[01-Mar-2010 07:23:13] <sektorNBA> hi

[01-Mar-2010 07:24:18] <sektorNBA> Daemon: zenhub starting...

[01-Mar-2010 07:24:18] <sektorNBA> Traceback (most recent call last):

[01-Mar-2010 07:24:18] <sektorNBA> File "/home/zenoss/Products/ZenHub/zenhub.py", line 22, in ?

[01-Mar-2010 07:24:18] <sektorNBA> from XmlRpcService import XmlRpcService

[01-Mar-2010 07:27:35] <sektorNBA> can anyone help me?

[01-Mar-2010 08:31:42] <malbon> hola folks,

[01-Mar-2010 08:33:17] <malbon> sektorNBA: how are you starting the zenhub daemon?

[01-Mar-2010 08:55:51] <sektorNBA> malbon i upgraded it from 2.4 but i had some zen packs installed. i guess i screw it up

[01-Mar-2010 08:57:18] <sektorNBA> now i am doing a fresh source install at freebsd

[01-Mar-2010 08:58:09] <malbon> I don't think it is something you've done. It looks like a path problem.

[01-Mar-2010 09:04:56] <jb> any ent 2.5.1 people hre?

[01-Mar-2010 09:04:57] <jb> here

[01-Mar-2010 09:06:16] <sektorNBA> i am installing it jb

[01-Mar-2010 09:06:32] <sektorNBA> but it got some problems with csh shell from freebsd.

[01-Mar-2010 09:07:19] <venturaville> we have it

[01-Mar-2010 09:07:44] <jb> are you using ACLs for your users?

[01-Mar-2010 09:09:59] <venturaville> nope

[01-Mar-2010 09:10:08] <jb> GRR

[01-Mar-2010 09:10:13] <jb> the user system is completely busted in 2.5.1 it looks like

[01-Mar-2010 09:11:57] <jb> oh.

[01-Mar-2010 09:16:04] <rmatte> jb: that might be something that got fixed in 2.5.2 if it is screwy

[01-Mar-2010 09:17:25] <rmatte> jb: what's broken about it?

[01-Mar-2010 09:18:27] <malbon> any news on the 2.5.2 release date?

[01-Mar-2010 09:20:55] <rmatte> they generally release on Wednesdays or Thursdays, and it's all packaged up and ready to be released, they are just waiting on Marketing...

[01-Mar-2010 09:21:03] <rmatte> so there's a good chance that it could get released later this week

[01-Mar-2010 09:21:14] <rmatte> if not later this week then next week would be likely

[01-Mar-2010 09:28:06] <malbon> ok, cool, if it isn't out this week I'll hassle rusty. ;)

[01-Mar-2010 09:32:38] <rmatte> well, that's for the Core release anyways...

[01-Mar-2010 09:32:49] <rmatte> the enterprise release will probably come out a few weeks after

[01-Mar-2010 09:32:51] <rmatte> per usual

[01-Mar-2010 09:33:10] <rmatte> mrayzenoss: good morning

[01-Mar-2010 09:33:17] <venturaville> guten moogen

[01-Mar-2010 09:35:10] <rmatte> :)

[01-Mar-2010 09:35:38] <mrayzenoss> morning all

[01-Mar-2010 09:36:03] <Skippy1> hello all

[01-Mar-2010 09:40:19] <Skippy1> quiet in here :) is that good cos no-one has problems ?

[01-Mar-2010 09:43:07] <Skippy1> anyone written a scrip to reset an ip of a particular device ? - i have one that is dhcp assigned and renews every 60 minutes and often gets a new IP and so falls off the monitoring until I manually reset it

[01-Mar-2010 09:43:45] <rmatte> well, I mean, setting the IP of a device is simple to do... but where are you going to pass the IP to the script from?

[01-Mar-2010 09:44:44] <Skippy1> i'm hoping just clearing it and forcing it to get it from dns is going to be good enough ?

[01-Mar-2010 09:46:57] <rmatte> you can't really do that once the device is added

[01-Mar-2010 09:47:03] <rmatte> you have to actually tell it what IP to use

[01-Mar-2010 09:47:12] <rmatte> so you're going to have to find some way of passing the IP to the script

[01-Mar-2010 09:47:32] <Skippy1> i can look it up first via dig or something and then parse it to the script

[01-Mar-2010 09:47:35] <rmatte> dns is only used during initial discovery

[01-Mar-2010 09:47:49] <rmatte> k...

[01-Mar-2010 09:47:55] <rmatte> well, what you're looking for is basically this...

[01-Mar-2010 09:49:03] <rmatte> http://fpaste.org/Fitr/

[01-Mar-2010 09:49:31] <rmatte> but you'll want to pass the IP to the script as a commandline argument...

[01-Mar-2010 09:49:32] <rmatte> so...

[01-Mar-2010 09:51:58] <rmatte> http://fpaste.org/OFHS/

[01-Mar-2010 09:54:21] <rmatte> so save that as a .py

[01-Mar-2010 09:54:34] <rmatte> replace devicename with the name of the device

[01-Mar-2010 09:54:43] <rmatte> and then execute it with the IP as an option

[01-Mar-2010 09:54:53] <rmatte> (needs to be executed on the zenoss box as the zenoss user)

[01-Mar-2010 09:55:00] <Skippy1> just tryting that - ta..

[01-Mar-2010 09:55:07] <rmatte> cool

[01-Mar-2010 09:55:59] * sektorNBA is finishing zenoss installation at freebsd.

[01-Mar-2010 09:56:17] <rmatte> sektorNBA: have fun, FreeBSD can be a nightmare with Zenoss lol

[01-Mar-2010 09:56:24] <Skippy1> ImportError: No module named Globals

[01-Mar-2010 09:56:34] <rmatte> Skippy1: then you're not executing it as the zenoss user

[01-Mar-2010 09:56:42] <rmatte> on the zenoss box

[01-Mar-2010 09:57:19] <rmatte> the zenoss user has a different python environment than the rest of the users on the system

[01-Mar-2010 09:58:29] <Skippy1> doh.. :-)

[01-Mar-2010 09:59:41] <Skippy1> its not complaining but its not working either

[01-Mar-2010 09:59:58] <rmatte> hmmm, add a commit() at the end of the script

[01-Mar-2010 10:00:17] <rmatte> http://fpaste.org/5xB3/

[01-Mar-2010 10:00:27] <rmatte> and make sure that you actually changed 'devicename' in the script

[01-Mar-2010 10:00:31] <rmatte> to match the name of the device

[01-Mar-2010 10:05:53] <Skippy1> genius ! - its working :)

[01-Mar-2010 10:06:57] <malbon> skippy1

[01-Mar-2010 10:07:02] <malbon> doh! :)

[01-Mar-2010 10:07:18] <malbon> Skippy1: yes, I have a script, just not documented it yet, runs in cron.

[01-Mar-2010 10:07:28] fOrsberg is now known as forsberg

[01-Mar-2010 10:08:02] <Skippy1> malbin - rmatte just knocked something together for me - but I'm keen to see some more examples - helps me understand what is happending and how it works

[01-Mar-2010 10:08:22] <malbon> I've got a script which does a bit more, I'll pastebin it in a mo.

[01-Mar-2010 10:09:40] <malbon> Skippy1: http://fpaste.org/1j1N/

[01-Mar-2010 10:10:24] <malbon> Skippy1: It goes and detects the current IP in dns, and the one in zenoss and resets using dns as a preference. If a box isn't in DNS it puts it into maintenance.

[01-Mar-2010 10:11:39] <malbon> rmatte: do you have zenhub / pbdaemon foo?

[01-Mar-2010 10:12:47] <Skippy1> malbon - that sounds like what I'm after - the box that changes its IP registeres its change with dns - its just updating zenoss of the change

[01-Mar-2010 10:13:54] <malbon> Skippy1: yes indeed, I use it for monitoring our workstations. :)

[01-Mar-2010 10:15:13] <malbon> Skippy1: you need to modify line 19 to the correct path in your environment, so it may be dmd.Devices.Server.Windows.getSubDevices() for example

[01-Mar-2010 10:15:37] <rmatte> malbon: zenhub / pbdaemon foo?

[01-Mar-2010 10:16:19] <rmatte> malbon: I just threw the script together to accept the IP as a commandline option and change the IP of whatever device, I did it on the spot in 2 mins lol

[01-Mar-2010 10:16:33] <rmatte> clearly you've put more time in to yours

[01-Mar-2010 10:17:05] <malbon> rmatte: yeah, how do I do the equivelent of a getConfigs() but for OSProcess items?

[01-Mar-2010 10:17:37] <malbon> rmatte: yeah, took a couple of hours all in, but then I'd not played with zendmd quite this way before, and I am paranoid so a lot of testing was done.

[01-Mar-2010 10:20:36] <rmatte> malbon: what value are you looking to have it return for processes?

[01-Mar-2010 10:22:56] <malbon> rmatte: I'm looking at writing my own collector for the HPUX process table which is similar to the HostResources one, but being proprietory using different oid's.

[01-Mar-2010 10:23:16] <rmatte> ok, but getConfig is used to retrieve properties, such as maxval

[01-Mar-2010 10:23:35] <malbon> rmatte: I'd also like to use this process long term for collecting vpn processes off of Cisco ASAs

[01-Mar-2010 10:24:20] <rmatte> Description of getConfig()

[01-Mar-2010 10:24:20] <rmatte>         """Return the config used by the collector to process simple min/max

[01-Mar-2010 10:24:20] <rmatte>         thresholds. (id, minval, maxval, severity, escalateCount)

[01-Mar-2010 10:24:28] <rmatte> that's what you're trying to do, but for processes?

[01-Mar-2010 10:25:20] <rmatte> Maybe explain what you're hoping to accomplish with the function that you're searching for?

[01-Mar-2010 10:25:29] <malbon> rmatte: I have mainly been looking at the Zenhub xmlrpc interface, which is slightly different.

[01-Mar-2010 10:25:43] <rmatte> right, but what does that have to do with getConfig()

[01-Mar-2010 10:25:45] <rmatte> ?

[01-Mar-2010 10:43:16] <theacolyte> rmatte: re my SSH vs SFTP question, I found that the SSH check isn't playing nice with my sftp server

[01-Mar-2010 10:43:26] <theacolyte> It doesn't appear to be closing the connection once done

[01-Mar-2010 10:47:00] <rmatte> theacolyte: I see

[01-Mar-2010 10:47:11] <theacolyte> It's almost certainly my sftp server's fault

[01-Mar-2010 10:47:19] <theacolyte> since it runs on Windows by nature it's hackish

[01-Mar-2010 10:47:28] <rmatte> what I was saying is that the sftp server IS the ssh server

[01-Mar-2010 10:47:31] <theacolyte> But, I was curious to see how the checks were performed

[01-Mar-2010 10:47:33] <rmatte> sftp is just ftp over ssh

[01-Mar-2010 10:47:37] <theacolyte> Yeah, no, I understand it

[01-Mar-2010 10:47:39] <rmatte> there's no distinction between the processes

[01-Mar-2010 10:47:41] <rmatte> k

[01-Mar-2010 10:48:06] <rmatte> so are you like specifically monitoring sftp somehow?

[01-Mar-2010 10:48:12] <rmatte> I don't even get where sftp comes in to this

[01-Mar-2010 10:48:22] <rmatte> all you need to monitor is ssh, if ssh goes down sftp goes down

[01-Mar-2010 10:48:30] <rmatte> as long as ssh is working, sftp will work

[01-Mar-2010 10:48:31] <theacolyte> I'd just be curious how the check is performed... SFTP is listed and SSH FYI

[01-Mar-2010 10:48:38] <theacolyte> It's possible I need to write something that logs in then disconnects

[01-Mar-2010 10:48:56] <rmatte> it probably just connects in different ways, but it's absolutely pointless to monitor sftp, since it IS ssh

[01-Mar-2010 10:48:59] <theacolyte> Is it just a socket connection on port 22?

[01-Mar-2010 10:49:11] <rmatte> it's just a client connecting to sshd with different options off the bat

[01-Mar-2010 10:49:32] <rmatte> it's just a connection to port 22 with a different header to kick-start the session than what's used with plain ssh

[01-Mar-2010 10:49:39] <rmatte> but it IS ssh

[01-Mar-2010 10:49:41] <rmatte> ;)

[01-Mar-2010 10:49:42] <theacolyte> ah got it

[01-Mar-2010 10:49:49] <theacolyte> I know what sftp is

[01-Mar-2010 10:50:07] <theacolyte> What I'm trying to figure out is how to not confuse my sftp server

[01-Mar-2010 10:51:06] <rmatte> ok, just so that I understand...

[01-Mar-2010 10:51:14] <rmatte> you're monitoring it in Zenoss as both ssh and sftp?

[01-Mar-2010 10:51:40] <rmatte> ...and if you monitor it for ssh, it doesn't cause any issues

[01-Mar-2010 10:51:46] <rmatte> but when you monitor it for both it does?

[01-Mar-2010 10:51:51] <rmatte> or am I missing something here?

[01-Mar-2010 10:52:39] <nickanderson> Hello

[01-Mar-2010 10:52:46] <theacolyte> No, no sorry

[01-Mar-2010 10:53:14] <rmatte> nickanderson: hello

[01-Mar-2010 10:53:23] <theacolyte> I've tried both SFTP and SSH (which I guess I expected to work the same), I'm trying to figure out where in Zenoss the IP services are set up so I can see if I can add stuff to it

[01-Mar-2010 10:53:29] <theacolyte> Like login, then quit

[01-Mar-2010 10:53:56] <rmatte> so is this IPService monitoring?

[01-Mar-2010 10:54:07] <rmatte> or actual SSH monitoring to substitute for SNMP?

[01-Mar-2010 10:54:31] <Simon4> theacolyte: you may have to make the ssh test a zencommand perhaps?

[01-Mar-2010 10:54:49] <theacolyte> ipservice monitoring

[01-Mar-2010 10:54:55] <theacolyte> yeah maybe

[01-Mar-2010 10:55:00] <rmatte> then you'll want to go look in the Services section

[01-Mar-2010 10:55:08] <Simon4> getting an ipservice check to negotiate ssh would be tricky I think

[01-Mar-2010 10:55:11] <theacolyte> all I know is I got a call on Friday saying our sftp server was down and I looked in the logs and zenoss was making connections that weren't dying

[01-Mar-2010 10:55:15] <nickanderson> I have some hardware that the provided mibs for are not correct, The mib has a definition for 1.3.6.1.4.1.2.6.158.5.1 but it does not send traps on that, it sends traps on 1.3.6.1.4.1.2.6.158.5.1.0. How should I go about transforming traps sent to 1.3.6.1.4.1.2.6.158.5.1.0 and drop that last .0?

[01-Mar-2010 10:56:07] <rmatte> theacolyte: Services -> IpService -> Privileged

[01-Mar-2010 10:56:18] <rmatte> search for sftp

[01-Mar-2010 10:56:43] <rmatte> you can search for ssh as well

[01-Mar-2010 10:56:53] <rmatte> you can enable/disable monitoring of those services from there, then remodel the device

[01-Mar-2010 10:57:18] <rmatte> you can also change whatever other settings (though there aren't many)

[01-Mar-2010 10:58:24] <theacolyte> Not monitoring it ah

[01-Mar-2010 10:58:35] <theacolyte> it probably just makes a socket connection

[01-Mar-2010 10:59:26] <rmatte> yeh

[01-Mar-2010 11:00:00] <rmatte> you would see it listed in the IP Services section on the processes tab for the device if you're monitoring it on that device

[01-Mar-2010 11:00:51] <theacolyte> yeah

[01-Mar-2010 11:03:29] <sektorNBA> i got an error building zenoss from source at freebsd 7.2 . can anyone help me?

[01-Mar-2010 11:04:04] <Skippy1> is there a cheat in python like there is in perl to run binary files ? using the `` ?

[01-Mar-2010 11:09:23] <rmatte> sektorNBA: do you absolutely have to use FreeBSD for your Zenoss server?

[01-Mar-2010 11:09:35] <rmatte> sektorNBA: It's much much easier to maintain Zenoss on a Linux server

[01-Mar-2010 11:09:59] <rmatte> I like FreeBSD but it's really not the way to go for Zenoss

[01-Mar-2010 11:10:45] <mrayzenoss> sektorNBA: unless you want to maintain a port :)

[01-Mar-2010 11:10:54] <rmatte> Like, you might get it installed and working on FreeBSD, but then just wait until you need to upgrade, you'll be crying.

[01-Mar-2010 11:10:55] <mrayzenoss> BTW, anyone using Netbotz?

[01-Mar-2010 11:11:18] <rmatte> Or you'll have problems that just magically pop up, some day one of your daemons suddenly won't start properly

[01-Mar-2010 11:11:24] <rmatte> I speak from experience in here

[01-Mar-2010 11:11:31] <rmatte> it's always the same story

[01-Mar-2010 11:11:46] <rmatte> and yes, if someone would maintain a port, that would help a lot

[01-Mar-2010 11:12:11] <mrayzenoss> really we need an experienced Zenoss and FreeBSD user to maintain a port

[01-Mar-2010 11:12:27] <mrayzenoss> I did get to talk to the guy who wrote the howto for FreeBSD at SCALE

[01-Mar-2010 11:12:47] <rmatte> definitely... once I get my lab server setup at home, I'll install FreeBSD on one of the VMs and screw around with it

[01-Mar-2010 11:13:13] <rmatte> first I need to rent a dedicated box and move everything off my home server to that, then I need to buy more RAM for my home server, and then install ESX on it.

[01-Mar-2010 11:13:34] <rmatte> could be a few months before I'm ready to go

[01-Mar-2010 11:14:02] <rmatte> I'm going to try to have 4 VMs running on it, one with redhat/rpm, one with ubuntu/stack, one with appliance, and one with FreeBSD

[01-Mar-2010 11:14:16] <rmatte> that'll allow me to develop and test for a wide range of stuff

[01-Mar-2010 11:20:51] <mrayzenoss> Last call for ZenPacks for the Cloud contest

[01-Mar-2010 11:24:57] <sektorNBA> rmatte that vm( appliance) pisses me off.

[01-Mar-2010 11:25:09] <sektorNBA> mrayzenoss my boss want it at my firewalls.

[01-Mar-2010 11:25:27] <rmatte> weird, I have a device which apparently doesn't have any zProperties that exist

[01-Mar-2010 11:25:37] <rmatte> even though I can see them in the UI, they've somehow been wiped from the db

[01-Mar-2010 11:26:06] <sektorNBA> mrayzenoss iam not that good to maitain a port *YET*

[01-Mar-2010 11:26:41] <mrayzenoss> sektorNBA: yeah, there are a couple of FreeBSD people who may get involved

[01-Mar-2010 11:26:58] <mrayzenoss> so the word I got from our build engineer is that SVN trunk switches to Python 2.6 this weekend

[01-Mar-2010 11:27:31] <ke4qqq> wonder how many zenpacks will break

[01-Mar-2010 11:27:47] <ke4qqq> any chance your build engineer will be churning out packages for us to beat on?

[01-Mar-2010 11:27:48] <mrayzenoss> we plan on getting alpha builds out soon

[01-Mar-2010 11:27:56] <mrayzenoss> definitely

[01-Mar-2010 11:28:12] <mrayzenoss> and I'm looking at a way to automatically install every ZenPack in SVN

[01-Mar-2010 11:28:26] <mrayzenoss> so I can at least unit test them that way

[01-Mar-2010 11:28:38] <rmatte> that wouldn't be overly hard to do, though you'd have to take in to account dependencies

[01-Mar-2010 11:28:57] <mrayzenoss> rmatte: agreed, I'd probably just make a list and install them in that order

[01-Mar-2010 11:29:53] <mrayzenoss> speaking of ZenPacks… anyone using Netbotz?  I've got a patch for the Netbotz ZenPack that completely changes the OIDs that are used, so I don't know if it's a separate template or not

[01-Mar-2010 11:31:54] <rmatte> can't say that I am

[01-Mar-2010 11:33:22] <ke4qqq> sadly no - though I wish I was - a consulting gig I did a year or so back had them - they were nice.

[01-Mar-2010 11:33:45] <mrayzenoss> the patch mentions which models were tested against, the version that's up doesn't

[01-Mar-2010 11:35:29] <rmatte> I love it when I have to wait half an hour for a "sudo su" to go through

[01-Mar-2010 11:35:38] * rmatte taps his foot

[01-Mar-2010 11:43:54] <forsberg> sudo bash!

[01-Mar-2010 11:43:55] <forsberg> ;>

[01-Mar-2010 11:45:06] <rmatte> :P

[01-Mar-2010 11:45:25] <rmatte> doesn't make a difference, same effect

[01-Mar-2010 11:46:08] <rmatte> rmatte@noc:~$ sudo su

[01-Mar-2010 11:46:08] <rmatte> root@noc:/home/rmatte# echo "$SHELL"

[01-Mar-2010 11:46:11] <rmatte> /bin/bash

[01-Mar-2010 11:46:14] <rmatte> :P

[01-Mar-2010 11:46:17] <rmatte> shorter to type

[01-Mar-2010 12:06:27] forsberg is now known as fOrsberg

[01-Mar-2010 12:09:05] <sektorNBA> I got ./libxml/encoding.h:29:19: error: iconv.h: No such file or directory

[01-Mar-2010 12:10:41] <ckrough> 'sudo su -'

[01-Mar-2010 12:10:43] fOrsberg is now known as forsberg

[01-Mar-2010 12:10:50] <ckrough> trailing slash gets you the env

[01-Mar-2010 12:11:15] forsberg is now known as fOrsberg

[01-Mar-2010 12:11:26] <rmatte> ckrough: I know, I generally do sudo su - zenoss

[01-Mar-2010 12:11:31] <rmatte> I just don't bother for root

[01-Mar-2010 12:11:39] fOrsberg is now known as forsberg

[01-Mar-2010 12:12:49] forsberg is now known as fOrsberg

[01-Mar-2010 12:15:49] fOrsberg is now known as forsberg

[01-Mar-2010 12:28:00] <pils> hello... i have the cisco mib pack (the very large one), installed... and i set the switches to report traps on several items.  when zenoss receives the trap, it's reporting it as the oid, such as:  summary: snmp trap 1.3.6.1.4.1.9.9.43.2.1

[01-Mar-2010 12:28:11] <pils> Is there a way to get that to display in more friendly format?

[01-Mar-2010 12:29:05] <MrWGW> got a question, what is the path to the Zenoss logo file?

[01-Mar-2010 12:29:15] <MrWGW> that displays on the top header?

[01-Mar-2010 12:29:35] <ckrough> I think it's under portal_skins in zope

[01-Mar-2010 12:30:11] <MrWGW> thanks

[01-Mar-2010 12:30:55] <ckrough>  /zport/portal_skins/zenui/logo.png/manage_main

[01-Mar-2010 12:52:11] <mrayzenoss> pils: in zendmd, you'll need to dmd.Mibs.reindex_all() and commit() if you're running 2.5.1 (bug that was fixed in 2.5.2)

[01-Mar-2010 12:58:56] <Skippy1> pils: I had this too - the solution for me is to use event translation

[01-Mar-2010 13:00:16] <malbon> pils: you could install the mib, that should give the incoming trap a name to translate too.

[01-Mar-2010 13:00:16] <pils> I did...

[01-Mar-2010 13:00:19] <pils> I can try it again

[01-Mar-2010 13:04:10] <Skippy1> anyone help with a weirdism I have - on a linux box I have the HD with /home monitored 0 it is a 400gb drive which reports -267.4gb free and 294% utilisation - the drive really has 30g free and 78% used, any ideas why zennoss on this drive only has got its thong in a knot

[01-Mar-2010 13:10:57] <pils> mrayzenoss: any other ideas?  i installed the zenpack, then ran those two commands in zendmd

[01-Mar-2010 13:32:48] <mrayzenoss> If anyone's looking to do some Zenoss contracting: thread/12893

[01-Mar-2010 14:03:25] <rmatte> that dual-horizon requirement wouldn't be the simplest thing to do

[01-Mar-2010 14:04:16] <rmatte> it'd have to be a daemon that runs in the Zenoss instance at each site, keeps track of things and reports back

[01-Mar-2010 14:07:13] <mrayzenoss> I believe he wants some sort of active/passive monitoring, where they monitor a Zenoss installation's health from another install and receive events from the monitored install

[01-Mar-2010 14:07:17] <mrayzenoss> a Zenoss ZenPack

[01-Mar-2010 14:35:02] <Skippy1> eveyone in bed ?

[01-Mar-2010 14:35:19] <mrayzenoss> I'm in my office :)

[01-Mar-2010 14:35:57] <Skippy1> hehe - when zenoss graphs a device - it goes and gets the info it wants - is it possible for the info to be given to zenoss to graph ?

[01-Mar-2010 14:37:00] <forsberg> I think you gotta be more specific :)

[01-Mar-2010 14:39:11] <Skippy1> i run a util that collects a series of data from various sources and comes back with a result - i wondered if I could give that result to zenoss and then it create a graph from that result - the problem is it runs at fairly random times and sometimes manually also

[01-Mar-2010 14:40:56] <forsberg> i think you are better off with just getting the data every 5 minutes, and do a proper graph, but hmmm :

[01-Mar-2010 14:41:40] <forsberg> or make an output file and make zenoss poll that file

[01-Mar-2010 14:41:46] <Skippy1> it often takes more than 5 minutes to contact all the sources and compute the results

[01-Mar-2010 14:42:35] <forsberg> well then just get the output into the file and then make datapoints from looking the file

[01-Mar-2010 14:43:15] <Skippy1> yeah I thought of that - its an option

[01-Mar-2010 14:43:27] <Skippy1> gotta run - on call and being called 8-)

[01-Mar-2010 14:49:02] <rmatte> well, I just got tasked with migrating all of the changes that were made to our ticketing system to a newer version, this shall be fun

[01-Mar-2010 14:54:22] <theacolyte> hope your documentation is good :>

[01-Mar-2010 14:55:28] <rmatte> well, whenever a file was edited it was copied to a .bak, so I have reference from that.  There are also comments for each modification

[01-Mar-2010 14:55:33] <rmatte> so it hopefully shouldn't be too painful

[01-Mar-2010 14:58:33] <theacolyte> ah, yeah

[01-Mar-2010 14:58:38] <theacolyte> 'just a few diffs and commnets away then

[01-Mar-2010 14:58:41] <theacolyte> starting the timer now

[01-Mar-2010 14:58:54] <rmatte> nah, it's going to take a little more than that lol

[01-Mar-2010 14:59:06] <rmatte> the code in the newer version is quite different

[01-Mar-2010 14:59:13] <theacolyte> You're usin RT right

[01-Mar-2010 15:01:43] <rmatte> nope

[01-Mar-2010 15:01:44] <rmatte> otrs

[01-Mar-2010 15:02:04] <theacolyte> ah, right

[01-Mar-2010 16:09:12] forsberg is now known as fOrsberg

[02-Mar-2010 00:00:30] [disconnected at Tue Mar  2 00:00:30 2010]

[02-Mar-2010 00:00:31] [connected at Tue Mar  2 00:00:31 2010]

[02-Mar-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[02-Mar-2010 01:49:36] fOrsberg is now known as forsberg

[02-Mar-2010 03:24:01] salawank is now known as salax

[02-Mar-2010 06:12:46] <baffle> How do I associate an IP address with a device, even if the this IP does not really belong to the device? I.e. a server has an IPMI BMC, and I want events from the BMC to be associated with a device.

[02-Mar-2010 08:27:57] <sektorNBA> hi

[02-Mar-2010 08:28:12] <sektorNBA> i am trying to install zenoss at freebsd. I got this at the end

[02-Mar-2010 08:28:12] <sektorNBA> installing conf/zope.conf.in

[02-Mar-2010 08:28:12] <sektorNBA> installing conf/hubpasswd

[02-Mar-2010 08:28:12] <sektorNBA> ERROR 1045 (28000): Access denied for user 'zenoss'@'localhost' (using password: YES)

[02-Mar-2010 08:28:13] <sektorNBA> During the MySQL install, the supplied MySQL user (zenoss) or password is incorrect

[02-Mar-2010 08:28:27] <sektorNBA> where i change the user/pass it use?

[02-Mar-2010 08:30:20] <sektorNBA> hello

[02-Mar-2010 08:34:47] <Simon4> sektorNBA: http://www.sysadminwiki.net/wiki/index.php?title=Common_Zenoss_Errors_-_Post_Install

[02-Mar-2010 08:35:18] <sektorNBA> ty

[02-Mar-2010 08:47:22] <sektorNBA> i need libcrypto.so.4 now

[02-Mar-2010 08:47:23] <sektorNBA> shit

[02-Mar-2010 08:50:58] <sektorNBA> Simon4 i build zenoss from source. how do i start it?

[02-Mar-2010 08:51:38] <Simon4> sektorNBA: hmm, $ZENHOME/bin/zenoss start

[02-Mar-2010 08:51:59] <Simon4> as user zenoss, most likely

[02-Mar-2010 08:52:17] <sektorNBA> k

[02-Mar-2010 08:52:58] <sektorNBA> no lucky. /usr/local/zenoss/bin/zenpackdaemons: line 21: 35682 Segmentation fault: 11 (core dumped) $PYTHON $ZENHOME/Products/ZenUtils/ZenPackDaemons.py "$CMD" $*

[02-Mar-2010 08:53:11] <Simon4> yukes

[02-Mar-2010 08:53:13] <Simon4> yikes even

[02-Mar-2010 08:59:20] <sektorNBA> do i need to set a $PYTHON variable ?

[02-Mar-2010 09:00:49] <Simon4> possibly - it should point to your python executable if so

[02-Mar-2010 09:01:14] <Simon4> I ahven't set up from compiled before, so will only be guessing

[02-Mar-2010 09:01:20] <Simon4> the segfault is a bit disturbing though

[02-Mar-2010 09:01:26] forsberg is now known as fOrsberg

[02-Mar-2010 09:05:46] <sektorNBA> i didnt see that in docs. which directory is it?

[02-Mar-2010 09:14:10] <mrayzenoss> $ZENHOME/python/

[02-Mar-2010 09:16:16] <sektorNBA> mrayzenoss i haven´t it. ;(

[02-Mar-2010 09:17:47] <sektorNBA> [zenoss@caos /usr/local/zenoss]$ ls

[02-Mar-2010 09:17:48] <sektorNBA> Products etc include share zenoss-2.5.1

[02-Mar-2010 09:17:48] <sektorNBA> bin export lib skel zenoss-2.5.1.tar.gz

[02-Mar-2010 09:17:48] <sektorNBA> doc extras libexec var

[02-Mar-2010 09:18:13] <mrayzenoss> is it in $ZENHOME/zenoss/python?

[02-Mar-2010 09:21:03] <rmatte> hmmmm, does anyone know what the best way would be to diff 2 directories (hopefully in some sort of way where I can browse through the diffs file by file)?

[02-Mar-2010 09:21:15] <mrayzenoss> I use a tool called dirdiff

[02-Mar-2010 09:21:32] <rmatte> ah, I have that installed too

[02-Mar-2010 09:21:35] <rmatte> I'll give it a try

[02-Mar-2010 09:21:50] <rmatte> thanks

[02-Mar-2010 09:22:08] <mrayzenoss> it's in TK but I haven't found anything I like better yet

[02-Mar-2010 09:22:26] <rmatte> I don't mind TK at all, some of my favourite apps are TK apps ;)

[02-Mar-2010 09:22:45] <rmatte> I used to do quite a bit of coding in TCL back in the day

[02-Mar-2010 09:22:48] <mrayzenoss> it'll do 3 way diffs which is handy

[02-Mar-2010 09:22:49] <sektorNBA> mrayzenoss there is no zenoss in $ZENHOME. weird.

[02-Mar-2010 09:22:58] <rmatte> looks like it can do up to 4 way diffs

[02-Mar-2010 09:23:06] <sektorNBA> mrayzenoss even the zenoss vm applicance ( rh ) has no python

[02-Mar-2010 09:23:37] <mrayzenoss> I believe on the VM it's the system Python

[02-Mar-2010 09:24:16] <sektorNBA> declare -x PYTHONPATH="/usr/local/zenoss/lib/python"

[02-Mar-2010 09:24:20] <sektorNBA> hummm

[02-Mar-2010 09:24:55] <rmatte> it shouldn't be $ZENHOME/zenoss

[02-Mar-2010 09:25:01] <sektorNBA> i guess i need to set $PYTHON to /usr/local/zenoss/lib/python, right ?

[02-Mar-2010 09:25:04] <rmatte> Zenoss won't event run with that path

[02-Mar-2010 09:25:06] <rmatte> even*

[02-Mar-2010 09:25:21] <rmatte> your $ZENHOME should be /usr/local/zenoss

[02-Mar-2010 09:25:29] <sektorNBA> rmatte it is.

[02-Mar-2010 09:25:31] <rmatte> k

[02-Mar-2010 09:25:36] <rmatte> then yes, you probably should

[02-Mar-2010 09:25:41] <sektorNBA> declare -x ZENHOME="/usr/local/zenoss"

[02-Mar-2010 09:27:03] <rmatte> declare, you using csh or something?

[02-Mar-2010 09:27:19] <sektorNBA> bash. i got that from export

[02-Mar-2010 09:27:27] <rmatte> oh

[02-Mar-2010 09:31:09] <rmatte> there really needs to be some sort of stack installer for FreeBSD

[02-Mar-2010 09:31:22] <sektorNBA> i set $PYTHON to /usr/local/zenoss/lib/python but i still got erros

[02-Mar-2010 09:31:29] <rmatte> people spend more time setting up Zenoss on FreeBSD than it would take to do a complete Linux install and install FreeBSD

[02-Mar-2010 09:31:42] <rmatte> erm, and install Zenoss rather

[02-Mar-2010 09:32:02] <rmatte> does /usr/local/zenoss/lib/python exist?

[02-Mar-2010 09:32:06] <sektorNBA> yes

[02-Mar-2010 09:32:12] <rmatte> same segmentation fault?

[02-Mar-2010 09:32:19] <sektorNBA> a lot.

[02-Mar-2010 09:32:31] <rmatte> hmmm

[02-Mar-2010 09:32:41] <sektorNBA> let me use pastebin

[02-Mar-2010 09:33:27] <sektorNBA> http://pastebin.com/85eUr1vP

[02-Mar-2010 09:33:28] <rmatte> http://www.google.ca/url?sa=t&source=web&ct=res&cd=1&ved=0CAYQFjAA&url=http%3A%2F%2Fcommunity.zenoss.org%2Fthread%2F12079%3Bjsessionid%3D91154E2D5F9B41B625CF237AAAEEF1C5.node0&rct=j&q=zenoss+freebsd+segmentation+fault&ei=my-NS4T1GcyXtgfz1KXwCA&usg=AFQjCNFyZUltf0prWg9oEMV54CXD93Uqag

[02-Mar-2010 09:33:30] <rmatte> oops

[02-Mar-2010 09:33:39] <rmatte> thread/12079;jsessionid=91154E2D5F9B41B625CF237AAAEEF1C5.node0

[02-Mar-2010 09:33:43] <rmatte> common issue apparently

[02-Mar-2010 09:34:44] <rmatte> Here's a link to the steps to install 2.4.1 on FreeBSD, there are some hacks involved: docs/DOC-3514.pdf;jsessionid=D0F2CFD909AF0A461DCF870F79FE5FA0.node

[02-Mar-2010 09:34:48] <rmatte> it's worth a read

[02-Mar-2010 09:35:22] <rmatte> but apparently even after following those steps, some people were still getting segmentation faults

[02-Mar-2010 09:35:34] <rmatte> do you have linux compatibility mode enabled in FreeBSD?

[02-Mar-2010 09:35:55] <rmatte> the python package that comes with the source is compiled for Linux I believe

[02-Mar-2010 09:39:33] <sektorNBA> rmatte yes i have it enabled.

[02-Mar-2010 09:46:59] <sektorNBA> http://www.freebsd.org/doc/en/books/handbook/linuxemu-lbc-install.html

[02-Mar-2010 09:47:05] <sektorNBA> i will install that package

[02-Mar-2010 09:47:40] <sektorNBA> ==> linux_base-f10-10_2 compat.linux.osrelease: 2.4.2 is not supported, please use 2.6.16. weird

[02-Mar-2010 09:52:01] <jb> why do you insist on using fbsd again?

[02-Mar-2010 09:52:01] <jb> :P

[02-Mar-2010 09:54:58] <sektorNBA> cause i dont like linux. i am running zenoss at vmware. it sucks

[02-Mar-2010 09:56:38] <rmatte> you may not likely linux, but when it comes to Zenoss it'll avoid tons of headaches

[02-Mar-2010 09:56:44] <rmatte> s/likely/like

[02-Mar-2010 09:57:26] <rmatte> I used to be a FreeBSD fanboy too, but I ended up learning Linux and honestly, I don't see a massive difference in performance or security

[02-Mar-2010 09:57:41] <rmatte> and I definitely see a huge difference in compatibility

[02-Mar-2010 09:57:43] <sektorNBA> i hate iptables.

[02-Mar-2010 09:58:00] <rmatte> honestly, after messing around with iptables, I prefer it to pf

[02-Mar-2010 09:58:21] <rmatte> but I guess it really depends on what you want to do with it

[02-Mar-2010 09:58:36] <jb> sektorNBA: well your life will be 100% easier with linux.

[02-Mar-2010 09:58:39] <jb> not to mention upgrades

[02-Mar-2010 09:58:59] <rmatte> yeh, installing Zenoss on Linux is braindead simple... 5 minutes and you're up and running

[02-Mar-2010 09:59:21] <rmatte> also, source upgrades are even more painful than the initial source install from what I've heard

[02-Mar-2010 09:59:38] <rmatte> If there was a reliable Zenoss port then it would make sense to use FreeBSD, but there isn't

[02-Mar-2010 09:59:46] <ckrough> rpm -ivh zenoss*rpm

[02-Mar-2010 09:59:50] <ckrough> service zenoss start

[02-Mar-2010 09:59:51] <ckrough> tada!

[02-Mar-2010 10:00:06] <jb> and upgrades are the same :)

[02-Mar-2010 10:00:10] <ckrough> yup

[02-Mar-2010 10:00:12] <rmatte> yeh, or ./zenoss*bin

[02-Mar-2010 10:00:13] <rmatte> tada

[02-Mar-2010 10:00:14] <jb> -Uvh :)

[02-Mar-2010 10:00:16] <rmatte> ;)

[02-Mar-2010 10:00:22] <ckrough> "The Way"

[02-Mar-2010 10:00:29] <jb> but anyways, upgrading an installation on FBSD is going to be far from trivial

[02-Mar-2010 10:00:35] <jb> not worth the headache..

[02-Mar-2010 10:00:53] <sektorNBA> someone will create a port soon. zenoss is a great tool.

[02-Mar-2010 10:01:04] <jb> doubt it

[02-Mar-2010 10:01:09] <sektorNBA> i was testing cacti, zabbix and others but zenos is best.

[02-Mar-2010 10:01:16] <rmatte> sektorNBA: also, if you don't like iptables, you can always install pf

[02-Mar-2010 10:01:43] <ckrough> zenoss' extensibility is where it really starts to shine

[02-Mar-2010 10:01:43] <rmatte> If someone does create a port it's going to have to be properly maintained

[02-Mar-2010 10:02:11] <rmatte> When I was using FreeBSD that was my biggest pet peeve, ports that were like 10 versions behind the current release version

[02-Mar-2010 10:02:17] <sektorNBA> rmatte linux support pf?

[02-Mar-2010 10:02:29] <rmatte> sektorNBA: I'm pretty sure that I heard it does

[02-Mar-2010 10:02:36] * rmatte checks

[02-Mar-2010 10:03:10] <sektorNBA> google find some docs.

[02-Mar-2010 10:03:26] <rmatte> Well, here's 1, not sure how old it is

[02-Mar-2010 10:03:28] <rmatte> http://glacier.lbl.gov/rsync.pf.Linux-i386.html

[02-Mar-2010 10:04:44] <ckrough> my gut tells me you would be better off learning iptables to save yourself the headache

[02-Mar-2010 10:04:55] <ckrough> unless you are doing really complex filter on the box

[02-Mar-2010 10:07:36] <rmatte> yeh

[02-Mar-2010 10:07:50] <rmatte> there's nothing wrong with iptables for the majority of tasks

[02-Mar-2010 10:08:17] <rmatte> If you find it difficult to configure you can always use webmin to configure it

[02-Mar-2010 10:08:33] <jb> gross :P

[02-Mar-2010 10:10:15] <rmatte> meh, webmin makes stuff quicker, I use it for managing ldap and dns (though I manage bind manually at home, we just do way too much dns changes to do it all by manually editing config files)

[02-Mar-2010 10:11:15] <rmatte> sektorNBA: gentoo is pretty close to FreeBSD in terms of package management and setup, it helped ease the transition from FreeBSD to Linux for me.

[02-Mar-2010 10:11:30] <jb> erm, i manage over 2000 zones

[02-Mar-2010 10:11:38] <rmatte> by hand?

[02-Mar-2010 10:11:40] <jb> and up until last year, it was done all by hand

[02-Mar-2010 10:11:42] <jb> and it wasn't hard :P

[02-Mar-2010 10:11:57] <rmatte> meh

[02-Mar-2010 10:12:07] <rmatte> I just find it tedious and prone to errors

[02-Mar-2010 10:12:13] <rmatte> but to each his own

[02-Mar-2010 10:12:17] <sektorNBA> i use different softwares. dns for example i use tinydns

[02-Mar-2010 10:12:21] <jb> yeah, guess im just used to it

[02-Mar-2010 10:13:26] fOrsberg is now known as forsberg

[02-Mar-2010 10:32:21] RobertLaptop_ is now known as RobertLaptop

[02-Mar-2010 10:32:27] RobertLaptop is now known as RobertLaptop_

[02-Mar-2010 11:10:13] <cgibbons> hmmm

[02-Mar-2010 11:15:54] <cgibbons> anyone online that's a Twisted fan/guru?

[02-Mar-2010 11:25:58] <rmatte> I'm a *twisted* fan/guru of Zenoss, but not in the sense of the software project ;)

[02-Mar-2010 11:26:48] <rmatte> ...and I am quickly starting to hate perl.

[02-Mar-2010 11:27:18] <cgibbons> good, you're sane then

[02-Mar-2010 11:27:30] <rmatte> lol

[02-Mar-2010 11:28:09] <rmatte> duplicating mods that we made to otrs 2.3.4 and moving them over to otrs 2.4.7, but they've rewritten large portions of it, so the mods are not as straight forward as they used to be

[02-Mar-2010 12:57:33] Simon4_ is now known as SimonH

[02-Mar-2010 12:57:55] <SimonH> hur... freenode irritating

[02-Mar-2010 12:57:59] SimonH is now known as Simon2

[02-Mar-2010 12:58:04] Simon2 is now known as Simon3

[02-Mar-2010 13:21:18] <sciolist> is there any graphical model of what permissions are available at which role (i.e., Manager vs. zenmanager vs zenuser)?

[02-Mar-2010 13:25:51] <ckrough> there is http://your_server/zport/dmd/manage , then click on the security tab

[02-Mar-2010 13:25:56] <ckrough> not sure if thats what you mean though

[02-Mar-2010 13:27:57] <sciolist> that will work... just trying to explain to folks who will have ZenManager on some objects, but with a default role of ZenUser

[02-Mar-2010 13:28:02] <sciolist> thanks

[02-Mar-2010 14:13:56] mrayzenoss1 is now known as mrayzenoss

[02-Mar-2010 15:07:36] <cgibbons> tada

[02-Mar-2010 15:07:46] <ckrough> no rabbit

[02-Mar-2010 15:40:43] <rmatte> eugh this is ridiculous, I need a javascript line to select a menu item from a dropdown menu that I created

[02-Mar-2010 15:40:54] <rmatte> I just want a line that highlights that item

[02-Mar-2010 15:41:11] <rmatte> basically I have a menu that says quick links, then you drop it down, select an item, and it opens in another window

[02-Mar-2010 15:41:19] <rmatte> I need it to then go back to quick links after that has happened

[02-Mar-2010 15:41:33] <rmatte> I've been googling for 15 minutes now, I would think this would be simple but there's nothing on it

[02-Mar-2010 15:41:43] <rmatte> all I've found is how to preserve form settings if submission fails

[02-Mar-2010 15:56:07] <nickanderson> Hello

[02-Mar-2010 15:56:46] <nickanderson> So I have this hardware thats sending traps

[02-Mar-2010 15:57:29] <nickanderson> I have the mib for the hardware that the manufacturer provided

[02-Mar-2010 15:58:44] <nickanderson> One of two things seems to be happening. Either zenoss is tacking on a .0 to the trap or the hardware is trapping to the .0 thats not defined in the mib

[02-Mar-2010 15:58:55] <nickanderson> its a generic trap

[02-Mar-2010 15:59:14] <nickanderson> so everything from powersupply failed to hard drive removed traps to the same oid

[02-Mar-2010 16:00:09] <nickanderson> I need to do transforms on it, any idea how I can make zenoss see that incoming trap as its definded in the mib (without that trailing .0)?

[02-Mar-2010 16:01:27] <rmatte> nickanderson: I've seen people with that issue before, and I know there's a way to fix that, but I'm not sure what it is, have you tried searching the forums?

[02-Mar-2010 16:01:34] <rmatte> I'm sure there are some posts on there about that

[02-Mar-2010 16:02:33] <rmatte> hmmmm, are you using v1 traps?

[02-Mar-2010 16:02:44] <rmatte> if so, does the device support using v2 traps?

[02-Mar-2010 16:03:53] <rmatte> also, what version of Zenoss are you using?

[02-Mar-2010 16:04:09] <rmatte> This appears to have been fixed in 2.4: http://dev.zenoss.com/trac/ticket/3965

[02-Mar-2010 16:04:15] <rmatte> so if you're using 2.5, that would be weird

[02-Mar-2010 16:04:24] <rmatte> but if you're using 2.3, you need to upgrade

[02-Mar-2010 16:04:42] <forsberg> :)

[02-Mar-2010 16:06:33] <mrayzenoss> 2.5.2 is up and in the process of being published

[02-Mar-2010 16:06:46] <nickanderson> im useing 2.5

[02-Mar-2010 16:07:24] <nickanderson> and snmpv1

[02-Mar-2010 16:07:31] <nickanderson> dosnt look like I have the option of sending v2 informs

[02-Mar-2010 16:09:12] <nickanderson> Yeah I saw a thread in the forms before, relating back to the 2.3 era

[02-Mar-2010 16:10:55] <rmatte> well, at this point, if there's a .0 being appended I doubt it's Zenoss doing it

[02-Mar-2010 16:11:21] <rmatte> you can always just make an event mapping against the untranslated OID and then transform it

[02-Mar-2010 16:11:25] <rmatte> it will work just as well

[02-Mar-2010 16:11:33] <rmatte> Mibs really just make traps look prettier

[02-Mar-2010 16:11:48] <nickanderson> yeah

[02-Mar-2010 16:12:03] <rmatte> but if you're transforming it, you won't even notice

[02-Mar-2010 16:12:11] <nickanderson> thanks for that link

[02-Mar-2010 16:12:16] <rmatte> np

[02-Mar-2010 16:12:16] <nickanderson> i can test it

[02-Mar-2010 16:12:18] <nickanderson> :)

[02-Mar-2010 16:12:26] <nickanderson> i think its the dumb hardware thats doing it

[02-Mar-2010 16:12:37] <nickanderson> might have to send a flame @ IBM

[02-Mar-2010 16:12:43] <rmatte> some hardware vendors are clueless when it comes to snmp

[02-Mar-2010 16:12:58] <nickanderson> IBM was second in line getting their enterprises OID

[02-Mar-2010 16:13:06] <nickanderson> you think they might know something

[02-Mar-2010 16:13:09] <rmatte> hehe

[02-Mar-2010 16:16:36] <nickanderson> uhhh

[02-Mar-2010 16:16:38] <nickanderson> wth

[02-Mar-2010 16:17:15] <nickanderson> snmptrap -v 1 -c public localhost 1.3.6.1.4.1.2.6.158.5.1 localhost 6 0 ''  1.3.6.1.4.1.2.6.158.5.1.9 s 'something happened'

[02-Mar-2010 16:17:26] <nickanderson> can you help me break that down

[02-Mar-2010 16:17:40] <nickanderson> i understand up to the first oid

[02-Mar-2010 16:17:51] <nickanderson> whats the second localhost mean?

[02-Mar-2010 16:19:34] <nickanderson> thats the agent

[02-Mar-2010 16:19:42] <nickanderson> then trap-type, whats that

[02-Mar-2010 16:19:47] <nickanderson> and whats specific-type?

[02-Mar-2010 16:19:57] <nickanderson> and uptime is the ''

[02-Mar-2010 16:31:53] <nickanderson> gah

[02-Mar-2010 16:31:57] <nickanderson> cool

[02-Mar-2010 16:32:13] <nickanderson> i think i see how to write transforms without the ease of a mib

[02-Mar-2010 17:04:22] <tuples> Well thats a bit strange

[02-Mar-2010 17:04:40] <tuples> 5 IPs, 4 http ports to monitor on each

[02-Mar-2010 17:04:56] <tuples> the second port monitored errors for each IP

[02-Mar-2010 17:05:03] <tuples> but the rest work

[02-Mar-2010 17:06:29] <tuples> sendto error, both are using the same host :?

[02-Mar-2010 17:43:18] <tuples> disabled ping for those and they all have cleared up for now

[02-Mar-2010 18:26:40] forsberg is now known as fOrsberg

[02-Mar-2010 21:28:03] <nickanderson> rmatte: I just saw your comment about the xenserver zenpack

[02-Mar-2010 21:28:37] <nickanderson> I looked at that zenpack, looks like its only for the opensource xen

[02-Mar-2010 21:29:21] <nickanderson> would need to add support for the xe command or the xapi (i think thats what its caled) for citrix xenserver

[02-Mar-2010 22:21:12] <nickanderson> anyone around?

[03-Mar-2010 00:00:30] [disconnected at Wed Mar  3 00:00:30 2010]

[03-Mar-2010 00:00:30] [connected at Wed Mar  3 00:00:30 2010]

[03-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[03-Mar-2010 01:32:51] fOrsberg is now known as forsberg

[03-Mar-2010 08:28:41] <ckrough> is it a good idea to reindex() before upgrades?

[03-Mar-2010 08:29:42] <rmatte> ckrough: doesn't make a difference

[03-Mar-2010 08:40:41] forsberg is now known as fOrsberg

[03-Mar-2010 09:32:04] <rmatte> well, looks like I may be setting up like 10+ more Zenoss boxes soon

[03-Mar-2010 09:57:32] <MrWGW> got a question

[03-Mar-2010 09:57:46] <MrWGW> what do the lines between locations visible in some zenoss screenshots signify, and how do I get them?

[03-Mar-2010 09:57:54] <MrWGW> on my network the locations are slef contained with no interconnecting lines

[03-Mar-2010 09:59:35] fOrsberg is now known as forsberg

[03-Mar-2010 10:07:54] <SEJeff_work> rmatte, Why that many?

[03-Mar-2010 10:09:28] <rmatte> SEJeff_work: that many new clients coming on

[03-Mar-2010 10:09:47] <rmatte> MrWGW: you mean on the google maps?

[03-Mar-2010 10:09:48] * SEJeff_work pats rmatte on the back

[03-Mar-2010 10:09:50] <SEJeff_work> Good stuff

[03-Mar-2010 10:10:08] <rmatte> MrWGW: all those lines means is that there are devices at both locations in the same IP subnet

[03-Mar-2010 10:10:29] <rmatte> MrWGW: so if you have different subnets per site you won't get those lines

[03-Mar-2010 10:11:05] <rmatte> SEJeff_work: well, I don't make the sales, I just setup the monitoring lol

[03-Mar-2010 10:21:31] <Brixius> Hello, I got a question on 2.5.1, I wan to add an existing template to a device, When I click on Template Add it gives me a dialog to enter a name, when I goto bind existing template(Which is what I would expect to use) I can only have one template on the device.

[03-Mar-2010 10:21:45] <Brixius> it replaces the existing template with the new one.

[03-Mar-2010 10:23:36] <cluther> You can ctrl-click to select multiple templates.

[03-Mar-2010 10:23:56] <Brixius> ok, figured it out, but looks like a bug, you have to select all templates you want to bind when you bind an existing template. It should just add the new ones and not overwrite the old ones.

[03-Mar-2010 10:24:44] <ckrough> is there an ETA on 2.5.2 Enterprise now that 2.5.2 core is out?

[03-Mar-2010 10:25:03] <cluther> Brixius: Maybe not the perfect UI, but that is by design. Otherwise you'd need another method for unbinding templates.

[03-Mar-2010 10:25:06] <SEJeff_work> ckrough, I was wondering the exact same thing

[03-Mar-2010 10:26:02] <cluther> ckrough, SEJeff_work: Checking on that..

[03-Mar-2010 10:33:02] <rmatte> cluther: long time no see

[03-Mar-2010 10:34:20] <rmatte> lol, the announcement on the community site that "Zenoss 2.5.1 now available!" should probably be changed.

[03-Mar-2010 10:34:40] <ckrough> hey, it's still true

[03-Mar-2010 10:46:58] * zenethian moseys in and puts down a large cauldron filled with espresso and a plate of chocolate oatmeal cookies.

[03-Mar-2010 10:54:42] <cluther> ckrough, SEJeff_work: Enterprise 2.5.2 bits are up at the usual location now.

[03-Mar-2010 10:55:18] <SEJeff_work> cluther, Thanks. Does that have the new features hinkle hinted about at SCALE? Where you upload the picture of the rack and it overlays status info ontop?

[03-Mar-2010 10:55:44] <Brixius> Anyone familer with ZenPacks.community.WMIDataSource I get the following error with anything that uses it. url[2], device.zWmiProxy = url[2].split('@')

[03-Mar-2010 10:55:44] <Brixius> ValueError: need more than 1 value to unpac

[03-Mar-2010 11:15:30] <rmatte> SEJeff_work: that feature is enterprise only, you an enterprise user?

[03-Mar-2010 11:15:40] <SEJeff_work> rmatte, Yes

[03-Mar-2010 11:15:55] <rmatte> k, not sure if that made it in to 2.5.2 or not

[03-Mar-2010 11:17:46] <malbon> it is a shame that the zenwinperf is enterprise only.

[03-Mar-2010 11:18:18] <rmatte> malbon: Egor's next WMI Data Source release will have the same functionality apparently

[03-Mar-2010 11:18:49] <malbon> rmatte: so it'll be a 'drop in replacement' so I only need to have 1 installed?

[03-Mar-2010 11:19:30] <rmatte> I'm not sure exactly how it'll work, I just know it'll do full WMI monitoring

[03-Mar-2010 11:19:38] <rmatte> including disks, interfaces, etc...

[03-Mar-2010 11:20:42] <SEJeff_work> rmatte, It works pretty well

[03-Mar-2010 11:21:03] <malbon> fair enough, still don't like the idea of running multiple WMI infrastructures, makes it difficult to diagnose faults.

[03-Mar-2010 11:21:06] <SEJeff_work> But zenoss has to pay the bills and keep making one of the best open sauce monitoring solutions in existence. No qualms here.

[03-Mar-2010 11:21:45] <rmatte> yeh, but honestly, it's more the support that sells, if you're a very large company you're going to buy the supported product hands down

[03-Mar-2010 11:21:50] <rmatte> or if you're government

[03-Mar-2010 11:22:14] <SEJeff_work> rmatte, And the support flippin' rocks. Their responsiveness was the only reason we decided to renew

[03-Mar-2010 11:22:19] <malbon> SEJeff_work: Yes I don't mind that either, I happily pay for support on core and get the Enterprise bits for free.

[03-Mar-2010 11:22:22] <rmatte> yeh, they are very good

[03-Mar-2010 11:22:39] <rmatte> the enterprise version also had more scalability, with things like device ACLs

[03-Mar-2010 11:22:51] <SEJeff_work> If you're into that :)

[03-Mar-2010 11:22:59] <SEJeff_work> and distributed collectors if you're global

[03-Mar-2010 11:23:06] <rmatte> but moving something like WMI monitoring to core wouldn't affect the bottom line in my opinion, if anything it would attract more pottential clients

[03-Mar-2010 11:23:12] <rmatte> potential*

[03-Mar-2010 11:23:33] <rmatte> well, I have a ZenPack which assists in configuring distributed collectors in core

[03-Mar-2010 11:23:40] <rmatte> I'm just waiting for approval from my management to release it

[03-Mar-2010 11:24:21] <rmatte> with the pack, it literally takes a couple of minutes to do the configuration

[03-Mar-2010 11:24:30] <SEJeff_work> rmatte, Sweet

[03-Mar-2010 11:26:27] <rmatte> It's commandline based but it works

[03-Mar-2010 11:26:42] <rmatte> you basically just install the pack, become zenoss user, type "zencollectortool" and follow the prompts

[03-Mar-2010 11:27:19] <rmatte> It eliminates all of the tedious config file editing, and walks you through everything step by step

[03-Mar-2010 11:27:38] <SEJeff_work> rmatte, I just click "New Collector" and away it goes but that is very cool :)

[03-Mar-2010 11:27:58] <rmatte> yeh, I sent my script to egor, he might try to develop something for the UI

[03-Mar-2010 11:28:04] <SEJeff_work> whenever they get a version that works with a newer python and unbundle some of the libs will be exciting

[03-Mar-2010 11:28:08] <rmatte> but for now at least it's something

[03-Mar-2010 11:28:14] <SEJeff_work> So they could do something like get in the fedora repos

[03-Mar-2010 11:28:20] <rmatte> yeh lol

[03-Mar-2010 11:28:29] <SEJeff_work> yum install zenoss

[03-Mar-2010 11:29:11] <rmatte> doubt enterprise will be avilable via repo

[03-Mar-2010 11:29:46] <SEJeff_work> Of course not

[03-Mar-2010 11:29:55] <SEJeff_work> But for getting more users it will be a huge selling point

[03-Mar-2010 11:30:01] <rmatte> for sure

[03-Mar-2010 11:33:39] <malbon> you probably could get it through a redhat channel if they provided it that way.

[03-Mar-2010 11:34:16] <SEJeff_work> one can only hope

[03-Mar-2010 11:34:16] <malbon> RHN supplies quite a bit of our software direct which is outside the distribution.

[03-Mar-2010 11:34:20] <rmatte> Brixius: sorry, just noticed your question...

[03-Mar-2010 11:34:30] <rmatte> Brixius: where are you seeing that error?

[03-Mar-2010 11:35:43] <SEJeff_work> malbon, Nice cloak

[03-Mar-2010 11:35:44] <rmatte> SEJeff_work: the other thing Zenoss has is the community, which can be beneficial to even the enterprise users

[03-Mar-2010 11:35:52] <SEJeff_work> indeed

[03-Mar-2010 11:36:07] <SEJeff_work> Growing the community by making it easier to install benefits all

[03-Mar-2010 11:36:40] <malbon> SEJeff_work: ???

[03-Mar-2010 11:36:41] <rmatte> yeh, you also have access to this chatroom for support if required

[03-Mar-2010 11:36:51] <rmatte> and most of the people helping out in here, such as myself, are core users

[03-Mar-2010 11:37:24] <SEJeff_work> I'm an enterprise user whos done a good bit of scripting / integration work on zenoss. It is always good to see how other people use it and perhaps help them too

[03-Mar-2010 11:37:43] <rmatte> yup

[03-Mar-2010 11:37:49] * malbon is like SEJeff_work too.

[03-Mar-2010 11:37:56] <SEJeff_work> malbon, ~mikea@openwrt/developer/malbon

[03-Mar-2010 11:37:57] <rmatte> we've done tons of integration work with it

[03-Mar-2010 11:38:06] <malbon> SEJeff_work: yeah :)

[03-Mar-2010 11:38:12] <rmatte> we have our multiple zenoss core instances feeding back to a central ticketing system

[03-Mar-2010 11:38:28] <rmatte> it also integrates with our custom cmdb

[03-Mar-2010 11:38:49] <malbon> SEJeff_work: did a lot of porting work in it's early days, looking forward to getting a guruplug so I can do that too.

[03-Mar-2010 11:39:51] <rmatte> well, time for me to get back to porting our customizations from the old version of our ticketing system to the new version

[03-Mar-2010 11:39:59] <rmatte> (since no one else wanted to do it lol)

[03-Mar-2010 11:40:36] <malbon> what service desk tools are you guys using, or more to the point is anyone using Hornbill Supportworks?

[03-Mar-2010 11:41:00] <rmatte> OTRS for ticketing

[03-Mar-2010 11:41:11] <SEJeff_work> rt

[03-Mar-2010 11:41:16] <rmatte> heavily modified

[03-Mar-2010 11:41:17] <malbon> nice and open source :) ok.

[03-Mar-2010 11:41:19] <SEJeff_work> But not for zenoss alerts

[03-Mar-2010 11:41:49] <rmatte> We have it so that the event count, first time, last time, etc... show up in each ticket

[03-Mar-2010 11:41:59] <rmatte> and get upgraded in the existing ticket as they change

[03-Mar-2010 11:42:11] <rmatte> s/upgraded/updated

[03-Mar-2010 11:42:23] <SEJeff_work> zenoss would murder rt in our environment so we don't do that

[03-Mar-2010 11:42:33] <rmatte> hehe

[03-Mar-2010 11:42:53] <malbon> yeah, I have something similar, but it also tracks in the event log when the ticket gets bounced from team to team, acknowledgements in tickets are in parity etc.

[03-Mar-2010 11:42:59] <rmatte> Our OTRS queue looks like: http://dmon.org/otrs.png

[03-Mar-2010 11:43:13] <rmatte> though the new version looks nicer, I'm not done all the mods on it though

[03-Mar-2010 11:43:38] <malbon> heh nice. :)

[03-Mar-2010 11:43:46] <SEJeff_work> If you can pull it off, Jira is one of the nicest things for ticketing

[03-Mar-2010 11:43:49] <SEJeff_work> Not cheap though

[03-Mar-2010 11:43:51] <rmatte> the colours represent the severity of the event

[03-Mar-2010 11:43:59] <rmatte> and all those tickets were generated automatically from Zenoss

[03-Mar-2010 11:44:02] <rmatte> it works well

[03-Mar-2010 11:44:18] <malbon> We have support works already, and I have written quite an extensive interface for it. Am thinking if it is worth packaging up or not.

[03-Mar-2010 11:44:31] <SEJeff_work> rmatte, "Device Reboot" How did you get that?

[03-Mar-2010 11:44:51] <malbon> SEJeff_work: SNMP system restart trap?

[03-Mar-2010 11:45:15] <rmatte> SEJeff_work: lol, I just put evt.summary = 'Device Reboot' as the transform for the snmp_coldStart trap

[03-Mar-2010 11:45:18] <SEJeff_work> malbon, eeeeewwwww. We setup an init script that only runs on shutdown

[03-Mar-2010 11:45:37] <SEJeff_work> It sends an event (xmlrpc) to zenoss with a serverside transform that creates a downtime event

[03-Mar-2010 11:46:02] <rmatte> SEJeff_work: also, keep in mind that some of the ticket descriptions there were modified by our Agents after they came in

[03-Mar-2010 11:46:03] <SEJeff_work> So if a linux admin every types "shutdown -h now" or reboot or something like that, it auto-creates a 10 minute maint window

[03-Mar-2010 11:46:03] <malbon> SEJeff_work: how does that work for routers / Windows boxes / appliances etc?

[03-Mar-2010 11:46:06] <rmatte> but "Device Reboot" is stock

[03-Mar-2010 11:46:27] <SEJeff_work> malbon, The xmlrpc is generic, but the init script is just for 'nix boxes

[03-Mar-2010 11:46:38] <SEJeff_work> Ah cool. Traps are evil. We dont use them

[03-Mar-2010 11:47:03] <malbon> intersting idea, traps are evil it is true, but then tcp snmp would be nice... :)

[03-Mar-2010 11:47:34] <rmatte> SEJeff_work: our position is that if an admin plans to do maintenance on a device, they are to notify us in advance which they do (we're an MSP, so each client is required to submit change requests when doing work that will affect device availability)

[03-Mar-2010 11:48:12] <SEJeff_work> rmatte, Different strokes for different folks :)

[03-Mar-2010 11:48:16] <rmatte> yup

[03-Mar-2010 11:48:18] <malbon> we have a similar policy to rmatte

[03-Mar-2010 11:48:55] <SEJeff_work> malbon, We actually wanted it for a "schedule_downtime" python script so we could shutup monitoring easily before doing maintenance on a server

[03-Mar-2010 11:49:13] <SEJeff_work> kells came up with that init script and we thought it was clever enough to install it on all of our boxes

[03-Mar-2010 11:49:30] <SEJeff_work> blogs/zenossblog/2009/12/23/tip-of-the-month-dynamically-creating-maintenance-windows

[03-Mar-2010 11:49:50] <rmatte> cool

[03-Mar-2010 11:49:56] <rmatte> yeh, it is clever indeed

[03-Mar-2010 11:50:18] <malbon> yeah it is a nice idea, but our Application admins are lazy enough as it is, I don't want to encourage them.

[03-Mar-2010 11:50:45] <SEJeff_work> rmatte, We've extended it quite a bit. It lets you cancel downtime, add downtime, and even do entire locations. Someday we'll publish a flood of stuff on the wiki

[03-Mar-2010 11:51:18] <malbon> funily enough though I have recently used a similar technique for directly injecting information from our datawarehousing appliance.

[03-Mar-2010 11:53:17] <rmatte> SEJeff_work: nice

[03-Mar-2010 11:53:47] <rmatte> Zenoss Inc. are definitely on track to be very successful

[03-Mar-2010 11:54:07] <rmatte> Once they are done with the new UI then they can start focussing on stuff that is horribly lacking, like reporting

[03-Mar-2010 11:54:34] <malbon> Crystal Reports integration would be nice.

[03-Mar-2010 11:54:49] <rmatte> yeh, I've heard that mentioned before

[03-Mar-2010 11:55:04] <rmatte> we ended up coding our own reports via some external python scripts and some php

[03-Mar-2010 11:55:11] <rmatte> they work quite nicely

[03-Mar-2010 11:55:39] <rmatte> we can provide device availability, interface availability, process availability, interface usage, cpu usage, memory usage, disk usage, etc...

[03-Mar-2010 11:56:26] <rmatte> the php runs each script with parameters then compiles everything in to a monthly report which we give to each client each month

[03-Mar-2010 11:57:15] <rmatte> anyways, I really need to get back to work, talk to you guys later

[03-Mar-2010 11:57:18] <SEJeff_work> malbon, It seems to work very well for us

[03-Mar-2010 11:57:43] <SEJeff_work> rmatte, later

[03-Mar-2010 11:59:00] <malbon> yeah I'd better get on with my stuff too.

[03-Mar-2010 12:27:53] <Brixius> rmatte: sorry was out to lunch, I'm seeing that error when I test the WMIPerfMSSQL2000 template, I which is in ZenPacks.community.WMIPerf_MSSQL

[03-Mar-2010 12:28:28] <Brixius> er when testing the WMIPerfMSSQL2000 template data sources.

[03-Mar-2010 12:29:02] forsberg is now known as fOrsberg

[03-Mar-2010 12:29:09] fOrsberg is now known as forsberg

[03-Mar-2010 13:49:54] <Brixius> Must be lunch time, it got quiet in here

[03-Mar-2010 13:54:02] forsberg is now known as fOrsberg

[03-Mar-2010 14:12:59] <rmatte> ok, that's when you're seeing the error, but where are you seeing it?

[03-Mar-2010 14:13:04] <rmatte> in the UI, console, log file?

[03-Mar-2010 14:24:19] <Brixius> I wasn't getting any perf data so I decided to go into the template and test the datasource, and that's where I'm seeing it.

[03-Mar-2010 14:32:27] <Brixius> From the UI

[03-Mar-2010 14:33:25] <Brixius> http://pastebin.ca/1821784

[03-Mar-2010 14:51:59] <mrayzenoss> blogs/zenossblog/2010/03/03/zenoss-qa-test-day-march-4--252-upgrades

[03-Mar-2010 14:54:10] <rmatte> Brixius: not sure about that one, you'd have to ask bigegor if you see him in here

[03-Mar-2010 14:56:19] <Brixius> ok, thanks

[03-Mar-2010 15:30:02] <Brixius> On to another device then, I'll come back to the wmi issues later.  Is there a way to lookup a device interface name from it's snmp index?  I'm getting linkdown/linkup traps and one of the varbinds is the ifindex

[03-Mar-2010 16:05:37] <rmatte> it doesn't actually pass the name of the interface in the trap?

[03-Mar-2010 16:29:05] <Brixius> no

[03-Mar-2010 16:29:10] <Brixius> unfortunatly

[03-Mar-2010 16:32:07] <Brixius> the varbind is "ifIndex: 10" which refers to an interface on os/Interfaces pane(in this case one called hk0.0) and I'd like to parse it out to say that in a transform.

[04-Mar-2010 00:00:30] [disconnected at Thu Mar  4 00:00:30 2010]

[04-Mar-2010 00:00:30] [connected at Thu Mar  4 00:00:30 2010]

[04-Mar-2010 00:00:42] [zenoss-logger (logger bot) has joined #zenoss]

[04-Mar-2010 01:06:19] fOrsberg is now known as forsberg

[04-Mar-2010 04:05:01] <Butch_> morning

[04-Mar-2010 05:27:10] <san4el> Hi everybody!

[04-Mar-2010 05:27:32] <san4el> ïðèâåò âñåì!

[04-Mar-2010 05:36:22] <san4el> hi

[04-Mar-2010 05:46:49] <forsberg> hi

[04-Mar-2010 05:48:17] <san4el> support of russian langauge redused again in zenoss 2.5.2?

[04-Mar-2010 06:01:33] <san4el> is there russian zenoss users?

[04-Mar-2010 06:42:31] <Butch_> hi

[04-Mar-2010 06:42:56] <Butch_> anyone using zenoss behind apache https?

[04-Mar-2010 06:45:42] <aclark> Butch_: I think that is fairly common

[04-Mar-2010 06:48:30] <Butch_> hi aclark

[04-Mar-2010 06:48:45] <Butch_> we just talked about this

[04-Mar-2010 06:48:51] <Butch_> a few weeks ago

[04-Mar-2010 06:49:03] <Butch_> ... still fightin with the same problem :(

[04-Mar-2010 06:49:46] <Butch_> all recipe found doesn't apply

[04-Mar-2010 06:50:32] <chemist> hello

[04-Mar-2010 06:51:02] <Butch_> hi chemist

[04-Mar-2010 06:52:18] <chemist> can anyone tell me how to download the source for a zenpack?

[04-Mar-2010 06:52:28] <chemist> apache monitor in particular

[04-Mar-2010 06:52:47] <chemist> or how to convert an installed zenpack to development

[04-Mar-2010 06:53:58] <chemist> I need to change the address that apache monitor looks at for it's data from /server-stauts?auto to something els

[04-Mar-2010 06:54:01] <chemist> e

[04-Mar-2010 07:08:02] <tehhobbit> anyone know if I can use regular perf templates with indexed snmpvalues ? (have switches with 1 .. 3 cpu's and want to make a generic template for cpu useage )

[04-Mar-2010 07:27:30] <NewBee> can zenoss handel very large Drive arrays? like 11TB?

[04-Mar-2010 07:28:53] <forsberg> sure, which filesystem ?`

[04-Mar-2010 07:29:57] <NewBee> it's a windows server

[04-Mar-2010 07:30:00] <NewBee> so ntfs

[04-Mar-2010 07:30:18] <NewBee> I would like to use snmp.

[04-Mar-2010 07:30:32] <forsberg> yrk

[04-Mar-2010 07:30:33] <forsberg> :=

[04-Mar-2010 07:30:44] <NewBee> it finds my C: and D: drive but not the F: drive.

[04-Mar-2010 07:31:19] <NewBee> if i snmp walk it i see that it's -

[04-Mar-2010 07:31:42] <NewBee> guessing because it's a 32bite problem.

[04-Mar-2010 07:31:45] <NewBee> HOST-RESOURCES-MIB::hrStorageUsed.4 = INTEGER: -1678839791

[04-Mar-2010 07:32:56] <NewBee> any ideas of what's up?

[04-Mar-2010 07:34:11] <NewBee> when i model it i get. -- 2010-03-04 08:33:12INFO         zen.ZenModelerSkipping F:\ Label:MESHFILES  Serial Number 1039d46d. 0 total blocks.

[04-Mar-2010 07:38:10] <tehhobbit> NewBee: its because the 32bit counter used in HOST-RESOURCES-MIB cant handle the values

[04-Mar-2010 07:38:38] <tehhobbit> I use ssh for disk monitoring instead but that might not be an option for you

[04-Mar-2010 07:38:55] <NewBee> ii guess i could use WMI

[04-Mar-2010 07:39:33] <tehhobbit> wmi might have something yes

[04-Mar-2010 07:40:39] <NewBee> any one know if someone has a powershell plugin for zenoss yet?

[04-Mar-2010 07:42:14] <ckrough> that thought made me a little queasy :)

[04-Mar-2010 07:43:35] <NewBee> ya I'm a linux guy. but i have to use some MS systems. and i just started using powershell.

[04-Mar-2010 07:43:37] <NewBee> WOW

[04-Mar-2010 07:43:40] <NewBee> is all i can say

[04-Mar-2010 07:44:06] <NewBee> the best thing since perl... and grep.

[04-Mar-2010 07:49:55] <ckrough> Yeah, just kidding. I haven't played with it much, but from what I've seen it's a great addition

[04-Mar-2010 07:55:03] <tehhobbit> dont really like the oo way of powershell to fond of pipes

[04-Mar-2010 07:55:13] <tehhobbit> but thats just habbit

[04-Mar-2010 07:56:49] <NewBee> ya i am a but pipe guy.

[04-Mar-2010 07:58:30] <forsberg> buttpipe?

[04-Mar-2010 08:00:19] <NewBee> big lol

[04-Mar-2010 08:00:47] <NewBee> althow that is what MS needs

[04-Mar-2010 08:02:34] <NewBee> tehhobbit: is that the client side of snmp or the server side? just thinking if there was a 64bit ver of snmp on the clint that would fix this.

[04-Mar-2010 08:11:34] <SEJeff_work> NewBee, No, even with a 64bit snmp, that doesn't work in my experience

[04-Mar-2010 08:12:02] <SEJeff_work> For our linux hosts with really big (2-12T) volumes, we use ssh modelling for the disk space and snmp for everything else

[04-Mar-2010 08:13:41] <tehhobbit> its the server side

[04-Mar-2010 08:14:24] <tehhobbit> SEJeff_work: the HOST-RESOURCES-MIB only have 32bit counters so doesnt mather if you normaly use v2c/3 '

[04-Mar-2010 08:17:36] <SEJeff_work> <SEJeff_work> NewBee, No, even with a 64bit snmp, that doesn't work in my experience

[04-Mar-2010 08:17:39] <SEJeff_work> Yes, I know that

[04-Mar-2010 08:35:24] <tehhobbit> I hate 3com stuff sometimes, polling cpu usuage from a L3 switch (3com descent edge) shows 19% usage with snmp v1 38% with v3, so question is it the "security" with v3 that eats cpu or is the mib b0rked

[04-Mar-2010 08:39:51] <venturaville> probably the security

[04-Mar-2010 08:39:57] <venturaville> 3com switches don't have very big CPUs......

[04-Mar-2010 08:40:27] <venturaville> and more than likely it is not in ASICs

[04-Mar-2010 08:41:11] <venturaville> of course the easy test is to snmpget it with both at the same time....

[04-Mar-2010 08:41:50] <tehhobbit> crap stuff anyways

[04-Mar-2010 08:46:13] <tehhobbit> venturaville: no its not asic

[04-Mar-2010 08:51:24] forsberg is now known as fOrsberg

[04-Mar-2010 08:53:48] <p4d4w4n> hi champs

[04-Mar-2010 08:54:07] <p4d4w4n> i need send sms the zenoss

[04-Mar-2010 09:02:04] <rmatte> sweet, just got the go-ahead to release my collector ZenPack

[04-Mar-2010 09:02:23] <tehhobbit> what does it do ?

[04-Mar-2010 09:02:40] <rmatte> automated most of the steps to setup remote collectors in core

[04-Mar-2010 09:03:02] <rmatte> basically if you wanted to setup a remote collector, you'd setup 2 servers, install Zenoss on each, install the ZenPack on each

[04-Mar-2010 09:03:20] <rmatte> then you'd type "zencollectortool" on each, and select the appropriate option

[04-Mar-2010 09:03:39] <rmatte> on the master Zenoss server you'd select "Setup master" or whatever, and on the collector you'd select "Setup collector"

[04-Mar-2010 09:03:49] <rmatte> then it walks you through the setup (mostly automated)

[04-Mar-2010 09:04:20] <rmatte> I've tested it and it works well

[04-Mar-2010 09:05:07] <SEJeff_work> rmatte, Thats vaguedly similar to the functionality build in enterprise with the exception it is fully automated from the zenoss server webui in enterprise

[04-Mar-2010 09:05:17] <rmatte> yeh

[04-Mar-2010 09:05:31] <rmatte> It's not quite fully automated, but pretty close

[04-Mar-2010 09:05:45] <rmatte> and it actually tells you what to do step by step and even gives you the lines to paste in to the collector configuration

[04-Mar-2010 09:05:50] <rmatte> so it's just copy and paste

[04-Mar-2010 09:06:17] <rmatte> remote collectors are a pain to setup by hand in Core

[04-Mar-2010 09:06:21] <rmatte> lots of config file editing

[04-Mar-2010 09:06:44] <tehhobbit> oooh rmatte been waiting for something like that

[04-Mar-2010 09:06:58] <tehhobbit> had quite a few issues with distributed collectors

[04-Mar-2010 09:07:19] <tehhobbit> esp. certain zenpacks needs to do more that just copy across

[04-Mar-2010 09:07:40] <rmatte> well, it doesn't help with ZenPacks... those still need to be installed on both by hand

[04-Mar-2010 09:07:49] <chemist> I am getting a 'Error formatting event: location' after 2.5.2 upgrade

[04-Mar-2010 09:07:59] <chemist> that is what is being emailed

[04-Mar-2010 09:08:03] <rmatte> or the ZenPack director can just be copied over

[04-Mar-2010 09:08:05] <tehhobbit> rmatte: do you use ZenJMX remote ?

[04-Mar-2010 09:08:26] <rmatte> tehhobbit: I haven't tried ZenJMX remote, no

[04-Mar-2010 09:09:04] <rmatte> tehhobbit: This release of zencollectortool is obviously a rough release, if additional setup functionality needs to be added I'll add it

[04-Mar-2010 09:09:21] <rmatte> right now it does configuration of all standard core daemons as well as the zenperfwmi daemon

[04-Mar-2010 09:09:24] <tehhobbit> rmatte: something seems to be missing the daemon wont start propperly

[04-Mar-2010 09:09:40] <rmatte> well, did you edit the monitor and hubhost lines in the daemon's config file?

[04-Mar-2010 09:09:48] <rmatte> on the remote collector?

[04-Mar-2010 09:09:56] <tehhobbit> thinking about doing a full install with both zenoss and zenpack rpm's and then just copy back the data

[04-Mar-2010 09:10:11] <tehhobbit> rmatte: will deffinatly try it out

[04-Mar-2010 09:10:19] <tehhobbit> yes

[04-Mar-2010 09:10:24] <rmatte> tehhobbit: If you haven't done that configuration then that's your issue

[04-Mar-2010 09:10:25] <bbibeault> @chemist - can you offer any more info on what you are seeing?

[04-Mar-2010 09:10:56] <tehhobbit> hubhost is set

[04-Mar-2010 09:11:08] <tehhobbit> but hubport is not ....

[04-Mar-2010 09:11:10] <tehhobbit> hmmm

[04-Mar-2010 09:11:19] <chemist> bbibeault: that is what I am seeing in the subject line of the email alerts

[04-Mar-2010 09:11:20] <venturaville> hrm ... the wbemdatasource version on the website is way out of date compare to what is in subversion....

[04-Mar-2010 09:11:57] <bbibeault> Every email alert, or just one alerting rule?

[04-Mar-2010 09:12:08] <chemist> just one

[04-Mar-2010 09:12:59] <rmatte> well, it's obviously not liking the location variable or something in the alerting rule

[04-Mar-2010 09:14:09] <chemist> it's a simple %(location)s

[04-Mar-2010 09:16:30] <tehhobbit> rmatte: next issue, zenrender dies every now and then without any log message

[04-Mar-2010 09:16:31] <venturaville> the wbemdatasource zenpack in subversion has support for proxying which allows me to finally get that IBM/LSI array zenpack working....

[04-Mar-2010 09:17:02] <chemist> so I can't see what's not to like about it

[04-Mar-2010 09:17:08] <rmatte> Weird, for 3 of my devices, when I go to save zProperties I get: BadRequest: The property zWinPassword does not exist

[04-Mar-2010 09:17:16] <rmatte> but it works fine for every other device

[04-Mar-2010 09:17:22] <rmatte> even 2 other devices in the same device class

[04-Mar-2010 09:21:06] <rmatte> ah, figured it out

[04-Mar-2010 09:21:19] <rmatte> had to do...

[04-Mar-2010 09:21:24] <rmatte> >>> a=find("devicename")

[04-Mar-2010 09:21:30] <rmatte> >>> a.__delattr__("zWinPassword")

[04-Mar-2010 09:21:30] <rmatte> >>> commit()

[04-Mar-2010 09:21:30] <rmatte> >>> a.__delattr__("zWinUser")

[04-Mar-2010 09:21:30] <rmatte> >>> commit()

[04-Mar-2010 09:21:30] <rmatte> >>> a.__delattr__("zMaxOIDPerRequest")

[04-Mar-2010 09:21:30] <rmatte> >>> commit()

[04-Mar-2010 09:21:42] <rmatte> just cycled through each attribute until the zProperties page saved without erroring

[04-Mar-2010 09:22:47] <tehhobbit> is there a "console" to zenoss to do that kinda stuff ?

[04-Mar-2010 09:22:54] <jb> zendmd

[04-Mar-2010 09:23:15] <rmatte> yeh, zendmd is incredibly awesome

[04-Mar-2010 09:23:32] <SEJeff_work> what jb and rmatte said

[04-Mar-2010 09:24:23] <tehhobbit> used to run a rails site so quite used to that kind of console

[04-Mar-2010 09:24:30] <tehhobbit> tnx for the tip =D

[04-Mar-2010 09:25:06] <rmatte> np

[04-Mar-2010 09:28:10] <tehhobbit> anyone know if I can use regular perf templates with indexed snmpvalues ? (have switches with 1 .. 3 cpu's and want to make a generic template for cpu useage

[04-Mar-2010 09:33:53] <chemist> can anyone tell me what time it is in EST?

[04-Mar-2010 09:34:24] <jb> 10:34

[04-Mar-2010 09:34:31] <jb> AM

[04-Mar-2010 09:34:37] <chemist> so devs will be here soon

[04-Mar-2010 09:34:50] <jb> in 25 minutes..

[04-Mar-2010 09:39:48] <rmatte> tehhobbit: you need to make sure your zenrender is configured correctly... like, are you seeing graphs when you browse devices that are on the remote collector?

[04-Mar-2010 09:40:32] <tehhobbit> yes I do, I just cant change scale

[04-Mar-2010 09:40:43] <tehhobbit> (if I dont restart it )

[04-Mar-2010 09:41:17] <rmatte> that's really weird, I've never had that issue

[04-Mar-2010 09:41:26] <bbibeault> @Chemist - just has someone in QA test this, the proper syntax is %(Location)s, and that works just fine in our testing.

[04-Mar-2010 09:41:35] <bbibeault> Can you please retry with a capital L and let me know what you get?

[04-Mar-2010 09:42:59] <chemist> one sec

[04-Mar-2010 09:43:06] <rmatte> case sensitivity ;)

[04-Mar-2010 09:43:25] <chemist> ..has been implemented in 2.5.2

[04-Mar-2010 09:43:44] <chemist> that worked thank

[04-Mar-2010 09:43:51] <rmatte> no, it's always been implemented, but it's possible that "location" got changed to "Location"

[04-Mar-2010 09:44:14] <chemist> but my other box with 2.5.1 works with %(location)s

[04-Mar-2010 09:44:14] <rmatte> bbibeault: it is kind of weird for the first letter to be capitalized in that, when looking at the other properties

[04-Mar-2010 09:44:17] <rmatte> device

[04-Mar-2010 09:44:19] <rmatte> component

[04-Mar-2010 09:44:21] <rmatte> summary

[04-Mar-2010 09:44:24] <rmatte> firstTime

[04-Mar-2010 09:44:28] <rmatte> severityString

[04-Mar-2010 09:44:30] <rmatte> Location

[04-Mar-2010 09:44:36] <rmatte> it's a bit of a gray sheep

[04-Mar-2010 09:44:49] <mrayzenoss> perhaps because it's an organizer?

[04-Mar-2010 09:45:10] <rmatte> mrayzenoss: I doubt that

[04-Mar-2010 09:45:33] <rmatte> it's not meant to be an organizer in that context, it's meant to be a device property

[04-Mar-2010 09:45:36] <chemist> I'd just created a trac tickect too

[04-Mar-2010 09:45:47] <chemist> #6251

[04-Mar-2010 09:46:43] <rmatte> chemist: If you can find proof that it was changed in the actual code then you could mark it as a regression

[04-Mar-2010 09:47:03] <rmatte> otherwise just close the trac ticket

[04-Mar-2010 09:47:15] <chemist> well, it works in 2.5.1 and not in 2.5.2

[04-Mar-2010 09:47:25] <chemist> that should be proof enough :)

[04-Mar-2010 09:49:27] * rmatte downloads 2.5.2 to compare

[04-Mar-2010 09:52:15] <rmatte> >>> d = dmd.Devices.findDevice('gen01.novanoc.com')

[04-Mar-2010 09:52:15] <rmatte> >>> d.location

[04-Mar-2010 09:52:15] <rmatte> <ToOneRelationship at /zport/dmd/Devices/Server/Linux/devices/gen01.novanoc.com/location>

[04-Mar-2010 09:52:15] <rmatte> >>> d.Location

[04-Mar-2010 09:52:15] <rmatte> Traceback (most recent call last):

[04-Mar-2010 09:52:16] <rmatte>   File "<console>", line 1, in ?

[04-Mar-2010 09:52:16] <rmatte> AttributeError: Location

[04-Mar-2010 09:52:20] <rmatte> that's in 2.4.5

[04-Mar-2010 09:52:28] <rmatte> going to wait for 2.5.2 to download then try it there

[04-Mar-2010 09:52:42] <jb> ugh i wish 2.5.2 would hurry up

[04-Mar-2010 09:53:00] <rmatte> the 2.5.2 download server is probably getting hit quite hard right now

[04-Mar-2010 09:53:12] <jb> no, its not out yet :)

[04-Mar-2010 09:53:14] <jb> enterprise at least

[04-Mar-2010 09:53:19] <ckrough> jb: it is

[04-Mar-2010 09:53:26] <jb> since when?

[04-Mar-2010 09:53:30] <ckrough> jb: chet said its in the download section of as yesterday

[04-Mar-2010 09:53:34] <jb> oh wtf

[04-Mar-2010 09:53:38] <ckrough> jb: ha

[04-Mar-2010 09:53:39] <rmatte> lol

[04-Mar-2010 09:53:39] <jb> checking now!

[04-Mar-2010 09:53:49] <jb> indeed it is

[04-Mar-2010 09:53:50] <ckrough> jb: we asked about it in dev chat, I got the impression he put it there during the chat

[04-Mar-2010 09:53:52] <jb> installing it right now

[04-Mar-2010 09:53:58] <jb> i -really- hope this fixes the event console

[04-Mar-2010 09:54:03] <ckrough> jb: beta test that for me and let me know what breaks :)

[04-Mar-2010 09:54:06] <rmatte> jb: it does

[04-Mar-2010 09:54:18] <rmatte> jb: I've tested the event console and there have been tons of fixes in 2.5.2

[04-Mar-2010 09:54:34] <SEJeff_work> jb, Are you referring to where the even console will say "Server Disconnect" or something along that?

[04-Mar-2010 09:54:41] <jb> SEJeff_work: yes, exactly

[04-Mar-2010 09:54:46] <SEJeff_work> Yup 2.5.2 fixes that for me

[04-Mar-2010 09:54:52] <rmatte> jb: yeh, that's fixed

[04-Mar-2010 09:54:56] <SEJeff_work> But you can just zenpatch to fix it I think

[04-Mar-2010 09:55:18] <mrayzenoss> I believe there are around 30 tickets tied to the event console for 2.5.2

[04-Mar-2010 09:55:45] <rmatte> yup

[04-Mar-2010 09:56:36] * rmatte watches as the 2.5.2 download chugs along

[04-Mar-2010 09:57:40] <jb> slow?

[04-Mar-2010 09:57:42] <jb> mine was fast :)

[04-Mar-2010 09:57:58] <rmatte> 100K/s which is quite slow considering it usually comes down at over 1M/s

[04-Mar-2010 09:58:14] * Simon4 files 2.5.2 for "after holiday"

[04-Mar-2010 09:58:31] <rmatte> Simon4: yeh, if it's not broken don't fix it before a holiday

[04-Mar-2010 09:58:36] <rmatte> ;)

[04-Mar-2010 09:58:41] <Simon4> it's broken, but managed ;)

[04-Mar-2010 09:58:45] <rmatte> hehe

[04-Mar-2010 09:58:54] * Simon4 is hoping for some event console fixes also, and some db integrity ones

[04-Mar-2010 09:59:02] <rmatte> mine is awesomely patched

[04-Mar-2010 09:59:06] <rmatte> ;)

[04-Mar-2010 10:01:08] <aclark> Anyone got a "how do I run trunk" doc handy (if there is such a thing)?

[04-Mar-2010 10:02:10] <emillerATzenoss> hello

[04-Mar-2010 10:04:41] <rmatte> mrayzenoss: [10:57am] <aclark> Anyone got a "how do I run trunk" doc handy (if there is such a thing)?

[04-Mar-2010 10:04:50] <rmatte> know of anything like that?

[04-Mar-2010 10:05:04] <mrayzenoss> The devs are doing their morning standup, emillerATzenoss is today's Dev on call.  He's relatively new, so be gentle :)

[04-Mar-2010 10:05:05] <rmatte> emillerATzenoss: hello

[04-Mar-2010 10:05:49] <mrayzenoss> aclark: trunk is not for the timid :)

[04-Mar-2010 10:05:53] <emillerATzenoss> i don't know of any how to run trunk doc

[04-Mar-2010 10:06:14] <mrayzenoss> we should have alpha installers out soon, but you can build from SVN trunk

[04-Mar-2010 10:06:20] <aclark> mrayzenoss: understood

[04-Mar-2010 10:06:28] <mrayzenoss> the plan is to switch to Python 2.6/Zope 2.12 for trunk this weekend

[04-Mar-2010 10:06:46] <mrayzenoss> and next QA Day is tentatively supposed to be source builds off of trunk

[04-Mar-2010 10:06:53] <mrayzenoss> getting people to help us kick the tires

[04-Mar-2010 10:07:04] <mrayzenoss> find build issues on other platforms and whatnot.

