[01-Apr-2010 00:00:30] [connected at Thu Apr  1 00:00:30 2010]

[01-Apr-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[01-Apr-2010 03:02:06] <fus10nx> Hey Simon4

[01-Apr-2010 03:06:43] <Simon4> heya fus10nx

[01-Apr-2010 03:07:56] <fus10nx> are you in the UK as well?

[01-Apr-2010 03:09:42] <Simon4> yeah

[01-Apr-2010 03:10:42] <fus10nx> how many devices is your ZenOSS monitoring ?

[01-Apr-2010 03:11:12] <Simon4> about 300 presently

[01-Apr-2010 03:11:16] <Simon4> will be much more soon

[01-Apr-2010 03:12:09] <fus10nx> are you doing anything w/ WMI ?

[01-Apr-2010 03:12:58] <Simon4> nothing sorry

[01-Apr-2010 03:13:04] <Simon4> we're all snmp

[01-Apr-2010 03:13:15] <Simon4> some of the US based guys in here do a bit of the WMI stuff I believe

[01-Apr-2010 03:13:17] <fus10nx> are you using the SNMP Informant on each machine as well ?

[01-Apr-2010 03:13:31] <Simon4> I haven't got any windows boxes to monitor

[01-Apr-2010 03:13:36] <fus10nx> ahhhh i see

[01-Apr-2010 03:13:45] <Simon4> I've used snmp informant in the past I think, long times ago

[01-Apr-2010 03:14:08] <fus10nx> ya, it just seems that WMI can collect more information w/o installing additonal software.

[01-Apr-2010 03:14:22] <Simon4> yeah, if I did it now I would use WMI

[01-Apr-2010 03:14:27] <fus10nx> really?

[01-Apr-2010 03:14:37] <Simon4> native data collection is win for me vs installing other software

[01-Apr-2010 03:15:00] <Simon4> note that I'm saying this without actually trying it, so reality may force me in the other direction ;)

[01-Apr-2010 03:15:00] <fus10nx> ya -- WMI has been a bit of a pain to get setup for me, get errors all over the place but trying to work it out

[01-Apr-2010 03:15:12] <Simon4> but I would certainly start with just WMI

[01-Apr-2010 03:15:38] <fus10nx> does your Zen properly detect CPU;s?

[01-Apr-2010 03:15:51] <Simon4> on linux devices, yeah

[01-Apr-2010 03:15:56] <fus10nx> I get lots of unknowns and rmatte was mentioned its because the CPU library in Zen is very small

[01-Apr-2010 03:16:08] <Simon4> I think - lemme look

[01-Apr-2010 03:16:47] <Simon4> so I get manufacturer unknown, model       Intel Xeon CPU L5430 at 2.66GHz

[01-Apr-2010 03:16:59] <Simon4> speed and cache are all correct

[01-Apr-2010 03:20:58] <fus10nx> interesting

[01-Apr-2010 03:21:07] <fus10nx> ya i just get unknowns on 2 of my quad core devices for everything

[01-Apr-2010 03:21:22] <fus10nx> but its alright, ill have a play w/ that later down the line, not too worried about it at the moment, more concerned w/ getting WMI stuff to work properly

[01-Apr-2010 03:34:03] <fus10nx> Simon4: let's say we have a device, a DLP Cinema Projector and it ships with a MIB file, can you just upload that MIB file to Zen ?

[01-Apr-2010 03:34:15] <Simon4> yup

[01-Apr-2010 03:34:24] <fus10nx> wow, why even buy what's up gold then

[01-Apr-2010 03:34:28] <fus10nx> this software is 10x better

[01-Apr-2010 03:34:38] <Simon4> oui :)

[01-Apr-2010 03:34:42] <fus10nx> haha

[01-Apr-2010 03:34:58] <fus10nx> do you use a lot of ZenPacks?

[01-Apr-2010 03:35:34] <Simon4> we have about 10 loaded I think, some from the community, some that we have written in-house

[01-Apr-2010 03:36:09] <fus10nx> thats cool...I still dont fully get how to get them working. The WMI ones require that we put our devices in an already made Class "/Devices/CMI/WMI"

[01-Apr-2010 03:36:36] <fus10nx> so i've just installed another but dont see it in the zProperties then zCollectorPlugins

[01-Apr-2010 03:39:09] <Simon4> you may be able to change that class by a) editing the zenpack before instlaling it, or b) looking at the zproperties for that /Devices/CMI/WMI class and cloning them elsewhere

[01-Apr-2010 03:39:38] <fus10nx> ya I'll certainly look into that

[01-Apr-2010 03:39:40] <Simon4> however if it's doing UI mods (extra tabs for devices etc) you'll need to stay under that tree structure or edit the zenpack, since it's defining a new class type

[01-Apr-2010 03:45:44] <fus10nx> thanks ill take a look. I've just figured out "Win Services" which is good

[01-Apr-2010 03:46:51] <Simon4> sweet

[01-Apr-2010 03:46:58] <fus10nx> where abouts in the UK are you?

[01-Apr-2010 03:47:23] <Simon4> London

[01-Apr-2010 03:47:35] <fus10nx> oh nice, I'm there quite often but I'm in the East Midlands

[01-Apr-2010 03:47:42] <fus10nx> was in Bath a few days ago, great place

[01-Apr-2010 03:47:51] <Simon4> ah cool, I still need to get there

[01-Apr-2010 03:48:29] <fus10nx> ive realized that most people form the UK dont travel much within the UK

[01-Apr-2010 03:48:49] <Simon4> heh, I'm a kiwi living over here, but have been a bit lazy for travel of late

[01-Apr-2010 03:48:49] <fus10nx> this girl at my office has lived here for ~30 years and has never been to London, I was shocked

[01-Apr-2010 03:48:53] <Simon4> woah

[01-Apr-2010 03:48:55] <fus10nx> haha ohhh ok

[01-Apr-2010 03:48:59] <Simon4> that's fairly special

[01-Apr-2010 03:49:00] <fus10nx> Ya I moved over from the states

[01-Apr-2010 03:49:13] <fus10nx> ya most cases aren't that severe but i've heard some interesting ones

[01-Apr-2010 03:49:23] <fus10nx> my first task here was traveling from the top (Scotland) allllllll the way down

[01-Apr-2010 03:49:28] <fus10nx> and stoping everywhere in between

[01-Apr-2010 04:02:19] <fus10nx> any idea how often it polls services to see if their up or down ?

[01-Apr-2010 04:02:44] <Simon4> once every 60 seconds

[01-Apr-2010 04:02:51] <Simon4> if it's the same as the ipservice checks

[01-Apr-2010 04:02:53] <fus10nx> strange, i had to remodel to get it to notice the closing of a service

[01-Apr-2010 04:03:01] <fus10nx> ill keep an eye on it

[01-Apr-2010 04:03:19] <Simon4> heading officeways, back in a while

[01-Apr-2010 04:03:24] <fus10nx> take care !

[01-Apr-2010 04:14:24] <fus10nx> ya for some reason if I stop a service it doesnt show until I remodel and same when I start the service back up again

[01-Apr-2010 06:38:40] <yesnomaybe> Does anyone else have issues w/ Win Services? I manually stop SNMP Trap but it still shows as running on Zen

[01-Apr-2010 06:39:05] <yesnomaybe> I manually refresh and it sitll doesnt show. Only if I remodel does it show as stopped

[01-Apr-2010 06:45:33] <yesnomaybe> sort of defeats the whole purpose of monitoring !

[01-Apr-2010 06:46:20] <tehhobbit> yesnomaybe: have you waited 5 mins which is the default check interval ?

[01-Apr-2010 06:46:29] <yesnomaybe> oh yes, well over that

[01-Apr-2010 06:46:34] <tehhobbit> ok

[01-Apr-2010 06:46:35] <yesnomaybe> then i did an entirly fresh install

[01-Apr-2010 06:46:48] <yesnomaybe> frehs ubuntu install, fresh zenoss install, still the same thing

[01-Apr-2010 06:46:54] <yesnomaybe> if i remodel it, then it shows

[01-Apr-2010 06:47:49] <yesnomaybe> im literally just going under classes, then services, select the service under WInService then gioing to edit and turn monitor to true

[01-Apr-2010 06:48:04] <yesnomaybe> then on the status it shows what instances have it

[01-Apr-2010 06:48:27] <yesnomaybe> and if i go into the SNMP trap service (not just my device status page) it still shows it as up

[01-Apr-2010 08:24:33] <yesnomaybe> hey mrayzenoss, thanks for yesterdeay

[01-Apr-2010 08:24:38] yesnomaybe is now known as fus10nx

[01-Apr-2010 08:25:22] <mrayzenoss> no problem

[01-Apr-2010 08:26:03] <fus10nx> I have a question for you that you mighht be able to help me out with...for service, if I shut something down such as SNMP trap on a windows computer, Zen should detect that w/o me having to remodel the device, right?

[01-Apr-2010 08:26:52] <mrayzenoss> not sure I follow.  If you shut down the Windows SNMP service and it's being monitored, Zenoss will event on that

[01-Apr-2010 08:27:28] <fus10nx> SNMP Trap, not even the whole SNMP service. I see it as green as if it's OK and I've been trying it all day and the only way it seems to detect the issue is if I re-model

[01-Apr-2010 08:27:56] <fus10nx> manually refreshing, waiting upwards of an hour, etc. -- only restart Zen or re-modeling seems to detect it and this is the case on multiple machines

[01-Apr-2010 08:28:17] <fus10nx> bur really it should detect it within 5 minutes, right ?

[01-Apr-2010 08:30:26] <eidolon> hi folks - i was going on a bad assumption her e- but zenoss doesn't store machine information in mysql, does it?  only events?

[01-Apr-2010 08:30:28] <mrayzenoss> well the trap should be sent to Zenoss, which will event on it

[01-Apr-2010 08:30:34] <mrayzenoss> eidolon: correct

[01-Apr-2010 08:31:02] <eidolon> okay, is there a way to reference the machine inventory externally (ala, odbc or the like)?  we're looking at machine inventory / asset tracking in an automated way.

[01-Apr-2010 08:31:05] <mrayzenoss> fus10nx: do you have your SNMP traps configured to send them to your Zenoss server?

[01-Apr-2010 08:31:24] <mrayzenoss> eidolon: you can access it via the zendmd and query it however you like

[01-Apr-2010 08:31:37] <eidolon> hmm, zendmd is the api, yes?

[01-Apr-2010 08:31:47] <fus10nx> mrayzenoss: just a sec please

[01-Apr-2010 08:31:47] <mrayzenoss> zendmd is the CLI for the API

[01-Apr-2010 08:31:51] * eidolon nods

[01-Apr-2010 08:32:05] <mrayzenoss> but yeah, you can access the calls remotely as well

[01-Apr-2010 08:32:15] * eidolon nods

[01-Apr-2010 08:32:51] Absenth_AWAY is now known as Absenth

[01-Apr-2010 08:32:51] <mrayzenoss> and you can make custom calls that are accessible remotely if you need to

[01-Apr-2010 08:33:06] <eidolon> we're trying to see if we can use zenoss for machine inventory and tracking (asset management) - how do we run reports against it, etc etc. (yes we know there's reporting in zenoss... but it's targeted differently )

[01-Apr-2010 08:33:33] <fus10nx> mrayzenoss: ahh ok, that did it then, it was the traps, I see now, thanks

[01-Apr-2010 08:35:18] <mrayzenoss> eidolon: so you can call lots of things via REST already, but if you want to get specific data, you can upload a script and hang it off of a URL and call that

[01-Apr-2010 08:35:25] <mrayzenoss> eidolon: https://sourceforge.net/projects/zenossdeki/ has examples

[01-Apr-2010 08:36:06] <eidolon> oooo

[01-Apr-2010 08:36:09] * eidolon doesn't know mindtouch

[01-Apr-2010 08:36:27] <mrayzenoss> mindtouch is a wiki on steroids, a mashup engine

[01-Apr-2010 08:38:00] * eidolon nods

[01-Apr-2010 08:38:15] <eidolon> we use mediawiki (the best known, yet one of the worst of the wikis)

[01-Apr-2010 08:38:46] <fus10nx> mrayzenoss: do you know a lot about the WMI stuff that Egor has done ?

[01-Apr-2010 08:39:58] <fus10nx> wmi.ProcessorMap never seems to work on my windows 7 machine, then on an XP machine a few of them don't seem to work, it's quite strange

[01-Apr-2010 08:40:15] <mrayzenoss> fus10nx: not really

[01-Apr-2010 08:40:22] <fus10nx> ill speak to him next time he's in here

[01-Apr-2010 08:40:25] <fus10nx> but thanks

[01-Apr-2010 08:40:30] <mrayzenoss> fus10nx: I know the gaps that he's filling in

[01-Apr-2010 08:40:52] <fus10nx> ya...I just like the fact of not having to install another 3rd party piece of software (SNMP Informant) and just using WMI is nice

[01-Apr-2010 08:43:20] <fus10nx> mrayzenoss: are you in the same timezone as Charles ?

[01-Apr-2010 08:44:22] <mrayzenoss> I don't know where Charles is :)

[01-Apr-2010 08:44:38] <mrayzenoss> I'm CDT in the US (Austin), he's been in this office before

[01-Apr-2010 08:44:46] <mrayzenoss> he may be completely virtual

[01-Apr-2010 08:44:59] <fus10nx> ahhh ok !

[01-Apr-2010 08:45:02] <fus10nx> CST time he is

[01-Apr-2010 08:45:10] <mrayzenoss> our other office is in Annapolis (EDT)

[01-Apr-2010 08:45:10] <fus10nx> how many people are at Zen

[01-Apr-2010 08:45:16] <mrayzenoss> 50-60

[01-Apr-2010 08:45:21] <fus10nx> oh wow

[01-Apr-2010 08:45:36] <fus10nx> i suppose most clients are the core ones

[01-Apr-2010 08:46:53] <mrayzenoss> yeah, since we're commercial open source, the vast majority of our users are using Zenoss Core

[01-Apr-2010 08:50:12] <mistich> good morning all

[01-Apr-2010 08:50:27] <mrayzenoss> morning

[01-Apr-2010 08:50:43] <mistich> how do I add a property to a class in a zenpack?

[01-Apr-2010 08:50:45] <fus10nx> morning mistich

[01-Apr-2010 08:51:37] <mistich> such as adding a second email address in the UserSettings class

[01-Apr-2010 08:51:56] <mistich> here is the code from the class

[01-Apr-2010 08:51:57] <mistich>     _properties = ZenModelRM._properties + (

[01-Apr-2010 08:51:57] <mistich>         {'id':'email', 'type':'string', 'mode':'w'},

[01-Apr-2010 08:51:57] <mistich>         {'id':'pager', 'type':'string', 'mode':'w'},

[01-Apr-2010 08:51:57] <mistich>         {'id':'defaultPageSize', 'type':'int', 'mode':'w'},

[01-Apr-2010 08:51:58] <mistich>         {'id':'defaultEventPageSize', 'type':'int', 'mode':'w'},

[01-Apr-2010 08:51:58] <mistich>         {'id':'defaultAdminRole', 'type':'string', 'mode':'w'},

[01-Apr-2010 08:51:59] <mistich>         {'id':'defaultAdminLevel', 'type':'int', 'mode':'w'},

[01-Apr-2010 08:51:59] <mistich>         {'id':'oncallStart', 'type':'int', 'mode':'w'},

[01-Apr-2010 08:52:00] <mistich>         {'id':'oncallEnd', 'type':'int', 'mode':'w'},

[01-Apr-2010 08:52:00] <mistich>         {'id':'escalationMinutes', 'type':'int', 'mode':'w'},

[01-Apr-2010 08:52:01] <mistich>         {'id':'dashboardState', 'type':'string', 'mode':'w'},

[01-Apr-2010 08:52:01] <mistich>         {'id':'netMapStartObject', 'type':'string', 'mode':'w'},

[01-Apr-2010 08:52:02] <mistich>         {'id':'eventConsoleRefresh', 'type':'boolean', 'mode':'w'},

[01-Apr-2010 08:52:02] <mistich>         {'id':'zenossNetUser', 'type':'string', 'mode':'w'},

[01-Apr-2010 08:52:03] <mistich>         {'id':'zenossNetPassword', 'type':'string', 'mode':'w'},

[01-Apr-2010 08:52:06] <fus10nx> mistich: pastebin :)

[01-Apr-2010 08:52:23] <fus10nx> is it OK to mix SNMP with WMI ? there won't be a conflict when Zen tries to report on things, will there?

[01-Apr-2010 08:53:36] <mistich> http://pastebin.com/ZwbYAbFz

[01-Apr-2010 08:53:37] <mrayzenoss> fus10nx: you can safely mix protocols, as long as they're not trying to write to the same thing

[01-Apr-2010 08:53:39] <mrayzenoss> the /Server/Windows device class already mixed WMI and SNMP

[01-Apr-2010 08:53:53] <fus10nx> ok, I'll be doing it to the /Devices/CIM/WMI

[01-Apr-2010 08:54:16] <mistich> such as adding a second email address in the UserSettings class

[01-Apr-2010 08:54:16] <mistich> here is the code from the class http://pastebin.com/ZwbYAbFz

[01-Apr-2010 08:54:21] <fus10nx> actually, is it simple enough to "clone" everything from /Devices/CIM/WMI to /Devices/Servers/Windows

[01-Apr-2010 08:55:21] <mrayzenoss> fus10nx: I usually discourage overwriting the default device classes

[01-Apr-2010 08:55:28] <mrayzenoss> so perhaps /Server/Windows/CIM

[01-Apr-2010 08:56:32] <fus10nx> sure sure that's fine

[01-Apr-2010 08:56:46] <fus10nx> really i just want the features from WMI in the /Server/Windows

[01-Apr-2010 08:56:55] <fus10nx> so Egor's stuff works but its not a huge deal to be fair

[01-Apr-2010 09:00:01] <mistich> anyone how do I add a property to a class in a zenpack?

[01-Apr-2010 09:01:35] <mrayzenoss> mistich: 1 sec

[01-Apr-2010 09:01:39] <mistich>  ok

[01-Apr-2010 09:02:25] <mrayzenoss> just looking for an example, since I've seen this before

[01-Apr-2010 09:02:35] <mistich> cool

[01-Apr-2010 09:05:50] <mrayzenoss> http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.WMIPerf_MSSQL/ZenPacks/community/WMIPerf_MSSQL/__init__.py

[01-Apr-2010 09:06:00] <mrayzenoss> just put it in the __init__.py

[01-Apr-2010 09:06:10] <mrayzenoss> that's the simplest example I could find

[01-Apr-2010 09:06:46] <mistich> I'll try it thanks and that will extend  the existing class or create a new one

[01-Apr-2010 09:06:48] <chudler> a tip, you can modify properties at runtime with

[01-Apr-2010 09:07:08] <chudler> theOrganizer.setZenProperty('zMyProperty', value)

[01-Apr-2010 09:07:25] <mrayzenoss> mistich: I believe the technique I showed you adds a global zProperty

[01-Apr-2010 09:08:19] <mistich> zProperty I'm don't need I need a class property

[01-Apr-2010 09:08:29] <rmatte> what are you talking about?

[01-Apr-2010 09:08:45] <rmatte> the properties of a class are set in zProperties

[01-Apr-2010 09:08:50] <rmatte> I don't get what you're asking?

[01-Apr-2010 09:08:50] <mistich> zProperty I don't need I need to extend a class property

[01-Apr-2010 09:08:59] <rmatte> define: class property

[01-Apr-2010 09:09:02] <rmatte> what are you referring to?

[01-Apr-2010 09:10:05] <rmatte> explain exactly what you're trying to do in detail so that we understand

[01-Apr-2010 09:10:13] <mistich> getting a example

[01-Apr-2010 09:10:15] <rmatte> k

[01-Apr-2010 09:12:16] <mistich> this code here http://dev.zenoss.org/trac/browser/trunk/Products/ZenModel/UserSettings.py after line 583 I want to add the following code {'id':'smsemail', 'type':'string', 'mode':'w'}, but do it in a zenpack

[01-Apr-2010 09:12:31] <mistich> I'm trying to turn some patches I have into zenpacks

[01-Apr-2010 09:13:33] <mistich> does that make more sense

[01-Apr-2010 09:13:56] <rmatte> you would need to put a line in __init__.py with executes the patch command and runs those patches

[01-Apr-2010 09:14:07] <rmatte> and package the patches wherever in the ZenPack directory

[01-Apr-2010 09:14:21] <rmatte> though it's kind of weird that you're including patches as part of a ZenPack

[01-Apr-2010 09:14:24] <mistich> but I don't want to run the patch

[01-Apr-2010 09:14:28] <rmatte> why not just include them separately?

[01-Apr-2010 09:14:40] <mistich> I want to monkey patch the class

[01-Apr-2010 09:14:59] <rmatte> like, dynamically patch it?

[01-Apr-2010 09:15:15] <mistich> now were talking

[01-Apr-2010 09:15:18] <mistich> yes

[01-Apr-2010 09:15:33] <rmatte> why wouldn't you just create the patch files, include them in the pack, then use sed to replace what you need to replace?

[01-Apr-2010 09:15:42] <rmatte> and then run the patch files after they have been modified?

[01-Apr-2010 09:16:04] <rmatte> just write up a bash script that does it, include it in the pack, and have the __init__.py execute it

[01-Apr-2010 09:16:09] <mistich> because there is so many of there when I upgrade a pain to patch

[01-Apr-2010 09:16:31] <mistich> then when we apply zenoss patches makes things worse

[01-Apr-2010 09:16:48] <rmatte> k, you lost me on that last sentence

[01-Apr-2010 09:17:01] <rmatte> "because there is so many of there"

[01-Apr-2010 09:17:01] <rmatte> ?

[01-Apr-2010 09:17:03] <mistich> so trying to do it without touching the true source so when we upgrade there are no issues

[01-Apr-2010 09:17:07] <mistich> them

[01-Apr-2010 09:17:23] <rmatte> oh

[01-Apr-2010 09:17:33] * rmatte has no idea

[01-Apr-2010 09:17:38] <rmatte> you'd have to ask an actual dev methinks

[01-Apr-2010 09:17:53] <mistich> I have like 5 patches I'm turning into zenpacks

[01-Apr-2010 09:18:18] <mistich> yeah thats what I figured but thought I'd ask anyway

[01-Apr-2010 09:18:18] <rmatte> I don't really see how you can patch something without patching the original source

[01-Apr-2010 09:19:00] <rmatte> you'll have problems when you do an upgrade no matter what, which is why re-applying the patches should be part of your upgrade steps

[01-Apr-2010 09:19:54] <mistich> it is just thought it might be nice to make a zenpack

[01-Apr-2010 09:20:07] <rmatte> yes, but I don't see how that's feasible

[01-Apr-2010 09:20:24] <mistich> I have everything for the most part working in a zenpack but a couple of lines of code

[01-Apr-2010 09:20:30] <rmatte> you can make a ZenPack that does a straight patch, but when it comes to Zenoss upgrades you would need to re-apply it

[01-Apr-2010 09:20:42] <mistich> @monkeypatch('Products.ZenModel.UserSettings.UserSettingsManager')

[01-Apr-2010 09:20:57] <mistich> is a wonderful method

[01-Apr-2010 09:21:49] <rmatte> I've never really worked with monkeypatches, not I'm not familliar with them

[01-Apr-2010 09:21:56] <rmatte> so I'm not*

[01-Apr-2010 09:24:13] <mistich> thanks

[01-Apr-2010 09:24:38] <mistich> the developers are here at 11 today

[01-Apr-2010 09:24:46] <mistich> is that correct

[01-Apr-2010 09:25:53] <mrayzenoss> yes

[01-Apr-2010 09:26:02] <mrayzenoss> assuming you're in San Antonio, 10am

[01-Apr-2010 09:26:31] <mistich> thanks

[01-Apr-2010 09:31:31] <rmatte> oh, monkeypatches actually dynamically patch code at runtime

[01-Apr-2010 09:31:33] <rmatte> I see

[01-Apr-2010 09:32:33] <mrayzenoss> yeah, monkeypatching is a technical term :)

[01-Apr-2010 09:32:40] <mistich> but what I can tell it only will add a function

[01-Apr-2010 09:32:49] <mrayzenoss> see also duck-typing

[01-Apr-2010 09:33:55] <rmatte> mrayzenoss: yeh, I'd heard of them before but hadn't looked in to it.

[01-Apr-2010 09:34:29] <mrayzenoss> it can make it really hard to debug code, if what you're looking at isn't the actual code

[01-Apr-2010 09:34:39] <rmatte> I can imagine

[01-Apr-2010 09:35:34] <mistich> but will save time not having to review code to make sure the line numbers didn't change

[01-Apr-2010 09:35:47] <mistich> just dumps the new method into the class

[01-Apr-2010 09:40:01] <otakup0pe> ok i can't be the only person who saw xkcd today and then immediately started making a zenoss interface

[01-Apr-2010 09:40:59] <Brixius> I think I'm going to like Egor's WMIDataSource... I can get rid of a bunch of script files I wrote that use wmic and return data using nagios style output.

[01-Apr-2010 09:42:28] <mrayzenoss> otakup0pe: awesome

[01-Apr-2010 09:42:45] <mrayzenoss> otakup0pe: if it could be mashed up with zendmd

[01-Apr-2010 09:42:56] <mrayzenoss> where you could pull rrd graphs up...

[01-Apr-2010 09:43:12] <otakup0pe> haha wow i didn't even know about zendmd

[01-Apr-2010 09:43:24] <otakup0pe> yeah i just happend to write a django app last night to cycle through things from the rrdserver

[01-Apr-2010 09:43:40] <otakup0pe> and it was easy as pie to tie into this js shell thing hehe.

[01-Apr-2010 09:48:19] <rmatte> yup

[01-Apr-2010 09:48:52] <rmatte> zendmd is awesome though

[01-Apr-2010 09:54:19] <Brixius> I wonder if google ever thought they would create a DOS attack against the topeka city website by there April Fools joke..

[01-Apr-2010 09:55:00] <otakup0pe> oh question for zenoss people :3 i managed to bump up the snmpd timing to 30 and my datapoints went throught the roof (as expected)

[01-Apr-2010 09:55:08] <otakup0pe> but how do i increase the resolution of the graphs themselves ?

[01-Apr-2010 09:55:14] <otakup0pe> is that a manual rrdserver config tweak or something

[01-Apr-2010 09:56:14] <mrayzenoss> otakup0pe: yeah, you'll need to adjust your RRD settings

[01-Apr-2010 09:56:19] <mrayzenoss> collector->localhost->edit

[01-Apr-2010 09:56:24] <otakup0pe> hmm

[01-Apr-2010 09:56:41] <otakup0pe> oh son of a

[01-Apr-2010 09:56:43] <otakup0pe> thanks :)

[01-Apr-2010 09:56:51] <mrayzenoss> the existing graphs will need to be updated or removed and regenerated

[01-Apr-2010 09:58:06] <venturaville> zenoss saved us this morning

[01-Apr-2010 09:58:35] <venturaville> especially my netapp zenpack (I'm happy)

[01-Apr-2010 09:58:52] <rmatte> cool

[01-Apr-2010 09:58:55] <mrayzenoss> hmm… I guess I should hurry up and publish the latest version of that

[01-Apr-2010 09:59:11] <venturaville> the stacked NFSv3 graph is killer

[01-Apr-2010 09:59:35] <venturaville> though from what someone else said there is also a latency version of that same thing in SNMP somewhere

[01-Apr-2010 09:59:41] <venturaville> going to go hunt for it

[01-Apr-2010 09:59:42] <rmatte> mrayzenoss: I renamed ZenPacks.Nova.Windows.SNMPPerfMonitor.Simple to ZenPacks.Nova.Windows.SNMPPerfMonitorSimple... how do you want to handle that in svn?

[01-Apr-2010 09:59:59] <rmatte> the paths are obviously all different

[01-Apr-2010 10:00:00] <mrayzenoss> SVN has a rename

[01-Apr-2010 10:00:13] <rmatte> right, but the commit to svn won't be accurate

[01-Apr-2010 10:00:14] <mrayzenoss> oh yeah… I guess I'll have to redo the paths

[01-Apr-2010 10:00:26] <rmatte> can I just send you the packs?

[01-Apr-2010 10:00:53] <rmatte> (the updated versions)

[01-Apr-2010 10:01:42] <mrayzenoss> sure

[01-Apr-2010 10:01:51] <rmatte> k, firing them over shortly

[01-Apr-2010 10:03:10] <rmatte> sent

[01-Apr-2010 10:03:32] <rmatte> those are wayyyy nicer than the packs currently listed on the page

[01-Apr-2010 10:03:44] <rmatte> ;)

[01-Apr-2010 10:04:04] <rmatte> and once I figure out how the heck to write a collector plugin for them it'll be even nicer

[01-Apr-2010 10:04:20] <rmatte> anyways, I'm under the gun for some stuff today so I need to get back to it

[01-Apr-2010 10:04:21] <rmatte> ttyl

[01-Apr-2010 10:05:10] <rmatte> (when I figure out the collector plugin I'm probably just going to merge those 2 packs in to 1

[01-Apr-2010 10:07:39] <cgibbons> hmmm

[01-Apr-2010 10:07:58] <mrayzenoss> jplouis is the dev on call, there are a couple more devs in the channel if you have questions

[01-Apr-2010 10:08:21] <mrayzenoss> also, today is a QA Test Day in #zenoss-testing if you want to help beat on our source builds

[01-Apr-2010 10:26:06] <mrayzenoss> this may be the quietest IRC session ever

[01-Apr-2010 10:27:37] <venturaville> I would contribute to the noise, but I am on-call this week so I'm a bit tied up :-P

[01-Apr-2010 10:35:54] <mistich> are there developers her now

[01-Apr-2010 10:36:43] <mistich> this code here http://dev.zenoss.org/trac/browser/trunk/Products/ZenModel/UserSettings.py after line 583 I want to add the following code {'id':'smsemail', 'type':'string', 'mode':'w'}, but do it in a zenpack

[01-Apr-2010 10:40:00] <mistich> knock knock

[01-Apr-2010 10:40:55] <cgibbons> wakie wakie

[01-Apr-2010 10:41:01] <jplouis> looking...

[01-Apr-2010 10:42:34] <jplouis> you can monkey patch the class

[01-Apr-2010 10:43:15] <mistich> yes I can monkey patch def but how do I do the code I posted

[01-Apr-2010 10:46:49] <jplouis> in a zenpack in __init__.py add the following:

[01-Apr-2010 10:46:49] <jplouis> from Products.ZenModel.UserSettings import UserSettings

[01-Apr-2010 10:46:50] <jplouis> EventManagerBase._properties += (

[01-Apr-2010 10:46:50] <jplouis>     {'id':'smsemail', 'type':'string', 'mode':'w'}

[01-Apr-2010 10:46:50] <jplouis>     )

[01-Apr-2010 11:07:39] <mrayzenoss> any solaris users here?

[01-Apr-2010 11:14:40] <mrayzenoss> hope everyone has their iPads ready: blogs/zenossblog/2010/04/01/zenoss-drops-linux-installs-moves-to-ipad

[01-Apr-2010 11:15:43] <rmatte> lol, almost forgot it was april fool's

[01-Apr-2010 11:15:55] <mistich> lets do it for real would love to try it in Dallas datacenter

[01-Apr-2010 11:16:01] <mistich> will you support it

[01-Apr-2010 11:16:17] <rmatte> lol

[01-Apr-2010 11:20:39] <mistich> any developers still around that want to look a zenpack for me can't find why it is not working

[01-Apr-2010 11:29:20] <mistich> jplouis still around

[01-Apr-2010 11:29:57] <mrayzenoss> he went to lunch, but I'll ask him to stop by when he gets back

[01-Apr-2010 11:30:09] <mistich> yeah his code didn't work

[01-Apr-2010 11:30:13] <mistich> from Products.ZenModel.UserSettings import UserSettings

[01-Apr-2010 11:30:13] <mistich> EventManagerBase._properties += (

[01-Apr-2010 11:30:13] <mistich>     {'id':'smsemail', 'type':'string', 'mode':'w'}

[01-Apr-2010 11:30:13] <mistich>     )

[01-Apr-2010 11:30:25] <mistich> ok thanks

[01-Apr-2010 11:30:27] <mrayzenoss> there's a copy paste error

[01-Apr-2010 11:30:40] <mrayzenoss> EventManagerBase should be UserSettings

[01-Apr-2010 11:31:23] <mistich> thought so but get a different error

[01-Apr-2010 11:31:27] <mistich> 2010-04-01T11:31:01 ERROR zen.ZenRelations Unable to set property descriptors: TypeError: can only concatenate tuple (not "dict") to tuple

[01-Apr-2010 11:32:23] <ska> Is there a free version of Zenoss, and what restrictions does it have?

[01-Apr-2010 11:32:46] <bigegor> mistich: {'id':'smsemail', 'type':'string', 'mode':'w'},

[01-Apr-2010 11:33:10] <mistich> ?

[01-Apr-2010 11:33:11] <chudler> Suppose I have a graph report, and I want the graph default "Range" to be maybe "Weekly" instead of "Hourly".  I couldn't find this option. ?

[01-Apr-2010 11:33:37] <mistich> I see the comma

[01-Apr-2010 11:33:40] <bigegor> mistich: comma at the end

[01-Apr-2010 11:34:06] <mistich> should have not drank so much last night

[01-Apr-2010 11:38:16] <rmatte> ska: Zenoss core

[01-Apr-2010 11:38:45] <rmatte> ska: The only real restrictions are that you don't have access to all of the exclusive Enterprise ZenPacks...

[01-Apr-2010 11:39:07] <rmatte> Enterprise also has built in functionality for configuring collectors and some other stuff

[01-Apr-2010 11:39:15] <rmatte> and there's also the support that you receive with enterprise

[01-Apr-2010 11:39:23] <rmatte> other than that, Zenoss Core is fairly full featured

[01-Apr-2010 11:42:10] * Simon4 snorks at mr Hinkle's tweet

[01-Apr-2010 11:42:38] <Simon4> blogs/zenossblog/2010/04/01/zenoss-drops-linux-installs-moves-to-ipad for those who haven't seen it

[01-Apr-2010 11:46:08] <rmatte> seen it

[01-Apr-2010 11:46:09] <rmatte> lol

[01-Apr-2010 11:47:25] <chudler> Of course it is apparently non-trivial to make it do what I want ARG

[01-Apr-2010 11:48:42] <chudler> blogs/zenossblog/2009/04/28/tip-of-the-month-changing-default-performance-graph-ranges/ works globally, but probably no easy way for an end-user to apply it per-graph/report?

[01-Apr-2010 11:54:44] <rmatte> chudler: nope, it's per collector

[01-Apr-2010 11:54:58] <rmatte> If you only want it on certain devices you need to setup another collector

[01-Apr-2010 11:57:50] <chudler> what?  are you talking about graph reports?

[01-Apr-2010 12:00:03] <chudler> for graph reports, I think it would be per zenoss install.  Obviously I'm not going to install another entire zenoss for doing somebody's report.  I'll just tell them it is "impossible"

[01-Apr-2010 12:02:49] <rmatte> oh, I thought you were talking about the RRD settings for the way that data was collected rather than the way it is displayed

[01-Apr-2010 12:03:29] <citrus> anyways to send an email out after an alert has been acknowledged? saying its been acknowledged by user

[01-Apr-2010 12:04:35] <rmatte> citrus: that function is not built-in

[01-Apr-2010 12:04:43] <rmatte> you'd have to script something in python to do that

[01-Apr-2010 12:05:10] <rmatte> or log a feature request in trac

[01-Apr-2010 12:05:23] <rmatte> (probably wouldn't be added for ages though)

[01-Apr-2010 12:12:44] <venturaville> kudos on the ipad article

[01-Apr-2010 12:12:55] <Brixius> ska: have you found the free version yet?  community/download

[01-Apr-2010 12:12:55] <venturaville> now I just need that ported down the iphone :-P

[01-Apr-2010 12:13:31] <Simon4> venturaville: but wouldn't you want an iphone zenpack so that you can monitor all your staff's phones from your ipad farm? ;)

[01-Apr-2010 12:13:45] <venturaville> *light bulb goes off*

[01-Apr-2010 12:14:37] <venturaville> argh don't tempt me! :-)

[01-Apr-2010 12:14:58] <venturaville> though I have actually run a webserver off a sharp Zaurus before........

[01-Apr-2010 12:16:39] <rmatte> I've run a webserver off my iPhone

[01-Apr-2010 12:16:44] <rmatte> apache

[01-Apr-2010 12:16:52] <chudler> for like a really long time, I thought this was a bug:  http://dev.zenoss.org/trac/ticket/6391  Geez, apparently it's a feature?? wow

[01-Apr-2010 12:17:52] <rmatte> chudler: lol

[01-Apr-2010 12:18:12] <rmatte> chudler: if there's an option to toggle that, it's a feature, if there isn't, it's a bug

[01-Apr-2010 12:18:14] <rmatte> in my opinion

[01-Apr-2010 12:18:56] <chudler> the "feature", if I understand correctly, is that events continue to queue while in maintenance mode.  After leaving maintenance, guess what happens

[01-Apr-2010 12:18:57] <venturaville> sounds like a bug

[01-Apr-2010 12:19:13] <venturaville> or a half way implemented feature

[01-Apr-2010 12:19:27] <rmatte> chudler: ah

[01-Apr-2010 12:19:37] <chudler> agreed, it might be that it was half-way done, yes

[01-Apr-2010 12:51:05] <Simon4> right, off the clock, beer and code time :)

[01-Apr-2010 12:58:57] <citrus> i have a machine that says its up in zenoss,  never got an alert   but i phisically shut this machine down,  i tried pinging it from the zenoss box and from my box with no response.    why would zenoss say it still up?

[01-Apr-2010 13:00:53] <Simon4> do you have the zProperty of "zPingMonitorIgnore" set to true?

[01-Apr-2010 13:01:08] <Simon4> or productionState of "Maintenance" ?

[01-Apr-2010 13:01:20] <venturaville> or the IP address changed?

[01-Apr-2010 13:04:37] <citrus> false,  production

[01-Apr-2010 13:04:55] <citrus> the ip address i am pinging was taken from zenoss

[01-Apr-2010 13:05:18] <Simon4> is zenping running?

[01-Apr-2010 13:07:30] <rmatte> man I hate misconfigured network equipment

[01-Apr-2010 13:07:48] <rmatte> "commission this stuff" --- 15 minutes later --- "this is all crap, needs to be configured"

[01-Apr-2010 13:08:09] <mrayzenoss> nuke the site from orbit, the only way to be sure

[01-Apr-2010 13:08:48] <rmatte> citrus: did any ping alerts ever come in for that device in the past?

[01-Apr-2010 13:08:52] <citrus> zenping was running but 4 others weren't      starting them now and i will see if that fixes it

[01-Apr-2010 13:09:07] <rmatte> mrayzenoss: hehe

[01-Apr-2010 13:09:24] <rmatte> yeh, daemons not running is not a good thing  ;)

[01-Apr-2010 13:50:41] <citrus> Yeah it finally failed about 20 minutes after i started those services.

[01-Apr-2010 13:59:50] <rmatte> you might want to check the log files for those services to see why they crashed

[01-Apr-2010 14:43:21] * Simon4 discovers his logic is rubbish

[01-Apr-2010 14:43:27] <Simon4> oops, etc

[01-Apr-2010 14:48:43] cluther_ is now known as cluther

[01-Apr-2010 15:36:16] * Simon4 rewrites his zenpack logic to be a little more sensible

[01-Apr-2010 15:48:29] <Simon4> rmatte: feel like testing this for a laugh?

[01-Apr-2010 15:48:39] <rmatte> testing what?

[01-Apr-2010 15:48:46] <Simon4> bladechassis zenpack

[01-Apr-2010 15:49:30] <Simon4> actually, better I commit it first

[01-Apr-2010 15:49:51] <rmatte> oh

[01-Apr-2010 15:50:00] <rmatte> I don't have time to test that today, but I can test it next week

[01-Apr-2010 15:50:05] <Simon4> sweet

[01-Apr-2010 15:50:37] <Simon4> yeah, I realised I have a C3000 to test it against in weekend also (we generally have C7000's), so quitely ignore me :)

[01-Apr-2010 15:52:05] <rmatte> k hehe

[01-Apr-2010 16:45:40] <Simon4> sweet

[01-Apr-2010 16:45:41] <Simon4> http://simonandmarie.smugmug.com/Other/ZenPack-Development/11694904_QEBya#825145908_FCyMY

[01-Apr-2010 16:48:27] <Simon4> that'll do for tonight

[01-Apr-2010 16:57:36] <mrayzenoss> that looks slick

[01-Apr-2010 17:28:16] <rmatte> that is slick as hell

[01-Apr-2010 17:28:40] <mrayzenoss> RoAkSoAx: did you get a chance to play with the build some?

[01-Apr-2010 17:28:47] <rmatte> I'm out, cheers folks

[01-Apr-2010 17:28:52] <mrayzenoss> later

[01-Apr-2010 17:28:57] <mrayzenoss> I did a Ubuntu 9.10 build, found some issues

[01-Apr-2010 17:29:31] <RoAkSoAx> mrayzenoss, not yet, I've been busy with school stuff, but it is in my todo list for tomorrow

[01-Apr-2010 17:30:05] <mrayzenoss> no worries, I muddled through it, I need to post a document in the Ubuntu group describing my work

[01-Apr-2010 17:30:15] <mrayzenoss> found some more issues, which is good

[01-Apr-2010 17:31:57] <RoAkSoAx> mrayzenoss, if you do so for tomorrow i can compare my findings to them

[01-Apr-2010 17:33:04] <mrayzenoss> I'll go do that now

[01-Apr-2010 17:33:11] <RoAkSoAx> awesome

[01-Apr-2010 17:38:18] <mrayzenoss> docs/DOC-5837

[01-Apr-2010 17:38:35] <mrayzenoss> kinda rough, but easy enough to follow

[01-Apr-2010 17:38:46] <mrayzenoss> I guess emacs isn't required for building :)

[01-Apr-2010 17:39:37] <mrayzenoss> RoAkSoAx: we'll probably do source build debugging discussions in #zenoss-testing

[01-Apr-2010 17:39:55] <RoAkSoAx> mrayzenoss, nor svn i presume :P

[01-Apr-2010 17:40:51] <mrayzenoss> you'll need subversion for trunk builds

[01-Apr-2010 17:41:03] <mrayzenoss> and I usually build from the stable branch rather than the tarball

[01-Apr-2010 17:41:12] <mrayzenoss> since it's tarball+fixes

[01-Apr-2010 17:41:24] <RoAkSoAx> mrayzenoss, but do you need it for the build or just to obtain the source code?

[01-Apr-2010 17:41:41] <mrayzenoss> you know… I don't know

[01-Apr-2010 17:41:55] <mrayzenoss> I don't think I've ever built the tarball

[01-Apr-2010 17:42:32] <RoAkSoAx> mrayzenoss, well I guess for debian packaging we'll use the tarball :)

[01-Apr-2010 17:43:30] <mrayzenoss> so you're interested in Debian or Ubuntu?

[01-Apr-2010 17:43:58] <mrayzenoss> because there's a guy in #zenoss-testing who's interested in Debian as well

[01-Apr-2010 17:44:02] <mrayzenoss> and bzed is too

[01-Apr-2010 17:44:45] <mrayzenoss> and now that I'm paying attention, I noticed you posted in the Debian Group

[01-Apr-2010 17:44:59] <mrayzenoss> I'll post my Debian notes there :)

[01-Apr-2010 17:45:37] <bzed> RoAkSoAx: you don't want to use the tarball if you want to build a proper dbeian package

[01-Apr-2010 17:45:46] <bzed> RoAkSoAx: as the tarball contains an insane amount of code copies

[01-Apr-2010 17:45:59] <bzed> RoAkSoAx: or you repackage the tarball....

[01-Apr-2010 17:46:07] <bzed> more easy to pull from svn then imho

[01-Apr-2010 17:46:26] <bzed> also you want to split things like the plugins out into a source on their own

[01-Apr-2010 17:46:38] <bzed> as you don't want to need to update zenoss just for an update of the plugins

[01-Apr-2010 17:46:42] <bzed> and such things...

[01-Apr-2010 17:46:55] <bzed> look into the pkg-zenoss repository on alioth for a start

[01-Apr-2010 17:48:18] <RoAkSoAx> bzed, that's true... we *don't* have to use the tarball, you can just use the svn

[01-Apr-2010 17:48:27] <RoAkSoAx> it's just how you maintain the packaging

[01-Apr-2010 17:48:53] <RoAkSoAx> mrayzenoss, Ubuntu is debian packaging :)

[01-Apr-2010 17:49:16] <bzed> RoAkSoAx: hint: you're talking to a debian developer who spent a lot of time on packaging zenoss properly. its not as easy as you think.

[01-Apr-2010 17:49:46] <RoAkSoAx> bzed, i know it is not easy :)

[01-Apr-2010 17:50:09] <mrayzenoss> RoAkSoAx: yeah, going through Debian into Ubuntu is the logical way

[01-Apr-2010 17:51:08] <mrayzenoss> Debian Source Build Notes: docs/DOC-5838

[01-Apr-2010 17:51:15] <mrayzenoss> feel free to update and add your findings

[01-Apr-2010 17:51:36] <RoAkSoAx> mrayzenoss, yep, It is the same packaging, there can be some small changes or big ones, but we usually like to keep the delta as minimum as possible or, better yet, no delta

[01-Apr-2010 17:52:14] <bzed> RoAkSoAx: do you have an alioth account?

[01-Apr-2010 17:53:03] <mrayzenoss> updated the Ubuntu to reflect the use of 'sudo' :)

[01-Apr-2010 17:53:12] <mrayzenoss> I gotta run guys, see you tomorrow

[01-Apr-2010 17:55:04] <RoAkSoAx> bzed, roaksoax-guest

[01-Apr-2010 17:55:08] <RoAkSoAx> mrayzenoss, l8r :)

[01-Apr-2010 17:55:21] <bzed> RoAkSoAx: if you want to I could give you access to the pkg-zenoss team

[01-Apr-2010 17:55:33] <RoAkSoAx> bzed, that would be nice :)

[01-Apr-2010 17:55:41] <bzed> looking at the old stuff should be a good start, although I think most of it needs to be redone

[01-Apr-2010 17:55:56] <bzed> zope2.12 stuff is pretty different from 2.8...

[01-Apr-2010 17:57:48] <RoAkSoAx> bzed, so we'll pretty much need to repackage it

[01-Apr-2010 17:57:54] <bzed> yes

[01-Apr-2010 17:58:02] <bzed> add the proper dependencies

[01-Apr-2010 17:58:05] <bzed> even worse

[01-Apr-2010 17:58:21] <bzed> see if there are patches flowing aroudn which are necessary for some of the dependencies

[01-Apr-2010 17:58:26] <bzed> and which were not sent tp upstream

[01-Apr-2010 17:58:38] <bzed> look trough all the code written by the zenoss people

[01-Apr-2010 17:58:44] <bzed> see if the copyrights are all fine

[01-Apr-2010 17:58:49] <bzed> rip out non-free stuff

[01-Apr-2010 17:59:04] <bzed> there is an old patch for that, the developers were pretty helpful to do so

[01-Apr-2010 17:59:13] <bzed> package the non-free stuff seperate

[01-Apr-2010 17:59:31] <bzed> find a proper way to install everythng at the right place

[01-Apr-2010 17:59:43] <bzed> my guess is a week of work

[01-Apr-2010 17:59:48] <bzed> at least

[01-Apr-2010 17:59:59] <RoAkSoAx> maybe more

[01-Apr-2010 18:00:01] <bzed> yeah

[01-Apr-2010 18:00:22] <bzed> the last time I've started I've opened a can of worms

[01-Apr-2010 18:00:57] <bzed> especially related to copyrights

[01-Apr-2010 18:02:22] <RoAkSoAx> well i guess we'll see how this goes THEN

[01-Apr-2010 18:02:26] <bzed> :)

[01-Apr-2010 18:03:22] <RoAkSoAx> bzed, though i consider myself a beginner packager :) since I mostly work on doing merges from Debian to Ubuntu

[01-Apr-2010 18:04:10] <bzed> RoAkSoAx: that will be a rough start then :)

[01-Apr-2010 18:04:43] <RoAkSoAx> bzed, that is what makes it more intersting because its the place where you learn the most :)

[01-Apr-2010 18:04:44] <bzed> I've packaged a lot, but zenoss is still one of the biggest things I've ever worked on

[01-Apr-2010 18:05:03] <bzed> too bad it took so long to migrate to zope 2.12

[01-Apr-2010 18:05:11] <RoAkSoAx> bzed, i feel the same way

[01-Apr-2010 18:05:23] <bzed> I've added you to pkg-zenoss

[01-Apr-2010 18:05:28] <bzed> have a look at the svn repository

[01-Apr-2010 18:05:40] <bzed> probably create a new branch, I think we should keep the old one as reference

[01-Apr-2010 18:05:59] <bzed> basically most things need to be done from scratch

[01-Apr-2010 18:06:12] <bzed> but things like the dbconfig-common stuff could be copied

[01-Apr-2010 18:06:24] <bzed> same for ideas on how to package it

[01-Apr-2010 18:06:28] <RoAkSoAx> i know that couple years ago the packaging of zenoss for Ubuntu/Debian was discussed in an Ubuntu Developer Summit, but was decided not to go for it due to use of root account and it would have been useless for quite a few releases due to python

[01-Apr-2010 18:06:37] <RoAkSoAx> bzed, ok I;ll take a look at it :)

[01-Apr-2010 18:06:44] <bzed> heh

[01-Apr-2010 18:06:56] <bzed> see, the root account stuff was fixed in the debian package

[01-Apr-2010 18:07:41] <bzed> actually all issues except zope were fixed

[01-Apr-2010 18:07:48] <bzed> it failed with zope 2.09

[01-Apr-2010 18:08:03] <bzed> not in very visible ways, but it failed

[01-Apr-2010 18:08:06] <bzed> same for 2.10

[01-Apr-2010 18:08:13] <bzed> otherwise the package was just fine

[01-Apr-2010 18:09:17] <RoAkSoAx> i see

[01-Apr-2010 18:11:42] <RoAkSoAx> ok then I guess i'll just have to import the new zenoss branch, take a look to your packaging and start working on it :)

[01-Apr-2010 20:14:16] <bzed> RoAkSoAx: it might make sense to use git and use git-svn on the zenoss stuff to import it and do the proper things with it

[01-Apr-2010 20:14:19] <bzed> not sure

[01-Apr-2010 21:54:38] <Kris__> I have two network interfaces, eth0 and eth1.  eth0 is private, eth1 is public addressing.  eth0 works, when I add the IP, GW and NM for eth1, the routing fails.  Any ideas?

[01-Apr-2010 22:25:09] <rhettardo> have you tried remodeling the device after the changes?

[01-Apr-2010 22:25:16] <rhettardo> and checking the route table via snmpwalk?

[02-Apr-2010 00:00:30] [disconnected at Fri Apr  2 00:00:30 2010]

[02-Apr-2010 00:00:30] [connected at Fri Apr  2 00:00:30 2010]

[02-Apr-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[02-Apr-2010 00:00:45] <anthony.freenode.net> [freenode-info] please register your nickname...don't forget to auto-identify! http://freenode.net/faq.shtml#nicksetup

[02-Apr-2010 04:02:20] <Simon4> heh

[02-Apr-2010 04:02:21] <Simon4>  # TODO: document what this means and why we care

[02-Apr-2010 04:02:25] <Simon4> I love finding comments like that

[02-Apr-2010 05:03:10] <danuvius> good morning...

[02-Apr-2010 05:10:47] <danuvius> wich deamon do i need to restart with a higher verbosity level to debug graph problems?

[02-Apr-2010 05:10:55] <danuvius> rgraph render problems?

[02-Apr-2010 05:12:10] <Simon4> zenrender, generally

[02-Apr-2010 05:43:53] <danuvius> zenrender is not running..

[02-Apr-2010 05:44:02] <danuvius> and I do ha graphs..

[02-Apr-2010 05:44:13] <danuvius> ha=have

[02-Apr-2010 05:44:51] <Simon4> have a looksee in event.log

[02-Apr-2010 05:45:02] <Simon4> I get rrd warnings/errors in there by default

[02-Apr-2010 05:45:30] <danuvius> hmmm ok... No render errors there

[02-Apr-2010 05:45:33] <danuvius> :(

[02-Apr-2010 05:46:38] <danuvius> only "INFO ZODB.Conflict database conflict error" and "INFO RenderServer File: /tmp/renderserver/graph-"

[02-Apr-2010 05:47:12] <danuvius> I made a custom graph adding values from different rrd files..

[02-Apr-2010 05:47:44] <danuvius> any idea how to increase verbosity?

[02-Apr-2010 05:47:54] <Simon4> the other thing to do, is grab the url to the graph, and paste it in here: http://sethrh.webfactional.com/zenoss_rrd_helper/

[02-Apr-2010 05:48:07] <Simon4> that'll show you the exact command it's trying to run, which you can then play with on the commandline

[02-Apr-2010 05:50:59] <danuvius> thnx

[02-Apr-2010 06:29:51] <danuvius> found my error. Thnx Simon4

[02-Apr-2010 06:29:57] <Simon4> no probs

[02-Apr-2010 09:51:35] <chemist> a bit off topic

[02-Apr-2010 09:51:54] <chemist> but does anyone know what the default user/pass for Dell OMSA is?

[02-Apr-2010 09:58:03] <eidolon> hey folks - i need a simple dump of all hosts + all IP addresses from zenoss.  ist ehre an easy way to do that?  I fond /reports/device reports / all devices -> export all, but that doesn't include the IP address (which I know i could 'host' lookup, but that seems clunky)

[02-Apr-2010 09:58:16] <eidolon> (goal here is i need to map all our subnets)

[02-Apr-2010 10:00:01] <mrayzenoss> easy zendmd script

[02-Apr-2010 10:00:13] <mrayzenoss> for d in dmd.Devices.getSubDevices():

[02-Apr-2010 10:00:34] <mrayzenoss> print d.id d.manageIp

[02-Apr-2010 10:00:56] <mrayzenoss> you'll need to work the spacing, but that's it

[02-Apr-2010 10:01:12] <mrayzenoss> and probably the output formatting to make it nicer

[02-Apr-2010 10:01:50] <mrayzenoss> community/documentation/wiki/zendmd

[02-Apr-2010 10:14:09] <eidolon> thanks mray :)

[02-Apr-2010 10:17:04] <eidolon> just checking.  zendmd < thatscript.txt ?

[02-Apr-2010 10:20:10] <mrayzenoss> somewhere on the wiki it has how to run it as a stand-alone script

[02-Apr-2010 10:20:29] <mrayzenoss> docs/DOC-3217

[02-Apr-2010 10:21:22] <mrayzenoss> quick informal poll, what are your 5 most favorite community ZenPacks

[02-Apr-2010 10:22:32] <Simon4> can we vote for our own? ;)

[02-Apr-2010 10:22:50] <mrayzenoss> yeah, I'm just trying to get a feel for which ones are most popular

[02-Apr-2010 10:23:34] <Simon4> only two community ones we currently use are HPBladeChassis and PointThreshold

[02-Apr-2010 10:23:43] * eidolon dunno 'communtiy zen packs' - i'm an enterprise site :)

[02-Apr-2010 10:24:10] <mrayzenoss> yeah, I'm working on an effort to get community zenpacks "supported" for Enterprise

[02-Apr-2010 10:24:19] <mrayzenoss> need to know which ones to work on

[02-Apr-2010 10:24:46] <Simon4> mrayzenoss: what qualifies as "unsupported" ? do they still work etc?

[02-Apr-2010 10:25:00] <mrayzenoss> well… there are going to be 3 categories

[02-Apr-2010 10:25:14] <mrayzenoss> "Experimental", "Testing" and "Stable"

[02-Apr-2010 10:25:24] <mrayzenoss> (shamelessly borrowed from Debian)

[02-Apr-2010 10:25:36] <Simon4> ah k

[02-Apr-2010 10:25:37] <mrayzenoss> and Testing is going to be everything on the Community ZenPack page

[02-Apr-2010 10:25:55] <mrayzenoss> until it's gone through a few hoops around documentation

[02-Apr-2010 10:26:03] <mrayzenoss> and people actually using it

[02-Apr-2010 10:26:10] <ke4qqq> fping, wmi datasource, adv. device vetails

[02-Apr-2010 10:26:47] <mrayzenoss> oh yeah, fping.  I use that one myself, and the LinuxMonitor extensions

[02-Apr-2010 10:27:59] <mrayzenoss> the long-term goal is to eventually allow you to install "Stable" ZenPacks directly though the application, without having to go download them separately

[02-Apr-2010 10:28:42] <ke4qqq> zenoss doing package management - interesting

[02-Apr-2010 10:28:44] <mrayzenoss> and for Zenoss Support to tell customers they'll take their phone calls for "Stable" ZenPacks, as opposed to anything else they find on the internet

[02-Apr-2010 10:29:11] <mrayzenoss> ke4qqq: they're already versioned .eggs, it's not a big leap

[02-Apr-2010 10:29:36] <eidolon> hey mrayzenoss?

[02-Apr-2010 10:29:37] <eidolon> http://pastebin.stonekeep.com/6725

[02-Apr-2010 10:29:44] <eidolon> 'import globals' is failing from that page

[02-Apr-2010 10:29:58] <ckrough> everyone seems to like the event console.

[02-Apr-2010 10:30:00] <mrayzenoss> are you using the 'zenoss' user with the proper PYTHONPATH?

[02-Apr-2010 10:30:03] <ke4qqq> right - but that means (I think) that you end up with something above the dpkg/rpm layer which is what you have now and have a synaptic/packagekit/yum/$foo level

[02-Apr-2010 10:30:06] <jb> so, if you have a data source with a type of COUNTER (or DERIVE).. what happens if zenoss tries to poll the datasource and it can't because the device is down.

[02-Apr-2010 10:30:17] <eidolon> ummm.  probably not :)

[02-Apr-2010 10:30:20] <jb> it looks like its messing up the perf data, and counting it as a "0"

[02-Apr-2010 10:30:30] <jb> instead of just skipping

[02-Apr-2010 10:30:30] <Simon4> mrayzenoss: presumably installing "Test" zenpacks isn't going to hget you the cold shoulder from support on unrelated things?

[02-Apr-2010 10:30:37] <jb> which is totally screwing up my graphs :/

[02-Apr-2010 10:30:51] <Simon4> jb: it ends up as NaN when I rrddump rrd files

[02-Apr-2010 10:30:53] <eidolon> okay, that worked.

[02-Apr-2010 10:30:55] <eidolon> thanks mray :)

[02-Apr-2010 10:31:00] <Simon4> so behaving correctly

[02-Apr-2010 10:31:12] <jb> hmm

[02-Apr-2010 10:31:13] <mrayzenoss> Simon4: right now Support dislikes seeing community ZenPacks, because they have no idea how good/bad they are

[02-Apr-2010 10:31:21] <jb> i can't figure out why my graphs have negative values then

[02-Apr-2010 10:31:22] <Simon4> mrayzenoss: understood

[02-Apr-2010 10:31:30] <mrayzenoss> Simon4: so I'm putting together a framework for them to judge the perceived quality

[02-Apr-2010 10:31:41] <mrayzenoss> "Stable" will definitely be supported, I've gotten their sign-off on that

[02-Apr-2010 10:31:50] <mrayzenoss> "Testing" will still be murky

[02-Apr-2010 10:32:24] <mrayzenoss> and "Experimental" are those from the forums and the Internet, that I haven't done even basic testing

[02-Apr-2010 10:49:07] <blahblah2> At what point does ZenOss draw the lines between locations ? It's drawn a line between 2 of my locations (all on the same server) but it hasn't drawn a line to the 3rd location and I am trying to figure out why it's doing that

[02-Apr-2010 10:50:24] <mrayzenoss> the lines between the locations are an overlay of the Network Map

[02-Apr-2010 10:50:50] <blahblah2> so they are automatically defined by the software because I haven't done anything additional

[02-Apr-2010 10:50:51] <mrayzenoss> so if you have devices that are connected on the Network Map on separate locations, they should be connected on the Google Map

[02-Apr-2010 10:51:06] <blahblah2> i see now, thats why the 3rd isnt there because interfaces didnt work when i used WMI

[02-Apr-2010 10:51:12] <blahblah2> windows xp machine too, quite strange

[02-Apr-2010 11:03:47] <blahblah2> is there a simple way to copy classes because I've got a lot of computers on a domain and some not on the domain so the user name and passwords are different, OR is it possible to set zProperties for specific devices ?

[02-Apr-2010 11:05:23] <mrayzenoss> yes and yes

[02-Apr-2010 11:05:35] <mrayzenoss> you can set the zProperties on the individual devices

[02-Apr-2010 11:05:44] <mrayzenoss> and override the device class-provided ones

[02-Apr-2010 11:06:57] <blahblah2> excellent, thanks.

[02-Apr-2010 11:07:22] <mrayzenoss> you can also just add a sub-organizer to the existing class and inherit the settings, overriding them as necessary

[02-Apr-2010 11:14:59] <blahblah2> this software always amazes me

[02-Apr-2010 11:15:53] <mrayzenoss> love a ringing endorsement like that

[02-Apr-2010 11:19:04] <theacolyte> You rang?

[02-Apr-2010 11:32:13] * Simon4 tests out his ZXTM zenpack on some real data

[02-Apr-2010 11:55:08] <Simon4> mrayzenoss: another endorsement for you - it's brilliant that you can write a custom modeller plugin in under a day to model load balancer pools dynamically and collect/graph performance for those pools. I have no idea of any other NMS that would even let you do that let alone do it so rapidly

[02-Apr-2010 12:44:20] <mrayzenoss> Simon4: thanks, we'll be hitting you up for a testimonial at some point :)

[02-Apr-2010 12:48:12] <QbY> I have an interesting problem--and its probably by design, but its hurting--i have a threshold that if its crossed I get an event, and alert is mailed.  However, if subsequent events come in, and the threshold crossing has gotten worse, it only mails the first one--no one realizes the problem has gotten WORSE.

[02-Apr-2010 12:57:34] <Simon4> QbY: change "state = new" in the alerting rule for the person

[02-Apr-2010 12:57:40] <Simon4> to not check the state

[02-Apr-2010 12:57:51] <Simon4> then it should email every time until the event is fixed I believe

[02-Apr-2010 12:58:08] <QbY> well its always emailing--its just emailing the FIRST event

[02-Apr-2010 12:58:15] <Simon4> ah

[02-Apr-2010 12:58:17] <QbY> not the current one that has worse metrcs

[02-Apr-2010 12:58:32] <Simon4> now I have to think ;)

[02-Apr-2010 13:00:32] <QbY> :)

[02-Apr-2010 13:00:54] <QbY> only way to get fresh stastitics is to log in, acknolwedge/history the event then the next one will show current state

[02-Apr-2010 13:03:00] <Simon4> you could make an event transform to update the event summary maybe?

[02-Apr-2010 13:03:20] <Simon4> it'll be the deduplicating that stops the update, but with a transform you may be able to force the update

[02-Apr-2010 13:03:26] <mrayzenoss> or a transform to stop them from de-duping

[02-Apr-2010 13:04:59] <QbY> hrmm...  i kinda like de-duping--but i'd like to get the freshest metrics in the notice

[02-Apr-2010 13:27:55] <Brixius> Is there a way to include other rrd graphs on the perf screen, I'd like to have the Network Throughput graphs on their, I know I can recreate them using either snmp or wmi data sources, but it seems a little silly to do so when the data is allready being gathered.

[02-Apr-2010 13:31:35] <mrayzenoss> Brixius: not that I'm aware of, but that would be a nifty ZenPack, to add an extra tab with every perf graph for that device on 1 screen

[02-Apr-2010 13:33:20] <Brixius> I really wish I was better with python, I'm learning but it's kindof slow going...

[02-Apr-2010 13:34:39] <Brixius> It's more just switching gears of the symantics between languages

[02-Apr-2010 14:55:25] <QbY> Ok, finally able to return to Zenoss...  When  I look at the De-Dup ID information its all about the interface and type of alarm--I don't think mangling that will help..  I just wish there was a way in the transform that i could say. "email this instancE"

[02-Apr-2010 15:00:15] <Simon4> you can mangle the dedupid in the transform, or modify the event summary or message

[02-Apr-2010 15:00:26] <Simon4> my vote is modify the summary/message to include the new data

[02-Apr-2010 15:00:38] <QbY> i looked at the de-dup id, and changng the summary message isn't a par tof it

[02-Apr-2010 15:00:48] <Simon4> it should work and not break de-dup

[02-Apr-2010 15:01:01] <QbY> Simon4.  It does already, but Zenoss is just resending the first event still

[02-Apr-2010 15:01:07] <QbY> i'll post the message

[02-Apr-2010 15:01:11] <QbY> err tranform

[02-Apr-2010 15:01:58] <QbY> http://pastebin.com/HjsRKsLT

[02-Apr-2010 15:03:30] <Simon4> hmm

[02-Apr-2010 15:03:33] <Simon4> yeah, that's what I would ahve done

[02-Apr-2010 15:03:46] <Simon4> this needs someone with more knowledge than I

[02-Apr-2010 15:07:15] <QbY> You ever mess with reportmail?

[02-Apr-2010 15:07:56] <Simon4> nope

[02-Apr-2010 15:08:22] <QbY> trying to fire it off from cron--but don't know which directory it needs to switch into first..

[02-Apr-2010 15:08:32] <QbY> /opt/zenoss/bin/reportmail: line 15: /bin/zenfunctions: No such file or directory

[02-Apr-2010 15:08:32] <QbY> /opt/zenoss/bin/reportmail: line 16: /Products/ZenReports/ReportMail.py: No such file or directory

[02-Apr-2010 15:08:49] <Simon4> running from cron as the zenoss user?

[02-Apr-2010 15:08:55] <QbY> yes

[02-Apr-2010 15:09:05] <Simon4> might be worth sourcing /home/zenoss/.profille or similar

[02-Apr-2010 15:09:20] <Simon4> since $ZENHOME will need to be set amongst other htings at a guess

[02-Apr-2010 15:09:25] <QbY> my cron tab is calling a bash script which is setting the path

[02-Apr-2010 15:09:52] <ckrough> I usually fire it off from within a script with

[02-Apr-2010 15:09:53] <ckrough> #!/bin/sh

[02-Apr-2010 15:09:57] <ckrough> cd $ZENHOME

[02-Apr-2010 15:10:07] <ckrough>  /opt/zenoss/bin/reportmail -v 10 --logpath=....

[02-Apr-2010 15:10:22] <ckrough> as the zenoss user from cron

[02-Apr-2010 15:10:34] <QbY> let me try that one

[02-Apr-2010 15:10:45] <QbY> i've got it setting the PATH and Zenhome now and its still complaining

[02-Apr-2010 15:11:05] <ckrough> it needs to be rum from $ZENHOME

[02-Apr-2010 15:11:22] <ckrough> s/rum/run/

[02-Apr-2010 15:11:34] <QbY> i vote for rum ;)

[02-Apr-2010 15:12:01] <Simon4> +1

[02-Apr-2010 15:15:55] <QbY> i built reports to satisfy my people who need visual represetnation

[02-Apr-2010 15:16:07] <QbY> but i'm under the gun to get the graphs into our provisioning system from zenoss

[02-Apr-2010 15:20:49] <QbY> ckrough: can you share your cron statement / script ??

[02-Apr-2010 15:53:45] <Brixius> This is more of a python question, I'm playing around with zendmd and want to explore an interface, I got a device, and I can loop through the interfaces, I'd like to just get a single instance of an interface out of the collection, but I can't seem to figure out how to refrence it, ie dev=dmd.Devices.findDevice('devname'), then interface=dev.os.interfaces???? I've tried (1), [1], {1}

[02-Apr-2010 15:57:50] <Simon4> interface = dev.os.interfaces()[0]

[02-Apr-2010 15:58:02] <Brixius> thank you.

[02-Apr-2010 15:58:09] <Simon4> or, for i in dev.os.interfaces(): to iterate through em

[02-Apr-2010 15:59:09] <Brixius> ya, I was doing that, but it seems that "i" only has scope in the loop and then it's gone so I can't explore the "i" object after the loop is over.

[02-Apr-2010 15:59:41] <Simon4> hmm, you can, but it points at the last item (or should do)

[02-Apr-2010 15:59:50] <Brixius> hmm, ok

[02-Apr-2010 16:00:01] <Brixius> maybe I was wrong on that one.

[02-Apr-2010 16:00:17] <Simon4> http://pastie.org/900916

[02-Apr-2010 16:01:23] <Brixius> hmm, I could sware I tried that

[02-Apr-2010 16:12:37] <Brixius> what is dev.os.interface.snmpindex used for? it's 0 on all my interfaces, I was thinkng this would be the snmpindex of the interface.

[02-Apr-2010 16:13:28] <Simon4> yeah, generally it is... basically it's the interface specific part of the OID

[02-Apr-2010 16:14:02] <Simon4> it's zero for all my interfaces also, not quite sure why

[02-Apr-2010 16:15:11] <Brixius> I'm wondering why it's 0 then on my interfaces of the device, which I'm trying to create a transform that will lookup the ifname, my device gives me linkdown and in the varbinds it sends the snmpindex of the link in question

[02-Apr-2010 16:22:39] <Brixius> so can someone tell me if ifindex is synomous to the "SNMP Index" shown on the devices interface properties screen.

[02-Apr-2010 16:25:15] <Simon4> seems that just interfaces use ifindex

[02-Apr-2010 16:25:21] <Simon4> http://pastie.org/900961

[02-Apr-2010 16:25:26] <Simon4> filesystems use snmpindex

[02-Apr-2010 16:25:52] <Brixius> seems to be an inconsistancy then...

[02-Apr-2010 16:25:57] <Simon4> yeah

[02-Apr-2010 16:26:05] <Simon4> the zenpacks I've written use snmpindex

[02-Apr-2010 16:26:20] <Simon4> (for tother things, not interfaces etc)

[02-Apr-2010 16:29:33] <gwb235> what is the best way to resolve "not in topology" warnings ?

[02-Apr-2010 18:37:07] <QbY> Anyone monitoring Asterisk?

[03-Apr-2010 00:00:30] [disconnected at Sat Apr  3 00:00:30 2010]

[03-Apr-2010 00:00:30] [connected at Sat Apr  3 00:00:30 2010]

[03-Apr-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[03-Apr-2010 05:09:10] <PhunkMaster> Anyone in here happen to be awake? I am trying to test the Zenoss core for my company. I am not quite sure how I broke it, but for some reason I am unable to use the Easy Add Device anymore. When I view the error details, it appears to be about a page long. The short version of it is Type: ValueError Value: too many values to unpack

[03-Apr-2010 12:00:48] <blahblah2> Afternoon  all

[03-Apr-2010 16:27:58] <seen1976> Have a question if someone would be so kind as to help me on a Saturday...........

[03-Apr-2010 16:30:12] <seen1976> We are trying to export the graphs for a customer from Zenoss into our own internal portal that we have.  Can somebody tell me how to go about this.  I really need to get this finished before Monday and would be grateful for any and all help

[04-Apr-2010 00:00:30] [disconnected at Sun Apr  4 00:00:30 2010]

[04-Apr-2010 00:00:31] [connected at Sun Apr  4 00:00:31 2010]

[04-Apr-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[04-Apr-2010 14:53:26] Simon4_ is now known as Simon4

[04-Apr-2010 22:40:06] MattD is now known as mdereus

[04-Apr-2010 22:51:54] MattD is now known as mdereus

[04-Apr-2010 23:19:02] MattD is now known as mdereus

[04-Apr-2010 23:30:58] MattD is now known as mdereus

[05-Apr-2010 00:00:30] [disconnected at Mon Apr  5 00:00:30 2010]

[05-Apr-2010 00:00:31] [connected at Mon Apr  5 00:00:31 2010]

[05-Apr-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[05-Apr-2010 05:02:41] fOrsberg is now known as forsberg

[05-Apr-2010 05:23:10] forsberg is now known as fOrsberg

[05-Apr-2010 10:14:03] <RoAkSoAx> mrayzenoss, heya!! Installed zenoss pretty much following your installation notes: http://paste.ubuntu.com/409536/

[05-Apr-2010 10:14:35] <RoAkSoAx> howerver, now it shows me the interface of zenoss interface, few minutes ago it was =showing zenoss core, but couldn't connect to dashbord because it was showing some kind of weird message

[05-Apr-2010 10:58:45] <mistich> good morning

[05-Apr-2010 10:58:52] <Simon4> morning

[05-Apr-2010 10:59:45] <Simon4> mrayzenoss: if writing ZenPacks to share with the community etc, should they be named ZenPacks.community.XXXX, or ZenPacks.authorname.XXXXX, or is there a reasoning behind doing either or a best practice to follow?

[05-Apr-2010 11:00:12] <mistich> I'm trying to create a CommandPlugin that will use ssh to do a show route on a firewall. it connects to the firewall but trys to run it in a non privileged mode.

[05-Apr-2010 11:00:31] <mistich> how do I tell CommandPlugin to run the command in enable mode

[05-Apr-2010 11:01:13] <mrayzenoss> Simon4: it's left to the discretion of the author

[05-Apr-2010 11:01:33] <mrayzenoss> if you're working with someone else, perhaps 'community'

[05-Apr-2010 11:02:02] <Simon4> *nod*

[05-Apr-2010 11:02:05] <mrayzenoss> 'zenoss' is the only reserved one, since that implies it's provided by zenoss

[05-Apr-2010 11:05:20] <mistich> nevermind

[05-Apr-2010 11:14:02] <mistich> ok I found where I set the enable in zproperties but still doesn't seem to work

[05-Apr-2010 11:15:16] <mistich> here is the log nowhere says it entered enable mode

[05-Apr-2010 11:15:24] <mistich> 2010-04-05 11:10:31,182 DEBUG zen.SshClient: x.x.x.x host fingerprint:

[05-Apr-2010 11:15:24] <mistich> 2010-04-05 11:10:31,228 DEBUG zen.SshClient: Creating new SSH connection...

[05-Apr-2010 11:15:24] <mistich> 2010-04-05 11:10:31,229 DEBUG zen.SshClient: Expanded SSH key path from zKeyPath ~/.ssh/id_dsa to /home/zenoss/.ssh/id_dsa

[05-Apr-2010 11:15:24] <mistich> 2010-04-05 11:10:31,229 DEBUG zen.SshClient: SSH key path /home/zenoss/.ssh/id_dsa doesn't exist

[05-Apr-2010 11:15:24] <mistich> 2010-04-05 11:10:31,250 DEBUG zen.SshClient: Sorted list of authentication methods: ['password']

[05-Apr-2010 11:15:25] <mistich> 2010-04-05 11:10:31,250 DEBUG zen.SshClient: Attempting method password

[05-Apr-2010 11:15:25] <mistich> 2010-04-05 11:10:31,270 INFO zen.SshClient: Connected to device x.x.x.x

[05-Apr-2010 11:15:26] <mistich> 2010-04-05 11:10:31,270 DEBUG zen.SshClient: Started the channel for command: show route

[05-Apr-2010 11:15:26] <mistich> 2010-04-05 11:10:31,281 DEBUG zen.SshClient: Opening command channel for show route

[05-Apr-2010 11:15:27] <mistich> 2010-04-05 11:10:31,460 DEBUG zen.ZenModeler: Running 1 clients

[05-Apr-2010 11:18:19] <mistich> Bueller

[05-Apr-2010 11:24:35] <mistich> matt

[05-Apr-2010 11:25:07] <mistich> is anyone around?

[05-Apr-2010 11:26:59] <Simon4> that looks like the enable pw is only for telnet stuff, vs ssh

[05-Apr-2010 11:32:18] <mistich> yeah went thru trac and found it the ticket http://dev.zenoss.org/trac/ticket/5711

[05-Apr-2010 11:32:34] <mistich> thought it was fixed but that was another ticket

[05-Apr-2010 11:33:07] <Simon4> a ha

[05-Apr-2010 11:34:51] <mistich> put both of them in at the same time

[05-Apr-2010 12:11:00] <RoAkSoAx> mrayzenoss, should zenoss trunk be displaying the interface of zenoss enterprise?

[05-Apr-2010 12:11:02] * Simon4 has a quick look at the source

[05-Apr-2010 12:11:17] <Simon4> mistich: damm, not a trivial hack to enable :)

[05-Apr-2010 12:13:33] <mrayzenoss> RoAkSoAx: there's a ticket opened for that

[05-Apr-2010 12:13:42] <mrayzenoss> it's just mid-reskinning

[05-Apr-2010 12:14:18] <RoAkSoAx> mrayzenoss, ok then. Other than that the trunk seems to be installing correctly. I'll blog about it soon

[05-Apr-2010 12:14:39] <mrayzenoss> excellent

[05-Apr-2010 12:27:41] <mistich> Simon4: I agree but firewalls do not do telnet so ssh only

[05-Apr-2010 12:28:02] <Simon4> mistich: yeah agreed... I also would like to see that ticket closed

[05-Apr-2010 12:28:18] <Simon4> I was just having a quick look to see whether it was easy to hack in or not ;)

[05-Apr-2010 12:29:19] <mistich> no I looked and will have to put some time into it

[05-Apr-2010 12:30:17] <mistich> I still have 8 more zenpacks to move to 2.5.2

[05-Apr-2010 12:30:24] <mistich> so far 6 done

[05-Apr-2010 12:32:07] * Simon4 needs to upgrade his test VM, maybe tonight

[05-Apr-2010 12:45:03] <RoAkSoAx> mrayzenoss, done: http://www.roaksoax.com/2010/04/installing-zenoss-in-ubuntu-10-04

[05-Apr-2010 12:49:08] <trashguy> hello

[05-Apr-2010 12:50:25] <trashguy> anyoen ever use zenoss with force10 gear?

[05-Apr-2010 12:57:26] <mrayzenoss> If it's standard SNMP stuff, shouldn't be a problem

[05-Apr-2010 12:58:16] <trashguy> i have all the mibs too, i am new to zenoss and we only have F10 gear so i  was just curious

[05-Apr-2010 12:58:34] <mrayzenoss> so a quick clarification, the MIBs are only used for translating traps

[05-Apr-2010 13:00:17] <mrayzenoss> assuming it's a standard router/switch, the /Network/Router or Switch it should find the interfaces

[05-Apr-2010 13:00:26] <mrayzenoss> as long as things are fairly standard

[05-Apr-2010 13:00:26] <trashguy> yea

[05-Apr-2010 13:00:34] <trashguy> they are cisco ish

[05-Apr-2010 13:00:57] <mrayzenoss> well, if not everything works, there are lots of examples of custom modelers

[05-Apr-2010 13:01:01] <trashguy> http://www.force10networks.com/

[05-Apr-2010 13:01:14] <mrayzenoss> yeah

[05-Apr-2010 13:01:30] <trashguy> I am reading the docs, irc is useful when someones liek yea ive done it, go read this :)

[05-Apr-2010 13:02:06] <citrus2> i have cacti and zenoss running on the same server.  can i make zenoss run on port 80  and create and alias in apache like cacti?

[05-Apr-2010 13:02:31] <citrus2> i know i could just create a redirect.. but that would be to easy

[05-Apr-2010 13:02:39] <mrayzenoss> citrus2: you can change the port in the zope.conf, but since it runs as the zenoss user you won't have port 80

[05-Apr-2010 13:03:00] <citrus2> so it has its own built in http service i take ti

[05-Apr-2010 13:03:01] <citrus2> it

[05-Apr-2010 13:03:07] <mrayzenoss> yeah, port 8080

[05-Apr-2010 13:03:09] <mrayzenoss> Zope

[05-Apr-2010 13:03:30] <citrus2> oh well,  i will just create a redirect then

[05-Apr-2010 14:54:29] <nickanderson> good day

[05-Apr-2010 14:55:38] <nickanderson> how much information is exposed in command templates?

[05-Apr-2010 14:56:27] <Brixius> Hello, I have the following trap http://pastebin.org/137669 and I'm trying to transform it. It seems like it should be simple, Here's ther transform I'm using "

[05-Apr-2010 14:56:27] <Brixius> evt.summary="CAC failure: nexToneRegId=" % (evt.nexToneRegId)

[05-Apr-2010 14:57:35] <Brixius> It's mapping to the correct event class, but for some reason the Mapping doesn't appear to be running on it.

[05-Apr-2010 14:58:36] <nickanderson> I need to monitor several websites across several hosts, my idea was to use the device classes, so something like /devices/websites/hostname/url I was thinking

[05-Apr-2010 14:58:43] <nickanderson> never mind i see flaw in my logic

[05-Apr-2010 14:59:33] <Brixius> let me reword that, It's showing the event under the mapping but the transform doesn't appear to be applying.  it still has the default snmp message in summary.

[05-Apr-2010 15:04:13] <Brixius> I guess I should've looked at the statement closer, I'm thinking " evt.summary="CAC failure: nexToneRegId= %s" % (evt.nexToneRegId) will work better

[05-Apr-2010 15:04:28] <nickanderson> yes

[05-Apr-2010 15:04:36] <nickanderson> sorry didnt see that before

[05-Apr-2010 15:04:46] <nickanderson> but that was incorrect substitution

[05-Apr-2010 15:04:56] <nickanderson> string replacement whatever you wanna call it

[05-Apr-2010 16:03:18] <Brixius> I have a command datasource on an old zenoss box, that is using snmpget to get an oid, then it filters it through sed because it the device returns a retarded value like "..... 52" is there any way to do this same sort of thing just using an snmp datasource?

[05-Apr-2010 16:17:54] <mrayzenoss> Brixius: I don't think so.

[05-Apr-2010 16:18:07] <mrayzenoss> some devices suck that way

[05-Apr-2010 16:21:11] <Brixius> ok, I guess I'll just stick to the command datasource then.  Was hoping to be able to just use an snmp datasource but oh well.

[05-Apr-2010 16:22:16] <citrus2> where do you set how often a system gets pinged?

[05-Apr-2010 16:22:55] <willwh> citrus2: just 1 specific device?

[05-Apr-2010 16:23:03] <willwh> or for a template for many devices?

[05-Apr-2010 16:23:11] <citrus2> for a template

[05-Apr-2010 16:23:18] <mrayzenoss> Ping Cycle Time (secs) is defined on the collector

[05-Apr-2010 16:23:36] <willwh> there's your answer :)

[05-Apr-2010 16:26:15] <citrus2> sorry i still learning

[05-Apr-2010 16:26:17] <citrus2> where is that

[05-Apr-2010 16:26:52] <mrayzenoss> on the left nav, Collectors

[05-Apr-2010 16:26:57] <mrayzenoss> then 'localhost'

[05-Apr-2010 16:27:02] <mrayzenoss> you can edit it if necessary

[05-Apr-2010 16:27:46] <citrus2> got it thank you

[05-Apr-2010 16:27:48] <citrus2> :)

[05-Apr-2010 16:31:27] <frozty_sa> evening

[05-Apr-2010 16:32:53] <frozty_sa> so before I head into this and possibly reinvent the wheel, I'd just like to ask once more quickly: anyone know if there's a plugin or alert method I could use for determining if something is different to a normal pattern?

[05-Apr-2010 16:33:05] <citrus2> where is the only one entry "local host" in the collectors?

[05-Apr-2010 16:33:25] <citrus2> its that kinda the defacto unless i create on specifically for a machine?

[05-Apr-2010 16:33:33] <frozty_sa> like, for instance, being able to create an alert showing that traffic on a specific interface has dropped when compared to its normal levels at a similar time

[05-Apr-2010 16:37:38] <Simon4> frozty_sa: check out holt-winter or "predictive" thresholds

[05-Apr-2010 16:38:11] <Simon4> citrus2: initial install has the performance collector on the same machine, i.e. "localhost"

[05-Apr-2010 16:38:28] * Simon4 goes back to tv

[05-Apr-2010 16:38:44] <citrus2> oh  i see  the "collector" could be on a different machine then the rest of the zenoss componants?

[05-Apr-2010 16:38:52] <Simon4> yup

[05-Apr-2010 16:38:58] <Simon4> and you can have more than one

[05-Apr-2010 16:39:09] <citrus2> what benifit would that give you?

[05-Apr-2010 16:39:30] <mrayzenoss> scalability

[05-Apr-2010 16:39:43] <Simon4> with a bunch of devices it lets you spread load, or it also lets you put a collector somewehre that's firewalled off from the main install

[05-Apr-2010 16:41:04] <citrus2> interesting,  so you could install the collector say at a different datacenter and have it "report back" /

[05-Apr-2010 16:41:05] <citrus2> ?

[05-Apr-2010 16:42:56] <mrayzenoss> yes

[05-Apr-2010 16:42:58] <Simon4> yup

[05-Apr-2010 16:43:06] <Simon4> mrayzenoss: you can take this one :)

[05-Apr-2010 16:43:22] <mrayzenoss> nah, I need to go work :)

[05-Apr-2010 16:43:32] <frozty_sa> Simon4: I'll try it out and see how effectively it works, thanks

[05-Apr-2010 16:43:36] <frozty_sa> just doing the reading now

[05-Apr-2010 16:43:46] <Simon4> frozty_sa: I haven't tried it, but seems like the way to do it

[05-Apr-2010 16:43:58] <frozty_sa> Simon4: yeah, we'll see

[05-Apr-2010 16:44:14] <frozty_sa> I just need to get as many sort of alerts as possible in place

[05-Apr-2010 16:44:37] <frozty_sa> and then continue dreaming about quitting my job and becoming some sort of monk

[05-Apr-2010 16:45:59] <Simon4> heh

[05-Apr-2010 16:46:07] <Simon4> life would be much simpler

[05-Apr-2010 16:46:09] <Simon4> for sure

[05-Apr-2010 16:47:34] <citrus2> why would my 5 webservers report HTTP is DOWN on Zenoss  but  port 80 is working fine..

[05-Apr-2010 16:49:35] <frozty_sa> navigate to the events for the webservers

[05-Apr-2010 16:49:41] <frozty_sa> and see what exactly they say

[05-Apr-2010 16:50:05] <frozty_sa> it'll lead you to the first point of seeing what's wrong with either the servers or the eventing

[05-Apr-2010 16:50:45] <citrus2> http://pastebin.org/137884

[05-Apr-2010 16:53:19] <citrus2> when i telnet into port 80 to the server from the zenoss box it connects just fine too

[05-Apr-2010 16:53:25] <citrus2> i don't see why this would fail

[05-Apr-2010 16:54:36] <Brixius> have you tried doing a wget from your device from the zenoss box and see what happens?

[05-Apr-2010 16:54:49] <Brixius> er to your device from zenoss

[05-Apr-2010 16:59:29] <Brixius> citrus2: one other thing you might check, does your server have multiple ip addresses? Is your web site on an address other then the one on the top of the device status screen?

[05-Apr-2010 17:01:27] <citrus2> Yes actually they are

[05-Apr-2010 17:03:26] <Brixius> I'm guessing zenoss is checking the wrong ip address for http then.  Might be easist to create a webserver device that monitors the web site seperate from the server device.

[05-Apr-2010 17:03:48] <Brixius> or there may be a way to handle this I'm not aware of.

[05-Apr-2010 17:04:12] <citrus2> how do i suppress these errors for now?

[05-Apr-2010 17:04:22] <citrus2> i don't want to just acknoledge them

[05-Apr-2010 17:04:28] <citrus2> i want them to just not report for now

[05-Apr-2010 17:04:38] <citrus2> sorry i am still in the middle of the zenoss admin guide

[05-Apr-2010 17:04:51] <Brixius> Disable the monitoring of tcp service http on the os tab of the device.

[05-Apr-2010 17:04:55] <citrus2> my add keeps me from reading it though :-)

[05-Apr-2010 17:05:19] <Brixius> it's probably listed under IP services on the os tab.

[05-Apr-2010 17:06:11] <Brixius> actually, looks like there's a way to set the check ip address on the ip service

[05-Apr-2010 17:08:38] <Brixius> LOL.. I've been trying to figure it out for a while now, most of my issues are due to the fact that I don't really know python.  something seems to have clicked today and it's making alot more sense.

[05-Apr-2010 17:17:46] <citrus2> how do i drill down to that?  everytime i click on events and drill down to a certain class i see no information other than devices that have spawn that event..

[05-Apr-2010 21:03:16] <salax> hi all, can i change/edit the zenoss banner at the top for the oss version? and where can i edit it?

[05-Apr-2010 21:22:41] <aclark> salax: you might try in the ZMI in portal_skins somewhere e.g. /portal_skins/zenui/img/zenoss_black_logo.png/manage_main

[05-Apr-2010 22:27:56] <salax> aclark: thanks.. i will look at it

[06-Apr-2010 00:00:30] [disconnected at Tue Apr  6 00:00:30 2010]

[06-Apr-2010 00:00:30] [connected at Tue Apr  6 00:00:30 2010]

[06-Apr-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[06-Apr-2010 00:42:56] <salax> i've did some research and it is at /usr/local/zenoss/zenoss/Products/ZenWidgets/skins/zenui  .. filename logo.png

[06-Apr-2010 01:03:48] <Troubadix09> morning all

[06-Apr-2010 01:46:48] fOrsberg is now known as forsberg

[06-Apr-2010 04:29:24] <penthium> hello

[06-Apr-2010 04:35:37] <fus10nx> morning :)

[06-Apr-2010 04:36:15] <penthium> I have a issue with zenoss 2.5.2, I can create Event, view new event generate by zenoss system, but I can't delete event. ( sorry for my english I a frogy french ;) )

[06-Apr-2010 04:36:42] <fus10nx> I'm new to ZenOSS so I won't be much help but I don't think you can ever fully delete the events

[06-Apr-2010 04:36:48] <fus10nx> I think they just move to some historical event area

[06-Apr-2010 04:40:21] <penthium> also in the history event area I can't delete old event.

[06-Apr-2010 04:40:40] <fus10nx> I'm not 100% sure, sorry. A lot of people come in here around 5PM London time

[06-Apr-2010 04:40:51] <penthium> this issue come just after I change the slq

[06-Apr-2010 04:41:00] <penthium> sql database server

[06-Apr-2010 04:41:38] <penthium> because I don't want to use the sql generate by zenoss.

[06-Apr-2010 04:43:40] <penthium> ok I'll try to come back a this time, thx.

[06-Apr-2010 04:44:01] <fus10nx> ok good luck

[06-Apr-2010 07:38:54] <ckrough> rmatte: you on?

[06-Apr-2010 08:33:23] <forsberg> not untill later

[06-Apr-2010 08:33:26] <forsberg> :P

[06-Apr-2010 08:33:32] <forsberg> unless hes up very early

[06-Apr-2010 08:33:59] <ckrough> slacker.

[06-Apr-2010 08:36:16] <forsberg> hehe well! in all fairness he is in another timezone than me anyways

[06-Apr-2010 08:49:50] <frederic_> hello, I've a issue since i've moved sql database to another server, I can't delete any events from the Gui. I can conecte to the sql database, view create and delete entry, but with the interface I've "Connection to server Eror" each time I try to delete ( close ) event. ( sory for my english, I'm french )

[06-Apr-2010 09:00:19] <Simon4> does your zenoss user on the new mysql server have appropriate permissions?

