[08-Jul-2009 12:15:46] [connected at Wed Jul  8 12:15:46 2009]
[08-Jul-2009 12:15:56] [zenoss-logger (logger bot) has joined #zenoss]
[08-Jul-2009 12:27:28] <mrayzenoss> The zenoss-logger appears to be back
[08-Jul-2009 12:27:58] <mrayzenoss> http://www.zenoss.com/
[08-Jul-2009 12:28:55] <mrayzenoss> hmm.... doesn't seem to be opening URLs for people though
[08-Jul-2009 12:39:32] <rmatte> hmmmm, if I go to http://30.30.241.100:8180/zport/dmd/ZenPackManager/manage and click on packs, I can see the list of ZenPacks, but if I go to delete one I get an error saying that it doesn't exist. I'm just trying to come up with a method to make Zenoss unaware that a ZenPack exists without actually removing what that ZenPack installed
[08-Jul-2009 12:39:47] mrayzenoss1 is now known as mrayzenoss
[08-Jul-2009 12:40:02] <rmatte> hmmmm, if I go to http://30.30.241.100:8180/zport/dmd/ZenPackManager/manage and click on packs, I can see the list of ZenPacks, but if I go to delete one I get an error saying that it doesn't exist. I'm just trying to come up with a method to make Zenoss unaware that a ZenPack exists without actually removing what that ZenPack installed
[08-Jul-2009 12:40:19] <rmatte> oops
[08-Jul-2009 12:40:47] <rmatte> I figure if I add an additional step to make Zenoss unaware of the Zenpack's existance after installation, that will eliminate the performance issues
[08-Jul-2009 12:54:39] <rmatte> twm1010: this is what I went through to build the Mibs, including links to the original Mibs from the Cisco FTP as well as a link to the set of Mibs that I modified bundled together: http://demonic.cc/?p=20
[08-Jul-2009 12:55:41] <twm1010> Very nice! thank you!
[08-Jul-2009 13:00:14] <twm1010> looks like libsmi2 is not in the default repository for RHEL5
[08-Jul-2009 13:00:27] <rmatte> I use Ubuntu, so now idea
[08-Jul-2009 13:00:34] <rmatte> I just do apt-get libsmi2
[08-Jul-2009 13:00:35] <rmatte> hehe
[08-Jul-2009 13:00:45] <rmatte> no idea* rather
[08-Jul-2009 13:00:52] <twm1010> and i'd imagine libsmi won't cut it
[08-Jul-2009 13:01:04] <rmatte> I don't believe so, hunt around for an rpm
[08-Jul-2009 13:02:46] <twm1010> yeah, got it... its still going to churn for a few hours loading manually i bet?
[08-Jul-2009 13:03:11] <rmatte> loading manually is much quicker
[08-Jul-2009 13:03:24] <rmatte> although you're probably going to run in to some errors the first time around
[08-Jul-2009 13:03:34] <rmatte> then you need to load certain Mibs in order (as detailed in my guide)
[08-Jul-2009 13:03:43] <rmatte> then run it through a third time
[08-Jul-2009 13:04:12] <twm1010> hrmm, it conflict with libsmi trying to install..
[08-Jul-2009 13:04:24] <rmatte> what does?
[08-Jul-2009 13:04:31] <twm1010> the rpm i found for libsmi2
[08-Jul-2009 13:04:52] <rmatte> did you also install libsmi?
[08-Jul-2009 13:05:01] <twm1010> yep
[08-Jul-2009 13:05:08] <rmatte> uninstall libsmi and install libsmi2
[08-Jul-2009 13:05:12] <rmatte> you only need libsmi2
[08-Jul-2009 13:06:14] <twm1010> k
[08-Jul-2009 13:10:17] <twm1010> yeah... that's gonna take a while
[08-Jul-2009 13:10:37] <rmatte> hehe
[08-Jul-2009 13:11:12] <rmatte> ok, so so far I have: for p in dd.ZenPackManager.packs():
[08-Jul-2009 13:11:16] <rmatte>  p
[08-Jul-2009 13:11:36] <rmatte> and it prints out all of the zenpacks (basically like doing a zenpack --list)
[08-Jul-2009 13:11:43] <rmatte> now I just need to figure out how to delete the items
[08-Jul-2009 13:11:58] <rmatte> <ZenPack at /zport/dmd/ZenPackManager/packs/ZenPacks.Nova.Cisco.Mibs>
[08-Jul-2009 13:13:08] <cluther> That's some dangerous business.
[08-Jul-2009 13:13:21] <cluther> dmd.ZenPackManager.packs._delObject(p.id)
[08-Jul-2009 13:13:25] <rmatte> meh, it's a lab box, if I screw it up I just throw another one up
[08-Jul-2009 13:13:36] <twm1010> I was about to say, maybe you should run this by Chet, you don't want a zenpack leaving orphaned references. could get nasty on you.
[08-Jul-2009 13:13:39] <rmatte> yeh, I was looking at dmd.ZenPackManager.packs._delObject(ZenPacks.Nova.Cisco.Mibs)
[08-Jul-2009 13:13:47] <rmatte> found it in zenpack.py
[08-Jul-2009 13:14:00] <cluther> That'd work if you made the ZenPack name a string.
[08-Jul-2009 13:14:20] <rmatte> yeh, I was looking at dmd.ZenPackManager.packs._delObject('ZenPacks.Nova.Cisco.Mibs')?
[08-Jul-2009 13:14:32] <cluther> Looks good.
[08-Jul-2009 13:14:43] <rmatte> k, but will that remove the contents of the ZenPack as well?
[08-Jul-2009 13:14:47] <rmatte> or just the reference to it?
[08-Jul-2009 13:15:09] <rmatte> basically I just want to remove it from the list without removing what it installed
[08-Jul-2009 13:15:31] <rmatte> that way Zenoss won't try to index it or do whatever it is doing that's causing the performance issues
[08-Jul-2009 13:15:45] <cluther> It will remove anything that is contained in the ZenPack. This should be almost nothing. Give it a shot and see. You can always go to /manage and undo your transaction if stuff is removed.
[08-Jul-2009 13:16:05] <rmatte> what do you mean by "contained in the zenpack"?
[08-Jul-2009 13:16:29] <rmatte> ah, forgot about the undo feature, that saved my arse one time
[08-Jul-2009 13:16:35] <cluther> I use contained in a purely technical sense. Every object in the ZODB has a primary path that contains it. For a MIB this would be /zport/dmd/Mibs/....
[08-Jul-2009 13:16:36] <rmatte> ok, guess I'll give it a go..
[08-Jul-2009 13:16:53] <cluther> So the MIB is contained within Mibs, not the ZenPack.
[08-Jul-2009 13:17:12] <rmatte> ok, so it technically shouldn't touch the Mibs then in theory
[08-Jul-2009 13:17:19] <twm1010> So if a Zenpack installs mibs, and you remove the Zenpack, it won't uninstall the mibs?
[08-Jul-2009 13:17:47] <rmatte> hmmm, this is taking a while, I was thinking it'd be more instant, this is basically like doing a --remove
[08-Jul-2009 13:17:58] <cluther> twm1010: By default on removal it will go through all of its "packables" and remove them too. rmatte is trying to avoid doing that.
[08-Jul-2009 13:18:03] <rmatte> twm1010: it will, that's what I'm trying to avoid
[08-Jul-2009 13:18:24] <rmatte> well it's running, hopefully it doesn't take hours to run through
[08-Jul-2009 13:18:30] <rmatte> I was hoping this would only take a few seconds
[08-Jul-2009 13:18:51] <cluther> rmatte: Even if it works it probably wouldn't be instantaneous. All of the MIB notes/etc still have to have one side of a relationship removed from them.
[08-Jul-2009 13:19:01] <rmatte> oooh, it's done
[08-Jul-2009 13:19:09] <rmatte> I assume I have to commit?
[08-Jul-2009 13:19:11] <cluther> yes
[08-Jul-2009 13:19:13] <rmatte> k
[08-Jul-2009 13:20:53] <rmatte> if this works then it will be quite nice, hopefully it boosts performance like I expect it will
[08-Jul-2009 13:20:59] <rmatte> ok, done the commit, now to test...
[08-Jul-2009 13:21:07] <rmatte> I shouldn't have to restart zope or anything?
[08-Jul-2009 13:21:20] <cluther> No.. it should just be fast now.
[08-Jul-2009 13:21:42] <rmatte> k
[08-Jul-2009 13:22:22] <rmatte> a zenpack --list certainly ran through much faster
[08-Jul-2009 13:22:55] <cluther> Some of the other things that should be faster would be adding a new data source, going to zProperties and starting zenhub.
[08-Jul-2009 13:23:07] <rmatte> adding new data sources hang forever sometimes
[08-Jul-2009 13:23:10] <rmatte> drives me nuts
[08-Jul-2009 13:23:34] <cluther> We're trying to figure out a better way to handle these massive ZenPacks.
[08-Jul-2009 13:23:59] <cluther> We really weren't thinking about ZenPacks with tens of thousands of objects at the time.
[08-Jul-2009 13:24:04] <rmatte> I'm sure it's not the easiest thing in the world to fix, seems to be well rooted in everything
[08-Jul-2009 13:24:24] <rmatte> wow, zProperties pages are loading mint fast now
[08-Jul-2009 13:24:32] <rmatte> and this lab box is running on a pretty slow VM too
[08-Jul-2009 13:24:35] <rmatte> looks like success
[08-Jul-2009 13:24:50] <rmatte> I'll do some more testing to be sure, but it's definitely looking good
[08-Jul-2009 13:28:28] <rmatte> yup, that's the fix
[08-Jul-2009 13:29:04] <twm1010> That's not too bad really.
[08-Jul-2009 13:29:11] <rmatte> I restarted zenoss, going to make sure these heartbeat failures go away
[08-Jul-2009 13:31:09] <rmatte> twm1010: that's a better solution, install the ZenPack (yeh, it'll take a while, but it's better than moving the Mibs to a Cisco folder by hand), then run...
[08-Jul-2009 13:31:18] <rmatte> In zendmd:
[08-Jul-2009 13:31:18] <rmatte> >>> dmd.ZenPackManager.packs._delObject('ZenPacks.Nova.Cisco.Mibs')
[08-Jul-2009 13:31:18] <rmatte> >>> commit()
[08-Jul-2009 13:31:18] <rmatte> >>>
[08-Jul-2009 13:32:20] <rmatte> I'll do some more testing and once I'm 100% positive that it hasn't negatively impacted anything else (which I doubt it will), I'll have Matt throw it up on the ZenPack page as a suggested workaround for the performance issues
[08-Jul-2009 13:32:25] <rmatte> should make quite a few people happy
[08-Jul-2009 13:33:04] <rmatte> Thanks for the assistance Chet, per usual
[08-Jul-2009 13:33:08] <twm1010> Does the Zenpack put the mibs into an organizer as well?
[08-Jul-2009 13:33:15] <rmatte> yeh, the ZenPack does
[08-Jul-2009 13:33:24] <twm1010> that alone is worth it
[08-Jul-2009 13:33:24] <rmatte> it puts them in an organizer called Cisco
[08-Jul-2009 13:33:39] <rmatte> yeh, when I moved them in to the organizer by hand initially it took me like a half hour
[08-Jul-2009 13:33:44] <rmatte> which is ridiculous
[08-Jul-2009 13:33:56] <twm1010> indeed, any more than 40 and you get a timeout or something?
[08-Jul-2009 13:34:08] <rmatte> pretty much
[08-Jul-2009 13:34:16] <rmatte> so I had to do it 40 at a time
[08-Jul-2009 13:34:22] <twm1010> i get timeouts changing device classes, haven't figured out why yet.
[08-Jul-2009 13:34:51] <rmatte> like just browsing between?
[08-Jul-2009 13:34:54] <rmatte> or moving them around?
[08-Jul-2009 13:35:13] <twm1010> moving a discovered device to it's proper class
[08-Jul-2009 13:35:21] <twm1010> seems to do it the worst with large switches
[08-Jul-2009 13:35:49] <rmatte> weird, I never have problems with it
[08-Jul-2009 13:36:02] <twm1010> neither does my test box
[08-Jul-2009 13:36:08] <twm1010> alsoinstalled from stack
[08-Jul-2009 13:36:20] <twm1010> my test box is SUSE, my prod is RHEL5
[08-Jul-2009 13:36:22] <rmatte> I find that putting a checkmark next to a device name and selecting to move it to a new class rather than doing it right from the device page tends to work better for whatever reason
[08-Jul-2009 13:37:00] <twm1010> it seems to have quieted down since i first installed, things move decently quick now, but i still can't really move more than one or two devices simultaneously
[08-Jul-2009 13:37:04] <rmatte> so far I'm loving Ubuntu Server 8.04 with stack installed Zenoss
[08-Jul-2009 13:37:06] <twm1010> in the past i could move 20+...
[08-Jul-2009 13:37:09] <rmatte> seems to work beautifully
[08-Jul-2009 13:37:25] <rmatte> and Ubuntu is so easy to maintain due to the kick ass package manager
[08-Jul-2009 13:37:33] <rmatte> aptitude++
[08-Jul-2009 13:37:44] <twm1010> Indeed, I like it as well... but the boss man said run it on RHEL, so I did.
[08-Jul-2009 13:37:51] <rmatte> ah
[08-Jul-2009 13:38:05] <rmatte> I'm not a fan of RedHat, they have too much proprietary crap in it
[08-Jul-2009 13:38:23] <twm1010> Well, you have to figure they have to, to be able to support it.
[08-Jul-2009 13:38:30] <rmatte> true enough
[08-Jul-2009 13:38:48] <rmatte> we test some stuff on CentOS from time to time, but we're currently an Ubuntu shop
[08-Jul-2009 13:39:17] <rmatte> time for a victory smoke, bbiab
[08-Jul-2009 13:44:32] <rmatte> back
[08-Jul-2009 13:45:58] <rmatte> twm1010: if you're running it on a fairly power server you may want to take a look at: http://www.zenoss.com/Members/fdeckert/how-to-tweak-zenoss/
[08-Jul-2009 13:47:02] <rmatte> not sure if you're made any changes like that already, but adding the 'parallel <number>' lines for daemons that are used heavily and adding 'workers <number>' to the zenhub config can drastically increase performance on multi-threaded systems
[08-Jul-2009 13:47:30] <rmatte> fairly powerful* server rather
[08-Jul-2009 13:54:04] <dberger> what's the difference between manager and zenmanager?
[08-Jul-2009 13:54:32] <rmatte> not sure, but I always use Manager since ZenManager never seems to work properly
[08-Jul-2009 14:10:24] <rmatte> mrayzenoss: you around?
[08-Jul-2009 14:10:34] <mrayzenoss> yeah
[08-Jul-2009 14:10:56] <rmatte> did you see the bits above about the workaround for the Cisco Mibs pack?
[08-Jul-2009 14:13:45] <rmatte> it works like a charm, it'd be worth throwing it up on the page for the ZenPack
[08-Jul-2009 14:16:30] <rmatte> could apply to any other large ZenPacks as well
[08-Jul-2009 14:20:06] <venturaville> what is the syntax for referring to another graph point in the RPN field?
[08-Jul-2009 14:20:20] <twm1010> rmatte: thanks for the performance link, looking at doing some of these now
[08-Jul-2009 14:20:36] <twm1010> I'm using a DL380G4 HP for my Zenoss box, 4GB ram
[08-Jul-2009 14:20:40] <rmatte> twm1010: no problem, they've helped a lot on the box that we're using to monitor 350 devices on
[08-Jul-2009 14:20:54] <rmatte> what's the processor in that thing?
[08-Jul-2009 14:21:47] <twm1010> shows up at 4 x 3.2Ghz Xeons
[08-Jul-2009 14:21:52] <twm1010> I would guess dual hyper-threaded
[08-Jul-2009 14:22:14] <rmatte> probably
[08-Jul-2009 14:23:14] <rmatte> the rackmount I have at home has quad xeon with hyperthreading & shows up as 8 1.6GHz CPUs when it's really 4 1.6GHz
[08-Jul-2009 14:23:43] <rmatte> that's still a nice chunk of processing power there, running multiple VMs on it or just a main OS?
[08-Jul-2009 14:23:54] <twm1010> RHEL runs native, no virtualization
[08-Jul-2009 14:24:08] <rmatte> cool, then those performance enhancements will definitely help you out
[08-Jul-2009 14:24:21] <twm1010> should do the trick, I almost got an IBM x3650, dual quad-cores, 8 GB ram, but it got snatched for another project
[08-Jul-2009 14:26:41] <rmatte> venturaville: I have 1 RPN thing that I did a while ago, let me check
[08-Jul-2009 14:28:42] <rmatte> venturaville: ok, you'll see "Available RRD Variables" at the bottom when you're viewing the graph point, you'll only see them provided that they are lower in the list in Graph Points
[08-Jul-2009 14:28:49] <venturaville> right
[08-Jul-2009 14:28:51] <venturaville> got that
[08-Jul-2009 14:28:52] <rmatte> sorry, higher in the list
[08-Jul-2009 14:29:15] <rmatte> so in this case I wanted the percent of memory used, so my RPN expression was: Total,/,100,*
[08-Jul-2009 14:30:03] <rmatte> basically it takes the total memory used (which would be right at the end of the expression, but you don't actually put it there), times 100, divided by the Total
[08-Jul-2009 14:30:11] <rmatte> it's done backwards
[08-Jul-2009 14:30:14] <venturaville> right
[08-Jul-2009 14:30:18] <venturaville> but
[08-Jul-2009 14:30:33] <venturaville> where is Total coming from?
[08-Jul-2009 14:30:39] <jb> bleh x's are a pain in the ass to deal with
[08-Jul-2009 14:30:39] <venturaville> a CDEF or a raw data point?
[08-Jul-2009 14:30:40] <jb> (ibm xseries)
[08-Jul-2009 14:30:43] <rmatte> it's another graph point
[08-Jul-2009 14:30:46] <venturaville> jb: amen to that
[08-Jul-2009 14:30:51] <rmatte> which is coming from a data point
[08-Jul-2009 14:30:56] <venturaville> jb: ibm blades are even worse
[08-Jul-2009 14:31:06] <jb> im running zenoss on a poweredge 2950 with two quad cores and 32GB of ram
[08-Jul-2009 14:31:10] <jb> local raid10 for zenoss
[08-Jul-2009 14:31:13] <venturaville> is that the actual name of the data point?
[08-Jul-2009 14:31:22] <rmatte> you take the data point that collects the total value, add it to the graph, and then you can reference it
[08-Jul-2009 14:31:36] <rmatte> venturaville: it's the name of the graph point
[08-Jul-2009 14:31:47] <venturaville> memBuffer-raw,+,memCached-raw,+,memShared-raw,+,1024,*,${here/hw/totalMemory},/,1,-,-100,*
[08-Jul-2009 14:31:47] <ckrough> jb: ditto
[08-Jul-2009 14:31:55] <venturaville> this is mine
[08-Jul-2009 14:31:56] <rmatte> the data point is mem_total_total
[08-Jul-2009 14:31:58] <venturaville> it doesn't like it
[08-Jul-2009 14:32:19] <rmatte> I just renamed the graphpoint to Total
[08-Jul-2009 14:32:50] <venturaville> error: invalid rpn expression in: memAvailReal-raw,memBuffer,+,memCached,+,memShared,+,1024,*,10478809088,/,1,-,-100,*
[08-Jul-2009 14:32:57] <rmatte> all those values show up in Available RRD Variables at the bottom I assume?
[08-Jul-2009 14:33:02] <venturaville> yep
[08-Jul-2009 14:33:11] <venturaville> does that whether I do -raw or not
[08-Jul-2009 14:33:19] <venturaville> for the data points
[08-Jul-2009 14:33:55] <venturaville> the tales expression one works as it gets interpreted into that 10478809088 number
[08-Jul-2009 14:34:03] <venturaville> (for total mem)
[08-Jul-2009 14:34:06] <rmatte> I'm no expert in RPN so can't really help you with syntax
[08-Jul-2009 14:34:12] <rmatte> but you definitely have some syntax issue there
[08-Jul-2009 14:35:01] <venturaville> found it!
[08-Jul-2009 14:35:03] <venturaville> stupid newline
[08-Jul-2009 14:37:16] <rmatte> lol
[08-Jul-2009 14:43:14] <twm1010> rmatte: this server is sure humming now
[08-Jul-2009 14:43:27] <rmatte> lol
[08-Jul-2009 14:43:30] <rmatte> nice
[08-Jul-2009 14:44:11] <rmatte> well, time to get back to reading this python book
[08-Jul-2009 14:44:22] <jb> ugh, i need to do that
[08-Jul-2009 14:44:40] <rmatte> yeh, I know enough to be dangerous, finally getting around to reading the book I have
[08-Jul-2009 14:44:48] <rmatte> gotta say though, very versatile language
[08-Jul-2009 14:44:58] <twm1010> I have a shell scripting book sitting on my shelf here I need to read...
[08-Jul-2009 14:45:15] <rmatte> I'm already pretty much an expert at shell scripting
[08-Jul-2009 14:45:39] <rmatte> it's not hard once you get used to all the pipes and redirects
[08-Jul-2009 14:45:56] <venturaville> have you learned awk yet?
[08-Jul-2009 14:46:38] <rmatte> venturaville: I haven't really worked on learning that much, one of my friends knows awk like the back of his hand but can't actually shell script worth anything, we should swap brains for a day at some point
[08-Jul-2009 14:47:10] <twm1010> I'll get there, they just keep piling stuff on to me. Our team of 8 is down to 3, and one of us is quitting, so it's down to two of us.
[08-Jul-2009 14:47:16] <rmatte> expect scripting is also very nice to learn
[08-Jul-2009 14:47:35] <venturaville> used to do way to much of that years ago (modem bank testing)
[08-Jul-2009 14:47:44] <twm1010> A guy on another team was running Solarwinds, he got let go today, so now I have to maintain that hunk of crap while building out my Zenoss system.
[08-Jul-2009 14:48:13] <jb> we run solarwinds for all network devices
[08-Jul-2009 14:48:26] <jb> around 1200 or so
[08-Jul-2009 14:48:29] <twm1010> Yes, for that purpose it is very good, and we may maintain it for that.
[08-Jul-2009 14:48:37] <rmatte> we have solarwinds instances that we use to monitor certain things, like VOIP
[08-Jul-2009 14:48:47] <jb> for system monitoring.. it sucks.
[08-Jul-2009 14:48:57] <rmatte> we are also going to primarily use it for reporting
[08-Jul-2009 14:49:03] <jb> yeah.. here too
[08-Jul-2009 14:49:07] <rmatte> since Zenoss' reporting system is pretty bad right now
[08-Jul-2009 14:49:25] <twm1010> Yep, the SW report writer is pretty nice.
[08-Jul-2009 14:49:41] <twm1010> I think a combination of two systems is needed. The fact that SW can't classify traps blew me away.
[08-Jul-2009 14:49:48] <rmatte> when Zenoss gets something similar to that it'll be gold, doesn't even really need to be quite as powerful
[08-Jul-2009 14:50:14] <rmatte> yeh, Zenoss handles traps beautifully
[08-Jul-2009 14:50:53] <twm1010> We have their application monitor too, I'm not a fan how they nickle and dime you for every component
[08-Jul-2009 14:51:05] <twm1010> An MS Exchange box consuming almost 20 out of 100 pricey licenses.
[08-Jul-2009 14:51:41] <eidolon> *sigh* where is the URL set that comes in email notifications? it's wrong in our install
[08-Jul-2009 14:52:22] <twm1010> eidolon: is it a property? I would think it would use the hostname of the zenoss server?
[08-Jul-2009 14:52:34] <jb> we monitor exchange with MOM and Zenoss
[08-Jul-2009 14:53:24] <ckrough> Im restoring an event database and its taking forever. processlist in mysql shows it spending a ton of time at "Repair with keycache", any suggestions on how to avoid that?
[08-Jul-2009 14:53:27] <eidolon> yeah, i don't want it the hostname of our server. 
[08-Jul-2009 14:53:29] <cluther> eidolon: It is "zopeurl" in $ZENHOME/etc/zenactions.conf
[08-Jul-2009 14:54:27] <twm1010> jb: how are you monitoring exchange? service monitoring plus WMI perf monitors?
[08-Jul-2009 14:54:46] <jb> yep
[08-Jul-2009 14:54:53] <jb> i added more to the exchange zenpack
[08-Jul-2009 14:55:25] <jb> cluther: hey, do you know where the temperature/fan thresholds are configured for the BigIP zenpack?.. I don't see them on the BigIPDevice or LTMVirtualServer templates
[08-Jul-2009 14:55:31] <twm1010> Ah... you're on enteprise?
[08-Jul-2009 14:55:35] <jb> yeah
[08-Jul-2009 14:55:45] <jb> but its easy enough just to monitor the perfmon counters
[08-Jul-2009 14:55:52] <cluther> jb: Fan and Temperature templates.
[08-Jul-2009 14:56:01] <jb> Oh.. doh.
[08-Jul-2009 14:56:33] <jb> i need to see about adding some more perf charts for my LTM's too
[08-Jul-2009 14:57:09] <twm1010> I think last time I did this I hacked up the WMI Zenpack to query the classes I needed
[08-Jul-2009 14:57:48] <twm1010> the perfmon zenpack usese winexe right? I was looking to avoid that and use wmic instead
[08-Jul-2009 14:58:07] <jb> not sure what the core zenpack uses
[08-Jul-2009 14:58:07] <cluther> ckrough: That can be rough. Supposedly you can restore just the schema, then "alter table ... disable keys" on some of the big tables like history, detail and log. Then load the data.
[08-Jul-2009 14:58:33] <cluther> ckrough: Of course you'll have to enable the keys again on those tables once done. This will take a while too, but not as long.
[08-Jul-2009 14:59:12] <cluther> ckrough: The best advice might be to just give the MySQL server a ton of RAM to work with. Then scale it back once you have the data in.
[08-Jul-2009 14:59:14] <ckrough> cluther: yeah, I was attempting to alter over to innodb in DFW and the altering didnt go well, now Im just working on getting back to a useable state, but the restore is brutal
[08-Jul-2009 14:59:37] <eidolon> cluther: interesting, zenactiosn.conf is empty.
[08-Jul-2009 14:59:51] <eidolon> just add zopeurl in there?
[08-Jul-2009 14:59:58] <eidolon> do i have to hup / reload / twiddle anything?
[08-Jul-2009 15:00:15] <cluther> eidolon: Add a line that just has "zopeutl http://yourhostname:8080"
[08-Jul-2009 15:00:27] <cluther> eidolon: Then run "zenactions restart" as the zenoss user.
[08-Jul-2009 15:00:34] <eidolon> k
[08-Jul-2009 15:01:26] <eidolon> restarting.
[08-Jul-2009 15:15:46] <kisielk> for running autodiscovery on our network, is zendisc through cron the preferred way to do this?
[08-Jul-2009 15:15:59] <kisielk> or is there a way to configure it through the web interface?
[08-Jul-2009 15:17:09] <twm1010> you want to schedule it?
[08-Jul-2009 15:18:11] <twm1010> There is a zProperty for Autodiscover.
[08-Jul-2009 15:19:30] <twm1010> which tells me that somewhere else in zenoss a scheduled autodiscovery takes place, and you can use the zprop to choose which networks get discovered.
[08-Jul-2009 15:23:09] <kisielk> twm1010: I couldn't find anywhere to schedule it
[08-Jul-2009 15:24:27] <kisielk> mrayzenoss: maybe you can answer this one
[08-Jul-2009 15:26:38] <cluther> Zenoss doesn't do discovery automatically. You have to put it in cron.
[08-Jul-2009 15:26:55] <cluther> 0 2 * * * bash -lc 'zendisc run'
[08-Jul-2009 15:26:57] <twm1010> cluther: what's the zprop for then
[08-Jul-2009 15:27:13] <cluther> twm1010: Controlling which networks will get discovered when you run that command.
[08-Jul-2009 15:27:22] <twm1010> ah... very cool.
[08-Jul-2009 15:31:39] <kisielk> cluther: alright. Are there any performance numbers on how long it takes to run on average? Or how much of a load it puts on? I'd like to know how frequently it would be sensible to run it
[08-Jul-2009 15:32:36] <twm1010> I think that would depend on how many networks you ahve
[08-Jul-2009 15:32:46] <twm1010> a full class B might take a while
[08-Jul-2009 15:33:16] <kisielk> well, presumably it doesn't actually do a full scan of every IP address
[08-Jul-2009 15:33:55] <kisielk> also, did the /Network hierarchy stuff get fixed up in 2.4 ?
[08-Jul-2009 15:34:05] <Juggie> is there a way to blow away all the information zenoss has gathered and restart from scratch?
[08-Jul-2009 15:34:06] <kisielk> ours is still looking pretty disjoint
[08-Jul-2009 15:34:15] <twm1010> Juggie: reinstall
[08-Jul-2009 15:35:03] <cluther> kisielk: I hesitate to say that the /Networks stuff has been perfected, but it is a lot better.
[08-Jul-2009 15:35:25] <kisielk> cluther: so should I delete my existing networks there, and rescan?
[08-Jul-2009 15:35:32] <cluther> kisielk: You can control how heavy a hand zendisc uses with the --chunk and --parallel options.
[08-Jul-2009 15:35:36] <cluther> kisielk: Yes.
[08-Jul-2009 15:35:59] <kisielk> because our network is in the 10/8 space, but right now at the top level I see all the different subnets
[08-Jul-2009 15:36:13] <kisielk> and then like, my /22 subnets have all the individual /24's in them, and all that kind of junk
[08-Jul-2009 15:36:13] <cluther> --chunk limits the number of simultaneous in-flight ping packs and I believe the default is 10, which is quite slow. I usually use 32 or 64 if I'm in more of a hurry.
[08-Jul-2009 15:36:34] <kisielk> presumably it should be some kind of hierarchy
[08-Jul-2009 15:37:38] <cluther> You can force a specific hierarchy with the zDefaultNetworkTree property on /Networks
[08-Jul-2009 15:37:47] <cluther> If you want to see /8, /16, /22, /24..
[08-Jul-2009 15:38:02] <kisielk> hm, I'm not sure I follow
[08-Jul-2009 15:38:32] <etank> cluther: do you need the / on the zDefaultNetworkTree ?
[08-Jul-2009 15:39:06] <cluther> No.. definitely not. One bitmask per line.
[08-Jul-2009 15:39:15] <kisielk> okay I see
[08-Jul-2009 15:39:23] <cluther> If you have a good sized network I recommend 8, 16, 24, 32
[08-Jul-2009 15:39:25] <kisielk> so basically, if I have 8, 16, 22, 24, 32
[08-Jul-2009 15:39:33] <kisielk> it will create those levels of the hierarchy
[08-Jul-2009 15:39:48] <etank> i have a case with /Networks where Zenoss is picking up the wrong mask on some of my boxes
[08-Jul-2009 15:39:57] <cluther> Yes, and if a /25 is found it will still create that breakdown within the proper /24
[08-Jul-2009 15:40:03] <etank> they have dual nics
[08-Jul-2009 15:40:23] <etank> i get something nasty looking like:
[08-Jul-2009 15:40:32] <kisielk> cluther: but in that case, the 16's will be created in the 8, and so on?
[08-Jul-2009 15:40:32] <etank>  /Networks  /172.31.252.0  /172.31.252.0  /172.31.252.0  /172.31.252.0
[08-Jul-2009 15:40:47] <cluther> kisielk: Yes.
[08-Jul-2009 15:41:00] <kisielk> okay, well I will try messing with that then
[08-Jul-2009 15:41:18] <etank> 24 25 24 25 repectively
[08-Jul-2009 15:42:54] <etank> so because of that the network map does not show all of our devices in the subnet
[08-Jul-2009 15:43:06] <cluther> etank: That is bad. Are you on 2.3 or 2.4?
[08-Jul-2009 15:43:19] <etank> cluther: 2.4.2 Core (for now)
[08-Jul-2009 15:43:35] <cluther> hmm.. did you start on that version or upgrade from something earlier?
[08-Jul-2009 15:43:50] <etank> upgrade from 2.4.0 i think
[08-Jul-2009 15:44:08] <etank> 2.4.0 -> 2.4.1 /> 2.4.2
[08-Jul-2009 15:44:28] <cluther> Try going into /Networks and deleting them all. Then remodel just your interfaces by running "zenmodeler run --collect=InterfaceMap" as the zenoss user.
[08-Jul-2009 15:44:44] <cluther> It'll take a bit, but you might have a more sensible network tree once its done.
[08-Jul-2009 15:45:36] <etank> running now
[08-Jul-2009 15:45:57] <etank> did you mean for me to delete just the network that was messed up or all of them cluther
[08-Jul-2009 15:46:09] <cluther> All of them.
[08-Jul-2009 15:46:21] <cluther> They'll be rebuilt.
[08-Jul-2009 15:46:59] <etank> getting rid of just the one did it too
[08-Jul-2009 15:52:31] <etank> cluther: thanks for the help
[08-Jul-2009 15:52:41] <etank> my network is all nice and clean again
[08-Jul-2009 15:52:43] HKhan is now known as Hamzah
[08-Jul-2009 16:00:16] <rmatte> clean network maps are always nice
[08-Jul-2009 16:11:46] ckrough is now known as ckrough_away
[08-Jul-2009 16:16:21] <kisielk> cluther: okay, that works much better now. I just wish I could have /22 and /24 networks at the same level within a /16
[08-Jul-2009 16:16:41] <kisielk> instead of having a bunch of empty /22's in there that just contain /24's
[08-Jul-2009 16:26:59] <cluther> If you took the 22 out of zDefaultNetworkTree and rebuilt again you could.
[08-Jul-2009 16:27:32] <kisielk> but left my existing 22 networks?
[08-Jul-2009 16:27:39] <cluther> yes
[08-Jul-2009 16:27:55] <kisielk> heh, that seems kind of ugly
[08-Jul-2009 16:28:15] <kisielk> it's too bad there's no way to modify the tree after it's created
[08-Jul-2009 16:28:30] <cluther> Yeah, it would be cool if it were dynamic.
[08-Jul-2009 16:36:29] mrayzenoss1 is now known as mrayzenoss
[08-Jul-2009 16:42:55] <bakermd> I want to monitor the toner on my Lexmark Network printers - I have a .mib file for it, is there an easy process to get the data graphed?
[08-Jul-2009 16:43:19] <rmatte> bakermd: there's actually a printer toner Zenpack, not sure if that would be adequate for your printers
[08-Jul-2009 16:43:43] <mrayzenoss> I'm actually checking out some complaints about the PrinterToner ZenPack right now... works for me
[08-Jul-2009 16:43:49] <bakermd> rmatte: Tried that one - and removed it - still cannot add Anything to the /Devices/Printers/Laser class (It killed it )
[08-Jul-2009 16:44:22] <bakermd> mrayzenoss: Totally destroyed /Devices/Printers/Laser on my side
[08-Jul-2009 16:45:18] <rmatte> ok, MIBs are not actually used for polling, only for trapping, you need to figure out what OIDs to poll for the toner and create a template then add them as datapoints
[08-Jul-2009 16:45:22] <rmatte> then graph the datapoints
[08-Jul-2009 16:47:53] <mrayzenoss> yeah, there is that too
[08-Jul-2009 16:48:11] <rmatte> http://ubuntuforums.org/archive/index.php/t-572593.html
[08-Jul-2009 16:48:17] <rmatte> that has some info on Lexmark toner OIDs
[08-Jul-2009 16:48:22] <mrayzenoss> bakermd: hmmmm... there was an issue with an earlier version of the PrinterToner ZenPack
[08-Jul-2009 16:48:34] <rmatte> some of them aren't complete OIDs though, they are translated, you'll need to figure out the full OIDs for the values
[08-Jul-2009 16:48:44] <mrayzenoss> but I'm using 1.40.4
[08-Jul-2009 16:48:45] <kisielk> bakermd: basically, you can just create your own template and bind it to the device. It's pretty simple
[08-Jul-2009 16:49:44] <bakermd> kisielk: Reading a tutorial now that is talking about importing the mibs via `zenmib run`
[08-Jul-2009 16:50:06] <mrayzenoss> bakermd: the MIB Browser ZenPack will make it much easier to install an import MIBs
[08-Jul-2009 16:50:47] <kisielk> bakermd: yeah but mibs won't do anything for you as far as performance templates go
[08-Jul-2009 16:50:58] <bakermd> kisielk: Right
[08-Jul-2009 16:51:10] <rmatte> yeh, he's looking to graph the toner info, so he needs to dig in and find the OIDs then create a template
[08-Jul-2009 16:52:00] <mrayzenoss> which is what the PrinterToner ZenPack does
[08-Jul-2009 16:52:09] <bakermd> mrayzenoss: But it broke my system
[08-Jul-2009 16:52:19] <mrayzenoss> bakermd: 2.4.2?
[08-Jul-2009 16:52:27] <bakermd> mrayzenoss: Installed the mib browser now...
[08-Jul-2009 16:52:36] <bakermd> 2.4.2 = ?
[08-Jul-2009 16:52:42] <mrayzenoss> the version of Zenoss
[08-Jul-2009 16:53:23] <rmatte> Settings -> Versions
[08-Jul-2009 16:53:35] <bakermd> 2.4.2
[08-Jul-2009 16:53:54] <rmatte> under software component versions
[08-Jul-2009 16:54:00] <rmatte> not under Available Zenoss Version
[08-Jul-2009 16:54:05] <rmatte> just to be clear
[08-Jul-2009 16:54:33] <bakermd> Gotcha - Software Component Versions -> Zenoss        Zenoss 2.4.2
[08-Jul-2009 16:54:38] <rmatte> k
[08-Jul-2009 16:58:47] <bakermd> Used Zenmib to add the mibs to the DB - it was successful - however the web interface is showing 0 items inside the mib
[08-Jul-2009 17:02:21] <mrayzenoss> bakermd: I'm still wondering about the failed PrinterToner install
[08-Jul-2009 17:02:41] <bakermd> mrayzenoss: Anything you would like me to try?
[08-Jul-2009 17:03:15] <mrayzenoss> so do a 'zenoss restart', then when it's done go to Settings->ZenPacks and tell me what's listed
[08-Jul-2009 17:04:50] <bakermd> mrayzenoss: Will be just a sec...
[08-Jul-2009 17:04:54] <bakermd> Thanks for the help
[08-Jul-2009 17:13:11] <rmatte> mrayzenoss: thinking it's broken?
[08-Jul-2009 17:14:37] <mrayzenoss> no, it works for me
[08-Jul-2009 17:14:52] <rmatte> no, I mean thinking it's broken on his install, as in didn't get removed properly
[08-Jul-2009 17:14:55] <mrayzenoss> I did recreate the issues in the forum thread
[08-Jul-2009 17:15:00] <mrayzenoss> but I got through them
[08-Jul-2009 17:15:03] <rmatte> ah
[08-Jul-2009 17:15:25] <rmatte> well, my ride is here so I'm out
[08-Jul-2009 17:15:28] <rmatte> cheers guys'
[08-Jul-2009 17:15:47] <mrayzenoss> later
[08-Jul-2009 17:25:17] <bakermd> mrayzenoss: Service has restarted, waiting on the web interface
[08-Jul-2009 17:25:48] <bakermd> For some reason my load average is Really high top - 18:25:44 up 4:13, 1 user, load average: 6.18, 4.62, 3.18
[08-Jul-2009 17:25:54] <mrayzenoss> yikes
[08-Jul-2009 17:26:13] <bakermd> 55% CPU = /usr/local/zenoss/python/bin/.python.bin /usr/local/zenoss/zenoss/lib/python/Zope2/Startup/run.py -C /usr/local/zenoss/zenoss/etc/zope.conf
[08-Jul-2009 17:26:32] <bakermd> Much of the remaining is used by /usr/local/zenoss/python/bin/.python.bin /usr/local/zenoss/zenoss/lib/python/ZEO/runzeo.py -C /usr/local/zenoss/zenoss/etc/zeo.conf
[08-Jul-2009 17:26:49] <mrayzenoss> what are the specs on the box?  How many devices are you monitoring?
[08-Jul-2009 17:27:28] <bakermd> 3GHz Zeon / 4 GB RAM / 6x 15kRPM SCSI drives  monitoring 24 devices
[08-Jul-2009 17:27:52] <jb> somethin definetely wrong there
[08-Jul-2009 17:28:05] <mrayzenoss> heh, I'm doing 15 devices with a netbook
[08-Jul-2009 17:28:20] <bakermd> Recently installed the Cisco MIB pack... took 8 hrs to load!
[08-Jul-2009 17:28:28] <jb> wtf
[08-Jul-2009 17:28:34] <bakermd> thinking I need to remove that - only 2 cisco devices anyway
[08-Jul-2009 17:28:34] <jb> where is that pack btw?
[08-Jul-2009 17:28:36] <mrayzenoss> ahhh... did you see rmatte's comments earlier today?
[08-Jul-2009 17:28:46] <bakermd> mrayzenoss: no...
[08-Jul-2009 17:28:54] <mrayzenoss> I _just_ updated the page http://www.zenoss.com/community/projects/zenpacks/ciscomibs
[08-Jul-2009 17:29:03] <mrayzenoss> you can delete the ZenPack and it'll leave the MIBs
[08-Jul-2009 17:29:10] <mrayzenoss> everything will get much faster
[08-Jul-2009 17:29:23] <bakermd> so delete the zenpack then?
[08-Jul-2009 17:29:36] <mrayzenoss> with the instructions on the page
[08-Jul-2009 17:29:39] <mrayzenoss> via zendmd
[08-Jul-2009 17:29:42] <bakermd> okay
[08-Jul-2009 17:30:06] <bakermd> ooohhh - did not pay attention
[08-Jul-2009 17:30:08] <bakermd> thanks
[08-Jul-2009 17:49:00] <edwardam> Anyone know why a customer 95th percentile VDEF wouldn't work ?
[08-Jul-2009 17:49:05] <edwardam> s/customer/custom/
[08-Jul-2009 19:39:25] <bakermd> Anyone use the NNTP monitor?
[08-Jul-2009 19:40:04] <bakermd> Instructions say to navigate to the device you want to monitor and bind the template, but the template is not in the bind list, despite the egg being installed cleanly
[08-Jul-2009 19:58:21] <bakermd> In All Templates I see things like NtpMonitor, HttpMonitor, JabberMonitor, etc - but no NNTPMonitor
[08-Jul-2009 20:04:43] <etank> bakermd: this may not be it but i think i remember having to restart the daemons with some ZenPacks
[08-Jul-2009 20:05:00] <bakermd> Gotcha - will try that now
[08-Jul-2009 20:05:17] <etank> it may not be the fix though
[08-Jul-2009 20:05:25] <etank> <-- still new to zenoss himself
[08-Jul-2009 20:08:22] <bakermd> I get some errors on the service reload - anyone recognize this problem? http://pastie.org/539476
[08-Jul-2009 20:09:54] <etank> did that only start happening once you installed the NNTP pack?
[08-Jul-2009 20:10:46] <etank> is there a best practice doc somewhere for setting up Systems and Groups
[08-Jul-2009 20:11:02] <etank> as in how best to use each
[08-Jul-2009 20:13:32] <mrayzenoss> etank: Systems and Groups are essentially the same thing under the covers. Groups are usually thought of as following your org chart (ie. QA, Accounting, Marketing, etc.) and Systems follow purpose (Web Farm, Email, Databases, etc.). Devices can belong to multiple Systems and multiple Groups. How you use them is up to you.
[08-Jul-2009 20:14:05] <etank> i think we kind of reversed that at our shop
[08-Jul-2009 20:14:20] <mrayzenoss> it's completely up to you
[08-Jul-2009 20:14:25] <etank> we have systems like TFS, SQL, IIS, etc
[08-Jul-2009 20:14:41] <etank> and groups as Production, Dev, Test, etc
[08-Jul-2009 20:14:48] <mrayzenoss> makes sense
[08-Jul-2009 20:14:58] <etank> i just want it to be logical for anyone using the site
[08-Jul-2009 20:16:12] <etank> mrayzenoss: any thoughts on the best way to monitor 2 nodes of a windows cluster?
[08-Jul-2009 20:17:11] <etank> my thought is to monitor the nodes by name and exclude the stuff (disks, services, etc) that are dedicated to the cluster
[08-Jul-2009 20:17:34] <etank> then monitor the cluster leaving out the stuff that is being monitored by the individual nodes
[08-Jul-2009 20:17:52] <mrayzenoss> so the cluster nodes have a shared IP and individual IPs?
[08-Jul-2009 20:17:55] <etank> we have at least 4 clusters that we want to use
[08-Jul-2009 20:17:59] <etank> yeah
[08-Jul-2009 20:18:26] <etank> node a's ip would be 100 b's 101 and the cluster 102
[08-Jul-2009 20:18:33] <etank> for instance
[08-Jul-2009 20:18:58] <mrayzenoss> I don't know if there's a best practice for cluster monitoring somewhere, I wish there was
[08-Jul-2009 20:19:22] <etank> maybe i can document what works for us so it can help others in the same situation then
[08-Jul-2009 20:19:30] <mrayzenoss> there's an enhancement request somewhere to monitor by MAC address instead of IP, that would resolve some issues
[08-Jul-2009 20:20:14] <etank> i think my biggest issues are going to be monitoring some systems that are behind a vpn firewall
[08-Jul-2009 20:20:35] <etank> and also getting zenoss to graph mssql backup and reindex times
[08-Jul-2009 20:20:54] <mrayzenoss> I've seen you asking about that, I just don't know the right answer.
[08-Jul-2009 20:21:03] <etank> we hope to start our enterprise trial week after next
[08-Jul-2009 20:21:06] <mrayzenoss> I think Chet had a solution he mentioned earlier
[08-Jul-2009 20:21:25] <mrayzenoss> but I think you could modify the MS SQL ODBC one as well
[08-Jul-2009 20:21:29] <etank> then if it works for us on our team zenoss may be used corp wide
[08-Jul-2009 20:21:48] <mrayzenoss> excellent
[08-Jul-2009 20:21:54] <etank> i have the link saved that Chet pasted
[08-Jul-2009 20:22:11] <etank> just trying to figure out how to get the data into the graph
[08-Jul-2009 22:46:34] <Guest4926> hello
[09-Jul-2009 00:00:46] [disconnected at Thu Jul  9 00:00:46 2009]
[09-Jul-2009 00:00:46] [connected at Thu Jul  9 00:00:46 2009]
[09-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[09-Jul-2009 00:16:55] <Guest4926> hello
[09-Jul-2009 00:21:12] <rhetts> hi
[09-Jul-2009 02:54:57] jabular_ is now known as jabular
[09-Jul-2009 04:16:30] <lowe> Anyone else running zenoss on debian from the dev.zenoss.org/deb-repository? Do apt complain about untrusted packages? Tried to add the gpg-key to the apt keychain but it doesn't seem to help
[09-Jul-2009 06:54:53] ckrough_away is now known as ckrough
[09-Jul-2009 06:55:17] <ckrough> morning
[09-Jul-2009 07:05:42] <Dr_Agon> hi all
[09-Jul-2009 07:06:05] <Dr_Agon> I've wrote my own collector plugin. It collects some file systems from asm.
[09-Jul-2009 07:06:21] <Dr_Agon> After modelling I see them in os section of device. But every file system is mapped to FileSystem Template. Can anybody tell me how can i map them to my own template?
[09-Jul-2009 07:06:58] <Dr_Agon> I mean at the modelling time.
[09-Jul-2009 07:09:15] <venturaville> I have one that I just did in the netapp combined zenpack
[09-Jul-2009 07:09:28] <venturaville> I don't think mrayzenoss has sent it out yet though
[09-Jul-2009 07:10:37] <Dr_Agon> erm.. so i can't see your code yet?
[09-Jul-2009 07:10:44] <venturaville> just a sec :-)
[09-Jul-2009 07:11:16] <venturaville> http://nicklay.com/public/ZenPacks.community.NetApp-1.12-py2.4.egg
[09-Jul-2009 07:11:49] <Dr_Agon> thnx!
[09-Jul-2009 07:12:12] <venturaville> take a look at objects.xml
[09-Jul-2009 07:13:07] <Dr_Agon> and one more question: what shoult be the name of variable that contains mount point of file system (to pass it for my ssh cmd in template) ?
[09-Jul-2009 07:13:49] <venturaville> the template does not use the mount point
[09-Jul-2009 07:14:09] <venturaville> your collector plugin does that
[09-Jul-2009 07:14:47] <venturaville> if you want to redo the information at the top of the filesystem view, then it gets way more complicated
[09-Jul-2009 07:14:53] <Dr_Agon> hm, but template defines data points that will collect data, aren't it?
[09-Jul-2009 07:14:59] <venturaville> performance data
[09-Jul-2009 07:15:19] <Dr_Agon> f.e. fs usage
[09-Jul-2009 07:15:25] <venturaville> the stuff at the top is either fetched in the modeler plugin, or it is calculated from that + the performance data
[09-Jul-2009 07:15:31] <venturaville> fs usage = ?
[09-Jul-2009 07:16:02] <Dr_Agon> file system free/used Blocks
[09-Jul-2009 07:16:15] <Dr_Agon> it's not a constant
[09-Jul-2009 07:17:42] <venturaville> usedBlocks in the text field is pulled from the RRD values in the performance data
[09-Jul-2009 07:18:10] <venturaville> totalBlocks is pulled from the modeler
[09-Jul-2009 07:18:18] <venturaville> (which actually irritates me)
[09-Jul-2009 07:18:56] <Dr_Agon> and how shoul i push it into rrd ? i've always thought that it will be pushed there by data point defined in template.
[09-Jul-2009 07:19:11] <venturaville> it should be, but the data point itself needs to be named usedBlocks
[09-Jul-2009 07:19:25] <Dr_Agon> of course
[09-Jul-2009 07:19:53] <Dr_Agon> f.e. i have to file systems after modelling
[09-Jul-2009 07:19:59] <venturaville> I had a time with mine for a while because I named my data source usedBlocks as well (which broke it)
[09-Jul-2009 07:20:18] <venturaville> quick way to test:
[09-Jul-2009 07:20:22] <venturaville> run zendmd
[09-Jul-2009 07:20:25] <venturaville> and do this:
[09-Jul-2009 07:20:32] <venturaville> d = find('mydevicename)
[09-Jul-2009 07:20:50] <venturaville> f = d.os.filesystems.myfilesystemname
[09-Jul-2009 07:20:59] <venturaville> (- is the name of the / filesystem).....
[09-Jul-2009 07:21:19] <venturaville> f.getRRDValue('usedBlocks')
[09-Jul-2009 07:22:31] <Dr_Agon> collected by collector plugin and maped to FileSystemMine (f.e.). In FileSystemMine i've added datapoint that will run ssh cmd. How ssh cmd will understand that it should print usedBlocks info about first or about second file system ?
[09-Jul-2009 07:27:58] <venturaville> in the modeler plugin it relies on an 'snmpindex' to tell it later
[09-Jul-2009 07:29:01] <venturaville> for instance / on one of mine has snmpindex 4
[09-Jul-2009 07:30:23] <Dr_Agon> hmm, ok - i'll read your code carefully 
[09-Jul-2009 07:30:41] <venturaville> I haven't looked at it closely, but I wonder if you could put something other than a number in snmpindex to use later.........
[09-Jul-2009 07:31:38] <venturaville> ./ZenModel/ManagedEntity.py:44:     {'id':'snmpindex', 'type':'string', 'mode':'w'},
[09-Jul-2009 07:31:40] <venturaville> looks like you can
[09-Jul-2009 07:31:42] <venturaville> it is just a string
[09-Jul-2009 07:34:12] <Dr_Agon> well, but it is in case of snmp
[09-Jul-2009 07:34:25] <Dr_Agon> and i have ssh cmd only
[09-Jul-2009 07:34:53] <chris_> ZenModeler: is there a way to initialize a custom property off an device during collection?
[09-Jul-2009 07:35:44] <Dr_Agon> even if i assign snmpindex there is no snmp like output from ssh cmd
[09-Jul-2009 07:36:39] <chris_> Getmap (.../'.1.3.6.1.4.1.231.2.10.2.2.5.10.3.1.28.0': 'cFirmwareVersionCabinet' ..) only works if the custom property cFirmwareVersionCabinet has a value already - but not if this value ist the initial value
[09-Jul-2009 07:38:06] <cluther> You can do this in 2.4, but it requires just a bit of extra work.
[09-Jul-2009 07:38:17] <venturaville> Dr_Agon: the snmpindex is just useful for the cmd to be able to know which filesystem the modeler collected that you need to fetch performance data for
[09-Jul-2009 07:38:27] <cluther> Replace cFirmwareVersionCabinet in your GetMap with _firmwareVersionCabinet
[09-Jul-2009 07:38:56] <venturaville> Dr_Agon: you will obviously need to pass it... the syntax for that is in objects.xml as well
[09-Jul-2009 07:39:28] <cluther> Then in your process method do om.setZenProperty = MultiArgs('cFirmwareVersionCabinet', data['_firmwareVersionCabinet'])
[09-Jul-2009 07:40:08] <cluther> Oh, you'll need to import MultiArgs too: from Products.DataCollector.plugins.DataMaps import MultiArgs
[09-Jul-2009 07:40:21] <Dr_Agon> venturaville: ok, thanks, i'll try
[09-Jul-2009 07:40:27] <venturaville> Dr_Agon: though to pass it the mount point you could do that as well
[09-Jul-2009 07:40:27] <chris_> ok - done in source
[09-Jul-2009 07:41:12] <chris_> wow - i never would have found a solution - I don't even understand this one fully
[09-Jul-2009 07:41:43] <Dr_Agon> venturaville, can i pass like this ${here/mount} ?
[09-Jul-2009 07:41:51] <cluther> cProperties are weird like zProperties in that they live on the hierarchy. If you set them like they're Python properties it causes problems.
[09-Jul-2009 07:42:17] <cluther> So this setZenProperty method takes two parameters, the property name, and its value. It then handles the hierarchical stuff for you.
[09-Jul-2009 07:42:33] <chris_> ah - ok I see thx!!
[09-Jul-2009 07:43:44] <venturaville> Dr_Agon: that looks right
[09-Jul-2009 07:44:08] <chris_> is there a way to dynamically modify the OIDs used in GetMap .... during collection?
[09-Jul-2009 07:44:34] <venturaville> the base modeler ties itself to the Filesystem data source template name I believe though so I think you will need to name your template exactly that.
[09-Jul-2009 07:45:51] <chris_> the problem: blades of fujitsu hardware are appending a number to the OID. something like GetMap ('1.2.3.4.5.5.6.6.*','propertyname') would do it (and * can be any single number)
[09-Jul-2009 07:46:28] <chris_> the alternative would be: look for this index number an replace "*" in the list
[09-Jul-2009 07:46:41] <Dr_Agon> and what if i have node which have standart file systems and mine from asm ?
[09-Jul-2009 07:46:54] <cluther> You can, but it might be easier to use snmpGetTableMaps instead and just pick out the one you want in the process method.
[09-Jul-2009 07:47:10] <chris_> cluther: during collection I get now: om.setZenProperty = MultiArgs('cFirmwareVersionCabinet', data['_firmwareVersionCabinet']) NameError: global name 'data' is not defined
[09-Jul-2009 07:47:13] <venturaville> Dr_Agon: that is a whole nother world of hurt...
[09-Jul-2009 07:47:48] <venturaville> is this oracle asm?
[09-Jul-2009 07:47:52] <Dr_Agon> year =)
[09-Jul-2009 07:47:52] <Dr_Agon> yeah*
[09-Jul-2009 07:47:52] <cluther> chris_: That example made a bunch of assumptions. Could you paste your plugin up somewhere?
[09-Jul-2009 07:47:54] <Dr_Agon> thx
[09-Jul-2009 07:47:59] <Dr_Agon> yup
[09-Jul-2009 07:48:57] <chris_> cluther: http://forums.zenoss.com/viewtopic.php?p=36820#36820
[09-Jul-2009 07:49:32] <cluther> chris_: ok, just use getdata instead of data
[09-Jul-2009 07:49:55] <cluther> om.setZenProperty = MultiArgs('cFirmwareVersionCabinet', getdata['_cFirmwareVersionCabinet'])
[09-Jul-2009 07:49:56] <venturaville> Dr_Agon: I will ask our app guy and see if he has looked at those yet.. he has his own oracle zenpack going
[09-Jul-2009 07:50:25] <chris_> ok - I understand. fog is clearing
[09-Jul-2009 07:50:49] <Dr_Agon> venturaville: cool
[09-Jul-2009 07:51:36] <chris_> _cFirmwareVersionCabinet is the temp var => then it is read from the table => and processed with the zenoss setZenProperty to do the actual assignment to the object instance - right?
[09-Jul-2009 07:52:03] <Dr_Agon> venturaville: look forward to
[09-Jul-2009 07:52:45] <cluther> chris_: Exactly.
[09-Jul-2009 07:58:26] <atrawog> Hi Chris and Cluther: Seems like we are currently doing the Austrian Zenoss User Q&A session
[09-Jul-2009 07:59:13] <atrawog> cluther: Do you have time to look at my Threshold problem too: http://forums.zenoss.com/viewtopic.php?t=10299
[09-Jul-2009 07:59:38] <chris_> hi atrawog - we should go for a coffee sometimes
[09-Jul-2009 08:01:08] <cluther> atrawog: You really need to model the total blocks. You can collect it in an RRD too, but you should model it to threshold against.
[09-Jul-2009 08:01:19] <atrawog> chris_: We have to do that for sure. Are you aware of any other Austrians using Zenoss?
[09-Jul-2009 08:01:31] <cluther> atrawog: Currently there's no way to threshold one data point against another.
[09-Jul-2009 08:02:57] <chris_> yep - I am
[09-Jul-2009 08:03:16] <chris_> maybe we should do something like a Austrian Zenoss USer Group?
[09-Jul-2009 08:03:37] <chris_> without Brüno
[09-Jul-2009 08:04:36] <cluther> Oddly enough one of my colleagues is speaking to a customer in german at the moment too.
[09-Jul-2009 08:05:58] <atrawog> cluther: Problem is that my AIX & NetApp admins are kicking me every time they resize a filesystem after an alert and instead of an clear message they end up getting false alerts until they manually remodel the device.
[09-Jul-2009 08:07:12] <chris_> atrawog: we have the same problem
[09-Jul-2009 08:07:19] <cluther> atrawog: One option would be to collect this data with COMMAND data source and do the percent utilized calculation in there.
[09-Jul-2009 08:10:10] <atrawog> cluther:Thats the next problem they kick me: Measuring long time data growth is a must. So every template we used is based on actual blocks used and not on percent calculations.
[09-Jul-2009 08:11:00] <cluther> atrawog: So have your COMMAND plugin return the totalBlocks, usedBlocks, blockSize and percentUsed
[09-Jul-2009 08:11:13] <cluther> Then you can create whatever graphs or thresholds you need for "now thresholds" and long-term utilization.
[09-Jul-2009 08:11:29] <venturaville> atrawog: the way I do it is simply to turn off the base filesystem threshold ... fetch totalBlocks in the datasource as well, and then call it with cacheRRDValue in my new threshold
[09-Jul-2009 08:12:18] <atrawog> cluther: The trick I'm currently using is defining threshold based on block available insted of blocks used.
[09-Jul-2009 08:12:40] <cluther> That seems smart.
[09-Jul-2009 08:14:52] <venturaville> sneaky ... I like
[09-Jul-2009 08:19:32] <atrawog> venturaville: That's exactly what I'm trying to accomplish. I got as far as defining my own Filesytem class and overloading getTotalBlocks(). Which works flawless in the WebGui, but I haven't managed to make it work with thresholds.
[09-Jul-2009 08:20:51] <venturaville> atrawog: I have numerous issues with the base filesystem class... it doesn't fully implement inodes/files for instance
[09-Jul-2009 08:22:54] <atrawog> cluther: My command parsers scans [r' (?P<totalBlocks>\d+) +(?P<usedBlocks>\d+) ' r'+(?P<availBlocks>\d+) +(?P<percentUsedBlocks>\d+)%']
[09-Jul-2009 08:26:24] <atrawog> venturaville: I found it suprisingly easy to subclass the Zenoss Filesystem and simply overload the bits I would like to change.
[09-Jul-2009 08:28:01] <chris_> cluther: the collector is running without an error - but: the custom property is not set during initial collection - even not if I enter an value manually befor the collection
[09-Jul-2009 08:28:02] <venturaville> I would love to see the code ....
[09-Jul-2009 08:31:20] <atrawog> ventureville: No problem. I spend some time preparing for the chat today: http://code.google.com/p/atrawog-zenoss-community/source/browse/zenpacks/ZenPacks.atrawog.HelloLinux/ZenPacks/atrawog/HelloLinux/zenmodel/LinuxFileSystem.py
[09-Jul-2009 08:33:55] <venturaville> I'll take a look at it on the express bus into work
[09-Jul-2009 08:34:01] <venturaville> later :-)
[09-Jul-2009 08:48:04] <MuseIT> looking to find out some more about L3 topology alerts. is what i see on the netmap what is it using?
[09-Jul-2009 08:49:08] <atrawog> chris: Any suggestions when and where to have the first AZUG (Austrian Zenoss User Group) meeting?
[09-Jul-2009 08:49:29] <MuseIT> also can there be a parent/child setup? i have a Virtual bearmetal system and i would like just one aler for it. no all the VMs on it.
[09-Jul-2009 09:05:49] <rocket> morning matt and chet
[09-Jul-2009 09:06:32] <atrawog> mrayzenoss: Thanks for the Zenoss 2.4 beta t-shirt. I got mine yesterday
[09-Jul-2009 09:07:22] <mrayzenoss> morning everybody, hanging out in the defect review and I'll be back in a bit
[09-Jul-2009 09:11:31] <Dieterbe> Hi, for windows servers does one _need_ snmp-informant or should it be possible to monitor cpu/memory/.. and per-process memory usage with the built-in snmp stuff in windows?
[09-Jul-2009 09:12:05] <Dieterbe> (i've checked the zenoss admin manual but didn't find the explanation too clear)
[09-Jul-2009 09:12:08] <ckrough> I think WMI is The Way (tm) with windows
[09-Jul-2009 09:12:57] <atrawog> Dieterbe: You can try using WMI instead of SNMP
[09-Jul-2009 09:13:47] <Dieterbe> aha, well i have WMI enabled, filled in the account stuff, but i'm not getting graphs (the (in)famous 'MISSING RRD FILE' problem)
[09-Jul-2009 09:13:48] <atrawog> Dieterbe: But SNMP is much more reliable and will use lesser resources on you Zenoss server.
[09-Jul-2009 09:14:42] <Dieterbe> hmm well i think the cleanyness of the "to be monitored" windows servers is in this case more important than the performance of the zenoss system (only 3 windows devices btw)
[09-Jul-2009 09:15:19] <Dieterbe> i just need to be able to monitor "the usual stuff" and also per-process memory usage
[09-Jul-2009 09:16:58] <atrawog> Dieterbe: Problem with WMI is that you "sometimes" end up getting connection errors and it's difficult to figure out why
[09-Jul-2009 09:17:21] <Dieterbe> but they recover automatically within.. say 10 minutes?
[09-Jul-2009 09:19:59] <atrawog> Dieterbe: They normaly recover themself, but they do make alerting a bit problematic.
[09-Jul-2009 09:20:32] <Dieterbe> i see. i'll keep it in mind
[09-Jul-2009 09:21:20] <Dieterbe> first i would like to get this running though. any idea why i wouldn't see any graphs even though wmi is enabled and the user/pass are filled in the zproperties (username prefixed by '.\') ?
[09-Jul-2009 09:21:38] <Dieterbe> also snmp should work, at least the snmpwalk works
[09-Jul-2009 09:21:54] <Dieterbe> maybe i should try to remodel explicitly using wmi?
[09-Jul-2009 09:22:37] <atrawog> Dieterbe: Does fetching the Windows event log work OK?
[09-Jul-2009 09:23:20] <Dieterbe> how do i do that? you don't mean the 'events' tab i suppose?
[09-Jul-2009 09:27:01] <atrawog> Dieterbe: If WMI is set up correctly ZenEventlog should start fetching your Windows event log http://www.zenoss.com/community/docs/zenoss-guide/2.3.3/ch16s06.html
[09-Jul-2009 09:30:25] <atrawog> Dieterbe: The full ZenWinPerf windows performance daemon is only available in Zenoss Professional or Enterprise ttp://www.zenoss.com/community/docs/zenoss-guide/2.3.3/ch17s01.html
[09-Jul-2009 09:31:32] <Dieterbe> atrawog: okay Main views -> Event console shows 8 entries. the evetclasses are /Status/Snmp and /Perf/-Filesystem
[09-Jul-2009 09:32:30] <Dieterbe> atrawog: but i don't need that zenpack, rigt?
[09-Jul-2009 09:32:34] <Dieterbe> right*
[09-Jul-2009 09:35:40] <atrawog> Dieterbe: What kind of hardware are you using? Most hardware vendors offer SNMP agents and we pretty successfully use the combination of Zenoss + HP Insight Agent
[09-Jul-2009 09:36:08] <MuseIT> where can i find a list of the vars like ${device/id} I'm trying to make a wimTest command and need zWinUser and passwords
[09-Jul-2009 09:36:16] <rmatte> yeh, HP Insight Agent is quite nice for detecting hardware issues
[09-Jul-2009 09:37:37] <rmatte> MuseIT: what are you scripting the command in?
[09-Jul-2009 09:37:56] <tehhobbit> rmatte: you use it, I cant seem to get it to respond propperly over snmp
[09-Jul-2009 09:38:11] <MuseIT> just adding a command to the zenoss drop down. so shell i guss
[09-Jul-2009 09:38:33] <rmatte> and you need to pass the zwinpass and zwinuser to it?
[09-Jul-2009 09:39:33] <rmatte> have you tried just ${device/zWinUser} ?
[09-Jul-2009 09:39:40] <MuseIT> yep
[09-Jul-2009 09:40:07] <rmatte> I know that the WMIPerformanceMonitor ZenPack passes those variable in it's template, I forget exactly how it passes them
[09-Jul-2009 09:40:23] <rmatte> let me load it up on my lab box and check
[09-Jul-2009 09:40:38] <MuseIT> something like wimc -U ${zUser}%${zwinpass} //${device/manageip} "select *........
[09-Jul-2009 09:40:59] <MuseIT> thank you
[09-Jul-2009 09:42:26] <MuseIT> just got it
[09-Jul-2009 09:42:31] <rmatte> k cool
[09-Jul-2009 09:42:49] <MuseIT> i didn't have the device/zWinUser
[09-Jul-2009 09:43:05] <rmatte> ah
[09-Jul-2009 09:44:07] <MuseIT> i got kicked
[09-Jul-2009 09:44:16] <MuseIT> bad inet connetion here.
[09-Jul-2009 09:45:23] <rmatte> hehe
[09-Jul-2009 09:45:34] <MuseIT> so my command looks like this now. wmic -U ${device/zWinUser}%{device/zWinpassword} //${device/manageIp} "select * from Win32_ComputerSystem"
[09-Jul-2009 09:46:02] <MuseIT> and it's working. thanks for the help.
[09-Jul-2009 09:46:17] <rmatte> no problem
[09-Jul-2009 09:58:58] <bwest> hello
[09-Jul-2009 10:01:34] <bwest> Hello all
[09-Jul-2009 10:02:40] <bwest> jason has this been canceled?
[09-Jul-2009 10:02:45] <Jane_Curry> Greetings bwest
[09-Jul-2009 10:03:03] <bwest> ok now people are showing up hello Jane
[09-Jul-2009 10:03:22] <bwest> Excellent work on the papers Jane
[09-Jul-2009 10:03:34] <Jane_Curry> Thanks
[09-Jul-2009 10:03:46] <Jane_Curry> Let me start with a goglie ball.....
[09-Jul-2009 10:04:19] <Jane_Curry> If I am creating a ZenPack with my own object classes..
[09-Jul-2009 10:05:34] <Jane_Curry> The 2.4 Dev Guide says (p71) "classes should be subclassed from ZenModel and objects initialised from within our ZenPack's --Init__.py
[09-Jul-2009 10:06:04] <Jane_Curry> I cannot see anything in any of the ZenPack __init__.py files (I think I've tried them all)
[09-Jul-2009 10:06:35] <bwest> I haven't seen anything either
[09-Jul-2009 10:06:49] <Jane_Curry> I set the object class for a device class by setting the zProperty zPythonClass to my object class..
[09-Jul-2009 10:07:15] <Jane_Curry> .. like ZenPacks.skills1st.bridge.BridgeDevice
[09-Jul-2009 10:07:27] <bwest> mray: this is an easy one I Cannot clear events from the events console? Can i force a clear from commandline?
[09-Jul-2009 10:07:30] <atrawog> Jane_Curry: What everybody seams to do is: from Globals import InitializeClass InitializeClass(LinuxFileSystem)
[09-Jul-2009 10:07:33] <__ian__> What exactly are you trying to do?
[09-Jul-2009 10:08:01] <mrayzenoss1> bwest: how can you not clear them? Do you not have the proper permissions as the current user?
[09-Jul-2009 10:08:29] <__ian__> Jane_Curry: That is, what's the symptom?
[09-Jul-2009 10:08:43] <bwest> mray: I am the admin. I check the events I want to move to history and and click the move to history option and nothing happens
[09-Jul-2009 10:09:20] <__ian__> Jane_Curry: What atrawog said, really. You don't have to do it in __init__.py...if I'm understanding you correctly.
[09-Jul-2009 10:10:03] <mrayzenoss1> bwest: very strange.  Which version and which browser?
[09-Jul-2009 10:10:15] <bwest> 2.4.1 firefox
[09-Jul-2009 10:10:40] <bwest> mray: I have restarted zenoss serveral times to no avail
[09-Jul-2009 10:10:47] <Jane_Curry> OK - I have the import InitializwClss and the end of my object class file has InitializeClass(BridgeDevice)
[09-Jul-2009 10:10:59] <Jane_Curry> Are we saying that that is all I need??
[09-Jul-2009 10:11:07] <__ian__> Not sure exactly what you're trying to do
[09-Jul-2009 10:11:24] <Jane_Curry> Do I NOT need to set the zPythonClass??
[09-Jul-2009 10:11:50] <__ian__> Is your goal to get the class associated with a particular DeviceClass?
[09-Jul-2009 10:12:31] <Jane_Curry> __ian__ I am trying to create a new device class called BridgeDevice which has a component class, BridgeInterface, which will show information for all ports on a Bridge (switch really)
[09-Jul-2009 10:12:38] <mrayzenoss1> bwest: I don't have any good suggestions really. Log out, stop zenoss, start zenoss, clear your browser cache, log back in as the 'admin'
[09-Jul-2009 10:12:49] <__ian__> That paragraph in the dev guide is confusing. I don't understand what it's saying.
[09-Jul-2009 10:13:03] <Jane_Curry> Oh good - thought it was me....
[09-Jul-2009 10:13:03] <__ian__> Right, and as long as you've created the class you have succeeded.
[09-Jul-2009 10:13:05] <mrayzenoss1> bwest: you can move them in mysql, but that's besides the point
[09-Jul-2009 10:13:34] <bwest> mray:There is no command I can use from the command line to delete all events?
[09-Jul-2009 10:13:49] <__ian__> If you want it to be used in a particular DeviceClass, yeah, you set the value of zPythonClass
[09-Jul-2009 10:13:58] <__ian__> on the DeviceClass
[09-Jul-2009 10:14:23] <Jane_Curry> __ian__ yup - the  goal is to get the class associated with a particular DeviceClass
[09-Jul-2009 10:14:29] <atrawog> Jane_Curry: You have to set the zPythonClass. But up to my knowledge the class won't work without initializing it via InitializeClass()
[09-Jul-2009 10:14:35] <mrayzenoss1> bwest: let me check
[09-Jul-2009 10:14:44] <ckrough> Has there been any work on a unified event console for bringing together events from multiple installation sites?
[09-Jul-2009 10:14:59] <__ian__> atrawog: That's sort of true. It'll probably work fine; InitializeClass() is a Zope thing that registers it so you can add it through the ZMI
[09-Jul-2009 10:15:16] <bwest> ckrough:I know that is an enterprise feature
[09-Jul-2009 10:15:37] <__ian__> Jane_Curry: What you want is a simple migrate script to make sure the property is set to the correct value.
[09-Jul-2009 10:15:47] <ckrough> bwest: it is? I'm an Enterprise user, havent seen that. Is this recent?
[09-Jul-2009 10:15:52] <Jane_Curry> .. so the good practice seems to be to set both the zPythonClass and to Initialize
[09-Jul-2009 10:16:32] <Jane_Curry> __ian__ - not sure what you mean about "a simple migrate script"
[09-Jul-2009 10:16:39] <bwest> ckrough:I should have said I think, I thought that was the whole point of the global dashboard
[09-Jul-2009 10:16:40] <__ian__> Jane_Curry: Actually, you don't even need a migrate script, since you're creating the device class in this pack, right?
[09-Jul-2009 10:16:57] <ckrough> bwest: Ill review the later versions again. Im still running 2.3.3
[09-Jul-2009 10:17:27] <__ian__> Jane_Curry: Migrate script was red herring. All you have to do is set the property value on your device class.
[09-Jul-2009 10:17:29] <bwest> ckrough: try looking at the info for 2.4.1 and later
[09-Jul-2009 10:17:48] <ckrough> bwest: heh, even 2.3.3 http://www.zenoss.com/community/docs/zenoss-guide/2.3.3/ch26s01.html
[09-Jul-2009 10:17:53] <ckrough> bwest: thanks
[09-Jul-2009 10:18:05] <Jane_Curry> got it - that's what I have but I was confused by the comments about __init__.py in the dev guide
[09-Jul-2009 10:18:20] <bwest> cluther:mray is looking for me but Can i force a clear from commandline?
[09-Jul-2009 10:18:31] <__ian__> Jane_Curry: Well, there's lots of stuff that can happen in __init__.py.
[09-Jul-2009 10:18:42] mrayzenoss1 is now known as mrayzenoss
[09-Jul-2009 10:18:58] <Jane_Curry> __ian__ but do I control it and do I care??????
[09-Jul-2009 10:19:11] <__ian__> Jane_Curry: I believe what the dev guide is talking about is the case where you want to add a new zProperty
[09-Jul-2009 10:19:13] <ckrough> OK, next question/request... is there any existing zenpacks/functionality to create a summary alert when a specified event threshold is reached? ie; rather than sending 2000 emails, can a single email be sent summarizing the crisis once N events are generated in under X seconds?
[09-Jul-2009 10:19:45] <bwest> ckrough use the filter for alerts
[09-Jul-2009 10:20:29] <bwest> cluther: mray is looking but I cannot Delete events from my event console. Can i force a clear from commandline?
[09-Jul-2009 10:20:38] <__ian__> Jane_Curry: if you dump something like this in the __init__.py of a ZenPack, it'll create new zProperties for you: http://pastebin.ca/1489726
[09-Jul-2009 10:20:50] <__ian__> Jane_Curry: But no, in the instant case I believe you don't care.
[09-Jul-2009 10:21:12] <mrayzenoss> bwest: mysql -uzenoss -pzenoss events -e "delete from status"
[09-Jul-2009 10:21:26] <__ian__> mrayzenoss: whoa there
[09-Jul-2009 10:21:31] <bwest> thx
[09-Jul-2009 10:21:42] <mrayzenoss> bwest: that will delete _ALL_ current events
[09-Jul-2009 10:21:44] <__ian__> mrayzenoss: nm, was thinking of history.
[09-Jul-2009 10:21:55] <bwest> yes i want to delete those to
[09-Jul-2009 10:22:02] <mrayzenoss> bwest: change status -> history to delete those
[09-Jul-2009 10:22:05] <ckrough> bwest: regarding the filter, I am referring to multiple device events, not a single device. Like when a row or two of a DC is lost and path outage doesnt apply, I get thousands of alerts when I could just use one that says "omg things are bad"
[09-Jul-2009 10:22:09] <__ian__> mrayzenoss: not quite
[09-Jul-2009 10:22:21] <__ian__> mrayzenoss: Deleting from history is not as simple as that
[09-Jul-2009 10:22:33] <mrayzenoss> __ian__: hmmm... I just wiped my history out with it
[09-Jul-2009 10:22:38] <__ian__> mrayzenoss: Because it needs to clean up the log and details tables too
[09-Jul-2009 10:22:56] <mrayzenoss> __ian__:  yeah
[09-Jul-2009 10:23:10] <mrayzenoss> bwest: also wipe out 'log' 'alert_state' 'detail"
[09-Jul-2009 10:23:27] <bwest> mray:with the same command
[09-Jul-2009 10:23:29] <mrayzenoss> bwest: here's the schema, go nuts: http://dev.zenoss.org/trac/browser/branches/zenoss-2.4.x/Products/ZenEvents/db/zenevents.sql
[09-Jul-2009 10:26:04] <__ian__> ckrough: I don't think that functionality exists yet.
[09-Jul-2009 10:26:08] <bwest> another question On the event console I want to add the field duration: that shows the time the event has held this status
[09-Jul-2009 10:26:22] <ckrough> __ian__: thanks
[09-Jul-2009 10:26:41] <__ian__> ckrough: I'll make a ticket for that enhancement. Sure would be cool.
[09-Jul-2009 10:27:22] <bwest> is that possible
[09-Jul-2009 10:28:07] <__ian__> bwest: Possible but not simple. That'll take some code.
[09-Jul-2009 10:28:14] <Jane_Curry> Thanks __ian__ - next one........
[09-Jul-2009 10:28:42] <mrayzenoss> bwest: so you have firstTime, lastTime and stateChange
[09-Jul-2009 10:29:26] <bwest> mray: if by stateChange you mean the time that the event has been in a particular state
[09-Jul-2009 10:29:30] <bwest> then yes
[09-Jul-2009 10:29:34] <mrayzenoss> bwest: and the default expiration for events is 4 hours
[09-Jul-2009 10:29:50] <mrayzenoss> bwest: stateChange is the time of the last change
[09-Jul-2009 10:29:53] <mrayzenoss> not duration
[09-Jul-2009 10:30:28] <bwest> I would like a field that shows how long an event has been in a it's particular state to replace the count field
[09-Jul-2009 10:30:34] <bwest> just like in nagios
[09-Jul-2009 10:31:10] <Jane_Curry> I have a method in my object definition file in my ZenPack that takes a local attribute (RemoteAddress) and then searches through the ZODB database for any devices that match this RemoteAddress (its a MAC address)
[09-Jul-2009 10:31:19] <mrayzenoss> bwest: you could change your sorting to be by firstTime
[09-Jul-2009 10:31:54] <mrayzenoss> bwest: the events wouldn't leave the console unless 4 hours had passed with no changes
[09-Jul-2009 10:32:03] <mrayzenoss> you'd have to do the math in your head though...
[09-Jul-2009 10:32:28] <mrayzenoss> you can also change the 4 hour Event Aging Threshold
[09-Jul-2009 10:32:45] <shakespear> cluther: the collector is running without an error - but: the custom property is not set during initial collection - even not if I enter an value manually befor the collection
[09-Jul-2009 10:32:46] <bwest> mray: yes but that doesn't quite achieve my goal. I want to look at an event and have a field that shows how long this device has had this problem.
[09-Jul-2009 10:33:04] <mrayzenoss> yeah, what __ian__ said then
[09-Jul-2009 10:33:12] <shakespear> cluther: it seems om.setZenProperty = MultiArgs('cFirmwareVersionCabinet', getdata['_cFirmwareVersionCabinet']) doesn't have the desired effect
[09-Jul-2009 10:33:16] <mrayzenoss> you could open an enhancement request
[09-Jul-2009 10:33:25] <bwest> ok.  thx
[09-Jul-2009 10:33:39] <__ian__> bwest: Something similar to that is planned for the next release
[09-Jul-2009 10:33:55] <mrayzenoss> yeah, the Event Console is getting some love from Ian for King Crab
[09-Jul-2009 10:34:01] <bwest> perfect
[09-Jul-2009 10:34:30] <Jane_Curry> ... what I want to do is to use the same code in my modeler plugin to populate a RemoteIpAddress attribute ion my new object class....
[09-Jul-2009 10:35:15] <bwest> _ian_: how does zPingInterfaceDescription and zPingInterfaceName work? will this allow me to create events off interfaces being down?
[09-Jul-2009 10:35:33] <Jane_Curry> ... but I can't find the right way to setup dmd - here is what I have in the object class file...
