[01-May-2010 03:09:30] <Zeppo> is it possible to run zendisc outside of zenoss? And if it is, what are the minimal requirements (i.e. zenhub running and so on)? I'd like to detect a network topology the way zenoss does, using these information in my own script. TIA

[01-May-2010 03:26:44] fOrsberg is now known as forsberg

[01-May-2010 05:06:06] forsberg is now known as fOrsberg

[01-May-2010 13:01:15] fOrsberg is now known as forsberg

[01-May-2010 14:52:01] <twm1010> cluther: you around?

[01-May-2010 14:54:20] <twm1010> Anyone which daemon handles maintenance windows?

[01-May-2010 14:54:28] forsberg is now known as fOrsberg

[02-May-2010 00:00:30] [disconnected at Sun May  2 00:00:30 2010]

[02-May-2010 00:00:31] [connected at Sun May  2 00:00:31 2010]

[02-May-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[02-May-2010 04:36:44] fOrsberg is now known as forsberg

[02-May-2010 08:04:01] <Zeppo> is it possible to run zendisc outside of zenoss? And if it is, what are the minimal requirements (i.e. zenhub running and so on)? I'd like to detect a network topology the way zenoss does, using these information in my own script. TIA

[02-May-2010 11:13:34] forsberg is now known as fOrsberg

[02-May-2010 12:32:41] fOrsberg is now known as forsberg

[02-May-2010 14:04:15] <voxter> Anyone use the xen virtual hosts zenpack?

[02-May-2010 16:04:59] <chemist> Hamzah: have you set up any alerting rules?

[02-May-2010 16:07:15] <Hamzah> I figured it out :) I had the alerts setup, but something was going odd with zen actions, restarting it fixed the problem

[02-May-2010 16:11:48] <chemist> cool :)

[02-May-2010 18:00:29] forsberg is now known as fOrsberg

[03-May-2010 00:00:30] [disconnected at Mon May  3 00:00:30 2010]

[03-May-2010 00:00:31] [connected at Mon May  3 00:00:31 2010]

[03-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[03-May-2010 02:29:17] <voxter> anyone awake?

[03-May-2010 02:32:55] <rhettardo> i yam

[03-May-2010 02:34:13] <voxter> Maybe you can shed some light on this for me, I just added a new device, expected it to use my ssh key i had set up to complete ssh modelling... it failed, and ive fixed what i think the problem was, the issue is when i tell zenoss to re model the device, it doesnt appear to try ssh again

[03-May-2010 02:34:32] <rhettardo> hm

[03-May-2010 02:34:44] <rhettardo> not sure about that one.  i dont use ssh modeling

[03-May-2010 02:35:20] <rhettardo> i can surely give it a try here though real quick

[03-May-2010 02:35:47] <voxter> hmm. I think its the collector plugin, this is specific to a xen monitoring xenpack

[03-May-2010 02:35:53] <voxter> i just added a zenoss.cmd.linux check and that worked.

[03-May-2010 02:35:54] <rhettardo> oh

[03-May-2010 02:38:25] <voxter> ill try deleting/recreating

[03-May-2010 02:38:54] <voxter> although that would prove as a bug i suppose, that this zenpack doesnt rescan for xen guests on a remodel. seems like a pretty significant flaw. although i suppose its entirely likely i could be 'doing it wrong' ?

[03-May-2010 02:38:54] <voxter> :P

[03-May-2010 02:40:00] <rhettardo> ha

[03-May-2010 02:40:35] <rhettardo> well if it wernt for the community finding issues and proving those to the dev's we would not have the wonderful tool to use for 'free'

[03-May-2010 02:43:48] <voxter> so very true.

[03-May-2010 02:49:49] <broken_puppy> So, I'm sure you hate this question, but....

[03-May-2010 02:50:22] <broken_puppy> Loading initial Zenoss objects into the Zeo database... (this can take a few minutes).  -- how many is a few?

[03-May-2010 02:50:43] <voxter> mine took 15 ish

[03-May-2010 02:51:14] <broken_puppy> 15!!! File a bug or show some activity.  Wow.

[03-May-2010 02:52:11] <broken_puppy> Guess I have time to un-install cacti.

[03-May-2010 02:52:35] <voxter> depends on the system i guess. my baseline was a xen domU with poorly performing disks

[03-May-2010 02:52:58] <broken_puppy> voxter, thanks for the point of ref.

[03-May-2010 02:53:20] <broken_puppy> I seriously started poking around the system to see why it failed.

[03-May-2010 02:58:43] * broken_puppy wonders how much longer...

[03-May-2010 03:04:46] <broken_puppy> Here's another question (as a first-time zenoss downloader and installer (pre-user))... Why does it come with it's own everything (python etc)?

[03-May-2010 03:05:24] <broken_puppy> Seems like it would get more traction if you could apt-get it and it wasn't in .opt

[03-May-2010 03:05:43] <voxter> broken_puppy: you can get it both ways

[03-May-2010 03:05:49] <voxter> broken_puppy: i presume you got the stack installer. its for noobs.

[03-May-2010 03:06:08] <voxter> so they dont have to field as many 'Why am i getting "python" command not found errors!!@?#"

[03-May-2010 03:06:13] <voxter> there are rpms and debs

[03-May-2010 03:07:11] <voxter> http://www.zenoss.com/download/links

[03-May-2010 03:07:20] <voxter> "native packages"

[03-May-2010 03:08:18] <voxter> oh those are only rpms!

[03-May-2010 03:08:27] <broken_puppy> voxter, yeah, I got the RPM for CentOS (RHEL).  Just feels kinda heavy, but I understand the support issues.

[03-May-2010 03:08:45] <voxter> i pulled a yum repo out of svn/git

[03-May-2010 03:09:45] <broken_puppy> voxter, where would I look to see that something is really happening? I'm nearing 20 mins.

[03-May-2010 03:10:05] <voxter> broken_puppy: hmm. honestly, i dont really know. i just know i left mine for a significant amount of time and came back and it was done.

[03-May-2010 03:10:16] <broken_puppy> oh, OK

[03-May-2010 03:10:26] <broken_puppy> I'll give it some more time.

[03-May-2010 03:39:45] <broken_puppy> All right.  It can't be taking this long.  Who can point me the the best way to trouble shoot a hung Zeo database load?

[03-May-2010 03:40:10] <broken_puppy> I was nearing 40 mins.

[03-May-2010 03:40:41] <broken_puppy> and the results of du -s /opt/zenoss/ wasn't changing.

[03-May-2010 03:55:50] <fragfutter> how can i monitor a clustered service? The simplest cluster i want to monitor has a virtual IP that is floating between two machines. The active machine is running a special mount and a webservice. Mount, IP, Service will migrate to/from the standby system.

[03-May-2010 05:17:14] <fragfutter> i have a machine where zenoss detects wrong total memory, how would i debug this?

[03-May-2010 05:27:16] <fragfutter> ah. the machine hasn't been modeled.

[03-May-2010 08:02:26] <TSCDan> When is Stone Crab supposed to be released? I was planning on redoing my zenoss install (it's a bit funky at the moment) when the new ubuntu version came out (last thursday), but then I got an email saying that stone crab is in alpha. Gah! So I'm trying to determine if I should wait for stone crab or just go forward with the current 2.5

[03-May-2010 08:42:25] <jb> hey ckrough.. you around?

[03-May-2010 08:54:07] <etank> is there a best practice doc for partitioning for performance with a zenoss install?

[03-May-2010 08:54:09] <ckrough> ya

[03-May-2010 08:54:19] <ckrough> jb: ya

[03-May-2010 08:54:55] <ckrough> etank: I included some partition info here: blogs/zenossblog/2010/02/25/tip-of-the-month-performance-tuning-for-zenoss-storage

[03-May-2010 08:56:24] <etank> thanks ckrough

[03-May-2010 09:02:13] <jb> ckrough: got a couple questions about dist collectors..

[03-May-2010 09:02:29] <jb> should you be able to run zenperfsnmp on the remote collector?

[03-May-2010 09:03:03] <ckrough> yup

[03-May-2010 09:03:16] <jb> hrmf.

[03-May-2010 09:03:33] <jb> doesn't even have a config

[03-May-2010 09:03:34] <jb> zenoss@fc-zencpci01:/opt/zenoss/log/zencpci01$ zenperfsnmp run -v10

[03-May-2010 09:03:34] <jb> WARN: unable to read config file /opt/zenoss/etc/zenperfsnmp.conf -- skipping

[03-May-2010 09:03:34] <jb> Traceback (most recent call last):

[03-May-2010 09:03:34] <jb> IOError: [Errno 2] No such file or directory: '/opt/zenoss/etc/zenperfsnmp.conf'

[03-May-2010 09:03:59] <ckrough> it should have pushed that when it updated the collector

[03-May-2010 09:04:08] <jb> its there, but its named collector_zenperfsnmp.conf

[03-May-2010 09:04:08] <ckrough> try doing an update to that collector from the gui again

[03-May-2010 09:04:11] <jb> k

[03-May-2010 09:04:15] <ckrough> ya, thats right

[03-May-2010 09:04:24] <ckrough> I assume the collector is named 'collector' in the gui?

[03-May-2010 09:04:35] <jb> no, its zencpci01_zenperfsnmp.conf

[03-May-2010 09:04:41] <jb> thats the name of the collector (zencpci01)

[03-May-2010 09:04:49] <ckrough> if you look in the daemons.txt its running 'zencpci01_zenperfsnmp' not 'zenperfsnmp'

[03-May-2010 09:04:55] <jb> sec

[03-May-2010 09:05:03] <jb> yep.

[03-May-2010 09:05:18] <ckrough> try the collector specific one, it should get the right config file

[03-May-2010 09:05:21] <jb> but when I manually run zenperfsnmp, it looks for zenpersnmp.conf

[03-May-2010 09:05:26] <jb> ohhh.

[03-May-2010 09:05:32] <ckrough> ya/ dont run the vanilla one

[03-May-2010 09:05:47] <jb> got it.

[03-May-2010 09:05:59] <jb> ok, so that was successful

[03-May-2010 09:06:03] <ckrough> cool

[03-May-2010 09:06:03] <jb> but my graphs aren't working.

[03-May-2010 09:06:37] <ckrough> check your render port, make sure you can get to the remote server on that port. graphs are served up via zenrender on the collector

[03-May-2010 09:07:01] <jb> hm

[03-May-2010 09:07:30] <jb> check the remote collectors zenrender conf?

[03-May-2010 09:07:42] <jb> hubhost/hubport?

[03-May-2010 09:07:59] <ckrough> yeah, make sure its running and make sure you can connect to that port from wherever you are

[03-May-2010 09:08:07] <ckrough> but not hosthub... sec

[03-May-2010 09:08:28] <ckrough> #http-port 8091

[03-May-2010 09:08:35] <ckrough> thats the listen port for the graphs

[03-May-2010 09:08:56] <jb> hrm, ok.. so I should be able to connect to that port on my remote collector -from- my hub?

[03-May-2010 09:09:11] <jb> that works..

[03-May-2010 09:09:11] <ckrough> from your desk

[03-May-2010 09:09:14] <ckrough> cool

[03-May-2010 09:09:58] <jb> 2010-05-03 09:06:31,509 INFO zen.zenrender: Connecting to 172.26.101.55:8789

[03-May-2010 09:09:58] <jb> 2010-05-03 09:06:31,514 INFO zen.zenrender: Connected to ZenHub

[03-May-2010 09:13:48] <jb> welp

[03-May-2010 09:19:24] <jb> hrm, i dunno whats going on.

[03-May-2010 09:21:58] <jb> Chapter 5. Distributed Name Server (DNS)

[03-May-2010 09:21:59] <jb> erm :)

[03-May-2010 09:22:06] <jb> distributed name server?

[03-May-2010 09:23:09] <ckrough> lol

[03-May-2010 09:23:28] <jb> mrayzenoss: might want to revise that (the extended monitoring doc) :)

[03-May-2010 09:23:32] <ckrough> jb: any errors in your zenrender logs on the collector?

[03-May-2010 09:23:38] <jb> none

[03-May-2010 09:27:51] <jb> and browsing devices on this collector is just slow

[03-May-2010 09:27:56] <mrayzenoss> jb: http://dev.zenoss.org/trac/ticket/6643

[03-May-2010 09:28:04] <jb> i clicked the "os" tab of one, and its taking like 20-30 seconds

[03-May-2010 09:46:01] <rmatte> bah, I lost my 200+ days uptime on my workstation, power failure

[03-May-2010 09:46:30] <ckrough> bummer

[03-May-2010 09:46:59] <ckrough> a guy here is at 900 now, used a UPS to move it between desks

[03-May-2010 09:47:01] <ckrough> slackware ftw

[03-May-2010 09:47:04] <twm1010> somewhere on this machine i have a screenshot of a netware 5 box that had a 4 yr

[03-May-2010 09:47:25] <twm1010> it ran our company intranet, 300mhz, 64mb ram, it was a beast!

[03-May-2010 09:47:32] <ckrough> they dont make them like they used to

[03-May-2010 09:53:34] <rmatte> lol

[03-May-2010 09:58:59] <rmatte> If anyone is able to help me out with this today it would be appreciated: thread/13431?tstart=0

[03-May-2010 10:00:57] <fragfutter> rmatte did you look at your object inside the zope management interface?

[03-May-2010 10:17:09] <rmatte> fragfutter: well, what am I looking for specifically?

[03-May-2010 10:36:45] <TSCDan> Does anybody have a rough idea of when stone crab is slated to be released?

[03-May-2010 10:37:59] <rmatte> like 5 months-ish

[03-May-2010 10:38:02] <rmatte> perhaps 4

[03-May-2010 10:38:06] <TSCDan> oh jeez

[03-May-2010 10:38:16] <rmatte> it's only in alpha right now, still lots of work to do

[03-May-2010 10:38:18] <TSCDan> ok, nvm then, guess I'll redo my 2.5 now

[03-May-2010 10:38:32] <rmatte> hehe

[03-May-2010 10:38:35] <TSCDan> :-)

[03-May-2010 10:38:38] <TSCDan> Thanks for the info

[03-May-2010 10:38:41] <rmatte> np

[03-May-2010 10:50:13] <fragfutter> rmatte: you can look at the device and see if it has a defined property or if it acquired it.

[03-May-2010 10:56:51] <twm1010> rmatte: bit the bullet, on 2.5.2 now ;)

[03-May-2010 10:56:55] <fragfutter> rmatte: and for your second post, you should be able to test with hasattr or just do it and pack it in a try except clause.

[03-May-2010 11:06:23] <rmatte> fragfutter: right, but is deleting the attribute going to cause any issues?

[03-May-2010 11:06:31] <rmatte> like, will the WMI pack cease to function properly

[03-May-2010 11:06:36] <rmatte> that's what I'm most concerned with

[03-May-2010 11:06:37] <fragfutter> rmatte: can't tell you.

[03-May-2010 11:06:47] <rmatte> yeh, I'll have to wait to speak with Egor

[03-May-2010 11:06:57] <fragfutter> rmatte: i suppose it is an acquired property on most of your nodes.

[03-May-2010 11:07:17] <rmatte> yes

[03-May-2010 11:07:38] <fragfutter> rmatte: so most nodes won't have it. they all acquire it from further up.

[03-May-2010 11:08:27] <rmatte> yeh, but each one is throwing the error and deleting it locally on that specific device clears it up

[03-May-2010 11:08:52] <rmatte> I don't even use the WMI pack yet, so properties aren't set locally anywhere

[03-May-2010 11:11:02] <fragfutter> rmatte: you could check it http://pastebin.com/mcU8wEZn

[03-May-2010 11:15:01] <rmatte> I'll try that out

[03-May-2010 11:15:02] <rmatte> thanks

[03-May-2010 11:16:47] <rmatte> it's all acquired

[03-May-2010 11:18:13] <fragfutter> rmatte: strange.

[03-May-2010 11:18:22] <rmatte> yup

[03-May-2010 11:18:52] <fragfutter> Then the propertyManager should not try to update it.

[03-May-2010 11:19:15] <rmatte> well, it seems to

[03-May-2010 11:19:21] <fragfutter> might be a bad keyindex in the zodb.

[03-May-2010 11:20:47] <rmatte> any suggestions?

[03-May-2010 11:22:12] <fragfutter> compress the zodb

[03-May-2010 11:22:21] <rmatte> zeopack?

[03-May-2010 11:22:48] <fragfutter> rmatte: should be it. i do it through the webinterface

[03-May-2010 11:23:19] <rmatte> I did that one of the boxes on Friday, didn't fix the issue

[03-May-2010 11:23:25] <fragfutter> ok

[03-May-2010 11:25:32] <rmatte> actually, let me doublecheck...

[03-May-2010 11:27:20] <rmatte> hmmm it might have actually... I'll try it on another one of the servers...

[03-May-2010 11:29:49] <fragfutter> rmatte: i'm picking in the dark. But you could try to rebuild the index. I have no idea about the internals of zenoss code. but 8080/zport/dmd/Devices/deviceSearch/manage_catalogAdvanced should be the right index. Update catalog and see if it helps

[03-May-2010 11:30:05] <fragfutter> i need to leave or HR will beat me out of the building ;)

[03-May-2010 11:35:47] <rmatte> nope, didn't fix it

[03-May-2010 12:52:06] <rmatte> egor!

[03-May-2010 13:02:23] <rmatte> npmccallum: If you have a second, any ideas about message/48377 ?

[03-May-2010 13:06:04] <npmccallum> rmatte: well, that is strange

[03-May-2010 13:06:15] <rmatte> :)

[03-May-2010 13:06:32] <rmatte> any idea on how to fix it without breaking anything?

[03-May-2010 13:06:56] <npmccallum> I've definitely seen the problem before, an in my case it was fixed by the same solution.  The root cause was someone manually setting the property on the object before it existed

[03-May-2010 13:07:19] <rmatte> hmmm, not sure how that could have happened

[03-May-2010 13:07:27] <npmccallum> keep in mind, that it might exist not just on devices, but really on any object in the tree

[03-May-2010 13:07:35] <npmccallum> device classes, components, etc

[03-May-2010 13:07:40] <rmatte> yeh

[03-May-2010 13:08:07] <npmccallum> it might have happened by a zenpack setting that attr in a dirty way

[03-May-2010 13:08:07] <rmatte> well, I ran some code that fragfutter game me to run...

[03-May-2010 13:08:17] <rmatte> http://pastebin.com/mcU8wEZn

[03-May-2010 13:08:20] <npmccallum> or really any number of things

[03-May-2010 13:08:36] <rmatte> game == gave*

[03-May-2010 13:08:43] <rmatte> and they all came back as acquired

[03-May-2010 13:09:04] <rmatte> so is the method I was using the correct way to fix it?

[03-May-2010 13:09:15] <rmatte> I'm not seeing the issue on any classes

[03-May-2010 13:09:19] <rmatte> just on the devices themselves

[03-May-2010 13:09:35] <npmccallum> i assume you're seeing a traceback in the UI?

[03-May-2010 13:09:49] <rmatte> yeh, in the UI

[03-May-2010 13:09:58] <rmatte> the traceback is what I pasted in the forum post

[03-May-2010 13:10:12] <npmccallum> the most direct way is to uncomment portions of the template until the issue goes away

[03-May-2010 13:10:32] <npmccallum> that will tell you what object is throwing that exception

[03-May-2010 13:10:43] <rmatte> which template?

[03-May-2010 13:10:57] <npmccallum> which ever template is rendering to the screen

[03-May-2010 13:10:57] <rmatte> you mean the zProperties page itself?

[03-May-2010 13:11:08] <rmatte> I see

[03-May-2010 13:11:34] <rmatte> so does __delattr__ break anything?

[03-May-2010 13:11:35] <npmccallum> hrm, you're looking at the zproperties page?

[03-May-2010 13:11:43] <npmccallum> IDK

[03-May-2010 13:11:46] <rmatte> yeh... here's what happens...

[03-May-2010 13:11:49] <rmatte> I go to a device

[03-May-2010 13:11:55] <rmatte> then I select More -> zProperties

[03-May-2010 13:12:05] <rmatte> then even without editing anything if I click Save I get the traceback

[03-May-2010 13:12:06] <npmccallum> traceback

[03-May-2010 13:12:16] <npmccallum> right, so its trying to do the write

[03-May-2010 13:12:30] <rmatte> If I __delattr__ for zWinPassword and zWinUser then it saves fine

[03-May-2010 13:12:32] <npmccallum> and failing because the attribute already exists but doesn't appear to be a zproperty

[03-May-2010 13:12:40] <rmatte> right

[03-May-2010 13:12:56] <rmatte> so deleting the already existing attribute shouldn't hurt anything because it'll be recreated if needed on next save?

[03-May-2010 13:12:59] <npmccallum> I don't believe delattr breaks anything

[03-May-2010 13:13:14] <npmccallum> I think you're safe to run it against all your devices

[03-May-2010 13:13:16] <npmccallum> HOWEVER

[03-May-2010 13:13:19] <npmccallum> back up first

[03-May-2010 13:13:25] <rmatte> k, so now I just need some zendmd code to go through and remove the attributes if they exist

[03-May-2010 13:13:38] <rmatte> I will

[03-May-2010 13:13:43] <npmccallum> I'd just do it for all of them

[03-May-2010 13:13:52] <npmccallum> and do except: pass

[03-May-2010 13:14:06] <rmatte> yeh, guess that would work too

[03-May-2010 13:14:33] <rmatte> k, I'll give it a go, thaks

[03-May-2010 13:14:35] <rmatte> thanks*

[03-May-2010 13:14:36] <npmccallum> np

[03-May-2010 13:15:55] <rmatte> the weirdest thing is that I have the problem on some devices that I added after I did the upgrade

[03-May-2010 13:16:04] <rmatte> it'll be interesting to see if it fixes that too

[03-May-2010 13:16:26] <rmatte> (for future device adds I mean)

[03-May-2010 13:16:31] <rmatte> I'll have to test afterwards

[03-May-2010 13:25:40] <voxter>  I guess it must be true that by default, zenoss does not do ping perf graphs? for latency and loss, etc?

[03-May-2010 13:27:04] <voxter> ahh. just throughput, packets, and errors. not latency

[03-May-2010 13:49:41] <rmatte> eugh, our link to the datacenter is horrible today

[03-May-2010 14:02:35] <twm1010> anyone setup SMS paging?

[03-May-2010 14:14:37] <rmatte> voxter: all the stuff it collects by default for traffic is via snmp (throughput, packets, and errors)

[03-May-2010 14:14:47] <rmatte> there are some zenpacks for doing ping graphing

[03-May-2010 14:15:21] <rmatte> such as: docs/DOC-3467

[03-May-2010 14:15:22] <voxter> Yep got 'er. I'm installing the fping zenpack right now, its unfortunate that it doesnt behave like smokeping

[03-May-2010 14:15:29] <rmatte> cool

[03-May-2010 14:15:38] <rmatte> well, what does smokeping do that's so special?

[03-May-2010 14:15:42] <voxter> If i knew enough about building graphs i would make it myself, but i am still just getting my feet wet

[03-May-2010 14:15:49] <voxter> it combines latency and packet loss into a single graph

[03-May-2010 14:15:55] <voxter> incredibly ingeniously

[03-May-2010 14:16:15] <rmatte> that would take 2 seconds to do with that fping zenpack

[03-May-2010 14:16:28] <rmatte> just add the packet loss graphpoint to the latency and loss graph

[03-May-2010 14:16:30] <rmatte> and voila

[03-May-2010 14:17:23] <voxter> Im not sure it can do this though, check this out

[03-May-2010 14:17:26] <voxter> http://oss.oetiker.ch/smokeping-demo/img/Customers/OP/johan~mipsrv01_last_10800.png

[03-May-2010 14:17:42] <voxter> you see how each reporting cycle is represented in a "bar" most are green, in this graph two are blue

[03-May-2010 14:17:50] <rmatte> ah, I see what you're saying

[03-May-2010 14:17:53] <voxter> blue = 1/20 packets lost (as represented by the graph at the bottom)

[03-May-2010 14:18:08] <voxter> the bar is drawn as a line in the graph where AVG RTT is

[03-May-2010 14:18:12] <rmatte> yeh, Zenoss doesn't support that type of graphing

[03-May-2010 14:18:15] <rmatte> just line or area

[03-May-2010 14:18:21] <voxter> MAX rtt and MIN rtt are above and below

[03-May-2010 14:18:26] <rhettardo> ah yes.  i use a template similar to smokeping in cacti

[03-May-2010 14:18:28] <rmatte> you could do it with lines but it wouldn't look quite as nice

[03-May-2010 14:18:31] <ckrough> you can do ticks in zenoss too

[03-May-2010 14:18:38] <rmatte> ckrough: oh?

[03-May-2010 14:18:40] <ckrough> hwpredict does it

[03-May-2010 14:18:45] <voxter> ticks = individual colored bars?

[03-May-2010 14:18:48] <ckrough> ya

[03-May-2010 14:18:53] <rmatte> cool, didn't know that

[03-May-2010 14:19:02] <rmatte> is it all done via UI or do you have to do some special config?

[03-May-2010 14:19:07] <rmatte> I've never seen the option

[03-May-2010 14:19:14] <ckrough> custom config in the graph

[03-May-2010 14:19:18] <rmatte> ah

[03-May-2010 14:19:19] <voxter> you should be able to achieve this then... decide your color based on a percentage calculated from loss

[03-May-2010 14:19:28] <voxter> ckrough: do you have an example that I might be able to follow that does this?

[03-May-2010 14:19:29] <ckrough> you can put just about anything rrdtool supports, to my knowledge

[03-May-2010 14:19:38] <rmatte> right, pain in the rear though

[03-May-2010 14:19:41] <ckrough> not anymore, but used to have them in the past

[03-May-2010 14:19:45] <rmatte> unless you're an rrdtool guru

[03-May-2010 14:19:52] <ckrough> its not too bad

[03-May-2010 14:20:06] <voxter> I wonder if you do it once, if you can save those custom mods in a zenpack?

[03-May-2010 14:20:17] <voxter> so that anyone else wishing to do it doesnt have to also do the 'custom mod'

[03-May-2010 14:20:22] <ckrough> yeah, they become part of the templat

[03-May-2010 14:20:26] <ckrough> +e

[03-May-2010 14:20:27] <voxter> and so that it lasts through upgrades.

[03-May-2010 14:20:34] <rmatte> voxter: yeh, there's a spot right in the graph to specify rrd options

[03-May-2010 14:20:34] <ckrough> yes

[03-May-2010 14:20:39] <rmatte> so it would be part of the ZenPack

[03-May-2010 14:20:43] <voxter> Hmm interesting.

[03-May-2010 14:21:32] <voxter> So presumably if you were able to do some math on the result of the packet loss (out of 100%) and you could specify varying colors based on varying percentage, then you could get your colored bar

[03-May-2010 14:21:44] <voxter> and plot that 'tick' using average RTT

[03-May-2010 14:21:54] <voxter> then use area bars to draw the 'smoke'

[03-May-2010 14:22:25] <voxter> it eliminates the need for smokeping 100%, and arguably smokeping is an untouchably valuable rrd tool right now

[03-May-2010 14:22:50] <rmatte> yeh, it could be done, would just take some dev work by someone knowledgeable enough to do it

[03-May-2010 14:23:31] <voxter> Ok, so either I can take a stab at it and take possibly up to 3 days and hours of yelling at zenoss/rrd due to my current lack of understanding, or maybe one of you with some familiarity can assist me? :)

[03-May-2010 14:23:48] <ckrough> hah, queue up them 3 days :)

[03-May-2010 14:23:57] <ckrough> but Im sure we can get you through the rough spots

[03-May-2010 14:24:16] <rmatte> I can assist you with anything to do with actually packaging the zenpack and creating the template, but I don't know rrdtool well enough to know how to do the actual config properly to get what you're wanting

[03-May-2010 14:24:42] <ckrough> its instant feedback, so you can toy and refresh until you get it right. watch the render.log for errors and bring up the rrdgraph man page and you should be golden

[03-May-2010 14:24:47] <rmatte> yeh

[03-May-2010 14:25:00] <rmatte> well, wouldn't he have to kill the old RRD each time?

[03-May-2010 14:25:06] <ckrough> also, somewhere in smokeping is the rrdgraph template that creates that graph... steal that

[03-May-2010 14:25:10] <rmatte> or does it just adjust it on the fly?

[03-May-2010 14:25:49] <ckrough> ah, yeah, I guess it depends on how many rrds hold the data. i was thinking it was all one

[03-May-2010 14:25:53] <rmatte> yeh, if you can get the rrdgraph template from smokeping that's 50% of the work already done

[03-May-2010 14:26:24] <rmatte> no point in reinventing the wheel

[03-May-2010 14:26:53] <rmatte> a smokeping zenpack would be a nice addition

[03-May-2010 14:27:29] <rmatte> see what you can do, if you run in to snags you can always ask in here, if you need someone to polish the pack up a bit when you're done with it I can do that for you

[03-May-2010 14:27:52] <ckrough> the fping zenpack shows packet loss by bar color, it's just not rendered like smokeping, so the rras should be there, its just a matter of rrdgraph

[03-May-2010 14:28:45] <rmatte> yeh, if you can modify that pack to do what you want then send it to me I can rename it and credit whoever helped (including the original fping developer)

[03-May-2010 14:28:57] <rmatte> fping pack developer rather

[03-May-2010 14:33:47] <voxter> ok Im gonna dig into it here and see what I can find... i guess i should do this as a mod to fping, ya?

[03-May-2010 14:34:06] <rmatte> it would be the best idea, most of the work will already be done that way

[03-May-2010 14:34:39] <rmatte> then when you're done modding it I can walk you through saving it, then you can send it to me and I'll rename it (there's quite a bit involved in renaming it)

[03-May-2010 14:34:57] <rmatte> then I'll fire it back to you and you can submit it

[03-May-2010 14:36:52] <voxter> OK great.  I'll start by getting my hands dirty in fping and smokeping's RRD.

[03-May-2010 14:36:57] <rmatte> cool

[03-May-2010 14:40:08] <twm1010> I'm so close :|

[03-May-2010 14:40:18] <twm1010> I can get qpage to send from command line, but zenoss just tells me "failed" :|

[03-May-2010 14:41:03] <rmatte> did you check zenactions.log?

[03-May-2010 14:41:09] <twm1010> good point

[03-May-2010 14:41:45] <rmatte> also, are you following this: http://docs.google.com/viewer?a=v&q=cache:cmJlvdLkAD8J:community.zenoss.org/docs/DOC-5819.pdf%3Bjsessionid%3DFA76943ED0FC18B23294F5AFC02DDB09.node0+zenoss+qpage&hl=en&gl=ca&pid=bl&srcid=ADGEESikMdDvAT9du7wonhe5-c37RHkmiQqwKTDrPvrSp0t6m7ZWY2MUS7QAqruNlHenu9LZAUfcrTaoOp-9kCEW2nW5OCOydk-LseWGAXftQ-7DcD-4yh-UeazCSwmBSbi0DnRbpgLX&sig=AHIEtbTnaXenRe9BBXXbd1eo8OSy-oVTIA

[03-May-2010 14:41:46] <rmatte> ?

[03-May-2010 14:41:58] <twm1010> Yes

[03-May-2010 14:42:12] <twm1010> everything works as expected all the way up to the testing from zenoss itself

[03-May-2010 14:42:30] <rmatte> k

[03-May-2010 14:42:39] <rmatte> zenactions.log is your best bet then

[03-May-2010 14:42:47] <rmatte> you can also try: zenactions run -v10 --cycle

[03-May-2010 14:42:52] <rmatte> not sure if it support --cycle or not

[03-May-2010 14:43:00] <rmatte> but if it does, that'll give you a running view of what the daemon is doing

[03-May-2010 14:43:35] <rmatte> yeh it does support --cycle

[03-May-2010 14:43:37] <rmatte> so do that

[03-May-2010 14:44:40] <twm1010> hrmm... nothing scrolling by on me here

[03-May-2010 14:45:33] <rmatte> hmmm

[03-May-2010 14:48:12] <twm1010> I do get some 'Lock wait timeout exceeded" messages , but thats not unusual

[03-May-2010 14:48:23] <twm1010> might have something to do with my 29GB mysql database :|

[03-May-2010 14:48:52] <rmatte> you're sure that $ZENHOME/bin/sendqpage.sh is executable as zenoss user?

[03-May-2010 14:51:34] <rmatte> and are you able to manually run qpage -dip [pager] [message] as the zenoss user?

[03-May-2010 14:51:54] <rmatte> need to make sure it's actually working via commandline before blaming zenoss ;)

[03-May-2010 14:56:10] <voxter> Question. I've added the fping zenpack.  It appears as though it attaches itself to the /Devices/Server path automatically. but going into the Perf tab of a server in that tree, i see no fping stuff.  Is that because i need to remodel all the devices now that the fping template has been added, or?

[03-May-2010 14:56:23] <rmatte> no

[03-May-2010 14:56:26] <rmatte> go to that class

[03-May-2010 14:56:32] <rmatte> then go to templates

[03-May-2010 14:56:38] <rmatte> then from the dropdown select bind templates

[03-May-2010 14:57:01] <rmatte> then bind that template to that class (best to hold down the ctrl key when doing it so that you don't unbind any templates that are already bound)

[03-May-2010 14:57:05] <voxter> oh i see. they are just listed as available, not necessarily bound.

[03-May-2010 14:57:10] <rmatte> correct

[03-May-2010 14:57:34] <rmatte> it gives more flexibility that way

[03-May-2010 14:57:47] <rmatte> you can create a template at the root level and only bind it to the classes that you want to

[03-May-2010 14:58:17] <rmatte> I would expect that template to be available at the root level

[03-May-2010 14:58:25] <rmatte> it would be weird if it was only available under the server class

[03-May-2010 14:58:26] <voxter> right.

[03-May-2010 14:58:33] <rmatte> cool

[03-May-2010 14:58:39] <voxter> interesting that when you go to Perf, you have to rely on the graph heading to see what template is providing it..

[03-May-2010 14:59:04] <voxter> For example i have a couple ping-like templates attached right now, and two of them extend a graph called "Packet loss" im not sure which one is from which. :P

[03-May-2010 14:59:07] <rmatte> well, it just looks cleaner that way

[03-May-2010 14:59:14] <rmatte> I've never personally lost track of what's providing what

[03-May-2010 14:59:36] <rmatte> yeh well, you wouldn't normally run in to a situation like that

[03-May-2010 14:59:39] <voxter> Fair enough. i guess its only confusing if you are testing out templates that each provide similar data

[03-May-2010 14:59:45] <rmatte> correct

[03-May-2010 15:00:02] <rmatte> you really should just unbind the other template to avoid confusion

[03-May-2010 15:00:26] <twm1010> rmatte: ah ha

[03-May-2010 15:00:34] <rmatte> twm1010: hm?

[03-May-2010 15:00:35] <twm1010> correct, i was doing my qpage tests as root

[03-May-2010 15:00:39] <rmatte> ah

[03-May-2010 15:00:51] <rmatte> yeh, permissions issue was the most likely thing

[03-May-2010 15:00:57] <twm1010> perhaps i should configure the daemon to run as zenoss user?

[03-May-2010 15:01:20] <rmatte> well... that doesn't seem to be necessary from what I'm reading in the guide

[03-May-2010 15:01:28] <rmatte> the zenoss user just needs to be able to execute the qpage command

[03-May-2010 15:01:45] <rmatte> but you could, whatever works really

[03-May-2010 15:02:05] <twm1010> here's my output:

[03-May-2010 15:02:07] <twm1010>  qpage -dip tmcnicho test

[03-May-2010 15:02:07] <twm1010> /etc/qpage.conf(8): cannot access /dev/ttyS0: Permission denied

[03-May-2010 15:02:07] <twm1010> /etc/qpage.conf(11): no device for modem=ttya

[03-May-2010 15:02:07] <twm1010> cannot access LockDir(/var/lock): Permission denied

[03-May-2010 15:02:07] <twm1010> Error reading configuration file

[03-May-2010 15:02:24] <rmatte> hmmm, one sec...

[03-May-2010 15:02:59] <rmatte> you have sudo installed on that box I assume?

[03-May-2010 15:03:17] <rmatte> can you execute: which qpage

[03-May-2010 15:03:23] <rmatte> and tell me the path to it?

[03-May-2010 15:04:10] <twm1010> /usr/local/bin/qpage

[03-May-2010 15:04:24] <rmatte> sudo is installed yes?

[03-May-2010 15:04:26] <twm1010> aye

[03-May-2010 15:04:34] <rmatte> add this line to your /etc/sudoers file

[03-May-2010 15:04:35] <rmatte> zenoss ALL=(ALL) NOPASSWD: /usr/local/bin/qpage

[03-May-2010 15:04:41] <rmatte> then try again

[03-May-2010 15:07:11] <rmatte> you should be able to successfully execute qpage as zenoss user now

[03-May-2010 15:10:08] <rmatte> if not you may have to try: zenoss ALL=(ALL) NOPASSWD: qpage

[03-May-2010 15:10:20] <rmatte> not sure how specific sudo is about exactly how you're executing it

[03-May-2010 15:10:32] <rmatte> although I'm pretty sure the full path will work fine

[03-May-2010 15:18:31] <voxter> here's a question, so i enabled fping graphs on a device i monitor right.. zenoss will then monitor that device's IP.  If i wanted, say, fping graphs to also be collected FROM that device (ie instead of the data coming from zenoss -> device, it goes from device -> other ips) i would set up "device" to be a collector and configure a bunch of stuff that way..

[03-May-2010 15:21:19] <rmatte> yes

[03-May-2010 15:21:29] <rmatte> provided the device is a linux box and can actually run as a zenoss collector

[03-May-2010 15:22:14] <voxter> correct ya. Thats integral for this smokeping thing i need to do, multiple viewpoints of the same data, to determine where problems might be.

[03-May-2010 15:22:17] <voxter> Thanks

[03-May-2010 15:22:25] <rmatte> np

[03-May-2010 15:23:04] <twm1010> rmatte: i can run qpage as zenoss user now, UI still gives me a failure, i'll pick this backup later, probably still a permissions issue somewhere

[03-May-2010 15:23:20] <rmatte> k

[03-May-2010 15:28:15] <twm1010> thanks for the help, as alwys

[03-May-2010 15:29:12] <rmatte> np

[03-May-2010 15:51:04] <rmatte> hmmm, I think I might know what caused this zProperties error... jane's jc_batch_load script

[03-May-2010 15:51:21] <rmatte> the 2 servers that I'm having the issue on both had devices loaded on using that script

[03-May-2010 15:53:55] <rmatte> yup, and there's one more server I used the script on, same problem

[03-May-2010 15:54:03] <rmatte> has to be related

[03-May-2010 15:55:02] <rmatte> yup, every device except for one that I added manually via the UI has the problem

[03-May-2010 15:55:15] <rmatte> looks like I'm going to have to check her script and see if I can figure out where it goes wrong

[03-May-2010 16:00:06] <mrayzenoss> I think zenbatchload has been updated since she wrote that script

[03-May-2010 16:00:24] <mrayzenoss> it's pretty useful now

[03-May-2010 16:03:23] <rmatte> ah

[03-May-2010 16:03:29] <rmatte> I'm just used to using that script lol

[03-May-2010 16:03:44] <rmatte> I'll have to check it out

[03-May-2010 16:06:08] ErikRose is now known as SlimyMarketer

[03-May-2010 16:07:31] <rmatte> yeh, I think I see what's happening with her script

[03-May-2010 16:07:46] <rmatte> it has a spot in the csv for zWinUser and zWinPassword values, but I always leave those blank when adding

[03-May-2010 16:07:56] <rmatte> zWinUser, zWinPassword = devicedata[17:19]

[03-May-2010 16:08:15] <rmatte> here's the part that's causing the issue

[03-May-2010 16:08:17] <rmatte> # Now for the bits from Scott's script - loadDevice doesn't handle Windows users / passwords...

[03-May-2010 16:08:17] <rmatte>       d=dmd.Devices.findDevice(deviceName)

[03-May-2010 16:08:17] <rmatte>       print "dmd device is ",d

[03-May-2010 16:08:17] <rmatte>       d.zWinUser=zWinUser

[03-May-2010 16:08:17] <rmatte>       d.zWinPassword=zWinPassword

[03-May-2010 16:08:37] <rmatte> I'll comment that out next time I run it

[03-May-2010 16:12:18] SlimyMarketer is now known as ErikRose

[03-May-2010 17:03:34] <rmatte> well, now that that's fixed my 2.5.2 installs are spotless

[03-May-2010 17:03:49] <rmatte> (well, after the path fix and removing the odd characters from device comments)

[03-May-2010 17:05:59] <mrayzenoss> are you going to start up a post-2.5.2 page like you did for 2.4.5?

[03-May-2010 17:07:13] <rmatte> Probably, though there aren't really patches, just little manual tweaks

[03-May-2010 17:07:45] <mrayzenoss> yeah, so far :)

[03-May-2010 17:07:47] <rmatte> that -l option for the start lines in the startup script needs to be included in the next version

[03-May-2010 17:08:02] <rmatte> to make sure that the path issue doesn't come back

[03-May-2010 17:10:34] <rmatte> man vmware really sucks sometimes

[03-May-2010 17:10:44] * rmatte waits half an hour for a snapshot to be deleted

[03-May-2010 17:11:03] <rmatte> I don't see how it takes half the time to take a snapshot than it does to delete one

[03-May-2010 17:16:10] <chemist> how do I get the sscpuidle and other grey messages to go away?

[03-May-2010 17:16:36] <rmatte> chemist: those are debug messages

[03-May-2010 17:16:47] <rmatte> it means it's not able to properly collect that value from that device

[03-May-2010 17:16:58] <rmatte> meaning either the snmp permissions are incorrect on the device

[03-May-2010 17:17:08] <rmatte> or the OID that it's trying to poll isn't supported

[03-May-2010 17:17:17] <rmatte> you getting those for your Linux servers?

[03-May-2010 17:17:29] <chemist> I know, I just wonder why I can see the value if I snmpwalk it

[03-May-2010 17:17:31] <chemist> yes

[03-May-2010 17:17:51] <rmatte> when you snmpwalk the value, do you see 1 line or more than 1?

[03-May-2010 17:18:03] <rmatte> (is this a custom template you made?)

[03-May-2010 17:18:26] <chemist> one line

[03-May-2010 17:18:34] <rmatte> paste?

[03-May-2010 17:19:00] <chemist> UCD-SNMP-MIB::ssCpuIdle.0 = INTEGER: 96

[03-May-2010 17:19:22] <rmatte> k... and the template is a custom one?

[03-May-2010 17:19:27] <chemist> no, vanilla zenoss

[03-May-2010 17:19:33] <rmatte> Linux template?

[03-May-2010 17:19:38] <chemist> yes

[03-May-2010 17:19:51] <chemist> net-snmp I believe

[03-May-2010 17:19:57] <rmatte> If you go to the template, select that datasource, add the device name in and click test, does it work?

[03-May-2010 17:21:12] <chemist> yes

[03-May-2010 17:21:29] <rmatte> k, does anything graph on the perf tab?

[03-May-2010 17:22:23] <chemist> yes, all 4 graphs

[03-May-2010 17:22:36] <rmatte> and you see data for ssCpu?

[03-May-2010 17:22:49] <rmatte> ssCpuIdle rather

[03-May-2010 17:23:17] <rmatte> (it should be graphed in light grey)

[03-May-2010 17:23:28] <chemist> I see the cpu idle time in the first graph

[03-May-2010 17:23:45] <chemist> sorry, second graph

[03-May-2010 17:23:49] <rmatte> k

[03-May-2010 17:23:57] <rmatte> then you shouldn't be receiving the debug message

[03-May-2010 17:24:04] <rmatte> have you tried clearing it to see if it comes back?

[03-May-2010 17:24:17] <chemist> that is what I thought

[03-May-2010 17:24:25] <chemist> yes I have

[03-May-2010 17:24:34] <rmatte> have you tried restarting Zenoss?

[03-May-2010 17:24:50] <rmatte> might be something screwy cached somewhere

[03-May-2010 17:25:31] <chemist> sorry, just realized the error of my ways

[03-May-2010 17:25:43] <rmatte> ?

[03-May-2010 17:25:57] <chemist> I was testing different snmpd.conf's on the device

[03-May-2010 17:26:05] <rmatte> ah

[03-May-2010 17:26:18] <chemist> one of them wouldn't allow that OID

[03-May-2010 17:26:39] <chemist> and I had not cleared the debugs since the last change

[03-May-2010 17:27:36] <chemist> can I add a datasource to display text instead ofgraphing it?

[03-May-2010 17:28:14] <rmatte> no, you can only graph the data

[03-May-2010 17:28:42] <rmatte> If you wanted to display text you'd need to code a custom skin template for the device page as well as a custom collector plugin

[03-May-2010 17:30:10] <chemist> I might just have to hack the status tab like someone suggested in teh forum

[03-May-2010 17:31:58] <rmatte> you could do that, not really the best idea though in terms of upgradeability

[03-May-2010 17:32:33] <rmatte> woohoo, all of my 2.5.2 installs are pristine

[03-May-2010 17:32:49] <rmatte> now I just need to figure out why my last server that needs to be upgraded won't do a zenmigrate properly

[03-May-2010 17:32:53] <rmatte> that's tomorrow fun project

[03-May-2010 17:32:58] <rmatte> tomorrow's*

[03-May-2010 17:33:19] <chemist> yes I know, my best option is to create a custom report, but documentation is scarce

[03-May-2010 17:33:32] <rmatte> well, what are you looking to display?

[03-May-2010 17:34:45] <chemist> I can't remeber the exact OID (not at work) but we run an in-house binary  and we need to know what version it is

[03-May-2010 17:34:56] <rmatte> I see

[03-May-2010 17:35:44] <rmatte> you might be able to get that to display in the software tab with some hackery

[03-May-2010 17:36:07] <mrayzenoss> I know themactech was working on something like that, I was trying to get him to write it up

[03-May-2010 17:36:18] <mrayzenoss> displaying a custom property on th status page

[03-May-2010 17:36:28] <rmatte> yeh, that would be a really nice feature to have

[03-May-2010 17:38:16] <mrayzenoss> maybe I'll get to that in my zenpack tutorial someday :0

[03-May-2010 17:38:28] <mrayzenoss> right now I'm trying to model memory and cpu on OSX

[03-May-2010 17:42:19] * rmatte is waiting for a snapshot to delete so that he can go home

[03-May-2010 17:42:40] <chemist> themactech posted on my thread in the forum how to do it

[03-May-2010 17:42:49] <rmatte> cool

[03-May-2010 17:43:13] <chemist> thread/12349?tstart=0

[03-May-2010 17:44:04] <chemist> I'm sure someone with more knowledge than me could use that info to create that function

[03-May-2010 17:44:12] <chemist> would make a good zenpack

[03-May-2010 17:45:11] <chemist> been able to display a range of OID's or other data that you don't want/need graphing

[03-May-2010 17:45:25] <chemist> but display a string instead

[03-May-2010 17:46:28] <rmatte> yeh

[03-May-2010 17:46:56] <rmatte> well, we'll wait and see how the new UI looks when it's finished and then they'll be open to adding stuff like that

[03-May-2010 17:47:20] <rmatte> though I'm still rooting for focus on reports

[03-May-2010 17:47:22] <chemist> the ability to view certain system info on the staus tab would make zenoss 100% what I need

[03-May-2010 17:47:40] <rmatte> the current reports system is really limited

[03-May-2010 17:47:49] <chemist> so I'm not alone on that either

[03-May-2010 17:47:56] <rmatte> no, you're not

[03-May-2010 17:48:13] <rmatte> we actually rigged up our own custom report scripts that run outside of Zenoss to generate our monthly reportsd

[03-May-2010 17:48:15] <rmatte> reports*

[03-May-2010 17:48:32] <rmatte> I've written some that do stuff like process availability, interface availability, etc...

[03-May-2010 17:48:41] <chemist> by interogatting zope externally?

[03-May-2010 17:48:52] <rmatte> well, I just use python to interface with it

[03-May-2010 17:48:53] <chemist> *interrogating

[03-May-2010 17:48:57] <rmatte> basically same as zendmd

[03-May-2010 17:49:23] <rmatte> then I dump it in csv format, and we have a php script that executes all the necessary scripts, grabs the data, parses it, and formats it

[03-May-2010 17:49:34] <chemist> nice

[03-May-2010 17:49:57] <rmatte> We do a disk usage reports, I added an extra columns for current threshold levels

[03-May-2010 17:50:04] <rmatte> so clients can see what the thresholds are set at at a glance

[03-May-2010 17:50:12] <rmatte> same for cpu, memory, etc...\

[03-May-2010 17:50:44] <rmatte> it'd be great if the majority of that was possible via the UI

[03-May-2010 17:51:14] <chemist> yes, specially for people like me (green on programming)

[03-May-2010 17:51:20] <rmatte> yeh

[03-May-2010 17:51:35] <rmatte> I wrote the python scripts that we use and my manager wrote the php

[03-May-2010 17:51:46] <rmatte> it works quite well, but it was a ton of work

[03-May-2010 17:53:20] <chemist> I can imagine

[03-May-2010 17:54:11] <rmatte> I have a page with the scripts on the community site, I need to update it since it has older versions of the scripts

[03-May-2010 17:54:37] <chemist> zenpack development might be too ambitious as my first python project

[03-May-2010 17:54:45] <rmatte> hehe

[03-May-2010 17:54:54] <rmatte> well, my first python projects were writing transforms

[03-May-2010 17:54:59] * chemist has started some python tutorials

[03-May-2010 17:55:07] <rmatte> after that I pretty much jumped straight to writing a zenoss daemon

[03-May-2010 17:55:08] <rmatte> lol

[03-May-2010 17:55:32] <rmatte> the daemons works great, couple of small bugs that I intend to fix in the next version

[03-May-2010 17:55:36] <rmatte> and the code is sloppier than hell

[03-May-2010 17:55:39] <rmatte> but it fits the bill

[03-May-2010 17:55:47] <rmatte> daemon*

[03-May-2010 17:55:54] <chemist> yes, I modified a couple of transforms

[03-May-2010 17:56:05] <rmatte> http://dmon.org/zenticket.html

[03-May-2010 17:56:38] <rmatte> that's when I really delved deep in to python for the first time

[03-May-2010 17:57:47] <chemist> what ticketing system?

[03-May-2010 17:58:00] <chemist> otrs? rt?

[03-May-2010 17:58:30] <rmatte> otrs

[03-May-2010 17:58:45] <rmatte> a heavily customized version of otrs

[03-May-2010 17:59:27] <rmatte> We're upgrading to the latest version soon, I was the one who ported all the customizations over from the old version to the new

[03-May-2010 17:59:32] <rmatte> also had to make some mods of my own

[03-May-2010 17:59:49] <chemist> they must pay you well :)

[03-May-2010 17:59:58] <rmatte> funny too since it's all written in perl and dotiac, and I don't really know either of those

[03-May-2010 18:00:10] <rmatte> nah, I actually get paid crap right now for what I do

[03-May-2010 18:00:22] <rmatte> still making the same as I was when I was a network analyst here

[03-May-2010 18:00:32] <rmatte> my 1 year review is coming up very soon though

[03-May-2010 18:00:32] <chemist> make a nice cv though

[03-May-2010 18:00:41] <rmatte> yeh, my cv is quite nice

[03-May-2010 18:01:17] <chemist> I use a transform to suppress events if certain device is down

[03-May-2010 18:01:25] <rmatte> I'm responsible for 13 Zenoss servers (more on the way), OTRS, and ZipTie (config backup system)

[03-May-2010 18:01:41] <rmatte> cool

[03-May-2010 18:01:49] <rmatte> the sky is the limit with transforms

[03-May-2010 18:02:03] <chemist> but it's random on it's success

[03-May-2010 18:02:18] <rmatte> hmmm

[03-May-2010 18:02:22] <chemist> some times it works and I only get one event

[03-May-2010 18:02:35] <chemist> sometimes it doesnt and I get flooded

[03-May-2010 18:02:42] <rmatte> well, it would really depend on whether the events come in before the transform is run or after

[03-May-2010 18:03:00] <rmatte> if the down event comes in, the transform runs, finishes, then the other events flood in it's not going to accomplish anything

[03-May-2010 18:03:24] <rmatte> actually sorry, the exact opposite

[03-May-2010 18:03:29] <rmatte> if the events come in before the down event

[03-May-2010 18:03:40] <rmatte> actually, no, I was the right the first time

[03-May-2010 18:03:48] <rmatte> but the whole thing is inherently faulty

[03-May-2010 18:04:05] <rmatte> there has to be a better way to do that

[03-May-2010 18:04:27] <chemist> proper dependencies would be a good start

[03-May-2010 18:04:38] <rmatte> yeh, that's another HUGE one on my list

[03-May-2010 18:04:47] <rmatte> right alongside reports

[03-May-2010 18:06:53] <rmatte> man, the memory usage on 2.5.2 is infinitely better than it was on 2.4.5

[03-May-2010 18:07:18] <rmatte> I'm only up to 2GB usage on my largest install after over 5 days, on 2.4.5 I'd be up to 3.5GB

[03-May-2010 18:07:31] <chemist> yes, those 2 and the custom data on the status tab would make zenoss not the best, but the perfect tool

[03-May-2010 18:08:05] <chemist> how do you deal with? a sheduled restart of zenoss?

[03-May-2010 18:08:19] <rmatte> "zenoss restart" lol

[03-May-2010 18:08:31] <chemist> hehe

[03-May-2010 18:08:48] <chemist> manually or cron'd?

[03-May-2010 18:08:52] <rmatte> manually

[03-May-2010 18:09:00] <rmatte> why would you want to cron it?

[03-May-2010 18:10:57] <rmatte> hmmm, the google map isn't working properly on my largest install after the upgrade, figures...

[03-May-2010 18:11:05] <chemist> I just find zenoss gets sluggish after 4-5 days

[03-May-2010 18:11:33] <chemist> so I have it restart at 4 am

[03-May-2010 18:11:54] <chemist> we are only mission critical at certain times

[03-May-2010 18:12:09] <rmatte> I see

[03-May-2010 18:12:16] <rmatte> yeh, I don't do automated restarts

[03-May-2010 18:17:15] <rhettardo> im getting all these alerts and i go check the event history and its blank

[03-May-2010 18:18:30] <rhettardo> i go click on event details in the emails and i get an error :(

[03-May-2010 18:18:34] <rhettardo> wtf

[03-May-2010 18:19:10] <chemist> rhettardo: when you click on the event history link it's empty?

[03-May-2010 18:19:17] <rhettardo> yes

[03-May-2010 18:19:26] <chemist> 2.5.2?

[03-May-2010 18:19:38] <rhettardo> coyeah

[03-May-2010 18:20:24] <rhettardo> yeah...

[03-May-2010 18:20:36] <rhettardo> damn cat was in front of the monitor and his some keys...

[03-May-2010 18:20:56] <chemist> Is there a date in the filter toolbar?

[03-May-2010 18:21:42] <chemist> I found that changing the date to earlier than today gave me the historic events

[03-May-2010 18:22:18] <chemist> "Last seen" date

[03-May-2010 18:23:50] <chemist> about the screwy event details, that is beyond me

[03-May-2010 18:25:48] <chemist> and the cat, get it it's own computer

[03-May-2010 18:25:57] <chemist> :)

[03-May-2010 18:27:53] <rhettardo> i have a first seen and last seen date in (yesterday and today) and its still blank

[03-May-2010 18:28:41] <chemist> is mysql running?

[03-May-2010 18:29:17] <rhettardo> yup

[03-May-2010 18:31:08] <rhettardo> im in the history table right now and sorted by date and i dont even see any of alerts in there im getting sent

[03-May-2010 18:32:00] <rhettardo> this looks like its going to be fun

[03-May-2010 18:32:17] <chemist> I guess you tried a restart?

[03-May-2010 18:48:54] <voxter> Is there a defacto guide i should be referencing for setting up a separate zenoss collector, or would this do it?

[03-May-2010 18:48:55] <voxter> docs/DOC-2496

[03-May-2010 18:50:17] <rmatte> voxter: oh, one sec

[03-May-2010 18:50:37] <rmatte> voila: docs/DOC-5861

[03-May-2010 18:51:04] <rmatte> I sent a collector setup ZenPack that I had made to Egor and he made one that's 100x better

[03-May-2010 18:51:19] <voxter> aha, when it started telling me to edit config files, i figured someone had done this work for me in a more efficient way ;)

[03-May-2010 18:51:33] <voxter> love zenpacks.

[03-May-2010 18:51:55] <rmatte> well, you're lucky actually, me and Egor only released our packs quite recently

[03-May-2010 18:52:08] <rmatte> mine has only been out for a few weeks and Egor's has been out for a few days hehe

[03-May-2010 18:52:43] <rmatte> Egor's is awesome because it automatically rsyncs the zenpack data as well as RRD data when devices are moved

[03-May-2010 18:52:55] <voxter> nice!

[03-May-2010 18:53:01] <rmatte> plus his has a UI component, mine was all commandline

[03-May-2010 18:53:07] <voxter> any other amazingly awesome zenpacks that you can suggest?

[03-May-2010 18:53:11] <rmatte> I think I'm just going to have Matt mark my pack is obsolete

[03-May-2010 18:53:23] <rmatte> well, it depends on what you're looking to do

[03-May-2010 18:53:38] <rmatte> Egor's WMI Data Source pack allows you to fully model windows servers via WMI

[03-May-2010 18:53:49] <rmatte> which is neat since it was previously an enterprise-only feature

[03-May-2010 18:54:03] <voxter> oh nice. i was just about to ask about modelling a couple windows servers!

[03-May-2010 18:54:06] <voxter> I'll check that out too

[03-May-2010 18:54:11] <rmatte> The predictive threshold pack can be nice too

[03-May-2010 18:54:20] <voxter> im eagerly waiting for the Xen zenpack that these university students are doing for their thesis

[03-May-2010 18:54:23] <voxter> it sounds pretty awesome.

[03-May-2010 18:54:31] <rmatte> yup

[03-May-2010 18:54:31] <voxter> I am using the one thats available now but arguably its pretty basic

[03-May-2010 18:54:42] <voxter> predictive threshold.. nice.. thats pretty important

[03-May-2010 18:54:47] <rmatte> actually, if you want to monitor Xen for the time being, the libvirt pack is great for that

[03-May-2010 18:55:10] <voxter> I was thinking about that, if you dont go and tell zenoss all your thresholds, you dont get useful alerts.

[03-May-2010 18:55:16] <rmatte> oh, and if you prefer to monitor windows via SNMP (as I do) you can use my packs...

[03-May-2010 18:55:17] <rmatte> docs/DOC-3386

[03-May-2010 18:55:21] <voxter> and at that, your thresholds might be complete guesses

[03-May-2010 18:55:22] <rmatte> docs/DOC-3570

[03-May-2010 18:55:41] <voxter> for example I want to be alerted if one ethernet link exceeds 4mbit which wouldnt be normal for people

[03-May-2010 18:55:55] <voxter> I'm curious about tresholds that are like... alert if this changes more than 40% than normal

[03-May-2010 18:55:56] <rmatte> yeh

[03-May-2010 18:56:02] <voxter> i guess thats what predictive thresholds do

[03-May-2010 18:56:13] <rmatte> well, sort of, I actually don't fully understand them yet

[03-May-2010 18:56:18] <rmatte> I'm going to look in to them more soon

[03-May-2010 18:56:34] <voxter> Im glad i checked zenoss out again. I tried it about 3 years ago and gave up immediately due to its structure complexity

[03-May-2010 18:56:42] <voxter> i just couldnt wrap my head around it in comparison to cacti and nagios

[03-May-2010 18:56:45] <rmatte> but yeh, there are lots of great community zenpacks out there

[03-May-2010 18:56:49] <voxter> but this time has been much more intuitive

[03-May-2010 18:56:55] <voxter> I think the zenpacks help alot.

[03-May-2010 18:57:08] <rmatte> some of them are really low quality though, make sure you try them out on a lab box or something before going prod with them

[03-May-2010 18:57:12] <rmatte> to make sure it's satisfactory

[03-May-2010 18:57:36] <rmatte> well, the documentation has gotten a lot better

[03-May-2010 18:57:41] <rmatte> and the community is a lot more active

[03-May-2010 18:57:47] <voxter> i just find that poking around in the UI makes more sense

[03-May-2010 18:57:52] <rmatte> yeh

[03-May-2010 18:57:58] <rmatte> well, wait until the revamped UI lol

[03-May-2010 18:58:08] <voxter> before you kinda had to know exactly how the structure worked before anything made sense

[03-May-2010 18:58:16] <rmatte> I know I'm going to spend a week just finding where they've stuck everything

[03-May-2010 18:58:33] <voxter> is there a demo/screenshots of the new UI? is it intended to make things more powerful or easier to navigate?

[03-May-2010 18:59:03] <rmatte> Are you going to be monitoring lots of Cisco stuff?

[03-May-2010 18:59:06] <voxter> What id like to see more of, is a tiny usage document along with each zenpack, for people starting out, its helpful

[03-May-2010 18:59:08] <rmatte> oh, yeh there is

[03-May-2010 18:59:09] <rmatte> one sec

[03-May-2010 18:59:23] <rmatte> http://public-demo.zenoss.com/

[03-May-2010 18:59:27] <rmatte> login with admin/zenoss

[03-May-2010 18:59:30] <voxter> For example, when i installed the fping zenpack, i didnt quite realize that it would be made available as a template, but not bound, and that i had to do that

[03-May-2010 18:59:44] <rmatte> yeh

[03-May-2010 18:59:48] <voxter> thats something you come to expect once youve done it 10 times, but

[03-May-2010 18:59:49] <rmatte> just part of the learning process

[03-May-2010 18:59:55] <rmatte> I've been working with Zenoss for just over a year

[03-May-2010 19:00:01] <voxter> oh yeah.... no left menu

[03-May-2010 19:00:06] <rmatte> yup

[03-May-2010 19:00:15] <voxter> wow. it looks great!

[03-May-2010 19:00:21] <rmatte> yeh, it looks really good

[03-May-2010 19:00:26] <rmatte> I personally can't wait

[03-May-2010 19:00:36] <voxter> i like it when interfaces for this purpose are arranged somewhat like a very interactive excel workbook

[03-May-2010 19:00:47] <rmatte> aha, they finally started working on the device status page

[03-May-2010 19:00:57] <voxter> yeah thats what im looking at here.

[03-May-2010 19:01:28] <voxter> i dunno if groups are a new feature altogether, but thats great how they're laid out

[03-May-2010 19:01:45] <voxter> Locations, Devices, fantastic

[03-May-2010 19:02:08] <rmatte> nah, that's in your current version

[03-May-2010 19:02:15] <rmatte> Groups, Locations, Systems, Devices

[03-May-2010 19:02:23] <rmatte> all different ways to categorize devices

[03-May-2010 19:02:46] <voxter> oh, yeah, i see it.

[03-May-2010 19:02:49] <rmatte> ;)

[03-May-2010 19:02:53] <voxter> the UI is just much much much nicer in the new version

[03-May-2010 19:03:00] <voxter> I suppose its probably dangerous to run that in production ;)

[03-May-2010 19:03:02] <rmatte> the new UI is mostly the same stuff, just new look and organization

[03-May-2010 19:03:11] <rmatte> at the moment, probably

[03-May-2010 19:03:21] <rmatte> they have an alpha version to download and mess around with

[03-May-2010 19:03:52] <rmatte> I just finished cleaning up the mess from my King Crab upgrades, I'm in no rush to go through that again

[03-May-2010 19:03:53] <rmatte> ;)

[03-May-2010 19:04:24] <rmatte> they are going to be moving to Python 2.6 and a newer version of Zope in that version

[03-May-2010 19:07:29] <rmatte> and yeh, back to what you were saying about usage documents, we're trying to improve on the quality of them

[03-May-2010 19:08:12] * rmatte thinks he should rebuild his Cisco Mibs ZenPack soon

[03-May-2010 19:23:17] <voxter> sorry one sec, phone!

[03-May-2010 19:34:58] <voxter> sorry back now..

[03-May-2010 19:35:15] <voxter> I use a decent amount of cisco stuff yeah. A company i consult for uses all cisco 6500 and 3500 stuff throughout their whole network

[03-May-2010 20:14:58] <etank> rmatte: is that the new zenoss?

[03-May-2010 20:23:21] <mrayzenoss> is what the new Zenoss?

[03-May-2010 20:24:12] <etank> mrayzenoss: rmatte pasted a link a little while ago

[03-May-2010 20:24:32] <etank> to http://public-demo.zenoss.com/

[03-May-2010 20:24:43] <etank> wondering if that is the new look

[03-May-2010 20:24:43] <mrayzenoss> yeah, that's it

[03-May-2010 20:24:50] <mrayzenoss> admin/zenoss is the login

[03-May-2010 20:25:04] <etank> yeah i checked it out.

[03-May-2010 20:25:09] <etank> looks pretty rockin

[03-May-2010 20:30:58] <MTecknology> is zenoss primarily system monitoring?

[03-May-2010 20:31:28] <mrayzenoss> yes

[03-May-2010 20:31:35] <mrayzenoss> network devices, servers, apps

[03-May-2010 20:31:58] <mrayzenoss> SNMP, SSH, WMI, JMX, etc.

[03-May-2010 20:32:07] <MTecknology> what's different between nagios and zenoss?

[03-May-2010 20:33:24] <mrayzenoss> Nagios is primarily availability montiroing

[03-May-2010 20:33:34] <mrayzenoss> so you're probably using Cactii for performance monitoring

[03-May-2010 20:33:40] <MTecknology> I was originally looking at landscape and was thrown off by the huge price tag; zenoss and nagios look like the only real alternatives

[03-May-2010 20:33:51] <mrayzenoss> Zenoss does both, plus a lot more

[03-May-2010 20:34:10] <mrayzenoss> I gotta run, check this tomorrow: blogs/zenossblog/2010/05/03/getting-started-with-zenoss-core-webinar-may-4

[03-May-2010 20:34:19] <MTecknology> probably a stretch, does it do any system management?

[03-May-2010 20:34:30] <MTecknology> ok- thanks

[03-May-2010 20:35:39] <etank> MTecknology: what kind of system management do you have in mind

[03-May-2010 20:36:52] <MTecknology> etank: landscape lets you do about everything you could want on an ubuntu system (ubuntu specific) - it's like a gui way to kill a task, install updates, install a new package, reboot a cluster of servers, etc.

[03-May-2010 20:37:03] <etank> aah

[03-May-2010 20:37:18] <MTecknology> etank: for the most part - I care very little- I'm just learning and asking

[03-May-2010 20:37:22] <etank> so kind of like spacewalk or rhn for centos / red hat systems

[03-May-2010 20:38:09] <etank> if you were meaning reacting to events and performing system tasks then zenoss can do that

[03-May-2010 20:38:15] <MTecknology> ya, I think so

[03-May-2010 20:38:30] <MTecknology> the spacewalk part

[03-May-2010 20:38:46] <etank> i.e. we have a system that we are monitoring. if a certain service stops that we care about zenoss restarts it for us

[03-May-2010 20:38:57] <MTecknology> cool

[03-May-2010 20:39:18] <etank> we just have a shell script that it uses to remote to the box and start it back up

[03-May-2010 20:39:28] <etank> it happens almost as soon as it goes down

[03-May-2010 20:39:54] <MTecknology> once the healthcheck runs?

[03-May-2010 20:39:58] <MTecknology> or something like that

[03-May-2010 20:40:01] <etank> as far as updates and stuff though i dont think it could push then

[03-May-2010 20:40:15] <etank> yeah. once every 5 minutes it does an snmp walk.

[03-May-2010 20:40:22] <etank> or maybe that is once a minute

[03-May-2010 20:40:29] <MTecknology> landscape is 5min

[03-May-2010 20:40:45] <etank> either way. as soon as it sees the stopped daemon it restarts it for us

[03-May-2010 20:40:53] <MTecknology> nice

[03-May-2010 20:41:21] <MTecknology> what's different with ose?

[03-May-2010 20:41:25] <etank> for patches and stuff i dont see it doing it

[03-May-2010 20:41:35] <etank> support basically

[03-May-2010 20:41:44] <etank> you dont get the enterprise zenpacks

[03-May-2010 20:41:45] <MTecknology> nice

[03-May-2010 20:41:55] <etank> but there are a ton of nice community packs available

[03-May-2010 20:42:20] <MTecknology> so.. get an extra server and install the daemon on each client?

[03-May-2010 20:42:31] <etank> no daemons

[03-May-2010 20:42:34] <etank> all snmp

[03-May-2010 20:42:37] <etank> or ssh

[03-May-2010 20:42:43] <etank> wmi for windows

[03-May-2010 20:42:49] <MTecknology> cool

[03-May-2010 20:43:11] <MTecknology> I'm going to have to try this out :D

[03-May-2010 20:43:13] <MTecknology> thanks :D

[03-May-2010 20:43:19] <MTecknology> gotta run - bbiab

[03-May-2010 20:43:28] <etank> the best way to see what it can do is join up for the seminar that matt talked about

[03-May-2010 20:43:36] <etank> or dl it and give it a go

[03-May-2010 20:54:49] <MTecknology> etank: sorry, something pretty urgent came up

[03-May-2010 20:55:33] <MTecknology> etank: I looked at the tco for landscape and that's in the neighborhood of $4k/yr :P

[03-May-2010 20:57:34] <MTecknology> etank: so every 5min the system, goes out, checks a system for healthstatus and various other info, then if something happens it can trigger on that?

[03-May-2010 20:58:02] <MTecknology> and it keeps pretty reports of the system mem/hd/cpu usage and uptime?

[03-May-2010 20:58:31] <etank> yeah pretty much

[03-May-2010 20:58:42] <etank> you can adjust it to make it check more often

[03-May-2010 20:58:52] <etank> and even set escalation on events

[03-May-2010 20:58:56] <MTecknology> 5min sounds pretty good

[03-May-2010 20:58:58] <MTecknology> hm?

[03-May-2010 20:59:02] <etank> and it does deduplication

[03-May-2010 20:59:29] <etank> so if an event comes in you get one email instead of once per each time the same event is seen

[03-May-2010 20:59:45] <MTecknology> nice

[03-May-2010 20:59:45] <etank> which is nice

[03-May-2010 20:59:48] <MTecknology> what's set escalation?

[03-May-2010 21:00:17] <etank> for instance. if an event comes in it sets a dedupe id and you get an email

[03-May-2010 21:00:33] <etank> when the event happens again it incriments the count on the event by 1

[03-May-2010 21:00:49] <etank> at first the event is set to a Warning

[03-May-2010 21:01:09] <MTecknology> so when you get around to looking at it, you see the count instead of a lot of the same errors?

[03-May-2010 21:01:17] <etank> you can set a rule that says if the count reaches a certain level then the event gets bumped to an Error

[03-May-2010 21:01:22] <etank> yes

[03-May-2010 21:01:26] <MTecknology> cool

[03-May-2010 21:01:37] <etank> keeps the email count down

[03-May-2010 21:02:00] <etank> and you can tell it to only alert after a certain time window

[03-May-2010 21:02:10] <etank> that keeps flapping reports to a minimum too

[03-May-2010 21:02:22] <MTecknology> can you tell it to send an arbitrary command to the server? when it runs?

[03-May-2010 21:02:52] <MTecknology> s/r?/r/

[03-May-2010 21:02:58] <etank> yeah you should be able to create a template and have it do that

[03-May-2010 21:03:35] <etank> and then you can set a timing on that template so that it happens every 5 mintues 20 mintues what ever

[03-May-2010 21:03:56] <MTecknology> so, aptitude update && yes y | aptitude full-upgrade ? :P

[03-May-2010 21:04:09] <MTecknology> I don't think I'd actually want to do it.. but..

[03-May-2010 21:04:21] <etank> yeah i wouldnt want to do that

[03-May-2010 21:04:51] <etank> you may be able to get an update count (like what byobu reports) and graph the numbers

[03-May-2010 21:05:05] <etank> if the graph is up then you have updates to apply

[03-May-2010 21:05:15] <etank> but you wouldnt know what they were really

[03-May-2010 21:05:39] <MTecknology> that's plenty of info for me

[03-May-2010 21:06:12] <etank> hope it helped some

[03-May-2010 21:06:13] <MTecknology> how secure is snmp?

[03-May-2010 21:06:27] <etank> snmp v3 can be encrypted

[03-May-2010 21:06:35] <MTecknology> I meant just seeing update count is plenty

[03-May-2010 21:06:42] <etank> but there was a bug in zenoss with snmp3 with the last release

[03-May-2010 21:06:56] <MTecknology> personally, I'd probably run it over ssh, just comfort :P

[03-May-2010 21:07:04] <MTecknology> I never dealt with snmp

[03-May-2010 21:07:06] <etank> and with snmp you can tell it to only talk to certain systems

[03-May-2010 21:08:02] <MTecknology> you mean like a group of systems?

[03-May-2010 21:08:07] <etank> i think that with ssh you would need to copy some zenoss scripts to the machine being monitored

[03-May-2010 21:08:21] <etank> so that zenoss has something to run on that box

[03-May-2010 21:08:44] <etank> well you can tell a snmp client to only allow a certain IP to make calls against it

[03-May-2010 21:09:17] <etank> and you set a community string that is not common

[03-May-2010 21:09:23] <MTecknology> oh

[03-May-2010 21:09:25] <etank> public is the common string name

[03-May-2010 21:09:34] <etank> we use something at work that is specific to our team

[03-May-2010 21:10:04] <MTecknology> is it like a password kinda thing?

[03-May-2010 21:10:15] <etank> kinda

[03-May-2010 21:10:29] <MTecknology> Then would something like 'kXD6B8sZpXQ5bWHQ5O8vLneCGaPCvPucdMSN38WC5pi0' work? :P

[03-May-2010 21:10:42] <etank> but you would prolly use the same string on every client

[03-May-2010 21:10:51] <etank> that make the roll out easier

[03-May-2010 21:11:09] <MTecknology> I use ufw for managing access usually

[03-May-2010 21:11:27] <etank> aah. just make sure you open port 161 for snmp then

[03-May-2010 21:11:36] <MTecknology> in addition to firewall rules - usually duplicate entries

[03-May-2010 21:11:38] <etank> udp 161 i think is the right port at least

[03-May-2010 21:12:28] <MTecknology> so install snmp on clients, setup string, setup access rules, configure zenoss, and bliss?

[03-May-2010 21:12:52] <etank> check out blogs/zenossblog/2010/05/03/getting-started-with-zenoss-core-webinar-may-4 tomorrow if you get a chance

[03-May-2010 21:13:04] <etank> it will help make everything make a lot more sense

[03-May-2010 21:13:17] <etank> pretty much though

[03-May-2010 21:13:33] <MTecknology> I'm planning on doing my best to make it to that

[03-May-2010 21:13:47] <etank> it is well worth it :)

[03-May-2010 21:13:54] <MTecknology> EDT is?

[03-May-2010 21:14:04] <etank> MTecknology: you are in Utah right?

[03-May-2010 21:14:08] <etank> gmt -5

[03-May-2010 21:14:09] <MTecknology> nope

[03-May-2010 21:14:34] <etank> hmm. you are the Ubuntu forums guy ?

[03-May-2010 21:14:38] <MTecknology> South Dakota

[03-May-2010 21:14:43] <etank> aah. ok

[03-May-2010 21:14:50] <MTecknology> I'm in Ubuntu Forums, ya

[03-May-2010 21:14:54] <etank> thought so

[03-May-2010 21:15:05] <MTecknology> OH!

[03-May-2010 21:15:07] <MTecknology> YOU!

[03-May-2010 21:15:07] <etank> seen your name a lot on the planet and forums

[03-May-2010 21:15:19] * etank

[03-May-2010 21:15:22] <MTecknology> all of a sudden you sond familiar

[03-May-2010 21:15:40] <etank> hope that is a good thing

[03-May-2010 21:16:00] <MTecknology> i don't recall why I remember you clearly - means you didn't make me hate you :P

[03-May-2010 21:16:23] <MTecknology> I don't remember why you sound so familiar now though

[03-May-2010 21:16:31] <etank> hmm. you set up the ubuntu member thing on my forum account ?

[03-May-2010 21:16:39] <etank> im from the Kentucky team

[03-May-2010 21:16:54] <MTecknology> i'm not a forum admin

[03-May-2010 21:17:04] <etank> hung out in the ubuntu-us chan forever

[03-May-2010 21:17:16] <MTecknology> that's probably where

[03-May-2010 21:17:20] <etank> ok

[03-May-2010 21:17:28] <etank> knew it was one of those things :)

[03-May-2010 21:17:31] <MTecknology> ya

[03-May-2010 21:17:44] <MTecknology> alrighty - I can sleep at ease tonight :D

[03-May-2010 21:18:41] <etank> the Utah thing ... i was thinking of Technoviking

[03-May-2010 21:18:46] <MTecknology> oh

[03-May-2010 21:18:48] <etank> he set up my ubuntu member thing

[03-May-2010 21:18:53] <MTecknology> what is 10am EDT in UTC?

[03-May-2010 21:18:55] <etank> but i do remember you too

[03-May-2010 21:19:04] <MTecknology> or - CST

[03-May-2010 21:19:17] <etank> that is 3PM UTC

[03-May-2010 21:20:14] <etank> maybe 2PM with the time change

[03-May-2010 21:20:55] <MTecknology> I like using UTC for things so much more :P

[03-May-2010 21:21:02] <etank> yeah 2PM

[03-May-2010 21:21:09] <etank> me too

[03-May-2010 21:21:21] <etank> took a while to get used to it but now that i am it is easier

[03-May-2010 21:22:00] <MTecknology> all that to figure out I'm 1hr behind EDT :P

[03-May-2010 21:23:03] <MTecknology> how long do the sessions usually last?

[03-May-2010 21:23:10] <etank> i think an hour

[03-May-2010 21:24:59] <MTecknology> heh.. may have found an ugly thing in the signup

[03-May-2010 21:25:12] <MTecknology> I have one webpage embedded in the same

[03-May-2010 21:25:31] <MTecknology> the homepage is inside the webinars page :P

[03-May-2010 21:25:42] <MTecknology> the topic is wrong..

[03-May-2010 21:26:27] <MTecknology> 3.0 aplha is released.

[03-May-2010 21:26:36] <MTecknology> or is that 3.0-alpha2?

[03-May-2010 21:27:12] <MTecknology> etank: hi, welcome to my twitter page, i'm more interesting on identi.ca

[03-May-2010 21:27:56] <etank> MTecknology: cool. i have not used identi.ca in a while

[03-May-2010 21:28:17] <etank> i need a good integrated client (twitter, identi.ca, facebook)

[03-May-2010 21:28:20] <MTecknology> i like it, except for their jabber users disappearing

[03-May-2010 21:28:28] <MTecknology> I just use bitlbee

[03-May-2010 21:30:11] <MTecknology> well... the #1 thing I care about is what the system offers; #2 is the community

[03-May-2010 21:30:18] <MTecknology> #3 is the price poitn

[03-May-2010 21:30:38] <MTecknology> so far, I don't see any reason this doesn't hit every one of those marks.

[03-May-2010 21:31:49] <MTecknology> looks very active too

[03-May-2010 21:32:06] <MTecknology> etank: btw - you were extremely helpful - thanks :D

[03-May-2010 21:32:23] <etank> MTecknology: any time

[03-May-2010 21:32:56] <etank> if the need is there too the enterprise product is very fairly priced

[03-May-2010 21:33:11] <etank> lots of helpful people in here too

[03-May-2010 21:33:20] <MTecknology> I'm looking at the cost of that

[03-May-2010 21:33:57] <etank> im sure it is different for every company but i think we got a very fair deal

[03-May-2010 21:34:05] <etank> if it is something that you need

[03-May-2010 21:34:50] <MTecknology> what would i expect for say 25 nodes?

[03-May-2010 21:35:17] <MTecknology> here we go

[03-May-2010 21:35:18] <MTecknology> http://www.zenoss.com/product/pricing

[03-May-2010 21:36:12] <MTecknology> that confuses me - the Min # resources?

[03-May-2010 21:36:21] <etank> hmm. i think there is a minimum

[03-May-2010 21:36:48] <etank> the min was lower in the past

[03-May-2010 21:36:55] <etank> i think we got it at 150 nodes

[03-May-2010 21:37:00] <MTecknology> does it mean you need to have at least 250 resources that you're paying for and you pay $100 each?

[03-May-2010 21:37:06] <etank> yeah

[03-May-2010 21:37:19] <MTecknology> so $25,000 min per year

[03-May-2010 21:37:36] <etank> sounds about right

[03-May-2010 21:37:50] <MTecknology> pricey

[03-May-2010 21:37:58] <etank> that is actually pretty cheap compaired to the competition

[03-May-2010 21:38:12] <etank> if you have 25 nodes then i would go with the community edition

[03-May-2010 21:38:34] <MTecknology> landscape is $150/node/year - but then there's no foss option

[03-May-2010 21:39:06] <MTecknology> the foss version sounds excellent though

[03-May-2010 21:40:57] <MTecknology> and when you get to that point the cost seems pretty low

[03-May-2010 21:41:14] <MTecknology> i guess the issue is getting to the point where you can justify that cost

[03-May-2010 21:42:38] <MTecknology> i mean if you're going to go that route - i likely will at some point.

[03-May-2010 21:44:20] <MTecknology> etank: now....... I need to figure out the physical system to host things on

[03-May-2010 21:46:50] <etank> check out people/ckrough/blog/2010/02/09/performance-tuning-for-zenoss-storage

[03-May-2010 21:47:09] <etank> im not sure if an ubuntu host is officially supported at this point

[03-May-2010 21:47:53] <etank> i think i saw that it is something being worked on

[03-May-2010 21:49:01] <MTecknology> you mean putting zenoss on ubuntu?

[03-May-2010 21:49:32] <etank> yeah. as the server running zenoss.

[03-May-2010 21:49:41] <etank> it is possible to do.

[03-May-2010 21:49:48] <etank> it is the first way i ran zenoss

[03-May-2010 21:50:02] <etank> but if you go enterprise at some point ...

[03-May-2010 21:50:16] <MTecknology> does it run decent on debian?

[03-May-2010 21:50:24] <etank> yeah

[03-May-2010 21:50:27] <etank> runs fine

[03-May-2010 21:50:42] <MTecknology> Memory

[03-May-2010 21:50:43] <MTecknology> Up to 1TB1, ECC DDR3

[03-May-2010 21:50:49] <MTecknology> JEBUS!

[03-May-2010 21:51:16] <etank> i think i have 4GB ram in mine and it is monitoring 65 nodes or so at this point

[03-May-2010 21:51:39] <etank> watch the disk I/O though for the rrd files

[03-May-2010 21:51:47] <MTecknology> I was looking at a server

[03-May-2010 21:52:03] <etank> heh. that is my server :)

[03-May-2010 21:52:05] <etank> for now.

[03-May-2010 21:52:25] <etank> im going to upgrade it soon i think

[03-May-2010 21:52:48] <MTecknology> is it just a web ui as far as zenoss is managed? aside from maybe hopping in cli once in a while?

[03-May-2010 21:53:06] <MTecknology> I can't imagine 1TB RAM

[03-May-2010 21:53:08] <etank> mostly web yeah

[03-May-2010 21:53:57] <etank> at least for me

[03-May-2010 21:59:10] <etank> thats it for me tonight.

[03-May-2010 21:59:25] <etank> night MTecknology. good luck with zenoss. good chatting with you.

[03-May-2010 22:00:12] <MTecknology> etank: aight, thanks - and thanks for the info

[03-May-2010 23:25:16] HCoyote_ is now known as HCoyote

[04-May-2010 00:00:30] [disconnected at Tue May  4 00:00:30 2010]

[04-May-2010 00:00:30] [connected at Tue May  4 00:00:30 2010]

[04-May-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[04-May-2010 00:04:29] <voxter> so weird, my fping just suddenly stopped collecting data... ?!

[04-May-2010 00:24:47] <voxter> If something suddenly stops recording data, how do i debug it/

[04-May-2010 04:28:27] <voxter> wtf is zenoss doing, i just installed it and its sitting at "Zenoss installation complete" and refused to return me to a prompt.

[04-May-2010 04:56:08] <fragfutter> voxter: on initial startup it needs to byte compile a lot of python files and it needs to populate the zodb

[04-May-2010 04:56:23] <voxter> I ctrl-c'ed it. it seems to be okay.

[04-May-2010 04:56:32] <voxter> It sat there doing nothing for a good 15 minutes.

[04-May-2010 04:56:41] <voxter> after completing the zodb import

[04-May-2010 04:57:04] <fragfutter> voxter: did you check if it was doing cpu or disk i/o?

[04-May-2010 04:57:39] <voxter> fragfutter: nope, indeed i did not.

[04-May-2010 04:58:02] <voxter> its possible i pissed it off, but so far it seems to be happy about what its doing

[04-May-2010 04:58:13] <voxter> its a remote collector so if somehow its screwed im not overly concerned, i can redo it

[04-May-2010 05:00:50] <fragfutter> i would guess on byte compiling files. this would fix itself

[04-May-2010 05:06:30] <voxter> Could have used a status other than "installation completed" to let me know it was still doing something :)

[04-May-2010 05:07:23] <fragfutter> you mean a progress bar? running from 0-99% in five seconds and then working there for 20minutes?

[04-May-2010 05:09:31] <voxter> no i mean like

[04-May-2010 05:09:39] <voxter> "Byte compiling, please wait..."

[04-May-2010 05:09:47] <voxter> not "Zenoss Installation Completed."

[04-May-2010 05:10:23] <fragfutter> (i was only guessing what it was doing)

[04-May-2010 05:10:37] <fragfutter> i know installers with a "completing installation" and a progress bar.