[04-Mar-2010 10:07:16] * mrayzenoss looks at the FreeBSD and OpenSolaris people in particular :p

[04-Mar-2010 10:08:05] <jb> hey rocket

[04-Mar-2010 10:08:48] <rocket> hey jb

[04-Mar-2010 10:11:12] <rmatte> rocket: wassap?

[04-Mar-2010 10:11:23] <rocket> heh lots .. :p

[04-Mar-2010 10:11:49] * rmatte thought you'd say "the ceiling", but "lots" works too.

[04-Mar-2010 10:13:14] <rocket> heh .. :)

[04-Mar-2010 10:17:40] <rmatte> bbibeault: the attribute is still "location" in 2.5.2, not sure why only "Location" suddenly works for alerts

[04-Mar-2010 10:18:13] <chemist> I've asked this before and I should probably trac it, but how difficult would it be to display an snmp value in the 'status' tab of a device?

[04-Mar-2010 10:18:42] <rmatte> chemist: I've never seen it done, so I'm going to go with 'very'

[04-Mar-2010 10:19:01] <bbibeault> Thanks guys, please make sure that the ticket includes this info, and we can discuss it in today's defect review.  That is a very odd thing to crop up...

[04-Mar-2010 10:19:03] <rmatte> though it would make for a nice feature in the new UI

[04-Mar-2010 10:19:05] <aclark> mrayzenoss: What I'm getting at is for Plone we use mr.developer (http://pypi.python.org/pypi/mr.developer/1.10) for core dev, and I'm interested in setting up a similar environment for Zenoss just so I can poke at it

[04-Mar-2010 10:19:28] <chemist> rmatte: it would, and I can think of many uses for it

[04-Mar-2010 10:19:37] <chemist> so over to the devs

[04-Mar-2010 10:19:46] <rmatte> chemist: open a feature request in trac

[04-Mar-2010 10:19:57] <aclark> I suppose I can just grab http://dev.zenoss.org/svn/trunk/Products for starters

[04-Mar-2010 10:19:59] <chemist> will do

[04-Mar-2010 10:21:07] <rmatte> anyways, time for me to get back to perl and html :P

[04-Mar-2010 10:24:47] <mrayzenoss> aclark: I let iancmcc know what you were talking about

[04-Mar-2010 10:25:31] <aclark> iancmcc: :-)

[04-Mar-2010 10:25:35] <aclark> mrayzenoss: thx

[04-Mar-2010 10:27:54] <jb> finally.. a working event console.

[04-Mar-2010 10:28:27] <bbibeault> Well that's a good endoresement.  :)

[04-Mar-2010 10:28:33] <jb> :)

[04-Mar-2010 10:29:31] <aclark> iancmcc: Do you guys give out repo access or is the community code just filtered through patches? (i.e. as described here: community/developers)

[04-Mar-2010 10:30:22] <jb> hrm you know

[04-Mar-2010 10:30:27] <jb> for the AIX SSH zenpack

[04-Mar-2010 10:30:34] <jb> the filesystem alerts are coming in as "blocks"

[04-Mar-2010 10:30:37] <jb> not MB/GB

[04-Mar-2010 10:30:49] <jb> i thought this was fixed in like 2.4.x, but it appears broken again in 2.5.x

[04-Mar-2010 10:31:19] <emillerATzenoss> i will take a look

[04-Mar-2010 10:31:27] <jb> thanks

[04-Mar-2010 10:31:30] <mrayzenoss> aclark: currently we're working off patches, but we keep talking about switching to Git

[04-Mar-2010 10:31:38] <mrayzenoss> so people can maintain their own branches

[04-Mar-2010 10:32:03] <aclark> iancmcc_: hankering to fiddle with a zenoss buildout again and would like to use mr.developer to build trunk, so i guess i'll start with http://dev.zenoss.org/svn/trunk/Products and see how far I get

[04-Mar-2010 10:32:10] <jb> emillerATzenoss: example -  threshold of high disk usage exceeded: current value 360356.00

[04-Mar-2010 10:32:15] <jb> not very helpful :)

[04-Mar-2010 10:32:34] <emillerATzenoss> no

[04-Mar-2010 10:32:50] <aclark> mrayzenoss: Yeah, have you considered other DVCS e.g. hg and bzr? (I hear those are more "Pythonic" choices, but I don't know much about DVCSes)

[04-Mar-2010 10:34:25] <mrayzenoss> aclark: yeah, there have been numerous discussions about Hg vs. Git.  Not sure why Git was eventually chosen, but I know several of the devs are already using the Git/SVN bridge

[04-Mar-2010 10:34:36] <mrayzenoss> including Ian, who seems to be out of the channel again

[04-Mar-2010 10:34:40] <aclark> nice

[04-Mar-2010 10:34:42] <aclark> heh

[04-Mar-2010 10:34:56] <iancmcc> aclark: Sorry, got a call. So mr.developer is just for eggy things, right?

[04-Mar-2010 10:35:04] <zenethian> I want me some bazaar.  I would settle with hg.

[04-Mar-2010 10:35:06] <iancmcc> It won't help with all the cmmi stuff?

[04-Mar-2010 10:35:21] <iancmcc> zenethian: we went through them all, and git was the best for our purposes.

[04-Mar-2010 10:35:42] <aclark> iancmcc: correct, but the cmmi stuff is more or less easily handled by cmmi recipes

[04-Mar-2010 10:35:56] <zenethian> Yah, I figured, I just sorta <3 bazaar for some fanatic zealotous reason.

[04-Mar-2010 10:36:08] <zenethian> and git is quite a bit better than svn by all means.

[04-Mar-2010 10:37:10] <iancmcc> aclark: So what's the benefit at this point in time? Sorry if you said this, I lost connection there for a minute

[04-Mar-2010 10:37:28] <zenethian> 10:30 < aclark> iancmcc_: hankering to fiddle with a zenoss buildout again and would like to use mr.developer to build trunk, so i guess i'll start with http://dev.zenoss.org/svn/trunk/Products and see how far I get

[04-Mar-2010 10:39:44] <aclark> iancmcc: For eggs, mr.developer makes it Hella-easy to develop a complex stack, e.g. http://svn.plone.org/svn/plone/buildouts/plone-coredev/branches/4.0/sources.cfg

[04-Mar-2010 10:40:11] <iancmcc> That's great, because zope 2.12 really threw us for a loop

[04-Mar-2010 10:40:21] <aclark> right

[04-Mar-2010 10:41:10] <aclark> iancmcc: I'm excited about Zenoss under Python 2.6/Zope 2.12

[04-Mar-2010 10:42:03] <iancmcc> It's all functional. We're doing the merge i think next week although I promise nothing.

[04-Mar-2010 10:42:41] <malbon> hi folks,

[04-Mar-2010 10:42:53] <iancmcc> Python and Zope each closed a couple loopholes we were relying on, which made it...fun.

[04-Mar-2010 10:43:09] <malbon> I can see from python-snmp that TimeTicks are declared as Unsigned, is that on every use or do I need to translate them?

[04-Mar-2010 10:43:42] <aclark> iancmcc: can you describe how to run what your developing against? or is it too much trouble… In other words, will Zope 2.12 and the right branches of all the products get me anything?

[04-Mar-2010 10:44:48] <iancmcc> aclark: I haven't tried dumping Products into Zope 2.12, but it /ought/ to, other things being equal

[04-Mar-2010 10:44:55] <iancmcc> the tricky bit is all the setup

[04-Mar-2010 10:45:22] <iancmcc> If you drop your existing Data.fs in it should work like gangbusters

[04-Mar-2010 10:45:54] <iancmcc> But all the db creation is hand-rolled right now and might rely on some script locations or something

[04-Mar-2010 10:46:35] <aclark> iancmcc: You know, that was my stumbling block last time. So maybe that's not a bad way to go (I could just do a "regular" install and use that Data.fs to test with)

[04-Mar-2010 10:46:54] <iancmcc> Yeah, I doubt you'll have problems there

[04-Mar-2010 10:46:59] <aclark> iancmcc: I'm kind of interested in buildout recipe-i-fying all that setup

[04-Mar-2010 10:47:06] <iancmcc> aclark: lord, me too

[04-Mar-2010 10:47:20] <mrayzenoss> everyone wants an easier build

[04-Mar-2010 10:47:54] <aclark> iancmcc, mrayzenoss OK I'll start in April, send over the contracts ;-)

[04-Mar-2010 10:47:56] * aclark ducks

[04-Mar-2010 10:47:58] <rocket> i dont :p

[04-Mar-2010 10:48:04] <aclark> hah

[04-Mar-2010 10:48:37] <mrayzenoss> aclark: all I can promise you is swag :p

[04-Mar-2010 10:48:55] <aclark> Anyway, I think I can fiddle again, thanks guys

[04-Mar-2010 10:49:21] <aclark> mrayzenoss: Heh yeah I've all but given up trying to get hired

[04-Mar-2010 10:50:46] * aclark will just hang out Zenoss consulting shingle at some point, hopefully in 2010

[04-Mar-2010 10:52:30] <mrayzenoss> aclark: did you see this? thread/12893?tstart=0

[04-Mar-2010 10:52:50] <mrayzenoss> I figure #2 and #3 are a bit Zope-oriented

[04-Mar-2010 10:53:38] <aclark> mrayzenoss: w00t! thanks will take a look

[04-Mar-2010 10:55:31] <mrayzenoss> and our Support Manager is always hunting for more L1/L2 guys

[04-Mar-2010 10:56:10] <mrayzenoss> speaking of new L2 guys...

[04-Mar-2010 10:56:49] <aclark> what do they do?

[04-Mar-2010 10:57:14] <Brixius> I'm having an issue with ZenPacks.community.WMIDataSource-2.0.98-py2.4.egg when I try to use it, (Actually I guess it's ZenPacks.community.WMIPerf_MSSQL that is trying to use it) I get the following error

[04-Mar-2010 10:57:17] <Brixius> File "/usr/local/zenoss/zenoss/ZenPacks/ZenPacks.community.WMIDataSource-2.0.98-py2.4.egg/ZenPacks/community/WMIDataSource/WMIClient.py", line 265, in WmiGet

[04-Mar-2010 10:57:17] <Brixius> url[2], device.zWmiProxy = url[2].split('@')

[04-Mar-2010 10:57:17] <Brixius> ValueError: need more than 1 value to unpack

[04-Mar-2010 10:57:55] <zenethian> level 1 and level 2 support.  You know, the kinda job that pays you big bucks for tearing your hair out.

[04-Mar-2010 10:58:00] <zenethian> ;-)

[04-Mar-2010 10:59:07] <mrayzenoss> L1 is "Client Support Engineer", L2 is "Senior Support Engineer" http://www.zenoss.com/about/jobs#4

[04-Mar-2010 10:59:29] <mrayzenoss> a little more flexible about working remote

[04-Mar-2010 10:59:40] <zenethian> So when I'm on Level 3 support, I'm an Senior Executive Support Engineer Mk V

[04-Mar-2010 11:00:31] <zenethian> hmm, I wish I could work remote more often.

[04-Mar-2010 11:00:58] <malbon> it doesn't say how much which is disappointing.

[04-Mar-2010 11:01:01] <cparlette> everybody wants to be an L1, they're the coolest

[04-Mar-2010 11:01:01] <mrayzenoss> well, we have a new L2 guy starting from Minnesota real soon

[04-Mar-2010 11:01:27] <mrayzenoss> so says the L1 who closes the most tickets

[04-Mar-2010 11:01:28] <zenethian> "a lot"

[04-Mar-2010 11:02:03] <zenethian> I hear those L1 guys all live in a bunker in the woods in South Dakota.

[04-Mar-2010 11:02:53] <malbon> I'd like to be L3 at least, if not higher, but you need a lot of gold to get there.

[04-Mar-2010 11:03:35] <zenethian> Right now I feel like I'm on L4 support.  because every time the L3 guy gets something he looks across the room and goes "hey Jon-Pierre..."

[04-Mar-2010 11:03:44] <mrayzenoss> heh

[04-Mar-2010 11:04:19] <emillerATzenoss> :P

[04-Mar-2010 11:05:35] <zenethian> I guess I had enough XP to level up.

[04-Mar-2010 11:05:35] <mrayzenoss> so anyone doing 2.5.2 upgrades?  #zenoss-testing is ready to help you out

[04-Mar-2010 11:06:00] <mrayzenoss> or cheat codes

[04-Mar-2010 11:06:33] * zenethian dances o/  _o/ _o_ o/ o o/ o /o/

[04-Mar-2010 11:07:30] <zenethian> Up Up Down Down Left Right Left Right B A Start!

[04-Mar-2010 11:07:43] <zenethian> I am so totally programming that into an easter egg in the new UI.

[04-Mar-2010 11:07:51] <ckrough> discussion on Fedora Infrastructure list about not being happy with Zabbix and nagios

[04-Mar-2010 11:08:06] <zenethian> Oh really

[04-Mar-2010 11:08:42] <chemist> I upgraded one instance already, apart from the %(location)s to %(Location)s problem its been pretty smooth

[04-Mar-2010 11:09:54] <aclark> mrayzenoss: are there debs out? docs/DOC-3240?noregister all link to the same stack installer…

[04-Mar-2010 11:10:55] <ckrough> zenethian: thread http://lists.fedoraproject.org/pipermail/infrastructure/2010-March/008562.html

[04-Mar-2010 11:11:33] <frozty_sa> mrayzenoss: I want to upgrade tonight, but the lack of easy debian package rollback makes me hesitate :)

[04-Mar-2010 11:11:36] <frozty_sa> and I don't have a test env to upgrade first :/

[04-Mar-2010 11:14:20] <mrayzenoss> yeah, I think there's a bug with the .deb, we may have to respin it

[04-Mar-2010 11:14:40] <mrayzenoss> this: message/46248

[04-Mar-2010 11:14:46] <mrayzenoss> only seems to happen with the .deb installer

[04-Mar-2010 11:16:40] <mrayzenoss> frozty_sa: we have some good contacts within Fedora, we should start working on Zenoss packaging once trunk has switched over to Python 2.6

[04-Mar-2010 11:17:54] <frozty_sa> mrayzenoss: there ain't no way in melting snowhell I'm ever really running fedora again unless some major changes happen

[04-Mar-2010 11:19:19] <mrayzenoss> frozty_sa: Fedora is a great place to get your packages hammered on.  They do good work slapping things into shape.  I wouldn't recommend them for production environments, but Fedora is the gateway to proper RHEL support

[04-Mar-2010 11:19:21] <frozty_sa> (in fedora, not zenoss)

[04-Mar-2010 11:19:38] <frozty_sa> mrayzenoss: oh, I know, I worked with fedora for a loooong time :)

[04-Mar-2010 11:21:29] <frozty_sa> mrayzenoss: mind if I pm?

[04-Mar-2010 11:21:42] <mrayzenoss> feel free

[04-Mar-2010 11:24:52] <nickanderson> Is the cycle time on a data source how often it will collect that data?

[04-Mar-2010 11:31:35] <rocket> heh sorry I was in a meeting .. but I can let everyone know I am a new level 2 support guy now .. :p

[04-Mar-2010 11:31:57] <Simon4> congrats :)

[04-Mar-2010 11:32:42] <nickanderson> congrats

[04-Mar-2010 11:33:09] <rocket> Thanks.... I plan on promoting zenoss .. err nagi... err matt which one was it again? :p

[04-Mar-2010 11:34:06] * mrayzenoss pulls out clue stick

[04-Mar-2010 11:35:13] <rocket> Zenoss! ftw! :)

[04-Mar-2010 11:35:21] <rocket> thanks Matt I needed that .. :P

[04-Mar-2010 11:36:54] <mrayzenoss> ok, I'm letting emillerATzenoss off the hook, I'm finally heading into the office

[04-Mar-2010 11:37:01] <theacolyte> oh boy

[04-Mar-2010 11:37:03] <theacolyte> 2.5.2

[04-Mar-2010 11:37:05] <theacolyte> GO!

[04-Mar-2010 11:37:55] <rocket> cparlette: L1 is pretty cool I will admit .. :p

[04-Mar-2010 11:38:14] <rocket> considering L1 cache is the fastest as well .. :p

[04-Mar-2010 11:39:15] <mrayzenoss> back after lunch

[04-Mar-2010 11:40:21] <Brixius> anyone here familear with ZenPacks.community.WMIDataSource-2.0.98-py2.4.egg

[04-Mar-2010 11:41:18] <theacolyte> Yes

[04-Mar-2010 11:41:19] <theacolyte> I use it

[04-Mar-2010 11:43:11] <Brixius> I'm having an issue with it where no mater what I do I get an error, I have a feeling it's something stupid, but I can't track it down.

[04-Mar-2010 11:44:44] <Brixius> I'm trying to move something that I did in a nagios style check that works using wmic to use WMIDataSource, but I keep geting the following.

[04-Mar-2010 11:44:52] <Brixius> File "/usr/local/zenoss/zenoss/ZenPacks/ZenPacks.community.WMIDataSource-2.0.98-py2.4.egg/ZenPacks/community/WMIDataSource/WMIClient.py", line 323, in ?

[04-Mar-2010 11:44:52] <Brixius> results = WmiGet(url, query, properties)

[04-Mar-2010 11:44:52] <Brixius> File "/usr/local/zenoss/zenoss/ZenPacks/ZenPacks.community.WMIDataSource-2.0.98-py2.4.egg/ZenPacks/community/WMIDataSource/WMIClient.py", line 265, in WmiGet

[04-Mar-2010 11:44:52] <Brixius> url[2], device.zWmiProxy = url[2].split('@')

[04-Mar-2010 11:44:52] <Brixius> ValueError: need more than 1 value to unpack

[04-Mar-2010 11:45:33] <zenethian> that means the URL doesn't have an @ in it

[04-Mar-2010 11:46:44] <theacolyte> Ok, taking a look

[04-Mar-2010 11:46:49] <rocket> zenethian: I agree .. :P

[04-Mar-2010 11:48:06] <zenethian> I'm not sure why the URL would have an @ in it, but hey.  I didn't write that.

[04-Mar-2010 11:48:18] <Brixius> I'd agree to that, it makes sense, but I'm not sure where I would put an @ in the ui so it has one, and not sure how to call this from zendmd either(I'm no python expert although I'm learning more everyday)

[04-Mar-2010 11:48:33] <rocket> input validation would be nice in that case me thinks .. :p

[04-Mar-2010 11:48:59] <theacolyte> That's a bit beyond what I can help you with, though egor is usually in this channel

[04-Mar-2010 11:49:08] <theacolyte> by the way, I'm biting the bullet - and installing 2.5.2

[04-Mar-2010 11:49:10] <zenethian> Well an @ in URLs specifies a username.  so http://rocket@inyourpocket.com/index.html//rocket@inyourpocket.com/index.html I think is what it's looking for.

[04-Mar-2010 11:49:11] <theacolyte> I'll let you know how it goes

[04-Mar-2010 11:49:47] <rocket> zenethian: heh and : usually indicate a password as well :p

[04-Mar-2010 11:49:57] <zenethian> Hopefully.

[04-Mar-2010 11:50:09] <rocket> if its in the right place that is

[04-Mar-2010 11:50:28] <zenethian> Well an @ in URLs specifies a username.  so http://rocket:abc123@inyourpocket.com/index.htmlabc123@inyourpocket.com/index.html I think is what it's looking for.

[04-Mar-2010 11:50:32] <zenethian> there

[04-Mar-2010 11:50:43] <zenethian> zomg I know rocket's password.

[04-Mar-2010 11:50:58] <zenethian> Now if only knew who rocket is.  :o

[04-Mar-2010 11:51:18] <rocket> someday we may meet :p

[04-Mar-2010 11:51:30] <zenethian> There can be only one.

[04-Mar-2010 11:51:37] <Brixius> ya, I'm trying to figure out where the bit of code is that's building that url to send to WMIClient.py

[04-Mar-2010 11:51:45] <theacolyte> Great, now I've got the highlander theme in my head

[04-Mar-2010 11:51:56] <zenethian> Mission complete!

[04-Mar-2010 11:52:17] <rocket> zenethian: I am sure they will do a formal picture thingy etc

[04-Mar-2010 11:52:21] <theacolyte> I hatey oU!

[04-Mar-2010 11:52:31] <zenethian> hey leave oU alone.

[04-Mar-2010 11:52:36] <rocket> theacolyte: which version of the theme? lol

[04-Mar-2010 11:52:44] <theacolyte> the TV show

[04-Mar-2010 11:52:45] <theacolyte> hehe

[04-Mar-2010 11:52:47] <zenethian> See the one that came to mind was the TV show.

[04-Mar-2010 11:52:52] <zenethian> it was so much better than the movie

[04-Mar-2010 11:52:56] <theacolyte> Not the queen version

[04-Mar-2010 11:53:01] <theacolyte> which I'd agree is much better

[04-Mar-2010 11:53:24] <theacolyte> actually wait

[04-Mar-2010 11:53:27] <theacolyte> they're both queen

[04-Mar-2010 11:54:15] <theacolyte> I think I need a dose of Christopher LAmbert

[04-Mar-2010 11:54:48] <SEJeff_work> needs more cowbell

[04-Mar-2010 11:56:02] <theacolyte> I have a feva

[04-Mar-2010 11:56:43] <theacolyte> I think I need to watch that clip again

[04-Mar-2010 11:56:46] <theacolyte> classic SNL

[04-Mar-2010 11:57:02] <zenethian> I'm sorry but Adrian Paul was so much better than Christopher Lambert as a highlander.

[04-Mar-2010 11:58:04] <theacolyte> I actually completel yagree with you, but lambert has a better accent

[04-Mar-2010 11:58:47] <zenethian> Oh I agree that in general, Christopher Lambert is awesome and indeed has a fantastic accent.  he was the only reason that Mortal Kombat was palletable.

[04-Mar-2010 12:06:24] <theacolyte> Well, I upgraded just fine

[04-Mar-2010 12:06:52] <theacolyte> I did get some heartbeat alerts right after it came up, but that could be because of the upgrade

[04-Mar-2010 12:27:24] <theacolyte> I spoke too soon, it hosed zenperfwmi

[04-Mar-2010 12:28:05] <theacolyte> localhost zenperfwmi heartbeat failure

[04-Mar-2010 12:28:19] <theacolyte> s/hosed/didn't start/g

[04-Mar-2010 12:28:36] <theacolyte> hmm, actually, I can't start it

[04-Mar-2010 14:03:48] fOrsberg is now known as forsberg

[04-Mar-2010 14:05:36] <forsberg> ello, anything to be aware of with 2.5.2 ? only have apache and mysql community zenpacks installed :P

[04-Mar-2010 14:08:29] mrayzenoss1 is now known as mrayzenoss

[04-Mar-2010 14:32:30] <rmatte> forsberg: If you haven't noticed any issues with it, then no

[04-Mar-2010 14:32:46] <rmatte> forsberg: there have been ZenPack related issues reported but no one has been able to reproduce them

[04-Mar-2010 14:34:36] <mrayzenoss> there's a path issue with the .deb

[04-Mar-2010 14:35:18] <venturaville> does anyone here have a doc explaining the syntax for the WBEMDataSource Instance field?

[04-Mar-2010 14:35:27] <venturaville> trying to mesh it up with what I am getting out of YAWN

[04-Mar-2010 14:35:43] <mrayzenoss> venturaville: keep an eye out for bigegor if/when he logs in

[04-Mar-2010 14:35:50] <venturaville> I have been :-)

[04-Mar-2010 14:36:40] <venturaville> all I really need is an example in both wbemcli and in WBEMDataSource and I can figure it out from there

[04-Mar-2010 14:38:47] <RoAkSoAx> mrayzenoss, will you be around tomorrow?

[04-Mar-2010 14:38:54] <Brixius> I wish bigegor would login, I have a question for him too...

[04-Mar-2010 14:43:07] <mrayzenoss> I'm always around

[04-Mar-2010 14:43:23] <RoAkSoAx> mrayzenoss, cool i'll bother you tomorrow with zenoss questions :)

[04-Mar-2010 14:43:32] <mrayzenoss> 'cept weekends.  That's family time (and catching up on ZenPack backlog)

[04-Mar-2010 14:44:18] <RoAkSoAx> cool, cause im gonna take this weekend of to, but I will have the class presentation this upcoming tuesday

[04-Mar-2010 14:44:53] <RoAkSoAx> mrayzenoss, i've already been playing with it and have basic config monitoring through snmp and ssh however i still have questions and Im pretty sure I will have more tomorrow

[04-Mar-2010 14:46:24] <forsberg> hehe ok thanks rmatte, everything looks fine :P

[04-Mar-2010 14:46:43] <forsberg> i couldnt wait for an answer :)

[04-Mar-2010 14:53:38] <nickanderson> Is there a way to have the zenpacks create the context paths?

[04-Mar-2010 14:54:39] <mrayzenoss> you mean use a relative path to a script within the ZenPack?  Yes

[04-Mar-2010 14:55:18] <nickanderson> I mean the event paths

[04-Mar-2010 15:13:48] mrayzenoss1 is now known as mrayzenoss

[04-Mar-2010 15:28:57] <mrayzenoss> Guess everyone's 2.5.2 upgrades went fine, quiet day in #zenoss-testing

[04-Mar-2010 15:29:55] <theacolyte> mrayzenoss: Mine went well-ish

[04-Mar-2010 15:30:19] <mrayzenoss> well-ish!  A ringing endorsement!  :p

[04-Mar-2010 15:31:21] <theacolyte> haha

[04-Mar-2010 15:31:30] <theacolyte> I had/have an issue with the wmi zenpack

[04-Mar-2010 15:31:45] <theacolyte> After upgrading the daemon didn't start automatically, and the start button doesn't wokr in the interface

[04-Mar-2010 15:31:51] <theacolyte> I actually had to start it from the command line

[04-Mar-2010 15:32:02] <theacolyte> works fine now that I've done that, I just need to r emember to do it every time I reboot

[04-Mar-2010 15:32:36] <theacolyte> But everything else is perfectly fine

[04-Mar-2010 15:34:54] <mrayzenoss> the WMI daemon from the ZenPack?

[04-Mar-2010 15:35:01] <mrayzenoss> hmm… wonder what might have changed

[04-Mar-2010 15:35:59] <theacolyte> yeah

[04-Mar-2010 15:36:00] <theacolyte> not sure

[04-Mar-2010 15:36:08] <theacolyte> I did a stack upgrade on my original stack install

[04-Mar-2010 15:36:19] <theacolyte> everything worked fine, but started getting heartbeat failure emails for zenperfwmi

[04-Mar-2010 15:36:32] <theacolyte> Took a look in the web itnerface, saw it was stopped, clicked the start button - nothint happened

[04-Mar-2010 15:36:34] <forsberg> wmi = :(

[04-Mar-2010 15:36:50] <theacolyte> Logged into the box, started it from the command line... ./zenperfwmi start

[04-Mar-2010 15:36:55] <theacolyte> then it showed fine in daemons

[04-Mar-2010 15:37:13] <theacolyte> forsberg: WMI is perfectly fine... the zenoss core implementation of it is getting better thanks to egor

[04-Mar-2010 15:37:27] <theacolyte> I actually like wmi a whoe lot more than snmp

[04-Mar-2010 15:39:49] <Brixius> theacolyte: I just wish I could make it work...

[04-Mar-2010 15:40:13] <theacolyte> hehe, I feel that way even when it is working

[04-Mar-2010 15:40:17] <theacolyte> What's wrong with it?

[04-Mar-2010 15:40:40] <Brixius> the can't find an @ in the url error

[04-Mar-2010 15:41:05] <Brixius> I think you looked at it and said I'd have to ask bigegor about it

[04-Mar-2010 15:48:05] <Brixius> I wonder if my problem is I installed WBEMDataSource from svn, where can I download the latest stable version, I had a link to community zenpacks but can't seem to find it now..

[04-Mar-2010 15:50:03] <theacolyte> ah yeah I have not a clue

[04-Mar-2010 15:52:51] <Brixius> What version of wbemdatasource do you have installed theacolyte?

[04-Mar-2010 15:52:59] <Brixius> and wmidatasource?

[04-Mar-2010 15:53:31] <theacolyte> Whatever the latest from the community site is

[04-Mar-2010 15:53:33] <theacolyte> (not SVN)

[04-Mar-2010 15:53:53] <Brixius> ok, thanks, I'll try installing that and see if it's better. :)

[04-Mar-2010 16:26:18] forsberg is now known as fOrsberg

[04-Mar-2010 16:32:51] <Brixius> well I think I did something bad by loading the SVN Version of the WMIDataSource.  After removing that one(version 2.0.98) and installing the one from the Community web site, I get errors anytime I try to use the WMI datasource.  Perhaps I should go back and reinstall with a clean install, since I was in the process of setting up a replacement for a really realy old zenoss(2.1.x version) and hadn't gotten real far into setting it up yet anyw

[04-Mar-2010 16:33:28] <theacolyte> what error?

[04-Mar-2010 16:34:25] <Brixius> Like a bug report screen.  I think I'm going to do a stack restart and see if that takes care of it.

[04-Mar-2010 16:35:47] fOrsberg is now known as forsberg

[04-Mar-2010 16:38:03] <Brixius> cool, zenoss-stack restart fixed it.

[04-Mar-2010 16:38:10] <theacolyte> Excellent

[04-Mar-2010 16:38:19] <Brixius> and wmi is working now...

[04-Mar-2010 16:38:20] <theacolyte> They say you should always do that after a zenpack install anwyay

[04-Mar-2010 16:38:26] <Brixius> ok

[04-Mar-2010 17:30:34] <OfficeCase> Looking for help with MIBS can anyone be of assistance?

[04-Mar-2010 17:36:06] <forsberg> irss

[04-Mar-2010 17:36:13] forsberg is now known as fOrsberg

[04-Mar-2010 17:44:18] <OfficeCase> opp

[04-Mar-2010 17:56:14] RoAk is now known as RoAkSoAx

[05-Mar-2010 00:00:30] [disconnected at Fri Mar  5 00:00:30 2010]

[05-Mar-2010 00:00:30] [connected at Fri Mar  5 00:00:30 2010]

[05-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[05-Mar-2010 02:21:48] fOrsberg is now known as forsberg

[05-Mar-2010 02:37:06] forsberg is now known as fOrsberg

[05-Mar-2010 02:37:35] fOrsberg is now known as forsberg

[05-Mar-2010 02:48:40] forsberg is now known as fOrsberg

[05-Mar-2010 02:52:11] fOrsberg is now known as forsberg

[05-Mar-2010 02:53:36] forsberg is now known as fOrsberg

[05-Mar-2010 04:10:55] <frozty_sa> does the squid zenpack still work for 2.5.x?

[05-Mar-2010 07:54:18] <frozty_sa> seems like it

[05-Mar-2010 07:54:25] <frozty_sa> (regarding my earlier squid question)

[05-Mar-2010 08:53:39] <dec3pti0n> why do I need to install MIBs within zenoss if that MIB is alread on the monitored nodes ?

[05-Mar-2010 08:56:06] <ckrough> dec3pti0n: zenoss needs to be able to read the mib to translate snmp traps to English

[05-Mar-2010 08:56:20] <ckrough> dec3pti0n: I believe that's the only thing it's used for. Without them you just get OID numbers in traps

[05-Mar-2010 08:58:35] <dec3pti0n> ah ok, thanks

[05-Mar-2010 09:09:39] <Michelle3> I'm new zenoss so if anyone can answer this, thanks in advance. I can 2019t figure out how to make a user have no global role if they have already been given one?  The list doesn 2019t give you way to choose no role or deselect the role they already have?

[05-Mar-2010 09:12:18] <ckrough> you should be able to control-click to enable and disable roles

[05-Mar-2010 09:12:58] <ckrough> When you hit save it just applies any roles that are highlighted. Any roles that are not longer highlighted will be taken away.

[05-Mar-2010 09:13:11] <ckrough> The changes for most Zenoss menus are atomic.

[05-Mar-2010 09:19:34] <Michelle3> That was way too easy. Thanks!

[05-Mar-2010 09:20:07] <ckrough> Your welcome

[05-Mar-2010 09:25:04] <Michelle3> I have noticed that when I delete a user and then recreate them later with the same name, the user retains the settings it had when it was deleted rather that being a "new user".  Is this normal?  It seems like the user is not really deleted. I've noticed this with organizers as well.

[05-Mar-2010 09:25:53] <ckrough> hmmm, never noticed that, but it sounds Zope-ish, which means its normal

[05-Mar-2010 09:26:14] <Michelle3> ha

[05-Mar-2010 09:27:13] <ckrough> You can gain some insight onto the inner workings by adding /manage to the url when you are in zenoss

[05-Mar-2010 09:27:29] <mrayzenoss> sounds like a bug to me

[05-Mar-2010 09:27:41] <ckrough> http://your_server/zport/dmd/manage

[05-Mar-2010 09:28:08] <ckrough> mrayzenoss: I was thining that type of stuff gets removed when you zeopack

[05-Mar-2010 09:28:19] <ckrough> *thinkig*

[05-Mar-2010 09:28:24] <ckrough> ...damn fingers

[05-Mar-2010 09:28:58] <Michelle3> k

[05-Mar-2010 09:46:28] <Brixius> well setting up the data collection and perf monitoring is making alot more sense to me now, I still have to figure out event settings (rules, regex, and transforms)

[05-Mar-2010 09:48:09] <ckrough> hmm, when I add information to a device's comment field (via Edit) I get this: 'ObjectNotFound: Object with id os not found on relation instances' in the event.log

[05-Mar-2010 09:48:29] <ckrough> and the same in the gui

[05-Mar-2010 09:48:42] <ckrough> OS Product has no options on the edit page, empty list

[05-Mar-2010 09:48:45] <ckrough> where is that list defined

[05-Mar-2010 09:48:53] <ckrough> (this magically started happening today)

[05-Mar-2010 09:53:06] <ckrough> Is it possible to reindex(), or otherwise repair things, for just one device?

[05-Mar-2010 10:32:16] <rmatte> ckrough: yeh, that's simple

[05-Mar-2010 10:32:29] <rmatte> d = dmd.Devices.findDevice('whatever')

[05-Mar-2010 10:32:32] <rmatte> d.reindex()

[05-Mar-2010 10:32:36] <rmatte> commit()

[05-Mar-2010 10:33:58] <rmatte> actually hmmm, thought that would work...

[05-Mar-2010 10:35:21] * rmatte checks some old notes

[05-Mar-2010 10:36:54] <ckrough> ok... IODrives are in production now. /crosses fingers

[05-Mar-2010 10:39:38] <rmatte> AHHHH

[05-Mar-2010 10:39:40] <rmatte> I figured it out

[05-Mar-2010 10:39:51] <rmatte> it's dmd.Devices.reIndex()

[05-Mar-2010 10:39:53] <rmatte> not reindex

[05-Mar-2010 10:39:57] <rmatte> (case sensitive)

[05-Mar-2010 10:40:13] <rmatte> now to see if that works for a single device

[05-Mar-2010 10:40:22] <rmatte> yup

[05-Mar-2010 10:40:33] <rmatte> so to reindex a single device:

[05-Mar-2010 10:40:37] <rmatte> d = dmd.Devices.findDevice('whatever')

[05-Mar-2010 10:40:44] <rmatte> d.reIndex()

[05-Mar-2010 10:40:47] <rmatte> commit()

[05-Mar-2010 10:41:21] * ckrough testing it out

[05-Mar-2010 10:41:44] <ckrough> I put together a zendmd standalone that can rebalance devices across collectors

[05-Mar-2010 10:41:51] <rmatte> cool

[05-Mar-2010 10:42:12] <ckrough> you set a MAX# of devices per collector and it iterates through them reassigning devices, including rsyncing all the RRD files

[05-Mar-2010 10:42:22] <ckrough> once I get it 'right' and trust it Ill upload it

[05-Mar-2010 10:42:35] <ckrough> I want it to balance by OID count rather than device count though

[05-Mar-2010 10:42:55] <rmatte> that's an awesome idea

[05-Mar-2010 10:43:04] <ckrough> Im using it in dev now, it works

[05-Mar-2010 10:43:19] <ckrough> I hope to get it to the point where it handles errors gracefully so I can just throw it in cron

[05-Mar-2010 10:43:36] <rmatte> what types of errors have you seen?

[05-Mar-2010 10:43:37] <ckrough> or zenjobs, though I havent checked that out yet since Im still on 2.3.3

[05-Mar-2010 10:43:56] <ckrough> not many yet, but a lot of what ifs. what if rsync fails, what if there is a opermission problem, etc...

[05-Mar-2010 10:44:04] <ckrough> I dont want to accidentally lose RRD data

[05-Mar-2010 10:44:10] <SEJeff_work> ckrough, IODrives being fusion-io?

[05-Mar-2010 10:44:27] <ckrough> there will always be a small gap during the rsync, but losing whole rrd files would be bad

[05-Mar-2010 10:44:30] <ckrough> SEJeff_work: yes

[05-Mar-2010 10:44:36] <SEJeff_work> Make _absolute sure_ you are using the latest firmware and drivers

[05-Mar-2010 10:44:43] <SEJeff_work> We've seen epic fail dataloss with older drivers

[05-Mar-2010 10:44:50] <ckrough> SEJeff_work: how much older?

[05-Mar-2010 10:45:08] <SEJeff_work> ckrough, I dunno, we've had them a year and some change. Just make sure you're on the latest as of now release

[05-Mar-2010 10:45:10] <ckrough> SEJeff_work: I've brought all mine up to whatever was current a couple months ago

[05-Mar-2010 10:45:15] <ckrough> Ill check

[05-Mar-2010 10:45:20] <ckrough> thanks

[05-Mar-2010 10:45:21] <SEJeff_work> just a heads up

[05-Mar-2010 10:45:26] <ckrough> are you using them too?

[05-Mar-2010 10:45:39] <SEJeff_work> And when they go offline hard, it might take an hour or so for them to do consistency checks

[05-Mar-2010 10:45:47] <SEJeff_work> fyi, the ramsan drives are much faster

[05-Mar-2010 10:46:05] <SEJeff_work> and have fully open source drivers fwiw

[05-Mar-2010 10:46:27] <ckrough> Im running them in software raid1 and I rsync back to a local RAID10 because I dont fully trust them

[05-Mar-2010 10:47:15] <SEJeff_work> ckrough, Also make sure to balance out your interrupts via irq affinity or irqbalance if you don't care much about the apps

[05-Mar-2010 10:47:41] <venturaville> SEJeff_work: what do the ramsan drives list for?

[05-Mar-2010 10:47:46] <ckrough> heh

[05-Mar-2010 10:48:26] <ckrough> SEJeff_work: why irqbalance?

[05-Mar-2010 10:50:05] <SEJeff_work> venturaville, I dunno, you can beat up the salesmen and get them cheaper :)

[05-Mar-2010 10:50:23] <venturaville> I'm sure.. was mostly just curious if they are in the ballpark of FusionIO

[05-Mar-2010 10:50:39] <ckrough> There is an interesting article online regarding how CCP uses RamSan for EVEOnline, good read.

[05-Mar-2010 10:51:14] <SEJeff_work> venturaville, Cheaper, and faster

[05-Mar-2010 10:51:57] <ckrough> thats surprising, ramsan used to be majorly expensive

[05-Mar-2010 10:52:32] <venturaville> http://storagemojo.com/2009/03/10/tms-announces-450-gb-pci-e-ssd/

[05-Mar-2010 10:52:38] <venturaville> $18K list price for the larger one

[05-Mar-2010 10:54:44] <venturaville> that would probably put the smaller one in the $12K list price range

[05-Mar-2010 10:54:48] <venturaville> which I think is in the ballpark

[05-Mar-2010 10:56:36] <SEJeff_work> venturaville, Can't speak list, but they are cheaper than fusion io

[05-Mar-2010 10:58:47] <venturaville> hrm probably ... that page was from a year ago

[05-Mar-2010 11:14:33] <chemist> blist

[05-Mar-2010 12:09:30] <mrayzenoss> Anyone with a Slashdot account, feel free to suggest this article: http://slashdot.org/submission/1186968/

[05-Mar-2010 12:45:33] <car1> has anybody seen this error when running commands on a remote server through ssh: RROR zen.ZenActions: Command restart_tomcat says ssh: /usr/local/zenoss/common/lib/libcrypto.so.0.9.8: no version information available (required by ssh)

[05-Mar-2010 12:45:37] <car1> Segmentation fault

[05-Mar-2010 12:45:38] <car1> ?

[05-Mar-2010 12:45:47] <car1> after this attempted to run zenstatus crashed

[05-Mar-2010 12:46:08] <car1> i have a public key on the server, and can ssh to it with no password from the zenoss server

[05-Mar-2010 12:47:37] <RoAkSoAx> How do I monitor devices that are under a NAT?

[05-Mar-2010 12:47:38] <car1> i tried unsetting LD_LIBRARY_PATH after reading this: message/22047#22047

[05-Mar-2010 12:47:42] <car1> but that made no difference

[05-Mar-2010 13:16:52] <venturaville> yeah

[05-Mar-2010 13:16:55] <venturaville> bigegor is on :-)

[05-Mar-2010 13:17:47] <bigegor> hello all

[05-Mar-2010 13:20:26] <venturaville> bigegor: I was hoping you could help me with the syntax for the queries field in your subversioned copy of WBEMDataSource

[05-Mar-2010 13:20:50] <venturaville> having trouble mapping what I have in yawn to it

[05-Mar-2010 13:21:37] <Brixius> Hello, I had a question for you on wmidatasource 2.0.98(I think it was) from svn if you got a moment... I actually downgraded to 1.5 which appears to work.

[05-Mar-2010 13:21:38] <bigegor> venturaville: i can try.

[05-Mar-2010 13:22:26] <bigegor> Brixius: i've put 2.0.99 on svn

[05-Mar-2010 13:22:26] <theacolyte> bigegor: Hey man, 2.5.2 is out!

[05-Mar-2010 13:22:59] <bigegor> theacolyte: i know ;)

[05-Mar-2010 13:23:29] <theacolyte> When are you going live with your new pack!

[05-Mar-2010 13:25:30] <Brixius> bigegor: one question, the error I was getting was an error on line 266, Something like  url[2], zWMIProxy = url[2].split('@')

[05-Mar-2010 13:25:45] <rmatte> Brixius: what version of Zenoss are you using again?

[05-Mar-2010 13:26:09] <Brixius> 2.5.1 at work, I installed 2.5.2 at home in a test

[05-Mar-2010 13:26:13] <rmatte> k

[05-Mar-2010 13:27:24] <rmatte> Brixius: so was it erroring on 2.5.1 or 2.5.2?

[05-Mar-2010 13:27:33] <Brixius> 2.5.1

[05-Mar-2010 13:27:36] <rmatte> ah

[05-Mar-2010 13:27:40] <bigegor> Brixius: have you change zWmiProxy?

[05-Mar-2010 13:27:49] <rmatte> Because people were complaining about it in 2.5.2: message/46365#46365

[05-Mar-2010 13:27:56] <Brixius> no, it's blank on my zproperitys

[05-Mar-2010 13:28:01] <Brixius> on the device

[05-Mar-2010 13:29:31] <Brixius> oh, ya, I was having that problem on 2.5.2 at home, zenpacks wouldn't install from the ui, I had to go out and run zenpack --install /tmp/(zenpackname) which would install them...

[05-Mar-2010 13:30:44] <Brixius> also had issues with wmic on 2.5.2, WMIDataSource 1.5 was complaining that it couldn't find wmic, but when I did a su - zenoss I could run wmic just fine.

[05-Mar-2010 13:31:04] <Brixius> from the command linek

[05-Mar-2010 13:33:59] <rmatte> eugh, wtf is going on today, all of my Zenoss servers are crashing and burning, I'm having to restart them to be able to hit the web interfaces

[05-Mar-2010 13:34:53] <bigegor> Brixius: have you in your username, password or in hostname symbols "@:/"?

[05-Mar-2010 13:35:18] <Brixius> hmm, well now that I think about it.... yes...

[05-Mar-2010 13:36:20] <Brixius> I didn't even think about that I do use an @ in a password... Perhaps that's my problem...

[05-Mar-2010 13:39:50] <bigegor> Brixius: @ used as delimiter for credentials and hostname

[05-Mar-2010 13:40:51] <Brixius> I'm changing my password to not have an @ in it and see if that fixes the problem.

[05-Mar-2010 13:41:07] <RoAkSoAx> rmatte, how to make zenoss nmap a server? :)

[05-Mar-2010 13:43:18] <rmatte> RoAkSoAx: http://dev.zenoss.org/trac/ticket/5238#comment:12

[05-Mar-2010 13:43:46] <RoAkSoAx> rmatte, thanks

[05-Mar-2010 13:44:59] <RoAkSoAx> rmatte, btw.. that also applies to zenoss 2.5.1?

[05-Mar-2010 13:46:58] <rmatte> yes

[05-Mar-2010 13:53:13] <Brixius> ok, I've changed my password in AD and zenoss zWwinPassword to not have an @ in it, and I still get the following from WMIDataSource.

[05-Mar-2010 13:53:16] <Brixius> Get WMI Instance SELECT UserCount FROM Win32_PerfFormattedData_MSExchangeIS_MSExchangeIS from //10.31.0.13/root/cimv2

[05-Mar-2010 13:53:16] <Brixius> Traceback (most recent call last):

[05-Mar-2010 13:53:16] <Brixius> File "/usr/local/zenoss/zenoss/ZenPacks/ZenPacks.community.WMIDataSource-2.0.99-py2.4.egg/ZenPacks/community/WMIDataSource/WMIClient.py", line 323, in ?

[05-Mar-2010 13:53:16] <Brixius> results = WmiGet(url, query, properties)

[05-Mar-2010 13:53:16] <Brixius> File "/usr/local/zenoss/zenoss/ZenPacks/ZenPacks.community.WMIDataSource-2.0.99-py2.4.egg/ZenPacks/community/WMIDataSource/WMIClient.py", line 265, in WmiGet

[05-Mar-2010 13:53:17] <Brixius> url[2], device.zWmiProxy = url[2].split('@')

[05-Mar-2010 13:53:17] <Brixius> ValueError: need more than 1 value to unpack

[05-Mar-2010 13:56:02] * Brixius bangs head on nearest hard surface... apparently the format for the username is domainusername not domain/username

[05-Mar-2010 14:08:17] <Brixius> Thanks for the help bigegor, you hint about the @:/ symbols fixed my problem.

[05-Mar-2010 14:08:49] <bigegor> Brixius: np

[05-Mar-2010 14:15:16] <mrayzenoss> bigegor: looks like a lot of fixes going into SVN, let me know when you want me to publish new versions

[05-Mar-2010 14:24:16] <RoAkSoAx> how can I log into the Zenoss virtual appliance using ssh?

[05-Mar-2010 14:26:58] <Simon4> mrayzenoss: is there a post 2.5.2 on the way?

[05-Mar-2010 14:27:38] <mrayzenoss> RoAkSoAx: you can login from the console as 'root', no password

[05-Mar-2010 14:27:53] <mrayzenoss> RoAkSoAx: then use 'passwd zenoss' to setup a password for the 'zenoss' user

[05-Mar-2010 14:27:57] <mrayzenoss> then you can SSH in

[05-Mar-2010 14:28:07] <RoAkSoAx> mrayzenoss, yeah i know but i can't access the console :)

[05-Mar-2010 14:28:42] <mrayzenoss> RoAkSoAx: hmmm… that's a bit of a problem

[05-Mar-2010 14:29:02] <RoAkSoAx> mrayzenoss, indeed i don't know why vmware is not allowing me :)

[05-Mar-2010 14:29:12] <mrayzenoss> someone should make a console portlet, that would be super awesome

[05-Mar-2010 14:30:02] <RoAkSoAx> mrayzenoss, does 2.5.2 includes the fix for portscan?

[05-Mar-2010 14:30:18] <mrayzenoss> RoAkSoAx: I don't think so.

[05-Mar-2010 14:30:57] <mrayzenoss> Simon4: we may respin the 2.5.2 .debs if we can nail down the cause of the issue, and there will be a 2.5.3 target for tickets, but we are hoping not to do a 2.5.3

[05-Mar-2010 14:31:12] <Simon4> mrayzenoss: cool

[05-Mar-2010 14:31:41] * Simon4 will be spinning up 2.5.2 on redhat in a week (when back from holiday), will be good to ahve a play

[05-Mar-2010 14:34:11] <RoAkSoAx> mrayzenoss, :( anyways. next question :)... How can I monitor servers that are under NAT?

[05-Mar-2010 14:36:48] <Simon4> RoAkSoAx: many servers?

[05-Mar-2010 14:36:58] <RoAkSoAx> Simon4, not really, just a couple

[05-Mar-2010 14:37:15] <mrayzenoss> if your NAT box can do port forwarding, you could use custom ports

[05-Mar-2010 14:37:30] <Simon4> easiest way I can think of is put in NAT port forwards for each host on a different UDP port, so say 162 goes to server1:161, 163 to server2:161 etc

[05-Mar-2010 14:37:46] <Simon4> then change zproperties for $host1 to be port 162 on $nat_ip

[05-Mar-2010 14:37:49] <Simon4> etc

[05-Mar-2010 14:37:51] <mrayzenoss> Simon4: RoAkSoAx: yeah, I do that with SSH to my home boxes

[05-Mar-2010 14:37:55] <Simon4> what mrayzenoss said in about 2 seconds :)

[05-Mar-2010 14:38:10] <RoAkSoAx> ok i guess that problem is resolved then :)

[05-Mar-2010 14:38:36] <mrayzenoss> it's not the easiest to scale solution, but it works

[05-Mar-2010 14:38:45] <mrayzenoss> ie. it's OK for a handful of machines

[05-Mar-2010 14:39:16] <Simon4> we ended up patching bits to support the snmp context - and talk to a custom proxy that then snmp queries individual devices based on the context name

[05-Mar-2010 14:39:26] <mrayzenoss> libvirt Virtualization ZenPack updated: docs/DOC-4687  Hot stuff

[05-Mar-2010 14:39:27] <Simon4> since we have places with >50 machines behind very restrictive firewalls

[05-Mar-2010 14:41:03] <Simon4> mrayzenoss: would be nice if the graphs screenshot on that page actually zoomed in when you clicked on it

[05-Mar-2010 14:41:09] <Simon4> the other images seem to

[05-Mar-2010 14:41:35] <Simon4> looks really cool, also

[05-Mar-2010 14:41:45] <RoAkSoAx> now, I have a LinksysWRT54GS running with dd-wrt. I want to get some info from it using telnet, since the router does not support SSH. I changed the router to class /Server/SSH/Linux, and in zProperties changed the port and the protocol to use telnet, what else do I need to change for it to work?

[05-Mar-2010 14:45:17] <RoAkSoAx> due to it keeps telling me: 2010-03-05 20:39:42     WARNING     zen.ZenModeler     Client DD-WRT timeout

[05-Mar-2010 15:08:33] <RoAkSoAx> and is it possible for two zenoss instances on different networks exchange information?

[05-Mar-2010 15:12:40] <mrayzenoss> ok, anyone here running 2.5.2 from the .deb installer?  I need some troubleshooting

[05-Mar-2010 15:12:48] <mrayzenoss> we have a theory about what is broken

[05-Mar-2010 15:12:52] <mrayzenoss> with the .deb installer

[05-Mar-2010 15:13:54] <bigegor> mrayzenoss: i'm installing it now

[05-Mar-2010 15:14:17] <mrayzenoss> have you done any customization to your shell environment for the zenoss user?

[05-Mar-2010 15:15:10] <RoAkSoAx> mrayzenoss, when are you guys gonna change to python 2.6+ so that we can include zenoss in Ubutnnu :)

[05-Mar-2010 15:15:20] <mrayzenoss> possibly this weekend

[05-Mar-2010 15:15:21] <RoAkSoAx> Ubuntu :) 8.10 and Up

[05-Mar-2010 15:15:51] <RoAkSoAx> mrayzenoss, i can possibly help to package it so that's included in the repositories :)

[05-Mar-2010 15:16:01] <bigegor> mrayzenoss: no

[05-Mar-2010 15:16:28] <mrayzenoss> bigegor: could you echo $SHELL for me?

[05-Mar-2010 15:21:36] <bigegor> /bin/bash

[05-Mar-2010 15:21:56] <mrayzenoss> hmm… that knocks our theory

[05-Mar-2010 15:22:26] <mrayzenoss> Ubuntu switched to dash as their /bin/sh link, we thought we might be making some bash assumptions

[05-Mar-2010 15:22:51] <mrayzenoss> bigegor: does /bin/sh link to dash?

[05-Mar-2010 15:22:59] <RoAkSoAx> it is

[05-Mar-2010 15:23:28] <bigegor> yes

[05-Mar-2010 15:30:18] <mrayzenoss> We may be assuming /bin/sh links to /bin/bash when on Debian and Ubuntu machines it links to /bin/dash, which could cause other problems

[05-Mar-2010 15:32:31] <RoAkSoAx> mrayzenoss, is it possible to make to zenoss machines "share" infor?

[05-Mar-2010 15:34:46] <mrayzenoss> yeah, with Enterprise there's a global dashboard

[05-Mar-2010 15:34:56] <mrayzenoss> and you can do distributed collectors

[05-Mar-2010 15:35:06] <mrayzenoss> rmatte has a script for doing those with Core

[05-Mar-2010 15:35:28] <Simon4> is that script hosted somewhere convenient?

[05-Mar-2010 15:35:40] <Simon4> (interested to compare with what I'm doing minus a script)

[05-Mar-2010 15:36:31] <RoAkSoAx> mrayzenoss, good to know for the presentation ;)

[05-Mar-2010 15:39:30] <mrayzenoss> RoAkSoAx: if you're interested in Enterprise features: http://www.zenoss.com/product/tours/enterprise

[05-Mar-2010 15:39:50] <mrayzenoss> 95% of that is in Core too

[05-Mar-2010 15:39:59] <RoAkSoAx> cool

[05-Mar-2010 15:40:10] <mrayzenoss> the new datacenter visualization stuff has some screenshots there

[05-Mar-2010 15:41:22] <RoAkSoAx> yeah I can see them

[05-Mar-2010 15:41:24] * Simon4 files away the visualisation stuff to show management next week

[05-Mar-2010 15:41:30] <RoAkSoAx> mrayzenoss, but the professor actually wants the real thing

[05-Mar-2010 15:41:35] <Simon4> (trying to convince em to drop the cash on enterprise)

[05-Mar-2010 15:42:24] <RoAkSoAx> Simon4, It's for a class presentation... however I would mention that Zenoss has both an Enterprise and OpenSource version and etc etc etc

[05-Mar-2010 15:43:13] <Simon4> mrayzenoss: do you know if the ESX stuff works with ESXi as well as full blown ESX?

[05-Mar-2010 15:43:27] <mrayzenoss> Simon4: in Enterprise or Core?

[05-Mar-2010 15:43:50] <Simon4> either really, if enterprise it's more ammo for me to take to management, and if in core I'll just be happy :)

[05-Mar-2010 15:44:12] <mrayzenoss> Enterprise is VMware ESX certified and works with vCenter as well

[05-Mar-2010 15:44:19] <mrayzenoss> it does a ton of stuff

[05-Mar-2010 15:44:33] <Simon4> sure, but will it talk to individual ESXi servers with no virtualcenter?

[05-Mar-2010 15:44:34] <mrayzenoss> for Core there are several Community ZenPacks for ESX and ESXi

[05-Mar-2010 15:44:39] <mrayzenoss> yes

[05-Mar-2010 15:44:42] <Simon4> cool

[05-Mar-2010 15:45:04] <mrayzenoss> I mentioned vCenter because you can point it at that and it'll pick up all your ESX installs as well

[05-Mar-2010 15:45:09] <Simon4> *nod*

[05-Mar-2010 15:45:12] <mrayzenoss> and track vmotions across ESX hosts

[05-Mar-2010 15:45:48] <theacolyte> I'm not sure if any of them work against vsphere

[05-Mar-2010 15:45:56] <theacolyte> in fact, I'm pretty sure they don't

[05-Mar-2010 15:47:30] <mrayzenoss> theacolyte: according to the libvirt docs, this should work with vSphere docs/DOC-4687

[05-Mar-2010 15:48:40] <theacolyte> not familiar with libvirt... interesting

[05-Mar-2010 16:10:50] <sciolist> what is the default timeout for a command check? 30 seconds?

[05-Mar-2010 16:11:33] <mrayzenoss> I think it's 60

[05-Mar-2010 16:11:39] <mrayzenoss> wait, timeout

[05-Mar-2010 16:11:48] <sciolist> is that the 'Cycle Time' parameter?

[05-Mar-2010 16:12:02] <mrayzenoss> zCommandCommandTimeout

[05-Mar-2010 16:12:04] <mrayzenoss> 15 seconds

[05-Mar-2010 16:12:23] <mrayzenoss> zCommandLoginTimeout

[05-Mar-2010 16:12:25] <mrayzenoss> 10 seconds

[05-Mar-2010 16:12:39] <mrayzenoss> and the cycle time is 60

[05-Mar-2010 16:13:20] <sciolist> hrm, I've got an email check command that's going over 75 seconds regularly

[05-Mar-2010 16:13:35] <sciolist> when I test it I get "Command timed out

[05-Mar-2010 16:13:35] <sciolist> DONE in 30 seconds"

[05-Mar-2010 16:16:06] <sciolist> so increase of zCommandCommandTimeout to 120 should get me there?

[05-Mar-2010 16:18:05] <sciolist> nope, no effect

[05-Mar-2010 16:18:25] <sciolist> still times out in 30-31 seconds

[05-Mar-2010 18:51:33] <chkraise> ls

[05-Mar-2010 19:41:34] fOrsberg is now known as forsberg

[05-Mar-2010 23:04:54] <mistich> anyone from zenoss around

[05-Mar-2010 23:05:03] <theacolyte> Not *from* zenoss, I doubt it

[06-Mar-2010 00:00:30] [disconnected at Sat Mar  6 00:00:30 2010]

[06-Mar-2010 00:00:30] [connected at Sat Mar  6 00:00:30 2010]

[06-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[07-Mar-2010 00:00:30] [disconnected at Sun Mar  7 00:00:30 2010]

[07-Mar-2010 00:00:30] [connected at Sun Mar  7 00:00:30 2010]

[07-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[07-Mar-2010 07:36:41] <Egyptian> afternoon all

[07-Mar-2010 07:36:57] <Egyptian> can anyone help me with creating a custom report? i dont want it to be multi-graph.. just data

[07-Mar-2010 07:37:41] <Egyptian> whenever i follow the zenoss admin guide (chp 12.8.2) it keeps telling me unkonw device to getId .. i am sure i m writing it wrong

[07-Mar-2010 07:37:48] <Egyptian> could somoene point me out where my error is?

[07-Mar-2010 07:38:28] <rmatte> what are the fields that you're trying to grab?

[07-Mar-2010 07:39:12] <Egyptian> just like the documentation

[07-Mar-2010 07:39:15] <Egyptian> getIp

[07-Mar-2010 07:39:19] <Egyptian> hiya rmatte

[07-Mar-2010 07:39:26] <Egyptian> zenoss wizard :)

[07-Mar-2010 07:39:28] <rmatte> just getIp?

[07-Mar-2010 07:39:50] <rmatte> it's getManageIp

[07-Mar-2010 07:40:01] <rmatte> you can use....

[07-Mar-2010 07:40:02] <rmatte> getId

[07-Mar-2010 07:40:02] <rmatte> getManageIp

[07-Mar-2010 07:40:02] <rmatte> getHWManufacturerName

[07-Mar-2010 07:40:02] <rmatte> getHWProductName

[07-Mar-2010 07:40:02] <rmatte> getHWSerialNumber

[07-Mar-2010 07:40:03] <rmatte> getPingStatusString

[07-Mar-2010 07:40:03] <rmatte> getSnmpStatusString

[07-Mar-2010 07:40:03] <Egyptian> no

[07-Mar-2010 07:40:04] <rmatte> getPriorityString

[07-Mar-2010 07:40:12] <Egyptian> and getManageIP

[07-Mar-2010 07:40:32] <Egyptian> getId and getManageIp

[07-Mar-2010 07:40:36] <Egyptian> i cant write thats all

[07-Mar-2010 07:40:40] <Egyptian> so what is the seperator

[07-Mar-2010 07:40:50] <Egyptian> coz when i tried getIP now by itself it works

[07-Mar-2010 07:41:08] <rmatte> they need to be on different lines

[07-Mar-2010 07:41:20] <rmatte> Columns should be...

[07-Mar-2010 07:41:21] <rmatte> getId

[07-Mar-2010 07:41:21] <rmatte> getManageIp

[07-Mar-2010 07:41:32] <rmatte> Column names should be...

[07-Mar-2010 07:41:33] <rmatte> Device

[07-Mar-2010 07:41:34] <rmatte> IP Address

[07-Mar-2010 07:41:36] <Egyptian> now it works

[07-Mar-2010 07:41:39] * Egyptian is puzzled

[07-Mar-2010 07:41:42] <rmatte> :P

[07-Mar-2010 07:42:10] <Egyptian> YOU!

[07-Mar-2010 07:42:16] <Egyptian> you cast a spell on my zenoss installation!

[07-Mar-2010 07:42:24] <Egyptian> duh

[07-Mar-2010 07:42:25] <Egyptian> which one is mac address ?

[07-Mar-2010 07:42:26] <rmatte> hehe

[07-Mar-2010 07:42:51] <rmatte> you can't do Mac Address, that's an interface property, not a device property

[07-Mar-2010 07:43:20] <rmatte> If you need to do Mac Address for interfaces you'd have to literally code a report in Python and Tales from scratch

[07-Mar-2010 07:43:31] <Egyptian> ack

[07-Mar-2010 07:43:44] <Egyptian> i need it for the network inventory .. ip to mac is what is required

[07-Mar-2010 07:44:00] <Egyptian> and i am definetly not a coder

[07-Mar-2010 07:44:02] <Egyptian> nyet!

[07-Mar-2010 07:44:23] <rmatte> I'm a coder, but I didn't know any Python before I started using Zenoss

[07-Mar-2010 07:44:34] <rmatte> It's very handy to learn

[07-Mar-2010 07:45:13] * Egyptian umms

[07-Mar-2010 07:45:16] * Egyptian haws

[07-Mar-2010 07:45:26] <rmatte> but yeh, the UI based report designing is very limited

[07-Mar-2010 07:45:41] <rmatte> they will be looking at rewriting the reports system after stone crab is released

[07-Mar-2010 07:47:31] <Egyptian> stone crab .. are we going back to the stone ages?

[07-Mar-2010 07:47:45] * Egyptian wishes somebody would name something after egypt

[07-Mar-2010 07:47:51] <Egyptian> and not use cairo

[07-Mar-2010 07:47:54] <Egyptian> maybe "traffic"

[07-Mar-2010 07:48:06] <Egyptian> or "ramsis"

[07-Mar-2010 07:48:11] <Egyptian> or "nefertiti"

[07-Mar-2010 07:48:12] <rmatte> I'm voting "ghost crab" for the release after that

[07-Mar-2010 07:48:18] <rmatte> or "coconut crab"

[07-Mar-2010 07:48:24] * Egyptian is running away with this egyptian theme

[07-Mar-2010 07:48:33] <Egyptian> not walking away from it ..definetly

[07-Mar-2010 07:49:02] <Egyptian> right .. my report works .. now to figure out what kind of reports i can do

[07-Mar-2010 07:49:03] <rmatte> egypt themes are fine, it just needs to be egyptian gods

[07-Mar-2010 07:49:15] <Egyptian> ramsis thought he was a god

[07-Mar-2010 07:49:20] <Egyptian> isnt that good enough?

[07-Mar-2010 07:49:27] <rmatte> osiris, set, ra, hathor, Apep, Anubis, etc...

[07-Mar-2010 07:49:41] <Egyptian> i vote for anubis

[07-Mar-2010 07:49:56] <Egyptian> :)

[07-Mar-2010 07:50:12] <Egyptian> gtg ..prayers then meeting

[07-Mar-2010 07:50:15] <Egyptian> no rest for the wicked

[07-Mar-2010 07:50:16] <Egyptian> bye

[07-Mar-2010 07:50:19] <rmatte> later

[07-Mar-2010 07:50:20] <Egyptian> oh .. thanks rmatte

[07-Mar-2010 07:50:24] <rmatte> np

[07-Mar-2010 07:50:44] * Egyptian puts some lime in the cocunut

[07-Mar-2010 14:33:37] zz_forsberg is now known as forsberg

[07-Mar-2010 14:50:32] forsberg is now known as zz_forsberg

[07-Mar-2010 14:51:22] zz_forsberg is now known as forsberg

[07-Mar-2010 14:57:45] forsberg is now known as fOrsberg

[07-Mar-2010 14:58:30] fOrsberg is now known as forsberg

[07-Mar-2010 15:00:07] <forsberg> sorry for spam, just setting up new bnc

[07-Mar-2010 15:00:08] <forsberg> ;>

[07-Mar-2010 18:07:39] forsberg is now known as fOrsberg

[08-Mar-2010 00:00:30] [disconnected at Mon Mar  8 00:00:30 2010]

[08-Mar-2010 00:00:31] [connected at Mon Mar  8 00:00:31 2010]

[08-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[08-Mar-2010 01:31:43] fOrsberg is now known as forsberg

[08-Mar-2010 02:24:33] forsberg is now known as fOrsberg

[08-Mar-2010 02:43:47] fOrsberg is now known as forsberg

[08-Mar-2010 07:37:11] forsberg is now known as fOrsberg

[08-Mar-2010 07:47:55] <Egyptian> hi

[08-Mar-2010 07:48:12] <Egyptian> is the yum repo fedora based?

[08-Mar-2010 07:55:08] <rmatte> well, it's not really "based" on anything

[08-Mar-2010 07:55:15] <rmatte> it's an RPM, it'll install on any Linux system

[08-Mar-2010 07:55:35] <rmatte> It's probably built on CentOS or RHEL

[08-Mar-2010 07:55:54] <rmatte> but it supports SuSe, Fedora, etc...

[08-Mar-2010 07:56:30] <rmatte> I've gotta go, heading to work

[08-Mar-2010 07:56:33] <rmatte> bbl

[08-Mar-2010 08:05:14] <Egyptian> well i am trying to yum update my zenoss installation but its 65M .. and the newer versions of create repo has support for delta rpms .. delta rpms only push the changes to yum which then "patches" the changes

[08-Mar-2010 08:08:55] <ke4qqq> Egyptian: I don't think they are using delta rpms

[08-Mar-2010 08:09:02] <ke4qqq> I'd honestly be shocked if they were

[08-Mar-2010 08:38:35] <Troubadix09> hi

[08-Mar-2010 08:39:08] <Egyptian> shocked why?

[08-Mar-2010 08:39:12] <Troubadix09> I have an issue with the dashboard after upgrading from zenoss 2.5.1 to 2.5.2

[08-Mar-2010 08:39:19] <Egyptian> and i am suggesting delta rpms

[08-Mar-2010 08:39:26] <Egyptian> instead of the 65m download

[08-Mar-2010 08:39:34] <Egyptian> it doesnt want to complete the download for me

[08-Mar-2010 08:40:39] <Troubadix09> in the left corner of the dashboard should written in red letters "Last Updated at 'Timestamp' PM"

[08-Mar-2010 08:41:36] <Troubadix09> but there is written in red letters "Lost connection to the Server"

[08-Mar-2010 08:41:57] <Troubadix09> but the dashboard refreshes every 60 sec.

[08-Mar-2010 08:43:32] <Troubadix09> and for a short time lower than one second there is "Last Updated at 'Timestamp' PM" and then changes to "Lost connection ..."

[08-Mar-2010 08:43:54] <Troubadix09> have someone else this behaviour?

[08-Mar-2010 08:44:15] <Troubadix09> it is zenoss core vm-appliance

[08-Mar-2010 08:53:49] <daveborg98> I am having trouble getting 3 daemons to start after a migration. These are zenping, zensyslog, and zentrap. This is CentoOS 5.4 64bit and Zenoss 2.5.1. I seem to be unable to get any logs to try and figure out how to fix it. Any ideas?

[08-Mar-2010 08:57:49] <mrayzenoss> check the permissions on the zensocket

[08-Mar-2010 08:58:20] <jb> hrm

[08-Mar-2010 08:58:32] <mrayzenoss> daveborg98: chown root:zenoss /usr/local/zenoss/bin/zensocket

[08-Mar-2010 08:58:38] <jb> i have a zen instance that is reporting devices down

[08-Mar-2010 08:58:39] <mrayzenoss> daveborg98: chmod 04750 /usr/local/zenoss/bin/zensocket

[08-Mar-2010 08:58:41] <jb> when they are actually up

[08-Mar-2010 09:01:06] <tehhobbit> jb: can you ping the device from the monitor host ?

[08-Mar-2010 09:05:10] <Troubadix09> mrayzenoss: are there any known problems with json.py after upgrading from 2.5.1 to 2.5.2?

[08-Mar-2010 09:05:46] <daveborg98> mrayzenoss: That fixed it. I think it was the one thing I did not try. Thank you.

[08-Mar-2010 09:05:48] <mrayzenoss> Troubadix09: I haven't heard of any

[08-Mar-2010 09:06:04] <mrayzenoss> the only things I'm aware of with 2.5.2 are related to the .deb installer

[08-Mar-2010 09:06:10] <Troubadix09> mrayzenoss: i get this error after upgrading zenoss-appliance via rpath-gui in zope-log

[08-Mar-2010 09:06:57] <Troubadix09> 2010-03-08T11:47:56 ERROR Zope.SiteErrorLog http://lv-tux-101.pari-nds.root.loc:8080/zport/dmd/ZenEventManager/getHeartbeatIssuesJSON

[08-Mar-2010 09:06:58] <Troubadix09> Traceback (innermost last):

[08-Mar-2010 09:06:58] <Troubadix09>   Module ZPublisher.Publish, line 119, in publish

[08-Mar-2010 09:06:58] <Troubadix09>   Module ZPublisher.mapply, line 88, in mapply

[08-Mar-2010 09:06:58] <Troubadix09>   Module ZPublisher.Publish, line 42, in call_object

[08-Mar-2010 09:06:58] <Troubadix09>   Module Products.ZenWidgets.browser.Portlets, line 200, in __call__

[08-Mar-2010 09:06:58] <Troubadix09>   Module Products.ZenUtils.json, line 41, in inner

[08-Mar-2010 09:06:59] <Troubadix09>   Module Products.ZenWidgets.browser.Portlets, line 216, in getHeartbeatIssuesJSON

[08-Mar-2010 09:06:59] <Troubadix09>   Module Products.ZenEvents.EventManagerBase, line 1071, in getHeartbeat

[08-Mar-2010 09:07:00] <Troubadix09>   Module MySQLdb.cursors, line 137, in execute

[08-Mar-2010 09:07:00] <Troubadix09>   Module MySQLdb.connections, line 33, in defaulterrorhandler

[08-Mar-2010 09:07:01] <Troubadix09> AttributeError: 'module' object has no attribute 'isoToTimestamp'

[08-Mar-2010 09:07:01] <Troubadix09> 2010-03-08T11:47:56 ERROR Zope.SiteErrorLog http://lv-tux-101.pari-nds.root.loc:8080/zport/dmd/ZenEventManager/getHeartbeatIssuesJSON

[08-Mar-2010 09:07:02] <Troubadix09> Traceback (innermost last):

[08-Mar-2010 09:07:18] <ke4qqq> Troubadix09: fpaste.org or pastebin.ca are your friends

[08-Mar-2010 09:07:27] <mrayzenoss> yeah, put it up there

[08-Mar-2010 09:08:26] <Troubadix09> sorry, it should only be the last line

[08-Mar-2010 09:10:58] <Troubadix09> paste here now http://www.fpaste.org/sO50/

[08-Mar-2010 09:12:32] <rmatte> mrayzenoss: when you say there are .deb installer issues, I assume that equates to issues with the standard stack installer as well?

[08-Mar-2010 09:13:00] <mrayzenoss> no

[08-Mar-2010 09:13:06] <rmatte> k

[08-Mar-2010 09:13:17] <mrayzenoss> rmatte: there seems to be a path issue with the .debs, not sure what is causing it

[08-Mar-2010 09:13:19] <rmatte> guess I'm lucky that I'm not using the .deb then

[08-Mar-2010 09:13:37] <rmatte> are the path issues what are causing those ZenPack install issues?

[08-Mar-2010 09:13:42] <rmatte> or is that something else?

[08-Mar-2010 09:13:43] <mrayzenoss> the path issue affects zenpack from the UI and zenperfwmi and zenjmx

[08-Mar-2010 09:13:57] <rmatte> ah ok, so the .deb was the commonality

[08-Mar-2010 09:13:59] <rmatte> interesting

[08-Mar-2010 09:14:16] <mrayzenoss> we believe it's related to the /bin/sh on Debian and Ubuntu pointing to dash instead of bash, we're still trying to track it down

[08-Mar-2010 09:14:26] <rmatte> ah

[08-Mar-2010 09:16:15] <mrayzenoss> Troubadix09: this was a 2.5.1 to 2.5.2 upgrade?

[08-Mar-2010 09:16:22] <mrayzenoss> which platform/installer?

[08-Mar-2010 09:16:25] <Troubadix09> mrayzenoss: yes

[08-Mar-2010 09:17:04] <Troubadix09> mrayzenoss: rpath zenoss-appliance

[08-Mar-2010 09:19:04] <Troubadix09> mrayzenoss: update via rpath-gui, then stop zenoss, delete all *.pid in $ZENHOME/var/, then restart the whole appliance via console "shutdown -r now"

[08-Mar-2010 09:19:48] <mrayzenoss> Troubadix09: so the event console is completely broken?

[08-Mar-2010 09:20:42] <Troubadix09> mrayzenoss: no, event-console works, but when i click an event for details, the details-window open with nothing in it and then i get "Server connection Error"

[08-Mar-2010 09:21:04] <mrayzenoss> any event?

[08-Mar-2010 09:21:09] <mrayzenoss> which browser

[08-Mar-2010 09:21:37] <Troubadix09> mrayzenoss: yes, on all events, firefox 3.6

[08-Mar-2010 09:22:16] <mrayzenoss> Troubadix09: try clearing out all your events, do new events do this too?

[08-Mar-2010 09:22:51] <mrayzenoss> and clear your browser cache

[08-Mar-2010 09:24:59] fOrsberg is now known as forsberg

[08-Mar-2010 09:25:07] <Troubadix09> mrayzenoss: yes, clear my events, clear event cache, clear browser cache, get new events, same error when clicking for event details

[08-Mar-2010 09:25:35] <mrayzenoss> nuts

[08-Mar-2010 09:26:53] <Troubadix09> mrayzenoss: top-right corner of the details-window the " (x) " to close it is greyed out and i can't close the details-window but in the background i can mark or delete events

[08-Mar-2010 09:27:36] <forsberg> k

[08-Mar-2010 09:29:56] <Troubadix09> mrayzenoss: or agknowlege them

[08-Mar-2010 09:30:20] <jb> hm

[08-Mar-2010 09:30:23] <Troubadix09> mrayzenoss: reload page would close the details-window

[08-Mar-2010 09:30:43] <jb> zenoss thinks that there are multiple daemons running for some things..

[08-Mar-2010 09:30:45] <jb> like zentrap

[08-Mar-2010 09:30:50] <jb> stale PID file or something maybe?

[08-Mar-2010 09:33:24] <Troubadix09> mrayzenoss: i don't know if it is important, but before upgrading always there was a zeo.zec -file in $ZENHOME/var/ after restart there isn't such a file

[08-Mar-2010 09:33:42] <rmatte> Troubadix09: that's just a cache file, it's not required

[08-Mar-2010 09:33:49] <jb> Daemon: zentrap program running; pid=21561

[08-Mar-2010 09:33:50] <Troubadix09> rmatte: thanks

[08-Mar-2010 09:33:53] <jb> Daemon: zentrap program running; pid=21561

[08-Mar-2010 09:33:58] <jb> "zenoss status" reports that

[08-Mar-2010 09:34:04] <jb> (twice)

[08-Mar-2010 09:34:33] <rmatte> Troubadix09: I actually find that running without .zec files is preferable, less issues

[08-Mar-2010 09:36:48] <Troubadix09> rmatte: but where can i turn it off? I haven't turn it of, it must be done between upgrading?!

[08-Mar-2010 09:36:57] <Troubadix09> of=off

[08-Mar-2010 09:39:35] <Troubadix09> mrayzenoss: i have found this error in zeo-log http://www.fpaste.org/Jx2Z/

[08-Mar-2010 09:40:11] <jb> ok, it was because of blogs/zenossblog/2009/01/27/tip-of-the-month-running-only-some-daemons/

[08-Mar-2010 09:45:20] <Troubadix09> mrayzenoss: and found this error in zenactions-log http://www.fpaste.org/vsWc/

[08-Mar-2010 09:49:39] <Troubadix09> mrayzenoss: the other logs have nothing special

[08-Mar-2010 09:56:18] <mrayzenoss> Troubadix09: I've got a meeting coming up.  Go open a ticket against 2.5.2 and put those traces in it.  We haven't recreated yet, but we can look at it

[08-Mar-2010 09:57:41] <Troubadix09> mrayzenoss: thanks for your help, i would open a ticket for it.

[08-Mar-2010 11:06:52] bryan_ is now known as Bryan

[08-Mar-2010 11:06:55] Bryan is now known as Bryanstein

[08-Mar-2010 11:50:33] <mr-m> how can i map snmp oids to a named oid? such as: .1.3.6.1.4.1.6431.1.1.1.2501 is bwApplicationServerUnreachable, my vendor gives me a map file and mibs separately.

[08-Mar-2010 12:20:37] <Troubadix09> mr-m: zenoss uses MIBs only for transform OIDs in readable form. Load the MIB-File into Zenoss and Zenoss will do it for you

[08-Mar-2010 12:22:34] <Troubadix09> mr-m: helpful would be to install the MIB-Util-Zenpack here docs/DOC-3412

[08-Mar-2010 12:23:33] <Troubadix09> mr-m: helpful for installing MIBs is this Document docs/DOC-2450

[08-Mar-2010 12:37:25] <rmatte> Troubadix09 meant to say it only uses it for Trap OIDs

[08-Mar-2010 12:37:32] <rmatte> not transform OIDs

[08-Mar-2010 12:37:59] <rmatte> the rest he stated is accurate

[08-Mar-2010 12:56:59] <Troubadix09> rmatte: thanks ;-)

[08-Mar-2010 12:57:35] <Troubadix09> rmatte: you are right, only for transforming trap-oids :)

[08-Mar-2010 12:59:40] <Troubadix09> rmatte: is there something new in the zenperfsnmp daemon in zenoss 2.5.2?

[08-Mar-2010 13:04:20] <Troubadix09> rmatte: forget my question, I restart the daemon and everything works now, except my event-console-problem :)

[08-Mar-2010 13:10:00] <Troubadix09> worktime is over, cu all

[08-Mar-2010 13:26:41] <ckrough> zendmd doesnt seem to respect '--pcachedir', anyone dealt with that?

[08-Mar-2010 13:26:51] <ckrough> I need to reindex() but dont have enough space in /tmp

[08-Mar-2010 13:30:24] forsberg is now known as fOrsberg

[08-Mar-2010 13:35:54] fOrsberg is now known as forsberg

[08-Mar-2010 13:44:33] forsberg is now known as fOrsberg

[08-Mar-2010 14:52:59] <rmatte> ckrough: honestly, no idea, never tried that

[08-Mar-2010 14:53:08] <rmatte> why is your /tmp so small?

[08-Mar-2010 14:53:15] <ckrough> 2g

[08-Mar-2010 14:53:32] <rmatte> and reindex is taking more than that?

[08-Mar-2010 14:53:39] <ckrough> ya

[08-Mar-2010 14:53:45] <ckrough> data.fs is 3g on that box

[08-Mar-2010 14:54:07] <ckrough> ended up unmounting /tmp and symlinking it to another location temporarily

[08-Mar-2010 15:05:49] <rmatte> ah

[08-Mar-2010 15:05:51] <rmatte> that works

[08-Mar-2010 15:08:28] <ckrough> yeah, seems to be working

[08-Mar-2010 18:18:03] straterr1 is now known as straterra

[08-Mar-2010 21:25:49] <mrayzenoss> New Redis Monitor Community ZenPack: docs/DOC-5333 for all you nosql lovers out there

[08-Mar-2010 21:26:09] <mrayzenoss> Thanks to chudler for that one!

[08-Mar-2010 21:26:22] * mrayzenoss is slowly killing zenpack backlog

[08-Mar-2010 21:29:43] <sergeymasushko> mrayzenoss: is there a way to download Zope objects tdit then and upload them back... just I'm looking for a solution where no need to make templates manually...

[08-Mar-2010 21:30:25] <mrayzenoss> sergeymasushko: did you see this? http://www.linuxdynasty.org/howto-add-multiple-datapoints-to-zenoss-using-the-zenoss-api.html

[08-Mar-2010 21:32:33] <sergeymasushko> mrayzenoss: oh! thanks, that is what I'm looking for.

[08-Mar-2010 21:34:34] <mrayzenoss> he's been updating that daily, there's a forum thread somewhere

[08-Mar-2010 23:10:08] bryan_ is now known as Bryanstein

[08-Mar-2010 23:21:06] <daveborg98> Would anyone know how to remove a zenpack if I cannot get to the WEB UI? I installed a WMIExchangeZenpack and it gave me an error on installing but I cannot seem to fix it.

[09-Mar-2010 00:00:30] [disconnected at Tue Mar  9 00:00:30 2010]

[09-Mar-2010 00:00:30] [connected at Tue Mar  9 00:00:30 2010]

[09-Mar-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[09-Mar-2010 01:20:09] fOrsberg is now known as forsberg

[09-Mar-2010 05:24:36] <Troubadix09> hello all

[09-Mar-2010 07:01:49] <sektorNBA> hi.

[09-Mar-2010 07:02:04] <sektorNBA> is there Device class path to freebsd?

[09-Mar-2010 07:27:25] <sektorNBA> hello

[09-Mar-2010 07:32:27] <Troubadix09> sektorNBA: i think "no", but you could dreate a Sub-Device-Class under whatever you want

[09-Mar-2010 07:34:43] <Troubadix09> sektorNBA: dreate=create :)

[09-Mar-2010 07:35:13] <sektorNBA> cause iam getting nan values when monitoring a freebsd 8.0

[09-Mar-2010 07:42:13] <Troubadix09> sektorNBA: snmpwalk works?

[09-Mar-2010 07:48:27] <sektorNBA> Troubadix09 yes it works.

[09-Mar-2010 07:54:37] <Troubadix09> hello mray

[09-Mar-2010 07:55:28] <mrayzenoss> good morning

[09-Mar-2010 07:57:59] <Troubadix09> sektorNBA: new Devices? because Graphs start drawing after a few polling rounds (5min cycle-time x 3 cycle minimum = 15min to start drawing)

[09-Mar-2010 07:59:29] <Troubadix09> mrayzenoss: thanks for having an eye on my ticket belongs to yesterday chat

[09-Mar-2010 08:00:21] <mrayzenoss> yeah, defect review is this afternoon, maybe someone will have an idea what that is

[09-Mar-2010 08:00:30] <sektorNBA> its is working. ;-)

[09-Mar-2010 08:00:35] <mrayzenoss> I couldn't recreate it, but I also didn't upgrade 2.5.1->2.5.2

[09-Mar-2010 08:00:47] <mrayzenoss> hmm… I guess I could do that

[09-Mar-2010 08:02:19] <Troubadix09> mrayzenoss: I don't know, if it is important, but the zenaction-daemon every 60 sec. creates the error in the log

[09-Mar-2010 08:02:47] <mrayzenoss> feel free to add that note

[09-Mar-2010 08:03:06] <sektorNBA> now i wanna monitor some os process.

[09-Mar-2010 08:04:58] <Troubadix09> mrayzenoss: i attached to the ticket, and i have two event-commands wich run a script (play a sound), if a new event with critical or  error severity coming in.

[09-Mar-2010 08:05:38] <Troubadix09> mrayzenoss: after upgrade the script never starts via zenoss

[09-Mar-2010 08:06:15] <Troubadix09> mrayzenoss: manual start via console as user zenoss work

[09-Mar-2010 08:15:03] <sektorNBA> how can i tell zenoss to monitor some running process?

[09-Mar-2010 08:24:35] <Troubadix09> sektorNBA: look here docs/DOC-3537

[09-Mar-2010 08:26:33] <Troubadix09> sektorNBA: and here docs/DOC-4810#d0e4180

[09-Mar-2010 08:37:31] <daveborg98> I am running into the same problem I had yesterday where the zenping, zensyslog, and zentrap daemons will not start. On this instance, I am running CentOS5.4 and Zenoss 2.5.2. I checked the permissions on zensocket (that fixed the issue yesterday). The last this the logs is Deleting PID file andthen shuts down.

[09-Mar-2010 08:38:03] <ckrough> daveborg98: have you tried strace with them?

[09-Mar-2010 08:48:25] <daveborg98> ckrough: Sorry, had a user support issue come up. I am not familiar with strace (I only know enough about Linux to be dangerous).

[09-Mar-2010 08:52:17] <jb> anyone recall that reportmail bug?

[09-Mar-2010 08:52:28] <jb> not parsing spaces correctly?

[09-Mar-2010 08:52:34] <jb> i know there is a patch, just can't find the ticket

[09-Mar-2010 09:01:39] <ckrough> daveborg98: ah. its fairly safe, google it, it may help. it shows you what the program is doing on a system level. If you can watch it with strace you may be able to see what it was doing when it crashed

[09-Mar-2010 09:01:59] <ckrough> daveborg98: are you running the daemon with  "zenblahblah.py run -v10" to see debugging info?

[09-Mar-2010 09:03:04] <daveborg98> ckrough: I was just getting ready to do that. I don't know how I miised it yesterday, but I was going to follow the troublshooting doc on the Zenoss website.

[09-Mar-2010 09:03:22] <daveborg98> ckrough: That was part of it.

[09-Mar-2010 09:09:36] <daveborg98> ckrough: Thank you for the help so far. It has pointed me to some useful info. It appears that the rrdtool package may not be installed. If I installed fromt he RPM, is that a possibility?

[09-Mar-2010 09:10:44] <ckrough> How did you isntall zenoss?

[09-Mar-2010 09:10:51] <ckrough> daveborg98: RPM, binary installer, etc.. ?

[09-Mar-2010 09:11:26] <daveborg98> ckrough: I installed via the CentOS 5 64 bit RPM.

[09-Mar-2010 09:11:29] <ckrough> daveborg98: RRDtool should be installed as part of zenoss, make sure you are the zenoss user when you look for it, you should find it in $ZENHOME/bin

[09-Mar-2010 09:17:13] <daveborg98> ckrough: It is there....I should be able to start the zensyslog daemon as root byt using '/opt/zenoss/bin/zensyslog run', correct?

[09-Mar-2010 09:18:22] <ckrough> as the zenoss user

[09-Mar-2010 09:18:32] <ckrough> almost everything should be done as the zenoss user

[09-Mar-2010 09:18:35] <ckrough> 'su - zenoss' from root

[09-Mar-2010 09:19:13] <ckrough> but yeah, as zenoss "/opt/zenoss/bin/zensyslog run -v10" would give you a foreground running zensyslog with debug output

[09-Mar-2010 09:21:07] * zenethian passes out

[09-Mar-2010 09:21:10] <zenethian> er

[09-Mar-2010 09:21:16] * zenethian passes out breakfast tacos

[09-Mar-2010 09:22:36] <daveborg98> Maybe I need to doublecheck my zensocket permissions. When I run the "/opt/zenoss/bin/zensyslog run -v10" it tells me "ZenSocket Error: zensocket needs to be run as root or setuid". I have set the zensocket permissions with "chmod 04750 /opt/zenoss/bin/zensocket".

[09-Mar-2010 09:22:49] <Troubadix09> bye all

[09-Mar-2010 09:23:16] <ckrough> daveborg98: who owns the file

[09-Mar-2010 09:25:43] <daveborg98> ckrough: zenoss owns it. "-rwsr-x---  1 zenoss root      9720 Feb 26 01:08 zensocket"

[09-Mar-2010 09:25:52] <ckrough> it should be root:zenoss

[09-Mar-2010 09:25:59] <ckrough> chown root:zenoss zensocket

[09-Mar-2010 09:29:16] <daveborg98> ckrough: There are several items in the /bin directory as zenoss:root and not root:zenoss. Should they all be changed?

[09-Mar-2010 09:29:33] <ckrough> I think zensocket is the only one that needs to be owned by root

[09-Mar-2010 09:31:43] <daveborg98> ckrough: crap, I still get the "zensocket needs to be run as root" error message. Would it be better to just rip and replace?

[09-Mar-2010 09:42:56] <daveborg98> ckrough: Thank you for the help. It is in a VM with a couple snapshots, I think I will revert to a previous snap and start over. This is not a production machine, so this may be easier to make sure I did not hose the install somehow.

[09-Mar-2010 09:43:09] <ckrough> daveborg98: heres the ls from a running box: "-rwsr-xr-x 1 root root 9688 Feb 18  2009 /opt/zenoss/bin/zensocket"

[09-Mar-2010 09:43:29] <ckrough> daveborg98: gl

[09-Mar-2010 09:44:16] <daveborg98> ckrough: Thanks. I can say that I am getting more familiar with Zenoss than I thought I would, which is a good thing.

[09-Mar-2010 10:34:51] <ckrough> mrayzenoss: thread/12959 is that puppet zenpack link working for you?

[09-Mar-2010 10:34:54] <ckrough> in the text

[09-Mar-2010 10:35:19] <mrayzenoss> hmm... nuts

[09-Mar-2010 10:35:25] <mrayzenoss> fixing it

[09-Mar-2010 10:35:54] <mrayzenoss> done

[09-Mar-2010 10:36:20] <ckrough> better, thanks!

[09-Mar-2010 10:58:16] <mrayzenoss> New AMQP ZenPack: docs/DOC-5817 Thanks ke4qqq!

[09-Mar-2010 11:00:10] <ckrough> I'd like to see more description about *what* is monitored in the zenpacks. Not picking on ke4qqq, this applies to almost all of them. Rather than 'provides the zenamqp daemon for monitoring', it would helpful if they said something along the lines of "this zenpack adds moniroting of process count, response time, blah blah, for AMPQ"

[09-Mar-2010 11:00:10] <theacolyte> hey mrayzenoss - what's that handy trac search link that shows changes between 2.5.1 and 2.5.2?