[09-Jul-2009 10:36:03] <Jane_Curry> def getIpRemoteAddress(self):
[09-Jul-2009 10:36:07] <Jane_Curry>         dmd=self.dmd
[09-Jul-2009 10:36:11] <Jane_Curry>         devmac=self.RemoteAddress
[09-Jul-2009 10:36:11] <Jane_Curry>         IpAddress=[]
[09-Jul-2009 10:36:11] <Jane_Curry>         Ips=dmd.ZenLinkManager.layer2_catalog(macaddress=devmac)
[09-Jul-2009 10:36:11] <Jane_Curry>         for i in Ips:
[09-Jul-2009 10:36:11] <Jane_Curry>             IpAddress=IpAddress + [i.getObject().manageIp]
[09-Jul-2009 10:36:12] <Jane_Curry>         return IpAddress
[09-Jul-2009 10:37:04] <Jane_Curry> ... "self" in this case is the class for the modeler plugin, not the class for the object itself...
[09-Jul-2009 10:37:48] <Jane_Curry> ... so it all barfs out in the modeler plugin
[09-Jul-2009 10:37:56] <__ian__> ah, right. Hang on, lemme check
[09-Jul-2009 10:39:45] <bwest> mray: is there a specific place where patches for a particular zenoss version are located
[09-Jul-2009 10:40:17] <mrayzenoss> bwest: yeah, http://dev.zenoss.org/trac/report
[09-Jul-2009 10:40:39] <__ian__> Jane_Curry: in what method on the modeler plugin are you trying to do this?
[09-Jul-2009 10:40:42] <mrayzenoss> Current Maintenance Release (All) is stuff getting fixed for 2.4.3
[09-Jul-2009 10:41:07] <mrayzenoss> bwest: so if you see something in there, with 'closed', you could grab the patch from the 2.4.x branch and apply it
[09-Jul-2009 10:41:09] <__ian__> Jane_Curry: All you need is access to some database object. If you have that, it'll get dmd by acquisition (so you could just call, for example, device.dmd)
[09-Jul-2009 10:42:00] <mrayzenoss> bwest: and whenever 2.4.3 comes out, it'll roll up those patches
[09-Jul-2009 10:42:56] <bwest> mray: I have seen alot of people having issues with upgrading to 2.4.2 I am running 2.4.1 is worth just waiting to 2.4.3
[09-Jul-2009 10:43:15] <mrayzenoss> 2.4.3 won't be out for awhile
[09-Jul-2009 10:43:56] <mrayzenoss> I think the goal for maintenance releases is usually 6-8 weeks
[09-Jul-2009 10:44:00] <mrayzenoss> 2.4.1 was an exception
[09-Jul-2009 10:44:11] <mrayzenoss> since the maps were broken
[09-Jul-2009 10:44:31] <bwest> mray: so it is worth upgrading to 2.4.2
[09-Jul-2009 10:44:47] <mrayzenoss> I would say yes, there are 50 some tickets closed
[09-Jul-2009 10:44:49] <etank> anyone reported yet that the google maps dont work in Chrome?
[09-Jul-2009 10:45:13] <rmatte> I think there's a ticket for Chrome in general
[09-Jul-2009 10:45:14] <mrayzenoss> Last time I used Chrome they did, is that new?
[09-Jul-2009 10:45:14] <etank> which seems really odd to me since both maps and Chrome are Google
[09-Jul-2009 10:45:25] <rmatte> kind of funny that google maps don't work in google's own browser
[09-Jul-2009 10:45:31] <__ian__> Yeah, maps are working in Chrome for me right now
[09-Jul-2009 10:45:33] <etank> i tried using chrome yesterday and it didnt work
[09-Jul-2009 10:45:35] <Jane_Curry> __ian__ Tried this dmd=device.dmd in the plugin code and get AttributeError: DeviceProxy instance has no attribute 'dmd' on remodel
[09-Jul-2009 10:46:10] <etank> i get just a big ugly gray box when in chrome
[09-Jul-2009 10:46:41] <bwest> mray:once last question to apply a specific patch what is the command
[09-Jul-2009 10:46:57] <__ian__> Jane_Curry: This might be a fruitless endeavor. It depends what you're getting passed in, since the plugins aren't database objects.
[09-Jul-2009 10:47:09] <Jane_Curry> __ian__ this code is simply part of the def process
[09-Jul-2009 10:47:22] <mrayzenoss> bwest: zenpatch revistion_number
[09-Jul-2009 10:47:58] <__ian__> etank: what version of zenoss? It's working in the trunk, at least.
[09-Jul-2009 10:48:10] <mrayzenoss> bwest: so in the ticket there's probably something like (In 14180) Fixes #5279
[09-Jul-2009 10:48:19] <bwest> mray:how does zPingInterfaceDescription and zPingInterfaceName work? will this allow me to create events off interfaces being down?
[09-Jul-2009 10:48:27] <mrayzenoss> bwest: so you'd use 'zenpatch 14180'
[09-Jul-2009 10:48:31] <etank> __ian__: 2.4.2 Core
[09-Jul-2009 10:48:32] <Jane_Curry> __ian__ Yeh - I think I know what you're saying - wanted to have a go at it though
[09-Jul-2009 10:49:35] <__ian__> Jane_Curry: It can probably be done, but at a different point in the process, and then you get into writing a special HubService, which has a database connection...it gets pretty tricky.
[09-Jul-2009 10:49:44] <etank> __ian__: this is what i get in chrome http://img224.imageshack.us/img224/4053/chrome.jpg
[09-Jul-2009 10:49:56] <__ian__> etank: and it works in FF or Safari?
[09-Jul-2009 10:50:06] <etank> FF and IE 8 yes
[09-Jul-2009 10:50:11] <mrayzenoss> etank: what happens if you resize the window?
[09-Jul-2009 10:50:12] <Jane_Curry> __ian__ Hmmm - not sure I want to try THAT hard
[09-Jul-2009 10:50:12] <etank> i dont have Safari loaded
[09-Jul-2009 10:50:14] <__ian__> etank: Have you upgraded recently?
[09-Jul-2009 10:50:30] <etank> __ian__: not since 2.4.2 came out
[09-Jul-2009 10:50:36] <__ian__> etank: Possibly a browser cache problem?
[09-Jul-2009 10:50:46] <etank> hmm
[09-Jul-2009 10:50:55] <etank> mrayzenoss: same gray box.
[09-Jul-2009 10:51:10] <etank> mrayzenoss: i assume you meant the maps portlet resizing
[09-Jul-2009 10:51:13] <__ian__> etank: Maps haven't changed between 2.4 and the trunk.
[09-Jul-2009 10:51:22] <__ian__> etank: So I'm guessing it's not a code issue.
[09-Jul-2009 10:51:50] <mrayzenoss> etank: yeah, occasionally I get a weird render with FF of the map, just resizing the application a tiny bit redraws it
[09-Jul-2009 10:51:52] <etank> clearing cache now to try again
[09-Jul-2009 10:52:31] <etank> i did have an issue where if i put in the full address to where i am located it showed that i am in the ocean east of Greenland
[09-Jul-2009 10:52:42] <etank> thats pretty far from Kentucky
[09-Jul-2009 10:52:55] <etank> had to tweak the address to get it just right
[09-Jul-2009 10:53:18] <Jane_Curry> I am still using Firefox 2 (2.0.0.19) and the interface is sometimes very jerky and also resource intensive
[09-Jul-2009 10:53:32] <Jane_Curry> Any comments from anyone who has compared FF2 with FF3 ??
[09-Jul-2009 10:53:46] <etank> clearing cache did not make a difference __ian__
[09-Jul-2009 10:53:56] <etank> i will just stick to FF for now
[09-Jul-2009 10:53:58] <mrayzenoss> Jane_Curry: I'm not sure we're still supporting FF2
[09-Jul-2009 10:54:03] <bwest> Jane: Can you answer this question for me zPingInterfaceDescription and zPingInterfaceName work? will this allow me to create events off interfaces being down in SSH?
[09-Jul-2009 10:54:07] <__ian__> Jane_Curry: FF3 has a much faster javascript engine.
[09-Jul-2009 10:54:13] <Jane_Curry> Zenoss is currently 2.4.1
[09-Jul-2009 10:54:35] <venturaville> ff.3.5 is really fast
[09-Jul-2009 10:56:48] <__ian__> bwest: I'm looking into those zProperties
[09-Jul-2009 10:57:02] <bwest> _ian_:ok thx
[09-Jul-2009 10:57:28] <jb> hi all..
[09-Jul-2009 10:57:34] <bwest> mray:how long did it take to clear your databdase tables my has been taking 30 minutes and still running
[09-Jul-2009 10:57:57] <bwest> mray:my history table was full
[09-Jul-2009 10:59:40] <mrayzenoss> bwest: mine was fast, small install
[09-Jul-2009 10:59:50] <__ian__> bwest: It's just mysql deletion. Depends on resources and number of rows.
[09-Jul-2009 10:59:56] <__ian__> bwest: Those zProperties are no longer used.
[09-Jul-2009 11:00:40] <bwest> _ian_: ok so that means I need to create a zenpack to monitor the status of the interfaces on SSH servers
[09-Jul-2009 11:03:01] <bwest> _ian_:Or more specific I need events created when interfaces go down and right now I only see events from the main host. although I have enabled monitoring on the interfaces
[09-Jul-2009 11:07:05] <daleamon> a quick dumb question... do I presume correctly that the 2.4.2 http://www.zenoss.com/community/docs/zenoss-guide is specific to a vendor version and not the debian one?
[09-Jul-2009 11:07:27] <daleamon> my initial screen looks nothing like the one in the manual, ie not a google map.
[09-Jul-2009 11:08:01] <__ian__> bwest: http://dev.zenoss.org/trac/ticket/5291
[09-Jul-2009 11:08:29] <__ian__> bwest: What does the interface template look like?
[09-Jul-2009 11:11:02] <mrayzenoss> daleamon: we try to make the documentation generic enough to cover all the platforms.
[09-Jul-2009 11:11:09] <mrayzenoss> daleamon: what are you seeing?
[09-Jul-2009 11:12:22] <bwest> _ian_: for snmp I use the OID sys up and set the threshold to 1 but I don't always get alerts
[09-Jul-2009 11:12:23] <daleamon> I have a dashboard with 4 quadrants, production state, Object watch list, device issues and messages.
[09-Jul-2009 11:12:47] <bwest> _ian_: for ssh I would guess I need a zenpack
[09-Jul-2009 11:13:17] <daleamon> Whereas the documentation fig 2.1 while claiming to be the same version looks like something much more advanced.
[09-Jul-2009 11:13:45] <mrayzenoss> daleamon: you can add and delete portlets for dashboard as well as change the layout
[09-Jul-2009 11:13:48] <mrayzenoss> daleamon: link?
[09-Jul-2009 11:13:51] <daleamon> I am assuming the fact that the debian package is 2.4.2 means it really is.
[09-Jul-2009 11:14:22] <daleamon> You mean the doc link? I poasted that up above.
[09-Jul-2009 11:15:08] <mrayzenoss> daleamon: ok, you can get that by clicking 'Configure layout..."
[09-Jul-2009 11:15:26] <mrayzenoss> pick the 2 columns
[09-Jul-2009 11:15:33] <daleamon> I'm sure you can configure it, but it would be useful for learning if what you see after the initial setup is same as what is in the docs you are using to learn it.
[09-Jul-2009 11:15:35] <bwest> _ian_ or mray:I just got this error when I ran the delete history command ERROR 1114 (HY000) at line 1: The table 'history' is full
[09-Jul-2009 11:16:00] <daleamon> I'll try that, but perhaps this is a debian problem that they selected a different default from the standard docs?
[09-Jul-2009 11:16:18] <mrayzenoss> nah, you're right.  The screenshot doesn't match the initial page for any install
[09-Jul-2009 11:17:31] <daleamon> thanks for the info. What happens when they do not match is that the new zenoss student will not trust that they have the correct docs
[09-Jul-2009 11:18:24] <mrayzenoss> yeah, I'll open a ticket
[09-Jul-2009 11:19:00] <daleamon> Thanks much. I will now go back to reading, secure in the knowledge that whatever I see actually exists :-^
[09-Jul-2009 11:20:29] <mrayzenoss> I'm going to go ahead and let __ian__ off the hook, thanks to everyone who stopped by. I'll have the transcript up later today.
[09-Jul-2009 11:23:53] <Jane_Curry> Many thanks.
[09-Jul-2009 11:23:55] <__ian__> bwest: Yeah, for ssh you need a zenpack. It's in enterprise.
[09-Jul-2009 11:24:43] <__ian__> bwest: You have a non-zenoss-related MySQL problem I don't know how to fix off the top of my head.
[09-Jul-2009 11:24:49] <__ian__> Later, all.
[09-Jul-2009 11:28:39] <mrayzenoss> bwest: There are several SSH community ZenPacks, which platform do you need?
[09-Jul-2009 11:39:33] <perr0> can someone ping when matt comes in the room please
[09-Jul-2009 11:39:57] <mrayzenoss> perr0: I'm here
[09-Jul-2009 12:15:08] <mrayzenoss> daleamon: http://dev.zenoss.org/trac/ticket/5292
[09-Jul-2009 12:21:05] <daleamon> thanks!
[09-Jul-2009 12:37:32] <daleamon> heh... after reading the google click-wrap I can well understand why Debian did not include it as default! Yech. Sharekespeare had the right idea about lawyers...
[09-Jul-2009 12:56:49] <rmatte> Debian did the same thing with firefox until they modified it
[09-Jul-2009 12:57:21] <rmatte> they were debating including iceweasel instead
[09-Jul-2009 13:00:08] <mrayzenoss> I expect once we start trying to get into Debian we'll have to remove the Network Map (Flash) and Google Maps and put them in a non-free .deb
[09-Jul-2009 13:04:47] <rmatte> probably
[09-Jul-2009 13:05:00] <rmatte> as long as they are available though, no big deal
[09-Jul-2009 13:07:22] <daleamon> Again, thanks all. I have to run, pick someone up and goto a customer site now...
[09-Jul-2009 13:22:02] <ckrough> mrayzenoss: It seems that ZenGlobe disappears from the documentation after 2.2.4. Do you know if that was that dropped or moved somewhere else?
[09-Jul-2009 13:22:41] <mrayzenoss> ckrough: Extended Monitoring Guide, chapter 52
[09-Jul-2009 13:22:49] <ckrough> mrayzenoss: thanks again
[09-Jul-2009 13:22:57] <mrayzenoss> the ZenPacks moved into their own doc
[09-Jul-2009 13:23:00] <mrayzenoss> with 2.4
[09-Jul-2009 13:47:18] <schwartz> Alright. Host was having an OS process monitored that I shut down. Got an alert. Removed it from the host's OS section. I keep getting alerts on it, move to history, clear heartbeats.. keeps alerting me.
[09-Jul-2009 13:47:22] <schwartz> Any ideas?
[09-Jul-2009 14:03:15] <rmatte> schwartz: you know zendmd I assume?
[09-Jul-2009 14:03:26] <rmatte> or know of, rather
[09-Jul-2009 14:03:27] <schwartz> a bit
[09-Jul-2009 14:03:34] <rmatte> ok, go in to zendmd and type:
[09-Jul-2009 14:03:37] <rmatte> reindex()
[09-Jul-2009 14:03:39] <rmatte> commit()
[09-Jul-2009 14:03:51] <rmatte> the errors should stop coming in, if they don't, restart Zenoss
[09-Jul-2009 14:05:36] <schwartz> trying it. will that auto clear an acknowledge an alert for something that's no longer being monitored?
[09-Jul-2009 14:05:57] <rmatte> nah, but it'll make Zenoss realize that it's no longer being monitored
[09-Jul-2009 14:06:20] <rmatte> rarely, but sometimes, Zenoss doesn't figure out process monitoring status for some reason
[09-Jul-2009 14:06:28] <rmatte> a reindex and commit clears it up
[09-Jul-2009 14:06:38] <schwartz> ok. moved it to history. we'll see if it pops back up.
[09-Jul-2009 14:06:41] <rmatte> k
[09-Jul-2009 14:06:50] <rmatte> did you do the reindex and commit?
[09-Jul-2009 14:08:02] <schwartz> yep. showed back up.
[09-Jul-2009 14:08:10] <schwartz> I did a rindex, commit, then moved to history.
[09-Jul-2009 14:08:33] <rmatte> try restarting zope first
[09-Jul-2009 14:08:38] <rmatte> zopectl restart
[09-Jul-2009 14:08:46] <rmatte> you may also need to push the changes to the collect
[09-Jul-2009 14:08:58] <rmatte> navigate to that device page and go to Manage -> Push Changes
[09-Jul-2009 14:09:07] <rmatte> collector*
[09-Jul-2009 14:14:28] <rmatte> what would be the easiest way to generate a Zenoss report with the total amount of errors accumulated on each interface?
[09-Jul-2009 14:15:25] <perr0> mrayzenoss: do you think I should forward my e-mail to the sales person who sent me the pricing?
[09-Jul-2009 14:20:52] <eidolon> hey folks - i'm tryign to add a device. 'messages' says "Modeling of the following devices has been scheduled: something...'
[09-Jul-2009 14:20:57] <eidolon> so how do i make it go? 
[09-Jul-2009 14:21:25] <mrayzenoss> if you go to Settings->Jobs you'll see the job listed
[09-Jul-2009 14:23:13] <eidolon> aha,.
[09-Jul-2009 14:23:14] <eidolon> thank you.
[09-Jul-2009 14:23:28] <eidolon> can i trigger it to run now?
[09-Jul-2009 14:24:57] <eidolon> i have one job that is about 3 hours old, hasn't run yet.
[09-Jul-2009 14:27:43] <mrayzenoss> is the zenjobs daemon running?
[09-Jul-2009 14:28:06] <eidolon> yes
[09-Jul-2009 14:28:17] <eidolon> maybe restart it?
[09-Jul-2009 14:28:27] <eidolon> the only daemon not showing green is 'zenactions'
[09-Jul-2009 14:28:40] <eidolon> though, view log on zenjobs isshowing n exception from about a month ago.
[09-Jul-2009 14:29:11] * eidolon restarts.
[09-Jul-2009 14:29:59] <eidolon> that cleared it.
[09-Jul-2009 14:30:01] <eidolon> the jobs are running now.
[09-Jul-2009 14:34:23] <rmatte> in the interface utilization report...
[09-Jul-2009 14:34:25] <rmatte>     exportFields python:['device', 'speed', 'input', 'output', 'total', 'percentUsed'];
[09-Jul-2009 14:34:31] <rmatte> where is it pulling those values from?
[09-Jul-2009 14:34:46] <rmatte> I see...
[09-Jul-2009 14:34:47] <rmatte>            <td tal:content="python: r.humanBits(r.speed)"/>
[09-Jul-2009 14:34:47] <rmatte>            <td tal:content="python: r.humanBits(r.input, 8)"/>
[09-Jul-2009 14:34:47] <rmatte>            <td tal:content="python: r.humanBits(r.output, 8)"/>
[09-Jul-2009 14:34:48] <rmatte>            <td tal:content="python: r.humanBits(r.total, 8)"/>
[09-Jul-2009 14:34:56] <rmatte> I want to duplicate the report but do it for interface errors
[09-Jul-2009 14:54:11] <kisielk> what is the "messages" portlet for?
[09-Jul-2009 14:54:15] <kisielk> I can't find it mentioned anywhere
[09-Jul-2009 14:54:37] <eidolon> it's showing me system messages- like 'such and such was added to the job queue'
[09-Jul-2009 14:55:57] <rmatte> I think that's pretty much all it's for, shows system messages
[09-Jul-2009 14:56:28] <etank> i dont think i have ever seen anything in the messages portlet
[09-Jul-2009 14:56:38] <etank> not even when doing a discovery on a large network
[09-Jul-2009 15:07:59] <tswicegood> I'm trying to install the MySqlMonitor ZenPack and getting an error that Google isn't helping me with. http://pastie.org/540485 Anyone have any ideas?
[09-Jul-2009 15:13:35] <rmatte> is there really any danger in leaving the zCollectorClientTimeout value at like 5000 instead of 300?
[09-Jul-2009 15:14:04] <rmatte> one of our clients has devices in europe that we run across a very slow link to
[09-Jul-2009 15:20:28] <rmatte> hey Matt, you wouldn't know of a way to create a report for total interface errors within say the last week, would you?
[09-Jul-2009 15:20:49] <rmatte> I'm trying to figure it out but since it's per interface, I'm assuming I'll need to do something similar to the utilization report
[09-Jul-2009 15:21:00] <rmatte> I'm just wondering if there's one that's already been made floating around somewhere
[09-Jul-2009 15:22:18] <mrayzenoss> not that I'm aware of.
[09-Jul-2009 15:22:34] <rmatte> k, this could be fun then
[09-Jul-2009 15:23:12] <rmatte> we actually found a bug today with the utilization report, if you export it to csv it doesn't export the interfaces column
[09-Jul-2009 15:27:49] <mrayzenoss> ugh, please open a ticket
[09-Jul-2009 15:28:09] <rmatte> my manager said he'd open one with his shiny new Trac account
[09-Jul-2009 15:29:14] <ckrough> in ZenGlobe, is the 'list of servers' supposed to be comma seperated, line feed, etc... single server works. add more and it breaks
[09-Jul-2009 15:30:24] <mrayzenoss> ckrough: you'll probably need to ask support about that one, I'm not very familiar with all the Enterprise features
[09-Jul-2009 15:30:36] <ckrough> will do
[09-Jul-2009 15:31:29] <mrayzenoss> rmatte: if you get that new report working, you can submit it in a ZenPack. There are a couple of Community ZenPacks with reports
[09-Jul-2009 15:32:33] <rmatte> cool, I'll probably wait until I have a bit of a collection before submitting a pack
[09-Jul-2009 15:32:47] <rmatte> though report writing is going to take a while for me to get my head wrapped around
[09-Jul-2009 15:33:11] <mrayzenoss> yeah, there are some improvements related to making Python calls coming in King Crab
[09-Jul-2009 15:33:36] <mrayzenoss> but the reports writing needs more work
[09-Jul-2009 15:34:07] <rmatte> it'd be nice to see it as more of a GUI based report builder with optional coding
[09-Jul-2009 15:34:15] <rmatte> since they are reworking the GUI anyways
[09-Jul-2009 15:35:49] <mrayzenoss> yeah, the GUI overhaul is getting spread over multiple releases since mostly Ian is working on it
[09-Jul-2009 15:36:17] <rmatte> cool
[09-Jul-2009 15:36:21] <mrayzenoss> King Crab will have a lot of stuff around the Event Console and Event Views getting much nicer, probably some other screens
[09-Jul-2009 15:36:25] <rmatte> can't wait to see what it looks like a year from now
[09-Jul-2009 15:36:36] <schwartz> rmatte: did the device -> manage / push changes, that seems to have cleared things up. Thanks!
[09-Jul-2009 15:36:58] <rmatte> schwartz: no problem, that's because the actual collector only comes back to Zenoss for updates every hour
[09-Jul-2009 15:37:14] <rmatte> schwartz: so you need to point out to it that something has changed by manually pushing the changes, it comes in very handy
[09-Jul-2009 15:37:18] <schwartz> interesting. I'll have to remember that.
[09-Jul-2009 15:38:07] <rmatte> well, we lost a week's worth of data because of the SAN failure that we had on tuesday
[09-Jul-2009 15:38:20] <rmatte> kind of dumb since the thing is brand new, it had only been running for 20 days
[09-Jul-2009 15:38:58] <rmatte> now we're working to implement vmotion between the SAN and the local hard disks
[09-Jul-2009 15:39:24] <rmatte> as well as a tape backup solution, which has been planned for a while but not yet implemented
[09-Jul-2009 15:47:05] <rmatte> I'm going to log a Trac tomorrow since in 2.3.3 when you looked at a graph it had the name of the threshold along with an indication of what it was set at, like "PercentMemoryUsed > 90", now in 2.4.2 you don't get that anymore, however the legend values actually work when you set them
[09-Jul-2009 16:41:05] Administrator___ is now known as swisher
[09-Jul-2009 16:41:29] swisher is now known as rcane
[09-Jul-2009 16:41:51] rcane is now known as agthurber
[09-Jul-2009 16:42:27] <agthurber> anyone out there?
[09-Jul-2009 16:50:48] <mrayzenoss> agthurber: I'm here
[09-Jul-2009 16:51:04] <agthurber> hello
[09-Jul-2009 16:51:57] <agthurber> i am having some trouble with zendcommand running after we upgraded to 2.4.2, do you think you might be able to help?
[09-Jul-2009 16:52:28] <agthurber> it is very strange it only appears to run for a couple hours a day instead of continiously like it did before the upgrade
[09-Jul-2009 16:52:38] <mrayzenoss> possibly.  Have you looked at the zencommand output with zencommand -v10 run?
[09-Jul-2009 16:53:27] <mrayzenoss> oops, typed that too fast
[09-Jul-2009 16:53:34] <mrayzenoss> is the daemon dying?
[09-Jul-2009 16:54:12] <agthurber> if i run zenocmmand run it does appear to stop, the error it shows is
[09-Jul-2009 16:54:31] <agthurber> exceptions.ValueError: invalid literal for float(): U
[09-Jul-2009 16:54:56] <mrayzenoss> paste the whole log into pastebin.com
[09-Jul-2009 16:55:58] <agthurber> http://pastebin.com/m30c4bec5
[09-Jul-2009 16:58:05] <agthurber> the daemon appears to be running form the zenoss daemon manager
[09-Jul-2009 16:58:13] <agthurber> i only see that when i manually run the command
[09-Jul-2009 17:01:10] <mrayzenoss> it appears that there's some sort of error parsing the value for BytesReceivedPersec... what protocol is sjc1pw002 being monitored with?
[09-Jul-2009 17:01:30] <mrayzenoss> like it's trying to store a character into a float field in RRD
[09-Jul-2009 17:01:31] <agthurber> wmi i believe
[09-Jul-2009 17:01:46] <agthurber> yeah i saw that, but i dont know why any of that would have changed
[09-Jul-2009 17:02:01] <agthurber> all we did was upgrade, we didnt make any changes to those settings
[09-Jul-2009 17:03:24] <mrayzenoss> and other Windows boxes are fine?
[09-Jul-2009 17:04:03] <agthurber> zenoss is blown up after the upgrade
[09-Jul-2009 17:04:18] <agthurber> zencommand works for an hour or two each day, all on it's own
[09-Jul-2009 17:04:22] <agthurber> and then stopps
[09-Jul-2009 17:04:31] <agthurber> i dont restart it but it has been like this for a week now
[09-Jul-2009 17:04:51] <agthurber> and i am thinking we may have to start over from scratch
[09-Jul-2009 17:05:44] <agthurber> one thing i did do was set a value for zsshconcurrentsessions because of a post i found on how that helped fix zenncommand in 2.4.2 but that only let to this problem
[09-Jul-2009 17:06:01] <agthurber> led* to this problem
[09-Jul-2009 17:06:13] <rmatte> is there a known zencommand issue in 2.4.2?
[09-Jul-2009 17:06:16] <mrayzenoss> so this is using one of the community WMI ZenPacks?
[09-Jul-2009 17:06:23] <rmatte> because if so, there's no way I'm upgrading to that version
[09-Jul-2009 17:07:01] <agthurber> check the forums, I have been having trouble with zencommand in 2.4.2
[09-Jul-2009 17:07:08] <agthurber> we are using the built in WMI monitoring
[09-Jul-2009 17:07:31] <rmatte> I don't know what could have possibly changed between 2.4.1 and 2.4.2, we're using it in 2.4.2 without any problems at all
[09-Jul-2009 17:07:45] <agthurber> heres the other part
[09-Jul-2009 17:07:59] <agthurber> we upgraded from 2.2.4 to 2.3.3 then to 2.4.2
[09-Jul-2009 17:08:13] <agthurber> so that may be part of this
[09-Jul-2009 17:09:00] <agthurber> mray do you have any other suggestions?
[09-Jul-2009 17:09:04] <rmatte> it may be, I never used anything below 2.3.3
[09-Jul-2009 17:09:33] <agthurber> we have been using zenoss for a long time, the ony problems we ever have iwth it is when we go to upgrade, which is why we fell so far behind
[09-Jul-2009 17:10:04] <rmatte> I always test the upgrades on a lab box, if they look good I go ahead with them
[09-Jul-2009 17:11:02] <mrayzenoss> I don't have any suggestions, I'm looking at your thread
[09-Jul-2009 17:11:07] <rmatte> anyways, done my shift for the day, so talk to you guys later
[09-Jul-2009 17:11:51] <agthurber> have a good one
[09-Jul-2009 17:12:26] <mrayzenoss> agthurber: sorry, I don't have any good suggestions. You could try to open a ticket but this seems hard to recreate
[09-Jul-2009 17:12:50] <agthurber> yeah, i think were in a holoe, might just have to start over
[09-Jul-2009 17:13:47] <mrayzenoss> you should be able to dump your devices and reimport them
[09-Jul-2009 17:14:02] <mrayzenoss> just make sure you have the same ZenPacks installed
[09-Jul-2009 17:15:55] <agthurber> do you have a link to docs explaining this?
[09-Jul-2009 17:17:10] <agthurber> also when i look at the log for zencommand running as the zenoss user when it appears to stop i get the same line over and over and over in the log
[09-Jul-2009 17:17:17] <agthurber> 2009-07-09 11:11:25 DEBUG zen.zencommand: Queueing event {'manager': 'localhost', 'timeout': 180, 'device': 'localhost', 'eventClass': '/Heartbeat', 'component': 'zencommand', 'agent': 'zencommand'}
[09-Jul-2009 17:17:18] <agthurber> 2009-07-09 11:11:25 DEBUG zen.zencommand: Total of 1 queued events
[09-Jul-2009 17:19:52] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch04s02.html#d4e1761
[09-Jul-2009 17:20:56] <agthurber> will exporting a device retain the commands associated with it?
[09-Jul-2009 17:21:09] <agthurber> and thank you, for that link
[09-Jul-2009 17:21:32] <mrayzenoss> If you have custom commands, you could probably export them into a ZenPack
[09-Jul-2009 17:21:45] <mrayzenoss> and import them into your new system
[09-Jul-2009 17:22:02] <mrayzenoss> before importing the devices
[09-Jul-2009 17:26:27] <agthurber> so if i have a device class and export the whole class as a zendpack, when i restore it it will not create the devices inside that class?
[09-Jul-2009 17:27:11] <mrayzenoss> no, it will have the device class and whatever templates you export in the ZenPack
[09-Jul-2009 17:27:52] <agthurber> so a zend=pack only saves the classes and the templates, not the devices
[09-Jul-2009 17:28:00] <mrayzenoss> correct
[09-Jul-2009 17:28:05] <agthurber> ok that makes more sense now
[09-Jul-2009 17:29:48] <agthurber> do i need to re-apply the templates to the devices after i import them?
[09-Jul-2009 17:31:36] <mrayzenoss> hmmm... I'm not sure
[09-Jul-2009 17:31:45] <mrayzenoss> You may have to rebind them
[09-Jul-2009 17:43:35] <agthurber> are there any updates to the WMI zendpacks? i un-bound them from the devices they were associated with and now zendcommand is working
[09-Jul-2009 17:44:38] <mrayzenoss> I've asked the author to submit them to the ZenPack repository, so other folks can update them, but he hasn't done so yet
[09-Jul-2009 17:45:00] <agthurber> i think that is where our problem is located
[09-Jul-2009 17:45:04] <mrayzenoss> and since they don't have a license, they're his ZenPacks so I can't update them
[09-Jul-2009 18:25:15] raw__ is now known as rmatte_mobile
[10-Jul-2009 00:00:46] [disconnected at Fri Jul 10 00:00:46 2009]
[10-Jul-2009 00:00:46] [connected at Fri Jul 10 00:00:46 2009]
[10-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[10-Jul-2009 00:01:11] <hackeron> hey, I'm just trying to install zenoss and in the INSTALL.txt file it says mysql -- can I not use postgres with zenoss?
[10-Jul-2009 00:13:46] <hackeron> also, just looking at the wikipedia monitoring systems comparison and it says zenoss doesn't have an "agent" - in my current environment I need to run a custom script on the remote host to query the database, analyse the data and report back some numbers to be checked for threshold and plotted - would this not be possible with zenoss?
[10-Jul-2009 02:54:39] <Garn32> I'm having problem with zenoss, all of the graphs show up with empty boxes for all of the text they should have, any ideas?
[10-Jul-2009 03:36:25] <Guest4926> hi
[10-Jul-2009 03:36:52] <Guest4926> anyone active?
[10-Jul-2009 04:07:45] <Dieterbe> yes
[10-Jul-2009 06:39:35] <etank> hackeron: your second question is one that i have too
[10-Jul-2009 06:40:23] <etank> but for me the remote box is a linux system
[10-Jul-2009 08:06:47] <hackeron> etank: my remote box is a linux system too
[10-Jul-2009 08:07:00] <hackeron> etank: did you find an answer?
[10-Jul-2009 08:29:05] <hackeron> anyone? - looking at the wikipedia monitoring system comparison, it says zenoss doesn't have an agent -- does that mean I can't have a host behind a firewall sending zenoss information or run custom scripts on the host to get various numbers?
[10-Jul-2009 08:32:57] <venturaville> snmpd is an agent, and can be extended to run custom scripts
[10-Jul-2009 08:33:34] <venturaville> zenoss uses standard agents, rather than write its own which is what wikipedia is referring to
[10-Jul-2009 08:34:11] <zenChild> hackeron: There are a number of ways to do what you want to do. For the firewall issue you can set up a distributed collector if you only want to open up one port to one host. This is easy with the Enterprise version, but possible with the core (http://www.zenoss.com/Members/fdeckert/how-to-install-distributed-collectors/)
[10-Jul-2009 08:34:43] <zenChild> hackeron: as for running commands on the host, you can collect information with SSH. That capability is built into the core product.
[10-Jul-2009 08:35:24] <zenChild> and like venturaville said, you can always extend snmpd on the host side.
[10-Jul-2009 08:35:41] <hackeron> zenChild: so I can get the client to ssh into the zenoss host and send it information?
[10-Jul-2009 08:36:15] <venturaville> that is more difficult, but it can be done
[10-Jul-2009 08:36:36] <hackeron> venturaville: any documentation for that?
[10-Jul-2009 08:37:51] <zenChild> http://www.zenoss.com/community/docs/zenoss-extended-monitoring/2.4.2/ch15.html
[10-Jul-2009 08:37:52] <hackeron> zenChild: that looks interesting, but what do you mean open up one port? -- I meant the box is behind NAT and has no open ports, it can only make outgoing connections
[10-Jul-2009 08:38:52] <venturaville> there is an alternative
[10-Jul-2009 08:39:06] <hackeron> zenChild: that shows how to set up a host to ssh to, not how to set up a host to accept ssh connections from?
[10-Jul-2009 08:39:07] <venturaville> zenoss has distributed collectors ... if you put it behind the NAT, and then only talk to it
[10-Jul-2009 08:39:28] <hackeron> venturaville: how would I talk to it?
[10-Jul-2009 08:39:47] <zenChild> hackeron: i was thinking you were talking about traffic between addressable DMZs.
[10-Jul-2009 08:41:21] <venturaville> http://www.zenoss.com/community/docs/zenoss-guide/2.3.3/ch28s01.html
[10-Jul-2009 08:41:46] <venturaville> " Collect performance and events from networks that cannot be reached by the Zenoss server "
[10-Jul-2009 08:42:06] <hackeron> zenChild: no, most of my hosts are behind a nat with no open ports and no way to talk to them -- currently I use logcheck to send me periodic emails of anything unexpected in the logs
[10-Jul-2009 08:42:32] <venturaville> your best bet is the distributed collector
[10-Jul-2009 08:43:03] <venturaville> that or just move your entire monitoring system to there (if most of your hosts are there anyways)
[10-Jul-2009 08:43:33] <zenChild> venturaville:  Have you gotten anywhere with the ZenPack for DS4k monitoring?
[10-Jul-2009 08:44:32] <hackeron> venturaville: no, every one of my hosts is on a separate LAN -- basically I deploy DVRs and firewall configuration is always tricky - normally can only get some obscure port for http -- here's an example: http://hackeron.dyndns.org:82/cctv/events - user/pass: demo/demo
[10-Jul-2009 08:44:55] <hackeron> venturaville: that's why I need an agent to send this information, not for zenoss to poll this information
[10-Jul-2009 08:45:25] <venturaville> zenChild: I will be starting work on that on Monday... too busy working on puppet lately
[10-Jul-2009 08:45:51] <hackeron> venturaville: currently I have postfix set up to connect to my smtp server with it's own user/pass and email me logcheck reports and a few other custom warnings
[10-Jul-2009 08:45:56] <venturaville> hackeron: you have some unusual requirements  :-)
[10-Jul-2009 08:45:57] <hackeron> venturaville: looks like maybe zenoss is not for me
[10-Jul-2009 08:46:20] <zenChild> venturaville: puppet.  Nice product.  I use it to deploy my TSM servers.
[10-Jul-2009 08:47:17] <venturaville> zenChild: it is better than opsware IMHO
[10-Jul-2009 08:47:18] <zenChild> hackeron: I would fire something off to the forums and see if anyone has an answer for you.
[10-Jul-2009 08:47:46] <hackeron> venturaville: well, any network monitoring that has an agent that will send information to the monitoring host will do -- is there really no way to achieve this without installing enterprice and using distributed collectors? - I want something very basic -- instead of zenoss -> client to have client -> zenoss
[10-Jul-2009 08:47:56] <zenChild> venturaville: agreed.  I met Luke at OSCON a couple years ago too.  He's a sharp guy.
[10-Jul-2009 08:48:51] <hackeron> zenChild: thanks, I'll try that, zenoss looks really nice and I'll be damned if I use java and opennms, lol -- why don't you guys have an agent though!
[10-Jul-2009 08:57:31] <zenChild> hackeron: I had one more thought... could you use a vpn client on the zenoss host to monitor hosts behind the NAT?
[10-Jul-2009 08:58:14] <zenChild> hackeron: We have a storage management network set up like that and we use OpenVPN on the Zenoss host to access hosts within that network.
[10-Jul-2009 08:59:28] <hackeron> zenChild: I was thinking about that actually - to have all the hosts behind nat connect to an openvpn - then the hosts appear local to zenoss -- but internet connectivity can be intermittent on the hosts and I need something that will "queue" all the monitoring stats until connectivity is back
[10-Jul-2009 09:01:00] <venturaville> hackeron: I am not sure any monitoring solution is going to be a perfect fit for you
[10-Jul-2009 09:01:37] <zenChild> If you did a remote collector in the NAT and then used OpenVPN to let the to collectors talk to eachother that would work.
[10-Jul-2009 09:02:30] <hackeron> venturaville: basically I want the monitoring solution to wait for the hosts to send it stuff - if it doesn't get a report in time, report host being down - and on the client side to have some kind of collector that will retry sending details to host, so when connectivity is down for say 3 hours, I don't lose any of the monitoring stats for that period
[10-Jul-2009 09:02:41] <hackeron> venturaville: so I get a kind of online uptime and actual dvr uptime
[10-Jul-2009 09:03:17] <venturaville> hackeron: yep had to do something like that over a decade ago for modem lines... I ended up having to write it :-?
[10-Jul-2009 09:03:41] <hackeron> venturaville: won't pretty much any monitoring solution with an agent do that though?
[10-Jul-2009 09:04:08] <venturaville> hackeron: not so much the queuing part... though depending on your programming skills you could probably do a hub to queue it up
[10-Jul-2009 09:04:36] <venturaville> if mrayzenoss ever gets on here he might have some better thoughts though
[10-Jul-2009 09:12:06] <hackeron> venturaville: I could write the queueing part - but is there some standard way to send stuff to zenoss?
[10-Jul-2009 09:12:27] <jb> hm i just noticed that when you export events to a CSV
[10-Jul-2009 09:12:31] <jb> that the dates are all wrong
[10-Jul-2009 09:12:33] <jb> date/times
[10-Jul-2009 09:12:35] <venturaville> I was thinking more along the lines of having the device template pull data from the hub
[10-Jul-2009 09:13:17] <hackeron> venturaville: bah, lol - what hub? - does zenoss have a hub?
[10-Jul-2009 09:13:25] <venturaville> I'm going to have to do something similar for the IBM DS4xxx array data
[10-Jul-2009 09:14:18] <venturaville> hackeron: I would suggest talking to mrayzenoss when he gets on
[10-Jul-2009 09:14:39] <hackeron> venturaville: heh, ok, I will
[10-Jul-2009 09:15:24] <jb> oh, it was excell.
[10-Jul-2009 09:15:41] <venturaville> mrayzenoss: hackeron here has an odd problem for you :-)
[10-Jul-2009 09:16:29] <hackeron> mrayzenoss: hi -- I have an odd problem as venturaville said. I've made a little CCTV over internet system and deployed around 20 DVRs I would like to monitor
[10-Jul-2009 09:16:54] <hackeron> mrayzenoss: here's what it looks like: http://hackeron.dyndns.org:82/cctv/events to give you some insight (user/pass: demo/demo)
[10-Jul-2009 09:17:28] <hackeron> mrayzenoss: problem is, every DVR is behind a NAT and port forwarding is normally impossible - I just about get away with forwarding port 80 to some obscure port
[10-Jul-2009 09:18:00] <hackeron> mrayzenoss: so I was hoping to instead of having zenoss pull information from the hosts, for the hosts to send zenoss information using ssh or some other method
[10-Jul-2009 09:19:29] <hackeron> reason for sending information is connectivity is also quite intermittent as cheap consumer dsl is used and I would like to have both DVR uptime and online uptime
[10-Jul-2009 09:21:00] <hackeron> I could write the queueing of monitoring info to be sent to zenoss, but I can't see any way to actually get information into zenoss without it pulling it
[10-Jul-2009 09:21:55] <mrayzenoss> interesting...
[10-Jul-2009 09:22:22] <mrayzenoss> rather than send your monitoring information, you could send events to the Zenoss system saying "everything's OK"
[10-Jul-2009 09:22:56] <mrayzenoss> and then perhaps some sort of trigger that says "if I haven't heard back in 30 minutes, everything is not OK"
[10-Jul-2009 09:23:10] <mrayzenoss> like using traps
[10-Jul-2009 09:23:48] <hackeron> hmm, ok - but how will zenoss get the monitoring information?
[10-Jul-2009 09:24:44] <mrayzenoss> well, that's the rub. You could try to go nuts with an event transform for parsing the inbound events and stuff everything into them
[10-Jul-2009 09:25:12] <mrayzenoss> let me think about it a bit, maybe there's a better way
[10-Jul-2009 09:25:30] <mrayzenoss> you can't just use SSH on an obfuscated port?
[10-Jul-2009 09:25:58] <mrayzenoss> or even SNMP v3?
[10-Jul-2009 09:26:58] <hackeron> mrayzenoss: I normally can only forward 1 port, sometimes I get away with forwarding 2, so I have http and ssh - but even then, connectivity is intermittent, -- say the DVR is offline for 3 hours but still up and running and local users are using it - how will I get information for the 3 missing hours?
[10-Jul-2009 09:27:28] <mrayzenoss> yeah, that's not something we handle very well yet
[10-Jul-2009 09:27:58] <mrayzenoss> you could monitor with SSH and send events from the DVR as heartbeats
[10-Jul-2009 09:28:12] <mrayzenoss> so you'd be missing monitoring data, but not alarm as long as you're getting heartbeats
[10-Jul-2009 09:28:29] <etank> zenChild: do you have a doc on how you set up the OpenVPN stuff? we have a requirement like that.
[10-Jul-2009 09:28:32] <hackeron> the event transform is interesting though - I could write a little python script that will try to send events for all monitoring information and if it can't, store in queue to retry in a bit -- is it possible to have events and monitoring information with a past date/time?
[10-Jul-2009 09:29:28] <mrayzenoss> hackeron: you can go really nuts with a transform, you have access to essentially everything in the system, so yes.
[10-Jul-2009 09:29:44] <mrayzenoss> it's just Python with access to the dmd
[10-Jul-2009 09:30:21] <hackeron> mrayzenoss: hmm, I'll look into that. That's very, very messy and backwards though, eh?
[10-Jul-2009 09:33:06] <mrayzenoss> so I went and talked to a couple of the devs, and they pointed out that you could probably just use syslog
[10-Jul-2009 09:33:51] <mrayzenoss> have the DVRs send their monitoring info in syslog, then set up rules based on what they're sending
[10-Jul-2009 09:34:25] <zenChild> etank:  one sec.  in a conversation.
[10-Jul-2009 09:34:38] <mrayzenoss> hackeron: you won't have the stats and graphs, but you'll be able to just send from the DVR the things you care about
[10-Jul-2009 09:35:28] <hackeron> mrayzenoss: hmm, interesting! - is there no way to have the stats and graphs though?
[10-Jul-2009 09:36:00] <etank> zenChild: np
[10-Jul-2009 09:36:02] <mrayzenoss> hackeron: probably if you do some work with the transforms...
[10-Jul-2009 09:36:04] <venturaville> actually I have a wicked idea ... you can have a data collector connect back to zendmd and fetch the last event from the device and grab data out of it
[10-Jul-2009 09:36:10] <venturaville> *grins evilly*
[10-Jul-2009 09:36:53] <mrayzenoss> hackeron: or a 2-phase solution where the syslog output is written to a log file and then you use a Command datasource to parse it
[10-Jul-2009 09:37:03] <mrayzenoss> you could get your graphs that way
[10-Jul-2009 09:37:41] <hackeron> mrayzenoss: that could work - I mean currently I use logcheck which does just that, except emails the information instead of sending it to another syslog server
[10-Jul-2009 09:38:07] <mrayzenoss> yeah, that would actually be easier to test and debug as well
[10-Jul-2009 09:38:39] <hackeron> mrayzenoss: so I'll write a script that will check all snmpd data as well as whatever else I'm interested in, log it to syslog - then use logcheck to get just the juicy, unparsed stuff, and then a separate script to retry endlessly to send it to a syslog server -- that could work!
[10-Jul-2009 09:39:09] <mrayzenoss> yeah, that sounds feasible
[10-Jul-2009 09:40:56] <mrayzenoss> Here's an example of a manipulating an external system and pushing it into RRD (in this case a database, but a shell command could be used) http://www.zenoss.com/Members/cluther/licenseEventCheck.py/view
[10-Jul-2009 09:41:24] <mrayzenoss> This example queries a database and monitors how many license keys are in use
[10-Jul-2009 09:43:29] <hackeron> mrayzenoss: thanks for that - so when I do ZenScriptBase(connect=True).dmd
[10-Jul-2009 09:43:45] <hackeron> mrayzenoss: I can just specify it to connect to the zemboss host online?
[10-Jul-2009 09:43:56] <rmatte> does anyone here have a Zenoss report build that reports interface errors?
[10-Jul-2009 09:44:05] <rmatte> built* rather
[10-Jul-2009 09:45:23] <mrayzenoss> hackeron: I'm guessing that's running as the 'zenoss' user
[10-Jul-2009 09:45:29] <mrayzenoss> but yeah
[10-Jul-2009 09:45:47] <zenChild> etank: I really didn't do anything out of the ordinary... just followed the OpenVPN docs. The zenoss server is just an OpenVPN client into the private network. The big thing is to put some thought behind your address ranges in case you have to be a client into many private networks. And if you don't have any CA tools to manage the certificates I would suggest something like TinyCA.
[10-Jul-2009 09:46:27] <hackeron> mrayzenoss: hmm, thanks for that, so for monitoring information I could use a series of scripts like that and create a class that will queue and retry
[10-Jul-2009 09:46:40] <etank> zenChild: i see.
[10-Jul-2009 09:46:46] <mrayzenoss> hackeron: good luck
[10-Jul-2009 09:47:45] <hackeron> mrayzenoss: are there any examples of using syslog data for events?
[10-Jul-2009 09:47:50] <zenChild> etank:  when you're setting it up you can ping me here if you are having issues.
[10-Jul-2009 09:49:18] <etank> zenChild: thanks. we have a cisco vpn device at the location. im wondering if i can just pass through that.
[10-Jul-2009 09:49:34] <etank> if so then i wouldnt even need OpenVPN
[10-Jul-2009 09:49:53] <mrayzenoss> hackeron: if you just point something generating syslog at your Zenoss server it'll show up in the Event Console
[10-Jul-2009 09:49:56] <zenChild> etank: should be able to. You could use vpnc or the cisco vpn client and things should work the same.
[10-Jul-2009 09:50:13] <mrayzenoss> hackeron: they usually go to INFO if they're /Unknown
[10-Jul-2009 09:50:46] <hackeron> mrayzenoss: ok, thanks - I'll probably be back
[10-Jul-2009 09:50:47] <mrayzenoss> take a look at Jane Curry's Event Management paper for further details: http://www.skills-1st.co.uk/papers/jcurry.html
[10-Jul-2009 09:50:50] <etank> zenChild: would you cron the connections though? or just leave them open. we are planning on using enterprise with remote collectors
[10-Jul-2009 09:51:49] <zenChild> I just have my OpenVPN connect at boot time and keep it running.
[10-Jul-2009 09:52:28] <etank> cool. we will have to pass that by IT security before we can set that up. but it seems like a plan.
[10-Jul-2009 09:57:36] <hackeron> mrayzenoss: thanks for that
[10-Jul-2009 09:58:18] <rmatte> all of Jane Curry's stuff is worth reading, very good
[10-Jul-2009 09:58:25] <rmatte> kind of fills in some gaps in the admin guide
[10-Jul-2009 09:58:32] <mrayzenoss> yeah, she's got some training sessions in Europe this summer
[10-Jul-2009 09:58:40] <rmatte> cool
[10-Jul-2009 09:58:44] <hackeron> mrayzenoss: oh yeah? where in europe?
[10-Jul-2009 09:59:10] <rmatte> say amsterdam, or ireland, I'll have my company fly me over
[10-Jul-2009 09:59:15] <mrayzenoss> http://www.skills-1st.co.uk/products/courses/zenoss-events.html
[10-Jul-2009 09:59:23] <mrayzenoss> London
[10-Jul-2009 09:59:29] <mrayzenoss> and something later in Austria
[10-Jul-2009 09:59:37] <hackeron> maidenhead!
[10-Jul-2009 09:59:42] <hackeron> that's a 40 minute drive for me, lol
[10-Jul-2009 09:59:57] <hackeron> if only I had 1200 quid, lol
[10-Jul-2009 10:00:08] <rmatte> the minute you have a week long workshop on writing collector plugins I'm there
[10-Jul-2009 10:00:32] <rmatte> I'd love developer workshops, would be nice if they were available via webex
[10-Jul-2009 10:05:19] <etank> rmatte: i agree
[10-Jul-2009 10:05:42] <rmatte> is there a guide somewhere that explains writing them?
[10-Jul-2009 10:06:01] <rmatte> or is it one of those: dig through other people's code and fly by the seat of your pants type thing?
[10-Jul-2009 10:07:23] <etank> i wish some of the community packs had better documentation :/
[10-Jul-2009 10:07:35] <etank> they install but now idea how to set them up properly
[10-Jul-2009 10:08:53] <rmatte> such as?
[10-Jul-2009 10:09:24] <rmatte> I haven't found the setup of any of them to be overly complicated, except for httpmon which was a bit of a trial and error type deal at first
[10-Jul-2009 10:12:51] <etank> httpmon was one and the other was the predictive threshold one
[10-Jul-2009 10:16:09] <etank> it could be because i am still fairly new to zenoss too
[10-Jul-2009 10:18:05] <eidolon> question. if i have emplates defined for, say, MySQL servers, and i add a new host, and i want that template to be used on the new host, I use More->Templates, then Bind Template, yes? and pick my /Device/MySQL tempalte? that will apply the mysql template to the device, not copy it? (i want to inherit the templat settings, not make a copy of them)
[10-Jul-2009 10:18:11] <eidolon> ala, use 'bind' as opposed to 'add'?
[10-Jul-2009 10:18:58] <jb> binding it is fine..
[10-Jul-2009 10:19:11] <jb> unless you want to make a local copy to change some settings of the template that only apply to that device
[10-Jul-2009 10:19:23] <eidolon> no, definitely don't want a local copy
[10-Jul-2009 10:19:29] <eidolon> that was done on a bunch of machiens here, and i need to un-do that.
[10-Jul-2009 10:19:33] <eidolon> because they're messed up
[10-Jul-2009 10:19:45] <jb>
[10-Jul-2009 10:20:14] <eidolon> should i remodel after adding the new template?
[10-Jul-2009 10:20:18] <eidolon> er, binding?
[10-Jul-2009 10:20:44] <eidolon> oh, looks like i don't need to
[10-Jul-2009 10:21:03] <eidolon> perf now has mysql statistics. but they're empty. i guess i'll wait for the polling interval to pass
[10-Jul-2009 10:22:20] <eidolon> crap
[10-Jul-2009 10:22:28] <eidolon> MySQL Error: (1045, "Access denied for user 'zenoss'@'
[10-Jul-2009 10:22:31] * eidolon fixes.
[10-Jul-2009 10:28:57] <eidolon> hrm. the zenpack docs are saying use a root login on the mysql server to gatehr stats. that seems odd.
[10-Jul-2009 10:43:46] <dollarbang> is there a write-up (or wiki) on the content of the logs, and maybe something on how to debug using them?
[10-Jul-2009 10:49:03] <ckrough> dollarbang: http://www.zenoss.com/community/docs/zenoss-guide/2.3.3/apfs01.html
[10-Jul-2009 10:49:19] <ckrough> dollarbang: I couldnt find that same section in 2.4 docs
[10-Jul-2009 10:49:40] <ckrough> dollarbang: I dont run 2.4, but I think there are some significant changes in log locations and content
[10-Jul-2009 10:53:54] <mrayzenoss1> hmm... I wonder where that section went
[10-Jul-2009 10:58:25] <ckrough> mrayzenoss1: best place for feature requests is a trac ticket?
[10-Jul-2009 10:58:38] <dberger> is there a zProperty that corresponds to snmp v3 username?
[10-Jul-2009 10:58:45] <mrayzenoss1> ckrough: yeah
[10-Jul-2009 10:58:50] <tehhobbit> snmpsecurityname
[10-Jul-2009 10:58:55] <dberger> thanks
[10-Jul-2009 11:09:11] <hackeron> hey, trying to install zenoss from source on latest ubuntu (9.04) - getting this: http://dpaste.com/65525/ - any ideas?
[10-Jul-2009 11:14:18] <mrayzenoss1> hackeron: http://dev.zenoss.org/trac/ticket/4994
[10-Jul-2009 11:14:31] <mrayzenoss1> hackeron: our not so secret shame
[10-Jul-2009 11:14:52] <mrayzenoss1> source builds currently suck for everything but Debian and RHEL/CentOS
[10-Jul-2009 11:23:20] <hackeron> mrayzenoss1: heh, annoying - any solution to this?
[10-Jul-2009 11:44:41] <eidolon> okay, i'm being dumb here. i have zenoss - it's modelled a host. i can see performance data. i just deliberately filled up a fixed disk to 96%. zenoss is not raising an alert. i'm -very- new to zenoss - where do i check what the alert thresholds for various states are?
[10-Jul-2009 11:54:37] <klinstifen> eidolon: go to the device, os tab
[10-Jul-2009 11:54:44] <klinstifen> click on the drive in question
[10-Jul-2009 11:54:53] <klinstifen> go to template tab
[10-Jul-2009 11:55:04] <klinstifen> and then filesystem and then look at the threshold
[10-Jul-2009 11:55:15] <klinstifen> that's where it is set
[10-Jul-2009 11:55:21] <klinstifen> you can adjust it from within there
[10-Jul-2009 11:55:26] <klinstifen> if need be
[10-Jul-2009 12:17:48] <dollarbang> ckrough: thanks, the Admin Guide for 2.4 doesn't have a troubleshooting section.
[10-Jul-2009 12:18:19] <mrayzenoss> ckrough: dollarbang: I've brought it up to see why it was removed, it'll move to a wiki page soon
[10-Jul-2009 13:00:23] <ckrough> mrayzenoss: is Trac an appropriate place for feature requests that only apply to enterprise installs?
[10-Jul-2009 13:00:58] <ckrough> mrayzenoss: It would be nice if it was in a location that others can comment/contribute, but I dont want to throw it out if it's irrelevant to most of the community
[10-Jul-2009 13:01:34] <ckrough> throw it out there, not throw it out...
[10-Jul-2009 13:02:39] <mrayzenoss> ckrough: there's probably something in the support portal for enterprise feature requests
[10-Jul-2009 13:02:50] <mrayzenoss> ckrough: of course, I don't have a login to look
[10-Jul-2009 13:06:43] <ckrough> Ill poke around
[10-Jul-2009 13:08:36] <eidolon> odd.  zenoss says this fs is at 90%
[10-Jul-2009 13:08:39] <eidolon> df shows it at 98%
[10-Jul-2009 13:17:56] <eidolon> dammit
[10-Jul-2009 13:18:07] <eidolon> okay, so we haev a class tempalte that says "throw a warning when this goes over 85%"
[10-Jul-2009 13:18:16] <eidolon> looking at the os->mount, it's at 93%
[10-Jul-2009 13:18:20] <eidolon> why didn't the alert trigger?
[10-Jul-2009 13:18:30] <eidolon> this is Sort Of Important :-/
[10-Jul-2009 13:18:35] <eidolon> because i want to turn off nagios ont hese hosts.
[10-Jul-2009 13:19:13] <eidolon> (note that lcicking on the filesystem does inf act show it at 93% - and looking at /Devices/Filesystem, there's an enabled WARN at 85%
[10-Jul-2009 13:19:16] <rmatte> eidolon: I believe there's a known bug that causes that, and there is a workaround buries somewhere
[10-Jul-2009 13:19:26] <rmatte> maybe Matt can point you do it
[10-Jul-2009 13:19:47] <eidolon> please. this is pretty important for us i can open a ticket i'm sure (we have an enterprise license)
[10-Jul-2009 13:20:10] <ckrough> eidolon: they are lightning fast on the enterprise tickets usually
[10-Jul-2009 13:20:23] <eidolon> my experience hasn't shown that.
[10-Jul-2009 13:20:27] <eidolon> they update around 9am each day.
[10-Jul-2009 13:20:28] <ckrough> hmm
[10-Jul-2009 13:20:40] <eidolon> i had an open ticket that was taking 20hrs turnaround on each exchange of information.
[10-Jul-2009 13:20:48] <ckrough> sounds like you need to thump some heads
[10-Jul-2009 13:21:09] <eidolon> so here's the rule in max Value
[10-Jul-2009 13:21:10] <eidolon> (here.totalBlocks * here.zFileSystemSizeOffset) * .85
[10-Jul-2009 13:21:18] <eidolon> is that right?
[10-Jul-2009 13:21:36] <eidolon> (this is here, btw: /Devices /Server /Templates /FileSystem /high disk usage warning
[10-Jul-2009 13:22:47] <ckrough> here.totalBlocks * .9 is what I have
[10-Jul-2009 13:22:51] <ckrough> (here.totalBlocks * .9)
[10-Jul-2009 13:23:01] <ckrough> what is here.zFileSystemSizeOffset
[10-Jul-2009 13:23:13] <eidolon> reading this: http://www.sysadminwiki.net/wiki/index.php?title=Linux_Filesystems_in_Zenoss
[10-Jul-2009 13:23:49] <rmatte> yeh, that's the article that I remember reading
[10-Jul-2009 13:23:51] <eidolon> the offiset is from hre:  http://dev.zenoss.org/trac/ticket/4378
[10-Jul-2009 13:24:18] <rmatte> did you create the zProperty in zendmd first as shown?
[10-Jul-2009 13:24:19] <eidolon> i wonder if our version of zenoss supports this.
[10-Jul-2009 13:24:25] <eidolon> sec.
[10-Jul-2009 13:24:29] <eidolon> i just inherited this system
[10-Jul-2009 13:24:29] <rmatte> # zendmd
[10-Jul-2009 13:24:29] <rmatte> dmd.Devices._setProperty("zFileSystemSizeOffset", 1.0, type="float")
[10-Jul-2009 13:24:29] <rmatte> commit()
[10-Jul-2009 13:24:54] <eidolon> yah i see it
[10-Jul-2009 13:24:58] <eidolon> how can i tell if it's there already?
[10-Jul-2009 13:25:03] * eidolon is not a python person
[10-Jul-2009 13:25:09] <ckrough> inheriting a production zenoss env can definitely try your patience
[10-Jul-2009 13:25:15] <eidolon> mmhmm.
[10-Jul-2009 13:25:16] <rmatte> just run the command, if it's already there it won't hurt anything
[10-Jul-2009 13:25:19] <eidolon> ok
[10-Jul-2009 13:25:21] <rmatte> but if you haven't done it then it probably isn't
[10-Jul-2009 13:25:36] <ckrough> rmatte:  so that sets it for the /Devices class?
[10-Jul-2009 13:25:38] <rmatte> just follow that guide step by step
[10-Jul-2009 13:26:02] * ckrough adds another todo to his list
[10-Jul-2009 13:26:05] <eidolon> i assume someone alreayd has followed this.
[10-Jul-2009 13:26:07] <eidolon> but i'll check
[10-Jul-2009 13:26:13] <eidolon> because that offset is already in the template.
[10-Jul-2009 13:26:16] <rmatte> either that, or lower the threshold to 85 percent or something
[10-Jul-2009 13:26:22] <rmatte> that would be the quick and easy way to do it
[10-Jul-2009 13:27:29] <eidolon> [zenoss@bos-admin06 ~]$ zendmd
[10-Jul-2009 13:27:29] <eidolon> 2009-07-10 14:27:00 WARNING OFS.Application Duplicate Product name
[10-Jul-2009 13:27:29] <eidolon> After loading Product 'Five' from '/opt/zenoss/Products',
[10-Jul-2009 13:27:29] <eidolon> I skipped the one in '/opt/zenoss/lib/python/Products'.
[10-Jul-2009 13:27:31] <eidolon> interesting
[10-Jul-2009 13:28:09] <eidolon> okay, looks like that property is already there.
[10-Jul-2009 13:28:15] <eidolon> BadRequest: Id zFileSystemSizeOffset is invalid or duplicate
[10-Jul-2009 13:29:48] <rmatte> k
[10-Jul-2009 13:30:00] <rmatte> follow the rest of that proc then and see if it's already done
[10-Jul-2009 13:30:04] <rmatte> maybe a step was missed
[10-Jul-2009 13:30:16] <eidolon> interesting
[10-Jul-2009 13:30:26] <eidolon> what's int eh transform is not the same as what's in that document.
[10-Jul-2009 13:30:51] <eidolon> http://pastebin.stonekeep.com/6021
[10-Jul-2009 13:30:51] <eidolon> tjhat
[10-Jul-2009 13:30:56] <eidolon> thats what's in my zenoss install
[10-Jul-2009 13:32:06] <eidolon> so... do i jettison what i have here and replace it with what's in that page?
[10-Jul-2009 13:32:54] <rmatte> yeh, there's no mention of zFileSystemSizeOffset in yours
[10-Jul-2009 13:33:27] <eidolon> except the offset is in the template rules elsewhere.
[10-Jul-2009 13:33:42] <eidolon> well, i supose i can just try it
[10-Jul-2009 13:33:46] <rmatte> yeh, put that one in there
[10-Jul-2009 13:33:47] <eidolon> i'll make a backup of the old code
[10-Jul-2009 13:33:49] <eidolon> ok
[10-Jul-2009 13:33:56] <rmatte> then you need to adjust that zProperties value for that device class
[10-Jul-2009 13:34:02] <rmatte> change it to .95 or something
[10-Jul-2009 13:34:06] <rmatte> leave the rest at 1
[10-Jul-2009 13:34:19] <eidolon> wait, what?
[10-Jul-2009 13:34:28] <eidolon> change what to .95?
[10-Jul-2009 13:34:33] <eidolon> the max?
[10-Jul-2009 13:34:39] <rmatte> zFileSystemSizeOffset os a zProperty on the zProperties page
[10-Jul-2009 13:34:42] <kisielk> Is it just me, or is the Network Map mostly just completely broken?
[10-Jul-2009 13:34:52] <eidolon> ok
[10-Jul-2009 13:35:00] <rmatte> the network map is fine provided you collected enough routes when modelling
[10-Jul-2009 13:35:08] <kisielk> I have a lot of routes
[10-Jul-2009 13:35:11] <kisielk> I think
[10-Jul-2009 13:35:21] <kisielk> but the network map doesn't seem to work at all for me
[10-Jul-2009 13:35:26] <rmatte> yeh, but you need routes from pretty much everything
[10-Jul-2009 13:35:44] <rmatte> and you need to set the starting point IP
[10-Jul-2009 13:35:49] <rmatte> in the top left
[10-Jul-2009 13:35:58] <rmatte> you can also specify a default in your settings
[10-Jul-2009 13:37:39] <eidolon> rmatte: so /Devices/Server - in zProperties, I have zFileSysdtemSizeOffset that's currently 1.0
[10-Jul-2009 13:37:42] <eidolon> you suggest changing that to .95?
[10-Jul-2009 13:37:55] <rmatte> for the Linux device class, yes
[10-Jul-2009 13:37:59] <eidolon> ok
[10-Jul-2009 13:38:08] <rmatte> I'm just following what the guide says
[10-Jul-2009 13:38:10] <rmatte> read the guide, lol
[10-Jul-2009 13:38:41] <eidolon> i am man.
[10-Jul-2009 13:38:57] <eidolon> i really dislike that he's hardcoding direct urls
[10-Jul-2009 13:39:04] <eidolon> because it interrupts the flow of what i'm looking at where.
[10-Jul-2009 13:39:15] <kisielk> rmatte: which settings do I specify that in?
[10-Jul-2009 13:39:16] <eidolon> and navigating the paths in zenoss is 90% of the difficulty in getting comfy with it, IMHO.
[10-Jul-2009 13:39:56] <rmatte> kisielk: specify what in?
[10-Jul-2009 13:40:10] <kisielk> rmatte: the default starting point
[10-Jul-2009 13:40:16] <rmatte> oh
[10-Jul-2009 13:40:17] <rmatte> one sec
[10-Jul-2009 13:41:33] <rmatte> go to Settings -> Users -> Select your username -> Fill in Network Map Start Object
[10-Jul-2009 13:42:12] <eidolon> WOOT!
[10-Jul-2009 13:42:17] * eidolon just got a critical alert.
[10-Jul-2009 13:42:25] <rmatte> eidolon: agreed, takes a while to get used to it, but once you're used to it you'll love it
[10-Jul-2009 13:42:35] <eidolon> yeah, i'm much better than i was 2 days ago
[10-Jul-2009 13:42:35] <rmatte> until they modify the whole GUI and we're back to square one
[10-Jul-2009 13:42:36] <rmatte>
[10-Jul-2009 13:42:45] <eidolon> is it in the works to do a UI rewrite?
[10-Jul-2009 13:42:59] <rmatte> yeh, that's their main focus for development right now, is revamping the GUI
[10-Jul-2009 13:43:10] <mrayzenoss> not quite a complete rewrite
[10-Jul-2009 13:43:11] <eidolon> okay... so.  next bit.  i have a nice screaming bright red alert.
[10-Jul-2009 13:43:16] <eidolon> but i dont' have email.
[10-Jul-2009 13:43:18] <eidolon> or a page.
[10-Jul-2009 13:43:29] <rmatte> I'm sure it won't be a complete rewrite, but yeh
[10-Jul-2009 13:43:37] <eidolon> pagin = true or me.
[10-Jul-2009 13:43:55] <rmatte> I'd actually like to see less Ajax, like for adding datapoints and stuff for example, half the time it doesn't even load
[10-Jul-2009 13:43:58] <mrayzenoss> With 2.4 we switched out the JavaScript libraries for extjs and added the add device wizard
[10-Jul-2009 13:44:17] <rmatte> (Maybe that was due to the problem with the Cisco ZenPack, I haven't worked with datapoints since I fixed that)
[10-Jul-2009 13:44:26] <mrayzenoss> for King Crab we're upgrading to Zope 2.11 and the Event Console pages will get rewritten
[10-Jul-2009 13:44:30] <eidolon> ajax is very tricky to work with, and zenoss does it 'okay'. much of the feedback and failpoints that HTTP solved ages ago are breaking again when folks start using ajax.
[10-Jul-2009 13:44:32] <eidolon> it's annoying.
[10-Jul-2009 13:44:42] <rmatte> yeh
[10-Jul-2009 13:44:50] <eidolon> oh, hey matt.  nicetameetcha.
[10-Jul-2009 13:44:54] <eidolon> 'a voice from inside!'
[10-Jul-2009 13:45:07] <rmatte> yeh, Matt is the man around these parts
[10-Jul-2009 13:45:07] <eidolon> matt, do you work with the enterprise support guys?  probably seen tickets from me
[10-Jul-2009 13:45:09] <mrayzenoss> for Stone Crab, we're planning on redoing the navigation to reduce the depth of the application, to make it easier to find stuff
[10-Jul-2009 13:45:33] <mrayzenoss> I actually don't do support at all and stay out of the Enterprise side
[10-Jul-2009 13:45:36] <rmatte> Stone Crab comes after King Crab right?
[10-Jul-2009 13:45:39] <eidolon> holy crap i just lit up most of my board.
[10-Jul-2009 13:45:40] <mrayzenoss> Yes
[10-Jul-2009 13:45:45] <eidolon> disk space warnings left and right.
[10-Jul-2009 13:45:48] <mrayzenoss> rmatte: Yes
[10-Jul-2009 13:45:49] <rmatte> So stone crab is maybe a year away?
[10-Jul-2009 13:45:58] <mrayzenoss> we've been pretty good about every 6 months
[10-Jul-2009 13:46:25] <mrayzenoss> 2.2 was April, 2.3 was November, 2.4 was May
[10-Jul-2009 13:46:34] <rmatte> eidolon: hmmmm, what'd you do right before that?
[10-Jul-2009 13:46:34] <mrayzenoss> King Crab is supposed to be Octoberish
[10-Jul-2009 13:46:47] <mrayzenoss> Stone Crab Aprilish
[10-Jul-2009 13:46:55] <eidolon> rmatte: hmm? not sure what you mean?
[10-Jul-2009 13:47:07] * eidolon ponders sending a resume to zenoss
[10-Jul-2009 13:47:12] <rmatte> yeh, so close enough to a year
[10-Jul-2009 13:47:14] <rmatte> 10 months or so
[10-Jul-2009 13:47:30] <rmatte> eidolon: before everything lit up, what was the last change you made?
[10-Jul-2009 13:47:34] <eidolon> hmmm
[10-Jul-2009 13:47:44] <rmatte> you set it to 0.95 right?
[10-Jul-2009 13:47:45] <rmatte> not 95?
[10-Jul-2009 13:47:46] <eidolon> changed the zproperty from 1.0 to 0.95
[10-Jul-2009 13:47:52] <eidolon> but
[10-Jul-2009 13:47:56] <eidolon> i also changed the script
[10-Jul-2009 13:48:00] <eidolon> so it could have been either one
[10-Jul-2009 13:48:12] <eidolon> bt i'm getting warnings across most of my servers - valid warnings.
[10-Jul-2009 13:48:16] <rmatte> well, the previous script didn't seem to do anything at all for the filesystem issues
[10-Jul-2009 13:48:30] <eidolon> awesome
[10-Jul-2009 13:48:40] <eidolon> i'm bothered that we're not getting alerts though
[10-Jul-2009 13:48:45] <eidolon> i mean, the dashboard is lit up
[10-Jul-2009 13:48:48] <eidolon> but no emails or pages.
[10-Jul-2009 13:48:50] <kisielk> rmatte: is there no way to set it globally?
[10-Jul-2009 13:49:07] <rmatte> kisielk: not that I've seen
[10-Jul-2009 13:49:23] <rmatte> kisielk: the idea is that each user is going to have their own preference
[10-Jul-2009 13:49:33] <eidolon> you knwo what i'd really wish to see?
[10-Jul-2009 13:49:36] <eidolon> a 'system log'
[10-Jul-2009 13:49:38] <eidolon> stuff goign on.
[10-Jul-2009 13:49:42] <eidolon> did zernoss send mail?
[10-Jul-2009 13:49:42] <rmatte> kisielk: network guys will want to start at a core router where as server guys will want to start at a device in their server cluster
[10-Jul-2009 13:49:48] <eidolon> i suppose i could look in /var/log/sendmail
[10-Jul-2009 13:49:58] <kisielk> hm okay. well I set it to my core router and get a blank map, heh
[10-Jul-2009 13:50:03] <rmatte> $ZENHOME/log
[10-Jul-2009 13:50:05] <rmatte> check those
[10-Jul-2009 13:50:17] <kisielk> ah, have to use the hostname
[10-Jul-2009 13:50:29] <rmatte> you should be able to use IP
[10-Jul-2009 13:50:32] <rmatte> I do it all the time
[10-Jul-2009 13:50:42] <rmatte> but if hostname works, whatever
[10-Jul-2009 13:50:51] <rmatte> I've gotta step away for a bit, I'll be back
[10-Jul-2009 13:51:42] <maxamillion> in the zenoss there is a report that will tell the percentage of availability, I was wondering where that is stored so that I may generate a differently formatted report for my boss
[10-Jul-2009 13:51:55] <maxamillion> in zenoss*
[10-Jul-2009 13:53:59] <mrayzenoss> maxamillion: http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch12s08.html for custom reports
[10-Jul-2009 13:54:15] <mrayzenoss> or if you just want to tweak the report, copy http://dev.zenoss.org/trac/browser/trunk/Products/ZenReports/reports/Performance_Reports/Availability_Report.rpt
[10-Jul-2009 13:54:19] <mrayzenoss> and edit it directly
[10-Jul-2009 13:54:42] <maxamillion> ah, awesome ....... thank you
[10-Jul-2009 13:54:50] <venturaville> doesn't work before 2.4 though to edit it directly.............
[10-Jul-2009 13:55:13] <venturaville> at least in zope
[10-Jul-2009 13:59:06] <maxamillion> is there by any chance a quick reference for the query syntax?
[10-Jul-2009 14:01:55] <eidolon> meh. email 'test' buttosn work, but i'm still not seeing any actual zenoss mails from real live alerts.
[10-Jul-2009 14:02:19] <eidolon>  /var/log/maillog definitely shows no mail going out.  so ti's in zenoss
[10-Jul-2009 14:02:25] <eidolon> (but did show the test mail)
[10-Jul-2009 14:21:04] <eidolon> so.  any suggestiosn from folks ont he notifications?
[10-Jul-2009 14:21:13] <eidolon> sort of sad having a monitoring system that can't tell us when there's a problem
[10-Jul-2009 14:23:31] <rmatte> we tried manually sending reports from Zenoss the other day via email and no emails showed up
[10-Jul-2009 14:23:41] <rmatte> the emailing seems to be hit and miss sometimes
[10-Jul-2009 14:24:47] <rmatte> eidolon: I can't really help you there, I've never used them
[10-Jul-2009 14:24:57] <rmatte> we have zenoss integrated with a ticketing system
[10-Jul-2009 14:25:05] <rmatte> so the ticketing system handles the notifications
[10-Jul-2009 14:26:29] <rmatte> have you tried restarting Zenoss?
[10-Jul-2009 14:27:37] <mrayzenoss> eidolon: you said you'd inherited the system, are there already alert schedules setup? Are the emails going to someone else?
[10-Jul-2009 14:29:51] <eidolon> mrayzenoss: we were getting CLEAR and other alerts up until about a week ago
[10-Jul-2009 14:30:02] <eidolon> right now i'm looking at the 'admin' schedule
[10-Jul-2009 14:30:06] <eidolon> and it should be wide open.
[10-Jul-2009 14:30:28] <eidolon> i have not tried a restart.
[10-Jul-2009 14:30:36] <eidolon> i'm not even sure how to do a restart
[10-Jul-2009 14:31:09] <mrayzenoss> log in as the 'zenoss' user and do a 'zenoss restart'
[10-Jul-2009 14:31:24] <eidolon> mmkay.
[10-Jul-2009 14:32:00] <eidolon> erstarting.
[10-Jul-2009 14:32:01] <eidolon> er
[10-Jul-2009 14:32:03] <eidolon> restarting
[10-Jul-2009 14:32:14] <rmatte> to restart ssh to the box, become root, then do "su - zenoss" and then "zenoss stop" wait a second then do "zenoss start", you can also do "zenoss restart", but I prefer to just stop and start. You can also do "zenoss status" to make sure all the daemons are running.
[10-Jul-2009 14:32:49] <eidolon> Daemon: zenactions stopping...
[10-Jul-2009 14:32:49] <eidolon> already stopped
[10-Jul-2009 14:32:53] <eidolon> is that relevant?
[10-Jul-2009 14:32:56] <mrayzenoss> bingo
[10-Jul-2009 14:32:58] <eidolon> heh
[10-Jul-2009 14:33:12] <eidolon> i wonder if we have a backlog of msgs.
[10-Jul-2009 14:33:16] * eidolon braces.
[10-Jul-2009 14:33:22] <eidolon> FOOM!!!!!!!!!!!!!
[10-Jul-2009 14:33:26] <eidolon> "yes.  yes we did."
[10-Jul-2009 14:34:02] <rmatte> another Zenoss system saved by the Zen Masters!
[10-Jul-2009 14:34:10] <mrayzenoss> heh
[10-Jul-2009 14:34:38] <mrayzenoss> FYI, I'm installing Zenoss on a Dell Mini right now, much faster than my dying PowerBook
[10-Jul-2009 14:34:46] <rmatte> nice
[10-Jul-2009 14:34:46] <eidolon> i'm listenign to pagers go off all around me
[10-Jul-2009 14:34:53] <rmatte> haha
[10-Jul-2009 14:34:55] <rmatte> awesome
[10-Jul-2009 14:35:00] <eidolon> the rest of the hosted ops team is getting the notificatiosn that were backlogged
[10-Jul-2009 14:35:05] <rmatte> people should use blackberries in this day and age
[10-Jul-2009 14:35:11] <mrayzenoss> eidolon: perhaps we should have put it into a maintenance window before doing this
[10-Jul-2009 14:35:23] <eidolon> wen i say 'pagers' i mean 'whatever portable electronic communciation device they ahve' - i have an iphone.
[10-Jul-2009 14:35:33] <rmatte> ah
[10-Jul-2009 14:35:35] <eidolon> mrayzenoss: eh.  i'm the guy in charge of monitoring
[10-Jul-2009 14:35:37] <rmatte> so do I
[10-Jul-2009 14:35:44] * eidolon is a god.
[10-Jul-2009 14:35:54] <mrayzenoss> heh
[10-Jul-2009 14:36:06] <rmatte> yeh, I'm in the same boat as you
[10-Jul-2009 14:36:15] <rmatte> what do you monitor, out of curiosity?
[10-Jul-2009 14:36:17] <rmatte> all internal?
[10-Jul-2009 14:36:18] <eidolon> man the pages on the iphone are butt-assed ugly though.
[10-Jul-2009 14:36:37] <eidolon> the Con't (5 of 8) thing - because it's trying to break it up into short msgs is annoying
[10-Jul-2009 14:36:48] <eidolon> i have 2 zenoss installs - the internal one is on hold while i get a handle on this.
[10-Jul-2009 14:36:57] <eidolon> but i'm monitoring a production environment of about 150+ machines.
[10-Jul-2009 14:37:51] <rmatte> there are "Plain Text" and "Send clear messages" options, maybe one of those would keep it from breaking it up
[10-Jul-2009 14:37:58] <rmatte> ah cool
[10-Jul-2009 14:38:08] <rmatte> we're an MSP so we monitor devices for other companies
[10-Jul-2009 14:38:27] <rmatte> my biggest Zenoss install right now is monitoring almost 350 devices
[10-Jul-2009 14:38:42] <ckrough> MSP = ?
[10-Jul-2009 14:38:51] <rmatte> Managed Services Provider
[10-Jul-2009 14:38:51] <eidolon> *nodndnod* we're a content provider for cell phones. we provide downloadable music for hansets, so not only do we have a raft of servers handling requests from cell phones, we have a BIG ASSED netapp holding all teh music.
[10-Jul-2009 14:38:55] <eidolon> several hundred terabytes.
[10-Jul-2009 14:38:55] <ckrough> ah
[10-Jul-2009 14:38:57] <rmatte> we're a NOC, basically
[10-Jul-2009 14:39:11] <ckrough> same
[10-Jul-2009 14:39:14] <eidolon> can i 'simulate' a page?
[10-Jul-2009 14:39:23] <eidolon> ala, i fiddle my config, can i force a pager notification so i can see it?
[10-Jul-2009 14:39:27] <rmatte> netapps are great, wish we had one, we just have a crappy HP SAN right now
[10-Jul-2009 14:39:29] <ckrough> eidolon: you can send a fake event
[10-Jul-2009 14:39:39] <eidolon> we have 2 netapps, a sunstore, and 3 other SANs
[10-Jul-2009 14:39:41] <eidolon> the netapps kick ass.
[10-Jul-2009 14:39:47] <eidolon> ckrough: how?