[06-Apr-2010 09:00:55] <Simon4> if you can see but not delete events it sounds like it may not have

[06-Apr-2010 09:04:59] <frederic_> Yes ! that's it ! thank alot ! the zenoss user must have the super privilege.

[06-Apr-2010 09:07:36] <frederic_> have a nice day, bye

[06-Apr-2010 09:31:47] <rmatte> mrayzenoss: blogs/zenossblog/2010/01/25/tip-of-the-month-monitoring-the-status-of-nfs-shares-with-zenoss

[06-Apr-2010 09:31:53] <rmatte> it got hit by spam bots

[06-Apr-2010 09:31:54] <rmatte> ;)

[06-Apr-2010 09:32:11] <rmatte> viagra and such

[06-Apr-2010 09:32:14] <mrayzenoss> hmm… interesting.  They didn't show up in my filters

[06-Apr-2010 09:32:22] * mrayzenoss is off to delete spam and update filters

[06-Apr-2010 09:32:35] <rmatte> hehe

[06-Apr-2010 09:35:08] <mrayzenoss> wait, those are trackbacks

[06-Apr-2010 09:35:12] <mrayzenoss> I'd disabled those

[06-Apr-2010 09:37:58] <mrayzenoss> ugh, someone turned them back on, Spamfest

[06-Apr-2010 09:40:34] <rmatte> lol

[06-Apr-2010 09:40:47] <rmatte> start crackin' skulls

[06-Apr-2010 09:43:25] <mrayzenoss> sorry everyone, you won't be able to go to the Zenoss blog to find where to get viagra

[06-Apr-2010 09:44:50] <ckrough> thats ok, as long as I can learn how to rip dvds to my iphone there

[06-Apr-2010 09:45:51] <mrayzenoss> I think there haven't been any of those posts since I added the keyword interceptors

[06-Apr-2010 09:46:23] <rmatte> lol

[06-Apr-2010 09:46:36] <rmatte> yeh, the Zenoss community site can be strangely educational

[06-Apr-2010 09:46:44] <rmatte> even more non-Zenoss related subject matter

[06-Apr-2010 09:46:47] <mrayzenoss> completely unintentional :p

[06-Apr-2010 09:46:51] <rmatte> who the heck needs google?

[06-Apr-2010 09:47:07] <rmatte> :)

[06-Apr-2010 09:47:09] <mrayzenoss> http://search.zenoss.com

[06-Apr-2010 09:47:33] <rmatte> haha, nice

[06-Apr-2010 10:01:45] <rmatte> hmmmm, 6 more days before I know whether or not the memory usage on 2.5.2 is better than on 2.4.5... so far it's looking like it is

[06-Apr-2010 10:07:02] <rmatte> venturaville: hey

[06-Apr-2010 10:08:07] <venturaville> hwody

[06-Apr-2010 10:08:10] <venturaville> howdy

[06-Apr-2010 10:10:12] <venturaville> mrayzenoss: I retested version 1.19 in subversion for netapp and it seems to work fine, including adding the Storage suborganizer.......

[06-Apr-2010 10:15:14] <rmatte> venturaville: has the HP Blade Chassis pack officially been released?

[06-Apr-2010 10:16:19] <rmatte> (the latest version I mean, the one you showed us screenshots of the other day)

[06-Apr-2010 10:17:06] * Simon4 wonders if they were the screenshots of the version he was working on

[06-Apr-2010 10:18:00] <venturaville> rmatte: what Simon4 said......

[06-Apr-2010 10:18:10] <venturaville> Simon4 is making some changes to  it

[06-Apr-2010 10:18:23] * Simon4 will commit em tonight  - has been a mental week

[06-Apr-2010 10:18:26] <venturaville> there are 3 ways to talk to them: ssh, SOAP API, snmp

[06-Apr-2010 10:18:31] <venturaville> and they don't provide all the same info ......

[06-Apr-2010 10:19:14] <venturaville> +1 to the mental week

[06-Apr-2010 10:19:23] <venturaville> last week's on-call for me was h***

[06-Apr-2010 10:19:35] <rmatte> oh right, it was you Simon4, sorry, slipped my mind

[06-Apr-2010 10:20:08] <rmatte> venturaville: crappy

[06-Apr-2010 10:21:15] <rmatte> venturaville: no chance to look at esx yet I assume?

[06-Apr-2010 10:23:45] <venturaville> no I barely got time to get any sleep

[06-Apr-2010 10:24:18] <venturaville> I was up all night one night dealing with a stack of arrays that lost their mind.....

[06-Apr-2010 10:24:31] <rmatte> ah

[06-Apr-2010 10:26:50] <rmatte> I'm only asking because I keep getting bugged about it, but I'm like "What do you want me to do?" lol

[06-Apr-2010 10:27:12] <rmatte> I still can't believe VMWare just suddenly decided to drop SMTP...

[06-Apr-2010 10:27:14] <rmatte> erm

[06-Apr-2010 10:27:17] <rmatte> SNMP

[06-Apr-2010 10:27:19] <rmatte> rather

[06-Apr-2010 10:27:32] <rmatte> (was working on setting up a mailserver yesterday, burned in to my brain)

[06-Apr-2010 10:28:02] <rmatte> and instead of at least maintaining level of SNMP support that they already had, they purposely crippled it

[06-Apr-2010 10:29:00] <venturaville> enterprise has a nice zenpack for it

[06-Apr-2010 10:29:09] <venturaville> *venturaville runs for cover*

[06-Apr-2010 10:29:18] <rmatte> yeh, a really nice one from what I hear...

[06-Apr-2010 10:29:24] <rmatte> and that's literally what I keep telling them

[06-Apr-2010 10:29:31] <rmatte> "If you'd buy enterprise, we wouldn't have this issue"

[06-Apr-2010 10:29:39] <venturaville> we are using it here

[06-Apr-2010 10:29:40] <venturaville> quite nice

[06-Apr-2010 10:29:47] <rmatte> I'd imagine that we will eventually once we grow large enough

[06-Apr-2010 10:29:48] <venturaville> pretty much 1 for 1 with vfoglight

[06-Apr-2010 10:30:01] <rmatte> cool

[06-Apr-2010 10:31:52] <rmatte> the main issue is that we're an MSP, so we'd need to get an MSP license of enterprise, and right now the price point would most likely be too high for us

[06-Apr-2010 10:32:01] <rmatte> but once we grow then we'll be able to look at doing it

[06-Apr-2010 11:14:56] <lthrasher> how is MSP licensing structured anyway? i don't see it mentioned on the zenoss site.

[06-Apr-2010 11:19:30] <ainc> how do i go about monitoirng disk space not by percentage, but say, not to let it go below 750MB?

[06-Apr-2010 11:28:14] <rmatte> lthrasher: you need to talk to the sales team directly for it

[06-Apr-2010 11:28:21] <rmatte> lthrasher: they don't advertise it on the site

[06-Apr-2010 11:29:24] <rmatte> ainc: checking...

[06-Apr-2010 11:31:16] <rmatte> ainc: hmmm, well, it's monitored in blocks...

[06-Apr-2010 11:33:52] <rmatte> working on finding the right property to use in the threshold...

[06-Apr-2010 11:37:24] <cluther> ainc: Set the max threshold on usedBlocks to "here.totalBlocks - ((750 * 1024 * 1024) / here.blockSize)"

[06-Apr-2010 11:38:15] <rmatte> Chet to the rescue

[06-Apr-2010 11:39:10] <rmatte> cluther: what is the equivalent of d = dmd.Devices.findDevice() for an interface or a disk?

[06-Apr-2010 11:39:28] <rmatte> like, is there a way to do the same thing without having to for loop through them all?

[06-Apr-2010 11:39:51] <rmatte> been wondering that for a while, no one has been able to give me an answer

[06-Apr-2010 11:40:18] <cluther> What do you know about the interface? It's name and the device it's on?

[06-Apr-2010 11:40:26] <rmatte> correct

[06-Apr-2010 11:40:53] <rmatte> I guess you'd do a for loop and set another variable based on that?

[06-Apr-2010 11:41:03] <rmatte> f = filesystem

[06-Apr-2010 11:41:04] <rmatte> or whatever

[06-Apr-2010 11:41:41] <cluther> Ideally you'd want to use the componentSearch catalog to make it fast. Unfortunately we don't have the component's name in there. The following should be quite fast though..

[06-Apr-2010 11:42:05] <rmatte> anything is faster than a for loop

[06-Apr-2010 11:46:32] <cluther> http://gist.github.com/357803

[06-Apr-2010 11:47:24] <rmatte> sweet

[06-Apr-2010 11:47:25] <rmatte> thanks

[06-Apr-2010 11:51:40] <ainc> ooops, sorry im back...trying it now

[06-Apr-2010 11:51:48] ainc is now known as cain22

[06-Apr-2010 11:57:29] <cain22> its currently above that threshold....can i change the 750 to lets say 3900 to make sure this test works?

[06-Apr-2010 11:59:23] <rmatte> cain22: yes, you can

[06-Apr-2010 11:59:36] <rmatte> it's a threshold, the value can be changed to whatever you like

[06-Apr-2010 11:59:49] <rmatte> as long as the formula remains in tact

[06-Apr-2010 12:00:01] <cain22> hmmm, gues its not creating an event...

[06-Apr-2010 12:00:16] <rmatte> it'll take 5 minutes

[06-Apr-2010 12:00:24] <rmatte> it gets created when Zenoss polls

[06-Apr-2010 12:00:33] <rmatte> just be patient and wait for it ;)

[06-Apr-2010 12:00:44] <cain22> and or model it

[06-Apr-2010 12:00:45] <cain22> :P

[06-Apr-2010 12:01:31] <rmatte> well, modelling isn't the same as zenperfsnmp, so I wasn't sure if that would kick it off or not

[06-Apr-2010 12:02:34] <cain22> well either it isn't the same, or it didn't work...i'll wait the five to make sure

[06-Apr-2010 12:02:47] <rmatte> yeh, it won't work, that's what I was saying

[06-Apr-2010 12:02:57] <rmatte> you can execute zenperfsnmp run from the commandline as the zenoss user

[06-Apr-2010 12:03:00] <rmatte> that would probably do it

[06-Apr-2010 12:03:22] <cain22> eh, i'll wait

[06-Apr-2010 12:03:27] <rmatte> up to you

[06-Apr-2010 12:04:07] <cain22> what does the 1024 stand for?

[06-Apr-2010 12:04:45] <rmatte> there are 1024 bytes in a megabyte

[06-Apr-2010 12:04:54] <rmatte> erm

[06-Apr-2010 12:04:56] <rmatte> in a kilobyte

[06-Apr-2010 12:04:57] <rmatte> rather

[06-Apr-2010 12:05:04] <rmatte> 1024 kilobytes in a megabyte

[06-Apr-2010 12:05:05] <rmatte> etc...

[06-Apr-2010 12:05:33] <rmatte> so 750 * 1024 * 1024 represents 750MB

[06-Apr-2010 12:05:43] <rmatte> in bytes

[06-Apr-2010 12:06:45] <rmatte> although looking at it now I'm not sure it'll behave the exact way that you expect

[06-Apr-2010 12:07:02] <rmatte> since he has it subtracting 750MB from the total

[06-Apr-2010 12:07:23] <cain22> hm...thinking i should just do the math and find out the percentage

[06-Apr-2010 12:07:52] <rmatte> oh nevermind, it's set on usedBlocks

[06-Apr-2010 12:07:54] <cain22> but this could be a helpful way of keeping tabs on the disk

[06-Apr-2010 12:07:55] <rmatte> yeh, it'll work

[06-Apr-2010 12:08:08] <rmatte> why would you bother to calculate the percentage?

[06-Apr-2010 12:08:22] <cain22> in case this doesn't work...  :P

[06-Apr-2010 12:08:28] <rmatte> it will work

[06-Apr-2010 12:08:34] <rmatte> what did you change 750MB to?

[06-Apr-2010 12:08:37] <rmatte> erm 750

[06-Apr-2010 12:08:44] <cain22> 3900

[06-Apr-2010 12:08:55] <rmatte> ok, and what's the current usage on the disk?

[06-Apr-2010 12:09:01] <cain22> 3.8GB

[06-Apr-2010 12:09:17] <rmatte> what's the total capacity?

[06-Apr-2010 12:10:16] <cain22> 18.6

[06-Apr-2010 12:10:20] <cain22> GB

[06-Apr-2010 12:11:13] <rmatte> ok, now... what he has given you is a threshold for how much free space there is, not how much used space, so you're confusing it

[06-Apr-2010 12:11:29] <rmatte> you're telling it to threshold when there's only 3.9GB free

[06-Apr-2010 12:11:44] <rmatte> but you have like 14GB free

[06-Apr-2010 12:12:19] <rmatte> comprende?

[06-Apr-2010 12:12:44] <cain22> negative...there is 3.8GB free space currently

[06-Apr-2010 12:12:57] <rmatte> oh, I asked for current usage

[06-Apr-2010 12:13:02] <rmatte> and you gave me free space :P

[06-Apr-2010 12:13:10] <rmatte> ok

[06-Apr-2010 12:13:16] <cain22> oopos

[06-Apr-2010 12:13:21] <QbY> does anyone know why my thruput graphs look this way? http://imagebin.ca/view/5npyRmb.html

[06-Apr-2010 12:13:21] <cain22> I think I'm dislexic

[06-Apr-2010 12:13:23] <rmatte> set it to something higher, like 4500 or something

[06-Apr-2010 12:13:26] <cain22> and a horrible speller

[06-Apr-2010 12:13:59] <rmatte> QbY: because the Zenoss box has too big of a workload to handle so it's blowing over polling cycle and dropping data

[06-Apr-2010 12:14:17] <QbY> what?!

[06-Apr-2010 12:14:22] <cluther> QbY: Actually it's because you're collecting a fast interface using 32bit counters.

[06-Apr-2010 12:14:29] <cluther> They're rolling more than once between collection cycles.

[06-Apr-2010 12:14:56] <QbY> ok, where would that be changed?

[06-Apr-2010 12:15:00] <cluther> Verify that the zSnmpVer is set to v2c and remodel the device. You should see the interface's type go from something to something_64

[06-Apr-2010 12:15:04] <mistich> thats only if he is not getting heart beat errors

[06-Apr-2010 12:15:12] <rmatte> QbY: what's the type listed for the interface when you click on it?

[06-Apr-2010 12:16:30] <rmatte> cluther: didn't think about the 64bit since I'm used to the 64bit template being automatically applied

[06-Apr-2010 12:17:35] <QbY> Type: ppp -- Its a Packet over Sonet Interface

[06-Apr-2010 12:18:13] <QbY> the packets graph on the same interface looks great

[06-Apr-2010 12:18:48] <rmatte> cluther: isn't that generally a workload issue if the 1 graph looks fine but the other doesn't?

[06-Apr-2010 12:19:04] <rmatte> I thought 64bit applied to both graphs

[06-Apr-2010 12:19:23] <cluther> QbY: No.. it's definitely a counter that's rolling too quickly. Notice how the gaps are worse when the traffic is higher. The reason the packet graph looks fine is that there are far fewer packets than bytes.

[06-Apr-2010 12:19:46] <QbY> ok..  so...........?

[06-Apr-2010 12:19:47] <cluther> What Zenoss version are you on? There is a fix in 2.5.2 that might resolve this for you.

[06-Apr-2010 12:19:52] <rmatte> QbY: k, do what Cluther said, make sure the snmp version is set to v2c in zproperties, then remodel

[06-Apr-2010 12:19:53] <cluther> http://dev.zenoss.org/trac/ticket/6067

[06-Apr-2010 12:19:53] <QbY> 2.5.2

[06-Apr-2010 12:19:59] <rmatte> and the type should change from ppp to ppp_64

[06-Apr-2010 12:20:12] <rmatte> then you need to copy the ethernetCsmacd_64 template and rename it to ppp_64

[06-Apr-2010 12:20:16] <rmatte> and you should be good after that

[06-Apr-2010 12:20:25] <mistich> seen this once before chet :)

[06-Apr-2010 12:20:41] <QbY> ok, just changed version and remodeling

[06-Apr-2010 12:20:47] <cluther> There is a chance that the interface doesn't support the HC counters even at SNMPv2c. In this case, you're screwed.

[06-Apr-2010 12:21:25] <QbY> ok, its now ppp_64

[06-Apr-2010 12:21:51] <cluther> You'll lose your historical graph data now, but it will fill in correctly now.

[06-Apr-2010 12:22:07] <QbY> ok, i don't see where to copy

[06-Apr-2010 12:22:09] <cluther> mistich: I don't believe it. :)

[06-Apr-2010 12:22:16] <rmatte> k, copy ethernetCsmacd_64 to ppp_64 under /Devices/Templates

[06-Apr-2010 12:22:18] <rmatte> and you're done

[06-Apr-2010 12:22:24] <cluther> rmatte: That's not necessary.

[06-Apr-2010 12:22:57] <cluther> There was a behavior change in 2.5.2 (see that previous ticket I linked) that makes this work without copying the template.

[06-Apr-2010 12:23:09] <QbY> the only copy i have is to apply it to another device class

[06-Apr-2010 12:23:23] <rmatte> cluther: k, I deal with boatloads of different interface types so I'm used to making lots of templates

[06-Apr-2010 12:23:27] <rmatte> just to have more control

[06-Apr-2010 12:23:46] <rmatte> many interface types don't support the standard data collection

[06-Apr-2010 12:23:57] <cluther> Yeah, nothing wrong with doing it that way. This is a lucky case because ppp interfaces have octets, packets and errors.

[06-Apr-2010 12:23:59] <cain22> hmm....no events listed from my disk space

[06-Apr-2010 12:24:05] <rmatte> yup

[06-Apr-2010 12:25:18] <rmatte> QbY: don't bother copying it, you're fine apparently

[06-Apr-2010 12:25:29] <QbY> k

[06-Apr-2010 12:37:36] <QbY> have i said today how much i loves my zenoss

[06-Apr-2010 12:37:44] <QbY> got my graphs into our intranet portal

[06-Apr-2010 12:37:50] <QbY> alerts mailed the right people

[06-Apr-2010 12:37:57] <QbY> my life is wonderful

[06-Apr-2010 12:38:08] <Simon4_> want to fix my modelling issue? ;)

[06-Apr-2010 12:38:14] <rhettardo> ha

[06-Apr-2010 12:48:27] <aclark> QbY: awesome!

[06-Apr-2010 13:05:10] <Simon4_> ERROR zen.ZenModeler: frexp() result out of range

[06-Apr-2010 13:05:12] <Simon4_> arrrgh

[06-Apr-2010 13:05:18] <QbY> if i could JUsT figure out how to always send the freshest information in an alert, i'd have the bestest system on the block

[06-Apr-2010 13:11:23] <rmatte> QbY: how'd you manage the integration in to the portal?

[06-Apr-2010 13:11:57] <QbY> let me find the link

[06-Apr-2010 13:11:59] <cain22> has anyone used Windows 2008 core R2 and have it talk with Zenoss via snmp?

[06-Apr-2010 13:12:06] <QbY> essentially, its a wget

[06-Apr-2010 13:12:36] <jb> i have 2008 R2 boxes that talk to zenoss via snmp and wmi

[06-Apr-2010 13:12:50] <QbY> we found it on the  forums, and just re-worked it for php

[06-Apr-2010 13:12:52] <QbY> let me post both

[06-Apr-2010 13:13:05] <rmatte> cain22: what's windows 2008 core?  you talking about windows server?

[06-Apr-2010 13:13:22] <QbY> rmatte: docs/DOC-4441;jsessionid=9B22424798885015861CBB932B8684AE.node0

[06-Apr-2010 13:13:34] <rmatte> QbY: cool

[06-Apr-2010 13:13:52] <mrayzenoss> sounds like this job I just added to zenoss-jobs: http://www.freelance-jobs-now.com/project-430806-zenoss-integration-elance-job.html

[06-Apr-2010 13:15:54] <QbY> rmatte: http://pastebin.com/8wgFzCqE  -- this one is ours..  that way we didn't hav ea cron job firing every 5 minutes and scping unnecessarily

[06-Apr-2010 13:15:58] <QbY> loads super fast

[06-Apr-2010 13:15:59] <venturaville> mrayzenoss: did you see what I said about the netapp zenpack?

[06-Apr-2010 13:16:09] <mrayzenoss> it worked for you no problems?

[06-Apr-2010 13:16:24] <mrayzenoss> I was trying it with the vanilla 2.5.2 vmware image

[06-Apr-2010 13:16:39] <mrayzenoss> I'll retrace my steps

[06-Apr-2010 13:18:36] <rmatte> too bad my php isn't up to par or I'd bid on it

[06-Apr-2010 13:19:17] <QbY> what do you mean?

[06-Apr-2010 13:19:22] <QbY> use my script..  super simple

[06-Apr-2010 13:19:30] <QbY> just replace it with your data

[06-Apr-2010 13:20:56] <venturaville> mrayzenoss: I don't have a vanilla 2.5.2 image, but it is close..  if need be I can blow away my install and try it from a fresh install of it

[06-Apr-2010 13:24:24] <rmatte> QbY: I think mrayzenoss was talking about something else ;)

[06-Apr-2010 13:24:30] <QbY> oh

[06-Apr-2010 13:26:23] <cain22> rmatte: Windows Server 2008 includes a variation of installation called Server Core. Server Core is a significantly scaled-back installation where no Windows Explorer shell is installed. All configuration and maintenance is done entirely through command line interface windows, or by connecting to the machine remotely using Microsoft Management Console.

[06-Apr-2010 13:27:12] <cain22> having an issue getting snmp to either work, or work with zenoss...

[06-Apr-2010 13:27:18] <cain22> with core

[06-Apr-2010 13:27:35] <jb> ?

[06-Apr-2010 13:27:39] <jb> can you snmpwalk it from your zenoss box?

[06-Apr-2010 13:27:53] <cain22> nope

[06-Apr-2010 13:28:05] <cain22> but, snmp service is on

[06-Apr-2010 13:28:13] <jb> did you configure the ACLs?

[06-Apr-2010 13:28:21] <cain22> and i've changed the reg key to use public authentication

[06-Apr-2010 13:28:44] <jb> it won't respond to anything out of the box

[06-Apr-2010 13:28:47] <jb> you have to configure an ACL

[06-Apr-2010 13:30:34] <rmatte> cain22: I see, sounds fun

[06-Apr-2010 13:30:41] <cain22> hmm...i'll have to check that out

[06-Apr-2010 13:30:45] <rmatte> cain22: I've never tried monitoring core before

[06-Apr-2010 13:31:05] <cain22> neither have i, at least till now

[06-Apr-2010 13:31:06] <cain22> lol

[06-Apr-2010 13:31:10] <rmatte> lol

[06-Apr-2010 13:31:17] <jb> its just like any other version of windows

[06-Apr-2010 13:31:24] <jb> where you have to configure an ACL :P

[06-Apr-2010 13:31:26] * rmatte whispers "should have just gone with linux"

[06-Apr-2010 13:31:35] <forsberg> :P

[06-Apr-2010 13:32:02] <rmatte> Unless you're running an exchange server or a domain server, there's no reason to use Windows :P

[06-Apr-2010 13:32:28] <forsberg> exchange is not an excuse anymore, use zarafa!

[06-Apr-2010 13:32:29] <forsberg> :)

[06-Apr-2010 13:32:32] <mistich> thats not even a good reason

[06-Apr-2010 13:32:40] <jb> quit being so closed minded :P

[06-Apr-2010 13:32:41] <rmatte> Zimbra is pretty pimp too

[06-Apr-2010 13:32:43] <jb> windows has it's place

[06-Apr-2010 13:32:47] <jb> just as linux does

[06-Apr-2010 13:32:50] <forsberg> denied

[06-Apr-2010 13:32:51] <forsberg> :>

[06-Apr-2010 13:32:51] <mistich> in the trash

[06-Apr-2010 13:32:58] <mrayzenoss> lol

[06-Apr-2010 13:33:02] <jb> you obviously don't work in a large shop

[06-Apr-2010 13:33:03] <jb> :P

[06-Apr-2010 13:33:15] <mrayzenoss> lolx2

[06-Apr-2010 13:33:24] <rmatte> hehe

[06-Apr-2010 13:33:32] <mistich> were just a little shop

[06-Apr-2010 13:33:40] <ckrough> heh

[06-Apr-2010 13:34:14] <ckrough> I'm not sure I've seen an equal for Active Directory

[06-Apr-2010 13:34:21] <jb> theres not one

[06-Apr-2010 13:34:31] <jb> novell has its thing..

[06-Apr-2010 13:34:42] <rmatte> honestly, that "big shop" argument is kind of BS... Linux can be used regardless of the size of the company, as long as the admins actually know what they are doing

[06-Apr-2010 13:34:49] <jb> heh

[06-Apr-2010 13:34:52] <rmatte> but it's harder to come by linux admins than it is to come by windows admins

[06-Apr-2010 13:34:57] <jb> no way..

[06-Apr-2010 13:35:02] <jb> a GOOD windows admin is not easy to find

[06-Apr-2010 13:35:13] <jb> a "OMG I JUST GOT MY MCSE I KNOW EVERYTHING" windows admin is easy to find :P

[06-Apr-2010 13:35:25] <rmatte> MCSE is garbage

[06-Apr-2010 13:35:36] <mistich> sorry I was just joking windows has a place

[06-Apr-2010 13:35:47] <jb> and, as ckrough said.. there really is no replacement for active directory

[06-Apr-2010 13:35:51] <jb> especially in a large enterprise :/

[06-Apr-2010 13:35:55] <venturaville> +1 to jb's statement about windows admins....

[06-Apr-2010 13:35:55] <rmatte> and I was speaking in the context of linux admins versus windows admins... there are obviously going to be a lot more windows admins since windows is currently more widely used

[06-Apr-2010 13:36:02] <rmatte> it doesn't mean you need to run windows in a big shop

[06-Apr-2010 13:36:02] <QbY> here we go..  rmatte, were you here the day i was talking about my threshold problem?

[06-Apr-2010 13:36:08] <venturaville> the best windows admins I know of are all linux admins who don't want to admit it

[06-Apr-2010 13:36:33] <rmatte> QbY: depends which problem that was

[06-Apr-2010 13:36:36] <QbY> i have an event that is going on right now, and even though it gets worse we only get email alerts from the first time it tripped--we don't get updated ones showing that the problem has gotten worse until i history it

[06-Apr-2010 13:36:41] <jb> until there is a decent replacement for AD, then I'll pretty much say windows is necessary in a large enterprise.

[06-Apr-2010 13:37:03] <jb> I am a sr linux admin that dabbles in windows..

[06-Apr-2010 13:37:06] <rmatte> QbY: yeh, that's just the way Zenoss behaves

[06-Apr-2010 13:37:18] <QbY> any solution

[06-Apr-2010 13:37:26] <rmatte> QbY: you can set a repeat time on the alerts to re-send an email every so many minutes until it's cleared

[06-Apr-2010 13:37:33] <rmatte> not sure if it would send updated data though

[06-Apr-2010 13:37:54] <rmatte> you'll see an input field called Repeat

[06-Apr-2010 13:37:57] <rmatte> in the alerting rules

[06-Apr-2010 13:38:00] <rmatte> with a 0 in it

[06-Apr-2010 13:38:20] <rmatte> I believe that represents minutes, so if you change the 0 to a 5, an alert will be sent every 5 minutes until the event is cleared

[06-Apr-2010 13:38:31] <QbY> rmatte: i have a repeat set..  and it does it every 5.. because its a very important alarm..  however, what's happening is, right now i have 100 modems that are offline. but when it first broke the threshold it was 20 modems offline..

[06-Apr-2010 13:38:39] <QbY> we need to know that its 100

[06-Apr-2010 13:38:45] <QbY> but zenoss keeps saying 20

[06-Apr-2010 13:39:00] <forsberg> make more thresholds maybe?

[06-Apr-2010 13:39:06] <rmatte> Zenoss is probably holding the original alert in some queue and working off that rather than the event itself

[06-Apr-2010 13:39:30] <rmatte> If that's how it's coded, there's not much you can do about that

[06-Apr-2010 13:39:39] <rmatte> not without development work anyways

[06-Apr-2010 13:39:40] <QbY> hrmm

[06-Apr-2010 13:39:54] <QbY> see, i'd like in each email alert to use the freshest status

[06-Apr-2010 13:40:11] <rmatte> yeh I understand, but there's no setting for that at the moment

[06-Apr-2010 13:40:18] <rmatte> and it's stupid that it doesn't use the freshest info

[06-Apr-2010 13:40:27] <rmatte> you can always log a bug report in trac and it'll get fixed eventually

[06-Apr-2010 13:40:37] <QbY> yeah..  well i gotta run out

[06-Apr-2010 13:40:40] <QbY> be back shortly

[06-Apr-2010 13:40:40] <mistich> eventually key word :)

[06-Apr-2010 13:40:51] <QbY> hehehe

[06-Apr-2010 13:40:57] <rmatte> yeh, not eventually but eveennnnnnnnnnnnnnnntuaaaaaaaaaalllly

[06-Apr-2010 13:41:10] <rmatte> yyyyyyyyyyyyyyyyyyyyyyyyy

[06-Apr-2010 13:41:12] <rmatte> yyyyyyyyyyyyyyyyyyyyyyyyy

[06-Apr-2010 13:41:17] <rmatte> y

[06-Apr-2010 13:41:21] <rmatte> :)

[06-Apr-2010 13:43:29] <bigegor> jb: Novell's eDirectory is an alternative for AD

[06-Apr-2010 13:43:43] <mistich> don't get him started :)

[06-Apr-2010 13:44:08] <Simon4_> these conversations always end well :)

[06-Apr-2010 13:44:14] <jb> yeah, I mentioned novell :)

[06-Apr-2010 13:45:26] <bigegor> but not eDirectory ;)

[06-Apr-2010 13:45:53] <jb> well, yeah.. sorry

[06-Apr-2010 13:46:09] <jb> i admittingly don't know anything about it..

[06-Apr-2010 13:46:49] <bigegor> really ???

[06-Apr-2010 13:46:56] <jb> nope :/

[06-Apr-2010 13:47:04] <jb> i know what it is, but not much beyond that

[06-Apr-2010 13:47:10] <ckrough> its like LDAP, but with an 'e' in front of it

[06-Apr-2010 13:47:17] * ckrough duck and cover

[06-Apr-2010 13:47:22] <jb> :P

[06-Apr-2010 13:47:47] <ckrough> I've never admin'd it, just been a victim

[06-Apr-2010 13:48:28] <bigegor> it commes first with NetWare 4, as NDS (Novell Directory Service), i thing in 1995 or 1994

[06-Apr-2010 13:49:46] <bigegor> LDAP is much younger

[06-Apr-2010 13:50:13] <jb> but its based on LDAP, no?

[06-Apr-2010 13:50:53] <bigegor> no

[06-Apr-2010 13:51:01] <jb> ah

[06-Apr-2010 13:51:32] <bigegor> it based on x500

[06-Apr-2010 13:54:28] <bigegor> but now eDirectory can operate as LDAP server too

[06-Apr-2010 13:56:54] <bigegor> http://en.wikipedia.org/wiki/Novell_eDirectory

[06-Apr-2010 13:57:26] <ckrough> 'released in 1993', LDAP RFC is dated 1997.

[06-Apr-2010 14:02:16] <bigegor> i've got my CNA certification in 1997 and Master CNE 1999

[06-Apr-2010 14:06:26] mrayzenoss1 is now known as mrayzenoss

[06-Apr-2010 14:24:11] <rmatte> AD uses LDAP at it's core as well

[06-Apr-2010 14:25:46] <mistich> they all do

[06-Apr-2010 14:26:53] <rmatte> yeh, it's pretty much the standard

[06-Apr-2010 14:27:10] <rmatte> eugh, I frigging hate dns sometimes

[06-Apr-2010 14:28:03] <rmatte> I moved my personal DNS server, but for some reason either the secondary or the third dns server (setup on a couple of my friend's servers) isn't receiving updates, even though their configuration has been modified properly

[06-Apr-2010 14:28:18] <rmatte> so my DNS server is sending out the correct new IP, and one of their servers is sending out the old IP

[06-Apr-2010 14:28:27] <rmatte> so the IP is flipping throughout the day

[06-Apr-2010 14:28:43] <mistich> lol

[06-Apr-2010 14:28:44] <ckrough> only one should be authoritative?

[06-Apr-2010 14:29:43] <rmatte> no, you can have more than 1 authoritative

[06-Apr-2010 14:30:43] <rmatte> although I don't even know why their servers are getting hit up for requests since my server is listed as the primary

[06-Apr-2010 14:30:54] <rmatte> theirs should only get queried if my server goes down

[06-Apr-2010 14:37:20] <sciolist> TTL and caching: two things that taste great together!

[06-Apr-2010 14:46:48] <willwh> zenoss saves the day again

[06-Apr-2010 14:47:33] <willwh> couple of http down notifications made me point our data centre guys to a switch - turns out it had started failing, approx a min prior to the notification

[06-Apr-2010 14:47:39] <willwh> replaced faster than you can say boo

[06-Apr-2010 14:47:42] <willwh> <3 you guys.

[06-Apr-2010 14:47:58] <willwh> I do have a little question about reports though

[06-Apr-2010 14:48:15] <willwh> The Performance Reports are fantastic.... to a point

[06-Apr-2010 14:48:50] <willwh> on the interface detail you can have a graph of usage - I just need to dig around to throw something like that together to plot all of our monitored interfaces average throughput on

[06-Apr-2010 14:48:56] <willwh> has anyone done anything similar?

[06-Apr-2010 14:49:01] <willwh> would love some pointers

[06-Apr-2010 14:49:46] <cluther> willwh: How's your Python?

[06-Apr-2010 14:52:20] <willwh> cluther: not wonderful

[06-Apr-2010 14:52:33] <willwh> I can likely figure it out

[06-Apr-2010 14:53:50] <cluther> willwh: There is a ZenPack demonstrating some example techniques in the community trac - http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.example.Techniques

[06-Apr-2010 14:54:28] <cluther> willwh: The specific changeset that introduced a complex custom report that shows an aggregate of hits on a web cluster is http://zenpacks.zenoss.org/trac-zenpacks/changeset/116

[06-Apr-2010 14:55:06] <cluther> willwh: You could use almost the exact same code to generate aggregate throughput graph reports.

[06-Apr-2010 15:01:04] <willwh> cluther: thank you kindly

[06-Apr-2010 15:01:09] <willwh> may be back with questions in a bit

[06-Apr-2010 15:06:34] <ckrough> when making changes to a large number of devices via zendmd, is there a risk of overwriting concurrent changes when you commit() at the end?

[06-Apr-2010 15:07:04] <ckrough> ie: if the zendmd script takes 4 hours, then I commit(), anything that was affected by the script but changed outside of the script would be clobbered. ya?

[06-Apr-2010 15:07:12] <cluther> No. You will get a ConflictError exception rather than overwriting anything.

[06-Apr-2010 15:07:47] <cluther> Most likely if your script takes 4 hours you should be committing intermittently or you'll certainly get to the end and have a massive transaction that is uncommitable.

[06-Apr-2010 15:08:12] <ckrough> k

[06-Apr-2010 15:12:12] <blull> Hello.  I'm working on deploying a zenoss core installation enterprise wide.  I've got a remote collector deployed and connecting back to the central zenhub, however I am having problems with figuring out how to get the remote collector to perform a discovery scan, and not the collector running on the same host as the zenhub.

[06-Apr-2010 15:12:33] <blull> I added the remote collector as a device, then added a new collector with the same name as the device I added.

[06-Apr-2010 15:13:03] <blull> I then specified that it should have 192.168.19.0/24 assigned to it for discovery

[06-Apr-2010 15:13:09] <blull> in the collector properties

[06-Apr-2010 15:13:20] <straterra> win 3

[06-Apr-2010 15:13:30] <blull> then when I kick of an auto-discovery job, it never gets pushed to the remote collector.

[06-Apr-2010 15:13:32] <blull> Any ideas?

[06-Apr-2010 15:13:40] <blull> been rtfm'ing and googling my head off, no luck :/

[06-Apr-2010 15:13:45] <cluther> blull: You have to run zendisc on the collector.

[06-Apr-2010 15:13:57] <cluther> Kicking it off from the UI will only ever use the local zendisc.

[06-Apr-2010 15:14:23] <blull> okay

[06-Apr-2010 15:14:24] <blull> so

[06-Apr-2010 15:14:32] <blull> i thought zenjobs handled that?

[06-Apr-2010 15:14:40] <blull> and kicked off zendisc on a as-needed basis?

[06-Apr-2010 15:15:06] <cluther> It's a reasonable thought, but zenjobs is unfortunately not that smart.

[06-Apr-2010 15:15:12] <blull> so on the remote collector for zendisc.conf, just give it my hubhost and monitor as i did my other daemons?

[06-Apr-2010 15:15:19] <cluther> Exactly.

[06-Apr-2010 15:15:25] <blull> okay cool, 1 sec let me try this.

[06-Apr-2010 15:15:27] <blull> btw

[06-Apr-2010 15:15:41] <blull> I have communication running through compressed ssl tunnels :)

[06-Apr-2010 15:15:53] <blull> stunnel is my friend :)

[06-Apr-2010 15:16:35] <blull> also, I had 1 other question, not sure if you guys will know.

[06-Apr-2010 15:16:48] <mistich> shoot

[06-Apr-2010 15:17:00] <ckrough> what did you break

[06-Apr-2010 15:17:07] <ckrough> :)

[06-Apr-2010 15:17:10] <blull> Considering zenoss is python, and there appears to be no open source zenpack that provides equiv functionality of the enterprise distributed collector zenpack

[06-Apr-2010 15:17:25] <jb> blull: someone -was- working on one..

[06-Apr-2010 15:17:28] <ckrough> rmatte is working on one

[06-Apr-2010 15:17:28] <mrayzenoss> blull: Collector Tool

[06-Apr-2010 15:17:33] <blull> ahh

[06-Apr-2010 15:17:46] <jb> did it ever get finished?

[06-Apr-2010 15:17:47] <mrayzenoss> blull: and Egor has one he's working on as well

[06-Apr-2010 15:17:49] <blull> mrayzenoss, you're talkinga bout one that auto configures the remote collector?

[06-Apr-2010 15:17:53] <blull> honestly

[06-Apr-2010 15:18:03] <mrayzenoss> docs/DOC-5831

[06-Apr-2010 15:18:20] <mrayzenoss> blull: Collector Tool generates configs, you manually copy them over

[06-Apr-2010 15:18:23] <blull> looking through the api, I think I could just basically hook all i/o and handle it all via a custom internal routing type system based on different props in the zeodb

[06-Apr-2010 15:18:50] <blull> mrayzenoss, thats not what I'm talking about.. im talking about via web ui... simple as cliking "add new remote collector" and throw in the ip :)

[06-Apr-2010 15:18:58] <mrayzenoss> yeah, Egor's working on that

[06-Apr-2010 15:19:09] <blull> he make any decent progress on it you know or need a hand?

[06-Apr-2010 15:19:18] <mrayzenoss> I've got a copy to put into SVN, my queue is kinda backed up right now

[06-Apr-2010 15:19:22] <blull> I would very much like something like that considering we're bringing in a ton of new datacenters in the near future heh

[06-Apr-2010 15:19:23] <mrayzenoss> email me and I'll send it to you

[06-Apr-2010 15:19:42] <blull> thanks, address?

[06-Apr-2010 15:19:46] <mrayzenoss> mray@zenoss.com

[06-Apr-2010 15:19:52] <blull> k

[06-Apr-2010 15:20:24] <jb> so what happens if egor's functionality is better than enterprise's? :P

[06-Apr-2010 15:20:25] <blull> i really actually may be able to provide some help if its required, I've been fluent with python and have developed enterprise class products for administration (ie. puppet management systems, etc.) :)

[06-Apr-2010 15:21:07] <mrayzenoss> jb: Enterprise will add something else

[06-Apr-2010 15:21:11] <jb> :P

[06-Apr-2010 15:21:28] <jb> blull: hm, what puppet management system did you develop?

[06-Apr-2010 15:21:50] <blull> jb, i spent about 4 months developing an appliance type system for puppet

[06-Apr-2010 15:22:00] <jb> whats it called?

[06-Apr-2010 15:22:05] <jb> or is it private?

[06-Apr-2010 15:22:07] <blull> 3 parts to it, a centralized "management" server, appliance servers that are plug and play, and a thick client

[06-Apr-2010 15:22:11] <blull> still haven't released it yet

[06-Apr-2010 15:22:19] <blull> was talking with the puppet guys about getting them involved

[06-Apr-2010 15:22:25] <blull> but they haev a different direction than I want to go

[06-Apr-2010 15:22:31] <blull> so financing it out of pocket at the moment

[06-Apr-2010 15:22:34] <jb> ie, dashboard?

[06-Apr-2010 15:22:37] <blull> nah

[06-Apr-2010 15:22:40] <blull> more remote management

[06-Apr-2010 15:22:42] <mrayzenoss> blull: there's a Puppet ZenPack too if you're interested

[06-Apr-2010 15:22:55] <jb> so the "management" server is a puppetmaster?

[06-Apr-2010 15:23:00] <blull> nah

[06-Apr-2010 15:23:10] <blull> management system controls all the appliance servers, that each run a puppet master.

[06-Apr-2010 15:23:15] <jb> oh.

[06-Apr-2010 15:23:17] <blull> picture a python -> ruby interface

[06-Apr-2010 15:23:19] <jb> interesting

[06-Apr-2010 15:23:26] <blull> for access all the info on the local puppetmaster appliance

[06-Apr-2010 15:23:34] <jb> and you have something that controls puppet clients too?

[06-Apr-2010 15:23:40] <jb> ie, classes/environments/etc?

[06-Apr-2010 15:24:17] <blull> Yup, it has a built in IDE as the moment using some pretty syntax highlighting for developing manifests... I'm working on a "manifest wizard" so you can at least create fairly useful blank slates to customize for your env

[06-Apr-2010 15:24:35] <jb> whoa that sounds awesome

[06-Apr-2010 15:24:36] <blull> it also allows you to easily apply manifests visually instead of just by basics as host name or subnet

[06-Apr-2010 15:24:40] <blull> ie

[06-Apr-2010 15:24:41] <jb> have you kept this secret or something?

[06-Apr-2010 15:24:49] * jb wonders why he hasn't heard of this :P

[06-Apr-2010 15:24:49] <blull> right click a folder that has say all your database servers in it

[06-Apr-2010 15:25:00] <blull> select the manifests to apply and the order priority

[06-Apr-2010 15:25:02] <blull> then boom

[06-Apr-2010 15:25:07] <blull> all the clients start using them

[06-Apr-2010 15:25:11] <blull> well

[06-Apr-2010 15:25:22] <blull> i spoke to the puppet devs on two different occassions at the end of 2008

[06-Apr-2010 15:25:32] <blull> they were dumb founded at why I chose python over ruby

[06-Apr-2010 15:25:42] <blull> and why I was doing things the way I was

[06-Apr-2010 15:25:44] <blull> when it came down to it

[06-Apr-2010 15:25:49] <blull> they have a programmer view of requirements

[06-Apr-2010 15:26:01] <blull> I have nearyl 15 years of sys admin experience in huge environments for fortune 100 companies

[06-Apr-2010 15:26:20] <jb> yeah it sounds nice

[06-Apr-2010 15:26:23] <blull> ie. when you would see cnn.com reports on paypal.com going down... I was usually the one who actually implemented or created the fix for the situation

[06-Apr-2010 15:26:32] <blull> so long story short I know where I'm coming from enterprise side

[06-Apr-2010 15:26:35] <blull> i just happen to program a lot to.

[06-Apr-2010 15:26:41] <blull> so we just clashed in direction

[06-Apr-2010 15:26:47] <blull> so i went my own way and have had it under wraps ever since.

[06-Apr-2010 15:26:52] <blull> they are doing something now that is web based.

[06-Apr-2010 15:26:55] <mrayzenoss> blull: Zenoss is always looking for Python devs :)

[06-Apr-2010 15:26:58] <blull> hehe

[06-Apr-2010 15:27:07] <jb> blull: they are doing puppet dashboard

[06-Apr-2010 15:27:10] <blull> yeah I was thinking about dedicating some time to a oss project

[06-Apr-2010 15:27:16] <jb> blull: but theres another product, foreman, which imo, is better..

[06-Apr-2010 15:27:17] <venturaville> could use some help with the puppet zenpack .....

[06-Apr-2010 15:27:22] <blull> jb, yeah, this is an enterprise management infrastructure... not a "dashboard"

[06-Apr-2010 15:27:32] <blull> a dashboard overview is just one of the many features I've incorporated.

[06-Apr-2010 15:27:37] <blull> anyway I'm calling the app Prototype

[06-Apr-2010 15:27:41] <jb> so you havethis in protection?

[06-Apr-2010 15:27:44] <jb> production rather

[06-Apr-2010 15:27:50] <blull> not where I work now

[06-Apr-2010 15:27:59] <blull> I've got a couple people in dev and lab environments

[06-Apr-2010 15:28:00] <jb> any info about it on the web?

[06-Apr-2010 15:28:08] <blull> ie. friend of mine over @ yelp.com is toying with it

[06-Apr-2010 15:28:13] <blull> nope

[06-Apr-2010 15:28:19] <blull> i've really had it underwraps for awhile

[06-Apr-2010 15:28:21] <blull> its pretty slick

[06-Apr-2010 15:28:31] <blull> the thick client is cross platform qt based

[06-Apr-2010 15:28:35] <venturaville> sounds a bit like a much better opsware

[06-Apr-2010 15:28:40] <jb> yeah

[06-Apr-2010 15:28:41] <blull> so it reall does natively work on windows, linux, osx

[06-Apr-2010 15:28:42] <venturaville> :-P

[06-Apr-2010 15:28:46] <jb> the thick client, runs on all puppet clients?

[06-Apr-2010 15:28:49] <jb> instead of puppetd?

[06-Apr-2010 15:28:52] <blull> nah

[06-Apr-2010 15:28:55] <blull> so

[06-Apr-2010 15:29:03] <blull> pupet gets installed on the individual node

[06-Apr-2010 15:29:06] <jb> ok

[06-Apr-2010 15:29:15] <blull> puppetmasterd is installed on what I refer to as the "appliance"

[06-Apr-2010 15:29:22] <venturaville> which reminds me... zenoss could really use an active monitoring daemon to run on OS clients...

[06-Apr-2010 15:29:27] <blull> the appliance runs puppetmasterd and my own python daemon w/ a local sql database

[06-Apr-2010 15:29:43] <venturaville> one that didn't require hacking away at SNMP sub agents or relying on pounding away at sshd

[06-Apr-2010 15:29:53] <blull> another daemon runs and pushes updates to the "management" server which also has a high level overview of the entire puppet infrastructure

[06-Apr-2010 15:30:25] <blull> when the daemon connects to "publish" it checks a queue for pending jobs assigned to it, and hands it off to the local "appliance" process which in turn updates puppetmasterd configs, etc.

[06-Apr-2010 15:30:36] <blull> works very well in a tiered environment

[06-Apr-2010 15:30:49] <blull> ie: you can go out but not backwards towards DB end.

[06-Apr-2010 15:31:05] <blull> but anyway yeah, if you want more info, email me.

[06-Apr-2010 15:31:08] <blull> ben@orbitalsecurity.com

[06-Apr-2010 15:31:14] <blull> I'm looking soon for people to get involved as well.

[06-Apr-2010 15:31:19] <jb> cool

[06-Apr-2010 15:31:21] <jb> i'll drop you a mail

[06-Apr-2010 15:31:22] <blull> Cause its becoming to big to manage myself.. its thousands of lines.

[06-Apr-2010 15:31:24] <blull> werd :)

[06-Apr-2010 15:35:09] <jb> but, that sounds very cool

[06-Apr-2010 15:35:32] <jb> a tree based hiearchy to apply classes/manifests to would be nice.

[06-Apr-2010 15:38:10] <bigegor> later

[06-Apr-2010 15:38:17] <jb> later

[06-Apr-2010 15:44:10] <wamcvey> Hey all... had a quick question... for monitoring Unix/Linux systems over SSH, rather than a static password associated with a username, is it possible to register an SSH private key into zenoss and then add that key to the authorized keys associated with the target accounts?

[06-Apr-2010 15:44:17] <mrayzenoss> yes

[06-Apr-2010 15:44:29] <venturaville> check the key path var as well

[06-Apr-2010 15:44:45] <mrayzenoss> the zProperty 'zKeyPath'

[06-Apr-2010 15:45:22] <wamcvey> Ahh... Ok. thanks all.

[06-Apr-2010 15:57:12] <mrayzenoss> venturaville: you around?

[06-Apr-2010 15:57:28] <blull> alright, bout to kick off a discovery from remote collector *crosses fingers*

[06-Apr-2010 15:58:55] <blull> omg its working! :)

[06-Apr-2010 16:00:33] <venturaville> mrayzenoss: for a short bit

[06-Apr-2010 16:00:59] <mrayzenoss> ok, just rebuilt the NetApp from SVN latest and all it has is a collector plugin

[06-Apr-2010 16:01:38] <mrayzenoss> I'm gonna go look at the diffs

[06-Apr-2010 16:02:14] <mrayzenoss> doesn't complain about the /Storage device class not being there either

[06-Apr-2010 16:02:26] <mrayzenoss> brand-new 2.5.2 vmware appliance

[06-Apr-2010 16:06:24] <venturaville> the Storage device class is now created via the init.py file

[06-Apr-2010 16:06:49] <venturaville> I'll have to go try a vanilla zenoss stack install tomorrow, and see if I can replicate it

[06-Apr-2010 16:07:01] <mrayzenoss> yeah, just noticed that

[06-Apr-2010 16:07:15] <mrayzenoss> but no NetApp sub-organizer

[06-Apr-2010 16:07:53] <blull> okay

[06-Apr-2010 16:07:54] <blull> well

[06-Apr-2010 16:08:04] <blull> the configs aren't being pushed to the remote detector

[06-Apr-2010 16:08:11] <blull> i added a new subnet to the collector

[06-Apr-2010 16:08:14] <blull> rather collector, not detector

[06-Apr-2010 16:08:28] <blull> anyway I added a new subnet to the remote collector, then kicked off an auto discovery job.

[06-Apr-2010 16:08:38] <blull> it kicked off on the main zenhub, not the remote collector

[06-Apr-2010 16:08:44] <blull> ideas?

[06-Apr-2010 16:09:30] <root____1> anyone have any input on creating query-able custom properties?

[06-Apr-2010 16:10:05] root____1 is now known as ptom

[06-Apr-2010 16:17:42] <ptom> anyone?

[06-Apr-2010 16:18:46] <rmatte> what do you mean by "queryable"?

[06-Apr-2010 16:21:07] <ptom> I want to assign a value to each device that I can query for. Like a link ID.

[06-Apr-2010 16:21:41] <ptom> but I need to be able to set/unset that custom property programitically.

[06-Apr-2010 16:24:42] <mrayzenoss> ptom: look at 4.2.1 in the Admin Guide: Managing Custom Device Properties

[06-Apr-2010 16:25:06] <mrayzenoss> you can access them programatically

[06-Apr-2010 16:25:21] <blull> how do I get zenoss zenhub to push config updates to a remote collector?

[06-Apr-2010 16:25:26] <blull> i guess what daemon would do that?

[06-Apr-2010 16:25:45] <ptom> but could I do something like this? return [ (d.id, d.getManageIp(), d.custom_link_id(), d.getLocationName()) for d in context.getSubDevices() ]

[06-Apr-2010 16:25:52] <blull> ie. I added a new network to the remote collector, than kicked off a discovery scan.  It ran the discovery scan on the zenhub, and not the remote collector.

[06-Apr-2010 16:26:33] <mrayzenoss> blull: that might be a bug in Enterprise too

[06-Apr-2010 16:27:13] <blull> mrayzenoss, oh, rmphf.  a restart caused the discovery to kick off without issue automatically on the remote collector however.

[06-Apr-2010 16:27:27] <wamcvey> So... My Linux systems don't have snmpd running, and I'd rather not install it (at least for now) just to get zenoss to shut up about snmp agent being down. I've removed the zenoss.snmp.* collectors for the device class I've moved my servers into. Anything else I should do? (I'm still seeing the alert/event about the agent being down).

[06-Apr-2010 16:27:36] <blull> i wouldn't need to run zenactions on the remote collector would I?

[06-Apr-2010 16:28:36] <mrayzenoss> wamcvey: zSnmpMonitorIgnore

[06-Apr-2010 16:28:38] <ptom> wamcvey: you can edit the properties to ignore snmp under the zproperties

[06-Apr-2010 16:29:07] <mrayzenoss> blull: I honestly don't know.  I don't mess with Enterprise so I don't know

[06-Apr-2010 16:29:27] <blull> i don't run enterprise :)

[06-Apr-2010 16:29:31] <blull> i run oss version

[06-Apr-2010 16:29:37] <blull> zenoss core right?

[06-Apr-2010 16:29:41] <mrayzenoss> yeah, but I work for Zenoss :p

[06-Apr-2010 16:29:53] <mrayzenoss> right

[06-Apr-2010 16:30:06] <blull> yeah core == open source non commercial supported version

[06-Apr-2010 16:30:11] <blull> thus im asking for non commercial support :)

[06-Apr-2010 16:30:33] <ptom> is zendisc running on the remote collector? check zendisc.conf

[06-Apr-2010 16:30:39] <blull> It is.

[06-Apr-2010 16:31:24] <blull> it only appears to work no the remote collector after a restart

[06-Apr-2010 16:31:43] <ptom> I don't have discovery turned on on my remote collectors, haven't messed with that

[06-Apr-2010 16:32:04] <blull> ooo what a second

[06-Apr-2010 16:32:08] <blull> when I add a device

[06-Apr-2010 16:32:22] <blull> and im doing an advanced add, i can select the collector

[06-Apr-2010 16:32:33] <blull> how would I "advance add" an entire subnet?

[06-Apr-2010 16:32:38] <blull> and then discover it?

[06-Apr-2010 16:33:02] <ptom> i would use either xml-rpc or REST calls

[06-Apr-2010 16:33:09] <rmatte> blull: no, zenactions is not required on the remote collector

[06-Apr-2010 16:33:23] <blull> ptom, so in other words, it doesn't work through the web ui?

[06-Apr-2010 16:33:33] <blull> well atleast not without a restart apparently

[06-Apr-2010 16:33:49] <ptom> haven't tried mysql. but I have done bulk adds using rest calls

[06-Apr-2010 16:33:56] <ptom> err myself, not mysql

[06-Apr-2010 16:33:59] <ptom> long day

[06-Apr-2010 16:34:55] <rmatte> blull: what you want to do can be scripted, but it's not supported directly in the UI

[06-Apr-2010 16:35:03] <blull> rmatte, rmphf, that sucks :/

[06-Apr-2010 16:35:08] <rmatte> you basically want to add devices before discovering them

[06-Apr-2010 16:35:11] <rmatte> correct?

[06-Apr-2010 16:35:16] <blull> no

[06-Apr-2010 16:35:17] <blull> i want to discover them

[06-Apr-2010 16:35:22] <blull> but kick off the discover from a remote collector

[06-Apr-2010 16:35:25] <blull> not from the zenhub server.

[06-Apr-2010 16:35:52] <rmatte> well, zenhub obviously has to do the data processing

[06-Apr-2010 16:35:54] <ptom> and you can specify performanceMonitor using a rest call

[06-Apr-2010 16:36:26] <blull> rmatte, when adding a device manually, i can assign i tto a collector.

[06-Apr-2010 16:36:36] <blull> rmatte, inside collectors, i can specify ranges to assign to that collector

[06-Apr-2010 16:36:40] <rmatte> oh, I see what you mean

[06-Apr-2010 16:36:44] <rmatte> you're just talking about assignment

[06-Apr-2010 16:36:47] <blull> yes

[06-Apr-2010 16:36:57] <blull> then kicking off a initial discovery on that remote collector

[06-Apr-2010 16:37:05] <blull> as the zenhub server can not route to that network.

[06-Apr-2010 16:37:10] <blull> its via tunnels.

[06-Apr-2010 16:37:57] <blull> right now, when I do an auto discover on a subnet I have mapped to a remote collector, it discovers it from the zenhub until I reboot both the zenhub and the remote collector.

[06-Apr-2010 16:38:04] <rmatte> hhmmm, yeh, that field appears to be missing in core

[06-Apr-2010 16:38:11] <blull> I am running core

[06-Apr-2010 16:38:18] <blull> core != enterprise :)

[06-Apr-2010 16:38:19] <blull> right?

[06-Apr-2010 16:38:22] <rmatte> correct

[06-Apr-2010 16:38:40] <rmatte> the autodiscover in the UI doesn't support specifying what collector to use for the devices

[06-Apr-2010 16:38:47] <blull> right.

[06-Apr-2010 16:38:52] <rmatte> which is most likely just an oversight

[06-Apr-2010 16:38:56] <blull> i c

[06-Apr-2010 16:39:02] <blull> i thought though that you specify it at the collector level?

[06-Apr-2010 16:39:20] <blull> Default Discovery Networks

[06-Apr-2010 16:39:32] <blull> is the setting when you are in the /Monitors/<remote collector> section

[06-Apr-2010 16:40:21] <mrayzenoss> Updated WMI and WBEM Data Source ZenPacks: docs/DOC-3392 docs/DOC-3409  Egor said they're good to go

[06-Apr-2010 16:40:50] <rmatte> blull: I'm not sure that that actually does what you think it does

[06-Apr-2010 16:40:55] <rmatte> trying to find info on it now...

[06-Apr-2010 16:41:09] <blull> rmatte, yeah i couldn't find info on google or zenoss about it, just a posting saying "this is obvious" heh

[06-Apr-2010 16:41:23] <blull> rmatte, well it appears to take effect after a restart

[06-Apr-2010 16:41:25] <blull> if i remove a subnet from there

[06-Apr-2010 16:41:33] <blull> it no longer gets used by the collector

[06-Apr-2010 16:41:36] <rmatte> yeh I just saw that posting

[06-Apr-2010 16:41:37] <blull> its acting like a bug =p

[06-Apr-2010 16:42:01] <rmatte> then file a but report: dev.zenoss.com/trac login with zenoss/zenoss

[06-Apr-2010 16:42:06] <rmatte> bug*

[06-Apr-2010 16:42:06] <blull> k

[06-Apr-2010 16:42:48] <blull> tomorrow though, right now I have to get this working as I have to let it run over night in parrallel with our existing nagios installation for comparison.

[06-Apr-2010 16:42:56] <blull> have to have results by friday :/

[06-Apr-2010 16:43:47] <ptom> so can anyone point me in the right direction for adding a custom property and being able to set/unset it programatically ?

[06-Apr-2010 16:44:37] <ptom> I would also like to be able to query this custom property with a script

[06-Apr-2010 16:44:51] <rmatte> ptom: what is the custom property going to store?

[06-Apr-2010 16:44:57] <ptom> ints

[06-Apr-2010 16:45:10] <rmatte> right, but for what?

[06-Apr-2010 16:45:19] <rmatte> I'm trying to get an idea of what exactly you're trying to do

[06-Apr-2010 16:45:20] <ptom> we are trying to link our monitoring devices with our inventory system

[06-Apr-2010 16:45:32] <ptom> i need them both to have the same linkid

[06-Apr-2010 16:45:51] <rmatte> oh, so you're not talking about device properties then

[06-Apr-2010 16:46:37] <ptom> no, I am. each device will have a custom property containing this unique id

[06-Apr-2010 16:46:46] <rmatte> you'd have to read up on zope, that's what Zenoss stores all of it's settings/properties in

[06-Apr-2010 16:46:51] <rmatte> oh

[06-Apr-2010 16:47:21] <mrayzenoss> ptom: you not see my message earlier?

[06-Apr-2010 16:47:30] <ptom> like, say I wanted to use ou'd have to read up on zope, that's what Zenoss stores all of it's settings/properties in

[06-Apr-2010 16:47:34] <ptom> 17:46 < rmatte> oh

[06-Apr-2010 16:47:36] <ptom> 17:47 <@mrayzenoss> ptom: you not see my message earlier?

[06-Apr-2010 16:47:38] <rmatte> go to a device page and in the dropdown select more -> custom

[06-Apr-2010 16:47:54] <rmatte> that's where you can see the custom properties...

[06-Apr-2010 16:47:58] <rmatte> now... to actually create them...

[06-Apr-2010 16:48:33] <mrayzenoss> stupid UI, they're called "Custom Schema"

[06-Apr-2010 16:48:54] <rmatte> Click on devices

[06-Apr-2010 16:49:08] <rmatte> and go to More -> custom Schema

[06-Apr-2010 16:49:27] <rmatte> then creat the property, make sure it starts with a lower case c

[06-Apr-2010 16:49:34] <rmatte> so something like cLinkID

[06-Apr-2010 16:49:36] <rmatte> or whatever

[06-Apr-2010 16:49:52] <ptom> mrayzenoss: missed it, sorry. i will check that out.

[06-Apr-2010 16:49:56] <rmatte> mrayzenoss: know of any documentation on playing with those in zendmd?

[06-Apr-2010 16:50:05] <mrayzenoss> same as a zProperty

[06-Apr-2010 16:50:16] <ptom> rmatte: can I query those?

[06-Apr-2010 16:50:38] <rmatte> ptom: yeh, you can

[06-Apr-2010 16:50:49] <mrayzenoss> ptom: rmatte: Admin Guide 6.3.2 has example in TALES

[06-Apr-2010 16:50:51] <rmatte> How are you planning to querry?

[06-Apr-2010 16:51:23] <ptom> something like this. return [ (d.id, d.getManageIp(), d.custom_link_id(), d.getLocationName()) for d in context.getSubDevices() ]

[06-Apr-2010 16:51:45] <rmatte> ok, so you're just doing it all in python... easy enough

[06-Apr-2010 16:51:55] <ptom> but if that won't work I can use whatever means

[06-Apr-2010 16:52:12] <rmatte> create the property like I said for starters

[06-Apr-2010 16:52:13] <ptom> if i don't have to create a function to query then that's even better

[06-Apr-2010 16:52:16] <ptom> k

[06-Apr-2010 16:52:17] <rmatte> and let me dig up the query you need

[06-Apr-2010 16:52:55] <ptom> ok, I created it.

[06-Apr-2010 16:53:48] <rmatte> hmmm, well I found how to set the property

[06-Apr-2010 16:53:57] <rmatte> now I just need to find how to get the property

[06-Apr-2010 16:54:22] <rmatte> setting it is like: d._setProperty("cWhatever", "value")

[06-Apr-2010 16:54:52] <rmatte> oh, actually getting them is even easier

[06-Apr-2010 16:54:56] <rmatte> just d.cWhatever

[06-Apr-2010 16:55:01] <ptom> awesome

[06-Apr-2010 16:55:08] <ptom> thank you, I will give it a try

[06-Apr-2010 16:55:13] <rmatte> np

[06-Apr-2010 17:10:04] <rmatte> well, I'm out, later guys

[06-Apr-2010 18:07:45] <wamcvey> So... I'm adding a host which has network filtering that drops ping packets. Where do I go to turn off (ideally at a class level, so I can add similar hosts to the class) ping checks?

[06-Apr-2010 18:40:35] <theacolyte> if I import MIBS, will past events translate?

[06-Apr-2010 18:46:48] <theacolyte> apparently the answer to that is no

[06-Apr-2010 19:25:54] forsberg is now known as fOrsberg

[07-Apr-2010 00:00:30] [disconnected at Wed Apr  7 00:00:30 2010]

[07-Apr-2010 00:00:30] [connected at Wed Apr  7 00:00:30 2010]

[07-Apr-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[07-Apr-2010 02:36:05] fOrsberg is now known as forsberg

[07-Apr-2010 02:40:49] <Troubadix09> morning all

[07-Apr-2010 05:28:12] forsberg is now known as fOrsberg

[07-Apr-2010 05:36:37] fOrsberg is now known as forsberg

[07-Apr-2010 07:54:10] <blull> Morning, quick question.  I performed a automated discovery sweep however it appears none of the dns names were resolved.  I think I fixed the issue on my side, im trying to figure out how to manually re-run a template assigned to a class?

[07-Apr-2010 07:57:49] <ckrough> I don't think you can rediscover by class

[07-Apr-2010 07:57:57] <ckrough> or remodel

[07-Apr-2010 07:58:28] <ckrough> you may be able to do a loop over that class in zendmd

[07-Apr-2010 07:59:45] <ckrough> blull: have you started toying with zendmd yet?

[07-Apr-2010 07:59:56] <ckrough> I mean, at all

[07-Apr-2010 08:00:56] <blull> nah not yet

[07-Apr-2010 08:01:09] <blull> just started playing with this the other

[07-Apr-2010 08:01:58] <ckrough> thats what I thought. you should check it out, opens up a lot of options, especially if you know python

[07-Apr-2010 08:03:17] <blull> yeah im pretty fluent with python

[07-Apr-2010 08:03:25] <blull> just been trying to see what capabilities I have through the web ui

[07-Apr-2010 08:03:26] <ckrough> you'll love using zendmd directly then

[07-Apr-2010 08:03:32] <ckrough> ya

[07-Apr-2010 08:03:40] <blull> wish I had more :)

[07-Apr-2010 08:03:46] <blull> i don't wanna have to script anything :)

[07-Apr-2010 08:04:52] <ckrough> docs/DOC-3187

[07-Apr-2010 09:03:03] forsberg is now known as fOrsberg

[07-Apr-2010 09:51:47] <mistich> good morning

[07-Apr-2010 09:55:43] <rmatte> g'day

[07-Apr-2010 09:56:36] <mistich> up for a question?

[07-Apr-2010 09:57:21] <mistich> I have a script /zport/dmd/findDeviceByIp I want to add it to the zenpack where would I add it in my init.py

[07-Apr-2010 09:57:22] <rmatte> sure

[07-Apr-2010 09:58:00] <rmatte> honestly I don't know off the top of my head... but you would have to write some python code which actually generates that object

[07-Apr-2010 09:58:10] <rmatte> from what I recall

[07-Apr-2010 09:58:32] <rmatte> have you checked around on the community site to see if there are any examples kicking around on the wiki?

[07-Apr-2010 09:58:45] <rmatte> maybe also checkout the examples ZenPack (not sure if it covers that sort of thing or not)

[07-Apr-2010 09:59:10] <rmatte> if cluther is around later he'd be a good person to ask

[07-Apr-2010 09:59:15] <mistich> yes not much there all for extending a particular class such as device or reports I want to extend dmd

[07-Apr-2010 09:59:30] <rmatte> yeh

[07-Apr-2010 09:59:59] <rmatte> I've never tried that or seen it done quite like what you're trying to do

[07-Apr-2010 10:00:15] <mistich> yes I'm special in that way :)

[07-Apr-2010 10:00:21] <rmatte> but you need to figure out python code to generate the object then throw that in to __init__.py

[07-Apr-2010 10:00:30] <mistich> thanks

[07-Apr-2010 10:00:45] <rmatte> you might be better to check zope documentation than zenoss documentation for that

[07-Apr-2010 10:01:35] <mrayzenoss> mistich: post that on zenoss-dev and I'll try to get a dev to respond

[07-Apr-2010 10:02:01] <rmatte> mrayzenoss to the rescue

[07-Apr-2010 10:02:20] <mrayzenoss> not a rescue yet, I'm just going for help :p

[07-Apr-2010 10:02:44] <rmatte> hehe

[07-Apr-2010 10:12:40] <mistich> boy sure they will let me in there :)

[07-Apr-2010 10:12:58] <mistich> no one in there but me

[07-Apr-2010 10:14:56] <mistich> #zenoss-dev on freenode correct

[07-Apr-2010 10:15:07] <mrayzenoss> meant the forums

[07-Apr-2010 10:15:14] <mistich> ahhh

[07-Apr-2010 10:15:41] <mistich> I gave up on them a long time ago will try again

[07-Apr-2010 10:18:13] <rmatte> well, he meant to post the question in the zenoss-dev section of the forums then pass the link to him

[07-Apr-2010 10:18:17] <rmatte> and he'll make sure a dev responds

[07-Apr-2010 10:18:26] <mistich> ahhh

[07-Apr-2010 10:22:49] <mistich> thread/13226

[07-Apr-2010 10:23:00] <mistich> ok posted thanks

[07-Apr-2010 10:36:07] <cluther> mistich: I just responded to the thread.

[07-Apr-2010 10:42:56] <mistich> @monkeypatch('Products.ZenModel.DataRoot.DataRoot')

[07-Apr-2010 10:43:01] <mistich> thats what I needed

[07-Apr-2010 10:43:12] <mistich> jos for people who are logging

[07-Apr-2010 10:43:13] fOrsberg is now known as forsberg

[07-Apr-2010 11:16:53] <jb> hmm

[07-Apr-2010 11:16:56] <jb> when configurting alert filters

[07-Apr-2010 11:17:01] <jb> i can't have one "Component Contains"

[07-Apr-2010 11:17:07] <jb> and one "Component doesn't contain"

[07-Apr-2010 11:17:09] <jb> ?

[07-Apr-2010 11:21:29] <rmatte> nope

[07-Apr-2010 11:21:40] <rmatte> it's limited to one operator per property

[07-Apr-2010 11:22:02] <rmatte> not sure if that's by design or just an oversight

[07-Apr-2010 11:22:26] <jb> hrm

[07-Apr-2010 11:22:31] <jb> ok so, if I add more than one "Message" filer

[07-Apr-2010 11:22:33] <jb> it uses "and"

[07-Apr-2010 11:22:38] <jb> no way to make that "or" ?

[07-Apr-2010 11:22:48] <rmatte> nope

[07-Apr-2010 11:22:51] <jb> grr

[07-Apr-2010 11:22:59] <jb> ok, so I can't do this.

[07-Apr-2010 11:22:59] <jb> hm

[07-Apr-2010 11:23:19] <rmatte> make another alerting rule for the other condition is the only way

[07-Apr-2010 11:23:36] <jb> ?

[07-Apr-2010 11:23:56] <jb> well, im trying to say "if message contains this, this, or that" don't send the alert

[07-Apr-2010 11:24:20] <rmatte> oh, how are you getting hung up on that?

[07-Apr-2010 11:24:26] <rmatte> the and isn't literally a python and

[07-Apr-2010 11:24:38] <jb> so it won't try to match all of them?

[07-Apr-2010 11:24:40] <rmatte> it's really an or

[07-Apr-2010 11:24:42] <jb> ok

[07-Apr-2010 11:24:48] <jb> perfect

[07-Apr-2010 11:25:13] <rmatte> actually...

[07-Apr-2010 11:25:31] <rmatte> I might be wrong there, just thinking about how I've configured filters in the past...

[07-Apr-2010 11:25:41] * rmatte checks

[07-Apr-2010 11:25:42] <jb> oh wow

[07-Apr-2010 11:25:43] <jb> hrm

[07-Apr-2010 11:25:46] <jb> it put "or"

[07-Apr-2010 11:25:50] <jb> and then 3 "ands"

[07-Apr-2010 11:26:41] <rmatte> I got it to do a bunch of ors

[07-Apr-2010 11:26:42] <jb> http://thunder.jbdesign.net/~jbaird/zenoss/zenossalert.jpg

[07-Apr-2010 11:26:57] <rmatte> just don't change the "contains" when you're adding the others

[07-Apr-2010 11:27:13] <rmatte> yeh, I don't see any ands

[07-Apr-2010 11:27:36] <rmatte> so I'm not sure how you're even getting that

[07-Apr-2010 11:27:42] <jb> yeah hrm

[07-Apr-2010 11:27:49] <jb> i didnt change the "does not contain"

[07-Apr-2010 11:28:22] <jb> now im only getting ands!

[07-Apr-2010 11:28:24] <jb> hmm

[07-Apr-2010 11:29:12] <jb> and now only or's using "contain"

[07-Apr-2010 11:29:14] <jb> might be a bug

[07-Apr-2010 11:29:32] <rmatte> well, add them as "contain" then change it to "does not"

[07-Apr-2010 11:29:36] <rmatte> hopefully that'll work

[07-Apr-2010 11:29:42] <jb> yeah thats what I did

[07-Apr-2010 11:29:42] <rmatte> and yes, it probably is a bug

[07-Apr-2010 11:29:48] <jb> haha

[07-Apr-2010 11:29:51] <rmatte> that filters section doesn't really operate as I'd expect it to

[07-Apr-2010 11:29:52] <jb> it changed them all to and's when I saved

[07-Apr-2010 11:29:53] <jb> damnit

[07-Apr-2010 11:30:06] <rmatte> no idea then

[07-Apr-2010 11:30:21] <rmatte> well, actually...

[07-Apr-2010 11:30:26] <rmatte> think of it this way...

[07-Apr-2010 11:30:37] <rmatte> those are conditions that must be met before the alert is sent

[07-Apr-2010 11:31:02] <rmatte> so if message does not contain whatever AND if message does not contain whatever AND if message does not contain whatever THEN send alert

[07-Apr-2010 11:31:06] <rmatte> which is what you want

[07-Apr-2010 11:31:41] <rmatte> the system obviously knows that message can't be 3 different things at once

[07-Apr-2010 11:31:58] <rmatte> so I doubt the label of and instead of or makes any difference behaviouraly at all

[07-Apr-2010 11:32:18] <jb> ok i'll try it

[07-Apr-2010 11:32:21] <rmatte> k

[07-Apr-2010 11:32:22] <jb> maybe it will work fine with "and"

[07-Apr-2010 11:32:28] <rmatte> probably

[07-Apr-2010 11:32:42] <rmatte> or actually wouldn't work too well

[07-Apr-2010 11:32:53] <rmatte> actually or wouldn't work at all

[07-Apr-2010 11:33:14] <jb> well, I it's never going to match all of the and's

[07-Apr-2010 11:33:15] <rmatte> if message does not contain blah OR if message does not contain BLAH or if message does not contain BLAH

[07-Apr-2010 11:33:21] <rmatte> it's always not going to contain 1 of the 3

[07-Apr-2010 11:33:23] <jb> it will only match one of them..

[07-Apr-2010 11:33:29] <rmatte> hence why you need to use AND

[07-Apr-2010 11:33:46] <rmatte> you want to make sure that it doesn't contain any of the 3 values and only continue if all 3 conditions are met

[07-Apr-2010 11:33:57] <rmatte> so I think it changing it from or to and is the system being intelligent

[07-Apr-2010 11:34:33] <rmatte> you're thinking of it backwards

[07-Apr-2010 11:35:20] <rmatte> you don't want it to contain any of the 3 pieces of text

[07-Apr-2010 11:35:31] <rmatte> and only if it meets all 3 conditions to continue

[07-Apr-2010 11:35:37] <rmatte> so you have to use AND

[07-Apr-2010 11:35:50] <jb> ah ok

[07-Apr-2010 11:35:55] <jb> hopefully this will work then :)

[07-Apr-2010 11:35:56] <jb> thanks

[07-Apr-2010 11:35:59] <rmatte> np

[07-Apr-2010 11:36:09] <rmatte> obviously and wouldn't work with contains

[07-Apr-2010 11:36:16] <rmatte> since it can't contain 3 things at once

[07-Apr-2010 11:36:27] <rmatte> but and works great with does not contain, since it can not contain 3 things at once

[07-Apr-2010 11:36:45] <rmatte> ;)

[07-Apr-2010 11:36:50] <rmatte> it's just a logic thing

[07-Apr-2010 11:43:24] <rhettardo> any one use mmm and monitor it w/ zenoss?

[07-Apr-2010 11:46:32] <rmatte> mmm?

[07-Apr-2010 11:47:12] <rhettardo> http://mysql-mmm.org/

[07-Apr-2010 11:47:42] <rmatte> ah, cool

[07-Apr-2010 11:47:47] <rmatte> nope, can't say that we do

[07-Apr-2010 11:47:57] <rhettardo> thought so

[07-Apr-2010 11:48:41] <rhettardo> it dosnt seem to be all that popular.  figured i'd ask since there are some zabbix scripts apparently.  i havnt checked them out though

[07-Apr-2010 11:48:46] <rhettardo> didnt really want to either

[07-Apr-2010 11:51:08] <rhettardo> its been a while...where do i change the DB auth/host settings at?

[07-Apr-2010 12:13:27] <forsberg> sorry to be so out of line, but any iscsi workers here?

[07-Apr-2010 12:14:06] <rhettardo> i use iscsi

[07-Apr-2010 12:17:04] <forsberg> im going insane, seems pretty secret which target is which device

[07-Apr-2010 12:18:46] <rhettardo> openfiler handles my iscsi setup :\

[07-Apr-2010 12:25:17] <forsberg> only the targets i guess ?

[07-Apr-2010 12:27:35] <rhettardo> yeah

[07-Apr-2010 12:28:10] <forsberg> yeah thats the easy part, anyways think i actually found it, in /dev/disk/by-path

[07-Apr-2010 12:28:14] <forsberg> wtf its retardo!

[07-Apr-2010 12:28:16] <forsberg> :)

[07-Apr-2010 13:08:05] <ptom> does anyone know of a way that I could search for a device by mac address?

[07-Apr-2010 13:08:32] <blull> well, i can only give one suggestion

[07-Apr-2010 13:08:38] <blull> it most likely involves some type of typing.

[07-Apr-2010 13:08:43] <blull> thats all the help I can be at this time :/

[07-Apr-2010 13:08:47] <ptom> heh

[07-Apr-2010 13:08:50] <ptom> ty

[07-Apr-2010 13:09:10] <ptom> I was wondering if this could be done by an indexed catalog

[07-Apr-2010 13:15:43] <mrayzenoss> ptom: docs/DOC-3512

[07-Apr-2010 13:16:02] <mrayzenoss> and there's a zendmd one about it on the wiki

[07-Apr-2010 13:23:45] <rhettardo> where do i go to change the mysql host

[07-Apr-2010 13:24:23] <ptom> mrayzenoss: thanks, I will have a look

[07-Apr-2010 13:25:15] <Simon4> rhettardo: not sure if there's somewhere on zenoss.com - mrayzenoss will correct me, but I've always ended up at http://www.sysadminwiki.net/wiki/index.php?title=Common_Zenoss_Errors_-_Post_Install

[07-Apr-2010 13:25:23] <Simon4> first thing is where the username/password are in dmd, etc

[07-Apr-2010 13:26:04] <Simon4> if you want gui - go to "event manager" on the left menu

[07-Apr-2010 13:26:12] <Simon4> tis in there

[07-Apr-2010 13:33:10] <rhettardo> ha

[07-Apr-2010 13:33:12] <rhettardo> thanks

[07-Apr-2010 13:33:26] <rhettardo> i knew it was simple.  its been a good 2.5 years since ive had to worry about this

[07-Apr-2010 13:38:19] <rmatte> Simon4: where does it even explain what he was asking in there?

[07-Apr-2010 13:39:28] <forsberg> mysql info in event manager?

[07-Apr-2010 13:39:30] <forsberg> :)

[07-Apr-2010 13:39:43] <forsberg> Connection Information

[07-Apr-2010 13:39:48] <rmatte> oh, I was focussing on the link you gave

[07-Apr-2010 13:40:17] <Simon4> it was somewhat of brain fail :/ has been a long day. thought the link had host in there as well as user/pass

[07-Apr-2010 13:40:22] <rmatte> ah

[07-Apr-2010 13:40:23] <rmatte> k

[07-Apr-2010 13:40:50] * Simon4 goes back to a beer and stops answering questions for the evening :)

[07-Apr-2010 13:40:55] <rmatte> hehe

[07-Apr-2010 14:05:56] <rmatte> man, I need to come up with a better way to configure memory on my zenoss servers

[07-Apr-2010 14:06:08] <rmatte> some of them chew deep in to swap for no reason

[07-Apr-2010 14:12:20] <Simon4> rmatte: I get that, occasionally end up with the run.py process eating gigs and gigs of ram

[07-Apr-2010 14:12:38] <Simon4> was hoping 2.5.2 might help that some

[07-Apr-2010 14:17:42] <rmatte> well, so far my 2.5.2 test box is decent with memory consumption

[07-Apr-2010 14:17:58] <rmatte> going to leave it running a couple more weeks and see where it's at

[07-Apr-2010 14:18:30] <mrayzenoss> I know there's a ticket opened about zencommand leaking memory

[07-Apr-2010 14:18:36] <mrayzenoss> it's getting looked at right now

[07-Apr-2010 14:18:43] <Simon4> hmm

[07-Apr-2010 14:19:32] <rmatte> zenperfsnmp seems to be one of the leaks in 2.4.5...

[07-Apr-2010 14:19:37] <rmatte> hoping it's fixed in 2.5.2

[07-Apr-2010 14:23:36] <cluther> rmatte: Any daemon that uses RRDtool to do evaluate thresholds, zenperfsnmp bigtime, will leak prior to 2.5.

[07-Apr-2010 14:23:55] <cluther> It's a known memory leak in the RRDtool fetch routine in the version of RRDtool we shipped with 2.4.x

[07-Apr-2010 14:26:02] <cluther> Simon4: run.py is Zope. Zope uses RRDtool's fetch routine a *lot* so it's likely that 2.5.2 will keep it from consuming so much memory.

[07-Apr-2010 14:26:21] <Simon4> cluther: we're on 2.5.1 - did that leak also?

[07-Apr-2010 14:26:39] <cluther> I think we shipped the fixed RRDtool in 2.5.1.. let me check.

[07-Apr-2010 14:26:50] <Simon4> thanks

[07-Apr-2010 14:27:20] <cluther> Yeah, bad news. We were on the good, 1.3.8, version of RRDtool in 2.5.1. So that couldn't be the issue.

[07-Apr-2010 14:27:30] <Simon4> heh doh

[07-Apr-2010 14:27:38] <Simon4> it seems to be behaving currently, so will monitor it

[07-Apr-2010 14:27:55] <Simon4> another quick question, cachesize and pcachesize in zenhub.conf - what are they measured in?

[07-Apr-2010 14:28:17] <cluther> pcachesize is in MB, cachesize is in number of objects. Typically you can estimate objects to be 3KB each.

[07-Apr-2010 14:28:32] <cluther> pcachesize relates to the disk cache and cachesize relates to the memory cache.

[07-Apr-2010 14:28:35] <Simon4> cool, thanks

[07-Apr-2010 14:28:43] <cluther> Each should be multiplied by the number of workers you configure.

[07-Apr-2010 14:28:48] <Simon4> *nod*

[07-Apr-2010 14:29:00] <Simon4> wondering if we've been configured out of the ballpark :)

[07-Apr-2010 14:29:33] <cluther> Zope and zenhub also have another way to "leak" memory. That's the transaction cache.

[07-Apr-2010 14:29:47] <Simon4> okay

[07-Apr-2010 14:30:01] <cluther> Basically if you run up huge transactions they will consume a lot of memory. Even after they're aborted or committed Python won't return the memory to the OS, although it will be reused within Python itself.

[07-Apr-2010 14:30:04] <Simon4> I've seen that run.py zope process at 15-16GB, so interested in causes

[07-Apr-2010 14:30:17] <Simon4> ahhh

[07-Apr-2010 14:30:27] <cluther> The usual suspect is reports. "All Components" most commonly.

[07-Apr-2010 14:30:50] <Simon4> heh, we have a zenpack that would make that more than likely

[07-Apr-2010 14:31:54] <cluther> If you're doing anything that "takes a really long time" to complete I would recommend intermittently doing a transaction.savepoint() to flush the transaction cache to disk.

[07-Apr-2010 14:32:10] <cluther> Or if it is entirely a read-only operation you can intermittently abort the transaction.

[07-Apr-2010 14:35:21] <Simon4> thanks

[07-Apr-2010 14:35:52] <Simon4> are there ways to get the memory back for the OS - or is it just a restart?

[07-Apr-2010 14:36:24] <rmatte> cluther: good to hear that it should be fixed in 2.5

[07-Apr-2010 14:36:36] <rmatte> or IS fixed, rather

[07-Apr-2010 14:36:42] <Simon4> rmatte: i don't see any daemons eating ram in 2.5.1

[07-Apr-2010 14:37:20] <cluther> Simon4: Only a restart will release the memory back to the OS unfortunately.

[07-Apr-2010 14:37:39] <Simon4> cluther: good to know - at least from that I can find ways to manage it

[07-Apr-2010 14:37:48] <Simon4> vs whinging about memory leak or whatnot ;)

[07-Apr-2010 14:47:27] <bigegor> cluther: i have problem with WMI, if i send query 'select * from HP_ProcessorCacheMemory' with namespace 'root/HPQ' to Windows 2003 server with HP SIM Agents installed, i got every time 'WERR_INSUFFICIENT_BUFFER' error, but if i send 'select DiviceID,BlockSize, NumberOfBlocks from HP_ProcessorCacheMemory' its works. How can i increase BUFFERS in pywbem?

[07-Apr-2010 14:50:06] <bigegor> cluther: sorry, in pysamba

[07-Apr-2010 14:51:48] <cluther> Nothing that looks easy. The only thing I can find is a compile-time #define in samba itself.

[07-Apr-2010 14:53:37] <cluther> Any chance you can work around it by querying just for the columns you need?

[07-Apr-2010 14:57:46] <bigegor> cluther: another problem if Property doesn't exist i got another errors :( But the same Property can be pressent in one Windows Version but not in other :(

[07-Apr-2010 14:59:56] <bigegor> it looks like M$ changed WMI classes in every Windows Version :(

[07-Apr-2010 15:00:57] <mrayzenoss> New Group Resources Report ZenPack: docs/DOC-5833 Thanks venturaville!

[07-Apr-2010 15:17:21] <bigegor> cluther: thanks, I will try to fix my code.

[07-Apr-2010 15:18:08] <blull> Hey I'm trying to determine a way to figure out remote operating system from a device, without logging into it using zenoss.

[07-Apr-2010 15:18:18] <blull> Not even OS, just if its a Windows or other device.

[07-Apr-2010 15:18:33] <blull> so basically i was thinking check for a listener on 3389, if its there, it'll be treated as windows.

[07-Apr-2010 15:18:51] <blull> Is there a zenpack that does a simple tcp port check or should I just define a nmap check as a command?

[07-Apr-2010 15:21:54] <bigegor> blull: Ithink TCP445 is more common on Windows

[07-Apr-2010 15:22:36] <Simon4> blull: you could always just go look at the ipservice TCP3389 (or 445) and see which devices are listed, since the ipservice check will add things there automagically on modelling

[07-Apr-2010 15:23:13] <blull> heh, i think i may have accidently deleted default templates I needed.

[07-Apr-2010 15:23:37] <blull> oo i see

[07-Apr-2010 15:23:52] <blull> so ipservice just kinda reports which ports it sees getting traffic?

[07-Apr-2010 15:24:38] <Simon4> there are two ipservice modllers from memory, one's snmp based, the other is nmap based (portscan)

[07-Apr-2010 15:24:54] <Simon4> when the device is modeled, ports that are seen to be open are added to the ipservice checks

[07-Apr-2010 15:25:02] <Simon4> i.e. if port = open, we check it

[07-Apr-2010 15:25:19] <Simon4> then you can rename and enable any checks you actually care about, and they're checked (by default) every 60 seconds for life

[07-Apr-2010 15:25:41] <blull> hmm.. the ipservice doesn't model remote collector data?

[07-Apr-2010 15:25:53] <blull> i only see 1 data center when i browse classes -> services -> ipservice

[07-Apr-2010 15:26:04] <blull> not the one im trying to do some testing in that i did a discovery scan in

[07-Apr-2010 15:36:36] <rmatte> Simon4: the nmap based one is broken

[07-Apr-2010 15:36:43] <rmatte> there's a trac ticket open for it, and I submitted a patch

[07-Apr-2010 15:36:47] <Simon4> rmatte: ah, crap

[07-Apr-2010 15:36:50] <rmatte> lord knows when it'll actually get patched though

[07-Apr-2010 15:37:07] <rmatte> nmap also isn't distributed with Zenoss as a binary (it should be)

[07-Apr-2010 15:37:16] <blull> hrmm

[07-Apr-2010 15:37:21] <blull> not sure how to address this

[07-Apr-2010 15:37:32] <blull> rather best practice to approach this

[07-Apr-2010 15:37:33] <rmatte> snmp is the default method, but it's quite flaky in my experience

[07-Apr-2010 15:37:34] <blull> after a discovery

[07-Apr-2010 15:37:40] <blull> i have a ton of systems in our production datacenter

[07-Apr-2010 15:37:52] <blull> right now I am not going to be adding local accounts

[07-Apr-2010 15:38:12] <rmatte> I patched up a test system and tested with nmap, it worked a million times better than the snmp method

[07-Apr-2010 15:38:13] <blull> im trying to sort them based on os version

[07-Apr-2010 15:38:17] <rmatte> so I hope they fix it at some point

[07-Apr-2010 15:38:37] <blull> is there anyway to easily determine OS at a very high level, win32 vs. nix is all I care about

[07-Apr-2010 15:38:43] <Simon4> rmatte: snmp has some advantages in that it picks up things listening on interfaces that zenoss can't see

[07-Apr-2010 15:38:57] <blull> so I can make two categories and then disable certain templates (ie. our ntp template against the windows systems requires some mods that don't apply to linux)

[07-Apr-2010 15:39:03] <Simon4> btu yeah, you're at the mercy of net-snmp's visibility of netstat -l etc

[07-Apr-2010 15:39:05] <rmatte> blull: like, automatically categorize them?

[07-Apr-2010 15:39:10] <blull> rmatte, yes basically

[07-Apr-2010 15:39:14] <mrayzenoss> rmatte: you don't happen to have your ticket number for that do you?

[07-Apr-2010 15:39:16] <blull> rmatte, or some way I can easily categorize them

[07-Apr-2010 15:39:22] <blull> don't mind it being manual

[07-Apr-2010 15:39:28] <rmatte> http://dev.zenoss.org/trac/ticket/5238#comment:12

[07-Apr-2010 15:39:56] <rmatte> mrayzenoss: I think my patch is missing a small modification to a second file which was included in a previous patch

[07-Apr-2010 15:40:02] <rmatte> but either way, all of the code is there

[07-Apr-2010 15:40:24] <mrayzenoss> rmatte: if you want to update the patch that's on there, I'll see about raising the priority so it gets rolled into 3.0

[07-Apr-2010 15:40:36] <mrayzenoss> I hate having patches left out

[07-Apr-2010 15:40:59] <rmatte> if I can remember how to put patches for 2 different files in to 1 patch

[07-Apr-2010 15:41:51] <rmatte> I need to merge http://dev.zenoss.org/trac/attachment/ticket/5238/patchfile.patch with the single modification to __init.py__ in http://dev.zenoss.org/trac/changeset/16617

[07-Apr-2010 15:43:40] <mrayzenoss> added myself to the CC, I'll see the patch get attached and remind myself to bring it up at the next defect review

[07-Apr-2010 15:45:01] <rmatte> k

[07-Apr-2010 15:48:50] <blull> okay

[07-Apr-2010 15:49:18] <blull> so I have a template NtpMonitor at /Devices.  I have Linux and Windows classes under /Devices.. I want to disable ntp for windows but not for any other devices.

[07-Apr-2010 15:49:30] <blull> How would I disable a template for 1 entire device class?