[09-Mar-2010 11:00:35] <ckrough> for some of the zenpacks I cant tell what monitoring they add for whatever it is that they are watching

[09-Mar-2010 11:01:05] <mrayzenoss> ckrough: yeah, I'm working on a framework for getting ZenPacks "certified".  Documentation is a key feature

[09-Mar-2010 11:01:17] <ckrough> cool

[09-Mar-2010 11:01:32] <mrayzenoss> in this case, AMQP is pretty well documented.  Events get captured by the daemon per device/class and show up on the console

[09-Mar-2010 11:01:38] <mrayzenoss> not much else

[09-Mar-2010 11:01:49] <ke4qqq> ckrough: yeah - I'll admit that's it has caused confusion - with people thinking that it monitors AMQP, but rather this essentially adds a daemon to permit listening for AMQP messages, and not monitoring AMQP itself - several people have asked that question

[09-Mar-2010 11:02:06] <mrayzenoss> we have a tool we're working on to generate documentation for ZenPacks

[09-Mar-2010 11:02:17] <ke4qqq> I'll try and add that documentation to the wikipage tonight

[09-Mar-2010 11:02:26] <mrayzenoss> so we can create a "Community Extended Monitoring Guide" of sorts

[09-Mar-2010 11:02:41] <ckrough> theacolyte: maybe http://bit.ly/bFswtD

[09-Mar-2010 11:03:07] <theacolyte> oh, that works, thanks

[09-Mar-2010 11:03:08] <mrayzenoss> probably around the beginning of April we'll start getting people to use it

[09-Mar-2010 11:03:55] <mrayzenoss> theacolyte: http://tr.im/core_2_5_1_tickets and http://tr.im/core_2_5_2_tickets

[09-Mar-2010 11:04:51] <theacolyte> cool, thanks :D

[09-Mar-2010 11:26:38] <sektorNBA> i wanna install squid zenpack. I am reading a doc http://tanso.net/zenoss/squid/ but i dunno what this mean: semanage port -a -t http_cache_port_t -p udp 3401

[09-Mar-2010 11:30:51] <RoAkSoAx> npmccallum, ping

[09-Mar-2010 11:31:00] <npmccallum> RoAkSoAx: pong

[09-Mar-2010 11:31:07] <RoAkSoAx> npmccallum, pm :)

[09-Mar-2010 11:32:18] <ke4qqq> sektorNBA: is selinux enabled?

[09-Mar-2010 11:33:06] <rmatte> sektorNBA: It literally explains what that means

[09-Mar-2010 11:33:28] <rmatte> sektorNBA: label port 3401/udp as a http_cache_port_t in the selinux-policy

[09-Mar-2010 11:33:33] <rmatte> semanage port -a -t http_cache_port_t -p udp 3401

[09-Mar-2010 11:33:56] <rmatte> semanage is the utility to manage selinux policies and the rest are the options

[09-Mar-2010 11:34:10] <rmatte> If you don't have selinux enabled you won't have access to said command (as ke4qqq stated)

[09-Mar-2010 11:39:37] <Kristopher424> Okay, okay, okay... I've follwed the admin guide on creating a mapped event.  My ? is... Under edit>Transform, I put: "evt.summary = evt.sysLocation Link Down"  Which should put the device location and the works Link Down afterward.  Does this change existing events or will I only see this on new incoming traps?

[09-Mar-2010 11:41:00] <mr-m> does anyone know off-hand how to globally update snmp communities in zenoss via zendmd?

[09-Mar-2010 11:47:27] <jb> hey matt?

[09-Mar-2010 11:47:34] <mrayzenoss> yeah?

[09-Mar-2010 11:49:10] <mr-m> for d in devices.Server.Linux.BroadWorks.devices(): d.zSnmpCommunity = "blah" commit() did it

[09-Mar-2010 11:51:12] <Kristopher424> Is the transform supposed to turn red when saved or does that mean syntax error?

[09-Mar-2010 12:01:26] <chkraise> syntax error

[09-Mar-2010 12:11:42] <sektorNBA> i installed zenpack to monitor squid. how do i use it?

[09-Mar-2010 12:12:28] <sektorNBA> nd bind the "Squid" template to your proxy servers

[09-Mar-2010 12:19:31] <rmatte> sektorNBA: this is really one of those RTFM moments: docs/DOC-4782#d0e3903

[09-Mar-2010 12:21:27] <rmatte> ...or don't

[09-Mar-2010 12:22:10] <forsberg> :)

[09-Mar-2010 12:37:56] <rmatte> There needs to be a bot in here that does a whole "Welcome to #zenoss, please check the Admin Guide (link) for the answer to your question before asking." message

[09-Mar-2010 12:38:08] <theacolyte> No

[09-Mar-2010 12:38:10] <theacolyte> No way

[09-Mar-2010 12:38:15] <rmatte> lol

[09-Mar-2010 12:38:17] <theacolyte> That would have precluded every question I've ever asked

[09-Mar-2010 12:38:18] <theacolyte> hahahaha

[09-Mar-2010 12:38:32] <rmatte> you've gotten better :P

[09-Mar-2010 12:38:38] <theacolyte> True!~

[09-Mar-2010 12:38:41] <rmatte> we have converted you

[09-Mar-2010 12:39:12] * rmatte chants "We are the borg, no RTFM is futile."

[09-Mar-2010 12:39:38] <chkraise> shooting from the hip is much more fun though

[09-Mar-2010 12:39:52] <rmatte> it can be, but it wastes people's time

[09-Mar-2010 12:40:16] <chkraise> well you fire away, then delete it all and RTFM and do it right

[09-Mar-2010 12:42:33] <forsberg> i agree with theacolyte

[09-Mar-2010 12:42:43] <rmatte> :)

[09-Mar-2010 12:46:34] <ckrough> I like to use you as guinea pigs when you dont rtfm. For things I dont want to test in my environment.

[09-Mar-2010 12:48:42] <rmatte> in the course of customizing the new version of OTRS I've fixed 2 of their bugs

[09-Mar-2010 12:49:06] <rmatte> ...and I don't even really know perl or dotiac, just using common sense

[09-Mar-2010 12:49:33] <rmatte> though perl reminds me a bit of python

[09-Mar-2010 12:55:32] <ckrough> http://www.voidspace.org.uk/python/weblog/images/chinese-python-poster.jpg

[09-Mar-2010 12:57:22] <mistich> anyone setup a script to execute on all critical events

[09-Mar-2010 13:01:56] <Mituc> hey guys, where should I look to figure out how to add zenoss users from a command line script, just like I add devices?

[09-Mar-2010 13:02:07] <rmatte> ok, whoever thought it would be a good idea to make ctrl-w a close shortcut for firefox should be shot

[09-Mar-2010 13:02:28] <rmatte> I'm so used to searching with ctrl-w in nano, but when I hit ctrl-w by accident instead of ctrl-f in firefox it closes

[09-Mar-2010 13:02:44] <mistich> anyone setup a script to execute on all critical events

[09-Mar-2010 13:02:44] <rmatte> garrr

[09-Mar-2010 13:02:53] <rmatte> mistich: that's easy to do

[09-Mar-2010 13:03:01] <mistich> not per event

[09-Mar-2010 13:03:10] <rmatte> hunh?

[09-Mar-2010 13:03:10] <mistich> but on all critical events

[09-Mar-2010 13:03:17] <rmatte> yeh, that's easy to do

[09-Mar-2010 13:03:24] <rmatte> Event Manager -> Commands

[09-Mar-2010 13:03:25] <mistich> well do tell sir :)

[09-Mar-2010 13:03:30] <rmatte> Add a command

[09-Mar-2010 13:03:44] <rmatte> Then configure the filter options for the command as you would for an alert

[09-Mar-2010 13:03:54] <rmatte> Make sure that a filter option for event severity is in place

[09-Mar-2010 13:03:56] <rmatte> and voila

[09-Mar-2010 13:04:18] <rmatte> the command will only execute on event count of 1 though

[09-Mar-2010 13:04:27] <rmatte> it won't execute again if the event count increases on the event

[09-Mar-2010 13:04:39] <mistich> filter options where

[09-Mar-2010 13:04:42] <rmatte> (It's by design, though it should be configurable in my opinion)

[09-Mar-2010 13:04:54] <rmatte> at the bottom of the command property window

[09-Mar-2010 13:04:58] <rmatte> where it says "Add filter"

[09-Mar-2010 13:05:07] <mistich> in settings->command

[09-Mar-2010 13:05:10] <rmatte> Select Severity from the dropdown

[09-Mar-2010 13:05:10] <forsberg> haha rmatte i do that alot

[09-Mar-2010 13:05:14] <forsberg> with the ctrl w

[09-Mar-2010 13:05:21] <rmatte> forsberg: yeh, it's so annoying

[09-Mar-2010 13:05:27] <rmatte> glad to know I'm not the only one

[09-Mar-2010 13:05:28] <rmatte> lol

[09-Mar-2010 13:05:34] <forsberg> enviromental braindamage

[09-Mar-2010 13:05:45] <forsberg> or something... didnt translate so well to english

[09-Mar-2010 13:05:50] <rmatte> hehe

[09-Mar-2010 13:06:07] <mistich> thanks found it

[09-Mar-2010 13:06:12] <rmatte> mistich: no problem

[09-Mar-2010 13:06:20] <rmatte> you can add multiple filters

[09-Mar-2010 13:06:23] <rmatte> if required

[09-Mar-2010 13:06:45] <mistich> thing is I set one up years ago so there is one there :)

[09-Mar-2010 13:07:20] <rmatte> ah

[09-Mar-2010 13:07:28] <mistich> old age I guess

[09-Mar-2010 13:07:34] <rmatte> well, if you don't use it you lose it

[09-Mar-2010 13:08:57] <mistich> oh yes

[09-Mar-2010 13:12:45] <Mituc> hey guys, where should I look to figure out how to add zenoss users from a command line script, just like I add devices? I'm looking at "/zport/portal_skins/zenmodel/dialog_addUser/manage_main" but I still can't figure out what URL to call..

[09-Mar-2010 13:19:58] <Mituc> ok, I'll take my chances with /zport/dmd/ZenUsers/dialog_addUser

[09-Mar-2010 13:28:06] <rmatte> that's probably it

[09-Mar-2010 13:28:20] <rmatte> although have you looked in zendmd?

[09-Mar-2010 13:28:36] <rmatte> I'm sure there's some function in there that can do it

[09-Mar-2010 13:29:03] <mr-m> does anyone know where there is documentation on translating traps and escalation to e-mail recipients?

[09-Mar-2010 13:30:19] <Mituc> rmatte: where should I look for zendmd?

[09-Mar-2010 13:35:00] <ckrough> Mituc: docs/DOC-3217

[09-Mar-2010 13:42:56] <Mituc> ckrough: thanks

[09-Mar-2010 13:58:36] <Mituc> ckrough, rmatte: ok, the thing is that I can't code in python (yet). /zport/dmd/ZenUsers/dialog_addUser doesn't seem to work either

[09-Mar-2010 13:58:47] <Mituc> I'll keep looking..

[09-Mar-2010 14:06:37] <rmatte> Mituc: well, coding in Python is really the only way to do it, unless you're using some sort of synthetic web transaction to do it

[09-Mar-2010 14:06:56] <rmatte> Mituc: I didn't even know Python until I started using Zenoss, it would be a good idea to start learning

[09-Mar-2010 14:07:43] <rmatte> Mituc: you are aware that Zenoss is capable of LDAP support, right?

[09-Mar-2010 14:08:03] <rmatte> Using LDAP would be more effective than adding users from the commandline

[09-Mar-2010 14:08:17] <rmatte> not sure what your end goal is though, so maybe commandline is better for what you intend to do

[09-Mar-2010 14:13:54] <ckrough> Mituc: you can use this to see the REST calls: https://addons.mozilla.org/en-US/firefox/addon/3829

[09-Mar-2010 14:14:03] <ckrough> if you use firefox that is

[09-Mar-2010 14:14:14] <Mituc> yep, I use firefox

[09-Mar-2010 14:15:01] <Mituc> indeed, probably moving users to ldap would be a smart move

[09-Mar-2010 14:29:45] <Kristopher424> Is it possible to not have the events count and actually show the each individual event?

[09-Mar-2010 14:31:33] <rmatte> Kristopher424: not really, event counts are designed to save space in the database

[09-Mar-2010 14:32:12] <rmatte> instead of adding a new entry for each event, it just updates the existing event with the new last time value and increases the count

[09-Mar-2010 14:32:34] <rmatte> it also keeps it neater

[09-Mar-2010 14:32:45] <rmatte> any particular reason why you need each to come in individually?

[09-Mar-2010 14:33:49] <Kristopher424> Solarwinds showed each trap's time individually, our engineers got pretty used to being able to see the exact times for each and every trap.

[09-Mar-2010 14:33:56] <Kristopher424> We miss a few timestamps this way.

[09-Mar-2010 14:43:12] <venturaville> hi bigegor

[09-Mar-2010 14:43:23] <venturaville> I almost have things working

[09-Mar-2010 14:43:37] <bigegor> venturaville: hi

[09-Mar-2010 14:43:44] <venturaville> oddly enough I can make it work perfectly from the command line, but it errors out in the test template screen

[09-Mar-2010 14:43:59] <venturaville> same args, url, properties, aliases (I tested using some debug print statements)

[09-Mar-2010 14:44:27] <venturaville> this is using the version you committed while I was gone (WBEMDataSource)

[09-Mar-2010 14:45:55] <venturaville> I get the "Error () received from query:.............  error whenever I do this in the template test screen

[09-Mar-2010 14:47:08] <bigegor> venturaville: wait, i will test it.

[09-Mar-2010 15:50:10] <sergeymasushko> hi guys. can someone guide my how to tune the zencommand daemon... I frequently receive fake alarms due to a command timeout...

[09-Mar-2010 15:53:31] <mrayzenoss> sergeymasushko: you've got a number of zProperties available to you

[09-Mar-2010 15:53:55] <mrayzenoss> zCommandCommandTimeout, zCommandCycleTime, zCommandLoginTimeout

[09-Mar-2010 15:54:44] <mrayzenoss> perhaps increase those and see if your timeouts go away?  I assume they're intermittent and you're running something that takes awhile?

[09-Mar-2010 15:55:28] <sergeymasushko> what about ?

[09-Mar-2010 15:55:28] <sergeymasushko> [zenoss@zenoss /usr/local/zenoss/etc]$ cat zencommand.conf

[09-Mar-2010 15:55:28] <sergeymasushko> parallel 3

[09-Mar-2010 15:56:14] <sergeymasushko> mrayzenoss: most of that alarms are timeouts of the mysql monitor zenpack

[09-Mar-2010 15:56:56] <mrayzenoss> You can probably bump those numbers up, most of our settings are pretty conservative

[09-Mar-2010 15:57:05] <venturaville> how can you force RRD creation?

[09-Mar-2010 15:57:23] <venturaville> e.g. there is no device in perf/Devices for it

[09-Mar-2010 15:57:56] <sergeymasushko> venturaville: remodel the device?

[09-Mar-2010 15:58:02] <venturaville> no dice

[09-Mar-2010 15:58:09] <venturaville> it is an ....interesting... device

[09-Mar-2010 15:58:24] <venturaville> e.g. not sshable, or snmpable

[09-Mar-2010 15:58:40] <venturaville> but I have some data points via WBEM allocated to it

[09-Mar-2010 15:58:46] <venturaville> and it isn't creating them

[09-Mar-2010 15:59:10] <sergeymasushko> venturaville: do you have binded template?

[09-Mar-2010 15:59:50] <venturaville> yep

[09-Mar-2010 15:59:51] <venturaville> no dice

[09-Mar-2010 15:59:57] <venturaville> I wonder if I unbind and rebind it

[09-Mar-2010 16:00:29] <sergeymasushko> venturaville: there datasources is of command type?

[09-Mar-2010 16:01:02] <venturaville> WBEM type...

[09-Mar-2010 16:07:19] <bigegor> venturaville: i have also one not sshable and snmpable WBEM only device.

[09-Mar-2010 16:07:35] <venturaville> still no dice on getting anything to show up in perf/Devices :-/

[09-Mar-2010 16:07:57] <venturaville> not even the top level hostname

[09-Mar-2010 16:11:06] <venturaville> I'll have to play with it tomorrow

[09-Mar-2010 16:11:08] <venturaville> later :-)

[09-Mar-2010 16:12:34] <venturaville> 2010-03-09 17:09:25,027 INFO zen.zenperfwbem: 0 devices processed (0 datapoints)

[09-Mar-2010 16:12:39] <venturaville> I'm guessing that is my problem......

[09-Mar-2010 16:18:54] <mrayzenoss> New Ganglia ZenPack: docs/DOC-5815

[09-Mar-2010 16:42:07] <gwb2352> grrrr... what/why does it take so long to Add IpService (OS Tab, IP Services).  The "OK" box sits there grayed out (doing something simple like tcp_00445 Ip Service Class)

[09-Mar-2010 17:03:35] <Kristopher424> If I don't want the "SNMP agent down" event to display, any way to remove it?

[09-Mar-2010 17:03:51] <mrayzenoss> you can acknowledge it

[09-Mar-2010 17:04:08] <mrayzenoss> you can remap the events or you can remove SNMP monitoring completely

[09-Mar-2010 17:04:53] <Kristopher424> I've tried to map the event and it says "1 event is not class /Unknown. Created 0 event mappings."

[09-Mar-2010 17:05:00] <Kristopher424> How do I remove snmp monitoring?

[09-Mar-2010 17:05:17] <mrayzenoss> do you want to remove it for a device or a class?

[09-Mar-2010 17:06:00] <Kristopher424> I need to monitor SNMP on the devices, but I get events that say "SNMP agent down" from /Status/Snmp, when the SNMP is working.. (might be momentary glitch in routing)

[09-Mar-2010 17:06:11] <mrayzenoss> under zProperties, set zSnmpMonitorIgnore to True

[09-Mar-2010 17:06:30] <mrayzenoss> oh, so you're just having timeouts?

[09-Mar-2010 17:06:35] <Kristopher424> Yeah.

[09-Mar-2010 17:06:48] <mrayzenoss> what kind of device?

[09-Mar-2010 17:07:53] <Kristopher424> It varies.  UPS's, Microwave Radios, WiMax units.

[09-Mar-2010 17:08:08] <mrayzenoss> which ones are timing out?

[09-Mar-2010 17:08:28] <Kristopher424> They all do, it's almost random.

[09-Mar-2010 17:08:51] <Kristopher424> I'll be getting SNMP replies and graphs, but Event Console will still give me an "SNMP agent down" event.

[09-Mar-2010 17:10:36] <Kristopher424> Does that zProperty need to be set on each device individually?

[09-Mar-2010 17:10:48] <mrayzenoss> that zProperty is to turn off SNMP

[09-Mar-2010 17:11:22] <Kristopher424> zSnmpMonitorIgnore would turn off all SNMP for that device? Or just ignore the Agent Down event?

[09-Mar-2010 17:11:28] <mrayzenoss> turn it off

[09-Mar-2010 17:11:50] <Kristopher424> Oh, definitely not what I need to do :)

[09-Mar-2010 17:12:41] <mrayzenoss> how many data points are you collecting?

[09-Mar-2010 17:12:54] <Kristopher424> 6, i believe.

[09-Mar-2010 17:14:32] <mrayzenoss> 6?  or 6K?

[09-Mar-2010 17:14:43] <Kristopher424> 6 per device.

[09-Mar-2010 17:15:02] <mrayzenoss> you can see your overall stats here: http://yourserver:8080/zport/dmd/Monitors/Performance/localhost/viewDaemonPerformance

[09-Mar-2010 17:16:48] <Kristopher424> Oh.. I see.  zenperfsnmp current is 5.18k

[09-Mar-2010 17:16:53] <mrayzenoss> you could also bump up the zSnmpTimeout a touch to see if that eases up your timeouts

[09-Mar-2010 17:17:01] <mrayzenoss> under zProperties

[09-Mar-2010 17:17:28] <mrayzenoss> if you were having gaps, I'd assume you were IO bound, but timeouts indicates the devices aren't responding fast enough

[09-Mar-2010 17:17:55] <Kristopher424> Which zProperties is this under?  /Events; /Status/Snmp, per device?

[09-Mar-2010 17:18:08] <mrayzenoss> you can set it per device class

[09-Mar-2010 17:18:38] <Kristopher424> Cool, the value is in seconds?

[09-Mar-2010 17:18:42] <mrayzenoss> if you go to your Event Console->Event History, you can search for those timeouts and get a feel for any particular devices or classes that seem most prone to timeouts

[09-Mar-2010 17:18:48] <mrayzenoss> yes

[09-Mar-2010 17:19:53] <Kristopher424> Cool, set.  Thanks.

[09-Mar-2010 17:22:33] forsberg is now known as fOrsberg

[09-Mar-2010 17:24:07] <mrayzenoss> New VMware ESXi ZenPack: docs/DOC-5816 Thanks ke4qqq!

[09-Mar-2010 17:24:40] <theacolyte> Ok, that one is worth checking out

[09-Mar-2010 17:24:51] <theacolyte> depending on the mechanism

[09-Mar-2010 17:25:05] <theacolyte> Oh that's nice

[09-Mar-2010 17:25:12] <theacolyte> I wouldn't really call that ESXi monitoring

[09-Mar-2010 17:25:23] <theacolyte> I would call  that ESXi Hardware Monitor

[09-Mar-2010 17:25:41] <mrayzenoss> I think he wants to add the Guests eventually

[09-Mar-2010 17:26:02] <theacolyte> Useful though

[09-Mar-2010 17:28:07] <ke4qqq> theacolyte: yeah it's possible to do guest monitoring, I just haven't yet - but the base code is there to be able to do it - representing it is another story.

[09-Mar-2010 17:28:46] <mrayzenoss> well, at least there are finally lots of examples for a tab+guests

[09-Mar-2010 17:28:49] <ke4qqq> but I figured I'd either use the virtual hosts zenpack or steal some of bigegors work.

[09-Mar-2010 17:29:26] <theacolyte> I'm just saying it's probably a better name for it

[09-Mar-2010 17:29:32] <theacolyte> Because otherwise people may think it's a VM monitor

[09-Mar-2010 17:29:36] <theacolyte> It's just semantics though

[09-Mar-2010 17:29:37] <ke4qqq> indeed

[09-Mar-2010 17:30:05] <theacolyte> If I wasn't using HP SIM I'd put it in

[09-Mar-2010 17:30:12] <theacolyte> And don't get me wrong, I hate HP SIM

[09-Mar-2010 17:30:30] <ke4qqq> I can completely empathize

[09-Mar-2010 17:30:52] <ke4qqq> though really - it needs lots of work to be where bigegor's hp zenpack is

[09-Mar-2010 17:31:21] <theacolyte> yeah, the only reason why I use SIM is it's easier to track ilo and do updates (and it does SSO for SMH/iLo)

[09-Mar-2010 17:31:35] <theacolyte> Beyond that I really hate it with a passion

[09-Mar-2010 17:31:39] <theacolyte> It's totally flakey

[09-Mar-2010 17:34:42] <ke4qqq> yep

[09-Mar-2010 17:34:52] <Kristopher424> Thanks again mrayzenoss.  I'm heading out.

[09-Mar-2010 17:43:24] <citrus2> i am getting zenoss all set up for the first time,  kinda fun putting all my devices in here and getting the ping alert working.  pretty easy stuff after i understood all the zenoss terms.

[09-Mar-2010 17:43:54] <citrus2> one question i have tho      how hard would it be to have zenoss monitor a website and make sure certain content is returned?

[09-Mar-2010 17:53:59] <ke4qqq> citrus2: super easy

[09-Mar-2010 17:54:24] <ke4qqq> let me give you a link

[09-Mar-2010 17:55:10] <ke4qqq> docs/DOC-5684

[09-Mar-2010 17:55:52] <ke4qqq> I sent you to the specific page- but you should read the entire chapter about http monitoring

[09-Mar-2010 18:05:20] <citrus2> thank you

[09-Mar-2010 18:17:18] <sunadmn> evening all

[09-Mar-2010 18:18:05] <sunadmn> have a question about installing the latest source bundle on Solaris

[09-Mar-2010 18:18:15] <sunadmn> anyone ever have luck doing that?

[09-Mar-2010 21:25:46] cwj_ is now known as cwj

[09-Mar-2010 21:28:48] <rayk3n> can anyone help me how can i enable alerts for load average

[09-Mar-2010 21:30:34] <rayk3n> everything seems to be monitoring fine except load average, i have my one alerting rules and everything shows on event console, except load average

[09-Mar-2010 21:39:16] <rayk3n> btw, i can see it perf graphs that a server increases its load avg, just dont know why it did not appear on my event console nor send me alerts.

[10-Mar-2010 00:00:30] [disconnected at Wed Mar 10 00:00:30 2010]

[10-Mar-2010 00:00:30] [connected at Wed Mar 10 00:00:30 2010]

[10-Mar-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[10-Mar-2010 01:29:17] <frozty_sa> oops

[10-Mar-2010 01:29:23] <frozty_sa> a few too many /win close's there

[10-Mar-2010 03:08:30] <Apachez> :)

[10-Mar-2010 03:15:20] <tehhobbit> morning Apachez

[10-Mar-2010 03:17:43] <Apachez> morrn'

[10-Mar-2010 03:31:56] salawank is now known as salax

[10-Mar-2010 03:38:29] <rayk3n> hi i've been doing some testing, can anyone telm me what's will value shall i put in max threshold of laLoadInt1, let say if i want to consider 1 as a high load avg.

[10-Mar-2010 03:41:24] <tehhobbit> rayk3n: be abit careful with alarms on load avg, I usually put the threshold on ssCpuRawWait and ssCpuUser instead

[10-Mar-2010 03:42:57] <rayk3n> tehhobbit can you give me an example? let say i want to consider 1 as a high load avg of a server

[10-Mar-2010 03:43:38] <tehhobbit> rayk3n: well load avg is a sum of alot of stuff thats why I dont use it

[10-Mar-2010 03:44:14] <tehhobbit> but I do get interested if a server has more than 50% io wait over the last 10 minutes

[10-Mar-2010 03:45:07] <rayk3n> hmmm, im talkng about simple load avg, 1min 5min 15min load average, which you can view by command "top" or "uptime",

[10-Mar-2010 03:45:21] <rayk3n> i know laLoadInt1 stands for 1 minute load average

[10-Mar-2010 03:47:56] <tehhobbit> yes, but what I'm trying to tell you is that an alarm on load avg isnt always a good thing. but sure if 1 is what you want på a alarm if its above 1 put 1000 in the max value field

[10-Mar-2010 03:48:37] <rayk3n> and what does 1200 stands for?

[10-Mar-2010 03:48:50] <rayk3n> that's what the default value of laLoadInt15

[10-Mar-2010 03:48:57] <rayk3n> sorry im a bit confused

[10-Mar-2010 03:51:00] <rayk3n> i have tried 1000 earlier , my server had a load avg of 2.08, 1.93, 1.60, nut somehow it does not show on my event console

[10-Mar-2010 03:54:07] <tehhobbit> those alarms are turned of by default have you checked that they are set to enabled = true ?

[10-Mar-2010 03:55:46] <rayk3n> yes, here is what i did. I added a threshold menu name "load avg" using laLoadInt1, set maxvalue to 1000 now, and enable = true, severity = warning. This is just for testing

[10-Mar-2010 03:55:52] <rayk3n> but still it does not show

[10-Mar-2010 03:55:59] <rayk3n> if i put a value of let say 300

[10-Mar-2010 03:56:18] <rayk3n> then an event will show that my server has a current value of 172 or something

[10-Mar-2010 03:58:32] <tehhobbit> which would be 0.3 afaik

[10-Mar-2010 04:00:10] <rayk3n> hmm but why does 1000 does not show anything which it should cause my server has at least 2 load average

[10-Mar-2010 04:04:30] <rayk3n> and why does if i set this to 300, alarm would say that my server has 172 current value

[10-Mar-2010 04:05:42] <tehhobbit> what value does a snmpwalk of tha oid give you

[10-Mar-2010 04:06:27] <rayk3n> can u tell me how can i generate that

[10-Mar-2010 04:07:04] <tehhobbit> bash# snmpwalk -v1 -c public <host> .1.3.6.1.4.1.2021.10.1.5.1

[10-Mar-2010 04:07:32] <rayk3n> ok wait

[10-Mar-2010 04:09:30] <rayk3n> UCD-SNMP-MIB::laLoadInt.1 = INTEGER: 157

[10-Mar-2010 04:09:59] <tehhobbit> and the load on the server is ?

[10-Mar-2010 04:10:33] <rayk3n> load average: 1.55, 1.25, 1.37

[10-Mar-2010 04:11:10] <tehhobbit> ok my bad used a 0 too much load 1 is 100

[10-Mar-2010 04:11:53] <rayk3n> ok ill try

[10-Mar-2010 04:12:06] <rayk3n> and what minvalue shall i put

[10-Mar-2010 04:12:15] <rayk3n> blank or 0

[10-Mar-2010 04:12:28] <AtuM> is it possible to create plugins in zenoss like in nagios?

[10-Mar-2010 04:18:41] <AtuM> I have some custom perl scripts that check Oracle services -pretty advanced stuff.. someone told me that Zenoss makes for a simpler and better interface

[10-Mar-2010 04:19:41] <tehhobbit> AtuM: look in /opt/zenoss/libexec and you'll find the standard nagios plugins

[10-Mar-2010 04:20:13] <tehhobbit> so yes porting it shouldnt be to much of problem, take the check_ntp as and example and work from that

[10-Mar-2010 04:22:26] <rayk3n> ei tehobbit thank i think i got, just one last question, what does Escalate Count stands for

[10-Mar-2010 04:24:50] <tehhobbit> havent used it , but basicly its used if there are X alarms without acc start sending the alarms to the person with the next paygrade

[10-Mar-2010 04:24:58] fOrsberg is now known as forsberg

[10-Mar-2010 04:27:01] <rayk3n> thanks mate

[10-Mar-2010 04:27:07] <rayk3n> that really helped

[10-Mar-2010 04:27:25] <tehhobbit> so dont tell your boss about it :P

[10-Mar-2010 04:29:24] <rayk3n> lol

[10-Mar-2010 04:41:12] <AtuM> tehhobbit, thanx.. will check it out

[10-Mar-2010 05:21:53] <swift> hi guys, i want to install zenoss core on my solaris10 x86 machine which has splunk... also, i want zenoss to monitor my windows, solaris and linux servers. Alerting must be possible, cananybody direct me to  a guide?

[10-Mar-2010 05:22:46] <tehhobbit> community/documentation/official_documentation/installation-guide/2.5.2

[10-Mar-2010 05:25:16] <swift> tehhobbit, il have to install zenoss core from source on solaris 10 x86?

[10-Mar-2010 05:30:59] <swift> tehhobbit, will i need to install zenoss agents on windows servers?

[10-Mar-2010 05:31:46] <tehhobbit> no zenoss uses wmi / snmp for monitoring

[10-Mar-2010 05:32:20] <swift> tehhobbit, so i must have the snmp port on my srvers unused?

[10-Mar-2010 05:32:32] <tehhobbit> and yes you need to install from source on solaris

[10-Mar-2010 05:33:07] <tehhobbit> swift: quite the other way , you must have a snmpagent listening on that port

[10-Mar-2010 05:33:37] <swift> i need to install some snmp agent on the servers?

[10-Mar-2010 05:34:11] <tehhobbit> no the one shipped with windows will do just fine

[10-Mar-2010 05:36:04] <swift> tehhobbit, snmp agent must be listening on UDP 161 right?

[10-Mar-2010 05:36:24] <tehhobbit> yes and allow connections from the monitor host

[10-Mar-2010 05:36:31] <swift> ok

[10-Mar-2010 05:36:57] <swift> also, i see that mysql on my sol server is Distro 4.0...which is less then the prerequisite for zenoss

[10-Mar-2010 06:11:44] <onexs> hi everybody

[10-Mar-2010 06:14:19] <tehhobbit> hejhej

[10-Mar-2010 06:17:43] <onexs> is there a plugin for smokeping for zenoss

[10-Mar-2010 06:23:41] <tehhobbit> smokeping ?

[10-Mar-2010 06:25:30] <onexs> yes

[10-Mar-2010 06:31:02] <Rolfs> i think it's only a fping plugin

[10-Mar-2010 06:31:28] <Rolfs> looking a little similar to the smokeping

[10-Mar-2010 06:35:50] <Troubadix09> onexs: i youse this one as command datasource docs/DOC-2513

[10-Mar-2010 06:35:59] <Troubadix09> youse=use

[10-Mar-2010 06:38:24] <onexs> thanks you

[10-Mar-2010 07:29:12] <swift> hi guys, zenoss uses snmp agent on the client side to send logs to the server... what if snmp service is being used by some other service?

[10-Mar-2010 07:29:16] <swift> is that a problem?

[10-Mar-2010 07:29:43] <swift> please advise

[10-Mar-2010 07:30:51] <swift> also, installing zenoss on the windows platform involves virtual appliance, so, will I have to keep this virtual appliance(window) running all the time on the windows machine??.. please advise

[10-Mar-2010 07:39:54] <sektorNBA> i am creating some new datagraphs. How do i add an alias?

[10-Mar-2010 07:40:04] <sektorNBA> like Aliases

[10-Mar-2010 07:40:04] <sektorNBA> memoryAvailable__bytes      |      1024,*

[10-Mar-2010 08:10:12] <sektorNBA> got it.

[10-Mar-2010 08:11:15] <sektorNBA> anyone around?

[10-Mar-2010 08:42:14] <chachan> Howdy, I'm having an issue with munin, a part of my munin-update.log is here: http://pastebin.com/hd6V4VBt

[10-Mar-2010 08:42:29] <chachan> can anybody help me with this?, please

[10-Mar-2010 08:45:33] forsberg is now known as fOrsberg

[10-Mar-2010 08:47:31] fOrsberg is now known as forsberg

[10-Mar-2010 10:12:43] forsberg is now known as fOrsberg

[10-Mar-2010 10:43:46] fOrsberg is now known as forsberg

[10-Mar-2010 11:47:25] <Kristopher424> Hmm.  Zenoss Events shows SNMP is down on all Trango Microwave Radios.  Graphing has stopped at 1pm yesterday but the Current value isn't NaN it's an actual value.

[10-Mar-2010 12:50:33] <willwh> forsberg: may I suggest http://quassel-irc.org/

[10-Mar-2010 12:51:58] <RoAkSoAx> npmccallum, mrayzenoss my presentation about zenoss went very well last night

[10-Mar-2010 12:52:09] <npmccallum> awesome!

[10-Mar-2010 12:52:12] <mrayzenoss> glad to hear it

[10-Mar-2010 12:52:23] <npmccallum> did you use lots of marketing speak?

[10-Mar-2010 12:52:45] <RoAkSoAx> npmccallum, off course!! My professor was very interested in sensor management though

[10-Mar-2010 12:53:03] <npmccallum> yes, we support some sensor systems :)

[10-Mar-2010 12:53:20] <npmccallum> most others are just a matter of setting the right oids

[10-Mar-2010 12:53:45] <RoAkSoAx> do you guys have sensors than can forward information to other sensors or stuff like that? since they are doing research on that

[10-Mar-2010 12:53:51] <RoAkSoAx> that's why she was interested

[10-Mar-2010 12:54:15] <forsberg> willwh sorry ?

[10-Mar-2010 12:54:46] <RoAkSoAx> npmccallum, or in such case, do you have a list of sensors that you support and what can be done with them?

[10-Mar-2010 12:55:10] <npmccallum> RoAkSoAx: I'm not really understanding the question

[10-Mar-2010 12:55:16] * forsberg slaps willwh around a bit with a large trout

[10-Mar-2010 12:55:22] <npmccallum> we basically gather data from snmp from sensors

[10-Mar-2010 12:55:31] <npmccallum> I'm not sure how you would "pass this to other sensors"

[10-Mar-2010 12:56:21] <RoAkSoAx> npmccallum, i don't either.. and I think that';s why she's doing research on that

[10-Mar-2010 12:57:24] <npmccallum> I'm not sure that really makes sense

[10-Mar-2010 12:57:39] <npmccallum> sensors are RO devices

[10-Mar-2010 12:58:38] <RoAkSoAx> yeah

[10-Mar-2010 12:58:47] <RoAkSoAx> npmccallum, anyways, do you have a list of sensors you support?

[10-Mar-2010 12:59:12] <npmccallum> technically we support no sensors directly

[10-Mar-2010 12:59:19] <npmccallum> we support sensor controllers

[10-Mar-2010 12:59:24] <npmccallum> like Falcon

[10-Mar-2010 12:59:50] <npmccallum> the Falcon zenpack is enterprise though

[10-Mar-2010 13:00:40] <RoAkSoAx> ok I'll look into how that works and tell her if she wants to test it, she needs enterprise

[10-Mar-2010 13:01:30] <RoAkSoAx> thanks for the info

[10-Mar-2010 13:02:17] <npmccallum> np

[10-Mar-2010 13:02:54] <mrayzenoss> I don't know how complex the Falcon ZenPack is, but writing a Community one might not be too hard either

[10-Mar-2010 13:03:06] <Kristopher424> Is upgrading from Community to Enterprise a full re-installation process?

[10-Mar-2010 13:04:26] <mrayzenoss> no

[10-Mar-2010 13:04:42] <mrayzenoss> you can get assistance migrating

[10-Mar-2010 13:04:51] <Kristopher424> Oh, cool.

[10-Mar-2010 13:05:31] <npmccallum> Falcon is pretty complex because the device sucks

[10-Mar-2010 13:05:40] <npmccallum> but most sensor controllers are fairly sane

[10-Mar-2010 13:09:13] <willwh> forsberg: why use a bouncer? :)

[10-Mar-2010 13:09:38] <willwh> screen & irssi is awesome, but, I've been using Quassel of late

[10-Mar-2010 13:09:39] <willwh> it rocks

[10-Mar-2010 13:12:48] <RoAkSoAx> npmccallum, If i understand correctly, you connect various sensors to a sensor controller, and zenoss reads snmp info from that controller, correct?

[10-Mar-2010 13:13:23] <npmccallum> yes

[10-Mar-2010 13:13:26] <npmccallum> that is correct

[10-Mar-2010 13:13:33] <npmccallum> that is pretty much the way that all sensors work

[10-Mar-2010 13:14:51] <RoAkSoAx> npmccallum, yeah iw ould guess so :)

[10-Mar-2010 13:15:04] <forsberg> why do you care? anyways i love a bouncer, im changing between alot of nets and lot clients etc, bnc is perfect for me and my friends to manage vhosts and shit and never care about loosing nicks and other stuff

[10-Mar-2010 13:15:56] <RoAkSoAx> npmccallum, and I'm guessing that sensor controllers can't share info between other sensor controllers or send snmp info to other controllers

[10-Mar-2010 13:16:10] <npmccallum> RoAkSoAx: I can't see what the point of doing that would be

[10-Mar-2010 13:16:18] <npmccallum> it smells of layer violation

[10-Mar-2010 13:16:53] <RoAkSoAx> npmccallum, all I know is that they are doing research in that

[10-Mar-2010 13:17:07] <npmccallum> sounds like stupid research :)

[10-Mar-2010 13:18:34] <RoAkSoAx> hahaha

[10-Mar-2010 13:19:00] <willwh> forsberg: I don't care - just mentioning :)

[10-Mar-2010 13:19:32] <forsberg> willwh: k

[10-Mar-2010 13:27:39] <Kristopher424> Just confirming- It's pretty impossible for Zenoss to stop getting SNMP replies from one brand of device, no matter where it's location on the network?  We have a bunch of Trango radios that have all decided to stop polling graphs... but Dragonwave and the others work fine.

[10-Mar-2010 13:30:24] <mrayzenoss> it's probably more likely that something is wrong with the devices

[10-Mar-2010 13:30:30] <mrayzenoss> perhaps bad firmware or something like that?

[10-Mar-2010 13:31:15] <mrayzenoss> are they all the same type of devices?  and they've all given up at the same time?

[10-Mar-2010 13:52:00] <Troubadix09> mrayzenoss: thanks for trying to reproduce my error in ticket 6262

[10-Mar-2010 13:52:30] <mrayzenoss> sorry I couldn't reproduce it

[10-Mar-2010 13:52:37] <mrayzenoss> we haven't heard anyone else with that issue

[10-Mar-2010 13:56:39] <Troubadix09> mrayzenoss: i would try a forum thread. have you any idea where to start with this issue? Python or MySql ?

[10-Mar-2010 14:00:32] <mrayzenoss> Troubadix09: I'd be curious if a backup from your upgraded 2.5.2 appliance would work if you restored it on a fresh 2.5.2 appliance

[10-Mar-2010 14:00:53] <mrayzenoss> that would let us know if this was a bad upgrade or bad data

[10-Mar-2010 14:02:29] <Troubadix09> mrayzenoss: oh, good idea, i will tryout that and give feedback.

[10-Mar-2010 14:03:48] <Troubadix09> but first i go back to a snapshot before upgrade and try upgrade again, maybe something broken thrue the upgrade

[10-Mar-2010 14:04:07] <mrayzenoss> yeah, VMs are nice that way

[10-Mar-2010 14:07:41] <Troubadix09> mrayzenoss: and on a test-server with esxi i installed a fresh centOS-VM, there i would install a fresh zenoss 2.5.2 and then move my appliance-backup and would see if it works

[10-Mar-2010 14:09:36] <mr-m> anyone have creative ways of customizing e-mails for specific classes of devices/events? i'm trying to display specific snmp oid's in an email based on certain trap conditions...

[10-Mar-2010 14:27:02] forsberg is now known as fOrsberg

[10-Mar-2010 14:28:52] <lthrasher> RoAkSoAx: where do you go to school?

[10-Mar-2010 14:30:06] <RoAkSoAx> lthrasher, Florida International University

[10-Mar-2010 16:16:08] RoAk is now known as RoAkSoAx

[10-Mar-2010 20:23:56] <mloven> can anyone on right now let me know if they've gotten a date range to work with reportmail?

[10-Mar-2010 21:17:07] <swift> hi guys, does zenoss core capture event viewer logs?.. is it able to filter these?

[11-Mar-2010 00:00:30] [disconnected at Thu Mar 11 00:00:30 2010]

[11-Mar-2010 00:00:30] [connected at Thu Mar 11 00:00:30 2010]

[11-Mar-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[11-Mar-2010 00:45:41] Bryanstein_ is now known as Bryanstein

[11-Mar-2010 01:08:02] fOrsberg is now known as forsberg

[11-Mar-2010 01:46:47] <Egyptian> morning all

[11-Mar-2010 01:47:03] <Egyptian> i am really on a thorn here: delta rpms for zenoss

[11-Mar-2010 01:47:28] <Egyptian> 1 68mb download when i can have something much smaller via delta rpms

[11-Mar-2010 01:47:32] <Egyptian> http://fedoraproject.org/wiki/QA:Testcase_YumPresto_createrepo

[11-Mar-2010 06:56:08] <Weetos> hey there - Is there any way to list currently logged in users in zenoss ?

[11-Mar-2010 06:58:06] <Weetos> I can list connections thanks to "netstat -tn | grep 8080" but this obsviously don't allow to know which users those connections belongs to

[11-Mar-2010 08:05:35] mrayzenoss1 is now known as mrayzenoss

[11-Mar-2010 08:33:11] <etank> i have the following transform that is meant to move events to history if they happen between 5:30am and 9:00am. it does not seem to be working though http://fpaste.org/EfVu/

[11-Mar-2010 08:33:18] <etank> what am i doing wrong in it

[11-Mar-2010 08:55:09] <ckrough> anyone know how to enable monitoring of a process for a whole class?

[11-Mar-2010 09:36:01] <rmatte> ckrough: you can't do it per class, it's all or nothing

[11-Mar-2010 09:36:17] <rmatte> if you enable it in the Process section then it's enabled system-wide whenever detected

[11-Mar-2010 09:37:47] <rmatte> etank: that's not going to work, you can't use >= operators on time tags, only on actual integers

[11-Mar-2010 09:37:52] <jb> hm, think im going to redo this squid zenpack

[11-Mar-2010 09:37:57] <jb> eggify it, and modify/add some data

[11-Mar-2010 09:38:07] <rmatte> etank: you would need to convert the date/time to epoch time (which is how it's actually stored in mysql for each event)

[11-Mar-2010 09:38:16] <rmatte> etank: then do the calculation based on the epoch time values

[11-Mar-2010 09:39:23] <mrayzenoss> jb: excellent.  The .Zip zenpacks are deprecated so I expect the useful ones will get repackaged

[11-Mar-2010 09:39:32] <mrayzenoss> I'm redoing the NetBotz one

[11-Mar-2010 09:39:49] <jb> i'll let you know once its complete

[11-Mar-2010 09:39:54] <mrayzenoss> thanks

[11-Mar-2010 09:40:17] <rmatte> the 3 zip packs I'm using are... APC UPS PowerNet, Fortigate, and HPTemperatureSensors

[11-Mar-2010 09:40:25] <jb> hey, does anyone remember the ticket/patch for zenreportmail?

[11-Mar-2010 09:40:42] <jb> that fixes the "space in subject, etc" issue?

[11-Mar-2010 09:40:47] <jb> my 2.5.2 upgrade re-broke it :/

[11-Mar-2010 09:41:12] <rmatte> jb: which one?  there are several

[11-Mar-2010 09:41:19] <rmatte> what issue are you trying to patch specifically?

[11-Mar-2010 09:41:26] <jb> one in particular that I added for 2.5.1

[11-Mar-2010 09:42:14] <rmatte> all I see is this

[11-Mar-2010 09:42:15] <rmatte> http://dev.zenoss.org/trac/ticket/5949

[11-Mar-2010 09:42:22] <rmatte> I don't see anything else relating to reportmail for 2.5.1

[11-Mar-2010 09:42:48] <rmatte> any idea what the title was?

[11-Mar-2010 09:43:27] <jb> one sec

[11-Mar-2010 09:43:35] <rmatte> you are talking about a trac ticket I assume?

[11-Mar-2010 09:43:39] <rmatte> not some internal support ticket?

[11-Mar-2010 09:43:40] <jb> yes

[11-Mar-2010 09:43:42] <rmatte> k

[11-Mar-2010 09:44:08] <rmatte> also, what trac account did you log it under?

[11-Mar-2010 09:44:11] <rmatte> what username did you use?

[11-Mar-2010 09:44:48] <jb> --subject="FMS DC Notification: Devices Down Report" \

[11-Mar-2010 09:44:55] <jb> i didnt log the ticket

[11-Mar-2010 09:44:59] <jb> chet did

[11-Mar-2010 09:45:04] <jb> the bug causes subject to be "FMS"

[11-Mar-2010 09:45:05] <rmatte> k, that helps...

[11-Mar-2010 09:45:21] <jb> as well as a few other things

[11-Mar-2010 09:46:43] <rmatte> do you know around what day he logged it?

[11-Mar-2010 09:49:04] <rmatte> The only thing that I've found that is even remotely close to what you're describing is: http://dev.zenoss.org/trac/ticket/5787

[11-Mar-2010 09:50:37] <rmatte> so either that's it, or he didn't actually log it

[11-Mar-2010 09:52:41] <rmatte> have you seen the ticket before?

[11-Mar-2010 09:52:48] <rmatte> or do you just assume that it exists?

[11-Mar-2010 09:55:36] <mloven> anyone know if it's possible to specify a date range in reportmail?

[11-Mar-2010 09:56:51] <ckrough> mloven: you would have to make a custom report in zope/zendmd and then call that with reportmail

[11-Mar-2010 09:58:41] <mloven> so no then.  :)    Understood.     Thanks ckrough.

[11-Mar-2010 09:59:00] <rmatte> mloven: did you think to check the admin guide? docs/DOC-4773#d0e9699

[11-Mar-2010 09:59:06] <ckrough> mloven: from my experience, no, but it maybe someone else knows a way

[11-Mar-2010 09:59:06] <rmatte> those are all of the options it supports

[11-Mar-2010 09:59:12] <rmatte> if it's not listed there, it's not supported

[11-Mar-2010 10:00:18] <mloven> understood.  Thanks rmatte.

[11-Mar-2010 10:00:19] <rmatte> once again, the admin guide covers 90% of the things anyone is likely to ask in here :P

[11-Mar-2010 10:00:22] <rmatte> np

[11-Mar-2010 10:00:23] <rmatte> ;)

[11-Mar-2010 10:00:40] <mloven> it's that 10% that worries me.

[11-Mar-2010 10:00:57] <rmatte> yeh, but you should always check first

[11-Mar-2010 10:01:29] <rmatte> that 10% is generally just questions about how to code transforms or whatever which can never really be fully documented as there are infinite possibilities

[11-Mar-2010 10:03:31] <rmatte> now time to try and figure out why my development version of OTRS is throwing SQL errors every few seconds

[11-Mar-2010 10:04:20] <rmatte> ah, it's erroring every time a page is loaded, hmmm

[11-Mar-2010 10:31:24] <ckrough> which modeling plugin handles OS Process mapping

[11-Mar-2010 10:39:33] <venturaville> mrayzenoss: is there any way to get zenhubworker to NOT use /tmp ......

[11-Mar-2010 10:41:04] <mrayzenoss> venturaville: I don't know where that's defined, but my assumption is we're using the TMPDIR shell variable

[11-Mar-2010 10:41:18] <mrayzenoss> or at least, I'd hope we are

[11-Mar-2010 10:41:23] <venturaville> apparently when zenhub starts up it hands them spots in /tmp

[11-Mar-2010 10:48:56] <venturaville> you are apparently calling mkstemp() to create the temp files

[11-Mar-2010 11:00:33] <venturaville> apparently it honors TMPDIR

[11-Mar-2010 11:05:03] <venturaville> it looks as though renderserver uses /tmp as well, but I don't see a way to override it without touching code

[11-Mar-2010 11:13:27] <venturaville> def __init__(self, id, tmpdir = '/tmp/renderserver', cachetimeout=300):

[11-Mar-2010 11:13:34] <venturaville> zenrender.py never sets the tmpdir, and provides no option to overwrite it without touching base code

[11-Mar-2010 11:17:34] <mrayzenoss> if you change that to '$TMPDIR/renderserver'

[11-Mar-2010 11:17:41] <mrayzenoss> will it work?

[11-Mar-2010 11:18:28] <mrayzenoss> venturaville: if it does feel free to open a ticket with a patch and we'll see about getting that fixed in the next release

[11-Mar-2010 11:35:25] <jb> anyone running IIS7/7.5?

[11-Mar-2010 11:35:32] <jb> im hoping the perf counters are the same

[11-Mar-2010 11:35:38] <jb> (as they were for IIS6)

[11-Mar-2010 11:36:13] <etank> thanks rmatte. i will look into how to do that. (epoch time)

[11-Mar-2010 11:43:42] <rmatte> np

[11-Mar-2010 11:53:05] <venturaville> tmpdir is set by zenrender.py, and it never consults anything to set it... so the patch would probably be a little more complicated :-/

[11-Mar-2010 11:53:27] <venturaville> I'll see what I can find

[11-Mar-2010 11:55:50] <venturaville> apparently import tempfile ................. tempfile.gettempdir() ....works

[11-Mar-2010 12:05:21] <jb> hmm

[11-Mar-2010 12:05:26] <jb> question regarding graphs

[11-Mar-2010 12:05:43] <jb> is it possible to turn a SNMP COUNTER into a "per second" type of value?

[11-Mar-2010 12:05:51] <jb> with 5 minute zenperfsnmp polls?

[11-Mar-2010 12:08:06] <venturaville> it does that by default?

[11-Mar-2010 12:09:40] <jb> really?

[11-Mar-2010 12:09:55] <venturaville> mrayzenoss: you should probably add notes about TMPDIR to the docs for when adding zenhubworkers...

[11-Mar-2010 12:10:00] <venturaville> jb: I believe so

[11-Mar-2010 12:10:36] <venturaville> or I have been doing it wrong all this time :-)

[11-Mar-2010 12:11:26] <venturaville> if you fetch it via getRRDValue it will still be the counter value

[11-Mar-2010 12:11:38] <venturaville> but when it goes to graph it it will calculate it out as a per second value

[11-Mar-2010 12:12:09] <jb> hmmm

[11-Mar-2010 12:12:11] <jb> let me check

[11-Mar-2010 12:13:06] <jb> i cant really tell what is being graphed

[11-Mar-2010 12:13:19] <venturaville> you can use zendmd

[11-Mar-2010 12:13:24] <venturaville> d = find('devicename')

[11-Mar-2010 12:13:29] <venturaville> d.getRRDDataPoints()

[11-Mar-2010 12:13:30] <mrayzenoss> venturaville: when you open a ticket with that patch, just put a note about doc'ing it and we'll get it into the documentation queue

[11-Mar-2010 12:13:39] <venturaville> then do d.getRRDValue('dpname')

[11-Mar-2010 12:13:49] <venturaville> mrayzenoss: k

[11-Mar-2010 12:15:39] <jb> http://thunder.jbdesign.net/~jbaird/zenoss/squid.jpg

[11-Mar-2010 12:15:49] <jb> im looking at the total requests

[11-Mar-2010 12:15:56] <jb> which is a COUNTER that increments

[11-Mar-2010 12:16:08] <venturaville> looks fine

[11-Mar-2010 12:16:15] <jb> so what is that telling me?

[11-Mar-2010 12:16:15] <venturaville> you are getting < 1 request / sec

[11-Mar-2010 12:16:19] <venturaville> m = milli

[11-Mar-2010 12:16:59] <jb> ahh

[11-Mar-2010 12:17:08] <venturaville> that one always gets me too

[11-Mar-2010 12:18:43] <jb> yeah

[11-Mar-2010 12:18:43] <jb> wow

[11-Mar-2010 12:18:47] <forsberg> use varnish not squid god damn it

[11-Mar-2010 12:18:49] <jb> i had no idea it did this.

[11-Mar-2010 12:18:50] <forsberg> ^^

[11-Mar-2010 12:18:56] <jb> can't use varnish here..

[11-Mar-2010 12:20:07] <venturaville> squid is alright .. but I wouldn't use it for reverse proxying these days

[11-Mar-2010 12:20:20] <forsberg> you can maybe * the number to get a better measurement number

[11-Mar-2010 12:21:21] <venturaville> what does that do?

[11-Mar-2010 12:21:44] <jb> im not using it for reverse

[11-Mar-2010 12:22:22] <jb> ok, so this basically says "poll every minute, and give me the average over each second during that 5 minutes" ?

[11-Mar-2010 12:22:26] <jb> sort of confused.

[11-Mar-2010 12:22:43] <forsberg> if you use derive ?

[11-Mar-2010 12:23:07] <jb> its a COUNTER

[11-Mar-2010 12:23:12] <forsberg> yeah same, but yes

[11-Mar-2010 12:23:30] <forsberg> so if you want pr 5 minute, you multiply up again with 300

[11-Mar-2010 12:23:30] <forsberg> 300,*

[11-Mar-2010 12:23:35] <forsberg> in RPN

[11-Mar-2010 12:23:54] <jb> i basically have a COUNTER that only increments

[11-Mar-2010 12:23:55] <jb> i'd like to get that in "per second" format

[11-Mar-2010 12:24:00] <jb> and apparently, zenoss is already doing this

[11-Mar-2010 12:24:07] <forsberg> not zenoss

[11-Mar-2010 12:24:11] <jb> well, RRD

[11-Mar-2010 12:24:16] <forsberg> yup :)

[11-Mar-2010 12:24:19] <jb> i poll this value every 5 minutes with zenoss though

[11-Mar-2010 12:24:31] <jb> so im having a hard time understanding how this value would be accurate.. but maybe I am over-thinking it

[11-Mar-2010 12:24:49] <forsberg> well when the number is so low it doesnt give much sense

[11-Mar-2010 12:25:09] <jb> yeah, like I know i am doing > 1 req/sec

[11-Mar-2010 12:25:14] <forsberg> i had the same problem, and i had to .... hmmm make it pr 5 minute to give any usefull meaning

[11-Mar-2010 12:25:46] <jb> i may have to do the same thing

[11-Mar-2010 12:25:53] <jb> per/sec just doesn't make any sense

[11-Mar-2010 12:26:07] <jb> i have used the 300,* RPN before to do this

[11-Mar-2010 12:26:09] <jb> with BIND stats, etc

[11-Mar-2010 12:26:13] <forsberg> at work it made perfect sense for our bind named server, but when i moved the template to home, it didnt

[11-Mar-2010 12:26:13] <forsberg> :)

[11-Mar-2010 12:26:27] <forsberg> excactly

[11-Mar-2010 12:26:36] <jb> yeah, 300,* on all of my BIND stuff

[11-Mar-2010 12:26:41] <jb> is that what you did?

[11-Mar-2010 12:26:44] <forsberg> yup

[11-Mar-2010 12:27:58] <jb> ok

[11-Mar-2010 12:28:02] <jb> think im going to the same

[11-Mar-2010 12:28:14] <jb> i can do the math to break it down per/sec if I really want, I guess..

[11-Mar-2010 12:28:22] <jb> value / 5 / 60

[11-Mar-2010 12:31:48] <venturaville> mrayzenoss: http://dev.zenoss.org/trac/ticket/6278

[11-Mar-2010 12:57:16] <mrayzenoss> venturaville: thanks

[11-Mar-2010 13:07:08] <jb> weird

[11-Mar-2010 13:07:13] <jb> i can't add this one datasource

[11-Mar-2010 13:07:20] <jb> everytime I try, zenoss errors..

[11-Mar-2010 13:07:25] <jb> right after typing the name of it

[11-Mar-2010 13:08:46] <jb> http://pastebin.com/QLSeVDea

[11-Mar-2010 13:08:49] <jb> that make any sense?

[11-Mar-2010 13:10:14] <venturaville> what are you naming it?

[11-Mar-2010 13:11:11] <jb> cacheCurrentFileDescrCnt

[11-Mar-2010 13:11:22] <venturaville> seems safe enough

[11-Mar-2010 13:11:26] <jb> yeah

[11-Mar-2010 13:12:02] <jb> oh well.. works with any other name

[11-Mar-2010 13:12:11] <venturaville> maybe it is already used?

[11-Mar-2010 13:12:32] <jb> nope

[11-Mar-2010 13:12:33] <venturaville> try cacheCurrentFileDescrCount

[11-Mar-2010 13:13:44] <jb> worked

[11-Mar-2010 13:33:41] <jb> ugh finally done with those graphs

[11-Mar-2010 13:33:48] <jb> now I just need to make sure they look right :)

[11-Mar-2010 14:00:55] forsberg is now known as fOrsberg

[11-Mar-2010 14:04:28] mrayzenoss1 is now known as mrayzenoss

[11-Mar-2010 15:02:41] <cgibbons> hmmm

[11-Mar-2010 16:13:07] <Apachez> http://www.lowbird.com/data/images/2010/03/77257814-2063-11de-9580-001ec929e6c4-31.png

[12-Mar-2010 00:00:30] [disconnected at Fri Mar 12 00:00:30 2010]

[12-Mar-2010 00:00:30] [connected at Fri Mar 12 00:00:30 2010]

[12-Mar-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[12-Mar-2010 01:29:22] fOrsberg is now known as forsberg

[12-Mar-2010 07:50:46] forsberg is now known as fOrsberg

[12-Mar-2010 08:24:04] <rmatte> good morning all

[12-Mar-2010 08:25:59] <tehhobbit> afternoon

[12-Mar-2010 08:26:36] <rmatte> that too\

[12-Mar-2010 08:36:29] <jb> what is "3.0G" as a graph value?

[12-Mar-2010 08:37:54] <venturaville> G = giga

[12-Mar-2010 08:38:17] <jb> ok

[12-Mar-2010 08:38:25] <jb> i dont know why my graphs have spikes

[12-Mar-2010 08:38:36] <jb> im doing COUNTERs with an RPN of 300,*

[12-Mar-2010 08:38:39] <jb> for these 5 minute graphs

[12-Mar-2010 08:38:51] <jb> and in the beginning, some have spikes into ~3.0G

[12-Mar-2010 08:39:26] <venturaville> I've had that happen when I first add a new data point

[12-Mar-2010 08:39:40] <jb> yeah

[12-Mar-2010 08:39:45] <jb> its annoying because it throws off the rest of the graph

[12-Mar-2010 08:39:48] <venturaville> I think there is a way to put a clamp on those, but I am not an RRD expert

[12-Mar-2010 08:40:28] <venturaville> according to one of the RRD pages, you can use DERIVE with rrd-min=0 instead of counter

[12-Mar-2010 08:40:52] <jb> ah

[12-Mar-2010 08:40:54] <venturaville> should be "good enough" for most things

[12-Mar-2010 09:25:11] <jb> bleh RRD pisses me off

[12-Mar-2010 09:26:23] <san4el> hih

[12-Mar-2010 09:26:24] <venturaville> it annoys most everyone

[12-Mar-2010 09:26:34] <jb> i guess i'll try derive

[12-Mar-2010 09:26:36] <venturaville> it isn't much fun to deal with directly

[12-Mar-2010 09:26:39] <jb> im still getting spikes

[12-Mar-2010 09:26:49] <san4el> my problem is russian language in monitored windows systems

[12-Mar-2010 09:26:49] <venturaville> make sure you are recreating your RRD files......

[12-Mar-2010 09:27:00] <jb> what do you mean?

[12-Mar-2010 09:27:08] <jb> like after I change them, delete the RRDs?

[12-Mar-2010 09:27:43] <venturaville> to be on the safe side

[12-Mar-2010 09:27:55] <jb> yeah

[12-Mar-2010 09:32:18] <jb> ok, we'll see if that fixes things

[12-Mar-2010 09:48:13] <venturaville> ok odd question ..what would cause something to say 'it was not idle' for a host when fetching perf data?

[12-Mar-2010 10:05:07] fOrsberg is now known as forsberg

[12-Mar-2010 10:19:14] <jb> venturaville: looks like DERIVE is working.

[12-Mar-2010 10:20:42] <venturaville> cool :-)

[12-Mar-2010 10:29:11] <sciolist> is there a 'best practice' guideline for running 'zenfixit' on a regular basis (i.e., weekly)?

[12-Mar-2010 10:35:03] <venturaville> bigegor: ... having trouble getting more than one WBEMdatasource to work at the same time

[12-Mar-2010 10:35:24] <venturaville> if I have 2 configured (each with 2 data points), it only ever does one set of them

[12-Mar-2010 10:36:54] <bigegor> is CIM class the same?

[12-Mar-2010 10:37:01] <venturaville> yep

[12-Mar-2010 10:37:14] <venturaville> different args

[12-Mar-2010 10:37:38] <venturaville> and I get a task is not idle note when I have them both enabled

[12-Mar-2010 10:39:13] <venturaville> scratch that I get them anyways: 2010-03-12 11:38:18,838 DEBUG zen.collector.scheduler: Task <mydevicename>  skipped because it was not idle

[12-Mar-2010 10:39:28] <venturaville> that may just a random useless note though

[12-Mar-2010 10:39:37] <venturaville> just -> just be

[12-Mar-2010 11:21:15] <rmatte> woohoo, finally done customizing and debugging the new version of our ticketing system, now I just need to plan deployment, then I can get back to Zenoss stuff ;)

[12-Mar-2010 12:27:00] forsberg is now known as fOrsberg

[12-Mar-2010 12:48:13] fOrsberg is now known as forsberg

[12-Mar-2010 14:52:04] <eidolon> hey folks - someone mentiond a while back that there was a config change in zenoss that would help with slow response time - sometihng about changing the numbe rof polling threads?  Anyone got a pointer?

[12-Mar-2010 14:52:58] <rmatte> eidolon: docs/DOC-2521

[12-Mar-2010 14:53:19] <rmatte> or are you talking slow network response time?

[12-Mar-2010 14:53:27] <eidolon> one moment

[12-Mar-2010 14:54:03] <eidolon> we're gtting incredibly slow response on teh web ui

[12-Mar-2010 15:38:59] forsberg is now known as fOrsberg

[12-Mar-2010 16:02:37] etank is now known as etank_afk

[12-Mar-2010 17:27:43] <mrayzenoss> FYI, the 2.5.2 .DEB installer has been updated.  message/46662

[12-Mar-2010 17:28:22] <RoAkSoAx> mrayzenoss, when you planing to migrate to python2.6?

[12-Mar-2010 17:29:18] <mrayzenoss> the next major release will be on Python 2.6

[12-Mar-2010 17:29:34] <mrayzenoss> we finally have trunk installers building with Python 2.6

[12-Mar-2010 17:29:45] <mrayzenoss> about to switch all of SVN to Python 2.6

[12-Mar-2010 17:29:48] <mrayzenoss> SVN trunk

[12-Mar-2010 17:31:38] <RoAkSoAx> mrayzenoss, awesome

[12-Mar-2010 17:31:48] <RoAkSoAx> we can now package to include it into ubuntu

[12-Mar-2010 17:31:53] <RoAkSoAx> in the ubuntu's repos

[12-Mar-2010 17:33:22] <mrayzenoss> that's the goal

[12-Mar-2010 17:34:37] <mrayzenoss> I'm heading out for a long weekend, but email me about it and I'll keep you in the loop as we start working on making the build more distro friendly in the next few weeks.  I'll be working with Debian, Ubuntu and Fedora so far

[12-Mar-2010 17:34:54] <mrayzenoss> and I expect FreeBSD and OpenSolaris as well

[12-Mar-2010 17:35:11] <mrayzenoss> but now I gotta go, wife's birthday dinner

[12-Mar-2010 20:10:45] etank_afk is now known as etank

[13-Mar-2010 00:00:30] [disconnected at Sat Mar 13 00:00:30 2010]

[13-Mar-2010 00:00:30] [connected at Sat Mar 13 00:00:30 2010]

[13-Mar-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[13-Mar-2010 02:48:02] fOrsberg is now known as forsberg

[13-Mar-2010 19:41:04] <Psi-Jack> Oh spiffy. Is this an official Zenoss support channel?

[13-Mar-2010 19:58:44] <theacolyte> Psi-Jack: yeah

[13-Mar-2010 19:58:47] <theacolyte> it is official

[13-Mar-2010 19:59:01] <Psi-Jack> Oh cool! theacolyte, I remember you from somewhere.. :)

[13-Mar-2010 19:59:02] <theacolyte> pretty dead on the weekends, but during the day it is active

[13-Mar-2010 19:59:07] <theacolyte> hehe yup

[13-Mar-2010 19:59:11] <theacolyte> ##windows-server

[13-Mar-2010 19:59:16] <Psi-Jack> Yep. ;)

[13-Mar-2010 19:59:17] <theacolyte> (and ##windows)

[13-Mar-2010 19:59:24] <theacolyte> anyways, headed for dinner, ttyl

[13-Mar-2010 19:59:28] <Psi-Jack> kk ;)

[13-Mar-2010 22:20:40] forsberg is now known as fOrsberg

[14-Mar-2010 00:00:30] [disconnected at Sun Mar 14 00:00:30 2010]

[14-Mar-2010 00:00:30] [connected at Sun Mar 14 00:00:30 2010]

[14-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[14-Mar-2010 01:42:17] fOrsberg is now known as forsberg

[14-Mar-2010 05:50:44] <Psi-Jack> Hmmm..

[14-Mar-2010 05:51:31] <Psi-Jack> How would I get Zenoss to be able to get the information as needed via SNMP for OS Processes and IP Services? Those two seem to be empty on every single one of my device listings I've setup.

[14-Mar-2010 09:53:19] forsberg is now known as fOrsberg

[14-Mar-2010 16:37:56] fOrsberg is now known as forsberg

[14-Mar-2010 21:01:36] bzed_ is now known as bzed

[15-Mar-2010 00:00:30] [disconnected at Mon Mar 15 00:00:30 2010]

[15-Mar-2010 00:00:30] [connected at Mon Mar 15 00:00:30 2010]

[15-Mar-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[15-Mar-2010 00:00:43] <holmes.freenode.net> [freenode-info] why register and identify? your IRC nick is how people know you. http://freenode.net/faq.shtml#nicksetup

[15-Mar-2010 02:36:20] forsberg is now known as fOrsberg

[15-Mar-2010 03:24:30] fOrsberg is now known as forsberg

[15-Mar-2010 04:22:18] <Troubadix09> hi all

[15-Mar-2010 07:35:44] <whitemice> I'm getting trap OID 1.3.6.1.2.1.2.2.1.8.15 from my NetGear switch;  and ZenOSS says this is undefined/unrecognized.  But isn't 1.3.6.1.2.1.2.2.1.8. a standard OID indicating the state of the interface?

[15-Mar-2010 08:01:06] Psi-Jack_ is now known as Psi-Jack

[15-Mar-2010 08:35:46] <mloven> Hey all, in 2.5.2, do cosmetic changes made in the event console still occur on the global level?   Meaning, they effect all users?

[15-Mar-2010 08:45:35] <mloven> nevermind.  Chet answered for me.

[15-Mar-2010 08:45:37] <mloven> :)

[15-Mar-2010 09:58:24] <fuzz> i heard that some/all "max recursion depth exceeded" problems should be fixed in 2.5.2 (like in our case: the network 10.0.0.0 is a child of itself, which causes everything from the gui to logging to zenping to die)

[15-Mar-2010 09:58:33] <fuzz> however, it does not seem to be fixed here

[15-Mar-2010 09:58:36] <fuzz> after upgrade to 2.5.2

[15-Mar-2010 09:58:57] <fuzz> i guess the cause is fixed, but it won't automatically fix an old objects that are broken?

[15-Mar-2010 09:59:02] <fuzz> -an

[15-Mar-2010 09:59:56] <mistich> it will not fix it just prevent it from happening

[15-Mar-2010 10:00:14] <fuzz> any idea how i can fix it?

[15-Mar-2010 10:00:26] <fuzz> when trying to delete the bad object, it dies with the same recursion error

[15-Mar-2010 10:00:35] <fuzz> both from the gui at /manage and from zendmd...

[15-Mar-2010 10:05:14] <fuzz> or do i have to add my ugly hacks to the code again? (if recursed more than 100 times, just use this object and hope for the best) :((

[15-Mar-2010 10:05:21] <fuzz> i can't imagine i'm the only one with this problem?

[15-Mar-2010 10:09:02] <mistich> we have run into it but just a couple of networks  so have been able to delete just those affected but last I talked to zenoss they suggested delete all networks and remodel

[15-Mar-2010 10:09:58] <venturaville> ok .. I need someone with some RRD skills to help

[15-Mar-2010 10:10:15] <venturaville> trying to preserve the maximums in a set of data going back months

[15-Mar-2010 10:10:25] <venturaville> and it looks like it is averaging peaks instead by default...

[15-Mar-2010 10:10:40] <venturaville> any thoughts on how to rectify that

[15-Mar-2010 10:10:41] <venturaville> ?

[15-Mar-2010 10:10:46] <fuzz> mistich: mkay.. but if i get an error when trying to delete the problem-network, how will i be able to delete all networks?

[15-Mar-2010 10:11:26] <mistich> good point not sure on that one sorry

[15-Mar-2010 10:14:44] forsberg is now known as fOrsberg

[15-Mar-2010 10:15:14] <fuzz> i fear i have to reinstall and set up everything from scratch to fix it :\

[15-Mar-2010 10:15:16] <fuzz> the horror!

[15-Mar-2010 10:15:46] <fuzz> ~400 devices.. *sigh*

[15-Mar-2010 10:31:46] <venturaville> bigegor: I got much of the fetching working with Engenio Provider.  We have a ticket open to fix the interaction between it and the arrays now

[15-Mar-2010 10:45:53] <mistich> fuzz I feel your pain

[15-Mar-2010 10:47:10] <fuzz> i added my hacks in _getNet() again.. so problem "solved" untill next upgrade ;)

[15-Mar-2010 10:47:36] <mistich> well hopefully its fixed in 2.5.2

[15-Mar-2010 13:57:09] <jb> hey cgibbons?

[15-Mar-2010 17:45:19] fOrsberg is now known as forsberg

[15-Mar-2010 17:47:55] <forsberg> anyone awake? :P

[15-Mar-2010 17:49:32] <aclark> o/

[15-Mar-2010 17:50:45] <forsberg> i wonder what the best way to monitor a website through zenoss is.....i want to be able to check for content

[15-Mar-2010 19:33:26] <sergeymasushko> is anyone around?

[15-Mar-2010 20:10:18] forsberg is now known as fOrsberg

[16-Mar-2010 00:00:30] [disconnected at Tue Mar 16 00:00:30 2010]

[16-Mar-2010 00:00:30] [connected at Tue Mar 16 00:00:30 2010]

[16-Mar-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[16-Mar-2010 02:47:57] <Troubadix09> morning all

[16-Mar-2010 03:58:37] fOrsberg is now known as forsberg

[16-Mar-2010 08:09:02] forsberg is now known as fOrsberg

[16-Mar-2010 08:40:17] <san4el> is there russian users&

[16-Mar-2010 08:40:19] <san4el> ?

[16-Mar-2010 09:03:56] <Egyptian[laptop]> help .. i just upgraded to 2.5.2 and i have lost my events view

[16-Mar-2010 09:07:29] <Egyptian[laptop]> oh thank god .. i came back after i refreshed the entire page

[16-Mar-2010 09:07:33] <Egyptian[laptop]> s/i/it

[16-Mar-2010 09:28:35] <rmatte> Egyptian[laptop]: part of the upgrade steps is to clear your browser cache

[16-Mar-2010 09:28:39] <rmatte> so you might want to do that

[16-Mar-2010 09:31:21] <Egyptian[laptop]> rmatte: np

[16-Mar-2010 09:31:23] <Egyptian[laptop]> thanks

[16-Mar-2010 09:36:17] <rmatte> np

[16-Mar-2010 10:22:57] <rmatte> hmmm, I need to figure out how to work with time in an event transform

[16-Mar-2010 10:23:11] <rmatte> like, have the event ignored for a specific device if the time is between something and something each day

[16-Mar-2010 10:23:21] <rmatte> this shall be fun

[16-Mar-2010 10:23:37] <rmatte> and a tip of the month candidate if I can get it working

[16-Mar-2010 10:35:00] <Egyptian[laptop]> rmatte: what is the definition of weekday in zenoss?

[16-Mar-2010 10:35:04] <Egyptian[laptop]> monday to friday?

[16-Mar-2010 10:35:12] <Egyptian[laptop]> if so how do i change it from sunday to thursday ?

[16-Mar-2010 10:40:40] <rmatte> I'm honestly not sure

[16-Mar-2010 10:40:46] <rmatte> I don't use that as a filter

[16-Mar-2010 10:40:58] <rmatte> well, I don't use schedules I should say

[16-Mar-2010 10:41:03] <rmatte> we're a 24x7 operation ;)

[16-Mar-2010 10:57:43] <rmatte> woohoo, figured out the code

[16-Mar-2010 10:57:53] <rmatte> import time

[16-Mar-2010 10:57:55] <rmatte> if time.localtime()[3] >= 11 and time.localtime()[3] < 12:

[16-Mar-2010 10:58:11] <rmatte> would do whatever action between 11 and 12 (as an example)

[16-Mar-2010 11:42:28] fOrsberg is now known as forsberg

[16-Mar-2010 12:22:33] <rmatte> voila

[16-Mar-2010 12:22:34] <rmatte> docs/DOC-5822

[16-Mar-2010 12:22:58] <Simon4> nice :)

[16-Mar-2010 12:24:01] <Simon4> someone was saying that "if 'ottawa' in evt.device' is lighter weight than re.match if you're worried about performance

[16-Mar-2010 12:26:44] <rmatte> it is lighter weight, I just do re.match as force of habbit

[16-Mar-2010 12:26:56] <rmatte> but honestly, I doubt it's THAT much more lightweight that it's a big deal

[16-Mar-2010 12:27:03] <rmatte> ;)

[16-Mar-2010 12:28:13] <Simon4> yeah, I haven't tried any actual tests

[16-Mar-2010 12:28:38] <Simon4> I'm about to dive head first into event transforms though, so examples like that are invaluable, cheers

[16-Mar-2010 12:29:12] <rmatte> :)

[16-Mar-2010 12:29:59] <rmatte> I'd never seen it done before and someone here mentioned that they wanted to drop events for a device during a certain period so I decided to dig, hopefully it'll be useful to others as well

[16-Mar-2010 12:30:19] <rmatte> transforms are nice because the possibilities are basically limitless

[16-Mar-2010 12:31:39] <Simon4> yeah, we have a need to serve up different reference url's based on server types and error types, so transforms should be useful for that kind of thing I think

[16-Mar-2010 12:33:36] <rmatte> yup

[16-Mar-2010 12:33:41] <rmatte> transforms are just python scripts

[16-Mar-2010 12:33:45] <rmatte> so you can do anything with them

[16-Mar-2010 12:34:29] <Simon4> do they have filesystem access? wondering about some kind of lookup that could be managed outside of zenoss itself

[16-Mar-2010 12:34:56] <rmatte> provided you import the correct libraries, yes

[16-Mar-2010 12:35:18] * Simon4 will have a play

[16-Mar-2010 12:35:37] <rmatte> I've never personally tried it so I don't have any examples to direct you to

[16-Mar-2010 12:37:10] <Simon4> if I make it work and it doesn't suck I'll write it up

[16-Mar-2010 12:38:30] <rmatte> cool

[16-Mar-2010 14:24:50] <venturaville> are there any agility users here?

[16-Mar-2010 14:25:27] <venturaville> (Anystream)

[16-Mar-2010 14:43:53] <rmatte> venturaville: noop

[16-Mar-2010 14:48:55] <venturaville> ah well... guess I am supporting this zenpack on my own :-)

[16-Mar-2010 14:49:05] <rmatte> hehe

[16-Mar-2010 14:49:55] <mistich> what zenpack

[16-Mar-2010 14:50:17] <venturaville> Anystream Agility monitor (using SOAP API)

[16-Mar-2010 14:50:55] <mistich> nope sorry

[16-Mar-2010 15:09:11] <rmatte> it's quiet in here today

[16-Mar-2010 15:09:56] <theacolyte>

[16-Mar-2010 15:09:58] <theacolyte> there!

[16-Mar-2010 15:11:52] <rmatte> ?

[16-Mar-2010 15:17:19] <theacolyte> Just keeping the  convo going brahhh

[16-Mar-2010 15:21:41] <forsberg> whats the best way to monitor a website in zenoss? i need to check http request for content....

[16-Mar-2010 15:22:16] <forsberg> im haxing a wget | grep blah return 0 or 1, but isnt there something smarter?

[16-Mar-2010 15:46:58] <rmatte> forsberg: you can always use the synthetic transactions pack

[16-Mar-2010 15:47:29] <rmatte> docs/DOC-3757

[16-Mar-2010 15:47:35] <rmatte> it uses twill and is very good

[16-Mar-2010 15:49:00] <forsberg> cool thanks!

[16-Mar-2010 15:49:04] <rmatte> np

[16-Mar-2010 15:49:17] <forsberg> just something like i knew was made - but the apache and http monitor was doing other stuff

[16-Mar-2010 15:49:32] <rmatte> I still use http monitor for basic monitoring

[16-Mar-2010 15:49:51] <rmatte> but I use the synthetic transactions pack when I need something a bit more advanced

[16-Mar-2010 15:49:54] <rmatte> it works like a charm

[16-Mar-2010 15:51:14] <forsberg> yeah well we are apperntly responsible for some loadbalances for a big customer, and i need something to make sure that the website is available

[16-Mar-2010 15:51:23] <forsberg> ie checking one or more "steps" is perfect

[16-Mar-2010 15:51:59] <forsberg> untill they decide to change any/every-thing behind the LB's wich we are not responsible of....

[16-Mar-2010 15:52:01] <forsberg> ;)

[16-Mar-2010 15:53:08] <rmatte> yeh

[16-Mar-2010 15:53:39] <rmatte> we used to monitor a car rental site and had a script that rented a car every half hour

[16-Mar-2010 15:53:46] <rmatte> :)

[16-Mar-2010 15:54:01] <rmatte> best way to ensure availability

[16-Mar-2010 15:54:39] <forsberg> right on :P

[16-Mar-2010 15:55:00] <forsberg> ill buy some stocks every half hour

[16-Mar-2010 15:55:02] <forsberg> and get rich

[16-Mar-2010 15:55:04] <forsberg> :P~

[16-Mar-2010 15:55:09] <rmatte> or so you think

[16-Mar-2010 15:55:12] <rmatte> lol

[16-Mar-2010 16:14:44] <rmatte> actually, that sort of automated stock buying has been done before but much more sophisticated

[16-Mar-2010 16:14:58] <rmatte> try and find a copy of the BBC documentary "The Midas Formula"

[16-Mar-2010 16:15:00] <rmatte> it's really good

[16-Mar-2010 16:16:32] <forsberg> :P

[16-Mar-2010 16:16:35] <forsberg> i saw

[16-Mar-2010 16:16:59] <rmatte> you saw that it's been done before, or you've seen the documentary?

[16-Mar-2010 16:17:26] <forsberg> the documentary

[16-Mar-2010 16:17:34] <rmatte> two of the biggest disasters in trading history were due to automated trading, that was one of them

[16-Mar-2010 16:18:28] <rmatte> the other was this: http://www.wired.com/threatlevel/2008/09/six-year-old-st/

[16-Mar-2010 16:22:30] <rmatte> they don't mention it in the story, but apparently a lot of automated trading systems scrape headlines and automatically react based on them

[16-Mar-2010 18:30:43] <chrismsnz> hello all, does anybody have any experience with monitoring processes on a linux server using SSH monitoring?

[16-Mar-2010 18:31:19] <chrismsnz> i added some, they're detected when modeled, but when i bring the process down there's no alert/notification/status update... zenoss is still telling me it's up

[16-Mar-2010 18:31:53] <chrismsnz> I've been reading docs all morning, can't find anything to help me... running zenoss 2.5.2

[16-Mar-2010 19:15:07] <theacolyte> zenos suses system time right?

[16-Mar-2010 19:20:52] forsberg is now known as fOrsberg

[17-Mar-2010 00:00:30] [disconnected at Wed Mar 17 00:00:30 2010]

[17-Mar-2010 00:00:30] [connected at Wed Mar 17 00:00:30 2010]

[17-Mar-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[17-Mar-2010 02:11:24] fOrsberg is now known as forsberg

[17-Mar-2010 04:56:44] <Troubadix09> morning all

[17-Mar-2010 05:48:56] Simon4_ is now known as Simon4

[17-Mar-2010 08:55:49] <rmatte> Troubadix09: g'mornin

[17-Mar-2010 08:57:36] <venturaville> morning

[17-Mar-2010 09:09:47] <rmatte> well, time to test upgrading my production image to 2.5.2...

[17-Mar-2010 09:10:26] <venturaville> ...keep wishing zenoss would use the system python instead of its builtin one..........

[17-Mar-2010 09:10:46] <rmatte> venturaville: it uses it's built in one because it needs to use python 2.4

[17-Mar-2010 09:10:52] <rmatte> though they are switching to 2.6

[17-Mar-2010 09:11:04] <Simon4> that's going to be nice :)

[17-Mar-2010 09:11:09] <venturaville> +1

[17-Mar-2010 09:11:17] <rmatte> but then python 3.0 will become standard and they'll be behind again

[17-Mar-2010 09:11:18] <rmatte> ;)

[17-Mar-2010 09:11:22] <venturaville> heh

[17-Mar-2010 09:11:42] <npmccallum> python3 isn't becoming standard anywhere anytime soon

[17-Mar-2010 09:11:49] <npmccallum> python2 will be supported for a long time

[17-Mar-2010 09:11:53] <rmatte> fair enough

[17-Mar-2010 09:11:58] <venturaville> yeah python3 breaks A LOT

[17-Mar-2010 09:12:00] <rmatte> I was more kidding than anything ;)

[17-Mar-2010 09:12:04] <venturaville> including the standard hello world app

[17-Mar-2010 09:12:41] <rmatte> my friend was using python 3.0 on his Linux system and was wondering why he was having problems

[17-Mar-2010 09:12:50] <rmatte> I was like "you're using 3.0, are you nuts?"

[17-Mar-2010 09:13:35] <npmccallum> python3 will be in all the major next releases

[17-Mar-2010 09:13:45] <rmatte> well, this will be nice if the upgrade doesn't break a bunch of stuff on the first attempt...

[17-Mar-2010 09:13:48] <npmccallum> but my guess is that there will be a parallel installable python2 for a while

[17-Mar-2010 09:14:03] <rmatte> npmccallum: yeh, there would have to be

[17-Mar-2010 09:14:38] <rmatte> one of our developers here tried learning Python and told me he hated it

[17-Mar-2010 09:14:43] <venturaville> I don't think it is in the new ubuntu

[17-Mar-2010 09:14:48] <Simon4> rmatte: is he a COBOL programmer?

[17-Mar-2010 09:14:49] <venturaville> or at least it isn't listed in the pool dir

[17-Mar-2010 09:15:00] <rmatte> Simon4: nah, he does lots of web and database stuff

[17-Mar-2010 09:15:17] <venturaville> and I would be shocked if redhat 6 will have it

[17-Mar-2010 09:15:41] <npmccallum> venturaville: rh6 will have it

[17-Mar-2010 09:15:56] <npmccallum> venturaville: they hired someone just to do that this last year

[17-Mar-2010 09:16:02] <venturaville> interesting

[17-Mar-2010 09:16:05] <venturaville> got a link for that?

[17-Mar-2010 09:16:42] <npmccallum> probably not

[17-Mar-2010 09:16:47] <npmccallum> but F13 has python3 already

[17-Mar-2010 09:17:18] <rmatte> mhmmmm, ...and we have errors after upgrade, now to try to sort these out...

[17-Mar-2010 09:17:18] <rmatte> Error running su zenoss -c '"/usr/local/zenoss/zenoss/bin/zenmigrate" ' :

[17-Mar-2010 09:17:18] <rmatte> mhmmmm, that's not good

[17-Mar-2010 09:17:18] <rmatte> oh, I think this is that error that I figured out the fix for not too long ago

[17-Mar-2010 09:17:19] * rmatte checks his notes

[17-Mar-2010 09:17:29] <rmatte> Yeh, this one: http://dev.zenoss.org/trac/ticket/6212

[17-Mar-2010 09:19:29] <venturaville> looking at F13 it looks as though the stock python will still be something pre 3

[17-Mar-2010 09:19:38] <venturaville> though they do have a python3 rpm on there

[17-Mar-2010 09:19:52] <venturaville> perhaps an alternate python root for it (ala debian)

[17-Mar-2010 09:20:42] <rmatte> unhunh, it's complaining about CMFCore

[17-Mar-2010 09:20:55] <npmccallum> yes, python3 won't be default for a while

[17-Mar-2010 09:21:04] <npmccallum> F14 at the earliest

[17-Mar-2010 09:21:06] <npmccallum> and it would be painful

[17-Mar-2010 09:21:19] <npmccallum> if we can get to python2.6, I would be very happy

[17-Mar-2010 09:21:22] <rmatte> Nate, any idea about this...

[17-Mar-2010 09:21:24] <rmatte> http://fpaste.org/HfzR/

[17-Mar-2010 09:21:34] <rmatte> I've seen that error for ZenPacks before

[17-Mar-2010 09:21:45] <rmatte> oh wait, nevermind

[17-Mar-2010 09:21:49] <rmatte> I know what it is, totally forgot

[17-Mar-2010 09:22:00] <rmatte> need to fix the namespace of one of my Windows SNMP Performance ZenPacks

[17-Mar-2010 09:22:12] * rmatte slaps forehead

[17-Mar-2010 09:23:53] <venturaville> yep looking at the macros.python3 file it is installing it as /usr/bin/python3 by default

[17-Mar-2010 09:38:13] <rmatte> eugh, that really really sucks, no idea how I'm going to fix this namespace issue without removing all of the windows devices from Zenoss

[17-Mar-2010 09:38:35] <rmatte> actually, I think I'll just change the name of the other ZenPack, screw it

[17-Mar-2010 09:43:39] <forsberg> rmatta the synthtransaction thing works super!!! thanks again

[17-Mar-2010 09:44:56] <rmatte> npmccallum: to rename a ZenPack, is it as simple as unzipping it, changing the directory structure, and changing any reference to the ZenPack name in the files, then rezipping it and renaming to .egg?

[17-Mar-2010 09:45:12] <rmatte> forsberg: np

[17-Mar-2010 09:45:30] <npmccallum> rmatte: I *think* so

[17-Mar-2010 09:46:51] <rmatte> k, guess I'll just have to try and see

[17-Mar-2010 09:48:06] <rmatte> well, time to try the upgrade again without colliding namespaces

[17-Mar-2010 09:51:19] <rmatte> hopefully no more UTC error

[17-Mar-2010 09:52:11] <rmatte> that seems to have fixed it, woot

[17-Mar-2010 09:57:07] <rmatte> now to decide what I want to do with these 3 .zip ZenPacks

[17-Mar-2010 10:01:57] <rmatte> lovely

[17-Mar-2010 10:01:59] <rmatte> zenpack --remove APC\ UPS\ PowerNet

[17-Mar-2010 10:02:01] <rmatte> doesn't work

[17-Mar-2010 10:02:08] <rmatte> and I can't remove the pack via the UI

[17-Mar-2010 10:02:16] <rmatte> guess it's time to resort to zendmd magic

[17-Mar-2010 10:03:44] <rmatte> that worked

[17-Mar-2010 11:07:32] <rmatte> mrayzenoss: aren't you supposed to be on vacation?

[17-Mar-2010 11:08:06] <mrayzenoss> easing back after a 4 day weekend

[17-Mar-2010 11:08:17] <rmatte> ah

[17-Mar-2010 11:08:21] <rmatte> nice

[17-Mar-2010 11:36:20] <rmatte> weird

[17-Mar-2010 11:36:43] <rmatte> when I use "/etc/init.d/zenoss-stack start" to start Zenoss after the upgrade I get: http://fpaste.org/SQCQ/

[17-Mar-2010 11:36:59] <rmatte> but if I start zenperfwmi or zenticket manually as the Zenoss user they start fine

[17-Mar-2010 11:37:53] <mrayzenoss> is that the updated version of Egor's pack?

[17-Mar-2010 11:38:03] <rmatte> it's version 1.5

[17-Mar-2010 11:38:17] <rmatte> but I'm wondering why my custom daemon is also complaining

[17-Mar-2010 11:38:29] <rmatte> I mean, it starts perfectly fine if I start it as the zenoss user

[17-Mar-2010 11:38:38] <rmatte> the startup script is the only time it errors out

[17-Mar-2010 11:41:26] <rmatte> it's not written like a standard zenoss daemon is though, so that may be part of the problem

[17-Mar-2010 11:41:34] <rmatte> I'll try upgrading egor's pack for starters

[17-Mar-2010 11:42:26] <rmatte> hmmm, the latest is 1.5 according to the ZenPack page

[17-Mar-2010 11:43:23] <rmatte> is there any newer version?

[17-Mar-2010 11:45:15] <theacolyte> Have any of you seen graphs be an hour off after daylight savings despite the system clock being correct?

[17-Mar-2010 11:45:47] <rmatte> yeh, everything starts perfectly fine when done by hand, there must be something funky with the startup script

[17-Mar-2010 11:46:24] <rmatte> theacolyte: does zenoss time in the top right corner of the UI show the correct time?

[17-Mar-2010 11:46:40] <theacolyte> yeah it does

[17-Mar-2010 11:46:47] <rmatte> weird

[17-Mar-2010 11:47:14] <theacolyte> It's just weird because I had an outage and I was looking at my graphs

[17-Mar-2010 11:47:28] <theacolyte> And the spike was exactly one out off from where I would have thought it would be

[17-Mar-2010 11:47:31] <theacolyte> LIke *exactly*

[17-Mar-2010 11:47:58] <rmatte> mine are all fine

[17-Mar-2010 11:53:03] <rmatte> eugh, they made changes to ctlscript.sh, which is probably what is causing the errors, just need to figure out what the heck they changed

[17-Mar-2010 12:08:29] RoAk is now known as RoAkSoAx

[17-Mar-2010 12:11:09] <rmatte> well, it's not caused by any changes made directly to ctlscript.sh

[17-Mar-2010 12:39:26] <car1> i've got 2 instances of mysql running on different ports on a bunch of servers, to monitor each instance i have to create a local copy of the mysqlmonitor template and update the port number and graphs, would it be possible to script this for all of my database servers?

