-Jan-2010 15:30:25] <rmatte> like, what am I going to do with an IRC ZenPack?
[21-Jan-2010 15:30:26] <rmatte> lol
[21-Jan-2010 15:30:32] <mrayzenoss1> hey, I wrote that one
[21-Jan-2010 15:30:38] <mrayzenoss1> it's super useless
[21-Jan-2010 15:30:45] <rmatte> it's good, it just doesn't have any practical uses for me
[21-Jan-2010 15:30:50] <mrayzenoss1> no, it's not good
[21-Jan-2010 15:31:00] <mrayzenoss1> who cares how many users are logged into a channel?
[21-Jan-2010 15:31:06] <rmatte> oh it only monitors how many users?
[21-Jan-2010 15:31:13] <mrayzenoss1> it was a training exercise for new developers
[21-Jan-2010 15:31:16] <rmatte> ah
[21-Jan-2010 15:31:28] <rmatte> I could extend the crap out of that pack, I might take a stab at it one day
[21-Jan-2010 15:31:45] <mrayzenoss1> yeah, there's lots of useful things that could be done, that ZenPack does none of them :)
[21-Jan-2010 15:31:51] <mrayzenoss1> and I can say that, since I wrote it
[21-Jan-2010 15:31:55] <rmatte> I'm an administrator on an irc network
[21-Jan-2010 15:32:02] <rmatte> so it could come in handy for us eventually
[21-Jan-2010 15:32:56] <dec3pti0n> are all the "Zenoss Core ZenPacks" already included in the zenoss stack install ?
[21-Jan-2010 15:33:09] <rmatte> no
[21-Jan-2010 15:33:12] <rmatte> none of them are
[21-Jan-2010 15:33:23] <rmatte> you need to install whichever ones you need
[21-Jan-2010 15:33:28] <dec3pti0n> cool thanks
[21-Jan-2010 15:34:04] mrayzenoss1 is now known as mrayzenoss
[21-Jan-2010 15:34:10] <dec3pti0n> perhaps I should have installed zenoss 2.4 , apache monitor requires 2.4 .
[21-Jan-2010 15:36:38] <baffle> Hau stange; If I do an snmpwalk on one of the devices where I have gaps in graphs, Ctrl-Cs it right away and tries to reissue the snmpwalk it just times out.. For many minutes.. W t f..
[21-Jan-2010 15:37:59] <mrayzenoss> dec3pti0n: that's 2.4 and later
[21-Jan-2010 15:38:06] <mrayzenoss> unless someone really screwed up
[21-Jan-2010 15:38:19] <baffle> rmatte: Public or closed network? I used to run a FreeNode server.
[21-Jan-2010 15:38:35] <dec3pti0n> that's what I was thinking , thanks
[21-Jan-2010 15:38:49] <baffle> Heh, irc.no.freenode.net still points to my IP, but it hasn't been up for 2? years.
[21-Jan-2010 15:40:14] <mchesmo3> New to Zenoss...downloaded and setting up the VM on Suse...It seems really slow to respond to do anything. is this just a VM issue. I can move it ot a much more powerful server if need be.
[21-Jan-2010 15:41:41] <baffle> mchesmo3: Define "really slow". :)
[21-Jan-2010 15:43:01] <mchesmo3> click on something and it takes 20-30 seconds to open something. My real question is...If I like Zenoss, should I use the VM or compile and install
[21-Jan-2010 15:44:37] <mrayzenoss> mchesmo3: does your VM have plenty of memory allocated to it?
[21-Jan-2010 15:44:40] <baffle> mchesmo3: Install the stack installer on a Ubuntu or RH/CentOS dedicated server with fast I/O and lots of gigs of memory.
[21-Jan-2010 15:44:52] <mrayzenoss> the 2.5 VM needs 1.5 gigs
[21-Jan-2010 15:44:56] <baffle> mchesmo3: No need to compile anything.
[21-Jan-2010 15:45:14] <baffle> mrayzenoss: In the real world it probably needs more. :)
[21-Jan-2010 15:45:33] <mrayzenoss> baffle: that's just what the VM is configured for that we deliver
[21-Jan-2010 15:45:37] <baffle> Like, I've seen zenmodeller eating 10G+ .. :)
[21-Jan-2010 15:45:41] <mrayzenoss> yeah, real world is more
[21-Jan-2010 15:49:29] <forsberg> whats tehe difference between stack, and native package on centos?
[21-Jan-2010 15:49:52] <mrayzenoss> forsberg: native package is an RPM
[21-Jan-2010 15:50:35] <mrayzenoss> forsberg: stack is a .bin with everything needed for Zenoss (Python, MySQL, libraries, Zenoss)
[21-Jan-2010 15:51:29] <forsberg> yes ok, but i was just thinking if there were any effectice difference compared to the rpm
[21-Jan-2010 15:51:51] <forsberg> effective
[21-Jan-2010 15:52:18] <baffle> mrayzenoss: Isn't native packages also called stack? I.e. zenoss-stack in Ubuntu (debian).
[21-Jan-2010 15:52:51] <mrayzenoss> the ubuntu ones are "native stacks", stack installers wrapped in .debs
[21-Jan-2010 15:53:06] <forsberg> maybe what im asking is, should is stack installer instead of rpm :) and why hehe
[21-Jan-2010 15:53:23] <forsberg> should i use stack, geeze my keyboard is broken
[21-Jan-2010 15:53:28] <mrayzenoss> Stack installer is brain-dead easy and upgrades are simple
[21-Jan-2010 15:53:40] <forsberg> i find it very simple on rpm also
[21-Jan-2010 15:53:41] <mrayzenoss> RPM is used by most customers and most-tested artifact
[21-Jan-2010 15:53:46] <forsberg> k
[21-Jan-2010 15:53:58] <mrayzenoss> VMware appliance is flakiest thing we provide
[21-Jan-2010 15:54:23] <mrayzenoss> since it's completely different from everything else
[21-Jan-2010 15:54:32] <forsberg> maybe my zenoss wouldnt break when i upgrade, if i use stack :P
[21-Jan-2010 15:54:43] <mrayzenoss> rmatte loves the stack installer
[21-Jan-2010 15:54:51] <forsberg> and i love rmatte
[21-Jan-2010 15:54:52] <forsberg> or wait
[21-Jan-2010 15:55:09] <forsberg> ;(
[21-Jan-2010 15:55:23] <mrayzenoss> don't worry, we archive the channel monthly :p
[21-Jan-2010 15:55:31] <forsberg> :[
[21-Jan-2010 15:55:54] <baffle> forsberg: So now that is searchable till the end of time. Or Google.
[21-Jan-2010 15:56:15] <forsberg> my keyboard is broke!
[21-Jan-2010 15:56:24] <mrayzenoss> heh
[21-Jan-2010 15:56:25] <forsberg> i ment something completely different
[21-Jan-2010 16:02:01] * rmatte is flattered
[21-Jan-2010 16:02:05] <rmatte> :P
[21-Jan-2010 16:03:08] <forsberg> ;)
[21-Jan-2010 16:20:31] <mchesmo3> Thanks all...I will play wth the VM then if we decide to move forward I will porbally use CentOS
[21-Jan-2010 16:20:34] <mchesmo3> Night
[21-Jan-2010 16:21:31] <forsberg> i can recomend that mchesmo3
[21-Jan-2010 16:21:38] <forsberg> takes 10 mins to have centos with zenoss running
[21-Jan-2010 16:22:21] <mchesmo3> Might take me a bit longer... I have little CentOS experiende..We are a FreeBSD shop...
[21-Jan-2010 16:22:43] <rmatte> CentOS is pretty braindead simple
[21-Jan-2010 16:22:57] <rmatte> A lot simpler than FreeBSD, I can tell you that for a fact
[21-Jan-2010 16:22:58] <rmatte> ;)
[21-Jan-2010 16:23:30] <mchesmo3> We get along with BSD pretty well for the most part...But our Network guys swear by CentOS
[21-Jan-2010 16:23:33] <forsberg> thats for sure, i <3 freebsd tho, but work uses centos
[21-Jan-2010 16:23:52] <rmatte> mchesmo3: CentOS is just simpler in the sense that you get a gui with a sort of setup wizard right off the bat
[21-Jan-2010 16:24:18] <rmatte> and it boots with genkernel so you don't have to touch the kernel (FreeBSD does too, but I have a habbit of customizing my kernel in BSD)
[21-Jan-2010 16:24:26] <forsberg> mchesmo3 just holla me when, ill help you make a nice minimal centos install with zenoss :)
[21-Jan-2010 16:24:50] <mchesmo3> forsberg....thanks man....this is why IRC works....
[21-Jan-2010 16:24:52] <mrayzenoss> forsberg: is there a bare-bones CentOS VM from CentOS?
[21-Jan-2010 16:25:14] <forsberg> i dont know, i just the netinstall, and remove _everything_
[21-Jan-2010 16:25:23] <forsberg> +use
[21-Jan-2010 16:25:33] <rmatte> http://www.vmware.com/appliances/directory/1029
[21-Jan-2010 16:25:40] <forsberg> so i dont get printservers and what not
[21-Jan-2010 16:25:47] <rmatte> that may be an ancient version of CentOS though
[21-Jan-2010 16:25:49] <forsberg> the next next next install is lame :P
[21-Jan-2010 16:26:05] <rmatte> ah nope, that's new enough
[21-Jan-2010 16:26:50] <rmatte> we're an Ubuntu shop though, so meh
[21-Jan-2010 16:36:12] <rmatte> yay, the day is almost over
[21-Jan-2010 16:39:55] <baffle> rmatte: Been tracing my SNMP problem; Not sure what it is yet. I've been doing monitor sessions on all ports from Zenoss -> Routers -> Switch-uplinks -> Bladecenter Switch Uplinks -> Blade Switchports now. :) And when I mirror the traffic going out on the switchport to the blade I see traffic on my monitor station but *not* on the server (blade) itself. How strange is *that*?
[21-Jan-2010 16:40:10] <baffle> rmatte: I.e. tcpdump doesn't output anything on the host itself..
[21-Jan-2010 16:40:36] <baffle> Just brainstorming here.
[21-Jan-2010 16:41:36] <baffle> rmatte: Hmm, my day was supposedly over 7 hours ago. :)
[21-Jan-2010 16:41:42] <forsberg> same here :P
[21-Jan-2010 16:41:48] <forsberg> cet timezone baffle? :P
[21-Jan-2010 16:42:08] <baffle> forsberg: Yes. Or +1 maybe.
[21-Jan-2010 16:42:12] <forsberg> actually now it starts, wife and baby asleep
[21-Jan-2010 16:42:14] <rmatte> baffle: lol
[21-Jan-2010 16:42:45] <baffle> forsberg: By the looks of it, the wife is about to fall asleep aswell..
[21-Jan-2010 16:43:30] <baffle> forsberg: Probably from catching a glimps of my screen.
[21-Jan-2010 16:43:50] <forsberg> ^^
[21-Jan-2010 16:44:46] <rmatte> lol
[21-Jan-2010 16:45:39] <forsberg> thats because you need http://media1.break.com/dnet/media/2010/1/55%20Finally!.jpg
[21-Jan-2010 16:46:49] <rmatte> lmao
[21-Jan-2010 16:46:59] <forsberg> mmmm :)))
[21-Jan-2010 16:47:22] <baffle> Me wants it.
[21-Jan-2010 16:47:33] <forsberg> uber for coding late night zenoss hax0rs
[21-Jan-2010 16:47:44] <rmatte> yeh, if you're not prone to seizures
[21-Jan-2010 16:48:01] <baffle> But why oh why is this not working.. Gaaaah.
[21-Jan-2010 16:48:31] <forsberg> you can ping something but not walk it ?
[21-Jan-2010 16:48:32] <baffle> It just has to be what is wrong. :) But I can't for the life of me figure out why it fails.
[21-Jan-2010 16:50:40] <baffle> Ho ho ho!
[21-Jan-2010 16:50:48] <baffle> It is a bond-issue!
[21-Jan-2010 16:50:58] <rmatte> like 007 bond?
[21-Jan-2010 16:51:46] <baffle> rmatte: Like in bonding/teaming/port-channel
[21-Jan-2010 16:51:54] <rmatte> I know, kidding
[21-Jan-2010 16:52:09] <rmatte> ;)
[21-Jan-2010 16:52:15] <baffle> rmatte: No you weren't.. ;)
[21-Jan-2010 16:52:24] <rmatte> lol
[21-Jan-2010 16:52:31] <baffle> rmatte: But again; Hau strange.
[21-Jan-2010 16:53:37] <baffle> Bond seems to be active/passive. eth0 is active, eth1 is passive. If I tcpdump eth0 the UDP packets stop arriving after a short while. If I tcpdump eth1 I see nothing. If I tcpdump bond0 I see the packets, but they don't seem to reach the rest of the OS.
[21-Jan-2010 16:53:49] <baffle> Yes.. Seems to be mode=1 bond.. So active/passive..
[21-Jan-2010 16:53:59] <baffle> eth0/eth1 doesn't share IPs.. That's strange..
[21-Jan-2010 16:54:06] <baffle> Err, doesn't share MACs I mean.
[21-Jan-2010 16:54:32] <rmatte> that is weird
[21-Jan-2010 17:05:07] <baffle> rmatte: After much debugging, it seems it actually might be snmpd that stops responding for too long due to polling /proc too much, this snmpget/walk times out and it doesn't handly any requests while it is doing /proc probing. Probably it is some parts of the network-mib.
[21-Jan-2010 17:08:14] <forsberg> timeout > 15 secs ;>

[21-Jan-2010 17:08:16] <forsberg> hehe
[21-Jan-2010 17:08:41] <baffle> Seems like it might be network connections etc. Ugh.
[21-Jan-2010 17:08:44] <forsberg> used an hour on that today, had 2 identical servers, one just didnt draw half the graphs, grrrrr :P
[21-Jan-2010 17:09:00] <baffle> Well, it is the webservers of norways largest website (a social network)
[21-Jan-2010 17:24:14] <tuvyz> my snmp collector plugin times out on a snmpv3 device. how can i debug it (see the actual command and/or snmp parameters it issues)?
[21-Jan-2010 17:29:59] <baffle> tuvyz: tcpdump?
[21-Jan-2010 17:30:59] <mrayzenoss> look at debug information in zenperfsnmp?
[21-Jan-2010 17:33:58] <tuvyz> i removed some of the collector plugins and i'm getting some collected data now. could it be that i'm using too many collector plugins for the collector machine to handle? it's a pentium m @ 800 MHz and 512 MB ram
[21-Jan-2010 17:34:21] <mrayzenoss> tuvyz: yes, that machine will be crushed by Zenoss
[21-Jan-2010 17:34:35] <mrayzenoss> needs at least a gig of RAM
[21-Jan-2010 17:34:49] <mrayzenoss> you can probably do 15 devices with that machine with more RAM
[21-Jan-2010 17:35:25] <mrayzenoss> I run on several machines with 1 to 2 gigs of RAM at home
[21-Jan-2010 17:35:34] <rocket> grrr just call me the skype bomb ... :p
[21-Jan-2010 17:35:50] <mrayzenoss> I think with my 12 devices they all hover around 800 megs of RAM used by Zenoss
[21-Jan-2010 17:37:29] <tuvyz> mrayzenoss: well this is my second device and the zenoss machine seems to be stressed only during modeling, otherwise mem and cpu usage is typical
[21-Jan-2010 17:37:57] <mrayzenoss> just warning you you don't have much headroom :)
[21-Jan-2010 17:41:03] <tuvyz> btw, is there an snmp lmSensors plugin for zenoss (there is a net-snmp lmsensors plugin, so it can be collected via snmp)
[21-Jan-2010 17:42:01] <mrayzenoss> no, but you could easily wrap a Nagios plugin with a command data source: http://freshmeat.net/projects/check_lm_sensors/
[21-Jan-2010 17:42:18] <cgibbons> hurm
[21-Jan-2010 17:43:01] <tuvyz> no commands for me.. just snmp :)
[21-Jan-2010 17:45:38] <mrayzenoss> if the lmsensors are exposed via SNMP, you could easily add a custom SNMP data source
[21-Jan-2010 17:47:20] <tuvyz> mrayzenoss: they are, and i collect them with cacti already. got a link to docs for adding a custom snmp data source?
[21-Jan-2010 17:48:00] <mrayzenoss> tuvyz: in fact, I was just writing a script for a video howto
[21-Jan-2010 17:48:07] <mrayzenoss> but yeah, it's in the Admin Guide
[21-Jan-2010 17:48:37] <mrayzenoss> Chapter 6.2, http://community.zenoss.org/docs/DOC-3908
[21-Jan-2010 17:51:56] <mrayzenoss> later
[21-Jan-2010 17:53:27] <tuvyz> thank mrayzenoss
[21-Jan-2010 18:06:41] <baffle> Gaaah, after all the changes I still get gaps..
[21-Jan-2010 18:07:14] <baffle> Someone should get around to writing a collector plugin for lmsensors..
[21-Jan-2010 20:03:44] <i-moet> Hai all. Need help, where i can find info about translation?
[21-Jan-2010 20:03:48] <i-moet> or language packs ?
[21-Jan-2010 21:54:05] <tuvyz> i create a service with a meaningful name, e.g. gkrellmd. i add it as an ipservice to a device but it is not added and i get a message to the effect that the service does not exist, although it poped up when i entered the first two letters. i go back to services and change the name of the service to tcp_19150 - bingo - it adds to devices. change the name back to gkrellmd. no add. change it back to tcp_19150. it adds. every single time. what's the deal?!
[21-Jan-2010 23:45:47] tuvyz is now known as tuv
[21-Jan-2010 23:47:27] <pi2t> anyone still wake up?
[22-Jan-2010 00:00:30] [disconnected at Fri Jan 22 00:00:30 2010]
[22-Jan-2010 00:00:30] [connected at Fri Jan 22 00:00:30 2010]
[22-Jan-2010 07:33:25] <ckrough> anyone know what font Zenoss uses in the logo?
[22-Jan-2010 07:41:16] <ckrough> anyone know what font Zenoss uses in the logo?
[22-Jan-2010 08:00:34] <rmatte> good morning all
[22-Jan-2010 08:03:16] <rmatte> ckrough: you can try this: http://new.myfonts.com/WhatTheFont/
[22-Jan-2010 08:03:24] <rmatte> it's overloaded right now apparently but eventually it'll work
[22-Jan-2010 08:11:56] <rmatte> http://failblog.files.wordpress.com/2010/01/test-fix-2.jpg?w=500&h=388
[22-Jan-2010 08:12:03] <rmatte> a little bit of humour for the morning
[22-Jan-2010 08:14:26] <rmatte> ckrough: think I might have found the font...
[22-Jan-2010 08:14:33] <rmatte> ckrough: I think it's Presicav Regular
[22-Jan-2010 08:14:48] <rmatte> but it looks like they manually made the edges on the letters angled
[22-Jan-2010 08:15:31] <rmatte> but yeh, it seems to match
[22-Jan-2010 08:15:33] <rmatte> http://new.myfonts.com/fonts/typodermic/presicav/regular/
[22-Jan-2010 08:16:42] <baffle> rmatte: Good morning. Sigh, I still have fucked graphs on a lot of devices..
[22-Jan-2010 08:17:14] <rmatte> baffle: yeh, I honestly don't know at this point, last thing I'd suggest is trying a reinstall if possible
[22-Jan-2010 08:17:29] <rmatte> but that's pretty extreme
[22-Jan-2010 08:18:29] <baffle> rmatte: And now I've changed snmpd.conf on all devices (to disable the tcpConnectionTable) (what triggered the timeout earlier), I've disabled syslog-logging (wich is default in RH/Ubuntu etc), I've mirrored all ports all the way to see if there is any packetloss.. Nothing.. :)
[22-Jan-2010 08:18:49] <baffle> rmatte: I've reinstalled before, and upgraded between servers too. :)
[22-Jan-2010 08:19:13] <rmatte> how many devices are you monitoring?
[22-Jan-2010 08:19:19] <rmatte> I know you have like 21000 datapoints
[22-Jan-2010 08:19:24] <rmatte> but what's the number of devices?
[22-Jan-2010 08:21:11] <baffle> rmatte: 396 I think.
[22-Jan-2010 08:22:07] <rmatte> ah, not much more than my largest install
[22-Jan-2010 08:22:57] <rmatte> you tried those performance tweaks?
[22-Jan-2010 08:23:05] <baffle> rmatte: Yes, all performance tweaks are in.
[22-Jan-2010 08:24:43] <baffle> Does ZenOss do one-off consulting? :)
[22-Jan-2010 08:25:41] <rmatte> they do do PS work
[22-Jan-2010 08:25:44] <cparlette> it's not out of the question, email sales@zenoss.com with your request
[22-Jan-2010 08:25:46] <rmatte> including that
[22-Jan-2010 08:25:55] <rmatte> It wouldn't be overly cheap though
[22-Jan-2010 08:26:05] <baffle> rmatte: No, ofcourse not. :)
[22-Jan-2010 08:26:08] <rmatte> lol
[22-Jan-2010 08:26:31] <rmatte> If you do go that route and they figure out what's wrong I'd be very very interested to know
[22-Jan-2010 08:26:45] <rmatte> Yours is the most obscure Zenoss problem I've ever seen
[22-Jan-2010 08:26:51] <cparlette> i'm not up to speed on your problem, got a cliffs notes version?
[22-Jan-2010 08:27:07] <rmatte> cparlette: cliffs notes version...
[22-Jan-2010 08:27:41] <rmatte> System specs: 20GB RAM, I forget how much processing power, but high, RAID 8 or 10 drive array...
[22-Jan-2010 08:27:58] <rmatte> Problem: He consistently has huge gaps in his performance graphs
[22-Jan-2010 08:28:14] <rmatte> he is monitoring 396 devices at the same location (datacenter)
[22-Jan-2010 08:28:18] <rmatte> about 21000 datapoints
[22-Jan-2010 08:28:24] <rmatte> What we've tried:
[22-Jan-2010 08:28:50] <rmatte> adjusting every timeout setting we could think of, dropping zMaxOIDPerRequest from 40 to 20 just in case firewalls were picking off oversized packets
[22-Jan-2010 08:29:00] <rmatte> he also tried the performance tweaks listed in the Wiki
[22-Jan-2010 08:29:16] <rmatte> He's checked hardware including disk io and network in pretty good detail
[22-Jan-2010 08:29:26] <cparlette> are all/most devices SNMP?
[22-Jan-2010 08:29:26] <rmatte> still can't figure out what's causing the loss of performance data
[22-Jan-2010 08:29:31] <rmatte> yes
[22-Jan-2010 08:29:34] <baffle> cparlette: It's all SNMP.
[22-Jan-2010 08:29:49] <baffle> cparlette: Or, I don't have gaps in anything not polled via SNMP.
[22-Jan-2010 08:29:56] <baffle> cparlette: I.e. things polled via ssh or simmilar.
[22-Jan-2010 08:30:19] <baffle> cparlette: Server is directly connected to the internet, only local firewall.
[22-Jan-2010 08:30:34] <cparlette> whats the cycle time on zenperfsnmp?
[22-Jan-2010 08:30:44] <cparlette> i'm guessing it's way over 300
[22-Jan-2010 08:30:58] <baffle> cparlette: That is in zenperfsnmp.log, right?
[22-Jan-2010 08:31:06] <rmatte> cparlette: you talking about the setting in zproperties or on the collector itself?
[22-Jan-2010 08:31:17] <cparlette> that, or go to the collector and look at the performance tab
[22-Jan-2010 08:31:28] <rmatte> oh
[22-Jan-2010 08:31:30] <rmatte> no it's not
[22-Jan-2010 08:31:41] <rmatte> he actually has very reasonable cycle times for zenperfsnmp
[22-Jan-2010 08:31:44] <cparlette> did you change that cycle time?
[22-Jan-2010 08:31:48] <rmatte> (we checked that yesterday)
[22-Jan-2010 08:31:49] <baffle> 2010-01-22 15:28:04,737 INFO zen.zenperfsnmp: ******** Cycle completed ********
[22-Jan-2010 08:31:49] <baffle> 2010-01-22 15:28:04,737 INFO zen.zenperfsnmp: Sent 21461 OID requests
[22-Jan-2010 08:31:49] <baffle> 2010-01-22 15:28:04,737 INFO zen.zenperfsnmp: Queried 367 devices
[22-Jan-2010 08:31:49] <baffle> 2010-01-22 15:28:04,737 INFO zen.zenperfsnmp: 0 in queue still unqueried
[22-Jan-2010 08:31:49] <baffle> 2010-01-22 15:28:04,738 INFO zen.zenperfsnmp: Successes: 361 Failures: 6 Not reporting: 0
[22-Jan-2010 08:31:52] <baffle> 2010-01-22 15:28:04,738 INFO zen.zenperfsnmp: Waited on 0 queries from previous cycles.
[22-Jan-2010 08:31:55] <baffle> 2010-01-22 15:28:04,738 INFO zen.zenperfsnmp: Successes: 0 Failures: 0 Not reporting: 0
[22-Jan-2010 08:31:57] <rmatte> we didn't touch any cycle time settings
[22-Jan-2010 08:31:58] <baffle> 2010-01-22 15:28:04,738 INFO zen.zenperfsnmp: Cycle lasted 84.32 seconds
[22-Jan-2010 08:32:00] <baffle> 2010-01-22 15:28:04,738 INFO zen.zenperfsnmp: *********************************
[22-Jan-2010 08:32:10] <rmatte> they should all be default
[22-Jan-2010 08:32:39] <rmatte> baffle: under overview for the localhost collector plugin...
[22-Jan-2010 08:32:43] <rmatte> is SNMP Performance Cycle Interval (secs) set to 300?
[22-Jan-2010 08:32:48] <cparlette> also something to keep in mind is that some settings you changed may require a deletion of the rrd files
[22-Jan-2010 08:33:15] <baffle> rmatte: Yes, it is set to 300.
[22-Jan-2010 08:33:17] <rmatte> cparlette: that would mean that he'd get no data at all, not sporadric data, wouldn't it?
[22-Jan-2010 08:33:17] <cparlette> i'm surprised at 84 seconds, thats fast
[22-Jan-2010 08:33:28] <rmatte> yeh, 84 seconds is fast
[22-Jan-2010 08:33:35] <cparlette> no, it would just condense the intervals in a weird way
[22-Jan-2010 08:33:40] <cparlette> which can cause gaps
[22-Jan-2010 08:33:47] <rmatte> I see
[22-Jan-2010 08:33:51] <cparlette> maybe try deleting 1/a few rrd's
[22-Jan-2010 08:33:54] <cparlette> or rename them temporarily
[22-Jan-2010 08:34:00] <baffle> cparlette: I've deleted a lot of RRD files without success.
[22-Jan-2010 08:34:13] <rmatte> cparlette: well, he could just copy his /perf/Devices directory
[22-Jan-2010 08:34:18] <rmatte> then go in and delete everything
[22-Jan-2010 08:34:29] <rmatte> If it doesn't fix his problem he could just move everything back
[22-Jan-2010 08:34:35] <cparlette> right
[22-Jan-2010 08:35:03] <cparlette> are the gaps huge, or just a bunch of tiny ones
[22-Jan-2010 08:35:11] <baffle> cparlette: Tiny, usually.
[22-Jan-2010 08:35:18] <rmatte> cparlette: also, 84 seconds is honestly not that fast, I collect 18000 datapoints on one of my servers in an average of 22.47 seconds
[22-Jan-2010 08:35:40] <rmatte> but 84 seconds is very very reasonable
[22-Jan-2010 08:36:30] <cparlette> is zenhub getting hammered?
[22-Jan-2010 08:36:39] <baffle> cparlette: Yes, ZenHub is a bit hammered.
[22-Jan-2010 08:37:12] <baffle> cparlette: Or, well, up and down really.
[22-Jan-2010 08:37:33] <baffle> cparlette: The Zenhub problem got a bit better after I upgraded the MySQL server.
[22-Jan-2010 08:38:10] <baffle> cparlette: (I always had a separate mysql, but before it was a 5 year old pizzabox, now it has 16G + 15K FC SAN + lots of CPU)
[22-Jan-2010 08:38:40] <baffle> cparlette: I do get some python tracebacks in zenperfsnmp, tho.
[22-Jan-2010 08:39:17] <cparlette> hm, can i see one?
[22-Jan-2010 08:39:20] <cparlette> in a pastebin
[22-Jan-2010 08:39:29] <baffle> cparlette: If zenperfsnmp hits a snag/is unable to parse an OID and does a traceback; Will it drop reporting other data it polls?
[22-Jan-2010 08:39:57] <rmatte> baffle: not in 2.5
[22-Jan-2010 08:40:00] <rmatte> it used to do that
[22-Jan-2010 08:40:27] <Abjo> Hey, someone can help me to setup a collector in zenoss ? i already started but now i'm out of idea's
[22-Jan-2010 08:41:05] <rmatte> Abjo: We can try to help but I don't think many of us have actually setup remote collectors
[22-Jan-2010 08:41:14] <rmatte> Abjo: What are you stuck on?
[22-Jan-2010 08:41:29] <baffle> cparlette: http://paste2.org/p/625828
[22-Jan-2010 08:41:39] <baffle> rmatte: Maybe it still does. That would explain.
[22-Jan-2010 08:41:48] <rmatte> baffle: ohhhh
[22-Jan-2010 08:41:51] <rmatte> that's different
[22-Jan-2010 08:41:56] <rmatte> I thought you meant something else
[22-Jan-2010 08:42:37] <rmatte> what the heck data are you polling that's returning "OFF" as a value?
[22-Jan-2010 08:42:58] <rmatte> what's happening is that Zenoss is expecting an integer
[22-Jan-2010 08:43:04] <rmatte> but the OID you're polling is returning a string
[22-Jan-2010 08:43:07] <rmatte> so it's choking
[22-Jan-2010 08:43:16] <Abjo> did i only have to copy the zenpacks from my main to the collector in the right directory or i have to do something more ?
[22-Jan-2010 08:43:27] <rmatte> Figure out what .1.3.6.1.4.1.27058.2.1.3.8 is for and remove that datapoint from the template
[22-Jan-2010 08:43:29] <rmatte> or fix it
[22-Jan-2010 08:44:17] <baffle> rmatte: It is stupid Netapps and other stuff that change from floats to string after updates. Yay. :)
[22-Jan-2010 08:44:29] <rmatte> baffle: Data was value= 26C (78F) ambient, <N/A>, 37C (98F)
[22-Jan-2010 08:44:34] <rmatte> that one is horrible too
[22-Jan-2010 08:44:55] <rmatte> Zenoss tries to store that in an RRD and chokes
[22-Jan-2010 08:45:04] <rmatte> Guess you found your problem
[22-Jan-2010 08:45:47] <ckrough> rmatte: thanks
[22-Jan-2010 08:46:02] <ckrough> rmatte: need to start making some custom logos. too many zenoss installs
[22-Jan-2010 08:48:04] <rmatte> hehe
[22-Jan-2010 08:48:05] <rmatte> no problem
[22-Jan-2010 08:49:07] <mchesmo3> I think the popular thought is to run Zenoss on CentOS but what would anyone think about runnng it on Suse....I have a box with lots of Ram and horsepower that is Suse already..
[22-Jan-2010 08:50:20] <ckrough> I think it's less work to reformat to RHEL/CentOS than it is to have to figure out the unique problems you could experience. Assuming it's a dedicated box and you know RHEL/CentOS
[22-Jan-2010 08:51:13] <ckrough> as a general rule I try to stick with what the community and/or devs use... community support is just better when you follow the herd on some things
[22-Jan-2010 08:54:36] <mchesmo3> I am pretty new to Zenoss but not to Linux/Unix We currently use MRTG, RRD and Nagios ....how does Zenoss compare ?
[22-Jan-2010 08:56:03] <baffle> rmatte: Yes, maybe. I'll report findings. :)
[22-Jan-2010 08:56:19] <rmatte> mchesmo3: it's much easier to use than Nagios
[22-Jan-2010 08:56:38] <rmatte> mchesmo3: and it'll do anything Nagios can, it even supports Nagios plugins, and it uses RRD
[22-Jan-2010 08:56:58] <mchesmo3> Sounds great...
[22-Jan-2010 08:57:11] <rmatte> honestly, it'll replace all the tools you're currently using in 1 package
[22-Jan-2010 08:57:29] <mchesmo3> Does it do all that with the free version or just the paid for one?
[22-Jan-2010 08:57:43] <rmatte> the free version does a lot
[22-Jan-2010 08:58:13] <rmatte> the main difference between the core and enterprise versions are that core doesn't do full WMI monitoring, only partial.
[22-Jan-2010 08:58:56] <rmatte> setting up multiple collectors is much easier on enterprise as well
[22-Jan-2010 08:59:08] <rmatte> and with enterprise you have access to some deluxe enterprise exclusive zenpacks
[22-Jan-2010 08:59:11] <rmatte> and the support obviously
[22-Jan-2010 08:59:16] <cgibbons> mchesmo3: for perspective, the original author of zenoss wrote it to make configuring nagios easier.
[22-Jan-2010 08:59:19] <rmatte> but core does do a heck of a lot
[22-Jan-2010 08:59:32] <mchesmo3> That was my next Q.....we have some budget for a system like this but we have gotten pretty comfortable with Nagios and MRTG
[22-Jan-2010 08:59:49] <mchesmo3> Does it scale to enterprise level well
[22-Jan-2010 08:59:50] <rmatte> mchesmo3: I assume your monitoring is pretty much all snmp based at the moment?
[22-Jan-2010 09:00:02] <rmatte> and define "enterprise level"
[22-Jan-2010 09:00:08] <rmatte> how many devices, and what types?
[22-Jan-2010 09:01:30] <mchesmo3> 5-10K devices, just about anything except desktops...All Cisco, windows and Unix servers, web app servers Mail, etc.
[22-Jan-2010 09:01:48] <rmatte> for 5-10K you would definitely want the enterprise version, and yes it could handle that
[22-Jan-2010 09:02:18] <rmatte> You could potentially do it with core but it would be a LOT of development work to get the remote collectors all setup and everything
[22-Jan-2010 09:02:37] <mchesmo3> does it do anything with logging. we use OSSEC currently...
[22-Jan-2010 09:03:12] <rmatte> it can handle traps and syslog, but it's syslog "maximum" is 5000 events per second
[22-Jan-2010 09:03:32] <rmatte> we use syslog-ng between Zenoss and the devices to redirect security logs elsewhere
[22-Jan-2010 09:04:19] <mchesmo3> thats cool we have a good solution in place for syslogs now..we use Syslog-ng into a syslog stream then split them back out with perl scripts
[22-Jan-2010 09:04:39] <mchesmo3> OK i am off to build a CentOS box.
[22-Jan-2010 09:04:43] <rmatte> have fun
[22-Jan-2010 09:05:10] <mchesmo3> thanks for the advice...Mike
[22-Jan-2010 09:05:33] <baffle> rmatte: Yes, NetApp and IBM Bladecenters etc seems to have changed their reporting. And are now reporting strings instead of proper values. How idiotic is that..
[22-Jan-2010 09:06:29] <rmatte> baffle: I hate stupid vendors who mess with SNMP
[22-Jan-2010 09:06:35] <rmatte> if it's not broken don't "fix" the damn thing
[22-Jan-2010 09:06:36] <rmatte> lol
[22-Jan-2010 09:06:51] <rmatte> If Cisco ever did that they'd have a firestorm on their hands
[22-Jan-2010 09:09:19] <baffle> rmatte: Not impressed about errorhandling, tho'. How about a try: save() except: log.error('Unable to save $oid in $rrd') or something. :)
[22-Jan-2010 09:09:38] <baffle> rmatte: If that turns out to be the problem. :)
[22-Jan-2010 09:09:49] <rmatte> I'm almost certain that it is
[22-Jan-2010 09:10:14] <rmatte> baffle: I do agree with the error handling though...
[22-Jan-2010 09:10:18] <rmatte> open a trac ticket for it
[22-Jan-2010 09:10:26] <rmatte> http://dev.zenoss.com/trac
[22-Jan-2010 09:10:30] <rmatte> username: zenoss
[22-Jan-2010 09:10:33] <rmatte> password: zenoss
[22-Jan-2010 09:10:58] <rmatte> (You can ask mrayzenoss for your own Trac account)
[22-Jan-2010 09:11:02] <cgibbons> and provide a patch if you really want it fixed :)
[22-Jan-2010 09:11:13] <rmatte> hehe
[22-Jan-2010 09:13:17] <rmatte> aloha rocket
[22-Jan-2010 09:15:18] <rocket> Hey rmatte
[22-Jan-2010 09:19:07] <ckrough> hmm.. I created a custom copy of the logo in Zope but it's still showing the old one
[22-Jan-2010 09:19:41] <baffle> cgibbons: True, true. :)
[22-Jan-2010 09:19:46] <rmatte> ckrough: restart Zenoss? it's probably cached
[22-Jan-2010 09:20:07] <baffle> cgibbons: Can you create a Trac account or only mray?
[22-Jan-2010 09:20:11] <ckrough> restarted zopectl
[22-Jan-2010 09:20:29] <rmatte> baffle: it's really Matt Ray's job to create Trac accounts lol
[22-Jan-2010 09:20:45] <rmatte> ckrough: you use zec files?
[22-Jan-2010 09:20:59] <cgibbons> my privs to do so were likely taken away, anyway... we had too many of us changing things in trac unintentionally
[22-Jan-2010 09:21:09] <ckrough> rmatte: nay
[22-Jan-2010 09:21:20] <rmatte> "whoops, there goes Matt Ray's account *whistle*"
[22-Jan-2010 09:21:28] <baffle> cgibbons: Wow, Zenoss going corporate then. Have to have access controls, and not everyone is root. :)
[22-Jan-2010 09:21:44] <cgibbons> a very good thing
[22-Jan-2010 09:22:14] <rmatte> well, our guys have "root" but it's via sudo and their ldap account
[22-Jan-2010 09:22:26] <rmatte> so if something gets messed up a quick look at the logs shows who did it
[22-Jan-2010 09:22:43] <rmatte> and by "our guys" I mean my development team
[22-Jan-2010 09:23:19] <baffle> I get a lot of "ERROR netsnmp: truncating signed value to 32 bits (1)" but I guess that isn't a problem?
[22-Jan-2010 09:25:51] <rmatte> hmmm
[22-Jan-2010 09:26:00] <rmatte> shouldn't be
[22-Jan-2010 09:26:04] <rmatte> might throw some values off
[22-Jan-2010 09:33:51] <baffle> I made a ticket. http://dev.zenoss.com/trac/ticket/6087
[22-Jan-2010 09:34:33] <rmatte> haha
[22-Jan-2010 09:34:38] <rmatte> I love the steps to reproduce
[22-Jan-2010 09:35:03] <rmatte> Impact is good too lol
[22-Jan-2010 09:35:08] <baffle> :-)
[22-Jan-2010 09:35:23] <baffle> Impact was initiall "Dead baby seals", but I changed it.
[22-Jan-2010 09:36:36] <rmatte> hehe
[22-Jan-2010 09:36:41] <rmatte> sticking with the marine theme
[22-Jan-2010 09:37:46] <rmatte> "due to moronic vendors"
[22-Jan-2010 09:37:50] <rmatte> love it
[22-Jan-2010 09:40:39] <baffle> rmatte: It's funny because it is true.
[22-Jan-2010 09:40:45] <rmatte> yup
[22-Jan-2010 09:44:32] <baffle> Hmm, does the "step" in a RRD define the polling intervall?
[22-Jan-2010 09:44:52] <rmatte> I do believe so
[22-Jan-2010 09:45:51] <mrayzenoss> I believe it's the other way. Polling interval defines step
[22-Jan-2010 09:46:56] <baffle> mrayzenoss: I found some RRDs with "step: 100" (from rrdtool info)
[22-Jan-2010 09:47:13] <mrayzenoss> hmmm, ok
[22-Jan-2010 09:47:40] <cgibbons_> hmm wtf is up with my network today
[22-Jan-2010 09:49:02] <luisfqs> Hello, I have zenoss 2.2.0 and net-snmp in rhel clients intel and ppc. In ppc servers zenoss doesn't show file systems, how can I fix this?