[10-Jul-2009 14:39:49] <rmatte> we have a sunstore sitting in our lab
[10-Jul-2009 14:39:53] <ckrough> eidolon: http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch02s03.html   3.1.2
[10-Jul-2009 14:40:00] <eidolon> the sunstore bites it.  it fails miserably under heavy load
[10-Jul-2009 14:40:16] <rmatte> we turned it off though since we don't have A/C in the lab right now and when we initially turned it on it was like miami beach in there
[10-Jul-2009 14:40:41] <eidolon> heh
[10-Jul-2009 14:40:58] <eidolon> hee hee
[10-Jul-2009 14:41:08] * eidolon listens to my manager downt he hall, hearing his phone start pinging.
[10-Jul-2009 14:41:18] <eidolon> "Yayyy!  thank yoU! n ow i can get woken up at 3am again!'
[10-Jul-2009 14:41:44] <rmatte> the client with 350 devices is fun since they have devices in the U.S, Mexico, Australia, and all over Europe, they used to have a site in Korea too, but it was decommed recently
[10-Jul-2009 14:42:00] <rmatte> http://dmon.org/graphics/zenoss.png
[10-Jul-2009 14:42:09] <rmatte> there's the google maps screenshot of that client's network
[10-Jul-2009 14:42:20] <eidolon> kewl.
[10-Jul-2009 14:42:34] <eidolon> what's 'Core'?
[10-Jul-2009 14:42:43] <rmatte> the free version of Zenoss
[10-Jul-2009 14:42:49] <eidolon> oops,l not that button.
[10-Jul-2009 14:42:50] <rmatte> we don't have an enterprise license
[10-Jul-2009 14:42:56] <rmatte> I just work my magic with what we have
[10-Jul-2009 14:43:15] <eidolon>
[10-Jul-2009 14:43:33] <rmatte> it's working pretty well, just having to overcome some monitoring hurdles here and there
[10-Jul-2009 14:43:51] <rmatte> I've got something patched together for now, but for stuff like windows monitoring I need to eventually write a proper collector plugin
[10-Jul-2009 14:44:05] <rmatte> since it's just using command based datapoints with a bunch of bash scripts I wrote up
[10-Jul-2009 14:47:18] <eidolon> hee hee
[10-Jul-2009 14:47:19] <rmatte> bah, we're getting flooded with syslogs from this one router because of a port config issue, I'm about ready to fly down to the states and start beating that company's network team in to submission. This has been happening since yesterday and we've notified them of it like ten times now
[10-Jul-2009 14:47:24] * eidolon just created a 'false' event.
[10-Jul-2009 14:47:37] <eidolon> component: clothing, eventclass: /Unknown, summary: My shirt is on fire!
[10-Jul-2009 14:47:47] <rmatte> I'm almost ready to just move it in to Decommed state
[10-Jul-2009 14:47:49] <rmatte> >:)
[10-Jul-2009 14:47:49] * eidolon acknowledges that
[10-Jul-2009 14:48:38] <rmatte> good thing our ticketing integration auto-correlates to the existing ticket
[10-Jul-2009 14:53:50] <etank> any way to get zenoss to hook into rt like it does rememdy?
[10-Jul-2009 14:54:30] <etank> we dont use remedy. we use crappy TFS. rt would be an improvement
[10-Jul-2009 15:08:32] <mrayzenoss> etank: I'm not super familiar with RT, but I assume you can send it emails in particular format and it will turn them into a ticket?
[10-Jul-2009 15:08:40] <mrayzenoss> or perhaps it has an API?
[10-Jul-2009 15:09:00] <etank> yeah i think you can do that
[10-Jul-2009 15:09:19] <mrayzenoss> I think the only magic for the Remedy integration is that it also modifies the events in Zenoss to link into Remedy
[10-Jul-2009 15:09:32] <mrayzenoss> but I'm not positive on that, since I don't use it
[10-Jul-2009 15:09:54] <etank> right. i think that with remedy it is two way.
[10-Jul-2009 15:10:13] <etank> update the ticket and the log is updated. update the log and the ticket is updated.
[10-Jul-2009 15:13:31] <mrayzenoss> you can customize the message format for outbound emails to match whatever RT is expecting inbound
[10-Jul-2009 15:13:58] <mrayzenoss> and you can configure Zenoss to accept incoming emails from another system to take the changes from RT
[10-Jul-2009 15:14:44] <mrayzenoss> here's a HOWTO for processing inbound email: http://www.zenoss.com/Members/shaskell/email-ack/email-acknowledgement-postfix-procmail-and-python
[10-Jul-2009 15:15:08] <mrayzenoss> err, that's just for email acking events
[10-Jul-2009 15:18:21] <mrayzenoss> rmatte: mind if I repost that image on the Zenoss Flickr?
[10-Jul-2009 15:18:45] <mrayzenoss> I can blur your login out
[10-Jul-2009 15:31:45] <rmatte> we have Zenoss passing info to a perl script which uses the OTRS API to create a ticket in OTRS when an event comes in
[10-Jul-2009 15:31:52] <rmatte> the script also handles correlation and such
[10-Jul-2009 15:31:59] <rmatte> we're going to make it much more intelligent in the future
[10-Jul-2009 15:32:51] <rmatte> and we are patiently waiting for the next version of OTRS
[10-Jul-2009 15:34:39] <mrayzenoss> how's OTRS compare to RT? I have no real frame of reference, I'm just curious because people ask me for Open Source ticketing systems
[10-Jul-2009 15:36:38] <mrayzenoss> also, we're planning on putting up a page eventually for Zenoss Integrations at some point, so if you want to put your OTRS integration there, it would have a place to live
[10-Jul-2009 15:37:07] <mrayzenoss> to go with the various Zenoss mashups and other integrations (iPhone, MindTouch, Network Weathermap, etc.)
[10-Jul-2009 15:38:27] <ckrough> so there is something Im not grokking, if I have a devive template, and there is a datasource (snmp) which isnt consistently available on devices bound to that template, do I just deal with the missing data, or is there a way that zenoss dymanically enables/disables that datasource during modelling?
[10-Jul-2009 15:40:06] <rmatte> OTRS is pretty good, and easy to customize (provided you know perl)
[10-Jul-2009 15:40:11] <rmatte> and it's web based, which is a plus
[10-Jul-2009 15:40:18] <rmatte> we've customized the hell out of ours
[10-Jul-2009 15:43:05] <mrayzenoss> ckrough: like this? trac-zenpacks/browser/zenpacks/ZenPacks.community.HPSIMMonitor/ZenPacks/community/HPSIMMonitor/modeler/plugins/community/snmp/HPDeviceTemplateMap.py
[10-Jul-2009 15:43:32] <mrayzenoss> I believe the HPSIM ZenPack applies a different template depending on what models
[10-Jul-2009 15:45:59] <rmatte> man, the snmp process monitoring can be pretty shotty, sometimes it's dead on, other times it'll show a process as being down when it's not down and then won't show it as being back up until I do a reindex() and commit() in zendmd
[10-Jul-2009 15:46:05] <rmatte> it's a bit annoying
[10-Jul-2009 16:00:26] <mrayzenoss> ckrough: FYI, the Troubleshooting Zenoss Daemons section is going back into the Admin Guide for next release. It had been planned to go into the Knowledge Base, but now it'll be in both locations
[10-Jul-2009 16:04:14] <rmatte> bless the person who wedged the door downstairs open with a stool, it's supposed to lock at 5:00pm, at which point I would not be able to scure these three deliciously cold beverages from the magical beverage dispenser
[10-Jul-2009 16:04:24] <rmatte> secure*
[10-Jul-2009 16:07:15] <maxamillion> zenoss is supposed to have a database in mysql correcT?
[10-Jul-2009 16:07:17] <maxamillion> correct*
[10-Jul-2009 16:07:43] <mrayzenoss> yeah, depending on how you installed it, the installer puts down it's own DB
[10-Jul-2009 16:08:01] <maxamillion> mrayzenoss: oh?
[10-Jul-2009 16:08:23] <mrayzenoss> the stack installers include a MySQL instance
[10-Jul-2009 16:09:03] <maxamillion> mrayzenoss: I thought I setup mysql before, but now zenoss is showing an error (I installed this thing over a year ago and our operations team works on it now) ... today all the sudden it shows an error on the dashboard page and when I do "mysql -u root -p" and 'show databases;' there's no zenoss database
[10-Jul-2009 16:13:01] <rmatte> maxamillion: Zenoss uses it's own MySQL server to store events and everything else (Settings, Devices, etc...) are stored in the Zope database.
[10-Jul-2009 16:13:21] <mrayzenoss> maxamillion: so if you had lost your Zenoss MySQL instance, you'd have no events
[10-Jul-2009 16:15:30] <maxamillion> mrayzenoss: I don't appear to have any events
[10-Jul-2009 16:15:46] <maxamillion> so is it safe to assume the database is gone?
[10-Jul-2009 16:18:24] <mrayzenoss> doesn't sound good
[10-Jul-2009 16:18:30] <mrayzenoss> did someone stop the DB?
[10-Jul-2009 16:19:12] <mrayzenoss> I'm recreating the effect on my home install to see what happens
[10-Jul-2009 16:19:32] <maxamillion> mrayzenoss: I haven't the slightest idea ... if someone did, it wasn't me
[10-Jul-2009 16:19:57] <mrayzenoss> seems to just hang on "Loading..." on the Event Console
[10-Jul-2009 16:20:00] <maxamillion> mrayzenoss: where does that zenoss mysql instance live? does it have an sys v init style control script inside the zenoss "world"?
[10-Jul-2009 16:20:09] <maxamillion> mrayzenoss: yeah, that's what i've got going on
[10-Jul-2009 16:20:17] <maxamillion> just sitting there loading forever
[10-Jul-2009 16:20:42] <mrayzenoss> what platform are you running on?  which installer?
[10-Jul-2009 16:21:03] <maxamillion> mrayzenoss: RHEL5, but we used the non-rpm installer if I remember correctly
[10-Jul-2009 16:21:17] <maxamillion> everything is in /usr/local/zenoss/zenoss if that helps
[10-Jul-2009 16:21:27] <maxamillion> (I seem to remember the rpm installing in /opt/)
[10-Jul-2009 16:32:00] <maxamillion> mrayzenoss: http://fpaste.org/paste/18294 <--- I think this might be an issue, do you know what user.group that needs to be?
[10-Jul-2009 16:43:38] <dberger> if you write a nagios-style plugin, will it ignore output printed to stderr?
[10-Jul-2009 16:47:56] <mrayzenoss> maxamillion: what happens if you try to restart the database? /etc/init.d/zenoss-stack restart
[10-Jul-2009 16:48:07] <mrayzenoss> dberger: I believe it only takes stdout
[10-Jul-2009 16:49:05] <mrayzenoss> dberger: and note this ticket I opened yesterday: http://dev.zenoss.org/trac/ticket/5294
[10-Jul-2009 16:49:08] <maxamillion> mrayzenoss: its restarting ... we'll find out in a moment
[10-Jul-2009 16:49:40] <maxamillion> mrayzenoss: http://fpaste.org/paste/18297 <--- it fails for one reason or another
[10-Jul-2009 16:49:58] <dberger> mrayzenoss, that's really good to know
[10-Jul-2009 16:50:19] <mrayzenoss> maxamillion: try doing a zenoss-stack stop and seeing what's running
[10-Jul-2009 16:50:31] <mrayzenoss> make sure everything is really stopped
[10-Jul-2009 16:50:39] <maxamillion> mrayzenoss: will do
[10-Jul-2009 16:50:42] <mrayzenoss> which user are you using?
[10-Jul-2009 16:50:51] <mrayzenoss> dberger: yeah, I learned that yesterday
[10-Jul-2009 16:50:54] <maxamillion> avahi 4025 0.0 0.0 2580 1372 ? Ss Jun12 0:00 avahi-daemon: running [zenoss.local]
[10-Jul-2009 16:51:24] <maxamillion> mrayzenoss: that's the only thing that appears to be running, but that is apparently in respect to the fact that zenoss is the hostname
[10-Jul-2009 16:52:24] <maxamillion> mrayzenoss: I think its starting up correctly this time, the permissions were wrong in /usr/local/zenoss/mysql for one reason or another
[10-Jul-2009 16:52:37] <mrayzenoss> weird
[10-Jul-2009 16:54:00] <rmatte> chown -R root:mysql /usr/local/zenoss/mysql
[10-Jul-2009 16:54:04] <rmatte>
[10-Jul-2009 16:54:50] <maxamillion> rmatte: I did 'chown mysql.mysql /usr/local/zenoss/mysql' and its working
[10-Jul-2009 16:55:58] <rmatte> hmmm, just found a rough collector plugin that cgibbons had written up for me like 5 or 6 months ago
[10-Jul-2009 16:56:03] <rmatte> totally forgot I had it
[10-Jul-2009 16:56:58] <rmatte> maxamillion: as long as it's working, Zenoss perms should be:
[10-Jul-2009 16:57:00] <rmatte> chown -R zenoss:users /usr/local/zenoss
[10-Jul-2009 16:57:00] <rmatte> chown -R root:mysql /usr/local/zenoss/mysql
[10-Jul-2009 16:57:00] <rmatte> chown root:zenoss /usr/local/zenoss/zenoss/bin/zensocket
[10-Jul-2009 16:57:01] <rmatte> chmod 04750 /usr/local/zenoss/zenoss/bin/zensocket
[10-Jul-2009 16:57:39] <rmatte> just in case your perms get royaly screwed at some point
[10-Jul-2009 17:01:47] <maxamillion> rmatte: thanks, I will make a note of that for future reference
[10-Jul-2009 17:01:56] <rmatte> np
[10-Jul-2009 17:14:42] <maxamillion> thanks again!!!
[10-Jul-2009 17:20:12] <rmatte> well, I guess they'll have to stick with their current version or upgrade
[10-Jul-2009 17:21:13] <mrayzenoss> quick question, anyone have opinions on posting to the Zenoss blog? Should I put stuff up now or wait until Monday morning?
[10-Jul-2009 17:28:43] <kisielk> opinions on what?
[10-Jul-2009 17:29:17] <mrayzenoss> just wondering if and when people read the blog, should I hold off on publishing something in hopes of getting more attention
[10-Jul-2009 17:29:42] <mrayzenoss> went ahead and published it: http://blog.zenoss.com/2009/07/10/help-zenoss-expand-our-testing/
[10-Jul-2009 17:30:15] <mrayzenoss> Our test plans are now publicly accessible, so you can add tests for our QA to use
[10-Jul-2009 17:30:26] <mrayzenoss> or do your own testing at home
[10-Jul-2009 17:30:34] <kisielk> just post any time
[10-Jul-2009 17:30:52] <kisielk> I (personally) don't care when, I'll eventually see it in my aggregator when I get around to reading it
[10-Jul-2009 17:30:58] <mrayzenoss> me too
[10-Jul-2009 17:32:52] <mrayzenoss> hopefully people have something they think needs more coverage, and they'll submit a test plan
[10-Jul-2009 17:34:14] <rmatte> lol, my friend has a car and a harley davidson, I ask him if I can get a ride home, he says "ok, but you'd better hold on tight", then I'm like "oh, so you have the bike today?" and he says "no" and points to his feet, and he has rollerblades on
[10-Jul-2009 17:34:29] <rmatte> bastard
[10-Jul-2009 17:34:41] <rmatte> anyways, time to start walking to the bus, talk to you guys on Monday
[10-Jul-2009 17:35:25] <mrayzenoss> later
[11-Jul-2009 00:00:46] [disconnected at Sat Jul 11 00:00:46 2009]
[11-Jul-2009 00:00:46] [connected at Sat Jul 11 00:00:46 2009]
[11-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[11-Jul-2009 08:51:40] <abdo> hi all
[11-Jul-2009 08:51:56] <abdo> from where RRD graph read its intial command
[11-Jul-2009 08:52:23] <abdo> for example: -F
[11-Jul-2009 08:52:23] <abdo> -E
[11-Jul-2009 08:52:23] <abdo> --height=100
[11-Jul-2009 08:52:23] <abdo> --width=500
[11-Jul-2009 08:52:23] <abdo> --lower-limit=0
[11-Jul-2009 08:52:24] <abdo> --rigid
[11-Jul-2009 08:52:26] <abdo> --vertical-label=percentage
[11-Jul-2009 08:52:28] <abdo> --upper-limit=100
[11-Jul-2009 08:52:30] <abdo> DEF:cpu-raw=rrdPath/cpu_cpu.rrd:ds0:AVERAGE
[11-Jul-2009 08:52:32] <abdo> CDEF:cpu-limit=cpu-raw,800,GT,UNKN,cpu-raw,IF
[11-Jul-2009 08:52:34] <abdo> CDEF:cpu=cpu-limit
[11-Jul-2009 08:52:36] <abdo> AREA:cpu-limit#00cc00ff:Used
[11-Jul-2009 08:52:38] <abdo> GPRINT:cpu-limit:LAST:cur\:%0.2lf%%
[11-Jul-2009 08:52:40] <abdo> GPRINT:cpu-limit:AVERAGE:avg\:%0.2lf%%
[11-Jul-2009 08:52:42] <abdo> GPRINT:cpu-limit:MAX:max\:%0.2lf%%\j
[11-Jul-2009 08:52:45] <abdo> from where import these commands?
[11-Jul-2009 09:50:52] <abdo> the same names, same traffic, what a lazy room!
[11-Jul-2009 11:08:59] <abdo> are hhere a live one?
[11-Jul-2009 11:09:05] <abdo> are here a live one?
[11-Jul-2009 18:10:27] <etank> for some reason a scheduled maintenance window that i have set on one server did not work today. i just got about 2500 emails from zenoss and then another 2500 when i cleared them :/
[12-Jul-2009 00:00:46] [disconnected at Sun Jul 12 00:00:46 2009]
[12-Jul-2009 00:00:46] [connected at Sun Jul 12 00:00:46 2009]
[12-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[12-Jul-2009 18:59:09] <hackeron> hey, I'm trying to install zenoss using the debian/ubuntu deb and I'm getting: Unknown error while running id zenoss | sed -e s/uid=//g -e s/\(.*//g
[12-Jul-2009 18:59:33] <hackeron> if I run id zenoss | sed -e s/uid=//g -e s/\(.*//g - I get 1002
[12-Jul-2009 19:01:58] <hackeron> any ideas?
[12-Jul-2009 19:11:18] <hackeron> and if I try the installer, I get a bunch of errors like Unable to create symbolic link /usr/local/zenoss/common/lib/libpangocairo-1.0.so.0 -> libpangocairo-1.0.so.0.2203.1
[12-Jul-2009 19:11:21] <hackeron>
[13-Jul-2009 00:00:46] [disconnected at Mon Jul 13 00:00:46 2009]
[13-Jul-2009 00:00:46] [connected at Mon Jul 13 00:00:46 2009]
[13-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[13-Jul-2009 05:08:48] <Dr_Agon> hi all
[13-Jul-2009 05:09:23] <venturaville> howdy ... forgot I had auto irc login set :-P
[13-Jul-2009 05:09:30] <venturaville> 6am change controls ... yikes
[13-Jul-2009 05:10:44] <Dr_Agon> do anybody know why Devices/Server/Linux could have disappeared from my zenoss ?
[13-Jul-2009 05:10:49] <Dr_Agon> =)
[13-Jul-2009 05:11:21] <venturaville> it usually has to be either manually deleted or maybe you had a zenpack attached to it that you uninstalled?
[13-Jul-2009 05:11:32] <Dr_Agon> it worked well and in 20 min i've logged into my zenoss again and haven't found most of my devices
[13-Jul-2009 05:11:49] <Dr_Agon> mmm
[13-Jul-2009 05:12:33] <Dr_Agon> how can i check what zenpacs are attached to devices? and is there any way to restore my devices
[13-Jul-2009 05:12:35] <Dr_Agon> ?
[13-Jul-2009 05:24:26] <venturaville> a lot of the performance data is probably still there
[13-Jul-2009 05:24:44] <venturaville> as to restoring them, you will probably need to make the group and readd the devices manually :-/
[13-Jul-2009 05:26:26] <Dr_Agon> heh, thx - starting to readding devices
[13-Jul-2009 05:27:30] <venturaville> if you look in settings -> ZenPacks ... and go to to individual zenpacks, they will usually tell you what they provide ... if they have something like /Devices/ in them, then they think they own some of those
[13-Jul-2009 05:28:06] <venturaville> if there is nothing else in that class when the zenpack is uninstalled, then it will usually delete it (assuming it is created in objects.xml in the zenpack)
[13-Jul-2009 05:28:34] <venturaville> I dislike that behaviour enough that I am probably going to move the instances where I do that in my zenpacks to a create only method
[13-Jul-2009 05:29:34] <Dr_Agon> i see
[13-Jul-2009 06:30:41] <zaafouri> hi all
[13-Jul-2009 06:30:55] <zaafouri> I'm using zenoss and I like it
[13-Jul-2009 06:31:05] <zaafouri> it's powerfull =)
[13-Jul-2009 06:53:23] <ckrough_> Yup
[13-Jul-2009 06:53:27] ckrough_ is now known as ckrough
[13-Jul-2009 08:38:13] <etank> anyone ever have an issue where you set up a maint window and you still got error messages?
[13-Jul-2009 09:51:02] <ckrough> anyone know if retroactively setting RRD Min or Max requires rebuildin the .rrd files?
[13-Jul-2009 10:24:00] <ckrough> is there a way to pause zenperf updates just for a particular device while that device's rrds are being modified?
[13-Jul-2009 10:24:02] <ErikRose> Is it normal to get occasional false alarms about SNMP agents or IPs being down, or is my network just flaky?
[13-Jul-2009 10:24:25] <ErikRose> ckrough: You could probably switch the box's state to Decommissioned.
[13-Jul-2009 10:26:44] <straterra> I'm trying to build zenoss on Gentoo..and it keeps dying on WMI.. here is the build output. http://downloads.fuhell.com/public/zenbuild.log
[13-Jul-2009 10:27:09] <raw> straterra: you could just use a stack installer
[13-Jul-2009 10:27:32] <straterra> Doesn't that include mysql and all of that?
[13-Jul-2009 10:28:38] <straterra> Also, I don't see a stack installer for Gentoo
[13-Jul-2009 10:30:03] <raw> yeh, it does include all of that, the stack installer is pretty universal, I'd put money that the ubuntu one would work fine on Gentoo
[13-Jul-2009 10:30:33] <raw> The stack installer includes all of the Zenoss dependencies
[13-Jul-2009 10:30:40] <raw> including the specific version of Python needed
[13-Jul-2009 10:31:30] <straterra> I already HAVE all of the dependencies
[13-Jul-2009 10:31:39] <straterra> I don't want two of everything..especially a SQL server
[13-Jul-2009 10:32:19] <raw> well then, let's see....
[13-Jul-2009 10:33:20] <raw> it's complaining about missing files, so it's probably a file structure issue, either that or a variable isn't set
[13-Jul-2009 10:33:46] <raw> do you have winexe, wmic, and *async_wmi_lib.so.0* files?
[13-Jul-2009 10:35:14] <straterra> Where would I find those?
[13-Jul-2009 10:35:14] <straterra> You mean, just on the file system..or in the build directory?
[13-Jul-2009 10:35:14] <raw> configure.ac:96: error: m4_defn: undefined macro: _AC_SUBST_VARS
[13-Jul-2009 10:35:14] <raw> hmmm
[13-Jul-2009 10:35:28] <raw> in the build directory
[13-Jul-2009 10:36:04] <raw> that doesn't look like the actual issue though
[13-Jul-2009 10:36:31] <straterra> I googled and found suggestions to comment out like three lines in a file
[13-Jul-2009 10:36:43] <straterra> http://forums.zenoss.com/viewtopic.php?p=28232
[13-Jul-2009 10:36:43] <raw> which 3 lines?
[13-Jul-2009 10:37:15] <straterra> AC_REQUIRE([AC_GNU_SOURCE])
[13-Jul-2009 10:37:15] <straterra> AC_REQUIRE([AC_AIX])
[13-Jul-2009 10:37:16] <straterra> AC_REQUIRE([AC_MINIX])
[13-Jul-2009 10:37:26] <straterra> In AC_REQUIRE([AC_GNU_SOURCE])
[13-Jul-2009 10:37:26] <straterra> AC_REQUIRE([AC_AIX])
[13-Jul-2009 10:37:29] <straterra> errr
[13-Jul-2009 10:37:31] <straterra> stupid copy
[13-Jul-2009 10:37:40] <raw> lol
[13-Jul-2009 10:37:41] <straterra> in file build/wmi-1.1.8/Samba/source/lib/replace/autoconf-2.60.m4
[13-Jul-2009 10:38:09] <raw> it's worth a shot
[13-Jul-2009 10:38:20] <straterra> I already did..
[13-Jul-2009 10:38:25] <straterra> And it didn't do anything helpful
[13-Jul-2009 10:38:36] <raw> no difference in the message you're getting?
[13-Jul-2009 10:38:41] <straterra> Correct
[13-Jul-2009 10:39:42] <raw> I'd say find the file configure.ac and look for line 96
[13-Jul-2009 10:39:45] <raw> for starters
[13-Jul-2009 10:40:54] <straterra> %config = (AC_FOREACH([AC_Var], m4_defn([_AC_SUBST_VARS]), [
[13-Jul-2009 10:41:14] <raw> search for _AC_SUBST_VARS in that files
[13-Jul-2009 10:41:24] <straterra> In what file?
[13-Jul-2009 10:43:11] <raw> ok hold on, I'm seeing what I can dig up
[13-Jul-2009 10:45:56] <raw> cd to /home/zenoss/zenoss-2.4.2/build/wmi-1.2.6/Samba/source
[13-Jul-2009 10:46:08] <raw> run the following commands:
[13-Jul-2009 10:46:09] <raw> aclocal
[13-Jul-2009 10:46:09] <raw> autoconf
[13-Jul-2009 10:46:09] <raw> autoheader
[13-Jul-2009 10:46:09] <raw> automake
[13-Jul-2009 10:46:16] <raw> tell me which one fails and what the output is
[13-Jul-2009 10:50:29] <raw> building from source can be such a pain, and so time consuming. I think I'm going to move my home server from Gentoo to Ubuntu Server at some point.
[13-Jul-2009 10:51:13] <straterra> aclocal fails
[13-Jul-2009 10:51:19] <straterra> configure.ac:96: error: m4_defn: undefined macro: _AC_SUBST_VARS
[13-Jul-2009 10:51:19] <straterra> configure.ac:96: the top level
[13-Jul-2009 10:51:20] <raw> ok, and the message is?
[13-Jul-2009 10:51:25] <raw> k
[13-Jul-2009 10:52:45] <raw> sounds like your automake and related tools may not be the right versions
[13-Jul-2009 10:52:58] <raw> autoconf
[13-Jul-2009 10:54:55] <straterra> 2.63
[13-Jul-2009 10:56:54] <raw> is there a newer version or is that the latest?
[13-Jul-2009 10:57:56] <raw> emerge autoconf && emerge automake
[13-Jul-2009 10:58:16] <straterra> That's the latest
[13-Jul-2009 10:58:18] <raw> ah, that is the latest
[13-Jul-2009 10:58:23] <straterra> Latest available, that is
[13-Jul-2009 10:58:55] <raw> it looks like the script in the source is made for 2.60
[13-Jul-2009 10:59:10] <raw> though I don't know why it wouldn't work for 2.63
[13-Jul-2009 11:00:00] <raw> hmmm, it might have been hard coded to require 2.60
[13-Jul-2009 11:00:01] <straterra> I don't know
[13-Jul-2009 11:00:23] <raw> I'm just looking at: http://archives.devshed.com/forums/samba-120/samba-4-0-0tp1-teething-pains-1751463.html
[13-Jul-2009 11:03:11] <eidolon> anyone know how to set an alert NOT to be raised until X number of repetitions?
[13-Jul-2009 11:05:18] <straterra> raw: so..what do you suggest?
[13-Jul-2009 11:08:39] <raw> I'm not really sure at this point, I don't have the source in front of me to dig through...
[13-Jul-2009 11:10:26] <raw> you could use the stack installer and not use the mysql that comes with it
[13-Jul-2009 11:10:39] <raw> just create the tables in your existing mysql
[13-Jul-2009 11:10:50] <raw> and set Zenoss not to start mysql automatically
[13-Jul-2009 11:11:05] <raw> there's obviously something wrong with the autoconfig script in that source
[13-Jul-2009 11:13:27] <straterra> Yeah
[13-Jul-2009 11:13:44] <straterra> I'm not trying to be a dickhole or anything..but..uhm..why is this release tagged stable if it doesnt even compile?
[13-Jul-2009 11:14:45] raw is now known as rmatte
[13-Jul-2009 11:15:13] <rmatte> it probably compiled on a couple of test systems which have older version of autoconf or something
[13-Jul-2009 11:15:25] <ckrough> eidolon: add the condition "count > x" in the alert rule?
[13-Jul-2009 11:15:28] <etank> is there anyway to do zooming on the graphs in zenoss like what is done with zabbix?
[13-Jul-2009 11:15:30] <rmatte> I don't work for Zenoss or anything, but I'll mention the issue to them
[13-Jul-2009 11:15:43] <ckrough> eidolon: assuming it's not autoclearning in the interim
[13-Jul-2009 11:15:47] <eidolon> ckrough: oohoh. wait, where?
[13-Jul-2009 11:15:48] <etank> drag over an area on the graph and you zoom to that time frame
[13-Jul-2009 11:15:49] * eidolon looks.
[13-Jul-2009 11:16:20] <rmatte> straterra: that's technically Samba not compiling, not Zenoss itself
[13-Jul-2009 11:16:33] <eidolon> when you say 'in the alert rule' - you're talking about alerting mechanism for users themselves, yes? i was thinkin in the template of the event.
[13-Jul-2009 11:16:35] <rmatte> straterra: you could try grabbing new source for Samba and dropping it in that directory
[13-Jul-2009 11:16:41] <eidolon> yes, i'm mixing event and alert together.
[13-Jul-2009 11:16:43] <rmatte> straterra: that would probably do the trick
[13-Jul-2009 11:16:51] <ckrough> etank: that rings a bell, may be in Trac soemwhere as a request
[13-Jul-2009 11:17:06] <ckrough> eidolon: yes, in the alert rule for the user/group
[13-Jul-2009 11:17:50] <eidolon> that's not so good - because that would set it for all my alerts.
[13-Jul-2009 11:17:52] <etank> ckrough: that would sure be a nice feature to have. on of the other admins here has already asked if it is possible because you can do it with our current monitoring software.
[13-Jul-2009 11:18:01] <eidolon> i know one of my servers occasioanlly times out on HTTP hits.
[13-Jul-2009 11:18:12] <eidolon> what i want to do is say "for thsi server - only raise an alert if this event repeats 3 times."
[13-Jul-2009 11:20:06] <eidolon> (brb)
[13-Jul-2009 11:22:21] <ckrough> eidolon: on thresholds you can do that with an escalate count, havent looked at doing it on a process/service
[13-Jul-2009 11:22:56] <rmatte> straterra: and if that does work, I'll get them to update the source package with the latest samba source
[13-Jul-2009 11:33:36] <straterra> I'll try it in a bit
[13-Jul-2009 11:33:41] <straterra> or now
[13-Jul-2009 11:33:56] <straterra> Isn't that a GPL violation?
[13-Jul-2009 11:38:30] <straterra> rmatte: except..I don't see anything wmi in the tarball for the latest samba
[13-Jul-2009 11:41:27] <straterra> rmatte: Though, latest samba works ok with autoconf on this system
[13-Jul-2009 11:52:57] <rmatte> I doubt it is, as long as you are providing the full source
[13-Jul-2009 11:53:30] <straterra> I provided the full source for samba
[13-Jul-2009 12:20:48] <daMaestro> does anyone know the best way to add a threashold to an osprocess monitor?
[13-Jul-2009 12:21:18] <daMaestro> i was thinking i could copy the osproccess template to the device and then add thresholds as needed, but I would like to add the threashold to my process class
[13-Jul-2009 12:22:29] <daMaestro> i don't see a way to add a cpu utilization threshold via net-snmpd
[13-Jul-2009 12:22:42] <daMaestro> already monitoring zCountProcs and that works correctly
[13-Jul-2009 12:33:55] <straterra> rmatte: just getting latest samba4 isn't enough, i dont think
[13-Jul-2009 12:43:39] <straterra> rmatte: btw..I temporarily downgraded autoconf
[13-Jul-2009 13:12:16] <straterra> New compiling error..
[13-Jul-2009 13:13:24] <straterra> http://pastebin.ca/1493810
[13-Jul-2009 13:35:21] <rmatte> sorry, just got back
[13-Jul-2009 13:36:39] <rmatte> my god
[13-Jul-2009 13:37:04] <rmatte> ok, so heimdal is an implementation of Kerberos 5
[13-Jul-2009 13:37:47] <rmatte> no idea why that failed to compile, there's a whole wackload of references to lines in that error
[13-Jul-2009 14:03:49] <straterra> Yeah..The issue is nameser.h and nameser_compat.h
[13-Jul-2009 14:03:58] <straterra> for whatever reason
[13-Jul-2009 14:04:47] <straterra> my glibc is fine
[13-Jul-2009 14:25:08] <rmatte> yeh, that's pretty messed up
[13-Jul-2009 14:25:39] <rmatte> anyways, I need to reformat this PC and install Ubuntu on it since I'm using Sabayon and it's all messed up, so I'll be back later
[13-Jul-2009 14:26:00] <straterra> ok..thanks for the help
[13-Jul-2009 14:28:34] <etank> one last check to make sure that i am correct on this. the zenoss rpm for core and enterprise install to /opt right?
[13-Jul-2009 14:29:05] <etank> i am building a box specifically for zenoss to run on now and want to make sure i partition it accordingly
[13-Jul-2009 14:32:08] <ckrough> is it /opt or /usr/local/zenoss?
[13-Jul-2009 14:33:12] <ckrough> ah nm, my core install was the stack installer
[13-Jul-2009 14:36:11] <etank> ckrough: my stack install defaulted to /usr/local too
[13-Jul-2009 14:36:22] <etank> but the rpm goes to /opt for sure?
[13-Jul-2009 14:37:12] <ckrough> if its relocatable you can use the --prefix /opt argument
[13-Jul-2009 14:37:47] <ckrough> sry, --prefix=/opt
[13-Jul-2009 14:38:47] <ckrough> rpm -qpi blahblah.rpm | grep -i relocation
[13-Jul-2009 14:39:46] <etank> thanks
[13-Jul-2009 14:39:50] daMaestro|isBack is now known as daMaestro
[13-Jul-2009 17:39:07] <potatodemon> Howdy, I am not able to track processes with zenoss core. Could it be a problem with my snmpd.conf ?
[13-Jul-2009 17:42:02] <potatodemon> I am trying to monitor processes on Linux machines. In particular a process that has the a word "foo" as on of the args.
[13-Jul-2009 17:47:21] <rhetts> it could definitely be
[13-Jul-2009 17:48:22] <rhetts> i would try manually walking the oid and see what is returned
[13-Jul-2009 17:49:31] <potatodemon> hum, when I run 'snmpwalk -v1 -c public <IP> from a remote machine and grep for a known process name I find it. So I think snmp is reporting it. Does it sound like a zenoss config problem?
[13-Jul-2009 17:50:30] <potatodemon> and what I look at the device OS tab the OS processes are not listed. even though I have monitored checked.
[13-Jul-2009 17:50:31] <rhetts> probably. zenoss then
[13-Jul-2009 17:50:57] <rhetts> snmpwalk -v 1 -c public HOST .1.3.6.1.2.1.25.4.2.1.2
[13-Jul-2009 17:51:06] <potatodemon> I just got my server up and running like 30 mins ago, will it take longer for the machines to remote their processes etc>
[13-Jul-2009 17:51:12] <rhetts> and .1.3.6.1.2.1.25.4.2.1.4
[13-Jul-2009 17:53:59] <potatodemon> here are the results for the first http://pastie.org/544833 and second command http://pastie.org/544835 . As done from a remote host. Looks like processes are being reported.
[13-Jul-2009 17:55:44] <potatodemon> I set the process name and regex to be 'apache2', and it has not found the process yet.
[13-Jul-2009 17:58:18] <potatodemon> I had the zProperty 'zCountProcs' set to false, I just set it to true, would that have been my problem?
[13-Jul-2009 18:20:20] daMaestro|isBack is now known as daMaestro
[13-Jul-2009 18:28:29] <kisielk> potatodemon: what happens when you walk hrSWRunPath ?
[13-Jul-2009 18:28:40] <kisielk> Zenoss needs that one as well
[13-Jul-2009 18:28:57] <kisielk> and I found a bug in Net-SNMP that's *still* not fixed, at least not on my openSUSE installs
[13-Jul-2009 18:39:21] <Hunner> Hello. Does zenoss have patch/lifecycle management tools for rhel or solaris?
[13-Jul-2009 19:08:29] bzed_ is now known as bzed
[13-Jul-2009 20:50:06] bzed_ is now known as bzed
[14-Jul-2009 00:00:46] [disconnected at Tue Jul 14 00:00:46 2009]
[14-Jul-2009 00:00:46] [connected at Tue Jul 14 00:00:46 2009]
[14-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[14-Jul-2009 02:01:14] <hackeron> anyone has any ideas? < http://forums.zenoss.com/viewtopic.php?p=36902#36902
[14-Jul-2009 02:03:57] <rhetts> have you tried from source or the .deb
[14-Jul-2009 02:04:44] <hackeron> rhetts: yes, same from deb - from source it won't compile because zenoss doesn't build on more recent distributions - I was pointed to a bug about it
[14-Jul-2009 02:05:23] <hackeron> rhetts: here's what happens from source: http://ubuntuforums.org/showthread.php?t=1169535
[14-Jul-2009 02:25:55] <Dieterbe_> Hi, anyone with experience monitoring windoze boxes? please see http://forums.zenoss.com/viewtopic.php?t=10317 thanks
[14-Jul-2009 07:30:12] <Dieterbe_> is it possible to monitor if a certain process (matching name ) uses more then x amount of ram, and if so, send an alert? and preferably also monitor the ram usage of these processes in the perf graphs
[14-Jul-2009 07:30:27] <Dieterbe_> i'm looking at zenoss' process monitoring feature but it doesn't seem to support this
[14-Jul-2009 08:20:21] <tripitaka> Hi, I'm using the community edition, is it possible to run two servers, using one as a backup?
[14-Jul-2009 08:23:27] <tripitaka> ie. active/active or active/standby clustered zenhubs
[14-Jul-2009 08:25:30] <jb> you can do a HA/drbd setup
[14-Jul-2009 08:25:41] <jb> i think there is an article on the wiki somewhre
[14-Jul-2009 08:28:56] <tripitaka> cheers
[14-Jul-2009 08:29:26] <tripitaka> jb: http://www.zenoss.com/Members/netdata/create-a-highavailable-zenoss/
[14-Jul-2009 08:29:33] <jb> there ya go
[14-Jul-2009 08:29:39] <jb> i would be careful about I/O with DRBD though
[14-Jul-2009 08:29:44] <tripitaka> hmm
[14-Jul-2009 08:30:22] <tripitaka> the standby should ideally be at a remote site, I'll have to see how much data needs to be transferred
[14-Jul-2009 08:30:38] <jb> yeah, if it's semi-high latency.. drbd might not be a good option
[14-Jul-2009 08:40:21] <fiXXXerMet> Can zenoss store the client data in the mysql database, or only in the zone db?
[14-Jul-2009 08:40:35] <ckrough> what do you mean by client data
[14-Jul-2009 08:40:35] <jb> all data except for events goes in zope
[14-Jul-2009 08:40:45] <fiXXXerMet> I meant events
[14-Jul-2009 08:41:10] <ckrough> yeah events go into the database, in the 'status' and 'history' tables
[14-Jul-2009 08:41:13] <jb> evnents are stored in mysql
[14-Jul-2009 08:41:14] <ckrough> sql database that is
[14-Jul-2009 08:41:22] <fiXXXerMet> So what is in zope?
[14-Jul-2009 08:41:28] <jb> everything else
[14-Jul-2009 08:41:48] <fiXXXerMet> lol, alright - can 'everything' be put into mysql?
[14-Jul-2009 08:41:48] <ckrough> Zope is an object database, all of the monitored devices, users, rules, etc.. are objects in zope
[14-Jul-2009 08:42:27] <jb> nope
[14-Jul-2009 08:42:41] <ckrough> the zope object database is the foundation of zenoss
[14-Jul-2009 09:08:23] <BCS-Satori> I have an HP server with snmpd configured for my HP MGMT alerts using rocommunity LocalROGroup 127.0.0.1 in the .conf file. Zenoss claims snmpd is not functioning which I believe is from not having rocommunity public. Is it possible to have zenoss look at a differnet community name or even have two communities in my conf file?
[14-Jul-2009 09:30:34] <mrayzenoss> Morning all
[14-Jul-2009 09:32:23] <frozty_sa> hi
[14-Jul-2009 09:44:07] <etank> sup mrayzenoss
[14-Jul-2009 09:46:39] <mrayzenoss> not much, just pushing out some ZenPacks today
[14-Jul-2009 09:54:00] <potatodemon> How can I setup Zenoss to check for processes maybe once a minute ?
[14-Jul-2009 10:06:25] <ShakeSpear> Has someone seen any other number than "1" for "Cores"?
[14-Jul-2009 10:07:17] <BCS-Satori> Is it possible to add a hw product to the list for the server? My HP Proliant server is not listed
[14-Jul-2009 10:24:36] <jb> ShakeSpear: no, same here..
[14-Jul-2009 10:26:12] <ShakeSpear> jb: what makes me wonder: there doesn't even exist an attribute to store the number of cores. this just seems to be "a fake" - but from where does the "1" come??
[14-Jul-2009 10:27:27] <twm1010> @mrayzenoss: what kind of zenpacks?
[14-Jul-2009 10:28:32] <mrayzenoss> Dell, Open Solaris, Brocade and NetApp
[14-Jul-2009 10:28:40] <jb> dell?
[14-Jul-2009 10:28:51] <jb> like detailed dell? like the HP pack?
[14-Jul-2009 10:28:51] <mrayzenoss> new Dell Monitor from Egor Puzanov
[14-Jul-2009 10:28:53] <twm1010> Excellent.. I've held off installing the existing NetApp ZPs waiting for something more thorough
[14-Jul-2009 10:29:21] <mrayzenoss> the NetApp is a roll-up of the 3 existing ones
[14-Jul-2009 10:29:41] <twm1010> excellent, i'll remove the others and throw it on my test box
[14-Jul-2009 10:30:00] <twm1010> the others were very incosistent, bad templates
[14-Jul-2009 10:30:09] <twm1010> it would be nice if someone did qtree modeling in the filesystem area
[14-Jul-2009 10:30:26] <twm1010> but traps should suffice for alerting on that
[14-Jul-2009 10:32:00] <ShakeSpear> currently I am working on the modeler code for Fujitsu Siemens (FTS now) hardware. How can a make a "clone" of the HPMon Package - simply rename?
[14-Jul-2009 10:32:37] <mrayzenoss> ShakeSpear: you can get the code and make the changes you need, is that what you mean?
[14-Jul-2009 10:32:52] <ShakeSpear> yes
[14-Jul-2009 10:33:13] <mrayzenoss> trac-zenpacks/wiki/ZenPackscommunityHPMon
[14-Jul-2009 10:33:27] <mrayzenoss> There's a link to the source there, and you can check out and build most of the community ZenPacks
[14-Jul-2009 10:33:48] <mrayzenoss> just copy the whole source tree, make name changes where necessary
[14-Jul-2009 10:34:00] <mrayzenoss> then update the code as necessary
[14-Jul-2009 10:35:46] <ShakeSpear> ok. would it work if I make a copy of ZenPacks.community.HPMon-1.3-py2.4.egg to ZenPacks.community.FTSMon-1.3-py2.4.egg and do the renaming / reworking there?
[14-Jul-2009 10:36:09] <ShakeSpear> (in my current zenoss installation in /oopt/zenoss/ZenPacks
[14-Jul-2009 10:36:40] <mrayzenoss> probably cleaner to start with the source
[14-Jul-2009 10:36:55] <ShakeSpear> ok
[14-Jul-2009 10:37:15] <mrayzenoss> all you need is the ZenPacks.community.HPMon source and the GNUMakefile and build-egg.sh to build
[14-Jul-2009 10:38:14] <ShakeSpear> i see
[14-Jul-2009 10:38:49] <mrayzenoss> so say you're changing it from HPMon to FTS, you'll just rename the HPMon directories and update the setup.py
[14-Jul-2009 10:39:09] <mrayzenoss> then you can 'make ZenPacks.community.FTS' and it'll spit out your new egg
[14-Jul-2009 10:39:30] <straterra> is there anyway to get an updated wmi in zenoss?
[14-Jul-2009 10:39:41] <straterra> the current one in there only works with older autoconf
[14-Jul-2009 10:39:58] <mrayzenoss> straterra: we've got a ticket to tackle that
[14-Jul-2009 10:40:09] <ShakeSpear> mrayzenoss: ok  thx
[14-Jul-2009 10:40:31] <ShakeSpear> does anyone now what is wrong with this: om.setZenProperty=MultiArgs('cCabinetControllerFirmware', om._cabinetControllerFirmware)
[14-Jul-2009 10:41:08] <frozty_sa> p/win 24
[14-Jul-2009 10:41:11] <frozty_sa> .....
[14-Jul-2009 10:41:24] <ShakeSpear> It doesn't work: cCabinetControllerFirmwara as Custom Property will only be set, if it already exists on the modelled device - but if not, the property will not be created
[14-Jul-2009 10:50:31] <straterra> mrayzenoss: good deal..im having issues with the wmi and my kernel headers..do you know of that too?
[14-Jul-2009 10:52:30] <mrayzenoss> http://dev.zenoss.org/trac/ticket/4994
[14-Jul-2009 11:08:27] <kisielk> I wish Zenoss were smart enough to watchdog its own daemons and restart them
[14-Jul-2009 12:03:51] <sto6ma9ch> Hello all.
[14-Jul-2009 12:04:05] <sto6ma9ch> I have a recurring issue with zenprocess and zenperfsnmp
[14-Jul-2009 12:04:45] <sto6ma9ch> After a while, I keep receiving "localhost zenprocess heartbeat failure" and "localhost zenperfsnmp heartbeat failure" messages in the Event Console
[14-Jul-2009 12:05:33] <sto6ma9ch> I'll restart both services and clear the heartbeats, but the events will come back after a few minutes.
[14-Jul-2009 12:06:16] <sto6ma9ch> The graphs are not displaying colected data, either, but I think these issues are related.
[14-Jul-2009 12:36:55] <venturaville> mrayzenoss: have you seen issues with threshold values involving getRRDValue() calculations being 'sticky'?
[14-Jul-2009 12:43:54] <mrayzenoss> nope
[14-Jul-2009 12:44:31] <venturaville> any idea which file does the recalculation of the min/max values?
[14-Jul-2009 13:15:08] <dberger> can someone tell me where the info on mapping snmp is in the 2.4 documentation?
[14-Jul-2009 13:57:34] <ckrough> in zendmd, how can I return all of the subobjects of a container? a la "dmd.Devices.Network.Switch.Cisco._getOb()"
[14-Jul-2009 13:57:39] <ckrough> which doesnt do the trick
[14-Jul-2009 14:03:06] <ckrough> .getSubOrganizers()
[14-Jul-2009 14:03:08] <ckrough> that does it
[14-Jul-2009 14:46:47] <mpi> hello everyone, lil question here... I installed httpmonitor yesterday and tried all I can to make it work without success....
[14-Jul-2009 14:47:49] <mpi> it seem to be installed, and I associated its template to a server... the graphs entries show up in the perf tab, but they remain empty even after a complete day, and even if I shut down the website the threshold doesnt fire up
[14-Jul-2009 14:48:12] <etank> mpi: i have the same issue
[14-Jul-2009 14:49:28] <mpi> sorry about that but Im glad im not the only one
[14-Jul-2009 14:49:38] <mrayzenoss1> I refer people to: http://www.zenoss.com/community/docs/howtos/monitoring-websites-with-httpmonitor and http://www.zenoss.com/community/docs/zenoss-extended-monitoring/2.4.2/ch16.html
[14-Jul-2009 14:50:14] <mpi> ive already read all of that actually...
[14-Jul-2009 14:58:18] <mpi> however the page assumes Im only monitoring the server with httpmonitor, but im also using snmp...
[14-Jul-2009 14:59:27] <mpi> so I found instructions that seemed more up to date elsewhere and followed them (basically only requiring me to bind the httpmonitor template to my already-existing device
[14-Jul-2009 15:00:09] <mpi> it's effectively bound since the perf tab shows the 2 additional graphs that the template provides
[14-Jul-2009 15:00:17] <mpi> but the graphs are never filled with any data
[14-Jul-2009 15:01:23] <mpi> it feels like the template is bound but there's no actions taken at regular intervals to update the data source, create the graphs and trigger the thresholds
[14-Jul-2009 15:02:16] <venturaville> mrayzenoss: so am I reading MinMaxThreshold.py right and minval and maxval only get updated in getGraphElements?
[14-Jul-2009 15:02:34] <venturaville> and then only for RPNs?
[14-Jul-2009 15:02:51] <mrayzenoss1> venturaville: asking the wrong guy, don't have my head in that code
[14-Jul-2009 15:02:59] <mrayzenoss1> rocket might know
[14-Jul-2009 15:03:11] <venturaville> rocket: you awake?
[14-Jul-2009 15:03:44] <rocket> be there in a sec
[14-Jul-2009 15:06:19] <rocket> let me look again
[14-Jul-2009 15:13:47] <venturaville> >>> talesEval('python:'+ugctest.minval,d)
[14-Jul-2009 15:13:48] <venturaville> 46.747799999999998
[14-Jul-2009 15:14:04] <venturaville> if I do the talesEval like the code manually it seems to work
[14-Jul-2009 15:14:34] <venturaville> it just never gets reevaluated automatically
[14-Jul-2009 15:15:06] <rocket> venturaville: if I remember right I think there is some funky initialization stuff going on with thresholds
[14-Jul-2009 15:15:37] <rocket> venturaville: eg the objects keep getting reinitialized etc ..
[14-Jul-2009 15:16:09] <rocket> venturaville: but the minimum value is being passed into getgraphElements via self
[14-Jul-2009 15:16:16] <rocket> venturaville: eg self.minimum
[14-Jul-2009 15:17:16] <rocket> venturaville: I think what you are seeing is either a bug .. or intended but a confusing result .. I am not sure yet
[14-Jul-2009 15:17:48] <rocket> venturaville: since the threshold keeps getting reinitialized and minval and maxval keep getting set from the template
[14-Jul-2009 15:27:18] <dberger> what's the difference between $ZENHOME/common/share/mibs and $ZENHOME/common/share/snmp/mibs ?
[14-Jul-2009 15:27:31] <venturaville> doesn't look like the threshold get reinitialized very often
[14-Jul-2009 15:27:33] <venturaville> maybe at modeling time?
[14-Jul-2009 15:29:10] <mrayzenoss1> dberger: probably some installer issue that got overlooked: http://dev.zenoss.org/trac/ticket/4413 ?
[14-Jul-2009 15:30:55] <dberger> mrayzenoss1, not sure what that would have to do with it
[14-Jul-2009 15:31:54] <mrayzenoss1> dberger: sorry, multitasking too much.  That ticket was about a missing directory
[14-Jul-2009 15:31:55] <dberger> mrayzenoss1, I'm just observing that share/mibs has subdirectories with a lot of mibs, whereas share/snmp/mibs has only a few core mibs
[14-Jul-2009 15:32:10] <mrayzenoss1> dberger: yeah, wasn't paying close enough attention
[14-Jul-2009 15:34:12] <mrayzenoss1> dberger: hmmm.. all I have is $ZENHOME/share/mibs on the VM
[14-Jul-2009 15:34:18] <mrayzenoss1> which version?
[14-Jul-2009 15:34:47] <dberger> 2.4.1
[14-Jul-2009 15:36:22] <mrayzenoss1> which installer?
[14-Jul-2009 15:37:26] <dberger> zenoss-stack-2.4.1-linux-x64.bin
[14-Jul-2009 15:38:07] <mrayzenoss1> dberger: if you could open a ticket, our build engineer will look at resolving it
[14-Jul-2009 15:39:00] <dberger> is it a problem? i just assumed it was something i didn't understand
[14-Jul-2009 15:39:17] <mrayzenoss1> probably not a problem, we just want to keep things consistent
[14-Jul-2009 15:39:41] <dberger> how do I open a ticket?
[14-Jul-2009 15:39:57] <mrayzenoss1> http://dev.zenoss.com/trac/wiki/HowToAddTicket
[14-Jul-2009 15:40:05] <mrayzenoss1> he's poking around on a install now
[14-Jul-2009 15:41:52] <mrayzenoss1> do you have any MIB ZenPacks installed?
[14-Jul-2009 15:42:22] <mrayzenoss1> or, have you installed some yourself?  And if so, are they all in the same directory?
[14-Jul-2009 15:47:40] <dberger> i have installed core zenpacks
[14-Jul-2009 15:48:47] <dberger> i am working on zenpacks, but I house them outside $ZENHOME and so far haven't copied anything on install
[14-Jul-2009 17:01:34] <mrayzenoss> Dell Monitor ZenPack is live: http://www.zenoss.com/community/projects/zenpacks/dellmon
[14-Jul-2009 17:16:39] <rmatte> cool
[14-Jul-2009 18:21:58] <sacpinball> sweet
[14-Jul-2009 21:07:31] <hackeron> anyone? < http://forums.zenoss.com/viewtopic.php?p=36929#36929
[15-Jul-2009 00:00:46] [disconnected at Wed Jul 15 00:00:46 2009]
[15-Jul-2009 00:00:46] [connected at Wed Jul 15 00:00:46 2009]
[15-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[15-Jul-2009 06:05:23] <vinzpr> hello
[15-Jul-2009 07:49:56] <zeus> hi all!
[15-Jul-2009 09:19:17] <rocket> morning matt
[15-Jul-2009 09:20:23] <mrayzenoss> morning
[15-Jul-2009 10:12:17] <hackeron> anyone? < http://forums.zenoss.com/viewtopic.php?p=36929#36929
[15-Jul-2009 10:16:32] <mrayzenoss> I've sent the link to our build engineer
[15-Jul-2009 10:34:07] <jb> hmm.. im not seeing the "Available Plugins" in the zColectorPlugins page
[15-Jul-2009 10:34:14] <jb> just the applied ones
[15-Jul-2009 10:34:27] <jb> i seem to remember seeing something about this somewhere?
[15-Jul-2009 10:34:38] <jb> oh, nevermind..
[15-Jul-2009 10:37:39] <mrayzenoss> hackeron: he suggested reset the system as best as possible and follow the install guide exactly. If you still have issues please open a ticket and we'll see if there's more we can do. How much memory does your system have?
[15-Jul-2009 10:38:06] <mrayzenoss> hackeron: follow the uninstall on p. 28 of the Installation Guide before installing again
[15-Jul-2009 11:04:55] <schwartz> okay, here's a question outside of the norm and likely pushing what zenoss was meant to do
[15-Jul-2009 11:05:26] <schwartz> so I have a set of servers. they do email. I monitor them individually for the amount of email coming in/out via snmp.
[15-Jul-2009 11:06:01] <schwartz> I want to alert based on the trending functionality in rrdtool. while this works on a per host basis, the trending in rrd doesn't work cross data sources
[15-Jul-2009 11:07:07] <schwartz> so: is there a way to get zenoos to consolidate a single data source definition across all hosts and stuff that into a single RRD?
[15-Jul-2009 11:07:54] <mrayzenoss> schwartz: you could probably script something that pulls the individual RRD files and pushes them into another file
[15-Jul-2009 11:08:36] <mrayzenoss> this may help: http://www.zenoss.com/Members/cluther/licenseEventCheck.py/view
[15-Jul-2009 11:08:44] <schwartz> I had thought about that, but I'm not seeing an easy way to get the last value
[15-Jul-2009 11:08:45] <schwartz> lemme look
[15-Jul-2009 11:09:24] <dberger> can someone point me to the documentation of how to use snmp traps with zenoss?
[15-Jul-2009 11:09:50] <schwartz> ah, so run something external that populates the data instead of zenoss doing the population? interesting idea.
[15-Jul-2009 11:10:26] <schwartz> I'm assuming I can still define something somewhere within zenoss to alert based on that rrd file? or do the events only occur from within the data collection routine?
[15-Jul-2009 11:10:31] <mrayzenoss> schwartz: yeah
[15-Jul-2009 11:10:51] <mrayzenoss> that yeah was to the first question
[15-Jul-2009 11:11:58] <schwartz> ok. still hate to have zenoss collect per host stats, then something else to collect the same stats, add them, and stuff them into another rrd, but life goes on thoughts on the alerting bit of it?
[15-Jul-2009 11:12:27] <mrayzenoss> dberger: traps go to the Event Console
[15-Jul-2009 11:12:42] <mrayzenoss> schwartz: I'm not sure
[15-Jul-2009 11:13:20] <rocket> jb how is it going?
[15-Jul-2009 11:16:44] <dberger> that's what i thought...i guess i was just having trouble finding it
[15-Jul-2009 11:16:52] <dberger> thanks
[15-Jul-2009 11:43:13] <straterra> So..is there any easy way to use the stack installer minus..say..mysql?
[15-Jul-2009 11:43:23] <straterra> I have all the deps I need for zenoss, wmi just doesn't compile
[15-Jul-2009 11:43:46] <rmatte> you just install using the stack installer then repoint Zenoss to your current MySQL server and create the tables
[15-Jul-2009 11:44:03] <straterra> Won't the init scripts try to start the bundled mysql?
[15-Jul-2009 11:44:20] <rmatte> yeh, you'll have to change some scripts too
[15-Jul-2009 11:44:27] <rmatte> but it shouldn't be overly complicated
[15-Jul-2009 11:44:32] <straterra> Hmm..
[15-Jul-2009 11:44:49] <straterra> It's weird. I can do in to the wmi stuff and try to compile stuff manually..and it compiles. But the installer fails
[15-Jul-2009 11:53:21] <jb> rocket: good.. whats going on?
[15-Jul-2009 12:00:56] <rocket> not much ... get any time to test the aix snmp zenpack?
[15-Jul-2009 12:02:31] <jb> ugh no.. ive been SO busy.
[15-Jul-2009 12:02:46] <rocket> no problems
[15-Jul-2009 12:02:50] <jb> still working on it?
[15-Jul-2009 12:02:57] <rocket> not much lately
[15-Jul-2009 12:03:13] <jb> but you guys are using it in prod?
[15-Jul-2009 12:05:13] <rocket> no for us the project has been on hold .. so not implemented in prod ..
[15-Jul-2009 12:05:41] <jb> ah
[15-Jul-2009 12:25:27] * ckrough zenhub makes ckrough sad
[15-Jul-2009 12:46:41] <rmatte> Where is the Zenoss upgrade guide located?  I lost the link to it.
[15-Jul-2009 12:50:54] <gutseb> hello
[15-Jul-2009 12:51:19] <gutseb> I am having trouble understanding a few concepts under zenoss
[15-Jul-2009 12:51:46] <gutseb> anyone up for helping a newb?
[15-Jul-2009 12:53:05] <schwartz> I think a number of people will answer what they can, so I'd say fire away
[15-Jul-2009 12:53:12] <gutseb> cool
[15-Jul-2009 12:54:43] <gutseb> so i was trying to follow along with a walkthrough for enhanced modeling of dell hardware.
[15-Jul-2009 12:54:43] <gutseb> 1. So i created a new class.
[15-Jul-2009 12:54:57] <gutseb> added the collector plugins to that class
[15-Jul-2009 12:55:06] <gutseb> saved the class
[15-Jul-2009 12:55:18] <gutseb> but when i try to add a device(host) it fails
[15-Jul-2009 12:55:55] <gutseb> so i think i am trying to add this in the wrong location. or i am missing where to apply a plugin.
[15-Jul-2009 12:56:12] <gutseb> so the question is where am i supposed to apply the collector plugin
[15-Jul-2009 12:56:37] <gutseb> to the class or to the device direclty?
[15-Jul-2009 12:56:54] <ckrough> The Class is the proper place
[15-Jul-2009 12:56:58] <gutseb> it seems that i should be able to set up a class and add devices to it.
[15-Jul-2009 12:57:25] <ckrough> are you specifying that class when you add the device?
[15-Jul-2009 12:57:31] <gutseb> no
[15-Jul-2009 12:57:40] <gutseb> i am trying to add a new class to apply to devices
[15-Jul-2009 12:57:55] <ckrough> let me reread your description
[15-Jul-2009 12:58:32] <ckrough> "but when i try to add a device(host) it fails" == what happens when it "fails"
[15-Jul-2009 13:01:06] <gutseb> sorry that was not very discriptive.
[15-Jul-2009 13:01:14] <gutseb> and it seems to be working now
[15-Jul-2009 13:01:18] <gutseb> sorry.
[15-Jul-2009 13:01:56] <gutseb> the first time i tried to add it it gave me a error that i may have read incorrectly
[15-Jul-2009 13:02:13] <gutseb> but it has started to collect data now.
[15-Jul-2009 13:02:15] <gutseb> thanks
[15-Jul-2009 13:03:40] <mrayzenoss> rmatte: Did you find the Installation Guide? http://downloads.sourceforge.net/zenoss/Zenoss_Core_Installation_2.4.0.pdf
[15-Jul-2009 13:04:19] <gutseb> i was looking at http://downloads.sourceforge.net/zenoss/Zenoss_Extended_Monitoring_2.4.2.pdf chapter 3
[15-Jul-2009 13:09:10] <rmatte> yeh, I managed to dig it up, thanks Matt
[15-Jul-2009 13:15:14] <ckrough> rmatte: was that you working on the mib dependency doc?
[15-Jul-2009 13:15:55] <eidolon> hi folks - how can i temporarily disable a monitor? there's a pulldown in the event viewer for filtering - it says "State: [acknowledged] [new] or [suppressed]" - we have an alert going off right now that means nothing, but it's part of an upstream template. i'd liek to suppress that alert for this host possible?
[15-Jul-2009 13:17:38] <mrayzenoss> eidolon: sounds like you want a maintenance window
[15-Jul-2009 13:18:37] <eidolon> eh. i wish it were that simple. for instance, we have a monitor that checks for a certain action on a host. the monitor is failing. i don't want to delete the monitor, i just want to silence / stop checking for a while.
[15-Jul-2009 13:18:49] <eidolon> where 'for a while' means 'until i get back to it'
[15-Jul-2009 13:19:09] <eidolon> actually
[15-Jul-2009 13:19:10] <eidolon> wait
[15-Jul-2009 13:19:51] <eidolon> (better would be a count-up. don't throw an alert until this command fails more than 5 times in a row)
[15-Jul-2009 13:19:52] <rmatte> ok, I haven't upgraded in a while, beyond backing up the data, there isn't any kind of restore necessary after the upgrade (provided everything goes smoothly) I assume?
[15-Jul-2009 13:20:06] <rmatte> like, you just backup the data, run the installer, and you are theoretically done afterwards right?
[15-Jul-2009 13:20:08] <eidolon> it appears that sometime in the morning, this one check_http checker times out.
[15-Jul-2009 13:26:51] <rmatte> eidolon: you could always try increasing the Timeout value on the check_http datapoint
[15-Jul-2009 13:27:26] <rmatte> it defaults to 60 seconds
[15-Jul-2009 13:30:50] <rmatte> I see a step of copying data back in here: http://www.sysadminwiki.net/wiki/index.php?title=Upgrading_your_Zenoss_Stack
[15-Jul-2009 13:31:04] <rmatte> is that step necessary? I don't see any mention of it in the upgrade procedures for stack.
[15-Jul-2009 13:32:39] <rmatte> oh right, you need to perform a zenmigrate at the end I believe?
[15-Jul-2009 13:32:44] <rmatte> or does the installer handle that
[15-Jul-2009 13:39:03] <ckrough> CISCO-PRODUCTS-MIB doesnt show in the mib browser, but it is under /opt/zenoss/share/mibs, can I assume it's loaded since it exists under share/mibs?
[15-Jul-2009 13:39:22] <rmatte> no, you need to actually load it in with zenmib
[15-Jul-2009 13:39:30] <rmatte> zenmib --install=<mib file>
[15-Jul-2009 13:39:33] <ckrough> thanks
[15-Jul-2009 13:39:37] <rmatte> as zenoss user
[15-Jul-2009 13:39:39] <rmatte> no prob
[15-Jul-2009 13:39:52] <rmatte> sorry
[15-Jul-2009 13:39:56] <rmatte> zenmib run rather
[15-Jul-2009 13:40:05] <rmatte> zenmib run <file> I believe
[15-Jul-2009 13:40:10] <rmatte> (haven't done it in a while)
[15-Jul-2009 13:40:13] <ckrough> zenmib help
[15-Jul-2009 13:40:19] <rmatte> yeh, that works hehe
[15-Jul-2009 13:41:01] <rmatte> mrayzenoss: ping
[15-Jul-2009 13:41:20] <mrayzenoss> pong
[15-Jul-2009 13:41:23] <rmatte> hehe
[15-Jul-2009 13:41:49] <rmatte> ok, this guide is kind of all over the place, there are steps listed to perform after certain types of upgrades, but others don't have them listed...
[15-Jul-2009 13:41:59] <ckrough> someone in here recently posted a link to the Cisco MIB dependency tree... anyone remember
[15-Jul-2009 13:42:01] <rmatte> after I run the stack installer to perform the upgrade, are there any additional steps?
[15-Jul-2009 13:42:06] <rmatte> or will it take care of the rest?
[15-Jul-2009 13:42:37] <rmatte> because the guide has no mention of running any additional commands after a stack upgrade
[15-Jul-2009 13:43:06] <rmatte> ckrough: those Cisco Mibs are brutal to get loaded lol
[15-Jul-2009 13:43:15] <rmatte> hence why I made the Cisco Mibs Zenpack
[15-Jul-2009 13:43:57] * ckrough searches. thanks
[15-Jul-2009 13:44:12] <mrayzenoss> rmatte: the stack installers run the zenmigrate step automatically
[15-Jul-2009 13:44:37] <rmatte> If you do install the Zenpack, be warned that it takes like 5 or 6 hours to install, and you also need to follow the steps listed on the ZenPack page in ZenDMD or your UI performance will suck
[15-Jul-2009 13:44:57] <rmatte> mrayzenoss: perfect, exactly what I wanted to hear, thanks
[15-Jul-2009 13:45:11] <rmatte> I'm liking the stack installer more and more
[15-Jul-2009 13:45:15] <mrayzenoss> ckrough: note that rmatte added the removal step of the ZenPack to leave the MIBs yet delete the ZenPack
[15-Jul-2009 13:45:26] <mrayzenoss> ckrough: for performance
[15-Jul-2009 13:46:36] <rmatte> yeh, that workaround only takes about 6 minutes to do, and it's very safe, you can do it while the system is running without requiring a restart
[15-Jul-2009 13:46:39] <ckrough> do we know the cause of the performance issue? Im hesitant to play that game of roulette
[15-Jul-2009 13:46:56] <ckrough> this is a large install and it's particularly finicky
[15-Jul-2009 13:47:32] <rmatte> it's not really finicky, I think the performance issues are caused by Zenoss trying to reindex the whole ZenPack list almost every time you try to load certain pages such as zProperties
[15-Jul-2009 13:47:53] <rmatte> that change takes the ZenPack out of Zenoss' ZenPack inventory, but leaves the Mibs installed
[15-Jul-2009 13:48:04] <rmatte> once that's done, everything runs beautifully
[15-Jul-2009 13:48:14] <rmatte> I'm using that pack on many production systems, it's quite safe
[15-Jul-2009 13:48:56] <rmatte> You can check out what I had to go through to install those mibs with zenmib: http://demonic.cc/?p=20
[15-Jul-2009 13:49:06] <ckrough> rmatte: that is awesome. thanks!
[15-Jul-2009 13:49:18] <ckrough> rmatte: good thinking on documenting that process.
[15-Jul-2009 13:49:52] <rmatte> ckrough: yeh, I didn't want to lose it since it took me over 8 hours to figure out
[15-Jul-2009 13:50:18] <rmatte> I also have the tarball on that page with the Mibs that I modified included in it
[15-Jul-2009 13:50:36] <rmatte> I don't know how well it'll load in to the current version of Zenoss since that was originally done with 2.3.3
[15-Jul-2009 13:51:24] <ckrough> Ill try it out in the dev environment. Running 2.2.4 in production now
[15-Jul-2009 13:51:36] <rmatte> k
[15-Jul-2009 13:52:26] <rmatte> we're upgrading our 2.3.3 machines to 2.4.2 soon, I'm working on the upgrade outline right now, though we have the luxury of being able to clone VMs from our production servers in to our lab and attempt the upgrades on clones before performing it on the actual production systems
[15-Jul-2009 13:53:43] <rmatte> ok, so that guide is done, now I get to have fun planning how I'm going to reorganize our ZenPacks since right now we have everything bundled up in to one, and every time we do an upgrade it changes the snmp community string in /Devices, which is extremely undesired behaviour
[15-Jul-2009 13:53:45] <ckrough> yeah that is nice. I use vmware for the dev stuff, but the prod boxes are meatspace
[15-Jul-2009 13:54:52] <rmatte> ah, we have big blade boxes for production, 20GHz with 32GB of RAM (Expandable to 64GB), so not using VMs would be a waste
[15-Jul-2009 13:55:35] <ckrough> How do you prevent the disk IO from manipulating the hv?
[15-Jul-2009 13:58:10] <ckrough> err, not manipulating but consuming
[15-Jul-2009 14:03:33] <mrayzenoss> FYI, HTML version of the Installation Guide is now available
[15-Jul-2009 14:03:41] <mrayzenoss> http://www.zenoss.com/community/docs/
[15-Jul-2009 14:03:59] <mrayzenoss> so you can send links instead of referring to page numbers
[15-Jul-2009 14:04:28] <ckrough> so, if I have a device that is showing unknown/.1.3.6.1.4.1.9.1.716 as the make/model, then I load the missing mib, how do I get zenoss to correct that info. remodel doesnt seem to do it
[15-Jul-2009 14:13:05] <straterra> Ok..so the stack installer..I install it, then edit the init scripts to not start mysql, right?
[15-Jul-2009 14:15:38] <rmatte> hv?
[15-Jul-2009 14:16:32] <ckrough> rmatte: sorry, hypervisor. I havent tried a large zen install on vmware or xen since they are so sensitive to disk IO
[15-Jul-2009 14:16:50] <ckrough> rmatte: was wondering if you did anything special or just threw hardware at it
[15-Jul-2009 14:17:17] <rmatte> not sure what you mean by consume
[15-Jul-2009 14:17:34] <rmatte> like, in what sense?
[15-Jul-2009 14:17:56] <ckrough> The disk IO generated by RRD consuming 100% util on your storage
[15-Jul-2009 14:18:10] <ckrough> and in turn slowing down all of the other guests on the box
[15-Jul-2009 14:18:15] <rmatte> oh
[15-Jul-2009 14:18:24] <rmatte> all of the blades are hooked in to a SAN via fiber
[15-Jul-2009 14:18:37] <rmatte> so it's pretty quick regardless
[15-Jul-2009 14:19:07] <ckrough> We had gone SAN and ended up moving back because the IO was affecting other people on the SAN, had to go to local storage
[15-Jul-2009 14:19:27] <rmatte> ckrough: make/model isn't controlled by Mibs...
[15-Jul-2009 14:19:29] <ckrough> so I assuming that the virtualization situation would be the same or worse. you may have bigger/better san available than I did though
[15-Jul-2009 14:19:33] <rmatte> click on "Products" in the left hand meny
[15-Jul-2009 14:19:35] <rmatte> menu*
[15-Jul-2009 14:19:37] <charlieS> that's pretty common SANs are evil.
[15-Jul-2009 14:19:38] <rmatte> they are defined in there
[15-Jul-2009 14:19:58] <rmatte> our SAN is dedicated to our monitoring environment
[15-Jul-2009 14:20:02] <rmatte> so it's not such a big deal
[15-Jul-2009 14:20:08] <ckrough> rmatte: not controlled by, by recognized and degined by, right?
[15-Jul-2009 14:20:21] <ckrough> rmatte: ah, ours was shared with other internal groups.
[15-Jul-2009 14:20:45] <rmatte> ckrough: "not controlled by, by recognized and degined by, right?" hunh?
[15-Jul-2009 14:21:00] <rmatte> oh
[15-Jul-2009 14:21:05] <rmatte> yeh, that's what I meant
[15-Jul-2009 14:21:13] <rmatte> the Mibs are used for Traps, that's it that's all
[15-Jul-2009 14:21:25] <rmatte> the make/model if discovered based on what is in the Products section
[15-Jul-2009 14:21:48] <rmatte> also, there is a bug in Zenoss where it picks up the OID instead of the proper product info, not sure if it's been fixed in the most recent release
[15-Jul-2009 14:21:57] <rmatte> been trying to get an answer about that for a while, no one apparently knows
[15-Jul-2009 14:22:06] <ckrough> yeah, that's what Im working on now
[15-Jul-2009 14:22:17] <rmatte> working on fixing it?
[15-Jul-2009 14:22:28] <ckrough> working on realizing its broken
[15-Jul-2009 14:22:31] <rmatte> ah
[15-Jul-2009 14:22:33] <rmatte> gotcha
[15-Jul-2009 14:22:41] <rocket> rmatte: I submitted a patch
[15-Jul-2009 14:22:48] <rmatte> yeh, for one of our boxes we just got fed up and went through and did them all by hand
[15-Jul-2009 14:23:00] <rmatte> rocket: awesome, you are THE man
[15-Jul-2009 14:23:03] <ckrough> I was thinking that I was getting OIDs because the CISCO-PRODUCT-MIB wasnt properly loaded and zenoss didnt have a definition for the OID
[15-Jul-2009 14:23:05] <rocket> rmatte: not sure if it was in 2.4.2 or not
[15-Jul-2009 14:23:19] <ckrough> rocket: are you rocket in trac?
[15-Jul-2009 14:23:25] <rocket> rmatte: and I believe you have to still delete all the old junk out of unknown
[15-Jul-2009 14:23:28] <rocket> ckrough: yes
[15-Jul-2009 14:23:39] <rmatte> ckrough: ah, yeh, loading that Mib won't help with Zenoss product recognition
[15-Jul-2009 14:25:03] <ckrough> rmatte: not to drag this out into a ckrough training session, but if the products MIB doesnt provide zenoss with model names to map to the OID, where does zenoss get the model names? (1.3.6.1.4.1.9.1.716 = catalyst296024TT)
[15-Jul-2009 14:25:22] <rmatte> you submitted that patch like 4 months ago?
[15-Jul-2009 14:25:36] <rocket> rmatte: it was quite a while ago
[15-Jul-2009 14:25:40] <rmatte> ckrough: no problem...
[15-Jul-2009 14:26:10] <rmatte> ok, in Zenoss, click on "Products" on the left under the "Classes" section
[15-Jul-2009 14:26:30] <rmatte> let me know when that page has loaded
[15-Jul-2009 14:26:32] <ckrough> kk
[15-Jul-2009 14:26:42] <rmatte> once it has, click on "Cisco", you'll see it in the list
[15-Jul-2009 14:26:48] <ckrough> yup
[15-Jul-2009 14:27:03] <rmatte> ok, now, notice that there's a list of model numbers with corresponding OID values?
[15-Jul-2009 14:27:08] <rmatte> that's how it recognizes them
[15-Jul-2009 14:27:10] <ckrough> yup
[15-Jul-2009 14:27:32] <rmatte> If you wanted to add a new product in, you'd have to find out what OID corresponds to that model, and add it to the list
[15-Jul-2009 14:27:33] <ckrough> and where does that list come from?
[15-Jul-2009 14:27:40] <ckrough> manually? eew
[15-Jul-2009 14:27:52] <ckrough> I was assuming it could come from http://www.oidview.com/mibs/9/CISCO-PRODUCTS-MIB.html
[15-Jul-2009 14:27:58] <rmatte> yeh, although a lot of stuff is already included with Zenoss
[15-Jul-2009 14:28:22] <rmatte> there's a huge list of Cisco devices already included, as you can see
[15-Jul-2009 14:28:32] <rmatte> so if you find any missing, you can add them, but I doubt it'll happen often
[15-Jul-2009 14:29:19] <ckrough> I think Cisco changes the numbers sometimes. for example, the one that caught my eye today is a 24 port 2960, 1.3.6.1.4.1.9.1.716
[15-Jul-2009 14:29:44] <ckrough> we have plenty of 2960 24's out there, but suddenly some new ones are .716
[15-Jul-2009 14:29:51] <ckrough> lemme check what some of the others return
[15-Jul-2009 14:30:01] <rmatte> it might be a 2960-S or whatever
[15-Jul-2009 14:30:09] <rmatte> might have some special model designation
[15-Jul-2009 14:30:14] <ckrough> yeah
[15-Jul-2009 14:32:35] <ckrough> yup, all the same number, just a handful that didnt get recognized. I was on the wrong track... thanks for the help
[15-Jul-2009 14:33:46] <rmatte> no problem
[15-Jul-2009 14:34:07] <rmatte> what version of Zenoss are you using by the way?
[15-Jul-2009 14:35:01] <rmatte> because I have a 2.3.3 box that hardly ever correctly recognizes the OIDs that are actually in the list, yours is apparently working fairly well, just wondering if you're using 2.4.1 or 2.4.2 (which would indicate that the patch was applied)
[15-Jul-2009 14:37:45] <ckrough> 2.2.4
[15-Jul-2009 14:38:00] <rmatte> ah, that might be before the problem popped up in the first place
[15-Jul-2009 14:38:19] <rmatte> because I assume that it did initially work at some point and then some change caused it to break
[15-Jul-2009 14:38:26] <ckrough> thats one of the reasons we are still on 2.2.4. we like to let you guys do the beta testing
[15-Jul-2009 14:38:37] <rmatte> well, that's why we have a lab now
[15-Jul-2009 14:38:47] <rmatte> I didn't have one before, but we finally picked up a bunch of equipment
[15-Jul-2009 14:39:12] <rmatte> we actually have a SAN for our lab
[15-Jul-2009 14:39:23] <ckrough> Yeah, Im running 2.4 in a lab, but there is just no way to replicate the production env accurately in the lab. we can get close, but its never a 100% test
[15-Jul-2009 14:39:30] <rmatte> we can't fire it up though until we get adequate A/C installed in that room
[15-Jul-2009 14:39:39] <rmatte> we fired it up the one day and it was like the carribean in there
[15-Jul-2009 14:39:47] <ckrough> heh, I feel your pain
[15-Jul-2009 14:40:29] <rmatte> it'd be fine in the winter, we'd just open the window, super ghetto server room.
[15-Jul-2009 14:40:45] <rmatte> good thing we don't host our production stuff in there
[15-Jul-2009 14:45:03] <ckrough> your at an ISP, correct?
[15-Jul-2009 14:45:50] <rmatte> nah, we're a managed services provider
[15-Jul-2009 14:46:09] <rmatte> we monitor equipment for clients
[15-Jul-2009 14:46:20] <rmatte> (a NOC)
[15-Jul-2009 14:46:26] <ckrough> ah, so your zenoss installs are segmented by customer?
[15-Jul-2009 14:46:40] <rmatte> for the most part, we do have 2 shared boxes
[15-Jul-2009 14:47:28] <rmatte> the problem with running the shared boxes is that generally smaller clients are on those boxes and they almost always have the standard 192.168 ip range
[15-Jul-2009 14:47:41] <ckrough> ouch
[15-Jul-2009 14:47:45] <rmatte> so far we've managed to cram 5 clients on the first box and 3 on the second
[15-Jul-2009 14:47:49] <ckrough> they allow for overlap in 2.4 right?
[15-Jul-2009 14:47:58] <rmatte> not in my experience
[15-Jul-2009 14:48:18] <rmatte> causes lots of fun issues
[15-Jul-2009 14:48:42] <rmatte> plus you wouldn't want overlap anyways, it wouldn't make sense
[15-Jul-2009 14:49:02] <ckrough> ah, thought I read about that somewhere. in the Service Provider release
[15-Jul-2009 14:49:14] <rmatte> well, we're just using core
[15-Jul-2009 14:49:23] <ckrough> ... which I dont see on the site now. mrayzenoss was the Service Prodiver concept dropped?
[15-Jul-2009 14:49:54] <rmatte> I've had to make all sorts of fun little mods and scripts to get the results that we want
[15-Jul-2009 14:50:26] <rmatte> I want to eventually get my monitoring method for the default windows SNMP agent packaged up in to a proper ZenPack with a collector plugin
[15-Jul-2009 14:50:47] <rmatte> but first I need to study the developmeny of collector plugins, and also study Rocket's AIX pack, since he's done some cool stuff with the UI
[15-Jul-2009 14:51:10] <mrayzenoss> ckrough: it's all Enterprise now
[15-Jul-2009 14:51:42] <rmatte> then once I'm done with that, I might take a shot at a proper community WMI pack, but they are both ambitious projects that are going to take a long time
[15-Jul-2009 14:51:56] <ckrough> yeah, Im wrapping up puppetizing our environment so I can spend more time customizing zenoss. ...crosses fingers...