[07-Apr-2010 15:49:35] <blull> promote and mark enabled = false?

[07-Apr-2010 15:50:37] <blull> or does it not take all inheritance by default and I have to promote to get a class working for a sub class?

[07-Apr-2010 15:51:30] <Simon4> bind the ntpmonitor template only to the /Devices/Linux class

[07-Apr-2010 15:51:37] <Simon4> it can still be defined in /Devices

[07-Apr-2010 15:51:49] <blull> so you have to bind manually to each class you define?

[07-Apr-2010 15:51:56] <blull> it doesn't auto-inherit it?

[07-Apr-2010 15:52:10] <Simon4> it will inherit as long as you don't bind something else

[07-Apr-2010 15:52:22] <blull> rmm

[07-Apr-2010 15:52:23] <blull> im confused

[07-Apr-2010 15:52:24] <blull> so

[07-Apr-2010 15:52:29] <blull> i have ntp bound to /devices

[07-Apr-2010 15:52:29] <Simon4> so if you bind to "Devices" both your windows and linux boxes will try and use it

[07-Apr-2010 15:52:34] <blull> right

[07-Apr-2010 15:52:42] <Simon4> if they're under /Devices/Windows and /Devices/Linux

[07-Apr-2010 15:52:51] <Simon4> if you bind it under /Devices/Linux

[07-Apr-2010 15:52:53] <blull> how do you exclude a class than?

[07-Apr-2010 15:52:58] <Simon4> and unbind it under /Devices/Windows

[07-Apr-2010 15:53:05] <blull> like say i want it for all Linux systems excluding the ones under Linux/specialstuff

[07-Apr-2010 15:53:16] <blull> how would I exclude that 1 class?

[07-Apr-2010 15:53:56] <Simon4> go to /Devices/Linux/Specialstuff, and change the bound templates to not include that template, that then overrides the bound templates for anything under that organiser

[07-Apr-2010 15:54:07] <Simon4> basically things will inherit unless overridden

[07-Apr-2010 15:54:32] <Simon4> the zProperty zDeviceTemplates is what you're after

[07-Apr-2010 15:54:52] <Simon4> you'll see it inheriting, and can override it in the gui

[07-Apr-2010 15:55:02] <blull> oh i c, I have to select multiple bindings at once.

[07-Apr-2010 15:55:24] <Simon4> I find it easier to just edit the zproperty vs playing in the templates tab, but ymmv

[07-Apr-2010 15:57:18] <rmatte> mrayzenoss: done - http://dev.zenoss.org/trac/attachment/ticket/5238/patchfile-new.2.patch

[07-Apr-2010 16:01:52] <rmatte> mrayzenoss: one of my comments explains some other stuff that needs to be done as well

[07-Apr-2010 16:02:31] <chudler> does anyone monitor CIFS (windows shares) from zenoss?  Any examples? Looking actually to do transactions, not the usual disk-space monitoring

[07-Apr-2010 16:05:16] <mrayzenoss> I take it you saw the zenpack forum post?

[07-Apr-2010 16:05:40] <chudler> I'm thinking its going to require mounting the share and scripting it up..  I wonder if anyone else has tackled this.  One of my cow-orkers probably did post something about this to the forum heh

[07-Apr-2010 16:06:00] <rmatte> zenpack forum post?

[07-Apr-2010 16:06:10] <mrayzenoss> message/47544#47544

[07-Apr-2010 16:06:44] <mrayzenoss> this is tangentially related: docs/DOC-3384

[07-Apr-2010 16:08:32] <chudler> mrayzenoss:  jmontgomery is a poor developer over here tasked with this, so he's leaning on me :-) Hence my question.

[07-Apr-2010 16:12:06] <rmatte> mrayzenoss: most of the patch wasn't even me, I just fixed some small issues that still existed with it

[07-Apr-2010 16:12:18] <mrayzenoss> is that a 2.5.2 patch?

[07-Apr-2010 16:12:49] <rmatte> I tested it in 2.4.5 but those files don't appear to have changed from 2.4.5 to 2.5.2

[07-Apr-2010 16:12:57] <rmatte> so it should work fine

[07-Apr-2010 16:13:38] <Simon4> rmatte: bladechassis updates have hit SVN... waiting on venturaville etc before actually releasing it, but if you're in the mood to be cutting edge it's there

[07-Apr-2010 16:13:53] <rmatte> Simon4: nah, I'll wait patiently

[07-Apr-2010 16:13:57] <rmatte> ;)

[07-Apr-2010 16:14:01] <Simon4> smart plan :)

[07-Apr-2010 16:14:10] <Simon4> he has lots of chassis, apparently

[07-Apr-2010 16:14:17] <rmatte> yeh, he does

[07-Apr-2010 16:14:20] <rmatte> we only have 1

[07-Apr-2010 16:14:22] * Simon4 tested on a C3000 today that he found lying about and it worked okay

[07-Apr-2010 16:14:30] <rmatte> cool

[07-Apr-2010 16:14:39] <rmatte> you just have blade chassis "laying around"?

[07-Apr-2010 16:15:10] <rmatte> ...and here I am fighting for another lab server pizzabox

[07-Apr-2010 16:15:14] <rmatte> since our current one is overloaded

[07-Apr-2010 16:15:17] <rmatte> lol

[07-Apr-2010 16:15:20] <Simon4> well... we were loaned it for another reason, btu I didn't see why I shouldn't spend 10 min pointing my dev VM at it to model it whilst it was up :)

[07-Apr-2010 16:15:29] <rmatte> ah

[07-Apr-2010 16:15:33] <rmatte> cool

[07-Apr-2010 16:15:44] <Simon4> have tested it in anger with about 16 C7000's

[07-Apr-2010 16:15:48] <rmatte> you made it sound like you found it under a staircase or something :P

[07-Apr-2010 16:15:50] <Simon4> so is behaving well, etc

[07-Apr-2010 16:15:56] <Simon4> under someone's desk? ;)

[07-Apr-2010 16:16:06] <rmatte> hehe

[07-Apr-2010 16:17:09] <Simon4> working on getting head around consolidating collection via ssh, vs soap and (unencrypted) snmp, since modelling now uses ssh. That or some kind of intelligent switching based on zProperties

[07-Apr-2010 16:17:25] <Simon4> will do some more work on it when sitting on a beach in two weeks and watching the world go by

[07-Apr-2010 16:17:46] <Simon4> (i.e. not distracted by real work heh)

[07-Apr-2010 16:20:09] <rmatte> lol

[07-Apr-2010 18:33:58] forsberg is now known as fOrsberg

[07-Apr-2010 22:33:04] <mrayzenoss> New Squid Monitoring ZenPack: docs/DOC-5832 Thanks jb!

[08-Apr-2010 00:00:30] [disconnected at Thu Apr  8 00:00:30 2010]

[08-Apr-2010 00:00:30] [connected at Thu Apr  8 00:00:30 2010]

[08-Apr-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[08-Apr-2010 02:12:06] fOrsberg is now known as forsberg

[08-Apr-2010 03:34:35] <penthium2> hello

[08-Apr-2010 03:38:36] <penthium2> how can I force a collection manualy ? ( ex : I just mount a nfs drive on a device, and I would like to force manualy the collecte to see the nfs-mount in the Os Tab, File systems setting )

[08-Apr-2010 03:39:23] <penthium2> I'm using SSH collect system and not the SNMP system

[08-Apr-2010 03:42:13] <forsberg> model device

[08-Apr-2010 03:43:57] <penthium2> thx

[08-Apr-2010 05:34:03] <penthium2> hello, I have a little issue, I lock a NFS drive from deletion with event alert. When I try to test the event alert ( umount device via ssh, and "manage" > "model device" ) the nfs drive stay in the "File systems" of zenoss but I haven't any event to alert me.

[08-Apr-2010 06:47:03] <ckrough> clear

[08-Apr-2010 07:05:38] <penthium2> hello, I have a little issue, I lock a NFS drive from deletion with event alert. When I try to test the event alert ( umount device via ssh, and "manage" > "model device" ) the nfs drive stay in the "File systems" of zenoss but I haven't any event to alert me.

[08-Apr-2010 07:15:01] <cluther> penthium2: Did you check the "Send event when actions are blocked by a lock" box when locking the file system?

[08-Apr-2010 07:15:29] <cluther> That would be one way to get an event like that.

[08-Apr-2010 07:16:37] <cluther> Are you monitoring via SNMP? If so, you may find a debug severity event related to "Error reading value for ...." because the OID for the file system no longer exists. You could transform this event into something more severe and obvious.

[08-Apr-2010 07:17:32] <penthium2> no I'm monitoring via ssh

[08-Apr-2010 07:18:29] <penthium2> and I check the : "send event when actions are blocked by a lock"

[08-Apr-2010 07:34:59] <JesperS> hi there, quick question. Am I right in assuming that Zenoss doesn't support asynchronious interfaces?

[08-Apr-2010 07:35:20] <JesperS> Eg. it detects my routers interface as Gbit, but in fact it is just a ADSL connection like 8Mbit/1Mbit. So can I pursurade Zenoss to accept both 8Mbit and 1Mbit as interface speeds?

[08-Apr-2010 07:36:34] <JesperS> as far as I can work out, there is only one Speed setting, in the Status/IpInterface section.

[08-Apr-2010 07:38:32] <JesperS> it would be great if Zenoss could (maybe it can?) be set up to alert on different Thresholds for in and out bandwidth usage.

[08-Apr-2010 07:38:50] <Simon4> you can

[08-Apr-2010 07:39:01] <JesperS> cool, how? :-)

[08-Apr-2010 07:39:46] <JesperS> right now I'm looking at /zport/dmd/Devices/rrdTemplates/ethernetCsmacd_64  and I guess it's there I need to change something, or?

[08-Apr-2010 07:40:27] <JesperS> like in the Thresholds section under   high utilization        MinMaxThreshold       ifHCInOctets_ifHCInOctets, ifHCOutOctets_ifHCOutOctets

[08-Apr-2010 07:42:10] <Simon4> that's the one

[08-Apr-2010 07:42:27] <Simon4> so you can make a local copy of the template into a device class that contains your adsl devices

[08-Apr-2010 07:42:35] <Simon4> then modify the thresholds in that local copy

[08-Apr-2010 07:43:20] <JesperS> and make two "high utilization" thresholds, one with InOctets and one with OutOctets, or?

[08-Apr-2010 07:45:10] <JesperS> or maybe create two templates?

[08-Apr-2010 09:03:13] forsberg is now known as fOrsberg

[08-Apr-2010 09:05:31] <ganiman> hello

[08-Apr-2010 09:06:11] <ganiman> I have a few devices which have mibs that change (for example wifi access points which change their mib each time you add or remove an ssid from it)

[08-Apr-2010 09:06:42] <ganiman> I need to do data collection on these devices, but dont want to keep changing a template every time the device reboots or changes it's config

[08-Apr-2010 09:06:58] <ganiman> opennms does this well - you can have opennms do snmpwalks instead of snmpgets

[08-Apr-2010 09:07:15] <ganiman> can this be done in zenoss?  I can't dig up documentation on this

[08-Apr-2010 09:11:54] <rmatte> ganiman: Zenoss doesn't even use snmpwalks unless you actually rig up a custom script as a command based datasource

[08-Apr-2010 09:12:01] <rmatte> it uses the twisted framework for SNMP

[08-Apr-2010 09:12:04] <rmatte> which is Python based

[08-Apr-2010 09:12:26] <ganiman> well, I am looking to achieve similar behavior

[08-Apr-2010 09:12:45] <rmatte> well, what exactly is changing on the devices when they reboot?

[08-Apr-2010 09:13:03] <ganiman> another example is, we have redline point to multipoint radios, and each radio which communicates with it gets a different id each time it connects

[08-Apr-2010 09:13:27] <ganiman> so if I want to graph SNR from a radio, the oid changes each time

[08-Apr-2010 09:13:50] <ganiman> if I have another radio connect, I dont want to have to do anything in zenoss - it should just be able to "walk" and see it

[08-Apr-2010 09:14:00] <rmatte> Well, you could create a custom script to collect the data... but is it always the same number of radio connections, or does it vary?

[08-Apr-2010 09:14:15] <ganiman> I have a hard time believing that zenoss does not already have a method to do this

[08-Apr-2010 09:14:38] <ganiman> the radios will var

[08-Apr-2010 09:14:41] <rmatte> The built in methods center around the OID not changing

[08-Apr-2010 09:14:42] <ganiman> vary*

[08-Apr-2010 09:14:52] <ganiman> well, that's kind of fail, isn't it?

[08-Apr-2010 09:15:05] <ganiman> many devices we work with have oids that change with each unique configuration

[08-Apr-2010 09:15:10] <rmatte> It doesn't stop you from doing the monitoring, you just need to write a simple script to do it

[08-Apr-2010 09:15:28] <rmatte> I've written scripts to collect CPU and Memory info from Windows devices... the OIDs are always different from device to device

[08-Apr-2010 09:15:30] <rmatte> and it works fine

[08-Apr-2010 09:15:50] <ganiman> especially when I have used other, less feature rich monitoring packages which have this kind of support out of the box

[08-Apr-2010 09:15:55] <rmatte> The main issue is that Zenoss doesn't support a variable number of OIDs (which is something they need to add)

[08-Apr-2010 09:16:23] <rmatte> Agreed, I'm pretty sure there are several feature request tickets open that center around that type of functionality

[08-Apr-2010 09:16:31] <rmatte> but right now they are focussing on a complete redesign of the UI

[08-Apr-2010 09:16:41] <rmatte> once that's done, they'll be able to get back to refining/enhancing features

[08-Apr-2010 09:17:09] <rmatte> But honestly, it's not that hard to write a little script that does what you're describing and integrate it with Zenoss

[08-Apr-2010 09:17:52] <rmatte> The main issue is the variable OIDs, but you could create graphs that have a certain amount of datapoints and just feed zeros to the ones that aren't active

[08-Apr-2010 09:18:32] <ganiman> I suppose,  but there is more to it than that

[08-Apr-2010 09:18:53] <ganiman> for example, one set of oids is the SSIDs an access point is serving

[08-Apr-2010 09:19:00] <ganiman> another set is the clients connected

[08-Apr-2010 09:19:09] <ganiman> another set is the mac addresses of clients connected

[08-Apr-2010 09:19:23] <ganiman> the last number in each oid is the same for the same device

[08-Apr-2010 09:19:53] <rmatte> see, that sort of stuff can be done, but zenoss collector plugin and UI components would need to be scripted and then packaged in to a ZenPack, which takes a fair amount of Python experience

[08-Apr-2010 09:20:01] <ganiman> so I can collected the name, mac, snr, and throughput (just an example) of a device by following it's ID

[08-Apr-2010 09:20:36] <ganiman> kind of a shame when something like opennms does this so well out of the box

[08-Apr-2010 09:21:02] <rmatte> Each monitoring system has it's strengths and weaknesses

[08-Apr-2010 09:21:09] <ganiman> well, not out of the box - have to tell it to walk oids in stead of get oids

[08-Apr-2010 09:21:32] <rmatte> yeh, I know what you mean, and that feature has been suggested for Zenoss several times...

[08-Apr-2010 09:21:39] <rmatte> and I'm sure it will eventually make it in to it

[08-Apr-2010 09:21:51] <rmatte> but they have a limited number of devs which are focussed on other aspects of Zenoss at the moment

[08-Apr-2010 09:21:53] <ganiman> yeah - we've been using zenoss for a couple years and have been considering moving to another  because of some of these snags

[08-Apr-2010 09:22:19] <rmatte> In your case it's not ideal for what you're wanting to do

[08-Apr-2010 09:22:26] <rmatte> but in the general amount of cases, it works very well

[08-Apr-2010 09:23:01] <rmatte> you could pay Zenoss Inc. to develop a ZenPack that does what you're wanting if you really wanted to

[08-Apr-2010 09:23:36] <rmatte> They'd be able to add a new tab to those devices that displays all of the different clients and such that are connected

[08-Apr-2010 09:23:40] <rmatte> then you could drill down on each

[08-Apr-2010 09:23:54] <ganiman> well that would be nice lol

[08-Apr-2010 09:24:06] <ganiman> that sounds like the requirement we are up against

[08-Apr-2010 09:24:24] <rmatte> It's completely doable, all of the framework is there, it just takes development work to put the necessary components together

[08-Apr-2010 09:25:02] <rmatte> you could also try posting it on the zenoss jobs section of the forum... some of the community members are quite good at that sort of thing as well

[08-Apr-2010 09:25:25] * Simon4 will write it for one million dollars :)

[08-Apr-2010 09:25:31] <rmatte> hehe

[08-Apr-2010 09:26:34] <rmatte> hopefully they will add the ability to walk an OID and gather data based on that in the future

[08-Apr-2010 09:26:40] <ganiman> I may be able to write it - but I need to learn more about the insides of zen

[08-Apr-2010 09:26:51] <rmatte> well, do you know Python well enough?

[08-Apr-2010 09:27:07] <ganiman> I know many other languages well enough - but I hardly use python

[08-Apr-2010 09:27:13] <Simon4> ganiman: have a look at some of the zenpacks at http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/

[08-Apr-2010 09:27:41] <rmatte> ganiman: also, read this: http://www.skills-1st.co.uk/papers/jane/zenpacks/zenpacks.pdf

[08-Apr-2010 09:27:53] <Simon4> lots of ideas in there for what's doable - look at the modeler/plugins directory for a zenpack to see how people custom model devices

[08-Apr-2010 09:28:10] <ganiman> excellent

[08-Apr-2010 09:28:24] <ganiman> and then I can sell my new zenpack for a million dollars?

[08-Apr-2010 09:28:32] <Simon4> once you get your head around it it's not too difficult

[08-Apr-2010 09:28:34] <rmatte> Theoretically, if you wanted to

[08-Apr-2010 09:28:38] <ganiman> ;)

[08-Apr-2010 09:28:53] <rmatte> though I'm sure the community would appreciate it if you released it GPL :P

[08-Apr-2010 09:29:14] <ganiman> I'm sure the place I work for would apprecaite it if I didn't sell their IP I create at work lol

[08-Apr-2010 09:29:22] <rmatte> hehe

[08-Apr-2010 09:29:35] <rmatte> I create lots of packs here at work and I always get approval to release it open source

[08-Apr-2010 09:29:40] <ganiman> I'm all for GPL, but it's not mine to GPL :/

[08-Apr-2010 09:30:07] <rmatte> ask your management... just make sure that you credit the company somehow in the name of the ZenPack and they'll probably be all for it

[08-Apr-2010 09:30:09] <rmatte> it's good publicity

[08-Apr-2010 09:34:08] <sciolist> so for zenoss daemons, logseverity option is a range from $what to $what

[08-Apr-2010 09:36:38] <sciolist> also... these 'levels' mean exactly $what in terms of  verbosity?

[08-Apr-2010 09:38:30] <rmatte> sciolist: seeing what I can dig up

[08-Apr-2010 09:39:27] <rmatte> did you try going to Settings -> Daemons and clicking on edit config next to a daemon?

[08-Apr-2010 09:39:37] <rmatte> it pretty much explains every config option, including that one

[08-Apr-2010 09:40:38] <rmatte> The default level is Info, meaning that informational messages are displayed... it can go all the way up to Debug, which shows all the debug info (intensive, but handy for troubleshooting)

[08-Apr-2010 09:40:51] <rmatte> There's also Critical, error, and warning, which are self explanatory

[08-Apr-2010 09:41:01] <sciolist> hmmm, no I just went to vview_config

[08-Apr-2010 09:41:06] <sciolist> s/vv/v/

[08-Apr-2010 09:41:09] <rmatte> yeh, edit config is more robust

[08-Apr-2010 09:41:11] <rmatte> ;)

[08-Apr-2010 09:41:13] <sciolist> ah

[08-Apr-2010 09:42:06] <sciolist> uh, nope: here's what I get from edit...

[08-Apr-2010 09:42:09] <sciolist> #

[08-Apr-2010 09:42:09] <sciolist> # Logging severity threshold, default: 20

[08-Apr-2010 09:42:09] <sciolist> # logseverity 10

[08-Apr-2010 09:42:34] <sciolist> my guess is lower == more verbose, but what's the range?

[08-Apr-2010 09:42:55] <rmatte> no no, you clicked the edit link under view config

[08-Apr-2010 09:43:02] <rmatte> you need to click the one right under the daemons tab

[08-Apr-2010 09:43:12] <sciolist> gaaaah!

[08-Apr-2010 09:43:28] <rmatte> viewconfig/edit allows you to edit the config file by hand... edit directly under the daemons tab is sort of a config wizard

[08-Apr-2010 09:44:25] <sciolist> hrm, that's not what I'm seeing... I get the same thing I posted earlier

[08-Apr-2010 09:44:29] <mrayzenoss> ganiman: sorry I missed the initial discussion, you could trigger a remodel on a restart of your device

[08-Apr-2010 09:44:32] <sciolist> from both links

[08-Apr-2010 09:45:07] <rmatte> mrayzenoss: that wouldn't do anything

[08-Apr-2010 09:45:13] <sciolist> this is in 2.4.5, btw

[08-Apr-2010 09:45:19] <rmatte> mrayzenoss: he's trying to graph performance data for wireless connections

[08-Apr-2010 09:45:25] <rmatte> as well as gather mac address info and such

[08-Apr-2010 09:45:38] <rmatte> the only way to currently do it is with a collector plugin and template

[08-Apr-2010 09:45:56] <rmatte> since Zenoss doesn't currently support actually walking an OID tree and dynamically gathering data for each OID

[08-Apr-2010 09:46:10] <mrayzenoss> yeah, a command data source could probably contain the logic for that

[08-Apr-2010 09:46:29] <rmatte> it could... the problem is that you can only have a fixed amount of datapoints on a graph

[08-Apr-2010 09:46:38] <rmatte> and he could have between 1 and 1000 clients connected or whatever

[08-Apr-2010 09:46:47] <mrayzenoss> true… unless the command in the data source fixed the RRD as well

[08-Apr-2010 09:46:57] <mrayzenoss> and called into Zenoss to add the data points

[08-Apr-2010 09:47:08] <mrayzenoss> quite hacky, but doable

[08-Apr-2010 09:47:12] <rmatte> it would have to fix the RRD and modify the graph for each device individually

[08-Apr-2010 09:47:27] <rmatte> meaning you'd have to have a ton of local copied of the template and have the script constantly modifying them

[08-Apr-2010 09:47:29] <rmatte> it's pretty sloppy

[08-Apr-2010 09:47:49] <rmatte> and intensive I would think

[08-Apr-2010 09:48:44] <rmatte> It is kind of a pain that that sort of thing isn't actually a feature in Zenoss... I've run across the limitation several times... in many cases it was a showstopper

[08-Apr-2010 09:49:12] <rmatte> in other cases, such as with my windows performance ZenPack, I had to creat 5 different templates for each number of CPUs

[08-Apr-2010 09:49:31] <rmatte> had the functionality been available, I could have just created 1 template and had it dynamically handle the graphing

[08-Apr-2010 09:50:17] <rmatte> I know the devs can't really do anything about it right now, but it should be on the radar in the near future methinks

[08-Apr-2010 09:50:37] <mistich> good morning have a python/zenoss issue I'm trying to print the output of a method to the browser window how can I do that

[08-Apr-2010 09:50:57] <rmatte> mistich, what method?

[08-Apr-2010 09:51:10] <mistich> return printed does it in a script but doesnt seem to work in a method I created

[08-Apr-2010 09:52:52] <rmatte> when you say return in to the screen, are you talking about doing a REST call?

[08-Apr-2010 09:53:33] <rmatte> like, calling on the function and having it print the output with no formatting?

[08-Apr-2010 09:55:01] <mistich> example a REST call returns ['this','is','a','test']

[08-Apr-2010 09:55:39] <mistich> I want the REST call to return each one on its own line

[08-Apr-2010 09:55:40] <mistich> this

[08-Apr-2010 09:55:42] <mistich> is

[08-Apr-2010 09:55:42] <mistich> a

[08-Apr-2010 09:55:44] <mistich> test

[08-Apr-2010 09:55:48] <rmatte> oh, hmmm

[08-Apr-2010 09:56:04] <rmatte> any particular reason why?

[08-Apr-2010 09:56:17] <rmatte> could you just have the script that's reading the values split them out?

[08-Apr-2010 09:56:47] <mistich> in the middle of upgrade old code one way new code other way

[08-Apr-2010 09:56:53] <rmatte> ah

[08-Apr-2010 09:57:14] <mistich> timesaver

[08-Apr-2010 09:57:22] <rmatte> well, if it's outputting like that, it's basically stored as a list, no?

[08-Apr-2010 09:58:16] <mistich> no currently it is a python script inside zope

[08-Apr-2010 09:58:24] <mistich> moved  it to a custom method

[08-Apr-2010 09:58:38] <rmatte> I'm saying the output format of ['this','is','a','test'] is a list

[08-Apr-2010 09:58:40] <mistich> a method cannot return a print s

[08-Apr-2010 09:58:45] <mistich> yes

[08-Apr-2010 09:59:05] <rmatte> what's the function that you have to use to actually return the data to the browser?

[08-Apr-2010 10:02:51] <mistich> one second trying something

[08-Apr-2010 10:03:14] <rmatte> k

[08-Apr-2010 10:34:11] <sciolist> hey, we're in the middle of an upgrade from 2.4.5 -> 2.5.2 and are seeing this error when upgrading the zenoss-core-zenpacks:

[08-Apr-2010 10:34:21] <sciolist> opt/zenoss/packs/ZenPacks.zenoss.LinuxMonitor-1.1.0-py2.4.egg...

[08-Apr-2010 10:34:21] <sciolist> INFO:zen.ZenPackCMD:Previous ZenPack exists with same name ZenPacks.zenoss.LinuxMonitor

[08-Apr-2010 10:34:21] <sciolist> ERROR: zenpack command failed. Reason: KeyError: 'titleOrId'

[08-Apr-2010 10:34:33] <sciolist> anyone seen this before?

[08-Apr-2010 10:38:04] <mrayzenoss> did you upgrade the zenpacks before zenoss?

[08-Apr-2010 10:38:34] <sciolist> negative

[08-Apr-2010 10:38:56] <sciolist> seeing this in thhe forums: message/42796#42796

[08-Apr-2010 10:39:34] <sciolist> was following the upgrade instructions from the support site

[08-Apr-2010 10:40:30] <sciolist> the forums suggest that 'zenmigrate --step=AddTitleOrIdToCatalogs' should do the trick

[08-Apr-2010 10:41:49] <sciolist> comments? caveats?

[08-Apr-2010 10:43:33] <sciolist> that also includes a 'rpm -Uvh zenoss-core-zenpacks... --force --nodeps' at the end

[08-Apr-2010 10:45:14] RoAk is now known as RoAkSoAx

[08-Apr-2010 10:45:38] <mrayzenoss> sciolist: post all your details on a forum post and I'll get someone to take a look

[08-Apr-2010 10:46:28] <sciolist> okay, but I'm in the middle of a downtime now for a production box... going to have to do something, even if it's roll back

[08-Apr-2010 10:47:18] <mrayzenoss> yeah, I'll try to get someone to respond quick, I just don't know myself

[08-Apr-2010 10:47:58] <sciolist> no prob... thanks for the info

[08-Apr-2010 11:05:52] <mistich> rmatte I got it had to make one long string and return the string

[08-Apr-2010 11:06:52] <mistich> another question self  returns the functions location what is the reserved word for returning the path you are at

[08-Apr-2010 11:08:00] <mrayzenoss> getDeviceUrl()?

[08-Apr-2010 11:08:12] <mrayzenoss> off the device?

[08-Apr-2010 11:08:44] <mistich> I want the path such /zport/dmd/Reports/Graph Reports but the method lives in /dmd

[08-Apr-2010 11:09:03] <theacolyte> I just got an update from a linked in group I'm in

[08-Apr-2010 11:09:09] <theacolyte> Zenoss Releases Service Assurance Monitoring Product for Private & Public Clouds based on Cisco UCS

[08-Apr-2010 11:09:12] <theacolyte> !!!

[08-Apr-2010 11:09:37] <mrayzenoss> theacolyte: http://www.zenoss.com/product/need-help/cisco-ucs-management

[08-Apr-2010 11:09:47] <mistich> self returns <DataRoot at dmd>

[08-Apr-2010 11:10:17] <theacolyte> excellent!

[08-Apr-2010 11:10:18] <mistich> I need  <DataRoot at dmd/Reports/Graph Reports>

[08-Apr-2010 11:12:37] <mistich> its not context

[08-Apr-2010 11:12:44] <mistich> or here

[08-Apr-2010 11:14:50] fOrsberg is now known as forsberg

[08-Apr-2010 11:18:15] <mistich> matt any suggetions?

[08-Apr-2010 11:18:59] <mrayzenoss> my expertise is limited, if you post on zenoss-dev I'll try to get someone to respond

[08-Apr-2010 11:19:15] <mrayzenoss> or you can just ping cluther and npmccallum and see if they're around

[08-Apr-2010 11:25:53] <npmccallum> mistich: getattr(self.Reports, "Graph Reports")

[08-Apr-2010 11:26:35] <mistich> thanks but thats not it chet is looking at it

[08-Apr-2010 11:44:57] <rmatte> mistich: ah

[08-Apr-2010 11:45:21] <rmatte> (in response to your previous issue, I was out to lunch)

[08-Apr-2010 11:46:48] <mistich> it looks like self.REQUEST['PATH_INFO'] will work to get you something like "/zport/dmd/Reports/...."

[08-Apr-2010 11:47:15] <mistich> chet found this I need to take "/zport/dmd/Reports/...." at find the object path now

[08-Apr-2010 11:47:58] <mistich> you dont a know of a function that you can input /dmd/Reports/Graph Reports and return the object?

[08-Apr-2010 11:57:02] <otakup0pe> so what can people tell me about accessing device performance information progrmatically via a web service

[08-Apr-2010 11:57:15] <otakup0pe> i know a little bit, for zproperties and templates

[08-Apr-2010 11:57:37] <mistich> found it getObjByPath('/zport/dmd/Devices/')

[08-Apr-2010 12:03:36] <otakup0pe> i guess time to read the source code again heh

[08-Apr-2010 12:12:04] <ptom> does anyone know if it's possible to link multiple devices to another device, much like a parent child relationship? Basically I have a wireless basestation which clients connect to. I would like to be able to link the children to the parent device. possible?

[08-Apr-2010 12:14:00] <ckrough> has anyone on used RemoveHoltData.py ?

[08-Apr-2010 12:15:22] <Michelle0> Upgrading to 2.52 and getting error updating additional collectors - ERROR zen.DistributedCollector global name 'excludeList' is not defined

[08-Apr-2010 12:16:01] <Michelle0> anybody seen this?

[08-Apr-2010 12:22:01] <rmatte> Upgrading from which version?

[08-Apr-2010 12:22:06] <mrayzenoss> Michelle0: are you an Enterprise customer?

[08-Apr-2010 12:22:13] <mrayzenoss> you might want to check with Support

[08-Apr-2010 12:22:27] <mrayzenoss> ckrough: where's that from?

[08-Apr-2010 12:23:01] <mrayzenoss> ptom: that sounds like how several of the virtualization devices work

[08-Apr-2010 12:23:45] <Michelle0> enterprise yes

[08-Apr-2010 12:23:52] <Michelle0> i am opening a support case

[08-Apr-2010 12:24:02] <Michelle0> just wondered in anyone had seen this

[08-Apr-2010 12:24:30] <ptom> mrayzenoss: can you point me in the direction of docs for this type of setup?

[08-Apr-2010 12:25:50] <mrayzenoss> do you want a network dependency or you're writing a zenpack to show the dependent devices on a separate tab?

[08-Apr-2010 12:27:09] <ptom> Not sure, I guess a zenpack would be the way to go, I just want to be able to show the relationship

[08-Apr-2010 12:27:15] <rmatte> Michelle0: I've done quite a few 2.5.2 upgrades as part of testing and I've seen quite a few errors but I've never seen that one

[08-Apr-2010 12:27:32] <rmatte> Michelle0: It's possible that it's somehow related to an enterprise ZenPack or something, support is your best bet

[08-Apr-2010 12:29:06] <rmatte> ptom: wouldn't the devices connected vary?

[08-Apr-2010 12:29:21] <ckrough> rmatte: docs/DOC-3408

[08-Apr-2010 12:29:27] <ckrough> Predictive Threshold zenpack

[08-Apr-2010 12:29:49] <mrayzenoss> ckrough: rocket wrote that

[08-Apr-2010 12:29:58] <rmatte> ckrough: how does that have anything to do with what ptom is asking?

[08-Apr-2010 12:29:58] <ckrough> k

[08-Apr-2010 12:30:06] <ckrough> rmatte: it doesnt

[08-Apr-2010 12:30:09] * rmatte is confused

[08-Apr-2010 12:30:20] <ckrough> rmatte: I was answering your question to me

[08-Apr-2010 12:30:51] <rmatte> which question?

[08-Apr-2010 12:30:52] <ckrough> rmatte: oh sorry, that was for mray

[08-Apr-2010 12:30:53] <ckrough> lol

[08-Apr-2010 12:30:59] <rmatte> ok lol

[08-Apr-2010 12:32:26] <otakup0pe> ohhhh i just had an epiphany or something and all of a sudden the ZenModel product makes sense :3

[08-Apr-2010 12:32:41] <otakup0pe> i only need to find one more web service - how to look up the actual url for templates for a given item

[08-Apr-2010 12:32:42] <ptom> rmatte: they are pretty much statically linked. It's not like an AP where random people connect.

[08-Apr-2010 12:32:45] <ckrough> mrayzenoss: counter values are stored with 6 decimal places, but then the script rebuilds the rrd without Holt it uses rrd update, which apparently only allows integers for counter DS's... which makes me wonder how the floats got in there in the first place. confused.

[08-Apr-2010 12:33:45] <rmatte> ptom: why not just group them in to groups then?

[08-Apr-2010 12:34:00] <rmatte> ptom: there is no built in method to directly link devices to other devices

[08-Apr-2010 12:34:39] <otakup0pe> there must be an easy way to link templates to models :(

[08-Apr-2010 12:34:46] <rmatte> If you create groups and put all of the devices along with the base station in to each group, it'll pretty much be the same

[08-Apr-2010 12:34:49] <otakup0pe> rather; graphs produced from templates

[08-Apr-2010 12:50:56] <otakup0pe> mmm i think i will need to add a custom method for this, or screen-scrape the actual  interface status page

[08-Apr-2010 12:52:12] <sciolist> so, where doesn the zenoss store the mysql passwd for the zenoss user?

[08-Apr-2010 12:55:15] <otakup0pe> oh god it worked zenoss you guys are the best

[08-Apr-2010 12:58:42] <rmatte> sciolist: you mean the password to the events database?

[08-Apr-2010 12:58:45] <otakup0pe> http://REDACTED/zport/dmd/Devices/Network/devices/REDACTED/os/interfaces/A1/getGraphDefUrl?graph=Throughput if anyone cares heh

[08-Apr-2010 12:58:50] <sciolist> rmatte:  yeah

[08-Apr-2010 12:59:06] <rmatte> sciolist: it's stored in mysql itself, just like any other mysql permissions

[08-Apr-2010 12:59:18] <rmatte> sciolist: it should have prompted you to enter the mysql password when you installed

[08-Apr-2010 12:59:29] <rmatte> sciolist: and if not, the password is probably "zenoss"

[08-Apr-2010 13:00:12] <sciolist> were screwed about six ways from Sunday, right now... had to back out of the upgrade and now the restore isn't  getting us back to ground zero

[08-Apr-2010 13:00:16] <rmatte> If you want to actually set the password in Zenoss, just click on Event Manager in the left hand meny

[08-Apr-2010 13:00:17] <rmatte> menu*

[08-Apr-2010 13:00:35] <sciolist> will try

[08-Apr-2010 13:00:39] <rmatte> so you don't remember your MySQL password?

[08-Apr-2010 13:00:55] <rmatte> do you at least know the root mysql password?

[08-Apr-2010 13:02:20] <Simon4> sciolist: if you want to see what zenoss thinks it should be

[08-Apr-2010 13:02:45] <Simon4> echo "dmd.ZenEventManager.password" | zendmd

[08-Apr-2010 13:02:49] <Simon4> on the commandline as zenoss user

[08-Apr-2010 13:03:37] <sciolist> yeah, I've got it back. now...

[08-Apr-2010 13:03:47] <sciolist> thanks for the help

[08-Apr-2010 13:04:22] <rmatte> np

[08-Apr-2010 13:06:54] <sciolist> still having problems starting stopping the daemons

[08-Apr-2010 13:07:04] <rmatte> Simon4: I love how it's that easy to extract sensored passwords from Zenoss... can do the same thing with zWinPassword or any other

[08-Apr-2010 13:07:16] <sciolist> nice

[08-Apr-2010 13:07:17] <rmatte> sciolist: seeing any errors?

[08-Apr-2010 13:07:18] <Simon4> rmatte: yeah - both useful and scary :)

[08-Apr-2010 13:07:35] <sciolist> rmatte:  yeah.. is there a pastiebot?

[08-Apr-2010 13:07:35] <rmatte> Simon4: the really scary part is that they are stored plaintext in a flat file

[08-Apr-2010 13:07:38] <rmatte> gotta love zope

[08-Apr-2010 13:07:46] <rmatte> sciolist: fpaste.org

[08-Apr-2010 13:07:50] <sciolist> thx

[08-Apr-2010 13:09:24] <sciolist> http://fpaste.org/Bvw2/

[08-Apr-2010 13:10:06] <sciolist> the thing that most concerns me is the :

[08-Apr-2010 13:10:10] <sciolist> "No handlers could be found for logger "root""

[08-Apr-2010 13:10:24] <rmatte> that's not as big a deal as it sounds

[08-Apr-2010 13:10:52] <rmatte> the fact that it's happening for every daemon is a bit odd though

[08-Apr-2010 13:11:01] <sciolist> ya

[08-Apr-2010 13:11:13] <rmatte> when you're executing sudo /sbin/service zenoss start...

[08-Apr-2010 13:11:18] <rmatte> what user are you doing that as?

[08-Apr-2010 13:12:23] <rmatte> I need some background on your issue as well...

[08-Apr-2010 13:12:27] <rmatte> what type of install is this?

[08-Apr-2010 13:12:34] <rmatte> appliance, rpm, stack, deb?

[08-Apr-2010 13:12:38] <sciolist> PRM on RHEL5

[08-Apr-2010 13:12:43] <rmatte> k

[08-Apr-2010 13:12:59] <rmatte> and these problems arose after an upgrade?

[08-Apr-2010 13:13:04] <sciolist> good read through my dyslxic typing

[08-Apr-2010 13:13:17] <sciolist> yeah: trying to roll back using this process:

[08-Apr-2010 13:13:45] <rmatte> It seems to be complaining most about ZenPacks.zenoss.EnterpriseCollector-1.0.0-py2.4.egg

[08-Apr-2010 13:13:55] <rmatte> do you know if there's an updated version of that pack?

[08-Apr-2010 13:14:07] <sciolist> 1) rpm -e upgraded RPMs, 2) install older version (2.4.5), 3) restore from backup

[08-Apr-2010 13:14:20] <sciolist> don't know...

[08-Apr-2010 13:14:27] <rmatte> I've never tried a zenoss downgrade

[08-Apr-2010 13:14:27] <sciolist> let me check the RPM manifest

[08-Apr-2010 13:14:35] <rmatte> what was the error after the upgrade?

[08-Apr-2010 13:14:35] <sciolist> this would be my first

[08-Apr-2010 13:14:56] <sciolist> several errors, culminating with an inibility to upgrade out distrib sollectors

[08-Apr-2010 13:15:03] <sciolist> err, collectors

[08-Apr-2010 13:15:14] <cluther> sciolist: It would appear that you have 2.5 ZenPacks installed on a 2.4 system.

[08-Apr-2010 13:15:32] <rmatte> cluther: ah, I thought it might have been the other way around

[08-Apr-2010 13:15:40] <rmatte> but that's pretty much what I figured based on the import error

[08-Apr-2010 13:15:48] <rmatte> it's trying to import something from the new framework

[08-Apr-2010 13:15:51] <rmatte> but can't find it obviously

[08-Apr-2010 13:15:58] <sciolist> hmmm, then how did the 'rpm -e' work, then?

[08-Apr-2010 13:16:15] <rmatte> ZenPacks have nothing to do with the rpm command

[08-Apr-2010 13:16:28] <rmatte> it's all just files to it

[08-Apr-2010 13:16:44] <mrayzenoss> rmatte: if you're using the core zenpacks RPM it kinda does

[08-Apr-2010 13:17:00] <rmatte> but once you fire up Zenoss and it tries to import stuff needed for the packs, but finds that it doesn't exist, it errors out

[08-Apr-2010 13:17:04] <sciolist> right, but the Enterprise ZenPacks install using the RPM process... it should have been removed with the same process

[08-Apr-2010 13:17:17] <rmatte> mrayzenoss: oh I see

[08-Apr-2010 13:17:21] <rmatte> sciolist: ah

[08-Apr-2010 13:18:02] <rmatte> mrayzenoss: would the RPM be intelligent enough to know that it won't be compatible with what's installed?

[08-Apr-2010 13:18:09] <sciolist> the pre- and post- script that initialize things in zenoss are run from the install/upgrade/erase switches

[08-Apr-2010 13:19:13] forsberg is now known as fOrsberg

[08-Apr-2010 13:20:06] <rmatte> sciolist: can't you get support to assist you?

[08-Apr-2010 13:20:28] <rmatte> sciolist: I've never used enterprise, so I'd be going at this relatively blind, since the issues seem to center around the enterprise zenpacks

[08-Apr-2010 13:20:36] <sciolist> got tickets in.... guess I need to contact them directly though.

[08-Apr-2010 13:20:53] <sciolist> np, thanks for the help

[08-Apr-2010 13:20:57] <rmatte> np

[08-Apr-2010 13:27:35] <mrayzenoss> rmatte: the RPM dependencies should work

[08-Apr-2010 13:27:50] <mrayzenoss> the 2.5 zenpacks RPM depends on the 2.5 zenoss RPM

[08-Apr-2010 13:27:55] <mrayzenoss> unless —nodeps was used

[08-Apr-2010 13:28:11] <rmatte> hmmm

[08-Apr-2010 13:30:50] <sciolist> for an upgrade the "--nodeps" option is part of the documneted process

[08-Apr-2010 13:31:46] <mrayzenoss> yeah, I meant that —nodeps would override the dependency of zenoss 2.5 being installed

[08-Apr-2010 13:31:52] <sciolist> ah

[08-Apr-2010 13:32:05] <mrayzenoss> so you'd be able to inadvertently shoot yourself in the foot

[08-Apr-2010 13:32:14] <sciolist> exactly

[08-Apr-2010 13:35:02] <sciolist> thanks for all your help, guys

[08-Apr-2010 14:57:26] <blull> afternoon

[08-Apr-2010 14:57:44] <blull> i ran a discovery from 1 collector that couldn't resolve dns for the ip's

[08-Apr-2010 14:57:55] <blull> i then moved the entire subnet i ran a discovery on to a remote collector which has dns.

[08-Apr-2010 14:58:02] <blull> I enabled dns checks for the entire subnet

[08-Apr-2010 14:58:53] <blull> I changed the Title for one of the servers the discovery found via ping to its hostname.. now i get alerts for that server saying "DNS Critical - expected 'ip addres here' but got 'hostname i set under Title'

[08-Apr-2010 14:58:59] <blull> title was empty before I set it to anything.

[08-Apr-2010 14:59:07] <blull> how do I associate a hostname with an IP?

[08-Apr-2010 14:59:28] <gwb235> grrr

[08-Apr-2010 15:13:49] <blull> hmmm

[08-Apr-2010 15:13:51] <blull> okay

[08-Apr-2010 15:13:54] <blull> im a little confused about something

[08-Apr-2010 15:14:03] <blull> would appreciate it if someone could give me some quick feedback

[08-Apr-2010 15:14:17] <blull> real easy question, how do I run a discovery from a remote collector?

[08-Apr-2010 15:20:59] <chemist> mrayzenoss: am I imagining it or did I see a zenpack submitted by yourself to set zenoss up as remote collector?

[08-Apr-2010 15:21:31] <rmatte> chemist: I submitted that one

[08-Apr-2010 15:21:44] <rmatte> My last name is Matte ;)

[08-Apr-2010 15:21:54] <chemist> oh, sorry :)

[08-Apr-2010 15:21:56] <rmatte> people tend to confuse me with Matt Ray

[08-Apr-2010 15:21:58] <rmatte> lol

[08-Apr-2010 15:22:09] <chemist> can you remind me what it's called?

[08-Apr-2010 15:23:24] <chemist> collector tool?

[08-Apr-2010 15:23:37] <rmatte> yup

[08-Apr-2010 15:24:09] <chemist> so, I have 2 independent zenoss installs, can I use this to 'merge' them?

[08-Apr-2010 15:25:39] <rmatte> well...

[08-Apr-2010 15:26:26] <chemist> one install is a lot more complex than the other, I would leave that as master

[08-Apr-2010 15:26:27] <rmatte> The idea is that you would have all of the devices on 1 install, then you would do a clean install on another server and install the same ZenPacks, then you would use the script to make that one in to a collector, and move some of the devices off the main install to that collector

[08-Apr-2010 15:26:40] <rmatte> a collector only runs the base collection daemons, it doesn't run zope

[08-Apr-2010 15:27:07] <chemist> ok, I don't mind redoing the small install (this would be the collector)

[08-Apr-2010 15:27:15] <rmatte> Basically the 1 install acting as the master would control absolutely everything

[08-Apr-2010 15:27:21] <rmatte> and the collector would just help with the actual load

[08-Apr-2010 15:27:28] <rmatte> so you could offload devices to the collector

[08-Apr-2010 15:28:10] <rmatte> yeh, you'd have to remove the devices from the small install, set it up as a collector, then add them back on the master but specify the other server as the collector when adding the devices

[08-Apr-2010 15:28:17] <chemist> ok, do I need to install your zenpack on both instances?

[08-Apr-2010 15:28:27] <rmatte> yes

[08-Apr-2010 15:28:41] <rmatte> you need to run the script on the master and select option 1

[08-Apr-2010 15:28:47] <rmatte> then run it on the collector and select option 2

[08-Apr-2010 15:28:51] <rmatte> and follow the instructions

[08-Apr-2010 15:29:04] <chemist> ok, will be trying it on saturday morning:)

[08-Apr-2010 15:29:07] <rmatte> although... egor made a much better collector pack based on mine

[08-Apr-2010 15:29:19] <rmatte> we're working on testing it right now and it should be released within the next couple of weeks

[08-Apr-2010 15:29:33] <rmatte> so if you wait for that one, it takes a lot of the manual steps out and does them automatically

[08-Apr-2010 15:29:44] <chemist> ah, ok

[08-Apr-2010 15:29:49] <rmatte> it's up to you

[08-Apr-2010 15:30:19] <rmatte> some people have had problems with the script is my pack freezing half-way through (I have no idea how it's freezing since I've looked over the code several times and can't find any reason)

[08-Apr-2010 15:30:27] <rmatte> the script IN my pack, rather

[08-Apr-2010 15:30:55] <chemist> I will give it a go I think

[08-Apr-2010 15:31:08] <rmatte> You can attempt it manually as well

[08-Apr-2010 15:31:09] <rmatte> docs/DOC-2496

[08-Apr-2010 15:31:13] <rmatte> with that guide

[08-Apr-2010 15:31:39] <chemist> ok, thanks

[08-Apr-2010 15:31:45] <rmatte> np

[08-Apr-2010 15:31:50] <mrayzenoss> yeah, bigegor told me he's still working on it, I'll try to find some time to check in his current version in the next day or so

[08-Apr-2010 15:32:04] <rmatte> yeh, I need to find time to actually test it

[08-Apr-2010 15:32:08] <kobalt> its not too bad to setup I have it setup here and am about to add a 2nd collector

[08-Apr-2010 15:32:16] <rmatte> I have too much other stuff on the go right now which takes priority

[08-Apr-2010 15:35:58] <rmatte> sweet, my reverse dns finally updated on my new server

[08-Apr-2010 15:37:23] <mrayzenoss> all my years of linux nerdom, I'm finally installing DD-WRT

[08-Apr-2010 15:37:29] <rmatte> hehe

[08-Apr-2010 15:37:36] <rmatte> I have that on my linksys router

[08-Apr-2010 15:37:47] <rmatte> beats the heck out of the default linksys fw

[08-Apr-2010 15:38:41] <mrayzenoss> doing a stack of them for use at the Zenoss Community Day

[08-Apr-2010 15:38:54] <rmatte> cool

[08-Apr-2010 15:39:25] <rmatte> using them for something to actually monitor or just for general connectivity?

[08-Apr-2010 15:40:57] <mrayzenoss> connectivity

[08-Apr-2010 15:41:05] <mrayzenoss> but when I'm done I'll have one to play with

[08-Apr-2010 15:41:18] <mrayzenoss> so someday I may finally write some of my own zenpacks

[08-Apr-2010 15:42:45] <rmatte> cool

[08-Apr-2010 15:45:09] <blull> no ideas how to assign an entire network range to a specific collector?

[08-Apr-2010 15:45:17] <blull> it appears you can only do it on a per device basis?

[08-Apr-2010 15:45:29] <blull> i don't see anything in zendmd that'll let me do it either.

[08-Apr-2010 15:45:37] <blull> atleast not under the dmd.* stuff

[08-Apr-2010 15:45:49] <rmatte> blull: if you add them to the main collector you can mass move them to the remote collector

[08-Apr-2010 15:45:55] <rmatte> that's the only way to do it from the UI

[08-Apr-2010 15:46:11] <blull> not even worried about doing it with the ui

[08-Apr-2010 15:46:13] <blull> how about zendmd?

[08-Apr-2010 15:46:21] <blull> i just don't want to have to run scripts to add a subnet

[08-Apr-2010 15:46:34] <blull> besides that, i'd have add 1 at a time and do the ping sweep externally

[08-Apr-2010 15:46:50] <rmatte> ah, think I might have found the way to do it

[08-Apr-2010 15:47:13] <blull> ?

[08-Apr-2010 15:47:20] <blull> really?

[08-Apr-2010 15:47:21] <rmatte> zendisc help

[08-Apr-2010 15:47:22] <blull> that would so rock

[08-Apr-2010 15:47:24] <blull> yeah

[08-Apr-2010 15:47:27] <blull> I kicked zendisc off manually

[08-Apr-2010 15:47:30] <blull> gave it a -net

[08-Apr-2010 15:47:44] <blull> actually ill do this

[08-Apr-2010 15:47:49] <blull> let me remove the entire subnet I did from the UI

[08-Apr-2010 15:47:57] <blull> ill re zendisc it via cli on the remote collector

[08-Apr-2010 15:48:11] <blull> and hopefully all the hosts it find are auto-assigned to the remote collector and not the main zenhub

[08-Apr-2010 15:48:12] <rmatte> --monitor is what you want

[08-Apr-2010 15:48:23] <blull> yeah, I always specify the remote collector as the monitor

[08-Apr-2010 15:48:27] <blull> when i ran the zendisc

[08-Apr-2010 15:48:33] <rmatte> then that should work

[08-Apr-2010 15:48:40] <blull> k

[08-Apr-2010 15:48:45] <blull> must go get my number for my raise :)

[08-Apr-2010 15:48:48] <blull> then ill try it in a sec.

[08-Apr-2010 15:48:54] <rmatte> lol

[08-Apr-2010 15:49:01] <blull> yay for being moved into your first 6 figure base salary position :)

[08-Apr-2010 15:49:23] <blull> i've made 116k in 1 year with a base salary of 65k while working in a NOC, but now my base is moving into the 6 figure range so I'me xcited++ :)

[08-Apr-2010 15:49:26] <blull> okay enough bragging lol

[08-Apr-2010 15:49:28] <blull> oh

[08-Apr-2010 15:49:32] <blull> and i have a 7th grade edumacation =p

[08-Apr-2010 15:49:35] <blull> okay now I brb =p

[08-Apr-2010 15:51:39] <blull> wow that was painless lol

[08-Apr-2010 15:51:40] <blull> and quick

[08-Apr-2010 15:51:47] <blull> yay for extra bonus money + a raise hehe

[08-Apr-2010 15:53:23] <blull> okay here we go..

[08-Apr-2010 16:00:03] <rmatte> lol

[08-Apr-2010 16:01:04] <blull> bin/zendisc run --hubhost=zenoss-hub --monitor=es-mon1 --now --net=172.21.89.0/24

[08-Apr-2010 16:01:05] <blull> okay

[08-Apr-2010 16:01:10] <blull> thats what I'm running as zenoss user.

[08-Apr-2010 16:01:12] <blull> that lookr ight to you?

[08-Apr-2010 16:01:28] <blull> and im running it on the remote collector...?

[08-Apr-2010 16:01:52] <blull> looks right to me, so im running it =p

[08-Apr-2010 16:03:16] <blull> hmm, maybe i have to run it on the zenhub

[08-Apr-2010 16:03:30] <blull> cause its not upping the event log on the zenhub with what its finding.

[08-Apr-2010 16:10:25] <rmatte> you don't run anything on the remote collector

[08-Apr-2010 16:10:31] <rmatte> you let the master do the work

[08-Apr-2010 16:10:46] <rmatte> as long as the remote collector is configured properly and talking to the master the master will handle everything

[08-Apr-2010 16:11:10] <rmatte> once the remote collector is setup you don't even touch that server

[08-Apr-2010 17:01:18] mrayzenoss1 is now known as mrayzenoss

[08-Apr-2010 20:09:40] <sciolist> anyone know what zenpack might provide /opt/zenoss/libexec/check_ajp.rb ?

[09-Apr-2010 00:00:30] [disconnected at Fri Apr  9 00:00:30 2010]

[09-Apr-2010 00:00:31] [connected at Fri Apr  9 00:00:31 2010]

[09-Apr-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[09-Apr-2010 01:34:16] fOrsberg is now known as forsberg

[09-Apr-2010 04:20:35] <tr_> ?

[09-Apr-2010 04:21:17] <tr_> Hey, anyone here with knowledge in Zenpacks and removing them when they are broken? Zenoss 2.5.2

[09-Apr-2010 04:22:05] <forsberg> you are not able to remove it or ?

[09-Apr-2010 04:22:50] <tr_> No, when I try to remove it, I get this error: zenpack command failed. Reason: AttributeError: 'MySqlMonitorDataSource' object has no attribute '__of__'

[09-Apr-2010 04:23:45] <forsberg> yrk, i had that excact thing

[09-Apr-2010 04:24:02] <forsberg> update broke my zenoss and i had to reinstall and start all over

[09-Apr-2010 04:24:04] <forsberg> BUT!

[09-Apr-2010 04:24:04] <tr_> My problem was that I had an old version of MySqlMonitor (version 1.1.0) installed which I have removed, and then there was also a v. 2.1.1 installed which is broken, so I want to remove it and clean install

[09-Apr-2010 04:24:09] <forsberg> yeah

[09-Apr-2010 04:24:23] <forsberg> but you also updated zenoss at some point i guess?

[09-Apr-2010 04:25:00] <tr_> Yep, our zenoss install is actually an old pre 2.0 which has been upgraded over time to 2.5.2

[09-Apr-2010 04:25:46] <forsberg> yeah excactly

[09-Apr-2010 04:25:54] <forsberg> you are very lucky, cause i had that

[09-Apr-2010 04:25:57] <forsberg> hold on, trying to find link

[09-Apr-2010 04:26:09] <forsberg> there came a fix shortly after

[09-Apr-2010 04:27:21] <forsberg> http://dev.zenoss.org/trac/ticket/5798

[09-Apr-2010 04:27:25] <forsberg> this is the closed dev ticket

[09-Apr-2010 04:27:29] <forsberg> i never tried it

[09-Apr-2010 04:27:41] <tr_> OK, sounds interesting.. I'd really like not to reinstall tne entire thing

[09-Apr-2010 04:28:11] <forsberg> yeah me2, well our installation kinda could benefit from it, since it was our first original zenoss install, and it could use a good cleaning anyways

[09-Apr-2010 04:28:12] <forsberg> BUT

[09-Apr-2010 04:28:27] <forsberg> it has been proven to take way longer than expecting, cause i have other stuff to do etc so....

[09-Apr-2010 04:28:39] <forsberg> we still have like 150 devices in the "old" installation etc :(

[09-Apr-2010 04:29:12] <tr_> we have close to 300 devices in our install

[09-Apr-2010 04:29:46] <forsberg> jup

[09-Apr-2010 04:29:49] <forsberg> i had that too

[09-Apr-2010 04:29:57] <forsberg> but anyways, read the ilnk there

[09-Apr-2010 04:30:24] <tr_> Unfortunately I already have commented the line in question

[09-Apr-2010 04:30:55] <forsberg> hmm :()

[09-Apr-2010 04:31:15] <forsberg> well then it might be some of the other stuff, like reindex in zendmd

[09-Apr-2010 04:31:30] <tr_> Tried that as well :-/

[09-Apr-2010 04:32:01] <forsberg> yrk, well then it sounds even more nasty, i recommend trying again a little later.

[09-Apr-2010 04:32:11] <forsberg> the most activity in here is in US timezone office hours

[09-Apr-2010 04:32:26] <tr_> Yeah, maybe I'll give it a try

[09-Apr-2010 04:32:39] <tr_> thanks for your help anyways! Gotta go for a while

[09-Apr-2010 05:35:10] <penthium2> I have a stupid question : What the diferent betwen "Templates" and "Collector Plugins" ?

[09-Apr-2010 06:54:18] <forsberg> i have no idea, i use templates all the time, and have to my knowledge never used a collector plugin

[09-Apr-2010 06:54:36] <forsberg> not in any specific manner anyways

[09-Apr-2010 07:08:02] <fus10nx> are any of you going the WMI route vs. SMTP ?

[09-Apr-2010 07:22:40] <tr_> forsberg: was it you who i talked to about strange errors when trying to remove mysqlmonitor ?

[09-Apr-2010 07:38:19] <Weetos> hey there

[09-Apr-2010 07:40:24] <Weetos> How would I go about listing devices set to a given priority ?

[09-Apr-2010 08:17:14] <rbd_> hey guys... I had zenoss autodetect the hosts in my network. that worked fine but it found the asterisk service on several of my servers and totally spams whenever asterisk is restarted. I'm having a problem finding out how to turn monitoring for asterisk off (I can't find it under services)

[09-Apr-2010 08:24:08] <sciolist> rbd_:  you want to stop monitoring or stop alerting?

[09-Apr-2010 08:25:12] <rbd_> well I'd like to stop alterting, but there's so many events and a lot of them have the machine's hostname in it... I don't have the time to comb through them with regexps to get it all right. given that, just turning off asterisk monitoring on all servers is fine with me

[09-Apr-2010 08:27:22] <sciolist> you could set your asterisk systems to a production state other than "Production" and change your alerting rules appropriately

[09-Apr-2010 08:28:51] <sciolist> depending on what/how you are monitoring, you could also set your device's zProperties to ignore Ping/SSH/SNMP monitoring which would pretty much turn things off, IIRC

[09-Apr-2010 08:29:11] <rbd_> well all the other services I'd like to be monitored at a production level...they are production servers

[09-Apr-2010 08:29:27] <sciolist> but I am not an expert, by far

[09-Apr-2010 08:30:06] <sciolist> there are probably many other ways to acomplish this

[09-Apr-2010 08:30:27] <sciolist> afk for a bit

[09-Apr-2010 08:35:53] forsberg is now known as fOrsberg

[09-Apr-2010 09:02:05] <Guest79056> is there a way i can easily add tcp/22 ssh ip services monitoring to all hosts in a class?

[09-Apr-2010 09:02:12] <Guest79056> kind of a hindsight issue here ;)

[09-Apr-2010 09:02:50] Guest79056 is now known as davidX

[09-Apr-2010 09:03:21] davidX is now known as Guest67874

[09-Apr-2010 09:03:37] Guest67874 is now known as davidX-

[09-Apr-2010 09:11:04] <jb> hrmf, i wish I could surpress all events except "IP is down" for windows hosts..

[09-Apr-2010 09:11:15] <jb> i get WMI is down, services are down, bleh, bleh..

[09-Apr-2010 09:12:47] <malbon> hola folks

[09-Apr-2010 09:13:04] <mdereus> jb: You can ignore those events.

[09-Apr-2010 09:13:11] <malbon> Is Matt Ray around today?

[09-Apr-2010 09:13:36] <mdereus> He's not on currently.

[09-Apr-2010 09:14:36] <malbon> yeah indeed. I'll ping him an email.

[09-Apr-2010 09:19:20] <blull> hey

[09-Apr-2010 09:19:21] <blull> question

[09-Apr-2010 09:19:29] <blull> i have a layer 3 switch i want to monitor all interfaces on.

[09-Apr-2010 09:19:35] <blull> each interface has snmp listening on it

[09-Apr-2010 09:19:54] <blull> so when zenoss pings and snmp checks each interface, it is associating the ip into 1 device by hostname

[09-Apr-2010 09:20:11] <blull> thus when I modify comments on 1 device i find by ip, it updates all devices with that same hostname

[09-Apr-2010 09:20:11] <malbon> cisco switch?

[09-Apr-2010 09:20:14] <blull> yeah

[09-Apr-2010 09:20:21] <blull> 37xx stacked series iirc

[09-Apr-2010 09:20:26] <malbon> you only need 1 ip address to manage it.

[09-Apr-2010 09:20:37] <blull> right, but I'm monitoring default gateways for different vlan's

[09-Apr-2010 09:20:50] <blull> want to ensure those don't go unreachable

[09-Apr-2010 09:20:53] <malbon> it'll sort out all the interfaces and routes (or at least should)

[09-Apr-2010 09:21:00] <blull> its not :/

[09-Apr-2010 09:21:25] <blull> i have 172.21.86.5 and 172.21.89.5 both different vlans, both different interfaces, both on the same switch with the same snmp results being returned.

[09-Apr-2010 09:21:36] <blull> thus when I update 86.5, 89.5 changes as well :/

[09-Apr-2010 09:21:49] <malbon> ah, ok, then set up /Ping devices for each of those IP addresses, because it'll tie up the interfaces on the device into a single object.

[09-Apr-2010 09:22:00] <blull> ahh okay.

[09-Apr-2010 09:22:29] <blull> oh wait wtf

[09-Apr-2010 09:22:30] <blull> i c

[09-Apr-2010 09:22:39] <davidX-> is there a more extensive document than this one on using REST? docs/DOC-3805#devmgmt_programmatically_adding_rest

[09-Apr-2010 09:22:40] <malbon> those pings are only for availability tracking then, because your metrics and configuration is still owned by the switch 'device'

[09-Apr-2010 09:22:41] <blull> so i was finding the device by ip under the Device/IP Search in the upper right side of the UI

[09-Apr-2010 09:22:58] <blull> but when I actually go to where its device class is, its referred to by hostname

[09-Apr-2010 09:23:21] <blull> so my question now is, if 2 ip's are associated with the same device class device, will it ping both of them?

[09-Apr-2010 09:23:24] <malbon> yeah, that's correct for finding normal devices, but your requirement is special not handled by the standard case.

[09-Apr-2010 09:23:43] <malbon> nope, there is only 1 ip pinged which is the one listed on the device page.

[09-Apr-2010 09:24:19] <malbon> davidX-: try the developers guide, it has some stuff in it.

[09-Apr-2010 09:28:00] <malbon> blull: why do you want to ping the individual ip's and not trust the interface state flag?

[09-Apr-2010 09:31:47] <rmatte> eugh, one of those morning, I get in and everything is broken it would seem

[09-Apr-2010 09:31:48] <rmatte> :P

[09-Apr-2010 09:32:59] <blull> malbon, because there may be other causes where the interface becomes unavailable, but the router doesn't detect it down.

[09-Apr-2010 09:33:03] <blull> malbon, i want both :)

[09-Apr-2010 09:33:09] <blull> hrmphf

[09-Apr-2010 09:33:09] <blull> okay

[09-Apr-2010 09:33:17] <blull> even manually adding a device, and assigning it to a remote collector.

[09-Apr-2010 09:33:31] <blull> it does not discover that device from the remote collector :/

[09-Apr-2010 09:33:36] <blull> or modal or whatever its called in zen

[09-Apr-2010 09:35:36] fOrsberg is now known as forsberg

[09-Apr-2010 09:51:55] <blull> okay

[09-Apr-2010 09:51:58] <blull> one final question

[09-Apr-2010 09:52:06] <blull> how would I configure zenoss to do sub 1 minute checks?

[09-Apr-2010 09:52:19] <blull> ie. i have some extremely critical real time devices I need to monitor every 1 - 5 seconds for icmp.

[09-Apr-2010 09:52:51] <blull> right now its only doing it every 60 or so seconds, even though I modified the template its using to have a cycle of 5 seconds.

[09-Apr-2010 09:54:19] <blull> Yeah just verified, data source command template is set to use a "Cycle Time" of 5 (down from 300)

[09-Apr-2010 09:55:01] <sciolist> rmatte:  I'll trade places with you.... =)

[09-Apr-2010 09:55:48] <rmatte> lol

[09-Apr-2010 09:56:03] <rmatte> sure, I'll take your salary

[09-Apr-2010 09:56:05] <rmatte> :P

[09-Apr-2010 09:57:10] <blull> ?

[09-Apr-2010 09:58:16] <blull> ooo

[09-Apr-2010 09:58:16] <blull> wtf

[09-Apr-2010 09:58:18] <blull> its not inheriting?

[09-Apr-2010 09:58:30] <blull> if I copy a temp[late from /Device to /Device/Datacenters/Servers

[09-Apr-2010 09:58:33] <blull> or something liek that

[09-Apr-2010 09:58:37] <rmatte> cycle time of 5 is 5 seconds

[09-Apr-2010 09:58:39] <rmatte> that's way too much

[09-Apr-2010 09:58:51] <blull> its not to much when you need immediate notification of the device

[09-Apr-2010 09:59:02] <rmatte> If you copy a template it doesn't inherit from what you copied it from, they exist as 2 separate templates

[09-Apr-2010 09:59:07] <blull> hrmm

[09-Apr-2010 09:59:08] <rmatte> which obviously makes sense

[09-Apr-2010 09:59:21] <blull> how do I get everything under where I copied it to, to inherit it?

[09-Apr-2010 09:59:35] <blull> cause i got a ton of different organizers under this level.

[09-Apr-2010 09:59:35] <rmatte> oh, to use the template you mean?

[09-Apr-2010 09:59:38] <blull> yeah

[09-Apr-2010 09:59:40] <blull> the one I just copied

[09-Apr-2010 09:59:54] <rmatte> as long as it's bound to that device class all of the devices in that class should automatically inherit it

[09-Apr-2010 09:59:55] <blull> i need Devices/Datacenter/Server/* to use it ... including all sub organizers.

[09-Apr-2010 10:00:05] <blull> yes that level of class, but not sub classes

[09-Apr-2010 10:00:06] <rmatte> provided that they don't already have local copies of the template at the device level (which is unlikely)

[09-Apr-2010 10:00:17] <rmatte> you have to bind it to the sub classes as well

[09-Apr-2010 10:00:22] <rmatte> go in to each class and bind the template

[09-Apr-2010 10:00:46] <blull> or maybe it does

[09-Apr-2010 10:00:51] <blull> nm now its working

[09-Apr-2010 10:01:05] <blull> i was doing packet captures... it took awhile for th template to start kicking in it looks like

[09-Apr-2010 10:01:13] <blull> now all of a sudden everything is using it heh

[09-Apr-2010 10:04:05] <rmatte> ah

[09-Apr-2010 13:27:59] <kerick> Hello, is there a way I can see exactly what the Zenoss UI is doing during an operation ( list of python commands and such )?

[09-Apr-2010 13:38:07] <ebroad> is there any way to set the component name used by zencommand based on the response from the command being executed

[09-Apr-2010 13:48:51] <rmatte> ebroad: there's a way to set the summary, but I've never seen it done with the component

[09-Apr-2010 13:49:46] <ebroad> rmatte, thanks, what I am trying to do is a bit convoluted, it basically involves checking ssl certificates for host with multiple dns names/same ip

[09-Apr-2010 13:50:15] <ebroad> so i want to set the component per dns name

[09-Apr-2010 13:51:39] <rmatte> I see

[09-Apr-2010 13:51:51] <rmatte> not sure how you'd do that, I've never done it myself

[09-Apr-2010 13:52:30] <rmatte> you could include it as part of the summary and then use a transform to extract it from the summary and set it as the components

[09-Apr-2010 13:52:33] <rmatte> component*

[09-Apr-2010 13:53:03] <ebroad> thats not a bad idea

[09-Apr-2010 13:53:05] <ebroad> thanx

[09-Apr-2010 13:53:26] <ebroad> is it possible to add multiple devices with the same ip at all though?

[09-Apr-2010 14:42:52] * rmatte yawns

[09-Apr-2010 14:42:54] <rmatte> hmm

[09-Apr-2010 14:51:05] <felbane> Hey all, any zenpack experts available for a quick question?

[09-Apr-2010 14:53:42] <felbane> I'm wondering if it's possible to extend the OS tab on a device via zenpack. For example, make a skin that can be injected after "routes" that shows a table.

[09-Apr-2010 14:54:49] <bigegor> easy

[09-Apr-2010 14:55:06] <felbane> Oh?

[09-Apr-2010 14:58:03] <bigegor> have a look on Advanced Device Details ZenPack

[09-Apr-2010 15:00:43] <bigegor> and RDBMS Monitoring ZenPack as example howto use it.

[09-Apr-2010 15:01:08] <felbane> okay, will do. Thanks!

[09-Apr-2010 15:02:13] <bigegor> np

[09-Apr-2010 15:06:20] <nickanderson> Hellop

[09-Apr-2010 15:07:10] <nickanderson> Im trying to get event commands working, seems straightforward but it dosnt appear to be working. Where can I look for errors?

[09-Apr-2010 15:12:53] <nickanderson> Im trying to do a simple /bin/echo ${devname} ${evt/summary} >> /tmp/mycmd.txt

[09-Apr-2010 15:18:33] <felbane> Have you tried $ZENHOME/log/zencommand.log

[09-Apr-2010 15:20:06] <felbane> You may need to turn on debug logging first. Settings > Daemons > zencommand "edit config" > logseverity

[09-Apr-2010 15:24:01] <nickanderson> sanity check here, severity >=0 will basically trigger on any event right?

[09-Apr-2010 15:24:27] <nickanderson> and thanks felbane im tailing that log now after changing to debug and restarting zencommand

[09-Apr-2010 15:24:55] <felbane> It might be zenactions instead of zencommand, now that I think about it

[09-Apr-2010 15:25:01] <felbane> Try both I guess

[09-Apr-2010 15:25:18] <felbane> Yes sev >=0 is any event

[09-Apr-2010 15:26:43] <felbane> Yeah its zenactions

[09-Apr-2010 15:27:00] <felbane> Also those commands are asynchronous

[09-Apr-2010 15:27:13] <felbane> Zenactions runs once per minute by default

[09-Apr-2010 15:32:43] <nickanderson> I saw the email send out

[09-Apr-2010 15:33:02] <nickanderson> but never saw anything relating to my event command

[09-Apr-2010 15:36:14] <nickanderson> any ideas?

[09-Apr-2010 15:41:16] <nickanderson> oh I guess I was staring an error int he face

[09-Apr-2010 15:41:51] <nickanderson> https://gist.github.com/05955090b44ef4f8a790

[09-Apr-2010 15:43:40] <felbane> :-)

[09-Apr-2010 15:50:35] <nickanderson> anyone else around?

[09-Apr-2010 15:57:52] <rmatte> yes

[09-Apr-2010 15:58:49] <rmatte> never seen that error before

[09-Apr-2010 16:06:51] <rhettardo> hows it going

[09-Apr-2010 16:07:59] <felbane> Bigegor you still here?

[09-Apr-2010 16:08:09] <bigegor> yes

[09-Apr-2010 16:09:17] <felbane> Is it really just as simple as specifying a new deviceOsDetail.pt file in your zenpack skins folder?

[09-Apr-2010 16:09:27] <felbane> Re: my earlier question

[09-Apr-2010 16:09:42] <felbane> Or do you have to tell Zenoss to use it somehow

[09-Apr-2010 16:14:56] <bigegor> you must do two things: first - create one relation under 'os' and second - create template in skinf folder with the name deviceOsCnameofrelationDetail.pt

[09-Apr-2010 16:17:39] <felbane> Ok, but how does your zenpack replace the default skin? Via the relationship?

[09-Apr-2010 16:20:24] <bigegor> deviceAdvDetail ZenPack replaced standard zenoss templates deviceOsDetail.pt, deviceHardwareDetail.pt and deviceSoftwareDetail.pt

[09-Apr-2010 16:20:52] <felbane> Sorry, I should clarify what I'm trying to do. Our company has a "minimum required external zenpacks" policy, so the zenpack I'm writing needs to be able to stand on its own if at all possible.

[09-Apr-2010 16:21:06] <felbane> Its a stupid policy but I have to do what the boss says

[09-Apr-2010 16:23:09] <sloof3> What would cause ZenOSS to suddenly lose all of it's sampled data?

[09-Apr-2010 16:24:18] <bigegor> you can try to create your own Device class like in AIX SNMPZenPack

[09-Apr-2010 16:25:29] <felbane> So I'm trying to figure out how it is that its doing the replacement. I see how the sub-lists are added but I don't see how you're telling Zenoss to use "deviceOsDetail.pt" from the zenpack instead of "deviceOsDetail.pt" from /Products

[09-Apr-2010 16:25:56] <felbane> Unless it just figures that out automatically? *confused*

[09-Apr-2010 16:25:59] <mrayzenoss> sloof3: perf data?

[09-Apr-2010 16:26:16] <bigegor> simply put it in skins folder

[09-Apr-2010 16:26:52] <felbane> I see. I'll give it a shot :)

[09-Apr-2010 16:28:45] <bigegor> but be patient with this function. If you create another zenpack with modified deviceOsDetail, only one will be used.

[09-Apr-2010 16:30:18] <felbane> That was going to be my next question, actually. If you have multiple zenpacks with that skin defined, does it just pick at random which one it uses?

[09-Apr-2010 16:30:39] <felbane> or is it just the last zenpack that's loaded

[09-Apr-2010 16:31:29] <bigegor> i think last - wins, but new Device class with new templates is more secure way to replace standard templates.

[09-Apr-2010 16:41:06] <felbane> Ok, that will work for our needs. Thanks for the help egor

[09-Apr-2010 16:42:06] <bigegor> np

[09-Apr-2010 17:19:05] <blull> before i spend anymore time on this

[09-Apr-2010 17:19:13] <blull> how do you do basic file system monitoring via snmp?

[09-Apr-2010 17:19:17] <blull> against a linunx host

[09-Apr-2010 17:19:21] <blull> running snmpd :)

[10-Apr-2010 00:00:30] [disconnected at Sat Apr 10 00:00:30 2010]

[10-Apr-2010 00:00:31] [connected at Sat Apr 10 00:00:31 2010]

[10-Apr-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[10-Apr-2010 00:00:45] <bartol.freenode.net> [freenode-info] if you're at a conference and other people are having trouble connecting, please mention it to staff: http://freenode.net/faq.shtml#gettinghelp

[10-Apr-2010 02:39:47] rhettardo_ is now known as rhettardo

[10-Apr-2010 07:54:00] <chemist> hell

[10-Apr-2010 07:54:02] <chemist> o

[10-Apr-2010 07:54:30] <chemist> can somenone tell me on a master/remote setup do I need to install MIB's on both systems?

[10-Apr-2010 12:29:31] * Simon4 writes a script that generates 800kb of zendmd commands

[10-Apr-2010 12:29:36] <Simon4> this might be a bit out of control ;)

[10-Apr-2010 19:51:05] <Snipe> evening all

[10-Apr-2010 19:52:05] <Snipe> I have a question regarding Python in the automated email alerts

[10-Apr-2010 19:52:57] <Snipe> Is there a way to have the alert take the time of the initial event and compare it with the time the event cleared and come up with a duration?\

[10-Apr-2010 19:57:52] Snipe is now known as WDC

[10-Apr-2010 19:58:42] WDC is now known as New2Zenoss

[10-Apr-2010 20:29:40] forsberg is now known as fOrsberg

[11-Apr-2010 00:00:30] [disconnected at Sun Apr 11 00:00:30 2010]

[11-Apr-2010 00:00:30] [connected at Sun Apr 11 00:00:30 2010]

[11-Apr-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[11-Apr-2010 05:43:51] fOrsberg is now known as forsberg

[11-Apr-2010 16:49:09] <QbY> I need to define "hosts" for Zenoss, but they wont' be polled in the normal snmp function..  I also need to create events for them in a way similar to traps but not with traps..  Is this possible?

[11-Apr-2010 16:49:45] <Simon4> you can create hosts under a specific device class with the zproperties for snmp monitoring turned off

[11-Apr-2010 16:49:51] <Simon4> same with ping monitoring, etc

[11-Apr-2010 16:50:04] <Simon4> (mod the zproperties for the class and stick hosts into it)

[11-Apr-2010 16:50:11] <QbY> For background, we run a large Asterisk server internally..  the "Hosts" would be handsets (telephones), and when my script that is monitoring asterisk detects that one has disappeared, it would somehow tell Zenoss that phone 123456789012 has went unreachable..  and then zenoss would treat that as a normal event

[11-Apr-2010 16:50:12] <Simon4> raising events can be done programatically

[11-Apr-2010 16:50:24] <Simon4> ah

[11-Apr-2010 16:50:45] <QbY> is this possible/

[11-Apr-2010 16:50:56] <Simon4> so you can get zenoss to just ping test the phones

[11-Apr-2010 16:51:18] <QbY> no, zenoss would not be able to ping the phone--some are behind firewalls/nat

[11-Apr-2010 16:51:50] <Simon4> you could set a template on the phone class that runs a zencommand to check the phone's "upness"

[11-Apr-2010 16:51:53] <Simon4> which is then your script

[11-Apr-2010 16:52:04] <QbY> ok..

[11-Apr-2010 16:52:10] <Simon4> that way each phone is a device, and events will be raised on it as per a normal template

[11-Apr-2010 16:52:18] <Simon4> so set a threshold for "phone down" say in the template

[11-Apr-2010 16:52:26] <Simon4> and your script returns values based on that

[11-Apr-2010 16:52:30] <QbY> do you have a link or something that would show me how to accomplish such

[11-Apr-2010 16:52:42] <Simon4> lemme have a dig

[11-Apr-2010 16:52:52] <Simon4> it's a fairly common thing

[11-Apr-2010 16:52:59] <QbY> and i didn't know if i really wanted zenoss to constantly "ping" i was just going to throw an alert back

[11-Apr-2010 16:53:36] <Simon4> there may be a way to create an alert for device "$device" from a base status script, especially if hte device names march in your script and zenoss

[11-Apr-2010 16:53:42] <Simon4> just programatically

[11-Apr-2010 16:53:57] <Simon4> so zenoss isn't polling, it's just aware of that device for creating events against

[11-Apr-2010 16:54:28] <Simon4> message/40767

[11-Apr-2010 16:54:37] <Simon4> example to use XML-RPC to create an event against a device

[11-Apr-2010 16:54:49] <Simon4> and a few more in the "more like this" bit on the right

[11-Apr-2010 16:55:09] <QbY> i've created the device class "Handset" i don't see where to tell Zenoss not to model or query it for snmp

[11-Apr-2010 16:55:20] <Simon4> ah

[11-Apr-2010 16:55:22] <Simon4> that bit's easy

[11-Apr-2010 16:55:22] <Simon4> \

[11-Apr-2010 16:55:27] <Simon4> when in the page for the device class

[11-Apr-2010 16:55:32] <Simon4> click the "zProperties" tab

[11-Apr-2010 16:55:32] <Simon4> \

[11-Apr-2010 16:55:38] <Simon4> stupid UK keyboard :/

[11-Apr-2010 16:55:39] <QbY> i'm there

[11-Apr-2010 16:55:42] <Simon4> right

[11-Apr-2010 16:55:48] <Simon4> set zPingMonitorIgnore to true

[11-Apr-2010 16:55:54] <Simon4> and zSNMPMonitorIgnore to true

[11-Apr-2010 16:56:13] <Simon4> and also, click the link beside zCollectorPlugins

[11-Apr-2010 16:56:19] <Simon4> and remove all the plugins from the list

[11-Apr-2010 16:56:28] <Simon4> that'll stop it modelling in any useful way, and stop collection of snmp/ping

[11-Apr-2010 16:57:00] <QbY> ok..  so to create my own "plugin" for reaching out to the server and checking..  what would i do?

[11-Apr-2010 16:57:05] <QbY> or an example doc would be nice

[11-Apr-2010 16:57:07] <Simon4> you can do it this way

[11-Apr-2010 16:57:07] <Simon4> docs/DOC-3072

[11-Apr-2010 16:57:42] <Simon4> which tehn gets bound to the handset class, and runs with a command-line argument of the handset name for each handset

[11-Apr-2010 16:57:54] <Simon4> it would mean you could graph stats for the handset too if they're available

[11-Apr-2010 16:58:10] <QbY> when i create the device is it possible to specify which "host" it would be on?

[11-Apr-2010 16:58:51] <Simon4> as in the asterisk server?

[11-Apr-2010 16:59:00] <QbY> yes

[11-Apr-2010 16:59:17] <QbY> probably location path would be the right one

[11-Apr-2010 16:59:23] <QbY> something that would be fed into my script

[11-Apr-2010 16:59:41] <Simon4> yeah, you could put it into one of the fields (notes, something) which could then be passed to the script when you define it

[11-Apr-2010 16:59:53] <QbY> nice

[11-Apr-2010 16:59:54] <Simon4> I have to head to bed sorry, but hopefully that gives you something to go on

[11-Apr-2010 16:59:58] <QbY> it does

[11-Apr-2010 17:00:00] <QbY> thanks simpon

[11-Apr-2010 17:00:02] <QbY> simon

[11-Apr-2010 17:00:06] <Simon4> no probs :)

[11-Apr-2010 17:00:07] <Simon4> night

[11-Apr-2010 17:00:09] <QbY> nite

[11-Apr-2010 19:25:40] <davetoo> any devs here?

[11-Apr-2010 19:26:57] <davetoo> I'm trying to build a native windows development environment, wondering what to look for as far as zope version incompatibilities for the older zenoss releases.

[11-Apr-2010 19:27:14] <davetoo> For example, since my running zenoss 2.4.2 reported that it was running on zope 2.8,

[11-Apr-2010 19:27:51] <davetoo> I installed windows zope 2.8.8, but that came with python 2.3.5 which barfed on some of the 2.4 syntax in some of the libs that zenoss uses

[11-Apr-2010 20:12:02] <Apachez> you are doing it wrong

[11-Apr-2010 21:07:47] <davetoo> well that was quite helpful

[12-Apr-2010 00:00:30] [disconnected at Mon Apr 12 00:00:30 2010]

[12-Apr-2010 00:00:30] [connected at Mon Apr 12 00:00:30 2010]

[12-Apr-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[12-Apr-2010 01:26:11] <Troubadix09> morning all

[12-Apr-2010 07:13:50] <mistich> good morning

[12-Apr-2010 07:14:21] <mistich> is dev.zenoss.org track password the same as your zenoss.com password

[12-Apr-2010 07:56:28] <ckrough> davetoo: http://blog.nuclearbunny.org/2009/10/22/my-zenoss-development-environment-part-1/

[12-Apr-2010 08:13:42] <blull> Morning

[12-Apr-2010 08:13:44] <blull> anyone awake?

[12-Apr-2010 08:13:48] <ckrough> yup

[12-Apr-2010 09:15:14] <jb> oh, cgibbons? :)

[12-Apr-2010 09:15:16] <jb> where are you :)

[12-Apr-2010 09:34:20] <nickanderson> Is there a way I can print out all the TALES expressions I have access to?

[12-Apr-2010 09:40:51] <ckrough> nickanderson: thats a common request, I don't believe you can

[12-Apr-2010 09:41:07] <ckrough> nickanderson: there are a couple pages in the dev guide that list some common ones

[12-Apr-2010 09:41:55] <nickanderson> Do I have access to any from event commands? Ive tried ${devname} and ${dev/devname} with no success

[12-Apr-2010 09:43:34] <ckrough> not sure, are you looking for TALES expressions, or just access to methods from event commands?

[12-Apr-2010 09:44:18] <nickanderson> Well all I want to do is run a script and pass in the device name that triggered the event

[12-Apr-2010 09:44:38] <nickanderson> so either way as long as I can get to the device variable

[12-Apr-2010 09:44:52] <ckrough> try stuff like ${here/dev.id} and ${self/blahblah}

[12-Apr-2010 09:45:02] <nickanderson> ok thanks

[12-Apr-2010 09:45:07] <ckrough> sometimes those work. You may want to look through some zenpacks for examples

[12-Apr-2010 09:50:02] <penthium> hello

[12-Apr-2010 09:51:53] <penthium> how can I monitor the route table for a server ? ( exemple : I want a monitoring who alarm deletion of route under my server )

[12-Apr-2010 10:14:14] <ckrough> anyone have a good script for removing Holt Winters RRAs from an RRD? I am trying RemoveHoltData.py but the resulting RRD tool isn't right.

[12-Apr-2010 10:18:15] <blull> hey

[12-Apr-2010 10:18:20] <blull> i just created a template on my zenhub

[12-Apr-2010 10:18:41] <blull> it is targeting an ip range I discovered on a remote collector, and have the ip ranges specified in the zproperties of the collector.

[12-Apr-2010 10:18:52] <blull> the template i created is a SNMP collector

[12-Apr-2010 10:19:00] <blull> however it is only being run on the zenhub, not on the remote collector

[12-Apr-2010 10:19:37] <blull> any ideas how to get that to automatically run off the remote collector?

[12-Apr-2010 10:22:12] <blull> i pushed the configs to the collector via the UI

[12-Apr-2010 10:22:16] <blull> not sure what else I need to do

[12-Apr-2010 10:25:36] <blull> hey

[12-Apr-2010 10:25:38] <blull> wake up :)

[12-Apr-2010 10:25:46] <blull> this shit is poorly documented like woah

[12-Apr-2010 10:25:48] <blull> :)

[12-Apr-2010 10:26:50] <ckrough> sorry, can't help you on that one. Are you running Enterprise?

[12-Apr-2010 10:28:11] <blull> nope

[12-Apr-2010 10:28:13] <blull> core

[12-Apr-2010 10:28:20] <blull> message/31847#31847  <-- apparently this guy has the same issue

[12-Apr-2010 10:30:21] <blull> really, someone needs to make this stuff actually usable beyond a single monitored location heh.

[12-Apr-2010 10:30:29] <blull> and document it properly

[12-Apr-2010 10:30:34] <blull> sorry just frustrated :/

[12-Apr-2010 10:30:58] <blull> I can't justify the 50k it'll cost to buy the enterprise version to monitor 5 dc's all of which only have a few physical devices heh.

[12-Apr-2010 10:31:35] <rmatte> blull: remote collectors aren't even officially supported in core, that's why it's poorly documented

[12-Apr-2010 10:32:00] <rmatte> it's one of those "you can do it but you're relatively on your own" sort of things

[12-Apr-2010 10:32:36] <blull> i c

[12-Apr-2010 10:32:46] <blull> would be nice if it was mentioned and not included int he core docs :0

[12-Apr-2010 10:33:02] <rmatte> it's not included in the official core docs

[12-Apr-2010 10:33:06] <rmatte> it's only included on the wiki

[12-Apr-2010 10:33:15] <rmatte> which are community written docs

[12-Apr-2010 10:33:44] <rmatte> If there's mention of it in the admin guide it's referring to Enterprise

[12-Apr-2010 10:33:49] <blull> rmm, iirc community.zenoss.org (is for core right?) has a manual linked from there that has a section discussing the setup procedure for distributed which isn't even applicable to core

[12-Apr-2010 10:34:05] <blull> right

[12-Apr-2010 10:34:28] <blull> just saying enterprise only info shouldn't be included in core docs provided by zenoss

[12-Apr-2010 10:34:37] <blull> even if its "an example", its just plain confusing :)

[12-Apr-2010 10:34:50] <rmatte> they don't distinguish between core and enterprise docs, but the enterprise sections are labeled as such

[12-Apr-2010 10:35:11] <rmatte> the admin guide covers both, but when I read through it I found it quite clear what was applicable to core and what wasn't

[12-Apr-2010 10:35:16] <blull> sorry just a little frustrated that I've spent good portion of the last week and a half deploying this to run into problem after problem :/

[12-Apr-2010 10:35:30] <rmatte> anyways... getting back to your actual issue...

[12-Apr-2010 10:35:55] <rmatte> you're saying a performance template isn't working on the collector... what indication are you seeing of this?

[12-Apr-2010 10:35:58] <rmatte> are graphs not generating?

[12-Apr-2010 10:37:43] <blull> no, the template I just created which uses an SNMP data source is running on the central zenhub, not the remote collector, even though the devices is assigned to the remote collector.

[12-Apr-2010 10:38:15] <rmatte> ok, I think you're failing to grasp the concept of how a remote collector works...

[12-Apr-2010 10:38:31] <rmatte> a remote collector is simply barebones monitoring daemons, there is no zenhub, there is no interface

[12-Apr-2010 10:39:05] <rmatte> those daemons would normally link back to the zenhub on that host, but instead, you reconfigure them to connect to the zenhub running on the master zenoss server

[12-Apr-2010 10:39:48] <rmatte> the zenhub on the master does all of the work in terms of applying configs, parsing data, interracting with the zope database and mysql...

[12-Apr-2010 10:39:56] <rmatte> all the daemons on the remote collector do is the actual monitoring

[12-Apr-2010 10:41:06] <rmatte> so when you create a template on the master and apply it to those devices, a long as the device is configured under the other collector, the master knows to use the daemons running on that server to perform the monitoring

[12-Apr-2010 10:42:28] <rmatte> so yes, it is going to run on that zenhub instance since there shouldn't be a zenhub instance running on the remote collector at all

[12-Apr-2010 10:42:47] <rmatte> does that answer your question, or is there more to it than that?

[12-Apr-2010 10:44:54] <rmatte> hmmm, I think I see what you're getting at... you're saying that there are still times when the main instance attempts to communicate with the devices directly

[12-Apr-2010 10:45:25] <rmatte> I've heard in the past that the main server still attempts to ping and performs the modelling directly... collectors are really designed to share the monitoring load

[12-Apr-2010 10:45:42] <rmatte> all zenoss instances still need to actually have access to all of the devices

[12-Apr-2010 10:45:59] <ckrough> I think zenping is handled on the remote collector for all recent versions

[12-Apr-2010 10:46:11] <ckrough> we had to patch to move it back to the main server

[12-Apr-2010 10:46:14] <rmatte> ckrough: right, but what about modelling?

[12-Apr-2010 10:46:46] <rmatte> I would think that it would be ideal to have the remote collector handle the monitoring

[12-Apr-2010 10:46:56] <rmatte> since collectors are often used on different segments of a network

[12-Apr-2010 10:47:10] <rmatte> erm, handle the modelling

[12-Apr-2010 10:47:11] <rmatte> rather

[12-Apr-2010 10:47:45] <rmatte> blull: is there any reason why the main hub can't have access to see the same devices?