[22-Jan-2010 09:49:48] <baffle> Whoooo.. Maybe this is my issue.. Somehow lots of my .rrds have "step = 100".
[22-Jan-2010 09:49:58] <baffle> Maybe it isn't ZenPerfSNMP crashing.
[22-Jan-2010 09:50:32] <rmatte> that's what cparlette was saying it might be
[22-Jan-2010 09:50:43] <rmatte> though I'm sure those non-float values didn't help
[22-Jan-2010 09:51:34] <rmatte> baffle: If you delete those particular RRDs it should regenerate them with the proper values
[22-Jan-2010 09:51:48] <rmatte> or you might be able to do some rrd magic to just fix them all
[22-Jan-2010 09:57:01] cgibbons_ is now known as cgibbons
[22-Jan-2010 10:00:47] <mrayzenoss> rmatte: don't you use OTRS?
[22-Jan-2010 10:01:05] <mrayzenoss> or was it some other ticket system?
[22-Jan-2010 10:01:38] <rmatte> OTRS, yessir
[22-Jan-2010 10:01:53] <mrayzenoss> whenever you feel like doc'ing that, http://dev.zenoss.org/trac/ticket/6086 :)
[22-Jan-2010 10:02:55] <rmatte> well... the problem is that our OTRS is heavily customized with all sorts of custom fields and stuff which our integration uses
[22-Jan-2010 10:03:01] <rmatte> so it most likely wouldn't work for them
[22-Jan-2010 10:03:19] <mrayzenoss> yeah, do you just do a lot of email customizations?
[22-Jan-2010 10:03:40] <rmatte> in terms of generating tickets you mean?
[22-Jan-2010 10:03:46] <mrayzenoss> yeah
[22-Jan-2010 10:03:48] <rmatte> we generate the tickets via the SOAP API
[22-Jan-2010 10:03:51] <mrayzenoss> do you have callbacks?
[22-Jan-2010 10:03:55] <mrayzenoss> oh, ok
[22-Jan-2010 10:03:55] <rmatte> with a perl script that one of the other guys wrote
[22-Jan-2010 10:04:04] <rmatte> and my zenticket daemon calls on that script
[22-Jan-2010 10:04:17] <QubeZ> hey rmatte, thanks for the transform
[22-Jan-2010 10:04:28] <rmatte> QubeZ: no problem, working as expected I hope?
[22-Jan-2010 10:04:50] <QubeZ> going to test later today... right now busy with crap load of asterisk configs :(
[22-Jan-2010 10:05:17] <rmatte> mrayzenoss: I could ask if I could relay the code to you guys to at least give you a guildeline of how we're doing it. Not sure if they'll want to though.
[22-Jan-2010 10:06:23] <mrayzenoss> I'm sure anything you post will be appreciated. The ticket is an Enhancement request, so it's probably just saying we need to make it easier or find some docs for it
[22-Jan-2010 10:06:33] <rmatte> I see
[22-Jan-2010 10:11:01] <luisfqs> Hello, I have zenoss 2.2.0 and net-snmp in rhel servers intel and ppc. In ppc servers zenoss doesn't show file systems, anyone can help me?
[22-Jan-2010 10:12:02] <rmatte> I assume by PPC you mean PowerPC?
[22-Jan-2010 10:12:02] <mrayzenoss> are the snmpd.conf files the same?
[22-Jan-2010 10:12:23] <luisfqs> rmatee: yes, PowerPC (ibm pseries)
[22-Jan-2010 10:12:40] <rmatte> running AIX I'm assuming?
[22-Jan-2010 10:13:00] <luisfqs> no, redhat enterprise 5
[22-Jan-2010 10:13:55] <rmatte> If you do an snmpwalk on the ppc servers for 1.3.6.1.2.1.25.2.3.1.6 do you get anything?
[22-Jan-2010 10:14:46] <rmatte> also, what mrayzenoss said, are you snmpd.conf files the same across the board?
[22-Jan-2010 10:16:33] <luisfqs> rmatte: snmpwalk returns all values with 0 in powerPC
[22-Jan-2010 10:16:44] <luisfqs> my snmpd.conf is the same
[22-Jan-2010 10:17:35] <luisfqs> e.g. HOST-RESOURCES-MIB::hrStorageUsed.1 = INTEGER: 0
[22-Jan-2010 10:18:41] <rmatte> well, that's your issue, net-snmp isn't reporting the values correctly
[22-Jan-2010 10:18:49] <rmatte> not a Zenoss issue
[22-Jan-2010 10:19:24] <luisfqs> ok, thanks for your help.
[22-Jan-2010 10:23:43] <rmatte> np
[22-Jan-2010 11:07:12] <baffle> rmatte: Wow. I think it actually might be wrong step in (a myriad) of RRD files. How that happend is beyond me.. :)
[22-Jan-2010 11:21:38] <rmatte> that is weird, yeh
[22-Jan-2010 11:21:58] <rmatte> you might have to just bite the bullet and kill it all
[22-Jan-2010 12:09:15] <stbain> Greetings all
[22-Jan-2010 12:20:42] <rmatte> hey
[22-Jan-2010 12:28:55] <stbain> OK, so I'm kind of at a buy/build/extend decision point here and was hoping to get some insight.
[22-Jan-2010 12:29:10] <stbain> I manage a stack of servers. Physical, virtual, etc.
[22-Jan-2010 12:29:48] <stbain> Already looking at Zenoss for monitoring and whatnot, but I also need to consider how to go about automating routine tasks for non-console-friendly employees who have to set up new Apache virtual hosts for customers and whatnot
[22-Jan-2010 12:30:03] <stbain> in the past, I've always just used some perl/python/shell scripts to do that stuff from the command line
[22-Jan-2010 12:30:32] <stbain> but I'm looking to roll it into a web interface. I looked at doing it from the ground up with Twisted, but now I'm seeing that I might be better off trying to extend Zenoss functionality. Does that make sense?
[22-Jan-2010 12:30:47] <rmatte> have you not looked at webmin?
[22-Jan-2010 12:30:51] <stbain> I have
[22-Jan-2010 12:30:54] <stbain> even looked at virtualmin
[22-Jan-2010 12:31:01] <stbain> (usermin, etc. etc.)
[22-Jan-2010 12:31:32] <rmatte> I find webmin fairly simple to use
[22-Jan-2010 12:31:46] <stbain> as do I
[22-Jan-2010 12:32:18] <rmatte> I don't personally see why you'd extend Zenoss functionality for that
[22-Jan-2010 12:32:42] <stbain> Because it's not just setting up traditional virtual hosts
[22-Jan-2010 12:32:51] <rmatte> Zenoss is a monitoring solution which is capable of kicking off scripts and what not based on triggers
[22-Jan-2010 12:32:53] <stbain> there are other products, software packages, etc.
[22-Jan-2010 12:32:59] <rmatte> but it wouldn't be the ideal solution for what you're describing
[22-Jan-2010 12:33:24] <stbain> I was looking more to the "management" side of the Zenoss solution than the "monitoring" side
[22-Jan-2010 12:33:39] <rmatte> right, but from an actual management perspective Zenoss is pretty bare bones, everything needs to be built up
[22-Jan-2010 12:33:56] <rmatte> it's much stronger in it's monitoring/alerting
[22-Jan-2010 12:34:04] <rmatte> there's really not much to it in terms of management
[22-Jan-2010 12:34:16] <stbain> sounds like just building ground up with Twisted would be a more sensible solution
[22-Jan-2010 12:34:28] <rmatte> you can use it to see what software you have installed on your servers but you can't use it to trigger installs or uninstalls for instance
[22-Jan-2010 12:34:44] <rmatte> it's not designed for that sort of purpose
[22-Jan-2010 12:34:53] <stbain> right, I would have to write my own jobs within the Zenoss system
[22-Jan-2010 12:34:58] <rmatte> what you're describing sounds a lot more like puppet, chef, or bcfg2
[22-Jan-2010 12:34:59] <stbain> all custom anyways,
[22-Jan-2010 12:35:13] <rmatte> you might want to check those out if you haven't already
[22-Jan-2010 12:35:27] <stbain> looking now
[22-Jan-2010 12:35:29] <etank> are there docs that have recommendations of how to layout zenoss. for instance, how to use Systems and Groups. also stuff like the device templates. there is a /Devices/Server/Windows/WMI/MSSQLServer. what if i have more than one domain. do i do a WMI/MSSQLServer/Dom1 WMI/MSSQLServer/Dom2 to set the zproperties
[22-Jan-2010 12:35:35] <rmatte> puppet is the most popular out of the 3
[22-Jan-2010 12:36:42] <etank> our would /WMI/Dom1/MSSQLServer /WMI/Dom2/MSSQLServer be better?
[22-Jan-2010 12:36:54] <rmatte> etank: there really is no best practices document
[22-Jan-2010 12:37:12] <rmatte> etank: it's really up to how you, and the zenpack developers decide to use the device classes
[22-Jan-2010 12:37:20] <rmatte> as for groups, you can use those however you wish
[22-Jan-2010 12:37:34] <rmatte> we use them to distinguish between different clients on multi-client boxes for instance
[22-Jan-2010 12:37:47] <rmatte> and then at the next layer to distinguish between server, network, security devices
[22-Jan-2010 12:37:52] <etank> or Prod, Test, Dev systems?
[22-Jan-2010 12:38:06] <rmatte> well, no, that would be more device production states
[22-Jan-2010 12:38:12] <rmatte> that's what the production states are there for
[22-Jan-2010 12:38:31] <rmatte> If you go to settings you'll see:
[22-Jan-2010 12:38:36] <rmatte> Production:1000
[22-Jan-2010 12:38:36] <rmatte> Pre-Production:500
[22-Jan-2010 12:38:36] <rmatte> Test:400
[22-Jan-2010 12:38:36] <rmatte> Maintenance:300
[22-Jan-2010 12:38:36] <rmatte> Decommissioned:-1
[22-Jan-2010 12:38:43] <rmatte> you can actually add more in there if you want
[22-Jan-2010 12:38:46] <rmatte> for instance I did:
[22-Jan-2010 12:38:53] <rmatte> Standby:-2
[22-Jan-2010 12:38:54] <rmatte> Spare:-3
[22-Jan-2010 12:40:22] <etank> makes sense. im just having a hard time deciding how to lay everything out.
[22-Jan-2010 12:40:41] <rmatte> everything kind of serves it's own purpose
[22-Jan-2010 12:40:53] <rmatte> device classes control what plugins and templates are assigned to devices
[22-Jan-2010 12:41:02] <rmatte> groups are just general categorization, whatever you want
[22-Jan-2010 12:41:05] <rmatte> systems is the same
[22-Jan-2010 12:41:09] <rmatte> and locations are self explanatory
[22-Jan-2010 13:42:24] <dec3pti0n> how long should take zenoss to see that an OS process is no longer running on the one box I have being monitored ?
[22-Jan-2010 13:47:22] <rmatte> 5 minutes
[22-Jan-2010 13:47:42] <rmatte> or less
[22-Jan-2010 13:48:47] <etank> if i want to extend the amount of time that data points are stored in rrd graphs before they start being averaged out where would i look?
[22-Jan-2010 13:49:51] <dec3pti0n> hmmm i think it has been over 5min now or close to it
[22-Jan-2010 13:50:15] <dec3pti0n> is that within a zproperty that can be changed ?
[22-Jan-2010 14:29:19] <ckrough> etank: in the collector page under edit
[22-Jan-2010 14:29:56] <ckrough> etank: changed the RRA values, but you cant do it to current RRD files. You have to recreate all of them. I think you can export and reimport with the new RRA definitions
[22-Jan-2010 14:30:14] <ckrough> etank: http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[22-Jan-2010 14:31:13] <etank> i see. now i have to figure out what to set them too :)
[22-Jan-2010 14:31:55] <ckrough> I have 1 year at 5min with no consolidation, and 1 month@5min + 11months @ 1hr. if you want them
[22-Jan-2010 14:33:39] <etank> ckrough: sure
[22-Jan-2010 14:34:13] <etank> so rrd does not start averaging until the datapoint is one year old?
[22-Jan-2010 14:34:42] <etank> i need to be able to look back at least 6 months with the data being as sharp as the day it was gathered
[22-Jan-2010 14:37:28] <ckrough> etank: actually, it never averages for some of our collectors
[22-Jan-2010 14:38:00] <ckrough> etank: http://pastebin.com/m78b42fcc
[22-Jan-2010 14:38:24] <ckrough> etank: so the second RRA is fairly pointless since the first line is 300s accuracy for 1 year already
[22-Jan-2010 14:38:45] <ckrough> etank: in the case of that one it "averages" 1 PDP... so technically no averaging at all
[22-Jan-2010 14:39:16] <philfry> does anyone know if its possible to create availability reports in zenoss core? the default availability report is very lacking
[22-Jan-2010 14:39:42] <ckrough> etank: http://pastebin.com/m65ca734b 1 month at 300s, then 11 more months at 1hr
[22-Jan-2010 14:41:34] <baffle> rmatte: Graphs seems to be working quite well now. But I'm noticing that interfaces on Ciscos with a lot of thruput (I.e. over 100M) seems to not work properly. Is this some 32bit vs 64bit counter issue? We have quite a few 10G links that I know there is a lot of traffic on, the graphs seems strange and spotty on "Thruput" while PPS seems to be nice..
[22-Jan-2010 14:42:27] <etank> ckrough: guess i need to understand what RRA:AVERAGE:0.5:1:105120 does
[22-Jan-2010 14:44:59] <etank> ckrough: so the first one never does averaging?
[22-Jan-2010 14:45:26] <ckrough> keep average 1 value , keep 105120 rows (your cycle time... prob 300s samples), allow a maximum of 50% (.5) of the samples (1) to be UNKNOWN
[22-Jan-2010 14:45:32] <ckrough> correct
[22-Jan-2010 14:45:57] <ckrough> its an average of 1 sample devided by 1, which is no averaging at all
[22-Jan-2010 14:45:58] <etank> ok cool
[22-Jan-2010 14:46:08] <etank> i think that is what i will use then
[22-Jan-2010 14:46:11] <rmatte> interesting, using swapoff and swapon to recover swap space on the fly and it's working surprisingly well
[22-Jan-2010 14:46:19] <ckrough> etank: be warned that your RRDs will be big
[22-Jan-2010 14:46:31] <etank> ckrough: about how big?
[22-Jan-2010 14:46:52] <ckrough> etank: sec
[22-Jan-2010 14:47:43] <venturaville> object has no attribute 'getZenPack' ]
[22-Jan-2010 14:47:51] <venturaville> how do I get access to the zenpack path from a modeler plugin?
[22-Jan-2010 14:48:07] <baffle> rmatte: Oh, never tought about that. Good call.
[22-Jan-2010 14:48:30] <rmatte> yeh, it worked well
[22-Jan-2010 14:48:33] <rmatte> just swapoff -a
[22-Jan-2010 14:48:36] <rmatte> then swapon -a
[22-Jan-2010 14:48:39] <rmatte> problem solved
[22-Jan-2010 14:48:45] <rmatte> I just freed up like 800MB of swap
[22-Jan-2010 14:48:45] <ckrough> etank: ~1.4 MB each it looks like
[22-Jan-2010 14:48:59] <etank> and the rrd would be / graph
[22-Jan-2010 14:49:08] <etank> not too bad actually
[22-Jan-2010 14:49:11] <etank> disk is cheap
[22-Jan-2010 14:49:15] <ckrough> rmatte: 'swapoff -a && swapon -a' is what Ive been using
[22-Jan-2010 14:49:25] <rmatte> ckrough: yeh, it works well
[22-Jan-2010 14:49:32] <ckrough> rmatte: got one box giving me troubles, toying with zope cache sizes
[22-Jan-2010 14:49:39] <ckrough> etank: problem isnt disk space it's disk IO
[22-Jan-2010 14:49:40] <rmatte> (I generally use && too since it waits for the previous command to complete successfully)
[22-Jan-2010 14:49:56] <rmatte> ckrough: my Zope is taking up like 50% memory combined
[22-Jan-2010 14:49:57] <ckrough> etank: bigger file = more read and write, which across a lot of RRDs has a major impact
[22-Jan-2010 14:50:00] <rmatte> which is about 3GB
[22-Jan-2010 14:50:08] <rmatte> I think I might have to drop the cache sizes down
[22-Jan-2010 14:50:29] <ckrough> etank: for example, on the exact same box I can do 50% less OIDs per cycle with the more aggressive RRA definition
[22-Jan-2010 14:50:55] <etank> ckrough: so if my goal is to be able to go back 6 months with no averaging this could be over kill
[22-Jan-2010 14:51:08] <ckrough> rmatte: how many zope threads do you start and how many objects cached?
[22-Jan-2010 14:51:15] <ckrough> rmatte: cant find good recommendations for those numbers
[22-Jan-2010 14:51:16] <etank> after 6 months i dont mind it to do some averging
[22-Jan-2010 14:51:27] <ckrough> etank: depends on how far you push your hardware
[22-Jan-2010 14:51:33] <rmatte> let me check
[22-Jan-2010 14:51:42] <ckrough> etank: its probably fine, I just didnt want you cursing me if smoke comes pouring out of the box :)
[22-Jan-2010 14:52:49] <rmatte> ckrough: 50 threads cache-size of 60000 and pool size of 50
[22-Jan-2010 14:53:14] <venturaville> no one knows about the zenpack pathin?
[22-Jan-2010 14:53:17] <venturaville> pathing
[22-Jan-2010 14:53:25] <rmatte> venturaville: hunh?
[22-Jan-2010 14:53:40] <venturaville> rmatte: trying to find out how to get to the zenpack path from within a modeler
[22-Jan-2010 14:53:43] <rmatte> ohhhhh
[22-Jan-2010 14:53:57] <etank> ckrough: im still trying to figure out the math for this
[22-Jan-2010 14:54:45] <rmatte> one second, I should be able to figure that out
[22-Jan-2010 14:55:06] * etank is trying to keep his brain from oozing otu of his hears
[22-Jan-2010 14:55:11] <etank> s/hears/ears
[22-Jan-2010 14:56:03] <rmatte> venturaville: well, you want it to dynamically grab the path but you can obviously define the actual zenpack name in there, right?
[22-Jan-2010 14:56:03] <venturaville> rmatte: fudge ..getting yanked into a meeting. . I'll ask again tomorrow :-)
[22-Jan-2010 14:56:05] <ckrough> etank: read this 5 times: http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[22-Jan-2010 14:56:15] <venturaville> rmatte: yeah I think I see where you are going
[22-Jan-2010 14:56:19] <Apachez> http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[22-Jan-2010 14:56:20] <Apachez> http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[22-Jan-2010 14:56:21] <Apachez> http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[22-Jan-2010 14:56:21] <etank> ckrough: yeah thats what im on right now :)
[22-Jan-2010 14:56:21] <Apachez> http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html
[22-Jan-2010 14:56:26] <rmatte> venturaville: it's really simple...
[22-Jan-2010 14:56:28] <Apachez> now he should have read it 5 times ;)
[22-Jan-2010 14:56:36] <rmatte> here's how you'd do it in a template
[22-Jan-2010 14:56:47] <rmatte> ${here/ZenPackManager/packs/ZenPacks.Nova.Windows.SNMPPerfMonitor/path}/libexec/wincpu.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"
[22-Jan-2010 14:57:06] <rmatte> so you'd want...
[22-Jan-2010 14:57:29] <venturaville> ah
[22-Jan-2010 14:57:39] <venturaville> I was trying to call a function that only exists in the Persistence module
[22-Jan-2010 14:57:48] <venturaville> I've got it
[22-Jan-2010 14:57:54] <rmatte> k cool
[22-Jan-2010 14:58:00] <venturaville> path is what I was after
[22-Jan-2010 14:58:01] <venturaville> thanks :-)
[22-Jan-2010 14:58:05] <rmatte> no problem
[22-Jan-2010 15:25:02] <etank> ckrough: one last question. does RRA:AVERAGE:0.5:12:8760 say to keep one hour averages for a yeahr?
[22-Jan-2010 15:25:25] <etank> i think i am starting to understand it some
[22-Jan-2010 15:46:01] <ckrough> etank: yes
[22-Jan-2010 15:47:05] <ckrough> 8760 rows, each of which represents 12 300s samples (averaged), so... 8760 * 12 * 300 = 1 year
[22-Jan-2010 15:48:25] <rmatte> that's hillarious
[22-Jan-2010 15:48:32] <rmatte> shows how many users they have on here
[22-Jan-2010 15:58:24] <etank> ckrough: awesome. im starting to get it now
[22-Jan-2010 16:04:00] <dec3pti0n> yeah I don't know what's wrong with my zenoss cause I just added an OSprocess to be monitored that is not running on this box and still no alerts !
[22-Jan-2010 16:04:13] <rmatte> restart zenprocess
[22-Jan-2010 16:04:13] <ckrough> rmatte: whats that
[22-Jan-2010 16:04:16] <ckrough> lol
[22-Jan-2010 16:04:19] * dec3pti0n must be doing something wrong
[22-Jan-2010 16:04:22] <rmatte> it's a fussy daemon, needs to be restarted when changes are made
[22-Jan-2010 16:04:33] <rmatte> hopefully that'll eventually be fixed
[22-Jan-2010 16:04:38] <rmatte> just go to settings -> daemons
[22-Jan-2010 16:04:41] <rmatte> and click restart on it
[22-Jan-2010 16:05:14] <rmatte> ckrough: the fact that they could that blog down within seconds of the global message being sent out by tomaw
[22-Jan-2010 16:05:42] <rmatte> the fact that they took that blog down rather
[22-Jan-2010 16:05:47] <dec3pti0n> cool thanks
[22-Jan-2010 16:05:53] <rmatte> np
[22-Jan-2010 16:06:11] <ckrough> oh. hadnt even looks at my status window in a while
[22-Jan-2010 16:06:17] <rmatte> dec3pti0n: oh, also, from the device page select Manage -> Push Changes in the dropdown menu
[22-Jan-2010 16:06:42] <rmatte> ckrough: I use a CLI IRC client so I see everything in one windows (except I split the channels up in to virtual windows)
[22-Jan-2010 16:06:51] <ckrough> same
[22-Jan-2010 16:06:53] <ckrough> irssi
[22-Jan-2010 16:06:59] <rmatte> bitchx here
[22-Jan-2010 16:07:01] <rmatte> lol
[22-Jan-2010 16:07:15] <rmatte> osiris-1c/bitchx-1.1-final
[22-Jan-2010 16:07:46] <rmatte> I've been using it for ages so it's a force of habbit
[22-Jan-2010 16:11:18] <rmatte> eugh, think I'm going to start using securecrt on here
[22-Jan-2010 16:11:26] <rmatte> getting too many boxes to login to manually all the time
[22-Jan-2010 16:31:40] rhettardo_ is now known as rhettardo
[22-Jan-2010 16:35:06] <rmatte> there we go, all setup, woot
[22-Jan-2010 16:38:42] <QubeZ> rmatte testing the nfs now.. got the transform in place, added the commmand to zenoss, put the remount script on relevant server, and exchanged ssh keys
[22-Jan-2010 16:38:49] <QubeZ> now time to umount the drive
[22-Jan-2010 16:45:05] <rmatte> it should work fine, it did when I tested it
[22-Jan-2010 16:47:59] <QubeZ> im waiting for graphs to draw... its been about 15 mins and still showing "nan" for values
[22-Jan-2010 16:48:05] <QubeZ> gonna wait til its actually monitored properly
[22-Jan-2010 16:51:50] <rmatte> takes about 20
[22-Jan-2010 16:59:28] <QubeZ> ok its umounted
[22-Jan-2010 16:59:32] <QubeZ> crossing-fingers
[22-Jan-2010 17:05:18] <QubeZ> hmmm still nothing
[22-Jan-2010 17:12:52] <QubeZ> DLR11|/mnt/test|/Perf/Snmp||1|Error reading value for \"/mnt/test\" on DLR11 (oid .1.3.6.1.2.1.25.2.3.1.6.6 is bad)
[22-Jan-2010 17:13:21] <QubeZ> which makes sense because thats around when i umounted it. but no event was triggered
[22-Jan-2010 17:27:08] <rmatte> click on the drive
[22-Jan-2010 17:27:12] <rmatte> what is "type" listed as?
[22-Jan-2010 17:27:34] <QubeZ> well sorry, event did triffer (an info one) saying the oid is bad
[22-Jan-2010 17:27:47] <QubeZ> so i changed my message to look for that oid and testing again
[22-Jan-2010 17:28:00] <QubeZ> type is networkDisk
[22-Jan-2010 17:29:51] <rmatte> hmmm
[22-Jan-2010 17:29:59] <rmatte> let me check a few things...
[22-Jan-2010 17:30:12] <QubeZ> i remounted and now the info message has went away
[22-Jan-2010 17:30:34] <rmatte> you shouldn't be getting an info message
[22-Jan-2010 17:30:40] <rmatte> you should be getting a debug message
[22-Jan-2010 17:31:10] <QubeZ> i didn't, its grey color... when i check/uncheck info messages in my event log, it disappears and reappearas
[22-Jan-2010 17:31:22] <rmatte> when you said you changed the OID you changed it where?
[22-Jan-2010 17:31:39] <rmatte> look...
[22-Jan-2010 17:31:40] <QubeZ> rmatte in my conditiosn for running the command: i said "when message contains <oid>"
[22-Jan-2010 17:31:48] <rmatte> I'll compare the OID to the one in the transform...
[22-Jan-2010 17:31:55] <rmatte> .1.3.6.1.2.1.25.2.3.1.6.6
[22-Jan-2010 17:31:56] <rmatte> 1.3.6.1.2.1.25.2.3.1.6
[22-Jan-2010 17:32:00] <QubeZ> its the same except the .6 at the end is missing
[22-Jan-2010 17:32:06] <rmatte> yeh, that's fine
[22-Jan-2010 17:32:09] <rmatte> I'm doing a re.search
[22-Jan-2010 17:32:15] <rmatte> so it's supposed to be missing in the transform
[22-Jan-2010 17:32:28] <rmatte> it should automatically apply to any hard disk related debug message
[22-Jan-2010 17:32:33] <rmatte> where did you put the trap?
[22-Jan-2010 17:32:37] <rmatte> erm, the transform
[22-Jan-2010 17:32:51] <QubeZ> under events/perf/snmp
[22-Jan-2010 17:33:16] <QubeZ> now that i remounted the drive, no graphs show up even after waiting awhile
[22-Jan-2010 17:33:30] <rmatte> they won't without the transform working properly
[22-Jan-2010 17:33:34] <rmatte> which for some reason it's not
[22-Jan-2010 17:33:40] <rmatte> it worked perfectly fine for me when I tested it
[22-Jan-2010 17:34:03] <QubeZ> hmm
[22-Jan-2010 17:34:04] <rmatte> what's the summary of the debug message?
[22-Jan-2010 17:34:14] <QubeZ> Error reading value for "/mnt/test" on DLR11 (oid .1.3.6.1.2.1.25.2.3.1.6.6 is bad)
[22-Jan-2010 17:34:29] <rmatte> you're 100% certain that you didn't somehow forget to click save or something when putting the transform in?
[22-Jan-2010 17:34:42] <QubeZ> yup because each time i go back to it, its there
[22-Jan-2010 17:34:47] <QubeZ> and i've hit save on that and my command portion
[22-Jan-2010 17:35:03] <rmatte> what version of Zenoss are you using?
[22-Jan-2010 17:35:17] <QubeZ> 2.5.1 i believe
[22-Jan-2010 17:35:23] <rmatte> hmmm
[22-Jan-2010 17:36:00] <rmatte> I'm blown away that it's not working then
[22-Jan-2010 17:36:43] <rmatte> I tested it extensively
[22-Jan-2010 17:36:43] <QubeZ> status for my /mnt/test shows grey under status of my device even though i mounted the nfs share again
[22-Jan-2010 17:37:03] <rmatte> like I said, that's because he transform didn't get to run through
[22-Jan-2010 17:37:12] <rmatte> you need to go to Manage -> Push changes on the device page
[22-Jan-2010 17:37:16] <rmatte> then wait 15 minutes
[22-Jan-2010 17:37:25] <rmatte> the transform does that automatically provided it runs through
[22-Jan-2010 17:37:32] <QubeZ> ahh k
[22-Jan-2010 17:37:33] <rmatte> I can't fathom why it's not running for you
[22-Jan-2010 17:37:50] <rmatte> your transform is exactly what was pasted on the site right?
[22-Jan-2010 17:37:55] <rmatte> you don't have any wacky formatting issues
[22-Jan-2010 17:38:00] <rmatte> like whitespace missing?
[22-Jan-2010 17:38:00] <QubeZ> yup, i went to your wiki page to copy/paste
[22-Jan-2010 17:38:08] <rmatte> can you pastebin from zenoss?
[22-Jan-2010 17:38:11] <QubeZ> sure
[22-Jan-2010 17:38:13] <rmatte> thanks
[22-Jan-2010 17:38:36] <QubeZ> http://pastebin.com/m687d16ea
[22-Jan-2010 17:40:36] <rmatte> you actually went to More -> Transform right? you didn't add it in as it's own event mapping or something?
[22-Jan-2010 17:40:53] <QubeZ> yup, its in more->transform
[22-Jan-2010 17:41:01] * rmatte scratches his head
[22-Jan-2010 17:41:09] <QubeZ> do i need to restart services or something for the transform to activate?
[22-Jan-2010 17:41:15] <QubeZ> saving it enough i'd assume
[22-Jan-2010 17:41:17] <rmatte> push changes, wait for it to start collecting again, then unmount it again
[22-Jan-2010 17:41:23] <QubeZ> k
[22-Jan-2010 17:41:23] <rmatte> no
[22-Jan-2010 17:41:28] <rmatte> it should work without a restart
[22-Jan-2010 17:41:47] <rmatte> I never had to restart Zenoss to get it to work
[22-Jan-2010 17:46:11] <rmatte> maybe there's something in the transform which zenoss 2.5 isn't liking, but I can't imagine what
[22-Jan-2010 17:46:17] <rmatte> it's all really basic stuff
[22-Jan-2010 17:47:00] <rmatte> I'm staring at the events that got transformed right now on my test box
[22-Jan-2010 17:48:29] <mrayzenoss> gotta go, let me know if this works for someone besides rmatte and I'll make it a tip of the month
[22-Jan-2010 17:48:38] <rmatte> k
[22-Jan-2010 17:48:47] <rmatte> If it doesn't I'll have to fix it
[22-Jan-2010 17:49:06] <QubeZ> this is a 2.5.1 install, just verified it
[22-Jan-2010 17:49:12] <rmatte> http://tinypic.com/view.php?pic=15xmxj6&s=6
[22-Jan-2010 17:49:16] <rmatte> like, check that out
[22-Jan-2010 17:49:27] <rmatte> the first event is the one prior to me having it set the eventClass
[22-Jan-2010 17:49:28] <QubeZ> rmatte and its basically mount <serverIP>:/home/zenoss /mnt/test
[22-Jan-2010 17:49:31] <QubeZ> thats all ive done
[22-Jan-2010 17:49:33] <rmatte> and the others are all the tests after that
[22-Jan-2010 17:50:28] <rmatte> they came in as critical as they should and kept coming in until the drive was remounted, then 15 minutes later it started collecting performance data for the drive again
[22-Jan-2010 17:50:37] <rmatte> so I can't possibly see how it's not working for you
[22-Jan-2010 17:50:52] <rmatte> I'll test it on my king crab box, give me 15 or 20 mins
[22-Jan-2010 17:51:16] <QubeZ> k
[22-Jan-2010 17:52:23] <rmatte> oooooh, weird
[22-Jan-2010 17:52:32] <rmatte> when you pasted it in and clicked save did the text turn red?
[22-Jan-2010 17:52:38] <QubeZ> yup
[22-Jan-2010 17:52:41] <rmatte> damn
[22-Jan-2010 17:52:46] <rmatte> that means it's choking on something
[22-Jan-2010 17:52:49] <rmatte> red text = bad
[22-Jan-2010 17:52:57] <QubeZ> ooh lol
[22-Jan-2010 17:52:59] <rmatte> let me try to figure out what it's complaining about...
[22-Jan-2010 17:53:47] <rmatte> oh crap, I might know what it is
[22-Jan-2010 17:54:21] <rmatte> hmmm, apparently not grrr
[22-Jan-2010 17:55:23] <rmatte> hmmmm, it's underlining potential problems but it's all super standard stuff
[22-Jan-2010 17:55:27] <rmatte> what the heck did they do to this
[22-Jan-2010 17:56:07] <rmatte> there's always some weird thing after every upgrade
[22-Jan-2010 17:56:09] <QubeZ> this is the line that makes it turn red: if re.search('1.3.6.1.2.1.25.2.3.1.6', evt.summary) and evt.severity == 1:
[22-Jan-2010 17:56:20] <rmatte> you sure?
[22-Jan-2010 17:56:32] <QubeZ> yup, just pasted all the code in there without that line and its blueish
[22-Jan-2010 17:56:35] <rmatte> yeh, you're right
[22-Jan-2010 17:56:38] <rmatte> ok, let me figure out why
[22-Jan-2010 17:56:52] <rmatte> oh shit, it is the spacing
[22-Jan-2010 17:56:55] <rmatte> wiki messed it up
[22-Jan-2010 17:57:00] <rmatte> one second
[22-Jan-2010 17:57:46] <rmatte> I'll have it fixed momentarily
[22-Jan-2010 17:58:21] <rmatte> damnit
[22-Jan-2010 17:58:59] <rmatte> hold on, I remember some trick that I used to post code properly before (wish there was just a code button to click)
[22-Jan-2010 18:00:49] <rmatte> eugh, it keeps flattening out the formatting, I am going to lose it in a second lol
[22-Jan-2010 18:01:03] <QubeZ> i fixed it
[22-Jan-2010 18:01:03] <rmatte> let me pastebin you the proper code so you can test while I fix this
[22-Jan-2010 18:01:11] <QubeZ> just spaced out the three lines after the first if:
[22-Jan-2010 18:01:12] <rmatte> k cool
[22-Jan-2010 18:01:15] <rmatte> yup
[22-Jan-2010 18:01:23] <rmatte> that's the issue, jive keeps insisting to remove the spaces
[22-Jan-2010 18:01:25] <QubeZ> i see now that is how your other if (2nd one) is properly setup
[22-Jan-2010 18:01:33] <rmatte> yeh
[22-Jan-2010 18:01:42] <rmatte> I hadn't noticed the problem before
[22-Jan-2010 18:01:49] <QubeZ> graphs still aren't drawing though
[22-Jan-2010 18:01:49] <rmatte> now I need to wrestle with jive to display it properly
[22-Jan-2010 18:02:06] <QubeZ> usage graphs just stopped even though i pushed changes
[22-Jan-2010 18:02:14] <rmatte> did you do Manage -> Push Changes?
[22-Jan-2010 18:02:21] <rmatte> remodel then
[22-Jan-2010 18:02:43] <QubeZ> k then will i have to wait again before umounting the filesystem?
[22-Jan-2010 18:02:45] <rmatte> the pushing of the changes need to be basically instant for it to work from what I saw
[22-Jan-2010 18:02:51] <rmatte> yeh, you will
[22-Jan-2010 18:02:56] <QubeZ> darn
[22-Jan-2010 18:05:37] <rmatte> now how the heck am I going to get this code in here properly
[22-Jan-2010 18:06:09] <QubeZ> but how did the other spaces work?
[22-Jan-2010 18:07:32] <rmatte> what?
[22-Jan-2010 18:07:50] <rmatte> http://pastebin.com/m7487273e
[22-Jan-2010 18:07:58] <rmatte> that's the way the code SHOULD be formatted
[22-Jan-2010 18:07:58] <QubeZ> the other spaces after the 2nd if
[22-Jan-2010 18:08:22] <QubeZ> ok i unmounted the nfs share
[22-Jan-2010 18:08:34] <rmatte> make sure the code is exactly as shown in that pastebin
[22-Jan-2010 18:08:36] <QubeZ> no email and no zencomand log entries
[22-Jan-2010 18:08:37] <rmatte> or it's not going to work
[22-Jan-2010 18:08:55] <QubeZ> i put two spaces after the if statement but ok ill copy yours exactly
[22-Jan-2010 18:09:07] <rmatte> yeh copy it exactly
[22-Jan-2010 18:09:15] <rmatte> python is EXTREMELY sensitive about spaces
[22-Jan-2010 18:09:59] <rmatte> AHA!
[22-Jan-2010 18:10:08] <rmatte> I finally figured it out, man that's not in an obvious spot
[22-Jan-2010 18:10:30] <rmatte> there, it's fixed on the page
[22-Jan-2010 18:10:47] <QubeZ> ok event triggered
[22-Jan-2010 18:11:03] <rmatte> cool
[22-Jan-2010 18:11:07] <QubeZ> i dont see my sh script executing though
[22-Jan-2010 18:11:10] <QubeZ> tailing the zencomand.log now
[22-Jan-2010 18:11:10] <rmatte> the transform worked I hope?
[22-Jan-2010 18:11:40] <QubeZ> ya the summary has changed
[22-Jan-2010 18:12:32] <QubeZ> got the email
[22-Jan-2010 18:12:36] <QubeZ> now when does the command run?
[22-Jan-2010 18:13:00] <QubeZ> doh, i need to change the message filter back
[22-Jan-2010 18:13:05] <rmatte> it should run instantaneously
[22-Jan-2010 18:13:09] <rmatte> but yeh, bad filter will do it
[22-Jan-2010 18:13:10] <rmatte> lol
[22-Jan-2010 18:13:28] <rmatte> after you unmount and get an event leave it unmounted
[22-Jan-2010 18:13:42] <rmatte> If it's working properly the event should be generated again and again until you've remounted it
[22-Jan-2010 18:13:51] <rmatte> I just want to make sure you see that behaviour
[22-Jan-2010 18:14:04] <QubeZ> i just changed the command parameters, so seeing as i messed it up the first time.. im leaving the drive unmounted so it'll catch it
[22-Jan-2010 18:14:09] <rmatte> then after you remount it it should start collecting data for it again 15 to 20 mins after
[22-Jan-2010 18:14:16] <rmatte> ok cool
[22-Jan-2010 18:14:22] <rmatte> you should see a second event come in
[22-Jan-2010 18:14:31] <QubeZ> count is up to 2
[22-Jan-2010 18:14:39] <rmatte> ok, now move it to history
[22-Jan-2010 18:14:46] <rmatte> the command will only generate on the first event count
[22-Jan-2010 18:14:57] <QubeZ> ok host key verification failed but thats fine i need to fix that
[22-Jan-2010 18:16:08] <QubeZ> i accidently exchanged the root key for my monitoring server instead of the zenoss user
[22-Jan-2010 18:17:55] <rmatte> lol
[22-Jan-2010 18:17:56] <rmatte> nice
[22-Jan-2010 18:18:10] <QubeZ> the lock is there
[22-Jan-2010 18:18:14] <rmatte> so you got 3 events, the third one kicked off the script and remounted it?
[22-Jan-2010 18:18:15] <QubeZ> its not triggering any command now
[22-Jan-2010 18:18:27] <rmatte> the lock should be there on the first event
[22-Jan-2010 18:18:30] <rmatte> so that's good
[22-Jan-2010 18:18:30] <QubeZ> i have 3 events now yes, wait 4 now
[22-Jan-2010 18:18:40] <rmatte> ok, but did the script successfully kick off?
[22-Jan-2010 18:18:46] <rmatte> like, can you confirm it remounted?