[15-Jul-2009 14:52:07] <rmatte> hehe
[15-Jul-2009 14:52:11] <rmatte> yeh, it's a big job
[15-Jul-2009 14:52:25] <rmatte> right now I'm designing an automated zenpack update system to push out updates across all our boxes
[15-Jul-2009 14:52:50] <rmatte> once we start getting up to like 20 servers it won't make sense to do it by hand anymore
[15-Jul-2009 14:52:50] <mrayzenoss> rmatte: sounds like you're writing Enterprise features
[15-Jul-2009 14:52:59] <rmatte> mrayzenoss: yeh, trying to
[15-Jul-2009 14:53:44] <rmatte> it's pretty complex stuff, especially since I'm just learning python as I go, but I'll get there eventually
[15-Jul-2009 14:54:12] <ckrough> rmatte: heh, Zope doesnt help in that situation
[15-Jul-2009 14:54:27] <rmatte> ckrough: yeh, you're telling me
[15-Jul-2009 14:54:27] <mrayzenoss> yeah, Zope gives Python a bad name
[15-Jul-2009 14:54:44] <ckrough> python I get. Zope makes my eyes bleed
[15-Jul-2009 14:54:52] <rmatte> Zope is nice in certain ways, but horrible in others
[15-Jul-2009 14:54:53] <mrayzenoss> you sound like our developers
[15-Jul-2009 14:55:53] <rmatte> hehe
[15-Jul-2009 14:56:35] <rmatte> I've come up with some neat stuff that I need to publish at some point
[15-Jul-2009 14:57:03] <rmatte> I made a script to have the hosts files on our monitoring servers update automatically from Zenoss every hour
[15-Jul-2009 14:57:29] <rmatte> I wrote it since everyone was complaining that hosts files always get outdated and are never properly maintained, so now it's just one less thing to worry about
[15-Jul-2009 14:57:35] <ckrough> rmatte: now your doing puppets work
[15-Jul-2009 14:57:57] <rmatte> yeh, I'm a big fan of automation
[15-Jul-2009 14:58:09] <ckrough> you have to be in this industry. that or go crazy.
[15-Jul-2009 14:58:25] <rmatte> also, the NOC guys were used to a command called "ovtopodump" for openview which would dump up/down status for each node
[15-Jul-2009 14:58:37] <rmatte> so I scripted up a "zendump" command that does the same thing
[15-Jul-2009 14:58:53] <mrayzenoss> ckrough: are you using the Zenoss module for Puppet?  How is it?
[15-Jul-2009 14:59:00] <ckrough> mrayzenoss: nope
[15-Jul-2009 14:59:17] * ckrough reads up on that module
[15-Jul-2009 14:59:51] <rmatte> instead of using something like puppet, I like writing my own stuff from scratch
[15-Jul-2009 14:59:53] <ckrough> ah, I had seen this before, but we are mostly networking (noc) so I dont add servers often enough to need it
[15-Jul-2009 14:59:57] <rmatte> cron++
[15-Jul-2009 15:00:34] <ckrough> rmatte: puppet has its rough spots, but I would definitely recommend checking it out if you manage lots boxes
[15-Jul-2009 15:00:58] <ckrough> mrayzenoss: that is slick though.
[15-Jul-2009 15:01:38] <rmatte> ckrough: I tend to just use automation via ssh
[15-Jul-2009 15:02:09] <rmatte> but I will take a look
[15-Jul-2009 15:02:39] <ckrough> I was mostly clusterssh and shell scripts, but it started to get difficult to manage
[15-Jul-2009 15:03:12] <rmatte> well, we're going to eventually do everything through a web interface
[15-Jul-2009 15:03:13] <straterra> mrayzenoss: BTW, svn compiles cleanly and loads perfectly (inregard to newer autoconf)
[15-Jul-2009 15:03:15] <ckrough> or I got lazy. kinda the same
[15-Jul-2009 15:03:16] <twm1010> @mrayzenoss: Did you release that netapp zenpack?
[15-Jul-2009 15:03:27] <rmatte> we'll check off what servers we want to do what to, click go, and it'll kick off the scripts
[15-Jul-2009 15:03:56] <rmatte> straterra: glad you finally got it working
[15-Jul-2009 15:03:58] <mrayzenoss> twm1010: it's in SVN, I'll get a page up for it now
[15-Jul-2009 15:04:19] <twm1010> gotcha
[15-Jul-2009 15:04:33] <straterra> rmatte: Me too
[15-Jul-2009 15:04:40] <straterra> I haven't configured it yet..just installed it
[15-Jul-2009 15:04:43] <rmatte> mrayzenoss: that brocade zenpack needs to be more specific in what type of Brocade switches it is for.
[15-Jul-2009 15:04:54] <rmatte> mrayzenoss: it doesn't work at all for the Brocade fiber switches
[15-Jul-2009 15:05:32] <rmatte> mrayzenoss: the old brocade pack works for fiber switches, but only monitors the first port on the switch
[15-Jul-2009 15:05:42] <twm1010> rmatte: i was about to test it against some hp branded brocades
[15-Jul-2009 15:06:04] <rmatte> yeh, we have HP branded brocade fiber switches, and the new pack doesn't work at all for us
[15-Jul-2009 15:06:12] <rmatte> the old pack works, but only on the first port hehe
[15-Jul-2009 15:06:24] <twm1010> mine are probably 4 years old though, early 4GB models
[15-Jul-2009 15:06:35] <rmatte> shouldn't matter
[15-Jul-2009 15:06:41] <twm1010> using 6.2 firmware
[15-Jul-2009 15:09:16] <mrayzenoss> sent the author an email asking for more specific models
[15-Jul-2009 15:09:33] <rmatte> cool
[15-Jul-2009 15:09:38] <mrayzenoss> I'll go work on the Open Solaris and NetApp ZenPacks now
[15-Jul-2009 15:10:40] <rmatte> grrr, one of our Zenoss boxes is spitting out duplicate packets
[15-Jul-2009 15:12:08] <straterra> Hmm..is there an easy way to change the port zenoss runs on?
[15-Jul-2009 15:12:44] <ckrough> listens on? apache frontend I guess http://www.zenoss.com/community/docs/howtos/setup-zenoss-with-apache/
[15-Jul-2009 15:13:08] <ckrough> doesnt free up the zope port, but makes it more user friendly
[15-Jul-2009 15:13:27] <rmatte> straterra: yeh, let me find where the config option is...
[15-Jul-2009 15:13:40] <straterra> I have another peice of software using 8080, thats all
[15-Jul-2009 15:13:46] <ckrough> ah
[15-Jul-2009 15:13:59] <rmatte> straterra: yeh, so do we, so we have Zenoss running on 8180
[15-Jul-2009 15:14:59] <straterra> I found it
[15-Jul-2009 15:15:02] <straterra> zope.conf
[15-Jul-2009 15:15:12] <rmatte> yeh
[15-Jul-2009 15:15:33] <rmatte> he line that says "address 8080"
[15-Jul-2009 15:15:37] <rmatte> the line*
[15-Jul-2009 15:15:40] <mrayzenoss> hmmm.. Chet converted the Celerra ZenPack to an Egg, yay
[15-Jul-2009 15:17:40] <rmatte> cool
[15-Jul-2009 15:20:15] <straterra> Hmm..does zenoss play ok from behind a proxy?
[15-Jul-2009 15:20:55] <straterra> It keeps trying to go to port 9999 now..hmmm
[15-Jul-2009 15:20:56] <schwartz> I had problems getting it to behave behind apache in 2.2 (I think) I haven't tried it yet under 2.4.2
[15-Jul-2009 15:22:34] <schwartz> If I remember correctly, it would have links/redirects/etc going to what it thinks its hostname is as well as what port it's running on, made it very difficult to proxy through apache.
[15-Jul-2009 15:22:46] <straterra> Yeah..that's what I'm finding
[15-Jul-2009 15:22:56] <schwartz> there were some zope work arounds, which almost got it to work
[15-Jul-2009 15:24:17] <ckrough> Im running 2.2.4 behind apache https, no issues
[15-Jul-2009 15:25:39] <daMaestro> does anyone know if there is a way to prevent a device modeling from picking up new processes?
[15-Jul-2009 15:25:47] <schwartz> lock the host
[15-Jul-2009 15:25:50] <daMaestro> aka.. we have a device that has a lot of start/stop/restart java stuff
[15-Jul-2009 15:26:03] <daMaestro> we'd like all other changes to be found
[15-Jul-2009 15:26:12] <daMaestro> just not keep adding OSProcess stuff
[15-Jul-2009 15:26:21] <daMaestro> as this particular machine starts and stops tons of java stuff
[15-Jul-2009 15:26:41] <schwartz> hmm. you may be able to just lock the os processes section
[15-Jul-2009 15:26:53] <daMaestro> you can only lock items
[15-Jul-2009 15:27:03] <daMaestro> each process looks different
[15-Jul-2009 15:27:34] <schwartz> hmm, on 2.4.2 I do have the option of locking os processes on it's pulldown tab
[15-Jul-2009 15:29:07] <daMaestro> that does not apply to "all OSProcesses"
[15-Jul-2009 15:29:13] <daMaestro> only the ones you select
[15-Jul-2009 15:29:18] <daMaestro> aka.. they already have to be there
[15-Jul-2009 15:29:38] <schwartz> dunno then
[15-Jul-2009 15:29:40] <daMaestro> i've just locked the device for now, thanks for that... but that makes it so we will manually have to go remodel things
[15-Jul-2009 15:29:48] <schwartz> *nod*
[15-Jul-2009 15:29:52] <daMaestro> all the alerts are really annoying
[15-Jul-2009 15:34:41] <ckrough> hmm pulling diskIOWrites 1.3.6.1.4.1.2021.13.15.1.1.6 causes 'diskio.c: don't know how to handle 11 request'
[15-Jul-2009 15:38:08] <ckrough> ah, apparently that is normal according to Red Hat
[15-Jul-2009 15:39:25] <straterra> Is there any way to make sure zenoss is actually...working? I added localhost in to it and I don't get any graphs..or any information at all
[15-Jul-2009 15:40:35] <ckrough> you can check to make sure processes and running. You can also browse to the rrd folder for that decice and check the 'rrdtool last' for a couple files to see if they are being updated
[15-Jul-2009 15:40:36] <straterra> Oh..its still in the 'jobs'
[15-Jul-2009 15:40:39] <straterra> For..whatever reason
[15-Jul-2009 15:40:54] <ckrough> it takes rrd 3 cycles to start graphing, so polling_time*3
[15-Jul-2009 15:41:24] <straterra> Yeah..still in the middle of 'DeviceCreationJob'. Been running for like 10 minutes
[15-Jul-2009 15:41:26] <ckrough> decice = device. the fingers are approaching end of day...
[15-Jul-2009 15:43:46] <straterra> and zenhub isn't running..for whatever reason
[15-Jul-2009 15:45:23] <rmatte> straterra: zenhub needs to be running
[15-Jul-2009 15:45:28] <rmatte> hence the hung modeling
[15-Jul-2009 15:45:37] <straterra> Yeah..I keep starting it, and it keeps stopping
[15-Jul-2009 15:45:53] <rmatte> go to a console, become the zenoss user, and type zenhub run
[15-Jul-2009 15:45:55] <rmatte> see what it says
[15-Jul-2009 15:46:04] <straterra> will do
[15-Jul-2009 15:46:22] <rmatte> you can also check out $ZENHOME/log/zenhub.log
[15-Jul-2009 15:47:06] <straterra> It's trying to listen on port 8081
[15-Jul-2009 15:47:07] <ckrough> may want to run it with --v10 if there isnt anything good in there
[15-Jul-2009 15:47:08] <straterra> which is in use
[15-Jul-2009 15:47:27] <rmatte> then that would be your problem
[15-Jul-2009 15:47:29] <rmatte> use 8180
[15-Jul-2009 15:47:46] <straterra> It can be an arbitrary value, right?
[15-Jul-2009 15:48:04] <rmatte> yes
[15-Jul-2009 15:48:19] <rmatte> I just tend to use 8180 since it doesn't interfere with Zenoss and it's easy to remember
[15-Jul-2009 15:48:22] <rmatte> but it's totally up to you
[15-Jul-2009 15:48:35] <straterra> I gotta hunt down how to change it now
[15-Jul-2009 15:49:10] <rmatte> well wait, is 8081 being used by the web interface?
[15-Jul-2009 15:49:26] <straterra> No, different software
[15-Jul-2009 15:49:32] <rmatte> ah
[15-Jul-2009 15:49:33] <rmatte> garbage
[15-Jul-2009 15:49:34] <rmatte> lol
[15-Jul-2009 15:49:46] <straterra> aparantly, I can use -x and the port
[15-Jul-2009 15:50:15] <rmatte> hmmm, you'll have to modify all the startup scripts to do that though
[15-Jul-2009 15:50:28] <straterra> http://www.zenoss.com/community/docs/zenoss-guide/2.3.3/ch23s06.html
[15-Jul-2009 15:50:30] <straterra> That still true?
[15-Jul-2009 15:50:59] <rmatte> yeh, should be
[15-Jul-2009 15:51:41] <rmatte> not sure if you need -x or -pbport
[15-Jul-2009 15:52:23] <straterra> nah..pb port is 8789 by default
[15-Jul-2009 15:52:28] <rmatte> ah
[15-Jul-2009 15:52:40] <straterra> the zenhub.conf thing isn't quite working though
[15-Jul-2009 15:52:53] <straterra> I have xport           8180
[15-Jul-2009 15:53:07] <straterra> and when I try to start... INFO: Commenting out unknown option 'xport' found on line 2 in config file
[15-Jul-2009 15:53:16] <schwartz> Alright. So working on populating an RRD file from an external script. Trying to figure out how to get zenoss to 'see' it, as a datasource so it can at least graph it. Thoughts on where to start on that? There doesn't seem to be an option in the 'add a new datasource dialog', it wants SNMP or some other (dns, mysql, etc) package to provide a DS.
[15-Jul-2009 15:54:17] <rmatte> you could probably modify zenhub.conf somehow to specify another port
[15-Jul-2009 15:54:22] <rmatte> which would be the preferable solution
[15-Jul-2009 15:54:43] <rmatte> just use -x
[15-Jul-2009 15:54:46] <rmatte> not -xport
[15-Jul-2009 15:54:58] <straterra> same message
[15-Jul-2009 15:55:04] <rmatte> oh wait
[15-Jul-2009 15:55:06] <rmatte> --xport
[15-Jul-2009 15:55:10] <rmatte> need 2 -'s
[15-Jul-2009 15:55:20] <rmatte> and you need an equal sign
[15-Jul-2009 15:55:25] <rmatte> --xport=8180
[15-Jul-2009 15:55:48] <straterra> If I do that, I get a warning about 'missing value on line 2'
[15-Jul-2009 15:56:18] <rmatte> I have something for you to try though
[15-Jul-2009 15:56:24] <straterra> ok
[15-Jul-2009 15:56:32] <rmatte> In the Zenoss UI, go to Settings -> Daemons
[15-Jul-2009 15:56:45] ckrough is now known as ckrough_away
[15-Jul-2009 15:56:47] <rmatte> click on the "edit config" link for zenhub
[15-Jul-2009 15:56:55] <rmatte> and add a line that says xport 8180
[15-Jul-2009 15:56:59] <rmatte> then click save
[15-Jul-2009 15:57:03] <rmatte> then try to start it
[15-Jul-2009 15:57:10] <straterra> heh..
[15-Jul-2009 15:57:11] <straterra> nice
[15-Jul-2009 15:57:18] <straterra> zenhub: command not found
[15-Jul-2009 15:57:24] <straterra> so sayeth the ui
[15-Jul-2009 15:57:40] <rmatte> ummmm, is Zenoss running as root right now?
[15-Jul-2009 15:57:52] <straterra> Nope
[15-Jul-2009 15:58:01] <straterra> running as the user zenoss
[15-Jul-2009 15:58:04] <rmatte> do echo "$ZENHOME"
[15-Jul-2009 15:58:09] <rmatte> from a commandline as the zenoss user
[15-Jul-2009 15:58:13] <rmatte> what shows up?
[15-Jul-2009 15:58:18] <straterra> /usr/local/zenoss
[15-Jul-2009 15:58:25] <rmatte> is that where you have it installed?
[15-Jul-2009 15:58:29] <straterra> yes
[15-Jul-2009 15:58:42] <straterra> zenhub is in bin/ too
[15-Jul-2009 15:58:43] <rmatte> so the directory /usr/local/zenoss/Products exists?
[15-Jul-2009 15:58:51] <straterra> correct
[15-Jul-2009 15:58:56] <rmatte> ok
[15-Jul-2009 15:59:09] <straterra> Can I just edit zenhub.py ?
[15-Jul-2009 15:59:15] <rmatte> nah
[15-Jul-2009 15:59:20] <straterra> that has the port in it o.O
[15-Jul-2009 15:59:27] <rmatte> well, you could
[15-Jul-2009 15:59:40] <rmatte> I guess that works too
[15-Jul-2009 15:59:51] <twm1010> hrmm, any way to do a discovery on an IP range, instead of a network ?
[15-Jul-2009 15:59:55] <rmatte> just if you ever do an upgrade you'll have to make that change again
[15-Jul-2009 16:00:07] <straterra> I imagine I have to change it in other places too
[15-Jul-2009 16:00:10] <rmatte> if you can find the zenhub.conf file the in the zenoss directory, you can add the line to it
[15-Jul-2009 16:00:13] <rmatte> xport 8180
[15-Jul-2009 16:00:22] <straterra> I tried that
[15-Jul-2009 16:00:33] <rmatte> try just port instead of xport?
[15-Jul-2009 16:00:39] <mrayzenoss> twm1010: yeah, the new Add Device Wizard supports that
[15-Jul-2009 16:00:50] <mrayzenoss> twm1010: go to Add Devices, then "Easy Add..."
[15-Jul-2009 16:01:03] <straterra> woah
[15-Jul-2009 16:01:07] <straterra> didnt compain on port
[15-Jul-2009 16:01:13] <rmatte> good
[15-Jul-2009 16:01:19] <rmatte> that must be the right option then
[15-Jul-2009 16:01:37] <straterra> I would have thought something else needs to be modified to tell of the new port
[15-Jul-2009 16:01:46] <straterra> You know..if the other daemons try to communicate with it
[15-Jul-2009 16:01:49] <rmatte> I've never changed the actual zenhub port, just the port on the UI, so I was just guessing
[15-Jul-2009 16:01:54] <twm1010> easy add.... look at that... thank you Matt
[15-Jul-2009 16:02:02] <rmatte> the other daemons probably find out what port to communicate with it from that config
[15-Jul-2009 16:02:20] <rmatte> but maybe there is something that needs to be changed
[15-Jul-2009 16:02:37] <rmatte> only way to tell is to model a device
[15-Jul-2009 16:02:46] <rmatte> if it says that it can't connect to zenhub then you know you have an issue
[15-Jul-2009 16:02:47] <straterra> So, try to add a device again?
[15-Jul-2009 16:02:51] <rmatte> if it connects then you're good
[15-Jul-2009 16:02:53] <rmatte> yup
[15-Jul-2009 16:03:16] <rmatte> sometimes it'll say it can't connect the finally connects
[15-Jul-2009 16:03:19] <rmatte> so give it a few minutes
[15-Jul-2009 16:03:30] <rmatte> then* finally connects, rather
[15-Jul-2009 16:04:34] <straterra> it cleared through the jobs
[15-Jul-2009 16:05:15] <rmatte> cool
[15-Jul-2009 16:05:26] <straterra> Of course..the snmp community didn't stick
[15-Jul-2009 16:05:29] <rmatte> by the way, you do know that you need to have snmp enabled or something for it to really model anything I assume?
[15-Jul-2009 16:05:31] <straterra> it went back to public..which isn't right
[15-Jul-2009 16:05:35] <rmatte> and you need the community string configured in Zenoss
[15-Jul-2009 16:05:37] <straterra> Yeah, snmp is enabled.
[15-Jul-2009 16:05:42] <rmatte> k
[15-Jul-2009 16:05:43] <straterra> I'm migrating from cacti/nagios to this
[15-Jul-2009 16:05:54] <rmatte> click on Devices on the left
[15-Jul-2009 16:06:08] <rmatte> then click on the zProperties tab
[15-Jul-2009 16:06:38] <rmatte> change the zSnmpCommunity value to the key that you want to use
[15-Jul-2009 16:06:51] <rmatte> (assuming you're using a global key, if not, then you'll have to set it on a per device basis)
[15-Jul-2009 16:07:24] <straterra> So..zenoss looks like with its graphing it can replace nagios and cacti
[15-Jul-2009 16:07:38] <rmatte> it can completely replace either, correct
[15-Jul-2009 16:07:50] <rmatte> it uses RRDTools just like the rest
[15-Jul-2009 16:08:45] <rmatte> it even supports nagios plugins, plus you can write custom scripts to collect data and as long as they output in nagios format you can use them in Zenoss
[15-Jul-2009 16:08:50] <rmatte> I do it all the time
[15-Jul-2009 16:09:17] <rmatte> nagios output format is like OK|value1=0 value2=4 value3=8
[15-Jul-2009 16:09:20] <straterra> I have a TON of nagios plugins
[15-Jul-2009 16:09:30] <straterra> Hand written ones, that is
[15-Jul-2009 16:09:37] <rmatte> yeh, you can use those
[15-Jul-2009 16:09:41] <rmatte> if you need to
[15-Jul-2009 16:09:52] <rmatte> but Zenoss' built in functionality may replace a lot of those
[15-Jul-2009 16:09:52] <straterra> I likely will
[15-Jul-2009 16:10:07] <straterra> So now..I look in perf to make sure snmp is working and all?
[15-Jul-2009 16:10:43] <rmatte> well, if SNMP is working it should have populated the OS tab with interfaces, disks, and such
[15-Jul-2009 16:10:48] <straterra> it didnt
[15-Jul-2009 16:10:54] <straterra> i didnt do anything but change the community, though
[15-Jul-2009 16:11:06] <rmatte> oh, right
[15-Jul-2009 16:11:10] <rmatte> you need to force it to remodel
[15-Jul-2009 16:11:38] <mrayzenoss> OpenSolaris ZenPack is now available: http://www.zenoss.com/community/projects/zenpacks/opensolaris
[15-Jul-2009 16:11:41] <rmatte> in the top left on the device window there is a downward arrow button
[15-Jul-2009 16:11:49] <rmatte> click on that, go to Manage -> Model Device
[15-Jul-2009 16:11:57] <straterra> Doing that now
[15-Jul-2009 16:12:00] <rmatte> k
[15-Jul-2009 16:12:08] <straterra> It's time for me to go home..thanks SO much for the help
[15-Jul-2009 16:12:20] <rmatte> no problem, cheers
[15-Jul-2009 16:12:52] <straterra> Error about zenhub..but I'll tackle it at home
[15-Jul-2009 16:13:06] <rmatte> k cool
[15-Jul-2009 16:24:21] <twm1010> @mrayzenoss: Has anyone thought about adding a field for a server's management NIC?
[15-Jul-2009 16:25:11] <twm1010> I suppose it is technically a separate host, but I'd like to correlate the management NIC with the server its tied to
[15-Jul-2009 16:26:07] <mrayzenoss> yeah, I follow what you're saying, I don't think I've seen that requested. Feel free to open an enhancement ticket, that'll get it into the feature queue
[15-Jul-2009 16:29:10] <schwartz> any takers on the externally populated rrd file thing?
[15-Jul-2009 16:34:01] <rmatte> schwartz: yeh...
[15-Jul-2009 16:34:08] <schwartz> ?
[15-Jul-2009 16:34:14] <rmatte> first off, what does the script do?
[15-Jul-2009 16:35:07] <schwartz> so far the script doesn't do anything yet. I'm needing to consolidate a single metric from a bunch of hosts into one rrd so I can effectively use rrd's predict functionality
[15-Jul-2009 16:35:35] <schwartz> so since there's no way to get zenoss to add those into a single rrd at collection time, I'm going to have to do it
[15-Jul-2009 16:36:51] <schwartz> actually maybe I'll make a separate collector for this and make it a zenpack.
[15-Jul-2009 16:38:43] <rmatte> you need multiple values in a single rrd file?
[15-Jul-2009 16:39:16] <schwartz> one data source in an rrd file, that's the value of all the hosts metrics added together
[15-Jul-2009 16:39:39] <rmatte> ok, that's fairly easy to do
[15-Jul-2009 16:39:41] <schwartz> it's email stats, and I've got about a dozen servers that do email, the prediction doesn't work so well when split up
[15-Jul-2009 16:39:46] <schwartz> glad to hear it
[15-Jul-2009 16:40:03] <rmatte> make the script to collect the data and add it up, then have the script output it in nagios format, so OK|whatever=<value>
[15-Jul-2009 16:40:25] <rmatte> then create a command based data source in Zenoss, and create a data point under it called whatever
[15-Jul-2009 16:40:25] <schwartz> *nod*
[15-Jul-2009 16:40:32] <rmatte> or whatever you want your variable to be called
[15-Jul-2009 16:40:39] <rmatte> then create a graph with that datapoint
[15-Jul-2009 16:40:40] <rmatte> and you're done
[15-Jul-2009 16:41:05] <rmatte> the script should output "Unknown" if it's not able to properly collect the data
[15-Jul-2009 16:41:10] <rmatte> but it's not an absolute requirement
[15-Jul-2009 16:41:29] <schwartz> normal nagios behavior then
[15-Jul-2009 16:41:36] <rmatte> correct
[15-Jul-2009 16:41:51] <schwartz> I was hoping to avoid having to query the same data twice, as I do want it stored per host as well (with no prediction)
[15-Jul-2009 16:42:03] <rmatte> as long as the script is able to collect the data, process it, and output the final value, then Zenoss will be able to pick up that value and graph it as it executes the command each polling cycle
[15-Jul-2009 16:42:26] <mrayzenoss> that's actually better than what I was suggesting
[15-Jul-2009 16:42:37] <schwartz> which half led me to say okay, I'll grab the values externally, add it up and have the script itself update all of the rrd files instead of zenoss, which flowed into how to get zenoss to use an rrdfile it isn't updating
[15-Jul-2009 16:42:38] <rmatte> well, you could have the script output each value as a datapoint
[15-Jul-2009 16:42:54] <rmatte> so OK|server1=<value> server2=<value> total=<value>
[15-Jul-2009 16:43:02] <schwartz> huh.
[15-Jul-2009 16:43:04] <rmatte> then have it graph each individual server as well as the total
[15-Jul-2009 16:43:07] <mrayzenoss> schwartz: yeah, I'm looking for Chet to followup on his link I sent you
[15-Jul-2009 16:43:24] <schwartz> thanks!
[15-Jul-2009 16:43:33] <schwartz> I'll check out the multiple value thing as well
[15-Jul-2009 16:44:05] <rmatte> that multiple value thing is just basically what I'm telling you to do, just less taylored towards what you want to do
[15-Jul-2009 16:44:37] <schwartz> *nod*
[15-Jul-2009 16:44:51] <rmatte> you can create one command based datasource that runs your script, then create multiple datapoints beneath it, such as server1, server2, etc...
[15-Jul-2009 16:45:21] <rmatte> then as long as the script outputs as OK|server1=24453 server2=2789 or whatever, Zenoss will create RRD files and put the data in to them
[15-Jul-2009 16:45:27] <rmatte> then you can graph based on any of those values
[15-Jul-2009 16:46:36] <rmatte> you could have one graph for the total, and a seperate graph for the servers, or a separate graph for each server, and the total, it's entirely up to you
[15-Jul-2009 16:47:46] <schwartz> yeah, I kind of like that. first attempt will be the nagios-like command one, but I would like to rewrite it in python, but will leave that for another day
[15-Jul-2009 16:47:55] <rmatte> the datapoint names need to correspond with the output of the script for it to work
[15-Jul-2009 16:48:24] <schwartz> good to know, thanks
[15-Jul-2009 16:48:29] <rmatte> you could do the same thing with a python script provided it outputs the same way
[15-Jul-2009 16:48:32] <rmatte> np
[15-Jul-2009 16:48:57] <rmatte> well, I must be going, later folks
[15-Jul-2009 16:49:08] <schwartz> yeah, trying to learn the internals of zenoss. one thing I don't like about nagios is that it forks off .. who knows how many process to get your data
[15-Jul-2009 16:49:12] <schwartz> night!
[15-Jul-2009 16:50:49] <rmatte> yeh, I'm working on figuring out how to write collector plugins, custom reports, and how to add things to the UI, going to take a while
[15-Jul-2009 16:51:00] * rmatte departs
[15-Jul-2009 16:51:55] <straterra> mrayzenoss: still around?
[15-Jul-2009 16:52:25] <mrayzenoss> yeah
[15-Jul-2009 16:52:47] <mrayzenoss> rmatte: schwartz: one more thing for Nagios plugins: http://dev.zenoss.org/trac/ticket/5294
[15-Jul-2009 16:52:47] <straterra> I changed the port for zenhub..but the other daemons are still trying to communicate with it on the old port
[15-Jul-2009 16:52:56] <mrayzenoss> did you restart everything?
[15-Jul-2009 16:52:59] <straterra> Is there anything else I have to change for the daemons to know the port has been moved?
[15-Jul-2009 16:53:12] <schwartz> pulling it up
[15-Jul-2009 16:53:16] <straterra> Yes, but I can restart it again..
[15-Jul-2009 16:56:09] <straterra> Didn't make a difference
[15-Jul-2009 16:56:17] <straterra> WARNING ZEO.zrpc: (30773) CW: error connecting to ('localhost', 8180): ECONNREFUSED
[15-Jul-2009 16:56:46] <mrayzenoss> straterra: what is your port-base set to in your zope.conf?
[15-Jul-2009 16:57:28] <straterra> It's not set to anything
[15-Jul-2009 16:57:58] <straterra> Oohh..I can offset all of the ports
[15-Jul-2009 16:58:00] <straterra> That's cool
[15-Jul-2009 16:58:16] <mrayzenoss> sorry, wasn't paying close attention earlier
[15-Jul-2009 16:58:24] <straterra> It's fine
[15-Jul-2009 16:58:25] <mrayzenoss> port-base 1000 will move you to 9080
[15-Jul-2009 16:58:28] <straterra> right
[15-Jul-2009 16:58:36] <straterra> I manually moved the http to 9999, but 9080 will work too
[15-Jul-2009 16:58:42] <straterra> 8080 and 8081 are both used though
[15-Jul-2009 17:01:38] <straterra> I removed the zenhub port change and the base port change, and just set port-base to 1000
[15-Jul-2009 17:01:42] <straterra> and now zenhub isn't starting at all
[15-Jul-2009 17:01:46] <straterra> But it throws no errors, too
[15-Jul-2009 17:02:20] <straterra> ah..it appears the port-base option doesn't change things for zenhub and such
[15-Jul-2009 17:04:06] <straterra> What other daemons communicate with zenhub?
[15-Jul-2009 17:04:21] <mrayzenoss> everything goes through zenhub
[15-Jul-2009 17:04:43] <straterra> So..I have to change the config for every daemon to use the new port?
[15-Jul-2009 17:04:56] <mrayzenoss> what exactly are you trying to do?
[15-Jul-2009 17:05:08] <mrayzenoss> the zope.cfg is good enough for the UI
[15-Jul-2009 17:06:14] <straterra> I have another daemon on the machine (Not Zenoss related) that's already bound to port 8081 and it needs to stay that way
[15-Jul-2009 17:06:20] <straterra> But zenhub is also trying to use port 8081
[15-Jul-2009 17:06:51] <straterra> I've sucessfully changed the port zenhub listens on to 9081. However, all of the other Zenoss daemons are still trying to connect to it via port 8081
[15-Jul-2009 17:10:38] <straterra> Looks like the only thing that uses the port is zensendevent
[15-Jul-2009 17:11:07] <mrayzenoss> hmm... since it sends events to the zenhub
[15-Jul-2009 17:11:15] <mrayzenoss> I'll go bug some devs
[15-Jul-2009 17:11:32] <straterra> I changed the port in zensendevent, but it still isn't working..I'll keep hacking at it
[15-Jul-2009 17:19:15] <mrayzenoss> ze
[15-Jul-2009 17:20:27] <mrayzenoss> zenhub conf
[15-Jul-2009 17:20:30] <mrayzenoss> oops,
[15-Jul-2009 17:20:44] <mrayzenoss> zenhub genconf will dump the options
[15-Jul-2009 17:21:19] <straterra> I already changed the config in zenhub
[15-Jul-2009 17:21:20] <mrayzenoss> and zeneventlog genconf
[15-Jul-2009 17:21:32] <mrayzenoss> genconf shows you the options for the daemons
[15-Jul-2009 17:22:11] <mrayzenoss> hubport looks like the options
[15-Jul-2009 17:22:11] <straterra> ok
[15-Jul-2009 17:22:25] <mrayzenoss> you'll probably have to edit each daemon's conf file manually
[15-Jul-2009 17:22:31] <straterra> That's fine
[15-Jul-2009 17:22:46] <straterra> and xmlrpcport for zenhub, right?
[15-Jul-2009 17:23:16] <mrayzenoss> heh, even easier just go to Settings->Daemons->Edit Config
[15-Jul-2009 17:23:22] <mrayzenoss> you can change them in the UI
[15-Jul-2009 17:23:34] <straterra> I'm faster with vi
[15-Jul-2009 17:24:00] <mrayzenoss> yeah, but we wouldn't have had to track them down if we'd gone straight to the UI
[15-Jul-2009 17:24:06] <mrayzenoss> they're documented there
[15-Jul-2009 17:24:18] <straterra> I tried to go that route..and got some error about zenhub
[15-Jul-2009 17:24:29] <straterra> something about the command not being found
[15-Jul-2009 17:24:45] <mrayzenoss> yeah, zenhub and zope have issues with that editor, but the rest of the daemons are good to go
[15-Jul-2009 17:25:30] <straterra> Question
[15-Jul-2009 17:25:44] <straterra> When I run zeneventlog genconf, it says the default port for hubport is 8789
[15-Jul-2009 17:25:49] <straterra> But..its not. It's 8081
[15-Jul-2009 17:25:58] <straterra> Is that just outdated or something?
[15-Jul-2009 17:28:26] <mrayzenoss> 8081 is the port that zenhub receives XML RPC calls on
[15-Jul-2009 17:28:34] <straterra> Right
[15-Jul-2009 17:28:36] <mrayzenoss> which is what zensendevent is doing
[15-Jul-2009 17:28:42] <straterra> Right
[15-Jul-2009 17:29:01] <mrayzenoss> 8789 is the TCP port used by Twisted for the daemons to talk to each other
[15-Jul-2009 17:29:42] <straterra> So, its the output in genconf wrong?
[15-Jul-2009 17:30:58] <mrayzenoss> no, it says "Port to use for XML-based Remote Procedure Calls (RPC)"
[15-Jul-2009 17:31:02] <straterra> I got it working now, btw.. Thank you very much
[15-Jul-2009 17:31:13] <straterra> # Port zenhub listens on.Default is 8789.,
[15-Jul-2009 17:31:14] <straterra> #  default: 8789
[15-Jul-2009 17:31:16] <straterra> #hubport 8789
[15-Jul-2009 17:31:26] <mrayzenoss> good.  I'm opening some tickets to get this doc'd better
[15-Jul-2009 17:31:39] <straterra> That's the block I'm talking about. Default is 8081, not 8789 for that
[15-Jul-2009 17:31:54] <straterra> That's spit out from zeneventlog genconf
[15-Jul-2009 17:33:42] <mrayzenoss> hmm... feel free to open lots of tickets around this
[15-Jul-2009 17:33:53] <mrayzenoss> or better yet, patches
[15-Jul-2009 17:34:03] <straterra> Sure
[15-Jul-2009 17:34:16] <straterra> Did you see what I said about getting wmi to work with modern distros?
[15-Jul-2009 17:34:19] <mrayzenoss> I don't think many people dig into this besides devs
[15-Jul-2009 17:34:41] <straterra> svn seems to work perfect on my gentoo machine, which is fully up to date
[15-Jul-2009 17:34:43] <mrayzenoss> yeah, there's a ticket for the WMI stuff
[15-Jul-2009 17:35:01] <mrayzenoss> svn trunk?
[15-Jul-2009 17:35:01] <mrayzenoss> or 2.4.x?
[15-Jul-2009 17:35:29] <straterra> let me double check
[15-Jul-2009 17:35:40] <straterra> trunk
[15-Jul-2009 17:35:52] <mrayzenoss> hmm.. maybe it's already gotten some love
[15-Jul-2009 17:36:10] <straterra> Sure seems like it
[15-Jul-2009 17:36:17] <mrayzenoss> I'm running 2.4.x from source at home, but it's on Debian
[15-Jul-2009 17:36:30] <straterra> I see
[15-Jul-2009 17:36:37] <straterra> I'll try trunk from Fedora at work tomorrow too..
[15-Jul-2009 17:37:07] <mrayzenoss> Debian and RHEL seem to be the only ones that work from source with 2.4.x
[15-Jul-2009 17:37:19] <mrayzenoss> but we plan on fixing everything in trunk for the next release
[15-Jul-2009 17:37:21] <straterra> Yeah..that's because they are based on some pretty old toolchains
[15-Jul-2009 17:37:25] <straterra> Sounds great
[15-Jul-2009 17:37:26] <mrayzenoss> yup
[15-Jul-2009 17:37:35] <mrayzenoss> we fixed OSX in 2.4.1
[15-Jul-2009 17:37:46] <straterra> You know the next milestone?
[15-Jul-2009 17:37:53] <straterra> Windows support
[15-Jul-2009 17:38:00] <straterra> (I'm really just kidding)
[15-Jul-2009 17:38:02] <mrayzenoss> King Crab will probably be starting beta in 2 months
[15-Jul-2009 17:38:11] <straterra> For windows?
[15-Jul-2009 17:38:24] <mrayzenoss> heh, probably not going to happen
[15-Jul-2009 17:39:21] <straterra> I'm already starting to like this better than nagios..even if I have no clue on how to turn off this swap alert
[15-Jul-2009 17:39:44] <mrayzenoss> zFileSystemMapIgnoreTypes
[15-Jul-2009 17:41:01] <straterra> virtualMemory is already in there
[15-Jul-2009 18:16:43] daMaestro|isBack is now known as daMaestro
[15-Jul-2009 18:39:45] <straterra> Is zenoss ipv6 capable?
[15-Jul-2009 19:55:47] <straterra> Can you make Zenoss use only wmi for windows machines as opposed to snmp too?
[15-Jul-2009 22:16:52] <ToKyNET> hello
[15-Jul-2009 22:18:02] <ToKyNET> what plugin would i need, to check some VIPs using a dig command?
[15-Jul-2009 22:54:31] <ToKyNET> anyone?
[15-Jul-2009 23:03:14] <edwardam> I setup a remote collector and have it working ... but the devices I moved there are being collected/polled by both the master/original collector and the remote one. Any ideas ?
[15-Jul-2009 23:21:46] <ToKyNET> no one seems to be up at this time..
[15-Jul-2009 23:22:04] <edwardam> Here's another one . I just setup DNS records for the master instance ... and I can't log in via the name ... but I can via the IP.
[15-Jul-2009 23:23:21] <ToKyNET> that sounds like its not resolving the name to the ip then
[15-Jul-2009 23:23:40] <ToKyNET> have you verified that it resolves to the IP?
[15-Jul-2009 23:24:44] <edwardam> it does.
[15-Jul-2009 23:24:47] <edwardam> I get the login page
[15-Jul-2009 23:24:59] <ToKyNET> then whats the problem?
[15-Jul-2009 23:25:02] <edwardam> It just says that the username/password I entered is incorrect.
[15-Jul-2009 23:25:11] <edwardam> But I *know* it's the right username/password
[15-Jul-2009 23:25:40] <edwardam> the admin password also doesn't work
[15-Jul-2009 23:26:20] <ToKyNET> hmm, not sure if zenoss has an acl system
[15-Jul-2009 23:26:41] <edwardam> it does, but it's zopes
[15-Jul-2009 23:26:53] <ToKyNET> maybe it does not like when the login comes from a hostname rather than an IP
[15-Jul-2009 23:27:10] <edwardam> that would be strange.
[15-Jul-2009 23:27:11] <ToKyNET> hostname=domain-name
[15-Jul-2009 23:27:58] <ToKyNET> i can look around on my install, but mine is vm-app im using to learn zenoss
[15-Jul-2009 23:31:19] <ToKyNET> reading zopectl.conf now
[15-Jul-2009 23:31:34] <edwardam> that was just messed up
[15-Jul-2009 23:31:42] <ToKyNET> huh?
[15-Jul-2009 23:32:03] <edwardam> I ended up modifying the python code ... although I think I may have been able to fix the issue by clearing my cookies too.
[15-Jul-2009 23:32:15] <ToKyNET> heh
[15-Jul-2009 23:34:06] <ToKyNET> this zopectl.conf is basically commented out %90
[15-Jul-2009 23:34:54] <edwardam> ToKyNET: That's standard for most zope setups
[15-Jul-2009 23:35:14] <edwardam> I am a little surprised that ZenOSS shipped that way though
[15-Jul-2009 23:35:40] <edwardam> I mostly just ended up doing: http://www.zenoss.com/community/docs/howtos/setting-up-multiple-zenperfsnmp-collectors-across-different-hosts
[15-Jul-2009 23:35:53] <edwardam> while it's working my master collector is still collecting :-(
[15-Jul-2009 23:36:03] <ToKyNET> ahh
[15-Jul-2009 23:36:09] <edwardam> I'll need to figure that out next week. ... time for a mini-vacation. l8r
[15-Jul-2009 23:36:15] <ToKyNET> k
[16-Jul-2009 00:00:46] [disconnected at Thu Jul 16 00:00:46 2009]
[16-Jul-2009 00:00:46] [connected at Thu Jul 16 00:00:46 2009]
[16-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[16-Jul-2009 00:06:41] <ToKyNET> any awake?
[16-Jul-2009 01:38:08] <rhetts> not for long
[16-Jul-2009 01:40:45] <rhetts> i'm quite late anyway
[16-Jul-2009 04:44:50] <TreeStump> hi all
[16-Jul-2009 04:45:21] <TreeStump> just got an install of zenoss core installed
[16-Jul-2009 04:45:31] <TreeStump> its nice
[16-Jul-2009 04:45:40] <TreeStump> but very different to nagios
[16-Jul-2009 04:47:01] <TreeStump> i keep having an error event come up "localhost zenjobs heartbeat failure" anyone know what that could be?
[16-Jul-2009 04:54:29] <Dieterbe> straterra> Can you make Zenoss use only wmi for windows machines as opposed to snmp too? <-- afaik yes
[16-Jul-2009 05:02:13] <Dieterbe> I don't really understand how the process monitoring is supposed to work. zenoss autodetects where the process runs, and if i configure it to alert, it will alert when a process doesn't run on a host where it first did run?
[16-Jul-2009 05:02:37] <Dieterbe> why can't i say something like "process foo must run on all devices from class bar, if it doesn't, raise alert"
[16-Jul-2009 05:44:22] <Diddi> hmm.. I just got zenoss installed from tarball, I went through the web wizard in the beginning, and now when I go to my dashboard, I get "Lost connection to the server."... I'm not sure where to start looking for the problem
[16-Jul-2009 05:47:26] <Diddi> when I go to "event console" I see that it's last updated.. quite recently.. but not in my dashboard.. what server did I loose connection to? (:
[16-Jul-2009 06:22:46] <Diddi> please hold any answers... I'll be back (:
[16-Jul-2009 06:48:17] <hackeron> anyone? < http://forums.zenoss.com/viewtopic.php?p=36929#36929
[16-Jul-2009 07:21:34] <etank> hackeron: i have the stack install on my jaunty system and it is working fine
[16-Jul-2009 07:21:41] <etank> i did not use the deb though
[16-Jul-2009 07:22:34] <etank> i used http://downloads.sourceforge.net/zenoss/zenoss-stack-2.4.2-linux.bin
[16-Jul-2009 07:22:46] <etank> it worked with no issues at all
[16-Jul-2009 07:25:09] <hackeron> etank: I tried the installer too and I get the same error Unknown error while running id zenoss | sed -e s/uid=//g -e s/\(.*//g
[16-Jul-2009 07:26:22] <hackeron> etank: I think I tried the 64bit one though, let me try the one you linked
[16-Jul-2009 07:26:59] <etank> hackeron: color me stupid but what is the command you pasted supposed to do?
[16-Jul-2009 07:27:18] <hackeron> etank: which one?
[16-Jul-2009 07:27:26] <etank> id zenoss | sed -e s/uid=//g -e s/\(.*//g
[16-Jul-2009 07:27:56] <hackeron> etank: well, it gets the uid and runs just fine, but for some reason the zenoss installer doesn't like it
[16-Jul-2009 07:28:05] <etank> just get the group id?
[16-Jul-2009 07:28:18] <etank> aah
[16-Jul-2009 07:29:00] <etank> the only issue that i have had with the 32bit stack installer is that if you uninstall it there are some pids that are not killed
[16-Jul-2009 07:29:15] <etank> so if you intend to reinstall it you have to kill them manually
[16-Jul-2009 07:29:25] <etank> zeoctl if i remember correctly
[16-Jul-2009 07:30:17] <etank> if you dont kill them then you will have to pick a port other than 8080 for the site
[16-Jul-2009 07:30:20] <hackeron> yeah, I never got that far - I can't get it to complete installation
[16-Jul-2009 07:30:42] <hackeron> should I try the 32bit installer on a 64bit system?
[16-Jul-2009 07:32:02] <etank> i only have a 32bbit box so that is all i have exp with
[16-Jul-2009 07:35:50] <hackeron> maybe this is an issue with the 64bit installer?
[16-Jul-2009 07:38:46] <hackeron> so what is this helperbinary and why does it use so much ram to install?
[16-Jul-2009 07:46:33] <etank> not sure
[16-Jul-2009 08:15:34] <straterra> Dieterbe: How? When I try to add it..The only Windows option is SNMP, not WMI
[16-Jul-2009 08:16:25] <Dieterbe> straterra: you need to disable the snmp templates, get the wmi zenpack from http://blog.dastrup.com/?p=13 and use that
[16-Jul-2009 08:16:28] <Dieterbe> straterra: that's how i do it
[16-Jul-2009 08:16:40] <straterra> I want to keep the ones for Linux..but yeah
[16-Jul-2009 08:16:42] <straterra> Ok
[16-Jul-2009 08:17:10] <Dieterbe> yeah just disable them for class /Device/Server/Windows
[16-Jul-2009 08:19:12] <straterra> Erhm..where do I go to do that?
[16-Jul-2009 08:21:49] <straterra> And that zenpack fails to install
[16-Jul-2009 08:21:57] <straterra> ERROR: zenpack command failed. Reason: 'module' object has no attribute '__file__'
[16-Jul-2009 08:26:08] <Dieterbe> hmmm odd. worked fine for me
[16-Jul-2009 08:28:34] <straterra> I'm on SVN too..so..maybe there was a change there
[16-Jul-2009 08:47:30] <etank> Dieterbe: just tested that pack with the 2.4.2 stack install and it installed fine
[16-Jul-2009 08:48:00] <Dieterbe> etank: yep for me too. it's straterra who had problems
[16-Jul-2009 08:48:09] <etank> i know
[16-Jul-2009 08:48:13] <etank> just letting you know
[16-Jul-2009 08:48:27] <etank> i can not seem to get it to bind at the device level though
[16-Jul-2009 08:48:34] <etank> i mean device class level
[16-Jul-2009 08:54:22] <etank> actually it is bound at the class but the devices are not showing the template :/
[16-Jul-2009 08:56:33] <etank> how do i make templates bound at the class level show up on the devices?
[16-Jul-2009 10:00:39] <rmatte> is there an easy way to backup devices from one Zenoss server then move them over to another?
[16-Jul-2009 10:05:39] <TreeStump> does anyone know if there is a way to get Zenoss to show bandwidth graphs of routers/servers and networking equipment just like the hardware performance graphs?
[16-Jul-2009 10:06:53] <ckrough_away> TreeStump: It does that natively.
[16-Jul-2009 10:07:49] <ckrough_away> TreeStump: Have you added a networking device that doesnt have graphs showing under Interfaces?
[16-Jul-2009 10:09:56] <ckrough_away> rmatte: zenobjectcopier.py may do it
[16-Jul-2009 10:09:59] ckrough_away is now known as ckrough
[16-Jul-2009 10:11:04] <ckrough> rmatte: $ZENHOME/bin/zenobjectcopier.py
[16-Jul-2009 10:11:38] <ckrough> rmatte: actually, that wont do it, becuase it just does the zope object. you need your rrds and stuff I assume
[16-Jul-2009 10:12:35] <ckrough> morning
[16-Jul-2009 10:12:41] <mrayzenoss> morning
[16-Jul-2009 10:12:56] <rmatte> ckrough: nah, just the devices, no performance data required
[16-Jul-2009 10:13:18] <ckrough> rmatte: May want to check it out then, I havent used it for that, but it may work
[16-Jul-2009 10:17:00] <TreeStump> ckrough_away: thank you sir, just noticed it does too!
[16-Jul-2009 10:17:20] <ckrough> TreeStump: glad its working for you
[16-Jul-2009 10:18:23] <etank> anyone know why templates are not showing up at the device level when they are bound to a class?
[16-Jul-2009 10:20:56] <ckrough> etank: do the template names exactly match the "thing" you want them to bind to?
[16-Jul-2009 10:21:36] <etank> how do you mean ckrough
[16-Jul-2009 10:22:22] <ckrough> etank: for example, a network interfaces is modeled as type "ethernetCsmacd" and the template that works with it is named "ethernetCsmacd"
[16-Jul-2009 10:22:34] <etank> i bound WMI Performance Monitors to /Devices/Server/Windows
[16-Jul-2009 10:22:36] <ckrough> what does your template do?
[16-Jul-2009 10:22:59] <etank> but it wasnt showing up on the devices in the class
[16-Jul-2009 10:23:18] <etank> i had to Reset Bindings on the devices for it to show up
[16-Jul-2009 10:23:25] <ckrough> etank: I'm not too familiar with the WMI/WIndows stuff
[16-Jul-2009 10:23:41] <mrayzenoss> ckrough: it's a community ZenPack
[16-Jul-2009 10:24:11] <ckrough> mrayzenoss: ok
[16-Jul-2009 10:54:59] <rmatte> ok, this is absolutely ridiculous
[16-Jul-2009 10:56:18] <rocket> rmatte: whats rediculous?
[16-Jul-2009 10:57:14] <rmatte> I duplicated the Interface Utilization report to the letter, I even made sure what was tabs in the original report is tabs, same amount, spacing, everything, yet it refuses to run this new custom report
[16-Jul-2009 10:58:05] <rmatte> IndentationError: expected an indented block (interfaceerrs.py, line 19)
[16-Jul-2009 10:58:14] <rmatte> the indentation of that block is identical to the other report
[16-Jul-2009 10:58:22] <rmatte> so I'm baffled as to why it won't run it properly
[16-Jul-2009 10:59:12] <rmatte> let me pastebin both report files so you can see..
[16-Jul-2009 11:00:03] <rocket> rmatte: what are you using for an editor?
[16-Jul-2009 11:00:06] <rmatte> interface utilization report (unmodified): http://pastebin.com/m48e61452
[16-Jul-2009 11:00:07] <rmatte> nano
[16-Jul-2009 11:00:27] <rocket> rmatte: you probably have a mixed combination of spaces and tabs
[16-Jul-2009 11:00:39] <rmatte> that's how it is in the original though
[16-Jul-2009 11:00:47] <rocket> rmatte: eg visually it looks the same .. but the python interpreter sees them differently
[16-Jul-2009 11:00:50] <rmatte> mixed combination of spaces and tabs
[16-Jul-2009 11:01:16] <rmatte> I've tried using all tabs, I've tried using all spaces on that block
[16-Jul-2009 11:01:18] <rmatte> nothing works
[16-Jul-2009 11:01:32] <rocket> rmatte: ick! .. I have my vimrc configured to show tabs vs spaces etc .. and to convert them to spaces
[16-Jul-2009 11:02:03] * rocket hates tabs in python code ...
[16-Jul-2009 11:02:59] * ckrough agrees with rocket. expand-tabs FTW
[16-Jul-2009 11:03:10] <rmatte> ok, so use all spaces then?
[16-Jul-2009 11:03:20] <ckrough> rmatte: what editor do you use?
[16-Jul-2009 11:03:25] <rmatte> nano
[16-Jul-2009 11:03:32] <rmatte> aka pico
[16-Jul-2009 11:03:41] <ckrough> http://www.python.org/dev/peps/pep-0008/
[16-Jul-2009 11:04:05] <ckrough> "Code lay-out"
[16-Jul-2009 11:04:49] <venturaville> rocket: did you ever get a chance to look at the minval thing in MinMaxThreshold.py?
[16-Jul-2009 11:05:11] <rocket> venturaville: not yet .. I will in about an hour .. meeting wife for lunch now
[16-Jul-2009 11:06:19] <ckrough> rocket: what vimrc line displays tabs vs spaces?
[16-Jul-2009 11:07:47] <rmatte> grrrrr, I can't possibly see what it's complaining about at this point
[16-Jul-2009 11:11:09] <rmatte> going to start over and try again
[16-Jul-2009 11:11:17] * ckrough python indentation within TAL/Templates makes me sad
[16-Jul-2009 11:14:58] <rmatte> it's still complaining, and this time it's IDENTICAL to the other report file
[16-Jul-2009 11:15:17] <int> hello! anyone knows if that possible to add "components" to device class without writing modeler plugins etc?
[16-Jul-2009 11:15:28] <rmatte> aha, I think I got it working
[16-Jul-2009 11:15:38] * rmatte waits
[16-Jul-2009 11:16:04] <ckrough> do you think line 19 is the same line 19 in the pastebin?
[16-Jul-2009 11:16:13] <ckrough> maybe nix all the space in 20 and 21
[16-Jul-2009 11:17:39] <int> ie i want to graph statistics for some groups, so i'm adding virtual device GroupsStatistics and want to add several "group" components - "group1", "group2", "group3"
[16-Jul-2009 11:18:03] <rmatte> hmmm, I think the report plugin is messed up now (indentation)... I got it from someone on the forums and the forums stripped out the identation
[16-Jul-2009 11:18:06] <rmatte> bah
[16-Jul-2009 11:18:29] <int> ie device has filesystem component and several filesystems mapped to it, i want to add another component type
[16-Jul-2009 11:18:34] <ckrough> rmatte: you could dump it into an IDE and have it auto-cleaup. Eclipse works wellf or that
[16-Jul-2009 11:20:50] <rmatte> I might just use IDLE
[16-Jul-2009 11:23:19] * int found http://georgefakhri.wordpress.com/2009/07/15/how-to-create-a-zenpack/
[16-Jul-2009 11:23:46] <ckrough> hmm, eclipse would probably choke on that mix of python/tal/html now that I think of it
[16-Jul-2009 11:23:55] <int> actually it's "How To create a Zenpack to extend device with new fields" fields = components
[16-Jul-2009 11:24:27] <ckrough> added to my ToRead tag, thanks int
[16-Jul-2009 11:25:00] <rmatte> looks like I finally fixed the thing
[16-Jul-2009 11:25:12] <ckrough> do tell
[16-Jul-2009 11:25:24] <rmatte> had to correct the indentation in the report plugin as well
[16-Jul-2009 11:25:39] <rmatte> copied the indentation from the plugin that it's based on
[16-Jul-2009 11:25:59] <rmatte> now to see how the report looks...
[16-Jul-2009 11:26:11] <ckrough> wonder of the zenoss devs deal with that, must come up frequently
[16-Jul-2009 11:26:27] <ckrough> mrayzenoss: any idea what your guys use for development?
[16-Jul-2009 11:26:38] <ckrough> mrayzenoss: ie: editor / IDE
[16-Jul-2009 11:28:08] <mrayzenoss> Eclipse, Textmate, Vi and Emacs
[16-Jul-2009 11:28:22] <ckrough> thanks
[16-Jul-2009 11:28:48] <mrayzenoss> I'm actually hacking on my .emacs right now to get tabs always as a different color
[16-Jul-2009 11:28:58] * mrayzenoss hates tabs
[16-Jul-2009 11:29:05] <schwartz> I'm with you there.
[16-Jul-2009 11:29:20] <rmatte> if you do...
[16-Jul-2009 11:29:21] <rmatte> i.getRRDValues(['ifInErrors',
[16-Jul-2009 11:29:21] <rmatte>                                           'ifOutErrors'],
[16-Jul-2009 11:29:22] <rmatte>                                          **summary)
[16-Jul-2009 11:29:29] <rmatte> does the **summary give you the total?
[16-Jul-2009 11:29:33] <mrayzenoss> I think there's a pre-processor on our SVN that rejects checkins with them
[16-Jul-2009 11:29:41] <mrayzenoss> or there should be
[16-Jul-2009 11:29:44] <ckrough> very nice
[16-Jul-2009 11:30:28] <rmatte> or would that only fetch the current value?
[16-Jul-2009 11:39:04] <rmatte> I'm basically trying to make the report show the total amount of errors ever seen based on what's in the RRD
[16-Jul-2009 11:44:00] <rmatte> nevermind, just noticed that it works based on a start date
[16-Jul-2009 11:49:48] <rmatte> ah no, that doesn't work, just does the average or maximum
[16-Jul-2009 11:49:53] <rmatte> I need to add a function to do the total
[16-Jul-2009 11:58:07] <rocket> ckrough:
[16-Jul-2009 11:59:28] <rocket> set tabstop=4
[16-Jul-2009 11:59:29] <rocket> set shiftwidth=4
[16-Jul-2009 11:59:29] <rocket> set expandtab
[16-Jul-2009 11:59:29] <rocket> set sts=4
[16-Jul-2009 11:59:29] <rocket> set list
[16-Jul-2009 11:59:31] <rocket> set listchars=tab:>.,trail:~
[16-Jul-2009 11:59:45] <rocket> ckrough: the listchars and list options
[16-Jul-2009 12:00:24] <rmatte> rocket: is there a way to do a total from an RRD file in Zenoss instead of MAXIMUM, MINIMUM, and AVERAGE?
[16-Jul-2009 12:01:15] <rocket> rmatte: if the rrd supports it then I would look at the code in the predictive threshold
[16-Jul-2009 12:01:44] <rocket> rmatte: eg that threshold sets new rrd data defaults for datapoints
[16-Jul-2009 12:02:13] <rocket> rmatte: what are you trying to total?
[16-Jul-2009 12:02:39] <rmatte> errors
[16-Jul-2009 12:03:05] <rmatte> so instead of doing a maximum value I want a total of all errors that came in during the specified period
[16-Jul-2009 12:03:28] <rmatte> so that I can make a report out of it
[16-Jul-2009 12:04:37] <rocket> rmatte: well the only way I can think to do something like that would be to fetch the data for the time window out of the rrd
[16-Jul-2009 12:04:45] <rocket> and do the manipulation with python
[16-Jul-2009 12:05:13] <rocket> rmatte: there isnt a total consolidation function for rrd .. :/
[16-Jul-2009 12:08:40] <rocket> rmatte: eg you would have to fetch the average data out of the highest resolution the rrd supports
[16-Jul-2009 12:10:07] <rocket> rmatte: a combination of rrdtool dump of an file will show you the data it contains and if its even possible
[16-Jul-2009 12:10:37] <rocket> then you can use the python bindings to do rrdtool.fetch()
[16-Jul-2009 12:11:47] <rmatte> sounds overly complicated
[16-Jul-2009 12:11:53] <rmatte> crappy
[16-Jul-2009 12:12:03] <rmatte> <tal:block metal:use-macro="here/reportMacros/macros/utilizationForm"/>
[16-Jul-2009 12:12:08] <rmatte> any idea where that macro is?
[16-Jul-2009 12:12:12] <rmatte> stored in Zope?
[16-Jul-2009 12:13:31] <rocket> nope
[16-Jul-2009 12:13:52] <rocket> Products/ZenModel/skins/zenmodel/reportMacros.pt
[16-Jul-2009 12:14:33] <rmatte> ah, unobvious place for that to be
[16-Jul-2009 12:14:34] <rmatte> lol
[16-Jul-2009 12:15:01] <venturaville> rocket: you are back?
[16-Jul-2009 12:15:17] <rocket> venturaville: I suppose
[16-Jul-2009 12:15:40] <rocket> venturaville: I cant remember what you were trying to do .. want to refresh me?
[16-Jul-2009 12:19:01] <rmatte> rocket: which file defines the way that a report is exported?
[16-Jul-2009 12:19:15] <rmatte> I created a new report based on an old one, but the export has elements of the old report
[16-Jul-2009 12:19:22] <rmatte> so I'm figuring there's some file that defines it?
[16-Jul-2009 12:19:55] <rmatte> (Just trying to get a grasp on this so that I can start creating a bunch of reports and make a ZenPack out of it eventually)
[16-Jul-2009 12:20:51] <rocket> rmatte: which report?
[16-Jul-2009 12:21:01] <rocket> rmatte: was the example one again?
[16-Jul-2009 12:21:22] <rmatte> I took the interface utilization report and modified it (someone else provided me with the code for the report plugin)
[16-Jul-2009 12:21:37] <rmatte> so it's the interface utilization report, modified to show errors
[16-Jul-2009 12:21:47] <rmatte> when I export it, it has elements of the interface utilization report
[16-Jul-2009 12:22:50] <rmatte> I figure that I can fix the interface utilization report while I'm at it
[16-Jul-2009 12:22:58] <rmatte> since it strips out the interface column when exporting
[16-Jul-2009 12:23:02] <rmatte> which basically makes it useless
[16-Jul-2009 12:23:21] <rocket> ok I havent done too much with the reports yet myself but I might guess its coming from one of the macros defined in the report
[16-Jul-2009 12:23:43] * rocket is looking
[16-Jul-2009 12:24:37] <rocket> rmatte: so when it exports do you see things like "device, speed, input, output, total, percentUsed' etc?
[16-Jul-2009 12:24:51] <rmatte> one sec, I'll do an export
[16-Jul-2009 12:25:07] <rmatte> as soon as the report decides to load...
[16-Jul-2009 12:25:10] * rmatte taps his foot
[16-Jul-2009 12:26:01] <ckrough> rmatte: you probably already know about this, but just in case: http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch14s03.html
[16-Jul-2009 12:26:06] <ckrough> might provide clues
[16-Jul-2009 12:26:37] <rmatte> no I didn't know about that lol
[16-Jul-2009 12:26:41] <rmatte> thanks, I'll have a read
[16-Jul-2009 12:26:49] <rocket> rmatte: I am thinking the exportFields header is missing 'interfaces'
[16-Jul-2009 12:27:03] <rmatte> yeh, that could be it
[16-Jul-2009 12:27:07] <rmatte> I'll have a look
[16-Jul-2009 12:27:46] <rmatte> all of a sudden my interface error report is hanging when I try to load it, bizarre, was working fine a few minutes ago
[16-Jul-2009 12:29:07] <rmatte> finally got it to load
[16-Jul-2009 12:30:42] <rmatte> you're right, that's all it was
[16-Jul-2009 12:33:07] <rocket> rmatte: file a patch and bug for the report?
[16-Jul-2009 12:33:25] <rmatte> yeh, I will
[16-Jul-2009 12:34:07] <rmatte> ok, when I export my error report, the device name, interface name, and speed import fine, but the error counts don't
[16-Jul-2009 12:34:21] <rmatte> (I managed to clean up the stuff from the other report)
[16-Jul-2009 12:34:36] <rocket> error counts dont what?
[16-Jul-2009 12:35:10] <rmatte>     exportFields python:['device', 'interface', 'speed', 'incoming', 'outgoing'];
[16-Jul-2009 12:35:12] <rmatte>         <th tal:replace="structure python:here.ZenTableManager.getTableHeader(
[16-Jul-2009 12:35:12] <rmatte>                             tableName,'incoming','Incoming Errors')"/>
[16-Jul-2009 12:35:12] <rmatte>         <th tal:replace="structure python:here.ZenTableManager.getTableHeader(
[16-Jul-2009 12:35:12] <rmatte>                             tableName,'outgoing','Outgoing Errors')"/>
[16-Jul-2009 12:35:24] <rmatte>            <td tal:content="python: r.fmt('%8.2f', r.inerror)"/>
[16-Jul-2009 12:35:25] <rmatte>            <td tal:content="python: r.fmt('%8.2f', r.outerror)"/>
[16-Jul-2009 12:35:29] <rmatte> they don't show up in the export
[16-Jul-2009 12:35:34] <rmatte> the columns are blank
[16-Jul-2009 12:35:44] <rmatte> the column titles show up, but there are no values
[16-Jul-2009 12:35:56] <hackeron> wow, looks like zenoss is running out of ram when installing on a VMS with 256mb ram - what's the ram requirement to install zenoss?
[16-Jul-2009 12:36:33] <rocket> rmatte: so these are new things you added?
[16-Jul-2009 12:37:57] <rmatte> it's a seperate report based on the original utilization report
[16-Jul-2009 12:37:57] <rocket> rmatte: is there actually data in the r.inerror variable?
[16-Jul-2009 12:38:09] <rmatte> rocket: yes, there is
[16-Jul-2009 12:38:16] <rmatte> when I run the report I see the data
[16-Jul-2009 12:38:19] <rmatte> when I export it, I don't
[16-Jul-2009 12:38:25] <rocket> rmatte: ok just a sec
[16-Jul-2009 12:38:26] <venturaville> rocket: I was looking at how or why minval is not getting reset in some cases on the thresholds
[16-Jul-2009 12:38:28] <rmatte> k
[16-Jul-2009 12:38:42] <venturaville> or at least where and how createThresholdInstance gets called again
[16-Jul-2009 12:40:30] <rocket> rmatte: do you have inerror in the exportfields thing?
[16-Jul-2009 12:40:40] <rocket> venturaville: ok just a sec
[16-Jul-2009 12:40:56] <rmatte> rocket: I have 'incoming' and 'outgoing'
[16-Jul-2009 12:41:04] <rmatte> since it goes by table header
[16-Jul-2009 12:41:25] <rocket> rmatte: the name might have to match the header?
[16-Jul-2009 12:41:52] <rmatte> oh, maybe
[16-Jul-2009 12:41:54] <hackeron> what's the minimum amount of ram have you guys successfully installed zenoss on? -- tried on 256mb free ram, ran out -- now trying on 512mb
[16-Jul-2009 12:41:56] <rocket> venturaville: createThresholdInstance is called in a couple places
[16-Jul-2009 12:42:09] <venturaville> zenhub for I believe
[16-Jul-2009 12:42:17] <venturaville> which I am guessing the main one
[16-Jul-2009 12:42:35] <rocket> venturaville: ZenModel/ThresholdGraphPoint.py is one
[16-Jul-2009 12:42:39] <rmatte> rocket: nope, that's not the case
[16-Jul-2009 12:42:50] <rocket> ZenModel/RRDGraph.py
[16-Jul-2009 12:42:56] <rocket> ZenModel/RRDView.py
[16-Jul-2009 12:43:01] <ckrough> hackeron: minimum RAM to function well is dependent on your environment
[16-Jul-2009 12:43:11] <ckrough> hackeron: just to install shouldnt need much I imagine
[16-Jul-2009 12:43:14] <mrayzenoss> hackeron: I've installed on 384 megs, but only monitored 7 things
[16-Jul-2009 12:43:14] <hackeron> ckrough: not to function, to install
[16-Jul-2009 12:43:15] <rmatte> rocket: the export name needs to match the table name, which it does
[16-Jul-2009 12:43:23] <hackeron> ckrough: installation is taking >300MB right now
[16-Jul-2009 12:43:32] <hackeron> mrayzenoss: ok, cool - thanks
[16-Jul-2009 12:43:48] <hackeron> mrayzenoss: it fails to install on 256mb
[16-Jul-2009 12:43:52] <mrayzenoss> hackeron: I know someone recently got Zenoss installed and running on a SheevaPlug, built from source
[16-Jul-2009 12:44:06] <ckrough> hackeron: we are talking RAM here, right?
[16-Jul-2009 12:44:11] <mrayzenoss> yeah
[16-Jul-2009 12:44:25] <mrayzenoss> I've got a G3 iBook running Debian PPC
[16-Jul-2009 12:44:25] <hackeron> mrayzenoss: yeah, but it's very difficult to build from source on recent distributions, right? (ubuntu jaunty here)
[16-Jul-2009 12:44:42] <mrayzenoss> hackeron: yeah, Debian works
[16-Jul-2009 12:44:58] <mrayzenoss> I just put it on Debian 5 on a 1 gig RAM box
[16-Jul-2009 12:45:03] <mrayzenoss> (from source)
[16-Jul-2009 12:45:27] <rocket> rmatte: I see a ZenReports/plugins/interfaces.py is there anything hidden in there that might be related?
[16-Jul-2009 12:45:33] <hackeron> heh, well, it's ubuntu jaunty here pre-installed
[16-Jul-2009 12:45:41] <rocket> rmatte: eg needing to implement python code for the export perhaps .. :/
[16-Jul-2009 12:45:45] <rmatte> weird, the report errors out if I change <td tal:content="python: r.fmt('%8.2f', r.inerror)"/> to <td tal:content="python: r.fmt(r.inerror)"/>
[16-Jul-2009 12:45:58] <hackeron> here we go, fingers crossed - so far went past the bit it died with 256mb ram
[16-Jul-2009 12:46:16] <hackeron> out of my 512mb ram, under 200mb left - come on, please be enough, lol
[16-Jul-2009 12:46:29] <hackeron> mysql started
[16-Jul-2009 12:46:39] <hackeron> encouraging
[16-Jul-2009 12:47:08] <hackeron> 160mb left...
[16-Jul-2009 12:47:17] <hackeron> 140mb left
[16-Jul-2009 12:48:15] <ckrough> hackeron: what tool are you using to show that
[16-Jul-2009 12:48:17] <hackeron> mrayzenoss: already using more than 386mb and still installing - hmm?
[16-Jul-2009 12:48:22] <hackeron> ckrough: free -m
[16-Jul-2009 12:48:32] <ckrough> hackeron: is it going into SWAP
[16-Jul-2009 12:48:34] <hackeron> ckrough: showing 0mb free ram, 130mb cached
[16-Jul-2009 12:48:44] <hackeron> ckrough: there is no swap, it's a VPS with no swap
[16-Jul-2009 12:48:47] <mrayzenoss> the installer isn't the same as running
[16-Jul-2009 12:49:03] * ckrough shreaks
[16-Jul-2009 12:49:03] <hackeron> mrayzenoss: ok, but why does it take >400mb of ram to isntall, lol?
[16-Jul-2009 12:49:20] <mrayzenoss> that's BitRock's deal, they provide the installers
[16-Jul-2009 12:49:36] <hackeron> <100mb left now - still installing
[16-Jul-2009 12:49:45] <hackeron> ok, installed! - yay!
[16-Jul-2009 12:49:53] <hackeron> back to 450MB free ram
[16-Jul-2009 12:50:01] <mrayzenoss> yay! 
[16-Jul-2009 12:50:12] <mrayzenoss> now you can run Zenoss in a really limited environment
[16-Jul-2009 12:50:27] <hackeron> 512mb ram is not enough?
[16-Jul-2009 12:50:50] <rmatte> rocket: ah, it's based on the value in the plugin apparently
[16-Jul-2009 12:51:18] <mrayzenoss> the lack of swap concerns me, but you should be able to run with tens(?) of devices
[16-Jul-2009 12:51:27] <rocket> rmatte: so you needed to define a plugin? or extend the plugin?
[16-Jul-2009 12:51:28] <mrayzenoss> really depends how much and what you monitor
[16-Jul-2009 12:52:20] <hackeron> mrayzenoss: hmm? - why does zenoss need so much ram?
[16-Jul-2009 12:53:00] <hackeron> I only have 6gb of space here, I guess I could dedicate a gb to a swap partition
[16-Jul-2009 12:55:21] <mrayzenoss> looking at my current home install, 10 devices being monitored, looks like MySQL is 120 megs and all the Zenoss daemons and Zope adds up to about 300 megs
[16-Jul-2009 12:55:38] <rmatte> rocket: the plugin defined those values as "inerror" and "outerror" instead of "incoming" and "outgoing". It turns out that the exportFields function uses the names as defined in the plugin rather than in the report
[16-Jul-2009 12:56:26] <rmatte> so I just changed the names of the tables to inerror and outerror, then added them to exportFields and voila
[16-Jul-2009 12:56:31] <hackeron> mrayzenoss: omg! -- why so much!
[16-Jul-2009 12:56:40] <mrayzenoss> hackeron: you could save memory by turning off daemons you won't be using http://www.zenoss.com/community/wiki/tips-and-tricks/RunningOnlySomeDaemons
[16-Jul-2009 12:56:58] <rmatte> hackeron: it's python, so it's going to use up memory
[16-Jul-2009 12:57:04] <mrayzenoss> hackeron: each daemon runs separately as well
[16-Jul-2009 12:57:20] <hackeron> rmatte: no, thats perl -- python is not bad with ram. I have a lot of python software
[16-Jul-2009 12:57:59] <mrayzenoss> 17 Zenoss processes on my box
[16-Jul-2009 12:59:39] <hackeron> mrayzenoss: I just started zenoss and I have 20
[16-Jul-2009 13:00:05] <hackeron> eating all of my ram and 400mb of my swap already
[16-Jul-2009 13:00:08] <hackeron> wtf
[16-Jul-2009 13:01:54] <hackeron> heh, why does it need 1GB of ram without monitoring a single host?
[16-Jul-2009 13:02:06] <int> can i manually add component instance to custom device? i added new device type and component type to zenpack, created device with new type, but can't figure how can i add new component to it
[16-Jul-2009 13:02:32] <mrayzenoss> hackeron: I wonder if the startup wizard is memory intensive?
[16-Jul-2009 13:02:49] <hackeron> I hope so! -- so far this is pretty shocking
[16-Jul-2009 13:02:59] <rocket> rmatte: great .. glad its working for you now
[16-Jul-2009 13:03:41] <int> i have dialog_addNewComponentType.pt - can i call it via web somehow?
[16-Jul-2009 13:06:56] <rmatte> rocket: thanks for the help
[16-Jul-2009 13:07:33] <rmatte> mrayzenoss: I found out why the interface utilization report doesn't include the interface column when it's exported, it's just one little thing that's missing in the rpt file. I'll file a trac for it soon.
[16-Jul-2009 13:07:58] <rocket> venturaville: ping
[16-Jul-2009 13:08:01] <mrayzenoss> rmatte: awesome, thanks
[16-Jul-2009 13:08:29] <venturaville> rocket: I am here... distracted :-)
[16-Jul-2009 13:08:48] <rocket> venturaville: ok so are you mostly interested in alerts on the thresholds?
[16-Jul-2009 13:08:53] <venturaville> nope
[16-Jul-2009 13:08:57] * ckrough is looking for scripts to put large numbers of devices into maintenance. Too many for GUI
[16-Jul-2009 13:09:00] <rocket> venturaville: or the graphing?
[16-Jul-2009 13:09:04] <venturaville> just the threshold checking of the minval
[16-Jul-2009 13:09:09] <venturaville> not graphing
[16-Jul-2009 13:09:11] <ckrough> I have something already, but want to merge/compare if others have done the same
[16-Jul-2009 13:09:41] <venturaville> I have heard before that sometimes you have to graph thresholds to get them to be updated properly
[16-Jul-2009 13:10:02] <venturaville> but I haven't had a good doc anywhere that cooberates that or explains it
[16-Jul-2009 13:10:31] <hackeron> hey, I added the wrong network for auto discovery in the setup - where do I cancel and add the correct network?
[16-Jul-2009 13:10:38] <rocket> venturaville: ok I think I have it remembered again
[16-Jul-2009 13:10:49] <mrayzenoss> hackeron: go to Settings->Jobs and cancel it
[16-Jul-2009 13:10:55] <rocket> venturaville: zenhub is doing the checking of the thresholds in the checkThresholds function
[16-Jul-2009 13:11:01] <venturaville> yep
[16-Jul-2009 13:11:09] <venturaville> I have gotten that far now at least
[16-Jul-2009 13:11:14] <rocket> venturaville: everytime checkThresholds is run it gets a list of datapoint
[16-Jul-2009 13:11:17] <rocket> s
[16-Jul-2009 13:11:26] <hackeron> mrayzenoss: thanks, cancelled - how do I run it again for the correct network?
[16-Jul-2009 13:11:56] <rocket> for that list of datapoints it grabs the rrdTemplate.threshold for it
[16-Jul-2009 13:12:14] <mrayzenoss> hackeron: go to Networks and add it there
[16-Jul-2009 13:12:38] <mrayzenoss> hackeron: or go to Add Device->Easy Add... and you'll be back in the Wizard
[16-Jul-2009 13:12:39] <rmatte> hmmm, the interface utilization report displays input and output as KB in the report, but appears to export as straight packets
[16-Jul-2009 13:12:47] <rocket> so basically its getting MinMaxThresholds object class
[16-Jul-2009 13:12:51] * rmatte reviews the code
[16-Jul-2009 13:12:52] <hackeron> mrayzenoss: how do I remove the old network?
[16-Jul-2009 13:13:09] <venturaville> rocket: basically someone here wrapped some items around getMinVal() in MinMaxThreshold.py and it only gets called once
[16-Jul-2009 13:13:22] <venturaville> items being ...some logging statements
[16-Jul-2009 13:14:12] <hackeron> mrayzenoss: ah, found it, never mind - thanks
[16-Jul-2009 13:14:44] <rmatte> hmmmm, when exporting the report it appears to export the straight r.speed, r.input, r.output, and r.total values instead of performing the r.humanBits function on them
[16-Jul-2009 13:15:35] <rmatte> though it exports the percentage perfectly fine
[16-Jul-2009 13:16:21] <rocket> venturaville: I would gather its run only once when zenhub starts up the first time and is checking that datapoints threshold instance?
[16-Jul-2009 13:16:32] <venturaville> yep
[16-Jul-2009 13:16:49] <venturaville> so it gets a number and uses it, and then never updates it again
[16-Jul-2009 13:17:16] <venturaville> unless someone manually hits save on the threshold page
[16-Jul-2009 13:17:18] <venturaville> .......
[16-Jul-2009 13:18:26] <rocket> venturaville: how else would it be updating that number in your case?
[16-Jul-2009 13:19:03] <venturaville> the minval filed in the threshold has an equation which includes a getRRDValue call.....
[16-Jul-2009 13:21:55] <rocket> venturaville: so its set via a rpn?
[16-Jul-2009 13:24:17] <venturaville> Min Value = here.getRRDValue('postsummary_total') * .99
[16-Jul-2009 13:24:19] <venturaville> in this case
[16-Jul-2009 13:24:29] <venturaville> in a Min/Max Threshold
[16-Jul-2009 13:24:51] <venturaville> which is fine ... it runs it correctly
[16-Jul-2009 13:24:58] <venturaville> and gets a number
[16-Jul-2009 13:25:04] <venturaville> but then never changes later.........
[16-Jul-2009 13:33:54] <ckrough> rmatte: heh, look what I just ran across: http://forums.zenoss.com/viewtopic.php?t=8664&highlight=bulk
[16-Jul-2009 13:33:59] <ckrough> 2) zendevicedump / zendeviceload in $ZENHOME/bin - though they are just harness files for ExportDevices.py and ImportDevices.py in $ZENHOME/Products/ZenRelations.
[16-Jul-2009 13:34:03] <ckrough> This gives you a way to dump existing devices out to an xml file, complete with information on device class, performance monitor server, locations, systems and groups.
[16-Jul-2009 13:34:12] <ckrough> you were looking for that earlier I think, no?
[16-Jul-2009 13:44:49] <hackeron> if snmpd wasn't installed on a device during auto detection - will I need to run autodetection again, or will it detect snmpd eventually? - if so how long will it take?
[16-Jul-2009 13:46:11] <ckrough> hackeron: I believe the default install will remodel (redetect/analyse) devices every 12 hours
[16-Jul-2009 13:46:23] <hackeron> ckrough: ok, thanks!
[16-Jul-2009 13:46:43] <ckrough> hackeron: it should pick up the details then. You will need to move the device from /Discovered to whatever class makes sense for that device[s] to get the full benefit
[16-Jul-2009 13:47:17] <ckrough> hackeron: /Monitors  /Hub  /localhost  /localhost  -> Modeler Cycle Interval (mins)
[16-Jul-2009 13:47:53] <ckrough> hackeron: if you are on a single-server install you should see that value on the Overview page at .../dmd/Monitors/Performance/localhost
[16-Jul-2009 13:48:51] <hackeron> ckrough: I have an openvpn and zenoss detected the 2 servers connected to the openvpn -- so I should move them from Discovered to Server?