[12-Apr-2010 10:48:02] <blull> sorry work problem

[12-Apr-2010 10:48:14] <rmatte> blull: there may still be a small amount of traffic seen from the hub to those devices, but the bulk of the monitoring will be done by the remote collector provided that it's configured correctly

[12-Apr-2010 10:52:56] <ckrough> rmatte: seen rocket on lately? is he using another name?

[12-Apr-2010 10:53:21] <rmatte> ckrough: I saw him on early last week, same name

[12-Apr-2010 10:53:26] <rmatte> but he hasn't been around lately

[12-Apr-2010 10:54:41] <tehhobbit> anyone got a good guide to setup hp insight stuff + snmp on rhel ?

[12-Apr-2010 10:58:45] <blull> wow that sucked.

[12-Apr-2010 10:59:28] <davetoo> Apachez: thanks for the help last night

[12-Apr-2010 10:59:48] <davetoo> </sarcasm>

[12-Apr-2010 11:00:11] <ckrough> davetoo: 12:56:28 | ckrough > davetoo: http://blog.nuclearbunny.org/2009/10/22/my-zenoss-development-environment-part-1/

[12-Apr-2010 11:00:12] <rmatte> lol

[12-Apr-2010 11:00:21] <ckrough> not sure if you were here for the response

[12-Apr-2010 11:00:26] <davetoo> no,

[12-Apr-2010 11:00:33] <blull> okay

[12-Apr-2010 11:00:36] <blull> now that I read everything you said

[12-Apr-2010 11:00:37] <ckrough> thats for MacOS, but its a start

[12-Apr-2010 11:00:42] <davetoo> I drew heavily on those articles

[12-Apr-2010 11:00:53] <davetoo> Google had found them for me

[12-Apr-2010 11:00:53] <blull> I think I just may be miscommunicating what i'm trying to do

[12-Apr-2010 11:01:20] <ckrough> davetoo: I tried applying them to Fedora but got tired of build errors... gave up. VM for the win.

[12-Apr-2010 11:01:50] <davetoo> If I can get work to buy me a copy of VMWare workstation, I'll have a running env

[12-Apr-2010 11:02:05] <davetoo> but I did finally get all the modules to import, last night, natively in windows

[12-Apr-2010 11:02:13] <davetoo> I don't expect it to run, just load

[12-Apr-2010 11:02:20] <rmatte> VMWare workstation is absolutely horrid, use virtualbox, it's free and blows it out of the water

[12-Apr-2010 11:02:23] <ckrough> or check out Xen or KVM, or even VirtualBox is you are windows bound

[12-Apr-2010 11:02:27] <blull> so rmatte, from what you're saying, any task that required "data collection for monitoring purposes is run on the remote collector"  Am i correct in that statement?

[12-Apr-2010 11:02:29] <davetoo> ok

[12-Apr-2010 11:02:41] <blull> as long as that device is configured to do so.

[12-Apr-2010 11:02:47] <rmatte> blull: yes

[12-Apr-2010 11:02:52] <ckrough> hist

[12-Apr-2010 11:02:56] <ckrough> wrong window :)

[12-Apr-2010 11:03:16] <davetoo> I will say that I've learned a lot about pydev and zope in the past 24 hours :)

[12-Apr-2010 11:03:25] <rmatte> blull: If you check the performance stats of that collector you should see the graphs showing poll cycle times, datapoints, etc...

[12-Apr-2010 11:03:33] <blull> rmatte, okay.  So with that in mind, I created a new template, added a Data Source that is SNMP to a specific UID, then linked that data source, to a threshold in the template.

[12-Apr-2010 11:03:36] <rmatte> provided it's collecting

[12-Apr-2010 11:03:57] <rmatte> ok...

[12-Apr-2010 11:04:03] * rmatte is listening

[12-Apr-2010 11:04:04] <blull> I then selected to bind this template to a specific class holding 10 devices.

[12-Apr-2010 11:04:14] <rmatte> with you so far

[12-Apr-2010 11:04:18] <blull> I verified all the devices in this class are set to use the remote collector.

[12-Apr-2010 11:04:26] * rmatte nods

[12-Apr-2010 11:04:49] <blull> When I tcpdump on the remote collector, I see no snmp packets going out.  When I tcpdump on the zenhub, I see snmp going to the hosts that should be going out (according to what was stated earlier) the remote collector.

[12-Apr-2010 11:05:45] <blull> see why I'm considering that an issue?  its not the expected behavior.

[12-Apr-2010 11:05:53] <rmatte> k, when you added the devices initially, were they added directly to the remote collector?

[12-Apr-2010 11:05:58] <rmatte> or did you add them to the master first?

[12-Apr-2010 11:05:59] <blull> yes.

[12-Apr-2010 11:06:01] <blull> remote

[12-Apr-2010 11:06:03] <rmatte> k

[12-Apr-2010 11:06:05] <blull> via a zendisc

[12-Apr-2010 11:06:17] <rmatte> If you login to the remote server and go to $ZENHOME/perf/Devices...

[12-Apr-2010 11:06:29] <rmatte> do you see directories for each of those devices?

[12-Apr-2010 11:06:36] <rmatte> and do those directories contain RRD files?

[12-Apr-2010 11:06:53] <blull> yup.

[12-Apr-2010 11:07:00] <blull> even has my rrd for the "Current User Count" I added.

[12-Apr-2010 11:07:13] <blull> rhmm

[12-Apr-2010 11:07:15] <blull> although

[12-Apr-2010 11:07:17] <blull> it called it

[12-Apr-2010 11:07:21] <blull> Current User Count_5 Minute Load.rrd

[12-Apr-2010 11:07:24] <blull> ?

[12-Apr-2010 11:07:33] <blull> this was a brand new template, and I called it Current User count heh.

[12-Apr-2010 11:07:41] <blull> wonder if it mixed up some settings somewhere?

[12-Apr-2010 11:07:50] <rmatte> that's because the datasource name is Current User Count and the datapoint name under the datasource is called 5 Minute Load

[12-Apr-2010 11:08:08] <rmatte> datasources can have multiple datapoints

[12-Apr-2010 11:08:12] <blull> i just checked the datasource, its also called Current User Count

[12-Apr-2010 11:08:20] <rmatte> click on the datasource

[12-Apr-2010 11:08:20] <blull> oh

[12-Apr-2010 11:08:23] <blull> its under the threshold

[12-Apr-2010 11:08:23] <blull> i c

[12-Apr-2010 11:08:30] <rmatte> ah

[12-Apr-2010 11:08:32] <blull> but yeah

[12-Apr-2010 11:08:51] <rmatte> anyways, ok, so it's obviously doing the collection since it's generating the RRDs

[12-Apr-2010 11:09:07] <rmatte> when you check the device in Zenoss I assume you see the graphs?

[12-Apr-2010 11:09:08] <blull> well not neccessarily

[12-Apr-2010 11:09:12] <blull> it may have generated the rrd

[12-Apr-2010 11:09:14] <blull> but is not collecting data.

[12-Apr-2010 11:09:24] <blull> + I see no snmp traffic going to the target host at all.

[12-Apr-2010 11:09:30] <rmatte> that's possible... but if you check the actual graphs in Zenoss do you see data?

[12-Apr-2010 11:09:35] <blull> let me double check

[12-Apr-2010 11:10:05] <rmatte> I'm heading to lunch, be back in about 15

[12-Apr-2010 11:10:22] <rmatte> check if the same RRDs are present on the master

[12-Apr-2010 11:10:25] <rmatte> they shouldn't be

[12-Apr-2010 11:10:33] <blull> i see where the graph should be

[12-Apr-2010 11:10:34] <rmatte> so if they aren't, and you're seeing data, then I'd say it's working

[12-Apr-2010 11:10:36] <blull> but its not displaying a graph.

[12-Apr-2010 11:10:45] <blull> just has the title and arrows, no gif dat

[12-Apr-2010 11:10:46] <blull> a

[12-Apr-2010 11:10:51] <blull> k

[12-Apr-2010 11:10:53] <blull> ahh

[12-Apr-2010 11:10:54] <blull> okay

[12-Apr-2010 11:11:06] <blull> so if I see the anything under Perf that should be on remote on master, remove it?

[12-Apr-2010 11:17:53] <blull> hrmphf, don't see any hosts in perf/Devices for the hosts I'm testing on.

[12-Apr-2010 11:18:00] <blull> on the zenhub

[12-Apr-2010 11:18:04] <blull> so its only on the remote collector.

[12-Apr-2010 11:18:35] <blull> i got zenperfsnmp running on both the hub and the remote collector, remote collector is going to the zenhub and the monitor is set to the remote monitor

[12-Apr-2010 11:31:51] <rmatte> if the RRDs are on the remote collector then it's definitely collecting snmp data

[12-Apr-2010 11:33:09] <davetoo> if they have the right dates

[12-Apr-2010 11:37:10] <rmatte> davetoo: the graphs render with data, so dates are irrelevant

[12-Apr-2010 11:41:24] <rmatte> blull: when you're doing the tcpdump, are you sure that you're not just picking up on the communications between the daemons and zenhub on the master server?

[12-Apr-2010 11:43:48] <blull> hey rmatte

[12-Apr-2010 11:43:52] <blull> i got it working mostly

[12-Apr-2010 11:44:03] <rmatte> cool, what was it?

[12-Apr-2010 11:44:05] <blull> deleteing the template and re-adding with same exact settings made it start working on the remote collector

[12-Apr-2010 11:44:08] <blull> so not sure what root cause was.

[12-Apr-2010 11:44:11] <blull> no errors in logs, etc.

[12-Apr-2010 11:44:15] <rmatte> ah, strange

[12-Apr-2010 11:44:30] <blull> well only thing I need to veriy is, in a collector, should the render URL be the url of the zenhub server?

[12-Apr-2010 11:44:38] <blull> ie: its accessed via the remote collector

[12-Apr-2010 11:44:47] <rmatte> renderurl is only set on the master

[12-Apr-2010 11:44:49] <blull> or is it the remote collector (ie. accessed via zenhub to pull data)

[12-Apr-2010 11:45:00] <rmatte> you shouldn't even have the UI running on the collector

[12-Apr-2010 11:45:04] <blull> I don't

[12-Apr-2010 11:45:06] <blull> im saying from the master

[12-Apr-2010 11:45:14] <rmatte> it's the address of the remote collector

[12-Apr-2010 11:45:27] <rmatte> so that it knows where to pull the render data from to render graphs properly

[12-Apr-2010 11:45:37] <blull> that was my question, pull :)

[12-Apr-2010 11:45:41] <blull> i have my tunnels setup the wrong direction

[12-Apr-2010 11:45:44] <blull> i thought render was a push function

[12-Apr-2010 11:45:53] <rmatte> nope

[12-Apr-2010 11:45:54] <blull> from collector -> hub not vice versa :)

[12-Apr-2010 11:46:01] <blull> explains why I can't see graphs, *fixes stunnel*

[12-Apr-2010 11:46:05] <rmatte> lol

[12-Apr-2010 11:46:17] <blull> i have ssl encrypted + compressed connections between collectors and hubs :)

[12-Apr-2010 11:46:27] <rmatte> push would be overly intensive

[12-Apr-2010 11:46:39] <rmatte> cool

[12-Apr-2010 11:46:43] <blull> yes agree, but doesn't it push the statistical data it collects anyway?

[12-Apr-2010 11:47:03] <blull> or does it only pull when the zenhub requests it?

[12-Apr-2010 11:47:07] <rmatte> some info gets pushed, but not RRD data

[12-Apr-2010 11:47:28] <blull> not raw rrd data, but I mean like the latest results, for example the value of a snmp query in a template?

[12-Apr-2010 11:47:51] <rmatte> nope, that's stored on the collector and accessed by the hub when required

[12-Apr-2010 11:49:35] <rmatte> Zenoss has a function to query the latest value in an rrd file

[12-Apr-2010 11:49:44] <rmatte> getRRDValue()

[12-Apr-2010 11:50:27] <blull> hmm okay

[12-Apr-2010 11:50:34] <rmatte> there's also cacheRRDValue()

[12-Apr-2010 11:50:37] <blull> i c

[12-Apr-2010 11:50:52] <rmatte> which is quicket but less accurate

[12-Apr-2010 11:50:56] <rmatte> quicker*

[12-Apr-2010 11:51:12] <blull> Once I change the render url

[12-Apr-2010 11:51:17] <blull> i should immediately see the graphs correct?

[12-Apr-2010 11:51:36] <rmatte> yes

[12-Apr-2010 11:51:49] <rmatte> actually, after entering the render url you may have to restart the hub

[12-Apr-2010 11:52:12] <rmatte> (restart Zenoss services on the master)

[12-Apr-2010 11:52:48] <blull> k

[12-Apr-2010 11:52:57] <blull> do I need to specify a unit for the datapoint?

[12-Apr-2010 11:53:00] <blull> or can that be blank?

[12-Apr-2010 11:53:09] <rmatte> defaults are fine for datapointsd

[12-Apr-2010 11:53:15] <rmatte> datapoints*

[12-Apr-2010 11:53:18] <blull> and I added a graph point, that links to my data point

[12-Apr-2010 11:53:27] <blull> k

[12-Apr-2010 11:53:29] <rmatte> the only thing you might change from time to time is type

[12-Apr-2010 11:53:44] <rmatte> but gauge is fine in most situations

[12-Apr-2010 11:54:01] <blull> yeah, I'm pretty familiar with rrd generation

[12-Apr-2010 11:54:02] <rmatte> I've personally never used anything but gauge

[12-Apr-2010 11:54:05] <rmatte> k

[12-Apr-2010 11:54:24] <rmatte> oh, you were talking about unit for the graph?

[12-Apr-2010 11:54:25] <blull> alright test time *crosses fingers*

[12-Apr-2010 11:54:27] <blull> yeah

[12-Apr-2010 11:54:37] <rmatte> that just gets shown on the left of the graph

[12-Apr-2010 11:54:41] <rmatte> it's a label

[12-Apr-2010 11:54:54] <rmatte> like, a graph for memory, you'd put "bytes"

[12-Apr-2010 11:54:56] <rmatte> or whatever

[12-Apr-2010 11:55:02] <blull> oh

[12-Apr-2010 11:55:03] <blull> yeah

[12-Apr-2010 11:55:04] <rmatte> it can be left blank if you want

[12-Apr-2010 11:55:06] <blull> i just noticed that under graph commands

[12-Apr-2010 11:55:12] <blull> actually not sure if it'll work blank?

[12-Apr-2010 11:55:18] <rmatte> it will work fine blank

[12-Apr-2010 11:55:18] <blull> cause it shows the graph command as --vertical-label=

[12-Apr-2010 11:55:20] <blull> no argument to it

[12-Apr-2010 11:55:21] <blull> oh okay

[12-Apr-2010 11:55:45] <rmatte> I usually fill it in though since it makes the graph look a bit cleaner

[12-Apr-2010 11:55:54] <blull> Damn the man, no graph data :/

[12-Apr-2010 11:56:06] <rmatte> ok, now hold up...

[12-Apr-2010 11:56:18] <blull> I just had the network admin open a firewall rule directly to the remote collector via port 8090 from the zenhub.

[12-Apr-2010 11:56:25] <rmatte> one very common mistake is people messing up when specifying the options in the conf files on the collector

[12-Apr-2010 11:56:32] <rmatte> the zenhub and monitor options

[12-Apr-2010 11:56:34] <blull> right

[12-Apr-2010 11:56:40] <blull> i've actually double checked those like 100 times

[12-Apr-2010 11:56:44] <blull> but ill check again :)

[12-Apr-2010 11:56:45] <rmatte> for monitor, did you use localhost or 127.0.0.1?

[12-Apr-2010 11:56:54] <blull> for monitor... neither.

[12-Apr-2010 11:56:57] <blull> zenoss-hub

[12-Apr-2010 11:57:02] <blull> which is defined in /etc/hosts to 127.0.0.1

[12-Apr-2010 11:57:11] <rmatte> yeh, you can't do that

[12-Apr-2010 11:57:12] <blull> to get around mysql default to localhost access via a socket.

[12-Apr-2010 11:57:19] <blull> uhh, really?

[12-Apr-2010 11:57:21] <rmatte> it has to be the IP of the actual interface on the server

[12-Apr-2010 11:57:30] <rmatte> you can't use 127.0.0.1

[12-Apr-2010 11:57:34] <blull> but lo is an actual interface on the server :)

[12-Apr-2010 11:57:43] <rmatte> PHYSICAL interface

[12-Apr-2010 11:57:44] <rmatte> :P

[12-Apr-2010 11:57:48] <rmatte> lol

[12-Apr-2010 11:57:49] <blull> its not a physical server though =p

[12-Apr-2010 11:57:51] <blull> its a virtual machine.

[12-Apr-2010 11:57:55] <blull> it has no "physical" interface =p

[12-Apr-2010 11:57:59] <rmatte> let's not get technical :P

[12-Apr-2010 11:58:01] <blull> hehe

[12-Apr-2010 11:58:02] <rmatte> so are mine

[12-Apr-2010 11:58:05] <blull> but yeah i get ya

[12-Apr-2010 11:58:07] <blull> why is that anyhow?

[12-Apr-2010 11:58:22] <blull> in my zenhub, its defined as localhost for the collector.

[12-Apr-2010 11:58:26] <rmatte> I don't know, it's just something I figured out when I was working with remote collectors for the first time

[12-Apr-2010 11:58:31] <blull> oooo

[12-Apr-2010 11:58:32] <blull> wait

[12-Apr-2010 11:58:35] <blull> they have to match don't they?

[12-Apr-2010 11:58:36] <rmatte> none of the graphs would render, then I changed the IP and voila, they rendered

[12-Apr-2010 11:58:40] <rmatte> match?

[12-Apr-2010 11:58:42] <blull> yeah

[12-Apr-2010 11:58:42] <blull> like

[12-Apr-2010 11:58:52] <rmatte> the zenhub one has to be the IP address of the master zenoss server

[12-Apr-2010 11:59:05] <rmatte> the monitor address has to be the IP of the physical interface on the local collector server

[12-Apr-2010 11:59:10] <davetoo> So... has the community worked out a repeatable distributed collector Core Zenoss setup now?

[12-Apr-2010 11:59:38] <blull> davetoo, I have code someone sent to me, im going to hack it up once i get this intial stuff into production to make it work

[12-Apr-2010 11:59:40] <rmatte> davetoo: My ZenPack got released for it, but I sent it over to Egor, and he made an even better one that we're working on testing...

[12-Apr-2010 11:59:50] <blull> hehe

[12-Apr-2010 11:59:51] <davetoo> gotcha

[12-Apr-2010 11:59:51] <rmatte> davetoo: the one that Egor made is pretty close to the Enterprise one

[12-Apr-2010 12:00:03] <blull> hrmm

[12-Apr-2010 12:00:12] <blull> well

[12-Apr-2010 12:00:13] <davetoo> How's the new 2.5.x Event Console working for people?

[12-Apr-2010 12:00:27] <davetoo> My Enterprise system .. is so slow, sometimes the page never even finishes loading

[12-Apr-2010 12:00:34] <rmatte> davetoo: besides the fact that sometimes it doesn't load at all and I have to restart my browser to get it to load, not too badly

[12-Apr-2010 12:00:37] <davetoo> (because I can't get people to clear their events)

[12-Apr-2010 12:00:37] <blull> the zenhub I have set to localhost, which looks like the master server because its running over tunnels to the master server.

[12-Apr-2010 12:00:51] <davetoo> rmatte: so is it any better than the old one or no?

[12-Apr-2010 12:00:57] <blull> and my monitor address, that is set to the hostname of the local collector

[12-Apr-2010 12:00:57] <rmatte> davetoo: yeh, we auto-clear events after an hour, so I'll never run in to that issue

[12-Apr-2010 12:01:06] <blull> which in turn resolves to the real ip

[12-Apr-2010 12:01:28] <rmatte> blull: well hold up...

[12-Apr-2010 12:01:39] <rmatte> <blull> the zenhub I have set to localhost, which looks like the master server because its running over tunnels to the master server.

[12-Apr-2010 12:01:45] <rmatte> so what is the actual IP set for zenhub?

[12-Apr-2010 12:02:27] <blull> an ip on 10.13.37.0/24

[12-Apr-2010 12:02:36] <blull> remote collector is on 172.21.86.0/24

[12-Apr-2010 12:03:01] <rmatte> k that should hopefully be fine (I never tested over a tunnel)

[12-Apr-2010 12:03:10] <blull> oh

[12-Apr-2010 12:03:11] <rmatte> try again without monitor set to localhost and hopefully it'll work

[12-Apr-2010 12:03:12] <blull> doh

[12-Apr-2010 12:03:19] <rmatte> ?

[12-Apr-2010 12:03:27] <blull> i accidently had my remote collector resolving to 127.0.0.1 on the zenhub

[12-Apr-2010 12:03:32] <blull> *smacks head*

[12-Apr-2010 12:03:36] <blull> dumb hosts entries.

[12-Apr-2010 12:03:37] <rmatte> woah woah

[12-Apr-2010 12:03:38] <rmatte> hold the phone

[12-Apr-2010 12:03:45] <rmatte> don't edit any configs on the master at all

[12-Apr-2010 12:03:52] <blull> i didn't, I edited os configs ;)

[12-Apr-2010 12:03:56] <blull> not zenoss configs

[12-Apr-2010 12:04:00] <rmatte> k

[12-Apr-2010 12:04:20] <rmatte> oh, I see what you're saying

[12-Apr-2010 12:04:27] <rmatte> the master thought the remote server was actually itself

[12-Apr-2010 12:04:37] <rmatte> yeh, that would explain the tcpdump problem

[12-Apr-2010 12:05:01] <rmatte> and changing the monitor options should fix the graph render problems

[12-Apr-2010 12:05:11] <blull> hmm

[12-Apr-2010 12:05:16] <blull> that doesn't sound right though

[12-Apr-2010 12:05:23] <blull> so basically

[12-Apr-2010 12:05:34] <blull> you can't have a collector using the same ip on each side of an isolated network?

[12-Apr-2010 12:05:44] <blull> then routed say via nat to a zenhub?

[12-Apr-2010 12:05:47] <blull> heh

[12-Apr-2010 12:06:42] <rmatte> well, you can't have it using 127.0.0.1 as the monitor IP for whatever reason, that's all that I know

[12-Apr-2010 12:06:50] <rmatte> as far as the rest of it goes, no idea, trial and error

[12-Apr-2010 12:09:17] <rmatte> I've got a new line for the karate kid: swapoff -a && swapon -a

[12-Apr-2010 12:10:45] * rmatte can't wait to upgrade to 2.5.2, getting sick of clearing swap

[12-Apr-2010 12:13:50] <davetoo> how does 2.5.2 fix your swap problem?

[12-Apr-2010 12:14:12] <rmatte> 2.4.5 used an older version of RRDtool which had a bad memory leak in it

[12-Apr-2010 12:14:19] <rmatte> 2.5.2 uses a newer version

[12-Apr-2010 12:14:48] <rmatte> though apparently 2.5.2 has a leak in zencommand, but that won't be nearly as bad

[12-Apr-2010 12:15:43] <davetoo> I'm waiting to see what Zenoss can do with RRDtool 1.4 and it's cache daemon

[12-Apr-2010 12:16:06] <blull> hmmm

[12-Apr-2010 12:16:19] <blull> well okay then.

[12-Apr-2010 12:16:25] <blull> guess ilc hange it to not using 127.0.0.1 heh

[12-Apr-2010 12:16:27] <blull> hopefully this works :)

[12-Apr-2010 12:16:40] <rmatte> hopefully

[12-Apr-2010 12:22:05] <blull> when I connect to a render url from the zenhub

[12-Apr-2010 12:22:12] <blull> or rather when the zenhub connects to a render url

[12-Apr-2010 12:22:20] <blull> it hits port 8090 or 8091?

[12-Apr-2010 12:22:54] <rmatte> 8090

[12-Apr-2010 12:22:59] <blull> hrmm k

[12-Apr-2010 12:23:02] <rmatte> is the port you need in the renderurl

[12-Apr-2010 12:23:13] <blull> yeah

[12-Apr-2010 12:23:34] <blull> it shows my collector listening on 8091 though?

[12-Apr-2010 12:23:37] <blull> for zenrender.py

[12-Apr-2010 12:24:12] <rmatte> that port is probably used too but for something else

[12-Apr-2010 12:25:06] <blull> k cause its not listening at all on 8090.

[12-Apr-2010 12:25:24] <blull> and when I send a get request, it respond as the renderer =p

[12-Apr-2010 12:26:00] <blull> does my render url have to match the hostname of my remote collector?

[12-Apr-2010 12:26:04] <blull> ie: can I use it by ip

[12-Apr-2010 12:26:11] <blull> or does it have to be the same as monitor?

[12-Apr-2010 12:26:11] <rmatte> hmmm

[12-Apr-2010 12:26:18] <rmatte> 8090 is the standard port as far as I know

[12-Apr-2010 12:26:37] <rmatte> maybe you changed it in the zenrender.conf file?

[12-Apr-2010 12:28:00] <rmatte> as long as it knows to hit the remote collector it should work

[12-Apr-2010 12:28:09] <rmatte> doesn't matter if you use the hostname or IP as far as I'm aware

[12-Apr-2010 12:28:19] <rmatte> though I don't know why you'd even bother using hostnames if they aren't in dns

[12-Apr-2010 12:28:33] <rmatte> I guess in case you need to change the IP in the future

[12-Apr-2010 12:28:37] <rmatte> though sed works wonders

[12-Apr-2010 12:28:45] <blull> hrmm, its listening on 8090 on the master.

[12-Apr-2010 12:28:52] <blull> nah i checked zenrender.conf

[12-Apr-2010 12:28:54] <blull> no ports defined

[12-Apr-2010 12:29:24] <rmatte> oh wait

[12-Apr-2010 12:29:45] <rmatte> I just double checked what I did in my script...

[12-Apr-2010 12:29:48] <rmatte> echo "3. Edit the newly created collector and change the Render URL to: \"http://$master:8090/$collector\"."

[12-Apr-2010 12:29:54] <rmatte> looks like it is the master

[12-Apr-2010 12:29:57] <rmatte> my bad

[12-Apr-2010 12:30:11] <blull> ahh so it pushes to the master than?

[12-Apr-2010 12:30:14] <rmatte> I guess it connects to the local zenrender which then communicates with the remote one

[12-Apr-2010 12:30:20] <blull> k

[12-Apr-2010 12:31:19] <rmatte> it's http://master-host-or-ip:8090/collector-host-or-ip

[12-Apr-2010 12:32:01] <rmatte> you really shouldn't put a limit on pushing or pulling... since Zenoss does a little of both

[12-Apr-2010 12:34:28] <blull> well when you're in a segmented environment, it really does matter.

[12-Apr-2010 12:34:51] <rmatte> right, but just make sure you're not crippling Zenoss in some way because of it

[12-Apr-2010 12:39:12] <blull> well

[12-Apr-2010 12:39:13] <blull> i changed it

[12-Apr-2010 12:39:19] <blull> the remote collector has no reference to 127.0.0.1 or localhost

[12-Apr-2010 12:39:54] <blull> it now references the hubhost as zenoss-hub which is a nont 127.0.0.0/8 ip and uses es-mon1 as the monitor (as before)

[12-Apr-2010 12:49:24] <rmatte> what does es-mon1 resolve to, not 127.0.0.1 I hope?

[12-Apr-2010 12:53:00] <blull> no, 172 address.

[12-Apr-2010 12:59:27] <rmatte> k

[12-Apr-2010 12:59:37] <rmatte> should hopefully work then, as long as the tunnel doesn't make it go wonky

[12-Apr-2010 13:22:57] <blull> uhh

[12-Apr-2010 13:23:00] <blull> what runs on 8789?

[12-Apr-2010 13:23:00] <rmatte> someone found out what the bug was in my collector tool zenpack, fixing it now...

[12-Apr-2010 13:23:19] <rmatte> blull: netstat -anp | grep 8789

[12-Apr-2010 13:23:24] <blull> oh duh

[12-Apr-2010 13:23:26] <blull> heh brainf art :)

[12-Apr-2010 13:23:27] <rmatte> ;)

[12-Apr-2010 13:23:36] <blull> oh wait I can't

[12-Apr-2010 13:23:39] <blull> i run stunnel on that port.

[12-Apr-2010 13:23:45] <blull> oh

[12-Apr-2010 13:23:46] <blull> hah!

[12-Apr-2010 13:23:47] <blull> i don't

[12-Apr-2010 13:23:50] <blull> i run it on 18789

[12-Apr-2010 13:23:53] <blull> hmm

[12-Apr-2010 13:24:04] <blull> nothing is listening on 8789

[12-Apr-2010 13:24:12] <rmatte> hmmm, not overly helpful when I check it...

[12-Apr-2010 13:24:22] <rmatte> tcp        0      0 127.0.0.1:41743         127.0.0.1:8789          ESTABLISHED 11745/.python.bin

[12-Apr-2010 13:24:30] <rmatte> tcp        0      0 127.0.0.1:35108         127.0.0.1:8789          ESTABLISHED 11739/.python.bin

[12-Apr-2010 13:24:30] <rmatte> tcp        0      0 127.0.0.1:41744         127.0.0.1:8789          ESTABLISHED 11769/.python.bin

[12-Apr-2010 13:24:38] <rmatte> tcp        0      0 127.0.0.1:35105         127.0.0.1:8789          ESTABLISHED 11646/.python.bin

[12-Apr-2010 13:24:38] <rmatte> tcp        0      0 127.0.0.1:35104         127.0.0.1:8789          ESTABLISHED 11542/.python.bin

[12-Apr-2010 13:24:38] <rmatte> tcp        0      0 127.0.0.1:41747         127.0.0.1:8789          ESTABLISHED 11956/.python.bin

[12-Apr-2010 13:24:38] <rmatte> tcp        0      0 127.0.0.1:41748         127.0.0.1:8789          ESTABLISHED 11996/.python.bin

[12-Apr-2010 13:24:41] <rmatte> and a few more

[12-Apr-2010 13:24:49] <rmatte> doesn't show exactly what python script though

[12-Apr-2010 13:25:08] <blull> youc an get that with a ps -efww | grep <pid>

[12-Apr-2010 13:25:11] <blull> it'll show the python script

[12-Apr-2010 13:25:39] <rmatte> zenhub listens on 8789

[12-Apr-2010 13:25:48] <blull> hrmm

[12-Apr-2010 13:25:52] <blull> that would explain wtf is wrong lol

[12-Apr-2010 13:25:55] <rmatte> hehe

[12-Apr-2010 13:26:18] <rmatte> good afternoon Chet

[12-Apr-2010 13:26:40] <blull> actually

[12-Apr-2010 13:26:45] <blull> that doesn't explain it :/

[12-Apr-2010 13:30:37] <blull> god damn dude I can't get the graphs working

[12-Apr-2010 13:30:45] <blull> everything is connecting to each other correctly

[12-Apr-2010 13:30:53] <blull> i changed from 127.0.0.1 to a non loopback ip

[12-Apr-2010 13:30:57] <blull> on the remote collector

[12-Apr-2010 13:31:03] <blull> still no graphs :(

[12-Apr-2010 13:36:24] <ckrough> thats for the hub, how about zenrender?

[12-Apr-2010 13:36:31] <ckrough>  (serves the graphs)

[12-Apr-2010 13:44:50] <ckrough> dear zenoss, please str.strip() the inputs. kthybye

[12-Apr-2010 13:52:14] <rmatte> blull: docs/DOC-5831

[12-Apr-2010 13:52:21] <rmatte> I just fixed my collector tool ZenPack

[12-Apr-2010 13:52:29] <rmatte> so it'll actually work now, just in case you need to use it in the future

[12-Apr-2010 13:52:49] <rmatte> though Egor's should be ready for release in a couple of weeks hopefully

[12-Apr-2010 13:52:54] <rmatte> and it'll blow mine away lol

[12-Apr-2010 14:03:45] <rmatte> it's a good tool for making sure that you didn't miss any steps when configuring the collector

[12-Apr-2010 14:03:47] <rmatte> and the master

[12-Apr-2010 14:08:07] <Jane_Curry> Am I missing something blindingly obvious about monitoring swap psace on Windows?

[12-Apr-2010 14:08:32] <blull> rmatte, i just looked @ your collector tool

[12-Apr-2010 14:08:32] <Jane_Curry> I dont see any values but the Host Resources MIB can provide info

[12-Apr-2010 14:08:35] <blull> the shell script under bin/

[12-Apr-2010 14:08:42] <blull> my setup is setup exactly as you do in that script.

[12-Apr-2010 14:08:45] <blull> on the collector.

[12-Apr-2010 14:10:07] <rmatte> Jane_Curry: you talking about via WMI or SNMP?

[12-Apr-2010 14:10:33] <rmatte> (I assume SNMP since you stated MIB)

[12-Apr-2010 14:10:36] <Jane_Curry> SNMP

[12-Apr-2010 14:10:51] <rmatte> yeh, it's pretty tricky... let me show you something I threw together...

[12-Apr-2010 14:11:50] <rmatte> Jane_Curry: http://fpaste.org/cxIj/

[12-Apr-2010 14:12:12] <rmatte> the OID for paging info is never the same, that's a bash script that hunts down the values and can be used as a command datasource

[12-Apr-2010 14:12:29] <rmatte> I have a couple of other scripts for memory and CPU

[12-Apr-2010 14:12:32] <rmatte> same idea

[12-Apr-2010 14:12:40] <rmatte> I hope to one day get them all converted to a collector plugin

[12-Apr-2010 14:13:21] <rmatte> blull: cool

[12-Apr-2010 14:14:11] <rmatte> it treats the paging as if it were a filesystem, it does the exact same with memory

[12-Apr-2010 14:14:26] <rmatte> and instead of making them the first 2 OIDs which would have been smart, they are always last

[12-Apr-2010 14:14:41] <rmatte> so the OID changes from server to server based on how many physical drives there are

[12-Apr-2010 14:17:01] <rmatte> oh sorry, that script is actually a bit off...

[12-Apr-2010 14:17:13] <rmatte> one second, there was a change that I made a while back, forgot to update that script

[12-Apr-2010 14:18:15] <rmatte> http://fpaste.org/uQwq/

[12-Apr-2010 14:18:16] <rmatte> there

[12-Apr-2010 14:18:21] <rmatte> 1.024 instead of 1024

[12-Apr-2010 14:18:32] <rmatte> makes it scale properly in Zenoss when graphing it

[12-Apr-2010 14:18:40] <Simon4> what's an order of magnitude between friends?

[12-Apr-2010 14:19:02] <Simon4> evening all

[12-Apr-2010 14:20:54] <rmatte> Simon4: hey

[12-Apr-2010 14:22:30] <Simon4> just snatching a few mins away from real work to do some dev :)

[12-Apr-2010 14:22:57] <rmatte> fun

[12-Apr-2010 14:23:08] <rmatte> I'm snatching a few mins away from dev to do some work

[12-Apr-2010 14:23:09] <rmatte> :P

[12-Apr-2010 14:29:06] <Jane_Curry> Thanks Ryan - I'll have a play with that

[12-Apr-2010 14:31:10] forsberg is now known as fOrsberg

[12-Apr-2010 14:34:31] <Jane_Curry> Any idea where the HR MIB gets the hrStorageSize value from for Windows?

[12-Apr-2010 14:34:48] <Jane_Curry> The value is always way off my c:\pagefile.sys..

[12-Apr-2010 14:35:03] <Jane_Curry> .. but I don't understand Windows that well

[12-Apr-2010 14:49:32] <rmatte> Jane_Curry: not sure

[12-Apr-2010 14:50:15] <rmatte> well, you're saying the value is way off, but it's listed in units, not bytes

[12-Apr-2010 14:50:56] <rmatte> If you'll notice in my script I gather another value which is units, then I take the value of hrStorageSize and multiply it by units then multiply again by 1.024 to get the actual value in bytes

[12-Apr-2010 14:51:25] <rmatte> or sorry, divide by 1.024

[12-Apr-2010 14:51:26] <rmatte> rather

[12-Apr-2010 14:55:21] <rmatte> so if you look at the hrStorageSize value and compare it directly to your swap size, it's not going to make sense

[12-Apr-2010 14:55:40] <rmatte> (or paging file size as they say in windowsland)

[12-Apr-2010 15:04:57] <Jane_Curry> Yeh - I know I have to do the arithmetic on allocationUnits and StorageSize

[12-Apr-2010 15:05:04] <Jane_Curry> but its still way off

[12-Apr-2010 15:05:25] <rmatte> really?

[12-Apr-2010 15:05:30] <rmatte> I find it's fairly accurate

[12-Apr-2010 15:09:12] <rmatte> then again, Microsoft's implementation of SNMP is pretty bad

[12-Apr-2010 15:09:48] <rmatte> ...and yet I still prefer it to WMI

[12-Apr-2010 15:31:03] <blull> hmm

[12-Apr-2010 15:31:07] <blull> when i cliek on perf

[12-Apr-2010 15:31:39] <blull> shouldn't it generate a packet to the remote collector for that server to say either pull the data or tell the collector that it needs to provide the data?

[12-Apr-2010 15:35:18] <rmatte> I would assume so, that's what zenrender handles on both the master and the collector

[12-Apr-2010 15:35:51] <rmatte> you have to actually load a graph though for it to happen

[12-Apr-2010 15:46:12] <sergeymasushko> hi guys. is it possible to collect perf data from the esxi v3 or v4?

[12-Apr-2010 15:47:48] <sloof3> mrayzenoss: Yes.  My perf data tends to disappear every so often.

[12-Apr-2010 15:48:37] <gwb235> I've been reading all the docs I can, but can't figure out why a command based datasource auto-clears and then creates a new event every time it's cycled.  What should I be looking at? (note: already verified no event transforms or Event zproperties are set that would auto-clear the event)

[12-Apr-2010 15:49:41] <gwb235> simple command, "/opt/zenoss/libexec/check_certs_Nagios.sh -x 30 -n -s  ${here/id} -p 443"

[12-Apr-2010 15:50:03] <sergeymasushko> gwb235: I've seen that bug earlier... there was a ticket about it. what version of the zenoss are you using?

[12-Apr-2010 15:50:10] <gwb235> 2.4.5

[12-Apr-2010 15:50:42] <gwb235> is there a simple web page to look at the patches to 2.4.5?

[12-Apr-2010 15:50:58] <sergeymasushko> gwb235: just a moment...

[12-Apr-2010 15:51:33] <rmatte> gwb235: docs/DOC-4434

[12-Apr-2010 15:51:54] <rmatte> there's a patch that fixes what you describe, there are also others listed there

[12-Apr-2010 15:52:03] <rmatte> though, I'd recommend just upgrading to 2.5.2 at this point

[12-Apr-2010 15:52:08] <sergeymasushko> gwb235: http://dev.zenoss.org/trac/ticket/5494

[12-Apr-2010 15:52:27] <gwb235> yeah... that's why we're at 2.4.5 right now... did not have a happy 2.5.2 upgrade and had to rollback

[12-Apr-2010 15:52:37] <rmatte> oh?

[12-Apr-2010 15:52:44] <rmatte> what went wrong with the 2.5.2 upgrade?

[12-Apr-2010 15:53:11] <sciolist> distributed collectors failed to upgrade

[12-Apr-2010 15:53:16] <rmatte> I see

[12-Apr-2010 15:53:35] <rmatte> well anyways, use the patches in the link that I pasted and it'll get your 2.4.5 installs workable

[12-Apr-2010 15:53:52] <sciolist> there is apprently a patch to apply to the rsync portion of the code, but we didn't have the docs available at the time and had to roll back

[12-Apr-2010 15:54:02] <rmatte> just copy and paste the zenpatch &&... command that's listed there

[12-Apr-2010 15:54:07] <rmatte> ah I see

[12-Apr-2010 15:54:08] <gwb235> zenoss support was (as always) outstanding, but we were outside our maintenance window by the time all players were on board and had already began the rollback

[12-Apr-2010 15:54:24] <rmatte> that sucks

[12-Apr-2010 15:54:43] <rmatte> do...

[12-Apr-2010 15:54:44] <rmatte> zenpatch 15003 && zenpatch 14397 && zenpatch 15214 && zenpatch 14049 && zenpatch 13304 && zenpatch 16733

[12-Apr-2010 15:54:49] <rmatte> then restart zenoss

[12-Apr-2010 15:55:07] <rmatte> you'll probably have to do it on the collectors as well

[12-Apr-2010 15:56:00] <gwb235> yup, thanks again for reminding me of the patches to 2.4.5, i remember applying them and forgot to reapply after rollback

[12-Apr-2010 15:56:22] <rmatte> no problem

[12-Apr-2010 15:56:47] <rmatte> my production stuff is still on 2.4.5 but I'm almost ready to upgrade

[12-Apr-2010 15:56:56] <rmatte> I have the upgrade steps down to a fine science

[12-Apr-2010 15:57:18] <sciolist> anything you can share?

[12-Apr-2010 15:57:42] <rmatte> well, it's mostly just the way that I add/remove the ZenPacks more than anything

[12-Apr-2010 15:57:53] <rmatte> it's fairly specific to our setup since most are custom ZenPacks

[12-Apr-2010 15:58:04] <sciolist> you using a stack install?

[12-Apr-2010 15:58:05] <sergeymasushko> our production version is 2.5.1

[12-Apr-2010 15:58:09] <rmatte> stack install, yes

[12-Apr-2010 15:58:20] <sergeymasushko> is there a need to upgrade to 2.5.2?

[12-Apr-2010 15:58:33] <sciolist> ah... we're RPm-based, so no cross-pollination goodness there

[12-Apr-2010 15:58:37] <rmatte> sergeymasushko: 2.5.2 has a LOT of bugfixes since 2.5.1

[12-Apr-2010 15:58:45] <rmatte> like 30+

[12-Apr-2010 15:58:52] <rmatte> so yes, it would be a good idea to upgrade

[12-Apr-2010 15:59:43] <sergeymasushko> rmatte: hm... since I have not noticed any issues... I guess that they are not critical for us :)

[12-Apr-2010 15:59:59] <rmatte> well, some are pretty bad but you may not notice them

[12-Apr-2010 16:00:25] <sergeymasushko> hm... need to look at the change list...

[12-Apr-2010 16:00:50] <rmatte> http://dev.zenoss.org/trac/query?status=assigned&status=backlog&status=closed&status=new&status=verification&group=patch_state&patch=2.5.2&col=id&col=summary&col=status&col=owner&col=priority&col=milestone&col=component&col=changetime&report=6&order=status

[12-Apr-2010 16:00:53] <lthrasher> does zenoss exec() the rrdtool binary for its rrd operations, or does it use the librrd-python bindings?

[12-Apr-2010 16:01:00] <rmatte> sergeymasushko: there's your list

[12-Apr-2010 16:01:32] <sergeymasushko> btw... sometimes zenoss reports about hartbeats... how can I find the cause?

[12-Apr-2010 16:01:38] <sergeymasushko> rmatte: thanks

[12-Apr-2010 16:01:58] <rmatte> sergeymasushko: how often does it report about heartbeats, every 2 hours or so?

[12-Apr-2010 16:02:24] <sergeymasushko> rmatte: once or twice a day maybe... .

[12-Apr-2010 16:07:56] <blull> does zenoss cache on the zenhub?

[12-Apr-2010 16:08:02] <blull> and if so, can I disable it temporarily for debugging?

[12-Apr-2010 16:08:09] <blull> everuy time I update the perf's

[12-Apr-2010 16:08:15] <blull> i see NO connections generated to ANY daemons :/

[12-Apr-2010 16:08:25] <blull> so im assuming its doing something internal inside the zenhub running on 8080

[12-Apr-2010 16:09:39] <rmatte> well, 8080 is just the web interface

[12-Apr-2010 16:09:56] <rmatte> but it uses other 80xx ports to connect to all the daemons

[12-Apr-2010 16:10:15] <rmatte> what do you mean by "update the perfs"?

[12-Apr-2010 16:11:18] <rmatte> afk for a few

[12-Apr-2010 16:30:04] <rmatte> back

[12-Apr-2010 17:15:21] fOrsberg is now known as forsberg

[12-Apr-2010 17:16:36] <gwb2352> rmatte: thanks for the link to post 2.4.5 patches.  everything happy again.

[12-Apr-2010 19:50:46] forsberg is now known as fOrsberg

[12-Apr-2010 23:10:45] <davetoo> Anybody here tested/using zenoss on RHEL 5.5?

[12-Apr-2010 23:53:13] <rhettardo> no but i use it on 5.4

[12-Apr-2010 23:54:26] <davetoo> aye

[12-Apr-2010 23:54:37] <davetoo> I'm building a vm for development, thought I'd give it a shot

[13-Apr-2010 00:00:30] [disconnected at Tue Apr 13 00:00:30 2010]

[13-Apr-2010 00:00:30] [connected at Tue Apr 13 00:00:30 2010]

[13-Apr-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[13-Apr-2010 03:53:14] <penthium> good morning, what is the period for the auto-modeling for the device ? can I change the value and where ?

[13-Apr-2010 04:13:26] fOrsberg is now known as forsberg

[13-Apr-2010 05:20:46] forsberg is now known as fOrsberg

[13-Apr-2010 07:19:49] fOrsberg is now known as forsberg

[13-Apr-2010 07:45:31] <jane_> test

[13-Apr-2010 07:46:40] <Simon4> ping

[13-Apr-2010 07:53:11] <penthium> how can I change the "Threshold session" for the apache monitoring zenpack ?

[13-Apr-2010 07:53:30] <penthium> pong

[13-Apr-2010 08:49:19] <penthium> how can I change the "Threshold session" for the apache monitoring zenpack ? please

[13-Apr-2010 08:52:00] <mrayzenoss> penthium: go to the Apache template, probably on /Devices/Server

[13-Apr-2010 08:52:07] <mrayzenoss> and the threshold is defined there

[13-Apr-2010 09:02:27] <penthium> I find the apache template but I don't find the threshold for the session number, only the CPU over 90 percent.

[13-Apr-2010 09:14:30] <penthium> I found the anwser :D I made a personal template who count the number of session for my apache servers, but  I class these event in the "/app/appache"

[13-Apr-2010 09:15:07] <penthium> so my issue doesn't form the Zenpack apache monitor but form my own script \o/

[13-Apr-2010 09:15:58] <penthium> thx marayzenoss for your little help, you open my tired eyes

[13-Apr-2010 09:17:24] <mrayzenoss> good

[13-Apr-2010 09:20:14] <penthium> HOOOOOooooooo it's so beautiful ! I see life in green :D

[13-Apr-2010 09:33:12] <marksteele> how does one scale zenoss core?

[13-Apr-2010 09:33:46] <marksteele> does it support proxies that aggregate check results?

[13-Apr-2010 09:35:15] <Jane_Curry> I have just put an append on the users forum about gathering  virtual memory / pagespace data

[13-Apr-2010 09:35:18] <Jane_Curry> thread/13276

[13-Apr-2010 09:35:44] <Jane_Curry> I would be interested in any feedback

[13-Apr-2010 09:37:45] <marksteele> website claims zenoss scales horizontally, anyone know how this is accomplished?

[13-Apr-2010 09:37:56] <Simon4> distributed collectors

[13-Apr-2010 09:37:59] <Simon4> search for that

[13-Apr-2010 09:38:07] <marksteele> thanks

[13-Apr-2010 09:38:08] <mrayzenoss> marksteele: how big do you need to scale?

[13-Apr-2010 09:38:27] <marksteele> don't know yet, but I tend to plan for 'big' :)

[13-Apr-2010 09:38:55] <marksteele> another question off the top of my head, but any replication/HA functionality?

[13-Apr-2010 09:39:37] <marksteele> or is the usual HA/replication route something like DRBD?

[13-Apr-2010 09:39:39] <mrayzenoss> Enterprise has a supported HA configuration, mostly based on known best-practices.

[13-Apr-2010 09:39:46] <mrayzenoss> yeah

[13-Apr-2010 09:39:50] <marksteele> gotcha

[13-Apr-2010 09:40:21] <mrayzenoss> depending on your hardware and what you're monitoring, a single box can probably do 1-2000 devices

[13-Apr-2010 09:40:29] <mrayzenoss> I've seen 8K on a single box though

[13-Apr-2010 09:42:17] <rmatte> If you have a beast of a server you can monitor tons

[13-Apr-2010 09:50:08] <penthium> what is the period for the auto-modeling for the device ? can I change the value and where ? ( I want to monitor route's table and have a event when a route disapear )

[13-Apr-2010 09:50:40] <penthium> because I don't see in gui the posibility to monitor the route table in the Os tab

[13-Apr-2010 10:52:00] <davetoo> phew

[13-Apr-2010 10:53:38] <Brixius> beeno perhaps?

[13-Apr-2010 10:58:44] <davetoo> Getting Eclipse Galileo running on RHEL 5 has been a pain in my neck for hours

[13-Apr-2010 10:59:07] <davetoo> trying to set up a zenoss development environment in a VM that I can clone

[13-Apr-2010 11:00:02] <Brixius> Question on an event transform.  http://pastebin.org/149313

[13-Apr-2010 11:00:17] <davetoo> (and that would be Beano)

[13-Apr-2010 11:00:59] <davetoo> hmm

[13-Apr-2010 11:01:08] <davetoo> too early, too many months since I've done one of those, sorry

[13-Apr-2010 11:02:25] <Brixius> I'm wondering if the hasattr will work for snmp trap vars.

[13-Apr-2010 11:04:34] <davetoo> I have .. not looked at those yet; don't know how they are represented

[13-Apr-2010 11:04:53] <rmatte> Brixius: it should

[13-Apr-2010 11:04:59] <rmatte> Brixius: an attribute is an attribute

[13-Apr-2010 11:05:54] <davetoo> well, 'evt.componant' won't work

[13-Apr-2010 11:06:01] <davetoo> evt.component might

[13-Apr-2010 11:06:05] <rmatte> the fact that you can do a getattr on them pretty much confirms that a hasattr will work as well

[13-Apr-2010 11:06:14] <rmatte> evt.component, correct

[13-Apr-2010 11:06:16] <rmatte> watch the spelling

[13-Apr-2010 11:06:54] <Brixius> I guess it's picky that way.

[13-Apr-2010 11:06:57] <Brixius> ;)

[13-Apr-2010 11:07:19] <davetoo> I'd rewrite that block using a dict to hold the component/summary data

[13-Apr-2010 11:07:21] <davetoo> easier to read

[13-Apr-2010 11:07:38] <davetoo> maybe

[13-Apr-2010 11:08:03] <rmatte> davetoo: meh, doesn't really matter, why overcomplicate?

[13-Apr-2010 11:08:07] <davetoo> I *hope* I get to dig into zenoss again

[13-Apr-2010 11:08:13] <rmatte> The code is quite easy to read as is

[13-Apr-2010 11:08:34] <davetoo> just personal preference; the other is easier for my eyes to read :)

[13-Apr-2010 11:08:36] <rmatte> yeh, I haven't had much of a chance to do any digging/dev work lately

[13-Apr-2010 11:08:41] <davetoo> well,

[13-Apr-2010 11:08:46] <rmatte> well, other than cleaning up ZenPacks

[13-Apr-2010 11:08:49] <davetoo> some of the . um, I'll be polite here,

[13-Apr-2010 11:09:03] <rmatte> lol

[13-Apr-2010 11:09:06] <davetoo> some folks at work are being penny-wise, pound-foolish IMO

[13-Apr-2010 11:09:24] <davetoo> and looking at alternatives to renewing our enterprise contract,

[13-Apr-2010 11:09:29] <davetoo> which .. bugs me no end

[13-Apr-2010 11:09:44] <davetoo> considering all the sleep I've forsaken to get where we are :)

[13-Apr-2010 11:10:05] <rmatte> I'll assume that penny-wise, pound-foolish is a british term for ill spending?

[13-Apr-2010 11:10:25] <davetoo> Maybe originally, but I'm California born and raised :)

[13-Apr-2010 11:10:30] <rmatte> I see

[13-Apr-2010 11:10:42] <rmatte> I've never heard that particular term used before, so was just curious

[13-Apr-2010 11:11:06] <rmatte> :)

[13-Apr-2010 11:11:25] <davetoo> I think they'll spend as much or more money from different budget categories, and meanwhile monitoring will suffer

[13-Apr-2010 11:11:35] <Brixius> I was playing around with zendmd trying to test the hasattr and I seem to be doing something wrong.

[13-Apr-2010 11:11:39] <davetoo> but.. they're gonna do what they 're gonna do

[13-Apr-2010 11:11:49] <Brixius> I do a evt=dmd.ZenEventManager.getEventDetailFromStatusOrHistory("88169294-98b9-4a3b-9683-eab0b7fe424c")

[13-Apr-2010 11:12:03] <Brixius> then I can't refrence evt.ifDescr

[13-Apr-2010 11:12:14] <davetoo> What I remember is that .. an evt is not always fully-formed

[13-Apr-2010 11:12:28] <davetoo> it can have different contents depending upon when you query it

[13-Apr-2010 11:12:33] <davetoo> maybe I'm wrong?

[13-Apr-2010 11:12:56] <davetoo> anyway, a trick I add to all of my scripts now:

[13-Apr-2010 11:13:01] <rmatte> Brixius: what version of Zenoss?

[13-Apr-2010 11:13:11] <davetoo>   import print;   pp = lambda x: pprint.pprint(x)

[13-Apr-2010 11:13:12] <Brixius> but I can find them in the collection ect.getEventDetails()

[13-Apr-2010 11:13:19] <davetoo> I use that endlessly in zendmd

[13-Apr-2010 11:13:20] <Brixius> I'm on 2.5.2

[13-Apr-2010 11:13:26] <rmatte> oh, you're testing in zendmd and it'[s not working?

[13-Apr-2010 11:13:31] <rmatte> that's weird

[13-Apr-2010 11:13:45] <Brixius> er evt.getEventDetails()

[13-Apr-2010 11:13:47] <rmatte> evt = dmd.ZenEventManager.getEventDetailFromStatusOrHistory() is the correct function

[13-Apr-2010 11:13:47] <davetoo> inspect module is also quite helpful for debugging this stuff

[13-Apr-2010 11:14:04] <rmatte> so it should be working

[13-Apr-2010 11:14:09] <rmatte> provided that's a good evid

[13-Apr-2010 11:14:31] <Brixius> yep it's a good id, I can print evt.summary and get the summary I expect

[13-Apr-2010 11:16:05] <davetoo> rmatte: I can't recall.. do you work for Zenoss?

[13-Apr-2010 11:16:18] <rmatte> no, I don't

[13-Apr-2010 11:16:20] <mrayzenoss> heh

[13-Apr-2010 11:16:30] <rmatte> ;)

[13-Apr-2010 11:17:01] <davetoo> I'm considering whether I want to move to Austin or MD :)

[13-Apr-2010 11:17:22] <rmatte> hehe

[13-Apr-2010 11:17:22] <davetoo> wouldn't care for the weather either place, though

[13-Apr-2010 11:17:48] <davetoo> hey, my employer has two ex-Zenoss people right now,

[13-Apr-2010 11:17:53] <rmatte> If I was a U.S citizen I would consider it, but there's way too much red tape when moving from Canada to work in the U.S

[13-Apr-2010 11:17:54] <mrayzenoss> right now Austin is beautiful, but come summer there will probably be another 3 months of 100+

[13-Apr-2010 11:17:54] <davetoo> maybe the door could swing the other way

[13-Apr-2010 11:18:20] <rmatte> one of my friends got hired by NBC Universal and was going to move to L.A but never made it across the border, so he just works for them from home here.

[13-Apr-2010 11:18:26] <davetoo> it gets hot here in the Silly Valleu but it's dry

[13-Apr-2010 11:18:31] <mrayzenoss> we've got 2 devs in Canada

[13-Apr-2010 11:18:32] <davetoo> heh

[13-Apr-2010 11:18:48] <davetoo> I have a friend who... let her visa expire and when she went home to T.O. for a visit,

[13-Apr-2010 11:19:00] <davetoo> got caught and .. hasn't been back to the US in four years :)

[13-Apr-2010 11:19:16] <rmatte> davetoo: rough

[13-Apr-2010 11:19:44] <rmatte> I live about 5-6 hours away from T.O.

[13-Apr-2010 11:25:01] <Brixius> so any ideas why I can't refrence evt.ifDescr from zendmd after getting the event with dmd.ZenEventManager.getEventDetailFromStatusOrHistory()

[13-Apr-2010 11:25:22] <davetoo> I wonder if anybod's come up with a Zenoss Development Appliance :)

[13-Apr-2010 11:25:22] <rmatte> well, you wouldn't be able to just do evt.ifDescr

[13-Apr-2010 11:25:32] <rmatte> you'd have to do the getattr

[13-Apr-2010 11:26:37] <rmatte> ifdescription = getattr(evt, 'ifDescr', 'Unknown')

[13-Apr-2010 11:26:39] <rmatte> or whatever

[13-Apr-2010 11:28:34] <Brixius> so in the transforms there must be some additional code running to allow the evt.varbind syntax to work then

[13-Apr-2010 11:29:05] <Brixius> well I mean zenoss runs some additional code to allow it.

[13-Apr-2010 11:29:23] <Brixius> and thats a ?

[13-Apr-2010 11:29:44] <rmatte> ok, I think you're confusing a couple things...

[13-Apr-2010 11:29:57] <Brixius> it wouldn't supprise me

[13-Apr-2010 11:30:05] <rmatte> when you check event details... there are certain event details that can be accessed directly with evt.whatever

[13-Apr-2010 11:30:20] <rmatte> such as device, component, evid, etc...

[13-Apr-2010 11:30:57] <rmatte> then, there are other ones which need to be accessed with the getattr function (those used to show up on a separate details tab, but in the latest event console they get jumbled in with everything else without distinction)

[13-Apr-2010 11:31:14] <rmatte> which I mentioned to Nate at some point in the past but it never got fixed.

[13-Apr-2010 11:32:09] <rmatte> so if ifIndex is an attribute passed form an SNMP trap (as I suspect it is)... then you need to use getattr to set a variable to that value

[13-Apr-2010 11:32:21] <rmatte> such as:

[13-Apr-2010 11:32:23] <rmatte> ifdescription = getattr(evt, 'ifDescr', 'Unknown')

[13-Apr-2010 11:32:32] <rmatte> then you would call on ifdescription when you need that value

[13-Apr-2010 11:32:47] <Brixius> is this in both zendmd and transforms?

[13-Apr-2010 11:33:02] <rmatte> if you're using dmd.ZenEventManager.getEventDetailFromStatusOrHistory() in zendmd, yes

[13-Apr-2010 11:33:08] <Brixius> ok

[13-Apr-2010 11:33:11] <rmatte> that's all that Zenoss is doing before it executes a transform really

[13-Apr-2010 11:33:25] <rmatte> (along with some other stuff that zendmd does automatically when it's started)

[13-Apr-2010 11:34:12] <rmatte> zendmd is python, and transform is python, the way that they access properties in an event doesn't change

[13-Apr-2010 11:34:36] <Brixius> Well this is what is confusing me then.  I have the following transform that does work.. "evt.summary = "CAC failure: nexToneRegId = %s" % (evt.nexToneRegId)" and as you can see it's revrencing just evt.nexToneRegId which is just a varbind.

[13-Apr-2010 11:35:29] <Brixius> I'm not using the getattr function to get that varbind from the snmp trap

[13-Apr-2010 11:35:54] <rmatte> hmmm, I've never had that work for me, so no idea

[13-Apr-2010 11:36:32] <rmatte> the other benefit of using getattr is that your transform/script won't crap out in the event that the attribute doesn't exist

[13-Apr-2010 11:37:25] <rmatte> If you reference evt.ifIndex and it doesn't exit, bye bye transform

[13-Apr-2010 11:37:31] <rmatte> getattr is just better practice

[13-Apr-2010 11:37:56] <rmatte> I didn't think you could directly call on those

[13-Apr-2010 11:38:24] <rmatte> s/exit/exist/g

[13-Apr-2010 11:39:11] <Brixius> ok I'll try the getattr and see what happens, I was trying the getattr you posted in zendmd and it allway's returns the "Unknown" part, never gives me the value shown in evt.getEventDatails()

[13-Apr-2010 11:39:36] <rmatte> if it's returning unknown it's because it's not able to properly retrieve the attribute

[13-Apr-2010 11:39:46] <rmatte> attribute names are case sensitive by the way

[13-Apr-2010 11:39:55] <Brixius> yep, checked that

[13-Apr-2010 11:39:59] <rmatte> k

[13-Apr-2010 11:40:24] <rmatte> double check the evid that you're using, double check everything

[13-Apr-2010 11:40:53] <rmatte> If that's the proper evid and that event actually does have an ifIndex attribute then there's absolutely no reason for it not to be working

[13-Apr-2010 11:40:59] <rmatte> it's coding, it's not magic ;)

[13-Apr-2010 11:47:48] <Brixius> looks like getattr(evt,'summary','u') is really just doing some error checking on evt.summary and if it doesn't exist return the second part.

[13-Apr-2010 11:47:59] <rmatte> that's correct

[13-Apr-2010 11:48:49] <rmatte> so I guess you should technically be able to grab it as just evt.ifIndex if it does exist

[13-Apr-2010 11:48:58] <Brixius> yep

[13-Apr-2010 11:49:14] <Brixius> which is why I was doing the hasattr first

[13-Apr-2010 11:49:29] <Brixius> but in zendmd it's not the same

[13-Apr-2010 11:51:17] <Brixius> in zendmd I have to loop through the evt.getEventDetails() get the values.

[13-Apr-2010 11:51:49] <rmatte> weird

[13-Apr-2010 11:52:08] <rmatte> wonder why that is

[13-Apr-2010 11:52:36] <rmatte> maybe zenoss is doing something else when it's running transforms

[13-Apr-2010 11:52:49] <Brixius> that's why I was thinking the transform processor is doing something to take the values in evt.getEventDetails() and turning them into properties of the event.

[13-Apr-2010 11:56:11] <Brixius> how do I use the value of a varable for a variable name, ie something like for i,j in evt.getEventDetails(): evt.i = j

[13-Apr-2010 11:57:38] <Brixius> nevermind found it, setattr

[13-Apr-2010 11:59:58] <Brixius> so as long as I run

[13-Apr-2010 12:00:03] <Brixius> for key,value in evt.getEventDetails():

[13-Apr-2010 12:00:10] <Brixius> setattr(evt,key,value)

[13-Apr-2010 12:00:57] <Brixius> I can refrence them in zendmd just like in transforms.(note the second line is indented but irc removed them)

[13-Apr-2010 12:11:46] <rmatte> ah, cool

[13-Apr-2010 12:11:59] <rmatte> guess that's what they do when a transform is executed then

[13-Apr-2010 12:21:01] forsberg is now known as fOrsberg

[13-Apr-2010 12:22:26] <bigegor> I just posted WMIPerf_Windows RC2, testers are welcomed.

[13-Apr-2010 12:42:20] fOrsberg is now known as forsberg

[13-Apr-2010 12:53:13] <ckrough> anyone know how to dump the events database password from dmd?

[13-Apr-2010 12:54:17] <rmatte> ckrough: yeh...

[13-Apr-2010 12:55:42] <rmatte> one second, just making sure it's right

[13-Apr-2010 12:56:16] <rmatte> dmd.ZenEventManager.password

[13-Apr-2010 12:56:24] <rmatte> that'll do it

[13-Apr-2010 12:56:26] <ckrough> heh

[13-Apr-2010 12:56:30] <ckrough> thanks

[13-Apr-2010 12:56:32] <rmatte> np

[13-Apr-2010 12:56:40] <Simon4> don't forget to decrypt.. oh wait

[13-Apr-2010 12:56:49] <rmatte> Simon4: :)

[13-Apr-2010 12:56:51] <Simon4> :)

[13-Apr-2010 12:56:53] <rmatte> good one

[13-Apr-2010 12:56:54] <rmatte> heheh

[13-Apr-2010 12:57:03] <Simon4> not sure that'll ever cease to be amusing

[13-Apr-2010 12:57:15] <rmatte> I still like the:

[13-Apr-2010 12:57:35] <rmatte> dev = dmd.Devices.findDevice('device')

[13-Apr-2010 12:57:39] <rmatte> dev.zWinPassword

[13-Apr-2010 12:57:46] <Simon4> hah

[13-Apr-2010 12:58:04] * Simon4 wonders just how many Administrator level passwords are avail that way

[13-Apr-2010 12:58:15] <rmatte> pretty much every single one

[13-Apr-2010 12:58:21] <rmatte> if it's stored in Zenoss anyways

[13-Apr-2010 12:58:55] <Simon4> monitoring systems were always goldmines for prospective hackers though

[13-Apr-2010 12:59:00] <rmatte> they are basically relying on security through obscurity for the time being

[13-Apr-2010 12:59:06] <Simon4> nice centralised database of every system on a network

[13-Apr-2010 12:59:13] <Simon4> passwords or not

[13-Apr-2010 12:59:14] <rmatte> yup

[13-Apr-2010 13:00:19] <rmatte> then again, a Zenoss box should be locked down well enough that that's not an issue

[13-Apr-2010 13:00:52] <Simon4> indeed

[13-Apr-2010 13:01:17] <rmatte> it would probably also be possible to run Zenoss from within a truecrypt container or something

[13-Apr-2010 13:01:42] <cgibbons> bounce bounce bounce

[13-Apr-2010 13:02:04] <rmatte> ?

[13-Apr-2010 13:02:40] <Simon4> rmatte: imagine the performance fun of that :)

[13-Apr-2010 13:02:45] <rmatte> yeh

[13-Apr-2010 13:02:53] <Simon4> and hey, if you have phys access to the box, just get the zope cache out of ram

[13-Apr-2010 13:03:01] <Simon4> most of the DB is in there if you have cache settings fairly high

[13-Apr-2010 13:03:16] <rmatte> true

[13-Apr-2010 13:11:07] <ckrough> hmm, which daemon needs to be restarted to recognize an event db host change?

[13-Apr-2010 13:13:03] <ckrough> nm, its automagic... just needs more patience fromt he admin

[13-Apr-2010 13:21:37] <rmatte> yup

[13-Apr-2010 13:24:15] <davetoo> so .. eclipse users...

[13-Apr-2010 13:24:27] <davetoo> do you run you development instance of zenoss as yourself?

[13-Apr-2010 13:24:42] <davetoo> or set eclipse up for 'zenoss'

[13-Apr-2010 13:25:10] <rmatte> pretty sure it has to be run as the zenoss user

[13-Apr-2010 13:25:21] <rmatte> there's too much stuff dependant on it

[13-Apr-2010 13:25:46] <rmatte> dependent*

[13-Apr-2010 13:32:35] <cgibbons> nope

[13-Apr-2010 13:32:38] <ckrough> I will be rebuilding my workstation soon. I think I am going to go with a CentOS5.4 VM guest and just run it as zenoss in the guest

[13-Apr-2010 13:32:40] <cgibbons> i never, ever run mine as zenoss user

[13-Apr-2010 13:32:50] <jb> hey cgibbons

[13-Apr-2010 13:33:01] <rmatte> ah, k

[13-Apr-2010 13:33:03] <jb> you played with 2008 R2 yet?

[13-Apr-2010 13:33:05] <cgibbons> (as a source build)

[13-Apr-2010 13:33:07] <cgibbons> nope

[13-Apr-2010 13:33:12] <jb> damn WMI memory leak is killing me.

[13-Apr-2010 13:33:28] <cgibbons> they introduced another one again? geez

[13-Apr-2010 13:33:44] <jb> wmi is constantly crashing on all of my R2 boxes

[13-Apr-2010 13:35:23] <rmatte> wmi == garbage

[13-Apr-2010 13:35:36] <rmatte> it worked fine in 2003, ever since 2008 it's been horrible

[13-Apr-2010 13:35:48] <rmatte> I hope they fix it

[13-Apr-2010 13:36:00] <Brixius> I think I'm using it in 2008 without any problems.

[13-Apr-2010 13:36:19] <jb> 2008 is fine, 2008 R2 is not.

[13-Apr-2010 13:36:25] <Brixius> ahh ok

[13-Apr-2010 13:36:28] <rmatte> we had 1 2008 box that we setup as a test box, very basic, running nothing on it, and WMI would crash on it at least once a week

[13-Apr-2010 13:36:31] <rmatte> drove me nuts

[13-Apr-2010 13:36:36] <rmatte> we tried reinstalling, same issue

[13-Apr-2010 13:36:43] <rmatte> then we installed 2003, and it worked fine

[13-Apr-2010 13:36:43] <Brixius> well that sucks

[13-Apr-2010 13:36:52] <cgibbons> there was a memory leak in the WMI library when 2008 came out, they fixed it with a hotfix & then the service pack

[13-Apr-2010 13:37:16] <rmatte> ah, I never tried it with the hotfix/service pack, so that was probably it

[13-Apr-2010 13:39:09] <cgibbons> it was specifically with 'notification queries' which is what Zenoss uses to watch the event log & service status

[13-Apr-2010 13:39:26] <cgibbons> jb - they do seem responsive to WMI bugs if you haven't reported it yet

[13-Apr-2010 13:39:37] <jb> yeah havent yet

[13-Apr-2010 13:39:38] <jb> but i will

[13-Apr-2010 13:43:12] <bigegor> how about WinRM?

[13-Apr-2010 13:51:15] <Brixius> I haven't looked at it yet

[13-Apr-2010 14:45:35] <Simon4> is device.id unique across an install? so that dmd.Devices.findDevice('device.x.y.com') only ever returns one "device.x.y.com" ?

[13-Apr-2010 14:46:27] <rmatte> well, you can't have devices named the same thing

[13-Apr-2010 14:46:48] <davetoo> what about "ip realms"?

[13-Apr-2010 14:47:02] <davetoo> which I haven't yet looked at (an Enterprise-only feature)

[13-Apr-2010 14:47:07] <rmatte> I've never dealt with them, I assume that's an enterprise feature

[13-Apr-2010 14:47:08] <rmatte> yeh

[13-Apr-2010 14:48:33] <Simon4> rmatte: that's the answer I need for now at least, thanks :)

[13-Apr-2010 14:49:09] <Simon4> need to "move" a bunch of devices from being "Devices" to being "CustomDevices", so scripting up removing them and re-adding em without removing events/perf/etc

[13-Apr-2010 14:50:00] <davetoo> CustomDevice?

[13-Apr-2010 14:50:16] <Simon4> davetoo: zenpack that extends the device type with more attributes

[13-Apr-2010 14:50:18] <davetoo> You made a subclass?

[13-Apr-2010 14:50:21] <davetoo> ah

[13-Apr-2010 14:50:22] <Simon4> yeah

[13-Apr-2010 14:50:28] * Simon4 is crap with terminology at times

[13-Apr-2010 14:50:30] <davetoo> what kind of attributes?

[13-Apr-2010 14:51:27] <davetoo> zenpack that you wrote?

[13-Apr-2010 14:51:30] <Simon4> davetoo: just some extra server/application things - tis used internally

[13-Apr-2010 14:51:33] <Simon4> people at work wrote it

[13-Apr-2010 14:51:44] <Simon4> but it seems that I can't just move devices into that class

[13-Apr-2010 14:51:54] <davetoo> hmm

[13-Apr-2010 14:52:29] <davetoo> Makes sense, in a way; they are a different object type

[13-Apr-2010 14:52:43] <Simon4> yeah exactly

[13-Apr-2010 14:52:49] * Simon4 will share script once it's hacked up

[13-Apr-2010 14:53:10] <davetoo> I'm trying to come up with a good development environment

[13-Apr-2010 14:53:23] <Simon4> I have a centos vm that I write zenpacks in

[13-Apr-2010 14:53:40] <Simon4> it's not a "good" development environment, but it works for some values of works :)

[13-Apr-2010 14:53:45] <davetoo> aye

[13-Apr-2010 14:53:56] <davetoo> I'm trying to build an RHEL5.5 vm right now

[13-Apr-2010 14:54:00] <rmatte> yeh, I pretty much do the same, just have a lab vm with Zenoss installed on it

[13-Apr-2010 14:54:05] <davetoo> mostly because I need experience with RHEL5.5

[13-Apr-2010 14:54:31] <rmatte> I'm not a big redhat fan

[13-Apr-2010 14:54:52] <davetoo> I figure I'll install zenoss from the RPM with will build the user account, then I'll mess itup by installing source on top of it :)

[13-Apr-2010 14:54:54] <davetoo> or maybe not

[13-Apr-2010 14:55:00] <rmatte> lol

[13-Apr-2010 14:55:23] <rmatte> or you could just follow the source install steps :P

[13-Apr-2010 14:55:55] <davetoo> rmatte: I'd rather not have to use it, but for one thing, Zenoss Enterprise support kind of requires it

[13-Apr-2010 14:55:59] <davetoo> rmatte: have, many many times,

[13-Apr-2010 14:56:01] <davetoo> it's quite slow

[13-Apr-2010 14:56:06] <davetoo> and easy to mess up

[13-Apr-2010 14:57:52] <davetoo> I need to be able to switch between 2.4.2 and 2.5.something

[13-Apr-2010 14:58:10] <davetoo> 'cause I still have several 2.4.2 Core systems and the API is different

[13-Apr-2010 15:07:55] <Simon4> rmatte: I'm not a huge redhat fan either, but work's env is rpm install on redhat, so keeps things consistent

[13-Apr-2010 15:09:09] <rmatte> yeh

[13-Apr-2010 15:32:45] <davetoo> who is chadwick?

[13-Apr-2010 15:35:38] <mrayzenoss> cgibbons?

[13-Apr-2010 15:36:00] <davetoo> The guy that wrote the three-page Eclipsedev evironment thing

[13-Apr-2010 15:36:11] <mrayzenoss> one and the same

[13-Apr-2010 15:36:13] <davetoo> http://blog.nuclearbunny.org/2009/10/22/my-zenoss-development-environment-part-1/

[13-Apr-2010 15:36:14] <davetoo> ah!

[13-Apr-2010 15:36:17] <davetoo> ok :)

[13-Apr-2010 15:36:19] <rmatte> [10:12]  * chadwick (n=cgibbons@rrcs-97-77-9-2.sw.biz.rr.com) has joined #zenoss

[13-Apr-2010 15:36:23] <rmatte> ;)

[13-Apr-2010 15:36:55] <davetoo> Then I won't clutter up his blog anymore, I'll ask here next time I see him

[13-Apr-2010 15:37:21] <rmatte> hehe

[13-Apr-2010 15:37:31] <cgibbons> blog is prolly better, I don't hang out here that much

[13-Apr-2010 15:37:37] <davetoo> So many questions :)

[13-Apr-2010 15:37:50] <davetoo> well, since you're here :)

[13-Apr-2010 15:37:54] <rmatte> yeh, he really doesn't

[13-Apr-2010 15:37:58] <rmatte> lol

[13-Apr-2010 15:38:03] <davetoo> when you do a "build", do you deploy to /opt/zenoss?

[13-Apr-2010 15:38:29] <davetoo> or only if it's a 'dist' build (or whatever terms you use)

[13-Apr-2010 15:38:35] <cgibbons> nope, I usually set my ZENHOME to ~/src/zenoss/zenoss-<version>/zenhome

[13-Apr-2010 15:38:49] <davetoo> 'k

[13-Apr-2010 15:39:07] <davetoo> I wish I had better svn chops :)

[13-Apr-2010 15:39:09] <cgibbons> but my needs are for developing/supporting multiple versions simultaneously, rather than using a source build in production

[13-Apr-2010 15:39:26] <davetoo> I actually need to develop for 2.4.2 and the latest release

[13-Apr-2010 15:39:45] <davetoo> because.. I haven't figured out a way to make my zendmd scripts work seamlessly in both :)

[13-Apr-2010 15:41:04] <davetoo> I need to figure out how to do both with the minimum of duplcated files/wasted disk space

[13-Apr-2010 15:46:19] <cgibbons> hmmm my 2.4.x, 2.5.x, and trunk directories are only 6 GB total, pretty tiny

[13-Apr-2010 15:47:46] <davetoo> yeah, I should know better than to try to take shortcuts, it always bites me

[13-Apr-2010 15:50:14] <davetoo> so do you know where the major zendmd api changes were in the 2.4 line?  If I'm writing tools to manage my soon-to-be-deprectated 2.4.2 systems, should I download a specific 2.4.x tag?  Or just the branch head?

[13-Apr-2010 15:50:45] <mrayzenoss> you should be able to get 2.4.2

[13-Apr-2010 15:51:46] <mrayzenoss> 2.4.x is the branch, http://dev.zenoss.org/trac/browser/tags/zenoss-2.4.2 is the tag

[13-Apr-2010 15:52:04] <davetoo> thanks

[13-Apr-2010 15:52:57] <davetoo> I've been sidetracked for months but I had some patches I wanted to test and submit

[13-Apr-2010 16:53:45] <davetoo> cgibbons: do you keep separate workspaces or do you use the same one and switch via svn?

[13-Apr-2010 17:06:14] <davetoo> wow

[13-Apr-2010 17:06:35] <davetoo> my current events table is too big for zenoss to manage, apparently

[13-Apr-2010 17:06:46] <davetoo> can't clear them from the UI

[13-Apr-2010 17:07:01] <mrayzenoss> Select-All is fixed in 2.5.2

[13-Apr-2010 17:07:48] <davetoo> I'm trying to clean it up before upgrading

[13-Apr-2010 17:07:51] <davetoo> from 2.4.5

[13-Apr-2010 17:08:03] <davetoo> I'm afraid that it's so busy/hosed that the migrate won't work well

[13-Apr-2010 17:08:14] <mrayzenoss> you can do a zenbackup and leave out events

[13-Apr-2010 17:08:23] <davetoo> ah

[13-Apr-2010 17:08:24] <mrayzenoss> you don't have to migrate them

[13-Apr-2010 17:08:32] <mrayzenoss> zenbackup —help has options

[13-Apr-2010 17:08:37] <davetoo> alternatively, is it safe for me to truncate the events table?

[13-Apr-2010 17:08:41] <mrayzenoss> yes

[13-Apr-2010 17:08:43] <davetoo> with mysql?

[13-Apr-2010 17:08:48] <mrayzenoss> assuming you don't want the events :p

[13-Apr-2010 17:09:17] <davetoo> no

[13-Apr-2010 17:10:24] <davetoo> should I stop zenoss?

[13-Apr-2010 17:11:05] <mrayzenoss> probably be safest

[13-Apr-2010 17:15:16] <davetoo> it's the 'status' table?

[13-Apr-2010 17:17:20] <mrayzenoss> yes

[13-Apr-2010 17:17:28] <mrayzenoss> http://dev.zenoss.org/trac/browser/trunk/Products/ZenEvents/db/zenevents.sql

[13-Apr-2010 17:18:21] <davetoo> yep, got it, thanks

[13-Apr-2010 17:18:40] <davetoo> I've spoiled/ruined myself with phpMysql :)

[13-Apr-2010 17:36:35] <davetoo> buh

[13-Apr-2010 17:36:42] <davetoo> maybe I didn't make this VM disk large enough

[13-Apr-2010 17:53:14] <davetoo> it's taking *forever* to truncate this history table :)

[13-Apr-2010 17:53:23] <davetoo> Maybe I should have just dropped it and let migrate fix it :)

[13-Apr-2010 18:01:12] <davetoo> so I did a source build and wanted it to install as my own username (not zenoss), but build-functions.sh defaulted to 'zenoss'

[13-Apr-2010 18:01:38] <davetoo> how deeply embedded is that username now? :)  How can I fix that without waiting another hour for it to rebuild on this vm, I wonder

[13-Apr-2010 18:37:12] <rhettardo> i have no idea

[13-Apr-2010 20:06:19] forsberg is now known as fOrsberg

[13-Apr-2010 20:51:22] RoAk is now known as RoAkSoAx

[13-Apr-2010 20:54:32] <davetoo> ugh

[13-Apr-2010 20:54:40] <davetoo> so ... I completely broke my database;

[13-Apr-2010 20:54:49] <davetoo> I had done an RPM install;

[13-Apr-2010 20:54:59] <davetoo> how can I rebuild the database without doing the entire install over :)

[13-Apr-2010 20:55:30] <davetoo> mysql -u root -p events < zenevents.sql

[13-Apr-2010 20:55:30] <davetoo>  1212  ls

[13-Apr-2010 20:55:30] <davetoo>  1213  mysql -u root -p events < zenprocs.sql

[13-Apr-2010 20:55:37] <davetoo> I'm not sure that was sufficient

[13-Apr-2010 21:07:29] <davetoo> fixed it

[14-Apr-2010 00:00:30] [disconnected at Wed Apr 14 00:00:30 2010]

[14-Apr-2010 00:00:30] [connected at Wed Apr 14 00:00:30 2010]

[14-Apr-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[14-Apr-2010 01:50:59] fOrsberg is now known as forsberg

[14-Apr-2010 07:50:34] <cain22> morning gents

[14-Apr-2010 07:51:45] <cain22> curious, what is the messages portlet for on the dashboard...i would have thought it was for someone to display messages for all the other users..but if thats the case i'm not sure on how to use it

[14-Apr-2010 08:28:43] <cain22> curious, what is the messages portlet for on the dashboard...i would have thought it was for someone to display messages for all the other users..but if thats the case i'm not sure on how to use it

[14-Apr-2010 08:58:52] <cain22> curious, what is the messages portlet for on the dashboard...i would have thought it was for someone to display messages for all the other users..but if thats the case i'm not sure on how to use it

[14-Apr-2010 09:00:38] <ckrough> meh, never used it... not sure

[14-Apr-2010 09:01:30] * Simon4 hasn't either

[14-Apr-2010 09:12:45] <jmp242> Neither have I

[14-Apr-2010 09:13:01] <jmp242> maybe a forum post cain22

[14-Apr-2010 09:34:07] <cain22> woulda figured that was an easy one...lol...well I guess I will post it..

[14-Apr-2010 09:37:43] <rmatte> cain22: first result on google: message/38671;jsessionid=17911F540E449B64483BFFAAAFBAF852.node0

[14-Apr-2010 09:37:46] <chemist> I had the messages portlet active when I first installed zenoss, I'm quite sure it reported auto-discovery jobs etc

[14-Apr-2010 09:37:51] <rmatte> when searching for "zenoss messages portlet"

[14-Apr-2010 09:38:15] <cain22>  curious, what is the messages portlet for on the dashboard...i would have

[14-Apr-2010 09:38:16] <cain22>    thought it was for someone to display messages for all the other users..but

[14-Apr-2010 09:38:16] <cain22>    if thats the case i'm not sure on how to use it

[14-Apr-2010 09:38:42] <cain22> ooppss

[14-Apr-2010 09:38:45] <rmatte> :P

[14-Apr-2010 09:39:05] <rmatte> like I said, first result on google, google is your friend

[14-Apr-2010 09:39:06] <rmatte> ;)

[14-Apr-2010 09:39:41] <rmatte> and chemist was correct

[14-Apr-2010 09:40:34] <cain22> yeah, woulda figured someone here would know though...google can be...lacking in information sometimes

[14-Apr-2010 09:40:50] <rmatte> yeh, but you could at least put in the effort to check before asking, is what I'm saying

[14-Apr-2010 09:41:45] <cain22> ha, way to try and bring me down..

[14-Apr-2010 09:42:15] <rmatte> Not trying to bring you down, I'm simply a strong advocate of the rtfm movement

[14-Apr-2010 09:46:13] <frozty_sa> hrm

[14-Apr-2010 09:46:28] <frozty_sa> can I model/monitor devices with both SNMP and SSH concurrently?

[14-Apr-2010 09:46:40] <frozty_sa> (he asks as he digs into documentation)

[14-Apr-2010 09:47:05] <rmatte> frozty_sa: you can mix and match datasources for performance to use ssh and SNMP, but you can't mix and match for anything collected on the OS tab

[14-Apr-2010 09:47:23] <rmatte> but generally it's one or the other

[14-Apr-2010 09:47:26] <frozty_sa> rmatte: I think that should be fine

[14-Apr-2010 09:47:38] <rmatte> what exactly are you wanting to do?

[14-Apr-2010 09:48:30] <frozty_sa> what I need to do is continue collecting environment information such as interface traffic, space, etc via SNMP, and get the output of some custom commands dedicated to testing specific sorts of infrastructure via SSH

[14-Apr-2010 09:48:58] <frozty_sa> one example via SSH would be collecting the time it took for a successful auth test to run

[14-Apr-2010 09:49:49] <frozty_sa> if it's a pain, I could probably just crowbar it into SNMP with snmpd's exec functionality

[14-Apr-2010 09:50:29] <rmatte> well, you could also just create custom scripts that do what you want and output the results in nagios format, then use the scripts as command datasources

[14-Apr-2010 09:50:37] <rmatte> there's lots of different ways you can do what you want

[14-Apr-2010 09:52:17] <frozty_sa> I've seen the nagios stuff mentioned, but haven't really fiddled with it

[14-Apr-2010 09:52:23] * frozty_sa does some more reading

[14-Apr-2010 09:52:39] <frozty_sa> (<3 how extensively I can make zenoss properly use SNMP)

[14-Apr-2010 09:53:22] <rmatte> hehe

[14-Apr-2010 09:54:22] <rmatte> yeh, it's pretty nice, there are a couple things that could be better, such as allowing an snmp datasource to walk an OID tree and generate datapoints from it automatically based on what it sees, then dynamically adjust the graph accordingly.

[14-Apr-2010 09:54:39] <rmatte> but that can kind of be done via custom scripts

[14-Apr-2010 10:01:44] <frozty_sa> yeah

[14-Apr-2010 10:01:56] <frozty_sa> I was actually hitting my head against that about two weeks back

[14-Apr-2010 10:02:28] <frozty_sa> then went after the code in the os.filesystem and os.interfaces stuff to see how it works, reckon that the "parsing" happens there during modelling

[14-Apr-2010 10:03:22] <frozty_sa> now that you mention dynamically adjusting graphs..you remind me that I still need to figure out why the Holt-Winters graphs aren't being made as necessary here

[14-Apr-2010 10:03:27] * frozty_sa adds it to the list

[14-Apr-2010 10:04:56] <rmatte> hehe

[14-Apr-2010 10:05:43] <frozty_sa> it's 14 tasks long now...

[14-Apr-2010 10:05:58] <frozty_sa> and I want to get them sorted by friday so I can take leave on monday+tuesday next week..

[14-Apr-2010 10:09:07] <rmatte> cool, better get crackin'

[14-Apr-2010 10:09:33] <rmatte> any other questions regarding things on your list, while you have me?

[14-Apr-2010 10:09:54] <frozty_sa> hehe, no, that's fine thanks :)

[14-Apr-2010 10:09:58] <rmatte> cool

[14-Apr-2010 10:10:05] <frozty_sa> the list is a global todo

[14-Apr-2010 10:10:15] <rmatte> ah

[14-Apr-2010 10:10:18] <rmatte> gotcha

[14-Apr-2010 10:10:20] <frozty_sa> there's only about 3 or 4 zenoss things on it, and some of those are purely slog work

[14-Apr-2010 10:10:39] <frozty_sa> such as making a report with all the important interface graphs being shown

[14-Apr-2010 10:10:50] <rmatte> yeh, simple but tedious

[14-Apr-2010 10:10:57] <rmatte> like more than half of the stuff I do

[14-Apr-2010 10:11:13] <rmatte> I need a scripting project soon or I shall go insane

[14-Apr-2010 10:11:21] <frozty_sa> hehe

[14-Apr-2010 10:11:48] <frozty_sa> want to take the little python UDP syslogger I made and turn it into a daemon? ;-)

[14-Apr-2010 10:11:59] <rmatte> :P

[14-Apr-2010 10:12:09] <rmatte> it actually wouldn't be as hard as you think

[14-Apr-2010 10:12:35] <frozty_sa> the main thing I want to do with it is just make sure it doesn't end up blocking or something

[14-Apr-2010 10:12:40] <rmatte> does it work with Zenoss as is?

[14-Apr-2010 10:13:09] <rmatte> automagically blocking?

[14-Apr-2010 10:13:12] <frozty_sa> I've got a lot of familiarity with the ideas around Twisted and such, but unfortunately little actual experience writing code around it

[14-Apr-2010 10:13:40] <frozty_sa> rmatte: it's supposed to be used in the place of the pager functionality (yet again some crowbarring)

[14-Apr-2010 10:13:49] <rmatte> ah, I don't have any real experience with writing code around twisted... I wrote a Zenoss daemon, but it's very unorthodox... it behaves like a Zenoss daemon, but it's not written to spec

[14-Apr-2010 10:14:18] <rmatte> it works quite well however sloppy it may be hehe

[14-Apr-2010 10:14:29] <rmatte> http://dmon.org/zenticket.html

[14-Apr-2010 10:14:33] <rmatte> that's details on it

[14-Apr-2010 10:15:01] <frozty_sa> will check, ta

[14-Apr-2010 10:15:40] <frozty_sa> if I have some Spare Timeâ„¢ sometime in the future, I'd like to work on extending the zenoss notification system to make it more flexible

[14-Apr-2010 10:15:55] <frozty_sa> rmatte: python 2.4?

[14-Apr-2010 10:17:28] <rmatte> yeh

[14-Apr-2010 10:17:51] <rmatte> has to be when it's written for Zenoss

[14-Apr-2010 10:17:58] <rmatte> for the time being anyways

[14-Apr-2010 10:19:03] <frozty_sa> well, this thing that I need wouldn't need to be in the zenoss environment

[14-Apr-2010 10:19:29] <frozty_sa> all it really needs to do is listen on a sock, take everything that's written to it there and shove it out on the network towards a syslog host

[14-Apr-2010 10:19:32] <rmatte> well, you want it integrated with Zenoss though you said

[14-Apr-2010 10:19:41] <rmatte> ah

[14-Apr-2010 10:19:51] <frozty_sa> then zenoss will be shoving messages to it via the pager functionality

[14-Apr-2010 10:19:59] <frozty_sa> like I said, crowbarred ;-)

[14-Apr-2010 10:20:06] <rmatte> very much so lol

[14-Apr-2010 10:20:08] <frozty_sa> basically, it's a faster notification method than mail

[14-Apr-2010 10:20:43] <rmatte> yeh

[14-Apr-2010 10:20:52] <frozty_sa> having the jabber bot around makes it less of a priority, though

[14-Apr-2010 10:21:10] <rmatte> I don't use pager or mail... the daemon I wroten opens a ticket in our ticketing system for events

[14-Apr-2010 10:21:21] <rmatte> s/wroten/wrote/g

[14-Apr-2010 10:21:37] <frozty_sa> speaking off, you wouldn't maybe know why the xmpp bot sometimes just disconnects? even though the eventloop for it continues running, it signs off and then signs back in when it gets new events to act on

[14-Apr-2010 10:22:00] <rmatte> no idea, I've never looked at the code, you'd have to ask the author

[14-Apr-2010 10:22:06] <frozty_sa> somewhat "irritating", as it just causes some minor-ish delays sometimes with the notification response time

[14-Apr-2010 10:22:15] <rmatte> you're aware that an update for it was released fairly recently?