[22-Jan-2010 18:18:52] <QubeZ> not mounted
[22-Jan-2010 18:18:59] <rmatte> ok, then the script failed for some reason
[22-Jan-2010 18:19:21] <rmatte> whenever you're ready to try having zenoss kick it off again just move the events to history
[22-Jan-2010 18:20:16] <rmatte> (not sure if you saw me say that before)
[22-Jan-2010 18:20:22] <rmatte> you need to move current events to history
[22-Jan-2010 18:20:31] <rmatte> then the next one that comes in will kick off the script
[22-Jan-2010 18:20:39] <rmatte> it only gets kicked off on an event count of 1
[22-Jan-2010 18:20:49] <QubeZ> i selected close events
[22-Jan-2010 18:20:53] <QubeZ> no new events yet
[22-Jan-2010 18:21:00] <rmatte> well give it time
[22-Jan-2010 18:21:08] <rmatte> it only checks once per polling cycle
[22-Jan-2010 18:21:47] <rmatte> afk for a minute
[22-Jan-2010 18:23:05] <rmatte> still no events?
[22-Jan-2010 18:23:13] <QubeZ> nope
[22-Jan-2010 18:23:17] <QubeZ> drive is umounted
[22-Jan-2010 18:23:27] <rmatte> k, wait a bit longer
[22-Jan-2010 18:24:13] <QubeZ> triggered
[22-Jan-2010 18:24:17] <rmatte> cool
[22-Jan-2010 18:24:24] <QubeZ> but no remount
[22-Jan-2010 18:24:28] <QubeZ> i run the script, it runs
[22-Jan-2010 18:24:44] <rmatte> did you run it as the zenoss user though?
[22-Jan-2010 18:25:00] <rmatte> it won't be able to run it if it needs to be run as root
[22-Jan-2010 18:25:14] <rmatte> (well, unless it has the key and is configured to do so)
[22-Jan-2010 18:25:45] <QubeZ> ok i messed it up again lol... fuck this friday
[22-Jan-2010 18:25:50] <QubeZ> it'll work now for sure, thanks rmatte
[22-Jan-2010 18:25:50] <rmatte> lol
[22-Jan-2010 18:25:55] <rmatte> np
[22-Jan-2010 18:25:59] <rmatte> what was messed?
[22-Jan-2010 18:26:02] <rmatte> key was in root?
[22-Jan-2010 18:26:49] <QubeZ> first time that was the problem, but second time it said again host key verificatin failed. I ran 'su zenoss; ssh root@MYIP /zenoss/remountNFS.sh' and it asked for verification and i hit yes so it added to known hosts
[22-Jan-2010 18:27:05] <rmatte> ah
[22-Jan-2010 18:27:06] <QubeZ> i guess before i use to run it as root (the command) so user was not entered into known hosts, only into authorized_keys
[22-Jan-2010 18:27:07] <rmatte> gotcha
[22-Jan-2010 18:27:16] <QubeZ> k event triggered again
[22-Jan-2010 18:27:17] <rmatte> makes sense
[22-Jan-2010 18:27:34] <QubeZ> mount successful!
[22-Jan-2010 18:27:36] <QubeZ> hell yea
[22-Jan-2010 18:27:39] <i-moet> anyone can help me with reference about translation or language ?
[22-Jan-2010 18:27:45] <rmatte> ok, now wait to make sure it starts collecting data in a good amount of time?
[22-Jan-2010 18:27:59] <i-moet> any url about it/howto, will help me
[22-Jan-2010 18:28:02] <QubeZ> ya but im going home for the weekend and going to look at graphs on monday
[22-Jan-2010 18:28:23] <rmatte> k cool
[22-Jan-2010 18:28:31] <rmatte> it should work, but let me know if there's a big gap
[22-Jan-2010 18:29:46] <QubeZ> ok
[22-Jan-2010 18:32:06] <QubeZ> k made sure the actual production servers now have the right keys hehe
[22-Jan-2010 18:32:17] <QubeZ> lock is still there thou
[22-Jan-2010 18:32:26] <rmatte> the lock will stay there, that's not an issue
[22-Jan-2010 18:32:48] <rmatte> it's a precaution to keep an automated remodel from potentially removing the share from monitoring
[22-Jan-2010 18:33:09] <QubeZ> k
[22-Jan-2010 18:33:32] <rmatte> if a remodel were to kick off while the share was unmounted it would delete the share from monitoring
[22-Jan-2010 18:33:36] <rmatte> which you obviously don't want
[22-Jan-2010 18:33:44] <rmatte> so I added that in to the transform for some added security
[22-Jan-2010 18:33:49] <QubeZ> nice
[22-Jan-2010 18:34:37] <rmatte> I'm glad we did this because I wouldn't have noticed the spacing issue on the site
[22-Jan-2010 18:34:49] <QubeZ> im your guinea pig
[22-Jan-2010 18:34:52] <rmatte> lol
[22-Jan-2010 18:35:04] <rmatte> well, Matt will be happy, he has a tip of the month now
[22-Jan-2010 18:35:23] <QubeZ> not sure why the event has disappeared, the nfs is fine
[22-Jan-2010 18:35:32] <rmatte> it won't dissapear
[22-Jan-2010 18:35:39] <rmatte> you'll have to move it to history manually
[22-Jan-2010 18:35:53] <QubeZ> ok, how about the server going back to green, its still red
[22-Jan-2010 18:36:08] <rmatte> it'll show as red until the event is moved to history
[22-Jan-2010 18:36:19] <rmatte> it's red because of the event
[22-Jan-2010 18:36:24] <i-moet> is it the zenoss or the zope will handle the i18n ?
[22-Jan-2010 18:36:34] <QubeZ> i'd think it would clear the event automatically
[22-Jan-2010 18:36:37] <rmatte> i18n?
[22-Jan-2010 18:36:39] <QubeZ> after the drive comes back
[22-Jan-2010 18:36:51] <rmatte> QubeZ: not in this case since the event is being transformed
[22-Jan-2010 18:37:06] <rmatte> QubeZ: if it were still a debug event it would
[22-Jan-2010 18:37:13] <QubeZ> so i guess thats phase 2, trying to figure out how to move the event to history
[22-Jan-2010 18:37:16] <rmatte> there's no way around it that I can see
[22-Jan-2010 18:37:30] <QubeZ> k thanks man, have a good weekend
[22-Jan-2010 18:37:35] <rmatte> you too
[22-Jan-2010 18:37:36] <rmatte> cheers
[22-Jan-2010 18:37:51] <i-moet> rmatte : Internationalization / translation to other language from English ?
[22-Jan-2010 18:38:15] <rmatte> i-moet: I don't believe Zenoss supports anything other than English unfortunately
[22-Jan-2010 18:38:31] <i-moet> sorry, but is these the zenoss (comercial) or the zenoss comunity ?
[22-Jan-2010 18:38:44] <i-moet> i mean these irc's channel
[22-Jan-2010 18:38:45] <rmatte> both of them only support english
[22-Jan-2010 18:38:51] <rmatte> this IRC channel pertains to both
[22-Jan-2010 18:38:58] <rmatte> but it is mostly core users in here
[22-Jan-2010 18:39:04] <rmatte> some developers are around as well sometimes
[22-Jan-2010 18:39:10] <i-moet> rmatte : ah, that's why googling it give me nothing :(
[22-Jan-2010 18:39:19] <rmatte> yes
[22-Jan-2010 18:40:40] <i-moet> http://bit.ly/657wQP --> give me point to the zope related.
[22-Jan-2010 18:40:45] <i-moet> ok never mind.
[22-Jan-2010 18:40:57] <i-moet> rmatte : thanks for your help :)
[22-Jan-2010 18:40:58] <rmatte> If they offered it in other languages they'd have to offer support in other languages. Their support engineers only speak english.
[22-Jan-2010 18:41:02] <rmatte> no problem
[22-Jan-2010 18:42:34] <i-moet> see u again all.
[22-Jan-2010 19:48:20] <pkqrahkk> Hi all. It seems we are again seeing javascript based flood spam. If you are experiencing this, please do not click the links in the messages as they will cause you to repeat the spam. More information is available at http://peoplesprimary.com. Thanks!
[22-Jan-2010 19:48:20] <pkqrahkk> Hi all. It seems we are again seeing javascript based flood spam. If you are experiencing this, please do not click the links in the messages as they will cause you to repeat the spam. More information is available at http://peoplesprimary.com. Thanks!
[22-Jan-2010 19:48:23] <pkqrahkk> Hi all. It seems we are again seeing javascript based flood spam. If you are experiencing this, please do not click the links in the messages as they will cause you to repeat the spam. More information is available at http://peoplesprimary.com. Thanks!
[23-Jan-2010 00:00:30] [disconnected at Sat Jan 23 00:00:30 2010]
[23-Jan-2010 00:00:30] [connected at Sat Jan 23 00:00:30 2010]
[23-Jan-2010 03:50:06] forsberg is now known as fOrsberg
[23-Jan-2010 03:50:21] fOrsberg is now known as forsberg
[23-Jan-2010 04:46:53] forsberg is now known as fOrsberg
[23-Jan-2010 07:35:42] fOrsberg is now known as forsberg
[23-Jan-2010 08:27:47] <forsberg> rmatte ?
[23-Jan-2010 12:28:35] forsberg is now known as fOrsberg
[23-Jan-2010 12:30:26] fOrsberg is now known as forsberg
[23-Jan-2010 12:46:00] <tuv> is an atom processor (e.g. dual-core atom 330) sufficient for smooth zenoss operation?
[23-Jan-2010 13:20:13] forsberg is now known as fOrsberg
[23-Jan-2010 13:24:25] <venturaville> anyone know how much about modeler plugins?
[23-Jan-2010 18:31:04] <tuv> is an atom processor (e.g. dual-core atom 330) sufficient for smooth zenoss operation?
[23-Jan-2010 19:24:14] <Apachez> zenoss will never be smooth as long as zope is being used
[23-Jan-2010 19:24:15] <Apachez> but sure
[24-Jan-2010 00:00:30] [disconnected at Sun Jan 24 00:00:30 2010]
[24-Jan-2010 00:00:30] [connected at Sun Jan 24 00:00:30 2010]
[24-Jan-2010 03:19:30] fOrsberg is now known as forsberg
[24-Jan-2010 13:34:30] <forsberg> idle idle?
[24-Jan-2010 13:34:55] <forsberg> anyone who made a nice /device/server/freebsd template? :)
[24-Jan-2010 18:21:07] forsberg is now known as fOrsberg
[24-Jan-2010 18:33:58] <tuvwx> what is the minimum required memory for zenoss?
[24-Jan-2010 22:47:51] <frozty_sa> tuvwx: roundabout 512MB to be safe
[25-Jan-2010 00:00:30] [disconnected at Mon Jan 25 00:00:30 2010]
[25-Jan-2010 00:00:30] [connected at Mon Jan 25 00:00:30 2010]
[25-Jan-2010 00:40:38] <tuv> frozty_sa: i ran it on a 512MB-ram machine, and it practically crippled the machine. browsing to the zenoss webpage took ages and eventually failed with a disconnected from server type of error
[25-Jan-2010 00:41:41] <tuv> it was running smoothly initiall though. this happened the next day. i guess python builds up its memory over time. even shutting zenoss down took like half an hour
[25-Jan-2010 00:41:47] <frozty_sa> I said minimum
[25-Jan-2010 00:41:56] <frozty_sa> but that's if the box is doing nothing else already
[25-Jan-2010 00:42:43] <tuv> tbh, i had cacti running there too, so a lightweight webserver, mysqld and a few php-cgi's
[25-Jan-2010 00:42:51] <frozty_sa> tuv: 512MB is more or less what zenoss needs to start up, in my experience. from that I'd probably say you want a gig or more for actual use
[25-Jan-2010 00:43:06] <frozty_sa> the already existing mysqld chomped some RAM then
[25-Jan-2010 01:18:56] nassy_ is now known as nassy
[25-Jan-2010 01:30:19] <Apachez> depends on how you configure your mysql
[25-Jan-2010 01:30:29] <Apachez> unfortunately zenoss lacks proper configuration of it
[25-Jan-2010 01:30:43] <Apachez> however the mysql is only used for events (syslog, snmptraps etc)
[25-Jan-2010 01:31:05] <Apachez> so if you have low of those you can use very moderate settings for the mysql like 16 meg innodbbuffer etc
[25-Jan-2010 02:29:26] <frozty_sa> Apachez: careful, I might just rebound to you any time I have zenoss architecture questions... :-p
[25-Jan-2010 03:06:26] baffle_ is now known as baffle
[25-Jan-2010 03:07:19] <baffle> Hmm, ZenOss is confused over different routingtables pointing to 10.0.0.0/24 and simmilar. Is there any good way to clean up the DB? Right now the Network-tab says that 10.0.0.0/24 has 860 subnetworks.
[25-Jan-2010 03:13:31] <frozty_sa> so uhm. zenoss. debian package, 2.5.1, from the http://dev.zenoss.org/deb stable repo
[25-Jan-2010 03:13:45] <frozty_sa> on an arb note, how on earth do I set the address for a location?
[25-Jan-2010 03:14:07] <frozty_sa> on a more serious line, where do I ask this thing for debugging information in the perf views?
[25-Jan-2010 03:14:27] <frozty_sa> I can get graphs drawn if I use the reports, but the perf views don't show anything
[25-Jan-2010 03:14:34] <frozty_sa> and all the RRDs are nicely populated with data
[25-Jan-2010 03:22:55] <baffle> frozty_sa: You go to /Locations, and click the small [edit] next to the address.
[25-Jan-2010 03:24:53] <baffle> Has anyone noticed that 2.5.1 (2.5.x?) is really bad at cleaning up after itself? I've stopped reporting of network connections from net-snmp (because it would make servers go max-load for a few minutes while it polled /proc) and now ZenOss has a lot of unreferenced objects I have to clean up with a brains-purge/reindex.
[25-Jan-2010 04:09:58] <frozty_sa> baffle: I don't have anything saying "address" on my /Locations
[25-Jan-2010 04:10:01] <frozty_sa> which is what confuses me
[25-Jan-2010 04:10:21] <frozty_sa> although I do have an API key for GMaps in, and I can zoom and fly around etc on the maps view
[25-Jan-2010 07:02:22] <baffle> frozty_sa: If you go to locations, and click your location, in the "Status" tab you have "Address [edit]".
[25-Jan-2010 07:23:39] <frozty_sa> bobrog: no, I don't
[25-Jan-2010 07:24:04] <frozty_sa> I have Summary as one big header, then I've got two columbs below that. Sub-locations, and Devices
[25-Jan-2010 07:24:31] <frozty_sa> Sub-Locations' menu only gives me the ability to create, move and delete organizers
[25-Jan-2010 07:25:05] <frozty_sa> but that's cosmetic, I'll worry about that sometime later
[25-Jan-2010 07:25:41] <frozty_sa> more pressing concerns are things like why it picks up 100Mbps interfaces as 10Mbps and why I can get interface graphs in one place but not the other :)
[25-Jan-2010 08:04:48] <chickenandbeans> Afternoon
[25-Jan-2010 08:51:39] <ckrough> hellp
[25-Jan-2010 08:51:54] <ckrough> hello I mean, though heeellllp usually applies as well...
[25-Jan-2010 08:54:56] <rocket> funny .. :p
[25-Jan-2010 08:58:13] <mistich> good morning
[25-Jan-2010 08:58:29] <rmatte> frozty_sa: if it's picking them up as 10Mbps it's because that's what they are being displayed at via SNMP
[25-Jan-2010 08:58:30] <mistich> anyone around for a threshold question?
[25-Jan-2010 08:58:36] <rmatte> frozty_sa: so it's some misconfiguration
[25-Jan-2010 08:58:42] <rmatte> mistich: ask away
[25-Jan-2010 08:59:05] <mistich> I'm trying to set a min/max threshold on percentage of free memory.
[25-Jan-2010 08:59:05] <mistich> such as 80% used
[25-Jan-2010 08:59:05] <mistich> the pix only has 2 snmp datapoints memory_free and memory_used.
[25-Jan-2010 08:59:05] <mistich> so I know how to do the math but not sure how to put it in zenoss
[25-Jan-2010 08:59:05] <mistich> memory_free + memory_used * 0.80
[25-Jan-2010 08:59:06] <mistich> this will get the 80% number but in zenoss when you go to threshold
[25-Jan-2010 08:59:08] <mistich> you can select the 2 datapoints
[25-Jan-2010 08:59:10] <mistich> memory_free and memory_used
[25-Jan-2010 08:59:12] <mistich> in the Max Value i put
[25-Jan-2010 08:59:14] <mistich> memory_free + memory_used * 0.80
[25-Jan-2010 08:59:16] <mistich> but if I do that what value does it compare the Max to
[25-Jan-2010 08:59:18] <mistich> memory_free or memory_used
[25-Jan-2010 08:59:40] <rmatte> mistich: I'll show you how I do it for my Cisco templates
[25-Jan-2010 08:59:51] <mistich> cool
[25-Jan-2010 09:00:10] fOrsberg is now known as forsberg
[25-Jan-2010 09:00:11] <mistich> so you add 2 data points together
[25-Jan-2010 09:00:22] <rmatte> (here.getRRDValue('mem5minUsed') + here.getRRDValue('mem5minFree')) * 0.9
[25-Jan-2010 09:00:26] <rmatte> that's how I do it
[25-Jan-2010 09:00:30] <mistich> ahhh
[25-Jan-2010 09:00:41] <rmatte> mistich: one second though, small detail...
[25-Jan-2010 09:00:45] <mistich> now were talking
[25-Jan-2010 09:01:11] <rmatte> rhettardo: zenpatch 16733 first
[25-Jan-2010 09:01:17] <mistich> then you select the datapoint you want to compare it too
[25-Jan-2010 09:01:17] <rmatte> bah
[25-Jan-2010 09:01:19] <rmatte> I meant do:
[25-Jan-2010 09:01:33] <rmatte> (IRC client autocompleted it as a name)
[25-Jan-2010 09:02:07] <frozty_sa> rmatte: yeah, I found that
[25-Jan-2010 09:02:08] <rmatte> it fixes: http://dev.zenoss.org/trac/ticket/5978
[25-Jan-2010 09:02:33] <rmatte> mistich: correct
[25-Jan-2010 09:02:35] <mistich> do what?
[25-Jan-2010 09:02:38] <frozty_sa> rmatte: one of net-snmp's fun special extras
[25-Jan-2010 09:02:45] <rmatte> mistich: apply zenpack 16733 first
[25-Jan-2010 09:02:52] <rmatte> to fix this: http://dev.zenoss.org/trac/ticket/5978
[25-Jan-2010 09:02:57] <rmatte> otherwise you could run in to issues
[25-Jan-2010 09:05:40] <rmatte> something is quite wonky memory-wise with Zenoss, guess I've never really done this before, but I cleared the swap space and left Zenoss running. Each time I do, within a few hours it's back up to using over a gig of swap, but the RAM usage isn't very high at all
[25-Jan-2010 09:05:44] <rmatte> strange behaviour
[25-Jan-2010 09:07:33] <mistich> we run into that issue all the time with zenoss using up swap
[25-Jan-2010 09:07:44] <ckrough> rmatte: (mistich and I work together)
[25-Jan-2010 09:08:21] <ckrough> dealing with a similar issue, ZOPE looks to be the culprit in our install. Is it a particular process on yours?
[25-Jan-2010 09:10:54] <rmatte> Zope takes up about 50% of the Memory
[25-Jan-2010 09:11:03] <rmatte> ZenPerfSnmp takes up 15%
[25-Jan-2010 09:11:14] <rmatte> the rest is spread amongst other Zenoss processes and system processes
[25-Jan-2010 09:12:21] <ckrough> if you run 'top' and sort by VIRT then SWAP, is it usually zope and zenperf at the top?
[25-Jan-2010 09:12:46] <rmatte> what's the hot-key for swap again?
[25-Jan-2010 09:12:53] <rmatte> erm
[25-Jan-2010 09:12:54] <rmatte> for sort
[25-Jan-2010 09:12:55] <rmatte> rather
[25-Jan-2010 09:13:01] <ckrough> F
[25-Jan-2010 09:13:08] <rmatte> aha, thanks
[25-Jan-2010 09:13:12] <ckrough> o and p
[25-Jan-2010 09:14:01] <ckrough> I like 'o' more than p since it shows which process is really eating total memory the most aggressively, whereas swap is just what happens to be swapped at the moment.
[25-Jan-2010 09:14:39] <venturaville> Does anyone know of a zenpack where datapoints/graphs are added at modeling time?
[25-Jan-2010 09:14:57] <rmatte> that looks like it's just to sort the fields though
[25-Jan-2010 09:15:00] <venturaville> (i.e. custom template per device/sub-component)
[25-Jan-2010 09:15:02] <rmatte> erm select the fields rather
[25-Jan-2010 09:15:05] <rmatte> not to sort by usage
[25-Jan-2010 09:15:17] <ckrough> upper F, case sensitive
[25-Jan-2010 09:15:18] <rmatte> oh nevermind
[25-Jan-2010 09:15:21] <rmatte> yeh, just noticed
[25-Jan-2010 09:15:21] <ckrough> f = select, F = sort
[25-Jan-2010 09:15:54] <rmatte> can't see the full names of the commands
[25-Jan-2010 09:16:01] <rmatte> just .python.bin
[25-Jan-2010 09:16:07] <ckrough> 'c'
[25-Jan-2010 09:16:10] <rmatte> (I usually use htop since it gives way more info
[25-Jan-2010 09:16:21] <rmatte> ah, there we go
[25-Jan-2010 09:16:22] <venturaville> rmatte: btw I finally figured out I could use __file__ in the modeler to get the zenpack path ... isn't available otherwise....
[25-Jan-2010 09:16:23] <ckrough> ya, same
[25-Jan-2010 09:17:07] <rmatte> zope is at the top with 25.7%
[25-Jan-2010 09:17:24] <rmatte> though when I run htop I see 2 zope processes
[25-Jan-2010 09:17:28] <rmatte> with 25% usage each
[25-Jan-2010 09:17:40] <ckrough> the other day when asked about your zope config for cached objects and threads... this is what I was digging into
[25-Jan-2010 09:17:54] <rmatte> ah, looks like it's just showing threads
[25-Jan-2010 09:18:00] <rmatte> but overall it's way above the others at the top
[25-Jan-2010 09:18:27] <ckrough> Im sure the 'sweet spot' for those depends on the workload of the box, but the documentation of those settings doesnt give recommended settings
[25-Jan-2010 09:18:37] <rmatte> yeh
[25-Jan-2010 09:18:50] <rmatte> and the defaults kind of suck for heavier load
[25-Jan-2010 09:28:53] <mistich> rmatte that might not work for me were on a older version
[25-Jan-2010 09:29:09] <mistich> Module Products.PageTemplates.ZRPythonExpr, line 47, in __call__
[25-Jan-2010 09:29:09] <mistich> __traceback_info__: (here.getRRDValue('mem5minUsed_mem5minUsed') + here.getRRDValue('mem5minFree_mem5minFree')) * 0.8 Module Python expression "(here.getRRDValue('mem5minUsed_mem5minUsed') + here.getRRDValue('mem5minFree_mem5minFree')) * 0.8", line 1, in <expression> TypeError: unsupported operand type(s) for +: 'NoneType' and 'NoneType'
[25-Jan-2010 09:29:45] <rmatte> mistich: what version are you on?
[25-Jan-2010 09:29:59] <mistich> 2.3.3
[25-Jan-2010 09:30:12] <rmatte> ah, you really should think about upgrading at some point
[25-Jan-2010 09:30:17] <ckrough> sush
[25-Jan-2010 09:30:21] <mistich> yes
[25-Jan-2010 09:30:22] <ckrough> shush I mean
[25-Jan-2010 09:30:26] <rmatte> you don't want to be 3 versions behind when stone crab comes out
[25-Jan-2010 09:30:30] <rmatte> very messuy
[25-Jan-2010 09:30:32] <rmatte> messy*
[25-Jan-2010 09:30:41] <rmatte> anyways, you can patch it yourself then...
[25-Jan-2010 09:30:41] <mistich> soon its on ckrough list of things to do :)
[25-Jan-2010 09:30:42] <ckrough> are you on 2.5 or 2.4.5 in production ryan?
[25-Jan-2010 09:30:53] <rmatte> mistich: go in to $ZENHOME/Products/ZenModel/RRDView.py
[25-Jan-2010 09:31:00] <rmatte> find the line with: threshold.manage_changeProperties(enabled=False)
[25-Jan-2010 09:31:02] <rmatte> comment it out
[25-Jan-2010 09:31:07] <rmatte> restart zenoss
[25-Jan-2010 09:31:12] <rmatte> (temporary workaround)
[25-Jan-2010 09:31:30] <rmatte> you'll receive the odd alert saying it's disabling the threshold but it won't
[25-Jan-2010 09:32:27] <rmatte> ckrough: 2.4.5 in prod, yes
[25-Jan-2010 09:32:35] <rmatte> but when 2.5.2 comes out I'll be testing to move to that
[25-Jan-2010 09:32:42] <rmatte> since 2.5.2 is looking like it's going to be good
[25-Jan-2010 09:32:57] <rmatte> they are spending a lot of time making sure that they fix a ton of bugs
[25-Jan-2010 09:33:04] <rmatte> hopefully it'll be the last 2.5 release
[25-Jan-2010 09:33:13] <rmatte> then they can focus on stone crab
[25-Jan-2010 09:33:42] <rmatte> my 2.4.5 has a bunch of patches applied to it though, it's not just straight 2.4.5
[25-Jan-2010 09:34:03] <mistich> no line threshold.manage_changeProperties(enabled=False) :)
[25-Jan-2010 09:34:22] <rmatte> mistich: guess that was added in 2.4 for some stupid reason
[25-Jan-2010 09:34:24] <rmatte> you're fine then
[25-Jan-2010 09:34:40] <mistich> yeah not sure why I'm getting a error
[25-Jan-2010 09:34:46] <rmatte> when 2.5.2 comes out you guys should quickly upgrade to 2.4.5 and then upgrade to 2.5.2
[25-Jan-2010 09:34:47] <ckrough> rmatte: if you have a list of 2.4.5 patches that you applied readily available I've love a copy. I will be going through all the 2.4.5 patch lists to see what we do and dont need anyway
[25-Jan-2010 09:34:52] <rmatte> since you need to follow the upgrade path
[25-Jan-2010 09:35:15] <rmatte> ckrough: one second
[25-Jan-2010 09:35:29] <rmatte> ckrough: http://community.zenoss.org/docs/DOC-4434
[25-Jan-2010 09:35:33] <ckrough> rmatte: k, only if it's convenient... I have to look at the list anyway
[25-Jan-2010 09:35:47] <ckrough> lol nice. thanks
[25-Jan-2010 09:35:49] <rmatte> it's quite convenient, I documented it in wiki
[25-Jan-2010 09:35:50] <rmatte> ;)
[25-Jan-2010 09:36:49] <rmatte> I'll add patch 16733 to that list
[25-Jan-2010 09:38:12] <rmatte> There, added it
[25-Jan-2010 09:49:00] <rmatte> good morning Matt
[25-Jan-2010 09:49:10] <mrayzenoss> morning
[25-Jan-2010 09:50:06] <forsberg> http://hesten.foppa.dk/files/Capture.JPG my new zenoss graph, the top is when my wife close the kitchen closet, and the valley is when she opens it full
[25-Jan-2010 09:50:20] <forsberg> i have to find some proper closet for that closet :)
[25-Jan-2010 09:50:25] <forsberg> ehh cooling
[25-Jan-2010 09:50:41] <rmatte> lol
[25-Jan-2010 09:50:59] <rmatte> install some vents at least
[25-Jan-2010 09:51:08] <mrayzenoss> nice. I used to have weather station in my office, you could see when the last person left the office at night, the A/C turning back on at 6am and the first people showing up in the morning
[25-Jan-2010 09:51:10] <forsberg> i have too, burned 2 hds last sommer
[25-Jan-2010 09:51:30] <forsberg> this is just smartctl on a HD
[25-Jan-2010 09:51:31] <forsberg> :)
[25-Jan-2010 09:51:31] <rmatte> lol, nice
[25-Jan-2010 09:51:50] <forsberg> but the effect of opening and closing it is really obvious...
[25-Jan-2010 09:52:24] <forsberg> i didnt have the nerves to continue the closed one for very long :)
[25-Jan-2010 09:57:20] <ckrough> I still need to have zenoss monitor the weight of the keg in my kegerator so I can order the replacement on time
[25-Jan-2010 09:57:51] <venturaville> mrayzenoss: do you know of any examples of a zenpack with a modeler that adds custom data points per device (or ideally per sub-component)
[25-Jan-2010 09:58:01] <forsberg> :P~~
[25-Jan-2010 09:58:25] <venturaville> I can do it after the fact in dmd, but doing so within the modeler has ...issues....
[25-Jan-2010 09:58:45] <mrayzenoss> custom data points? I'm not sure I follow.
[25-Jan-2010 09:59:47] <venturaville> custom data points ...e.g. for my case I have some guests on a virtual host, and I want to add a different graph for each interface or disk
[25-Jan-2010 09:59:56] <venturaville> and that list can change per guest on the host...
[25-Jan-2010 10:00:28] <mrayzenoss> so based on what modeling turns up, apply different templates?
[25-Jan-2010 10:03:35] <venturaville> not quite different templates... I actually want to create the template on the fly
[25-Jan-2010 10:03:53] <venturaville> or make the generic one local to the device, and then add my own data points
[25-Jan-2010 10:04:14] <venturaville> I had it working via script last night, but I can't get it to work in my modeler
[25-Jan-2010 10:14:48] <mrayzenoss> I know this ZenPack dynamically selects the template, it seems you could extend that idea to dynamically create datapoints instead: http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.HPSIMMonitor/ZenPacks/community/HPSIMMonitor/modeler/plugins/community/snmp/HPDeviceTemplateMap.py
[25-Jan-2010 10:16:12] <venturaville> this is just selecting a template via the zDeviceTemplates, but so far as I know the deviceproxy thing gets in the way of actually editing the datapoints using the function calls
[25-Jan-2010 10:24:40] <rmatte> he's talking about dynamic datapoints
[25-Jan-2010 10:24:46] <rmatte> which is a fairly common request
[25-Jan-2010 10:25:35] <rmatte> for example, if I were to create a command based datasource and wanted it to automatically generate datapoints based on what's found on the device
[25-Jan-2010 10:25:54] <venturaville> BINGO
[25-Jan-2010 10:25:59] <rmatte> so I could use the same template for a device where I'm collecting data for 4 CPUs or 8 CPUs
[25-Jan-2010 10:26:24] <venturaville> I think the biggest issue is the modeler not having access to the raw device object, or some level having access to it in the zenpack
[25-Jan-2010 10:26:43] <venturaville> e.g. an optional plugin level, post modeler maybe
[25-Jan-2010 10:26:58] <ckrough> add my vote too
[25-Jan-2010 10:30:28] <rmatte> well, looks like I'm going to have a chance to create a collector/modeler zenpack sooner than I anticipated
[25-Jan-2010 10:30:58] <venturaville> which one?
[25-Jan-2010 10:31:15] <rmatte> going to make a zenpack to automate setting up remote collectors for core
[25-Jan-2010 10:31:34] <rmatte> I haven't had a chance to really work with remote collector setups at all, but I was just told that we need to start looking in to it
[25-Jan-2010 10:31:47] <rmatte> so I'm going to setup 2 lab VMs within a day or two and get started on it
[25-Jan-2010 10:31:50] <rmatte> should be fun
[25-Jan-2010 10:33:55] <venturaville> yeah we are using them but we have enterprise
[25-Jan-2010 10:34:14] <theacolyte> only 900 unread messages from vacation
[25-Jan-2010 10:34:30] <rmatte> venturaville: yeh, well that's the trick, they aren't widely used with core because of the complexity of the setup
[25-Jan-2010 10:34:36] <rmatte> venturaville: I'm hoping to alleviate that
[25-Jan-2010 10:35:04] <rmatte> and I don't see any reason why I shouldn't be able to, with some research and dev time
[25-Jan-2010 10:35:46] <rmatte> The first version will be commandline based but hopefully once I have the key automation completed I may be able to get Egor lend a hand with implementing UI configuration
[25-Jan-2010 10:36:47] <rmatte> This is ridiculous... middle of January and we have rain/fog
[25-Jan-2010 10:37:26] <ckrough> 70 and sunny :)
[25-Jan-2010 10:37:34] <patzer> heh, where do you live?
[25-Jan-2010 10:37:34] <rmatte> lol
[25-Jan-2010 10:37:39] <rmatte> Ottawa, Canada
[25-Jan-2010 10:37:48] <patzer> my climate has rain/fog ... all months of the year
[25-Jan-2010 10:37:55] <ckrough> patzer: where's that
[25-Jan-2010 10:37:58] <patzer> Halifax
[25-Jan-2010 10:38:16] <rmatte> Generally our January is -22 Farenheit weather
[25-Jan-2010 10:38:19] <patzer> on the north atlantic, similar latitude and geograpy to London
[25-Jan-2010 10:38:29] <rmatte> It's 39.2 Farenheit out right now
[25-Jan-2010 10:38:59] <patzer> yeah, there is a heat wave coming through
[25-Jan-2010 10:39:13] <rmatte> patzer: we might be getting some of that from you guys
[25-Jan-2010 10:39:38] <patzer> I think it goes the other way, tomorrow will be 10 degrees warmer than today (in C)
[25-Jan-2010 10:39:53] <rmatte> ah
[25-Jan-2010 10:40:12] <rmatte> that's crazy
[25-Jan-2010 10:40:14] <patzer> I lived in Ottawa for a winter once, I am happy to not have two weeks of -40 everyday again
[25-Jan-2010 10:40:26] <patzer> although I could also do without the endless grey skies
[25-Jan-2010 10:40:40] <rmatte> patzer: yeh, it gets quite frigid here. We've had some really cold days but overall this has been an extremely mild winter
[25-Jan-2010 10:41:21] <patzer> I was in 2003 I think
[25-Jan-2010 10:41:38] <patzer> I am also interested in this dynamic datapoint discussion
[25-Jan-2010 10:42:02] <patzer> what is the verdict, that it is not easily doable?
[25-Jan-2010 10:42:50] <rmatte> patzer: I'm sure it would be difficult to implement, but it would be extremely powerful to have
[25-Jan-2010 10:43:10] <rmatte> patzer: It would make it so that even the average user could generate powerful dynamic templates
[25-Jan-2010 10:43:35] <rmatte> patzer: unlikely to see it until quite a while after stone crab though
[25-Jan-2010 10:43:40] <patzer> my problem is more graph management
[25-Jan-2010 10:43:45] <rmatte> I'd imagine that they are shying away from too many UI enhancements
[25-Jan-2010 10:44:02] <patzer> right now we have boxes with ~16 slots, and 4-16 ports on each slot
[25-Jan-2010 10:44:03] <rmatte> patzer: graph management in what sense?
[25-Jan-2010 10:44:19] <patzer> but only two-three slots are used on each box
[25-Jan-2010 10:44:48] <patzer> so the problem for me is how I do setup it up so that people can add new boxes, and there is a sensible data collection and presentation for them
[25-Jan-2010 10:45:08] <patzer> I don't really want to poll 16*16 data points for every box when only maybe 10 are actually collecting data
[25-Jan-2010 10:45:49] <rmatte> patzer: that's why you disable monitoring on the ports which you don't want to collect data for
[25-Jan-2010 10:46:03] <patzer> well the ports aren't collecting interface data
[25-Jan-2010 10:46:06] <rhettardo> stupid autocomplete
[25-Jan-2010 10:46:15] <rmatte> rhettardo: yup
[25-Jan-2010 10:46:29] <patzer> they are collecting other things, we are an ISP so for example atm virtual circuits, and authenticated subscribers, etc
[25-Jan-2010 10:46:39] <rmatte> I see
[25-Jan-2010 10:46:56] <patzer> the oid structure is such that it isn't indexed by ifindex, but by the actual port number
[25-Jan-2010 10:47:08] <patzer> oid.9.0.3.subs, stuff like that
[25-Jan-2010 10:47:23] <patzer> but modelling the device to see what ones are used is no problem
[25-Jan-2010 10:48:59] <patzer> but you do give an idea, maybe there is a way to extend the interface template to include and poll stuff based on its name
[25-Jan-2010 10:49:23] <patzer> that way disabling the interface would indeed disable collecting on stuff that isn't used
[25-Jan-2010 10:50:28] <cparlette> just jumping in here so it might be incorrect, but sounds like you're talking about using zInterfaceMapIgnoreNames
[25-Jan-2010 10:51:34] <rmatte> yeh, it does sound like that
[25-Jan-2010 11:12:19] <venturaville> am I correct in my reading of the source code, that zenoss CommandPlugin only works with remote ssh commands?
[25-Jan-2010 11:12:33] <venturaville> (not to be confused with datasources)
[25-Jan-2010 11:12:43] <mrayzenoss> they can be local shell calls
[25-Jan-2010 11:12:51] * mrayzenoss was confused
[25-Jan-2010 11:15:15] <venturaville> in zenmodeler.py in cmdCollect if the transport protocol is anything other than ssh or telnet it generates an error.....
[25-Jan-2010 11:15:25] <Egyptian[laptop]> hello all
[25-Jan-2010 11:15:32] <Egyptian[laptop]> how do i disable network discovery?
[25-Jan-2010 11:16:24] <rmatte> hmmm, I'm telling Zenoss to drop an event and it's not dropping it
[25-Jan-2010 11:16:28] * rmatte scratches his head
[25-Jan-2010 11:16:42] <rmatte> Egyptian[laptop]: what do you mean by "network discovery"?
[25-Jan-2010 11:17:41] <rmatte> aha, had to add a transform to drop it rather than zProperties, weird
[25-Jan-2010 11:18:20] <Egyptian[laptop]> rmatte: the function that goes and does an search for all ip in a vlan registered with zenoss and then does snmp on that discovered ip
[25-Jan-2010 11:18:35] <rmatte> Egyptian[laptop]: Zenoss only does that if you specifically tell it to
[25-Jan-2010 11:18:39] <rmatte> there is no automation
[25-Jan-2010 11:19:11] <rmatte> or are you talking about route discovery?
[25-Jan-2010 11:19:16] <Egyptian[laptop]> rmatte: ???? .. on my machine its does it .. as far as i can see .. once a day ..
[25-Jan-2010 11:19:41] <Egyptian[laptop]> rmatte: i m no expert in zenoss
[25-Jan-2010 11:19:45] <rmatte> ok, describe in more detail what's happening
[25-Jan-2010 11:19:51] <rmatte> I really don't understand based on your explanation
[25-Jan-2010 11:20:29] <Egyptian[laptop]> rmatte: the symptoms are .. the networks guys call me and scream at me coz i m doing snmp "attacks" on the backbone and want me to switch off this function
[25-Jan-2010 11:20:41] <rmatte> you're probably talking about remodelling
[25-Jan-2010 11:20:44] <rmatte> not discovery
[25-Jan-2010 11:21:09] <Egyptian[laptop]> and the ip for the backbone is _not_ registered in zenoss
[25-Jan-2010 11:21:10] <rmatte> by default Zenoss goes out and attempts to remodel devices to see if there are any new components on them
[25-Jan-2010 11:21:30] <rmatte> nah, it shouldn't be doing modelling at all for devices that aren't in it's inventory
[25-Jan-2010 11:21:40] <Egyptian[laptop]> oh .. so someone added the backbone to zenoss is that what you are saying?
[25-Jan-2010 11:21:42] <rmatte> It shouldn't even be touching them with snmp
[25-Jan-2010 11:21:56] <rmatte> well, is the backbone in Zenoss?