[16-Jul-2009 13:49:23] <ckrough> hackeron: order a more specific subclass of /Server, yes
[16-Jul-2009 13:49:56] <hackeron> ckrough: ok, well, my server is remote, linux and ssh -- should I put them under all 3?
[16-Jul-2009 13:50:10] <ckrough> hackeron: server/linux is probably what you want
[16-Jul-2009 13:50:17] <rmatte> is there any quick way for me to output a list of devices that have local thresholds applied to them?
[16-Jul-2009 13:50:31] <mrayzenoss> rmatte: that would be handy
[16-Jul-2009 13:50:36] <rmatte> there were a bunch of local thresholds applied to certain servers but our engineer lost the lists of which ones he tuned
[16-Jul-2009 13:50:43] <rmatte> and I don't want to have to go through them 1 by 1
[16-Jul-2009 13:50:47] <rmatte>
[16-Jul-2009 13:51:11] <ckrough> hackeron: zenoss collects information (details, performance) based on templates that are applied and inherited by the /Class, so a server in /Linux will be inspected for Linuxy things, etc...
[16-Jul-2009 13:51:38] <mrayzenoss> rmatte: hmmm... a report would be handy like http://www.zenoss.com/community/projects/zenpacks/transform-report
[16-Jul-2009 13:51:44] <hackeron> ckrough: cool, very useful -- ok, it also detected the localhost server on 10.8.0.1 -- should I delete that device?
[16-Jul-2009 13:51:59] <hackeron> ckrough: (I told it to autodetect on 10.8.0.0/24)
[16-Jul-2009 13:52:23] <ckrough> hackeron: I will redetect it in <12hrs if you do But you probably want that so you can see performance data, etc...
[16-Jul-2009 13:52:39] <ckrough> hackeron: so no, I wouldnt delete it
[16-Jul-2009 13:53:10] <hackeron> ckrough: ok, cool -- also, servers connecting to the openvpn can randomly get assigned different IP addresses each time they connect - how does zenoss handle that?
[16-Jul-2009 13:53:28] * ckrough looks at mrayzenoss
[16-Jul-2009 13:53:36] <ckrough> I should know that answer, but it slips my mind
[16-Jul-2009 13:53:37] <mrayzenoss> hackeron: poorly
[16-Jul-2009 13:53:49] <mrayzenoss> hackeron: Zenoss uses the IP address to track machines
[16-Jul-2009 13:54:22] <mrayzenoss> so if the machines are going to get random IPs each time they restart, you may need to do some hacking
[16-Jul-2009 13:54:26] <hackeron> mrayzenoss: so if the hostname, all hardware and everything else suddenly changes, it won't care?
[16-Jul-2009 13:54:44] <mrayzenoss> that's why it remodels every 12 hours, to stay on top of changes
[16-Jul-2009 13:54:55] <mrayzenoss> the idea being small changes of course
[16-Jul-2009 13:55:23] <hackeron> mrayzenoss: so say disk utilization keeps changing from 30% on a 2tb drive to 80% on a 300gb drive - it will just silently accept that?
[16-Jul-2009 13:55:29] <rmatte> remodeling is so intensive though when you're dealing with devices that you are monitoring across slow circuits
[16-Jul-2009 13:55:52] <mrayzenoss> hackeron: I'm not exactly sure, it's more likely it would throw alerts before getting remodeled
[16-Jul-2009 13:56:01] <rmatte> the modelling should only take place once and then Zenoss should check very specific things for changes
[16-Jul-2009 13:56:05] <rmatte> such as interfaces and disks
[16-Jul-2009 13:56:10] <hackeron> mrayzenoss: ah, ok, I'll check that out
[16-Jul-2009 13:56:43] <mrayzenoss> hackeron: assuming you have some sort of control over your DHCP server, you can probably configure it to give the same machines the same IPs every time
[16-Jul-2009 13:57:10] <mrayzenoss> dhcpd.conf has those settings
[16-Jul-2009 13:57:16] <mrayzenoss> and most routers do as well
[16-Jul-2009 13:57:28] * mrayzenoss uses an AirPort Extreme configured that way
[16-Jul-2009 13:57:32] <hackeron> mrayzenoss: it's not dhcp, it's handled by openvpn
[16-Jul-2009 13:58:02] <hackeron> mrayzenoss: and it seems there are 2 configs, 1 limiting one where you have a separate config for each host, and one where any host with a valid public key can connect and see the server
[16-Jul-2009 14:00:25] <mrayzenoss> I'm not super familiar with OpenVPN, but it sounds like the separate config is probably the way to go, probably something you could script
[16-Jul-2009 14:01:32] <mrayzenoss> Might I recommend http://oreilly.com/catalog/9780596515829/ since they gave me a free copy?
[16-Jul-2009 14:01:48] <gemineye> Is there a way to get Zenoss to report up/down status not based on pings but instead on SNMP queries? We have high network traffic that clogs access to our switches and the ICMP ping packets drop. I've looked through the forums but can't find any help there,
[16-Jul-2009 14:02:53] <mrayzenoss> etank: looks like you're the first review of the Python book I just recommended
[16-Jul-2009 14:03:16] <rmatte> gemineye: you could just disable ping monitoring in zProperties, then if you get an SNMP Agent Down alert, there's a good chance the device is down
[16-Jul-2009 14:04:45] <gemineye> But those types of alerts aren't considered in the uptime metrics, also when a system is considered down, other services that go down aren't alerted on.
[16-Jul-2009 14:06:26] <hackeron> mrayzenoss: thanks, I'll get it
[16-Jul-2009 14:06:57] <gemineye> Our network team is telling me they won't use Zenoss to monitor their switches because of this. They've gotten tired of all the ping failures that occur because the ICMP packets are dropped by the switches due to bandwidth constraints.
[16-Jul-2009 14:07:21] <hackeron> hmmm, so I don't get it, why does zenoss use 800MB of ram monitoring 2 devices? - I mean requesting snmp data takes what, 500KB of ram? -- pretty big leap from 500KB to 800MB...
[16-Jul-2009 14:09:46] <rmatte> hackeron: it is running the Zope framework, plus 14 daemons, plus a MySQL database, and the whole Zenoss/Zope systems are based on python, which needs an interpreter to execute the code, so it's fairly memory intensive
[16-Jul-2009 14:10:19] <hackeron> rmatte: I'm excluding the mysql database here - also, why does each of the daemons have to take up to 90MB of ram??
[16-Jul-2009 14:10:29] <hackeron> rmatte: what's it doing with all that ram?
[16-Jul-2009 14:10:53] <rmatte> Zenoss caches a heck of a lot of stuff to improve performance
[16-Jul-2009 14:11:06] <hackeron> rmatte: like what?
[16-Jul-2009 14:11:29] <rmatte> I'd have to go through the code to give you specifics, I know that it caches ZenPack info for instance
[16-Jul-2009 14:11:36] <rmatte> It caches RRD values
[16-Jul-2009 14:11:37] <hackeron> rmatte: 800MB -- ok, caching zope (hmm, why?) - 20MB -- where's the 780MB gone?
[16-Jul-2009 14:12:07] <rmatte> like I said, it's a wackload of python scripts, python is not as efficients, as say something like C
[16-Jul-2009 14:12:52] <rmatte> either way, I don't see what the big deal is, most people have 2 to 4GB of RAM in their workstations these days, let alone their servers
[16-Jul-2009 14:12:58] <hackeron> rmatte: it's not a python issue, I have python processes running 20 threads each collecting a hell of a lot of data and writing to database taking 20MB of ram
[16-Jul-2009 14:13:15] <rocket> hackeron: it consumes more ram because the enterprise equivalent uses alot of network and daemon tools etc
[16-Jul-2009 14:13:27] <hackeron> rmatte: not on a VPS and I'd rather use that memory to cache my large database
[16-Jul-2009 14:13:28] <rocket> hackeron: eg it is built to be distributed
[16-Jul-2009 14:13:47] <hackeron> rocket: I still dont understand where all that ram is going...
[16-Jul-2009 14:13:53] <rocket> hackeron: zope....
[16-Jul-2009 14:14:17] <hackeron> time to migrate to django maybe
[16-Jul-2009 14:14:19] <mrayzenoss> zope uses an object database to store everything in the UI minus perf data and events
[16-Jul-2009 14:14:24] <rocket> hackeron: basically each daemon loads quite a bit of zope more than one time ..
[16-Jul-2009 14:14:35] <rmatte> I'm monitoring over 350 devices on one Zenoss box, and I have a ton of custom command based scripts being executed by Zenoss for various tasks, I'm pulling in at just under 2GB usage
[16-Jul-2009 14:15:25] <rocket> hackeron: I am not saying that there arent areas that could be redone however
[16-Jul-2009 14:15:50] <rmatte> hackeron: yeh, agreed that it's not as efficient as it could be, but it's honestly not bad
[16-Jul-2009 14:15:52] <hackeron> mrayzenoss: Im not familiar with zope, but sounds like what django does, it has ORM - except it takes almost no ram
[16-Jul-2009 14:16:32] <hackeron> rmatte: heh, zenoss is fantastic but this ram use is just scarming me
[16-Jul-2009 14:17:00] <rocket> hackeron: flexibility comes at a cost ..
[16-Jul-2009 14:17:22] <mrayzenoss> trust me, nobody is crazy about Zope around here, but we use it for almost everything (UI, object DB, permissions, hierarchies)
[16-Jul-2009 14:17:38] <mrayzenoss> and not much effort has been put into reducing our memory footprint
[16-Jul-2009 14:17:44] <hackeron> mrayzenoss: not to start a flamewar but why not django?
[16-Jul-2009 14:17:52] <mrayzenoss> where was django in 2002?
[16-Jul-2009 14:17:59] <hackeron> ok, good point
[16-Jul-2009 14:18:19] <rmatte> they'd have to redesign the whole system to use django, which would be ridiculous lol
[16-Jul-2009 14:18:28] <mrayzenoss> it's been discussed by dev to switch to something else
[16-Jul-2009 14:18:51] <rmatte> Zope has some nice stuff, like the LDAP Authentication plugins
[16-Jul-2009 14:18:52] <mrayzenoss> right now we're getting off of ancient versions of Zope and Python, then we'll start seeing what our options are
[16-Jul-2009 14:19:05] <rmatte> django would need to support LDAP (if it doesn't already)
[16-Jul-2009 14:19:11] <mrayzenoss> King Crab will have Zope 2.11.2
[16-Jul-2009 14:19:22] <mrayzenoss> Stone Crab will have Zope 2.12 and Python 2.6 in theory
[16-Jul-2009 14:19:29] <cgibbons> xfingers
[16-Jul-2009 14:19:43] <rmatte> I hope the Zope upgrades go smoothly
[16-Jul-2009 14:19:47] <rmatte> I don't want to lose data
[16-Jul-2009 14:19:51] <mrayzenoss> then we'll be on the latest junk and can start looking at pieces to rewrite or move to other stuff
[16-Jul-2009 14:20:05] <hackeron> rmatte: django supports ldap
[16-Jul-2009 14:20:07] <rmatte> "the latest junk" haha
[16-Jul-2009 14:20:29] <rmatte> hackeron: cool, then it would be suited for the job
[16-Jul-2009 14:20:30] <rocket> hackeron: its easy to ask why .. its hard to dig in and understand design considerations over a period of time ..
[16-Jul-2009 14:21:05] <hackeron> rocket: heh, yeah - no, zope was pretty much the only sane choice available for a long time
[16-Jul-2009 14:21:40] <hackeron> rocket: now though, I donno, not a fan - zenoss is pretty impressive as long as I don't run top and see what it's doing to my resources
[16-Jul-2009 14:22:13] <rocket> hackeron: I am sure you could ask anyone in here if zenoss could do things better and they would say yes .. but compared to what else is out there it does fit a specific need ..
[16-Jul-2009 14:22:16] <hackeron> haha, it just gave me an error about my swap space being low -- guess it's still munching ram
[16-Jul-2009 14:22:54] <hackeron> rocket: oh absolutely, no way I would have nagios going and monitoring just about everything within what, 30 minutes
[16-Jul-2009 14:24:09] <rmatte> hackeron: it starts off at a certain amount of ram and keeps munching for a while up until a point
[16-Jul-2009 14:25:19] <gemineye> Does anyone use sendpage with Zenoss?
[16-Jul-2009 14:25:23] <rmatte> we use 2gb ram minimum on our Zenoss boxes
[16-Jul-2009 14:25:25] <etank> mrayzenoss: cool
[16-Jul-2009 14:25:58] <hackeron> rmatte: that's out of my price range at the moment -- I'm a broke little startup firm, lol
[16-Jul-2009 14:26:48] <cgibbons> hmmm
[16-Jul-2009 14:27:05] <rmatte> hackeron: you can't even work out 300 bucks to put together a workstation with 2GB of RAM in it?
[16-Jul-2009 14:27:17] <rmatte>
[16-Jul-2009 14:28:30] <hackeron> rmatte: not on the shitty, unreliable ADSL I have, no -- so I'm using a VPS hosted by these guys: prgmr.com -- $115 per year for 512mb, 12gb disk and 80gb transfer VPS
[16-Jul-2009 14:29:30] <hackeron> rmatte: problem isn't the hardware, is the internet connection - so I either get a rack and put my own hardware which is way out of price range or get a VPS
[16-Jul-2009 14:30:17] * etank loves getting free books (t-shirts too)
[16-Jul-2009 14:33:16] <rmatte> hackeron: you could rent a rackmount box from a colo company for like $70/month
[16-Jul-2009 14:34:12] <hackeron> rmatte: yikes!
[16-Jul-2009 14:34:32] <mrayzenoss> maybe ckrough could get you a discount
[16-Jul-2009 14:34:51] <hackeron> rmatte: also $70 per month will get me some kind of crappy little celeron -- here I have an SMP 2 x quad core opteron -- but not much ram
[16-Jul-2009 14:34:52] <rmatte> http://order.1and1.com/xml/order/ServerPremium;jsessionid=83D8FEC551208205C2436271146A8B33.TC61a?__frame=_top&__lf=Static&linkType=btn&linkOrigin=Home
[16-Jul-2009 14:35:10] <rmatte> 1and1 has some of the best prices I've seen, check out that page
[16-Jul-2009 14:35:30] <rmatte> quite honestly, it's not the processing power that kills you with Zenoss, it's the RAM
[16-Jul-2009 14:35:42] <mrayzenoss> followed by the IO
[16-Jul-2009 14:35:47] <rmatte> less processing power just means the monitoring cycles take a bit longer
[16-Jul-2009 14:35:50] <rmatte> and yeh, the IO
[16-Jul-2009 14:36:01] <hackeron> rmatte: what kind of bandwidth?
[16-Jul-2009 14:36:23] <rmatte> 1and1 is 100mbit I believe
[16-Jul-2009 14:36:35] <rmatte> you could contact them to verify
[16-Jul-2009 14:36:57] <mrayzenoss> their Linux VPS looks promising
[16-Jul-2009 14:37:33] <hackeron> rmatte: what's the advantage of paying $50 per month when I can have 1gb of ram for $16 per month?
[16-Jul-2009 14:37:40] <hackeron> mrayzenoss: who's prgmr?
[16-Jul-2009 14:38:34] <mrayzenoss> where's that?
[16-Jul-2009 14:38:38] <rmatte> because you can't realistically do anything useful with 1GB of RAM when you're talking about monitoring
[16-Jul-2009 14:39:01] <rmatte> unless you want to use nagios and spend forever setting it up to be super lightweight
[16-Jul-2009 14:39:39] <rmatte> I'd imagine you'd be pulling in more than $50 each month for the monitoring services, no?
[16-Jul-2009 14:40:03] <hackeron> rmatte: I'm not pulling anything for the monitoring servers
[16-Jul-2009 14:40:12] <hackeron> rmatte: it's all in development stages atm
[16-Jul-2009 14:40:15] <hackeron> rmatte: did I show you what I do?
[16-Jul-2009 14:40:20] <rmatte> right, but eventually you will be
[16-Jul-2009 14:40:24] <rmatte> nope
[16-Jul-2009 14:40:33] <hackeron> rmatte: when I can afford it, I'll get more ram
[16-Jul-2009 14:40:43] <hackeron> rmatte: here: http://hackeron.dyndns.org:82/cctv/events
[16-Jul-2009 14:40:46] <hackeron> rmatte: user/pass: demo/demo
[16-Jul-2009 14:41:11] <mrayzenoss> what does the motion box?
[16-Jul-2009 14:41:17] <rmatte> did you write that?
[16-Jul-2009 14:41:23] <hackeron> rmatte: yeah, in python
[16-Jul-2009 14:41:29] <rmatte> nice, I like
[16-Jul-2009 14:42:10] <rmatte> so it only takes photos when it detects movement?
[16-Jul-2009 14:42:11] <hackeron> rmatte: I have 20 python threads running there taking 20MB - pulling data at 200 fps -- actual motion processing is C, but motion info is fed to python via a unix socket
[16-Jul-2009 14:42:14] <rmatte> or is this all proof of concept?
[16-Jul-2009 14:42:33] <hackeron> rmatte: it's a finished product, but currently I sell it as just a DVR I come and set up -- no monitoring or anything
[16-Jul-2009 14:42:55] <hackeron> rmatte: so thinking of having both monitoring with zenoss and set up the openvpn to give people access even if the DVR is behind a NAT
[16-Jul-2009 14:43:00] <rmatte> oooh, I like the drag and drop feature
[16-Jul-2009 14:43:02] <hackeron> rmatte: and eventually charging for that
[16-Jul-2009 14:43:04] <rmatte> and I see it's live, very cool
[16-Jul-2009 14:43:13] <hackeron> rmatte: yep, want me to go outside and wave?
[16-Jul-2009 14:43:23] <rmatte> sure lol
[16-Jul-2009 14:43:24] <rmatte>
[16-Jul-2009 14:43:50] <rmatte> lol, I see him
[16-Jul-2009 14:44:02] <rmatte> pretty cool system
[16-Jul-2009 14:44:27] <hackeron> yeah, so I don't get why that works on 128mb of ram with 8 h264 encoding processes
[16-Jul-2009 14:44:31] <hackeron> and zenoss needs 2gb, lol
[16-Jul-2009 14:44:42] <rmatte> the answer is: Zope
[16-Jul-2009 14:44:50] <hackeron> heh, fair enough
[16-Jul-2009 14:44:52] <rmatte> 'tis a beast
[16-Jul-2009 14:44:54] <hackeron> yeah, I use django there
[16-Jul-2009 14:45:12] <rmatte> the response times are quite good
[16-Jul-2009 14:45:34] <mrayzenoss> hackeron: these guys use zenoss with zlinks from devices to the camera images: http://www.mts-iss.com/blog/
[16-Jul-2009 14:45:55] <hackeron> mrayzenoss: very cool!
[16-Jul-2009 14:46:01] <rmatte> hmmm, found a bit of a bug
[16-Jul-2009 14:46:11] <rmatte> I dragged one of the cameras out of the frame and it just vanished
[16-Jul-2009 14:46:14] <mrayzenoss> you could probably embed your DVR images into zlinks to show current images on the status pages
[16-Jul-2009 14:46:15] <rmatte> now I can't get it back
[16-Jul-2009 14:46:31] <hackeron> rmatte: hmm?
[16-Jul-2009 14:47:04] <hackeron> can you tell me how to reproduce?
[16-Jul-2009 14:47:04] <rmatte> like, I dragged a camera in to a frame in the layout, then dragged it out a ways and let go, and it vanished
[16-Jul-2009 14:47:14] <rmatte> then the item to drag and drop on the left was greyed out
[16-Jul-2009 14:47:32] <hackeron> rmatte: can't seem to reproduce
[16-Jul-2009 14:47:38] <hackeron> rmatte: it is grayed out when you drop it in a layout
[16-Jul-2009 14:48:18] <hackeron> btw: http://hackeron.dyndns.org:82/cctv/videos/168853
[16-Jul-2009 14:48:39] <mrayzenoss> BBIAB
[16-Jul-2009 14:48:43] <rmatte> yeh, when I drop it in it goes grayed out on the left, but once it's dropped in, I drag it out a ways and it vanishes from the layout, I also can't select it again to drag it in
[16-Jul-2009 14:48:55] <hackeron> but this is a debug version, so video is a bit choppy
[16-Jul-2009 14:49:07] <rmatte> did it create that video automatically?
[16-Jul-2009 14:49:53] <rmatte> even has sound, neat
[16-Jul-2009 14:50:03] <hackeron> rmatte: yes, it created the video automatically -- it has a pretty substantial post capture (anything from 5 seconds to 300 minutes) -- but if there is no repeat motion after the postcapture (in this case 2 minutes), it deletes the 2 minutes
[16-Jul-2009 14:50:25] <hackeron> rmatte: so you get a nice short little event video you can download and email
[16-Jul-2009 14:50:33] <rmatte> very cool
[16-Jul-2009 14:50:57] <hackeron> rmatte: and for contineous motion, or motion where a person steps out of sight of the camera for a few minutes, you still don't lose the video in between (say person is talking on the phone)
[16-Jul-2009 14:51:13] <hackeron> rmatte: plus you have timelapse anyway
[16-Jul-2009 14:51:50] <rmatte> yeh
[16-Jul-2009 14:52:18] <hackeron> rmatte: yeah, so point is only thing here that takes ram is the precapture (uncompressed video) and encoder processes for motion estimation - the python bits take maybe 10MB of ram altogether
[16-Jul-2009 14:52:30] <rmatte> you have a cat or something out back?
[16-Jul-2009 14:52:42] <hackeron> rmatte: heh, probably - he's ill at the moment, should really be indoors
[16-Jul-2009 14:53:35] <hackeron> where'd you see him?
[16-Jul-2009 14:53:39] <rmatte> I saw a white cat, then there was a black one in your yard a while back
[16-Jul-2009 14:53:49] <hackeron> mine's the white one
[16-Jul-2009 14:53:59] <rmatte> he was on one of the in-progress recordings
[16-Jul-2009 14:54:01] <rmatte> he's gone now
[16-Jul-2009 14:54:10] <rmatte> rear door
[16-Jul-2009 14:54:47] <hackeron> I'll go look, lol
[16-Jul-2009 14:56:52] <hackeron> got him
[16-Jul-2009 14:57:17] <rmatte> you know, you might be able to rig it up where you have Zenoss running on a box at home but have the actual collector on the VPS, that might help
[16-Jul-2009 14:57:30] <rmatte> just a thought
[16-Jul-2009 14:58:43] <rmatte> it's nice how the system only records events of interest and timestamps them then provides video clips
[16-Jul-2009 14:58:53] <rmatte> it's the ultimate surveillance system
[16-Jul-2009 14:59:03] <hackeron> there are a lot of really neat features on the system
[16-Jul-2009 14:59:33] <hackeron> some of them are noted here: http://fusetech.co.uk
[16-Jul-2009 15:00:05] <hackeron> it's basically the zenoss of physical security and survaillance
[16-Jul-2009 15:00:10] <hackeron> (minus the ram use)
[16-Jul-2009 15:00:44] <hackeron> but that's the thing, on the dev box I have very little ram to test, but in production I put loads of ram so the entier database is in ram -- I don't want to waste that ram needlessly
[16-Jul-2009 15:01:36] <rmatte> you could always just contract the monitoring out to us
[16-Jul-2009 15:01:41] <rmatte> muahaha
[16-Jul-2009 15:01:57] <hackeron> heh, I probably will when I can afford it
[16-Jul-2009 15:02:08] <hackeron> at the moment I have a pretty neat product but no clients
[16-Jul-2009 15:02:14] <rmatte> yeh
[16-Jul-2009 15:02:18] <rmatte> it is pretty neat
[16-Jul-2009 15:02:31] <rmatte> you should go after businesses like restaurants and stuff
[16-Jul-2009 15:02:40] <rmatte> is the camera equipment quite expensive?
[16-Jul-2009 15:03:25] <hackeron> heh, well, what's really nice is highly optimised h264 encoding (thank you #x264!) - but that means putting in core i7 920 minimum to handle 16 analog cameras
[16-Jul-2009 15:03:35] <hackeron> and only around 12 ip cameras
[16-Jul-2009 15:03:57] <hackeron> so the restaurants can use their existing cameras, but for me to make any kind of profit, the DVR ends up quite expensive
[16-Jul-2009 15:04:10] <rmatte> ah
[16-Jul-2009 15:04:36] <hackeron> but the compression is quite extreme - we're talking 3 months of contineous video
[16-Jul-2009 15:04:50] <hackeron> some installations I did keep a whole year from 8 cameras on 750gb
[16-Jul-2009 15:05:02] <rmatte> crazy
[16-Jul-2009 15:05:08] <hackeron> http://leom.homeip.net/cctv/events - user/pass: demo/demo
[16-Jul-2009 15:05:25] <hackeron> here only 3 cameras are visible for the demo, but 8 are recoded with almost conteneous motion
[16-Jul-2009 15:05:32] <hackeron> check out how far back you can jump in the calendar
[16-Jul-2009 15:06:02] <hackeron> and there an actual door sensor is used to trigger cameras, so you get a short video every time someone opens the door
[16-Jul-2009 15:06:59] <rmatte> scrolling is messed up for your app in ie7
[16-Jul-2009 15:07:24] <hackeron> yeah, I know -- broke it quite recently, not sure how
[16-Jul-2009 15:07:27] <hackeron> need to fix
[16-Jul-2009 15:07:41] <rmatte> ah
[16-Jul-2009 15:07:43] <hackeron> safari and firefox are fine though
[16-Jul-2009 15:08:19] <rmatte> it's night time there right now?
[16-Jul-2009 15:08:23] <rmatte> oh right
[16-Jul-2009 15:08:24] <rmatte> UK
[16-Jul-2009 15:08:52] <hackeron> it's 9pm
[16-Jul-2009 15:08:53] <rmatte> actually, you still have sunlight, I guess the cameras on that other demo are located inside and the lights are turned off
[16-Jul-2009 15:09:06] <hackeron> well, it just got darker
[16-Jul-2009 15:09:12] <hackeron> in the last 5 minutes or so
[16-Jul-2009 15:09:26] <rmatte> yeh, it generally just starts getting dark around 9pm here too
[16-Jul-2009 15:09:32] <rmatte> it's only 4:00pm right now
[16-Jul-2009 15:10:12] <hackeron> ok, added the demo DVR to the openvpn -- so in 12 hours time, it should auto appear in zenoss?
[16-Jul-2009 15:10:38] <rmatte> no, you have to add it
[16-Jul-2009 15:10:42] <rmatte> or do an autodiscovery
[16-Jul-2009 15:11:03] <hackeron> ah, so it doesn't autodiscovery automatically?
[16-Jul-2009 15:11:07] <rmatte> zenoss just automatically remodels every 12 hours (which I disable since it's annoying)
[16-Jul-2009 15:11:15] <hackeron> ah, ok, I see
[16-Jul-2009 15:11:35] <rmatte> nah, it doesn't do it automatically, since in larger environments you don't want it picking up every single network device on the network
[16-Jul-2009 15:11:51] <rmatte> if we did an autodiscovery on our largest client we'd pick up thousands of workstations and printers
[16-Jul-2009 15:11:54] <rocket> hackeron: you should add a bes/blackberry component to your product
[16-Jul-2009 15:12:45] <hackeron> rocket: what do you mean?
[16-Jul-2009 15:13:09] <rocket> eg like the iphone thing .. add a blackberry client
[16-Jul-2009 15:13:26] <hackeron> ah - I will eventually -- now working on PTZ support
[16-Jul-2009 15:13:32] <rocket> eg if you have the box installed inside a corporation and that corporation has a bes
[16-Jul-2009 15:13:52] <rocket> then blackberry users can access the info without needing holes in a firewall or dmz etc
[16-Jul-2009 15:13:57] <hackeron> well, can always email events with a link back to the sytem that will display a blackberry interface
[16-Jul-2009 15:14:16] <hackeron> ah, I see what you mean
[16-Jul-2009 15:24:13] <rmatte> well, I'm off to a poker tournament, later gents
[16-Jul-2009 15:27:44] <schwartz> mrayzenoss: did you find Chet?
[16-Jul-2009 15:28:24] <hackeron> rmatte: have fun
[16-Jul-2009 15:28:29] <hackeron> rmatte: don't lose your house
[16-Jul-2009 15:35:43] <gemineye> Are there detailed instructions on using SNPP with Zenoss other than the few paragraphs in the Admin docs?
[16-Jul-2009 15:41:00] <schwartz> gemineye: I haven't seen any, though once I figured out that it didn't actually work with my phone, I gave up on it
[16-Jul-2009 15:43:13] mrayzenoss1 is now known as mrayzenoss
[16-Jul-2009 15:44:04] <schwartz> mrayzenoss: not sure if you saw my last one, but did you find Chet on how to get zenoss to use/see an rrd file it's not actively populating from a datasource within zenoss/
[16-Jul-2009 15:45:52] <mrayzenoss> schwartz: I didn't see it, but what I got back from Chet was that a Command data source would work
[16-Jul-2009 15:46:38] <mrayzenoss> I'm not sure how to get an RRD file from a cron job into a perf graph, but I do know how to use a Command data source to do it
[16-Jul-2009 15:46:56] <straterra> Weird
[16-Jul-2009 15:47:20] <straterra> Do I have to add a Windows server using SNMP first, then later use WMI? I don't see a WMI option when I go to add the server..
[16-Jul-2009 15:47:39] <schwartz> trying to think through how that would work..
[16-Jul-2009 15:50:55] <cgibbons> in a regular core install, WMI is only used for Event Log and Windows Services monitoring; modeling & performance data collection is done via SNMP.
[16-Jul-2009 15:51:37] <cgibbons> and there the WMI stuff gets enabled when you fill in the two zProperties for the username & password for the devices or device classes
[16-Jul-2009 15:51:49] <cgibbons> enterprise, and some other community zenpacks, do some performance monitoring via WMI
[16-Jul-2009 15:52:15] <cgibbons> (WMI is being overloaded in that context, since enterprise really does it via perfmon over MS-RPC)
[16-Jul-2009 15:54:59] <straterra> Hmm
[16-Jul-2009 15:55:21] <straterra> I installed this http://blog.dastrup.com/?p=13
[16-Jul-2009 15:55:32] <straterra> I'll use that and SNMP..that's fine.
[16-Jul-2009 15:55:50] <straterra> I looked at the cost of Enterprise today..
[16-Jul-2009 15:56:02] <cgibbons> yeah, that one gathers the performance metrics via a few WMI calls. it has a few quirks and security problems IMO, but it does work
[16-Jul-2009 15:56:03] <straterra> If I read that right, its a minimum of 250 machines..at like $100 a machine?
[16-Jul-2009 15:56:45] <straterra> So..minimum $25,000?
[16-Jul-2009 15:57:44] <cgibbons> beats me
[16-Jul-2009 15:58:27] <straterra> hehe
[16-Jul-2009 15:58:31] <straterra> Maybe I'll call a sales guy
[16-Jul-2009 15:58:48] <mrayzenoss> cgibbons is a developer, I'm the community manager, we're not sales guys
[16-Jul-2009 15:58:58] <straterra> That makes sense
[16-Jul-2009 15:59:04] <mrayzenoss> I can answer basic questions about Sales
[16-Jul-2009 15:59:11] <mrayzenoss> yes, list price is $100
[16-Jul-2009 15:59:12] <cgibbons> if you drop matt's name though the sales guy will add a 50% premium so be careful
[16-Jul-2009 15:59:16] <mrayzenoss> lol
[16-Jul-2009 15:59:17] <straterra> lol
[16-Jul-2009 15:59:29] <straterra> A starting price of $25,000 is a bit much for me, thats all
[16-Jul-2009 15:59:34] <straterra> So I'm fine with the community edition
[16-Jul-2009 15:59:46] <mrayzenoss> yeah, they raised their minimum install size from 100 to 250 devices
[16-Jul-2009 15:59:49] <straterra> For some reason, that cost is ok for vmware and zimbra..but not zenoss
[16-Jul-2009 16:00:12] <mrayzenoss> there is discounted pricing for .EDUs
[16-Jul-2009 16:00:25] <straterra> I wish we were
[16-Jul-2009 16:00:47] <straterra> Lots of discounts
[16-Jul-2009 16:01:09] <mrayzenoss> and being "Enterprise" software, you negotiate with the Sales guy to get things like training and discounted pricing thrown in
[16-Jul-2009 16:03:25] <straterra> Yeah..we don't need training, persay
[16-Jul-2009 16:03:31] <straterra> Or really any of the extra stuff..just WMI
[16-Jul-2009 16:03:47] <mrayzenoss> talk to rmatte, he wants to write a new WMI ZenPack
[16-Jul-2009 16:04:00] <straterra> The only phone support I've ever used was for Yosemite Backup..which has AMAZING support, btw
[16-Jul-2009 16:04:03] <schwartz> I wouldn't have minded some training before I dove head first into zenoss
[16-Jul-2009 16:04:07] <straterra> You talk to someone who actually knows their software
[16-Jul-2009 16:04:08] <cgibbons> is that one you've got installed not working for you?
[16-Jul-2009 16:04:20] <straterra> cgibbons: Haven't tried it yet, tbh
[16-Jul-2009 16:04:20] <cgibbons> (the WMI zenpack)
[16-Jul-2009 16:04:24] <mrayzenoss> our support guys are pretty sharp
[16-Jul-2009 16:04:33] <straterra> I actually haven't started moving things over from nagios yet
[16-Jul-2009 16:04:53] <mrayzenoss> and we offered some free training at SCALE and there will be more at LISA
[16-Jul-2009 16:04:55] <cgibbons> our support guys rock, most of them actually used the product before being hired
[16-Jul-2009 16:05:00] <straterra> I just know that when a Windows machine gets heavily loaded, it doesn't respond to SNMP..so..I get pages that a machine is 'down' when its not (from nagios). I just assumed it would be similar.
[16-Jul-2009 16:05:12] <mrayzenoss> yeah, they were hired from the Community
[16-Jul-2009 16:05:23] <cgibbons> if you do use it, i suggest making sure your zWinUser zProperties use the forward slash syntax, not the conventional backslash, i.e. ZENOSS-DEV/MyUser or ./MyUser
[16-Jul-2009 16:05:38] <straterra> Why's that?
[16-Jul-2009 16:05:49] <cgibbons> that zenpack spawns command-line wmic calls but doesn't quote the username field and the shell happily eats the backslash
[16-Jul-2009 16:05:58] <straterra> guh
[16-Jul-2009 16:06:18] <straterra> Should I use /\ then? Or is / sufficient?
[16-Jul-2009 16:06:25] <cgibbons> just / - windows is happy with that
[16-Jul-2009 16:06:41] <straterra> Fair enough
[16-Jul-2009 16:06:49] <straterra> I don't think I even explicitely put a domain
[16-Jul-2009 16:06:54] <cgibbons> would just also be a matter of editing the zenpack to do wmic -U '${here/zWinUser}'
[16-Jul-2009 16:07:40] <straterra> The support here has been phenominal
[16-Jul-2009 16:10:07] <mrayzenoss> I tend to think we've got a great community, but I'm biased of course
[16-Jul-2009 16:17:52] <straterra> I do notice that zenoss is a bit of a memory hog..anyway to tune that?
[16-Jul-2009 16:18:23] <mrayzenoss> you could turn off some of the daemons
[16-Jul-2009 16:20:18] <mrayzenoss> I'm not aware of any efforts to really reduce memory usage, or flags for that. I wonder if there are Python flags for that?
[16-Jul-2009 16:20:38] <mrayzenoss> cgibbons: Any word on how the new Zope behaves?
[16-Jul-2009 16:21:21] <cgibbons> haven't really looked that closely. it never seems like it uses much memory to me...
[16-Jul-2009 16:24:04] <schwartz> Man, this rrdtool trending is fecking rocket science. Anyone learned this particular rocket yet?
[16-Jul-2009 16:25:24] <mrayzenoss> The docs are pretty thorough http://oss.oetiker.ch/rrdtool/  What are you looking for?
[16-Jul-2009 16:25:50] <schwartz> An understanding of the algorithms so I can bend it to my superior will.
[16-Jul-2009 16:26:10] <mrayzenoss> heh
[16-Jul-2009 16:26:28] <schwartz> The docs give the technical details, but I haven't found any good .. explanation of how it works.
[16-Jul-2009 16:26:41] <cgibbons> it sure does seem like it could be explained easier
[16-Jul-2009 16:26:44] <mrayzenoss> like the syntax or the concept?
[16-Jul-2009 16:27:28] <schwartz> concept, alpha beta and gamma values. if the number of rows for the different RRA's trend forward, or if it's limited to the 'season'
[16-Jul-2009 16:28:04] <mrayzenoss> yeah, I had to read up quite a bit when I started playing with the Predictive Thresholds ZenPack
[16-Jul-2009 16:28:56] <schwartz> at least you're familiar with the brain twisting associated with it.
[16-Jul-2009 16:30:07] <mrayzenoss> if you're looking at the Default RRD Create Command under Collectors->localhost->Edit
[16-Jul-2009 16:30:23] <mrayzenoss> the 4 RRA:AVERAGE lines are how much average data is stored in the DB
[16-Jul-2009 16:30:37] <schwartz> main goal: Come up with values such that it the FAILUREs fall inline where we actually did have problems, but not too many other spots so I don't get paged every 2 hours.
[16-Jul-2009 16:30:40] <mrayzenoss> the RRA:MAX are the maximum values (obviously)
[16-Jul-2009 16:31:11] <schwartz> *nod* I'm actually doing my testing outside of zenoss at the moment, so when I stuff it into zenoss it'll be a one time deal
[16-Jul-2009 16:31:25] <schwartz> doing the testing I'm doing now with the different values would be impossible from within zenoss
[16-Jul-2009 16:31:50] <mrayzenoss> an interesting thing I learned recently was that the step of the RRD values is set by the collection interval of the monitoring daemon
[16-Jul-2009 16:32:16] <mrayzenoss> so SNMP Performance Cycle Interval of 300 means RRD step of 300 for SNMP-based RRD files
[16-Jul-2009 16:32:31] <mrayzenoss> Command data sources are 60 by default
[16-Jul-2009 16:32:39] <schwartz> huh.
[16-Jul-2009 16:32:50] <schwartz> for this I am going down to 60 seconds instead of 300.
[16-Jul-2009 16:32:58] <mrayzenoss> yeah, I kept digging through the UI trying to find the -step argument
[16-Jul-2009 16:33:37] <schwartz> though, at this point, I wonder if that's actually doing me a disservice since deviations are obviously going to be much larger in a smaller period of time, whereas if they were smoothed out over 5 minutes this may actually work better.
[16-Jul-2009 16:33:51] <schwartz> nice. I'll make a note of that.
[16-Jul-2009 16:35:17] <straterra> Is there a template or something I can use to monitor network usage and disk usage (in terms of how full the disk is) ?
[16-Jul-2009 16:39:04] <MuseIT> anyone getting zendevicedump to work on 2.4.2?
[16-Jul-2009 16:40:13] <mrayzenoss> MuseIT: what are you seeing?
[16-Jul-2009 16:40:31] <MuseIT> trackback
[16-Jul-2009 16:40:43] <MuseIT> like 229 in ExportDevice.py
[16-Jul-2009 16:40:51] <MuseIT> like = line
[16-Jul-2009 16:41:11] <mrayzenoss> straterra: like the FileSystem template?
[16-Jul-2009 16:41:26] <mrayzenoss> MuseIT: pastebin it?
[16-Jul-2009 16:41:44] <straterra> Maybe..I don't think I have that installed
[16-Jul-2009 16:41:53] <MuseIT> xlm.parsers.expat.ExpatError: not well-formed
[16-Jul-2009 16:41:54] <mrayzenoss> it's one of the defaults
[16-Jul-2009 16:42:13] <straterra> o.O
[16-Jul-2009 16:42:16] <straterra> I'm not seeing it at all
[16-Jul-2009 16:42:25] <MuseIT> sorry what is pastbin it? are you looking for me to past the error?
[16-Jul-2009 16:42:48] <mrayzenoss> MuseIT: yes please. http://pastebin.com/
[16-Jul-2009 16:43:09] <mrayzenoss> you paste your stack trace and perhaps we can debug it a bit
[16-Jul-2009 16:43:31] <straterra> I have the zenoss zenpacks rpm installed too
[16-Jul-2009 16:43:33] <mrayzenoss> straterra: if you go to your device class it should be one of the templates
[16-Jul-2009 16:43:47] <straterra> It's not
[16-Jul-2009 16:44:08] <mrayzenoss> straterra: from the table menu, Add Template
[16-Jul-2009 16:44:41] <mrayzenoss> what device class are you looking at?
[16-Jul-2009 16:44:50] <straterra> I was just under device
[16-Jul-2009 16:44:56] <straterra> Does that work for linux and windows?
[16-Jul-2009 16:45:11] <MuseIT> ok
[16-Jul-2009 16:45:19] <mrayzenoss> yeah, it's a standard SNMP
[16-Jul-2009 16:45:30] <mrayzenoss> from Net-SNMP
[16-Jul-2009 16:45:33] <straterra> Ok, its under Devices/Server
[16-Jul-2009 16:45:47] <mrayzenoss> yeah, I guess /Devices doesn't have Filesystem
[16-Jul-2009 16:46:19] <straterra> So then I go to my Windows server, go to templates, go to bind templates..and its not there
[16-Jul-2009 16:46:47] <mrayzenoss> Under /Devices/Server/Windows?
[16-Jul-2009 16:46:58] <straterra> yeah
[16-Jul-2009 16:47:18] <mrayzenoss> wth?
[16-Jul-2009 16:47:21] <MuseIT> http://pastebin.com/m59017c04
[16-Jul-2009 16:47:25] <straterra> You and me both lol
[16-Jul-2009 16:47:53] <mrayzenoss> MuseIT: are you using any non-English encodings?
[16-Jul-2009 16:48:02] <MuseIT> nope
[16-Jul-2009 16:48:09] <mrayzenoss> fudge
[16-Jul-2009 16:48:36] <straterra> And..now its there o.O
[16-Jul-2009 16:49:11] <mrayzenoss> straterra: you running on a Pentium MMX or something?
[16-Jul-2009 16:49:20] <straterra> No..but a VERY bogged down server
[16-Jul-2009 16:49:23] <straterra> :/
[16-Jul-2009 16:49:33] <mrayzenoss> straterra: using LDAP?
[16-Jul-2009 16:49:37] <straterra> Nope
[16-Jul-2009 16:49:45] <mrayzenoss> have you bumped up your zenworkers?
[16-Jul-2009 16:49:55] <MuseIT> centos 5 rpm install. tryed the zenpatch 13671 but asks me about a reversed patch.
[16-Jul-2009 16:50:03] <straterra> No. It's on a VMWare machine that has way too many vm's on it
[16-Jul-2009 16:50:07] <straterra> I'm waiting more hardware, hehe
[16-Jul-2009 16:50:15] <mrayzenoss> ahh, yeah that'll suck
[16-Jul-2009 16:50:24] <straterra> This will eventually be moved to a real machine, I think
[16-Jul-2009 16:50:35] <mrayzenoss> MuseIT: I've got nothing.  Any other weirdness?
[16-Jul-2009 16:50:53] <straterra> Sweet..so..I added that graph..and now under 'perf' for the windows server, i get this
[16-Jul-2009 16:50:55] <MuseIT> no system is working fine
[16-Jul-2009 16:50:56] <straterra> User-supplied Python expression ((here.totalBlocks * here.zFileSystemSizeOffset) * .9) for maximum value caused error: ['usedBlocks_usedBlocks']
[16-Jul-2009 16:51:09] <MuseIT> just want to get a list of the devices out.
[16-Jul-2009 16:52:32] <MuseIT> i do have a conflict database error
[16-Jul-2009 16:52:37] <mrayzenoss> and you're just using zendevicedump?
[16-Jul-2009 16:53:09] <mrayzenoss> ahh, perhaps this will help http://www.zenoss.com/Members/cluther/fix-a-broken-devicesearch-or-componentsearch-catalog/
[16-Jul-2009 16:53:16] <MuseIT> zendevicedump -o test.xml
[16-Jul-2009 16:56:39] <MuseIT> what do i do to get out of zendmd?
[16-Jul-2009 16:56:47] <mrayzenoss> Ctrl-D
[16-Jul-2009 16:56:50] <MuseIT> k
[16-Jul-2009 16:57:12] <MuseIT> do i need to restart zenoss? or just give it a try
[16-Jul-2009 16:57:39] <mrayzenoss> can't hurt to restart, but I don't think you need to
[16-Jul-2009 16:58:08] <straterra> I'm getting MISSING RRD FILE for the Perf tab..should I just wait a while?
[16-Jul-2009 16:59:02] <MuseIT> didn't work. I'm restarting zenoss.
[16-Jul-2009 16:59:19] <mrayzenoss> straterra: remodel the device then wait 5 minutes
[16-Jul-2009 16:59:25] <straterra> kk
[16-Jul-2009 17:01:15] <MuseIT> still getting the conflict and the dump is not working.
[16-Jul-2009 17:02:50] <mrayzenoss> MuseIT: my searches came up pretty empty, perhaps a forum post would help. zendevicedump works for me
[16-Jul-2009 17:03:08] <MuseIT> What are the db conflict?
[16-Jul-2009 17:03:26] <mrayzenoss> I'd have to see the error
[16-Jul-2009 17:04:51] <MuseIT> 2009-07-16T18:01:56 INFO ZODB.Conflict database conflict error (oid 0x06, class Products.TemporaryFolder.TemporaryFolder.SimpleTemporaryContainer) at /zport/RenderServer/render (2 conflicts, of which 0 were unresolved, since startup at Thu Jul 16 18:00:55 2009)
[16-Jul-2009 17:06:03] <straterra> For some reason..the Linux machine is trying to do WMI
[16-Jul-2009 17:06:18] <mrayzenoss> umm... which device class is it under?
[16-Jul-2009 17:06:32] <straterra> It was because I had stuff set in zproperties, I think
[16-Jul-2009 17:06:35] <straterra> like wmi username/pw
[16-Jul-2009 17:06:43] <mrayzenoss> MuseIT: try doing a reindex() in zendmd followed by a commit()
[16-Jul-2009 17:15:18] <MuseIT> didn't do it.
[16-Jul-2009 17:15:42] <MuseIT> still get the conflict
[16-Jul-2009 17:16:10] <mrayzenoss> I guess post to the forums with a lot of detail, perhaps someone else has seen your issue
[16-Jul-2009 17:16:39] <MuseIT> ok thanks for your help
[16-Jul-2009 17:16:55] <mrayzenoss> sorry I couldn't help more
[16-Jul-2009 17:37:39] <mrayzenoss> later all
[17-Jul-2009 00:00:46] [disconnected at Fri Jul 17 00:00:46 2009]
[17-Jul-2009 00:00:46] [connected at Fri Jul 17 00:00:46 2009]
[17-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[17-Jul-2009 00:28:09] <Diddi> hi, in my dashboard it says "Lost connection to the server.". I have no idea to what server i lost connection to as everything else seem to work properly..
[17-Jul-2009 00:28:22] <Diddi> the logs doesn't say anything about it either... so where do I start looking? (:
[17-Jul-2009 01:32:29] <Diddi> hm.. could it be that zenoss is incompatible with mysql 5.1.x ?
[17-Jul-2009 01:38:00] <Diddi> all I can get is that it's a problem with the mysql-server, but I can connect and view all the tables from the zenoss account without any problems... I run mysql on a remote computer, and everything is setup in the zenoss configs
[17-Jul-2009 01:39:26] <Diddi> the installation script created all the tables by itself... so I can't see any reason for it not to work :/
[17-Jul-2009 01:49:25] <Diddi> uhm... where does zenoss save mysql connection settings?
[17-Jul-2009 02:23:25] <Diddi> could it be related to wmi? *thinking out loud*
[17-Jul-2009 03:48:14] <Diddi> uhm, what's up with this wmi-problem? I see alot about it in the forums, and the fix with commenting out a few lines... but why isn't there a permanent fix yet?
[17-Jul-2009 03:48:33] <Diddi> the AC_REQUIRE thingy
[17-Jul-2009 04:26:16] tripitaka_ is now known as tripitaka
[17-Jul-2009 05:41:22] <RDWestSr> good moring on the east coast
[17-Jul-2009 05:41:58] <RDWestSr> anyone up that could answer a quick question about centos 5.3 x86_64 ?
[17-Jul-2009 05:42:04] <Diddi> morning to you to (:
[17-Jul-2009 05:42:24] <RDWestSr> Just woke up - Time for coffee
[17-Jul-2009 05:46:15] <RDWestSr> I have centos5-64 and I have tried setting snmp.conf as admin guide says, also as default and I still cannot get snmp working - any ideas on centos?
[17-Jul-2009 06:18:40] <TBC_Ly0n> Hello I would have a question about pricing... What is the pricing model when we have less than 250 resources to manage?
[17-Jul-2009 06:43:29] <Dieterbe> is it possible to monitor oracle with zenoss Core ?
[17-Jul-2009 07:11:18] <TBC_Ly0n> Dieterbe: You have to get the Oracle Zenpack. To get it, you need to have access to Zenoss Support portal...
[17-Jul-2009 08:47:51] <Dieterbe> TBC_Ly0n: and for that, you need to be a zenoss enterprise customer i assume?
[17-Jul-2009 08:48:40] <SilverFox> what am I doing wrong with this snmptrap statement: http://pastebin.ca/1498242
[17-Jul-2009 08:48:51] <TBC_Ly0n> Dieterbe: I think so. http://www.zenoss.com/community/docs/zenoss-extended-monitoring/2.4.2/ch40s02.html
[17-Jul-2009 08:52:31] <RDWestSr> can anyone point me to any default setups for centos5.3 x86_64 ?
[17-Jul-2009 08:52:44] <RDWestSr> the docs does not work
[17-Jul-2009 08:53:54] <RDWestSr> my first go at this, any help greatly appreciated!
[17-Jul-2009 08:53:56] <RDWestSr>
[17-Jul-2009 08:54:45] <TBC_Ly0n> http://www.zenoss.com/download/links?creg=null
[17-Jul-2009 08:54:53] <TBC_Ly0n> RDWestSr: What do you not understand?
[17-Jul-2009 08:56:14] <RDWestSr> umm, I have went by docs, installed all centos net-snmp - then tried to set community like it said but snmp connection refused
[17-Jul-2009 08:56:41] <RDWestSr> i just not stopped iptables on server - let me see if that was it
[17-Jul-2009 08:56:49] <RDWestSr> not = now*
[17-Jul-2009 09:02:43] <TBC_Ly0n> Does your net-snmp is listening on all interfaces ?
[17-Jul-2009 09:03:39] <TBC_Ly0n> On a debian server, I can configure listening interfaces in /etc/default/snmpd...
[17-Jul-2009 09:03:59] <TBC_Ly0n> It can be this variable: SNMPDOPTS
[17-Jul-2009 09:06:35] <twm1010> RDWestSr: Are you trying to setup Zenoss on the CentOS box or are you trying to monitor the CentOS box with Zenoss?
[17-Jul-2009 09:08:57] <TBC_Ly0n> For my licensing question, I asked sales team... Just have to wait right now to see if they can do an acceptable offer...
[17-Jul-2009 09:16:43] <twm1010> @mrayzenoss: morning
[17-Jul-2009 09:16:45] <TBC_Ly0n> Hi Mr Ray
[17-Jul-2009 09:17:12] <mrayzenoss> morning
[17-Jul-2009 09:17:25] <venturaville> wow sudden mental flash back to Finding Nemo
[17-Jul-2009 09:17:33] <rmatte> lol
[17-Jul-2009 09:17:58] <rmatte> Matt "Stingray" Ray
[17-Jul-2009 09:18:11] <mrayzenoss> it's like elementary school all over again
[17-Jul-2009 09:18:18] <rmatte> haha
[17-Jul-2009 09:18:35] <rmatte> could be a good name for you if you ever decide to participate in UFC
[17-Jul-2009 09:18:50] <mrayzenoss> I think Matt "Tanker" Ray would be good
[17-Jul-2009 09:18:59] <rmatte> hehe
[17-Jul-2009 09:19:17] <rmatte> sounds too much like you're an alcoholic with that one
[17-Jul-2009 09:19:31] <rmatte> "Frank the tank!"
[17-Jul-2009 09:19:51] <twm1010> why oh why must our telecom department have 5 different SNMP communities...
[17-Jul-2009 09:19:56] <twm1010> sigh.
[17-Jul-2009 09:20:16] <rmatte> twm1010: standardize
[17-Jul-2009 09:20:38] <twm1010> heh, all MY stuff IS standardized, group policy FTW!
[17-Jul-2009 09:20:48] <rmatte> hehe
[17-Jul-2009 09:21:19] <twm1010> The Cisco stuff gets confusing, all the devices are "servers" but some run windows and run either an IOS or some *nix flavor.
[17-Jul-2009 09:32:01] <rmatte> you mean they run both?
[17-Jul-2009 09:32:05] <rmatte> or one of the 3?
[17-Jul-2009 09:33:13] <rmatte> I assume you mean they run either/or
[17-Jul-2009 09:35:13] <TBC_Ly0n> I assume he has to fire some teammates
[17-Jul-2009 09:35:32] <rmatte> lol
[17-Jul-2009 09:36:11] <RDWestSr> <twm1010> I am trying to monitor one of my remote centos servers from a centos server at home with zenoss installed
[17-Jul-2009 09:42:11] <rmatte> RDWestSr: what are you having problems with?
[17-Jul-2009 09:42:13] <rmatte> the installation?
[17-Jul-2009 09:42:15] <mrayzenoss> RDWestSr: depending on which protocols you plan on using, you'll probably have to open up some ports
[17-Jul-2009 09:42:29] <rmatte> oh I see, SNMP connection refused
[17-Jul-2009 09:43:22] <rmatte> yeh, I think CenOS is pretty locked down by default
[17-Jul-2009 09:46:02] <twm1010> i found with CentOS/redhat its easier to use either the gui or tui network screens to open ports rather than mess with IPtables configs
[17-Jul-2009 09:46:07] <twm1010> open port161
[17-Jul-2009 09:49:55] <RDWestSr> i killed firewall and it did help and i got all blue events now
[17-Jul-2009 09:50:05] <twm1010> blue events are informational
[17-Jul-2009 09:50:15] <RDWestSr> i had firwall where my ip is open to everything i thought
[17-Jul-2009 09:50:37] <RDWestSr> blehh, i hate reading!
[17-Jul-2009 09:50:53] <RDWestSr> specialy when I am behind on work - lol
[17-Jul-2009 09:51:03] <rmatte> well, better learn to like it quick, you'll be doing a lot of it if you're using Zenoss
[17-Jul-2009 09:51:21] <twm1010> Yeah... mucking through Zenoss is a good way to create more work for yourself.
[17-Jul-2009 09:51:50] <RDWestSr> i'm a 1 man show with 20+ hosted accounts, design/development, and 5 seo clients - now I have to read! - blahhhh - i will get to it later - tx for the help
[17-Jul-2009 09:52:18] <rmatte> lol
[17-Jul-2009 09:52:35] <RDWestSr> hehe - l8r guys - i got to get some new quotes - cya
[17-Jul-2009 09:53:30] <rmatte> apparently hyphens are the new comma
[17-Jul-2009 09:53:35] <rmatte>
[17-Jul-2009 09:55:53] <twm1010> Hey - it works
[17-Jul-2009 09:56:05] <rmatte> Yes - it appears to
[17-Jul-2009 09:56:14] <SilverFox> I'm trying to follow the steps found on http://www.zenoss.com/community/docs/zenoss-guide/2.2.4/ch11s15.html however, I seem to get an error when trying the snmptrap command, and I'm not sure why: http://pastebin.ca/1498242
[17-Jul-2009 09:58:28] <rmatte> hmmm
[17-Jul-2009 09:59:38] <rmatte> put -O before the s
[17-Jul-2009 10:00:07] <rmatte> you might have to move it before "localhost"
[17-Jul-2009 10:01:00] <twm1010> @mrayzenoss: any chance of seeing that NetApp zenpack today?
[17-Jul-2009 10:01:47] <SilverFox> -O: Unknown Object Identifier (Sub-id not found: (top) -> -O)
[17-Jul-2009 10:01:55] <rmatte> SilverFox: actually, looking at the help list for snmptrap, the s option can be used with -O -I and -L, not sure which you should be using
[17-Jul-2009 10:02:29] <hackeron> is there a list somewhere of what each zenoss daemon does?
[17-Jul-2009 10:02:50] <SilverFox> seems to work with -I
[17-Jul-2009 10:03:42] <rmatte> SilverFox: good, past the command so that Mr. Ray can take note to have that command updated on that page
[17-Jul-2009 10:03:47] <rmatte> paste*
[17-Jul-2009 10:04:03] <twm1010> hackeron: No condensed list, but the admin guide does describe almost all of them
[17-Jul-2009 10:04:06] <twm1010> might be stuff in the wiki too
[17-Jul-2009 10:04:26] <rmatte> was it just: snmptrap -v 2c -c public localhost '' 1.3.6.1.4.1.2021.991.1.3.6.1.2.1.1.6 -I s "Device in Annapolis" ?
[17-Jul-2009 10:04:40] <hackeron> twm1010: can't find. Basically zenoss is killing me box with it's ram use - I want to disable the daemons I don't need - but can't find what each does
[17-Jul-2009 10:04:58] <hackeron> I assume zenwin is for windows servers?
[17-Jul-2009 10:05:04] <rmatte> correct
[17-Jul-2009 10:05:05] <twm1010> hackeron: i don't know that you can disable any, unless you're not having it send any alerts, perhaps you can kill the windows once
[17-Jul-2009 10:05:09] <SilverFox> snmptrap -v 2c -c public localhost '' 1.3.6.1.4.1.2021.991.1.3.6.1.2.1.1.6 -I s "Device in Annapolis"
[17-Jul-2009 10:05:41] <twm1010> I find Zenoss likes memory, ALOT of it, it will eat swap too. I would say 2GB is minimum to dedicate just to Zenoss, so I'd never do less than 3GB on a Zenoss server
[17-Jul-2009 10:05:50] <SilverFox> coincidentally, I had to run that command as root (sudo) so it wouldn't squawk about being able to read snmpapp.conf (probably something to do with my OSX config)
[17-Jul-2009 10:06:45] <hackeron> what does zeneventlog do?
[17-Jul-2009 10:06:57] <hackeron> twm1010: yeah, I only have 512mb here
[17-Jul-2009 10:07:02] <twm1010> pretty sure it checks windows server event logs over WMI
[17-Jul-2009 10:07:11] <rmatte> twm1010: correct
[17-Jul-2009 10:07:34] <twm1010> hackeron: even if you disable what you can, 512 is pretty lean
[17-Jul-2009 10:07:40] <mrayzenoss> twm1010: http://www.zenoss.com/community/projects/zenpacks/netapp
[17-Jul-2009 10:07:42] <hackeron> what about zenprocess? - the logs show it isn't pulling anything
[17-Jul-2009 10:07:44] <twm1010> you're going to need a seriously large swap file
[17-Jul-2009 10:08:05] <hackeron> twm1010: yeah, I have a 1gb one - it's all good
[17-Jul-2009 10:09:17] <rmatte> hackeron: I'd go to a 2GB swap file
[17-Jul-2009 10:10:22] <twm1010> Are you running this on a Desktop machine?
[17-Jul-2009 10:10:36] <twm1010> how many devices are you going to monitor?
[17-Jul-2009 10:10:38] <hackeron> twm1010: no, a VPS
[17-Jul-2009 10:10:49] <hackeron> twm1010: so far 10 - all linux servers
[17-Jul-2009 10:10:56] <twm1010> i think you'll pull it off then.
[17-Jul-2009 10:11:16] <hackeron> so what can other daemons can I disable if I'm only monitoring linux servers?
[17-Jul-2009 10:11:26] <hackeron> I disabled zeneventlog zenwin zenprocess and zentrap
[17-Jul-2009 10:11:52] <rmatte> hackeron: zenmodeler since you probably won't need it automatically remodeling every 12 hours
[17-Jul-2009 10:11:54] <twm1010> zenprocess is for checking proc status zentrap is probably the trap listener
[17-Jul-2009 10:12:28] <rmatte> yeh, you need zentrap
[17-Jul-2009 10:12:42] <rmatte> if you're not going to be sending syslog events to Zenoss you can disable zensyslog
[17-Jul-2009 10:12:53] <hackeron> what about zensyslog?
[17-Jul-2009 10:13:14] <twm1010> if you're not going to send syslogs from your linux boxes to zenoss, you could stop that as well
[17-Jul-2009 10:13:26] <rmatte> depends whether or not the linux boxes you're monitoring are running a syslogd, and whether or not you want to send the info to Zenoss, probably not
[17-Jul-2009 10:14:00] <hackeron> ok, thanks, disabled. Also, any configs I can change in zope to make it cache less?
[17-Jul-2009 10:19:09] <rmatte> http://www.zenoss.com/Members/fdeckert/how-to-tweak-zenoss/
[17-Jul-2009 10:19:17] <rmatte> that has some zope tuning options listed
[17-Jul-2009 10:19:24] <hackeron> thanks
[17-Jul-2009 10:19:43] <rmatte> zope.conf
[17-Jul-2009 10:19:44] <rmatte>   cache-size 60000
[17-Jul-2009 10:19:44] <rmatte>   pool-size 50
[17-Jul-2009 10:20:08] <rmatte> decrease whatever values are listed for those
[17-Jul-2009 10:20:08] <cgibbons> how many devices are you trying to monitor again, hackeron?
[17-Jul-2009 10:20:18] <hackeron> cgibbons: currently 10
[17-Jul-2009 10:20:22] <cgibbons> cool
[17-Jul-2009 10:23:42] <hackeron> right, lets see
[17-Jul-2009 10:26:36] <dollarbang> Good day all, can someone tell me which release versions of Sun Solaris that the Enterprise version supports? thanks.
[17-Jul-2009 10:27:16] * rocket submits fedora zenpack to matt ..
[17-Jul-2009 10:27:25] <mrayzenoss> dollarbang: Enterprise is not currently supported running on Solaris, but it can monitor it via SNMP or SSH
[17-Jul-2009 10:28:48] * rmatte votes for Ubuntu Zenpack next
[17-Jul-2009 10:29:24] <SilverFox> he should also suggest that the 'demonstration MIB' file be named TRAP-TEST-MIB.mib
[17-Jul-2009 10:30:10] <rmatte> SilverFox: that documentation you're using is old, 2.2.4, so it's outdated
[17-Jul-2009 10:31:45] <cgibbons> ancient history
[17-Jul-2009 10:32:05] <rocket> rmatte: what package manager does ubuntu use again?
[17-Jul-2009 10:32:16] <rmatte> rocket: aptitude, like debian
[17-Jul-2009 10:32:30] <dollarbang> I'm sorry, I may not have been clear in what I was asking. We are running a Sun box on Solaris 8. Zenoss Core is currently running under Linux. Does the Enterprise Zenpack monitor Solaris release 8.
[17-Jul-2009 10:32:35] <rmatte> Ubuntu is basically a modified version of debian
[17-Jul-2009 10:32:44] <rocket> rmatte: what is the command to show the packages installed
[17-Jul-2009 10:32:50] <rocket> or do I need to look that up?
[17-Jul-2009 10:33:00] <TBC_Ly0n> hackeron: I have the same problem than U !!! Zenoss is eating all my RAM and swap...
[17-Jul-2009 10:33:09] <rmatte> rocket: you can do a dpkg -l
[17-Jul-2009 10:33:12] <hackeron> TBC_Ly0n: I think everyone has that problem
[17-Jul-2009 10:33:15] <TBC_Ly0n> How do you disable a service definitly?
[17-Jul-2009 10:33:29] <TBC_Ly0n> Without raisins alarms
[17-Jul-2009 10:33:37] <rmatte> rocket: it'll be a damn long list though
[17-Jul-2009 10:33:53] <rocket> rmatte: does that just show installed packages ?
[17-Jul-2009 10:34:02] <TBC_Ly0n> Yes
[17-Jul-2009 10:34:09] <TBC_Ly0n> Package, version and description.
[17-Jul-2009 10:34:40] <TBC_Ly0n> (and the state of it...)
[17-Jul-2009 10:34:41] <SilverFox> rmatte: aha
[17-Jul-2009 10:34:52] <mrayzenoss> dollarbang: I believe Solaris 8 works fine via SNMP, I don't see an Enterprise ZenPack for it
[17-Jul-2009 10:34:53] <TBC_Ly0n> The command can be run with a "unprivileged" user.
[17-Jul-2009 10:35:22] <rmatte> rocket: yeh
[17-Jul-2009 10:35:30] <mrayzenoss> dollarbang: and the Enterprise SSH ZenPack for Solaris is being written right now, I'm not sure if Solaris 8 is a requirement for it
[17-Jul-2009 10:35:35] <rmatte> rocket: alternatively you can do: aptitude search '~i'
[17-Jul-2009 10:35:49] <TBC_Ly0n> Solaris 8 is not supported anymore by Sun....
[17-Jul-2009 10:35:52] <cgibbons> hasn't solaris 8 been end-of-life for a while now? yeah
[17-Jul-2009 10:35:59] <TBC_Ly0n> So, I would not spend anything on that...
[17-Jul-2009 10:36:10] <rmatte> rocket: dpkg -l is probably better though since it actually gives version info for the packages
[17-Jul-2009 10:36:11] <TBC_Ly0n> cgibbons: lol
[17-Jul-2009 10:36:12] <mrayzenoss> dollarbang: try the OpenSolaris ZenPack and see what works/doesn't and report back
[17-Jul-2009 10:36:13] <cgibbons> we've still got a bunch of users running Windows 2000 devices, too.... er...
[17-Jul-2009 10:36:26] <mrayzenoss> we had NT4 at 1 customer
[17-Jul-2009 10:36:42] <rocket> rmatte: is there a way to get install time?
[17-Jul-2009 10:36:51] <cgibbons> that's just crazy
[17-Jul-2009 10:37:06] <rmatte> rocket: hmmm, good question
[17-Jul-2009 10:37:15] <mrayzenoss> cgibbons: I believe it was a hospital
[17-Jul-2009 10:37:51] <dollarbang> mrayzenoss: Ok, once we have iron'd out what we want to be monitored. Still making a list, checking it twice, looking for.....I digress. thanks.
[17-Jul-2009 10:38:08] <rmatte> rocket: no, doesn't look like that is possible
[17-Jul-2009 10:38:16] <mrayzenoss> dollarbang: http://www.zenoss.com/community/projects/zenpacks/opensolaris
[17-Jul-2009 10:38:20] <SilverFox> 2.3.3 docs had the same thing, but 2.4 appears to have axed that subsection out. (event monitoring -> snmp traps and event transforms)
[17-Jul-2009 10:39:26] <SilverFox> Looks like the docs got a major overhaul for this release?
[17-Jul-2009 10:41:08] <mrayzenoss> SilverFox: yeah, we've been trying to make them much better, wondering if that missing section was removed for a reason
[17-Jul-2009 10:41:42] <rmatte> mrayzenoss: if it works properly it'd be good to have it updated and re-added
[17-Jul-2009 10:41:42] <SilverFox> its very helpful/important, I think. Its possible the same information is in a different section. I'll have to look more.
[17-Jul-2009 10:42:00] <mrayzenoss> SilverFox: if you think we need it back, open a ticket and we'll put it back http://dev.zenoss.com/trac/wiki/HowToAddTicket
[17-Jul-2009 10:42:13] <mrayzenoss> again, I don't know why it was removed
[17-Jul-2009 10:42:19] * rmatte nods
[17-Jul-2009 10:44:23] <mrayzenoss> hmm... it looks like the new Chapter 7 is where the section would have been
[17-Jul-2009 10:44:56] <mrayzenoss> and all the zenmail stuff is gone
[17-Jul-2009 10:45:56] <mrayzenoss> typically things got removed when they were wrong or redundant. Some of the event stuff is getting rewritten for King Crab, since there will be a new Event Console
[17-Jul-2009 10:46:32] <SilverFox> http://www.zenoss.com/Members/fdeckert/how-to-tweak-zenoss/ needs updating I think.
[17-Jul-2009 10:46:36] <rmatte> so zenmail isn't documented at all now?
[17-Jul-2009 10:46:46] <mrayzenoss> not in the admin guide, wondering where it went
[17-Jul-2009 10:48:01] <rmatte> doesn't anyone here actually understand format code like "%8.2f"?
[17-Jul-2009 10:48:06] <rmatte> does anyone*
[17-Jul-2009 10:48:25] <rmatte> I'm looking for someone to explain it to me, since I wouldn't know what to google for
[17-Jul-2009 10:48:25] <mrayzenoss> 8 digits, 2 decimal places, floating point?
[17-Jul-2009 10:48:31] <SilverFox> that looks familier.  printf?
[17-Jul-2009 10:48:33] <mrayzenoss> just a guess
[17-Jul-2009 10:49:11] <rmatte> mrayzenoss: hmmm
[17-Jul-2009 10:49:55] <rocket> f, F The double argument is rounded and converted to decimal notation in the style [-]ddd.ddd, where the number of digits after the
[17-Jul-2009 10:49:58] <rocket> decimal-point character is equal to the precision specification. If the precision is missing, it is taken as 6; if the preci-
[17-Jul-2009 10:50:02] <mrayzenoss> http://www.informit.com/articles/article.aspx?p=28790&seqNum=2
[17-Jul-2009 10:50:02] <rocket> sion is explicitly zero, no decimal-point character appears. If a decimal point appears, at least one digit appears before it.
[17-Jul-2009 10:50:12] <rocket> man 3 printf   .. if you have gcc installed
[17-Jul-2009 10:50:26] * SilverFox the noob guessed it!
[17-Jul-2009 10:51:20] <rocket> rmatte: rrdtool is using printf strings to display the info .. thats ultimately where you need to go to understand it .. however they might have a better writeup on the rrdtool website
[17-Jul-2009 10:51:47] <rmatte> rocket: k
[17-Jul-2009 10:52:46] <rmatte> rocket: here's what I'm trying to do, I have a report that displays an error count, but the guy who wrote it had it display the count as a decimal for some reason, so 21 errors shows up as 0.21, the function used in the report to display the data is <td tal:content="python: r.fmt('%8.2f', r.inerror)"/>
[17-Jul-2009 10:53:23] <rocket> have you tried dropping the .2?
[17-Jul-2009 10:53:37] <rocket> just %8f
[17-Jul-2009 10:53:38] <rmatte> rocket: wouldn't that just round the decimal value?
[17-Jul-2009 10:53:47] <rmatte> I'll try it out
[17-Jul-2009 10:54:30] <rmatte> dropping the 2 just made the decimal places go out: 0.000000
[17-Jul-2009 10:54:53] <rocket> %8d
[17-Jul-2009 10:54:55] <rmatte> I'm wondering if fmt is the wrong function to call
[17-Jul-2009 10:55:02] <rocket> d, i The int argument is converted to signed decimal notation. The precision, if any, gives the minimum number of digits that must
[17-Jul-2009 10:55:06] <rocket> appear; if the converted value requires fewer digits, it is padded on the left with zeros. The default precision is 1. When 0
[17-Jul-2009 10:55:09] <rocket>               is printed with an explicit precision 0, the output is empty.
[17-Jul-2009 10:55:23] <rocket> or %8i
[17-Jul-2009 10:55:33] <rocket> where 8 is the width of the field
[17-Jul-2009 10:56:00] <rmatte> yeh, that just ends up displaying no data at all
[17-Jul-2009 10:56:27] <rmatte> I think r.fmt is the wrong function to use, I'll have to check out the other functions
[17-Jul-2009 10:57:22] <rmatte> I don't know if I should use something like r.__getitem__
[17-Jul-2009 10:57:41] <rmatte> (if that's even possible)
[17-Jul-2009 10:58:09] <rmatte> the functions are listed in $ZENHOME/Products/ZenReports/Utils.py
[17-Jul-2009 10:59:04] <kerick> is there a way to get Zenoss to show me a list of process it sees on a box?
[17-Jul-2009 11:00:50] <mrayzenoss> opened http://dev.zenoss.org/trac/ticket/5335 and http://dev.zenoss.org/trac/ticket/5334
[17-Jul-2009 11:02:24] <SilverFox> oh, thanks mrayzenoss.
[17-Jul-2009 11:03:47] <rmatte> yeh, 8i doesn't work either, it'd take 100 errors before any of the values would show as 1
[17-Jul-2009 11:05:28] <rmatte> is fmt a general python function?
[17-Jul-2009 11:06:57] <rmatte> hmmm, apparently, just having trouble finding documentation on it
[17-Jul-2009 11:07:07] <rmatte> ah, got it
[17-Jul-2009 11:08:58] <rmatte> it actually appears to be converting it in to a percent, not sure what it's calculating the percent with
[17-Jul-2009 11:09:07] <rmatte>     def fmt(self, fmt, value):
[17-Jul-2009 11:09:07] <rmatte>         if value is None:
[17-Jul-2009 11:09:07] <rmatte>             return UNAVAILABLE
[17-Jul-2009 11:09:07] <rmatte>         return fmt % value
[17-Jul-2009 11:11:46] <SilverFox> by default, will zenoss display events for ALL traps recieved? There isn't a default filter that ignores certain informational traps (like config change on a router) ?
[17-Jul-2009 11:12:03] <rmatte> SilverFox: nope, you have to tweak all of that yourself
[17-Jul-2009 11:12:16] <cgibbons> rmatte: that's the built-in python string formatting operator. http://docs.python.org/library/stdtypes.html#string-formatting
[17-Jul-2009 11:12:18] <rmatte> they don't assume that everyone is going to want the same things blocked
[17-Jul-2009 11:12:36] <rmatte> or filtered I should say
[17-Jul-2009 11:12:46] <rmatte> cgibbons: thanks
[17-Jul-2009 11:13:31] <SilverFox> rmatte: good, that means my router is filtering it. just wanted to know where to focus my efforts.
[17-Jul-2009 11:13:40] <cgibbons> the tricky thing being if you need to pass in multiple arguments to the string you have to give it a tuple... like: "Device: %s - Event: %s" % (deviceId, eventText)
[17-Jul-2009 11:14:17] <rmatte> right
[17-Jul-2009 11:14:17] <rocket> rmatte: I wonder what the fmt arg in that is getting set to
[17-Jul-2009 11:14:26] <rocket> rmatte: might want to add some prints to see
[17-Jul-2009 11:14:36] <rocket> eg in the interpreter I did
[17-Jul-2009 11:14:42] <rocket> fmt="%8d"
[17-Jul-2009 11:14:48] <rocket>  print fmt % 21
[17-Jul-2009 11:14:54] <rocket> 21
[17-Jul-2009 11:15:45] <rmatte> yeh, I'll have to screw around with it, not even sure if I want the value run through fmt at all
[17-Jul-2009 11:15:48] <rocket> rmatte: I would bet that the value is coming in as a percent first
[17-Jul-2009 11:15:58] <rmatte> it looks like the author of the report used it for lack of a better option in Utils.py
[17-Jul-2009 11:16:21] <rmatte> rocket: the value is being taken straight from the RRD file
[17-Jul-2009 11:16:39] <rocket> are you sure its not getting munged somewhere in between?
[17-Jul-2009 11:16:45] <rocket> rmatte: which report?
[17-Jul-2009 11:18:55] <gwb2351> howdy everyone, new zenoss customer here
[17-Jul-2009 11:20:16] <rocket> gwb2351: Hello gwb2351! Thank you for joining zenoss addicts anonymous! Would you like to share your story with us .. or would you like to sit out this round?
[17-Jul-2009 11:21:12] <gwb2351> my story is simple right now... plodding through adding devices and getting SNMP to work properly
[17-Jul-2009 11:21:29] <gwb2351> did have one question though...
[17-Jul-2009 11:23:23] <gwb2351> when trying to add "ZenPacks.Nagios.CheckPing-1.0-py2.4.zip" (through web gui, following intructions at http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch13s02.html
[17-Jul-2009 11:23:45] <gwb2351> INFO:zen.HookReportLoader:loading reports from:/opt/zenoss/Products/EGG-INFO/reports ERROR: zenpack command failed. Reason: Not a registered directory: EGG-INFO/skins
[17-Jul-2009 11:24:10] <gwb2351> only matches i found in search was .zip vs. .egg (which was fixed a while back)
[17-Jul-2009 11:24:16] <rocket> change from zip to egg
[17-Jul-2009 11:24:22] <rocket> eg rename from zip to egg
[17-Jul-2009 11:24:31] <rocket> your browser renamed it on you
[17-Jul-2009 11:25:45] <gwb2351> ah yes
[17-Jul-2009 11:26:04] <gwb2351> that worked
[17-Jul-2009 11:27:38] <rocket> rmatte: I will make a zenpack for ubuntu .. however it will not have the dates packages are installed
[17-Jul-2009 11:27:52] <gwb2351> thanks!
[17-Jul-2009 11:28:19] <rocket> rmatte: I dont really feel like looking at timestamps of files to get that info especially when the time stamps could be modified and its no longer accurate
[17-Jul-2009 11:28:41] <rocket> rmatte: I might have something later today
[17-Jul-2009 11:56:16] pgega is now known as CTO
[17-Jul-2009 12:21:59] <mrayzenoss> New Community NetApp ZenPack: http://www.zenoss.com/community/projects/zenpacks/netapp
[17-Jul-2009 12:26:17] <TBC_Ly0n> rocket: If you can create your zenpack to be fully-compatible with pure debian, it would be perfect
[17-Jul-2009 12:26:26] <TBC_Ly0n> Ubuntu are based on Debian, so it should be easy.
[17-Jul-2009 13:07:11] <rmatte> rocket: cool
[17-Jul-2009 14:19:23] <twm1010> Anyone have a link to a eye-friendly article on configuring a linux host to send syslog information to a syslog server?
[17-Jul-2009 14:23:43] <HKhan> twm1010 if its sysklogd running on the linux host, *.*   @syslog_server_address
[17-Jul-2009 14:48:25] <twm1010> thank you
[17-Jul-2009 14:48:39] <twm1010> so you send everything, let the other host sort out the priority?
[17-Jul-2009 15:04:35] <rmatte> twm1010: yes
[17-Jul-2009 15:04:53] <rmatte> twm1010: Zenoss will ignore lower priority syslogs by default and only alert on the higher severity
[17-Jul-2009 15:05:03] <rmatte> and you can tune out specific events with transforms as well
[17-Jul-2009 15:06:24] <straterra> Ok..I see the FileSystem template in /Devices/Server, but not in /Devices/Server/Windows
[17-Jul-2009 15:06:37] <straterra> When I go to add bindings for Windows, FileSystem is NOT in the list of templates
[17-Jul-2009 15:06:43] <rmatte> straterra: you have to click on one of the drives then go to templates
[17-Jul-2009 15:06:52] <rmatte> the template is a component of the drives, not of the device
[17-Jul-2009 15:07:06] <rmatte> same thing with interface templates, you need to click on an interface then go to templates to see it
[17-Jul-2009 15:07:11] <straterra> Where do I go to one of the drives?
[17-Jul-2009 15:07:17] <rmatte> on the OS tab
[17-Jul-2009 15:07:27] <rmatte> provided the device is properly modelled, you should see the hard drives
[17-Jul-2009 15:07:37] <straterra> Let me try that
[17-Jul-2009 15:07:39] <rmatte> k
[17-Jul-2009 15:07:52] <straterra> Ok, I see the drives
[17-Jul-2009 15:08:22] <rmatte> click on one, then go to templates
[17-Jul-2009 15:08:27] <straterra> So then..I click on C:\..and see the drive usage
[17-Jul-2009 15:08:36] <rmatte> correct
[17-Jul-2009 15:08:39] <straterra> Nice..can I add that graph to perf?
[17-Jul-2009 15:08:47] <rmatte> no
[17-Jul-2009 15:09:03] <rmatte> it's a filesystem template, so it's bound to that item
[17-Jul-2009 15:09:09] <straterra> Ok..thats fine
[17-Jul-2009 15:09:14] <straterra> It's just training to view them then
[17-Jul-2009 15:09:24] <rmatte> but if you click on the template tab from there, you'll see the filesystem template
[17-Jul-2009 15:09:26] <straterra> And the network graphs are under the nics then..right?
[17-Jul-2009 15:09:31] <rmatte> correct
[17-Jul-2009 15:09:33] <straterra> Right
[17-Jul-2009 15:10:27] <straterra> One other thing..I see 6 "grey" items under this server..yet I go to Events, and nothing is there
[17-Jul-2009 15:10:43] <rmatte> see the dropdown that says "Sev"?