[04-May-2010 05:10:59] <voxter> Yeah, i see what you're saying. I suspect it wasnt actually doing anything at all.

[04-May-2010 07:07:44] * fragfutter checks the watch. yes its later.

[04-May-2010 07:58:34] * MTecknology is excited for the seminar

[04-May-2010 07:58:54] * MTecknology needs to get done with his final before it starts

[04-May-2010 07:59:27] <mrayzenoss> there's video of a previous session here: docs/DOC-2606

[04-May-2010 07:59:45] <MTecknology> thanks

[04-May-2010 08:01:51] <fragfutter> if i set zFileSystemMapIgnoreNames, do i need to remodel the device(s)

[04-May-2010 08:01:58] <ckrough> morning

[04-May-2010 08:02:12] <ckrough> fragfutter: probably, doesnt hurt to do it and see

[04-May-2010 08:03:43] <fragfutter> ok. remodelling. cleaning filesystem event. waiting 15minutes if it shows up again.

[04-May-2010 08:04:09] <fragfutter> i just wanted to skip the waiting part...

[04-May-2010 08:05:18] <ckrough> if your cycle times are default it should reappear within 5min

[04-May-2010 08:08:56] <ckrough> does zenoss create this Data.fs.old or is that something I left there?

[04-May-2010 08:13:23] <ckrough> mrayzenoss: when you guys set up your zenoss dev environment, do you exclude Data.fs from svn/git?

[04-May-2010 08:13:38] <mrayzenoss> correct

[04-May-2010 08:28:32] <eidolon> hi folks - i have a zenoss installation that's showing heartbeat failures - but i just did a full /etc/init.d/zenoss restart, and the settings->daemons page is showing everything running.

[04-May-2010 08:29:58] <ckrough> eidolon: how is the load on the box? If the daemons aren't finishing their cycles in time they won't update the heartbeat table in time.

[04-May-2010 08:30:23] <eidolon> dbs@lwmzen03:~$ uptime

[04-May-2010 08:30:23] <eidolon>  09:30:16 up 11 days, 22:46,  1 user,  load average: 1.58, 1.47, 0.97

[04-May-2010 08:30:26] <eidolon> 'so so'

[04-May-2010 08:30:37] <eidolon> should i manually restart the processes in the daemon page?

[04-May-2010 08:35:40] <ckrough> dunno, I'd spend some time trying to figure out why the heartbeats are failing

[04-May-2010 08:35:49] <ckrough> otherwise its just going to come back and wake you up at 3am

[04-May-2010 08:36:10] <ckrough> I *think* running zenhub in debug will show heartbeat stuff in the logs

[04-May-2010 08:42:16] <fragfutter> ckrough: just making sure and waiting more then one round.

[04-May-2010 08:42:23] <fragfutter> ckrough: yes, one needs to remodel.

[04-May-2010 08:57:48] <MTecknology> I'm trying to join it :(

[04-May-2010 08:58:28] <MTecknology> first it complained about java not being installed so I'm trying the limited version; and that's just taking forever to load..

[04-May-2010 08:59:13] <ckrough> java?

[04-May-2010 08:59:41] <MTecknology> the webex session

[04-May-2010 09:00:25] <ckrough> ah

[04-May-2010 09:00:27] <MTecknology> I'm starting a windows vm to see if that works...

[04-May-2010 09:00:29] <MTecknology> :(

[04-May-2010 09:00:32] <MTecknology> i iz late

[04-May-2010 09:08:02] <fragfutter> is webex still distributing the audio part of a conference by telephone?

[04-May-2010 09:09:33] <twm1010> morning people

[04-May-2010 09:11:01] <ckrough> morning

[04-May-2010 09:12:06] <twm1010> If I were to replace my ibdata file for MySQL with one from a fresh stack-install, would that work?

[04-May-2010 09:12:12] <ckrough> anyone know the name of the dotfile that forces fresh install scripts? ie: $ZENHOME/.fresh???

[04-May-2010 09:12:23] <ckrough> twm1010: work for what?

[04-May-2010 09:12:30] <MTecknology> fragfutter: ya

[04-May-2010 09:12:54] <twm1010> ckrough: I want to go back to a blank events DB

[04-May-2010 09:13:00] <twm1010> mine is 36GB, and growing :|

[04-May-2010 09:13:04] <twm1010> i don't care about the data

[04-May-2010 09:13:17] <twm1010> but i also need to shrink the file, so i can't just truncate the table

[04-May-2010 09:14:01] <ckrough> hmm

[04-May-2010 09:14:05] <eidolon> oh, this is not good.

[04-May-2010 09:14:17] <eidolon> som, zenvmwaremodeler is failing to start up - here's the error message:  http://pastebin.stonekeep.com/6861

[04-May-2010 09:14:27] <ckrough> I would truncate the table, then mysqldump (w/ --routines), then renamed the innodb files and reimport the data.

[04-May-2010 09:14:35] <ckrough> I think that would work, but YMMV

[04-May-2010 09:14:57] <twm1010> hrmm..

[04-May-2010 09:14:59] <ckrough> twm1010: obviously back it up first :)

[04-May-2010 09:15:17] <ckrough> beware that it takes a loooong time to restore 36gb if you need to go back

[04-May-2010 09:15:21] <ckrough> mine is 36gb as well

[04-May-2010 09:15:36] <twm1010> i definitely don't need to save the data

[04-May-2010 09:15:50] <twm1010> i could care less about old events right now

[04-May-2010 09:16:20] <ckrough> I think you would be ok then

[04-May-2010 09:16:37] <fragfutter> twm1010: then dump only the structure

[04-May-2010 09:16:52] <ckrough> if you renamed/delete the innodb files mysql will rebuild them when it restarts, then import the .sql from the dump

[04-May-2010 09:17:11] <ckrough> fragfutter: good idea

[04-May-2010 09:17:28] <ckrough> make sure you use --routines, there are stored procedures you need in the databases

[04-May-2010 09:17:40] <twm1010> that would do the trick, if only i could login to mysql :)

[04-May-2010 09:18:00] <twm1010> the stack set it up, so i'm unsure which credentials to use

[04-May-2010 09:18:09] <ckrough> sec

[04-May-2010 09:18:20] <fragfutter> twm1010: most likely socket communication without password

[04-May-2010 09:18:28] * eidolon opens a new support ticket with zenoss.

[04-May-2010 09:18:30] <twm1010> ah, i'm in ;)

[04-May-2010 09:19:30] <ckrough> twm1010: you can get it in zendmd

[04-May-2010 09:19:44] <twm1010> are there system tables in there that might get truncated as well if i do this?

[04-May-2010 09:19:51] <twm1010> don't want to lose configs right?

[04-May-2010 09:20:00] <ckrough> system tables?

[04-May-2010 09:20:11] <fragfutter> twm1010: mysql tables are in myisam. if you have other databases inside of mysql they can be in there.

[04-May-2010 09:20:11] <ckrough> dont mess up the mysql database, it has all your permissions

[04-May-2010 09:21:00] <twm1010> gotcha... so i want to run mysqldump with a few switches that will back up the schema and routines, basically

[04-May-2010 09:21:20] <fragfutter> twm1010: you want to check which databases live in your mysql instance.

[04-May-2010 09:21:46] <fragfutter> then you want to backup the structure and the procedures of the events database (or whatever database you connected with zenoss)

[04-May-2010 09:22:03] <fragfutter> then you want to see if any other database is defined and has innodb tables

[04-May-2010 09:22:24] <twm1010> there are three DB's information_schema, events, and test

[04-May-2010 09:22:30] <fragfutter> if none is left, stop mysql. delete innodb files. start mysql. import structure.

[04-May-2010 09:22:47] <fragfutter> information_schema is no real database. test is ... test ;) so nothing to worry.

[04-May-2010 09:23:21] <twm1010> excellent, so i just need the right syntax for the dump and restore commands

[04-May-2010 09:23:53] <ckrough> if you structure dump the whole thing you'll lose your mysql permissions, right?

[04-May-2010 09:24:04] <ckrough> you want to get a full dump of mysql and a structure of events

[04-May-2010 09:24:26] <fragfutter> mysqldump --no-data --opt --routines events > events.sql

[04-May-2010 09:25:10] <fragfutter> mysql stores its internal data in myisam files. /var/lib/mysql/mysql

[04-May-2010 09:25:34] <fragfutter> you only want to kill ibdata1, ib_logfile*

[04-May-2010 09:26:09] <ckrough> what about ibdata2 if it exists?

[04-May-2010 09:26:09] <fragfutter> or be extra careful. Shutdown mysql. make a tarball of the mysql datadir. start again and then do the shrinking

[04-May-2010 09:26:20] <fragfutter> ckrough: ok ibdata* ;)

[04-May-2010 09:26:43] <ckrough> fragfutter: had me scared :)

[04-May-2010 09:27:21] <ckrough> twm1010: if the old db has been running uninterrupted, you may want to run this first just to get tuning ideas: http://blog.mysqltuner.com/

[04-May-2010 09:28:06] <fragfutter> default mysql config says something like: create a big bad ibdata file and expand as necessary. If too big, start next. place all innodb tables into this big bad datafile. you can reconfigure mysql to use one innodb file per table.

[04-May-2010 09:29:51] <ckrough> the '--opt' argument doesnt affect the running db in any way right? just adds drop/adds to the sql that it produces?

[04-May-2010 09:29:58] <fragfutter> ckrough: yes

[04-May-2010 09:30:19] <fragfutter> ckrough: i'm just too lazy to look up all the options hidden behide --opt

[04-May-2010 09:32:05] <twm1010> hrmm.. i'll have to read up on that script

[04-May-2010 09:32:28] <ckrough> twm1010: its safe, and very useful

[04-May-2010 09:32:38] <fragfutter> twm1010: increase handles, increase key cache, increase query cache

[04-May-2010 09:32:38] <ckrough> twm1010: written by a coworker

[04-May-2010 09:33:07] <fragfutter> twm1010: the events database is very plain. no big joins, no complex relations.

[04-May-2010 09:34:06] <twm1010> cool, i think we have a DB guy here who can help me with the dump/restore syntax then

[04-May-2010 09:35:27] <ckrough> here is my my.cnf if you want to compare/critique, let me know if anything looks dumb http://pastebin.com/iKscU78x

[04-May-2010 09:35:39] <ckrough> thats on a dedicated 32GB/8core box

[04-May-2010 09:35:46] <ckrough> events db is around 37GB

[04-May-2010 09:36:13] <twm1010> see mine shouldn't be that size really, the first year we've used it here we didn't really care about events so i kinda let them stack up instead of classify them

[04-May-2010 09:36:26] <fragfutter> i have an event that started today 15:21 and ended 16:29. In the log for this event there is an entry that says "auto cleared" for 18:30. (current time is 16:36). So the log entries get translated to the wrong timezone or translated twice.

[04-May-2010 09:36:30] <ckrough> do you have it set to auto-expire old events?

[04-May-2010 09:36:32] <twm1010> i monitored about 200 devices, on an older DL380G4

[04-May-2010 09:36:40] <twm1010> yeah, i recently adjusted it from 365 to 90 days

[04-May-2010 09:36:54] <ckrough> thats a lot of events for 200devices :)

[04-May-2010 09:37:07] <twm1010> but certain components of the GUI are ridiculously slow

[04-May-2010 09:37:15] <twm1010> say.. clickin on a device you haven't touched in a while

[04-May-2010 09:37:24] <twm1010> takes about 2-3 minutes to come up

[04-May-2010 09:37:39] <rmatte> twm1010: I've noticed that since I upgraded to 2.5.2

[04-May-2010 09:37:47] <rmatte> twm1010: used to be lightning fast with 2.4.5

[04-May-2010 09:37:58] <fragfutter> twm1010: but that is not mysql database related, that is zodb

[04-May-2010 09:38:05] <rmatte> no, that's mysql related

[04-May-2010 09:38:14] <rmatte> it's calculating availability when you load the status page

[04-May-2010 09:38:15] <ckrough> fragfutter: mysql has a big impact on the gui as well

[04-May-2010 09:38:17] <twm1010> potentially, but if its having to calculate availability on events, and the database is running like crap

[04-May-2010 09:38:19] <rmatte> it uses the mysql db for that

[04-May-2010 09:38:30] <twm1010> see why I want to start over? :)

[04-May-2010 09:38:31] <fragfutter> rmatte: ah. ok.

[04-May-2010 09:39:11] <rmatte> I find that everything to do with events is just generally slower in 2.5.2

[04-May-2010 09:39:22] <twm1010> trying to get phpmyadmin to work against the stack mysql

[04-May-2010 09:39:32] <twm1010> gets a socket error

[04-May-2010 09:39:34] <rmatte> hmmm, I should check if the upgrade knocked out my custom mysql config

[04-May-2010 09:41:18] <twm1010> I added: $cfg['Servers'][$i]['socket'] = '/apps/zenoss/mysql/tmp/mysql.sock';

[04-May-2010 09:41:25] <twm1010> to my phpmyadmin config :|

[04-May-2010 09:41:59] <twm1010> does stack mysql run on default mysql port?

[04-May-2010 09:43:43] <fragfutter> twm1010: netstat -lnpt

[04-May-2010 09:44:13] <rmatte> hmmm it did knock out the custom config, nice

[04-May-2010 09:45:30] <twm1010> ah, looks like zenoss-stack mysql uses 3307

[04-May-2010 09:45:32] <fragfutter> the viewHistoryDetail template is bugged. The fields start-time and stop-time are not timezone sensitive. The entries in log are.

[04-May-2010 09:45:51] <twm1010> probalby so it can run in parallel with a normal install

[04-May-2010 09:46:31] <rmatte> what's the easiest way to restart mysql is a stack install without restarting all of Zenoss?

[04-May-2010 09:46:38] <rmatte> in a stack install*

[04-May-2010 09:46:51] <rmatte> (I've never had need to do it until now lol)

[04-May-2010 09:48:10] <twm1010> i was pondering that last night myself :|

[04-May-2010 09:53:29] <rmatte> lol

[04-May-2010 09:53:35] <rmatte> well, let's figure it out...

[04-May-2010 09:54:46] <MTecknology> So.. SNMP is insecure but that's fine because you can't actually interact with the host system without changing the configurations?

[04-May-2010 09:55:26] <fragfutter> MTecknology: snmp v2 is insecure. v3 has all the fance enryption and authentication stuff you want.

[04-May-2010 09:55:34] <fragfutter> s/fance/fancy/

[04-May-2010 09:55:56] <fragfutter> where do i report a bug for zenoss core?

[04-May-2010 09:56:01] <rmatte> MTecknology: If SNMP is enabled by default it'll have the community string "public"

[04-May-2010 09:56:05] <rmatte> and will be wide open

[04-May-2010 09:56:25] <fragfutter> rmatte: depends on the system.

[04-May-2010 09:56:30] <rmatte> you have to specifically configure your snmp agents to only accept connections from certain IPs and change the default community string

[04-May-2010 09:56:32] <ckrough> fragfutter: http://dev.zenoss.org/trac

[04-May-2010 09:56:44] <fragfutter> ckrough: thanks

[04-May-2010 09:56:46] <MTecknology> thanks

[04-May-2010 09:56:47] <rmatte> fragfutter: true, but I'm pointing out the general case

[04-May-2010 09:57:04] <rmatte> fragfutter: obviously a Linux system doesn't have it enabled by default

[04-May-2010 09:57:35] <rmatte> fragfutter: login to trac with zenoss/zenoss

[04-May-2010 09:57:44] <rmatte> or ask Matt Ray for your own account next time you see him

[04-May-2010 09:57:46] <rmatte> (mrayzenoss)

[04-May-2010 09:57:56] <fragfutter> rmatte: ah. thanks.

[04-May-2010 09:58:00] <rmatte> np

[04-May-2010 10:07:45] <MTecknology> mrayzenoss: ya.. that was worth showing up to :)

[04-May-2010 10:09:14] <fragfutter> http://dev.zenoss.org/trac/ticket/6652

[04-May-2010 10:09:19] <fragfutter> hope its understandable

[04-May-2010 10:09:46] <rmatte> garr, it should not be this difficult to restart mysql lol

[04-May-2010 10:10:53] <fragfutter> rmatte: is mysqld_safe running? if yes just send a kill to mysqld. It should restart mysql then.

[04-May-2010 10:11:32] <ckrough> zensocket is the only file that needs special permissions, ya?

[04-May-2010 10:12:29] <fragfutter> ckrough: zenoss will write to the performance data rrd.

[04-May-2010 10:12:56] <rmatte> fragfutter: ah, good idea

[04-May-2010 10:13:03] <fragfutter> rmatte: and logfiles

[04-May-2010 10:13:18] <ckrough> fragfutter: you lost me on that one

[04-May-2010 10:13:18] <fragfutter> rmatte: and it wants to bytecompile its python files

[04-May-2010 10:14:05] <rmatte> killall mysqld.bin did the trick

[04-May-2010 10:14:41] <rmatte> yeh, you lost me too

[04-May-2010 10:14:43] <rmatte> lol

[04-May-2010 10:14:50] <rmatte> not sure what that has to do with restarting mysql

[04-May-2010 10:14:54] <rmatte> :P

[04-May-2010 10:15:01] <fragfutter> rmatte: then you have a second process that watches for mysqld(.bin) and if that one crashes restarts it. This is normaly called mysqld_safe

[04-May-2010 10:15:20] <rmatte> yeh, I didn't know about the safe process

[04-May-2010 10:15:23] <rmatte> it's basically just a watchdog

[04-May-2010 10:15:29] <fragfutter> rmatte: yes.

[04-May-2010 10:15:34] <rmatte> that simplifies things

[04-May-2010 10:15:52] <fragfutter> and i mixed rmatte and ckrough on a few lines ;)

[04-May-2010 10:15:58] <rmatte> actually, hmmm, that didn't work so well

[04-May-2010 10:16:11] <rmatte> can't connect to the mysql server message in the UI

[04-May-2010 10:16:53] <fragfutter> rmatte: something still listening on 3307?

[04-May-2010 10:20:17] <rmatte> hmmm, my aggregate reports aren't working on one of my servers

[04-May-2010 10:20:25] <rmatte> fragfutter: I just decided to completely restart Zenoss, meh

[04-May-2010 10:22:33] <twm1010> i wonder if i could make this easier, and just delete the mysql directory, and re-run the stack installer

[04-May-2010 10:24:07] <rmatte> twm1010: that would probably work

[04-May-2010 10:24:16] <rmatte> though I'd just move the directory, not delete it

[04-May-2010 10:24:18] <rmatte> just in case

[04-May-2010 10:24:18] <twm1010> so.. i'll rename it... see what happens,

[04-May-2010 10:24:22] <rmatte> yeh

[04-May-2010 10:26:25] <ckrough> Does anyone know the name of the dotfile that forces zenoss to run thought initial setup? something like $ZENHOME/.freshinstall

[04-May-2010 10:27:04] <fragfutter> fresh_install

[04-May-2010 10:27:09] <ckrough> danke

[04-May-2010 10:27:27] <fragfutter> cd $ZENHOME; grep -r -H * -e '\.fresh'

[04-May-2010 10:27:35] <twm1010> you know, i'm getting very punchy, hopefully i didn't just nuke my original setup here

[04-May-2010 10:28:22] <twm1010> not too worried, i have a zenbackup, but, ya know.

[04-May-2010 10:30:19] <twm1010> yeah, pretty sure it is :|

[04-May-2010 10:33:05] <twm1010> nope, that did it

[04-May-2010 10:37:39] <rmatte> hehe

[04-May-2010 10:38:23] <twm1010> wow, its like someone hit the turbo button

[04-May-2010 10:38:33] <twm1010> anything i click on, instant response

[04-May-2010 10:40:41] <ckrough> twm1010: do you run it on a dedicated box?

[04-May-2010 10:41:06] <twm1010> yeah

[04-May-2010 10:41:19] <twm1010> it has some config deficiencies

[04-May-2010 10:42:32] <rmatte> eugh, this is so dumb, every one of my servers is capable of generating an aggregate network graph except for 1

[04-May-2010 10:44:42] <rmatte> When I input the URL with the arguments to generate the graph it just spits that back out as text

[04-May-2010 11:09:03] <rmatte> nothing in the logs that would explain it, other than the usual db conflict errors (resolved)

[04-May-2010 11:41:19] <rmatte> anyone know where RenderServer would log to?

[04-May-2010 11:41:34] <rmatte> I'm guessing events.log

[04-May-2010 11:41:43] <rmatte> but I don't see anything useful in there

[04-May-2010 11:53:28] <rmatte> damnit, I accidentally left my phone playing music and now the battery is almost dead

[04-May-2010 12:42:42] RoAk is now known as RoAkSoAx

[04-May-2010 13:00:25] <voxter> Is it possible using distributed collectors for a host to have some performance graphs collected by one remote collector, and some other stuff (like snmp data) collected from the local collector?

[04-May-2010 13:01:06] <rmatte> nope

[04-May-2010 13:01:27] <rmatte> each device is assigned a collector and that's the collector that does the work

[04-May-2010 13:02:21] <voxter> Ya i kinda figured. and you cant add a device twice either hmm... the problem is i have only one graph (latency and loss) that i want graphed from the remote collector, all the other data i would like our main zenoss to do. can you think of a workaround?

[04-May-2010 13:03:21] <rmatte> you can actually add a device twice

[04-May-2010 13:03:37] <rmatte> you need to add it with a fake IP then rename it and change the IP by hand

[04-May-2010 13:03:52] <rmatte> not sure if it'd work if you did that and assigned the second device to a different collector

[04-May-2010 13:19:36] <ckrough> argh, does zenoss resolve and cache the event database IP? I overrode it in /etc/hosts but it still sent events to the production database...

[04-May-2010 13:20:13] <rmatte> hmmm

[04-May-2010 13:20:37] <rmatte> try restarting zenhub?

[04-May-2010 13:22:03] <rmatte> caching would make sense since you don't want Zenoss to lose connectivity to the event database all of a sudden because a dns server drops

[04-May-2010 13:26:27] <twm1010> rmatte: yeah, my events database was seriously nerfing my performance

[04-May-2010 13:39:06] <rmatte> npmccallum: that zProperties issue I was having was so dumb, found out it was caused by me using Jane Curry's jc_batch_load script

[04-May-2010 13:39:16] <rmatte> there was a section not properly coded which caused the issue

[04-May-2010 13:39:33] <npmccallum> rmatte: glad you found it :)

[04-May-2010 13:39:34] <rmatte> so I ran delattr across the board and it's all good now

[04-May-2010 13:39:44] <npmccallum> rmatte: you should give her a patch

[04-May-2010 13:39:44] <rmatte> thanks again for the help

[04-May-2010 13:40:08] <rmatte> now I get to tackle why zenmigrate refuses to work on one of my servers

[04-May-2010 13:42:51] <rmatte> I love vmware, I cloned the server so that I can leave it in production and boot up the clone to troubleshoot the zenmigrate issue

[04-May-2010 13:46:15] <twm1010> one day i'll have a vmware zenoss instance, one day... :(

[04-May-2010 13:46:29] <twm1010> right after we have SAN storage ;)

[04-May-2010 13:47:19] <npmccallum> rmatte, twm1010: btrfs snapshots can get you the same thing without the performance loss of virtualization :)

[04-May-2010 13:47:51] <npmccallum> not that I would *ever* recommend anyone run zenoss on an unstable filesystem

[04-May-2010 13:47:58] <npmccallum> ;)

[04-May-2010 13:48:32] <mrayzenoss> heh, I'm using VMware on my Mac, with a shared filesystem for zenpack development

[04-May-2010 13:48:45] <mrayzenoss> definitely an unstable filesystem

[04-May-2010 13:49:24] <npmccallum> mistake #1 - vmware

[04-May-2010 13:49:30] <ckrough> ooohhhh

[04-May-2010 13:49:31] <npmccallum> mistake #2 - mac

[04-May-2010 13:49:31] <npmccallum> ;)

[04-May-2010 13:49:35] <npmccallum> JK

[04-May-2010 13:49:52] <rmatte> lol

[04-May-2010 13:50:15] <rmatte> vmware ESX is great, vmware workstation is crap

[04-May-2010 13:50:31] <rmatte> I actually have a valid vmware workstation license and never use it

[04-May-2010 13:50:34] <rmatte> virtualbox++

[04-May-2010 13:50:52] <npmccallum> yes, i've been quite impressed with the virtualbox performance

[04-May-2010 13:51:00] <npmccallum> and 3.2 is going to add Mac guest support

[04-May-2010 13:51:09] <rmatte> nice

[04-May-2010 13:53:14] <rmatte> voila, zenmigrate error

[04-May-2010 13:53:15] <rmatte> Products.ZenRelations.Exceptions.ZenSchemaError: Schema for relation job not

[04-May-2010 13:53:16] <rmatte> found on JobStatus

[04-May-2010 13:53:47] <rmatte> I tried zenmigrate --level=2.3.3 but that didn't work either

[04-May-2010 14:00:25] <rmatte> hmmm, mysql is taking forever to start on the server, could be a clue

[04-May-2010 14:02:58] <rmatte> ah, there were a multiple instances of it running

[04-May-2010 14:06:04] <rmatte> zenmigrate still failing with same message

[04-May-2010 14:06:18] mrayzenoss1 is now known as mrayzenoss

[04-May-2010 14:06:52] <rmatte> anyone have any idea about what to try for this: http://fpaste.org/E5ji/

[04-May-2010 14:06:53] <rmatte> ?

[04-May-2010 14:08:02] <twm1010> it fails on fix bad jobs?

[04-May-2010 14:08:08] <rmatte> yup

[04-May-2010 14:08:15] <rmatte> it's the only one of my servers that does that

[04-May-2010 14:08:23] <twm1010> do you even have any jobs, discovery, etc?

[04-May-2010 14:08:27] <rmatte> nope

[04-May-2010 14:08:29] <rmatte> none

[04-May-2010 14:09:24] <twm1010> doh, zentray doesn't like Win7-64 :(

[04-May-2010 14:09:30] <rmatte> lol

[04-May-2010 14:09:51] <npmccallum> rmatte: what was the first version you installed on this server? what was your upgrade path?

[04-May-2010 14:10:27] <twm1010> bbl

[04-May-2010 14:10:31] <rmatte> npmccallum: 2.4.1 -> 2.4.5 -> 2.5.2

[04-May-2010 14:10:42] <rmatte> npmccallum: same upgrade path that I followed for my other servers

[04-May-2010 14:10:55] <rmatte> this server is cloned from the same template as the others were

[04-May-2010 14:11:46] <npmccallum> ok that is a bit wierd

[04-May-2010 14:11:49] <rmatte> yup

[04-May-2010 15:26:22] <rmatte> eugh, this zenmigrate issue is brutal

[04-May-2010 15:51:06] <MTecknology> zenmigrane?

[04-May-2010 15:51:31] <rmatte> yup

[04-May-2010 15:52:23] <MTecknology> interesting

[04-May-2010 16:05:13] <rmatte> there are so many functions called on leading up to this zenrender error eugh, been going through file by file seeing if I can figure out how to at least do some troubleshooting in zendmd

[04-May-2010 16:23:50] <twm1010> sure am getting a lot of false zen heartbeat failures since 2.5.2

[04-May-2010 16:29:02] <rmatte> twm1010: me too, TONS

[04-May-2010 16:29:27] <rmatte> I don't have Zenoss configured to alert on them but it is annoying non-the-less

[04-May-2010 16:29:51] <rmatte> I get a heartbeat failure for basically every daemon every few minutes

[04-May-2010 16:30:44] <rhettardo> gross

[04-May-2010 16:31:05] <rmatte> yup

[04-May-2010 16:33:33] <rmatte> they really need to tune those properly for the next release

[04-May-2010 16:33:46] <rmatte> they are useless as is

[04-May-2010 16:34:04] <rmatte> might as well not even have them at all

[04-May-2010 16:42:24] <MTecknology> zenoss is a massive install

[04-May-2010 16:42:35] <rmatte> I wouldn't call it "massive"

[04-May-2010 16:42:45] <MTecknology> very large

[04-May-2010 16:42:54] <rmatte> lol

[04-May-2010 16:43:04] <rmatte> well, what's very large for you?

[04-May-2010 16:43:06] <MTecknology> no... massive is appropriate

[04-May-2010 16:43:10] <twm1010> rmatte: you ever get to talking about voice gateway snmp config?

[04-May-2010 16:43:24] <rmatte> not yet

[04-May-2010 16:43:25] <MTecknology> rmatte: anything over 10mb is big

[04-May-2010 16:43:28] <MTecknology> :)

[04-May-2010 16:43:35] <rmatte> MTecknology: you're living in the past lol

[04-May-2010 16:43:38] <rmatte> :P

[04-May-2010 16:43:49] <twm1010> i'm thinking the global linkup linkdown statements would be removed

[04-May-2010 16:43:52] <MTecknology> rmatte: http://imagebin.ca/img/bA13SQ7O.png

[04-May-2010 16:44:04] <twm1010> and then on your D channels and gig interfaces you set snmp trap link-status

[04-May-2010 16:44:18] <rmatte> nice wallpaper, what wm, fluxbox?

[04-May-2010 16:44:45] <MTecknology> openbox

[04-May-2010 16:44:49] <rmatte> ah

[04-May-2010 16:45:01] <MTecknology> I kinda liked fluxbox but I kinda didn't

[04-May-2010 16:45:07] <rmatte> twm1010: yeh, I don't have any devices to configure yet, but soon

[04-May-2010 16:45:29] <rmatte> fluxbox is my favourite wm

[04-May-2010 16:45:55] <rmatte> and I use the wm in distinction from desktop environments (gnome, kde)

[04-May-2010 16:45:55] <MTecknology> I've been installing the .deb for about 10min now

[04-May-2010 16:45:59] <rmatte> I use gnome personally

[04-May-2010 16:46:05] <rmatte> I used to be a big kde user

[04-May-2010 16:46:11] <rmatte> the first wm I used was windowmaker

[04-May-2010 16:46:16] <rmatte> and I've used fluxbox and blackbox

[04-May-2010 16:46:26] <MTecknology> i no likey windowmaker

[04-May-2010 16:46:36] <MTecknology> it makes me sad and hateful :P

[04-May-2010 16:46:36] <rmatte> it shouldn't take 10 mins, the stack install takes me like 5 minutes

[04-May-2010 16:46:58] <rmatte> I didn't say I liked it lol, it just happened to be the first that I used

[04-May-2010 16:47:04] <rmatte> when I was just starting to learn FreeBSD

[04-May-2010 16:47:16] <rmatte> I was 14 or 15 at the time

[04-May-2010 16:47:19] <rmatte> ;)

[04-May-2010 16:47:22] <MTecknology> it's still setting up zenoss-stack

[04-May-2010 16:47:40] <rmatte> well, what are the specs on the box you're installing on though?

[04-May-2010 16:47:45] <rmatte> not a 486 or something I hope?

[04-May-2010 16:47:46] <mrayzenoss> MTecknology: what kind of hardware?

[04-May-2010 16:47:50] <mrayzenoss> jinx

[04-May-2010 16:47:53] <rmatte> ;)

[04-May-2010 16:48:35] <MTecknology> 2yr old laptop vm 768mb ram

[04-May-2010 16:48:43] <rmatte> that'd explain it

[04-May-2010 16:48:46] <MTecknology> I'm going to go eaat while I wait

[04-May-2010 16:48:49] <MTecknology> ttyal :)

[04-May-2010 16:48:58] <rmatte> good luck running Zenoss on 768mb of RAM

[04-May-2010 16:49:04] <rmatte> that'll be eaten up the second you start it up

[04-May-2010 16:49:21] <rmatte> MySQL and Zope alone will consume that

[04-May-2010 16:49:43] <rmatte> you need a minimum of 2GB really

[04-May-2010 16:50:07] <mrayzenoss> I run Zenoss on a pair of laptops with a gig and 1.25 gigs

[04-May-2010 16:50:14] <MTecknology> i have 2GB on the whole system

[04-May-2010 16:50:16] <mrayzenoss> they're pokey

[04-May-2010 16:50:21] <rmatte> mrayzenoss: yeh, but are you monitoring like 2 devices?

[04-May-2010 16:50:32] <rmatte> lol

[04-May-2010 16:50:32] <MTecknology> I have 25 to monitor

[04-May-2010 16:50:51] <rmatte> what level of monitoring, full SNMP?

[04-May-2010 16:51:12] <MTecknology> I'm not sure - but gf says food time

[04-May-2010 16:51:13] <MTecknology> ttyl

[04-May-2010 16:51:14] <rmatte> anything under a gig definitely won't cut it, so you should probably allocate more memory to it

[04-May-2010 16:51:19] <rmatte> later

[04-May-2010 16:56:52] <rmatte> well, I guess I'll post this zenrender error on the forums and hope someone has some idea of how to fix it

[04-May-2010 16:57:00] <rmatte> zenmigrate, rather

[04-May-2010 16:57:12] <mrayzenoss> yeah, 10 devices

[04-May-2010 17:00:18] <rmatte> thread/13448

[04-May-2010 17:00:28] <rmatte> If anyone can help me out, awesome

[04-May-2010 17:00:50] <rmatte> If not, I shall continue tearing my hair out trying to figure this out lol

[04-May-2010 17:01:17] <mrayzenoss> posted internally, hopefully someone will speak up

[04-May-2010 17:01:22] <rmatte> thanks

[04-May-2010 17:02:25] <rmatte> hmmmm, hockey game is on tonight, think I'll get going in about 15 to 20 mins

[04-May-2010 17:52:56] <MTecknology> mrayzenoss: so a gig per 10 devices?

[04-May-2010 17:54:18] <mrayzenoss> no

[04-May-2010 17:54:42] <mrayzenoss> you can probably do 50 devices with 2 gigs

[04-May-2010 17:54:50] <MTecknology> ok, thanks

[04-May-2010 17:54:52] <mrayzenoss> perhaps 1000 with 4 gigs

[04-May-2010 17:54:59] <MTecknology> oh, wow

[04-May-2010 17:55:14] <mrayzenoss> gotta run

[04-May-2010 17:55:25] <MTecknology> I'll allocate 1.5GB - I need the other 0.5GB for my system..

[04-May-2010 17:55:27] <MTecknology> mrayzenoss: thanks

[04-May-2010 17:55:52] <MTecknology> so.. I installed it..

[04-May-2010 17:56:46] <MTecknology> running the init script..

[04-May-2010 17:58:55] <MTecknology> allocating more ram..

[04-May-2010 17:58:59] <twm1010> :)

[04-May-2010 17:59:32] <MTecknology> i only have 2g on this thing

[04-May-2010 17:59:48] <twm1010> 2 gigs of ram? i ran on that for a long time

[04-May-2010 17:59:57] <twm1010> did up to 150 devices at one point, but you do start hitting walls

[04-May-2010 18:00:15] <MTecknology> this is my only system - my laptop

[04-May-2010 18:00:39] <twm1010> and you're running it as a VM, or native?

[04-May-2010 18:00:48] <MTecknology> vm

[04-May-2010 18:01:38] <MTecknology> On the System page, you have assigned more than 50% of your computer's memory (1.95 GB) to the virtual machine. There might not be enough memory left for your host operating system. Continue at your own risk.

[04-May-2010 18:01:42] <MTecknology> :P

[04-May-2010 18:02:39] <twm1010> is there a way to set a default dashboard?

[04-May-2010 18:02:50] <MTecknology> ok.. 1236MB - I'm gonna try that..

[04-May-2010 18:03:02] <mistich> anyone see this error before 67.0.0.0: object /zport/dmd/Networks/67.0.0.0/67.192.0.0/67.192.128.0/ipaddresses/67.192.157.97 not found on relation Vlan62/ipaddresses

[04-May-2010 18:03:17] <mistich> oh hello everyone

[04-May-2010 18:03:28] <twm1010> mistich what is generating said error/

[04-May-2010 18:03:30] <twm1010> ?

[04-May-2010 18:03:51] <mistich> when trying to delete a network

[04-May-2010 18:03:56] <MTecknology> happy -> echo 3 > /proc/sys/vm/drop_caches

[04-May-2010 18:04:17] <mistich> sync; echo 3 > /proc/sys/vm/drop_caches

[04-May-2010 18:04:24] <mistich> little cleaner

[04-May-2010 18:04:26] <twm1010> hrmm... no idea, maybe some weird bug with CIDR notation did something weird?

[04-May-2010 18:04:34] <MTecknology> mistich: makes sense

[04-May-2010 18:05:09] <mistich>  object 74.205.89.130/24 was not found on <ToOneRelationship at nexthop>

[04-May-2010 18:05:13] <mistich> how about this one

[04-May-2010 18:05:19] <MTecknology> ok.. let's startup openvpn and try out zenoss

[04-May-2010 18:08:44] <MTecknology> ok.. now to setup SNMP on the systems....

[04-May-2010 18:09:02] <MTecknology> is there anything to make that easy?

[04-May-2010 18:09:22] <rhettardo> you can write a script

[04-May-2010 18:09:38] <rhettardo> thats waht i do

[04-May-2010 18:09:39] <MTecknology> I mean like a guide that says what I need to do - I've never used it

[04-May-2010 18:09:45] <rhettardo> oh

[04-May-2010 18:09:53] <rhettardo> what OS

[04-May-2010 18:09:55] <MTecknology> any chance I could borrow yours with private info stripped?

[04-May-2010 18:09:58] <MTecknology> ubuntu

[04-May-2010 18:10:17] <MTecknology> aptitude isntall snmpd; that's as far as I can guess :P

[04-May-2010 18:10:20] <rhettardo> i can surely give you a dumbed down config

[04-May-2010 18:10:30] <twm1010> hrmm... there is one other thing with ubuntu

[04-May-2010 18:10:43] <twm1010> some config file you have to change that let's a host other than 127.0.0.1 query snmp

[04-May-2010 18:10:59] <twm1010> i can't remember where though, its not in /etc/snmpd.conf though, its somewhere else

[04-May-2010 18:11:05] <MTecknology> oh

[04-May-2010 18:11:09] <MTecknology> I'm sure I can find it

[04-May-2010 18:11:56] <MTecknology> I'm setting this up on six of my nodes atm

[04-May-2010 18:12:19] <rhettardo> http://pastebin.org/202032

[04-May-2010 18:12:29] <MTecknology> i love terminator...

[04-May-2010 18:12:54] <MTecknology> ssh into 6 machines, become root, group all the terminals, do everything at the same time :)

[04-May-2010 18:13:25] <rhettardo> then change public to whatever community string you would like

[04-May-2010 18:13:37] <rhettardo> and change mynetwork to whatever you'd like to give access to

[04-May-2010 18:16:59] <MTecknology> default is com2sec paranoid  default         public

[04-May-2010 18:20:06] <MTecknology> rhettardo: the second column is a profile thing?

[04-May-2010 18:20:23] <MTecknology> then the last is the grouping part?

[04-May-2010 18:22:07] <MTecknology> rhettardo:

[04-May-2010 18:22:08] <MTecknology> view all included .1 80

[04-May-2010 18:22:14] <MTecknology> what's the 80 part?

[04-May-2010 18:22:20] <MTecknology> the default config doesn't have that

[04-May-2010 18:22:59] <MTecknology> and what's this part? view system included iso.org.dod.internet.mgmt.mib-2.system

[04-May-2010 18:23:18] <MTecknology> oh - I get that 80 part

[04-May-2010 18:23:27] <MTecknology> the default for that last one i sview system included  .1.3.6.1.2.1.1

[04-May-2010 18:26:34] <rhettardo> the 80 part is the mask

[04-May-2010 18:26:58] <rhettardo> the second include line is what we are going to output if we only tell it to show us system

[04-May-2010 18:27:11] <rhettardo> i.e. snmpwalk -v 2c -c public web1 system

[04-May-2010 18:28:09] <MTecknology> So I should just change it to what you had

[04-May-2010 18:28:19] <rhettardo> if you'd like

[04-May-2010 18:29:34] <MTecknology> It look slike the default config on mine has system.sysDescr.0

[04-May-2010 18:29:43] <MTecknology> system.sysDescr.0 = "blah"

[04-May-2010 18:30:48] <MTecknology> looks like those pieces just add friendly info

[04-May-2010 18:33:49] <MTecknology> rhettardo: is        access MyROGroup ""      any       noauth    exact  all    none   none        the same as          s/exact/0/ ?

[04-May-2010 18:41:25] <twm1010> i don't have an ubuntu box in front of me, but if it serves me correctly, there was another snmp configuration file, maybe in sysconfig or some other location that determined what other hosts could hit the daemon

[04-May-2010 18:41:31] <twm1010> it was specific to ubuntu for sure

[04-May-2010 18:42:23] <twm1010> here

[04-May-2010 18:42:23] <twm1010> https://help.ubuntu.com/community/Zenoss

[04-May-2010 18:42:28] <twm1010> under Troubleshooting

[04-May-2010 18:42:43] <twm1010> that's what i'm referring to, its what interface SNMPD listens on

[04-May-2010 18:42:49] <twm1010> obviously loopback ain't gonna do crap for you

[04-May-2010 18:45:15] <MTecknology> alrighty... time to try this out

[04-May-2010 18:45:35] <MTecknology> no errors... that has to be good

[04-May-2010 18:48:58] <MTecknology> ok... giving it that amount of ram is making it run not so happy

[04-May-2010 18:53:48] <MTecknology> how can I check the progress of a discovery?

[04-May-2010 18:58:00] <MTecknology> yay - it's working

[04-May-2010 19:05:48] <MTecknology> wow.... this is sucking the ram :P

[04-May-2010 19:07:43] <MTecknology> first they were up - now they're showing as down

[04-May-2010 19:07:46] <MTecknology> :(

[04-May-2010 19:19:23] <MTecknology> musta been lagging that bad

[04-May-2010 19:19:47] * MTecknology is trying to remember how to change the classes the system is in - like /Discovered/Servers/Linux

[04-May-2010 19:23:39] <MTecknology> ok.. drag/drop into the server section

[04-May-2010 19:33:27] <MTecknology> ok.....

[04-May-2010 19:33:31] <MTecknology> I LOVE THIS THING

[04-May-2010 19:33:47] <MTecknology> Aside from it being heavy - it's awesoem

[04-May-2010 19:38:18] <MTecknology> networking keeps dying on my system - no guesses why....

[04-May-2010 19:38:59] <MTecknology> anyway - I saw what I needed to - and very happy with it

[04-May-2010 20:47:06] <voxter> with the windows WMI monitor, should i use either WMI or SNMP? or is it proper to use both? or..

[04-May-2010 20:54:46] <etank> voxter: i have used a combo of both in the past

[04-May-2010 20:54:55] <etank> with no real issue

[04-May-2010 20:54:58] <voxter> Appears with the new WMI pack i can eliminate snmp

[04-May-2010 20:54:59] <voxter> Thanks!

[04-May-2010 20:55:32] <etank> i have had issues where the wmi junk on a windows box was so screwed up that it was easier to go snmp

[05-May-2010 00:00:30] [disconnected at Wed May  5 00:00:30 2010]

[05-May-2010 00:00:30] [connected at Wed May  5 00:00:30 2010]

[05-May-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[05-May-2010 06:57:16] ckrough is now known as crow

[05-May-2010 06:57:24] crow is now known as coofamani

[05-May-2010 08:50:48] <tehhobbit> hmmm got a problem, cant scale graphs for devices that isnt on the same collector as the gui is running on

[05-May-2010 09:08:19] <ptom> does anyone know which function I should be using to set custom properties? I have tried _setProperty and _updateProperty but I get the following error. NameError: name 'cTest' is not defined.

[05-May-2010 09:08:48] <ptom> yet, when I do something like this: print node.custPropertyIds() It's lists the property

[05-May-2010 09:12:30] <fragfutter> ptom: if d is your device object, then  d._setProperty('foo', 'bar')

[05-May-2010 09:12:53] <ptom> tried that

[05-May-2010 09:13:37] <fragfutter> dmd.Devices.findDevice('something')._setProperty('foo','bar')

[05-May-2010 09:13:40] <fragfutter> works.

[05-May-2010 09:13:45] <fragfutter> otherwise show code ;)

[05-May-2010 09:13:45] <ptom> hmmm

[05-May-2010 09:14:01] <fragfutter> (and don't paste it to the channel)

[05-May-2010 09:14:27] <ptom> lol

[05-May-2010 09:16:07] <ptom> http://pastebin.com/bZV3mSeQ

[05-May-2010 09:18:29] <fragfutter> you are talking via RPC. I would guess that all functions starting with _ are hidden from rpc (as the underscore indicates private functions)

[05-May-2010 09:18:49] <fragfutter> do you need to use rpc?

[05-May-2010 09:18:54] <ptom> yeah

[05-May-2010 09:19:46] <fragfutter> your script needs to run remote?

[05-May-2010 09:20:28] <ptom> it does. this is only part of the script. the other part is pulling snmp data and comparing stuff in a DB, then setting this field.

[05-May-2010 09:21:41] <fragfutter> ptom: i can only tell how you do it with direct access to dmd

[05-May-2010 09:22:15] <ptom> there has to be another way

[05-May-2010 09:22:21] <ptom> with rpc

[05-May-2010 09:23:57] <ptom> thanks fragfutter, at least I know about the _ being hidden now.

[05-May-2010 09:24:13] <fragfutter> its only a guess.

[05-May-2010 09:25:08] <ptom> seems accurate

[05-May-2010 09:26:35] <sektorNBA> I am monitoring some windows servers and I wanna monitor the conectivity from this servers to a specific IP. How can i do that using zenoss ?

[05-May-2010 09:44:14] <fragfutter> ptom: i think you would need to expand the Device class with a getter/setter function

[05-May-2010 09:52:41] <ptom> hmmm, I was kind of affraid of that

[05-May-2010 09:53:23] <MTecknology> ptom: me too... not sure how I can track down the issue though

[05-May-2010 09:53:35] <fragfutter> or work directly with the zeo

[05-May-2010 09:54:58] <ptom> thanks for the input, i will keep hacking around and see what I can come up with

[05-May-2010 09:56:02] <fragfutter> if you don't want to touch the existing Device class, you could write a proxy class, that accepts devicename, propertyname, value and does it for oyu.

[05-May-2010 09:56:34] <ptom> do you know where I can find docs on that? I am a python Newb

[05-May-2010 09:56:43] <fragfutter> ptom: then don't do it

[05-May-2010 09:56:52] <fragfutter> ptom: you need to know python and zope.

[05-May-2010 09:56:53] <ptom> ahahaha

[05-May-2010 09:57:04] <fragfutter> ptom: hard to swallow in one go.

[05-May-2010 09:57:19] <ptom> what's life without a challenge?

[05-May-2010 09:58:00] <fragfutter> zope.org and plone.org are good information sources. Or poke around inside of http://<host>:8080/manage

[05-May-2010 09:58:30] <ptom> will do, even if i don't figure it out, I will at least learn something

[05-May-2010 10:03:05] <ptom> I got it!

[05-May-2010 10:03:32] <fragfutter> so lets here it ;)

[05-May-2010 10:03:32] <ptom>  node.setZenProperty('cTest', '1234')

[05-May-2010 10:03:47] <ptom> how or why that worked is beyond me

[05-May-2010 10:03:52] <ptom> but it worked

[05-May-2010 10:06:23] <fragfutter> Device is a subclass of ZenPropertyManager. ZenPropertyManager defines the function. Be careful acquisition is involved

[05-May-2010 10:07:12] <ptom> so I could inadvert be setting something else?

[05-May-2010 10:07:43] <fragfutter> setting will run on the node you chose. but getting can also return values higher up in the acquisition chain

[05-May-2010 10:07:55] <ptom> hmmm

[05-May-2010 10:27:24] <rmatte> eugh, I hate how even changing commands in Zenoss requires a password now, and it doesn't even work, have to go right in to the zope management interface to make changes

[05-May-2010 11:05:57] <rmatte> I'm finally starting to sort of get somewhere with this: message/48481#48481

[05-May-2010 11:06:00] <rmatte> thanks to Ian

[05-May-2010 11:06:07] <rmatte> still haven't completely figured it out yet though

[05-May-2010 11:36:28] <rmatte> YES!

[05-May-2010 11:36:34] <rmatte> I finally figured this out!

[05-May-2010 11:36:42] <rmatte> solution is posted here message/48484#48484 if anyone is curious

[05-May-2010 11:36:59] <willwh> hi rmatte :)

[05-May-2010 11:37:14] <rmatte> hi

[05-May-2010 11:37:14] <willwh> I was talking to someone in this channel re: remote control in windows RDP sessions

[05-May-2010 11:37:16] <willwh> from linux

[05-May-2010 11:37:40] <rmatte> not sure if it was me or not lol

[05-May-2010 11:37:43] <willwh> haha

[05-May-2010 11:37:49] <willwh> well - have you had any success there?

[05-May-2010 11:38:00] <willwh> someone said that the terminal services app didn't support something

[05-May-2010 11:38:01] <rmatte> you mean with connecting to a windows server via RDP from Linux?

[05-May-2010 11:38:09] <willwh> well, that works fine

[05-May-2010 11:38:19] <willwh> but then having someone remote control *that* session

[05-May-2010 11:38:23] <willwh> always hangs me up

[05-May-2010 11:39:27] <rmatte> oh

[05-May-2010 11:39:43] <rmatte> Well, how are they trying to remote control it?

[05-May-2010 11:39:58] <rmatte> only 1 person can be logged in to a session via RDP at a time

[05-May-2010 11:40:08] <rmatte> unless you're using some third party app for the second connection

[05-May-2010 11:40:17] <rmatte> something like sharemypc.com's app or whatever

[05-May-2010 11:41:00] <rmatte> Otherwise the second the other person connects via RDP and authenticates they take over the session and you get booted out

[05-May-2010 11:41:03] <rmatte> that's normal behaviour

[05-May-2010 11:45:24] <rmatte> now, time to test the upgrade...

[05-May-2010 11:54:00] <rhettardo> we have a  client that uses a terminal server on OS X and it lets you do screen sharing

[05-May-2010 11:54:06] <rhettardo> thats about the only part i like

[05-May-2010 11:59:32] <rmatte> ah

[05-May-2010 12:02:29] <rhettardo> its a bit comical that all my issues are with windows clients using the terminal server on a mac

[05-May-2010 12:44:13] <coofamani> you can share termserv sessions between two connected users in server

[05-May-2010 12:44:33] <coofamani> bring up the task manager and click users, you can share a single session there

[05-May-2010 12:45:31] <rhettardo> :D

[05-May-2010 13:21:40] <rmatte> coofamani: cool, didn't know about that

[05-May-2010 13:22:24] <jb> is it possible to monitor processes from just ONE machine?

[05-May-2010 13:22:32] <jb> ie, 100 boxes may have an instance of blah.exe

[05-May-2010 13:22:39] <jb> but i only want it to be monitored on one box

[05-May-2010 13:23:06] <rhettardo> manually add the process and lock it

[05-May-2010 13:23:10] <jb> it seems like when you add a process, next time the device remodels, if that process is detected on that system, it gets added..

[05-May-2010 13:23:22] <jb> oops, had this window on ignore :/

[05-May-2010 13:28:32] <jb> WMI process discovery is currently not supported.

[05-May-2010 13:33:19] <rhettardo> hm

[05-May-2010 13:33:35] <rhettardo> i cant remove anything from event console

[05-May-2010 13:34:01] <coofamani> is your mysql database working, and does the zenoss user have the correct permissions in mysql?

[05-May-2010 13:34:42] <coofamani> mysql> show grants for zenoss

[05-May-2010 13:34:54] <coofamani> or whatever username you use for the db

[05-May-2010 13:40:13] <rmatte> sometimes it just takes a while to remove it

[05-May-2010 13:40:21] <rmatte> I sometimes have to wait like 30 seconds

[05-May-2010 13:40:59] <rhettardo> db perms are fine.  its only started happening since my vm host failed hard yesterday

[05-May-2010 13:41:28] <rhettardo> so far all i've noticed is i cant remove events.

[05-May-2010 13:41:40] <rhettardo> no errors in any logs that i see either

[05-May-2010 13:43:17] <rhettardo> a few seconds after i click the X button i get a 'server connection error' modal that disappears quickly

[05-May-2010 13:43:56] <rmatte> ok, so reload the console

[05-May-2010 13:44:00] RoAk is now known as RoAkSoAx

[05-May-2010 13:44:02] <rmatte> the events should have been removed anyways

[05-May-2010 13:44:05] <rmatte> if not, try it again

[05-May-2010 13:44:13] <rmatte> that happens to me too sometimes on one of my busier servers

[05-May-2010 13:44:28] <rmatte> just takes a few tries

[05-May-2010 13:44:36] <rmatte> it even happened with the old console

[05-May-2010 13:47:53] <rhettardo> no dice

[05-May-2010 13:54:01] <rmatte> weird

[05-May-2010 13:54:10] <rmatte> you've tried doing a full restart of Zenoss and MySQL?

[05-May-2010 13:54:16] <rhettardo> yeah

[05-May-2010 13:54:31] <rmatte> also, make sure you don't have any extra mysql processes running (that's happened to me a couple times lately)

[05-May-2010 13:54:37] <rmatte> ps aux | grep mysqld.bin

[05-May-2010 13:54:42] <rmatte> or just mysqld

[05-May-2010 13:54:47] <rmatte> there should only be 1 instance

[05-May-2010 13:59:23] <rhettardo> yeah just restarted everything again and checked for dupe processes (there were none) and same thing

[05-May-2010 13:59:57] <rmatte> hmmm

[05-May-2010 14:00:35] <rmatte> If your VM failed hard yesterday there might be some lock file or something that didn't get removed but should have

[05-May-2010 14:07:15] mrayzenoss1 is now known as mrayzenoss

[05-May-2010 14:33:32] <MTecknology> What's the difference between "Stack Installers" and "Native Stacks (Debian/Ubuntu)"

[05-May-2010 14:35:42] <rmatte> Stack Installer is basically just a .bin file that you chmod +x then execute

[05-May-2010 14:35:49] <rmatte> the native stacks are .debs

[05-May-2010 14:36:06] <rmatte> basically the only difference is that you'll be able to interact with the package via aptitude if you install via the .deb

[05-May-2010 14:36:20] <rmatte> so to remove the package you'd do aptitude remove zenoss for instamce

[05-May-2010 14:36:21] <MTecknology> thanks :)

[05-May-2010 14:36:41] <rmatte> with the stack installer you'd do /usr/local/zenoss/uninstall

[05-May-2010 14:38:32] <MTecknology> nifty :)

[05-May-2010 14:38:56] <rmatte> yup

[05-May-2010 14:39:13] <MTecknology> packaging isn't much fun..

[05-May-2010 14:39:19] <twm1010> rmatte: oddly enough when i installed on top of the existing zenoss install, it force me to use  a different startup script than zenoss-stack

[05-May-2010 14:39:28] <MTecknology> I enjoy it - but dang it's hard to learn

[05-May-2010 14:40:01] <MTecknology> I wonder how hard it is to setup snmp on my router...

[05-May-2010 14:42:07] <MTecknology> Setting up zenoss-stack (2.5.2-1) ...

[05-May-2010 14:42:07] <MTecknology> Problem running post-install step. Installation may not complete correctly Error renaming /usr/local/zenoss/mysql/my.cnf-backup to /usr/local/zenoss/mysql/my.cnf

[05-May-2010 14:42:17] <MTecknology> Is that because it's a fresh install?

[05-May-2010 14:47:33] <coofamani> twm1010: in zendmd: dmd.ZenEventManager.password

[05-May-2010 14:47:48] <coofamani> twm1010: thats how you get the password for the db, I knew I had it around here

[05-May-2010 14:48:57] <rmatte> twm1010: odd

[05-May-2010 14:49:22] <rmatte> MTecknology: you are installing it as root I hope?

[05-May-2010 14:49:41] <MTecknology> there - purged package, wiped system, and installed again - looks better now :)

[05-May-2010 14:49:43] <MTecknology> rmatte: ya

[05-May-2010 14:49:53] <rmatte> k

[05-May-2010 14:50:12] <MTecknology> wiped /usr/local/zenoss - not system :P

[05-May-2010 14:51:32] <MTecknology> rmatte: I thnk I did something to break the install - not sure - but it's better now :)

[05-May-2010 14:51:59] <twm1010> you still trying to run this in a VM?

[05-May-2010 14:52:00] <twm1010> :)

[05-May-2010 14:52:17] <MTecknology> ya

[05-May-2010 14:52:27] <MTecknology> but now it's on my vm host

[05-May-2010 14:52:36] <MTecknology> there's 8GB available for all 5 VM's

[05-May-2010 14:52:45] <MTecknology> so I'm giving this one 2.3GB

[05-May-2010 14:52:53] <twm1010> ohhhh i thought you were doing this from a laptop

[05-May-2010 14:52:57] <MTecknology> i was

[05-May-2010 14:53:08] <rhettardo> that sounds painful

[05-May-2010 14:53:09] <MTecknology> - when i was deciding what I thought :)

[05-May-2010 14:53:19] <MTecknology> what's the default user/pass?

[05-May-2010 14:53:46] <MTecknology> somehow I don't get to the setup screen right after install

[05-May-2010 14:53:47] <rhettardo> zenoss/zenoss ?

[05-May-2010 14:53:58] <MTecknology> I tried that

[05-May-2010 14:54:03] <mrayzenoss> admin/zenoss

[05-May-2010 14:54:04] <rhettardo> wait i didnt really read waht you said

[05-May-2010 14:54:07] <rhettardo> yeah

[05-May-2010 14:54:21] <MTecknology> yay

[05-May-2010 14:54:22] <rhettardo> my head was elsewhere, was thinking db

[05-May-2010 14:54:25] <MTecknology> thanks

[05-May-2010 14:55:39] <MTecknology> discovery time :D

[05-May-2010 14:55:53] <MTecknology> and this time it doesn't need to run over openvpn either

[05-May-2010 14:57:49] <MTecknology> this this is MUCH happier now

[05-May-2010 14:58:22] <MTecknology> it also feels a lot different..

[05-May-2010 14:58:26] <rhettardo> i wish mine was happy :(

[05-May-2010 15:00:27] <MTecknology> Now I'm trying to remember how mr mrayzenoss showed me how to set the template

[05-May-2010 15:01:19] <MTecknology> there we go :)

[05-May-2010 15:01:29] <MTecknology> this is ever prettier when runnin correctly

[05-May-2010 15:02:30] <MTecknology> it's only complaining about ssh now..

[05-May-2010 15:02:42] <MTecknology> can I tell it to stop scanning all devices?

[05-May-2010 15:03:25] <rmatte> well, you would have had to specifically tell it to do the scan in the first place

[05-May-2010 15:03:40] <rmatte> you can cancel it by going to Settings -> Jobs

[05-May-2010 15:05:21] <MTecknology> rmatte: I only wan tto keep it from running for a little bit

[05-May-2010 15:08:01] <MTecknology> I'm trying to fix the error "SshUserAuth: no password found -- has zCommandPassword been set?"

[05-May-2010 15:08:14] <MTecknology> I suppose I can deal with that when I get back..

[05-May-2010 15:09:26] <rmatte> that's just because they are in the /Discovered class

[05-May-2010 15:09:30] <rmatte> move them all to proper classes

[05-May-2010 15:09:45] <rmatte> and the error will go away

[05-May-2010 15:10:08] <rmatte> hmmm, the alterpoint website is down, lovely

[05-May-2010 15:10:29] <MTecknology> rmatte: I moved them to either /Server/Linux or /Network/Router

[05-May-2010 15:11:28] <MTecknology> I need to add a shared key for root I guess :P

[05-May-2010 15:11:34] <MTecknology> I don't allow root login

[05-May-2010 15:11:57] <rmatte> no, it shouldn't even be trying to do SSH monitoring in those classes

[05-May-2010 15:12:07] <MTecknology> oh..

[05-May-2010 15:12:09] <MTecknology> ok

[05-May-2010 15:12:50] <MTecknology> I'll just wait

[05-May-2010 15:25:19] <MTecknology> Is there any point in monitoring the system that this is running on?

[05-May-2010 15:27:44] <MTecknology> It looks like snmp is bing found where ti needs to be - unfortunately, it also looks like things are too restricted - no information about the devices is found other than their OS-version and name

[05-May-2010 15:32:08] <MTecknology> Error reading value for "....." on .....

[05-May-2010 15:33:40] <MTecknology> Error reading value for "" on  (old .1.3.6.1.4.1.2021.x.x.x is bad)

[05-May-2010 15:34:10] <MTecknology> yay for protocols that I fully don't understand

[05-May-2010 15:35:22] <MTecknology> Whould should I set that value to?

[05-May-2010 15:45:39] <rmatte> well hold up

[05-May-2010 15:45:47] <rmatte> what type of device are you seeing that on?

[05-May-2010 15:46:12] <rmatte> that's snmp by the way, you'll want to learn about it

[05-May-2010 15:47:55] <MTecknology> Linux Hosts - Ubuntu 10.04

[05-May-2010 15:52:06] <MTecknology> I do intend to learn about snmp

[05-May-2010 15:54:57] <MTecknology> rmatte: any idea offhand what I'm doing wrong? I can post my snmo config

[05-May-2010 15:57:55] <MTecknology> http://dpaste.com/191313/

[05-May-2010 15:58:55] <MTecknology> Should I s/com2sec paranoid/com2sec readonly/ ?

[05-May-2010 16:02:42] <twm1010> I usually just add the one liner rocommunity mystringhere

[05-May-2010 16:03:00] <twm1010> unless you're super concerned about what info that could expose, i wouldn't worry

[05-May-2010 16:03:11] <MTecknology> oh

[05-May-2010 16:03:34] <rmatte> it's probably your snmp config

[05-May-2010 16:03:38] <twm1010> you can use the file, to restrict a community string, or host, or user for snmpv3 to certain OID's in the tree

[05-May-2010 16:04:03] <twm1010> if restriction is not an issue, skip that section, and just add read-only community

[05-May-2010 16:04:26] <rmatte> This is the config I use: http://pastebin.com/BVzbA01y

[05-May-2010 16:04:28] <twm1010> and if it's ubuntu don't forget to change the interface the daemon listens on

[05-May-2010 16:04:47] <MTecknology> twm1010: oh?

[05-May-2010 16:04:50] <rmatte> Change the trapsink IPs to the IP of your zenoss server, change your-key to the snmp string you want

[05-May-2010 16:04:57] <rmatte> and change the syscontact and location details

[05-May-2010 16:05:03] <rmatte> then restart snmpd and you're in business

[05-May-2010 16:05:28] <twm1010> yeah, exactly, see how he has rocommunity and rwcommunity there? that's reall all you need, if you want traps then fill out the trapsink lines

[05-May-2010 16:05:35] <twm1010> you can also use snmpconf to build this file for you

[05-May-2010 16:06:32] <MTecknology> oh

[05-May-2010 16:07:18] <MTecknology> I'm just waiting for it to go check the systems again

[05-May-2010 16:07:40] <MTecknology> that paranoid part is probably what was breaking it..

[05-May-2010 16:07:52] <rmatte> yup

[05-May-2010 16:16:26] <MTecknology> Now while I wait I just need figure out how to set the locatin on a map of a location

[05-May-2010 16:17:42] <mrayzenoss> after you've added the Location, edit the address

[05-May-2010 16:17:55] <mrayzenoss> put something that Google Maps will recognize

[05-May-2010 16:18:11] * mrayzenoss heads home

[05-May-2010 16:27:44] <MTecknology> Well.. the errors about SNMP went away. Too bad it's still not showing any information gathered.

[05-May-2010 16:28:25] <rhettardo> it will take a few polling cycles

[05-May-2010 16:28:26] <MTecknology> I am getting a warning about  threshold of low swap not met: current value .....

[05-May-2010 16:28:32] <MTecknology> ok

[05-May-2010 16:29:17] <MTecknology> This message popping up looks like an easy to fix one

[05-May-2010 16:30:41] <MTecknology> there we go :)

[05-May-2010 16:31:01] <MTecknology> ok - I'm with the going home idea :D - ttyal

[05-May-2010 16:31:06] <MTecknology> thanks for all the help :)

[05-May-2010 16:47:46] <rmatte> np

[06-May-2010 00:00:30] [disconnected at Thu May  6 00:00:30 2010]

[06-May-2010 00:00:31] [connected at Thu May  6 00:00:31 2010]

[06-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[06-May-2010 01:25:12] <voxter> Hey guys, whats shakin'

[06-May-2010 01:25:23] <fragfutter> nothing but the rain

[06-May-2010 02:40:31] <voxter> Any of you guys use the new WMI zenpack?

[06-May-2010 10:22:38] <bbibeault> Zenoss 3.0 Alpha #3 has launched, and the QA Test Day is currently in progress in #zenoss-testing.

[06-May-2010 10:24:25] <aclark> bbibeault: GO

[06-May-2010 11:32:30] <mistich> anyone from zenoss around

[06-May-2010 11:50:36] <kericksonzenoss> Most of the Zenoss dev people are in a "properly using third party code" meeting - so our attention is diverted

[06-May-2010 11:54:53] <mistich> I need to get ahold of eric asap he's killing my box

[06-May-2010 12:00:11] <kericksonzenoss> @mistich -> edahl?

[06-May-2010 12:01:34] <kericksonzenoss> @mistich -> or eedgar?

[06-May-2010 12:02:41] <mistich> I got eedgar thanks

[06-May-2010 12:03:24] <kericksonzenoss> @mistich -> edahl stopped it

[06-May-2010 12:03:39] <mistich> eedgar stopped it

[06-May-2010 12:04:02] <mistich> edahl don't think he takes support calls any more :)

[06-May-2010 12:04:28] <kericksonzenoss> nope

[06-May-2010 12:30:23] <cgibbons> hmmm

[06-May-2010 12:40:58] <coofamani> http://www.puppetlabs.com/blog/zenoss-and-puppet-labs-partner-to-deliver-integrated-it-monitoring-and-automation-for-next-generation-datacenters/

[06-May-2010 12:41:01] <coofamani> heh

[06-May-2010 12:41:25] <venturaville> interesting

[06-May-2010 12:42:59] <coofamani> Did you know about it?

[06-May-2010 12:43:08] <venturaville> first I heard about it

[06-May-2010 12:43:27] <venturaville> it is light on details though

[06-May-2010 12:43:41] <coofamani> lol, it links back to your zenpack. thanks for playing :)

[06-May-2010 12:45:43] <venturaville> this article sounds quite a bit more involved than the little zenpack I made

[06-May-2010 12:46:20] <coofamani> Maybe its 'forward thinking'. Only meat I see is your pack

[06-May-2010 12:46:33] <venturaville> that is my guess

[06-May-2010 12:46:45] <venturaville> does explain why recruiters started calling my home phone like crazy though

[06-May-2010 12:46:54] <coofamani> did they? nice

[06-May-2010 12:48:05] <coofamani> aw man, Guest40826 is here.. I cant stand that guy

[06-May-2010 12:48:22] <coofamani> 40825 is OK, but man this guy

[06-May-2010 12:48:32] <venturaville> heh

[06-May-2010 12:48:37] <Guest40826> Perhaps, I'll log out and then back in.  That'll show you.

[06-May-2010 12:48:48] <coofamani> ha

[06-May-2010 12:48:54] <Guest40826> Lurk engaged.

[06-May-2010 12:52:22] <chemist> I am seeing lots of these errors in my zeo.log

[06-May-2010 12:52:23] <chemist> http://pastebin.com/KhzC9LgW

[06-May-2010 12:57:56] <chemist> anyone can offer any pointers?

[06-May-2010 12:58:07] <coofamani> I've never seen that

[06-May-2010 12:58:59] <chemist> me neither :)

[06-May-2010 13:24:09] <chemist> guess no one is here

[06-May-2010 13:43:30] <twm1010> whats up?

[06-May-2010 13:44:04] <chemist> ever seen this? http://pastebin.com/KhzC9LgW

[06-May-2010 13:49:06] <twm1010> hrmm, no ideas here

[06-May-2010 14:08:03] mrayzenoss1 is now known as mray

[06-May-2010 14:08:14] mray is now known as mrayzenoss

[06-May-2010 14:08:17] mrayzenoss is now known as mray

[06-May-2010 14:30:42] <voxter> anyone using the public alpha?

[06-May-2010 14:30:54] <rhettardo> yes

[06-May-2010 14:31:22] <voxter> rhettardo: I am in love with the new interface.  Is it compatible with most zenpacks? or have there been significant architectural changes?

[06-May-2010 14:31:41] <voxter> rhettardo: basically, i am contemplating putting the alpha on my main zenoss collector i like it so much.

[06-May-2010 14:32:06] <rhettardo> not sure.  i dont use many zenpacks and i've only been using it very lightly and starting from scratch when a new release comes out

[06-May-2010 14:33:48] <mray> most ZenPacks should continue to work, we need people to alpha/beta test them

[06-May-2010 14:35:10] <voxter> mray: so then, you would encourage me to try this. :)

[06-May-2010 14:35:34] <mray> yes

[06-May-2010 14:39:37] <rhettardo> ha

[06-May-2010 14:52:38] <mray> FYI, bigegor's NetWare SNMP ZenPack has been updated: docs/DOC-3499

[06-May-2010 15:39:09] <voxter> mrayzenoss: doh, it appears as though the alpha is not available in the yum repository yet?

[06-May-2010 15:39:31] <mrayzenoss> no, we don't want anyone accidentally upgrading

[06-May-2010 15:40:42] <voxter> even in the /unstable/ branch?

[06-May-2010 15:40:45] <voxter> (which is empty)

[06-May-2010 15:41:00] <voxter> do you have an rpm i can manually install then?

[06-May-2010 15:41:12] <mrayzenoss> docs/DOC-5854

[06-May-2010 15:41:56] <voxter> sweet! ill give it a whirl

[06-May-2010 15:42:09] <voxter> are there any known major bugs that i should watch for? i use a decent amount of zenpack stuff

[06-May-2010 15:43:40] <mrayzenoss> it's an alpha, it's full of bugs :)

[06-May-2010 15:44:07] <mrayzenoss> http://dev.zenoss.org/trac/report/3

[06-May-2010 15:44:22] <voxter> good to know!

[06-May-2010 15:55:08] <klinstifen> afternoon all

[06-May-2010 15:55:30] <klinstifen> anyone have this issue: multigraphs suddenly stop graphing?

[06-May-2010 15:55:43] <klinstifen> the device they are attached to _is_ graphing correctly

[06-May-2010 15:56:08] <chemist> mrayzenoss: how do I change my trac password?

[06-May-2010 16:01:11] <mrayzenoss> chemist: you can't, only the Trac admin can

[06-May-2010 16:02:36] <chemist> oh, ok

[06-May-2010 16:02:50] <chemist> can I request a password?

[06-May-2010 16:03:55] <mrayzenoss> sure, pm me

[06-May-2010 16:07:59] <chemist> when setting up a remote collector, do I just sync the zenpack folder from master to collector?

[06-May-2010 16:13:40] <voxter> mrayzenoss: can i request a trac account? there doesnt seem to be a way to set one up myself.

[06-May-2010 16:16:04] <mrayzenoss> voxter: email me at mray@zenoss.com with the user/pass you want and I'll hook you up

[06-May-2010 16:27:54] <MTecknology> so.. now that I'm seeing this thing up and running after a while I like it a lot. The only thing I'm not found of is how heavy it is on the system.

[06-May-2010 16:28:09] <MTecknology> making snmp secure enough was easy too

[06-May-2010 16:32:02] <voxter> MTecknology: what'd you do for secure snmp?. I totally forgot it was transmitted in plaintext.

[06-May-2010 16:33:00] <MTecknology> voxter: well... secure being loosely defined :P

[06-May-2010 16:33:38] <MTecknology> voxter: I want to use snmp 3 so I can at least encrypt the traffic - but iptables helps

[06-May-2010 16:35:03] <voxter> wish i knew why sometimes -some- graphs stop collecting

[06-May-2010 16:35:19] <voxter> for instance fping stopped collecting for a day yesterday while the snmp pollers worked fine

[06-May-2010 16:35:39] <MTecknology> can I add graphs to the dashboard?

[06-May-2010 16:35:44] <rhettardo> v3 dosnt have to be plaintext

[06-May-2010 16:36:23] <MTecknology> I wonder what version I have installed

[06-May-2010 16:36:47] <MTecknology> odd..

[06-May-2010 16:37:52] <MTecknology> aptitude is telling me I have 5.4.2.1~dfsg0ubuntu1-0ubuntu2

[06-May-2010 16:38:15] <MTecknology> aside from being a wacky version number :S - isn't v3 the newest?

[06-May-2010 16:38:55] <rhettardo> your snmpd should surely support using v3.  you'd just need to configure it.  i personally dont use it

[06-May-2010 16:40:45] <MTecknology> YAY!

[06-May-2010 16:40:47] <rhettardo> it does support authentication and encryption though whereas v1 and v2c do not

[06-May-2010 16:40:53] <MTecknology> My network map is finally filled out

[06-May-2010 16:41:00] <rhettardo> nice

[06-May-2010 16:45:09] <chemist> when setting up a remote collector, do I just sync the zenpack folder from master to collector?

[06-May-2010 17:35:33] <mrayzenoss> Just a heads up, I've added a second entry to my SSH ZenPack tutorial series: docs/DOC-5865

[06-May-2010 17:35:44] <mrayzenoss> feel free to play along at home

[06-May-2010 17:37:54] <willwh> nice

[06-May-2010 17:40:15] <mrayzenoss> oh sweet, just found the steps for part #3 of the tutorial.  system_profiler is a powerful tool for OSX

[06-May-2010 20:30:10] <voxter> you guys use the windows event log conenctor for zenoss? curious on your opinion of something...

[06-May-2010 20:44:09] <voxter> This might seem absurd, but how do i add a new device in the alpha release??

[06-May-2010 20:44:29] <voxter> nm, found it! lol

[06-May-2010 20:50:24] <voxter> I thought the idea was that i could drag and drop things around in the new gui? in order to change a class i still have to go into each individual device?

[06-May-2010 20:57:11] <voxter> Is everyone gone for the day? must be.

[07-May-2010 00:00:30] [disconnected at Fri May  7 00:00:30 2010]

[07-May-2010 00:00:31] [connected at Fri May  7 00:00:31 2010]

[07-May-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[07-May-2010 00:20:53] <MTecknology> May  6 10:30:46 obtinio snmpd[29405]: Connection from UDP: [69.164.192.143]->[208.107.60.226]:-19796

[07-May-2010 00:20:59] <MTecknology> so..?

[07-May-2010 00:21:30] <MTecknology> oh..

[07-May-2010 00:21:53] <MTecknology> How can I allow access from a remote IP instead of restricing to LAN?

[07-May-2010 01:47:22] <fragfutter> i want to lower the priority of some perf/interface events (not all). Can someone tell me if i got this right? Goto /Events/Perf/Interface, Add Mapping, Event Class Key=defaultmapping, Rule "evt.device=="foobar" and evt.component=="interfacename", set zProperties zEventSeverity = Info

[07-May-2010 01:48:02] <fragfutter> Will this rule be evaluated for every incoming event (defaultmapping) or only events in the class perf/interface?

[07-May-2010 05:24:09] <JesperS> hi guys, quick question. What's the easiest way to move a device to a specific class, from command line? If it could be done by eg. wget it would be great.

[07-May-2010 05:33:42] <fragfutter> JesperS: you could go through a small python script

[07-May-2010 05:34:02] <fragfutter> JesperS: should be possible through ZEO or RPC interfaces

[07-May-2010 05:36:05] * fragfutter -> lunch

[07-May-2010 05:36:08] <JesperS> ok

[07-May-2010 05:36:13] <JesperS> I'll try to look at that

[07-May-2010 06:27:20] <JesperS> fragfutter (and others) any pointers on what kind of python script? Are there any examples/templates available?

[07-May-2010 06:27:39] <JesperS> I've been googl'ing some, and can't quite find anything usefull

[07-May-2010 06:44:33] <fragfutter> JesperS: http://pastebin.com/VjPLAFT6

[07-May-2010 06:45:06] <JesperS> Goddie, thank you.

[07-May-2010 06:45:32] <fragfutter> i classify my devices using that (with a longer deviceClassmap)

[07-May-2010 07:10:01] <fus10nx> off topic - but - Content Aware on Photoshop CS5 is prob. the most insane thing I've ever seen. Everyone should check it out. NOW.

[07-May-2010 07:10:38] <JesperS> hehe yea, it's pretty sweet, from what I've seen, haven't tried it yet.

[07-May-2010 07:10:56] <fus10nx> I just used it, its just insanely accurate

[07-May-2010 07:11:05] <fus10nx> if you do it properly you can get it exact, i mean its nuts

[07-May-2010 07:11:10] <JesperS> the first usefull new feature since PS3...

[07-May-2010 07:11:18] <fus10nx> haha yup

[07-May-2010 07:34:03] <klinstifen> morning all

[07-May-2010 07:34:14] <klinstifen> anyone ever have their multigrahs stop graphing?

[07-May-2010 07:41:19] <fus10nx> Not I

[07-May-2010 08:03:25] klinstifen_ is now known as klinstifen

[07-May-2010 10:06:25] mrayzenoss1 is now known as mrayzenoss

[07-May-2010 10:06:28] <felbane> hi all... is there a function/method that i can use in an event transform to correlate an ip address to a device?

[07-May-2010 10:10:43] <rmatte> felbane: well, explain a bit more

[07-May-2010 10:10:51] <rmatte> define: correlate

[07-May-2010 10:10:56] <rmatte> could mean a few things

[07-May-2010 10:11:26] <felbane> okay i have a syslog event coming in thats passing through an intermediary

[07-May-2010 10:12:11] <felbane> so i send the originating source ip as part of the message and then parse it out with a regex

[07-May-2010 10:12:55] <felbane> it works fine but when i set device to that ip, zenoss doesnt automatically match that with the device ip

[07-May-2010 10:13:20] <felbane> soim wondering if there is a getDeviceByIp(ipaddress) method i can use

[07-May-2010 10:20:07] <rmatte> You mean you want Zenoss to resolve the IP to a device name?

[07-May-2010 10:20:29] <felbane> yes

[07-May-2010 10:21:21] <felbane> i do evt.device = match.group(1) but zenoss doesnt match it to a device

[07-May-2010 10:22:33] <rmatte> Why wouldn't you do something like...

[07-May-2010 10:22:42] <rmatte> d = dmd.Devices.findDevice('30.30.1.25')

[07-May-2010 10:22:48] <rmatte> d.getDeviceName()

[07-May-2010 10:22:56] <rmatte> to get the actual hostname

[07-May-2010 10:23:16] <rmatte> obviously replace the IP with your IP variable

[07-May-2010 10:23:48] <rmatte> the whole thing would look like...

[07-May-2010 10:24:04] <rmatte> d = dmd.Devices.findDevice(ipVariable)

[07-May-2010 10:24:13] <rmatte> evt.device = d.getDeviceName()

[07-May-2010 10:25:11] <rmatte> set the evt.ipAddress as well

[07-May-2010 10:26:19] <rmatte> evt.ipAddress = ipVariable

[07-May-2010 10:30:03] <felbane> i will try that

[07-May-2010 10:58:19] <twm1010> ahoy

[07-May-2010 10:59:16] <twm1010> has anyone configured alerting for when the operational status of an interface changes?

[07-May-2010 11:09:51] <rmatte> twm1010: yeh, I have that configured

[07-May-2010 11:10:02] <rmatte> twm1010: you're talking about actually polling operational status I assume?

[07-May-2010 11:11:53] <rmatte> twm1010: docs/DOC-2494

[07-May-2010 11:11:56] <rmatte> works like a charm

[07-May-2010 11:12:22] <MTecknology> May  6 10:30:46 obtinio snmpd[29405]: Connection from UDP: [69.164.192.143]->[208.107.60.226]:-19796    -   I guess my snmp config isn't allowing anything other than LAN. How can I allow this?

[07-May-2010 11:12:25] <rmatte> plus with the new way that interface status is handled, it'll actually change the status of any interfaces that goes down to red on the OS tab

[07-May-2010 11:12:28] <rmatte> which is quite nice

[07-May-2010 11:12:55] <rmatte> though it relies on the name of the interface, so on some of my devices that have multiple interfaces with the same name it shows them as all being down

[07-May-2010 11:13:03] <rmatte> would be nice if it did it with ifIndex instead

[07-May-2010 11:13:04] <rmatte> but oh well

[07-May-2010 11:13:11] <rmatte> it works properly in the majority of cases

[07-May-2010 11:14:30] <rmatte> MTecknology: what do you mean isn't allowing anything but LAN?

[07-May-2010 11:14:37] <rmatte> you mean on the device that you're monitoring?

[07-May-2010 11:14:41] <MTecknology> ya

[07-May-2010 11:14:48] <rmatte> what kind of device is it?

[07-May-2010 11:14:52] <rmatte> Linux box?

[07-May-2010 11:14:56] <MTecknology> ya

[07-May-2010 11:15:10] <rmatte> then it would have to either be something in your iptables or something in the snmpd config file

[07-May-2010 11:15:12] <rmatte> check both

[07-May-2010 11:15:30] <MTecknology> it's the config - I don't know what to change though

[07-May-2010 11:15:44] <rmatte> pastebin it?

[07-May-2010 11:15:55] <rmatte> Hard to troubleshoot something like that blind ;)

[07-May-2010 11:16:05] <MTecknology> alrighty

[07-May-2010 11:18:21] <MTecknology> http://pastebin.com/6SQxb0sn

[07-May-2010 11:18:52] <rmatte> remove your paranoid lines for starters

[07-May-2010 11:19:15] <rmatte> see if that fixes it

[07-May-2010 11:19:24] <rmatte> Zenoss can't monitor properly with paranoid lines anyways

[07-May-2010 11:19:30] <MTecknology> those were in there by default - I figured I'd keep the default as close to the same as possible - give me a second

[07-May-2010 11:19:30] <rmatte> it won't have access to all of the OIDs that it needs

[07-May-2010 11:20:32] <twm1010> rmatte: thanks fo rhte info, that's perfect

[07-May-2010 11:20:35] <MTecknology> I thought that com2sec readonly  default xxx meant that it only used the readonly lines

[07-May-2010 11:20:54] <rmatte> twm1010: no problem, I've been using that since 2.3.3, it works great

[07-May-2010 11:21:15] <MTecknology> they're out of there now

[07-May-2010 11:21:37] <twm1010> interesting... so i'd really have to do this for every interface type template is well

[07-May-2010 11:21:43] <rmatte> MTecknology: not sure... here's the config that we use: http://pastebin.com/192DgdxJ

[07-May-2010 11:21:47] <twm1010> so if i want to know if a T1 controller goes down, instead of a serial interface, etc

[07-May-2010 11:22:20] <MTecknology> rmatte: what OS you on there?

[07-May-2010 11:22:25] <rmatte> yeh, you need to do it to every interface template

[07-May-2010 11:22:31] <rmatte> MTecknology: Ubuntu Server

[07-May-2010 11:22:34] <MTecknology> oh

[07-May-2010 11:22:56] <MTecknology> what's trapsink?

[07-May-2010 11:23:02] <rmatte> to send traps back to Zenoss

[07-May-2010 11:23:14] <rmatte> you set the IP of your Zenoss server

[07-May-2010 11:23:25] <MTecknology> what's a trap?

[07-May-2010 11:23:29] <rmatte> and set the SNMP string anywhere that it says "your-key"

[07-May-2010 11:23:49] <rmatte> I generally just comment out the rwcommunity unless absolutely necessary

[07-May-2010 11:24:05] <rmatte> an SNMP trap is an alert send from the device to the monitoring system

[07-May-2010 11:24:24] <rmatte> instead of polling the device via SNMP, the device sends an SNMP trap back to alert on various events

[07-May-2010 11:24:30] <MTecknology> oh

[07-May-2010 11:24:31] <MTecknology> OH!.....

[07-May-2010 11:24:38] <rmatte> for instance, if the device reboots you'll get an snmp_coldStart trap

[07-May-2010 11:24:38] <MTecknology> com2sec readonly  default

[07-May-2010 11:24:45] <MTecknology> that 'default' is the network

[07-May-2010 11:25:28] <MTecknology> the default must be lan only

[07-May-2010 11:25:41] <rmatte> it's possible

[07-May-2010 11:26:15] <MTecknology> let's find out :)

[07-May-2010 11:26:27] <rmatte> doing SNMP over the public internet isn't exactly recommended though

[07-May-2010 11:26:43] <rmatte> unless it's snmpv3

[07-May-2010 11:26:53] <rmatte> otherwise you should be using a tunnel

[07-May-2010 11:27:45] <MTecknology> Once I get it working without spewing errors all over I figure I'll worry about encryption

[07-May-2010 11:28:56] <MTecknology> I got >2M logs in errors last night

[07-May-2010 11:37:14] <rmatte> 2MB

[07-May-2010 11:37:17] <rmatte> or 2 million?

[07-May-2010 11:37:18] <rmatte> lol

[07-May-2010 11:38:35] <MTecknology> hu?

[07-May-2010 11:38:48] <rmatte> what is >2M?

[07-May-2010 11:38:55] <MTecknology> MB

[07-May-2010 11:38:57] <rmatte> k

[07-May-2010 11:39:02] <felbane> rmatte: findDevice doesnt work by IP address

[07-May-2010 11:39:16] <rmatte> felbane: it worked for me in zendmd

[07-May-2010 11:39:29] <felbane> which version? im on 2.4.5

[07-May-2010 11:39:41] <rmatte> felbane: ohhhhh, one second, there's a patch you'll need...

[07-May-2010 11:40:16] mrayzenoss1 is now known as mrayzenoss

[07-May-2010 11:40:26] <rmatte> felbane: zenpatch 14049

[07-May-2010 11:40:29] <rmatte> then restart zenoss

[07-May-2010 11:40:34] <rmatte> it'll work after that

[07-May-2010 11:40:44] <MTecknology> This is beating the crap out of me...