[17-Mar-2010 12:39:45] <rmatte> car1: anything is possible with scripting, but I don't know how you'd do it exactly off the top of my head

[17-Mar-2010 12:39:55] <rmatte> hunt around on the community site for scripts that do similar things

[17-Mar-2010 12:40:48] <Simon4> car1: if the port numbers are the same across servers (say 3306 and 3307), just make a copy of template #1,  as mysql_port2, change the graphs, and then apply that template wherever you like further down in your device tree

[17-Mar-2010 12:41:02] <Simon4> so the templates sit under /devices/server or whatnot

[17-Mar-2010 12:41:11] <Simon4> that way you only make the mod once, but can apply it to several servers

[17-Mar-2010 12:41:18] <Simon4> (we do this for the apache zenpack)

[17-Mar-2010 12:41:24] <car1> where are the device configurations (templates, and such) stored?

[17-Mar-2010 12:41:31] <car1> i only see an 'events' database in the server

[17-Mar-2010 12:41:38] <car1> i don't know that much about the zenoss architecture

[17-Mar-2010 12:41:42] <rmatte> it is all stored in Zope

[17-Mar-2010 12:41:47] <Simon4> they're in zopedb, which needs some pythonfu to play with

[17-Mar-2010 12:41:47] <rmatte> MySQL is only used for events

[17-Mar-2010 12:41:49] <car1> Simon4: unfortunately the port is different for each instance

[17-Mar-2010 12:42:09] <car1> i'm totally unfamiliar with zope

[17-Mar-2010 12:42:12] <car1> damn

[17-Mar-2010 12:42:19] <rmatte> it's a python based relational object database

[17-Mar-2010 12:42:33] <rmatte> I didn't know about Zope or Python until I started using Zenoss

[17-Mar-2010 12:42:38] <rmatte> you need to do some studying man

[17-Mar-2010 12:43:08] <car1> so i could possibly write a script to update the zope database directly?

[17-Mar-2010 12:43:13] <mrayzenoss> car1: are there a lot of devices?  Would it make sense to have separate device classes depending on which port is being monitored?

[17-Mar-2010 12:43:15] <car1> to apply local templates

[17-Mar-2010 12:43:28] <car1> there are 113 instances

[17-Mar-2010 12:43:32] <car1> different port for each

[17-Mar-2010 12:43:38] <car1> corresponding to a unique customer id

[17-Mar-2010 12:43:44] <mrayzenoss> ahhh

[17-Mar-2010 12:43:49] <Simon4> def needs scripting :)

[17-Mar-2010 12:43:56] <mrayzenoss> sounds scriptable

[17-Mar-2010 12:43:58] <Simon4> and yeah, doable for sure

[17-Mar-2010 12:44:17] <mrayzenoss> here's something in that vein: docs/DOC-2532  Copy_Datasources

[17-Mar-2010 12:44:19] <car1> anybody have a good zope tutuorial link?

[17-Mar-2010 12:44:23] <rmatte> car1: read up on zendmd for starters

[17-Mar-2010 12:44:39] <mrayzenoss> car1: lots of examples here: community/documentation/wiki/zendmd

[17-Mar-2010 12:44:49] <car1> roger that

[17-Mar-2010 12:44:53] <car1> thanks gentlemen

[17-Mar-2010 12:44:55] <rmatte> np

[17-Mar-2010 12:50:42] <forsberg> :)=

[17-Mar-2010 12:50:50] <rmatte> bah, the one day that egor hasn't dropped by yet

[17-Mar-2010 13:35:59] <rmatte> bigegor: hey

[17-Mar-2010 13:36:03] <rmatte> bigegor: question for you...

[17-Mar-2010 13:36:20] <bigegor> rmatte: hey

[17-Mar-2010 13:36:36] <bigegor> rmatte: ask

[17-Mar-2010 13:37:59] <rmatte> http://fpaste.org/SQCQ/

[17-Mar-2010 13:38:26] <rmatte> I installed the RC for your pack and the problem went away for that pack

[17-Mar-2010 13:38:41] <rmatte> but I still have the problem with my daemon that I wrote not importing libraries

[17-Mar-2010 13:38:53] <rmatte> and it only happens when I execute the startup script in /etc/init.d

[17-Mar-2010 13:38:58] <rmatte> any idea what I need to do to fix that?

[17-Mar-2010 13:39:21] <rmatte> I've been digging through scripts for the last hour and a half trying to figure this out

[17-Mar-2010 13:40:02] <rmatte> my daemons works fine in 2.4.5, so I'm not sure what they changed, I figured you might know since you have a lot of experience with this stuff

[17-Mar-2010 13:40:09] <rmatte> daemon*

[17-Mar-2010 13:43:16] <rmatte> and just to clarify, if I do "zenticket start" as the zenoss user it starts up no problem

[17-Mar-2010 13:43:34] <rmatte> but it refuses to start when I use the /etc/init.d/zenoss-stack script to start it as root

[17-Mar-2010 13:46:50] <bigegor> in my daemons i have change line MYPATH=`python -c "import os.path; print os.path.realpath('$0')"` to MYPATH=`readlink -f $0`

[17-Mar-2010 13:47:22] <rmatte> right, that was to fix the path issue, but I don't have a path issue, I have issues importing libraries for some reason

[17-Mar-2010 13:47:28] <rmatte> just wondering if you might know why that would be

[17-Mar-2010 13:48:18] <rmatte> (that readlink solution is great by the way, I adjusted my startup script to use that too)

[17-Mar-2010 13:48:53] <rmatte> the code for my daemon is available here: http://dmon.org/zenticket.html

[17-Mar-2010 13:49:08] <rmatte> not the cleanest code but it works, not sure if something in there is responsible for the imports failing

[17-Mar-2010 13:49:14] <rmatte> or if it's something else entirely

[17-Mar-2010 13:49:43] <rmatte> os is a pretty basic module though

[17-Mar-2010 13:49:49] <rmatte> surprising that it's failing to import it

[17-Mar-2010 13:51:57] <bigegor> it can by problem with PATH variable???

[17-Mar-2010 13:52:44] <rmatte> maybe

[17-Mar-2010 13:52:47] <rmatte> but which one?

[17-Mar-2010 13:53:14] <rmatte> and why only for my script

[17-Mar-2010 13:53:59] <rmatte> hmmm, then again you were getting import error for os too

[17-Mar-2010 13:54:54] * rmatte scratches his head

[17-Mar-2010 13:54:54] <rmatte> wonder why os is no longer supported all of a sudden

[17-Mar-2010 13:54:54] <rmatte> npmccallum: you alive?

[17-Mar-2010 13:55:22] <bigegor> oe, it try to import os from local installed python but not from /usr/local/zenoss

[17-Mar-2010 13:56:01] <rmatte> oh

[17-Mar-2010 13:56:10] <npmccallum> rmatte: nope :)

[17-Mar-2010 13:56:19] <rmatte> npmccallum: aha, there ye are

[17-Mar-2010 13:56:40] <rmatte> npmccallum: any idea why an import os fails when ctlscript.sh is executed?

[17-Mar-2010 13:56:59] <rmatte> but works fine when the script is executed as the zenoss user

[17-Mar-2010 13:57:04] <rmatte> seems like a bug

[17-Mar-2010 13:57:42] <rmatte> charlieS_: http://fpaste.org/SQCQ/

[17-Mar-2010 13:57:42] <rmatte> bah

[17-Mar-2010 13:57:42] <rmatte> meant to say IE: http://fpaste.org/SQCQ/

[17-Mar-2010 13:57:42] <rmatte> stupid autocomplete

[17-Mar-2010 13:58:52] <npmccallum> rmatte: no idea

[17-Mar-2010 13:59:31] <rmatte> anything with import os started fine with the startup script in 2.4.5 but doesn't in 2.5.2

[17-Mar-2010 14:32:38] <bigegor> i've put on SVN WMIDataSource 2.1 and WBEMDataSource 2.1 zenpacks, I will publish them on sunday if no errors be reported.

[17-Mar-2010 14:33:10] <rmatte> cool

[17-Mar-2010 14:33:37] <rmatte> I'm planning a 2.5.2 upgrade on my prod systems, part of it will be updating the WMIDataSource pack, so I look forward to that

[17-Mar-2010 14:33:53] <rmatte> when are you releasing the new WMI Performance Template pack?

[17-Mar-2010 14:37:07] <bigegor> next week, I think.

[17-Mar-2010 14:37:12] <rmatte> cool

[17-Mar-2010 14:37:25] <rmatte> well, once I figure out this import issue and you release those packs I'll be good to update

[17-Mar-2010 14:37:33] <rmatte> this has to be a path thing

[17-Mar-2010 14:38:26] <rmatte> bigegor: is that release version the one that supports full WMI modelling/monitoring?

[17-Mar-2010 14:38:30] <rmatte> or is that a later release?

[17-Mar-2010 14:39:07] <bigegor> rmatte: yes, modelling is supported.

[17-Mar-2010 14:39:13] <rmatte> sweet

[17-Mar-2010 14:41:25] <rmatte> that's going to be some groundbreaking stuff for core

[17-Mar-2010 14:48:01] <bigegor> I have more ideas ;)

[17-Mar-2010 14:50:17] <etank> rmatte: i was asking you about this a few days ago http://dpaste.com/173069/ and you said that it shouldnt work. why would that be again?

[17-Mar-2010 15:03:48] <rmatte> one second

[17-Mar-2010 15:04:16] <rmatte> etank: aha, you're the person I wanted to talk to

[17-Mar-2010 15:04:43] <etank> rmatte: awesome.

[17-Mar-2010 15:04:47] <etank> i think

[17-Mar-2010 15:04:59] <rmatte> You'll be happy to know that I needed this for something too so I figured out how to do it

[17-Mar-2010 15:05:00] <rmatte> docs/DOC-5822

[17-Mar-2010 15:05:17] <rmatte> you'll have to modify the code a bit to take in to account the half hour

[17-Mar-2010 15:05:25] <rmatte> but it's simple to do

[17-Mar-2010 15:06:02] <rmatte> hmmm

[17-Mar-2010 15:06:02] <etank> why would what i have not work?

[17-Mar-2010 15:06:11] <rmatte> actually, you're doing it based on timestamps of the events

[17-Mar-2010 15:06:14] <rmatte> so that's different

[17-Mar-2010 15:06:33] <rmatte> the reason it won't work the way you're doing it is because 05:30:00 is not an integer

[17-Mar-2010 15:06:38] <etank> http://dpaste.com/173082/  <-- that seems to work

[17-Mar-2010 15:06:51] <rmatte> Python can understand the concept of something like if whatever >= 100

[17-Mar-2010 15:06:59] <rmatte> but it has no idea what 05:30:00 is

[17-Mar-2010 15:07:02] <rmatte> it's just a string to it

[17-Mar-2010 15:07:24] <rmatte> really?

[17-Mar-2010 15:07:28] <etank> yup

[17-Mar-2010 15:07:31] <etank> give it a try

[17-Mar-2010 15:07:41] <rmatte> k

[17-Mar-2010 15:07:45] <etank> change the times around some too

[17-Mar-2010 15:08:19] <etank> my thought is that if i do this

[17-Mar-2010 15:08:21] <etank> datestamp, timestamp = str(evt.firstTime.split(' '))

[17-Mar-2010 15:08:36] <etank> then it would force timestamp to be a string and will allow it to work

[17-Mar-2010 15:08:52] <rmatte> wow, never thought that would work

[17-Mar-2010 15:08:58] <rmatte> apparently python is even cooler than I knew

[17-Mar-2010 15:08:59] <rmatte> lol

[17-Mar-2010 15:09:03] <etank> :)

[17-Mar-2010 15:09:10] <rmatte> good work

[17-Mar-2010 15:09:22] <etank> i think that part of my problem was evt.action as opposed to evt._action

[17-Mar-2010 15:09:40] <rmatte> were you just going to use that in a transform though?

[17-Mar-2010 15:09:51] <etank> i had evt._action='history' and it didnt seem to be working

[17-Mar-2010 15:09:52] <rmatte> you basically want certain events to be dropped at certain times?

[17-Mar-2010 15:09:53] <etank> yeah

[17-Mar-2010 15:10:00] <etank> i have an event class

[17-Mar-2010 15:10:07] <rmatte> My method is a bit simpler honestly

[17-Mar-2010 15:10:18] <etank> and when those events come in between certain times i want to throw them in history

[17-Mar-2010 15:10:20] <rmatte> but whatever works

[17-Mar-2010 15:10:27] <etank> ok

[17-Mar-2010 15:10:32] <etank> let me ask this then

[17-Mar-2010 15:11:00] <etank> if i want a transform that will bump the sev when a count reaches a certain number how would i do that?

[17-Mar-2010 15:11:19] <rmatte> if evt.count == 3:

[17-Mar-2010 15:11:27] <etank> ok

[17-Mar-2010 15:11:37] <etank> then set the sev

[17-Mar-2010 15:11:43] <rmatte> then under that if you'd do...

[17-Mar-2010 15:11:56] <rmatte> if evt.severity != 5:

[17-Mar-2010 15:12:06] <rmatte>    evt.severity = evt.severity + 1

[17-Mar-2010 15:12:27] <rmatte> or just set it to something specific

[17-Mar-2010 15:12:39] <etank> on that event class i have all events come in as warning. when the count is above 10 within one hour i want to bump the sev to error. if it is outside the hour then throw it in history

[17-Mar-2010 15:12:51] <rmatte> ah

[17-Mar-2010 15:12:59] <rmatte> yeh, do something like I just showed

[17-Mar-2010 15:13:07] <etank> the hour thing i think i can do easily

[17-Mar-2010 15:13:07] <rmatte> it'll work fine

[17-Mar-2010 15:13:28] <etank> i set it to purge events after one hour that have a sev of warn or lower

[17-Mar-2010 15:13:41] <etank> may not be the best way to do it but i think it would work

[17-Mar-2010 15:13:43] <rmatte> well, the hour thing you could use your technique against evt.firstTime and evt.lastTime

[17-Mar-2010 15:14:00] <rmatte> if less than an hour has elapsed between the 2 and the count is 10 do whatever

[17-Mar-2010 15:14:11] <etank> would i need to use timedelta?

[17-Mar-2010 15:14:36] <rmatte> you'd have to do it the way you showed me

[17-Mar-2010 15:14:45] <etank> ok cool

[17-Mar-2010 15:14:47] <rmatte> but also compare the dates to make sure that one isn't higher than the other

[17-Mar-2010 15:14:56] <rmatte> because that would mean that a whole day has elapsed

[17-Mar-2010 15:15:10] <etank> if datetime == datetime:

[17-Mar-2010 15:15:16] <etank> blah

[17-Mar-2010 15:15:22] <rmatte> lol, not quite

[17-Mar-2010 15:15:38] <etank> datestamp, timestamp = str(evt.firstTime.split(' '))

[17-Mar-2010 15:15:40] <rmatte> if date == date: more like it

[17-Mar-2010 15:15:43] <rmatte> after splitting them

[17-Mar-2010 15:15:47] <etank> if datestamp == datestamp

[17-Mar-2010 15:15:52] <etank> thats what i meant

[17-Mar-2010 15:16:00] <rmatte> actually, you'll need 2 variables

[17-Mar-2010 15:16:07] <rmatte> 1 for the firstTime and 1 for the lastTime

[17-Mar-2010 15:16:09] <rmatte> then compare those 2

[17-Mar-2010 15:16:12] <rmatte> and then compare the 2 times

[17-Mar-2010 15:16:19] <etank> your right

[17-Mar-2010 15:16:27] <etank> easy enough i think

[17-Mar-2010 15:16:38] <rmatte> yeh, shouldn't be too bad, you've figured out the tricky part

[17-Mar-2010 15:16:53] <rmatte> Feel free to add your method to my wiki entry

[17-Mar-2010 15:16:57] <etank> :)

[17-Mar-2010 15:17:06] <rmatte> it's another good way of doing it

[17-Mar-2010 15:18:14] <rmatte> make sure you do evt._action = 'history'

[17-Mar-2010 15:18:17] <rmatte> not evt.action

[17-Mar-2010 15:18:33] <etank> evt._action didnt seem to be working for some reason

[17-Mar-2010 15:18:45] <rmatte> evt._action is the only way that should work

[17-Mar-2010 15:18:46] <etank> not sure why

[17-Mar-2010 15:19:02] <etank> thats what made me think that my time stuff was broken

[17-Mar-2010 15:19:05] <rmatte> if it wasn't working then it was probably the code leading up to it which was wrong

[17-Mar-2010 15:19:15] <rmatte> hmmm

[17-Mar-2010 15:19:27] <jb> anyone here understand MS licensing by any chance?:)

[17-Mar-2010 15:19:27] <rmatte> does evt.action even work for you?

[17-Mar-2010 15:19:52] <rmatte> jb: expensive, unreasonable, decent service

[17-Mar-2010 15:20:00] <etank> rmatte: i would have to check tomorrow morning when it happens again to be sure i think

[17-Mar-2010 15:20:23] <rmatte> etank: k, evt._action = is really the proper way to do it

[17-Mar-2010 15:28:37] <rmatte> eugh, this is definitely a python library path issue of some sort

[17-Mar-2010 15:28:49] <rmatte> I can run any of the python environments as root and import os

[17-Mar-2010 15:51:47] <rmatte> think I half figured this out

[17-Mar-2010 15:51:51] <rmatte> when the startup script is running it's looking for os.py in trying /usr/local/zenoss/zenoss/lib/python/os.py

[17-Mar-2010 15:52:31] <rmatte> when I run it as the Zenoss user it's looking for it in /usr/local/zenoss/python/lib/python2.4/os.py

[17-Mar-2010 15:54:23] <rmatte> but I still haven't found any differences in the environment variables

[17-Mar-2010 15:54:29] <rmatte> so I'm stumped

[17-Mar-2010 15:55:02] <ke4qqq> rmatte: can you change pythons path in that script to look elsewhere first (really just changing the order of path)

[17-Mar-2010 15:55:32] <ke4qqq> I've had to do that when dealing with bundled python (speaking of bundled python is evil - just saying :) )

[17-Mar-2010 15:57:04] <rmatte> ke4qqq: I'm going to try that...

[17-Mar-2010 15:57:29] <rmatte> I was just kind of hoping to make this as dynamic as possible

[17-Mar-2010 15:57:55] <rmatte> actually, I still can

[17-Mar-2010 16:04:00] <rmatte> well, that fixed the os issue, now it's complaining about the module time

[17-Mar-2010 16:04:02] * rmatte rolls his eyes

[17-Mar-2010 16:04:45] <rmatte> mhmmm, because apparently there isn't one, wtf

[17-Mar-2010 16:05:32] <rmatte> ah, think I found it possibly

[17-Mar-2010 16:11:47] <rmatte> eugh, this is ridiculous, now the only time module I can find is in $ZENHOME/../python/lib/python2.4/lib-dynload

[17-Mar-2010 16:12:05] <rmatte> but I get: ImportError: /usr/local/zenoss/python/lib/python2.4/lib-dynload/time.so: undefined symbol: Py_InitModule4

[17-Mar-2010 16:13:17] <rmatte> AHA

[17-Mar-2010 16:13:25] <rmatte> finally got it

[17-Mar-2010 16:14:02] <rmatte> Had to add the following to the startup script for my daemon...

[17-Mar-2010 16:14:03] <rmatte> PYTHONPATH=$ZENHOME/../python/lib/python2.4:/usr/lib/python2.5/lib-dynload:$PYTHONPATH

[17-Mar-2010 16:14:03] <rmatte> export PYTHONPATH

[17-Mar-2010 16:15:16] <rmatte> dirty solution since I have to use the system time module, but at least it's working

[17-Mar-2010 16:16:17] <rmatte> mhmmm, actually now it doesn't work properly as the zenoss user

[17-Mar-2010 16:16:19] <rmatte> this is fantastic

[17-Mar-2010 16:28:12] <rmatte> there, now it's fixed

[17-Mar-2010 16:28:33] <rmatte> Had to add:

[17-Mar-2010 16:28:34] <rmatte> PATH=$ZENHOME/../common/sbin:$ZENHOME/../python/bin:$ZENHOME/../mysql/bin:$ZENHOME/bin:$ZENHOME/../common/bin:$PATH

[17-Mar-2010 16:28:34] <rmatte> export PATH

[17-Mar-2010 16:30:15] <mrayzenoss> so it was your PATH, not your PYTHONPATH?

[17-Mar-2010 16:30:28] <rmatte> it was both

[17-Mar-2010 16:30:41] <rmatte> I had to add the following to my daemon startup script...

[17-Mar-2010 16:30:42] <rmatte> PYTHONPATH=$ZENHOME/../python/lib/python2.4:$ZENHOME/../python/lib/python2.4/lib-dynload:$PYTHONPATH

[17-Mar-2010 16:30:43] <rmatte> export PYTHONPATH

[17-Mar-2010 16:30:43] <rmatte> PATH=$ZENHOME/../common/sbin:$ZENHOME/../python/bin:$ZENHOME/../mysql/bin:$ZENHOME/bin:$ZENHOME/../common/bin:$PATH

[17-Mar-2010 16:30:43] <rmatte> export PATH

[17-Mar-2010 16:30:52] <rmatte> which is stupid, because it worked perfectly fine in 2.4.5

[17-Mar-2010 16:31:06] <rmatte> not sure what changed, but the environment variables aren't being set properly

[17-Mar-2010 16:31:16] <rmatte> when the startup script kicks off it's using the system python

[17-Mar-2010 16:31:20] <rmatte> which is not good

[17-Mar-2010 16:32:14] <rmatte> not sure if it actually runs the other daemons with the system python or not

[17-Mar-2010 16:32:19] <rmatte> but it would be worth investigating

[17-Mar-2010 16:33:06] <rmatte> at least this is a workaround of sorts for now

[17-Mar-2010 16:33:34] <rmatte> guess I'll be releasing an update to my zenticket daemon sooner than originally planned

[17-Mar-2010 16:34:30] <rmatte> egor basically just masked the problem by changing the way he was finding the path

[17-Mar-2010 16:34:53] <rmatte> and since he doesn't use the os or time modules apparently, he never gets errors when loading the daemon

[17-Mar-2010 16:34:58] <rmatte> same with the other daemons

[17-Mar-2010 16:35:20] <rmatte> but I suspect that they are being loaded with the system python (could be wrong, but it certainly seems that way)

[17-Mar-2010 16:35:56] <rmatte> unless there is some mechanism now which forces them to load with python 2.4

[17-Mar-2010 16:38:15] <rmatte> checking right now...

[17-Mar-2010 16:38:59] <rmatte> well, they are all running off of /usr/local/zenoss/python/bin/.python.bin

[17-Mar-2010 16:39:09] <rmatte> so maybe there is something that forces it to use that

[17-Mar-2010 16:39:18] <rmatte> it's just weird that it doesn't seem to apply to other daemons

[17-Mar-2010 16:40:23] <rmatte> I'm going to open a trac ticket for this tomorrow (leaving in about 5 mins)

[17-Mar-2010 16:45:40] <rmatte> what a hassle lol

[17-Mar-2010 16:45:57] <rmatte> well, I'm out for the day, gnite folks

[17-Mar-2010 16:46:22] <mrayzenoss> later

[17-Mar-2010 18:33:53] <rhettardo> hello, i have two HP switches that have the wrong formula for the ports and are showing traffic in the Petabits/s.  How do i go about changing that?  i know how to change the math in the templates but not sure how to apply the template so this will stop.

[18-Mar-2010 00:00:30] [disconnected at Thu Mar 18 00:00:30 2010]

[18-Mar-2010 00:00:30] [connected at Thu Mar 18 00:00:30 2010]

[18-Mar-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[18-Mar-2010 05:48:10] Simon4_ is now known as Simon4

[18-Mar-2010 05:54:39] Simon4_ is now known as Simon4

[18-Mar-2010 09:52:15] <rmatte> woohoo, fun new bug: http://dev.zenoss.org/trac/ticket/6302

[18-Mar-2010 10:46:26] <venturaville> howdy mray

[18-Mar-2010 10:51:53] <rmatte> howdy howdy

[18-Mar-2010 10:53:53] <mrayzenoss> greetings

[18-Mar-2010 10:54:22] <venturaville> mray: I started adding pools/volumes support to the libvirt module (ala VMWare)

[18-Mar-2010 10:54:29] <mrayzenoss> badass

[18-Mar-2010 10:55:19] <mrayzenoss> I'm trying to encourage more testing on the stuff that hasn't been tested yet (OpenVZ, Xen, etc.)

[18-Mar-2010 10:55:27] <venturaville> I need some good documentation on the security model in zenoss, though, before I can add in the management code

[18-Mar-2010 10:56:16] <venturaville> a good example of how to setup a new user and tie it to certain things would be helpful

[18-Mar-2010 10:56:38] <venturaville> e.g. create/destroy/start/stop VMs...

[18-Mar-2010 10:58:49] <mrayzenoss> I think some of that is changing in the next release

[18-Mar-2010 10:59:17] <mrayzenoss> I'm not sure there's a good document on the current stuff

[18-Mar-2010 10:59:21] <venturaville> I'll probably wait until the next release to enable that part of the code then

[18-Mar-2010 11:01:31] <mrayzenoss> supposedly the UI team is writing lots of developer docs as they overhaul the UI

[18-Mar-2010 11:02:18] <mrayzenoss> there should be good developer docs for transitioning ZenPacks to the new UI if you've made UI changes.  There's also a backwards compatibility layer that covers a lot of things

[18-Mar-2010 11:02:34] <venturaville> das ist guht

[18-Mar-2010 11:46:41] <rmatte> hmmm

[18-Mar-2010 11:46:46] <rmatte> /usr/local/zenoss/python/lib/python2.4/site-packages/Products.CMFCore-2.1.2-py2.4.egg/Products/CMFCore/DirectoryView.py:497: UserWarning: DirectoryView HPTemperatureSensors refers to a non-existing path 'HPTemperatureSensors/skins/HPTemperatureSensors'

[18-Mar-2010 11:46:52] <rmatte> wonder where that info is kept

[18-Mar-2010 11:46:59] <rmatte> somehow it's remembering old directories

[18-Mar-2010 11:47:18] <mrayzenoss> I think the path changed in 2.5

[18-Mar-2010 11:47:26] <mrayzenoss> for a couple of libraries

[18-Mar-2010 11:47:54] <rmatte> well, this particular error is related to directories that I actually removed myself

[18-Mar-2010 11:48:01] <rmatte> but somehow something is remembering they were there

[18-Mar-2010 11:53:23] <rmatte> ah, seems to be somewhere in Zope

[18-Mar-2010 12:04:12] <rmatte> hmmm, I'm seeing negative values in graphs for packets and traffic regardless of using 64bit counters for those interfaces

[18-Mar-2010 12:04:17] <rmatte> any ideas as to why?

[18-Mar-2010 12:09:26] <rmatte> eugh, this whole no longer supporting zip zenpacks thing hurts, the HPTemperatureSensors ZenPack needs a lot of work done to it to convert it since it's datasources are like: $$ZENHOME/Products/HPTemperatureSensors/libexec/check_snmp_temperature.pl -H ${dev/manageIp} -C ${dev/zSnmpCommunity} -T hp -A'*'

[18-Mar-2010 12:17:05] <rmatte> eugh, I can't uninstall my zip ZenPacks from the UI or the commandline

[18-Mar-2010 12:17:07] <rmatte> lovely

[18-Mar-2010 12:21:18] <rmatte> oh, I can uninstall 2 of them, but not APC UPS PowerNet

[18-Mar-2010 12:21:25] <rmatte> presumably because there are spaces in the name

[18-Mar-2010 12:24:25] <rmatte> aha, finally got it removed

[18-Mar-2010 12:31:08] <etank> crowbar?

[18-Mar-2010 12:32:04] <venturaville> sledge hammer?

[18-Mar-2010 12:32:16] <etank> another good choice

[18-Mar-2010 12:32:16] <venturaville> trebuchet?

[18-Mar-2010 12:32:29] <etank> +1

[18-Mar-2010 12:32:59] <QbY> afternoon gents

[18-Mar-2010 12:33:02] <QbY> and glas

[18-Mar-2010 12:33:04] <QbY> gals

[18-Mar-2010 12:34:33] <QbY> i found zenoss last night browsing the web, and it appears to be the answer to my needs..  one question..  i need to monitor thousands of devices in different geogrpahic areas..  can i place remote monitors that feed the "mothership" -- when one of these devices fail i just wnat a simple report--not a major alarm.

[18-Mar-2010 12:35:57] <Simon4> Yes.

[18-Mar-2010 12:36:33] <Simon4> have a read/search on the website for either a) Enterprise, or b) remote collectors in core, and you should find some useful info

[18-Mar-2010 12:36:34] <QbY> how fast is it?  from each device (which is a cable modem) i need like four or five snmp queries

[18-Mar-2010 12:36:49] <Simon4> as in four or five oids?

[18-Mar-2010 12:37:18] <QbY> yeah

[18-Mar-2010 12:37:23] <QbY> and maybe a ping

[18-Mar-2010 12:37:26] <Simon4> should be very very fast

[18-Mar-2010 12:37:36] <QbY> this thing looks nice

[18-Mar-2010 12:37:44] <QbY> and i like that i can specify events

[18-Mar-2010 12:39:43] RoAk is now known as RoAkSoAx

[18-Mar-2010 12:40:43] <Simon4> QbY: one of our collectors gets through 13375 oid requests in 7.49 seconds across 146 boxes

[18-Mar-2010 12:42:06] <QbY> nice..............

[18-Mar-2010 12:42:13] <QbY> we have an old old old version of jffnms

[18-Mar-2010 12:42:20] <QbY> she's been good .. but its time to move on

[18-Mar-2010 12:42:26] <rmatte> nah I had to do zenpack --remove="APC UPS PowerNet"

[18-Mar-2010 12:42:36] <rmatte> I was originally trying it without the =

[18-Mar-2010 12:42:47] <rmatte> which generally works, but apparently not in this case

[18-Mar-2010 12:49:38] <rmatte> QbY: Zenoss is godly, and yes, it will do what you want it to

[18-Mar-2010 12:50:53] <rmatte> QbY: just one thing to clarify, when you say "remote monitoring" are you referring to a monitoring server at each site, or actual agents installed on each device?

[18-Mar-2010 12:51:05] <rmatte> If you're talking about agents, Zenoss is agentless monitoring, so the answer to that would be no

[18-Mar-2010 12:51:28] <rmatte> It is capable of working with remote collectors (remote servers collecting data and feeding it back to a central server)

[18-Mar-2010 12:55:12] <QbY> rmatte: what i want to do, is install a server at each site, that monitors 1000+ cable modems (4-5 oids + ping each)..  and that server graph levels, and report offline modems back to our main system

[18-Mar-2010 12:55:28] <QbY> but in the main system it not be a true "emergency" just creates a ticket in our ticket system

[18-Mar-2010 12:56:34] <rmatte> well, what is your ticketing system?

[18-Mar-2010 13:01:38] <QbY> its one built internally

[18-Mar-2010 13:01:50] <QbY> but from my reading last night it looks like i can define an event that would call a shell scrpt

[18-Mar-2010 13:02:02] <QbY> so from that shell script it would just call a stored procedure on the sql server

[18-Mar-2010 13:02:09] <QbY> man, i want to install this now

[18-Mar-2010 13:02:24] <QbY> plus i'm gonna hook it into our phone system, so that when a major event happens, it will call :)

[18-Mar-2010 13:29:54] <QbY> wish this thing sat on postgres

[18-Mar-2010 13:30:45] <mrayzenoss> QbY: it's open source, I'd love to see a patchset

[18-Mar-2010 13:31:12] <mrayzenoss> the MySQL code is pretty compartamentalized

[18-Mar-2010 13:31:46] <mrayzenoss> Most is in http://dev.zenoss.org/trac/browser/trunk/Products/ZenEvents

[18-Mar-2010 13:32:13] <mrayzenoss> someone recently asked me about porting to Oracle, gave them the same answer :)

[18-Mar-2010 13:36:50] <rhettardo> hi

[18-Mar-2010 13:37:45] <rhettardo> im having some issues w/ two of my cheap hp switches and bits/s being reported wrong

[18-Mar-2010 13:38:23] <rhettardo> to the tune of petabits a second when a port is only pushing megabits a second

[18-Mar-2010 13:39:36] <rhettardo> i can see when i fetch the rrd that the its all screwedup, some times its fine (1268934300: 2.9126219205e+05) and other times its way off (1268937000: 2.0900690515e+14)

[18-Mar-2010 13:40:38] <rhettardo> any one know what is up with that wackiness? for what its worth i monitor all the same devices in cacti on custom templates and dont have an issue

[18-Mar-2010 13:42:34] <rmatte> rhettardo: perhaps you need to use 64bit counters versus 32bit or vice versa?

[18-Mar-2010 13:42:55] <QbY> the zenpacks--we can design our own too..  nice

[18-Mar-2010 13:42:55] <rmatte> I monitor TONS of HP switches, and I've never seen an issue like that with Zenoss

[18-Mar-2010 13:42:59] <QbY> ilove this

[18-Mar-2010 13:43:18] <rmatte> QbY: yeh, it's a really nice product

[18-Mar-2010 13:47:35] * forsberg <3 zenoss

[18-Mar-2010 13:50:04] <rhettardo> rmatte: i dont recall how to change from 64 to 32bit counters per device.  could you point me in the right direction?

[18-Mar-2010 13:50:24] <rmatte> rhettardo: it's not done per device

[18-Mar-2010 13:50:27] <rmatte> it's done per interface

[18-Mar-2010 13:50:33] <rmatte> and it's based on the template you're using

[18-Mar-2010 13:50:47] <rmatte> by default there's the ethernetCsmacd and ethernetCsmacd_64

[18-Mar-2010 13:50:58] <rmatte> it auto-assigns templates based on the interface type listed

[18-Mar-2010 13:51:04] <rmatte> you'll see the type when you click on the interface

[18-Mar-2010 13:51:24] <rmatte> The counter idea is just a guess though, it could be something else entirely

[18-Mar-2010 13:51:28] <rmatte> is it happening on every switch?

[18-Mar-2010 13:51:40] <rhettardo> no, just these two hp 1800-24's

[18-Mar-2010 13:52:30] * Simon4 coped ethernetCsmacd_64 to ethernetCsmacd under /Devices/Network and left it at that, meant all switches/routers defaulted to 64 bit counters

[18-Mar-2010 13:52:36] <Simon4> that might not be the "correct" way however

[18-Mar-2010 14:04:20] <rmatte> Simon4: well, if it works then fine

[18-Mar-2010 14:09:48] mrayzenoss1 is now known as mrayzenoss

[18-Mar-2010 14:09:57] <Simon4> rmatte: yeah, seems a nice global way to do it for network kit anyay

[18-Mar-2010 14:10:01] <Simon4> anyway, even

[18-Mar-2010 14:10:24] * Simon4 had awful issues with cisco kit and 32 bit counters a long time ago

[18-Mar-2010 14:13:25] <rhettardo> maybe its an issue w/ this switch reporting incorrectly w/ the 64 bit counters, i just changed it over to the 32bit oid's so we'll see what happens

[18-Mar-2010 14:13:28] <forsberg> is there some obvious way that i dont get, to have more devices with same ip ?

[18-Mar-2010 14:20:04] <QbY> does anyone know of a fully working demo of zenoss on the web?

[18-Mar-2010 14:21:38] <forsberg> it takes almost no hardware and no time to set it up  for something basic

[18-Mar-2010 14:21:59] <forsberg> small vm somewhere and like 1 hour and you are adding devices

[18-Mar-2010 14:22:42] <mrayzenoss> forsberg: no, that doesn't work well

[18-Mar-2010 14:22:50] <forsberg> ok?

[18-Mar-2010 14:22:51] <forsberg> :)

[18-Mar-2010 14:23:13] <forsberg> worked fine for me for testing it

[18-Mar-2010 14:23:23] <mrayzenoss> eventually Zenoss will notice you have dupes

[18-Mar-2010 14:23:27] <forsberg> ahh sorry

[18-Mar-2010 14:23:34] <forsberg> with the devices, yes

[18-Mar-2010 14:23:46] <mrayzenoss> Egor put this up: people/bigegor/blog/2010/02/22/virtualip-zenpack

[18-Mar-2010 14:24:10] <mrayzenoss> might be useful depending on how you're monitoring the same IP

[18-Mar-2010 14:24:42] <forsberg> i have a linux device (my mailserver) and wanted to use the synthetictransaction thing on the mailserver also

[18-Mar-2010 14:25:14] <forsberg> and i have other places where i have a loadbalancer where i want to monitor some ip services on the vip only

[18-Mar-2010 14:25:37] <forsberg> but thanks, ill try it out

[18-Mar-2010 14:32:32] * Simon4 found a way to add devices with duplicate IP's, but it needs some zendmd and zenperfsnmp has a whinge, but still works fine

[18-Mar-2010 14:32:43] <Simon4> you can't do it via the GUI though

[18-Mar-2010 14:33:08] <Simon4> we have a patch to support snmp contexts, so had to be able to add a bunch of devices with the same target IP, but a different context

[18-Mar-2010 14:35:34] <forsberg> k

[18-Mar-2010 14:36:34] <Simon4> I also used it for something similar to what you're suggesting

[18-Mar-2010 14:37:04] <forsberg> yeah i mean, i cant put proper ip service monitoring on a device if it only have the ip sometimes

[18-Mar-2010 14:51:08] <Brixius> Question, I've got a "Custom" tab on my device nodes that just gives me an error when I click on it, I've installed a bunch of zenpacks from the community packs(and uninstalled a bunch) I'm wondering if there's an easy way to get rid of it, or find out what put it there.  I can post the error text to pastebin if that would help.

[18-Mar-2010 14:51:41] <bmac2> anyone know how to make the ?status/Heartbeat  localhost zenperfwbem heartbeat failure error go away on a debian linux monitoring?

[18-Mar-2010 14:51:58] <rmatte> Brixius: what packs have you installed?

[18-Mar-2010 14:52:03] <bmac2> I tried every item I could find online

[18-Mar-2010 14:52:11] <forsberg> !! :)

[18-Mar-2010 14:52:19] <bmac2> and all different monitoring schemes

[18-Mar-2010 14:52:28] <rmatte> bmac2: you mean after removing the pack?

[18-Mar-2010 14:52:37] <bmac2> no

[18-Mar-2010 14:52:44] <bmac2> haven't removed anything

[18-Mar-2010 14:52:48] <bmac2> trying to set it all up

[18-Mar-2010 14:52:50] <Brixius> rmatte: I don't remember exactly as I've installed and removed a bunch.

[18-Mar-2010 14:53:10] <bmac2> I installed the zenplugin.py on the debian box

[18-Mar-2010 14:53:26] <rmatte> well, it's weird that you're even getting the heartbeat errors...

[18-Mar-2010 14:53:30] <rmatte> is it working despite the errors?

[18-Mar-2010 14:53:36] <bmac2> yep

[18-Mar-2010 14:53:45] <rmatte> well, there is one way to eliminate them

[18-Mar-2010 14:54:01] <rmatte> mysql -uzenoss -pzenoss events -e "delete from heartbeat where component in ('zenperfwbem')"

[18-Mar-2010 14:54:03] <bmac2> I go to status and everything is green eXCEPT zenperfwbem

[18-Mar-2010 14:54:08] <bmac2> k

[18-Mar-2010 14:54:11] <bmac2> let me do that

[18-Mar-2010 14:54:16] <rmatte> well wait

[18-Mar-2010 14:54:23] <bmac2> k

[18-Mar-2010 14:54:26] <rmatte> what do you mean you go to status?

[18-Mar-2010 14:54:32] <rmatte> where exactly are you looking?

[18-Mar-2010 14:54:38] <bmac2> when I click on the tab

[18-Mar-2010 14:54:41] <bmac2> in the webpage

[18-Mar-2010 14:54:44] <rmatte> which tab

[18-Mar-2010 14:54:48] <Brixius> I did a bad thing buy installing every zenpack in the community collection at one point(just installed everything from a loop in the command line) After doing that I realized it was a mistake.

[18-Mar-2010 14:54:53] <bmac2> the very first one taht says status

[18-Mar-2010 14:54:59] <rmatte> right, but for what?

[18-Mar-2010 14:55:00] <bmac2> click on device list

[18-Mar-2010 14:55:04] <bmac2> then click on the debian machine

[18-Mar-2010 14:55:09] <rmatte> every device has a status tab

[18-Mar-2010 14:55:10] <bmac2> then click on status for that machine

[18-Mar-2010 14:55:15] <rmatte> ok

[18-Mar-2010 14:55:19] <rmatte> so if you go to events

[18-Mar-2010 14:55:23] <rmatte> do you see an event there?

[18-Mar-2010 14:55:28] <bmac2> I have an orange error

[18-Mar-2010 14:55:32] <rmatte> which says?

[18-Mar-2010 14:55:34] <bmac2> that I posted above, and it sends an email

[18-Mar-2010 14:55:41] <rmatte> I see

[18-Mar-2010 14:55:48] <rmatte> so just to clarify...

[18-Mar-2010 14:55:56] <rmatte> you actually added the monitoring server itself in to Zenoss?

[18-Mar-2010 14:56:09] <bmac2>   !  zenperfwbem   /Status/Heartbeat  localhost zenperfwbem heartbeat failure

[18-Mar-2010 14:56:12] <bmac2> yes

[18-Mar-2010 14:56:17] <bmac2> is that the issue?

[18-Mar-2010 14:56:21] <bmac2> I got it monitoring itself?

[18-Mar-2010 14:56:26] <rmatte> well, that's why you're getting the emails for one thing

[18-Mar-2010 14:56:30] <rmatte> since it sees it as a valid device

[18-Mar-2010 14:56:45] <rmatte> if you removed it, you'd still see the heartbeat errors but it wouldn't email

[18-Mar-2010 14:56:52] <rmatte> basically you have 2 options...

[18-Mar-2010 14:56:53] <bmac2> I want ot monitor the disk space on the debain box

[18-Mar-2010 14:56:59] <bmac2> and a couple of other things

[18-Mar-2010 14:57:04] <rmatte> you could add a filter to your alerting rule so that it doesn't alert for that one specific error

[18-Mar-2010 14:57:08] <rmatte> you could remove the device

[18-Mar-2010 14:57:19] <rmatte> or you can use the command I pasted above to eliminate the heartbeat errors

[18-Mar-2010 14:57:21] <rmatte> your call

[18-Mar-2010 14:57:29] <bmac2> which would be a more clean solution"

[18-Mar-2010 14:57:42] <rmatte> adding the filter honestly

[18-Mar-2010 14:57:45] <bmac2> k

[18-Mar-2010 14:57:47] <bmac2> let me try that

[18-Mar-2010 14:57:52] <bmac2> cause it is the only error I get

[18-Mar-2010 14:57:53] <rmatte> though it's weird that you're event getting those heartbeat errors

[18-Mar-2010 14:57:57] <rmatte> how often do they occur?

[18-Mar-2010 14:57:58] <bmac2> and I am monitoring a ton of stuff

[18-Mar-2010 14:58:05] <rmatte> how much stuff?

[18-Mar-2010 14:58:08] <bmac2> I have got 37 of them

[18-Mar-2010 14:58:15] <rmatte> hmmm, 37 isn't excessive

[18-Mar-2010 14:58:19] <bmac2> between 13:21 and 13:57

[18-Mar-2010 14:58:35] <bmac2> so around one a minute

[18-Mar-2010 14:58:35] <Brixius> rmatte: ok I did some more searching and found that there is a cutomTab in ZenPacks.example.Techniques which I had installed then uninstalled. but I think it left something behind.

[18-Mar-2010 14:59:18] <rmatte> bmac2: just use the command I pasted, if you ever need the heartbeat errors back you can just reinstall the Zenpack

[18-Mar-2010 14:59:26] <bmac2> k

[18-Mar-2010 14:59:30] <bmac2> so delete it from the database

[18-Mar-2010 14:59:45] <rmatte> Brixius: so install it again, then remove it again and see if the tab goes away

[18-Mar-2010 14:59:54] <rmatte> Brixius: that ZenPack is really only intended for developers though

[18-Mar-2010 14:59:55] <bmac2> I have 12 items I am monitoring

[18-Mar-2010 15:00:11] <rmatte> Brixius: not sure why you installed it in the first place unless you were intending to code a complex ZenPack

[18-Mar-2010 15:00:27] <bmac2> a juniper firewall cluster, 7 freebsd vms, 2 Xservers and debian box

[18-Mar-2010 15:00:35] <rmatte> cool

[18-Mar-2010 15:05:56] <bmac2> hmmm

[18-Mar-2010 15:06:11] <bmac2> won't let em connect ot the database from the command line on that machine

[18-Mar-2010 15:14:52] <Brixius> rmatte: ya it was a stupid thing to do, but hey, I've been stupid before.... it errors when I try to install it again.

[18-Mar-2010 15:15:03] <rmatte> what's the error?

[18-Mar-2010 15:15:29] <Brixius> ERROR: zenpack command failed. Reason: Exception: Unable to create object using the following attributes:

[18-Mar-2010 15:15:29] <Brixius>   * id: /zport/dmd/Reports/Example Reports/Cluster Requests

[18-Mar-2010 15:15:29] <Brixius>   * module: Products.ZenModel.Report

[18-Mar-2010 15:15:29] <Brixius>   * class: Report

[18-Mar-2010 15:15:49] <rmatte> restart zenoss and then try installing again

[18-Mar-2010 15:25:02] <Brixius> nope same thing

[18-Mar-2010 15:25:53] <rmatte> ssh in to the server and become the zenoss user, then navigate to $ZENHOME/Products/ZenReports/reports

[18-Mar-2010 15:26:10] <rmatte> See if there's a folder in there called Example_Reports

[18-Mar-2010 15:28:32] <rmatte> also, just fyi, I'm on a tight schedule to get stuff done today, so the quicker the better please

[18-Mar-2010 15:28:59] <Brixius> no, there's no folder there called Example_Reports

[18-Mar-2010 15:29:10] <rmatte> k, as the zenoss user execute: python $ZENHOME/Products/ZenReports/ReportLoader.py --force

[18-Mar-2010 15:29:18] <rmatte> then try installing the pack again

[18-Mar-2010 15:30:00] <Brixius> ok, I'll give it a try

[18-Mar-2010 15:30:07] <Brixius> just ran that

[18-Mar-2010 15:30:13] <Brixius> I'll install now,

[18-Mar-2010 15:30:29] <rmatte> k

[18-Mar-2010 15:31:07] <rmatte> If it doesn't install then I'd like you to create a directory called Example_Reports in $ZENHOME/Products/ZenReports/reports

[18-Mar-2010 15:31:21] <rmatte> then execute: python $ZENHOME/Products/ZenReports/ReportLoader.py --force

[18-Mar-2010 15:31:25] <rmatte> then try installing again

[18-Mar-2010 15:31:31] <Brixius> k

[18-Mar-2010 15:31:43] <rmatte> and if it fails again, restart zenoss, then try

[18-Mar-2010 15:31:49] <rmatte> if all of that doesn't work, I'm stumped

[18-Mar-2010 15:32:53] <Brixius> ok, if this doesn't work, it's a fairly new install, and I've learned alot doing it, so I might just reinstall and be more careful as to the zenpacks I install...

[18-Mar-2010 15:34:19] <rmatte> k

[18-Mar-2010 15:34:23] <rmatte> that might be a good idea

[18-Mar-2010 15:34:33] <rmatte> be careful to only install ZenPacks that you actually need

[18-Mar-2010 15:34:44] <rmatte> if you just go randomly installing stuff before reading you'll end up in situations like this

[18-Mar-2010 15:35:17] <Brixius> that be one of the things I learned

[18-Mar-2010 15:35:23] <rmatte> hehe

[18-Mar-2010 15:35:24] <Simon4> + backup the zopedb before installing anything

[18-Mar-2010 15:35:31] <Simon4> a restore is so much easier than backing out

[18-Mar-2010 15:35:48] <Simon4> in my experience at least

[18-Mar-2010 15:38:34] <forsberg> Brixius thats what i learned, to be abit cautios with the zenpacks if you are in "production" :)

[18-Mar-2010 15:41:48] <Brixius> well at least this isn't a production install "yet" I was preping a box to replace our old production system running 2.1.92...

[18-Mar-2010 15:42:31] <Simon4> Brixius: costs nothing to run a vm that contains a restore of your live ssytem to test installing crap on... ignore me if I'm preaching to the choir btu that's what we do here

[18-Mar-2010 15:42:43] <forsberg> me2 now

[18-Mar-2010 15:43:41] <Brixius> ya, I'm running on ESX as well. I'll make sure to snapshot the system before test installing anything from now on.

[18-Mar-2010 15:44:27] <Brixius> I do that with prod stuff all the time, but since this wasn't prod, I didn't... Live and learn

[18-Mar-2010 15:45:38] <rmatte> yeh, I use snapshots religiously

[18-Mar-2010 15:45:42] <rmatte> great for testing

[18-Mar-2010 15:45:51] <Simon4> if you want to keep perfdata (rrd files) a zenbackup --no-perfdata --no-eventsdb gets a quick copy of the config db that you can tthen restore

[18-Mar-2010 15:45:56] <Simon4> that's if perfdata is on same box

[18-Mar-2010 15:46:07] <Simon4> since a restore of snapshot may kill that depending on filesystem layout

[18-Mar-2010 15:46:27] <rmatte> well, that's why I do the snapshot right beforehand

[18-Mar-2010 15:46:32] <Simon4> people seem to get grumpy at gaps in their graphs :)

[18-Mar-2010 15:46:34] <rmatte> rather than use a really old one

[18-Mar-2010 15:46:55] <Simon4> rmatte: yeah, our tenants whinge at even tiny gaps, makes life... entertaining

[18-Mar-2010 15:47:05] <rmatte> hehe

[18-Mar-2010 15:47:36] <Simon4> wonder sometimes if they do nothing else but look at reports :D

[18-Mar-2010 15:51:22] <forsberg> we have standing joke also, boss is yelling and everyone laughing even for a 5-10 min gap

[18-Mar-2010 15:51:49] <forsberg> always my fault :/

[18-Mar-2010 15:51:54] <forsberg> :)

[18-Mar-2010 15:52:18] <rmatte> lol

[18-Mar-2010 15:52:42] <forsberg> (im doing 99% of the zenoss work)

[18-Mar-2010 15:53:09] <Simon4> forsberg: we get "Incident" tickets raised for "my requests graph had a gap, was there an outage?"

[18-Mar-2010 15:53:21] <forsberg> lol

[18-Mar-2010 15:53:26] <Simon4> I either need better education or a larger hammer ;)

[18-Mar-2010 15:53:39] <forsberg> well we only used it internally, so its only my boss

[18-Mar-2010 15:53:40] <forsberg> heheh

[18-Mar-2010 15:53:58] <forsberg> and now i appriciate its only boss

[18-Mar-2010 15:54:02] * Simon4 has about 10 blog posts just waiting to be written, one of these days

[18-Mar-2010 15:54:36] <Simon4> currently working on how to make a system with up to 1000 datapoints per device scale nicely

[18-Mar-2010 15:57:05] <forsberg> i would do it the bofh way Simon4 "you want graphs without gaps ? *clickedeclick" ---- "yeah...ups now there is no graphs at all????????!?" -- "thats right"

[18-Mar-2010 15:57:22] <Simon4> heheehe :)

[18-Mar-2010 15:57:34] <Simon4> "where did those rrd files go?"

[18-Mar-2010 15:58:14] <forsberg> http://bofh.ch/ to those who forgot the bastard operator from hell

[18-Mar-2010 15:58:22] <Simon4> he was a kiwi and all

[18-Mar-2010 15:58:33] <Simon4> from memory

[18-Mar-2010 16:47:40] <mrayzenoss> Anyone in Texas?  More free Zenoss training: blogs/zenossblog/2010/03/18/zenoss-community-day-april-9th--austin-tx

[18-Mar-2010 16:47:52] <Simon4> want to fly people to texas?

[18-Mar-2010 16:47:54] <Simon4> :)

[18-Mar-2010 16:48:06] <Simon4> although, texas lacks snow

[18-Mar-2010 16:48:54] <davidX-> spiderhouse is pretty sweet

[18-Mar-2010 16:49:02] <davidX-> i'll try to make it down (in dfw)

[18-Mar-2010 16:49:11] <mrayzenoss> we've had people fly across country for the free training, since it's cheaper than the pay training

[18-Mar-2010 16:49:22] <mrayzenoss> plus the Texas LinuxFest is the next day

[18-Mar-2010 16:49:34] <Simon4> mrayzenoss: I'm in the UK, but tempting

[18-Mar-2010 16:50:01] <mrayzenoss> heh

[18-Mar-2010 16:50:07] <mrayzenoss> Jane Curry does some training

[18-Mar-2010 16:50:11] <mrayzenoss> in the UK

[18-Mar-2010 16:50:13] <Simon4> more to meet people I think vs learning things looking at the overview

[18-Mar-2010 16:50:17] <Simon4> yeah, I've read her papers

[18-Mar-2010 16:50:18] <Simon4> very cool

[18-Mar-2010 16:50:33] <mrayzenoss> http://www.skills-1st.co.uk/

[18-Mar-2010 16:50:47] <mrayzenoss> 3 days of Event Training and 1 day Getting Started

[18-Mar-2010 16:52:01] <mrayzenoss> sweet, they're near Slough, like in The Office

[18-Mar-2010 16:52:19] <Simon4> yeah

[18-Mar-2010 16:52:30] * Simon4 is at the one and only BBC

[18-Mar-2010 16:52:48] <Simon4> spent a bit of time with Nathaniel in the past

[18-Mar-2010 16:52:52] <Simon4> good guy

[18-Mar-2010 16:53:34] <mrayzenoss> Simon4: is this you? http://dev.zenoss.org/trac/ticket/5894

[18-Mar-2010 16:53:54] <Simon4> that's one of our devs, not me

[18-Mar-2010 16:53:54] <mrayzenoss> hmm.. probably not now that I've looked at the email address

[18-Mar-2010 16:54:02] * Simon4 is responsible for the install/architecture

[18-Mar-2010 16:54:15] <Simon4> matus is one of our monitoring engineering guys

[18-Mar-2010 16:54:22] <Simon4> who writes our custom zenpacks/does other dev

[18-Mar-2010 16:54:39] <Simon4> I'm the guy who keeps it running and tries to make it scale :D

[18-Mar-2010 17:36:32] bzed_ is now known as bzed

[18-Mar-2010 17:40:16] <Brixius> is there a way to tell zenmib to load mibs into a folder?

[18-Mar-2010 18:00:39] <QbY> Simon, probably 100 devices all over the place (country) .. mainly routers, etc..

[18-Mar-2010 18:01:10] <Simon4> if you're looking at a single install, for 100 devices you don't need a heap given today's hardware

[18-Mar-2010 18:01:44] <QbY> now, in the VERY near future, i want to add remove sensors/collectors to monitor cable modems at the local site (each site will have about 1000-2000 modems each)

[18-Mar-2010 18:01:47] <Simon4> single 4-core proc + meh... 4 GB ram would be enough, main thing I've found is to put the rrd /zopedb/mysql on seperate spindles

[18-Mar-2010 18:01:56] <Simon4> since they are all databases in the end

[18-Mar-2010 18:02:07] <Simon4> so sharing spindles = pain, as per normal scaling rules

[18-Mar-2010 18:02:20] <QbY> spindles?

[18-Mar-2010 18:02:30] <Simon4> hard drive spindles

[18-Mar-2010 18:02:34] <QbY> oh ok

[18-Mar-2010 18:02:53] <QbY> i'm telling you, the more i read about this thing, the more i like it

[18-Mar-2010 18:03:04] <QbY> groundwork seems like its just a cute front end to nagios

[18-Mar-2010 18:03:35] <Simon4> basically it's an awesome front end to rrdtool

[18-Mar-2010 18:03:43] <Simon4> if you want to take it at it's most basic

[18-Mar-2010 18:04:02] <Simon4> but the layers on top are the bits that make the people who pay us happy :)

[18-Mar-2010 18:04:27] <QbY> hehe..  now one question..  and i haven't got to play with a demo, so  may be asking a dumb question...  zenoss generates rrd tools--can i link that graphic into another website?

[18-Mar-2010 18:05:10] <Simon4> basically, you can expose the rrd files anywhere you like

[18-Mar-2010 18:05:18] <QbY> nice

[18-Mar-2010 18:05:37] <Simon4> all the perf data storage is just in rrd files

[18-Mar-2010 18:05:42] <QbY> cause currently i have rrd generated by our current nms, then one that our customers see generated by another script

[18-Mar-2010 18:05:49] <Simon4> so yeah, any fronend you can think of is possible

[18-Mar-2010 18:06:28] <QbY> so, what's the bad..

[18-Mar-2010 18:06:32] <QbY> there has to be somethin gbad

[18-Mar-2010 18:06:35] <QbY> other than mysql

[18-Mar-2010 18:07:25] <Simon4> heh

[18-Mar-2010 18:07:36] <Simon4> mysql is "manageable"

[18-Mar-2010 18:07:42] <Simon4> as in normal situations

[18-Mar-2010 18:08:03] <QbY> hehehe..  we are a postgresql house

[18-Mar-2010 18:08:12] <QbY> but, will make exceptions

[18-Mar-2010 18:08:48] <Simon4> we've found the time to create things like templates and reports can be a lot when you get a whole lot of oids - but that's looking at multiple "tenants' with multiple "applications" with multiple "kpis"

[18-Mar-2010 18:08:53] <Simon4> so we're a special case

[18-Mar-2010 18:09:10] <Simon4> asnd even then we've found scripting ways to create all that dynamically which is awesome

[18-Mar-2010 18:09:19] <QbY> let me get this installed

[18-Mar-2010 18:11:04] <Simon4> and there are a bunch of people in here who can pretty much help with anything, given the right question  :)

[18-Mar-2010 18:12:38] <QbY> Simon, any issue running this inside a virtual machine?

[18-Mar-2010 18:15:27] <Simon4> not at all, other than maybe disk IO if your install grows

[18-Mar-2010 18:15:39] <Simon4> biut with 100 devices you should be fine

[18-Mar-2010 18:15:51] <Simon4> disk IO seemed to be the biggest hit for us

[18-Mar-2010 18:16:00] <QbY> remember my remote collectors

[18-Mar-2010 18:16:20] <Simon4> yup

[18-Mar-2010 18:16:52] <QbY> I can put a 1x Quad-Core 2.5Ghz 4GB RAM 320GB server... how's that?

[18-Mar-2010 18:17:07] <Simon4> sounds reasonable to me

[18-Mar-2010 18:17:31] <Simon4> basically, the things that made zenoss struggle for me were

[18-Mar-2010 18:17:58] <Simon4> a) untuned mysql events db - makes it really really slow to navigate around the UI and heartbeats fail occasionally

[18-Mar-2010 18:18:13] <QbY> OS--64bit or 32bit?

[18-Mar-2010 18:18:21] <Simon4> and b) slow disk means you can't update a bunch of rrd files in a hurry

[18-Mar-2010 18:18:31] <Simon4> 64bit, no reason not to these days

[18-Mar-2010 18:19:17] <Simon4> even with remote collectors we rsync the rrd data from  collectors to the master server, so that multi-graph reports across multiple datacentres work just fine

[18-Mar-2010 18:19:34] <Simon4> not something you'll get as standard buit easy to make work

[18-Mar-2010 18:27:00] <QbY> what kinda disk space do you have allocated?

[18-Mar-2010 18:41:19] * Simon4 looks

[18-Mar-2010 19:04:56] <Simon4> zenoss@mon001:~$ du -shc /opt/zenoss/perf/Devices

[18-Mar-2010 19:04:56] <Simon4> 1.2G    /opt/zenoss/perf/Devices

[18-Mar-2010 19:04:57] <Simon4> 1.2G    total

[18-Mar-2010 19:05:12] <Simon4> that's about 150 devices with about 200 oids/device

[18-Mar-2010 19:05:22] <Simon4> for rrd at elast

[18-Mar-2010 19:13:54] <QbY> well i got a lot more than that.. :)  should be good on storage--starting my install nwo

[18-Mar-2010 19:18:50] <QbY> hey Simon.  Brand new server running CenOS 64-bit.  Which would you recommend,   Source, RPM, or Virtual Appliance?

[18-Mar-2010 19:19:54] <Simon4> I can't see any issue with RPM

[18-Mar-2010 19:20:05] <Simon4> that way you can tune your mysql DB etc

[18-Mar-2010 19:20:15] <Simon4> can't see any gain with source tbh

[18-Mar-2010 19:20:36] <Simon4> if you don't have any mysql dba's, I guess go the appliance/stack

[18-Mar-2010 19:21:33] <QbY> hrmm

[18-Mar-2010 19:21:46] <QbY> i don't think i've ever installed an appliance anything

[18-Mar-2010 19:23:51] * Simon4 isn't a fan

[18-Mar-2010 19:24:08] <Simon4> but

[18-Mar-2010 19:24:26] <Simon4> mysql tuning for the events db has turned out to be a huge thing(tm) for us

[18-Mar-2010 19:24:45] <Simon4> so if you're going to go with a default mysql install, the appliance may be better

[18-Mar-2010 19:25:00] <Simon4> maybe try the appliance and steal the my.cnf settings then go the rpm?

[18-Mar-2010 19:25:14] <Simon4> gives you an instant "dba" type install :)

[18-Mar-2010 19:39:59] <QbY> per the docs... this says, "Install the Core ZenPacks" -- where do I find the "Core ZenPack"

[18-Mar-2010 19:47:29] <rhettardo> what os are you running

[18-Mar-2010 19:48:09] <rhettardo> http://sourceforge.net/projects/zenoss/files/

[18-Mar-2010 19:48:53] <QbY> i found the "Bundle"

[18-Mar-2010 19:48:56] <QbY> file name matched

[18-Mar-2010 19:49:03] <QbY> it didn't really make sense int he guide

[18-Mar-2010 19:49:14] <QbY> its all isntalled--now for my first device..

[18-Mar-2010 20:44:18] <QbY> can someone help me with locations/systems/groups

[18-Mar-2010 20:44:27] <QbY> or point me to something that explains their purpose

[18-Mar-2010 20:52:54] <salax> QbY: location : for positioning / location purpose to be mapped with google map api

[18-Mar-2010 20:53:09] <salax> location of servers /datacenters .. etc2

[18-Mar-2010 20:53:27] <QbY> we manage networks for people all over the country

[18-Mar-2010 20:53:44] <QbY> so, a location would be the city?

[18-Mar-2010 20:53:51] <QbY> and group would be the customer name?

[18-Mar-2010 20:54:11] <salax> QbY: a location would be something like address that are recognized by google map

[18-Mar-2010 20:55:56] <QbY> or should i get real specific, create a location for South Carolina, and then a sub location for columbia, sc

[18-Mar-2010 21:00:53] <QbY> for a "Device" -- What is "Title"

[18-Mar-2010 21:52:45] <QbY> is it possible to set  the timeout for the snmpwalk when adding a device?

[19-Mar-2010 00:00:30] [disconnected at Fri Mar 19 00:00:30 2010]

[19-Mar-2010 00:00:30] [connected at Fri Mar 19 00:00:30 2010]

[19-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[19-Mar-2010 01:02:06] <QbY> Trying to configure my Zenoss; couple of questions..  First, trying to add a CMTS (a type of router--Cisco) and it never adds all of the interfaces--snmp times out..  Do I need to build a special class for this?

[19-Mar-2010 01:03:23] <QbY> Also, creating some data sources for my CMTSes..  I want to build--capturing two values (i have the oids) one is total modems, one is registered modems..  I need to write a threshhold if the number of registered modems is less than 1/2 of the total modems..  where is the "language" for Min/Max value found..

[19-Mar-2010 06:08:10] forsberg is now known as fOrsberg

[19-Mar-2010 06:30:24] fOrsberg is now known as forsberg

[19-Mar-2010 07:38:49] forsberg is now known as fOrsberg

[19-Mar-2010 07:41:11] tehhobbi1 is now known as tehhobbit

[19-Mar-2010 08:37:06] <whitemice> Is it possible to query ZenOSS for an monitored value (via SOAP, XML-RPC?)

[19-Mar-2010 09:19:59] <felbane> Anyone available to answer a zenpack question?

[19-Mar-2010 09:20:50] <rmatte> well, depending on what the question is, yes

[19-Mar-2010 09:21:07] <felbane> I am building a page template (skin) to allow editing of my custom datasource, but I can't figure out how to make Zenoss recognize changes in the file

[19-Mar-2010 09:21:29] <rmatte> I've never done any work with skins, so not sure

[19-Mar-2010 09:21:36] <rmatte> have you read Jane Curry's ZenPack paper?

[19-Mar-2010 09:21:37] <felbane> Do I have to restart something?

[19-Mar-2010 09:21:44] <rmatte> well yeh

[19-Mar-2010 09:21:51] <rmatte> you need to restart after any advanced zenoss install

[19-Mar-2010 09:21:55] <rmatte> (restart zenoss that is)

[19-Mar-2010 09:22:06] <felbane> Yeah but it doesn't tell you how to iteratively develop skins files :-)

[19-Mar-2010 09:22:14] <rmatte> you can't

[19-Mar-2010 09:22:26] <rmatte> it's literally modify, restart, see how it looks, modify, restart, repeat

[19-Mar-2010 09:22:55] <rmatte> which is something that they are going to look at developing a solution for in the future

[19-Mar-2010 09:23:04] <felbane> I guess this is why we have development boxes o_O

[19-Mar-2010 09:23:10] <rmatte> correcto

[19-Mar-2010 09:23:21] <rmatte> you don't want to be doing that type of development on prod installs

[19-Mar-2010 09:23:55] <felbane> Okay, just wanted to make sure I wasn't missing something

[19-Mar-2010 09:23:59] <felbane> Ty

[19-Mar-2010 09:25:12] <rmatte> np

[19-Mar-2010 09:52:37] <tuples> last night a Device template was applied to all servers that pretty much wiped the performance template applied to servers

[19-Mar-2010 09:53:06] <tuples> I can remove it in the servers zproperties by deleting the zDeviceTemplates local property

[19-Mar-2010 09:53:13] <tuples> but I cannot find its source :(

[19-Mar-2010 10:00:45] <rmatte> tuples: you understand the concept of binding templates?

[19-Mar-2010 10:24:13] <QbY> morning/afternoon/evening all

[19-Mar-2010 10:24:31] <Simon4> hiya

[19-Mar-2010 10:24:40] <QbY> i got my install working :)

[19-Mar-2010 10:24:57] <QbY> however, gots some issues--i  was hoping you'd be here to guide me

[19-Mar-2010 10:25:02] <Simon4> heh

[19-Mar-2010 10:25:09] <Simon4> I can try :)

[19-Mar-2010 10:25:51] <QbY> first, a cmts is a router, but has specialized interfaces--problem is, when you snmpwalk a cmts you can wait 30 minutes for it to complete..  how can i have it only walk interfaces when adding it, so that it doesn't time out and quit--adding the device but not the interfaces

[19-Mar-2010 10:38:04] <rmatte> QbY: remove all collector plugins except for the interfacemap plugin on that device

[19-Mar-2010 10:38:21] <QbY> ok

[19-Mar-2010 10:38:37] <QbY> where can i find a description of those collector plugins

[19-Mar-2010 10:39:24] <QbY> by the way--just for the record, zenoss is a nice NICE piece of software--so any devs in the room, next time you're in ATL find me for a round of beers

[19-Mar-2010 10:50:35] <rmatte> hehe

[19-Mar-2010 10:50:49] <rmatte> wait until you see the new interface in the next release

[19-Mar-2010 10:50:54] <QbY> oh shit

[19-Mar-2010 10:51:06] <QbY> i like this one already--i can only imagine improvemnet

[19-Mar-2010 10:51:25] <rmatte> http://public-demo.zenoss.com/zport/dmd/itinfrastructure?submitted=true

[19-Mar-2010 10:51:29] <rmatte> login with admin/zenoss

[19-Mar-2010 10:51:35] <QbY> so, can anyone point me to the functionliaty of zenoss.snmp.NewDeviceMap/DeviceMap/InterfceMap/RouteMap

[19-Mar-2010 10:51:40] <rmatte> it's obviously not done yet, but sneak peek

[19-Mar-2010 10:51:40] <QbY> actually Interface and Route, I already know

[19-Mar-2010 10:51:56] <rmatte> RouteMap obviously maps routes

[19-Mar-2010 10:52:05] <rmatte> ;)

[19-Mar-2010 10:52:16] <QbY> Interface maps map the interface s(which I need)

[19-Mar-2010 10:52:26] <QbY> but I'm trying to see if device is something i need

[19-Mar-2010 10:52:35] <QbY> what's the username/password for this demo

[19-Mar-2010 10:52:49] <rmatte> admin/zenoss

[19-Mar-2010 10:53:08] <rmatte> Device maps the details you see in the Device Information on the Status tab

[19-Mar-2010 10:53:12] <rmatte> you can leave that

[19-Mar-2010 10:53:23] <rmatte> the RouteMap is probably what's taking ages to model

[19-Mar-2010 10:53:25] <QbY> ohyeah--i like this

[19-Mar-2010 10:53:28] <QbY> i like this new interface

[19-Mar-2010 10:53:29] <rmatte> just drop that plugin and you should be fine

[19-Mar-2010 10:53:35] <rmatte> yeh, the new interface is pure sex

[19-Mar-2010 10:54:14] <QbY> one thing i love about this new one, is the ability to create groups--throw a device in a group, and alert only those people in that group

[19-Mar-2010 10:55:52] <QbY> now, if i add something to the template after adding devices--will that automatically apply to existing interfaces?

[19-Mar-2010 10:56:51] <rmatte> you mean add something to the interface template?

[19-Mar-2010 10:56:56] <rmatte> there are different types of templates

[19-Mar-2010 10:57:02] <rmatte> device templates, interface, filesystem

[19-Mar-2010 10:57:07] <rmatte> processes

[19-Mar-2010 10:57:09] <rmatte> etc...

[19-Mar-2010 10:57:29] <QbY> well the "Device" type

[19-Mar-2010 10:57:29] <QbY> o

[19-Mar-2010 10:57:41] <QbY> i'm building a new device type, that is just for my docsis devices

[19-Mar-2010 11:37:05] fOrsberg is now known as forsberg

[19-Mar-2010 11:56:57] <QbY> arg.. i think i'm gonna scream!

[19-Mar-2010 11:57:02] <QbY> i cannot get this thing to ad

[19-Mar-2010 11:57:07] <QbY> walk it out without timing out

[19-Mar-2010 11:58:36] <QbY> WARNING         zen.ZenDiscClient cmts.... timeout...

[19-Mar-2010 11:59:04] <Simon4> how long does a snmpwalk take on the commandline?

[19-Mar-2010 11:59:20] <QbY> 20 minutes

[19-Mar-2010 12:00:41] <Simon4> wow

[19-Mar-2010 12:00:51] <QbY> they have a lot of data

[19-Mar-2010 12:01:14] <QbY> i was only using the InterfaceMap

[19-Mar-2010 12:02:53] <Simon4> I wouldn't have thought it would have timed out if it was still getting data back

[19-Mar-2010 12:03:39] <Simon4> try becoming the zenoss user, and running zenmodeler run -d cmts -v3 - you'll get a bunch of output but should be able to see more what it's doing

[19-Mar-2010 12:03:48] <Simon4> (assuming your device is called cmts)

[19-Mar-2010 12:04:19] <QbY> i'm running tcpdump at the same time, so i see it working

[19-Mar-2010 12:17:17] <QbY> is version 3 faster than 2c?

[19-Mar-2010 12:17:47] <QbY> i'm running a dump and saving it to a text file .. just to time it now

[19-Mar-2010 12:17:49] <QbY> with 2c

[19-Mar-2010 12:18:17] <rmatte> QbY...

[19-Mar-2010 12:18:21] <QbY> yes

[19-Mar-2010 12:18:35] <rmatte> change zCollectorClientTimeout to 5000 for those devices

[19-Mar-2010 12:18:37] <rmatte> then model

[19-Mar-2010 12:18:45] <rmatte> it'll take the full 20 mins but won't time out

[19-Mar-2010 12:18:55] <QbY> ok

[19-Mar-2010 12:19:12] <Simon4> heh, I was just asking colleague if that was the one we tweaked

[19-Mar-2010 12:19:44] <rmatte> we monitor stuff in europe that is realllllly slow to respond, you can literally read the snmpwalk line by line as it comes in

[19-Mar-2010 12:19:53] <rmatte> so I tweaked that and I was able to model them

[19-Mar-2010 12:20:01] <Simon4> yeah, we have it at 1800 for some kit

[19-Mar-2010 12:20:25] <rmatte> I just used 5000 for the heck of it, it can probably be set lower, but whatever

[19-Mar-2010 12:20:34] <QbY> my routers load fast--but in an snmpwalk--you'll find when hitting a cmts there are so many entires for each cable modem connected

[19-Mar-2010 12:20:47] <QbY> i'm timing it.. we are at 14 minutes now..

[19-Mar-2010 12:20:59] <rmatte> I love how I ask for a medium sized combo at Wendy's and I still get a drink the size of Texas

[19-Mar-2010 12:21:25] <QbY> hehehheheheehee

[19-Mar-2010 12:24:49] * Simon4 makes his first foray into ZenPack land

[19-Mar-2010 12:25:07] <QbY> Simon4: I started making one last nigt

[19-Mar-2010 12:25:09] <Simon4> the HP Bladechassis zenpack is very ... basic, and could do with some more detail about each blade me thinks

[19-Mar-2010 12:50:08] <QbY> 45 minutes and still walking

[19-Mar-2010 12:50:11] <QbY> geez

[19-Mar-2010 12:57:45] <QbY> i've got to find a better way to add equipment--if i have to wait 45+ minutes for each cmts i add, i won't be able to start monitoring for a freaking week

[19-Mar-2010 12:58:25] <rmatte> well, is that actually how long it takes to snmpwalk one?

[19-Mar-2010 12:58:40] <rmatte> I think Zenoss does a full walk even if you disable whatever plugins

[19-Mar-2010 12:58:49] <rmatte> so you sort of don't have a choice

[19-Mar-2010 12:58:54] <QbY> this is at 52 minutes and still walking

[19-Mar-2010 12:59:01] <Simon4> QbY: are there other ways to get the information?

[19-Mar-2010 12:59:11] <QbY> its a cisco router

[19-Mar-2010 12:59:25] <Simon4> you could possibly write a custom modeller plugin to fetch just hte bits you want and avoid the snmpwalk, not sure

[19-Mar-2010 12:59:44] <Simon4> so you enumerate the interfaces another way

[19-Mar-2010 12:59:46] <rmatte> Simon4: it sounds simpler when you say it like that, but modeler plugins are not simple to write :P

[19-Mar-2010 12:59:55] <Simon4> rmatte: shush, I was doing so well!

[19-Mar-2010 13:00:03] * Simon4 wasn't going to offer to write it ;)

[19-Mar-2010 13:00:07] <rmatte> lol

[19-Mar-2010 13:00:18] <QbY> i dunno what to do, other than just add each in manually

[19-Mar-2010 13:00:22] <venturaville> what are you modelling?

[19-Mar-2010 13:00:23] <QbY> but there's 45+ interfaces each

[19-Mar-2010 13:00:29] <rmatte> that's like saying "Oh, you're getting stuck in traffic on the way to work?  All you have to do is take a rocket ship!"

[19-Mar-2010 13:00:42] <Simon4> actually, as it happens, I have modeller code open on my terminal, so I'll refine my suggestion in an hour or two

[19-Mar-2010 13:00:43] <rmatte> :P

[19-Mar-2010 13:00:48] <Simon4> since I'm about to do basically that

[19-Mar-2010 13:00:51] <rmatte> hehe

[19-Mar-2010 13:00:56] <QbY> venturaville: a CMTS.  A CMTS is essntially a Cisco 7204VXR..  But has cable interfaces--each one has thousands of modems hanging off of it..

[19-Mar-2010 13:01:04] <venturaville> ouch

[19-Mar-2010 13:01:12] <QbY> so when it walks, it has a LOT of data

[19-Mar-2010 13:01:23] <rmatte> How many CMTS do you have?

[19-Mar-2010 13:01:39] <QbY> oh, about 50

[19-Mar-2010 13:01:53] <rmatte> well, you might just have to suck it up and continue the way you are now

[19-Mar-2010 13:01:59] <rmatte> until you get them all added

[19-Mar-2010 13:02:06] <Simon4> hmm

[19-Mar-2010 13:02:14] <Simon4> there's a nightly remodel to think of too

[19-Mar-2010 13:02:20] <Simon4> depending on how long it ends up taking

[19-Mar-2010 13:02:21] <QbY> geez

[19-Mar-2010 13:02:24] <rmatte> I disable zenmodel

[19-Mar-2010 13:02:32] <Simon4> yeah, was goig to say you can kill that

[19-Mar-2010 13:02:42] <Simon4> I leave it on for the configuration management aspect

[19-Mar-2010 13:02:46] <QbY> jffnms used to let me add them in about 1 minute

[19-Mar-2010 13:02:48] <rmatte> QbY: blogs/zenossblog/2009/01/27/tip-of-the-month-running-only-some-daemons/

[19-Mar-2010 13:02:51] <Simon4> but it can be a pita

[19-Mar-2010 13:02:58] <rmatte> you'll want to do that, and add everything except for zenmodel to the list

[19-Mar-2010 13:03:07] <rmatte> that way you won't have automated remodels of your devices

[19-Mar-2010 13:03:08] <QbY> i can't tell it in the snmpwalk command to just walk for interfaces

[19-Mar-2010 13:03:43] <rmatte> Zenoss isn't actually using snmpwalk, if that's what you mean

[19-Mar-2010 13:04:03] <rmatte> it used twisted's snmp implementation

[19-Mar-2010 13:04:05] <rmatte> uses*

[19-Mar-2010 13:04:12] <QbY> i really want to use this software--based on what i've seen to date.. but this going to be too painful

[19-Mar-2010 13:04:39] <rmatte> QbY: you'll probably run in to the same issue no matter what software you use, honestly

[19-Mar-2010 13:04:53] <QbY> true

[19-Mar-2010 13:05:10] <rmatte> so you tried removing collector plugins and that didn't work any faster?

[19-Mar-2010 13:05:11] <QbY> we used to use jffnms.. and when i'd add a cmts it would hang for about a minute, but that was it

[19-Mar-2010 13:05:24] <QbY> i removed everyting except interfacemap

[19-Mar-2010 13:05:46] <rmatte> I think this is how Zenoss does it...

[19-Mar-2010 13:06:01] <rmatte> it performs a full walk, then stores that in memory, then the collector plugins are simply used to parse the data

[19-Mar-2010 13:06:10] <rmatte> it's kind of a crappy way of doing it, but that's the deal

[19-Mar-2010 13:06:16] <QbY> hrmm

[19-Mar-2010 13:06:27] <QbY> i wonder if i could make snmp view for just interfac ecs

[19-Mar-2010 13:08:13] <QbY> let me create a IF-Mib view only

[19-Mar-2010 13:08:31] <mrayzenoss> QbY: do you have zRouteMapCollectOnlyLocal set to true?

[19-Mar-2010 13:08:45] <QbY> if that is the default yes

[19-Mar-2010 13:08:51] <QbY> i haven't messed wih that

[19-Mar-2010 13:09:01] <QbY> let me create a speical view for just the IF-Mib

[19-Mar-2010 13:09:02] <mrayzenoss> not sure what the default is, you may be collecting too much info on the routing tables

[19-Mar-2010 13:09:54] <rmatte> mrayzenoss: that setting does absolutely nothing for modelling time in my experience

[19-Mar-2010 13:10:00] <rmatte> it just means that only local routes are shown

[19-Mar-2010 13:10:07] <rmatte> but it still walks the entire route table

[19-Mar-2010 13:10:26] <QbY> i can set the cmts to only respond to certain mibs

[19-Mar-2010 13:10:33] <QbY> so a walk will only return certain mibs

[19-Mar-2010 13:10:39] <rmatte> I'm pretty sure zenoss collects absolutely everything, stores it in memory, then parses it with the collector plugins

[19-Mar-2010 13:10:49] <rmatte> so there's no way to only have it collect partial info when modelling

[19-Mar-2010 13:10:55] <rmatte> in my own personal experience anyways

[19-Mar-2010 13:11:04] <rmatte> and if that is the case, it's a stupid way of modelling devices

[19-Mar-2010 13:11:45] <rmatte> QbY: well then get rid of the Mibs that have to do with routes

[19-Mar-2010 13:11:53] <QbY> its not the routes though

[19-Mar-2010 13:12:06] <QbY> if its walking the entire thing, that's a LOT of data

[19-Mar-2010 13:12:08] <Simon4> it's just the sheer number of interfaces, it sounds like

[19-Mar-2010 13:12:08] <QbY> a ton

[19-Mar-2010 13:12:16] <QbY> there are 45 actual interfaces

[19-Mar-2010 13:12:27] <QbY> but for each modem there is at least 10 different values

[19-Mar-2010 13:12:31] <QbY> so 2,000x10

[19-Mar-2010 13:12:43] <QbY> counters, flap data, etc

[19-Mar-2010 13:12:57] <rmatte> I see

[19-Mar-2010 13:13:06] <rmatte> you're wanting to monitor all of that?

[19-Mar-2010 13:14:14] <QbY> no,

[19-Mar-2010 13:14:24] <QbY> just interface statuses

[19-Mar-2010 13:15:12] <mrayzenoss> I was looking around through the email archives, found the recommendation that you remove the zenoss.snmp.RouteMap collector plugin

[19-Mar-2010 13:15:48] <QbY> mrayzenoss: it's been removed.. the only thing i'm looking at is the interfacemap

[19-Mar-2010 13:15:49] <mrayzenoss> looking for a good link explaining why

[19-Mar-2010 13:18:14] <rmatte> QbY: perhaps all those modems are seen as "interfaces"?

[19-Mar-2010 13:19:25] <QbY> i've never seen them in any of our other nms'es as interfaces..  but when you do a walk, they are definitely there

[19-Mar-2010 13:19:37] <QbY> i'm going to modify the config on the cmts to only report back the interfaces mib

[19-Mar-2010 13:20:18] <QbY> and ignore the DOCSIS--IF-MIB

[19-Mar-2010 13:35:03] * rmatte scratches his head

[19-Mar-2010 13:50:58] <Simon4> mrayzenoss: random dev question that you might know the answer to - if I want to check out http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.HPBladeChassis from svn, what's the location?

[19-Mar-2010 13:51:32] <mrayzenoss> Simon4: docs/DOC-2371

[19-Mar-2010 13:51:49] <mrayzenoss> more tips at community/developers/zenpack_development

[19-Mar-2010 13:51:54] <Simon4> ah, legend, cheers

[19-Mar-2010 14:17:33] <Simon4> schweet, got zenpack checked out, and building

[19-Mar-2010 14:17:36] <Simon4> time to break it :D

[19-Mar-2010 14:18:02] * Simon4 fetches some more wine

[19-Mar-2010 14:18:38] <venturaville> Simon4: feel free to add some more stuff to it :-)

[19-Mar-2010 14:19:02] <venturaville> I've got on my list to pickle and cache the SOAP XML data so it doesn't kill the chassis

[19-Mar-2010 14:19:22] <Simon4> venturaville: cheers :) it's oging to be a learning experience, but I want to be able to see cpu/memory in each blade + ILO ip and NIC mac addresses

[19-Mar-2010 14:19:34] <Simon4> since that gives us great configuration management/audit/data for cobbler etc

[19-Mar-2010 14:20:00] <Simon4> can get the info fine, just need to find a tidy way to stick it into zenoss, so your zenpack became the target :)

[19-Mar-2010 14:20:09] <Simon4> since it does 99% of what I want brilliantly

[19-Mar-2010 14:20:16] <venturaville> be my guest

[19-Mar-2010 14:47:15] <QbY> gents

[19-Mar-2010 14:48:03] <QbY> for any of you who encounter someone with my same issue on a cisco router..  have them investigate... snmp-server view zenoss interfaces included

[19-Mar-2010 14:48:03] <QbY> snmp-server view zenoss ipAddrEntry included

[19-Mar-2010 14:48:12] <QbY> that way they don't have to walk it all to make them happy

[19-Mar-2010 14:50:27] <Simon4> QbY: brilliant

[19-Mar-2010 14:50:40] <Simon4> that speeds it up and sorts the issue?

[19-Mar-2010 14:52:19] * Simon4 attempts to install his slightly modified zenpack

[19-Mar-2010 14:55:40] <QbY> Simon4: ..  adds it in 10 seocnds..

[19-Mar-2010 14:55:53] <Simon4> excellent!

[19-Mar-2010 14:57:31] <QbY> my last question--i'm adding custom data points..  i know how to do it..

[19-Mar-2010 14:58:02] <QbY> i want to create a threshold.  based on a mathemetical check of two values--where can i find tutorial on it..

[19-Mar-2010 14:58:31] <Simon4> I haven't managed yet to make a threshold based on two separate datapoints

[19-Mar-2010 14:59:01] <QbY> what i want to do is: value 1 / value 2.. if that is < 70 ..  i need it to throw an alarm

[19-Mar-2010 14:59:14] <Simon4> mainly I think because thresholds are evaluated at collect-time, so there isn't visibility of the other datapoints

[19-Mar-2010 14:59:26] * Simon4 would love someone to stop him right now and tell him that he's wrong

[19-Mar-2010 14:59:43] <QbY> oooo

[19-Mar-2010 14:59:52] <venturaville> CDEF:ruse=ruse-raw CDEF:ravgwait=ruse-raw,rio-raw,0.00000000001,+,/

[19-Mar-2010 14:59:57] <venturaville> anyone know why that would break?

[19-Mar-2010 15:00:47] <Simon4> it's breaking my brain looking at it, but that might be the wine :)

[19-Mar-2010 15:00:48] <venturaville> ruse-raw and rio-raw are defined

[19-Mar-2010 15:01:01] <venturaville> CDEF:ravgwait=ruse-raw,rio-raw,/

[19-Mar-2010 15:01:07] <venturaville> simpler version of it....also break

[19-Mar-2010 15:01:08] <venturaville> breaks

[19-Mar-2010 15:01:30] <venturaville> I have 2 data points ...both hidden

[19-Mar-2010 15:01:34] <Simon4> it looks okay

[19-Mar-2010 15:01:38] <venturaville> and then a custom datapoints CDEF

[19-Mar-2010 15:01:48] <venturaville> I can graph the 2 points individually in another graph

[19-Mar-2010 15:01:58] <venturaville> but when I actually try to calculate something it breaks

[19-Mar-2010 15:02:10] <Simon4> this is in multi-graph report? or?

[19-Mar-2010 15:02:30] <venturaville> straight up template

[19-Mar-2010 15:02:37] <venturaville> hrm ... even putting a 1 in there breaks it

[19-Mar-2010 15:02:43] <venturaville> CDEF:ravgwait=1

[19-Mar-2010 15:02:43] <Simon4> k, so no multiple host sillyness

[19-Mar-2010 15:05:05] * Simon4 ssh tunnels his way into making his dev vm look like a blade chassis

[19-Mar-2010 15:05:33] <venturaville> ah.. I'm an idiot apparently

[19-Mar-2010 15:05:44] <venturaville> can't generate a graph with CDEF data points... have to add an AREA

[19-Mar-2010 15:05:51] <venturaville> did I mention how much I dislike RRD

[19-Mar-2010 15:05:54] <Simon4> heh

[19-Mar-2010 15:06:05] <Simon4> RRD is a bunch of pain at times

[19-Mar-2010 15:15:30] <rmatte> unhunh, that was fun

[19-Mar-2010 15:16:46] <Simon4> heh

[19-Mar-2010 15:16:54] <Simon4> wild irc rollercoaster fun

[19-Mar-2010 15:21:44] * Simon4 goes hunting for ssh modeler examples

[19-Mar-2010 15:22:09] <mrayzenoss> Simon4: the LinuxMonitor and OpenSolaris are both good

[19-Mar-2010 15:22:19] <Simon4> mrayzenoss: cheers

[19-Mar-2010 15:22:28] <mrayzenoss> the LinuxMonitor has lots of extensions off of it

[19-Mar-2010 15:22:42] <mrayzenoss> Ubuntu, Fedora, Mandriva, Suse

[19-Mar-2010 15:22:45] <Simon4> can get bunch of info about blades via ssh to the OA on a chassis, so using the same techniques would be smart etc

[19-Mar-2010 15:23:09] <Simon4> and if I can use the same relevant zProperties it makes life really easy

[19-Mar-2010 15:24:46] <Simon4> w00t

[19-Mar-2010 15:24:53] <Simon4> my skin and db changes worked :D

[19-Mar-2010 15:25:01] <Simon4> now to actually populate them (the hard bit lol)

[19-Mar-2010 15:28:30] <mrayzenoss> I think the VMware ESX Filesystems ZenPack models via SSH and monitors via SNMP

[19-Mar-2010 15:28:39] <Simon4> *nod*

[19-Mar-2010 15:28:45] <Simon4> sounds like what this could do

[19-Mar-2010 15:29:10] <Simon4> mainly my changes are modeler related, since it's all about configuration and blade/chassis utilisation management

[19-Mar-2010 15:29:22] <Simon4> so our infrastructure team can bin their horrible spreadsheet

[19-Mar-2010 15:29:28] <Simon4> (which is _always_ out of date)

[19-Mar-2010 15:43:54] <rmatte> ohhhh my gawwwwd I am going to throw something, I finally get libvirt compiled, installed, get the libvirt python module installed, but now it's complaining about a libvirtmod modules

[19-Mar-2010 15:43:57] <rmatte> module*

[19-Mar-2010 15:45:30] <Simon4> gaaah

[19-Mar-2010 15:45:44] <Simon4> I hate those kind of yak shaves

[19-Mar-2010 15:47:24] <Simon4> hmm, the wine is working, I'm understanding this code

[19-Mar-2010 15:51:42] <rmatte> lol

[19-Mar-2010 15:55:30] <Simon4> oh man... this actually is easy

[19-Mar-2010 15:59:47] <Simon4> class VmwareEsxDf(CommandPlugin):

[19-Mar-2010 15:59:47] <Simon4>     """

[19-Mar-2010 15:59:47] <Simon4>     Run vdf to model filesystem information. Should work on all ESX servers.

[19-Mar-2010 15:59:48] <Simon4>     """

[19-Mar-2010 15:59:48] <Simon4>     maptype = "FilesystemMap"

[19-Mar-2010 15:59:48] <Simon4>     command = '/usr/sbin/vdf'

[19-Mar-2010 16:00:00] <Simon4> so you just ask zenoss to ssh in, and run the command for you, and then you can play with the output

[19-Mar-2010 16:00:07] <Simon4> love it.