[14-Apr-2010 10:22:18] <frozty_sa> no problemo..will wait until I see chudler around again

[14-Apr-2010 10:22:23] <frozty_sa> rmatte: how recently?

[14-Apr-2010 10:22:29] * frozty_sa checks up on his version

[14-Apr-2010 10:22:33] <rmatte> within the last few weeks from what I recall

[14-Apr-2010 10:23:02] <rmatte> I heard him discussing it in here one day

[14-Apr-2010 10:23:13] <rmatte> maybe he didn't get around to actually doing the release though, I never checked

[14-Apr-2010 10:23:20] <frozty_sa> version 1.1 here

[14-Apr-2010 10:23:56] <rmatte> ah, that appears to be the latest one

[14-Apr-2010 10:24:04] <rmatte> yeh, have a chat with him next time you see him

[14-Apr-2010 10:24:06] <frozty_sa> not seeing an update on the page, so maybe just a dev that he's working on

[14-Apr-2010 10:24:21] <rmatte> probably

[14-Apr-2010 10:24:57] <rmatte> hmmm, looks like I'll finally have a chance to test Egor's collector pack today

[14-Apr-2010 10:24:58] <rmatte> woot

[14-Apr-2010 10:25:51] <frozty_sa> *blink*

[14-Apr-2010 10:26:23] <rmatte> I sent him my collector tool pack and he made an even better one that works via the actual UI

[14-Apr-2010 10:26:31] <rmatte> just need to do some testing with it before he can release it

[14-Apr-2010 10:27:01] <frozty_sa> ah

[14-Apr-2010 10:27:13] * frozty_sa must still get around to collectors too..

[14-Apr-2010 10:27:19] <rmatte> from what he tells me, it's quite slick

[14-Apr-2010 10:27:36] <rmatte> and I believe him, if his other work is any indication ;)

[14-Apr-2010 10:29:53] <chemist> rmatte: any eta from egor on release?

[14-Apr-2010 10:30:16] <rmatte> chemist: as soon as we're done testing it... I'm working on setting up a couple of lab boxes right now to test with

[14-Apr-2010 10:30:51] <chemist> excellent, I'd be happy to help testing it

[14-Apr-2010 10:30:55] <rmatte> chemist: probably not by the end of this week, but early to mid next week

[14-Apr-2010 10:31:08] <chemist> look forward to it

[14-Apr-2010 10:31:12] <rmatte> pass me your email address, I'll forwarding it to you

[14-Apr-2010 10:31:19] <rmatte> forward*

[14-Apr-2010 10:31:39] <rmatte> the more testers the better

[14-Apr-2010 10:32:24] <rmatte> sent

[14-Apr-2010 10:32:28] <chemist> :)

[14-Apr-2010 10:33:13] <rmatte> according to egor, you just need to install the pack, then setup public/private ssh keys so that the master box can connect to the collector as the zenoss user without having to use a password

[14-Apr-2010 10:33:24] <chemist> cool, will try it out tomorrow

[14-Apr-2010 10:33:31] <rmatte> the collector just needs a clean install of Zenoss on it

[14-Apr-2010 10:33:40] <rmatte> then you kick off the config from the Master and it handles the rest

[14-Apr-2010 10:35:07] <chemist> zenpacks aswell?

[14-Apr-2010 10:36:16] <chemist> or are they installed initially?

[14-Apr-2010 10:36:37] <rmatte> it copies them over automatically

[14-Apr-2010 10:36:45] <chemist> wicked

[14-Apr-2010 10:36:49] <rmatte> all you need on the collector is the ZenPack directories

[14-Apr-2010 10:36:56] <rmatte> the templates are handled by the master

[14-Apr-2010 10:37:17] <rmatte> (along with plugins, and whatever else that is stored in zope)

[14-Apr-2010 10:37:31] <chemist> good work

[14-Apr-2010 10:37:45] <rmatte> well, you can tell him that lol

[14-Apr-2010 10:38:01] <chemist> good work on kickstarting it :)

[14-Apr-2010 10:38:17] <rmatte> I made this, which is another way of doing it: docs/DOC-5831

[14-Apr-2010 10:38:28] <rmatte> more manual steps required, but still cuts the time down drastically

[14-Apr-2010 10:38:43] <rmatte> thanks

[14-Apr-2010 10:38:50] <rmatte> I'm glad Egor decided to code it

[14-Apr-2010 10:38:52] <chemist> that is what I need, less setup time

[14-Apr-2010 10:39:16] <rmatte> yeh, if Egor's pack isn't out by the time you need it you can use my pack in the meantime

[14-Apr-2010 10:39:29] <chemist> we have many sites and we have many idle machines at theses sites

[14-Apr-2010 10:39:55] <chemist> a quick and easy way to configure them as collectors would be ideal

[14-Apr-2010 10:40:17] <chemist> and take the strain of the master

[14-Apr-2010 10:47:08] <rmatte> yup

[14-Apr-2010 10:47:21] <rmatte> idle machines you say, send some my way ;)

[14-Apr-2010 10:47:31] <chemist> hehe

[14-Apr-2010 10:48:55] <felbane> Hello all

[14-Apr-2010 10:49:11] <felbane> Question for the zenpack wizards:

[14-Apr-2010 10:49:51] <felbane> In my __init__.py, I'm trying to import a class within my zenpack

[14-Apr-2010 10:50:19] <felbane> Like so: from ZenPacks.MyCompany.MyPack import MyClass

[14-Apr-2010 10:51:11] <felbane> I think it's failing upon install because Zenoss doesn't know of the zenpack yet (since it hasn't been installed)

[14-Apr-2010 10:52:11] <felbane> The methods I'm trying to grab can be copied out to __init__ if need be, but I hate duplication of code and was hoping I was just doing something wrong

[14-Apr-2010 10:54:12] <felbane> The error I'm getting says: ERROR: zenpack command failed. Reason: 'ZenPack' object has no attribute '__of__'

[14-Apr-2010 11:01:34] <felbane> Bah nevermind. Figured it out by looking at one of egor's packs

[14-Apr-2010 11:06:39] <rmatte> cool

[14-Apr-2010 11:10:26] <felbane> Well sorta. Now I have to figure out why I'm getting a 'resource not found'

[14-Apr-2010 11:26:34] <davetoo> hmm

[14-Apr-2010 11:26:46] <davetoo> zendmd tab completion is not like IPython's :)

[14-Apr-2010 12:17:25] <frozty_sa> also, ipython and zendmd don't like each other much.. :D

[14-Apr-2010 12:27:59] <davetoo> which is why I"m rewriting it

[14-Apr-2010 12:53:55] <davetoo> w00t!

[14-Apr-2010 12:54:04] <davetoo>     from IPython.Shell import IPShellEmbed

[14-Apr-2010 12:54:04] <davetoo>     ipshell = IPShellEmbed()

[14-Apr-2010 12:54:04] <davetoo>     ipshell.IP.user_ns = _customStuff()

[14-Apr-2010 12:54:04] <davetoo>     ipshell()

[14-Apr-2010 12:54:15] <davetoo> converted zendmd to use IPython

[14-Apr-2010 12:55:13] <davetoo> better tab completion

[14-Apr-2010 12:57:50] <mrayzenoss> davetoo: tab completion was added in 2.5

[14-Apr-2010 12:58:34] <davetoo> mrayzenoss: I know, but his is better :)

[14-Apr-2010 12:58:53] <davetoo> In [6]: dev.get

[14-Apr-2010 12:58:53] <davetoo> Display all 217 possibilities? (y or n)

[14-Apr-2010 12:59:18] <davetoo> much better

[14-Apr-2010 12:59:22] <mrayzenoss> yeah, reading through the docs now

[14-Apr-2010 12:59:29] <davetoo> there are a couple of things missing, but once I get that ironed out...

[14-Apr-2010 12:59:46] <davetoo> e.g. it's missing some functions of a plain ipython interpreter

[14-Apr-2010 13:00:22] <davetoo> when it loads the modules, it then can do tab completion on any variable/method in those libraries

[14-Apr-2010 13:00:44] <mrayzenoss> cool.  Throw that up on the wiki, might make a good tip-of-the-month candidate

[14-Apr-2010 13:01:05] <davetoo> will do

[14-Apr-2010 13:01:08] <mrayzenoss> thanks

[14-Apr-2010 13:01:14] <davetoo> man

[14-Apr-2010 13:01:17] <davetoo> you're going to like this :)

[14-Apr-2010 13:02:23] <cgibbons> so davetoo (follow up from my blog): i think you're making something harder on yourself. you should never have to edit build-functions. the zenoss user definition in there is mostly for non-source build installs. there's a couple of places where it tries to  use it but those aren't an issue for a local source build.

[14-Apr-2010 13:02:40] <davetoo> What's the best method to get a device's status in zendmd?  i.e. what's the highest-level event in the status table

[14-Apr-2010 13:02:47] <cgibbons> i'm not sure what you're doing wrong or if my instructions are wrong. if you'd like I can do a 2.4.2 build on my box and save the output to a log file for reference.

[14-Apr-2010 13:03:05] <davetoo> I may have done something out of sequence, not sure

[14-Apr-2010 13:03:07] <davetoo> was pretty tired

[14-Apr-2010 13:03:46] <davetoo> I'd be interested in just your bash shell history :)

[14-Apr-2010 13:04:28] <cgibbons> just mydevice.getStatus() will do what you need

[14-Apr-2010 13:05:01] <davetoo> thanks, was looking through the list of methods that my new zendmd returns :)

[14-Apr-2010 13:05:44] <davetoo> I have a tool that walks the device tree and applies an arbitrary lambda to each device, the lambda supplied on the commandline

[14-Apr-2010 13:05:51] <davetoo> so I can snarf devices

[14-Apr-2010 13:05:52] <cgibbons> also getStatusString, and getStatusImgSrc and some other methods should not be that class but are

[14-Apr-2010 13:06:25] <davetoo> In [3]: dev.getStatusImgSrc?

[14-Apr-2010 13:06:25] <davetoo> Type:           instancemethod

[14-Apr-2010 13:06:25] <davetoo> Base Class:     <type 'instancemethod'>

[14-Apr-2010 13:06:25] <davetoo> String Form:    <bound method Device.getStatusImgSrc of <Device at /zport/dmd/Devices/Server/Linux/detailed/devices/zenhubrtp01.rtp.netapp.com>>

[14-Apr-2010 13:06:25] <davetoo> Namespace:      Interactive

[14-Apr-2010 13:06:26] <davetoo> File:           /opt/zenoss/Products/ZenModel/EventView.py

[14-Apr-2010 13:06:26] <davetoo> Definition:     dev.getStatusImgSrc(self, status)

[14-Apr-2010 13:06:27] <davetoo> Docstring:

[14-Apr-2010 13:06:27] <davetoo>     Return the image source for a status number

[14-Apr-2010 13:06:34] <davetoo> you guys are really going to like zendmd with IPYthon :)

[14-Apr-2010 13:12:24] <rmatte> lol

[14-Apr-2010 13:23:52] <davetoo> It looks like my use of ZenScriptBase all this time has probably been overkill

[14-Apr-2010 13:24:07] <davetoo> now that I understand Python better :)

[14-Apr-2010 13:24:13] <davetoo> I can just import zendmd

[14-Apr-2010 13:33:41] tang is now known as dgc

[14-Apr-2010 13:34:03] dgc is now known as tang

[14-Apr-2010 13:43:36] <cgibbons> davetoo: 2.4.2, right?

[14-Apr-2010 13:43:48] <davetoo> yes

[14-Apr-2010 13:53:22] <cgibbons> ok building/logging now

[14-Apr-2010 15:01:41] <blull> Hey

[14-Apr-2010 15:01:46] <blull> How do I monitor a generic tcp service?

[14-Apr-2010 15:01:50] <blull> ie, port 22 tcp?

[14-Apr-2010 15:01:52] <blull> with zenoss core

[14-Apr-2010 15:03:01] <jmp242> You should be able to add it as an IP service

[14-Apr-2010 15:05:29] <blull> okay

[14-Apr-2010 15:05:40] <blull> I added a new service called ssh

[14-Apr-2010 15:05:55] <blull> i then set monitor to true under Edit tab

[14-Apr-2010 15:05:59] <blull> what are Service Keys?

[14-Apr-2010 15:09:47] <jmp242> Hmmm, I don't know

[14-Apr-2010 15:09:55] <jmp242> haven't actually done this myself

[14-Apr-2010 15:13:07] <blull> okay

[14-Apr-2010 15:13:14] <blull> how do I actually monitor the service now that I added it?

[14-Apr-2010 15:13:16] <blull> via a template?

[14-Apr-2010 15:15:56] <bigegor> blull: http://yourserver:8080/zport/dmd/Services/IpService/Privileged/serviceclasses/ssh/ipServiceClassEdit

[14-Apr-2010 15:16:22] <bigegor> change monitor to True

[14-Apr-2010 15:16:34] <bigegor> and remodel your server

[14-Apr-2010 15:17:24] <blull> oh nice

[14-Apr-2010 15:17:25] <blull> k

[14-Apr-2010 15:17:37] <blull> udp_00022 is there?

[14-Apr-2010 15:17:40] <blull> i thought ssh was only tcp?

[14-Apr-2010 15:18:23] <blull> so services are automatically modeled then?

[14-Apr-2010 15:18:30] <bigegor> yes

[14-Apr-2010 15:18:37] <blull> well, heres the thing

[14-Apr-2010 15:18:46] <blull> i don't want to model on a per device basis as we have thousands all over the place

[14-Apr-2010 15:19:04] <blull> and i only want certain collections to be modeled to use ssh, while others should use 3389, 443, etc.

[14-Apr-2010 15:19:05] <bigegor> in zCollectorPlugins list you can use zenoss.snmp.IpServiceMap or zenoss.portscan.IpServiceMap

[14-Apr-2010 15:19:22] <blull> how do I set the frequency of checks on there?

[14-Apr-2010 15:19:36] <blull> cause some of the systems I need 3 second visibility into the http 3 way handshake.

[14-Apr-2010 15:20:35] <blull> hrmm

[14-Apr-2010 15:20:47] <blull> there is no zenoss.tcp.IpServiceMap?

[14-Apr-2010 15:20:59] <blull> ie. zenoss.snmp.IpServiceMap I'm assuming only uses snmp?

[14-Apr-2010 15:24:54] <bigegor> zenoss.snmp.IpServiceMap only reads tcptable and udptable OIDs.

[14-Apr-2010 15:25:19] <blull> how do I do tcp only 3 way handshake tests than?

[14-Apr-2010 15:26:01] <blull> for all hosts inside a class?

[14-Apr-2010 15:27:27] <blull> wow

[14-Apr-2010 15:27:31] <blull> i have to do it on a per host basis

[14-Apr-2010 15:27:33] <blull> that fucking sucks.

[14-Apr-2010 15:27:37] <blull> or I have to port scan my hosts, lame.

[14-Apr-2010 15:27:48] <blull> otherwise I have to write a zenpack for it :/

[14-Apr-2010 15:28:53] <bigegor> if you set monitor to True for  IpService it will be enabled for all devices

[14-Apr-2010 15:30:32] <rmatte> yeh, just go in to Services on the left, find the services you want and set enabled to true, then when you remodel devices if the service is available on the device it will automatically be monitored

[14-Apr-2010 15:31:24] <blull> rmm

[14-Apr-2010 15:31:31] <blull> how do I kick off a class wide remodel?

[14-Apr-2010 15:31:38] <blull> i need to remodel an entire dc heh

[14-Apr-2010 15:32:37] <Simon4_> zenmodeler run -p /Device/Class

[14-Apr-2010 15:32:42] <bigegor> you need to wait 720 mins ;)

[14-Apr-2010 15:32:44] <blull> heh

[14-Apr-2010 15:32:50] <blull> i can't do it through the webui?

[14-Apr-2010 15:32:53] <Simon4_> damm, bigegor's answer was better :)

[14-Apr-2010 15:33:02] tang is now known as dgc

[14-Apr-2010 15:33:02] <blull> i refuse to wait 720 minutes :)

[14-Apr-2010 15:33:13] <rmatte> no you can't do it through the webui

[14-Apr-2010 15:33:14] <Simon4_> blull: sure, go to each device, click down the menus and select "model device"

[14-Apr-2010 15:33:31] <rmatte> but I don't see what difference it makes

[14-Apr-2010 15:33:49] <rmatte> unless you're allergic to commandline or something :P

[14-Apr-2010 15:34:28] <Simon4_> rmatte: it doesn't give you the shakes? ;)

[14-Apr-2010 15:34:35] <rmatte> not generally

[14-Apr-2010 15:35:48] <chudler> ui access does not imply system access

[14-Apr-2010 15:36:07] * Simon4_ imagines a slightly scary "commandline" zenpack

[14-Apr-2010 15:36:49] <blull> im not alergic to command line

[14-Apr-2010 15:36:53] <blull> the purpose of this system is to offer a ui.

[14-Apr-2010 15:37:02] <blull> and we have different levels of access allowed to it.

[14-Apr-2010 15:37:12] <blull> I don't want a noc monkey mucking around on prod systems.

[14-Apr-2010 15:37:20] <blull> I want them to click a button in a web interface :)

[14-Apr-2010 15:37:30] <Simon4_> blull: sure, but they shouldn't need to bulk remodel things ad-hoc

[14-Apr-2010 15:37:47] <blull> no they shouldn't, nor should they.

[14-Apr-2010 15:37:56] <blull> However if we push out a new process enterprise wide for example

[14-Apr-2010 15:38:02] <blull> we will have to wait or remodel correct?

[14-Apr-2010 15:38:12] <blull> hrmphf

[14-Apr-2010 15:38:30] <blull> Zenoss needs a dev who's worked in a environment with more systems then there are people in some countries :)

[14-Apr-2010 15:39:13] <blull> all about roi and time invested.. I shouldn't have to take the time to bounce through bastion hosts to run a cli command when it should be a 1 click thing as I already have the connectivity and acl's in place :)

[14-Apr-2010 15:39:27] <blull> just saying hehe

[14-Apr-2010 15:39:30] <Simon4_> sounds like a good basis for a zenpack :)

[14-Apr-2010 15:39:37] <blull> nah

[14-Apr-2010 15:39:38] <blull> not a pack.

[14-Apr-2010 15:40:54] <blull> wait

[14-Apr-2010 15:40:55] <blull> i just had a thought

[14-Apr-2010 15:40:59] <blull> after I remodel these devices.

[14-Apr-2010 15:41:04] <blull> and they pickup that ssh and http are running

[14-Apr-2010 15:41:21] <blull> how do I kick off a tcp port check to that service every 3 seconds if I need to?

[14-Apr-2010 15:44:44] <bigegor> 'Status Cycle Interval' in collector config

[14-Apr-2010 16:08:23] <davetoo> In zendmd, how can I determine the highest-severity active event for a device?

[14-Apr-2010 16:12:01] <bigegor> davetoo: dev.ZenEventManager.getMaxSeverity(dev)

[14-Apr-2010 16:12:46] <davetoo> thanks

[14-Apr-2010 16:12:57] <bigegor> np

[14-Apr-2010 16:15:33] <davetoo> dev.getEventManager()

[14-Apr-2010 16:16:54] <bigegor> ???

[14-Apr-2010 16:20:12] <bigegor> davetoo: dmd.ZenEventManager.getMaxSeverity(device) also work

[14-Apr-2010 16:24:12] <davetoo> yep

[14-Apr-2010 16:24:28] <davetoo> The way I'm using it I need to get it from the device

[14-Apr-2010 16:24:42] <tpdd> is it possible to triggger an alert based on a windows login record in the event log?

[14-Apr-2010 16:24:59] <mrayzenoss> if that shows up in the event console, yes

[14-Apr-2010 16:26:54] <bigegor> davetoo: dmd.ZenEventManager.getMaxSeverity(find('yourdevicename'))

[14-Apr-2010 16:31:31] <tpdd> can you point me to doc that has an example of setting up a new event?

[14-Apr-2010 16:32:08] <mrayzenoss> tpdd: do you see the event in the console?  mapped to /Unknown or something?

[14-Apr-2010 16:35:09] <tpdd> no my mistake i can see it in the event log on the windows box im monitoring but not in the event log

[14-Apr-2010 16:36:19] <blull> hrmm

[14-Apr-2010 16:36:20] <blull> okay

[14-Apr-2010 16:36:22] <blull> your idea isn't working

[14-Apr-2010 16:36:30] <blull> I enabled ssh, http, and https IpServices

[14-Apr-2010 16:36:37] <blull> i then went to /Devices/My Datacenter

[14-Apr-2010 16:36:45] <blull> added the portscan plugin to the collector

[14-Apr-2010 16:36:52] <blull> I went to the collector and told it to set the cyclet imeout to 5 seconds

[14-Apr-2010 16:37:12] <blull> and now I'm running  bin/zenmodeler on -p /Devices/My Datacenter --now

[14-Apr-2010 16:37:25] <blull> not seeing ANY of the few hundred systems showing http, ssh, or https in use :/

[14-Apr-2010 16:37:29] <blull> any ideas what I missed?

[14-Apr-2010 16:41:22] <bigegor> blull: try first with one device: zenmodeler run -d yourserver

[14-Apr-2010 16:57:14] <bigegor> blull:you can try also: zenmodeler run -p '/Devices/My Datacenter' --collect=zenoss.portscan.IpServiceMap

[14-Apr-2010 16:59:11] <mrayzenoss> New Xen Virtual Machines Report: docs/DOC-5834

[14-Apr-2010 17:00:32] <Simon4_> mrayzenoss: heya... do you know if a "table" like in that report can be sorted by more than one field?

[14-Apr-2010 17:01:05] * Simon4_ has a bladechassis one that would really like to be sorted by chassis, then position

[14-Apr-2010 17:01:22] <mrayzenoss> Simon4_: yeah, that report can be sorted by any of the columns

[14-Apr-2010 17:01:28] <mrayzenoss> I'm clicking on it now :)

[14-Apr-2010 17:01:37] <Simon4_> mrayzenoss: sorry... not by any

[14-Apr-2010 17:01:43] <Simon4_> but by this, AND this

[14-Apr-2010 17:01:52] <mrayzenoss> ahh… dunno

[14-Apr-2010 17:01:52] <Simon4_> so you get blades in order, and chassis grouped

[14-Apr-2010 17:01:58] <mrayzenoss> definitely more likely in the new UI

[14-Apr-2010 17:02:13] * Simon4_ has been trying to solve this, and it hasn't been easy :)

[14-Apr-2010 17:02:17] <davetoo> so in the new Event console/list, where's the old Detail view that used to be in a popup?

[14-Apr-2010 17:05:54] <mrayzenoss> davetoo: double click on the event

[14-Apr-2010 17:06:11] <mrayzenoss> Simon4_: the word from dev is it's not supported by the JS widget we're using

[14-Apr-2010 17:06:36] <mrayzenoss> Simon4_: but… you could have the query pre-sorted so the first view would work

[14-Apr-2010 17:06:46] <Simon4_> mrayzenoss: crap... I was thinking as a workaround to have a table per blade chassis, but that fails paginating

[14-Apr-2010 17:07:14] <mrayzenoss> Simon4_: well it wouldn't be stored in SQL anyway right?

[14-Apr-2010 17:07:19] <Simon4_> mrayzenoss: yeah, I guess if you model blades in order they should stay that way

[14-Apr-2010 17:07:31] * mrayzenoss brainfart table != SQL table

[14-Apr-2010 17:07:44] <Simon4_> hang on

[14-Apr-2010 17:07:48] <Simon4_> zopedb = sql?

[14-Apr-2010 17:07:51] <Simon4_> :)

[14-Apr-2010 17:07:51] <mrayzenoss> no

[14-Apr-2010 17:07:56] <Simon4_> yeah

[14-Apr-2010 17:08:57] <Simon4_> so you're saying that if I model stuff in order, if I order by chassis they should stay in order?

[14-Apr-2010 17:09:12] <mrayzenoss> that would be my assumption

[14-Apr-2010 17:09:29] <mrayzenoss> but I could see all sorts of ways the ordering would get hosed

[14-Apr-2010 17:09:40] <Simon4_> yeah

[14-Apr-2010 17:10:13] <mrayzenoss> of course, if you're really eager, you could patch ExtJS to do what you want

[14-Apr-2010 17:10:18] <Simon4_> :D

[14-Apr-2010 17:11:05] * Simon4_ is on 3.5 weeks holiday, maybe after that

[14-Apr-2010 17:11:35] <mrayzenoss> I can think of many things I'd rather do on vacation

[14-Apr-2010 17:11:50] <bigegor> Simon4_: or you can set comtonent.id to something like "chassisid_bladeid"

[14-Apr-2010 17:12:24] <bigegor> component.id*

[14-Apr-2010 17:13:08] <Simon4_> bigegor: then sort by that?

[14-Apr-2010 17:13:18] * Simon4_ likes this

[14-Apr-2010 17:13:22] <bigegor> by id

[14-Apr-2010 17:19:54] <Simon4_> that would work well

[14-Apr-2010 17:20:15] <Simon4_> thanks :)

[14-Apr-2010 17:20:20] <bigegor> np

[14-Apr-2010 17:20:23] <mrayzenoss> yeah, good idea

[14-Apr-2010 17:38:38] <davetoo> hmm

[14-Apr-2010 19:15:45] <felbane> hello everybody

[14-Apr-2010 19:22:37] <felbane> So, I'm trying to figure out how to implement a relatively complex ZenPack (I think). So far I've been able to work out how to extend the 'os' tab with another table, populate that tab via a WMI query, and allow "user created" entries. What I'm having trouble working out is how to set it up to poll the items and generate RRDs.

[14-Apr-2010 19:25:43] <felbane> Essentially briging the gap between "device/queues/(list of queues)" and "gather lengths of (list of queues) regularly and update this RRD"

[14-Apr-2010 19:26:05] <felbane> I've been at this zenpack all day, my brain's kind of aching

[14-Apr-2010 19:45:30] forsberg is now known as fOrsberg

[14-Apr-2010 19:50:58] <felbane> bah laptop died

[14-Apr-2010 19:51:13] <felbane> if anyone responded to previous question, please re-send?

[15-Apr-2010 00:00:30] [disconnected at Thu Apr 15 00:00:30 2010]

[15-Apr-2010 00:00:30] [connected at Thu Apr 15 00:00:30 2010]

[15-Apr-2010 00:00:45] [zenoss-logger (logger bot) has joined #zenoss]

[15-Apr-2010 00:00:45] <calvino.freenode.net> [freenode-info] channel flooding and no channel staff around to help? Please check with freenode support: http://freenode.net/faq.shtml#gettinghelp

[15-Apr-2010 04:17:56] fOrsberg is now known as forsberg

[15-Apr-2010 06:09:14] <blull> morning

[15-Apr-2010 06:10:44] <blull> what is the difference between zenoss.portscan.IpServiceMap and zenoss.nmap.IpServiceMap?  Just the method of the portscan (builtin vs. external)?

[15-Apr-2010 06:12:24] <blull> also I modified an entire class of devices (with lots of sub classes) to use the zenoss.portscan.ipservicemap plugin, kicked off a "zenmodeler run --now -p /Devices/Datacenter --hubhost=zenoss-hub --monitor=my-monitor" so it would remodel all devices in that datacenter and pick up the ipservices I enabled, however it missed A LOT of them :/

[15-Apr-2010 06:12:42] <blull> only picked up about 10 or so systems instead of the 50+ it should have picked up for 1 specific port.

[15-Apr-2010 08:00:05] <Jesper> hi guys. Could someone here please point me in the right direction!   I have a Zenoss box that monitors a lot of routers, and I would like to change the performance template for all of them.

[15-Apr-2010 08:00:19] <Jesper> How do I go about this (no, using the web frontend is not an option...). Is it possible to do through commandline, or make a zenpack, or something like that?

[15-Apr-2010 08:00:42] <ckrough> what do you mean by "change the performance template"

[15-Apr-2010 08:00:44] <mrayzenoss> zendmd is how you'd do it non-UI

[15-Apr-2010 08:00:49] <mrayzenoss> but what are you trying to do?

[15-Apr-2010 08:01:23] <forsberg> and why not through ui :)

[15-Apr-2010 08:01:27] <Jesper> split the template up in in and out going traffic. For an ADSL connection that is eg. 8Mbit/512Kbit

[15-Apr-2010 08:01:37] <Jesper> forsberg: 2000+ routers...

[15-Apr-2010 08:01:59] <ckrough> can you just make the change to the whole class? rather than using local copies?

[15-Apr-2010 08:02:04] <forsberg> but you change the template of the device class. not every single router

[15-Apr-2010 08:02:31] <forsberg> ...

[15-Apr-2010 08:02:49] <Jesper> oh ok, nice. Did not know that. But the next problem is probably the speed. There's a lot of different speeds on the various routers/connections

[15-Apr-2010 08:03:00] <Jesper> is it possible to make different classes with different speeds?

[15-Apr-2010 08:03:01] <tehhobbit> anyone that could lend som ideas, I want to monitor several websites using httpmonitor (vhosts) but I cant add them since there is a machine with the same ip ( the physical machine where I do cpu and hw monitoring)

[15-Apr-2010 08:03:23] <ckrough> Jesper: you can use variables in the thresholds, if thats what you are worried about. ie: 80% of max. etc...

[15-Apr-2010 08:03:38] <tehhobbit> can I somehow make zenoss ignore the fact that there will be more than 1 ?

[15-Apr-2010 08:03:54] <ckrough> Jesper: that way the modeler takes care of detecting the speeds and the template works for all

[15-Apr-2010 08:04:23] <forsberg> tehhobbit people/bigegor/blog/2010/02/22/virtualip-zenpack and http://nettikconsulting.wordpress.com/2010/03/19/zenoss-more-than-one-device-with-a-single-ip/ was the links i found on that topic

[15-Apr-2010 08:04:43] <forsberg> or maybe the synthetic transaction pack can help you

[15-Apr-2010 08:05:03] <Jesper> ckrough: nope, it's not the problem. The problem is that there is only one speed definition, and it's set to eg. 100Mbit (because of the speed on the Ethernet interface). I can't set it to both in-speed and out-speed.

[15-Apr-2010 08:05:04] <forsberg> i heard in here that new version of zenoss will be able to manage more devices with same ip

[15-Apr-2010 08:05:20] <ckrough> Jesper: I see

[15-Apr-2010 08:06:20] otakup0p1 is now known as otakup0pe

[15-Apr-2010 08:06:51] <Jesper> But if I'm understanding this correct. It's possible to make a specific class, with let's say 8M/512K and make a template that fits that. And another class with eg. 10M/768K.  Or?

[15-Apr-2010 08:10:16] <forsberg> indeed

[15-Apr-2010 08:11:08] <Jesper> ok cool, I'll have a look at that. Thanks

[15-Apr-2010 08:16:08] <tehhobbit> second link was kind of a hack but cleaner

[15-Apr-2010 08:17:16] <ckrough> Jesper: I *think* you could also write a modeler plugin that correctly 'types' each speed interface, then you can create templates that match that interface type

[15-Apr-2010 08:18:26] <ckrough> Jesper: the same way that it tags interfaces as 'ethernetCsmacd' and templates match that

[15-Apr-2010 08:20:34] <forsberg> tehhobbit yes its a "hack"

[15-Apr-2010 08:21:00] <forsberg> it really depends on your need

[15-Apr-2010 08:21:14] <forsberg> im not gonna use the hack, because i dont want to maintain/fix stuff when i upgrade etc

[15-Apr-2010 08:21:37] <Jesper> ckrough: ok, sounds about right. Not sure how to go about that though. :-/

[15-Apr-2010 08:22:34] <ckrough> Jesper: unfortunately I cant point you in the right direction on that one, but today is developer chat day... so score!

[15-Apr-2010 08:27:24] <Jesper> ET = Eastern Time (New York, etc.), right?

[15-Apr-2010 08:27:30] <mrayzenoss> yup

[15-Apr-2010 08:28:20] <Jesper> in about 1½ hours?

[15-Apr-2010 08:30:43] <Jesper> right?

[15-Apr-2010 08:32:09] <mrayzenoss> 1.5 hours

[15-Apr-2010 08:32:54] <Jesper> Thanks.    ½ is more fun, when you've got it on the keyboard. ;-)

[15-Apr-2010 09:30:22] <sektorNBA> hi

[15-Apr-2010 09:30:54] <sektorNBA> anyone use zenoss monitoring mikrotik? I cant have cpu, mem graphs working at zenoss.

[15-Apr-2010 09:51:21] <bedwards> I'm a Zenoss Developer

[15-Apr-2010 09:51:42] <themactech> Hi, can I ask you a few questions?

[15-Apr-2010 09:52:54] <bedwards> yes

[15-Apr-2010 09:54:17] <themactech> I have some questions regarding types of data, I have made some custom zProperties, using zendmd, with devices._setProperty('zWarExp','',type='date') as an example

[15-Apr-2010 09:54:42] <themactech> now how is such a property different from say, the comment field

[15-Apr-2010 09:55:24] <themactech> I have set these custom properties at the root device location, as I was made to understand that is the only place you can create custom zProperties

[15-Apr-2010 09:56:07] <themactech> they take the default values when that was specified in the zendmd command, but I can't seem to overwrite the values at lower branches in the device tree, it gives me an error message

[15-Apr-2010 09:56:31] <bedwards> what is the message?

[15-Apr-2010 09:56:39] <themactech> some datetime error

[15-Apr-2010 09:57:39] <themactech> also, i have written custom modelers, that would say map and OID value to the 'comments' field, and I can't seem to map that data to my custom zProperties

[15-Apr-2010 09:59:38] <themactech> I have a zProperty called zCustom1, and when I try to map a value to it from a collector plugin with snmpGetMap = GetMap '.1.3.6.x.x.x.x.x' : 'zCustom1' it does not work, though it works fine if I put 'comment' field, then that value goes into the comment field with no issue

[15-Apr-2010 10:02:32] <mrayzenoss> themactech: don't you have a forum post that talks about what you're trying to do?

[15-Apr-2010 10:02:44] <themactech> Yes I do

[15-Apr-2010 10:06:07] <Jesper> question. (I talked a bit about this earlier today) I have 2000+ routers with various Internet connection speeds, and I would like to I'm trying to set up my Zenoss installation to support speeds on ADSL connections. What's the easiest/best way to go about this?

[15-Apr-2010 10:06:58] <Jesper> I've tried to make a subclass with routers that has eg. a 4Mbit/768Kbit connection, but I can't quite seem to figure out how to add a correct template to it.

[15-Apr-2010 10:07:50] <Jesper> that is, other than making a local copy. And that's not really a viable approach on more than a few routers.

[15-Apr-2010 10:07:51] <bedwards> themactech, we don't have any date zProps so you are in uncharted territory.  Maybe you need a Zope date instead of a datetime date.

[15-Apr-2010 10:08:25] <themactech> I have the same problem with my other zProperties, one is a integer and the others are strings

[15-Apr-2010 10:08:44] <themactech> and what is the difference between an added zProperty, and a custom schema, or are they the same

[15-Apr-2010 10:09:03] <cgibbons> do any of our plugins actually set/modify zProperties via ApplyDataMap, Brian? I'm trying to think of an example.

[15-Apr-2010 10:09:04] <themactech> and are these different from the other fields that show in the status tab, i.e. comment, tag #, serial #

[15-Apr-2010 10:09:20] <cgibbons> Looking at the code I'm not sure it would work right either.

[15-Apr-2010 10:09:39] <themactech> What I'm trying to do is make some custom fields to show device-personalized data in the status tab

[15-Apr-2010 10:09:58] <bedwards> I don't know of any.  I don't think ApplyDataMap makes a setZenProperty call

[15-Apr-2010 10:10:18] <cgibbons> looks like it does an aq_base(obj) when it tries to get the attr, too

[15-Apr-2010 10:10:40] <themactech> for example the promise raid boxes I need to monitor have 2 controllers, having the same firmware on both is critical so I want to show the firmware version of both controllers in the status tab and maybe even set a threshold if both are not identical, which can happen when you swap out a defective

[15-Apr-2010 10:11:28] <bedwards> You should subclass Device and add a new attribute.

[15-Apr-2010 10:11:42] <themactech> if a custom zProperty is not the way to go, what are my options, are their any differences vs a zProperty and a custom schema?

[15-Apr-2010 10:12:17] <themactech> the zWarExp was for me to try and implement warranty expiration tracking and alerts for gear

[15-Apr-2010 10:13:05] <themactech> usually if you let a warranty expire on a piece of gear you cannot renew it, and this happens a lot and clients are notorious for not tracking this properly, we would implement this in zenoss if possible

[15-Apr-2010 10:13:06] <bedwards> basically zProperty might be acquired, an attribute will always be directly on the instance.  I don't think that custom schema is what you want either.  I'm looking for an example.

[15-Apr-2010 10:14:54] <themactech> how do I add an attribute, and how is an attribute different from a zProperty, also keep in mind I want to set these with custom modelers, I have a custom modeler for Apple devices that will get the serial number via SNMP and query Apple's website to pull the warranty expiration, amongst other things

[15-Apr-2010 10:15:28] <themactech> Dont want to have to input these manually for large apple deployments

[15-Apr-2010 10:16:00] <themactech> also if you renew your warranty, your warranty expiration date would be automatically update as Apple updates their database, same can be done with Dell

[15-Apr-2010 10:17:57] slank is now known as slank_

[15-Apr-2010 10:18:12] slank_ is now known as slank

[15-Apr-2010 10:19:24] <bedwards> check out EC2Manager in the ZenAWS zenpack.  It subclasses Device and adds an access_id attribute.

[15-Apr-2010 10:19:38] <themactech> is that available for core?

[15-Apr-2010 10:19:42] <themactech> i'm using core

[15-Apr-2010 10:19:52] <mrayzenoss> yeah, it's the AWS ZenPack

[15-Apr-2010 10:19:57] <themactech> alright will do

[15-Apr-2010 10:21:09] <themactech> now when I want to modify the UI skin for this data, I know how to insert a <dt>data here</dt> table, and russell at the community day in austin told me that the text in the <dt></dt> could hold a zenoss variable, but I can't find the syntax to do so

[15-Apr-2010 10:21:52] <themactech> because what I am trying to do is create say 2 custom fields, called custom1 and custom2, and have 2 other fields called customheader1 customheader2

[15-Apr-2010 10:22:17] <themactech> and have the <dt></dt> table display the value in customheader1 instead of static text

[15-Apr-2010 10:22:57] <bedwards> Jesper, are you writing a new zenpack for a type of device that isn't currently supported?

[15-Apr-2010 10:23:51] <themactech> this way for a class of devices, for example my promise raids, it would display "FW CTRL 1" and then the firmware, but for another device class, I could set the customheader1 value to "codec license" and then it would display "codec license" and for example mpeg, avi as the value

[15-Apr-2010 10:24:18] <mrayzenoss> themactech: would it be easier just to stuff everything into zLinks?

[15-Apr-2010 10:24:26] <mrayzenoss> wouldn't have to edit the apges

[15-Apr-2010 10:24:31] <mrayzenoss> s/apges/pages/

[15-Apr-2010 10:24:58] <themactech> but then setting threshold would be complicated no? I would have to parse thru the zlink field to extract actual data

[15-Apr-2010 10:25:12] <mrayzenoss> threshold?

[15-Apr-2010 10:25:50] <themactech> yes, if I set the custom1 field to hold firmware of controller1, and custom2 to hold firmware of controller2, i want to be able to test if both are equal and alert if not

[15-Apr-2010 10:26:16] <themactech> so if a client replaces a controller and it shipped from Promise Inc with older firmware and client does not check it, I will know and warn him

[15-Apr-2010 10:26:29] <themactech> also I do want alerts tied to the warranty expiration

[15-Apr-2010 10:26:53] <themactech> I would really like to make a nice zenpack that integrates warranty status tracking into zenoss

[15-Apr-2010 10:26:55] <mrayzenoss> ahh, I just meant for presenting the data on the status page, not for alerting on it

[15-Apr-2010 10:27:02] <themactech> our clients would appreciate that feature

[15-Apr-2010 10:27:21] <mrayzenoss> I know I've seen people link to their Dell pages by the service tag

[15-Apr-2010 10:27:22] <Jane_Curry> hiya

[15-Apr-2010 10:27:23] <bedwards> themactech, check out ec2ManagerStatus.pt.  It displays the access_id attribute.

[15-Apr-2010 10:27:31] <bedwards> hello Jane

[15-Apr-2010 10:27:38] <themactech> oh, well i can already display data with a custom schema, and have already successfully modified the skin to show it

[15-Apr-2010 10:28:05] <themactech> I just don't know how to set that custom schema from a modeler plugin

[15-Apr-2010 10:29:01] <themactech> as for what russell told me in austin last week, can the text in a <dt></dt> be dynamic with a zenoss variable?

[15-Apr-2010 10:29:09] <themactech> or was that info wrong?

[15-Apr-2010 10:29:35] <Jane_Curry> Does anyone know what the collector parameter "Maximum ping failures" actually does in 2.5.1 / 2.5.2?

[15-Apr-2010 10:29:46] <themactech> I've tried about every combo of syntax i.e. <dt>here/zCustom1</dt>

[15-Apr-2010 10:30:03] <Jane_Curry> Devices definitely still get pinged after they have been down after this number of failed pings

[15-Apr-2010 10:30:29] <Jesper> bedwards: no, I don't think so. It's a Cisco router that get detected just fine and everything. The problem is the WAN connection, it's detected as 100Mbit (Ethernet link speed) and it is in fact just 4Mbit/768Kbit (ADSL). I want to make a "high utilization" template for it, that supports both up- and download. I can do that on one unit, but I can't figure out how to make it as a template for multiple routers/units.

[15-Apr-2010 10:31:07] <themactech> make your template and bind it to the device class

[15-Apr-2010 10:31:17] <Jesper> I'm uncertain if it's a ZenPack I need to make, or if it's a Template, or....

[15-Apr-2010 10:31:18] <themactech> then put all these devices in that class

[15-Apr-2010 10:31:29] <Jesper> ok

[15-Apr-2010 10:32:34] <themactech> that is something that threw me off as well when i was testing, I made a template directly in the subclass I wanted it to apply to, and it ONLY showed in that subclass, yet it would not apply to it until i chose Bind template and bound it to the device class i wanted

[15-Apr-2010 10:33:08] <themactech> that made no sense to me as I figured if I made the template IN the class it should have been bound to it by default but its not the case

[15-Apr-2010 10:33:42] <themactech> so go to /devices/routes/jasper's_kick_ass_routes/ add a template with name of your choise

[15-Apr-2010 10:33:44] <themactech> choice

[15-Apr-2010 10:34:00] <bedwards> we are working on making template definition/binding clearer in the UI for the next release

[15-Apr-2010 10:34:00] <jplouis> themactech: unfortunately you can't modify custom schema values or zProperties from a modeler plugin. bedwards was correct, you will have to subclass Device to add a new type

[15-Apr-2010 10:34:09] <themactech> do all your settings datapoints, graphs, thresholds, etc, when done, bind it to that class

[15-Apr-2010 10:34:46] <themactech> alright thanks, I will look into it, now I can still show those attributes in the status page?

[15-Apr-2010 10:35:14] <bedwards> you have to override the default status page like the AWS zenpack does

[15-Apr-2010 10:35:18] <jplouis> yes, if you create you own version of the status page

[15-Apr-2010 10:35:21] <themactech> ok

[15-Apr-2010 10:35:49] <themactech> now, for the cherry on top, can the text in a <dt></dt> entry contain dynamic zenoss data

[15-Apr-2010 10:36:05] <themactech> or only static text?

[15-Apr-2010 10:36:30] <jplouis> I'm not sure what you are referring to, don't know what <dt> is

[15-Apr-2010 10:36:46] <Brixius> I think you have to goto zProperties/zDeviceTemplates of the device class and add it in there.

[15-Apr-2010 10:36:48] <themactech> when you modify the skin of the status page to show a new field

[15-Apr-2010 10:37:05] <themactech> you have to add a dt table

[15-Apr-2010 10:37:18] <rocket> page templates in general have access to some dynamic data via tales expressions etc

[15-Apr-2010 10:37:51] <themactech> that's what I was told and it works fine for the DATA of the field, however I can't seem to have the HEADER of the field contain such data

[15-Apr-2010 10:37:58] <themactech> (dt)TestFoo(/dt)

[15-Apr-2010 10:37:58] <themactech> (dd)

[15-Apr-2010 10:37:59] <themactech> (div tal:content="here/cTestFoo"/>

[15-Apr-2010 10:37:59] <themactech> (/dd)

[15-Apr-2010 10:38:04] <themactech> this is an example

[15-Apr-2010 10:38:34] <themactech> what I want is to make (dt)TestFoo(/dt) use here/insertzenossvarhere

[15-Apr-2010 10:38:46] <Jesper> themactech: ok, thank you very much. Gotta run now, but I'll have a look at that later.    and bedwards, sounds good about the UI in the next release.

[15-Apr-2010 10:39:07] <themactech> in the (dd)(/dd) section this works fine but I can't seem to do it in the (dt)(/dt)

[15-Apr-2010 10:39:17] Jesper is now known as JesperAway

[15-Apr-2010 10:40:36] <themactech> that would allow me to have custom fields, that I can set the header for, so the layout in the status tab would dynamically change to specifically taylor to a device class

[15-Apr-2010 10:41:34] <themactech> I want to show firmware version for both controllers for my Promise boxes, set customheader1 and customheader2 to "FW CTRL 1" and "FW CTRL 2" and voilà ! my status page is customized for my promise boxes

[15-Apr-2010 10:42:42] <jplouis> does <dt tal:content="here/foo"> not work?

[15-Apr-2010 10:42:45] <themactech> for another device class, say my Vbrick streaming devices, I need to show what video codecs are registered on that device, so I set the customheader1 and 2 to what I need and presto! status tab is customized for my vbricks, all of this with only one custom status page

[15-Apr-2010 10:42:53] <rocket> themactech: can you do <dt tal:content="here/foo"> ?

[15-Apr-2010 10:42:58] <rocket> jplouis: you beat me to it :p

[15-Apr-2010 10:43:19] <themactech> I think I tried that but I give it another go, my zenoss virtualbox is firing up now

[15-Apr-2010 10:46:22] <themactech> OMG it works, you guys rock, customized status pages PER device class are now mine!!! MUHAHAHAHAHA

[15-Apr-2010 10:47:05] <themactech> I had the syntax wrong before, I'm not very well versed in HTML

[15-Apr-2010 10:47:10] <themactech> thanks a bunch

[15-Apr-2010 10:47:14] <jplouis> dt tags are just html tags, nothing special.  You can do the tal:content for any html tag

[15-Apr-2010 10:47:15] <jplouis> NP

[15-Apr-2010 10:47:59] <themactech> this IRC thing is a weekly event right?

[15-Apr-2010 10:48:04] <mrayzenoss> biweekly

[15-Apr-2010 10:48:06] <jplouis> 2 weeks

[15-Apr-2010 10:48:09] <themactech> perfect

[15-Apr-2010 10:48:13] <mrayzenoss> themactech: looking forward to any ZenPacks you want to send in :)

[15-Apr-2010 10:48:30] <themactech> I've gotten more answers in 5 minutes then in 3 months of posting in the forums

[15-Apr-2010 10:49:03] <bedwards> The dt is not the issue here, Dude. I'm talking about drawing a line in the sand, Dude. Across this line, you DO NOT...

[15-Apr-2010 10:49:18] <themactech> Well, I can't give away the Apple secret sauce but the warranty tracking and the custom field per class is something I will be able to post in the community

[15-Apr-2010 10:49:36] <themactech> if I get it all done right

[15-Apr-2010 10:50:52] <Brixius> themactech: Wouldn't that be AppleSauce... What's the secret... lol

[15-Apr-2010 10:51:11] <Brixius> should say, where's the secret...

[15-Apr-2010 10:51:15] <themactech> I am working on templates for very detailed monitoring of Apple xsans

[15-Apr-2010 10:51:40] <citrus2> its funny that bi weekly can mean twice a week or once every 2 weeks

[15-Apr-2010 10:51:45] <citrus2> english...

[15-Apr-2010 10:52:04] <themactech> yeah, well french is my native tongue so I beat you on complexity right there

[15-Apr-2010 10:52:16] <bedwards> brixius: you should be in product marketing. also, what you said about zDeviceTemplates is correct. The "Bind Templates" and "Reset Bindings" menu items just set that zProperty.

[15-Apr-2010 10:52:20] <themactech> though I never got why you anglos drive on a parkway and park in a driveway

[15-Apr-2010 10:52:39] <bedwards> semiweekly = twice a week

[15-Apr-2010 10:53:47] <themactech> ok, well I guess I have a lot of testing to do, hopefully I can do most of this within 2 weeks so I can be here to bug you guys some more in the next session

[15-Apr-2010 10:53:57] <themactech> when do these end, I assume you're all not here till 11pm

[15-Apr-2010 10:54:41] <bedwards> In six minutes the official Developer chat is done, but people are on here all the time.

[15-Apr-2010 10:55:22] <themactech> alright, well thank you all very much for the info, its much appreciated, I will try out the leads you gave me, have a great day fellas

[15-Apr-2010 10:57:48] <themactech> hey, I just remembered a question I wanted to ask

[15-Apr-2010 10:58:23] <themactech> I have looked up how zenoss sets model name and os name, it does so via the relationship set by the OSproductkey and HWproductkey

[15-Apr-2010 10:58:51] <themactech> that means you have to keep an updated db of os and hw products

[15-Apr-2010 10:59:02] <themactech> could a modeler add entries in there automatically?

[15-Apr-2010 10:59:19] <Jane_Curry> Anyone seen my earlier querstion...

[15-Apr-2010 10:59:22] <Jane_Curry> Does anyone know what the collector parameter "Maximum ping failures" actually does in 2.5.1 / 2.5.2?

[15-Apr-2010 11:00:01] <themactech> so my apple modeler that pulls the apple model name from apple's web site, could I have the modeler check models in the Apple manufacturer, and if it does not find it, add the proper entry

[15-Apr-2010 11:00:16] <themactech> that way my modeler would automatically update the product and os database

[15-Apr-2010 11:00:40] <bedwards> it adds new products/manufacturers by default

[15-Apr-2010 11:00:59] <bedwards> Jane: I have not tested Maximum ping failures

[15-Apr-2010 11:01:45] <themactech> thanks

[15-Apr-2010 11:06:56] <mrayzenoss> Jane_Curry: http://dev.zenoss.org/trac/ticket/5809

[15-Apr-2010 11:16:05] <Jane_Curry> Thanks Matt - v helpful

[15-Apr-2010 11:16:46] <Simon4_> evening all

[15-Apr-2010 11:16:59] <Simon4_> volcanic ash is fail.

[15-Apr-2010 11:17:11] <mrayzenoss> Simon4_: interfere with your vacation?

[15-Apr-2010 11:17:22] <Simon4_> mrayzenoss: I'm supposed to be on a plane to NZ right now :/

[15-Apr-2010 11:17:36] <mrayzenoss> awwww, that's terrible

[15-Apr-2010 11:20:28] <Simon4_> in other, more zenoss stuff... ssh collectors. Do these have to be a command on the remote server that returns stuff in nagios format? or can you extend a class in a similar way to a ssh modeler where you can run commands and parse the output?

[15-Apr-2010 11:20:42] <Simon4_> I've been trying to find a zenpack that I can copy from, but failing so far

[15-Apr-2010 11:22:20] <Simon4_> basically want it to ssh to the device, run some "show xxxx" commands, parse the output and use those values for collection

[15-Apr-2010 11:25:15] <mrayzenoss> Simon4_: you can definitely do modeling and collection, take a look at the LinuxMonitor or OpenSolaris

[15-Apr-2010 11:25:43] <Simon4_> thanks :) off for a look now

[15-Apr-2010 11:26:19] <mrayzenoss> Simon4_: and this: docs/DOC-2381

[15-Apr-2010 11:28:35] <Simon4_> cool... modelling I have down (hpbladechassis uses it), just want to get collection working via ssh also so we're not using 4 different ways of talking to the chassis

[15-Apr-2010 11:36:05] <Simon4_> ahhh, commandParser is what I'm after I think

[15-Apr-2010 11:36:23] <Simon4_> thanks mrayzenoss

[15-Apr-2010 11:36:47] <mrayzenoss> good luck

[15-Apr-2010 11:37:01] <Simon4_> for the coding? or the flight? ;)

[15-Apr-2010 11:37:34] <mrayzenoss> well, I guess both

[15-Apr-2010 11:37:38] <Simon4_> heh, thanks

[15-Apr-2010 11:37:52] <Simon4_> all going well we're out of here on tomorrow's flight, but will see what happens

[15-Apr-2010 11:38:05] <mrayzenoss> do you get a refund when they cancel it for volcanic ash?

[15-Apr-2010 11:38:18] <Simon4_> seems the airlines will. Travel insurance is worthless though

[15-Apr-2010 11:46:50] <rsher51> must be a first: cancellations due to volcanic ash....

[15-Apr-2010 11:47:38] <Simon4_> yeah, not something we thought about till this am

[15-Apr-2010 11:49:52] <Simon4_> mrayzenoss: the opensolaris pack with it's unit tests etc makes great reference material

[15-Apr-2010 11:50:05] <mrayzenoss> be sure to tell rocket

[15-Apr-2010 11:50:38] <Simon4_> cool, will do.

[15-Apr-2010 14:29:09] <cgibbons> hmm

[15-Apr-2010 15:16:13] <kobalt> anyone else see this in their zenperfsnmp logs "zen.zenperfsnmp: Error loading config for devices..."

[15-Apr-2010 15:16:36] <Simon4_> kobalt: I've hit it before

[15-Apr-2010 15:17:27] <kobalt> Simon4_, any Idea what is causing it, I see it on a almost weekly basis sometimes more

[15-Apr-2010 15:17:46] <Simon4_> kobalt: I've found that a reindex() in zendmd fixes it

[15-Apr-2010 15:17:55] <kobalt> yeah I do that each time...

[15-Apr-2010 15:18:02] <Simon4_> I don't have a permanent fix :(

[15-Apr-2010 15:18:03] <kobalt> same devices report most of the time

[15-Apr-2010 15:18:07] <Simon4_> yeah

[15-Apr-2010 15:18:25] <Simon4_> there's a bug somewhere, but I haven't had time to try and dig it out

[15-Apr-2010 15:18:39] <kobalt> when you have over 2k devices its takes a bit to reindex all the time

[15-Apr-2010 15:18:43] * Simon4_ nods

[15-Apr-2010 15:19:05] <Simon4_> I was going to upgrade to 2.5.2, then dig further if it still happens

[15-Apr-2010 15:19:26] <kobalt> I found it affects other devices when I move a device from one class to another

[15-Apr-2010 15:19:35] <kobalt> but it seems to hit a select few when I do

[15-Apr-2010 15:19:37] <kobalt> dont

[15-Apr-2010 15:19:56] <kobalt> I will be going to 2.5.2 when I get back from vacation

[15-Apr-2010 15:35:43] <kobalt> so if I wanted to modify/create a collector to gather certian mib information and store it on a section of the status page for instance the serial # feild what file would I need to look at to see how its done

[15-Apr-2010 15:36:07] <Simon4_> you're looking at making a modeler plugin

[15-Apr-2010 15:37:24] <rmatte> kobalt: http://www.mail-archive.com/zenoss-users@zenoss.org/msg31186.html//www.mail-archive.com/zenoss-users@zenoss.org/msg31186.html

[15-Apr-2010 15:37:29] <rmatte> that might be helpful to you

[15-Apr-2010 15:37:35] <rmatte> in troubleshooting the error

[15-Apr-2010 15:38:37] <mrayzenoss> kobalt: that same discussion was held in here 5 hours ago: docs/DOC-5852

[15-Apr-2010 15:38:52] <mrayzenoss> you might want to trade notes with themactech

[15-Apr-2010 15:40:45] <kobalt> mrayzenoss, Sweet Ill have to take a look

[15-Apr-2010 15:46:45] <kobalt> rmatte, the devices do generate graphs and I get no error from the zendmd code that chet put in, I have noticed I get it every 6 hours (360 model cycle time)

[15-Apr-2010 15:48:16] <blull> hey

[15-Apr-2010 15:48:18] <blull> quick question

[15-Apr-2010 15:48:26] <blull> i enabled the zenoss.nmap.IpServiceMap plugin

[15-Apr-2010 15:48:40] <blull> i symlinked /opt/zenoss/libexec/nmap to /usr/bin/nmap (which exsists) on my remote collector

[15-Apr-2010 15:48:47] <blull> I made sure the device is assigned to the remote collector.

[15-Apr-2010 15:49:09] <blull> but im getting a pythonclient traceback saying /opt/zenoss/libexec/nmap does not exist.

[15-Apr-2010 15:49:10] <blull> it does.

[15-Apr-2010 15:51:05] <blull> /opt/zenoss/bin/zenmodeler run --now --monitor es-mon1 -F -d es-ws50.chicago.flextrade.com --weblog

[15-Apr-2010 15:51:12] <blull> es-mon1 is my remote monitor.

[15-Apr-2010 15:54:32] <blull> okayw eird

[15-Apr-2010 15:54:38] <blull> now all of a sudden my zenhub stopped taking connections.

[15-Apr-2010 15:54:43] <blull> and nothing in the logs saying it broke.

[15-Apr-2010 15:54:49] <blull> just started getting mad shit going red in events :/

[15-Apr-2010 15:56:52] <Simon4_> blull: are you sure the pythonclient isn't doing a "is this a file" test on the symlink and failing?

[15-Apr-2010 15:56:59] <Simon4_> random thought

[15-Apr-2010 15:58:26] <blull> pretty much unless the execvepe call in os.py in python is doing a if file check.

[15-Apr-2010 16:00:09] <rmatte> blull: just copy the nmap binary instead of symlinking, it should still work

[15-Apr-2010 16:00:20] <blull> yeah trying that now.

[15-Apr-2010 16:00:22] <rmatte> it did when I tried it

[15-Apr-2010 16:00:34] <rmatte> also, did you apply the patch in that trac ticket?

[15-Apr-2010 16:00:34] <blull> nah same thing man.

[15-Apr-2010 16:00:42] <blull> what patch :)

[15-Apr-2010 16:01:02] <rmatte> http://dev.zenoss.org/trac/ticket/5238

[15-Apr-2010 16:01:14] <rmatte> read through that

[15-Apr-2010 16:01:32] <blull> oh i c whats going on.

[15-Apr-2010 16:01:33] <rmatte> it's still not patched in 2.5.2, so you'll have to do it yourself

[15-Apr-2010 16:01:36] <blull> its using wrong flags to nmap.

[15-Apr-2010 16:02:40] <rmatte> well, that was part of the problem

[15-Apr-2010 16:02:46] <blull> -p 1-4000 -oG --open -sT

[15-Apr-2010 16:02:47] <blull> oh

[15-Apr-2010 16:02:47] <blull> heh

[15-Apr-2010 16:02:49] <rmatte> but there's also a patch that needs to be applied for it to work properly

[15-Apr-2010 16:02:52] <blull> hmm

[15-Apr-2010 16:02:52] <rmatte> read through the trac ticket

[15-Apr-2010 16:02:53] <blull> okay

[15-Apr-2010 16:02:54] <blull> k

[15-Apr-2010 16:03:23] <rmatte> you can skip down to my comments since I basically explain step by step how to fix it

[15-Apr-2010 16:03:50] <rmatte> I also submitted a patch recent but it's not available via zenpatch yet

[15-Apr-2010 16:05:09] <rmatte> I need to speak with cholden about how to submit the patch properly since he submitted the original one (though it wasn't 100% good)

[15-Apr-2010 16:05:55] <rmatte> If you zenpatch 16617

[15-Apr-2010 16:06:06] <rmatte> Then follow this comment: http://dev.zenoss.org/trac/ticket/5238#comment:12

[15-Apr-2010 16:06:10] <rmatte> It'll work

[15-Apr-2010 16:06:19] <rmatte> just to simplify things for you

[15-Apr-2010 16:09:08] <rmatte> Matt is working on getting the patch actually merged

[15-Apr-2010 16:10:57] <sciolist> hrm... I'm trying to get the LDAP Monitor template to check on one of our LDAP servers

[15-Apr-2010 16:11:58] <rmatte> ok...

[15-Apr-2010 16:12:04] <sciolist> grrr....nm, I forgot to save the zLDAPBaseDN correctly

[15-Apr-2010 16:12:08] <rmatte> lol

[15-Apr-2010 16:13:35] <blull> hmm

[15-Apr-2010 16:13:40] <blull> got the patch applied and the changes to zprefs

[15-Apr-2010 16:13:42] <blull> still fails.

[15-Apr-2010 16:14:23] <blull> lame

[15-Apr-2010 16:14:28] <blull> its kicking nmap off on the main zenhub

[15-Apr-2010 16:14:31] <blull> not the remote collector

[15-Apr-2010 16:14:42] <blull> this host is configured for the remote collector though?

[15-Apr-2010 16:15:44] <rmatte> you changed the nmap flags as described?

[15-Apr-2010 16:15:55] <rmatte> then you need nmap on the remote collector obviously

[15-Apr-2010 16:16:01] <rmatte> it's not going to use it on the master

[15-Apr-2010 16:16:11] <blull> yup, i have it on the remote system

[15-Apr-2010 16:16:12] <rmatte> I'd put it on both to be safe

[15-Apr-2010 16:16:16] <rmatte> and apply the patch to both

[15-Apr-2010 16:16:19] <rmatte> and changes

[15-Apr-2010 16:16:20] <blull> i just ran the command it was saying it was kicking off, on the remote collector, and it worekd.

[15-Apr-2010 16:16:24] <blull> yup i did.

[15-Apr-2010 16:16:46] <blull> i applied it to both, changed the zprefs at the /device level and the level of device im doing /devices/datacenter/citrix zpref's

[15-Apr-2010 16:17:03] <rmatte> did you make the manual changes that I detailed?

[15-Apr-2010 16:17:25] <rmatte> to $ZENHOME/Products/DataCollector/plugins/zenoss/nmap/IpServiceMap.py

[15-Apr-2010 16:18:01] <rmatte> the 16617 patch is incomplete, so those changes have to be made by hand for now

[15-Apr-2010 16:18:43] <blull> i c

[15-Apr-2010 16:18:49] <blull> i assumed they were included in the 8 day old patch though?

[15-Apr-2010 16:18:55] <blull> cause thast a 2 month old patch

[15-Apr-2010 16:19:02] <blull> you may want to make a note about still having to do that.

[15-Apr-2010 16:19:33] <rmatte> oh, yeh, they are included in the 8 day old patch, but the 8 day old patch expects the files to be in /usr/local/zenoss/zenoss

[15-Apr-2010 16:19:37] <rmatte> which I'm guessing yours aren't

[15-Apr-2010 16:19:46] <rmatte> like I said, the latest patches aren't 100% good to go get

[15-Apr-2010 16:19:58] <rmatte> hence why I said to use znpatch 16617 for now

[15-Apr-2010 16:20:03] <rmatte> and do those changes manually

[15-Apr-2010 16:20:34] <blull> yeah

[15-Apr-2010 16:20:37] <rmatte> I never tested with a remote collector though

[15-Apr-2010 16:20:40] <blull> so I patched with -p5

[15-Apr-2010 16:20:43] <rmatte> so that's another matter entirely

[15-Apr-2010 16:20:46] <blull> so it applied fine

[15-Apr-2010 16:20:49] <rmatte> k

[15-Apr-2010 16:21:00] <blull> also, it looks like it applied that additional readyopts line

[15-Apr-2010 16:21:02] <blull> as its already there.

[15-Apr-2010 16:21:04] <rmatte> what if you do the same thing with nmap on the master?

[15-Apr-2010 16:21:06] <rmatte> does it work then?

[15-Apr-2010 16:21:09] <blull> nope.

[15-Apr-2010 16:21:27] <blull> let me restart both collector and zenhub again just to make sure I didn't foobar something.

[15-Apr-2010 16:21:28] <rmatte> there are supposed to be 2 readyopts lines

[15-Apr-2010 16:21:32] <blull> there are.

[15-Apr-2010 16:21:36] <rmatte> k

[15-Apr-2010 16:21:40] <blull> it looks exactly like what it says it needs to be changed to.

[15-Apr-2010 16:21:49] <rmatte> k

[15-Apr-2010 16:22:27] <rmatte> so you have nmap copied to that same location on the master and the collector?

[15-Apr-2010 16:22:36] <blull> nope, just the collector.

[15-Apr-2010 16:22:40] <blull> Why should it have to be on the master?

[15-Apr-2010 16:22:41] <kobalt> so question if I copy CiscoMap.py and rename it and change the oid in it, is there anything else I need to do to get it to run?

[15-Apr-2010 16:22:43] <rmatte> do it on the master too, just to rule that out

[15-Apr-2010 16:22:45] <blull> im only using the nmap from the remote collector.

[15-Apr-2010 16:22:46] <blull> k

[15-Apr-2010 16:22:56] <rmatte> if it works when you copy it on the master then we know that it's trying to run it from the master instead

[15-Apr-2010 16:23:04] <rmatte> then at least we have a starting point for fixing that behaviour

[15-Apr-2010 16:23:23] <rmatte> and yes, you need to restart zenoss on both the master and the collector after applying the patch

[15-Apr-2010 16:23:39] <rmatte> so if you hadn't done that, that's very likely the issue

[15-Apr-2010 16:23:47] <rmatte> you should restart any time any patch is applied

[15-Apr-2010 16:23:49] <blull> alright here we go.

[15-Apr-2010 16:23:52] <blull> yeah

[15-Apr-2010 16:24:37] <blull> yeah im doing a tcpdump on the zenhub

[15-Apr-2010 16:24:42] <blull> it just did the entire model from the zenhub

[15-Apr-2010 16:24:45] <blull> not the remote collector.

[15-Apr-2010 16:24:48] <rmatte> figures

[15-Apr-2010 16:25:04] <blull> so it should work from the remote collector when the device is assigned to it right?

[15-Apr-2010 16:25:13] <rmatte> in theory yes

[15-Apr-2010 16:25:17] <blull> in theory?

[15-Apr-2010 16:25:28] <rmatte> core is not extensively tested with collectors

[15-Apr-2010 16:25:37] <blull> I thought in practice that was the entire purpose of the remote collector =p

[15-Apr-2010 16:25:38] <blull> hehe yeah

[15-Apr-2010 16:25:39] <sciolist> where would the LDAPMonitor ZenPack log to? zencommand.log?

[15-Apr-2010 16:25:42] <rmatte> so it's possible that something is messed up which would cause it to only model from the master

[15-Apr-2010 16:25:49] <blull> but core is that different than enterprise?

[15-Apr-2010 16:25:57] <rmatte> sciolist: depends on how it's actually doing the monitoring

[15-Apr-2010 16:26:01] <blull> hrmm

[15-Apr-2010 16:26:04] <blull> thats gonna be an issue

[15-Apr-2010 16:26:08] <blull> every 720 minutes

[15-Apr-2010 16:26:10] <blull> it remodals

[15-Apr-2010 16:26:13] <blull> remodels rather

[15-Apr-2010 16:26:17] <rmatte> sciolist: if it's using SNMP datasources then zenperfsnmp, if it's using command datasources then zencommand

[15-Apr-2010 16:26:23] <rmatte> sciolist: and so on

[15-Apr-2010 16:26:25] <blull> but it will take place from the zenhub, which is completely firewalled off from all these other collectors.

[15-Apr-2010 16:26:29] <sciolist> should be command-based (I think)

[15-Apr-2010 16:26:33] <sciolist> thank

[15-Apr-2010 16:26:34] <sciolist> s

[15-Apr-2010 16:26:49] <blull> which will then mark everything down thats an ipservice :/

[15-Apr-2010 16:26:56] <blull> in theory at least.

[15-Apr-2010 16:26:58] <blull> heh

[15-Apr-2010 16:26:59] <rmatte> blull: you sure zenmodel on the remote collector is configure properly?

[15-Apr-2010 16:27:07] <rmatte> blull: to connect back to zenhub on the master?

[15-Apr-2010 16:27:09] <blull> ill double check 1 sec.

[15-Apr-2010 16:27:12] <rmatte> k

[15-Apr-2010 16:27:48] <blull> [zenoss@es-mon1 etc]$ cat /opt/zenoss/etc/zenmodeler.conf

[15-Apr-2010 16:27:49] <blull> monitor es-mon1

[15-Apr-2010 16:27:49] <blull> hubhost zenoss-hub

[15-Apr-2010 16:27:55] <rmatte> blull: well, it wouldn't mark it as down, it would just set them to not be monitored unless they are locked

[15-Apr-2010 16:28:03] <blull> es-mon1 resolve to ip (not 127.0.0.1) of the remote modeler

[15-Apr-2010 16:28:09] <rmatte> k

[15-Apr-2010 16:28:21] <rmatte> and hubhost is the ip of the hub obviously

[15-Apr-2010 16:28:21] <blull> zenoss-hub resolves to the real ip of the remote collector as well

[15-Apr-2010 16:28:27] <rmatte> k

[15-Apr-2010 16:28:28] <rmatte> hmmm

[15-Apr-2010 16:28:42] <blull> as it goes via a stunnel connection to the zenhub in a different data center.

[15-Apr-2010 16:28:43] <rmatte> sounds like a possible bug, because it should obviously be doing the remodelling from the collector

[15-Apr-2010 16:28:58] <rmatte> I'm out of here soon but I'll ask around about that tomorrow and get back to you

[15-Apr-2010 16:29:03] <blull> if i kick a zenmodel off from the remote collector, it updates correctly on the zenhub.

[15-Apr-2010 16:29:12] <blull> k thanks.

[15-Apr-2010 16:29:14] <rmatte> You could post on the forums for the time being and see if anyone has any ideas

[15-Apr-2010 16:29:14] <blull> im curious

[15-Apr-2010 16:29:23] <blull> does enterprise have these type of issues on a regular basis?

[15-Apr-2010 16:29:28] <blull> or is it something much different than core?

[15-Apr-2010 16:29:43] <rmatte> enterprise's collector system is much more robust than Core's

[15-Apr-2010 16:29:55] <rmatte> so I doubt it

[15-Apr-2010 16:30:16] <rmatte> If this is a bug though, I'll open a ticket to have it fixed

[15-Apr-2010 16:30:26] <rmatte> I just want to confirm first

[15-Apr-2010 16:30:40] <sektorNBA> hi

[15-Apr-2010 16:30:41] <rmatte> you're not possibly blocking traffic going from the zenhub to the collector?

[15-Apr-2010 16:30:47] <blull> nah

[15-Apr-2010 16:30:51] <rmatte> k

[15-Apr-2010 16:30:53] <blull> i have everyting going via tunnels

[15-Apr-2010 16:30:58] <blull> verified the firewall rules myself on the cisco.

[15-Apr-2010 16:31:03] <rmatte> k

[15-Apr-2010 16:31:03] <sektorNBA> I wanna add some datasources to monitor some FreeBSD servers under zenoss. Where can I find all freeBSD available oids?

[15-Apr-2010 16:31:05] <blull> and i've dumped traffic on both ends.

[15-Apr-2010 16:31:10] <blull> so in the short term

[15-Apr-2010 16:31:16] <blull> i guess writing a tcp port check plugin?

[15-Apr-2010 16:31:19] <blull> and add it as a command?

[15-Apr-2010 16:31:23] <rmatte> sektorNBA: If you install net-snmp on your FreeBSD servers you can just use the Linux device template

[15-Apr-2010 16:31:24] <blull> nagios style

[15-Apr-2010 16:31:36] <blull> cause I have this datacenter nearly deployed minus a port 22 and port 3389 status check

[15-Apr-2010 16:31:43] <rmatte> sektorNBA: just copy it to a different class called FreeBSD

[15-Apr-2010 16:31:45] <blull> once thats done I am turning nagios off for this dc and moving on to the next dc.

[15-Apr-2010 16:32:19] <sektorNBA> net-snmp-5.4.2.1_6 An extendable SNMP implementation

[15-Apr-2010 16:32:21] <sektorNBA> i have it.

[15-Apr-2010 16:32:27] <rmatte> k

[15-Apr-2010 16:32:33] <rmatte> make sure it's configured correctly

[15-Apr-2010 16:32:41] <rmatte> and use the Linux device template

[15-Apr-2010 16:32:58] <sektorNBA> yes its /Devices /Server /Linux /10.100.0.125

[15-Apr-2010 16:33:06] <sektorNBA> but some of my graphs are weird.

[15-Apr-2010 16:33:17] <rmatte> blull: you could do that, yes

[15-Apr-2010 16:33:28] <rmatte> blull: I guess the SNMP port monitoring doesn't work for you?

[15-Apr-2010 16:33:44] <rmatte> sektorNBA: which ones?

[15-Apr-2010 16:34:22] <sektorNBA> rmatte cpu and memory

[15-Apr-2010 16:35:12] <rmatte> hmmm, does it graph at all for those?

[15-Apr-2010 16:35:31] <blull> we don't run snmp on every device.

[15-Apr-2010 16:35:41] <blull> ie. we have citrix appliance servers that don't give us the ability to do that .

[15-Apr-2010 16:35:53] <blull> + some old legacy appliance servers.

[15-Apr-2010 16:35:53] <rmatte> blull: ah

[15-Apr-2010 16:36:13] <rmatte> blull: yeh, we have the same issue here, but I'm waiting for the nmap issue to be properly patched before using it

[15-Apr-2010 16:36:19] <blull> im just concerned about 21, 22, 80, 443, and 3389 really.

[15-Apr-2010 16:36:23] <rmatte> mostly because I have 13 Zenoss servers (and more soon)

[15-Apr-2010 16:36:32] <rmatte> so I don't want to screw around with patching until it's solid

[15-Apr-2010 16:36:37] <sektorNBA> rmatte yes it does. CPU is 100% idle always and buffered always

[15-Apr-2010 16:36:45] <blull> yeah I will have 7 here when I am done deploying.

[15-Apr-2010 16:36:53] <blull> 8 if you count the display server.

[15-Apr-2010 16:36:59] <rmatte> sektorNBA: I'm not sure what the proper OIDs for that are...

[15-Apr-2010 16:37:16] <rmatte> sektorNBA: I'm surprised Zenoss doesn't include a FreeBSD template by default, or at least a ZenPack

[15-Apr-2010 16:37:18] <blull> i wrote some nifty little cross platform COMMAND scripts today though :)

[15-Apr-2010 16:37:29] <rmatte> sektorNBA: If you figure it out let me know and I'll help you put a ZenPack together

[15-Apr-2010 16:37:48] <sektorNBA> ok. i will do a nat to show it from internet. hold on plz

[15-Apr-2010 16:37:51] <rmatte> sektorNBA: see if you can find some cacti templates for it that contain the correct OIDs or something\

[15-Apr-2010 16:38:14] <sektorNBA> rmatte i am out of cacti. My boss loves zenoss !

[15-Apr-2010 16:38:26] <blull> $ZENOSS/libexec/check_disk_snmp <host> <drive or path>  -- ie   check_disk_snmp foobar123 C:\  reports max size, current utilization, and percent utilized :)  same script automagically works in linux based systems too :)

[15-Apr-2010 16:38:28] <blull> okay I go home, ttyl.

[15-Apr-2010 16:38:43] <blull> my boss has no idea of what realistic for deploying a 5 datacenter monitoring solution

[15-Apr-2010 16:38:51] <blull> its a minimum of a 2 - 3 month project at least.

[15-Apr-2010 16:39:02] <blull> i've had 7 business days so far and im already "it was due last week" lol

[15-Apr-2010 16:39:11] <sektorNBA> [root@Trinity ~]# snmpwalk -v1 -c Argos 10.100.0.125 1.3.6.1.4.1.2021.4.6.0

[15-Apr-2010 16:39:11] <sektorNBA> UCD-SNMP-MIB::memAvailReal.0 = INTEGER: 79428 kB

[15-Apr-2010 16:39:31] <sektorNBA> rmatte thats memAvailReal

[15-Apr-2010 16:39:40] <rmatte> sektorNBA: yeh I know, I'm just saying that there may be some cacti templates kicking around online that you can just use as a reference

[15-Apr-2010 16:40:12] <sektorNBA> rmatte ok. i will look for some

[15-Apr-2010 16:40:16] <rmatte> blull: yeh, that sound like our management too

[15-Apr-2010 16:40:23] <rmatte> sounds*

[15-Apr-2010 16:40:41] <sektorNBA> rmatte DO i need this 1024,*,${here/hw/totalMemory},/,1,-,-100,* ?

[15-Apr-2010 16:40:48] <rmatte> sektorNBA: look for hrProcessorLoad

[15-Apr-2010 16:40:53] <sektorNBA> Its under RPN

[15-Apr-2010 16:40:54] <rmatte> see if that's available in an snmpwalk

[15-Apr-2010 16:41:14] <sektorNBA> [root@Trinity ~]# cat trinity.snmp.txt |grep hrProcessorLoad

[15-Apr-2010 16:41:15] <sektorNBA> HOST-RESOURCES-MIB::hrProcessorLoad.768 = INTEGER: 1

[15-Apr-2010 16:41:29] <rmatte> sektorNBA: they are using that to calculate the memory, so probably

[15-Apr-2010 16:41:49] <rmatte> sektorNBA: if you add -0n when you query the memAvailReal OID you'll see the true untranslated value

[15-Apr-2010 16:41:54] <rmatte> or sorry, -On

[15-Apr-2010 16:41:55] <rmatte> rather

[15-Apr-2010 16:42:25] <rmatte> ok, so it does offer hrProcessorLoad, that's the percentage of CPU in use

[15-Apr-2010 16:42:31] <rmatte> which is currently 1% from the looks of it

[15-Apr-2010 16:42:39] <rmatte> so your description of 100% idle might actually be accurate

[15-Apr-2010 16:42:53] <rmatte> might want to run something like cat /dev/urandom > /dev/null on the box to generate fake load

[15-Apr-2010 16:43:03] <rmatte> then see how the graph is affected after about 15 to 30 minutes

[15-Apr-2010 16:43:14] <sciolist> hrProcessorLoad is on ly the load average, correct? not the CPU utilization

[15-Apr-2010 16:43:32] <sektorNBA> ok rmatte. i am trying this one

[15-Apr-2010 16:43:36] <rmatte> no, that's the CPU utilization

[15-Apr-2010 16:43:43] <rmatte> not the load average

[15-Apr-2010 16:43:58] <sektorNBA> rmatte is there an oid to uptime?

[15-Apr-2010 16:44:20] <rmatte> there's a standard uptime OID, just bind the /Devices/Templates/Device template

[15-Apr-2010 16:44:26] <rmatte> it's used to collect uptime

[15-Apr-2010 16:44:31] <rmatte> or reference it if you need the OID

[15-Apr-2010 16:44:55] <rmatte> make sure it's named exactly as you see it with the exact same OID

[15-Apr-2010 16:45:03] <rmatte> that's what Zenoss uses to show uptime on the status page

[15-Apr-2010 16:45:07] <rmatte> and in reports or whatever

[15-Apr-2010 16:45:28] <rmatte> well, I'm done for the day, I'll be back tomorrow

[15-Apr-2010 16:45:30] <rmatte> cheers

[15-Apr-2010 16:45:33] <sektorNBA> ty 1.3.6.1.2.1.1.3.0

[15-Apr-2010 16:45:41] <rmatte> that's the one

[15-Apr-2010 16:45:42] <rmatte> later

[15-Apr-2010 16:45:49] <sektorNBA> rmatte 1 more sec plz

[15-Apr-2010 16:45:55] <rmatte> (make sure you name the OID the same if you're adding it to a new template)

[15-Apr-2010 16:45:58] <rmatte> k, make it quick

[15-Apr-2010 16:45:58] <rmatte> lol

[15-Apr-2010 16:46:07] <sektorNBA> rmatte how can i rename the linux template to FreeBSD?

[15-Apr-2010 16:46:32] <sektorNBA> not raname.. create a new one

[15-Apr-2010 16:46:44] <rmatte> did you copy it over?

[15-Apr-2010 16:47:02] <rmatte> go to the templates tab, put a checkmark next to it and select copy template

[15-Apr-2010 16:47:12] <rmatte> create a new device class called /Server/FreeBSD beforehand

[15-Apr-2010 16:47:16] <rmatte> then select to copy it to that

[15-Apr-2010 16:47:17] <sektorNBA> rmatte i was at a device, i createa new copy to this device

[15-Apr-2010 16:47:26] <rmatte> don't copy it right to the device

[15-Apr-2010 16:47:27] <sektorNBA> ok

[15-Apr-2010 16:47:28] <rmatte> copy it to a class

[15-Apr-2010 16:47:36] <rmatte> you don't want a local copy on the device itself

[15-Apr-2010 16:48:01] <sektorNBA> ok, i am at /devices click at templates

[15-Apr-2010 16:48:05] <sektorNBA> there is no linux there

[15-Apr-2010 16:49:44] <sektorNBA> rmatte ?

[15-Apr-2010 16:50:15] <rmatte> go to /Server/Linux

[15-Apr-2010 16:52:09] <mrayzenoss> anyone using Chef?  Just got off a call with them and looking for guinea pigs

[15-Apr-2010 16:53:04] <rmatte> noop

[15-Apr-2010 19:29:43] forsberg is now known as fOrsberg

[15-Apr-2010 20:24:15] <Simon4> anyone used changePythonClass? it's a method on a device object

[15-Apr-2010 21:26:00] <namezk> anybody else experiencing gaps in graphs?

[15-Apr-2010 21:27:50] <Simon4> namezk: not interface graphs, by chance?

[16-Apr-2010 00:00:30] [disconnected at Fri Apr 16 00:00:30 2010]

[16-Apr-2010 00:00:30] [connected at Fri Apr 16 00:00:30 2010]

[16-Apr-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[16-Apr-2010 00:00:47] <jordan.freenode.net> [freenode-info] please register your nickname...don't forget to auto-identify! http://freenode.net/faq.shtml#nicksetup

[16-Apr-2010 03:40:18] <JesperAway> morning

[16-Apr-2010 03:40:24] JesperAway is now known as Jesper

[16-Apr-2010 03:40:35] <Jesper> I have made the following class/sub-device. Classes/Devices/Network/Router/Cisco/ADSL 4Mbit-768Kbit and I have made the template "copy_of_ethernetCsmacd_64).

[16-Apr-2010 03:40:36] <Jesper> I would like to bind it to a device class, but when I try to bind, I'm only able to select "Device [/Devices]" That doesn't seem to be right?

[16-Apr-2010 03:41:01] <Jesper> shouldn't I be able to select the device class I added it in?

[16-Apr-2010 03:46:47] fOrsberg is now known as forsberg

[16-Apr-2010 07:40:23] <sektorNBA> morning

[16-Apr-2010 07:40:49] <ckrough> morning

[16-Apr-2010 07:41:48] <sektorNBA> I am doing a cpu-load graph and i want to set min and max values, can anyone help me?

[16-Apr-2010 07:50:58] <ckrough> you want a min/max value to sanity check data returned by polls, or you want threshold alerts?

[16-Apr-2010 07:52:01] <sektorNBA> i got it.

[16-Apr-2010 07:52:11] <ckrough> gratz

[16-Apr-2010 07:52:15] <sektorNBA> snmpwalk -cArgos -v1 172.16.0.4 1.3.6.1.2.1.1.3.0

[16-Apr-2010 07:52:25] <sektorNBA> that´s uptime. which format is that?

[16-Apr-2010 07:52:31] <sektorNBA> SNMPv2-MIB::sysUpTime.0 = Timeticks: (5676400) 15:46:04.00

[16-Apr-2010 07:52:43] <ckrough> what do you mean by format?

[16-Apr-2010 07:52:58] <sektorNBA> I wanna show my graphs in days

[16-Apr-2010 07:53:27] <ckrough> you would pick that when viewing the report, up top in the drop down that says "hourly, daily" etc

[16-Apr-2010 07:54:06] <ckrough> time series uptime graph might look odd, since it just keeps going up if there are no reboots

[16-Apr-2010 07:55:46] <sektorNBA> I guess 5676400 is seconds.

[16-Apr-2010 08:12:58] <sektorNBA> how do i rename a template?

[16-Apr-2010 09:36:57] <sciolist> grrr, how do I reset the admin password on the Zope instance? The Settings-> Users change doesn't seem to do it

[16-Apr-2010 09:44:00] <Jesper> sektorNBA: as far as I know you can only rename your own templates, eg. if you made a copy of one of the default templates.

[16-Apr-2010 09:45:01] <Jesper> and the field you need to change is the one on the top

[16-Apr-2010 10:32:16] pipit is now known as pipit_a-w-a-y

[16-Apr-2010 10:37:10] ke4qqq_ is now known as ke4qqq

[16-Apr-2010 11:37:04] <blull> hi

[16-Apr-2010 11:37:06] <blull> quick question

[16-Apr-2010 11:37:09] <blull> in zproperties

[16-Apr-2010 11:37:36] <blull> there is a CommandCycleTimeout, and CommandCommandTimeout, are those just related to running commands on the remote system?

[16-Apr-2010 11:37:42] <blull> or are those the timeouts applied to checks?

[16-Apr-2010 11:37:56] <blull> I thought each template has a cycle timeout that it uses, but I have it pinging 64 hosts every 5 seconds

[16-Apr-2010 11:38:01] <blull> i don't see that much traffic :/

[16-Apr-2010 11:38:42] <blull> wait nm

[16-Apr-2010 11:38:43] <blull> it is.

[16-Apr-2010 11:47:22] <sektorNBA> Jesper thanks, let me try raname it

[16-Apr-2010 12:09:43] <blull> hey

[16-Apr-2010 12:09:52] <blull> whats the easiest way to disable an alert for a single device?

[16-Apr-2010 12:10:04] <blull> specifically I need to disable the memcheck on a host that I implemented as an external command

[16-Apr-2010 12:10:47] <blull> define a maintenance window for it?

[16-Apr-2010 12:19:34] <blull> grr

[16-Apr-2010 12:19:35] <blull> well

[16-Apr-2010 12:19:38] <blull> that puts it in a maint window

[16-Apr-2010 12:19:46] <blull> but i just need to temporarily supress 1 alert for a couple of days.

[16-Apr-2010 12:19:55] <blull> there doesn't appear to be an easy way to do that besides ack'ing it?

[16-Apr-2010 12:27:39] mrayzenoss is now known as mray

[16-Apr-2010 12:27:48] mray is now known as mrayzenoss

[16-Apr-2010 12:46:31] <otakup0pe> quick random question. i remember readinng once about a flash app which acted as like a traffic light view of zenoss

[16-Apr-2010 12:47:03] <mrayzenoss> otakup0pe: not ringing a bell

[16-Apr-2010 12:50:48] <otakup0pe> ahh i was thinking the iphone event console

[16-Apr-2010 12:51:09] <mrayzenoss> ahh, that's a PHP app that's read-only

[16-Apr-2010 12:51:32] <otakup0pe> ack

[16-Apr-2010 12:51:47] <otakup0pe> zenoss goes through all the trouble to make decent web services and this app uses the raw db heh

[16-Apr-2010 13:00:06] <ckrough> got a link to the php app?

[16-Apr-2010 13:00:15] <ckrough> sounds like NOC projection candy

[16-Apr-2010 13:05:57] <mrayzenoss> ckrough: http://sourceforge.net/projects/zipec/

[16-Apr-2010 13:07:32] <ckrough> ahh, seen that, I thought it was something else. Thanks.

[16-Apr-2010 15:26:00] RoAk is now known as RoAkSoAx

[16-Apr-2010 16:24:11] <Simon4> so any of you US people want to send me an ipad? ;)

[16-Apr-2010 17:11:30] arodriguez is now known as RoAkSoAx

[16-Apr-2010 17:17:26] RoAk is now known as prueba

[16-Apr-2010 17:27:14] RoAk is now known as testing

[16-Apr-2010 17:27:28] testing is now known as testintnt

[16-Apr-2010 20:19:16] forsberg is now known as fOrsberg

[16-Apr-2010 22:15:14] RoAk is now known as andreserl

[16-Apr-2010 22:26:30] RoAkSoAx is now known as andreserl

[16-Apr-2010 22:29:52] andreserl is now known as roaksoax

[16-Apr-2010 22:33:39] roaksoax is now known as andreserl

[17-Apr-2010 00:00:30] [disconnected at Sat Apr 17 00:00:30 2010]

[17-Apr-2010 00:00:30] [connected at Sat Apr 17 00:00:30 2010]

[17-Apr-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[17-Apr-2010 01:44:11] fOrsberg is now known as forsberg

[17-Apr-2010 03:48:26] forsberg is now known as fOrsberg

[17-Apr-2010 14:35:06] RoAkSoAx_ is now known as RoAkSoAx

[17-Apr-2010 14:46:44] rhettardo_ is now known as rhettardo

[17-Apr-2010 20:26:58] <sloof3> Once again I lose all my perf data for an unknown reason

[18-Apr-2010 00:00:30] [disconnected at Sun Apr 18 00:00:30 2010]

[18-Apr-2010 00:00:30] [connected at Sun Apr 18 00:00:30 2010]

[18-Apr-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[18-Apr-2010 06:42:25] fOrsberg is now known as forsberg

[18-Apr-2010 19:37:40] forsberg is now known as fOrsberg

[18-Apr-2010 20:58:13] bzed_ is now known as bzed

[19-Apr-2010 00:00:30] [disconnected at Mon Apr 19 00:00:30 2010]

[19-Apr-2010 00:00:31] [connected at Mon Apr 19 00:00:31 2010]

[19-Apr-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[19-Apr-2010 01:50:39] dgc_ is now known as dgc

[19-Apr-2010 04:26:04] fOrsberg is now known as forsberg

[19-Apr-2010 07:04:04] <blull_> morning

[19-Apr-2010 07:04:09] <blull_> saw a weird thing with zenoss this morning

[19-Apr-2010 07:04:21] <blull_> it auto discovered a device that was already in zenoss.

[19-Apr-2010 07:04:45] <blull_> apparently, if DNS is broken during a zendisc run, it will add hosts that are already added but associated with a dns name?

[19-Apr-2010 07:04:48] <blull_> is that a known bug?

[19-Apr-2010 07:08:28] <cluther> blull_: That's a possibility for devices that are multi-homed where Zenoss is unable to model the network interfaces. Basically Zenoss has duplication detection based on name and IP address.. if duplicates can't be found for either of those, Zenoss assumes it is a different device.

[19-Apr-2010 07:08:44] <blull_> right

[19-Apr-2010 07:08:46] <blull_> but

[19-Apr-2010 07:09:22] <blull_> rmm nm

[19-Apr-2010 09:16:33] <blull_> hey

[19-Apr-2010 09:16:36] <blull_> under zproperties

[19-Apr-2010 09:16:49] <blull_> there are a few different possible properties that may control timeouts of checks

[19-Apr-2010 09:16:57] <blull_> specifically I'm trying to control the timeout to icmp

[19-Apr-2010 09:17:01] <blull_> and some tcp port checks

[19-Apr-2010 09:17:07] <blull_> that I run as external COMMAND's

[19-Apr-2010 09:17:18] <blull_> CommandCommandTimeout <-- is that for remote COMMAND templates?

[19-Apr-2010 09:17:28] <blull_> what is CommandLoginTimeout <-- is that for ssh sessions?