[07-May-2010 11:40:45] <MTecknology> May  7 11:30:44 obtinio snmpd[8226]: Connection from UDP: [69.164.192.143]->[208.107.60.226]:-26648

[07-May-2010 11:40:52] <rmatte> Also, if you're still using 2.4.5 you should apply these patches: docs/DOC-4434

[07-May-2010 11:41:02] <MTecknology> my zenoss system is inside 208.107.60.226

[07-May-2010 11:41:19] <rmatte> MTecknology: no where in that message does it say that it's being blocked

[07-May-2010 11:41:22] <MTecknology> but obtinio is 69.164.192.143

[07-May-2010 11:41:37] <MTecknology> oh...

[07-May-2010 11:41:42] <rmatte> it's just stating that a connection is being made

[07-May-2010 11:41:49] <MTecknology> ok..

[07-May-2010 11:42:18] <rmatte> If you're getting lots of debug messages relating to SNMP then it's because of the paranoid lines

[07-May-2010 11:42:50] <MTecknology> the paranoid lines are gone

[07-May-2010 11:43:00] <rmatte> what exactly are you seeing that's an issue?

[07-May-2010 11:43:09] <rmatte> you've never really stated the specifics of your issue

[07-May-2010 11:43:35] <MTecknology> I thought those were blocks..

[07-May-2010 11:43:43] <rmatte> no

[07-May-2010 11:43:49] <MTecknology> yay

[07-May-2010 11:44:33] <rmatte> lol

[07-May-2010 11:45:33] <MTecknology> so now I want it not logging those..

[07-May-2010 11:45:50] <MTecknology> mostly because that's a log of a packate flying around and I don't care :P

[07-May-2010 11:45:55] <MTecknology> I have this - SNMPDOPTS='-Lsd -Lf /dev/null -u snmp -g snmp -I -smux -p /var/run/snmpd.pid'

[07-May-2010 11:45:59] <rmatte> double click on one of them, how are they coming in, syslog?

[07-May-2010 11:46:17] <MTecknology> double click?

[07-May-2010 11:46:26] <rmatte> where are you seeing those?

[07-May-2010 11:46:30] <MTecknology> oh - that's the line I get in syslog

[07-May-2010 11:46:31] <rmatte> I assumed you're seeing them in Zenoss?

[07-May-2010 11:46:45] <MTecknology> I get nightly emails of all logs

[07-May-2010 11:47:05] <rmatte> that connection tracking is meant for security purposes

[07-May-2010 11:47:13] <rmatte> 2MB is nothing really when it comes to logs

[07-May-2010 11:47:16] <MTecknology> ok

[07-May-2010 11:47:24] <rmatte> the system should just auto-roll them over anyways

[07-May-2010 11:48:04] <MTecknology> odd...

[07-May-2010 11:48:16] <MTecknology> it look slike logcheck should ignore those

[07-May-2010 11:48:30] <MTecknology> ^\w{3} [ :0-9]{11} [._[:alnum:]-]+ snmpd\[[0-9]+\]: Connection from UDP: \[[.0-9]{7,15}\]:[0-9]{4,5}$

[07-May-2010 11:49:39] <rmatte> http://gskinner.com/RegExr/

[07-May-2010 11:49:49] <rmatte> paste the log line in and then paste the regex at the top

[07-May-2010 11:49:50] <rmatte> no match

[07-May-2010 11:50:12] <MTecknology> I should fix it and file a logcheck bug :P

[07-May-2010 11:50:40] <rmatte> guess so

[07-May-2010 11:56:33] <felbane>  rmatte: just to be clear, d = dmd.Devices.findDevice('192.168.0.1') should make 'd' an object?

[07-May-2010 11:56:59] <felbane> rmatte: when i 'print d' after that, i get 'None'

[07-May-2010 11:57:53] <MTecknology> There we go :D  ^\w{3} [ :0-9]{11} [._[:alnum:]-]+ snmpd\[[0-9]+\]: Connection from UDP: \[[.0-9]{7,15}\]\-\>\[[.0-9]{7,15}\]:\-[0-9]{4,5}$

[07-May-2010 11:58:08] <MTecknology> That site doesn't seem to like ^ or $

[07-May-2010 12:01:53] <MTecknology> wow... zenoss spits out a butt load of snmp traffic

[07-May-2010 12:02:30] <MTecknology> generates a lot of it*

[07-May-2010 12:05:37] <rmatte> well obviously, it's a monitoring system

[07-May-2010 12:05:51] <MTecknology> ya

[07-May-2010 12:06:06] <rmatte> one of my Zenoss boxes polls 17000 OIDs every 5 mins

[07-May-2010 12:06:13] <rmatte> lots of traffic

[07-May-2010 12:06:45] <MTecknology> wow

[07-May-2010 12:14:42] <rmatte> 2.5.2 sure is heartbeat land

[07-May-2010 12:15:10] <rmatte> I think I'm just going to tune them out, absolutely useless

[07-May-2010 14:41:23] <twm1010> rmatte: around?

[07-May-2010 14:53:58] <rmatte> yeh

[07-May-2010 14:54:04] <rmatte> twm1010: what's up?

[07-May-2010 14:54:18] <twm1010> i've copied ethernetcsMacd over to ../Routers/Cisco

[07-May-2010 14:54:35] <twm1010> following the article, it says to create the datasource, and then the datapoint

[07-May-2010 14:54:59] <rmatte> you don't need to make datapoints anymore for standard SNMP datasources

[07-May-2010 14:55:02] <rmatte> that's done automatically now

[07-May-2010 14:55:04] <twm1010> ah, ok

[07-May-2010 14:55:21] <twm1010> but it would still be  ifOperStatus_ifOperStatus for the threshold right?

[07-May-2010 14:55:22] <rmatte> the only time you make datapoints is for things like command datasources

[07-May-2010 14:55:27] <rmatte> correcto

[07-May-2010 14:55:42] <rmatte> it'll automatically create a datapoints with the same name as the datasource

[07-May-2010 14:55:47] <rmatte> datapoint*

[07-May-2010 14:55:58] <rmatte> so ifOperStatus_ifOperStatus will be it

[07-May-2010 14:57:12] <twm1010> and you have to create the /Status/IpInterface event class

[07-May-2010 14:57:24] <twm1010> and set the zprop of that class to critical

[07-May-2010 14:58:22] <twm1010> ah no, severity on the threshold is sufficient

[07-May-2010 14:59:25] <rmatte> yup

[07-May-2010 15:02:05] <twm1010> maybe warning + 1 escalation is better... maybe.

[07-May-2010 15:02:10] <twm1010> but this is a 5 minute poll right?

[07-May-2010 15:03:16] <twm1010> is that OID valid for all types of interfaces?

[07-May-2010 15:03:30] <twm1010> t1 controllers, ds3 controllers, etc

[07-May-2010 15:05:58] <twm1010> haha, i just got 793 e-mails for every down PRI channel

[07-May-2010 15:07:50] <rmatte> man I hate catalyst switches, such a pain to find OIDs for

[07-May-2010 15:08:05] <rmatte> yes, 5 minute poll

[07-May-2010 15:08:13] <rmatte> valid for all types of interfaces

[07-May-2010 15:09:53] <twm1010> OK here's a pickle

[07-May-2010 15:10:03] <twm1010> I have to exclude the channels from voice T1's

[07-May-2010 15:10:11] <twm1010> which are of course, interfaces

[07-May-2010 15:10:44] <twm1010> so for each PRI T1, you have Serial 0/0/0:0 through Serial 0/0/0:22

[07-May-2010 15:11:00] <twm1010> and the D channel on Serial 0/0/0:23

[07-May-2010 15:11:04] <rmatte> ok, do they have a specific interface type?

[07-May-2010 15:11:15] <twm1010> propPointToPointSerial

[07-May-2010 15:11:24] <twm1010> but that's similar to a standard point to point T right?

[07-May-2010 15:11:32] <rmatte> do any interfaces that you actually want to monitor have that same type?

[07-May-2010 15:11:54] <twm1010> sure, a simple point to point would

[07-May-2010 15:12:26] <rmatte> hmmm

[07-May-2010 15:12:32] <rmatte> well, you'll have to get crafty

[07-May-2010 15:12:53] <rmatte> do you need to monitor the channels at all?

[07-May-2010 15:13:11] <rmatte> like, for utilization, or can you outright disable monitoring on them?

[07-May-2010 15:13:12] <twm1010> no, the ifOperStatus from the T1 controller is suffucient to tell if the T1 is up or down

[07-May-2010 15:13:51] <twm1010> here's the solution, i do the opposite on the voice gateways

[07-May-2010 15:14:03] <twm1010> i exclude them entirely, and go to the gig interfaces and set a local tempalte

[07-May-2010 15:14:08] <twm1010> for ethernetCSmacd

[07-May-2010 15:14:34] <rmatte> cool

[07-May-2010 15:21:15] <twm1010> I want as few duplicates of ethernetCSmacd as possible though

[07-May-2010 15:21:59] <rmatte> the flexibility of the interface templates is a bit limited unfortunately

[07-May-2010 15:33:50] <twm1010> that was awesome my team received a wopping 1500 emails in about 5 minutes

[07-May-2010 15:36:32] <twm1010> I'm giong to have the same kind of problem with say... an analog gateway, i think

[07-May-2010 15:37:17] <rmatte> lol

[07-May-2010 15:37:33] <twm1010> I'm trying to visualize the easiest template nesting

[07-May-2010 15:37:43] <twm1010> whether i want to create a sub-class under /Network/Router/Cisco

[07-May-2010 15:37:48] <twm1010> or a new one under /Network

[07-May-2010 15:37:55] <twm1010> because some devices are dedicated voice gateways and do only that

[07-May-2010 15:38:05] <twm1010> others are site routers that also happen to have one or two PRIs

[07-May-2010 15:38:47] <twm1010> I want the fewest "exceptions" as possible, I suppose.

[07-May-2010 15:42:09] <rmatte> yup

[07-May-2010 15:42:30] <mrayzenoss> wasn't there someone trying to do a ZenPack to do Linux filesystem IO?

[07-May-2010 15:42:51] <mrayzenoss> I seem to remember that, and can't find any trace

[07-May-2010 15:43:06] <rmatte> hmmm

[07-May-2010 15:43:06] <mrayzenoss> adding per/filesystem IO would be handy, especially for reporting

[07-May-2010 15:43:17] <rmatte> yeh

[07-May-2010 15:43:23] <rmatte> never heard of it

[07-May-2010 15:43:36] <mrayzenoss> and especially for tying it to virtualized devices

[07-May-2010 15:43:36] <rmatte> I'm working on a cisco catalyst template right now

[07-May-2010 15:43:50] <rmatte> think I finally found the correct CPU OID to use

[07-May-2010 15:46:02] <rmatte> Then I need to make a Cisco Content Switch template

[07-May-2010 15:46:39] <mrayzenoss> hmm… I got ZenPacks.community.CiscoASAVPN-1.1-py2.4.egg in my inbox today

[07-May-2010 15:46:59] <rmatte> how is that different from the current ASA pack?

[07-May-2010 15:47:15] <twm1010> yes,i'm interested, we're putting in two more ASA's soon :)

[07-May-2010 15:47:31] <rmatte> I monitor like 10 ASAs combined

[07-May-2010 15:47:53] <mrayzenoss> I'll forward it to you, I was going to get to it next week

[07-May-2010 15:47:54] <rmatte> the current pack works fine (I modified it a bit to look nicer)

[07-May-2010 15:48:28] <rmatte> k, I might not get around to looking at it all that soon though, lots of templates to make for a new client we have

[07-May-2010 15:48:54] <mrayzenoss> aight, no worries

[07-May-2010 15:49:24] <rmatte> I'll have some new ZenPacks for you in a couple of weeks

[07-May-2010 15:50:01] <mrayzenoss> yay

[07-May-2010 15:50:07] <mrayzenoss> I'm having fun writing my OS X one

[07-May-2010 15:50:10] <mrayzenoss> gonna do Chef next

[07-May-2010 15:50:18] <rmatte> APC environmental monitor, Liebert environmental monitor, Liebert AC, Cisco Catalyst, etc...

[07-May-2010 15:50:24] <mrayzenoss> then an SNMP one with a custom modeler and an extra tab

[07-May-2010 15:50:29] <rmatte> cool

[07-May-2010 15:50:30] <mrayzenoss> yay

[07-May-2010 15:50:56] <rmatte> I have some suped up cisco templates for router, switch, and wireless that you may be interested in including

[07-May-2010 15:51:04] <rmatte> the graphs just look nicer

[07-May-2010 15:51:12] <rmatte> and it does memory thresholding

[07-May-2010 15:51:24] <rmatte> and displays the threshold line at the correct spot on the graph

[07-May-2010 15:51:26] <rmatte> it's pretty nice

[07-May-2010 15:51:34] <mrayzenoss> looking forward to it

[07-May-2010 15:52:33] <twm1010> rmatte: sweet.... prettier is nice

[07-May-2010 15:52:59] <rmatte> yes, yes it is

[07-May-2010 15:53:21] <rmatte> plus the memory thresholding is an absolute must for our clients, took me a while to master making templates to the point where I could do it

[07-May-2010 15:53:25] <twm1010> do you use alot of aggregate graphs?

[07-May-2010 15:53:39] <twm1010> i use muli-graphs, but no aggregate reports

[07-May-2010 15:53:40] <rmatte> we use the network aggregate graph in our monthly reports

[07-May-2010 15:53:45] <rmatte> and the filesystem graph

[07-May-2010 15:53:47] <rmatte> but nothing else

[07-May-2010 15:54:24] <rmatte> aggregate graphs should be buildable from the UI

[07-May-2010 15:54:34] <rmatte> instead of having presets, some of which don't even work properly

[07-May-2010 15:54:45] <rmatte> but oh well, patience

[07-May-2010 15:55:03] <twm1010> i'll be on a plane this evening, i'll wrap my brain around how to handles these PRI's while i'm there and figure out the right class nesting

[07-May-2010 15:55:10] <rmatte> sweet, I have a working catalyst template

[07-May-2010 15:55:24] <rmatte> cool

[07-May-2010 15:55:55] <twm1010> here's my current interface map ignore names

[07-May-2010 15:55:56] <twm1010> ^(Null|Foreign|unrouted|rif0|lo|VLAN-|sit|EFXS|FXO|Voice|.*-Bearer Channel)

[07-May-2010 15:56:32] <twm1010> do you ignore any types?

[07-May-2010 15:57:55] <rmatte> just the defaults really

[07-May-2010 15:58:43] <twm1010> I'll get it sorted, hopefully i don't have any voice gateways with point to points data circuits :)

[07-May-2010 15:58:59] <rmatte> g

[07-May-2010 15:59:01] <rmatte> bah

[07-May-2010 15:59:07] <rmatte> ^(Voice Encapsulation|Voice Over IP Peer|Foreign Exchange Station)

[07-May-2010 15:59:09] <rmatte> that's mine

[07-May-2010 15:59:30] <rmatte> although we adjust it as need be per client

[07-May-2010 15:59:52] <rmatte> I have 13 different boxes so there's always some tweaking

[07-May-2010 16:04:15] <twm1010> sure... i'm about to add my second :)

[07-May-2010 16:04:19] <twm1010> that hold redundancy thing

[07-May-2010 16:05:20] <twm1010> i've not yet decided whether to setup a remote collector, or a second instance entirely

[07-May-2010 16:05:33] <twm1010> I'm tempted to setup the second box with higher polling values, etc.

[07-May-2010 16:05:38] <twm1010> faster, i mean to say

[07-May-2010 16:05:47] <twm1010> as it will watch fewer devices, much fewer

[07-May-2010 16:23:23] <mrayzenoss> Later on folks, have a good weekend

[07-May-2010 16:23:29] <willwh> o/

[07-May-2010 17:36:59] fOrsberg is now known as forsberg

[07-May-2010 19:26:09] forsberg is now known as fOrsberg

[07-May-2010 20:45:51] cgibbons_ is now known as cgibbons

[08-May-2010 00:00:30] [disconnected at Sat May  8 00:00:30 2010]

[08-May-2010 00:00:30] [connected at Sat May  8 00:00:30 2010]

[08-May-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[08-May-2010 03:43:27] fOrsberg is now known as forsberg

[08-May-2010 09:16:23] <chemist> anyone around

[08-May-2010 10:31:37] <forsberg> semi

[08-May-2010 10:31:37] <forsberg> ;>

[08-May-2010 17:54:03] rhettardo_ is now known as rhettardo

[08-May-2010 21:40:34] forsberg is now known as fOrsberg

[09-May-2010 00:00:30] [disconnected at Sun May  9 00:00:30 2010]

[09-May-2010 00:00:30] [connected at Sun May  9 00:00:30 2010]

[09-May-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[09-May-2010 03:40:24] fOrsberg is now known as forsberg

[09-May-2010 18:06:12] forsberg is now known as fOrsberg

[10-May-2010 00:00:30] [disconnected at Mon May 10 00:00:30 2010]

[10-May-2010 00:00:30] [connected at Mon May 10 00:00:30 2010]

[10-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[10-May-2010 01:03:19] fOrsberg is now known as forsberg

[10-May-2010 05:38:37] <chemist> anybody run cron jobs as the zenoss user?

[10-May-2010 05:53:24] <JesperS> chemist: yea, it does so by default. Check the cron.daily and .weekly

[10-May-2010 05:54:38] <chemist> JesperS: I am trying to restart the xmppBot from a cron job but I keep getting emails saying 'command not found'

[10-May-2010 05:54:51] <JesperS> looks like a path issue

[10-May-2010 05:54:56] <chemist> I have put $ZENHOME/bin in the path

[10-May-2010 05:55:33] <JesperS> and is it there?

[10-May-2010 05:55:50] <JesperS> try    ls -l $ZENHOME/bin/xmppBot

[10-May-2010 05:56:16] <JesperS> I don't have that one on my install, I guess it's some addon you've installed?

[10-May-2010 05:56:35] <chemist> lrwxrwxrwx 1 zenoss zenoss 113 2010-04-30 10:54 /usr/local/zenoss/zenoss/bin/xmppBot -> /usr/local/zenoss/zenoss/ZenPacks/ZenPacks.chudler.xmppBot-1.1-py2.4.egg/ZenPacks/chudler/xmppBot/daemons/xmppBot

[10-May-2010 05:56:48] <chemist> yeah, the xmppBot zenpack

[10-May-2010 05:56:51] <JesperS> k

[10-May-2010 05:57:08] <JesperS> try   ls -l /usr/local/zenoss/zenoss/ZenPacks/ZenPacks.chudler.xmppBot-1.1-py2.4.egg/ZenPacks/chudler/xmppBot/daemons/xmppBot

[10-May-2010 05:57:14] <JesperS> is it executable?

[10-May-2010 05:58:02] <JesperS> aslo try a  echo $ZENHOME   in your script, and try to run it. Is the path set (correctly) ?

[10-May-2010 05:58:28] <chemist> it is executable

[10-May-2010 06:01:23] <JesperS> Maybe $ZENHOME isn'

[10-May-2010 06:01:25] <JesperS> woops

[10-May-2010 06:01:42] <JesperS> Maybe $ZENHOME isn't polupated, as far as I recall it's from .bashrc right?

[10-May-2010 06:01:54] <JesperS> maybe .bashrc isn't run in scripts, but only in interactive mode.

[10-May-2010 06:02:08] <JesperS> I've had some issues with that before

[10-May-2010 06:04:35] <JesperS> I have a small'ish question now...

[10-May-2010 06:04:37] <JesperS> Is it possible to make alerts from events in the history part. I mean, I have a router that send SNMP traps when a interface goes down/up. When the interface goes down I get an event, but it's cleared again when the interface goes up.

[10-May-2010 06:04:42] <JesperS> I would like to set up an alert if the interface is "flapping" like it does this 5 times, but when the event is cleared the count stays at 1 on new events.

[10-May-2010 06:04:48] <JesperS> Do any of you guys know if it's possible to do this at all?

[10-May-2010 06:58:06] forsberg is now known as fOrsberg

[10-May-2010 07:55:11] fOrsberg is now known as forsberg

[10-May-2010 08:36:58] <Simon4> afternoon all

[10-May-2010 09:39:27] <jb> so, i created a new device class, moved some existing devices that had individual templates bound to them

[10-May-2010 09:39:33] <jb> and then bound a new template to the new device class

[10-May-2010 09:39:38] <jb> the devices are not inheriting the new template

[10-May-2010 09:39:41] <jb> even after a remodel..

[10-May-2010 09:39:43] <jb> whats the trick>

[10-May-2010 09:39:44] <jb> ?

[10-May-2010 09:41:30] <mrayzenoss> have you overridden the zCollectorPlugins locally?

[10-May-2010 09:43:15] <jb> nope

[10-May-2010 09:43:46] <mrayzenoss> any custom templates?

[10-May-2010 09:47:07] <jb> yeah

[10-May-2010 09:47:20] <jb> i created a new one, bound it to the class that the devices are in

[10-May-2010 09:48:25] <mrayzenoss> the custom template was there before the device was added, or after?

[10-May-2010 09:58:28] <jb> after

[10-May-2010 10:00:24] <rmatte> jb: If you've done local bindings on those devices the class level binding won't work

[10-May-2010 10:00:28] <rmatte> and it sounds like you might have?

[10-May-2010 10:00:52] <rmatte> You can either have device level binding or class level binding, you can't have both

[10-May-2010 10:01:26] <rmatte> mrayzenoss: that link you pasted to me the other day with an example of resetting bindings for a class would help him

[10-May-2010 10:01:35] <rmatte> the zendmd example

[10-May-2010 10:01:39] <rmatte> I can't seem to find it

[10-May-2010 10:03:03] <rmatte> ah, I think I might have found it...

[10-May-2010 10:04:09] <rmatte> Ah, here it is: docs/DOC-3219

[10-May-2010 10:04:22] <mrayzenoss> oh yeah, that's handy

[10-May-2010 10:06:33] <jb> ahh ok

[10-May-2010 10:06:40] <jb> i'll reset all of the local bindings

[10-May-2010 10:08:53] <rmatte> cool

[10-May-2010 10:18:43] <jb> hrm

[10-May-2010 10:18:48] <jb> ok, so I unbound everything locally

[10-May-2010 10:18:53] <jb> and rebound the templates on the class

[10-May-2010 10:19:06] <jb> do I need to do anything else?  no templates are being inheritied

[10-May-2010 10:19:16] <rmatte> Sweet

[10-May-2010 10:19:18] <rmatte> http://dev.zenoss.com/trac/ticket/6092 is fixed

[10-May-2010 10:19:30] <rmatte> the patch should be compatible with 2.5.2

[10-May-2010 10:19:43] <rmatte> jb: no, it should just work

[10-May-2010 10:19:46] <Simon4> jb: if it's just that collection isn't happening for those templaates/devices, it may be worth restarting zenperfsnmp (if they're snmp ones) since it caches the data, that or push changes to the device

[10-May-2010 10:20:06] <rmatte> jb: like, are you not seeing the graphs at all on those devices?

[10-May-2010 10:25:47] <mrayzenoss> Just read this http://www.ntop.org/blog/?p=112  If you want more network information integrated into Zenoss, this ntop Python extension looks pretty straightforward

[10-May-2010 10:27:27] <jb> rmatte: correct.. no graphs

[10-May-2010 10:28:45] <rmatte> jb: hmmm

[10-May-2010 10:29:08] <rmatte> jb: try resetting bindings on the class then re-applying?

[10-May-2010 10:29:35] <rmatte> when you bind to the class it binds to the sub-objects, so if you haven't rebound it since removing the local bindings from the devices, try that

[10-May-2010 10:30:04] <rmatte> I see that you rebound the templates on the class already

[10-May-2010 10:30:15] <rmatte> but did you use the reset bindings option in the menu when doing it?

[10-May-2010 10:37:34] <jb> no, I didn't

[10-May-2010 10:37:37] <jb> hrm let me try that

[10-May-2010 10:37:53] <jb> ah that pulled them all in

[10-May-2010 10:38:42] <jb> thanks!

[10-May-2010 10:39:19] <rmatte> np

[10-May-2010 10:51:45] <Simon4> removing devices in zendmd - I've been doing something like the following in test, but not sure it's "correct" since I don't think it sorts catalog stuff?

[10-May-2010 10:51:45] <Simon4> >>> dev = dmd.Devices.findDevice('devicename.here')

[10-May-2010 10:51:46] <Simon4> >>> dev.deleteDevice()

[10-May-2010 10:53:24] <rmatte> that's weird as hell

[10-May-2010 10:53:36] <rmatte> Zenoss is slower than hell for anyone here with a windows computer

[10-May-2010 10:53:42] <rmatte> but on my Linux box it's very fast

[10-May-2010 10:55:18] <rmatte> wonder what could be causing that

[10-May-2010 10:55:25] <rmatte> it's the same story on my windows VM

[10-May-2010 10:56:02] <Weetos> Does anyone know how to get a device name from an ip address in a event transform ? (I mean the device name as set in zenoss, not the PTR record)

[10-May-2010 10:57:51] <rmatte> Weetos: yes...

[10-May-2010 10:58:22] <rmatte> d = dmd.Devices.findDevice(evt.ip)

[10-May-2010 10:58:26] <rmatte> d.ipAddress

[10-May-2010 10:58:37] <rmatte> or sorry

[10-May-2010 10:59:13] <rmatte> just need to get the right function

[10-May-2010 10:59:23] <Weetos> rmatte> ok

[10-May-2010 10:59:33] <Weetos> thanks for taking the time to hel^p

[10-May-2010 10:59:38] <rmatte> nop

[10-May-2010 10:59:40] <rmatte> np*

[10-May-2010 11:00:22] <rmatte> yeh ok... so

[10-May-2010 11:00:31] <rmatte> d = dmd.Devices.findDevice(evt.ip)

[10-May-2010 11:00:40] <rmatte> name = d.getDeviceName()

[10-May-2010 11:00:50] <rmatte> that'll work

[10-May-2010 11:00:55] <rmatte> what version of Zenoss are you using?

[10-May-2010 11:02:58] <Weetos> I'm using 2.5.2

[10-May-2010 11:03:08] <rmatte> k, then that'll work fine

[10-May-2010 11:03:18] <Weetos> ok thank you rmatte

[10-May-2010 11:03:27] <Weetos> I owe you big time !

[10-May-2010 11:03:57] <rmatte> hehe, no problem

[10-May-2010 11:04:41] <Weetos> oh last thing do I have to import something special before using those lines of code ?

[10-May-2010 11:04:55] <rmatte> nope

[10-May-2010 11:04:58] <rmatte> it's imported automatically

[10-May-2010 11:05:08] <Weetos> okay fine =)

[10-May-2010 11:06:28] <rmatte> interesting, disabling the windows firewall seems to make the event console load normally

[10-May-2010 11:11:19] <mrayzenoss> rmatte: odd indeed

[10-May-2010 11:11:39] <mrayzenoss> so I assume the poor performance was FF & IE?

[10-May-2010 11:12:07] <rmatte> hmmm, apparently the windows firewall isn't the solution

[10-May-2010 11:12:33] <rmatte> mrayzenoss: There are 3 people here who no matter what they do can't get the event console to load in Firefox or IE

[10-May-2010 11:12:39] <rmatte> yet I can get it to load fine in Linux

[10-May-2010 11:12:48] <rmatte> and I even somehow managed to get it to load on my windows VM

[10-May-2010 11:14:36] <Weetos> Do the people running FF use FireBug ? and if they do, is there anything interesting reported ?

[10-May-2010 11:14:48] <rmatte> nah, they don't use firebug

[10-May-2010 11:15:42] <Weetos> I would start there, it helps a lot debugging Ajax stuff

[10-May-2010 11:16:51] <Weetos> oops sorry guys I have to go - see you later

[10-May-2010 11:16:59] <rmatte> later

[10-May-2010 11:23:30] <jb> do you guys know if zenwin polls perf data every 300 seconds by default?

[10-May-2010 11:23:39] <jb> i don't see a value for that in the collector options

[10-May-2010 11:23:55] <jb> Windows Modeler Cycle Interval (secs)60

[10-May-2010 11:23:58] <jb> or is that it?

[10-May-2010 11:24:14] <theacolyte> I believe so , yes, but iirc it is 300 seconds

[10-May-2010 11:24:27] <jb> that is set to 60

[10-May-2010 11:26:55] <rmatte> windows modeler cycle is the correct one for zenwin

[10-May-2010 11:27:07] <jb> so it actually polls every 60 seconds

[10-May-2010 11:27:12] <jb> vs a default of 300 seconds for SNMP

[10-May-2010 11:27:23] <rmatte> It would appear so

[10-May-2010 11:27:27] <jb> i was trying to find a document to confirm that

[10-May-2010 11:27:33] <rmatte> I don't even use zenwin, so never ran across that

[10-May-2010 11:27:34] <jb> i dont see it in the admin guide

[10-May-2010 11:27:36] <theacolyte> Let me look atm y logs

[10-May-2010 11:28:22] <jb> zenwinperf

[10-May-2010 11:28:45] <theacolyte> It's 5 minutes on mine

[10-May-2010 11:28:50] <theacolyte> and I did not change the value off default

[10-May-2010 11:28:51] <jb> hmmm

[10-May-2010 11:29:46] <jb> yea same here

[10-May-2010 11:29:47] <jb> 2010-05-10 11:08:07,331 INFO zen.zenwinperf: 58 devices processed (48402 datapoints)

[10-May-2010 11:29:47] <jb> 2010-05-10 11:13:07,335 INFO zen.zenwinperf: 58 devices processed (50663 datapoints)

[10-May-2010 11:29:47] <jb> 2010-05-10 11:18:07,339 INFO zen.zenwinperf: 58 devices processed (53558 datapoints)

[10-May-2010 11:29:47] <jb> 2010-05-10 11:23:07,344 INFO zen.zenwinperf: 58 devices processed (55925 datapoints)

[10-May-2010 11:29:47] <jb> 2010-05-10 11:28:07,348 INFO zen.zenwinperf: 58 devices processed (58180 datapoints)

[10-May-2010 11:30:03] <jb> i wonder if that value is configurable per collector or device

[10-May-2010 11:30:08] forsberg is now known as fOrsberg

[10-May-2010 11:30:23] <rmatte> have you checked in zProperties to see if there's a value in there?

[10-May-2010 11:30:51] <jb> ah yes it is!

[10-May-2010 11:30:53] <jb> awesome

[10-May-2010 11:31:34] <jb> zWinPerfCycleSeconds

[10-May-2010 11:31:41] <rmatte> there you go

[10-May-2010 11:31:44] <rmatte> ;)

[10-May-2010 11:33:12] <jb> so i can set that per device class

[10-May-2010 11:33:16] <jb> i wish zenperfsnmp was like that

[10-May-2010 11:34:19] <rmatte> there are zenperfsnmp timeout values in zProperties

[10-May-2010 11:34:23] <rmatte> just not cycle values

[10-May-2010 11:34:39] <jb> yeah thats collector wide, right?

[10-May-2010 11:34:45] <rmatte> there's a good reason why it's done that way, because if you adjust the cycle values you need to adjust the RRD create values

[10-May-2010 11:34:52] <rmatte> yes, collector wide

[10-May-2010 11:35:26] <jb> yeah.. you would think there would be the same issue with RRDs and zenwin

[10-May-2010 11:35:28] <jb> if you change the value

[10-May-2010 11:35:53] <rmatte> yeh

[10-May-2010 11:43:56] <voxter> hi guys!

[10-May-2010 11:44:53] <voxter> Using alpha2, Is there a known issue with event acknowledgement buttons' tool tips not showing up? and/or clicking the button i believe also does nothing

[10-May-2010 11:45:17] <voxter> When i say the button, i mean the checkmark with the X, which i believe to be an 'acknowledge and go away' button?

[10-May-2010 11:45:31] <voxter> oh. thats not what it is

[10-May-2010 11:45:34] <voxter> its unacknowledge.

[10-May-2010 11:46:11] <voxter> In that case,  what is the proper use case for acknowledging an event? checkmark? X? checkmark then X a certain number of days later?

[10-May-2010 11:46:21] <voxter> if i "X" out an event, can i still see it in an event history log someplace?

[10-May-2010 11:46:50] <rmatte> If you absolutely need to acknowledge an event before closing it then click the checkmark

[10-May-2010 11:46:59] <rmatte> otherwise, if you just want to move it to history, click the button with the X

[10-May-2010 11:47:14] <rmatte> If you want to do both then obviously click the checkmark then the x

[10-May-2010 11:49:45] <bigegor> Hi, Have someone HP EVA?

[10-May-2010 11:52:23] <rmatte> well, firebug doesn't show anything helpful to assist in troubleshooting these event console response issues

[10-May-2010 11:53:26] <rmatte> voxter: and yes, you would use the Event History link in the bottom left to view events

[10-May-2010 11:57:51] <voxter> rmatte: are you seeing that the tooltips dont work though?

[10-May-2010 12:00:31] <Simon4> hmm, this removing devices lark is harder, time to look at other methods

[10-May-2010 12:05:15] <voxter> rmatte: interesting. one event that i cleared, an smtp error of "expected and didnt get 220 back from smtp server" i look for it after i 'X" it in the event log and its not coming up in event history

[10-May-2010 12:06:05] <voxter> oh, i had to take 'last seen' field and empty it.

[10-May-2010 12:06:17] <voxter> first seen/last seen is not necessarily show me date range x -> y

[10-May-2010 12:07:03] <rmatte> Just set the date for the last seen and that'll give you a range from that date to present

[10-May-2010 12:07:29] <rmatte> I don't recommend emptying the field since that'll display the entire event database

[10-May-2010 12:07:36] <rmatte> which is very intensive when you have lots of events

[10-May-2010 12:07:58] <voxter> perfect! thanks!

[10-May-2010 12:08:01] fOrsberg is now known as forsberg

[10-May-2010 12:10:04] <rmatte> np

[10-May-2010 12:13:10] <rmatte> hmmm

[10-May-2010 12:13:11] <rmatte> Ext.ux.grid is undefined

[10-May-2010 12:13:11] <rmatte> http://delta01.novanoc.com:8180/zenui/js/zenoss.js

[10-May-2010 12:13:11] <rmatte> Line 103

[10-May-2010 12:16:19] <bigegor> Ok, I'll ask the question another way, if somebody want test HP EVA 4400/6400/8400 ZenPack?

[10-May-2010 12:18:07] <rmatte> I don't have any HP EVA unfortunately

[10-May-2010 12:58:46] <rmatte> I finally figured out how to just restart mysql via the zenoss-stack script

[10-May-2010 12:59:13] <rmatte> /etc/init.d/zenoss-stack restart mysql

[10-May-2010 13:13:28] forsberg is now known as fOrsberg

[10-May-2010 13:28:03] <jb> im not sure this zWinPerfCycleSeconds is doing anything

[10-May-2010 13:28:31] <jb> you would think the graphs would show more data

[10-May-2010 13:28:40] <jb> for a box with that value set to 60

[10-May-2010 13:28:42] <jb> vs one set to 300

[10-May-2010 13:28:48] <jb> right?

[10-May-2010 13:32:24] <rmatte> you probably need to adjust the RRD creation values

[10-May-2010 13:38:13] <jarcuri> hi, Iwould like to know where can  I find zenoss plugin

[10-May-2010 13:38:46] <jarcuri> for downloading

[10-May-2010 13:39:24] <jarcuri> or how to monitoring HP-UX filesystem without needing this plugin

[10-May-2010 13:41:20] <rmatte> without needing what plugin?

[10-May-2010 13:42:07] <rmatte> Download this: docs/DOC-3462

[10-May-2010 13:42:09] <rmatte> unzip it

[10-May-2010 13:42:38] <rmatte> and then install it as the zenoss user with: zenpack --install ZenPacks.community.HPUXMonitor-2.1.2-py2.4.egg

[10-May-2010 13:43:17] <jarcuri> i wonder how to monitoring HP-UX filesystem, but through SNMP is not working, so I have heard that u can use Zenoss Plugin to do that, which uses SSH command, is that true ?

[10-May-2010 13:44:59] <jarcuri> i have ZenPacks.community.HPUXMonitor-2.1.2-py2.4.egg already installed, and template for filesystem monitoring is not working

[10-May-2010 13:47:22] <rmatte> You're sure that SNMP is enabled with the proper permissions on the device?

[10-May-2010 13:48:26] <jarcuri> i guess, cuz If i run snmpwalk against OID 1.3.6.1.4.1.11.2.3.1.2.2.1.6, i have the answer

[10-May-2010 13:48:28] <rmatte> You sure the FileSystem template for HPUX is under the device class that you put the devices under?

[10-May-2010 13:48:41] <rmatte> k, then you don't have the devices in the right place

[10-May-2010 13:48:45] <rmatte> servlet/JiveServlet/showImage/102-3462-5-2949/HPUX-filesystem.png

[10-May-2010 13:48:57] <rmatte> that's a screenshot of the filesystem template that comes with the ZenPack

[10-May-2010 13:49:11] <rmatte> the devices need to be under a class that uses that template instead of the standard FileSystem template

[10-May-2010 13:49:12] <jarcuri> thats the same I have

[10-May-2010 13:49:31] <rmatte> then there's no reason for it not to be working

[10-May-2010 13:49:34] <jarcuri> oh

[10-May-2010 13:50:00] <rmatte> what class do you have the devices under right now?

[10-May-2010 13:50:28] <jarcuri> let me check

[10-May-2010 13:50:31] <rmatte> k

[10-May-2010 13:51:43] <jarcuri> \/Devices /Server /HPUX /Templates /FileSystem

[10-May-2010 13:52:02] <rmatte> ok, and you have all of your devices under /Devices/Server/HPUX?

[10-May-2010 13:52:06] <jarcuri> target class: Products.ZenModel.Device

[10-May-2010 13:52:13] <rmatte> all of your HPUX devices I should say

[10-May-2010 13:52:16] <jarcuri> yes

[10-May-2010 13:52:23] <rmatte> and you remodeled them under that class?

[10-May-2010 13:52:38] <jarcuri> remodeled ? how do I do that

[10-May-2010 13:52:56] <rmatte> Go to the device page for a device and select Manage -> Remodel Device

[10-May-2010 13:53:14] <rmatte> from the dropdown

[10-May-2010 13:53:26] <rmatte> do that for each device

[10-May-2010 13:53:35] <rmatte> that'll allow it to discover the filesystems on them

[10-May-2010 13:57:47] <jarcuri> could't find this options "remodel device" under Manage

[10-May-2010 13:58:08] <jarcuri> just model device

[10-May-2010 13:58:37] <rmatte> same thing

[10-May-2010 13:58:38] <rmatte> click that

[10-May-2010 13:58:53] <rmatte> (I'm doing this off the top of my head lol)

[10-May-2010 13:59:12] <jarcuri> lol

[10-May-2010 14:00:45] <jarcuri> one if the INFO message was

[10-May-2010 14:00:46] <jarcuri> plugins: zenoss.snmp.NewDeviceMap, zenoss.snmp.DeviceMap, HPDeviceMap, DellDeviceMap, zenoss.snmp.InterfaceMap, zenoss.snmp.RouteMap, zenoss.snmp.IpServiceMap, zenoss.snmp.HRSWInstalledMap, zenoss.snmp.HRSWRunMap, zenoss.snmp.CpuMap, HPCPUMap, DellCPUMap, DellPCIMap

[10-May-2010 14:00:53] <jarcuri> one of

[10-May-2010 14:01:18] <rmatte> it always does that

[10-May-2010 14:01:23] <rmatte> that's normal

[10-May-2010 14:01:33] <jarcuri> but no device was added

[10-May-2010 14:01:43] <rmatte> hunh?

[10-May-2010 14:01:49] <rmatte> what do you mean no device was added?

[10-May-2010 14:01:54] <rmatte> the device is supposed to already be there

[10-May-2010 14:02:02] <rmatte> you're supposed to be modelling an existing device

[10-May-2010 14:06:12] <jarcuri> Problem loading plugin ZenPacks.community.HPUXMonitor.modeler.plugins.community.snmp.HPUXFileSystemMap

[10-May-2010 14:06:24] <jarcuri> Traceback (most recent call last): File "/opt/zenoss/Products/DataCollector/Plugins.py", line 102, in create self.modPath) File "/opt/zenoss/Products/DataCollector/Plugins.py", line 171, in importPlugin return importClass(modPath) File "/opt/zenoss/Products/ZenUtils/Utils.py", line 360, in importClass raise ex ImportError: No module named CollectorPlugin

[10-May-2010 14:06:31] <jarcuri> Due to import errors, removing the ZenPacks.community.HPUXMonitor.modeler.plugins.community.snmp.HPUXFileSystemMap plugin from this collection cycle

[10-May-2010 14:07:27] jarcuri is now known as jeffi

[10-May-2010 14:09:25] <rmatte> hmmmm

[10-May-2010 14:09:39] <rmatte> looks like that was never updated to be compatible with Zenoss 2.5.2

[10-May-2010 14:10:07] <rmatte> contact the author of the ZenPack and ask them to update it

[10-May-2010 14:11:02] <jeffi> ok, thanks for the help

[10-May-2010 14:11:07] <rmatte> np

[10-May-2010 14:13:40] <rmatte> well, so far zero people here are happy with the new event console

[10-May-2010 14:13:42] <rmatte> :(

[10-May-2010 14:14:01] <rmatte> due to performance issues

[10-May-2010 14:16:36] <jeffi> thats right, since 've upgrade zenoss from 2.4 to 2.5 branch, this feature just stop of working

[10-May-2010 14:17:11] <rmatte> I've noticed that scrolling through events is much better with it, but loading or moving events to history is horribly slow

[10-May-2010 14:19:59] <rmatte> hopefully it'll be improved in stone crab

[10-May-2010 14:55:54] <zykes-> question, does zenoss support autodiscovery outside of snmp ?

[10-May-2010 15:04:33] <mrayzenoss> zykes-: yes

[10-May-2010 15:04:58] <mrayzenoss> zykes-: discovery is based on the /Discover device class, which attempts to use SNMP, WMI and SSH

[10-May-2010 15:05:14] <mrayzenoss> you could add additional modeling plugins to discover devices

[10-May-2010 15:08:17] <Simon4> mrayzenoss: do you know of any guides to removing devices programatically?

[10-May-2010 15:08:41] <Simon4> (and, Hi!)

[10-May-2010 15:08:51] <rmatte> Simon4: why wouldn't you just call on the function to remove a device?

[10-May-2010 15:08:56] <mrayzenoss> Simon4: no, but that shouldn't be too hard

[10-May-2010 15:09:10] <Simon4> rmatte: I've been using device.deleteDevice()

[10-May-2010 15:09:16] <Simon4> but it seems to leave the catalogue in a mess :/

[10-May-2010 15:09:24] <rmatte> oh?

[10-May-2010 15:09:26] <Simon4> so hoping there's some better way that i just haven't seen

[10-May-2010 15:09:52] <rmatte> why don't you figure out which Zenoss python file is responsible for removing devices and check what it does?

[10-May-2010 15:11:09] <Simon4> yeah, that was where I started today.. it just seems to call _delObject(id)

[10-May-2010 15:11:24] <rmatte> well, I'm sure it does a bit more than that

[10-May-2010 15:11:29] <mrayzenoss> Products/ZenModel/Device.py

[10-May-2010 15:11:34] <mrayzenoss> deleteDevice()

[10-May-2010 15:11:36] <Simon4> mrayzenoss: yeah, that's where I'm digging

[10-May-2010 15:12:05] <rmatte> that's the literal function you want

[10-May-2010 15:12:11] <rmatte> deleteDevice()

[10-May-2010 15:12:21] <Simon4> but after removing the device, if you navigate to say... Collectors/localhost

[10-May-2010 15:12:35] <Simon4> you get a stacktrace saying there's a keyerror for the device you just hoovered

[10-May-2010 15:12:50] <Simon4> so there are remnants still somewhere

[10-May-2010 15:13:04] <Simon4> same if you go to an ipservice that the host had defined, etc

[10-May-2010 15:13:12] <Simon4> it's like not all the relations are cleaned up

[10-May-2010 15:13:44] <rmatte> Then there may be some other function that it's calling before or after

[10-May-2010 15:13:54] <rmatte> you need to find the code to remove a device from the GUI

[10-May-2010 15:13:59] <rmatte> and check example what it's calling on

[10-May-2010 15:14:14] <rmatte> example == exactly

[10-May-2010 15:15:15] <rmatte> oh hold up

[10-May-2010 15:15:24] <rmatte> after you delete the device, are you doing a reindex()?

[10-May-2010 15:15:29] <rmatte> if not, that's your problem

[10-May-2010 15:16:15] <rmatte> to delete all devices you'd do...

[10-May-2010 15:16:17] <rmatte> for dev in dmd.Devices.getSubDevices():

[10-May-2010 15:16:17] <rmatte>   print dev.id

[10-May-2010 15:16:17] <rmatte>   dev.deleteDevice()

[10-May-2010 15:16:17] <rmatte> commit()

[10-May-2010 15:16:17] <rmatte> reindex()

[10-May-2010 15:16:18] <rmatte> commit()

[10-May-2010 15:16:29] <rmatte> so you need to remove the device, commit, reindex, and commit again

[10-May-2010 15:16:51] <rmatte> try it and let us know if it works

[10-May-2010 15:17:42] <Simon4> rmatte: ahhh

[10-May-2010 15:17:47] * Simon4 tries with the reindex

[10-May-2010 15:29:33] <rmatte> mrayzenoss: I assume that the dev guys are aware that the public demo of stone crab doesn't work at all in IE?

[10-May-2010 15:29:51] <rmatte> IE 8 to be specific (even if I use compatibility mode)

[10-May-2010 15:29:53] <mrayzenoss> I'm sure

[10-May-2010 15:29:56] <rmatte> k

[10-May-2010 15:30:03] <mrayzenoss> QA is testing on IE7 and 8

[10-May-2010 15:30:14] <mrayzenoss> 3.0 is supposed to be IE8 compatible

[10-May-2010 15:30:20] <rmatte> k

[10-May-2010 15:30:28] <rmatte> none of the pages load at all as it stands

[10-May-2010 15:30:32] <rmatte> so that'll be fun for them

[10-May-2010 15:30:48] <mrayzenoss> I'm sure it's something transient

[10-May-2010 15:30:53] <zykes-> mrayzenoss: any user needed to use wmi on hosts ?

[10-May-2010 15:31:09] <rmatte> yes, you need an account that has access to poll WMI

[10-May-2010 15:31:17] <zykes-> where do i set that ?

[10-May-2010 15:31:27] <rmatte> easiest way, make the user an admin

[10-May-2010 15:31:50] <rmatte> then that username and password needs to be set in zWinUser and zWinPassword under zProperties in Zenoss

[10-May-2010 15:32:01] <zykes-> ah

[10-May-2010 15:32:14] <rmatte> there's a section in the Zenoss Admin guide that explains all of this

[10-May-2010 15:32:28] <mrayzenoss> docs/DOC-3745 is a nice wiki article for XP

[10-May-2010 15:33:49] <zykes-> Also i'm trying to add a OS Product for windows 2008, it says it adds it but it doesn't ?

[10-May-2010 15:46:32] <rmatte> yeh, that's a bug

[10-May-2010 15:46:40] <rmatte> apparently it's not going to get fixed in 2.5.2 though

[10-May-2010 15:46:51] <zykes-> that sucks...

[10-May-2010 15:46:55] <rmatte> you can add product models perfectly fine, but anything to do with software can't be added

[10-May-2010 15:47:20] <rmatte> yup

[10-May-2010 15:49:15] <rmatte> oh wait

[10-May-2010 15:49:18] <rmatte> I was able to add one

[10-May-2010 15:49:35] <rmatte> ok, you're probably making a really simple mistake

[10-May-2010 15:49:52] <rmatte> see at the bottom where there's a "show all" button?

[10-May-2010 15:49:55] <zykes-> i can add one, but it doesn't show in the overview.

[10-May-2010 15:49:59] <rmatte> (the bug that I was thinking of is something else)

[10-May-2010 15:50:05] <zykes-> i mean, when i create a new device i can't select it

[10-May-2010 15:50:09] <rmatte> oh

[10-May-2010 15:50:12] <rmatte> yeh, that's the bug

[10-May-2010 15:50:15] <zykes-> under microsoft

[10-May-2010 15:50:23] <rmatte> doesn't prevent you from creating them, but prevents you from seeing them in that list

[10-May-2010 15:50:31] <rmatte> but what you can do is...

[10-May-2010 15:50:41] <rmatte> make sure that the product has the proper OID in it when you create it

[10-May-2010 15:50:45] <rmatte> then remodel the device

[10-May-2010 15:50:51] <zykes-> oid ?

[10-May-2010 15:50:56] <rmatte> it should hopefully assign it even though you can't manually assign it

[10-May-2010 15:51:03] <rmatte> yes, the snmpd oid that represents that product

[10-May-2010 15:51:11] <mrayzenoss> rmatte: if you clear your cache, does IE8 work?

[10-May-2010 15:51:21] <rmatte> mrayzenoss: I'll try in a sec

[10-May-2010 15:51:42] <rmatte> zykes-: when you first modeled the device, under the software details, did you not see numbers with dots between them?

[10-May-2010 15:51:46] <zykes-> rmatte: does that work on windows ?

[10-May-2010 15:51:56] <rmatte> yes, I guarantee it does

[10-May-2010 15:52:02] <zykes-> hmm, k

[10-May-2010 15:53:11] <rmatte> I love how every modern browser has to bury the place where you can clear your cache

[10-May-2010 15:53:20] <rmatte> it used to just be options -> clear cache

[10-May-2010 15:53:27] <rmatte> now I need to hunt around for it each time I want to do it

[10-May-2010 15:53:28] <rmatte> lol

[10-May-2010 15:54:15] <rmatte> aha, found it

[10-May-2010 15:56:15] <rmatte> zykes-: Actually, you're monitoring via WMI?

[10-May-2010 15:56:29] <rmatte> Won't work then

[10-May-2010 15:56:33] <zykes-> i'm wanting to do so..

[10-May-2010 15:56:35] <rmatte> so yeh, you're pretty much stuck

[10-May-2010 15:56:35] <zykes-> wit what ?

[10-May-2010 15:56:40] <zykes-> auto discovery ?

[10-May-2010 15:56:55] <rmatte> the OID I was talking about as a workaround to have the OS name set is SNMP only

[10-May-2010 15:56:59] <zykes-> ah

[10-May-2010 15:57:01] <rmatte> it's not going to work if you're using WMI

[10-May-2010 15:57:09] <zykes-> i know :p

[10-May-2010 15:57:18] <mrayzenoss> egor's wmi modeler is in beta

[10-May-2010 15:57:21] <zykes-> mrayzenoss: can i delete os products ?

[10-May-2010 15:57:28] <zykes-> wmi modeler?

[10-May-2010 15:57:37] <rmatte> mrayzenoss: his WMI modeler has been out of beta for a while hasn't it?

[10-May-2010 15:57:40] <rmatte> since 1.97?

[10-May-2010 15:57:57] <rmatte> you can delete os products

[10-May-2010 15:58:01] <rmatte> and you can add them

[10-May-2010 15:58:05] <rmatte> you just can't manually set them

[10-May-2010 15:58:05] <bigegor> it RC for now ;)

[10-May-2010 15:58:20] <mrayzenoss> docs/DOC-3500

[10-May-2010 15:58:22] <rmatte> they claim that they can't reproduce the issue in stone crab, they'd better be 100% sure of that

[10-May-2010 15:58:23] <mrayzenoss> RC3

[10-May-2010 15:58:29] <rmatte> because of not I shall be screaming about it :)

[10-May-2010 15:59:07] <zykes-> i go under products

[10-May-2010 15:59:26] <zykes-> Microsoft where i find server 2008

[10-May-2010 15:59:43] <zykes-> but can't find a way to del it

[10-May-2010 16:00:12] <rmatte> ohhhh

[10-May-2010 16:00:15] <rmatte> there is a way around that

[10-May-2010 16:00:23] <rmatte> This is the original ticket that they closed: http://dev.zenoss.org/trac/ticket/5473

[10-May-2010 16:00:31] <rmatte> add the OS Product, then go in to zendmd and do a reindex()

[10-May-2010 16:00:34] <rmatte> then it'll show up

[10-May-2010 16:00:41] <rmatte> only works for OS Products, not for Software Products

[10-May-2010 16:00:56] <rmatte> I discovered that when testing

[10-May-2010 16:01:11] <rmatte> when I say reindex I mean do the following as the zenoss user from the commandline:

[10-May-2010 16:01:12] <rmatte> zendmd

[10-May-2010 16:01:16] <rmatte> reindex()

[10-May-2010 16:01:19] <rmatte> commit()

[10-May-2010 16:01:20] <rmatte> ctrl-d

[10-May-2010 16:01:26] <rmatte> do that after you add the OS Product

[10-May-2010 16:02:51] <zykes-> But again, can i delete products ?

[10-May-2010 16:03:04] <rmatte> hunh?

[10-May-2010 16:03:11] <rmatte> you can if you want, but why would you want to?

[10-May-2010 16:03:11] <zykes-> can I delete a product ?

[10-May-2010 16:03:28] <zykes-> i added the wrong thing

[10-May-2010 16:03:32] <rmatte> you can delete anything you can create

[10-May-2010 16:03:46] <rmatte> ok, so find it in the list, put a checkmark next to it, then from the dropdown select delete

[10-May-2010 16:03:51] <rmatte> or remove, or whatever they labeled it

[10-May-2010 16:04:17] <zykes-> i see very little dropdowns :p

[10-May-2010 16:04:26] <rmatte> ?

[10-May-2010 16:04:37] <mrayzenoss> insufficient privileges?

[10-May-2010 16:04:43] <zykes-> i can't see a dropdown that gives me any actions for that

[10-May-2010 16:04:54] <rmatte> I'm staring right at it

[10-May-2010 16:05:11] <rmatte> it's the downward arrow to the left of the "Manufacturers" title, or "Microsoft" in your case

[10-May-2010 16:05:52] <rmatte> sorry, to the left of Products

[10-May-2010 16:05:53] <rmatte> rather

[10-May-2010 16:05:57] <rmatte> (I just loaded it to check)

[10-May-2010 16:06:07] <zykes-> ah, lol

[10-May-2010 16:06:08] <zykes-> :p

[10-May-2010 16:06:14] <zykes-> i thought it was a sort thingy

[10-May-2010 16:06:17] <rmatte> There's a Delete Products option

[10-May-2010 16:06:24] <rmatte> no, it's not

[10-May-2010 16:06:36] <rmatte> there are context menus like that all over the place in Zenoss

[10-May-2010 16:06:39] <rmatte> and they are always menus

[10-May-2010 16:07:08] <rmatte> I'd strongly suggest that you read the Zenoss Admin guide top to bottom

[10-May-2010 16:07:12] <rmatte> it explains all of this stuff

[10-May-2010 16:07:14] <rmatte> and more

[10-May-2010 16:08:03] <zykes-> guess i'll do tmrw at work ;)

[10-May-2010 16:08:20] <rmatte> k

[10-May-2010 16:08:22] <mrayzenoss> to be fair, the UI isn't completely intuitive, hence the overhaul for 3.0

[10-May-2010 16:08:55] <zykes-> mrayzenoss: i guess the upgrade is going to be without any fuzz ?

[10-May-2010 16:08:57] <rmatte> yeh, but honestly, the 3.0 UI looks almost as confusing lol

[10-May-2010 16:09:05] <rmatte> but it is a bit better

[10-May-2010 16:09:09] <rmatte> and it definitely looks nicer

[10-May-2010 16:09:34] <rmatte> clearing the cache didn't help by the way

[10-May-2010 16:09:47] <rmatte> I get an error loading page when browsing to Events, or Infrastructure

[10-May-2010 16:09:54] <rmatte> Reports as well

[10-May-2010 16:10:02] <rmatte> and Advanced

[10-May-2010 16:10:22] <mrayzenoss> rmatte: word from QA was there was something IE-bad only in yesterday's build

[10-May-2010 16:10:23] <rmatte> the Dashboard loads, but with the old headers on each portlet

[10-May-2010 16:10:26] <rmatte> rather than the new ones

[10-May-2010 16:10:31] <cgibbons> there was a comma problem in one of the javascript files on Friday or Thursday & it was fixed, but it might not have been pushed to public-demo yet

[10-May-2010 16:10:31] <rmatte> ah

[10-May-2010 16:10:41] <mrayzenoss> what he said :)

[10-May-2010 16:10:43] <rmatte> k, that might explain it then

[10-May-2010 16:11:03] <rmatte> man I hate IE, that browser needs to be erradicated

[10-May-2010 16:11:31] <rmatte> what I hate even more is that people insist on using it

[10-May-2010 16:45:35] mrayzenoss is now known as mattray

[10-May-2010 17:01:16] <rmatte> well, I'm out for the day, cheers guys

[10-May-2010 17:02:19] <Simon4> seu, thanks for the help, I think I'm onto the issue

[10-May-2010 17:02:56] <rmatte> cool

[10-May-2010 17:03:03] <rmatte> :)

[10-May-2010 17:36:26] <Simon4> oh ffs

[10-May-2010 17:36:59] <Simon4> dev.changeDeviceClass() works a charm, including changing the python class type etc

[10-May-2010 17:37:22] * Simon4 uninvents the wheel

[10-May-2010 17:38:08] * Simon4 was sure that never worked in the past, at least through the gui

[10-May-2010 22:34:08] <jimbosyn> Hi folks

[10-May-2010 22:34:25] <jimbosyn> Got a quick question that I could not find an answer to on the forums

[10-May-2010 22:34:49] <jimbosyn> Anyone in here?

[10-May-2010 22:38:06] <jimbosyn> So I'm trying to scale the web front end.

[10-May-2010 22:38:42] <jimbosyn> Is it safe to have multiple servers running zope pointing back to a centralized zeo server?

[11-May-2010 00:00:30] [disconnected at Tue May 11 00:00:30 2010]

[11-May-2010 00:00:30] [connected at Tue May 11 00:00:30 2010]

[11-May-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[11-May-2010 01:33:06] fOrsberg is now known as forsberg

[11-May-2010 01:59:35] <mike1> hello all

[11-May-2010 05:29:30] <JesperS> Hi guys. I found the following docs/DOC-5493 - Extended Capabilities for Cisco Devices

[11-May-2010 05:29:32] <JesperS> It mentions /Network/Cisco/ASA device class, but there is no such class in my install. If I create a ASA sub device, would it get the correct configuration parameters, or am I missing something here?

[11-May-2010 05:29:33] <zykes-> http://pastebin.com/iN1f1Pe6 < anyone got a clue on that ?

[11-May-2010 05:30:42] <zykes-> i get an error when trying to view my zenpacks

[11-May-2010 05:59:45] <zykes-> does the zenoss-stack include zenoss-packs for 2.5.2 on debian ?

[11-May-2010 07:21:45] <LW> Good morning, I'm running into an issue where for some reason zenoss now seems to not process my maintenance windows.  Looking at zenactions.log, seemed to keep going back and forth on a particular maintenance window, and then after that window ended, it said starting another window, and then proceeded to alarm for a number of servers all night.  We ran into the same thing Sunday morning with a single group, which appeared to be fixed b

[11-May-2010 07:33:16] <LW> what is interestingly enough, it is reporting that "standard window" is affecting servers in other system groupings that would have exited their windows by now

[11-May-2010 07:33:28] <LW> standard window being our standard maintenance window name

[11-May-2010 07:49:34] irc is now known as Guest91406

[11-May-2010 07:52:55] <tornado> I'm just to check if anybody alive on the zenoos channel...

[11-May-2010 07:53:42] <Simon4> morning mrayzenoss

[11-May-2010 07:54:27] <mrayzenoss> morning

[11-May-2010 07:58:07] <LW> sweet people who are alive

[11-May-2010 07:58:32] <LW> ran into an issue twice so far where zenoss stops processing maintenance windows, any hints?

[11-May-2010 08:01:52] <mrayzenoss> stops processing?  are they set up as repeating windows?

[11-May-2010 08:02:13] <LW> yes, they are, and some have worked for over a year without problems

[11-May-2010 08:02:20] <LW> i'm also seeing another issue, lemmie grab the link

[11-May-2010 08:03:16] <LW> thread/13386;jsessionid=DEE6C9925443ECBC6BD3BE655D249EA9.node0?decorator=print&displayFullThread=true I'm also seeing this :(

[11-May-2010 08:04:13] <LW> mrayzenoss: everything was working fine, it stated that it was starting a window i think around 11pm last night, and then about a half hour later my engineers wanted to kill me :)

[11-May-2010 08:04:23] <LW> when it started alarming for servers

[11-May-2010 08:11:15] <mrayzenoss> and these windows had worked previously?

[11-May-2010 08:22:51] <LW> mrayzenoss: without any problems, for quite some time, minus the occassional adjustment due to the windows needing to be resized due to the end devices

[11-May-2010 08:23:07] * Simon4 finally gets his script working again

[11-May-2010 08:23:22] <Simon4> three weeks off slows things down too much purely from forgetting things :/

[11-May-2010 08:24:10] * LW ponders deleting all events from the history table where the ip address is localhost

[11-May-2010 08:31:15] <LW> bah, 8.6 million event class change entries in my history table

[11-May-2010 08:52:45] <jb> hrm, how do I enable file system monitoring via WMI with winperf?

[11-May-2010 08:52:51] <jb> i have the collector plugin enabled

[11-May-2010 08:53:54] <jb> ah, nm.

[11-May-2010 09:28:46] <cgibbons> hmmm

[11-May-2010 09:30:30] <rmatte> indeed

[11-May-2010 09:31:06] <mrayzenoss> that's what cgibbons says when he sees something on our code he doesn't like :p

[11-May-2010 09:31:15] <rmatte> haha

[11-May-2010 09:31:17] <mrayzenoss> I saw something yesterday in Software.py

[11-May-2010 09:31:28] <mrayzenoss> the version numbers aren't hooked up at all

[11-May-2010 09:31:36] <rmatte> nice

[11-May-2010 09:32:04] <rmatte> mrayzenoss: you'll be a zenoss dev soon enough :)

[11-May-2010 09:32:06] <mrayzenoss> I was trying to do a very thorough job with collecting OS X software inventory, all for naught

[11-May-2010 09:32:39] <mrayzenoss> rmatte: I actually interviewed for that job at first and chose Community Manager instead

[11-May-2010 09:33:28] <rmatte> cool

[11-May-2010 09:33:55] <mrayzenoss> guess I should open a ticket for that

[11-May-2010 09:37:38] <rmatte> If you expect it to get fixed, yup :)

[11-May-2010 09:39:20] <mrayzenoss> http://dev.zenoss.org/trac/ticket/6711

[11-May-2010 09:39:45] <rmatte> no "failwhale"?

[11-May-2010 09:40:03] <mrayzenoss> I'm sure that will resurface in another ticket soon

[11-May-2010 09:40:49] <rmatte> :)

[11-May-2010 09:54:43] forsberg is now known as fOrsberg

[11-May-2010 10:32:09] <coofamani> rascd7009

[11-May-2010 10:32:13] <coofamani> woops

[11-May-2010 10:32:52] <mrayzenoss> FYI, if that's a password, this channel is logged ;)

[11-May-2010 10:33:15] <coofamani> heh, already changed :)

[11-May-2010 10:33:20] <coofamani> thanks

[11-May-2010 10:33:20] <rmatte> lol

[11-May-2010 10:33:33] <rmatte> probably changed to rascd7010 :P

[11-May-2010 10:33:36] <coofamani> it happens... never did it in IRC before. Gonna be a great week.

[11-May-2010 10:33:37] <rmatte> (kidding)

[11-May-2010 10:33:43] <coofamani> keepassx ftw

[11-May-2010 10:33:44] <aclark> hah

[11-May-2010 10:34:19] fOrsberg is now known as forsberg

[11-May-2010 10:45:58] <LW> question, does the mysql status table maintain all status information for every single host?

[11-May-2010 10:46:24] <LW> (can't look, deleting form another table in the db and... its not returned on a select yet from the status table"

[11-May-2010 10:48:44] <forsberg> its only for all the events

[11-May-2010 10:48:47] <forsberg> _all_ the events

[11-May-2010 10:57:33] <gwb2351> what's the latest/best method for batch loads/additions of devices to zenoss?  wget?  curl?  zenbatchload?

[11-May-2010 11:11:32] <mrayzenoss> gwb2351: zenbatchload

[11-May-2010 11:24:41] <LW> i guess running into issue that I'm seeing spamming my logs and history table with maintenance window changes started after daylight savings time change earlier this year, which would explain the increase in disk utilization that I've seen as well, i wonder if I should just go re-define all my maintenance windows with a more recent date in that case

[11-May-2010 11:31:50] <mrayzenoss> LW: looks like this? http://dev.zenoss.com/trac/ticket/5997

[11-May-2010 11:34:37] <mrayzenoss> updated the ticket to get a 2.5.x branch patch made available

[11-May-2010 11:39:14] <rmatte> mrayzenoss: nice

[11-May-2010 11:40:46] <mrayzenoss> yeah, that's been a fairly constant bug.  Everytime we deal with DST something else breaks

[11-May-2010 11:44:41] <Serideru> is anyone using the domain/ssl expiration zenpack?

[11-May-2010 11:45:31] <mrayzenoss> I've used it before when I tested it

[11-May-2010 11:48:03] <Serideru> I'm having an issue with it populating the number of days until domain expiration.  jwhois is working and i'm not getting an error that says the command failed.  but I get a MISSING RRD FILE: for the Domain_Days graph

[11-May-2010 11:48:15] <Serideru> Days to ssl cert expire works fine

[11-May-2010 11:48:26] <mrayzenoss> hmm… that's odd

[11-May-2010 11:48:32] <mrayzenoss> 1 command works, the other doesn't?

[11-May-2010 11:48:46] <Serideru> yep

[11-May-2010 11:50:16] <Serideru> originally it was a problem with my firewall.  I had my network guys open port 43 for whois and then the jwhois command actually returned data.

[11-May-2010 11:50:56] <mrayzenoss> so from the cli, the check_domain.sh works as the zenoss user?

[11-May-2010 11:51:05] <mrayzenoss> anything of interest in the zencommand.log?

[11-May-2010 11:53:56] <Serideru> just manually ran check_domain.sh as zenoss and it worked.  let me check the zencommand log

[11-May-2010 11:56:11] <Serideru> ahh ok there's the problem.  it's running the check_domain.sh on the full name of the site and not just the domain.  example: members.test.com does a lookup for members.test.com and not just test.com

[11-May-2010 12:09:30] <LW> mrayzenoss: ours our generally >1hr, but it sounds like kind of what its seeing, it initially faulted on one that was 30 minutes long and since broke just after that one window last night

[11-May-2010 12:10:23] <gwb2351> should I be able to use these lines to zenbatchload (current: ERROR: unable to parse the entry for...)

[11-May-2010 12:10:31] <gwb2351> /Devices/Server/Passive

[11-May-2010 12:10:32] <gwb2351>     0022P-canon_3570-Generic_Text_Only manageIp=172.16.43.166, comments="RPM Elite Print Queues. Only contact David Anthony and Gary Henderson M-F 8AM - 5PM on alerts", systemPaths=['/Printing/Mainframe-Print-Servers/RPM Elite (hisprint02)'], groupPaths=['/ITS/Systems'], performanceMonitor='cornea'

[11-May-2010 12:52:48] <LW> any way to get a complete list of schedules?

[11-May-2010 13:00:13] <gwb2351> ok, the manageIp bit, even though in the --sample_configs output, is not valid to zenbatchload

[11-May-2010 13:00:50] <mrayzenoss> LW: hmmm… there's a report in Enterprise to show all maintenance windows

[11-May-2010 13:06:57] <LW> rats

[11-May-2010 13:07:45] <LW> i used uniq and cut on the zenactions logs to get me a list of likely groups, i'm going to move their window start dates to 2010 and see if the problem goes away

[11-May-2010 13:16:32] <gwb2351> oddly enough /opt/zenoss/Products/ZenModel/BatchDeviceLoader.py has code for manageIp parsing, but I must be doing something wrong

[11-May-2010 13:19:56] <gwb2351> ah ha... need to put quotes about the manageIp.  Bad --sample_config, bad!

[11-May-2010 13:20:35] <mrayzenoss> gwb2351: please open a ticket, that's bad

[11-May-2010 13:23:43] <gwb2351> ticket created, simple fix

[11-May-2010 14:35:31] <chemist> anyody use xmppBot zenpack?

[11-May-2010 14:53:49] <chemist> or even anybody

[11-May-2010 15:00:46] <mrayzenoss1> chudler's your man, don't see him here

[11-May-2010 15:00:51] mrayzenoss1 is now known as mrayzenoss

[11-May-2010 15:01:56] <chemist> does he come here at all? can't remember if I've seen him or not

[11-May-2010 15:02:55] <mrayzenoss> he definitely shows up here sometimes

[11-May-2010 15:03:01] <mrayzenoss> just not today apparently :(

[11-May-2010 15:03:29] <chemist> I'll keep an eye out for him :)

[11-May-2010 15:07:20] <chemist> mrayzenoss: how's the alpha testing going?

[11-May-2010 15:07:37] <chemist> wish I had time to take part

[11-May-2010 15:10:46] <mrayzenoss1> testing is going well, alphas keep rolling out and bugs are getting found and fixed

[11-May-2010 15:12:42] <bigegor> is it possible to set zProperty by device modelling? I want to set zLinks property.

[11-May-2010 15:13:15] <twm1010> you mean, model the device, and based on information returned set a zProperty at the device level?

[11-May-2010 15:13:23] <bigegor> yes

[11-May-2010 15:13:50] mrayzenoss1 is now known as mrayzenoss

[11-May-2010 15:13:53] <twm1010> I don't see why not, during device auto discovery it will set zSnmpCommunity to whichever one worked

[11-May-2010 15:14:16] <eon> tetsu0

[11-May-2010 15:14:35] <twm1010> but that's a one time thing, i've never seen an automated re-model set any zProperty

[11-May-2010 15:17:05] <bigegor> it doesn't work. I can not set zProperty with ObjectMap :(

[11-May-2010 16:14:42] <cgibbons> You can always swizzle up a patch to ApplyDataMap.py that supports that and it has a better chance of making it in. Not that it helps you short-term.

[11-May-2010 16:26:02] <bigegor> thanks

[11-May-2010 18:35:23] <jeffi> hi, does anyone here have already used Oracle Zenpack ?

[11-May-2010 18:58:13] forsberg is now known as fOrsberg

[12-May-2010 00:00:30] [disconnected at Wed May 12 00:00:30 2010]

[12-May-2010 00:00:30] [connected at Wed May 12 00:00:30 2010]

[12-May-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[12-May-2010 01:40:20] fOrsberg is now known as forsberg

[12-May-2010 03:13:39] <zykes-> Connection was refused by other side: 111: Connection refused. < what does that mean ?

[12-May-2010 03:13:57] <JesperS> where do you get it?

[12-May-2010 03:14:06] <zykes-> . /Events /Cmd /Fail / Event Console

[12-May-2010 03:14:07] <JesperS> but it looks like the server you are trying to connect to is refusing it

[12-May-2010 03:17:55] <zykes-> zenmodeler is complaining

[12-May-2010 04:18:39] <JesperS> The other day I asked about how to move a device into a specific class, from command-line that is. I was suggested to have a look at doing it via python but I can't seem to get it working.

[12-May-2010 04:18:53] <JesperS> I'm hoping that one of you guys have a example script for this purpose?

[12-May-2010 04:19:49] <Simon4> dev = dmd.Devices.findDevice('devicename')

[12-May-2010 04:20:30] <Simon4> dev.changeDeviceClass('/Devices/Where/You/Choose')

[12-May-2010 04:20:32] <Simon4> commit()

[12-May-2010 04:20:42] <Simon4> worked for me yesterday

[12-May-2010 04:20:45] <JesperS> sweet, that looks promising

[12-May-2010 04:21:13] <JesperS> kind of a uphill struggle here, I'm fairly new at Zenoss AND I've never touched python before... :-/

[12-May-2010 04:24:42] <JesperS> brilliant, it works. Thank you very much Simon4, who would have thought it was so easy!   I was trying to use moveDevices and other functions like that.

[12-May-2010 04:25:09] <Simon4> no probs... I wrote a large part of a "delete and re-add" script before finding that function call, so can feel your pain :)

[12-May-2010 04:25:20] <JesperS> damn google for not finding the thing I needed, istead of the thing I searched for! ;-)

[12-May-2010 04:25:45] <tehhobbit> JesperS: google isnt quite that agile .... yet

[12-May-2010 07:01:29] <zykes-> anyone here able to use WMI ?

[12-May-2010 07:01:36] <zykes-> without being local admin og domain admin

[12-May-2010 07:36:35] forsberg is now known as fOrsberg

[12-May-2010 08:37:03] <Egyptian[Laptop]> afternoon all

[12-May-2010 08:37:16] <Egyptian[Laptop]> i am tryign to do some dns monitoring and following the extended guide

[12-May-2010 08:37:41] <Egyptian[Laptop]> i have the dig monitor zenpack enabled but when i come to add teh templeate i onl y find the dns monitor .. no dig monitor .. help

[12-May-2010 08:47:55] <Egyptian[Laptop]> started zenoss .. even then .. no luck

[12-May-2010 08:51:33] <baronvaile> Really quick, what units are the errors/sec graph for interfaces in?

[12-May-2010 08:55:55] <rmatte> you pretty much just answered the question

[12-May-2010 08:56:02] <rmatte> errors per second

[12-May-2010 08:56:28] <rmatte> each section on the graph represents a 5 minute polling window

[12-May-2010 08:56:41] <rmatte> so it's the amount of errors per second seen during 5 minutes

[12-May-2010 08:57:08] <baronvaile> so 111.69m would be in millions of errors?

[12-May-2010 08:57:13] <rmatte> nope

[12-May-2010 08:57:22] <rmatte> 111.69M would be millions

[12-May-2010 08:57:29] <rmatte> m is millionths

[12-May-2010 08:57:35] <rmatte> as in, millimeters for instance

[12-May-2010 08:58:01] <rmatte> so if you take 111.69 and multiply that by 300 seconds (5 minutes)

[12-May-2010 08:58:20] <rmatte> you'll get the total amount of errors over that cycle

[12-May-2010 08:58:31] <rmatte> well, not quite actually

[12-May-2010 08:58:46] <rmatte> more like multiple 0.11169 or something

[12-May-2010 08:58:55] <rmatte> in short, it's not much

[12-May-2010 08:59:12] <rmatte> maybe an error or 2 over that 5 minute cycle

[12-May-2010 09:00:59] <baronvaile> ok, I'm new to monitoring T1 interfaces on our equipment and didn't know the math. Our enterprise network guy usually does that, but he is not at this location so I put in equipment to track.

[12-May-2010 09:01:02] <baronvaile> Thank you.

[12-May-2010 09:01:15] <rmatte> yeh, for a T1 interface that's more than acceptable

[12-May-2010 09:01:28] <rmatte> anything below 1 or 2% is considered acceptable by carriers

[12-May-2010 09:02:22] <baronvaile> wow, thanks again. I didn't know the SLA agreement levels.

[12-May-2010 09:02:53] <rmatte> np, I'm sysadmin for a NOC, so I run across this sort of stuff a lot

[12-May-2010 09:03:21] <rmatte> anyways, last night was change night and I've only had 5 hours of sleep so I'm going back to bed

[12-May-2010 09:03:23] <rmatte> ttyl

[12-May-2010 09:03:54] <baronvaile> Maybe I can pick your brain later about how you got to that level?

[12-May-2010 09:04:08] <baronvaile> but thanks for now.

[12-May-2010 09:05:34] <rmatte> sure

[12-May-2010 09:06:03] <baronvaile> I see your on via a Madison WI address, is that correct?

[12-May-2010 09:06:32] <baronvaile> I'm Baronvaile@verizon.net

[12-May-2010 09:06:39] <rmatte> It's actually Atlanta, GA, but I'm from Ottawa, Canada.  I'm just connected via my private server.

[12-May-2010 09:07:01] <baronvaile> ah, ok.

[12-May-2010 09:07:36] <baronvaile> oh, your freenode server is in Madison. My bad.

[12-May-2010 09:07:43] <rmatte> ah

[12-May-2010 09:07:52] <rmatte> yup, appears to be

[12-May-2010 09:08:18] <baronvaile> anyway, go get some sleep; or what ever you want to do off shift.

[12-May-2010 09:08:36] <rmatte> hehe

[12-May-2010 09:08:38] <rmatte> later

[12-May-2010 09:08:46] <venturaville> rmatte: you are in Atlanta?

[12-May-2010 09:09:15] <rmatte> venturaville: no, my private server is, hosted by sagonet

[12-May-2010 09:09:18] <venturaville> ah

[12-May-2010 09:09:21] <rmatte> ;)

[12-May-2010 09:09:30] <venturaville> probably right up the street from us then

[12-May-2010 09:09:36] <rmatte> probably

[12-May-2010 09:09:41] <rmatte> :)

[12-May-2010 09:09:55] <rmatte> ok, I need sleep or I shall die today, ttyl folks

[12-May-2010 10:14:33] <coofamani> anyone know if availability calculations take maintenance windows into account?

[12-May-2010 10:15:05] <gwb2351> grrr... I just added 300+ devices and meant to make sure they didn't use one of the templates from the device class they were added to.  What's the best way to take a list of device names and say "remove this template" ?

[12-May-2010 10:18:16] <rmatte> coofamani: yes, they do

[12-May-2010 10:18:27] <rmatte> coofamani: availability is only affected if a device is in a Production state

[12-May-2010 10:21:30] <coofamani> ok, so I'm trying to install this zenpack (docs/DOC-4642) but its trying to install to $ZENHOME/Products instead of $ZENHOME/ZenPacks. I've looked through the source and objects.xml but dont see where it's getting that path from

[12-May-2010 10:21:36] <coofamani> any suggestions?

[12-May-2010 10:22:04] <coofamani> 'ERROR: zenpack command failed. Reason: OSError: [Errno 20] Not a directory: '/opt/zenoss/Products/ZenPacks.community.AvailabilityReportPerGroup-1.0-py2.4.egg/skins' '

[12-May-2010 10:22:43] <coofamani> additionally, /opt/zenoss/Products does exist... but why is it looking for it there

[12-May-2010 10:25:53] <rmatte> coofamani: let me guess, it's a .egg.zip

[12-May-2010 10:26:03] <rmatte> and you didn't unzip it so that it's just a .egg before installing?

[12-May-2010 10:26:16] * coofamani damnit

[12-May-2010 10:26:18] <rmatte> (The zenoss site zips them automatically, but they aren't supposed to be zipped)

[12-May-2010 10:26:32] <coofamani> thanks

[12-May-2010 10:26:45] <rmatte> no problem

[12-May-2010 10:26:50] <rmatte> well, time for me to head in to work

[12-May-2010 10:26:51] <rmatte> bbl

[12-May-2010 10:29:47] fOrsberg is now known as forsberg

[12-May-2010 10:33:40] <nzle> I'm trying to setup a template to run a command on a remote server.  The command returns 0 or 1 and no output.  I added a template for the host, and added a COMMAND datasource, and checked the 'Use SSH' and put the path and filename in the Command Template.   When I hit test though it returns no such file or dir, and the ocmmand output doesnt look like it sshed anywhere.  creds are in the hosts zproperties, do I need to enable somehting else??

[12-May-2010 11:07:21] <gwb2351> I have a /Systems hierarchy with 362 devices.  About 20 of them have a red alert next to the device, but the Summary events only shows 6 red alerts.  Ideas?

[12-May-2010 11:08:21] <gwb2351> (summary events for that part of the /Systems hierarchy.  the summary events in /Devices looks shows all the devices properly)

[12-May-2010 11:25:12] <twm1010> mmm..

[12-May-2010 11:25:24] <twm1010> so you're looking at the top level organizer of /Systems?

[12-May-2010 11:25:39] <twm1010> are there suborganizers?

[12-May-2010 13:00:29] <gwb2351> twm1010: yes, there are sub organizers.  The question is now moot; I added an event transform, and suddenly everything appears where it should be.  thanks

[12-May-2010 13:06:48] forsberg is now known as fOrsberg

[12-May-2010 13:11:00] <gwb2351> A bigger PITA I'm dealing with now is trying to add a System Administered Object to a ZenUser and getting Proxy Error.  Sigh.

[12-May-2010 13:20:10] <gwb2351> the event.log has this "ReadConflictError: database read conflict error (oid 0x09, class Products.Transience.Transience.Increaser)"

[12-May-2010 13:24:51] <gwb2351> reindex the DB? http://www.google.com/#hl=en&q=ReadConflictError%3A+database+read+conflict+error+zenoss&aq=f&aqi=&aql=&oq=&gs_rfai=&fp=826e5216022ce3e9

[12-May-2010 13:41:19] <rmatte> yes, reindex

[12-May-2010 13:41:31] <rmatte> reindex, commit, restart zope

[12-May-2010 13:41:38] <gwb2351> zopectl restart?

[12-May-2010 13:41:41] <rmatte> yes

[12-May-2010 14:00:42] <gwb2351> boo.  no love.  may it have something to do that the System Administered Object has ~350 devices?

[12-May-2010 14:00:59] <rmatte> doubt it

[12-May-2010 14:01:12] <gwb2351> (2 conflicts, of which 0 were unresolved, since startup at Wed May 12 13:58:16 2010)

[12-May-2010 14:01:48] <gwb2351> try restarting zeo as well?

[12-May-2010 14:02:11] <coofamani> '0 were unresolved' = life is good, no?

[12-May-2010 14:02:19] <rmatte> that's fine

[12-May-2010 14:02:22] <rmatte> that's just routine

[12-May-2010 14:02:25] <coofamani> si si

[12-May-2010 14:02:55] <gwb2351> alas, immediately after that message the traceback happened "ReadConflictError: database read conflict error (oid 0x13, class BTrees._OOBTree.OOBTree)"

[12-May-2010 14:03:13] <rmatte> I had the same thing this morning when remodeling a device

[12-May-2010 14:03:23] <rmatte> so I reindex, commited, restarted zope, then remodeled a different device

[12-May-2010 14:03:27] <rmatte> then remodeled that one again

[12-May-2010 14:03:30] <rmatte> and the error went away

[12-May-2010 14:03:52] <coofamani> I would shut down all of the zenhubs and workers then reindex()

[12-May-2010 14:04:18] <rmatte> yeh, that might be a plan

[12-May-2010 14:04:25] <rmatte> actually, I restarted Zenoss entirely

[12-May-2010 14:04:28] <rmatte> rather than just zope

[12-May-2010 14:04:32] <rmatte> so maybe give that a shot

[12-May-2010 14:04:46] <gwb2351> i'll have to wait for off hours to shutdown everything, but thanks for the ideas.

[12-May-2010 14:04:54] <rmatte> k

[12-May-2010 14:17:37] <QbY> does anyone know if there is a way to monitor authoritative dns servers?  we had one crash and knew nothing about it until it was too late..  i'd like to set up a few hosts that i want zenoss to query against specific servers..  suggestions?

[12-May-2010 14:17:59] <rmatte> custom script?

[12-May-2010 14:18:01] <twm1010> Well, I see no way to differentiate between monitoring of authoritative and non-authoritative servers

[12-May-2010 14:18:15] <twm1010> by default, i mean, authoritative defined by how, a record lookup?

[12-May-2010 14:18:22] <QbY> well auth/non-auth doesn't really matter.. i just need to know that server A is answering queries

[12-May-2010 14:18:33] <twm1010> sure, there is a nagio check_dns script, i'm sure of it

[12-May-2010 14:18:35] <rmatte> I think he just means he wants to do queries against multiple nameservers

[12-May-2010 14:18:39] <rmatte> and if a query fails, alert

[12-May-2010 14:18:46] <QbY> tep

[12-May-2010 14:18:48] <QbY> yep

[12-May-2010 14:19:50] <twm1010> read the admin guide for setting up command datasources

[12-May-2010 14:19:58] <twm1010> and set one up to hit http://nagiosplugins.org/man/check_dns

[12-May-2010 14:21:22] <twm1010> hrmm.. thanks for bringing this up, been meaning to do it :)

[12-May-2010 14:22:16] <twm1010> wait a tick, docs/DOC-3511

[12-May-2010 14:22:25] <twm1010> there's a core zenpack for that

[12-May-2010 14:29:53] <rmatte> well, there you go

[12-May-2010 14:29:58] <rmatte> grab the dns zenpack and you're set

[12-May-2010 14:30:35] <twm1010> And almost made him do a whole bunch of work :)

[12-May-2010 14:31:28] <rmatte> :)

[12-May-2010 14:58:04] <sektorNBA> nice topic

[12-May-2010 15:22:47] <gwb2351> rmatte: there's a knowledge base document "Database Conflict Errors" that shows how to change the "Timeout resolution(in seconds)" value from 20 (default) to a 1200 (suggested) to resolve errors.  It Works For Me(tm)!