[17-Jul-2009 15:10:51] <rmatte> change it from Info to debug
[17-Jul-2009 15:11:03] <rmatte> grey events are debug messages
[17-Jul-2009 15:11:18] <straterra> Aaaah, ok
[17-Jul-2009 15:11:26] <straterra> Error reading value for "cpuPercentProcessorTime" on mifs1.indy.marian.local (oid .1.3.6.1.4.1.9600.1.1.5.1.5.6.95.84.111.116.97
[17-Jul-2009 15:11:31] <straterra> and other ones for memory and cpu
[17-Jul-2009 15:11:41] <straterra> Oh..those are old
[17-Jul-2009 15:11:43] <straterra> nevermind
[17-Jul-2009 15:11:59] <straterra> This is more intuative that I thought
[17-Jul-2009 15:12:46] <straterra> Hmm..but under perf, only one graph shows up..and it says MISSING RRD FILE
[17-Jul-2009 15:12:53] <straterra> the rest of the graphs there don't even show
[17-Jul-2009 15:13:19] <rmatte> those are probably left over from when you were modeling
[17-Jul-2009 15:13:29] <rmatte> oh
[17-Jul-2009 15:13:41] <rmatte> sounds like your SNMP permissions might be set wrong
[17-Jul-2009 15:13:52] <straterra> On the server side?
[17-Jul-2009 15:13:57] <rmatte> yeh
[17-Jul-2009 15:14:01] <straterra> It's set to...hold on
[17-Jul-2009 15:14:02] <rmatte> try to snmpwalk .1.3.6.1.4.1.9600.1.1.5.1.5.6.95.84.111.116.97
[17-Jul-2009 15:14:18] <straterra> read create
[17-Jul-2009 15:14:22] <rmatte> actually wait, I know why
[17-Jul-2009 15:14:38] <rmatte> you installed that snmp informant zenpack I assume?
[17-Jul-2009 15:14:44] <straterra> Let me check
[17-Jul-2009 15:14:47] <straterra> I installed just the core ones
[17-Jul-2009 15:15:07] <straterra> Core ones plus a wmi one
[17-Jul-2009 15:15:26] <rmatte> then where are those graphs being applied from?
[17-Jul-2009 15:15:45] <rmatte> I don't believe Zenoss comes with predefines graphs for Windows via SNMP
[17-Jul-2009 15:15:50] <rmatte> predefined*
[17-Jul-2009 15:16:36] <straterra> The template is Device
[17-Jul-2009 15:16:37] <rmatte> I had to design a monitoring system to monitor CPU and memory performance on windows servers using the base SNMP, but it's a bit complex right now, I need to convert it in to an actual collector plugin, and wrap it up in a ZenPack, I just don't have time for it right now
[17-Jul-2009 15:16:40] <straterra> "Windows template that requires Informant MIB"
[17-Jul-2009 15:16:46] <rmatte> yeh, exactly
[17-Jul-2009 15:16:56] <rmatte> you need to install snmp informant on your server.
[17-Jul-2009 15:17:10] <rmatte> http://www.snmp-informant.com/
[17-Jul-2009 15:17:10] <straterra> Really? :/
[17-Jul-2009 15:17:35] <straterra> Is it its own snmp server?
[17-Jul-2009 15:17:58] <rmatte> yeh, that's why I had to do create scripts to collect the performance data from the standard microsoft SNMP agent
[17-Jul-2009 15:18:15] <rmatte> it's an add-on to the local snmp server I believe, but I'm not sure
[17-Jul-2009 15:18:17] <rmatte> never used it
[17-Jul-2009 15:18:33] <straterra> Oh..but..I install that and I should get performance graphs?
[17-Jul-2009 15:18:38] <rmatte> correct
[17-Jul-2009 15:18:49] <rmatte> the OIDs it's trying to poll are snmp-informant specific
[17-Jul-2009 15:18:58] <straterra> Oh, gotcha
[17-Jul-2009 15:19:31] <rmatte> we work off the values from the default windows SNMP agent, but our solution isn't pretty, it works though
[17-Jul-2009 15:20:46] <straterra> I hope its in MSI format..It'd be easy to push to the servers
[17-Jul-2009 15:20:57] <straterra> aaaaand it isnt
[17-Jul-2009 15:20:58] <rmatte> the planning for the microsoft SNMP agent was garbage, the OID for RAM changes dynamically based on how many physical drives are in the system, so it's different from server to server, then the CPU OIDs tend to wander, so they aren't always the same
[17-Jul-2009 15:21:33] <rmatte> so I've had to create scripts to hunt down the OID on each server and then collect the data from it
[17-Jul-2009 15:24:30] <straterra> That's lame
[17-Jul-2009 15:24:36] <rmatte> very
[17-Jul-2009 15:25:04] <straterra> I notice you can configure 'paging', as opposed to email
[17-Jul-2009 15:25:10] <straterra> Does that require some kind of physical modem?
[17-Jul-2009 15:25:12] <rmatte> yeh
[17-Jul-2009 15:25:16] <rmatte> no
[17-Jul-2009 15:25:35] <straterra> Right now, we send emails to the cell providors sms gateway
[17-Jul-2009 15:25:36] <rmatte> I think you need to send it to the email address that corresponds with the pager
[17-Jul-2009 15:25:44] <rmatte> and it just changes the way it formats the messages when they are sent
[17-Jul-2009 15:26:06] <rmatte> makes it briefer to better suit mobile devices I believe
[17-Jul-2009 15:26:09] <straterra> Nifty
[17-Jul-2009 15:28:13] FooDog is now known as chimes
[17-Jul-2009 15:28:42] chimes is now known as chord
[17-Jul-2009 15:29:07] chord is now known as ding_
[17-Jul-2009 15:29:13] ding_ is now known as notify
[17-Jul-2009 15:29:20] <twm1010> rmatte: i've always wondered about that
[17-Jul-2009 15:29:33] <twm1010> thats a great explanation
[17-Jul-2009 15:29:40] <straterra> Ok..its installed..should I remodel?
[17-Jul-2009 15:29:43] notify is now known as FooDog
[17-Jul-2009 15:30:03] <twm1010> straterra: yeah
[17-Jul-2009 15:30:47] <straterra> So now..I wait a few, and the graphs should work/appear in perf then
[17-Jul-2009 15:33:00] <twm1010> yeah, let it do a few collection cycles, wait about 6-10 minutes
[17-Jul-2009 15:33:18] <twm1010> filesystem stuff should start populating on the O/S tab too
[17-Jul-2009 15:42:42] <straterra> rmatte: I still get this as a debug message.. Error reading value for "lDiskDiskReadBytesPerSec" on mifs1.indy.marian.local (oid .1.3.6.1.4.1.9600.1.1.1.1.15 is bad)
[17-Jul-2009 15:42:49] <straterra> With informant installed
[17-Jul-2009 15:44:41] <twm1010> You put in /Server/Windows device class right?
[17-Jul-2009 15:45:02] <twm1010> Are you sure you're still getting that message and its not old?
[17-Jul-2009 15:45:34] <straterra> Doh, it is old
[17-Jul-2009 15:45:38] <straterra> I just cleared them all though o.O
[17-Jul-2009 15:45:54] <straterra> Unless the system time is off
[17-Jul-2009 15:46:03] <straterra> It is
[17-Jul-2009 15:46:44] <straterra> So..yeah, its not old
[17-Jul-2009 15:47:22] <straterra> And its right..the IO Bytes graph under perf is nonexistent
[17-Jul-2009 15:47:52] <rmatte> you may need to push changes to the collector
[17-Jul-2009 15:48:06] <straterra> How do I do that? o.O
[17-Jul-2009 15:48:06] <rmatte> go to the device page, access the dropdown menu, and select Manage -> Push Changes
[17-Jul-2009 15:48:20] <rmatte> the collector only updates it's configuration from Zenoss once every hour
[17-Jul-2009 15:48:26] <rmatte> but you can force it to update by doing that
[17-Jul-2009 15:48:32] <straterra> Aaaaah, ok
[17-Jul-2009 15:48:44] <twm1010> rmatte: so thats why when you delete a device you still get events for a short time after..
[17-Jul-2009 15:48:55] <rmatte> yeh
[17-Jul-2009 15:50:38] <straterra> so..is there any difference between the hardware and os views of my hard drives?
[17-Jul-2009 15:54:06] <straterra> I did the push..and the debug messages keep coming back
[17-Jul-2009 15:54:54] <straterra> I'll just delete that graph..its graphed under hardware correctly
[17-Jul-2009 15:58:22] <straterra> oops..I didn't mean to delete that template :/
[17-Jul-2009 15:58:36] <straterra> Oh well
[17-Jul-2009 15:59:07] <rmatte> lol
[17-Jul-2009 15:59:13] <rmatte> you deleted the whole template?
[17-Jul-2009 15:59:38] <straterra> Yeah o.O
[17-Jul-2009 15:59:42] <straterra> I didn't mean to do that
[17-Jul-2009 16:00:22] <straterra> I..don't think I need it though
[17-Jul-2009 16:00:27] <straterra> It was the harddisk template
[17-Jul-2009 16:02:17] <straterra> I do notice that damn near nothing is being collected for this linux mahcine
[17-Jul-2009 16:02:18] <straterra> machine^
[17-Jul-2009 16:02:34] <straterra> No files systems, no routes, etc
[17-Jul-2009 16:05:54] <twm1010> fyi, don't need todelete templates, you just unbind them
[17-Jul-2009 16:07:16] <straterra> Yeah :/
[17-Jul-2009 16:07:41] <rmatte> the filesystem template you mean?
[17-Jul-2009 16:07:51] <rmatte> you won't have disk threshold monitoring without it, so it's fairly important
[17-Jul-2009 16:07:51] <straterra> No, HardDisk was the name of it
[17-Jul-2009 16:07:58] <rmatte> ah
[17-Jul-2009 16:08:05] <straterra> I still have filesystem
[17-Jul-2009 16:08:10] <rmatte> must have just been for disk IO monitoring then
[17-Jul-2009 16:08:15] <rmatte> so it shouldn't matter
[17-Jul-2009 16:08:58] <straterra> Right
[17-Jul-2009 16:09:41] <straterra> I'm getting TONS of messages like this.. Error reading value for "lo" on mims1.indy.marian.local (oid .1.3.6.1.2.1.2.2.1.17.1 is bad)
[17-Jul-2009 16:09:48] <straterra> And thats a Linux box running snmpd
[17-Jul-2009 16:10:07] <rmatte> looks like it's complaining about not being able to monitor the local interface
[17-Jul-2009 16:10:16] <straterra> That's not the only one, though
[17-Jul-2009 16:10:24] <rmatte> if you click on that interface you should see a interface type listed
[17-Jul-2009 16:10:31] <rmatte> it'll probably be called something like loopback
[17-Jul-2009 16:10:50] <rmatte> all you have to do is go to the /Devices section, go in to templates, and you'll see the ethernetS<whatever> template
[17-Jul-2009 16:11:02] <rmatte> just copy that template, then go in to it, and rename it to exactly what the type shows up as
[17-Jul-2009 16:11:11] <rmatte> then go in to it and delete all the datapoints and graphs
[17-Jul-2009 16:11:51] <rmatte> Zenoss determines what interface template to use per interface based on which matches with the type of interface (based on it's name)
[17-Jul-2009 16:11:57] <straterra> What about like..error reading value memcached..and eth0..and such
[17-Jul-2009 16:11:59] <rmatte> so interfaces don't support the same OIDs as others
[17-Jul-2009 16:12:18] <rmatte> you shouldn't be getting any errors for eth0, if you are then it's a permissions issues on the Linux box
[17-Jul-2009 16:12:22] <straterra> It's like it can't read much of anything
[17-Jul-2009 16:12:29] <rmatte> you need to give that snmp community string all permissions
[17-Jul-2009 16:12:34] <straterra> I did o.O
[17-Jul-2009 16:13:00] <rmatte> I'm monitoring many Linux boxes and I don't have any issues like that, make sure the snmp community string is set properly in Zenoss for that device
[17-Jul-2009 16:13:07] <rmatte> go to the device's zProperties and check
[17-Jul-2009 16:13:26] <rmatte> I need to head home, I'll be back on Monday
[17-Jul-2009 16:13:27] <rmatte> cheers.
[17-Jul-2009 16:13:54] <straterra> thanks for the help
[17-Jul-2009 16:14:05] <straterra> Have a good weekend
[17-Jul-2009 16:14:40] <kerick> straterra: make sure iptables isn't running, that one catches me sometimes
[17-Jul-2009 16:16:37] <straterra> [root@mims1 ~]# snmpwalk -v 2c -c marian 10.1.1.51
[17-Jul-2009 16:16:38] <straterra> SNMPv2-SMI::mib-2 = No more variables left in this MIB View (It is past the end of the MIB tree)
[17-Jul-2009 16:16:42] <straterra> weeeeeird
[17-Jul-2009 16:21:35] <straterra> bad snmpd config
[17-Jul-2009 19:00:57] <SilverFo2> ok, so I've verified that a trap is reaching the box by using ngrep, but zenoss isn't listing it in event console. Any ideas why not?
[17-Jul-2009 19:55:36] SilverFo2 is now known as SilverFox
[17-Jul-2009 20:45:24] <straterra> Is there any way to get the HardDisk template back?
[17-Jul-2009 23:36:02] <ToKyNET> anyone awake?
[17-Jul-2009 23:43:15] <SilverFox> barely...
[18-Jul-2009 00:00:46] [disconnected at Sat Jul 18 00:00:46 2009]
[18-Jul-2009 00:00:46] [connected at Sat Jul 18 00:00:46 2009]
[18-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[18-Jul-2009 00:51:08] ToKyNET is now known as ToKyNET-afk
[18-Jul-2009 07:01:59] <Diddi> hi! When I go to my dashboard, it says "Lost connection to the server." but I can't figure out what's wrong.. Everything else seem to work probperly, and I can access mysql database with the configured account
[18-Jul-2009 07:02:17] <Diddi> zenoss 2.4.2 tar-ball install without wmi (if that matters)
[18-Jul-2009 07:03:34] <Diddi> I don't even know where to start look
[18-Jul-2009 07:07:24] <Diddi> any hint? (:
[18-Jul-2009 08:16:10] <int> hello! i wonder if anyone looked on ZenPacks.example.Techniques ?
[18-Jul-2009 08:21:09] <int> it adds menu item with action="webServerDeviceDetail", but it seems that webServerDeviceDetail is not declared
[18-Jul-2009 08:59:54] <int> can i somehow bind template to component from python shell?
[18-Jul-2009 09:06:42] <int> ok, found http://www.zenoss.com/Members/cluther/how-zenoss-chooses-templates
[18-Jul-2009 09:25:06] <int> how can i use zProperty value in COMMAND template?
[18-Jul-2009 11:46:28] <int> anyone knows how to reinstall zenpack?
[18-Jul-2009 12:29:47] <int> or maybe howto remove component from device?
[18-Jul-2009 16:45:21] <otakup0pe> so why is it that when i try to install zenoss it hangs at the very end
[18-Jul-2009 16:45:25] <otakup0pe> i've tried three different methods.
[18-Jul-2009 16:45:31] <otakup0pe> the vmware appliance is sweet tho
[18-Jul-2009 18:07:17] <otakup0pe> *crickets*
[18-Jul-2009 20:00:19] <straterra> zenoss installs flawlessly and easily in centos and rhel
[18-Jul-2009 20:15:55] <otakup0pe> i have a variety of debian installs
[18-Jul-2009 20:16:10] <otakup0pe> stracing one of them shows the installer hanging on connecting to localhost:8080
[18-Jul-2009 22:25:29] <SilverFox> I'd sure like to figure this out: http://forums.zenoss.com/viewtopic.php?p=37047#37047
[18-Jul-2009 22:25:40] <SilverFox> been at it for a few days, and am not gaining any ground.
[19-Jul-2009 00:00:46] [disconnected at Sun Jul 19 00:00:46 2009]
[19-Jul-2009 00:00:46] [connected at Sun Jul 19 00:00:46 2009]
[19-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[19-Jul-2009 02:19:09] <Diddi> anyone awake?
[19-Jul-2009 02:20:07] <Diddi> err nvm
[20-Jul-2009 00:00:46] [disconnected at Mon Jul 20 00:00:46 2009]
[20-Jul-2009 00:00:46] [connected at Mon Jul 20 00:00:46 2009]
[20-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[20-Jul-2009 03:05:32] <int> morning!
[20-Jul-2009 03:12:05] <Dieterbe> hi
[20-Jul-2009 03:13:16] <Dieterbe> is it possible to do something like "this process should run on every host in class foo, otherwise alert me" ? i'm confused by the way zenoss seems to work: when you use process monitoring, it seems to autodetect where it runs and alerts if doesn't run anymore on a host where it ran first, which is not what i want
[20-Jul-2009 06:34:24] <Dieterbe> i created some network checks at Classes/services, eg http on port 10080 for example. i have hosts that run this, but when i remodel them, the servicecheck is not added for it? (count stays 0)
[20-Jul-2009 06:34:57] <Dieterbe> also, i cannot add the check manually, when i do " add ip service" on a device, my custom checks are not in the list
[20-Jul-2009 07:08:09] <hackeron> "threshold of low swap not met" < what does that mean?
[20-Jul-2009 07:14:17] <LW> ok... every week for the last four weeks, my data.fs file has apparently been growing ~10% a week.... its now 40GB. Any way to shrink?
[20-Jul-2009 07:26:38] <Dieterbe> hackeron: in the default templates for /Devices/Server, somewhere some thresholds are configured for swap
[20-Jul-2009 07:26:49] <Dieterbe> hackeron: you should look around in the performance templates stuff
[20-Jul-2009 07:27:35] <hackeron> Dieterbe: I'm looking, can't find the tresholds?
[20-Jul-2009 07:29:02] <hackeron> Dieterbe: I'm in /Devices /Server /Linux /Templates /Device /memAvailSwap -- but it's empty, I don't see where to change the threshold?
[20-Jul-2009 07:29:21] <Dieterbe> well it should be there somewhere. i don't know from the top of my head
[20-Jul-2009 07:30:08] <hackeron> oh wait, think I found it
[20-Jul-2009 08:08:44] <int> hello! anyone knows how to reinstall zenpack or how to remove device component?
[20-Jul-2009 08:12:20] HKhan_ is now known as HKhan
[20-Jul-2009 08:32:36] <Dieterbe> int: yes. Management -> settings, there you can administer zenpacks
[20-Jul-2009 08:35:54] <int> Dieterbe: i mean i have zenpack installed in development mode, i updated it and want run upgrade method (like in "trac-zenpacks/browser/zenpacks/ZenPacks.example.Techniques/ZenPacks/example/Techniques/__init__.py")
[20-Jul-2009 08:36:23] <Dieterbe> int: sorry, can't help you with that
[20-Jul-2009 08:37:52] <int> Dieterbe: np, thx for looking
[20-Jul-2009 08:59:18] <rmatte> int: you want to convert an installed zenpack to development mode?
[20-Jul-2009 08:59:49] <rmatte> int: here's the method that we developed: http://blog.zenoss.com/2009/03/29/tip-of-the-month-restoring-zenpacks-to-development-mode/
[20-Jul-2009 09:00:10] <rmatte> oh wait, you want to run an "upgrade"?
[20-Jul-2009 09:00:38] <int> rmatte: yep, want to run upgrade method of zenpack installed in development mode
[20-Jul-2009 09:01:56] <rmatte> if there isn't an option for it in the zenpack script, no idea
[20-Jul-2009 09:02:09] <rmatte> I believe it just intelligently decides when to do an upgrade
[20-Jul-2009 09:02:16] <rmatte> i.e: if it detects new items
[20-Jul-2009 09:02:54] <rmatte> to force it to do one you'd have to write a new script, import the function, and execute it
[20-Jul-2009 09:04:37] <int> rmatte: cool idea, i'll look on zenpack.py sources, looks like it run upgrade in some cases, thx!
[20-Jul-2009 09:04:57] <int> rmatte: do you know how can i remove component of device?
[20-Jul-2009 09:05:23] <int> rmatte: i mean on python level, not from gui
[20-Jul-2009 09:16:54] <rmatte> int: unfortunately I don't
[20-Jul-2009 09:18:11] <int> rmatte: any idea whom i should ask about ZenPacks.example.Techniques ?
[20-Jul-2009 09:19:42] <rmatte> if you see rocket in here, he's a wizard
[20-Jul-2009 09:19:56] <rmatte> not sure if he'll be around today or not though
[20-Jul-2009 09:21:30] <int> rmatte: ok, will ty to ask him
[20-Jul-2009 09:37:31] <rocket> int: have you tried zenpack --link --install <Installed_Zenpack>
[20-Jul-2009 09:37:34] <rocket> from the command line
[20-Jul-2009 09:38:00] <rocket> int: you might need to increment the version number first though for the upgrade method to work
[20-Jul-2009 09:38:57] <int> rocket: afair i tried that, will try that again to be sure
[20-Jul-2009 09:39:10] <int> rocket: any ide how to remove component from device in python?
[20-Jul-2009 09:39:11] <rocket> int: incremented the numbers too?
[20-Jul-2009 09:39:35] <rocket> int: just a single component?
[20-Jul-2009 09:40:12] <rocket> int: I am not exactly sure with zenoss since they have built so many relations into the components .. its a bit tricky if I am thinking correctly
[20-Jul-2009 09:40:29] <int> rocket: trac-zenpacks/browser/zenpacks/ZenPacks.example.Techniques/ZenPacks/example/Techniques/WebServerDevice.py
[20-Jul-2009 09:40:49] <int> rocket: manage_addWebSite defined there, but no manage_deleteWebSite
[20-Jul-2009 09:41:36] <rocket> want to totally delete the website I am assuming?
[20-Jul-2009 09:41:44] <rocket> or component?
[20-Jul-2009 09:42:51] <int> rocket: i thought that website(s) is components of device
[20-Jul-2009 09:43:51] <int> rocket: there is webserver device that has several websites added to it
[20-Jul-2009 09:44:00] <int> so i want to remove one website
[20-Jul-2009 09:44:12] <rocket> int: in this case I dont think it is .. see the line
[20-Jul-2009 09:44:23] <rocket> website = self.websites._getOb(id)
[20-Jul-2009 09:44:36] <rocket> that is actually searching zope for the website object
[20-Jul-2009 09:45:06] <rocket> "del website" ... should probably make it go away
[20-Jul-2009 09:46:31] <int> rocket: del(self.websites._getOb(id)) ?
[20-Jul-2009 09:47:00] <rocket> int: try that first .. you might need to do this too
[20-Jul-2009 09:47:05] <rocket> del(self.websites._getOb(id))
[20-Jul-2009 09:47:20] <rocket> int: actually that probably will do it
[20-Jul-2009 09:47:40] <rocket> might want to see if there is a _delObject method
[20-Jul-2009 09:52:37] <int> rocket: self.websites._delObject('siteid') works! thx!
[20-Jul-2009 09:52:55] <rocket> int: not a problem
[20-Jul-2009 09:53:07] <int> will try to upgrade zenpack now
[20-Jul-2009 09:53:35] <rocket> were you looking for the ubuntu zenpack?
[20-Jul-2009 09:57:41] <int> no
[20-Jul-2009 09:57:51] <int> hmm, it breaks with "ERROR: zenpack command failed. Reason: getattr(): attribute name must be string" now
[20-Jul-2009 10:00:08] <int> found error in my zenpack
[20-Jul-2009 10:01:10] <int> but doing try: run() except exception as e: print str(e); exit(1) is not very friendly, stacktrace would be much better
[20-Jul-2009 10:01:30] <rocket> int: any more than that error? ... when do you see it .. when you are upgrading?
[20-Jul-2009 10:01:57] <rocket> int: I agree (at least for developers a stack trace is much nicer )
[20-Jul-2009 10:02:15] <int> btw, does someone has write access to comunity svn repo? there is typo in trac-zenpacks/browser/zenpacks/ZenPacks.example.Techniques/ZenPacks/example/Techniques/WebServerDevice.py
[20-Jul-2009 10:03:06] <int> rocket: yep, on upgrading, zenpack.py tries to run installing pack and catch all exceptions and print only str(e)
[20-Jul-2009 10:06:32] <int> about typo: it's "**kwargs" in definition and "**kw" on next line
[20-Jul-2009 10:09:51] <int> also bug in trac-zenpacks/browser/zenpacks/ZenPacks.example.Techniques/ZenPacks/example/Techniques/__init__.py in def remove()
[20-Jul-2009 10:10:20] <int> ZenPackBase.remove(self.app, leaveObjects) should probably be ZenPackBase.remove(self, app, leaveObjects)
[20-Jul-2009 10:31:26] <hackeron> I keep getting "ip is down" emails for machines on busy networks and recover in under a minute - is there anyway to get zenoss to retry for a minute before sending an email?
[20-Jul-2009 11:00:40] <LW> is there any reason my data.fs file would be growing by about 10% a week?
[20-Jul-2009 11:16:48] <int> i wonder can i add virtual device and store it perf counters as sum/avg of real devices? ie for example if i has cluster of web servers, then i would like to look on total requests per second graph (across all cluster nodes)
[20-Jul-2009 11:18:07] <int> or only create "fake" device same type and mangle it counters with external script? ideas?
[20-Jul-2009 11:26:16] <int> ZenPacks/example/Techniques has Cluster Requests.rpt, probably that can helps too
[20-Jul-2009 11:27:27] <int> i wonder can zenoss "cache" telnet/ssh connections?
[20-Jul-2009 11:30:44] <kisielk> int: why not create a report for that?
[20-Jul-2009 11:31:22] <kisielk> you can do exactly your example using Multi-Device Reports
[20-Jul-2009 11:32:20] <int> kisielk: with vitual device it would be possible to set thresholds too
[20-Jul-2009 11:33:44] <int> kisielk: also its easier if you look on server1/2/3 and total graphs in one place and not jumps between devices/reports
[20-Jul-2009 11:33:59] <kisielk> sure
[20-Jul-2009 11:34:18] <kisielk> but setting up the multi-device report is like 5 minutes of work, so that's why I suggested it
[20-Jul-2009 11:35:41] <int> kisielk: yep, i'll look on such reports too, thx
[20-Jul-2009 11:36:15] <rocket> hrmm looks like matt is out today ..
[20-Jul-2009 11:42:52] <int> rocket: any idea if zenoss caches ssh/telnet connectons?
[20-Jul-2009 11:43:10] <rocket> as far as I know it does not .. but I could have missed a setting that does that
[20-Jul-2009 11:43:18] <rmatte> hackeron: you should be able to put delays on alerts\
[20-Jul-2009 11:43:25] <rmatte> hackeron: I know you can do it with commands
[20-Jul-2009 11:43:34] <hackeron> rmatte: what's the command?
[20-Jul-2009 11:43:57] <rmatte> no, I mean when you configure commands that kick off when alerts come in, you can set a delay for the command to kick off
[20-Jul-2009 11:44:04] <rmatte> so there must be the same option for alerts, checking now...
[20-Jul-2009 11:44:33] <hackeron> rmatte: thanks
[20-Jul-2009 11:44:37] <rmatte> hackeron: yeh, go to the alert that you configured
[20-Jul-2009 11:44:46] <rmatte> and there's "Delay (secs)"
[20-Jul-2009 11:44:46] <hackeron> rmatte: I didn't configure any - I'm using defaults
[20-Jul-2009 11:44:53] <rmatte> ah
[20-Jul-2009 11:45:07] <hackeron> rmatte: oh, wait, I did configure one, to email when warning
[20-Jul-2009 11:45:24] <rmatte> ok, and those are the emails you are receiving I assume?
[20-Jul-2009 11:45:46] <rmatte> go to that alert, and change Delay (secs) to something like 260 or 300, or whatever
[20-Jul-2009 11:45:48] <hackeron> I believe so
[20-Jul-2009 11:45:58] <rmatte> it defaults to 0 second delay
[20-Jul-2009 11:46:10] <hackeron> ah, there it is, thanks!
[20-Jul-2009 11:46:12] <rmatte> even 120 seconds should be sufficient
[20-Jul-2009 11:46:16] <rmatte> no problem
[20-Jul-2009 11:47:24] <hackeron> trying with 60 seconds for now
[20-Jul-2009 11:48:36] <rmatte> 60 seconds probably won't work too well since it'll kick off the alert right before the event gets cleared, there's like a 50/50 chance of it clearing before it kicks off, you'd be better with like 90 seconds or something
[20-Jul-2009 11:49:04] <hackeron> rmatte: ah, ok
[20-Jul-2009 11:52:18] <int> rocket: it seems it cache, at least for ssh
[20-Jul-2009 12:03:40] <rocket> int: good to know .. I havent played with it enough yet to know for sure .. other than to use it for some development stuff .. thanks for the update ..
[20-Jul-2009 12:16:46] <blocky> does zenoss support snmp contexts?
[20-Jul-2009 12:17:24] <blocky> is it possible to add two devices with the same ip but poll different contexts
[20-Jul-2009 12:33:14] <rmatte> blocky: don't believe so
[20-Jul-2009 12:33:27] <rmatte> blocky: Zenoss won't let you add two devices with the same IP
[20-Jul-2009 12:44:00] <blocky> so i cant do monitoring through a NAT whatsoever?
[20-Jul-2009 12:44:55] <rmatte> not really
[20-Jul-2009 12:45:05] <rmatte> it's agentless monitoring, point to point
[20-Jul-2009 12:47:44] <blocky> so if i needed to monitor multiple servers behind multiple firewalls, what would my best option be?
[20-Jul-2009 12:48:26] <rocket> blocky: not sure .. but you might try setting up a distributed collector behind each firewall
[20-Jul-2009 12:48:41] <rocket> blocky: I am not sure if they are designed to handle this situation ...
[20-Jul-2009 12:48:48] <rmatte> blocky: or setting up tunnels
[20-Jul-2009 12:49:02] <blocky> well i know snmp supports distributed collecting through contexts
[20-Jul-2009 12:49:30] <blocky> wouldn't tunneling still require multiple zenoss devices to have the same ip?
[20-Jul-2009 12:49:55] <rmatte> blocky: not unless it's the exact same subnet behind each firewall?
[20-Jul-2009 12:51:06] <blocky> hmm
[20-Jul-2009 12:51:16] <blocky> still, it's less than ideal
[20-Jul-2009 12:51:29] <rmatte> less than ideal?
[20-Jul-2009 12:51:39] <schwartz> ow.
[20-Jul-2009 12:51:53] <rmatte> we're monitoring 5 different clients' sites on one Zenoss instance with tunnels and it's working fine
[20-Jul-2009 13:09:09] <blocky> you mean sending the snmp requests over a vpn?
[20-Jul-2009 13:09:16] <blocky> what exactly are you tunnelling?
[20-Jul-2009 13:11:37] <ErikRose> Can anyone think of why my CPU graph would be grossly incorrect?
[20-Jul-2009 13:12:05] <ErikRose> It used to be fine, but, ever since I drastically decreased the CPU use on these boxes (2 of them are doing it), the graph hasn't caught up.
[20-Jul-2009 13:12:27] <ErikRose> top shows 0% practically all the time, but the CPU graph still reads 40.
[20-Jul-2009 13:12:39] <ErikRose> Could RRD be stubborn somehow?
[20-Jul-2009 13:14:35] <administrator_> hello all new user of zenoss , the maybe a a silly question but where can i find the zendisk agent?
[20-Jul-2009 13:22:05] <rmatte> ErikRose: you might have to delete the RRD file and let Zenoss recreate it, not sure what might have caused that, have you tried restarting Zenoss?
[20-Jul-2009 13:22:25] <ErikRose> rmatte: I have not restarted Zenoss. I'll give that a shot before deleting anything. Thanks!
[20-Jul-2009 13:22:31] <rmatte> np
[20-Jul-2009 13:29:02] <venturaville> are there any big zenoss events/meetings/etc... in southeast US over the next 6 months or so?
[20-Jul-2009 13:34:13] <gwb2351> easy question: Trying to install http://www.zenoss.com/community/projects/zenpacks/brocade-switches and errors with
[20-Jul-2009 13:34:28] <gwb2351> WARNING:zen.AddToPack:Unable to find context path /zport/dmd/Devices/Blade Switches/rrdTemplates (line 4 ?) for BrocadeSwitch WARNING:zen.AddToPack:Not committing any changes ERROR: zenpack command failed. Reason: Unable to create object using the following attributes: * id: /zport/dmd/Devices/Blade Switches/rrdTemplates/BrocadeSwitch * module: Products.ZenModel.RRDTemplate * class: RRDTemplate
[20-Jul-2009 13:34:40] <gwb2351> suggestions?
[20-Jul-2009 13:41:37] * ErikRose restarts snmpd, too.
[20-Jul-2009 13:51:18] <ErikRose> Huh. Zenoss is fine. snmpwalk shows that SNMP is lying to me.
[20-Jul-2009 13:56:36] <eidolon> hey folks - can i add a user defined alerting mechanism? we're pairing up with a third party monitoring company. right now alerts can only be pushed 'out system' via 'pager' or 'email'. can i addsomething that will, say, do a REST post call to a remote server?
[20-Jul-2009 13:56:41] <eidolon> (we have zenoss enterprise)
[20-Jul-2009 13:56:45] <cgibbons> do you have multiple CPUS, Erik?
[20-Jul-2009 13:56:47] <rmatte> gwb2351: just a warning, that ZenPack isn't really that great, it's only setup to monitor the first port on the switch
[20-Jul-2009 13:56:58] <ErikRose> cgibbons: It's a VMWare VM with one logical CPU. Other VMs are behaving.
[20-Jul-2009 13:57:01] administrator_ is now known as kd4zay
[20-Jul-2009 13:57:05] <cgibbons> huh weird
[20-Jul-2009 13:57:09] <rocket> eidolon: if not that sounds like a decent enhancement or a zenpack candidate
[20-Jul-2009 13:57:15] <ErikRose> cgibbons: For some reason, SNMP just keeps reporting 19.
[20-Jul-2009 13:57:32] <eidolon> rocket: i'd just have to learn Python.  again. 
[20-Jul-2009 13:57:41] <rmatte> gwb2351: also, to install that ZenPack you need to unzip the .zip, then you should see a .egg, use that to install the pack, you also need to create /Devices/Blade Switches by hand.
[20-Jul-2009 13:58:25] <cgibbons> does it do it if you restart the snmp agent on the box? does the VM have the VMware tools installed?
[20-Jul-2009 13:59:08] <ErikRose> cgibbons: Yes, I restarted snmpd—no difference. We're using the openvm-tools instead of the stock VMWare tools, but yes.
[20-Jul-2009 13:59:09] <eidolon> i suppose since we're not using SNNP i could just use that.
[20-Jul-2009 13:59:53] <eidolon> er.  SNPP.  the pager htingy.
[20-Jul-2009 13:59:59] <cgibbons> weird
[20-Jul-2009 14:00:17] <ErikRose> cgibbons: That's what I think, too. :-)
[20-Jul-2009 14:10:45] <rmatte> eidolon: instead of an alert you can specify commands to run when events come in
[20-Jul-2009 14:11:37] <rmatte> eidolon: You mentioned "user defined", what exactly are you shooting for?
[20-Jul-2009 14:11:45] <gwb2351> i'll see if I can get what I want from the BrocadeMonitor zenpack
[20-Jul-2009 14:12:03] <rmatte> eidolon: if you go to "Event Manager" on the left, then click the Commands tab you can specify commands (execute a script and pass values to it or whatever).
[20-Jul-2009 14:22:29] <eidolon> rmatte: we're setting up a mechanism to alerta third party monitoring system whenever we have internal issues.
[20-Jul-2009 14:23:28] <eidolon> rmatte: what i want is something akin to 'paging' and 'email' - but this one should be 'RPC call' or omething. it forwards the alert out to another system.
[20-Jul-2009 14:24:17] <eidolon> an example is there's a 'page command' that's hooked to the 'pager' function in alerting rules.
[20-Jul-2009 14:24:33] <eidolon> i want to add to 'email' and 'page' something like 'remotealert' - that would be the best arrangement.
[20-Jul-2009 14:26:49] <eidolon> brb.  meeting, i'll loig in from there.
[20-Jul-2009 14:56:59] <kd4zay> is there a way to delete host entrys?
[20-Jul-2009 14:58:16] <rmatte> host entries?
[20-Jul-2009 14:58:24] <rmatte> could you be a bit more specific?
[20-Jul-2009 15:00:03] <kd4zay> can you delete a host entry once its been added
[20-Jul-2009 15:00:27] <rmatte> when you say "host entry" what are you referring to?
[20-Jul-2009 15:00:43] <rmatte> that could mean like five different things
[20-Jul-2009 15:00:46] <kd4zay> servers that have been discovered
[20-Jul-2009 15:00:57] <rmatte> you mean delete a device from Zenoss?
[20-Jul-2009 15:01:03] <kd4zay> yes
[20-Jul-2009 15:01:35] <rmatte> of course, just go to the Device List, select the device you want to delete, go to the dropdown, and select Delete Devices
[20-Jul-2009 15:01:52] <rmatte> you can delete multiple devices at once
[20-Jul-2009 15:02:10] <kd4zay> im new to zenoss
[20-Jul-2009 15:02:45] <kd4zay> also is there a agent called zendisc?
[20-Jul-2009 15:03:12] <rmatte> no, there's no daemon called zendisc
[20-Jul-2009 15:03:42] <rmatte> there's an executable called zendisc
[20-Jul-2009 15:04:00] <kd4zay> is that a service for windows servers
[20-Jul-2009 15:04:22] <rmatte> if you ssh to the server then become the zenoss user ("su - zenoss" as root) then type zendisc you'll see the options
[20-Jul-2009 15:04:41] <rmatte> no, zendisc is the script used by Zenoss to discover any kind of device
[20-Jul-2009 15:04:45] <rmatte> it's not windows specific
[20-Jul-2009 15:04:45] <kd4zay> im trying to get the hard drive capacity and other info from my windows servers
[20-Jul-2009 15:05:16] <kd4zay> so far i just have snmp enabled
[20-Jul-2009 15:06:03] <rmatte> make sure that the snmp service is configured properly on them, then in Zenoss on the left click on Devices, then go to zProperties tab, scroll down and find zSnmpCommunity and set it to your SNMP community string, then click Save at the bottom of the page.
[20-Jul-2009 15:06:25] <kd4zay> ok
[20-Jul-2009 15:06:47] <rmatte> then click on Devices on the left, and click on Discovered, since that's the default class that the devices end up in when you do autodiscovery
[20-Jul-2009 15:07:51] <rmatte> actually, go to Device List, select your windows devices, go to the dropdown, and select Move to Class
[20-Jul-2009 15:08:01] <rmatte> then move them to the /Devices/Server/Windows class
[20-Jul-2009 15:08:47] <rmatte> then for each device, click on it to open the device page, then go to the dropdown on the device page and select Manage -> Model Device
[20-Jul-2009 15:08:55] <rmatte> then wait for it to model, and it should pick up all the info
[20-Jul-2009 15:09:33] <rmatte> you'll also need to install snmp-informant on your servers if you want to easily gather the performance info via SNMP, otherwise you'll have to download the WMI Performance ZenPack by James Dastrup and use that
[20-Jul-2009 15:09:56] <kd4zay> ok
[20-Jul-2009 15:17:05] <bobrog> anybody use the /Ping device class? i would like to but it is no present on my 2.4.2 appliance. The 2.4.2 admin guide says /Ping should be one of the default device classes.
[20-Jul-2009 15:20:10] <mrayzenoss> bobrog: so if you go to Devices, you have Discovered, KVM, Network, Power, Printer and Server but no Ping?
[20-Jul-2009 15:22:14] <bobrog> mrayzenoss: thats correct
[20-Jul-2009 15:22:37] <mrayzenoss> Well, I guess you can add it back, but I'm curious how it got deleted
[20-Jul-2009 15:22:40] <bobrog> i have a couple of other classes you did not mention probably becasue its enterprise edition but no Ping
[20-Jul-2009 15:23:26] <mrayzenoss> If you go to the page menu you can "Add Organizer" and re-add Ping
[20-Jul-2009 15:23:38] <bobrog> what are the properties of the Ping class?  is it just one with no collector plugins?
[20-Jul-2009 15:23:43] <bobrog> ok
[20-Jul-2009 15:23:44] <mrayzenoss> I'm looking
[20-Jul-2009 15:24:22] <mrayzenoss> Looks like it has a custom Template for Device that is blank
[20-Jul-2009 15:25:04] <mrayzenoss> that template is bound
[20-Jul-2009 15:26:01] <mrayzenoss> everything gets pinged, so I don't even think you need the blank template
[20-Jul-2009 15:26:39] <mrayzenoss> just add a new organizer for it and make sure that zPingMonitorIgnore is False
[20-Jul-2009 15:27:08] <mrayzenoss> yeah, I just added Ping2 as a device class, it just works
[20-Jul-2009 15:27:47] <bobrog> ok. do i need to remove the default snmp collector plugins? so it doesnt even try to use that, or maybe set a zproperty to not use snmp?
[20-Jul-2009 15:27:48] <mrayzenoss> set zSnmpMonitorIgnore to True
[20-Jul-2009 15:27:56] <bobrog> cool
[20-Jul-2009 15:28:00] <bobrog> ill try this now
[20-Jul-2009 15:31:38] <twm1010> Afternoon people...
[20-Jul-2009 15:32:10] <twm1010> Could I trouble someone to help a newb with a regular expression
[20-Jul-2009 15:32:50] * rocket has mastered regular expressions in grade school ... eg "Who cut the cheese?"
[20-Jul-2009 15:32:52] <bobrog> mrayzenoss: this seems to have the desired effect, thanks for the assistance
[20-Jul-2009 15:33:08] <twm1010> Probably a simple question...
[20-Jul-2009 15:33:21] <twm1010> I use the following RegEx to filter out some interfaces I don't need to monitor
[20-Jul-2009 15:33:22] <twm1010> ^(unrouted|rif0|lo|VLAN-|sit|EFXS|FXO|Voice)
[20-Jul-2009 15:34:09] <twm1010> Now, I also want it to filter out interfaces ending -Bearer Channel
[20-Jul-2009 15:34:56] <rocket> .*-Bearer Channel       ???
[20-Jul-2009 15:36:00] <twm1010> Sure, and to combine the two into one RegEx?
[20-Jul-2009 15:36:31] <rocket> ^(unrouted|rif0|lo|VLAN-|sit|EFXS|FXO|Voice|.*-Bearer Channel)    ???
[20-Jul-2009 15:36:55] <twm1010> That easy huh?
[20-Jul-2009 15:37:04] <rmatte> bobrog: just go through the zproperties for that class and disable everything except for ping monitoring
[20-Jul-2009 15:40:25] <twm1010> rocket: Thanks.
[20-Jul-2009 15:40:46] <rocket> twm1010: that work for you?
[20-Jul-2009 15:41:21] <twm1010> Indeed. between the part I added and that last touch my Voice gateways now show only 200 interfaces instead of 1000+
[20-Jul-2009 15:41:40] <twm1010> it was showing every channel of a PRI of course, and then it's bearer channel
[20-Jul-2009 15:41:46] <twm1010> adds up fast.
[20-Jul-2009 15:43:57] <twm1010> Rest of them should remodel tonight... that'll reduce load on the collector quite a bit.
[20-Jul-2009 19:33:38] <blocky> anybody know where i can find the oid for an hp dl360 g5 to monitor temperature
[20-Jul-2009 20:24:50] <|malibu|> can zenoss do ssh-based monitoring of aix?
[21-Jul-2009 00:00:46] [disconnected at Tue Jul 21 00:00:46 2009]
[21-Jul-2009 00:00:46] [connected at Tue Jul 21 00:00:46 2009]
[21-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[21-Jul-2009 03:55:54] <Mituc> hi guys
[21-Jul-2009 08:19:13] <klinstifen> morning all
[21-Jul-2009 08:19:34] <klinstifen> anyone have this issue: hosts randomly report snmp is down even though you can snmpwalk the host ?
[21-Jul-2009 08:20:14] <rmatte> sounds like link issues
[21-Jul-2009 08:29:19] <klinstifen> link issues?
[21-Jul-2009 08:30:01] <klinstifen> like the physical link?
[21-Jul-2009 08:30:34] <klinstifen> if that's what you mean i don't think it is it as there are no connectivity issues with this host
[21-Jul-2009 08:33:43] <jb> odd.. like 10 of my windows hosts had their snmpd crash last night
[21-Jul-2009 08:36:58] <rmatte> klinstifen: how many boxes you monitoring?
[21-Jul-2009 08:37:25] <rmatte> you might try going in to zProperties for /Devices and increasing the snmp timeout value
[21-Jul-2009 08:37:25] <klinstifen> 16
[21-Jul-2009 08:37:38] <klinstifen> hmm that's an idea...might just be a timeout issue
[21-Jul-2009 08:37:41] <klinstifen> i'll give that a go
[21-Jul-2009 08:37:52] <rmatte> you should also check the timeouts in your collector settings
[21-Jul-2009 08:37:57] <rmatte> if that doesn't work
[21-Jul-2009 08:46:32] <jb> <rmatte> if that doesn't work
[21-Jul-2009 08:46:34] <jb> oops
[21-Jul-2009 08:46:39] <jb> evt.component = "Node: %s:%s" % (gethostbyaddr(evt.bigipNotifyObjNode), evt.bigipNotifyObjPort)
[21-Jul-2009 08:46:57] <jb> Node: ('isatest02.corp.follett.com', [], ['172.26.100.46']):0
[21-Jul-2009 08:47:12] <jb> does anyone know why the node is being displayed like this?
[21-Jul-2009 08:47:16] <jb> array-ish
[21-Jul-2009 08:47:39] <jb> i simply want to display Node: isatest02.corp.follett.com:80
[21-Jul-2009 08:47:56] <jb> if possible
[21-Jul-2009 09:23:09] <rmatte> Does anyone in here have a fresh Zenoss install available? I just need to check if certain items exist by default
[21-Jul-2009 09:24:42] <rmatte> jb: have you tried setting those as variable prior to setting them in to evt.component?
[21-Jul-2009 09:25:27] <rmatte> like host=gethostbyaddr(evt.bigipNotifyObjNode) ip=evt.bigipNotifyObjPort then do evt.component = "Node: %s:%s" % (host,ip)
[21-Jul-2009 09:25:37] <rmatte> that might work better
[21-Jul-2009 09:26:16] <jb> nope.. I'll try that
[21-Jul-2009 09:26:17] <jb> thanks
[21-Jul-2009 09:26:37] <rmatte> jb: np, I always tend to do it that way since it tends to yield less unexpected results
[21-Jul-2009 09:27:40] <rmatte> also, what are you trying to do exactly, just display the hostname and IP of the device that the alert came in for?
[21-Jul-2009 09:28:23] <jb> yeah
[21-Jul-2009 09:28:29] <jb> this is for a Bigip
[21-Jul-2009 09:28:35] <rmatte> Bigip?
[21-Jul-2009 09:28:39] <jb> F5 LTM
[21-Jul-2009 09:28:40] <jb> load balancer
[21-Jul-2009 09:28:49] <jb> basically, I get a trap when a "service" goes down
[21-Jul-2009 09:29:04] <jb> bigipNotifyObjNode/ObjPort tell me the IP:Port of that node
[21-Jul-2009 09:29:11] <rmatte> ah
[21-Jul-2009 09:29:45] <rmatte> when the event comes in, if you click on the details link (the magnifying glass on the right), then go to the details tab, do you need the info in there?
[21-Jul-2009 09:30:03] <rmatte> you might not be doing this the best way possible
[21-Jul-2009 09:30:17] <rmatte> do you see the info there*
[21-Jul-2009 09:30:18] <rmatte> rather
[21-Jul-2009 09:30:35] <jb> i really just need it in the alerts via email
[21-Jul-2009 09:30:39] <jb> and in the event console as well
[21-Jul-2009 09:30:50] <rmatte> right, I meant to say see
[21-Jul-2009 09:31:00] <rmatte> if you go to the details tab, is the info that you're looking for listed there?
[21-Jul-2009 09:31:08] <rmatte> If it is, I have an easier solution for you
[21-Jul-2009 09:31:29] <jb> yep, its there
[21-Jul-2009 09:31:38] <jb> oh wow
[21-Jul-2009 09:31:39] <twm1010> Morning!
[21-Jul-2009 09:31:43] <jb> bigipNotifyObjMsg  Pool member 172.26.103.92:80 monitor status down.
[21-Jul-2009 09:31:43] <jb> bigipNotifyObjNode 172.26.103.92
[21-Jul-2009 09:31:44] <jb> bigipNotifyObjPort 80
[21-Jul-2009 09:31:52] <rmatte> ok, what's it listed as?  bigipNotifyObjNode and bigipNotifyObjPort ?
[21-Jul-2009 09:31:56] <jb> yep
[21-Jul-2009 09:32:00] <jb> but i need the hostname
[21-Jul-2009 09:32:06] <jb> so i need to perform a gethosbyaddr() on the node
[21-Jul-2009 09:32:13] <rmatte> ah, I see, it doesn't actually provide the hostname
[21-Jul-2009 09:32:15] <rmatte> gotcha
[21-Jul-2009 09:32:36] <jb> im curious as to what causes those details to get there?
[21-Jul-2009 09:32:45] <rmatte> your Mibs
[21-Jul-2009 09:32:51] <jb> ok cool
[21-Jul-2009 09:32:56] <rmatte> they tell Zenoss how to interpret the data from the trap
[21-Jul-2009 09:33:27] <rmatte> so did it work with the variables?
[21-Jul-2009 09:34:03] <jb> .
[21-Jul-2009 09:34:09] <jb> sec
[21-Jul-2009 09:34:10] <rmatte> ..
[21-Jul-2009 09:34:12] <jb> \
[21-Jul-2009 09:34:30] <jb> <jb> sec
[21-Jul-2009 09:34:32] <jb> argh
[21-Jul-2009 09:34:36] <rmatte> hehe
[21-Jul-2009 09:34:40] <jb> can't even paste today
[21-Jul-2009 09:34:44] <jb> evt.component = host=gethostbyaddr(evt.bigipNotifyObjNode) ip=evt.bigipNotifyObjPort then do evt.component =
[21-Jul-2009 09:34:46] <jb>           "Node: %s:%s" % (host,ip)"
[21-Jul-2009 09:34:59] <rmatte> yeh
[21-Jul-2009 09:34:59] <jb> no evt.component
[21-Jul-2009 09:35:06] <jb> sec
[21-Jul-2009 09:35:21] <jb> hmm.. not valid
[21-Jul-2009 09:35:34] <rmatte> pastebin the code
[21-Jul-2009 09:35:43] <twm1010> curious: In regards to interface alerting, do you let Zenoss do a poll and see that the interface is down, or do you use a trap to do so?
[21-Jul-2009 09:35:50] <rmatte> actually, it should technically only be single quotes
[21-Jul-2009 09:35:52] <rmatte> ' '
[21-Jul-2009 09:35:53] <jb> http://pastebin.ca/1502075
[21-Jul-2009 09:36:02] <jb> k
[21-Jul-2009 09:36:17] <jb> no go
[21-Jul-2009 09:36:26] <rmatte> oh no no
[21-Jul-2009 09:36:42] <rmatte> I wasn't actually typing it in the correct format, I was lazy and put it all on one line lol
[21-Jul-2009 09:36:51] <jb> oh
[21-Jul-2009 09:36:51] <jb> hah
[21-Jul-2009 09:36:57] <rmatte> host=gethostbyaddr(evt.bigipNotifyObjNode)
[21-Jul-2009 09:36:59] <jb> sorry, im clueless when it comes to this.
[21-Jul-2009 09:36:59] <rmatte> ip=evt.bigipNotifyObjPort
[21-Jul-2009 09:37:11] <rmatte> evt.component = "Node: %s:%s" % (host,ip)"
[21-Jul-2009 09:37:23] <rmatte> and sorry, space out between the =
[21-Jul-2009 09:38:07] <rmatte> host = gethostbyaddr(evt.bigipNotifyObjNode)
[21-Jul-2009 09:38:07] <rmatte> ip = evt.bigipNotifyObjPort
[21-Jul-2009 09:38:08] <rmatte> evt.component = 'Node: %s:%s' % (host,ip)
[21-Jul-2009 09:38:10] <rmatte> like that
[21-Jul-2009 09:38:44] <rmatte> and if that doesn't work I have another idea
[21-Jul-2009 09:40:16] <jb> hmm.. it's still red
[21-Jul-2009 09:40:28] <rmatte> ok... one sec
[21-Jul-2009 09:41:10] <twm1010> So... I could set a critical threshold on the ethernetCSmacD template, if ifOperstatus = whatever the failed value is, right?
[21-Jul-2009 09:41:35] <rmatte> jb: try this:
[21-Jul-2009 09:41:36] <rmatte> ip = getattr(evt, 'bigipNotifyObjPort', 'Unknown')
[21-Jul-2009 09:41:36] <rmatte> host = gethostbyaddr(ip)
[21-Jul-2009 09:41:36] <rmatte> evt.component = 'Node: %s:%s' % (host,ip)
[21-Jul-2009 09:42:01] <rmatte> if that doesn't work I'm stumped
[21-Jul-2009 09:42:10] <jb> there we go
[21-Jul-2009 09:42:22] <jb> ok, lets see how it turns out
[21-Jul-2009 09:42:24] <jb> thanks
[21-Jul-2009 09:42:29] <rmatte> always use the getattr function for setting variable from details data
[21-Jul-2009 09:42:35] <jb> k
[21-Jul-2009 09:42:37] <rmatte> seems to work the best
[21-Jul-2009 09:44:15] <int> i wonder if someone uses telnet to get perf data?
[21-Jul-2009 09:44:53] <rmatte> int: never heard of that being done, though you could write a script to connect via telnet, and pull the data, then output it in nagios format and have Zenoss run it as a command based datapoint
[21-Jul-2009 09:44:59] <rmatte> you'd have to write an expect script or something
[21-Jul-2009 09:45:28] <rmatte> not really the best solution in the world though
[21-Jul-2009 09:45:47] <rmatte> having telnet usernames and password flying out plaintext every 5 mins is a security risk
[21-Jul-2009 09:46:44] <int> rmatte: yep, i tried that way, but that seems sub-optimal
[21-Jul-2009 09:46:52] <int> it would be better not to login every time
[21-Jul-2009 09:47:18] <rmatte> int: which would mean you'd have to have telnet configured as passwordless on the device
[21-Jul-2009 09:47:21] <rmatte> which is even worse
[21-Jul-2009 09:47:40] <rmatte> unless you run some special app on the telnet port that collects the data and spits it out on connect
[21-Jul-2009 09:47:51] <rmatte> then have a command to connect and capture the data
[21-Jul-2009 09:47:51] <int> rmatte: i mean login once and then use that connection (and reconnect in case network issues)
[21-Jul-2009 09:48:20] <rmatte> int: even the ssh based performance collection logs in each time, it doesn't maintain an active connection
[21-Jul-2009 09:48:39] <rmatte> using telnet for monitoring is not a practical solution, snmp was invented for a reason
[21-Jul-2009 09:48:45] <int> rmatte: it seems it reuses connection
[21-Jul-2009 09:49:34] <rmatte> int: not from what I've seen
[21-Jul-2009 09:51:45] <rmatte> http://dev.zenoss.org/trac/ticket/4273
[21-Jul-2009 09:52:24] <int> rmatte: there is no messages about avery minute ssh logins in logs, but it gets data somehow, and there is active ssh proccess
[21-Jul-2009 09:53:10] <rmatte> ok, maybe it does, just seems a bit intensive to leave all those ssh connections open
[21-Jul-2009 09:53:29] <rmatte> if you were monitoring like 1000 servers it'd get pretty intense
[21-Jul-2009 09:54:03] <Diddi> ok this is driving me crazy... I've installed zenoss 2.4.2 on a freshly installed debian with apt-get (according to the docs). everything seems to be up and running, and I can access the web interface... but in the dashboard it says "Lost connection to the server".. I can't figure out what's wrong
[21-Jul-2009 09:55:11] <rmatte> Diddi: That's indicating that there's an error connection to the MySQL database
[21-Jul-2009 09:55:17] <rmatte> connecting*
[21-Jul-2009 09:55:23] <rmatte> make sure that MySQL is running properly
[21-Jul-2009 09:55:24] <int> rmatte: is that possible to push perf data from external scripts to zenoss?
[21-Jul-2009 09:56:04] <Diddi> rmatte: mysql is running on port 3307.. I hope zenoss is confed to use it... what's the default credentials, so I can try myself?
[21-Jul-2009 09:56:31] <rmatte> int: not really possible to push it, Zenoss needs to execute the script, I suppose you could have scripts actively running pushing to file(s) and then have Zenoss executing a script that scrapes the file(s) for data
[21-Jul-2009 09:57:00] <rmatte> Diddi: was MySQL installed already when you installed Zenoss?
[21-Jul-2009 09:57:06] <Diddi> nope
[21-Jul-2009 09:57:36] <Diddi> fresh install with debian, had nothing from the beginning... I've only installed zenoss-stack with apt-get
[21-Jul-2009 09:57:44] <Diddi> and all dependencies is included there
[21-Jul-2009 09:57:53] <int> rmatte: yep, but then you will hit issues with locks etc, that's what i like to avoid
[21-Jul-2009 09:58:30] <rmatte> int: not necessarily, it's like performing a tail -f on a log file that's being actively written to
[21-Jul-2009 09:58:46] <rmatte> Diddi: hmmm
[21-Jul-2009 09:59:37] <Diddi> had the same problem when I tried from source on another computer... where I definitly know the database-connection was working properly
[21-Jul-2009 09:59:42] <Diddi> no errors in any logs
[21-Jul-2009 09:59:46] <rmatte> Diddi: I'm trying to remember where the MySQL config is stored
[21-Jul-2009 10:02:53] <straterra> Is it easy to move an install of zenoss?
[21-Jul-2009 10:03:06] <straterra> I have it configured in a virtual machine right now..and I'm going to be moving it to a 'real' machine
[21-Jul-2009 10:04:50] <rmatte> yeh, 3307 is the default
[21-Jul-2009 10:05:02] <rmatte> are you able to connect to mysql on that port?
[21-Jul-2009 10:05:45] <Diddi> yeah
[21-Jul-2009 10:06:03] <Diddi> altough I don't have any user/pass, but netcat still approves
[21-Jul-2009 10:06:05] <Diddi> (:
[21-Jul-2009 10:06:18] <rmatte> then it's probably an authentication issue
[21-Jul-2009 10:06:27] <rmatte> did you set a username and pass on install or just take the defaults?
[21-Jul-2009 10:06:54] <Diddi> the apt-get install is quite non-interactive
[21-Jul-2009 10:07:07] <Diddi> so I'd bet defaults
[21-Jul-2009 10:07:32] <rmatte> ok, do: mysql -u zenoss -p
[21-Jul-2009 10:07:35] <rmatte> then type in zenoss
[21-Jul-2009 10:07:39] <rmatte> see if you get a prompt
[21-Jul-2009 10:09:19] <Diddi> that works fine
[21-Jul-2009 10:09:37] <rmatte> k, super weird then, how did you start zenoss?
[21-Jul-2009 10:09:47] <Diddi>  /etc/init.d/zenoss-stack start
[21-Jul-2009 10:09:54] <twm1010> but if you installed from stack, it runs its own MySQL
[21-Jul-2009 10:09:59] <Diddi> indeed
[21-Jul-2009 10:10:02] <Diddi> and that I do
[21-Jul-2009 10:10:07] <twm1010> probably with it's own usernames and passwords, preset.
[21-Jul-2009 10:10:10] <rmatte> become root, then do: su - zenoss
[21-Jul-2009 10:10:19] <rmatte> then type "zenoss stop"
[21-Jul-2009 10:10:25] <rmatte> then type "zenoss start"
[21-Jul-2009 10:10:28] <rmatte> and check it again
[21-Jul-2009 10:10:38] <twm1010> it would probably be zenoss-stack stop/start
[21-Jul-2009 10:10:47] <rmatte> twm1010: they both work
[21-Jul-2009 10:10:53] <twm1010> ah, as zenoss user right..
[21-Jul-2009 10:11:01] <rmatte> twm1010: the zenoss-stack is just a startup script
[21-Jul-2009 10:11:06] <rmatte> twm1010: yeh, as zenoss user
[21-Jul-2009 10:11:15] <twm1010> does the zenoss command as zenoss user start and stop the built-in sql instance tho?
[21-Jul-2009 10:11:22] <twm1010> i don't think so, that threw me off building my SUSE install
[21-Jul-2009 10:11:25] <rmatte> twm1010: yes, it does
[21-Jul-2009 10:11:43] <Diddi> same error
[21-Jul-2009 10:12:13] <rmatte> twm1010: actually hmmm, it doesn't lol, I never noticed that before
[21-Jul-2009 10:12:40] <twm1010> as root, try service zenoss-stack stop/start
[21-Jul-2009 10:12:47] <twm1010> or, just restart would probably do it
[21-Jul-2009 10:13:24] <rmatte> I just do stop to make sure everything has a proper chance to exit
[21-Jul-2009 10:13:38] <rmatte> the restart it quite quick
[21-Jul-2009 10:13:41] <rmatte> is*
[21-Jul-2009 10:13:44] <Diddi> twm1010: service doesn't exist..
[21-Jul-2009 10:13:53] <Diddi> command that is
[21-Jul-2009 10:14:02] <rmatte> Diddi: nah, /etc/init.d/zenoss-stack restart
[21-Jul-2009 10:14:20] <rmatte> service is something redhat specific I assume?
[21-Jul-2009 10:14:24] <Diddi> yeah I think so
[21-Jul-2009 10:14:43] <rmatte> redhat/centos/suse specific
[21-Jul-2009 10:15:20] <rmatte> that's why I'm not a fan of any of those distros, they set they don't follow many standards
[21-Jul-2009 10:15:35] <Diddi> I'm only running debian to get zenoss working
[21-Jul-2009 10:15:47] <Diddi> but that didn't help
[21-Jul-2009 10:15:54] <rmatte> I'm running Zenoss on Ubuntu Server and it works fine
[21-Jul-2009 10:16:15] <rmatte> Ubuntu Server 8.04, and I did the install from the stack installer manually
[21-Jul-2009 10:17:04] <Diddi> but it should work with debian
[21-Jul-2009 10:17:16] <rmatte> yes, it should work fine with Debian
[21-Jul-2009 10:17:21] <rmatte> it sounds like you're cursed or something
[21-Jul-2009 10:17:23] <rmatte> lol
[21-Jul-2009 10:17:35] <rmatte> hold on, I'm trying to find where Zenoss stores it's MySQL config
[21-Jul-2009 10:18:32] <Diddi> take your time (: i'm on my second week trying to get this working
[21-Jul-2009 10:18:37] <Diddi> hehe
[21-Jul-2009 10:18:45] <rmatte> that's crazy
[21-Jul-2009 10:18:49] <Diddi> indeed
[21-Jul-2009 10:18:59] <rmatte> we didn't have any problems installing it at first, just took a while to figure out some stuff
[21-Jul-2009 10:19:03] <rmatte> like compiling Cisco Mibs
[21-Jul-2009 10:19:13] <Diddi> student.. no time... eager to get it working from source...
[21-Jul-2009 10:19:18] <rmatte> hehe
[21-Jul-2009 10:19:21] <Diddi> no time == too much time
[21-Jul-2009 10:21:23] <rmatte> if one of the devs or engineers were around here they'd know this right away, I've just never had to reconfig MySQL on my Zenoss installs
[21-Jul-2009 10:23:00] <Diddi> hehe
[21-Jul-2009 10:35:02] <rmatte> yeh, I can't find it, it must be stored in Zope somewhere
[21-Jul-2009 10:36:46] <Diddi> hmm
[21-Jul-2009 10:47:37] <sto6ma9ch> Hello.
[21-Jul-2009 10:47:40] <rmatte> great, I've got a device that I can snmpwalk perfectly fine, and that I've verified that Zenoss has the right settings for, but it's throwing SNMP agent down notifications for it
[21-Jul-2009 10:47:57] <sto6ma9ch> I have one server reporting negative disk space.
[21-Jul-2009 10:48:54] <rmatte> sto6ma9ch: remodel the device
[21-Jul-2009 10:51:30] <sto6ma9ch> Done. I'll see if that takes care of it after a cycle.
[21-Jul-2009 10:51:51] <rmatte> sto6ma9ch: something probably changed on the server and Zenoss didn't remodel to pick up on it. Either that, or the initial modelling didn't work properly
[21-Jul-2009 10:52:02] <rmatte> 99% of the time that's the issue when it comes to drive space being reported incorrectly
[21-Jul-2009 10:52:12] <rmatte> you may need to delete the RRD file and have Zenoss regenerate it as well
[21-Jul-2009 10:52:21] <rmatte> but just leave it at a remodel for now
[21-Jul-2009 11:13:24] <rmatte> hey Matt, any idea what would cause Zenoss to generate SNMP agent down notifications for a device after changing the community string and renaming the device from the IP address to the name of the device?
[21-Jul-2009 11:13:35] <rmatte> I've tried remodelling, pushing changes, doing a reindex, nothing is working
[21-Jul-2009 11:14:15] <rmatte> the device is fully snmpwalkable
[21-Jul-2009 11:15:51] <rmatte> and it was working fine before, all we did was change the string on the device, updated it in Zenoss, and renamed the device
[21-Jul-2009 11:15:52] <mrayzenoss> if you move the event to history does it come back?
[21-Jul-2009 11:16:01] <rmatte> mrayzenoss: yeh, it does
[21-Jul-2009 11:16:49] <rmatte> I've seen people in here with this type of issue before, just haven't had it actually happen to us before
[21-Jul-2009 11:17:10] <kisielk> my zenprocess keeps dying
[21-Jul-2009 11:17:32] <kisielk> even though the state is reported as green
[21-Jul-2009 11:17:43] <kisielk> it shows up in the Zenoss issues, and it doesn't collect data
[21-Jul-2009 11:17:46] <kisielk> I have to restart it every day
[21-Jul-2009 11:17:51] <venturaville> mrayzenoss: are there any conventions/training things coming up in the zenoss world soon ... my boss is making budget for those things for the rest of the year............
[21-Jul-2009 11:18:01] <rmatte> kisielk: check $ZENHOME/logs/zenprocess.log
[21-Jul-2009 11:18:32] <kisielk> ya, it just says 2009-07-21 09:15:22 ERROR zen.zenprocess: performance scan job not finishing: 10 jobs running 82 jobs waiting 3 jobs finished
[21-Jul-2009 11:18:37] <mrayzenoss> venturaville: we'll be at the Red Hat Summit, VMworld, Ohio Linux Fest and USENIX:LISA
[21-Jul-2009 11:18:38] <kisielk> over and over again
[21-Jul-2009 11:18:46] <jb> hm
[21-Jul-2009 11:18:49] <jb> hey rmatte
[21-Jul-2009 11:18:54] <mrayzenoss> venturaville: we may do our free training at OLF or LISA
[21-Jul-2009 11:19:09] <rmatte> jb: yeh?
[21-Jul-2009 11:19:27] <jb> the "Component" ended up being blank
[21-Jul-2009 11:19:33] <jb> evt.component = 'Node: %s:%s' % (host,ip)
[21-Jul-2009 11:19:41] <jb> wonder if its because im missing that trailing '
[21-Jul-2009 11:19:56] <rmatte> you shouldn't need a trailing '
[21-Jul-2009 11:19:58] <jb> oh, no.. thats not it
[21-Jul-2009 11:21:09] <rmatte> pastebin what you have please, I closed the window that I had it stored in
[21-Jul-2009 11:22:19] <venturaville> mrayzenoss: I added LISA to the list I handed up
[21-Jul-2009 11:23:00] <venturaville> mrayzenoss: Its been a while since I last went to LISA. one of the organizers used to work here
[21-Jul-2009 11:23:37] <jb> http://pastebin.ca/1502188
[21-Jul-2009 11:24:29] <jb> actually
[21-Jul-2009 11:24:32] <jb> its not even using evt.bigipNotifyObjNode
[21-Jul-2009 11:25:07] <jb> evt.component = "Node: %s:%s" % (gethostbyaddr(evt.bigipNotifyObjNode), evt.bigipNotifyObjPort)
[21-Jul-2009 11:25:09] <rmatte> ohhhh, it's using Port
[21-Jul-2009 11:25:17] <jb> that was my orig transform
[21-Jul-2009 11:25:46] <rmatte> what's the value for evt.bigipNotifyObjNode ?
[21-Jul-2009 11:25:55] <jb> its an IP address
[21-Jul-2009 11:26:20] <jb> bigipNotifyObjNode  172.26.103.92
[21-Jul-2009 11:26:25] <jb> that i need to gethostbyaddr()
[21-Jul-2009 11:26:28] <twm1010> transforms really are the big secret to Zenoss, huh.
[21-Jul-2009 11:26:31] <rmatte> and what's evt.bigipNotifyObjPort ?
[21-Jul-2009 11:26:37] <jb> bigipNotifyObjPort  80
[21-Jul-2009 11:26:47] <rmatte> ah, my bad, you need the port displayed I assume?
[21-Jul-2009 11:26:52] <jb> yeah
[21-Jul-2009 11:26:55] <jb> hostname:port
[21-Jul-2009 11:27:02] <rmatte> gotcha, I misunderstood
[21-Jul-2009 11:27:04] <rmatte> let me fix it...
[21-Jul-2009 11:27:14] <twm1010> rmatte: its gotta take FOREVER to move all these mibs into folders
[21-Jul-2009 11:27:31] <rmatte> twm1010: which Mibs?
[21-Jul-2009 11:27:46] <twm1010> the Cisco ones you helped me import last week, 1000+
[21-Jul-2009 11:27:52] <rmatte> ah
[21-Jul-2009 11:27:59] <rmatte> you could have just installed my ZenPack
[21-Jul-2009 11:28:50] <twm1010> that was before ya fixed it
[21-Jul-2009 11:29:09] <rmatte> jb:
[21-Jul-2009 11:29:10] <rmatte> from socket import gethostbyaddr
[21-Jul-2009 11:29:10] <rmatte> evt.summary = "LTM Service Down"
[21-Jul-2009 11:29:10] <rmatte> ip = getattr(evt, 'bigipNotifyObjNode', 'Unknown')
[21-Jul-2009 11:29:10] <rmatte> port = getattr(evt, 'bigipNotifyObjPort', 'Unknown')
[21-Jul-2009 11:29:10] <rmatte> host = gethostbyaddr(ip)
[21-Jul-2009 11:29:12] <rmatte> evt.component = 'Node: %s:%s' % (host,port)
[21-Jul-2009 11:29:14] <rmatte> try that
[21-Jul-2009 11:29:14] <twm1010> I wonder if I install it after the fact if it will move the existing ones or duplicate them
[21-Jul-2009 11:29:18] <jb> will do.
[21-Jul-2009 11:29:18] <jb> thanks
[21-Jul-2009 11:30:12] <rmatte> twm1010: doubt it, since they aren't in the same location, it might duplicate them
[21-Jul-2009 11:30:23] <rmatte> twm1010: you could delete what's there then install the ZenPack
[21-Jul-2009 11:30:59] <twm1010> deleting is taking just as long as "moving" so... just gonna keep a browser up on another machine and move them a little at a time
[21-Jul-2009 11:31:18] <rmatte> have fun
[21-Jul-2009 11:31:27] <rmatte> took me like 30 mins to an hour just moving them
[21-Jul-2009 11:31:38] <twm1010> well, i don't have the fastest box, so...
[21-Jul-2009 11:31:50] <rmatte> omg, I think that SNMP problem just resolved itself
[21-Jul-2009 11:31:52] <twm1010> i'm currently trying to iron out all the devices not responding to SNMP quickly..
[21-Jul-2009 11:31:59] <rmatte> or something I did resolved it finally
[21-Jul-2009 11:32:10] <twm1010> also... having an odd issue with two sever on the same segment as my zenoss box
[21-Jul-2009 11:32:24] <rmatte> what's the issue with them?
[21-Jul-2009 11:33:02] <twm1010> Well, my Zenoss server, and these two Netapp Filers are dual connected to our core switches that have a backplane between them
[21-Jul-2009 11:33:30] <twm1010> ever since i changed my Zenoss server to bond the two interfaces (instead of using just one)... I keep getting ICMP timeouts to the Netapp boxes every so often
[21-Jul-2009 11:38:15] <twm1010> no interface errors, none of my other linux servers setup this way have an issue
[21-Jul-2009 11:38:29] <twm1010> perhaps its got to do with the link aggregation on the filer heads seeing a path to that IP from both nics
[21-Jul-2009 11:41:49] <rmatte> that may be it
[21-Jul-2009 11:43:04] <twm1010> In a bonded interface scenario, should both nics AND the bond interface all share the same MAC?
[21-Jul-2009 11:50:58] <kisielk> twm1010: no
[21-Jul-2009 11:51:23] <etank> in an event transform how do i specify a device name?
[21-Jul-2009 11:52:59] <etank> evt.device i think it is
[21-Jul-2009 11:53:30] <grumpie> has anyone had any issues with customizing reports? I am getting an error when saving a custom report "Site error
[21-Jul-2009 11:53:30] <grumpie> An error was encountered while publishing this resource. The requested resource does not exist."
[21-Jul-2009 11:54:31] <etank> grumpie: i have seen that too.
[21-Jul-2009 11:55:01] <etank> lately when i try to open some reports my cpu spikes to 100% and never comes down until i restart the stack.
[21-Jul-2009 11:55:54] <grumpie> all the reports run just fine. i just need to edit them. but getting error when saving
[21-Jul-2009 12:04:03] <grumpie> the Request Information section reveals "AUTHENTICATED_USER Anonymous User" regardless how i log in
[21-Jul-2009 12:05:35] <twm1010> kisielk: I added primary=eth0 to the options line for the bonding in modprobe.conf, hopefully that gets things going
[21-Jul-2009 12:08:58] <twm1010> if not, perhaps i'll switch to failover instead of loadbalance
[21-Jul-2009 12:14:18] <rmatte> grumpie: are these coded reports?
[21-Jul-2009 12:14:26] <rmatte> grumpie: or are they listed under custom reports?
[21-Jul-2009 12:14:40] <grumpie> they are the builtin reports
[21-Jul-2009 12:16:32] <grumpie> i also tried to create a new report and save it. same errors
[21-Jul-2009 12:23:50] <sacpinball> grumpie - i tried working with the anonymous user before and had the same problem
[21-Jul-2009 12:23:59] <sacpinball> no matter how you login it always says anonymous user
[21-Jul-2009 12:24:03] <rmatte> grumpie: ok, where are you editing them?
[21-Jul-2009 12:25:42] <rmatte> grumpie: the best method is to SSH to the server, become root, then become the zenoss user by typing "su - zenoss", then navigate to $ZENHOME/Products/ZenReports/reports
[21-Jul-2009 12:26:18] <rmatte> grumpie: edit the .rpt file in whichever folder. once you're done the edit execute "python $ZENHOME/Products/ZenReports/ReportLoader.py --force" as the zenoss user
[21-Jul-2009 12:43:12] <grumpie> rmatte: ok. i'll try that out
[21-Jul-2009 12:43:17] <grumpie> thatnks
[21-Jul-2009 12:46:31] <kd4zay> can some one help , i no longer have a windows class , im not sure what happened to it
[21-Jul-2009 12:47:50] <kd4zay> is there a way to add it back?
[21-Jul-2009 12:49:17] <twm1010> kd4zay: Sure
[21-Jul-2009 12:49:35] <twm1010> navigate to /Devices/Servers
[21-Jul-2009 12:49:47] <twm1010> and add it back, however, nuking the class probably nuked some templates too
[21-Jul-2009 12:50:06] <kd4zay> ok
[21-Jul-2009 12:51:03] <rmatte> kd4zay: you could always install Zenoss on a separate box, then make a ZenPack with the Windows class in it, and install the ZenPack on that box
[21-Jul-2009 12:51:05] <kd4zay> only thing i did was install a couple of zenpack not sure if that would have any thing to do with it
[21-Jul-2009 12:51:15] <rmatte> which ZenPacks?
[21-Jul-2009 12:52:16] <kd4zay> ZenPacks.BlakeDrager.fping BlakeDrager Blake Drager 1.0 Yes ZenPacks.PBnJSolutions.Barracuda.Watcher PBnJSolutions Paul Winkeler 1.3 Yes
[21-Jul-2009 12:52:36] <rmatte> never used either, so no idea
[21-Jul-2009 12:52:48] <rmatte> ToKyNET-afk: lunch
[21-Jul-2009 12:52:52] <rmatte> eugh
[21-Jul-2009 12:52:56] <rmatte> afk - lunch
[21-Jul-2009 12:53:02] <rmatte> stupid name auto-complete
[21-Jul-2009 12:53:07] <rmatte>
[21-Jul-2009 12:53:47] <chudler> zenoss/zope METAL/macros are making my head spin at the moment :-/
[21-Jul-2009 12:54:58] <chudler> I never realized before, but I have no idea how <table metal:use-macro="here/viewPerformanceDetail/macros/objectperf" /> expands into a graph
[21-Jul-2009 13:00:06] tripitaka_ is now known as tripitaka
[21-Jul-2009 13:02:49] <rmatte> Are the following included in a clean Zenoss install?:
[21-Jul-2009 13:03:13] <rmatte> /Devices/Power/UPS/APC
[21-Jul-2009 13:03:13] <rmatte> /Devices/Security/Firewall/CheckPoint
[21-Jul-2009 13:03:13] <rmatte> /Devices/Security/Firewall/HP
[21-Jul-2009 13:03:13] <rmatte> /Devices/Security/Firewall/Nokia
[21-Jul-2009 13:03:41] <twm1010> I don't have a /Devices/Security
[21-Jul-2009 13:03:45] <twm1010> stack-install from 2.4.2
[21-Jul-2009 13:03:55] <rmatte> k, how about power?
[21-Jul-2009 13:03:57] <twm1010> I do have the APC one, but i think i put it there.
[21-Jul-2009 13:04:15] <twm1010> let me look at my dev box
[21-Jul-2009 13:04:53] <rmatte> thanks
[21-Jul-2009 13:05:33] <twm1010> Yeah, APC is there on my dev too
[21-Jul-2009 13:05:37] <twm1010> so, it probably came with zenoss
[21-Jul-2009 13:05:39] <rmatte> k cool, thanks
[21-Jul-2009 13:08:03] <twm1010> Ideas what might cause a missing RRD file on a performance page?
[21-Jul-2009 13:08:12] <rmatte> twm1010: wrong OID?
[21-Jul-2009 13:08:14] <twm1010> the datapoints look fine, the OID's answer
[21-Jul-2009 13:08:27] <rmatte> does the OID answer with more than one result?
[21-Jul-2009 13:08:30] <twm1010> perhaps i need to lead oid's awith a .?
[21-Jul-2009 13:08:43] <rmatte> no, it'll automatically put the leading ., or it should
[21-Jul-2009 13:08:58] <rmatte> run a test on the datapoint against a device
[21-Jul-2009 13:09:05] <rmatte> what's the output?
[21-Jul-2009 13:09:49] <twm1010> SNMPv2-SMI::mib-2.33.1.3.3.1.3.1 = INTEGER: 117
[21-Jul-2009 13:09:55] <twm1010> input voltage of a tripp-lite UPS
[21-Jul-2009 13:10:05] <rmatte> hmmm, looks alright
[21-Jul-2009 13:10:17] <rmatte> how long has it been since you added that device?
[21-Jul-2009 13:10:23] <twm1010> about 24 hrs now
[21-Jul-2009 13:10:37] <rmatte> and the RRD file isn't even there?
[21-Jul-2009 13:11:13] <rmatte> like, if you actually go check on the server, the file doesn't exist at all?
[21-Jul-2009 13:12:03] <twm1010> nope, not there
[21-Jul-2009 13:12:12] <rmatte> you created a datapoint under your data source right?
[21-Jul-2009 13:12:20] <rmatte> If you didn't, it won't work
[21-Jul-2009 13:12:23] <twm1010> I see the file for sysUpTime_sysUpTime.rrd coming from the device template
[21-Jul-2009 13:13:07] <twm1010> care to elaborate that last bit?
[21-Jul-2009 13:13:37] <rmatte> If you click on the data source to see the settings (the place where you ran the test from), you'll see a section at the bottom called "DataPoints"
[21-Jul-2009 13:13:51] <rmatte> did you add a datapoint in that section when you configured the data source?
[21-Jul-2009 13:15:00] <twm1010> I don't see that section, honestly.
[21-Jul-2009 13:15:19] <twm1010> I see an aliases section
[21-Jul-2009 13:15:22] <rmatte> you went in to the template, then click on the data source to open the options right?
[21-Jul-2009 13:15:34] <rmatte> Aliases?