[19-Apr-2010 09:17:53] <blull_> What is zProdStateThreshold?

[19-Apr-2010 09:18:17] <blull_> and how is this all affeced by zCollectorClientTimeout?

[19-Apr-2010 09:22:00] <blull_> zCollectorClientTimeout     int     Allows you to set the timeout time of the collector client in seconds

[19-Apr-2010 09:22:02] <blull_> so

[19-Apr-2010 09:22:21] <blull_> does that mean its a global max execution for EVERYTHING in total for the lifetime of the current profiling of the client?

[19-Apr-2010 09:25:47] <blull_> also

[19-Apr-2010 09:25:53] <blull_> if my cycletime is set to say 60 in zproperties

[19-Apr-2010 09:26:05] <blull_> but the cycletime in the template is < zproperties cyceltime

[19-Apr-2010 09:26:08] <blull_> which one wins out?

[19-Apr-2010 09:26:15] <blull_> I'm thinking template?

[19-Apr-2010 09:40:08] <rmatte> blull_: you basically correctly answered each of your questions

[19-Apr-2010 09:52:14] <blull_> sweet.

[19-Apr-2010 09:52:16] <blull_> so in otherwords

[19-Apr-2010 09:52:49] <blull_> templates run for the specified # of cycle seconds specified in the template, as long as it does not exceed the time specified in the zproperties, and "thresholds" are ALWAYS specified in the template correct?

[19-Apr-2010 09:53:15] <blull_> just making sure cause I'm moving out main prod datacenter over to zenoss today and turning off nagios for it, so I wan to make sure 100% I understood that correctly.

[19-Apr-2010 10:21:12] <evilliger> question regarding zenoss core and mysql.

[19-Apr-2010 10:21:46] <evilliger> is anyone running mysql for zenoss on a seperate server?

[19-Apr-2010 10:22:01] <Simon4> I do

[19-Apr-2010 10:22:28] <venturaville> same here

[19-Apr-2010 10:22:51] <evilliger> is it simple.

[19-Apr-2010 10:23:09] <evilliger> how's the performance on your zenoss server WRT web UI?

[19-Apr-2010 10:26:33] <rmatte> blull_: correct

[19-Apr-2010 10:26:35] <Simon4> it's simple

[19-Apr-2010 10:26:49] <Simon4> there's a place in the gui to set the hostname of your mysql server

[19-Apr-2010 10:26:53] <mistich> have a question if you set a event transform at the top level in 2.5.2 it populates down to other events.

[19-Apr-2010 10:27:04] <Simon4> and our performance increased, since we took load off the disks on the zenoss server by doing it

[19-Apr-2010 10:27:05] <rmatte> mistich: correct

[19-Apr-2010 10:27:37] <mistich> so if that is true and you set a transform in /events/perf which one does it follow

[19-Apr-2010 10:28:08] <rmatte> mistich: whichever is first in the sequence

[19-Apr-2010 10:28:25] <rmatte> there's a sequence order that you can set

[19-Apr-2010 10:29:01] <mistich> not in the transforms

[19-Apr-2010 10:29:18] <rmatte> well, are you talking about mapping or literally More -> Transform?

[19-Apr-2010 10:29:23] <mistich> I'm talking about the class level

[19-Apr-2010 10:29:27] <rmatte> I see

[19-Apr-2010 10:29:33] <mistich> More -> Transforms

[19-Apr-2010 10:29:41] <rmatte> It would apply the one on the class itself I believe

[19-Apr-2010 10:29:55] <rmatte> generally you try not to have conflicting transforms though by adding if statements

[19-Apr-2010 10:29:56] <mistich> I'll test it just wanted to check first

[19-Apr-2010 10:30:58] <mistich> oh yes but want to turn off alerting this way I have a alert rule if class = networking send mail and page

[19-Apr-2010 10:31:17] <rmatte> why would you do that with a transform?

[19-Apr-2010 10:31:31] <rmatte> that's what alert filters are for

[19-Apr-2010 10:31:35] <mistich> so I want to set class to maintance  in the transform and this will bypass the alert rule

[19-Apr-2010 10:32:08] <rmatte> based on what?

[19-Apr-2010 10:32:36] <mistich> no I already have a filter but during a maintenance window want to shut off all emails and paging easily

[19-Apr-2010 10:33:19] <rmatte> I assume you've seen this: blogs/zenossblog/2009/12/23/tip-of-the-month-dynamically-creating-maintenance-windows ?

[19-Apr-2010 10:33:56] <mistich> so currently I only email and page on if class == networking I should be able to set a transform to set class = maintenance

[19-Apr-2010 10:34:22] <rmatte> Why are you just adding another filter based on the device class?

[19-Apr-2010 10:34:32] <rmatte> erm, why are you not, rather

[19-Apr-2010 10:34:58] <rmatte> can you please explain the scenario and your end goal?

[19-Apr-2010 10:35:03] <rmatte> I'm not getting the big picture here

[19-Apr-2010 10:35:25] <mistich> Goal shut off all emails and pages during a time frame

[19-Apr-2010 10:35:54] <rmatte> ok, and this timeframe is daily recurring?

[19-Apr-2010 10:36:05] <Simon4> mistich: so you still want the events, just not the alerting?

[19-Apr-2010 10:36:12] <mistich> yes

[19-Apr-2010 10:36:17] <mistich> Simon4 yes

[19-Apr-2010 10:36:22] <rmatte> Simon4: a maintenance window would do that wouldn't it?

[19-Apr-2010 10:36:45] <rmatte> set a maintenance window to move devices to Maintenance during a timeframe each day

[19-Apr-2010 10:36:52] <venturaville> rmatte: BTW adding stats to the ESX driver to libvirt is going to be troublesome

[19-Apr-2010 10:36:56] <mistich> but to put 11107 devices in maintenance kills zope

[19-Apr-2010 10:37:00] <rmatte> then add a filter to the alerting rule to only alert on devices if they are in a state of Production

[19-Apr-2010 10:37:11] <rmatte> venturaville: that sucks lol

[19-Apr-2010 10:37:21] <rmatte> mistich: I see

[19-Apr-2010 10:37:26] <mistich> the problem is when I put all 11107 devices in maintenance kills zope

[19-Apr-2010 10:37:31] <mistich> see the problem

[19-Apr-2010 10:37:32] <Simon4> you could take a hammer to it, and kill zenactions maybe

[19-Apr-2010 10:37:36] <rmatte> that's a lot of devices, hmmm

[19-Apr-2010 10:37:41] <Simon4> since that's what sends the emails/etc

[19-Apr-2010 10:37:44] <Simon4> it's ugly though :)

[19-Apr-2010 10:38:34] <mistich> yeah but will they Que up

[19-Apr-2010 10:38:45] <rmatte> mistich: ummm, ok, here's a question for you...

[19-Apr-2010 10:38:51] <mistich> shoot

[19-Apr-2010 10:39:00] <Simon4> mistich: not sure, not hard to test though

[19-Apr-2010 10:39:00] <rmatte> have you considered using the Schedule tab on the alerting rule to do this?

[19-Apr-2010 10:39:12] <rmatte> you can schedule when the rule is active or not daily

[19-Apr-2010 10:39:20] <rmatte> which will accomplish what you want without all the mess

[19-Apr-2010 10:39:42] <mistich> ahhhhh there you go will have to look at that

[19-Apr-2010 10:39:55] <rmatte> ;)

[19-Apr-2010 10:40:17] <rmatte> no point in reinventing the wheel

[19-Apr-2010 10:41:49] <mistich> only bad thing there are 5 zenoss servers with 16 groups each and 2 alerts per group = 160 schedules to set

[19-Apr-2010 10:42:19] <rmatte> hmmm

[19-Apr-2010 10:42:26] <mistich> turning off send actions seems the easiest

[19-Apr-2010 10:42:39] <rmatte> zenactions you mean

[19-Apr-2010 10:42:58] <mistich> but the schedule is cleaner

[19-Apr-2010 10:43:05] <rmatte> the schedule is much cleaner

[19-Apr-2010 10:43:07] <mistich> yes zenactions send actions  :)

[19-Apr-2010 10:43:11] * Simon4 agrees

[19-Apr-2010 10:43:19] <Simon4> and it's a schedule per alerting rule, not per device

[19-Apr-2010 10:43:21] <rmatte> maybe you can figure out how to use zendmd to add schedules in bulk

[19-Apr-2010 10:43:27] <Simon4> ah, read your comment again

[19-Apr-2010 10:43:33] <Simon4> you can edit alerting rules with zendmd

[19-Apr-2010 10:43:42] * Simon4 has messed with that some in the past

[19-Apr-2010 10:43:47] <mistich> thats what flunky's are for

[19-Apr-2010 10:43:53] <rmatte> well, you can do anything with zendmd, the question is how to code it to do it in bulk

[19-Apr-2010 10:44:27] <mistich> yes zendmd will work but more fun to have the new guy do it :)

[19-Apr-2010 10:44:30] <Simon4> I probably have code that would be close, or could find it fairly quickly

[19-Apr-2010 10:44:45] * Simon4 wrote code ages ago to migrate users/groups and rules from one install to another

[19-Apr-2010 10:44:51] <Simon4> so I must have walked every rule

[19-Apr-2010 10:45:21] <Simon4> if new guy has issues give me a yell and I can dig something out

[19-Apr-2010 10:45:33] <rmatte> mistich: having the new guy do it leaves the door open for human error

[19-Apr-2010 10:45:41] <rmatte> especially when it's being done 160 times

[19-Apr-2010 10:45:57] <mistich> I know just giving him a hard time

[19-Apr-2010 10:46:01] <rmatte> lol

[19-Apr-2010 10:46:02] eric is now known as Guest40826

[19-Apr-2010 10:46:14] <Guest40826> Guest40826 doesn't appreciate that, I'm sure...

[19-Apr-2010 10:46:31] <mistich> oh trust me I know it has burned us several times by doing it manually

[19-Apr-2010 10:46:31] <rmatte> too bad there's no support for a global alerting schedule

[19-Apr-2010 10:46:41] <mistich> zendmd is my friend

[19-Apr-2010 10:47:01] <Simon4> rmatte: yeah - I've wanted that a lot also

[19-Apr-2010 10:47:14] <Simon4> that or at least "turn all alerts off", with ability to reenable

[19-Apr-2010 10:47:22] <rmatte> well, we'll have to wait for the new UI before any new features like that get coded

[19-Apr-2010 10:47:38] <Simon4> need to write some dmd script to do that, just need to save state of which alerts were enabled when you disable em

[19-Apr-2010 10:47:52] <Simon4> since we have a bunch kicking around under user accounts that are disabled in live

[19-Apr-2010 10:48:00] <mrayzenoss> seems like a report similar to the Notification Schedules report would be nice

[19-Apr-2010 10:48:01] <rmatte> ah

[19-Apr-2010 10:48:51] <mistich> not much too it just need time to do it

[19-Apr-2010 10:49:27] <mistich> will look into it more thanks folks for the sound board

[19-Apr-2010 10:49:39] <rmatte> np

[19-Apr-2010 11:33:59] <Lazesharp> hi guys, anyone have any experience setting up monitoring for Java JMX attributes?

[19-Apr-2010 12:29:33] <davetoo> 'morning

[19-Apr-2010 12:38:21] forsberg is now known as fOrsberg

[19-Apr-2010 12:46:03] <chemist> rmatte: hiya, how do I run the zenpack you sent me?

[19-Apr-2010 12:46:22] <chemist> I've installed it on the master

[19-Apr-2010 12:48:19] <evilliger> is anyone running zenoss core on mysql 5.1?

[19-Apr-2010 12:59:46] <rmatte> chemist: which one?

[19-Apr-2010 12:59:52] <rmatte> chemist: the collector tool one?

[19-Apr-2010 13:00:15] <chemist> rmatte: yes

[19-Apr-2010 13:00:29] <rmatte> what version did you install?

[19-Apr-2010 13:00:32] <rmatte> 1.0?

[19-Apr-2010 13:00:36] <chemist> yes

[19-Apr-2010 13:00:42] <chemist> you have a newer one?

[19-Apr-2010 13:00:42] <rmatte> Download 1.1: docs/DOC-5831

[19-Apr-2010 13:00:47] <chemist> cool

[19-Apr-2010 13:00:50] <rmatte> it has a bugfix in it

[19-Apr-2010 13:01:01] <rmatte> then you need to install it on both the master and the collector

[19-Apr-2010 13:01:16] <rmatte> then as the zenoss user execute zencollectortool on the master and select option 1

[19-Apr-2010 13:01:31] <rmatte> then execute zencollectortool on the collector and select option 2 then follow/answer the prompts

[19-Apr-2010 13:01:44] <chemist> cool, thanks

[19-Apr-2010 13:01:47] <rmatte> make sure the same ZenPacks that are installed on the master are installed on the collector before selecting option 2

[19-Apr-2010 13:02:16] <chemist> ok, what happens if I want to install zenpacks afterwards?

[19-Apr-2010 13:02:52] <rmatte> chemist: you install them on the master and then copy the zenpack directory from $ZENHOME/ZenPacks over to $ZENHOME/ZenPacks on the collector

[19-Apr-2010 13:03:10] <chemist> just a straight copy?

[19-Apr-2010 13:03:22] <chemist> so I could rsync it

[19-Apr-2010 13:03:28] <rmatte> correct

[19-Apr-2010 13:03:33] <chemist> excellent

[19-Apr-2010 13:03:40] <rmatte> that's how Egor's new collector pack will do it once it's released

[19-Apr-2010 13:04:05] <chemist> ok, I'll give this a go

[19-Apr-2010 13:04:08] <rmatte> k cool

[19-Apr-2010 13:04:21] <bigegor> chemist: do not foget to create symlinks for ZenPacks daemons

[19-Apr-2010 13:04:41] <rmatte> yeh, egor makes a good point

[19-Apr-2010 13:04:42] <chemist> on the collectors, right?

[19-Apr-2010 13:04:47] <rmatte> yeh, on the collector

[19-Apr-2010 13:05:06] <rmatte> any non-standard daemons will need symlinks manually created for them

[19-Apr-2010 13:05:54] <chemist> is it advisable to have a fresh install on the collector?

[19-Apr-2010 13:06:09] <rmatte> chemist: doesn't really matter

[19-Apr-2010 13:06:14] <chemist> cool

[19-Apr-2010 13:06:20] <rmatte> as long as the ZenPacks are the same

[19-Apr-2010 13:07:21] <rmatte> just realize that if you're monitoring devices on what is to be the collector you'll need to remove them and re-add them after the conversion

[19-Apr-2010 13:07:29] <rmatte> since the master will be the devices database at that point

[19-Apr-2010 13:07:37] <bigegor> if someone want try my ZenPack, its now in SVN.

[19-Apr-2010 13:07:50] <chemist> yeah,I thought that might be the case

[19-Apr-2010 13:08:01] <bigegor> http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.DistributedCollectors

[19-Apr-2010 13:08:13] <rmatte> all a remote collector is is a bunch of monitoring daemons that connect back to the zenhub on the master

[19-Apr-2010 13:08:55] <rmatte> Updates to my windows SNMP performance ZenPacks have been released: docs/DOC-3386

[19-Apr-2010 13:08:58] <rmatte> docs/DOC-3570

[19-Apr-2010 13:09:44] <chemist> bigegor: is that the same as rmatte's?

[19-Apr-2010 13:10:41] <bigegor> chemist: no, its my with Zenoss UI

[19-Apr-2010 13:10:59] <chemist> where do I find the .egg file?

[19-Apr-2010 13:12:34] * chemist doesn't know how to install from source

[19-Apr-2010 13:15:15] <bigegor> chemist: docs/DOC-5861

[19-Apr-2010 13:15:51] <chemist> thanks :)

[19-Apr-2010 13:15:59] <bigegor> np

[19-Apr-2010 13:17:58] <bigegor> chemist: you need to set up ssh public key authentication on remote collector

[19-Apr-2010 13:25:00] <citrus2> i notice that my zenoss box consume a lot of memory after a few days and starts to act sluggish in the web portal,  is there a way to keep this under control a bit better?

[19-Apr-2010 13:25:14] <rmatte> citrus2: what version are you running?

[19-Apr-2010 13:25:51] <citrus2> 2.5.1  i believe

[19-Apr-2010 13:26:12] <rmatte> upgrade to 2.5.2 for starters, it has lots of bug fixes

[19-Apr-2010 13:26:22] <rmatte> how much RAM do you have in the server?

[19-Apr-2010 13:26:30] <citrus2> 1.5 GB

[19-Apr-2010 13:26:39] <citrus2> 25 monitored machines

[19-Apr-2010 13:26:53] <rmatte> what type of monitoring?

[19-Apr-2010 13:27:07] <citrus2> just ping right now  but it grabs snmp and wmi info i believe

[19-Apr-2010 13:27:15] <citrus2> ping alerting i should say

[19-Apr-2010 13:27:32] <rmatte> so all you're doing is ping monitoring?

[19-Apr-2010 13:27:42] <rmatte> oh, you only alert on ping you mean?

[19-Apr-2010 13:27:53] <rmatte> but you're doing full SNMP and WMI monitoring

[19-Apr-2010 13:28:04] <citrus2> yes

[19-Apr-2010 13:28:16] <rmatte> how many days before you start to notice performance issues?

[19-Apr-2010 13:28:24] <citrus2> about 7 or 8

[19-Apr-2010 13:28:46] <chemist> is there a default password for the user zenoss?

[19-Apr-2010 13:28:48] <rmatte> I don't have any Zenoss servers with less than 2GB of RAM since I've found that's really the bare minimum

[19-Apr-2010 13:28:54] <rmatte> you're kind of pushing it with only having 1.5

[19-Apr-2010 13:29:17] <rmatte> chemist: no, it's randomized

[19-Apr-2010 13:29:26] <rmatte> chemist: you have to set one if you need one

[19-Apr-2010 13:29:30] <citrus2> yeah it started as a virtual machine, and still is until i can get some dedicated hardware for it

[19-Apr-2010 13:29:41] <chemist> ok, and it wont break anything?

[19-Apr-2010 13:29:43] <rmatte> chemist: though you're supposed to use a public/private key which doesn't involve a password

[19-Apr-2010 13:30:10] <chemist> I'll just copy the keys over

[19-Apr-2010 13:31:09] <rmatte> just copy it over to a regular user account and then copy it from there as the zenoss user

[19-Apr-2010 13:31:33] <rmatte> chemist: and no, you wouldn't break anything by changing the zenoss user password

[19-Apr-2010 13:31:41] <rmatte> nothing relies on the password

[19-Apr-2010 13:40:00] <bigegor> chemist: do not set zenoss user password

[19-Apr-2010 13:40:50] <rmatte> bigegor: he's not, he's just copying the key

[19-Apr-2010 14:02:25] <citrus2> anyone in here use cacti? i am trying to decide if i need it or not since zenoss has built in graphs

[19-Apr-2010 14:03:57] <venturaville> cacti+nagios is duplicated functionality for zenoss

[19-Apr-2010 14:04:30] <venturaville> in short you don't need cacti

[19-Apr-2010 14:09:27] <rmatte> yeh, definitely don't need cacti with Zenoss

[19-Apr-2010 14:18:13] <chemist> bigegor: if the device I want to monitor with the remote collector already exists on the master, can I just change the collector property?

[19-Apr-2010 14:18:28] <rmatte> chemist: no, you need to copy the RRD data over as well

[19-Apr-2010 14:18:48] <rmatte> $ZENHOME/perf

[19-Apr-2010 14:19:58] <bigegor> chemist: rmatte is right

[19-Apr-2010 14:20:19] <chemist> copy the whole devicename folder over?

[19-Apr-2010 14:20:20] <rmatte> bigegor: wonder if there's a way to get your pack to do that automatically

[19-Apr-2010 14:20:25] <rmatte> chemist: yes

[19-Apr-2010 14:20:28] <chemist> cool

[19-Apr-2010 14:20:52] <bigegor> its not so easy :(

[19-Apr-2010 14:21:13] <rmatte> bigegor: or maybe you could just add a form where the person can add the device name, click submit, and it copies the data

[19-Apr-2010 14:21:22] <rmatte> wouldn't exactly be automated, but would be nicer than doing it by hand

[19-Apr-2010 14:22:08] <chemist> so, copy the RRD data over, then change the collector property?

[19-Apr-2010 14:22:19] <rmatte> chemist: yes

[19-Apr-2010 14:22:30] <chemist> seems the logic way, moot point really :)

[19-Apr-2010 14:22:30] <rmatte> then push changes

[19-Apr-2010 14:22:32] <bigegor> right way is to extend 'change monitor' function

[19-Apr-2010 14:22:45] <rmatte> bigegor: ah

[19-Apr-2010 14:23:36] <bigegor> rmatte: i will look, if its possible

[19-Apr-2010 14:23:43] <rmatte> bigegor: cool

[19-Apr-2010 14:24:37] <chemist> also, can someone tell me if there is a better way to monitor vpn tunnels? at the moment I monitor them by pinging the endpoints

[19-Apr-2010 14:25:08] <rmatte> IPSEC tunnels with no interfaces associated with them?

[19-Apr-2010 14:25:23] <chemist> I think so

[19-Apr-2010 14:25:36] <rmatte> You can enable traps or syslogs to alert on status change

[19-Apr-2010 14:25:39] <rmatte> that's what we do with ours

[19-Apr-2010 14:25:54] <rmatte> Cisco devices?

[19-Apr-2010 14:25:59] <chemist> I wish

[19-Apr-2010 14:26:03] <rmatte> lol

[19-Apr-2010 14:26:21] <rmatte> what kind?

[19-Apr-2010 14:26:39] <chemist> Astaro ASG here, going to Draytek routers at the remote sites

[19-Apr-2010 14:27:08] <felbane> Hello, question about creating a command datasource

[19-Apr-2010 14:27:48] <felbane> How can I reference my zenpack libexec path in the command definition?

[19-Apr-2010 14:29:34] <chemist> (I hate Astaros)

[19-Apr-2010 14:33:29] <rmatte> felbane: one sec...

[19-Apr-2010 14:34:29] <rmatte> So if your ZenPack was called ZenPacks.Nova.Windows.SNMPPerfMonitor.Simple and the script were called winmem.sh you'd do: ${here/ZenPackManager/packs/ZenPacks.Nova.Windows.SNMPPerfMonitor.Simple/path}/libexec/winmem.sh

[19-Apr-2010 14:34:44] <rmatte> and you can obviously add options:

[19-Apr-2010 14:34:45] <rmatte> ${here/ZenPackManager/packs/ZenPacks.Nova.Windows.SNMPPerfMonitor.Simple/path}/libexec/winmem.sh "${dev/manageIp}" "${dev/zSnmpCommunity}"

[19-Apr-2010 14:35:06] <felbane> Excellent. Let me try that

[19-Apr-2010 14:35:10] <rmatte> k

[19-Apr-2010 14:38:58] <felbane> Ok, that works. Now a followup

[19-Apr-2010 14:39:36] <felbane> My zenpack has created relationships with the 'os' component of a device.

[19-Apr-2010 14:40:06] <felbane> Fx, device/myserver/os/msmq

[19-Apr-2010 14:40:58] <felbane> What I'm hoping to do is set up the template so that RRDs will be created for each of the 'monitored' queues

[19-Apr-2010 14:41:39] <felbane> So how would I tell the data source to use the id of the queue

[19-Apr-2010 14:42:15] <felbane> ${here/os/msmq/id} ?

[19-Apr-2010 14:42:40] <felbane> Or ${here/os/msmq/myqueue/id} ?

[19-Apr-2010 14:42:59] <felbane> Or am I going about this wrong (do I need to create my own datasource)

[19-Apr-2010 14:43:33] <bigegor> felbane: if you want bind template to component use ${here/id}

[19-Apr-2010 14:44:59] <felbane> Ok so, in the component field I canput ZenPacks.community.MsmqMonitor.MessageQueue and then ${here/id} will reference each queue?

[19-Apr-2010 14:46:01] <bigegor> felbane: yes, if you want reference device use ${dev/something}

[19-Apr-2010 14:50:53] <felbane> Hmm. It's not behaving as expected. I might just have to post my code on the forums. Bigegor, I've been using your ODBC zenpacks as a reference... going to look in the objects.xml again to see if there's something I'm doing wrong

[19-Apr-2010 14:52:16] <bigegor> felbane: ok, i will look

[19-Apr-2010 14:53:05] <felbane> I will give a link when I have it posted

[19-Apr-2010 15:07:41] <felbane> Ok message/47848

[19-Apr-2010 15:08:44] <felbane> I have the modeler working, the queue entries are being created under os

[19-Apr-2010 15:09:14] <felbane> Some of the skins files are not fully complete because I have been trying to get the RRD created

[19-Apr-2010 15:11:00] <bigegor> felbane: i've got it.

[19-Apr-2010 15:11:58] <felbane> Thanks for looking at it

[19-Apr-2010 15:14:14] <bigegor> felbane: you can try to use WMIDataSource, look at new WMIPerf_Windows fro WMI modelers plugins

[19-Apr-2010 15:16:22] <felbane> I looked in to that but (1) I'm stuck using 2.4.5 for now and also it needs to be very very very easy for the user to set up which is why I wanted to do the table-style in 'os' tab.

[19-Apr-2010 15:19:15] <bigegor> felbane:WMIDataSource 1.5 works with Zenoss 2.4.5

[19-Apr-2010 15:19:31] <felbane> Also my company (stupidly) wants as few daemons as possible on our collectors

[19-Apr-2010 15:19:46] <felbane> Oh I was only looking at the most recent release

[19-Apr-2010 15:19:48] <bigegor> felbane: ok

[19-Apr-2010 15:20:33] <felbane> I don't agree with their decisions but I have to do what the boss says. At least it let's me do fun stuff like making zenpacks

[19-Apr-2010 15:20:36] <felbane> :-)

[19-Apr-2010 15:23:12] <felbane> If I could figure out how to test my datasource (maybe in zendmd) I could try to see why its not creating RRDs

[19-Apr-2010 15:23:50] <felbane> I'm just not sure how. It imports ok

[19-Apr-2010 15:25:22] <bigegor> felbane: one more question. why not using Command DataSource?

[19-Apr-2010 15:26:49] RoAkSoAx is now known as RoAk

[19-Apr-2010 15:27:29] <felbane> Bigegor: my understanding is that if I made it with command datasource (like a cacti/nagios plugin) then the users would have to go to each device and make the templates

[19-Apr-2010 15:27:41] <chemist> when using remote collectors, does the master have to have full (firewall) access to a device added to the collector? or just the collector needs access?

[19-Apr-2010 15:27:41] <felbane> The queues are different on evry machine

[19-Apr-2010 15:28:30] <felbane> So this way they should be able to go to device, check boxes, and enable monitoring just like win services. Simple

[19-Apr-2010 15:28:36] RoAkSoAx_ is now known as RoAkSoAx

[19-Apr-2010 15:29:00] <cluther> felbane: You can do that with a COMMAND datasource without adding a new collector daemon. That's how the enterprise MSMQ ZenPack does it.

[19-Apr-2010 15:29:10] <cluther> Well.. not with a COMMAND data source, but the idea is the same.

[19-Apr-2010 15:30:26] <cluther> chemist: Discovery/modeling (via the UI) is still done from the master. So it's best to make sure the master has access via the same ports as the collector if at all possible. Otherwise you have to do some things via the command line on the collector.

[19-Apr-2010 15:31:08] <felbane> Cluther: that's where I'm stuck. I'm not sure how to get either command datasource or my custom datasource to create the rrds

[19-Apr-2010 15:32:02] <cluther> You've already extended the model and have the queues modeled?

[19-Apr-2010 15:32:23] <felbane> Does anybody have tips on how to test out a custom datasource? I can import it in zendmd without errors but I don't know how to get it to make a test run

[19-Apr-2010 15:32:30] <felbane> Yes the queues are modeling fine

[19-Apr-2010 15:32:56] <felbane> Its just the queue -> collector jump that I'm having trouble with

[19-Apr-2010 15:34:55] <cluther> You've put your datasource in a template named "MessageQueue" and not bound it to anything, right?

[19-Apr-2010 15:35:10] <felbane> It is not creating the RRDs in $ZENHOME/perf/Devices/device/os/msmq/

[19-Apr-2010 15:35:36] <cluther> felbane: When you run zencommand with -v10 do you even see it executing your commands?

[19-Apr-2010 15:35:40] <felbane> Correct

[19-Apr-2010 15:35:45] <felbane> Let me try that

[19-Apr-2010 15:36:40] <citrus2> can i use data from cacti to replace date graphs in zenoss?

[19-Apr-2010 15:36:57] <chemist> cluther: thanks

[19-Apr-2010 15:36:58] <citrus2> date*

[19-Apr-2010 15:37:14] <bigegor> cluther: is CommandParser class was in 2.4.5?

[19-Apr-2010 15:37:39] <cluther> bigegor: Yeah.

[19-Apr-2010 15:38:54] <felbane> Cluther: when I run zencommand -v10 it is executing the data source for each queue but it gives error

[19-Apr-2010 15:39:06] <felbane> Code: 1 msg: general error

[19-Apr-2010 15:39:21] <cluther> Great! Then you've made it through the hardest parts.

[19-Apr-2010 15:39:28] <felbane> Lol

[19-Apr-2010 15:39:51] <bigegor> felbane: you can use Command DataSource + custom CommandParser (for wmic output parsing)

[19-Apr-2010 15:40:53] <felbane> Oh my goodness I feel really really silly right now

[19-Apr-2010 15:40:57] <cluther> Right.. what bigegor said. Your data source is likely using the Auto parser which thinks your command's output looks like Cacti.

[19-Apr-2010 15:41:46] <felbane> So I originally had my script returning in the format myquename: 2

[19-Apr-2010 15:42:17] <felbane> Ie "%s: %s" % (queuename,queuecount)

[19-Apr-2010 15:42:47] <felbane> I changed it when I went to the command datasource to "queueCount: %s"

[19-Apr-2010 15:43:05] <felbane> But I didn't remove "queuename" from the tuple

[19-Apr-2010 15:43:11] <felbane> Let me see if this works now

[19-Apr-2010 15:44:50] <cluther> felbane: If you're going to use Cacti formatted output from your script like that you should be sure to always exit with code 0 or you'll get the "general error" messages.

[19-Apr-2010 15:47:01] * felbane does a little dance

[19-Apr-2010 15:47:30] <felbane> Okay so zencommand run -v10 is showing the commands executing okay

[19-Apr-2010 15:48:03] <felbane> Now I guess I just need to figure out how to get the count to display properly on the table

[19-Apr-2010 15:54:55] <bigegor> chemist: is DistributedCollector Zenpack working for you?

[19-Apr-2010 15:57:42] <felbane> Well actually now I need to make sure my datasource is set up to inject the right type (Gauge, not counter d'oh)

[19-Apr-2010 15:58:10] <felbane> Bigegor and cluther: thank you very much for the help so far. I might have more questions later :-)

[19-Apr-2010 15:59:34] <chemist> bigegor: I think so, just trying to find devices I can access fromboth locations

[19-Apr-2010 16:08:46] <mrayzenoss> citrus2: I think you could theoretically pull over your Cactii RRD files if the create settings were exactly the same and they were named properly

[19-Apr-2010 16:09:08] <mrayzenoss> also have to be the same 32/64-bit

[19-Apr-2010 16:21:02] RoAkSoAx_ is now known as RoAk

[19-Apr-2010 16:21:46] RoAk is now known as RoAkSoAx

[19-Apr-2010 16:24:55] <chemist> bigegor: http://pastebin.ca/1868545

[19-Apr-2010 16:25:20] <chemist> I can snmpwalk the device from the master and the collector

[19-Apr-2010 16:26:14] <bigegor> chemist: restart zenoss

[19-Apr-2010 16:34:20] <chemist> that's better :) I restarted zope earlier, but not zenoss

[19-Apr-2010 16:42:07] <rmatte> I always restart Zenoss after install ZenPacks

[19-Apr-2010 16:42:15] <rmatte> Restarting Zope rarely works

[19-Apr-2010 16:42:43] <chemist> I should have known

[19-Apr-2010 16:45:17] <rmatte> well, 2.5.2 still appears to have a memory leak but not nearly as bad as 2.4.5 did

[19-Apr-2010 16:45:56] <rmatte> I look forward to the day when all of the memory leaks are gone

[19-Apr-2010 16:48:06] * HCoyote turns off rmatte's computer.  "There you go!  No more memory leaks!"  :-D

[19-Apr-2010 16:48:24] <rmatte> lol

[19-Apr-2010 16:48:48] <rmatte> well, I'm out of here, later

[19-Apr-2010 16:52:33] <davidX-> ahhh

[19-Apr-2010 16:52:49] <davidX-> listening to our cfo try to understand server density per rack

[19-Apr-2010 16:52:51] <davidX-> yessss

[19-Apr-2010 16:53:08] <davidX-> hi HCoyote

[19-Apr-2010 16:53:15] <HCoyote> hey man.  ltns.

[19-Apr-2010 16:53:22] <davidX-> no doubt

[19-Apr-2010 16:53:37] <HCoyote> still hanging out at layeredtech?

[19-Apr-2010 16:53:46] <davidX-> still hanging out at layeredtech ;)

[19-Apr-2010 16:53:51] <HCoyote> :-)

[19-Apr-2010 16:53:52] <davidX-> you?

[19-Apr-2010 16:54:00] <HCoyote> I'm at UT

[19-Apr-2010 16:54:02] <HCoyote> for now.

[19-Apr-2010 16:54:04] <HCoyote> :-)

[19-Apr-2010 16:54:08] <davidX-> awesome

[19-Apr-2010 16:54:11] <davidX-> laid back?

[19-Apr-2010 16:54:25] <HCoyote> heh.

[19-Apr-2010 16:55:01] <HCoyote> some days, it's like giving a football to a monkey and watching it rape the pigskin.  Other days, it's just bad.

[19-Apr-2010 16:55:36] <HCoyote> lots of churn, lots of confusion, lots of chaos.

[19-Apr-2010 16:56:11] <davidX-> ahh, for now indeed then..

[19-Apr-2010 16:56:36] <davidX-> lots of churn is not fun.

[19-Apr-2010 16:56:47] <HCoyote> Not a place for someone with a heavy background in corporate IT work ... where, you know, you have to understand a bit about goals, timelines, budgets, "the bottom line".

[19-Apr-2010 16:56:50] <davidX-> i can deal with a lot of things but that can be difficult

[19-Apr-2010 16:56:55] <HCoyote> yeah.

[19-Apr-2010 16:57:15] <davidX-> is EA hiring systems people down there yet?

[19-Apr-2010 16:57:19] <HCoyote> yeah

[19-Apr-2010 16:57:31] <HCoyote> and rumors that facebook is coming too

[19-Apr-2010 16:59:38] <davidX-> EA would be fun

[19-Apr-2010 16:59:52] <davidX-> i spent a lot of time last year working with a gaming studio

[19-Apr-2010 16:59:54] <HCoyote> I hear EA is a righteously evil place to work ...

[19-Apr-2010 16:59:57] <davidX-> nice culture

[19-Apr-2010 17:00:09] <davidX-> i've heard that and heard it's amazing

[19-Apr-2010 17:00:17] <davidX-> i think it depends what tiny umbrella you're under

[19-Apr-2010 17:01:26] <HCoyote> possibly.

[19-Apr-2010 17:07:27] <davidX-> well, if i hear of anything decent down there i'll ping you

[19-Apr-2010 17:07:41] <HCoyote> sweet. :-)

[19-Apr-2010 17:07:41] <davidX-> quite a bit of 'cloud' positions open down there.

[19-Apr-2010 17:07:42] <HCoyote> thanks.

[19-Apr-2010 17:08:21] <HCoyote> that'd be right up my alley I think.

[19-Apr-2010 17:42:39] <felbane> Bigegor, you still around?

[19-Apr-2010 17:42:51] <bigegor> yes

[19-Apr-2010 17:43:28] <felbane> Do you still have the code I posted earlier? I wanted to see if you could look at something

[19-Apr-2010 17:43:53] <bigegor> yes, i have it

[19-Apr-2010 17:44:23] <felbane> Okay in the __init__py in the main folder I have the addMessageQueue methods defined

[19-Apr-2010 17:45:12] <felbane> And then slightly below I have the part where it creates the relationship and then assigns the OperatingSystem.addMessageQueue to the function

[19-Apr-2010 17:45:36] <felbane> Is there anything else I have to do beyond that to register the method? Maybe a commit()?

[19-Apr-2010 17:45:51] <felbane> Or am I making the reference wrong?

[19-Apr-2010 17:47:02] <bigegor> wait, let me look on your code.

[19-Apr-2010 17:51:19] <bigegor> felbane: if you want extend class with new method, you can use 'monkeypatch'

[19-Apr-2010 17:52:41] <bigegor> like here: http://zenpacks.zenoss.org/trac-zenpacks/browser/zenpacks/ZenPacks.community.DistributedCollectors/ZenPacks/community/DistributedCollectors/__init__.py

[19-Apr-2010 17:54:38] <felbane> Ok I will try it that way. I was just trying to do it like in RDBMS zenpack and it wasn't working right

[19-Apr-2010 17:54:42] <felbane> Thanks

[19-Apr-2010 18:03:23] <bigegor> felbane: in MessageQueue class i can not find definition of DeviceStatus

[19-Apr-2010 18:04:13] <felbane> Oh sorry that should be HWStatus

[19-Apr-2010 18:04:30] <felbane> From your deviceAdvDetail pack

[19-Apr-2010 18:11:12] <bigegor> felbane: its working now?

[19-Apr-2010 18:12:10] <felbane> Weird. If I go into ZenDMD and do for attr in dir(dmd.Devices.Server.Windows.devices.myserver.os): print getattr(

[19-Apr-2010 18:12:15] <felbane> Etc etc

[19-Apr-2010 18:12:22] <felbane> It shows up

[19-Apr-2010 18:13:12] <felbane> addMessageQueue <bound method OperatingSystem.addMessageQueue>

[19-Apr-2010 18:13:46] <felbane> But when I click "add queue" in the menu and enter the info in the dialog, it gives an error after I click 'OK'

[19-Apr-2010 18:14:18] <felbane> "Resource not found" on myserver/os/addMessageQueue

[19-Apr-2010 18:17:33] <bigegor> you have no dtml folder and addMessageQueue.dtml file in your ZenPack

[19-Apr-2010 18:19:01] <felbane> Is that what is needed? I will add it. I thought that was just for making changes from the Zope interface

[19-Apr-2010 18:19:58] <bigegor> yes, you need it.

[19-Apr-2010 18:25:47] <felbane> Same error after adding it

[19-Apr-2010 18:26:36] <bigegor> restart zenoss

[19-Apr-2010 18:26:46] <felbane> I did

[19-Apr-2010 18:27:01] <bigegor> hmm

[19-Apr-2010 18:34:47] <bigegor> try in zendmd: dmd.Devices.Server.Windows.devices.myserver.os.addMessageQueue('something', True)

[19-Apr-2010 18:36:05] <bigegor> or: find('myserver').os.addMessageQueue('something', True)

[19-Apr-2010 18:36:55] <felbane> AttributeError: addMessageQueue

[19-Apr-2010 18:39:25] <felbane> If I do this though

[19-Apr-2010 18:39:52] <felbane> From ZenPacks.community.MonitorMsmq import addMessageQueue

[19-Apr-2010 18:40:08] <felbane> os = find('myserver').os

[19-Apr-2010 18:40:28] <felbane> addMessageQueue(os, 'somwthing', True)

[19-Apr-2010 18:40:32] <felbane> It works

[19-Apr-2010 18:40:48] <felbane> Well

[19-Apr-2010 18:40:54] <felbane> It doesn't give an error

[19-Apr-2010 18:41:21] <bigegor> have you try it with monkeypatch also?

[19-Apr-2010 18:41:48] <felbane> I'm still adding that in on another branch

[19-Apr-2010 18:52:17] <felbane> Still nothing. I think I'm going to take a break from it for a while and look at it again tomorrow

[19-Apr-2010 18:52:37] <felbane> I've been at this one all day

[19-Apr-2010 19:19:25] <felbane> bigegor: if you're interested in playing with the code, I posted the latest copy (with RRDs working) at thread/13323

[19-Apr-2010 19:20:28] <felbane> I'm going home now :-) thanks for all the help today

[19-Apr-2010 20:40:06] mrayzenoss is now known as mattray

[20-Apr-2010 00:00:30] [disconnected at Tue Apr 20 00:00:30 2010]

[20-Apr-2010 00:00:31] [connected at Tue Apr 20 00:00:31 2010]

[20-Apr-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[20-Apr-2010 04:39:32] <tehhobbit> how can I debug the output from a certain collector plugin (cant understand why it doesnt collect info clearly available and just one device that doesnt)

[20-Apr-2010 04:40:18] fOrsberg is now known as forsberg

[20-Apr-2010 04:43:20] <carl86> hi guys

[20-Apr-2010 04:44:12] <carl86> was just wondering if anybody has had any joy with getting zenoss to monitor jmx values specifically in cassandra?

[20-Apr-2010 05:28:07] forsberg is now known as fOrsberg

[20-Apr-2010 07:57:43] <mrayzenoss> Morning

[20-Apr-2010 07:59:43] <Simon4> morning

[20-Apr-2010 09:07:11] <mistich> Morning

[20-Apr-2010 09:08:14] <Simon4> heya

[20-Apr-2010 09:08:25] <mistich> have a question about adding a property like availableVersion under dmd. how would I do this in a zenpack tried the following code in the __init__.py

[20-Apr-2010 09:08:30] <mistich> from Products.ZenModel import DataRoot

[20-Apr-2010 09:08:30] <mistich> DataRoot.dcname = "dcname"

[20-Apr-2010 09:08:34] <mistich> but didn't work

[20-Apr-2010 09:08:51] * Simon4 runs screaming from the channel

[20-Apr-2010 09:09:15] <mistich> what :)

[20-Apr-2010 09:09:28] <Simon4> :) sorry

[20-Apr-2010 09:09:43] <Simon4> a more interesting way of saying "no idea"

[20-Apr-2010 09:10:05] <Simon4> interested in the answer though

[20-Apr-2010 09:10:22] <mistich> yeah was hoping one of the zenoss folks would know

[20-Apr-2010 09:12:20] <mistich> think I got it

[20-Apr-2010 09:12:51] * Simon4 has a play with a custom ssh collector

[20-Apr-2010 09:13:25] <mistich> will it use the enable password you set in zproperties?

[20-Apr-2010 09:13:48] <Simon4> the ssh password? yeah, that's the hope

[20-Apr-2010 09:14:00] <Simon4> I think it's actually a "parser" that I'm defining

[20-Apr-2010 09:14:09] <Simon4> copying from the opensolaris zenpack atm

[20-Apr-2010 09:15:24] <mistich> no the option to use enable mode  then use the enable password set in zpropreties

[20-Apr-2010 09:18:01] fOrsberg is now known as forsberg

[20-Apr-2010 09:29:17] <mistich> where is here/templates/macros/page1 live in the following page template ? <tal:block metal:use-macro="here/templates/macros/page1">

[20-Apr-2010 09:29:48] <mistich> nevermind found it

[20-Apr-2010 09:34:31] <carl86> guys have any of you managed to get zenoss monitoring cassandra? im having real trouble with it

[20-Apr-2010 10:02:57] <rmatte> I've never used Cassandra, so no idea

[20-Apr-2010 10:05:11] <mrayzenoss> carl86: haven't tried, what sort of information does Cassandra expose?

[20-Apr-2010 10:06:58] <venturaville> http://blog.justindorfman.com/2010/03/how-to-monitor-cassandra-with-nagios.html

[20-Apr-2010 10:07:02] <venturaville> looks like JMX

[20-Apr-2010 10:07:48] <venturaville> "This tutorial will show you how to start monitoring 'HeapMemoryUsage' in Nagios using a plugin called" check_jmxhttp://exchange.nagios.org/directory/Plugins/Java-Applications-and-Servers/check_jmx/details aka jmxquery. Basically Java Management Extensions (JMX) allow us to monitor the JVMs running on a Cassandra cluster via port 8080 (in this case)."

[20-Apr-2010 10:08:13] <venturaville> http://www.mahalo.com/how-to-monitor-cassandra-with-nagios

[20-Apr-2010 10:08:16] <venturaville> more readable lin

[20-Apr-2010 10:08:17] <venturaville> link

[20-Apr-2010 10:11:58] <mrayzenoss> so you could either build and use the check_jmx command as a Command data source

[20-Apr-2010 10:12:21] <mrayzenoss> or you could install ZenJMX and use that data source

[20-Apr-2010 10:12:30] <mrayzenoss> docs/DOC-3396

[20-Apr-2010 10:13:25] <mrayzenoss> Chapter 10 of the Extended Monitoring Guide covers it

[20-Apr-2010 10:14:23] <Simon4> mrayzenoss: know of any doco that shows the link between commandParsers and collector plugins/modelling plugins? I understand what the parser is doing, but not how to define the command whose output the parser parses.

[20-Apr-2010 10:15:00] <mrayzenoss> just the stuff that's here: community/developers/zenpack_development

[20-Apr-2010 10:15:29] <carl86> yeah i have followed that document, and the placeholder appears for the graph, but no data populates the graph

[20-Apr-2010 10:15:45] <carl86> the zenoss extended monitoring one

[20-Apr-2010 10:15:53] <mrayzenoss> carl86: anything of interest in the zenjmx log?

[20-Apr-2010 10:16:04] <mrayzenoss> are you using the Sun 1.6 JDK?

[20-Apr-2010 10:16:14] <carl86> yeah i am using sun 1.6

[20-Apr-2010 10:16:39] <carl86> im just checking the zenjmx log now

[20-Apr-2010 10:17:31] <carl86> where is the zenjmx log

[20-Apr-2010 10:17:45] <mrayzenoss> I assume in $ZENHOME/log?

[20-Apr-2010 10:17:55] * mrayzenoss needs to play with ZenJMX more obviously

[20-Apr-2010 10:18:18] <carl86> yeah thats right lol i was thinking in the gui for some reason

[20-Apr-2010 10:18:35] <mrayzenoss> in the UI it's Settings->Daemons

[20-Apr-2010 10:20:00] <carl86> cheers

[20-Apr-2010 10:20:46] <carl86> it says it was shut down

[20-Apr-2010 10:21:04] <carl86> which could of  course be the problem

[20-Apr-2010 10:21:15] <mrayzenoss> if you do a 'zenjmx status' what do you get?

[20-Apr-2010 10:22:55] <carl86> it has definately started now

[20-Apr-2010 10:23:51] <carl86> but it is not referring to unreachable nodes in the logs

[20-Apr-2010 10:24:58] <carl86> if i do a zenjmx status it says it cannot find the comman

[20-Apr-2010 10:25:06] <carl86> even though im in the zenoss directory

[20-Apr-2010 10:25:21] <mrayzenoss> are you logged in as 'zenoss'

[20-Apr-2010 10:26:15] <carl86> i am now

[20-Apr-2010 10:26:26] <carl86> it says program running and pid

[20-Apr-2010 10:26:41] <mrayzenoss> ok...

[20-Apr-2010 10:27:54] <mrayzenoss> so try 'zenjmx run -v10 -d cassandraserver' and see what the output looks like

[20-Apr-2010 10:29:33] <Simon4> mrayzenoss: having read stuff under that link... is it a safe assumption that the command used for modelling ("command = /bin/do something --here") is then the command used for collection also for a parser with the same name?

[20-Apr-2010 10:29:47] <Simon4> Step 4 in docs/DOC-2381 is fairly brief :)

[20-Apr-2010 10:30:06] * Simon4 is happy to add to that once he works out what he's doing

[20-Apr-2010 10:30:30] <mistich> have another question in 2.3.2 you could edit templates.pt but in 2.5.2 zenoss seems to ignore the custom templates.pt anyone else

[20-Apr-2010 10:31:54] <carl86> the output is vast and says at the end it is removing service EventService and zenpack.zenoss.zenjm.services.zenjmxconfigservice

[20-Apr-2010 10:33:14] <mrayzenoss> carl86: try going back to the UI and checking your graphs, wasn't zenjmx stopped before?

[20-Apr-2010 10:34:28] <carl86> yeah i just checked it the graph still does not display

[20-Apr-2010 10:34:45] <carl86> just a place holder for the graph

[20-Apr-2010 10:36:43] <mrayzenoss> if you do 'zenjmx debug' it will put the daemon in debug mode, then 'tail -f zenjmx.log' and you can watch it as it attempts to run against your device

[20-Apr-2010 10:39:37] <carl86> there is no output after setting log level to debug

[20-Apr-2010 10:39:56] <carl86> o some inpute just came in

[20-Apr-2010 10:40:03] <mrayzenoss> pastebin it

[20-Apr-2010 10:40:16] <carl86> 2010-04-20 16:39:39,261 DEBUG zen.zenjmx: handleFinish(): success [{'component': 'JMX', 'summary': 'DataSource unreachable nodes; Error calling mbean org.apache.cassandra.service.StorageService.Attributes.UnreachableNodes: Exception: object name is malformed: org.apache.cassandra.service.StorageService.Attributes.UnreachableNodes', 'datasourceId': 'unreachable nodes', 'eventKey': '', 'device': 'nimbus.internal', 'eventClass': '/Status/JMX/Connection'}]

[20-Apr-2010 10:40:54] <Simon4> mrayzenoss: sorted it, nm :)

[20-Apr-2010 10:41:18] <mrayzenoss> Simon4: yay

[20-Apr-2010 10:41:32] <mrayzenoss> carl86: is Remote JMX Access enabled?

[20-Apr-2010 10:41:35] <Simon4> much using of source

[20-Apr-2010 10:41:48] <mrayzenoss> Simon4: feel free to add more docs

[20-Apr-2010 10:41:52] <mrayzenoss> or edit those there

[20-Apr-2010 10:41:58] <mrayzenoss> hmm… I should move those to the wiki

[20-Apr-2010 10:41:59] <Simon4> will do once i get something working

[20-Apr-2010 10:42:40] <carl86> on cassandra yes it is

[20-Apr-2010 10:45:34] <mrayzenoss> carl86: can you reach the cassandra server with JConsole from the Zenoss server?

[20-Apr-2010 10:45:35] <carl86> it is configured to port 8080

[20-Apr-2010 10:45:56] <carl86> i dont have x running on the zenoss server

[20-Apr-2010 10:46:50] <mrayzenoss> just export the display back to a box that does

[20-Apr-2010 10:47:32] <mrayzenoss> I went and asked the developer who wrote it and he said the way to debug is to verify JConsole works from the Zenoss server, since ZenJMX is a wrapper for it

[20-Apr-2010 10:48:15] <carl86> o ok i will have to give that a go then, i could remember how to do it so thought that i would just try it with zenoss

[20-Apr-2010 10:48:44] <mrayzenoss> yeah, unfortunately JMX data sources don't have a 'Test' button like some of the others

[20-Apr-2010 10:48:55] <mrayzenoss> so the debug tip I got is for JConsole

[20-Apr-2010 10:49:16] <mrayzenoss> once you get that up, whatever works there should work in the JMX data source

[20-Apr-2010 10:49:26] <carl86> yeah thats fair enough

[20-Apr-2010 10:49:35] <mrayzenoss> if that doesn't work, post to the forums and I'll get the dev to look at it

[20-Apr-2010 10:49:51] <carl86> the zenoss forums?

[20-Apr-2010 10:49:56] <mrayzenoss> yes please

[20-Apr-2010 10:50:03] <mrayzenoss> perhaps zenpacks

[20-Apr-2010 10:50:06] <mrayzenoss> I'll see it

[20-Apr-2010 10:50:11] <carl86> ok thats great thanks

[20-Apr-2010 10:50:32] <carl86> iv gotta go and do a deply now but thank you for you help

[20-Apr-2010 10:50:39] <mrayzenoss> good luck

[20-Apr-2010 10:51:28] <carl86> cheers

[20-Apr-2010 11:06:45] <venturaville> mrayzenoss: what was it you were having trouble with with the netapp zenpack in subversion?

[20-Apr-2010 11:08:12] <mrayzenoss> when I built it and installed it, there wasn't anything there

[20-Apr-2010 11:08:19] <mrayzenoss> I think it put some MIBs in

[20-Apr-2010 11:08:26] <mrayzenoss> but no device class or templates

[20-Apr-2010 11:08:39] <mrayzenoss> using the 2.5.2 appliance

[20-Apr-2010 11:10:27] <venturaville> I used the 2.5.2 debian stack installer

[20-Apr-2010 11:10:33] <venturaville> it is working fine from a vanilla install

[20-Apr-2010 11:10:50] <venturaville> the only thing I did after installing the zenpack was restart zenoss

[20-Apr-2010 11:11:09] <mrayzenoss> fudge

[20-Apr-2010 11:11:16] <venturaville> zenoss@treebeard:~/zenoss-svn/zenpacks$ zenpack --install ZenPacks.community.NetApp-1.19-py2.4.egg

[20-Apr-2010 11:11:16] <venturaville> 2010-04-20 12:00:29,618 INFO zen.ZPLoader: Loading /usr/local/zenoss/zenoss/ZenPacks/ZenPacks.community.NetApp-1.19-py2.4.egg/ZenPacks/community/NetApp/objects/objects.xml

[20-Apr-2010 11:11:16] <venturaville> 2010-04-20 12:01:07,371 INFO zen.AddToPack: End loading objects

[20-Apr-2010 11:11:16] <venturaville> 2010-04-20 12:01:07,372 INFO zen.AddToPack: Processing links

[20-Apr-2010 11:11:16] <venturaville> 2010-04-20 12:01:11,124 INFO zen.AddToPack: Loaded 2345 objects into the ZODB database

[20-Apr-2010 11:11:16] <venturaville> 2010-04-20 12:01:11,309 INFO zen.HookReportLoader: loading reports from:/usr/local/zenoss/zenoss/ZenPacks/ZenPacks.community.NetApp-1.19-py2.4.egg/ZenPacks/community/NetApp/reports

[20-Apr-2010 11:12:23] <mrayzenoss> firing up my VM and resetting it to a fresh install

[20-Apr-2010 11:14:50] <mrayzenoss> Output: zenpack --install /tmp/ZenPacks.community.NetApp-1.19-py2.4.egg Installed /home/zenoss/ZenPacks/ZenPacks.community.NetApp-1.19-py2.4.egg 2010-04-20 16:14:29,847 INFO zen.HookReportLoader: loading reports from:/home/zenoss/ZenPacks/ZenPacks.community.NetApp-1.19-py2.4.egg/ZenPacks/community/NetApp/reports Done installing ZenPack.

[20-Apr-2010 11:15:00] <mrayzenoss> different output

[20-Apr-2010 11:15:05] <venturaville> makes no sense

[20-Apr-2010 11:15:16] <mrayzenoss> any path assumptions being made?

[20-Apr-2010 11:15:29] <venturaville> not that I am aware of

[20-Apr-2010 11:15:49] <venturaville> check_* doesn't even come into play on an install

[20-Apr-2010 11:16:14] <venturaville> is there a difference between the VM and the stack installer?

[20-Apr-2010 11:16:25] <mrayzenoss> location of zenhome

[20-Apr-2010 11:16:45] <mrayzenoss> I'm digging up the diff

[20-Apr-2010 11:17:39] <mrayzenoss> http://zenpacks.zenoss.org/trac-zenpacks/changeset/594

[20-Apr-2010 11:17:50] <mrayzenoss> I'm going to see what version 1.13 did

[20-Apr-2010 11:34:15] <mrayzenoss> And VMware crashed everything

[20-Apr-2010 11:35:45] <venturaville> eh?

[20-Apr-2010 11:44:07] <rmatte> He opened vmware and it killed his computer lol

[20-Apr-2010 11:44:11] <rmatte> from the sounds of it

[20-Apr-2010 11:44:27] <rmatte> virtualbox > vmware

[20-Apr-2010 11:47:19] <chudler> kvm > *

[20-Apr-2010 12:09:30] <rmatte> hehe

[20-Apr-2010 12:10:29] <venturaville> I like KVM

[20-Apr-2010 12:10:46] <venturaville> it was what motivated me to get a libvirt zenpack going

[20-Apr-2010 12:12:13] <rmatte> VMWare need to smarten up about ESX monitoring, they've made it ridiculously difficult to do

[20-Apr-2010 12:12:37] <rmatte> If they hadn't purposely trashed the SNMP monitoring which was already developed then it'd be fine

[20-Apr-2010 12:13:40] <mrayzenoss> my current desktop is OSX, but I'm constantly debating going back to Ubuntu

[20-Apr-2010 12:13:56] <rmatte> I use Ubuntu for my desktop at home and at work

[20-Apr-2010 12:14:03] <mrayzenoss> how're the desktop virtualization technologies?

[20-Apr-2010 12:14:13] <rmatte> VirtualBox works great

[20-Apr-2010 12:14:21] <rmatte> better than VMWare

[20-Apr-2010 12:14:38] <rmatte> I use it at home and at work, no problems with it

[20-Apr-2010 12:15:27] <aclark> I've got a Mac Book Pro w/Snow Leopard and Parallels running Win 7, Ubuntu, and two Centos', amazing.

[20-Apr-2010 12:15:45] <mrayzenoss> aclark: sounds like you've got more RAM than me

[20-Apr-2010 12:15:47] <rmatte> aclark: that is pretty neat

[20-Apr-2010 12:15:50] * Simon4 uses vmware fusion, it works and I can mess with other virtual machines in it

[20-Apr-2010 12:16:04] <Simon4> which is great if you have esx kicking about at work

[20-Apr-2010 12:16:17] <aclark> mrayzenoss: 2.66 / 4GB

[20-Apr-2010 12:16:27] <aclark> i only usually run one or two VMs at once

[20-Apr-2010 12:16:32] <rmatte> ah

[20-Apr-2010 12:16:32] <aclark> but having them all is key

[20-Apr-2010 12:17:00] <mrayzenoss> aclark: that's my setup, but ever since I went to the new Fusion things have gotten much worse

[20-Apr-2010 12:17:12] <Simon4> mrayzenoss: I had awful perf issues iwth fusion 3

[20-Apr-2010 12:17:18] <aclark> mrayzenoss: too bad…

[20-Apr-2010 12:17:21] <Simon4> then I switched to 64 bit kernel on OS X and they went away

[20-Apr-2010 12:17:40] <Simon4> workmate had same issue

[20-Apr-2010 12:18:17] <mrayzenoss> here we go again, firing up my 2.5.2 clean VM again

[20-Apr-2010 12:19:35] <mrayzenoss> we should have the first alpha out later today (but no VM yet)

[20-Apr-2010 12:19:57] <Simon4> cool

[20-Apr-2010 12:31:39] <rmatte> alpha of stone crab?

[20-Apr-2010 12:32:34] <rmatte> the interface is looking good

[20-Apr-2010 12:32:46] <mrayzenoss> yup

[20-Apr-2010 12:33:18] <rmatte> the one thing that looks absolutely horrid is the device overview page

[20-Apr-2010 12:33:25] <rmatte> I hope they plan to spruce that up a bit lol

[20-Apr-2010 12:34:34] <Simon4> would be good to get hands on an alpha for our zenpack devs to play with.. i managed to terrify them showing them the preview :D

[20-Apr-2010 12:35:22] <rmatte> lol

[20-Apr-2010 12:35:40] <Simon4> they've written a bunch of UI stuff hehe

[20-Apr-2010 12:35:41] <mrayzenoss> Simon4: there's definitely going to be an effort to ensure ZenPacks keep working

[20-Apr-2010 12:37:20] <Simon4> mrayzenoss: yeah, they're worried about their code (custom dashboard portlets/template changes/etc)

[20-Apr-2010 12:37:28] <Simon4> so faster we can get em a copy to play with the better

[20-Apr-2010 12:37:45] <Simon4> rmatte: it was fairly amusing however when I showed them

[20-Apr-2010 12:38:03] <rmatte> hehe

[20-Apr-2010 12:40:43] <rmatte> It's going to be fun figuring out where everything is once it's released

[20-Apr-2010 12:41:00] <rmatte> since I subconciously know what to click on to get where in the current UI

[20-Apr-2010 12:41:09] <Simon4> yeah, agreed

[20-Apr-2010 12:41:18] <Simon4> it's a bit of a fundamental rethink

[20-Apr-2010 12:41:23] <rmatte> yup

[20-Apr-2010 12:41:44] <rmatte> I just can't wait until they can get back to enhancements related to actual monitoring and reporting

[20-Apr-2010 12:42:07] <venturaville> +1

[20-Apr-2010 12:42:16] * Simon4 is still waiting for someone to solve the "service monitoring" problem, across any nms

[20-Apr-2010 12:42:35] <venturaville> which service monitoring problem are you referring to?

[20-Apr-2010 12:42:40] <rmatte> mrayzenoss: no news on that nmap patch?

[20-Apr-2010 12:42:56] <mrayzenoss> rmatte: ticket #?

[20-Apr-2010 12:43:04] <Simon4> venturaville: bring able to drill down from an application level, i.e. how a service management department would want to

[20-Apr-2010 12:43:11] <rmatte> mrayzenoss: http://dev.zenoss.org/trac/ticket/5238#comment:12

[20-Apr-2010 12:43:19] <Simon4> i.e. "so appxxx is down, what services/applications are affected by that?"

[20-Apr-2010 12:43:42] <rmatte> the patch needs to be applied, the zNmapPortscanOptions needs to be updated, and the nmap binary needs to be packaged with Zenoss in $ZENHOME/libexec

[20-Apr-2010 12:43:47] <venturaville> Simon4: yeah we had some contracts with the zenoss folk for zenoss additions for us along that lines

[20-Apr-2010 12:44:01] <Simon4> you can approximate it in zenoss with groups say.. but it's totally manual so that falls very quickly

[20-Apr-2010 12:44:44] <venturaville> the aggregate device additions we got from them were quite useful

[20-Apr-2010 12:44:45] <rmatte> Simon4: yeh, another thing I'd like to see is object tagging... so that you can create tags and then apply them to objects

[20-Apr-2010 12:45:01] <rmatte> for instance, I may want to label certain disks as "backup" or certain interfaces as "wan"

[20-Apr-2010 12:45:05] <Simon4> rmatte: yeah, that would help also

[20-Apr-2010 12:45:10] <rmatte> and then generate reports while using those as filters

[20-Apr-2010 12:45:20] <rmatte> it would make it much more powerful and it's a relatively simple addition

[20-Apr-2010 12:45:45] <Simon4> you could almost monkeypatch that in as a zenpack?

[20-Apr-2010 12:45:50] <venturaville> rmatte: +1 to that as well.  I have my group resources report that I would very much like to use to break out things like filesystems on netapps by client.....

[20-Apr-2010 12:46:03] <rmatte> you could, but I'd like to see it as an actual addition

[20-Apr-2010 12:46:09] <Simon4> for sure

[20-Apr-2010 12:46:25] <mrayzenoss> rmatte: the 2-1 defects are first in line to get fixed in the next hardening iteration

[20-Apr-2010 12:46:35] <rmatte> venturaville: yeh, we're using a php script that my manager wrote that calls on python scripts which I wrote that dump report info in csv format

[20-Apr-2010 12:46:49] <rmatte> then it parses the list and determines what interfaces are WAN interfaces based on a statically defined list

[20-Apr-2010 12:46:57] <rmatte> then generates the final report

[20-Apr-2010 12:47:04] <rmatte> mrayzenoss: k

[20-Apr-2010 12:47:15] <mrayzenoss> rmatte: the P1s have to get fixed first

[20-Apr-2010 12:47:25] <rmatte> mrayzenoss: understandable

[20-Apr-2010 12:47:31] <venturaville> yeah the group resources report was my first attempt to do something like that... keeping it generic without the tag support is difficult though

[20-Apr-2010 12:47:40] <rmatte> yeh

[20-Apr-2010 12:48:05] <rmatte> I wrote a paper on how the reporting system could be improved... I think I'm going to add some more info to it and re-submit it

[20-Apr-2010 13:20:31] <mrayzenoss> venturaville: so I finally got the 1.13 version installed on the VM, same results...

[20-Apr-2010 13:20:43] <venturaville> interesting

[20-Apr-2010 13:20:50] <venturaville> it works just fine in the stack installer

[20-Apr-2010 13:20:51] <mrayzenoss> I'll see if I can get more debugging information

[20-Apr-2010 13:21:03] <venturaville> and the only thing that references a path is the check_* script in the objects.xml

[20-Apr-2010 13:21:19] <venturaville> perhaps it doesn't like the install/upgrade code in __init__.py????

[20-Apr-2010 13:21:56] <mrayzenoss> that's my guess, but it seems fine

[20-Apr-2010 13:33:20] <mrayzenoss> venturaville: I'm going to see if this is a 2.5.2 VMware regresssion by testing against 2.5.1 and 2.4.5

[20-Apr-2010 13:33:44] <mrayzenoss> I remember this working, and I test 90% of the time with the VMware image

[20-Apr-2010 13:35:22] <venturaville> we are using 1.13 on our production one

[20-Apr-2010 13:35:27] <venturaville> it is working there

[20-Apr-2010 13:35:31] <venturaville> *shrug*

[20-Apr-2010 13:40:02] <mrayzenoss> grrr, doesn't work on 2.4.5 either.  Something strange going on

[20-Apr-2010 13:41:21] <venturaville> I'll say

[20-Apr-2010 13:41:33] <venturaville> can someone else with a vanilla install (not VM) try this zenpack out

[20-Apr-2010 13:41:33] <venturaville> ?

[20-Apr-2010 13:42:58] <mrayzenoss> or conversely, if someone has a nice stripped-down CentOS or Ubuntu I could use as a base image, I'll start testing with that as well

[20-Apr-2010 13:43:58] <venturaville> none that aren't customized out the wazoo.... :-P

[20-Apr-2010 13:45:48] <mrayzenoss> hmm.. turns out I have ISO installers for Ubuntu I could use

[20-Apr-2010 13:49:20] <davetoo> umm...

[20-Apr-2010 13:49:35] <davetoo> when did this change to ZenUser management happen:

[20-Apr-2010 13:49:44] <davetoo> I'm both a ZenManager and a Zope Manager,

[20-Apr-2010 13:49:56] <davetoo> but I can't make any changes to a user I just added

[20-Apr-2010 13:50:48] <davetoo> I make special local users for alerting rules/filtering

[20-Apr-2010 13:51:21] <davetoo> I upgraded to 2.5.2 last week and just now found this behavior.

[20-Apr-2010 13:56:26] <davetoo> oh what the heck

[20-Apr-2010 13:56:39] <davetoo> this is quite sub-optimal

[20-Apr-2010 13:57:05] <davetoo> now it's just sent out the new local password for the "alerting" account to the mailing list where the alerts are sent

[20-Apr-2010 13:57:08] <davetoo> blowing that password :0

[20-Apr-2010 13:57:10] <davetoo> :)

[20-Apr-2010 14:00:50] <rmatte> haha

[20-Apr-2010 14:01:12] <rmatte> weird that it sent that out

[20-Apr-2010 14:01:21] <davetoo> well, it's the email for the "user"

[20-Apr-2010 14:01:33] <rmatte> I see

[20-Apr-2010 14:01:38] <davetoo> and when I hit "reset password" (which I didn't read closely)

[20-Apr-2010 14:01:42] <rmatte> ah

[20-Apr-2010 14:01:42] <davetoo> it did what I asked it do

[20-Apr-2010 14:01:43] <davetoo> to

[20-Apr-2010 14:01:46] <rmatte> lol

[20-Apr-2010 14:02:08] <Simon4> that "reset password" button is a huge UX fail

[20-Apr-2010 14:02:09] <rmatte> that problem sounds familliar

[20-Apr-2010 14:02:10] <rmatte> hmmm

[20-Apr-2010 14:02:27] <rmatte> (the not being able to edit users as administrator issue)

[20-Apr-2010 14:02:31] <davetoo> but.. the real problem is this new "confirmation password" thing at the bottom

[20-Apr-2010 14:02:36] <Simon4> so many of our users type in new passwords, then hit that button, and wonder why they can't log in again

[20-Apr-2010 14:03:04] <rmatte> the confirmation password is supposed to be the password of the account that you're using I believe

[20-Apr-2010 14:03:07] <rmatte> such as the admin account

[20-Apr-2010 14:03:22] <rmatte> so that's probably why you're not able to edit them as the admin account, because you're not filling that in with the admin password

[20-Apr-2010 14:03:30] <rmatte> they need to add an explanation of that to that page

[20-Apr-2010 14:04:07] <Simon4> rmatte: yeah, it is

[20-Apr-2010 14:04:15] <davetoo> um

[20-Apr-2010 14:04:15] <Simon4> and so totally not obvious

[20-Apr-2010 14:04:23] <rmatte> yup

[20-Apr-2010 14:04:33] <davetoo> how is the Admin going to know the users' password?

[20-Apr-2010 14:04:38] <Simon4> nahh

[20-Apr-2010 14:04:58] <davetoo> I'm logged in as myself,

[20-Apr-2010 14:05:01] <davetoo> with LDAP auth,

[20-Apr-2010 14:05:03] <Simon4> you edit the users details, then put _your_ password in the bottom password field, then click update

[20-Apr-2010 14:05:06] <rmatte> davetoo: you're not able to just use the admin password for the confirmation?

[20-Apr-2010 14:05:12] <rmatte> yeh, YOUR password

[20-Apr-2010 14:05:15] <rmatte> not the user's password

[20-Apr-2010 14:05:21] <rmatte> it's very unobvious though

[20-Apr-2010 14:05:22] <davetoo> and I'm a ZenManager

[20-Apr-2010 14:05:24] <davetoo> and a Manager

[20-Apr-2010 14:05:29] <davetoo> and I do put in *my* password

[20-Apr-2010 14:05:31] <davetoo> and it does'nt work

[20-Apr-2010 14:05:37] <rmatte> You only need to be a Manager, no point in being a ZenManager too

[20-Apr-2010 14:05:43] <davetoo> erm,

[20-Apr-2010 14:05:46] <Simon4> we hacked this to remove the requirement, it was so horribly nonintuitive

[20-Apr-2010 14:05:47] <davetoo> I disagree

[20-Apr-2010 14:06:03] <davetoo> I want people who are ZenManagers but not Zope managers

[20-Apr-2010 14:06:08] <davetoo> oh,

[20-Apr-2010 14:06:19] <rmatte> I'm saying theys hould be one or the other

[20-Apr-2010 14:06:20] <rmatte> not both

[20-Apr-2010 14:06:21] <davetoo> well, perhaps Manager is a superset, but not the other way 'round

[20-Apr-2010 14:06:24] <davetoo> 'k

[20-Apr-2010 14:06:26] <rmatte> correct

[20-Apr-2010 14:06:44] <davetoo> anyway, I have an immediate need to fix these accounts

[20-Apr-2010 14:06:55] <rmatte> If the confirmation isn't working with your password then I have no idea what the issue is

[20-Apr-2010 14:07:28] <davetoo> Maybe what I need to change is a simple property that the ZMI can safely change

[20-Apr-2010 14:08:02] <davetoo> yay, it didn't explode

[20-Apr-2010 14:08:09] <rmatte> lol

[20-Apr-2010 14:08:37] <davetoo> I can't always predict when using the ZMI to change things will backfire

[20-Apr-2010 14:08:50] <davetoo> in this case I just needed to change the user's Email alert

[20-Apr-2010 14:10:06] <davetoo> This stuff sure can be confusing.. Zope plus file-based products plus ...

[20-Apr-2010 14:10:29] bzed_ is now known as bzed

[20-Apr-2010 14:20:10] <rmatte> yup

[20-Apr-2010 14:48:21] <gwb235> we've recently been unable to "Add IpService".  The "OK" button remains gray when trying to enter things like "smtp" into the Ip Service Class.  Any ideas?

[20-Apr-2010 14:50:40] <rmatte> gwb235: click cancel and try it again, the UI can be finicky

[20-Apr-2010 14:50:55] <rmatte> are you talking about from the IP Services page?

[20-Apr-2010 14:50:58] <rmatte> or from the device page?

[20-Apr-2010 14:51:12] <gwb235> from the device page, OS tab, IP Services section

[20-Apr-2010 14:51:38] <rmatte> right, that's what I figured...

[20-Apr-2010 14:51:50] <rmatte> you need to type it in and wait to see a dropdown menu to select it from

[20-Apr-2010 14:52:02] <rmatte> if you don't see the menu within a few seconds, hit cancel, then do it again

[20-Apr-2010 14:52:15] <rmatte> once you see the menu, select what you want then click ok

[20-Apr-2010 14:52:29] <gwb235> I've seen it work (auto-matching/completing your entry with the dropdown) but not recently (no dropdown).  I'll do the fill-in, cancel routine a few more times.

[20-Apr-2010 14:52:33] <rmatte> the dropdown menu will drop down from the input field

[20-Apr-2010 14:52:43] <rmatte> yeh, it never works for me on the first try

[20-Apr-2010 14:52:49] <rmatte> I always have to click cancel then do it again

[20-Apr-2010 14:52:54] <rmatte> it works on the second or third try

[20-Apr-2010 14:53:43] <rmatte> afk for a second

[20-Apr-2010 14:53:58] <gwb235> no such love for me :(  I'll try again later, thanks

[20-Apr-2010 14:57:22] <rmatte> try clearing your browser cache

[20-Apr-2010 14:57:26] <rmatte> and restarting your browser

[20-Apr-2010 14:57:44] <rmatte> also make sure you're using a supported browser

[20-Apr-2010 14:59:00] <gwb235> IE and firefox show me the same behavior

[20-Apr-2010 14:59:28] <rmatte> yeh, that part of the UI sucks

[20-Apr-2010 14:59:33] <rmatte> hopefully it'll be fixed in the new UI

[20-Apr-2010 15:01:49] <venturaville> bigegor: howdy

[20-Apr-2010 15:02:06] <bigegor> hello

[20-Apr-2010 15:02:36] <venturaville> I have pretty much given up on trying to do the IBM/LSI arrays with CIM... too unreliable

[20-Apr-2010 15:02:59] <venturaville> I'm going to be looking at the IBM SVCs which talk the same way though

[20-Apr-2010 15:03:44] <venturaville> hate having wasted all that time on trying to get it working...just too frustrating :-/

[20-Apr-2010 15:05:13] * rmatte chants "We want ESX!" while banging a stick against a tree

[20-Apr-2010 15:05:16] <rmatte> :P

[20-Apr-2010 15:05:21] <bigegor> i'm also trying to get working CIM Hardware Monitoring of IBM, Dell and HP servers. HORROR :(

[20-Apr-2010 15:05:42] <venturaville> at least the HP ones work reliably :-)

[20-Apr-2010 15:06:20] <Simon4> rmatte: general rule, swapping "stick against tree" with "credit card against reader" makes more effect :)

[20-Apr-2010 15:06:21] <venturaville> rmatte: I hear you.  have to work it on my own free time for ESX as we already have monitoring for those through enterprise.... (can't justify the work time for it)

[20-Apr-2010 15:06:32] <bigegor> but Dell and IBM not :(

[20-Apr-2010 15:06:53] <rmatte> HP tend to have the best monitoring solutions

[20-Apr-2010 15:06:54] <venturaville> I only get about 10% of my time to do monitoring...

[20-Apr-2010 15:07:17] <rmatte> venturaville: yeh, understandable

[20-Apr-2010 15:07:34] <venturaville> bigegor ... the SOAP API for the chassis works pretty well .. have you looked around for that for the servers through iLO or some such?

[20-Apr-2010 15:07:48] <bigegor> IBM Deirector reported AMD Opteron processor as k6

[20-Apr-2010 15:07:56] <venturaville> ouch

[20-Apr-2010 15:08:09] <rmatte> haha, K6

[20-Apr-2010 15:08:11] <rmatte> that's awesome

[20-Apr-2010 15:08:27] <bigegor> Processor Family=15

[20-Apr-2010 15:08:38] <rmatte> that's some pretty hefty processing power there in that K6

[20-Apr-2010 15:08:44] <rmatte> and it's not even a K6-2

[20-Apr-2010 15:08:46] <rmatte> :P

[20-Apr-2010 15:08:59] <venturaville> What I wouldn't have given for a K6 with that much oomph back when I actually had one

[20-Apr-2010 15:09:06] <rmatte> same here

[20-Apr-2010 15:09:46] <rmatte> people would have sacrificed children back then for Opteron level computing performance

[20-Apr-2010 15:10:14] <venturaville> *brings back memories of compiling linux kernel on 7MHz processor

[20-Apr-2010 15:10:32] <rmatte> tonight is going to be fun, I have to upgrade 13 Zenoss servers from 2.4.5 to 2.5.2

[20-Apr-2010 15:10:47] <bigegor> Dell used Deprecated Status (string) property instead of OperatingStatus (int list). :(

[20-Apr-2010 15:11:01] <rmatte> The first time I compiled a kernel was the FreeBSD kernel on a 150MHz processor

[20-Apr-2010 15:11:11] <rmatte> but I'm only 24 and I was 14 or 15 at the time

[20-Apr-2010 15:11:55] <rmatte> My Zenoss farm is growing soon, at least 2 new servers on the way

[20-Apr-2010 15:11:56] <venturaville> bigegor: having the same fun trying to unite the HP and IBM blade chassis zenpacks

[20-Apr-2010 15:12:14] <rmatte> did that HP blade chassis pack get officially released?

[20-Apr-2010 15:12:42] <venturaville> there is a version out there, but I think Simon4 was still working on his additions

[20-Apr-2010 15:12:48] <rmatte> ah

[20-Apr-2010 15:12:55] <rmatte> yeh, I'm anxiously waiting his additions

[20-Apr-2010 15:13:03] <venturaville> been swamped for 2 weeks and haven't been able to ask him about it

[20-Apr-2010 15:13:06] <Simon4> rmatte: they're in svn

[20-Apr-2010 15:13:16] <venturaville> there you go ... :-)

[20-Apr-2010 15:13:19] <Apachez> venturaville: you are just shy, admit it - its ok ;)

[20-Apr-2010 15:13:21] * Simon4 has also been swamped and lately stuck trying to find flights

[20-Apr-2010 15:13:22] <venturaville> (haven't tested them myself)

[20-Apr-2010 15:13:22] <rmatte> Simon4: yeh, I'd rather wait for them to be eggified so that I can throw the pack in to production

[20-Apr-2010 15:13:30] <Simon4> rmatte: svn co, then make :)

[20-Apr-2010 15:13:39] <rmatte> meh, I'll wait

[20-Apr-2010 15:13:56] <rmatte> I've done enough ZenPack building/rebuilding while prepping for this 2.5.2 upgrade

[20-Apr-2010 15:13:59] <rmatte> not in a rush to do more

[20-Apr-2010 15:14:04] <venturaville> heheh

[20-Apr-2010 15:14:11] * Simon4 will try and get things solidified in the next couple of days, now I'm stuck in the country

[20-Apr-2010 15:14:15] <venturaville> Simon4: I'll update my tree and take a look

[20-Apr-2010 15:14:23] <Simon4> venturaville: awesome, please do

[20-Apr-2010 15:14:27] <rmatte> stuck in the country?

[20-Apr-2010 15:14:48] <Simon4> rmatte: I'm in the UK, was supposed to be best man at a mate's wedding in New Zealand on Friday

[20-Apr-2010 15:14:54] <rmatte> ahhhh

[20-Apr-2010 15:14:56] <Simon4> volcano says no :(

[20-Apr-2010 15:15:02] <rmatte> yeh, I heard about that

[20-Apr-2010 15:15:06] <rmatte> crazy

[20-Apr-2010 15:15:16] <Simon4> has been a fairly rough week

[20-Apr-2010 15:15:22] <venturaville> my wife was talking to some people adopting in Poland that have been stuck there for quite some time

[20-Apr-2010 15:16:33] <venturaville> no sign of them getting out soon

[20-Apr-2010 15:17:32] <davetoo> Venturaville?

[20-Apr-2010 15:17:38] <davetoo> where's that name from

[20-Apr-2010 15:17:41] <Simon4> yeah, it's a huge mess over here

[20-Apr-2010 15:18:08] <davetoo> well, what I wonder is, where are the statements from the *pilots*

[20-Apr-2010 15:26:50] <venturaville> davetoo: long since lost in obscurity ..had the same net alias for > 15 years

[20-Apr-2010 15:27:04] <rmatte> Simon4: http://scienceblogs.com/startswithabang/upload/2010/04/volcanic_lightning_eyjafjallaj/See%20Explanation.%20%20Clicking%20on%20the%20picture%20will%20download%250A%20the%20highest%20resolution%20version%20available.jpeg

[20-Apr-2010 15:27:08] <davetoo> are you from Ventura CA?

[20-Apr-2010 15:27:27] <rmatte> That's the volcano erupting and causing electrical storms

[20-Apr-2010 15:29:06] <Simon4> rmatte: heh, that's my desktop background atm :)

[20-Apr-2010 15:29:19] <Simon4> there have been some amazing images

[20-Apr-2010 15:29:37] <mrayzenoss> I've been looking for larger versions

[20-Apr-2010 15:29:47] <mrayzenoss> for a desktop image that is

[20-Apr-2010 15:30:01] <Simon4> ah... urm

[20-Apr-2010 15:30:03] <rmatte> yeh, it's pretty neat, I'd love to get a 1440x900 version of it

[20-Apr-2010 15:30:08] * Simon4 tries to find

[20-Apr-2010 15:30:17] <mrayzenoss> Simon4: I've looked a bit :)

[20-Apr-2010 15:30:39] <Simon4> this was well bigger than my 1440x900 before I desktop'ed it

[20-Apr-2010 15:31:09] <venturaville> davetoo: nope

[20-Apr-2010 15:31:13] <Simon4> http://apod.nasa.gov/apod/image/1004/icevolcano_fulle_big.jpg

[20-Apr-2010 15:31:22] <mrayzenoss> sweet

[20-Apr-2010 15:31:27] <mrayzenoss> Simon4: thanks

[20-Apr-2010 15:31:28] <Simon4> not quite full, but damn close

[20-Apr-2010 15:31:39] <mrayzenoss> that'll fit my laptop screen

[20-Apr-2010 15:31:46] <Simon4> :)

[20-Apr-2010 15:32:11] <Simon4> others I would love full size off = http://img72.imageshack.us/img72/8261/ejafjalla16apr2010mfull.jpg and http://img689.imageshack.us/img689/1026/e1300000002.jpg

[20-Apr-2010 15:34:26] <rmatte> yeh, those are pretty crazy

[20-Apr-2010 15:42:24] <ckrough> I want to loop through the zProperties of a set of organizers, like: 'for i in dmd.Devices.Network.Router.Cisco.getSubOrganizers():'

[20-Apr-2010 15:42:43] <ckrough> problem is that what it returns for i is what I expected, but I cant seem to modify it

[20-Apr-2010 15:42:50] <ckrough> ie; commit() doesnt make a change

[20-Apr-2010 15:43:04] <ckrough> Im wondering if getSubOrganizers() is not what I want

[20-Apr-2010 15:43:59] <rmatte> well, how are you trying to set it?

[20-Apr-2010 15:44:25] <ckrough> for i in dmd.Devices.Network.Router.Cisco.getSubOrganizers():

[20-Apr-2010 15:44:31] <ckrough> i.zCollectorPlugins.remove(oldPlugin)

[20-Apr-2010 15:44:52] <ckrough> looking at the resulting i.zCollectorPlugins, oldPlugin is gone, but after a commit() and restart of dmd its back

[20-Apr-2010 15:45:01] <ckrough> no errors in event.log

[20-Apr-2010 15:45:10] <rmatte> so you're trying to actually completely remove a zProperty?

[20-Apr-2010 15:45:19] <ckrough> ys

[20-Apr-2010 15:45:56] <rmatte> the proper way to remove a zProperty is...

[20-Apr-2010 15:46:00] <rmatte> >>> ch = devices

[20-Apr-2010 15:46:00] <rmatte> >>> ch._delProperty('zTest')

[20-Apr-2010 15:46:00] <rmatte> >>> commit()

[20-Apr-2010 15:46:18] <ckrough> how would I do that for acollector plugin?

[20-Apr-2010 15:46:31] <ckrough> they are a list (.zCollectorPlugins)

[20-Apr-2010 15:46:36] <rmatte> oh, hmmm

[20-Apr-2010 15:46:52] <ckrough> I figured I just have to modify the list and commit

[20-Apr-2010 15:47:09] <rmatte> I would assume so...

[20-Apr-2010 15:47:17] <rmatte> but it might be grabbing the list on startup from elsewhere

[20-Apr-2010 15:47:24] <rmatte> so you're just modifying a temporary list

[20-Apr-2010 15:47:47] <ckrough> nah, modifying the .zCollectorPlugins directly

[20-Apr-2010 15:48:24] <ckrough> like so: http://pastebin.com/gMTWqCZA

[20-Apr-2010 15:48:52] <ckrough> when I step through one iteration everything looks right

[20-Apr-2010 15:48:57] <ckrough> but commit() does nothing

[20-Apr-2010 15:49:22] <rmatte> right, but what I'm saying is that that list might be generated from something else, so while you're editing the list it's grabbing it from elsewhere on restart

[20-Apr-2010 15:49:33] <rmatte> commit() can't just magically not be working

[20-Apr-2010 15:49:39] <ckrough> exactly what I was thinking

[20-Apr-2010 15:50:25] <ckrough> so getSubOrganizers() is not the method I want

[20-Apr-2010 15:51:26] <rmatte> Your best bet is to check $ZENHOME/Products/ZenModel/migrate/zCollectorPlugins.py

[20-Apr-2010 15:51:32] <rmatte> to see how it does it

[20-Apr-2010 15:52:31] <ckrough> thanks. very differnt

[20-Apr-2010 15:52:33] <rmatte> there's probably some tiny detail you're missing

[20-Apr-2010 15:52:43] <rmatte> they seem to be using a set function

[20-Apr-2010 15:52:49] <rmatte> rather than doing append

[20-Apr-2010 15:55:01] <ckrough> i.getProperty('zCollectorPlugins')

[20-Apr-2010 15:55:05] <ckrough> looks promising

[20-Apr-2010 15:55:23] <rmatte> yup

[20-Apr-2010 15:56:23] <rmatte> well, I have to get going, I'll be back around noon tomorrow

[20-Apr-2010 15:56:27] <rmatte> cheers folks

[20-Apr-2010 17:29:56] <felbane> Hello all.

[20-Apr-2010 17:31:03] <felbane> I'm about to the point of losing hair by force :-)

[20-Apr-2010 17:31:49] <felbane> I cannot figure out why my custom dialog => add object process is failing.

[20-Apr-2010 17:36:25] <felbane> Http://community.zenoss.org/thread/13323

[20-Apr-2010 17:36:51] <felbane> If anyone is interested.

[20-Apr-2010 17:48:37] <felbane> Updated the thread to be more descriptive of the current problem.

[20-Apr-2010 19:25:17] <voxter> In centos, am i better off grabbing the rpm or the "stack" binary?

[20-Apr-2010 19:26:11] <voxter> maybe a better question, why does zenoss not have a yum repository?

[20-Apr-2010 19:27:22] cgibbons_ is now known as cgibbons

[20-Apr-2010 19:32:31] <voxter> Nevermind, I found a repository. :)

[20-Apr-2010 19:42:58] <rmatte> eugh, 2.5.2 takes forever to load device pages

[20-Apr-2010 19:43:03] <rmatte> it's snappy for everything else

[20-Apr-2010 19:50:04] <rmatte> hmmm, now it's magically working faster

[20-Apr-2010 20:11:42] <straterra> o

[20-Apr-2010 20:44:39] forsberg is now known as fOrsberg

[20-Apr-2010 22:26:32] <xpot-mobile> Question: I have a device using satellite and zenmodeler times out due to the latency... is there a way I can increase the timeout for zenmodeler on this device?

[20-Apr-2010 23:56:13] <rmatte> this is driving me nuts, I tested upgrading a 2.4.5 box to 2.5.2 and it worked fine, LDAP authentication still worked... now I upgrade my prod boxes and LDAP authentication stops working on every one of them

[20-Apr-2010 23:56:24] <rmatte> and there is literally no difference between them and the test box

[20-Apr-2010 23:57:54] <rmatte> hmmm, actually it only appears to be 2 servers that I'm having the issue with

[21-Apr-2010 00:00:30] [disconnected at Wed Apr 21 00:00:30 2010]

[21-Apr-2010 00:00:31] [connected at Wed Apr 21 00:00:31 2010]

[21-Apr-2010 00:00:44] [zenoss-logger (logger bot) has joined #zenoss]

[21-Apr-2010 00:00:44] <gibson.freenode.net> [freenode-info] channel trolls and no channel staff around to help? please check with freenode support: http://freenode.net/faq.shtml#gettinghelp

[21-Apr-2010 00:00:54] <rmatte> all the others were fresh 2.4.1 installs and it's working fine on those

[21-Apr-2010 00:03:44] <rmatte> ah sweet, I figured it out

[21-Apr-2010 00:03:51] <rmatte> have to recompile

[21-Apr-2010 04:31:47] <fus10nx> If I am in the device view - /Devices/Server/Windows/<my server name> for instance and the status says up w/ the green icon. If I unplug the ethernet cable from that server, how long (without me manually refreshing the screen), will it take for it to show up as down ? Or does a refresh need to happen ?

[21-Apr-2010 06:27:45] fOrsberg is now known as forsberg

[21-Apr-2010 07:49:50] <cgibbons> device screens don't auto-refresh in the current shipping versions

[21-Apr-2010 07:50:09] <venturaville> yeah I've noticed that in some situations too

[21-Apr-2010 08:08:46] <fus10nx> what screen do auto-refresh ? and will 3.0 offer that feature?

[21-Apr-2010 08:10:45] <cgibbons> the event consoles refresh

[21-Apr-2010 08:10:54] <fus10nx> ok thank you

[21-Apr-2010 08:11:43] <cgibbons> and 3.0 won't change the device status page refreshing

[21-Apr-2010 08:12:51] <cgibbons> dashboard auto-refreshes too

[21-Apr-2010 08:12:58] <fus10nx> hmmm ok

[21-Apr-2010 08:12:59] <cgibbons> which has a device problems portlet, so that may give you what you are after

[21-Apr-2010 08:13:06] <fus10nx> i figured it was an ajax thing but obv. not

[21-Apr-2010 08:13:14] <fus10nx> so device page requires manual refrehws

[21-Apr-2010 08:14:00] <cgibbons> in general dashboard + event console are generally used for an operations console view, whereas device status is something you bring up when you're troubleshooting

[21-Apr-2010 08:15:16] <fus10nx> so how often will the dash auto update

[21-Apr-2010 08:15:35] <fus10nx> ive just unplugged my network cable - waiting for an error to show in "Device Issues" on the dashboard

[21-Apr-2010 08:16:24] <cgibbons> i'm curious about that myself. i'm gonna guess @ 60 seconds.