[12-May-2010 15:23:01] <rmatte> cool

[12-May-2010 15:23:30] <rmatte> I wonder what kind of performance hit you'll take from that though

[12-May-2010 15:24:29] <gwb2351> it doesn't seem like anything -- it's mainly a timeout, which was happening faster than the write

[12-May-2010 15:26:22] <gwb2351> did you see the private msg with the paste (would be annoying to others if I pasted it here)

[12-May-2010 15:30:00] <rmatte> I just saw (paste)

[12-May-2010 15:30:16] <rmatte> oh nevermind

[12-May-2010 15:30:20] <rmatte> you literally pasted that

[12-May-2010 15:30:22] <rmatte> yes, I saw it

[12-May-2010 15:30:57] <gwb2351> so I don't think there will be a perf hit, unless some other long-running query is actually infinite (taking CPU resources)

[12-May-2010 15:36:02] <rmatte> yeh

[12-May-2010 15:50:17] <mrayzenoss> so I've got an RRD question

[12-May-2010 15:51:27] <mrayzenoss> where do I set the Y-axis scale (not the label)?  I've got load values from 0-.9 but my y-axis is going 0-800

[12-May-2010 15:53:02] <Simon4> in a graph definition? there's a min Y and max Y value you can set

[12-May-2010 15:53:18] <Simon4> by default rrdgraph tries to autofit these to the data

[12-May-2010 15:53:37] <mrayzenoss> but that's the thing, the graph is taking the .9 value and graphing it at 900

[12-May-2010 15:54:20] <Simon4> ah, you could put a RPN of 900,/ in the graph point definition

[12-May-2010 15:54:23] <Simon4> gah

[12-May-2010 15:54:24] <Simon4> sorry

[12-May-2010 15:54:27] <Simon4> 100,/

[12-May-2010 15:54:33] <Simon4> brain not working :)

[12-May-2010 15:55:43] <mrayzenoss> that would fix the graph, would that change the values being listed (cur, avg, max)?

[12-May-2010 15:56:02] <Simon4> It hink it does, I'm pretty sure that's what we do here

[12-May-2010 15:56:06] <Simon4> 2 secs I'll check

[12-May-2010 15:57:45] <Simon4> it sorts that also

[12-May-2010 15:58:02] <Simon4> our "Device" template in /Devices/Server does the 100,/ RPN for load and the cur/avg/max are all correct

[12-May-2010 15:58:55] <mrayzenoss> yeah, that's what I'm looking at

[12-May-2010 15:59:04] <mrayzenoss> since I'm doing load via SSH

[12-May-2010 15:59:24] <Simon4> looks like the RPN is applied before any summary calculations so you should be fine

[12-May-2010 16:00:00] <mrayzenoss> hmm… ok, I'll try putting those in and see what graphs

[12-May-2010 16:00:27] <Simon4> it dosn't affect the content of the RRD files, so there's no loss in trying it anyway

[12-May-2010 16:02:01] <mrayzenoss> I wasn't sure if the SNMP value stored for load was a float or not

[12-May-2010 16:03:42] <mrayzenoss> when I saw the 100,/ I assumed it wasn't

[12-May-2010 16:05:26] <Simon4> we get load via snmp, and yeah, I'm pretty sure it's just an integer

[12-May-2010 16:05:48] <mrayzenoss> hmm… I'm getting mine via SSH and I can see in the zencommand.log it's stored as a float

[12-May-2010 16:08:35] <Simon4> <!-- 2010-03-30 11:00:00 BST / 1269943200 --> <row><v> 1.4314803486e+02 </v></row>

[12-May-2010 16:08:37] <Simon4> I tell a lie

[12-May-2010 16:08:39] <Simon4> sorry

[12-May-2010 16:08:55] <Simon4> float it is

[12-May-2010 16:09:32] <rmatte> lol, I'm working my way through upgrading an Ubuntu Server 8.04 install to 9.10 right now

[12-May-2010 16:09:40] <rmatte> it's boatloads of fun so far

[12-May-2010 16:09:50] <rmatte> :)

[12-May-2010 16:09:53] <mrayzenoss> no 10.04?

[12-May-2010 16:10:06] <rmatte> actually yeh, forgot 10.04 is out

[12-May-2010 16:10:15] <rmatte> I need to upgrade both my workstations to that

[12-May-2010 16:10:52] <rmatte> unhunh, well so far this upgrade isn't going so well, tons of failures during bootup

[12-May-2010 16:11:22] <rmatte> I might just end up sticking with 8.04

[12-May-2010 16:12:13] <mrayzenoss> worst part about messing with RRD is waiting on new graphs

[12-May-2010 16:12:21] <rmatte> yup

[12-May-2010 16:12:29] <Simon4> oui

[12-May-2010 16:17:41] <mrayzenoss> hmm… my scale now goes 0-8, but the values listed are all 0.01, when I know the load is really around .6

[12-May-2010 16:18:23] <mrayzenoss> stupid RRD… be back in a few minutes

[12-May-2010 16:18:24] <Simon4> mrayzenoss: have you had a look at the values in the rrd file itself with rrdtool dump filename.rrd ?

[12-May-2010 16:18:37] <rmatte> you removed the %s in format?

[12-May-2010 16:18:49] <mrayzenoss> Simon4:  I saw this in the log: 2010-05-12 14:08:03,388 DEBUG zen.zencommand: RRD save result: 0.62

[12-May-2010 16:19:00] <mrayzenoss> rmatte: what's that?

[12-May-2010 16:19:30] <mrayzenoss> rmatte: format for my Graph Points: %6.2lf

[12-May-2010 16:20:28] <mrayzenoss>                <!-- 2010-05-12 12:12:00 PDT / 1273691520 --> <row><v> NaN </v></row>

[12-May-2010 16:22:56] <mrayzenoss> Simon4:                <!-- 2010-05-12 12:12:00 PDT / 1273691520 --> <row><v> NaN </v></row>

[12-May-2010 16:23:21] <mrayzenoss>                <!-- 2010-05-12 14:12:00 PDT / 1273698720 --> <row><v> 6.3000000000e-01 </v></row>

[12-May-2010 16:23:24] <mrayzenoss> data :)

[12-May-2010 16:23:52] <mrayzenoss> so mine's stored as a float…

[12-May-2010 16:23:57] <mrayzenoss> gotta run, be back in a bit

[12-May-2010 16:24:16] <Simon4> hurrah :)

[12-May-2010 16:31:33] <rmatte> mrayzenoss: you already removed it

[12-May-2010 16:31:42] <rmatte> default is %5.2lf%s

[12-May-2010 16:38:41] <mrayzenoss> and I'm back

[12-May-2010 17:03:08] <mrayzenoss> a-ha!

[12-May-2010 17:03:15] <mrayzenoss> it appears that setting the Max Y fixed it

[12-May-2010 17:07:35] <rmatte> nice

[12-May-2010 17:07:41] <rmatte> well, I'm out, later

[12-May-2010 17:08:10] <mrayzenoss> I'll be adding this to my OS X ZenPack tutorial later tonight

[12-May-2010 17:08:36] <Simon4> os x zenpack? for monitoring os x?

[12-May-2010 17:13:47] <mrayzenoss> yes

[12-May-2010 17:14:05] <mrayzenoss> Simon4: docs/DOC-5865

[12-May-2010 17:14:24] <Simon4> mrayzenoss: awesome

[12-May-2010 17:14:50] <mrayzenoss> I'm doing uptime and load right now

[12-May-2010 17:15:07] <mrayzenoss> so #4 will be up tonight

[12-May-2010 17:15:27] * Simon4 will be stealin gsome

[12-May-2010 17:15:27] <Simon4>

[12-May-2010 17:15:29] <Simon4> gah

[12-May-2010 17:15:36] <mrayzenoss> and as I write the ZenPack, I'm trying to doc the steps and changes, and what needs to be done to make sure ZenPacks work with 3.0

[12-May-2010 17:15:46] <Simon4> stealing info from that tutorial for my bladechassis ssh collection

[12-May-2010 17:15:53] <Simon4> so very nicely timed :)

[12-May-2010 17:16:05] <mrayzenoss> well, I'm just stealing from the OpenSolaris and LinuxMonitor ZenPacks

[12-May-2010 17:16:14] <Simon4> *nod*

[12-May-2010 17:16:21] <mrayzenoss> but documenting it, which is where the value is :)

[12-May-2010 17:16:24] * Simon4 has spent a fair bit of time digging in those also

[12-May-2010 17:16:53] <mrayzenoss> so each tutorial goes with a feature, and gets a tagged version and you can diff versions

[12-May-2010 17:17:14] <mrayzenoss> it's the ZenPack tutorial I've always wanted, so I decided to write it myself

[12-May-2010 17:17:22] <Simon4> yeah

[12-May-2010 17:17:34] <Simon4> I started a blog post along the same lines about three weeks ago, then went on holiday

[12-May-2010 17:17:57] <aclark> mrayzenoss: nice

[12-May-2010 17:18:01] <Simon4> devs in tomorrow?

[12-May-2010 17:18:05] <mrayzenoss> yup

[12-May-2010 17:18:13] <Simon4> cool

[12-May-2010 17:18:25] * Simon4 has a hopefully easy question to help tidy things up at work before enterprise

[12-May-2010 17:19:10] * Simon4 got back from holiday to a meeting at 9am Monday "we want enterprise yesterday please"

[12-May-2010 17:24:58] <straterra> i havent upgraded zenoss in a loooong time

[12-May-2010 17:42:46] fOrsberg is now known as forsberg

[12-May-2010 20:46:10] forsberg is now known as fOrsberg

[12-May-2010 20:48:20] <cgibbons> shew

[13-May-2010 00:00:30] [disconnected at Thu May 13 00:00:30 2010]

[13-May-2010 00:00:31] [connected at Thu May 13 00:00:31 2010]

[13-May-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[13-May-2010 02:44:50] fOrsberg is now known as forsberg

[13-May-2010 05:33:17] <zykes->  what's the difference between systems and groups ?

[13-May-2010 05:58:56] <forsberg> if nothing else, its 2 different "filters" you can use in alerting rules and others

[13-May-2010 07:13:37] <smbambling> Hi All.

[13-May-2010 07:16:03] <smbambling> I'm having a little trouble I hope someone might be able to help with.  I added a new OS Manufacturer for CentOS and a new OS Product of 5.4.  When trying to edit or added these settings to a device it doesn't seem to take.  After I select the OS Manufacturer drop down for CentOS the OS product drop down is empty.

[13-May-2010 07:17:15] <smbambling> When I select the Products  (/dmd) from the pin menu I see that CentOS was created and that it has a Type Operating System set for 5.4

[13-May-2010 07:30:41] <fus10nx> Is ZenOSS avil. in any other languages or just English ?

[13-May-2010 07:38:56] <fus10nx> Anyone who's just joined know if ZenOSS avil. in any other languages or just English ?

[13-May-2010 07:39:20] <fus10nx> Simon4: was it you I was speaking w/ who had integrated ZenOSS and a ticket management system together ?

[13-May-2010 07:39:30] <Simon4> not me, sorry

[13-May-2010 07:39:54] <fus10nx> I'm trying to find people who have done more w/ ZenOSS than just the stock installation - some integration into other systems

[13-May-2010 07:40:08] <coofamani> fus10nx: I think ryan did (rmatte)

[13-May-2010 07:40:14] * Simon4 needs to get alerts tuned better before having it open tickets

[13-May-2010 07:40:24] <fus10nx> thanks coofamani

[13-May-2010 07:40:39] * Simon4 plays with zenoss discovery

[13-May-2010 07:40:39] <coofamani> Im going to be tying it in with Splunk soon, but havent started the planning yet

[13-May-2010 07:40:42] <fus10nx> we have a german company interested but it appears its English only

[13-May-2010 07:40:43] <Simon4> lets see what mess it makes :)

[13-May-2010 07:40:47] <coofamani> tieing

[13-May-2010 07:40:50] <coofamani> man that looks odd

[13-May-2010 07:42:04] <fus10nx> I saw Nagios the other day at a trade show, lots of companies demoing it in some form of integration into their software but it looks like crap compared to ZenOSS

[13-May-2010 07:42:41] <cgibbons> imagine how bad it'll look once the new UI is out

[13-May-2010 07:43:24] <fus10nx> whats that? nagios ?

[13-May-2010 07:43:28] <fus10nx> or ZenOSS

[13-May-2010 07:43:36] <cgibbons> zenoss

[13-May-2010 07:43:50] <cgibbons> zenoss's new ui, the nagios stuff will look even worse

[13-May-2010 07:45:07] <smbambling> Has anyone had issues adding additional OS Manufacturers ?

[13-May-2010 07:50:01] <fus10nx> is there screenshots of the new UI?

[13-May-2010 07:58:50] <cgibbons> even better, public-demo.zenoss.com

[13-May-2010 07:59:08] <cgibbons> and yes there are, although I do not know where they live (mattray will once he arrives)

[13-May-2010 08:00:08] <smbambling> login for public demo ?

[13-May-2010 08:01:07] <fus10nx> ya, what's the login ?

[13-May-2010 08:01:17] <cgibbons> hrm it is not what i would have expected. let's ask matt when he shows up :)

[13-May-2010 08:01:39] <cgibbons> shouldn't be too much longer

[13-May-2010 08:02:23] <cgibbons> it gets horked up pretty frequently, too, as you can imagine :)

[13-May-2010 08:03:50] <fus10nx> admin / zenoss

[13-May-2010 08:04:19] <fus10nx> ya its brutally slow, naturally haha - prob. loads of people connecting all the time

[13-May-2010 08:04:56] <fus10nx> i kinda like the old interface better how the navigation is on the left not the top

[13-May-2010 08:04:57] <zykes-> any reason for not being able to lineup two things besides each other in the dashboard ?

[13-May-2010 08:05:25] <fus10nx> although the new interface is quite slick on the tree menu in infrastrucutre

[13-May-2010 08:06:18] <cgibbons> I'm sure anyone used to the old interface will probably feel like a fish out of water for a long while. But we worked with one of the top usability engineers available to simplify the entire product and hence the UI, so I think the end result is going to be a big win, and certainly will be for anyone new (easier ramp up time).

[13-May-2010 08:06:41] <fus10nx> ya it makes more sense for sure even after playing with it for a minute or two

[13-May-2010 08:06:48] <fus10nx> how locations break down and stuff

[13-May-2010 08:11:05] <fus10nx> cgibbons: is tehre plans to have native WMI support?

[13-May-2010 08:11:22] <fus10nx> instead of it being a ZenPack I mean

[13-May-2010 08:11:40] <zykes-> any reason for not being able to lineup two things besides each other in the dashboard ?

[13-May-2010 08:11:54] <cgibbons> You mean a full set of monitoring templates, etc. for doing performance monitoring via WMI?

[13-May-2010 08:11:59] <fus10nx> zykes-: should line up just keep trying to adjust properly

[13-May-2010 08:12:05] <fus10nx> or send a screenshot so I can see what you mean

[13-May-2010 08:12:11] <fus10nx> cgibbons: i suppose, ya

[13-May-2010 08:12:29] <fus10nx> I am trying to avoid paying thousands to get the snmp software you guys recommend

[13-May-2010 08:13:03] <cgibbons> No plans short-term. The Enterprise version of the product has a full suite of Windows monitoring that uses WMI & PerfMon for the performance collection, so we've let the community augment what core has in it already.

[13-May-2010 08:13:04] <smbambling> besides the admin guide that zenoss post are they any other documents that I should read for configuring and setting up zenoss fully ?

[13-May-2010 08:14:17] <fus10nx> ya snmp informare

[13-May-2010 08:14:20] <fus10nx> informer*

[13-May-2010 08:14:38] <fus10nx> http://www.snmp-informant.com/

[13-May-2010 08:16:11] <fus10nx> i guess im just looking to get as much info. as possible out of the system

[13-May-2010 08:16:24] <fus10nx> what more can enterprise do that core can't ?!

[13-May-2010 08:16:49] <cgibbons> It has a whole bunch of ZenPacks included that core doesn't

[13-May-2010 08:17:26] <zykes-> is it due to licening or ?

[13-May-2010 08:17:27] <cgibbons> I think you're probably better off using egor's community zenpacks for Windows than using snmp-informant. he's done a good job with them.

[13-May-2010 08:17:36] <smbambling> From my readings it supports a more granular detail of monitoring (I looked mostly at windows monitoring)

[13-May-2010 08:18:17] <fus10nx> ya I am using them but not all of them seem to work for one reason or another and ZenOSS gives back a ton of alerts because of it

[13-May-2010 08:18:54] <fus10nx> reports in new ZenOSS are awesome

[13-May-2010 08:19:10] <fus10nx> is public-demo.zenoss.com core or enterprise ?

[13-May-2010 08:19:20] <cgibbons> core

[13-May-2010 08:19:24] <fus10nx> awesome !

[13-May-2010 08:19:29] <fus10nx> cgibbons: do you work for ZenOSS?

[13-May-2010 08:19:33] <cgibbons> yeah

[13-May-2010 08:20:23] <fus10nx> ya i ALWAYS get 3 "Errors" when using Egor's stuff -- Problem while executing plugin community.wmi.InterfaceMap, Problem while executing plugin community.wmi.DiskDriveMap, Problem while executing plugin community.wmi.ProcessorMap

[13-May-2010 08:20:43] <fus10nx> always happens on multiple machines, xp, server 2k8, win7, vista

[13-May-2010 08:21:36] <gordy> Are there alternative skins, themes for zenoss?

[13-May-2010 08:21:49] <fus10nx> gordy: make your own ! it's open source !

[13-May-2010 08:22:25] <fus10nx> i changed all my colors around just by editing the CSS, it was pretty simple using firefox and firebug

[13-May-2010 08:22:26] <cgibbons> I must admit to having never used egor's zenpacks :) but he's on here pretty regularly

[13-May-2010 08:22:35] <fus10nx> ill have to check w/ him

[13-May-2010 08:22:40] <fus10nx> do you know what his nick is ?

[13-May-2010 08:22:44] <cgibbons> bigegor

[13-May-2010 08:22:51] <fus10nx> cool, thanks !

[13-May-2010 08:23:34] <fus10nx> out of curisoity, to get the performance graphs, do you need Egor's stuff

[13-May-2010 08:23:41] <fus10nx> or does ZenOSS out of the box do that?

[13-May-2010 08:24:26] <cgibbons> I mostly know the Enterprise version, but I think the deal with Windows stuff in core is that the only way to get performance graphs would be with SNMP Informant. The Windows SNMP agent providers very little performance information that way.

[13-May-2010 08:24:40] <cgibbons> Core has WMI functionality built-in, but only for doing service monitoring and event log monitoring.

[13-May-2010 08:24:44] <fus10nx> ahhh so maybe that's why it works on some but not all

[13-May-2010 08:24:46] <fus10nx> thanks for that

[13-May-2010 08:24:58] <fus10nx> any idea what the price of enterprise is?

[13-May-2010 08:25:05] <cgibbons> Enterprise augments that with a bunch of new modeling plugins and a feature to collect performance data via perfmon and it has a bunch of templates for that.

[13-May-2010 08:25:26] <cgibbons> I'm not, no. Once mrayzenoss shows up (soon! he's on skype now!) he can hook  you up.

[13-May-2010 08:25:37] <fus10nx> haha ok

[13-May-2010 08:25:52] <cgibbons> I know it's a whole lot cheaper than the big 4 :)

[13-May-2010 08:25:59] <fus10nx> who's the big 4?

[13-May-2010 08:26:02] <fus10nx> what's up gold ?

[13-May-2010 08:26:23] <fus10nx> I actually spoke w/ Tomi and Charles Dykes a while back thanks to Matt but I'll have to ask him for a little more info. re: pricing

[13-May-2010 08:26:44] <cgibbons> BMC PATROL, HP OpenView, CA's UniCenter, and Tivoli

[13-May-2010 08:26:59] <fus10nx> computer assosiates

[13-May-2010 08:27:15] <fus10nx> do you think ZenOSS is better than the CA stuff?

[13-May-2010 08:27:28] <cgibbons> I'm biased :) and I hate CA...

[13-May-2010 08:27:32] <fus10nx> haha

[13-May-2010 08:27:37] <fus10nx> I know Fujitsu use CA's stuff and they complain that it's too expensive

[13-May-2010 08:27:48] <cgibbons> a bunch of us @ Zenoss used to work on the BMC PATROL product, and yes in many respect's Zenoss is much better than that (and in others it isn't)

[13-May-2010 08:28:32] <fus10nx> so would it be safe to say that ZenOSS could be the "core" of monitoring a network operation center?

[13-May-2010 08:28:47] <cgibbons> oh sure

[13-May-2010 08:29:18] <fus10nx> cool

[13-May-2010 08:29:35] <cgibbons> For some history, one of the co-founders & the guy that did most of the early work, Zenoss was written to make setting up nagios easier to build & manage based upon his own experience working in a NOC

[13-May-2010 08:29:37] <fus10nx> so you know what was interesting to me, that snmp informant makes all their MIB files available online

[13-May-2010 08:30:05] <fus10nx> http://www.wtcs.org/informant/mibs.htm   -- so does that mean you can simply download from their site and upload into ZenOSS ?

[13-May-2010 08:30:41] <cgibbons> you could, although uploading a MIB into zenoss isn't super useful. we don't use it for performance collection, it's only used if you want to get text for events.

[13-May-2010 08:30:45] <coofamani> fus10nx: zenoss for a NOC - definitely

[13-May-2010 08:31:24] <fus10nx> cgibbons: awww that's too bad. We have an LCD unit that provides a MIB file - but maybe it is simply just for reporting/data collection

[13-May-2010 08:31:57] <fus10nx> wait then im so confused

[13-May-2010 08:31:59] <cgibbons> it's pretty trivial to create a new template but you'd just use the OIDs directly for that.

[13-May-2010 08:32:13] <gordy> Is there a zenoss mailing list ?

[13-May-2010 08:32:55] <fus10nx> so snmp informant sells an "advanced version" of their software and its kinda pricey if you were to roll out to 1000 computers -- since they make the MIB available for free you are saying I could just use the MIB file in place of the software and load it into ZenOSS and then i'd get all those aditionally features at no cost?

[13-May-2010 08:33:15] <cgibbons> no, so it breaks down like this

[13-May-2010 08:33:28] <cgibbons> MIBs are just documents, they just describe the structure of information available

[13-May-2010 08:33:44] <cgibbons> snmp informant's free version is just crippleware, it has a bunch of features but not all

[13-May-2010 08:34:04] <cgibbons> but if MIB says OID 1.2.3.4.5 has data but their SNMP agent isn't providing it, then you can't get the data even if you know hte OID

[13-May-2010 08:34:31] <cgibbons> so the MIB is really irrelevant to anything, it's just a matter of what data is available from any given source

[13-May-2010 08:38:20] <fus10nx> i still dont fully get it but thanks for trying to explain. It seems like MIB files are pointless then

[13-May-2010 08:38:29] <cgibbons> really just view them as documentation

[13-May-2010 08:38:37] <fus10nx> ahhh ok

[13-May-2010 08:38:51] <fus10nx> that's interesting so our LCD must be able to send snmp traps somewhere

[13-May-2010 08:38:56] <fus10nx> i wonder if ZenOSS would handle it

[13-May-2010 08:39:01] <cgibbons> sure

[13-May-2010 08:39:06] <fus10nx> because its not like a cisco device or a windows PC

[13-May-2010 08:39:13] <fus10nx> i figure it would require custom programming then

[13-May-2010 08:39:35] <fus10nx> its got things like backlight status, temperature sensor readngs, etc.

[13-May-2010 08:39:53] <cgibbons> for example, SNMP OID 1.3.6.1.2.1.1.5 is really the name of the system... if you load a MIB that has it defined properly, it would have a verbose name of iso.org.dod.internet.mgmt.mib-2.system.sysName and a description of what it is, along with datatype information.

[13-May-2010 08:41:07] <fus10nx> so if I clearly understand, SNMP does nothing with control. It simply just reports on stuff but doesn't allow you to reboot a machine or something like that

[13-May-2010 08:41:16] <fus10nx> its strickly just a transport method for providing information

[13-May-2010 08:42:32] <cgibbons> it can do more than that, since you can set values in addition to read them or trap on them... many devices can be rebooted by setting a specific value, etc.

[13-May-2010 08:42:49] <fus10nx> ok thats good to know

[13-May-2010 08:42:53] <gordy> So is there no 'mailman' style mailing list? Just the jive-based web thing on the zenoss site ?

[13-May-2010 08:42:56] <fus10nx> im looking at a MIB file sent to us from the LCD manufacturer

[13-May-2010 08:43:20] <fus10nx> i see lots of "read-only" on status' so I'm guessing it's just to provide information to some other piece of software

[13-May-2010 08:43:26] <cgibbons> yep

[13-May-2010 08:43:34] <zykes-> so you would say that ZenOSS is better then BMC ?:)

[13-May-2010 08:44:11] <cgibbons> gordy, check with mrayzenoss when he shows up. i think you can subscribe to the forums via a mailing list but i'm not sure of the semantics anymore. it used to be mailman before we moved to jive... but he would be the one to know.

[13-May-2010 08:44:12] <fus10nx> and ZenOSS could be that "other piece of software" to data collect from my LCD, correct?

[13-May-2010 08:44:24] <cgibbons> yep & yep

[13-May-2010 08:44:51] <fus10nx> would it require additional programming

[13-May-2010 08:44:52] <fus10nx> ?

[13-May-2010 08:45:11] <cgibbons> nah, just UI-level customization of a template

[13-May-2010 08:45:19] <fus10nx> nice ! thanks for that !

[13-May-2010 08:45:31] <cgibbons> depending upon how fancy you get

[13-May-2010 08:45:53] <cgibbons> where you wind up needing to write code is if you've got some sort of custom protocol, or if you need to represent a more complex device.

[13-May-2010 08:46:26] <cgibbons> but if this thing is just rather straight-forward and there are a few performance metrics you want to monitor (like say temperature), then you can do it all via UI

[13-May-2010 08:46:56] <fus10nx> thats great to hear. Now I need to find someone who's done a full blown install w/ SNMP information premium and see what that've done w/ it and ZenOSS

[13-May-2010 08:47:03] <fus10nx> i.e. registry viewing/command execution/etc

[13-May-2010 08:47:57] <zykes-> cgibbons: my question, zenoss vs bmc, what poison ?

[13-May-2010 08:48:46] <gordy> Is there method of having devices but not have them report in the event console (until needed) ? (I've  inherited management of this zenoss system and have used zabbix in the past and there is much to relearn)

[13-May-2010 08:49:35] <cgibbons> the BMC product suite is much more mature and has more features, but it's less flexible and doesn't scale as well. but the classic BMC product is agent-based (they have an agentless one, too) so in some cases an agent solution might be better for your uses.

[13-May-2010 08:50:00] <fus10nx> so ZenOSS is NOT a replacement for WHat'sup Gold?

[13-May-2010 08:50:28] <cgibbons> no idea, i've seen that product before (worked with a big telco on replacing their use of what's up gold with the BMC product a few  years ago), but i don't know it's feature set.

[13-May-2010 08:57:41] <fus10nx> thanks for youre help cgibbons

[13-May-2010 08:58:01] <cgibbons> sure thing

[13-May-2010 08:59:20] <fus10nx> oh, one last thing

[13-May-2010 08:59:24] <fus10nx> does this keep track of uptime/downtime?

[13-May-2010 08:59:40] <fus10nx> so if a client says "when was it off, for how long was it off, when did it turn back on, how long has it been on" ?

[13-May-2010 09:00:36] <cgibbons> yeah, there's an 'up' attribute on a device and some reporting of an availability metric for a device. I'm not sure how rich the reports for available are though.

[13-May-2010 09:27:16] <cgibbons> there he is

[13-May-2010 09:27:24] * mrayzenoss ducks

[13-May-2010 09:34:18] * coofamani cringes at maint window interface

[13-May-2010 09:35:46] <cgibbons> the chicken blood interface

[13-May-2010 09:38:14] <coofamani> one down, 18 to go.

[13-May-2010 10:00:09] <mrayzenoss> Today's developer on call is mtzenoss

[13-May-2010 10:01:19] <mtzenoss> Hello

[13-May-2010 10:01:42] <smbambling> was a user/pass listed for public-demo.zenoss.com

[13-May-2010 10:02:06] <mrayzenoss> admin/zenoss

[13-May-2010 10:04:18] <smbambling> what version is this ?

[13-May-2010 10:04:51] <mrayzenoss> 3.0 alpha

[13-May-2010 10:05:09] <smbambling> cool.

[13-May-2010 10:05:32] <LarsN> is there an easy(ish) method to monitor the status of an nfs mount?

[13-May-2010 10:06:22] <LarsN> ie, it is, or is not mounted.

[13-May-2010 10:08:21] <smbambling> mray in the current version of zenoss I'm having some issue with adding additional OS Manufacturers.

[13-May-2010 10:09:07] <smbambling> When added to a device it does not show anything in the OS Procut drop down

[13-May-2010 10:09:30] <smbambling> Also is there any way to have zenoss automatically pull the centos release in this place ?

[13-May-2010 10:09:38] <Simon4> mtzenoss: I have a bunch of performance templates provided by a ZenPack (the zenpack was created on another install of zenoss for migration purposes). Is there a way to now remove the ZenPack from our current install without removing the performance templates? Basically want our current install to "forget" the zenpack was ever installed

[13-May-2010 10:11:50] <mrayzenoss> Simon4: I'm not sure you can do that

[13-May-2010 10:11:56] <mrayzenoss> I'll ask

[13-May-2010 10:13:07] <themactech> hi, since I have a crashed Zenoss, most of my questions today will be on fsrecover and zeopack

[13-May-2010 10:13:50] <mrayzenoss> smbambling: via SSH or SNMP?

[13-May-2010 10:13:57] <smbambling> SNMP

[13-May-2010 10:14:29] <themactech> i try fsrecover with the pack option and it still seems to try and copy over the full 88GB Data.fs file, is that because it is failing at the packing part or is it because its a two step process, where it must first do the recovery then pack it?

[13-May-2010 10:15:57] <themactech> also, I am now copying this huge file over, what will I need to load it in a test zenoss for recovery testing, what specific settings must match so this Data.fs file will load in a new box built just for recovering it?

[13-May-2010 10:16:35] <mrayzenoss> themactech: I believe zeopack makes a backup for recovery purposes, so that explains the copying

[13-May-2010 10:16:58] <mrayzenoss> themactech: I think you need to be on the same 32/64 bit and same version of Zenoss for the Data.fs to work

[13-May-2010 10:16:58] <themactech> I have not been able to use zeopack yet

[13-May-2010 10:17:30] <themactech> fsrecover asks for a target file and a destination, zeopack seems to require using the active Data.fs

[13-May-2010 10:18:00] <themactech> since my zenoss won't start up because it filled the disk, and is causing my linux to go fubar, I can't use zeopack

[13-May-2010 10:18:31] <themactech> That's why I'm now looking at transplanting this 88GB Data.fs into a test machine to try and fix it

[13-May-2010 10:20:06] <LarsN> themactech: if you have the same version of zope on another machine.  you should be able to use zeopack there on the data.fs.

[13-May-2010 10:20:08] <themactech> I made another linux/zenoss install on a smaller partition on that machine, once I got everything configured, I replace the /usr/local/zenoss/zenoss/var folder with a simlink to the same folder on the original parition.  Zenoss did not start up with the 88GB Data.fs file, I did not write down the error though.

[13-May-2010 10:20:42] <themactech> but zope is not enough right, I have to get zenoss up and running and USING this Data.fs file correct?

[13-May-2010 10:21:04] <themactech> zeopack asks for a port, I take it that means the Data.fs must be in use by the service

[13-May-2010 10:21:21] <LarsN> themactech: the Data.fs is just a zope thing.  There's no reason I can think of that it would require the entire zenoss installation.

[13-May-2010 10:22:08] <LarsN> themactech: doing a quick search on offline packing.

[13-May-2010 10:22:10] <mrayzenoss> themactech: you'll need to make sure all your ZenPacks are installed too

[13-May-2010 10:22:35] <mrayzenoss> before bringing over your Data.fs

[13-May-2010 10:22:38] <themactech> I will try tonight to copy this 88GB file over the internet (the zenoss server is in another state), VPN timeout issues have hampered this but I just figured out a workaround, might take 16 hours to copy this but then I will be able to kick it around much easier

[13-May-2010 10:22:49] <mrayzenoss> ouch

[13-May-2010 10:23:00] <mrayzenoss> bittorrent?  :p

[13-May-2010 10:23:00] <themactech> ok, maybe the zenpacks caused the Data.fs to not load up

[13-May-2010 10:24:05] <LarsN> themactech: I'd recommend making sure you have a backup of the Data.fs before trying this....  but: http://pypi.python.org/pypi/z3c.offlinepack/0.2

[13-May-2010 10:24:27] <LarsN> scratch that, that's for zope3

[13-May-2010 10:24:45] <LarsN> there's a similar zope2 package, I'm just not finding it.

[13-May-2010 10:25:04] <themactech> I will look into it, might be a good utility to keep installed on my zenoss boxes

[13-May-2010 10:25:19] <themactech> somewhat miffed that my zenoss DB blew up to 88GB and fubared my ubuntu

[13-May-2010 10:25:48] <themactech> worst thing is  i was schedule to clone the entire this last week, but that got put off by a week due to an urgent project

[13-May-2010 10:26:07] <LarsN> if it makes you feel any better, I had a customer try to insert ~ 65,000 records into a plone site (zope backed) then, when he realized that wasn't going to work, he programatically went to remove the 36k or so that he previously installed.

[13-May-2010 10:26:25] <mtzenoss> @Simon4 You can remove the packable relationship from the zeodb for the templates, .pack.remove()

[13-May-2010 10:26:38] <LarsN> the Data.fs grew to 127gb in under a day, filled /var and broke things good.

[13-May-2010 10:27:25] <LarsN> usually to pack a Data.fs you need 2x the mount of disk space.  ie, a 100gb Data.fs will require 200gb of space since it packs it, and creates a Data.fs.old that's otherwise untouched.

[13-May-2010 10:27:34] <LarsN> s/mount/amount

[13-May-2010 10:27:55] <themactech> I will initiate this large copy tonight, then I will try to add the missing zenpacks to my secondary install and try again simlinking the original Data.fs into the new install to see if it loads up.

[13-May-2010 10:28:25] <themactech> space will not be an issue when I get the file off this server

[13-May-2010 10:28:35] <LarsN> themactech: my sr. sysadmin is telling me the z3c.offlinepack should work on Zope2 databases.

[13-May-2010 10:28:45] <zykes-> anyone tried using fortigate on zenoss ?

[13-May-2010 10:28:49] <themactech> cool, i will give it a shot

[13-May-2010 10:28:55] <zykes-> i tried the zenpack but it fucks it up

[13-May-2010 10:29:43] <themactech> BTW I build all my Zenoss servers on Dell poweredge servers with DRAC interfaces, the DRAC interface saved me a trip to new york on this failure, was able to do a bare metal restore remotely with it

[13-May-2010 10:29:45] <mrayzenoss> zykes-: 1.0 or 1.1?

[13-May-2010 10:30:33] <zykes-> 1.1 i think

[13-May-2010 10:30:38] <LarsN> themactech: we use console redirection and some of 1995's greatest technology....  the 48port Xyplex terminal server :)

[13-May-2010 10:30:45] <LarsN> times three

[13-May-2010 10:31:10] <LarsN> how well do those DRAC cards work?

[13-May-2010 10:32:01] <themactech> The DRAC has that, plus you can mount virtual media, I put a systemrescueCD in my MacBook Pro (booted into windows, the DRAC is not very mac friendly), told the DRAC to make the server believe this CD was in the drive and booted my Dell in New York off the rescueCD in my laptop

[13-May-2010 10:32:53] <themactech> The console redirection gives you complete KVM access, you can even get into bios settings at boot

[13-May-2010 10:33:28] <LarsN> we regularly PXE boot to install OSs remotely.

[13-May-2010 10:33:39] <themactech> we do the same on XServes

[13-May-2010 10:33:43] <LarsN> PXE + Console redirection works pretty well in our 98% FreeBSD shop.

[13-May-2010 10:34:03] <themactech> we setup a netboot server and tell the down server via IPMI to boot off of it

[13-May-2010 10:34:55] <LarsN> nod.

[13-May-2010 10:35:14] <themactech> We argue with vendors all the time to include IPMI in their server builds

[13-May-2010 10:35:56] <themactech> Digital rapids doesn't have it, Harris doesn't have it, so we build their services on Dell boxes so we can do remote service

[13-May-2010 10:37:25] <themactech> I am trying the new 3.0 alpha and I can't figure out how to add a network, the new interface has things more condensed but I find it less intuitive.

[13-May-2010 10:37:53] <themactech> I'm in the network pane, and there is no 'add' button, just a lot of fields and a save button

[13-May-2010 10:38:02] <mtzenoss> button at the bottom left

[13-May-2010 10:38:30] <themactech> Oh I assume I only had to fill in the fields and save, but that gives me a pop up error

[13-May-2010 10:38:42] <mtzenoss> It's a "+" below the network listing tree

[13-May-2010 10:38:53] <mtzenoss> yeah, that page is still in progress

[13-May-2010 10:39:49] <themactech> added my network and doing discovery now, the new interface will need getting used to, no big deal

[13-May-2010 10:43:18] <Simon4> mtzenoss: sorry for delayed response - thanks heaps I'll give that a go :)

[13-May-2010 10:43:38] <mtzenoss> np

[13-May-2010 10:44:10] <mrayzenoss> if anyone is checking out the Alpha #4, our QA team is in #zenoss-testing if you have questions/comments

[13-May-2010 10:46:00] <themactech> I am still looking for a way to add new fields to the database, I want to add fields with the same class/behavior than the 'comments' field.

[13-May-2010 10:46:11] <themactech> does mtzenoss know how I can do this?

[13-May-2010 10:49:21] <mtzenoss> What item are you trying to add new fields to?

[13-May-2010 10:50:15] <LarsN> mrayzenoss: I will hopefully be building an alpha 4 test server tonight at home.

[13-May-2010 10:50:24] <mrayzenoss> LarsN: cool

[13-May-2010 10:50:38] <mrayzenoss> we may be entering beta in about 2 weeks

[13-May-2010 10:50:42] <themactech> I am running alpha 4 on ubuntu 10.04 on a VirtualBox image

[13-May-2010 10:50:57] <LarsN> I'll be deploying to FreeBSD 8.0-release-p2

[13-May-2010 10:51:11] <themactech> I want to have a few more fields that can hold customized data in the Status tab

[13-May-2010 10:51:32] <themactech> I have gear where we need to track firmware info, some other gear needs us to track the licensing of the box

[13-May-2010 10:52:11] <themactech> I have figured how to customize the field headers on the status page per device class, I just need a few fields that can hold the custom info I want to display

[13-May-2010 10:52:42] <mtzenoss> Are you talking about the Event console?

[13-May-2010 10:52:48] <themactech> I need those to be accessible from a modeler plugin, that is why I use the 'comments' field as a template since that field does everything I need, just need to clone that field a few times

[13-May-2010 10:53:12] <themactech> no, the device status page

[13-May-2010 10:53:15] <mrayzenoss> themactech: I believe you'll have to extend the model to get that

[13-May-2010 10:53:22] <themactech> I agree

[13-May-2010 10:54:08] <themactech> I was given examples that do this in Zenpacks and will extend model for only a device class, I will need this further down the road, but now I want to extend the DB with these few fields so that it applies to all devices

[13-May-2010 10:54:25] <themactech> I want a Custom1 and Custom2 field that all device classes can push data into

[13-May-2010 10:54:37] <mtzenoss> oh, the device page, let me check

[13-May-2010 10:54:42] <themactech> I will then customize the header for these fields per device class

[13-May-2010 10:54:56] <themactech> I have already posted on how to do the field header customization in the forums

[13-May-2010 10:55:08] <themactech> all i need now is to add the fields that will hold the data

[13-May-2010 10:55:42] <themactech> this process would make it very easy for anyone to customize the device status page for any device class they create

[13-May-2010 10:57:30] <mtzenoss> @themactech add to the _properties of the Device model

[13-May-2010 10:58:10] <themactech> I know how to add zproperties and custom schemas, but I was told these cannot be pushed data to from a modeler

[13-May-2010 10:58:23] <themactech> And my testing seems to confirm that info

[13-May-2010 10:58:27] <mtzenoss> From a ZenPack you can just do Device._properties += {}

[13-May-2010 10:58:36] <mtzenoss> hm

[13-May-2010 10:58:47] <themactech> again its why i mention the 'comments' field, it has all the behavior I am looking for

[13-May-2010 10:59:22] <mtzenoss> comments is just a regular property

[13-May-2010 10:59:23] <themactech> The only things I can do with ZenPacks right now are from the GUI

[13-May-2010 10:59:44] <themactech> alright can you give me a walkthrough on how to make a 'comments' clone?

[13-May-2010 11:00:07] <mtzenoss> Let me see if there's a way to do it from the UI

[13-May-2010 11:00:36] <themactech> that would be preferable but if not, give me the hard liquor, I have to figure this out

[13-May-2010 11:03:00] joseph is now known as Guest23093

[13-May-2010 11:08:22] <mtzenoss> @themactech have you tried using the ZMI?

[13-May-2010 11:08:59] <themactech> ZMI?  have no clue what you are talking about

[13-May-2010 11:09:17] <mtzenoss> http://zenosshost:8080/manage

[13-May-2010 11:10:17] <themactech> I had my alpha virtualbox fired up, let me switch to my 2.5.2 version

[13-May-2010 11:11:20] <mtzenoss> it can be done from the alpha also

[13-May-2010 11:11:32] <mtzenoss> are you trying to do this with the alpha or 2.5?

[13-May-2010 11:12:02] <themactech> tried it in the alpha, asked me to authenticate and my zenoss username and password was rejected

[13-May-2010 11:12:15] <mtzenoss> hm

[13-May-2010 11:12:43] <mrayzenoss> the /manage trick only works off of dmd in the alpha I believe

[13-May-2010 11:13:21] <mtzenoss> really? working fine at the root for me from trunk

[13-May-2010 11:14:28] <mrayzenoss> root meaning zport/dmd/manage

[13-May-2010 11:16:45] <mrayzenoss> themactech: Egor was asking about similar stuff yesterday and was pointed towards hacking on ApplyDataMaps.py

[13-May-2010 11:18:59] <themactech> alright I'm in /zport/dmd/manage, what am I looking for?

[13-May-2010 11:19:43] <mtzenoss> Consulting with the rest of the team...

[13-May-2010 11:19:50] <themactech> Just go to property tab and adding a property?

[13-May-2010 11:20:53] <mtzenoss> yes, you can at least add properties that way, looking for a way for you to get data in there from the modeler

[13-May-2010 11:22:28] <themactech> I now use this syntax:  snmpGetMap = GetMap({

[13-May-2010 11:22:28] <themactech>      '.1.3.6.1.4.1.x.x.x.x.x' : 'setHWTag',

[13-May-2010 11:22:28] <themactech>      '.1.3.6.1.4.1.x.x.x.x.y' : 'comments'

[13-May-2010 11:22:29] <themactech>      })

[13-May-2010 11:27:29] <mtzenoss> It's something Zenoss doesn't support natively. Have you asked about it on the forums? I've asked the rest of the team for input.

[13-May-2010 11:28:24] <themactech> I was given the lead of checking some zenpacks that add this for specific device classes within the zenpack, but I am looking for something simpler right now

[13-May-2010 11:29:06] <themactech> to me this would be a simple database issue, there is a field in a database called 'comments' I want to duplicate that field, this does not seem to me like a complex database command

[13-May-2010 11:29:51] <themactech> all the leads that I get to investigate this seem to point to rather complex solutions to a simple problem

[13-May-2010 11:30:57] <mtzenoss> What all do you want to duplicate about comments? Adding a property is not a problem, duplicating the functionality around it is a bigger task

[13-May-2010 11:31:47] <themactech> I need a field I can push data to from a modeler plugin, and that I can display the data in the device status tab

[13-May-2010 11:32:02] <themactech> I was told zproperties cannot be addressed in modelers

[13-May-2010 11:32:09] <themactech> and in fact I tried to and could not

[13-May-2010 11:32:39] <themactech> what is the difference between a zproperty and the comments field, I do not see or understand the distinction between those two kinds of fields

[13-May-2010 11:33:04] <themactech> I need the behavior of the comments field, but the only thing I can add easily is a zproperty

[13-May-2010 11:33:58] <mtzenoss> In zenmodeler you don't have a direct connection to the ZODB, so you don't have access to a real-live Zope object -- just a shadow

[13-May-2010 11:34:31] <mrayzenoss> looks like the comments is defined in Device.py

[13-May-2010 11:34:43] <mtzenoss> The data that fills out the 'device' object is whatever gets pumped out from a zenhub service.

[13-May-2010 11:34:47] <mtzenoss> So since you don't have the real device object, only a sanitized and small represenation, you are necessarily limitied in what you can do.

[13-May-2010 11:35:02] <themactech> how do you determine which fields you can address within the modeler then, is there and intermediate process that acts as a directory of some sorts

[13-May-2010 11:36:18] <mtzenoss> DeviceProxy, which can be monkey-patched, but not an easy task

[13-May-2010 11:38:11] <themactech> Device.py has the following call:

[13-May-2010 11:38:20] <themactech> device.manage_editDevice(

[13-May-2010 11:38:38] <mtzenoss> Check out the DataCollector/plugins/CollectorPlugin.py file for an example of what will be available for the modeler to *read*

[13-May-2010 11:38:59] <themactech> tag, serialnumber, zsnmpcommunity, zsnmpport, comment, hwmanufacturer, etc

[13-May-2010 11:39:08] <themactech> is this the dictionary that modelers can access

[13-May-2010 11:39:22] <themactech> can it be expanded in Device.py?

[13-May-2010 11:40:11] <mtzenoss> DeviceProxy is

[13-May-2010 11:43:53] <mtzenoss> A developer suggests "If they are super, super interested in getting a modeler plugin to drop junk onto a device, then they would probably have to craft an event and then create an event transform that mangles the device based on the event data."

[13-May-2010 11:47:03] <themactech> I already have modelers pushing 'junk' to the device status page, just that right now I hijack the comments field and the service tag # field, You guys should make a few custom fields available to users, especially since it is really easy to make a custom header for that field that will adapt its value per device class.  So the same Custom1 field will show "firmware version" for a device class that needs  that info,

[13-May-2010 11:47:03] <themactech> "licensing" for another device class that will need that info, all the while showing the value in field Custom1

[13-May-2010 11:47:57] <themactech> SNMP monitoring applies to such a broad spectrum of devices that I don't see how you expect no1 to try and deviate from the data type you include by default in the device status page.

[13-May-2010 11:48:24] <themactech> I have seen very frequent requests in the forums from folks who want to display other info in the device status page

[13-May-2010 11:49:14] <themactech> if you put a few more 'junk' fields in there, it would make it REALLY easy for people to do so

[13-May-2010 11:49:50] <themactech> But from what I get from the forums, all the folks who want to do this are faced with the complexity of doing brain surgery while wearing boxing gloves

[13-May-2010 11:50:12] <themactech> If the process of adding a few field for custom user use is very complex then please add a couple from the get go

[13-May-2010 11:50:50] <themactech> you don't even need to show them by default, if some1 wants to use them he can mod the device status page skin, and that is very easy

[13-May-2010 11:51:30] <zykes-> wondering, does zenoss support BNT switches ?

[13-May-2010 11:51:52] <zykes-> and compellent san

[13-May-2010 12:14:42] <mistich> good morning

[13-May-2010 12:14:51] <gordy> Will zenoss integrate with SCADA device protocols ?

[13-May-2010 12:15:56] <mistich> how can you test a groups alerts without creating a event

[13-May-2010 12:34:44] <zykes-> mrayzenoss: tested fortigate 1.1 with zenoss 2.5.2 ?

[13-May-2010 12:35:23] <mrayzenoss> not personally, I was just wondering which version was causing you trouble

[13-May-2010 12:36:12] <mrayzenoss> gordy: I haven't seen that, but I'm sure something could be written, since lots of other protocols have been added

[13-May-2010 12:37:35] <mrayzenoss> zykes-: I haven't seen BNT or Compellent ZenPacks, but I assume they do SNMP so they can probably be monitored

[13-May-2010 12:38:11] <zykes-> bnt is nice same with compellent

[13-May-2010 12:38:49] <cgibbons> ouf

[13-May-2010 12:43:43] <willwh> hey guys - trying to install the JMX Zenpack

[13-May-2010 12:44:09] <willwh> do I just use the zenpack tab & upload "ZenPacks.zenoss.ZenJMX-3.3.2-py2.4.egg" ?

[13-May-2010 12:44:13] <willwh> it's erroring out for me.

[13-May-2010 12:44:29] <willwh> sec I'll have an error.. :)

[13-May-2010 12:45:10] <willwh> ERROR: zenpack command failed. Reason: SystemExit: usage: zenpack.py [options] requirement_or_url ...

[13-May-2010 12:45:37] <willwh> error: invalid command 'easy_install' :)

[13-May-2010 12:45:56] <mrayzenoss> which installer?

[13-May-2010 12:46:08] <mrayzenoss> are you doing that as root?

[13-May-2010 12:59:27] <willwh> well, just using the "install zenpack" in settings

[13-May-2010 12:59:39] <willwh> and it's a hosted zenoss install.... so whichever user zenoss is running as I guess.

[13-May-2010 13:01:54] <mrayzenoss> weird

[13-May-2010 13:02:03] <mrayzenoss> who's hosting it?

[13-May-2010 13:13:40] <willwh> mrayzenoss: need to wait on my CTO

[13-May-2010 13:14:15] <willwh> looks like it's on the amazon cloud though

[13-May-2010 13:14:20] <willwh> quick reverse lookup ;)

[13-May-2010 13:14:29] <willwh> once he's alive I'll ask and let you know

[13-May-2010 13:14:34] <willwh> it's possible they'll just install for me anyway

[13-May-2010 13:14:35] <willwh> :)

[13-May-2010 13:14:57] <willwh> JMX looks really nice though - we run a TON of wowza media servers

[13-May-2010 13:15:49] <willwh> so it's going to be fantastic to get my fingers dirty monitoring those :)

[13-May-2010 13:15:56] <willwh> even just heap memory usage is going to be gold

[13-May-2010 13:19:09] <zykes-> JMX is ?

[13-May-2010 13:35:18] <willwh> http://java.sun.com/javase/technologies/core/mntr-mgmt/javamanagement/

[13-May-2010 14:02:52] LarsN is now known as LarsN_Away

[13-May-2010 15:04:56] <zykes-> anyone here that's activated snmp on a fortigate before ?

[13-May-2010 19:49:40] <forsberg> zykes- yes

[13-May-2010 19:50:01] <forsberg> to the fortigate question

[13-May-2010 20:44:40] forsberg is now known as fOrsberg

[13-May-2010 21:51:36] <georgesweeney> arriving

[13-May-2010 22:47:38] <imcdona> anyone know how to get the zenoss vmware image to work on esxi?

[14-May-2010 00:00:30] [disconnected at Fri May 14 00:00:30 2010]

[14-May-2010 00:00:31] [connected at Fri May 14 00:00:31 2010]

[14-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[14-May-2010 01:48:07] <zykes-> fOrsberg: got time to help a bit ?

[14-May-2010 03:26:05] <Simon4_> guessing most people are asleep, but has anyone considered/done putting rrdfiles and zenrender on a separate "reporting" server to spread load away from the collectors?

[14-May-2010 04:43:36] fOrsberg is now known as forsberg

[14-May-2010 05:41:24] <chemist> Simon4: no, but it would be a good idea

[14-May-2010 05:42:32] <Simon4> chemist: yeah, the idea is growing on me fairly fast

[14-May-2010 06:23:35] <chemist> Simon4: let me know if you progress it

[14-May-2010 07:59:28] <smbambling> morning

[14-May-2010 07:59:43] <mrayzenoss> morning

[14-May-2010 08:00:49] <smbambling> mrayszenoss are you up in Annapolis ?

[14-May-2010 08:02:06] <mrayzenoss> nope, Austin

[14-May-2010 08:05:31] <coofamani> morning

[14-May-2010 08:25:33] LarsN_Away is now known as LarsN

[14-May-2010 08:26:39] <LarsN> If I have the "device" template set at the /devices/servers/ level

[14-May-2010 08:27:04] <LarsN> how would I go about creating a copy of that, with different parameters at the /devices/servers/freebsd/jails level?

[14-May-2010 08:27:32] <LarsN> without clicking on "create local copy" on each device.

[14-May-2010 08:29:01] <mrayzenoss> LarsN: go to /Server/Templates and select the Device template

[14-May-2010 08:29:49] <mrayzenoss> select Copy Templates and copy it over to /Servers/FreeBSD/Jails

[14-May-2010 08:30:03] <mrayzenoss> go over to the Jails class, make your modifications

[14-May-2010 08:30:06] <mrayzenoss> bind the template

[14-May-2010 08:30:13] <mrayzenoss> remodel any devices in Jails

[14-May-2010 08:30:18] <LarsN> Brilliant!

[14-May-2010 08:30:27] <zykes-> do you have any training in europe ?

[14-May-2010 08:30:34] <LarsN> thanks mrayzenoss, that's one of the tasks I've been meaning to figure out for a long time now.

[14-May-2010 08:30:50] <mrayzenoss> zykes-: Jane Curry offers training through Skills 1st, I've heard really good things about that

[14-May-2010 08:31:00] <mrayzenoss> zykes-: we do offer remote web-based training too

[14-May-2010 08:31:06] <zykes-> any certifications

[14-May-2010 08:31:07] <zykes-> `?

[14-May-2010 08:31:14] <mrayzenoss> not yet

[14-May-2010 08:31:36] <zykes-> too bad :d

[14-May-2010 08:31:41] <mrayzenoss> right now it's mostly based on exhibited reputation

[14-May-2010 08:31:55] <mrayzenoss> like, people showing up in the forums and IRC and showing they know what they're talking about

[14-May-2010 08:32:09] <mrayzenoss> Jane for example, knows more than a lot of Zenoss employees

[14-May-2010 08:32:10] * LarsN doesn't have to worry about that.  :)

[14-May-2010 08:32:17] <mrayzenoss> so I feel comfortable referring people to her

[14-May-2010 08:32:17] <zykes-> ;)

[14-May-2010 08:32:25] <LarsN> the best thing about me is.

[14-May-2010 08:32:33] * LarsN is almost completely stupid.

[14-May-2010 08:32:35] <mrayzenoss> heh, just noticed Jane_Curry is in the channel

[14-May-2010 08:33:00] <LarsN> I'm looking forward to Zenoss 3.

[14-May-2010 08:33:09] <mrayzenoss> me too

[14-May-2010 08:33:21] <mrayzenoss> lot of good developments besides the new UI

[14-May-2010 08:33:23] <LarsN> no more unsupported python 2.4 with it's awful memory management.

[14-May-2010 08:33:25] <mrayzenoss> Python 2.6

[14-May-2010 08:33:31] <mrayzenoss> Zope 2.12

[14-May-2010 08:33:38] <mrayzenoss> documented developer API

[14-May-2010 08:33:39] <LarsN> I work for a plone shop.

[14-May-2010 08:33:54] <LarsN> we're looking forward to Plone 4 which also moves to python 2.6 and zope 2.12

[14-May-2010 08:34:15] <mrayzenoss> LarsN: did you see this? http://au.last-bastion.net/software/plone4zenoss

[14-May-2010 08:34:44] <LarsN> mrayzenoss: I hadn't....  that's pretty slick :)

[14-May-2010 08:35:46] <LarsN> thanks for the tip.

[14-May-2010 08:43:55] <zykes-> hmm, does zenoss support sms ?

[14-May-2010 08:44:31] <mrayzenoss> docs/DOC-5819

[14-May-2010 08:44:52] <mrayzenoss> lots of different ways to implement SMS, that's one technique

[14-May-2010 08:52:19] <LarsN> we purchased an SMS gateway. :)

[14-May-2010 08:52:38] <LarsN> it's got 4 cellular radios in it.  Currently we have 2 sim cards installed.

[14-May-2010 08:52:52] <LarsN> each radio can send up to 10sms's a minute.

[14-May-2010 08:53:13] <mrayzenoss> cool

[14-May-2010 08:53:31] <LarsN> it's a lot more reliable than using, 111-222-3333@tmobile.net

[14-May-2010 08:54:02] <mrayzenoss> I know one of our customers went to the AT&T store, bought a cheap cellphone supported by Linux, got an unlimited SMS plan and duct-taped it to the back of their rack :)

[14-May-2010 08:54:04] <LarsN> or whatever other e-mail gateway/sms thingy the various providers give you.  some of them start delaying delivery if you get a "lot" sent.

[14-May-2010 08:54:11] <LarsN> mrayzenoss: NICE!

[14-May-2010 08:54:23] <mrayzenoss> enterprise IT :p

[14-May-2010 08:55:26] <LarsN> the gateway works bidirectionally which is also pretty slick.

[14-May-2010 08:55:44] <LarsN> we can send an SMS to the gateway, and it'll "do things" based on who sent what message.

[14-May-2010 09:29:32] <coofamani> LarsN: yeah, I've seen throttling as bad as 3 -4 hours at times

[14-May-2010 09:30:32] <LarsN> coofamani: that's what prompted us to purchase the SMS gateway

[14-May-2010 09:30:56] <coofamani> who did you go with?

[14-May-2010 09:31:19] <LarsN> I think the device is a Microway....  with two TMobile sims

[14-May-2010 09:31:27] <LarsN> I'll confirm the device quick.

[14-May-2010 09:32:18] <LarsN> MultiTech SF400-G

[14-May-2010 09:32:21] <LarsN> is the device.

[14-May-2010 09:32:25] <coofamani> then you just have to deal with the fact that all the engineers cell phones go belly up when there is a big issue.

[14-May-2010 09:32:35] <LarsN> yeah.

[14-May-2010 09:32:49] <LarsN> coofamani: we actually send SMSs to engineers, as well as e-mails.

[14-May-2010 09:32:55] <coofamani> same

[14-May-2010 09:33:01] <LarsN> coofamani: I have a second e-mail address, that's an alias.

[14-May-2010 09:33:21] <LarsN> it copies the message to my inbox, but also causes our asterisk server to call my phone.

[14-May-2010 09:33:56] <LarsN> the asterisk server calls every 60 seconds for 10 minutes, or until the engineer answers and presses a key on the keypad

[14-May-2010 09:34:00] <coofamani> hmm, I swore I had a trac ticket requesting event summarization

[14-May-2010 09:34:02] <coofamani> cant find it

[14-May-2010 09:34:25] <LarsN> the oncall, and backup oncall engineers get the phone notifications above and beyond SMS and E-Mail.

[14-May-2010 09:34:50] <LarsN> I have a hate, love, hate relationship with Zenoss :)

[14-May-2010 09:35:26] <coofamani> Im watching their sales team get beat up in the office across the hall

[14-May-2010 09:35:46] <coofamani> I think mrayzenoss is going to be washing my car when this is over.

[14-May-2010 09:35:52] <coofamani> sorry matt

[14-May-2010 09:40:15] <LarsN> coofamani: how well do you know zope?

[14-May-2010 09:40:27] <coofamani> meh

[14-May-2010 09:40:31] <LarsN> heh.

[14-May-2010 09:40:39] <coofamani> just what I'ved learned through using Zenoss

[14-May-2010 09:41:02] <LarsN> I'd love to see if I could set zenoss up to use, a zeocluster.

[14-May-2010 09:41:11] <coofamani> I think you can

[14-May-2010 09:41:14] <Jane_Curry> zykes- do you want to ask questions directly about training / consultancy in Europe?

[14-May-2010 09:41:18] <coofamani> thats been on my extended to-play-with list

[14-May-2010 09:41:20] <Jane_Curry> We are UK based

[14-May-2010 09:41:29] <LarsN> where the zodb is separated from the zope application server.  And then see if I can insert relstorage

[14-May-2010 09:41:39] <LarsN> so the zodb gets put into mysql, or psql.

[14-May-2010 09:41:50] <LarsN> so real sysadmin tools can be used to back it up,

[14-May-2010 09:41:59] <LarsN> or replicate it across the network.

[14-May-2010 09:42:06] <Simon4> LarsN: crazy talk!

[14-May-2010 09:42:08] <Simon4> :)

[14-May-2010 09:42:27] * LarsN is going to be giving a talk on relstorage at the Plone Symposium East later this month.

[14-May-2010 09:42:38] * Simon4 has a read about ti

[14-May-2010 09:42:41] <Jane_Curry> mrayzenoss - is anyone awake for input on the zenoss-testing channel (like bbibeault1)?? got a bit more input on the alpha

[14-May-2010 09:42:59] <cgibbons> relstorage seemed pretty hacky; does it actually work well?

[14-May-2010 09:43:11] <mrayzenoss> I'll go bug him

[14-May-2010 09:43:23] <LarsN> cgibbons: it seems to work pretty well for the projects we're using it on currently.

[14-May-2010 09:43:41] <LarsN> cgibbons: in both cases we went that route so we could cluster database servers for higher availability.

[14-May-2010 09:44:16] <LarsN> we're running FreeBSD, mysql is bound to a CARP address, shared across the two servers.  And then using mysql replication between the two actual mysql instances

[14-May-2010 09:44:50] <LarsN> the zeo's (12 of them across two servers) talk to mysql on CARP.  and some ifstated magic takes care of promoting a slave to master as necessary.

[14-May-2010 09:45:23] <LarsN> 1.4 relstorage has a lot of improvements added in part because of the projects we're working on.

[14-May-2010 09:46:06] <Jane_Curry> Thanks Matt

[14-May-2010 09:47:46] <LarsN> cgibbons: once I get my Zenoss 3 test instance up and running

[14-May-2010 09:48:02] <LarsN> I'm going to look into inserting RelStorage and see how well it works, or fails.

[14-May-2010 09:48:03] <LarsN> :)

[14-May-2010 09:48:38] <cgibbons> chet here has done the work to use it already. we haven't reviewed it yet or put it into the product (priorities) but there should be no real issues.

[14-May-2010 09:49:05] <LarsN> cgibbons: the other thing I've been meaning to try

[14-May-2010 09:49:17] <LarsN> is inserting the cpickle bits from the unladen swallow project.

[14-May-2010 09:49:27] <LarsN> it's "lots" faster than the default cpickle

[14-May-2010 09:49:38] <LarsN> and is supposed to "drop in"

[14-May-2010 09:49:46] <LarsN> according to someone working on the project.

[14-May-2010 09:51:23] forsberg is now known as fOrsberg

[14-May-2010 09:59:41] <mrayzenoss> QA is back in #zenoss-testing if you're playing with the alpha

[14-May-2010 10:06:29] <rmatte> I hope they plan to change that start/stop icon on the graphs page

[14-May-2010 10:09:00] <mrayzenoss> rmatte: what where?

[14-May-2010 10:28:53] <rmatte> on the graph page in the new UI

[14-May-2010 10:29:08] <rmatte> next to the input field where you can set auto-refresh time

[14-May-2010 10:29:12] <rmatte> it's using the old start/stop icon

[14-May-2010 10:29:18] <rmatte> and it's very ugly

[14-May-2010 10:30:06] <rmatte> It sticks out like a transvestite in a bikini competition

[14-May-2010 10:30:13] <rmatte> :P

[14-May-2010 10:41:13] <rmatte> I'm loving what they are doing with the device page

[14-May-2010 10:41:27] <rmatte> can't wait to see the final result

[14-May-2010 10:43:41] <mrayzenoss> you should open a ticket with that line in there… "It sticks out like a transvestite in a bikini competition"

[14-May-2010 10:49:55] <kobalt> so if I have 2.5.1 install with the zenpacks I need and I want to upgrade to 2.5.2 I just do the upgrade as listed in the documents, is there a need to do the zenpack update, they have not changed from 2.5.1. to 2.5.2

[14-May-2010 10:50:41] <mrayzenoss> kobalt: they should be fine

[14-May-2010 10:50:57] <mrayzenoss> kobalt: as always, make backups before upgrading

[14-May-2010 10:51:04] <kobalt> yeah doing that now

[14-May-2010 11:40:55] <mrayzenoss> rmatte: I posted your comment earlier in the dev chat:  that the old UI wrapped in the new UI "sticks out like a transvestite in a bikini competition"

[14-May-2010 11:41:08] <mrayzenoss> response: So they want more or less of it?

[14-May-2010 11:42:05] <mrayzenoss> I could hear people laughing down the hall

[14-May-2010 11:42:33] <rmatte> haha

[14-May-2010 11:42:50] <rmatte> nice

[14-May-2010 12:28:41] <rmatte> and to respond, preferably less of the old :P

[14-May-2010 12:46:13] <Apachez> atlantis: T minus 9 minutes and holding...

[14-May-2010 12:57:13] fOrsberg is now known as forsberg

[14-May-2010 13:14:50] LarsN is now known as LarsN_Away

[14-May-2010 13:27:03] <rmatte> ok great, so I have a perf directory that is not being created

[14-May-2010 13:30:42] <rmatte> ah I see why

[14-May-2010 13:46:48] <forsberg> ;)

[14-May-2010 14:06:21] forsberg is now known as fOrsberg

[14-May-2010 14:14:41] <bigegor> zWbem

[14-May-2010 14:24:28] <zykes-> zwbem is for what ?

[14-May-2010 14:29:45] <bigegor> Nothing. Wrong input.

[14-May-2010 14:30:13] <rmatte> hehe

[14-May-2010 14:34:05] <twm1010> So are zenoss users referred to as zenites or zenossians?

[14-May-2010 14:34:27] <rmatte> not really

[14-May-2010 14:34:35] <rmatte> "zenoss users" is fine

[14-May-2010 14:34:36] <rmatte> lol

[14-May-2010 14:36:46] <rmatte> though there is reference to "Zenossian" here: blogs/zenossblog/tags/zapplet

[14-May-2010 14:40:30] <twm1010> you can tell it's friday :)

[14-May-2010 14:41:12] <rmatte> yeh

[14-May-2010 14:41:17] <rmatte> It's also my birthday

[14-May-2010 14:41:22] <rmatte> so I'm even less inclined to do work

[14-May-2010 14:41:27] <rmatte> :)

[14-May-2010 14:44:29] <coofamani> zuzers

[14-May-2010 14:46:27] <rmatte> hehe

[14-May-2010 14:46:52] <rmatte> zuperusers

[14-May-2010 14:49:23] <cgibbons> i think zenossian is usually intended to mean an employe

[14-May-2010 14:49:35] <rmatte> that's what I figured since it was referring to Nate

[14-May-2010 14:49:46] <twm1010> so we're zenites or zenots ?

[14-May-2010 14:49:49] <twm1010> :)

[14-May-2010 14:50:02] <coofamani> zintrels?

[14-May-2010 14:50:36] <rmatte> zentinels

[14-May-2010 14:50:40] <rmatte> :P

[14-May-2010 14:50:54] <coofamani> wow we are bored.

[14-May-2010 14:51:02] <coofamani> it has been a looong week

[14-May-2010 14:51:13] <rmatte> yup

[14-May-2010 14:51:24] <coofamani> you get all your 2.5.2s done?

[14-May-2010 14:51:30] <rmatte> yup, finally all done

[14-May-2010 14:51:45] <rmatte> now I've been informed that I need to completely rebuild our Zenoss VM template image

[14-May-2010 14:51:56] <coofamani> any aftermath? we still need work on zenhub on the larger distrib installs... its finicky

[14-May-2010 14:51:58] <rmatte> since we can't get it to upgrade to the latest version of the OS (it's quite broken)

[14-May-2010 14:52:09] <coofamani> whats os are you runnign again?

[14-May-2010 14:52:10] <rmatte> I've noticed that performance isn't quite as great

[14-May-2010 14:52:18] <rmatte> despite less memory usage

[14-May-2010 14:52:21] <rmatte> other than that, not really

[14-May-2010 14:53:03] <coofamani> watch the per-worker ram usage on zenhub if you are monitoring big networks. each worker will build a ping tree when/if zenping hits it, but it won't release the ram

[14-May-2010 14:53:18] <coofamani> I had them climbing to 5GB each (6 workers) over time

[14-May-2010 14:53:32] <rmatte> nah, they are actually all using less RAM than they were with 2.4.5

[14-May-2010 14:53:37] <rmatte> but the overall performance is worse

[14-May-2010 14:53:43] <rmatte> UI performance

[14-May-2010 14:53:54] <rmatte> especially with the new event console

[14-May-2010 14:53:54] <coofamani> heh, that improved for me, bug time

[14-May-2010 14:54:00] <rmatte> loads times are insane sometimes

[14-May-2010 14:54:03] <rmatte> load*

[14-May-2010 14:54:06] <coofamani>  /big/ tie I mean

[14-May-2010 14:54:08] <coofamani> time

[14-May-2010 14:54:14] <coofamani> damnit. hands have gone home for the weekend

[14-May-2010 14:54:18] <rmatte> lol

[14-May-2010 14:55:24] <rmatte> the os we're running on the vm template is Ubuntu Server 8.04

[14-May-2010 14:55:34] <rmatte> I tried to upgrade it to 10.04 but it won't do it

[14-May-2010 14:55:42] <rmatte> tons of issues

[14-May-2010 14:55:54] <rmatte> I didn't do the original install, someone else did

[14-May-2010 14:56:04] <rmatte> so I'm actually anxious to do it myself from a clean slate

[14-May-2010 14:56:32] <rmatte> I know I won't screw anything up :)

[14-May-2010 14:58:56] <rmatte> speaking of which, time to upgrade my workstation to 10.04...

[14-May-2010 15:13:26] <rmatte> wow, I'm so glad that the ubuntu servers are lightning fast

[14-May-2010 15:13:33] <twm1010> :)

[14-May-2010 15:13:36] <rmatte> files are coming down at a whopping 4,051B/s

[14-May-2010 15:13:40] <twm1010> nice

[14-May-2010 15:13:47] <rmatte> oh now, down to 1,800B/s

[14-May-2010 15:25:13] <zykes-> linux monitoring on zenoss - does it get os processes via snmp ?

[14-May-2010 15:25:37] <davetoo> usually

[14-May-2010 15:25:50] <zykes-> hmmm k

[14-May-2010 15:25:54] <twm1010> brb

[14-May-2010 15:26:04] <davetoo> "usually" = if you haven't set it up for ssh monitoring instead

[14-May-2010 15:26:04] <zykes-> a anything special i need to have ?

[14-May-2010 15:26:30] <zykes-> what's the better thing to use ?

[14-May-2010 15:26:41] <rmatte> I use snmp personally

[14-May-2010 15:26:47] <davetoo> These days the default snmp.conf is usually locked down and only provides system info, so you have to open it up

[14-May-2010 15:27:23] <davetoo> All my systems have been internal, so I just replace snmp.conf with the one-liner "rocommunity <somestring>"

[14-May-2010 15:28:02] <zykes-> public community ?

[14-May-2010 15:28:18] <rmatte> Here's an snmp config you can use: http://pastebin.com/x8mjtbzs

[14-May-2010 15:28:31] <rmatte> replace the trapsink IPs, and change your-key to the snmp string you want to use

[14-May-2010 15:28:38] <rmatte> also set the syscontact and syslocation info

[14-May-2010 15:29:02] <rmatte> The trapsink IPs should be the IP of your Zenoss server (that is, if you want to receive SNMP traps)

[14-May-2010 15:29:05] <zykes-> my key ?

[14-May-2010 15:29:24] <rmatte> it literally says "your-key" in the config file that I pasted

[14-May-2010 15:29:48] <zykes-> what's that used for ?

[14-May-2010 15:29:50] <rmatte> I'm saying find any instance of "your-key" in that config file and change it to the snmp string that you want to use

[14-May-2010 15:30:03] <rmatte> for authentication when polling for snmp

[14-May-2010 15:30:08] <rmatte> default would be "public"

[14-May-2010 15:30:16] <zykes-> ah

[14-May-2010 15:30:37] <rmatte> make sure you set the key in Zenoss via zProperties

[14-May-2010 15:30:48] <rmatte> you should really stay away from using "public", always change it to something else

[14-May-2010 15:31:47] <zykes-> how come ?

[14-May-2010 15:31:53] <rmatte> for security reasons

[14-May-2010 15:31:59] <rmatte> everyone knows "public"

[14-May-2010 15:32:13] <rmatte> so if you leave it at that and someone gets on to your network they can find out anything about any of your devices

[14-May-2010 15:32:49] <rmatte> so change it to something else

[14-May-2010 15:33:00] <rmatte> unless you don't care about security, then just set it to public :P

[14-May-2010 15:40:26] <Rolfs> any reason why zenoss doesn't include processmonitoring over SSH ?

[14-May-2010 15:41:02] <rmatte> Rolfs: are you certain that it doesn't?

[14-May-2010 15:41:08] <rmatte> I was under the impression that it did

[14-May-2010 15:41:23] <Rolfs> the documentation i saw.. was that it did only over snmp

[14-May-2010 15:41:50] <Rolfs> but i think it's not that hard to write a parser/addon for it to do processmonitoring over ssh

[14-May-2010 15:42:25] <rmatte> probably not

[14-May-2010 15:43:05] <davetoo> Maybe it was only Enterprise, can't remember

[14-May-2010 15:43:09] <rmatte> you're against using snmp or something?

[14-May-2010 15:43:20] <rmatte> enterprise has the same Linux monitoring as Core

[14-May-2010 15:44:30] <mrayzenoss> I think the OpenSolaris ZenPack does process monitoring

[14-May-2010 15:44:40] <mrayzenoss> docs/DOC-3414

[14-May-2010 15:44:47] <mrayzenoss> and http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.OpenSolaris/ZenPacks/community/OpenSolaris/modeler/plugins/zenoss/cmd/solaris/process.py

[14-May-2010 15:44:57] <mrayzenoss> it's on my todo list for the OS X one

[14-May-2010 15:45:21] <Rolfs> mrayzenoss: Ok.. that might be something for me too look at :)

[14-May-2010 15:45:24] <zykes-> does the file you posted rmatte use public for anyting?

[14-May-2010 15:45:25] <rmatte> mrayzenoss: I thought process monitoring was already supported for Linux via ssh, no?

[14-May-2010 15:45:36] <mrayzenoss> I think it is, I just haven't used it

[14-May-2010 15:45:46] <rmatte> zykes-: no, the file I posted has "your-key" as the snmp string

[14-May-2010 15:45:52] <rmatte> you need to change that to what you want it to be

[14-May-2010 15:46:16] <zykes-> com2sec then is used for ?

[14-May-2010 15:46:25] <rmatte> ignore everything at the bottom

[14-May-2010 15:46:34] <rmatte> the setting you want to change are in the top half

[14-May-2010 15:46:34] <zykes-> how come ?

[14-May-2010 15:46:55] <rmatte> because, the settings at the bottom need to stay as they are so that you can properly monitor with Zenoss

[14-May-2010 15:46:59] <rmatte> just trust me on this

[14-May-2010 15:47:05] <rmatte> change what I told you to change earlier

[14-May-2010 15:47:08] <rmatte> and you'll be good

[14-May-2010 15:47:37] <zykes-> and i need to change the community strings in zenoss as well then i guess from public to my own key ?

[14-May-2010 15:47:43] <rmatte> correct

[14-May-2010 15:47:49] <rmatte> under zProperties

[14-May-2010 15:48:08] <rmatte> you can either do it for everything, for a specific class, or for a specific device

[14-May-2010 15:48:12] <rmatte> depending on what level you set it at

[14-May-2010 15:49:01] <mrayzenoss> hmm… it looks like SSH process monitoring for Linux is done with the zenplugin, which is deprecated

[14-May-2010 15:49:16] <mrayzenoss> but nobody bothered to rewrite it

[14-May-2010 15:49:20] <mrayzenoss> (popular feature)

[14-May-2010 15:49:29] <Rolfs> ok

[14-May-2010 15:49:58] <rmatte> yeh, would be nice to get that rewritten

[14-May-2010 15:50:20] <mrayzenoss> I figure I'll look at it when I do it in my OS X zenpack

[14-May-2010 15:50:26] <rmatte> cool

[14-May-2010 15:50:35] <Rolfs> ;)

[14-May-2010 15:52:43] <mrayzenoss> looks like there is a community SSH process monitor with OpenSolaris, but no Core examples

[14-May-2010 15:52:53] <mrayzenoss> and the Enterprise Linux one doesn't have it either

[14-May-2010 15:52:57] * mrayzenoss rolls eyes

[14-May-2010 15:55:04] <Rolfs> I think I have to upgrade to 3.0 version soon to keep track with Zenoss rollouts.. I also think we need to get some faster disks for the RRD files ;)

[14-May-2010 15:56:02] <zykes-> zSnmpCommunity < is that the thing i should change rmatte ?

[14-May-2010 16:03:20] <rmatte> Rolfs: 3.0 of what?

[14-May-2010 16:03:27] <rmatte> zykes-: yes

[14-May-2010 16:04:30] <zykes-> rmatte: should i need to add some MIB files for linux or ?

[14-May-2010 16:04:50] <rmatte> zykes-: no, pretty sure it already translates those

[14-May-2010 16:05:00] <zykes-> Unable to get data for 10.10.1.15 from either ipAddrTable or ipNetToMediaTable

[14-May-2010 16:05:11] <rmatte> and you're using my config file?

[14-May-2010 16:05:23] <zykes-> yes

[14-May-2010 16:05:26] <rmatte> that's odd

[14-May-2010 16:05:45] <zykes-> indeed

[14-May-2010 16:05:48] <rmatte> I have a bug ticket open for that error since it's not handled as well as it should be

[14-May-2010 16:05:53] <rmatte> one second...

[14-May-2010 16:06:06] <Rolfs> 3.0 of zenoss.. when it gets stable ;)

[14-May-2010 16:06:32] <rmatte> http://dev.zenoss.org/trac/ticket/5691

[14-May-2010 16:06:43] <rmatte> I listed a workaround for that error in that ticket

[14-May-2010 16:06:45] <davetoo> I'm looking forward to being able to write for python 2.6

[14-May-2010 16:06:55] <rmatte> just modify the file as described, restart zenoss, and you'll be able to model the device

[14-May-2010 16:07:00] <rmatte> I hope they fix that soon

[14-May-2010 16:07:55] <zykes-> so this applies for linux boxes as well ?

[14-May-2010 16:08:01] <rmatte> yes

[14-May-2010 16:08:11] <rmatte> in certain cases

[14-May-2010 16:08:18] <rmatte> it happens on one of my Linux boxes

[14-May-2010 16:08:28] <rmatte> wait

[14-May-2010 16:08:38] <rmatte> when you added the config, did you restart the snmpd daemon?

[14-May-2010 16:08:47] <zykes-> yes

[14-May-2010 16:08:50] <rmatte> k

[14-May-2010 16:17:24] <zykes-> hmms, didn't help that much :/

[14-May-2010 16:17:43] <zykes-> Changed /opt/zenoss/Products/DataCollector/plugins/zenoss/snmp/InterfaceMap.py

[14-May-2010 16:17:48] <zykes-> commented out the line

[14-May-2010 16:18:18] <rmatte> you'll still see the same message when you do but the modelling will complete

[14-May-2010 16:18:26] <rmatte> make sure you restart Zenoss after modifying that

[14-May-2010 16:19:47] <zykes-> why hasn't it been fixed i wonder..

[14-May-2010 16:19:58] <rmatte> because they only have so many developers

[14-May-2010 16:20:06] <rmatte> the ticket is open, they'll get to it when they can

[14-May-2010 16:20:26] <rmatte> They are rewriting the whole UI for the next version, very very big job

[14-May-2010 16:20:33] <rmatte> they have 5 out of 7 developers just working on that

[14-May-2010 16:21:29] <zykes-> :p

[14-May-2010 16:21:35] <rmatte> once that's done then they can get back to fixing bugs

[14-May-2010 16:22:05] <zykes-> something is a mess

[14-May-2010 16:22:23] <zykes-> now after modeling a device it doesn't give me anything, like no cpu no nothing

[14-May-2010 16:24:29] mrayzenoss1 is now known as mattray

[14-May-2010 16:25:31] <zykes-> 2010-05-15 07:24:17,155 WARNING zen.zenperfsnmp: Error reading value for "ssCpuIdle" on xx (oid .1.3.6.1.4.1.2021.11.11.0 is bad)

[14-May-2010 16:26:26] <rmatte> hmmm

[14-May-2010 16:26:32] <rmatte> something is still messed up then

[14-May-2010 16:26:48] <rmatte> look around online for different config files and try them

[14-May-2010 16:26:51] <zykes-> snmp related or mib's on my install ?

[14-May-2010 16:27:03] <rmatte> snmp related

[14-May-2010 16:27:07] <rmatte> has nothing to do with Mibs

[14-May-2010 16:27:13] <zykes-> k

[14-May-2010 16:27:14] <rmatte> permissions issue

[14-May-2010 16:27:40] <rmatte> I'm out in 5 minuts, but I have a Linux box acting the same way with that config, even though the same config is working fine on a bunch of others

[14-May-2010 16:27:46] <rmatte> if I figure out why I'll let you know

[14-May-2010 16:30:17] mattray is now known as mrayzenoss

[14-May-2010 16:34:01] <rmatte> I'm done for the day, have a good weekend guys

[14-May-2010 17:07:14] <mrayzenoss> New Cisco ASA VPN ZenPack from malbon: docs/DOC-5881

[14-May-2010 17:23:37] <mrayzenoss> New Sun MIB Process Monitor ZenPack: docs/DOC-5882  Thanks malbon!

[14-May-2010 17:27:03] <theacolyte> Now that could be good

[14-May-2010 17:27:08] * theacolyte looks at hte ASA zenpack

[14-May-2010 17:27:41] <theacolyte> hmm

[14-May-2010 17:27:43] <theacolyte> basic, but good

[14-May-2010 17:28:39] <mrayzenoss> I would like a comparison with docs/DOC-3465

[14-May-2010 17:28:55] <mrayzenoss> still waiting on more documentation for the new one

[14-May-2010 17:29:00] <mrayzenoss> but it's up

[14-May-2010 17:29:32] <theacolyte> That's the one I have actually

[14-May-2010 17:30:28] <mrayzenoss> I mean, they look complimentary, but I don't know which hardware they support

[14-May-2010 17:30:59] <theacolyte> All the new one appears to do is list interface statistics

[14-May-2010 17:31:06] <theacolyte> ad a VPN connection count

[14-May-2010 17:31:25] <theacolyte> the other one is a bit more granular

[14-May-2010 17:31:29] <mrayzenoss> hmm.. so maybe they could be combined

[14-May-2010 17:31:35] <theacolyte> and doesn't seem to do if statistics

[14-May-2010 17:31:40] <mrayzenoss> or maybe keeping them separate is a good idea

[14-May-2010 17:31:54] <theacolyte> I'm not sure if the if statistcs are for the vpn connections, or the interfaces on the ASA itself though

[14-May-2010 17:32:45] <theacolyte> not sure

[14-May-2010 17:33:17] <mrayzenoss> feel free to attach screenshots to the Cisco ASA one if you've got something better

[14-May-2010 17:55:59] fOrsberg is now known as forsberg

[14-May-2010 21:24:07] forsberg is now known as fOrsberg

[15-May-2010 00:00:30] [disconnected at Sat May 15 00:00:30 2010]

[15-May-2010 00:00:30] [connected at Sat May 15 00:00:30 2010]

[15-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[15-May-2010 03:24:38] fOrsberg is now known as forsberg

[15-May-2010 09:41:57] <zykes-> hmmm, how can i monitor processes in linux via zenoss ?

[15-May-2010 10:31:41] <forsberg> you go to processes and add new ones

[15-May-2010 10:31:43] <forsberg> regex

[15-May-2010 10:32:01] <forsberg> and then remodel device and check if you did it right

[15-May-2010 10:32:02] <forsberg> :)

[15-May-2010 15:35:29] <Rolfs> zykes: you have use snmp for monitoring them

[15-May-2010 15:42:25] <zykes-> forsberg: doesn't it get stuff automatically ?

[15-May-2010 15:42:48] <zykes-> Rolfs: i allready got snmp up and running... do i need to config anything manually ?

[15-May-2010 16:18:19] * cahill Discounts!! Our Special Limited Time Offers Up To May,22!!!New BranD!! Notebooks,Plasma and LCD TV's.Buy your electronic needs at our unique prices. Laptop Sony VAIO® VGN-FW590FFD-575,57$!!!Apple MacBook® Air MC234LL/A-695,27$!!! http://www.elplace.com/

[15-May-2010 18:33:34] <Rolfs> you have to add a process to watch under process in the menu.. and then model the device again and see that that process gets into the OS-page

[15-May-2010 18:33:42] <Rolfs> of the device

[15-May-2010 20:22:55] <benw|laptop> hi, I'm having a problem getting Zenoss turned on. I set the sql details and stuff, then did the 'service zenoss start' but I get this....

[15-May-2010 20:23:02] <benw|laptop> Zenoss not initialized.  Performing first-boot initialization...

[15-May-2010 20:23:02] <benw|laptop> Fresh install pre steps

[15-May-2010 20:23:02] <benw|laptop> : No such file or directory

[15-May-2010 20:23:02] <benw|laptop> root@vps [~]#

[15-May-2010 21:50:51] forsberg is now known as fOrsberg

[16-May-2010 00:00:30] [disconnected at Sun May 16 00:00:30 2010]

[16-May-2010 00:00:31] [connected at Sun May 16 00:00:31 2010]

[16-May-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[16-May-2010 04:34:57] fOrsberg is now known as forsberg

[16-May-2010 08:30:44] <zykes-> does zenoss support mysql 5.1 ?

[16-May-2010 12:09:01] <pfo> are there some screen shots from the 3.0 alpha version's new UI?

[16-May-2010 12:24:43] <zykes-> Egor around ? Anyone here having problems with PerfWMI ?

[16-May-2010 12:28:39] <zykes-> i get: MISSING RRD FILE: hostname Processor_LoadPercentage_avg

[16-May-2010 14:44:43] pfo_ is now known as pfo

[16-May-2010 18:10:18] RoAkSoAx_ is now known as andreserl

[16-May-2010 18:26:42] forsberg is now known as fOrsberg

[16-May-2010 19:34:41] andreserl is now known as RoAkSoAx

[17-May-2010 00:00:30] [disconnected at Mon May 17 00:00:30 2010]

[17-May-2010 00:00:31] [connected at Mon May 17 00:00:31 2010]

[17-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[17-May-2010 00:47:31] pfo_ is now known as pfo

[17-May-2010 02:01:40] fOrsberg is now known as forsberg

[17-May-2010 04:28:15] pfo_ is now known as pfo

[17-May-2010 05:03:18] <voxter> where does zenoss get its list of daemons it knows to start ?