[19-Mar-2010 16:01:42] <mrayzenoss> Simon4: did you see this? docs/DOC-2381

[19-Mar-2010 16:02:04] <Simon4> mrayzenoss: I have now :)

[19-Mar-2010 16:02:39] <Simon4> but yeah, was reading the esx one, and it's very clear and easy to follow

[19-Mar-2010 16:03:02] <Simon4> far easier than I expected it to be

[19-Mar-2010 16:03:16] * Simon4 has the parser code for the output already written from another job

[19-Mar-2010 16:10:31] <rmatte> omg, I think I might have finally figured out this libvirt thing

[19-Mar-2010 16:10:35] * rmatte crosses his fingers

[19-Mar-2010 16:10:54] <rmatte> wow: http://thereifixedit.files.wordpress.com/2010/03/129121045718603313.jpg

[19-Mar-2010 16:13:53] <Simon4> hehe

[19-Mar-2010 16:14:32] <Simon4> in a similar vein

[19-Mar-2010 16:14:33] <Simon4> http://twitpic.com/19iz35

[19-Mar-2010 16:14:40] <Simon4> was what I arrived at the DC to today

[19-Mar-2010 16:15:06] <rmatte> lol

[19-Mar-2010 16:15:52] * Simon4 attempts his "custom" modeler plugin

[19-Mar-2010 16:15:59] <Simon4> which does not much more than debug log results

[19-Mar-2010 16:16:56] <Simon4> however if it successfully ssh's in etc, I'm home and hosed

[19-Mar-2010 16:21:28] <Simon4> Using SSH collection method for device localhost.localdomain

[19-Mar-2010 16:21:31] <Simon4> a good start!

[19-Mar-2010 16:25:08] <Simon4> :D :D :D :D :D :D :D :D

[19-Mar-2010 16:25:10] <theacolyte> I knew I was missing a channel on my list

[19-Mar-2010 16:25:21] <theacolyte> I finally upgraded my irssi box to Debian 5

[19-Mar-2010 16:25:22] <Simon4> ssh modeling win, four lines of code, and works a charm

[19-Mar-2010 16:25:31] <theacolyte> Debian 4 was EOL'd last month... :'(

[19-Mar-2010 16:25:44] <Simon4> rmatte: I might have to disagree with you on the whole "custom modelers are hard" thing

[19-Mar-2010 16:26:08] <Simon4> (and I'll def blog about my experiences/how I did it)

[19-Mar-2010 16:26:21] <Simon4> theacolyte: debian 4 = pre-lenny, yeah?

[19-Mar-2010 16:27:00] <theacolyte> Yeah

[19-Mar-2010 16:27:04] <theacolyte> 5 = lenny

[19-Mar-2010 16:27:12] <Simon4> cool

[19-Mar-2010 16:27:19] <Simon4> I have a lenny box kicking about somewhere

[19-Mar-2010 16:27:34] <Simon4> which I pay little attention to :O

[19-Mar-2010 16:30:23] <rmatte> great, so I spend half a day getting these libvirt libraries installed, and they don't seem to work

[19-Mar-2010 16:30:36] <rmatte> guess I'll wait until next time I see venturaville

[19-Mar-2010 16:32:23] <rmatte> Simon4: it's easy if you're just modifying an existing modeller

[19-Mar-2010 16:32:31] <rmatte> Simon4: for the modeler I want to write it's extremely complex

[19-Mar-2010 16:32:38] <Simon4> rmatte: for sure

[19-Mar-2010 16:32:49] <Simon4> but it seems that there are enough examples out there to do most things

[19-Mar-2010 16:33:22] <Simon4> I took the ESX ssh bit, learned about doing ssh modeling, then found the om points for the blade chassis bit, and now have most of a ssh blade chassis modeller

[19-Mar-2010 16:33:46] <rmatte> cool

[19-Mar-2010 16:33:48] <Simon4> if you wanted to use something other than ssh or snmp, yeah, I can see it being more effort

[19-Mar-2010 16:34:06] <rmatte> well, I want to use snmp, but it's the way I want to use snmp that makes it difficult

[19-Mar-2010 16:34:42] <Simon4> we have an inhouse modeler written by our "actual" engineering guys, which uses snmp in a totally non-standard way

[19-Mar-2010 16:34:56] * Simon4 is an admin/architect who dabbles in code occasionally

[19-Mar-2010 16:35:16] <Simon4> the devs at work just ran their own snmpwalk I believe

[19-Mar-2010 16:35:18] <rmatte> I want to write a modeler plugin for my windows snmp performance which collects CPU, Memory, and Paging file stats from windows without snmp-informant, but to do that it has to dynamically hunt down OIDs which are in different positions from server to server

[19-Mar-2010 16:35:29] <Simon4> which is a horrible hack, but works

[19-Mar-2010 16:35:33] <rmatte> It's not like I can just walk a tree and grab every value

[19-Mar-2010 16:35:42] <rmatte> ah

[19-Mar-2010 16:35:58] <rmatte> yeh, I'm an admin/developer

[19-Mar-2010 16:36:05] <rmatte> more admin than developer

[19-Mar-2010 16:36:09] <Simon4> it's doable in the plugin though if you don't mind the dirty, just exec snmpwalk and process the output :)_

[19-Mar-2010 16:36:11] <Simon4> yeah, similar

[19-Mar-2010 16:36:20] <rmatte> The most complex thing I've written for zenoss is a custom daemon

[19-Mar-2010 16:36:25] <Simon4> only dev I do is to make life easier :)

[19-Mar-2010 16:36:31] <rmatte> hehe

[19-Mar-2010 16:38:33] <theacolyte> Screw SNMP on windows

[19-Mar-2010 16:38:36] <theacolyte> WMI is the wya to do it

[19-Mar-2010 16:38:49] * Simon4 more than likely agrees

[19-Mar-2010 16:38:59] <Simon4> thankfully there are no windows boxes in my vicinity

[19-Mar-2010 16:39:23] <Simon4> but using native data collection method on any platform = win

[19-Mar-2010 16:39:48] * Simon4 tries to work out just how much effort zenoss is doing for him

[19-Mar-2010 16:40:37] <Simon4> hur

[19-Mar-2010 16:40:53] <Simon4> multiple lines turned into a single string, already strip()'ed

[19-Mar-2010 16:40:58] <Simon4> THANKS

[19-Mar-2010 16:41:01] <theacolyte> The only real way you'll ever get anything worthwhile out of SNMP on windows is with something that integrats with it, be it informant, or ideally HP agents like I do it

[19-Mar-2010 16:41:12] <theacolyte> But as rmatte can attest, I'm big on WMI

[19-Mar-2010 16:41:33] <Simon4> I'm just big on whatever doesn't need additional software, especially on windows

[19-Mar-2010 16:41:43] <Simon4> since it's an additional layer of "support"

[19-Mar-2010 16:41:48] <theacolyte> same

[19-Mar-2010 16:41:55] <theacolyte> having to install any sort of agent makes me unahppy

[19-Mar-2010 16:44:14] * Simon4 tries to work out how to delimit this "results" thing

[19-Mar-2010 16:44:26] <Simon4> since zenoss was so nice to remove \n, and all other things

[19-Mar-2010 16:45:30] <Simon4> I guess I could string.startswith() or whatnot, then truncate the string after finding it

[19-Mar-2010 16:47:42] <Simon4> might be the only way

[19-Mar-2010 16:48:06] <forsberg> god its annoying i cant add more devices with same ip :)

[19-Mar-2010 16:51:45] <Simon4> forsberg: give me 10 mins, lemme blog how we did it (the 10 mins = me removing company specific stuff :) )

[19-Mar-2010 16:52:11] <forsberg> :)

[19-Mar-2010 16:55:50] <rmatte> eugh, so close to getting this working

[19-Mar-2010 17:00:17] <rmatte> Need to figure out how to fix this now...

[19-Mar-2010 17:00:18] <rmatte> libvirt.libvirtError: internal error Expecting VMX entry 'virtualHW.version' to be 7 for VI API version 4.0 but found 4

[19-Mar-2010 17:03:36] <rmatte> going to try a newer version of libvirt, apparently that problem is fixed in 0.7.7

[19-Mar-2010 17:04:21] <mrayzenoss> it's like I always say… with Open Source you get what you pay for :p

[19-Mar-2010 17:04:26] <rmatte> yup

[19-Mar-2010 17:04:34] <rmatte> too true

[19-Mar-2010 17:04:51] <rmatte> I'm documenting the hell out of the process I followed once I'm done with this and throwing it up on the community site

[19-Mar-2010 17:04:57] * Simon4 utterly agrees :)

[19-Mar-2010 17:04:57] <rmatte> hopefully it'll save other people the same headache

[19-Mar-2010 17:05:04] <mrayzenoss> of course, when libvirt gets closer and closer to 1.0 it's going to rock so much

[19-Mar-2010 17:05:10] <rmatte> yup

[19-Mar-2010 17:05:20] <rmatte> it definitely sounds nice

[19-Mar-2010 17:05:27] <rmatte> can't wait until I can experience it for myself :P

[19-Mar-2010 17:05:41] * rmatte taps his foot while it compiles

[19-Mar-2010 17:06:03] <rmatte> if this doesn't fix it I'm going to go smash my head through a wall or something :|

[19-Mar-2010 17:06:21] <rmatte> going on 5 hours now, with a fried/restored kernel

[19-Mar-2010 17:06:42] <rmatte> at least the day flew by

[19-Mar-2010 17:09:51] <rmatte> great, now it's "Missing essential config entry 'ide0:0.fileName'"

[19-Mar-2010 17:10:21] <Simon4> rmatte: you can at least be happy that for me it's 10pm Friday

[19-Mar-2010 17:10:28] <Simon4> and I'm still doing zenoss + wine

[19-Mar-2010 17:10:44] <rmatte> lol

[19-Mar-2010 17:10:53] <rmatte> why wine?

[19-Mar-2010 17:11:01] <Simon4> I like it? :)

[19-Mar-2010 17:11:12] <rmatte> oh, I thought you meant wine as in wine is not an emulator

[19-Mar-2010 17:11:15] <rmatte> :P

[19-Mar-2010 17:11:16] <Simon4> ahahaa

[19-Mar-2010 17:11:16] <Simon4> no

[19-Mar-2010 17:11:19] <Simon4> drink :)

[19-Mar-2010 17:11:23] <rmatte> ah, gotcha

[19-Mar-2010 17:11:27] <rmatte> I wish I had a beer right now

[19-Mar-2010 17:11:38] <rmatte> I intend to go home and pound a few back after this whole libvirt thing

[19-Mar-2010 17:12:16] <Simon4> :)

[19-Mar-2010 17:12:17] <Simon4> do it

[19-Mar-2010 17:13:40] <forsberg> haha :)

[19-Mar-2010 17:16:12] <rmatte> this is the first ZenPack install which has driven me to drink

[19-Mar-2010 17:16:13] <rmatte> lol

[19-Mar-2010 17:16:42] <mrayzenoss> Hmm.. out of the 140 or so I've processed, I can say lots of them make me drink

[19-Mar-2010 17:17:05] <rmatte> hehe

[19-Mar-2010 17:17:21] <forsberg> :)

[19-Mar-2010 17:17:23] <rmatte> well, I don't generally use the crappy ones

[19-Mar-2010 17:17:23] <mrayzenoss> new NetBotz and updated NetBotz on the way

[19-Mar-2010 17:18:13] <rmatte> the VMWareESXMonitor pack works fine, besides the fact that VMWare don't really support SNMP much anymore, the most I could do was get the guests tab to populate

[19-Mar-2010 17:18:19] <rmatte> other than that, no performance data for any VMs

[19-Mar-2010 17:18:40] <mrayzenoss> hmm… I know I've seen that working

[19-Mar-2010 17:18:53] <rmatte> It works for VMWare 3.5

[19-Mar-2010 17:18:54] <mrayzenoss> but again, SNMP sucks on VMware

[19-Mar-2010 17:18:56] <rmatte> not for 4.0

[19-Mar-2010 17:19:12] <rmatte> the SNMP on 4.0 is garbage compared to 3.5

[19-Mar-2010 17:22:36] <rmatte> ah, it looks like it's failing because of one of the config files on the actual vmware server

[19-Mar-2010 17:22:37] <rmatte> figured

[19-Mar-2010 17:22:40] <rmatte> figures*

[19-Mar-2010 17:23:13] <rmatte> time to try to hunt down which file

[19-Mar-2010 17:32:10] * Simon4 has to say that any day working with Zenoss drives him to drink

[19-Mar-2010 17:32:17] <Simon4> but I'm not sure it's Zenoss :)

[19-Mar-2010 17:32:51] <Simon4> randomly dying zenperfsnmp for three hours grouped with a mysql heartbeat table performance problem however

[19-Mar-2010 17:32:56] <Simon4> ARRRRGH DRINK

[19-Mar-2010 17:33:14] <Simon4> this is why you have major graphs on your 24/7 team's dashboard

[19-Mar-2010 17:33:19] <Simon4> "hey, the graphs stopped"

[19-Mar-2010 17:33:59] <forsberg> :/

[19-Mar-2010 17:34:37] <Simon4> ohshit

[19-Mar-2010 17:34:41] <Simon4> forsberg: sorry

[19-Mar-2010 17:34:48] <Simon4> 10 min turned into a while

[19-Mar-2010 17:35:16] <Simon4> 2 secs

[19-Mar-2010 17:36:11] <mrayzenoss> FYI, updated NetBotz ZenPack: docs/DOC-3479

[19-Mar-2010 17:36:31] <mrayzenoss> and new NetBotz RackBotz/WallBotz ZenPack: docs/DOC-5798

[19-Mar-2010 17:36:41] <mrayzenoss> lotz of botz

[19-Mar-2010 17:37:22] <forsberg> :P

[19-Mar-2010 17:37:50] <rmatte> my god, we have like 100 vmx files that I have to go through

[19-Mar-2010 17:39:05] <rmatte> why can't it just ignore vmx errors and just move on

[19-Mar-2010 17:39:08] <rmatte> garrrr

[19-Mar-2010 17:41:39] <Simon4> forsberg: http://nettikconsulting.wordpress.com/2010/03/19/zenoss-more-than-one-device-with-a-single-ip/

[19-Mar-2010 17:41:53] <Simon4> my first blog post LOL!!!1111etc

[19-Mar-2010 17:42:21] <mrayzenoss> delicious'ed for inclusion in the link blog

[19-Mar-2010 17:42:50] <Simon4> heh, cheers :)

[19-Mar-2010 17:43:59] <mrayzenoss> with 2.5 you can have multiple devices with the same name

[19-Mar-2010 17:44:27] <Simon4> for clusters?

[19-Mar-2010 17:44:30] <mrayzenoss> there's now a title and id fields

[19-Mar-2010 17:44:38] <mrayzenoss> it was added for VMs

[19-Mar-2010 17:45:15] <mrayzenoss> but I don't think it got much notice

[19-Mar-2010 17:45:37] * Simon4 has a whole bunch of blog posts that will hopeuflly turn up in the next month or so

[19-Mar-2010 17:45:38] <mrayzenoss> instead of calling 'id' off of devices, we now call 'titleorid'

[19-Mar-2010 17:45:53] <Simon4> we'll see, I've never been much of a blogger,but this stuff is fun to share

[19-Mar-2010 17:47:32] <mrayzenoss> Egor sent me this ZenPack on a related note: people/bigegor/blog/2010/02/22/virtualip-zenpack

[19-Mar-2010 17:49:50] <Simon4> yeah, that looked cool also

[19-Mar-2010 17:51:01] <forsberg> thanks!

[19-Mar-2010 17:52:51] * Simon4 has about another urm... 15-20 open posts that should be finished sometime

[19-Mar-2010 17:52:57] <Simon4> all zendmd related

[19-Mar-2010 17:53:09] <Simon4> but want to do one about tonight's zenpack stuff too

[19-Mar-2010 17:53:20] <Simon4> and some architecture stuff too as we get our new kit in

[19-Mar-2010 17:53:36] <Simon4> (very muhc a work in progress)

[19-Mar-2010 17:53:41] <mrayzenoss> sounds good

[19-Mar-2010 17:53:46] <mrayzenoss> time for me to head home

[19-Mar-2010 17:53:59] <Simon4> seu - cheers again

[19-Mar-2010 17:54:10] * Simon4 moves onto a Mt Difficulty Pinot Noir

[19-Mar-2010 17:54:16] <Simon4> great Kiwi wine

[19-Mar-2010 20:08:07] forsberg is now known as fOrsberg

[20-Mar-2010 00:00:30] [disconnected at Sat Mar 20 00:00:30 2010]

[20-Mar-2010 00:00:30] [connected at Sat Mar 20 00:00:30 2010]

[20-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[20-Mar-2010 02:56:49] fOrsberg is now known as forsberg

[20-Mar-2010 06:28:21] rhettardo_ is now known as rhettardo

[20-Mar-2010 07:20:21] forsberg is now known as fOrsberg

[20-Mar-2010 08:08:48] <Simon4> awesome, custom modeller is working just fine, now to work out how to hand the data back to zenmodeler

[20-Mar-2010 08:12:44] <rmatte> cool

[20-Mar-2010 08:12:59] <Simon4> http://pastie.org/878432 :)

[20-Mar-2010 08:29:04] <Simon4> woot

[20-Mar-2010 08:29:10] <Simon4> and excusing the non-initalised values

[20-Mar-2010 08:29:11] <Simon4> http://lol.nzdance.net/bladeservers1.png

[20-Mar-2010 08:35:15] <sergeymasushko> is there a way to suppress events on the first device in there is an event on the other device?

[20-Mar-2010 08:37:06] <Simon4> yeah, event transforms will let you do that

[20-Mar-2010 08:37:15] <Simon4> I saw a good example of it somewhere, give me a sec

[20-Mar-2010 08:37:30] <sergeymasushko> Simon4: thanks...

[20-Mar-2010 08:38:19] <Apachez> -1 cpus

[20-Mar-2010 08:38:27] <Apachez> quantum cpus? ;)

[20-Mar-2010 08:38:34] <Simon4> Apachez: we have some awesome kit ;)

[20-Mar-2010 08:39:27] * Simon4 is putting the rest of the parser code in now, just happy that I'm creating the objects/etc all correctly

[20-Mar-2010 08:42:11] <sergeymasushko> I know that event class transforms is able to suppress events for the class by the ping status command getPingStatus() ... I'm interesting in the same behavior for events... e.g. if command check failed for the radius server I would not like to receive alarms for another device...

[20-Mar-2010 08:44:39] <Simon4> hmm

[20-Mar-2010 08:44:44] <Simon4> maybe that's what I read, vs other status

[20-Mar-2010 08:46:45] <sergeymasushko> Simon4: is it that what you seen? docs/DOC-3215

[20-Mar-2010 08:46:58] <Simon4> yeah, that's the one

[20-Mar-2010 08:47:04] <Simon4> and yeah, pingStatus :/

[20-Mar-2010 12:27:38] <QbY> anyone know of a good place to download mibs?

[20-Mar-2010 13:50:19] <QbY> I am creating a performance template for an interface that i am querying several oids ..  I have two data sources: TotalModems (Gauge) and Active Modems (Gauge).  I need to create a Threshold where if the number of Active Modems is <70% i want to be alarmed....  I cannot find an example of how to do this--anyone?

[20-Mar-2010 13:51:14] <QbY> I need it to calculate ActiveModems / Total Modems -- and if that is <70 I need an alarm.

[20-Mar-2010 15:38:03] <QbY> ARGH!!!!!!!!!!!!!111

[20-Mar-2010 15:54:56] <QbY> getRRDValue(self,'TotalModems_TotalModems',start=None,end=None, function='LAST'') -- Does anyone see a problem with that?

[20-Mar-2010 16:18:11] Egyptian[LapTop1 is now known as Egyptian[Laptop]

[21-Mar-2010 00:00:30] [disconnected at Sun Mar 21 00:00:30 2010]

[21-Mar-2010 00:00:30] [connected at Sun Mar 21 00:00:30 2010]

[21-Mar-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[21-Mar-2010 13:18:45] <Simon4> gah

[21-Mar-2010 13:18:56] <Simon4> Position is an integer, yet I get this on sort? http://lol.nzdance.net/bladeservers4.png

[21-Mar-2010 13:31:52] <QbY> I have Zenoss capturing traps/informs from my gear..  But the Event is listed as: "snmp trap 1.3.6.1.4.1.9.0.1" -- I've loaded the Cisco MIBS..  What else must I do to make it log something human readable?

[21-Mar-2010 13:35:52] <Simon4> QbY: this could be useful http://www.google.co.uk/url?sa=t&source=web&ct=res&cd=1&ved=0CAYQFjAA&url=http%3A%2F%2Fwww.skills-1st.co.uk%2Fpapers%2Fjane%2Fzenoss_event_management_paper.pdf&ei=-mamS7W4DI_QjAfDxaS6BQ&usg=AFQjCNHI1EeSLGWrPJDiVp161662AKJjtQ&sig2=FcpO0QysV2bzq8ec2nFI6w

[21-Mar-2010 13:35:57] <Simon4> gah, sorry about hte link

[21-Mar-2010 13:36:01] <Simon4> pdf is a good read though

[21-Mar-2010 13:36:15] <Simon4> from memory it talks about traps a fair bit

[21-Mar-2010 13:36:42] <QbY> And, I would LOVE an answer to this: message/46925#46925

[21-Mar-2010 13:37:36] <Simon4> as would I :)

[21-Mar-2010 13:38:56] <Simon4> crap, seems I have broken the sorting in my zenpack somewhere, the original version works perfectly :/

[21-Mar-2010 13:44:26] <QbY> i'm going to delete all these mibs and reimport them..  i don't have to attach a mib to a device or anything right?

[21-Mar-2010 13:45:18] <Simon4> you shouldn't have to, no

[21-Mar-2010 13:51:29] <QbY> do these have to be imported by zenmib or will just having them accessible to net-snmp work?

[21-Mar-2010 13:53:19] <Simon4> you need to zenmib em

[21-Mar-2010 14:03:00] <QbY> yeah, once i get these traps human readable, and the answer to my question--i've got a badass nms

[21-Mar-2010 14:03:49] * Simon4 is just sorting the finishing touches to his zenpack mods, and then has a badass cha

[21-Mar-2010 14:03:56] <Simon4> chassis management system, even

[21-Mar-2010 14:04:17] <Simon4> nothing I love more than replacing shitty static documentation with decent dynamic management

[21-Mar-2010 14:04:24] <QbY> although i did notice that even though i have my devices in groups--and put folks into distribution lists for that particular group--that people were getting alarms for crap they weren't supposed to

[21-Mar-2010 14:42:59] <QbY> Yo Simon4 or anyone, are you familiar with loading MIBS?

[21-Mar-2010 14:43:33] <QbY> http://pastebin.com/cRTDYYBM

[21-Mar-2010 14:43:50] <QbY> but in my interface; i'm snowing CISCO-SMI as an installed mib

[21-Mar-2010 14:45:12] <Simon4> QbY: sorry, I've only done it once and haven't spent enough time on it to be helpful

[21-Mar-2010 14:45:24] <Simon4> it's on my list of "things to sort out real soon"

[21-Mar-2010 18:53:07] fOrsberg is now known as forsberg

[21-Mar-2010 18:53:45] forsberg is now known as fOrsberg

[21-Mar-2010 18:54:06] fOrsberg is now known as forsberg

[21-Mar-2010 18:56:09] <QbY> anyone know how to embed the rrd graphs into another web page?  '

[21-Mar-2010 19:17:22] <QbY> i think i need to re-install zen..  poo!  anyone know of a way to do it without losing all my customizations??

[21-Mar-2010 19:22:50] <Simon4> you can zenbackup your configuration, and restore when you reinstall

[21-Mar-2010 19:37:31] <QbY> backup made

[21-Mar-2010 19:40:26] <QbY> where do i download the file from

[21-Mar-2010 19:40:52] <Simon4> backup hangs out in /opt/zenoss/backups

[21-Mar-2010 19:41:18] <QbY> i see..

[21-Mar-2010 19:41:18] * Simon4 double checks that

[21-Mar-2010 19:41:37] <Simon4> INFO:zenbackup:Backup written to /opt/zenoss/backups/zenbackup_20100227.tgz

[21-Mar-2010 19:41:42] <Simon4> yup, it does

[21-Mar-2010 19:42:06] <QbY> now to uninstall, i'd do rpm --remove

[21-Mar-2010 19:42:11] <QbY> will that kill that backup?

[21-Mar-2010 19:42:35] <Simon4> I would suggest moving the backup out of the way before doing that... I don't htink the rpm -e will kill it, but tis good to be safe

[21-Mar-2010 19:42:46] <QbY> i've already made a copy

[21-Mar-2010 19:42:46] <QbY> hehe

[21-Mar-2010 19:43:43] <QbY> i'm wondering though if this reinstall is going to fix the issue..  i keep getting: Unable to parse information from /opt/zenoss/share/mibs/irtf/IRTF-NMRG-SMING-TYPES whenever i try to install a mib

[21-Mar-2010 19:43:48] <QbY> and some just aren't installing

[21-Mar-2010 19:45:19] <Simon4> hmm... I doubt it will help.... really you need to find someone in here with good mib/trap skills to answer the question

[21-Mar-2010 19:47:07] * Simon4 heads off to bed, night

[21-Mar-2010 21:45:03] bzed_ is now known as bzed

[22-Mar-2010 00:00:30] [disconnected at Mon Mar 22 00:00:30 2010]

[22-Mar-2010 00:00:30] [connected at Mon Mar 22 00:00:30 2010]

[22-Mar-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[22-Mar-2010 00:10:58] <swift> hi guys, what's the default file format for alerts in zenoss?

[22-Mar-2010 02:37:10] swift_ is now known as swift

[22-Mar-2010 08:55:30] mplex_ is now known as mplex

[22-Mar-2010 09:25:19] <rmatte> venturaville: you around?

[22-Mar-2010 09:25:34] <venturaville> rmatte: what's up?

[22-Mar-2010 09:25:59] <rmatte> ok, so I've made progress with this libvirt thing, I had to upgrade to libvirt 0.7.7 because of some version 4 instead of 7 error

[22-Mar-2010 09:26:03] <rmatte> now I'm seeing...

[22-Mar-2010 09:26:10] <rmatte> libvirt.libvirtError: internal error Missing essential config entry 'ide0:0.fileName'

[22-Mar-2010 09:26:22] <rmatte> does that mean that one of the vmx files on the ESX shares is messed up?

[22-Mar-2010 09:26:56] <rmatte> ah, I guess it does because I just added one of my servers and it worked fine

[22-Mar-2010 09:27:00] <rmatte> but the other 2 it's not working on

[22-Mar-2010 09:27:09] <rmatte> guess I'll have to figure out what vmx file is messed up on them

[22-Mar-2010 09:30:02] <rmatte> lol

[22-Mar-2010 09:30:23] <rmatte> yeh, now that I actually see this libvirt pack in action, I like it a lot

[22-Mar-2010 09:30:58] <rmatte> just waiting to see some graphs now...

[22-Mar-2010 09:31:25] <rmatte> I don't even see graphs with 'nan' yet

[22-Mar-2010 09:32:38] <venturaville> let me see what all I can get working on mine

[22-Mar-2010 09:32:45] <rmatte> k thanks

[22-Mar-2010 09:35:36] <venturaville> I am getting data for my guests

[22-Mar-2010 09:35:50] <venturaville> I'm not yet certain how to do things for the host yet

[22-Mar-2010 09:36:10] <venturaville> I may have a workaround to get more data for the host itself

[22-Mar-2010 09:36:55] <venturaville> i.e. totalling up all interfaces and disk stats and summarizing them on the host (or something similar)

[22-Mar-2010 09:37:04] <venturaville> mainly only an issue with ESX

[22-Mar-2010 09:37:34] <venturaville> all the others I have a means to get data from the hosts via normal means

[22-Mar-2010 09:39:07] <rmatte> yeh

[22-Mar-2010 09:39:15] <rmatte> ESX can be a bit tricky

[22-Mar-2010 09:39:56] <rmatte> I'm getting no performance data for my guests

[22-Mar-2010 09:40:06] <rmatte> I see them in the list, but no graphs

[22-Mar-2010 09:40:23] <venturaville> hrm

[22-Mar-2010 09:40:40] <rmatte> I'm starting to get relatively discouraged lol, all I want to see is guest performance data, second ZenPack i've tried and still no go ;)

[22-Mar-2010 09:40:53] <rmatte> although yours does seem to show more promise so far

[22-Mar-2010 09:42:16] <venturaville> hrm ..running my check_libvirt script on it is returning 0s

[22-Mar-2010 09:42:22] <venturaville> which version of ESX are you running btw?

[22-Mar-2010 09:42:25] <rmatte> 4.0

[22-Mar-2010 09:42:34] <venturaville> yeah I think they just upgraded this one to 4.0 too

[22-Mar-2010 09:42:37] <venturaville> need to check

[22-Mar-2010 09:42:54] <rmatte> k

[22-Mar-2010 09:43:33] <rmatte> I see no new folder or RRD data for the hosts

[22-Mar-2010 09:43:39] <rmatte> for the guests, rather

[22-Mar-2010 09:43:50] <rmatte> guess I'll try remodelling the device

[22-Mar-2010 09:43:51] <venturaville> nope this one is still 3

[22-Mar-2010 09:43:58] <venturaville> hopefully I didn't break the check script

[22-Mar-2010 09:44:26] <rmatte> where should the RRDs for guest performance be stored?

[22-Mar-2010 09:44:42] <rmatte> $ZENHOME/perf/<device>/guests/ ?

[22-Mar-2010 09:45:05] <rmatte> because I'm not seeing any folders that contain RRDs for the guests

[22-Mar-2010 09:46:51] <venturaville> looks like my copy is broken too

[22-Mar-2010 09:47:07] <venturaville> give me a bit to push out a copy that works for ESX

[22-Mar-2010 09:47:12] <rmatte> thankya sir

[22-Mar-2010 09:48:01] <rmatte> also, not sure if there's some way to work around this error: libvirt.libvirtError: internal error Missing essential config entry 'ide0:0.fileName'

[22-Mar-2010 09:48:02] <rmatte> ?

[22-Mar-2010 09:48:10] <rmatte> seems like something that should just be ignored

[22-Mar-2010 09:48:20] <rmatte> but maybe that's a libvirt thing that's unavoidable

[22-Mar-2010 09:55:01] <venturaville> I haven't seen that one yet

[22-Mar-2010 09:55:12] <venturaville> can you do this:

[22-Mar-2010 09:55:22] <venturaville> virsh -c esx://myhostname/?no_verify=1

[22-Mar-2010 09:55:29] <venturaville> and then run some commands for me?

[22-Mar-2010 09:55:54] <venturaville> I don't have a vsphere 4.0 that I can query at the moment

[22-Mar-2010 09:56:49] <rmatte> sure

[22-Mar-2010 09:56:56] <rmatte> one second

[22-Mar-2010 09:57:46] <rmatte> ok I'm in

[22-Mar-2010 09:57:49] <rmatte> wutcha need me to run?

[22-Mar-2010 09:58:06] <venturaville> list

[22-Mar-2010 09:58:06] <rmatte> (did you want this done on one of the servers throwing the error, or on the one not generating RRDs?

[22-Mar-2010 09:58:22] <venturaville> one not generating RRDs

[22-Mar-2010 09:58:24] <rmatte> k

[22-Mar-2010 09:58:26] <rmatte> one second

[22-Mar-2010 09:58:26] <venturaville> net-list

[22-Mar-2010 09:58:28] <venturaville> vol-list

[22-Mar-2010 09:58:43] <venturaville> pool-list

[22-Mar-2010 09:59:50] <rmatte> sent it to you via privmsg

[22-Mar-2010 10:00:14] <venturaville> k

[22-Mar-2010 10:00:24] <venturaville> brb ... have to switch meeting rooms ...be back in about 5 minutes

[22-Mar-2010 10:05:36] <venturaville> rmatte: I'm back

[22-Mar-2010 10:13:31] <rmatte> hey

[22-Mar-2010 10:13:44] <rmatte> ok, let me know of any other commands you need me to run

[22-Mar-2010 10:17:49] <venturaville> unfortunately I can't get any of them to return data now

[22-Mar-2010 10:17:59] <venturaville> I had it working from home on one I was trying

[22-Mar-2010 10:18:18] <venturaville> I'm wondering if that was the development branch of ESX driver I was using :-/

[22-Mar-2010 10:19:04] <venturaville> apparently I am going to have to do some digging in the libvirt base to get it working

[22-Mar-2010 10:19:17] <rmatte> k

[22-Mar-2010 10:19:29] <venturaville> how about I get back to you later this week ... I am finishing up some pools/volumes code that I need to commit first :)

[22-Mar-2010 10:19:40] <rmatte> k cool

[22-Mar-2010 10:21:03] * rmatte ywans

[22-Mar-2010 10:21:11] <rmatte> s/ywans/yawns

[22-Mar-2010 10:22:45] <Michelle1> Anybody have any advice on testing Dell OpenManage traps for windows?  I setup everything and am attempting to test the traps by changing the System Board Ambient Temp Max Warning threshhold so that a warning is generated but I don't get an event in Zenoss.  I do get an event if I restart the SNMP service on the client though.

[22-Mar-2010 10:23:14] <Michelle1> I'm not sure how to make sure the traps are working properly.

[22-Mar-2010 10:23:47] <rmatte> well, I assume you're talking about 2 separate snmp agents?

[22-Mar-2010 10:24:11] <Michelle1> I am?

[22-Mar-2010 10:24:36] <Michelle1> The documentation on setting this up is not very good so feel free to elaborate.

[22-Mar-2010 10:24:57] <Michelle1> I'm also new to the whole Zenoss and SNMP thing

[22-Mar-2010 10:25:52] <Michelle1> We use WMI to monitor the Windows boxes but also want to be able to use Zenoss to alert on the Dell OpenManage alerts too.

[22-Mar-2010 10:26:35] <Michelle1> I followed the info out there on enabling SNMP and setting the trap destination and community in both the SNMP service and in OpenManage.

[22-Mar-2010 10:27:38] <rmatte> ok...

[22-Mar-2010 10:28:02] <rmatte> Dell OpenManage is a piece of software running on the server?

[22-Mar-2010 10:28:09] <Michelle1> yep

[22-Mar-2010 10:28:38] <rmatte> ok, the main question is, does openmanage know how to send traps and respond to SNMP without interfering with system SNMP

[22-Mar-2010 10:28:49] <rmatte> I'm not familliar with OpenManage myself, but that's the main question

[22-Mar-2010 10:29:06] <Michelle1> that's where i'm confused as well

[22-Mar-2010 10:29:15] <rmatte> Basically, you've got 2 SNMP agents running on the same system, and if one is not able to jive with the other, then you'll have issues

[22-Mar-2010 10:29:38] <rmatte> you'd have to check the documentation to see how it handles that sort of thing

[22-Mar-2010 10:37:45] <rmatte> My guess is that the system SNMP agent already has the trap port in use, so the OpenManage software can't send traps

[22-Mar-2010 10:37:57] <rmatte> just speculation though

[22-Mar-2010 12:39:19] <NewBee> What is the best way to get a list of all the devices install in my zenoss system from the command line?

[22-Mar-2010 12:41:11] <rmatte> you'd have to script something to do that, but it wouldn't be hard

[22-Mar-2010 12:41:42] <rmatte> you literally just want a list of device names with nothing else, or what?

[22-Mar-2010 12:41:58] <NewBee> name and ip

[22-Mar-2010 12:43:01] <rmatte> http://fpaste.org/muQN/

[22-Mar-2010 12:43:11] <rmatte> save that as a .py file, chmod +x it, and execute it

[22-Mar-2010 12:43:33] <rmatte> (as the zenoss user obviously)

[22-Mar-2010 12:45:19] <mrayzenoss> I just renamed a couple of the zendmd examples in community/documentation/wiki/zendmd

[22-Mar-2010 12:45:52] <mrayzenoss> from things like "Useful tips" to "Ddocs/DOC-2525", figured specific is better

[22-Mar-2010 12:46:10] <rmatte> cool

[22-Mar-2010 12:46:20] <mrayzenoss> feel free to add that example

[22-Mar-2010 12:47:00] <NewBee> is that %-30s

[22-Mar-2010 12:47:06] lthrashe1 is now known as lthrasher

[22-Mar-2010 12:47:09] <rmatte> that's just spacing

[22-Mar-2010 12:47:15] <NewBee> it's not happy with that line.

[22-Mar-2010 12:47:21] <rmatte> oh?

[22-Mar-2010 12:47:24] <rmatte> oh wait, my bad

[22-Mar-2010 12:47:26] <rmatte> one sec

[22-Mar-2010 12:47:29] <NewBee> np

[22-Mar-2010 12:48:29] <rmatte> http://fpaste.org/PJqY/

[22-Mar-2010 12:48:31] <rmatte> there

[22-Mar-2010 12:48:32] <rmatte> that'll work

[22-Mar-2010 12:49:19] <NewBee> That is perfect.

[22-Mar-2010 12:50:06] <NewBee> is there some place i can see what other things  dmd.Devices.getSubDevices has in it. I'm new to python.

[22-Mar-2010 12:51:10] <rmatte> you mean the device properties that you can call on?

[22-Mar-2010 12:51:19] <NewBee> yes

[22-Mar-2010 12:51:31] <NewBee> so you have d.id

[22-Mar-2010 12:51:50] <NewBee> and d.manageIp

[22-Mar-2010 12:51:54] <NewBee> is there d.status

[22-Mar-2010 12:52:05] <rmatte> yes, there is

[22-Mar-2010 12:52:07] <rmatte> do this...

[22-Mar-2010 12:52:16] <rmatte> become the zenoss user and type zendmd

[22-Mar-2010 12:52:34] <NewBee> ok

[22-Mar-2010 12:52:56] <rmatte> then type...

[22-Mar-2010 12:53:16] <rmatte> d = dmd.Devices.findDevice('NameOrIpOfADevice')

[22-Mar-2010 12:53:27] <rmatte> obviously replace that with the name or IP of a device

[22-Mar-2010 12:53:40] <rmatte> then type d. and hit tab

[22-Mar-2010 12:53:47] <rmatte> you'll see a bunch of functions

[22-Mar-2010 12:53:54] <NewBee> cool

[22-Mar-2010 12:54:41] <rmatte> sorry, it's not status

[22-Mar-2010 12:54:44] <rmatte> it's d.getPingStatus()

[22-Mar-2010 12:54:50] <rmatte> type d.get then hit tab

[22-Mar-2010 12:54:52] <rmatte> to narrow it down

[22-Mar-2010 12:55:12] <NewBee> that is awesome

[22-Mar-2010 12:55:13] <rmatte> always put a () after the function

[22-Mar-2010 12:55:18] <rmatte> so that it retrieves the value

[22-Mar-2010 12:55:22] <NewBee> ok

[22-Mar-2010 12:55:33] <rmatte> zendmd is the best spot to test stuff when you're writing scripts for Zenoss

[22-Mar-2010 12:55:56] <rmatte> the ping status is in numerical values, but you can do something like...

[22-Mar-2010 12:56:39] <rmatte> http://fpaste.org/GLZ0/

[22-Mar-2010 12:56:59] <rmatte> that'll print out the name, IP, ping status in english (UP,DOWN,etc...) and uptime of each device

[22-Mar-2010 12:59:21] <NewBee> perfect

[22-Mar-2010 13:04:04] <rmatte> python is pretty simple to learn, you just have to dig in

[22-Mar-2010 13:44:57] <venturaville> why is it command datasource test with ssh enabled goes to the localhost?

[22-Mar-2010 13:45:21] <rmatte> you specified a host in the test I assume?

[22-Mar-2010 13:45:48] <rmatte> like, you didn't just click test without entering a device?

[22-Mar-2010 13:46:18] <venturaville> yep

[22-Mar-2010 13:46:28] <rmatte> weird, must be a bug

[22-Mar-2010 13:47:56] <venturaville> looks like it

[22-Mar-2010 13:48:14] <venturaville> took me a bit to figure out why the numbers I was getting back from /proc/diskstats were not matching up with what I was seeing

[22-Mar-2010 13:53:35] <davidX-> while using the server/cmd class should i see any performance graphs?

[22-Mar-2010 13:59:14] <mrayzenoss> davidX-: what type of device?

[22-Mar-2010 13:59:50] <venturaville> linux device

[22-Mar-2010 14:00:15] <mrayzenoss> using SSH or SNMP?

[22-Mar-2010 14:00:17] <venturaville> /bin/grep sda3 /proc/diskstats | awk ' { print "|rio=" $$4 " ruse=" $$7 " wio=" $$8 " wuse=" $$11 } '

[22-Mar-2010 14:00:18] <venturaville> ssh

[22-Mar-2010 14:00:43] <mrayzenoss> venturaville: I was asking davidX- :)

[22-Mar-2010 14:01:07] <davidX-> mrayzenoss: luckily the answer is the same :D linux server

[22-Mar-2010 14:01:52] <mrayzenoss> venturaville: I'm not sure your case is a bug or just intended behavior.  SSH'ing to localhost to run the command with no host could either be a bug or a feature

[22-Mar-2010 14:02:00] <mrayzenoss> depending how you feel

[22-Mar-2010 14:02:16] <mrayzenoss> davidX-: if you're monitoring via SNMP, put it in /Server/Linux

[22-Mar-2010 14:02:33] <davidX-> i'm giving ssh monitoring a shot

[22-Mar-2010 14:02:36] <mrayzenoss> davidX-: if you're using SSH, make sure you have the LinuxMonitor ZenPack installed and put it in /Server/SSH/Linux

[22-Mar-2010 14:02:42] <davidX-> ahhh

[22-Mar-2010 14:02:44] <davidX-> thx

[22-Mar-2010 14:03:00] <mrayzenoss> davidX-: you may want the LinuxMonitorAddOn and your distro-specific ZenPacks as well

[22-Mar-2010 14:03:06] <mrayzenoss> for more goodness

[22-Mar-2010 14:03:20] * mrayzenoss is testing OpenSolaris ZenPack patches

[22-Mar-2010 14:04:16] <venturaville> mrayzenoss: I am putting in a hostname....

[22-Mar-2010 14:04:31] <mrayzenoss> venturaville: is the hostname == a device you have?

[22-Mar-2010 14:04:34] <venturaville> yep

[22-Mar-2010 14:04:42] <venturaville> It is already modeled

[22-Mar-2010 14:04:49] <mrayzenoss> and it's still going to localhost?

[22-Mar-2010 14:04:52] <venturaville> and I can ssh to it from the command line as well

[22-Mar-2010 14:04:54] <venturaville> yep

[22-Mar-2010 14:05:12] <venturaville> I can verify it because the /proc/diskstats numbers that are returned are from the zenoss host instead of the other one

[22-Mar-2010 14:05:25] <davidX-> mrayzenoss: thx, i'll give it a shot

[22-Mar-2010 14:05:31] <mrayzenoss> venturaville: 2.5.2?

[22-Mar-2010 14:05:54] <venturaville> 2.5.1 on this particular guy

[22-Mar-2010 14:07:05] <mrayzenoss> venturaville: and you hit 'save' before you tried the test?

[22-Mar-2010 14:07:11] <venturaville> yep

[22-Mar-2010 14:07:26] <mrayzenoss> what's the command?

[22-Mar-2010 14:07:35] <venturaville> /bin/grep sda3 /proc/diskstats | awk ' { print "|rio=" $$4 " ruse=" $$7 " wio=" $$8 " wuse=" $$11 } '

[22-Mar-2010 14:07:51] <venturaville> that is what goes in the box anyways

[22-Mar-2010 14:07:53] <mrayzenoss> simple enough…

[22-Mar-2010 14:19:27] <davidX-> mrayzenoss: how long should it take for graphs to generate/populate with the linux addon?

[22-Mar-2010 14:19:53] <mrayzenoss> davidX-: after installing the zenpack, restart zope

[22-Mar-2010 14:20:07] <mrayzenoss> davidX-: move your linux device into /Server/SSH/Linux and remodel it

[22-Mar-2010 14:20:18] <mrayzenoss> davidX-: after 15 minutes you should have data points

[22-Mar-2010 14:23:13] <davidX-> danke

[22-Mar-2010 14:34:33] <chewbranca> anyone use zenoss to monitor rails apps in addition to normal server monitoring?

[22-Mar-2010 14:35:34] <mrayzenoss> I've seen people doing process monitoring with rails and http or synthetic web transactions of their sites

[22-Mar-2010 14:36:28] <chewbranca> mrayzenoss, hrmm.. interesting, haven't found a whole lot of information on zenoss and rails, but I did find out that you can script events to zenoss with xmlrpc, which is very intriguing

[22-Mar-2010 14:36:39] <chewbranca> I'm trying to find a full stack monitoring system right now

[22-Mar-2010 14:36:58] <chewbranca> exploring zenoss as an option right now

[22-Mar-2010 14:37:12] <mrayzenoss> I did see this recently: http://rubygems.org/gems/zenoss_client

[22-Mar-2010 14:37:21] <mrayzenoss> speaking of Ruby

[22-Mar-2010 14:38:03] <chewbranca> ahhh nice, I'll take a peak at that

[22-Mar-2010 14:54:19] <rmatte> chewbranca: the synthetic transactions ZenPack is really good for monitoring web apps as well, it uses twill which is braindead to script in

[22-Mar-2010 14:56:57] <chewbranca> rmatte, hrmm... interesting

[22-Mar-2010 14:57:07] <chewbranca> checkign out the docs for that now

[22-Mar-2010 15:06:51] <rmatte> you can script it so that it automatically fills out forms and submits them, then checks to make sure that they submitted correctly

[22-Mar-2010 15:07:26] <rmatte> I've got a few scripts, one logs in to a site then clicks the logout button, just to make sure that the backend of the site is working (since the login screen rarely fails but the backend fails often enough)

[22-Mar-2010 15:18:46] <chewbranca> ahhh nice

[22-Mar-2010 15:19:49] <chewbranca> that could definitely be useful, I've got a lot of random things that aren't part of the web facing side of the app that need to be monitored, so I really like the ability to send events through xmlrpc, found an interesting article here: http://dancingpenguinsoflight.com/2009/05/send-events-to-zenoss-from-scripts/

[22-Mar-2010 15:20:38] <rmatte> yup

[22-Mar-2010 15:20:45] <rmatte> you can do tons of stuff with the Zenoss API

[22-Mar-2010 15:20:56] <davidX-> mrayzenoss: do you know of any extensive docs for the virtual host xen monitor zenpack?

[22-Mar-2010 15:21:15] <rmatte> davidX-: you talking about the libvirt pack?

[22-Mar-2010 15:21:36] <rmatte> or some other xen pack?

[22-Mar-2010 15:22:06] <rmatte> oh, I see which one

[22-Mar-2010 15:22:17] <rmatte> you should try out the libvirt pack, it does xen monitoring and does it well apparently

[22-Mar-2010 15:22:22] <davidX-> ahh

[22-Mar-2010 15:22:48] <rmatte> oh actually, it's untested with Xen

[22-Mar-2010 15:22:56] <rmatte> so you might want to just use the pack you were looking at

[22-Mar-2010 15:23:13] <davidX-> i don't mind testing it

[22-Mar-2010 15:23:18] <davidX-> i'm testing the xen pack against a dev env

[22-Mar-2010 15:23:44] <rmatte> davidX-: but to answer your question, read the Zenpack page

[22-Mar-2010 15:23:47] <rmatte> it says "This ZenPack is documented in the Zenoss Extended Monitoring Guide."

[22-Mar-2010 15:23:53] <rmatte> with a link to the extended monitoring guide

[22-Mar-2010 15:24:14] <davidX-> yeah, the documentation there is, "here's how you add it."

[22-Mar-2010 15:25:09] <rmatte> docs/DOC-5541

[22-Mar-2010 15:25:14] <rmatte> there's more there than just how to add it

[22-Mar-2010 15:25:26] <davidX-> ahh!

[22-Mar-2010 15:25:30] <davidX-> many thanks

[22-Mar-2010 15:25:33] <rmatte> np

[22-Mar-2010 15:31:09] <davidX-> libvirt looks pretty itneresting

[22-Mar-2010 15:31:15] <davidX-> interesting, too..

[22-Mar-2010 15:34:16] <rmatte> yeh, it is

[22-Mar-2010 15:34:35] <rmatte> the pack isn't working well for ESX right now but the author (venturaville) said he'd try to fix that by end of week

[22-Mar-2010 15:34:51] <rmatte> not sure how it's working for Xen as it appears to have not been tested yet

[22-Mar-2010 15:35:00] <rmatte> but it's nice because it actually uses the API of each virtualization platform

[22-Mar-2010 15:36:52] <rmatte> Once he fixes the ESX stuff it'll be awesome for VMWare monitoring

[22-Mar-2010 15:42:21] <davidX-> we have a ton of virt platforms

[22-Mar-2010 15:42:30] <davidX-> hyperv, esx, esxi, xen, etc

[22-Mar-2010 15:42:43] <davidX-> so, this is definitely exciting

[22-Mar-2010 15:42:49] <rmatte> cool, you could really help out in testing that zenpack then

[22-Mar-2010 15:42:56] <rmatte> and once it's all polished up it'll be really useful for you

[22-Mar-2010 15:43:11] <rmatte> I'm helping venturaville with ESX testing

[22-Mar-2010 15:43:12] <davidX-> i'll test against xen today

[22-Mar-2010 15:43:17] <rmatte> sweet

[22-Mar-2010 15:44:11] <rmatte> by the way, you'll need to install libvirt 0.7.0 or higher (0.7.7 worked best for me and it's the latest)

[22-Mar-2010 15:44:24] <rmatte> if you need help installing it let me know, I spent tons of time figuring it out ;)

[22-Mar-2010 15:45:05] <rmatte> but 0.7 or higher wasn't available in my package manager

[22-Mar-2010 15:45:11] <rmatte> so I had to compile from source

[22-Mar-2010 15:49:33] <davidX-> grabbing the tarball now

[22-Mar-2010 15:49:37] <rmatte> cool

[22-Mar-2010 15:50:03] <rmatte> the dependencies for compiling from source are...

[22-Mar-2010 15:50:10] <rmatte> libxml2-dev

[22-Mar-2010 15:50:15] <rmatte> libdevmapper-dev

[22-Mar-2010 15:50:23] <rmatte> libcurl4-gnutls-dev

[22-Mar-2010 15:50:29] <rmatte> python-dev

[22-Mar-2010 15:50:47] <rmatte> you'll also need some xen package, I didn't compile mine with xen support because we don't use Xen

[22-Mar-2010 15:50:56] <rmatte> I'm sure you'll figure that part out

[22-Mar-2010 15:51:20] <rmatte> make sure you do ./compile --with-esx so that you have esx support as well

[22-Mar-2010 15:51:28] <rmatte> all the other support is active by default

[22-Mar-2010 15:53:29] <rmatte> getting libvirt installed is 95% of the battle with that pack

[22-Mar-2010 15:54:42] <davidX-> got it, thx

[22-Mar-2010 15:54:52] <rmatte> np

[22-Mar-2010 15:55:25] <QbY> is their a mailing list for zenoss users  or are the discussion forums on the page the best place to post question/

[22-Mar-2010 15:55:38] <rmatte> discussion forums are the best spot

[22-Mar-2010 15:55:46] <QbY> darn

[22-Mar-2010 15:55:47] <rmatte> besides this chatroom

[22-Mar-2010 15:55:51] <QbY> 85 views and not a response et

[22-Mar-2010 15:55:52] <QbY> yet

[22-Mar-2010 15:55:57] <rmatte> link?

[22-Mar-2010 15:56:05] <QbY> thread/13067?tstart=0

[22-Mar-2010 15:56:51] <rmatte> oh, I know how to do that

[22-Mar-2010 15:57:04] <QbY> oh..  you are now my new best friend

[22-Mar-2010 15:57:11] <QbY> you wanna tell me or point me to docs

[22-Mar-2010 15:57:17] <rmatte> I'm a Zenoss master, I know quite a bit ;)

[22-Mar-2010 15:57:24] <QbY> if this is solved, along with one other issue--i can deploy it

[22-Mar-2010 15:57:32] <mrayzenoss> rmatte: lol

[22-Mar-2010 15:57:33] <rmatte> I'll show you an example of a threshold that I've setup...

[22-Mar-2010 15:57:38] <QbY> ok

[22-Mar-2010 15:57:39] <jb> rmatte: please dont ever say that again :P

[22-Mar-2010 15:57:50] <rmatte> mrayzenoss: I'm allowed to reference myself as such :)

[22-Mar-2010 15:58:03] <rmatte> jb: :P

[22-Mar-2010 15:58:14] <mrayzenoss> rmatte: absolutely

[22-Mar-2010 15:58:17] * Simon4_ boggles

[22-Mar-2010 15:58:30] <QbY> Sup Simon

[22-Mar-2010 15:58:32] * Simon4_ doesn't want to see rmatte's threshold

[22-Mar-2010 15:58:35] <Simon4_> that's for sure :)

[22-Mar-2010 15:58:55] <Simon4_> heya QbY :)

[22-Mar-2010 15:59:02] <Simon4_> back shortly

[22-Mar-2010 15:59:06] <mrayzenoss> Simon4_: I ran your blog post by a couple of the developers, the consensus is that it will break on Windows

[22-Mar-2010 15:59:23] <jb> what wiill?

[22-Mar-2010 15:59:25] <rmatte> Well, the max value for my template is set to: (here.getRRDValue('mem5minUsed') + here.getRRDValue('mem5minFree')) * 0.9

[22-Mar-2010 15:59:30] <Simon4_> mrayzenoss: very possible, I have no windows machines - need to have a test of it

[22-Mar-2010 15:59:36] <rmatte> just to give you an idea of how you could do yours

[22-Mar-2010 15:59:56] <QbY> rmatte: so those are two different data points?

[22-Mar-2010 15:59:58] * Simon4_ will edit to mention that however - thanks :)

[22-Mar-2010 16:00:20] <Simon4_> rmatte: you might just have solved me a bunch of issues also with that

[22-Mar-2010 16:00:23] <rmatte> QbY: yes

[22-Mar-2010 16:00:26] <Simon4_> (the two datapoint threshold thing)

[22-Mar-2010 16:01:10] <QbY> rmatte: if i edit it in the template, it will automatically deploy to all my devices/interfaces of that type correct?

[22-Mar-2010 16:01:37] <rmatte> I also put the following transform in the /Events/Perf/Memory event class: http://fpaste.org/rjHr/

[22-Mar-2010 16:01:44] <rmatte> Since the event comes in for that class

[22-Mar-2010 16:01:55] <rmatte> The transform makes the memory percentage human readable

[22-Mar-2010 16:02:02] <rmatte> rather than just showing the bits used

[22-Mar-2010 16:02:12] <QbY> ah

[22-Mar-2010 16:02:27] <rmatte> QbY: correct, as long as those devices or interfaces are inheriting the template

[22-Mar-2010 16:02:33] <QbY> i have that with several of these, the values are actually integers.  but they need to be decimales (/10)

[22-Mar-2010 16:03:37] <rmatte> ok, but I assume you're ok with the examples I just gave?

[22-Mar-2010 16:03:43] <rmatte> any questions?

[22-Mar-2010 16:03:52] <QbY> i'm waiting on the next poll--i set it at 100% so that it will definitely fail

[22-Mar-2010 16:03:59] <rmatte> k

[22-Mar-2010 16:04:22] <QbY> rmatte: so, i have another--i need to alarm if a certain condition is met, but not every poll--maybe once a day

[22-Mar-2010 16:05:15] <QbY> so even though it exceeds my threshold--i don't want it to be an event at every poll

[22-Mar-2010 16:05:33] <rmatte> QbY: so only during a certain time of the day?

[22-Mar-2010 16:05:49] <QbY> maybe once a day--or even once a week for this one

[22-Mar-2010 16:06:14] <rmatte> hmmmm

[22-Mar-2010 16:06:36] <QbY> we do cable modems--so for the two data points, i need to see if 30% of the modems on an interface are gone--if so, i need an alarm every 5 minutes until someone acknolwedges it

[22-Mar-2010 16:06:37] <rmatte> you'll have to do it with a transform... but that'll be tricky if you clear the event it won't know

[22-Mar-2010 16:06:44] <rmatte> docs/DOC-5822

[22-Mar-2010 16:06:56] <rmatte> you could use that, but only have the event come in for a certain minute

[22-Mar-2010 16:07:08] <QbY> rmatte: !!!!!!!!  YOU ARE THE ZENOSS MASTER!!!!!!!!!!!!

[22-Mar-2010 16:07:24] <rmatte> jb: toldja

[22-Mar-2010 16:08:27] <rmatte> but yeh, your options are to do it with a transform, or to mess around with your alerting rules

[22-Mar-2010 16:08:36] <rmatte> maybe setup a custom alert schedule/repeat time for just that 1 event

[22-Mar-2010 16:08:46] <rmatte> and then have the rest of the events follow a different alerting rule

[22-Mar-2010 16:09:35] <QbY> what about alerting folks--i've got my different locations set up, and devices in groups--i want to alert the group if one of their devices is failing

[22-Mar-2010 16:09:55] <QbY> i added them in as a group with an email alert, but everyone was getting everyone's alerts

[22-Mar-2010 16:09:55] <rmatte> just create as many rules as you need with appropriate filters

[22-Mar-2010 16:10:38] <QbY> man, Zenoss is the s*%!

[22-Mar-2010 16:10:39] <rmatte> maybe use groups to categorize the devices by ownership

[22-Mar-2010 16:10:57] <rmatte> then have multiple alerting rules with filters based on each group

[22-Mar-2010 16:11:02] <rmatte> that send alerts to the appropriate people

[22-Mar-2010 16:11:21] <rmatte> yeh, zenoss is quite nice

[22-Mar-2010 16:11:33] <diatonic> Just upgraded to 2.5.2 (upgraded to latest WMIDataSource & WMIPerf_Windows ZenPacks first) and my event console is flooded with 'warnings' from $ZENHOME/Products/WMIPerformanceMonitor/libexec/wmi_stats.pl not executing. It doesn't appear to exist on the filesystem. Thoughts?

[22-Mar-2010 16:11:51] <rmatte> well, time to upgrade my production test instance to 2.5.2

[22-Mar-2010 16:12:12] <rmatte> diatonic: did you restart zenoss after installing those packs?

[22-Mar-2010 16:12:20] <rmatte> If not, then that's your problem

[22-Mar-2010 16:12:21] <diatonic> rmatte, I did :(

[22-Mar-2010 16:12:36] <rmatte> so you installed the packs and completely restarted zenoss after installing them?

[22-Mar-2010 16:12:48] <diatonic> rmatte, Yep, even rebooted the server

[22-Mar-2010 16:12:49] <rmatte> actually, hold up a second

[22-Mar-2010 16:13:07] <QbY> rmatte: i am concerned about traps/informs--a) the event is the oid even though the mibs are installed; and b) i believe i'm missing informs

[22-Mar-2010 16:13:09] <rmatte> wmi_stats.pl isn't from any of Egor's packs

[22-Mar-2010 16:13:15] <rmatte> that's from some ancient WMI pack

[22-Mar-2010 16:13:22] <rmatte> remove any WMI packs that aren't authored by Egor

[22-Mar-2010 16:13:32] <rmatte> then download Egor's packs from the community site and install those

[22-Mar-2010 16:14:15] <rmatte> his packs make all the other WMI packs obsolete (except for the performance packs developed by other people but which have WMI Data Source ZenPack listed as a dependency

[22-Mar-2010 16:14:31] <diatonic> rmatte, Thanks, I'll give that a shot. I had removed the 2 WMI packs that were installed before the upgrade ... but I'll see what I can track down.

[22-Mar-2010 16:14:39] <QbY> is it possible to reclassify events in the event console?

[22-Mar-2010 16:14:53] <rmatte> QbY: be 100% sure that you have all of the MIBs installed, and not just 1 out of a set of MIBs

[22-Mar-2010 16:15:08] <rmatte> define reclassify

[22-Mar-2010 16:15:31] <rmatte> diatonic: k, you may have old templates laying around too

[22-Mar-2010 16:15:42] <QbY> rmatte:  -- they came in as "Unknown" so now i want to put them in the right Class that I've built

[22-Mar-2010 16:15:50] <rmatte> diatonic: those old WMI packs are zip zenpacks, which suck at cleaning themselves up when removed

[22-Mar-2010 16:15:51] <diatonic> rmatte, Think remodeling the devices would help?

[22-Mar-2010 16:15:59] <rmatte> diatonic: check /Server/Windows -> Templates

[22-Mar-2010 16:16:00] <QbY> rmatte: yes, following Cisco's website, I installed all of the required dependencies

[22-Mar-2010 16:16:12] <rmatte> you'll probably find an old template referencing that .pl file

[22-Mar-2010 16:16:22] <diatonic> rmatte, Thanks!

[22-Mar-2010 16:16:40] <rmatte> QbY: Ummm, perhaps you should try installing my Cisco MIBs ZenPack instead?

[22-Mar-2010 16:16:49] <rmatte> There are tons of Cisco MIBs, it would be very easy to miss one

[22-Mar-2010 16:17:07] <rmatte> diatonic: no problem

[22-Mar-2010 16:17:40] <rmatte> QbY: you have to reclassify them by mapping them to a specific event class

[22-Mar-2010 16:17:52] <rmatte> QbY: what types of events are we talking here, traps, syslog?

[22-Mar-2010 16:18:25] <QbY> rmatte: yes, i have already build the class-but these came in before the class--can they be classified after the fact?

[22-Mar-2010 16:18:31] <rmatte> no

[22-Mar-2010 16:18:36] <rmatte> once they are in they are in

[22-Mar-2010 16:18:44] <diatonic> rmatte, You are correct, there was an obsolete template there.

[22-Mar-2010 16:18:46] <rmatte> unless you feel like going in to MySQL and mucking about with them directly

[22-Mar-2010 16:18:50] <rmatte> but I'd advise against it

[22-Mar-2010 16:19:01] <rmatte> diatonic: cool

[22-Mar-2010 16:19:58] <QbY> i'll live with them..

[22-Mar-2010 16:20:17] <rmatte> k hehe

[22-Mar-2010 16:20:21] <rmatte> afk for a sec

[22-Mar-2010 16:22:13] <rmatte> back

[22-Mar-2010 16:24:27] <QbY> rmatte: you spoke of transforms earlier.  i have snmp values i'm reading that are integers for example 150..  however its actually 15.0 dBmV..  everyone is used to seeing 15.0..  would a transform be the way to fix that?

[22-Mar-2010 16:24:56] <rmatte> well, a transform is one way, but you really just want it to be visible as 15.0 on the graphs right?

[22-Mar-2010 16:25:02] <rmatte> or in the event as well?

[22-Mar-2010 16:25:19] <QbY> both

[22-Mar-2010 16:25:53] <rmatte> ok... you're going to need a transform and an RPN expression for the graph

[22-Mar-2010 16:26:03] <rmatte> what does the event look like when it comes in?

[22-Mar-2010 16:26:49] <QbY> well for the most part its not an event until it falls below a certain level

[22-Mar-2010 16:26:59] <rmatte> ok, so it's a threshold event

[22-Mar-2010 16:27:03] <QbY> yes

[22-Mar-2010 16:27:09] <rmatte> what event class does it come in for?

[22-Mar-2010 16:27:14] <QbY> below 200 integer or 20.0 dBmV I need to alarm

[22-Mar-2010 16:27:23] <QbY> DOCSIS/SignalQuality

[22-Mar-2010 16:28:22] <rmatte> k, well, you saw the example I pasted earlier of how I did my transform to make the memory human readable right?

[22-Mar-2010 16:28:37] <rmatte> you'd have to do something similar, but even simpler

[22-Mar-2010 16:28:44] <QbY> let me scroll back--i'm so excited about this mass de-registration

[22-Mar-2010 16:28:47] <QbY> that you made it work

[22-Mar-2010 16:28:50] <rmatte> hehe

[22-Mar-2010 16:29:02] <QbY> i kept hearing all weekend "see--open source isn't it"

[22-Mar-2010 16:29:06] <chewbranca> anyone know approximately how much resources is required to run zenoss monitoring on a server?

[22-Mar-2010 16:29:15] <rmatte> open source is it, it just takes time to learn it

[22-Mar-2010 16:29:27] <rmatte> I've spent a year learning Zenoss, once you really learn it well it's extremely powerful

[22-Mar-2010 16:29:36] <rmatte> it's just not super straightforward for first time users

[22-Mar-2010 16:30:11] <rmatte> our management is quite impressed with what I've been able to squeeze out of it

[22-Mar-2010 16:30:19] <QbY> rmatte: .. no, it isn't, but its kinda easy to figure out--i am going to update (if you don't) my post and show how it was fixed

[22-Mar-2010 16:30:28] <rmatte> k

[22-Mar-2010 16:30:30] <QbY> because i'm sure i'm not the only person

[22-Mar-2010 16:30:37] <rmatte> yup, good idea

[22-Mar-2010 16:30:50] <rmatte> always good to follow up so that people coming across it in the future don't just hit a dead end ;)

[22-Mar-2010 16:31:04] <QbY> if you can, i want to make sure that i can get this event set right

[22-Mar-2010 16:31:11] <QbY> becuase this is the most important alarm we have

[22-Mar-2010 16:31:41] <rmatte> k

[22-Mar-2010 16:32:08] <rmatte> well, show me what you have when you're done and I'll tell you whether it should work or not

[22-Mar-2010 16:32:11] <QbY> the threshold is going into event class "Service/Availbility"

[22-Mar-2010 16:32:23] <rmatte> k

[22-Mar-2010 16:32:35] <QbY> so would i start by adjusting it there--also, i don't kno wif this matters

[22-Mar-2010 16:32:41] <QbY> lets say there are 200 modems on an interface

[22-Mar-2010 16:33:05] <QbY> if there are only 170 active, that's an alarm.

[22-Mar-2010 16:33:10] <rmatte> by the way, sorry to interrupt, but how did you end up fixing your modelling issue?

[22-Mar-2010 16:33:17] <QbY> mine?

[22-Mar-2010 16:33:27] <rmatte> yeh, where it was taking like 5 years to model a device

[22-Mar-2010 16:33:39] <QbY> snmp-server view

[22-Mar-2010 16:33:52] <rmatte> oh, so it was a change on the device side?

[22-Mar-2010 16:33:54] <QbY> the cisco cmts has waaaaaaaaaaaaaay too much info--most of which is not necessary for monitoring/modeling

[22-Mar-2010 16:33:56] <QbY> yes

[22-Mar-2010 16:33:59] <rmatte> ah ok cool

[22-Mar-2010 16:34:26] <rmatte> ok, so what is the name of the datasource that represents the amount of modems

[22-Mar-2010 16:34:27] <QbY> it was just choking on that..  so i created a view for what was really needed--excluding the rest

[22-Mar-2010 16:34:38] <rmatte> and what is the name of the datasource that represents the amount of active?

[22-Mar-2010 16:35:21] <QbY> TotalModems is the number of modems seen on that niterface in the past 24 hours

[22-Mar-2010 16:35:30] <rmatte> k

[22-Mar-2010 16:35:33] <QbY> RegisteredModems are the total number of online / active modems

[22-Mar-2010 16:35:44] <rmatte> k, so those are the 2 we're working with then...

[22-Mar-2010 16:35:51] <QbY> yes

[22-Mar-2010 16:36:01] <QbY> they are both graphed too if that makes a difference

[22-Mar-2010 16:36:05] <rmatte> so you want to be alerted if the amount of active drops below 70% of the total amount

[22-Mar-2010 16:36:12] <QbY> yes

[22-Mar-2010 16:36:14] <rmatte> k...

[22-Mar-2010 16:36:17] <QbY> 85% actually

[22-Mar-2010 16:36:21] <rmatte> then you'd do...

[22-Mar-2010 16:37:25] <rmatte> well, first you'd create a new threshold

[22-Mar-2010 16:37:36] <QbY> which i have one..

[22-Mar-2010 16:37:36] <rmatte> and make the threshold against registered modems

[22-Mar-2010 16:37:44] <QbY> which i do

[22-Mar-2010 16:37:46] <rmatte> so select that from the list of datasources

[22-Mar-2010 16:37:52] <rmatte> then in the Min value (not max)

[22-Mar-2010 16:37:55] <rmatte> you'd put: here.getRRDValue('TotalModems') * 0.85

[22-Mar-2010 16:38:02] <QbY> that's exactly what i have

[22-Mar-2010 16:38:10] <rmatte> that'll work

[22-Mar-2010 16:38:26] <rmatte> That's telling Zenoss that the minimum amount of modems has to be 85% of the total

[22-Mar-2010 16:38:32] <rmatte> if it drops below, it'll trigger the threshold

[22-Mar-2010 16:38:46] <QbY> and it did (i set it for 100% earlier just to check)

[22-Mar-2010 16:38:52] <rmatte> good stuff

[22-Mar-2010 16:39:29] <QbY> now.. before you go further..

[22-Mar-2010 16:39:34] <rmatte> but you want it to be 17.0 instead of 170?

[22-Mar-2010 16:39:42] <QbY> at 100 modems, if only 84 are online, its going to cross the threshold and alarm

[22-Mar-2010 16:39:49] <rmatte> that's right

[22-Mar-2010 16:40:00] <QbY> however, some deregs start at 84.. but can go down and down and down

[22-Mar-2010 16:40:05] <QbY> so next poll it could only be 40 online

[22-Mar-2010 16:40:14] <QbY> would that be a new event, or a continuation of the first

[22-Mar-2010 16:40:34] <QbY> the 17.0 vs 170 is for signal measurements

[22-Mar-2010 16:40:56] <rmatte> as long as it's for the same set of datapoints it would be a continuation of the first event, it would increase the event count on the event and update the current value in the summary

[22-Mar-2010 16:41:34] <rmatte> so for instance, first poll if it was 84, then you'd see "threshold of whatever not met: current value: 84"

[22-Mar-2010 16:41:37] <QbY> gotcha..  is it possible to in the email let them know the total missing--currently (our perl script) says, "there should be X but there are only Y"

[22-Mar-2010 16:41:57] <rmatte> then on the next poll, event count would go to 2 and the event summary would be automatically updated to "threshold of whatever not met: current value: 40"

[22-Mar-2010 16:42:21] <rmatte> hmmm

[22-Mar-2010 16:42:42] <rmatte> well, you could use a transform to figure that out

[22-Mar-2010 16:42:57] <rmatte> and update the summary with the amount missing

[22-Mar-2010 16:43:43] <rmatte> missing = here.getRRDValue('TotalModems') - here.getRRDValue('RegisteredModems')

[22-Mar-2010 16:44:01] <rmatte> evt.summary = 'Total missing: %s' % (missing)

[22-Mar-2010 16:44:03] <rmatte> as an example

[22-Mar-2010 16:44:29] <QbY>  when you say "transform" where is that at?

[22-Mar-2010 16:44:57] <rmatte> I recommend that you read the section of the Zenoss admin guide to get a basic understanding of it

[22-Mar-2010 16:45:05] <rmatte> it's a huge mouthful to explain everything in here

[22-Mar-2010 16:45:15] <rmatte> there's a whole section covering transforms

[22-Mar-2010 16:45:23] <rmatte> there is also tons of wiki documentation

[22-Mar-2010 16:45:28] <QbY> ok..

[22-Mar-2010 16:45:51] <QbY> so, in my emails..  which i need it to go every 5 minutes until someone silences it

[22-Mar-2010 16:45:59] <rmatte> a transform is just a python script

[22-Mar-2010 16:46:00] <QbY> do i do that in transforms also?

[22-Mar-2010 16:46:08] * QbY knows no python

[22-Mar-2010 16:46:09] <rmatte> but it's used to manipulate events automatically when they come in

[22-Mar-2010 16:46:26] <rmatte> that's fine, the small amount of python that you need to learn to do transforms is almost nothing

[22-Mar-2010 16:46:30] <rmatte> braindead simple

[22-Mar-2010 16:46:58] <rmatte> No, for the emails where you need it to go 5 minutes before someone silences it, you need to set the repeat time in your alerting rule

[22-Mar-2010 16:47:12] <rmatte> alert every 5 minutes, rather

[22-Mar-2010 16:48:06] <rmatte> and that example I gave before was a bit off, it would actually be...

[22-Mar-2010 16:48:09] <QbY> so settings-> users..  create a group

[22-Mar-2010 16:48:25] <rmatte> dev = dmd.Devices.findDevice(evt.ipAddress)

[22-Mar-2010 16:48:34] <rmatte> missing = dev.getRRDValue('TotalModems') - dev.getRRDValue('RegisteredModems')

[22-Mar-2010 16:48:39] <rmatte> evt.summary = 'Total missing: %s' % (missing)

[22-Mar-2010 16:48:55] <rmatte> well, you can either create a group or do alerts per individual user

[22-Mar-2010 16:48:57] <QbY> rmatte: but these are checked on the Interfaces of a device

[22-Mar-2010 16:49:31] <rmatte> oh, well, you can still do it, it'll just be finding an interface instead of the device itself

[22-Mar-2010 16:49:41] <rmatte> read up on transforms first so that you have a basic understanding

[22-Mar-2010 16:49:51] <rmatte> once you have that I'll help you out with the transform

[22-Mar-2010 16:50:55] <QbY> ok, last question--because you've given me enough to keep me busy all night

[22-Mar-2010 16:51:08] <QbY> i would like to use the graphs in an existing portal application we have

[22-Mar-2010 16:51:26] <QbY> is there a way i can put a url in the <img> tag?

[22-Mar-2010 16:52:03] <rmatte> well, that's a bit tough, I mean, you can get the link to the image by right clicking and doing copy link location...

[22-Mar-2010 16:52:09] <rmatte> or copy image location rather

[22-Mar-2010 16:52:16] <rmatte> but you need to be authenticated with Zenoss to see the image

[22-Mar-2010 16:52:28] <rmatte> so using the images in a portal would be challenging

[22-Mar-2010 16:52:38] <rmatte> the other option is to access the RRD files directly using RRDtools

[22-Mar-2010 16:52:43] <rmatte> but that's development work

[22-Mar-2010 16:53:29] <QbY> ah

[22-Mar-2010 16:53:41] <QbY> maybe a script on the server that would generate the rrd and feed it out

[22-Mar-2010 16:53:51] <QbY> i just don't want to go giving access to everyone to the system

[22-Mar-2010 16:53:57] <QbY> unless its a very restricted access

[22-Mar-2010 16:54:17] <rmatte> well, if you give them regular user access, they won't be able to modify anything

[22-Mar-2010 16:54:21] <rmatte> they'll just be able to view stuff

[22-Mar-2010 16:54:43] <rmatte> afk, brb

[22-Mar-2010 17:01:40] <rmatte> back

[22-Mar-2010 17:02:14] <rmatte> my Zenoss instances are integrated with LDAP and I can control user access by adding them to specific LDAP groups that I've created

[22-Mar-2010 17:02:31] <rmatte> so our frontline NOC guys get added to the FrontLine group and only get regular user access

[22-Mar-2010 17:02:45] <rmatte> Engineers get added to the Engineers group and get full access

[22-Mar-2010 17:03:11] <rmatte> ...but I'm dealing with multiple Zenoss instances, doesn't make much sense if you only have 1 instance

[22-Mar-2010 17:04:06] <rmatte> you can also create custom roles which only have access to specific things, but you have to do it in Zope and it's a bit complex

[22-Mar-2010 17:04:13] <rmatte> I just stick with the basic roles

[22-Mar-2010 17:09:28] <sergeymasushko> hi to all

[22-Mar-2010 17:10:00] <rmatte> hey

[22-Mar-2010 17:10:11] <Simon4_> hi hi

[22-Mar-2010 17:11:04] <sergeymasushko> I'm trying to write a transform... that transform should suppress events for that class if there is an events on the other devices...

[22-Mar-2010 17:12:49] <sergeymasushko> what do you think about it:

[22-Mar-2010 17:12:49] <sergeymasushko> http://pastebin.com/kK05hsi6

[22-Mar-2010 17:13:24] <rmatte> similar to a transform that I did

[22-Mar-2010 17:13:27] <rmatte> does it work?

[22-Mar-2010 17:13:30] <rmatte> or have you not tested it?

[22-Mar-2010 17:13:46] <sergeymasushko> looks like it work...

[22-Mar-2010 17:13:51] <rmatte> cool

[22-Mar-2010 17:14:09] <sergeymasushko> but I'm not sure if is correct... maybe there is a better solution :)

[22-Mar-2010 17:15:18] <chewbranca> are there any android apps for tying into zenoss management?

[22-Mar-2010 17:15:45] <rmatte> chewbranca: not that I'm aware of, the only mobile app that I know of is for iphone

[22-Mar-2010 17:15:47] <sergeymasushko> rmatte: what is the your transform?

[22-Mar-2010 17:15:58] <rmatte> http://sourceforge.net/projects/zipec/

[22-Mar-2010 17:16:00] <mrayzenoss> chewbranca: I've heard rumblings of people looking at it, but not yet

[22-Mar-2010 17:16:02] <chewbranca> rmatte, bummer

[22-Mar-2010 17:16:16] <mrayzenoss> OpenSolaris ZenPack updated: docs/DOC-3414

[22-Mar-2010 17:16:16] <rmatte> sergeymasushko: one sec...

[22-Mar-2010 17:17:16] <chewbranca> hrmmm yeah the iphone app looks exactly as you would expect, I see there are planning blackberry support but no mention of android

[22-Mar-2010 17:17:27] <rmatte> sergeymasushko: http://fpaste.org/jd3V/

[22-Mar-2010 17:17:38] <rmatte> I need to work on it some more, it kind of works but doesn't work as I would expect it to

[22-Mar-2010 17:18:00] <chewbranca> mrayzenoss, ahhh yeah would be nice

[22-Mar-2010 17:19:33] <mrayzenoss> Various people have told me they're working on iPhone interfaces, hopefully I'll have something solid on that front soon.  An Android app would be similarly nice

[22-Mar-2010 17:19:46] <sergeymasushko> rmatte: heh... your transform a bit "smarter" :)

[22-Mar-2010 17:20:10] <rmatte> a bit, but it needs to be in my case

[22-Mar-2010 17:20:26] <rmatte> the thing about ZiPEC is that they haven't submitted it to the app store

[22-Mar-2010 17:20:32] <rmatte> so you need to jailbreak your phone to use it

[22-Mar-2010 17:20:36] <rmatte> which kind of sucks

[22-Mar-2010 17:20:37] <rmatte> lol

[22-Mar-2010 17:21:02] <rmatte> actually, it looks like it's just an ajax app

[22-Mar-2010 17:21:07] <rmatte> I could be wrong

[22-Mar-2010 17:21:16] <Simon4_> webapp it seems

[22-Mar-2010 17:21:24] <rmatte> yeh

[22-Mar-2010 17:21:33] <rmatte> it'd be cooler if it was an actual iphone app in the app store

[22-Mar-2010 17:21:35] <rmatte> less load time

[22-Mar-2010 17:22:14] <Simon4_> slightly trickier to make work I guess, would need to support http/https, then access one of the API's I guess

[22-Mar-2010 17:22:29] <Simon4_> (we hide public zenoss behind https on apache)

[22-Mar-2010 17:22:58] <rmatte> unhunh, well, so far Zenoss 2.5.2 is reporting that interfaces are down on interfaces that are clearly up

[22-Mar-2010 17:23:04] <rmatte> guess I'll have to tackle that one tomorrow

[22-Mar-2010 17:23:24] <chewbranca> problem with a webapp is that its request driven rather than event driven, I'm basically looking for a more advanced notifier than just getting an email

[22-Mar-2010 17:23:38] <rmatte> don't want to upgrade and end up with an interface monitoring nightmare

[22-Mar-2010 17:23:43] <rmatte> anyways, I'm out of hear, ttyl guys

[22-Mar-2010 17:23:55] <rmatte> out of here* rather

[22-Mar-2010 17:23:58] <chewbranca> although its looking like just writing an android app to parse your emails and respond based on filters is probably the simplest and most flexible solution

[22-Mar-2010 17:23:59] * sergeymasushko is sad due to there is no official apple store in his country :(

[22-Mar-2010 17:24:00] <rmatte> my brain has already left it seems

[22-Mar-2010 17:24:02] <rmatte> gnite folks

[22-Mar-2010 17:24:10] <chewbranca> later rmatte thanks for helping out with my questions

[22-Mar-2010 17:24:14] <rmatte> np

[22-Mar-2010 17:24:16] <rmatte> cheers

[22-Mar-2010 17:24:32] <mrayzenoss> I was talking with a couple guys at SCALE who said they're working on a native iPhone app, mostly focused on the Event Console

[22-Mar-2010 17:24:55] <Simon4_> mrayzenoss: that would be cool, with some push support would be even cooler

[22-Mar-2010 17:25:01] <chewbranca> mrayzenoss, nice, no iphone for me though

[22-Mar-2010 17:25:17] <Simon4_> since as chewbranca says, it's all about getting relevant stuff in a timely manner, which doesn't work on an iphone singletasking

[22-Mar-2010 17:25:22] <mrayzenoss> Simon4_: yeah, I told them to make sure they take advantage of all the hooks provided

[22-Mar-2010 17:25:45] <QbY> rmatte: in my mass de-registration (totalmodems * .85) how would i write the clear event?

[22-Mar-2010 17:26:14] <chewbranca> Simon4_, yeap, which is why I think basically just an email filter system with notifications and sounds attached is probably the best bet, then you're also not limited to one particular platform, I get lots of different notifications and would like to tie them all together

[22-Mar-2010 17:46:10] <mrayzenoss> Cisco MDS ZenPack updated: docs/DOC-3397

[22-Mar-2010 17:46:34] <lronPaw> hi

[22-Mar-2010 17:46:39] <lronPaw> anyone around?

[22-Mar-2010 17:47:45] <lronPaw> i need some help debugging why the postgresql zenpack doesnt work for a particular host, it does for others already setup but not new hosts

[22-Mar-2010 17:47:56] <lronPaw> and when you run the .py file directly its able to collect stats

[22-Mar-2010 17:48:06] <mrayzenoss> I gotta run, but which Postgres zenpack?

[22-Mar-2010 17:48:19] <lronPaw> um the learning one

[22-Mar-2010 17:48:22] <lronPaw> i'll find it now

[22-Mar-2010 17:48:30] <lronPaw> can you run the zenpack command via zenoss

[22-Mar-2010 17:48:34] <lronPaw> charlieS2: zencommand

[22-Mar-2010 17:48:42] <lronPaw> charlieS2: zencommand?

[22-Mar-2010 17:48:44] <lronPaw> argh

[22-Mar-2010 17:48:47] <lronPaw> nick completyion

[22-Mar-2010 17:49:31] <lronPaw> enPacks.LearningObjects.PostgresqlMonitor-1.1-py2.4.egg

[22-Mar-2010 17:50:42] <lronPaw> looked through all the logs for hostname references

[22-Mar-2010 17:51:15] <lronPaw> and havent been able to find anything

[22-Mar-2010 17:51:18] <sergeymasushko> lronPaw: so the zenpack is working for that host directly from the command line and didn't work for the host from the zenoss?

[22-Mar-2010 17:51:27] <lronPaw> yep

[22-Mar-2010 17:51:37] <lronPaw> and it works for other hosts (at least 2) through zenoss

[22-Mar-2010 17:52:01] <sergeymasushko> have you increased debug level logging for the zencommand ?

[22-Mar-2010 17:52:13] <lronPaw> i did a -d 10

[22-Mar-2010 17:52:15] <lronPaw> i think

[22-Mar-2010 17:52:35] <lronPaw> but like i said i couldn't find how to run the zenpack via zencommand

[22-Mar-2010 17:53:00] <sergeymasushko> after you added the zenpack to the device have you restarted the zencommand daemon?

[22-Mar-2010 17:53:10] <lronPaw> oh no i haven't

[22-Mar-2010 17:53:13] <lronPaw> recommended?

[22-Mar-2010 17:53:53] <sergeymasushko> please try... and wait some time... 5-10 minutes :) usually it helps me :)

[22-Mar-2010 17:54:07] <lronPaw> ok i'll give it a whirl

[22-Mar-2010 17:54:42] <lronPaw> is there a way you can run the commandline zenpack .py file through zenoss on the commandline?

[22-Mar-2010 17:54:50] <lronPaw> dibbzeh

[22-Mar-2010 17:54:55] <dibbz> raaa

[22-Mar-2010 17:55:06] <lronPaw> cos it would be useful to see if you can so you can debug the process

[22-Mar-2010 17:55:12] <lronPaw> or at least trace where its failin gin the logs

[22-Mar-2010 17:55:30] <lronPaw> its just instigating that command through zenoss thats the hard bit for me at this poiint

[22-Mar-2010 17:58:45] <sergeymasushko> lronPaw: if the command is ran through zenoss and there is an error so you should be able to find this info in the log... if the log is empty than you are looking in the wrong place of the command was not run.

[22-Mar-2010 18:04:45] <lronPaw> yeh well i am thinking the commands not being run for that device

[22-Mar-2010 18:04:56] <lronPaw> we have a considerable number of devices and only 1 collector at this stage

[22-Mar-2010 18:05:12] <lronPaw> thinking of adding another collector on a neighbouring host

[22-Mar-2010 18:05:38] <lronPaw> these devices that arent working have only recently been added, infact thye are probably the last devices to be added to zenoss

[22-Mar-2010 18:08:54] <lronPaw> so im wondering if zenoss never reaches that device

[22-Mar-2010 18:09:07] <lronPaw> we have ntp perf mon template active for this device and thats working

[22-Mar-2010 18:09:18] <lronPaw> same wwith the device template

[22-Mar-2010 18:11:51] rhettardo_ is now known as rhettardo

[22-Mar-2010 18:25:38] straterr1 is now known as straterra

[22-Mar-2010 18:48:52] <QbY> i'm testing some stuff--is it possible to force a poll on a device?

[22-Mar-2010 18:55:39] <lronPaw> remodel the device?

[22-Mar-2010 18:56:35] <QbY> k.

[22-Mar-2010 19:30:17] <QbY> hrmm..  this test is a pain..  if i manually add an event, will it follow the rules i've built?

[22-Mar-2010 23:59:38] <MrMyke> Hello all I am new here and just downloaded tool

[23-Mar-2010 00:00:18] <MrMyke> is anyone really here?

[23-Mar-2010 00:00:30] [disconnected at Tue Mar 23 00:00:30 2010]

[23-Mar-2010 00:00:30] [connected at Tue Mar 23 00:00:30 2010]

[23-Mar-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[23-Mar-2010 02:08:51] <Troubadix09> morning all

[23-Mar-2010 02:17:50] <forsberg> mornings

[23-Mar-2010 02:17:51] <forsberg> :)

[23-Mar-2010 02:39:24] <Egyptian[web]> good morning all

[23-Mar-2010 02:39:56] <Egyptian[web]> i would like to view the report that tells me what alerts have been sent in the last month

[23-Mar-2010 02:40:02] <Egyptian[web]> is there one like that? where is it located ?

[23-Mar-2010 02:40:49] <Egyptian[web]> or do i have to make my own :(

[23-Mar-2010 02:41:18] <forsberg> pretty sure you have to make your own, it shouldnt take too long to look through the available reports?

[23-Mar-2010 07:03:59] <ke4qqq>  /win 24

[23-Mar-2010 08:53:48] <pokui> hi all, how do I report spam/abuse via pms in the zeonoss forums?

[23-Mar-2010 08:54:14] <ckrough> you dont report them in the zenoss forums :)

[23-Mar-2010 08:54:40] <pokui> ckrough: meant the spam was comming via pms in the forums

[23-Mar-2010 08:54:49] <pokui> not the reports :D

[23-Mar-2010 08:54:53] <ckrough> best bet is to just /ignore them

[23-Mar-2010 08:55:04] <pokui> heh. cool.

[23-Mar-2010 08:55:10] <ckrough> maybe the freenode website has something... but irc is basically lawless

[23-Mar-2010 08:55:39] <pokui> nope. not talking about irc. talking about the community.zenoss.org site, but i'll see if there are filters there as well

[23-Mar-2010 08:55:48] <ckrough> ahhh

[23-Mar-2010 08:56:01] <ckrough> talk with mrayzenoss when he is on

[23-Mar-2010 08:56:16] <pokui> ckrough: thanks.

[23-Mar-2010 08:56:21] <ckrough> np

[23-Mar-2010 08:58:14] <rmatte> wow, they really messed up on the way interfaces are handled when remodelling on 2.5

[23-Mar-2010 08:58:38] <ckrough> rmatte: do tell, Im getting ready to go 2.5.2 in production, dont want any surprises

[23-Mar-2010 08:59:18] <rmatte> well first off, there's no O/A anymore, they've been replaced with "Status", which is fine...

[23-Mar-2010 08:59:42] <rmatte> the problem is that it automatically sets any interface that's administratively up to monitored when modelling, regardless of the operational status

[23-Mar-2010 09:00:14] <rmatte> which is a backwards way of doing it

[23-Mar-2010 09:00:17] <ckrough> what do you mean "monitored"? as in, it creates an even if it flaps or goes down?

[23-Mar-2010 09:00:27] <rmatte> correct

[23-Mar-2010 09:00:35] <rmatte> and since it's already technically down, you get events

[23-Mar-2010 09:00:39] <ckrough> I remember us discussing this in IRC with the devs

[23-Mar-2010 09:00:45] <rmatte> I've doing active interface status polling though, most people rely on traps

[23-Mar-2010 09:00:47] <ckrough> I think chet had a workaround for larger installs

[23-Mar-2010 09:00:53] <rmatte> yeh, I had a long discussion of this with cluther

[23-Mar-2010 09:01:14] <rmatte> yeh, not sure if the changes that Chet and I discussed even got implemented, doesn't appear that they did

[23-Mar-2010 09:01:35] <rmatte> hopefully it'll be done properly in stone crab, I've got something that I can modify to make it a bit easier on myself with 2.5 in the meantime

[23-Mar-2010 09:01:47] <ckrough> He's working with me on the upgrades, I'll be sure to ask before we move to 2.5x

[23-Mar-2010 09:01:51] <rmatte> need to move my operational status threshold to a different event class

[23-Mar-2010 09:02:00] <rmatte> /Status/IpInterface

[23-Mar-2010 09:02:09] <ckrough> isnt that where it already lives?

[23-Mar-2010 09:02:12] <rmatte> that should automatically update the status icon on interfaces that are down

[23-Mar-2010 09:02:21] <rmatte> no, the original location was /Net/Link

[23-Mar-2010 09:02:24] <ckrough> ah

[23-Mar-2010 09:02:36] <rmatte> and it's not a default thing either...

[23-Mar-2010 09:02:38] <ckrough> ...they all sound familiar at this point

[23-Mar-2010 09:02:46] <rmatte> docs/DOC-2494

[23-Mar-2010 09:02:50] <rmatte> that's what I'm using

[23-Mar-2010 09:03:03] <rmatte> I'll have to test it out then modify the guide to correspond to the new status handling

[23-Mar-2010 09:05:40] <ckrough> yeah... I'll want to disable updown alerting for pretty much all of them. Dont want an alert every time ops moves a server

[23-Mar-2010 09:05:49] <rmatte> the other noticeable thing with 2.5.2 is that it's running significantly slower than 2.4.5 on a copy of the same VM image

[23-Mar-2010 09:06:09] <rmatte> whereas there was supposed to be improved performance

[23-Mar-2010 09:06:23] <ckrough> support warned me that data.fs will be much larger in the later version, not sure which version starts the change as Im still 2.3.3

[23-Mar-2010 09:06:38] <ckrough> not sure if that would be the problem with the vm

[23-Mar-2010 09:06:45] <rmatte> possibly

[23-Mar-2010 09:06:53] <ckrough> supposedly zenhub is faster and more efficient though

[23-Mar-2010 09:07:01] <rmatte> the new zope version is in 2.5

[23-Mar-2010 09:07:02] <ckrough> what hypervisor do you run? vmwarE?

[23-Mar-2010 09:07:10] <rmatte> ESX 4.0

[23-Mar-2010 09:07:29] <rmatte> 3 of them, with a vSphere server to manage them

[23-Mar-2010 09:07:34] <venturaville> we found out belatedly with our zenhub that when you configure it for workers that it was using local disk on /tmp for FIFOs...