[21-Jul-2009 13:15:38] * rmatte scratches his head
[21-Jul-2009 13:16:03] <rmatte> ok, let's walk through, this, go to the templates section for that device, then click on the template
[21-Jul-2009 13:16:10] <twm1010> that's where I'm at
[21-Jul-2009 13:16:26] <rmatte> let me know when you see the Data Sources, Thresholds, and Graph Definitions sections
[21-Jul-2009 13:17:01] <twm1010> Yep, there.
[21-Jul-2009 13:17:13] <rmatte> ok, now under Data Sources, what's listed there?
[21-Jul-2009 13:17:42] <twm1010> three lines
[21-Jul-2009 13:17:50] <rmatte> which one is for the voltage?
[21-Jul-2009 13:17:52] <twm1010>  tl_inputvoltage        1.3.6.1.2.1.33.1.3.3.1.3       SNMP       True
[21-Jul-2009 13:18:03] <twm1010>  tl_outputload        1.3.6.1.2.1.33.1.4.4.1.5       SNMP       True
[21-Jul-2009 13:18:03] <rmatte> ok, click on that one
[21-Jul-2009 13:18:05] <twm1010> k
[21-Jul-2009 13:18:21] <rmatte> now you should see the DataPoints section at the bottom, correct?
[21-Jul-2009 13:20:12] <rmatte> actually weird
[21-Jul-2009 13:20:28] <rmatte> I just logged on to my lab box which is 2.4.2 and they removed that whole section, wtf
[21-Jul-2009 13:20:37] <twm1010> Yeah, then you see what I see
[21-Jul-2009 13:20:52] <rmatte> they must have removed it for everything other than command based datapoints
[21-Jul-2009 13:21:03] <rmatte> it used to always be visible before, weird
[21-Jul-2009 13:21:06] <rmatte> ok, that's definitely not the problem then
[21-Jul-2009 13:21:17] <rmatte> hmmm
[21-Jul-2009 13:21:35] <rmatte> are you getting any debug messages for that device (grey events)?
[21-Jul-2009 13:21:38] <twm1010> I built my template for my Chlorite UPS almost the exact same way
[21-Jul-2009 13:21:45] <twm1010> lemme check
[21-Jul-2009 13:22:02] <rmatte> you need to change the Sev dropdown to debug
[21-Jul-2009 13:22:10] <jmp242> hey, does anyone know how Zenoss applies the template inheritance (is it just at device creation, or does it eventually trickle down)
[21-Jul-2009 13:22:52] <rmatte> jmp242: If a device is in a group that a template is bound to then all sub-objects of that group have the template applied to them
[21-Jul-2009 13:22:58] <twm1010> none that i can see
[21-Jul-2009 13:23:34] <jmp242> hmmm - I take it this doesn't take effect if you apply it to /Devices though
[21-Jul-2009 13:23:34] <rmatte> jmp242: If I move a device from /Network/Cisco to /Server/Windows all Windows templates, plugins, etc... get applied to that device, and the Cisco stuff gets removed
[21-Jul-2009 13:23:47] <jmp242> (or it doesn't seem to in my testing)
[21-Jul-2009 13:24:12] <rmatte> jmp242: yeh, a template applied to /Devices should trickle down, unless a specific group already has a locally applied template with the same name
[21-Jul-2009 13:24:37] <jmp242> mmm - so if I'm using the b_fping template as a local copy on my test device (single device)
[21-Jul-2009 13:24:39] <rmatte> jmp242: there are inherited templates, and local templates
[21-Jul-2009 13:25:01] <jmp242> it could prevent the b_fping applied from the original template at /Devices from working
[21-Jul-2009 13:25:12] <twm1010> No, it shouldn't prevent it.
[21-Jul-2009 13:25:25] <rmatte> jmp242: you can't have both, it's either local or inherited, not a mixture of the two
[21-Jul-2009 13:25:26] <twm1010> But just because you put it at /Devices, doesn't necessarily mean it binds it all the way down.
[21-Jul-2009 13:25:56] <rmatte> yeh, you need to play with the binding at each level until it works properly
[21-Jul-2009 13:26:07] <jmp242> Hmmm
[21-Jul-2009 13:26:08] <jmp242> ok...
[21-Jul-2009 13:26:43] <twm1010> rmatte: I'm going to try and just add another datasource with the same OID, but a different name
[21-Jul-2009 13:28:42] <twm1010> the template doesn't HAVE to be in the same device class right? I placed the template at /Dev/Power/UPS and then bound it at ../UPS/Tripplite, that shouldn't matter should it?
[21-Jul-2009 13:29:53] <jmp242> I see, for some reason it worked everywhere except my second test location /Devices/Server/Linux
[21-Jul-2009 13:29:57] <jmp242> ... that's life
[21-Jul-2009 13:29:59] <jmp242> now it's working
[21-Jul-2009 13:30:17] <twm1010> It could be that the collector hadn't picked up the changes yet.
[21-Jul-2009 13:33:04] <twm1010> man.. no dice :|
[21-Jul-2009 13:37:26] <rmatte> twm1010: no, that should work
[21-Jul-2009 13:38:10] <rmatte> twm1010: no debug messages?
[21-Jul-2009 13:38:18] <twm1010> nope, no debug msgs
[21-Jul-2009 13:38:39] <twm1010> could it be the names I gave my datasources?
[21-Jul-2009 13:38:44] <twm1010> the _ throwing things off?
[21-Jul-2009 13:38:56] <twm1010> ah, but i just created another simply called inputvoltage, no dice
[21-Jul-2009 13:39:19] <rmatte> it should show up as inputvoltage_inputvoltage when you put it in the graph
[21-Jul-2009 13:39:50] <twm1010> indeed, thats what it shows as missing, the RRD files simply aren't there
[21-Jul-2009 13:39:59] <rmatte> it also takes 3 polling cycles before it starts graphing
[21-Jul-2009 13:40:13] <twm1010> i wonder if... these devices are snmpv1 only?
[21-Jul-2009 13:40:16] <rmatte> not sure if it takes 3 polling cycles before it actually creates the RRD, but I'd imagine it would create it on the first cycle
[21-Jul-2009 13:40:27] <twm1010> in the past it's taken 5-10 mins at most
[21-Jul-2009 13:40:29] <rmatte> I doubt it, your test wouldn't have worked if that was the case
[21-Jul-2009 13:40:31] <twm1010> its been 24 hours
[21-Jul-2009 13:40:35] <rmatte> but you could try with snmpv1
[21-Jul-2009 13:41:21] <twm1010> eh, device answers to both from snmpget..
[21-Jul-2009 13:41:43] <rmatte> mrayzenoss: is there a good guide somewhere on how to add actual files, like scripts and stuff to ZenPacks, and have them install to the Products directory?
[21-Jul-2009 13:42:06] <rmatte> twm1010: yeh, most devices support v2 these days
[21-Jul-2009 13:42:08] <mrayzenoss> rmatte: that's actually discouraged, the old .zip ZenPacks used to do that
[21-Jul-2009 13:42:23] <mrayzenoss> but you can link to the scripts in the ZenPack directory
[21-Jul-2009 13:42:24] <rmatte> mrayzenoss: so how do you do it now then?
[21-Jul-2009 13:42:36] <mrayzenoss> 1 sec
[21-Jul-2009 13:42:39] <rmatte> thanks
[21-Jul-2009 13:42:40] <twm1010> Odd.. nothing special about this device class, no zprops set, just like the class next to it that works fine.
[21-Jul-2009 13:43:28] <rmatte> twm1010: I'm out of ideas
[21-Jul-2009 13:43:30] <mrayzenoss> rmatte: take a look at line 28 trac-zenpacks/browser/zenpacks/ZenPacks.community.VMwareEsx/ZenPacks/community/VMwareEsx/objects/objects.xml
[21-Jul-2009 13:43:49] <mrayzenoss> this refers to the location of the ZenPack, then into the libexec directory within it
[21-Jul-2009 13:43:57] <twm1010> @mrayzenoss: any ideas why a clearly working SNMP datasource is not creating an RRD file?
[21-Jul-2009 13:44:31] <rmatte> mrayzenoss: right, so you'd just use a line like that as the command for the command based datasource then?
[21-Jul-2009 13:44:43] <mrayzenoss> twm1010: usual culprits are permissions, remodeling/pushing changes and restarting zenhub and zenperfsnmp
[21-Jul-2009 13:44:52] <mrayzenoss> rmatte: yes
[21-Jul-2009 13:44:55] <rmatte> mrayzenoss: which location would I drop the files in the ZenPack then?
[21-Jul-2009 13:45:21] <mrayzenoss> ZenPacks/community/YourZenPack/libexec/
[21-Jul-2009 13:45:32] <rmatte> perfect, thanks
[21-Jul-2009 13:45:41] <twm1010> i'll play a little more, lest i dent this desk with my head
[21-Jul-2009 13:46:10] <mrayzenoss> twm1010: does anything show up when you run zenperfsnmp -v10 run ?
[21-Jul-2009 13:49:26] <twm1010> no failures
[21-Jul-2009 13:49:34] <twm1010> that i can see in the zenperf log
[21-Jul-2009 13:52:09] <mrayzenoss> twm1010: had wrong syntax, it's 'zenperfsnmp run -v10 yourhost'
[21-Jul-2009 13:54:06] <twm1010> hrmm still doesn't seem like its focusing on that one host
[21-Jul-2009 14:04:55] <blocky> would somebody give me a hand with snmpd
[21-Jul-2009 14:05:08] <blocky> the log file shows it receiving requests but it doesnt say why its not replying'
[21-Jul-2009 14:05:28] <twm1010> is iptables running?
[21-Jul-2009 14:06:08] <blocky> i dont see the module loaded
[21-Jul-2009 14:06:17] <blocky> the policy is ACCEPT on each table tho
[21-Jul-2009 14:06:35] <blocky> oh nm it is
[21-Jul-2009 14:06:42] <twm1010> try stopping the service entirely, and see if your problem goes away, assuming other devices are working without issue
[21-Jul-2009 14:09:21] <jmp242> So I've installed the zenpack NTPMonitor in Zenoss 2.3.3, and when I click on the command it gives me in the template (to see what it looks like) <NtpMonitorDataSource at broken>
[21-Jul-2009 14:09:39] <jmp242> do I need to do something else to use this Zenpack (and examine it)?
[21-Jul-2009 14:09:50] <mrayzenoss> jmp242: did you restart zope after installation?
[21-Jul-2009 14:10:24] <jmp242> ahh no. So I need to restart Zenoss then
[21-Jul-2009 14:10:39] <jmp242> do the Zenpacks say somewhere what we need to restart to start using them?
[21-Jul-2009 14:10:56] <jmp242> I.E. am I just being dense
[21-Jul-2009 14:11:06] <mrayzenoss> jmp242: I'm not exactly sure what the trigger is, but sometimes when you install a ZenPack it says to restart zope
[21-Jul-2009 14:11:22] <mrayzenoss> but I've found that it's erratic, so to be safe you should always restart zope
[21-Jul-2009 14:11:49] <mrayzenoss> and we don't make the system restart zope automatically, because it would kick you out of the UI when you install a ZenPack
[21-Jul-2009 14:12:23] <mrayzenoss> if you're adding additional functionality to an existing device, you'll need to restart zenhub as well
[21-Jul-2009 14:12:54] <mrayzenoss> so in a nutshell, restarting zope every time is good, everything is safest
[21-Jul-2009 14:15:44] <blocky> what is the first step to try when a graph does not appear
[21-Jul-2009 14:17:36] <jmp242> ok, I've restarted Zenoss and I do see the graph, still NAN, but it's probably just waiting to pull stuff up
[21-Jul-2009 14:17:48] <rmatte> it'll be "nan" for 3 polling cycles
[21-Jul-2009 14:17:49] <blocky> on the perf page only my Load Average graph appears and it says missing RRD files
[21-Jul-2009 14:17:51] <rmatte> that's normal
[21-Jul-2009 14:17:58] <rmatte> the fact that the graph is showing is a good sign
[21-Jul-2009 14:17:59] <blocky> ive just added a new device and run the modeller on it
[21-Jul-2009 14:18:02] <blocky> and i know the oids are available
[21-Jul-2009 14:18:16] <rmatte> blocky: Linux server?
[21-Jul-2009 14:19:25] <rmatte> blocky: Make sure that the community string is set right in Zenoss for that device, I've noticed that devices put in to the Linux device group tend to create a local version of the SNMP string
[21-Jul-2009 14:19:28] <jmp242> note - in the template itself, I now get some odd error trying to pull up the command
[21-Jul-2009 14:19:33] <rmatte> go in to zProperties, scroll to the bottom, and delete it
[21-Jul-2009 14:21:28] <rmatte> blocky: zProperties for the device, scroll to the bottom of the page, under Delete Local Property, delete the zSnmpCommunity, either that or set the right community string under zSnmpCommunity in zProperties
[21-Jul-2009 14:21:42] <rmatte> blocky: then remodel the device and you should be good
[21-Jul-2009 14:22:05] <jmp242> working now
[21-Jul-2009 14:28:47] <rmatte> good
[21-Jul-2009 14:45:00] <kd4zay> rmatte i am now getting information from my servers but still have one error poping up for all the servers > Could not read the Windows event log (NT_STATUS_ACCESS_DENIED). Check your username/password settings and verify network connect
[21-Jul-2009 14:46:05] <rmatte> mrayzenoss: that trick for removing zenpacks from the list without removing the content is great for reorganizing ZenPacks (splitting larger ZenPacks in to smaller ones)
[21-Jul-2009 14:50:03] <rmatte> kd4zay: it's trying to monitor the event log using WMI, if you don't want it doing this go in to zProperties for the class and change zWinEventLog to False
[21-Jul-2009 14:50:33] <rmatte> If you want it doing this you need to enable WMI on the servers, then specify an admin account in zWinUser and zWinPassword
[21-Jul-2009 14:50:38] <kd4zay> actually i do want to monitor the eventlog
[21-Jul-2009 14:50:58] <kd4zay> ahh ok
[21-Jul-2009 14:51:04] <rmatte> event log messages will come in as snmp traps, so you can monitor it that way if you want, just install the Windows Mibs
[21-Jul-2009 14:51:09] <kd4zay> in the actual service itself?
[21-Jul-2009 14:51:42] <rmatte> no, you enable the service, create an account on the servers with admin access to the boxes, then put the zWinUser and password in zProperties for the class
[21-Jul-2009 14:52:10] <kd4zay> ahh ok , thanks
[21-Jul-2009 14:52:21] <jmp242> rmatte: how are event logs coming in as SNMP Traps? I've not seen that - do you mean using something like epilog
[21-Jul-2009 14:53:26] <rmatte> no, if you setup the windows servers to trap back to your Zenoss box it'll send certain event log messages (any critical messages) as traps
[21-Jul-2009 14:53:40] <rmatte> you need to install WINDOWS-NT-PERFORMANCE.mib
[21-Jul-2009 14:54:04] <rmatte> then the traps will come in looking like snmp trap enterprises.8.4.23.45.3
[21-Jul-2009 14:54:14] <jmp242> oh - I didn't actually know that. I'm personally happy with WMI, but that's me
[21-Jul-2009 14:54:30] <rmatte> and if you go to the details tab for the trap after it has come in, you can get an idea of what the trap came in for an apply transforms to each trap as they come in
[21-Jul-2009 14:54:35] <rmatte> k, well then use WMI
[21-Jul-2009 14:54:38] <rmatte>
[21-Jul-2009 14:55:19] <twm1010> Yeah, the WMI logging is very nice
[21-Jul-2009 14:55:53] <jmp242> I'm also quite happy with the PerfMon Zenpack now that I'm playing with it... heck there's a bunch of fun Zenpacks since I last seriously installed any
[21-Jul-2009 14:55:57] <rmatte> yeh, you're going to have to tune the heck out of it though, you're going to get a lot of useless messages coming in
[21-Jul-2009 14:56:20] <rmatte> the PerfMon zenpack is alright, it has it's flaws
[21-Jul-2009 14:56:29] <rmatte> I'm going to try to improve upon it
[21-Jul-2009 14:56:46] <rmatte> I'm using SNMP for performance monitoring on our windows boxes
[21-Jul-2009 14:59:12] <jmp242> Mmmm - I liked the WMI Zenpack... still chugging along well for us.
[21-Jul-2009 14:59:18] <jb> i use SNMP and WMI
[21-Jul-2009 14:59:24] <jb> most of my custom stuff is all WMI
[21-Jul-2009 14:59:26] <jb> perfmon counters
[21-Jul-2009 15:01:14] <rmatte> I just don't like the fact that when a command fails with that WMI pack, it spits the username and password out in the command failed event
[21-Jul-2009 15:01:42] <rmatte> I modified the script to pass the username and password from Zope, but it makes the script overly intensive to run over a large number of servers
[21-Jul-2009 15:01:52] <jmp242> I suppose that could be changed - we don't really care, the only people who can access Zenoss are people who have domain admin accounts anyway.
[21-Jul-2009 15:01:55] <rmatte> A collector plugin would be needed, which I'm slowly learning how to write
[21-Jul-2009 15:11:46] <johns55> zenoss.com is down
[21-Jul-2009 15:12:09] <chudler> works for me
[21-Jul-2009 15:13:15] <johns55> This website is temporarily unavailable, please try again later.
[21-Jul-2009 15:13:22] <jb> its not down..
[21-Jul-2009 15:13:26] <johns55> what's your DNS resolving to?
[21-Jul-2009 15:13:41] <chudler> 204.13.248.125
[21-Jul-2009 15:13:44] <johns55> 64.202.189.170
[21-Jul-2009 15:13:45] <mrayzenoss> we're having some website weirdness right now
[21-Jul-2009 15:13:46] <jb> www.zenoss.com has address 98.129.21.196
[21-Jul-2009 15:14:04] <johns55> damn u guys got some nation-wide load balancing going on?
[21-Jul-2009 15:14:29] <johns55> from the looks of it the site's being hosted by godaddy
[21-Jul-2009 15:14:55] <jb> (or rackspace)
[21-Jul-2009 15:15:01] <mrayzenoss> RackSpace
[21-Jul-2009 15:15:10] <mrayzenoss> but yes, there is some sort of weirdness going on
[21-Jul-2009 15:15:50] <chudler> 64.202.189.170 does respond with "This website is temporarily unavailable...", although it is a 200 message.
[21-Jul-2009 15:16:17] <johns55> a whois on that IP resolves to a godaddy IP 64.202.189.170
[21-Jul-2009 15:16:46] <johns55> trying to access 98.129.21.196 from my end asks me for a pass
[21-Jul-2009 15:17:01] <johns55> might be some sort of geo load balancing going on
[21-Jul-2009 15:21:11] <johns55> does anyone know what the best upgrade path is if I have version 2.1.2?
[21-Jul-2009 15:21:36] <johns55> Do I really need to do that 5 step upgrade path where I have to jump each version level?
[21-Jul-2009 15:21:42] <johns55> or can I jump straight to 2.4?
[21-Jul-2009 15:22:53] <mrayzenoss> johns55: nope, you'll have to go 2.1.2->2.1.3->2.2.4->2.3.3->2.4.2 http://www.zenoss.com/community/docs/installation-guide/2.4.2/ch08s01.html#d4e866
[21-Jul-2009 15:23:17] <johns55> oh man thats what I'm afraid of
[21-Jul-2009 15:23:36] <mrayzenoss> yeah, it's kinda far back
[21-Jul-2009 15:23:46] <johns55> I was looking at your source forge site and the oldest tar ball you have is 2.3? Where are the older 2.1.3 stored?
[21-Jul-2009 15:24:06] <mrayzenoss> just a sec
[21-Jul-2009 15:27:03] <mrayzenoss> SourceForge recently moved their stuff around, we had an Archive for them somewhere, we're tracking it down
[21-Jul-2009 15:27:20] <johns55> ok cool thanks for your help
[21-Jul-2009 15:29:30] <mrayzenoss> johns55: here's an archive we keep: http://dev.zenoss.org/downloads/zenoss/
[21-Jul-2009 15:29:56] <johns55> perfect
[21-Jul-2009 15:30:11] <johns55> I haven't done an upgrade before, is it safe? I initially installed it from source
[21-Jul-2009 15:30:42] <mrayzenoss> to be honest, unless you have a compelling reason not to, I'd install clean with 2.4.2 and migrate your devices
[21-Jul-2009 15:31:09] <mrayzenoss> QA tests back 2 releases for upgrades, so ostensibly it should work
[21-Jul-2009 15:32:08] <mrayzenoss> I wouldn't upgrade a production server, I'd clone it and test it with a mirror
[21-Jul-2009 15:32:37] <johns55> hrm that would be tough in my case since we're limited in test servers and cloning mechanisms
[21-Jul-2009 15:33:29] <johns55> migrating devices wouldnt be a bad idea however I have some custom scripts and plugins and it would take awhile to move over
[21-Jul-2009 15:35:00] <jb> yeah.. that upgrade will probably be risky
[21-Jul-2009 15:35:11] <jb> it's really not good to get that far behind
[21-Jul-2009 15:35:27] <johns55> heh it's been working so well I never thought of upgrading
[21-Jul-2009 15:35:34] <mrayzenoss> upgrades are getting better, but 2.1 is before we had a QA team
[21-Jul-2009 15:36:09] <johns55> I think I'll just run the backup script, tarball the zenoss dir, and do a dump of the mysql database
[21-Jul-2009 15:36:16] <johns55> then do the upgrade a step at a time
[21-Jul-2009 15:36:37] <johns55> if anything I'll just perform a restore
[21-Jul-2009 15:37:22] <johns55> ok guys i gotta go thanks for your help and advice
[21-Jul-2009 15:43:44] <rmatte> johns55: package all your custom stuff in to a zenpack
[21-Jul-2009 15:44:04] <rmatte> johns55: the upgrade process for source is pretty brutal compared to something like stack
[21-Jul-2009 15:44:28] <blocky> how long should it be before the rrd files are created
[21-Jul-2009 15:45:12] <twm1010> no more than 5-6 minutes usually
[21-Jul-2009 15:45:25] <twm1010> but i have a stubborn template right now that doesn't create jack diddly, for no apparent reason
[21-Jul-2009 15:45:41] <blocky> what does it mean when the graph doesnt appear at all
[21-Jul-2009 15:45:49] <blocky> my load average says missing rrd files
[21-Jul-2009 15:46:04] <blocky> but the other perf graphs, and the ethernetCsmacd have no graph at all
[21-Jul-2009 15:46:19] <rmatte> blocky: did you check any of the stuff I mentioned before?
[21-Jul-2009 15:46:27] <blocky> yea
[21-Jul-2009 15:46:40] <rmatte> so you're 100% positive that the community string is set properly then
[21-Jul-2009 15:46:50] <blocky> i tried deleting the local community property
[21-Jul-2009 15:47:03] <blocky> i believe so, but is there a way i can view the output of the modeller to make sure?
[21-Jul-2009 15:47:07] <rmatte> and did you make sure that after doing it, the right string was populated to the field?
[21-Jul-2009 15:47:07] <blocky> my snmp logs are less than helpful
[21-Jul-2009 15:47:25] <rmatte> go to the device page and do a Manage -> Push Changes
[21-Jul-2009 15:47:29] <rmatte> to push the changes to the collector
[21-Jul-2009 15:47:30] <twm1010> Did you stop iptables to see if that was interfering?
[21-Jul-2009 15:47:51] <rmatte> you should also make sure that the snmp config on the server allows for all access
[21-Jul-2009 15:47:54] <blocky> snmpwalk is working now, iptables is clear
[21-Jul-2009 15:48:05] <twm1010> if snmpwalk is working, then you're good
[21-Jul-2009 15:48:19] <blocky> do i want to delete zSnmpCommunity and zSnmpCommunities?
[21-Jul-2009 15:48:20] <twm1010> unless the snmpd.conf has limited your zenoss host to certain OIDs
[21-Jul-2009 15:48:23] <rmatte> snmpwalk is working manually, or from the commands menu in Zenoss?
[21-Jul-2009 15:48:27] <jb> unless you misconfigured snmpd
[21-Jul-2009 15:48:28] <blocky> manually
[21-Jul-2009 15:48:35] <jb> wc -l your snmpwalk
[21-Jul-2009 15:48:38] <rmatte> try it from the commands meny
[21-Jul-2009 15:48:40] <jb> how many lines does it return?
[21-Jul-2009 15:48:45] <rmatte> menu*
[21-Jul-2009 15:49:04] <rmatte> go to the device page, dropdown menu, commands, snmpwalk
[21-Jul-2009 15:49:10] <blocky> should i delete both of those local copies
[21-Jul-2009 15:49:10] <rmatte> make sure that goes through for starters
[21-Jul-2009 15:49:16] <blocky> 5361 lines
[21-Jul-2009 15:49:20] <rmatte> what's the second local copy?
[21-Jul-2009 15:49:27] <blocky> zSnmpCommunity and zSnmpCommunities
[21-Jul-2009 15:49:36] <rmatte> yeh, remove both local properties
[21-Jul-2009 15:49:42] <rmatte> zSnmpCommunity is the main one
[21-Jul-2009 15:49:47] <rmatte> but you should just remove both
[21-Jul-2009 15:49:59] <rmatte> nothing should be local in that regards unless needed
[21-Jul-2009 15:50:03] <blocky> okay they are not set to the correct string now
[21-Jul-2009 15:50:10] <twm1010> zSnmpCommunities is usually only set at /Devices, and is used during Discovery, if I am correct
[21-Jul-2009 15:50:10] <blocky> where do i correct that, in the device template?
[21-Jul-2009 15:50:11] <rmatte> that's your problem then
[21-Jul-2009 15:50:16] <rmatte> change the string in /Devices
[21-Jul-2009 15:50:29] <rmatte> in the templates section of the root of /Devices
[21-Jul-2009 15:50:37] <rmatte> erm, zProperties section
[21-Jul-2009 15:50:38] <rmatte> rather
[21-Jul-2009 15:50:50] <rmatte> (sorry, I'm working with templates right now hehe)
[21-Jul-2009 15:50:54] <blocky> np
[21-Jul-2009 15:51:23] <blocky> im trying the modeller again
[21-Jul-2009 15:51:26] <blocky> should i push changes first
[21-Jul-2009 15:51:42] <rmatte> yes
[21-Jul-2009 15:52:23] <blocky> snmpwalk worked from the menu
[21-Jul-2009 15:52:29] <rmatte> good sign
[21-Jul-2009 15:52:33] <rmatte> try a model
[21-Jul-2009 15:52:51] <rmatte> pastebin the model output too so we can see it
[21-Jul-2009 15:53:36] <blocky> there's a few metrics, ram and swap etc, which were blocked due to zFileSystemMapIgnoreTypes
[21-Jul-2009 15:53:55] <rmatte> yeh, that's normal
[21-Jul-2009 15:54:09] <rmatte> pastebin the output, I'm more interested in other possible messages
[21-Jul-2009 15:54:20] <blocky> okay hold on
[21-Jul-2009 15:54:27] <rmatte> k
[21-Jul-2009 15:55:58] <blocky> http://pastebin.com/d164d796d
[21-Jul-2009 15:56:30] <rmatte> checking...
[21-Jul-2009 15:57:50] <rmatte> yeh, that looks good
[21-Jul-2009 15:58:02] <rmatte> the performance info should populate now at some point
[21-Jul-2009 15:58:22] <rmatte> "Changes in configuration applied" is a good sign, because it indicates that it picked up more info than the last time you modeled
[21-Jul-2009 16:01:24] <blocky> so far i noticed the filesystem info populated where it wasnt befoer
[21-Jul-2009 16:01:35] <rmatte> yeh, that's a good sign
[21-Jul-2009 16:01:42] <rmatte> check the perf tab, you should see all graphs
[21-Jul-2009 16:01:44] <blocky> ill give it a few more minutes and see if the rrds/graphs get created
[21-Jul-2009 16:01:47] <rmatte> they may have "nan" values
[21-Jul-2009 16:01:52] <blocky> not yet
[21-Jul-2009 16:02:09] <rmatte> ah, it needs to hit a polling cycle first, so give it a bit of time
[21-Jul-2009 16:02:26] <blocky> ok cool
[21-Jul-2009 16:02:37] <blocky> once i get this working i need to figure out how HP's snmp agent works
[21-Jul-2009 16:03:01] <grumpie> can the Availability Report report on specific times of the day, rather than just whole days?
[21-Jul-2009 16:06:40] <rmatte> grumpie: it's availability between the dates specified, but there's no way to break it down by day
[21-Jul-2009 16:06:56] <rmatte> or by parts of days rather
[21-Jul-2009 16:07:37] <blocky> still no graphs here rmatte
[21-Jul-2009 16:07:42] <grumpie> ok. thanks
[21-Jul-2009 16:07:49] <blocky> load average is appearing with the rrds missing
[21-Jul-2009 16:07:53] <blocky> but its the only one
[21-Jul-2009 16:08:59] <rmatte> blocky: push changes to the collector, also, check to make sure the RRD files are actually still missing
[21-Jul-2009 16:09:12] <rmatte> $ZENHOME/perf
[21-Jul-2009 16:10:06] <blocky> there's no folder at all for that device
[21-Jul-2009 16:11:31] <rmatte> that could be an issue
[21-Jul-2009 16:12:13] <rmatte> mrayzenoss: any idea what would keep a performance folder from being created for a device?
[21-Jul-2009 16:12:20] <rmatte> blocky: was it under a different name initially?
[21-Jul-2009 16:12:39] <blocky> well not exactly
[21-Jul-2009 16:12:43] <rmatte> blocky: is this a new Zenoss install (if so, how did you install it?)
[21-Jul-2009 16:13:07] <blocky> i had the same device modelled previously under a different name but the same ip
[21-Jul-2009 16:13:30] <rmatte> ok, and is there a performance folder left over from that device?
[21-Jul-2009 16:13:31] <blocky> i just deleted the line in /etc/hosts
[21-Jul-2009 16:13:44] <blocky> yep
[21-Jul-2009 16:13:49] <rmatte> delete that folder
[21-Jul-2009 16:13:57] <blocky> done
[21-Jul-2009 16:14:04] <blocky> remodel?
[21-Jul-2009 16:14:06] <rmatte> now remodel the device
[21-Jul-2009 16:14:11] <rmatte> and see if the folder gets created
[21-Jul-2009 16:14:49] <blocky> if it looked up the host i was using now to an ip, and then backwards referenced that ip to the first host it was associated with and stopped there
[21-Jul-2009 16:15:17] <rmatte> that may be what the issue was
[21-Jul-2009 16:15:22] <blocky> no folder was created
[21-Jul-2009 16:15:39] <rmatte> there's no perf folder for the IP as well or anything I assume?
[21-Jul-2009 16:15:45] <blocky> nope
[21-Jul-2009 16:15:51] <rmatte> try removing and re-adding the device
[21-Jul-2009 16:16:01] <rmatte> make sure the snmp info is correct when adding it
[21-Jul-2009 16:16:11] <blocky> oh lemme delete one other folder as well
[21-Jul-2009 16:16:16] <rmatte> lol
[21-Jul-2009 16:16:35] <rmatte> delete that folder, remodel, if that doesn't work you might consider restarting zenoss and trying again
[21-Jul-2009 16:16:36] <blocky> i wish zenoss could abstract device names from their host
[21-Jul-2009 16:17:24] <blocky> okay im blowing out all the devices that had anything to do with that ip
[21-Jul-2009 16:17:32] <rmatte> hehe
[21-Jul-2009 16:17:39] <rmatte> there were more than 1?
[21-Jul-2009 16:17:58] <blocky> okay it just deleted the wrong device :|
[21-Jul-2009 16:18:06] <rmatte> yikes
[21-Jul-2009 16:18:12] <rmatte> how'd you manage that?
[21-Jul-2009 16:18:27] <blocky> i checked the box and selected delete from the menu
[21-Jul-2009 16:18:58] <blocky> the notification icon in the corner said the one i picked was deleted, but the device list shows it still tehre and another one gone
[21-Jul-2009 16:19:15] <rmatte> sounds like you've got something pretty screwed up there
[21-Jul-2009 16:19:23] <rmatte> are you familliar with zendmd?
[21-Jul-2009 16:19:33] <blocky> ive tried to play with it a bit
[21-Jul-2009 16:19:38] <rmatte> ok, go in to it
[21-Jul-2009 16:19:38] <blocky> far from familiar
[21-Jul-2009 16:19:43] <rmatte> type: reindex()
[21-Jul-2009 16:19:47] <rmatte> then: commit()
[21-Jul-2009 16:19:51] <rmatte> and exit
[21-Jul-2009 16:19:58] <rmatte> it sounds like your zope db is out of wack
[21-Jul-2009 16:20:21] <rmatte> reindexing it will probably solve the problem
[21-Jul-2009 16:20:25] <blocky> okay
[21-Jul-2009 16:20:34] <blocky> i have been messing with it a fair bit
[21-Jul-2009 16:20:45] <blocky> i spent some time trying to get a sort of hybrid report that shows graphs and tables
[21-Jul-2009 16:20:54] <blocky> i reindexed
[21-Jul-2009 16:20:57] <blocky> now restart zenoss?
[21-Jul-2009 16:21:17] <rmatte> you reindexed, and commited right?
[21-Jul-2009 16:21:21] <blocky> yes sir
[21-Jul-2009 16:21:26] <rmatte> ok, yeh, restart zenoss now
[21-Jul-2009 16:21:32] <rmatte> then re-add whatever device got deleted
[21-Jul-2009 16:21:40] <rmatte> and then try to delete the one you tried to remove in the first place
[21-Jul-2009 16:21:51] <rmatte> or actually, try to delete that one before adding the other one
[21-Jul-2009 16:22:59] <blocky> ive got another of these bandaids in place with 3 devices redirecting to the same ip
[21-Jul-2009 16:23:03] <blocky> but its an entirely separate ip
[21-Jul-2009 16:23:15] <rmatte> lol
[21-Jul-2009 16:23:26] <rmatte> bandaids like that are bad news with Zenoss
[21-Jul-2009 16:24:12] <blocky> okay it deleted properly that time
[21-Jul-2009 16:24:26] <blocky> im gonna try to readd my device
[21-Jul-2009 16:24:46] <rmatte> k
[21-Jul-2009 16:24:53] <blocky> why doesn't it appear in the device list right away
[21-Jul-2009 16:25:09] <rmatte> ?
[21-Jul-2009 16:25:12] <rmatte> it should
[21-Jul-2009 16:25:17] <blocky> not yet
[21-Jul-2009 16:25:24] <rmatte> well, let it run through
[21-Jul-2009 16:26:33] <blocky> okay the graphs look the same as before
[21-Jul-2009 16:26:37] <blocky> do i have to run the modeller manually?
[21-Jul-2009 16:26:50] <rmatte> well, do you see the hard drive info?
[21-Jul-2009 16:27:02] <blocky> yea
[21-Jul-2009 16:27:18] <blocky> still no perf folder
[21-Jul-2009 16:27:24] <rmatte> I'm stumped then
[21-Jul-2009 16:27:34] <rmatte> you'd need to talk to someone like cluther for this one
[21-Jul-2009 16:27:42] <rmatte> I've never seen anything like that before
[21-Jul-2009 16:28:33] <blocky> how do i blow out mysql
[21-Jul-2009 16:28:39] <blocky> ill reinstall zenoss
[21-Jul-2009 16:29:01] <rmatte> did you install using stack installer?
[21-Jul-2009 16:30:06] <blocky> im not sure
[21-Jul-2009 16:30:17] <blocky> yes
[21-Jul-2009 16:30:18] <rmatte> was it a .bin file that you executed?
[21-Jul-2009 16:30:24] <blocky> yea i did
[21-Jul-2009 16:30:35] <rmatte> ok, so you want to completely uninstall Zenoss?
[21-Jul-2009 16:30:36] <blocky> 2.4.1
[21-Jul-2009 16:30:46] <blocky> yea
[21-Jul-2009 16:30:53] <rmatte> execute /usr/local/zenoss/uninstall
[21-Jul-2009 16:31:14] <rmatte> that'll wipe out all the devices and everything though
[21-Jul-2009 16:31:27] <blocky> thats fine
[21-Jul-2009 16:31:27] <rmatte> and you might as well download and install 2.4.2 now
[21-Jul-2009 16:31:37] <rmatte> just to save an upgrade later
[21-Jul-2009 16:31:43] <blocky> im a noob to sql
[21-Jul-2009 16:31:55] <rmatte> Zenoss only stores events in sql
[21-Jul-2009 16:32:00] <rmatte> it uses Zope for everything else
[21-Jul-2009 16:32:06] <blocky> k
[21-Jul-2009 16:32:16] <blocky> does the uninstaller remove those events?
[21-Jul-2009 16:32:30] <rmatte> yes
[21-Jul-2009 16:32:37] <rmatte> it removes MySQL
[21-Jul-2009 16:32:48] <blocky> is the tar.gz a zipped up bin
[21-Jul-2009 16:32:56] <rmatte> no
[21-Jul-2009 16:33:01] <rmatte> the .tar.gz is source
[21-Jul-2009 16:33:04] <blocky> oh
[21-Jul-2009 16:33:06] <blocky> screw that
[21-Jul-2009 16:33:22] <rmatte> make sure you get the .bin
[21-Jul-2009 16:33:30] <blocky> trying to find one on google
[21-Jul-2009 16:33:41] <rmatte> what Linux are you running?
[21-Jul-2009 16:33:50] <blocky> centos 5
[21-Jul-2009 16:33:56] <blocky> get it
[21-Jul-2009 16:33:59] <blocky> err got it
[21-Jul-2009 16:34:02] <blocky> not get centos
[21-Jul-2009 16:34:05] <rmatte> you're sure you didn't install the rpm?
[21-Jul-2009 16:34:16] <blocky> pretty sure
[21-Jul-2009 16:34:18] <blocky> what if i did though
[21-Jul-2009 16:34:38] <rmatte> it'd put the files in a different spot
[21-Jul-2009 16:34:53] <rmatte> if you have /usr/local/zenoss/zenoss then you used the bin installer
[21-Jul-2009 16:35:09] <blocky> is this not correct syntax: find / -name *zenoss*
[21-Jul-2009 16:35:35] <rmatte> "*zenoss*"
[21-Jul-2009 16:35:43] <rmatte> you need the quotes when using the wildcards
[21-Jul-2009 16:35:48] <blocky> k cool
[21-Jul-2009 16:35:51] <blocky> yea im sure it was the bin
[21-Jul-2009 16:35:59] <blocky> 16 mins left on the new version
[21-Jul-2009 16:36:07] <rmatte> you can also just do "updatedb" then "locate zenoss"
[21-Jul-2009 16:36:12] <blocky> how late will you be on today?
[21-Jul-2009 16:36:21] <rmatte> I'm leaving in 1 minute
[21-Jul-2009 16:36:24] <rmatte> just finished my shift
[21-Jul-2009 16:36:29] <blocky> ah
[21-Jul-2009 16:36:33] <blocky> well thanks for the help
[21-Jul-2009 16:36:33] <rmatte> I'll be around tomorrow though
[21-Jul-2009 16:36:36] <rmatte> no problem
[21-Jul-2009 16:36:37] <blocky> hopefully this works a bit better
[21-Jul-2009 16:36:40] <blocky> you east coast?
[21-Jul-2009 16:36:42] <rmatte> hope so
[21-Jul-2009 16:36:46] <rmatte> east coast, yeh
[21-Jul-2009 16:36:52] <blocky> im in vancouver
[21-Jul-2009 16:36:56] <rmatte> I'm in Ottawa
[21-Jul-2009 16:37:33] <blocky> you might enjoy this
[21-Jul-2009 16:37:35] <blocky> http://www.ottawacitizen.com/news/Saucy+illicit+sign+points+Constance+charms/1776726/story.html
[21-Jul-2009 16:37:54] <rmatte> yeh, I've seen that article hehe
[21-Jul-2009 16:37:58] <rmatte> pretty good
[21-Jul-2009 16:38:19] <rmatte> I guess that has spread around the net like wildfire
[21-Jul-2009 16:38:32] <blocky> it was on the front page of reddit today
[21-Jul-2009 16:38:49] <rmatte> nice lol
[21-Jul-2009 16:39:03] <blocky> quite a good illustration actually
[21-Jul-2009 16:39:04] <blocky> very professional
[21-Jul-2009 16:39:19] <blocky> love the comment from the counciller "i thought he was teaching her to water ski"
[21-Jul-2009 16:39:37] <rmatte> haha, yeh, I love that comment too
[21-Jul-2009 16:40:21] <rmatte> It's funny how they used a bunch of sexual innuendos in the article too
[21-Jul-2009 16:40:57] <blocky> erected
[21-Jul-2009 16:40:58] <blocky> har
[21-Jul-2009 16:41:23] <rmatte> sandblasted back it it's >>>virgin<<< state. vandals >>>had their way<<< with the sign
[21-Jul-2009 16:41:44] <rmatte> but yeh, I really have to get going, I'll see you around tomorrow if you're here
[21-Jul-2009 16:41:46] <rmatte> cheers
[21-Jul-2009 16:42:52] <blocky> have a good night
[21-Jul-2009 16:43:24] <rmatte> you too
[21-Jul-2009 16:47:08] schwartz_ is now known as schwartz
[21-Jul-2009 17:32:15] <straterra> I keep getting pages that all of my linux servers have some crazy offset (ntp)..but the time on all of them is accurate and fine..
[21-Jul-2009 17:32:28] <straterra> roughly 180 seconds for each
[21-Jul-2009 19:17:17] <otakup0pe> anyone knowledgable in the zenoss REST api around ?
[21-Jul-2009 19:17:40] <otakup0pe> is there a method i could use to pull down all datapoints of a given type ?
[21-Jul-2009 19:18:30] <otakup0pe> i think i need the name of every interface ?
[21-Jul-2009 19:18:41] <otakup0pe> oh nm :3
[21-Jul-2009 23:49:05] <Diddi> rmatte: FYI. regarding my problem with lost connection... I found it's related to my browser. running it in firefox instead of opera works perfectly (:
[22-Jul-2009 00:00:46] [disconnected at Wed Jul 22 00:00:46 2009]
[22-Jul-2009 00:00:46] [connected at Wed Jul 22 00:00:46 2009]
[22-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[22-Jul-2009 07:19:07] <Dieterbe_> Hi, is it just me or do you only see your nagios plugin result if it's not ok? when the status returns ok i cannot find the check on my device page
[22-Jul-2009 07:19:56] <jmp242> ?
[22-Jul-2009 07:20:10] <jmp242> I'm not at all sure what you mean Diddi
[22-Jul-2009 07:20:11] <jmp242> err
[22-Jul-2009 07:20:16] <jmp242> Dieterbe_:
[22-Jul-2009 07:20:41] <jmp242> Are you expecting to see something somewhere when everything is "OK"?
[22-Jul-2009 07:21:34] <Dieterbe_> yes
[22-Jul-2009 07:21:59] <Dieterbe_> When i am at: /Devices  /Server  / Linux / mydevice / os
[22-Jul-2009 07:22:18] <Dieterbe_> err sorry, not os, but status (eg the main view of the device)
[22-Jul-2009 07:22:34] <Dieterbe_> it shows "Component Type" with "Status"
[22-Jul-2009 07:22:44] <Dieterbe_> nagios checks are only shown there if they are not ok
[22-Jul-2009 07:22:52] <jmp242> Ahh. I don't believe that Zenoss will display anything if the checks succeed
[22-Jul-2009 07:23:17] <jmp242> I don't know how you'd get anything there without some extra work (and I'm not sure what that would be)
[22-Jul-2009 07:23:35] <jmp242> You could of course generate info level events from your sagios check
[22-Jul-2009 07:23:39] <jmp242> That might show up
[22-Jul-2009 07:25:01] <jmp242> I do see that Warning level events will show up as other... But it sounds like you might want to post in the forum regarding what might need to be done to the UI to get what you want
[22-Jul-2009 08:31:57] <straterra> I hate AIX..
[22-Jul-2009 08:32:07] <straterra> SNMP on older AIX is horribly broken
[22-Jul-2009 08:57:54] <jb> hah yes
[22-Jul-2009 08:57:59] <jb> AIX SNMPD is -terrible-
[22-Jul-2009 09:02:06] <straterra> Hmm..and 'paging' doesn't seem to be working for me in zenoss
[22-Jul-2009 09:02:47] <straterra> I have 1234567890@messaging.sprintpcs.com set for my paging adress
[22-Jul-2009 09:03:00] <straterra> And an alerting rule using page instead of email
[22-Jul-2009 09:05:16] <straterra> Oh..paging uses snpp
[22-Jul-2009 09:07:44] <nzle> How would one in ZenossCore, creat a user who is limited primarily to just submitting events via the methods decribed at http://www.zenoss.com/community/docs/howtos/send-events/
[22-Jul-2009 09:11:56] <nzle> How would one in ZenossCore, creat a user who is limited primarily to just submitting events via the methods decribed at http://www.zenoss.com/community/docs/howtos/send-events
[22-Jul-2009 09:13:03] <twm1010> I don't think you can do that in Core.
[22-Jul-2009 09:13:12] <twm1010> role based access controls is a feature of enterprise
[22-Jul-2009 09:14:01] <nzle> I was afraid of that.
[22-Jul-2009 09:16:44] <nzle> Another question related to what Im trying to do is where is there a good description of setting up heartbeats. I have a remote system with ssh tunnels which I want to know if the tunnels are down. I want to setup a wget to submit an event every few minutes at that end, and have a hertbeat event that goes critical if it doesnt get one in a few minutes.
[22-Jul-2009 09:22:27] <jmp242> nzle
[22-Jul-2009 09:22:29] <jmp242> see:
[22-Jul-2009 09:22:40] <jmp242> trac-zenpacks/wiki/CommunityFAQGeneral?version=72#HowdoIuseHeartbeatstylemonitoring
[22-Jul-2009 09:26:44] <nzle> Yep, I had found that.  Hoping for somthing with a little more meat to it.  thanx
[22-Jul-2009 09:32:02] <jmp242> Mmm. It seemed pretty clear to me - what question did you have?
[22-Jul-2009 09:45:13] <rmatte> jmp242: All you need to do is map an event that you get however works for your device.
[22-Jul-2009 09:45:22] <rmatte> the first line doesn't event make sense
[22-Jul-2009 09:45:24] <rmatte>
[22-Jul-2009 09:45:33] <rmatte> even*
[22-Jul-2009 09:50:18] <jmp242> ahh
[22-Jul-2009 09:50:26] <jmp242> well, it sor of does rmatte
[22-Jul-2009 09:50:52] <jmp242> I guess I could clarify - I'm just saying however you like to get the event (SNMP, Command based, Syslog, Trap, whatever)
[22-Jul-2009 09:51:05] <jmp242> you then map the event properly so you can apply the transforms to that event
[22-Jul-2009 09:51:35] <jmp242> of course, *I* know what I mean
[22-Jul-2009 09:54:20] <rmatte> hehe
[22-Jul-2009 10:03:32] <rmatte> man, they really need to establish standards for ZenPacks to be met before accepting them, the number one standard should be, if you include Mibs, put them in a container, don't leave them sitting in the root of /Mibs, it gets very messy after a while.
[22-Jul-2009 10:18:54] <nzle> I'm trying to submit an event using wget as described at http://www.zenoss.com/community/docs/howtos/send-events. Have the fields changed i recent version of Zenoss Core? My event is coming in but the eventkey, component, and summary are coming in null.
[22-Jul-2009 10:19:39] <kd4zay> hi all , got a question, i have a event that keep popin up on my linux boxes >threshold of low swap not met: current value 0.00 is this a setting that need to be modified if so where ?
[22-Jul-2009 10:24:59] <twm1010> rmatte: i'm still moving mibs
[22-Jul-2009 10:25:29] <twm1010> any more than 15 at a time, it times out.
[22-Jul-2009 10:26:21] <rmatte> kd4zay: yes, it is, go to Devices -> Server -> Linux, then click on the templates tab, then click on the link to the Device template, you'll see a thresholds section, click on the low swap threshold and change the enabled option to False
[22-Jul-2009 10:26:36] <rmatte> twm1010: toldja, you could have installed my pack and it'd be done by now
[22-Jul-2009 10:26:49] <rmatte> I was able to do 40 at a time
[22-Jul-2009 10:27:12] <kd4zay> thanks rmatte
[22-Jul-2009 10:27:20] <rmatte> kd4zay: no problem
[22-Jul-2009 11:09:14] <twm1010> rmatte: Yes, but that was before you figured out your little performance trick, it takes me just as much time to delete than it does to move so... same speed
[22-Jul-2009 11:12:48] <rmatte> true enough
[22-Jul-2009 11:13:43] <twm1010> it seems odd though, my SLES box never did this, most of my GUI actions involving moving things, be it one server to another device class, a mib to an organizer, take a long time.
[22-Jul-2009 11:14:06] <twm1010> Do you think latency from me on the GUI to the Zenoss server has an impact? My Zenoss box is at HQ, I'm at a remote site.
[22-Jul-2009 11:14:17] <twm1010> I might try it over an RDP connection
[22-Jul-2009 11:14:27] <rmatte> it would probably have a bit of an impact
[22-Jul-2009 11:14:44] <rmatte> The Zenoss UI is largely Ajax based
[22-Jul-2009 11:15:17] <rmatte> but your server might have something to do with it as well
[22-Jul-2009 11:15:55] <twm1010> perhaps, nothing too special with it, RHEL5 fully patched, stack install 2.4.2
[22-Jul-2009 11:16:40] <twm1010> ah ha, it took 40 that time..
[22-Jul-2009 11:16:41] <rmatte> I meant hardwarewise
[22-Jul-2009 11:16:51] <twm1010> yeah, its an older DL380 HP
[22-Jul-2009 11:18:42] <twm1010> I should fire up the HP tools and configure the RAID controller optomize for writes, i'd imagine
[22-Jul-2009 11:18:56] <twm1010> Also saw something about configuring the filesystem to not update access times
[22-Jul-2009 11:19:35] <RediXe> Does anyone know if there is a link to a 64 bit Ubuntu server 9.04 binary?
[22-Jul-2009 11:19:59] <mrayzenoss> RediXe: the .deb listed for Ubuntu should work with 9.04
[22-Jul-2009 11:20:15] <rmatte> RediXe: the 32 bit binary will run on a 64bit system regardless
[22-Jul-2009 11:20:25] <rmatte> RediXe: as long as it's pre-compiled
[22-Jul-2009 11:21:20] <rmatte> but yeh, just grab the 64bit installer for 8.04
[22-Jul-2009 11:21:23] <rmatte> and it should work fine
[22-Jul-2009 11:21:55] <RediXe> Yeah, we just run 64 bit and the list 6.06, 8.04 as the supported. I believe we have the 8.04 installed but when trying to auto discover I checked the log of the job and it said zenhub was not running
[22-Jul-2009 11:22:19] <rmatte> that .bin stack installer is quite generic, you can run that on almost any flavour of Linux, so just grab that
[22-Jul-2009 11:22:39] <rmatte> it has all of the dependencies and components bundled in to it
[22-Jul-2009 11:23:16] <mrayzenoss> so does the .deb for Debian/Ubuntu
[22-Jul-2009 11:23:30] <mrayzenoss> it's not a proper distro-friendly .deb
[22-Jul-2009 11:23:39] <rmatte> If it's complaing that zenhub is not running you need to investigate why, maybe a process that you are already running is taking up a port that Zenoss needs to use to communicate
[22-Jul-2009 11:23:51] <rmatte> Is this a clean install of Linux?
[22-Jul-2009 11:24:57] <RediXe> Pretty much, just apache 2, php, and mysql I believe are the only other things installed so far
[22-Jul-2009 11:26:12] <rmatte> so you had mysql already installed prior to Zenoss running?
[22-Jul-2009 11:26:20] <mrayzenoss> the Zenoss install puts down another MySQL instance, it's possible it's conflicting with your previous install
[22-Jul-2009 11:28:03] <RediXe> Well it appears to be working just fine .. minus the zenhub "ERROR:zen.ZenDisc:Timeout connecting to zenhub: is it running?"
[22-Jul-2009 11:29:57] <rmatte> RediXe: you can ssh in, become the zenoss user (su - zenoss), then execute: zenhub run -v10
[22-Jul-2009 11:30:09] <rmatte> that might give you an indication of why it's not starting
[22-Jul-2009 11:30:20] <rmatte> you can also check $ZENHOME/logs/zenhub.log
[22-Jul-2009 11:31:32] <RediXe> http://pastebin.org/3590   when trying to run the zenhub run -v10
[22-Jul-2009 11:32:31] <rmatte> ok, so it's complaining that port 8081 is already in use by something
[22-Jul-2009 11:32:43] <rmatte> running anything on that box that would be using that port?
[22-Jul-2009 11:32:57] <chudler> pkill -9 -f zenhub
[22-Jul-2009 11:33:23] <chudler> ;-)
[22-Jul-2009 11:33:41] <RediXe> ps aux doesn't say it's running so should I still try to kill it
[22-Jul-2009 11:33:49] <rmatte> sure
[22-Jul-2009 11:33:50] <rmatte> go for it
[22-Jul-2009 11:33:54] <RediXe> lol
[22-Jul-2009 11:34:07] <chudler> not likely to do anything, obviously.  This is a common problem on my install, though.
[22-Jul-2009 11:34:11] <rmatte> then try starting it again, if you get the same message then you've got something running that's hogging that port
[22-Jul-2009 11:34:54] <RediXe> Yeah something is hogging that port ... I'll play around some... thank you for the help so far though ... really appreciate it
[22-Jul-2009 11:35:02] <rmatte> no problem
[22-Jul-2009 11:35:05] <rmatte> lunch time!
[22-Jul-2009 11:35:19] <chudler> Redixe, if you want to see what it is, try "netstat -tulpna | grep 8081" as root, or sudo.
[22-Jul-2009 11:36:33] <RediXe> pdns server .. .... can fix it
[22-Jul-2009 11:39:18] <RediXe> Yup, ZenHub appears to be running just fine now
[22-Jul-2009 11:40:24] <rmatte> double cheeseburger++
[22-Jul-2009 11:41:19] <rmatte> netstat -tulpna is what netstat should display by default in my opinion
[22-Jul-2009 11:41:59] <chudler> heh, yeah. I have long since forgotten what each one does. I just like that it looks like "tulip"
[22-Jul-2009 11:42:24] <rmatte> I remember some commands like that, for instance du -sh
[22-Jul-2009 11:42:28] <rmatte> dush
[22-Jul-2009 11:42:37] <rmatte>
[22-Jul-2009 11:42:40] <RediXe> Trying to lose weight So grilled chicken ..... Is there way to set up Zenoss to deal with vlans? (I'm not a network guru but it was a concern brought up by the network guy)
[22-Jul-2009 11:43:02] <rmatte> define "deal" with vlans
[22-Jul-2009 11:43:58] <chudler> I think du -h is deprecated. Its now du -si -s, which sucks, but I could be totally wrong I dont remember where I got that idea.
[22-Jul-2009 11:44:14] <RediXe> ummm, the server that is running on a vlan and we have other vlans I guess was curious if zenoss would be able to some how discover devices on the other vlans
[22-Jul-2009 11:44:22] <rmatte> it may be deprecated, but it still works
[22-Jul-2009 11:44:44] <rmatte> and I doubt -h is deprecated, it works on an array of commands\
[22-Jul-2009 11:44:52] <rmatte> df, du, ls, etc...
[22-Jul-2009 11:45:34] <rmatte> RediXe: provided you don't overlap IPs and the server can actually see all the Vlans, you shouldn't have an issue
[22-Jul-2009 11:45:40] <straterra> Sweet. zenjmx keeps crashing
[22-Jul-2009 11:46:05] <rmatte> straterra: check the log file for clues
[22-Jul-2009 11:46:11] <straterra> Which one?
[22-Jul-2009 11:46:24] <rmatte> I think it's called zmx.log or something
[22-Jul-2009 11:46:31] <rmatte> in $ZENHOME/logs
[22-Jul-2009 11:47:02] <chudler> its -H I was thinking of. it is deprected in gnu-utils. I think -h is not likely to go anywhere.
[22-Jul-2009 11:47:22] <rmatte> yeh, lol, -h is pretty standard
[22-Jul-2009 11:48:08] <straterra> http://pastebin.ca/1503380
[22-Jul-2009 11:48:46] <mrayzenoss> straterra: which JDK are you using?
[22-Jul-2009 11:49:02] <straterra> uhm..whatever comes with CentOS
[22-Jul-2009 11:49:08] <straterra> I didn't change anything..I rebooted the server o.O
[22-Jul-2009 11:49:38] <mrayzenoss> somewhere it's doc'd that you have to use the Sun JDK
[22-Jul-2009 11:49:52] <straterra> As far as I know..I am
[22-Jul-2009 11:50:03] <straterra> I haven't switched JDK's
[22-Jul-2009 11:50:50] <straterra> All I did was power down zenoss, reboot the server, and bring it back up
[22-Jul-2009 11:51:34] <rmatte> zenoss is so powerful it needs to be powered down before actually rebooting the server
[22-Jul-2009 11:51:45] <rmatte> (Just kidding around)
[22-Jul-2009 11:51:56] <mrayzenoss> straterra: as the Zenoss user, type 'which java'
[22-Jul-2009 11:52:01] <rmatte> and nothing was upgraded prior to the reboot?
[22-Jul-2009 11:52:34] <mrayzenoss> straterra: http://www.zenoss.com/community/docs/zenoss-extended-monitoring/2.4.2/ch09s02.html#zenjmx_jre_prereq
[22-Jul-2009 11:52:39] <rmatte> I've had it where I've upgraded something and it only took effect after a reboot and ended up messing something up, so it's possible
[22-Jul-2009 11:53:09] <straterra> [zenoss@mims1 ~]$ which java
[22-Jul-2009 11:53:09] <straterra> /usr/bin/which: no java in (/opt/zenoss/bin:/usr/kerberos/bin:/usr/local/bin:/bin:/usr/bin:/home/zenoss/bin)
[22-Jul-2009 11:53:13] <straterra> \O/
[22-Jul-2009 11:53:21] <rmatte> there be your issue good sir
[22-Jul-2009 11:53:32] <straterra> Wonder why it took a reboot to do that
[22-Jul-2009 11:53:59] <rmatte> java was possibly still running even after being removed
[22-Jul-2009 11:54:56] <straterra> Possible..but..I don't remember uninstalling anything
[22-Jul-2009 11:54:58] <rmatte> you're sure it was working before the reboot?
[22-Jul-2009 11:55:02] <straterra> Yeah
[22-Jul-2009 11:55:10] <rmatte> pretty weird
[22-Jul-2009 11:55:20] * rmatte adds this to his list of reasons why not to use CentOS
[22-Jul-2009 11:55:34] <rmatte> or RedHat
[22-Jul-2009 11:55:36] <rmatte>
[22-Jul-2009 11:55:36] <straterra> Well..RHEL is expensive
[22-Jul-2009 11:55:37] <straterra> lol
[22-Jul-2009 11:55:46] <rmatte> yes, that is reason number 1
[22-Jul-2009 11:55:50] <rmatte>
[22-Jul-2009 11:56:22] <straterra> You wouldn't happen to know what package it is, would you?
[22-Jul-2009 11:56:30] <straterra> java-1.6.0-openjdk ?
[22-Jul-2009 11:56:42] <rmatte> something that contains "jdk" and ends in .rpm?
[22-Jul-2009 11:56:49] <rmatte> not openjdk
[22-Jul-2009 11:57:13] <chudler> I could be totally off, but I thought java license didnt allow re-distribution?
[22-Jul-2009 11:57:20] <rmatte> that's open source version, you need the sun version
[22-Jul-2009 11:57:26] <chudler> oh, yeah I am totally off.
[22-Jul-2009 11:57:46] <rmatte> chudler: it allows redistribution, but you have to manually agree to the license upon installation
[22-Jul-2009 11:57:49] <RediXe> From the server that runs zenoss we can not ping a machine on another vlan ... I am guessing that means zenoss won't be able to see it?
[22-Jul-2009 11:58:28] <rmatte> chudler: heck, back in the day when I was using FreeBSD I had to go to a web link, accept the license agreement, download the file, throw it in to the package manager's file directory, and then I could continue installing it
[22-Jul-2009 11:59:02] <rmatte> RediXe: correct
[22-Jul-2009 11:59:15] <rmatte> RediXe: you'd need to somehow tie the server in to each vlan
[22-Jul-2009 11:59:28] <rmatte> RediXe: that's not really a Zenoss issue though, that's a network/server issue
[22-Jul-2009 12:00:37] <rmatte> doesn't matter what software you throw on that server, if you can't ping something, the software won't see it
[22-Jul-2009 12:00:49] <straterra> so..I'll just install the RPM
[22-Jul-2009 12:00:53] <RediXe> Yeah that's what I figured
[22-Jul-2009 12:00:58] <rmatte> straterra: that works
[22-Jul-2009 12:01:14] <rmatte> RediXe: your network guy should know this stuff
[22-Jul-2009 12:01:18] <straterra> I just don't know why it worked before o.O
[22-Jul-2009 12:01:26] <rmatte> straterra: black magic
[22-Jul-2009 12:01:34] <straterra> Yeah, I guess so
[22-Jul-2009 12:01:36] <rmatte> Zenoss comes bundled with it, but you only get one dose
[22-Jul-2009 12:01:43] <rmatte> so the reboot spent your dose, and you're done
[22-Jul-2009 12:01:47] <rmatte> now you need to install Java
[22-Jul-2009 12:01:48] <rmatte>
[22-Jul-2009 12:01:48] <straterra> well damn
[22-Jul-2009 12:02:36] <RediXe> rmatte, I'm sure he does ... he just tasked me with playing with this stuff so I am
[22-Jul-2009 12:02:45] <rmatte> RediXe: nice hehe
[22-Jul-2009 12:03:06] <rmatte> RediXe: That's how I got in to managing Zenoss, I was tasked with messing around with it, 7 months later I'm the expert
[22-Jul-2009 12:03:45] <rmatte> still have lots to learn though, I'm getting in to the development aspect of things now
[22-Jul-2009 12:04:16] <RediXe> rmatte, lol yeah, I just mentioned it since I used it a little at my last job and so I got stuck with playing with it again.
[22-Jul-2009 12:04:25] <rmatte> haha
[22-Jul-2009 12:04:37] <rmatte> It's a neat product, bit of a learning curve though
[22-Jul-2009 12:05:42] <RediXe> Yeah, this whole vlan thing throws a pretty good sized curve ball though .... so not sure what we will end up doing
[22-Jul-2009 12:06:51] <rmatte> the server needs to be exposed to each vlan somehow
[22-Jul-2009 12:06:57] <rmatte> you might have to setup tunnels or something
[22-Jul-2009 12:07:28] <rmatte> which might not be practical depending on the size of the vlans and the number of vlans
[22-Jul-2009 12:12:35] <blocky> so im having the same issue i had yesterday with a new server
[22-Jul-2009 12:12:45] <blocky> no graphs appearing except load average, which says missing rrd files
[22-Jul-2009 12:13:06] <rmatte> blocky: is that the only Linux server having that issue?
[22-Jul-2009 12:13:27] <blocky> no the one that was having it yesterday is now working
[22-Jul-2009 12:13:31] <blocky> this is another linux server i added
[22-Jul-2009 12:13:33] <rmatte> ah
[22-Jul-2009 12:13:42] <blocky> same issue, different box
[22-Jul-2009 12:13:44] <rmatte> well do everything we did yesterday
[22-Jul-2009 12:13:51] <blocky> reinstall zenoss?
[22-Jul-2009 12:13:52] <blocky> lol
[22-Jul-2009 12:13:56] <rmatte> nah
[22-Jul-2009 12:14:04] <rmatte> on the OS tab, have the drives showed up?
[22-Jul-2009 12:14:08] <blocky> i checked for local copies of community strings
[22-Jul-2009 12:14:21] <blocky> no, no os info
[22-Jul-2009 12:14:37] <rmatte> if you go to the dropdown menu, Commands -> snmpwalk
[22-Jul-2009 12:14:40] <rmatte> does it work?
[22-Jul-2009 12:14:57] <blocky> no
[22-Jul-2009 12:14:58] <blocky> lol
[22-Jul-2009 12:15:07] <blocky> brb
[22-Jul-2009 12:15:10] <rmatte> there's your problem
[22-Jul-2009 12:15:12] <rmatte>
[22-Jul-2009 12:18:26] <straterra> Sweet.
[22-Jul-2009 12:18:31] <straterra> that daemon is working again..thanks
[22-Jul-2009 12:19:06] <rmatte> np
[22-Jul-2009 12:19:29] <rmatte> Matt deserves 99% of the credit on that solution
[22-Jul-2009 12:19:45] <rmatte> I told you to check the log file, so that was my contribution
[22-Jul-2009 12:19:56] <johns55> hey guys, is it better to install Zenoss from source code or from the stack installer? I'm planning to do this on a Fedora 11 distribution
[22-Jul-2009 12:20:09] <rmatte> stack installer hands down, avoid the headache
[22-Jul-2009 12:20:29] <johns55> i would still need to install mySQL before I run the installer correct?
[22-Jul-2009 12:20:41] <rmatte> nope, the stack installer comes bundled with everything
[22-Jul-2009 12:20:44] <rmatte> just install and go
[22-Jul-2009 12:20:50] <johns55> wow incredible
[22-Jul-2009 12:20:53] <johns55> okay I will give it a shot
[22-Jul-2009 12:20:56] <johns55> thanks for the tip
[22-Jul-2009 12:20:57] <rmatte> cool
[22-Jul-2009 12:20:58] <rmatte> np
[22-Jul-2009 12:21:13] <mrayzenoss> yeah, I recommend stack installers until we get our act together with proper distro support
[22-Jul-2009 12:22:15] <rmatte> yeh, distro support will be nice, but if it works, it works
[22-Jul-2009 12:22:30] <mrayzenoss> there is going to be some work on our build real soon for LSB compatibility
[22-Jul-2009 12:22:52] <chudler> GREAT.  I had a few issues with hearbeat and LSB re zenoss, but nothing serious.
[22-Jul-2009 12:23:25] <rmatte> LSB, the one thing that RedHat and CentOS tend to ignore
[22-Jul-2009 12:23:28] <mrayzenoss> King Crab should have LSB compliance and the build straightened up
[22-Jul-2009 12:23:38] <mrayzenoss> Any word on RHEL 6 for LSB?
[22-Jul-2009 12:24:16] <rmatte> not that I know of
[22-Jul-2009 12:25:22] * chudler thought RHEL6 is still vaporware
[22-Jul-2009 12:25:31] <rmatte> it is
[22-Jul-2009 12:25:37] <mrayzenoss> chudler: some day
[22-Jul-2009 12:26:02] <rmatte> quite honestly, I'm just waiting for RedHat to die off and for Ubuntu to take it's place
[22-Jul-2009 12:26:58] <rmatte> much cheaper, and better package system
[22-Jul-2009 12:27:01] <mrayzenoss> it might be awhile
[22-Jul-2009 12:27:28] <rmatte> yeh, RedHat is everywhere, but if they don't get their act together something like that will eventually happen
[22-Jul-2009 12:27:42] <dberger> Anyone know what I have to do to avoid getting "Only ZenPacks that have been installed in development mode can be exported" when I try to export a ZenPack?
[22-Jul-2009 12:27:55] <rmatte> they were lucky in basically being the first commercial version of Linux, so everyone jumped on it early on because of the initial hype
[22-Jul-2009 12:28:03] <mrayzenoss> dberger: trac-zenpacks/wiki/DevelopmentMode
[22-Jul-2009 12:29:38] <rmatte> that was fun figuring out
[22-Jul-2009 12:34:40] <dberger> thanks
[22-Jul-2009 12:39:41] <mrayzenoss> heh, just saw a Red Hat presentation listing RHEL 5.4, 6.0 and 6.1 road maps
[22-Jul-2009 12:40:07] <mrayzenoss> it's got RHEL 6 / F12, so I'm guessing RHEL 6 is in 6 months
[22-Jul-2009 12:40:37] <chudler> grrr, I thought they'd use F11. I think it would make a nice base. I dont see how they could have time for a beta if they use F12.. Oh well.
[22-Jul-2009 12:40:45] <mrayzenoss> F11 is already out
[22-Jul-2009 12:40:55] <mrayzenoss> ahh, I see what you're saying
[22-Jul-2009 12:41:37] <mrayzenoss> well, the presentation is about the libvirt roadmap, so RHEL6 and F12 will have the same libvirt release
[22-Jul-2009 12:42:25] <jb> ok
[22-Jul-2009 12:47:09] <rmatte> hmmm, I just received a "must be 18 or older" message on my iphone when installing an RSS reader of all things
[22-Jul-2009 12:47:26] <rmatte> just in case you sign up to a porn feed?
[22-Jul-2009 12:48:17] <chudler> there's porn feeds?  of course there are... /me takes a lunch break
[22-Jul-2009 12:48:26] <rmatte> hehe
[22-Jul-2009 12:49:28] <rmatte> ooooh, dev session tomorrow
[22-Jul-2009 12:50:06] * rmatte pokes mrayzenoss and hints to include notice in the channel topic
[22-Jul-2009 12:50:09] <rmatte> ah, too quick
[22-Jul-2009 12:50:13] <mrayzenoss> heh
[22-Jul-2009 12:50:14] <rmatte>
[22-Jul-2009 12:53:36] <rmatte> hmmm, I have some trac tickets to log that I've been putting off
[22-Jul-2009 12:53:46] <rmatte> might as well do that now
[22-Jul-2009 12:54:45] <rmatte> is there no way to change the Trac password?
[22-Jul-2009 12:56:56] <mrayzenoss> rmatte: alas no, we use a shared password between svn and trac
[22-Jul-2009 12:57:00] <mrayzenoss> some day
[22-Jul-2009 12:57:22] <mrayzenoss> rmatte: you can email me your password and I'll set the new one
[22-Jul-2009 12:57:33] <mrayzenoss> since I already have your old password
[22-Jul-2009 13:00:14] <rmatte> http://dev.zenoss.org/trac/ticket/5350
[22-Jul-2009 13:00:31] <rmatte> yeh, I have the old password too, it's just hard to remember
[22-Jul-2009 13:06:40] <rmatte> Here's a big one: http://dev.zenoss.org/trac/ticket/5351
[22-Jul-2009 13:10:21] <rmatte> I just somehow pulled a muscle in my leg, man that hurts
[22-Jul-2009 13:10:43] <rmatte> I need to excercise or something, I'm managing to injure myself while sitting at a desk
[22-Jul-2009 13:12:18] <rmatte> hmmm, trying to think of what the third trac case I wanted to open is for
[22-Jul-2009 13:15:48] <johns55> interesting, you run the stack installer as root, does this mean it somehow creates the zenoss user and installs the application to have that user run it?
[22-Jul-2009 13:16:00] <mrayzenoss> yes
[22-Jul-2009 13:16:16] <johns55> wow this is way easier than source installation
[22-Jul-2009 13:16:45] <rmatte> yeh, upgrading with it is way easier too
[22-Jul-2009 13:16:51] <rmatte> you just run the new .bin and it does everything
[22-Jul-2009 13:17:52] <johns55> interesting. it only allows alpha numeric for mysql root pass
[22-Jul-2009 13:17:54] <eidolon> hi folks - is there a zenoss monitor that will 'watch' a log file for a string? i'd like to tell our developers to use syslog and set specific messages in there to throw alerts to zenoss. possible?
[22-Jul-2009 13:18:22] <johns55> eidolon: CMD script?
[22-Jul-2009 13:18:37] <eidolon> i don't know aht that is, sorry.
[22-Jul-2009 13:18:49] <rmatte> johns55: correct, I just use the default username and pass, since if someone gets in to the box they've already owned it anyways
[22-Jul-2009 13:19:00] <johns55> make a little bash script that returns an exit status when it greps a specific word in the file
[22-Jul-2009 13:19:45] <eidolon> johns55: hrm. how do you know how far back to grep? if i have a syslog that's 4 million rows long, i don't want to grep that file repeatedly, and possibly keep throwing erros when it finds it in that file.
[22-Jul-2009 13:19:45] <rmatte> eidolon: are you effectively trying to filter syslog messages?
[22-Jul-2009 13:19:55] <eidolon> rmatte: errrrr.
[22-Jul-2009 13:19:57] <jmp242> why not just use syslog to Zenoss and then use event mapping?
[22-Jul-2009 13:19:58] <eidolon> dunno? 
[22-Jul-2009 13:20:09] <johns55> eidolon, use logrotate to reduce log size
[22-Jul-2009 13:20:10] <rmatte> eidolon: what is your end goal in doing this?
[22-Jul-2009 13:20:30] <rmatte> If it's what I think it is I may have a better solution to suggest
[22-Jul-2009 13:20:59] <eidolon> so. we ahve a bzillion perl apps deployed. and some java apps as well. all of them know how to generate a syslog message (same with bash scripts). if there's a problem, i want them to go "PAHTUI!" and spit a message to syslog. that gets picked up by zenoss and generates an event.
[22-Jul-2009 13:21:13] * eidolon is all ears.
[22-Jul-2009 13:21:38] <rmatte> Zenoss has syslog support built in, so provided you're running a syslogd on the server that's set to forward syslog messages to Zenoss, then you're all set
[22-Jul-2009 13:21:49] <eidolon> hmmm.
[22-Jul-2009 13:22:09] <eidolon> the syslog framework is... painful. first, it's syslog-ng. second, it was written by a mad monk 4-5 years ago, and mostly doesn't work.
[22-Jul-2009 13:22:21] <rmatte> lol
[22-Jul-2009 13:22:31] <eidolon> but. if i can fire syslog messages back to the zenoss host, then set up mappings, that may make the most sense.
[22-Jul-2009 13:22:53] <rmatte> that's what we're doing, but we do it for switches, routers, firewalls, servers, etc...
[22-Jul-2009 13:23:01] * eidolon nods
[22-Jul-2009 13:23:03] <eidolon> i.. hm.
[22-Jul-2009 13:23:07] <eidolon> i think we may have a loghost.
[22-Jul-2009 13:23:30] <eidolon> but. i'll check the docs for syslog targets in zenoss. i hadn't considered syslogging directly to zenoss.
[22-Jul-2009 13:23:35] <mrayzenoss> if you're just modifying scripts, it might be easier to skip to the chase and use zensendevent
[22-Jul-2009 13:24:17] <eidolon> cept that would require zensendevetn on all the hosts, neh?
[22-Jul-2009 13:24:23] <eidolon> we have hundreds of machines.
[22-Jul-2009 13:24:33] <mrayzenoss> true
[22-Jul-2009 13:24:44] <eidolon> okay, let me look at hte logging path.  i like this.
[22-Jul-2009 13:26:17] <twm1010> mrayzenoss: http://forums.zenoss.com/viewtopic.php?t=10410 think you could persuade some to weigh in on that?
[22-Jul-2009 13:27:22] <rmatte> twm1010: those are the two best solutions, and I use both
[22-Jul-2009 13:28:07] <twm1010> Why both?
[22-Jul-2009 13:28:33] <rmatte> more reliable, we're a NOC who monitors some pretty large clients, we can't afford to miss events
[22-Jul-2009 13:28:46] <blocky> is anyone familiar with the zenpack for HP proliant servers
[22-Jul-2009 13:28:49] <twm1010> fair enough.
[22-Jul-2009 13:28:59] <twm1010> do you use syslog or traps, or both?
[22-Jul-2009 13:29:09] <rmatte> blocky: yeh, I use it, doesn't seem to do much other than showing server temperature
[22-Jul-2009 13:29:09] <blocky> i created a new subdevice /server/linux/proliant and bound the templates for fan speed, temperature and a couple others
[22-Jul-2009 13:29:20] <rmatte> twm1010: both
[22-Jul-2009 13:29:21] <blocky> well id like to see that
[22-Jul-2009 13:29:30] <rmatte> twm1010: syslogs, traps, and a threshold for operational status
[22-Jul-2009 13:29:35] <blocky> but now my perf tab just shows User-supplied Python expression (here.threshold) for maximum value caused error: ['TemperatureCelsius_temperature_celsius']
[22-Jul-2009 13:29:49] <twm1010> doesn't that cause duplicate events though? I'm looking to avoid that.
[22-Jul-2009 13:30:20] <rmatte> blocky: not sure about that one
[22-Jul-2009 13:31:23] <rmatte> twm1010: yeh it does, but we operate a bit differently, we have zenoss integrated with a central ticketing system, so if multiple events come in the analysts just relate all of the events back to a single ticket and work out of that. In your case, quite honestly the most reliable that I've found is using the threshold.
[22-Jul-2009 13:31:47] <rmatte> twm1010: sometimes we don't receive syslog or traps for ports going down but the threshold always works
[22-Jul-2009 13:31:51] <eidolon> well interesting.  apparently we -have- a logging host.
[22-Jul-2009 13:31:55] <eidolon> but someone's not keeping it particularly clean
[22-Jul-2009 13:31:57] <eidolon> dbs@bos-log01:/var/log$ ls -lh messages.log
[22-Jul-2009 13:31:57] <eidolon> -rw-r--r--  1 root root 5.0G Jul 22 14:31 messages.log
[22-Jul-2009 13:32:10] <rmatte> 5gig file, my lord
[22-Jul-2009 13:32:45] <chudler> -rw-r--r--  1 root root 15G Jul 22 13:32 /var/log/messages
[22-Jul-2009 13:32:47] <chudler> got you beat
[22-Jul-2009 13:32:52] * eidolon loses.
[22-Jul-2009 13:33:17] <rmatte> tail -100000 messages.log > messages.new && rm messages.log && mv messages.new messages.log
[22-Jul-2009 13:33:18] <rmatte>
[22-Jul-2009 13:33:23] <mrayzenoss> twm1010: I think we try to treat them equally, since some devices don't support one or the other
[22-Jul-2009 13:33:28] <twm1010> well sure, the threshold is from the OID, which will either answer or it won't
[22-Jul-2009 13:33:31] <chudler> it rotates every day.  just lots and lots of things logging there
[22-Jul-2009 13:34:22] <rmatte> twm1010: it's not whether it'll answer or not, it just gives a value between 0 and 2
[22-Jul-2009 13:35:38] <twm1010> So do you go through each interface on the device and add a local template with the threshold?
[22-Jul-2009 13:35:52] <twm1010> or do you place a template with a threshold in the device class
[22-Jul-2009 13:36:08] <rmatte> http://www.zenoss.com/Members/cluther/polling-interface-status/
[22-Jul-2009 13:36:13] <rmatte> There's the guide
[22-Jul-2009 13:36:19] <blocky> can i take a look at the python it's choking on?
[22-Jul-2009 13:36:27] <johns55> after installing the stack installer, it says...Info: To access the Zenoss Application, go to http://localhost:8080 from your
[22-Jul-2009 13:36:28] <johns55> browser.
[22-Jul-2009 13:36:28] <twm1010> I'm supposing I could dice things up some, create subclasses under ../Cisco for WAN/LAN/Edge routers..
[22-Jul-2009 13:36:30] <johns55> does that sound right?
[22-Jul-2009 13:36:47] <twm1010> johns55: assuming your viewing it FROM the localhost, yes
[22-Jul-2009 13:36:52] <rmatte> johns55: well, replace localhost with the IP of the server
[22-Jul-2009 13:37:07] <rmatte> johns55: http://ip.of.server:8080
[22-Jul-2009 13:37:18] <johns55> rmatte: I can fire up lynx on the current server and it will load but not from a remote computer
[22-Jul-2009 13:37:40] <rmatte> johns55: it won't load from a remote computer using the actual IP of the server?
[22-Jul-2009 13:37:46] <johns55> nope
[22-Jul-2009 13:37:52] <johns55> and it's listening to all ports
[22-Jul-2009 13:37:54] <johns55> .python.b 29318 zenoss    3u  IPv4  54226      0t0  TCP *:8080 (LISTEN)
[22-Jul-2009 13:37:56] <rmatte> johns55: then you have a firewall issue on your hands
[22-Jul-2009 13:38:06] <johns55> hrm okay let me check iptables
[22-Jul-2009 13:38:22] <johns55> rmatte: you're absolutely correct
[22-Jul-2009 13:38:30] <johns55> thanks
[22-Jul-2009 13:38:37] <rmatte> no prob
[22-Jul-2009 13:39:10] <twm1010> rmatte: thanks, I'll probably end up using a combination here, but grouping my routers will probably save me a lot of work
[22-Jul-2009 13:39:27] <rmatte> yeh, grouping does save work indeed
[22-Jul-2009 13:41:41] <blocky> how do i edit a zenpack?
[22-Jul-2009 13:42:41] <rmatte> edit in what sense?
[22-Jul-2009 13:43:35] <rmatte> If you're talking about putting a ZenPack in to development mode: http://blog.zenoss.com/2009/03/29/tip-of-the-month-restoring-zenpacks-to-development-mode/
[22-Jul-2009 13:56:45] <twm1010> Is there a Zenpack that simply adds a bunch of pre-built interface templates? That would be handy.