[17-May-2010 05:03:26] <voxter> mine seems to be missing zenrender and i keep having to start it by hand

[17-May-2010 05:05:58] <voxter> oh. interesting. zenrender is commented out in /opt/zenoss/bin/zenoss

[17-May-2010 05:06:00] <voxter> I wonder why?

[17-May-2010 05:06:18] <voxter> (running 3.0 alpha)

[17-May-2010 06:43:01] <tehhobbit> anyone good with dmd that could help me with a task, need to move a bunch of servers on a certain iprange to another collector

[17-May-2010 07:00:37] <Simon4> are they all in a single iprange under /Networks in Zenoss?

[17-May-2010 07:06:52] <tehhobbit> yes

[17-May-2010 07:07:45] <Simon4> hokay... so you could do something like this:

[17-May-2010 07:07:56] <Simon4> net = dmd.Networks.findNet('172.23.32.0')

[17-May-2010 07:08:03] <Simon4> for ip in net.ipaddresses():

[17-May-2010 07:08:32] <Simon4>     ip.device().setPerformanceMonitor('perf.monitor.name')

[17-May-2010 07:08:36] <Simon4> commit()

[17-May-2010 07:09:09] <Simon4> switch out the net address with yours, and the perf monitor name, of course.

[17-May-2010 07:10:51] <tehhobbit> sweet tnx :)

[17-May-2010 07:11:24] <Simon4> np :)

[17-May-2010 07:12:57] LarsN_Away is now known as LarsN

[17-May-2010 07:15:24] <LarsN> do any of you know what I would need to put in the "Send String" in the FTPMonitor Zenpack if I wanted to authenticate to the remote FTP server, and put a test file?

[17-May-2010 07:38:54] LarsN is now known as LarsN_Away

[17-May-2010 08:41:45] <Jane_Curry> voxter - I think you'll find that you only need zenrender if you have multiple Zenoss collectors

[17-May-2010 08:50:17] <coofamani> hello

[17-May-2010 09:43:51] <bb0x> hi

[17-May-2010 09:44:15] <bb0x> does anyone know why zenhub wont start if the time was changed on the system ?

[17-May-2010 09:44:28] <bb0x> I'm using 2.3.3

[17-May-2010 09:49:33] <noob_> Has anyone looked at the memory leak on windoes systems in  WMI when zeneventlog in enabled?

[17-May-2010 09:50:11] <noob_> have have all 2008R2 systems and see the leak when i have the zeneventlog enabled.

[17-May-2010 09:51:02] <tehhobbit> well dont think its due to zeneventlog, chances are pretty good that you would get the same result with anything polling the eventlog at a certain interval

[17-May-2010 09:52:23] <noob_> but if i use powershell and only get the last 100 events every 5mins no problem. so i was thinking that it might be that zeneventlog was doing a full log scan. and maybe that could be changed.

[17-May-2010 09:56:01] <bb0x> does anyone know why zenhub wont start if the time was changed on the system ?

[17-May-2010 09:57:03] <tehhobbit> noob_: since the wmi querys kinda look like sql there should be a LIMIT somewhere

[17-May-2010 09:57:20] <tehhobbit> noob_: but still windows that leaking

[17-May-2010 09:59:53] <noob_> that's what i was thinking also(leaking) but I started to think that if my logs where to large to finsh a scan before the next poll .then that isn't a leak but a Q problem.

[17-May-2010 10:01:32] <tehhobbit> noob_: true, being a unix person I'm utterly confused by windows eventlog in general

[17-May-2010 10:01:34] <noob_> And I also found out in power shall that if you do anything other than the last X number of entries.  Say at a "before day" or "with"  the stupid thing does a full scan

[17-May-2010 10:01:55] <tehhobbit> thinking about running a syslog on the windows hosts too instead

[17-May-2010 10:02:14] <noob_> Yet you can count me in the crowd.  The event log is a mess

[17-May-2010 10:07:32] <bb0x> does anyone know why zenhub wont start if the time was changed on the system ?

[17-May-2010 10:08:09] <Simon4> bb0x: nothing in zenhub.log?

[17-May-2010 10:08:35] <bb0x> nope ... it just says after a 2-3 minutes that is not running

[17-May-2010 10:08:42] <bb0x> nothing in the logs

[17-May-2010 10:09:09] <bb0x> also I've deleted zeo1-1.zec and the same problems

[17-May-2010 10:09:32] <bb0x> I;ve rebooted the system and when it comes back up the time wasn't syncronized

[17-May-2010 10:09:45] <bb0x> zenoss was started with the old time

[17-May-2010 10:09:53] <bb0x> I've stop zenoss and sync the time

[17-May-2010 10:10:02] <bb0x> and since then zenhub doesn't want to start

[17-May-2010 10:10:35] <bb0x> I've search on forums and other people had the same problems ... and they said they waited the time that was on server before

[17-May-2010 10:10:41] <bb0x> and restart zenoss after

[17-May-2010 10:11:57] <tehhobbit> bb0x: did the clock go forward or backwards

[17-May-2010 10:12:21] <tehhobbit> could be a version of "timestamp in the future"

[17-May-2010 10:12:35] <bb0x> theacolyte, backward

[17-May-2010 10:12:40] <bb0x> with a few hours ...

[17-May-2010 10:13:01] <bb0x> in a few minutes will be the time which was on server when it started

[17-May-2010 10:13:15] <tehhobbit> try then

[17-May-2010 10:13:43] <bb0x> I think it wrote something on zope database ... or on events database  ... and make some checks based on that

[17-May-2010 10:14:42] <tehhobbit> I'm thinking its takes timestamps and use them to determin what to check first

[17-May-2010 10:15:04] <tehhobbit> instead of starting from scratch

[17-May-2010 10:15:14] <bb0x> where I can check this timestamps ?

[17-May-2010 10:15:31] <tehhobbit> good question

[17-May-2010 10:15:34] <tehhobbit> zendmd

[17-May-2010 10:18:32] <bb0x> tehhobbit, I don't know how to use zendmd ... what kind of commands can I use ?

[17-May-2010 10:20:54] <tehhobbit> bb0x: there is prolly a option you can put it the config file to make it ignore old timestamps

[17-May-2010 10:24:38] <bb0x> the time is now the time that was before on the system .... before sync it backward

[17-May-2010 10:24:44] <bb0x> and zenhub starts ok now

[17-May-2010 10:24:56] <bb0x> thanks, I will check if these timestamps can be ignored

[17-May-2010 10:25:54] <fragfutter> zenoss is modeling my loopback interfaces with 10Mbs. For this reason i get regular performance warnings. How can i tell zenoss to model them with gigabit?

[17-May-2010 10:33:53] <Kristopher> I'm trying to load new MIBS and I'm getting "zen.Zenmib: Unable to find a file that defines SNMPv2-CONF"  It looks like the MIBS have loaded, and I can't find anything on this in the forums.

[17-May-2010 11:41:05] <davetoo> how much luck do people have these days running zenoss under freebsd?

[17-May-2010 11:41:15] <willwh> haven't tried

[17-May-2010 11:41:47] <forsberg> i should try, but havent either

[17-May-2010 11:42:37] <davetoo> I'm not going to spend too much time trying, just curious

[17-May-2010 11:49:55] <davetoo> uh oh, I seem to have exploded my freebsd vm :(

[17-May-2010 11:50:14] <forsberg> :)

[17-May-2010 11:51:12] <davetoo> worse, the esxi host isn't answering pings

[17-May-2010 12:25:59] <mrayzenoss> Time for some new ZenPack updates

[17-May-2010 12:26:25] <mrayzenoss> Cisco ASA VPN: docs/DOC-5881 Thanks malbon!

[17-May-2010 12:26:48] <mrayzenoss> HP EVA Monitor: docs/DOC-5867 Thanks bigegor!

[17-May-2010 12:27:07] <mrayzenoss> Sun MIB Process Monitor: docs/DOC-5882 Thanks again malbon!

[17-May-2010 12:27:39] <mrayzenoss> and a slew of updates from bigegor

[17-May-2010 12:27:56] <mrayzenoss> updated MySQL ODBC Database Monitor: docs/DOC-3388

[17-May-2010 12:28:11] <mrayzenoss> updated WBEM Data Source: docs/DOC-3409

[17-May-2010 12:28:32] <mrayzenoss> WMI Windows Performance has now hit 2.0: docs/DOC-3500

[17-May-2010 12:28:51] <mrayzenoss> and rmatte updated the Collector Tool ZenPack: docs/DOC-5831

[17-May-2010 12:41:20] cgibbons_ is now known as cgibbons

[17-May-2010 12:43:05] <mrayzenoss> and 1 last ZenPack update for today, bigegor's NetWare SNMP is up to 1.2: docs/DOC-3499

[17-May-2010 12:43:48] <rmatte> cool

[17-May-2010 12:44:54] <jro> any release schedule yet for 3.0?

[17-May-2010 12:45:39] <mrayzenoss> probably going to go to beta in the next week

[17-May-2010 12:46:01] <chemist> bigegor: where does your distributed collectors ZenPack write it's log to?

[17-May-2010 12:46:03] <jro> liking the new interface thus far but feel a bit lost :)

[17-May-2010 12:46:10] cgibbons_ is now known as cgibbons

[17-May-2010 12:46:32] <mrayzenoss> jro: I believe the alpha->beta transition indicates that all the screens have mostly finalized

[17-May-2010 12:46:36] <bigegor> chemist: it doesn't

[17-May-2010 12:47:38] <chemist> ok, I am getting some command not found when I tryto update

[17-May-2010 12:47:44] <jro> mrayzenoss: thanks, I need to move our box to a new platform so figured I'd just wait for 3.0 to save me an upgrade

[17-May-2010 12:47:56] <chemist> but the output doesn't show what command fails

[17-May-2010 12:48:16] <bigegor> chemist: ithink its cpio

[17-May-2010 12:48:18] <mrayzenoss> jro: I can tell you you're at least 5 or 6 weeks away

[17-May-2010 12:48:37] <mrayzenoss> the beta is going to be awhile, we really want to iron out upgrade issues

[17-May-2010 12:48:52] <jro> as expected being a major relase :)

[17-May-2010 12:48:59] <jro> thanks for the update

[17-May-2010 12:49:09] <chemist> bigegor: do I have to put cpio in zenoss user path?

[17-May-2010 12:50:49] <bigegor> chemist: no, put it in /usr/bin or in /bin

[17-May-2010 12:51:37] <chemist> ok

[17-May-2010 12:55:40] <chemist> I just restarted the zenoss server, all the daemons are running but I can't access the UI

[17-May-2010 12:55:47] <chemist> or zendmd

[17-May-2010 12:56:47] <chemist> I get this in zendmd:  18:56:21 WARNING ZEO.zrpc (3834) CW: error connecting to ('localhost', 8100): ECONNREFUSED

[17-May-2010 12:58:28] <bigegor> chemist: you can not get zendmd on remote collector, without changing zeo.conf

[17-May-2010 12:58:45] <chemist> this is on master

[17-May-2010 12:59:25] <bigegor> check if zeoctl is running

[17-May-2010 13:00:00] <chemist> it is

[17-May-2010 13:00:26] <bigegor> hmmm

[17-May-2010 13:00:57] <chemist> it's back now

[17-May-2010 13:01:02] <chemist> after another restart

[17-May-2010 13:04:58] <mrayzenoss> Sweet, just got confirmed for FLOSS Weekly this week

[17-May-2010 13:11:28] <rmatte> cool

[17-May-2010 13:12:14] <rmatte> Ah, that's one of Leo Laporte's things

[17-May-2010 13:12:17] <rmatte> interesting

[17-May-2010 13:12:24] <chemist> is there a way to get the ALL the installed software with WMIPerf?

[17-May-2010 13:13:11] <rmatte> chemist: direct that question at Egor since he'd know best

[17-May-2010 13:13:35] <rmatte> you're talking about having it show up on the software tab in Zenoss I assume?

[17-May-2010 13:14:36] <chemist> rmatte: yes

[17-May-2010 13:15:05] <chemist> at present it only shows stuff installed with MSI

[17-May-2010 13:15:35] <rmatte> ah

[17-May-2010 13:16:43] <rmatte> mrayzenoss: Confirmed for Wednesday I see

[17-May-2010 13:16:53] <rmatte> mrayzenoss: I'll be sure to check it out

[17-May-2010 13:17:29] <mrayzenoss> I've watched it live before, there were 800 people in the live-chat

[17-May-2010 13:18:21] <rmatte> cool

[17-May-2010 13:18:47] <rmatte> is that 9:30am EST?

[17-May-2010 13:18:55] <rmatte> or CST?

[17-May-2010 13:18:59] <mrayzenoss> PDT

[17-May-2010 13:19:02] <rmatte> ah

[17-May-2010 13:21:44] <rmatte> so that's 6:30am EST?

[17-May-2010 13:21:53] <bigegor> chemist: what do you mean 'ALL'?

[17-May-2010 13:22:35] <mrayzenoss> rmatte: 12:30pm EDT

[17-May-2010 13:22:49] <rmatte> bigegor: he's saying that only software which was installed via an msi installer is visible

[17-May-2010 13:23:18] <rmatte> ah

[17-May-2010 13:23:19] <mrayzenoss> rmatte: Pacific is Eastern minus 3 hours

[17-May-2010 13:23:29] <mrayzenoss> not the other way

[17-May-2010 13:23:38] <mrayzenoss> I've been burned repeatedly by timezones

[17-May-2010 13:24:08] <rmatte> hmmm, I did "current local time PDT" on google and the first result was: http://www.timeanddate.com/worldclock/city.html?n=224

[17-May-2010 13:24:19] <rmatte> which shows 11:24am right now

[17-May-2010 13:24:30] <rmatte> and it's 2:24pm here

[17-May-2010 13:24:34] <rmatte> so yeh, you're right

[17-May-2010 13:24:37] <mrayzenoss> heh

[17-May-2010 13:24:40] <chemist> bigegor: as rmatte said, is there a way to collect the other soft?

[17-May-2010 13:25:06] <mrayzenoss> chemist: I think SNMP and WMI return different results on Windows

[17-May-2010 13:25:50] <mrayzenoss> I can't remember where I heard that, perhaps a ticket somewhere

[17-May-2010 13:25:57] <bigegor> I will check this.

[17-May-2010 13:26:18] <cgibbons> Normally to get the other stuff you have to look through the registry.

[17-May-2010 13:26:33] <chemist> I think SNMPsearches the Uninistall reg key

[17-May-2010 13:26:40] <zykes-> what's WNEM ?

[17-May-2010 13:26:58] <zykes-> WBEM

[17-May-2010 13:27:24] <rmatte> zykes-: google it, first result

[17-May-2010 13:27:25] <chemist> zykes-: Web Based Enterprise Management

[17-May-2010 13:27:45] <zykes-> hmm k

[17-May-2010 13:30:31] <rmatte> afk for a few

[17-May-2010 13:35:43] <mrayzenoss> If you're writing any Storage ZenPacks, take a look at Egor's HP EVA Monitor: docs/DOC-5867  Very cool

[17-May-2010 13:35:55] <bigegor> hmmm, Win32_Product is realy so bad :( it dog slow and return only MSI installed software :( typical M$

[17-May-2010 13:38:03] <cgibbons> Nothing written in the past 10 year should have been written without it :)

[17-May-2010 13:39:45] <zykes-> bigegor: any hints to why cpu monitoring doesn't work with wmiperf in your pack ?

[17-May-2010 13:40:59] <bigegor> it works. Tested with Win2003, Win2000 and XP

[17-May-2010 13:42:27] <zykes-> i just get "MISSING RRD FILE"

[17-May-2010 13:42:51] <zykes-> <hostname> Processor_LoadPercentage_avg

[17-May-2010 13:43:02] <zykes-> CPU Utilization over 90 shows

[17-May-2010 13:44:18] <bigegor> Check zenperfwmi.log for errors?

[17-May-2010 13:56:48] <zykes-> none

[17-May-2010 13:57:42] <mrayzenoss> zykes-: did you restart everything?

[17-May-2010 13:57:55] <mrayzenoss> adding the WMI Daemon requires a zenoss restart

[17-May-2010 13:59:17] <zykes-> yes

[17-May-2010 14:00:06] <bigegor> try run zenperfwmi daemon with debug options: zenperfwmi -v10 -d yourserver

[17-May-2010 14:04:15] <zykes-> what to look for ?

[17-May-2010 14:04:27] <bigegor> for errors

[17-May-2010 14:13:24] <chemist> I installed WMIDatasource and WMIPerf_Windows on the master, updated the collector but I get this in zenperfwmi log

[17-May-2010 14:13:28] <chemist> http://pastebin.com/E6td5fSH

[17-May-2010 14:13:37] <chemist> and zenperfwmi wont start

[17-May-2010 14:17:16] <rmatte> reinstall the WMI Data Source pack and then restart Zenoss again

[17-May-2010 14:18:00] <chemist> do I have to remove it first? or just reinstall?

[17-May-2010 14:19:55] <bigegor> install last WMIDataSource version (2.3)

[17-May-2010 14:20:03] LarsN_Away is now known as LarsN

[17-May-2010 14:20:52] <LarsN> docs/DOC-4673;jsessionid=8732D52E8BEEDE114F3A8674C4655FF6.node0  can someone assist me with the second part of these instructions?  Specifically I don't understand where it's trying to get me to go when it says "You will then need to navigate to /Events/Perf/Snmp and select More -> Transform from the dropdown.  Insert the following transform and click save."

[17-May-2010 14:26:51] <LarsN> Also Zenoss is reporting my nfs mount as a total size of 1.6T, while df -h is showing more correctly 3.6T

[17-May-2010 14:26:57] <LarsN> which seems somewhat sub-optimal.

[17-May-2010 14:36:24] <zykes-> no errors that i can see

[17-May-2010 14:42:05] <bigegor> zykes: can you see returned results?

[17-May-2010 14:59:21] <chemist> bigegor: this is what I get when I "Update Remote Monitors"

[17-May-2010 14:59:22] <chemist> http://pastebin.com/crjtejtg

[17-May-2010 15:02:35] <bigegor> it seems you have problem with ssh. Try as zenoss user connect to remote computer over ssh.

[17-May-2010 15:03:59] <chemist> I get the libcrypto message but it still logs me in

[17-May-2010 15:04:15] <chemist> from master to collector

[17-May-2010 15:08:30] <bigegor> do you have zenoss installed on remote collector?

[17-May-2010 15:08:41] <chemist> yes

[17-May-2010 15:08:58] <chemist> and I receive events from its ddevices

[17-May-2010 15:09:31] <bigegor> it seems like you loged in to remote collector not as zenoss user

[17-May-2010 15:11:14] <chemist> monitor sshd[5159]: Accepted publickey for zenoss from 192.168.66.17 port 35739 ssh2

[17-May-2010 15:16:29] <bigegor> check zenoss environment on remote collector

[17-May-2010 15:17:11] <chemist> ok

[17-May-2010 15:31:07] <nwilkens> Running into a ZenPing issue in 2.5.2.  Hosts are responding fine to ping, but Zenoss reports them as down.  I see a few other posts like this in the forum with no solutions.  Any ideas?

[17-May-2010 15:31:33] <jb> anyone have a graph using COUNTER or DERIVE datasources?

[17-May-2010 15:32:26] <coofamani> I think the interface graphs are COUNTER, no?

[17-May-2010 15:32:52] <jb> perhaps

[17-May-2010 15:33:02] <jb> i have a problem with some of my custom datasources for squid which are COUNTER.

[17-May-2010 15:33:12] <jb> when squid restarts on the host, the COUNTER goes back to 0.

[17-May-2010 15:33:24] <jb> which causes my graphs to show a funky negative value

[17-May-2010 15:33:32] <Simon4> jb: make it a DERIVE with a minimum of zero

[17-May-2010 15:33:32] <coofamani> set a min/max value

[17-May-2010 15:33:35] <coofamani> thats normal behavior

[17-May-2010 15:33:43] <Simon4> it's what we do for such things

[17-May-2010 15:33:48] <jb> hm one sec

[17-May-2010 15:33:51] <coofamani> same thing happens when a network switch crashes and the counters reset

[17-May-2010 15:34:05] <jb> that doesn't happen on when I restart BIND for example (those are COUNTER sources, too)

[17-May-2010 15:34:19] <coofamani> sometimes youll also get spikes from that reset to 0, the MAX prevents those, whereas MIN should prevent the negs

[17-May-2010 15:34:36] <coofamani> jb: probably depends on whether the service has a persistent counter betweenr eboots

[17-May-2010 15:34:37] <jb> ok, i'll try a RRD Min of 0

[17-May-2010 15:36:25] <jb> thanks :)

[17-May-2010 15:41:23] <jb> hmm.. it still went negative

[17-May-2010 15:41:27] <jb> maybe I need to delete the RRD

[17-May-2010 15:41:46] <Simon4> you'll need to rrdtune it, or delete it (rrdtool tune)

[17-May-2010 15:41:50] <jb> thanks

[17-May-2010 15:41:56] <coofamani> the MIN/MAX will onlu apply to future values unless you rebuild it

[17-May-2010 15:44:44] <jb> i just wiped the RRD's in question

[17-May-2010 15:51:39] <rmatte> nwilkens: Zenoss doesn't actually use the "ping" command

[17-May-2010 15:51:47] <rmatte> check the ZenPing logfile for errors

[17-May-2010 15:52:31] <nwilkens> right.

[17-May-2010 15:52:57] <nwilkens> ZenPing shows WARNING zen.ZenPing: ip xxx is down.

[17-May-2010 15:53:14] <nwilkens> But a ping indicates the host is up.

[17-May-2010 15:53:19] <nwilkens> and the host is up.

[17-May-2010 15:56:27] <rmatte> I understand...

[17-May-2010 15:56:36] <rmatte> are there errors in $ZENHOME/log/zenping.log?

[17-May-2010 15:57:25] <rmatte> any time I've seen that behaviour there have been errors in the zenping logfile

[17-May-2010 15:58:36] <rmatte> also, when you say a ping indicates that the host is up, are you performing that test ping from the Zenoss box itself?

[17-May-2010 15:58:40] <rmatte> or from your workstation?

[17-May-2010 15:59:43] <nwilkens> performing the ping From the server.

[17-May-2010 15:59:55] <nwilkens> brb

[17-May-2010 15:59:58] <rmatte> k

[17-May-2010 16:14:02] <gwar9999> hello all.  Does anybody have a working snipped of xmlrpc code (preferable in python) to add a device?  I tried following the 2.5.2 documentation but it doesn't work for me.

[17-May-2010 16:14:08] <gwar9999> snippet

[17-May-2010 16:16:22] <rmatte> dmd.DeviceLoader.loadDevice

[17-May-2010 16:16:25] <rmatte> is one

[17-May-2010 16:17:31] <rmatte> http://fpaste.org/cDEo/

[17-May-2010 16:17:39] <rmatte> example including variables

[17-May-2010 16:19:00] <rmatte> that's straight python though and needs to be executed locally on the Zenoss server itself

[17-May-2010 16:19:51] <gwar9999> rmatte: thanks but I need to use xmlrpc since the code will reside on a different host.

[17-May-2010 16:19:54] <rmatte> docs/DOC-3805

[17-May-2010 16:19:56] <rmatte> read that

[17-May-2010 16:20:09] <rmatte> covers REST, XMLRPC, and some others

[17-May-2010 16:20:42] <gwar9999> rmatte: I've read through it a few times and things aren't working for me.  If I know the group I can query and delete.  However, I've had no success in "adding" a device.

[17-May-2010 16:21:11] <rmatte> you're sure you've been specifying all of the options correctly?

[17-May-2010 16:24:20] <gwar9999> Oh I think I got it working... finally.  Type-o on this try, however, it still results in an exception (junk after document element: line 6) when parsing the xmlrpc response, but the device does make it.  Thanks for helping.

[17-May-2010 16:25:17] <mrayzenoss> nuts, my FLOSS Weekly interview got postponed again

[17-May-2010 16:37:04] <nwilkens> rmatte, Sorry for the delay, .. No errors in the zenping log, except for the 'ip down'.  Enabling debug shows 'unexpected pkt' for a few hosts, but not any of the ones that are down.

[17-May-2010 16:37:50] <nwilkens> Deleting them and recreating them has solved the problem in the past

[17-May-2010 16:43:39] <QubeZ> hello all

[17-May-2010 16:44:06] <QubeZ> I have 2 Serial links bonded into a multlink. However, when one of the serial lines goes down, Zenoss doesn't not alert. Any ideas? It is set to monitoring true.

[17-May-2010 16:53:03] <rmatte> nwilkens: have you tried that yet?

[17-May-2010 16:55:56] <nwilkens> I havn't yet deleted any nodes and re-added them.

[17-May-2010 16:56:28] <rmatte> have you tried just restarting Zenoss?

[17-May-2010 16:56:55] <rmatte> also, perhaps a reindex in zendmd

[17-May-2010 16:57:05] <rmatte> you'd be surprised what kind of stuff that has fixed for me in the past

[17-May-2010 16:57:36] <LarsN> do any of you have any better/alternative suggestions for monitoring the status of an NFS mount than: docs/DOC-4673;jsessionid=8732D52E8BEEDE114F3A8674C4655FF6.node0

[17-May-2010 16:58:13] <nwilkens> How do you reindex in zendmd?

[17-May-2010 16:58:50] <rmatte> LarsN: not really, that was the best way I could find

[17-May-2010 16:58:50] <LarsN> What I'd idealy like to see is as simple as:  server:/path/mount /mnt/path (nfs)

[17-May-2010 16:58:53] <forsberg> QubeZ you would propably have to make some kind of manuel check on the links then

[17-May-2010 16:58:57] <LarsN> if that goes away, we get an alert.

[17-May-2010 16:59:24] <rmatte> LarsN: well, you could develop a custom SSH datasource to do it

[17-May-2010 16:59:30] <QubeZ> forsberg: For example, Serial 3/3:3 is showing graphs and set to monitoring true, but when it goes down... Zenoss doesn't alert me. I find that strange.

[17-May-2010 16:59:48] <rmatte> LarsN: but honestly, that method that's on that page works just as well

[17-May-2010 16:59:57] <LarsN> rmatte: I'll continue to follow the above path, and if we find it lacking I guess maybe we'll roll our own.  Thanks.

[17-May-2010 17:00:03] <QubeZ> LarsN: it works for me great too.

[17-May-2010 17:00:13] <LarsN> rmatte: just need to change the event level from 1, to 4 or 5 I assume in the transofrm

[17-May-2010 17:00:16] <LarsN> transform.

[17-May-2010 17:00:21] <QubeZ> I've been running that piece of code for awhile, actually my situation is what inspired rmatte to write it and it works

[17-May-2010 17:00:25] <rmatte> yeh, customize it however you need to

[17-May-2010 17:00:31] <LarsN> rgr.

[17-May-2010 17:00:40] <LarsN> danke.

[17-May-2010 17:00:52] <LarsN> (that's roger, for the non ham types :) )

[17-May-2010 17:01:04] <rmatte> actually, the transform already does that

[17-May-2010 17:01:05] <rmatte> evt.severity = 5

[17-May-2010 17:01:12] <QubeZ> I have a server that sometimes loses the NFS mount and so I have that to monitor and run a ssh script to remount the share.

[17-May-2010 17:01:32] <LarsN> I must be doing something wrong then.

[17-May-2010 17:02:09] <rmatte> why's that?

[17-May-2010 17:02:11] <LarsN> I get an "info" alert saying it can't read /mnt/path  and randomly the NFS mount just vanishes from the OS tab if it wasn't mounted for "some" amount of time.

[17-May-2010 17:02:36] <rmatte> well, my transform automatically locks the mount so that it can't vanish

[17-May-2010 17:02:36] <LarsN> if it's mounted up, everything's fine.  It sees the mount in the OS tab, although it incorrectly reports the size.

[17-May-2010 17:02:49] <LarsN> df -h shows 3.6T  the os tab shows 1.6T

[17-May-2010 17:02:56] <rmatte> so I'm not sure if you have a problem with the spacing of the transform or something

[17-May-2010 17:03:13] <rmatte> If it's showing incorrect info, remodel

[17-May-2010 17:03:13] <LarsN> that's possible.

[17-May-2010 17:03:33] <rmatte> though SNMP may have some limit on hard drive sizes, not sure

[17-May-2010 17:03:42] <LarsN> that's also possible.

[17-May-2010 17:04:03] * LarsN is running ancient net-snmp.

[17-May-2010 17:04:23] <rmatte> net-snmp is pretty good

[17-May-2010 17:04:29] <rmatte> you mean you're running an old version of it?

[17-May-2010 17:05:09] <LarsN> 5.4.2.1_6

[17-May-2010 17:05:12] <LarsN> on FreeBSD

[17-May-2010 17:05:19] <rmatte> ah

[17-May-2010 17:05:49] <rmatte> you sure that info event that you're seeing isn't a debug event?

[17-May-2010 17:05:58] <LarsN> it is a debug event.

[17-May-2010 17:06:01] <LarsN> the problem is

[17-May-2010 17:06:10] <rmatte> k, then you haven't applied my transform properly

[17-May-2010 17:06:15] <LarsN> for the last hour, the mount has been detached, and zenoss never triggered an event :)

[17-May-2010 17:06:22] <rmatte> also, make sure the OID in the debug event matches the OID in the transform

[17-May-2010 17:06:25] <LarsN> i knew I must have been doing something wrong.

[17-May-2010 17:06:26] <rmatte> I hadn't tested with FreeBSD

[17-May-2010 17:06:36] <rmatte> it's possible that it's slightly different and needs to be changed in the transform

[17-May-2010 17:06:41] <LarsN> Error reading value for "/mnt/lllyapp-ftp" on lllyprxy01.sixfeetup.com (oid .1.3.6.1.2.1.25.2.3.1.6.36 is bad)

[17-May-2010 17:06:50] <LarsN> so I want to put that oid in the transform.

[17-May-2010 17:06:58] <LarsN> in place of the one provided on the link above.

[17-May-2010 17:08:43] <LarsN> adjustments made.  Now I get to give Zenoss it's important time to think about what it's done :)

[17-May-2010 17:09:18] <rmatte> well hold up

[17-May-2010 17:09:33] <rmatte> I think I only had a partial OID in the transform

[17-May-2010 17:09:45] <rmatte> which is how it should be

[17-May-2010 17:09:49] <LarsN> yours ends at the .6 before the .36

[17-May-2010 17:09:51] * rmatte verifies

[17-May-2010 17:09:54] <LarsN> otherwise ti matches.

[17-May-2010 17:09:59] <LarsN> s/ti/it

[17-May-2010 17:10:01] <rmatte> yeh, then it's fine

[17-May-2010 17:10:04] <rmatte> remove the .36

[17-May-2010 17:10:13] <rmatte> because that's the part that changes from drive to drive

[17-May-2010 17:10:17] <LarsN> and the preceeding .

[17-May-2010 17:10:19] <LarsN> I assme

[17-May-2010 17:10:20] <LarsN> assume

[17-May-2010 17:10:30] <rmatte> make it basically look like it did before

[17-May-2010 17:10:31] <rmatte> lol

[17-May-2010 17:10:46] <LarsN> done.

[17-May-2010 17:10:48] <rmatte> also, if you click on the drive on the OS tab

[17-May-2010 17:10:54] <rmatte> what is the "Type:" shown as?

[17-May-2010 17:11:11] <LarsN> networkDisk

[17-May-2010 17:11:16] <rmatte> k

[17-May-2010 17:11:22] <LarsN> I see

[17-May-2010 17:11:22] <rmatte> the transform should work as is then

[17-May-2010 17:11:26] <LarsN> the nsmp index is 36

[17-May-2010 17:11:27] <LarsN> :)

[17-May-2010 17:11:34] <LarsN> which is where the last entry in the oid comes from.

[17-May-2010 17:11:48] <rmatte> correct

[17-May-2010 17:11:50] <LarsN> <--- needs new drivers for his fingers.  Apparently these are faulty.

[17-May-2010 17:12:16] <rmatte> but yeh, you put the transform exactly where it says to in the guide?

[17-May-2010 17:12:36] <rmatte> Also, when you saved the transform, the text didn't turn red (indicating a syntax error)?

[17-May-2010 17:12:40] <LarsN> http://our.zenoss//zport/dmd/Events/Perf/Snmp

[17-May-2010 17:12:46] <LarsN> correct.

[17-May-2010 17:12:53] <rmatte> k

[17-May-2010 17:12:57] <rmatte> so it should work then

[17-May-2010 17:13:02] <rmatte> wait

[17-May-2010 17:13:06] <rmatte> what version of Zenoss are you using?

[17-May-2010 17:13:24] <LarsN> that is a fantastic question.

[17-May-2010 17:13:26] * LarsN checks.

[17-May-2010 17:13:29] <LarsN> 2.4.x iirc.

[17-May-2010 17:13:34] <rmatte> Settings -> Versions

[17-May-2010 17:13:57] <LarsN> 2.4.5

[17-May-2010 17:17:10] <LarsN> you're going to tell me I need to upgrade to 2.5.x aren't you :)

[17-May-2010 17:19:31] <rmatte> no

[17-May-2010 17:19:35] <rmatte> one second...

[17-May-2010 17:20:07] <rmatte> (you really should consider upgrading to 2.5.2, but I'll give you some stuff to tide you over for the time being...)

[17-May-2010 17:20:23] <LarsN> i do appreciate the help.

[17-May-2010 17:20:36] <rmatte> docs/DOC-4434

[17-May-2010 17:20:47] <rmatte> as the zenoss user do this: zenpatch 15003 && zenpatch 14397 && zenpatch 15214 && zenpatch 14049 && zenpatch 13304 && zenpatch 16733

[17-May-2010 17:20:50] <rmatte> then restart Zenoss

[17-May-2010 17:20:58] <rmatte> one of those patches will allow that transform to work properly

[17-May-2010 17:21:08] <rmatte> and the others are good to have

[17-May-2010 17:21:41] <LarsN> I'm going to have to sit down and determine what the safe way to upgrade our installation is :)

[17-May-2010 17:22:01] <rmatte> What type of install did you do?

[17-May-2010 17:22:06] <LarsN> I doubt our Sr. Sysadmin who's traditionally done it is going to want to this time.

[17-May-2010 17:22:21] <rmatte> I can help you out with the upgrade, I just finished upgrading my 13 Zenoss servers from 2.4.5 to 2.5.2 last week

[17-May-2010 17:22:30] <LarsN> it's a core install, with a lot of unsupported bits.....

[17-May-2010 17:22:34] <rmatte> and ran in to just about every problem possible along the way

[17-May-2010 17:22:51] <LarsN> for instance we have additional collectors working.  which is unsupported in core.

[17-May-2010 17:22:51] <rmatte> We have unsupported bits too...

[17-May-2010 17:22:59] <rmatte> not sure how complex yours are compared to ours

[17-May-2010 17:23:06] <LarsN> should be simple.

[17-May-2010 17:23:12] <LarsN> :)

[17-May-2010 17:23:13] <rmatte> nah, it's supported, they just don't make it easy to configure it

[17-May-2010 17:23:16] <rmatte> that's not a big deal

[17-May-2010 17:23:17] <LarsN> I just need to read the docs.

[17-May-2010 17:23:33] <LarsN> and make sure we have a solid backup before I do it.

[17-May-2010 17:23:44] <rmatte> yeh, that's the essential part

[17-May-2010 17:23:51] <rmatte> if you run in to any funky issues come find me in here

[17-May-2010 17:24:04] <rmatte> I came across some really weird issues after the upgrade but managed to figure them all out

[17-May-2010 17:24:06] <LarsN> I'll start with the patches tonight. and look into the upgrade this week.

[17-May-2010 17:24:12] <rmatte> cool

[17-May-2010 17:24:13] <LarsN> cool, thanks rmatte

[17-May-2010 17:24:19] <rmatte> no problem

[17-May-2010 17:24:56] <rmatte> well, I'm out for the day

[17-May-2010 17:24:57] <rmatte> later

[17-May-2010 17:25:27] <LarsN> cya later rmatte thanks again.

[17-May-2010 17:27:56] pfo_ is now known as pfo

[17-May-2010 17:40:58] LarsN is now known as LarsN_Away

[17-May-2010 18:33:45] <gwar9999> can I use the REST or xmlrpc interface to query a device regardless of it's class?  That is, if I don't know the class of a host (maybe it's linux, maybe it's windows)?  Seems like the examples shown always include the class (aka devicepath).

[17-May-2010 19:01:49] LarsN_Away is now known as LarsN

[17-May-2010 19:02:12] <LarsN> can someone point me to where I actually download the above listed zenpatches?

[17-May-2010 19:03:01] <LarsN> or do I really just type...  zenpatch <patch#> from the server?

[17-May-2010 19:04:27] <LarsN> nevermind :)

[17-May-2010 19:13:45] <LarsN> rmatte: side note....  applying those patches to 2.4.5 fixed the transform.  It's now working as expected.  Many thanks!

[17-May-2010 19:24:12] <sergeymasushko> hi 2 all. I have a weird situation. I have 2 device classes called "Server" and "VPNs". I tried to create the same thershold "/FileSystem      /high disk usage" as in the "Server" class but I get the following error: zenhub|User-supplied Python expression (here.getTotalBlocks() * .85) for maximum value caused error: ['usedBlocks_usedBlocks'] The threshold high disk usage in template /zport/dmd/Devices/VPNs/rrdTemplates/FileSystem has been disabled.

[17-May-2010 19:24:12] <sergeymasushko> What I'm doing wrong?

[17-May-2010 19:31:38] <LarsN> how frequently does Zenoss remodel devices automatically, if at all?

[17-May-2010 19:32:02] <sergeymasushko> 270 minutes as far as I remember

[17-May-2010 19:32:32] <sergeymasushko> you can force remodeling of a particular device if needed of adjust modeling cycle time

[17-May-2010 19:33:02] <sergeymasushko> LarsN: do you need steps for above actions?

[17-May-2010 19:35:28] <LarsN> manual modling I'm familiar with

[17-May-2010 19:35:44] <LarsN> I just wanted to make sure, as we make changes to the server's NFS mount points that eventually they'll get picked up

[17-May-2010 19:35:47] <LarsN> :)

[17-May-2010 19:35:50] <LarsN> thanks.

[17-May-2010 19:35:54] <LarsN> every 4 hours or so is acceptable.

[17-May-2010 19:36:19] <LarsN> I'm manually updating 12 servers, that aren't quite right yet, but getting them close none the less.

[17-May-2010 19:36:34] <sergeymasushko> :)

[17-May-2010 19:37:19] <LarsN> told my boss once I'm done with this bit.  I'm drinking a six pack

[17-May-2010 19:37:21] <LarsN> and going to bed.

[17-May-2010 19:37:23] <LarsN> :)

[17-May-2010 20:01:04] LarsN is now known as LarsN_Away

[18-May-2010 00:00:30] [disconnected at Tue May 18 00:00:30 2010]

[18-May-2010 00:00:31] [connected at Tue May 18 00:00:31 2010]

[18-May-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[18-May-2010 02:25:15] Simon4_ is now known as Simon4

[18-May-2010 09:59:53] <cgibbons> creepily quiet today

[18-May-2010 10:07:16] <fragfutter> i could ask some questions about hp-eva ...

[18-May-2010 10:07:30] <mrayzenoss> fragfutter: bigegor logs on frequently

[18-May-2010 10:07:46] <fragfutter> but lives in a different timezone like the rest of the channel ;)

[18-May-2010 10:09:00] <mrayzenoss> apparently he's a nightowl, he was here yesterday at 5pm my time, which I assume was around midnight

[18-May-2010 10:21:29] <fragfutter> looks like his zenpack can't handle my eva. and after a day with HP-SIM i can't wrap my head around zenoss again.

[18-May-2010 10:25:33] <fragfutter> SIM, CIM, SMI-S, WBEM, wcproxy, cve. Hey snmp would be so nice...

[18-May-2010 10:42:26] frozty_s1 is now known as frozty_sa

[18-May-2010 11:00:03] <kobalt> so question, with snmp traffic graphs, I have a graph that in the daly graph it shows a 16mb spike but if I zoom out to a month the highest its shows is 400 k is this normal?

[18-May-2010 11:01:01] <kobalt> well daily shows 2mb and weekly shows 800k etc

[18-May-2010 11:02:22] <kobalt> but the max at the bottom shows the max data on the circuit

[18-May-2010 11:02:37] <kobalt> its weird its almost like the graph is averaging the data

[18-May-2010 11:05:34] <cgibbons> it does indeed as you zoom out depending upon the way the graph was defined

[18-May-2010 11:06:30] <kobalt> so I want the data to not average but give me actuall data if on the 1st they hit 29mb I want to see it in the graph

[18-May-2010 11:06:51] <kobalt> how would I change the graph definition

[18-May-2010 11:07:14] <gwb2353> monkey about with RRD?

[18-May-2010 11:07:45] <zykes-> any howto on getting https for zenoss ?

[18-May-2010 11:07:48] <gwb2353> http://oss.oetiker.ch/rrdtool/tut/rrd-beginners.en.html

[18-May-2010 11:15:24] <gwb2353> there's a knowledge base document "Using HTTPS/SSL for your Zenoss UI" that you could probably convice mrayzenoss to send to you; haven't found a good subsitute for that on the web

[18-May-2010 11:20:52] <zykes-> mrayzenoss: ?

[18-May-2010 13:24:02] LarsN_Away is now known as LarsN

[18-May-2010 13:24:11] <LarsN> rmatte: ayt? :)

[18-May-2010 14:03:15] <SEJeff> WHat would be the best way to access all events for a given device from the dmd commandline?

[18-May-2010 14:26:20] <chemist> bigegor: are you about?

[18-May-2010 14:46:21] <bigegor> chemist: hey

[18-May-2010 14:46:23] LarsN is now known as LarsN_Away

[18-May-2010 14:52:02] <zykes-> bigegor: about yesterday the error about not getting load average to work, what was the command again to test with ?

[18-May-2010 14:52:58] <bigegor> zykes: zenperfwmi -v10 -d monitoreddevicename

[18-May-2010 14:58:46] <bigegor> yesterday, I was trying to play a bit with 3.0 alpha. Is it possible to install zenpacks on it?

[18-May-2010 15:00:00] <mrayzenoss> so the catch for installing ZenPacks on 3.0 is they have to be compiled with Python 2.6

[18-May-2010 15:00:14] <mrayzenoss> which means I have a huge task ahead of me to repackage all the Community ones

[18-May-2010 15:02:31] <bigegor> thanks Matt, i will try to convert some of my zenpacks.

[18-May-2010 15:02:57] <mrayzenoss> we're putting together a document for upgrading ZenPacks for 3.0, just reviewed a final draft of it

[18-May-2010 15:03:17] <mrayzenoss> bigegor: custom data sources will need to be upgraded

[18-May-2010 15:04:34] <bigegor> hmmm, looks like I've got a huge task...

[18-May-2010 15:07:15] <mrayzenoss> well, I'm going to be working on a lot of them

[18-May-2010 15:27:14] <chemist> bigegor: hi egor

[18-May-2010 15:27:48] <bigegor> hi

[18-May-2010 15:27:51] <chemist> can you take a look at this and see what you make of it?

[18-May-2010 15:28:21] <chemist> http://pastebin.com/yjN69Sef

[18-May-2010 15:29:18] <chemist> bigegor: this is in the log whenever zenperfwmi is restarted

[18-May-2010 15:33:11] <bigegor> try model device without community.wmi.DiskDriveMap plugin

[18-May-2010 15:37:01] <chemist> still the same

[18-May-2010 15:37:16] <chemist> I removed the plugin and restarted zenperfwmi

[18-May-2010 15:38:46] <bigegor> this device monitored by remote collector?

[18-May-2010 15:42:14] <chemist> http://pastebin.com/zU9ny56u

[18-May-2010 15:42:16] <chemist> yes

[18-May-2010 15:43:06] <chemist> initial error is gone, but I still can not get zenperfwmi to run on the remote collector

[18-May-2010 15:45:04] <chemist> this is the error on the remote http://pastebin.com/VbeysDms

[18-May-2010 15:45:20] <bigegor> i have the same problem with HPEVAMon zenpack. If in components class exist method which used self.cacheRRDValue function, it blocked remote collector

[18-May-2010 15:47:01] <chemist> and that stops zenperfwmi from running on remote?

[18-May-2010 15:47:39] <bigegor> yes.

[18-May-2010 15:48:04] <chemist> any ideas?

[18-May-2010 15:50:58] <bigegor> remove community.wmi.DiskDriveMap plugin from /CIM/WMI. Remove and add your device.

[18-May-2010 15:53:25] <chemist> ok, will try

[18-May-2010 15:53:31] <chemist> another question

[18-May-2010 15:53:55] <chemist> can I remove a remote collector and add it again later?

[18-May-2010 15:54:09] <chemist> will I have to re-add the devices?

[18-May-2010 15:56:05] <bigegor> move all devices from remote collector to master, and then remove remote collector.

[18-May-2010 15:56:41] <chemist> ok, thanks

[18-May-2010 15:56:50] <chemist> great zenpacks by the way

[18-May-2010 15:57:00] <bigegor> thanks

[18-May-2010 16:00:33] <bigegor> Zenoss Devs: what's the relation between cacheRRDValue and getRRDNames components methods and remote collectors?

[18-May-2010 16:02:29] <bigegor> Zenoss Devs: how shall i using these methods?

[18-May-2010 16:55:25] forsberg is now known as fOrsberg

[18-May-2010 17:04:03] <zykes-> hmmm, mrayzenoss around ?

[18-May-2010 17:04:15] <mrayzenoss> yeah

[18-May-2010 17:04:15] <zykes-> Wondering if there would be a chance to use "Alert" templates ?

[18-May-2010 17:04:32] <mrayzenoss> what are Alert templates?

[18-May-2010 17:05:09] <zykes-> i mean, instead of re-creating alert settings could one use templates  ?

[18-May-2010 17:05:29] <zykes-> alert rule i mean

[18-May-2010 17:06:11] <mrayzenoss> not sure I follow what you're asking.  Do you want sets of alert rules you can reuse for multiple users?

[18-May-2010 17:07:01] <zykes-> or groups yes, so you don't have to redefine for each time

[18-May-2010 17:07:23] <mrayzenoss> yeah, you can create User Groups, assign the Alert Rules to them and assign multiple users to the group

[18-May-2010 17:08:00] <zykes-> but you can't create what i wanted ? :p

[18-May-2010 17:11:34] <mrayzenoss> well, you can create a group, set your alert rules on it, and add users to it.  Users can belong to multiple groups

[18-May-2010 17:13:17] <zykes-> guess so yeh

[18-May-2010 17:19:48] <zykes-> bigegor: does your wmi script support interface statistics ?

[18-May-2010 17:20:24] <bigegor> zykes: i have no wmi scripts

[18-May-2010 17:20:55] <zykes-> what was yours again then bigegor ?

[18-May-2010 17:21:39] <sergeymasushko> hi 2 all. I have a weird situation. I have 2 device classes called "Server" and "VPNs". I tried to create the same thershold "/FileSystem      /high disk usage" as in the "Server" class but I get the following error: zenhub|User-supplied Python expression (here.getTotalBlocks() * .85) for maximum value caused error: ['usedBlocks_usedBlocks'] The threshold high disk usage in template /zport/dmd/Devices/VPNs/rrdTemplates/FileSystem has been disabled.

[18-May-2010 17:21:48] <bigegor> WMIDataSource and WMIPerf_Windows Zenpack with templates

[18-May-2010 17:22:38] <zykes-> yeh bigegor that I meant, trying to use the interface report but it works for the linux machines but not the wmi one ?

[18-May-2010 17:26:24] <bigegor> can be, i will have a look on interface report for wmi interfaces.

[18-May-2010 17:28:58] <mrayzenoss> without looking at it, I'd assume the WMI interfaces need to alias their data points so they'll show up on the report

[18-May-2010 17:34:00] <bigegor> Matt, why cacheRRDValue methon doesn't work if device monitored by remote collector?

[18-May-2010 17:34:23] <mrayzenoss> I don't know, haven't ever dug into that myself

[18-May-2010 17:37:08] <bigegor> getRRDNames method must return all RRD names or only used by cacheRRDValue?

[18-May-2010 17:47:04] <bigegor> sergeymasushko: try to copy FileSystem template from Server to VPNs

[18-May-2010 17:48:19] <sergeymasushko> bigegor: I tried but I do not see that copy in the "bind" list of the templates..

[18-May-2010 17:49:53] <bigegor> sergeymasushko: its normal you can not see components template in "bind" list of device.

[18-May-2010 17:50:36] <sergeymasushko> bigegor: then how can I apply this template?

[18-May-2010 17:52:14] <bigegor> sergeymasushko: zenoss will bind it to FileSystem component of your device automatically

[18-May-2010 17:53:58] <sergeymasushko> ok thanks. will try to delete currnt template and copy existing one

[18-May-2010 17:55:03] <bigegor> np

[18-May-2010 18:05:52] <Kristopher> For RPNs: If I wanted to divide the value by 10, would it be 10,/ or /,10?

[18-May-2010 18:06:00] <sergeymasushko> bigegor: I tried. the template was copied successfully and then the "high disk usage"  Threshold in the template gone to false again :(

[18-May-2010 18:06:43] <bigegor> Kristopher:10,/

[18-May-2010 18:06:59] <Kristopher> bigegor, thanks!

[18-May-2010 18:10:21] <bigegor> sergeymasushko: but you don't bind FileSystem to device, right?

[18-May-2010 18:10:30] <sergeymasushko> sure

[18-May-2010 18:13:01] <bigegor> your Template listed on Template page of monitored filesystem?

[18-May-2010 18:13:08] <Kristopher> Hmm. I remember someone telling me that OIDs should end in .1, because Zenoss may not like the .0.  If the OID is ...2.4.0, do I change the 0 to 1 or make it ...2.4.0.1?

[18-May-2010 18:13:54] <sergeymasushko> I see the perf graph and the thresold on the graph. I see that value of the /tmp filesystem (from OS tab) is over the threshold... but there is no event due to threshold is disbled and I can not enable it ^(

[18-May-2010 18:15:47] <bigegor> Kristopher: 2.4.0 is right.

[18-May-2010 18:17:27] <sergeymasushko> bigegor: yes, I see the template by this path: /Devices  /VPNs  /mydevice  /os  /tmp (Template tab)

[18-May-2010 18:18:24] <Kristopher> Now it's working, thanks again bigegor.

[18-May-2010 18:19:16] <bigegor> sergeymasushko: try set 'Maximum Value' to some number (calculate it your self)

[18-May-2010 18:20:52] <sergeymasushko> bigegor: it's not a solution... this template is working for different boxes with different filesytem partiotions' sizes...

[18-May-2010 18:23:03] <bigegor> sergeymasushko: i know, it not a solution, but troubleshooting.

[18-May-2010 18:24:26] <sergeymasushko> bigegor: just I do not understand what is special there if it works in one place and doesn't work in the another...

[18-May-2010 18:24:36] * sergeymasushko scratching his head

[18-May-2010 18:26:02] <sergeymasushko> heh... looks like come here too late...

[18-May-2010 18:28:12] <bigegor> sergeymasushko: chech zFileSystemSizeOffset zProperty it used in here.getTotalBlocks() method

[18-May-2010 18:28:19] <bigegor> check

[18-May-2010 18:38:47] <sergeymasushko> bigegor: it is equal to 1.0

[18-May-2010 18:39:26] <bigegor> for both device classes?

[18-May-2010 18:39:47] <sergeymasushko> yes.

[18-May-2010 18:40:39] <sergeymasushko> btw. I'm looking zproperties direcly on particular devices... it has highest priority, right?

[18-May-2010 18:41:39] <bigegor> if you set it on device, yes.

[18-May-2010 18:41:49] <sergeymasushko> ok

[18-May-2010 18:42:50] <bigegor> wait, i have one idea.

[18-May-2010 18:43:39] <sergeymasushko> sure

[18-May-2010 18:46:16] <bigegor> try 'here.totalBlocks * .85' as Maximum value

[18-May-2010 18:49:54] <sergeymasushko> the same thing

[18-May-2010 18:51:41] <bigegor> but if you put number as Maximum value?

[18-May-2010 18:52:01] <bigegor> try it.

[18-May-2010 18:53:29] <sergeymasushko> bigegor: with the number it's OK

[18-May-2010 18:53:49] <bigegor> try: here.getTotalBlocks() * 85 / 100

[18-May-2010 18:56:19] <sergeymasushko> 2010-05-18 19:54:34,294 ERROR zen.MinMaxCheck: User-supplied Python expression (here.getTotalBlocks() * 85 / 100) for maximum value caused error: ['usedBlocks_usedBlocks']

[18-May-2010 18:56:28] <sergeymasushko> still the same

[18-May-2010 18:56:38] <bigegor> simply:here.getTotalBlocks()

[18-May-2010 18:58:07] <Kristopher> (one last thing) I have to convert Celcius to Fahrenheit.  I'm using "1.8,*,32,+" as my RPN.   I also need to divide the Celcius by 10 (because the device automatically multiplies by 10, so zenoss needs to divide by 10).  I put "10,/,1.8,*,32,+" which doesn't seem right...

[18-May-2010 19:01:48] <bigegor> Kristopher:10,/,9,*,5,/,32,+

[18-May-2010 19:01:53] <sergeymasushko> bigegor: the same

[18-May-2010 19:02:23] <bigegor> have you remote collector?

[18-May-2010 19:02:30] <sergeymasushko> no

[18-May-2010 19:03:04] <bigegor> no idea :(

[18-May-2010 19:03:20] <sergeymasushko> I guess I know the root of the issue :)

[18-May-2010 19:03:32] <bigegor> ?

[18-May-2010 19:03:37] <sergeymasushko> freebsd :)

[18-May-2010 19:04:34] <bigegor> try linux ;)

[18-May-2010 19:04:41] <sergeymasushko> zenoss I sinstalled from tarboll... probably something went wrong...

[18-May-2010 19:06:16] <bigegor> try: in zendmd reindex() and then commit()

[18-May-2010 19:06:18] <sergeymasushko> bigegor: it would be nice, but it is not in my privileges to choose the OS. the all our servers are on FreeBSD...

[18-May-2010 19:06:57] <sergeymasushko> bigegor: do you think it will help?

[18-May-2010 19:09:27] <bigegor> here.getTotalBlocks() doesn't return correct value (here.totalBlocks) totalBlocks is property of filesystem. may be something wrong with filesystem instance

[18-May-2010 19:09:53] <sergeymasushko> ]could be

[18-May-2010 19:10:48] <bigegor> you can alse remodel your device

[18-May-2010 19:13:47] <sergeymasushko> it doesn't help =(

[18-May-2010 19:16:29] <bigegor> can you create VPNs device class under Server device class?

[18-May-2010 19:16:57] <sergeymasushko> actually yes...

[18-May-2010 19:17:34] <bigegor> it wil inherit all performance templates from Server device class

[18-May-2010 19:17:35] <sergeymasushko> I'd like to try one more thing...

[18-May-2010 19:17:52] <sergeymasushko> sure I understand it...

[18-May-2010 19:18:30] <sergeymasushko> going to try to restart zenhub...

[18-May-2010 19:18:52] <bigegor> good idea :)

[18-May-2010 19:21:46] <sergeymasushko> even restart doesn't help.

[18-May-2010 19:22:18] <bigegor> reinstall ALL ;)

[18-May-2010 19:22:40] <sergeymasushko> I have no sudo permissions :(

[18-May-2010 19:23:23] <bigegor> later, good luck.

[18-May-2010 19:23:37] <sergeymasushko> heh... will install zenoss at home on my ubuntu server...

[18-May-2010 19:23:50] <sergeymasushko> bigegor: see ya, thanks.

[18-May-2010 19:24:03] <bigegor> np

[19-May-2010 00:00:30] [disconnected at Wed May 19 00:00:30 2010]

[19-May-2010 00:00:30] [connected at Wed May 19 00:00:30 2010]

[19-May-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[19-May-2010 01:23:43] fOrsberg is now known as forsberg

[19-May-2010 02:17:31] <zykes-> Jane_Curry: how much for training ?

[19-May-2010 02:35:35] <Jane_Curry> zykes - for public classes we charge £400 (GB pounds) / day

[19-May-2010 02:36:05] <Jane_Curry> zykes - we have a 1-day FastStart class for those who haven't used Zenoss before

[19-May-2010 02:36:21] <Jane_Curry> .. and a 3-day Event Management class that goes a lot deeper

[19-May-2010 02:36:55] <Jane_Curry> .. details at http://www.skills-1st.co.uk/products/courses/

[19-May-2010 02:37:29] <Jane_Curry> .. or we can do on-site or remote 1-company skills transfer and/or consultancy

[19-May-2010 02:38:04] <Jane_Curry> .. £850/day on-site or £800 remote

[19-May-2010 02:38:45] <Jane_Curry> Also considering offering the event management class as a remote offering using cloud-based machines

[19-May-2010 02:39:04] <Jane_Curry> I would appreciate input from anyoone as to hwther this would be of interest

[19-May-2010 02:40:27] <Wolf> we have zenoss enterprise at our company, I went to the advanced training at zenoss HQ in MD :)

[19-May-2010 02:41:08] <Wolf> got the little zebra on my desk at work hehe

[19-May-2010 02:47:02] <Wolf> sadly we have MySQL performance issues that nobody has been able to resolve yet

[19-May-2010 02:51:24] <zykes-> hmm, i've created an alert with a schedule called "ON DUTY" which should start at 16 pm each day and last 16 hours until 8 the next day, but I get alerts in e-mail which are tagged "ON DUTY" in the work hours ?

[19-May-2010 02:54:45] <fragfutter> Wolf: tuning mysql is voodoo.

[19-May-2010 03:00:09] * Wolf nods

[19-May-2010 03:37:49] <rhettardo> ndb cluster w/ mysql-mmm

[19-May-2010 03:46:47] <zykes-> Jane_Curry: got a bit of time ?

[19-May-2010 05:43:34] <Lantizia> o_O a little zebra

[19-May-2010 06:26:18] <zykes-> Error reading value for "ssCpuUser" on xx (oid .1.3.6.1.4.1.2021.11.9.0 is bad) < anyone got a clue on this ?

[19-May-2010 07:29:36] <MrWGW> hey, on the zenoss virtual appliance, where is the webdir?

[19-May-2010 07:29:38] <MrWGW> or the lighttpd.conf?

[19-May-2010 08:46:04] LarsN_Away is now known as LarsN

[19-May-2010 08:46:19] <LarsN> If I want to change the IP address of a device, without removing and re-adding it.

[19-May-2010 08:46:34] <LarsN> what do I change the URL to in order to manually edit the fields in zope?

[19-May-2010 08:47:00] <LarsN> nevermind, add /manage :)

[19-May-2010 08:51:28] forsberg is now known as fOrsberg

[19-May-2010 09:03:25] <Jane_Curry> Hi zykes...

[19-May-2010 09:03:55] <twm1010> Hrmm.. 3.0alpha UI is nice

[19-May-2010 09:13:11] <gwb2351> anybody use zenoss to monitor/trend statistics on EMC gear (CX4, CX3, etc. etc.)?

[19-May-2010 09:14:22] <gwb2351> Looking for service time, response time, dirty pages, disk utilization, average busy queue length that kind of stuff

[19-May-2010 09:24:21] <rmatte> eugh, that is so lame, Ubuntu 10.04 does not ship with python 2.4, and it's impossible to install it from aptitude

[19-May-2010 09:24:52] <twm1010> why do you need to install it?

[19-May-2010 09:24:53] <fragfutter> rmatte: so we wait for zenoss 3 ;)

[19-May-2010 09:25:08] <twm1010> isn't python in the stack?

[19-May-2010 09:25:16] <rmatte> twm1010: for the python2.4-ldap plugin

[19-May-2010 09:25:23] <twm1010> oh, to do AD auth?

[19-May-2010 09:25:30] <rmatte> LDAP auth

[19-May-2010 09:25:31] <rmatte> but yes

[19-May-2010 09:25:57] <rmatte> I'm working on upgrading all my boxes from ubuntu 8.04 to 10.04

[19-May-2010 09:26:06] <rmatte> this is the only real roadblock I've run in to

[19-May-2010 09:26:10] <twm1010> interesting...

[19-May-2010 09:26:12] <rmatte> I guess I might have to build it from source

[19-May-2010 09:27:40] <Jane_Curry> zykes - need to dip out for about 90 mins - back just after 5:00 (UK time)

[19-May-2010 09:33:09] fOrsberg is now known as forsberg

[19-May-2010 09:43:32] <twm1010> rmatte: is there not a python2.6-ldap plugin?

[19-May-2010 09:44:16] <LarsN> rmatte: I wanted to thank you a couple days ago btw.

[19-May-2010 09:44:31] <LarsN> 1) for writing that fantastic doc on monitoring the status of NFS mounts.

[19-May-2010 09:44:49] <LarsN> and 2) for assisting me in getting that monitoring working.  applying the patches you suggested fixed it right up :)

[19-May-2010 09:44:59] <twm1010> i'll have to take a look at that myself!

[19-May-2010 09:46:31] <LarsN> I've got to ask, is the push towards Zope 2.12 / Python 2.6 driven by the abandoned support for python 2.4 now?

[19-May-2010 09:46:48] <LarsN> FreeBSD doesn't even ship python 2.4 in the ports tree anymore :(

[19-May-2010 09:46:59] <Wolf> I thought Zenoss was moving away from zope

[19-May-2010 09:47:32] <SEJeff> That would be awful difficult as much functionality as they rely on from zope

[19-May-2010 09:47:48] <SEJeff> Zenoss as a django ui would be pretty, but it would be an awful lot of heavy lifting they get for free via zope

[19-May-2010 09:48:22] <Wolf> I donno, seem to remember someone at Zenoss saying that

[19-May-2010 09:50:54] <rmatte> Wolf: it's impossible for Zenoss to move away from Zope

[19-May-2010 09:51:02] <rmatte> the entire project is heavily dependant on it

[19-May-2010 09:51:20] <Wolf> maybe they just ment moving to the newer version

[19-May-2010 09:51:30] <rmatte> yes, they moved to a newer version in 2.5.2

[19-May-2010 09:51:41] <LarsN> I'd like to do some testing with relstorage.

[19-May-2010 09:51:43] <rmatte> and they're going to be moving to another version in 3.0

[19-May-2010 09:51:52] <LarsN> get away from the filestorage based zodb

[19-May-2010 09:51:56] <Wolf> when is 3.0 coming out?

[19-May-2010 09:52:08] <rmatte> 4 or 5 months?

[19-May-2010 09:52:13] <Wolf> ya I donno why they use zodb when mysql is sitting there

[19-May-2010 09:52:17] <rmatte> it's not even in Beta yet

[19-May-2010 09:53:00] <rmatte> honestly, for the way that they use Zope, it's better suited

[19-May-2010 09:53:23] <SEJeff> Wolf, They use it for object storage. rmatte is right there

[19-May-2010 09:53:27] <rmatte> it's a heavily relational object database

[19-May-2010 09:53:37] <LarsN> right.

[19-May-2010 09:53:45] <rmatte> they'd have to duplicate tons of functionality which Zope already supplies

[19-May-2010 09:54:05] <Wolf> I seee

[19-May-2010 09:54:07] <LarsN> relstorage continues to use "zodb" it just pickles the data, and drops it into a relational database.  it's nearly transparent to the zope application server.

[19-May-2010 09:54:49] <Wolf> oh, so relstorage would make MySQL the real backend but Zope would think it is zodb?

[19-May-2010 09:54:52] <LarsN> the advantage being, tools normal to system administrators work for administration.  and you can set up replication of the database across multiple servers :)

[19-May-2010 09:55:03] <LarsN> Wolf: roughly, yes.

[19-May-2010 09:55:13] <Wolf> that  would be nice

[19-May-2010 09:55:15] <LarsN> mysql, oracle, or psql

[19-May-2010 09:55:19] <LarsN> are supported currently.

[19-May-2010 09:55:35] <Wolf> tho I still haven't figured out this one performance issue with MySQL

[19-May-2010 09:55:36] <rmatte> mysql would make the most sense

[19-May-2010 09:55:48] <SEJeff> LarsN, I won't what the speed would be like in relation to zodb

[19-May-2010 09:55:51] <LarsN> for zenoss I agree since it's already there and running.

[19-May-2010 09:56:00] <SEJeff> s/won't/wonder/

[19-May-2010 09:56:40] <LarsN> SEJeff: we've got a pretty large site running with RelStorage.  Performance is at least equal, and in many ways faster than filestorage was.

[19-May-2010 09:56:59] <Wolf> I have a 10 disk RAID 10 consisting of 450GB 15K 6gpbs SAS drives internal to the box its on, and the OS gets great IOPS but MySQL maxes out at about 300

[19-May-2010 09:57:02] <LarsN> SEJeff: however we're on FreeBSD, and there are some differences with UFS2 vs say, ext4 in terms of file performance.

[19-May-2010 09:57:23] <SEJeff> That is true

[19-May-2010 09:57:33] <LarsN> SEJeff: what mysql did allow us to do is, coupled with FreeBSD...

[19-May-2010 09:57:47] <SEJeff> FBSD handles threads better than Linux

[19-May-2010 09:57:59] <rmatte> lol, the last time I used FreeBSD the standard was still nfs

[19-May-2010 09:58:03] <SEJeff> and since mysql is better at thread concurrency than Linux *still*...

[19-May-2010 09:58:05] <LarsN> db01 - 10.0.0.5 & carp address of 10.0.0.6       db02 - 10.0.0.6 & carp address of 10.0.0.6

[19-May-2010 09:58:29] <SEJeff> LarsN, Do you use the OBSD carp daemon or one of the other ones?

[19-May-2010 09:58:35] <LarsN> set up mysql on DB01 as a master, DB02 as a slave.  use ifstated and some scripts to promote the active CARP address to master.

[19-May-2010 09:58:48] <LarsN> FreeBSD has carp in the kernel

[19-May-2010 09:58:53] <SEJeff> Oh shiney

[19-May-2010 09:58:56] <LarsN> just have to recompile it in.

[19-May-2010 09:59:10] <LarsN> set up the zope application servers to talk to mysql on the carp address

[19-May-2010 09:59:18] <rmatte> carp is for doing failover?

[19-May-2010 09:59:22] <LarsN> and we can lose a database server, and drop on average 5 packets.

[19-May-2010 09:59:33] <rmatte> cool

[19-May-2010 10:00:38] <Wolf> anyone have any ideas on MySQL performance? I've done a lot of tweaking and beating on it, even opened a few tickets with zenoss :P

[19-May-2010 10:01:29] <LarsN> Wolf: Best $25 I ever spent was on:  http://tinyurl.com/yxsymg

[19-May-2010 10:01:42] <LarsN> well

[19-May-2010 10:01:45] <Wolf> ya I have that book

[19-May-2010 10:01:52] <LarsN> that and Time Management for Systems Administrators :)

[19-May-2010 10:02:51] <Wolf> well I'm technically the network guy here, but Zenoss is my project so I do everything with it at every level

[19-May-2010 10:03:09] <Wolf> it has its own dedicated uber-box even :P

[19-May-2010 10:03:30] <LarsN> our zenoss installation is on a pair of xen Dom-Us

[19-May-2010 10:03:42] <LarsN> 2 virtual cores, 2gb of ram each.

[19-May-2010 10:04:00] <LarsN> imho sub optimal, but I didn't get to make the decisions.

[19-May-2010 10:04:09] <LarsN> tuning zope helped a lot.  2 threads :)

[19-May-2010 10:04:57] <LarsN> how many devices are you monitoring Wolf ?

[19-May-2010 10:05:49] <Wolf> ours is a 12 core * 2.8Ghz, 64GB RAM, RAID 10 = 10 * 450GB 15K 6g SAS, RAID 1 = 2 * 450GB 15K 6g SAS monitoring about 100 devices....

[19-May-2010 10:06:04] <SEJeff> LarsN, You know, that is from Percona, they also run www.mysqlperformanceblog.com

[19-May-2010 10:06:04] <LarsN> I've only got ~ 150 currently.  across two datacenters.  and another ~50 at a third datacenter with it's own Zenoss monitor.

[19-May-2010 10:06:47] <LarsN> SEJeff: I didn't.  Thanks for the tip :)

[19-May-2010 10:07:03] <SEJeff> LarsN, We brought them in for some mysql consulting. Those guys are gurus.

[19-May-2010 10:07:05] <Wolf> however the zenevents database is around 200GB total.....

[19-May-2010 10:07:26] <SEJeff> That is way overspecc'd Wolf

[19-May-2010 10:07:37] <SEJeff> We have it distributed on 5 boxes using the "distributed collectors" setup

[19-May-2010 10:07:39] <Wolf> yes it is, but I still get crap for IOPS

[19-May-2010 10:08:00] <Wolf> I have 5 distributed collectors on that box (VMware ESXi 4)

[19-May-2010 10:08:02] <SEJeff> Wolf, http://github.com/facebook/flashcache :)

[19-May-2010 10:08:08] <fragfutter> Wolf: did you check for slow queries?

[19-May-2010 10:08:22] <SEJeff> What is the point of distributed collectors on the same physical machine?

[19-May-2010 10:08:25] <SEJeff> Seems pointless

[19-May-2010 10:08:27] <LarsN> the Dom-0 that's running our Zenoss master is a quad core, with 16gb of ram.  the Dom-0 running a second collector is nearly the same specs

[19-May-2010 10:08:43] <Wolf> on a machine with 12 cores? I don't think so

[19-May-2010 10:08:53] <fragfutter> Wolf: Wolf and what is mysqladmin processlist saying?

[19-May-2010 10:09:04] <twm1010> jeebus...

[19-May-2010 10:09:11] <Wolf> fragfutter: the slow query is anything touching the history table :)

[19-May-2010 10:09:20] <twm1010> i'm monitoring ~250 nodes with much less horsepower

[19-May-2010 10:09:28] <LarsN> I'm looking at maybe setting up Zenoss with a zeo cluster.

[19-May-2010 10:09:28] <Wolf> processlist is essentially sleeping most of the time

[19-May-2010 10:09:36] <LarsN> and dropping the zope threads to 1 per zeo

[19-May-2010 10:09:54] <twm1010> i saw performance go up massively when i fixing all the debug messages and removing query of unnecessary OIDs

[19-May-2010 10:10:17] <SEJeff> twm1010, How do you get rid of all of the debug messages?

[19-May-2010 10:10:18] * LarsN isn't sure it's worth the effort at this point.

[19-May-2010 10:10:20] <Wolf> see, Zenoss enterprise support said we need more hardware to fix the issues we were having

[19-May-2010 10:10:21] <SEJeff> The OID is bad crap?

[19-May-2010 10:10:33] <twm1010> SEJeff: it depends on the device right?

[19-May-2010 10:10:33] <Wolf> so we got an insane box :D

[19-May-2010 10:10:35] <SEJeff> Wolf, with 100 servers being monitored?

[19-May-2010 10:10:38] <fragfutter> Wolf: i got better speeds when splitting innodb to seperate files

[19-May-2010 10:10:48] <Wolf> fragfutter: we have seperate files, and partitions!

[19-May-2010 10:10:55] <twm1010> Zenoss tries to automatically use ethernetCSmacD template to monitor any "interface" it doesn't have a named template type for

[19-May-2010 10:11:11] <twm1010> so i created blank templates for the usual suspects that are incompatible, like ds1, ds3, for our T1 and T3 controllers

[19-May-2010 10:11:20] <SEJeff> twm1010, Very clever

[19-May-2010 10:11:34] <SEJeff> Can you write that up in the wiki somewhere?

[19-May-2010 10:11:34] <twm1010> at some point i want to go back and add OID's that are valid, like clocking errors and whatnot

[19-May-2010 10:11:40] <Wolf> Here is the thing, our company must monitor EVERY login/logout/authentication/authorization event from EVERY system and store it..... ;)

[19-May-2010 10:11:55] <twm1010> excluding interfaces you don't want to monitor is very important to

[19-May-2010 10:11:56] <LarsN> twm1010: actually that's a brilliant idea.  it'd be very handy to have that written up in a wiki.

[19-May-2010 10:11:56] <fragfutter> Wolf: so lots of writes

[19-May-2010 10:12:00] <SEJeff> Seems like zenoss is the wrong tool for that job

[19-May-2010 10:12:02] <SEJeff> yuck

[19-May-2010 10:12:10] <twm1010> by using the zproperty to set ignore types and names

[19-May-2010 10:12:12] <Wolf> fragfutter: true, but the writes are not the issue

[19-May-2010 10:12:24] <fragfutter> Wolf: they invalidate all caches

[19-May-2010 10:12:36] <LarsN> twm1010: with a "why we did this" "performance before" "performance after" etc :)

[19-May-2010 10:12:54] <twm1010> think about it, if you had say a voice gateway with 6 voice T1's connected, all those virtual VOIP interfaces +  the PRI channels you might have a device with 500+ interfaces when only really 7 need to be monitored

[19-May-2010 10:13:07] <twm1010> if all those interfaces assume the standard ethernetCSmacD template, which has like 6 oid's

[19-May-2010 10:13:16] <twm1010> that's 3000 oid's you're trying to query, that are bad :)

[19-May-2010 10:13:24] <twm1010> that keeps zenperfsnmp VERY busy now doesn't it

[19-May-2010 10:13:26] <SEJeff> twm1010, That is a problem we have as well. We have pretty large switch stacks

[19-May-2010 10:13:29] <Wolf> I have an 8GB innodb buffer pool, we don't get 8GB of writes in a day, actually not even in a few days with the way I changed how it handles logging those events

[19-May-2010 10:13:29] * LarsN agrees with twm1010 :)

[19-May-2010 10:13:37] <SEJeff> We just used distributed collectors and horizontally scaled out

[19-May-2010 10:13:43] <twm1010> now in our case mmultiply that times 5 voice gateways that really have 10 PRI's, etc.... etc.... etc....

[19-May-2010 10:13:46] <twm1010> its all about tuning

[19-May-2010 10:13:53] <SEJeff> The rrd writes were murdering our box even on really fast hardware

[19-May-2010 10:13:59] <LarsN> SEJeff: we're running A distributed collector for a remote site.

[19-May-2010 10:14:07] <LarsN> figure not monitoring over a Point to Point makes sense.

[19-May-2010 10:14:21] <SEJeff> LarsN, Thats our problem. Lots of remote sites all over the world :/

[19-May-2010 10:14:28] <SEJeff> Going over the wan seems silly for that

[19-May-2010 10:14:59] <LarsN> the problem I'm running into now is on machines running FreeBSD Jails, when the periodic processes run, the jails server slows down enough to not respond to the snmp queries.

[19-May-2010 10:15:22] <LarsN> triggering ~ 50 alerts, one for each Virtual Private Server on the host.

[19-May-2010 10:15:34] <SEJeff> LarsN, Can you do the fbsd equivalent of changing the scheduling priority of the snmp daemon?

[19-May-2010 10:15:40] <LarsN> yes.

[19-May-2010 10:15:43] <LarsN> or,

[19-May-2010 10:15:49] <LarsN> we stagger when periodic runs :)

[19-May-2010 10:16:01] <SEJeff> Both good ideas

[19-May-2010 10:16:15] * LarsN will brb

[19-May-2010 10:16:30] <fragfutter> Wolf: even with you raid 10 did you check if the mysql is disk or cpu bound?

[19-May-2010 10:19:02] <rmatte> ubuntu removing python 2.4 and 2.5 from latest version == epic fail

[19-May-2010 10:20:23] <fragfutter> rmatte: yes ubuntu moves into a desktop direction. not caring about longtail but stuffing features (ubuntu one everywhere, wtf)

[19-May-2010 10:20:52] <fragfutter> rmatte: but i think there were some ports of 2.4 and 2.5 in a ppa.

[19-May-2010 10:21:47] <LarsN> rmatte: everyone is removing python 2.4

[19-May-2010 10:21:53] <LarsN> rmatte: it's officially unsupported.

[19-May-2010 10:23:25] <SEJeff> Ubuntu isn't an enterprise distribution. Those that think it is have good jokes.

[19-May-2010 10:23:47] <SEJeff> This desktop is Ubuntu, but not any servers of mine. Thats just silly

[19-May-2010 10:23:48] * LarsN grins.

[19-May-2010 10:23:55] <Wolf> fragfutter: I think its CPU since I see iowait

[19-May-2010 10:24:10] <LarsN> distro wars are like arguing religion.  Nobody's going to convince me anything with a linux kernel is enterprise ready.

[19-May-2010 10:24:14] * LarsN is a FreeBSD guy :)

[19-May-2010 10:24:21] * LarsN also kids.

[19-May-2010 10:24:39] <SEJeff> Look at the engineering / qa of Redhat or Novell vs Canonical. It isn't even in the same league. For real "enterprise-y" stuff, you need something engineering heavy

[19-May-2010 10:24:53] <LarsN> ie, FreeBSD ;)

[19-May-2010 10:24:56] <SEJeff> And FBSD has a zillion years of development already

[19-May-2010 10:24:57] <SEJeff> So yes

[19-May-2010 10:25:07] <Wolf> so CentOS counts as well?

[19-May-2010 10:25:09] <LarsN> technically, our zenoss installation is on CentOS though

[19-May-2010 10:25:18] <LarsN> development of CentOS is awesome....

[19-May-2010 10:25:21] <SEJeff> Yes, because it is a verbatim rebuild of RHEL

[19-May-2010 10:25:39] <SEJeff> But RHEL5 is getting really long in the tooth. Hopefully RHEL6 / centos6 will be out soon

[19-May-2010 10:25:40] <Wolf> yep thats what we use for zenoss here, no point in paying monthly for "support"

[19-May-2010 10:25:42] <LarsN> RHEL releases.  CentOS grabs sRPMS, re-brands it, compiles.

[19-May-2010 10:25:50] <SEJeff> Yup

[19-May-2010 10:26:34] <Wolf> but I prefer Ubuntu myself, I cut my teeth on Slackware (ages ago) and moved to Debian, then Ubuntu ... natural progression it felt to me :P

[19-May-2010 10:26:51] <SEJeff> my desktops are all ubuntu

[19-May-2010 10:27:06] <Wolf> my new little theforest.us play server is Ubuntu 10.04

[19-May-2010 10:27:21] * LarsN is running slackware on my desktops.

[19-May-2010 10:27:27] <LarsN> well, 2 laptops, and a desktop.

[19-May-2010 10:27:35] <Wolf> whats slackware use for package management nowadays?

[19-May-2010 10:27:39] <LarsN> heh

[19-May-2010 10:27:43] <LarsN> package management.

[19-May-2010 10:27:47] <LarsN> you're funny :)

[19-May-2010 10:27:50] <Wolf> ya, thats what I thought :P

[19-May-2010 10:28:07] <LarsN> in some cases, it actually solves a lot of problems.

[19-May-2010 10:28:09] <Wolf> I don't have time for that model anymore, which is why I moved to a system with apt

[19-May-2010 10:28:26] <LarsN> I was an ubuntu desktop user since 4.10 iirc

[19-May-2010 10:28:26] <Wolf> true, I just don't have time to deal with that sort of system maintenance

[19-May-2010 10:28:44] <LarsN> in server land I still use FreeBSD.  I rather like ports.

[19-May-2010 10:29:16] <LarsN> really just about everything has a purpose.

[19-May-2010 10:29:41] <SEJeff> Use the right tool for the job. If you know freebsd well then it is likely the right tool for the job.

[19-May-2010 10:29:52] <Wolf> I like FreeBSD, used it in the past a few times, but I "grew up" on Linux so its all warm and fuzzy familiar

[19-May-2010 10:30:11] <rmatte> I'm the exact opposite, grew up on FreeBSD, but use Linux now

[19-May-2010 10:30:16] <SEJeff> lxc is really nice and a good "competitor" with jails now that some enterprise distros are shipping kernels with lxc support

[19-May-2010 10:30:19] <Wolf> almost 12 years now, damn....

[19-May-2010 10:30:21] <LarsN> I grew up on Solaris.

[19-May-2010 10:30:24] <SEJeff> OpenBSD

[19-May-2010 10:30:25] <LarsN> just sayin;

[19-May-2010 10:31:01] <LarsN> I had an OpenBSD mail server.  Solid little box.

[19-May-2010 10:31:18] <LarsN> MySQL backed Postfix

[19-May-2010 10:31:40] <Wolf> neat

[19-May-2010 10:32:32] <Wolf> if I had to spin up a server that needed to run a lot and face the world I'd consider FBSD

[19-May-2010 10:32:48] <LarsN> heading to lunch

[19-May-2010 10:32:53] <LarsN> have a release in an hour and a half

[19-May-2010 10:32:54] <LarsN> :)

[19-May-2010 10:33:36] <Wolf> lunch is a good idea....

[19-May-2010 10:36:11] <zykes-> RHEL6 rel date june / july

[19-May-2010 10:54:48] <rmatte> eugh, trying to get python-ldap working with Zenoss in ubuntu 10.04 is a pain

[19-May-2010 11:01:21] <SEJeff> rmatte, Just use setup.py install as user zenoss so it installs in the zenoss specific python

[19-May-2010 11:14:31] <rmatte> SEJeff: hmmm

[19-May-2010 11:14:48] <rmatte> setup.py install python-ldap?

[19-May-2010 11:14:55] <SEJeff> yes

[19-May-2010 11:15:33] <SEJeff> zenoss has their bundled python that zenoss uses. You'll need to install the module in that version of python. Hopefully with zenoss 3.0 they can use the os install of python. Bundling it is sad.

[19-May-2010 11:16:20] <cgibbons> It was mandatory given the limitation from the version of Zope that was being used.

[19-May-2010 11:16:44] <rmatte> SEJeff: yeh I know, before I was just linking the system files back to that location, per the guide

[19-May-2010 11:17:25] <SEJeff> rmatte, But that probably won't work so well with Lucid's python 2.6 and zenoss's 2.4

[19-May-2010 11:18:11] <rmatte> yeh, but what I might do is just copy the old files back in to where they were

[19-May-2010 11:18:18] <rmatte> it's not going to affect anything

[19-May-2010 11:18:31] <rmatte> which setup.py are you talking about by the way?

[19-May-2010 11:18:35] <rmatte> there are tons of them

[19-May-2010 11:18:51] <SEJeff> rmatte, Download python-ldap from upstream pypy

[19-May-2010 11:19:01] <SEJeff> untar the file, and run setup.py in the bundle

[19-May-2010 11:19:03] <SEJeff> It isn't hard

[19-May-2010 11:19:07] <rmatte> ah ok

[19-May-2010 11:19:50] <SEJeff> rmatte, Get them from here: http://pypi.python.org/pypi/python-ldap/

[19-May-2010 11:20:15] <SEJeff> Just wget the file as user zenoss on the server, then untar it and run setup.py install. That should take care of everything for you so long as you do it as user zenoss.

[19-May-2010 11:20:18] <rmatte> will do

[19-May-2010 11:22:28] <rmatte> build failed

[19-May-2010 11:33:51] <LarsN> rmatte: the other option is...

[19-May-2010 11:34:02] <LarsN> this is a prime opportunity to move to the alpha.

[19-May-2010 11:34:21] <LarsN> a former Operations Manager of mine used to say:  "You know what they say........  God hates a coward."

[19-May-2010 11:34:24] <LarsN> :)

[19-May-2010 11:35:28] <rmatte> lol

[19-May-2010 11:35:36] <rmatte> think I'll pass on that one :P

[19-May-2010 11:36:06] <rmatte> aha

[19-May-2010 11:36:09] <rmatte> got it working

[19-May-2010 11:40:50] <SEJeff> rmatte, What was the magic sequence?

[19-May-2010 11:43:35] <rmatte> I just copied over the files from my older distribution

[19-May-2010 11:43:46] <rmatte> and then recompiled some of the python files

[19-May-2010 11:44:05] <rmatte> works fine now

[19-May-2010 12:46:42] <msg4al> does anyone here have experience using zenoss in a remote environment?

[19-May-2010 12:46:56] <msg4al> i just need a little guidance to get my solution going

[19-May-2010 12:47:02] <msg4al> please help

[19-May-2010 12:47:05] <rmatte> msg4al: define remote environment

[19-May-2010 12:47:55] <msg4al> rmatte: zenoss is behind a firewall and network to be monitored is at a remote location behind another firewall (VPN is not a option)

[19-May-2010 12:48:23] <msg4al> NRPE is one option but i dont want to reinvent the wheel, looking for a proven method

[19-May-2010 12:48:34] <rmatte> the method I use is VPN

[19-May-2010 12:48:36] <rmatte> so no idea

[19-May-2010 12:48:47] <msg4al> vpn can work for 2 or 3 networks

[19-May-2010 12:48:54] <rmatte> Zenoss is an agentless solution, so it needs some way of hitting those hosts

[19-May-2010 12:48:55] <msg4al> but i plan to install it for about 20

[19-May-2010 12:49:11] <Wolf> install a distributed collector and connect back via SSH

[19-May-2010 12:49:11] <msg4al> NRPE gives you that ability (Nagios plugin)

[19-May-2010 12:50:18] <msg4al> Wolf: can you elaborate? is a dedicated machine required for the distributed collector? do you have any links for the setup?