[23-Mar-2010 09:07:49] <venturaville> we ended up using tmpfs to mount on /tmp

[23-Mar-2010 09:07:56] <rmatte> ah

[23-Mar-2010 09:08:18] <ckrough> and most of the daemons seemt to ignore the --pcachedir argument

[23-Mar-2010 09:09:12] <ckrough> can I safely assume the FIFOs stay small

[23-Mar-2010 09:09:32] <rmatte> not sure

[23-Mar-2010 09:10:16] <venturaville> they are small, unfortunately they get written and read from quite a lot

[23-Mar-2010 09:13:11] <rmatte> I'm personally getting sick of the memory leak in Zenoss, guess I'll see how 2.5 performs in that regard

[23-Mar-2010 09:13:26] <ckrough> I've been using this for fstab in some setups

[23-Mar-2010 09:13:26] <ckrough> tmp                         /tmp        tmpfs rw,size=8G,mode=1777,fscontext=system_u:object_r:tmp_t:s0 0 0

[23-Mar-2010 09:13:29] <ckrough> vartmp                      /var/tmp    tmpfs rw,size=8G,mode=1777,fscontext=system_u:object_r:tmp_t:s0 0 0

[23-Mar-2010 09:22:34] <rmatte> mrayzenoss: good morning

[23-Mar-2010 09:23:36] <mrayzenoss> morning.  Just finished yammering for an hour in the Getting Started Webinar

[23-Mar-2010 09:23:50] <mrayzenoss> now I'm banning users and adding new spam filters

[23-Mar-2010 09:23:52] <mrayzenoss> woohoo

[23-Mar-2010 09:26:00] <rmatte> lol

[23-Mar-2010 09:26:01] <rmatte> nice

[23-Mar-2010 09:27:14] <rmatte> I'm messing around with my 2.5.2 test instance, so far I'm not liking the way automated interface monitoring is handled

[23-Mar-2010 09:27:33] <rmatte> everything else seems fine

[23-Mar-2010 09:32:29] <mrayzenoss> that sounds pretty good

[23-Mar-2010 09:33:01] <rmatte> overall, yes

[23-Mar-2010 09:33:28] <rmatte> it's running a bit slower than 2.4.5 as well

[23-Mar-2010 09:33:36] <rmatte> probably has to do with the new version of Zope

[23-Mar-2010 09:38:46] <rmatte> the interface status icons reflecting the actual status (after a bit of configuration) is definitely nice

[23-Mar-2010 09:39:29] <mrayzenoss> wasn't your 2.4.5 install fairly optimized?

[23-Mar-2010 09:39:55] <mrayzenoss> I just hadn't heard many people say 2.5 was slower

[23-Mar-2010 09:39:57] <rmatte> I'm comparing my 2.5.2 test image directly to the 2.4.5 image that it was cloned from

[23-Mar-2010 09:40:11] <rmatte> they are setup exactly the same, monitoring the exact same devices

[23-Mar-2010 09:40:18] <rmatte> the only difference is that one is running 2.5.2

[23-Mar-2010 09:40:34] <mrayzenoss> interesting

[23-Mar-2010 09:40:38] <rmatte> the 2.5.2 instance is noticeably slower, and the collector performance graphs look odd

[23-Mar-2010 09:40:57] <rmatte> the datapoint rate is shown as sort of ramps, rather than a constant count as it was in 2.4.5

[23-Mar-2010 09:41:37] <rmatte> I'm going to let it run for a few more hours and check back in to it, I just finished reconfiguring the interface monitoring

[23-Mar-2010 09:43:39] <rmatte> I just updated this docs/DOC-2494 so that it says to use /Status/IpInterface rather than /Net/Link

[23-Mar-2010 09:43:52] <rmatte> works like a charm in 2.5.2, can actually see current interface status.

[23-Mar-2010 09:44:13] <rmatte> our front line guys are going to like that

[23-Mar-2010 09:48:58] <rmatte> hmmm, looks like the 2.5.2 install is settling down/speeding up a bit

[23-Mar-2010 09:49:03] <rmatte> guess I'll wait and see if it continues

[23-Mar-2010 09:49:22] <mrayzenoss> as always, open tickets if you find anything concrete :)

[23-Mar-2010 09:49:40] <rmatte> I shall, I opened a ticket for that path issue

[23-Mar-2010 09:49:46] <rmatte> doesn't appear to have been looked at yet

[23-Mar-2010 09:50:02] <mrayzenoss> what's the number?

[23-Mar-2010 09:50:51] <rmatte> http://dev.zenoss.org/trac/ticket/6302

[23-Mar-2010 09:51:12] <mrayzenoss> it got assigned to Glenn

[23-Mar-2010 09:51:13] <rmatte> apparently they think it's the same as http://dev.zenoss.org/trac/ticket/6250

[23-Mar-2010 09:51:50] <mrayzenoss> well, 6250 definitely broke the .deb, but touched other bits

[23-Mar-2010 09:55:26] <rmatte> well, 602 has nothing to do with the .deb specifically

[23-Mar-2010 09:55:38] <rmatte> since I just use the plain jane stack installer

[23-Mar-2010 09:55:43] <rmatte> 6302 rather

[23-Mar-2010 09:56:11] <mrayzenoss> yeah, but the stuff that BitRock messed with in 6250 affected more than just the .deb, since the .deb is just a variant of the .bin

[23-Mar-2010 09:56:20] <mrayzenoss> anyway, Glenn's still tracking it

[23-Mar-2010 09:56:33] <rmatte> cool

[23-Mar-2010 09:56:42] <rmatte> I found workarounds for it anyways, so I'm good for now

[23-Mar-2010 09:56:47] <mrayzenoss> right now he's getting stack installers of trunk ready so we can switch to Python 2.6 and Zope 2.12

[23-Mar-2010 09:56:55] <rmatte> cool

[23-Mar-2010 09:56:55] <mrayzenoss> RPMs are built

[23-Mar-2010 09:57:16] <rmatte> does Glenn work for you guys or BitRock?

[23-Mar-2010 09:57:22] <mrayzenoss> us

[23-Mar-2010 09:57:24] <rmatte> ah

[23-Mar-2010 09:57:26] <mrayzenoss> he's our build engineer

[23-Mar-2010 09:57:29] <rmatte> cool

[23-Mar-2010 09:57:38] <rmatte> sounds fun

[23-Mar-2010 09:57:43] <mrayzenoss> he sends bugs to BitRock related to the stack

[23-Mar-2010 09:57:54] <mrayzenoss> they changed some stuff upstream from us and caused 6250

[23-Mar-2010 09:58:18] <rmatte> oh well, minor detail which will be fixed

[23-Mar-2010 09:58:26] <rmatte> the stack installer overall is a joy

[23-Mar-2010 10:01:36] <rmatte> hmmm, it's going to end up taking me about 6 hours to do the upgrade to 2.5.2 across all of my servers

[23-Mar-2010 10:01:43] <rmatte> guess I'll have to spread it out over 2 change windows

[23-Mar-2010 10:02:12] forsberg is now known as fOrsberg

[23-Mar-2010 10:02:22] <rmatte> then I have our ticketing system upgrade to orchestrate

[23-Mar-2010 10:19:26] <skipzoid> hello all

[23-Mar-2010 10:20:46] <skipzoid> no one here ?

[23-Mar-2010 10:21:19] <mrayzenoss> it's been kinda quiet this morning

[23-Mar-2010 10:22:39] <rmatte> yeh, it has

[23-Mar-2010 10:22:43] <rmatte> but yes, people are here

[23-Mar-2010 10:23:47] <mistich> good morning all

[23-Mar-2010 10:24:42] <mistich> anyone know if zenoss syslog daemon can Encrypting Syslog Traffic with TLS such as in RFC5424/5425

[23-Mar-2010 10:24:52] <skipzoid> Q: i'm playing with event translations - i'm tying to change evt.summary into someting human meaning full by checking a field and its corrosponding data as displayed on the event details tab: i have ccsAlarmParameter and its value is - (1, 3, 6, 1, 4, 1, 3142, 2, 7, 294, 2, 5) , my script reads -  err_val = getattr(evt, 'ccsAlarmParameter','Unknown') evt.summary = ' %s' % (err_val)  if err_val == '(1, 3, 6, 1, 4, 1, 3142, 2, 7, 2

[23-Mar-2010 10:25:34] <rmatte> mistich: no, it can't do encryption

[23-Mar-2010 10:25:42] <rmatte> mistich: it's just a plain jane syslog server

[23-Mar-2010 10:26:01] <skipzoid> but the evt.summary neither changes to the value or text ?

[23-Mar-2010 10:26:16] <mistich> I assumed so but haven't checked version 5.2 our yeat

[23-Mar-2010 10:26:45] <rmatte> skipzoid: it's (1, 3, 6, 1, 4, 1, 3142, 2, 7, 294, 2, 5) because it's a dictionary

[23-Mar-2010 10:27:16] <skipzoid> sorry - very new to this - not sure what you mean

[23-Mar-2010 10:27:45] <rmatte> it appears to be multiple values stored in a single variable, think of an araay

[23-Mar-2010 10:27:47] <rmatte> array*

[23-Mar-2010 10:28:02] <rmatte> so to get at the first value it would be ccsAlarmParameter[0]

[23-Mar-2010 10:28:09] <rmatte> I could be wrong, but that's what it looks like

[23-Mar-2010 10:28:15] <skipzoid> ah, not just a string stored in the varible ccsAlarmParameter ?

[23-Mar-2010 10:28:30] <rmatte> yeh, I doubt that it is

[23-Mar-2010 10:28:47] <rmatte> there's a way to compare the whole thing, I just don't remember what it is

[23-Mar-2010 10:28:54] <rmatte> look up examples of python dictionaries

[23-Mar-2010 10:30:19] <rmatte> you can see an example of what I mean by going in to zendmd

[23-Mar-2010 10:30:36] <rmatte> and typing time.localtime()

[23-Mar-2010 10:30:40] <rmatte> you'll see the same type of output

[23-Mar-2010 10:30:53] <rmatte> actually, import time

[23-Mar-2010 10:30:57] <rmatte> before you type that

[23-Mar-2010 10:31:05] <skipzoid> this is the event - http://picpaste.com/event1.bmp

[23-Mar-2010 10:31:08] <rmatte> then you can retrieve individual values with time.localtime()[0]

[23-Mar-2010 10:31:15] <rmatte> 1, 2, 3, etc...

[23-Mar-2010 10:31:45] <rmatte> yeh, looks like dictionaries

[23-Mar-2010 10:31:57] <rmatte> not sure what those values are meant to represent

[23-Mar-2010 10:32:43] <skipzoid> the kit is meant to be monitored by bespoke snmp software that translates those numbers into something meaning full, the manufcture wil not povide the MIB thoughs

[23-Mar-2010 10:32:47] <rmatte> cssAlarmLevel and AlarmSlot seem like they'd be more pertinent

[23-Mar-2010 10:33:21] <skipzoid> alarm level is like the severity

[23-Mar-2010 10:33:28] <rmatte> right

[23-Mar-2010 10:33:50] <skipzoid> the slot defines the frame complainting, its the parameter that defines the actual error within the frame

[23-Mar-2010 10:34:19] <rmatte> I see

[23-Mar-2010 10:35:17] <skipzoid> do you think i need someting like evt.ccsAlarmPrameter[0] etc. to read each value ?

[23-Mar-2010 10:36:35] <rmatte> it's possible

[23-Mar-2010 10:36:44] <rmatte> or you could try...

[23-Mar-2010 10:37:21] <rmatte> import string

[23-Mar-2010 10:37:41] <skipzoid> ?

[23-Mar-2010 10:37:42] <rmatte> I'm not done, hang on

[23-Mar-2010 10:37:45] <rmatte> :P

[23-Mar-2010 10:38:15] <rmatte> err_val = str(getattr(evt, 'ccsAlarmParameter','Unknown'))

[23-Mar-2010 10:38:18] <rmatte> see if that works

[23-Mar-2010 10:38:49] <rmatte> that will hopefully convert the output to a string when setting the variable

[23-Mar-2010 10:38:52] <skipzoid> lets try, jus a sec..

[23-Mar-2010 10:38:54] <rmatte> k

[23-Mar-2010 10:40:29] <ckrough> anyone know how to stop a script from within zendmd without exiting zendmd?

[23-Mar-2010 10:40:39] <rmatte> ctrl-c

[23-Mar-2010 10:40:40] <ckrough> ie: I want my script to sanity check and just stop if it fails

[23-Mar-2010 10:40:42] <ckrough> lol

[23-Mar-2010 10:40:47] <rmatte> :P

[23-Mar-2010 10:40:50] <ckrough> sorry, I mean how can I have the script kill itself

[23-Mar-2010 10:40:57] <ckrough> sys.exit() drops out of zendmd

[23-Mar-2010 10:40:57] <rmatte> import sys

[23-Mar-2010 10:41:01] <rmatte> oh

[23-Mar-2010 10:41:14] <ckrough> ya... I want to stop processign and fall back to the >>>

[23-Mar-2010 10:41:29] <rmatte> hmmm

[23-Mar-2010 10:42:06] <ckrough> meh, lots of threads on google with same unanswered question

[23-Mar-2010 10:42:19] <ckrough> Ill just deal with sys.exit

[23-Mar-2010 10:43:09] <skipzoid> rmatte !!!!!!!!!!! GENIUS !!!!!!!!

[23-Mar-2010 10:43:14] <rmatte> if you created a class within zendmd and ran it within that, it'd work

[23-Mar-2010 10:43:30] <Simon4> ckrough: maybe have script (if it's a loop) break on error and drop through to the end?

[23-Mar-2010 10:43:40] <rmatte> breaking is definitely an option

[23-Mar-2010 10:43:53] <ckrough> Simon4: its not a loop at that stage, but Im jsut going to do a big if/else

[23-Mar-2010 10:44:43] <skipzoid> rmatte : its working a treat now - thanks very much

[23-Mar-2010 10:44:48] <rmatte> np

[23-Mar-2010 10:44:59] <rmatte> you can convert variables in python quite easily

[23-Mar-2010 10:45:05] <rmatte> int() str() float() etc...

[23-Mar-2010 10:45:35] <skipzoid> i need to spend some more time on python - i'm a pearl person and its a bit different

[23-Mar-2010 10:46:10] <skipzoid> tea time now..

[23-Mar-2010 10:51:19] <rmatte> "If Python is executable pseudo code, perl is executable line noise"

[23-Mar-2010 10:51:21] <rmatte> ;)

[23-Mar-2010 10:51:38] <rmatte> my favorite quote

[23-Mar-2010 10:51:53] <Simon4> class :)

[23-Mar-2010 10:52:16] <Simon4> anyone in here used manage.addCollectionItem in anger?

[23-Mar-2010 10:52:22] <Simon4> (for multi-graph report wrangling)

[23-Mar-2010 10:52:53] <rmatte> can't say I have

[23-Mar-2010 10:53:32] * Simon4 has a bunch of reports where the component names are all about to change by a predictable amount, so scripting the change would be good

[23-Mar-2010 10:53:41] <Simon4> off to the source for a read, methinks :)

[23-Mar-2010 10:54:09] <rmatte> yop

[23-Mar-2010 11:02:26] <Brixius> well this is kindof fustrating, I just installed 2.5.2 on a fresh ubuntu 9.10 install using the debian stack installer.  Now I have to use 2 browsers to use the full functionality, Firefox doesn't display the events, and IE doesn't operate the collector plugins screen(there's no "Add Fields" to click on in IE)

[23-Mar-2010 11:03:15] <rmatte> Brixius: Those screens work perfectly fine in both browsers for me, clear your browser cache

[23-Mar-2010 11:03:45] <rmatte> I assume you used those browsers to view a different version of Zenoss in the past?

[23-Mar-2010 11:04:22] <Brixius> yes

[23-Mar-2010 11:08:03] <Brixius> well, the events now show in Firefox, but I never was able to get the Collector plugins to work in ie, the "Add Fields" has allway's been missing... It was one of the stumbling blocks I ran into learning Zenoss, I could never find the place to add them until I used firefox.

[23-Mar-2010 11:08:58] <mrayzenoss> Brixius: yeah, I've always disliked that UI section

[23-Mar-2010 11:09:02] <mrayzenoss> the "hidden panel"

[23-Mar-2010 11:09:14] <mrayzenoss> as we like to say in the defect reviews… "fixed in new UI"

[23-Mar-2010 11:09:18] <rmatte> Brixius: If you're using IE 8, enable compatibility mode

[23-Mar-2010 11:09:43] <rmatte> IE is garbage in my opinion anyways

[23-Mar-2010 11:09:47] <rmatte> FireFox is the way to go

[23-Mar-2010 11:10:23] <Brixius> That's my thought, couldn't it just be Enabled Plugins, and Disabled plugins, then you highlight them and have enable and disable buttons to move them to the other panel..

[23-Mar-2010 11:10:59] <rmatte> probably, the whole UI is being completely revamped in the next version

[23-Mar-2010 11:11:04] <rmatte> so it's not going to be an issue

[23-Mar-2010 11:11:29] <Brixius> ok, guessing that's going to be like a 3.x release then?

[23-Mar-2010 11:11:35] <rmatte> 2.6

[23-Mar-2010 11:11:41] <Brixius> ok

[23-Mar-2010 11:11:46] <rmatte> unless they decide to release it as 3.x

[23-Mar-2010 11:12:03] <rmatte> either way, it's coming in 6 months ish

[23-Mar-2010 11:13:02] <Brixius> ok, I was just going to ask on an eta.  Wasn't sure if one would be given, I know many people doen't like to speculate as many timelines are subject to slippage.

[23-Mar-2010 11:13:28] <rmatte> Brixius: yeh well, 2.5.2 was basically just released, so now they can focus on 2.6 now that they have a stable 2.5 release

[23-Mar-2010 11:13:33] <rmatte> http://public-demo.zenoss.com/zport/dmd/itinfrastructure?submitted=true

[23-Mar-2010 11:13:37] <rmatte> login with admin/zenoss

[23-Mar-2010 11:13:48] <rmatte> that's a demo of the new UI (partial demo)

[23-Mar-2010 11:17:12] <Brixius> it's different thats for sure

[23-Mar-2010 11:17:33] <rmatte> yup

[23-Mar-2010 11:18:19] <Brixius> that'll take a bit to find all the old stuff, but I think it;ll be nicer.

[23-Mar-2010 11:18:38] <rmatte> yeh, it's going to take some getting used to for sure

[23-Mar-2010 11:20:39] <mistich> anyone know in zendmd how to list all the thresholds set in a template

[23-Mar-2010 11:20:40] <mistich> I get all the templates like so

[23-Mar-2010 11:20:40] <mistich> for template in dmd.Devices.getAllRRDTemplates():

[23-Mar-2010 11:20:40] <mistich>    print template

[23-Mar-2010 11:21:23] <mistich> I need to find out what the names are for each threshold and what the severity is set to

[23-Mar-2010 11:22:11] <rmatte> well...

[23-Mar-2010 11:22:22] <rmatte> I know how to list actual properties for them

[23-Mar-2010 11:22:24] <rmatte> such as...

[23-Mar-2010 11:22:35] <rmatte> for f in d.os.filesystems():

[23-Mar-2010 11:22:38] <rmatte>  maxval = f.getRRDTemplateByName('FileSystem').thresholds.findObjectsById('high disk usage')[0].maxval

[23-Mar-2010 11:22:39] <rmatte>  print '%s        %s' % (f.id, maxval)

[23-Mar-2010 11:23:15] <rmatte> so I'd say you could do...

[23-Mar-2010 11:23:25] <mistich> let me play with that

[23-Mar-2010 11:23:29] <mistich> thansk

[23-Mar-2010 11:23:55] <rmatte> for t in f.getAllRRDTemplates().thresholds:

[23-Mar-2010 11:23:58] <rmatte>  print t.id

[23-Mar-2010 11:24:00] <rmatte> or something

[23-Mar-2010 11:24:08] <rmatte> just guessing

[23-Mar-2010 11:24:09] <mistich> yes I'm there

[23-Mar-2010 11:24:14] <rmatte> np

[23-Mar-2010 11:54:50] <venturaville> has anyone else noticed how slow zenhub is under load?

[23-Mar-2010 11:55:17] <rmatte> depends on your settings

[23-Mar-2010 11:55:19] <venturaville> it acts like a bottleneck for everything

[23-Mar-2010 11:55:35] <rmatte> well, that would make sense since it handles everything

[23-Mar-2010 11:55:37] <venturaville> 6 workers

[23-Mar-2010 11:55:45] <rmatte> how many processors/cores?

[23-Mar-2010 11:55:59] <venturaville> to the point of waiting a minute or more for datapoint changes to propogate...and everything comes to a screeching halt while it does

[23-Mar-2010 11:56:07] <venturaville> 4 cores

[23-Mar-2010 11:56:13] <rmatte> I'd drop your workers to 3 then

[23-Mar-2010 11:56:23] <rmatte> no point in having 6 with 4 cores

[23-Mar-2010 11:56:41] <rmatte> what are your zserver-threads set at for Zope?

[23-Mar-2010 11:56:43] <rmatte> default?

[23-Mar-2010 11:57:16] <venturaville> default of 4

[23-Mar-2010 11:57:18] <rmatte> good

[23-Mar-2010 11:57:33] <rmatte> drop the workers, it'll probably help

[23-Mar-2010 11:58:37] <rmatte> one of the guys here saw what you've done with your ZenPack so far and commented that it's pretty cool

[23-Mar-2010 11:58:42] <rmatte> ;)

[23-Mar-2010 11:59:04] <venturaville> which one?

[23-Mar-2010 11:59:08] <rmatte> libvirt

[23-Mar-2010 11:59:15] <venturaville> yeah it is coming along

[23-Mar-2010 11:59:20] <rmatte> for sure

[23-Mar-2010 11:59:46] <venturaville> I'm a bit disappointed by the ESX driver support in stock libvirt, but everything open source has pretty nice support in it :-)

[23-Mar-2010 12:00:21] <rmatte> yeh

[23-Mar-2010 12:00:22] <venturaville> I got pools/volumes working yesterday, but haven't committed it back to the SVN yet

[23-Mar-2010 12:00:29] <rmatte> cool

[23-Mar-2010 12:00:38] <rmatte> any idea if you'll be able to actually get ESX working properly?

[23-Mar-2010 12:00:47] <venturaville> I'm at the mercy of libvirt

[23-Mar-2010 12:00:55] <rmatte> ah

[23-Mar-2010 12:01:00] <venturaville> I have access to the API docs though for vsphere 4.0 so we shall see..

[23-Mar-2010 12:01:08] <rmatte> cool

[23-Mar-2010 12:01:09] <venturaville> perhaps I can do some low level coding as well :-P

[23-Mar-2010 12:01:15] <rmatte> hehe

[23-Mar-2010 12:01:58] <venturaville> I have to do most of my coding outside of work for it, since I can't justify a lot of hours on it otherwise

[23-Mar-2010 12:02:03] <venturaville> so a bit slower than I would like

[23-Mar-2010 12:02:08] <rmatte> ah

[23-Mar-2010 12:02:16] <rmatte> I do pretty much all of my coding while at work

[23-Mar-2010 12:02:47] <rmatte> though I'm going to setup some development servers at home and start doing some work at home as well

[23-Mar-2010 12:03:43] <venturaville> I keep a zenoss-stack and a bunch of KVM guests on my laptop, then I can just fire it all up and do coding on the hour long bus ride to and from work

[23-Mar-2010 12:04:01] <rmatte> cool

[23-Mar-2010 12:04:22] <rmatte> I've got a rackmount server sitting on my desk at home with 4 1.6GHz processors in it

[23-Mar-2010 12:04:45] <rmatte> I'm getting a dedicated server to move all of the stuff off of that to soon, then I'm going to throw ESXi on it and setup some development VMs

[23-Mar-2010 12:07:28] <venturaville> argh ... so on a loaded down zenoss server, even with 3 workers.. if I make a couple of graph point edits, and then hit the perf tab on the only system bound to that template... it sits there for many minutes

[23-Mar-2010 12:08:29] <rmatte> I've never seen that behaviour before

[23-Mar-2010 12:08:51] <rmatte> you sure you're not just getting caught up in the middle of a busy polling cycle?

[23-Mar-2010 12:09:13] <venturaville> it doesn't matter when I do it

[23-Mar-2010 12:09:18] <venturaville> I can replicate it quite reliably

[23-Mar-2010 12:09:47] <venturaville> it acts as though everything in zenhub is being blocked while the updates are run

[23-Mar-2010 12:10:51] <rmatte> weird

[23-Mar-2010 12:11:48] <rmatte> do you have the HP ProCurve or Cisco MIBs ZenPack installed on that system?

[23-Mar-2010 12:12:00] <venturaville> the cisvo mibs yes

[23-Mar-2010 12:12:06] <venturaville> but then most everyone does

[23-Mar-2010 12:12:09] <rmatte> did you perform the step listed on the Zenpack page?

[23-Mar-2010 12:12:44] <venturaville> which step?

[23-Mar-2010 12:12:57] <rmatte> the step that the ZenPack page says to perform in zendmd

[23-Mar-2010 12:13:21] <rmatte> dmd.ZenPackManager.packs._delObject('ZenPacks.Nova.Cisco.Mibs')

[23-Mar-2010 12:13:23] <rmatte> commit()

[23-Mar-2010 12:13:45] <venturaville> and what does that do?

[23-Mar-2010 12:14:22] <rmatte> the pack has a ton of objects, and there's some bug in Zenoss which causes it to scan through zenpack objects when certain pages are loaded and certain actions are performed which makes the UI sluggish as all hell

[23-Mar-2010 12:14:47] <rmatte> performing that step removes the ZenPack from the list while leaving the MIBs installed

[23-Mar-2010 12:14:55] <rmatte> which eliminates the performance issues

[23-Mar-2010 12:15:42] <rmatte> It took me quite a while to figure out that fix

[23-Mar-2010 12:16:00] <rmatte> The performance issues were unbearable until I did

[23-Mar-2010 12:16:12] <rmatte> 15 minutes to load zProperties pages sometimes for example

[23-Mar-2010 12:16:30] <rmatte> performance issues with anything to do with editing templates as well

[23-Mar-2010 12:19:12] <venturaville> apparently we don't have that one installed on this particular zenoss server

[23-Mar-2010 12:19:23] <rmatte> ah, then that's not the issue

[23-Mar-2010 12:19:33] <rmatte> what packs do you have installed on there?

[23-Mar-2010 12:19:36] <venturaville> interestingly it is almost precisely a minute time lag added per data point updated

[23-Mar-2010 12:19:40] <venturaville> dozens

[23-Mar-2010 12:19:50] <venturaville> we are a very heterogenous environment........

[23-Mar-2010 12:19:55] <rmatte> zenpack --list > fpaste.org

[23-Mar-2010 12:20:49] <rmatte> man, what the heck did they do to the Data Point Rate graph in 2.5.2?

[23-Mar-2010 12:20:58] <rmatte> instead of nice steady data I get wedges

[23-Mar-2010 12:21:32] <rmatte> that bug I was talking about also affects installs with large numbers of ZenPacks installed

[23-Mar-2010 12:21:48] <rmatte> The fact that you have 72 installed is most likely a contributing factor

[23-Mar-2010 12:22:35] <venturaville> got a trac ticket # for that bug?

[23-Mar-2010 12:22:42] <rmatte> lol, IRCd monitor, you guys really did install everything and anything almost ;)

[23-Mar-2010 12:22:57] <rmatte> let me dig it up (I filed it)

[23-Mar-2010 12:23:04] <venturaville> and we are using nearly all of them heavily

[23-Mar-2010 12:23:29] <rmatte> ah

[23-Mar-2010 12:23:49] <rmatte> http://dev.zenoss.org/trac/ticket/5071

[23-Mar-2010 12:23:56] <rmatte> that's the bug ticket I opened 10 months ago for that issue

[23-Mar-2010 12:24:31] <rmatte> The first comment by bbibeault is "Not limited to this zenpack; the more zenpacks and the more packables, the worse this problem becomes."

[23-Mar-2010 12:28:10] <venturaville> we have the HP proliant MIBs loaded

[23-Mar-2010 12:28:16] <venturaville> going to see if dropping some of those helps

[23-Mar-2010 12:29:25] <venturaville> geesh ... 199 mibs and 10s of thousands of nodes in the HP Proliant MIB stack :-/

[23-Mar-2010 12:33:42] <rmatte> you don't need to drop them

[23-Mar-2010 12:33:52] <rmatte> just do the same as what's listed on the Cisco page for the pack

[23-Mar-2010 12:34:03] <rmatte> the same steps are actually listed on the HP Proliant MIB page

[23-Mar-2010 12:34:34] <rmatte> it's not the fact that the objects are in Zope, it's the fact that the pack is indexed as a ZenPack

[23-Mar-2010 12:34:42] <rmatte> once Zenoss no longer sees it in the ZenPack list you're good

[23-Mar-2010 12:34:55] <rmatte> drop any other packs that you know you're not using, it'll help with performance

[23-Mar-2010 12:35:21] <rmatte> Hopefully they'll fix that bug at some point

[23-Mar-2010 12:36:26] <venturaville> it looks as though we have a lot of the HP MIBs, but they were installed singly instead of the big zenpack

[23-Mar-2010 12:36:29] <venturaville> still craziness

[23-Mar-2010 12:37:15] <rmatte> ah

[23-Mar-2010 12:37:28] <rmatte> so they were installed via zenmib you mean?

[23-Mar-2010 12:37:37] <venturaville> that is what I think yes

[23-Mar-2010 12:37:52] <rmatte> ah

[23-Mar-2010 12:38:14] <rmatte> but yeh, I'm pretty sure that the number of ZenPacks that you have installed is causing the response issue

[23-Mar-2010 12:49:28] <mistich> rmatte how do I set a Property on a threshold in zendmd

[23-Mar-2010 12:51:17] <mistich> nevermind

[23-Mar-2010 12:52:23] <rmatte> ;)

[23-Mar-2010 12:53:46] <rmatte> I like people who answer their own questions

[23-Mar-2010 12:57:46] <mistich> me too

[23-Mar-2010 13:11:30] <cain23> hello

[23-Mar-2010 13:13:56] <cain23> can you set alerting rules on the devices themselves?

[23-Mar-2010 13:14:32] <rmatte> no, but you can set filters on the alerting rules to limit them to certain devices

[23-Mar-2010 13:18:36] <cain23> yeah i know that...

[23-Mar-2010 13:20:04] <cain23> but can you have the alerting rules limit by the eventGroup id?

[23-Mar-2010 13:23:57] <citrus2> so a couple weeks ago i got zenoss ping monitoring all my servers in 3 different datacenters.     so the internet connection goes out at 3am this morning here where the zenoss box is and it spams my email with every single server is down every 5 minutes for 50 servers,     how do i make a rule to avoid this and just let me know my internet connection at the office went down

[23-Mar-2010 13:24:05] <rmatte> you can have them filtered by absolutely anything cain23

[23-Mar-2010 13:24:19] <rmatte> create an alert and check the filters

[23-Mar-2010 13:25:11] <cain23> i have, but not seen eventgroup id

[23-Mar-2010 13:25:13] <rmatte> citrus2: that's one of the tricky things with Zenoss, it only supports layer3 topology, and that's useless if it doesn't completely understand the topology (which is basically impossible in most cases)

[23-Mar-2010 13:25:32] <cain23> Looking to just filter by application event log errors....

[23-Mar-2010 13:25:34] <rmatte> cain23: "group"

[23-Mar-2010 13:25:57] <rmatte> you mean event group?

[23-Mar-2010 13:26:04] <rmatte> that's event class

[23-Mar-2010 13:26:22] <rmatte> it's all there, just read more carefully

[23-Mar-2010 13:26:23] <cain23> on the event its listed as eventGroup

[23-Mar-2010 13:26:49] <citrus2> rmatte,  what do you do in this case?  or others?

[23-Mar-2010 13:27:12] <rmatte> citrus2: nothing, we just put up with the flood of alerts in the event that that happens

[23-Mar-2010 13:27:12] <cain23> eventGroup  Application       ----thats striaght from my event viewer

[23-Mar-2010 13:28:00] <rmatte> I don't even see eventGroup populated in the majority of events, can't you use something else?

[23-Mar-2010 13:28:39] <rmatte> eventClass would be better

[23-Mar-2010 13:28:54] <cain23> event class is /Unknown

[23-Mar-2010 13:29:41] <rmatte> eventGroup is not filterable, so you're going to have to pick something else

[23-Mar-2010 13:30:08] <rmatte> are there other commonalities between the events?

[23-Mar-2010 13:30:24] <cain23> ha, I noticed...but people are requesting all application event log errors, or all system errors...

[23-Mar-2010 13:30:36] <cain23> for certain devices

[23-Mar-2010 13:31:02] <rmatte> What about Agent?

[23-Mar-2010 13:31:09] <citrus2> rmatte,   is their nothing conditional that i can set?    such as   alert on this rule unless a certain other event has occurred   that event being something like device, ISP down

[23-Mar-2010 13:32:05] <rmatte> citrus2: no, the filters don't work that way, you could try to do it with a transform but it wouldn't be easy and wouldn't be 100% effective

[23-Mar-2010 13:32:44] <rmatte> citrus2: they just really need to get layer2 topology and manual relations integrated in to Zenoss

[23-Mar-2010 13:34:19] <citrus2> sad

[23-Mar-2010 13:34:23] <rmatte> citrus2: some material for you... thread/8783

[23-Mar-2010 13:34:47] <cain23> rmatte: just as a side note, i've looked at 20 different events from the system and they all had something in the eventGroup field...

[23-Mar-2010 13:35:27] <rmatte> cain23: that's fine, but it's not a valid filter... "agent" is a valid filter, and it's always populated

[23-Mar-2010 13:35:32] <cain23> I'll just have to request that capability from them

[23-Mar-2010 13:35:41] <rmatte> so what are you seeing for "agent" ?

[23-Mar-2010 13:36:03] <cain23> zeneventlog

[23-Mar-2010 13:36:31] <rmatte> ok, and I'm guessing all of the events you're looking at are zeneventlog events

[23-Mar-2010 13:36:36] <rmatte> hence why the eventGroup is populated

[23-Mar-2010 13:36:47] <rmatte> it's populated based on the type of eventlog message

[23-Mar-2010 13:37:21] <jb> mrayzenoss: new squid zenpack is ready.. just waiting on some data to populate my graphs for screenshots.

[23-Mar-2010 13:37:25] <rmatte> you're sure there's absolutely nothing else that you could use to filter by?

[23-Mar-2010 13:37:32] <rmatte> jb: nice

[23-Mar-2010 13:37:47] <cain23> im sure, for these two people and this particualr request...

[23-Mar-2010 13:38:52] <cain23> "System Event Logs for any entries with an Error level"

[23-Mar-2010 13:40:24] <rmatte> ok, so what's so difficult about that?

[23-Mar-2010 13:40:52] <rmatte> in the filters... if Agent is zeneventlog and if Severity = Error

[23-Mar-2010 13:41:43] <cain23> that won't give back application events too?

[23-Mar-2010 13:41:46] <rmatte> probably want if State = Production as well

[23-Mar-2010 13:42:04] <rmatte> then you can add filters for device, device group, device class, whatever

[23-Mar-2010 13:42:21] <rmatte> there is no mention of application events in that request

[23-Mar-2010 13:42:38] <rmatte> oh wait, I see what you're saying

[23-Mar-2010 13:42:39] <cain23> no, not that particular one...i have other requests...lol

[23-Mar-2010 13:42:48] <rmatte> they only want to see stuff from the System section of the event log

[23-Mar-2010 13:42:57] <cain23> correct

[23-Mar-2010 13:43:07] <cain23> or application..depending on person and device

[23-Mar-2010 13:43:15] <rmatte> can you give me a screenshot of the details of one of those events?

[23-Mar-2010 13:43:30] <rmatte> (I'm trying to pull up some ancient ones but I haven't used event log monitoring in ages)

[23-Mar-2010 13:50:32] <rmatte> I wonder if I still have any laying around

[23-Mar-2010 13:51:23] <venturaville> weird ... any changes to a template block everything in the UI for almost exactly a minute

[23-Mar-2010 13:53:05] <straterra> Sweet.. I changed my zenoss server's IP..and now I get this error on that device

[23-Mar-2010 13:53:06] <straterra> 10.0.1.13 sendto error Host mims1.indy.marian.local and mims1.twtc.marianinc.com are both using ip 10.0.1.13

[23-Mar-2010 13:53:21] <rmatte> IP address conflict obviously

[23-Mar-2010 13:53:44] <straterra> its the same machine

[23-Mar-2010 13:53:56] <straterra> the machine has both of those hostnames

[23-Mar-2010 13:54:17] <rmatte> you're seeing that as an event?

[23-Mar-2010 13:54:24] <straterra> Yup

[23-Mar-2010 13:54:35] <rmatte> did you update the hosts file on that box after changing the IP?

[23-Mar-2010 13:54:41] <straterra> Yup

[23-Mar-2010 13:54:51] <rmatte> do you have that device in Zenoss at all?

[23-Mar-2010 13:55:02] <rmatte> like, is it monitoring itself?

[23-Mar-2010 13:55:02] <straterra> I do

[23-Mar-2010 13:55:06] <straterra> Yup

[23-Mar-2010 13:55:09] <rmatte> did you change it's IP in Zenoss?

[23-Mar-2010 13:55:23] <straterra> I did

[23-Mar-2010 13:55:32] <rmatte> do you have it added twice in Zenoss?

[23-Mar-2010 13:55:51] <rmatte> under different names but same IP?

[23-Mar-2010 13:56:08] <straterra> Nope

[23-Mar-2010 13:56:24] <rmatte> Is the hostname that you're using for it in Zenoss the same as the hostname in the hosts file?

[23-Mar-2010 13:56:43] <straterra> yes

[23-Mar-2010 13:57:08] <straterra> But, like I said earlier, the machine has two hostnames

[23-Mar-2010 13:57:15] <straterra> in the host file, that is

[23-Mar-2010 13:58:01] <rmatte> are those hostnames also available via dns?

[23-Mar-2010 13:58:17] <straterra> yes

[23-Mar-2010 13:58:24] <rmatte> then remove one of them from the hosts file

[23-Mar-2010 13:58:35] <rmatte> (the one that doesn't match the name in Zenoss)

[23-Mar-2010 13:59:12] <straterra> I have to go change mysql too..

[23-Mar-2010 13:59:28] <straterra> a warning/error about the host not being able to connect to this mysql server

[23-Mar-2010 13:59:33] <rmatte> basically this is what I think is happening...

[23-Mar-2010 13:59:53] <rmatte> Zenoss checks the hosts file to see the hostname of the server and takes the first hostname

[23-Mar-2010 14:00:03] <rmatte> and it uses that for monitoring itself

[23-Mar-2010 14:00:14] <straterra> ah..the alert went away

[23-Mar-2010 14:00:32] <rmatte> you honestly probably didn't even have to remove it, just switch the order around

[23-Mar-2010 14:00:45] <rmatte> the name that you added in to Zenoss should be first in the hosts file

[23-Mar-2010 14:01:13] <straterra> Ah, ok

[23-Mar-2010 14:01:15] <rmatte> so if you added hosts2 in to zenoss and in hosts it was showing as 127.0.0.1 hosts1 hosts2 you'd change it to look like 127.0.0.1 hosts2 hosts1

[23-Mar-2010 14:01:56] <rmatte> Zenoss uses the socket.getfqdn() function which grabs the first FQDN from the hosts file

[23-Mar-2010 14:02:28] <straterra> that makes sense then

[23-Mar-2010 14:02:49] <rmatte> If you go in to zendmd and type socket.getfqdn() you'll see what it's grabbing

[23-Mar-2010 14:14:39] <straterra> Do I have to restart zenoss for that change?

[23-Mar-2010 14:14:46] <rmatte> no

[23-Mar-2010 14:18:41] <SKULL2006> hi

[23-Mar-2010 14:19:06] <SKULL2006> how are you guys

[23-Mar-2010 14:19:08] <SKULL2006> ?

[23-Mar-2010 14:25:20] <rmatte> if he had stuck around for more than 2 minutes he would have found out

[23-Mar-2010 14:30:53] <rmatte> mrayzenoss: well, the view event history link in the event console doesn't appear to work properly

[23-Mar-2010 14:32:00] <rmatte> mrayzenoss: ah nevermind, it just takes a second to load up but doesn't have a loading indicator

[23-Mar-2010 14:36:11] <Simon4_> rmatte: was it venturaville who originally wrote the community HP bladechassis zenpack?

[23-Mar-2010 14:36:21] <rmatte> Simon4_: not sure

[23-Mar-2010 14:36:57] <rmatte> according to the ZenPack page, yes it was

[23-Mar-2010 14:37:05] <Simon4_> cool, cheers

[23-Mar-2010 14:37:20] * Simon4_ has completely rewritten the modeler, so will have a chat to him about such things

[23-Mar-2010 14:37:25] <rmatte> cool

[23-Mar-2010 14:37:44] <rmatte> he's been focussing on his libvirt pack lately

[23-Mar-2010 14:37:52] <Simon4_> yeah, that sounds most good

[23-Mar-2010 14:38:11] <rmatte> once you're done with that pack I can test it out, we have an HP blade chassis so that could be very useful for us

[23-Mar-2010 14:38:53] <Simon4_> I'm just removing some changes to get it to a "stable" state since I started some new features, so will have something installable (that I'm planning to test at work) tonight, and will let you know how that goes

[23-Mar-2010 14:39:06] <rmatte> cool

[23-Mar-2010 14:39:16] <Simon4_> once I've hurled it at 20 chassis I mgiht be confident enough to let others play ;)

[23-Mar-2010 14:39:26] <rmatte> hehe

[23-Mar-2010 14:40:12] * Simon4_ needs to find/read the contribution agreement thing too I think

[23-Mar-2010 14:41:04] <Simon4_> not sure if that applies for zenpacks as well as core patches - mrayzenoss will know I'm sure

[23-Mar-2010 14:42:39] <rmatte> yeh, it applies to all that stuff, but you really only have to read it once

[23-Mar-2010 14:42:48] <Simon4_> :)

[23-Mar-2010 14:42:51] <rmatte> they just want to make sure that you're aware of it

[23-Mar-2010 14:45:14] <Simon4_> hmm, wonder if I've broken the here/snmpindex value since I'm modelling via ssh

[23-Mar-2010 14:45:16] * Simon4_ has a looksee

[23-Mar-2010 14:46:17] <Simon4_> >>> fred.bladeservers()[0].snmpindex

[23-Mar-2010 14:46:17] <Simon4_> -1

[23-Mar-2010 14:46:19] <Simon4_> ruh oh

[23-Mar-2010 14:46:41] <Simon4_> seems I have a fun job for this evening then

[23-Mar-2010 14:47:06] <rmatte> lol

[23-Mar-2010 14:47:34] <Simon4_> hopefully it's related to the blade position, which will then be easy

[23-Mar-2010 14:47:45] <rmatte> probably

[23-Mar-2010 14:48:03] <rmatte> HP are usually relatively logical with their stuff

[23-Mar-2010 14:48:05] <Simon4_> yeah

[23-Mar-2010 14:48:18] <Simon4_> should be a snmp table with indexing based on either bladeNum or N+bladeNum

[23-Mar-2010 14:48:25] <rmatte> yeh

[23-Mar-2010 14:48:27] <Simon4_> can check against the orig zenpack in the offic

[23-Mar-2010 14:48:28] <Simon4_> +e

[23-Mar-2010 14:51:27] <rmatte> rhettardo: aloha

[23-Mar-2010 14:51:34] <rhettardo> hello

[23-Mar-2010 14:51:36] <rhettardo> how are you

[23-Mar-2010 14:51:40] <rmatte> good, you?

[23-Mar-2010 14:51:55] <rhettardo> pretty good

[23-Mar-2010 14:52:03] <rhettardo> just editing some setup scripts

[23-Mar-2010 14:52:07] <rmatte> cool

[23-Mar-2010 14:52:35] <rhettardo> still havnt figured out what the deal is w/ my two hp switches and the abnormal data

[23-Mar-2010 14:52:58] <rmatte> oh, on the interfaces?

[23-Mar-2010 14:53:03] <rhettardo> didnt do any more investigation after i changed the template to use the 32bit counter oid's

[23-Mar-2010 14:53:04] <rmatte> yeh, that's messed up

[23-Mar-2010 14:53:05] <rhettardo> yeah

[23-Mar-2010 14:53:24] <rhettardo> but its screwed up with either the 64 or 32 bit counters

[23-Mar-2010 14:53:26] <rmatte> actually, what were you seeing again? negative values?

[23-Mar-2010 14:53:47] <rhettardo> no, abnormally high and inconsistent numbers

[23-Mar-2010 14:53:50] <rmatte> ah

[23-Mar-2010 14:54:03] <rmatte> I had weird issues on 1 64bit interface the other day, but just 1

[23-Mar-2010 14:54:10] <rhettardo> i compared the rrd's from my cacti box to the rrd's for zenoss and thers a large difference

[23-Mar-2010 14:54:30] <rmatte> did you post in the forum about it?

[23-Mar-2010 14:54:51] <rhettardo> for certains times everything looks fine and then the decimal point will be moved way off at other times

[23-Mar-2010 14:54:53] <rhettardo> nah

[23-Mar-2010 14:55:02] <rhettardo> that will be the last step

[23-Mar-2010 14:55:05] <rmatte> you should post, and include screenshots and whatever else is pertinent

[23-Mar-2010 14:55:09] <rmatte> someone will probably know

[23-Mar-2010 14:55:19] <rhettardo> yeah

[23-Mar-2010 14:55:26] <rhettardo> i'll get to that soon

[23-Mar-2010 14:55:35] <rmatte> cool

[23-Mar-2010 14:55:56] <rhettardo> i did recently update to 2.5.2 and found out that the xenhost template dosnt work w/ xenserver

[23-Mar-2010 14:56:20] <rhettardo> not a big deal but sucks

[23-Mar-2010 14:57:29] <rmatte> yeh

[23-Mar-2010 15:00:28] <Simon4_> schweet, blade position = snmpindex

[23-Mar-2010 15:00:30] <Simon4_> I'm saved :)

[23-Mar-2010 15:03:56] <rmatte> nice

[23-Mar-2010 15:10:29] <rmatte> anyone know off the top of their head how to go about changing the summarization for RRD files?

[23-Mar-2010 15:10:39] <rmatte> like, if I wanted it to take 30 days before data is summarized

[23-Mar-2010 15:14:30] willwh|afk is now known as willwh

[23-Mar-2010 15:15:43] <Simon4_> rmatte: it's in the rrd create command

[23-Mar-2010 15:16:54] <rmatte> yeh, I'm digging through the documentation right now

[23-Mar-2010 15:17:15] <rmatte> was just hoping someone may have been adept enough to say: "you do this: <example>"

[23-Mar-2010 15:17:17] <Simon4_> I did it

[23-Mar-2010 15:17:21] <Simon4_> hang on

[23-Mar-2010 15:17:32] <rmatte> k

[23-Mar-2010 15:18:09] <Simon4_> go to collectors/edit

[23-Mar-2010 15:18:13] <rmatte> k

[23-Mar-2010 15:18:18] <rmatte> already there

[23-Mar-2010 15:20:44] * Simon4_ is just calculating what's there to work out what he made it :)

[23-Mar-2010 15:20:52] <rmatte> k

[23-Mar-2010 15:20:57] <Simon4_> basically the last number is "number of rows"

[23-Mar-2010 15:21:16] <Simon4_> so if you increase that, you get more rows at that resolution before it starts consolidating from my understanding

[23-Mar-2010 15:21:25] <Simon4_> http://oss.oetiker.ch/rrdtool/doc/rrdcreate.en.html being the reference I used

[23-Mar-2010 15:21:38] <rmatte> well, has it worked ok for you in practice?

[23-Mar-2010 15:22:01] <Simon4_> yup, only thing is it doesn't affect current rrd files

[23-Mar-2010 15:22:35] <rmatte> yeh, that's fine

[23-Mar-2010 15:22:39] <rmatte> this is for a new client, clean slate

[23-Mar-2010 15:22:50] <rmatte> but I obviously want to get it right the first time ;)

[23-Mar-2010 15:23:19] <rmatte> they expected polling cycles of 1 minutes 30 seconds with RRD data going back 30 days before consolidation apparently

[23-Mar-2010 15:23:24] <rmatte> so I need to make this work (yay)

[23-Mar-2010 15:23:34] <Simon4_> so I doubled the first line to 1440, but you'll need more than that by the sounds

[23-Mar-2010 15:23:53] <rmatte> 600 to 1440?

[23-Mar-2010 15:24:00] <Simon4_> *cough*

[23-Mar-2010 15:24:02] <Simon4_> kinda doubled

[23-Mar-2010 15:24:03] <Simon4_> :)

[23-Mar-2010 15:24:10] <rmatte> Here are the lines I have

[23-Mar-2010 15:24:11] <rmatte> straterra:AVERAGE:0.5:1:600

[23-Mar-2010 15:24:11] <rmatte> straterra:AVERAGE:0.5:6:600

[23-Mar-2010 15:24:11] <rmatte> straterra:AVERAGE:0.5:24:600

[23-Mar-2010 15:24:11] <rmatte> straterra:AVERAGE:0.5:288:600

[23-Mar-2010 15:24:11] <rmatte> straterra:MAX:0.5:6:600

[23-Mar-2010 15:24:12] <rmatte> straterra:MAX:0.5:24:600

[23-Mar-2010 15:24:12] <rmatte> straterra:MAX:0.5:288:600

[23-Mar-2010 15:24:17] <rmatte> so do I need to add something?

[23-Mar-2010 15:24:49] <Simon4_> so the 600 in the top line is 600 rows of 5 min samples (default collection interval)

[23-Mar-2010 15:24:55] <Simon4_> which is about 2 days, give or take

[23-Mar-2010 15:25:02] <rmatte> I see

[23-Mar-2010 15:25:10] <Simon4_> we decided we wanted four days of full res, to cover long weekends and the like

[23-Mar-2010 15:25:26] <Simon4_> so 1440 = 5 days

[23-Mar-2010 15:25:30] <rmatte> so only the first line needs to be changed, everything else can be left?

[23-Mar-2010 15:25:32] <Simon4_> just ot be safe

[23-Mar-2010 15:26:08] <Simon4_> the first line is the "full res", the next line is the first lot of consolidation, the next line the next etc, so if they want 30 days at full res, they may want longer at the next resolution?

[23-Mar-2010 15:26:53] <Simon4_> if not, then you don't need to modify any other lines

[23-Mar-2010 15:27:05] <rmatte> so I need 8640

[23-Mar-2010 15:27:17] <rmatte> but the other problem is, that I need 1 minute 30 second polling cycles

[23-Mar-2010 15:27:22] <rmatte> so I need to adjust the frames

[23-Mar-2010 15:27:31] <rmatte> or columns, whatever they are called

[23-Mar-2010 15:27:33] <Simon4_> yeah, the step length needs to change

[23-Mar-2010 15:27:42] <rmatte> which is where?

[23-Mar-2010 15:28:04] <Simon4_> looks like it's set by zenoss elsewhere, or automatically based on the "snmp Performance Cycle Interval" on the same page

[23-Mar-2010 15:28:07] <Simon4_> that would be my guess

[23-Mar-2010 15:28:24] <Simon4_> if you create a dev, and change that, a rrdtool dump on a file will tell you if it's picked it up

[23-Mar-2010 15:28:26] <rmatte> ah ok cool

[23-Mar-2010 15:29:57] <rmatte> thanks

[23-Mar-2010 15:30:04] <Simon4_> will be interesting to see how rrdtool graph likes the larger files

[23-Mar-2010 15:30:15] <rmatte> yeh, it will indeed

[23-Mar-2010 15:30:28] <rmatte> I hope it doesn't crash and burn

[23-Mar-2010 15:31:04] <Simon4_> mirror rrd files to ramdisk, serve rrdgraph out of ram :)

[23-Mar-2010 15:31:32] <rmatte> don't have ramdisk

[23-Mar-2010 15:31:44] <rmatte> and don't want to waste RAM setting one up

[23-Mar-2010 15:31:51] <rmatte> lol

[23-Mar-2010 15:32:30] <Simon4_> I ahve a mad idea to have Data.fs in ram with a 5 min writeback to disk - seeded by someone else a while ago... keen to give it a go

[23-Mar-2010 15:32:47] <Simon4_> doing same for rrdfiles could be a laugh, that or at least make sure the linux filesystem cache keeps em around

[23-Mar-2010 15:33:23] <rmatte> right, but if the server fails you lose 5 minutes of data

[23-Mar-2010 15:33:49] <Simon4_> data.fs isn't a big deal, just config, and rrdfiles you possibly seed the other way using the filesystem cache

[23-Mar-2010 15:34:05] <Simon4_> thtat or rrdfiles on ssd would be equally good

[23-Mar-2010 15:34:27] <rmatte> true, but say you make a bunch of config changes, then the server dies, have fun going back and checking to make sure they are all still there

[23-Mar-2010 15:34:37] <rmatte> it's unlikely, but possible

[23-Mar-2010 15:34:37] <Simon4_> yeah, it's a risk for sure

[23-Mar-2010 15:34:58] <Simon4_> again, maybe SSD is the smarter option

[23-Mar-2010 15:35:09] <Simon4_> tis slowly getting more affordable

[23-Mar-2010 15:35:52] <nonsenso> doesn't reconnoiter do something similar with not averaging old data?

[23-Mar-2010 15:38:18] <Simon4_> rmatte: I'm up to 250-300,000 datapoints on a set of spindles so disk perf is starting to get fun

[23-Mar-2010 15:40:35] <rmatte> I can imagine

[23-Mar-2010 15:40:43] <rmatte> that's a lot of datapoints

[23-Mar-2010 15:41:16] <Simon4_> tenant applications with tenant defined datapoints via a custom zenpack

[23-Mar-2010 15:41:19] <Simon4_> give them rope.....

[23-Mar-2010 15:41:44] <rmatte> and zey hang

[23-Mar-2010 15:41:45] <Simon4_> some applications come up with >1000 kpi's they want measured... and take some convincing that maybe they really dont' care that much

[23-Mar-2010 15:41:56] <rmatte> hehe

[23-Mar-2010 15:45:32] <Simon4_> gah, vmware's NAT engine has timeouts :/

[23-Mar-2010 15:46:28] <Brixius> anyone have a good idea how to handle HP ILO interfaces in zenoss?

[23-Mar-2010 15:47:38] <Simon4_> Brixius: it's next on my list once I sort this BladeChassis stuff

[23-Mar-2010 15:47:55] <Simon4_> what are you wanting to get out of the ILO?

[23-Mar-2010 15:48:43] <Brixius> I'm not sure what the best way to handle them is as they send traps for the server they are in as well.  It'd be nice to have a way to associate them with the server they belong to so that traps recieved are tied correctly in zenoss

[23-Mar-2010 15:49:19] <Simon4_> true

[23-Mar-2010 15:50:04] <straterra> rmatte: hmm?

[23-Mar-2010 15:51:20] <rmatte> straterra: sorry, auto-complete

[23-Mar-2010 15:51:26] <straterra> ah, ok

[23-Mar-2010 15:51:41] <Brixius> hmm another problem, event rollup doesn't appear to be working in 2.5.2, I can goto individual devices and see events, but when I look at events under the orginizer there's nothing listed.

[23-Mar-2010 15:51:46] <rmatte> meant to paste: RRA:AVERAGE:0.5:1:600

[23-Mar-2010 15:51:48] <rmatte> like that

[23-Mar-2010 15:51:59] <rmatte> but my client auto-completed rra to straterra

[23-Mar-2010 15:52:00] <rmatte> lol

[23-Mar-2010 15:52:10] <straterra> well, that works

[23-Mar-2010 15:52:19] <straterra> You should use straterra graphs

[23-Mar-2010 15:52:47] <rmatte> hehe

[23-Mar-2010 15:53:52] <Simon4_> outsourcing

[23-Mar-2010 15:53:54] <Simon4_> love it :)

[23-Mar-2010 15:56:33] <rmatte> :P

[23-Mar-2010 16:10:30] <straterra> For a low, low price..My team of small asian children will hand draw all of your graphs in real time!

[23-Mar-2010 16:10:55] <rmatte> lol

[23-Mar-2010 16:12:07] <Simon4_> :D

[23-Mar-2010 16:21:26] * Simon4_ unleashes his zenpack mods on 10 blade chassis

[23-Mar-2010 16:21:42] <rmatte> woot

[23-Mar-2010 16:21:58] <rmatte> "Unleash, the horde"

[23-Mar-2010 16:22:04] <Simon4_> :D

[23-Mar-2010 16:22:14] <rmatte> *evil laugh*

[23-Mar-2010 16:22:29] <Simon4_> I need code for "this is the standby OA, doin't worry about there being no blades"

[23-Mar-2010 16:22:36] <Simon4_> for those who don't run OA shared IP mode

[23-Mar-2010 16:22:43] <rmatte> ah

[23-Mar-2010 16:22:56] <Simon4_> it doesn't die, it just returns nothing, and a nice message would be good

[23-Mar-2010 16:23:29] <Simon4_> will have to try snmp on a standby OA to see if it's similar or not

[23-Mar-2010 16:25:40] <Simon4_> I need to sort the bladeserverlist report too, since you can sort by a) position, or b) chassis, but not chassis then position

[23-Mar-2010 16:25:55] <Simon4_> which is kinda dumb, and I think a limitation of the table design

[23-Mar-2010 16:26:07] <rmatte> youp

[23-Mar-2010 16:26:19] <Simon4_> breaking it out by chassis seems sensible though

[23-Mar-2010 16:26:24] <Simon4_> so one table per blade chassis

[23-Mar-2010 16:27:39] <Simon4_> in better news, modelling of 10 chassis went perfectly, got all details into the db... will work on the other things this week :)

[23-Mar-2010 16:28:02] <rmatte> sweet

[23-Mar-2010 16:29:28] <Simon4_> did you ever use the "height/width/depth" values?

[23-Mar-2010 16:29:50] * Simon4_ is guessing they were added because the snmp data gave them, vs for any requrements reason

[23-Mar-2010 16:32:12] <QbY> Questions about a reboot--first, after rebooting my installation didn't poll until I went and started several of the daemons..  also, sveeral interfaces which I had disabled for monitoring were re-enabled..

[23-Mar-2010 16:33:39] <Simon4_> QbY: I've run into disabled monitoring things reenabling after a model, wihch happens automatically every 12 hours unless configured otherwise

[23-Mar-2010 16:33:51] <QbY> ah

[23-Mar-2010 16:33:59] <QbY> what's "Lock IPInterfaces"

[23-Mar-2010 16:34:01] <Simon4_> you might be able to lock such things after disabling them

[23-Mar-2010 16:34:02] <Simon4_> yeah, that

[23-Mar-2010 16:34:14] <Simon4_> I _believe_ it stops modeling messing with it again

[23-Mar-2010 16:34:21] <Simon4_> which is most likely what you're after

[23-Mar-2010 16:36:11] <QbY> ok

[23-Mar-2010 16:36:53] <QbY> also---bad bad bad..  i had an event earlier..  when it cleared--its deleted forever

[23-Mar-2010 16:36:56] <QbY> i can't have that

[23-Mar-2010 16:37:02] <QbY> i never want a event to delete

[23-Mar-2010 16:37:19] <Simon4_> there was a discussion about that in here a week or so back

[23-Mar-2010 16:37:27] <Simon4_> rmatte: do you remember that? suppressing clear events?

[23-Mar-2010 16:38:34] <Simon4_> QbY: from memory someone wrote an event transform to basically kill the clear events, which achieved such things

[23-Mar-2010 16:38:51] <QbY> i want it to clear--but don't want it to delete

[23-Mar-2010 16:38:55] <QbY> we need historical information sometimes

[23-Mar-2010 16:40:18] <Simon4_> hmmm, cleared events and other stuff still end up in the history table

[23-Mar-2010 16:40:32] <Simon4_> can't find where in the new events console however that you access the event history

[23-Mar-2010 16:41:39] <Simon4_> ah

[23-Mar-2010 16:41:45] <Simon4_> bottom left, click on "event history"

[23-Mar-2010 16:41:55] <Simon4_> will show you all the evnts, and the clear events

[23-Mar-2010 16:42:00] * Simon4_ was blind :)

[23-Mar-2010 16:42:46] <Simon4_> and under the "Event Manager" link on the left menu, you can set the timescales for aging to history, and how long history is kept for

[23-Mar-2010 17:05:27] <Brixius> everytime I run zenmib I get a bunch of warnings from like the following WARNING zen.ZenMib: Unable to parse information from /usr/local/zenoss/common/share/mibs/irtf/.index -- skipping

[23-Mar-2010 17:14:11] * Simon4_ attempts to get his head around report template

[23-Mar-2010 17:14:14] <Simon4_> +s

[23-Mar-2010 17:20:22] <Simon4_> failing to find any reports with more than one zentable, also... dammit will actually have to do some work

[23-Mar-2010 17:20:39] <QbY> i need to figure out how to tell zenoss to only email an event once a day--even if it happens on every poll..

[23-Mar-2010 17:42:48] <rmatte> QbY: events never get deleted (well, unless you have "Delete Historical Events Older Than" set in Event Manager), they just get moved to history

[23-Mar-2010 17:43:07] <QbY> ok i found it..

[23-Mar-2010 17:43:40] <rmatte> QbY: as far as only emailing an event once a day, you're going to have to get fairly crafty... you'll need to create an alerting rule without a repeat time for just that event...

[23-Mar-2010 17:43:56] <rmatte> then have an event transform that automatically moves the event to history at a certain time each day

[23-Mar-2010 17:44:11] <QbY> ok

[23-Mar-2010 17:44:13] <rmatte> after being moved to history, when it regenerates it'll alert

[23-Mar-2010 17:44:23] <QbY> gotcha

[23-Mar-2010 17:45:03] <rmatte> if the event auto-clears that method won't work

[23-Mar-2010 17:45:13] <rmatte> and if anyone moves it to history manually, it won't work

[23-Mar-2010 17:45:17] <rmatte> but it's the only way to do it

[23-Mar-2010 17:47:32] <rmatte> honestly, if you just don't enable a repeat time on the alert, and tell people to manually clear the event once they've resolved it (there's a clear link right in the alert email) that's really the intended way to handle it

[23-Mar-2010 17:48:05] <rmatte> the auto-clearing of it daily is just another option if need be

[23-Mar-2010 17:54:58] <QbY> rmatte: you know what sucks is..  it can clear on its own

[23-Mar-2010 17:55:20] <QbY> one poll it could be bad, next poll good, so...  i'm going to get an endless stream of them if the conditions are right

[23-Mar-2010 17:56:39] <Simon4_> QbY: you can stop alerts from flapping by specifying "count >2" or similar in alert rults

[23-Mar-2010 17:56:41] <Simon4_> rules

[23-Mar-2010 17:57:05] <Simon4_> so they still get stored/recorded, but emails don't go out unless there's more than one consecutive etc

[23-Mar-2010 17:57:18] <QbY> hrmm..  my life is beginning to become hll

[23-Mar-2010 17:57:20] <QbY> hell

[23-Mar-2010 17:57:30] <QbY> can i build all the alerting rules and then put them in a "zenpack"

[23-Mar-2010 17:57:33] <QbY> or template

[23-Mar-2010 17:57:42] <QbY> so that i don't have to recreate 10 rules for every group

[23-Mar-2010 17:59:12] <Simon4_> you could script it with zendmd

[23-Mar-2010 18:00:23] <MrMyke>  I am looking for some info / help in getting started with this tool.   Heck maybe even pay a person to walk me through getting it going.

[23-Mar-2010 18:01:55] <Simon4_> QbY: http://pastie.org/883836 - code to create a user and alerting rule for a user, you can then set rule.where() for whatever your standard rules are

[23-Mar-2010 18:02:48] <Simon4_> MrMyke: people in here can certainly help with answering questions (or try), and there's a bunch of docs/stuff in the forums on the zenoss site also

[23-Mar-2010 18:03:53] <MrMyke> Simon thank you for that  I have been working with the tool for about 6 hrs and I can't get it to montor even one server.

[23-Mar-2010 18:04:15] <MrMyke> i am really bad at this so far.

[23-Mar-2010 18:04:24] <Simon4_> hmmm, that's not a good start. What kind of server are you trying to monitr?

[23-Mar-2010 18:04:30] <Simon4_> gah, monitor

[23-Mar-2010 18:04:40] <QbY> i also botched by install somewhere using .com instead of .net for the server name--where do i correct that?

[23-Mar-2010 18:05:38] <MrMyke>  I have 10 server only  8 are windoes 2003 2 are BSD   i also have ESX hosts that i would like to monitor as well  then I what to monitor 2 websites and a VPN tunnel.

[23-Mar-2010 18:06:04] <Simon4_> QbY: for the zenoss server name?

[23-Mar-2010 18:06:31] <QbY> yes

[23-Mar-2010 18:06:46] <QbY> in the links its showing .com

[23-Mar-2010 18:07:23] <Simon4_> the links in alert emails?

[23-Mar-2010 18:08:21] <QbY> Simon4_: yes

[23-Mar-2010 18:08:37] <Simon4_> QbY: check out /opt/zenoss/etc/zenactions.conf for a line starting with "zopeurl"

[23-Mar-2010 18:09:14] <Simon4_> replace /opt/zenoss with wherever your $ZENHOME is

[23-Mar-2010 18:09:21] <QbY> not there

[23-Mar-2010 18:09:29] <QbY> nothing in that file

[23-Mar-2010 18:09:34] <Simon4_> try adding this:

[23-Mar-2010 18:09:50] <Simon4_> zopeurl<tab>http://host.name.of.zenoss.server

[23-Mar-2010 18:09:59] <Simon4_> then restart zenoss

[23-Mar-2010 18:10:08] <QbY> k

[23-Mar-2010 18:10:32] <Simon4_> zenactions, which sends out the emails looks in that file to see what the url of the zenoss server is

[23-Mar-2010 18:10:46] <Simon4_> otherwise it uses the hostname of the box

[23-Mar-2010 18:10:57] <QbY> yeah, i was assuming that's where it was, i just changed the hostname

[23-Mar-2010 18:15:40] <MrMyke> ok 1st prob i have is i can't seem to get the right key from goggle for the map.  I go to teh website past the host server name in   place a check in the  i have read and agree then click Generate API Key

[23-Mar-2010 18:15:47] <MrMyke> I get a bunch of code back

[23-Mar-2010 18:16:21] <MrMyke> wow okay I just fixed that I found the Key :)

[23-Mar-2010 18:17:56] <Simon4_> :)

[23-Mar-2010 18:18:26] <Simon4_> I have to head to bed (in the UK here), but I woudl start with the windows servers, which I believe use WMI if you add them under devices/server/windows or whatnot

[23-Mar-2010 18:18:28] <MrMyke> next i would like to remove any and all deviced i have added by hand so i can tell it to auto scan teh network   how would I go about that from teh device list ?

[23-Mar-2010 18:19:16] <Simon4_> click on "Device List" click on "select all" then click on the down arrow for teh menu, and select delete

[23-Mar-2010 18:19:22] <Simon4_> *gone*

[23-Mar-2010 18:20:31] <MrMyke> ha ha :)

[23-Mar-2010 18:20:51] <MrMyke>  that worked :)

[23-Mar-2010 18:21:31] <MrMyke> okay i would like to auto discover a whjole network for Windsoes server and give it a  ID and Password to tray loggin on

[23-Mar-2010 18:38:59] <MrMyke> ok i am doign a discovery

[23-Mar-2010 19:03:42] <dibbz> hi all, last night my postgres perf graphs stopped and I look in the zencommand log and I see "INFO zen.zencommand: Deleting command" .. i'm at a loss to understand why

[23-Mar-2010 19:06:25] <dibbz> i did add a graph with a lot of datapoints yesterday, could repeated failures to render this cause it?

[23-Mar-2010 19:27:31] <QbY> can anyone point me to a tutorial on useing the rendering engine to output the graphs to a directory (so my webserver can use them on a site)?

[23-Mar-2010 19:40:38] <dibbz> i saved my cookie and used wget

[23-Mar-2010 19:40:43] <dibbz> to save the graphs for reports

[23-Mar-2010 20:04:28] <cain22> hello all

[23-Mar-2010 20:04:54] <cain22> question, anyone else run into zenmigrate issues when they upgraded to 2.5.2?

[23-Mar-2010 21:00:35] <rhettardo> not me

[23-Mar-2010 21:16:44] <cain22> i did...it blows..

[24-Mar-2010 00:00:30] [disconnected at Wed Mar 24 00:00:30 2010]

[24-Mar-2010 00:00:30] [connected at Wed Mar 24 00:00:30 2010]

[24-Mar-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[24-Mar-2010 02:39:08] <jaco> Hey Guys, how do I install zenplugins for the tar.gz file?

[24-Mar-2010 02:48:48] <Troubadix09> morning all

[24-Mar-2010 04:09:00] fOrsberg is now known as forsberg

[24-Mar-2010 04:31:56] forsberg is now known as fOrsberg

[24-Mar-2010 04:35:31] fOrsberg is now known as forsberg

[24-Mar-2010 05:53:50] citrus_ is now known as citrus

[24-Mar-2010 06:55:59] Simon4_ is now known as Simon4

[24-Mar-2010 07:39:15] <NewBee> is there a webpage testing app that works with zenoss? i thought  i saw something but i can't find it.

[24-Mar-2010 07:39:32] <NewBee> It would record what you do on a page and use that as the test case.

[24-Mar-2010 07:48:40] <forsberg> NewBee there is a zenpack

[24-Mar-2010 07:49:43] <forsberg> docs/DOC-3757

[24-Mar-2010 07:49:50] <forsberg> its really nice for testing web pages

[24-Mar-2010 07:51:29] <NewBee> Thank you

[24-Mar-2010 07:52:26] <NewBee> humm guessing no Flash support?

[24-Mar-2010 08:13:24] <NewBee> http://wiki.alertfox.com/First_Steps Looks like they have a system that will let zenoss check Flash. Need Pro Version. uses imacros.

[24-Mar-2010 08:35:18] <ckrough> any rrd pro's on?

[24-Mar-2010 08:35:30] <Simon4> hmm, not pro, but what are you trying to do?

[24-Mar-2010 08:35:41] <ckrough> I just need a second set of eyes on this rrdcreate:

[24-Mar-2010 08:35:42] <ckrough> RRA:AVERAGE:0.5:1:8928

[24-Mar-2010 08:35:43] <ckrough> RRA:AVERAGE:0.5:12:8760

[24-Mar-2010 08:35:50] <ckrough> sorry, not that one... sec

[24-Mar-2010 08:36:00] <ckrough> RRA:AVERAGE:0.5:1:105120

[24-Mar-2010 08:36:01] <ckrough> RRA:AVERAGE:0.5:12:8760

[24-Mar-2010 08:36:17] <ckrough> that second line is redundant/pointless for this config right? since line 1 covers the same time period

[24-Mar-2010 08:39:13] <Simon4> I think you're right

[24-Mar-2010 08:39:26] <Simon4> just reading the rrd create page again to get my head around it

[24-Mar-2010 08:39:59] <Simon4> yeah, both those RRA's would cover the same timeframe

[24-Mar-2010 08:40:10] <ckrough> so the second would never be used in the graph, right?

[24-Mar-2010 08:40:19] <Simon4> yah

[24-Mar-2010 08:40:23] <Simon4> it'll go for the freshest data

[24-Mar-2010 08:40:26] <ckrough> cool, thats what I thought as well

[24-Mar-2010 08:40:42] <ckrough> need to remove that second line and get that data out of the existing rrds, just wasting space I think

[24-Mar-2010 08:53:25] <Simon4> yeah

[24-Mar-2010 08:54:47] <tlir> guys

[24-Mar-2010 08:55:14] <tlir> I'm evaluating a couple of projects and I'm looking for a specific functionality which I couldn't figure out if it's supported in zenoss or not

[24-Mar-2010 08:55:35] <tlir> is it possible to schedule "jobs" to run plugins/scripts for a specific host or a group of hosts?

[24-Mar-2010 08:56:05] <venturaville> technically yes, but there are other tools that are probably better used for that purpose

[24-Mar-2010 08:56:22] <ckrough> you may want puppet or something more along those lines

[24-Mar-2010 08:56:31] <venturaville> that was about to be my suggestion :-)

[24-Mar-2010 08:57:03] <tlir> I'm not looking for that option only, but ofcourse also to monitor hosts though the scheduling functionality is required

[24-Mar-2010 08:57:03] <ckrough> or http://www.capify.org/index.php/Capistrano

[24-Mar-2010 08:57:11] <ckrough> I havent used it, but  I think thats what its for

[24-Mar-2010 08:59:12] <ckrough> zenoss is very flexible if you know python

[24-Mar-2010 08:59:27] <venturaville> capistrano is more of a deployment engine, but it can usually play nicely with puppet anyways

[24-Mar-2010 09:00:13] <ckrough> Im just upgrading to the current version, but there is a newer daemon called zenjobs that may do what you want

[24-Mar-2010 09:30:28] <tlir> venturaville: you mentioned that zenoss has support for job scheduling. that's supported in the gui?

[24-Mar-2010 09:30:46] <venturaville> ckraought mentioned it not me

[24-Mar-2010 09:30:49] <venturaville> err ckrough

[24-Mar-2010 09:31:21] <tlir> (03:55:35 PM) tlir: is it possible to schedule "jobs"

[24-Mar-2010 09:31:27] <tlir> 03:56:05 PM) venturaville: technically yes

[24-Mar-2010 09:31:40] <tlir> you mentioned that zenoss supports such scheduling or am I wrong?

[24-Mar-2010 09:31:53] <venturaville> yes as in it is possible to do it ... (it has a full on programming API to do anything you want)

[24-Mar-2010 09:32:05] <tlir> ahh

[24-Mar-2010 09:32:10] <venturaville> and from what ckrough was saying there is a zenjobs setup that supports it (I don't use it personally)

[24-Mar-2010 09:32:21] <tlir> it doesn't have GUI screens natively in the zenoss install then?

[24-Mar-2010 09:32:24] <venturaville> might want to catch mrayzenoss when he gets on

[24-Mar-2010 09:32:32] <tlir> alright

[24-Mar-2010 09:32:33] <venturaville> he works for zenoss and probably can help you more than I can on that one

[24-Mar-2010 09:32:48] <tlir> there's just so much info, I'd just need some direction to a documentation on this topic

[24-Mar-2010 09:32:49] <tlir> thanks

[24-Mar-2010 09:32:59] <venturaville> zenoss.com has some decent online docs

[24-Mar-2010 09:56:20] <Simon4> tlir: it's under settings->Jobs tab, never used it though

[24-Mar-2010 09:56:40] <Simon4> seems I can't even add a job, so maybe it's a coming soon type thing

[24-Mar-2010 09:56:59] <ckrough> Simon4: what version are you running?

[24-Mar-2010 09:57:17] <ckrough> if its in the gui, it should be there

[24-Mar-2010 09:57:40] <Simon4> ckrough: 2.5.1.. maybe I just don't know how to add jobs?

[24-Mar-2010 09:57:54] <tlir> Simon4: uhh thanks

[24-Mar-2010 09:57:56] <Simon4> I get a list with "No records found" and no drop down menus etc

[24-Mar-2010 09:58:07] <tlir> Simon4: I'd appreciate if you could try mocking up some simple script to run

[24-Mar-2010 10:03:19] <Simon4> tlir: I'm learning as fast as you ;) docs/DOC-3201 was something I did just find

[24-Mar-2010 10:03:30] <Simon4> but it looks like you'll need some code to create your job of some sort

[24-Mar-2010 10:03:33] <tlir> :-)

[24-Mar-2010 10:04:11] <tlir> just a sec, reading that

[24-Mar-2010 10:07:10] <tlir> uhmm, it seems like an API that provides this functionality

[24-Mar-2010 10:07:32] <tlir> while that's probably great in flexibility it might mean that I need to develop my own screens to configure jobs and such and that might be a little bit out of the scope

[24-Mar-2010 10:08:28] <Simon4> yeah, that's the feeling I'm getting

[24-Mar-2010 10:50:26] <venturaville> okay .. a question for someone with some zenpack debugging abilities:

[24-Mar-2010 10:50:34] <venturaville> ERROR: zenpack command failed. Reason: AttributeError: 'libvirtHost' object has no attribute 'id'

[24-Mar-2010 10:50:54] <venturaville> (from libvirt in zenpacks subversion tree

[24-Mar-2010 10:51:10] <venturaville> it only happens the first time I up the version number and do an install

[24-Mar-2010 10:51:13] <venturaville> the second time it works

[24-Mar-2010 10:51:53] <venturaville> I know what it is saying, but it doesn't make sense for what is in the code

[24-Mar-2010 11:03:38] <Simon4> venturaville: only thing I would think is that there's something in the objects being added or in the install section that's referencing id before the object exists

[24-Mar-2010 11:04:09] <venturaville> the only thing I know if is the modeler

[24-Mar-2010 11:04:30] <venturaville> and it doesn't set it for the host, only the guests, pools, volumes

[24-Mar-2010 14:26:44] <tlir> guys

[24-Mar-2010 14:26:57] <tlir> docs/DOC-4737 -> shows how to list jobs

[24-Mar-2010 14:27:16] <tlir> is it also possible to define jobs (to run scripts on scheduled times) from the GUI too?

[24-Mar-2010 14:32:08] <rmatte> why wouldn't you just use cron?

[24-Mar-2010 14:32:17] <tlir> manually?

[24-Mar-2010 14:32:23] <rmatte> like, what exactly are you trying to do?

[24-Mar-2010 14:32:38] <tlir> perform ftp operations on a host/group of hosts

[24-Mar-2010 14:32:54] <tlir> like daily do ftp backup, etc

[24-Mar-2010 14:32:54] <rmatte> ok, are you wanting that to happen based on an event coming in?

[24-Mar-2010 14:32:58] <rmatte> or on a specific schedule?

[24-Mar-2010 14:33:10] <rmatte> Zenoss isn't really designed for that sort of thing

[24-Mar-2010 14:33:18] <rmatte> create a script and cron it

[24-Mar-2010 14:33:22] <rmatte> that's what cron is there for

[24-Mar-2010 14:33:23] <tlir> I'm not too familiar with the platform to understand the events system (which sounds interesting) though I'd like to do it on a specific schedule

[24-Mar-2010 14:33:30] <rmatte> it's really not complex to setup cron job

[24-Mar-2010 14:33:32] <rmatte> jobs*

[24-Mar-2010 14:33:33] <tlir> right but I can't just use cron for that

[24-Mar-2010 14:33:48] <rmatte> why?

[24-Mar-2010 14:34:24] <venturaville> puppet is your best bet

[24-Mar-2010 14:34:26] <venturaville> (really)

[24-Mar-2010 14:34:31] <venturaville> it can do cron job setups and all of that

[24-Mar-2010 14:34:40] <rmatte> yeh, puppet is definitely more along the lines of what you're looking for

[24-Mar-2010 14:34:58] <venturaville> zenoss == monitoring

[24-Mar-2010 14:35:05] <tlir> because we're not talking about writing a 5 lines script and adding a cronjob, it's a more complicated process - I need to build the web pages to allow to configure the scheduling, then read the zenoss database of hosts/group to act upon, etc... we're talking about a complete interface to manage that

[24-Mar-2010 14:35:09] <venturaville> it can/could do more, but it isn't the best tool for that

[24-Mar-2010 14:35:32] <venturaville> you can have something that queries zenoss and spits out a puppet config .........

[24-Mar-2010 14:35:45] <tlir> venturaville: you're pushing in puppet again when I already explained that I need both of these functions: 1. monitoring of hosts 2. running scheduled jobs

[24-Mar-2010 14:35:52] <rmatte> You just said you only need it to do the backup once a day, so how much are you really going to be modifying the schedule?

[24-Mar-2010 14:36:06] <tlir> I'm just wondering if I can get both of these functions with an app out of the box

[24-Mar-2010 14:36:30] <tlir> ofcourse I can also install both zenoss and puppet or something else and use 2 screens to configure the different things (one for monitoring and one for scheduling) but that's just ugly

[24-Mar-2010 14:36:32] <rmatte> I'm sure there's some expensive enterprise app out there that can do both

[24-Mar-2010 14:36:40] <rmatte> but not with any open source app that I've seen

[24-Mar-2010 14:36:46] <rmatte> they are completely different playing fields

[24-Mar-2010 14:36:46] <tlir> really?

[24-Mar-2010 14:36:50] <venturaville> yes

[24-Mar-2010 14:36:57] <tlir> I'm actually surprised

[24-Mar-2010 14:37:23] <rmatte> why would a monitoring tool handle ftp backups?

[24-Mar-2010 14:37:34] <tlir> the ftp backups is an example

[24-Mar-2010 14:37:39] <rmatte> I can see it maybe handling backups of cisco devices and such via telnet or ssh

[24-Mar-2010 14:37:41] <rmatte> or snmp

[24-Mar-2010 14:37:45] <rmatte> but not ftp backups

[24-Mar-2010 14:37:48] <tlir> the script would be doing other things as well (gather data, etc)

[24-Mar-2010 14:38:05] <tlir> I found a post close to the subject here: message/10521#10521

[24-Mar-2010 14:38:10] <rmatte> so what are you doing, backing up device configs, or what?

[24-Mar-2010 14:38:34] <tlir> rmatte: currently pushing file configs to devices via ftp

[24-Mar-2010 14:38:43] <rmatte> I see

[24-Mar-2010 14:38:50] <venturaville> that page is talking about running the command graphing the return values

[24-Mar-2010 14:38:51] <tlir> rmatte: I am guessing I will need the opposite as well (getting the files - meaning backup or sort of)

[24-Mar-2010 14:39:05] <rmatte> tlir: you could definitely make a script that pulls device names from Zenoss, that's very easy to do

[24-Mar-2010 14:39:08] <tlir> venturaville: I think you are too strict with your definitions

[24-Mar-2010 14:39:09] <rmatte> then use that list to do something

[24-Mar-2010 14:39:18] <tlir> venturaville: firstly, it talks about executing a command which is already half way

[24-Mar-2010 14:39:25] <rmatte> but as for actually scheduling the stuff from a UI, that's not possible out of the box with Zenoss

[24-Mar-2010 14:39:35] <venturaville> tlir: what rmatte said

[24-Mar-2010 14:39:41] <tlir> rmatte: you mean a standalone script that queries the database to pull off the hosts info ?

[24-Mar-2010 14:39:43] <rmatte> you might want to hunt around to see if there's some php system out there to assist in configuring cron jobs

[24-Mar-2010 14:40:11] <ckrough> just read back through... you really want puppet with templates :)

[24-Mar-2010 14:40:19] <tlir> rmatte: uhh... yeah I thought of using one thing for monitoring and the another for the cronjobs thing... just a bit ugly :)

[24-Mar-2010 14:40:21] <venturaville> tlir: yes you can use the zendmd scripts or the SOAP API to query the list of devices

[24-Mar-2010 14:40:21] <rmatte> tlir: yes, something that uses the Zenoss API to pull hostnames, ips, and whatever else you need

[24-Mar-2010 14:40:29] <rmatte> very simple to do in python, and can also be done in other languages

[24-Mar-2010 14:40:47] <ckrough> there is a nice JSON method for pulling device lists from zenoss

[24-Mar-2010 14:40:52] <tlir> alright

[24-Mar-2010 14:41:03] <tlir> I thought maybe I could catch both of these functions in one app :)

[24-Mar-2010 14:41:05] <rmatte> I don't personally call using 2 separate systems ugly, as long as they are properly integrated

[24-Mar-2010 14:41:19] <rmatte> we have tons of different pieces of software including Zenoss all integrated together and it works fine

[24-Mar-2010 14:41:28] <venturaville> same here

[24-Mar-2010 14:41:37] <ckrough> same

[24-Mar-2010 14:42:02] <rmatte> I would much rather integrate a bunch of systems that do what they are designed to do very well rather than get an out of the box system that doesn't do any one thing particularily well

[24-Mar-2010 14:42:20] <tlir> yeah ofcourse

[24-Mar-2010 14:42:32] <ckrough> though there is RANCID with zenoss, but thats really specific

[24-Mar-2010 14:42:43] <rmatte> yeh, RANCID is great for backing up device configs

[24-Mar-2010 14:42:47] <tlir> It's just that there will be extensive work required for the integration process, anyway, more than I expected since I was looking to have an app with both of these functions

[24-Mar-2010 14:43:12] <rmatte> tlir: you're not really much of a coder I take it?

[24-Mar-2010 14:43:13] <venturaville> once you get puppet working, all you need is a simple script that runs to query zenoss and then dump some configs out

[24-Mar-2010 14:43:21] <ckrough> ^^this

[24-Mar-2010 14:43:31] <venturaville> could be done in a handful of lines of scripting

[24-Mar-2010 14:43:34] <ckrough> that coupled with ERB templates and you may be there

[24-Mar-2010 14:43:34] <tlir> rmatte: lol I am, that's not the issue :)

[24-Mar-2010 14:43:43] <rmatte> k

[24-Mar-2010 14:43:58] <venturaville> ckrough: true true... I just started using the ERB ones over perl template toolkit today

[24-Mar-2010 14:44:06] <venturaville> not that bad...

[24-Mar-2010 14:44:09] <tlir> whats erb guys?

[24-Mar-2010 14:44:35] <venturaville> you can template files from within puppet

[24-Mar-2010 14:45:13] <venturaville> http://projects.reductivelabs.com/projects/puppet/wiki/Puppet_Templating

[24-Mar-2010 14:45:14] <rmatte> Puppet supports templates and templating via ERB, which is part of the Ruby standard library and is used for many other projects including Ruby on Rails.

[24-Mar-2010 14:45:16] <venturaville> (happened to have it open)

[24-Mar-2010 14:45:50] <tlir> ok

[24-Mar-2010 14:45:53] <ckrough> tlir: and it can pull variables from the server its acting on, so the template can customize itself to the box its going to configure, great for scripts, configs, crons, etc...

[24-Mar-2010 14:47:35] <venturaville> for example:

[24-Mar-2010 14:47:39] <venturaville> root@ubuntu1:~# facter | grep -i model

[24-Mar-2010 14:47:39] <venturaville> hardwaremodel => x86_64

[24-Mar-2010 14:48:19] <venturaville> you'll never go back......

[24-Mar-2010 14:48:30] <tlir> :)

[24-Mar-2010 14:48:51] <tlir> I'm just not familiar with zenoss enough to see the big picture

[24-Mar-2010 14:48:59] <tlir> but I'll probably poke around if I decide to use it

[24-Mar-2010 14:49:26] <rmatte> zenoss is incredible for monitoring and simple to integrate with other stuff, so I'd definitely recommend taking a good look

[24-Mar-2010 14:52:54] <cgibbons> and for every device you monitor with zenoss, I *won't* kill a kitten

[24-Mar-2010 14:54:13] <ckrough> Ive saved a lot of kittens

[24-Mar-2010 14:54:57] <rmatte> me too

[24-Mar-2010 14:55:29] <ckrough> somewhere in Annapolis is a field of kittens

[24-Mar-2010 14:55:56] <rmatte> yeh, cgibbons' basement

[24-Mar-2010 14:56:09] <rmatte> (he has a big basement)

[24-Mar-2010 14:56:13] <cgibbons> i do like me some kitten soup

[24-Mar-2010 14:56:19] <rmatte> nom nom

[24-Mar-2010 15:18:52] <rhettardo> kitten pho

[24-Mar-2010 16:12:35] <mrayzenoss> rmatte: Thanks! blogs/zenossblog/2010/03/24/tip-of-the-month-time-sensitive-event-transforms

[24-Mar-2010 16:35:21] <rmatte> mrayzenoss: haha, cool

[24-Mar-2010 16:36:01] <rmatte> I've moved up from stalwart to fixture ;)

[24-Mar-2010 16:37:17] <Simon4> rmatte: nice one :)

[24-Mar-2010 16:37:19] <mrayzenoss> :p

[24-Mar-2010 16:37:37] <rmatte> tnx

[24-Mar-2010 16:37:47] <mrayzenoss> next time I'll hit the thesaurus up

[24-Mar-2010 16:37:52] <rmatte> hehe

[24-Mar-2010 16:38:05] <rmatte> well, I'm out for the day (sick, working from home, and done for the day)

[24-Mar-2010 16:38:10] <rmatte> talk to you guys later

[24-Mar-2010 16:38:12] <Simon4> seu

[24-Mar-2010 16:38:42] <mrayzenoss> I was just updating the HP Temperature Sensors

[24-Mar-2010 16:38:46] <mrayzenoss> thanks for that too

[24-Mar-2010 16:38:50] <rmatte> np

[24-Mar-2010 16:39:05] <Simon4> gah, keep missing venturaville

[24-Mar-2010 16:39:08] <rmatte> I'll do the juniper pack sometime this week (I might be taking a sick day tomorrow)

[24-Mar-2010 16:39:31] <rmatte> cheers

[24-Mar-2010 16:39:56] * Simon4 unleashes his blade chassis toy on the other half of the blade servers

[24-Mar-2010 16:40:31] <mrayzenoss> anyone using Cfengine?

[24-Mar-2010 16:40:47] <mrayzenoss> I'm talking to them about writing a monitoring ZenPack and then an integration one

[24-Mar-2010 16:42:15] <Simon4> mrayzenoss: we're a puppet shop, thanks for the idea to have zenoss monitor puppet status ;)

[24-Mar-2010 16:42:36] <mrayzenoss> Simon4: you see this? docs/DOC-5818

[24-Mar-2010 16:43:01] <Simon4> hmm, I may have ages ago, but it wasn't like that

[24-Mar-2010 16:43:03] <Simon4> thanks!

[24-Mar-2010 16:43:12] <mrayzenoss> yeah, that's pretty new

[24-Mar-2010 16:43:18] <mrayzenoss> from venturaville as well

[24-Mar-2010 16:43:25] <Simon4> nice

[24-Mar-2010 16:43:51] <mrayzenoss> I'm trying to get Cfengine and Chef so we have a well-rounded provisioning/monitoring integration story

[24-Mar-2010 16:43:52] <Simon4> I've hacked up his blade chassis zenpack, so need to chat to him about how to integrate changes nicely, since I kind of replaced the entire modeler :)

[24-Mar-2010 16:44:16] <Simon4> moved to ssh to the OA, vs snmp, which didn't give a lot of info

[24-Mar-2010 16:45:40] <Simon4> mrayzenoss: can you think of any custom reports you've seen that have more than one zentable? trying to create a blade chassis report with one table per chassis, showing all the blade in that chassis

[24-Mar-2010 16:45:50] <Simon4> I can't find any in the default install to copy from :/

[24-Mar-2010 16:47:23] <mrayzenoss> Simon4: community/zenpacks?view=tags&tags=reports are the ones I've seen

[24-Mar-2010 16:47:45] <Simon4> cheers - will have a looksee

[24-Mar-2010 16:47:56] <mrayzenoss> haven't written any myself

[24-Mar-2010 16:49:23] * Simon4 might have to write some python to do it yet

[24-Mar-2010 16:49:48] <Simon4> since it might be the only way to iterate over multiple tables via a device class

[24-Mar-2010 16:50:28] <mrayzenoss> if you see npmccallum in the channel, ping him :)

[24-Mar-2010 16:50:35] <Simon4> heh, cheers :)

[24-Mar-2010 17:08:33] <mrayzenoss> HP Temperature Sensors ZenPack updated: docs/DOC-3459 Thanks rmatte!