[25-Jan-2010 11:22:04] <Egyptian[laptop]> well .. the password is public may be some one played with it .. its still in testing
[25-Jan-2010 11:22:11] <rmatte> I'd check
[25-Jan-2010 11:22:14] <Egyptian[laptop]> rmatte: will check and get back to you
[25-Jan-2010 11:22:18] <rmatte> k
[25-Jan-2010 11:23:28] <rmatte> If you do need to disable remodelling I can walk you through it, but it's better to have it turned on if you can manage it
[25-Jan-2010 11:24:09] <rmatte> I personally disable it since we monitor stuff across slow links and I don't need Zenoss going out and collecting a full snmp snapshot of each device every few hours
[25-Jan-2010 11:26:01] <patzer> venturaville: there is an option when you use a COMMAND datasource
[25-Jan-2010 11:26:34] <patzer> venturaville: if you set "Use SSH: False", then it will run locally on the zenoss host
[25-Jan-2010 11:27:03] <patzer> (but you can still pass options such as ${dev/manageIp} if you want to connect on your own to a device, via snmp or whatever else)
[25-Jan-2010 11:27:58] <venturaville> patzer: I am specifically talking about the modeler :-)
[25-Jan-2010 11:29:31] <rmatte> patzer: venturaville is actually coding a collector plugin
[25-Jan-2010 11:29:32] <patzer> the same thing applies
[25-Jan-2010 11:29:45] <rmatte> so it's not exactly the same
[25-Jan-2010 11:30:18] <rmatte> woohoo, I think I finally have one of my new Zenoss boxes tuned
[25-Jan-2010 11:30:31] <rmatte> the client had the noisient damn wireless lan controller I've ever seen
[25-Jan-2010 11:30:46] <rmatte> noisiest*
[25-Jan-2010 11:31:37] <patzer> ah sorry
[25-Jan-2010 11:31:55] <patzer> venturaville: in that case make your modeler a "PythonPlugin" and not a "CommandPlugin"
[25-Jan-2010 12:03:24] <venturaville> patzer: I am doing a PythonPlugin and calling out to the command manually, so we have come to the same conclusion :-)
[25-Jan-2010 12:19:19] <rmatte> quiet in here today lol
[25-Jan-2010 12:19:27] <rmatte> which is weird for a Monday
[25-Jan-2010 12:22:45] <patzer> venturaville: sure, I just meant which parent class you subclass from ;)
[25-Jan-2010 12:23:03] <venturaville> :-)
[25-Jan-2010 12:29:26] <rmatte> netsplit
[25-Jan-2010 12:42:43] <eidolon> odd, no one was on #znoss.
[25-Jan-2010 12:43:10] <rmatte> ?
[25-Jan-2010 12:43:15] <rmatte> netsplit
[25-Jan-2010 12:43:30] <rmatte> the irc server that you are on split from the network
[25-Jan-2010 12:43:48] <eidolon> dude, i was on the wrong channel name :)
[25-Jan-2010 12:43:51] <eidolon> it was a joke :)
[25-Jan-2010 12:43:53] <rmatte> oh
[25-Jan-2010 12:43:54] <rmatte> :P
[25-Jan-2010 12:44:05] <eidolon> so Rusty just said i need to run zenfixit to fix my CMD problem.
[25-Jan-2010 12:44:09] * eidolon is downloading now.
[25-Jan-2010 12:44:44] <rmatte> my god futureshop sucks sometimes... I've been calling and calling and no one ever answers
[25-Jan-2010 12:44:53] * rmatte grrr
[25-Jan-2010 12:45:11] <rmatte> zenfixit?
[25-Jan-2010 12:45:40] <eidolon> yeah
[25-Jan-2010 12:45:53] <eidolon> KBDocumentsDocument Name:Using ZenFixIt to Correct Common Zenoss Issues
[25-Jan-2010 12:45:55] <theacolyte> bleh
[25-Jan-2010 12:46:04] <theacolyte> None of my issues are resolved from my vacation
[25-Jan-2010 12:46:09] <theacolyte> I had hoped for fixes :(
[25-Jan-2010 12:46:13] <rmatte> lol
[25-Jan-2010 12:49:31] <rmatte> vacations don't generally generate work
[25-Jan-2010 12:49:34] <rmatte> erm
[25-Jan-2010 12:49:38] <rmatte> they generally generate work
[25-Jan-2010 12:49:39] <rmatte> rather
[25-Jan-2010 12:50:03] <rmatte> I wish I could go on vacation and come back to less work than when I left :)
[25-Jan-2010 12:51:17] <forsberg> :P
[25-Jan-2010 12:51:44] <forsberg> it happened for me.......in a parallel universe
[25-Jan-2010 12:52:05] <theacolyte> lol
[25-Jan-2010 13:07:02] <rmatte> Well, I'd like to take a vacation to said parallel universe
[25-Jan-2010 13:07:09] <rmatte> an extended vacation
[25-Jan-2010 13:11:50] <rmatte> mhmmmm, just heard something which sounded like a star trek phaser going off
[25-Jan-2010 13:12:06] <rmatte> probably someone's car alarm
[25-Jan-2010 13:14:21] <eidolon> welllllll, zenfixit didn't.
[25-Jan-2010 13:14:22] * eidolon grumps.
[25-Jan-2010 13:14:26] * eidolon schedules a call with rusty.
[25-Jan-2010 13:18:31] <rmatte> lol
[25-Jan-2010 14:03:07] <LarsN> Any idea why when attempting to install the zenpack "ZenPacks.zenoss.ZenJMX-3.3.1-py2.4.egg.zip" I would get a ERROR: zenpack command failed. Reason: [Errno 20] Not a directory: '/usr/local/zenoss/zenoss/Products/ZenPacks.zenoss.ZenJMX-3.3.1-py2.4.egg/skins'
[25-Jan-2010 14:03:28] <theacolyte> Unzip
[25-Jan-2010 14:03:32] <theacolyte> And then install the .egg
[25-Jan-2010 14:04:32] <LarsN> that appears to have worked, better :)
[25-Jan-2010 14:04:33] <LarsN> thanks.
[25-Jan-2010 14:10:02] <venturaville> looks like remote libvirt on redhat 5.4 has issues even to other redhat 5.x hosts
[25-Jan-2010 14:10:37] <venturaville> but works just fine when sourced from ubuntu 9.10
[25-Jan-2010 14:11:00] <venturaville> I'm guessing that is due to the difference in libvirt versions 0.6.3 -> 0.7.0
[25-Jan-2010 14:17:19] <sergeymasushko> hi guys. is it possible to rename all devices by a mask e.g. I need to rename "device_name.old_comapy_name.net" to "device_name.new_company_name.com"?
[25-Jan-2010 14:18:30] <rmatte> sergeymasushko: you could do that in zendmd
[25-Jan-2010 14:18:52] <sergeymasushko> rmatte: hm... can you suggest how?
[25-Jan-2010 14:19:00] <rmatte> well, let's see...
[25-Jan-2010 14:19:03] <rmatte> give me a few minutes
[25-Jan-2010 14:19:06] <sergeymasushko> sure
[25-Jan-2010 14:22:54] <rmatte> I just need to find a function to do the actual renaming
[25-Jan-2010 14:23:00] <rmatte> after I find that this'll be simple
[25-Jan-2010 14:23:09] <patzer> rmatte: device.rename('newname')
[25-Jan-2010 14:23:20] <rmatte> patzer: thanks
[25-Jan-2010 14:23:58] <rmatte> sergeymasushko: is it always devicename.domain.suf ?
[25-Jan-2010 14:24:11] <rmatte> you don't have any weird one like dev.ice.name.domain.suf?
[25-Jan-2010 14:24:41] <rmatte> (I'm just trying to figure out the best way to split the name out)
[25-Jan-2010 14:25:15] <patzer> probably could just use str.replace()
[25-Jan-2010 14:25:58] <rmatte> hmmm, good idea
[25-Jan-2010 14:27:44] <rmatte> so if I capture the device name to a variable called name...
[25-Jan-2010 14:28:03] <gwb2351> are people using / is this a good idea? http://community.zenoss.org/blogs/zenossblog/2009/05/28/tip-of-the-month-event-suppression ?
[25-Jan-2010 14:28:10] <rmatte> wondering how I can have str.replace work against that variable
[25-Jan-2010 14:29:11] <rmatte> gwb2351: doesn't Zenoss already do that now?
[25-Jan-2010 14:29:12] <venturaville> does anyone here have redhat running a libvirtd host that can help me test something?
[25-Jan-2010 14:29:33] <gwb2351> rmatte: not if the alert is being generated from a command (in a template)
[25-Jan-2010 14:29:44] <rmatte> oh I see
[25-Jan-2010 14:29:55] <rmatte> gwb2351: you'd have to use it in /
[25-Jan-2010 14:30:02] <rmatte> wouldn't make sense to apply it anywhere else
[25-Jan-2010 14:32:01] <rmatte> sergeymasushko: almost done
[25-Jan-2010 14:32:47] <rmatte> sergeymasushko: http://pastebin.com/m2988abf1
[25-Jan-2010 14:33:07] <rmatte> obviously replace old_comapy_name.net and new_company_name.com with whatever
[25-Jan-2010 14:35:57] <sergeymasushko> rmatte: thanks a lot.
[25-Jan-2010 14:36:32] <sergeymasushko> rmatte: one more question: will all old rrd files continue to work after this change?
[25-Jan-2010 14:36:53] <rmatte> sergeymasushko: yes, they will
[25-Jan-2010 14:37:03] <rmatte> the only thing is, they'll still be in folders with the old names
[25-Jan-2010 14:37:11] <rmatte> the directory names don't change (though I wish they did)
[25-Jan-2010 14:38:10] <rmatte> but your main device names aren't changing, only the domain, so finding them won't really be an issue for you
[25-Jan-2010 14:41:17] <sergeymasushko> rmatte: if I reanme device through the web interface will it change the directory name ?
[25-Jan-2010 14:41:45] <rmatte> nope
[25-Jan-2010 14:41:57] <rmatte> it's the exact same function as the web interface uses
[25-Jan-2010 14:42:13] <rmatte> the directory names are set when you first add a device and don't change
[25-Jan-2010 14:42:24] <venturaville> I found a work around for the redhat libvirt hosts, but I am not sure it works everywhere
[25-Jan-2010 14:42:32] <venturaville> anyone have redhat 5.x?
[25-Jan-2010 14:42:37] <ckrough> 5.4
[25-Jan-2010 14:42:50] <sergeymasushko> ok, so to rename devices I need to run zendmd and then paste code from pastebin?
[25-Jan-2010 14:42:51] <venturaville> ckrough: do you have libvirt installed anywhere?...
[25-Jan-2010 14:43:15] <ckrough> venturaville: at home I was running it with KVM on centos 5.4, but not at work
[25-Jan-2010 14:43:15] <rmatte> sergeymasushko: right, but like I said, change the old and new domains in the code first
[25-Jan-2010 14:43:24] <rmatte> or else it's not going to work
[25-Jan-2010 14:43:32] <sergeymasushko> rmatte: sure =)
[25-Jan-2010 14:43:46] <rmatte> also, do a commit()
[25-Jan-2010 14:43:48] <rmatte> when you're done
[25-Jan-2010 14:44:00] <sergeymasushko> thanks
[25-Jan-2010 14:44:38] <sergeymasushko> rmatte: do I need to restart something when it will be done?
[25-Jan-2010 14:44:53] <rmatte> nope
[25-Jan-2010 14:45:06] <sergeymasushko> rmatte: ok, thank you a lot :)
[25-Jan-2010 14:45:13] <rmatte> no problem
[25-Jan-2010 14:58:16] <rocket> rmatte: comeon quite being so nice and helping people .. :p
[25-Jan-2010 15:00:26] <LarsN> Helping people?
[25-Jan-2010 15:00:28] <LarsN> that's allowed?
[25-Jan-2010 15:00:29] <LarsN> :)
[25-Jan-2010 15:00:37] <rmatte> rocket: :P
[25-Jan-2010 15:00:56] <LarsN> generally speaking, #zenoss is one of the more helpful rooms on Freenode :)
[25-Jan-2010 15:01:02] <LarsN> in my experience anyway.
[25-Jan-2010 15:01:05] <rmatte> we try
[25-Jan-2010 15:01:08] <Apachez> LarsN: ever tried a perl channel? ;)
[25-Jan-2010 15:01:11] <rmatte> well, I do anyways
[25-Jan-2010 15:01:14] <ckrough> lol
[25-Jan-2010 15:01:18] <LarsN> I try not to Perl.
[25-Jan-2010 15:01:36] <LarsN> it's bad enough we just stuck Java in zenoss.
[25-Jan-2010 15:01:39] <Egyptian[laptop]> how do i register an rfe?
[25-Jan-2010 15:01:41] <rocket> I think I had a similar experience in a python channel ..
[25-Jan-2010 15:01:43] <ckrough> s/perl/python/g
[25-Jan-2010 15:01:45] <LarsN> if we add perl, we will have completed the unholy trinity.
[25-Jan-2010 15:01:46] <ckrough> ya
[25-Jan-2010 15:01:47] <rmatte> #python is a nightmare, every time I ask a question in there it's met with attitude
[25-Jan-2010 15:01:52] <LarsN> at which point the world will end.
[25-Jan-2010 15:02:17] <ckrough> seems to be the way with most channels
[25-Jan-2010 15:02:26] <Egyptian[laptop]> LarsN: only when you add oracle support ;)
[25-Jan-2010 15:02:43] <rmatte> "Yeh, the topic says we're a help channel... but we're not going to help you"
[25-Jan-2010 15:02:45] <ckrough> we would probably do it here as well if we had the same volume... those popular channels get flooded with RTFM type stuff
[25-Jan-2010 15:02:57] <rmatte> "Oh, and don't paste more than 3 lines because that's flooding and you'll get banned"
[25-Jan-2010 15:03:00] <LarsN> #ubuntu attempts to be helpful. although determining who's an idiot and who's actually smart can be more difficult.
[25-Jan-2010 15:03:07] <rmatte> one time I pasted 4 lines and the channel went in to an uproar
[25-Jan-2010 15:03:13] <ckrough> people in here have been great about researching their issue and showing up prepared
[25-Jan-2010 15:03:13] <rmatte> and I was like "Seriously guys!?"
[25-Jan-2010 15:03:30] <ckrough> try #fedora-admin
[25-Jan-2010 15:03:31] <LarsN> Egyptian[laptop]: I don't swing that way. (oracle) :)
[25-Jan-2010 15:03:31] <Egyptian[laptop]> rmatte: umm.. fedora has an fpaste tool that automatically pastes to fpaste.org
[25-Jan-2010 15:03:48] <rmatte> ckrough: not all people, I've had to tell people to RTFM a few times
[25-Jan-2010 15:04:02] <venturaville> kvm and libvirt channels get like that ...lots of wars about who is reponsible for what....
[25-Jan-2010 15:04:05] <a21> LarsN: I've spen a few hours yesterday on #ubuntu and there was nobody who even tried to help..
[25-Jan-2010 15:04:09] <rmatte> Egyptian[laptop]: so does #python
[25-Jan-2010 15:04:20] <ckrough> rmatte: of course, in their defense, zenoss needs to WTFM too :)
[25-Jan-2010 15:04:44] <rmatte> ckrough: the admin guide is pretty good though, and it covers a LOT of stuff
[25-Jan-2010 15:05:10] <rmatte> when people come in here and ask really basic stuff about creating device templates I always tell them to have a read
[25-Jan-2010 15:05:10] <ckrough> yes. documentation seems to have improved a lot over the past year or so
[25-Jan-2010 15:05:15] <LarsN> a21: that seems to be the exception. but you might be right :)
[25-Jan-2010 15:05:15] * Egyptian[laptop] hears someone talking about a concerted documentation effort
[25-Jan-2010 15:05:46] <LarsN> I'll admit, Zenoss annoys me more often than I'm plesently surprised by it.
[25-Jan-2010 15:05:47] <rmatte> ckrough: but there are tons of little tricks and secrets that I've learned along the way which aren't documented. That's the stuff I generally end up helping people with
[25-Jan-2010 15:06:03] <LarsN> then again I've not found a monitoring system that didn't annoy me. so Zenoss isn't exactly singled out. :)
[25-Jan-2010 15:06:31] <rmatte> LarsN: I don't find it overly annoying. There are things that I find from time to time which drive me nuts but they are great about fixing it.
[25-Jan-2010 15:06:36] <ckrough> is there a general network monitoring channel out there?
[25-Jan-2010 15:06:42] <rocket> rmatte: no secrets ... quit holding out on me .. :p
[25-Jan-2010 15:06:48] <LarsN> I'll use an example from today.
[25-Jan-2010 15:06:54] <LarsN> add a device. set status to pre-production
[25-Jan-2010 15:06:56] <rmatte> rocket: lol
[25-Jan-2010 15:07:11] <LarsN> once it's finished adding, the device decided to be in production..... and start alerting.
[25-Jan-2010 15:07:20] <LarsN> not a huge deal, but annoying. ;)
[25-Jan-2010 15:07:23] <rmatte> LarsN: I've never had that happen to me
[25-Jan-2010 15:07:31] <rmatte> LarsN: and I add devices in Pre-Production all the time
[25-Jan-2010 15:07:33] <Apachez> I will go for zenoss 100% once zenoss gets rid of zope :)
[25-Jan-2010 15:07:50] <rmatte> LarsN: you sure you didn't accidentally change the setting to production somehow before clicking add?
[25-Jan-2010 15:07:53] <rocket> Apachez: uhmm keep dreaming ... that will probably never happen
[25-Jan-2010 15:08:16] <rmatte> Apachez: that won't happen, the entire product is built on zope
[25-Jan-2010 15:08:32] <rocket> Apachez: they are leveraging zope features at this point not found in other cms frameworks .. eg they have looked at django etc and it doesnt have the feature set they need at this point
[25-Jan-2010 15:08:33] <rmatte> Apachez: that's like telling Microsoft to get rid of NT
[25-Jan-2010 15:08:42] <rocket> Apachez: too much is written against it ..
[25-Jan-2010 15:08:57] <ckrough> I think they are going to .NET in 3.0
[25-Jan-2010 15:09:13] <rmatte> ckrough: I'll believe it when I see it
[25-Jan-2010 15:09:13] <ckrough> lol
[25-Jan-2010 15:09:18] <ckrough> no way.
[25-Jan-2010 15:09:32] <ckrough> coldfusion all the way!
[25-Jan-2010 15:09:48] <rmatte> meh
[25-Jan-2010 15:10:23] <venturaville> coldfusion *choke* *gasp* *~-~-~....*
[25-Jan-2010 15:10:37] <rmatte> well, I just cleared swap on one of my Zenoss boxes, now to watch and see how long it takes before it starts massively munching swap space again
[25-Jan-2010 15:11:46] <ckrough> rmatte: I set up low swap thresholds on the box that is having that issue and I've been tweaking the zope threads and cache settings... lowering them seems to increase the time between swap issues
[25-Jan-2010 15:12:00] <ckrough> rmatte: so I'm still looking into the sanity of my zope.conf settings. FYI.
[25-Jan-2010 15:12:13] <rmatte> ckrough: well, keep at it, maybe with enough troubleshooting we can pinpoint the cause
[25-Jan-2010 15:12:14] <venturaville> coldfusion isn't that bad I supposed, but I haven't seen it scale very far
[25-Jan-2010 15:12:33] <rmatte> ckrough: because this is majorly detrimental to Zenoss performance
[25-Jan-2010 15:12:43] <ckrough> venturaville: I think it's abandoned at this point
[25-Jan-2010 15:12:45] <forsberg> <3 zenoss btw
[25-Jan-2010 15:12:50] <ckrough> venturaville: it was OK back in the day
[25-Jan-2010 15:12:52] <rmatte> forsberg: glad to hear
[25-Jan-2010 15:14:40] <LarsN> my whole job surrounds hosting zope.
[25-Jan-2010 15:14:41] <LarsN> ;)
[25-Jan-2010 15:14:47] <LarsN> primarily for Plone sites.
[25-Jan-2010 15:15:05] <rmatte> I remember this simple coldfusion attack where you'd go to the coldfusion login page, save the html file, modify the form length for password to be like a billion characters, then paste that many in and click submit... you'd pin the box at 100% CPU
[25-Jan-2010 15:15:06] <LarsN> which means I'm even stuck supporting old Zope.
[25-Jan-2010 15:15:10] <ckrough> LarsN: do you know of any good resources for tuning zope?
[25-Jan-2010 15:15:27] <LarsN> ckrough: we rarely tune zope.
[25-Jan-2010 15:15:35] <LarsN> the only thing we set is threads down to maybe 2.
[25-Jan-2010 15:15:40] <venturaville> it works quite well if you have fast disk behind it
[25-Jan-2010 15:15:44] <LarsN> and then run more zopes.
[25-Jan-2010 15:15:50] <Apachez> threads down to 1 seems to be the option for zope
[25-Jan-2010 15:15:50] <ckrough> venturaville: what does?
[25-Jan-2010 15:15:54] <LarsN> with something like haproxy or pound in front of it.
[25-Jan-2010 15:15:56] <LarsN> them
[25-Jan-2010 15:15:58] <venturaville> ckrough: the zope db
[25-Jan-2010 15:16:13] <venturaville> lots of memory helps too....
[25-Jan-2010 15:16:18] <rmatte> I have threads set at 50, think that's too high?
[25-Jan-2010 15:16:19] <LarsN> lots of memory.
[25-Jan-2010 15:16:22] <LarsN> yes
[25-Jan-2010 15:16:23] <rmatte> (this is on a busy busy box)
[25-Jan-2010 15:16:26] <LarsN> by about 48/49
[25-Jan-2010 15:16:32] <LarsN> less threads, more zopes.
[25-Jan-2010 15:16:37] <LarsN> two zopes per core.
[25-Jan-2010 15:16:38] <ckrough> venturaville: I have some fairly big boxes running it, but with the number of devices we monitor walking the zope db gets slow
[25-Jan-2010 15:16:49] <rmatte> hmmm, I might try bringing that down them
[25-Jan-2010 15:16:51] <rmatte> then*
[25-Jan-2010 15:17:00] <venturaville> ckrough: how many devices?
[25-Jan-2010 15:17:06] <LarsN> plone.org has 8 physical cores, is running 8 zope instances, with Pound, and Varnish.
[25-Jan-2010 15:17:06] <rmatte> by threads you mean the zserver-threads option in the config right?
[25-Jan-2010 15:17:06] <ckrough> 12k
[25-Jan-2010 15:17:33] <Apachez> what does splunk use?
[25-Jan-2010 15:17:42] <Apachez> since splunk also have python as backend
[25-Jan-2010 15:17:53] <Apachez> and is gazillion times faster than zenoss
[25-Jan-2010 15:17:57] <Apachez> and takes less mem
[25-Jan-2010 15:18:29] <rmatte> hmmm, I'll try knocking the threads down to 5
[25-Jan-2010 15:18:31] <rmatte> erm 4
[25-Jan-2010 15:18:37] <ckrough> LarsN: 25 threads, cache-size 200000, pool-size 75 on an 8 core 32GB box
[25-Jan-2010 15:19:07] <LarsN> ckrough: which zope?
[25-Jan-2010 15:19:16] <LarsN> version.
[25-Jan-2010 15:19:51] <ckrough> 2.8.8
[25-Jan-2010 15:20:25] <LarsN> so, python 2.3?
[25-Jan-2010 15:20:30] <ckrough> 2.4.3
[25-Jan-2010 15:20:36] <LarsN> nod
[25-Jan-2010 15:20:45] <venturaville> ckrough: what kind of disk I/O are you doing on it?
[25-Jan-2010 15:20:54] <LarsN> we skipped 2.8 all together most of our sites are 2.7 or 2.9/2.10
[25-Jan-2010 15:21:45] <ckrough> venturaville: that box is *just* the zenoss ui and zope, no performance collection, its a raid1+0 on 6 spindles
[25-Jan-2010 15:21:46] <venturaville> Apachez: splunk can be used in conjunction with zenoss to do pre-filtering of logs
[25-Jan-2010 15:21:46] <rmatte> I'll test with 4 threads to make sure that the box can still keep up with monitoring
[25-Jan-2010 15:22:13] <LarsN> ckrough: are you running 1 zope instace then?
[25-Jan-2010 15:22:23] <LarsN> or are you running a zeo cluster, with a detached zodb
[25-Jan-2010 15:22:40] <ckrough> LarsN: single, default Zenoss architectyre
[25-Jan-2010 15:22:51] <LarsN> I ask because zope is single threaded by default.
[25-Jan-2010 15:22:58] <LarsN> so it'll eat 100% of one core. :(
[25-Jan-2010 15:23:03] <Apachez> venturaville: doesnt help when you get a cardiac arrest each time you need to get into the zenoss slow gui :P
[25-Jan-2010 15:23:11] <LarsN> that's typically why we "go wide"
[25-Jan-2010 15:23:21] <LarsN> using pound or haproxy between the web, and "zope"
[25-Jan-2010 15:26:44] <ckrough> LarsN: thanks. I;ve been planning on checking out haproxy in front of it, also some sort of caching
[25-Jan-2010 15:27:17] <LarsN> ckrough: the hosting stack we just built for a customer is something like this.
[25-Jan-2010 15:27:33] <rmatte> well, guess I'll leave this at 4 threads and see where I'm at next week
[25-Jan-2010 15:28:09] <LarsN> [internet] <--> [nginx :80/443] <--> [varnish :3180] <--> [nginx :3280] <--> [haproxy :3380] <--> [zope^n] <--> [MySQL]
[25-Jan-2010 15:30:43] <LarsN> rmatte: it didn't fall over I take it
[25-Jan-2010 15:30:56] <rmatte> LarsN: doesn't seem like it
[25-Jan-2010 15:31:13] <LarsN> rmatte: :)
[25-Jan-2010 15:31:26] <LarsN> rmatte: I've never seen lowering the threads kill zope
[25-Jan-2010 15:31:26] <rmatte> Murphy's law dictates that it'll happen the second I walk out the door though
[25-Jan-2010 15:31:31] <LarsN> I HAVE seen the reverse happen.
[25-Jan-2010 15:31:47] <LarsN> one customer set his zope threads to, maybe 250?.
[25-Jan-2010 15:32:05] <rmatte> LarsN: I'm monitoring over 18000 datapoints, on 354 devices, so I was worried if it'd still be able to keep up
[25-Jan-2010 15:32:14] <rmatte> but apparently it can
[25-Jan-2010 15:32:25] <LarsN> rmatte: zope isn't doing the monitoring, so you should be fine :)
[25-Jan-2010 15:32:53] <rmatte> yeh I know, but it's accessing zope while monitoring
[25-Jan-2010 15:33:45] <LarsN> rmatte: if Zenoss can be configured with a Zeo Cluster
[25-Jan-2010 15:33:48] <LarsN> that'd be ideal
[25-Jan-2010 15:33:56] <LarsN> multiple zopes running in parallel.
[25-Jan-2010 15:34:05] <rmatte> not sure if it can or not
[25-Jan-2010 15:34:16] <LarsN> that is how we host high traffic sites. I might have to set up a zenoss instance at home and test it.
[25-Jan-2010 15:34:22] <LarsN> I've got an 8 core desktop. it can take it ;)
[25-Jan-2010 15:34:26] <rmatte> hehe
[25-Jan-2010 15:34:37] <LarsN> if it works I'll write a blog post about how I did it and share here.
[25-Jan-2010 15:35:24] <Muti> does anyone know how to return from an event mapping transform (short-circuiting the rest of the transform)?
[25-Jan-2010 15:35:29] <rmatte> cool
[25-Jan-2010 15:35:39] <rmatte> Muti: continue
[25-Jan-2010 15:36:02] <Muti> simple enough, thanks. =)
[25-Jan-2010 15:36:14] <rmatte> "The continue statement, also borrowed from C, continues with the next iteration of the loop. "
[25-Jan-2010 15:36:23] <Muti> hmm, it didn't like that
[25-Jan-2010 15:36:30] <rmatte> ok, pastebin the code
[25-Jan-2010 15:36:37] <rmatte> and explain where you're trying to have it exit from
[25-Jan-2010 15:37:05] <rmatte> you could try: break
[25-Jan-2010 15:37:09] <Muti> http://pastey.net/131901-37bh
[25-Jan-2010 15:37:27] <rmatte> you could also try...
[25-Jan-2010 15:37:28] <Muti> basically, I want to be able to check for sane values before processing the event
[25-Jan-2010 15:37:29] <rmatte> import sys
[25-Jan-2010 15:37:33] <rmatte> sys.exit()
[25-Jan-2010 15:37:57] <Muti> will that exit the parent process though?
[25-Jan-2010 15:38:11] <rmatte> I'd try break
[25-Jan-2010 15:38:22] <rmatte> looks like it should work
[25-Jan-2010 15:38:25] <Muti> doesn't like break either. =/
[25-Jan-2010 15:38:40] <Muti> I tried break, continue, return
[25-Jan-2010 15:38:51] <rmatte> then try sys.exit
[25-Jan-2010 15:38:56] <rmatte> it shouldn't kill the parent process
[25-Jan-2010 15:39:27] <Muti> ok, I'll give it a shot
[25-Jan-2010 15:39:37] <rmatte> k
[25-Jan-2010 15:39:56] <rmatte> actually hold up
[25-Jan-2010 15:40:00] <rmatte> is that your entire transform?
[25-Jan-2010 15:40:03] <rmatte> or just part of it?
[25-Jan-2010 15:40:36] <rmatte> I'm just wondering if your logic is flawed
[25-Jan-2010 15:40:42] <rmatte> for example, why would you not just do:
[25-Jan-2010 15:40:48] <rmatte> if trapType:
[25-Jan-2010 15:40:51] <Muti> that's just the first part...
[25-Jan-2010 15:40:53] <rmatte> and then put the rest of what you want to do?
[25-Jan-2010 15:40:55] <Muti> and yeah, I could do that
[25-Jan-2010 15:41:04] <rmatte> then do that, much safer and simpler
[25-Jan-2010 15:41:14] <Muti> I just don't like having to nest if I can short-circuit. =)
[25-Jan-2010 15:41:26] <rmatte> meh, if it works it works
[25-Jan-2010 15:41:29] <Muti> yeah
[25-Jan-2010 15:41:37] <Muti> I'd rather do that than sys.exit()
[25-Jan-2010 15:41:41] <rmatte> yeh
[25-Jan-2010 15:43:52] LarsN is now known as LarsN_Away
[25-Jan-2010 15:47:29] <rmatte> I'm going to edit this document tomorrow... a lot of the advice in it is flawed: http://community.zenoss.org/docs/DOC-2521
[25-Jan-2010 15:47:57] <rmatte> parallel 4 actually decreases, not increases since the default is 10
[25-Jan-2010 15:48:03] <rmatte> but I've found that 5 is the magic number with that
[25-Jan-2010 15:48:16] <rmatte> zserver-threads should be left at the default of 4
[25-Jan-2010 15:48:21] <rmatte> pool-size is probably fine
[25-Jan-2010 15:48:32] <rmatte> workers should really be 4, 8 is overkill
[25-Jan-2010 15:48:43] <cparlette> making zenhub workers 8 as a blanket rule is bad
[25-Jan-2010 15:48:50] <rmatte> yup
[25-Jan-2010 15:48:54] <Muti> I was just looking at that document earlier today
[25-Jan-2010 15:48:57] <cparlette> really any blanket number is bad
[25-Jan-2010 15:49:02] <cparlette> its highly dependant on your system
[25-Jan-2010 15:49:26] <rmatte> right, but the document is useful, but it's better to shoot low than to shoot very high
[25-Jan-2010 15:49:39] <rmatte> I mean, the zserver-threads at 150 is a bit ludicrous
[25-Jan-2010 15:50:03] <rmatte> anyways, I'll mod it tomorrow
[25-Jan-2010 16:03:03] <forsberg> i have a question, i have a couple of game servers one with qw and one with l4d2, now there is several ports on each server, and i cannot figure out wich way is bestest to show them in a graph
[25-Jan-2010 16:03:44] <forsberg> one is monitored with a sh script returning the numbers, the other new one is with snmp oid perl script thing
[25-Jan-2010 16:04:08] <forsberg> but with 4 ports on one graph the lines are on each other, and i tried with area/stacking but.... hmz!
[25-Jan-2010 16:04:14] <rmatte> forsberg: so you create a command based datasource which runs the script
[25-Jan-2010 16:04:28] <rmatte> then you have the script output like OK|value1=67 value2=73
[25-Jan-2010 16:04:37] <forsberg> yes thats the old solution
[25-Jan-2010 16:04:41] <rmatte> then you create 2 datapoints called value1 and value2 under the datasource (case sensitive)
[25-Jan-2010 16:04:49] <forsberg> and the new solution is proper snmp oid
[25-Jan-2010 16:04:57] <rmatte> well, that's the easiest solution
[25-Jan-2010 16:05:07] <rmatte> what's the issue with doing it that way?
[25-Jan-2010 16:05:10] <forsberg> sec
[25-Jan-2010 16:06:12] <forsberg> http://hesten.foppa.dk/files/Capture.JPG
[25-Jan-2010 16:06:29] <forsberg> the lines are ontop of each other etc, just interested in if there was a better way to show it :)
[25-Jan-2010 16:12:28] <rmatte> split it up in to 4 different graphs?
[25-Jan-2010 16:12:39] <rmatte> or do you want them stacked?
[25-Jan-2010 16:12:50] <rmatte> (like one on top of the other)
[25-Jan-2010 16:13:12] <rmatte> also, is that counting number of connections?
[25-Jan-2010 16:15:02] <rmatte> if that's number of connections then stacked would be the best option
[25-Jan-2010 16:15:45] * rmatte waits for a response
[25-Jan-2010 16:16:00] <forsberg> sorry, no its taken from qstat
[25-Jan-2010 16:16:02] <forsberg> and greb and cut
[25-Jan-2010 16:16:13] <rmatte> so what does it represent?
[25-Jan-2010 16:16:22] <forsberg> players on each server( port )
[25-Jan-2010 16:16:28] <rmatte> number of players right?
[25-Jan-2010 16:16:30] <forsberg> and 4 graphs yeah...but to much
[25-Jan-2010 16:16:36] <forsberg> number of players on each yes
[25-Jan-2010 16:16:49] <rmatte> so what you really want is just an overall tally of how many players are connected overall?
[25-Jan-2010 16:17:10] <forsberg> yes
[25-Jan-2010 16:17:13] <rmatte> simple...
[25-Jan-2010 16:17:21] <rmatte> one second, just loading a template to walk you through this
[25-Jan-2010 16:17:31] <rmatte> go to the graph properties
[25-Jan-2010 16:17:51] <rmatte> click on the first graphpoint
[25-Jan-2010 16:18:02] <forsberg> yeah just wanna stack them?
[25-Jan-2010 16:18:14] <rmatte> well, and make one other adjustment
[25-Jan-2010 16:18:17] <rmatte> just bare with me
[25-Jan-2010 16:18:24] <forsberg> oki doki yes
[25-Jan-2010 16:18:26] <rmatte> click on a graph point
[25-Jan-2010 16:18:44] <rmatte> where it says format, it should say %5.2lf%s, correct?
[25-Jan-2010 16:18:55] <rmatte> change that to %5.0lf
[25-Jan-2010 16:18:57] <forsberg> yup
[25-Jan-2010 16:18:59] <forsberg> ok
[25-Jan-2010 16:19:13] <rmatte> (that'll eliminate the .00 or whatever after the number, since you can't have half a player
[25-Jan-2010 16:19:20] <forsberg> ahh :P
[25-Jan-2010 16:19:24] <rmatte> for line type select area
[25-Jan-2010 16:19:30] <rmatte> and set stacked to true
[25-Jan-2010 16:19:38] <rmatte> then do the same for the other 3 graphpoints
[25-Jan-2010 16:19:41] <forsberg> yes yes excellent
[25-Jan-2010 16:19:47] <rmatte> and take a look at your graph, it will be much spiffier
[25-Jan-2010 16:20:27] <rmatte> post a screenshot after, curious to compare
[25-Jan-2010 16:20:39] <cparlette> rmatte: i was just about to ask for that too
[25-Jan-2010 16:20:49] <cparlette> as a gamer, that graph is pretty sweet
[25-Jan-2010 16:20:50] <theacolyte> I will never have pretty graphs rmatte
[25-Jan-2010 16:20:52] <mrayzenoss> BTW, the Zenoss community needs an L4D2 ZenPack :)
[25-Jan-2010 16:20:52] <theacolyte> never :(
[25-Jan-2010 16:20:59] <forsberg> lol :P
[25-Jan-2010 16:21:02] <rmatte> theacolyte: lol
[25-Jan-2010 16:21:12] <theacolyte> mrayzenoss personally hosed and closed my ticket!
[25-Jan-2010 16:21:17] <forsberg> well im hehehe i can make mine a zenpack, its very simple tho
[25-Jan-2010 16:21:26] <rmatte> theacolyte: why?
[25-Jan-2010 16:21:37] <theacolyte> hehe, it's a dupe apparantly, but didn't see the fix
[25-Jan-2010 16:21:42] <rmatte> ah
[25-Jan-2010 16:21:42] <theacolyte> wrt the transparency
[25-Jan-2010 16:21:46] <rmatte> yeh
[25-Jan-2010 16:22:00] <theacolyte> I could probably just manually change the RRD files
[25-Jan-2010 16:22:15] <rmatte> probably, seems a bit extreme though
[25-Jan-2010 16:22:29] <rmatte> I personally like the transparency
[25-Jan-2010 16:22:32] <rmatte> but to each his own
[25-Jan-2010 16:22:54] <theacolyte> I don't see how you could actually get it to work unless you happened to have a graph that al3ways overlapped
[25-Jan-2010 16:23:11] <rmatte> theacolyte: for CPU graphs it works fine
[25-Jan-2010 16:23:15] <rmatte> theacolyte: same for memory
[25-Jan-2010 16:23:35] <theacolyte> http://dev.zenoss.org/trac/attachment/ticket/6037/transparency.png
[25-Jan-2010 16:23:38] <rmatte> you can mix area and line graphpoints too
[25-Jan-2010 16:23:45] <rmatte> to get the desired look
[25-Jan-2010 16:23:58] <rmatte> such as they've done with traffic graphs
[25-Jan-2010 16:24:07] <theacolyte> It's going to sound sad, but that's the only reason why I haven't released any zenpacks
[25-Jan-2010 16:24:17] <rmatte> lol
[25-Jan-2010 16:24:25] <rmatte> stickler for graph quality
[25-Jan-2010 16:24:29] <forsberg> now the graphs not drawing ;(
[25-Jan-2010 16:24:30] <theacolyte> yeah
[25-Jan-2010 16:24:39] <rmatte> forsberg: seriously!?
[25-Jan-2010 16:24:44] * rmatte scratches head
[25-Jan-2010 16:24:50] <forsberg> hehe gonan double check
[25-Jan-2010 16:24:53] <forsberg> dont scratch yet
[25-Jan-2010 16:24:58] <rmatte> you sure you didn't mess up the format string somewhere?
[25-Jan-2010 16:25:19] <rmatte> messy format strings == no graph
[25-Jan-2010 16:25:22] <mrayzenoss> I'm serious about putting the ZenPack up, we actually have a number of game companies using Zenoss
[25-Jan-2010 16:25:34] <rmatte> nice
[25-Jan-2010 16:25:46] <mrayzenoss> if you think about it, most of them are MSPs
[25-Jan-2010 16:26:11] <theacolyte> mrayzenoss: I wish I had time to drop by when I was out there
[25-Jan-2010 16:26:17] <forsberg> mrayzenoss mine?
[25-Jan-2010 16:26:18] <theacolyte> I didn't even have time to see *my* friends
[25-Jan-2010 16:26:42] <mrayzenoss> forsberg: yeah, it'll be nice to have the words "Left 4 Dead" on the ZenPacks page :)
[25-Jan-2010 16:26:58] <forsberg> even nicer with my name
[25-Jan-2010 16:27:01] <forsberg> jk
[25-Jan-2010 16:27:10] <mrayzenoss> of course we list the author
[25-Jan-2010 16:27:40] <rmatte> lol
[25-Jan-2010 16:27:48] <rmatte> provided he can get the graph to draw
[25-Jan-2010 16:27:50] <rmatte> :P
[25-Jan-2010 16:27:53] <theacolyte> I don't want my name attached to ugly graphs!