[19-May-2010 12:50:33] <rmatte> yes, a dedicated machine is required for the collector

[19-May-2010 12:50:50] <Wolf> unless you are running CentOS and have a beefy box

[19-May-2010 12:50:59] <Wolf> or RHEL5

[19-May-2010 12:51:00] <rmatte> The Collector Tool and Distributed Collector ZenPacks assist with distributed collector setuip

[19-May-2010 12:51:02] <rmatte> setup*

[19-May-2010 12:51:39] <rmatte> (use one or the other, not both)

[19-May-2010 12:54:04] <Wolf> ya what he said :P

[19-May-2010 12:56:51] <rmatte> I'

[19-May-2010 12:57:00] <rmatte> I've never heard of NRPE working with Zenoss

[19-May-2010 12:57:06] <rmatte> though it should in theory with some dev work

[19-May-2010 12:57:33] <rmatte> I hate agent based monitoring though

[19-May-2010 12:59:16] <bigegor_> hi, why as "Render URL" we using "http://masterserver:8090/remotecollector" and not "http://remotecollector:8091"

[19-May-2010 12:59:44] <mrayzenoss> rmatte: actually I've heard nrpe works just fine

[19-May-2010 12:59:54] <rmatte> oh?

[19-May-2010 13:00:00] <rmatte> someone should wiki a guide for it's use

[19-May-2010 13:00:07] <rmatte> and make some ZenPack for it

[19-May-2010 13:00:50] <rmatte> bigegor_: because it's connection to the local renderserver and telling it to point to the remote renderserver

[19-May-2010 13:00:55] <rmatte> connecting*

[19-May-2010 13:01:08] <rmatte> bigegor_: that's the correct way of doing it

[19-May-2010 13:01:12] <mrayzenoss> rmatte: message/45960#45960

[19-May-2010 13:01:57] <rmatte> ah

[19-May-2010 13:02:14] <bigegor_> but it doesn't realy work. I have problem to get cacheRRDValue from http://masterserver:8090/remotecollector

[19-May-2010 13:02:32] <rmatte> bigegor_: hmmm

[19-May-2010 13:03:00] <rmatte> bigegor_: but you can use getRRDValue, right?

[19-May-2010 13:03:14] <rmatte> If so, it sounds more like a limitation of zenrender than anything

[19-May-2010 13:03:19] <bigegor_> but if i use http://remotecollector:8091 all works (example HPEVAMon)

[19-May-2010 13:03:22] <rmatte> only being able to remotely provide getRRDValue rather than cache

[19-May-2010 13:03:29] <rmatte> hmmmm

[19-May-2010 13:03:33] <bigegor_> rmatte: the same problem

[19-May-2010 13:03:33] * Simon4 listens in to this

[19-May-2010 13:03:45] <rmatte> well, if it works like that then I guess that's the way to go

[19-May-2010 13:03:56] <rmatte> I was just following that original guide

[19-May-2010 13:04:17] <rmatte> I figured they knew what they were talking about

[19-May-2010 13:04:26] <msg4al> ok the remote collector service is less than ideal for small environments

[19-May-2010 13:05:00] <bigegor_> it hang on xmlrpclib request for RRD summary

[19-May-2010 13:05:02] <msg4al> nrpe is a good alternative as long as it works well with zenoss

[19-May-2010 13:05:03] <rmatte> I still don't see why tunnels aren't an option

[19-May-2010 13:05:11] <rmatte> we have tunnels going to some sites that only have like 3 or 4 devices

[19-May-2010 13:05:28] <rmatte> we have about 25 tunnels total

[19-May-2010 13:05:47] <rmatte> we'll have quite a few more eventually

[19-May-2010 13:06:04] <rmatte> as long as you don't have IP address overlapping

[19-May-2010 13:06:23] <msg4al> rmatte: I still don't see why tunnels aren't an option.........overlapping ip's

[19-May-2010 13:06:42] <msg4al> most of our network are on the 192.168.10.0 subnet

[19-May-2010 13:06:55] <rmatte> ah, that's just bad network planning lol, but ok

[19-May-2010 13:07:11] <rmatte> then I guess you're stuck

[19-May-2010 13:07:27] <rmatte> unless you were to do some form of natting over the tunnels

[19-May-2010 13:07:38] <msg4al> that can be a little confusing

[19-May-2010 13:07:41] <msg4al> thats why

[19-May-2010 13:07:55] <msg4al> i was thinking about nrpe which in essence installs a agent

[19-May-2010 13:08:02] <rmatte> yes, it does

[19-May-2010 13:08:08] <rmatte> but then you're back to agent monitoring

[19-May-2010 13:08:24] <msg4al> im just not sure how well zenoss would interpret the results obtained from nrpe

[19-May-2010 13:08:29] <rmatte> what if you need to monitor network equipment as well down the road?

[19-May-2010 13:08:35] <rmatte> NRPE is fine for servers

[19-May-2010 13:08:57] <twm10101> si there a way to get zenoss to ungroup events?

[19-May-2010 13:08:59] <msg4al> nagios has snmp plugins, i think

[19-May-2010 13:09:14] <msg4al> using nrpe i can use those plugins

[19-May-2010 13:09:18] <rmatte> twm10101: no, that's static behaviour

[19-May-2010 13:09:27] <rmatte> twm10101: it's meant to save database space

[19-May-2010 13:09:37] <rmatte> since it just increments the counter instead of adding a new entry each time

[19-May-2010 13:09:46] <msg4al> again, my concern is zenoss will be able to obtain the data but may not read it right.

[19-May-2010 13:10:20] <rmatte> msg4al: nagios plugins works fine with Zenoss, I doubt that'll be an issue provided you configure it correctly

[19-May-2010 13:10:41] <rmatte> according to mrayzenoss people have gotten it working

[19-May-2010 13:10:53] <rmatte> message/45960#45960

[19-May-2010 13:11:08] <mrayzenoss> I just searched for it a few times on the community.zenoss.org site

[19-May-2010 13:12:06] <msg4al> ok so in theory...once i obtain the info from host using nrpe..zenoss will know how to interpret the data?

[19-May-2010 13:12:24] <rmatte> zenoss understands nagios formatted output

[19-May-2010 13:12:25] <rmatte> so yes

[19-May-2010 13:12:39] <rmatte> in theory

[19-May-2010 13:13:35] <mrayzenoss> msg4al: but Zenoss won't get anything other than those performance metrics

[19-May-2010 13:13:36] <msg4al> sweet....integrating zenoss and nagios can provide you with the best of both worlds. i am very interested to see how well they work togather

[19-May-2010 13:13:43] <mrayzenoss> msg4al: no modeling of the device

[19-May-2010 13:13:55] <rmatte> yeh, so not disk stats for instance

[19-May-2010 13:14:00] <rmatte> no*

[19-May-2010 13:14:05] <rmatte> or interface stats

[19-May-2010 13:14:35] <rmatte> which is why agentless monitoring via NRPE isn't that great

[19-May-2010 13:14:56] <mrayzenoss> it's really only handy if there's something NRPE was doing that you can't recreate

[19-May-2010 13:14:56] <rmatte> I still think tunnels with natting is the optimum solution

[19-May-2010 13:15:15] <rmatte> just make sure the devices are named correctly and that the natting makes sense, follows location numbers or something easy to remember

[19-May-2010 13:15:59] <msg4al> ok.....rmatte: can you give me a simple natting example?

[19-May-2010 13:16:22] <rmatte> natting varies based on where you do it, we do it on a router, you can do it right on the linux box

[19-May-2010 13:16:48] <rmatte> all natting is is telling the system that 192.168.10.1 at the remote site correspond to 10.0.10.1 at the local box

[19-May-2010 13:16:49] <rmatte> for example

[19-May-2010 13:16:55] <rmatte> google around, there's tons of examples

[19-May-2010 13:17:27] <msg4al> ok but how does that address the overlap issue? would i need to nat each ip?

[19-May-2010 13:17:39] <rmatte> you can nat a range

[19-May-2010 13:17:50] <rmatte> so you'd nat one site to a range, another site to a different range, and so on

[19-May-2010 13:17:58] <rmatte> then use those ranges in Zenoss instead of the actual IPs

[19-May-2010 13:18:04] <rmatte> which takes care of your overlap issue

[19-May-2010 13:18:39] <rmatte> since the server knows that when sending to this IP range, direct the data out whichever tunnel corresponds to that range, then translate it to/from the actual range for that site

[19-May-2010 13:19:08] <rmatte> I'm sure there are lots of guides on google for doing that sort of a setup

[19-May-2010 13:19:37] <mistich> good day

[19-May-2010 13:19:43] <rmatte> g'day

[19-May-2010 13:20:00] <msg4al> ok that makes sense

[19-May-2010 13:20:08] <mistich> what daemon runs event commands

[19-May-2010 13:20:21] <rmatte> zenactions

[19-May-2010 13:20:38] <rmatte> unless you're talking about command datasources

[19-May-2010 13:20:42] <rmatte> then that would be zencommand

[19-May-2010 13:20:53] <mistich> ok

[19-May-2010 13:20:53] <rmatte> but the commands section under event manager is zenactions

[19-May-2010 13:21:34] <mistich> no I see it had my filter wrong

[19-May-2010 13:22:27] <rmatte> ah

[19-May-2010 13:26:18] <msg4al> rmatte: how many different sites do you have vpn;s setup for? is it bandwidth intensive?

[19-May-2010 13:30:00] <msg4al> also, if vpn goes down you probably get flooded with alerts

[19-May-2010 13:30:51] <Simon4> vpn down would give you a heartbeat failure, not much else

[19-May-2010 13:31:02] <Simon4> (if I'm reading rmatte's config correctly)

[19-May-2010 13:32:51] <msg4al> hearbeat failure for all the hosts that are being monitored

[19-May-2010 13:33:05] <Simon4> no, just the daemons on the collector at the end of the vpn

[19-May-2010 13:33:21] <Simon4> I believe he has a collector at the end of each vpn

[19-May-2010 13:33:25] <msg4al> no collector, we are talking about natthing

[19-May-2010 13:33:34] <msg4al> natting*** f

[19-May-2010 13:33:40] <msg4al> oops natting*

[19-May-2010 13:33:47] <Simon4> k, I may be miles out then, sorry

[19-May-2010 13:33:55] <rmatte> if a tunnel dropped you'd get alerted for every device at the end of that tunnel

[19-May-2010 13:34:09] <rmatte> and no, I don't have a collector for each tunnel, I have a seperate Zenoss install for each tunnel

[19-May-2010 13:34:10] <msg4al> no problem simon, thanks for the input

[19-May-2010 13:34:16] <rmatte> but only because I monitor different clients

[19-May-2010 13:34:22] <rmatte> and they each need their own server for security reasons

[19-May-2010 13:34:34] <rmatte> we do have some servers with multiple clients on them

[19-May-2010 13:35:16] <msg4al> how is the bandwidth utlization? have you come across any issues that were bandwidth related?

[19-May-2010 13:35:30] <rmatte> not really, tunnels are pretty lightweight

[19-May-2010 13:35:55] <rmatte> it obviously depends on the connection that you're going out over and the connection that you're connecting to

[19-May-2010 13:36:05] <msg4al> ok when you say tunnel you are referring to a site to site vpn connection right?

[19-May-2010 13:36:07] <rmatte> if you're connecting via a 56k modem then you'll get 56k performance obviously

[19-May-2010 13:36:13] <rmatte> correct

[19-May-2010 13:37:18] <msg4al> ok cool. thanks rmatte. i'll try that in my lab and see how it goes. i would also like to test nrpe but it seems like the natting will work better for me

[19-May-2010 13:42:05] <rmatte> yeh, you'll get more data from the natting

[19-May-2010 13:43:57] <msg4al> do you know which ports need to be opened?

[19-May-2010 13:47:32] <rmatte> there's a pretty big range

[19-May-2010 13:47:36] <rmatte> I don't have an exact list

[19-May-2010 13:47:51] <rmatte> I just open everything over the tunnel

[19-May-2010 13:47:59] <rmatte> it's encrypted and secure anyways

[19-May-2010 13:49:05] <msg4al> ok, i'll check on the specifics. i dont want to just open everything since i will be using the same server for multiple clients

[19-May-2010 13:54:39] <Kristopher> Do you guys know why I'd be getting the dmd errors when loading the 'OS' tab on a device? message/48986  Thanks.

[19-May-2010 14:11:38] <rmatte> Kristopher: have you tried a reindex?

[19-May-2010 14:12:54] <Kristopher> Yep.  reindex() and commit() then restarted zenprocess.

[19-May-2010 14:13:27] <rmatte> try restarting all of zenoss

[19-May-2010 14:13:32] <rmatte> instead of just zenprocess

[19-May-2010 14:13:38] <Kristopher> As in reboot the entire server, OS and all?

[19-May-2010 14:13:42] <rmatte> no

[19-May-2010 14:13:49] <rmatte> as in zenoss stop/zenoss start

[19-May-2010 14:14:32] <twm10101> or if you did stack install, execute service zenoss-stack restart

[19-May-2010 14:15:35] LarsN is now known as LarsN_Away

[19-May-2010 14:16:12] <Kristopher> zenoss stop/zenoss start completed but didn't rectify.

[19-May-2010 14:17:56] <Kristopher> uptime

[19-May-2010 14:20:59] <rmatte> does it do that for every device?

[19-May-2010 14:22:36] <Kristopher> No, only a select few.

[19-May-2010 14:23:31] <Kristopher> Same model, SNMP version, etc as an exactly identical device...one shows the dmd error, the other goes to the OS page.'

[19-May-2010 14:35:07] <coofamani> zenoss ever planning on going to syslog format for logs?

[19-May-2010 14:36:21] <mrayzenoss> coofamani: output our logs via syslog?

[19-May-2010 14:36:32] <mrayzenoss> or adhere to a format?

[19-May-2010 14:36:51] <mrayzenoss> rmatte: this should brighten your day: http://dev.zenoss.org/trac/ticket/5106

[19-May-2010 14:40:35] <coofamani> mrayzenoss: syslog RFC format, for consumption by other apps like splunk/etc

[19-May-2010 14:41:36] <zykes-> splunk is useful for ?

[19-May-2010 14:42:08] <SEJeff> zykes-, Doing real time log analysis and visualization

[19-May-2010 14:42:14] <coofamani> correlation and investigation, etc...

[19-May-2010 14:42:37] <coofamani> zenoss is one of the few apps I manage that doesnt produce syslog format logs, so it's a one-off...

[19-May-2010 14:42:37] <mrayzenoss> coofamani: I don't see anything related to that in the open tickets, please open an enhancement request

[19-May-2010 14:42:42] <coofamani> kk

[19-May-2010 14:43:10] <SEJeff> coofamani, But you should be able to teach splunk the zenoss log format

[19-May-2010 14:43:29] <coofamani> if your running agents yeah

[19-May-2010 14:43:50] <coofamani> I want everything relayed and consolidated via syslog standard, keeps my options open

[19-May-2010 14:44:04] <SEJeff> And I'll guess you aggregate it all with syslog-ng

[19-May-2010 14:44:11] <coofamani> yup :)

[19-May-2010 14:44:15] <mrayzenoss> SEJeff: showing my own ignorance, I assume our logging follows a standard?

[19-May-2010 14:44:24] <SEJeff> coofamani, We do the same :)

[19-May-2010 14:44:26] <mrayzenoss> or is at least consistent

[19-May-2010 14:44:32] <coofamani> your logging follows crazy Matt :)

[19-May-2010 14:44:43] <SEJeff> mrayzenoss, Well syslog compliant is the best way for parsing by non-zenoss tools

[19-May-2010 14:44:53] <mrayzenoss> yeah, I understand that

[19-May-2010 14:45:21] <mrayzenoss> I meant, do the various logs coming out of Zenoss all follow the same format, even if it's not the syslog standard?

[19-May-2010 14:45:43] <mrayzenoss> that'll make it easier to fix

[19-May-2010 14:45:53] <coofamani> havent looked that closely

[19-May-2010 14:46:06] <mrayzenoss> me neither :)

[19-May-2010 14:46:20] <coofamani> the biggest problem I see is that zenoss outputs multi-line 'transactions' and that doesnt play well with most syslog based things

[19-May-2010 14:48:51] <SEJeff> Yeah that won't work

[19-May-2010 14:49:04] <SEJeff> But a python traceback on 1 line would really suck to read

[19-May-2010 14:52:02] <mrayzenoss> sounds like Perl

[19-May-2010 14:52:08] <coofamani> agreed. Im hoping the log app can read it and make it pretty for general consumption

[19-May-2010 17:14:09] forsberg is now known as fOrsberg

[19-May-2010 18:10:09] rhettardo_ is now known as rhettardo

[20-May-2010 00:00:30] [disconnected at Thu May 20 00:00:30 2010]

[20-May-2010 00:00:30] [connected at Thu May 20 00:00:30 2010]

[20-May-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[20-May-2010 01:59:15] fOrsberg is now known as forsberg

[20-May-2010 05:52:20] <baffle> Anyone have any idea what is happening here? http://paste2.org/p/840093

[20-May-2010 05:55:16] <baffle> Seems to happen a few other places in the /Processes tree as well, especially if there is more than "a handful" of process instances.

[20-May-2010 06:50:23] <JesperS> Is there a timeline on when 3.0 final is expected to be released? Maybe just a estimate?   I'm asking because we are going to reinstall our 2.5.2 in a (fairly) short while, and was thinking about skipping the reinstall and go directly to 3.0 if it's just around the corner.

[20-May-2010 06:50:59] <fragfutter> later

[20-May-2010 07:32:25] Simon4_ is now known as Simon4

[20-May-2010 07:39:57] Simon4_ is now known as Simon4

[20-May-2010 07:59:39] <jimbosyn> anyone in the channel?

[20-May-2010 08:00:44] <JesperS> nope, all 63 are ghosts!

[20-May-2010 08:01:04] <jimbosyn> yeah, It's just real quite :)

[20-May-2010 08:01:18] <jimbosyn> Got a question regarding zeniping.

[20-May-2010 08:02:47] <jimbosyn> I know that zenperfsnmp stores performance data in rrd files

[20-May-2010 08:02:54] <jimbosyn> Where does zenping store it's data?

[20-May-2010 08:03:17] <sergeymasushko> zenping doesn't store this data

[20-May-2010 08:03:30] <sergeymasushko> you should install a zenpack for it

[20-May-2010 08:03:43] <sergeymasushko> e.g. nagios ping or something similar

[20-May-2010 08:04:08] <jimbosyn> Yeah, I'm using fping for some devices, but I have around 3k devices that I only want to ping

[20-May-2010 08:04:35] <sergeymasushko> create a special device class

[20-May-2010 08:04:45] <jimbosyn> I'm just trying to figure out how  many devices can be served by a single collector if all I'm monitoring is ping

[20-May-2010 08:04:48] <sergeymasushko> move devices there and aplly the template for the class

[20-May-2010 08:05:34] <jimbosyn> sergeymasushko: so for ping only devices, what would be the bottleneck

[20-May-2010 08:06:08] <jimbosyn> for snmp devices, it's disk speed since all data is written out to rrds

[20-May-2010 08:06:19] <sergeymasushko> re collector - dunno I have no experience with such installations... I monitor about 300-400 devices...

[20-May-2010 08:08:17] <jimbosyn> sergeymasushko: ok.  so would it be safe to say that zenping only records information if the ping fails?

[20-May-2010 08:08:40] <jimbosyn> And that that data is stored in the ZEO db?

[20-May-2010 08:10:52] <jimbosyn> I'm currently getting around 900 snmp devices per collector.  I'm trying to figure out how many ping only devices I can fit on each collector.

[20-May-2010 08:50:55] <jimbosyn> Can anyone tell me if zenping writes data to disk for each polling cycle?

[20-May-2010 09:02:00] <JesperS> <sergeymasushko> zenping doesn't store this data    <- Looks like a no to me. I suppose it only keeps state in ram?

[20-May-2010 09:04:13] <jimbosyn> Ok guys, thanks for the help.  I'll do some testing to see how many devices can comfortable fit on a single collector for ping only.

[20-May-2010 09:09:40] LarsN_Away is now known as LarsN

[20-May-2010 09:09:55] <LarsN> rmatte: you around this morning?

[20-May-2010 09:16:48] <mrayzenoss> jimbosyn: I've heard of a single box doing 8K SNMP, so I'm guessing ping is even bigger

[20-May-2010 09:17:32] <fragfutter> i would think it also depence what you are collecting for each device.

[20-May-2010 09:18:23] * fragfutter . o O (depence? need more coffee...)

[20-May-2010 09:19:36] <jimbosyn> fragfutter: I'm looking to do around 3000 ping only devices

[20-May-2010 09:19:49] <LarsN> ok, that's annoying.

[20-May-2010 09:20:05] <LarsN> add a device.  Choose pre-production from the drop down at the time I'm adding it.

[20-May-2010 09:20:43] <LarsN> first polling interval after....  triggered events with alerts.  Zenoss knows I didn't mean to put it in pre-production, and went ahead and promoted the device to production for me.  :)

[20-May-2010 09:21:35] <fragfutter> jimbosyn: why do you want to use zenoss? Is there any feature you need? If you only ping and don't collect any performance data i would consider nagios.

[20-May-2010 09:22:37] <jimbosyn> I'm also monitoring Cisco routers and wanted to do some syslog colleciton as well

[20-May-2010 09:22:41] <fragfutter> jimbosyn: in nagios i had 5k services on 100 machines. So simply pinging 3k machines should be no problem.

[20-May-2010 09:23:07] <fragfutter> jimbosyn: ah. ok.

[20-May-2010 09:23:32] <jimbosyn> I have about 2500 Cisco routers at this point on distributed collectors

[20-May-2010 09:23:47] <jimbosyn> I can do around 850-900 per collector for SNMP Devices

[20-May-2010 09:24:03] <jimbosyn> I'm just trying to calculate capacity for ping only devices.

[20-May-2010 09:24:25] <sergeymasushko> mrayzenoss: morning.

[20-May-2010 09:24:31] <mrayzenoss> good morning

[20-May-2010 09:25:00] <sergeymasushko> mrayzenoss: I have a weird situation. I have 2 device classes called "Server" and "VPNs". I tried to create the same thershold "/FileSystem      /high disk usage" as in the "Server" class but I get the following error: zenhub|User-supplied Python expression (here.getTotalBlocks() * .85) for maximum value caused error: ['usedBlocks_usedBlocks'] The threshold high disk usage in template /zport/dmd/Devices/VPNs/rrdTemplates/FileSystem has been disable

[20-May-2010 09:28:12] <mrayzenoss> sergeymasushko: version?

[20-May-2010 09:28:55] <sergeymasushko> mrayzenoss: 2.5.1

[20-May-2010 09:29:25] <sergeymasushko> mrayzenoss: installed on FreeBSD :)

[20-May-2010 09:30:35] <mrayzenoss> sergeymasushko: sounds like this bug: http://dev.zenoss.org/trac/ticket/5978

[20-May-2010 09:30:58] <mrayzenoss> fixed in 2.5.2 :)

[20-May-2010 09:31:55] <sergeymasushko> mrayzenoss: is it possible do a patch without upgrade? :)

[20-May-2010 09:32:05] <mrayzenoss> yeah, zenpatch

[20-May-2010 09:32:39] <mrayzenoss> zenpatch 16775

[20-May-2010 09:34:03] <sergeymasushko> looks like it is not so simple :)

[20-May-2010 09:35:52] <sergeymasushko> mrayzenoss: http://pastebin.com/TwfPdsGH

[20-May-2010 09:58:27] <LarsN> jimbosyn: how powerful are the collector machines?

[20-May-2010 09:58:35] <LarsN> at 850-900 snmp devices per?

[20-May-2010 09:58:57] <LarsN> we're monitoring 150 or so snmp + http_monitor + additional. with a xen dom-u :)

[20-May-2010 09:59:48] <jimbosyn> LarsN: They are Amazon EC2 m1.large instances -> 8 cpu 8GB RAM

[20-May-2010 10:00:00] <LarsN> jimbosyn: rgr

[20-May-2010 10:01:32] <jimbosyn> LarsN: rgr?

[20-May-2010 10:01:36] <LarsN> roger,

[20-May-2010 10:01:43] <LarsN> as in, I understand :)

[20-May-2010 10:01:47] <jimbosyn> cool

[20-May-2010 10:02:08] <jimbosyn> LarsN: Im storing rrd file in tmpfs ramdisk.   It is extremely fast.

[20-May-2010 10:02:27] Simon4_ is now known as Simon4

[20-May-2010 10:02:30] <LarsN> that's not a bad idea.

[20-May-2010 10:02:42] <LarsN> that plus a script to dump it to actual disk every X.

[20-May-2010 10:02:44] <jimbosyn> I have a startup/shutdown script to backup/restore rrd files.  I also cron periodic backups about every 15 min

[20-May-2010 10:02:56] <LarsN> I like it.

[20-May-2010 10:03:17] <jimbosyn> I get great peformance and it's a cloud install, so adding a collector takes about 10 minutes.

[20-May-2010 10:06:53] <LarsN> I've taken over our zenoss installation here.

[20-May-2010 10:08:33] <LarsN> there are a lot of adjustments I am planning on making now.

[20-May-2010 10:12:57] <LarsN> including upgrading to 2.5.2

[20-May-2010 10:16:42] <mrayzenoss> jimbosyn: somewhere there's a ticket about disk writes being too slow on cloud installs, I guess this is your work-around?

[20-May-2010 10:17:59] <jimbosyn> mrayzenoss: I saw slow disk writes on the amazone EBS.  I was able to get fast writes with raid0, striping across up to 8 volumes, but this was very complex to manage

[20-May-2010 10:18:32] <jimbosyn> mrayzenoss: and much faster

[20-May-2010 10:26:56] <fragfutter> i want to change the priority of the events a switch port performance warning generates

[20-May-2010 10:27:25] <fragfutter> for a specific switch, and a specific port on the switch i only want to have informational events.

[20-May-2010 10:28:14] <fragfutter> i would add an eventclass mapping at /zport/dmd/Events/Perf/Interfaces

[20-May-2010 10:28:34] <fragfutter> is that correct?

[20-May-2010 10:35:42] <rayk3n> hi all can somebody help me monitor memAvailReal and put the correct values for it, my goal is to send alerts when 5G of memory is being used

[20-May-2010 10:40:19] <rmatte> rayk3n: what type of device are you referring to?

[20-May-2010 10:40:43] <rayk3n> linux device

[20-May-2010 10:41:05] <rmatte> what other mem values are available?

[20-May-2010 10:41:22] <rmatte> actually, I'll check my templates...

[20-May-2010 10:42:22] <rayk3n> i want to set the memAvailReal since it shows the available memory on the device

[20-May-2010 10:42:23] <rmatte> hmmmm

[20-May-2010 10:42:41] <rmatte> right

[20-May-2010 10:43:25] <rayk3n> here is what my device memory looks like

[20-May-2010 10:43:27] <rayk3n> [paul@XVM-Mysql-Master ~]$ free

[20-May-2010 10:43:27] <rayk3n>              total       used       free     shared    buffers     cached

[20-May-2010 10:43:27] <rayk3n> Mem:      16627312    2801724   13825588          0     171092    184592

[20-May-2010 10:43:33] <rmatte> are you able to just use here.memAvailReal?

[20-May-2010 10:43:49] <rayk3n> when i ran memavailfree it shows

[20-May-2010 10:43:50] <rayk3n> Executing command

[20-May-2010 10:43:50] <rayk3n> snmpwalk -cuprint1nG -v2c 206.126.137.228 1.3.6.1.4.1.2021.4.6.0

[20-May-2010 10:43:50] <rayk3n> against MySQL Master

[20-May-2010 10:43:51] <rayk3n> UCD-SNMP-MIB::memAvailReal.0 = INTEGER: 13924928 kB

[20-May-2010 10:43:51] <rmatte> or here.hw.memAvailReal

[20-May-2010 10:43:53] <rayk3n> DONE in 0 seconds

[20-May-2010 10:44:10] <rmatte> right, that has nothing to do with what I'm asking

[20-May-2010 10:44:19] <rmatte> give me a few minutes...

[20-May-2010 10:44:33] <rayk3n> alriight

[20-May-2010 10:47:19] <rmatte> now...

[20-May-2010 10:47:35] <rmatte> are you ok with setting the threshold as a percentage of utilization?

[20-May-2010 10:48:00] <rayk3n> hmm

[20-May-2010 10:48:05] <rayk3n> let me try

[20-May-2010 10:48:09] <rmatte> well hold up

[20-May-2010 10:48:19] <rayk3n> can i use that on a specific device only?

[20-May-2010 10:48:27] <rmatte> no

[20-May-2010 10:48:39] <rayk3n> i guess i can't use that

[20-May-2010 10:48:39] <rmatte> ok, let me explain something...

[20-May-2010 10:48:47] <rayk3n> sure

[20-May-2010 10:48:52] <rmatte> memAvailReal only gives you the amount of available memory on the device

[20-May-2010 10:48:53] <rayk3n> btw thanks for the time

[20-May-2010 10:48:59] <rmatte> which is basically useless unless you also know the total

[20-May-2010 10:49:09] <rmatte> now the total is stored in here.hw.totalMemory

[20-May-2010 10:49:48] <rmatte> so we need to use the two to set a proper threshold

[20-May-2010 10:50:19] <rayk3n> hmm i think i do not have totalMemory in my template

[20-May-2010 10:50:25] <rmatte> the problem is, if we had something that provided used instead of available it would be much more helpful

[20-May-2010 10:50:27] <rayk3n> i just use the default linux device

[20-May-2010 10:50:42] <rmatte> right, the default linux device doesn't have any properly built thresholds

[20-May-2010 10:50:53] <rmatte> hence why you need to develop them yourself

[20-May-2010 10:51:18] <rmatte> either that or just hardcode the amount of bytes in to a threshold

[20-May-2010 10:51:22] <rmatte> but that's not very dynamic

[20-May-2010 10:51:45] <rayk3n> ic

[20-May-2010 10:52:13] <rayk3n> can i just use a min or max value? on the memAvailReal

[20-May-2010 10:52:20] <rayk3n> then add a warning or critical alert

[20-May-2010 10:52:23] <rmatte> it'll have to be in bytes if you do

[20-May-2010 10:52:35] <rayk3n> its fine by me

[20-May-2010 10:52:40] <rmatte> it's not smart enough to know it's a percentage unless you actually provide an equation that works out to be a percentage

[20-May-2010 10:52:51] <rayk3n> just need to know what bytes like example. 1392492

[20-May-2010 10:53:02] <rmatte> k, well then just specify the proper number of bytes of available memory for when 5GB is in use

[20-May-2010 10:53:04] <rayk3n> if it goes 700000 or lower

[20-May-2010 10:53:12] <rmatte> and apply the threshold against memAvailReal

[20-May-2010 10:53:26] <rmatte> you'll need more than 700000 lol

[20-May-2010 10:53:38] <fragfutter> memTotalReal-memAvailReal-memBuffer= memused

[20-May-2010 10:53:40] <rmatte> 1024000000 is 1GB

[20-May-2010 10:53:57] <rayk3n> yes that's what i need to know

[20-May-2010 10:54:03] <rmatte> fragfutter: there is no memtotalreal, he'd have to add that

[20-May-2010 10:55:00] <fragfutter> rmatte: can't you add this as a datasource?

[20-May-2010 10:55:14] <rmatte> fragfutter: of course

[20-May-2010 10:55:22] <fragfutter> fairly static, but should work.

[20-May-2010 10:55:34] <rmatte> nah, you could make it non-static

[20-May-2010 10:56:10] <fragfutter> rmatte: sure. i ment it would poll the same value most of the times.

[20-May-2010 10:56:25] <rayk3n> so if i want to send alert if a 13G memory uses 5G so that's 7G ill just input 1024000000(x7) in the min ?

[20-May-2010 10:57:39] <rmatte> rayk3n: for a rough way of doing it, correct

[20-May-2010 10:58:10] <rayk3n> hmm weird i just added 14000000 on the min. and it gives me alert

[20-May-2010 10:59:01] <fragfutter> can i have a threshold on a calculated value?

[20-May-2010 10:59:07] <rayk3n> i mean 140000000

[20-May-2010 10:59:22] <rayk3n> and my current memory is 1392492

[20-May-2010 10:59:25] <rmatte> fragfutter: what do you mean?

[20-May-2010 11:00:06] <rmatte> rayk3n: you're doing something wrong, the threshold is set against memAvailReal, yes?

[20-May-2010 11:00:25] <rayk3n> yes

[20-May-2010 11:00:47] <rmatte> snmpwalk that OID to see what the current value is

[20-May-2010 11:00:53] <rmatte> then you'll understand why it's not working

[20-May-2010 11:01:05] <rmatte> using the "free" command isn't going to cut it

[20-May-2010 11:01:09] <rayk3n> <rayk3n> UCD-SNMP-MIB::memAvailReal.0 = INTEGER: 13924928 kB

[20-May-2010 11:01:09] <rayk3n> <rayk3n> DONE in 0 seconds

[20-May-2010 11:01:34] <fragfutter> rmatte: if i define a threshold, it is defined against a datapoint, which represents a single oid. If i could send a threshold against something like pointA+pointB

[20-May-2010 11:01:52] <rmatte> ok, so you set the min to 140000000 and it's actually 13924928

[20-May-2010 11:02:01] <rmatte> you have an extra 0 that shouldn't be there apparently

[20-May-2010 11:02:06] <rmatte> actually

[20-May-2010 11:02:20] <rmatte> add -On as a flag to the snmpwalk

[20-May-2010 11:02:34] <rmatte> pretty sure that's the problem though

[20-May-2010 11:03:22] <twm1010> rmatte:  do you pull any perf data from ds1 and ds3 controllers?

[20-May-2010 11:03:40] <twm1010> i have blank templates right now, so they don't use ethernetCSmacd, but surely there are metrics to be watches there

[20-May-2010 11:03:44] <twm1010> *watched

[20-May-2010 11:03:44] <rayk3n> hmm where can i find that

[20-May-2010 11:04:15] <fragfutter> the trailing zero is fine for a netsnmp server

[20-May-2010 11:05:42] <rayk3n> rmatte i think 14000000 is the value im looking for

[20-May-2010 11:07:10] <rmatte> looks like

[20-May-2010 11:07:35] <rmatte> I guess they cut it down to KB instead of bytes, my bad

[20-May-2010 11:08:02] <rmatte> snmpwalk is your friend when troubleshooting stuff like that

[20-May-2010 11:08:20] <rayk3n> ic thanks

[20-May-2010 11:09:26] <rmatte> np

[20-May-2010 11:12:56] <fragfutter> rmatte: any idea about a threshold against a calculated value?

[20-May-2010 11:13:05] <rmatte> define: calculated value

[20-May-2010 11:13:15] <rmatte> I can do basically anything with thresholds, just be more specific :)

[20-May-2010 11:17:35] <fragfutter> rmatte: i have two data source. Both are SNMp values.

[20-May-2010 11:17:41] <rmatte> ok

[20-May-2010 11:18:00] <fragfutter> rmatte: no i want to have a threshold, that complaiins if data_source1+data_source2 > 100

[20-May-2010 11:18:28] <rmatte> I see...

[20-May-2010 11:19:59] <rmatte> I'm trying to think of the best way to do that...

[20-May-2010 11:20:26] <fragfutter> can i add a data source that calculates its value from the two other datasources?

[20-May-2010 11:21:20] <mrayzenoss> fragfutter: kinda

[20-May-2010 11:21:31] <mrayzenoss> you can do that with a threshold or a graph point

[20-May-2010 11:21:42] <fragfutter> i'm listening

[20-May-2010 11:21:46] <mrayzenoss> I'm looking :)

[20-May-2010 11:22:59] <rmatte> I know how to do it with a graphpoint, but that doesn't help him with his threshold

[20-May-2010 11:23:32] <rmatte> I know how to do calculations on the min and max values themselves

[20-May-2010 11:23:41] <rmatte> but he wants to apply the threshold against the sum of 2 datasources

[20-May-2010 11:23:51] <fragfutter> rmatte: exactly

[20-May-2010 11:24:08] <rmatte> There's a way to do it, I just don't know off the top of my head...

[20-May-2010 11:25:55] <mrayzenoss> hmm… it seems like you should be able to put the data points in the min/max value fields

[20-May-2010 11:29:09] <fragfutter> mrayzenoss: looking at the source, yes this might work.

[20-May-2010 11:30:26] <rmatte> mrayzenoss: you can, but that doesn't work for what he wants to do

[20-May-2010 11:30:48] <rmatte> the threshold needs to be applied against a value

[20-May-2010 11:30:59] <rmatte> and that value needs to be data_source1+data_source2

[20-May-2010 11:31:05] <rmatte> adding them in to the min or max doesn't help

[20-May-2010 11:31:15] <mrayzenoss> hmm… yeah

[20-May-2010 11:31:43] <rmatte> that's one thing I'd love to see fixed, since I've had to use command datasources for no real good reason just because of that

[20-May-2010 11:31:51] <rmatte> it works fine the other way around

[20-May-2010 11:32:19] <mrayzenoss> yeah, command data sources make it easy

[20-May-2010 11:32:27] <fragfutter> i could define minmax-threshold against datasource2 a use a min-value of 100-datasource1

[20-May-2010 11:32:27] <mrayzenoss> sounds like a good enhancement ticket

[20-May-2010 11:32:49] <rmatte> hmmm, you could

[20-May-2010 11:32:50] <fragfutter> (max-value that is)

[20-May-2010 11:33:01] <rmatte> if you want to use the datasources in min/max that's really easy...

[20-May-2010 11:33:03] <rmatte> just do...

[20-May-2010 11:33:13] <rmatte> here.getRRDValue("nameOfDataSource")

[20-May-2010 11:33:33] <rmatte> you're using Zenoss 2.5.2 I hope?

[20-May-2010 11:33:44] <fragfutter> yes

[20-May-2010 11:33:47] <rmatte> k, then yeh

[20-May-2010 11:33:50] <rmatte> just do that and you'll be good

[20-May-2010 11:34:21] <fragfutter> yeah, but it is a rather strange way to define the threshold.

[20-May-2010 11:34:30] <rmatte> whatever works

[20-May-2010 11:34:31] <rmatte> lol

[20-May-2010 11:35:34] <fragfutter> define a custom oid on the monitored-server side might be easier on my brain if i look at it in a few months

[20-May-2010 11:36:17] <rmatte> the method you just described makes sense and would work fine

[20-May-2010 11:36:31] <rmatte> 100 - here.getRRDValue("datasource1")

[20-May-2010 11:36:36] <rmatte> as the min

[20-May-2010 11:36:55] <rmatte> actually, wouldn't you want it set as the max?

[20-May-2010 11:37:07] <fragfutter> yes max, i corrected myself.

[20-May-2010 11:37:11] <rmatte> ah, k

[20-May-2010 11:38:02] <fragfutter> i think i will file an enhancement request for a threshold that allows a python snippet.

[20-May-2010 11:38:27] <rmatte> that would be cool

[20-May-2010 11:38:30] <fragfutter> would that be possible?

[20-May-2010 11:38:34] <rmatte> probably

[20-May-2010 11:38:41] <rmatte> something similar to event handling

[20-May-2010 11:39:19] <fragfutter> i also have questions about event handling, but i need to leave...

[20-May-2010 11:39:21] <rmatte> afk for a few

[20-May-2010 12:00:11] <mistich> rmatte you update the cisco mibs latley

[20-May-2010 12:07:35] <rmatte> mistich: not recently, why do you ask?

[20-May-2010 12:10:06] <mistich> tried loading the zenpack hour and a half still not loaded

[20-May-2010 12:10:26] <mistich> just diffed what was on cisco to yours nothing added

[20-May-2010 12:10:42] <mistich> installing mibs manual now

[20-May-2010 12:10:52] <rmatte> it says right on the ZenPack page that the ZenPack takes 5 to 7 hours to install

[20-May-2010 12:10:58] <rmatte> or something along those lines

[20-May-2010 12:11:15] <mistich> Performing these steps will eliminate the performance issues, while  leaving the MIBs installed. The first line will take about 15 minutes to  complete, the commit may take another 7-10.

[20-May-2010 12:11:26] <rmatte> wrong part

[20-May-2010 12:11:28] <mistich> docs/DOC-3404

[20-May-2010 12:11:32] <rmatte> read it properly

[20-May-2010 12:11:37] <mistich> NOTE: There are a very large number of MIBs included in  this ZenPack (1128), so it is advised that you install with the  command-line rather than the UI so it does not time-out. It may still  take 4 or 5 hours to process.

[20-May-2010 12:11:41] <rmatte> correct

[20-May-2010 12:11:53] <rmatte> "It may still  take 4 or 5 hours to process."

[20-May-2010 12:12:16] <rmatte> anyways, if you want to take a stab at installing them manually all good

[20-May-2010 12:12:20] <mistich> yes going to run it again

[20-May-2010 12:12:23] <rmatte> let me know if they install without any issues

[20-May-2010 12:12:35] <rmatte> I had lots of problems when I was first installing them

[20-May-2010 12:12:41] <rmatte> which is why I ended up making the ZenPack

[20-May-2010 12:12:51] <rmatte> though zenmib was apparently rewritten based on my guide for building the Mibs

[20-May-2010 12:12:54] <mistich> yes will try the zenpack again

[20-May-2010 12:13:01] <rmatte> k

[20-May-2010 12:13:45] <rmatte> I will update the pack eventually, I just need to find the time to do it

[20-May-2010 12:13:57] <rmatte> but it shouldn't need updating at the moment

[20-May-2010 12:14:12] <mistich> well did you get all the mibs in there or is there some missing

[20-May-2010 12:14:20] <rmatte> I got all of them in

[20-May-2010 12:14:29] <mistich> ok

[20-May-2010 12:14:31] <rmatte> not sure if there are any new ones since then though

[20-May-2010 12:14:35] <rmatte> but I haven't had any issues

[20-May-2010 12:14:55] <rmatte> I had to modify some of them so that they'd load properly

[20-May-2010 12:15:09] <rmatte> there's a link on the ZenPack page to a guide that explains what I went through to get them all loaded

[20-May-2010 12:15:13] <rmatte> it was quite involved

[20-May-2010 12:15:15] <mistich> is this is the link of the ones you used its the same as ciscos  There are a very large number of MIBs included in this ZenPack (1128),  so it is advised that you install with the command-line rather than the  UI so it does not time-out. It may still take 4 or 5 hours to process

[20-May-2010 12:15:32] <mistich> is this is the link of the ones you used its the same as ciscos  http://demonic.cc/downloads/zenoss/cisco-mibs-zenoss.tar.gz

[20-May-2010 12:15:49] <rmatte> that has the Mibs that I modifed

[20-May-2010 12:15:51] <rmatte> modified*

[20-May-2010 12:15:55] <rmatte> so that they would load

[20-May-2010 12:16:03] <mistich> ok then no new updates from cisco

[20-May-2010 12:16:14] <rmatte> though the modifications are probably not necessary anymore

[20-May-2010 12:16:29] <rmatte> not unless cisco updated theirs since I originally made the pack

[20-May-2010 12:16:36] <rmatte> The pack is a year old

[20-May-2010 12:16:39] <mistich> nope

[20-May-2010 12:16:48] <rmatte> though I can't see them having done anything overly groundbreaking since then

[20-May-2010 12:18:04] <rmatte> make sure you follow the zendmd process listed on the ZenPack page after it's done installing

[20-May-2010 12:18:10] <rmatte> or your UI is going to be very slow

[20-May-2010 12:18:32] <mistich> why does it make it slow?

[20-May-2010 12:18:43] <rmatte> it's some bug in Zenoss

[20-May-2010 12:18:54] <rmatte> the more ZenPack content installed, the slower the UI gets

[20-May-2010 12:18:56] <mistich> doesn't surprise me

[20-May-2010 12:19:19] <rmatte> so by performing those steps, it leaves the Mibs installed but makes Zenoss unaware of the ZenPack

[20-May-2010 12:19:23] <rmatte> which resolves the issue

[20-May-2010 12:19:45] <mistich> ok

[20-May-2010 12:20:05] <mistich> well started zenpack install over well see how it goes

[20-May-2010 12:20:13] <rmatte> yeh, give it a lot of time lol

[20-May-2010 12:20:16] <rmatte> it will finish

[20-May-2010 12:20:44] <mistich> I will just miss read and got back from lunch and still running

[20-May-2010 12:20:55] <rmatte> hehe

[20-May-2010 12:21:05] <mistich> and 100% cpu

[20-May-2010 12:21:16] <mistich> but have 7 cores to spare

[20-May-2010 12:22:26] <mistich> wish zenoss would be more multi-threaded friendly

[20-May-2010 12:22:56] <Simon4_> I wonder if you could run zenmib 8 times with 1/8'th of the mibs each :)

[20-May-2010 12:23:03] <rmatte> lol

[20-May-2010 12:23:31] <mistich> trust me I have done that with zendisc all the time

[20-May-2010 12:24:13] <mistich> scan a whole /16 in less than 5 mins

[20-May-2010 12:24:56] RoAk is now known as RoAkSoAxB

[20-May-2010 14:29:20] <Hamzah> hmmm For a Cisco Catalyst switch, is it better to use the /Devices/Network/Router/Cisco Class (Ie does it get any extra information line it does for Cisco routers) or will I get the same details if I use /Devices/Network/Switch ?

[20-May-2010 14:32:01] LarsN is now known as LarsN_Away

[20-May-2010 14:37:12] <rmatte> Hamzah: I created my own switch template

[20-May-2010 14:37:22] <rmatte> and then I created another specifically for catalyst switches

[20-May-2010 14:37:30] <rmatte> since they use slightly different performance OIDs

[20-May-2010 14:37:46] <Hamzah> hmm I see

[20-May-2010 14:37:51] <rmatte> Hamzah: Zenoss is really more of a blank slate than a fully ready to go monitoring tool, you'll have to learn how to develop templates

[20-May-2010 14:37:56] <rmatte> it's really not hard to do

[20-May-2010 14:38:17] <rmatte> plus once you figure it out you'll be able to monitor anything

[20-May-2010 14:38:33] <Hamzah> I've been meaning to learn how to do that for some time :) I currently only have it monitoring the "default" things

[20-May-2010 14:39:37] <Hamzah> The administration guide has everything regarding developing templates, right?

[20-May-2010 14:40:02] <rmatte> correct

[20-May-2010 14:40:19] <rmatte> one of the first things I did when I started using Zenoss was to read the admin guide from top to bottom

[20-May-2010 14:40:19] <Hamzah> ok, I'll go through that and see if I can figure it out :)

[20-May-2010 14:40:23] <rmatte> and it was extremely helpful

[20-May-2010 14:40:27] <rmatte> k

[20-May-2010 14:41:01] <rmatte> I prefer it when people read through that first before trying to walk them through it

[20-May-2010 14:41:08] <rmatte> it's the whole teach a man to fish type thing :)

[20-May-2010 14:41:19] <rmatte> After you've read it, if you run in to snags I'll be happy to help

[20-May-2010 14:43:22] <Hamzah> I've only read through parts of it so far, not top to bottom. Will read through it tonight.... I have nothing to do and am stuck on night shift anyway :P

[20-May-2010 14:44:04] <rmatte> hehe

[20-May-2010 14:44:05] <rmatte> cool

[20-May-2010 14:49:26] <msg4al> hello, i am really confused about event mapping, specifically rules, regex, transforms.  can anyone explain or point me in the right direction?

[20-May-2010 14:57:19] <mrayzenoss> msg4al: besides the Admin Guide, this has the best write-up: http://www.skills-1st.co.uk/papers/jcurry.html

[20-May-2010 14:57:40] <msg4al> thanks!

[20-May-2010 14:57:41] <mrayzenoss> Specifically "Zenoss Event Management"

[20-May-2010 15:01:21] <mrayzenoss> gotta run

[20-May-2010 15:01:57] <rmatte> well, that's one of the better typos I've made...

[20-May-2010 15:01:59] <rmatte> lsd: command not found

[20-May-2010 15:09:37] <msg4al> should i be concerned about event class mappings in the beggining or is this a advanced function?  i checked the admin guide and resource recommended by mrayzenoss and im still very confused

[20-May-2010 15:12:28] <twm1010> msg4al: once your system is tuned and working, event class mappings are the heart of you knowing what's going on

[20-May-2010 15:12:58] <twm1010> so in the beginning, focus on making zenoss work and getting information into it, then dig into event class mapping so you know what to do with the information you're getting

[20-May-2010 15:18:37] <msg4al> twm1010: thanks for the info. i have a bunch of info already collected through snmp. i just dont understand the syntax...if i go to edit a event for example, all i see is event class key and example fields filled in

[20-May-2010 15:24:59] <twm1010> so you're looking to classify the events you've already received?

[20-May-2010 15:28:00] <msg4al> yes, but i am more interested in seeing the rules that zenoss uses to classify existing events.  for example, /events/status/wmi has a ton of events.  i wanna check and understand the logic it used to place those alerts there. maybe im making it too complicated

[20-May-2010 15:30:34] <twm1010> hrmm..i would read the admin guide nice and slow

[20-May-2010 15:30:46] <twm1010> it can use examples, regular expressions, and event class keys to do  it

[20-May-2010 15:30:54] <twm1010> it can transform those events as well

[20-May-2010 16:16:39] <chemist> can I install zenpacks from the command line on a collector?

[20-May-2010 16:19:55] <chemist> I'll explain, I used bigegor's Distributed Collector zenpack to set up a remote collector. I think the WMI DataSource and WMI Perf plugins have somehow broke

[20-May-2010 16:20:26] <chemist> reinstalling on the master and updating the collector had no effect,

[20-May-2010 16:20:55] <chemist> so I am thinking of installing them from the command line on the collector itself

[20-May-2010 16:31:58] <Simon4_> chemist: I've done it, it doesn't reeeealy install it, but it's fixed issues with some zenpacks

[20-May-2010 16:32:22] <Simon4_> zenpack --install=zenpackname.x.y.z --host=masterzenossserver --port=8100 or similar

[20-May-2010 16:32:35] <Simon4_> you get a bunch of errors, but it fixed my issue with my zenpacks

[20-May-2010 16:32:50] <chemist> you ran that fromthe collector?

[20-May-2010 16:32:55] <Simon4_> yeah

[20-May-2010 16:33:12] <Simon4_> seems to be an issue with maybe zenpacks that do things outside the $ZENHOME/ZenPacks dir

[20-May-2010 16:33:36] <Simon4_> we hit the same with an in-house zenpack under enterprise today

[20-May-2010 16:34:13] <Simon4_> (had had same issue in core)

[20-May-2010 16:37:19] Simon4_ is now known as Simon4

[20-May-2010 16:38:05] * Simon4 has had thoughts that to create a collector mirroring $ZENHOME to the collector may be the way to do it, it's a bit like nuking from orbit but hey

[20-May-2010 16:38:27] <Simon4> you would just need to exclude $ZENHOME/etc

[20-May-2010 16:38:58] <Simon4> oh, chemist, I proof of concept'ed the remote zenrender idea today

[20-May-2010 16:39:09] <chemist> and?

[20-May-2010 16:39:36] <chemist> you broke it? :)

[20-May-2010 16:39:53] <Simon4> worked a charm with some apache rewrites in the way :)

[20-May-2010 16:40:10] <chemist> excellent news

[20-May-2010 16:40:11] <Simon4> I couldn't get the http://zenhub:hubport/render.server.here type url

[20-May-2010 16:40:36] <Simon4> but with a rewrite or proxy straight to http;//renderserver:8091 it worked perfectly

[20-May-2010 16:41:22] <Simon4> will be playing more with the "through the zenhub" idea yet, but at least I've proved it can be done

[20-May-2010 16:44:47] <chemist> Simon4: just trying your method on a couple of zenpacks

[20-May-2010 16:45:19] <chemist> will that be trying to install from the master?

[20-May-2010 17:48:01] <MWT> Hello! Anyone around?

[20-May-2010 22:03:58] <Wolf> is there a document that I can access which lists the new features and changes coming in 3.0?

[21-May-2010 00:00:30] [disconnected at Fri May 21 00:00:30 2010]

[21-May-2010 00:00:30] [connected at Fri May 21 00:00:30 2010]

[21-May-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[21-May-2010 04:25:23] <Jane_Curry> Wolf - there is some draft documentation for 3.0 referenced here - message/48232#48232

[21-May-2010 05:18:06] <Jane_Curry> zykes-  are you still looking at notification schedules?

[21-May-2010 05:19:36] <zykes-> yeh Jane_Curry

[21-May-2010 05:20:35] <Jane_Curry>  I find your mileage may vary

[21-May-2010 05:20:47] <Jane_Curry> Sometimes it gets duration correct in the report and sometimes it gets it wrong

[21-May-2010 05:21:02] <Jane_Curry> I believe it DOES actually honour your schedule for alerts - it just seems to be the report that sometimes has the wrong numbers in it

[21-May-2010 05:21:16] <Jane_Curry> Does this match what you see????

[21-May-2010 05:23:36] <zykes-> well it says that duration is 2 days

[21-May-2010 05:23:44] <zykes-> i can't get that to match what i mean it should be..

[21-May-2010 05:36:30] <Jane_Curry> try changing the duration and then go back and llok at the report

[21-May-2010 05:36:51] <Jane_Curry> sometimes I find the report than has the correct value, sometimes not

[21-May-2010 05:37:03] <Jane_Curry> .. not sure why yet...

[21-May-2010 05:37:44] <zykes-> where's ray or cgi

[21-May-2010 05:37:45] <zykes-> :p

[21-May-2010 06:24:03] <Jane_Curry> I think you will find when the value is incorrect that it shows duration as the time to when the NEXT active schedule starts

[21-May-2010 06:24:31] <Jane_Curry> I think it will always get the duration value correct if your alerting schedule is not currently active

[21-May-2010 06:24:40] <Jane_Curry> Any way you can confirm this??

[21-May-2010 06:24:52] <Jane_Curry> PS. I am on Zenoss 2.5.2 - what's your level?

[21-May-2010 06:26:35] <Jane_Curry> Can anyone else confirm that they are seeing incorrect values for Duration in Reports -> User Reports -> Notification values

[21-May-2010 06:27:57] <Jane_Curry> I think the problem is in the Products.ZenEventsActionRule in the nextDurationNice method

[21-May-2010 06:33:00] <zykes-> same Jane_Curry

[21-May-2010 06:52:40] <Jane_Curry> anyone else confirm this????

[21-May-2010 08:18:34] <rmatte> anyone have any idea what would cause this: http://fpaste.org/OiEt/ ?

[21-May-2010 08:19:27] <rmatte> it just keeps cycling through that error over and over

[21-May-2010 08:19:55] LarsN_Away is now known as LarsN

[21-May-2010 08:21:23] <LarsN> I restarted my zenoss server this morning for the first time in months.  Wanted to add another cpu to the virtual machine which required a reboot.  After bringing the machine back online. I am getting "unable to connect to the MySQL server."

[21-May-2010 08:21:44] <LarsN> I also notice that zenhub and zenactions aren't running, and won't start.

[21-May-2010 08:21:59] <rmatte> what type of install, stack install?

[21-May-2010 08:22:22] <LarsN> i'm not sure I know the answer to that.  I didn't build the machine.

[21-May-2010 08:22:31] <LarsN> this is the central server, with one additional collector.

[21-May-2010 08:22:38] <rmatte> does the file /etc/init.d/zenoss-stack exist?

[21-May-2010 08:22:51] <LarsN> it does.

[21-May-2010 08:22:57] <rmatte> k, then it's a stack install

[21-May-2010 08:23:04] <rmatte> do /etc/init.d/zenoss-stack stop

[21-May-2010 08:23:09] <rmatte> then /etc/init.d/zenoss-stack start

[21-May-2010 08:23:17] <rmatte> and http://fpaste.org any errors

[21-May-2010 08:24:56] <LarsN> apparently that was all it took.

[21-May-2010 08:25:25] <LarsN> no errors, and things appear to be working again.  I didn't realize that init script was there.  I was restarting via ../zenoss/bin/zenoss restart

[21-May-2010 08:25:45] <rmatte> ah

[21-May-2010 08:25:51] <rmatte> yeh, that init script controls mysql

[21-May-2010 08:25:58] <rmatte> zenoss restart only controls the daemons themselves

[21-May-2010 08:26:12] <LarsN> it was strange, I was able to log into mysql on the port it's running user, as the zenoss user.

[21-May-2010 08:26:15] <rmatte> if you just want to start mysql via the init script you can do: /etc/init.d/zenoss-stack start mysql

[21-May-2010 08:26:28] <rmatte> or it's /etc/init.d/zenoss-stack start mysqld

[21-May-2010 08:26:34] <rmatte> one or the other, can't remember which

[21-May-2010 08:26:45] <LarsN> that's good information to have, thanks rmatte

[21-May-2010 08:26:49] <rmatte> np

[21-May-2010 08:27:04] <LarsN> learned two things today.  1) I have a stack install.  and 2) there's an init script that's pretty important :)

[21-May-2010 08:27:16] <rmatte> hehe

[21-May-2010 08:27:18] <LarsN> while I've got you here.

[21-May-2010 08:27:32] <LarsN> I had been working with that transform you wrote about for nfs monitoring.

[21-May-2010 08:27:46] <rmatte> yup

[21-May-2010 08:27:53] <LarsN> my share's are showing up under the OS tab, which is great.  However I don't see the lock next to them.

[21-May-2010 08:28:46] <rmatte> it only locks them when they go down

[21-May-2010 08:28:51] <LarsN> ahhh

[21-May-2010 08:28:52] <rmatte> though you can do it by hand if you want

[21-May-2010 08:29:04] <rmatte> if a share goes down the transform automatically locks it

[21-May-2010 08:29:12] * LarsN takes one down to make sure it's still functioning as expected.

[21-May-2010 08:29:18] <rmatte> hehe

[21-May-2010 08:29:21] <LarsN> I did a lot of work on 6 servers yesterday. :)

[21-May-2010 08:29:29] <rmatte> :)

[21-May-2010 08:29:54] <rmatte> One of my Zenoss servers is giving me grief when I try to remodel solaris devices

[21-May-2010 08:30:09] <rmatte> might be a bug in 2.5.2, not sure

[21-May-2010 08:30:19] <rmatte> haven't remodeled any of these devices since I upgraded to 2.5.2

[21-May-2010 08:30:26] <LarsN> I have three open solaris machines.

[21-May-2010 08:30:32] forsberg is now known as fOrsberg

[21-May-2010 08:30:36] <LarsN> each of which is it's own, significant challenge.

[21-May-2010 08:30:42] <rmatte> yeh

[21-May-2010 08:30:49] <rmatte> they are a huge pain

[21-May-2010 08:30:49] <rmatte> lol

[21-May-2010 08:31:00] <rmatte> they aren't even my solaris servers, they belong to one of our clients

[21-May-2010 08:31:03] <rmatte> we just monitor them

[21-May-2010 08:31:25] <LarsN> we built three storage servers on OpenSolaris for zfs

[21-May-2010 08:32:15] <LarsN> hmmm...

[21-May-2010 08:32:20] <LarsN> well that's not good.

[21-May-2010 08:33:37] <rmatte> ?

[21-May-2010 08:34:03] <LarsN> currently the NFS share is going to time out.

[21-May-2010 08:34:15] * LarsN hates nfs.

[21-May-2010 08:34:45] <LarsN> anyway.  zenoss doesn't currently think NFS has any problems. although I'm on the machine and can confirm it's pretty broken.

[21-May-2010 08:35:30] <rmatte> well, I guess it depends on whether the share is unmounted or not

[21-May-2010 08:35:49] <rmatte> it needs to timeout and unmount before it'll see it

[21-May-2010 08:36:14] <LarsN> I'm going to have to write a plugin that uses ssh, and looks for the existance of a file in the share I think.

[21-May-2010 08:36:20] <rmatte> the only other solution would be to develop a script that actually needs to ssh in to the machine and check the shares

[21-May-2010 08:36:30] <rmatte> yup

[21-May-2010 08:36:57] <rmatte> how did you even manage to get nfs to fail like that?

[21-May-2010 08:39:30] <LarsN> I'm willing to put money on our amazingly terrible Dell PowerConnect 6548 switches.

[21-May-2010 08:40:45] <rmatte> lol

[21-May-2010 08:40:59] <rmatte> the only switches anyone should ever buy are Cisco and HP

[21-May-2010 08:41:10] <rmatte> and maybe Enterasys if you're adventurous

[21-May-2010 08:41:42] <LarsN> we've been looking at Junipers.  The dell's were attractive since they were ~ $1800 each, while a similar Juniper was $9000

[21-May-2010 08:41:51] <LarsN> pretty sure you get what you pay for.

[21-May-2010 08:41:52] <LarsN> :/

[21-May-2010 08:44:19] <rmatte> Juniper does make good stuff

[21-May-2010 08:44:36] <LarsN> we are using three of their SRX series routers.

[21-May-2010 08:44:39] <rmatte> although I'm not a massive fan of their switches

[21-May-2010 08:44:43] <rmatte> their netscreens are awesome though

[21-May-2010 08:44:59] <Simon4> their SRX's are cool

[21-May-2010 08:45:00] <LarsN> I think the srx is a junpier router + a netscreen / two

[21-May-2010 08:45:01] <LarsN> :)

[21-May-2010 08:45:07] <rmatte> when it comes to routers I'm Cisco all the way

[21-May-2010 08:45:12] <rmatte> switches can be Cisco or HP

[21-May-2010 08:45:17] <rmatte> security can be Cisco or Juniper

[21-May-2010 08:45:26] * Simon4 would like to make a large pile of HP switches and burn them

[21-May-2010 08:45:28] <LarsN> these dell switches show a lot of promise.

[21-May-2010 08:45:28] <rmatte> just from experience

[21-May-2010 08:45:42] <Simon4> as it happens, I'm going to have a large pile soon :)

[21-May-2010 08:45:45] <LarsN> there are just some very VERY strange things they do.

[21-May-2010 08:45:59] <rmatte> Simon4: If they are decent switches ship them my way :P

[21-May-2010 08:46:21] <Simon4> see, I struggle to use "HP" and "decent switches" in the same sentence ;)

[21-May-2010 08:46:36] <LarsN> 10.1.1.5/26 & 10.1.1.25/26  traceroute from .5 --> .25 the first hop is 10.1.1.1 (the switch ip.)

[21-May-2010 08:46:39] <rmatte> Simon4: I find they are quite good, but I guess to each his own

[21-May-2010 08:46:41] <Simon4> just a bunch of blade ones

[21-May-2010 08:46:47] <rmatte> the HP management console is great

[21-May-2010 08:47:00] <rmatte> their blade switches are a different story

[21-May-2010 08:47:06] <LarsN> it's idiotic that, on the same network, the switch would assume it needs to operate at layer 3.

[21-May-2010 08:47:07] <rmatte> I hate HP blade switches

[21-May-2010 08:47:19] <Simon4> rmatte: yeah, that's the only HP switch exp I have, so probably explains the hate

[21-May-2010 08:47:25] <rmatte> yeh

[21-May-2010 08:47:35] <rmatte> they must have had a different team making their blade switches

[21-May-2010 08:47:41] <rmatte> because their other switches are rock solid

[21-May-2010 08:48:20] <rmatte> I used to not mind Nortel switches but the company was uber unstable and finally died

[21-May-2010 08:50:52] <rmatte> eugh, what the hell could possibly be causing these twisted errors

[21-May-2010 08:51:15] <LarsN> rmatte: if I'm not doing a lot of crazy things with my installation

[21-May-2010 08:51:30] <LarsN> how likely is the 2.5.2 upgrade to "just work" from 2.4.5 (w/patches)

[21-May-2010 08:52:02] <rmatte> well, the upgrade should go smoothly enough but you might run in to little problems

[21-May-2010 08:52:14] <rmatte> which I have the solution for most of

[21-May-2010 08:52:32] <LarsN> does the order I do the upgrade in matter?

[21-May-2010 08:52:38] <LarsN> ie, master --> collectors

[21-May-2010 08:52:48] <rmatte> nah, that shouldn't be an issue

[21-May-2010 08:52:51] <LarsN> k

[21-May-2010 08:52:58] <rmatte> just upgrade the collector, and the master at the same time

[21-May-2010 08:56:27] <rmatte> cgibbons: question... prior to 2.5.2 I was able to properly model devices in /Server/Solaris... after the upgrade I can model any devices except for those in that class... I get the following error that just repeatedly continues when modelling: http://fpaste.org/oGQm/ --- It also hangs up zenhub so that it ignores the stop command and zenmodel isn't able to connect to it.  I end up having to kill -9 the zenhub process and then start it again.

[21-May-2010 08:56:44] <rmatte> any idea what would be causing that?

[21-May-2010 09:01:17] <rmatte> ignore the note at the top about the RouteMap plugin, I thought that might be causing the issue but it wasn't

[21-May-2010 09:06:04] <rmatte> This is the exact same problem I'm having message/48069

[21-May-2010 09:10:12] * rmatte is going to open a Trac ticket

[21-May-2010 09:18:36] <LarsN> rmatte: is it normal behaviour for zenoss to do the following:  click add device, setup all the fields (including selecting pre-production) from the production state dropdown.  click save.

[21-May-2010 09:18:46] <LarsN> device get's modeled, and stuck in state production.

[21-May-2010 09:23:02] <rmatte> no

[21-May-2010 09:23:20] <rmatte> it briefly goes in to a state of Production for some reason but the end result should be Pre-Production

[21-May-2010 09:35:41] fOrsberg is now known as forsberg

[21-May-2010 09:37:16] <rmatte> there

[21-May-2010 09:37:20] <rmatte> http://dev.zenoss.com/trac/ticket/6779

[21-May-2010 09:40:08] <forsberg> nast

[21-May-2010 09:40:10] <forsberg> y

[21-May-2010 09:49:45] <rmatte> very

[21-May-2010 09:49:58] <rmatte> good thing I don't run zenmodeler on a schedule

[21-May-2010 09:50:03] <rmatte> or it would really be messing me up

[21-May-2010 09:50:46] <LarsN> rmatte: real solaris, or opensolaris?

[21-May-2010 09:50:48] <LarsN> and version?

[21-May-2010 09:50:58] <rmatte> real solaris with net-snmp installed on it

[21-May-2010 09:51:22] <LarsN> rmatte: if I do my upgrade this weekend, I may end up modeling three opensolaris machines also running net-snmp.

[21-May-2010 09:51:32] <LarsN> I'll contribute to your ticket if I run into any problems.

[21-May-2010 09:52:14] <rmatte> sounds good

[21-May-2010 09:52:17] <rmatte> the more the merrier

[21-May-2010 09:53:15] <LarsN> are you around at all on the weekends?

[21-May-2010 09:53:33] <rmatte> not really

[21-May-2010 09:53:35] <rmatte> just weekdays

[21-May-2010 09:53:37] <LarsN> rgr.

[21-May-2010 09:53:49] <LarsN> I'm at least a green belt in googlefu.

[21-May-2010 09:53:49] <rmatte> I'm camping this weekend :)

[21-May-2010 09:53:57] <rmatte> hehe

[21-May-2010 09:54:05] <LarsN> if I run into troubles with my upgrade this weekdn I should be able to resolve most of them :

[21-May-2010 09:54:06] <rmatte> my googlefu is pretty darn good

[21-May-2010 09:54:23] <rmatte> if you do it probably won't be anything showstopping

[21-May-2010 09:54:32] <rmatte> my issues were more annoyances than anything

[21-May-2010 09:54:58] <rmatte> LarsN: you using rpm or stack install?

[21-May-2010 09:55:13] <mistich> just make a backup before you upgrade

[21-May-2010 09:56:15] <rmatte> LarsN: better question, does /etc/init.d/zenoss-stack exist on your servers?

[21-May-2010 09:56:37] <LarsN> oh rmatte I did find a package for *nix machines I like.  surfraw.

[21-May-2010 09:56:42] <LarsN> I have the zenoss-stack install.

[21-May-2010 09:56:52] <rmatte> k

[21-May-2010 09:57:03] <rmatte> after you do the upgrade it'll ask you if you want to start Zenoss...

[21-May-2010 09:57:04] <rmatte> say no

[21-May-2010 09:57:08] <rmatte> then issue this command...

[21-May-2010 09:57:08] <LarsN> ok.

[21-May-2010 09:57:25] <rmatte> sed -i "s/su zenoss -c/su zenoss -l -c/g" /usr/local/zenoss/zenoss/scripts/ctl.sh

[21-May-2010 09:57:39] <rmatte> that'll fix a path issue that exists with the stack install

[21-May-2010 09:57:53] <rmatte> then execute /etc/init.d/zenoss-stack start by hand afterwards

[21-May-2010 09:58:03] <rmatte> everything else should be non-showstopping like I said

[21-May-2010 09:58:23] <LarsN> perfect, thanks.

[21-May-2010 09:58:26] <rmatte> np

[21-May-2010 09:58:36] <LarsN> enjoy the camping :)

[21-May-2010 09:58:41] <rmatte> I shall hehe

[21-May-2010 09:58:45] <LarsN> this weekend, will hopefully be dry

[21-May-2010 09:58:50] <LarsN> it's been raining all week.

[21-May-2010 09:58:59] <rmatte> we're bringing enough alcohol to bring down a herd of elephants

[21-May-2010 09:59:01] <LarsN> which is cramping the Indianapolis Motor Speedway's style pretty bad.

[21-May-2010 09:59:02] <rmatte> so it should be good

[21-May-2010 09:59:15] <rmatte> ah

[21-May-2010 09:59:25] <rmatte> yeh, I can imagine

[21-May-2010 09:59:30] <rmatte> can't race in the rain

[21-May-2010 10:05:32] <Simon4> racing is more fun in the rain :)

[21-May-2010 10:05:36] <Simon4> crashes++

[21-May-2010 10:07:22] <rmatte> hehe

[21-May-2010 10:09:19] <LarsN> need zenoss to alert whenever that happens.

[21-May-2010 10:12:09] <rmatte> haha, the google logo rocks today

[21-May-2010 10:12:20] <LarsN> yeah

[21-May-2010 10:12:21] <LarsN> it does

[21-May-2010 10:14:55] <rmatte> it's cool that it's actually playable

[21-May-2010 10:15:38] <LarsN> kinda like the xkcd shell on april 1

[21-May-2010 10:15:39] <rmatte> I just beat it :)

[21-May-2010 10:18:07] <rmatte> xkcd shell?

[21-May-2010 10:18:26] <rmatte> do they have that archived somewhere?

[21-May-2010 10:18:37] <LarsN> I'm not sure

[21-May-2010 10:18:42] <Simon4> the xkcd shell is awesome

[21-May-2010 10:18:46] <LarsN> they had changed the front page to be an interactive shell :)

[21-May-2010 10:18:51] <LarsN> you could ls for comics, etc.

[21-May-2010 10:19:01] <Simon4> http://uni.xkcd.com/

[21-May-2010 10:19:05] <Simon4> there it is

[21-May-2010 10:19:23] * Simon4 suggests typing "make me a sandwich"

[21-May-2010 10:19:44] <rmatte> ah

[21-May-2010 10:19:45] <cgibbons> and they didn't even use flash for the game, good for them

[21-May-2010 10:20:21] <rmatte> google never uses flash

[21-May-2010 10:20:40] <cgibbons> they do

[21-May-2010 10:21:41] <rmatte> really?

[21-May-2010 10:21:56] <rmatte> I've never seen anything done in flash by them

[21-May-2010 10:21:57] <cgibbons> yeah in Google Finance, the interactive chart they have

[21-May-2010 10:22:01] <rmatte> ah

[21-May-2010 10:22:05] <cgibbons> very subtle, but flash

[21-May-2010 10:22:11] <rmatte> I don't use those apps so I guess that's why

[21-May-2010 10:22:57] <LarsN> isn't google.com/videos flash?

[21-May-2010 10:23:11] <LarsN> I won't count youtube since that was purchased, not developed.

[21-May-2010 10:23:20] <rmatte> it's a flash video player, but that's pretty much a standard on the net these days

[21-May-2010 10:24:23] <cgibbons> they're moving it to html5 with h.264 anyway, so that flash is short-lived. depending upon how loose you are with short :)

[21-May-2010 10:24:33] <cgibbons> the interactive chart will go the html5 route too

[21-May-2010 10:26:14] <LarsN> that's going to ruin youtube on linux :(

[21-May-2010 10:27:03] <rmatte> how so?

[21-May-2010 10:27:24] <LarsN> have you seen many linux systems that ship an h.264 decoder?

[21-May-2010 10:27:33] <LarsN> ubuntu does...

[21-May-2010 10:27:38] <rmatte> nah, but they will if everything moves over to it

[21-May-2010 10:28:02] <rmatte> they never used to even have flash as an option

[21-May-2010 10:28:49] <bigegor> this news also interesting: http://www.phoronix.com/scan.php?page=news_item&px=ODI2OA

[21-May-2010 10:29:41] <rmatte> nice

[21-May-2010 10:29:55] <rmatte> I love how google acquires stuff then opensources it

[21-May-2010 10:31:43] <rmatte> google is the biggest thing that Linux has going for it right now in terms of potentially becoming a mainstream OS

[21-May-2010 10:32:02] <rmatte> mainstream desktop OS I should say

[21-May-2010 10:32:30] <LarsN> I'd generally agree with that.

[21-May-2010 10:32:52] <LarsN> Google, and Canonical

[21-May-2010 10:32:57] <rmatte> yeh

[21-May-2010 10:33:19] <LarsN> love or hate ubuntu, they really have tried to focus on the user experience.  usually at the expense of some of the things we sysadmins look for in a server.

[21-May-2010 10:33:29] <rmatte> yeh

[21-May-2010 10:33:49] <rmatte> well, I use it for my servers and it's not that bad, there is the odd thing that drives me nuts but I can always overcome it

[21-May-2010 10:33:57] <LarsN> Fedora seems to have started down that road as well since ~ 9.  however they still have a strong focus on their open source principal.

[21-May-2010 10:34:19] <LarsN> if it's not GPL, or GPL compatible, fedora won't ship it.

[21-May-2010 10:34:26] <rmatte> yeh

[21-May-2010 10:34:27] <LarsN> which is annoying on a desktop.

[21-May-2010 10:34:33] <rmatte> very

[21-May-2010 10:34:45] <LarsN> *says the guy currently running slackware on his macbookpro*

[21-May-2010 10:34:51] <rmatte> so they won't ship flash player?

[21-May-2010 10:34:58] <LarsN> no, nor mp3

[21-May-2010 10:35:03] <rmatte> lame

[21-May-2010 10:35:15] <LarsN> you can get them via the rpmfusion repo.

[21-May-2010 10:35:20] <rmatte> I used to use gentoo as my desktop OS

[21-May-2010 10:35:25] <rmatte> and FreeBSD before that

[21-May-2010 10:35:26] <LarsN> but it's not as simple as Ubuntu's "ubuntu-restricted" package.

[21-May-2010 10:35:28] <LarsN> one shot and done.

[21-May-2010 10:35:35] <rmatte> Ubuntu is a breeze compared to those

[21-May-2010 10:35:36] <rmatte> :)

[21-May-2010 10:35:38] <LarsN> I keep trying to run FreeBSD on the desktop.

[21-May-2010 10:35:51] <LarsN> I'd really like it to work.  and every major version gets a little closer.

[21-May-2010 10:35:52] <rmatte> I used FreeBSD as my desktop for like 5 or 6 years

[21-May-2010 10:36:11] <LarsN> in desktop land, it just feels slow in ways it shouldn't.

[21-May-2010 10:36:17] <rmatte> It works fine as long as you don't mind editing source and config files to get stuff to load 50% of the time lol

[21-May-2010 10:36:33] <rmatte> I was still learning unix at the time though, so it was a good learning experience for me

[21-May-2010 10:36:47] <rmatte> I have better knowledge of FreeBSD and Linux than someone who started out on Ubuntu

[21-May-2010 10:37:13] <LarsN> same.

[21-May-2010 10:37:15] <rmatte> I didn't find it very slow in desktopland

[21-May-2010 10:37:19] <rmatte> it was comparable to Linux

[21-May-2010 10:37:21] <LarsN> I got started with slackware 0.(something)

[21-May-2010 10:37:28] <LarsN> in 93 iirc.

[21-May-2010 10:37:35] <rmatte> cool

[21-May-2010 10:38:03] <LarsN> moved to FreeBSD for servers when I came to my current employeer, but have dipped my toe in the water numerous times before then.

[21-May-2010 10:38:07] <rmatte> well, I'm just past year 10 from when I first started messing around with FreeBSD

[21-May-2010 10:38:23] <rmatte> but I'm 25, I started when I was 14 or 15

[21-May-2010 10:38:24] <LarsN> spent five or so years on OpenBSD before that.

[21-May-2010 10:38:29] <aclark> LarsN: go go go

[21-May-2010 10:38:34] <LarsN> zomg

[21-May-2010 10:38:36] <LarsN> aclark:

[21-May-2010 10:38:36] <rmatte> OpenBSD is nice

[21-May-2010 10:38:55] <LarsN> rmatte: whatever that guy tells you...  don't beleive it.  aclark is shifty.  :)

[21-May-2010 10:39:00] * LarsN kids:)

[21-May-2010 10:39:03] <aclark> heh

[21-May-2010 10:39:06] <rmatte> haha

[21-May-2010 10:39:07] <LarsN> aclark: will you be at PSE?

[21-May-2010 10:39:22] <aclark> LarsN: just for the day i am teaching unfortunately (tues)

[21-May-2010 10:39:27] <LarsN> rgr.

[21-May-2010 10:39:35] <aclark> you?

[21-May-2010 10:39:48] <LarsN> presenting on friday at the end of the day on RelStorage

[21-May-2010 10:39:54] <LarsN> 5:15-6

[21-May-2010 10:40:15] <LarsN> still not sure what I'm going to say about it :)

[21-May-2010 10:41:02] <LarsN> oops

[21-May-2010 10:41:14] <LarsN> good thing I looked....  10-10:45 am friday :)

[21-May-2010 10:41:22] <aclark> ah nice

[21-May-2010 10:41:25] <aclark> heh

[21-May-2010 10:41:27] <Simon4> LarsN: hurry up and get RelStorage working on zenoss. kthx

[21-May-2010 10:41:43] <aclark> The title of your talk should be "WTF is RelStorage?" :-)

[21-May-2010 10:41:49] <rmatte> haha

[21-May-2010 10:41:59] <rmatte> "I'll tell you WTF it is!"

[21-May-2010 10:42:04] <aclark> hah

[21-May-2010 10:42:05] <LarsN> aclark: RelStorage are you nuts!?!?!

[21-May-2010 10:42:08] <rmatte> "It's storage!"

[21-May-2010 10:42:13] <rmatte> "MONSTER STORAGE!"

[21-May-2010 10:42:19] <rmatte> then the pyrotechnics kick off

[21-May-2010 10:42:38] <LarsN> rmatte: ala the brawndo youtube commercials?

[21-May-2010 10:42:40] <aclark> relative storage…

[21-May-2010 10:43:01] <rmatte> LarsN: more like those new old spice commercials: "Kick! Block! EXPLOSIONNNNNN!"

[21-May-2010 10:43:52] <rmatte> http://www.youtube.com/watch?v=3tI4CbCniBI&feature=PlayList&p=60FA3565B68C63DF&playnext_from=PL&index=0&playnext=1

[21-May-2010 10:44:42] <rmatte> awesome

[21-May-2010 10:59:31] <coofamani> what did I just walk into

[21-May-2010 11:00:50] <rmatte> lol

[21-May-2010 11:00:59] <rmatte> Friday conversation

[21-May-2010 11:06:45] <jb> does anyone get any swap/paging graphs from linux hosts?

[21-May-2010 11:06:46] <jb> out of the box..

[21-May-2010 11:09:02] <coofamani> jb: is that in the default template? I dont have it

[21-May-2010 11:09:19] <coofamani> swap memory usage yes

[21-May-2010 11:09:23] <coofamani> swapping/paging no

[21-May-2010 11:13:45] <jb> nah its not

[21-May-2010 11:13:57] <jb> have you added any additional graphs for generic linux hosts/.

[21-May-2010 11:13:58] <jb> ?

[21-May-2010 11:19:28] <rmatte> the default linux template graphs swap

[21-May-2010 11:19:41] <rmatte> paging is the microsoft term for swap space

[21-May-2010 11:19:44] <rmatte> so it's irrelevant

[21-May-2010 11:20:37] <cgibbons> it's not just the microsoft term for it...

[21-May-2010 11:23:27] <rmatte> well, either way, it's the term that they wish to use for it whereas linux calls it swap

[21-May-2010 11:23:34] <coofamani> swap used vs swap rate is what I was referring to

[21-May-2010 11:23:35] <rmatte> I was just making the point that they are the same thing

[21-May-2010 11:23:47] <rmatte> ah

[21-May-2010 11:23:55] <rmatte> swap used is graphed by default but not swap rate

[21-May-2010 11:24:32] <rmatte> If swap rate is provided by net-snmp then you can easily add it

[21-May-2010 11:24:53] <coofamani> not sure it would be very valuable. you are either out of ram or your not

[21-May-2010 11:25:10] <coofamani> if its happening at all, then its probably an issue.

[21-May-2010 11:25:35] <rmatte> yeh

[21-May-2010 11:26:05] <rmatte> since Linux is actually intelligent about it's use of swap

[21-May-2010 11:26:10] <coofamani> heh

[21-May-2010 11:26:14] <coofamani> sshhh dont tell the msft people

[21-May-2010 11:26:18] <rmatte> unlike windows which can't tell the difference between physical and swap

[21-May-2010 11:26:24] <rmatte> lol

[21-May-2010 11:26:39] <rmatte> at least windows isn't limited to 3.25GB of RAM anymore

[21-May-2010 11:27:00] <coofamani> Windows: welcome to 1975

[21-May-2010 11:27:05] <rmatte> hehe

[21-May-2010 11:27:22] <coofamani> should we warn them about the y2k bug?

[21-May-2010 11:27:43] <rmatte> Microsoft: we don't just know where the limit is, we ARE the limit!

[21-May-2010 11:27:58] <rmatte> yeh, really

[21-May-2010 11:28:05] <SEJeff> rmatte, The low end of the limit perhaps

[21-May-2010 11:28:09] <SEJeff> Like the bottom of the limit :)

[21-May-2010 11:28:20] <coofamani> tho, I still cant get EVE running correctly in Wine...

[21-May-2010 11:28:29] <coofamani> meh, it has its good parts

[21-May-2010 11:28:36] <jb> paging is not just for microsoft :P

[21-May-2010 11:28:39] <SEJeff> Active Directory and exchange

[21-May-2010 11:28:44] <SEJeff> thats what they have going for them

[21-May-2010 11:28:45] <coofamani> deja vu

[21-May-2010 11:28:57] <SEJeff> The rest...not so much

[21-May-2010 11:29:26] <coofamani> exchange is debatable. it's just the least evil platform in a horribly done service

[21-May-2010 11:29:54] <coofamani> imagine if google released gmail for local installation. exchange would vanish in year

[21-May-2010 11:29:58] <jb> i'd say exchange is pretty valuable.. albeit overpriced and overly complex

[21-May-2010 11:30:01] <coofamani> gmail + calendar

[21-May-2010 11:30:08] <jb> but gapps is up there too..

[21-May-2010 11:30:13] <cgibbons> doubt that, google's calendar isnt as good yet

[21-May-2010 11:30:20] <coofamani> its getting there fast

[21-May-2010 11:30:38] <jb> the thing about exchange is.. it integrates with -everything- ms

[21-May-2010 11:30:51] <coofamani> thats their primary directive right there

[21-May-2010 11:30:53] <jb> which is usually important for a large corporation using AD

[21-May-2010 11:30:54] <rmatte> yeh, and the integration is done so poorly that you don't want to use it

[21-May-2010 11:30:57] <jb> no its not

[21-May-2010 11:30:59] <jb> it works.

[21-May-2010 11:31:02] <coofamani> it works

[21-May-2010 11:31:10] <coofamani> its the reason so many shops are like 98% MS

[21-May-2010 11:31:12] <rmatte> clearly you haven't seen the CRM integration

[21-May-2010 11:31:18] <coofamani> if you stick in there product set its rokken

[21-May-2010 11:31:22] <rmatte> then again Microsoft CRM is the biggest POS I've ever seen/used

[21-May-2010 11:31:31] <jb> uhm, considering we operate an exchange environment with 15,000 mailboxes.. i'd say I know if it works or not :)

[21-May-2010 11:31:41] <coofamani> yeah, but thats a small market compared to AD/exchange/fileserver

[21-May-2010 11:32:15] <cgibbons> i'd love it if we could ditch our zimbra junk & use exchange :)

[21-May-2010 11:32:33] <rmatte> meh, mail is mail

[21-May-2010 11:32:40] <jb> not really :)

[21-May-2010 11:32:42] <jb> it used ot be

[21-May-2010 11:32:45] <jb> to be..

[21-May-2010 11:32:52] <coofamani> we are over a million on it now I think

[21-May-2010 11:32:55] <jb> exchange is much more than "just mail" though

[21-May-2010 11:33:13] <jb> large corporations need more than "just mail"

[21-May-2010 11:33:17] <SEJeff> coofamani, As a pro Linux Admin, I'd have to agree with jb

[21-May-2010 11:33:28] <SEJeff> And then when you throw in sharepoint, it does everything

[21-May-2010 11:33:30] <jb> and I am a linux admin too, btw :)

[21-May-2010 11:33:35] <coofamani> same

[21-May-2010 11:33:39] <SEJeff> There is still nothing that compares with it or active directory

[21-May-2010 11:33:44] <jb> but, I understand the need for MS in a corporat eenvironment

[21-May-2010 11:33:44] <coofamani> agreed

[21-May-2010 11:33:55] <SEJeff> FreeIPAv2 from redhat will soon be equivalent from a linux pov to AD

[21-May-2010 11:34:00] <SEJeff> But it is in alpha at the moment

[21-May-2010 11:34:18] <SEJeff> It will support full on AD trust relationships via krb :)

[21-May-2010 11:34:22] <jb> nice

[21-May-2010 11:34:30] <coofamani> I just spend 2 hours battling ldap/edir/ncsd/pam.d

[21-May-2010 11:34:32] <coofamani> gimme AD

[21-May-2010 11:34:40] <jb> coofamani: no shit.. its a royal pain in the ass.