[24-Mar-2010 17:10:11] <Simon4> mrayzenoss: I'm being lazy, but got a link to the contributors agreement? guessing I should read/sort before contributing to community zenpacks?

[24-Mar-2010 17:10:49] <mrayzenoss> community/developers

[24-Mar-2010 17:10:57] <Simon4> cheers

[24-Mar-2010 17:11:01] <mrayzenoss> thanks

[24-Mar-2010 17:11:26] <mrayzenoss> just ping me after you fill that out and I'll hook you up with an SVN account for zenpacks.zenoss.org

[24-Mar-2010 17:11:32] <Simon4> ah, awesome

[24-Mar-2010 17:11:35] <Simon4> that was next question :)

[24-Mar-2010 17:11:46] <mrayzenoss> email me at mray@zenoss.com and I'll remember

[24-Mar-2010 17:12:15] <mrayzenoss> that also clears any patches for Zenoss

[24-Mar-2010 17:13:35] <Simon4> *nod*

[25-Mar-2010 00:00:30] [disconnected at Thu Mar 25 00:00:30 2010]

[25-Mar-2010 00:00:31] [connected at Thu Mar 25 00:00:31 2010]

[25-Mar-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[25-Mar-2010 02:54:57] <Troubadix09> morning all

[25-Mar-2010 03:31:59] <danuvius> morning

[25-Mar-2010 05:32:15] <Egyptian[web]> morning all

[25-Mar-2010 05:32:45] <Egyptian[web]> is there a ssh command in zenoss as part of the drop down menu (or did i create it?)

[25-Mar-2010 05:33:54] <forsberg> part of wich drop down menu ?

[25-Mar-2010 05:34:51] <Egyptian[web]> also .. i seem to hve problems with ssh modeling .. zCommandPassword and zCommandUsername are set but it doesnt work

[25-Mar-2010 05:35:15] <Egyptian[web]> forsberg: the drop down menu for the device .. along with snmpwalk and dns reverse

[25-Mar-2010 05:36:10] <Egyptian[web]> ERROR       zen.SshClient     SSH login to apple with SSH keys failed

[25-Mar-2010 05:36:23] <Egyptian[web]> but i dont want to use ssh keys .. !

[25-Mar-2010 05:46:33] <jocelyn_> Hi, I'd like to know if Zenoss core is translated in any language or if it's english-only ?

[25-Mar-2010 05:50:15] <Egyptian[web]> ah .. i see .. it tries first with ssh keys then with the password

[25-Mar-2010 07:58:42] <ckrough> anyone have good resources for reconfiguring existing RRD files? I need to remove holt winters and an RRA from a lot of them

[25-Mar-2010 08:01:47] <Simon4> looks like removing an RRA is hard work(tm), however you can resize the rra to one row with http://oss.oetiker.ch/rrdtool/doc/rrdresize.en.html

[25-Mar-2010 08:02:39] <Simon4> I have seen people in the past use rrdtool dump, edit the resultant xml, then rrdtool restore em back, but it'll take a bit of scripting

[25-Mar-2010 08:08:46] <ckrough> thats probably what i'll have to do

[25-Mar-2010 08:10:04] <mrayzenoss> morning all

[25-Mar-2010 08:13:29] <Simon4> heya

[25-Mar-2010 08:14:43] <Simon4> I need to be able to combine uk and us timezones, that way I could get 20 hours work done in a day and still get some sleep

[25-Mar-2010 08:17:34] <mrayzenoss> I'm only getting 4 or 5 hours of sleep anyway

[25-Mar-2010 08:17:48] <mrayzenoss> of course, I'm just wandering around in the dark trying to put a baby back to sleep

[25-Mar-2010 08:17:57] <Simon4> ah, self inflicted then ;)

[25-Mar-2010 08:17:59] <mrayzenoss> not doing something productive like hanging out on IRC

[25-Mar-2010 08:41:41] * Simon4 finds a bug in his parser and hurredly fixes it

[25-Mar-2010 09:00:46] <venturaville> are you still working on the blade chassis zenpack?

[25-Mar-2010 09:02:27] <Simon4> venturaville: yeah, I need to chat to you about that sometime actually :)

[25-Mar-2010 09:02:38] <venturaville> anytime

[25-Mar-2010 09:03:15] <Simon4> have replaced the modeller with a ssh one that grabs details, so you get a bunch more info, and updated the templates to reflect that.

[25-Mar-2010 09:03:32] <Simon4> am currently stress-testing it at work on our farm of chassis :)

[25-Mar-2010 09:03:52] <mrayzenoss> venturaville: did you want me to refresh the HP Blade Chassis ZenPack?  I've got it on my todo list for today

[25-Mar-2010 09:04:10] <mrayzenoss> 1.23 is on the page, 1.27 is in SVN

[25-Mar-2010 09:04:18] <venturaville> hrm

[25-Mar-2010 09:04:21] <venturaville> let me look real quick

[25-Mar-2010 09:06:27] <venturaville> it should be, however you might want to let Simon4 and I merge things before pushing out another one

[25-Mar-2010 09:06:42] <venturaville> also the one in subversion changes the Device Class location....

[25-Mar-2010 09:06:44] <mrayzenoss> whatever you prefer

[25-Mar-2010 09:06:51] <mrayzenoss> I can wait

[25-Mar-2010 09:06:55] <venturaville> k

[25-Mar-2010 09:07:26] <venturaville> mrayzenoss: I have a bug I could use some help with when you get a chance though

[25-Mar-2010 09:08:01] <venturaville> in the libvirt I have in subversion, if I have an installed version of it...make some changes change the version and do --install on it, I get an Attribute Error about id in libvirtHost

[25-Mar-2010 09:08:06] <venturaville> if I install it again it works.....

[25-Mar-2010 09:08:41] <venturaville> ./ZenUtils/ZenPackCmd.py:444:            if str(e) == "'ZenPack' object has no attribute '__of__'":

[25-Mar-2010 09:08:45] <venturaville> that is where it comes from....

[25-Mar-2010 09:11:43] <venturaville> Simon4: is there a way to get the fan data from the shell?

[25-Mar-2010 09:11:54] <Simon4> venturaville: as in how many/etc?

[25-Mar-2010 09:12:17] <venturaville> how fast they are spinning etc..

[25-Mar-2010 09:12:32] <venturaville> I am already using snmp and the SOAP API

[25-Mar-2010 09:12:39] <venturaville> ideally we could limit it to 2 methods

[25-Mar-2010 09:12:49] <Simon4> I think so

[25-Mar-2010 09:12:51] <venturaville> one would be better but

[25-Mar-2010 09:12:55] <venturaville> probably not workable for these

[25-Mar-2010 09:13:00] * Simon4 has a quick look

[25-Mar-2010 09:13:50] <venturaville> sweet

[25-Mar-2010 09:13:51] <venturaville> found it

[25-Mar-2010 09:13:56] <venturaville> show enclosure fan all

[25-Mar-2010 09:14:21] <Simon4> :)

[25-Mar-2010 09:14:34] <Simon4> I'm parsing show server info all for the blade details

[25-Mar-2010 09:14:35] <venturaville> now just need a way to cache this data in a pickle string somewhere

[25-Mar-2010 09:14:42] <venturaville> so we don't have to log into it a thousand times

[25-Mar-2010 09:14:50] <venturaville> per cycle

[25-Mar-2010 09:15:04] <venturaville> maybe a tmp file in /tmp or something?

[25-Mar-2010 09:16:15] <Simon4> yeah... I found with the modeler I could crap multiple commands into the one ssh pass, just by delimiting em with \n in the command string, so that only logs in once currently

[25-Mar-2010 09:16:20] <Simon4> haven't gone near collection yet

[25-Mar-2010 09:18:05] <Simon4> I've also created a "Chassis" tab to hold stuff like number of psu's, psu serial numbers, chassis serial numbers and all that good info, but haven't populated it yet

[25-Mar-2010 09:22:00] <venturaville> collection is the part I am wondering about caching the data from

[25-Mar-2010 09:22:35] <venturaville> zenoss doesn't lend itself easily to doing that, or at least there aren't a lot of examples out there

[25-Mar-2010 09:23:14] <venturaville> I have a disk latency monitor for Linux that I use, but it has to run over ssh and so I can't use it everywhere

[25-Mar-2010 09:23:20] <venturaville> (same problem)

[25-Mar-2010 09:39:09] forsberg is now known as fOrsberg

[25-Mar-2010 10:04:29] <zells> just curious if anyone has run into the network modeling issue on google maps with point to multipoint interfaces? or how to work around it?

[25-Mar-2010 10:04:49] <zells> its treating the /24 as a full mesh between all routers

[25-Mar-2010 10:19:37] <Brixius> I'm getting a bunch of warnings when I run zenmib, Unable to parse information from /usr/local/zenoss/common/share/mibs/<changes>/.index -- skipping

[25-Mar-2010 10:20:14] <Brixius> is it safe to delete these files?

[25-Mar-2010 10:31:35] <Brixius> I'm getting a bunch of warnings when I run zenmib, Unable to parse information from /usr/local/zenoss/common/share/mibs/<changes>/.index -- skipping

[25-Mar-2010 10:31:40] <Brixius> is it safe to delete these files?

[25-Mar-2010 11:54:53] <cornbread> Is it possible to send arbitrary data to zenoss and have it make graphs of this data based on a metric name?  Example: I start monitoring how many requests per second my server is processing, can I send this data to zenoss in XML or JSON Format and make new graphs from it?

[25-Mar-2010 11:55:29] <ckrough> not exactly, but you can write a python script that will pull the number from your server and record that

[25-Mar-2010 11:55:54] <mrayzenoss> yeah, I'm having trouble thinking of an example of a push-based graph

[25-Mar-2010 11:56:11] <mrayzenoss> but Zenoss can definitely pull from anything

[25-Mar-2010 11:56:15] <ckrough> without writing some sort of listener (ugh)

[25-Mar-2010 11:56:34] <mrayzenoss> hmm… there is a AMPQ ZenPack, that may graph off of a listener

[25-Mar-2010 11:56:47] <mrayzenoss> no, it's just events

[25-Mar-2010 11:57:03] <cornbread> So zenoss has to do the pulling

[25-Mar-2010 11:57:14] <ckrough> thats the easiest/normal way

[25-Mar-2010 11:57:28] <ckrough> it can digest syslog and snmptraps natively

[25-Mar-2010 11:58:42] <cornbread> So let's say I have a java process keeping track of how many new websites it indexes per second, what would be the best way to store this so zenoss could read it?  I would prefer to dump the data as soon as I calculate it

[25-Mar-2010 11:59:24] <mrayzenoss> cornbread: if the java app is exposing metrics via JMX you're good to go

[25-Mar-2010 12:00:13] <mrayzenoss> docs/DOC-3396

[25-Mar-2010 12:03:13] <cornbread> Thanks mrayzenoss I'll try that.  Never used JMX before

[25-Mar-2010 12:04:14] <citrus> Can i change the Time format for  %(firstTime)s

[25-Mar-2010 12:15:39] <citrus> everyone sleeping in there?

[25-Mar-2010 12:15:48] <citrus> here*

[25-Mar-2010 12:16:17] <mrayzenoss> lunchtime?

[25-Mar-2010 12:16:24] <mrayzenoss> Where do you want that Time format changed?

[25-Mar-2010 12:16:36] <citrus> in the alert message

[25-Mar-2010 12:18:07] <citrus> right now it shows 2010/03/25 10:14:48.000  i want something a bit more like        10:14:48 on 03/24/2010

[25-Mar-2010 12:20:10] <citrus> mrayzenoss: do you know?

[25-Mar-2010 12:21:17] <mrayzenoss> I'm looking a bit, I don't know

[25-Mar-2010 13:02:52] <mrayzenoss> citrus: it looks like it's Python calling str() on the dateTime

[25-Mar-2010 13:03:14] <mrayzenoss> I'm not sure how to format that though with the %(dateTime)s

[25-Mar-2010 13:34:35] <mrayzenoss> citrus: looks like that's not currently changeable

[25-Mar-2010 14:07:26] <citrus> thank you

[25-Mar-2010 14:35:32] <venturaville> would anyone be interested in a per group rollup report of usage/allocation numbers across devices (storage, cpu, memory, etc.)

[25-Mar-2010 14:36:05] <venturaville> I've got one that is marginally specific to us, but it might be more useful in a generic role....

[25-Mar-2010 14:42:22] <venturaville> no takers eh?

[25-Mar-2010 14:43:36] <mrayzenoss> venturaville: Yes, I would be interested

[25-Mar-2010 14:44:06] <mrayzenoss> Even if it's not generally useful, the more examples I have the better

[25-Mar-2010 14:51:37] <venturaville> its in svn now

[25-Mar-2010 14:51:44] <venturaville> this one is fairly generic

[25-Mar-2010 14:52:00] <venturaville> I have some changes that are more specific to make to it, but I am hoping that they will be minimal

[25-Mar-2010 14:56:40] <slops17> hey all

[25-Mar-2010 14:57:00] <venturaville> ZenPacks.community.groupreporting

[25-Mar-2010 14:57:03] <slops17> i just installed zenoss on my ubuntu machine

[25-Mar-2010 14:57:35] <slops17> i am trying to figure out what i need for it to auto discover my nodes on my network

[25-Mar-2010 14:57:52] <venturaville> for groupreporting to work you will need to give your devices a Group tag of some sort

[25-Mar-2010 14:57:58] <slops17> do i need snmp enabled on all devices including workstations?

[25-Mar-2010 14:58:39] <slops17> i have never used snmp before so not quite sure what i need to do

[25-Mar-2010 14:58:58] <venturaville> snmp or ssh

[25-Mar-2010 14:59:33] <venturaville> slops17: there should be some docs in the manual talking about autodiscover

[25-Mar-2010 14:59:44] <slops17> venturaville, do i need to just install snmp or actually set up the community settings?

[25-Mar-2010 14:59:54] <venturaville> you will need it to be accessible

[25-Mar-2010 15:00:02] <venturaville> more than likely you aren't going to be happy with the default settings

[25-Mar-2010 15:01:00] <slops17> what i am trying to do is monitor the bandwidth usage between an other location and the main office

[25-Mar-2010 15:01:25] <slops17> so i can get reports to the vp of it to justify getting a bigger pipe

[25-Mar-2010 15:02:17] <venturaville> if you have access to the network switches/routers via SNMP, then those might be your best bet

[25-Mar-2010 15:02:39] <slops17> we do but dont have snmp set up on the switches yet

[25-Mar-2010 15:02:52] <slops17> i know nothing about snmp yet

[25-Mar-2010 15:02:58] <slops17> just started to look at it

[25-Mar-2010 15:03:11] <venturaville> cisco/3com/nortel?

[25-Mar-2010 15:03:18] <slops17> dell

[25-Mar-2010 15:03:26] <venturaville> eww :-/

[25-Mar-2010 15:04:07] <venturaville> might be on your own on those

[25-Mar-2010 15:04:10] <mrayzenoss> slops17: docs/DOC-2606 has video from a Getting Started session I did

[25-Mar-2010 15:04:13] <slops17> have to work with what i have right

[25-Mar-2010 15:04:30] <venturaville> sorry ...Dell just brings back bad memories for me :-)

[25-Mar-2010 15:04:31] <slops17> mrayzenoss, thanks ill take a look

[25-Mar-2010 15:04:45] <mrayzenoss> and the Getting Started PDF is here: community/documentation

[25-Mar-2010 15:05:24] * slops17 looking thnaks

[25-Mar-2010 15:05:38] <mrayzenoss> slops17: basically you'll need to have some way for Zenoss to monitor your gear (SNMP, SSH, WMI, etc.).  For network devices, that's usually SNMP

[25-Mar-2010 15:07:11] <slops17> do i need an snmp server setup as well?

[25-Mar-2010 15:10:57] <mrayzenoss> so SNMP is a service running on your device, whether it's a router, Linux or Windows box.  It exposes metrics that Zenoss polls

[25-Mar-2010 15:11:11] <mrayzenoss> like enabling SSH

[25-Mar-2010 15:11:14] <mrayzenoss> with SSHD

[25-Mar-2010 15:12:05] <slops17> ok

[25-Mar-2010 15:12:27] <mrayzenoss> and the community string is essentially the password used to pull those metrics

[25-Mar-2010 15:12:29] <slops17> so the zenoss box is the snmp server then?

[25-Mar-2010 15:13:21] <mrayzenoss> not exactly.  The Zenoss box is a client to every device that runs the snmp daemon

[25-Mar-2010 15:13:30] <mrayzenoss> so it's the client, everything else is a server

[25-Mar-2010 15:13:54] <slops17> ah ok now i understand

[25-Mar-2010 15:14:03] <slops17> that makes more sense now

[25-Mar-2010 15:14:07] <mrayzenoss> except when you're dealing with SNMP traps, which are messages sent by the device to an endpoint

[25-Mar-2010 15:14:15] <slops17> i was confused

[25-Mar-2010 15:14:19] <mrayzenoss> where Zenoss would be the endpoint

[25-Mar-2010 15:14:48] <mrayzenoss> yeah, it's a little different, but once you get into it it's a very widespread standard

[25-Mar-2010 15:14:57] <slops17> so system name would be the device name then

[25-Mar-2010 15:15:03] <mrayzenoss> yeah

[25-Mar-2010 15:15:17] <slops17> system location is where the device is

[25-Mar-2010 15:15:30] <slops17> and the system contact is the admin person?

[25-Mar-2010 15:15:56] <mrayzenoss> yup

[25-Mar-2010 15:17:27] <slops17>  Community Strings read only?

[25-Mar-2010 15:22:17] <mrayzenoss> yes

[25-Mar-2010 15:36:12] <slops17> ok so i think i got it now

[25-Mar-2010 15:36:18] <slops17> thanks for the help guys

[25-Mar-2010 15:53:55] <theacolyte> traps are usually panic alerts

[25-Mar-2010 16:01:23] <QbY> quick question--where does one go to transform an alarm?

[25-Mar-2010 16:01:29] <QbY> err event

[25-Mar-2010 16:04:55] <QbY> disregard..  found it..

[25-Mar-2010 16:04:57] <mrayzenoss> docs/DOC-4766#d0e7149

[25-Mar-2010 16:05:44] <QbY> however; i have groups (my customers) that get alarms based on events going on that is their repsonsibility.  if they hit the "acknowledge" link in the email it makes them login--they have no login--is there a way to allow acknowledgements  WITHOUT logging in?

[25-Mar-2010 16:06:39] <mrayzenoss> QbY: docs/DOC-3237

[25-Mar-2010 16:08:55] <QbY> last question--i know its possible in the enterprise version, but i don't have $25k to drop for it..  there's no way to limit what people can see?  ie, if they are in a certain group they'd only see their group data?

[25-Mar-2010 16:09:37] <mrayzenoss> Jane Curry's got a paper that covers adjusting privileges on the users in the Zope interface.

[25-Mar-2010 16:09:51] <mrayzenoss> http://www.skills-1st.co.uk/papers/jane/users_events_zproperties_paper.pdf January 2009.

[25-Mar-2010 16:09:57] <mrayzenoss> http://www.skills-1st.co.uk/papers/jcurry.html

[25-Mar-2010 16:10:13] <mrayzenoss> that's the only documentation I've seen covering Zope permissions and Zenoss

[25-Mar-2010 16:10:19] <QbY> awesome..  now i have my nightly reading/work set out

[25-Mar-2010 16:24:39] <QbY> mrayzenoss: you have any documentation on all of the variables functions i have available in transforms?

[25-Mar-2010 16:26:03] <mrayzenoss> well, it's Python

[25-Mar-2010 16:26:33] <mrayzenoss> and you get the evt and device when you start

[25-Mar-2010 16:26:56] <mrayzenoss> essentially whatever you can do in zendmd or in the code

[25-Mar-2010 16:27:21] <mrayzenoss> community/documentation/wiki/event_transforms

[25-Mar-2010 16:27:41] <mrayzenoss> you might want to look at community/documentation/wiki/zendmd too

[25-Mar-2010 16:27:41] <QbY> k.  i'm reading this one here on supressing subsequent events--a definite need.

[25-Mar-2010 16:38:10] <QbY> i just noticed even though i have a router configured with "logging ip.ad.dd.ress" of my zenoss installation--there are no events associated with log entries

[25-Mar-2010 16:40:25] <mrayzenoss> they're Unknown?

[25-Mar-2010 16:40:35] <mrayzenoss> or not mapped to a device?

[25-Mar-2010 16:40:38] <QbY> i don't see entries anywhere--including configuration changes

[25-Mar-2010 16:41:06] <mrayzenoss> syslog messages aren't going from your router to your zenoss install?

[25-Mar-2010 16:41:22] <QbY> negative

[25-Mar-2010 16:41:25] <QbY> don't see any

[25-Mar-2010 17:11:29] <Brixius> I'm getting a bunch of warnings when I run zenmib, Unable to parse information from /usr/local/zenoss/common/share/mibs/<changes>/.index -- skipping

[25-Mar-2010 17:11:30] <Brixius> is it safe to delete these files?

[25-Mar-2010 17:14:21] <tuples> I'm trying to track down the format a data source needs to return as

[25-Mar-2010 17:14:43] <tuples> I've found NAME:VAR NAME=VAR NAME-VAR

[25-Mar-2010 17:14:46] <tuples> none seem to work

[25-Mar-2010 17:16:11] <tuples> its a data source with multiple data points, so I've tried NAME1 = VAR;NAME2 = VAR2

[25-Mar-2010 17:16:21] <tuples> I get no errors, everything seems to work great

[25-Mar-2010 17:16:45] <tuples> but there are no graphs being created so I suspect that my data return is malformated

[25-Mar-2010 17:28:06] <mrayzenoss> tuples: which parser are you using?

[25-Mar-2010 17:51:30] <stopher424> Hey guys, I checked out docs/DOC-3044 to see if I had paging setup correctly; Now I'm starting to wonder...How in the world does Zenoss send pages?  In the user prefs I set the pager so the number is 9285550155; Can it really page to a number? How does this all work?

[25-Mar-2010 18:07:32] <QbY> in my custom data points, the information is extracted via snmp as an integer--but i need to divide that integer by 10 for both the graphs and the warnings..  what do you guys suggest, a transform?

[26-Mar-2010 00:00:30] [disconnected at Fri Mar 26 00:00:30 2010]

[26-Mar-2010 00:00:31] [connected at Fri Mar 26 00:00:31 2010]

[26-Mar-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[26-Mar-2010 02:26:47] fOrsberg is now known as forsberg

[26-Mar-2010 02:46:29] <frozty_sa> hiya

[26-Mar-2010 02:46:47] <frozty_sa> how would I make a zenoss alert that'd act on something like traffic suddenly dropping a lot?

[26-Mar-2010 03:06:17] <voxter> Debating Zenoss, Cacti+Nagios (or OpsView), or Zabbix - I tried zenoss once about 2.5 years ago and it was a resource hog and the UI was very cumbersome to use/grasp when adding checks/systems/groups

[26-Mar-2010 03:06:45] <voxter> Can someone convince me otherwise? Im reading multiple "review" blog posts comparing each and they all seem very personal opinion driven based on what they already knew/liked

[26-Mar-2010 04:55:21] <danuvius> morning...

[26-Mar-2010 04:55:44] <danuvius> someone around that could me help out a little with ZenPack development?

[26-Mar-2010 04:58:44] <danuvius> frozty_sa: you're not looking for a minimal value threshold, but say if it drops certain percentage below avarage?

[26-Mar-2010 05:37:18] <frozty_sa> yeah

[26-Mar-2010 05:38:05] <frozty_sa> danuvius: what I pretty much want to accomplish is a something that'd freak out if, for instance, a route to somewhere important would go down but the interface stay up

[26-Mar-2010 05:38:17] <frozty_sa> one of the easiest ways of monitoring that is by checking traffic

[26-Mar-2010 05:41:10] <danuvius> hmmm in that cas a min threshold should work...

[26-Mar-2010 05:42:33] <danuvius> are you using core or enterprise?

[26-Mar-2010 05:45:06] <danuvius> hmmmm. disregard that last question, i thought the predictive filter was an enterprise zenpack, but i see it is a community zenpack

[26-Mar-2010 05:50:44] <danuvius> frozty_sa: are you using some dynamic routing protocol? For this cases I use syslog information that peers are gone and use alerting based on Syslog

[26-Mar-2010 05:55:50] <danuvius> frozty_sa: or snmp traps. If static routing is used a simple ping from or to the right ip will do the trick..

[26-Mar-2010 05:57:01] <frozty_sa> danuvius: that's the thing, those won't necessarily reflect properly

[26-Mar-2010 05:57:39] <frozty_sa> danuvius: let's say one of my upstream providers suddenly doesn't serve as much traffic as I should be doing, and I'm sitting at 3/10s of the capacity I'm supposed to have

[26-Mar-2010 05:57:55] <frozty_sa> I won't receive a trap or syslog message of any kind in that sort of situation

[26-Mar-2010 05:58:06] <frozty_sa> I'll look into the predictive filter zenpack

[26-Mar-2010 05:58:50] <danuvius> good luck with it...

[26-Mar-2010 07:49:58] <whitemice> any howto for how to connect to the MySQL instance used by ZenOSS?  (username, password, database name)

[26-Mar-2010 07:51:22] <ckrough> username and db name are listed under the event manager

[26-Mar-2010 07:51:56] <ckrough> db name is eventsthe db name is 'events'

[26-Mar-2010 07:52:11] <ckrough> woops

[26-Mar-2010 07:52:19] <ckrough> typed that twice. it's still early :)

[26-Mar-2010 07:53:06] <whitemice> arg;  so users and groups are going to be some zope thing.... </groans>

[26-Mar-2010 07:53:36] <whitemice> password is beaded out in  /ZenEventMqanager anyway.

[26-Mar-2010 07:53:52] <whitemice> now how to get under ZenOSS into Zope's admin interface

[26-Mar-2010 07:54:13] <whitemice> tips? (I know I've been there once before;  should have written it down)

[26-Mar-2010 07:54:28] <ckrough> just add /manage to the url

[26-Mar-2010 07:54:51] <ckrough> http://localhost:8080/zport/dmd/manage

[26-Mar-2010 07:55:00] <ckrough> or whatever you server name and port are

[26-Mar-2010 07:55:17] <whitemice> Bingo!  Thanks.

[26-Mar-2010 07:55:19] <ckrough> what are you trying to do? Most user/group stuff can be done in the gui

[26-Mar-2010 07:55:31] <whitemice> If you are going to OLF I'll buy you a bear.

[26-Mar-2010 07:55:43] <ckrough> be sure you know what you are changing if you do that in zope, they may not be what they seem

[26-Mar-2010 07:55:46] <ckrough> what's olf?

[26-Mar-2010 07:55:47] <whitemice> Yes, but the user/group app is gone belly up.

[26-Mar-2010 07:55:51] <whitemice> Ohio Linux Fest

[26-Mar-2010 07:55:52] <ckrough> ouch

[26-Mar-2010 07:55:59] <ckrough> ah, nah:) you'll have to drink it for me

[26-Mar-2010 07:56:09] * ckrough twists arm

[26-Mar-2010 07:57:21] <whitemice> going to the users tab under settings just "A Zenoss error has occurred"

[26-Mar-2010 07:57:34] <whitemice> hmmm, maybe there is an event log for errors in Zope somewhere.

[26-Mar-2010 07:57:40] <ckrough> yeah

[26-Mar-2010 07:57:48] <whitemice> Zope's managment interface makes Sharepoint look tidy!

[26-Mar-2010 07:57:54] <ckrough> $ZENHOME/log/Z2.log

[26-Mar-2010 07:57:58] <ckrough> and event.log

[26-Mar-2010 07:58:17] <ckrough> event.log is probably the one you want

[26-Mar-2010 07:58:59] <whitemice> We use LDAP auth for users; which works.  But once I added a user to a group - Blam!  No more user's tab. :(

[26-Mar-2010 07:59:16] <ckrough> odd. Im using ldap as well, no problems with users and groups

[26-Mar-2010 07:59:26] <ckrough> you may want to try reindexing zope

[26-Mar-2010 07:59:38] <whitemice> "reindexing zope"?

[26-Mar-2010 07:59:51] <ckrough> as the zenoss user run 'zendmd' on the command line then run 'reindex()' then 'commit()'

[26-Mar-2010 08:00:02] <ckrough> that shouldnt cause any problems, your mileage may vary

[26-Mar-2010 08:00:06] <whitemice> shutdown zenoss first?

[26-Mar-2010 08:00:10] <ckrough> nah

[26-Mar-2010 08:00:14] <whitemice> ok.....

[26-Mar-2010 08:01:03] <ckrough> if run that on large running setups with no issues

[26-Mar-2010 08:01:20] <ckrough> if you are particulary nervous you could backup ZENHOME/var/data.fs first

[26-Mar-2010 08:01:21] <whitemice> we are only ~100 devices

[26-Mar-2010 08:01:22] <ckrough> thats the zope database

[26-Mar-2010 08:02:40] <whitemice> done [reindex() commit()]

[26-Mar-2010 08:02:59] <whitemice> "Users" tab still poops.

[26-Mar-2010 08:03:03] <ckrough> damn

[26-Mar-2010 08:03:10] <ckrough> not sure what else to suggest

[26-Mar-2010 08:08:14] <whitemice> thread/13005   Sadly, no responses.

[26-Mar-2010 08:09:43] <ckrough> that is odd

[26-Mar-2010 08:10:06] <ckrough> in a little bit rmatte and mrayzenoss will probably be here, one of them may have more suggestions

[26-Mar-2010 08:10:14] <whitemice> I can't find the groups anywhere in zope, to remove people, back to how it was before.

[26-Mar-2010 08:11:45] <ckrough>  /zport/dmd/ZenUsers/manage

[26-Mar-2010 08:11:56] <ckrough> thats where the groups live, but I dont think you can edit the memberships there

[26-Mar-2010 09:28:43] <venturaville> mrayzenoss: can you update groups from the modeler?

[26-Mar-2010 09:28:49] <venturaville> (I noticed your dev email)

[26-Mar-2010 09:34:00] <mrayzenoss> venturaville: I don't know, I assume it can be done

[26-Mar-2010 09:34:22] <venturaville> mrayzenoss: it would be a lot easier if we could get full dmd access from the modeler for sure

[26-Mar-2010 09:34:31] <mrayzenoss> yeah

[26-Mar-2010 09:34:36] <venturaville> btw I put in some fixes for the NetApp ZenPack in SVN

[26-Mar-2010 09:34:41] <mrayzenoss> cool

[26-Mar-2010 09:35:47] <mrayzenoss> I spent too much time rewriting this ZenPack yesterday: docs/DOC-4538

[26-Mar-2010 09:36:08] * Simon4 arrives

[26-Mar-2010 09:36:11] <Simon4> hello all

[26-Mar-2010 09:36:12] <venturaville> I'll pass that on to Josh

[26-Mar-2010 09:36:12] <mrayzenoss> it's kinda fun to actually code for a change :)

[26-Mar-2010 09:36:13] <venturaville> he will like that

[26-Mar-2010 09:36:24] <venturaville> mrayzenoss: tell me about it ... I have to do much of my coding outside of work

[26-Mar-2010 09:36:34] <Simon4> our "service management" scary people would like that too

[26-Mar-2010 09:36:41] <venturaville> which reminds me... I just got the pools/volumes code working yesterday for libvirt

[26-Mar-2010 09:36:46] <mrayzenoss> haven't put up the new version yet, but I can't remember the last time I did any SQL or Python

[26-Mar-2010 09:36:49] <mrayzenoss> awesome

[26-Mar-2010 09:36:57] <venturaville> have to pretty up a bit though

[26-Mar-2010 09:37:34] <venturaville> mrayzenoss: I do need some help with a bug I keep hitting

[26-Mar-2010 09:37:53] <mrayzenoss> what's that?

[26-Mar-2010 09:37:54] <venturaville> I keeping getting an AttributeError about id, the first time I install a new version

[26-Mar-2010 09:38:01] <venturaville> it works fine the second time I do the install

[26-Mar-2010 09:38:15] <venturaville> darned if I can figure it out

[26-Mar-2010 09:38:26] <mrayzenoss> hmm… post it on the zenoss-dev forum and I'll make the support Dev take a look at it

[26-Mar-2010 09:38:31] <venturaville> k

[26-Mar-2010 09:38:36] <mrayzenoss> I haven't bugged him enough this iteration

[26-Mar-2010 09:38:41] <whitemice> ckrough: I just uncommented the assert in Products/PluggableAuthService/plugins/ZODBGroupManager.py restart zen and now it works

[26-Mar-2010 09:39:25] <ckrough> gratz

[26-Mar-2010 09:44:21] <venturaville> mrayzenoss: I sent out that email

[26-Mar-2010 09:47:30] <mrayzenoss> ok, got it into his queue :)

[26-Mar-2010 09:55:39] <venturaville> mrayzenoss: are they ever going to fix the Available Files/Capacity Files items in the filesystem status pages?

[26-Mar-2010 09:56:22] <mrayzenoss> Is there a ticket for that somewhere?  I can followup if there is

[26-Mar-2010 09:56:23] <venturaville> I've had the data sitting in netapp for a long while, and it still doesn't look like it ever gets picked up

[26-Mar-2010 09:56:35] <venturaville> don't know

[26-Mar-2010 09:56:39] <venturaville> it is a base zenoss thing

[26-Mar-2010 09:57:38] <venturaville> do you want one for it?

[26-Mar-2010 10:01:21] <venturaville> having trouble with IRC?

[26-Mar-2010 10:41:42] <mrayzenoss1> meetings

[26-Mar-2010 10:41:47] mrayzenoss1 is now known as mrayzenoss

[26-Mar-2010 10:53:31] <QbY> I'm trying to create a transform, I entered the transform code, and then created an event manually, but it doesn't drop it as desired--how can I verify that the transform is even being looked at?

[26-Mar-2010 11:05:05] <Brixius> Is there a way to manually edit an ethernet ports speed, I have a device that is misreporting the speed as 10mb and it's actually 1gb.

[26-Mar-2010 11:05:23] <zenossirc> #zenoss

[26-Mar-2010 11:07:21] slops17 is now known as slops_

[26-Mar-2010 11:08:04] slops_ is now known as slopslaptop

[26-Mar-2010 11:08:52] slopslaptop is now known as slops17

[26-Mar-2010 11:14:41] <mrayzenoss> QbY: http://dev.zenoss.org/trac/ticket/5572

[26-Mar-2010 11:15:02] <mrayzenoss> QbY: put zenhub in debug and the event traces are in there

[26-Mar-2010 11:15:20] <QbY> ok.

[26-Mar-2010 11:16:06] <QbY> would    evt._action = "drop" give the event a severity of -1

[26-Mar-2010 11:16:46] <mrayzenoss> no, it would drop it

[26-Mar-2010 11:16:50] <mrayzenoss> severity is 0-5

[26-Mar-2010 11:16:58] <mrayzenoss> dropped means not written to the DB

[26-Mar-2010 11:17:22] <QbY> that's what i thought; however, it's still showing up in our event console, AND emailing out

[26-Mar-2010 11:17:25] <QbY> but with a severity of -1

[26-Mar-2010 11:17:43] <QbY> i'm wondering if manual event entries would still get caught up in the transform

[26-Mar-2010 11:18:16] <QbY> sorry, priority is -1

[26-Mar-2010 11:20:48] <mrayzenoss1> rmatte: you around?

[26-Mar-2010 11:21:50] <QbY> per the zenuhb log, you don't even see my manually created event

[26-Mar-2010 11:21:51] <QbY> hrmmm

[26-Mar-2010 11:23:21] mrayzenoss1 is now known as mrayzenoss

[26-Mar-2010 11:37:51] <QbY> i'm beginning to believe that manually entered (via the web) events do not get transforms applied to them

[26-Mar-2010 11:39:39] <mrayzenoss> that would suck

[26-Mar-2010 11:39:46] <mrayzenoss> try using zensendevent from the CLI

[26-Mar-2010 11:39:51] <QbY> ok

[26-Mar-2010 11:40:14] <QbY> i'm going to lower one of my threshold alarms which will make it generate more alarms

[26-Mar-2010 12:15:41] <venturaville> mrayzenoss: what is storageDevice in the filesystem attributes supposed to represent... is that supposed to be represent the physical device the filesystem is attached to?

[26-Mar-2010 12:16:01] <venturaville> it is blank for every filesystem I have looked at thus far

[26-Mar-2010 12:18:06] <mrayzenoss> hmm… dunno.  I'll ask

[26-Mar-2010 12:26:02] Simon4_ is now known as Simon4

[26-Mar-2010 12:49:36] <ckrough> any idea what would cause this: ProtocolError: <ProtocolError for xxx.xxx.xxx.xxx:8080/zport/RenderServer: 401 Unauthorized>

[26-Mar-2010 12:49:44] <ckrough> getting that in event.log

[26-Mar-2010 13:01:39] slops17 is now known as slopslaptop

[26-Mar-2010 13:02:20] slopslaptop is now known as slopslaptop_

[26-Mar-2010 13:02:31] slopslaptop_ is now known as slopslaptop__

[26-Mar-2010 13:02:48] slopslaptop__ is now known as slops17

[26-Mar-2010 13:53:06] mrayzenoss1 is now known as mrayzenoss

[26-Mar-2010 13:57:45] <skipzoid> hey all

[26-Mar-2010 13:59:51] <Zaphodthe3rd> Hi just downloaded the vmware zenoss and imported it into esx4. Having a few problems. Anyone know how I enable SSH

[26-Mar-2010 14:00:02] <skipzoid> trigger = dict(getattr(evt, 'ccsAlarmTrigger',14) - will this work ???? I'm trying to read a value in what I think is a dictionary - ccsAlarmTrigger       (1, 3, 6, 1, 4, 1, 3142, 2, 7, 294, 2, 4, 1, 18)

[26-Mar-2010 14:01:45] <Zaphodthe3rd> Having a few funnies actually, tryin to run zenpack it says its not recognised command

[26-Mar-2010 14:01:51] <Zaphodthe3rd> from the cli

[26-Mar-2010 14:02:30] <Zaphodthe3rd> Maybe i should have used a different install other than the vm one

[26-Mar-2010 14:08:28] mrayzenoss1 is now known as mrayzenoss

[26-Mar-2010 14:12:09] <Zaphodthe3rd> anyone know why i cant ssh to the zenoss server?

[26-Mar-2010 14:12:14] <Zaphodthe3rd> keeps saying bad password?

[26-Mar-2010 14:12:22] <Zaphodthe3rd> and i know its not

[26-Mar-2010 14:14:36] <rhettardo> check your logs

[26-Mar-2010 14:14:55] <Simon4_> ssh -v hostname

[26-Mar-2010 14:15:56] Simon4_ is now known as Simon4

[26-Mar-2010 14:25:09] <Zaphodthe3rd> cheers i'll try that

[26-Mar-2010 14:34:11] <Simon4> venturaville: did you write check_bladechassis?

[26-Mar-2010 14:34:16] <venturaville> yes

[26-Mar-2010 14:34:40] <venturaville> quite a mess now with SNMP + SOAP API eh? :-)

[26-Mar-2010 14:35:12] <Simon4> I'm about to ahve a read, and get my head around all the curly braces ;)

[26-Mar-2010 14:43:11] <Simon4> so currently, in terms of collection/graphing, there's power/temp for each blade, and power/temp for the chassis (at least in what's graphed)

[26-Mar-2010 14:43:30] <venturaville> yep

[26-Mar-2010 14:43:42] <venturaville> and the SOAP API provides a little more detail on the power as well as fan setup and speed

[26-Mar-2010 14:44:01] <venturaville> try:

[26-Mar-2010 14:44:07] <venturaville> mybladename/xmldata?item=all

[26-Mar-2010 14:44:10] <venturaville> using https://

[26-Mar-2010 14:44:35] <venturaville> there are also height, width, depth, x, y offsets

[26-Mar-2010 14:44:49] <Simon4> yeah

[26-Mar-2010 14:44:50] <venturaville> so in theory you could actually draw what the system looked like, right down to the blinking LEDs.... :-P

[26-Mar-2010 14:45:00] <Simon4> how much do you care about height/width/depth by the way?

[26-Mar-2010 14:45:02] <theacolyte> With ascii characters

[26-Mar-2010 14:45:10] <venturaville> not enough to stress about it right now

[26-Mar-2010 14:45:16] <Simon4> yeah, I was tempted to draw a picture ;) maybe later

[26-Mar-2010 14:45:34] <Simon4> venturaville: cool, I used the real estate for other info

[26-Mar-2010 14:45:38] <mrayzenoss> venturaville: you've seen the new DataCenter View right?

[26-Mar-2010 14:45:47] <venturaville> nope

[26-Mar-2010 14:46:00] <mrayzenoss> heh, we wrote it for you guys

[26-Mar-2010 14:46:01] <mrayzenoss> lol

[26-Mar-2010 14:46:07] <venturaville> ah that one

[26-Mar-2010 14:46:11] <venturaville> yeah I have seen that one

[26-Mar-2010 14:46:19] <venturaville> thought he meant in one of the HP tools

[26-Mar-2010 14:46:23] <Simon4> I'm going to work on making what I've done less work centric this weekend, since it's probably a bit specialised. I've still only modified the modeler, and all the collection/etc keeps working (I set the snmpindex since it's the same as the slot number :D)

[26-Mar-2010 14:46:37] <theacolyte> Got a link to what this all looks like?

[26-Mar-2010 14:46:51] <venturaville> theacolyte: which part?

[26-Mar-2010 14:46:53] <Simon4> theacolyte: mray's stuff? or?

[26-Mar-2010 14:47:25] <venturaville> Simon4: in theory we could make the check script have an option to choose between SNMP, SSH, and SOAP API

[26-Mar-2010 14:47:34] <theacolyte> Ah, I'm seeing what you're saying... your bladecenter provides the info and you're trying to parse it using their SOAP API

[26-Mar-2010 14:47:46] <venturaville> supposedly there is a CIMOM one too, but that is just too traumatizing if you aren't ready for it :-)

[26-Mar-2010 14:47:52] <Simon4> venturaville: yeah, i was wondering about that

[26-Mar-2010 14:48:00] <venturaville> theacolyte: bingo

[26-Mar-2010 14:48:11] <Simon4> we have strict infosec requirements, so ssh/snmpv3 are good, other things less so

[26-Mar-2010 14:48:26] <Simon4> but we could make it switch based on zproperties or something I'm sure

[26-Mar-2010 14:48:44] <venturaville> yep

[26-Mar-2010 14:48:56] <venturaville> really really need to learn how to get zenoss to cache a pickled stats object

[26-Mar-2010 14:49:01] <venturaville> mrayzenoss: ?

[26-Mar-2010 14:49:36] <mrayzenoss> I can't find it on our .com website

[26-Mar-2010 14:49:37] <Simon4> if we do collection via ssh, is that with a zencommand called that does the ssh connection? or can we do it more smoothly than that (using zensshclient or similar)

[26-Mar-2010 14:50:04] <Simon4> passing auth parameters as command line arguments to a zencommand kinda defeats the purpose of using ssh a bit

[26-Mar-2010 14:50:24] <venturaville> if you break it into a separate modeling plugin then they can coexist

[26-Mar-2010 14:50:25] <Simon4> (I'm asking questions before reading examples, so thsi may be easily answered)

[26-Mar-2010 14:51:38] <mrayzenoss> http://www.zenoss.com/product/tours/enterprise has the screenshots

[26-Mar-2010 14:51:50] <mrayzenoss> "Datacenter Visualization" and "Rack Elevation View"

[26-Mar-2010 14:52:34] <Apachez> mrayzenoss: does the enterprise version have the same crappy (nonexistent) mysql config as the freeware version?

[26-Mar-2010 14:52:39] <Simon4> mrayzenoss: so we'll be able to plug in a "chassis elevation view" then :)

[26-Mar-2010 14:52:43] <Simon4> (hopefully)

[26-Mar-2010 14:52:45] <mrayzenoss> Apachez: yes

[26-Mar-2010 14:53:07] <Apachez> :S

[26-Mar-2010 14:53:47] <mrayzenoss> Apachez: it just means with a few good tickets you can suggest what we should be using

[26-Mar-2010 14:54:03] <mrayzenoss> the MySQL db isn't usually a problem

[26-Mar-2010 14:54:06] <mrayzenoss> it's the RRD

[26-Mar-2010 14:54:35] <venturaville> the mysql DB gets to be a problem when you have very large numbers of events.....

[26-Mar-2010 14:54:40] <venturaville> (which is our situation)

[26-Mar-2010 14:54:53] <venturaville> and that can be mitigated by very fast disks

[26-Mar-2010 14:55:17] <Simon4> mrayzenoss: heh, the mysqldb is the thing that kills us every time

[26-Mar-2010 14:55:38] <mrayzenoss> yeah, AFAIK we're just using a vanilla installation

[26-Mar-2010 14:55:46] <mrayzenoss> even in Enterprise

[26-Mar-2010 14:55:50] * Simon4 just had dbops at work build a new one

[26-Mar-2010 14:56:01] <Simon4> and it seems much better, even with 12 milllion rows in the history table

[26-Mar-2010 14:56:07] <mrayzenoss> Simon4: feel free to send that back as a wiki article or a ticket

[26-Mar-2010 14:56:07] <Simon4> *cough*

[26-Mar-2010 14:56:25] <Simon4> mrayzenoss: I will, giving it a little time to bed in etc

[26-Mar-2010 14:56:27] <mrayzenoss> I'm hoping to push for Drizzle in the post-Stone Crab release

[26-Mar-2010 14:57:25] <mrayzenoss> and the RRDTool caching daemon

[26-Mar-2010 14:58:07] <Simon4> mrayzenoss: is there a way to manually kick off history event aging?

[26-Mar-2010 14:58:26] <Simon4> just dropped the days to keep history for, and want to get it to age things now whilst it's quiet

[26-Mar-2010 14:59:02] <Apachez> are there any good reason for why innodb is being used?

[26-Mar-2010 14:59:12] <mrayzenoss> well, you can adjust the "Delete Historical Events Older Than (days)" and the "Event Aging Threshold (hours)"

[26-Mar-2010 14:59:13] <Apachez> myisam should be more lightweight for the purpose its used for

[26-Mar-2010 14:59:21] <Simon4> mrayzenoss: yeah, I adjusted that

[26-Mar-2010 14:59:30] <Simon4> want to make it actually go and do that though

[26-Mar-2010 14:59:37] <mrayzenoss> Apachez: we switched from myisam->innodb about 2 years ago, I don't know the story there

[26-Mar-2010 14:59:40] <Simon4> vs it running at some unknown time when I'm in bed

[26-Mar-2010 14:59:54] <Simon4> we only just switched from myisam, stupid error meant we were still using it

[26-Mar-2010 15:00:07] <Simon4> full table locks on things like aging events would destroy the entire box

[26-Mar-2010 15:00:15] <Apachez> not really

[26-Mar-2010 15:00:30] <Apachez> innodb is really slow to gather large amount of data like the events will store

[26-Mar-2010 15:00:42] <Simon4> Apachez: I'm no DBA, but I certainly saw the effects when the aging job was running

[26-Mar-2010 15:00:46] <mrayzenoss> I think the problem is we haven't taken the time to tune the MySQL side of things.

[26-Mar-2010 15:01:07] <ckrough> myisam doesnt allow field locking, innodb does, or something like that

[26-Mar-2010 15:01:14] <mrayzenoss> the devs who know DBs were hard-core Oracle guys

[26-Mar-2010 15:01:23] <mrayzenoss> and MySQL "just works" ;P

[26-Mar-2010 15:01:40] <Apachez> I have done some tweaks to the config but im too lazy to commit it back to zenoss :P

[26-Mar-2010 15:01:50] <Apachez> basically since I switched to splunk + munin :P

[26-Mar-2010 15:01:52] <ckrough> what he said: 'it's said that InnoDB works faster because it uses row locking instead of table locking'

[26-Mar-2010 15:01:54] <mrayzenoss> I think since then we've gotten some more MySQL experience in-house, it just hasn't gotten prioritized

[26-Mar-2010 15:01:56] <whitemice> buggers: every RDBMS "just works"

[26-Mar-2010 15:02:04] <Apachez> but I will have some spare time during easter to do some more tests with zenoss

[26-Mar-2010 15:02:27] <Apachez> ckrough: again, not true :)

[26-Mar-2010 15:02:31] <mrayzenoss> whitemice: which means everytime someone asks me about running on PostgreSQL/Oracle/blah I can tell them "knock yourself out" :)

[26-Mar-2010 15:02:49] <Apachez> but in short terms innodb performs better (IF and I say IF tweaked settings) during high concurrency compared to myisam

[26-Mar-2010 15:02:50] <mrayzenoss> we're also still on 5.0, which is kinda old

[26-Mar-2010 15:03:02] <Apachez> but for logging as with events myisam will most likely perform far better

[26-Mar-2010 15:03:04] <whitemice> no; it means that argument for MySQL is crap.  As a multi-DB admin I just find it very annoying "Well, MySQL just works -so I used that".

[26-Mar-2010 15:03:20] <mrayzenoss> whitemice: I agree

[26-Mar-2010 15:03:33] <whitemice> which means "I couldn't be bothered to think about it"

[26-Mar-2010 15:03:43] <Apachez> by the way isnt zenoss enterprise violating the mysql license? ;)

[26-Mar-2010 15:03:54] <mrayzenoss> nope, we already went around with them

[26-Mar-2010 15:04:06] <Apachez> they have some retarded thing that mysql binaries are not allowed to be included unless you buy a specific redistribution license

[26-Mar-2010 15:04:09] <mrayzenoss> we use the MySQL the OS provides

[26-Mar-2010 15:04:27] <Apachez> but its ok if you include just an install script that will fetch the mysql stuff from mysql.com for the clients

[26-Mar-2010 15:04:46] <Simon4> only time you get mysql is if you use the stack installer

[26-Mar-2010 15:04:50] <Apachez> I thought mysql was included in the stack?

[26-Mar-2010 15:04:54] <venturaville> it is

[26-Mar-2010 15:04:56] <Simon4> the rpm installer assumes you've already gotten yourself some sql

[26-Mar-2010 15:04:59] <mrayzenoss> it is, most customers use RPM

[26-Mar-2010 15:05:14] * Simon4 was assuming that enterprise would have been RPM, not stack

[26-Mar-2010 15:05:15] <mrayzenoss> again, more fuel for Drizzle :)

[26-Mar-2010 15:05:39] <ckrough> heh

[26-Mar-2010 15:05:57] <ckrough> every time I hear it I think of Master Shake

[26-Mar-2010 15:06:03] <ckrough> The Driiiiizzzle

[26-Mar-2010 15:06:11] <Apachez> I would expect a deb as enterprise :)

[26-Mar-2010 15:11:34] <Simon4> hurrah, events finally aging, and it's not blocking the whole table

[26-Mar-2010 15:12:42] <mrayzenoss> seriously though, I'm pretty sure there's been next to no tuning on the DB so please submit tickets or wiki articles and we'll try to squeeze a little more out of the db

[26-Mar-2010 15:12:57] <mrayzenoss> and Stone Crab will still be on MySQL 5.0

[26-Mar-2010 15:13:29] <mrayzenoss> even though some users are using 5.1 and 5.4

[26-Mar-2010 15:30:37] <Simon4> mmmm, kiwi beer and python... my Friday nights are getting a bit geeky of late

[26-Mar-2010 15:33:04] <Zaphodthe3rd> anyone know why i get this  /bin/bash: /root/.bashrc: Permission denied when i run anything logged in as zenoss

[26-Mar-2010 15:34:03] <Zaphodthe3rd> so installing a zenpack from a fresh install is a bit of a challenge

[26-Mar-2010 15:34:09] <mrayzenoss> are you using su?

[26-Mar-2010 15:34:23] <Zaphodthe3rd> yep

[26-Mar-2010 15:34:27] <Zaphodthe3rd> su zenoss

[26-Mar-2010 15:34:30] <mrayzenoss> 'su - zenoss'

[26-Mar-2010 15:35:11] <mrayzenoss> without the '-' you're trying to use the root's env

[26-Mar-2010 15:36:10] <Zaphodthe3rd> ahh cheers sorry im a bit of a noob

[26-Mar-2010 15:36:22] <Zaphodthe3rd> i'll get there eventually

[26-Mar-2010 15:36:31] <mrayzenoss> I'm just glad it was something I knew

[26-Mar-2010 15:36:35] <Zaphodthe3rd> that works great

[26-Mar-2010 15:36:57] <Zaphodthe3rd> well be carefull i may have plently more if this zenpack doesnt show

[26-Mar-2010 15:37:06] <Zaphodthe3rd> thanks for helping

[26-Mar-2010 15:37:07] <Simon4> mrayzenoss: with modeler plugins, is there something somewhere in the zenpack to register them? or do they just need to exist under modeler/plugins/whatnot.py?

[26-Mar-2010 15:38:43] <Zaphodthe3rd> do you know if i can monitor windows services on boxes on the vanilla install or do i need a wmi zenpack

[26-Mar-2010 15:38:54] <mrayzenoss> sevices are out of the box

[26-Mar-2010 15:39:07] <mrayzenoss> Simon4: don't know, haven't dug into that yet

[26-Mar-2010 15:39:18] <Simon4> mrayzenoss: hokay, I'll do a quick experiment

[26-Mar-2010 15:42:34] <Zaphodthe3rd> i take it if your zenpack doesnt show in the gui zenpacks installed bit its not installed properly

[26-Mar-2010 15:43:13] <mrayzenoss> that's correct

[26-Mar-2010 15:43:20] <Zaphodthe3rd> lol great

[26-Mar-2010 15:43:45] <Zaphodthe3rd> ive installed it through cli and it said installed and done a zenoss restart

[26-Mar-2010 15:46:07] <Zaphodthe3rd> its put it here b the looks of things/home/zenoss/ZenPacks/zenpack.egg/ZenPacks/community/NetApp

[26-Mar-2010 15:50:58] <Simon4> mrayzenoss: answer = put it in the directory, and it's in the "Available Fields" list once you've installed the zenpack

[26-Mar-2010 15:52:29] <mrayzenoss> cool

[26-Mar-2010 15:52:46] <Simon4> now to actually put contents into the modeller code

[26-Mar-2010 16:11:36] <Simon4> w00t, now modelling all the fans within a bladechassis

[26-Mar-2010 16:11:57] <Simon4> now for PSU details

[26-Mar-2010 16:23:11] <Zaphodthe3rd> im not sure the vmware zenoss has wmi monitoring out of the box

[26-Mar-2010 16:23:32] <Zaphodthe3rd> cant see wmi mentioned anywhere and no where to enter creds

[26-Mar-2010 16:23:48] <Zaphodthe3rd> looks like its just snmp

[26-Mar-2010 16:24:02] <Simon4> Zaphodthe3rd: head to /Devices/Server/Windows

[26-Mar-2010 16:24:06] <Simon4> then click on the zproperties tab

[26-Mar-2010 16:24:16] <Simon4> down the bottom is place for zWinUser and zWinPassword etc

[26-Mar-2010 16:24:22] <Simon4> add devices into that deviceclass

[26-Mar-2010 16:26:24] <Zaphodthe3rd> ahh so its in zproperties

[26-Mar-2010 16:27:10] <Simon4> yeah, tis the place to set most values of that type

[26-Mar-2010 16:27:47] <Zaphodthe3rd> i'll give it a shot

[26-Mar-2010 16:29:33] * Simon4 gets bogged down in debugging tal stuff

[26-Mar-2010 16:36:18] <Zaphodthe3rd> well ive found it

[26-Mar-2010 16:36:49] <Zaphodthe3rd> its a bit crazy to navigate

[26-Mar-2010 16:37:05] <mrayzenoss> yeah, that's getting fixed in the next major release

[26-Mar-2010 16:43:19] <Zaphodthe3rd> im going to have to quit and i havent even managed to get one windows device bringing anything back except for that its up

[26-Mar-2010 16:44:10] <Zaphodthe3rd> whens the major release out

[26-Mar-2010 16:47:18] <Zaphodthe3rd> could there possibly be anything else to do other than give it its name\ip add it to the windows class and specify a user account with admin access to the box

[26-Mar-2010 16:48:26] <Zaphodthe3rd> right time for a beer

[26-Mar-2010 16:50:01] * Simon4 manages to add his "fans" tab, but destroy his "blades" tab

[26-Mar-2010 16:50:11] <Simon4> I hate these kind of bugs

[26-Mar-2010 17:06:53] <mrayzenoss> Updated the Event Histograms ZenPack: docs/DOC-4538

[26-Mar-2010 17:08:27] <Simon4> a ha!

[26-Mar-2010 17:08:31] * Simon4 works out his stupidity

[26-Mar-2010 17:19:16] * QbY wishes he could be playing with his Zenoss

[26-Mar-2010 17:27:12] <mrayzenoss> later all!

[26-Mar-2010 18:49:59] <Simon4> http://lol.nzdance.net/bladezenpack1.png

[26-Mar-2010 18:50:02] <Simon4> schweeeet!

[26-Mar-2010 18:58:46] * Simon4 heads to bed, having sorted his stuff for tonight

[27-Mar-2010 00:00:19] <frozty_sa> Simon4: omnomnom at that screenshot

[27-Mar-2010 00:00:30] [disconnected at Sat Mar 27 00:00:30 2010]

[27-Mar-2010 00:00:30] [connected at Sat Mar 27 00:00:30 2010]

[27-Mar-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[27-Mar-2010 02:50:51] <Simon4> morning all

[27-Mar-2010 03:31:39] <skipzoid> anyone else here ?

[27-Mar-2010 03:32:35] * Simon4 is

[27-Mar-2010 03:36:26] <skipzoid> :-)

[27-Mar-2010 03:36:50] <skipzoid> are you any good with python scripting ?

[27-Mar-2010 03:40:07] <Simon4> I'm okay.. what are you trying to do?

[27-Mar-2010 03:41:18] <skipzoid> trigger = dict(getattr(evt, 'ccsAlarmTrigger',14) - will this work ???? I'm trying to read a value in what I think is a dictionary - ccsAlarmTrigger   (1, 3, 6, 1, 4, 1, 3142, 2, 7, 294, 2, 4, 1, 18)

[27-Mar-2010 03:42:14] <skipzoid> what I want is to test the value 18 - the device reports 18 for a set fault and 19 for a clear fault

[27-Mar-2010 03:45:24] <Simon4> so you're trying to see if the value 18 is in the set (1,3,6,x,x,x,18)?

[27-Mar-2010 03:46:41] <skipzoid> yep - exactly - the goal is to be able to set the evt.severity so if 18 is present it sets it to critical and if 19 then sets it clear

[27-Mar-2010 03:47:01] <skipzoid> i'm assuming this will clear it from the active events list

[27-Mar-2010 03:47:23] <Simon4> >>> fred = (1,3,6,4,3,2,18)

[27-Mar-2010 03:47:23] <Simon4> >>> fred

[27-Mar-2010 03:47:23] <Simon4> (1, 3, 6, 4, 3, 2, 18)

[27-Mar-2010 03:47:24] <Simon4> >>> 18 in fred

[27-Mar-2010 03:47:24] <Simon4> True

[27-Mar-2010 03:47:24] <Simon4> >>> 10 in fred

[27-Mar-2010 03:47:24] <Simon4> False

[27-Mar-2010 03:47:35] <Simon4> so you can just say "if 18 in setName:"

[27-Mar-2010 03:52:43] <skipzoid> something like : trigger = str(getattr(evt, 'ccsAlarmTrigger')

[27-Mar-2010 03:52:56] <skipzoid> if "18" in trigger

[27-Mar-2010 03:57:55] <Simon4> yeah

[27-Mar-2010 04:00:01] <skipzoid> is -  if "18" in trigger            valid syntax ?

[27-Mar-2010 04:00:41] <Simon4> yup

[27-Mar-2010 04:00:44] <Simon4> it'll be

[27-Mar-2010 04:00:48] <Simon4> if "18" in trigger:

[27-Mar-2010 04:01:17] <Simon4> presuming trigger is something like ("1", "3", "18")

[27-Mar-2010 04:01:22] <Simon4> since you're making strings

[27-Mar-2010 04:01:27] <skipzoid> evt.severity = 0                     is a clear ?

[27-Mar-2010 04:02:44] <Simon4> I think so, yeah

[27-Mar-2010 04:03:03] <skipzoid> lets try it then, back in a mo - need to walk into the other room to fire off a trap

[27-Mar-2010 04:10:14] <skipzoid> simon4: thank you very much - I'm learning all the time and with your help its now setting and clearing the events when the actual fault clears !! genius !

[27-Mar-2010 04:10:27] <Simon4> skipzoid: no problems, glad it worked :)

[27-Mar-2010 05:34:01] <sergeymasushko> can some one guide me how to make correct regex for the threshold?

[27-Mar-2010 05:52:15] <frozty_sa> skipzoid: if you need to only evaluate on the last value, you can do 'if "18" == trigger[-1]:'

[27-Mar-2010 05:52:50] <sergeymasushko> ?

[27-Mar-2010 05:55:54] <sergeymasushko> I added the following threshold Min Value: here.memAvailSwap * .5

[27-Mar-2010 05:55:54] <sergeymasushko> but I receive the error like: User-supplied Python expression (here.memAvailSwap * .5) for minimum value caused error: ['memAvailSwap_memAvailSwap']

[27-Mar-2010 05:55:54] <sergeymasushko> The threshold Swap_utilization in template /zport/dmd/Devices/Server/rrdTemplates/Device has been disabled.

[27-Mar-2010 05:55:54] <sergeymasushko> What is wrong with it?

[27-Mar-2010 10:05:21] <Brixius> I have a device reporting the incorrect port speed, it's reporting 10mb when it's actually 1gb, is there a way to override that on the interface settings, ie manually set it?

[27-Mar-2010 12:14:27] <QbY> is it possible to create a database with a dynamically generated oid?  i'm querying cable modems and some of the information i need comes from the cmts they are connected to--but the end of the oid in the  cmts is the cable modem's mac address in decimal format

[27-Mar-2010 13:11:23] <rmatte> QbY: only by creating a command based datasource with a custom script to collect the data

[27-Mar-2010 13:11:36] <QbY> ok

[27-Mar-2010 13:11:41] <rmatte> or by writing a collector plugin, but the command datasource would be easier

[27-Mar-2010 13:11:56] <rmatte> I've done similar for monitoring CPU and Memory on windows boxes via SNMP

[27-Mar-2010 13:12:14] <rmatte> since the OIDs are very dynamic

[27-Mar-2010 13:13:08] <QbY> rmatte: do you know of some bang up documentation on creating transforms--mine aren't working when i create a manual event

[27-Mar-2010 13:17:54] <QbY> I tried this one--placed it in Perf/Snmp-- and it doesn't drop

[27-Mar-2010 13:17:56] <QbY> docs/DOC-2554

[27-Mar-2010 13:18:09] <QbY> the one for Dropping Events --Error reading value for...

[27-Mar-2010 13:19:12] <rmatte> they won't work with manual events from my experience

[27-Mar-2010 13:19:16] <rmatte> it has to be an actual event

[27-Mar-2010 13:19:18] <QbY> oh ok

[27-Mar-2010 13:19:25] <QbY> will there be a log of the transform

[27-Mar-2010 13:19:51] <rmatte> when you're creating manual events I'm pretty sure it just directly creates the event in MySQL without having it pass through the same steps as actual events do

[27-Mar-2010 13:20:05] <rmatte> no, you won't see a log that the transform was performed

[27-Mar-2010 13:20:11] <rmatte> you'll simply see the results of the transform

[27-Mar-2010 13:20:16] <rmatte> such as if you have it modify something

[27-Mar-2010 13:20:56] <rmatte> If you were to run the right daemon in debug mode you might see some log info relating to the transformation, but it's not worth it

[27-Mar-2010 13:21:27] <rmatte> I'll check out your post...

[27-Mar-2010 13:21:40] <rmatte> oh, thos are just examples

[27-Mar-2010 13:21:53] <rmatte> error reading value for what?

[27-Mar-2010 13:21:57] <rmatte> what's the full error message?

[27-Mar-2010 13:22:42] <QbY> Error reading value for "T1 1/7" on ... (oid .1.3.6.1.2.1.2.2.1.16.12 is bad)

[27-Mar-2010 13:23:02] <rmatte> oh, try adding a .0 to the end of the OID

[27-Mar-2010 13:23:05] <QbY> That OID is looking for traditional interface performance information--but that particular interface won't have it.

[27-Mar-2010 13:23:11] <QbY> sub interfaces, etc

[27-Mar-2010 13:23:13] <rmatte> Zenoss doesn't support OID trees

[27-Mar-2010 13:23:22] <rmatte> oh I see what you mean

[27-Mar-2010 13:23:56] <rmatte> yeh, I usually just create a local copy of the template on the interface and remove those OIDs... the other thing you can do is to create new interface templates for the different types of interfaces

[27-Mar-2010 13:24:29] <rmatte> if you copy the ethernetCsmacd template and change the name to match the "type" listed when you click on the interface then it'll automatically apply that template to those types of interfaces

[27-Mar-2010 13:24:33] <QbY> ok..  is it normal for the text in the transform editor to be red?

[27-Mar-2010 13:24:38] <QbY> or does that indicate a failure

[27-Mar-2010 13:24:42] <QbY> err syntax error

[27-Mar-2010 13:24:49] <rmatte> so if you know that interfaces with type of "other" don't support any of those OIDs, you can just make a template called other and remove those OIDs

[27-Mar-2010 13:24:55] <rmatte> no

[27-Mar-2010 13:25:03] <rmatte> if the text is red it means you have a syntax error in your python

[27-Mar-2010 13:25:06] <QbY> ok, i've added it, i'm going to see if it corects it

[27-Mar-2010 13:25:15] <rmatte> fpaste.org the text and I'll have a look

[27-Mar-2010 13:25:30] <QbY> i've added that one -- just as it appears on the web page and it turned blue after save

[27-Mar-2010 13:25:45] <rmatte> it shouldn't even turn blue, it should just be black

[27-Mar-2010 13:25:46] <rmatte> lol

[27-Mar-2010 13:25:55] <rmatte> I don't even know what blue would mean, never seen that happen

[27-Mar-2010 13:26:33] <rmatte> those error reading value messages are just debug messages, relatively harmless, all you can try to do is cut them down as best you can

[27-Mar-2010 13:26:52] <QbY> in transforms is the here.getRRDValue('name') available to me?

[27-Mar-2010 13:27:06] <rmatte> no, you'd need to do...

[27-Mar-2010 13:27:26] <rmatte> dev = dmd.Devices.findDevice(evt.device)

[27-Mar-2010 13:27:35] <rmatte> dev.getRRDValues('name')

[27-Mar-2010 13:27:41] <rmatte> dev.getRRDValue('name')

[27-Mar-2010 13:27:43] <rmatte> rather

[27-Mar-2010 13:27:58] <rmatte> and if it's for an interface you have to query that interface name

[27-Mar-2010 13:28:01] <rmatte> something like...

[27-Mar-2010 13:28:03] <QbY> i may be jumping hte gun, but these events happen on particular interfaces on the device

[27-Mar-2010 13:28:32] <QbY> do you remember the total active users problem i had the other day?

[27-Mar-2010 13:29:00] <rmatte> http://dev.zenoss.com/trac/ticket/5376

[27-Mar-2010 13:29:06] <rmatte> look at the transform I posted in that trac ticket

[27-Mar-2010 13:29:10] <rmatte> it relates to specific interfaces

[27-Mar-2010 13:29:11] <QbY> one thing that would stop my constant barrage of questions is if i could find a great documentation of the variables and methods available

[27-Mar-2010 13:29:24] <rmatte> it doesn't actually work because of the way Zenoss handles interface monitoring, but the code is fine

[27-Mar-2010 13:29:30] <rmatte> maybe that'll make you understand how to do it

[27-Mar-2010 13:29:52] <rmatte> instead of interface.operStatus = 2 you could do interface.getRRDValue

[27-Mar-2010 13:30:12] <QbY> ok

[27-Mar-2010 13:30:20] <rmatte> there isn't really one be all end all piece of documentation

[27-Mar-2010 13:30:27] <rmatte> you need to hunt around until you understand

[27-Mar-2010 13:30:27] <QbY> yeah, i've noticed that

[27-Mar-2010 13:30:59] <rmatte> by the way, you don't need all that code at the top of the transform leading up to dmd = (in my example)

[27-Mar-2010 13:31:04] <rmatte> that used to be required but isn't anymore

[27-Mar-2010 13:31:32] <rmatte> what you are more interested in is...

[27-Mar-2010 13:31:33] <rmatte> d = dmd.Devices.findDevice(evt.ipAddress)

[27-Mar-2010 13:31:46] <rmatte> and for interface in d.os.interfaces():

[27-Mar-2010 13:32:05] <QbY> last question--and them i'm off to make all these "dreams" come true--so i can prove zenoss to mgmt on monday and drop the old methods

[27-Mar-2010 13:32:09] <rmatte> then you can do if interface.getInterfaceName() evt.component:

[27-Mar-2010 13:32:18] <rmatte> if interface.getInterfaceName() == evt.component:

[27-Mar-2010 13:32:19] <rmatte> rather

[27-Mar-2010 13:32:25] <QbY> i need to extract performance graph data..  so that it can be included in our existing web portals

[27-Mar-2010 13:32:27] <rmatte> assuming the component of the event is the interface name

[27-Mar-2010 13:32:28] <QbY> suggestions?

[27-Mar-2010 13:32:44] <rmatte> no idea, that's extremely complex as I explained the other day

[27-Mar-2010 13:33:15] <QbY> can i build a report and then a special page that doesn't require a login?

[27-Mar-2010 13:33:15] <rmatte> you'd basically have to code something to display the RRD data per device and go at the RRD files directly

[27-Mar-2010 13:33:19] <rmatte> that would be the simplest method

[27-Mar-2010 13:33:43] <rmatte> I don't think you can just make 1 page not require a login

[27-Mar-2010 13:33:56] <rmatte> but maybe, you'd have to screw around in the zope management interface

[27-Mar-2010 13:34:28] <rmatte> but like I said, honestly, easiest method would be to build something that works with the RRD files directly

[27-Mar-2010 13:34:34] <rmatte> they are located in $ZENHOME/perf

[27-Mar-2010 13:34:42] <rmatte> but it's still a lot of coding

[27-Mar-2010 13:35:09] <QbY> yeah

[27-Mar-2010 13:35:18] <QbY> i guess i need to install apache on the same box

[27-Mar-2010 13:35:21] <rmatte> there is no simple way to do it, is the short answer

[27-Mar-2010 13:35:25] <rmatte> yeh

[27-Mar-2010 13:35:53] <rmatte> unless you could somehow make the report, and have some proxy auto-login to Zenoss and cache that, then serve it up

[27-Mar-2010 13:35:59] <rmatte> no idea how you'd configure that though

[27-Mar-2010 13:36:25] <rmatte> anyways, I've gotta go, I'll be back on Monday

[27-Mar-2010 13:36:26] <rmatte> cheers

[27-Mar-2010 13:37:17] <QbY> ok, have a good one

[27-Mar-2010 16:28:25] <cornbread> Is it possible to /join #django

[27-Mar-2010 16:28:41] <cornbread> fuck, sorry

[27-Mar-2010 16:28:42] <cornbread> heh

[27-Mar-2010 22:52:40] forsberg is now known as fOrsberg

[28-Mar-2010 00:00:30] [disconnected at Sun Mar 28 00:00:30 2010]

[28-Mar-2010 00:00:30] [connected at Sun Mar 28 00:00:30 2010]

[28-Mar-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[28-Mar-2010 11:59:15] bzed_ is now known as bzed

[28-Mar-2010 12:31:31] HCoyote_ is now known as HCoyote

[28-Mar-2010 12:57:11] straterr1 is now known as straterra

[28-Mar-2010 15:01:52] fOrsberg is now known as forsberg

[28-Mar-2010 15:41:24] dibbz_ is now known as dibbz

[28-Mar-2010 17:27:18] <QbY> howdy all

[28-Mar-2010 17:28:13] <QbY> Trying to make my event summary a bit more informative using a transform--can someone tell me why this isn't working: http://pastebin.com/xkkG1Z8m

[28-Mar-2010 22:37:14] forsberg is now known as fOrsberg

[29-Mar-2010 00:00:30] [disconnected at Mon Mar 29 00:00:30 2010]

[29-Mar-2010 00:00:30] [connected at Mon Mar 29 00:00:30 2010]

[29-Mar-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[29-Mar-2010 00:40:15] <ahowchin> Wow. Hi all

[29-Mar-2010 00:41:43] <ahowchin> I'm looking to implement a server monitoring system at work. We want basic host info (disk, cpu, memory) as well as info about the application servers running on the host (Heap usage, GC's etc.). We're currently looking at both Hyperic and Zenoss (both open source versions). Just wondering - does Zenos work with application servers (specifically JBoss) out of the box?

[29-Mar-2010 00:55:37] <rhettardo> i believe so but am not certain because i dont use those

[29-Mar-2010 06:52:08] <Dieterbe> hello guys

[29-Mar-2010 06:52:47] <Dieterbe> is there a way to make errors look the same as criticals on the event console? (ie both in red)

[29-Mar-2010 06:52:57] <Dieterbe> using 2.4.5

[29-Mar-2010 08:31:45] MattD is now known as mdereus

[29-Mar-2010 08:50:15] <ckrough> morning

[29-Mar-2010 08:53:03] <Dieterbe> hi

[29-Mar-2010 08:55:49] fOrsberg is now known as forsberg

[29-Mar-2010 09:42:09] <rmatte> good morning all

[29-Mar-2010 09:43:48] <ckrough> morning Ryan

[29-Mar-2010 09:44:35] <venturaville> good morning :-)

[29-Mar-2010 09:45:48] <Brixius> Good Morning

[29-Mar-2010 09:46:48] <ckrough> venturaville: running 2.5.2 yet?

[29-Mar-2010 09:47:41] <venturaville> ckrough: yep

[29-Mar-2010 09:47:50] <ckrough> any major problems?

[29-Mar-2010 09:48:00] <venturaville> lots

[29-Mar-2010 09:48:07] <ckrough> heh, great. like?

[29-Mar-2010 09:48:10] <ckrough> Im upgrading today

[29-Mar-2010 09:48:44] <rmatte> I'm testing 2.5.2 for upgrade, so far the test box that I upgraded from 2.4.5 to 2.5.2 has been running fine...

[29-Mar-2010 09:48:51] <venturaville> we have one nasty one that seems to affect us because of the sheer number of zenpacks/mibs

[29-Mar-2010 09:48:52] <rmatte> the upgrade steps are fairly specific though

[29-Mar-2010 09:49:21] <rmatte> venturaville: yeh, but that bug has been around for ages

[29-Mar-2010 09:49:56] <ckrough> well, we'll see how it behaves. This setup has 9 collectors plus the head unit

[29-Mar-2010 09:50:08] <rmatte> fun fun

[29-Mar-2010 09:50:10] <ckrough> roughly 1 million OIDs

[29-Mar-2010 09:50:49] <ckrough> I tested the upgrade process in dev several times and didnt see anthing major, but doing it in Prod is always a different animal

[29-Mar-2010 09:51:07] * rmatte you should have put your little finger to the corner of your mouth when you said that, and finished it off with a maniacal laugh

[29-Mar-2010 09:51:17] <ckrough> how do you know i didnt ;)

[29-Mar-2010 09:51:24] <rmatte> "1 million OIDs, muahahaha"

[29-Mar-2010 09:51:26] <ckrough> one miiiilion oids

[29-Mar-2010 09:51:28] <rmatte> hehe

[29-Mar-2010 09:52:04] <rmatte> I find 2.5.2 is running a bit slower than 2.4.5

[29-Mar-2010 09:52:19] <rmatte> the UI is anyways

[29-Mar-2010 09:52:42] <venturaville> yeah, unfortunately we are having a hard time distinguishing increased load on zenoss overall versus the issues with 2.5.2

[29-Mar-2010 09:52:47] <rmatte> I cloned one of my 2.4.5 production servers then upgraded the clone, and the 2.4.5 install is performing much better

[29-Mar-2010 09:52:52] <venturaville> fusionIO helps a bit

[29-Mar-2010 09:52:57] <venturaville> but we don't have them on our collectors yet

[29-Mar-2010 09:53:22] <rmatte> fusionIO on the collectors would really only help with the RRDs

[29-Mar-2010 09:53:26] <ckrough> do you get the 320s or the 160s?

[29-Mar-2010 09:53:40] <ckrough> go for the 320s if they make an SLC in that size

[29-Mar-2010 09:53:44] <ckrough> 160 fills up fast

[29-Mar-2010 09:53:44] <venturaville> 80s at the moment

[29-Mar-2010 09:54:17] <ckrough> witha  160 Ive got around 300k OIDs on one server, but you run out of disk space... not sure what the max could be on a 320gb ssd

[29-Mar-2010 09:57:22] <ckrough> venturaville: do you mount all of /opt/zenoss on the ssd, or just /opt/zenoss/var? or is it for mysql

[29-Mar-2010 10:15:19] <rmatte> cluther: g'day

[29-Mar-2010 10:16:09] <rmatte> ckrough: you'd probably want to mount /opt/zenoss/var and /opt/zenoss/perf at the very least

[29-Mar-2010 10:16:28] <rmatte> using it for MySQL as well wouldn't be a bad idea if you've got the space

[29-Mar-2010 10:16:36] <ckrough> I do perf on the collectors, not sure where teh biggest bang for your buck is on the master, probably all of /opt/zenoss

[29-Mar-2010 10:17:28] <rmatte> yeh

[29-Mar-2010 10:18:05] <rmatte> In the interest of UI responsiveness, doing MySQL on the master is a good idea

[29-Mar-2010 10:18:44] <cluther> rmatte: Same to you.

[29-Mar-2010 10:21:08] <venturaville> we do all of /opt/zenoss

[29-Mar-2010 10:21:43] <venturaville> mysql on the master tends to eat into your CPU quite a bit

[29-Mar-2010 10:22:09] <venturaville> I'm not sure that you would get back in latency between the 2 what you put into it by putting it on the same box as the UI

[29-Mar-2010 10:22:50] <rmatte> possibly

[29-Mar-2010 10:23:14] <ckrough> it ate too much cpu, having on the same gig switch on its own server is good

[29-Mar-2010 10:25:42] <venturaville> the biggest problem we have is with the UI

[29-Mar-2010 10:26:02] <venturaville> the collectors can keep up, the event database can keep up

[29-Mar-2010 10:26:13] <venturaville> but it seems like zenhub is a giant bottleneck

[29-Mar-2010 10:26:29] <venturaville> especially if you are running workers and don't have /tmp on tmpfs or something

[29-Mar-2010 10:26:50] <venturaville> from doing an strace on one of the workers, it looks like it is passing the entire zope database back and forth

[29-Mar-2010 10:27:47] <venturaville> when I edit a global template everything just pauses for a minute

[29-Mar-2010 10:28:27] <rmatte> yeh, they have been changing code in the other daemons to try and put less stress on zenhub

[29-Mar-2010 10:28:47] <rmatte> such as having zensyslog actually do some processing rather than just straight feeding everything to zenhub for processing

[29-Mar-2010 10:28:50] <rmatte> same with zentrap

[29-Mar-2010 10:29:28] <rmatte> I don't think the changes made it in to 2.5.2 but they'll be in stone crab

[29-Mar-2010 10:29:35] <venturaville> I keep meaning to put in a feature request for a hook into the preparser to clean up events before they make it into the system

[29-Mar-2010 10:29:48] <ckrough> venturaville: I was looking at mine and I dont see FIFOs in /tmp related to hubs

[29-Mar-2010 10:29:54] <venturaville> an event flood can take us out

[29-Mar-2010 10:30:06] <rmatte> venturaville: yeh, I can imagine

[29-Mar-2010 10:30:18] <venturaville> ckrough: how many workers are you running

[29-Mar-2010 10:30:20] <ckrough> venturaville: are you running a single hub with multiple workers? I've moved to multiple hubs with multiple workers... seems better

[29-Mar-2010 10:30:29] <venturaville> you can tell where it is writing to by doing ps -aelfw | grep worker

[29-Mar-2010 10:31:14] <venturaville> the -C /tmp/tmp@@@@@ thing is where it is going back and forth to

[29-Mar-2010 10:31:23] <ckrough> a total of 8 workers across 5 hubs

[29-Mar-2010 10:31:31] <venturaville> ckrough: how are you coalescing that setup?

[29-Mar-2010 10:31:38] <ckrough> gotcha, I am probably going to ramdisk the /tmp then

[29-Mar-2010 10:31:42] <ckrough> hard to show in IRC

[29-Mar-2010 10:31:52] <ckrough> when I get this upgrade done Ill draw it in pastebin for ya

[29-Mar-2010 10:31:55] <venturaville> k

[29-Mar-2010 10:32:03] <venturaville> we need a single UI.....

[29-Mar-2010 10:32:52] <ckrough> YES, been asking for a unified 2-way gui

[29-Mar-2010 10:33:07] <venturaville> btw zenrender likes to write to /tmp too .....

[29-Mar-2010 10:33:20] <rmatte> yeh, it does

[29-Mar-2010 10:33:26] <ckrough> I notices it copies its config there, but havent done much lsof poking

[29-Mar-2010 10:34:20] <venturaville> you might not see it in lsof

[29-Mar-2010 10:34:37] <venturaville> it opens it, writes it, and immediately deletes it (unless you have caching turned on)

[29-Mar-2010 10:34:38] <ckrough> blasphemy!

[29-Mar-2010 10:34:41] <venturaville> strace will show it

[29-Mar-2010 10:35:00] <venturaville> i.e. it happens too fast

[29-Mar-2010 10:36:02] <rmatte> strace is godly

[29-Mar-2010 10:37:10] <venturaville> +1

[29-Mar-2010 10:37:50] <ckrough> have you played with SystemTap or OProfile?

[29-Mar-2010 10:39:13] <venturaville> oprofile yes

[29-Mar-2010 10:40:01] <venturaville> not systemtap

[29-Mar-2010 10:40:28] <venturaville> moving rooms ...losing my connection for a bit

[29-Mar-2010 11:51:37] <etank> anyone in here creating bar graphs with zenoss instead of just line or area?

[29-Mar-2010 11:52:38] <etank> im looking for a way to graph how long an event took to happen having the X axis be the amount of time and the Y axis be the time of day that the event happened

[29-Mar-2010 11:54:22] <rmatte> well, what do you mean by "how long"

[29-Mar-2010 11:54:30] <rmatte> you mean the duration between start time and end time?

[29-Mar-2010 11:54:43] <etank> a backup happened at 9:30am and took 47 minutes

[29-Mar-2010 11:54:52] <etank> X = 47 Y = 9:30am

[29-Mar-2010 11:54:58] <rmatte> and how is Zenoss going to know that it took 47 minutes?

[29-Mar-2010 11:55:20] <rmatte> the Y axis is always time

[29-Mar-2010 11:55:21] <etank> i can write a sql query to pull the data from the database

[29-Mar-2010 11:55:36] <rmatte> or sorry, the x axis rather

[29-Mar-2010 11:55:59] <rmatte> ok, so just make a command based datasource that runs that command and passes the value back to Zenoss

[29-Mar-2010 11:56:00] <etank> rmatte: one sec and i will get a screen shot of what i am trying to duplicate in zenoss

[29-Mar-2010 11:56:20] <rmatte> the command will run every 5 minutes and graph the data

[29-Mar-2010 11:56:27] <etank> it may not even have to be a graph but we do need to display the data somehow

[29-Mar-2010 11:56:44] <etank> we are trying to get rid of one piece of software and roll the funcion into zenoss

[29-Mar-2010 11:56:46] <rmatte> as far as doing a bar graph, I've never seen that done

[29-Mar-2010 11:56:57] <rmatte> you'd have to mess around with the RRDtool values

[29-Mar-2010 11:57:14] <rmatte> (I'm not sure RRD supports bar graphs, but it should)

[29-Mar-2010 11:57:48] <rmatte> ah

[29-Mar-2010 11:57:49] <rmatte> http://forums.cacti.net/post-64224.html

[29-Mar-2010 11:58:00] <rmatte> "This is a NOGO. cacti is an rrdtool frontend. And rrdtool does not provide real bar charts. "

[29-Mar-2010 11:58:44] <rmatte> Zenoss uses rrdtool as well

[29-Mar-2010 11:58:52] <rmatte> bar graphs aren't possible

[29-Mar-2010 11:58:57] <etank> rmatte: http://img265.imageshack.us/img265/3444/grapha.jpg  <-- something that is what i am trying to relace

[29-Mar-2010 11:59:16] <rmatte> yeh, you can't make a bar graph like that in Zenoss

[29-Mar-2010 11:59:22] <etank> :/

[29-Mar-2010 11:59:27] <rmatte> rrdtool doesn't support that

[29-Mar-2010 11:59:52] <etank> any idea how i could use zenoss to present data like that? even if not in a graph

[29-Mar-2010 12:00:16] <rmatte> store the duration as events

[29-Mar-2010 12:00:22] <ckrough> you may be able to feed rrd the total time at the end of every cycle as a guage

[29-Mar-2010 12:00:43] <ckrough> it would just look like a gappy line graph, but it may represent the data the way you want

[29-Mar-2010 12:01:17] <ckrough> they would be aligned with the end of the job rather than the beginning, since rrd cant see into the future

[29-Mar-2010 12:01:23] <rmatte> yeh, you can use a command based datasource as I suggested earlier, gauge type

[29-Mar-2010 12:01:27] <rmatte> and feed it in and graph it

[29-Mar-2010 12:01:37] <rmatte> though it may not necessarilly looks exactly as you expect

[29-Mar-2010 12:01:41] <rmatte> but it should be decent enough

[29-Mar-2010 12:02:08] <mrayzenoss> etank: docs/DOC-4538 does a MySQL query and graphs it

[29-Mar-2010 12:02:16] <etank> ok. i will look into that then

[29-Mar-2010 12:02:31] <mrayzenoss> it's against the Zenoss DB, but it's just a command datasource

[29-Mar-2010 12:03:40] <rmatte> yeh, it's basically the same thing that you'll want

[29-Mar-2010 12:09:16] <rmatte> mrayzenoss: any idea why the datapoint rate graph for zenprocess looks like this in 2.5.2: http://i39.tinypic.com/2i97dsh.jpg ?

[29-Mar-2010 12:09:33] <rmatte> mrayzenoss: the same install that I upgraded from which is still running 2.4.5 doesn't have a zenprocess graph like that

[29-Mar-2010 12:09:42] <ckrough> lol Ive never seen that rmatte

[29-Mar-2010 12:09:44] <rmatte> also notice the datapoint dropoffs which correspond with it

[29-Mar-2010 12:09:44] <mrayzenoss> dunno, I'll ask

[29-Mar-2010 12:10:27] <mrayzenoss> hmm… I've got the same thing on one of my installs

[29-Mar-2010 12:10:30] <rmatte> It doesn't seem to affect the actual graphing of performance data for processes as far as I know, but it's quite strange

[29-Mar-2010 12:14:53] <mrayzenoss> rmatte: http://dev.zenoss.com/trac/ticket/6092

[29-Mar-2010 12:15:24] <rmatte> ah

[29-Mar-2010 12:17:21] <rmatte> k cool, as long as it's not going to affect actual data collection, no big deal

[29-Mar-2010 12:22:09] <rmatte> other than that, 2.5.2 has been running really well

[29-Mar-2010 12:23:20] <rmatte> viewing event history in the new event console is a bit tempermental

[29-Mar-2010 12:34:49] <rmatte> selecting the time for first seen and last seen sometimes doesn't reload the lis

[29-Mar-2010 12:34:51] <rmatte> list*

[29-Mar-2010 12:35:00] <rmatte> so I then have to reload the entire event console and try again

[29-Mar-2010 12:35:46] <rmatte> annoying and time consuming

[29-Mar-2010 12:37:48] <eidolon> hey folks - we have a zenoss server that's getting sort of sluggish.  it has about 300 hosts being monitored in it, 4gig of RAM, anda  single Xeno 2.5gig CPU.  loads don't spike much, but a suggestion has come up to up the ram.  4gig seems like a lot though.  suggestions for performance tuning a largeish Zenoss install?

[29-Mar-2010 12:38:21] <rmatte> eidolon: docs/DOC-2521

[29-Mar-2010 12:38:48] <rmatte> you say 300 devices, but how many data points?

[29-Mar-2010 12:39:01] <rmatte> you could have 300 devices with tons of datapoints or with almost none

[29-Mar-2010 12:39:35] <mrayzenoss> there's an 80K device install will about 240K datapoints

[29-Mar-2010 12:39:35] <eidolon> we have a lot of datapoints :)  is there a simple way to count?

[29-Mar-2010 12:40:14] <mrayzenoss> Collectors->localhost->performance

[29-Mar-2010 12:40:14] <rmatte> yes, very simple way

[29-Mar-2010 12:40:19] <rmatte> ^what he said

[29-Mar-2010 12:40:24] <rmatte> bottom graph

[29-Mar-2010 12:40:26] <eidolon> k, sec

[29-Mar-2010 12:40:30] <rmatte> add up the values from the different daemons

[29-Mar-2010 12:40:35] <rmatte> k means thousand

[29-Mar-2010 12:40:48] <eidolon> hmm.

[29-Mar-2010 12:41:19] <eidolon> http://www.screencast.com/users/eidolon/folders/Jing/media/52e6be62-71ce-435e-86eb-e43d332801dc

[29-Mar-2010 12:41:22] <eidolon> 5000?

[29-Mar-2010 12:41:37] <rmatte> about 5200

[29-Mar-2010 12:41:41] <rmatte> but yeh

[29-Mar-2010 12:41:50] <rmatte> that's almost nothing, so 4GB should be plenty of RAM

[29-Mar-2010 12:42:16] <eidolon> mmkay...  we're getting enormous delays on clicks.  like, search for a device, and it takes 20-30 seconds to come back

[29-Mar-2010 12:42:18] <rmatte> Be aware that Zenoss does have some memory leaks, I generally have to restart some of the daemons once in a while to get the memory usage back down

[29-Mar-2010 12:42:29] <mrayzenoss> eidolon: LDAP?

[29-Mar-2010 12:42:33] <eidolon> nosir, local

[29-Mar-2010 12:42:44] <rmatte> follow that tweaking guide

[29-Mar-2010 12:43:00] <rmatte> tuning the MySQL and Zope settings will help with the response time

[29-Mar-2010 12:43:03] <eidolon> rest of that performance page:  http://www.screencast.com/users/eidolon/folders/Jing/media/83b0d055-17e9-451c-926b-94897d22ab32

[29-Mar-2010 12:43:07] <rmatte> more zenhub workers will also help

[29-Mar-2010 12:43:34] <rmatte> those values look fine

[29-Mar-2010 12:43:34] <eidolon> k

[29-Mar-2010 12:44:12] * eidolon types a name into device /ip search

[29-Mar-2010 12:44:15] * eidolon is waiting for an answer

[29-Mar-2010 12:44:17] <eidolon> 14 seconds.