[25-Jan-2010 16:28:02] <theacolyte> I have my color scheme all picked out
[25-Jan-2010 16:28:12] <forsberg> but i dont feel its done, graphs are ugly ( not drawing) and i use either a sh script or perl and snmp conf and both solutions need qstat
[25-Jan-2010 16:28:15] <mrayzenoss> heh, I can take "anonymous" ones as well
[25-Jan-2010 16:28:33] <rmatte> I think my graphs look pretty good: http://dmon.org/graphics/zenoss/windowsperf.png
[25-Jan-2010 16:28:44] <mrayzenoss> rmatte: very pretty
[25-Jan-2010 16:28:47] <forsberg> and im not sure how to make it universal, i kinda hardcore how many ports i haev going
[25-Jan-2010 16:28:58] <forsberg> not hardcore
[25-Jan-2010 16:29:01] <forsberg> hardcode!
[25-Jan-2010 16:29:08] <rmatte> mrayzenoss: and I'll finally be able to release my updated packs once 2.5.2 is out
[25-Jan-2010 16:29:11] <mrayzenoss> forsberg: it's a good example though, other users could customize as necessary
[25-Jan-2010 16:29:21] <theacolyte> Those graphs are fine to look at, but not that great for information
[25-Jan-2010 16:29:35] <rmatte> mrayzenoss: and then I need to figure out how to write a collector plugin for it
[25-Jan-2010 16:29:46] <rmatte> theacolyte: they are fine for information
[25-Jan-2010 16:30:54] <forsberg> ok, gotcha, i will try to make it into something
[25-Jan-2010 16:30:54] <rmatte> theacolyte: pretty easy to figure out which processor/core is spiking a lot, the memory is very easy to read
[25-Jan-2010 16:31:18] <theacolyte> I guess it does work better in your scenario
[25-Jan-2010 16:31:22] <rmatte> yeh
[25-Jan-2010 16:32:05] <theacolyte> I think I may just release this zenpack, but remember this is the one where the OID is specific to the device
[25-Jan-2010 16:32:15] <rmatte> right
[25-Jan-2010 16:32:23] <rmatte> you'll have to put instructions on the page
[25-Jan-2010 16:32:33] <rmatte> or write a script that does the work
[25-Jan-2010 16:32:56] <theacolyte> do you know of any zenpacks I can steal a script from?
[25-Jan-2010 16:33:06] <rmatte> my Windows Performance Monitor packs
[25-Jan-2010 16:33:22] <rmatte> they'll just give you an idea of how I did it though
[25-Jan-2010 16:33:48] <rmatte> but it's honestly not that difficult
[25-Jan-2010 16:34:57] <rmatte> forsberg: any luck with the graph?
[25-Jan-2010 16:35:15] <theacolyte> rmatte: Yes, but I'm an idiot, so it's easy for yo uto say!
[25-Jan-2010 16:35:17] <theacolyte> :P
[25-Jan-2010 16:35:21] <rmatte> lol
[25-Jan-2010 16:36:26] <theacolyte> man, I can't believe this
[25-Jan-2010 16:36:57] <theacolyte> but it looks like you can't get specific error #'s from cisco routers
[25-Jan-2010 16:37:05] <theacolyte> just aggregates
[25-Jan-2010 16:37:06] <rmatte> ?
[25-Jan-2010 16:37:17] <theacolyte> 73072 input errors, 73066 CRC, 12806 frame, 4814 overrun, 0 ignored, 11203 abort
[25-Jan-2010 16:37:23] <theacolyte> and yes that's a horribly functioning T1
[25-Jan-2010 16:37:49] <rmatte> well yeh, it derives from the last amount
[25-Jan-2010 16:37:53] <rmatte> it's the only way you can graph it
[25-Jan-2010 16:38:05] <rmatte> you could do a tally but you'd have to do some rrd magic or something
[25-Jan-2010 16:38:34] <theacolyte> I need to figure out how to set a threshold to tell me when that T1 is having issues
[25-Jan-2010 16:38:49] <rmatte> just set a threshold against the error OIDs
[25-Jan-2010 16:39:02] <forsberg> yeah i had %%5.0lf in one of em :)
[25-Jan-2010 16:39:07] <rmatte> I set mine at 300 * 0.07 which is 21 errors per second over a period
[25-Jan-2010 16:39:12] <rmatte> you might want to set yours lower
[25-Jan-2010 16:39:21] <rmatte> anyways, I'm done for the day, talk to you guys later
[25-Jan-2010 16:40:05] <theacolyte> later man
[25-Jan-2010 16:47:35] <forsberg> what to put in as Max Y in graph definition if i want it to auto max?
[25-Jan-2010 16:48:16] <mrayzenoss> leave it blank?
[25-Jan-2010 16:50:21] <forsberg> hmm cannot be blank, it puts the older number back in
[25-Jan-2010 16:52:07] <Muti> so I have an rrd of type GAUGE (for load average) whose values are decimals (like 0.03). When zenperfsnmp inserts the value into the RRD it's multiplied by 100, is there a reason for that?
[25-Jan-2010 16:59:19] <gwb2351> any DMD hackers out there that can tell me how to script dumping all events (history) for a device in the database?
[25-Jan-2010 17:02:58] <mrayzenoss> gwb2351: sounds more like a SQL query from the status and history tables
[25-Jan-2010 17:14:29] <mrayzenoss> rmatte: published http://community.zenoss.org/blogs/zenossblog/2010/01/25/tip-of-the-month-monitoring-the-status-of-nfs-shares-with-zenoss Thanks!
[25-Jan-2010 17:26:56] forsberg is now known as fOrsberg
[25-Jan-2010 17:46:33] <HCoyote> mrayzenoss: gwb2351 was asking for me ... I'd prefer to use the API where possible instead of grubbing through the backend DB to find the list of events that have occurred for a device.
[25-Jan-2010 17:51:18] <mrayzenoss> HCoyote: gotta run, ping me tomorrow and I'll see if I can followup
[25-Jan-2010 17:51:24] <HCoyote> sure thing.
[25-Jan-2010 17:51:26] <HCoyote> thanks.
[25-Jan-2010 19:58:46] <InitX> hi all i'am newbie to zenoss
[25-Jan-2010 19:59:52] <sergeymasushko> InitX: hi
[25-Jan-2010 20:18:49] <InitX> I want to translate zenoss but i don't know where to begin
[25-Jan-2010 20:26:19] <a21> InitX: what is the goal?
[25-Jan-2010 21:37:01] <InitX> a21: trasnlate the english to localization
[25-Jan-2010 21:38:56] <a21> InitX: a gyu who works with zenoss should know english, at least a bit, I think..
[25-Jan-2010 21:48:35] <InitX> a21 I want to translate Zenoss from english to localization like japanese, chinese etc. etc. but I don't konw where to begin this work
[25-Jan-2010 21:50:26] <a21> InitX: what about documentation?
[25-Jan-2010 21:56:18] <sergeymasushko> InitX_: as far as I know there is no possibility to localize the zenoss to another language yet. Matt Ray worked on it... You should look for 'mrayzenoss' here at EST daytime.
[26-Jan-2010 00:00:30] [disconnected at Tue Jan 26 00:00:30 2010]
[26-Jan-2010 00:00:30] [connected at Tue Jan 26 00:00:30 2010]
[26-Jan-2010 00:00:59] <Apachez> http://www.youtube.com/watch?v=mKtdTJP_GUI
[26-Jan-2010 01:02:34] <Egyptian[web]> morning all
[26-Jan-2010 03:54:26] <tehhobbit> Message:
[26-Jan-2010 03:54:26] <tehhobbit> threshold of zenperfsnmp cycle time exceeded: current value 300.00
[26-Jan-2010 03:54:39] <tehhobbit> what can I do about this ?
[26-Jan-2010 04:59:01] <abou_7anash_009> hi trying to change treshold for disk alerts as per my search I should do it under the OS tab but my devices dont have an OS tab
[26-Jan-2010 04:59:39] <abou_7anash_009> nm found it under templates
[26-Jan-2010 05:00:49] <abou_7anash_009> hmm..did change it from .9 to .8 but descriptoin says 90% for template
[26-Jan-2010 05:00:52] <abou_7anash_009> wher to change that ?
[26-Jan-2010 07:10:06] <ckrough> abou_7anash_009: its in the filesystem template
[26-Jan-2010 07:12:11] <ckrough> oh... that was hours ago :)
[26-Jan-2010 07:12:17] <ckrough> still waking up
[26-Jan-2010 08:29:41] <rocket> good morning ...
[26-Jan-2010 08:29:47] <mrayzenoss> yup
[26-Jan-2010 08:31:03] <ckrough> hello
[26-Jan-2010 08:59:42] <ckrough> hmm... in zendmd, this doesnt list anything: dmd.Devices.Network.Router.getSubDevices()
[26-Jan-2010 08:59:51] <ckrough> but it should list all the devices in that organizer, right?
[26-Jan-2010 09:00:29] <venturaville> I think there is a recursive one
[26-Jan-2010 09:00:32] <venturaville> try that one....
[26-Jan-2010 09:55:00] <rmatte> wow, bringing the zope threads down to 4 from 50 really did improve performance
[26-Jan-2010 09:55:19] <rmatte> my collection cycles for zenperfsnmp are like 6 seconds now instead of 20
[26-Jan-2010 09:55:21] <ckrough> really?
[26-Jan-2010 09:55:36] <rmatte> and the UI is overall more responsive
[26-Jan-2010 09:55:50] <rmatte> hasn't caused any issues with monitoring, quite the contrary
[26-Jan-2010 09:55:54] <ckrough> nice. This is the zserver-threads config right?
[26-Jan-2010 09:56:02] <rmatte> less heartbeat failures too
[26-Jan-2010 09:56:06] <rmatte> correct
[26-Jan-2010 10:03:04] <rmatte> the memory usage is decent so far, but I'll see where it's at in a week
[26-Jan-2010 10:03:11] <rmatte> it always has a tendency to just eat more and more
[26-Jan-2010 10:05:40] <rmatte> hopefully the tweak will fix that
[26-Jan-2010 10:19:38] <rmatte> man, sometimes the community site freezes up for me, like it's waiting to load some item but can't
[26-Jan-2010 10:19:47] <rmatte> anyone else notice that?
[26-Jan-2010 10:21:35] <rmatte> it's stuck at: Transferring data from www.google-analytics.com
[26-Jan-2010 10:21:43] * rmatte yawns and waits
[26-Jan-2010 10:21:53] * theacolyte comforts rmatte
[26-Jan-2010 10:22:08] <rmatte> it's annoying having to wait 20 minutes to reply to a forum post
[26-Jan-2010 10:22:20] <theacolyte> you should come unglued and smash a few things
[26-Jan-2010 10:22:23] <theacolyte> always works for me
[26-Jan-2010 10:22:38] <rmatte> nah, company is too cheap to replace the smashed thins
[26-Jan-2010 10:22:39] <theacolyte> at the very least it'll entertain your coworkers
[26-Jan-2010 10:22:41] <rmatte> things*
[26-Jan-2010 10:22:42] <rmatte> lol
[26-Jan-2010 10:22:42] <theacolyte> lol
[26-Jan-2010 10:23:02] <ckrough> if you answer any more forums posts Im going to have to make an rmatte folder in gmail
[26-Jan-2010 10:23:09] <rmatte> lol
[26-Jan-2010 10:23:26] <rmatte> I actually haven't answered that many lately
[26-Jan-2010 10:23:28] <theacolyte> I've lost a little zenoss steam
[26-Jan-2010 10:26:30] <theacolyte> trying to figure out how Ican get through today and manage to look like I'm doing something
[26-Jan-2010 10:30:25] <mrayzenoss> FYI, the WMI MSSQL ZenPack now supports 2000/2005/2008: http://community.zenoss.org/docs/DOC-4641
[26-Jan-2010 10:30:45] <theacolyte> Yes, but does it support >1 instances
[26-Jan-2010 10:30:52] <rmatte> mrayzenoss: why does Enterprise appear to have more collector perf graphs than core does?
[26-Jan-2010 10:31:09] <rmatte> there are some really useful looking ones, like memory usage per zenoss process
[26-Jan-2010 10:31:11] <mrayzenoss> zenwinperf & zenvmwarepef?
[26-Jan-2010 10:31:36] <rmatte> I'm talking about the graphs the were posted here: http://community.zenoss.org/message/44928#44928
[26-Jan-2010 10:31:49] <rmatte> they appear to be collector performance graphs that aren't included in core
[26-Jan-2010 10:32:05] <rmatte> well, not the system graphs
[26-Jan-2010 10:32:10] <rmatte> but thememory per process graph
[26-Jan-2010 10:32:22] <cgibbons> enterprise is kinda supposed to have more stuff, no?
[26-Jan-2010 10:32:46] <rmatte> cgibbons: well, obviously, but those are some pretty basic looking performance graphs, and they'd be extremely useful in Core
[26-Jan-2010 10:33:06] <cgibbons> it's just a template, it can be added
[26-Jan-2010 10:33:21] <rmatte> Enterprise already has a ton of extra features, we're just talking about a graph here
[26-Jan-2010 10:33:23] <rmatte> ;)
[26-Jan-2010 10:34:46] <ckrough> actually CORE is getting very close
[26-Jan-2010 10:35:03] <ckrough> if you consider all of the community zenpacks
[26-Jan-2010 10:35:12] <rmatte> ckrough: it doesn't have full WMI monitoring, it doesn't have easily configured remote collectors (though I hope to fix that)
[26-Jan-2010 10:35:21] <rmatte> plus there's all the enterprise ZenPacks
[26-Jan-2010 10:36:19] <rmatte> overall though, if I were to purchase enterprise it would be for the support (if I were monitoring really critical stuff)
[26-Jan-2010 10:37:31] <rmatte> Oh, it doesn't have easily installed LDAP either, which is another ZenPack that I hope to eventually work on
[26-Jan-2010 10:39:00] <cgibbons> i'm not sure enterprise even changes that template hrm
[26-Jan-2010 10:43:33] fOrsberg is now known as forsberg
[26-Jan-2010 10:44:11] <rmatte> really?
[26-Jan-2010 10:44:23] <rmatte> it could be some custom graph that he made, but it looks Zenoss collector style
[26-Jan-2010 10:46:12] <etank> anyone know if it is OK to add comments to the "Default RRD Create Command" area of a collectors settings?
[26-Jan-2010 10:46:31] <etank> i want to say "# this does x"
[26-Jan-2010 10:47:00] <ckrough> etank: not sure, but I suspect its not safe
[26-Jan-2010 10:47:43] <etank> ckrough: that was my fear as well.
[26-Jan-2010 10:47:45] <gwb235> in the graph custom definition, you can add stuff like
[26-Jan-2010 10:47:58] <gwb235> COMMENT:noResourceConservation(1)
[26-Jan-2010 10:47:58] <gwb235> COMMENT:memoryShortage(2)
[26-Jan-2010 10:48:07] <etank> ckrough: thank you for your help again with understanding the RRA stuff
[26-Jan-2010 10:48:31] <etank> gwb235: i am making the change in the default RRD create command though
[26-Jan-2010 10:48:33] <etank> not a custom
[26-Jan-2010 10:48:34] <ckrough> you're welcome
[26-Jan-2010 10:48:49] <etank> but i see what you mean gwb235
[26-Jan-2010 10:49:34] <gwb235> so i do it on some graphs, but not all, using the templates (not exactly what you were asking)
[26-Jan-2010 11:06:45] <rmatte> cgibbons: ah, apparently the guy actually created those graphs... I'm going to get him to make it in to a ZenPack, or send it to me so that I can
[26-Jan-2010 11:52:33] <forsberg> rmatte? any way to get a graph to auto max value?
[26-Jan-2010 11:52:38] <forsberg> or any other
[26-Jan-2010 11:55:44] <rmatte> what do you mean by auto max value?
[26-Jan-2010 11:56:10] <forsberg> -1 seemed to almost work
[26-Jan-2010 11:56:23] <rmatte> I don't really get what you're asking
[26-Jan-2010 11:56:42] <forsberg> so the Max Y is auto compared to the max reading in the given period of time
[26-Jan-2010 11:57:29] <rmatte> still don't quite understand... you want it to auto-zoom?
[26-Jan-2010 11:58:08] <forsberg> http://hesten.foppa.dk/files/Capture.JPG
[26-Jan-2010 11:59:43] <forsberg> hmm ok, dont know how to explain it then.. the theoretical max here is 4 x 8 but if i put 32 as Max Y then graph is just small bumps at the bottom
[26-Jan-2010 12:00:34] <rmatte> oh I see
[26-Jan-2010 12:00:38] <rmatte> just don't set a max at all
[26-Jan-2010 12:00:42] <rmatte> and it'll auto-scale
[26-Jan-2010 12:00:52] <rmatte> that's what I do
[26-Jan-2010 12:01:20] <forsberg> new value is invalid
[26-Jan-2010 12:01:30] <forsberg> it should be type INT
[26-Jan-2010 12:02:42] <rmatte> hmmm, let me check something
[26-Jan-2010 12:03:28] <rmatte> yeh, -1 is the default, leave max and min at -1
[26-Jan-2010 12:03:34] <rmatte> and it'll auto-scale
[26-Jan-2010 12:04:10] <forsberg> ok yes, just looked like it misses the top monday at 18
[26-Jan-2010 12:04:26] <rmatte> screenshot?
[26-Jan-2010 12:04:47] <forsberg> same as before
[26-Jan-2010 12:04:54] <forsberg> last ss was with -1
[26-Jan-2010 12:04:55] <rmatte> you sure it's missing anything?
[26-Jan-2010 12:04:59] <rmatte> it's showing as about 20
[26-Jan-2010 12:04:59] <forsberg> nopes
[26-Jan-2010 12:05:26] <rmatte> so you had 20 people at peak on Monday
[26-Jan-2010 12:05:30] <forsberg> well it must be all there :) im not sure that layout is better tho, but it was what i was asking for the other day
[26-Jan-2010 12:05:44] <forsberg> and 0 decimals certainly was important also
[26-Jan-2010 12:05:45] <rmatte> that's the best layout in my opinion
[26-Jan-2010 12:05:57] <rmatte> you could play with the colours to make it a bit prettier lol
[26-Jan-2010 12:06:16] <rmatte> for sure, you don't want decimals where they don't belong
[26-Jan-2010 12:06:28] <forsberg> yes to the colors hehe
[26-Jan-2010 12:06:46] <rmatte> sometimes the colors work out well with the defaults, other times it's just hideous
[26-Jan-2010 12:07:53] <rmatte> hmmm, Matt ran off
[26-Jan-2010 12:07:54] <rmatte> bleh
[26-Jan-2010 12:08:50] <forsberg> wich way is regarded the right way, also if i should make a zenpack out of it, using command or snmp ? the sh/command thing seems much better here, since then the script and qstat only needs to be present at the zenoss machine
[26-Jan-2010 12:09:45] <rmatte> do whatever works
[26-Jan-2010 12:10:05] <rmatte> and what do you mean by "right way"?
[26-Jan-2010 12:10:11] <rmatte> what are you referring to?
[26-Jan-2010 12:10:13] <forsberg> if there was some kind of best practice
[26-Jan-2010 12:10:19] <rmatte> for?
[26-Jan-2010 12:10:29] * rmatte is lost & confused
[26-Jan-2010 12:10:41] <forsberg> monitoring stuff
[26-Jan-2010 12:10:55] <rmatte> oh, no, that's really based on personal preference
[26-Jan-2010 12:11:13] <rmatte> each monitoring system is going to have different settings as they have different requirements
[26-Jan-2010 12:11:24] <rmatte> there really is not blanket best practice solution
[26-Jan-2010 12:11:38] <rmatte> the closest thing is to check best practices for each manufacturer and follow those if you need a guideline
[26-Jan-2010 12:14:00] <robo> hi: I am getting this error when I'm trying to add an OID to monitor: Error reading value for "ssCpuRawUser" on dsbigip (oid .1.3.6.1.4.1.2021.11.50 is bad). I believe that OID is indeed correct because using snmpwalk to query that OID on that system returns an expected value
[26-Jan-2010 12:14:27] <robo> what i'm doing is following http://www.sysadminwiki.net/wiki/index.php?title=How_to_Setup_Ironport_Monitoring_with_the_IronportMonitor_ZenPack.
[26-Jan-2010 12:14:56] <robo> Well, not following but using it as a guide.
[26-Jan-2010 12:15:42] <rmatte> robo: you're probably querrying an snmp tree rather than a single value
[26-Jan-2010 12:15:48] <rmatte> try adding a .0 to the end of that OID
[26-Jan-2010 12:16:07] <rmatte> (Zenoss knows the difference between a tree (a.k.a index) and a single value)
[26-Jan-2010 12:16:21] <rmatte> but an snmpwalk won't reveal that difference unless you look very closely
[26-Jan-2010 12:16:21] <robo> interesting, let me try that
[26-Jan-2010 12:16:35] <rmatte> k
[26-Jan-2010 12:18:08] <robo> added. I remodeled the device now I just need to wait for it to show
[26-Jan-2010 12:19:20] <rmatte> cool
[26-Jan-2010 12:19:35] <rmatte> test the datapoint to make sure it returns a value with the .0 added
[26-Jan-2010 12:19:38] <rmatte> if it does you should be good
[26-Jan-2010 12:19:58] <robo> It did return value with .0 on the end. Though, it also succeeded without the .0
[26-Jan-2010 12:20:10] <robo> not sure if it matters, but here is what snmpwalk returns: UCD-SNMP-MIB::ssCpuRawUser.0 = Counter32: 535044199
[26-Jan-2010 12:20:10] <rmatte> yeh, try to think of it this way...
[26-Jan-2010 12:20:23] <rmatte> without the .0 it's like listing the contents of a directory
[26-Jan-2010 12:20:24] <robo> I wonder if zenoss supports UCD-SNMP MIB's
[26-Jan-2010 12:20:31] <robo> ah, okay
[26-Jan-2010 12:20:34] <rmatte> and in this case only 1 file exists
[26-Jan-2010 12:20:42] <rmatte> by adding the .0 you're specifically targetting that file
[26-Jan-2010 12:20:56] <rmatte> if more values existed then you'd get multiple lines returned without the .0
[26-Jan-2010 12:20:57] <robo> still not showing the new datapoint in the gui
[26-Jan-2010 12:21:07] <rmatte> hunh?
[26-Jan-2010 12:21:11] <rmatte> what do you mean by showing it?
[26-Jan-2010 12:21:33] <rmatte> if it's an snmp type datasource the datapoint is automatically created and is invisible
[26-Jan-2010 12:21:41] <rmatte> (the documentation needs to be updated)
[26-Jan-2010 12:21:51] <robo> when i click on "status" then under "Component Type" is where i'm used to seeing the datapoints
[26-Jan-2010 12:21:58] <rmatte> the datapoint used to be visible, but now it's only visible for certain datasource types such as command based datasources
[26-Jan-2010 12:22:25] <rmatte> just go add it to a graph
[26-Jan-2010 12:23:06] <robo> ah, okay
[26-Jan-2010 12:24:11] <dwoodard> join #hexxeh
[26-Jan-2010 12:24:25] <theacolyte> errr, no?
[26-Jan-2010 12:24:33] <forsberg> .b dwoodard
[26-Jan-2010 12:24:34] <forsberg> ;>

[26-Jan-2010 12:24:44] <robo> added the graph but it's not collecting any values
[26-Jan-2010 12:24:55] <robo> i'm pretty sure i will get an error in the event tab soon
[26-Jan-2010 12:25:21] <forsberg> not collection values as in no graph showing or no rrd file in the perf/Device/blah dir
[26-Jan-2010 12:25:34] <rmatte> robo: it takes 3 polling cycles (15 minutes) before you'll see any values on the graph
[26-Jan-2010 12:25:50] patzer_ is now known as patzer
[26-Jan-2010 12:25:58] <rmatte> robo: does it currently show "nan" for all the values?
[26-Jan-2010 12:26:49] <robo> yeah, it shows nan for cur, avg, and max
[26-Jan-2010 12:26:53] <forsberg> \o/
[26-Jan-2010 12:28:22] <rmatte> robo: then you're good
[26-Jan-2010 12:28:24] <rmatte> just be patient
[26-Jan-2010 12:29:19] <robo> yaay, i just started to get numbers :-)
[26-Jan-2010 12:29:26] <robo> looks like the .0 is the trick! ty rmatte @
[26-Jan-2010 12:31:11] <rmatte> np
[26-Jan-2010 12:31:49] <rmatte> eugh, this is driving me nuts
[26-Jan-2010 12:31:50] <rmatte> port 31-Excessive CRC/alignment errors. See help.
[26-Jan-2010 12:32:01] <rmatte> trying to write a regex to extract 31 from that
[26-Jan-2010 12:32:41] <patzer> /^\w+ (\d+)/
[26-Jan-2010 12:32:57] <rmatte> aha
[26-Jan-2010 12:32:59] <patzer> or you could say /port (\d+)/
[26-Jan-2010 12:32:59] <robo> /^port\s(\d{2})-.*?$/
[26-Jan-2010 12:33:21] <rmatte> actually I used \s((.*))-
[26-Jan-2010 12:33:23] <rmatte> and it worked too
[26-Jan-2010 12:33:39] <robo> that can break if you get another "-" in there
[26-Jan-2010 12:33:52] <patzer> you usually want .*? for non-greedy matching
[26-Jan-2010 12:34:19] <patzer> sir robo is correct ;)
[26-Jan-2010 12:34:59] <rmatte> robo: there never is another - in there
[26-Jan-2010 12:41:28] <Muti> it will also break if you have a port between 1 and 9, or 100+
[26-Jan-2010 12:41:49] <rmatte> why's that?
[26-Jan-2010 12:41:55] <Muti> \d{2}
[26-Jan-2010 12:42:10] <rmatte> I'm using \s((.*?))-
[26-Jan-2010 12:42:14] <Muti> ah
[26-Jan-2010 12:42:15] <rmatte> which shouldn't break
[26-Jan-2010 12:42:23] <Muti> I thought you were using robo's
[26-Jan-2010 12:42:26] <rmatte> nah
[26-Jan-2010 12:42:50] <Muti> why not just use (\d+)
[26-Jan-2010 12:43:04] <Muti> that will find the first number in the string
[26-Jan-2010 12:43:24] <rmatte> I'd rather be more specific than that
[26-Jan-2010 12:48:47] <robo> yeah, if you're matching 1, 2 or 3 numbers I'd use \d+ or \d{1,3}
[26-Jan-2010 12:49:54] <robo> i however would never use (.*?)- , because i wouldn't feel confident i'm always getting what i want
[26-Jan-2010 12:50:11] <robo> i like to have much more control over what i'm grabbing than that provides
[26-Jan-2010 12:50:19] <rmatte> the message is always formatted in a very specific way
[26-Jan-2010 12:50:27] <rmatte> so doing it my way isn't going to cause any issues
[26-Jan-2010 12:50:30] <robo> cool
[26-Jan-2010 12:50:45] <rmatte> but if it weren't then yes I'd be looking for a more robust solution
[26-Jan-2010 12:51:20] LarsN_Away is now known as LarsN
[26-Jan-2010 12:59:43] * eidolon1 gets off the phone with rusty.
[26-Jan-2010 12:59:46] eidolon1 is now known as eidolon
[26-Jan-2010 12:59:58] <eidolon> rmatte: that 'no graphs' problem?
[26-Jan-2010 13:00:08] <eidolon> it was a typo in the COMMAND entry. one $, not $$
[26-Jan-2010 13:00:30] <rmatte> ah
[26-Jan-2010 13:00:46] <eidolon> we had to do a 'walk the tree' command in zendmd, and that pointed it out. it wasn't really logging anywhere.
[26-Jan-2010 13:00:47] <rmatte> you mean for zenhome
[26-Jan-2010 13:00:53] <rmatte> $ZENHOME instead of $$ZENHOME
[26-Jan-2010 13:00:53] <rmatte> ?
[26-Jan-2010 13:01:03] <eidolon> no, it was in the command:
[26-Jan-2010 13:01:09] <rmatte> ahh
[26-Jan-2010 13:01:23] <rmatte> fun fun
[26-Jan-2010 13:01:33] <eidolon> /usr/bin/curl -s http://status.REDACTED.net | grep `date +%x` | awk -F' ' '{print "|Ingestion="$$5}'
[26-Jan-2010 13:01:37] <eidolon> there was only one $ on the $5
[26-Jan-2010 13:01:46] <rmatte> ah, I see
[26-Jan-2010 13:20:02] LarsN is now known as LarsN_Away
[26-Jan-2010 13:27:19] <robo> hi: so I created a new F5 organizer, and in there I created a bunch of templates for datapoints and graphs. Instead of creating my own CPU graph is there a way I can import the /Server/Linux ones and I'll just change the OIDs?
[26-Jan-2010 13:28:49] <mrayzenoss> yeah, you can copy them
[26-Jan-2010 13:29:13] <robo> hmm
[26-Jan-2010 13:30:26] <robo> ah, ty. Found it -- had to adjust my google search
[26-Jan-2010 13:52:49] * rmatte is wondering when Nate is going to fix zapplet so that it doesn't show an error icon when a Zenoss server has 0 events in the event console
[26-Jan-2010 13:53:05] <rmatte> I've gone through the code but I can't figure out how to go about fixing it yet
[26-Jan-2010 13:54:34] <mrayzenoss> npmccallum: you get that?
[26-Jan-2010 13:54:46] <rmatte> I mentioned it to him ages ago
[26-Jan-2010 13:54:56] <rmatte> He either forgot about it or has been too busy to fix it
[26-Jan-2010 13:55:01] <rmatte> it's probably a relatively simple fix
[26-Jan-2010 13:55:35] <npmccallum> I just haven't gotten to it
[26-Jan-2010 13:55:55] <npmccallum> its my hope to do some zapplet hacking @ SCaLE in a few weeks
[26-Jan-2010 13:56:01] <rmatte> cool
[26-Jan-2010 13:56:04] <npmccallum> I'll be on the road mostly until then
[26-Jan-2010 13:56:40] <rmatte> the other thing which would be a nice addition is for event views to show the name of the zenoss server at the top and have an x to close the view
[26-Jan-2010 13:56:50] <rmatte> the only way to close the view right now is to click the icon for that server again
[26-Jan-2010 13:57:01] <rmatte> and sometimes I get a bunch of event views open and can't remember which they are
[26-Jan-2010 13:57:07] <rmatte> so it's trial and error to get them to close
[26-Jan-2010 13:57:40] <rmatte> or maybe have the last view auto-close when a new view is clicked
[26-Jan-2010 13:58:06] <npmccallum> rmatte: I suspect that this will all be changing with appindicators anyway
[26-Jan-2010 13:58:06] <rmatte> if the bug were fixed and those were implemented it would be good to go
[26-Jan-2010 13:58:15] <rmatte> appindicators?
[26-Jan-2010 13:58:23] <npmccallum> http://people.canonical.com/~ted/libappindicator/current/AppIndicator.html
[26-Jan-2010 13:58:28] <npmccallum> http://www.notmart.org/misc/statusnotifieritem/
[26-Jan-2010 13:58:43] <npmccallum> its the newish API for applets
[26-Jan-2010 13:58:54] <rmatte> ah cool
[26-Jan-2010 14:00:26] <npmccallum> there are huge benefits and some downsides
[26-Jan-2010 14:00:37] <npmccallum> the current method of popping up the window will no longer work
[26-Jan-2010 14:00:46] <npmccallum> and we'll be limited to a menu
[26-Jan-2010 14:01:40] <rmatte> I see
[26-Jan-2010 14:01:59] <rmatte> sounds kind of limited
[26-Jan-2010 14:02:05] <npmccallum> yes, it is
[26-Jan-2010 14:02:13] <npmccallum> but I think the UI of zapplet could be better anyway
[26-Jan-2010 14:03:23] <rmatte> I shall wait and see
[26-Jan-2010 14:04:01] <rmatte> man, knocking the amount of zope threads down to 4 from 50 massively improved my poll cycles
[26-Jan-2010 14:04:41] <rmatte> memory usage seems better, but I need to give it a week to see if it creeps in to swap like it always does
[26-Jan-2010 14:07:59] <rmatte> like, check this out: http://tinypic.com/view.php?pic=2njbvar&s=6
[26-Jan-2010 14:08:09] <rmatte> I changed the setting on the 25th
[26-Jan-2010 14:08:36] <rocket> grrr ... why .. why .. cant I get my head around ruby programming?
[26-Jan-2010 14:09:29] <gwb235> here's a question to drive you crazy... "You must implement WMI monitoring via a secure, encrypted connection". Anybody do that in the past?
[26-Jan-2010 14:10:08] <gwb235> related info: http://msdn.microsoft.com/en-us/library/aa393266(VS.85).aspx
[26-Jan-2010 14:13:33] <venturaville> rocket: ...try erlang sometime ... ruby will look easy in comparison :-)
[26-Jan-2010 14:15:07] <rocket> venturaville: heh .. I guess I have gotten to used to python .. :p
[26-Jan-2010 14:15:16] <jmp242> Hey, has anyone had any better luck with Process monitoring yet?
[26-Jan-2010 14:15:50] <rocket> venturaville: ruby and metaprogramming just seems weird to me at the moment .. but I have to figure it out to understand how a certain program is written
[26-Jan-2010 14:16:01] <rmatte> jmp242: in what sense?
[26-Jan-2010 14:19:57] <rmatte> 16894 OID requests
[26-Jan-2010 14:20:07] <rmatte> Successes: 338 Failures: 0 Not reporting: 0
[26-Jan-2010 14:20:12] <rmatte> lasted 31.20 seconds
[26-Jan-2010 14:20:16] <rmatte> not bad
[26-Jan-2010 14:21:11] forsberg is now known as fOrsberg
[26-Jan-2010 14:21:21] rmatte is now known as rm4tte
[26-Jan-2010 14:21:50] rocket is now known as r0ck3t
[26-Jan-2010 14:21:54] <r0ck3t> ;)
[26-Jan-2010 14:21:55] <rm4tte> hehe
[26-Jan-2010 14:22:04] <rm4tte> I was waiting for you to chime in too
[26-Jan-2010 14:22:15] <r0ck3t> sure you were ...
[26-Jan-2010 14:22:20] r0ck3t is now known as rocket
[26-Jan-2010 14:22:28] <rm4tte> we're so 1337 it's cr4zy
[26-Jan-2010 14:22:50] <rocket> lolz ...
[26-Jan-2010 14:22:55] * rm4tte prances around, completely content with his 1337n3$$
[26-Jan-2010 14:23:01] rm4tte is now known as rmatte
[26-Jan-2010 14:23:09] * rocket stabs ruby with a fork ...
[26-Jan-2010 14:23:13] <rmatte> lol
[26-Jan-2010 14:23:19] <rmatte> poor ruby
[26-Jan-2010 14:23:34] <rocket> why couldnt puppet have been written in python .. it might have made sense then .. :p
[26-Jan-2010 14:23:46] <rmatte> rocket: there's always bcfg2 :P
[26-Jan-2010 14:24:15] <rocket> I am just trying to write a new providor ...
[26-Jan-2010 14:24:39] <jmp242> rmatte
[26-Jan-2010 14:24:47] <jmp242> generally in the sense of getting it to notice a process
[26-Jan-2010 14:24:48] <jmp242> going down
[26-Jan-2010 14:25:46] <jmp242> or actually for it to properly identify a process
[26-Jan-2010 14:25:51] <jmp242> I've got a process called taskDaemon.py
[26-Jan-2010 14:25:57] <jmp242> that's a python process on a remote server
[26-Jan-2010 14:26:00] <rmatte> jmp242: works fine for me, but I monitor more windows processes than anything
[26-Jan-2010 14:26:15] <rmatte> I am monitoring my zenoss processes on all of my zenoss servers, and the instance doing the monitoring works fine
[26-Jan-2010 14:26:19] <rmatte> it sees when processes go down
[26-Jan-2010 14:26:22] <jmp242> mmm
[26-Jan-2010 14:26:32] <jmp242> so I'm guessing you don't have processes like
[26-Jan-2010 14:26:39] <venturaville> rocket: you can work off the existing puppet example providers out there
[26-Jan-2010 14:26:46] <venturaville> just look through some of the recipes
[26-Jan-2010 14:26:48] <jmp242> this /usr/bin/python /opt/cds-invenio/bin/bibsched -q s
[26-Jan-2010 14:27:00] <rmatte> jmp242: nope
[26-Jan-2010 14:27:02] <jmp242> but more like myprocess.exe
[26-Jan-2010 14:27:14] <rmatte> jmp242: that's the sort of thing which zenprocess seems to not do very well right now
[26-Jan-2010 14:27:22] <rmatte> jmp242: well...
[26-Jan-2010 14:27:26] <rocket> venturaville: where? .. I have been looking at the source code .. but thats quite the mess .. :p
[26-Jan-2010 14:27:28] <jmp242> We've been having the devil's own time getting these processes that run under vms or whatever, java
[26-Jan-2010 14:27:30] <jmp242> python etc
[26-Jan-2010 14:27:37] <jmp242> to monitor right
[26-Jan-2010 14:27:42] <rmatte> for instance, in the processes section I have...
[26-Jan-2010 14:28:01] <rmatte> mysqld.bin, zenactions, zencommand, zeneventlog, etc...
[26-Jan-2010 14:28:01] <jmp242> hmm, maybe a command based datasource
[26-Jan-2010 14:28:04] <rmatte> just defined like that
[26-Jan-2010 14:28:11] <rocket> venturaville: I am just trying to make a zypper package manager providor for suse .. :p
[26-Jan-2010 14:28:14] <rmatte> then on a device when they are discovered they look like...
[26-Jan-2010 14:28:35] * rmatte waits for the page to load
[26-Jan-2010 14:28:40] <jmp242> lol
[26-Jan-2010 14:29:30] <rmatte> /usr/local/zenoss/mysql/bin/mysqld.bin --defaults-file=/usr/local/zenoss/mysql/my.cnf --basedir=/usr/local/zenoss/mysql --datadir=/usr/local/zenoss/mysql/data --user=m
[26-Jan-2010 14:29:41] <rmatte> /usr/local/zenoss/python/bin/.python.bin /usr/local/zenoss/zenoss/Products/Jobber/zenjobs.py --configfile /usr/local/zenoss/zenoss/etc/zenjobs.conf --cycle --daemon
[26-Jan-2010 14:29:44] <rmatte> stuff like that
[26-Jan-2010 14:29:47] <rmatte> and it works fine
[26-Jan-2010 14:30:05] <rmatte> if you just specified
[26-Jan-2010 14:30:07] <rmatte> erm
[26-Jan-2010 14:30:21] <rmatte> if you just specified bibsched as the process name/regex it would work for you
[26-Jan-2010 14:30:38] <rmatte> I suspect you're overcomplicating things
[26-Jan-2010 14:30:50] <venturaville> rocket: lots of examples: http://github.com/reductivelabs/puppet/tree/master/lib/puppet/provider/package/
[26-Jan-2010 14:31:02] <rmatte> also, after you make any config change to zenprocess (discover processes on a device for instance) restart zenprocess
[26-Jan-2010 14:31:09] <rmatte> otherwise it'll take forever to pick up the new config
[26-Jan-2010 14:31:13] <rmatte> and it'll keep using the old one
[26-Jan-2010 14:31:20] <jmp242> I forgot about that
[26-Jan-2010 14:31:44] <rmatte> jmp242: I just do it as a second nature thing these days, and it does work, it's just annoying to have to do it
[26-Jan-2010 14:32:01] <rmatte> jmp242: luckily it can be done from the GUI so it's not that painful
[26-Jan-2010 14:33:04] <jmp242> yea, trying now
[26-Jan-2010 14:33:09] <jmp242> I can't believe I forgot about that
[26-Jan-2010 14:33:24] <rocket> venturaville: i have been looking there .. :p
[26-Jan-2010 14:33:43] <rmatte> hehe
[26-Jan-2010 14:38:54] <rmatte> jmp242: another trick is if you're monitoring windows processes (whatever.exe) it's a good idea to set "Ignore Parameters" to True for the process
[26-Jan-2010 14:40:42] <rmatte> but not for linux processes
[26-Jan-2010 14:45:13] <jmp242> indeed rmatte
[26-Jan-2010 14:45:19] <jmp242> The Zenprocess restart did it
[26-Jan-2010 14:45:29] <jmp242> I'd add that to the FAQ if I could still get the edits to work right
[26-Jan-2010 14:45:46] <rmatte> lol
[26-Jan-2010 14:46:10] <rmatte> I wonder if there's a trac ticket open for that
[26-Jan-2010 14:46:23] <rmatte> (obviously shouldn't have to manually restart zenprocess)
[26-Jan-2010 14:47:14] <jmp242> It seems like it's related to picking up config changes mostly
[26-Jan-2010 14:47:39] <jmp242> but they really should also add some GUI and/or parsing support for java and python etc apps
[26-Jan-2010 14:47:55] <jmp242> which generally have a second argument that points to the actual app
[26-Jan-2010 14:47:56] <rmatte> yeh
[26-Jan-2010 14:48:16] <jmp242> you want to search trac/open the ticket or should I?