[21-Apr-2010 08:16:51] <cgibbons> ah  you can configure each portlet's refresh interval

[21-Apr-2010 08:17:05] <fus10nx> ya thats true

[21-Apr-2010 08:17:10] <fus10nx> its been over 60 seconds, still nothing shows

[21-Apr-2010 08:17:11] <cgibbons> click on the asterisk on the portlet you want to configure the refresh timer for

[21-Apr-2010 08:17:52] <fus10nx> ya its set for 60

[21-Apr-2010 08:18:13] <fus10nx> even manual refresh it isnt showing

[21-Apr-2010 08:18:16] <fus10nx> just on the device list

[21-Apr-2010 08:18:25] <fus10nx> oh wait

[21-Apr-2010 08:18:26] <fus10nx> haha

[21-Apr-2010 08:34:37] <ckrough> anyone have a clue why this only passes through the first iteration? http://pastebin.com/dByZ0AAN

[21-Apr-2010 08:35:52] <fus10nx> Is anyone using the full blown version of SNMP informant ? I'm just wondering, if they provide the MIB files, instead of buying the entire piece of software, can't you just load the MIB files into ZenOSS?

[21-Apr-2010 09:11:38] <cluther> ckrough: No, I don't see why that is.. but try something like this instead. http://gist.github.com/373848

[21-Apr-2010 09:12:30] <ckrough> cluther: thanks!

[21-Apr-2010 09:12:48] <cluther> fus10nx: The MIB doesn't do you any good without the agent.

[21-Apr-2010 09:40:36] <davetoo> cgibbons: (or other devs) how important are the unit tests to you? :)

[21-Apr-2010 09:40:59] <cgibbons> we fix'em when we break 'em

[21-Apr-2010 09:41:05] <davetoo> heh

[21-Apr-2010 09:42:15] <davetoo> so I don't know when I'll get around to making it publishable but I'm essentialy condensing that eclipse stuff into a bash script

[21-Apr-2010 09:43:46] <davetoo> I haven''t done much with a running zenoss yet but I have downloaded and built (successfully) 2.4.2 and 2.5.2.

[21-Apr-2010 09:46:08] <davetoo> Internally, when you (or maybe support or Professional Services) have to load up a batch of devices, are you using zenbatchload or are you using different tools?

[21-Apr-2010 09:46:38] <cluther> davetoo: Our QA group uses zenbatchload a lot.

[21-Apr-2010 09:47:09] <davetoo> the BatchDeviceLoader code makes my brain hurt a bit, but it looks like the split function is fixed

[21-Apr-2010 09:47:27] <davetoo> don't remember if I sent in the (trivial) patch for that

[21-Apr-2010 09:47:32] <davetoo> how's it going cluther, long time

[21-Apr-2010 09:49:10] <cluther> No complaints here.

[21-Apr-2010 09:52:04] <davetoo> I checked out trunk by mistake yesterday; is it worth trying to build from that?

[21-Apr-2010 09:52:59] <cluther> davetoo: Yeah, it's up to Python2.6 and a newer Zope now.

[21-Apr-2010 09:53:12] <cluther> It is definitely in a buildable state.

[21-Apr-2010 09:53:24] <davetoo> great

[21-Apr-2010 09:54:17] <davetoo> I like @property decorators :)

[21-Apr-2010 09:55:24] <davetoo> so is it at all feasible to have multiple versions running, simultaneously, on the same box?  Or is it too hard to track down all the config changes?

[21-Apr-2010 09:55:59] <davetoo> I still have a bunch of 2.4.x boxes and I need to test that what I'm writing still works there in addition to 2.5.2+

[21-Apr-2010 09:56:39] <venturaville> does anyone know where the thresholds for the 95% critical events are generated for /Perf/Filesystem?

[21-Apr-2010 09:57:03] <davetoo> yes, I know exactly where 'cause I disabled mine last night :)

[21-Apr-2010 09:57:30] <davetoo> in the filesystem template

[21-Apr-2010 09:57:35] <venturaville> nope not there

[21-Apr-2010 09:57:38] <venturaville> the default is 90% there

[21-Apr-2010 09:57:42] <davetoo> what device class?

[21-Apr-2010 09:57:51] <venturaville> this is for some NetApp ones

[21-Apr-2010 09:57:56] <davetoo> Enterprise?

[21-Apr-2010 09:57:58] <venturaville> which are also set for 90% on the filesystems

[21-Apr-2010 09:57:59] <venturaville> yep

[21-Apr-2010 09:58:02] <davetoo> Those are what I turned off :)

[21-Apr-2010 09:58:16] <venturaville> strange voodo going on that is marking 95% ones

[21-Apr-2010 09:58:23] <davetoo> well, I saw a 90% and a 98%

[21-Apr-2010 09:58:57] <cluther> venturaville: If it is a snapshot you should checkout the FileSystemSnapshot template.

[21-Apr-2010 09:59:00] <davetoo> ok, 90% and 98% I know about, 95, no

[21-Apr-2010 09:59:28] <davetoo> My version of the snapshot template has 120%, that's it

[21-Apr-2010 09:59:40] <cluther> davetoo: It's too much pain to try to run multiple versions simultaneously due to all of the port conflicts. I prefer to have them all installed and only run one at a time.

[21-Apr-2010 09:59:54] <davetoo> but I think there is stuff in the command templates

[21-Apr-2010 10:00:23] <davetoo> I havent' looked yet, but have you looked in the check_netapp.py code?

[21-Apr-2010 10:00:37] <davetoo> ${here/ZenPackManager/packs/ZenPacks.zenoss.NetAppMonitor/path}/libexec/check_netapp.py

[21-Apr-2010 10:01:01] <davetoo> cluther: 'k, good to know, thanks.

[21-Apr-2010 10:02:19] <venturaville> none are set in command templates

[21-Apr-2010 10:02:33] <venturaville> I wrote the check_netapp.py code.... :-P

[21-Apr-2010 10:02:45] <davetoo> d'oh!

[21-Apr-2010 10:02:50] <davetoo> seriously?

[21-Apr-2010 10:02:55] <venturaville> yeah

[21-Apr-2010 10:03:15] <venturaville> we only saw this show up when we moved the filesystem events to /Perf/Filesystem/NAS

[21-Apr-2010 10:03:21] <davetoo> well, I'll keep quiet then :)

[21-Apr-2010 10:03:33] <venturaville> it doesn't happen when they go to /Storage/Full

[21-Apr-2010 10:03:59] <davetoo> You checked for inherited class transforms?

[21-Apr-2010 10:04:22] <cluther> venturaville: I think there's more than one check_netapp.py out there. I wrote the one in the enterprise NetAppMonitor pack.

[21-Apr-2010 10:04:32] <davetoo> (oh)

[21-Apr-2010 10:04:34] <cluther> But it doesn't have any thresholding stuff in it.

[21-Apr-2010 10:04:35] <venturaville> yeah we are using the one I wrote at the moment

[21-Apr-2010 10:04:43] <cluther> ah

[21-Apr-2010 10:04:55] <venturaville> cluther = enterprise netapp zenpack

[21-Apr-2010 10:05:23] <venturaville> anyways I'm more curious about the /perf/Filesystem and what is so special about it

[21-Apr-2010 10:05:44] <cluther> venturaville: It has an elaborate transform to make the events look nice.

[21-Apr-2010 10:06:13] <davetoo> ..which I've been looking at how to modify without breaking;

[21-Apr-2010 10:06:25] <davetoo> we tend to let our systems run at such high utilization

[21-Apr-2010 10:07:15] <venturaville> cluther: any pointers to where that gets applied?

[21-Apr-2010 10:07:16] <davetoo> that I want to add a set of tags in /Groups/Ignore and have the transforms check those, and drop some or all threshold events if found

[21-Apr-2010 10:07:20] <venturaville> code wise and UI wise

[21-Apr-2010 10:07:54] <cluther> venturaville: All events coming into /Perf/Filesystem and its subclasses get the transform applied to them. You can find it under menu -> more -> transform on the /Perf/Filesystem event class.

[21-Apr-2010 10:08:42] <davetoo> cluther: how/where is that maintained in the source?

[21-Apr-2010 10:08:49] <cluther> Nowhere.

[21-Apr-2010 10:08:55] <cluther> It's in the object database.

[21-Apr-2010 10:09:25] <davetoo> ok, so you just use the transform window as an editor to create them.

[21-Apr-2010 10:09:36] <cluther> I suppose it is also under ZenModel/data/events.xml for initial database creation too.

[21-Apr-2010 10:11:24] <venturaville> I see it now

[21-Apr-2010 10:11:32] <venturaville> when were those transforms added?

[21-Apr-2010 10:11:36] <venturaville> was that started in 2.5.2?

[21-Apr-2010 10:11:50] <cluther> It's been built up since 2.4 I think.

[21-Apr-2010 10:12:40] <venturaville> this was the first I think I have tried to put events in there for non-Linux filesystems

[21-Apr-2010 10:13:05] <lthrasher> where can i find the source for the netapp enterprise zenpack?  it doesn't seem to be in the sourceforge svn repo.

[21-Apr-2010 10:16:19] <venturaville> have to buy enterprise license for that one :-)

[21-Apr-2010 10:19:06] <lthrasher> oh, like the red hat model?

[21-Apr-2010 10:20:47] <venturaville> not quite

[21-Apr-2010 10:20:48] <davetoo> I just upgraded to 2.5.2 last week... I sure do like this event console, especially the easy filtering

[21-Apr-2010 10:20:58] <venturaville> red hat has the source code open for most everything

[21-Apr-2010 10:21:18] <lthrasher> yeah, 100% of everything i think

[21-Apr-2010 10:21:21] <venturaville> but you can't use 2 of the copyrighted binary RPMs without a license

[21-Apr-2010 10:21:36] <venturaville> redhat release and redhat logos I believe

[21-Apr-2010 10:22:03] <venturaville> which is essentially what centos is doing only they recompile everything from source

[21-Apr-2010 10:25:00] <xpot> Question: I have a device that is over a satellite connection, so it has high latency.  i increased the collector timeout from 180 to  a higher value so that it can do its collection without timeout.  however, I still do not get any performance graphs.  All graphs just show the broken image icon.  any suggestions?

[21-Apr-2010 10:25:02] <lthrasher> so the enterprise bits are what, commercially licensed?

[21-Apr-2010 10:29:28] * lthrasher relurks, still a bit confused

[21-Apr-2010 10:33:37] <cluther> lthrasher: Yes, the enterprise bits are commercially licensed.

[21-Apr-2010 10:39:37] <cgibbons> hurm

[21-Apr-2010 10:39:44] <davetoo> what's the easiest way to enumerate all the local properties for a device (or an Organizer)?  I'm looking at something like

[21-Apr-2010 10:39:44] <davetoo> [ foo.update({k:v})  for k,v  in dev.zenPropertyItems() if dev.hasProperty(x) ]

[21-Apr-2010 10:40:18] <davetoo> or does that demonstrate a clear lack of understanding? :)

[21-Apr-2010 10:40:37] <cluther> There's a lot going on there. You're trying to build this foo dictionary only if the device has a property called x?

[21-Apr-2010 10:40:49] <davetoo> oh,

[21-Apr-2010 10:40:52] <davetoo> s/x/k

[21-Apr-2010 10:41:07] <davetoo> in other words, I just want what's been defined/overridden at that node

[21-Apr-2010 10:41:23] <cluther> Looks good to me in that case.

[21-Apr-2010 10:41:32] <davetoo> I'm ... building a list for input to BatchDeviceLoader

[21-Apr-2010 10:41:54] <davetoo> dumping from old 2.3.x systems, making some changes to the results,

[21-Apr-2010 10:42:04] <davetoo> and using that to  load into my 2.5.2 systrem

[21-Apr-2010 10:42:09] <davetoo> 'k, thanks

[21-Apr-2010 10:55:29] <felbane> Does anybody know why Zenoss would fail to see an object that zendmd is able to see just fine?

[21-Apr-2010 10:56:07] <felbane> I'm getting this "The requested resource does not exist" error, but the object is there

[21-Apr-2010 10:57:00] <felbane> message/47854

[21-Apr-2010 10:57:36] <felbane> Day three and I still can't figure out what's wrong. I'm not sure where to begin for debugging this.

[21-Apr-2010 10:58:16] <cluther> felbane: There are several reasons for that. The first one is to make sure you've committed in zendmd if that's where you created the object.

[21-Apr-2010 10:58:32] <cluther> The second would be the immediate_view in the object's factory_type_information. If that skin doesn't exist, you'll get the error.

[21-Apr-2010 10:59:27] <felbane> If the definition for the object is in a zenpack, do I have to explicitly call commit()?

[21-Apr-2010 10:59:33] <cluther> A useful troubleshooting step would be to go to /manage within the web interface and remove NotFound from the list of ignored exceptions. Now when a NotFound error occurs a traceback will be sent to the event.log.

[21-Apr-2010 11:00:12] <cluther> You only have to call commit if you added or modified the object within zendmd. If it was added via zenmodeler or the UI commit will be called automatically.

[21-Apr-2010 11:00:58] <felbane> Where are the ignored exceptions defined in manage?

[21-Apr-2010 11:03:32] <felbane> Nvm found it

[21-Apr-2010 11:03:56] <felbane> Zenoss:8080/error_log/manage

[21-Apr-2010 11:04:49] <felbane> Omfg I feel so dumb

[21-Apr-2010 11:05:06] <felbane> "Objects must have a docstring to be published"

[21-Apr-2010 11:05:13] <felbane> Son of a...

[21-Apr-2010 11:05:20] * felbane slaps self

[21-Apr-2010 11:05:55] <cluther> Ahh.. the docstring caveat consumes another soul.

[21-Apr-2010 11:06:10] <felbane> Three. Days.

[21-Apr-2010 11:06:16] <felbane> -_-

[21-Apr-2010 11:06:29] <felbane> Well that one is definitely going in my guide.

[21-Apr-2010 11:08:24] <cluther> I think this is Zope's way of punishing people who don't provide docstrings for their methods.

[21-Apr-2010 11:08:47] <blull_> morning

[21-Apr-2010 11:08:53] <blull_> i just brought up a new remote collector online

[21-Apr-2010 11:08:58] <aclark> cluther: hah

[21-Apr-2010 11:09:07] <blull_> for some reason, it is running checks against hosts assigned to a different remote collector.

[21-Apr-2010 11:09:07] <aclark> indeed

[21-Apr-2010 11:09:21] <felbane> It is true punishment. I was holding off on the docstrings until I got everything working.

[21-Apr-2010 11:09:34] <felbane> Lesson learned.

[21-Apr-2010 11:09:36] <blull_> i just verified, one of the hosts I see it running ping checks against, is assigned to a remote collector called es-mon1, while this collector is called ec-mon1 (two physical locations)

[21-Apr-2010 11:09:39] <cluther> aclark! How handy.. we have someone here to blame.

[21-Apr-2010 11:09:50] <blull_> but i am seeing the ping go out ec-mon1, not es-mon1 as it should.

[21-Apr-2010 11:09:58] <aclark> hah

[21-Apr-2010 11:10:39] <blull_> Also under /opt/zenoss/perf/Devices I see a bunch of devices I monitor in this other datacenter .  All I did was add this new collector after configuring it to connect to the central zenhub :/

[21-Apr-2010 11:10:49] * felbane glares at aclark...

[21-Apr-2010 11:11:12] <felbane> Aclark: stop forcing me to be a good coder!

[21-Apr-2010 11:11:17] <felbane> :-P

[21-Apr-2010 11:11:22] <cluther> My theory is that Zope is using this trick to collect enough pounds of flesh to create a leviathan and take over the world.

[21-Apr-2010 11:12:42] <felbane> Well it's a lot closer now... despite my best efforts I'm not exactly a small guy.

[21-Apr-2010 11:12:49] * aclark is not really responsible for Zope 2 ;-) Though a core Plone guy is now the release manager

[21-Apr-2010 11:14:37] <felbane> Guilt by association?

[21-Apr-2010 11:14:40] <felbane> ;-)

[21-Apr-2010 11:15:05] <blull_> grr wtf when i turn on this new remote collector

[21-Apr-2010 11:15:14] <felbane> I feel like a scolded 5th grader right now.

[21-Apr-2010 11:15:17] <blull_> it runs a discover on an IP range ASSIGNED TO A DIFFERENT COLLECTOR :(

[21-Apr-2010 11:18:13] <cluther> blull_: It sounds like your collector is having an identify crisis. Can you verify that you have the right "monitor" name set in the collector's .conf files?

[21-Apr-2010 11:18:30] <Nube> Hi!

[21-Apr-2010 11:18:38] <blull_> yup, its ec-mon1 on the new one in every *.conf file, as well as on the central zenhub.

[21-Apr-2010 11:19:14] <blull_> wait

[21-Apr-2010 11:19:15] <Nube> I don´t stand nathing

[21-Apr-2010 11:19:15] <blull_> that made no sense.

[21-Apr-2010 11:19:32] <blull_> its ec-mon1 on the new collector, which is its hostname, in all the *.conf files on the new remote collector.

[21-Apr-2010 11:19:44] <blull_> or it was rofl...

[21-Apr-2010 11:19:49] <blull_> omg someone didn't use rcs

[21-Apr-2010 11:19:54] <blull_> i hate user error :/

[21-Apr-2010 11:20:11] <Nube> I speak spanish

[21-Apr-2010 11:21:20] <Nube> and a little more english

[21-Apr-2010 11:21:23] <Nube> ;)

[21-Apr-2010 12:21:50] <chickenandbeans> hello all

[21-Apr-2010 12:22:37] <chickenandbeans> Bit of a problem have a graph showing "missing rrd file"... but the rrd files are present on the system... anyone had this problem?

[21-Apr-2010 12:24:23] <cluther> chickenandbeans: Are the rrd files owned by zenoss:zenoss?

[21-Apr-2010 12:24:58] <chickenandbeans> yeah

[21-Apr-2010 12:25:40] <chickenandbeans> It's a bit of a new one for me :)

[21-Apr-2010 12:26:40] <cluther> Using distributed collectors?

[21-Apr-2010 12:28:06] <chickenandbeans> yes, but the rrd files in question are local :)

[21-Apr-2010 12:28:21] <chickenandbeans> As in local to the main zenoss instance

[21-Apr-2010 12:28:41] <cluther> And the device they're associated with is assigned to the localhost collector? What's the localhost collector's render url?

[21-Apr-2010 12:31:22] <chickenandbeans> yup collector is localhost and the render url is /zport/RenderServer

[21-Apr-2010 12:38:37] <chickenandbeans> Have written a script that checks for stage catalogue references (these are just processes though), could these have a knock on effect for graph rendering?

[21-Apr-2010 12:52:40] <cgibbons> hurm

[21-Apr-2010 12:56:42] <Brixius> yeah, just as I finish upgrading to 2.6 looks like I might have to do it again soon...   ;)  hopefully it will be an easier time then upgrading from 2.2 VMWare appliance to 2.6 Stack was.

[21-Apr-2010 13:13:44] <Brixius> I installed pyodbc on my ubuntu system as root, and if I go into python and type import pyodbc, it all works, but if I switch to another user and type import pyodbc in python I get the following error: pyodbc.so: undefined symbol: __gxx_personality_v0

[21-Apr-2010 13:14:33] <cluther> Brixius: If another user is the zenoss user you're likely running into the fact that Zenoss has its own python. Try installing pyodbc as the zenoss user instead.

[21-Apr-2010 13:15:45] <Brixius> I think it has something to do with the fact that root get's python 2.6.4 and zenoss is running 2.4.4, I tried building it as the zenoss user, but still get the same thing.  I think It's still using the python-dev from the system, and so linking it for 2.6.4 still.

[21-Apr-2010 13:17:45] <Brixius> just not sure how to fix it, but since it's not directly zenoss related, more system, I'll have to do some digging and find out how to fix it.

[21-Apr-2010 13:20:09] <chickenandbeans> Does anyone know a way of programmatically fetching all the catalogue names in zenoss? I've found 7, not sure if i've got them all... :)

[21-Apr-2010 13:23:44] willwh|afk is now known as willwh

[21-Apr-2010 13:56:43] <chudler> how does zenoss3 differentiate component templates from normal ones?  Like ethernetCsmacd, FileSystem, etc?  Oh actually, I guess it doesn't differentiate, like previous versions

[21-Apr-2010 13:57:20] <chudler> I thought it was only because of how they are displayed on the screen on a clean install.

[21-Apr-2010 14:06:24] <ckrough> are there any screenshots of v3 online?

[21-Apr-2010 14:06:30] <ckrough> dont have time to toy but i would like to see it

[21-Apr-2010 14:07:17] <cluther> http://public-demo.zenoss.com/

[21-Apr-2010 14:07:27] <chudler> http://public-demo.zenoss.com/ user:admin pw:zenoss

[21-Apr-2010 14:07:46] <chudler> ack, I guess its still up?  I just saw that in the forum somewheres

[21-Apr-2010 14:09:37] <Brixius> is there a way to get zenoss to send a report in e-mail at a regular basis, ie every friday e-mail a report.

[21-Apr-2010 14:12:33] <cgibbons> we just released an alpha of 3.0, too

[21-Apr-2010 14:15:12] <mrayzenoss> Brixius: reportmail + cron

[21-Apr-2010 14:16:57] <Brixius> thanks

[21-Apr-2010 14:25:04] <nonsenso> v3 looks great.

[21-Apr-2010 15:00:50] <blull_> wow lol

[21-Apr-2010 15:00:52] <blull_> just fyi

[21-Apr-2010 15:00:54] <blull_> in zendisc

[21-Apr-2010 15:00:56] <blull_> there should be a check added

[21-Apr-2010 15:01:06] <blull_> that if the argument to range is a single subnet in cidr notation

[21-Apr-2010 15:01:12] <blull_> it should prompt "did you really mean --net"

[21-Apr-2010 15:01:28] <blull_> cause i just spent like 4 hours typing zendisc --range=172.20.199.0/24 instead of zendisc --net heh

[21-Apr-2010 15:01:31] <blull_> and it would return immediately

[21-Apr-2010 15:01:35] <blull_> but then start populating systems anyway

[21-Apr-2010 15:01:39] <blull_> lol

[21-Apr-2010 15:02:05] <mrayzenoss> is that in 2.5.2 or in the alpha?

[21-Apr-2010 15:06:24] <chudler> parse_iprange in IpUtil could raise InvalidIPRangeError in this case, but it doesn't :-p

[21-Apr-2010 15:25:44] <tripitaka> Hi, I want to have switch up/down events reported, what's the best approach?

[21-Apr-2010 16:01:18] <gwb235> has anybody taken pulvers bit here (thread/9385#37475) and made it work with Netapp ?

[21-Apr-2010 16:01:57] <gwb235> I end up with "Disk space low: 816.7% used (-3583.40 GB free)".  What's the best way to step-thru/debug an event transform ?

[21-Apr-2010 16:04:19] <gwb235> maybe just steal the transform from /Perf/Filesystem and put into /Storage/Full ?

[21-Apr-2010 16:11:13] <gwb235> that seemed to work

[21-Apr-2010 18:28:34] <QbY> I know that I can create a template with the name of the type of Interface and have that applied to devices with that interface type..  However, if I want to create a template for environmental metrics--what do i do?

[21-Apr-2010 18:46:10] forsberg is now known as fOrsberg

[21-Apr-2010 18:54:26] RoAkSoAx_ is now known as andreserl

[21-Apr-2010 19:04:48] RoAkSoAx_ is now known as andreserl

[21-Apr-2010 20:14:02] <davetoo> so I need help doing somethign bad :)

[21-Apr-2010 20:14:29] <davetoo> How do I find/grab a zenPropertyManager in a running instance so that I can test a monkey-patch?

[22-Apr-2010 00:00:30] [disconnected at Thu Apr 22 00:00:30 2010]

[22-Apr-2010 00:00:31] [connected at Thu Apr 22 00:00:31 2010]

[22-Apr-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[22-Apr-2010 01:44:02] fOrsberg is now known as forsberg

[22-Apr-2010 03:49:53] <Dieterbe> does this error look familiar to anyone?

[22-Apr-2010 03:49:54] <Dieterbe> /usr/lib/nagios/plugins/check_nrpe: /usr/local/zenoss/common/lib/libssl.so.0.9.8: no version information available (required by /usr/lib/nagios/plugins/check_nrpe)

[22-Apr-2010 03:50:15] <Dieterbe> this somewhat suprises me:

[22-Apr-2010 03:50:16] <Dieterbe> $ echo $LD_LIBRARY_PATH

[22-Apr-2010 03:50:17] <Dieterbe> /usr/local/zenoss/python/lib:/usr/local/zenoss/mysql/lib:/usr/local/zenoss/zenoss/lib:/usr/local/zenoss/common/lib::

[22-Apr-2010 03:50:39] <Dieterbe> why would a normal system program need to search for libraries first in /usr/local/zenoss ?

[22-Apr-2010 04:05:56] <tehhobbit> nrpe in zenoss ?

[22-Apr-2010 07:32:07] forsberg is now known as fOrsberg

[22-Apr-2010 08:18:45] fOrsberg is now known as forsberg

[22-Apr-2010 08:26:40] <ckrough> hmm... 2.5.2 template fails over to  ethernetCsmacd_64 only if  ethernetCsmacd has problems (due to 64 bit counters) right?

[22-Apr-2010 08:27:02] <ckrough> so if my  ethernetCsmacd is 64 bit it should never fail over to  ethernetCsmacd_64?

[22-Apr-2010 09:14:35] <chickenandbeans> Found a zenoss gem :)

[22-Apr-2010 09:15:56] <chickenandbeans> re my problem yesterday of Graphs showing 'rrd file missing' but the actual rrd files are there

[22-Apr-2010 09:21:28] <rmatte> well, my 2.5.2 updates were almost a complete success

[22-Apr-2010 09:21:36] <rmatte> one of the boxes errored out on zenmigrate so I had to roll it back

[22-Apr-2010 09:22:00] <rmatte> I'm getting a lot of heartbeat errors on one of my servers

[22-Apr-2010 09:22:28] <rmatte> and people are complaining that they can no longer copy and paste multiple events from the new event console (which isn't a big deal, I'll just script something that outputs it in plaintext)

[22-Apr-2010 09:23:27] <rmatte> overall, not bad

[22-Apr-2010 09:29:45] <psurge> What is the best method and Zenpack to monitoring CPU/memory on Windows machines?

[22-Apr-2010 09:30:02] <psurge> WMI or SNMP?

[22-Apr-2010 09:30:40] <davetoo> heh

[22-Apr-2010 09:31:07] <davetoo> I kind of invalidated my 2.5.2 "before vs. after" testing

[22-Apr-2010 09:31:30] <davetoo> my events history database was so large that even just *dropping* the history table took 1/2 hour

[22-Apr-2010 09:31:53] <davetoo> eventually I dropped the entire db and started over,

[22-Apr-2010 09:32:06] <davetoo> so I'm not comparing apples to apples in the event console but it sure is snappier

[22-Apr-2010 09:35:51] <ckrough> rmatte: Im wrapping up upgrade # 4 of 6 now, hasnt been too bad

[22-Apr-2010 09:36:32] <ckrough> Im going to do HKG today, you around?

[22-Apr-2010 09:36:37] <ckrough> mt

[22-Apr-2010 09:40:08] <rmatte> psurge: you can do either

[22-Apr-2010 09:40:42] <rmatte> psurge: I personally use SNMP: docs/DOC-3570 docs/DOC-3386

[22-Apr-2010 09:41:02] <rmatte> psurge: but if you want to use WMI then have a look at Egor's WMI Data Source and WMI Performance packs

[22-Apr-2010 09:47:37] <rmatte> ckrough: HKG?

[22-Apr-2010 09:48:52] <rmatte> Has anyone come across this error when upgrading to 2.5.2: http://fpaste.org/n1Og/ ?

[22-Apr-2010 09:49:09] <rmatte> Only one of my Zenoss boxes has that problem when performing the zenmigrate

[22-Apr-2010 09:49:18] <rmatte> I can't try upgrading it again until I can figure out how to fix that

[22-Apr-2010 09:49:35] <cluther> rmatte: I've seen that before. I think one of the 2.4.x series migrate scripts never ran to get into this position.

[22-Apr-2010 09:49:46] <cluther> rmatte: You could "zenmigrate --level=2.3.3" to resolve it most likely.

[22-Apr-2010 09:50:07] <rmatte> so I do that then re-run zenmigrate with no arguments afterwards?

[22-Apr-2010 09:50:13] <rmatte> or just run that?

[22-Apr-2010 09:50:15] <cluther> Just that.

[22-Apr-2010 09:50:17] <rmatte> k

[22-Apr-2010 09:50:24] <rmatte> thanks, I'll give it a whirl during our next change window

[22-Apr-2010 09:51:18] <rmatte> cluther: I also had reports of this error: http://fpaste.org/JwiN/

[22-Apr-2010 09:51:29] <rmatte> can that be caused if someone just didn't clear their browser cache?

[22-Apr-2010 09:51:41] <rmatte> I checked the zope.conf file and utf8 is defined in there

[22-Apr-2010 09:51:51] <mrayzenoss> aclark: you see this? message/47920?tstart=0#47920

[22-Apr-2010 09:55:10] <aclark> mrayzenoss: cool! I saw that guy's email re: buildout the other day…

[22-Apr-2010 09:55:40] <rmatte> that plone4zenoss looks really neat

[22-Apr-2010 09:56:09] <cluther> rmatte: There are a lot of causes for that utf8 error. Search for utf8 at http:// dev.zenoss.org/trac/

[22-Apr-2010 09:56:27] <cluther> Some are related to events, others to Google Maps. Some are fixed in 2.5.2, and at least one is not.

[22-Apr-2010 09:56:45] <rmatte> cluther: k, thanks

[22-Apr-2010 09:57:22] <fragfutter> how can i prevent zendisc to log an Cmd/Fail error during discovery. I don't want to set an ssh password.

[22-Apr-2010 09:58:26] <rmatte> mrayzenoss: do you know if plone4zenoss is going to be open source or are they eventually going to be requiring licenses?

[22-Apr-2010 09:59:10] <rmatte> fragfutter: you mean you just want it to not attempt connecting via ssh at all during modelling?

[22-Apr-2010 10:00:18] <fragfutter> rmatte: not during discovery. If i move the discovered device to a class where zCommandPassword is set, it can try the SSH during remodelling

[22-Apr-2010 10:00:21] <mrayzenoss> rmatte: no idea, feel free to ask him message/47920?tstart=0#47920

[22-Apr-2010 10:00:39] <rmatte> fragfutter: oh, you mean during discovery, I read that wrong

[22-Apr-2010 10:01:07] <fragfutter> rmatte: it annoys me, that i run a discovery and for every device it finds i get a warning event.

[22-Apr-2010 10:01:34] <fragfutter> rmatte: wmi at least fails silently.

[22-Apr-2010 10:01:40] <mrayzenoss> fragfutter: I guess you could disable it in the /Discovered device class

[22-Apr-2010 10:01:55] <rmatte> fragfutter: you could write a transform which automatically suppressed those events if the device is in the /Discovered class

[22-Apr-2010 10:01:56] <psurge> rmatte: Thanks. I can get the ZenPacks installed without errors but I'm still trying to figure out how to apply them to a device. Noob trying to figure out what does what.

[22-Apr-2010 10:02:15] <rmatte> mrayzenoss: would he just do that in zProperties?

[22-Apr-2010 10:02:26] <fragfutter> mrayzenoss, rmatte: so not possible without writing code?

[22-Apr-2010 10:02:28] <mrayzenoss> rmatte: yes

[22-Apr-2010 10:02:35] <rmatte> mrayzenoss: which option?

[22-Apr-2010 10:02:56] <rmatte> I don't see a zCommandMonitorIgnore or zSSHMonitorIgnore

[22-Apr-2010 10:03:31] <mrayzenoss> rmatte: me neither, nuts

[22-Apr-2010 10:03:32] <rmatte> and I also don't see an ssh plugin in collector plugins for that class

[22-Apr-2010 10:03:47] <rmatte> so the transform would really be the only way of doing it

[22-Apr-2010 10:03:57] <aclark> mrayzenoss: Dear gstreiff, why not hire aclark to help with all of these build issues including buildout? :-)

[22-Apr-2010 10:03:57] <mrayzenoss> rmatte: zenoss.cmd.uname

[22-Apr-2010 10:04:14] <aclark> mrayzenoss: glad he got back to the guy at least

[22-Apr-2010 10:04:15] <rmatte> that's the plugin?

[22-Apr-2010 10:04:32] <mrayzenoss> the 'cmd' ones are the SSH calls

[22-Apr-2010 10:04:34] <rmatte> I only see snmp plugins in mine under /Discovered

[22-Apr-2010 10:04:37] <rmatte> and I haven't changed it

[22-Apr-2010 10:05:00] <rmatte> and I doubt he would have added it

[22-Apr-2010 10:05:01] <mrayzenoss> rmatte: under http://server:8080/zport/dmd/Devices/Discovered/zSortableProperty?prop=zCollectorPlugins

[22-Apr-2010 10:05:29] <rmatte> yeh I'm there

[22-Apr-2010 10:05:31] <rmatte> I don't see it

[22-Apr-2010 10:05:37] <rmatte> it's in the add fields section

[22-Apr-2010 10:05:45] <mrayzenoss> weird, I'm on a vanilla RPM 2.5.2 install

[22-Apr-2010 10:05:59] <rmatte> oh, well, this particular install dates back to 2.4.1

[22-Apr-2010 10:06:01] <fragfutter> i have a clean zenoss installed from rpm zenoss-2.5.2.el5.x86_64.rpm and in discovered there are five zenoss.cmd plugins

[22-Apr-2010 10:06:04] <rmatte> so maybe it changed

[22-Apr-2010 10:06:16] <rmatte> fragfutter: go to the Discovered class under Devices

[22-Apr-2010 10:06:21] <mrayzenoss> fragfutter: if you remove those collector plugins, it won't attempt to use SSH

[22-Apr-2010 10:06:24] <rmatte> then from the dropdown select More -> Collector Plugins

[22-Apr-2010 10:06:33] <fragfutter> mrayzenoss: ah. thats fine

[22-Apr-2010 10:06:41] <rmatte> do you see any zenoss.cmd.* plugins listed there?

[22-Apr-2010 10:06:49] <rmatte> if so, click the x next to them to remove them from the list

[22-Apr-2010 10:06:53] <fragfutter> rmatte: thanks already there. yes five of them.

[22-Apr-2010 10:06:58] <rmatte> k, coo

[22-Apr-2010 10:06:59] <fragfutter> rmatte: you can stop typing ;)

[22-Apr-2010 10:07:02] <rmatte> hehe

[22-Apr-2010 10:08:58] <fragfutter> it worked. thanks.

[22-Apr-2010 10:12:54] <aclark> mrayzenoss: is that Jive running community.z.o ?

[22-Apr-2010 10:13:08] <mrayzenoss> aclark: yes

[22-Apr-2010 10:13:30] <fragfutter> i defined a template at /devices/network/switch/foundry that graphs the CPU usage. The newly discovered device has been moved to this class. Shouldn't it now also collect performance data for this device?

[22-Apr-2010 10:13:42] <mrayzenoss> fragfutter: did you bind the template?

[22-Apr-2010 10:14:19] <fragfutter> mrayzenoss: no. Do i need to bind it for every device i move into this class? it is not automaticly inherited?

[22-Apr-2010 10:14:32] <mrayzenoss> it needs to be bound for the device class

[22-Apr-2010 10:14:48] <mrayzenoss> when devices get added to the class they'll get it

[22-Apr-2010 10:15:27] <fragfutter> mrayzenoss: the device-class is devices/network/switch/foundry?

[22-Apr-2010 10:15:31] <mrayzenoss> yes

[22-Apr-2010 10:16:27] <aclark> mrayzenoss: what does that cost roughly?

[22-Apr-2010 10:16:46] <mrayzenoss> aclark: I'm not exactly sure

[22-Apr-2010 10:17:35] <fragfutter> mrayzenoss: i think i got it. There is no indicator if the templates are bound or not?

[22-Apr-2010 10:19:51] <mrayzenoss> fragfutter: if you click on bind templates, you should see the bound ones selected

[22-Apr-2010 10:21:17] <fragfutter> mrayzenoss: ok. i expected an indicator in the overview. but ok.

[22-Apr-2010 10:22:14] <mrayzenoss> fragfutter: I think in the 3.0 release it's more apparent

[22-Apr-2010 10:24:23] <baffle> Hmm, I think zeopack is not working correctly/launching from cron in my installation. Data.fs is 47G.

[22-Apr-2010 10:25:43] <baffle> It's not supposed to, is it? :)

[22-Apr-2010 10:25:44] <tehhobbit> mrayzenoss: how far away is zenoss 3 ?

[22-Apr-2010 10:25:56] <mrayzenoss> tehhobbit: we put out the first alpha 2 days ago

[22-Apr-2010 10:26:04] <tehhobbit> oooh

[22-Apr-2010 10:26:10] <tehhobbit> where can I get it

[22-Apr-2010 10:26:16] <tehhobbit> would be fun to have a look at

[22-Apr-2010 10:26:26] <baffle> tehhobbit: Topic?

[22-Apr-2010 10:26:31] <fragfutter> so i can define a template somewhere in my device tree. Bind it somewhere below this tree. all devices below this bind point will automaticly pick it up.

[22-Apr-2010 10:26:41] <mrayzenoss> tehhobbit: http://bit.ly/9xmYxX

[22-Apr-2010 10:27:47] <mrayzenoss> fragfutter: I'd verify that… I know when you define the template it's available everywhere further down the hierarchy, but I don't believe binding at 1 level binds below

[22-Apr-2010 10:28:05] <tehhobbit> mrayzenoss baffle tnx

[22-Apr-2010 10:28:09] <forsberg> yes fragfutter :> try it out :)

[22-Apr-2010 10:28:11] <tehhobbit> sry missed topic

[22-Apr-2010 10:28:41] <forsberg> oh, what mrayzenoss said

[22-Apr-2010 10:31:11] <fragfutter> so how do the device/.../node/os/interface performance data get to their template?

[22-Apr-2010 10:31:37] <baffle> mrayzenoss: Is 47G Data.fs a normal size?

[22-Apr-2010 10:32:12] <fragfutter> baffle: not for a normal zope instance, but i know nothing about zenoss

[22-Apr-2010 10:32:21] <tehhobbit> new interface have a nice feeling to it

[22-Apr-2010 10:32:24] <mrayzenoss> baffle: sounds unusually large

[22-Apr-2010 10:32:35] <baffle> mrayzenoss: That's what she said!

[22-Apr-2010 10:33:14] <fragfutter> baffle: you can try and pack it through the web

[22-Apr-2010 10:33:28] <baffle> mrayzenoss: Also, is this the "correct" way to run zeopack? -> /usr/local/zenoss/python/bin/python /usr/local/zenoss/zenoss/bin/zeopack.py -h localhost -p 8100 -d 1

[22-Apr-2010 10:33:55] <baffle> Just doing zeopack as user "zenoss" gives this:

[22-Apr-2010 10:33:55] <baffle> zenoss@zenoss:~$ zeopack

[22-Apr-2010 10:33:55] <baffle> 'import site' failed; use -v for traceback

[22-Apr-2010 10:33:55] <baffle> warning: Not importing directory '/usr/local/zenoss/common/lib/gettext': missing __init__.py

[22-Apr-2010 10:33:59] <baffle> -su: zeopack: command not found

[22-Apr-2010 10:34:11] <baffle> Err, my bad. :)

[22-Apr-2010 10:34:57] <baffle> Doing zeopack as user "zenoss" without specifying path to python gives this: http://paste2.org/p/787954

[22-Apr-2010 10:35:16] <rmatte> baffle: It always baffles my mind (no pun intended) how people get Data.fs files that large, the one on my largest Zenoss install is 967MB

[22-Apr-2010 10:35:38] <baffle> rmatte: Well, the Data.fs.old is 97G.

[22-Apr-2010 10:35:46] <rmatte> that's massive

[22-Apr-2010 10:35:52] <baffle> rmatte: I seriously do *not* understand why it is so huge.

[22-Apr-2010 10:36:15] <mrayzenoss> baffle: according to the Admin Guide, the syntax is $ZENHOME/bin/zeopack.py -p 8100

[22-Apr-2010 10:36:31] <rmatte> did ou change a Zope setting that might affect the size?

[22-Apr-2010 10:36:39] <rmatte> s/ou/you

[22-Apr-2010 10:37:33] <baffle> rmatte: Ofcourse, you monitor networks, and I monitor (mostly) servers. I think one of the problems is that by default snmpd outputs the full tcp/udp state table, and I think zenoss might registere these.. :-/ Not funny when you monitor large websites with thousands of connections pr. host..

[22-Apr-2010 10:38:03] <baffle> rmatte: Uhm. Not sure?

[22-Apr-2010 10:38:09] <rmatte> well, I monitor a fair amount of servers and networks

[22-Apr-2010 10:38:25] <rmatte> yeh, true

[22-Apr-2010 10:38:35] <rmatte> I don't monitor many webservers

[22-Apr-2010 10:39:04] <baffle> ZODB-stuff from zope.conf: http://paste2.org/p/787963

[22-Apr-2010 10:39:28] <baffle> mrayzenoss: Mmm, guess it is mostly the same, except I have specified -d 1 wich means pack everything older than 1 day..

[22-Apr-2010 10:39:39] <baffle> mrayzenoss: But it seems the cronjob doesn't work. :-/

[22-Apr-2010 10:39:44] <fragfutter> baffle: try to pack it through the web.

[22-Apr-2010 10:40:07] <fragfutter> and see if it shrinks

[22-Apr-2010 10:42:06] <baffle> Is there any way to observe the shrinkage?

[22-Apr-2010 10:42:14] <baffle> I.e. does it create a new file that grows?

[22-Apr-2010 10:42:34] <mrayzenoss> baffle: I know it backs up the old one

[22-Apr-2010 10:42:43] <fragfutter> baffle: the webinterface of zope shows the active size.

[22-Apr-2010 10:43:32] <baffle> mrayzenoss: Hmm, so maybe it would move Data.fs to Data.fs.old and start with a zeroed Data.fs if it worked?

[22-Apr-2010 10:44:01] <mrayzenoss> baffle: dunno, let me know :)

[22-Apr-2010 10:46:37] <baffle> Total number of objects in the database

[22-Apr-2010 10:46:38] <baffle> 12842184

[22-Apr-2010 10:47:16] <baffle> Hmm, pressing "Pack" results in a "Site Error".

[22-Apr-2010 10:48:41] <baffle> Oh, it is packing. Yay, I found the secret place to find that out. :)

[22-Apr-2010 10:48:45] <fragfutter> baffle: diskspace?

[22-Apr-2010 10:48:55] <fragfutter> baffle: or web timeout?

[22-Apr-2010 10:48:56] <baffle>   File "/usr/local/zenoss/zenoss/lib/python/ZODB/FileStorage/FileStorage.py", line 1233, in pack

[22-Apr-2010 10:49:00] <baffle>     raise FileStorageError('Already packing')

[22-Apr-2010 10:49:35] <baffle> So, it is somehow allready packing. Would be cool to have some kind of progress, tho.

[22-Apr-2010 10:50:15] <felbane> is there a problem with exporting zenpacks in 2.4.5?

[22-Apr-2010 10:50:19] <baffle> Yeah, the disk is reading 30MB/sec, so I guess that is it.

[22-Apr-2010 10:50:29] <felbane> i.e. not zipping up all subdirectories?

[22-Apr-2010 10:52:42] <mrayzenoss> felbane: are they .zip or .egg zenpacks?

[22-Apr-2010 10:53:28] <felbane> it's a custom "development mode" zenpack that i built in my dev zenoss 2.4.5 vm

[22-Apr-2010 10:53:33] <felbane> and now i want to export it

[22-Apr-2010 10:53:45] <felbane> but when i do all i get are the py files

[22-Apr-2010 10:53:58] <felbane> no objects.xml, no skins, no dtml, etc

[22-Apr-2010 10:55:49] <felbane> i havent tried it yet but i think if i do a setup.py build and then copy everything into the build folder it might work

[22-Apr-2010 10:56:01] <mrayzenoss> felbane: yeah, I was going to suggest that

[22-Apr-2010 10:56:17] <felbane> is that a problem in 2.5 as well?

[22-Apr-2010 10:56:23] <mrayzenoss> I'm not sure

[22-Apr-2010 10:56:50] <felbane> if so, major pita for zenpack devs.

[22-Apr-2010 10:57:07] <felbane> idk where the setup.py's code comes from

[22-Apr-2010 10:57:36] <felbane> i guess setuptools isn't made to recurse entire build directories?

[22-Apr-2010 10:59:12] <felbane> its like include_package_data is ignoring everything that isnt *.py

[22-Apr-2010 10:59:33] <felbane> OH.

[22-Apr-2010 10:59:44] <felbane> my manifest got deleted

[22-Apr-2010 10:59:48] <felbane> thats probably it

[22-Apr-2010 11:00:17] <rmatte> man the filtering in finicky in this new event console

[22-Apr-2010 11:00:44] <felbane> rmatte: 3.0?

[22-Apr-2010 11:01:33] <rmatte> no, 2.5.2

[22-Apr-2010 11:01:51] <rmatte> when I go to event history and put something in to the summary filter it takes like a day to load anything

[22-Apr-2010 11:02:01] <baffle> Oh, yeah, does winexe work for people by the way? The included one never has for me, I have a static binary I got from a forum once that I always end up using..

[22-Apr-2010 11:02:05] <felbane> mrayzenoss: yep, MANIFEST.in didnt have the "graft ZenPacks" line

[22-Apr-2010 11:02:48] <fragfutter> mrayzenoss: templates are in the zDeviceTemplates property. The moment i define a leaf with this property it will be valid for all leafs underneath, until i overwrite it again.

[22-Apr-2010 11:03:05] <fragfutter> mrayzenoss: typical zope property inheritance (without merging)

[22-Apr-2010 11:04:25] <baffle> I.e. if I use the supplied winexe (0.80) I get "ERROR: Failed to install service winexesvc - NT code 0x00000424", but the 0.90 I got from a forum works.

[22-Apr-2010 11:05:27] <baffle> Ah, yes, because the machine is 64bit, and winexe 0.80 doesn't support 64bit.

[22-Apr-2010 11:05:39] <rmatte> hehe

[22-Apr-2010 11:06:39] <felbane> baffle: not so baffled now?

[22-Apr-2010 11:06:50] <felbane> *sad trombone*

[22-Apr-2010 11:06:58] <rmatte> nice, now I'm getting a connection error every time I want to view event history

[22-Apr-2010 11:07:03] * rmatte shakes fists

[22-Apr-2010 11:09:05] * davetoo yawns

[22-Apr-2010 11:09:09] <davetoo> I shouldn

[22-Apr-2010 11:09:15] <davetoo> not stay up so late working on my zenoss tools

[22-Apr-2010 11:14:38] <baffle> Hmm, wonder if zenoss-stack works in Lucid.

[22-Apr-2010 11:40:06] <davetoo> so if I were to build from the 2.5.x-eval branch.. that's where this 3.0 alpha is from?

[22-Apr-2010 11:40:35] <bbibeault> @davetoo - I don't think so, I don't even know what that branch is.  Do you see a 2.5.70 branch?

[22-Apr-2010 11:41:03] <davetoo> (I'm looking via the trac browser.. lemme look with tortoise)

[22-Apr-2010 11:42:09] <davetoo> branches, not tag, right?

[22-Apr-2010 11:42:26] <davetoo> Not in the public repo

[22-Apr-2010 11:42:55] <bbibeault> I'm sorry, I am not familiar enough with the layout in this, but I do know that what we are doing the new UI with is the 2.5.70 branch.

[22-Apr-2010 11:43:30] <davetoo> ok, I'll try a release package

[22-Apr-2010 11:56:03] <mrayzenoss> davetoo: SVN trunk is the current alpha

[22-Apr-2010 11:56:12] <mrayzenoss> davetoo: 2.5.x is the current stable branch

[22-Apr-2010 11:57:29] <davetoo> ok, thanks

[22-Apr-2010 11:59:54] <davetoo> trying to figure out what this new "Enter password:" line is from when I start the source install (after the mysql server port line)

[22-Apr-2010 12:02:04] <rmatte> hmmm, this utf8 error seems to only be present in IE

[22-Apr-2010 12:02:09] <rmatte> and clearing the cache does nothing to fix it

[22-Apr-2010 12:02:14] <rmatte> this will be fun to figure out

[22-Apr-2010 12:08:08] <rmatte> it's also only present on one specific server

[22-Apr-2010 12:12:35] <rmatte> and only for certain devices

[22-Apr-2010 12:22:43] <rmatte> eugh, nothing can ever just work, there always has to be something

[22-Apr-2010 12:51:45] <rmatte> bah, everything I'm finding on the zenoss site is in relation to the event console, which loads perfectly fine... it's when I try to go to certain device pages that I get the error

[22-Apr-2010 12:51:56] <rmatte> UnicodeDecodeError: 'utf8' codec can't decode byte 0x96 in position 15: unexpected code byte

[22-Apr-2010 13:02:49] * ckrough /topic 'eugh, nothing can ever just work, there always has to be something'

[22-Apr-2010 13:04:43] <rmatte> lol

[22-Apr-2010 13:04:57] <rmatte> ckrough: that netsplit just proved my point

[22-Apr-2010 13:18:33] <rmatte> mrayzenoss: http://dev.zenoss.org/trac/ticket/6568

[22-Apr-2010 13:19:23] <mrayzenoss> hmmm… dunno.

[22-Apr-2010 13:19:33] <mrayzenoss> defect review is in 40 minutes, we'll look at it

[22-Apr-2010 13:19:40] <rmatte> It's not like any other utf8 errors that I could find

[22-Apr-2010 13:19:46] <rmatte> thanks

[22-Apr-2010 13:32:33] <tuxknowledge> quick Q: is ZenOss free for commercial use to be used to monitor internal Servers only?

[22-Apr-2010 13:33:31] <mrayzenoss> it's free for any use

[22-Apr-2010 13:33:36] <mrayzenoss> GPLv2

[22-Apr-2010 13:35:31] <tuxknowledge> ok, I got a bit confused cost wise looking at the pricing guide at http://www.zenoss.com/product/pricing

[22-Apr-2010 13:35:52] <mrayzenoss> that's for Zenoss Enterprise.

[22-Apr-2010 13:35:55] <tuxknowledge> but I assume that is only for ZenOss Enterprise

[22-Apr-2010 13:36:03] <tuxknowledge> ok, but no cost at all for core?

[22-Apr-2010 13:36:19] <mrayzenoss> Zenoss, or Zenoss Core is free and open source software

[22-Apr-2010 13:36:21] <mrayzenoss> zenoss.org

[22-Apr-2010 13:36:33] <tuxknowledge> excellent

[22-Apr-2010 13:36:50] <tuxknowledge> I am one of those users that believes just because it is open source does not mean it is free

[22-Apr-2010 13:37:05] <tuxknowledge> Thanks for clearing that up for me

[22-Apr-2010 13:37:36] <rmatte> when in doubt check the licensing

[22-Apr-2010 13:39:33] <tuxknowledge> does zenoss core support autodetection based on a range of IP to scan?

[22-Apr-2010 13:39:41] <rmatte> yes

[22-Apr-2010 13:40:13] <rmatte> though it's not a job that you can schedule or anything, it's a one time thing

[22-Apr-2010 13:40:28] <rmatte> you can always setup a cron job to do it at intervals if required

[22-Apr-2010 13:41:02] <tuxknowledge> is it auto detection based on manual runs?

[22-Apr-2010 13:41:15] <tuxknowledge> not like live auto detection

[22-Apr-2010 13:41:19] <rmatte> well, perhaps you could explain in detail what you're expecting?

[22-Apr-2010 13:41:30] <rmatte> then I can explain whether it can do it or not

[22-Apr-2010 13:41:40] <tuxknowledge> Like I build out a new server and zenoss finds it without any user interaction

[22-Apr-2010 13:42:28] <rmatte> it doesn't do that right out of the box

[22-Apr-2010 13:42:36] <rmatte> you could get it to with some development work

[22-Apr-2010 13:42:59] <mrayzenoss> tuxknowledge: if you're using Puppet there's a module to push it into Zenoss automatically

[22-Apr-2010 13:43:03] <rmatte> autodiscovery is really messy though generally, unless that particular subnet will just be used for servers or whatever

[22-Apr-2010 13:43:13] <mrayzenoss> tuxknowledge: and I'm working on one for Chef

[22-Apr-2010 13:44:08] <tuxknowledge> well we do cloud computing so all of our stuff is server instances, but we only get select IP ranges like, 208.113.68.2-14

[22-Apr-2010 13:44:24] <tuxknowledge> so obviously I dont want to scan 208.113.68.1-255

[22-Apr-2010 13:44:33] <mrayzenoss> right, so you could discover that manually

[22-Apr-2010 13:44:53] <mrayzenoss> just enter 208.113.68.2-14 in the add device wizard

[22-Apr-2010 13:44:59] <tuxknowledge> I see

[22-Apr-2010 13:45:16] <mrayzenoss> and then you could have zenoss periodically rescan that range for new devices with cron

[22-Apr-2010 13:45:23] <tuxknowledge> ok

[22-Apr-2010 13:45:26] <rmatte> or he could cron zendisc to run

[22-Apr-2010 13:45:33] <mrayzenoss> and the existing devices would get remodeled periodicially

[22-Apr-2010 13:45:34] <rmatte> yeh

[22-Apr-2010 13:45:35] <tuxknowledge> is there an API or something that gets used

[22-Apr-2010 13:45:49] <mrayzenoss> API?  For your devices or for Zenoss?

[22-Apr-2010 13:45:57] <tuxknowledge> for Zenoss

[22-Apr-2010 13:46:04] <rmatte> Zenoss supports SOAP API and REST

[22-Apr-2010 13:46:10] <tuxknowledge> oh ok

[22-Apr-2010 13:46:17] <rmatte> you can also use straight python

[22-Apr-2010 13:46:18] <tuxknowledge> I was wondering how I would cron it

[22-Apr-2010 13:46:20] <mrayzenoss> not SOAP, XML-RPC

[22-Apr-2010 13:46:40] <mrayzenoss> and there are CLI interfaces for lots of common things, like discovering a network

[22-Apr-2010 13:46:45] <tuxknowledge> ok

[22-Apr-2010 13:46:45] <rmatte> cronning it would be really simply, there's an executable for the discovery that supports commandline options

[22-Apr-2010 13:47:02] <rmatte> run zendisc as the zenoss user

[22-Apr-2010 13:47:07] <tuxknowledge> this company I recently started working with has zenoss core, but I have never used it

[22-Apr-2010 13:47:20] <rmatte> It's quite nice

[22-Apr-2010 13:47:40] <tuxknowledge> I come from a Nagios/openNMS monitoring system

[22-Apr-2010 13:47:51] <mrayzenoss> tuxknowledge: community/about/webinars

[22-Apr-2010 13:47:53] <tuxknowledge> zenoss looks pretty but dont really understand it yet

[22-Apr-2010 13:47:59] <mrayzenoss> next one is in 10 days

[22-Apr-2010 13:48:09] <rmatte> Zenoss is like Nagios and Cacti wrapped in to 1 product

[22-Apr-2010 13:48:15] <tuxknowledge> I see

[22-Apr-2010 13:48:17] <mrayzenoss> or… if you're at LinuxFest NorthWest this weekend, I'll be happy to give you a hands-on demo

[22-Apr-2010 13:48:21] <tuxknowledge> Cacti is cool

[22-Apr-2010 13:48:26] <tuxknowledge> Never used it but looks useful

[22-Apr-2010 13:48:39] <tuxknowledge> Nah, I am in the SouthEast

[22-Apr-2010 13:48:43] <rmatte> yeh, Zenoss does the same type of graphing using rrdtool as well

[22-Apr-2010 13:48:50] <tuxknowledge> cool

[22-Apr-2010 13:49:00] <tuxknowledge> The more I hear the more interested I am

[22-Apr-2010 13:49:19] <tuxknowledge> I think I may dive more into it and play with the zenoss core install that my company has

[22-Apr-2010 13:49:21] <rmatte> what types of servers?

[22-Apr-2010 13:49:23] <rmatte> Linux, windows?

[22-Apr-2010 13:49:25] <tuxknowledge> CentOS

[22-Apr-2010 13:49:28] <rmatte> ah

[22-Apr-2010 13:49:41] <tuxknowledge> I am thinking SNMP is going to be my best friend

[22-Apr-2010 13:49:48] <rmatte> Zenoss' Linux monitoring is really good

[22-Apr-2010 13:50:04] <rmatte> There are a couple of annoying default thresholds that you'll want to delete from the template at first

[22-Apr-2010 13:50:12] <rmatte> takes 2 seconds to do

[22-Apr-2010 13:50:34] <tuxknowledge> I see

[22-Apr-2010 13:50:34] <rmatte> yeh, I do all of my monitoring via SNMP, even for windows hosts

[22-Apr-2010 13:51:08] <tuxknowledge> I ahve an OpenNMS server I added to my companies zenoss core. It detected openNMS is running snmp but says it is down

[22-Apr-2010 13:51:15] <tuxknowledge> but doesnt detect anything else

[22-Apr-2010 13:51:35] <tuxknowledge> thats a bit disconcerting

[22-Apr-2010 13:51:39] <rmatte> did you specify the snmp string in zProperties?

[22-Apr-2010 13:51:44] <tuxknowledge> yes

[22-Apr-2010 13:52:01] <rmatte> are you able to snmpwalk the OpenNMS box from the Zenoss server with that string?

[22-Apr-2010 13:52:11] <tuxknowledge> wait, maybe I didnt

[22-Apr-2010 13:52:20] <tuxknowledge> I specified the community string somewhere

[22-Apr-2010 13:52:47] <rmatte> It needs to be specified in zSnmpCommunity not in zSnmpCommunities

[22-Apr-2010 13:52:48] <tuxknowledge> yea it is in there

[22-Apr-2010 13:52:58] <rmatte> zSnmpCommunities is only used for autodiscovery

[22-Apr-2010 13:53:16] <rmatte> people tend to mix those up at first

[22-Apr-2010 13:53:18] <tuxknowledge> it is a default install of Net-snmp

[22-Apr-2010 13:53:28] <rmatte> default with no configuration done?

[22-Apr-2010 13:53:34] <rmatte> that's not going to get you very far

[22-Apr-2010 13:53:39] <tuxknowledge> zSnmpCommunity is set to public

[22-Apr-2010 13:53:49] <tuxknowledge> no it isnt

[22-Apr-2010 13:53:53] <rmatte> yeh, but the permissions will be all wrong

[22-Apr-2010 13:54:11] <tuxknowledge> can you walk me through it to understand how it works?

[22-Apr-2010 13:54:17] <tuxknowledge> or do you not wanna do that

[22-Apr-2010 13:54:28] <rmatte> http://pastebin.com/sp25cMif

[22-Apr-2010 13:54:31] <rmatte> take that config

[22-Apr-2010 13:54:36] <rmatte> replace your snmpd.conf with that

[22-Apr-2010 13:54:59] <rmatte> then change the trapsink and trap2sink IPs to the IP of the Zenoss server (so that it knows where to send SNMP traps to)

[22-Apr-2010 13:55:11] <rmatte> then set your-key to the snmp community string of your choice

[22-Apr-2010 13:55:28] <rmatte> there are 4 instances of your-key to change in that config

[22-Apr-2010 13:55:35] <rmatte> Then change syslocation to whatever

[22-Apr-2010 13:55:40] <rmatte> and change syscontact to whatever

[22-Apr-2010 13:55:46] <rmatte> save it, then restart net-snmp

[22-Apr-2010 13:55:53] <tuxknowledge> ok give me a second

[22-Apr-2010 13:55:54] <rmatte> then put that community string in to Zenoss

[22-Apr-2010 13:55:57] <rmatte> k

[22-Apr-2010 13:58:17] <tuxknowledge> ok, snmp restarted with that config

[22-Apr-2010 13:58:29] <tuxknowledge> trap sent to 74.3.224.67 with key public

[22-Apr-2010 13:58:39] <tuxknowledge> obviously I dont care about security

[22-Apr-2010 13:58:46] <tuxknowledge> not on this system at least

[22-Apr-2010 13:58:59] <rmatte> I still wouldn't use public, what's the point?

[22-Apr-2010 13:59:04] <rmatte> it takes 2 seconds to change it

[22-Apr-2010 13:59:26] <tuxknowledge> it is testing

[22-Apr-2010 13:59:32] <tuxknowledge> who care what it is

[22-Apr-2010 13:59:43] <tuxknowledge> snmpwalk gets no response

[22-Apr-2010 14:00:00] <rmatte> firewall?

[22-Apr-2010 14:00:05] <rmatte> are you able to at least ping it?

[22-Apr-2010 14:00:08] <tuxknowledge> I wouldnt think

[22-Apr-2010 14:00:24] <tuxknowledge> I know iptables alows the port 161

[22-Apr-2010 14:00:33] <tuxknowledge> I personally added the exception

[22-Apr-2010 14:00:40] <tuxknowledge> ping resonds

[22-Apr-2010 14:00:48] <rmatte> did you add it for the wrong protocol per chance?

[22-Apr-2010 14:00:52] <rmatte> was it added for tcp or udp?

[22-Apr-2010 14:00:58] <tuxknowledge> tcp

[22-Apr-2010 14:01:25] <tuxknowledge> -A RH-Firewall-1-INPUT -m state --state NEW -m tcp -p tcp --dport 161 -j ACCEPT

[22-Apr-2010 14:01:26] <rmatte> snmp is udp

[22-Apr-2010 14:01:29] <tuxknowledge> ah

[22-Apr-2010 14:02:06] <tuxknowledge> there it is

[22-Apr-2010 14:02:09] <rmatte> ;)

[22-Apr-2010 14:02:11] <tuxknowledge> snmpwalk works now

[22-Apr-2010 14:02:24] <tuxknowledge> it is always simple things

[22-Apr-2010 14:02:33] <tuxknowledge> I swear it is always a checkbox

[22-Apr-2010 14:02:38] <rmatte> hehe

[22-Apr-2010 14:02:47] <rmatte> if snmpwalk works now then zenoss will model it correctly

[22-Apr-2010 14:02:50] <gaieges> can someone show me how to disable http polling on a given host ?

[22-Apr-2010 14:03:11] <rmatte> gaieges: well, what HTTP polling are you referring to?

[22-Apr-2010 14:03:29] <tuxknowledge> Thanks rmatte

[22-Apr-2010 14:03:37] <rmatte> tuxknowledge: no problem

[22-Apr-2010 14:04:08] <gaieges> oh ya know what.. i've been playin with this for like a half hour, and right as i come here to ask - I found exactly what i want to do!

[22-Apr-2010 14:04:13] <gaieges> sorry, and thanks rmatte

[22-Apr-2010 14:04:45] <tuxknowledge> I tell people all the time it is broke until you ask for help. Then the problem goes away

[22-Apr-2010 14:04:50] <tuxknowledge> LOL

[22-Apr-2010 14:05:01] <rmatte> gaieges: haha, nice, np

[22-Apr-2010 14:05:21] <rmatte> I like it when problems fix themselves and people answer their own questions :)

[22-Apr-2010 14:05:24] <tuxknowledge> can I for zenoss to do a serice scan on this server or o I have to wait

[22-Apr-2010 14:05:32] <tuxknowledge> force*

[22-Apr-2010 14:05:45] <tuxknowledge> wow typos

[22-Apr-2010 14:05:55] <rmatte> tuxknowledge: go to the device page for the server, and from the dropdown select More -> Model Device

[22-Apr-2010 14:06:03] <rmatte> or sorry, Manage -> Model Device

[22-Apr-2010 14:06:23] <tuxknowledge> got it

[22-Apr-2010 14:06:28] <tuxknowledge> Thanks

[22-Apr-2010 14:06:32] <rmatte> np

[22-Apr-2010 14:07:51] <rmatte> If you want it to monitor certain processes and automatically discover them on your servers, click on Processes on the left, then add a new process with the name of it

[22-Apr-2010 14:08:04] <rmatte> for example, if I wanted to automatically detect/monitor cron I'd create one called cron

[22-Apr-2010 14:08:18] <tuxknowledge> i see

[22-Apr-2010 14:08:39] <rmatte> make sure the regex matches what you want and that Monitor is set to True for the process

[22-Apr-2010 14:09:17] <tuxknowledge> I see

[22-Apr-2010 14:09:26] <tuxknowledge> I hate regex, I just dont get em

[22-Apr-2010 14:09:30] <rmatte> There is also an ignore parameters option.  That tells Zenoss to ignore any options after the process name (if you don't want it to differentiate between multiple processes running with the same name)

[22-Apr-2010 14:09:36] <tuxknowledge> I have tried to understand but get lost everytime

[22-Apr-2010 14:09:43] <rmatte> well, you don't have to use actual regex, just the name of the process is enough

[22-Apr-2010 14:09:52] <tuxknowledge> ok

[22-Apr-2010 14:09:55] <rmatte> if you ever do need to use actual regex syntax you can do that too, I never bother personally

[22-Apr-2010 14:10:25] <tuxknowledge> I like what I am seeing

[22-Apr-2010 14:10:40] <tuxknowledge> seems like it is definately more user friendly than Nagios or OpenNMS

[22-Apr-2010 14:10:58] <tuxknowledge> I can see I will spend less time in config files and more in the web console

[22-Apr-2010 14:11:11] <rmatte> by the way, if you go to Devices/Server/Linux then click on the Templates tab and click on the "Device" template...

[22-Apr-2010 14:11:40] <rmatte> you can disable some of the more annoying thresholds, such as low swap

[22-Apr-2010 14:11:52] <rmatte> it's harcoded for 1GB for whatever reason, which is useless

[22-Apr-2010 14:11:54] <tuxknowledge> got it

[22-Apr-2010 14:11:55] <rmatte> so I just disable it

[22-Apr-2010 14:12:03] <tuxknowledge> This is useful

[22-Apr-2010 14:12:17] <rmatte> that's the interface used to create device templates

[22-Apr-2010 14:12:20] <rmatte> it's relatively simple

[22-Apr-2010 14:12:38] <tuxknowledge> Sweet

[22-Apr-2010 14:12:46] <rmatte> the whole interface is being redone in the next version by the way

[22-Apr-2010 14:12:50] <rmatte> so it'll be even slicker looking

[22-Apr-2010 14:12:52] <tuxknowledge> I will definatel spend some time reading doco and learnin the system

[22-Apr-2010 14:13:20] <tuxknowledge> I am not worried about how it looks, I just want it to work. But making it look better doesnt hurt

[22-Apr-2010 14:13:28] <tuxknowledge> I am a no fluff kind of guy

[22-Apr-2010 14:13:47] <rmatte> I'm the same way

[22-Apr-2010 14:13:56] <rmatte> but they wanted to focus on the interface

[22-Apr-2010 14:14:11] <rmatte> I can't wait until it's done so that they can get back to fixing stuff that actually matters more to me

[22-Apr-2010 14:16:37] <tuxknowledge> lol

[22-Apr-2010 14:17:39] <mrayzenoss> I actually think the new UI will make the application work better, since there will be less confusion

[22-Apr-2010 14:17:58] <rmatte> yeh, I'm liking what I'm seeing so far

[22-Apr-2010 14:18:26] <tuxknowledge> I do like the customizable home page

[22-Apr-2010 14:18:28] <rmatte> I'm not saying it's a bad thing, I just would have preferred to see a bit more work done on things like stability and reporting beforehand, but oh well

[22-Apr-2010 14:18:39] <tuxknowledge> we do Cloud Computing so a map doesn't work for us

[22-Apr-2010 14:18:44] <tuxknowledge> but seeing info is nice

[22-Apr-2010 14:18:57] <rmatte> yeh, the map is nice in my case

[22-Apr-2010 14:19:14] <rmatte> http://dmon.org/graphics/zenoss/gen2.png

[22-Apr-2010 14:19:56] <rmatte> mrayzenoss: haha, I see that they ommitted the ugly overview page

[22-Apr-2010 14:20:04] <tuxknowledge> Thats a lot going on

[22-Apr-2010 14:20:10] <rmatte> tuxknowledge: yup

[22-Apr-2010 14:21:16] <mrayzenoss> tuxknowledge: you need to be able to use http://farm5.static.flickr.com/4027/4498415130_22a092693b_b.jpg for your google map

[22-Apr-2010 14:21:39] <rmatte> haha

[22-Apr-2010 14:21:48] <rmatte> bad joke :P

[22-Apr-2010 14:22:20] <tuxknowledge> haha

[22-Apr-2010 14:22:29] <tuxknowledge> Thats about it

[22-Apr-2010 14:40:04] <gaieges> ok i lied im still having trouble removing the http status monito

[22-Apr-2010 14:40:06] <gaieges> monitor

[22-Apr-2010 14:40:16] <gaieges> anyone ever had to do that ?

[22-Apr-2010 14:45:24] <bigegor> Hi All, If somebody using my Distributed Collectors ZenPack? I need feedback.

[22-Apr-2010 14:46:54] <tuxknowledge> Im off for the day. Thanks for the help/info

[22-Apr-2010 14:47:04] <tuxknowledge> I am sure I will  be back soon

[22-Apr-2010 14:47:23] <gaieges> any thoughts on how to disable a http monitor ?

[22-Apr-2010 14:48:31] <rmatte> gaieges: are you referring to http under ip services?

[22-Apr-2010 14:48:37] <rmatte> or the httmon ZenPack?

[22-Apr-2010 14:48:47] <gaieges> http under ip servers rmatte

[22-Apr-2010 14:49:10] <rmatte> just set monitor to disabled on it or remove it

[22-Apr-2010 14:49:15] <gaieges> i did

[22-Apr-2010 14:49:16] <rmatte> then lock your ip services

[22-Apr-2010 14:49:20] <gaieges> oh lock em too ?

[22-Apr-2010 14:49:28] <rmatte> so that it doesn't add it back when it remodels

[22-Apr-2010 14:49:32] <gaieges> ahhhh

[22-Apr-2010 14:49:40] <rmatte> then go to Manage -> Push changes

[22-Apr-2010 14:49:48] <rmatte> so that the collector is aware of the change

[22-Apr-2010 14:49:59] <rmatte> otherwise it'll keep monitoring it until it updates which takes up to an hour

[22-Apr-2010 14:50:02] <gaieges> hmmm

[22-Apr-2010 14:50:05] <gaieges> interesting

[22-Apr-2010 14:50:06] <gaieges> thanks!

[22-Apr-2010 14:50:07] <rmatte> then you should be good to go

[22-Apr-2010 14:50:08] <rmatte> np

[22-Apr-2010 15:28:03] <gaieges> rmatte

[22-Apr-2010 15:28:14] <gaieges> i'm still having issues with the notifications

[22-Apr-2010 15:28:26] <gaieges> they're still going out even though i've disabled pretty much everything in that section

[22-Apr-2010 15:28:34] <gaieges> turned off monitoring

[22-Apr-2010 15:28:39] <gaieges> removed the ipservices

[22-Apr-2010 15:28:52] <gaieges> and they're still coming through

[22-Apr-2010 15:30:19] <rmatte> hmmm

[22-Apr-2010 15:30:34] <rmatte> try restarting zenoss

[22-Apr-2010 15:30:37] <gaieges> ya i am

[22-Apr-2010 15:30:45] <rmatte> and if that doesn't work then as the zenoss user type zendmd

[22-Apr-2010 15:30:50] <rmatte> and in the zendmd console do:

[22-Apr-2010 15:30:52] <rmatte> reindex()

[22-Apr-2010 15:30:54] <rmatte> commit()

[22-Apr-2010 15:30:57] <rmatte> one of those should fix it

[22-Apr-2010 15:31:02] <rmatte> I have to leave for the day, ttyl

[22-Apr-2010 15:39:30] <gaieges> thank you sir

[22-Apr-2010 15:39:32] <gaieges> i'll give that a try

[22-Apr-2010 16:01:52] <gaieges> can anyone help with my zenoss issue?

[22-Apr-2010 16:02:06] <gaieges> my http monitor was turned off, and the checkbox is gon

[22-Apr-2010 16:02:10] <gaieges> but still sends messages

[22-Apr-2010 16:23:36] <mrayzenoss> Anyone have any good images for docs/DOC-5334

[22-Apr-2010 16:42:16] <rhettardo> i think i may be able to contribute

[22-Apr-2010 17:26:25] <gaieges> someone hellppp meee

[22-Apr-2010 17:26:38] <gaieges> my http component won't stop sending alerts

[22-Apr-2010 17:26:44] <gaieges> and i just want to disable it for a given device

[22-Apr-2010 17:30:44] <rhettardo> i've created a multipgraph report but i only want to show data for the past hour.   i have added --start=-3600 and --end=-60 to the custom graph definition but those do not seem to be sticking

[22-Apr-2010 17:30:50] <rhettardo> where have i gone wrong?

[22-Apr-2010 17:33:12] <rhettardo> when i refresh the graph page it shows the proper values for maybe 1/2 second then show the past day and a half

[22-Apr-2010 17:37:05] <rhettardo> if i change the start and end points in the url i get a TypeError, Value: int() argument must be a string or a number

[22-Apr-2010 18:33:25] forsberg is now known as fOrsberg

[22-Apr-2010 23:38:49] <dibbz> so my event history is 14 million rows

[22-Apr-2010 23:39:52] <dibbz> i will just drop it, however i thought id check there is no ramifications

[23-Apr-2010 00:00:30] [disconnected at Fri Apr 23 00:00:30 2010]

[23-Apr-2010 00:00:30] [connected at Fri Apr 23 00:00:30 2010]

[23-Apr-2010 00:00:47] [zenoss-logger (logger bot) has joined #zenoss]

[23-Apr-2010 01:30:10] fOrsberg is now known as forsberg

[23-Apr-2010 02:57:20] <fragfutter> on my foundry switches all ethernet interface are displayed with the same mac address. any ideas?

[23-Apr-2010 02:59:43] <fragfutter> oh. hp opennnm does the same. weird.

[23-Apr-2010 03:02:08] <rhettardo> at least your switches dotn show speed randomly in the petabit/s range

[23-Apr-2010 03:02:14] <rhettardo> *speeds

[23-Apr-2010 03:02:32] <fragfutter> no only six gigabits ;)

[23-Apr-2010 03:02:48] <rhettardo> im only working with gigabit ports...

[23-Apr-2010 03:03:09] <rhettardo> lacp and gigabit ports in some cases but mostly gigabit fd!

[23-Apr-2010 04:08:00] <frozty_sa> eh, that's a nice bug

[23-Apr-2010 04:11:34] <rhettardo> surely

[23-Apr-2010 04:28:20] <surgpm> Hi. The CPU, free memory and paging graphs don't work for /Server/Windows in Zenoss Core 2.5.2. Is this a known issue?

[23-Apr-2010 04:34:01] <rhettardo> unsure.  i dont even monitor my windows box

[23-Apr-2010 04:41:51] <surgpm> Is there a way to disable graphs in a template?

[23-Apr-2010 04:46:06] <rhettardo> you can greate a template per class and remove the graphs as needed

[23-Apr-2010 04:46:15] <rhettardo> or greate a local copy per device and edit it as well

[23-Apr-2010 04:46:18] <rhettardo> create

[23-Apr-2010 04:46:26] <surgpm> Ah okay.

[23-Apr-2010 04:47:21] <rhettardo> which ever is more logical for your situation.  just keep in mind you may need to copy the template to the appropriate class

[23-Apr-2010 04:47:39] <rhettardo> and then double check your template bindings for that class

[23-Apr-2010 04:48:23] <surgpm> Thanks. Will have to read and play - new to Zenoss. Looks a lot easier to manage than Nagios though.

[23-Apr-2010 04:48:33] <rhettardo> yes

[23-Apr-2010 04:48:53] <rhettardo> most of what i just said will be accomplished by using the submenu's when in a class

[23-Apr-2010 04:49:32] <rhettardo> when trying to do per device templates just go to the submenu once you've navigated to the device and find the templates option

[23-Apr-2010 04:50:29] <rhettardo> its all fairly intuitive if you're familiar w/ this sort of thing

[23-Apr-2010 04:52:07] <rhettardo> im out.  good luck.

[23-Apr-2010 04:53:29] <surgpm> Fixed. Thanks!

[23-Apr-2010 04:56:38] <penthium> hello

[23-Apr-2010 04:57:13] <penthium> how can I configure zenoss to execute a command when i've a alert ?

[23-Apr-2010 05:03:04] <penthium> Ok, problem solved, it's in the event manager.

[23-Apr-2010 07:18:31] <fragfutter> i can't delete an event mapping

[23-Apr-2010 07:18:38] <fragfutter> it raises a zope error

[23-Apr-2010 08:27:50] forsberg is now known as fOrsberg

[23-Apr-2010 08:38:22] <mistich> good day

[23-Apr-2010 08:38:55] <mistich> to create a new organizer in zenddmd I just do dmd.Devices.Network.Router.createOrganizer('test')

[23-Apr-2010 08:49:29] <blull_> morning

[23-Apr-2010 08:49:34] <blull_> I got a couple of questions if someone has a second

[23-Apr-2010 08:49:39] <blull_> more deployment/management

[23-Apr-2010 08:49:43] <blull_> than anything thats an issue.

[23-Apr-2010 08:50:17] <blull_> First question is, right now, all of our alerts come from our production, internal IP space which can not be accessed remotely.  We do not allow vpn access into our corp environment (or prod for that) from our mobile devices that receive sms and email notifications.

[23-Apr-2010 08:51:08] <blull_> I was wondering, is there any way to ack and/or delete messages via remote methods such as e-mail, or via a DMZ situated service that allows zenoss to get the alerts either via a push or pull?

[23-Apr-2010 08:51:28] <blull_> Cause right now, we're having to manually ack all alerts via VPN on pc's, which becomes a pain when everyone is out at once :)

[23-Apr-2010 08:52:45] <blull_> my second, more important question is.

[23-Apr-2010 08:53:07] <blull_> right now all my remote collectors go to a central location of course.

[23-Apr-2010 08:53:29] <blull_> My question is, are SMS and E-mail alerts always sent from the zenhub?  or are they sent from the remote collectors?

[23-Apr-2010 08:53:41] <blull_> I ask because if we have a power outage at our zenhub location, we will not receive alerts obviously.

[23-Apr-2010 08:54:02] <blull_> I need a backup notification method, pref via sms from the actual collector or a secondary site.

[23-Apr-2010 08:54:16] <blull_> how is that done in zenoss?

[23-Apr-2010 09:24:54] <blull_> so

[23-Apr-2010 09:25:07] <blull_> how do i generate an email from a remote collector if it can't connect back to the zenhub for example?

[23-Apr-2010 09:30:11] <rmatte> blull_: it has to be able to connect back

[23-Apr-2010 09:30:17] <rmatte> there's no way around that

[23-Apr-2010 09:30:32] <rmatte> the only other way would be to have a completely separate instance of Zenoss rather than a remote collector

[23-Apr-2010 09:30:54] <blull_> isn't that kinda bad?

[23-Apr-2010 09:31:05] <blull_> what if you for example lost power at your zenhub site?

[23-Apr-2010 09:31:12] <blull_> no notifications go out at all

[23-Apr-2010 09:31:48] <blull_> if thats not possible at all

[23-Apr-2010 09:31:51] <blull_> i see 2 solutions

[23-Apr-2010 09:32:23] <blull_> 1 write a script that runs via cron on each collector and monitors devices across links to ensure remote site and ALL functionality at it required for notification are working... i see this as risky and a hack.

[23-Apr-2010 09:32:36] <blull_> other one would be to write a "zen hub mirror" type of daemon

[23-Apr-2010 09:32:44] <rmatte> yeh, but the point is that the master server is a complete zenoss install, the collectors are not

[23-Apr-2010 09:32:54] <rmatte> they are just the monitoring daemons collecting back to the master

[23-Apr-2010 09:32:56] <blull_> that takes zen data, and distributes it to two 2 seperate installations of zenoss @ 2 seperate sites.

[23-Apr-2010 09:33:03] <blull_> right..

[23-Apr-2010 09:33:04] <blull_> but

[23-Apr-2010 09:33:10] <rmatte> collectors have no databases, nothing beyond basic monitoring

[23-Apr-2010 09:33:15] <blull_> if you don't have a secondary notification method.

[23-Apr-2010 09:33:22] <blull_> whats the point in the monitoring?

[23-Apr-2010 09:33:24] <blull_> its not reliable.

[23-Apr-2010 09:33:35] <blull_> its a single point of failure

[23-Apr-2010 09:33:40] <rmatte> the idea of having a master monitoring server go down is bad to begin with

[23-Apr-2010 09:33:45] <blull_> uhh yeah

[23-Apr-2010 09:33:46] <blull_> but

[23-Apr-2010 09:33:50] <blull_> you can't control disaster

[23-Apr-2010 09:33:51] <rmatte> the server should obviously have battery backup

[23-Apr-2010 09:33:55] <blull_> you can mitigate risk

[23-Apr-2010 09:34:00] <rmatte> and if it's that critical, generator

[23-Apr-2010 09:34:03] <blull_> right

[23-Apr-2010 09:34:06] <blull_> you've never had a DC go down?

[23-Apr-2010 09:34:08] <blull_> an AC go out?

[23-Apr-2010 09:34:15] <blull_> or a generator freeze over during winter?

[23-Apr-2010 09:34:16] <blull_> it happens.

[23-Apr-2010 09:34:31] <rmatte> We host all of our stuff at the best datacenter in this city, and they literally have zero downtime

[23-Apr-2010 09:34:34] <rmatte> so no, we haven't

[23-Apr-2010 09:34:46] <rmatte> the only thing that can happen is hardware failure such as bad RAM, but the chances are very slim

[23-Apr-2010 09:34:55] <rmatte> and we have a separate server monitoring of main Zenoss servers

[23-Apr-2010 09:35:01] <rmatte> s/of/our

[23-Apr-2010 09:35:02] <blull_> okay

[23-Apr-2010 09:35:03] <blull_> there we go.

[23-Apr-2010 09:35:08] <blull_> that answers it.

[23-Apr-2010 09:35:16] <blull_> setup a secondary zenoss installation to monitor the first.

[23-Apr-2010 09:35:20] <blull_> at a secondary site.

[23-Apr-2010 09:35:21] <rmatte> yes

[23-Apr-2010 09:35:26] <rmatte> and have the first monitor the second

[23-Apr-2010 09:35:32] <rmatte> that way you know if either of them goes down

[23-Apr-2010 09:35:39] <blull_> but i can't just monitor the zenoss system as well

[23-Apr-2010 09:35:44] <rmatte> that's really the only way to do it

[23-Apr-2010 09:35:44] <blull_> have to monitor all supporting infrastructure.

[23-Apr-2010 09:35:55] <blull_> for delivery of notifications

[23-Apr-2010 09:35:56] <rmatte> yeh, so do that

[23-Apr-2010 09:36:02] <blull_> yeah I'm going to.

[23-Apr-2010 09:36:04] <rmatte> hehe

[23-Apr-2010 09:37:04] <blull_> just fyi though, if you work for the zenoss guys, you really may want to point out that a fortune 5000 (not 500) which loses 1,000,000 a minute (yes, thats realistic) -- they will absolutely require a reliable notification method as a core component of zenoss.

[23-Apr-2010 09:37:53] <blull_> just my 2 cents based on 12 years of working in lead positions in multi country millions a minute losses environments :)

[23-Apr-2010 09:45:15] fOrsberg is now known as forsberg

[23-Apr-2010 09:50:57] <rmatte> hehe

[23-Apr-2010 09:51:04] <rmatte> I don't work for Zenoss

[23-Apr-2010 09:51:44] <rmatte> the notification method is reliable enough provided it's setup correctly

[23-Apr-2010 09:52:05] <rmatte> the enterprise version has a lot more redundancy built in to the collector model as well

[23-Apr-2010 09:52:22] <rmatte> you're working with the open source version, so you can't really complain that it's not suitable for multi-million dollar companies hehe

[23-Apr-2010 10:01:17] * rmatte wonders if there's a feature request open yet to allow direct copying/pasting from the new event console

[23-Apr-2010 10:01:38] <rmatte> our noc guys are complaining about that, god forbid they have to do an export :P

[23-Apr-2010 10:02:12] <rmatte> anyways, time to troubleshoot this UnicodeDecodeError some more

[23-Apr-2010 10:32:34] <tripitaka> blull: how about you have a system at a remote site generate traps to the zenoss server each hour, and receive the emailed alerts. If that remote system generates a trap and does not recieve the email alert within a few minutes, it generates an alert telling you that the zenoss service is not working

[23-Apr-2010 10:43:14] <rmatte> basically heartbeats

[23-Apr-2010 10:43:34] <rmatte> decent idea

[23-Apr-2010 10:43:54] <rmatte> then again, he really should just monitor all of the zenoss processes via snmp

[23-Apr-2010 10:44:03] <rmatte> it's possible for zenperfsnmp to crash but for zentrap to keep working

[23-Apr-2010 10:44:08] <rmatte> so that's not the most effective means

[23-Apr-2010 11:16:58] <rmatte> I love how one of my zenoss boxes insists on continuing to monitor/alert on interfaces that are maked as monitored

[23-Apr-2010 11:17:04] <rmatte> even after I've pushed changes several times

[23-Apr-2010 11:17:36] * rmatte is starting to think that this 2.5.2 upgrade is going to turn in to a disaster

[23-Apr-2010 11:18:02] <rmatte> it certainly filled my schedule up, because I can't do anything else until I fix all these issues

[23-Apr-2010 11:23:48] <rmatte> that are not marked as monitored rather

[23-Apr-2010 11:25:48] <rmatte> shit, now zope won't start on one of my server

[23-Apr-2010 11:25:52] * rmatte pulls hair out

[23-Apr-2010 11:33:22] <rmatte> well that's great, this production box is hosed now and there are no devs around :(

[23-Apr-2010 11:36:34] <rmatte> venturaville: you wouldn't happen to know the location of the zope log file?

[23-Apr-2010 11:36:46] <rmatte> it doesn't appear to be in $ZENHOME/log and I've never had a need for it until now

[23-Apr-2010 11:37:27] <rmatte> zope suddenly won't start on one of the servers that I upgraded to 2.5.2 (which is not encouraging my faith in this new version at all)

[23-Apr-2010 11:37:57] <rmatte> everything else starts except zope, which makes the server basically useless

[23-Apr-2010 11:39:47] <rmatte> this probably has something to do with it: http://fpaste.org/cDRI/

[23-Apr-2010 11:40:15] <rmatte> ah, it's event.log

[23-Apr-2010 11:40:40] <rmatte> all I did was a reindex and now it won't start

[23-Apr-2010 11:40:43] <venturaville> yeah that one always confuses me too

[23-Apr-2010 11:40:56] <rmatte> I'm seeing lots of database conflict errors

[23-Apr-2010 11:41:40] <venturaville> the zope database and interactions with it seem the most fragile part of zenoss unfortunately

[23-Apr-2010 11:42:16] <rmatte> there aren't even errors showing up when i try to start it

[23-Apr-2010 11:42:23] <rmatte> it's just completely dead, no sign of what's wrong with it

[23-Apr-2010 11:42:33] <rmatte> I'm going to get crucified if I can't fix this

[23-Apr-2010 11:42:33] <venturaville> strace...?

[23-Apr-2010 11:42:45] <rmatte> yeh, I'll try that

[23-Apr-2010 11:44:21] <rmatte> last few lines: http://fpaste.org/CrUR/

[23-Apr-2010 11:44:29] <rmatte> the problem is that it's spawning the daemon off

[23-Apr-2010 11:44:34] <rmatte> which isn't getting straced

[23-Apr-2010 11:44:43] <rmatte> and the daemon is what's crashing it would seem

[23-Apr-2010 11:44:57] <venturaville> strace -f

[23-Apr-2010 11:45:00] <rmatte> k

[23-Apr-2010 11:45:17] <rmatte> running

[23-Apr-2010 11:45:22] <rmatte> 2.5.2 is a nightmare so far

[23-Apr-2010 11:45:28] <rmatte> I'm having nothing but problems after the upgrade

[23-Apr-2010 11:45:41] <rmatte> it seemed fine, but I keep running in to problem after problem

[23-Apr-2010 11:45:52] <rmatte> this is the worst I've seen

[23-Apr-2010 11:48:10] <rmatte> I can live with the odd error and what not, but zope completely rolling over and dying is not acceptable

[23-Apr-2010 11:48:24] * rmatte waits for the strace to finish

[23-Apr-2010 11:49:42] <rmatte> I hate it when it's my butt on the line and stupid things like this happen

[23-Apr-2010 11:49:55] <venturaville> don't we all

[23-Apr-2010 11:50:43] <rmatte> this strace is taking forever lol

[23-Apr-2010 11:53:56] <davetoo> rmatte: really?

[23-Apr-2010 11:54:06] <davetoo> how big is your installation?

[23-Apr-2010 11:54:18] <rmatte> davetoo: this particular server isn't big at all

[23-Apr-2010 11:54:26] <rmatte> like under 35 devices

[23-Apr-2010 11:54:52] <davetoo> wow

[23-Apr-2010 11:54:53] <rmatte> but it's a prod server, I need to fix it

[23-Apr-2010 11:54:53] <davetoo> well...

[23-Apr-2010 11:55:18] <davetoo> I upgraded my Enterprise system to 2.5.2 a couple of weeks ago and it's running fine

[23-Apr-2010 11:55:35] <rmatte> davetoo: I have 13 zenoss servers, some are running fine after the upgrade, some are not

[23-Apr-2010 11:55:39] <davetoo> interesting

[23-Apr-2010 11:55:46] <rmatte> I've done reindexes on others and not run in to any problems

[23-Apr-2010 11:55:51] <rmatte> but on this server a reindex killed it

[23-Apr-2010 11:55:53] <davetoo> back to the "good old days" of instability

[23-Apr-2010 11:55:57] <rmatte> yup

[23-Apr-2010 11:56:01] <rmatte> it's quite lovely

[23-Apr-2010 11:56:12] <rmatte> I should have just stuck with 2.4.5, it was actually working

[23-Apr-2010 11:56:25] <davetoo> We needed the improved event console performance

[23-Apr-2010 11:56:26] <rmatte> they try to change too much stuff between releases and it ends up messing things up

[23-Apr-2010 11:56:40] <davetoo> ours was.. too many events, it would never even load the page

[23-Apr-2010 11:56:41] <rmatte> We barely even use the event console lol

[23-Apr-2010 11:57:23] <rmatte> I've also got this problem: http://dev.zenoss.org/trac/ticket/6568

[23-Apr-2010 11:57:27] <rmatte> on a couple of my servers

[23-Apr-2010 11:57:27] <davetoo> I need to learn how to profile zope

[23-Apr-2010 11:57:35] <rmatte> It's a never-ending problem after this upgrade

[23-Apr-2010 11:57:48] <davetoo> well, we put our system together and then never got Ops to agree to use/maintain it,

[23-Apr-2010 11:58:01] <davetoo> so when I decided to drop the gigantic history database for the upgrade,

[23-Apr-2010 11:58:24] <davetoo> it took 20 minutes just to *drop* the friggin' entire database

[23-Apr-2010 11:58:31] <rmatte> nice

[23-Apr-2010 11:58:44] <davetoo> which I decided to do after giving up on truncating the table, after an hour.