[21-May-2010 11:34:44] <SEJeff> Exactly

[21-May-2010 11:34:47] <jb> coofamani: likewise.com :)

[21-May-2010 11:35:44] <SEJeff> jb, Honestly you can do it with pam_ldap and setting up the client's krb.conf

[21-May-2010 11:36:01] <jb> yeah its doable

[21-May-2010 11:36:08] <SEJeff> They made me setup a POC before I screamed loud enough to keep using our own openldap infrastructure

[21-May-2010 11:36:19] <coofamani> POC?

[21-May-2010 11:36:21] <SEJeff> All monitored by Zenoss of course

[21-May-2010 11:36:22] <jb> proof of concept

[21-May-2010 11:36:25] <SEJeff> bingo

[21-May-2010 11:36:50] <jb> out linux/unix infrastructure is not tied to LDAP currently..

[21-May-2010 11:36:57] <SEJeff> oof

[21-May-2010 11:37:00] <jb> i manage admin users with puppet and call it a day.

[21-May-2010 11:37:01] <SEJeff> How many servers?

[21-May-2010 11:37:06] <jb> around 100.

[21-May-2010 11:37:18] <SEJeff> Ah thats not too big of an environment

[21-May-2010 11:37:19] <jb> 100 linux/aix, and about 600 windows.

[21-May-2010 11:37:23] <SEJeff> That would certainly work

[21-May-2010 11:37:28] <jb> yeah any larger and it would be a nightmare

[21-May-2010 11:37:32] <jb> it already is in some cases

[21-May-2010 11:37:35] <rmatte> We have everything tied in to LDAP

[21-May-2010 11:37:51] <jb> most of our linux/unix stuff isn't user based stuff.. so there are only a few local users (the admins) on each box

[21-May-2010 11:37:57] <jb> and i can manage those accounts with puppet

[21-May-2010 11:38:23] <rmatte> well, I also have multiple zenoss servers

[21-May-2010 11:38:27] <rmatte> so I need LDAP for auth

[21-May-2010 11:38:28] <coofamani> their new website is such a huge improvement

[21-May-2010 11:38:37] <jb> zenoss ldap leaves a lot to be desired :/

[21-May-2010 11:38:43] <coofamani> heh

[21-May-2010 11:39:11] <rmatte> jb: it works well enough once you screw around with it enough to figure it out

[21-May-2010 11:39:36] <jb> yeah, ive had a lot of problems

[21-May-2010 11:39:48] <jb> i'd like to add a AD user to zenoss, and pre-assign them roles/systems/etc

[21-May-2010 11:39:58] <coofamani> does it do group mapping in 2.5.x? havent tried

[21-May-2010 11:40:08] <jb> coofamani: I could never get it working correctly

[21-May-2010 11:40:13] <coofamani> bummer

[21-May-2010 11:40:51] <SEJeff> rmatte, I have a lot of problems with ZenOSS + ldap auth for the interface

[21-May-2010 11:40:59] <SEJeff> Even with zenoss enterrpise

[21-May-2010 11:41:31] <SEJeff> You'll randomly see nasty ldap tracebacks all through the webui. Especially during modelling when a new device is added. Everything else works very well however.

[21-May-2010 11:41:43] <rmatte> SEJeff: really?  I never see that

[21-May-2010 11:41:55] <coofamani> Hmmm. I havent seen that and use it across multiple installs

[21-May-2010 11:42:08] <rmatte> I use it across 14 installs currently

[21-May-2010 11:42:17] <rmatte> never have problems with it

[21-May-2010 11:42:22] <SEJeff> rmatte, I think it has something to do with the distributed collectors. Not sure

[21-May-2010 11:42:24] <coofamani> with a good number of eyeballs on it so ldap gets enough of a workout

[21-May-2010 11:42:30] <SEJeff> We have 1 master and 5 distributed collectors

[21-May-2010 11:42:42] <coofamani> thast the config we use and havent seen those problems

[21-May-2010 11:42:49] <SEJeff> I dunno maybe 2k monitored servers / network gear?

[21-May-2010 11:42:54] <rmatte> SEJeff: well, I've tested with distributed collectors and it worked fine, but maybe Enterprise is different

[21-May-2010 11:42:55] <coofamani> nope

[21-May-2010 11:43:27] <coofamani> unless your master is overloaded and just timing out? do you have redundant ldap servers listed in zope?

[21-May-2010 11:44:02] <SEJeff> coofamani, No, we only have 1 I think, but it isn't overloaded. Our ldap load is very small

[21-May-2010 11:44:29] <coofamani> SEJeff: I've had pain with 1 listed, no idea of the root cause, but adding a second really helped

[21-May-2010 11:44:54] <coofamani> SEJeff: that was most likely due to load on the ldap server though

[21-May-2010 11:46:51] <SEJeff> coofamani, Thanks for the tip. I'll try it out

[21-May-2010 11:46:56] <jb> SEJeff: lots of problems here too.

[21-May-2010 11:46:57] <jb> (with LDAP)

[21-May-2010 11:47:15] <coofamani> SEJeff: of course, I was referring to two *different* ldap server :)

[21-May-2010 11:47:18] <jb> it tries to authenticate with the LDAP server everytime you click throughout the interface..

[21-May-2010 11:47:23] <SEJeff> coofamani, Right

[21-May-2010 11:47:24] <coofamani> jb SEJeff what version of zen?

[21-May-2010 11:47:26] <SEJeff> Yup exactly jb

[21-May-2010 11:47:27] <jb> 2.5.2

[21-May-2010 11:47:30] <SEJeff> 2.5.2

[21-May-2010 11:47:36] <coofamani> how odd

[21-May-2010 11:47:39] <jb> so

[21-May-2010 11:47:45] <jb> if you login with your local "admin" user

[21-May-2010 11:47:46] <jb> in zenoss

[21-May-2010 11:47:58] <jb> even though thats a local user, it tries to authenticate against your LDAP server using "admin"

[21-May-2010 11:48:11] <jb> thus locking the "admin" user account out on your LDAP infrastructure

[21-May-2010 11:48:15] <coofamani> lol

[21-May-2010 11:48:39] <rmatte> ah, we use "root" instead of "admin" for our admin user in ldap

[21-May-2010 11:48:48] <rmatte> so we don't have that issue

[21-May-2010 11:50:00] <rmatte> the LDAP auth is a Zope plugin though, it's not really Zenoss specific

[21-May-2010 11:50:06] <jb> yeah

[21-May-2010 11:50:08] <rmatte> so there's nothing they can really do about that

[21-May-2010 11:50:16] <cgibbons> stop using that Zope POS

[21-May-2010 11:50:19] <rmatte> unless they were to take it and make their own version of it

[21-May-2010 11:50:23] <cgibbons> er, that was supposed to be on the inside...

[21-May-2010 11:50:27] <rmatte> haha

[21-May-2010 11:50:45] <rmatte> cgibbons: whenever you have time: http://dev.zenoss.com/trac/ticket/6779 :)

[21-May-2010 11:51:02] <SEJeff> cgibbons, Make zenoss into a giant django app!

[21-May-2010 11:51:13] <rmatte> I'll work on figuring out what's causing that

[21-May-2010 12:11:57] <LarsN> no django.

[21-May-2010 12:19:22] RoAk is now known as RoAkSoAxB

[21-May-2010 12:19:28] RoAkSoAxB is now known as andreserl

[21-May-2010 12:28:22] <LarsN> I'm content with the Zope model.  The Hierarchial nature actually lends itself well to this task :)

[21-May-2010 12:43:46] <mistich> anyone use the pythonplugin for a modeler

[21-May-2010 12:51:56] <cgibbons> yep

[21-May-2010 12:52:13] <mistich> have a example and where is the docs on it

[21-May-2010 12:53:03] <cgibbons> dunno about docs, where i'm doing it is in enterprise so nothing to share, lemme see if i can summarize

[21-May-2010 12:53:59] <mistich> I under stand

[21-May-2010 12:55:08] <cgibbons> do you have a specific question about it or starting from scratch?

[21-May-2010 12:55:36] <mistich> scratch have created snmp modelers etc.. but not a pythonplugin

[21-May-2010 12:56:00] <mistich> would love for a doc but can't find it

[21-May-2010 12:57:03] <cgibbons> top-level is you make a class that subclasses PythonPlugin. provide a collect method that does the data collection, and a process method that then handles the data returned and turns it into ObjectMap instances

[21-May-2010 12:57:56] <mistich> thanks will read thru the code

[21-May-2010 12:58:50] <cgibbons> and one thing to keep in mind is if you return a twisted Deferred from the collect method then the framework will handle the whole twisted asynchronous I/O problem for ou

[21-May-2010 13:37:14] Simon4_ is now known as Simon4

[21-May-2010 13:59:23] <jb> rocket, please fix the aix zenpack :P

[21-May-2010 13:59:26] <jb> whereever you are!

[21-May-2010 14:02:15] <rmatte> he can't

[21-May-2010 14:02:27] <rmatte> he works for Zenoss now so he can't work on community material

[21-May-2010 14:02:48] <Simon4> jb: sounds like a good excuse to learn some zenpack stuff :)

[21-May-2010 14:03:00] <SEJeff> Zenoss loves to hire directly from the community. Thats how a lot of their employees got their jobs

[21-May-2010 14:03:04] <SEJeff> Clever scheme

[21-May-2010 14:03:07] <rmatte> yup

[21-May-2010 14:03:16] <rmatte> well it makes sense, hire people who know the product

[21-May-2010 14:03:41] <rmatte> what's wrong with the AIX pack?

[21-May-2010 14:03:54] <jb> the aix zenpack is enterprise :P

[21-May-2010 14:03:57] <jb> well, the official one.

[21-May-2010 14:04:01] <jb> and its broken :/

[21-May-2010 14:04:03] <rmatte> oh

[21-May-2010 14:04:12] <rmatte> I thought you mean the community AIX pack

[21-May-2010 14:04:17] <rmatte> meant*

[21-May-2010 14:04:21] <jb> no, although he wrote that one too

[21-May-2010 14:04:25] <jb> the enterprise one is SSH based

[21-May-2010 14:04:30] <jb> due to aix's snmp suckiness

[21-May-2010 14:04:41] <jb> Simon4: ive written several zenpacks :)

[21-May-2010 14:04:55] <rmatte> ah

[21-May-2010 14:04:59] <Simon4> jb: apologies, didn't realise you were talking about an enterprise one :)

[21-May-2010 14:05:02] <jb> np

[21-May-2010 14:05:45] <Simon4> that and it's friday, I've had a glass of wine, and am bound to be a smartarse :)

[21-May-2010 14:51:29] <mistich> cgibbons thanks got it working

[21-May-2010 15:00:38] LarsN is now known as LarsN_Away

[21-May-2010 16:01:46] Simon4_ is now known as Simon4

[21-May-2010 19:58:19] forsberg is now known as fOrsberg

[22-May-2010 00:00:30] [disconnected at Sat May 22 00:00:30 2010]

[22-May-2010 00:00:30] [connected at Sat May 22 00:00:30 2010]

[22-May-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[22-May-2010 02:57:54] fOrsberg is now known as forsberg

[22-May-2010 08:57:57] <zykes-> jane__: around ?

[22-May-2010 12:47:24] forsberg is now known as fOrsberg

[22-May-2010 14:03:45] _sektorNBA is now known as sektorNBA

[23-May-2010 00:00:30] [disconnected at Sun May 23 00:00:30 2010]

[23-May-2010 00:00:31] [connected at Sun May 23 00:00:31 2010]

[23-May-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[23-May-2010 04:53:57] fOrsberg is now known as forsberg

[23-May-2010 08:52:36] forsberg is now known as fOrsberg

[23-May-2010 09:44:08] LarsN_Away is now known as LarsN

[23-May-2010 09:45:33] LarsN is now known as LarsN_Away

[23-May-2010 11:49:29] fOrsberg is now known as forsberg

[23-May-2010 12:09:02] <fus10nx> anyone here used Zabbix?

[23-May-2010 13:02:29] <frozty_sa> fu[tab][tab] I have, and it's a piece of shit

[23-May-2010 13:02:37] <frozty_sa> but hey, they're gone..

[23-May-2010 13:43:00] RoAk is now known as andreserl

[23-May-2010 19:42:32] forsberg is now known as fOrsberg

[23-May-2010 20:21:32] bzed_ is now known as bzed

[24-May-2010 00:00:30] [disconnected at Mon May 24 00:00:30 2010]

[24-May-2010 00:00:30] [connected at Mon May 24 00:00:30 2010]

[24-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[24-May-2010 03:38:26] fOrsberg is now known as forsberg

[24-May-2010 08:38:32] <nickanderson> community/about/webinars

[24-May-2010 08:38:34] <nickanderson> broken

[24-May-2010 08:38:54] <nickanderson> "passenger" Rails deployment that just works ;)

[24-May-2010 09:06:54] forsberg is now known as fOrsberg

[24-May-2010 09:34:42] Simon4_ is now known as Simon4

[24-May-2010 09:40:57] Simon4_ is now known as Simon4

[24-May-2010 10:16:45] <Kristopher> Good Morning all.

[24-May-2010 10:19:35] <mrayzenoss> morning

[24-May-2010 10:24:42] <gwb2351> I have an "OS Processes" defined (in the OS tab of the device) but zencommand (via ssh) is not checking the process... steps to debug?  (Note:  the File Systems _are_ being checked, can see that in the zencommand debug log)

[24-May-2010 10:25:16] tehhobbit is now known as l

[24-May-2010 10:25:42] l is now known as tehhobbit

[24-May-2010 10:40:39] <gwb2351> note that running zencommand on the command line w/ the same debug level os the zencommand process gives vastly different output

[24-May-2010 10:44:25] <mrayzenoss> gwb2351: what now?

[24-May-2010 10:44:48] <mrayzenoss> running zencommand run -v10 has different output than putting zencommand into debug mode?

[24-May-2010 10:45:42] <gwb2351> have I annoyed you enough yet?  :)  still muddling through debugging steps

[24-May-2010 10:46:12] <gwb2351> I have the zencommand process that zenoss launches with loglevel 10, tail -f zencommand.log

[24-May-2010 10:46:14] <mrayzenoss> since I'm about to dive back into an SSH ZenPack, I'm curious what you're seeing

[24-May-2010 10:46:38] <gwb2351> I have the zencommand command-line output for a device using -v10

[24-May-2010 10:46:55] <gwb2351> comparing the zencommand.log vs. zencommand command-line output, different things

[24-May-2010 10:47:17] <gwb2351> ...trying to debug why zencommand process is not picking up on the fact that the process is running on device

[24-May-2010 10:47:34] <gwb2351> (that a process is running on device)

[24-May-2010 10:51:46] <gwb2351> hmmm... I'll change the Cycle Time of the Tempates/OSPRocess/ps ; it's set to 300 which may be why I see df output more often than ps output of zencommand

[24-May-2010 10:53:12] <gwb2351> OK, that's one thing; the /Templates/FileSystem/disk is set to 60

[24-May-2010 10:53:22] <gwb2351> the /Templates/OSProcess/ps is set to 300

[24-May-2010 10:53:38] <gwb2351> which is why zencommand command-line is different than tail -f zencommand.log

[24-May-2010 10:53:58] * gwb2351 is slowly learning how zencommand builds it queue

[24-May-2010 10:54:23] * Simon4 is hacking around in zencommand today also

[24-May-2010 10:54:25] <Simon4> must be the day for it

[24-May-2010 10:54:46] <mrayzenoss> I'm looking at other stuff :)

[24-May-2010 10:54:55] <mrayzenoss> new Hyper-V and memcached ZenPacks

[24-May-2010 10:55:07] <mrayzenoss> and I'm working on some Chef stuff

[24-May-2010 10:55:19] <Simon4> it needs to start raining here again

[24-May-2010 10:55:33] <Simon4> I was going to do a bunch of BladeChassis zenpack stuff this weekend but it waas 26 degrees and sunny all weekend

[24-May-2010 10:55:41] <Simon4> so that went out the window :)

[24-May-2010 10:56:04] <gwb2351> is the "ps" Parser a regex or an exact match?

[24-May-2010 10:56:42] <Simon4> mrayzenoss: has your os x zenpack got to the "writing parsers and unit tests for parsers" stage yet?

[24-May-2010 10:57:44] <mrayzenoss> Simon4: sadly no.  I got to reuse the uptime parser

[24-May-2010 10:57:52] <Simon4> all good

[24-May-2010 10:58:07] <mrayzenoss> I've only done custom modeling so far

[24-May-2010 10:58:08] * Simon4 has a fair idea from the opensolaris zenpack

[24-May-2010 10:58:13] <gwb2351> I have /Processes/zencommand.py defined as Name:zencommand Regex zencommand.py but an alert is generated even though zencommand.py is running

[24-May-2010 10:58:34] <gwb2351> (Ignore Parameters: True)

[24-May-2010 11:00:11] <gwb2351> DEBUG:zen.zencommand:Queueing event {'severity': 4, 'component': 'zencommand', 'agent': 'zencommand', 'summary': 'Process not running: zencommand', 'manager': 'localhost', 'eventGroup': 'Process', 'device': 'cornea.its.utexas.edu', 'eventClass': '/Status/OSProcess'}

[24-May-2010 11:01:18] <Simon4> gwb2351: I'm guessing that the process checker is seeing "python"

[24-May-2010 11:01:27] <Simon4> as opposed to zencommand.py, which is in fact an argument

[24-May-2010 11:01:41] <Simon4> ps auxw |grep zencommand makes it look that way anywa

[24-May-2010 11:01:59] <Simon4> zenoss   10222  7.5  0.3 355028 65060 ?        Sl   17:00   0:03 /opt/zenoss/bin/python /opt/zenoss/Products/ZenRRD/zencommand.py --configfile /o

[24-May-2010 11:02:00] <Simon4> etc

[24-May-2010 11:02:12] <mrayzenoss> New Hyper-V Server R2 ZenPack: docs/DOC-5886

[24-May-2010 11:02:19] <gwb2351> ah, so set Ignore Parameters to false and see what happens

[24-May-2010 11:09:52] RoAk is now known as andreserl

[24-May-2010 11:20:25] <Jordack> Is there any differance between the Native Packages from the downlaod page and the latest stable SVN?

[24-May-2010 11:28:08] <nickanderson> I am creating a zenpack and it includes a script that is executed when a failure is detected. I copied that script into the zenpack lib directory. What path should I give my event command so that it executes the script included in the zenpack without having having to do anything on the system but install the zenpack?

[24-May-2010 11:52:28] <mrayzenoss> Jordack: there are a few post-2.5.2 changes in SVN

[24-May-2010 11:53:25] <mrayzenoss> Jordack: we've tagged a bunch of tickets as 2.5.3 candidates, but there is not a plan to actually do a 2.5.3 unless we really have to

[24-May-2010 11:53:42] <mrayzenoss> so even if they're tagged, there may or may not be 2.5.x changes

[24-May-2010 11:53:51] <mrayzenoss> checked in already

[24-May-2010 11:54:19] <mrayzenoss> I did make sure http://dev.zenoss.org/trac/ticket/5997 got backported

[24-May-2010 11:56:01] <mrayzenoss> nickanderson: You can path to your script with ${here/ZenPackManager/packs/ZenPacks.community.YOURZENPACK/path}/libexec/yourscript.sh  ${dev/manageIp}

[24-May-2010 11:56:35] <nickanderson> mrayzenoss: perfect, thanks

[24-May-2010 11:56:38] <mrayzenoss> nickanderson: I know things put in libexec get chmod +x, I'm not positive about /lib

[24-May-2010 11:56:49] <nickanderson> ok even better to know

[24-May-2010 11:56:58] <nickanderson> many thanks

[24-May-2010 11:57:04] <mrayzenoss> no problem

[24-May-2010 12:01:39] <Kristopher> Anyone around to check into this dmd error?  I didn't see this in other Zenoss installations, not sure exactly what's going on. thread/13576?tstart=0

[24-May-2010 13:37:39] <jane__> zykes- - I have raised ticket 6793 ( http://dev.zenoss.com/trac/ticket/6793 ) for the broken duration field in the notification report

[24-May-2010 14:06:30] <nickanderson> I have noticed that on 2.5.2 when adding new devices, my entries for Production State and Title do not seem to stick

[24-May-2010 14:06:50] <nickanderson> is that correct behavior?

[24-May-2010 14:07:31] <nickanderson> After adding the device, changing the values does save properly. It just doesn't seem to work to set them at the time of device addition.

[24-May-2010 14:08:54] <nickanderson> sorry, I am running 2.5.1 not 2.5.2

[24-May-2010 14:12:43] <mrayzenoss> nickanderson: I think that's an open ticket

[24-May-2010 14:12:54] <cgibbons> hmmm i hear an alien

[24-May-2010 14:13:15] <nickanderson> ok mrayzenoss, i wasnt sure if it was intended or if I was crzy

[24-May-2010 14:13:58] <nickanderson> btw, community/about/webinars

[24-May-2010 14:14:09] <nickanderson> seems the ruby app is busted

[24-May-2010 14:14:34] <mrayzenoss> ruh-roh

[24-May-2010 14:14:58] <nickanderson> i say blame it on event testing :)

[24-May-2010 14:15:15] <nickanderson> it was intentionally broken to test alarms

[24-May-2010 14:15:56] <mrayzenoss> I was unaware of any Ruby, so that's news to me

[24-May-2010 14:16:25] <SEJeff> mrayzenoss, You didn't hear? Stone Crab will be based on RoR

[24-May-2010 14:17:26] <mrayzenoss> I'm excited, I've been working with Ruby over the weekend, I think I'm ready

[24-May-2010 14:17:31] <nickanderson> find /zenoss-core | xargs sed -i 's/python/ruby/'

[24-May-2010 14:17:43] * nickanderson pukes

[24-May-2010 14:17:45] <nickanderson> ;)

[24-May-2010 14:17:47] <SEJeff> ditto :)

[24-May-2010 14:18:08] <SEJeff> But if you're just trying to sed the interpreter, you might want to tighten up that search a bit

[24-May-2010 14:18:47] <nickanderson> i figure its prolly good to update all the marketing fluf from being a python app to ruby as well

[24-May-2010 14:19:17] <mrayzenoss> Marketing would probably be excited about it, since they all came from a Ruby shop

[24-May-2010 14:19:23] <nickanderson> heh

[24-May-2010 14:19:39] <nickanderson> apparently the marketing dept invaded your webinar

[24-May-2010 14:36:50] <gwb2353> is anybody monitoring the zenoss processes (i.e. zencommand.py) using the OS tab/OS Processes bits for a device?

[24-May-2010 14:40:15] <Simon4> gwb2353: I can test it on a box here?

[24-May-2010 14:48:00] <mrayzenoss> New Memcached ZenPack: docs/DOC-5887

[24-May-2010 14:48:14] <Simon4> hurrah!

[24-May-2010 14:48:26] <Simon4> go Braudel :)

[24-May-2010 14:48:37] <mrayzenoss> he's got a Tokyo Tyrant one coming soon

[24-May-2010 14:48:43] <Simon4> mrayzenoss: broken image links it seems

[24-May-2010 14:49:09] <mrayzenoss> grrr… I can't get them off of his server for some reason

[24-May-2010 14:49:17] <Simon4> anti-leech stuff at a guess

[24-May-2010 14:49:20] <mrayzenoss> yeah

[24-May-2010 14:49:26] <mrayzenoss> luckily I'm enterprising

[24-May-2010 14:49:32] <Simon4> :)

[24-May-2010 14:50:39] <mrayzenoss> aight, FF isn't downloading them, but Chrome is

[24-May-2010 14:50:46] <Simon4> hmm

[24-May-2010 14:50:48] <Simon4> I'm in chrome

[24-May-2010 14:50:50] <Simon4> on OS X

[24-May-2010 14:51:04] <Simon4> broken image links here

[24-May-2010 14:51:26] <mrayzenoss> yeah, they had links back to his site

[24-May-2010 14:51:31] <mrayzenoss> replacing them now

[24-May-2010 14:57:09] <mrayzenoss> fixed

[24-May-2010 14:57:36] <Simon4> perfect

[24-May-2010 15:06:29] <Simon4> mrayzenoss: do you know if there's any work/roadmap to let you custom name OS Processes? if you set a regex on a commandline option (like "zencommand.py") the process gets called a dumb name "/opt/zenoss/bin/python /opt/zenoss/Products/ZenRRD"

[24-May-2010 15:08:14] <Simon4> same frustration for tomcat processes, where they all look like "/usr/bin/jdkxxx/java something something something"

[24-May-2010 15:11:02] <gwb2353> Simon4 came up with the solution; needed to add zenoss.cmd.linux.process to zproperties zcollectors

[24-May-2010 15:13:03] <Simon4> gwb2353: just to add to that, unless you're monitoring a different zenoss instanace, the failure of a daemon should result in a "$daemonname heartbeat failure" alert in zenoss also

[24-May-2010 15:18:53] <gwb2353> Simon4: I believe the "Name" is populated by the Process Class Regex, but can be changed to be more readable in the event console without mucking up what zenoss should be doing.

[24-May-2010 15:19:03] <gwb2353> and not affect the alert/monitoring

[24-May-2010 15:19:30] <Simon4> gwb2353: yeah, we have grumpy tenants who want it to look nice in the OS tab :/

[24-May-2010 15:19:43] <Simon4> would be nice to have a "display name" field as well as a "regex" field

[24-May-2010 15:20:07] <Simon4> or just list it in the OS tab as the name you configure under /Processes

[24-May-2010 15:22:34] <mrayzenoss> Simon4: haven't seen anything about that feature request

[24-May-2010 15:22:49] <Simon4> mrayzenoss: I'll hassle the enterprise people some :)

[24-May-2010 15:31:42] * Simon4 does some testing of his ZXTM zenpack

[24-May-2010 17:24:52] <zykes-> test

[24-May-2010 17:28:54] <mrayzenoss> 1.2.3

[24-May-2010 17:56:07] <MattRK> Quick ZenMib question if anyone has a second. Does Zenmib have a way to automatically clean up duplicate mibs? Everytime i run zenmib i get about 50 warnings telling me i have two coplies of [insertmibname].

[24-May-2010 17:57:04] <MattRK> I started with a fresh zenoss install. The first thing i did was add the cisco mibs zenpack. After i was done i used ZenMib to import all the mibs that came with zenoss. (Under /opt/zenoss/share/mibs/) After that i started getting these warnings everytime import mibs.

[25-May-2010 00:00:30] [disconnected at Tue May 25 00:00:30 2010]

[25-May-2010 00:00:30] [connected at Tue May 25 00:00:30 2010]

[25-May-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[25-May-2010 00:24:28] <MrWGW> I'm having some trouble SNPM modelling a Juniper EX3200

[25-May-2010 00:24:40] <MrWGW> intiially it worked, then I decided I wanted to monitor it using a different IP address it has

[25-May-2010 00:24:53] <MrWGW> and now, it detects the device and basic info about it like its location descriptor, but not the interfaces etc

[25-May-2010 01:26:37] fOrsberg is now known as forsberg

[25-May-2010 03:27:50] <fus10nx> Does ZenOSS have anything like Zabbix for SLA?

[25-May-2010 06:27:00] <fragfutter> i have some issues with event transformation. If an event already has an event class (in this case /Perf/Interface) my event mappings (that would degrade the severity for some interfaces) are not run.

[25-May-2010 06:27:35] <fragfutter> for test events generated with snmptrap and pushed into unknown, transformation work fine.

[25-May-2010 07:31:59] <Simon4> anyone know how to remove networks via zendmd?

[25-May-2010 07:32:06] * Simon4 is having a slow brain day

[25-May-2010 07:49:09] <fragfutter> Simon4: dmd.Networks.manage_delObjects('name')

[25-May-2010 07:50:50] * fragfutter is having a HP pain day

[25-May-2010 07:51:24] <mistich> good morning

[25-May-2010 07:51:45] <venturaville> dzien dobry

[25-May-2010 07:52:02] <mistich> anyone use jsonGetDeviceNames

[25-May-2010 07:52:22] <mistich> where is the code for this located?

[25-May-2010 07:52:55] <Simon4> fragfutter: yeah, found manage_deleteOrganizer works

[25-May-2010 07:53:03] * Simon4 may have removed "/" though - just reaching for the backup :)

[25-May-2010 07:53:10] <Simon4> hurrah for test boxes

[25-May-2010 07:53:34] <fragfutter> Simon4: if you did not commit afterwards just close your python shell

[25-May-2010 07:53:43] <Simon4> fragfutter: oh I committed :)

[25-May-2010 07:53:55] <fragfutter> Simon4: horay for backups.

[25-May-2010 07:53:58] <Simon4> oui

[25-May-2010 07:54:10] <Simon4> it's my scratch host anyway, I kill it fairly often

[25-May-2010 07:55:02] <fragfutter> Simon4: what about a virtual machine with snapshots (or a san with snapshots)

[25-May-2010 07:55:40] <Simon4> fragfutter: yeah, I would go down that path if zenrestore wasn't so convenient

[25-May-2010 07:55:53] <Simon4> it's a 30 sec task to just restore the zopedb from before I broke it

[25-May-2010 07:56:03] <fragfutter> Simon4: wow. ok.

[25-May-2010 07:57:39] <fragfutter> mistich: you mean the code generating the list or the url?

[25-May-2010 08:02:13] <fus10nx> So I was checking out manageengine and I was wondering if Zenoss had anything like thier probe technology - this way you can monitor devices on other networks w/o VPN's setup everywhere

[25-May-2010 08:09:30] <mistich> yes but found it thanks

[25-May-2010 08:22:38] <fragfutter> mistich: there was an or in the question ;) but if you found it.

[25-May-2010 08:23:02] <mistich> sorry doing ten things at once

[25-May-2010 08:23:16] LarsN_Away is now known as LarsN

[25-May-2010 08:23:19] <mistich> I found what I was looking for

[25-May-2010 08:56:28] <QubeZ> hello all

[25-May-2010 08:57:10] <QubeZ> i have two serial lines monitored in Zenoss (they are also bonded into a multilnk), however, when one serial line goes down I see no alerts. Any ideas? It does show monitoring true and status down but no alerts.

[25-May-2010 09:25:24] <rmatte> QubeZ: 2 options, setup that device to trap back to Zenoss when interfaces go up or down on that device

[25-May-2010 09:25:35] <rmatte> or you can do: docs/DOC-2494 (which is what I use)

[25-May-2010 09:27:26] <QubeZ> thanks

[25-May-2010 09:43:37] LarsN is now known as LarsN_Away

[25-May-2010 10:02:10] <QubeZ> rmatte: everytime I try to add the datasource following that tutorial, I lose web connectivity to my zenoss machine on port 8080. I have to restart Zenoss.

[25-May-2010 10:03:09] <rmatte> QubeZ: that's not normal

[25-May-2010 10:03:14] <rmatte> check your Zenoss log files

[25-May-2010 10:03:38] <rmatte> also, are you sure that you're just not being patient enough?

[25-May-2010 10:04:01] <QubeZ> nope, it says "error" in red at the bottom right of the web interface then i cannot telnet localhost 8080 on the machine

[25-May-2010 10:04:05] <rmatte> do you have any large ZenPacks installed such as the Cisco Mibs and ProCurve Mibs ZenPacks?

[25-May-2010 10:05:53] <QubeZ> yup

[25-May-2010 10:07:16] <rmatte> and they show up in the ZenPack list?

[25-May-2010 10:07:24] <rmatte> (which one do you have installed?)

[25-May-2010 10:08:17] <QubeZ> checking, have to restart zenoss again

[25-May-2010 10:08:33] <QubeZ> i should remove the cisco mibs, not sure what i even have them in there for

[25-May-2010 10:08:41] <rmatte> no

[25-May-2010 10:08:45] <rmatte> you shouldn't

[25-May-2010 10:08:53] <QubeZ> i have the cisco and procurve though, let me check

[25-May-2010 10:08:59] <rmatte> k

[25-May-2010 10:09:05] <rmatte> do they actually show up in your ZenPacks list?

[25-May-2010 10:09:34] <QubeZ> no they do not

[25-May-2010 10:09:57] <rmatte> ok, so you followed the zendmd steps on the ZenPack pages then when installing them?

[25-May-2010 10:10:05] <QubeZ> i have the asterisk, windows, snmptcp, snmpudp, hpmon, deviceadvdetail and the rest are standard ones

[25-May-2010 10:10:23] <QubeZ> rmatte: yup, even the step on how to avoid all the slow downs by implementing the removal process or whatever that was

[25-May-2010 10:10:31] <rmatte> k, nothing out of the ordinary then that would cause the issue you're seeing

[25-May-2010 10:10:39] <rmatte> have you checked your daemon logs?

[25-May-2010 10:10:53] <QubeZ> ill have to check them

[25-May-2010 10:10:59] <rmatte> k

[25-May-2010 10:11:40] <QubeZ> just got my power supply replaced for my 3560, wanna config the switch now... damn A.D.D. hehe

[25-May-2010 10:11:44] <QubeZ> ill be back to zenoss later

[25-May-2010 10:11:50] <rmatte> k

[26-May-2010 00:00:30] [disconnected at Wed May 26 00:00:30 2010]

[26-May-2010 00:00:30] [connected at Wed May 26 00:00:30 2010]

[26-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[26-May-2010 00:37:37] forsberg is now known as fOrsberg

[26-May-2010 01:02:07] fOrsberg is now known as forsberg

[26-May-2010 05:35:56] Simon4_ is now known as Simon4

[26-May-2010 07:51:07] <nickanderson> Good Morning

[26-May-2010 08:17:43] <fragfutter> no its raining since morning...

[26-May-2010 08:18:16] <Simon4> warm and sunny here

[26-May-2010 08:19:11] <Disorganized> a little too warm and sunny here

[26-May-2010 08:19:33] <davetoo> Raining in the Silly Valley

[26-May-2010 08:19:37] <fragfutter> Simon4: sunny in the UK? or for UK weather its sunny ;)

[26-May-2010 08:19:58] <Simon4> fragfutter: heh, slightly overcast, but warm and the sun pokes out occasionally ;)

[26-May-2010 08:20:08] <Simon4> the weekend however was stunning, blue skies and 27

[26-May-2010 08:20:15] * fragfutter has seen UK girls with barely anything one, while he though about getting a warmer jacket

[26-May-2010 08:54:15] <smbambling> morning.

[26-May-2010 08:56:20] <nickanderson> mornin

[26-May-2010 08:56:36] <smbambling> When installing the Oracle Zenpack I'm having a few issues.  Is there any log to tail to see what errors I might be getting

[26-May-2010 08:56:50] <smbambling> tnsping is working and graphing correctly but nothing else shows.

[26-May-2010 08:57:17] <nickanderson> not sure which log

[26-May-2010 08:57:32] <nickanderson> to get output you might need to turn up the log level

[26-May-2010 08:57:44] <nickanderson> or to get more useful output

[26-May-2010 08:57:56] <nickanderson> are you installing the zenpack with the gui or on the cli

[26-May-2010 08:58:10] <smbambling> the zenpack installed no issue…via the gui

[26-May-2010 08:58:26] <nickanderson> ah

[26-May-2010 08:59:15] <nickanderson> look at the logs in $ZENHOME/logs

[26-May-2010 08:59:15] <smbambling> If I run the same command from the cli it get output…but nothing seems to be graphing for the datasource

[26-May-2010 08:59:37] <nickanderson> try the zencommand.log

[26-May-2010 09:05:12] <smbambling> cool thanks nick

[26-May-2010 09:30:28] <nickanderson> How would I access the zenoss configuration settings from a script?

[26-May-2010 09:30:44] <nickanderson> Looking to use the smtp and from address settings

[26-May-2010 09:38:56] <rmatte> nickanderson: explain a bit more exactly what you're trying to do?

[26-May-2010 09:45:50] <fragfutter> nickanderson: dmd.email_from_address

[26-May-2010 09:46:26] <nickanderson> Im replacing zensnpp with a new script that just sends an email using the page field

[26-May-2010 09:46:49] <nickanderson> I dont have access to a divergent path for alerts, but i do have a desire for decreasing what buzzes my phone

[26-May-2010 09:47:26] <nickanderson> Instead of duplicating those config values I want to just use what was already input

[26-May-2010 09:48:49] <fragfutter> nickanderson: inside of zope you have a MailHost object at /zport/MailHost this allows sending of eMails

[26-May-2010 09:50:30] <fragfutter> i have no idea at which point you need to access mail, but you could open a connection to zope, get the MailHost object and use send

[26-May-2010 10:03:13] <rmatte> oh, you want to copy the alert filters over?

[26-May-2010 10:04:36] <nickanderson> all i want to do is replace zensnpp with a command that sends email, so i can escelate to the pager "number" (in my case an alternate email address) without re-entering pager numbers all the time manually

[26-May-2010 10:07:44] <rmatte> I see, why don't you just create a second alerting rule and set that email address for it, then duplicate the filter settings?

[26-May-2010 10:08:11] <rmatte> you're making this super complicated when it doesn't need to be

[26-May-2010 10:09:35] <nickanderson> It seems like a lot more work each time I have to create an alert for multiple people, also if someones "pager" number changes I dont want to go and edit all the duplicated alterts

[26-May-2010 10:10:30] <rmatte> yeh, you lost me again, I still don't really 100% get what your end goal is

[26-May-2010 10:10:36] <nickanderson> :)

[26-May-2010 10:10:52] <rmatte> are you trying to get the alerts to go to an email address which then passes the alerts along to the appropriate pager numbers or what?

[26-May-2010 10:11:10] <nickanderson> i want to escelate to sms

[26-May-2010 10:11:20] <fragfutter> and if nobody reacts it escalates and pages someone

[26-May-2010 10:11:26] <nickanderson> yeh

[26-May-2010 10:11:43] <rmatte> hmmm

[26-May-2010 10:11:49] <nickanderson> since people dont deal with setting up their own alerts, and I have to do it

[26-May-2010 10:11:54] <nickanderson> I want less data input

[26-May-2010 10:12:24] <nickanderson> and i really dont want to have a bunch of alert filters that need changed when someone changes their cell phone number or soemthing

[26-May-2010 10:12:40] <nickanderson> Id rather just update that pager field

[26-May-2010 10:13:00] <nickanderson> then all the alerts that are set to pager dont need edited again

[26-May-2010 10:13:23] <nickanderson> does that make any sense?

[26-May-2010 10:14:14] <rmatte> not much lol

[26-May-2010 10:14:18] <nickanderson> lol

[26-May-2010 10:14:30] <rmatte> without actually knowing how you currently have things setup compared to how you want them setup

[26-May-2010 10:14:54] <rmatte> you're just saying that you only want to change 1 field, that doesn't really help me to see the big picture

[26-May-2010 10:16:11] <nickanderson> perhaps im just setting stuff up wrong

[26-May-2010 10:16:15] <nickanderson> here is an example

[26-May-2010 10:16:30] <nickanderson> I have a group with 3 people in it

[26-May-2010 10:16:46] <nickanderson> and i have 4 alerting rules

[26-May-2010 10:16:47] <rmatte> k

[26-May-2010 10:16:53] <fragfutter> nickanderson: do i get it right? you create an account. you create an alert rule with a delay of 0seconds that would send an email. Then you create a second alert rule with a delay of 10minutes that sends a page.

[26-May-2010 10:17:39] <nickanderson> one that sends an email if the service is down, and one for each user in the group that has a delay of 15 minutes that sends an sms

[26-May-2010 10:17:43] <nickanderson> exactly fragfutter

[26-May-2010 10:17:55] <nickanderson> i could have just had 2 rules

[26-May-2010 10:17:58] <nickanderson> one for email

[26-May-2010 10:18:02] <nickanderson> and one for "sms"

[26-May-2010 10:18:08] <nickanderson> utalizing that page field

[26-May-2010 10:18:43] <nickanderson> and I have that working after replacing the page command with one that just sends an email to the pager address

[26-May-2010 10:19:15] <nickanderson> forget that I have it working

[26-May-2010 10:19:15] <rmatte> ah I see

[26-May-2010 10:19:25] <nickanderson> one user changes his "pager" number

[26-May-2010 10:19:29] <nickanderson> cell phone number

[26-May-2010 10:19:32] <nickanderson> secondary email address

[26-May-2010 10:19:36] <nickanderson> whatever you want to call it

[26-May-2010 10:20:02] <nickanderson> now I have to go find every alert rule that I have setup for him and change them

[26-May-2010 10:20:16] <nickanderson> I would prefer to just update the pager field on the user

[26-May-2010 10:20:43] <nickanderson> So I am aware that i can replace the pager command "zensnpp"

[26-May-2010 10:21:04] <fragfutter> so you want the zenoss-pager command to look up the pager field of the user and send an email to this address

[26-May-2010 10:21:16] <rmatte> he wants the pager command to send an email to a secondary address

[26-May-2010 10:21:17] <nickanderson> i just thought it would be nice to use the config for mail server and from user that was already input into zenoss, instead of duplicating the config variables

[26-May-2010 10:21:29] <rmatte> so that he can edit both addresses on the user

[26-May-2010 10:21:39] <rmatte> I understand now

[26-May-2010 10:21:44] <nickanderson> right

[26-May-2010 10:21:47] <fragfutter> and "abuses" the pager field to store it

[26-May-2010 10:21:50] <nickanderson> yes

[26-May-2010 10:21:55] <nickanderson> exactly

[26-May-2010 10:22:01] <nickanderson> i do not have access to a divergent path

[26-May-2010 10:22:08] <nickanderson> for alerting

[26-May-2010 10:22:46] <nickanderson> in a few searches it seems pretty common that people want to abuse that pager field :)

[26-May-2010 10:23:00] <rmatte> yeh

[26-May-2010 10:23:29] <kobalt> anyone use the nagios check ping?

[26-May-2010 10:23:34] <rmatte> well, you'll probably want to replace it with the mail command

[26-May-2010 10:23:54] <rmatte> mail -s "w00t" me@you.com

[26-May-2010 10:23:57] <rmatte> or whatever

[26-May-2010 10:26:21] <fragfutter> run a snpp-to-mail server that mails your providers mail-to-sms server ;)

[26-May-2010 10:31:43] <kobalt> looking to see if others get the same response time from check_ping

[26-May-2010 10:36:25] <fragfutter> nickanderson: you could hack Products/ZenEvents/zenactions.py

[26-May-2010 10:59:57] <Simon4> does anyone know why zenoss sometimes models switches as interface names of "GigabitEthernet2_11" and sometimes as "Gi2_11" ?

[26-May-2010 11:00:34] * Simon4 is fairly sure the switch configuration hasn't changed in the last 24 hours

[26-May-2010 11:04:37] <nickanderson> This is what I have working https://gist.github.com/c4acafaf723907eb6c53 i suppose ill just leave it like that, i dont see how to get access to the smtp variables in zenoss

[26-May-2010 11:18:37] <SEJeff> nickanderson, Doesn't zenoss just use the mta on localhost?

[26-May-2010 11:19:54] <nickanderson> its configurable SEJeff, by default yes. I was going to default my settings to what was configured in zenoss and allow specific override

[26-May-2010 11:20:18] <nickanderson> but i cant figure out howto get to those settings

[26-May-2010 11:20:33] <nickanderson> oh well, it works its just not as nice as it could be]

[26-May-2010 11:20:39] <SEJeff> nickanderson, I can help /me thinks

[26-May-2010 11:20:52] <SEJeff> Are you running this directly from the zenoss server? ie: have access directly to the dmd?

[26-May-2010 11:21:25] <SEJeff> nickanderson, Where do you configure it at?

[26-May-2010 11:21:31] <nickanderson> yes, its a replacement for the zensnpp command, so it gets executed however zensnpp pager command is

[26-May-2010 11:21:47] <Simon4> nickanderson:

[26-May-2010 11:21:49] <Simon4> >>> dmd.smtpHost

[26-May-2010 11:21:50] <Simon4> 'localhost'

[26-May-2010 11:21:50] <Simon4> >>> dmd.smtpPass

[26-May-2010 11:21:50] <Simon4> ''

[26-May-2010 11:21:50] <Simon4> >>> dmd.smtpPort

[26-May-2010 11:21:51] <Simon4> 25

[26-May-2010 11:21:51] <nickanderson> Under management, settings

[26-May-2010 11:22:02] <Simon4> same for dmd.smtpUseTLS and dmd.smtpUser

[26-May-2010 11:22:40] <nickanderson> ok how do i get the dmd into my script :)

[26-May-2010 11:23:02] <nickanderson> im not a programmer ;)

[26-May-2010 11:23:33] <nickanderson> and dmd.emailFrom has the other setting i would use

[26-May-2010 11:26:25] <Simon4> ah

[26-May-2010 11:29:05] <Simon4> nickanderson: https://gist.github.com/6d4dfd3b19451fde7cb6 - excusing the rough edit :)

[26-May-2010 11:29:22] <Simon4> it needs to run as the zenoss user obviously

[26-May-2010 11:29:39] <nickanderson> sweet

[26-May-2010 11:29:52] <nickanderson> thaks ill take that and run

[26-May-2010 11:30:08] <Simon4> that github thing is quite cool :) hadn't tried forking/etc before on there

[26-May-2010 11:30:15] <nickanderson> yeh

[26-May-2010 11:30:18] <nickanderson> its dandy

[26-May-2010 11:30:25] * Simon4 needs to play more with this stuff

[26-May-2010 11:30:48] <Simon4> if you're just reading from dmd you won't need the import of commit obviously

[26-May-2010 11:31:07] <nickanderson> http://gist.github.com/414714

[26-May-2010 11:31:18] <nickanderson> handy little shell script

[26-May-2010 11:31:25] <nickanderson> echo blah | gist

[26-May-2010 11:31:31] <Simon4> oh nice :)

[26-May-2010 11:31:46] <nickanderson> i guess the url to the original source is in there

[26-May-2010 11:31:59] <nickanderson> i was lazy and did cat ~/bin/gist | gist

[26-May-2010 11:32:03] <Simon4> :D

[26-May-2010 11:32:07] <Simon4> meta-gist

[26-May-2010 11:42:30] <rmatte> Simon4: in reply to your question about modelling, that would indicate that it's pulling the names down from ifName rather than ifDescr

[26-May-2010 11:42:42] <rmatte> lord knows why

[26-May-2010 11:42:51] <rmatte> I find the code that pulls names for interfaces can be a bit funky

[26-May-2010 11:43:08] <Simon4> cheers for the lead

[26-May-2010 11:43:17] * Simon4 will go have a dig in that code at some point

[26-May-2010 11:44:19] <rmatte> This trac ticket that I opened relates to that code: http://dev.zenoss.org/trac/ticket/5691

[26-May-2010 11:44:27] <rmatte> the code has been changed a bit since I logged the ticket

[26-May-2010 12:37:00] <mistich> good day

[26-May-2010 12:37:23] <mistich> anyone know the zendmd command to delete a datapoint out of a template

[26-May-2010 12:42:17] <mistich> I take that as a NO

[26-May-2010 12:47:07] <smbambling> sorry

[26-May-2010 13:19:10] <rmatte> mistich: one sec, let me dig it up

[26-May-2010 13:20:32] <rmatte> ah, I only have the info for deleting a template

[26-May-2010 13:21:00] <rmatte> it's probably similar though

[26-May-2010 13:21:19] <rmatte> for example, if you wanted to delete the FileSystem template you'd do...

[26-May-2010 13:21:21] <rmatte> t = dmd.getObjByPath('Devices/Server/rrdTemplates/FileSystem')

[26-May-2010 13:21:26] <rmatte> t.getPrimaryParent()._delObject(t.id)

[26-May-2010 13:21:53] <rmatte> a datapoint is technically just another object under that template

[26-May-2010 13:22:35] <rmatte> to delete the datsource under that template you'd probably just do...

[26-May-2010 13:22:48] <rmatte> t = dmd.getObjByPath('Devices/Server/rrdTemplates/FileSystem/datasources/usedBlocks')

[26-May-2010 13:22:52] <rmatte> t.getPrimaryParent()._delObject(t.id)

[26-May-2010 13:23:07] <rmatte> or whatever the path is to the datasource that you want to delete

[26-May-2010 13:23:39] <rmatte> you'd have to test it, but I think that's right

[26-May-2010 15:04:58] <rmatte> hmmm, my zenperfsnmp keeps crashing on my largest install

[26-May-2010 15:05:04] <rmatte> bah

[26-May-2010 15:16:17] mrayzenoss1 is now known as mrayzenoss

[26-May-2010 15:19:56] <rmatte> my datapoints graph is all out of wack too

[26-May-2010 15:22:37] <rmatte> seems to be recovering, I'll give it a bit and hopefully it'll be back to normal

[26-May-2010 15:58:51] <rmatte> well, I'm up to about 18500 datapoints on one of my installs now

[26-May-2010 16:09:52] <gwb2351> sometimes I wonder whay my zenperfsnmp is doing... it's at 16,1000 and there's no way I'm using that much SNMP

[26-May-2010 16:10:40] <gwb2351> I guess I do have 5 cisco 6509 switches, that could be it

[26-May-2010 16:10:44] <davetoo> um

[26-May-2010 16:10:48] <davetoo> YEAH!

[26-May-2010 16:10:51] <davetoo> duh! :)

[26-May-2010 16:11:12] <gwb2351> what production state tells zenoss to not collect data?

[26-May-2010 16:13:05] <rmatte> Decommissioned

[26-May-2010 16:13:43] <rmatte> or any custom production state below -1

[26-May-2010 16:14:54] <gwb2351> snmp seems pretty snappy so i'm not planning on worrying about it.  zenperfsnmp tells me "Sent 12315 OID requests... Cycle lasted 2.81 seconds"

[26-May-2010 16:16:05] <gwb2351> note: putting switches into maintenance decreased snmp from 12315 to 1816 OID requests

[26-May-2010 16:16:22] <gwb2351> (for the heaviest collector)

[26-May-2010 16:22:24] <rmatte> ah

[26-May-2010 16:22:33] <rmatte> I thought maintenance kept monitoring them but didn't alert

[26-May-2010 16:22:35] <rmatte> guess not

[26-May-2010 16:23:19] <rmatte> eugh, there's a ticket that's been sitting in our stack for over 3 hours and no one has even looked at it

[26-May-2010 16:23:23] <mrayzenoss>  rmatte that's some other production state… PreProduction maybe?

[26-May-2010 16:23:24] * rmatte gets his beating stick ready

[26-May-2010 16:23:40] <rmatte> mrayzenoss: yeh, Pre-Prod does that, but I thought Maintenance was similar

[26-May-2010 16:23:42] <rmatte> guess not

[26-May-2010 16:24:12] <mrayzenoss> I think the options are monitoring, alerting, dashboard

[26-May-2010 16:24:33] <mrayzenoss> err, not that order

[26-May-2010 16:24:56] <mrayzenoss> you can have monitoring with stuff show up on the dashboard with no alerts

[26-May-2010 16:25:06] <mrayzenoss> or just monitoring, no alerts or dashboard

[26-May-2010 16:28:44] <rmatte> ah

[26-May-2010 16:29:09] <rmatte> I'm up to 392 devices on my largest install now

[26-May-2010 16:29:17] <rmatte> just a few more before I break 400 :)

[26-May-2010 16:30:43] <rmatte> That particular client has a huge server farm which we'll hopefully eventually be monitoring

[26-May-2010 16:30:58] <rmatte> I'd have to use collectors for that though since that would bring us over 1000 devices

[26-May-2010 16:32:01] <rmatte> I applied the patch to get rid of the sawtooth collector graphing today

[26-May-2010 16:32:13] <rmatte> just need to restart the zenoss servers next tuesday to have it take effect

[26-May-2010 16:34:27] <rmatte> I'm also done writing upgrade scripts to upgrade my zenoss boxes to ubuntu 10.04

[26-May-2010 16:34:38] <rmatte> so I'll be doing that next Tuesday if testing goes well this week

[26-May-2010 16:34:46] <rmatte> lots of fun

[26-May-2010 16:40:14] <rmatte> so is 3.0 going to come packages with it's own install of python 2.6 in the stack installer, or is it going to use the system python?

[26-May-2010 16:40:19] <rmatte> packaged*

[26-May-2010 16:41:36] * davetoo thinks "it's about time" to the python 2.6 news :)

[26-May-2010 16:42:01] <rmatte> yeh, for sure

[26-May-2010 16:42:30] <rmatte> I'm having to copy python-ldap files in to my boxes after the upgrade so that zenoss ldap support still works in ubuntu 10.04

[26-May-2010 16:42:35] <rmatte> since they removed python 2.4 and 2.5

[26-May-2010 16:42:40] <forsberg> echo `telnet bofh.jeffballard.us 666 2>/dev/null` |grep --color -o "Your excuse is:.*$"

[26-May-2010 16:42:56] <rmatte> hehe

[26-May-2010 16:43:08] <mrayzenoss> the installers are essentially unchanged

[26-May-2010 16:43:20] <rmatte> right, but is python 2.6 going to be bundled with

[26-May-2010 16:43:21] <mrayzenoss> so they still bundle whatever they previously bundled

[26-May-2010 16:43:23] <mrayzenoss> yes

[26-May-2010 16:43:27] <rmatte> ah ok

[26-May-2010 16:43:28] <rmatte> cool

[26-May-2010 16:43:28] <mrayzenoss> not the RPM though

[26-May-2010 16:43:33] <rmatte> I actually personally prefer it that way

[26-May-2010 16:43:51] <mrayzenoss> I'm slowly working on Ubuntu packaging, but it's a tough slog

[26-May-2010 16:43:55] <rmatte> I don't want aptitude to know anything about Zenoss, that way it can't screw with it during upgrades

[26-May-2010 16:44:05] <rmatte> ;)

[26-May-2010 16:44:08] <mrayzenoss> makes sense

[26-May-2010 16:44:11] <rmatte> I like aptitude, but not that much

[26-May-2010 16:44:14] <rmatte> hehe

[26-May-2010 16:44:26] <rmatte> well, I don't trust it that much, I should say

[26-May-2010 16:44:47] <mrayzenoss> using the OS Python would make it easier to add libraries though

[26-May-2010 16:44:54] <rmatte> true

[26-May-2010 16:45:00] <rmatte> but oh well, symlinks work

[26-May-2010 16:45:01] <mrayzenoss> like ssl, odbc and libvirt

[26-May-2010 16:45:36] <rmatte> yeh, libvirt was a pain

[26-May-2010 16:45:43] <rmatte> and now I can't even use it lol

[26-May-2010 16:45:54] <rmatte> I still haven't tried that trick that someone posted about the SNMP pack

[26-May-2010 16:46:01] <rmatte> I'm going to make some time next week to try that

[26-May-2010 17:16:36] <SEJeff> mrayzenoss, Honestly I'd love the ability to yum or apt-get install zenoss

[26-May-2010 17:17:26] <mrayzenoss> deb http://dev.zenoss.org/deb main stable

[26-May-2010 17:17:33] <mrayzenoss> put that in your sources.list

[26-May-2010 17:18:05] <mrayzenoss> http://dev.zenoss.org/yum/

[26-May-2010 17:18:14] <mrayzenoss> for whatever you do on RHEL/CentOS

[26-May-2010 17:18:27] <mrayzenoss> those just re-use our existing .deb and RPM

[26-May-2010 17:18:33] <SEJeff> mrayzenoss, Can we get enterprise that way? ;)

[26-May-2010 17:18:34] <cgibbons> argh

[26-May-2010 17:18:41] <SEJeff> A stack installer in an rpm doesn't count as a real rpm

[26-May-2010 17:18:43] <SEJeff> That is a sham

[26-May-2010 17:18:48] <mrayzenoss> I know

[26-May-2010 17:18:58] <SEJeff> But a good start

[26-May-2010 17:19:12] <mrayzenoss> and if you got enterprise that way, that would imply we were supporting Debian or Ubuntu

[26-May-2010 17:19:33] <mrayzenoss> which afaik is either 0 or 1 customers

[26-May-2010 17:19:53] <SEJeff> mrayzenoss, I was using a generality

[26-May-2010 17:20:10] <mrayzenoss> understood, I like to respond to rhetorical questions as well

[26-May-2010 17:20:12] <SEJeff> But would love to mirror a yum repo of zenoss and install that on my servers

[26-May-2010 17:20:23] <SEJeff> Vs putting it in one of our internal repos like I do now :)

[26-May-2010 17:22:43] <cgibbons> all about priorities

[26-May-2010 17:23:16] <SEJeff> 3.0 is clearly priority

[26-May-2010 17:23:34] <cgibbons> and making this upgrade work is starting to become a whole lot less of a priority to me after today. stupid zope. :)

[26-May-2010 17:23:57] <mrayzenoss> maybe we can just ask all users to do fresh installs? :p

[26-May-2010 17:24:01] <SEJeff> Once you're able to use os installed python, then you can think about re-orging the files from /opt/zenoss to something FHS compliant. Then you can work on real packages and getting into repos

[26-May-2010 17:24:03] <cgibbons> sounds good!

[26-May-2010 17:24:11] <SEJeff> Then... PROFIT!

[26-May-2010 17:24:20] <mrayzenoss> SEJeff: that's on the roadmap for our build engineer

[26-May-2010 17:24:41] <mrayzenoss> like, "real soon now"

[26-May-2010 17:26:36] <SEJeff> Sweet

[26-May-2010 17:26:45] <SEJeff> See you tomorrow gents *waves*

[26-May-2010 17:27:16] <mrayzenoss> later

[26-May-2010 17:30:01] forsberg is now known as fOrsberg

[27-May-2010 00:00:30] [disconnected at Thu May 27 00:00:30 2010]

[27-May-2010 00:00:30] [connected at Thu May 27 00:00:30 2010]

[27-May-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[27-May-2010 00:00:43] <calvino.freenode.net> [freenode-info] please register your nickname...don't forget to auto-identify! http://freenode.net/faq.shtml#nicksetup

[27-May-2010 01:37:58] fOrsberg is now known as forsberg

[27-May-2010 03:40:09] Simon4_ is now known as Simon4

[27-May-2010 03:52:04] <zykes-> would there be any reason for that I can't copy a template from VMware / ESXi to a new sub-devices under Virtla Machine Host ?

[27-May-2010 04:17:37] <forsberg> anyone have qnap nas boxes in zenoss somewhere? :)

[27-May-2010 06:09:47] forsberg is now known as fOrsberg

[27-May-2010 06:34:03] <zykes-> if i wanna monitor a specific webpage

[27-May-2010 06:34:09] <zykes-> is there any way to do this in a smart way ?

[27-May-2010 06:42:32] fOrsberg is now known as forsberg

[27-May-2010 07:38:26] <zykes-> anyone know where i can view the "graph" for "Event Histograms" ?

[27-May-2010 08:02:34] LarsN_Away is now known as LarsN

[27-May-2010 08:32:24] mrayzenoss is now known as mray

[27-May-2010 08:38:09] <jonny330> I need help setting up the HP proliant zenpack anyone know how that works?

[27-May-2010 08:52:13] <chemist> zykes-: http monitor or if you want to be more aggressive synthetic web transactions

[27-May-2010 08:58:55] mrayzenoss is now known as mray

[27-May-2010 09:00:34] forsberg is now known as fOrsberg

[27-May-2010 09:01:10] mray is now known as mrayzenoss

[27-May-2010 09:03:28] LarsN is now known as LarsN_Away

[27-May-2010 09:03:36] mrayzenoss is now known as mray

[27-May-2010 09:06:59] LarsN_Away is now known as LarsN

[27-May-2010 09:10:01] mrayzenoss is now known as mray

[27-May-2010 09:14:37] <mray> maybe that'll stick this time

[27-May-2010 09:16:57] * fragfutter hands mray a pushpin

[27-May-2010 09:17:26] <mray> a bit of a hassle to change your registered account

[27-May-2010 09:19:05] LarsN is now known as LarsN_Away

[27-May-2010 09:27:45] LarsN_Away is now known as LarsN

[27-May-2010 09:41:43] <jonny330> I need help setting up the HP proliant zenpack anyone know how that works?

[27-May-2010 09:49:07] LarsN is now known as LarsN_Away

[27-May-2010 10:00:32] <mray> There are a couple of Zenoss developers here today, nzook is on the hook, but he's pretty new, so ask away

[27-May-2010 10:05:34] <Filbert> I'm seeing an issue where various component services are failing to start

[27-May-2010 10:05:44] <Filbert> they're failing with errors that seem to reference non-existant devices

[27-May-2010 10:06:13] <Filbert> I've tried a reindex, increased the zenhub connection timeouts for the most frequent offenders (I was getting timeout issues first)

[27-May-2010 10:06:23] <Filbert> but services still seem to fail with a big traceback in the logs

[27-May-2010 10:06:31] <Filbert> anyone got any advice on how I can go about debugging this further?

[27-May-2010 10:09:39] <mray> could you post the traceback on pastebin.org?

[27-May-2010 10:09:46] <Filbert> sure,one mo

[27-May-2010 10:10:27] <Filbert> http://pastebin.org/285881

[27-May-2010 10:10:53] <Filbert> ldnbh1gxendb02 is a decomissioned server

[27-May-2010 10:11:05] <Filbert> so should have been removed

[27-May-2010 10:11:09] <Filbert> I also can't find it anywhere in the webui

[27-May-2010 10:11:34] <chemist> tips for speeding up web ui anyone?

[27-May-2010 10:12:18] <rocket> chemist: what zenoss version?

[27-May-2010 10:12:35] <chemist> rocket: 2.5.2

[27-May-2010 10:13:05] <rocket> there are a few performance patches you should apply

[27-May-2010 10:13:15] pmcguire is now known as ptmcg

[27-May-2010 10:13:47] <mray> chemist: this seems to have a few good tips: docs/DOC-2521

[27-May-2010 10:14:14] <nzook> Filbert:  Have you looked for the server in dmd?

[27-May-2010 10:14:44] <rocket> @chemist: zenpatch 18589;zenpatch 18890;zenpatch 18908;zenpatch 18941

[27-May-2010 10:15:26] <Filbert> nzook: nope, is there a useful guide on how to do that?

[27-May-2010 10:15:49] <rocket> These patches were found with the following trac query.

[27-May-2010 10:15:51] <rocket> http://dev.zenoss.com/trac/query?status=closed&status=verification&group=patch_state&patch=2.5.3&col=id&col=summary&col=status&col=owner&col=priority&col=milestone&col=component&col=changetime&report=6&order=status

[27-May-2010 10:16:17] fOrsberg is now known as forsberg

[27-May-2010 10:18:00] <nzook> Filbert:  from the command line, type "zendmd".  This will get you a command prompt (it may take a bit).  Do find("ldnbh1gxendb02").

[27-May-2010 10:18:39] <Filbert> nothing returned...

[27-May-2010 10:19:14] <rocket> are you sure that was the managed name?

[27-May-2010 10:19:36] <Filbert> nope :(

[27-May-2010 10:19:44] <Filbert> it's not my original installation I'm afraid :(

[27-May-2010 10:19:49] <Filbert> lemme try by it's IP

[27-May-2010 10:19:54] * chemist applies patches

[27-May-2010 10:20:11] <Filbert> nothing returned for the IP either

[27-May-2010 10:20:23] <Filbert> or FQDN

[27-May-2010 10:20:28] <Filbert> it'd have been one of those

[27-May-2010 10:20:29] <rocket> Filbert: did you run zencheckrels to make sure your relationships are fixed?

[27-May-2010 10:20:40] <Filbert> nope

[27-May-2010 10:20:56] <Filbert> how do I do that?

[27-May-2010 10:21:09] <Filbert> oh, its' a comment

[27-May-2010 10:21:11] <Filbert> er, command

[27-May-2010 10:21:13] <rocket> Filbert: also make sure you are running 2.5.2 if you haven't upgraded .. earlier versions have problems with broken relationships

[27-May-2010 10:21:21] <Filbert> ah, we're on 2.5.1

[27-May-2010 10:21:28] <rocket> get to 2.5.2 asap

[27-May-2010 10:21:38] <rocket> 2.5.1 will eat your lunch....

[27-May-2010 10:21:52] <Filbert> lol

[27-May-2010 10:22:02] <Filbert> ok, was planning ona n upgrade earlier

[27-May-2010 10:22:14] <Filbert> but everything is such a mess here, I was hoping I could put it off for a few weeks ;)

[27-May-2010 10:22:50] <Filbert> zenchkrels running now

[27-May-2010 10:24:50] <Filbert> it seems to be fixing some stuff

[27-May-2010 10:31:42] * Filbert taps his fingers

[27-May-2010 10:31:47] <Filbert> not a very quick process... :/

[27-May-2010 10:31:53] <Filbert> although that might be the size of our ZenOss install

[27-May-2010 10:39:49] <rocket> unfortunately it may crap out if other things access the db at the same time

[27-May-2010 10:42:58] <Simon4> it only commits at the end?

[27-May-2010 10:43:21] <Simon4> or does it commit after each rel fix?

[27-May-2010 10:44:09] <rocket> cant remember .. looking...

[27-May-2010 10:45:49] <rocket> just at the end .. but it could be changed pretty easily

[27-May-2010 10:46:01] <rocket> Products/ZenUtils/CheckRelations.py

[27-May-2010 10:46:23] <Simon4> cool

[27-May-2010 10:46:48] <rocket> it does set savepoints though .. so it should resume ..

[27-May-2010 10:47:16] <rocket> Ideally you do this on a system where nothing is happening though .. no modelling all other daemons are down etc

[27-May-2010 10:47:25] <rocket> but it can take a long time

[27-May-2010 10:47:46] <Simon4> yeah, considering doing it on a dedicated spare box

[27-May-2010 10:47:57] <Simon4> jsut after zeoctl start

[27-May-2010 10:48:02] <mray> rocket: is zenchkrels documented anywhere?

[27-May-2010 10:48:32] <rocket> mray: I am not sure if it is .. ( from our perspective we dont use it too much as it is pretty slow )

[27-May-2010 10:48:58] <rocket> mray: but it is important to anyone who was running 2.5.1 to run it to fix their corrupt zodb

[27-May-2010 10:48:59] <mray> rocket: I'm gonna open a ticket for it

[27-May-2010 10:49:16] <Simon4> rocket: so suggestion is, run zenchkrels, then upgrade to 2.5.2?

[27-May-2010 10:49:20] <rocket> mray: we need to improve it ..

[27-May-2010 10:49:24] <rocket> Simon4: yes

[27-May-2010 10:49:27] <Simon4> before restarting any other daemons etc

[27-May-2010 10:49:33] <Simon4> awesome, will try early next week

[27-May-2010 10:49:43] * Simon4 would love to not have to repair relations every second breath

[27-May-2010 10:49:45] <mray> rocket: I didn't even know what it did

[27-May-2010 10:50:04] <rocket> mray: it checks zope relations and tries to repair them

[27-May-2010 10:50:31] <rocket> mray: eg there is a flaw in 2.5.1 where half of the relation could be deleted and the other half exists

[27-May-2010 10:50:45] <rocket> in certain instances

[27-May-2010 10:50:56] <rocket> specifically if making changes from zendmd in 2.5.1

[27-May-2010 10:51:29] <rocket> then when you try and follow certain relationships and it crosschecks you get tracebacks etc

[27-May-2010 10:52:00] <rocket> internally for enterprise customers we are a bit more surgical about fixing these relationships and have an adhoc tool to do so

[27-May-2010 10:52:28] <rocket> but we *always* recommend getting off 2.5.0/1 and moving to 2.5.2

[27-May-2010 10:52:43] <Filbert> looks like I know what I'm doing tomorrow then :)

[27-May-2010 10:53:24] <mray> rocket: feel free to spice up http://dev.zenoss.org/trac/ticket/6809

[27-May-2010 10:53:27] <rocket> 2.5.2 specifically fixes the issues as far as we know that the zodb can get into an inconsistent state

[27-May-2010 10:58:36] <theacolyte> Oh, hello

[27-May-2010 10:59:14] <theacolyte> Are there any plans of building WMI into core?

[27-May-2010 10:59:15] * Simon4 plots a 2.5.2 upgrade sooner than going to enterprise after reading those comments

[27-May-2010 11:00:18] <rocket> Simon4: have a support contract?

[27-May-2010 11:00:29] <mray> theacolyte: WMI is already in Core, that's how Windows event log and services are monitored

[27-May-2010 11:00:33] <Simon4> rocket: oui

[27-May-2010 11:00:42] <mray> theacolyte: or do you mean the WMI Data Source?

[27-May-2010 11:00:57] <Simon4> yes, even

[27-May-2010 11:01:19] <rocket> Simon4: let me look for a KB article for a tool that you should run instead

[27-May-2010 11:01:24] <Simon4> rocket: thanks

[27-May-2010 11:01:25] <theacolyte> mray: Ah, I thought event logs were provided by egor's plugin

[27-May-2010 11:01:38] <theacolyte> But regardless, yes the WMI datasource

[27-May-2010 11:01:53] <mray> theacolyte: nothing solid yet, but there have been discussions

[27-May-2010 11:02:13] <rocket> Simon4: "Using ZenFixIt to Correct Common Zenoss Issues"

[27-May-2010 11:02:40] * Simon4 goes hunting

[27-May-2010 11:02:43] <theacolyte> Hyperic really did it right for WMI I think - the counters are all in there, you just select them

[27-May-2010 11:02:45] <rocket> Simon4: thats a more surgical repair .. but may not catch everything

[27-May-2010 11:02:51] <theacolyte> No need to look up classes to remember where it's located

[27-May-2010 11:03:01] <Simon4> k, i"ll try it out on a copy of the db early next week

[27-May-2010 11:03:05] <rocket> zenchkrels is really intended to be the broad stroke brush

[27-May-2010 11:04:52] <rocket> mray: added some notes to that ticket for you

[27-May-2010 11:10:40] <mray> New FDS 389 LDAP Server ZenPack: docs/DOC-5889

[27-May-2010 11:36:28] <theacolyte> oh

[27-May-2010 11:36:29] <theacolyte> I have one

[27-May-2010 11:36:45] <theacolyte> Up/down monitoring for interfaces - is there any movement in that direction?

[27-May-2010 11:40:52] <Simon4> didn't rmatte write a doc on how to do that?

[27-May-2010 11:42:22] <rmatte> Simon4: I didn't write the doc, I just modified it to be more current

[27-May-2010 11:42:27] <rmatte> Chet Luther wrote the doc originally

[27-May-2010 11:42:57] <Simon4> a ha

[27-May-2010 11:43:07] <rmatte> docs/DOC-2494

[27-May-2010 11:43:09] <rmatte> that's the doc

[27-May-2010 11:43:33] <rmatte> It works great, I use it to do all my port monitoring

[27-May-2010 11:43:42] <rmatte> it's more effective than traps in my opinion

[27-May-2010 12:03:41] <theacolyte> hmm

[27-May-2010 12:03:46] <theacolyte> yeah I'll have to revisit it

[27-May-2010 12:03:55] <theacolyte> Just seems that it's a big hole in the current out of box setup

[27-May-2010 12:03:59] <theacolyte> Maybe a checkbox for up/down?

[27-May-2010 12:07:56] <rmatte> yeh, that would be good

[27-May-2010 12:08:36] <rmatte> since I can't select to monitor the utilization of ports without monitoring up/down unless I create local copies of interface templates and remove the up/down monitoring, which is basically impossible

[27-May-2010 12:08:57] <rmatte> so hopefully that gets added as a feature in the future

[27-May-2010 12:12:51] <theacolyte> yea

[27-May-2010 12:15:44] <jonny330> I need help setting up the HP proliant zenpack anyone know how that works?

[27-May-2010 12:18:09] <rmatte> jonny330: did you read the ZenPack page?

[27-May-2010 12:18:55] <jonny330> yeah i did waht it said but its not woking

[27-May-2010 12:19:00] <jonny330> *working

[27-May-2010 12:19:08] <rmatte> jonny330: it's not hard, just install the pack and drop the devices in the class that the pack creates

[27-May-2010 12:19:18] <rmatte> and make sure the templates are bound to that class

[27-May-2010 12:19:24] <rmatte> what's not working about it?

[27-May-2010 12:19:39] <jonny330> no information

[27-May-2010 12:19:46] <rmatte> did you remodel?

[27-May-2010 12:19:50] <jonny330> yes

[27-May-2010 12:19:57] <rmatte> are you sure snmp is configured properly on your devices?

[27-May-2010 12:20:10] <jonny330> now let me ask you this i used the ip address of the iLo port on the server

[27-May-2010 12:20:15] <rocket> restart zenoss after install zenpack?

[27-May-2010 12:20:20] <rmatte> I doubt that's going to work

[27-May-2010 12:20:25] <rmatte> you need the IP of the server itself

[27-May-2010 12:20:27] <jonny330> am i supposed to do is that way or point it at windows?

[27-May-2010 12:20:50] <rmatte> well, are you able to snmpwalk the ilo?

[27-May-2010 12:21:02] <jonny330> didn't try let me check

[27-May-2010 12:21:34] <rmatte> I'm honestly not sure if that pack is supposed to be pointed at the server itself or not

[27-May-2010 12:21:59] <jonny330> yeah thats what people keep telling me

[27-May-2010 12:22:19] <jonny330> i don't see why i cant just point it at the ilo port isan't that what it is there for?

[27-May-2010 12:22:21] <rmatte> bigegor wrote the pack

[27-May-2010 12:22:50] <jonny330> yeah i have been trying to talk to him

[27-May-2010 12:22:59] <rmatte> you can't "just point something" at an IP unless you know that it's actually going to dish back details

[27-May-2010 12:23:04] <rmatte> and I assume the pack uses snmp

[27-May-2010 12:23:28] <jonny330> well the ip had all the information in there it does ssh, telnet, http, etc

[27-May-2010 12:23:50] <rocket> HP Insight Management Agents on monitored system ... so that implies you connect to windows

[27-May-2010 12:24:01] <rocket> and need to have that tool installed

[27-May-2010 12:24:05] <rmatte> aha

[27-May-2010 12:24:05] <rocket> for it to work

[27-May-2010 12:24:06] <rmatte> I see

[27-May-2010 12:24:22] <rmatte> yeh I was just about to say what rocket said

[27-May-2010 12:24:28] <rmatte> it says that right on the zenpack page

[27-May-2010 12:24:29] <rmatte>  External Dependencies:

[27-May-2010 12:24:30] <rmatte>          HP Insight Management Agents on monitored system

[27-May-2010 12:24:48] <rmatte> so install that on the server, configure it properly, and point it at the server itself, not the iLO

[27-May-2010 12:25:18] <jonny330> oh i guess i didn't read it carefully

[27-May-2010 12:25:24] <rmatte> yeh, lol

[27-May-2010 12:25:34] <jonny330> well thats why we have irc right

[27-May-2010 12:25:43] <jonny330> help each other out

[27-May-2010 12:25:44] <rmatte> indeed

[27-May-2010 12:26:09] <rmatte> rocket: long time no talk, how are things in zenoss-land?

[27-May-2010 12:26:29] <jonny330> i wanted to use the dell open mange port on a couple other servers do i have to do the same thing?

[27-May-2010 12:26:35] <rocket> busy

[27-May-2010 12:26:40] <rocket> very very busy

[27-May-2010 12:26:46] <rmatte> jonny330: well, dell would be a different pack I would assume

[27-May-2010 12:27:02] <rmatte> rocket: that's good though, makes the days fly by

[27-May-2010 12:27:21] <rocket> yes it does ... always an interesting day being detective

[27-May-2010 12:27:27] <rmatte> hehe

[27-May-2010 12:28:04] <jonny330> yeah dell has a serate pack just wondering if anyone used it

[27-May-2010 12:28:39] <rmatte> nope, I haven't

[27-May-2010 12:28:45] <rmatte> though it's probably the same thing

[27-May-2010 12:28:53] <rmatte> need to make sure that you have whatever dell utility installed

[27-May-2010 12:28:55] <jonny330> that stinks

[27-May-2010 12:29:13] <rmatte> well, you don't want to have 2 devices in Zenoss for each server

[27-May-2010 12:29:21] <jonny330> i really wanted to just get the reports directly out of the iLo and open manage ports

[27-May-2010 12:29:23] <rmatte> you want to be able to monitor the hardware and software at the same time

[27-May-2010 12:29:26] <rmatte> so it only makes sense

[27-May-2010 12:29:44] <jonny330> right but those ports have all the information i would need

[27-May-2010 12:29:58] <rocket> jonny you can always create your own zenpack .. if you can grab the info from the ilo ..

[27-May-2010 12:29:59] <rmatte> not really, let's say you want to monitor processes on the servers

[27-May-2010 12:30:04] <rmatte> you can't do that via iLO

[27-May-2010 12:30:14] <jonny330> i guess i am not too concered with the software

[27-May-2010 12:30:26] <rmatte> kind of weird, but ok

[27-May-2010 12:30:37] <rmatte> how would you know if the server froze up but the hardware kept rolling

[27-May-2010 12:30:41] <rmatte> ?

[27-May-2010 12:30:51] <jonny330> i guess you wouldn't

[27-May-2010 12:30:56] <rmatte> exactly

[27-May-2010 12:30:58] <rmatte> and that is quite bad

[27-May-2010 12:31:13] <jonny330> yeah so it still wouldnt tell me what i needed

[27-May-2010 12:31:49] <jonny330> the whole point is to make sure the server is working. I just didnt' want to have to reliy on windows for anything

[27-May-2010 12:32:01] <jonny330> i hate windows

[27-May-2010 12:32:05] <rmatte> yeh well, if windows isn't working, the server isn't working

[27-May-2010 12:32:06] <rmatte> lol

[27-May-2010 12:32:33] <jonny330> right which is my biggest concern since windows is its own problem

[27-May-2010 12:32:49] <rmatte> yeh, but that just means that it needs monitoring more than anything

[27-May-2010 12:32:57] <rmatte> if you don't like windows, don't use it

[27-May-2010 12:33:00] <rmatte> if you use it, monitor it

[27-May-2010 12:33:01] <rmatte> lol

[27-May-2010 12:34:41] <jonny330> i dont' have a choice, if it was up to me i wouldn't be using widnows

[27-May-2010 12:35:12] <jonny330> all the developers here are doing sharepoint and .net apps with sql backends

[27-May-2010 12:41:39] <meatwad64> has anyone had an issue with data.fs file growing too large?

[27-May-2010 12:41:53] <rmatte> yeh, we're a very windows shop too, but my group is all linux

[27-May-2010 12:41:57] <meatwad64> the data.fs file is like 6 gb and mysql is about the same size

[27-May-2010 12:42:06] <rmatte> meatwad64: yeh, you can zeopack it to decrease the size...

[27-May-2010 12:42:13] <meatwad64> yeah how long does it take?

[27-May-2010 12:42:18] <meatwad64> i've been trying to run it

[27-May-2010 12:42:18] <rmatte> not too long

[27-May-2010 12:42:21] <meatwad64> and it just sits there

[27-May-2010 12:42:25] <rmatte> yes

[27-May-2010 12:42:27] <rmatte> let it sit there

[27-May-2010 12:42:29] <rmatte> it will finish

[27-May-2010 12:42:32] <meatwad64> ok

[27-May-2010 12:42:34] <meatwad64> thanks

[27-May-2010 12:42:34] <theacolyte> rmatte: That's OK - we're not all perfect

[27-May-2010 12:42:40] <rmatte> theacolyte: hehe

[27-May-2010 12:43:03] <rmatte> meatwad64: it'll move the old data.fs to data.fs.old

[27-May-2010 12:43:11] <rmatte> which you can remove once you confirm that the new one is working ok

[27-May-2010 12:43:27] <meatwad64> so how much space should it need?

[27-May-2010 12:43:30] <rmatte> it'll swap them on the fly, so there's no need to restart zenoss or anything after it's done

[27-May-2010 12:43:36] <rmatte> not much

[27-May-2010 12:43:45] <rmatte> what size is your Data.fs right now?

[27-May-2010 12:43:48] <rmatte> prior to the zeopack

[27-May-2010 12:44:39] <meatwad64> ok i'm getting this error

[27-May-2010 12:44:40] <meatwad64> [zenoss@zenoss bin]$ python zeopack.py -h localhost -p 8100

[27-May-2010 12:44:40] <meatwad64> [Errno 2] No usable temporary directory found in ['/tmp', '/var/tmp', '/usr/tmp', '/usr/local/zenoss/zenoss/bin']

[27-May-2010 12:44:52] <meatwad64> does that mean it doesn't have enough space to do it?

[27-May-2010 12:44:57] <rmatte> you don't have a /tmp directory?

[27-May-2010 12:45:08] <meatwad64> no i have one

[27-May-2010 12:46:23] <rmatte> how much space is available in it?

[27-May-2010 13:06:40] <rmatte> ?

[27-May-2010 14:05:04] <mray>  /msg nickserv info

[27-May-2010 14:05:09] <mray> heh

[27-May-2010 14:05:13] <rmatte> nice

[27-May-2010 14:05:14] <rmatte> :P

[27-May-2010 14:05:20] <rmatte> lol

[27-May-2010 14:07:53] mray is now known as mrayzenoss

[27-May-2010 14:07:59] mrayzenoss is now known as mray

[27-May-2010 14:16:33] mray is now known as mrayzenoss

[27-May-2010 14:21:32] <theacolyte>  /msg mray hello

[27-May-2010 14:21:53] mrayzenoss is now known as mray

[27-May-2010 14:21:57] * mray sighs

[27-May-2010 14:22:34] <theacolyte> hehe

[27-May-2010 14:22:53] <mray> if I login as mray, I don't get my operator roles

[27-May-2010 14:23:02] <mray> even though I've grouped mray and mrayzenoss

[27-May-2010 14:23:12] <rmatte> check the channel access list

[27-May-2010 14:23:23] <rmatte> it's probably specifically defined for mrayzenoss

[27-May-2010 14:23:37] <mray> yeah, when I'm logged in as mrayzenoss I can't add mray

[27-May-2010 14:23:45] <mray> and when I'm mray, it says I'm already ops

[27-May-2010 14:23:46] <rmatte> weird

[27-May-2010 14:23:51] <mray> meh, whatever

[27-May-2010 14:23:57] <rmatte> you're showing as op right now

[27-May-2010 14:24:57] <rmatte> well, freenode has failed you, looks like we're going to have to move the whole channel over to my IRC network

[27-May-2010 14:25:02] <rmatte> come along people, single file

[27-May-2010 14:25:12] <rmatte> >:)

[27-May-2010 14:30:03] <mray> yeah, I'm op because I switched to mrayzenoss and then back to mray

[27-May-2010 14:31:17] <theacolyte> haha

[27-May-2010 14:31:23] <theacolyte> Your network lacks sir

[27-May-2010 14:34:00] <rmatte> you haven't seen my network :P

[27-May-2010 14:36:01] <theacolyte> hehe

[27-May-2010 14:55:36] <tecs> zenoss has really improved since the last time i used it

[27-May-2010 14:55:49] <rmatte> wait until the next update, totally new interface

[27-May-2010 14:56:12] <tecs> yeah, im not sure what to think about the new interface

[27-May-2010 14:56:45] <tecs> what were the issues with the old interface?

[27-May-2010 14:57:14] <rmatte> usability and looks

[27-May-2010 15:00:33] <tecs> fair enough

[27-May-2010 15:02:05] <rmatte> though I honestly don't have any issues with it

[27-May-2010 15:02:08] <rmatte> but the new one is slick

[27-May-2010 15:12:07] <rmatte> man, the snmp ip service monitoring is horrible

[27-May-2010 15:12:16] <rmatte> I get so many false positives with it, it's basically useless

[27-May-2010 15:12:56] <rmatte> can't wait until the nmap plugin is patched

[27-May-2010 15:21:50] mray is now known as mrayzenoss

[27-May-2010 15:21:51] mrayzenoss is now known as mray

[27-May-2010 15:24:43] <tecs> rmatte, it uses snmp?

[27-May-2010 15:24:55] <tecs> its not just, making a tcp connection and seeing what happens

[27-May-2010 15:27:26] <rmatte> no

[27-May-2010 15:27:27] <rmatte> it's not

[27-May-2010 15:27:37] <rmatte> that's what the nmap plugin does, but it's broken

[27-May-2010 15:27:48] <rmatte> and there's been a ticket open to fix it with a patch for quite a while

[27-May-2010 16:02:33] <meatwad64> is 3.0 just a UI change? i was using the demo before and the reports didn't seem any different

[27-May-2010 16:06:47] meatwad64 is now known as akertis

[27-May-2010 16:07:32] <cgibbons> reports are one of the things that are using backwards compatibility hooks into the new UI; so yes they're pretty much the same, bug fixes aside.

[27-May-2010 16:13:12] <mray> akertis: reporting is being considered as the #1 priority for the post 3.0 release

[27-May-2010 16:13:26] <rmatte> thank the heavens lol

[27-May-2010 16:13:54] <akertis> great to hear!