[26-Jan-2010 14:49:47] ideopathic_ is now known as ideopathic
[26-Jan-2010 14:54:02] <rmatte> go for it :)
[26-Jan-2010 15:03:21] <Muti> heh: threshold of high load exceeded: current value 294.51
[26-Jan-2010 15:05:41] <rmatte> high load?
[26-Jan-2010 15:05:46] <rmatte> like on a linux server?
[26-Jan-2010 15:05:50] <Muti> yep
[26-Jan-2010 15:05:54] <rmatte> brutal
[26-Jan-2010 15:05:59] <rmatte> what the heck are you running on there?
[26-Jan-2010 15:06:04] <rmatte> or is it running on a 386 or something?
[26-Jan-2010 15:06:31] <Muti> heh, no.. it's a dual cpu xeon
[26-Jan-2010 15:06:39] <Muti> running apache
[26-Jan-2010 15:06:56] <rmatte> just apache?
[26-Jan-2010 15:07:10] <Muti> that's the only thing that uses resources
[26-Jan-2010 15:07:17] <rmatte> crazy
[26-Jan-2010 15:07:27] <rmatte> is it getting a billion hits a second or something?
[26-Jan-2010 15:08:21] <Muti> heh, it gets quite a bit and the data it's serving is stored in a storage cluster
[26-Jan-2010 15:08:39] <patzer> I used to have an old solaris7 pizza box a decade ago that routinely had loads that high when we got high traffic loads
[26-Jan-2010 15:08:50] <patzer> but they were very high loads for the late 90s I think
[26-Jan-2010 15:12:33] <Muti> looks like it's serving about 22 MB/s
[26-Jan-2010 15:13:06] <cgibbons> hurm
[26-Jan-2010 15:21:43] <Muti> so anyone have any experience w/ graphing CPU usage of FreeBSD servers?
[26-Jan-2010 15:34:23] <rmatte> Muti: as long as you have net-snmp installed I would think it would be the same OIDs as used in the Linux template
[26-Jan-2010 15:34:26] <rmatte> I could be wrong though
[26-Jan-2010 15:45:13] <rmatte> Muti: I don't have access to an snmp enabled FreeBSD box or I'd take a closer look
[26-Jan-2010 15:45:51] <Muti> yeah, it's the same OIDs
[26-Jan-2010 15:46:00] <rmatte> k cool
[26-Jan-2010 15:46:07] <Muti> but I was having issues w/ the ssCpuRaw* values being weird
[26-Jan-2010 15:46:16] <rmatte> define weird
[26-Jan-2010 15:46:25] <Muti> well, they're measured in total ticks
[26-Jan-2010 15:47:21] <Muti> so when I'd graph say, ssCpuRawIdle
[26-Jan-2010 15:47:38] <Muti> even though idle was at like 99%, it came out to about 51%
[26-Jan-2010 15:49:50] <Muti> but apparently this device (an Isilon storage node) also exposes CPU usage in percentage in the enterprise MIB
[26-Jan-2010 15:49:55] <Muti> so I'm just using that now
[26-Jan-2010 15:52:33] <rmatte> ah
[26-Jan-2010 15:53:54] <Muti> I think it may have been because the CPU ticks/s on FreeBSD is different
[26-Jan-2010 15:55:01] <rmatte> probably
[26-Jan-2010 16:21:39] <cgibbons> jane is my hero
[26-Jan-2010 16:26:34] <Muti> is it not possible to add a device to a device class that doesn't have any collector plugins?
[26-Jan-2010 16:26:51] <rmatte> cgibbons: why?
[26-Jan-2010 16:27:06] <rmatte> Muti: you should be able to, it just won't really do anything
[26-Jan-2010 16:27:18] <cgibbons> she does a good job explaining things that i don't know how they work
[26-Jan-2010 16:27:26] <cgibbons> in this context, the event mapping magic
[26-Jan-2010 16:27:36] <rmatte> yeh, she has definitely dug in real deep to all that stuff
[26-Jan-2010 16:27:48] <rmatte> oh my freaking god I hate php sometimes
[26-Jan-2010 16:28:04] <rmatte> I'm digging through this "cake php" that some of the devs here are using trying to fix one little thing
[26-Jan-2010 16:28:15] <rmatte> there are buttons which post to another file
[26-Jan-2010 16:28:33] <rmatte> and they basically go out and instruct zenoss to perform a command (ping or snmp) then return the output
[26-Jan-2010 16:28:45] <rmatte> but when I check the file that it's calling all that's in that file is...
[26-Jan-2010 16:28:57] <rmatte> <?php echo $commandOutput; ?>

[26-Jan-2010 16:29:12] <rmatte> so I have absolutely no idea where it's getting the rest of it's instructions from when the button is clicked
[26-Jan-2010 16:29:20] <rmatte> there's no reference to anything else in the code that I can see
[26-Jan-2010 16:29:30] * rmatte scratches his head
[26-Jan-2010 16:51:33] <rmatte> aha, finally figured it out
[26-Jan-2010 17:01:11] <kisielk> how do I fix missing RRD files?
[26-Jan-2010 17:01:18] <kisielk> is there a way to just get them to be recreated?
[26-Jan-2010 17:01:36] <rmatte> kisielk: if you have missing RRD files it's because your datasources aren't correct or Zenoss isn't able to gather data for the device
[26-Jan-2010 17:01:55] <rmatte> you can't manually replace them, Zenoss creates them provided that the data is correctly being gathered
[26-Jan-2010 17:02:27] <kisielk> well, I have 100 identical machines
[26-Jan-2010 17:02:27] <kisielk> s
[26-Jan-2010 17:02:30] <rmatte> well, you could technically create them manually but it would be tedious and pointless
[26-Jan-2010 17:02:34] <kisielk> ome of them have the RRD files for the perf graphs
[26-Jan-2010 17:02:38] <kisielk> some of them do not
[26-Jan-2010 17:02:47] <kisielk> SNMP collections works just the same for all of them
[26-Jan-2010 17:02:57] <rmatte> perhaps you should pick an OID that Zenoss is having problems with on some and snmpwalk them
[26-Jan-2010 17:03:04] <rmatte> see if it's returning correct values
[26-Jan-2010 17:03:05] <kisielk> I did that, they all work
[26-Jan-2010 17:03:13] <kisielk> this is just basic stuff like load average etc
[26-Jan-2010 17:03:19] <rmatte> do zenperfsnmp run -v10 --cycle
[26-Jan-2010 17:03:33] <rmatte> and wait for a summary of how long it took, and how many failures there were
[26-Jan-2010 17:04:26] <kisielk> seems to work
[26-Jan-2010 17:04:47] <rmatte> what was the summary?
[26-Jan-2010 17:05:09] <kisielk> hm actually it's weird
[26-Jan-2010 17:05:10] <kisielk> INFO zen.zenperfsnmp: Configured 97 of 0 devices
[26-Jan-2010 17:05:12] <rmatte> you should see a line like: Successes: 338 Failures: 0 Not reporting: 0
[26-Jan-2010 17:05:49] <kisielk> hm okay
[26-Jan-2010 17:05:50] <kisielk> s
[26-Jan-2010 17:05:52] <kisielk> omething is fucked with it
[26-Jan-2010 17:05:54] <kisielk> not surprised
[26-Jan-2010 17:06:01] <rmatte> have you tried restarting Zenoss?
[26-Jan-2010 17:07:18] <kisielk> well, I'm sure it's been restarted a few times since this has been running
[26-Jan-2010 17:07:26] <rmatte> I see
[26-Jan-2010 17:07:36] <kisielk> but it's been monitoring these hosts for a few months now, and there's no RRD files for the load average for some of them
[26-Jan-2010 17:07:41] <kisielk> which doesn't make any sense to me
[26-Jan-2010 17:08:06] <rmatte> but they do all return results for that particular oid?
[26-Jan-2010 17:08:10] <kisielk> yes
[26-Jan-2010 17:08:16] <theacolyte> wow, that sounds familiar.
[26-Jan-2010 17:08:44] <rmatte> do you see any errors when you do zenperfsnmp run -v10 --cycle?
[26-Jan-2010 17:09:08] <kisielk> well, like I said, it seems kind of screwed
[26-Jan-2010 17:09:16] <rmatte> why do you say that?
[26-Jan-2010 17:09:35] <kisielk> because it has a bunch of lines that say "Error loading config for devices"
[26-Jan-2010 17:09:37] <kisielk> with a list of devices
[26-Jan-2010 17:09:46] <kisielk> however it does that for every device in Zenoss
[26-Jan-2010 17:10:56] <theacolyte> rmatte: remind me your thold for errors on T1's
[26-Jan-2010 17:11:16] <rmatte> I just do 300 * 0.07 on all my stuff (which is 21 errors per second)
[26-Jan-2010 17:11:19] <rmatte> but that's pretty high
[26-Jan-2010 17:11:39] <rmatte> it's just because carriers generally consider anything below that percentage to be acceptable
[26-Jan-2010 17:11:47] <rmatte> so we really only want to be alerted on the really critical stuff
[26-Jan-2010 17:12:18] <rmatte> kisielk: sounds like something is messed in your database
[26-Jan-2010 17:12:20] <rmatte> INFO:zen.zenperfsnmp:Configured 338 of 338 devices
[26-Jan-2010 17:12:27] <rmatte> yours says "of 0 devices"
[26-Jan-2010 17:12:40] <kisielk> yes
[26-Jan-2010 17:12:52] <kisielk> but somehow most of the perf stuff gets collected
[26-Jan-2010 17:13:06] <kisielk> I guess I just have to delete everything
[26-Jan-2010 17:13:09] <rmatte> have you tried a reindex() and commit() in zendmd?
[26-Jan-2010 17:13:12] <kisielk> yes
[26-Jan-2010 17:13:13] <kisielk> I do that like
[26-Jan-2010 17:13:14] <kisielk> 3 times a day
[26-Jan-2010 17:13:15] <kisielk> sin
[26-Jan-2010 17:13:20] <kisielk> ce that's the only suggestion for every problem
[26-Jan-2010 17:13:23] <rmatte> well, you shouldn't be doing it nearly that often
[26-Jan-2010 17:13:32] <theacolyte> Hmmm I don't even see how to enter that 300 * 0.07
[26-Jan-2010 17:13:36] <rmatte> it's ok once in a while, but doing it really often like that can cause more harm than good
[26-Jan-2010 17:13:37] <theacolyte> there's just a min/max
[26-Jan-2010 17:13:38] <kisielk> heh, well, that's the first thing people always suggest
[26-Jan-2010 17:13:46] <rmatte> theacolyte: you enter that right in to max
[26-Jan-2010 17:13:55] <rmatte> 300 * 0.07
[26-Jan-2010 17:13:55] <theacolyte> ahhh
[26-Jan-2010 17:13:59] <theacolyte> and set the min to 0
[26-Jan-2010 17:14:02] <rmatte> no
[26-Jan-2010 17:14:06] <rmatte> leave the min unset
[26-Jan-2010 17:14:09] <theacolyte> :P
[26-Jan-2010 17:14:15] <rmatte> seriously
[26-Jan-2010 17:14:19] <theacolyte> I learned the min to 0 one the hard way
[26-Jan-2010 17:14:28] <aclark> hah
[26-Jan-2010 17:14:32] <rmatte> otherwise you're going to get threshold alert for every interface with 0 errors
[26-Jan-2010 17:14:32] * aclark files that one for later
[26-Jan-2010 17:14:32] <rmatte> :P
[26-Jan-2010 17:14:40] <theacolyte> yeah, I noticed
[26-Jan-2010 17:14:43] <rmatte> hehe
[26-Jan-2010 17:14:50] <theacolyte> got that one about 2 weeks ago wheN Iwent on vacation
[26-Jan-2010 17:14:55] <theacolyte> I'm like hmm... this is weird
[26-Jan-2010 17:14:59] <rmatte> lol
[26-Jan-2010 17:15:22] <rmatte> "It's complaining that there aren't enough errors!? Quick, guys, switch some duplex settings!"
[26-Jan-2010 17:15:32] <theacolyte> is there a way to create a local copy of a template for multiple interfaces for a device?
[26-Jan-2010 17:15:36] <theacolyte> lol, no joke
[26-Jan-2010 17:15:48] <rmatte> theacolyte: nope, gotta do it one by one by hand
[26-Jan-2010 17:15:53] <theacolyte> loverly
[26-Jan-2010 17:16:17] <rmatte> theacolyte: well, actually...
[26-Jan-2010 17:16:33] <rmatte> you want it applied to every interface on said device?
[26-Jan-2010 17:16:35] <theacolyte> could just do it for all my routers
[26-Jan-2010 17:16:37] <theacolyte> no
[26-Jan-2010 17:16:37] <theacolyte> hehe
[26-Jan-2010 17:16:50] <rmatte> well, like every interface that matches I mean
[26-Jan-2010 17:17:04] <theacolyte> If it contains Serial in it, yes
[26-Jan-2010 17:17:18] <rmatte> well, physical contain serial
[26-Jan-2010 17:17:22] <rmatte> just as sub do
[26-Jan-2010 17:17:27] <rmatte> so that wouldn't work all that well
[26-Jan-2010 17:17:32] <theacolyte> right
[26-Jan-2010 17:17:35] <rmatte> lol
[26-Jan-2010 17:17:36] <theacolyte> is fine
[26-Jan-2010 17:17:43] <theacolyte> I'm just going to do it for this one site
[26-Jan-2010 17:17:47] <theacolyte> only 4 there
[26-Jan-2010 17:17:48] <rmatte> I know what you're doing... you'll have to do it one by one by hand
[26-Jan-2010 17:17:57] <rmatte> that's how I did it
[26-Jan-2010 17:18:03] <rmatte> the other option...
[26-Jan-2010 17:18:15] <rmatte> is to copy the frame-relay template to the device itself
[26-Jan-2010 17:18:19] <rmatte> and then modify the template
[26-Jan-2010 17:18:23] <theacolyte> ahhhhhhhhhhhhhhhhhhhhh
[26-Jan-2010 17:18:25] <theacolyte> right
[26-Jan-2010 17:18:44] <rmatte> As long as you're not being super picky about which interfaces have what then it'll work
[26-Jan-2010 17:18:58] * theacolyte sighs, I should just fix my equallogic zenpack
[26-Jan-2010 17:19:02] <theacolyte> mrayzenoss ignore me now
[26-Jan-2010 17:19:04] <rmatte> lol
[26-Jan-2010 17:19:04] <theacolyte> hehe
[26-Jan-2010 17:19:11] <theacolyte> he's like that damn acolyte guy
[26-Jan-2010 17:19:48] <rmatte> :)
[26-Jan-2010 17:19:59] <theacolyte> It's a pretty damn good template if you ask me sans that stupid OID stuff
[26-Jan-2010 17:20:08] <rmatte> well, decreasing the zope threads is making performance a million times better on this Zenoss box
[26-Jan-2010 17:20:19] <rmatte> normally it would be sitting at 3GB utilization after 8 hours
[26-Jan-2010 17:20:25] <theacolyte> wow
[26-Jan-2010 17:20:27] <rmatte> it's been stable at 2040MB
[26-Jan-2010 17:20:30] <theacolyte> how many devices?
[26-Jan-2010 17:20:42] <rmatte> 352
[26-Jan-2010 17:20:52] <rmatte> though about 25 of those are ping only
[26-Jan-2010 17:21:11] <rmatte> it's about 17500 datapoints overall
[26-Jan-2010 17:21:18] <theacolyte> I think I'm at about 1200 now
[26-Jan-2010 17:21:23] <theacolyte> Using 1250mb
[26-Jan-2010 17:21:32] <rmatte> yeh, but your datapoints are probably quite low
[26-Jan-2010 17:21:41] <theacolyte> they aren't all that heavy duty
[26-Jan-2010 17:21:43] <theacolyte> yeah
[26-Jan-2010 17:21:57] <rmatte> I mean, if you're only monitoring 3 datapoints on 1200 devices then you're only monitoring 3600 datapoints
[26-Jan-2010 17:22:01] <theacolyte> I need to figure out some colors that don't suck
[26-Jan-2010 17:22:06] <rmatte> lol
[26-Jan-2010 17:22:35] <theacolyte> I may just switch back to line graphs for the graphs that have >2 datapoints
[26-Jan-2010 17:23:07] <rmatte> that's an option
[26-Jan-2010 17:23:56] <theacolyte> Is there a way to align all the txt up in the legend?
[26-Jan-2010 17:24:10] <theacolyte> if the datapoint is different lengths it borks it a bit
[26-Jan-2010 17:24:11] <rmatte> I'm out
[26-Jan-2010 17:24:11] <rmatte> later
[26-Jan-2010 17:24:14] <theacolyte> later man
[26-Jan-2010 17:26:46] <mrayzenoss> theacolyte: http://dev.zenoss.org/trac/ticket/4541
[26-Jan-2010 17:27:19] <theacolyte> ah
[26-Jan-2010 17:27:21] <theacolyte> still open
[26-Jan-2010 17:27:25] <mrayzenoss> sadly yes
[26-Jan-2010 17:27:31] <theacolyte> man, come on stonecrab!
[26-Jan-2010 17:27:33] <theacolyte> hehe
[26-Jan-2010 17:28:03] <theacolyte> how can you tell if something will be fixed in the next release or not?
[26-Jan-2010 17:28:43] <mrayzenoss> wellâ€¦ there are reports available in Trac: http://dev.zenoss.org/trac/report
[26-Jan-2010 17:29:12] <mrayzenoss> specifically the Current Release ones
[26-Jan-2010 17:29:22] <theacolyte> ahhhhhh
[26-Jan-2010 17:30:45] <mrayzenoss> Closed and Awaiting Verification are pretty much in
[26-Jan-2010 17:31:00] <mrayzenoss> Developer Queue is much more iffy
[26-Jan-2010 17:36:10] <cgibbons> it's usually a lot easier to tell if something is going into a maint release versus a major release, thosse are somewhat more random
[26-Jan-2010 17:40:02] <mrayzenoss> later
[26-Jan-2010 19:42:14] tuvyz is now known as tuv
[26-Jan-2010 19:56:25] nonsenso_ is now known as nonsenos
[26-Jan-2010 19:56:27] nonsenos is now known as nonsenso
[26-Jan-2010 23:24:49] <rmatte> eugh, I'm going to throw something... just completed a change where I restarted zenoss on all of my servers... now zenprocess on the main server that is monitoring them just decides to completely stop working
[26-Jan-2010 23:25:01] <rmatte> no matter what I try I can't get it to work again (and I've tried everything)
[26-Jan-2010 23:25:38] <rmatte> I've remodeled the devices several times, pushed changes several times, deleted the processes and rediscovered them, restarted zenprocess several times
[26-Jan-2010 23:25:47] <rmatte> nothing is working, I keep getting hundreds of process down events
[26-Jan-2010 23:26:05] <rmatte> I should have been in bed like an hour ago (that's how long I've been working to try and fix this)
[26-Jan-2010 23:50:27] <rmatte> I just tried a reindex since that's helped in the past, no improvement though
[26-Jan-2010 23:50:29] <rmatte> grrrr
[27-Jan-2010 00:00:30] [disconnected at Wed Jan 27 00:00:30 2010]
[27-Jan-2010 00:00:30] [connected at Wed Jan 27 00:00:30 2010]
[27-Jan-2010 00:03:29] <rmatte> finally figured this out, somehow one of the processes got changed to ignore parameters and it screwed up monitoring for all of the other python based processes
[27-Jan-2010 00:20:02] <rmatte> well, bed time is long overdue for me, ciao
[27-Jan-2010 01:03:41] fOrsberg is now known as forsberg
[27-Jan-2010 06:26:56] <cebalrai> My debian box is giving me a "Unable to create zenoss user" when installing using the deb package....
[27-Jan-2010 06:27:15] <cebalrai> Couldn't find anyone with the same problem
[27-Jan-2010 06:27:47] <tehhobbit> check that the user isnt already there
[27-Jan-2010 06:35:51] <cebalrai> hmm
[27-Jan-2010 06:35:53] <cebalrai> good point
[27-Jan-2010 06:41:15] <cebalrai> Is the user created using the helper binary? I was looking through the post/preinst scripts and couldn't find the lines
[27-Jan-2010 06:53:40] <tehhobbit> It should be created with the prescript otherwise you cant install the binaries with correct owner
[27-Jan-2010 07:02:27] <cebalrai> apperently my server is unable to create home directories...fun
[27-Jan-2010 07:04:08] <cebalrai> so it's not zenosses fault. Thanks though
[27-Jan-2010 07:46:53] <aprigio> hi After the installation of zenoss, the account user admin does not work. It is admin and password zenoss?
[27-Jan-2010 07:56:15] <forsberg> if its a new install you should be asked to put in password yourself?
[27-Jan-2010 08:00:24] <aprigio> no
[27-Jan-2010 08:00:55] <forsberg> what version did you install?
[27-Jan-2010 08:03:02] <forsberg> installed it a couple of times lately, and each time i was asked to set admin pass first time i browsed to server:8080
[27-Jan-2010 08:05:15] <ckrough> morning all
[27-Jan-2010 08:05:18] <aprigio> yes I know.Â But the account user admin not enter.Â This account is in Mysql?Â For the daemon creates an account in / etc / passwd with the name of zenoss.
[27-Jan-2010 08:05:18] <aprigio> I can not access the admin account.
[27-Jan-2010 08:06:22] <forsberg> hmm im not sure i understand wich account you really mean then, sorry :)
[27-Jan-2010 08:22:35] <forsberg> the account is not in mysql tho, i can tell you that, only thing in mysql is the event history
[27-Jan-2010 08:23:07] <forsberg> well not history, but events
[27-Jan-2010 08:56:14] <ckrough> I want to 'rebrand' my dev environment, overriding images in portal_skins/custom doesnt seem to do it
[27-Jan-2010 08:56:31] <ckrough> how would I override the stuff in /zenoss/Products/ZenModel/skins/zenmodel?
[27-Jan-2010 09:16:55] <rocket> venturaville: Eureka!!! .. :p
[27-Jan-2010 09:21:49] <venturaville> what's up?
[27-Jan-2010 09:22:10] <rocket> I have a working puppet type and provider :p
[27-Jan-2010 09:25:50] LarsN_Away is now known as LarsN
[27-Jan-2010 09:58:58] <rmatte> ckrough: pretty sure that some of that stuff is stored right in Zope
[27-Jan-2010 09:59:20] <rmatte> no idea how to override it, but if/when you figure it out be sure to make a wiki post about it
[27-Jan-2010 10:01:10] <forsberg> anyone who uses zenoss to monitor freebsd servers? :)
[27-Jan-2010 10:03:56] <ckrough> rmatte: found it, gotta modify the CSS in the skin in zope
[27-Jan-2010 10:09:59] <venturaville> rocket: congrats
[27-Jan-2010 10:12:57] <rmatte> ckrough: ah
[27-Jan-2010 10:13:21] <rmatte> rocket: congrats on finally getting that working lol
[27-Jan-2010 10:13:36] <rmatte> rocket: you've been wrestling with that for like over a month
[27-Jan-2010 10:21:20] <rocket> rmatte: well considering I had to learn ruby to do it thats not terrible .. and its not been full time
[27-Jan-2010 10:21:31] <rocket> rmatte: eg just when I had some free time etc .. ;p
[27-Jan-2010 10:22:34] <cgibbons> @#$!print
[27-Jan-2010 10:23:44] <rocket> cgibbons: Tree Killer ... !@#$#@eprint! :p
[27-Jan-2010 10:42:32] <theacolyte> Skipping Physical Memory (ram) as it matches zFileSystemMapIgnoreTypes.
[27-Jan-2010 10:42:46] <theacolyte> is that normal when snmp mapping, or is that maybe why I'm having issues at the moment with my wmi ram plugin
[27-Jan-2010 10:42:56] <rmatte> that's normal
[27-Jan-2010 10:43:06] <rmatte> ramdisk is ignored by default
[27-Jan-2010 10:43:07] <theacolyte> loverly
[27-Jan-2010 10:43:26] <rmatte> you're really supposed to do your memory monitoring on the perf tab, not in the filesystem section
[27-Jan-2010 10:43:45] <rmatte> hence why it's ignored
[27-Jan-2010 10:43:54] <theacolyte> makes sense
[27-Jan-2010 10:49:45] <dec3pti0n> I'm reading over section 6.1 of DOC-3915 , zenos administration 2.5.0, and it mentions "Monitors" as being part of the left side nav bar but I see nothing.
[27-Jan-2010 10:50:13] <dec3pti0n> I'm using 2.5.1 .. is Monitors now called the collectors ? cause that the only thing I can think of
[27-Jan-2010 10:53:04] <Muti> yes
[27-Jan-2010 10:54:24] <Muti> has anyone come up w/ a good solution for monitoring something w/ a dynamic IP? I need to monitor a storage cluster that has no master, just a hostname which is mapped to active nodes in the cluster
[27-Jan-2010 10:55:33] <Muti> the only thing I can think of is to create a COMMAND performance monitor that does a DNS lookup and updates the IP address each time it runs
[27-Jan-2010 11:05:32] <dec3pti0n> anyone here has encountered issues with zenos and apps that show up on netstat listening on tcp6 ?
[27-Jan-2010 11:06:03] <dec3pti0n> I'm not using ipv6 but debian seems to show some apps listening on tcp6
[27-Jan-2010 12:09:19] <chezgi> where can i find some documentations about designing scale-able high performance monitoring.
[27-Jan-2010 12:14:33] <rocket> chezgi: as in .. thats pretty vague .. eg I could just say zenoss.com as a reponse .. or http://en.wikipedia.org/wiki/C_%28programming_language%29 :p
[27-Jan-2010 12:18:36] <chezgi> rocket: zenoss used python, and it is expensive (cpu).
[27-Jan-2010 12:19:19] <chezgi> rocket: and i only asked for some design docs not software.
[27-Jan-2010 12:20:39] <rocket> chezgi: I would say look for general software design implementations for scalable high performance coding and you have your answer
[27-Jan-2010 12:21:18] <rocket> chezgi: you probably wont find anything specific to this domain unless your willing to pay big bucks for it
[27-Jan-2010 12:21:42] <chezgi> rocket: there is some issues in monitoring.
[27-Jan-2010 12:22:06] <rocket> chezgi: as far as zenoss code it trades off performance for flexibility .. as no single vendor implements monitoring the same way ..
[27-Jan-2010 12:22:30] <rocket> eg you have snmp, snmp traps, custom software, email alerts ... you name it
[27-Jan-2010 12:22:49] <chezgi> rocket: i developed custom monitoring software for my distributed application but it can't be scaled good.
[27-Jan-2010 12:22:59] <rmatte> eugh, wish I knew perl better... trying to fix a ziptie adapter for HP ProCurve switches.
[27-Jan-2010 12:23:47] <rocket> chezgi: ok so you are looking for docs on how to improve your custom monitoring software?
[27-Jan-2010 12:24:11] <chezgi> rocket: yes.
[27-Jan-2010 12:24:26] <rocket> chezgi: what language?
[27-Jan-2010 12:25:14] <chezgi> rocket: i used python as main but core sections are C (connected with swig).
[27-Jan-2010 12:26:03] <chezgi> rocket: i need only theories not implementation.
[27-Jan-2010 12:26:22] <rocket> chezgi: then I would say google is your new best friend ..
[27-Jan-2010 12:26:45] <rocket> because even though you are monitoring .. you seem to have a specific problem domain outside of simple monitoring
[27-Jan-2010 12:59:49] <rmatte> k
[27-Jan-2010 12:59:52] <rmatte> erm
[27-Jan-2010 13:00:35] <x-ip> hi ... i'm having problems with zenoss and sending the alerts emails
[27-Jan-2010 13:01:24] <x-ip> i configured the smtp at zenoss ... and at the logs from the smtp server (a functional postfix server) i'm gettinz zenoss connecting to the port 25 and disconnecting to fast, without sending the email
[27-Jan-2010 13:01:30] <x-ip> any ideas about this situation ? :-S
[27-Jan-2010 13:02:31] <x-ip> http://pastebin.ca/1768327 <-- here are the smtp server logs
[27-Jan-2010 13:03:50] <eidolon> hi folks. i have a machine in zenoss that is tmporarily off line.
[27-Jan-2010 13:03:55] <eidolon> i've set it's State to 'Decommissioned'
[27-Jan-2010 13:04:09] <eidolon> but i'm still getting monitoring alerts for it. like, i cleared all the alerts for it, moved to history... and i'm still getting:
[27-Jan-2010 13:04:23] <eidolon> Could not read the status of Windows services (NT_STATUS_HOST_UNREACHABLE). Check your username/password settings and verify n
[27-Jan-2010 13:08:48] <mrayzenoss> eidolon: I don't see any tickets that look like that, I wonder if anyone else has seen that
[27-Jan-2010 13:09:13] <eidolon> i wonder if something just didn't get 'updated' when i changed state. unfortunately, the answer may be the Microsoft answer. "Reboot zenoss"
[27-Jan-2010 13:32:41] <cgibbons> have you tried the 'push config' menu option at all?
[27-Jan-2010 13:51:16] <gwb235> what / is there a "transform" that drops an event entirely (i.e. doesn't make an entry in the mysql database?)
[27-Jan-2010 13:51:33] <rmatte> gwb235: evt._action = 'drop'
[27-Jan-2010 13:51:41] <rmatte> or set action in zproperties for the event mapping to drop
[27-Jan-2010 14:01:09] <gwb235> if evt.summary = "calling function 'setProductKey' with": evt._action = 'drop' ?
[27-Jan-2010 14:04:08] <rmatte> no
[27-Jan-2010 14:04:11] <rmatte> import re
[27-Jan-2010 14:04:55] <rmatte> if re.match("calling function 'setProductKey' with", evt.summary): evt._action = 'drop'
[27-Jan-2010 14:05:12] <rmatte> you can't compare string with an if whatever =
[27-Jan-2010 14:05:40] <gwb235> ahh... and if the event class is /Change/Set use zProperties to drop
[27-Jan-2010 14:05:52] <gwb235> (in that /events/change/set hierarchy)
[27-Jan-2010 14:06:19] <gwb235> which is "better" ?
[27-Jan-2010 14:06:58] <HCoyote> is it safe to run the following sql delete on events.history?
[27-Jan-2010 14:06:59] <HCoyote> delete from history where summar like "calling function 'setProductKey' with%";
[27-Jan-2010 14:07:07] <HCoyote> s/summar/summary/
[27-Jan-2010 14:07:32] <gwb235> note: HCoyote and I are working together to clear the mysql DB
[27-Jan-2010 14:10:29] forsberg is now known as fOrsberg
[27-Jan-2010 14:10:40] <rmatte> clear it of what?
[27-Jan-2010 14:11:05] <HCoyote> the 5.9 million rows matching "Calling function 'setProductKey' with"
[27-Jan-2010 14:11:32] <HCoyote> considering there are only 6.2 million rows in history, this information seems ... superfluous.
[27-Jan-2010 14:12:54] <gwb235> the event seems to be created from ssh linux rpm -qa type output, foreach host, foreach rpm installed.
[27-Jan-2010 14:13:13] <HCoyote> and it seems to do it everyday.
[27-Jan-2010 14:13:56] <rmatte> generally those types of events are generated from modelling devices I believe
[27-Jan-2010 14:14:02] <Apachez> optimize table history; is your friend
[27-Jan-2010 14:14:14] <Apachez> however you need to do more to recall the empty space
[27-Jan-2010 14:14:40] <gwb235> i'm not too worried about recapturing the used space (only 10G at this point) just don't want the same growth rate in the future
[27-Jan-2010 14:15:55] <HCoyote> optimize doesn't really recover any space on innodb if you're using a single tablespace for everything.
[27-Jan-2010 14:16:33] <Apachez> as I said
[27-Jan-2010 14:16:38] <Apachez> hwoever it will fix your index
[27-Jan-2010 14:16:49] <Apachez> switch to one file per table instead to optimize diskusage
[27-Jan-2010 14:16:54] fOrsberg is now known as forsberg
[27-Jan-2010 14:16:58] <Muti> does optimize write lock the table?
[27-Jan-2010 14:16:59] <Apachez> you might also want to tweak some other settings aswell
[27-Jan-2010 14:16:59] <HCoyote> there's nothign wrong with the index right now. data hasn't been deleted yet.
[27-Jan-2010 14:17:11] <Apachez> Muti: only on myisam
[27-Jan-2010 14:17:15] <Apachez> innodb is transactional
[27-Jan-2010 14:17:38] <Apachez> index is most likely not sorted if you never sorted it
[27-Jan-2010 14:18:08] <HCoyote> so, anyway, is it safe to run that delete in history or are there additional bits in the database that need to be cleaned out when I do this?
[27-Jan-2010 14:21:36] <rmatte> no additional bits
[27-Jan-2010 14:21:45] <HCoyote> thanks!
[27-Jan-2010 14:21:47] <rmatte> they are single entries, they don't link to anything
[27-Jan-2010 14:23:14] <mrayzenoss> http://dev.zenoss.org/trac/browser/trunk/Products/ZenEvents/db/zenevents.sql
[27-Jan-2010 14:23:34] <mrayzenoss> you could probably clear up the log and details
[27-Jan-2010 14:33:00] <dec3pti0n> under the template for Linux device, there is the threshold for high load. The max value shows 1200 , I'm a bit confused on that number
[27-Jan-2010 14:33:25] <dec3pti0n> the 5min load needs to reach 1200 in order to alert ?
[27-Jan-2010 14:33:36] <HCoyote> mrayzenoss: I'll look at those too. they aren't anywhere near as large as history at this point.
[27-Jan-2010 14:35:31] <rmatte> I wish Zenoss had a feature to collect and store configs for network devices
[27-Jan-2010 14:35:47] <rmatte> then I wouldn't be mucking around with ZipTie lol
[27-Jan-2010 14:36:18] <rmatte> does anyone know of any other open source config management tools?
[27-Jan-2010 14:36:51] <dec3pti0n> there is one called puppet if this is what you are talking about
[27-Jan-2010 14:37:03] <rmatte> no, network configuration management
[27-Jan-2010 14:37:10] <dec3pti0n> ah sorry
[27-Jan-2010 14:37:14] <rmatte> like, for automatically backing up the configs off network devices
[27-Jan-2010 14:37:31] <rmatte> the only free one that I know of is ziptie, but I'm surprised that's the only one
[27-Jan-2010 14:39:18] <Muti> Apachez: OPTIMIZE TABLE does put a write lock on the table by the way
[27-Jan-2010 14:41:00] <ckrough> http://blog.sflow.com/2010/01/open-vswitch.html
[27-Jan-2010 14:46:47] <Apachez> Muti: as I know the other transactions will continue to work during the optimize process
[27-Jan-2010 14:56:33] <theacolyte> rmatte: rancid
[27-Jan-2010 14:56:42] <theacolyte> rmatte: I use kiwi cattools tho
[27-Jan-2010 14:56:50] <rmatte> rancid is open source?
[27-Jan-2010 14:57:16] <theacolyte> cha
[27-Jan-2010 14:57:39] <rmatte> hmmmm, does it have a gui though?
[27-Jan-2010 14:58:04] <rmatte> doesn't seem to, just looks like plain cvs
[27-Jan-2010 14:58:16] <ckrough> it integreates with zenoss
[27-Jan-2010 14:58:29] <rmatte> ckrough: oh? how?
[27-Jan-2010 14:58:49] <ckrough> http://community.zenoss.org/docs/DOC-4270
[27-Jan-2010 14:59:00] <ckrough> meh, Enterprise :)
[27-Jan-2010 14:59:05] <rmatte> will it backup actual device configs?
[27-Jan-2010 14:59:08] <ckrough> yes
[27-Jan-2010 14:59:14] <rocket> rmatte: yes
[27-Jan-2010 14:59:15] <ckrough> except for vlan.dat
[27-Jan-2010 14:59:17] <rmatte> is there a list of device types it supports?
[27-Jan-2010 14:59:28] <ckrough> you can modify the scripts it uses to support just about anything
[27-Jan-2010 14:59:41] <ckrough> basically a 'show run' and the config gets checked into SVN
[27-Jan-2010 14:59:59] <ckrough> I imagine it could work with any OS that can dump its config to tftp or console
[27-Jan-2010 15:00:29] <rmatte> we are using ziptie right now and it's working great, but we need to be able to backup the config for HP switches, and the adapter script for that isn't working properly
[27-Jan-2010 15:00:31] <ckrough> so, for example, I can find a device in zenoss and click on the rancid link to see all of it's backups
[27-Jan-2010 15:00:37] <rmatte> I've been pouring over the code trying to fix it
[27-Jan-2010 15:00:50] <rmatte> hmmm, that's pretty cool
[27-Jan-2010 15:01:05] <ckrough> yeah, and you can set up events to trigger the backup from snmptraps
[27-Jan-2010 15:01:11] <ckrough> and/or cron
[27-Jan-2010 15:01:23] <ckrough> or any zenoss event for that matter
[27-Jan-2010 15:01:31] <rmatte> rancid doesn't look like it has hp procurve support out of the box
[27-Jan-2010 15:01:31] <theacolyte> I may actually buy the commercial version
[27-Jan-2010 15:01:42] <theacolyte> be back in a bit salesforce meeting
[27-Jan-2010 15:01:55] <rmatte> oh it does
[27-Jan-2010 15:03:57] <rmatte> oh, the rancid integration is for enterprise only
[27-Jan-2010 15:03:59] <rmatte> bleh
[27-Jan-2010 15:08:03] <rmatte> would have been nice
[27-Jan-2010 15:08:50] <Egyptian[Home]> evening all
[27-Jan-2010 15:09:59] <theacolyte> rmatte: is that ziptie app free?
[27-Jan-2010 15:12:40] <ckrough> 'the ZipTie product has matured and stabilized enough to become part of the NetworkAuthority family of products.'
[27-Jan-2010 15:12:44] <venturaville> how do you flip axis for a graph point without making it a negative number?
[27-Jan-2010 15:12:48] * ckrough queues sad trombone
[27-Jan-2010 15:16:54] <Apachez> http://www.youtube.com/watch?v=CzbURUrgQao
[27-Jan-2010 15:24:14] LarsN is now known as LarsN_Away
[27-Jan-2010 15:25:12] <rmatte> Apachez: lol
[27-Jan-2010 15:25:25] <rmatte> theacolyte: well, it's free but the project was discontinued
[27-Jan-2010 15:25:46] <rmatte> theacolyte: the company funding the project merged the code in to one of their products which is free to use but only for personal use (hence as a company we can't use it)
[27-Jan-2010 15:25:59] <rmatte> theacolyte: so we're using the last version which was available
[27-Jan-2010 15:26:17] <rmatte> and if I could get it to backup procurve devices then we'd be good
[27-Jan-2010 15:28:54] <rmatte> but the "adapter" code for HP ProCurve devices is screwed
[27-Jan-2010 15:29:12] <rmatte> one of the guys here did a great job of fixing it so far, but we just can't figure out how to finish fixing it
[27-Jan-2010 15:30:11] <theacolyte> rmatte: zip it up and send it over :)
[27-Jan-2010 15:30:47] <rmatte> zip what up?