[23-Apr-2010 11:58:48] <rmatte> well, looks like I'll be working late tonight

[23-Apr-2010 11:58:52] <rmatte> woohoo

[23-Apr-2010 11:59:11] <rmatte> I'm tempted to drop databases to resolve the utf-8 error, but I can't

[23-Apr-2010 11:59:35] <rmatte> brb

[23-Apr-2010 11:59:39] <davetoo> I tell people here not to expect IE to work correctly with zenoss

[23-Apr-2010 12:01:01] <rmatte> lol

[23-Apr-2010 12:01:09] <rmatte> well, either way, it's listed as a supported browser

[23-Apr-2010 12:01:14] <rmatte> so that's a really dumb problem to be having

[23-Apr-2010 12:01:26] <rmatte> lots of people around here use IE (unfortunately)

[23-Apr-2010 12:14:03] <rmatte> now zope is stuck and I can't kill it

[23-Apr-2010 12:14:10] <rmatte> even a kill -9 doesn't work

[23-Apr-2010 12:15:06] <venturaville> ouch

[23-Apr-2010 12:15:25] <rmatte> I think I'm just going to reboot the box then fry an fsrecover as egor suggested

[23-Apr-2010 12:15:30] <rmatte> looks promising

[23-Apr-2010 12:16:14] <venturaville> not a bad idea

[23-Apr-2010 12:16:28] <venturaville> do you use LVM?

[23-Apr-2010 12:16:46] <rmatte> LVM?

[23-Apr-2010 12:16:59] <rmatte> no

[23-Apr-2010 12:17:26] <rmatte> haha, the reboot is stuck at shutting down zope

[23-Apr-2010 12:17:30] * rmatte cries

[23-Apr-2010 12:17:44] <rmatte> this is really something

[23-Apr-2010 12:17:48] <venturaville> if you still have an ssh console to it, you can force reboot it

[23-Apr-2010 12:18:27] <rmatte> I do

[23-Apr-2010 12:18:49] <rmatte> how do you force it?

[23-Apr-2010 12:19:04] <rmatte> I don't see any option for shutdown

[23-Apr-2010 12:19:26] <venturaville> echo 1 > /proc/sys/kernel/sysrq

[23-Apr-2010 12:19:29] <cgibbons> that sort of behavior is always 'fun'

[23-Apr-2010 12:19:31] <venturaville> echo b > /proc/sysrq-trigger

[23-Apr-2010 12:20:00] <rmatte> cgibbons: I'm hoping to fix this within the next couple of hours, otherwise it's going to be a long night

[23-Apr-2010 12:20:42] <rmatte> venturaville: you are the man

[23-Apr-2010 12:21:24] <venturaville> :-)

[23-Apr-2010 12:21:37] <venturaville> LVM snapshots are a good way to make a snapshot of your setup before you upgrade

[23-Apr-2010 12:21:40] <venturaville> nice and quick to restore too

[23-Apr-2010 12:22:25] <venturaville> course tarballs work too

[23-Apr-2010 12:22:31] <mistich> good day

[23-Apr-2010 12:22:42] <rmatte> I had snapshots, I figured everything was ok so I removed them

[23-Apr-2010 12:22:52] <rmatte> since they cause performance issues

[23-Apr-2010 12:23:02] <mistich> have a problem running the zenobjectcopier.py and doesn't seem to be importing

[23-Apr-2010 12:23:05] <rmatte> I didn't expect to come across this many problems

[23-Apr-2010 12:23:22] <mistich> $ zenobjectcopier.py --import --file=./tmpTemplates/@zport@dmd@Devices@Web@SugarCRM@rrdTemplates@SugarCRM.xml

[23-Apr-2010 12:23:25] <mistich> Importing objects from ./tmpTemplates/@zport@dmd@Devices@Web@SugarCRM@rrdTemplates@SugarCRM.xml.

[23-Apr-2010 12:23:25] <mistich> 2010-04-23 17:21:46,193 INFO zen.ImportRM: Calling reIndex /zport/dmd/Devices/Web/SugarCRM

[23-Apr-2010 12:24:00] <mistich> no errors but not showing up in the templates

[23-Apr-2010 12:24:17] <rmatte> haha

[23-Apr-2010 12:24:24] <rmatte> so after the reboot, zenoss is working again

[23-Apr-2010 12:24:26] <rmatte> go figure

[23-Apr-2010 12:24:52] <rmatte> well that saves me time

[23-Apr-2010 12:25:28] <mistich> yeah but if you reboot again what happends

[23-Apr-2010 12:25:55] <rmatte> well, wouldn't be any different than doing zenoss stop/start

[23-Apr-2010 12:25:59] <rmatte> which I plan to try in a moment

[23-Apr-2010 12:26:04] <rmatte> after I reindex again

[23-Apr-2010 12:26:16] <mistich> yes I now upgrading to 2.5.2

[23-Apr-2010 12:26:38] <mistich> lets try that again

[23-Apr-2010 12:26:50] <mistich> you upgrading to 2.5.2

[23-Apr-2010 12:26:51] <rmatte> fsrecover is good to know about though

[23-Apr-2010 12:27:01] <rmatte> I've already upgraded most of my servers to 2.5.2

[23-Apr-2010 12:27:22] <rmatte> my main problem is: http://dev.zenoss.org/trac/ticket/6568

[23-Apr-2010 12:27:42] <mistich> if you are be careful with the template_64 really messed us up yesteday

[23-Apr-2010 12:28:01] <mistich> just don't use ie :)

[23-Apr-2010 12:28:02] <rmatte> really?

[23-Apr-2010 12:28:04] <rmatte> how so?

[23-Apr-2010 12:28:13] <rmatte> that's what I was just having problems with

[23-Apr-2010 12:28:37] <rmatte> what kinds of problems?

[23-Apr-2010 12:28:47] <mistich> we had a  ethernetCsmacd_64 in the root

[23-Apr-2010 12:29:29] <mistich> then in say 2 classes we had 2 custom ethernetCsmacd one for switches and one for routers routers get more oids

[23-Apr-2010 12:29:59] <mistich> when the devices remodeled they all used ethernetCsmacd_64 in the root

[23-Apr-2010 12:30:06] <rmatte> reindexed and restarted zenoss, now it's working fine

[23-Apr-2010 12:30:07] <rmatte> weird

[23-Apr-2010 12:30:16] <rmatte> ok well, I'm happy now

[23-Apr-2010 12:30:22] <rmatte> thanks for the help folks

[23-Apr-2010 12:30:46] <mistich> so now if we wanted a 64 and a 32 bit template. it would double our ethernetCsmacd template count from 30 to 60

[23-Apr-2010 12:31:27] <mistich> so for every custom ethernetCsmacd you need a custom ethernetCsmacd_64

[23-Apr-2010 12:31:55] <mistich> just wanted to let you know

[23-Apr-2010 12:32:50] <mistich> anyone use zenobjectcopier.py

[23-Apr-2010 12:38:20] <mistich> anyone

[23-Apr-2010 12:39:27] <rmatte> mistich: oh, that's weird

[23-Apr-2010 12:39:39] <rmatte> mistich: it's not supposed to be doing that

[23-Apr-2010 12:39:48] <mistich> yeah 2.5.2

[23-Apr-2010 12:39:52] <rmatte> mistich: bring that to cluther's attention when you see him

[23-Apr-2010 12:40:05] <mistich> it just stops after INFO zen.ImportRM: Calling reIndex /zport/dmd/Devices/Web/SugarCRM

[23-Apr-2010 12:40:07] <rmatte> the whole new 64bit template handling was his brainchild

[23-Apr-2010 12:40:38] <mistich> oh on the 64 yeah I just removed it and back to normal

[23-Apr-2010 12:40:59] <mistich> still trying to figure out the  zenobjectcopier.py issue

[23-Apr-2010 12:55:47] <davetoo> I'd never heard of that before you mentioned it

[23-Apr-2010 13:02:49] <rmatte> great, now webmin stopped working on all of my boxes, this is broken server week I swear

[23-Apr-2010 13:03:01] <rmatte> iptables-restore doesn't work too, which is bizarre

[23-Apr-2010 13:12:22] <rmatte> ah, it because someone moved it to port 20000 apparently

[23-Apr-2010 13:12:23] <rmatte> grrrr

[23-Apr-2010 13:18:01] <cgibbons> i give in, it was me

[23-Apr-2010 13:24:01] <lthrasher> rmatte: you said earlier "you're working with the open source version, so you can't really complain that it's not suitable for multi-million dollar companies" -- curious how you figure that

[23-Apr-2010 13:24:53] <davetoo> cgibbons:  I tried to build trunk yesterday following most of your tag recipe... had some troubles :)

[23-Apr-2010 13:25:16] <cgibbons> yeah that recipe doesn't work for trunk really. lots of stuff changed there just recently.

[23-Apr-2010 13:25:37] <cgibbons> i haven't been doing much work in 3.0 myself so i haven't had to deal with builds there much

[23-Apr-2010 13:25:39] <lthrasher> rmatte: my employer should clear two million in revenues this year, but if we managed just our two largest customers' stuff (60k devices) with a silver zenoss enterprise subscription and finagled a 50% volume discount, it'd cost 3 million per year.

[23-Apr-2010 13:25:41] <davetoo> ah

[23-Apr-2010 13:26:11] <cgibbons> i general it's the same, but with some changes because of the long overdue upgrade to zope & pytho

[23-Apr-2010 13:26:43] <davetoo> aye

[23-Apr-2010 13:27:02] <davetoo> ..which I'm happy for

[23-Apr-2010 13:53:21] <davetoo> Who would be most likely to be able to capture a new-from-scratch trunk build in a typescript session?

[23-Apr-2010 13:53:50] <rmatte> lthrasher: they have MSP licensing

[23-Apr-2010 13:53:56] <rmatte> lthrasher: so volume wouldn't be a factor

[23-Apr-2010 13:54:07] <davetoo> I'm trying to turn what I got from you (cgibbons) into a shell script :)

[23-Apr-2010 13:54:40] <rmatte> lthrasher: my point was that complaining about software that you are getting for free not being as robust as software that you would pay for is pointles

[23-Apr-2010 13:54:57] <davetoo> heh

[23-Apr-2010 13:54:59] <rmatte> pointless*

[23-Apr-2010 13:55:14] <davetoo> the Napster Generation

[23-Apr-2010 13:55:17] <rmatte> and I find it hard to dispute that

[23-Apr-2010 13:58:06] <rmatte> that being said, Zenoss has generally good quality as far as open source products go, we don't run in to many problems with it

[23-Apr-2010 14:05:46] <davetoo> A year ago, I was scared of new versions :P)

[23-Apr-2010 14:06:04] <davetoo> "core" seemed really unstable to me

[23-Apr-2010 14:06:07] <davetoo> daemons crashing

[23-Apr-2010 14:06:13] <davetoo> well, 1.5 years

[23-Apr-2010 14:06:26] <davetoo> these days, I'm pretty impressed by the daemon stability

[23-Apr-2010 14:06:33] <mistich> note do not run zenobjectcopier.py --import it mess up your whole index

[23-Apr-2010 14:06:46] <mistich> note do not run zenobjectcopier.py --import it mess up your whole index in 2.5.2

[23-Apr-2010 14:26:13] <blull_> is there a way to ack zenoss alerts via email?

[23-Apr-2010 14:26:18] <blull_> by responding or something

[23-Apr-2010 14:51:29] <kobalt> I have been having a difficult time finding the zenoss in a active/passive cluster documentation that netdata did anyone know where it is

[23-Apr-2010 14:52:36] <davetoo> So is there even a draft of an admin guide for 3.x?

[23-Apr-2010 15:02:03] <rmatte> blull_: the events have a link to acknowledge them by default right in the email

[23-Apr-2010 15:02:11] <rmatte> the emails rather

[23-Apr-2010 15:07:01] <blull_> right

[23-Apr-2010 15:07:09] <blull_> but its behind firewalls, etc.

[23-Apr-2010 15:07:11] <blull_> on our prod network

[23-Apr-2010 15:07:15] <blull_> can't hit it from my cell phone :)

[23-Apr-2010 15:07:28] <blull_> trying to avoid having to login to a pc and vpn in just to ack a full disk alert lol

[23-Apr-2010 15:07:47] <davetoo> yeah a few of my friends have been surprised that you can't just "reply" to an alert sent to a blackberry, etc.

[23-Apr-2010 15:10:54] <kobalt> anyone know where the docs are? for clustering

[23-Apr-2010 15:13:31] <kobalt> woot found it

[23-Apr-2010 15:13:45] <rmatte> blull_: ah, yeh, there's no way without actualy connectivity to the box

[23-Apr-2010 15:14:02] <rmatte> actual*

[23-Apr-2010 15:14:09] <blull_> k

[23-Apr-2010 15:14:11] <blull_> well

[23-Apr-2010 15:14:41] <blull_> ill hack something together on the zenhub that connects into dmz and queries a file of events to ack that gets written to by a daemon that monitors incoming email to that dmz system :)

[23-Apr-2010 15:14:55] <rmatte> it's definitely doable

[23-Apr-2010 15:15:00] <blull_> yup yup :)

[23-Apr-2010 15:15:05] <dec3pti0n> is there a way to change a device name ?

[23-Apr-2010 15:15:12] <rmatte> yes

[23-Apr-2010 15:15:17] <rmatte> Manage -> Rename Device

[23-Apr-2010 15:15:27] <rmatte> from the device page

[23-Apr-2010 15:15:37] <dec3pti0n> :( sorry for asking .. should have seen that

[23-Apr-2010 15:15:59] <rmatte> hehe

[23-Apr-2010 15:17:44] <rmatte> Well, I'm finally done the other stuff that I was doing, now back to this fun: http://dev.zenoss.org/trac/ticket/6568

[23-Apr-2010 15:17:52] <lthrasher> rmatte: i'm getting RHEL as free software but red hat is happy to listen to me complain as long as i keep paying my RHN.

[23-Apr-2010 15:18:48] <rmatte> lthrasher: I didn't say that you can't complain

[23-Apr-2010 15:20:09] <rmatte> and I didn't say that Zenoss wouldn't listen

[23-Apr-2010 15:20:42] <rmatte> I was stating that enterprise has more enterprise level features

[23-Apr-2010 15:21:06] <rmatte> and the whole conversation was centered around redundency

[23-Apr-2010 15:22:22] <rmatte> The enterprise version has a more robust redundency when it comes to a collector setup

[23-Apr-2010 15:24:22] <rmatte> I used the words "can't complain" but what I ultimately meant is that the core version is designed with smaller shops in mind

[23-Apr-2010 15:24:32] <rmatte> so let's not turn this in to a flame war

[23-Apr-2010 15:24:34] <rmatte> ;)

[23-Apr-2010 15:27:07] <lthrasher> no flames here, i just like my freedom respected and the enterprise license is actually hostile toward that (anti-forking "code of conduct" clause)

[23-Apr-2010 15:28:01] <rmatte> hey, I'm an open source advocate too... but the core version of Zenoss is not designed for really large installs out of the box

[23-Apr-2010 15:28:16] <rmatte> I run multiple core boxes

[23-Apr-2010 15:28:18] <davetoo> Can ANYbody tell me how to drive 3.x? :)    How do I initialte modeling/discovery of a network or an individual device from the gui?

[23-Apr-2010 15:28:48] <rmatte> davetoo: hmmm

[23-Apr-2010 15:29:19] <bigegor> lthrasher: you are free to extend Zenoss Core functionality  ;)

[23-Apr-2010 15:29:34] <rmatte> yeh, bigegor does a great job extending core

[23-Apr-2010 15:30:30] <rmatte> lthrasher: it all seems to center around the bar at the bottom

[23-Apr-2010 15:30:35] <rmatte> oops

[23-Apr-2010 15:30:36] <davetoo> yea it's going to take a while to learn my way around this new interface (which I suspect is missing a bunch of widgets too)

[23-Apr-2010 15:30:38] <rmatte> that was meant for davetoo

[23-Apr-2010 15:30:48] <davetoo> right,

[23-Apr-2010 15:30:56] <rmatte> so if you go to infrastructure...

[23-Apr-2010 15:31:01] <HCoyote> rmatte:  I forget ... are you in austin?

[23-Apr-2010 15:31:02] <rmatte> does it give you the option to add a device?

[23-Apr-2010 15:31:10] <rmatte> HCoyote: nope, Ottawa Canada

[23-Apr-2010 15:31:13] <HCoyote> ah, ok.

[23-Apr-2010 15:31:25] <lthrasher> bigegor: not if i've had an enterprise subscription in the past 12 months

[23-Apr-2010 15:31:32] <davetoo> I just found the way to discover devices on a network

[23-Apr-2010 15:32:19] <rmatte> lthrasher: that keeps people from grabbing an enterprise license, digging in and seeing how all the enterprise-specific packs were coded, then writing their own and dropping enterprise

[23-Apr-2010 15:32:31] <rmatte> lthrasher: people would abuse it were that not in place

[23-Apr-2010 15:33:11] <lthrasher> rmatte: right, exercising the freedoms that open source software is supposed to guarantee

[23-Apr-2010 15:33:30] <rmatte> lthrasher: the enterprise ZenPacks are not considered open source

[23-Apr-2010 15:33:55] <rmatte> Zenoss core is open source, all of the other enterprise bells and whistles are not

[23-Apr-2010 15:34:14] <davetoo> The whole Commercial Open Source model is still a social and economic experiment as far as I can tell;

[23-Apr-2010 15:34:30] <davetoo> not sure the jury has even entered deliberation yet :)

[23-Apr-2010 15:34:33] <lthrasher> just seems to me like a case of doing it wrong

[23-Apr-2010 15:35:45] <rmatte> lthrasher: that's obviously based on personal view

[23-Apr-2010 15:36:51] <lthrasher> rmatte: as is yours, we're both entitled to 'em and i respect yours despite disagreeing.

[23-Apr-2010 15:37:55] <rmatte> I personally feel that nitpicking that a company that is releasing an open source version of their product should be more open source isn't overly productive.  The fact that they are offering such a good product for free at all is quite significant in my opinion.

[23-Apr-2010 15:38:25] <rmatte> If it's not appealing enough there are always other similar projects which are completely open source

[23-Apr-2010 15:38:33] <davetoo> One drawback is when they end up hiring all of the active developers from the community :)

[23-Apr-2010 15:38:41] <davetoo> And they get paid to work on the closed stuff,

[23-Apr-2010 15:38:49] <davetoo> there's nothing "wrong" with that,

[23-Apr-2010 15:38:59] <davetoo> just an unfortunate side-effect sometimes

[23-Apr-2010 15:39:02] <rmatte> davetoo: true, but that's just smart hiring

[23-Apr-2010 15:39:07] <lthrasher> rmatte: i agree it's good that a company creates any useful free software.  "open-core" just seems like a more honest way to market something with that kind of model.

[23-Apr-2010 15:39:25] <davetoo> right, it's actually a good sign for the future of the product

[23-Apr-2010 15:39:56] <rmatte> davetoo: yeh, and the fact that they are making lots of money from the enterprise version ensures that the core version is going to be well kept and properly maintained in the future

[23-Apr-2010 15:40:06] <rmatte> which actually gives me more confidence in the open source version

[23-Apr-2010 15:40:12] <davetoo> yep

[23-Apr-2010 15:40:23] <lthrasher> i just worry about the fate of the community if a snoracle should happen.

[23-Apr-2010 15:40:37] <rmatte> lthrasher: anything is possible

[23-Apr-2010 15:40:43] <davetoo> now if I could only convince the [unkind plural omitted] at work to renew the goddamned enterprise license

[23-Apr-2010 15:40:54] <rmatte> lthrasher: I'd suspect that if something like that did happen, a fork may occur

[23-Apr-2010 15:41:13] <davetoo> but I may not be on the team that wins that battle :(

[23-Apr-2010 15:41:20] <rmatte> davetoo: that sucks

[23-Apr-2010 15:41:30] <davetoo> indeed.  Bad resource allocation IMO.

[23-Apr-2010 15:42:00] <davetoo> So maybe I need to find some other big Zenoss customers near here :)

[23-Apr-2010 15:42:23] <davetoo> and go show them my chops

[23-Apr-2010 15:42:29] <rmatte> ah, I figured out how to add a single device davetoo

[23-Apr-2010 15:42:33] <davetoo> 'k

[23-Apr-2010 15:42:36] <rmatte> click on infrastructure

[23-Apr-2010 15:42:48] <davetoo> (I'm running a network discovery on my alpha right now)

[23-Apr-2010 15:42:50] <rmatte> then above the device list there's a button with a square and a + in the middle of it

[23-Apr-2010 15:42:54] <rmatte> click that

[23-Apr-2010 15:43:08] <davetoo> oh, I saw that...

[23-Apr-2010 15:43:08] <rmatte> and there's "Add single device" or "Add multiple devices"

[23-Apr-2010 15:43:14] <rmatte> the Add multiple sounds enticing

[23-Apr-2010 15:43:26] <davetoo> maybe that's the way to remodel an existing device

[23-Apr-2010 15:43:52] <rmatte> doesn't appear to be

[23-Apr-2010 15:43:59] <rmatte> you probably need to remodel from a device page itself

[23-Apr-2010 15:44:18] <rmatte> hmmm

[23-Apr-2010 15:44:24] <rmatte> yeh, remodelling is a good question

[23-Apr-2010 15:44:26] * lthrasher simmers down a bit thinking about her next tasty beverage

[23-Apr-2010 15:44:29] <rmatte> pushing changes is another thing

[23-Apr-2010 15:44:52] <rmatte> lthrasher: ah, didn't know you were a woman, don't get many around here

[23-Apr-2010 15:45:03] <rmatte> s/were/are :P

[23-Apr-2010 15:45:13] <rmatte> my brain is fried from today

[23-Apr-2010 15:45:32] <rmatte> I do not look forward to Monday

[23-Apr-2010 15:45:36] <lthrasher> rmatte: try showering more frequently ;)

[23-Apr-2010 15:45:47] <rmatte> bleh

[23-Apr-2010 15:45:49] <rmatte> lol

[23-Apr-2010 15:46:27] <rmatte> Jane Curry is the only other woman I've seen in here, perhaps we need to spruce the channel up a bit

[23-Apr-2010 15:47:49] <lthrasher> jane's a sharp one, you're lucky she hates java so much :)

[23-Apr-2010 15:47:59] <rmatte> I despise Java

[23-Apr-2010 15:49:02] <rmatte> I only learned python because of Zenoss, but I like it a lot

[23-Apr-2010 15:49:15] <davetoo> same here

[23-Apr-2010 15:49:19] <davetoo> I was going to try Ruby

[23-Apr-2010 15:49:25] <davetoo> I'm glad I didnt' :)

[23-Apr-2010 15:49:35] <lthrasher> i can't get past the whitespace thing.  just as i'm sure you can't get past the stacktraces.

[23-Apr-2010 15:49:41] <davetoo> I'm glad my boss-at-the-time found zenoss

[23-Apr-2010 15:49:56] <davetoo> lthrasher: are you a skater chick? :)

[23-Apr-2010 15:49:58] <davetoo> I..

[23-Apr-2010 15:50:11] <davetoo> the whitespace thing has caused me to explore beyond vi as an editor

[23-Apr-2010 15:50:16] <lthrasher> davetoo: you mean like roller derby?

[23-Apr-2010 15:50:19] <davetoo> no,

[23-Apr-2010 15:50:25] <davetoo> never heard of Thrasher Magazine?

[23-Apr-2010 15:50:30] <lthrasher> like avril lavigne ;)

[23-Apr-2010 15:50:38] <davetoo> Google that and then you'll get it,

[23-Apr-2010 15:50:42] <davetoo> well, that's her costume

[23-Apr-2010 15:51:08] <davetoo> I... both of my brothers skate and surf, but I never mastered either so I'm an outsider :)

[23-Apr-2010 15:52:02] * lthrasher is a thrasher in name only

[23-Apr-2010 15:52:29] <lthrasher> can't drive a skateboard, don't play hockey for atlanta, etc.

[23-Apr-2010 15:52:57] <rmatte> hehe

[23-Apr-2010 15:53:15] <rmatte> the whitespace thing isn't bad, I hate it when people don't properly indent code

[23-Apr-2010 15:53:23] <rmatte> so I have no problem with a language that forces you to

[23-Apr-2010 15:53:42] <rmatte> I've seen bash scripts that one of the engineers here wrote and I almost cried

[23-Apr-2010 15:53:48] <rmatte> 500+ lines, no indentatio

[23-Apr-2010 15:53:50] <rmatte> n

[23-Apr-2010 15:54:00] <lthrasher> me too... i also hate when people don't drive in their lane, but i don't think that removing the lane markers is the solution ;)

[23-Apr-2010 15:54:09] <rmatte> haha

[23-Apr-2010 15:54:55] <davetoo> Before I started using "smart", code-folding editors I really, really missed curly braces and

[23-Apr-2010 15:54:56] <davetoo> %

[23-Apr-2010 15:54:58] <davetoo> ' matching

[23-Apr-2010 15:55:31] <lthrasher> ah, a vi user - there's hope for you yet

[23-Apr-2010 15:55:57] <davetoo> yeah I've been using gvim for the past four years or so

[23-Apr-2010 15:55:59] <bigegor> davetoo: http://en.wikipedia.org/wiki/Why%27s_%28poignant%29_Guide_to_Ruby

[23-Apr-2010 15:56:27] <davetoo> I started reading that;

[23-Apr-2010 15:56:32] <davetoo> I may play with Ruby sometime yet,

[23-Apr-2010 15:56:39] <davetoo> but I *needed* python for zenoss,

[23-Apr-2010 15:57:00] <davetoo> and I'm so pleased with all the good stuff in the standard libraries

[23-Apr-2010 15:57:25] <bigegor> i like python too

[23-Apr-2010 15:58:09] <kobalt> anyone figure out how to zenoss to do a model on a device that is already added (like to manually update the current info) on 3.0A

[23-Apr-2010 15:58:15] <lthrasher> wonder if zenoss would run under jython ;)

[23-Apr-2010 15:58:34] <davetoo> kobalt: we werjust tryhing to figure that out 1/2 hour ago

[23-Apr-2010 15:58:44] <kobalt> hehe

[23-Apr-2010 15:59:02] <kobalt> I just got it up and running here so was playing with it

[23-Apr-2010 15:59:17] <rmatte> I am out of here

[23-Apr-2010 15:59:19] <davetoo> yeah I'm doing some network discoveries right now

[23-Apr-2010 15:59:23] <davetoo> have a good weekend rmatte

[23-Apr-2010 15:59:26] <rmatte> you too

[23-Apr-2010 15:59:32] <kobalt> Later rmatte

[23-Apr-2010 16:03:09] <rmatte> later

[23-Apr-2010 16:06:22] <cgibbons> can you just zenmodel from the command-line, kobalt?

[23-Apr-2010 16:07:14] <kobalt> let me try one sec

[23-Apr-2010 16:09:09] <kobalt> I turned on the interface descrips in the zproperties and its not showing up, I ran the zenmodeler and it did not add the ifdescr

[23-Apr-2010 16:10:28] <kobalt> its seems like it collected it but I dont see it in the components

[23-Apr-2010 16:11:13] <kobalt> using zenmodeler run -v 10 -d <device hostname> --collect='InterfaceMap' --force

[23-Apr-2010 19:25:22] forsberg is now known as fOrsberg

[24-Apr-2010 00:00:30] [disconnected at Sat Apr 24 00:00:30 2010]

[24-Apr-2010 00:00:30] [connected at Sat Apr 24 00:00:30 2010]

[24-Apr-2010 00:00:43] [zenoss-logger (logger bot) has joined #zenoss]

[24-Apr-2010 03:41:59] <stoked> anyone know of a good way to monitor mutiple windows processes that are consistntly restarted?

[24-Apr-2010 05:10:02] fOrsberg is now known as forsberg

[24-Apr-2010 13:13:07] forsberg is now known as fOrsberg

[24-Apr-2010 13:37:18] rhettardo_ is now known as rhettardo

[24-Apr-2010 17:00:47] <seen1976> can someone point me to information on event commands (ie. shell scripts we can fire when an event occurs or clears)

[24-Apr-2010 18:18:37] fOrsberg is now known as forsberg

[24-Apr-2010 19:05:19] <baffle> rmatte: So you're having lots of problems with 2.5.2? Actually, for me, it is much more stable than 2.5.0/2.5.1.

[24-Apr-2010 21:51:56] forsberg is now known as fOrsberg

[25-Apr-2010 00:00:30] [disconnected at Sun Apr 25 00:00:30 2010]

[25-Apr-2010 00:00:30] [connected at Sun Apr 25 00:00:30 2010]

[25-Apr-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[25-Apr-2010 05:07:39] fOrsberg is now known as forsberg

[25-Apr-2010 17:48:30] forsberg is now known as fOrsberg

[25-Apr-2010 22:16:39] <mrayzenoss> Thanks to rmatte for updating his Collector Tool ZenPack: docs/DOC-5831

[25-Apr-2010 22:32:06] <mrayzenoss> Thanks to malbon for updating his HPUX ZenPack: docs/DOC-3462

[25-Apr-2010 22:45:36] <rhettardo> you're here late

[25-Apr-2010 22:45:39] <rhettardo> its not even a work day

[25-Apr-2010 22:54:28] mrayzenoss1 is now known as mrayzenoss

[25-Apr-2010 22:54:42] <mrayzenoss> rhettardo: I'm at LinuxFest Northwest

[25-Apr-2010 22:54:52] <rhettardo> ah

[25-Apr-2010 22:55:12] <mrayzenoss> in my hotel room, finishing up a few things before going to bed, waking up way too early and flying home

[25-Apr-2010 22:55:20] <rhettardo> working tomorrow?

[25-Apr-2010 22:55:28] <mrayzenoss> only on the plane

[25-Apr-2010 22:55:37] <rhettardo> nice

[25-Apr-2010 22:55:44] <rhettardo> at least you have that going for you

[25-Apr-2010 22:56:09] <mrayzenoss> so figure 2 hours or so until my laptop battery dies, unless I'm lucky enough to have a power plug on my flight, like I did on the way up

[25-Apr-2010 22:56:11] <rhettardo> i added some shit to trac for alpha but i dont have a name so i just used zenoss. not sure if it really matter but throwing it out there

[25-Apr-2010 22:56:17] <rhettardo> errr stuff to trac

[25-Apr-2010 22:56:18] <mrayzenoss> thanks

[25-Apr-2010 22:56:33] <mrayzenoss> we should be doing an alpha release every week

[25-Apr-2010 22:56:36] <rhettardo> nice.  never had power on a flight

[25-Apr-2010 22:56:41] <rhettardo> cool

[25-Apr-2010 22:56:43] <mrayzenoss> yeah, it was my first time too

[25-Apr-2010 22:56:58] <rhettardo> well i've found plenty of issues but i've only posted the one on the forum and submitted the two to trac

[25-Apr-2010 22:56:59] <mrayzenoss> on a flight last year I noticed an AC jack, so I could at least charge my phone

[25-Apr-2010 22:57:19] mrayzenoss1 is now known as mrayzenoss

[25-Apr-2010 22:57:23] <mrayzenoss> hotel internet sucks

[25-Apr-2010 22:57:31] <rhettardo> i've really only flown mostly within california using southwest so its been very minimalistic

[25-Apr-2010 22:57:34] <rhettardo> yes they do

[25-Apr-2010 22:58:05] <rhettardo> over powered AP's with horrible density make for nice noisefloor

[25-Apr-2010 22:58:40] <mrayzenoss> this is ethernet, I figured hardwire would be more stable

[25-Apr-2010 22:58:50] <mrayzenoss> might have to go back to the jailbroken iPhone

[25-Apr-2010 22:59:05] <rhettardo> oh.  yeah i'd hope it would be

[25-Apr-2010 23:00:40] <mrayzenoss> anyway, 1 more ZenPack and I'm going to bed

[25-Apr-2010 23:00:53] <rhettardo> word

[25-Apr-2010 23:01:01] <rhettardo> well have a good flight

[25-Apr-2010 23:01:27] <mrayzenoss> thanks

[25-Apr-2010 23:01:33] <mrayzenoss> you're in CA?

[25-Apr-2010 23:01:49] <rhettardo> yeah

[25-Apr-2010 23:01:50] <rhettardo> Burbank

[25-Apr-2010 23:01:56] <mrayzenoss> did you go to SCALE?

[25-Apr-2010 23:02:01] <rhettardo> no :\

[25-Apr-2010 23:02:19] <rhettardo> we talked about going to a few different shows but never actually do

[25-Apr-2010 23:02:23] <mrayzenoss> well, we'll be back for Velocity/DevOps day probably

[25-Apr-2010 23:02:30] <mrayzenoss> in June

[25-Apr-2010 23:02:33] <rhettardo> cool

[25-Apr-2010 23:02:37] <rhettardo> where is it being held at

[25-Apr-2010 23:02:40] <rhettardo> if you know off hand

[25-Apr-2010 23:03:20] <mrayzenoss> Santa Clara I believe

[25-Apr-2010 23:03:32] <mrayzenoss> or Mountain View

[25-Apr-2010 23:03:46] <rhettardo> ah

[25-Apr-2010 23:19:07] <mrayzenoss> Last ZenPack of the night

[25-Apr-2010 23:19:24] <mrayzenoss> VMware ESX SNMP ZenPack: docs/DOC-5856  Thanks to ke4qqq and rmatte!

[25-Apr-2010 23:31:37] mrayzenoss1 is now known as mrayzenoss

[26-Apr-2010 00:00:30] [disconnected at Mon Apr 26 00:00:30 2010]

[26-Apr-2010 00:00:31] [connected at Mon Apr 26 00:00:31 2010]

[26-Apr-2010 00:00:46] [zenoss-logger (logger bot) has joined #zenoss]

[26-Apr-2010 01:21:45] fOrsberg is now known as forsberg

[26-Apr-2010 01:57:26] <baffle> rmatte: Do you use 1 master Zenoss + lots of collectors or a bunch of separate masters + some of them have extra collectors? Is the distribution to separate collectors due to physical location or load?

[26-Apr-2010 02:18:49] <forsberg> rmatte wont be out of bed for another 6-8 hours

[26-Apr-2010 02:18:50] <forsberg> ;>

[26-Apr-2010 05:27:51] <chemist> mornng all

[26-Apr-2010 05:28:27] <chemist> anyone know where I can find documentation regarding vpn tunnel monitoring with zenoss?

[26-Apr-2010 05:43:46] <biertie> is it possible to change the port of the web interface in centos?

[26-Apr-2010 05:54:06] <forsberg> yes biertie

[26-Apr-2010 05:54:15] <forsberg> and chemist what about the tunnel do you want to monitor?

[26-Apr-2010 05:54:21] <forsberg> besides the obvious if its up :)

[26-Apr-2010 05:55:49] <biertie> forsberg: well, I installed zenoss via the rpm, and I can start the service and blah blah

[26-Apr-2010 05:55:54] <biertie> but I don't find anymore -_-

[26-Apr-2010 05:56:24] <biertie> *any files

[26-Apr-2010 05:57:31] <forsberg> im searching on google, cant remember, it should be possible to find the info tho

[26-Apr-2010 05:57:44] <biertie> well, I can't find it :(

[26-Apr-2010 05:58:10] <forsberg> message/40513#40513

[26-Apr-2010 05:58:22] <forsberg> searched for "port 8080" on www.zenoss.org

[26-Apr-2010 05:58:24] <forsberg> first link

[26-Apr-2010 05:59:10] <biertie> $ZENHOME

[26-Apr-2010 05:59:14] <biertie> seems empty :)

[26-Apr-2010 05:59:44] <forsberg> /opt/zenoss

[26-Apr-2010 05:59:54] <forsberg> on centos

[26-Apr-2010 06:00:05] <forsberg> havent tried anything else

[26-Apr-2010 06:00:25] <biertie> aha, thx

[26-Apr-2010 06:00:36] <biertie> my locate must be borken :(

[26-Apr-2010 06:00:41] <forsberg> updatedb

[26-Apr-2010 06:00:42] <forsberg> :P

[26-Apr-2010 06:01:03] <forsberg> have to updatedb to reindex for recent files

[26-Apr-2010 06:01:19] <biertie> ach right

[26-Apr-2010 06:01:25] <biertie> that's why I'm still 'junior' I think

[26-Apr-2010 06:01:31] <biertie> because I forget these things way to fast :p

[26-Apr-2010 06:01:52] <forsberg> next time you might remember ;>

[26-Apr-2010 06:03:17] <forsberg> chemist, if you just want to check if its up, you can maybe just ping an ip in the other end, and process for vpn software

[26-Apr-2010 06:03:19] <forsberg> anyways afk

[26-Apr-2010 06:03:27] <biertie> bye, and thx again :)

[26-Apr-2010 06:12:03] <chemist> forsberg: that is what I do atm, I was wandering if there was a way of doing it with snmp

[26-Apr-2010 06:15:12] <chemist> hardware is Astaro security appliance at one end and Draytek 2820's at the other

[26-Apr-2010 06:16:37] <forsberg> well if the device has snmp support, it should be able to find the values to check stuff, but...

[26-Apr-2010 06:17:13] <forsberg> basically the tunnel will be up if you can ping, and it wont be up if you cant ping

[26-Apr-2010 06:17:40] <forsberg> but its mostly first active in here, in like 3-5+ hours

[26-Apr-2010 06:17:43] <forsberg> so try again there :)

[26-Apr-2010 06:22:59] * chemist wishes the bosses would pay for Cisco stuff

[26-Apr-2010 06:24:23] <forsberg> ;)

[26-Apr-2010 06:24:27] <forsberg> ciscock

[26-Apr-2010 06:29:49] <biertie> chemist: cisco isn't that good

[26-Apr-2010 06:31:13] <chemist> hp?

[26-Apr-2010 06:31:34] <chemist> astaro appliances certainly suck

[26-Apr-2010 06:48:45] <biertie> dnno, I'm happy with astaro :)

[26-Apr-2010 06:49:01] <biertie> but I would never buy a cisco switch

[26-Apr-2010 06:53:45] <biertie> god dammed

[26-Apr-2010 06:53:50] <biertie> web interface still doesn't work -_-

[26-Apr-2010 07:56:31] <fragfutter> biertie: it looks like the port is hardcoded in more then one place.

[26-Apr-2010 08:14:57] <biertie> fragfutter: yes, so it seems

[26-Apr-2010 08:15:09] <biertie> I reinstalled it, checked it with lynx localhost:8080

[26-Apr-2010 08:15:12] <biertie> and it worked

[26-Apr-2010 08:15:17] <biertie> changed to port in zope.conf

[26-Apr-2010 08:15:21] <biertie> and it's broken -_-

[26-Apr-2010 08:17:26] <biertie> and I am not amused :p

[26-Apr-2010 08:33:29] <fragfutter> biertie: its also in zenctl and i found it in some libraries.

[26-Apr-2010 08:33:43] <fragfutter> biertie: place an apache with mod_proxy in front of it.

[26-Apr-2010 08:34:12] <fragfutter> biertie: typical deployment of a zope service ;)

[26-Apr-2010 08:34:50] <biertie> but I already have a service on port 8080

[26-Apr-2010 08:34:53] <biertie> that's the main problem :(

[26-Apr-2010 08:34:57] <fragfutter> biertie: ah, ok.

[26-Apr-2010 08:36:48] <fragfutter> i want to classify and change the severity of a snmp trap received by zenoss. To get started i wrote a python script, that pulls the existing event from dmd. Then changes the summary and commits it. this doesn't change the presentation in the event console. So i must have a missunderstanding or a brain twister. Any hints?

[26-Apr-2010 08:42:18] <fragfutter> and how can i debug event transformation during event mapping?

[26-Apr-2010 09:04:17] <biertie> lol, ok, according to the manual I should get in a wizzard like thing

[26-Apr-2010 09:04:19] <biertie> didn't happen :')

[26-Apr-2010 09:16:24] <mrayzenoss> biertie: was this a new install or an upgrade?

[26-Apr-2010 09:16:46] <biertie> new

[26-Apr-2010 09:16:47] <biertie> :)

[26-Apr-2010 09:16:56] <mrayzenoss> hmmm… are you logged in?

[26-Apr-2010 09:17:42] <biertie> with what pass? :o

[26-Apr-2010 09:17:55] <biertie> ach

[26-Apr-2010 09:17:56] <biertie> now I see it

[26-Apr-2010 09:18:07] <biertie> lol, I was looking in lynx to it

[26-Apr-2010 09:18:11] <biertie> that's why I didn't work prolly :P

[26-Apr-2010 09:18:13] <mrayzenoss> LOL

[26-Apr-2010 09:18:29] <mrayzenoss> lynx + AJAX == fail

[26-Apr-2010 09:19:23] <chickenandbeans> links should work though :)

[26-Apr-2010 09:19:40] <chickenandbeans> well kinda

[26-Apr-2010 09:33:16] <mrayzenoss> New EmailPing ZenPack for monitoring your email-based alerting mechanism: docs/DOC-5853

[26-Apr-2010 09:37:25] <davidX-> nice

[26-Apr-2010 09:48:55] <rmatte> I have a ton of this in my zentrap logfile on one of my servers...

[26-Apr-2010 09:48:56] <rmatte> WARNING zen.ZenTrap: Queue exceeded maximum length

[26-Apr-2010 09:49:05] <rmatte> 2010-04-26 10:12:31,271 WARNING zen.ZenTrap: Queue exceeded maximum length: 5074/5000. Trimming

[26-Apr-2010 09:49:33] <rmatte> I think it's causing traps to get dropped since a drive failed on a server last night which we usually get traps for, yet we didn't get anything

[26-Apr-2010 09:51:55] <rmatte> When I click on edit config for zentrap I get /bin/sh: zentrap: no found

[26-Apr-2010 09:51:57] <rmatte> hmmm

[26-Apr-2010 09:59:26] <rmatte> I'm going to have to organize reboots for all of my Zenoss servers tomorrow night, the one that I did reboot is working fine, all of the others have weird problems

[26-Apr-2010 10:09:03] <ckrough> rmatte: I bumped my maxqueue up to 20000, shouldnt cause a problem if you have enough ram

[26-Apr-2010 10:09:50] <ckrough> rmatte: 'maxqueuelen 20000' in the config for the daemon thats hitting its limit

[26-Apr-2010 10:10:25] <ckrough> on first restart zenperf will go to ~28000 and drop 8000, but its fine after the first pass

[26-Apr-2010 10:21:18] mrayzenoss1 is now known as mrayzenoss

[26-Apr-2010 10:28:42] <rmatte> ckrough: thanks

[26-Apr-2010 10:31:28] <rmatte> weird

[26-Apr-2010 10:31:37] <rmatte> now that the maxqueue is increased: 2010-04-26 11:31:07,121 ERROR netsnmp: truncating unsigned value to 32 bits (2)

[26-Apr-2010 10:32:52] <rmatte> oh nevermind, that's some other box

[26-Apr-2010 10:39:40] <twm1010> hello everyone

[26-Apr-2010 10:42:14] <ckrough> hello

[26-Apr-2010 10:42:59] <ckrough> rmatte: zenoss warned me that it could use a lot of RAM if you go over 10000. I havent had any problems with that setting on 32GB dedicated collectors

[26-Apr-2010 10:56:25] <rmatte> I'm barely going over

[26-Apr-2010 10:56:32] <rmatte> I'm going over by like 50ish

[26-Apr-2010 10:56:39] <rmatte> and that's over 5000 I mean

[26-Apr-2010 10:56:59] <rmatte> I might just set it at 10000 for now

[26-Apr-2010 10:59:59] <ckrough> makes sense

[26-Apr-2010 11:25:57] <rmatte> Now I just need to figure out why I'm seeing: ERROR netsnmp: truncating unsigned value to 32 bits (2)

[26-Apr-2010 11:26:10] <rmatte> in the zentrap logfile on some of my servers

[26-Apr-2010 12:28:07] <jb> hrm, i'm having some problems with some graphs

[26-Apr-2010 12:28:16] <jb> i have some COUNTER datasources, which work fine typically

[26-Apr-2010 12:28:30] <jb> but the value that they are counting gets reset of the particular daemon that generates them is restarted

[26-Apr-2010 12:28:41] <jb> which obviously screwd up my graphs, and makes them have wacky negative values

[26-Apr-2010 12:28:43] <jb> any ideas?

[26-Apr-2010 13:02:10] <rmatte> jb: no idea

[26-Apr-2010 13:02:23] <rmatte> jb: I assumed it would use the previous value from the RRD

[26-Apr-2010 13:02:26] <rmatte> sounds like a bug

[26-Apr-2010 13:07:54] <aclark> woo! alpha installers

[26-Apr-2010 13:08:07] <aclark> based on Zope 2.12 / Python 2.6 I presume ;-)

[26-Apr-2010 13:09:56] <rmatte> probably

[26-Apr-2010 13:12:19] <aclark> definitely

[26-Apr-2010 13:13:17] <aclark> (based on what's in zenossinst/externallibs)

[26-Apr-2010 13:43:45] <rmatte> I'm still trying to figure out how to fix: http://dev.zenoss.org/trac/ticket/6568

[26-Apr-2010 13:45:16] <rmatte> it's something to do with events in the MySQL database

[26-Apr-2010 13:45:25] <rmatte> since it seems to error out when calculating the uptime

[26-Apr-2010 13:45:46] <rmatte> but I can't just wipe the events

[26-Apr-2010 13:46:11] <ckrough> jb try setting sane MIN/MAX values in the RRD, it should reject rollovers then

[26-Apr-2010 13:46:26] <rmatte> erm, when calculating availability rather

[26-Apr-2010 14:24:42] <rrittenhouse> I have ZenOSS (v 2.5.x) installed on a server. I havent accessed it in a while but now when I do the server starts to slow down and become unresponsive. Is there a way to export the data out of zenoss via SSH and import it into a new install?

[26-Apr-2010 14:26:52] <rrittenhouse> As soon as I boot it up it starts to consume all 2GB of ram

[26-Apr-2010 14:31:33] <rmatte> mhmmm, so they are using iso-8859-1 encoding when dumping events to MySQL but utf8 when reading them

[26-Apr-2010 14:31:42] <rmatte> that's lovely

[26-Apr-2010 14:57:33] <rmatte> should still work though, there's something funky with the way IE is handling it

[26-Apr-2010 15:08:09] <blull_> hey

[26-Apr-2010 15:08:10] <blull_> very quick

[26-Apr-2010 15:08:23] <blull_> what service do I need to run in order to scrape event logs locally on a window sserver?

[26-Apr-2010 15:08:34] <blull_> zeneventlog?

[26-Apr-2010 15:08:40] <blull_> zenwin?

[26-Apr-2010 15:08:44] <blull_> zenwinmodeler?

[26-Apr-2010 15:18:20] <rmatte> blull_: zeneventlog

[26-Apr-2010 15:18:29] <blull_> werd

[26-Apr-2010 15:18:38] <blull_> hmm

[26-Apr-2010 15:18:39] <blull_> i run it

[26-Apr-2010 15:18:41] <rmatte> what do you mean by scrape them locally though?

[26-Apr-2010 15:18:46] <blull_> via logins

[26-Apr-2010 15:18:47] <blull_> locally on the box

[26-Apr-2010 15:18:52] <blull_> not via remote logging

[26-Apr-2010 15:18:54] <blull_> wait

[26-Apr-2010 15:19:01] <blull_> can you stream from windows natively to zeneventlog?

[26-Apr-2010 15:19:10] <blull_> without installing 3rrd party utils on windows?

[26-Apr-2010 15:24:05] <rmatte> zeneventlog uses WMI I believe

[26-Apr-2010 15:24:28] <rmatte> Zenoss is agentless, so there isn't anything to install on the windows box other than configuring WMI

[26-Apr-2010 15:29:23] <ckrough> ahhh all upgrades to 2.5.2 done

[26-Apr-2010 15:29:25] * ckrough collapse

[26-Apr-2010 15:31:28] <rmatte> ckrough: my upgrades are an ongoing thing

[26-Apr-2010 15:31:37] <rmatte> this utf8 error is driving me nuts

[26-Apr-2010 15:31:58] <rmatte> I know there's some event(s) in the database that's causing it, but tracking down the specific events and fixing/deleting them is a whole other issue

[26-Apr-2010 15:32:06] <rmatte> and then there's no guarantee that it won't just happen again

[26-Apr-2010 15:32:20] <rmatte> it was never an issue with 2.4.5 though and it's the exact same events, so not sure what changed to cause this

[26-Apr-2010 16:11:34] <otakup0pe> ugh can someone tell me why the json returned by getEventSummary is 'invalid'

[26-Apr-2010 16:13:07] <otakup0pe> oh wait it uses single quotes, json requires double

[26-Apr-2010 16:13:10] <otakup0pe> funny that

[26-Apr-2010 16:13:55] <npmccallum> otakup0pe: I don't believe that it is JSON, I believe that it is python

[26-Apr-2010 16:18:31] <otakup0pe> oh it's kind of hard to tell ^_^

[26-Apr-2010 16:18:37] <otakup0pe> makes sense though. it's pickle or something ?

[26-Apr-2010 16:21:56] <npmccallum> no

[26-Apr-2010 16:22:23] <npmccallum> do this in python: >>>print [1, 2, "hi"]

[26-Apr-2010 16:22:31] <npmccallum> it should look the same

[26-Apr-2010 16:43:56] <rmatte> Hey Nate, any idea how I might go about fixing this: http://dev.zenoss.org/trac/ticket/6568 ?

[26-Apr-2010 16:44:24] <rmatte> Pretty sure it's some event in MySQL that's causing it, it only affects IE, and I had the same events in 2.4.5 with no issues

[26-Apr-2010 16:44:46] <rmatte> not sure what was changed to cause that, or how to go about fixing or deleting the specific events causing the issue

[26-Apr-2010 16:45:01] <rmatte> I also don't know whether or not I should expect this to happen again once I have actually resolved it

[26-Apr-2010 16:45:38] <rmatte> If you could even point me in the right directory of what files I should be looking at code in

[26-Apr-2010 16:45:52] <rmatte> I can't figure out which file actually calculates the availability when loading the status page

[26-Apr-2010 16:47:26] <npmccallum> so basically this is a python error

[26-Apr-2010 16:48:29] <npmccallum> rmatte: I would start by finding the template and removing each section of the code that prints python (ie comment out most of the page)

[26-Apr-2010 16:48:45] <npmccallum> then uncomment each section where python is used and try again

[26-Apr-2010 16:48:54] <rmatte> k

[26-Apr-2010 16:48:55] <npmccallum> this will help you find what section of the code is triggering the error

[26-Apr-2010 16:49:09] <npmccallum> then you'll have an idea where the bad string is coming from

[26-Apr-2010 16:49:29] <npmccallum> and you can check zdb/mysql

[26-Apr-2010 16:49:50] <npmccallum> my guess is that it is a modeler being feed bad data

[26-Apr-2010 16:49:56] <npmccallum> but it could be literally anything

[26-Apr-2010 16:50:08] <rmatte> it's strange how it works in firefox but not IE

[26-Apr-2010 16:50:48] <npmccallum> yeah, that is a bit wierd

[26-Apr-2010 16:51:10] <rmatte> but yeh, I'll have to make a clone of the VM and bring it up under another IP tomorrow so that I can mess with it, since it's a production system

[26-Apr-2010 16:51:13] <npmccallum> something like this typically happens when an invalid ascii string is generated

[26-Apr-2010 16:51:41] <npmccallum> for instance, I wrote a modeler and sometimes the devices would return "0xff0xff0xff..."

[26-Apr-2010 16:51:47] <npmccallum> that is of course not valid ascii

[26-Apr-2010 16:51:48] <rmatte> any idea why it's only appearing after the upgrade?

[26-Apr-2010 16:51:59] <npmccallum> no idea

[26-Apr-2010 16:52:09] <rmatte> right

[26-Apr-2010 16:52:23] <rmatte> k, I'll do more digging tomorrow

[26-Apr-2010 16:52:35] <npmccallum> alright, let me know if you have any other questions

[26-Apr-2010 16:52:40] <rmatte> thanks

[26-Apr-2010 16:52:40] <npmccallum> sorry I couldn't help more

[26-Apr-2010 16:52:56] <rmatte> all good, you can't make bricks without clay

[26-Apr-2010 16:53:14] <rmatte> at least I have some stuff to try tomorrow now

[26-Apr-2010 17:15:02] <rmatte> well, I'm out for the night, later folks

[26-Apr-2010 18:17:55] forsberg is now known as fOrsberg

[26-Apr-2010 20:54:36] * ckrough lurks

[27-Apr-2010 00:00:30] [disconnected at Tue Apr 27 00:00:30 2010]

[27-Apr-2010 00:00:30] [connected at Tue Apr 27 00:00:30 2010]

[27-Apr-2010 00:00:48] [zenoss-logger (logger bot) has joined #zenoss]

[27-Apr-2010 01:44:10] fOrsberg is now known as forsberg

[27-Apr-2010 06:59:55] <chemist> hello

[27-Apr-2010 07:00:17] <chemist> I am trying to set up synthetic trasactions

[27-Apr-2010 07:01:42] <chemist> when I run the "test against device" I get a "ImportError: No module named time"

[27-Apr-2010 07:02:43] <ckrough> did you try to import time into one of your scripts?

[27-Apr-2010 07:03:10] <chemist> twill script? no, its a one line script

[27-Apr-2010 07:03:20] <chemist> and it works on another zenoss instance

[27-Apr-2010 07:03:27] <ckrough> odd

[27-Apr-2010 07:03:48] <ckrough> I was just asking because I remember having issues with importing Time in something I was working on

[27-Apr-2010 07:04:58] <chemist> http://pastebin.com/US7cria8

[27-Apr-2010 07:05:03] <chemist> the complete errot

[27-Apr-2010 07:05:07] <chemist> error

[27-Apr-2010 07:08:05] <chemist> the strange thing is that if I execute the command manually I get a Transaction Success

[27-Apr-2010 07:14:29] <ckrough> it may be a permission issues with importing

[27-Apr-2010 07:14:45] <ckrough> are you importing 'time' in your script, or is that some dependency that is trying to import it

[27-Apr-2010 07:15:06] <chemist> I'm guessing its some dependency

[27-Apr-2010 07:15:19] <ckrough> I recall trying to import 'time' in a zope Python Script and having the same issue

[27-Apr-2010 07:15:22] <ckrough> no solution

[27-Apr-2010 07:15:38] <ckrough> you execute it manually as the zenoss user and it's fine?

[27-Apr-2010 07:15:49] <chemist> because the script is a simple go script

[27-Apr-2010 07:15:55] <chemist> with an address

[27-Apr-2010 07:16:32] <chemist> and as I said, the same script works on another zenoss installation

[27-Apr-2010 07:19:27] <chemist> yes, I can execute it manually and it work

[27-Apr-2010 07:50:30] <forsberg> hmmm

[27-Apr-2010 07:52:08] <ckrough> indeed

[27-Apr-2010 07:52:43] <Simon4> definitely

[27-Apr-2010 07:54:53] <forsberg> :)

[27-Apr-2010 08:42:48] <twm1010> morning everyone

[27-Apr-2010 08:43:44] <twm1010> hey matt

[27-Apr-2010 08:45:22] <mrayzenoss> good morning

[27-Apr-2010 08:45:41] <twm1010> i'm going to bring up a 3.0 test VM today ;)

[27-Apr-2010 08:45:57] <mrayzenoss> cool, lemme see when we're generating the next revision

[27-Apr-2010 08:46:01] <mrayzenoss> should have at least 1 a week

[27-Apr-2010 08:51:12] <twm1010> ok, i might wait then, i'll probably get the VM ready

[27-Apr-2010 08:55:33] <felbane> good morning all

[27-Apr-2010 08:56:24] <mrayzenoss> good morning

[27-Apr-2010 08:58:02] <chemist> I am trying to set up synthetic trasactions

[27-Apr-2010 08:58:13] <chemist> when I run the "test against device" I get a "ImportError: No module named time"

[27-Apr-2010 08:58:20] forsberg is now known as fOrsberg

[27-Apr-2010 08:58:22] <chemist> http://pastebin.com/US7cria8

[27-Apr-2010 08:58:29] <chemist> the strange thing is that if I execute the command manually I get a Transaction Success

[27-Apr-2010 09:00:46] <mrayzenoss> hmm… I wonder if there's some path issue?

[27-Apr-2010 09:01:40] <chemist> I am running 2.52 on debiab

[27-Apr-2010 09:01:47] <chemist> *debian

[27-Apr-2010 09:03:34] <chemist> even more odd is that the same synthetic transaction works on another zenoss instance, same version, sameOS

[27-Apr-2010 09:07:51] <felbane> Does anyone know how I would go about accessing the dmd within a modeler plugin? I'm trying the "from Products.ZenUtils.ZenScriptBase import ZenScriptBase   dmd=ZenScriptBase(connect=True).dmd" method but the modeler doesnt work with that line in there.

[27-Apr-2010 09:12:22] <felbane> or, alternatively, how can i use the deviceproxy already being passed to the modeler to check if an object within device/os already exists?

[27-Apr-2010 09:18:25] <mrayzenoss> felbane: this might help: http://pastebin.org/185332

[27-Apr-2010 09:19:11] <mrayzenoss> felbane: check out the zenpacks SVN repo and grep against it, that's my technique when I need to find something

[27-Apr-2010 09:19:35] <felbane> ok will do

[27-Apr-2010 09:19:56] <mrayzenoss> one of the nice things about the repo hitting critical mass on checkins

[27-Apr-2010 09:20:00] <mrayzenoss> 120+ examples

[27-Apr-2010 09:22:41] <blull_> rmm lame question but, how do I tell zenoss to not attempt to login via ssh when mapping a host?

[27-Apr-2010 09:23:34] <fragfutter> blull_: remove it from the modeller plugins

[27-Apr-2010 09:24:31] <fragfutter> blull_: Classes -> Devices -> ... -> dropdown -> more -> collector plugins

[27-Apr-2010 09:24:37] <fragfutter> zport/dmd/Devices/Discovered/zSortableProperty?prop=zCollectorPlugins

[27-Apr-2010 09:25:01] <fragfutter> blull_: remove all zenoss.cmd plugins

[27-Apr-2010 09:25:17] <blull_> ahh okay

[27-Apr-2010 09:25:24] <blull_> thanks

[27-Apr-2010 09:25:35] <blull_> actualyl

[27-Apr-2010 09:25:47] <blull_> i need to just disable ssh logins, not windows logins

[27-Apr-2010 09:25:51] <blull_> lemme see if that works

[27-Apr-2010 09:26:58] <blull_> actually all I have added is snmp stuff

[27-Apr-2010 09:27:00] <blull_> no cmd stuff

[27-Apr-2010 09:27:13] <blull_> I have none of the zenoss.cmd.*

[27-Apr-2010 09:27:45] <blull_> wait i lied.

[27-Apr-2010 09:27:46] <blull_> nm :)

[27-Apr-2010 09:33:44] <rmatte> lol

[27-Apr-2010 09:34:37] <rmatte> I'm cloning one of my production boxes so that I can mess around and figure out what the cause of some errors is

[27-Apr-2010 09:34:51] <rmatte> hopefully this will lead to a solution

[27-Apr-2010 09:44:16] <twm1010> bah, solutions are for panzies!

[27-Apr-2010 09:44:20] <twm1010> reboot until it works!

[27-Apr-2010 09:47:30] <felbane> mrayzenoss: I'm sure I'm not the first person to try to do this. Basically, I want to check if a modeled object is already there and if so, ignore it. That way, I don't override any user settings (monitoring status, etc)

[27-Apr-2010 09:47:44] <felbane> do you know if any of the core code does this?

[27-Apr-2010 09:48:17] <mrayzenoss> felbane: you want to lock it from remodeling?

[27-Apr-2010 09:48:37] <mrayzenoss> twm1010: there should be another alpha drop by Thursday

[27-Apr-2010 09:48:55] <felbane> Yes. my modeler sets om.monitor = False

[27-Apr-2010 09:49:12] <felbane> so i dont want it to disable monitoring when remodeling

[27-Apr-2010 09:49:22] <felbane> for objects that already exist

[27-Apr-2010 10:11:24] <cgibbons> bouncy

[27-Apr-2010 10:11:59] <chemist> is the timeout field in the http monitor the amount of time  it will wait for the command to execut

[27-Apr-2010 10:12:04] <chemist> e?

[27-Apr-2010 10:13:38] fOrsberg is now known as forsberg

[27-Apr-2010 10:20:22] <C_Kode> Q: I have version 2.3.2 and I want to download the 2.5.2 CentOS installer.  Will this upgrade on its own, or is there something I else I need to do to upgrade?

[27-Apr-2010 10:20:42] <mrayzenoss> C_Kode: 2.3.2 to 2.5.2?  Stack or RPM install?

[27-Apr-2010 10:20:52] <mrayzenoss> you'll have to go to 2.4.5 first

[27-Apr-2010 10:21:24] <C_Kode> Thanks, I just found that.  But I can use the CentOS .bin file?

[27-Apr-2010 10:21:50] <mrayzenoss> yeah, get the proper 32/64 bit 2.4.5 .bin and the 2.5.2 .bin files

[27-Apr-2010 10:23:00] <C_Kode> I'm not seeing 2.4.2 on the downloads page or a link to a respository.

[27-Apr-2010 10:23:39] <C_Kode> Found the sourceforge link

[27-Apr-2010 10:24:16] <mrayzenoss> 2.4.5 is here: https://sourceforge.net/projects/zenoss/files/

[27-Apr-2010 10:25:23] <ErikRose>  Does anybody know offhand how to get Zenoss to let go of the memory of a process? E.g., it's used to Squid being run with -s, and now that it's not anymore, Zenoss continually freaks out about the process being down, despite a perfectly good process matching ^/usr/sbin/squid.

[27-Apr-2010 10:30:21] <aclark> ErikRose! Sounds like a bug…

[27-Apr-2010 10:30:25] <C_Kode> I have a 2.2.0 version too, but I don't see the 2.2.4 or 2.3.x version on SourceForge.

[27-Apr-2010 10:30:41] <C_Kode> Guessing I should just blow this one up hehe

[27-Apr-2010 10:31:10] <ErikRose> aclark: You followed me! :-) I'm going to assume it's just surprising UI.

[27-Apr-2010 10:31:13] * ErikRose heads to the manual.

[27-Apr-2010 10:32:52] <ganiman> hello

[27-Apr-2010 10:33:11] * ganiman pokes ptom

[27-Apr-2010 10:33:57] <ptom> Does anyone know of a way that I can do a snmp getnext in a custom plugin?

[27-Apr-2010 10:34:06] <twm1010> @mrayzenoss: cool, i'll download it then

[27-Apr-2010 10:34:10] <aclark> ErikRose: I hang in here to feed my obsession ;-)

[27-Apr-2010 10:47:01] <C_Kode> Is there away to name the tmp location to use during zenbackup?

[27-Apr-2010 10:48:00] <rmatte> C_Kode: all of the arguments are listed here: http://docs.huihoo.com/zenoss/admin-guide/2.4.2/ch14s04.html

[27-Apr-2010 10:48:11] <C_Kode> Thanks

[27-Apr-2010 10:48:24] <rmatte> pretty sure you're looking for the --file option

[27-Apr-2010 10:48:42] <mrayzenoss> or the —temp-dir?

[27-Apr-2010 10:48:52] <rmatte> --file=/path/to/file.tgz

[27-Apr-2010 10:49:02] <rmatte> is --temp-dir an actual option?

[27-Apr-2010 10:49:09] <mrayzenoss> http://dev.zenoss.org/trac/changeset/5533

[27-Apr-2010 10:49:22] <rmatte> ah

[27-Apr-2010 10:49:36] <rmatte> --temp-dir it is then

[27-Apr-2010 10:51:35] <C_Kode> Great.  I was having a space issuse on /tmp that was on the root fs trying to back it up

[27-Apr-2010 10:51:54] <rmatte> ah

[27-Apr-2010 10:54:08] <rmatte> mrayzenoss: You mentioned a command a while back to clear the MySQL database

[27-Apr-2010 10:54:17] <rmatte> zenclear or something

[27-Apr-2010 10:54:19] <rmatte> can't remember

[27-Apr-2010 10:55:41] <C_Kode> rmatte: Is that to clear old data and truncate the db?

[27-Apr-2010 10:55:52] <rmatte> nah, to literally wipe the whole db

[27-Apr-2010 10:55:56] <rmatte> I need to do it for testing

[27-Apr-2010 10:55:57] <mrayzenoss> zenwipe?

[27-Apr-2010 10:56:07] <rmatte> does that wipe just the db?

[27-Apr-2010 10:56:10] <rmatte> or does it do zope as well?

[27-Apr-2010 10:56:12] <mrayzenoss> lemme check

[27-Apr-2010 10:56:15] <rmatte> thanks

[27-Apr-2010 10:56:17] <mrayzenoss> I think it's both

[27-Apr-2010 10:56:22] <C_Kode> I would like to purge old data. My db appears to be several gbs in size

[27-Apr-2010 10:56:41] <rmatte> ah

[27-Apr-2010 10:56:42] <rmatte> "Located in trunk/inst. It will rebuild your dmd and clear the events

[27-Apr-2010 10:56:42] <rmatte> database. It also loads a google key for localhost into the

[27-Apr-2010 10:56:44] <rmatte> system. Handy for developing."

[27-Apr-2010 10:56:52] <rmatte> yeh, it's both

[27-Apr-2010 10:56:54] <mrayzenoss> yeah, that's the one

[27-Apr-2010 10:57:05] <rmatte> oh well, I think I'll just set the rollover to 1 day, should work

[27-Apr-2010 10:57:09] <mrayzenoss> http://dev.zenoss.org/trac/browser/trunk/inst/zenwipe.sh

[27-Apr-2010 10:57:12] <rmatte> I could go in and drop everything but I'm too lazy

[27-Apr-2010 10:58:24] forsberg is now known as fOrsberg

[27-Apr-2010 10:58:30] <mrayzenoss> the Data.fs is all you need for your objects

[27-Apr-2010 10:59:44] <chemist> if the txt in a transform is red, does that indicate a syntax error?

[27-Apr-2010 11:00:50] <rmatte> chemist: yes

[27-Apr-2010 11:01:58] <chemist> would someone check this for me? http://pastebin.com/LP1L7Q2r

[27-Apr-2010 11:08:48] <rmatte> chemist: hmmm

[27-Apr-2010 11:09:11] <chemist> rmatte: that bad?

[27-Apr-2010 11:09:20] <rmatte> well, I'd do it a bit differently... hold on...

[27-Apr-2010 11:09:27] <chemist> I found it on the wiki somewhere

[27-Apr-2010 11:09:29] <rmatte> it's obviously the quotes that are messing you up

[27-Apr-2010 11:09:41] <rmatte> actually, you need a space before evt.severity = 5

[27-Apr-2010 11:10:03] <chemist> after the 0: ?

[27-Apr-2010 11:10:23] <chemist> or on the evt.severity line?

[27-Apr-2010 11:10:23] <rmatte> yeh, evt.severity = 5 is within the if loop

[27-Apr-2010 11:10:29] <rmatte> so it needs to have a space before it

[27-Apr-2010 11:10:31] <rmatte> to indent it

[27-Apr-2010 11:10:48] <rmatte> but give me a second...

[27-Apr-2010 11:11:40] * chemist wishes he could download python into his brain

[27-Apr-2010 11:12:15] <rmatte> I'd do it like this:

[27-Apr-2010 11:12:16] <rmatte> http://pastebin.com/DWJMDTMs

[27-Apr-2010 11:12:21] fOrsberg is now known as forsberg

[27-Apr-2010 11:12:55] <chemist> what's the difference/advantage?

[27-Apr-2010 11:13:51] <rmatte> there's no real "advantage"...

[27-Apr-2010 11:14:03] <rmatte> I'm just not sure of the way you had it because of the quotes

[27-Apr-2010 11:14:18] <chemist> it worked nyhow

[27-Apr-2010 11:14:19] <rmatte> I think the .find needs to be an exact match, whereas re.search can be a partial match

[27-Apr-2010 11:14:30] <chemist> so thanks :)

[27-Apr-2010 11:14:35] <rmatte> and I think the single quotes in the event summary that you put in were messing things up

[27-Apr-2010 11:14:38] <rmatte> no problem

[27-Apr-2010 11:14:39] <chemist> I'll try your method

[27-Apr-2010 11:14:43] <rmatte> k

[27-Apr-2010 11:15:16] <ckrough> you'll need to escape those '.'s

[27-Apr-2010 11:15:34] <ckrough> well, it will work anyway but.

[27-Apr-2010 11:16:18] <rmatte> yeh, if he escapes them it would work, I just forgot how to do escaping in python

[27-Apr-2010 11:16:23] <rmatte> it's not just a backslash is it?

[27-Apr-2010 11:16:32] <chemist> i tried that

[27-Apr-2010 11:16:39] <ckrough> I dont think python cares in taht situation

[27-Apr-2010 11:16:42] <ckrough> its just regex rules

[27-Apr-2010 11:16:45] <ckrough> so \.

[27-Apr-2010 11:16:54] <rmatte> ah

[27-Apr-2010 11:17:26] <rmatte> pretty sure that re.search won't care about the .'s

[27-Apr-2010 11:17:29] <rmatte> but I could be wrong

[27-Apr-2010 11:17:40] <rmatte> he said it works as is

[27-Apr-2010 11:17:42] <ckrough> well, regex . matches '.' so it will work anyway

[27-Apr-2010 11:17:50] <rmatte> yeh

[27-Apr-2010 11:20:09] <ptom> help with a custom plugin?

[27-Apr-2010 11:23:45] <C_Kode> You said zenwipe will wipe all old events, but not damage the install?

[27-Apr-2010 11:24:24] <mrayzenoss> zenwipe removes everythign

[27-Apr-2010 11:24:39] <mrayzenoss> C_Kode: http://dev.zenoss.org/trac/browser/trunk/inst/zenwipe.sh

[27-Apr-2010 11:24:50] <mrayzenoss> you can see it deletes the object database and drops the event table

[27-Apr-2010 11:25:28] <C_Kode> Data.fs is the object database?

[27-Apr-2010 11:25:33] <mrayzenoss> yes

[27-Apr-2010 11:25:49] <C_Kode> Ojbect I'm guessing are the checks themselves (devices, etc)

[27-Apr-2010 11:26:00] <rmatte> devices, locations, any objects

[27-Apr-2010 11:26:08] <rmatte> MySQL only stores events

[27-Apr-2010 11:26:14] <C_Kode> Ok, so running that is like installing fresh.

[27-Apr-2010 11:26:19] <rmatte> correct

[27-Apr-2010 11:26:23] <C_Kode> Thanks

[27-Apr-2010 11:27:27] <otakup0pe> random question... is there any kind of convention for packaing non-zenpack zenoss add-ons ?

[27-Apr-2010 11:28:13] <rmatte> zenoss add-ons are generally packaged as zenpacks

[27-Apr-2010 11:28:18] <rmatte> explain?

[27-Apr-2010 11:29:03] <otakup0pe> i have a suite of tools which are designed to run in an immersive virtual environment and retrieve information from a zenoss install entierly via web services.

[27-Apr-2010 11:29:20] <otakup0pe> so mostly 3d bar graphs and sprites which represent events. very eyecandy stuff.

[27-Apr-2010 11:29:57] <otakup0pe> alternatively, picture that scene in the matrix when the racks of weapons appear and stretch off as as far as the eye can see. except instead of racks of weapons it's bw utilization charts from switches.

[27-Apr-2010 11:30:19] <otakup0pe> so yes, i don't know how it would fit as a zenpack at all seeing as it has some intense 3rd party dependencies.

[27-Apr-2010 11:32:51] <C_Kode> Will truncating the tables in the "events" database cause any issues?

[27-Apr-2010 11:33:33] <mrayzenoss> C_Kode: you'll lose all your events, perhaps orphaning their history

[27-Apr-2010 11:33:38] <mrayzenoss> http://dev.zenoss.org/trac/browser/trunk/Products/ZenEvents/db/zenevents.sql

[27-Apr-2010 11:34:05] <mrayzenoss> C_Kode: oops.  I thought you asked about the 'status' table, my bad

[27-Apr-2010 11:34:12] <mrayzenoss> no, no problems

[27-Apr-2010 11:34:30] <mrayzenoss> just losing all your events

[27-Apr-2010 11:34:31] <C_Kode> I want to wipe all history, status, etc but not the objects themselves.

[27-Apr-2010 11:34:37] <C_Kode> Great thanks

[27-Apr-2010 11:35:04] <mrayzenoss> otakup0pe: I've been meaning to put together a landing page for "Integrations", that fall between the cracks of what a ZenPack does

[27-Apr-2010 11:35:16] <C_Kode> The DB is huge and I don't need all that old data

[27-Apr-2010 11:35:38] <mrayzenoss> otakup0pe: since you probably have code that isn't installed into Zenoss

[27-Apr-2010 11:35:57] <otakup0pe> yeah it's all via web service.

[27-Apr-2010 11:36:13] <rmatte> you can make a zenpack that's a component of that

[27-Apr-2010 11:36:22] <otakup0pe> where would teh zenpack come in ?

[27-Apr-2010 11:36:37] <otakup0pe> given the cpu load provided by the 3d environment, you probably don't want it to interact with your monitoring hardware at all

[27-Apr-2010 11:36:38] <rmatte> well, I guess there's really nothing to be installed in to Zenoss at all in that case

[27-Apr-2010 11:36:45] <rmatte> so you can rig up whatever installation method you like

[27-Apr-2010 11:36:56] <rmatte> an rpm perhaps?

[27-Apr-2010 11:37:04] <mrayzenoss> otakup0pe: you can put the integration stuff here: community/documentation/wiki/integrations and I'll come up with a pretty landing page to list them and get more notice

[27-Apr-2010 11:37:24] <otakup0pe> ok i can start poking in the wiki :)

[27-Apr-2010 11:37:26] <mrayzenoss> and I'll get it linked from the ZenPacks page

[27-Apr-2010 11:37:39] <otakup0pe> i'm only thinking about this now as i just registered the github project heh

[27-Apr-2010 11:38:10] <mrayzenoss> well, I've been meaning to organize things like the zenoss/mindtouch mashup, the puppet module, the php-weathermap, the zipec and zapplets

[27-Apr-2010 11:38:59] <otakup0pe> awesome

[27-Apr-2010 11:39:00] <mrayzenoss> otakup0pe: If you put it on the wiki, I'll see it and remember to link it up when I add that landing page

[27-Apr-2010 11:39:13] <otakup0pe> yeah and i'll keep you posted. it's a very pretty way of viewing zenoss data i promise.

[27-Apr-2010 11:39:25] <mrayzenoss> otakup0pe: I remember seeing some of the early screenshots

[27-Apr-2010 11:40:30] <otakup0pe> yeah the bargraphs look mostly the same. i'm certainly not a 3d artist lol.

[27-Apr-2010 11:45:01] <chemist> if I add a device to /Devices/Ping on a remote collector, will the ping itself be from the master or the collector?

[27-Apr-2010 12:15:54] RoAk is now known as RoAkSoAx

[27-Apr-2010 12:17:37] <ErikRose> aclark: I had to restart zenprocess, btw. That cleared up the bad process entries.

[27-Apr-2010 12:19:19] <aclark> ErikRose: Ah nice

[27-Apr-2010 12:34:02] <ckrough> Im getting high cpu load on some larger routers during modelling. Does zenoss try to pull down cam tables?

[27-Apr-2010 12:34:53] <ckrough> Im not running the bridge-mib zenpack

[27-Apr-2010 13:17:38] <rmatte> how the heck do you comment a .pt file (skin file)?

[27-Apr-2010 13:17:44] <rmatte> # doesn't seem to work

[27-Apr-2010 13:19:15] <rmatte> ah <!-- and -->

[27-Apr-2010 13:40:30] forsberg is now known as fOrsberg

[27-Apr-2010 13:45:51] <ckrough> is there a debug more verbose than -v10? I want to see exactly what the modeler is doing

[27-Apr-2010 13:48:42] <rmatte> nope

[27-Apr-2010 13:48:46] <rmatte> v10 is max I believe

[27-Apr-2010 13:48:51] <rmatte> you can do an strace

[27-Apr-2010 13:48:55] <ckrough> meh

[27-Apr-2010 13:49:09] <ErikRose> $ doesn't appear to match at the end of a process-matching regexp. :-(

[27-Apr-2010 13:49:28] <ckrough> ErikRose: make sure you are matching against what snmp returns and not what ps returns

[27-Apr-2010 13:49:34] <ErikRose> Good point.

[27-Apr-2010 13:49:51] <ErikRose> I just relaxed the expression; that's good enough for me.

[27-Apr-2010 13:53:37] <ErikRose> Thanks, btw

[27-Apr-2010 14:00:12] <davetoo> well OK then,

[27-Apr-2010 14:00:19] <davetoo> how well do "we" have LDAP working with Core now?

[27-Apr-2010 14:00:24] <davetoo> LDAP auth

[27-Apr-2010 14:00:33] <rmatte> I have LDAP auth working fine in my core installs

[27-Apr-2010 14:02:21] <davetoo> I never found the role-based controls sufficiently granular/flexible

[27-Apr-2010 14:03:16] <davetoo> also..

[27-Apr-2010 14:03:22] <davetoo> why did check_snmp disappear???

[27-Apr-2010 14:03:27] <davetoo> :(

[27-Apr-2010 14:03:38] <davetoo> That broke some of my templates.

[27-Apr-2010 14:03:57] <rmatte> the role based rules aren't that unflexible

[27-Apr-2010 14:04:15] <rmatte> you can generate custom roles then assign them to LDAP groups

[27-Apr-2010 14:05:02] <rmatte> and I have no idea about check_snmp, what did it do?

[27-Apr-2010 14:07:35] <davetoo> In the simple case, acted like snmpget with formatted output, but also do nagios return codes based on varios tests

[27-Apr-2010 14:07:46] <davetoo> either string/regex matches, or numerical comparisons

[27-Apr-2010 14:12:45] <miamiseb> I've managed to add my cisco device through autodiscovery and have configured a RW string that works enough that snmpwalk display the correct output, but I don't see a way to perform remote backups of the config, the integration with rancid is not clearly explained, anyone played with it?

[27-Apr-2010 14:22:09] <rmatte> AHA!

[27-Apr-2010 14:22:17] <ckrough> SCIENCE!

[27-Apr-2010 14:22:22] <rmatte> I finally found the part of the status page that is causing the page not to load in IE...

[27-Apr-2010 14:24:08] <bigegor> rmatte: you are the man

[27-Apr-2010 14:26:17] <rmatte> lol

[27-Apr-2010 14:26:27] <rmatte> it's the comments field that appears to cause problems

[27-Apr-2010 14:26:37] <rmatte> when I comment that out the page loads fine

[27-Apr-2010 14:32:02] <ptom> is it possible to set a custom property using a rest call ?

[27-Apr-2010 14:38:23] <rmatte> some of the comments have garbage characters in them, glad this is an easy fix

[27-Apr-2010 14:55:25] <mrayzenoss1> how'd the garbage get there?

[27-Apr-2010 14:55:45] mrayzenoss1 is now known as mrayzenoss

[27-Apr-2010 14:56:07] <rmatte> no idea

[27-Apr-2010 14:56:13] <rmatte> I wasn't the one who added those comments

[27-Apr-2010 14:56:31] <rmatte> and the person who did used IE while doing it I'm sure, since he always uses IE for everything

[27-Apr-2010 14:56:37] <rmatte> so maybe it was IE that did it

[27-Apr-2010 14:56:43] <mrayzenoss> copy & paste from Word?

[27-Apr-2010 14:56:50] <rmatte> that could be it actually

[27-Apr-2010 14:56:56] <rmatte> I'll have to ask him tomorrow

[27-Apr-2010 14:57:28] <rmatte> anyways, I fixed them up, so that's that

[27-Apr-2010 14:57:35] <mrayzenoss> feel free to update your ticket

[27-Apr-2010 14:57:41] <rmatte> already have

[27-Apr-2010 14:58:31] <rmatte> now I just need to get the server that I had zenmigrate problems with upgraded tonight and reboot/reindex all of the servers (since I can't actually use the "edit config" option for any of the daemons until I've done that for whatever reason)

[27-Apr-2010 14:58:35] <rmatte> and then I'll be golden

[27-Apr-2010 14:59:03] <rmatte> our front line guys are complaining that there's no way to do a copy/paste from the event console

[27-Apr-2010 14:59:26] <rmatte> so I'll have to make a command that they can run which dumps the event console contents

[27-Apr-2010 14:59:41] <rmatte> might be a nice feature to include for the stone crab release though

[27-Apr-2010 15:00:54] <mrayzenoss> You mean like the export button?

[27-Apr-2010 15:01:11] <rmatte> they just want to be able to straight copy to clipboard so that they can paste in to tickets

[27-Apr-2010 15:01:15] <rmatte> they don't want to have to export first

[27-Apr-2010 15:01:28] <rmatte> so something like the export button but which would copy directly to clipboard would be awesome

[27-Apr-2010 15:02:31] <rmatte> so that you could highlight multiple items, click the button, then paste

[27-Apr-2010 15:02:33] <mrayzenoss> yeah, good suggestion, open a ticket please :)

[27-Apr-2010 15:02:38] <rmatte> I shall

[27-Apr-2010 15:54:41] <chemist> if I add a device to /Devices/Ping on a remote collector, will the ping itself be from the master or the collector?

[27-Apr-2010 15:55:35] <rmatte> chemist: both I believe

[27-Apr-2010 15:55:48] <rmatte> though it's supposed to only be from the collector

[27-Apr-2010 15:55:53] <rmatte> test it and see

[27-Apr-2010 16:03:31] <davetoo> Anybody know how to copy a perf template in version 3?

[27-Apr-2010 16:12:46] <davetoo> and how do I get to a Device Class's zPropertyManager

[27-Apr-2010 16:18:39] <chemist> I tested it and itwouldappear it is the master that does it

[27-Apr-2010 16:19:25] <chemist> not what I was hoping for

[27-Apr-2010 16:28:53] <rmatte> yeh, that's kind of lame

[27-Apr-2010 16:29:16] <davetoo> um

[27-Apr-2010 16:29:39] <davetoo> you sure? Is the collector set right?

[27-Apr-2010 16:29:41] <davetoo> is this Enterprise?

[27-Apr-2010 16:29:53] <chemist> no, core

[27-Apr-2010 16:30:05] <davetoo> oh, I don't know how the Core distributed-collector hack works

[27-Apr-2010 16:30:14] <chemist> Iset thecollectorup with egor's zenpack

[27-Apr-2010 16:30:15] <davetoo> with Enterprise, zenping on the remote collector does that job

[27-Apr-2010 16:32:01] * chemist hammers space bar

[27-Apr-2010 16:33:07] <davetoo> Using Zenoss version 3 is like learning to drive a car in a country with opposite-hand drive

[27-Apr-2010 16:33:15] <davetoo> e.g. for me it'd be like driving in the UK

[27-Apr-2010 16:33:32] <chemist> I know how that feels :)

[27-Apr-2010 16:33:52] <chemist> but I learned to drive here

[27-Apr-2010 17:21:47] <davetoo> I sure would like to know how to print/dump the topology/pingtree from a collector

[27-Apr-2010 17:23:08] RobertLaptop_ is now known as RobertLaptop

[27-Apr-2010 17:42:43] <miamiseb> Anyone happen to know if the WMI login collector plugin is compatible with Server 2008?

[27-Apr-2010 17:45:39] <miamiseb> I ask because I have an administrator username and password configured that I've tested but upon modeling I get "Could not read the windows event log (NT_STATUS_CONNECTION_REFUSED)

[27-Apr-2010 18:00:23] <miamiseb> Looks like I may need the WMI Data Source zenpack which as part of it's description states "The advantage for the WMI Data Source is that credentials should be properly handled and performance may be improved." from my understanding then credentials have a problem without this zenpack.

[27-Apr-2010 18:04:17] <twm1010> miamiseb: what are you trying to do?

[27-Apr-2010 18:05:06] <miamiseb> Grab as much info as possible from a windows server (which right now, is no info) although, really, I care more about being able to backup cisco's based upon syslog messages.

[27-Apr-2010 18:05:37] <twm1010> ok, that's a bunch of loaded questions

[27-Apr-2010 18:06:05] <twm1010> zenoss, in a default install, requires you to load SNMP informant to gather information via SNMP

[27-Apr-2010 18:06:33] <twm1010> you also need to set a zWinUser and a zWinPassword in zProperties of /Servers/Windows

[27-Apr-2010 18:06:51] <twm1010> if you don't want to use SNMP, you can install Egor's WMI Datasource zenpack, and it's child zenpacks

[27-Apr-2010 18:07:26] <twm1010> once you do that, you'll need to go to the assigned templates on /Server/Windows, and change the performance template bindings to use the new WMI based ones

[27-Apr-2010 18:07:28] <miamiseb> Yeah, I've removed the snmp collector plugins from this hosts, and added the default wmi one, and set the proper zProperties, but haven't installed the zenpack

[27-Apr-2010 18:07:47] <twm1010> you need the WMI datasource zenpack for any of egor's other WMI zenpacks to work

[27-Apr-2010 18:07:53] <twm1010> exchange/iis, etc

[27-Apr-2010 18:08:14] <twm1010> it will install a new zenoss daemon, zenperfwmi

[27-Apr-2010 18:08:21] <twm1010> i would suggest you restart zenoss after installing his zenpack

[27-Apr-2010 18:09:02] <miamiseb> Yeah, I was trying to get a default collector plugin zenoss.wmi.WinServiceMap to go and thats the one that been reporting issues, I'll download his zenpack and give it a go

[27-Apr-2010 18:09:50] <twm1010> ermm... zenoss.wmi.WinServiceMap

[27-Apr-2010 18:09:54] <twm1010> is a default collector

[27-Apr-2010 18:10:12] <twm1010> used by Zenoss to gather information on windows services

[27-Apr-2010 18:10:21] <twm1010> probably want to leave that there.

[27-Apr-2010 18:10:29] <miamiseb> right, thats what I was saying, that one isn't working, it saying NT_CONNECTION_REFUSED

[27-Apr-2010 18:10:44] <miamiseb> but not firewall between the hosts blocking WMI ports

[27-Apr-2010 18:11:13] <twm1010> NT_Connection_Refused, means the host throwing the error is being reached, and refusing it past any firewall

[27-Apr-2010 18:11:29] <miamiseb> it means the port is closed generally

[27-Apr-2010 18:11:30] <twm1010> so... either A) username/password ain't workin

[27-Apr-2010 18:11:38] <miamiseb> but its not, and I've double checked the user/pass

[27-Apr-2010 18:11:43] <twm1010> ok, what O/S

[27-Apr-2010 18:11:48] <miamiseb> it is however, server 2008, and I don't know if anything's changed

[27-Apr-2010 18:11:49] <twm1010> XP, by chance?

[27-Apr-2010 18:11:51] <twm1010> hrmmm..

[27-Apr-2010 18:13:23] <twm1010> hrmm.. i would suspect something else then

[27-Apr-2010 18:13:32] <miamiseb> Looking for a wmi browsing tool to test from the linux host

[27-Apr-2010 18:13:52] <twm1010> the account your using is an admin account?

[27-Apr-2010 18:14:01] <miamiseb> "Administrator" in fact

[27-Apr-2010 18:14:07] <twm1010> there is a string we can use to test from the zenoss server command line

[27-Apr-2010 18:14:09] <miamiseb> and its a workgroup environment

[27-Apr-2010 18:14:10] <twm1010> let me look it up

[27-Apr-2010 18:15:46] <miamiseb> apt has a package for wbemcli, nice.

[27-Apr-2010 18:18:41] <twm1010> zenoss has one built-in

[27-Apr-2010 18:18:48] <twm1010> i can run this logged in as zenoss user

[27-Apr-2010 18:19:03] <twm1010>  wmic -U Administrator%MyPassword //MyServer "select * from Win32_computerSystem"

[27-Apr-2010 18:19:33] <miamiseb> The program 'wmic' is currently not installed.  You can install it by typing:

[27-Apr-2010 18:19:46] <twm1010> hrmmm

[27-Apr-2010 18:20:09] <twm1010> you can try $ZENHOME/bin/wmic -U Administrator%MyPassword //MyServer "select * from Win32_computerSystem"

[27-Apr-2010 18:20:36] <twm1010> dunno where your zenhome is, could be /opt/zenoss....

[27-Apr-2010 18:22:14] <miamiseb> not in that dir, but I think wmi is just not allowing remote monitoring on that host

[27-Apr-2010 18:23:03] <rhettardo> echo $ZENHOME

[27-Apr-2010 18:23:57] <miamiseb> nope, its enabled for remote monitoring. /usr/local/zenoss/zenoss

[27-Apr-2010 18:24:09] <miamiseb> but that path /bin doesn't contain wmic

[27-Apr-2010 18:24:13] <twm1010> perhaps, i've had to muck iwth settings before, on say... old POS win2k boxes

[27-Apr-2010 18:24:23] <twm1010> its a little finicky sometimes

[27-Apr-2010 18:24:39] <twm1010> can you use WMI explorer or WBEMtest from a windows box successfully?

[27-Apr-2010 18:25:44] <miamiseb> im downloading wmic, wbemtest requires a wmi over http proxy

[27-Apr-2010 18:26:34] <miamiseb> and it failed

[27-Apr-2010 18:26:50] <miamiseb> wait no

[27-Apr-2010 18:26:51] <miamiseb> it worked

[27-Apr-2010 18:27:10] <miamiseb> admin password status and a whole bunch of other things

[27-Apr-2010 18:28:32] <miamiseb> sweet, I added COMPUTERNAME\Administrator as the user and it seems to be working

[27-Apr-2010 18:29:53] <miamiseb> Thanks twm1010

[27-Apr-2010 18:32:25] <twm1010> try using .\Administrator

[27-Apr-2010 18:32:39] <twm1010> that way you can be more global

[27-Apr-2010 18:32:59] <miamiseb> Yeah, I suppose I should be doing this at the class template level rather than the device

[27-Apr-2010 18:33:13] <twm1010> right, work smart ;)

[27-Apr-2010 18:33:56] <miamiseb> I'm pretty familiar with solarwinds but there is certainly a learning curve here. Working through the administration manual slowly.

[27-Apr-2010 18:35:27] <miamiseb> I think once I learn how to write my own collectors and execute actions based on syslog triggers I'll have a full solution to our current woes.

[27-Apr-2010 18:47:12] <twm1010> I'm currently moving us away from Solarwinds except for core network devices

[27-Apr-2010 18:47:27] <twm1010> i don't like solarwind's requirement for you to be explicit about anythin and everything you want to monitor

[27-Apr-2010 18:47:43] <twm1010> i also found it inadequate in trap and syslog processing

[27-Apr-2010 19:07:13] <miamiseb> Yeah, I can't get it to fire backups based on syslog messages for nothing, I'm using a java tool called NetworkAuthority inventory to take and revision control my cisco configurations and the "simple event correlator" (SEC) to watch the syslog txt file and fire off a backup via NAI

[27-Apr-2010 19:07:48] <miamiseb> I'm trying to replace that mechanism with zenoss, because adding more listeners and syslog levels doesn't really scale.

[27-Apr-2010 19:59:29] <miamiseb> Gnight all

[28-Apr-2010 00:00:30] [disconnected at Wed Apr 28 00:00:30 2010]

[28-Apr-2010 00:00:31] [connected at Wed Apr 28 00:00:31 2010]