[27-May-2010 16:14:29] <cgibbons> you have greater faith than I do that the priority wont' change again, matt :)

[27-May-2010 16:14:48] <mray> well, moving something out of #1 means it's likely to stay near the top

[27-May-2010 16:15:21] <cgibbons> or canceling it completely :)

[27-May-2010 16:15:26] <mray> touche

[27-May-2010 16:15:42] <cgibbons> but I agree, I think it's likely to happen

[27-May-2010 16:15:58] <mray> and with every developer dedicated to working on it ;)

[27-May-2010 16:20:35] * Simon4 will donate beer to the reporting cause

[27-May-2010 16:20:59] * mray will drink said beers on behalf of the developers, they should be working

[27-May-2010 16:21:04] <Simon4> :)

[27-May-2010 16:22:28] * rmatte will donate beer hats so that the developers can drink their beer while they type, thus thwarting mray's plan

[27-May-2010 16:22:41] <mray> lol

[27-May-2010 16:23:34] <rmatte> well, I need to head off to an appointment, later folks

[27-May-2010 16:23:53] <Simon4> need to limit the beer to keep them on the ballmer peak: http://xkcd.com/323/

[27-May-2010 16:24:06] <rmatte> I was going to mention the ballmer peak

[27-May-2010 16:24:08] <rmatte> lol

[27-May-2010 16:28:16] <mray> hmm… I think I found the peak many times when I was in college

[27-May-2010 16:28:26] <mray> now I just get tired and go to sleep

[27-May-2010 16:28:54] <Simon4> I find it occasionally, but it's fairly short lived nowadays - I also hit the sleep thing

[27-May-2010 16:29:47] <mray> someday I'll get a full night's sleep

[27-May-2010 16:29:59] * mray has a 6-month old

[27-May-2010 16:30:18] <rocket> drink more beer matt

[27-May-2010 16:33:05] <mray> I'm very unlikely to get to Ballmer's Peak

[27-May-2010 16:33:22] <mray> that BAC takes several beers

[27-May-2010 16:33:23] <Simon4> I think he was more thinking getting to sleep ;)

[27-May-2010 16:33:42] <Simon4> enough beers and even a 6 month old won't wake you

[27-May-2010 17:25:31] <theacolyte> mray: DEVELOPERS DEVELOPERS DEVELOPERS DEVELOPERS

[27-May-2010 17:26:01] <theacolyte> oh, that xkcd is different

[27-May-2010 17:26:22] <theacolyte> http://www.youtube.com/watch?v=8To-6VIJZRE

[27-May-2010 17:26:29] <theacolyte> it's old... but still good

[27-May-2010 17:41:45] <chemist> if Zenoss knows the network topology and the dependency between devices, shouldn't event suppression just work?

[27-May-2010 17:42:03] <mray> not every monitoring daemon checks the topology

[27-May-2010 17:42:23] <rmatte> Static topology really needs to get implemented at some point

[27-May-2010 17:42:31] <rmatte> where you can manually set the topology

[27-May-2010 17:42:37] <chemist> yes

[27-May-2010 17:42:44] <mray> yeah, I put that in as a feature request for the next release

[27-May-2010 17:43:00] <theacolyte> that'd be nice

[27-May-2010 17:43:10] <rmatte> I mean, layer 2 would be awesome and all, but we really just need something that works

[27-May-2010 17:43:16] <chemist> so can it be done with transforms?

[27-May-2010 17:43:21] <rmatte> and manual dependencies are the simplest way

[27-May-2010 17:43:26] <theacolyte> In some topologies, the way zenoss does it won't work

[27-May-2010 17:43:31] <theacolyte> like vpn tunnels

[27-May-2010 17:43:35] <rmatte> chemist: it could, but that would be uber messy and time consuming

[27-May-2010 17:44:00] <chemist> theacolyte: that is my problem

[27-May-2010 17:44:10] <rmatte> anyways, I have to go BBQ, bbl

[27-May-2010 17:44:53] <mray> chemist: blogs/zenossblog/2009/05/28/tip-of-the-month-event-suppression

[27-May-2010 17:47:10] <chemist> mray: I use the Location transform wiht mixed results

[27-May-2010 17:48:00] <mray> gotta run

[27-May-2010 17:48:27] <chemist> bye

[27-May-2010 18:23:03] forsberg is now known as fOrsberg

[27-May-2010 19:07:49] <rmatte> back

[27-May-2010 22:08:41] LarsN_Away is now known as LarsN

[27-May-2010 22:10:06] LarsN is now known as LarsN_Away

[27-May-2010 22:43:31] chowmein is now known as chowmeined

[28-May-2010 00:00:30] [disconnected at Fri May 28 00:00:30 2010]

[28-May-2010 00:00:30] [connected at Fri May 28 00:00:30 2010]

[28-May-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[28-May-2010 00:00:44] <verne.freenode.net> [freenode-info] channel flooding and no channel staff around to help? Please check with freenode support: http://freenode.net/faq.shtml#gettinghelp

[28-May-2010 02:40:50] fOrsberg is now known as forsberg

[28-May-2010 03:37:17] <tehhobbit> why doesnt all of zenoss scripts respect the configfiles (when setting zenhub to anything but localhost the upgrade scripts and zenpack stuff still wants to use localhost

[28-May-2010 04:55:14] <fragfutter> where is the definition of severity? Error=4, Warning=3 or whatever the values are.

[28-May-2010 05:05:53] <chemist> fragfutter: settings

[28-May-2010 05:05:57] <fragfutter> ZenEvents/ZenEventClasses.py (for the record)

[28-May-2010 05:07:04] <fragfutter> chemist: thanks. but i meant the event severity. that is hard coded.

[28-May-2010 05:10:04] igustin is now known as hidden

[28-May-2010 05:10:22] hidden is now known as igustin

[28-May-2010 08:08:36] <kobalt> hey guys morning

[28-May-2010 08:09:49] <kobalt> after upgrading to 2.5.2 I am getting error connecting to server when I look at event history

[28-May-2010 08:19:30] <kobalt> anyone awake =)

[28-May-2010 08:19:55] <fragfutter> yes, but i can only guess

[28-May-2010 08:20:36] <fragfutter> the channel normaly wakes up in two hours

[28-May-2010 08:20:53] <kobalt> yeah figured that Im in early do this upgrade

[28-May-2010 08:21:08] <Simon4> kobalt: only thought from a memory long times ago is maybe try a reindex() in zendmd?

[28-May-2010 08:21:15] <mrayzenoss> kobalt: what did you upgrade from?

[28-May-2010 08:21:19] <kobalt> 251

[28-May-2010 08:21:23] <kobalt> 2.5.1

[28-May-2010 08:21:24] <fragfutter> do you have the same problem when looking at the dashboard? Is the database running?

[28-May-2010 08:21:27] <kobalt> to 2.5.1

[28-May-2010 08:21:31] <kobalt> err 2.5.2

[28-May-2010 08:21:34] <mrayzenoss> kobalt: is dropping events from the DB an option?

[28-May-2010 08:21:40] <mrayzenoss> my guess is there's an encoding error

[28-May-2010 08:21:57] <kobalt> let me export the db then I can drop it

[28-May-2010 08:22:06] <rocket> turn on debugging in zenhub

[28-May-2010 08:22:08] <kobalt> regular console is fine

[28-May-2010 08:22:10] <rocket> eg zenhub debug

[28-May-2010 08:22:23] <rocket> and see if there are any related tracebacks or messages in the logs

[28-May-2010 08:24:06] <fragfutter> now that i see that mray is awake...

[28-May-2010 08:24:22] <kobalt> I dont see anything that stands out in zenhub.log

[28-May-2010 08:24:31] <kobalt> dumping the data from history

[28-May-2010 08:24:33] <rocket> restart mysql

[28-May-2010 08:24:37] <rocket> restart zenoss

[28-May-2010 08:24:39] <kobalt> I have

[28-May-2010 08:25:04] <kobalt> thought it was this bug

[28-May-2010 08:25:06] <kobalt> http://dev.zenoss.org/trac/ticket/6050

[28-May-2010 08:26:29] <fragfutter> if i want to have an event transformation for events that are already mapped (perf/interfaces), can i do this with an additional mapping (/zport/dmd/Events/Perf/Interface/instances/foobar/eventClassInstEdit)? Or do i need to edit (/zport/dmd/Events/Perf/Interface/editEventClassTransform)

[28-May-2010 08:26:42] <rocket> kobalt: what is the error you see?

[28-May-2010 08:27:12] <kobalt> fragfutter, error connecting to server when I click on the event history

[28-May-2010 08:27:43] <kobalt> err lol

[28-May-2010 08:27:48] <kobalt> that was for rocket

[28-May-2010 08:28:00] <kobalt> event console is fine

[28-May-2010 08:31:12] <rocket> kobalt: exact message please

[28-May-2010 08:31:56] <kobalt> its a black box that says error server connection error.

[28-May-2010 08:32:01] <kobalt> I have not found it in the logs

[28-May-2010 08:33:25] <kobalt> I think I found it sec

[28-May-2010 08:35:43] <kobalt> I found in even.log http://pastebin.com/VaMB2h4r

[28-May-2010 08:36:21] <kobalt> I changed the history default sort to lastTime desc

[28-May-2010 08:36:25] <kobalt> but now I get nothing

[28-May-2010 08:37:50] <rocket> I am wondering if some reason the database may not have migrated properly.  I cannot remember if there was a change between 2.5.1 and 2.5.2

[28-May-2010 08:38:07] <rocket> if you can start with a clean events database I would go that route to start troubleshooting

[28-May-2010 08:38:46] <kobalt> I think there was a change to how it reads it, they changed it to search on a index field IIRC

[28-May-2010 08:39:20] <kobalt> I am still doing the dump.... we get a lot of events hehe, when thats done I can drop it and recreate it

[28-May-2010 08:39:43] <fragfutter> kobalt: you could check the current structure

[28-May-2010 08:39:52] forsberg is now known as fOrsberg

[28-May-2010 08:39:53] <kobalt> mrayzenoss, is there a sql script to recreate the db

[28-May-2010 08:40:39] <mrayzenoss> kobalt: http://dev.zenoss.org/trac/browser/trunk/inst/zenwipe.sh wipes everything, the SQL bits are in the script

[28-May-2010 08:41:13] <kobalt> does it wipe the event console also?

[28-May-2010 08:41:32] <mrayzenoss> yeah, event console is the MySQL DB

[28-May-2010 08:41:47] <mrayzenoss> do you want the Event Console or the Event History or everything?

[28-May-2010 08:41:59] <mrayzenoss> cause here's the schema: http://dev.zenoss.org/trac/browser/trunk/Products/ZenEvents/db/zenevents.sql

[28-May-2010 08:42:10] <fragfutter> anyone an idea about my event transformation?

[28-May-2010 08:42:18] <mrayzenoss> those are stored in ZODB

[28-May-2010 08:42:29] <kobalt> just the history

[28-May-2010 08:42:36] <mrayzenoss> you can just login to the MySQL DB and prune the history

[28-May-2010 08:42:51] <mrayzenoss> 'history' is the table

[28-May-2010 08:43:15] <kobalt> yeah

[28-May-2010 08:43:18] <mrayzenoss> just 'delete * from history'

[28-May-2010 08:43:31] <mrayzenoss> or truncate it

[28-May-2010 08:43:36] <fragfutter> truncate table would be better

[28-May-2010 08:43:48] <kobalt> probably be faster

[28-May-2010 08:43:52] <kobalt> to truncate

[28-May-2010 08:44:00] <kobalt> thanks

[28-May-2010 08:45:04] <kobalt> fragfutter, it would be easiest to do it with a transform there are examples on the wiki

[28-May-2010 08:45:36] <fragfutter> kobalt: the question is if i can have it outside of the existing transform.

[28-May-2010 08:46:30] <fragfutter> that way i could add multiple transforms as seperate entries and won't need to lump them together in one textbox with a set of if-then statements

[28-May-2010 08:50:44] <kobalt> not sure on that

[28-May-2010 08:51:09] <kobalt> woot

[28-May-2010 08:51:13] <kobalt> it was the sorting

[28-May-2010 08:51:26] <kobalt> the reason it was taking so long was I was doing a backup

[28-May-2010 08:51:29] <kobalt> =)

[28-May-2010 08:52:18] <kobalt> its takes a little bit to go through 65726 records

[28-May-2010 08:52:43] <kobalt> thanks for you help all

[28-May-2010 08:53:36] <fragfutter> kobalt: shouldn't take that long. 65k records with an index on lastTime

[28-May-2010 08:53:55] <kobalt> it only takes a few seconds

[28-May-2010 08:56:38] <fragfutter> kobalt: should be milliseconds. i have 80k records. The request for the dataset (sorted) is 290ms.

[28-May-2010 08:58:36] <kobalt> hrmm

[28-May-2010 08:58:48] <kobalt> maybe I dont have something setup correctly then

[28-May-2010 09:02:37] <fragfutter> q!

[28-May-2010 09:02:54] <kobalt> lol

[28-May-2010 09:02:55] <fragfutter> wront window but closing anyway ;) bye

[28-May-2010 09:11:19] <jimbosyn> Anyone in here want to take a stab at helping me figure out how to do a custom multigraph report?

[28-May-2010 09:12:23] <jimbosyn> I have read both the admin manual and the zenoss book and neither address my need

[28-May-2010 10:24:14] fOrsberg is now known as forsberg

[28-May-2010 10:33:03] <rmatte> wasn't there a community solaris ZenPack other than the OpenSolaris one?

[28-May-2010 10:33:57] <rocket> rmatte did you have that trac ticket again I was looking at

[28-May-2010 10:34:05] <rocket> I cant find it in my history now .. :/

[28-May-2010 10:34:17] <fus10nx> This may be a silly question but I was looking at a piece of software like ZenOSS and they somehow had the ability to monitor their machines over non-directly connected networks (so no VPN or anything). Does ZenOSS have any ability to do this so you aren't sending SNMP over the internet in plain site ?

[28-May-2010 10:35:42] <fus10nx> this is the software that does it - http://www.manageengine.com/

[28-May-2010 10:35:44] <rmatte> yeh, one second

[28-May-2010 10:36:35] <rmatte> fus10nx: well, they obviously aren't connecting it by magic, so they are probably just generating a tunnel using the product itself

[28-May-2010 10:36:39] <rmatte> which is essentially the same thing

[28-May-2010 10:36:52] <fus10nx> I figured, sort of like how LogMeIn works

[28-May-2010 10:36:56] <fus10nx> so does ZenOSS have that ability ?

[28-May-2010 10:37:06] <rmatte> rocket: http://dev.zenoss.org/trac/ticket/5238#comment:12

[28-May-2010 10:37:14] <rocket> or a ssh tunnel ..

[28-May-2010 10:37:20] <rmatte> that'll bring you right to the comment where it explains how to get it rolling

[28-May-2010 10:37:22] <rmatte> yeh, or ssh

[28-May-2010 10:37:33] <fus10nx> I'm just trying to advise a client on the least expensive way to monitor devices w/o putting a cisco VPN at each location

[28-May-2010 10:38:01] <rmatte> well, you don't really need a cisco vpn, just a linux box with openswan or something

[28-May-2010 10:38:21] <rmatte> it can be done very cheaply

[28-May-2010 10:38:38] <rmatte> we just use plain IPSEC tunnels to connect to our clients, which most network devices support

[28-May-2010 10:38:44] <fus10nx> ya i guess i was trying to get to the point of not having to add extra hardware and using the existing stuff

[28-May-2010 10:38:55] <rmatte> well, what do they have currently?

[28-May-2010 10:39:03] <fus10nx> they have like 600 remote sites and at each site is just like 1 or 2 pc's

[28-May-2010 10:39:12] <fus10nx> nothing, they dial into a terminal server

[28-May-2010 10:39:16] <rmatte> I see

[28-May-2010 10:39:20] <rocket> its a lot of overhead to build into a product .. I personally am a fan of letting dedicated tools do that job for me

[28-May-2010 10:39:26] <fus10nx> so litreally just computers and a simple router and internet connection

[28-May-2010 10:39:30] <rmatte> you'd be better off with agent based hosting in that case

[28-May-2010 10:39:52] <rmatte> agent based monitoring rather

[28-May-2010 10:39:52] <fus10nx> Do you know of any free ones? I wonder if Zabbix does it

[28-May-2010 10:40:00] <rmatte> well, nagios can do it

[28-May-2010 10:40:10] <fus10nx> nagios isnt that great imo

[28-May-2010 10:40:12] <rmatte> and you can actually use the nagios agent with Zenoss, or so I've heard

[28-May-2010 10:40:16] <fus10nx> ZenOSS is the best

[28-May-2010 10:40:35] <fus10nx> well that was my next question, I suppose I could just spend some money and have someone program something so it's a one off cost

[28-May-2010 10:40:47] <rmatte> you could

[28-May-2010 10:41:05] <fus10nx> although LogMeIn has that VPN based software, Hamachi or somethinbg

[28-May-2010 10:41:08] <fus10nx> that might just do it as well

[28-May-2010 10:41:52] <fus10nx> its like $200 a year for up to 256 clients

[28-May-2010 10:42:22] <forsberg> openvpn on the zenoss box?

[28-May-2010 10:42:22] <forsberg> :P

[28-May-2010 10:42:22] <forsberg> but ok 600 locations

[28-May-2010 10:42:46] <rmatte> rocket: apply patch 16617 and then follow the instruction in comment 12

[28-May-2010 10:42:57] <rocket> I would not have openvpn on the zenoss box that is monitoring 600 locations

[28-May-2010 10:43:04] <fus10nx> the agent idea would be nice so at each location just install an agent which setups the tunnel and does the extra snmp stuff so you can get more stats/information from the box w/ the possibility of being able to run commands on the box such as reboots

[28-May-2010 10:43:41] <fus10nx> the perfect tool would be ZenOSS+LogMeIn

[28-May-2010 10:43:46] <rmatte> yeh, monitoring 600 locations each with only 2 computers or whatever is a bit ridiculous to begin with

[28-May-2010 10:43:48] <forsberg> :(

[28-May-2010 10:43:53] <rmatte> but if it must be done, agent is the way to go

[28-May-2010 10:43:56] <forsberg> ye

[28-May-2010 10:43:58] <rmatte> feeding back to a public address

[28-May-2010 10:44:05] <rmatte> via sslk

[28-May-2010 10:44:06] <rmatte> ssl*

[28-May-2010 10:44:12] <rocket> rmatte: this ticket will be defect reviewed tuesday

[28-May-2010 10:44:23] <rmatte> rocket: thanks

[28-May-2010 10:44:26] <rocket> mray already bumped its priority

[28-May-2010 10:44:27] <fus10nx> ill look into the nagios agent thing or just pain someone to write something

[28-May-2010 10:44:31] <rmatte> cool

[28-May-2010 10:44:37] <fus10nx> thx guys

[28-May-2010 10:44:43] <rocket> the devs will take it from there... crossing fingers it will make 3.0

[28-May-2010 10:44:51] <rmatte> me too

[28-May-2010 10:44:58] <rmatte> it's a pretty quick fix

[28-May-2010 10:45:06] <rmatte> most of the work is already done

[28-May-2010 10:46:41] <rmatte> rocket: didn't you write a community pack that used the Sun SNMP Agent?

[28-May-2010 10:46:56] <rocket> nope .. aix snmp

[28-May-2010 10:46:58] <rmatte> ah

[28-May-2010 10:47:15] <rmatte> man I wish I could get my hands on enterprise packs hehe

[28-May-2010 10:47:31] <rmatte> hopefully once we start making more cash I can push to get enterprise licenses

[28-May-2010 10:47:52] <rocket> ;) .. I am spoiled now .. :p

[28-May-2010 10:47:53] <mrayzenoss> rmatte: docs/DOC-5882

[28-May-2010 10:47:56] <fus10nx> here's another question, why does ZenOSS run nativly on Windows if it's all python based ?!

[28-May-2010 10:48:01] <mrayzenoss> that adds an SNMP process monitor for Solaris

[28-May-2010 10:48:30] <rmatte> mrayzenoss: ah, cool

[28-May-2010 10:48:30] <mrayzenoss> fus10nx: Zenoss is a webapp, the server is on a Linux or Unix box

[28-May-2010 10:48:54] <rmatte> yeh, Zenoss doesn't really run on windows at all

[28-May-2010 10:49:01] <mrayzenoss> it monitors Windows

[28-May-2010 10:49:03] <fus10nx> was just curious why

[28-May-2010 10:49:06] <fus10nx> but ya i gotcha

[28-May-2010 10:49:12] <rmatte> there was one project to package it for windows using some sort of emulated environment

[28-May-2010 10:49:16] <rmatte> I forget what it was called

[28-May-2010 10:49:24] <fus10nx> the virtual appliance ?

[28-May-2010 10:49:28] <rocket> fus10nx: because zenoss has many compiled binaries its not all python

[28-May-2010 10:49:30] <rmatte> nah, something else

[28-May-2010 10:49:37] <fus10nx> rocket: thanks, i gotcha

[28-May-2010 10:49:37] <rmatte> the virtual appliance is just a vm

[28-May-2010 10:49:44] <rmatte> the thing I'm talking about ran locally on the windows box

[28-May-2010 10:49:52] <rmatte> but it used some sort of emulated backend

[28-May-2010 10:49:55] <cgibbons> There are a lot of assumptions in the current Zenoss code about the operating environment being a Linux box. The binaries that it depends upon would almost all compile on Windows, but the environment assumptions are the harder problem.

[28-May-2010 10:50:07] <mrayzenoss> fus10nx: the Virtual appliance can run on VMware Player for Windows

[28-May-2010 10:50:17] <rmatte> cgibbons: yeh, it's more the paths and account

[28-May-2010 10:50:35] <fus10nx> i gotcha

[28-May-2010 10:50:45] <fus10nx> whats the recommened linux OS? Ubuntu ?

[28-May-2010 10:51:10] <mrayzenoss> Ubuntu or CentOS

[28-May-2010 10:51:20] <mrayzenoss> that's 80% of users

[28-May-2010 10:51:28] <fus10nx> thanks Matt

[28-May-2010 10:51:56] <Doublewood> Using it on FreeBSD  just fine.

[28-May-2010 10:52:14] <rmatte> Doublewood: bet you wanted to kill yourself while you were setting it up though lol

[28-May-2010 10:52:17] <mrayzenoss> Doublewood: oh yeah?  We could use a good write-up for that

[28-May-2010 10:52:29] <rmatte> yeh, good write-ups are good to have

[28-May-2010 10:52:35] <rmatte> they also help in future packaging efforts

[28-May-2010 10:52:42] <rmatte> which I intend to tackle at some point

[28-May-2010 10:52:46] <Doublewood> You know I may have a couple interns this summer that we could help out with that.

[28-May-2010 10:52:49] <rmatte> once I get my home server converted to a lab box

[28-May-2010 10:53:00] <rmatte> that would be cool

[28-May-2010 10:53:25] <rmatte> I haven't worked with ports in ages, it'll be fun to get back in to it

[28-May-2010 10:53:41] <rmatte> is FreeBSD still using ports as the default or has it moved to portage?

[28-May-2010 10:53:53] <rocket> if you ever want to be an enterprise customer use one of the supported enterprise distros it will save you money in the end

[28-May-2010 10:54:16] <rmatte> rocket: which are RHEL, CentOS and Fedora?

[28-May-2010 10:54:23] <Doublewood> rmatte...you know I set it up last summer and did not documment shit..We are HUGE BSD users so I don't remember it being that bad..

[28-May-2010 10:54:31] <rmatte> ah

[28-May-2010 10:54:44] <rmatte> I used to be a big BSD user, but I haven't touched it in years

[28-May-2010 10:54:49] <Doublewood> rmatte...they are stil using ports although we build our own now

[28-May-2010 10:54:54] <rmatte> ah

[28-May-2010 10:54:56] <rmatte> nice

[28-May-2010 10:54:57] <mrayzenoss> rmatte: RHEL/CentOS and SLES/OpenSuSE

[28-May-2010 10:54:59] <Doublewood> Still ROCK solid...

[28-May-2010 10:55:03] <rmatte> mrayzenoss: ah

[28-May-2010 10:55:13] <rocket> thanks mrayzenoss beat me to it

[28-May-2010 10:55:19] <mrayzenoss> there may be an Enterprise Ubuntu user

[28-May-2010 10:55:22] <Doublewood> I run all my SNORT IDS stuff on BSD...

[28-May-2010 10:55:23] <rmatte> we'd have a heck of a time converting to Enterprise lol

[28-May-2010 10:55:27] <mrayzenoss> and 1 OS X

[28-May-2010 10:55:37] <rmatte> unless Enterprise runs on Ubuntu

[28-May-2010 10:55:50] <mrayzenoss> rmatte: SLES/OpenSuSE is via the stack installer

[28-May-2010 10:55:55] <rmatte> ah

[28-May-2010 10:56:05] <rmatte> so technically I could use that same stack on Ubuntu?

[28-May-2010 10:56:43] <mrayzenoss> yes

[28-May-2010 10:57:30] <mrayzenoss> it's just a matter of getting Support on-board

[28-May-2010 10:57:47] <mrayzenoss> they try to keep it to a small set of platforms, so they don't have to master everything

[28-May-2010 10:58:01] <mrayzenoss> but if you can make a good enough case, they'd probably let you stay on Ubuntu

[28-May-2010 10:58:19] <mrayzenoss> like, if it's an OS problem, you'll handle it

[28-May-2010 10:58:38] <rmatte> good enough case would be "We already have 15+ VMs on Ubuntu and don't feel like re-doing all of them"

[28-May-2010 10:58:45] <rmatte> ;)

[28-May-2010 10:59:02] <rmatte> yeh, we handle our OS problems as is anyways

[28-May-2010 10:59:04] <rmatte> no big deal

[28-May-2010 10:59:26] <mrayzenoss> not every customer is as savvy as you :)

[28-May-2010 10:59:37] <forsberg> orly :)

[28-May-2010 10:59:46] <rmatte> true

[28-May-2010 11:00:22] <Doublewood> rmatte...I just shelled in to my Zenoss box...now I remember why it was so easy to set up.....It is a Suse box..sorry bout that dude...I really thought I remembered it being a BSD box

[28-May-2010 11:00:30] <rmatte> ahhh

[28-May-2010 11:00:57] <rmatte> I've seen Zenoss running on FreeBSD but people who do always end up running in to all sorts of fun problems

[28-May-2010 11:00:57] <Doublewood> But my intern offer still stands...If I get some labor I will put them to work.

[28-May-2010 11:01:12] <rmatte> and not just during setup, problems have a habbit of magically materializing

[28-May-2010 11:01:31] <rmatte> "My daemons were running fine last week, nothing has changed, but now they crash every time I try to start them"

[28-May-2010 11:01:33] <rmatte> sort of thing

[28-May-2010 11:03:29] <mrayzenoss> Doublewood: much appreciated, since I know we have about 50 FreeBSD users

[28-May-2010 11:03:47] <mrayzenoss> and would have more with better documentation (or a port)

[28-May-2010 11:04:13] <Doublewood> Cool ... maybe we will work on a port...that sounds really frustrating but kind of fun.

[28-May-2010 11:04:41] <mrayzenoss> well, this summer will have Zenoss 3.0, which is on Python 2.6 so that should be a better starting point

[28-May-2010 11:04:58] <Doublewood> cool

[28-May-2010 11:05:20] <Doublewood> anyone going to DEFCON this summer

[28-May-2010 11:07:45] <mrayzenoss> hmm… my schedule doesn't go out that far yet

[28-May-2010 11:08:07] <mrayzenoss> I'm doing CloudCamp Austin in 2 weeks, and SouthEast LinuxFest

[28-May-2010 11:08:22] <mrayzenoss> pretty sure I'm not going to Velocity

[28-May-2010 11:08:50] <mrayzenoss> DEFCON does look fun though

[28-May-2010 11:09:25] <rmatte> defcon always looks like a blast, I'll go some day

[28-May-2010 11:11:55] <Doublewood> DEFCON just finished up quals for CTF..talk about 50 hours of getting your ass kicked....we finished in 45th out of about 550 teams

[28-May-2010 11:12:20] <rmatte> cool

[28-May-2010 11:12:24] <Doublewood> They only take the top 10 team for the actual CTF event at DEFCON

[28-May-2010 11:12:33] <rmatte> ah

[28-May-2010 11:12:57] <Doublewood> everyone should play on a team at least once.  very humbling experience....

[28-May-2010 11:13:53] <Doublewood> makes you realize very quickly that you are just one little guy with some skills, and there is a whole big world out there with truly l33t skills....

[28-May-2010 11:16:14] <rmatte> yup

[28-May-2010 11:18:42] <forsberg> :P

[28-May-2010 11:19:07] <forsberg> CTF in what game?

[28-May-2010 11:21:44] <rmatte> probably some hacking challenge

[28-May-2010 11:22:29] <rmatte>  a few hacker teams cram into a huge smoky room every summer to show their stuff.

[28-May-2010 11:22:29] <rmatte> The contestants have their eyes on a group of about a dozen computers, each prepared lovingly to attract attention and repel intrusion attempts. Successfully penetrating the security of one of these machines is worth a point. The team with the most points in 48 hours wins a little cash and a lot of bragging rights.

[28-May-2010 11:22:32] <rmatte> that's the description

[28-May-2010 11:22:56] <rmatte> sounds quite fun

[28-May-2010 11:23:02] <Doublewood> CTF = Capture the Flag...

[28-May-2010 11:23:16] <rmatte> yeh, I knew that lol

[28-May-2010 11:23:18] <forsberg> ahh

[28-May-2010 11:23:23] <rmatte> but it's obviously not in the literal sense

[28-May-2010 11:23:28] <forsberg> yes yes ok, that ctf

[28-May-2010 11:23:35] <Doublewood> HUGE braggin rights...

[28-May-2010 11:23:41] <rmatte> hehe

[28-May-2010 11:24:28] <Doublewood> There is a little twist, that you can only bring like 7 or 8 of your team to Vegas...

[28-May-2010 11:24:49] <Doublewood> The winning team this year had 40 plus members in 4 different countries

[28-May-2010 11:25:20] <Doublewood> 40 that they would admit to .

[28-May-2010 11:25:41] <rmatte> ah

[28-May-2010 11:25:46] <rmatte> so they are able to connect remotely?

[28-May-2010 11:26:19] <rmatte> or are they just there to make suggestions and the 7 or 8 do the actual hacking?

[28-May-2010 11:26:26] <Doublewood> The real fun comes when another team gets an answer that your team does not have yet....Then instead of solving the problem you simply hack the other team and steal their answer...perfectly accaptable behavior in this game

[28-May-2010 11:26:37] <rmatte> haha

[28-May-2010 11:26:39] <rmatte> that's awesome

[28-May-2010 11:26:42] <Doublewood> yep

[28-May-2010 11:27:15] <Doublewood> moral of the story is not only do you have to be good but you gotta protect your stuff as well.

[28-May-2010 11:27:23] <rmatte> yup

[28-May-2010 11:27:28] <Doublewood> I think they can connect remotely.

[28-May-2010 11:27:56] <rmatte> I'd just have a printer ready, print out the answer, throw it in your back pocket, delete

[28-May-2010 11:27:57] <rmatte> :)

[28-May-2010 11:28:10] <Doublewood> My idea to win in Vegas was to hire like 100 strippers to hang out with the other teams.

[28-May-2010 11:28:15] <rmatte> lmao

[28-May-2010 11:28:19] <rmatte> that would be very expensive

[28-May-2010 11:28:24] <rmatte> especially at vegas rates

[28-May-2010 11:28:41] <Doublewood> maybe I could get a vendor to pay for it...

[28-May-2010 11:28:53] <rmatte> you'd be better off hiring 1 tranny to walk around systematically distracting them

[28-May-2010 11:29:03] <mloven> I'd like to see that expense report....

[28-May-2010 11:29:04] <Doublewood> now thats funny

[28-May-2010 11:29:48] <Doublewood> it is frowned upon but you can buy answers from other teams....

[28-May-2010 11:30:27] <rmatte> lol

[28-May-2010 11:30:34] <rmatte> for cash?

[28-May-2010 11:30:50] <Doublewood> "Dude,,,,I'll trade you my girlfriend for that wicked cool shell code"...

[28-May-2010 11:30:59] <rmatte> haha

[28-May-2010 11:31:20] <Doublewood> yea for cash seriously

[28-May-2010 11:31:29] <rmatte> fun

[28-May-2010 11:32:23] <Doublewood> OK enough fun...back to work...have a great weekend.

[28-May-2010 11:33:43] <rmatte> you too

[28-May-2010 11:57:21] <Doublewood> we have never used the e-mail feature before...when i put in our internal e-mail address and send a test it goes nowhere...is there a error log somewhere

[28-May-2010 11:59:33] <forsberg> do you have any mta on the machine ?

[28-May-2010 12:00:38] <forsberg> i mean, if it can deliver to the relay you specified, it will

[28-May-2010 12:11:23] <rmatte> Doublewood: make sure your mailserver isn't blocking the message

[28-May-2010 12:11:34] <rmatte> we had issues with our mailserver filtering messages coming from zenoss

[28-May-2010 12:11:38] <rmatte> we had to have an exception added

[28-May-2010 12:42:14] LarsN_Away is now known as LarsN

[28-May-2010 12:49:34] <Doublewood> rmatte...naw...its a little internal one we run in security...more than likely it is an MTA issue...will check in a few

[28-May-2010 12:52:44] <otakup0pe> is it possible to put an email in the pager field ? i remember being able to do this before.... i think.

[28-May-2010 12:57:54] <rmatte> k

[28-May-2010 12:58:12] <rmatte> otakup0pe: you need to replace the pager command with a mail command, but yeh

[28-May-2010 13:00:41] <otakup0pe> ah-ha ! i knew i was missing something

[28-May-2010 13:10:47] <otakup0pe> hmm what would it get set to ? i tried /usr/bin/mail $RECIPIENT but no joy

[28-May-2010 13:11:05] <rmatte> not sure, I've never done it myself

[28-May-2010 13:11:11] <rmatte> I just know the basics of what's involved

[28-May-2010 13:12:44] <otakup0pe> i guess a friday afternoon is as good a time as any to grep the zenoss source :3

[28-May-2010 13:14:30] <Doublewood> I think i have a bead on the problem...in the mail error log,,,it looks like Zenoss is trying to send mail as Zenossuser_mchesmo3  is there any way to change that to just mchesmo3

[28-May-2010 13:15:00] <Doublewood> then gets bounced as "unknown user"

[28-May-2010 13:22:13] <mrayzenoss> otakup0pe: why not just use an Event Command?

[28-May-2010 13:22:30] <otakup0pe> because i am stubborn i think. hehe.

[28-May-2010 13:22:51] <otakup0pe> i'm going to use event commands with growl though

[28-May-2010 13:23:15] <rmatte> Doublewood: the override is in Settings

[28-May-2010 13:23:20] <rmatte> From Address For Emails

[28-May-2010 13:23:41] <rmatte> is it the address that's the problem?

[28-May-2010 13:26:12] <mrayzenoss> otakup0pe: one of our QA guys uses growl that way

[28-May-2010 13:26:19] <mrayzenoss> he also has it hooked up to Trac

[28-May-2010 13:26:28] <mrayzenoss> so whenever a ticket changes he gets a notification

[28-May-2010 13:26:39] <otakup0pe> nice

[28-May-2010 13:27:02] <otakup0pe> on my dev server it plays sad-trombone.wav when web services fail. this is my favorite use thus far.

[28-May-2010 13:27:10] <rmatte> lol

[28-May-2010 13:27:49] <rmatte> that could get annoying fairly quickly

[28-May-2010 13:27:55] <otakup0pe> i turn it off when debugging ;)

[28-May-2010 13:27:59] <rmatte> hehe

[28-May-2010 13:34:09] LarsN is now known as LarsN_Away

[28-May-2010 13:48:33] <cgibbons> argh

[28-May-2010 13:49:05] <rocket> argh?

[28-May-2010 13:52:06] <cgibbons> doing some nasty hacks. i feel dirty.

[28-May-2010 13:57:08] <rmatte> lol

[28-May-2010 13:58:00] <rmatte> It would be so nice if Zenoss remembered when model info was set manually and not reset the info when remodelling

[28-May-2010 13:58:33] <rmatte> I just went through and set model info for a bunch of stuff but it insists on replacing it when remodelling

[28-May-2010 13:59:09] <Simon4_> yup, that feature is definitely "character building"

[28-May-2010 13:59:18] <rmatte> hehe

[28-May-2010 14:00:44] <rmatte> it's one of the reasons why I don't run zenmodeler as a daemon

[28-May-2010 14:10:03] <rmatte> hmmm, Zenoss is refusing to model interfaces on one out of 37 wireless controllers

[28-May-2010 14:10:09] <rmatte> it errors out when modelling, strange

[28-May-2010 15:15:13] <rmatte> upgrading my first prod box from 8.04 to 10.04, hopefully it goes smoothly so that I can do some during our next official change window

[28-May-2010 15:17:09] <theacolyte> eww

[28-May-2010 15:17:10] <theacolyte> Wait

[28-May-2010 15:17:15] <theacolyte> Wait a little longer

[28-May-2010 15:17:16] <rmatte> ?

[28-May-2010 15:17:24] <theacolyte> 8.04 LTS was a lot less buggy than I'm finding 10.04

[28-May-2010 15:17:34] <theacolyte> You may want to do the old microsoft addage, wait for SP2

[28-May-2010 15:17:35] <theacolyte> heh

[28-May-2010 15:17:36] <rmatte> I'm finding 10.04 is fine personally

[28-May-2010 15:17:44] <theacolyte> ymmv

[28-May-2010 15:17:50] <rmatte> I did the upgrade on 2 of my lab boxes and they are actually running faster after the upgrade

[28-May-2010 15:18:02] <rmatte> I wrote scripts to automatically apply some fixes after the upgrade

[28-May-2010 15:18:10] <theacolyte> ah cool

[28-May-2010 15:18:13] <rmatte> but after that's done it works fine

[28-May-2010 15:18:17] <theacolyte> it does boot very, very quickly

[28-May-2010 15:18:21] <theacolyte> Like... 2 seconds

[28-May-2010 15:18:24] <theacolyte> literally

[28-May-2010 15:18:35] <rmatte> well, mine don't boot quite that fast, Zenoss takes some time to fire up and whatever

[28-May-2010 15:18:46] <rmatte> the initial boot is quite fast though

[28-May-2010 15:18:50] <theacolyte> How are you going to do the upgrade?

[28-May-2010 15:18:58] <rmatte> what do you mean?

[28-May-2010 15:19:58] <theacolyte> moving files or updating core files or whatever

[28-May-2010 15:20:45] <rmatte> well, to perform the initial upgrade I'm just doing...

[28-May-2010 15:21:00] <rmatte> sed -i "s/(unknown,/('unknown',/g" /usr/bin/lsb_release

[28-May-2010 15:21:12] <rmatte> to patch a bug that I found in lsb_release which was preventing the upgrade

[28-May-2010 15:21:13] <rmatte> and then...

[28-May-2010 15:21:23] <rmatte> do-release-upgrade --proposed -m server

[28-May-2010 15:21:36] <rmatte> then I let that run through, selecting all default options when prompted...

[28-May-2010 15:21:51] <rmatte> and then I run a script to fix some issues

[28-May-2010 15:22:31] <rmatte> http://fpaste.org/OoGI/

[28-May-2010 15:22:37] <rmatte> that's the script to fix stuff

[28-May-2010 15:23:35] <rmatte> after all that's done the system is good to go on 10.04

[28-May-2010 15:25:00] <willwh> 10.4 <3

[28-May-2010 15:25:22] <theacolyte> ah got it

[28-May-2010 15:25:25] <rmatte> There's some stuff that I hadn't checked since I hadn't done it on a prod box yet, but I'll be checking that after this upgrade is done

[28-May-2010 15:25:25] <rmatte> if anything is messed up I can roll back to a snapshot I just took

[28-May-2010 15:25:44] <theacolyte> I ran into some issues with likewise

[28-May-2010 15:25:44] <rmatte> though I don't anticipate any issues

[28-May-2010 15:25:57] <theacolyte> Ah is there a VMCI patch?

[28-May-2010 15:26:08] <rmatte> yeh

[28-May-2010 15:26:12] <rmatte> let me find the page that I got it from

[28-May-2010 15:26:40] <theacolyte> I am totally stealing your script

[28-May-2010 15:26:47] <rmatte> go for it

[28-May-2010 15:27:00] <rmatte> parts of it link back to files on our nfs share though

[28-May-2010 15:27:08] <rmatte> but not much of it

[28-May-2010 15:27:57] <rmatte> here you go: http://rfid-fantasy.blogspot.com/2010/05/vmware-tools-install-on-ubuntu-1004.html

[28-May-2010 15:28:03] <rmatte> works like a charm

[28-May-2010 15:28:10] <rmatte> it's such a small fix too

[28-May-2010 15:29:13] <theacolyte> heh, neat

[28-May-2010 15:29:18] <theacolyte> there was one other one that failed that I didn't care about

[28-May-2010 15:29:22] <theacolyte> (vmware module)

[28-May-2010 15:29:24] <theacolyte> hgfs

[28-May-2010 15:29:24] <rmatte> really?

[28-May-2010 15:29:32] <rmatte> that's the only one that I had fail out of the lot

[28-May-2010 15:29:41] <theacolyte> I've only done one so far

[28-May-2010 15:30:08] <rmatte> ah

[28-May-2010 15:30:18] <rmatte> you don't use ldap for zenoss authentication I assume?

[28-May-2010 15:30:20] <rmatte> or do you?

[28-May-2010 15:30:31] <theacolyte> I do not

[28-May-2010 15:30:35] <theacolyte> I probably should set it up though

[28-May-2010 15:30:38] <rmatte> if not you can just remove that section of code from the script

[28-May-2010 15:30:40] <rmatte> the rest is relevant

[28-May-2010 15:30:40] <theacolyte> I'm just one dude

[28-May-2010 15:31:12] <rmatte> I'm basically just copying the python-ldap files for python2.4 back in to the system so that the symlinks to the python2.4 in the zenoss directory work

[28-May-2010 15:31:24] <rmatte> since they removed python2.4 and 2.5 from 10.04

[28-May-2010 15:31:30] <rmatte> stupid move on their part, but oh well

[28-May-2010 15:31:35] <theacolyte> 2.6

[28-May-2010 15:31:36] <theacolyte> ?

[28-May-2010 15:31:50] <rmatte> yeh, they only have 2.6 by default now

[28-May-2010 15:31:54] <theacolyte> weird

[28-May-2010 15:31:55] <rmatte> no 2.4 or 2.5

[28-May-2010 15:32:06] <theacolyte> I run SVN and RoR/apache on mine

[28-May-2010 15:32:07] <rmatte> and you can't install 2.4 or 2.5 packages anymore

[28-May-2010 15:32:42] <rmatte> but yeh, once Zenoss moves to 2.6 it won't matter anyways

[28-May-2010 15:32:47] <rmatte> so it's just a temporary fix

[28-May-2010 15:34:44] <rmatte> apparmor was making the servers run uber sluggish so I decided to remove it

[28-May-2010 15:35:49] <rocket> apparmor isnt being supported much anymore anyways

[28-May-2010 15:35:58] <rmatte> yeh, I remember seeing an article about that

[28-May-2010 15:36:20] <rocket> Crispin left for windows

[28-May-2010 15:36:25] <rmatte> yeh

[28-May-2010 15:36:40] <rmatte> I guess if the paycheck is big enough people will do anything hehe

[28-May-2010 15:37:26] <rmatte> and it's not like microsoft couldn't use his expertice

[28-May-2010 15:38:01] <rmatte> $10 says he's banging his head on his desk at this very moment while reviewing windows security code

[28-May-2010 15:38:03] <rmatte> :)

[28-May-2010 15:39:13] <rmatte> finally, we got the go ahead to migrate all of our dev stuff off of the SAN

[28-May-2010 15:39:18] <rmatte> been having major IO issues lately

[28-May-2010 15:39:51] <rocket> rmatte: with zenoss?

[28-May-2010 15:40:00] <rmatte> in general, but yes, with zenoss in part

[28-May-2010 15:40:09] <rmatte> zenoss is responsible for most of it I'm sure

[28-May-2010 15:40:11] <rocket> 2.5.2 install?

[28-May-2010 15:40:16] <rmatte> all 2.5.2 installs yeh

[28-May-2010 15:40:24] <rmatte> I've noticed they are sluggish after the upgrade

[28-May-2010 15:40:43] <rmatte> it wasn't so sluggish with 2.4.5

[28-May-2010 15:41:00] <rmatte> though the memory usage has improved a LOT in 2.5.2

[28-May-2010 15:41:10] <rmatte> not eating up swap after 10 days anymore

[28-May-2010 15:41:18] <rocket> zenpatch 18589;zenpatch 18908;zenpatch 18890;zenpatch 18941

[28-May-2010 15:41:28] <rmatte> ooooh

[28-May-2010 15:41:31] <rmatte> what do those do?

[28-May-2010 15:41:43] <rocket> fix alot of zodb churn

[28-May-2010 15:41:52] <rmatte> sweet

[28-May-2010 15:42:03] <rmatte> they've been tested and won't break anything I assume?

[28-May-2010 15:42:45] <rocket> http://dev.zenoss.com/trac/ticket/6521;http://dev.zenoss.com/trac/ticket/6522;http://dev.zenoss.com/trac/ticket/6529; http://dev.zenoss.com/trac/ticket/6413

[28-May-2010 15:43:20] <rocket> considering support applies them at customers yes its been tested and code reviewed.  NOT through full regression testing of our QA dept yet though

[28-May-2010 15:43:32] <rocket> they are part of trunk/3.0 however

[28-May-2010 15:43:33] <theacolyte> apparmor... what a load of fun that is

[28-May-2010 15:43:38] <theacolyte> better than selinux I guess

[28-May-2010 15:43:45] <cgibbons> some of those changes explicitly break a few things, but I doubt you'll be exposed to what they do break

[28-May-2010 15:43:56] <rocket> and they havent made qa complain yet .. :p

[28-May-2010 15:44:10] <cgibbons> yeah, they have, just within context

[28-May-2010 15:44:12] <rmatte> rocket: thanks, I appreciate it

[28-May-2010 15:44:30] <rmatte> what do they break?

[28-May-2010 15:44:49] <rmatte> just a quick overview would be nice so that I'm informed

[28-May-2010 15:44:59] <cgibbons> some stuff inside the Cisco UCS ZenPack specifically

[28-May-2010 15:45:09] <rmatte> k, I don't even use that

[28-May-2010 15:45:17] <rmatte> isn't that enterprise?

[28-May-2010 15:45:20] <cgibbons> yes, and in beta

[28-May-2010 15:45:24] <rmatte> ah k

[28-May-2010 15:45:26] <theacolyte> I would really like access to a Cisco UCS system, so that I could be able to use that ZenPack

[28-May-2010 15:45:28] <rmatte> is that all it breaks?

[28-May-2010 15:45:31] <theacolyte> plox make it happen

[28-May-2010 15:45:32] <cgibbons> so far

[28-May-2010 15:45:35] <rmatte> k

[28-May-2010 15:45:42] * Simon4_ has a UCS system, but nowhere to rack it at present :/

[28-May-2010 15:45:49] <theacolyte> I will make space in mine

[28-May-2010 15:45:56] <Simon4_> heh

[28-May-2010 15:49:50] <cgibbons> pwd

[28-May-2010 15:50:11] <rmatte> /home/cgibbons

[28-May-2010 15:50:13] <Simon4_> you would be in /zenoss, it seems

[28-May-2010 15:50:56] <rmatte> hehe

[28-May-2010 15:51:25] <theacolyte> hehe

[28-May-2010 16:09:14] <rmatte> I wish this upgrade could be run without user interraction

[28-May-2010 16:09:30] <rmatte> it's annoying to have to answer prompts as it goes through

[28-May-2010 16:10:26] <zykes-> rmatte: ubuntu / debian ?

[28-May-2010 16:10:30] <rmatte> ubuntu

[28-May-2010 16:10:34] <zykes-> :p

[28-May-2010 16:10:52] <zykes-> looking forward until spacewalk supports ubuntu as well as redhat based sys

[28-May-2010 16:11:03] <rmatte> yeh, spacewalk looks really nice

[28-May-2010 16:11:06] <zykes-> any smart way to monitor web pages that doesn't react to ping ?

[28-May-2010 16:11:21] <rmatte> zykes-: yup, the synthetic transactions ZenPack

[28-May-2010 16:11:33] <rmatte> or the HTTP Monitor ZenPack

[28-May-2010 16:11:43] <rocket> rmatte can you use expect?

[28-May-2010 16:11:45] <rmatte> I use a combination of both for some sites

[28-May-2010 16:11:50] <zykes-> well yeh, but it still gives critical events on ping

[28-May-2010 16:11:58] <rmatte> rocket: It'd be too much of a pain to do lol

[28-May-2010 16:12:02] <rocket> k

[28-May-2010 16:12:08] <rmatte> rocket: too many possible combinations of prompts

[28-May-2010 16:12:28] <zykes-> redhat's way of doing it is much easier imo

[28-May-2010 16:12:33] <zykes-> no promts! :D

[28-May-2010 16:12:47] <zykes-> so wanting to get rhel6 final

[28-May-2010 16:14:43] <rmatte> yeh, I'm not a fan of prompts unless you specifically ask for them

[28-May-2010 16:14:52] <rmatte> but oh well

[28-May-2010 16:15:10] <rmatte> it'll be nice to have my Zenoss boxes on a current version of Linux

[28-May-2010 16:15:15] <rmatte> rather than an ancient version

[28-May-2010 16:17:34] <zykes-> ancient meaning ?

[28-May-2010 16:22:59] <rmatte> meaning over a year since last update

[28-May-2010 16:23:24] <rmatte> after I do these upgrades we'll be doing upgrades regularly

[28-May-2010 16:23:41] <rmatte> a year ago we were focussing on learning the tools and getting everything setup so we weren't really worried about it

[28-May-2010 16:23:56] <rmatte> but now that everything is setup we're getting everything up to date

[28-May-2010 16:24:48] <rmatte> upgrading our dns/nfs/ldap server is going to be fun since everything relies on that

[28-May-2010 16:25:11] <rmatte> we need to configure proper failover for it at some point with a floating IP

[28-May-2010 16:25:12] <zykes-> ldap > rhds

[28-May-2010 16:25:12] <zykes-> :p

[28-May-2010 16:25:23] <rmatte> and get the nfs share automatically rsynced each night

[28-May-2010 16:25:37] <zykes-> openfiler ? .

[28-May-2010 16:26:01] <rmatte> ?

[28-May-2010 16:26:08] <rmatte> is that some sort of nfs utility?

[28-May-2010 16:26:18] <zykes-> heh, i mistook

[28-May-2010 16:26:34] <rmatte> nah, what I meant was, we have this nfs share that has scripts which all of the zenoss boxes uses

[28-May-2010 16:26:42] <rmatte> we keep them on that so that we only have to update them in one place

[28-May-2010 16:26:56] <rmatte> so we need proper failover to nfs on our second server if it fails

[28-May-2010 16:26:59] <zykes-> public ones or ?

[28-May-2010 16:27:00] <zykes-> :p

[28-May-2010 16:27:04] <rmatte> no, private

[28-May-2010 16:27:07] <rmatte> lol

[28-May-2010 16:27:17] <zykes-> heh, check scripts or ?

[28-May-2010 16:27:44] <rmatte> well, for example, I have a zenoss daemon that I wrote which calls on a create_ticket.pl script to create tickets in our ticketing system

[28-May-2010 16:27:49] <rmatte> and the script is stored on the share

[28-May-2010 16:28:23] <zykes-> how does the daemon work ?

[28-May-2010 16:28:24] <rmatte> I also have a script which is cronned on each Zenoss box to keep the hosts file up to date with the hosts that are in zenoss

[28-May-2010 16:28:46] <rmatte> http://dmon.org/zenticket.html

[28-May-2010 16:28:57] <rmatte> the code there isn't the latest, but you'll get the just of it

[28-May-2010 16:29:52] <rmatte> I'm not a pro at python either, that's the biggest project I've done with it

[28-May-2010 16:29:53] <rmatte> it works well

[28-May-2010 16:30:33] <rmatte> I need to make an update to it soon to add some more error checking

[28-May-2010 16:30:49] <rmatte> it's quite stable but does crash under certain rare conditions, I've been fixing the issues as I find them

[28-May-2010 16:32:41] <rmatte> I need to go through and comment more of the code too

[28-May-2010 17:18:37] <cgibbons> huh, I never noticed the Windows 7 (Vista, too?) network map before.

[28-May-2010 17:35:30] <rmatte> ?

[28-May-2010 17:38:01] forsberg is now known as fOrsberg

[28-May-2010 17:54:09] <theacolyte> haha

[28-May-2010 17:54:15] <theacolyte> in the process of trying to figure out what you were talking about

[28-May-2010 17:54:25] <theacolyte> I never realized that win7 had a c redential manger

[28-May-2010 17:55:14] <SEJeff> http://www.newgrounds.com/portal/view/534416 random link-o-the-day

[28-May-2010 17:55:40] <SEJeff> You can play NES mario brothers with players from other classic games like mega man, or samus from metroid. Addictive for classic gamers

[28-May-2010 18:02:45] <rmatte> SEJeff: that is pretty awesome

[28-May-2010 18:09:41] <SEJeff> Yup. It killed my productivity for the day

[29-May-2010 00:00:30] [disconnected at Sat May 29 00:00:30 2010]

[29-May-2010 00:00:30] [connected at Sat May 29 00:00:30 2010]

[29-May-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[29-May-2010 02:24:27] fOrsberg is now known as forsberg

[29-May-2010 02:53:36] forsberg is now known as fOrsberg

[29-May-2010 07:04:55] fOrsberg is now known as forsberg

[29-May-2010 07:07:05] forsberg is now known as fOrsberg

[29-May-2010 07:14:42] fOrsberg is now known as forsberg

[29-May-2010 09:28:20] LarsN_Away is now known as LarsN

[29-May-2010 11:47:11] LarsN is now known as LarsN_Away

[29-May-2010 12:14:17] <rmatte> man, the new event console is soooooo slowwwwww when viewing really large numbers of events

[29-May-2010 12:14:24] <rmatte> and I get server connection errors all the time

[29-May-2010 12:14:34] <rmatte> then zenoss completely locks up and I have to restart it

[29-May-2010 12:15:16] <rmatte> yup, it just happened again

[29-May-2010 12:15:17] <rmatte> grrr

[29-May-2010 12:19:44] <rmatte> I used to be able to view hundreds of thousands of events at once with the old one

[29-May-2010 12:19:53] <rmatte> can't even view a few thousand with the new one

[29-May-2010 12:36:40] <theacolyte> It does really start to slow down the more tyou get

[29-May-2010 15:20:13] <Xajin> Howdy

[29-May-2010 15:30:46] <Xajin> I have setup OS Processes and IP services for Mysqld, I down mysqld but only ip services reflects it, os processes does not, any thoughts

[29-May-2010 16:40:39] <chemist> why am I seeing - (minus) values in zenoss daemons heartbeats?

[29-May-2010 17:04:25] <Xajin> Anyone get the Device does not publish HOST-RESROUCES-MIB ?

[29-May-2010 17:13:36] LarsN_Away is now known as LarsN

[29-May-2010 19:52:06] forsberg is now known as fOrsberg

[29-May-2010 22:49:31] LarsN is now known as LarsN_Away

[29-May-2010 22:49:36] LarsN_Away is now known as LarsN

[29-May-2010 23:39:03] rhettardo_ is now known as rhettardo

[30-May-2010 00:00:30] [disconnected at Sun May 30 00:00:30 2010]

[30-May-2010 00:00:31] [connected at Sun May 30 00:00:31 2010]

[30-May-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[30-May-2010 04:29:54] fOrsberg is now known as forsberg

[30-May-2010 06:28:30] forsberg is now known as fOrsberg

[30-May-2010 06:33:21] fOrsberg is now known as forsberg

[30-May-2010 12:59:48] * Simon4 has a look at trac

[30-May-2010 12:59:51] <Simon4> 8 weeks since I did any

[30-May-2010 12:59:52] <Simon4> work

[30-May-2010 12:59:52] <Simon4> on

[30-May-2010 13:00:03] <Simon4> hmm... colloquy bung?

[30-May-2010 13:00:12] <Simon4> that was a lot of carriage returns

[30-May-2010 13:00:46] <Simon4> anyway, far too long since I did any work on the BladeChassis ZenPack

[30-May-2010 14:35:25] <zykes-> ERROR: zenpack command failed. Reason: Exception: Unable to create object using the following attributes: * id: /zport/dmd/Devices/Storage/IBM * module: Products.ZenModel.DeviceClass * class: DeviceClass

[30-May-2010 14:35:30] <zykes-> anyone seen that ?

[30-May-2010 16:37:18] <zykes-> cgibbons: around ?

[30-May-2010 17:25:52] forsberg is now known as fOrsberg

[30-May-2010 19:41:13] <rmatte> zykes-: this is another case of reading the dependencies before installing the pack

[30-May-2010 19:41:28] <rmatte> you need to create the Devices/Storage/IBM device class by hand prior to installing it

[30-May-2010 19:41:56] <rmatte> ZenPack pages generally explain that sort of thing, so I'm guessing it's listed (too lazy to check)

[30-May-2010 19:43:19] <rmatte> so I just checked...

[30-May-2010 19:43:20] <rmatte> # Installation:

[30-May-2010 19:43:20] <rmatte>     * Create /Mibs/IBM subfolder

[30-May-2010 19:43:20] <rmatte>     * Create /Devices/Storage device class

[30-May-2010 19:43:33] <rmatte> create /Devices/Storage THEN install the pack

[30-May-2010 19:43:47] <rmatte> also, read zenpack pages more closely next time

[30-May-2010 19:44:18] <rmatte> everyone seems to download them, try to install them, then ask in here or on the forums if they get an error when 90% of the time the answer is on the ZenPack page itself

[30-May-2010 19:46:34] <rmatte> also creat the /Mibs/IBM subfolder as well

[31-May-2010 00:00:30] [disconnected at Mon May 31 00:00:30 2010]

[31-May-2010 00:00:30] [connected at Mon May 31 00:00:30 2010]

[31-May-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[31-May-2010 01:31:07] fOrsberg is now known as forsberg

[31-May-2010 02:20:54] <voxter> Soo, I'm running alpha 3 (i think ill have to check) and deleted a device, but its still attempting to monitor and alert a service on it. How can i dig in there and clean the rest out?

[31-May-2010 03:06:17] <forsberg> if nothing else works, i restart zenoss

[31-May-2010 03:06:19] <forsberg> sounds like an old bug

[31-May-2010 03:06:20] <forsberg> :)

[31-May-2010 04:22:28] <Simon4> grr, Template's "Test against device" feature seems to ignore the "Use SSH" option :/

[31-May-2010 04:24:48] <zykes-> hmmm, has there been any thought to make a script that reports events to irc ?

[31-May-2010 04:25:32] <Simon4> heh

[31-May-2010 04:25:36] <Simon4> I was considering it the other day

[31-May-2010 04:26:03] <Simon4> wouldn't be hard to knock up a bot in python that could be tickled up by an event command, or even just watch mysql

[31-May-2010 06:41:14] <zykes-> anyone running plone4zenoss ?

[31-May-2010 08:45:27] tehhobbit is now known as TillBakaRullare

[31-May-2010 08:47:39] TillBakaRullare is now known as tehhobbit

[31-May-2010 08:55:34] forsberg is now known as fOrsberg

[31-May-2010 10:21:25] <frozty_sa> anyone around here who successfully use the memcached zenpack? I've loaded it as per the instructions on the site, and the RRDs "seem" to be populating with data, but going to the perf screen I don't get any graphs being drawn for it (added one graph that has all the datapoints)

[31-May-2010 10:24:26] <frozty_sa> actually, no, I don't think the RRDs are really populating. memCached_memCached is the one for cached memory, right?

[31-May-2010 10:40:44] <chachan> Hi everyone, I'm having a problem when a try to add a device through ssh, I removed all snmp zCollertorPlugins, added zenoss.cmd.uname and set zCommandUsername and zCommandPassword, am I forgetting something?. thanks

[31-May-2010 10:41:11] fOrsberg is now known as forsberg

[31-May-2010 10:41:40] <chachan> the problem is that in events tab I get "snmp agent down"

[31-May-2010 10:42:24] <chachan> I want use ssh instead of snmp

[31-May-2010 10:47:17] <chachan> is this the channel for this kind of questions?

[31-May-2010 11:36:33] <rmatte> chachan: what type of device?

[31-May-2010 11:36:44] <chachan> rmatte, a Linux server

[31-May-2010 11:37:33] <rmatte> then put it in the /Server/Linux/SSH class instead of /Server/Linux

[31-May-2010 11:37:53] <rmatte> read the Zenoss admin guide, it explains SSH monitoring

[31-May-2010 11:37:55] <chachan> I remove a nmap Plugin and now I dont get "snmp agetn down", but the device has not all information

[31-May-2010 11:38:20] <rmatte> read the guide and if you still have questions after that then ask

[31-May-2010 11:38:21] <chachan> rmatte, I already read it, and follow its intructions

[31-May-2010 11:38:33] <rmatte> what type of info are you missing?

[31-May-2010 11:38:48] <chachan> operative system info

[31-May-2010 11:38:52] <chachan> file system

[31-May-2010 11:38:56] <chachan> I dont get anything

[31-May-2010 11:39:00] <rmatte> also, the nmap plugin has nothing to do with SNMP, so that's a coincidence that you stopped getting the messages when you removed it

[31-May-2010 11:39:18] <rmatte> make sure that the ssh filesystem modeler plugin is applied to the class

[31-May-2010 11:39:26] <rmatte> it probably isn't

[31-May-2010 11:39:28] <chachan> It says me that the device is UP, but I dont see that it polls the info

[31-May-2010 11:39:37] <chachan> mmm

[31-May-2010 11:39:45] <rmatte> has it modeled any info at all?

[31-May-2010 11:39:50] <rmatte> such as interfaces?

[31-May-2010 11:39:51] <chachan> I will check it

[31-May-2010 11:40:14] <rmatte> if not, then you have a more general issue, did you define the username and password under zProperties as described in the admin guide?

[31-May-2010 11:40:21] <chachan> do I need add plugin for operative system info?

[31-May-2010 11:40:41] <rmatte> I'm not sure about that part, honestly

[31-May-2010 11:40:42] <chachan> Yes, I set the username and the password

[31-May-2010 11:40:48] <rmatte> I don't do ssh monitoring, I just know of it

[31-May-2010 11:41:00] <rmatte> I know that you should at least get CPU, Memory, interfaces, and filesystems

[31-May-2010 11:41:08] <chachan> well, I'll check it

[31-May-2010 11:41:10] <rmatte> k

[31-May-2010 11:41:22] <chachan> I dont added these plugins :s

[31-May-2010 11:41:37] <chachan> It could be my mistake

[31-May-2010 11:41:48] <chachan> thanks for the help

[31-May-2010 11:41:49] <chachan> :)

[31-May-2010 11:41:50] <rmatte> np

[31-May-2010 11:47:07] <chachan> interfaces are working right now :)

[31-May-2010 11:47:32] <rmatte> cool

[31-May-2010 11:47:34] <chachan> the plugin was missing, well, I'm a newbie in zenoss

[31-May-2010 11:52:35] <rmatte> all good

[31-May-2010 11:52:37] <rmatte> lots to learn

[31-May-2010 11:54:45] <rmatte> the new event console makes me want to scream sometimes

[31-May-2010 11:55:17] <chachan> hehehe

[31-May-2010 11:56:06] <chachan> I think that I'll need to add this channel to my default list of irc channels

[31-May-2010 11:56:49] <rmatte> yeh, it's a handy place

[31-May-2010 11:57:12] <chachan> well, now I'm testing zenoss to propouse it in my company

[31-May-2010 11:58:22] <chachan> one boss wants to develop and application like zenoss, why do it if zenoss exists? hehehe

[31-May-2010 11:58:56] <rmatte> yeh

[31-May-2010 11:58:58] <chachan> other boss dont really care about, just want to know that everything it's ok

[31-May-2010 11:59:18] <rmatte> Zenoss is great and very extendable since it's written in python

[31-May-2010 11:59:41] <rmatte> it has it's minor annoyances, but so does every product

[31-May-2010 12:03:03] <chachan> yep, it's understandable

[31-May-2010 12:07:46] <willwh> chachan: from a random user, to another random tester - Zenoss rocks

[31-May-2010 12:08:04] <chachan> willwh, :)

[31-May-2010 12:08:07] <willwh> I can't even begin to count the number of hours it's saved me since it's been running :)

[31-May-2010 12:08:37] <chachan> cool

[31-May-2010 12:09:18] <chachan> That's what I want to say very soon :)

[31-May-2010 12:13:30] <willwh> it takes a ilttle bit of time to get stuff tuned down the way you need it

[31-May-2010 12:13:35] <willwh> depending on what you're monitoring

[31-May-2010 12:13:48] <willwh> but once you get the hang of what to look for, it hardly takes any time at all :)

[31-May-2010 12:14:47] <chachan> :)

[31-May-2010 12:14:57] <chachan> that's what I looking for

[31-May-2010 12:19:30] <rmatte> yeh, I used to suck at tuning stuff, but now that I've learned all the little tricks it's really easy

[31-May-2010 12:19:49] <rmatte> you'll want to learn a bit of python so that you can do transforms if you need top

[31-May-2010 12:19:50] <rmatte> top*

[31-May-2010 12:19:53] <rmatte> errr

[31-May-2010 12:19:54] <rmatte> to****

[31-May-2010 12:19:56] <rmatte> lol

[31-May-2010 12:20:40] <rmatte> well, time to figure out how to get my syslog clients not to send a syslog message for every damn UDP connection that is made since it's owning the events database on the zenoss server that monitors our other zenoss servers

[31-May-2010 12:21:06] <rmatte> I can't even load up the event history without the UI completely freezing up

[31-May-2010 12:24:29] <chachan> rmatte, I know python, I'm not a guru, but I understand the code and can write a little :p

[31-May-2010 12:24:55] <rmatte> cool

[31-May-2010 12:25:00] <rmatte> that's all you really need

[31-May-2010 12:25:08] <rmatte> I'm not a guru but I'm fairly proficient

[31-May-2010 12:25:59] <chachan> do anyone use zenoss.cmd.linux.process?

[31-May-2010 12:26:51] <chachan> I'm getting an error when I model the device

[31-May-2010 12:28:18] <chachan> I'm going to lunch, I'll back soon, cya guys and thanks for the comments

[31-May-2010 12:32:29] * Simon4 tries to find a way to sneak more than one command into a zencommand via ssh

[31-May-2010 12:32:31] <otakup0pe> ugh is the sequence editing page ever going to get an update

[31-May-2010 12:32:49] <otakup0pe> i only have 15 entries and it's already quite onerous

[31-May-2010 12:33:39] <otakup0pe> no matter what i do it seems to end up just taking a random order :(

[31-May-2010 12:45:22] <otakup0pe> i hope i'm doing something wrong and this isn't a flaw of zenoss :v

[31-May-2010 12:51:57] <Simon4> otakup0pe: has always worked for me

[31-May-2010 12:52:14] <otakup0pe> with a sizable amount of event mappings i'm guessing

[31-May-2010 12:52:18] <otakup0pe> i have 16 which doesn't seem like a lot :(

[31-May-2010 12:53:01] <rmatte> woohoo, time to delete 100 million events from my event database on one of my servers

[31-May-2010 12:53:54] <otakup0pe> it's completely random too heh

[31-May-2010 12:55:03] <otakup0pe> i can literally hit save without modifying any of the numbers and it comes out different each time O_o

[31-May-2010 12:55:57] <Simon4> hmm, I haven't tried it in event mappings, but it works fine for resequencing things like graphpoints

[31-May-2010 12:57:25] <otakup0pe> it's only giving me problems on one of the eventkey's

[31-May-2010 12:58:18] <otakup0pe> hmm does sequence apply to event classes or event id's

[31-May-2010 12:58:39] <rmatte> otakup0pe: just change the value of the one you want to move if you're just moving 1

[31-May-2010 12:58:44] <otakup0pe> doesn't work

[31-May-2010 12:59:04] <rmatte> hmmm, you said it works in graphs but not for transforms?

[31-May-2010 12:59:11] <otakup0pe> i haven't tried it with graphs

[31-May-2010 12:59:13] <rmatte> I haven't tried resequencing transforms for quite a while

[31-May-2010 12:59:16] <otakup0pe> and in some of my sets of event mappings it works

[31-May-2010 12:59:25] <rmatte> try it with graphs and see if it works

[31-May-2010 12:59:31] <rmatte> it might be specific to transforms

[31-May-2010 12:59:38] <otakup0pe> it's working for some transforms

[31-May-2010 12:59:58] <rmatte> maybe it's intelligently ordering them to some degree as well?

[31-May-2010 13:00:10] <otakup0pe> it's certainly trying.

[31-May-2010 13:00:15] <rmatte> lol

[31-May-2010 13:00:17] <otakup0pe> but based on what ?

[31-May-2010 13:00:40] <otakup0pe> it's essentially rendered zenoss's event mapping feature useless to me. which is a shame because i was enjoying it.

[31-May-2010 13:00:53] <otakup0pe> maybe it's time to finally go for the commercial support ;)

[31-May-2010 13:01:15] <rmatte> lol

[31-May-2010 13:01:21] <rmatte> you're a commercial client?

[31-May-2010 13:01:44] <rmatte> well, enterprise client, I should say

[31-May-2010 13:02:46] <otakup0pe> no still using zenoss core. this seems like something that must actually work and i'm just doing something wrong.

[31-May-2010 13:03:28] <otakup0pe> i seem to be able to resequence graphs in templates

[31-May-2010 13:11:37] <Simon4> hmm, collecting via ssh on a single Blade Chassis does a fairly good job of attempting to DOS the OA

[31-May-2010 13:16:54] <rmatte> lol

[31-May-2010 13:17:12] <rmatte> otakup0pe: it's very possible that it's a bug

[31-May-2010 13:17:16] <rmatte> new ones pop up all the time

[31-May-2010 13:17:22] <otakup0pe> :3

[31-May-2010 13:17:40] <otakup0pe> i'll see if i can distill it into a coherent report

[31-May-2010 13:17:42] <rmatte> one release you can set hardware info for devices, the next you can't, the next you sort of can

[31-May-2010 13:17:43] <rmatte> lol

[31-May-2010 13:17:52] <otakup0pe> i really don't have time to come up with a repro though

[31-May-2010 13:23:21] <rmatte> eugh, does anyone know if there's a place to crank up the timeout value on the new event console?

[31-May-2010 13:23:36] <rmatte> when I click to load the event history it cranks for about 1 minute then I get a server connection error

[31-May-2010 13:23:40] <rmatte> it's driving me absolutely insane

[31-May-2010 13:35:29] <Simon4> not sure, I think it's loaded via ajax, so you may need to dig into javascript (ugh)

[31-May-2010 13:45:08] <rmatte> lame lol

[31-May-2010 13:45:18] <rmatte> I'll see if clearing the entire events database improves the situation

[31-May-2010 13:45:25] <rmatte> I'm not having the same problem on any of my other servers

[31-May-2010 13:46:58] * Simon4 plays with ZenPack skins... possibly my least favourite bit of ZenPack dev

[31-May-2010 13:53:05] <rmatte> I've never tried to do skins yet

[31-May-2010 13:54:07] <Simon4> I just end up finding something that looks like what I want, and copy/paste it, not very scientific

[31-May-2010 14:08:04] <rmatte> hehe

[31-May-2010 14:41:17] <rmatte> hmmm, I think something is seriously wrong with mysql on this Zenoss server

[31-May-2010 14:45:49] <rmatte> yup it is, great

[31-May-2010 14:46:18] <rmatte> on other servers I can do "select * from detail;" or "select * from historyl;", on this one it just freezes up, which explains why I can't load the event history on it

[31-May-2010 14:50:04] <rmatte> now the question is, how the heck am I going to fix that...

[31-May-2010 14:51:53] <Simon4> drop table history? :)

[31-May-2010 14:53:37] <rmatte> yeh but then I lose the whole table and need to recreate it

[31-May-2010 14:53:43] <rmatte> but that is an option on this one particular server

[31-May-2010 14:53:47] <rmatte> it wouldn't be on any other

[31-May-2010 14:54:03] <rmatte> could I just do something like delete * from history; first?

[31-May-2010 14:54:06] <rmatte> just to try?

[31-May-2010 14:54:21] <Simon4> yup

[31-May-2010 14:54:23] <rmatte> hmmm, the log table is fine

[31-May-2010 14:54:42] <chachan> truncate table history

[31-May-2010 14:54:47] <rmatte> same with heartbeat and alert

[31-May-2010 14:54:52] <chachan> I think that's the command

[31-May-2010 14:54:55] <Simon4> when you say "freezes up" do you just mean appears to never come back? or does the mysql process actually die?

[31-May-2010 14:55:06] <rmatte> Simon4: appears to never come back

[31-May-2010 14:55:28] <rmatte> Simon4: I have TONS of events in the database, but I told Zenoss to delete events older than 1 day so it should have gone through and deleted them all

[31-May-2010 14:55:42] <Simon4> hmm, is there still a delete process happening?

[31-May-2010 14:55:49] <Simon4> show processlist on the mysql commandline

[31-May-2010 14:56:10] <rmatte> nope

[31-May-2010 14:56:27] <Simon4> it would be worth checking that whilst doing a select count(*) from history; also, see what your select process is up to

[31-May-2010 14:56:43] <rmatte> k

[31-May-2010 14:57:01] <rmatte> running that right now

[31-May-2010 14:57:14] <rmatte> I have problems querrying the detail table as well

[31-May-2010 14:57:17] <rmatte> all the others work fine

[31-May-2010 14:57:46] <Simon4> a huge flood of recent events could cause this I guess, fill both tables to the point where mysql struggles

[31-May-2010 14:58:44] <rmatte> | 11 | zenoss | localhost | events | Query   |   97 | Sending data | select count(*) from history |

[31-May-2010 14:59:00] <Simon4> k, it thinks it's running okay

[31-May-2010 14:59:11] <rmatte> it's not even recent, I was receiving tons of UDP connection syslogs from servers, which I just fixed today

[31-May-2010 14:59:22] <rmatte> but now I need to clean up the mess and get it back to normal

[31-May-2010 14:59:34] <rmatte> there we go

[31-May-2010 14:59:35] <rmatte> mysql> select count(*) from history;

[31-May-2010 14:59:35] <rmatte> +----------+

[31-May-2010 14:59:35] <rmatte> | count(*) |

[31-May-2010 14:59:35] <rmatte> +----------+

[31-May-2010 14:59:35] <rmatte> |  3173847 |

[31-May-2010 14:59:36] <rmatte> +----------+

[31-May-2010 14:59:36] <rmatte> 1 row in set (2 min 31.32 sec)

[31-May-2010 14:59:48] <rmatte> quite a few

[31-May-2010 14:59:56] <Simon4> yeah, not stupidlarge though

[31-May-2010 15:00:07] <Simon4> we got up to 12 million once when someone forgot to set an aging interval

[31-May-2010 15:00:24] <rmatte> lol

[31-May-2010 15:00:34] <rmatte> guess I should kill all the events

[31-May-2010 15:00:35] <Simon4> try select * from history LIMIT 10;

[31-May-2010 15:00:38] <rmatte> is truncate the command do that?

[31-May-2010 15:00:44] <rmatte> LIMIT 10?

[31-May-2010 15:00:45] <Simon4> see if that retursn

[31-May-2010 15:00:47] <Simon4> yeah

[31-May-2010 15:00:55] <Simon4> says give me the first 10 entries from that table;

[31-May-2010 15:01:01] <rmatte> yes, it does

[31-May-2010 15:01:20] <rmatte> same for detail

[31-May-2010 15:01:43] <Simon4> what does select count(*) from detail; give you?

[31-May-2010 15:02:08] <Simon4> (may take a minute or two to finish)

[31-May-2010 15:02:17] <rmatte> it's nice to have a refresher on mysql, haven't really worked with it much since college lol

[31-May-2010 15:02:27] <rmatte> yeh, I'll give it some time

[31-May-2010 15:04:36] <Simon4> truncate looks like the fast way to empty the table, basically in the background mysql just drops the existing one and creates a new one with the same schema

[31-May-2010 15:05:09] <Simon4> http://dev.mysql.com/doc/refman/5.1/en/truncate-table.html

[31-May-2010 15:05:09] <rmatte> ah ok

[31-May-2010 15:05:20] <rmatte> the select on detail is still rolling

[31-May-2010 15:05:33] <forsberg> truncate cleans it up yes :P

[31-May-2010 15:08:11] <rmatte> mysql> select count(*) from detail;

[31-May-2010 15:08:11] <rmatte> +----------+

[31-May-2010 15:08:11] <rmatte> | count(*) |

[31-May-2010 15:08:11] <rmatte> +----------+

[31-May-2010 15:08:11] <rmatte> |  5139659 |

[31-May-2010 15:08:12] <rmatte> +----------+

[31-May-2010 15:08:12] <rmatte> 1 row in set (6 min 2.82 sec)

[31-May-2010 15:08:16] <Simon4> heh

[31-May-2010 15:08:25] <rmatte> so I have more details than events

[31-May-2010 15:08:29] <rmatte> that's nice

[31-May-2010 15:09:03] <rmatte> ok, so I'm wanting to truncate the details table and the history table from the looks of it

[31-May-2010 15:09:37] <Simon4> yeah

[31-May-2010 15:09:54] <rmatte> what's the exact command again?

[31-May-2010 15:10:13] <rmatte> truncate table history;

[31-May-2010 15:10:21] <rmatte> truncate table detail;

[31-May-2010 15:10:21] <rmatte> ?

[31-May-2010 15:10:30] <Simon4> yup

[31-May-2010 15:10:38] <Simon4> mysql> truncate table detail;

[31-May-2010 15:10:38] <Simon4> Query OK, 9144405 rows affected (4.03 sec)

[31-May-2010 15:10:43] <Simon4> from a test box I had kicking about

[31-May-2010 15:10:45] <Simon4> so it's quick

[31-May-2010 15:10:57] <rmatte> going to stop zenoss while I do it anyways

[31-May-2010 15:11:23] <rmatte> very quick

[31-May-2010 15:11:44] <rmatte> looks good

[31-May-2010 15:12:06] <rmatte> now to wait for a few events and see if I can load the history

[31-May-2010 15:15:28] <rmatte> nice, it works again

[31-May-2010 15:15:29] <rmatte> woot

[31-May-2010 15:15:42] <rmatte> now to do more tuning since there's still quite a bit of garbage coming in

[31-May-2010 15:17:32] <chachan> You could create a job for this :)

[31-May-2010 15:17:43] <chachan> but, I have a question

[31-May-2010 15:17:58] <chachan> are history and detail in relationship?

[31-May-2010 15:19:48] <rmatte> they should be I'd assume

[31-May-2010 15:22:11] <chachan> so, it's needly to truncate both to clean a little the database, I guess, cool, nice trick

[31-May-2010 15:26:56] <rmatte> yup

[31-May-2010 15:26:59] <rmatte> it worked great for me

[31-May-2010 15:27:06] <rmatte> Simon4: thanks for the help

[31-May-2010 15:28:29] <Simon4> no probs

[31-May-2010 15:32:06] <rmatte> with the amount of events that I had coming in before it's entirely possible that those 3 million events were just from today lol

[31-May-2010 15:32:48] <rmatte> I've been meaning to tune this stuff but didn't get around to it until today

[31-May-2010 15:40:12] <rmatte> wow, I've actually gone 3 minutes without seeing a new event, that's new

[31-May-2010 15:40:19] * Simon4 now knows how to write custom zencommand parsers

[31-May-2010 15:40:24] <Simon4> has been a productive weekend

[31-May-2010 15:40:28] <rmatte> nice

[31-May-2010 15:41:13] <rmatte> I think the detail table is only used for traps and event log details

[31-May-2010 15:41:38] <rmatte> (ones which show details on what used to be a details tab and is now mashed in with all the other info)

[31-May-2010 15:41:52] <rmatte> I hope they fix that, I want to see a second link just for event details

[31-May-2010 15:42:01] <rmatte> rather than having to comb through all the other garbage

[31-May-2010 15:42:28] <rmatte> it was great when it was on it's own tab

[31-May-2010 15:42:44] <Simon4> yeah

[31-May-2010 15:42:56] <Simon4> http://lol.nzdance.net/zenpack/bladeDetail.png

[31-May-2010 15:43:24] <rmatte> nice

[31-May-2010 15:43:42] <rmatte> quite slick

[31-May-2010 15:44:04] <Simon4> couple of small issues to iron out, but then I'll document and actually release, I'll keep putting it off for features otherwise

[31-May-2010 15:49:43] <rmatte> cool

[31-May-2010 15:49:54] <rmatte> HP Blade Chassis pack right?

[31-May-2010 15:50:03] <Simon4> yeah

[31-May-2010 15:52:16] <chachan> that tab looks like if it has CCS problems, am I wrong?

[31-May-2010 15:52:47] <chachan> I use Google Chrome over linux and I see some boxes out of place

[31-May-2010 15:53:18] <Simon4> chachan: yeah, I get that on all zenoss tabs though

[31-May-2010 15:53:39] <rmatte> chachan: the whole interface will be redone in the next version anyways

[31-May-2010 15:53:39] <Simon4> using Chrome in OS X

[31-May-2010 15:54:42] <chachan> rmatte, cool, it's a great notice, but this one is cool too

[31-May-2010 15:57:04] <chachan> mmm, I have another question, do you know about munin?, I wish to make a command in a remote server and the results will be set on a graphs, is it possible?

[31-May-2010 15:57:13] * Simon4 narrowly avoids checking in some pyc files

[31-May-2010 15:57:18] <Simon4> that would ahve been embarassing

[31-May-2010 15:57:50] <rmatte> chachan: yeh, it's called a command based datasource

[31-May-2010 15:58:20] <rmatte> which you can do with a script on the local server or with a command via SSH

[31-May-2010 15:58:30] <rmatte> It's explained in the admin guide

[31-May-2010 15:58:40] <chachan> rmatte, :)

[31-May-2010 15:59:25] <chachan> I'm checking the guide

[31-May-2010 15:59:51] <chachan> Escalation of alterting rocks

[31-May-2010 16:01:09] <rmatte> yeh, the alerting system could be a bit better, but it's pretty good as is

[31-May-2010 16:03:44] <zykes-> too bad there isn't a module like that for IBM

[31-May-2010 16:03:46] <zykes-> as well :p

[31-May-2010 16:03:56] <zykes-> just HP / Dell .

[31-May-2010 16:04:05] <Simon4> zykes-: for Blade Chassis?

[31-May-2010 16:04:53] <zykes-> yeh

[31-May-2010 16:04:58] <zykes-> and x series

[31-May-2010 16:05:06] <Simon4> if I had access to an IBM bladecenter

[31-May-2010 16:05:27] <zykes-> heh, what data you need ?

[31-May-2010 16:07:44] * Simon4 finds the mib and has a read

[31-May-2010 16:15:56] <Simon4> the output of a snmpwalk 1.3.6.1.4.1.2.3.51.2.2 would probably be enough to knock something up fairly easily

[31-May-2010 16:18:23] <Simon4> Dear HP, take a leaf out of IBM's book and do SNMP properly, love, SImon

[31-May-2010 16:21:31] <rmatte> lol

[31-May-2010 16:21:40] <rmatte> HP do SNMP horribly

[31-May-2010 16:21:48] <rmatte> specially with interface names for blade switches

[31-May-2010 16:21:53] <rmatte> especially*

[31-May-2010 16:22:07] <rmatte> they just name then copper ethernet (10/100) or whatever

[31-May-2010 16:22:16] <rmatte> and that's all that's available via snmp

[31-May-2010 16:22:21] <Simon4> gack

[31-May-2010 16:22:24] <rmatte> no indication as to port numbers or anything in the names

[31-May-2010 16:22:32] <rmatte> ifDescr is that, and ifName is blank

[31-May-2010 16:22:35] <rmatte> super super stupid

[31-May-2010 16:23:45] <rmatte> one of the guys here complained that there weren't proper interface names for the blade switches in Zenoss, so I asked him what he wanted me to do beyond mailing a letter to HP to ask them to stop being idiots

[31-May-2010 16:24:01] <chachan> I have a noob question, is snmp secure for use on public and production servers?

[31-May-2010 16:24:01] <rmatte> the SNMP on their regular user port switches isn't bad at all

[31-May-2010 16:24:06] <rmatte> it's all their higher end stuff that tends to suck

[31-May-2010 16:24:07] <forsberg> have linkss switch with same problem -         Ethernet Interface

[31-May-2010 16:24:18] <forsberg> can give them descr tho

[31-May-2010 16:24:21] <rmatte> yeh, that's no annoying

[31-May-2010 16:24:25] <rmatte> so*

[31-May-2010 16:24:32] <forsberg> linksys switch

[31-May-2010 16:24:58] <rmatte> yeh, but giving a description doesn't really help when Zenoss tells you that utilization is exceeded on Ethernet Interface

[31-May-2010 16:25:04] <Simon4> chachan: you need to run snmpv3 to get any kind of decent authentication and encryption

[31-May-2010 16:25:04] <rmatte> jeez, thanks

[31-May-2010 16:25:06] <rmatte> lol

[31-May-2010 16:25:16] <forsberg> je :)

[31-May-2010 16:25:28] <rmatte> I actually wish Zenoss would allow you to manually assign names to interfaces

[31-May-2010 16:25:38] <rmatte> it would be a nice workaround for dumb devices

[31-May-2010 16:25:42] <bigegor> HP is good, but IBM is nightmare ;)

[31-May-2010 16:25:56] <rmatte> neither are good

[31-May-2010 16:26:04] <rmatte> they both need to take a page from Cisco in my opinion

[31-May-2010 16:26:10] <rmatte> Cisco SNMP is actually done properly

[31-May-2010 16:26:13] <chachan> Simon4, mm, in this case, which one do you recommend me, ssh or snmp?

[31-May-2010 16:26:37] <rmatte> chachan: for what type of device?

[31-May-2010 16:27:58] <bigegor> for real man only: http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.IBMMon

[31-May-2010 16:27:58] <chachan> linux servers

[31-May-2010 16:28:37] <chachan> they are running centos

[31-May-2010 16:29:02] <forsberg> snmp works fine :P