[29-Mar-2010 12:44:51] <rmatte> 14 seconds is fairly long, I have a box with almost 17000 datapoints, and it only takes a couple of seconds

[29-Mar-2010 12:44:59] <rmatte> but it's all tweaked up as that page describes

[29-Mar-2010 12:45:36] <rmatte> stock Zenoss Core can be quite slow, you need those tweaks if you intend to handle larger numbers of devices

[29-Mar-2010 12:46:09] <eidolon> we have itself being monitored

[29-Mar-2010 12:46:09] <eidolon> http://www.screencast.com/users/eidolon/folders/Jing/media/0af0a1c1-2537-4a7f-a9de-beb2b4feaebb

[29-Mar-2010 12:46:19] <eidolon> i'm a little worried about that memory display :)

[29-Mar-2010 12:46:44] <rmatte> how long has your Zenoss been running for?

[29-Mar-2010 12:46:51] <rmatte> go to Settings -> Versions

[29-Mar-2010 12:47:04] <rmatte> What does it show for zope uptime?

[29-Mar-2010 12:47:19] <eidolon> Zope16 days 20 hours 51 min 30 sec

[29-Mar-2010 12:47:25] <rmatte> try this...

[29-Mar-2010 12:47:38] <rmatte> login to the box, become the zenoss user, and do zenperfsnmp restart

[29-Mar-2010 12:47:49] <rmatte> then keep an eye on the memory

[29-Mar-2010 12:47:52] <rmatte> you should see it drop

[29-Mar-2010 12:47:58] <eidolon> k

[29-Mar-2010 12:48:20] <rmatte> zenperfsnmp has a memory leak, and Zope seems to have one as well, but not as bad as zenperfsnmp

[29-Mar-2010 12:48:42] <rmatte> I'm debating croning restarts of it every 10 days or so

[29-Mar-2010 12:48:47] <eidolon> ok

[29-Mar-2010 12:49:03] <eidolon> restarted.

[29-Mar-2010 12:49:23] <eidolon> this is zenoss 2.4.5 btw

[29-Mar-2010 12:49:43] <eidolon> spikey load average

[29-Mar-2010 12:50:19] <nonsenso> rmatte: what version are you running?

[29-Mar-2010 12:53:37] <rmatte> eidolon: my servers are running 2.4.5

[29-Mar-2010 12:53:45] <rmatte> eidolon: and I have 1 upgraded to 2.5.2 as a test

[29-Mar-2010 12:54:24] * eidolon nods

[29-Mar-2010 12:54:34] <eidolon> how's 2.5.x going?  last check it was still unstable.

[29-Mar-2010 12:54:41] <rmatte> working fine

[29-Mar-2010 12:54:51] <rmatte> 2.5.2 is pretty stable

[29-Mar-2010 12:55:02] <rmatte> the new event console is a bit temperamental though

[29-Mar-2010 12:55:17] <rmatte> mostly when viewing event history

[29-Mar-2010 13:06:38] <Brixius> If I had hair it would be pulled out by now... I have an event that I moved from /Change/Set/Status to /Ignore.  Then restarted zenoss 2.5.2 and yet the new events goto the old eventclass...

[29-Mar-2010 13:08:10] <rmatte> I assume you mean that you moved the event mapping?

[29-Mar-2010 13:08:18] <Brixius> yep

[29-Mar-2010 13:08:28] <rmatte> and how did you go about doing that?

[29-Mar-2010 13:09:39] <Brixius> went into classes/events/change/set/status and clicked on the mapping.  Then selected move mapping from the menu and selected /Ignore

[29-Mar-2010 13:16:58] <eidolon> hmm, the comment in the doc says 'workers 3' is optimized for 4cores.  this is a single core host.

[29-Mar-2010 13:17:12] <Brixius> Nevermind, I think It's working now, kindof fustrating though....  I found one mapping I missed moving...

[29-Mar-2010 13:17:21] <Brixius> so my bad

[29-Mar-2010 13:17:41] <rmatte> eidolon: ah, single core can be a bit rough performancewise on a server

[29-Mar-2010 13:17:50] <rmatte> that's probably part of the reason for your response time issues

[29-Mar-2010 13:18:13] * eidolon nods

[29-Mar-2010 13:18:23] <rmatte> either way, skip the workers part then

[29-Mar-2010 13:18:34] <eidolon> mmk

[29-Mar-2010 13:18:35] <rmatte> do the rest of the steps

[29-Mar-2010 13:22:47] <eidolon> okay all done, restarting zenoss and mysql

[29-Mar-2010 13:27:02] <rmatte> cool

[29-Mar-2010 13:33:46] <Brixius> Ok apparently I lied, While the event mapping defination I wanted to move, moved the event is still going to the old event class, even though there is nothing mapped to that class under Events/Mapping tab

[29-Mar-2010 13:34:19] <Brixius> Wait, I didn't type that well, so it doesn't read well.

[29-Mar-2010 13:35:11] <rmatte> that doesn't even make sense, are you sure that there's nothing in the old class if you go to More -> Transform in the dropdown menu?

[29-Mar-2010 13:35:23] <Brixius> I moved an event mapping from /Change/Set/Status to /Ignore, and checked and the Mapping Defination did move to /Ignore.

[29-Mar-2010 13:35:34] <rmatte> the only reason the event should be coming in with the old class is if a mapping of some kind exists

[29-Mar-2010 13:35:53] <Brixius> However, I still have the events going to the old class.

[29-Mar-2010 13:36:20] <rmatte> you're sure you didn't accidentally copy instead of move?

[29-Mar-2010 13:36:35] <rmatte> you're sure there's no leftover mapping in the old class?

[29-Mar-2010 13:36:53] <Brixius> yep, went back and checked ant they are not listed under /change/set/status

[29-Mar-2010 13:37:09] <rmatte> what does the mapping look like?

[29-Mar-2010 13:37:15] <rmatte> what are the properties?

[29-Mar-2010 13:40:15] <rmatte> also, how is the event being created?

[29-Mar-2010 13:40:24] <rmatte> snmp polling, trap, command datasource?

[29-Mar-2010 13:40:38] <Brixius> Name: "cpqIdeController" Event Class Key: "cpqIdeController" Rule: "evt.component == 'cpqIdeController0'"

[29-Mar-2010 13:41:00] <Brixius> Agent is zenperfsnmp

[29-Mar-2010 13:41:19] <rmatte> ok, is the event a threshold event?

[29-Mar-2010 13:41:31] <Brixius> yes

[29-Mar-2010 13:41:35] <rmatte> then that's your problem

[29-Mar-2010 13:41:47] <rmatte> find the performance template that is polling for that

[29-Mar-2010 13:41:51] <rmatte> then click on the threshold

[29-Mar-2010 13:41:57] <Brixius> ok

[29-Mar-2010 13:42:02] <rmatte> and change the event class for the threshold

[29-Mar-2010 13:42:11] <Brixius> ok

[29-Mar-2010 13:42:13] <rmatte> thresholds are handled much differently than other events

[29-Mar-2010 13:42:19] <Brixius> ahh, ok

[29-Mar-2010 13:46:16] <Brixius> I think it's a collector plugin, from the Zenpacks.community.HPMon  I don't see a performance template under the device for it.

[29-Mar-2010 13:56:52] <Brixius> Found it, :)

[29-Mar-2010 14:00:19] <rmatte> cool

[29-Mar-2010 14:02:14] <Brixius> I'm slowly catching on, still have a device that sends me an ifindex for a linkdown event that I would like to resolve to a name, any ideas how I'd go about doing that in a transform?

[29-Mar-2010 14:08:04] cluther_ is now known as cluther

[29-Mar-2010 14:15:51] <rmatte> Brixius: you really should read up on transforms on the community site

[29-Mar-2010 14:15:57] <rmatte> and also read up on zendmd

[29-Mar-2010 14:16:13] <rmatte> there's more than enough material floating around for you to figure that out

[29-Mar-2010 14:16:20] <Brixius> ok

[29-Mar-2010 14:16:24] <rmatte> it's the whole "teach a man to fish" scenario

[29-Mar-2010 14:16:35] <rmatte> I could write the transform for you, but it's not going to help you

[29-Mar-2010 14:17:12] <rmatte> also, read the Zenoss Admin guide in it's entirety if you haven't

[29-Mar-2010 14:17:54] <rmatte> It'll explain about 90% of the stuff you're likely to ask in here

[29-Mar-2010 14:20:20] <rmatte> Just a hint, if you click on an interface in Zenoss you'll notice that Zenoss is aware of both the ifindex and the name

[29-Mar-2010 14:20:36] <rmatte> so you can make use of that in the transform

[29-Mar-2010 14:27:03] <Brixius> I get that and saw that a while ago, in fact I was exploring zendmd and was able to get the ifname, but when I put what I thought should work in a transform it wouldn't give me anything, so I'm obviously doing something wrong but can't track down what.  I've sense deleted what I had in there, I'll see what I can come up with.

[29-Mar-2010 14:33:59] <rmatte> k

[29-Mar-2010 14:34:10] <rmatte> once you come up with something I'll be happy to help you debug it

[29-Mar-2010 14:52:27] <QbY> Can someone take a look at this transform and tell me why it doesn't wanna work for me: ) http://pastebin.com/jnPRdZDP

[29-Mar-2010 14:52:42] <venturaville> ckrough: did you ever get that diagram put together?

[29-Mar-2010 14:53:05] <rmatte> because you can't do dev = evt.device

[29-Mar-2010 14:53:08] <rmatte> you need to do...

[29-Mar-2010 14:53:18] <rmatte> dev = dmd.Devices.findDevice(evt.device)

[29-Mar-2010 14:53:55] <Brixius> Ah ha, that's what I was doing wrong as well I think...

[29-Mar-2010 14:54:12] <rmatte> lol

[29-Mar-2010 14:54:32] <rmatte> QbY: also...

[29-Mar-2010 14:54:42] <rmatte> you're doing for interface in dev.os.interfaces():

[29-Mar-2010 14:54:49] <rmatte> but you're doing tModems = int.getRRDValue('TotalModems')

[29-Mar-2010 14:54:55] <rmatte> int is not going to work

[29-Mar-2010 14:55:06] <rmatte> need to be: tModems = interface.getRRDValue('TotalModems')

[29-Mar-2010 14:55:10] <QbY> ok..  i need to get teh RRD values from the interface that is alarming

[29-Mar-2010 14:55:20] <QbY> so scratch the for

[29-Mar-2010 14:55:33] <rmatte> the for loop will work fine if I you do it properly

[29-Mar-2010 14:55:41] <venturaville> cacheRRDValue ... if you can....

[29-Mar-2010 14:56:06] <rmatte> change int.getRRDValue to interface.getRRDValue and it'll work

[29-Mar-2010 14:56:16] <QbY> hit save, and it went red

[29-Mar-2010 14:56:25] <rmatte> repaste what you have now

[29-Mar-2010 14:56:42] <rmatte> there's probably some stupid little syntax error somewhere

[29-Mar-2010 14:56:49] <QbY> http://pastebin.com/k7A2cfAj

[29-Mar-2010 14:56:58] <QbY> i'm in a rush because its happening live

[29-Mar-2010 14:57:14] <rmatte> your spacing is messed

[29-Mar-2010 14:57:23] <rmatte> you have an extra space in front of the evt.message line

[29-Mar-2010 14:57:29] <rmatte> Python is extremely particular about spacing

[29-Mar-2010 14:57:34] <rmatte> each level should be 4 spaces

[29-Mar-2010 14:57:45] <rmatte> (it can be a single space but 4 is best practice)

[29-Mar-2010 14:57:58] <rmatte> oh also

[29-Mar-2010 14:58:04] <QbY>          threshold of MassDeregWarning not met: current value 0.00 ...  still giving me the standard message

[29-Mar-2010 14:58:06] <rmatte> evt.message = "Interface: %s Registered: %s Expected Total: %s" % (int, rModems, tModems)

[29-Mar-2010 14:58:10] <rmatte> change int to interface

[29-Mar-2010 14:58:11] <rmatte> in that

[29-Mar-2010 14:58:29] <rmatte> well, where are you putting the transform?

[29-Mar-2010 14:59:10] <QbY> i go to events, docsis / massdreg (my own class) and then hit more / transform

[29-Mar-2010 14:59:28] <rmatte> is that the class that the threshold is coming in for?

[29-Mar-2010 15:00:06] <QbY> yes

[29-Mar-2010 15:00:48] <rmatte> did you fix the spacing and change int to interface on the last line?

[29-Mar-2010 15:00:56] <QbY> yes

[29-Mar-2010 15:00:58] <QbY> posting new

[29-Mar-2010 15:01:06] <rmatte> k

[29-Mar-2010 15:01:34] <QbY> http://pastebin.com/xi9feFQ1

[29-Mar-2010 15:01:59] <rmatte> give me a minute...

[29-Mar-2010 15:03:22] <QbY> i'm giving zenoss a minute too.. should hit it again in just about 1 minutes

[29-Mar-2010 15:04:02] Guest16351 is now known as ckrough

[29-Mar-2010 15:06:02] <rmatte> http://pastebin.com/Nb9WYu2C

[29-Mar-2010 15:06:04] <rmatte> try that

[29-Mar-2010 15:06:30] <rmatte> what version of Zenoss by the way?

[29-Mar-2010 15:09:37] <QbY> 2.5

[29-Mar-2010 15:09:49] <rmatte> 2.5.2?

[29-Mar-2010 15:09:54] <QbY> yep

[29-Mar-2010 15:09:56] <rmatte> k

[29-Mar-2010 15:11:12] <QbY> this transform will apply on the next one even though its acknolwedged?

[29-Mar-2010 15:11:34] <rmatte> yes, it should come in as a different event

[29-Mar-2010 15:11:43] <rmatte> but you might as well just move the existing one to history

[29-Mar-2010 15:12:07] <rmatte> you did set that threshold to come in under that event class right?

[29-Mar-2010 15:12:12] <QbY> yes

[29-Mar-2010 15:12:14] <rmatte> k

[29-Mar-2010 15:12:20] <rmatte> no reason why it shouldn't work now then

[29-Mar-2010 15:12:49] <QbY> Interface:  Registered: 45.8 Expected Total: 51.0

[29-Mar-2010 15:13:09] <rmatte> oh, that's easy

[29-Mar-2010 15:13:15] <QbY> hrmm

[29-Mar-2010 15:13:21] <rmatte> change interface on the last line to interface.getInterfaceName()

[29-Mar-2010 15:13:25] <QbY> hrmmm

[29-Mar-2010 15:13:31] <rmatte> then it'll work fine

[29-Mar-2010 15:13:39] <QbY> that should be 146 total and 146 registered

[29-Mar-2010 15:13:41] <QbY> on c3

[29-Mar-2010 15:14:05] <rmatte> well, those are the current RRD values that it pulled

[29-Mar-2010 15:14:19] <QbY> hrmmm

[29-Mar-2010 15:15:47] <rmatte> change interface on the last line to interface.getInterfaceName() so that it at least shows the right interface name

[29-Mar-2010 15:16:03] <QbY> is ther e away to force a check of the threshold

[29-Mar-2010 15:16:17] <rmatte> no, it has to kick off

[29-Mar-2010 15:16:23] <rmatte> well

[29-Mar-2010 15:16:23] <QbY> k

[29-Mar-2010 15:16:30] <rmatte> you could try: zenperfsnmp run

[29-Mar-2010 15:16:34] <rmatte> that might do it

[29-Mar-2010 15:16:41] <rmatte> (as the zenoss user of course)

[29-Mar-2010 15:16:42] <QbY> i will wait

[29-Mar-2010 15:16:56] <rmatte> k

[29-Mar-2010 15:17:02] <rmatte> afk for a few minutes

[29-Mar-2010 15:17:12] <QbY> k

[29-Mar-2010 15:17:55] <QbY> is it safe to delete from mysql

[29-Mar-2010 15:17:59] <QbY> the history table in mysql

[29-Mar-2010 15:19:16] <QbY>          Interface: Cable3/0-upstream2 Registered: 0.0 Expected Total: 105.0

[29-Mar-2010 15:19:19] <QbY> BOOYAH!

[29-Mar-2010 15:19:22] <QbY> thanks rmatte

[29-Mar-2010 15:33:54] <rmatte> np\

[29-Mar-2010 15:34:11] <rmatte> Why would you delete it from MySQL?

[29-Mar-2010 15:35:51] <mrayzenoss> you probably don't want to delete the table, maybe just the contents?

[29-Mar-2010 15:37:20] <QbY> i just  want to delete/update some of the transforms..  because they were showing the wrong thing..

[29-Mar-2010 15:37:28] <QbY> and some are bogus

[29-Mar-2010 15:37:36] <QbY> from when i set the threshold to a level that couldn't be met

[29-Mar-2010 15:39:07] <rmatte> well, if you know how to safely delete the entries then go for it

[29-Mar-2010 15:39:15] <rmatte> I personally wouldn't even bother

[29-Mar-2010 15:54:15] * QbY really loves Zenoss

[29-Mar-2010 15:56:08] <rmatte> yup, it's a nice product

[29-Mar-2010 16:03:13] <QbY> rmatte: on my interfaces sometimes i have nice "Destrctiptions" -- shows on the OS Page as "Descr" How can I get that in my transform?

[29-Mar-2010 16:03:25] <QbY> I

[29-Mar-2010 16:03:35] <QbY> I'm going to try and write a script that will call from our PBX

[29-Mar-2010 16:03:42] <QbY> text to speech for the summary--and then call

[29-Mar-2010 16:05:41] <rmatte> one second

[29-Mar-2010 16:06:26] <rmatte> instead of interface.getInterfaceName() you just do interface.description

[29-Mar-2010 16:16:58] Diddi_ is now known as Diddi

[29-Mar-2010 16:28:43] <rmatte> eugh, I've been noticing that some events that should be auto-clearing aren't auto-clearing lately, getting annoying

[29-Mar-2010 16:39:40] <rmatte> I'm out, later

[29-Mar-2010 18:31:19] forsberg is now known as fOrsberg

[29-Mar-2010 21:05:33] <theacolyte> popcorn

[29-Mar-2010 21:10:47] <mrayzenoss> night all

[29-Mar-2010 21:37:40] RobertLaptop_ is now known as RobertLaptop

[30-Mar-2010 00:00:30] [disconnected at Tue Mar 30 00:00:30 2010]

[30-Mar-2010 00:00:30] [connected at Tue Mar 30 00:00:30 2010]

[30-Mar-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[30-Mar-2010 03:41:54] <jaco> Hey Guys, hope you can help.  I'm really battling to get SNMP to work... I've followed all the tutorials I could find but I'm just not getting it up and running?!

[30-Mar-2010 08:40:55] <ckrough> anyone have a script for removing holt-winters from an rrd?

[30-Mar-2010 08:43:08] <mrayzenoss> ckrough: this is from the Community ZenPack: http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.PredictiveThreshold/RemoveHoltData.py

[30-Mar-2010 08:45:47] <ckrough> thanks

[30-Mar-2010 08:49:54] <ckrough> ahh zenping...

[30-Mar-2010 08:54:27] <ckrough> mrayzenoss: whats the svn repo address for that?

[30-Mar-2010 08:54:38] <ckrough> I dont see it in zenpacks under http://dev.zenoss.org/svn/trunk

[30-Mar-2010 08:55:36] <eidolon> hey folks - is there a way to put a device into a mode where all events are suppressed?  we thought that Maintenance mode would do that - but the events are still happening - they're not -alerting- (which is good), but our status board goes red, etc etc.

[30-Mar-2010 08:55:49] <ckrough> mrayzenoss: I guess I dont have the community zenpack repo in eclipse

[30-Mar-2010 09:00:56] <mrayzenoss> ckrough: svn co http://zenpacks.zenoss.org/svn/zenpacks

[30-Mar-2010 09:01:04] <mrayzenoss> or get the specific one you want

[30-Mar-2010 09:01:55] <ckrough> thanks

[30-Mar-2010 09:02:38] <ckrough> got it, thansk

[30-Mar-2010 09:19:13] * eidolon tugs on mrayzenoss's pantsleg.

[30-Mar-2010 09:19:54] <mrayzenoss> eidolon: Chapter 8 of the Admin Guide: Production States

[30-Mar-2010 09:20:08] * eidolon googles.

[30-Mar-2010 09:20:11] <mrayzenoss> You can toggle monitoring, alerting, dashboard

[30-Mar-2010 09:20:32] <eidolon> goodamn i hate scribd.

[30-Mar-2010 09:20:35] * Simon4 plays "why doesn't my switch collect from portchannel interfaces even though it's modelling them" after only 2 hours sleep

[30-Mar-2010 09:20:36] <mrayzenoss> so you want maintenance probably: monitoring with no alerts or dashboard

[30-Mar-2010 09:20:38] <Simon4> this isn't going well :/

[30-Mar-2010 09:20:44] <eidolon> i wish i could tell google to elminiate scribd from ALL my search results.

[30-Mar-2010 09:21:02] <eidolon> except our 'dashboard' is the event log

[30-Mar-2010 09:21:03] <mrayzenoss> eidolon: community/documentation/official_documentation/zenoss-guide

[30-Mar-2010 09:21:09] * eidolon finds the dashboard worthless, frankly.

[30-Mar-2010 09:21:20] <eidolon> thanks fro the link

[30-Mar-2010 09:21:23] <Simon4> eidolon: search for searchstring -inurl(scribd)

[30-Mar-2010 09:21:28] <mrayzenoss> eidolon: lots of admins just watch the event console

[30-Mar-2010 09:21:35] <eidolon> Simon4: i have to append that to all my searches.

[30-Mar-2010 09:21:42] <eidolon> mrayzenoss: then they are FOOLS!  I will CRUSH THEM!

[30-Mar-2010 09:21:42] <eidolon> i mean

[30-Mar-2010 09:21:46] <eidolon> "oh isn't that interesting."

[30-Mar-2010 09:21:47] <Simon4> eidolon: or make a firefox keyword search that does it for you

[30-Mar-2010 09:21:48] <Simon4> or similar

[30-Mar-2010 09:22:31] <eidolon> man this is NOT my day.  http://www.screencast.com/users/eidolon/folders/Jing/media/04489a9b-503a-4db6-932c-e238aa8f33a6

[30-Mar-2010 09:23:09] * eidolon tries again

[30-Mar-2010 09:23:17] <eidolon> crashed again.  awesome.

[30-Mar-2010 09:23:29] * eidolon is now just whining.

[30-Mar-2010 09:24:18] <eidolon> btw, this is our 'big board' for system monitoring and status:  http://twitpic.com/x3w2d  - we run the event console on the big screen with auto-refresh so we can see all the issues that are currently active.

[30-Mar-2010 09:24:25] <eidolon> it shows all unacknowledged alerts.

[30-Mar-2010 09:26:08] <rmatte> lol

[30-Mar-2010 09:26:33] <rmatte> the dashboard isn't "useless", but an event console portlet would probably be a nice addition

[30-Mar-2010 09:26:45] <rmatte> then you could have the event console and the google map, or whatever you want on the same screen

[30-Mar-2010 09:27:05] <eidolon> rmatte: i was a bit free with the hyperbole, perhaps - i find it useless :)

[30-Mar-2010 09:27:21] <rmatte> our "big screen" just cycles through google maps on each server since it's more meant to be a flashy thing to show touring clients rather than actually useful

[30-Mar-2010 09:27:30] <eidolon> "SOMETHING IS WRONG!" "what is? "can't tell ya.  just one of the machines in this group has thrown an error." "oh, good.  what error?" "Can't tell you that either."

[30-Mar-2010 09:27:59] <rmatte> that's when you take the initiative and login to the server to check :P

[30-Mar-2010 09:29:02] <rmatte> it would be nice if Zenoss had a dashboard similar to solarwinds

[30-Mar-2010 09:29:15] <rmatte> the only reason we use solarwinds is because of the flashy way it displays stats

[30-Mar-2010 09:29:45] <rmatte> and I'm stuck painfully commissioning 260 devices in to it apparently (was handed that assignment yesterday)

[30-Mar-2010 09:29:56] <rmatte> and I personally can't stand the administrative side of it's web interface

[30-Mar-2010 09:29:58] <rmatte> it is horrid

[30-Mar-2010 09:30:03] <eidolon> never used it.

[30-Mar-2010 09:30:11] <eidolon> most of my network monitoring has been in nagios and zenoss.

[30-Mar-2010 09:30:17] <rmatte> it has the stupidest and I mean STUPIDEST quirks...

[30-Mar-2010 09:30:25] <eidolon> and i'm still in therapy after maintaining a fairly large nagios install.

[30-Mar-2010 09:30:40] <rmatte> like to toggle an option, there's no checkbox or dropdown, it's literally an input text field, and you have to change "False" to "True" or vice versa

[30-Mar-2010 09:30:44] <rmatte> drives me insane

[30-Mar-2010 09:30:52] <rmatte> like, would it have been so hard to code dropdown menus instead?

[30-Mar-2010 09:30:54] <eidolon> ew.

[30-Mar-2010 09:31:22] <rmatte> yeh, I've heard Nagios can be quite fun

[30-Mar-2010 09:31:27] <rmatte> I've never really used it

[30-Mar-2010 09:31:42] <rmatte> I messed around with groundworks for about 5 minutes one time

[30-Mar-2010 09:32:36] <eidolon> it's a maze of twisty little configuration files.

[30-Mar-2010 09:32:41] <eidolon> you spend a lot of time in vim.

[30-Mar-2010 09:32:53] <Simon4> "fun"

[30-Mar-2010 09:32:57] <Simon4> for some values of fun

[30-Mar-2010 09:33:00] <Simon4> that's for sure

[30-Mar-2010 09:33:19] <eidolon> where 'fun' in this case is on a par with repeatedly hitting yourself in the head with a stick.

[30-Mar-2010 09:33:56] <aclark> hah

[30-Mar-2010 09:42:14] <rmatte> mrayzenoss: any idea where the default dashboard layout is stored?

[30-Mar-2010 09:42:28] <rmatte> I want to be able to configure a default layout for users on all of my Zenoss servers

[30-Mar-2010 09:42:40] <mrayzenoss> that's a good question, I'd like to know that myself

[30-Mar-2010 09:42:44] * rmatte hopes it's not hardcoded somewhere

[30-Mar-2010 09:42:55] <rmatte> I know where it's stored in Zope per user

[30-Mar-2010 09:43:02] <rmatte> but it would be great to be able to configure a default

[30-Mar-2010 09:43:03] * mrayzenoss assumes it's hardcoded somewhere

[30-Mar-2010 09:43:23] <rmatte> even if it is hardcoded somewhere, if it could be changed it'd be good

[30-Mar-2010 09:43:28] <mrayzenoss> right

[30-Mar-2010 09:43:41] * Simon4 knows where it is

[30-Mar-2010 09:43:43] <Simon4> please hold

[30-Mar-2010 09:43:47] <rmatte> ty

[30-Mar-2010 09:44:38] <rmatte> when our clients login to Zenoss I'm sure they won't want to see the site portlet with guides on using Zenoss :P

[30-Mar-2010 09:45:05] <Simon4> dmd.ZenUsers._getOb('username').dashboardState

[30-Mar-2010 09:45:08] <Simon4> tis a string

[30-Mar-2010 09:45:38] <rmatte> yeh, but that's only for each user

[30-Mar-2010 09:45:38] <Simon4> or you can just use  dmd.ZenUsers.shelson.dashboardState if your username doesn't have special chars, but less safe

[30-Mar-2010 09:45:45] <rmatte> I'm talking about the global setting

[30-Mar-2010 09:45:48] <Simon4> oh gah, sorry

[30-Mar-2010 09:45:50] <rmatte> where that gets defined from by default

[30-Mar-2010 10:01:43] <mrayzenoss> according to cluther: Ok.. after actually looking it up. In the "restoreDefaults" method in $ZENHOME/Products/ZenWidgets/skins/zenui/javascript/portlet.js

[30-Mar-2010 10:15:35] <rmatte> mrayzenoss: cool

[30-Mar-2010 10:27:32] <rmatte> mrayzenoss: I just sent you the eggified Juniper pack, you should be receiving it shortly.

[30-Mar-2010 10:27:59] fOrsberg is now known as forsberg

[30-Mar-2010 10:28:04] <Simon4> rmatte: juniper switch? or firewall?

[30-Mar-2010 10:29:20] <rmatte> Netscreen firewall

[30-Mar-2010 10:56:35] <hav0x> question about multigraphs

[30-Mar-2010 10:56:41] <hav0x> is this a help channel?

[30-Mar-2010 10:56:46] <hav0x> =P

[30-Mar-2010 11:00:42] <mrayzenoss> we try to be helpful

[30-Mar-2010 11:09:04] <willwh|afk> mrayzenoss: you are

[30-Mar-2010 11:09:09] willwh|afk is now known as willwh

[30-Mar-2010 11:09:59] <mrayzenoss> hav0x: http://www.youtube.com/watch?v=qME_7Zh4yOc&feature=player_embedded

[30-Mar-2010 11:14:03] <rmatte> haha, youtube videos now, nice

[30-Mar-2010 11:23:12] <hav0x> thx mrayzenoss

[30-Mar-2010 11:23:15] <hav0x> checking it out

[30-Mar-2010 11:29:07] <willwh> mrayzenoss: screw that :)

[30-Mar-2010 11:29:10] <willwh> youtube is horrid

[30-Mar-2010 11:29:13] <willwh> come use us :D

[30-Mar-2010 11:29:17] <willwh> netromedia.com

[30-Mar-2010 11:29:28] <willwh> 1gb delivery per month if you need it

[30-Mar-2010 11:30:49] <hav0x> most usefull mrayzenoss

[30-Mar-2010 11:30:57] <mrayzenoss> hav0x: glad to hear it

[30-Mar-2010 11:31:15] <hav0x> my problem was with naming

[30-Mar-2010 11:31:32] <hav0x> uppercase apparently matters in grphpoint names

[30-Mar-2010 11:31:33] <mrayzenoss> willwh: we just use what we can embed on this page: docs/DOC-2606  No favorites really

[30-Mar-2010 11:31:44] <willwh> :)

[30-Mar-2010 11:31:59] <willwh> we have a commercial version of flowplayer

[30-Mar-2010 11:32:04] <willwh> licensed for re-use

[30-Mar-2010 11:32:17] <willwh> embed code ready to go if you set your service to be publicly listed

[30-Mar-2010 11:32:27] <willwh> if you want to give it a whirl

[30-Mar-2010 11:32:28] <willwh> let me know

[30-Mar-2010 11:32:40] <willwh> new portal build coming out this morning too

[30-Mar-2010 11:33:44] <willwh> if youtube suits you fine, whatever :)

[30-Mar-2010 11:33:50] <willwh> some people don't like branded players etc...

[30-Mar-2010 11:33:56] <willwh> just thought I'd offer.

[30-Mar-2010 13:19:38] <cgibbons> hmmm

[30-Mar-2010 13:27:58] <Zenethian> mmhmm

[30-Mar-2010 13:29:08] <rmatte> hmmmm

[30-Mar-2010 13:31:50] <ckrough> hmmmm

[30-Mar-2010 13:33:14] <mrayzenoss> fuhhhmmnn

[30-Mar-2010 13:33:52] <rmatte> got my new dedicated server today, nice download speed...

[30-Mar-2010 13:33:54] <rmatte> 95,545,644  10.6M/s   in 9.0s

[30-Mar-2010 13:35:20] <rmatte> venturaville: aloha

[30-Mar-2010 13:35:31] <venturaville> guten tag

[30-Mar-2010 13:44:24] <Zenethian> where'd you get one from?

[30-Mar-2010 13:50:24] <rmatte> sagonet

[30-Mar-2010 13:50:34] <rmatte> had it setup in their Atlanta datacenter

[30-Mar-2010 13:50:51] <venturaville> ?

[30-Mar-2010 13:51:05] <rmatte> venturaville: got my dedicated server that I ordered

[30-Mar-2010 13:51:18] <venturaville> how much a month?

[30-Mar-2010 13:51:23] <rmatte> 109

[30-Mar-2010 13:52:06] <rmatte> 2.4GHz, 4GB RAM, 2x500GB in RAID 1, 10Mbps port with 3TB monthly data throughput

[30-Mar-2010 13:52:33] * eidolon has a linode right now - pretty happy with it

[30-Mar-2010 13:52:56] <rmatte> I got it as one of their server specials on this month.

[30-Mar-2010 13:53:06] <rmatte> and paid a 1 time fee for the extra ram and second HD

[30-Mar-2010 13:53:19] <rmatte> so the setup fee was a bit pricy, but the monthly fee is great for what I'm getting

[30-Mar-2010 13:53:57] <venturaville> yeah I have played with the linode ones before, for the lower priced ones the performance was pretty bad though

[30-Mar-2010 13:54:27] <rmatte> I'm not a fan of VPS hosting for anything relatively intensive

[30-Mar-2010 13:54:39] <rmatte> I'm going to be running a bunch of stuff off of this server

[30-Mar-2010 13:54:43] <rmatte> so it's worth going dedicated

[30-Mar-2010 13:55:09] <venturaville> I'm paying $5 a month to rollernet so I can keep my server at home, and just have my mail forwarded in and out of AT&Ts SMTP ACL blockage...

[30-Mar-2010 13:56:04] <rmatte> I'm running a server at home right now, but I want to use it as a Zenoss development server, so I decided to get a dedicated one

[30-Mar-2010 13:56:25] <rmatte> I need to move everything over from my home server, going to take a while to do

[30-Mar-2010 13:56:42] <rmatte> need to move my mailserver, nameserver, webserver, my mysql databases, irc server, etc...

[30-Mar-2010 13:56:45] <venturaville> I have a lifetime account with dyndns which is nice

[30-Mar-2010 13:57:10] <rmatte> meaning you don't have to keep clicking every month to keep it active?

[30-Mar-2010 13:57:21] <rmatte> I have a static IP for my home internet, so I don't bother with that anymore

[30-Mar-2010 13:57:32] <venturaville> yeah I got in early when dyndns first started

[30-Mar-2010 13:57:35] <rmatte> I used to use dyndns back in the day

[30-Mar-2010 13:57:38] <rmatte> cool

[30-Mar-2010 13:57:38] <mrayzenoss> rmatte: there's a daemon for dyndns, you never have to touch anything

[30-Mar-2010 13:57:48] <rmatte> mrayzenoss: yeh, I used to run it

[30-Mar-2010 13:57:52] <venturaville> I don't have to either since I have a static IP

[30-Mar-2010 13:57:52] <mrayzenoss> I've been using it for 6 or 7 years

[30-Mar-2010 13:58:13] <venturaville> going on 10 for me  I believe

[30-Mar-2010 13:58:23] <rmatte> but when I was running it it was on a 150MHz PC which I was using as a FreeBSD server, connected to a 56k modem

[30-Mar-2010 13:58:36] <rmatte> I think I was 15 or 16 at the time

[30-Mar-2010 14:24:29] <ckrough> I even have dyndns on my android phone

[30-Mar-2010 14:36:07] <rmatte> ckrough: nice

[30-Mar-2010 14:42:30] <ckrough> 2.5.2 seems to like memory more than 2.3.3 did, spending a lot of time fighting off swap usage

[30-Mar-2010 14:43:04] <rmatte> ckrough: yeh, there are definitely leaks

[30-Mar-2010 14:43:16] <rmatte> ckrough: I have the same problem in both 2.4.5 and 2.5.2

[30-Mar-2010 14:43:24] <rmatte> but honestly, even 2.3.3 had an issue with that as well

[30-Mar-2010 14:43:29] <ckrough> doesnt take long, I am using up 32GB of ram and going into swap within ~30min

[30-Mar-2010 14:43:36] <rmatte> I find that doing the following helps...

[30-Mar-2010 14:43:43] <rmatte> zenperfsnmp restart

[30-Mar-2010 14:43:46] <ckrough> I've taking my cachesize down to 2500 now

[30-Mar-2010 14:43:51] <rmatte> then as root: swapoff -a && swapon -a

[30-Mar-2010 14:43:59] <rmatte> zenperfsnmp has the leak

[30-Mar-2010 14:44:04] <ckrough> its zenhub on the master thats killing me

[30-Mar-2010 14:44:06] <ckrough> zenperf is ok

[30-Mar-2010 14:44:06] <rmatte> Zope has a bit of a leak too but not nearly as bad

[30-Mar-2010 14:44:13] <ckrough> I think it's the ping tree rebuilds

[30-Mar-2010 14:44:22] <rmatte> I've never seen a problem with zenhub when it comes to memory

[30-Mar-2010 14:44:38] <ckrough> 'sync; echo 3 > /proc/sys/vm/drop_caches' before the swapoff -a will write out disk cache too

[30-Mar-2010 14:45:11] <rmatte> cool

[30-Mar-2010 14:45:41] <ckrough> I can scroll the event console again. havent been able to do that in forever

[30-Mar-2010 14:45:56] <rmatte> hehe

[30-Mar-2010 15:04:41] <car1> for mysql passwords stored in zproperties, are they stored in the zopedb in plain text?

[30-Mar-2010 15:05:02] <car1> our monitoring server was compromised and i'm concerned that they may have gotten access to our customer's dbs

[30-Mar-2010 15:10:34] <QbY> anyone know if there is any performance impact on loading thousands of locations and thousands of devices into zenoss.. or should they be loaded into a separate installaton?  this is for cable modems in the field-

[30-Mar-2010 15:14:39] <ckrough> QbY: everything adds some overhead, it will depend on what you are monitoring

[30-Mar-2010 15:15:11] <QbY> they are cable modems that are at the subscriber prem..  in order for me to get a nice map, i'd have to add a location to each..

[30-Mar-2010 15:23:25] <rmatte> well, the map is going to be extremely intensive to load with thousands of locations

[30-Mar-2010 15:23:47] <rmatte> the rest is dependant on what you're monitoring on each modem

[30-Mar-2010 15:23:53] <rmatte> if it's ping only then it'll be fine

[30-Mar-2010 15:24:05] <rmatte> if you're doing a fair bit of snmp monitoring for each you'll want to look in to remote collectors

[30-Mar-2010 15:28:02] <ckrough> when you are using workers, does the parent count as a hub process? IE: 2 workers = 3 total with the parent doing some processing

[30-Mar-2010 15:28:12] <ckrough> or does the parent just deligate to the workers

[30-Mar-2010 15:28:45] <ckrough> *delegate*

[30-Mar-2010 15:29:21] <rmatte> delegates I think, but not 100% sure

[30-Mar-2010 15:30:24] <venturaville> confirmed

[30-Mar-2010 15:31:59] <rmatte> venturaville: made any progress with esx?

[30-Mar-2010 15:34:26] <venturaville> haven't had time to get back to it yet, though I did get the pools/volumes code working for libvirt... so should have more time later this week

[30-Mar-2010 15:34:51] <rmatte> cool

[30-Mar-2010 15:35:03] <venturaville> spent almost an entire day trying to get zenoss UI performing better last week :-/

[30-Mar-2010 15:35:19] <rmatte> yeh, it's getting slower with each release it seems

[30-Mar-2010 15:37:18] <venturaville> hopefully them taking some of the load off of the hub in the next release will help

[30-Mar-2010 15:40:57] <rmatte> yeh

[30-Mar-2010 15:42:21] <ckrough> venturaville: finished up 2.5.2 upgrades, zenhub seems more memory hungry

[30-Mar-2010 15:45:14] <venturaville> ckrough: that is about what I have seen too

[30-Mar-2010 15:45:24] <venturaville> did you ever get that diagram online?

[30-Mar-2010 15:45:48] <ckrough> nah, that whole design is out the window now

[30-Mar-2010 15:45:58] <venturaville> how so?

[30-Mar-2010 15:46:10] <ckrough> for now, 2.5.2 is behaving much differently

[30-Mar-2010 15:46:56] <ckrough> one of the problems is that zenping rebuilds its tree per-daemon, so when you are running multiple daemons and hubs, the hubs eat all your ram and swap

[30-Mar-2010 15:47:04] <ckrough> didnt happen in dev, but it's happening in prod...

[30-Mar-2010 15:47:51] <rmatte> that blows because I already have memory issues with 2.4.5

[30-Mar-2010 15:47:52] <ckrough> we patched PingConfig.py so that the master box does all the pining

[30-Mar-2010 15:47:57] <rmatte> and I don't want worse memory issues with 2.5.2

[30-Mar-2010 15:48:06] <ckrough> rather than each collector, but still ahving memory issues

[30-Mar-2010 15:48:11] <rmatte> it's getting to the point where it's ridiculous

[30-Mar-2010 15:48:35] <QbY> rmatte: sorry for the delay, on each modem we'd colect about 4 snmp points. plus, one is gotta come from the cmts, so i don't know how i'm going to do that

[30-Mar-2010 15:49:04] <ckrough> after all the configs and ping trees are built it seems ok, but it takes over an hour to get going if you restart

[30-Mar-2010 15:49:08] <ckrough> thats no good

[30-Mar-2010 15:50:49] <venturaville> we haven't seen that particular problem yet

[30-Mar-2010 15:50:58] <venturaville> but then we aren't adding in large routing tables either

[30-Mar-2010 15:51:16] <venturaville> mostly servers

[30-Mar-2010 15:52:19] <rmatte> QbY: 4000 datapoints isn't much

[30-Mar-2010 15:53:02] <QbY> i'm debating if i want to install a fresh zenoss on another separate server just for cable modems

[30-Mar-2010 15:53:06] <QbY> or mix it in with my regular

[30-Mar-2010 15:54:01] <venturaville> get up into the 6 figures and we can talk :-)

[30-Mar-2010 15:54:33] <venturaville> have to go ..later all

[30-Mar-2010 16:10:07] <rmatte> later

[30-Mar-2010 16:20:40] <jb_> what needs to be running for zenoss to get traps?

[30-Mar-2010 16:20:45] <jb_> its coming in (verified via tcpdump)

[30-Mar-2010 16:20:47] <jb_> zentrap is running

[30-Mar-2010 16:20:52] <jb_> is there any reliance on net-snmp?

[30-Mar-2010 16:38:10] <rmatte> no

[30-Mar-2010 16:38:11] <rmatte> none

[30-Mar-2010 16:38:15] <rmatte> zentrap handles everything

[30-Mar-2010 16:38:31] <rmatte> net-snmp is just an snmp agent

[30-Mar-2010 16:39:04] <rmatte> it's possible that you have something which is running on the box and interfering with zentrap

[30-Mar-2010 16:39:08] <rmatte> I've seen it many times before

[30-Mar-2010 16:39:16] <rmatte> make sure you don't have any other snmp servers running on there

[30-Mar-2010 16:42:13] <packer> Having problems saving properties on custom object (DeviceComponent) - anyone ?

[30-Mar-2010 16:50:48] <mrayzenoss> Well that went smoothly.  10.6.3 upgrade wasted an hour of my life

[30-Mar-2010 16:52:43] <ckrough> argh

[30-Mar-2010 16:53:06] <ckrough> each zenhubworker is using 5GB of ram, seemingly regardless of my cachesize setting

[30-Mar-2010 16:53:15] <ckrough> anyone seen that in 2.5.2 before?

[30-Mar-2010 17:03:23] <rhettardo> not i

[30-Mar-2010 17:03:44] <ckrough> bummer, I run out of swap before the collectors get their configs done

[30-Mar-2010 20:34:16] cgibbons_ is now known as cgibbons

[30-Mar-2010 23:05:12] forsberg is now known as fOrsberg

[31-Mar-2010 00:00:30] [disconnected at Wed Mar 31 00:00:30 2010]

[31-Mar-2010 00:00:30] [connected at Wed Mar 31 00:00:30 2010]

[31-Mar-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[31-Mar-2010 03:01:00] <fus10nx> I am just trying to figure out exactly how Zenoss works and if I understand correctly the server end will sit on a linux server nativily i.e. Ubuntu and then on each client, linux or windows let's say, an SNMP client piece of software must be installed. Am I correct in saying that ?

[31-Mar-2010 03:49:18] <fus10nx> Anyone?

[31-Mar-2010 04:27:35] <frozty_sa> uhhhm

[31-Mar-2010 04:28:07] <frozty_sa> so while I'm waiting for my password reset mail, does the community site now entirely run the mailing lists in a forum-like manner?

[31-Mar-2010 04:28:23] <frozty_sa> or do you just enable mail notifications and then get mails like a mailing list would normally work?

[31-Mar-2010 04:29:54] <mrayzenoss> you can use it as a web forums or an email list, I think you just have to enable notifications for the forums you care to receive emails for

[31-Mar-2010 04:33:12] <frozty_sa> okay

[31-Mar-2010 04:33:14] <frozty_sa> that suits me

[31-Mar-2010 04:33:35] <mrayzenoss> I use a mix myself, but I'm also subscribed to everything

[31-Mar-2010 04:35:33] <frozty_sa> I mainly want -users, -dev and -zenpacks at this time, I think

[31-Mar-2010 04:38:14] <mrayzenoss> yeah, I'm probably the special case, since it's my job to keep an eye on everything ;)

[31-Mar-2010 04:38:17] <frozty_sa> what does the "Follow this community" button mean? :)

[31-Mar-2010 04:38:23] <mrayzenoss> get emails

[31-Mar-2010 04:38:24] <frozty_sa> mrayzenoss: true true :)

[31-Mar-2010 04:38:53] <mrayzenoss> you'll probably want some filters for those too

[31-Mar-2010 04:39:23] <mrayzenoss> me, I've got about 30 mail folders and twice as many rules

[31-Mar-2010 04:39:44] <frozty_sa> so what's the difference between that one and enable "Receive email notifications"?

[31-Mar-2010 04:40:03] * frozty_sa just clicks both anyway

[31-Mar-2010 04:40:20] <frozty_sa> mrayzenoss: yar, I'm probably sitting at around 45 filters anyhow

[31-Mar-2010 04:40:27] <frozty_sa> a bit more won't really hurt :D

[31-Mar-2010 04:40:44] <frozty_sa> kmail's filtering ftw , as far as desktop clients go

[31-Mar-2010 04:41:37] <mrayzenoss> hmm… don't know what the difference is

[31-Mar-2010 04:42:19] <mrayzenoss> looks like I've got both clicked for everything, again, I'm the special case

[31-Mar-2010 04:46:06] <danuvius> is it possiple to make a graph with sums all values from multiple interfaces. (as if it were one interface?).

[31-Mar-2010 04:46:57] <danuvius> we do some load balancing ad we would like to have a report that sums the traffic of all the interfaces is one graph as it was one interface

[31-Mar-2010 04:47:13] <frozty_sa> in about 3 months I'll probably be on most of the lists anyway

[31-Mar-2010 04:47:43] <mrayzenoss> danuvius: multigraphs http://www.youtube.com/watch?v=qME_7Zh4yOc&feature=player_embedded

[31-Mar-2010 04:48:08] <mrayzenoss> it's in the Admin Guide

[31-Mar-2010 04:55:17] <danuvius> I have been playing with that, but it is not what I'm looking for (or I'm missing something).

[31-Mar-2010 04:56:57] <danuvius> Say I have two interfaces that have incoming and outgoing packet. And in my graph I want two lines one with the sum of the incoming packets and one line with the sum of the outgoing packets

[31-Mar-2010 04:57:25] <danuvius> and with multigraph reports I get four lines...(not summed)

[31-Mar-2010 04:58:40] <mrayzenoss> I haven't done any lately, but it seems you should be able to get that

[31-Mar-2010 04:59:15] <mrayzenoss> 12.7 in the Admin Guide

[31-Mar-2010 04:59:25] * mrayzenoss really should go back to bed

[31-Mar-2010 04:59:47] <mrayzenoss> and there's the baby again

[31-Mar-2010 04:59:53] <mrayzenoss> night all (again)

[31-Mar-2010 05:00:07] <Simon4> danuvius: create a collection with both interfaces, a graph with two datapoints, one for in, one for out, change the datapoints to be "stacked", then a graphgroup of the collection and datapoints with "all devices on same graph". I tend to make one datapoint an area too (so you get area vs line for easy identification)

[31-Mar-2010 05:00:49] <frozty_sa> that video is a good reference point

[31-Mar-2010 05:00:57] <frozty_sa> I'd have hated something quite a lot without that

[31-Mar-2010 05:04:22] <danuvius> but then I get four "sources" in the label (and four areas)

[31-Mar-2010 05:04:44] <danuvius> i only want two lines (and in the label the summed values)

[31-Mar-2010 05:09:06] <fus10nx> Does anyone here know how the ZenOSS license works ? For instance there are tons of things we love about ZenOSS that we want to integrate into a web based piece of software that we commerically sell, is this possible or do we need to buy some type of license from ZenOSS ?

[31-Mar-2010 05:10:16] <Simon4> danuvius: you can do that, it's a wee bit more complex but doable.. give me a minute or two and I'll just remind myself of how

[31-Mar-2010 05:10:49] <fus10nx> Also, when I add a device, you can't choose a windows desktop for instance, can the software do this? I only see /Server/Windows

[31-Mar-2010 05:11:09] <fus10nx> Or does ZenOSS class Windows XP / Windows Vista / Windows 7 as "Servers"

[31-Mar-2010 05:26:23] <chemist> fus10nx: it does treat windows machines as servers

[31-Mar-2010 05:30:59] <fus10nx> ok, thanks

[31-Mar-2010 05:31:10] <fus10nx> for some reason i keep getting snmp down but its installed and started on my windows 7 machine

[31-Mar-2010 05:31:24] <chemist> firewall?

[31-Mar-2010 05:32:16] <fus10nx> shouldnt be

[31-Mar-2010 05:32:18] <fus10nx> local netowrk

[31-Mar-2010 05:32:42] <chemist> can you snmpwalk it from zenoss?

[31-Mar-2010 05:33:19] <fus10nx> i've just installed ZenoSS and read the readme but dont know what snmpwalk is

[31-Mar-2010 05:34:22] <chemist> if you go to the page for your device you will see the tabs on top, 'Status' OS' 'Hardware' etc

[31-Mar-2010 05:34:48] <chemist> next to them is a drop down arrow

[31-Mar-2010 05:35:18] <chemist> go to 'Run Commands' and snmpwalk

[31-Mar-2010 05:47:20] <fus10nx> sorry

[31-Mar-2010 05:47:24] <fus10nx> been trying to find this drop down

[31-Mar-2010 05:47:30] <fus10nx> oohhh

[31-Mar-2010 05:48:23] <fus10nx> Timeout: No Response from 192.168.0.47

[31-Mar-2010 05:48:56] <fus10nx> ill try to open port 161 on this windows firewall and see if it makes a difference

[31-Mar-2010 05:50:18] <fus10nx> bah, that still didn't do it...perhaps i've insatlled SNMP wrong or something, I just turned it on and start it in Windows Features

[31-Mar-2010 05:51:39] <chemist> you have to set up the community strings

[31-Mar-2010 05:51:49] <fus10nx> is there a doc on this for windows

[31-Mar-2010 05:52:13] <chemist> one second

[31-Mar-2010 05:55:03] <chemist> you need to open your services console on the windows machine

[31-Mar-2010 05:55:07] <fus10nx> ya i see it now

[31-Mar-2010 05:55:09] <fus10nx> i found a doc

[31-Mar-2010 05:55:13] <fus10nx> community name under traps

[31-Mar-2010 05:55:15] <fus10nx> "public" ?

[31-Mar-2010 05:55:56] <fus10nx> then accept SNMP packets from any host

[31-Mar-2010 05:56:16] <fus10nx> rather ill just specify my ZenOSS server IP address

[31-Mar-2010 05:56:28] <fus10nx> there we go !

[31-Mar-2010 05:56:32] <fus10nx> snmpwalk worked now

[31-Mar-2010 05:57:37] <fus10nx> so is there a way to change that "public" name to something specific or is it always public ?

[31-Mar-2010 05:57:54] <chemist> you can change it to whatever you want

[31-Mar-2010 05:58:05] <fus10nx> very cool

[31-Mar-2010 05:58:07] <chemist> so long as you change it in zenoss too

[31-Mar-2010 05:58:12] <fus10nx> yup, i gotcha

[31-Mar-2010 05:58:35] <fus10nx> ill have to play w/ this software quite a bit. I need to email the owner(s) to see if we can integrate this into our software that we sell if we buy a license fee or something

[31-Mar-2010 05:58:40] <fus10nx> it seems to be really amazing software

[31-Mar-2010 05:58:47] <chemist> it is

[31-Mar-2010 05:58:49] <chemist> :)

[31-Mar-2010 05:59:20] <fus10nx> so can this monitor Mac's as well ?

[31-Mar-2010 05:59:24] <chemist> yes

[31-Mar-2010 05:59:26] <fus10nx> man

[31-Mar-2010 05:59:27] <fus10nx> amazing !

[31-Mar-2010 05:59:50] <fus10nx> do you use it in a work enviornment or personal ?

[31-Mar-2010 05:59:54] <chemist> work

[31-Mar-2010 05:59:58] <fus10nx> how many devices ?

[31-Mar-2010 06:00:22] <chemist> we monitor about 450 devices at 35 locations

[31-Mar-2010 06:00:25] <fus10nx> I've been looking into solutions like this because we are working w/ Sony to build them a network operations center (NOC) and this seems like a perfect fit

[31-Mar-2010 06:00:30] <fus10nx> HOLY SHIT !

[31-Mar-2010 06:00:39] <chemist> that is small

[31-Mar-2010 06:00:41] <fus10nx> I was going to ask about seperate locaations...how does it get the SNMP messages? VPN tunnel ?

[31-Mar-2010 06:00:46] <chemist> yes

[31-Mar-2010 06:00:51] <fus10nx> very cool

[31-Mar-2010 06:01:19] <chemist> there are peopleon this channel running bigger than we are

[31-Mar-2010 06:01:53] <fus10nx> thats amazing man

[31-Mar-2010 06:01:59] <fus10nx> so really its for monitoring not controlling

[31-Mar-2010 06:02:16] <fus10nx> also, if snmpwalk is working, the system should be working, ya ?

[31-Mar-2010 06:02:34] <chemist> yes

[31-Mar-2010 06:03:10] <chemist> you can control to acertain degree, as in you can get zenoss to run scripts basedon events

[31-Mar-2010 06:03:14] <fus10nx> so really I do only need "READ ONLY" for the accepted community stuff

[31-Mar-2010 06:03:20] <fus10nx> oh wow, very cool

[31-Mar-2010 06:04:02] <chemist> the channel gets very busy after 17:00 gmt

[31-Mar-2010 06:04:10] <fus10nx> I'm in the UK

[31-Mar-2010 06:04:14] <chemist> same as

[31-Mar-2010 06:04:17] <fus10nx> from the US but living in the UK now

[31-Mar-2010 06:04:21] <fus10nx> where about in the UK are you?

[31-Mar-2010 06:04:29] <chemist> but most users are canada/US

[31-Mar-2010 06:04:30] <fus10nx> I'm East Midlands area

[31-Mar-2010 06:04:33] <chemist> Bristol

[31-Mar-2010 06:04:38] <fus10nx> ohh ok ive been there a few times

[31-Mar-2010 06:05:01] <fus10nx> its pretty slick the whoel system, I must say

[31-Mar-2010 06:05:24] <chemist> yes, I've been using it in production for about six months

[31-Mar-2010 06:05:31] <chemist> still learning though :)

[31-Mar-2010 06:05:36] <fus10nx> I'm not sure if you can answer this but can you customize the color and all that on the interface, its open source right ?

[31-Mar-2010 06:05:48] <chemist> it helps if you know python

[31-Mar-2010 06:05:53] <fus10nx> ya

[31-Mar-2010 06:06:00] <fus10nx> well i figured there is some CSS element in there

[31-Mar-2010 06:06:08] <chemist> it is open source and written in python

[31-Mar-2010 06:06:23] <chemist> but I don't know about the UI

[31-Mar-2010 06:06:39] <fus10nx> no worries

[31-Mar-2010 06:06:46] <fus10nx> how long does it take on average to collect information from a device

[31-Mar-2010 06:06:53] <fus10nx> I did the snmpwalk but it still hasn't populated things

[31-Mar-2010 06:07:10] <fus10nx> nevermind !

[31-Mar-2010 06:07:11] <fus10nx> haha

[31-Mar-2010 06:07:22] <fus10nx> holy. shit. that is amazing.

[31-Mar-2010 06:07:24] <chemist> from the same drop down menu

[31-Mar-2010 06:07:29] <chemist> Manage

[31-Mar-2010 06:07:33] <chemist> Maodeldevice

[31-Mar-2010 06:07:46] <fus10nx> ohhh model queries it

[31-Mar-2010 06:07:46] <chemist> *Model Device

[31-Mar-2010 06:07:51] <chemist> yes

[31-Mar-2010 06:07:53] <fus10nx> i gotta go through and read the manual more indepth now

[31-Mar-2010 06:07:58] <chemist> yes :)

[31-Mar-2010 06:07:59] <danuvius> Simon4: any ideas?

[31-Mar-2010 06:08:01] <fus10nx> thanks SO much for your help man, this is AMAZING

[31-Mar-2010 06:08:05] <chemist> np

[31-Mar-2010 06:08:07] <fus10nx> wait wait

[31-Mar-2010 06:08:12] <fus10nx> your using this in your organization you said

[31-Mar-2010 06:08:18] <chemist> yes

[31-Mar-2010 06:08:18] <fus10nx> and you dont have to pay to use it ?

[31-Mar-2010 06:08:20] <danuvius> Simon4: did you remomeber again?

[31-Mar-2010 06:08:27] <chemist> no, core is free

[31-Mar-2010 06:08:30] <fus10nx> jesus

[31-Mar-2010 06:08:36] <fus10nx> im off to play then :)

[31-Mar-2010 06:08:39] <chemist> enterprise you have to pay

[31-Mar-2010 06:08:51] <chemist> have fun :)

[31-Mar-2010 06:08:53] <Simon4> danuvius: sorry, in a meeting, but I looked, and I had to patch some code to make that work from memory

[31-Mar-2010 06:08:56] <Simon4> has been a while

[31-Mar-2010 06:09:00] <fus10nx> zendisc and zenmodeler shouldnt be lit up green for windows, only linux right ?

[31-Mar-2010 06:09:20] <chemist> no, for all

[31-Mar-2010 06:09:22] <fus10nx> snmp, zendisc and zenmodelete are all like orange looking ciricles and the IP address of the device is blue

[31-Mar-2010 06:09:37] <fus10nx> under the devices --> stastus

[31-Mar-2010 06:09:53] <chemist> blue is info

[31-Mar-2010 06:09:59] <chemist> orange is an error

[31-Mar-2010 06:10:17] <chemist> so either snmp is down or not configured properly

[31-Mar-2010 06:10:30] <fus10nx> well its collecting all the info .

[31-Mar-2010 06:10:45] <chemist> if you change anything in snmp service on windows you have to restart the service

[31-Mar-2010 06:10:55] <fus10nx> ahh ok, ill give that a shot

[31-Mar-2010 06:11:12] <fus10nx> do I need to do anything w/ SNMP trap in the windows services ?

[31-Mar-2010 06:12:09] <chemist> you can ask windows to send traps for certain events

[31-Mar-2010 06:12:43] <chemist> I have snmp traps set for application crash and application hang events

[31-Mar-2010 06:13:05] <fus10nx> ahhh okkkkkkk

[31-Mar-2010 06:13:15] <fus10nx> man i really gotta learn about SNMP apparently

[31-Mar-2010 06:13:19] <fus10nx> this is like, the key to it all

[31-Mar-2010 06:14:08] <chemist> if you have many windows boxes you might want to use WMI also

[31-Mar-2010 06:14:38] <fus10nx> so just quickly...for the SNMP Service -- Under traps tab I have the community name, 'public', under the Security tab I have "Accept community names - public - READ ONLY' and then accept snmp packets form these hosts "192.168.0.x, localhost"

[31-Mar-2010 06:14:57] <fus10nx> the WMI performance Monitor ?

[31-Mar-2010 06:15:07] <fus10nx> Performance Adapater I mean

[31-Mar-2010 06:15:29] <chemist> you need to install the WMI zenpack first

[31-Mar-2010 06:15:43] <fus10nx> will it provide more information ?

[31-Mar-2010 06:15:48] <fus10nx> on the Zen interface I mean

[31-Mar-2010 06:15:59] <chemist> yes, a lot more

[31-Mar-2010 06:16:30] <chemist> I would change the 'public' string to something else anyway

[31-Mar-2010 06:20:59] <fus10nx> ok

[31-Mar-2010 06:21:15] <fus10nx> i wish this would run on windows nativly, the server end

[31-Mar-2010 06:21:27] <fus10nx> that would just be the best, then i wouldnt need to get linux involved at all

[31-Mar-2010 06:21:43] <fus10nx> not that i am against linux, i love it personally, but all of our clients are windwos based

[31-Mar-2010 06:49:10] <fus10nx> where do I modify the community name?

[31-Mar-2010 06:49:18] <fus10nx> i tried in /etc/snmp/snmpd.conf

[31-Mar-2010 07:25:15] <fus10nx> Simon4: do you where I can change the SNMP community name ?

[31-Mar-2010 07:31:48] <Simon4> fus10nx: in the zproperties of the device, or the device class

[31-Mar-2010 07:31:57] <Simon4> if you set the device class ones, it affects all devices in that class

[31-Mar-2010 07:32:36] <fus10nx> what do you mean zproperties ? is that on the web interface?

[31-Mar-2010 07:32:55] <Simon4> yeah, on the web interface

[31-Mar-2010 07:33:07] <Simon4> say if you navigate to "devices/windows/server"

[31-Mar-2010 07:33:21] <Simon4> you'll see a tab called "zProperties"

[31-Mar-2010 07:33:39] <fus10nx> oh oh oh

[31-Mar-2010 07:33:58] <fus10nx> ahhhh ok i see how this works now

[31-Mar-2010 07:34:20] <fus10nx> so there's some predefined classes but you can make your own

[31-Mar-2010 07:40:56] <Simon4> yeah

[31-Mar-2010 07:40:58] <Simon4> exactly

[31-Mar-2010 07:41:20] <Simon4> the predefined ones are a good starter, since they have both zproperties set appropriately, and also appropriate performance templates bound

[31-Mar-2010 07:43:03] <fus10nx> ya I'll play around with that

[31-Mar-2010 07:43:13] <fus10nx> for some reason zendisc always reigsters as orange

[31-Mar-2010 07:43:15] <fus10nx> and i cant figure out why

[31-Mar-2010 07:43:25] <fus10nx> IpRouteEntry is Green

[31-Mar-2010 07:43:51] <fus10nx> IpInterface = Green , IP address (ping I think) is blue, zendisc is orange

[31-Mar-2010 08:08:13] <fus10nx> Anyone else alive who can give a hand ?

[31-Mar-2010 08:13:07] <Simon4> fus10nx: when you add/model a device, see if there are any errors in the output

[31-Mar-2010 08:13:21] * Simon4 runs away again, stupid work getting in the way

[31-Mar-2010 08:13:52] <fus10nx> haha ok

[31-Mar-2010 08:14:02] <fus10nx> well ive just been using autodiscover

[31-Mar-2010 08:34:19] fOrsberg is now known as forsberg

[31-Mar-2010 08:37:38] <fus10nx> so pretty much to get any useful information from windows like hardware, software, etc. you need a zenpack

[31-Mar-2010 08:40:46] forsberg is now known as fOrsberg

[31-Mar-2010 08:42:11] fOrsberg is now known as forsberg

[31-Mar-2010 08:42:18] <fus10nx> so pretty much to get any useful information from windows like hardware, software, etc. you need a zenpack

[31-Mar-2010 08:42:21] <fus10nx> err sorry

[31-Mar-2010 08:50:10] <fus10nx> ohhhhhhhhh so ZenOSS is like What's Up Gold kind of

[31-Mar-2010 08:55:36] <Apachez> fus10nx: unless its standard snmp

[31-Mar-2010 08:55:52] <Apachez> however ossim might be something useful aswell

[31-Mar-2010 08:56:34] <fus10nx> ossim

[31-Mar-2010 08:56:43] <Apachez> the downside with zenoss (in my opinion) is the non optimized db settings along with some old python which uses (whatever the name is of that application webserver) singlethreaded stuff so its really slow compared to other solutions

[31-Mar-2010 08:56:45] <fus10nx> hmmm ok, I would just like to populate like software, hardware, event, etc.

[31-Mar-2010 08:57:07] <fus10nx> for a free solution its absolutely amazing

[31-Mar-2010 08:57:08] <Apachez> but if you can live with that (and most people can) then zenoss is great :)

[31-Mar-2010 08:57:10] <fus10nx> much better than nagios

[31-Mar-2010 08:57:14] <Apachez> yup :)

[31-Mar-2010 08:57:22] <fus10nx> so do you use the wmi stuff? the zenpacks?

[31-Mar-2010 08:57:29] <fus10nx> im just trying to figure out how to install them

[31-Mar-2010 08:57:35] <Apachez> specially since both syslogging and graphing is included

[31-Mar-2010 08:57:46] <Apachez> like with cacti you only get graphing (along with some bugs=

[31-Mar-2010 08:57:49] <fus10nx> ya i cant seem to get my event log information or hardware or software

[31-Mar-2010 08:58:05] <Apachez> nope I just used vanilla zenoss

[31-Mar-2010 08:58:09] <Apachez> the stack install

[31-Mar-2010 08:58:20] <fus10nx> so you get like super basic info. from windows devices

[31-Mar-2010 08:58:22] <Apachez> I tried a couple of zenpacks but non that got my attention

[31-Mar-2010 08:58:35] <fus10nx> i was told that hte WMI stuff would provide loads more information

[31-Mar-2010 08:58:51] <Apachez> in my case I monitor a bunch of switches/routers (snmp both polling and traps) and get logs from them (syslog)

[31-Mar-2010 08:58:59] <fus10nx> ahhh i see

[31-Mar-2010 08:59:07] <fus10nx> 95% of our stuff would be windows based

[31-Mar-2010 08:59:13] <Apachez> and then I have a couple of "SLA" monitors like ping latency towards a couple of sites

[31-Mar-2010 08:59:18] <fus10nx> are installing zenpacks a bit of a pain ?

[31-Mar-2010 08:59:26] <fus10nx> ahhh i see

[31-Mar-2010 09:01:10] <Apachez> well backup is always good if you want to store history data

[31-Mar-2010 09:01:21] <Apachez> since not all packs are that good in install/uninstall

[31-Mar-2010 09:01:29] <Apachez> but usually its really easy

[31-Mar-2010 09:01:35] <Apachez> you upload the pack and activate it

[31-Mar-2010 09:01:36] <fus10nx> ya i've already imaged my working machine just a few minutes ago

[31-Mar-2010 09:01:43] <fus10nx> simple as that ?

[31-Mar-2010 09:01:49] <Apachez> sometimes you need to restart the zenoss to make the pack appear

[31-Mar-2010 09:02:07] <Apachez> (not restart the whole box but the zenoss core process which restarts all zenoss related stuff)

[31-Mar-2010 09:02:07] <fus10nx> ill give it a shot, if i break it, i've got a backup any how

[31-Mar-2010 09:02:17] <fus10nx> ya

[31-Mar-2010 09:02:17] <Apachez> clonezilla is your friend ;)

[31-Mar-2010 09:02:21] <fus10nx> haha

[31-Mar-2010 09:02:23] <fus10nx> acronis here

[31-Mar-2010 09:02:38] <fus10nx> so is there a specific place they need to be uploaded to ?

[31-Mar-2010 09:02:54] <Apachez> ehm

[31-Mar-2010 09:02:56] <Apachez> readme :)

[31-Mar-2010 09:03:50] <fus10nx> haha ok ill give it a shot

[31-Mar-2010 09:04:01] <fus10nx> egg files :)

[31-Mar-2010 09:08:07] <fus10nx> well, thats frickin easy though the interface

[31-Mar-2010 09:08:11] <fus10nx> i love simple software

[31-Mar-2010 09:11:33] <fus10nx> I just install the WMI Data Source -- should it now just automatically start collecting all of the more advanced information from Windows?

[31-Mar-2010 09:18:17] <rmatte> well, you have to install the performance pack that goes with it

[31-Mar-2010 09:18:32] <rmatte> to get the device template

[31-Mar-2010 09:20:26] <fus10nx> there doesnt seem to be any documentation on this page for it

[31-Mar-2010 09:20:32] <fus10nx> so its more than just the egg file?

[31-Mar-2010 09:21:05] <rmatte> no

[31-Mar-2010 09:21:22] <rmatte> k, Egor Puzanov has made several WMI related ZenPacks

[31-Mar-2010 09:21:28] <fus10nx> ya

[31-Mar-2010 09:21:41] <rmatte> all the WMI Data Source pack is is a daemon to collect WMI info, which adds a new datasource type to Zenoss

[31-Mar-2010 09:21:56] <fus10nx> ok gotcha

[31-Mar-2010 09:21:59] <rmatte> you still actually need a device template that uses that datasource type to collect the data

[31-Mar-2010 09:22:11] <fus10nx> so you mean installing additional software on windows ?

[31-Mar-2010 09:22:15] <rmatte> Hence you also want to install the WMI Windows Performance pack by Egor

[31-Mar-2010 09:22:18] <rmatte> no

[31-Mar-2010 09:22:22] <fus10nx> ya i installed both of them

[31-Mar-2010 09:22:28] <rmatte> well, you'll obviously need WMI configured on your windows boxes

[31-Mar-2010 09:22:48] <rmatte> and you'll need the WMI username/password set in zProperties as zWinUser and zWinPassword

[31-Mar-2010 09:22:50] <fus10nx> isn't that just enabling the service

[31-Mar-2010 09:23:01] <fus10nx> bah ok ill have to read up a lot more on WMI then

[31-Mar-2010 09:23:10] <fus10nx> I thought it was just enable the service, install zenpack, off you go

[31-Mar-2010 09:23:15] <rmatte> well...

[31-Mar-2010 09:23:29] <rmatte> the other 2 steps are to get an account that has admin access to all your boxes

[31-Mar-2010 09:23:48] <fus10nx> ya, domain account which I do have

[31-Mar-2010 09:23:51] <rmatte> and then putting the username and password for that account as zWinUser and zWinPassword in Zenoss

[31-Mar-2010 09:24:09] <rmatte> you also need to make sure that the devices are in a group with the WMI device template bound to it

[31-Mar-2010 09:24:26] <rmatte> did you just start using Zenoss or have you been using it for a while?

[31-Mar-2010 09:24:36] <fus10nx> just started, read through entire manual and now trying to use

[31-Mar-2010 09:24:48] <rmatte> by the entire manual you mean the Admin Guide?

[31-Mar-2010 09:24:54] <fus10nx> literally just started about 3 hours ago, read the admin guide through and through

[31-Mar-2010 09:25:00] <rmatte> k

[31-Mar-2010 09:25:05] <fus10nx> but still unclear to me with a lot of stuff, the wording of classes vs. groups and stuff

[31-Mar-2010 09:25:22] <rmatte> so you understand the concept of creating device classes, binding templates to those classes, and setting zProperties for those classes?

[31-Mar-2010 09:25:26] <fus10nx> before i even installed the software i read through the admin guide fully then re-read most of it and then installed

[31-Mar-2010 09:25:41] <fus10nx> i get it for the most part I suppose, its a matter of puttin git into practive

[31-Mar-2010 09:25:44] <fus10nx> practice*

[31-Mar-2010 09:25:48] <rmatte> well, it's kind of better to follow along in Zenoss as you read the guide

[31-Mar-2010 09:25:50] <fus10nx> to me I want to keep calling classes, groups

[31-Mar-2010 09:25:51] <rmatte> you absorb more that way

[31-Mar-2010 09:26:07] <fus10nx> ya thats my next approach, before i fully am sold on it i just wnated to make sure it can in fact do what we need it to

[31-Mar-2010 09:26:10] <rmatte> if you read the whole guide without even having touched the software your brain won't absorb the info well

[31-Mar-2010 09:26:20] <fus10nx> which was pretty much the case :)

[31-Mar-2010 09:26:28] <rmatte> well, what exactly do you need it to do?

[31-Mar-2010 09:26:30] <fus10nx> i get the general idea of it, thats for sure

[31-Mar-2010 09:26:49] <fus10nx> monitor a lot of windows devices and produce good reports for those devices. Also, restart software if they fail and a fe wother things

[31-Mar-2010 09:26:55] <fus10nx> inventory of PC and stuff like that

[31-Mar-2010 09:27:34] <rmatte> well, the restarting software thing isn't available by default, you could script it in

[31-Mar-2010 09:27:43] <Absenth> in Core, in the devices template, where would I need to go to disable auto-discovery of attached IP addresses?

[31-Mar-2010 09:27:45] <rmatte> the event manager can perform commands based on events

[31-Mar-2010 09:27:57] <fus10nx> ya which i was told, that would all come later though

[31-Mar-2010 09:28:06] <fus10nx> for now its collection and reporting of data

[31-Mar-2010 09:28:06] <rmatte> zenoss' reporting is a bit weak, but it's decent

[31-Mar-2010 09:28:25] <Absenth> I have a FreeBSD server that's running a number of Jails.  The Jails server ends up with the IP being monitored there, which prevents me from monitoring the Jail directly.

[31-Mar-2010 09:29:09] <rmatte> Absenth: you mean it gets monitored as an interface?

[31-Mar-2010 09:29:35] <rmatte> fus10nx: what Zenoss shines at is monitoring and alerting

[31-Mar-2010 09:29:53] <Absenth> rmatte: If I setup up jailsserver in zenoss using the FQDN, it binds not only that IP address, but every other IP address on that machine to that device in Zenoss.

[31-Mar-2010 09:30:05] <rmatte> fus10nx: it's extremely flexible, and easy to create new monitoring templates, but it has a learning curve like anything else

[31-Mar-2010 09:30:17] <Absenth> rmatte: which means once I get the Jail/VM started, I can't model it, because Zenoss thinks Jailsserver has that IP.

[31-Mar-2010 09:30:29] <fus10nx> sure, im gonna play a bit more to see if I can get WMI to work and then go back and re-read the admin guide again along side w/ messing w/ it

[31-Mar-2010 09:30:47] <rmatte> right, what I'm asking is, when you're saying that it binds all those IPs to that FQDN, is it doing it as interfaces?

[31-Mar-2010 09:31:05] <rmatte> it has to be storing the IPs somewhere in Zenoss, and devices in Zenoss can't have more than 1 main IP

[31-Mar-2010 09:31:27] <Absenth> rmatte: let me look again to be sure.

[31-Mar-2010 09:31:40] <rmatte> check on the OS tab to see if there are interfaces listed with those IPs

[31-Mar-2010 09:31:59] <rmatte> if so, then it's simple, just remove the interfaces, lock the interfaces section so that they don't get added back, then add the other IPs

[31-Mar-2010 09:32:05] <fus10nx> so there's a way to copy existing classes as they are and then expand on them, correct?

[31-Mar-2010 09:32:12] <fus10nx> or edit existing, pre-defined classes

[31-Mar-2010 09:32:14] <rmatte> you're not talking about an FQDN that is a pool of multiple IPs I assume?

[31-Mar-2010 09:32:31] <rmatte> fus10nx: are you confusing classes with templates?

[31-Mar-2010 09:32:34] <Absenth> rmatte: you're right it's adding them to the interfaces.

[31-Mar-2010 09:32:36] <fus10nx> blah possibly

[31-Mar-2010 09:32:51] <rmatte> fus10nx: basically a class is just a container that has several attributes defined to it

[31-Mar-2010 09:32:56] <Absenth> rmatte: no, I was assuming if I configured the device via FQDn it should only use the IP assigned to that name.

[31-Mar-2010 09:33:17] <rmatte> the attributes are zProperties, Templates, Collector Plugins, and a couple of others

[31-Mar-2010 09:33:53] <rmatte> to answer your question... you could copy /Server/Linux to /Server/Whatever and then create local copies of templates and modify them to your liking

[31-Mar-2010 09:33:55] <fus10nx> see to me it's like on the navigation bar.... Classes then under it is "Events" -- then events are basically the actual events that the device is reporting ?

[31-Mar-2010 09:34:02] <rmatte> you could also specific different collector plugins, and zProperties

[31-Mar-2010 09:34:13] <rmatte> you could simply create a brand new blank class and do the same

[31-Mar-2010 09:34:30] <fus10nx> ya, i think I'd go w/ just a fresh new class then move devices into that class

[31-Mar-2010 09:34:35] <rmatte> Event Classes are mappings which tell Zenoss how to handle the events

[31-Mar-2010 09:34:36] <fus10nx> or assign that "class" to those group of devices

[31-Mar-2010 09:34:44] <fus10nx> am I correct in saying that ?

[31-Mar-2010 09:34:55] <rmatte> you can create an event mapping for an event, when that event comes in Zenoss can modify the event based on rules

[31-Mar-2010 09:35:12] <rmatte> you don't assign event classes to device classes

[31-Mar-2010 09:35:17] <rmatte> they are 2 completely different things

[31-Mar-2010 09:35:25] <rmatte> event classes control the way events are handled when they come in

[31-Mar-2010 09:35:34] <rmatte> device classes control the way devices are monitored

[31-Mar-2010 09:35:36] <fus10nx> can you give a quick example?

[31-Mar-2010 09:35:46] <rmatte> oh event classes?

[31-Mar-2010 09:35:48] <rmatte> of*

[31-Mar-2010 09:35:48] <fus10nx> so an event class is like an applicaiton error

[31-Mar-2010 09:35:56] <fus10nx> let's say microsoft word crashses

[31-Mar-2010 09:36:00] <fus10nx> thats an event class, ya ?

[31-Mar-2010 09:36:05] <rmatte> well sort of

[31-Mar-2010 09:36:28] <rmatte> let's say you're monitoring any kind of service, such as IIS

[31-Mar-2010 09:36:32] <rmatte> and IIS crashes

[31-Mar-2010 09:36:44] <fus10nx> ok ok

[31-Mar-2010 09:36:46] <rmatte> an event will come in saying Windows Service 'IIS' is down or whatever

[31-Mar-2010 09:36:57] <Absenth> I'd just like to confirm quickly.  If I select bce0 and bce1, then remove interface(s), followed by lock interface(s).  The rest of my monitoring will continue to work.  (snmp etc.)

[31-Mar-2010 09:37:02] <rmatte> now let's say you want to automatically up the severity on events for IIS

[31-Mar-2010 09:37:22] <rmatte> you could then create a mapping that matches IIS events, and apply a rules to the mapping to up the severity of the event when that event comes in

[31-Mar-2010 09:37:41] <rmatte> there are also event transforms, which are python scripts that are applied against events when they come in (they are extremely powerful)

[31-Mar-2010 09:38:00] <rmatte> this stuff is all explained in the admin guide, you really should take another look

[31-Mar-2010 09:38:30] <rmatte> with event mappings you can taylor the look of events to your liking

[31-Mar-2010 09:38:55] <rmatte> you can change the severity, summary, message, component, class, and whatever else

[31-Mar-2010 09:38:55] <fus10nx> ahhh ok so this takes quite a bit of customization

[31-Mar-2010 09:38:55] <fus10nx> configuration rather

[31-Mar-2010 09:38:55] <rmatte> yes, it does

[31-Mar-2010 09:38:56] <fus10nx> ya i can see that now

[31-Mar-2010 09:39:07] <rmatte> it's aimed that someone who wants a very flexible and powerful monitoring tool

[31-Mar-2010 09:39:20] <rmatte> but it's certainly not the type of tool where you just drop stuff in and it does everything for you

[31-Mar-2010 09:39:28] <rmatte> aimed at*

[31-Mar-2010 09:40:06] <Absenth> rmatte: I'll second that statement.  We're working on scripts so that if a specific service on a machine dies, zenoss with either attempt to restart the service, or drop the CARP interface on that server to force all service to failover to the secondary.

[31-Mar-2010 09:40:07] <rmatte> the discovery of devices is automated to some degree, but you will need to customize certain things to get it to how you want it

[31-Mar-2010 09:40:27] <rmatte> Absenth: that shouldn't be overly difficult to do

[31-Mar-2010 09:40:29] <Absenth> <--- isn't the one responsible for those bits though.

[31-Mar-2010 09:40:52] <fus10nx> ya perhaps this is too much for us then, we are looking for a good solution to collect data and potentially down the line do the more powerful stuff but right away that's not our intention. Really we just wanted to collect data to start off with like CPU usage, Hard Disk usage, etc. etc.

[31-Mar-2010 09:41:04] <Absenth> is my assumption above correct?  dropping and then locking the interfaces won't result in the rest of my monitoring ceasing to work?

[31-Mar-2010 09:41:13] <Absenth> on the jails server that is.

[31-Mar-2010 09:42:24] <rmatte> correct

[31-Mar-2010 09:42:47] <Absenth> thanks :)

[31-Mar-2010 09:43:03] <rmatte> Zenoss will just stop monitoring those interfaces...

[31-Mar-2010 09:43:10] <rmatte> I assume those interfaces did have the jail IPs on them?

[31-Mar-2010 09:43:16] <Absenth> they do.

[31-Mar-2010 09:43:23] <rmatte> were they sub-interfaces, like eth0:1 or whatever?

[31-Mar-2010 09:43:31] <Absenth> FreeBSD doesn't work that way.

[31-Mar-2010 09:43:36] <rmatte> oh right

[31-Mar-2010 09:43:37] <Absenth> that would have been really nice though

[31-Mar-2010 09:43:38] <Absenth> :)

[31-Mar-2010 09:43:43] <rmatte> freebsd just pancakes them all on a single interface

[31-Mar-2010 09:43:54] <rmatte> from what I recall anyways

[31-Mar-2010 09:44:08] <rmatte> it's been years since I've done multiple IPs on a FreeBSD box

[31-Mar-2010 09:44:19] <Absenth> you're exactly right.

[31-Mar-2010 09:44:38] <Absenth> lots of alises, but they're all bce# (or whatever nic driver...)

[31-Mar-2010 09:44:39] <rmatte> I started with FreeBSD before I even touched Linux, I just haven't used it recently ;)

[31-Mar-2010 09:44:44] <rmatte> yeh

[31-Mar-2010 09:45:02] <Absenth> we've got ~ 180 or so FreeBSD machines, and ~ 20 linux VMs

[31-Mar-2010 09:45:09] <rmatte> nice

[31-Mar-2010 09:45:15] <Absenth> where we needed Xen, we're running Centos, otherwise it's FreeBD.

[31-Mar-2010 09:45:24] <Absenth> our Zenoss for instance is on a Centos box:)

[31-Mar-2010 09:45:29] <rmatte> yeh, good call

[31-Mar-2010 09:45:35] <rmatte> it's a pain to get it running on BSD

[31-Mar-2010 09:45:50] <rmatte> and then upgrades are a nightmare afterwards

[31-Mar-2010 09:46:15] <rmatte> but yeh, remove the interface and then you should be able to add the jails as separate devices

[31-Mar-2010 09:46:26] <Absenth> perfect!  Thanks rmatte

[31-Mar-2010 09:46:30] <rmatte> np

[31-Mar-2010 09:47:45] <fus10nx> ya rmatte thanks for your help man, ill read through it again but i need to make sure that the WMI stuff will work and then ill get way more into ZenOSS if I can see that it is collecting all the info. we need -- disk space, cpu, ram usage, etc. etc.

[31-Mar-2010 09:48:46] <fus10nx> there isn't a ton of info. in the admin guide on ZenPacks so thats just why I was asking...but if it requires a full configuration of the system just see if this one option we need will work, then that's a bit of a pain but so be it

[31-Mar-2010 09:54:54] <rmatte> np

[31-Mar-2010 09:55:21] <rmatte> well, the newest WMI pack is supposed to allow for full WMI monitoring, but I haven't tested it

[31-Mar-2010 09:55:29] <rmatte> you might be better to go with SNMP rather than WMI

[31-Mar-2010 09:55:37] <fus10nx> well I'm using SNMP already

[31-Mar-2010 09:55:39] <rmatte> that's how I monitor windows servers and it works great

[31-Mar-2010 09:55:42] <fus10nx> but it only gives small details

[31-Mar-2010 09:55:47] <rmatte> small details?

[31-Mar-2010 09:55:54] <fus10nx> ya it doesnt seem to report on hardware or anything

[31-Mar-2010 09:55:55] <fus10nx> or software

[31-Mar-2010 09:56:03] <rmatte> it reports on software just fine for me

[31-Mar-2010 09:56:11] <fus10nx> v2c ?

[31-Mar-2010 09:56:16] <rmatte> v1 or v2c

[31-Mar-2010 09:56:35] <fus10nx> strange, my hardware, software, events and so on are blank for multiple PC's

[31-Mar-2010 09:56:44] <fus10nx> but I can hostname, OS version, HW model, HW make

[31-Mar-2010 09:57:31] <fus10nx> if you go to one of your devices and it shows the component types like zenmodeler, IpRouteEntry, IpInterface ..... what color is your round icon for zenmodeler ?

[31-Mar-2010 09:58:33] <fus10nx> like /Devices/Discovered/<device name>

[31-Mar-2010 09:58:36] <fus10nx> then under the status

[31-Mar-2010 10:01:20] <rmatte> I don't run zenmodeler at all

[31-Mar-2010 10:01:44] <fus10nx> OHHHHH I see why now

[31-Mar-2010 10:01:48] <fus10nx> when you use autodiscover

[31-Mar-2010 10:02:03] <fus10nx> oh wow, if i manually enter the device i get LOADS more information

[31-Mar-2010 10:02:21] <fus10nx> holy crap, wow, ya

[31-Mar-2010 10:04:36] <rmatte> yes

[31-Mar-2010 10:04:46] <fus10nx> so autodiscover is what's been messing me up so bad then

[31-Mar-2010 10:04:57] <rmatte> well, that depends

[31-Mar-2010 10:05:10] <rmatte> autodiscover usually drops the devices in /Discovered class

[31-Mar-2010 10:05:14] <fus10nx> ya

[31-Mar-2010 10:05:14] <ckrough> ^^

[31-Mar-2010 10:05:18] <rmatte> you need to move them to the correct class

[31-Mar-2010 10:05:19] <fus10nx> OHHHHHHH

[31-Mar-2010 10:05:31] <rmatte> and then go to Manage -> Remodel from the device page

[31-Mar-2010 10:05:35] <rmatte> to collect all the data for them

[31-Mar-2010 10:05:36] <fus10nx> because now I see CPU, FileSystem, etc.

[31-Mar-2010 10:05:41] <rmatte> yup

[31-Mar-2010 10:05:42] <fus10nx> so is that the WMI stuff ?

[31-Mar-2010 10:05:52] <rmatte> well, what class are the devices in right now?

[31-Mar-2010 10:05:53] <fus10nx> because when it was adding the device it was showing a bunch of WMI things

[31-Mar-2010 10:06:04] <fus10nx> ...  /server/windows

[31-Mar-2010 10:06:16] <fus10nx> i.e. /Devices  /Server  /Windows  / <device name>

[31-Mar-2010 10:06:17] <rmatte> /Server/Windows is SNMP by default

[31-Mar-2010 10:06:24] <fus10nx> i see and Discover is not

[31-Mar-2010 10:06:28] <fus10nx> im getting it

[31-Mar-2010 10:06:33] <fus10nx> so this STILL isn't WMI, its all SNMP

[31-Mar-2010 10:06:47] <fus10nx> zeneventlog is showing as orange but it's OK

[31-Mar-2010 10:06:50] <rmatte> afk

[31-Mar-2010 10:07:20] <fus10nx> thnx a lot man

[31-Mar-2010 10:18:26] <rmatte> np

[31-Mar-2010 10:18:32] <rmatte> what about your performance tab?

[31-Mar-2010 10:18:37] <rmatte> you seeing CPU and Memory details?

[31-Mar-2010 10:18:44] <fus10nx> I figured out the WMI stuff too i belive

[31-Mar-2010 10:18:51] <fus10nx> i didnt realize zProperties were for each and everything

[31-Mar-2010 10:18:54] <fus10nx> like each class has it's own

[31-Mar-2010 10:19:09] <rmatte> well, you can define local properties per class, yes

[31-Mar-2010 10:19:16] <fus10nx> I keep getting access denied on the WMI and I think it's just cus I havent put in the domain correct i.e. MYDOMAIN/DomainAcct

[31-Mar-2010 10:19:33] <rmatte> yeh, you need the domain/whatever

[31-Mar-2010 10:19:42] <fus10nx> MYDOMAIN\Admin or MYDOMAIN/Admin

[31-Mar-2010 10:20:05] <fus10nx> in windows it'd be the first but not sure on this

[31-Mar-2010 10:21:21] <rmatte> use the second

[31-Mar-2010 10:21:31] <rmatte> they both technically work but people have better luck with the forward slash (/)

[31-Mar-2010 10:22:28] <fus10nx> I had the eureka moment and now its all making a LOT more sense

[31-Mar-2010 10:22:39] <rmatte> Zenoss tends to do that to you

[31-Mar-2010 10:22:45] <rmatte> you'll have many more, believe me

[31-Mar-2010 10:22:47] <fus10nx> zProperties is like incred. important to the whoel system

[31-Mar-2010 10:22:50] <fus10nx> ya

[31-Mar-2010 10:23:01] <rmatte> yeh, zProperties is the bread and butter when it comes to configuring

[31-Mar-2010 10:23:15] <fus10nx> I'm going to work on the map/locations stuff shortly

[31-Mar-2010 10:23:18] <fus10nx> wanna populate that data

[31-Mar-2010 10:24:54] <fus10nx> so it's ajax based right?

[31-Mar-2010 10:25:04] <fus10nx> if i have the dash up and theres an error i dont need to refresh my screen

[31-Mar-2010 10:26:00] <fus10nx> its funny it detects how many cores I have but not the manufacturer, model, speed, ext speed, l1, l2, volts

[31-Mar-2010 10:26:05] <fus10nx> it just all shows as unknown or 0

[31-Mar-2010 10:26:19] <fus10nx> but it detects the number or cores fine cus i am running this on 6 different machinesd now

[31-Mar-2010 10:33:58] <rmatte> yes, there's Ajax in the UI

[31-Mar-2010 10:34:18] <rmatte> the dashboard auto-refreshes every 5 mins by default

[31-Mar-2010 10:34:42] <rmatte> it detects all that info for certain models of CPU, but not all

[31-Mar-2010 10:35:05] <rmatte> it detects the number of physical CPUs, not cores, from what I recall

[31-Mar-2010 10:39:00] <fus10nx> well its a single quad core, another single quad core then the rest are dual core

[31-Mar-2010 10:39:16] <fus10nx> as in 1 physical CPU in each machine

[31-Mar-2010 10:39:30] <rmatte> ah

[31-Mar-2010 10:39:35] <fus10nx> so when reading the admin guide i remember it saying something about install some other SNMP software, did you do that?

[31-Mar-2010 10:39:38] <fus10nx> some 3rd party software

[31-Mar-2010 10:39:46] <rmatte> I guess it was hyperthreading that I found threw it off

[31-Mar-2010 10:40:19] <fus10nx> ya, not sure why it's not getting the model numbers of any but eh, its fine

[31-Mar-2010 10:40:38] <fus10nx> the WMI stuff must be working because its getting my events and stuff

[31-Mar-2010 10:43:04] <rmatte> well, event log monitoring is included in Zenoss core

[31-Mar-2010 10:43:10] <rmatte> it's not part of that WMI pack

[31-Mar-2010 10:50:02] <fus10nx> hmm ok

[31-Mar-2010 10:50:10] <fus10nx> well do you know that 3rd party thing i'm refering to?

[31-Mar-2010 10:50:48] <fus10nx> basically i did the zProperties for the /Server/Windows class and added in a bunch of community.wmi items