[27-Jan-2010 15:30:53] <theacolyte> the old version
[27-Jan-2010 15:31:10] <rmatte> http://dmon.org/downloads/ziptie/
[27-Jan-2010 15:31:17] <rmatte> there's the servers and clients for both windows and linux
[27-Jan-2010 15:31:26] <rmatte> oh, and mac
[27-Jan-2010 15:31:47] <theacolyte> does iti nstall its own webserver?
[27-Jan-2010 15:31:51] <theacolyte> i guess ill spin up a vm
[27-Jan-2010 15:31:59] <rmatte> there is no webserver, you connect with the client
[27-Jan-2010 15:31:59] <theacolyte> after this meeting with salesforce
[27-Jan-2010 15:32:03] <theacolyte> ah
[27-Jan-2010 15:32:09] <rmatte> but the client works very well
[27-Jan-2010 15:32:24] <theacolyte> and it does archiving, what else?
[27-Jan-2010 15:32:28] <theacolyte> ios upgrades?
[27-Jan-2010 15:32:40] <rmatte> not sure about upgrades, but you can have it perform config changes
[27-Jan-2010 15:32:59] <theacolyte> do you run it on your zenoss servers or seperately?
[27-Jan-2010 15:33:13] <rmatte> on my zenoss servers
[27-Jan-2010 15:33:20] <theacolyte> cool
[27-Jan-2010 15:38:25] <rmatte> it's been working flawlessly for cisco gear
[27-Jan-2010 15:38:29] <rmatte> just not for HP
[27-Jan-2010 15:38:30] <rmatte> lol
[27-Jan-2010 15:38:47] <theacolyte> that'll teach you to buy crappy hp networking gear
[27-Jan-2010 15:38:49] <theacolyte> hehe
[27-Jan-2010 15:38:57] <rmatte> it's our clients who have it, not us
[27-Jan-2010 15:39:08] <rmatte> we have a new client coming on who has like 40 procurve switches
[27-Jan-2010 15:39:11] <theacolyte> oh, right
[27-Jan-2010 15:39:12] <rmatte> so we need this to work
[27-Jan-2010 15:39:15] <theacolyte> ouch
[27-Jan-2010 15:39:25] <theacolyte> it's not free, but cattools is pretty good
[27-Jan-2010 15:39:45] <rmatte> yeh, I've used kiwi cattools before
[27-Jan-2010 15:39:58] <rmatte> but we want to avoid spending money if we can
[27-Jan-2010 15:40:14] <theacolyte> I want acar that doesn't get dirty - go!
[27-Jan-2010 15:40:31] <rmatte> kiwi cattools also only runs on windows doesn't it?
[27-Jan-2010 15:40:36] <theacolyte> yea
[27-Jan-2010 15:40:47] <theacolyte> rancid baby
[27-Jan-2010 15:40:48] <rmatte> yea, that's not ideal, since that means we'd need an extra vm for every client
[27-Jan-2010 15:40:50] <theacolyte> that's about it
[27-Jan-2010 15:41:11] <rmatte> well, I really like ziptie, hopefully we can get the script to work
[27-Jan-2010 15:41:19] <theacolyte> looks like I got the $ for enterprise :)
[27-Jan-2010 15:41:26] <theacolyte> is mrayzenoss sales?
[27-Jan-2010 15:41:29] <ckrough> nah
[27-Jan-2010 15:41:52] <rmatte> If we do I'm going to create a website called ziptie-project.org or something and host all of the files and a forum and stuff
[27-Jan-2010 15:42:03] <theacolyte> lol
[27-Jan-2010 15:42:12] <rmatte> maybe get some interest back in to it, because honestly it's a pretty completed project
[27-Jan-2010 15:42:21] <rmatte> they've just stopped hosting anything about it
[27-Jan-2010 15:42:21] <mrayzenoss> theacolyte: nope, I'm community/product management
[27-Jan-2010 15:42:25] <theacolyte> I'll have to check it
[27-Jan-2010 15:42:31] <theacolyte> mrayzenoss: Got it, I'll just call your sales $
[27-Jan-2010 15:42:43] <mrayzenoss> rmatte: the parent company got merged with another company recently
[27-Jan-2010 15:42:47] <mrayzenoss> Alterpoint
[27-Jan-2010 15:42:50] <theacolyte> After I pay, you will *have* to listen to me! :)
[27-Jan-2010 15:42:52] <theacolyte> hehe
[27-Jan-2010 15:43:14] <rmatte> mrayzenoss: yeh, and the code got pwned because of it
[27-Jan-2010 15:43:21] <Rolfs1> hmm.. the validation email from zenoss went into the greylisting
[27-Jan-2010 15:43:31] <rmatte> mrayzenoss: but the original project is open source, so I can host it
[27-Jan-2010 15:43:38] <theacolyte> honestly, how much "better" is the WMI in enterprise vs Egors plugin?
[27-Jan-2010 15:43:45] <rmatte> (although I'd have to dig to find the original sourcecode, I only have the binaries
[27-Jan-2010 15:43:50] <rmatte> but the binaries are enough really)
[27-Jan-2010 15:44:10] <rmatte> theacolyte: much better
[27-Jan-2010 15:44:20] <rmatte> theacolyte: you can monitor interfaces and harddisks via WMI with it
[27-Jan-2010 15:44:20] <Rolfs1> mm.. also in the reset password email: "We have received a request from 127.0.0.1 to reset your password"
[27-Jan-2010 15:44:30] <theacolyte> can you make your own graphs using it?
[27-Jan-2010 15:44:35] <rmatte> theacolyte: yeh
[27-Jan-2010 15:44:40] <theacolyte> awesome
[27-Jan-2010 15:44:53] <theacolyte> Gotta get some pricing
[27-Jan-2010 15:44:56] <theacolyte> must get out of this meeting
[27-Jan-2010 15:45:01] <rmatte> though egor was talking about getting interface/disk monitoring implemented in core via WMI in the future
[27-Jan-2010 15:45:04] <rmatte> so we'll see what happens
[27-Jan-2010 15:45:05] <mrayzenoss> theacolyte: cgibbons is the man to bug about that
[27-Jan-2010 15:45:27] <theacolyte> which part? sales or wmi?
[27-Jan-2010 15:45:41] <mrayzenoss> WMI
[27-Jan-2010 15:45:46] <theacolyte> got it
[27-Jan-2010 15:46:16] <mrayzenoss> basically Enterprise has WMI modeling and perfmon monitoring, community has Egor's WMI monitoring
[27-Jan-2010 15:46:33] <theacolyte> right
[27-Jan-2010 15:46:53] <theacolyte> I have run into (as yuou know) issues with the WMI zenpack so that's my main motivation
[27-Jan-2010 15:47:04] <theacolyte> well that and support
[27-Jan-2010 15:47:12] <theacolyte> not that you guys don't rock
[27-Jan-2010 15:47:32] <theacolyte> what's the upgrade path like?
[27-Jan-2010 15:47:41] cgibbons_ is now known as cgibbons
[27-Jan-2010 15:48:49] <Muti> theacolyte: if you're using RPMs you just install the enterprise RPM
[27-Jan-2010 15:49:00] <theacolyte> I used the stack installer
[27-Jan-2010 15:49:05] <straterra> I just moved my zenoss server to a new ip, but check_ntp is still trying to use the old IP
[27-Jan-2010 15:49:13] <theacolyte> I don't really feel like recreating all my stuff
[27-Jan-2010 15:49:24] <straterra> I changed the IP in zenoss and I've NO idea where its pulling the old addy
[27-Jan-2010 15:49:38] <mrayzenoss> theacolyte: client services is very familiar with migrating from Core to Enterprise
[27-Jan-2010 15:49:54] <theacolyte> mrayzenoss: excellent, included in purchase?
[27-Jan-2010 15:49:57] <rmatte> straterra: from your hosts file buddy
[27-Jan-2010 15:50:02] <theacolyte> or is it additional professional sercices
[27-Jan-2010 15:50:03] <straterra> hosts file?
[27-Jan-2010 15:50:10] <mrayzenoss> theacolyte: included I'm sure
[27-Jan-2010 15:50:12] <rmatte> actually, you mean you changed the IP on a device in Zenoss?
[27-Jan-2010 15:50:16] <straterra> You're right
[27-Jan-2010 15:50:16] <theacolyte> excellent
[27-Jan-2010 15:50:19] <theacolyte> I be calling
[27-Jan-2010 15:50:22] <straterra> It was the host file
[27-Jan-2010 15:50:25] <rmatte> ;)
[27-Jan-2010 15:51:01] <straterra> Thanks
[27-Jan-2010 15:51:04] <rmatte> np
[27-Jan-2010 16:00:08] <sciolist> when applying patches with zenpatch, is it normal for hunks to fail?
[27-Jan-2010 16:00:55] <sciolist> I've been trying to apply patch #15257 and three hunks failed when running zenpatch
[27-Jan-2010 16:02:05] <gwb235> i didn't have any failures :)
[27-Jan-2010 16:02:24] <gwb235> urgh
[27-Jan-2010 16:02:27] <sciolist> well aren't you special
[27-Jan-2010 16:02:31] <mrayzenoss> that patch is against trunk, so it's probably not going to apply cleanly
[27-Jan-2010 16:02:38] <mrayzenoss> what version are you running/
[27-Jan-2010 16:02:40] <mrayzenoss> ?
[27-Jan-2010 16:02:47] <sciolist> 2.4.2 IIRC
[27-Jan-2010 16:02:52] <gwb235> 2.4.5
[27-Jan-2010 16:02:58] <sciolist> doh!
[27-Jan-2010 16:03:10] <sciolist> yeah, I *did* upgrade it, didn't I?
[27-Jan-2010 16:03:25] <gwb235> it's that number after "Version" on the login page :)
[27-Jan-2010 16:04:06] <sciolist> different window, different views
[27-Jan-2010 16:10:12] <sciolist> mrayzenoss: so things should still work, even though the patch didn't apply all the way?
[27-Jan-2010 16:16:14] <cgibbons> magic 8 ball says Outlook not so good
[27-Jan-2010 16:16:23] <rmatte> lol
[27-Jan-2010 16:17:18] <forsberg> :)
[27-Jan-2010 16:17:44] <sciolist> but then we knew about Outlook....
[27-Jan-2010 16:37:51] <rmatte> I think I might have figured out why this HP template isn't working, it's trying to look in an ancient repository location (to store/retrieve the configs) for some reason
[27-Jan-2010 16:38:00] <rmatte> and the location doesn't exist so it won't work
[27-Jan-2010 16:40:10] <rmatte> now I just need to figure out why it's going to the old location instead of the new one
[27-Jan-2010 16:46:20] <theacolyte> Is there an easy way to get a device count?
[27-Jan-2010 16:46:39] <theacolyte> besides clicking on devices and reading the number, i mean
[27-Jan-2010 16:46:45] <theacolyte> :(
[27-Jan-2010 16:48:39] <rmatte> in where?
[27-Jan-2010 16:48:54] <rmatte> Zenoss?
[27-Jan-2010 16:49:04] <theacolyte> Forget it, I answered my on really stupid question
[27-Jan-2010 16:49:09] <rmatte> lol
[27-Jan-2010 16:51:12] <theacolyte> checking out ziptie now
[27-Jan-2010 16:52:32] <theacolyte> or rather, I would if I could find my VM
[27-Jan-2010 16:55:08] <theacolyte> How could zenoss know the % used on this device, but not know available bytes
[27-Jan-2010 16:55:13] <theacolyte> or capacity
[27-Jan-2010 16:57:14] <rmatte> % disk used?
[27-Jan-2010 16:57:23] <theacolyte> yea
[27-Jan-2010 17:10:30] forsberg is now known as fOrsberg
[27-Jan-2010 17:17:20] <rmatte> I can't for the life of me figure out why ziptie is trying to use a different (non-existant) repository for HP devices, but uses the correct repository for Cisco devices
[28-Jan-2010 00:00:30] [disconnected at Thu Jan 28 00:00:30 2010]
[28-Jan-2010 00:00:31] [connected at Thu Jan 28 00:00:31 2010]
[28-Jan-2010 00:23:53] <Apachez> http://cache.gawker.com/assets/images/comment/4/2010/01/5973697c09649e669716803dd9820485/original.jpg
[28-Jan-2010 01:29:30] fOrsberg is now known as forsberg
[28-Jan-2010 02:23:02] <frozty_sa> man, that URL restructuring on the zenoss site really messes with trying to get info out of the old forum and list archives
[28-Jan-2010 03:50:02] <zenoss> hai
[28-Jan-2010 03:50:21] <zenoss> how to send zenoss alerts to irc?
[28-Jan-2010 05:16:02] <Rolfs> how can I loop through a template check and change all datapoints to be derive instead of counter?
[28-Jan-2010 05:16:09] <Rolfs> with zendmd
[28-Jan-2010 05:54:48] snowgod is now known as admgre
[28-Jan-2010 08:51:01] <straterra> Ouch..I'm still on Zenoss 2.4.5
[28-Jan-2010 09:28:12] <rmatte> straterra: so am I
[28-Jan-2010 09:29:56] <gwb2353> me too!
[28-Jan-2010 09:30:17] <gwb2353> although I'm thinking 2.5.2 will be a cutover point
[28-Jan-2010 09:33:39] <ckrough> 2.3.3
[28-Jan-2010 09:46:16] <rmatte> yeh, 2.5.2 will be the cutover point for me
[28-Jan-2010 09:46:28] <rmatte> it's looking pretty good, and they fixed an issue which I needed fixed
[28-Jan-2010 09:56:17] <Apachez> is the slow zope fixed in 2.5.2? :)
[28-Jan-2010 09:57:24] <rmatte> Apachez: highly doubt it, and I honestly don't find it overly slow for the amount of relations that it's handling
[28-Jan-2010 09:57:58] <ckrough> what are your zserver-threads set at :)
[28-Jan-2010 09:58:54] <rmatte> yeh, I changed my threads from 50 to 4 and it massively improved performance
[28-Jan-2010 09:59:41] <ckrough> I went from 25 to 7, same
[28-Jan-2010 09:59:47] <ckrough> and SWAP issues seem to be resolved as well
[28-Jan-2010 10:04:06] forsberg is now known as fOrsberg
[28-Jan-2010 10:04:54] <Apachez> I have tried 1 aswell without much of improvement
[28-Jan-2010 10:07:29] fOrsberg is now known as forsberg
[28-Jan-2010 10:08:37] <rmatte> ckrough: yeh, same here, the memory issue seems to possibly be gone, though we won't truly know until a couple weeks from now
[28-Jan-2010 10:08:48] <rmatte> since the memory usage has a tendency to creep, and it may just be creeping more slowly now
[28-Jan-2010 10:09:19] <rmatte> Apachez: the threads should match the number of cores in the box
[28-Jan-2010 10:09:27] <rmatte> unless it's a single core system, then the default of 4 works fine
[28-Jan-2010 10:10:46] <Apachez> however zope is singlethreaded
[28-Jan-2010 10:10:52] <Apachez> or rather doesnt to threading that well
[28-Jan-2010 10:10:59] <Apachez> according to some google sources
[28-Jan-2010 10:11:49] <ckrough> may want to leave one free for OS/other
[28-Jan-2010 10:12:16] <rmatte> Apachez: well, it's handling it fine with the default setting of 4
[28-Jan-2010 10:12:26] <ckrough> also note that the number of connections to the zope database is hardcoded at 7: http://www.zope.org/Documentation/Books/ZopeBook/2_6Edition/MaintainingZope.stx
[28-Jan-2010 10:12:41] <ckrough> so if you go over 6 threads you should change that maximum in the code
[28-Jan-2010 10:12:52] <rmatte> ah
[28-Jan-2010 10:13:19] <rmatte> I don't know why you'd want to honestly, the performance increase with dropping it was quite significant
[28-Jan-2010 10:17:08] <straterra> rmatte: no real reason to upgrade?
[28-Jan-2010 10:20:07] <rmatte> straterra: there's a big reason to upgrade, so that you're not 2 releases behind by the time stone crab comes out
[28-Jan-2010 10:20:21] <rmatte> because you always have to follow a release path so you'll be doing the 2.5 upgrade one way or another
[28-Jan-2010 10:20:34] <rmatte> and I personally think that doing 2 or more upgrades at the same time can be quite messy
[28-Jan-2010 10:20:57] <rmatte> also, 2.5 offers improved zenperfsnmp performance over 2.4
[28-Jan-2010 10:21:03] <straterra> hmm
[28-Jan-2010 10:21:04] <rmatte> and there's a wackload of other fixes
[28-Jan-2010 10:21:13] <straterra> Is the upgrade difficult?
[28-Jan-2010 10:21:13] <eidolon> hi folks, me again.
[28-Jan-2010 10:21:22] <rmatte> straterra: how did you install?
[28-Jan-2010 10:21:22] <venturaville> modeling is much faster under 2.5 as well.....
[28-Jan-2010 10:21:30] <straterra> rpms, I think
[28-Jan-2010 10:21:31] <straterra> On CentOS
[28-Jan-2010 10:21:42] <rmatte> venturaville: that's excellent to hear
[28-Jan-2010 10:21:51] <eidolon> is there a seperate template for windows / WMI boxes for checking disk space thresholds than linux boxen? APpears to all be under /Devices/Server. we'd like to set thresholds different on Windows boxes than on Linux boxen.
[28-Jan-2010 10:21:52] <straterra> Yup, rpm's
[28-Jan-2010 10:22:06] <rmatte> straterra: then it's obviously not difficult to update, just grab the newer RPMs and install them
[28-Jan-2010 10:22:11] <straterra> And will my zenpacks and stuff work?
[28-Jan-2010 10:22:16] <rmatte> straterra: but obviously wait for the full 2.5.2 release
[28-Jan-2010 10:22:22] <rmatte> they are only on beta right now I believe
[28-Jan-2010 10:22:46] <straterra> Ok. I'll wait for 2.5.2 and upgrade. I don't have to do anything with the existing data or anything?
[28-Jan-2010 10:22:52] <ckrough> straterra: make a copy of $ZENHOME before you upgrade just in case.
[28-Jan-2010 10:23:19] <rmatte> straterra: well, there is an install guide which details precautions you can take beforehand (backing up data and such)... is your Zenoss server a VM by chance?
[28-Jan-2010 10:23:20] <ckrough> straterra: most of what you need is in the mysql db and in data.fs under $ZENHOME
[28-Jan-2010 10:23:33] <ckrough> straterra: I believe there are instructions for backing those up in the upgrade documentation
[28-Jan-2010 10:23:45] <straterra> rmatte: nope..physical
[28-Jan-2010 10:23:56] <rmatte> straterra: ah, then yeh, you'll want to read the upgrade guide
[28-Jan-2010 10:24:04] <rmatte> I have VMs so I just take snapshots beforehand
[28-Jan-2010 10:33:46] * eidolon asks again just to check
[28-Jan-2010 10:34:07] <eidolon> can i set disk space thresholds on my windows boxen different than the linux boxen? they seem to both be using /Devices/Server for 'FileSystem'
[28-Jan-2010 10:34:46] <eidolon> which is odd. /devices/server/templates/filesystem/high disk usage - max value is here.getTotalBlocks() * .9
[28-Jan-2010 10:34:49] <eidolon> throws a warning
[28-Jan-2010 10:34:56] <eidolon> but i'm seeing, on the board now, warnings for 80% thresholds.
[28-Jan-2010 10:35:31] <ckrough> if they are in /Devices/Server/Windows you can copy the template at /Windows and just customize it there.
[28-Jan-2010 10:36:12] <eidolon> AHA!
[28-Jan-2010 10:36:13] <eidolon> /DevicesÂ /ServerÂ /WindowsÂ /WMIÂ /TemplatesÂ /FileSystem
[28-Jan-2010 10:36:15] <eidolon> there it is :)
[28-Jan-2010 10:36:19] <theacolyte> There's a make a local copy button
[28-Jan-2010 10:36:24] <ckrough> ^^
[28-Jan-2010 10:36:26] * eidolon steps on theacolyte
[28-Jan-2010 10:37:07] <theacolyte> =/
[28-Jan-2010 10:37:43] <eidolon> okee, i think i change dit to warning on 90% and alert on 95%
[28-Jan-2010 10:37:59] <eidolon> i dont need to remodel my hosts, do i? (altered the template definition)
[28-Jan-2010 10:39:03] <mrayzenoss> FYI, the community site just got reskinnedâ€¦ looking for feedback
[28-Jan-2010 10:39:17] <eidolon> community.zenoss.org ?
[28-Jan-2010 10:39:39] <eidolon> looks very Joomla :)
[28-Jan-2010 10:39:44] <theacolyte> It looks exactly like vmware's
[28-Jan-2010 10:39:49] <theacolyte> (of course they use jive, too)
[28-Jan-2010 10:39:50] <rmatte> mrayzenoss: I liked the old look
[28-Jan-2010 10:39:58] <theacolyte> I actually prefer this
[28-Jan-2010 10:40:00] <mrayzenoss> rmatte: that was the default
[28-Jan-2010 10:40:15] <rmatte> actually
[28-Jan-2010 10:40:17] <rmatte> hmmm
[28-Jan-2010 10:40:18] <mrayzenoss> now we have the same skin as the .com site
[28-Jan-2010 10:40:25] <rmatte> let me browser a bit, I'm sure I'll come around ;)
[28-Jan-2010 10:40:29] <rmatte> browse*
[28-Jan-2010 10:41:02] <ckrough> is unlegacy a word :)
[28-Jan-2010 10:41:02] <theacolyte> Thankfully that, unlike vmware, your site doesn't suck
[28-Jan-2010 10:41:09] <theacolyte> no
[28-Jan-2010 10:41:15] <ckrough> that download button feels out of place.
[28-Jan-2010 10:41:18] <theacolyte> if it's not legacy, there's no reason to say it
[28-Jan-2010 10:41:23] <rmatte> I think it would look better if the top bar had a bit of a gradient effect like the bottom one, other than that it's actually pretty nice
[28-Jan-2010 10:41:46] <ckrough> still lacking on the pony factor
[28-Jan-2010 10:42:19] <rmatte> ^^ I agree
[28-Jan-2010 10:42:27] <ckrough> otherwise looking good
[28-Jan-2010 10:43:17] <mrayzenoss> the bottom bar is going to get redone with more navigation shortcuts for the community site
[28-Jan-2010 10:44:06] <rmatte> the top and bottom bars should both have some form of gradient, the solid colour just doesn't do it for me
[28-Jan-2010 10:44:41] <mrayzenoss> I think they had a gradient at one point and zebra stripes at another, I'm sure there was a reason it's solid now
[28-Jan-2010 10:45:06] <mrayzenoss> maybe the gradient made the links harder to read?
[28-Jan-2010 10:45:08] <rmatte> well, the zebra stripes would be a bit much
[28-Jan-2010 10:45:10] <mrayzenoss> I don't recall
[28-Jan-2010 10:45:20] <rmatte> that's why you fade upwards, not downward
[28-Jan-2010 10:45:29] <rmatte> so that the background for the links is black
[28-Jan-2010 10:45:30] <forsberg> fairy trails after mouse would be uber
[28-Jan-2010 10:45:31] <forsberg> ;(
[28-Jan-2010 10:45:35] <rmatte> but it fades in to gray
[28-Jan-2010 10:45:52] <mrayzenoss> forsberg: lol
[28-Jan-2010 10:46:04] <forsberg> sorry had to say something stupid, have been busy and serious all day ;)
[28-Jan-2010 10:46:45] <rmatte> lol
[28-Jan-2010 10:47:58] <mrayzenoss> I think the "Unlegacy" was supposed to be "Open Source", so that'll be fixed soon
[28-Jan-2010 10:48:11] <rmatte> yeh, Unlegacy sounds kind of off
[28-Jan-2010 10:48:17] <ckrough> agreed
[28-Jan-2010 10:48:26] <rmatte> mrayzenoss: instead of Open Source, just put Open
[28-Jan-2010 10:48:34] <rmatte> it'll fit better
[28-Jan-2010 10:48:38] <rmatte> Open IT Management
[28-Jan-2010 10:49:00] <ckrough> wandering into opennms confusion possibly
[28-Jan-2010 10:49:12] <rmatte> doubt it
[28-Jan-2010 10:49:32] <rmatte> If anything it'll attract people trying to find the OpenNMS site to the Zenoss site ;)
[28-Jan-2010 10:49:58] <rmatte> Plus I'm pretty sure the huge "ZENOSS" above it will trump any confusion
[28-Jan-2010 10:51:14] <rmatte> Oooh, I'm a "stalwart"
[28-Jan-2010 10:51:16] <rmatte> http://community.zenoss.org/blogs/zenossblog/2010/01/25/tip-of-the-month-monitoring-the-status-of-nfs-shares-with-zenoss
[28-Jan-2010 10:51:37] <rmatte> ;)
[28-Jan-2010 10:51:46] <mrayzenoss> I think we've already got the graphics that say "Open Source"
[28-Jan-2010 10:51:53] <rmatte> oh
[28-Jan-2010 10:51:56] <rmatte> well then that'll work
[28-Jan-2010 10:52:17] <cgibbons> so what problems can I not solve for you today?
[28-Jan-2010 10:52:25] <ckrough> global desktop
[28-Jan-2010 10:52:28] <ckrough> make it so
[28-Jan-2010 10:52:39] <theacolyte> cgibbons: Tell me about WMI in enterprise vs the community plugin
[28-Jan-2010 10:52:55] <cgibbons> woo something i know about
[28-Jan-2010 10:52:58] <cgibbons> whatcha wanna know?
[28-Jan-2010 10:53:10] <theacolyte> Differences for one
[28-Jan-2010 10:53:36] <ckrough> is there a dev chat today?
[28-Jan-2010 10:53:55] <rmatte> is it a QA test day?
[28-Jan-2010 10:54:05] * rmatte points to the topic
[28-Jan-2010 10:54:07] <cgibbons> okay so in the Core product there is a low-level WMI connectivity package. the community WMI ZenPacks use this, as well as the Core zenwin (Service Monitoring) and zeneventlog (Event Log Monitoring) daemons.
[28-Jan-2010 10:54:44] <ckrough> rmatte: that says #zenoss-testing
[28-Jan-2010 10:54:46] <cgibbons> Core doesn't have modeling of a Windows device using WMI data, just SNMP. Enterprise adds that, as well as many device classes for Windows, Exchange, SQL Server, ActiveDirectory, etc.
[28-Jan-2010 10:55:02] <rmatte> ckrough: I'm aware of that, I idle in that channel all the time
[28-Jan-2010 10:55:11] <theacolyte> Does it allow for easier template building, or do yo ustill need to know what classes to monitor?
[28-Jan-2010 10:55:16] <theacolyte> Say, if I wanted to add stuff
[28-Jan-2010 10:55:22] <rmatte> ah, that was set on the 21st apparently
[28-Jan-2010 10:55:24] <cgibbons> Enterprise also extends that WMI connectivity package to do performance monitoring via PerfMon data sources so most of the Enterprise Windows templates use that.
[28-Jan-2010 10:55:31] * rmatte nudges mrayzenoss
[28-Jan-2010 10:56:04] <theacolyte> Does it also completely remove the need for snmp for windows machines?
[28-Jan-2010 10:56:07] <cgibbons> yep
[28-Jan-2010 10:56:12] <cgibbons> Except for 1 thing
[28-Jan-2010 10:56:27] <cgibbons> if you need it, process monitoring (but most? people do service monitoring instead of Windows)
[28-Jan-2010 10:56:50] <theacolyte> Not bad
[28-Jan-2010 10:57:17] <cgibbons> I personally find PerfMon metrics easier to find & add to templates because you can just run the PerfMon GUI tool on a Windows box and find all that's there... WMI classes usually require digging in the doc.
[28-Jan-2010 10:57:52] <cgibbons> The process monitoring stuff is an oversight when we converted that layer and it hasn't bubbled back up on the priority list yet.
[28-Jan-2010 10:58:09] <rmatte> cgibbons: nor has fixing zenprocess apparently, lol
[28-Jan-2010 10:58:24] <cgibbons> yeah, well... no one is sure what it is supposed to do either :) maybe Jane
[28-Jan-2010 10:58:27] <theacolyte> cgibbons: do you provide any interface for that in zenoss, or is it just copy and paste?
[28-Jan-2010 10:58:55] <cgibbons> just a text field, so if you know you wanna monitor the \\Processor(_Total)\% Processor Time metric you just type that in
[28-Jan-2010 10:59:08] <rmatte> cgibbons: it does what it's supposed to do, it just really needs some of the regex handling cleaned up/fixed as well as having it actually pull down configs properly without having to be restarted
[28-Jan-2010 10:59:24] <rmatte> cgibbons: any time I add or remove processes I have to restart zenprocess, otherwise it keeps working with the old stuff
[28-Jan-2010 10:59:36] <cgibbons> It was rewritten in 2.5, which you aren't running, so maybe that would be better.
[28-Jan-2010 10:59:49] <rmatte> I didn't know that it got a rewrite
[28-Jan-2010 10:59:58] <rmatte> I had only heard about some changes to zenperfsnmp and a rewrite of zenmib
[28-Jan-2010 11:00:13] <cgibbons> just to the new collector framework, but a few things were cleaned up... design wasn't changed so if the issues are design related it probably still has them.
[28-Jan-2010 11:00:31] <cgibbons> zenwin, zeneventlog, zenprocess and zenwinperf were all converted to the new collector framework in 2.5.0
[28-Jan-2010 11:00:34] <rmatte> well, I guess I'll see when 2.5.2 comes out
[28-Jan-2010 11:00:42] <theacolyte> hmmm
[28-Jan-2010 11:00:53] <theacolyte> How good are your WMI templates for say, exch2007?
[28-Jan-2010 11:00:59] <theacolyte> What classes are you lookng at?
[28-Jan-2010 11:01:13] <cgibbons> lemme look
[28-Jan-2010 11:03:18] <cgibbons> so it looks like it doesn't pull any modeling data for Exchange itself (it uses the base Windows Device template so there's modeling there), but has many perfmon datapoints... like \MSExchangeTransport Queues(_Total)\Active Mailbox Delivery Queue Length
[28-Jan-2010 11:03:26] <cgibbons> I don't know how good it is, not being an exchange admin
[28-Jan-2010 11:05:19] <theacolyte> Hmmm
[28-Jan-2010 11:15:29] <rmatte> bah, regexr told me this regex should work but it doesn't
[28-Jan-2010 11:16:01] <rmatte> what's the regex to select the first integer in a string?
[28-Jan-2010 11:16:09] <rmatte> someone mentioned it a few days ago
[28-Jan-2010 11:16:21] <Muti> (\d+)
[28-Jan-2010 11:16:29] <rmatte> thanks
[28-Jan-2010 11:19:52] <Egyptian[laptop]> evening all
[28-Jan-2010 11:20:23] <Egyptian[laptop]> i just added my first test linux box and i got this mesage Error reading value for "ssCpuIdle" on owner (oid .1.3.6.1.4.1.2021.11.11.0 is bad) - is there a problem with my snmp mib?
[28-Jan-2010 11:22:32] <rmatte> Egyptian[laptop]: your snmp is not configured correctly on the server
[28-Jan-2010 11:23:31] <rmatte> Egyptian[laptop]: it's an snmp permissions issue, it has nothing to do with Mibs.
[28-Jan-2010 11:23:57] <rmatte> you haven't given enough permissions to the community string which you are using to monitor with
[28-Jan-2010 11:25:05] <rmatte> :)
[28-Jan-2010 11:25:10] <rmatte> it is indeed
[28-Jan-2010 11:25:11] <forsberg> :
[28-Jan-2010 11:25:44] <Egyptian[laptop]> rmatte: ah ok - this is my first try .. guess i will play around with snmpconf then
[28-Jan-2010 11:26:06] <rmatte> Egyptian[laptop]: one second, I'll pastebin a skeleton config for you
[28-Jan-2010 11:26:19] <forsberg> rocommunity public ip
[28-Jan-2010 11:26:20] <forsberg> :)
[28-Jan-2010 11:26:40] <Muti> Egyptian[laptop]: you can test whether SNMP is working on your linux box by running: snmpwalk -v 2c -c <community> <hostname>

[28-Jan-2010 11:27:19] <Muti> so anyone here make their own ZenPacks?
[28-Jan-2010 11:27:30] <rmatte> Egyptian[laptop]: http://pastebin.com/m5ae4d6ca
[28-Jan-2010 11:27:50] <forsberg> i did when moving templates from old broken install to new, but not published any
[28-Jan-2010 11:27:51] <Muti> I'm trying to decide whether I should add the device class to my ZenPack, since uninstalling the ZenPack deletes all the devices under said device class
[28-Jan-2010 11:28:03] <forsberg> yeah i didnt do that, for that reason
[28-Jan-2010 11:28:21] <rmatte> Egyptian[laptop]: replace anything that says "your-key" also replace the contact and location strings, then you're ready to go
[28-Jan-2010 11:28:22] <forsberg> rahter just have the template in the root, then bind it to whatever
[28-Jan-2010 11:28:34] <Muti> okay, yeah.. that's what I was leaning toward doing
[28-Jan-2010 11:28:43] <rmatte> Muti: that's bad advice, snmpwalk can only return partial output based on the config
[28-Jan-2010 11:28:53] <rmatte> Muti: it doesn't mean you'll be able to monitor it properly
[28-Jan-2010 11:30:25] <forsberg> i only used one line in snmpd.conf besides the description and random pass stuff, and thats rocommunity <community> <zenossip>

[28-Jan-2010 11:30:26] <forsberg> :>
[28-Jan-2010 11:31:03] <rmatte> forsberg that's a bit too simple for my taste
[28-Jan-2010 11:31:47] <forsberg> yeah its abit slack i guess, but all our stuff is behind firewalls etc, so...
[28-Jan-2010 11:32:15] <rmatte> also, where's your trap2sink config?
[28-Jan-2010 11:32:40] <rmatte> you won't get traps unless you've actually configured them
[28-Jan-2010 11:33:14] <forsberg> trapsink <zenossip> <community>

[28-Jan-2010 11:33:17] <forsberg> sorry i was lying
[28-Jan-2010 11:33:25] <rmatte> lol
[28-Jan-2010 11:33:38] <rmatte> :P
[28-Jan-2010 11:33:46] <Egyptian[laptop]> rmatte: thanks
[28-Jan-2010 11:33:49] <forsberg> and sysloc and sysdescr
[28-Jan-2010 11:33:59] <Egyptian[laptop]> how do i get zenoss to redo the device again?
[28-Jan-2010 11:34:34] <Egyptian[laptop]> other than deleting it ;)
[28-Jan-2010 11:34:59] <forsberg> the top dropdown - manage - model device
[28-Jan-2010 11:35:57] <Egyptian[laptop]> forsberg: heh the only one i dint try
[28-Jan-2010 11:36:02] <forsberg> :P
[28-Jan-2010 11:36:37] <Egyptian[laptop]> yay!! !it picked up my virbr interface!
[28-Jan-2010 11:36:46] <forsberg> gj \o/
[28-Jan-2010 11:59:51] <Apachez> http://i.imgur.com/MnEDT.jpg
[28-Jan-2010 12:00:51] <Apachez> http://www.missilebases.com/adironback
[28-Jan-2010 12:24:40] <rmatte> time to clean up swap space on my workstation
[28-Jan-2010 12:29:22] <rmatte> amazing how fast my workstation runs when the swap space isn't being used
[28-Jan-2010 12:29:26] <rmatte> :P
[28-Jan-2010 12:32:36] <forsberg> :)
[28-Jan-2010 12:32:41] <forsberg> runnign win7 on ssd
[28-Jan-2010 12:33:10] <forsberg> and not swapping
[28-Jan-2010 12:33:27] <a21> rmatte: what OS do you use?
[28-Jan-2010 12:33:35] <rmatte> a21: Ubuntu
[28-Jan-2010 12:33:56] <rmatte> a21: With a vista VM running in virtualbox, which I will be upgrading to windows 7 soon
[28-Jan-2010 12:34:42] <forsberg> thats hardcore man
[28-Jan-2010 12:34:43] <forsberg> vista!
[28-Jan-2010 12:34:59] <a21> rmatte: same thing with swap. one day I just disbled it at all and now feel much better
[28-Jan-2010 12:35:01] <rmatte> rmatte@noc:~$ uptime
[28-Jan-2010 12:35:01] <rmatte> 13:42:30 up 86 days, 2:41, 3 users, load average: 1.61, 1.04, 0.91
[28-Jan-2010 12:35:05] <rmatte> uptime on my workstation
[28-Jan-2010 12:35:06] <rmatte> ;)
[28-Jan-2010 12:35:36] <forsberg> at work i run osX with xp in vm, but osx wants to reboot everytime it updates itunes and what not
[28-Jan-2010 12:35:38] <forsberg> :(
[28-Jan-2010 12:35:43] <a21> rmatte: you need some specifis software?
[28-Jan-2010 12:35:57] <rmatte> a21: in windows you mean?
[28-Jan-2010 12:36:04] <a21> rmatte: yeah
[28-Jan-2010 12:36:13] <rmatte> yeh, vSphere client for ESX
[28-Jan-2010 12:36:25] <rmatte> and I use Outlook too
[28-Jan-2010 12:36:32] <forsberg> same :)
[28-Jan-2010 12:36:47] <rmatte> those are pretty much all I ever have running on that VM
[28-Jan-2010 12:36:53] <rmatte> I do end up using it for the odd thing other times
[28-Jan-2010 12:37:08] <rmatte> we also have some printers in the building which don't have Linux drivers available
[28-Jan-2010 12:37:12] <rmatte> so I use the VM to print sometimes
[28-Jan-2010 12:37:21] <forsberg> i must confess i love putty too tho, compared to osx terminal etc
[28-Jan-2010 12:37:31] <rmatte> SecureCRT owns all
[28-Jan-2010 12:37:37] <a21> rmatte: I see
[28-Jan-2010 12:37:38] <rmatte> destroys putty and anything else
[28-Jan-2010 12:37:46] <theacolyte> woarkstation > virtualbox
[28-Jan-2010 12:37:54] <theacolyte> workstation, rather
[28-Jan-2010 12:38:00] <theacolyte> securecrt is great, but not free
[28-Jan-2010 12:38:01] <rmatte> theacolyte: virtualbox works great
[28-Jan-2010 12:38:15] <rmatte> theacolyte: yeh, I have a licensed copy of SecureCRT
[28-Jan-2010 12:38:36] <rmatte> well worth it in my opinion
[28-Jan-2010 12:38:41] <rmatte> and it runs quite well in wine
[28-Jan-2010 12:39:04] <rmatte> (with some little quirks here or there)
[28-Jan-2010 12:40:20] <theacolyte> I can't use linux
[28-Jan-2010 12:40:25] <theacolyte> Well I take that back
[28-Jan-2010 12:40:27] <rmatte> why?
[28-Jan-2010 12:40:30] <theacolyte> I can't use it exclusively
[28-Jan-2010 12:40:36] <theacolyte> I do prefer it as a workstation
[28-Jan-2010 12:40:40] <theacolyte> In a word: Outlook
[28-Jan-2010 12:40:53] <rmatte> I could if it weren't for vmware being idiots and only making the vSphere client for windows
[28-Jan-2010 12:41:09] <rmatte> Outlook is nice, but I could do fine without it
[28-Jan-2010 12:41:10] <theacolyte> And actually I got this new laptop and was going to put ubuntu on it, but it really doesn't like my displayport
[28-Jan-2010 12:41:19] <theacolyte> I couldn't live without Outlook
[28-Jan-2010 12:41:20] <rmatte> I'd just use thunderbird or something
[28-Jan-2010 12:41:31] <theacolyte> And re: viclient, yes, it's retarded, viclient is retarded - we all hate it
[28-Jan-2010 12:41:40] <rmatte> yeh
[28-Jan-2010 12:41:55] <ckrough> when I was on windows I starting using this instead of securecrt http://www.netsarang.com/
[28-Jan-2010 12:41:58] <theacolyte> actually, not running in linux is the least of it's problems
[28-Jan-2010 12:42:00] <rmatte> I'm running a whole Vista VM to run 2 applications, it's pretty ridiculous
[28-Jan-2010 12:42:02] <ckrough> ssh client + it's a windows x server
[28-Jan-2010 12:42:40] <theacolyte> I need something tabbed like securecrt, but free
[28-Jan-2010 12:42:47] <rmatte> I wonder how good that Xmanager is
[28-Jan-2010 12:42:50] <theacolyte> boss won't buy securecrt -0 it's too damn expensive when putty exists
[28-Jan-2010 12:43:12] <rmatte> theacolyte: SecureCRT is cheap as dirt, tell him to stop pocket the nickles
[28-Jan-2010 12:43:21] <rmatte> stop pocketing*
[28-Jan-2010 12:43:36] <theacolyte> lol
[28-Jan-2010 12:43:41] <theacolyte> He did just approve a new desktop for me
[28-Jan-2010 12:43:42] <forsberg> whats wrong with putty, i mean whats better with securecrt or netsarang or whatnot?
[28-Jan-2010 12:43:48] <theacolyte> So I'm going to buy a stupidly fast desktop
[28-Jan-2010 12:43:50] <theacolyte> forsberg: Tabs, for one
[28-Jan-2010 12:43:52] <theacolyte> lack of automation
[28-Jan-2010 12:44:03] <ckrough> rmatte: try it out. I loved it. well written
[28-Jan-2010 12:44:21] <ckrough> <-- urxvt
[28-Jan-2010 12:44:26] <theacolyte> There's a way to tab putty, but I couldn't ever get it to work
[28-Jan-2010 12:44:34] <rmatte> forsberg: SecureCRT has better session options, supports macros, supports tabs, supports logging output
[28-Jan-2010 12:44:43] <theacolyte> well, putty does logging, but yeah
[28-Jan-2010 12:44:57] <forsberg> ok tabs i wont use, but others i could see i could use some time
[28-Jan-2010 12:45:02] <ckrough> there's always cygwin with kterm or gnome-terminal as well
[28-Jan-2010 12:45:07] <ckrough> thatrs all free
[28-Jan-2010 12:45:11] <theacolyte> you don't manage more than 1 device at a time?
[28-Jan-2010 12:45:16] <rmatte> wow man, I manage 13 Zenoss servers, with more coming, tabs are an absolute must for me
[28-Jan-2010 12:45:19] <forsberg> i manage alot at one time
[28-Jan-2010 12:45:21] <theacolyte> same
[28-Jan-2010 12:45:27] <forsberg> but i just open a new one etc
[28-Jan-2010 12:45:27] <ckrough> same
[28-Jan-2010 12:45:29] <rmatte> If I were only managing one or two I could get away with just commandline ssh
[28-Jan-2010 12:45:34] <ckrough> but I use clusterssh instead of tabs
[28-Jan-2010 12:45:37] <forsberg> close group :)
[28-Jan-2010 12:45:42] <ckrough> there is also dssh
[28-Jan-2010 12:45:43] <theacolyte> clusterssh hehe
[28-Jan-2010 12:46:10] <forsberg> but i wuoldnt benefit from tabs in same way as a browser
[28-Jan-2010 12:46:17] <forsberg> terminal in osx has tabs, hate it
[28-Jan-2010 12:46:39] <rmatte> clusterssh looks pretty cool actually
[28-Jan-2010 12:46:44] <rmatte> I'd never seen that before
[28-Jan-2010 12:46:44] <forsberg> but ill try securecrt right away :P
[28-Jan-2010 12:46:53] <venturaville> I use gnome terminal which has tabs
[28-Jan-2010 12:47:04] <rmatte> forsberg: the tabs in securecrt are obviously optional
[28-Jan-2010 12:47:06] <venturaville> I have all my key sequences setup to flip back and forth between tabs
[28-Jan-2010 12:47:11] <venturaville> I have 9 open at the moment
[28-Jan-2010 12:47:12] <rmatte> forsberg: I just hate a cluttered taskbar
[28-Jan-2010 12:47:29] <ckrough> problem with gnome-terminal is that you cant tell it to send all input to all tabs
[28-Jan-2010 12:47:34] <ckrough> I think kterm does that
[28-Jan-2010 12:47:34] <theacolyte> less mouse movement > key combos to switch tabs
[28-Jan-2010 12:47:35] <theacolyte> heh
[28-Jan-2010 12:47:45] <theacolyte> ah someone said that already
[28-Jan-2010 12:47:45] <forsberg> im on screen from jumphost, thats my tabs ;>

[28-Jan-2010 12:47:49] <venturaville> no but I can paste with my mouse on all 9 tabs in about 5 seconds
[28-Jan-2010 12:48:02] <rmatte> lol
[28-Jan-2010 12:48:10] <forsberg> but but nono i hear you, ill try it right away
[28-Jan-2010 12:48:12] <ckrough> whats a mouse
[28-Jan-2010 12:48:14] <ckrough> :)
[28-Jan-2010 12:48:21] <ckrough> j/k
[28-Jan-2010 12:48:27] <theacolyte> I'll take ridiculous for $2000, Alex
[28-Jan-2010 12:49:21] <rmatte> I don't think securecrt supports sending input to all tabs either
[28-Jan-2010 12:49:48] <ckrough> the only ones Ive seen that do are cssh and kterm
[28-Jan-2010 12:49:56] <ckrough> and possibly secpanel
[28-Jan-2010 12:50:14] <rmatte> I like securecrt's clone session option
[28-Jan-2010 12:50:28] <ckrough> the reason I end up sticking with cssh is that I can edit conf files on all boxes at once
[28-Jan-2010 12:50:53] <rmatte> ckrough: yeh, that would be sic actually
[28-Jan-2010 12:51:38] <forsberg> make a bit fat macro and do the job i have with updating 20 voip servers at one go, gonna try! ;)
[28-Jan-2010 12:51:44] <ckrough> and you can define clusters in ~/.clusterssh, $cssh myclustertag
[28-Jan-2010 12:52:17] <ckrough> forsberg: http://rackerhacker.com/2010/01/20/crash-course-in-dsh/
[28-Jan-2010 12:53:00] <cgibbons> hurm
[28-Jan-2010 12:53:08] <cgibbons> anyone good with linux router setup?
[28-Jan-2010 12:53:26] <forsberg> thanks ;> i was only kidding, not big balls enough to try all in one go, gonna do one at a time, and make sure it went a ok :P
[28-Jan-2010 12:53:34] <rmatte> cgibbons: well, in what aspect?
[28-Jan-2010 12:53:42] <Egyptian[laptop]> havent tried it yet .. i will have to soon but openchange is a connector between evolution(?) and exchange .. maybe it might help some here
[28-Jan-2010 12:54:16] <cgibbons> doing an open-vpn setup.... the link is active & happy, but the box isn't forwarding over the tunnel, even though ip_forward is enabled and iptables is configured. usually that just 'works'
[28-Jan-2010 12:54:44] <rmatte> cgibbons: route issue?
[28-Jan-2010 12:54:58] <rmatte> like, you're sure the iptables config is good?
[28-Jan-2010 12:55:28] <cgibbons> it's wide-open, accept on input, output, forward chains
[28-Jan-2010 12:55:48] <rmatte> how about your actual linux routes?
[28-Jan-2010 12:56:20] <rmatte> (type "route" in case you didn't know)
[28-Jan-2010 12:56:24] <rmatte> :P
[28-Jan-2010 12:56:47] <forsberg> netstat -r you mean ?
[28-Jan-2010 12:56:48] <forsberg> :>

[28-Jan-2010 12:56:48] <cgibbons> route on vpn server is right.... (10.175.211.0/24 going over the tun0 gateway device), on the LAN 10.175.211.0 router is the IP of the openvpn box, and it's getting to that gateway when i do a traceroute from my client test
[28-Jan-2010 12:57:01] <rmatte> forsberg: my way is easier :P
[28-Jan-2010 12:57:12] <forsberg> -bash: route: command not found
[28-Jan-2010 12:57:50] <forsberg> [foppa@rotten ~]$ route
[28-Jan-2010 12:57:50] <forsberg> usage: route [-dnqtv] command [[modifiers] args]
[28-Jan-2010 12:57:52] <forsberg> yrgh
[28-Jan-2010 12:58:20] <forsberg> centos and freebsd doesnt work too well hehe
[28-Jan-2010 12:58:29] <cgibbons> headscratcher hrm
[28-Jan-2010 12:58:49] <rmatte> forsberg: echo "#!/bin/bash" > /sbin/route && echo "netstat -r" >> /sbin/route && chmod +x /sbin/route
[28-Jan-2010 12:58:51] <rmatte> :P
[28-Jan-2010 12:58:54] <ckrough> cgibbons: iptables rules allowing traffic to tun0?
[28-Jan-2010 12:59:21] <ckrough> cgibbons: in addition to outbound to the otherside
[28-Jan-2010 12:59:36] <rmatte> ckrough: yeh, that's a good question
[28-Jan-2010 12:59:37] <cgibbons> yeah it's wide open, all 3 chains say "ACCEPT all -- anywhere anywhere"
[28-Jan-2010 12:59:42] <rmatte> ah
[28-Jan-2010 12:59:47] <theacolyte> Excellent
[28-Jan-2010 12:59:49] <theacolyte> What's your IP
[28-Jan-2010 12:59:58] <rmatte> :P
[28-Jan-2010 13:00:02] <rmatte> try bouncing the tunnel?
[28-Jan-2010 13:00:49] <cgibbons> yep no dice
[28-Jan-2010 13:01:00] <rmatte> there's no firewall on the other end?
[28-Jan-2010 13:01:42] <cgibbons> might be but wouldn't that stop me from being able to connect from the vpn router, too?
[28-Jan-2010 13:01:53] <rmatte> not necessarily
[28-Jan-2010 13:01:57] <cgibbons> or, does my internal LAN IP addresses cause confuse in this case i wonder? hmmm
[28-Jan-2010 13:02:30] <rmatte> It's worth checking if you can
[28-Jan-2010 13:03:39] <ckrough> cgibbons: also, if you are natting out your final router, that could confuse things depending on where your vpn terminates
[28-Jan-2010 13:05:05] <cgibbons> my co-worker with the same setup had to change out is 192.168.x.x. to a 10, so that's probably my problem too, lemme try that out
[28-Jan-2010 13:14:31] <cgibbons> shoot no change
[28-Jan-2010 13:19:23] <rmatte> I still vote firewall on other end
[28-Jan-2010 13:21:11] <Egyptian[laptop]> how do i make the default network ip range in network map something else?
[28-Jan-2010 13:21:25] <Egyptian[laptop]> i do not have any 192.168.1.0 networks!
[28-Jan-2010 13:21:49] <rmatte> Egyptian[laptop]: in settings
[28-Jan-2010 13:22:25] <rmatte> Settings -> Users -> <your user>

[28-Jan-2010 13:22:36] <rmatte> "Network Map Start Object"
[28-Jan-2010 13:27:59] <Egyptian[laptop]> what is EC2Manager ?
[28-Jan-2010 13:28:57] <Egyptian[laptop]> Network Map Start Object ?
[28-Jan-2010 13:31:39] <Egyptian[laptop]> and yes .. i change the default vlan for the flash thing ..thanks .. zenoss is fun :)
[28-Jan-2010 13:34:57] <theacolyte> mrayzenoss: pretty big usse:
[28-Jan-2010 13:35:01] <theacolyte> http://community.zenoss.org/index.jspa
[28-Jan-2010 13:35:08] <theacolyte> Jan 28, 2010
[28-Jan-2010 13:35:09] <theacolyte> Zenoss Core 2.1.91 (Zenoss 2.2 Beta 2) Available for Download
[28-Jan-2010 13:35:22] <theacolyte> all the news shows todays date
[28-Jan-2010 13:48:32] <dec3pti0n> I'm trying to perform my first zenoss backup and it's giving me an error, first I thought it was a gui thing but the same happens on cli
[28-Jan-2010 13:49:05] <ckrough> dec3pti0n: whats the error
[28-Jan-2010 13:49:09] <dec3pti0n> http://pastebin.ca/1769727
[28-Jan-2010 13:49:40] <dec3pti0n> is it failing just because that directory does not exist ?
[28-Jan-2010 13:49:51] <ckrough> or it cant read/write it
[28-Jan-2010 13:50:35] <dec3pti0n> well it doesn't exist ... I guess it should have been created during install ! cause I have not tried to install any zenpacks yet
[28-Jan-2010 13:50:59] <rmatte> Egyptian[laptop]: EC2Manager is to monitor amazon cloud
[28-Jan-2010 13:51:26] <rmatte> Egyptian[laptop]: did you not find the network map start object? I gave instruction on how to get to it.
[28-Jan-2010 13:51:42] <rmatte> Egyptian[laptop]: you don't set a "Range" you set a single IP to start at by default
[28-Jan-2010 13:51:43] <dec3pti0n> yeah it was failing just because of that directory ... all good now
[28-Jan-2010 13:59:43] <dec3pti0n> I stop ftp on the box I have to test out monitoring alert rules. I got the status changed for the IP Services but the OS processes that I also have monitoring the vsftpd process has not changed !
[28-Jan-2010 14:00:41] <ckrough> dec3pti0n: 1) if you've updated the process monitoring recently you may need to restart zenprocess to get it to update its config
[28-Jan-2010 14:00:55] <ckrough> dec3pti0n: b) are your process and service cycles the same
[28-Jan-2010 14:01:06] <ckrough> dec3pti0n: it may have just not happened yet
[28-Jan-2010 14:01:10] <dec3pti0n> ckrough: that's what I'm checking now
[28-Jan-2010 14:03:08] <Egyptian[laptop]> rmatte: yes i found it and set it up correctly .. its a beautiful map
[28-Jan-2010 14:03:23] <Egyptian[laptop]> thanks
[28-Jan-2010 14:04:30] <rmatte> np
[28-Jan-2010 14:05:09] <rmatte> ckrough: funny how you switched from numbers to letters when enumerating
[28-Jan-2010 14:05:18] <rmatte> 1) ...
[28-Jan-2010 14:05:18] <ckrough> :)
[28-Jan-2010 14:05:22] <dec3pti0n> ckrough: I can't seem to find where that is located, I thought it would be under zProperties for services and Processes, no ?
[28-Jan-2010 14:05:22] <rmatte> b) ...
[28-Jan-2010 14:05:43] <ckrough> hmm... probably under collectors > edit
[28-Jan-2010 14:05:49] <ckrough> or somewhere around there
[28-Jan-2010 14:05:54] <rmatte> dec3pti0n: yeh, collectors -> localhost -> edit
[28-Jan-2010 14:09:06] <dec3pti0n> ok found the Process Cycle Interval which is 2min, the only Service one is "windows service Cycle Interval" but no windows here :) .. probably SNMP perf cycle interval which is 5min
[28-Jan-2010 14:10:23] * dec3pti0n going back to the manual to go over OS processes
[28-Jan-2010 14:10:24] <rmatte> I have my process cycle at 300 I think
[28-Jan-2010 14:10:47] <rmatte> or no, 360
[28-Jan-2010 14:10:50] <rmatte> which is 6 minutes
[28-Jan-2010 14:13:34] <dec3pti0n> gonna try to restart the zenprocess and try the test again
[28-Jan-2010 14:14:06] <ckrough> dec3pti0n: delete any .zec files too if it seems to be ignoring your attempts
[28-Jan-2010 14:14:12] <cgibbons> ouf
[28-Jan-2010 14:14:17] <ckrough> no?
[28-Jan-2010 14:18:56] <ckrough> zendmd is The Way (tm)
[28-Jan-2010 14:28:39] <dec3pti0n> hehe i got the proc alert msg now after restarting the zenprocess
[28-Jan-2010 14:31:18] <dec3pti0n> cool now I got both thanks
[28-Jan-2010 14:48:19] <ckrough> just lowered zserver-threads to 5 in development and the gui got even faster
[28-Jan-2010 14:48:29] <ckrough> I think the zenoss default install uses 10 or 20
[28-Jan-2010 14:48:35] <ckrough> may want to reconsider that setting...
[28-Jan-2010 14:49:24] <ckrough> can someone with a virgin zope.conf check the following settings: zserver-threads, cache-size, and pool-size?
[28-Jan-2010 14:49:35] <ckrough> I may create a ticket to have those defaults re-evaluated
[28-Jan-2010 14:50:39] <ckrough> nm, the default is 4
[28-Jan-2010 14:51:07] <forsberg> where do i find this server threads :>

[28-Jan-2010 14:51:18] <ckrough> $ZENHOME/etc/zope.conf
[28-Jan-2010 14:57:13] <eidolon> ckrough: we're having serious lag in our server during heavy syncing... wonder if that would be a good change.
[28-Jan-2010 14:57:55] <rmatte> ckrough: put it at 4, it's working amazingly for me and it's the default anyways
[28-Jan-2010 14:58:55] <rmatte> the default is to have that line commented out, which makes Zope default to 4
[28-Jan-2010 14:59:50] <rmatte> ah, didn't see that you already figured that out
[28-Jan-2010 15:01:04] <ckrough> yeah.
[28-Jan-2010 15:14:15] <dec3pti0n> anyone here uses the apache monitor on 2.5.1 ?
[28-Jan-2010 16:18:32] <dec3pti0n> what's the easiest way to install a python module which is a zenpack dependency ?
[28-Jan-2010 16:18:46] <rmatte> which module?
[28-Jan-2010 16:19:23] <dec3pti0n> python-pgsql .... I was gonna download the source and try the command I found on the README.txt under the zenoss python directory
[28-Jan-2010 16:19:41] <rmatte> you could try: easy_install python-pgsql
[28-Jan-2010 16:19:44] <rmatte> as the zenoss user
[28-Jan-2010 16:19:48] <rmatte> it may or may not work
[28-Jan-2010 16:27:21] <dec3pti0n> cool thanks, i'll try that later
[28-Jan-2010 16:27:38] <rmatte> np
[28-Jan-2010 16:58:43] forsberg is now known as fOrsberg
[28-Jan-2010 16:58:47] fOrsberg is now known as forsberg
[28-Jan-2010 17:06:09] forsberg is now known as fOrsberg
[28-Jan-2010 17:17:50] <frozty_sa> uhm
[28-Jan-2010 17:17:56] <rmatte> uhm
[28-Jan-2010 17:18:00] <frozty_sa> so let's suppose I've got a new zenoss install
[28-Jan-2010 17:18:07] <rmatte> ok, let's suppose that
[28-Jan-2010 17:18:15] <rmatte> :)
[28-Jan-2010 17:18:29] <frozty_sa> and let's suppose it managed to clock up about 120000 items by way of some fun hosts sending STP logs towards it
[28-Jan-2010 17:18:39] <rmatte> k
[28-Jan-2010 17:19:09] <frozty_sa> how do we quickly get rid of said 120000 theoretical eventlog items without theoretically clicking in the eventlog in ~10k batches? :D
[28-Jan-2010 17:19:43] <mrayzenoss> frozty_sa: go straight to the db
[28-Jan-2010 17:19:49] <rmatte> well, do you just want them moved from history or completely purged for the database?
[28-Jan-2010 17:19:52] <rmatte> from*
[28-Jan-2010 17:20:01] <rmatte> moved to history rather
[28-Jan-2010 17:20:11] <frozty_sa> I'm "closing" them, so I think that's moved to history?
[28-Jan-2010 17:20:25] <rmatte> yeh, but are you really going to want all of those sitting around in the DB?
[28-Jan-2010 17:20:32] <mrayzenoss> yeah, they're in your history now, so searching history will be painful (until 2.5.2)
[28-Jan-2010 17:20:38] <frozty_sa> ideally for now, yes
[28-Jan-2010 17:20:53] <frozty_sa> I'm lucky enough to be able to use the space-is-cheap excuse here
[28-Jan-2010 17:20:55] <rmatte> frozty_sa: ok, I can whip something up for you, give me a couple minutes
[28-Jan-2010 17:21:05] <frozty_sa> rmatte: ta, that'd be appreciated
[28-Jan-2010 17:21:15] <frozty_sa> would also start giving me an insight into how the db is structured
[28-Jan-2010 17:21:35] <rmatte> http://pastebin.com/m6675a6b6
[28-Jan-2010 17:21:38] <rmatte> use that in zendmd
[28-Jan-2010 17:21:42] <rmatte> that'll clear the active events
[28-Jan-2010 17:21:53] <frozty_sa> although, coming from a bunch of zabbix installs (which was the previous monitoring framework I had to oversee in this usecase), zenoss is <3
[28-Jan-2010 17:21:53] <rmatte> (I wrote that a while back for someone else who had the same issue)
[28-Jan-2010 17:22:23] <rmatte> when it says "Deleting event" it actually means moving to history
[28-Jan-2010 17:22:31] <rmatte> but yeh, that'll work, just let it run until it's done
[28-Jan-2010 17:22:39] <rmatte> I'm out, cheers
[28-Jan-2010 17:22:44] * Muti just discovered that 2.5 million of the events in our history table are /Change/Set events of set attribute snmpindex on httpd processes
[28-Jan-2010 17:24:47] <Muti> that's 65% of our history table right there.. heh
[28-Jan-2010 17:24:59] <mrayzenoss> might want to tweak the rules to drop them
[28-Jan-2010 17:25:12] <Muti> yeah, time for a transform on /Change/Set
[28-Jan-2010 17:25:57] <frozty_sa> rmatte: ta :)
[28-Jan-2010 17:26:33] <frozty_sa> and now that I know where to find the CLI interface, I can automate a few of these things quite nicely
[28-Jan-2010 17:31:03] <Muti> P.S. Totally dig the new community site CSS
[28-Jan-2010 17:31:50] <mrayzenoss> glad to hear it
[28-Jan-2010 18:05:25] <theacolyte> man this is so weird
[28-Jan-2010 18:05:34] <theacolyte> 2 of my ESX servers don't have thier interfaces displayed
[28-Jan-2010 18:05:37] <theacolyte> 1 does
[28-Jan-2010 18:28:49] <theacolyte> interestingly that same server doesn't work with HP SIM that I'm putting in
[28-Jan-2010 18:28:49] <theacolyte> lol
[28-Jan-2010 18:50:33] <theacolyte> nm that last correlation
[28-Jan-2010 21:01:30] <theacolyte> wow
[28-Jan-2010 21:01:39] <theacolyte> I haven't seen GNAA in a really long time
[28-Jan-2010 21:01:53] <theacolyte> Next time that happens, PM me, I'll see to it they are removed
[29-Jan-2010 00:00:30] [disconnected at Fri Jan 29 00:00:30 2010]
[29-Jan-2010 00:00:30] [connected at Fri Jan 29 00:00:30 2010]
[29-Jan-2010 02:29:30] <frozty_sa> what style regexes are the EventClass mappings, pcre?
[29-Jan-2010 02:31:16] <frozty_sa> I want to do 'cpu (wait|system) usage of [0-9]*\.[0-9]?% matches resource limit \[cpu (wait|system) usage>[0-9]*\.[0-9]?%\]' so I can classify monit CPU alerts (from syslog), but they don't seem to match
[29-Jan-2010 02:31:38] <frozty_sa> if I take tone it down to either wait or system, then it's fine
[29-Jan-2010 02:32:45] <frozty_sa> s/take//
[29-Jan-2010 03:00:46] <oivind> could anyone please tell me the Data Point alias syntax to divide the given value by 10?
[29-Jan-2010 04:26:23] <oivind> what are possible reasons when a graph does not receive data? it only says cur:0.0 avg:0.0 max:nan? the datasource is working when it is tested, and no rrd max is set, and included in a graph point like i've done with other devices.
[29-Jan-2010 04:28:32] <oivind> everything looks as it should with legends in the graph, just no data :/
[29-Jan-2010 04:38:27] <tehhobbit> may I suggest channelmode +[rR]
[29-Jan-2010 09:44:57] <rmatte> does anyone have a procurve that they aren't using and would like to mail to me? :P
[29-Jan-2010 09:49:50] <rmatte> aha, think I found a solution
[29-Jan-2010 09:55:29] <rocket> rmatte: nope .. not gonna happen :p
[29-Jan-2010 09:56:14] <rmatte> hehe
[29-Jan-2010 09:58:54] <rocket> whats the solution?
[29-Jan-2010 09:59:48] <cgibbons> demo time!
[29-Jan-2010 10:00:09] <rocket> cgibbons: mc hammer flash back! :p
[29-Jan-2010 10:00:30] <rocket> Stop! demo time ... :p
[29-Jan-2010 10:02:56] <rmatte> lol
[29-Jan-2010 10:13:59] <theacolyte> rmatte: So that ESX network interface thing
[29-Jan-2010 10:14:07] <theacolyte> 3 hosts, 1 works, same snmp configuration for all 3
[29-Jan-2010 10:14:18] <theacolyte> the other ones don't read network or disk
[29-Jan-2010 10:19:31] fOrsberg is now known as forsberg
[29-Jan-2010 12:00:26] <forsberg> ehhh
[29-Jan-2010 12:00:33] <rmatte> nice to see that the opers are so quick to k-line idiots
[29-Jan-2010 12:00:44] <aclark> i just pinged a guy in #plone about it
[29-Jan-2010 12:00:54] <aclark> annoying
[29-Jan-2010 12:01:00] <rmatte> yeh
[29-Jan-2010 12:01:12] <rmatte> my client auto-ignores ctcp floods luckily
[29-Jan-2010 12:01:29] <forsberg> ahx
[29-Jan-2010 12:01:29] <aclark> hmmm i should set that
[29-Jan-2010 12:01:30] <forsberg> hax
[29-Jan-2010 12:01:36] <forsberg> 84.237.210.63
[29-Jan-2010 12:01:37] <forsberg> nuke go
[29-Jan-2010 12:01:38] <forsberg> :P
[29-Jan-2010 12:02:22] <rmatte> lol
[29-Jan-2010 12:02:31] <forsberg> :)
[29-Jan-2010 12:02:33] <forsberg> så døde han
[29-Jan-2010 12:02:40] <forsberg> ups wrong language, then he died
[29-Jan-2010 12:02:40] <forsberg> :)
[29-Jan-2010 12:04:08] <rmatte> well, 84.237.210.63 is a Mac
[29-Jan-2010 12:04:44] <rmatte> [+] Host 84.237.210.63 Running OS: "Apple Mac OS X 10.4.1" (Guess probability: 100%)
[29-Jan-2010 12:05:39] <forsberg> :P
[29-Jan-2010 12:06:16] <rmatte> seriously
[29-Jan-2010 12:06:17] <rmatte> lol
[29-Jan-2010 12:06:32] <rmatte> seems like there are no open services running on it though
[29-Jan-2010 12:06:35] <rmatte> double-checking
[29-Jan-2010 12:09:51] <rmatte> All 1000 scanned ports on 84.237.210.63 are filtered
[29-Jan-2010 12:10:37] <rmatte> too bad, coulda had some fun
[29-Jan-2010 12:23:14] <theacolyte> hey guys
[29-Jan-2010 12:23:17] <theacolyte> when that happens, PM me
[29-Jan-2010 12:25:54] <mistich> hello all
[29-Jan-2010 12:26:14] <mistich> anyone done any alerting on firewall failovers
[29-Jan-2010 12:26:23] <theacolyte> oh
[29-Jan-2010 12:26:27] <theacolyte> hey kloeri, didn't see you joined
[29-Jan-2010 12:28:50] <kloeri> hiya theacolyte
[29-Jan-2010 12:29:08] <theacolyte> For some reason this channel got targetted
[29-Jan-2010 12:29:09] <theacolyte> Weird
[29-Jan-2010 12:29:20] <kloeri> lots of random channels are being targeted
[29-Jan-2010 12:29:21] <theacolyte> Maybe it was just last on the /list and they were like hey, let's do 20 up from the b ottom!
[29-Jan-2010 12:29:24] <theacolyte> hehe
[29-Jan-2010 12:30:06] <kloeri> makes it impossible to follow all the channels being attacked but it's easier picking out the bots without joining tons of channels anyway :p
[29-Jan-2010 12:30:15] <theacolyte> true
[29-Jan-2010 13:04:28] forsberg is now known as fOrsberg
[29-Jan-2010 13:31:17] <Apachez> http://www.youtube.com/watch?v=lQnT0zp8Ya4
[29-Jan-2010 14:17:56] <rmatte> grrr, thought I finally had this ziptie adapter fixed but apparently not
[29-Jan-2010 14:20:56] <rocket> kloeri: its all your fault :)
[29-Jan-2010 14:24:49] <kloeri> rocket: I'll fully admit to that :)
[29-Jan-2010 14:24:59] <rocket> ;)
[29-Jan-2010 15:15:26] <mistich> rmatte what version of zenoss does using getRRDValue work
[29-Jan-2010 15:16:08] <mistich> in a threshold
[29-Jan-2010 15:16:18] <mistich> I cannot get this to work as max (here.getRRDValue('mem5minUsed_mem5minUsed')+here.getRRDValue('mem5minFree_mem5minFree'))*0.8
[29-Jan-2010 15:17:44] <venturaville> I've used it since at least as far back as 2.4.....
[29-Jan-2010 15:17:55] <mistich> how about 2.3
[29-Jan-2010 15:18:09] <venturaville> I don't have anything around that old :-/
[29-Jan-2010 15:18:13] <venturaville> sorry
[29-Jan-2010 15:18:19] <mistich> ok thanks
[29-Jan-2010 15:26:00] <rmatte> mistich: put a space between the + and the calls
[29-Jan-2010 15:26:20] <theacolyte> I'm in monitoring overload
[29-Jan-2010 15:26:23] <theacolyte> I put HP SIM back up
[29-Jan-2010 15:26:23] <mistich> been there done that
[29-Jan-2010 15:26:31] <rmatte> (here.getRRDValue('mem5minUsed') + here.getRRDValue('mem5minFree')) * 0.8
[29-Jan-2010 15:26:31] <mistich> I'm on 2.3.3
[29-Jan-2010 15:26:41] <rmatte> you weren't doing the getRRDValue calls properly either
[29-Jan-2010 15:27:53] <mistich> same error
[29-Jan-2010 15:27:55] <mistich> __traceback_info__: (here.getRRDValue('mem5minUsed') + here.getRRDValue('mem5minFree')) * 0.8 Module Python expression "(here.getRRDValue('mem5minUsed') + here.getRRDValue('mem5minFree')) * 0.8", line 1, in <expression> TypeError: unsupported operand type(s) for +: 'NoneType' and 'NoneType'
[29-Jan-2010 15:31:26] <rmatte> mistich: no idea then, you really should upgrade anyways
[29-Jan-2010 15:31:42] <rmatte> it's ridiculous that they are about to release 2.5.2 and you're still on 2.3
[29-Jan-2010 15:31:59] <rmatte> at least upgrade to 2.4.5
[29-Jan-2010 15:32:22] <rmatte> then you'll be 1 upgrade away when 2.5.2 comes out, at which point you should upgrade to that
[29-Jan-2010 15:32:35] <rmatte> (I will be)
[29-Jan-2010 15:33:53] <mistich> going to try it on one instance but we are working with zenoss on upgrading
[29-Jan-2010 15:34:23] <rmatte> it sounds like those calls are probably not supported in 2.3
[29-Jan-2010 15:34:36] <rmatte> so you'll have to sit tight until you get upgraded
[29-Jan-2010 15:34:50] <mistich> thanks for all the help
[29-Jan-2010 15:35:01] <rmatte> np
[29-Jan-2010 16:02:48] <mrenz> hello devs!
[29-Jan-2010 16:03:36] <mrenz> quick question
[29-Jan-2010 16:03:59] <mrenz> zenoss 2.4 - I unmounted a drive on one of my servers and zenoss hasn't alerted.....
[29-Jan-2010 16:04:40] <mrenz> is there any way for zenoss to alert when a drive gets unmounted?
[29-Jan-2010 16:05:05] <mrenz> (in 2.4)
[29-Jan-2010 16:11:37] <mrenz> hello?
[29-Jan-2010 16:19:05] <theacolyte> I don't know the answer, but patience is a virtue in smaller sized channels
[29-Jan-2010 16:19:20] <theacolyte> Forums are a good resource as well
[29-Jan-2010 16:19:49] <mrenz> yesyes
[29-Jan-2010 17:00:52] <rmatte> mrenz: that was literally the tip of the month this month
[29-Jan-2010 17:01:48] <theacolyte> woah... zenoss just lost it
[29-Jan-2010 17:01:52] <rmatte> mrenz: http://community.zenoss.org/blogs/zenossblog/2010/01/25/tip-of-the-month-monitoring-the-status-of-nfs-shares-with-zenoss
[29-Jan-2010 17:02:21] <rmatte> mrenz: If you want it to apply to all drives except for just nfs shares, remove the part " and re.match('networkDisk', f.type)"
[29-Jan-2010 17:02:35] <rmatte> s/except/instead of
[29-Jan-2010 17:02:58] <theacolyte> it's running really slowly and thinks everything is down
[29-Jan-2010 17:03:12] <rmatte> theacolyte: maybe you actually have a network problem?
[29-Jan-2010 17:03:30] <theacolyte> definately not this time
[29-Jan-2010 17:03:31] <theacolyte> I'm on it
[29-Jan-2010 17:04:12] <theacolyte> awhile back though I got some heartbeat errors tho
[29-Jan-2010 17:05:35] <theacolyte> I'm rebooting the VM, there's something wrong with something
[29-Jan-2010 17:05:45] <theacolyte> not sure what, but even sudo zenoss-stack restart took forever
[29-Jan-2010 17:06:53] <rmatte> hmmm
[29-Jan-2010 17:07:00] <rmatte> disk IO maybe?
[29-Jan-2010 17:07:50] <theacolyte> I'm going to take a look at the logs when it comes back up
[29-Jan-2010 17:07:53] <theacolyte> That wasn't healthy though
[29-Jan-2010 17:08:05] <theacolyte> In it's defense though I'm doing a LOT of updates right now
[29-Jan-2010 17:10:01] <rmatte> lol
[29-Jan-2010 17:14:22] <theacolyte> although nothing that would cause all my stuff to fail heartbeat??
[29-Jan-2010 17:24:08] <rmatte> well, I have heartbeat failures from zentrap fairly often on one of my servers, but it doesn't really mean anything, just means that zentrap is working hard
[29-Jan-2010 17:24:24] <theacolyte> it wasn't just zendtrap hehe
[29-Jan-2010 17:24:29] <theacolyte> it was half of the daemons
[29-Jan-2010 17:24:31] <theacolyte> i'm back though
[29-Jan-2010 17:24:39] <rmatte> well that's because the system was choking
[29-Jan-2010 17:24:41] <theacolyte> it looks like a nic dropped when I vmotioned and something bad happened
[29-Jan-2010 17:24:43] <rmatte> so they couldn't keep up
[29-Jan-2010 17:34:23] <rmatte> well, I'd better get out of here before it gets even colder outside
[29-Jan-2010 17:35:07] <rmatte> Weather for Ottawa, ON
[29-Jan-2010 17:35:07] <rmatte> Temp: -2Â°F
[29-Jan-2010 17:35:07] <rmatte> Current: Clear
[29-Jan-2010 17:35:07] <rmatte> Wind: W at 24 km/h
[29-Jan-2010 17:35:07] <rmatte> Humidity: 43%
[29-Jan-2010 17:35:25] <rmatte> the wind isn't helping either
[29-Jan-2010 17:41:35] <mrayzenoss> have a good weekend, mind the ice
[29-Jan-2010 17:41:49] <theacolyte> Balmy
[29-Jan-2010 17:43:38] <mrayzenoss> see everyone at OpsCamp tomorrow http://opscamp.org/austin/2010-01-30
[29-Jan-2010 20:35:47] snowgod is now known as admgre
[29-Jan-2010 20:39:54] etank is now known as elake
[29-Jan-2010 20:44:35] elake is now known as etank
[29-Jan-2010 20:50:29] [disconnected at Fri Jan 29 20:50:29 2010]
[30-Jan-2010 00:00:30] [connected at Sat Jan 30 00:00:30 2010]
[30-Jan-2010 00:57:39] fOrsberg is now known as forsberg
[30-Jan-2010 02:33:28] [disconnected at Sat Jan 30 02:33:28 2010]
[31-Jan-2010 00:00:30] [connected at Sun Jan 31 00:00:30 2010]
[31-Jan-2010 02:49:47] <frozty_sa> jeez
[31-Jan-2010 02:49:59] <frozty_sa> freenode's been quite bumpy :)
[31-Jan-2010 03:17:53] fOrsberg is now known as forsberg
[31-Jan-2010 06:46:58] <Egyptian[web]> after noon all
[31-Jan-2010 06:47:29] <Egyptian[web]> i am currently categorizing my servers and couldnt find a blade chassis class .. or is there ?
[31-Jan-2010 15:10:03] <bitmonk> I recall someone saying that the rrd databases are initiated with a static interval of sorts, and changing the poller interval(s) can break graphs or make them unintelligible. Is there a straightforward way to clear them?
[31-Jan-2010 15:10:17] <davetoo> "clear" them?
[31-Jan-2010 15:10:31] <davetoo> if you break them, you'll just have to delete (or mv) the .rrd files
[31-Jan-2010 15:11:04] <bitmonk> and they'll just be repopulated?
[31-Jan-2010 15:11:18] <davetoo> yes, the default interval is 5 minutes (300 seconds) and ... is set by the "default create command" in the Collector configs
[31-Jan-2010 15:11:19] <davetoo> yep
[31-Jan-2010 15:11:22] <forsberg> yes, they are in $ZENHOME/perf/Devices/devicename
[31-Jan-2010 15:11:35] <bitmonk> cool, thanks guys, that'll do..
[31-Jan-2010 15:11:47] * davetoo has been neck-deep in RRD for a couple of weeks
[31-Jan-2010 15:12:31] <forsberg> :]
[31-Jan-2010 15:12:52] <davetoo> at the moment I'm working to compare sar() ldAvg-5 data with zenoss' laLoad5 data.
[31-Jan-2010 15:13:42] <davetoo> I've written a bunch of shell scripts to dump the rrd to xml, put each RRA into it's own column in a .csv
[31-Jan-2010 15:13:51] <davetoo> with timestamps
[31-Jan-2010 15:14:18] <davetoo> Sometimes, I don't want RRDTools' "help" :)
[31-Jan-2010 15:14:29] <forsberg> :P
[31-Jan-2010 15:14:49] <bitmonk> davetoo: what's your goal? sounds interesting..
[31-Jan-2010 15:15:05] <davetoo> Plotting the data with ploticus :)
[31-Jan-2010 15:15:21] <davetoo> and doing statistical analysis, eventually
[31-Jan-2010 15:15:46] <davetoo> comparing the system load across about 40 servers vs. time of day to look for variances
[31-Jan-2010 15:15:59] <davetoo> doing things that Zenoss/RRDTool aren't designed for, basically :)
[31-Jan-2010 15:16:15] <bitmonk> interesting..
[31-Jan-2010 15:16:51] <bitmonk> no reason zenoss shouldn't be able to do that by some kind of 'role group' or something
[31-Jan-2010 15:16:59] <davetoo> I've learned a lot of new sed(1) tricks :) Probably should have just done it with python, though.
[31-Jan-2010 15:17:27] <davetoo> well... I'm going to calculate mean/std-deviation and such for each timestamp across all 40 servers
[31-Jan-2010 15:17:57] <davetoo> and visualize it a couple of ways. One is just to do a scatter plot with time as the x-axis
[31-Jan-2010 17:12:04] forsberg is now known as fOrsberg
[31-Jan-2010 17:44:40] <swygue> Hello, I'm looking for a quick way to move all acknowledge events to history using zendmd or from mysql
[01-Feb-2010 00:00:30] [disconnected at Mon Feb 1 00:00:30 2010]