[22-Jul-2009 13:57:17] <rmatte> not that I'm aware of, but I'll be making one eventually
[22-Jul-2009 13:57:51] <rmatte> I just need to get the correct OIDs for some of the interface types
[22-Jul-2009 14:01:28] <twm1010> example: I see Zenoss monitoring Serial 3/0 on one my routers, a DS3 to our call center, but there is also an interface T3/0 I'd assume this is the controller
[22-Jul-2009 14:02:34] <rmatte> right
[22-Jul-2009 14:06:44] <rmatte> I assume you know how to create interface templates?
[22-Jul-2009 14:16:38] <twm1010> sure, you just drop it in the class named after the "type" right?
[22-Jul-2009 14:16:48] <twm1010> so if i drop a template called ds3 under ../Cisco voila
[22-Jul-2009 14:29:01] <twm1010> Now if you drops said template, is it picked up immediate, on next collector changes, or next remodel?
[22-Jul-2009 14:32:20] <johns55> holy cow, it takes like 5 minutes to do an snmpwalk on my distribution switch and when I try to add this on zenoss it times out, any tips on what I should do?
[22-Jul-2009 14:35:35] <rmatte> twm1010: nah, I just drop them right under /Devices
[22-Jul-2009 14:35:45] <rmatte> since 1 interface template is good for any interface of that type
[22-Jul-2009 14:36:30] <rmatte> where you run in to issues is with sub interfaces for serial circuits, or whatever, since you need to copy the template locally and modify it for each interface, otherwise you get debug messages coming in complaining about the errors OID not being available
[22-Jul-2009 14:37:06] <twm1010> right, i had a ton of those, a lot less since i set my interface ignore zproperty and got rid of VLANs and other things i don't care about
[22-Jul-2009 14:37:21] <rmatte> ah, we have to monitor vlans
[22-Jul-2009 14:37:47] <twm1010> well, I thought about it a bit. The "vlan" interface shows the bandwidth going over the vlan right
[22-Jul-2009 14:38:03] <twm1010> but it doesn't tell you which physical ports are doing to the chatting on said VLAN
[22-Jul-2009 14:38:14] <twm1010> so, that doesnt' tell much other than which VLAN might be chatting the most
[22-Jul-2009 14:38:17] <rmatte> nah, I more monitor the vlan in case it drops
[22-Jul-2009 14:38:29] <twm1010> Gotcha.
[22-Jul-2009 14:38:33] <rmatte> vlans dropping on cisco devices == bad news
[22-Jul-2009 14:38:45] <twm1010> sure, no joke, didn't know that could happen
[22-Jul-2009 14:39:42] <rmatte> yeh, we were monitoring RBC for a while (with OpenView at the time, not Zenoss), and we'd see vlans drop occasionally, we'd have to call them to notify them when it happened
[22-Jul-2009 14:40:46] <rmatte> vlans dropping at a bank usually resulted in critical systems going down
[22-Jul-2009 14:41:14] <twm1010> sure, no joke, the VLAN originates somewhere, I suppose a remote switch dropping a VLAN would indicate a problem with VTP or something like that...
[22-Jul-2009 14:41:25] <rmatte> yeh
[22-Jul-2009 14:41:29] <twm1010> We basically have two monster switches that have a backplane, so... doesn't fit for me.
[22-Jul-2009 14:41:37] <rmatte> yeh
[22-Jul-2009 14:41:44] <rmatte> monster as in 48 port?
[22-Jul-2009 14:41:52] <twm1010> monster as in 200+ port
[22-Jul-2009 14:42:05] <rmatte> Cisco, Enterasys?
[22-Jul-2009 14:42:07] <twm1010> each has seven 48 port blades?
[22-Jul-2009 14:42:15] <twm1010> CAT6509
[22-Jul-2009 14:42:19] <rmatte> ah
[22-Jul-2009 14:42:30] <rmatte> does it run IOS or CatOS?
[22-Jul-2009 14:42:35] <twm1010> IOS
[22-Jul-2009 14:42:38] <rmatte> cool
[22-Jul-2009 14:42:54] <rmatte> so you have two enclosures then?
[22-Jul-2009 14:43:35] <twm1010> Yep
[22-Jul-2009 14:43:40] <rmatte> that's actually 336 ports per enclosure
[22-Jul-2009 14:43:53] <twm1010> something like that, we have open blade slots
[22-Jul-2009 14:43:56] <rmatte> so 672 ports total
[22-Jul-2009 14:43:58] <twm1010> and not all blades are 48 ports
[22-Jul-2009 14:44:02] <rmatte> ah
[22-Jul-2009 14:44:04] <twm1010> some have fiber
[22-Jul-2009 14:44:23] <rmatte> fiber are generally 24
[22-Jul-2009 14:44:54] <twm1010> each server and/or remote switch will generally have a connection to each
[22-Jul-2009 14:45:06] <twm1010> you lose one whole CAT, everything keeps ticking.
[22-Jul-2009 14:45:36] <rmatte> yeh
[22-Jul-2009 14:46:00] <rmatte> sort of a mesh
[22-Jul-2009 14:47:01] <twm1010> Yeah, its a decent setup. I'm no VLAN guru, so I'll let the admins tell ME if they need VLAN monitoring.
[22-Jul-2009 14:47:10] <rmatte> hehe
[22-Jul-2009 14:47:27] <twm1010> until then, that's a ton less OID's keeping the collector busy.
[22-Jul-2009 14:48:57] <twm1010> I'm thinking the threshold monitoring is going to be the way to go here for the interfaces
[22-Jul-2009 14:49:41] <rmatte> yeh, it's been working the best for us
[22-Jul-2009 14:50:14] <rmatte> and I've got one box monitoring 350 devices over 3 different continents, lots of switches, routers, and servers, it's working very well for that
[22-Jul-2009 14:54:46] <twm1010> and really, for a lot of interfaces that are that critical, you probably want to set custom utilization thresholds too, right? so why not.
[22-Jul-2009 14:55:59] <jb> 153 devices / 7k zenperfsnmp data points here
[22-Jul-2009 14:56:15] <jb>  14:56:11 up 62 days, 22:19,  1 user,  load average: 0.17, 0.19, 0.13
[22-Jul-2009 14:58:11] <twm1010> ~240 device, 10.5k datapoints... but alot of those are ping only devices
[22-Jul-2009 14:58:25] <jb> i wonder how many zenwinperf datapoints I have..
[22-Jul-2009 14:58:29] <jb> no graphs for that
[22-Jul-2009 14:58:37] <twm1010> that would be handy
[22-Jul-2009 14:58:46] <sacpinball> 157 devices up 239 days, 13:22,  1 user,  load average: 0.37, 0.27, 0.35
[22-Jul-2009 14:59:04] <jb> my box is a quad core xeon with 32GB of ram
[22-Jul-2009 14:59:12] <mrayzenoss> Updated the NetApp, Reports Portlet and SuSE ZenPacks are now available
[22-Jul-2009 14:59:26] <twm1010> mrayzenoss: the same netapp one from the other day?
[22-Jul-2009 14:59:46] <mrayzenoss> yeah, got a fix from venturaville
[22-Jul-2009 15:00:05] <johns55> any of you guys noticed an error when running zenoss off IE8?
[22-Jul-2009 15:00:12] <mrayzenoss> yeah, IE8 is unsupported
[22-Jul-2009 15:00:15] <twm1010> excellent, it failed install the other day.
[22-Jul-2009 15:00:18] <mrayzenoss> we'll add it in the next release
[22-Jul-2009 15:00:26] <twm1010> johns55: recommend FF3 or 3.5
[22-Jul-2009 15:00:49] <mrayzenoss> 3.5 isn't supported yet, but seems to work just fine
[22-Jul-2009 15:00:50] <johns55> mrayzenoss: okay cool, yeah I'm running FF 3.5 now but the event menu looks not aligned
[22-Jul-2009 15:00:52] <mrayzenoss> IE8 however...
[22-Jul-2009 15:01:00] <johns55> looks like a CSS mistake or something
[22-Jul-2009 15:01:16] <johns55> everyone else looks perfect though
[22-Jul-2009 15:01:23] <mrayzenoss> yeah, FF3.5 and IE8 should be supported officially in the next release
[22-Jul-2009 15:01:52] <mrayzenoss> Chrome and Safari may get officially supported someday
[22-Jul-2009 15:02:15] <jb> what about ie6?!
[22-Jul-2009 15:02:40] <twm1010> and lynx?!
[22-Jul-2009 15:02:50] <mrayzenoss> IE6 never
[22-Jul-2009 15:03:32] <rmatte> ie6 == upgrade to something newer or use a real web browser like firefox
[22-Jul-2009 15:03:32] <mrayzenoss> gotta run, BBIAB
[22-Jul-2009 15:04:02] <rmatte> the only thing that I liked about ie6 is that you could actually upload files to ftp servers with it, they took that functionality out in ie7 and ie8 for whatever reason
[22-Jul-2009 15:04:20] <twm1010> ah, cheers, NetApp ZP installs now...
[22-Jul-2009 15:04:21] <jb> ie6 is terrible from a webdev's perspective
[22-Jul-2009 15:04:21] <twm1010>
[22-Jul-2009 15:04:43] <rmatte> yeh, ie6 seems to do everything it's own way
[22-Jul-2009 15:07:06] <blocky> guys where can i find info about the hp-snmp-agent oids
[22-Jul-2009 15:07:28] <twm1010> blocky: the mibs probably come in a folder in the insight manager installation
[22-Jul-2009 15:07:29] <blocky> like there are 8 different temperature sensors, where are they located etc
[22-Jul-2009 15:07:53] <blocky> i unzipped the mibs to /usr/share/snmp/mibs
[22-Jul-2009 15:08:00] <blocky> but theres a ton of em and i have no idea how to read them
[22-Jul-2009 15:08:04] <straterra> Is there anyway to rename a notification?
[22-Jul-2009 15:08:09] <rmatte> hmmm, our cycle collector times dropped recently
[22-Jul-2009 15:08:22] <rmatte> could be a good thing or a bad thing
[22-Jul-2009 15:08:25] <straterra> Like.. CPU Idle notification..I'd rather get a page that a CPU is used x%, instead of free 0%
[22-Jul-2009 15:09:18] <twm1010> well. to search them and find OID's you should get a mibexplorer app that lets you explore them and validate them against a host
[22-Jul-2009 15:09:30] <rmatte> straterra: you'd have to make an event transform for it
[22-Jul-2009 15:09:41] <twm1010> http://www.serverscheck.com/mib_browser/ is what I use
[22-Jul-2009 15:09:45] <twm1010> pretty handy, but a little buggy
[22-Jul-2009 15:10:26] <blocky> text mode?
[22-Jul-2009 15:11:04] <blocky> cool thanks
[22-Jul-2009 15:11:12] <twm1010> nah, its a windows program
[22-Jul-2009 15:11:32] <twm1010> lets you drill down the trees of the OID's and then hit them against a host, like doing an SNMP walk
[22-Jul-2009 15:11:51] <twm1010> I use it when im designing templates and need to figure out what to query
[22-Jul-2009 15:15:21] <straterra> rmatte: is there a guide on that at all?
[22-Jul-2009 15:15:29] <straterra> Or..is one even needed?
[22-Jul-2009 15:15:48] <rmatte> straterra: you need to read up on event transforms in general, if you're planning on using traps
[22-Jul-2009 15:15:52] <rmatte> you'll be writing a lot of them
[22-Jul-2009 15:15:59] <straterra> Hmm..I don't use traps
[22-Jul-2009 15:16:05] <rmatte> I'd recommend reading through the Zenoss Admin guide
[22-Jul-2009 15:16:09] <straterra> Ok
[22-Jul-2009 15:18:32] <straterra> err..I just read the term regex
[22-Jul-2009 15:18:38] <straterra> God help us all if I have to use regex
[22-Jul-2009 15:20:05] <blocky> lol this mib zip from hp came with 244 mibs and their respective cfg file
[22-Jul-2009 15:20:15] <blocky> and none that says proliant
[22-Jul-2009 15:21:00] <blocky> and zero readme
[22-Jul-2009 15:24:49] <straterra> rmatte: Do you have any I can see as an example, per chance?
[22-Jul-2009 15:25:15] <straterra> I just read the section on it..but it didn't have very many actual code blocks and seemed to be more of a top-level understanding thing
[22-Jul-2009 15:26:35] <twm1010> straterra: the people in here are fairly patient, and can help you with regex
[22-Jul-2009 15:26:44] <straterra> like..this is an example.. if device and "Hawaii" in device.getLocationName():
[22-Jul-2009 15:26:57] <straterra> I'd like to see a list of properties for device, and stuff like that
[22-Jul-2009 15:27:17] <twm1010> I find this tool here, quite useful for learning RegEx...
[22-Jul-2009 15:27:18] <twm1010> http://gskinner.com/RegExr/
[22-Jul-2009 15:27:46] <rmatte> blocky: I believe the proliant zenpack comes with proliant mibs
[22-Jul-2009 15:28:00] <straterra> twm1010: That's a very cool website
[22-Jul-2009 15:28:08] <rmatte> straterra: yeh, one sec
[22-Jul-2009 15:29:05] <rmatte> straterra: http://pastebin.com/m1109e661
[22-Jul-2009 15:29:45] <straterra> Guh
[22-Jul-2009 15:30:26] <rmatte> that regexr site is neat
[22-Jul-2009 15:30:38] <rmatte> straterra: I'll break that transform down so that you understand
[22-Jul-2009 15:31:39] <rmatte> If you go to an event that comes in as a trap and click the details link on the right (magnifying glass), and then go to the details tab, you'll see values that came in (provided you have the Mibs) installed. The top of the transform is setting variables to those values
[22-Jul-2009 15:32:11] <rmatte> so when the trap comes in clogHistSeverity is the severity, clogHistMsgText is text describing the event, etc...
[22-Jul-2009 15:32:20] <straterra> err..
[22-Jul-2009 15:32:22] <blocky> rmatte: yea the temperature sensor template is causing an error
[22-Jul-2009 15:32:24] <straterra> Let me go take a look
[22-Jul-2009 15:32:26] <blocky> and thats the monitor i really want
[22-Jul-2009 15:32:30] <rmatte> straterra: k
[22-Jul-2009 15:32:34] <blocky> do you know how i can take a look at it and see if i see the error?
[22-Jul-2009 15:32:48] <straterra> rmatte: I don't have any traps configured, unfortunately
[22-Jul-2009 15:32:51] <straterra> I..don't really use them
[22-Jul-2009 15:32:55] <rmatte> blocky: what's the entire error?  pastebin it
[22-Jul-2009 15:33:37] <twm1010> Mmm... getting good info from my NetApp boxes, now if only it modeled the filesystems..
[22-Jul-2009 15:34:16] <rmatte> straterra: http://devarthur.blogspot.com/2008/04/zenoss-percentage-used-for.html
[22-Jul-2009 15:34:25] <rmatte> check that out, that might be basically what you're looking for
[22-Jul-2009 15:34:30] <straterra> Alright, will do
[22-Jul-2009 15:34:39] <blocky> User-supplied Python expression (here.threshold) for maximum value caused error: ['TemperatureCelsius_temperature_celsius']
[22-Jul-2009 15:34:43] <blocky> thats it
[22-Jul-2009 15:34:56] <blocky> thats all my perf page shows
[22-Jul-2009 15:35:03] <rmatte> blocky: is there a details link you can click for more?
[22-Jul-2009 15:35:32] <rmatte> that error in itself is basically useless
[22-Jul-2009 15:35:52] <rmatte> unzip the zenpack file, then look for python files
[22-Jul-2009 15:35:57] <rmatte> one of them will be the collector plugin
[22-Jul-2009 15:36:07] <rmatte> that's what you should check
[22-Jul-2009 15:36:18] <blocky> alright
[22-Jul-2009 15:36:29] <rmatte> then look for the same file in your install
[22-Jul-2009 15:36:32] <twm1010> where under $ZENHOME are the check_commandgoeshere plugins kept?
[22-Jul-2009 15:37:15] <straterra> rmatte: Ok..that makes sense..I just have to look up like how I would get CPU percentage..what properties
[22-Jul-2009 15:37:32] <straterra> I'm reading over what you said before..and I'm looking for the details tab
[22-Jul-2009 15:37:42] <rmatte> in $ZENHOME/bin
[22-Jul-2009 15:37:43] <rmatte> I believe
[22-Jul-2009 15:38:35] <rmatte> straterra: why would you want to transform cpu usage alerts?
[22-Jul-2009 15:38:36] <blocky> i found "HPTemperatureSensor.py"
[22-Jul-2009 15:38:46] <rmatte> blocky: that's the one
[22-Jul-2009 15:38:48] <blocky> but i think there must be more, this file has no mention of OIDs, for example
[22-Jul-2009 15:38:53] <straterra> rmatte: Because I want a page on how much the CPU is using, not how idle the cpu is
[22-Jul-2009 15:38:56] <rmatte> really?
[22-Jul-2009 15:38:58] <straterra> The page I get now is that the cpu is 0% idle
[22-Jul-2009 15:39:12] <straterra> That isn't an issue for me personally..but I know it's going to get brought up by like the CIO
[22-Jul-2009 15:39:31] <rmatte> straterra: you just need to poll a different OID for that
[22-Jul-2009 15:39:37] <straterra> Oh
[22-Jul-2009 15:40:02] <rmatte> straterra: what type of device is this?
[22-Jul-2009 15:40:02] <straterra> I'm sure that's easy to find
[22-Jul-2009 15:40:11] <straterra> Windows/linux
[22-Jul-2009 15:40:32] <rmatte> straterra: ah, snmp monitoring is tricky in Windows, you need to install snmp-informant to make it easier
[22-Jul-2009 15:40:39] <straterra> It's installed
[22-Jul-2009 15:40:49] <straterra> snmp-informant is installed on all Windows servers now
[22-Jul-2009 15:40:55] <rmatte> I'm doing it with the basic windows snmp agent, but I'm got a bunch of scripts collecting the data
[22-Jul-2009 15:41:00] <rmatte> cool
[22-Jul-2009 15:41:13] <straterra> But, I imagine it's not the same OID on both Linux and windows
[22-Jul-2009 15:41:18] <rmatte> find out what OID to poll for CPU info from snmp-informant
[22-Jul-2009 15:42:06] <blocky> oh
[22-Jul-2009 15:42:06] <blocky> okay i found another python file with the same name
[22-Jul-2009 15:42:52] <straterra> rmatte: I should be able to just look in the MIB files on one of the Windows servers for snmp-informant, right?
[22-Jul-2009 15:43:00] <blocky> what is the string i append to a url in zenoss to edit that page
[22-Jul-2009 15:43:24] <rmatte> straterra: just google for it
[22-Jul-2009 15:43:39] <rmatte> blocky: /manage at the end of the url
[22-Jul-2009 15:43:54] <rmatte> blocky: but be very careful, easy to nuke your Zenoss using the management console
[22-Jul-2009 15:44:24] <rmatte> I've seen people remove all permissions from the admin account before and totally fubar their zenoss lol
[22-Jul-2009 15:44:44] <blocky> lol okay ill be careful
[22-Jul-2009 15:44:50] <blocky> im gonna try disabling one or both threshholds
[22-Jul-2009 15:45:01] <blocky> cause its complaining about a max value formula
[22-Jul-2009 15:45:17] <rmatte> why are you going in to the management console to change threshold settings?
[22-Jul-2009 15:45:51] <rmatte> you do that right from the template page
[22-Jul-2009 15:46:08] <blocky> oh i found the problem
[22-Jul-2009 15:46:11] <jb> yeah, do that from teh template
[22-Jul-2009 15:46:11] <blocky> there was no value defined for max
[22-Jul-2009 15:46:16] <blocky> it just said here.threshhold
[22-Jul-2009 15:46:48] <rmatte> yeh, but here.threshhold might be pulled from the .py file?
[22-Jul-2009 15:46:53] <rmatte> did you see any reference to it?
[22-Jul-2009 15:47:44] <rmatte> I'm amazed that zenpack isn't working for you, I just dropped it in and it started working right away
[22-Jul-2009 15:47:45] <straterra> rmatte: .1.3.6.1.4.1.9600.1.1.5.1.5.1, according to http://www.opennms.org/wiki/SNMP_Informant_How-To
[22-Jul-2009 15:47:56] <rmatte> straterra: yeh, that looks right
[22-Jul-2009 15:48:03] <blocky> rmatte: dont think so
[22-Jul-2009 15:48:26] <blocky> capacity used and fan speed have no graph
[22-Jul-2009 15:48:30] <blocky> temperature has missing rrd files
[22-Jul-2009 15:48:31] <blocky> lol
[22-Jul-2009 15:48:39] <blocky> ive seen this problem before...
[22-Jul-2009 15:48:50] <bigegor> blocky: check name of DataPoint in TemperatureCelsius threshold it must be "celsius" not "temperature"
[22-Jul-2009 15:49:37] <blocky> it's TemperatureCelsius
[22-Jul-2009 15:49:45] <blocky> im not sure if i have insight installed though
[22-Jul-2009 15:50:18] <blocky> i do have all the hp mibs
[22-Jul-2009 15:51:45] <straterra> rmatte: Looks like the Windows template is already properly using CPU usage %
[22-Jul-2009 15:51:51] <straterra> It's just the linux one that isn't
[22-Jul-2009 15:53:56] <straterra> net-snmp doesn't give out a user time + system time type of output..so..I may just tell the CIO that page is going to have to stay that way
[22-Jul-2009 15:54:28] <rmatte> really?  mine shows cpu utilization
[22-Jul-2009 15:54:43] <straterra> net-snmp gives user time, system time, and idle time
[22-Jul-2009 15:54:55] <rmatte> Load Average, CPU Utilization, Memory Utilization, and IO
[22-Jul-2009 15:55:01] <rmatte> all over SNMP
[22-Jul-2009 15:55:09] <straterra> The cpu 'usage' page for Linux looks at the idle time (since that is 100 - system - user)
[22-Jul-2009 15:55:11] <rmatte> and this is a stock template
[22-Jul-2009 15:55:49] <rmatte> yeh, and the other values are the total if you add them up
[22-Jul-2009 15:56:05] <straterra> Right
[22-Jul-2009 15:56:12] <jb> i have system/user/wait/nice/interrupt
[22-Jul-2009 15:56:17] <jb> on my linux CPU Utilization graphs
[22-Jul-2009 15:56:28] <rmatte> I just have system, user, wait, and idle
[22-Jul-2009 15:56:45] <straterra> But instead of it notifying me on a low idle %, I'd rather it notify me on a high total cpu usage %
[22-Jul-2009 15:58:00] <rmatte> If you find an OID that gives you the total CPU usage, then just modify the template, add it in, and threshold against that
[22-Jul-2009 15:58:35] <rmatte> I personally don't spend time on stuff like that since it's the same result either way, you're getting notified about high CPU usage one way or another
[22-Jul-2009 15:58:56] <straterra> Right
[22-Jul-2009 15:59:03] <straterra> I'm on the same page as you in that regard
[22-Jul-2009 16:00:00] <straterra> I think I'll just tell them that's how it is
[22-Jul-2009 16:00:41] <blocky> okay when i query the temperature i get 7 differetn values
[22-Jul-2009 16:00:45] <blocky> are those 7 sensors?
[22-Jul-2009 16:00:57] <rmatte> it's pretty simple, an alert comes in, something is wrong, just explain that Zenoss doesn't do false positives in terms of thresholds
[22-Jul-2009 16:01:03] <rmatte> if they get a threshold, there's an issue
[22-Jul-2009 16:01:22] <straterra> Right
[22-Jul-2009 16:01:35] <rmatte> blocky: yeh, probably
[22-Jul-2009 16:01:41] <bigegor> blocky: rename DataPoint in /manage page back to "temperature_celsius"
[22-Jul-2009 16:02:37] <blocky> and then update the thresholds and graph?
[22-Jul-2009 16:03:51] <blocky> now there's just two missing rrd files :|
[22-Jul-2009 16:04:09] <bigegor> no, simple rename datapoint, but not datasource
[22-Jul-2009 16:04:28] <blocky> well i did
[22-Jul-2009 16:04:35] <blocky> and now the thresholds and graph say missing
[22-Jul-2009 16:05:53] <blocky> how do i rename a data point
[22-Jul-2009 16:05:54] <bigegor> what version of Zenoss do you have?
[22-Jul-2009 16:06:15] <blocky> 2.4.2
[22-Jul-2009 16:07:00] <bigegor> in 2.4.2 you can see name of datapoint ONLY in /manage page of DataSource
[22-Jul-2009 16:07:04] <blocky> i dont understand what you're asking me to do
[22-Jul-2009 16:07:35] <blocky> okay the data point is already named temperature_celsius
[22-Jul-2009 16:08:01] <bigegor> DataPoint or DataSource?
[22-Jul-2009 16:08:08] <blocky> point
[22-Jul-2009 16:08:21] <blocky> the source is TemperatureCelsius
[22-Jul-2009 16:08:44] <straterra> It's home time..thanks for all the help
[22-Jul-2009 16:09:24] <bigegor> now its right.
[22-Jul-2009 16:10:24] <blocky> okay
[22-Jul-2009 16:10:27] <blocky> how do i get the rrd files created
[22-Jul-2009 16:11:56] <jb> just wait for a few cycles..
[22-Jul-2009 16:12:07] <blocky> im pretty sure thats what it was before
[22-Jul-2009 16:13:09] <bigegor> can you seen temperature sensors on Hardware Tab?
[22-Jul-2009 16:16:17] <blocky> nope
[22-Jul-2009 16:16:23] <blocky> memory and cpu only
[22-Jul-2009 16:18:42] <bigegor> add community.snmp.HPTemperatureSensorMap plugin to zCollectorPlugins and remodel your Device
[22-Jul-2009 16:22:18] <blocky> done
[22-Jul-2009 16:22:51] <blocky> okay i see 4 cpu temp sensors and a few others but they all say the temp is unknown
[22-Jul-2009 16:22:58] <blocky> should i give it a few minutes to poll
[22-Jul-2009 16:23:08] <blocky> the rrd files are still missing
[22-Jul-2009 16:23:09] <bigegor> wait 5 min
[22-Jul-2009 16:25:54] <bigegor> unbind HPTemperatureSensor template from device
[22-Jul-2009 16:26:35] <blocky> okay
[22-Jul-2009 16:28:56] <bigegor> read "Installation Notes" here trac-zenpacks/wiki/ZenPackscommunityHPMon
[22-Jul-2009 16:35:21] <blocky> temperatuers are showing up now
[22-Jul-2009 16:35:27] <blocky> thanks bigegor
[22-Jul-2009 16:35:43] <blocky> now do i have to readd my own template to set thresholds and alerts?
[22-Jul-2009 16:36:32] <blocky> my io board is at 45 degrees celsius, is that normal?
[22-Jul-2009 16:36:41] <bigegor> You're welcome.
[22-Jul-2009 16:37:10] <rmatte> 45 sounds fine
[22-Jul-2009 16:37:24] <bigegor> yes
[22-Jul-2009 16:37:52] <rmatte> most production server equipment can go as high as like 75 degrees before it becomes dangerous, and 60 is high but safe, so 45 is fine
[22-Jul-2009 16:39:01] <blocky> what do you think i should set the thresholds at
[22-Jul-2009 16:39:23] <Blehk> Does zenoss run through apache as a wsgi? Or is it a plone application? (I've heard multiple answers so trying to find out)
[22-Jul-2009 16:41:25] <rmatte> blocky: threshold for like 60 probably
[22-Jul-2009 16:41:28] <rmatte> or 65
[22-Jul-2009 16:42:44] <bigegor> you can find actual threshold on /manage site from temperature sensor, this value from SIM agent.
[22-Jul-2009 16:46:46] <rmatte> eugh, there needs to be an option to globally push settings to the collector, instead of having to do it on a device per device basis
[22-Jul-2009 16:47:52] <rmatte> you can push changes in /Devices, but I don't think it's a global push
[22-Jul-2009 16:50:38] <rmatte> ah, apparently that does actually work, nice
[22-Jul-2009 17:06:17] <rmatte> Blehk: it's a Zope application
[22-Jul-2009 17:06:41] <Blehk> Perfect, thanks.
[22-Jul-2009 17:07:10] <rmatte> np
[22-Jul-2009 17:08:11] <johns55> hey guys, I just finished install the zenoss stack installer, and now testing out the alerts, when I try to do a email test it says socket error, does this mean I need to install sendmail?
[22-Jul-2009 17:13:24] <johns55> yup sendmail it was
[22-Jul-2009 17:26:52] <rmatte> man, they need to get a new server hosting the downloads for Zenoss, it is going so slow, I'm on a link capable of downloading upwards of 7MB/s, and it's coming down at just under 100KB/s
[22-Jul-2009 17:39:21] <johns55> wowsers 7 MB /s
[22-Jul-2009 17:39:24] <johns55> thats insane
[22-Jul-2009 17:39:29] <johns55> 1 mp3 in half a second?
[22-Jul-2009 17:40:11] <rmatte> yeh, I was downloading an Ubuntu CD image from the university of waterloo mirror (I'm in Ottawa, Canada), and it was coming down at between 5 and 7 MB/s
[22-Jul-2009 17:41:00] <rmatte> I had the whole image in like 2 or 3 minutes
[22-Jul-2009 17:41:02] <rmatte> hehe
[22-Jul-2009 17:41:38] <johns55> thats insane, edu connections are pretty awesome
[22-Jul-2009 17:41:53] <johns55> what isp are you on may I ask?
[22-Jul-2009 17:42:12] <rmatte> it's actually a fiber link, I forget who the provider is
[22-Jul-2009 17:42:24] <rmatte> (my work's connection)
[22-Jul-2009 17:42:49] <johns55> oo lol fibre i presume then
[22-Jul-2009 17:43:11] <johns55> nice
[22-Jul-2009 17:43:24] <rmatte> yeh
[22-Jul-2009 17:46:15] <blocky> how do i get a threshold to trigger an email
[22-Jul-2009 17:46:41] <rmatte> go to Settings -> Users -> click on a username -> Alerts tab
[22-Jul-2009 17:46:47] <rmatte> you can setup alerts based on filters
[22-Jul-2009 17:47:04] <rmatte> so like, if event message contains: threshold - email to this address
[22-Jul-2009 17:47:11] <rmatte> anyways, I'm out of here, cheers
[22-Jul-2009 17:47:46] <blocky> ciao
[23-Jul-2009 00:00:46] [disconnected at Thu Jul 23 00:00:46 2009]
[23-Jul-2009 00:00:46] [connected at Thu Jul 23 00:00:46 2009]
[23-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[23-Jul-2009 07:03:03] <tehhobbit> anyone got a good idea how to create agregate graphs unders the perf tab for switches ( total traffic from the backplane ) ?
[23-Jul-2009 07:05:16] <jmp242> I don't know that you could do it on the perf tab without a bunch of work, but you probably could use multi-graph reports
[23-Jul-2009 07:06:29] <tehhobbit> jmp242: reason for doing under the perf tab is that I want to automaticly get all the switch ports when I add a new switch
[23-Jul-2009 07:59:13] <stefanlsd> When are the devs in?
[23-Jul-2009 07:59:26] <W9ZEB> stefanlsd: according to the topic, 11:00 EDT
[23-Jul-2009 07:59:47] <W9ZEB> 15:00 UTC
[23-Jul-2009 08:01:03] <stefanlsd> W9ZEB: thanks. i never know the time zones. thats another 2 hours it seems
[23-Jul-2009 08:05:15] <W9ZEB> I have a problem that I'm hoping someone might be able to assist me with. One of my servers, running FreeBSD, has ~30 Jails. The root server has 2 IP addresses aliased per jail, plus it's own two IP addresses. If I put the base server in monitoring, I am not allowed to add any of the jails, as the IP address is discovered by snmp, and already assigned to the root server.
[23-Jul-2009 08:05:35] <W9ZEB> any suggestions on how to monitor the jails individually, and still be able to monitor the root server?
[23-Jul-2009 08:17:44] <tehhobbit> W9ZEB: hmmmm didnt have that problem with solaris and zones, seems asif snmp doesnt report the ip's of the zones
[23-Jul-2009 08:22:13] <W9ZEB> tehhobbit: do you alias the zone's IP's on the base server?
[23-Jul-2009 08:22:23] <W9ZEB> tehhobbit: or are they only assigned inside the zone?
[23-Jul-2009 08:22:32] * W9ZEB isn't familiar with Solaris & Zones
[23-Jul-2009 08:23:50] <dollarbang1> good morning all, can someone tell me how to temporarily stop monitoring a device?
[23-Jul-2009 08:24:01] <tehhobbit> W9ZEB: you assigne a alias of a nic to the zone
[23-Jul-2009 08:24:11] <tehhobbit> still shows up from the base server
[23-Jul-2009 08:24:17] <tehhobbit> but cant be accessed
[23-Jul-2009 08:24:36] <tehhobbit> I think you could do something simular to jails
[23-Jul-2009 08:25:03] <W9ZEB> dollarbang1: edit the device, and put it in Maintanence mode.
[23-Jul-2009 08:26:33] <W9ZEB> tehhobbit: I'll have to do additional reading I suppose.
[23-Jul-2009 08:26:52] <W9ZEB> the alternative is create a template for Jails servers that doesn't auto-discover IP addresses perhaps?
[23-Jul-2009 08:27:43] <tehhobbit> or maybe you can tweak the snmp confignot to show nics with a higher inderx than what the base servers nic have
[23-Jul-2009 08:27:51] <tehhobbit> not to show*
[23-Jul-2009 08:28:17] <tehhobbit> W9ZEB: I removedeverything but app monitor for the zones
[23-Jul-2009 08:28:49] <tehhobbit> since performance and such if fine to get from base server
[23-Jul-2009 08:28:54] <W9ZEB> I'm actually interested in monitoring everything, except swap from the jails (as jails don't report swap properly) which means I need to monitor swap on the base server
[23-Jul-2009 08:29:13] <stefanlsd> Maybe someone can help me 2 - I've written a custom command to log to our helpdesk. Im using a transform on a ping to suppress any further notifications on a device which works, but it seems like it still runs the custom command for each device it should suppress. anyone have any ideas? ie. how can i suppress the custom command also...
[23-Jul-2009 08:30:28] <tehhobbit> W9ZEB: Products/DataCollector/plugins/zenoss/snmp/InterfaceMap.py is something you should checkout and work from there
[23-Jul-2009 08:36:59] <W9ZEB> tehhobbit: thanks, I'll take a look
[23-Jul-2009 08:40:01] <nzle> I setup a heartbeat per the notes at http://www.zenoss.com/community/wiki/tips-and-tricks/HowToUseZenossHeartbeatsForYourOwnMonitoredDevices and it seems to be generating heartbeat failures now, but the failures dont clear when another heartbeat comes in. Do I need to setup somthing to clear these in the heartbeat event?
[23-Jul-2009 08:40:32] <twm1010> anyone have the linky handy to the command line syntax to test WMI?
[23-Jul-2009 08:45:13] <twm1010> Odd... the domain admin account is getting denied WMI access to a box
[23-Jul-2009 08:55:57] <dollarbang1> W9ZEB: thanks, did that, works like a charm.
[23-Jul-2009 08:58:01] <W9ZEB> dollarbang1: in Maint mode, it'll continue to capture failures.
[23-Jul-2009 08:58:04] <W9ZEB> but it won't alert on them.
[23-Jul-2009 08:59:59] <dollarbang1> W9ZEB: the problem was the amount of SNMP requests, there were too many, and since they were failures, Zenoss would attempt every 5 minutes. That maybe a setting issue, but for now it is alleviated.
[23-Jul-2009 09:13:19] <nzle> I think I see my heartbeat failure problem, for testing I had my timeout set low, 30 seconds. I think the incoming beat takes a bit of time to be noticed by the heartbeat monitor, and it was failing again before it cleared. Now that I increased it to 300 it seems to be behaving properly. Now how do I get it to fail red instead of orange. Up the severity on the beat event?
[23-Jul-2009 09:38:31] <twm1010> Anyone have any insight to the following error? Can't locate Net/SNMP.pm
[23-Jul-2009 09:38:55] <jb> ..
[23-Jul-2009 09:39:16] <deppy> wtf
[23-Jul-2009 09:39:39] <deppy> i'm trying to get some info on how to install nagios
[23-Jul-2009 09:39:59] <jb> o rly
[23-Jul-2009 09:40:36] <twm1010> A part of the NetApp Zenpack is failing to use the check_netapp script
[23-Jul-2009 09:40:39] <twm1010> That's my error
[23-Jul-2009 09:41:37] <nzle> twm010: looks like you need to install the Net::SNMP perl module on that box.
[23-Jul-2009 09:44:20] <twm1010> yeah, i found an RPM, but of course it has dependencies...
[23-Jul-2009 09:44:26] <twm1010> and i bet those dependencies have dependencies
[23-Jul-2009 09:47:20] <twm1010> nzle: groovy, i'm square now
[23-Jul-2009 09:52:19] <twm1010> Where do you set the default FROM users for zenoss email notifications?
[23-Jul-2009 09:52:29] <twm1010> The sender address.
[23-Jul-2009 09:58:39] <jb> sup matt
[23-Jul-2009 09:58:53] <mrayzenoss> morning all
[23-Jul-2009 09:59:08] <bedwards> hello
[23-Jul-2009 09:59:08] <jmp242> morning
[23-Jul-2009 09:59:10] <mrayzenoss> bedwards and I will be answering questions this morning, other Zenossians may join in
[23-Jul-2009 09:59:32] <bwest> hello all
[23-Jul-2009 10:01:31] <jb> mrayzenoss: it would be nice to get zenwinperf stats on the collector graphs..
[23-Jul-2009 10:02:37] <twm1010> jb: I bet you can add it.
[23-Jul-2009 10:02:42] <toninog> zenback/zenrestore... does anyone use it ?
[23-Jul-2009 10:03:07] <mrayzenoss> jb: that sounds like a good ticket or enhancement suggestion, you can enter it on the customer portal
[23-Jul-2009 10:03:42] <jb> k
[23-Jul-2009 10:04:01] <toninog> hi all...
[23-Jul-2009 10:04:06] <twm1010> but zenwinperf is a daemon in enterprise
[23-Jul-2009 10:04:16] <twm1010> specifically for collecting perf data over WMI?
[23-Jul-2009 10:04:23] <toninog> I was wondering if anyone had any sucess stories for using drbd ?
[23-Jul-2009 10:04:29] <stefanlsd> What kind of questions will you guys be answering? I have a technical question...
[23-Jul-2009 10:04:41] <chudler> toninog: we use drbd with success
[23-Jul-2009 10:04:49] <LiamW> They answer all questions
[23-Jul-2009 10:04:53] <mrayzenoss> twm1010: yes
[23-Jul-2009 10:05:01] <mrayzenoss> stefanlsd: whatever you want to ask
[23-Jul-2009 10:05:16] <toninog> chudler: mike here is frightfully afraid of it... is there any compelling evidence I can use to attack his fear?
[23-Jul-2009 10:05:42] <LiamW> mray: When monitoring SSH devices how do I get zenoss to create an event when the interface is down?
[23-Jul-2009 10:06:01] <jb> http://dev.zenoss.org/trac/ticket/5357
[23-Jul-2009 10:06:09] <chudler> probably not, it is a terrible situation to be afraid of technology. Perhaps if mike could elucidate on his fears.
[23-Jul-2009 10:06:32] <stefanlsd> thanks! I've written a custom command to log to our helpdesk. Im using a transform on a ping to suppress any further notifications on a device which works, (we get one email) but it seems like it still runs the custom command for each device it should suppress. ie. how can i suppress the custom command also...
[23-Jul-2009 10:06:40] <chudler> drbd is not just something someone cooked up in their basement, FWIW
[23-Jul-2009 10:06:56] <toninog> chudler: hehe - hence is fear
[23-Jul-2009 10:07:22] <toninog> its is a pity that the zoDB cannot be distributed like a mysql (or any other sql server) can
[23-Jul-2009 10:08:03] <bedwards> LiamW: you need to add an event to the result object returned by the parser
[23-Jul-2009 10:08:14] <mrayzenoss> LiamW: is there a ping down event associated with the interface?
[23-Jul-2009 10:08:17] <chudler> it is very well thought out, clean and performs well. There are probably thousands of happy users out there. I think it will be in the mainline kernel Real Soon Now
[23-Jul-2009 10:09:06] <LiamW> mray: I am not receiveing any events when an interface goes down, whether it is SNMP or SSH
[23-Jul-2009 10:09:07] <toninog> chudler: have you had a time when you actually tested its failover and did so successfully?
[23-Jul-2009 10:09:15] deppy is now known as christel
[23-Jul-2009 10:09:16] <chudler> toninog: yes
[23-Jul-2009 10:09:30] <toninog> chudler: is your drbd systems geographically dispersed ?
[23-Jul-2009 10:09:33] christel is now known as deppy
[23-Jul-2009 10:10:18] <jb> be careful with DRBD if your boxes aren't local to each other.. DRBD isn't very happy over high latency links.
[23-Jul-2009 10:10:35] <chudler> toninog:  no
[23-Jul-2009 10:10:50] <toninog> jb: have you had experience with the drbd over low latency links ?
[23-Jul-2009 10:10:57] <jb> yep
[23-Jul-2009 10:11:03] <jb> i use DRBD for several heartbeat clusters
[23-Jul-2009 10:11:12] <jb> im talking < 5ms though
[23-Jul-2009 10:11:25] <toninog> jb: ok - and do you have any scary stories ?
[23-Jul-2009 10:11:27] <rmatte> when using the stack installer to upgrade, it shouldn't write over my current zenoss config files (in $ZENHOME/etc)? Or will it?
[23-Jul-2009 10:11:42] <jb> nope
[23-Jul-2009 10:11:48] <jb> occasinally a split brain occurs..
[23-Jul-2009 10:11:52] <toninog> jb: excellent.. is it just lag then ?
[23-Jul-2009 10:12:38] <jb> yeah, just test it..
[23-Jul-2009 10:13:14] <toninog> jb: excellent... I think I will
[23-Jul-2009 10:13:19] <LiamW> rocket: will you be adding memory to the Linux monitor addon zenpack?
[23-Jul-2009 10:13:34] <mrayzenoss> rmatte: there may be a ticket about that
[23-Jul-2009 10:13:59] <mrayzenoss> LiamW: what kind of memory?
[23-Jul-2009 10:14:07] <mrayzenoss> LiamW: like a perf graph?
[23-Jul-2009 10:14:12] <rmatte> mrayzenoss: just looking for a will or won't, that way I'll know whether or not I need to back them up and restore them afterwards (no big deal if I do)
[23-Jul-2009 10:14:26] <rmatte> I guess I'll know after this upgrade lol
[23-Jul-2009 10:14:51] <LiamW> buffer, cache, used  etc
[23-Jul-2009 10:15:37] <toninog> chudler: jb: thanks for the info ...
[23-Jul-2009 10:15:58] <mrayzenoss> rmatte: http://dev.zenoss.org/trac/ticket/4967 Looks like it's supposed to work and we haven't had customer issues with it, but someone lost their etc on a failed upgrade
[23-Jul-2009 10:16:03] <mrayzenoss> so back it up to be safe
[23-Jul-2009 10:16:13] <rmatte> mrayzenoss: ok, sounds good, thanks
[23-Jul-2009 10:17:11] <rmatte> I finally got this clones vm to work on our test box, it was geared up to run on one of our super servers in our blade enclosure at a data center, then I copied it down to a server in our lab environment and it kept crashing lol
[23-Jul-2009 10:17:19] <rmatte> cloned*
[23-Jul-2009 10:17:22] <mrayzenoss> LiamW: You can open tickets against community ZenPacks in the Community ZenPack repository: http://community.zenoss.org
[23-Jul-2009 10:17:35] <LiamW> ok
[23-Jul-2009 10:17:37] <mrayzenoss> LiamW: and submit patches
[23-Jul-2009 10:17:44] <twm1010> Is there a good tutorial/guide on creating custom device reports?
[23-Jul-2009 10:18:16] <twm1010> I'm looking to create one that shows me all the gigabit ports on my core switches that are currently linked at 100Mb.
[23-Jul-2009 10:18:26] <bedwards> stefanlsd: I don't know of a way to suppress custom commands. you can open an enhancement request.
[23-Jul-2009 10:18:45] <rmatte> twm1010: you'd probably have to code that
[23-Jul-2009 10:19:02] <rmatte> twm1010: using the easy report interface under the edit tab in custom reports is very limited
[23-Jul-2009 10:19:17] <mrayzenoss> twm1010: http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch14.html and http://www.zenoss.com/community/docs/zenoss-guide/2.4.2/ch12s08.html
[23-Jul-2009 10:19:30] <mrayzenoss> twm1010: King Crab will have some more hooks to make zendmd calls easier
[23-Jul-2009 10:19:34] <LiamW> Where do I check to ensure that the zenoss cron jobs are running?
[23-Jul-2009 10:20:09] <rmatte> mrayzenoss: any chance that the report exporting gets fixed by King Crab?
[23-Jul-2009 10:20:26] <mrayzenoss> rmatte: Chet took your tickets that you opened, so those should get fixed
[23-Jul-2009 10:20:36] <rmatte> good stuff
[23-Jul-2009 10:21:08] <rmatte> the one is like a two second fix, the other one is really just Zenoss directly accessing the values in Zope rather than taking in to account the final values displayed in the report
[23-Jul-2009 10:21:31] <twm1010> mrayzenoss: any good examples out there I can use to backwards build what i need? like someone who created an interface inventory report or something?
[23-Jul-2009 10:22:16] <rmatte> twm1010: you could just look at the interface utilization report, it includes speeds of interfaces
[23-Jul-2009 10:22:17] <mrayzenoss> LiamW: not sure, you could check the stuff in $ZENHOME/logs for whatever you're calling, or I like to add "MARK-CRON" stuff to logs as a notifier that my cronjobs are running
[23-Jul-2009 10:22:37] <rmatte> twm1010: you then just have to add something to the report that only lists interfaces that are 100mbps
[23-Jul-2009 10:22:54] <twm1010> now how did i miss that...
[23-Jul-2009 10:23:09] <rmatte>
[23-Jul-2009 10:23:42] <twm1010> ah, well, excel can do that easy enough
[23-Jul-2009 10:24:07] <rmatte> yeh, the problem is that when you export to excel, it doesn't show the same values as in the report
[23-Jul-2009 10:24:16] <rmatte> instead of 100mbps you'll see like 100000000
[23-Jul-2009 10:24:19] <rmatte> or something
[23-Jul-2009 10:24:22] <twm1010> eh, thats okay
[23-Jul-2009 10:24:22] <mrayzenoss> the MAC Address report has more stuff with Interfaces for another example
[23-Jul-2009 10:24:48] <mrayzenoss> trac-zenpacks/wiki/ZenPackssodonnellMACReport
[23-Jul-2009 10:25:00] <twm1010> im looking for gigabit ports linked at 100Mbps... we've had people using up our expensive reports for management cards that only link at 100
[23-Jul-2009 10:25:00] <rmatte> twm1010: also, that export is broken for exporting, you won't see the interfaces column, but it's easy to fix
[23-Jul-2009 10:25:02] <twm1010> nice huh?
[23-Jul-2009 10:25:35] <twm1010> that particular report won't export is what you're saying?
[23-Jul-2009 10:25:36] <rmatte> twm1010: http://dev.zenoss.org/trac/ticket/5350
[23-Jul-2009 10:25:37] <mrayzenoss> feel free to send in any custom reports as ZenPacks
[23-Jul-2009 10:25:45] <rmatte> it'll export, the interfaces column will be missing
[23-Jul-2009 10:25:58] <rmatte> read that track ticket I posted, it explains how to fix it
[23-Jul-2009 10:26:12] <twm1010> mrayzenoss: I'd love to write some, but the problem is that I don't know the variables/column names to query?
[23-Jul-2009 10:26:40] <rmatte> after you edit the file, run: python $ZENHOME/Products/ZenReports/ReportLoader.py --force
[23-Jul-2009 10:26:40] <twm1010> rmatte: thanks, I'll edit the report
[23-Jul-2009 10:26:47] <rmatte> as zenoss user, to load the report
[23-Jul-2009 10:26:49] <rmatte> then it'll work
[23-Jul-2009 10:27:20] <rmatte> the report writing is pretty brutal, I'm just starting to wrap my mind around it
[23-Jul-2009 10:27:26] <jb> rocket: around?
[23-Jul-2009 10:29:39] <twm1010> Yeah I jumped over to our SolarWinds box to try and get what I need, but thanks to licensing, we only monitor some of the ports, not what I need here
[23-Jul-2009 10:30:29] <rmatte> ah, yeh, we're using solarwinds for reporting as well
[23-Jul-2009 10:30:39] <jb> we are too
[23-Jul-2009 10:30:51] <rmatte> I'd prefer to use Zenoss, but the reporting in it is a bit shaky at the moment
[23-Jul-2009 10:31:00] <rmatte> I'm sure it'll improve down the road
[23-Jul-2009 10:31:03] <jb> we monitor all network stuff with solarwinds..
[23-Jul-2009 10:31:17] <rmatte> for actual monitoring we use Zenoss
[23-Jul-2009 10:31:21] <rmatte> for reporting we use solarwinds
[23-Jul-2009 10:35:10] <rmatte> the fact that solarwinds only runs in windows is a bit of a deal breaker for the monitoring aspect
[23-Jul-2009 10:48:54] <mrayzenoss> kinda quiet after the first bunch of questions...
[23-Jul-2009 10:52:46] <jmp242> 4
[23-Jul-2009 10:53:48] <rmatte> yeh lol
[23-Jul-2009 10:54:54] <rmatte> I'm running a test upgrade right now, waiting to see how well it works. We're going to be deploying it to our production servers provided it works properly
[23-Jul-2009 10:55:02] <mrayzenoss> ok, let's see. There will probably be a 2.4.3 in a few weeks. We're still shooting for a beta testing for King Crab around late next month
[23-Jul-2009 10:55:33] <rmatte> cool
[23-Jul-2009 10:55:59] <rmatte> I'll probably load up the beta version of King crab on a VM in our lab
[23-Jul-2009 10:56:24] <mrayzenoss> excellent.  We sent out a ton of t-shirts to beta testers last release
[23-Jul-2009 10:56:51] <rmatte> are there t-shirts for this release?
[23-Jul-2009 10:57:07] <mrayzenoss> yeah, there will be beta tester t-shirts again
[23-Jul-2009 10:57:11] <rmatte> nice
[23-Jul-2009 10:57:12] <mrayzenoss> we should be putting up a demo server for feedback on the new event console soon
[23-Jul-2009 10:57:25] <mrayzenoss> Ian demoed it Monday, it's very nice
[23-Jul-2009 10:57:37] <rmatte> yeh, I can't wait to see that, does that mean the device list view is going to be a bit different too? (since it's similar)
[23-Jul-2009 10:57:43] <mrayzenoss> not yet
[23-Jul-2009 10:57:53] <rmatte> ah
[23-Jul-2009 10:57:55] <mrayzenoss> We're tackling one screen at a time
[23-Jul-2009 10:58:01] <rmatte> gotcha
[23-Jul-2009 10:58:11] <mrayzenoss> this will clear up the Event Console and Event Views
[23-Jul-2009 10:58:37] <mrayzenoss> let you save queries, better filtering and ordering, it's very slick
[23-Jul-2009 10:58:52] <rmatte> actually, I just remembered a trac ticket that I wanted to log for Zenoss auto-clearing events... It's not possible to auto-clear above error, so the Critical events never get cleared
[23-Jul-2009 11:00:11] <mrayzenoss> don't know the history of that, wonder if that's intentional
[23-Jul-2009 11:00:28] <rmatte> possibly, but seems like the user should be able to do it if they wish to
[23-Jul-2009 11:00:34] <mrayzenoss> agreed
[23-Jul-2009 11:01:15] <rmatte> in our case, we have it sending events to our ticketing system, so no one actually goes in to Zenoss to acknowledge or clear events, so we have it set to auto-clear them every hour
[23-Jul-2009 11:01:35] <rmatte> by "clear" I mean move to history
[23-Jul-2009 11:01:38] <mrayzenoss> well that's a shame, you won't get the full glory of the new Event Console
[23-Jul-2009 11:02:17] <rmatte> well, I still use the event console for tuning, but the analysts just use the ticketing system, easier to work with, plus it allows us to run multiple instances of Zenoss and have all the events displayed in one spot
[23-Jul-2009 11:02:25] <rmatte> and we can sort alerts out by queue if need be
[23-Jul-2009 11:03:37] <rmatte> where we're lacking right now is in the reporting aspect, which is something we're working on
[23-Jul-2009 11:04:19] <mrayzenoss> yeah, reporting just missed the cut for King Crab, it's a known weakness
[23-Jul-2009 11:04:36] <rmatte> well, the upgrade didn't trash my ldap settings, that's good news
[23-Jul-2009 11:05:15] <rmatte> it'd be nice to see a better gui based report designer
[23-Jul-2009 11:05:38] <rmatte> where you can pick common components, such as intefaces, disk, etc... and build reports then apply filters or whatever and save the report
[23-Jul-2009 11:05:50] <mrayzenoss> that would be nice
[23-Jul-2009 11:06:17] <rmatte> By the way, when I get the message that the google maps API is not valid, the map loads anyways, so how necessary is that actual key?
[23-Jul-2009 11:06:28] <mrayzenoss> that seems odd
[23-Jul-2009 11:06:35] <mrayzenoss> do you have an API key in there?
[23-Jul-2009 11:06:37] <rmatte> I've never had the map not load from a bad key
[23-Jul-2009 11:06:42] <rmatte> yeh, there's a key in there
[23-Jul-2009 11:07:04] <rmatte> it just complains that it's not valid (since I'm accessing the box using a different IP/hostname than it is on the prod version of the image)
[23-Jul-2009 11:07:07] <rmatte> but the map loads anyways
[23-Jul-2009 11:07:28] <mrayzenoss> perhaps the API key was made by IP and it changed?  Or by hostname and the IP changed?
[23-Jul-2009 11:07:34] <mrayzenoss> yup
[23-Jul-2009 11:07:37] <rmatte> yeh, that's probably it
[23-Jul-2009 11:07:54] <rmatte> but the fact that it still works means you could potentially just use 1 key, and just find a way to eliminate that error message
[23-Jul-2009 11:07:54] <mrayzenoss> minor annoyance, I don't know if there's a fix
[23-Jul-2009 11:07:58] <rmatte> provided it works regardless
[23-Jul-2009 11:08:09] <mrayzenoss> possibly, I don't know
[23-Jul-2009 11:08:19] Hunner_ is now known as Hunner
[23-Jul-2009 11:08:20] <mrayzenoss> maybe they figure the annoyance is enough
[23-Jul-2009 11:08:21] <rmatte> does the enterprise version just generate keys on the fly?
[23-Jul-2009 11:08:46] <mrayzenoss> I'm not sure, I think they provide the API key for you
[23-Jul-2009 11:09:04] <rmatte> ah
[23-Jul-2009 11:09:08] <mrayzenoss> I really don't have any exposure to that part
[23-Jul-2009 11:09:18] <rmatte> alright hehe, just curious
[23-Jul-2009 11:09:36] <rmatte> ummm weird
[23-Jul-2009 11:10:05] <rmatte> oh wait, nevermind, it worked
[23-Jul-2009 11:10:20] <rmatte> I was looking at "Available Zenoss Version" which still says 2.3.3
[23-Jul-2009 11:10:40] <rmatte> but the upgrade worked, very smoothly
[23-Jul-2009 11:11:14] <mrayzenoss> is it firewalled?  that's weird.  Mine says 2.4.2
[23-Jul-2009 11:16:47] <mrayzenoss> Alright, I'm gonna let bedwards off the hook and call it a session. I'll be around per usual, but thanks to everyone who stopped by
[23-Jul-2009 11:19:17] <rmatte> it's firewalled for incoming, not outgoing
[23-Jul-2009 11:19:34] <rmatte> so I'm surprised that it's not able to go out and see what the latest version is, either way, doesn't matter
[23-Jul-2009 11:19:56] <mrayzenoss> yeah, not really a big deal
[23-Jul-2009 11:20:39] <rmatte> now I need to break my monster zenpacks down in to smaller ones, fun fun
[23-Jul-2009 11:21:27] <rmatte> I have 1 Zenpack that is basically all templates in /Devices, but it changes the snmp community string when installed, which is quite bad, so I need to break it down, then I have another with all transforms which I'm going to leave as a single pack for now
[23-Jul-2009 11:24:18] <rmatte> Do you know if the bug with the network map, (where if devices are outside of the screen and you click "Fit to window" it makes all of the lines that are drawn between devices remain at 1 static location) has been reported?
[23-Jul-2009 11:24:39] <mrayzenoss> doesn't sound familiar
[23-Jul-2009 11:24:42] <rmatte> It basically makes it so that the lines stay in 1 spot, you can move devices around but you can't see the links between them properly
[23-Jul-2009 11:24:57] <rmatte> I'll do a search to see and log a trac for it if not
[23-Jul-2009 11:25:32] <rmatte> actually, it's kind of working, hmmm
[23-Jul-2009 11:26:07] <rmatte> hmmm, it's not as bad as I remember it being in the past
[23-Jul-2009 11:26:49] <rmatte> ah, it's actually working properly now
[23-Jul-2009 11:31:44] <dberger> is there any reason why an import that works in zendmd will fail in a daemon module's header?
[23-Jul-2009 11:34:48] <rmatte> dberger: the only thing that I can think of is that the daemon module doesn't have certain variables that zendmd loads
[23-Jul-2009 11:34:50] <rmatte> see: http://www.zenoss.com/Members/netdata/use-a-zendmd-script-as-standalone-program/
[23-Jul-2009 11:35:03] <rmatte> vars/functions
[23-Jul-2009 11:38:27] <dberger> rmatte, thanks...i guess i forgot to import Globals
[23-Jul-2009 11:38:34] <rmatte> no problem
[23-Jul-2009 11:47:07] <jb> hmm
[23-Jul-2009 11:47:12] <jb> you guys doing any SSH key based stuff?
[23-Jul-2009 11:47:13] <jb> with zen
[23-Jul-2009 11:48:16] <rmatte> I'm not personally
[23-Jul-2009 11:48:31] <rmatte> snmp is more suited for the type of monitoring we do
[23-Jul-2009 11:48:33] <jb> wonder if it is documented anywhere..
[23-Jul-2009 11:48:43] <rmatte> mrayzenoss: the upgrade didn't write over the config files, just FYI
[23-Jul-2009 11:48:43] <jb> well us too.. except for aix
[23-Jul-2009 11:48:44] <jb>
[23-Jul-2009 11:48:50] <rmatte> hehe
[23-Jul-2009 11:49:06] <rmatte> have you looked at the AIX zenpack that rocket made?
[23-Jul-2009 11:49:09] <rmatte> it's pretty badass
[23-Jul-2009 11:49:22] <jb> yeah
[23-Jul-2009 11:49:25] <jb> but its snmp based
[23-Jul-2009 11:49:35] <jb> which is unstable, and a bitch to configure :/
[23-Jul-2009 11:49:39] <rmatte> yeh, I guess you're more worried about the crappy AIX snmp agent crashing?
[23-Jul-2009 11:49:47] <jb> yep.. and the configuration of it
[23-Jul-2009 11:50:01] <jb> he hopefully might start transforming it to an ssh based pack
[23-Jul-2009 11:50:01] <rmatte> you could always install the port of net-snmp for AIX on your boxes
[23-Jul-2009 11:50:02] <sergeymasushko> hi 2 all I've a stupid question. how zenoss monitors cpu utilization by default... just I've broken my default template
[23-Jul-2009 11:50:08] <jb> thats what I do now..
[23-Jul-2009 11:50:10] <jb> and it still sucks
[23-Jul-2009 11:50:13] <rmatte> ah
[23-Jul-2009 11:50:14] <rmatte> lol
[23-Jul-2009 11:50:28] <rmatte> crashes a lot?
[23-Jul-2009 11:50:31] <jb> yeah
[23-Jul-2009 11:50:38] <rmatte> remind me never to use AIX
[23-Jul-2009 11:50:40] <jb> and doesn't provide all of the stuff that you need
[23-Jul-2009 11:50:41] <rmatte>
[23-Jul-2009 11:50:45] <jb> you have to use both AIX's snmpd and net-snmp
[23-Jul-2009 11:50:49] <rmatte> ah
[23-Jul-2009 11:50:50] <jb> and have net-snmp proxy certain oid's to aix's
[23-Jul-2009 11:50:52] <rmatte> garbage
[23-Jul-2009 11:51:02] <jb> so.. an SSH rocket pack would be the best
[23-Jul-2009 11:51:23] <rmatte> pretty sad that AIX is that crappy for snmp
[23-Jul-2009 11:52:00] <jb> yeah
[23-Jul-2009 11:52:15] <rmatte> Microsoft's SNMP agent isn't all that great either
[23-Jul-2009 11:52:22] <rmatte> horrible planning for the OIDs
[23-Jul-2009 11:52:45] <rmatte> I've had to write scripts to hunt down OID values for CPU and Memory
[23-Jul-2009 11:52:51] <rmatte> since they wander
[23-Jul-2009 11:53:29] <rmatte> if they were smart they would have always made physical memory the first value and paging the second, but instead they put them at the bottom, so the OID changed based on the number of physical drives in the system
[23-Jul-2009 11:53:47] <rmatte> the CPU OIDs also just tend to wander on their own, one day the first CPU will be .1, the next it's suddenly .2
[23-Jul-2009 11:53:57] <blocky> wow
[23-Jul-2009 11:54:13] <blocky> that was poor planning eh
[23-Jul-2009 11:54:28] <rmatte> So I've made scripts that address that, I'm going to turn them in to a collector plugin at some point
[23-Jul-2009 11:54:47] <twm1010> isn't that what informant addresses?
[23-Jul-2009 11:54:57] <twm1010> piping perfmon counters to OID's for reliable query?
[23-Jul-2009 11:55:02] <rmatte> twm1010: correct, but our clients didn't want informant installed on their systems
[23-Jul-2009 11:55:18] <twm1010> probably the name, it does sound kinda shady
[23-Jul-2009 11:55:32] <rmatte> and possibly the website that looks a thousand years old
[23-Jul-2009 11:55:41] <jb> haha yeah
[23-Jul-2009 11:55:43] <jb> it does
[23-Jul-2009 11:56:11] <rmatte> I should send them an email with a subject of: "HIRE A WEB DEVELOPER!!! KTHX!!!"
[23-Jul-2009 11:57:02] <rmatte> twm1010: but yeh, once I have the collector plugin built for that sort of thing, then Zenoss will be able to do effective performance monitoring for windows hosts.
[23-Jul-2009 11:57:28] <rmatte> my scripts are written in bash though, so I need to figure out how to turn that in to python
[23-Jul-2009 11:57:38] <rmatte> I'm still learning python
[23-Jul-2009 11:57:59] <twm1010> interesting, so you're going to use the built-in OID's or trying to?
[23-Jul-2009 11:58:30] <rmatte> yeh, I'm using them right now successfully, but it's with a couple of bash scripts to collect the info, so it's not very efficient
[23-Jul-2009 11:59:22] <rmatte> CPU: http://pastebin.com/m76f1adbc
[23-Jul-2009 11:59:43] <rmatte> Memory: http://pastebin.com/m50d2fe0d
[23-Jul-2009 12:00:09] <twm1010> so you do these scripts as a command datasource?
[23-Jul-2009 12:00:15] <rmatte> correct
[23-Jul-2009 12:00:20] <twm1010> the ultimate goal to use built-in SNMP, and not WMI, or use informant?
[23-Jul-2009 12:00:36] <rmatte> correct
[23-Jul-2009 12:01:06] <rmatte> the main downside right now is that if you want to monitor each separate CPU you need to create a template for each number of CPUs
[23-Jul-2009 12:01:21] <twm1010> Man, it would be nice if the report you select can come up first to let you select your criteria, instead of running with some kind of default
[23-Jul-2009 12:01:26] <rmatte> so I have /Devices/Server/1 Processor, 2 Processors, 4 Processors, etc...
[23-Jul-2009 12:01:28] <twm1010> i've got 1500 interfaces
[23-Jul-2009 12:01:34] <rmatte> then I just put the devices in the appropriate group
[23-Jul-2009 12:01:46] <twm1010> Ah... I gotcha, makes sense
[23-Jul-2009 12:01:59] <rmatte> twm1010: yeh, I hear you, I have a box with well over 3000 interfaces, it takes forever
[23-Jul-2009 12:02:09] <twm1010> I think I ran into that with Netware originally, whoever wrote the Netware Zenpack overcame that
[23-Jul-2009 12:03:05] <rmatte> overcame the having to make multiple groups thing?
[23-Jul-2009 12:03:18] <rmatte> or the report thing?
[23-Jul-2009 12:03:26] <twm1010> Well, overcame having it track usage among the CPU's
[23-Jul-2009 12:03:31] <rmatte> ah
[23-Jul-2009 12:03:34] <twm1010> instead of an aggregate value
[23-Jul-2009 12:04:11] <rmatte> but yeh, this method is working quite nicely for the moment
[23-Jul-2009 12:04:28] <rmatte> I'm monitoring over 50 windows servers on one Zenoss instance with it, and it works fine
[23-Jul-2009 12:09:01] <rmatte> actually 62 to be exact
[23-Jul-2009 12:11:07] <twm1010> I have quite a few Windows machines, at least 100...
[23-Jul-2009 12:11:41] <rmatte> zenoss@*****:/usr/local/zenoss/zenoss/Products/SNMPWindowsPerfMonitor/libexec$ ./wincpu.sh <device> <community>
[23-Jul-2009 12:11:41] <rmatte> OK|CPU1=1 CPU2=3 CPU3=1 CPU4=3 Total=2
[23-Jul-2009 12:11:41] <rmatte> zenoss@*****:/usr/local/zenoss/zenoss/Products/SNMPWindowsPerfMonitor/libexec$ ./winmem.sh <device> <community>
[23-Jul-2009 12:11:42] <rmatte> OK|MemoryTotal=4191872 MemoryUsed=3675456 PercentMemoryUsed=88
[23-Jul-2009 12:11:48] <rmatte> that's what the output looks like
[23-Jul-2009 12:13:28] <rmatte> takes about 5-7 seconds for both scripts to run per device
[23-Jul-2009 12:14:10] <rmatte> that's what I want to improve on
[23-Jul-2009 12:17:25] <rmatte> actually, it's only about 3 seconds
[23-Jul-2009 12:18:58] <twm1010> Not bad
[23-Jul-2009 12:19:34] <twm1010> You think Python would be faster?
[23-Jul-2009 12:19:44] <rmatte> probably
[23-Jul-2009 12:19:59] <rmatte> with bash I'm having to spawn all sorts of additional processes
[23-Jul-2009 12:20:23] <rmatte> spawning snmpwalk a couple times, the cut command, etc...
[23-Jul-2009 12:21:11] <twm1010> man, my zopectl is just choking running these interface reports
[23-Jul-2009 12:21:18] <twm1010> taking nearly 10 minutes at a time
[23-Jul-2009 12:21:33] <rmatte> brutal
[23-Jul-2009 12:22:56] <twm1010> I suppose it's having to go to each device, list all the interfaces, then pull the RRD information from each because utilization is there
[23-Jul-2009 12:23:09] <rmatte> yeh, that's what it's doing
[23-Jul-2009 12:23:24] <twm1010> a simpler report, only needing speeds, which is stored in zope already would be lightning fast.
[23-Jul-2009 12:23:34] <rmatte> yeh
[23-Jul-2009 12:24:01] <rmatte> you could make a simpler report fairly easily
[23-Jul-2009 12:24:35] <rmatte> just need to copy that .rpt file to some other name, then remove the fields you don't want, run the command to load the report, and you're good to go
[23-Jul-2009 12:25:04] <twm1010> hrmm, how does it get added to the GUI I wonder? or does that command do that
[23-Jul-2009 12:25:11] <rmatte> the command does it
[23-Jul-2009 12:25:30] <rmatte> python $ZENHOME/Products/ZenReports/ReportLoader.py --force
[23-Jul-2009 12:29:07] <rmatte> mrayzenoss: is there a list of the fixes in 2.4.3?
[23-Jul-2009 12:34:02] <rmatte> I'm trying to determine if it's worth waiting for 2.4.3 or just doing the 2.4.2 upgrades now
[23-Jul-2009 12:34:12] <mrayzenoss> rmatte: http://dev.zenoss.org/trac/report/6
[23-Jul-2009 12:34:14] <rmatte> I guess we might as well do the upgrades now
[23-Jul-2009 12:34:25] <rmatte> thanks
[23-Jul-2009 12:35:13] <mrayzenoss> our goal is to do a maintenance release every 6-8 weeks, just to roll up fixes
[23-Jul-2009 12:35:26] <rmatte> ah
[23-Jul-2009 12:35:59] <twm1010> testing my little interface inventory report now
[23-Jul-2009 12:37:45] <rmatte> wow
[23-Jul-2009 12:37:46] <rmatte> http://dev.zenoss.org/trac/ticket/5279
[23-Jul-2009 12:37:50] <rmatte> that's kind of a big one
[23-Jul-2009 12:38:18] <jb> heh yea
[23-Jul-2009 12:41:17] <twm1010> zope still cranking away... didn't seem to speed much up
[23-Jul-2009 12:43:10] <twm1010> i chopped it down a little, but problem some script or process is still pulling utilization data
[23-Jul-2009 12:43:11] <twm1010> http://pastebin.com/d6d1892c6
[23-Jul-2009 12:44:01] <rmatte> is that the whole file you pasted?
[23-Jul-2009 12:44:06] <rmatte> paste a cat, you're missing lines
[23-Jul-2009 12:44:39] <twm1010> k, one sec
[23-Jul-2009 12:45:30] <twm1010> http://pastebin.com/d5ab36815
[23-Jul-2009 12:46:51] <rmatte> it's still executing the report plugin though, which is probably collecting all the data regardless
[23-Jul-2009 12:47:02] <rmatte> you'd have to modify the plugin as well, then restart Zenoss
[23-Jul-2009 12:47:14] <twm1010> right, make a copy, have it call that instead... hrmm
[23-Jul-2009 12:47:23] <rmatte> objects python:here.ReportServer.plugin('interface', tableState);
[23-Jul-2009 12:47:35] <rmatte> it should be interface.py right now, rename it to interfacespeed.py
[23-Jul-2009 12:47:41] <rmatte> and change that line in the report to match
[23-Jul-2009 12:52:55] <twm1010> I see a .py and a .pyc for each plugin
[23-Jul-2009 12:53:02] <twm1010> assume pyc is a compiled version/
[23-Jul-2009 12:53:03] <twm1010> ?
[23-Jul-2009 12:53:50] <mrayzenoss> yeah, just delete it when you make changes to the .py
[23-Jul-2009 12:54:22] <rmatte> nah, he's copying it
[23-Jul-2009 12:54:28] <rmatte> so he doesn't need to delete the pyc
[23-Jul-2009 12:54:47] <twm1010> interesting, hopefully this does what i need, i waited it out earlier and got my export, took almost 15 minutes :O
[23-Jul-2009 12:54:55] <rmatte> lol
[23-Jul-2009 12:57:29] <twm1010> hrmm, didn't like that...
[23-Jul-2009 12:58:11] <twm1010> i think that's enough for today
[23-Jul-2009 12:58:19] <rmatte> try just executing it manually
[23-Jul-2009 12:58:25] <rmatte> (the plugin)
[23-Jul-2009 12:58:34] <rmatte> for the command line, it'll give you a better idea of what's wrong with it
[23-Jul-2009 12:58:49] <rmatte> once it runs through without complaining then you're ready to go
[23-Jul-2009 12:59:11] <rmatte> you also need to restart zenoss to load the plugin in properly
[23-Jul-2009 12:59:25] <rmatte> and make sure the rpt file references that plugin, and not the old one
[23-Jul-2009 12:59:40] <rmatte> objects python:here.ReportServer.plugin('interfacespeed', tableState); or whatever
[23-Jul-2009 12:59:58] <twm1010> i'll mess with it later, got some other things to get done here
[23-Jul-2009 13:00:03] <rmatte> cool
[23-Jul-2009 13:02:48] <rmatte> Good thing I noticed http://dev.zenoss.org/trac/ticket/5279 before upgrading to 2.4.2, since the boxes that are still on 2.3.3 are available via the net lol
[23-Jul-2009 13:02:51] <rmatte> could have been bad
[23-Jul-2009 13:03:03] <rmatte> we're going to wait for 2.4.3
[23-Jul-2009 13:03:16] <jb> so SSH OS process monitoring is completely broken?
[23-Jul-2009 13:03:34] <chudler> ouch, that is an ugly bug
[23-Jul-2009 13:03:55] <rmatte> yup
[23-Jul-2009 13:04:00] <mrayzenoss> rmatte: wizard wasn't in 2.3.x
[23-Jul-2009 13:04:17] <rmatte> yeh, I know, I tried the link on one of them
[23-Jul-2009 13:04:25] <rmatte> that's why we're going to stick with 2.3.3 for the time being
[23-Jul-2009 13:04:32] <rmatte>
[23-Jul-2009 13:04:55] <rmatte> you should almost send out a security advisory about that
[23-Jul-2009 13:05:16] <mrayzenoss> yeah, I think they discussed that when I wasn't around.
[23-Jul-2009 13:05:34] <rmatte> behind your back eh? tsk tsk
[23-Jul-2009 13:05:36] <twm1010> Where oh where do I set the "from" address for email notifications
[23-Jul-2009 13:05:53] <twm1010> right now its prefixing zenossuser_admin@ourdomain.com
[23-Jul-2009 13:06:12] <twm1010> nevermind...
[23-Jul-2009 13:06:20] <mrayzenoss> I believe I was in another meeting when we had that defect review, I would've raised a bigger stink probably
[23-Jul-2009 13:06:28] <rmatte> It's right under settings lol
[23-Jul-2009 13:06:46] <twm1010> Maybe I need a bigger screen.
[23-Jul-2009 13:07:03] <rmatte> There's only one thing worse than having a big security vulnerability in your software, and that's having a security vulnerability but not making it publicly known
[23-Jul-2009 13:07:24] <rmatte> twm1010: 30 inch!!!
[23-Jul-2009 13:07:48] <rmatte> that's what I'm eventually buying for my home computer down the line
[23-Jul-2009 13:08:18] <rmatte> once I pay off my credit line, which should be close to doomsday at this point
[23-Jul-2009 13:08:37] <jb> mrayzenoss: do you know how OS process monitoring works when using SSH?
[23-Jul-2009 13:08:56] <rmatte> I thought the SSH monitoring was primarily for performance monitoring
[23-Jul-2009 13:09:27] <jb> hm, so native OS process monitoring is not supported via SSH?
[23-Jul-2009 13:09:58] <rmatte> I could be wrong, but I don't think so
[23-Jul-2009 13:18:45] <blocky> how can i increase polling frequency from 5 minutes to 1
[23-Jul-2009 13:18:51] <blocky> do i need to recreate the rrd files
[23-Jul-2009 13:21:37] <jmp242> yes
[23-Jul-2009 13:22:19] <jmp242> blocky - search the forums, there is a thread on this. Polling more frequently will have performance implications. And you'll have to re-figure all the graph definitions I believe.
[23-Jul-2009 13:22:33] <blocky> manually?
[23-Jul-2009 13:22:42] <rmatte> yes
[23-Jul-2009 13:22:45] <jmp242> yes
[23-Jul-2009 13:22:47] <blocky> ugh
[23-Jul-2009 13:22:48] <blocky> why?
[23-Jul-2009 13:23:10] <jmp242> Because RRD generates how it's going to do the graphs when the graph /RRD is created
[23-Jul-2009 13:23:16] <jmp242> and Zenoss uses RRDTool for graphing
[23-Jul-2009 13:23:31] <rmatte> yeh, the RRD file knows the "units" it's using
[23-Jul-2009 13:23:34] <jmp242> It's pretty complicated, I'd check out the RRDTool site on scaling the graphs
[23-Jul-2009 13:24:20] <blocky> i never understand why some forums have an option to search message text only but not thread title only
[23-Jul-2009 13:24:33] <mrayzenoss> so I went and asked about the SSH process monitoring, it was working in 2.4.1 and broke in 2.4.2
[23-Jul-2009 13:24:34] <rmatte> blocky: yeh, it's pretty dumb
[23-Jul-2009 13:24:44] <mrayzenoss> it'll be fixed again
[23-Jul-2009 13:24:46] <rmatte> ah, is that on the list for 2.4.3 then?
[23-Jul-2009 13:24:51] <mrayzenoss> not sure yet
[23-Jul-2009 13:24:59] <mrayzenoss> I know it's been confirmed working in trunk
[23-Jul-2009 13:25:04] <rmatte> ah
[23-Jul-2009 13:25:25] <rmatte> well then it's just a matter of applying a fix that's already been made to 2.4.3 before release
[23-Jul-2009 13:25:26] <rmatte>
[23-Jul-2009 13:25:39] <mrayzenoss> yeah, same as that security one
[23-Jul-2009 13:25:44] <mrayzenoss> we talked about that as well
[23-Jul-2009 13:25:51] <rmatte> cool
[23-Jul-2009 13:26:03] <rmatte> when I saw that one I was like "wow"
[23-Jul-2009 13:26:05] <rmatte>
[23-Jul-2009 13:26:13] <mrayzenoss> I got to talk to Mozilla, Joomla and Drupal QA last weekend, they keep all their security tickets private
[23-Jul-2009 13:26:21] <mrayzenoss> this one slipped through the cracks
[23-Jul-2009 13:26:41] <rmatte> ah
[23-Jul-2009 13:26:43] <mrayzenoss> zenpatch 14181 to fix it
[23-Jul-2009 13:27:08] <rmatte> so that patch can be applied to 2.42?
[23-Jul-2009 13:27:11] <mrayzenoss> yeah
[23-Jul-2009 13:27:12] <rmatte> 2.4.2 rather
[23-Jul-2009 13:27:15] <rmatte> k cool
[23-Jul-2009 13:27:29] <mrayzenoss> in the ticket, you see the "Backporting r14180 to the 2.4.x branch"
[23-Jul-2009 13:27:35] <twm1010> boy am i a goober
[23-Jul-2009 13:27:37] <mrayzenoss> that's where you want to zenpatch from
[23-Jul-2009 13:27:50] <twm1010> the only programming i've done is COBOL, and that was 6 years ago...
[23-Jul-2009 13:27:59] <twm1010> still, i should be able to spot a missing close bracket, now shouldn't i
[23-Jul-2009 13:28:10] <blocky> lol
[23-Jul-2009 13:29:38] <blocky> so if im collecting data 5x as freqently, do i have to scale all my graphs horizontally to compress them
[23-Jul-2009 13:30:09] <twm1010> zoom out, basically?
[23-Jul-2009 13:30:20] <mrayzenoss> the step in the RRD graphs is set by the collector interval
[23-Jul-2009 13:30:25] <rmatte> zenoss needs to be stopped to apply the patch I assume?
[23-Jul-2009 13:30:34] <blocky> zoom out, yea
[23-Jul-2009 13:30:37] <mrayzenoss> rmatte: that would be safest
[23-Jul-2009 13:30:40] <rmatte> k
[23-Jul-2009 13:30:53] <blocky> mrayzenoss: so if i change the collector interval will the rrds get created automatically?
[23-Jul-2009 13:31:00] <mrayzenoss> alas, no
[23-Jul-2009 13:31:19] <mrayzenoss> well, it won't update existing ones
[23-Jul-2009 13:31:32] <mrayzenoss> so yes they'll get created if they don't exist, but it won't retrofit
[23-Jul-2009 13:31:47] <blocky> if i delete the rrd files will it recreate them?
[23-Jul-2009 13:31:51] <mrayzenoss> yes
[23-Jul-2009 13:31:56] <blocky> i dont have a ton of historical data i need to worry about
[23-Jul-2009 13:32:05] <blocky> do i need to modify the graph definitions?
[23-Jul-2009 13:32:44] <mrayzenoss> if you reduce the step size (by decreasing the time between collections), you'll store fewer days of that precision
[23-Jul-2009 13:33:16] <blocky> i dont actually know how many im storing now, what is the default?
[23-Jul-2009 13:33:23] <mrayzenoss> so you'll want to edit the CREATE commands if you still want 50 hours at 5 minutes as opposed to 10 hours at 1 minute
[23-Jul-2009 13:33:39] <mrayzenoss> assuming we're talking SNMP, the default is 300 seconds
[23-Jul-2009 13:33:51] <mrayzenoss> I think for Commands it's 60
[23-Jul-2009 13:34:10] <blocky> its snmp
[23-Jul-2009 13:34:30] <blocky> if i edit the create command in the Device template will that affect all my performance graphds
[23-Jul-2009 13:34:40] <mrayzenoss> yeah, so when you change the frequency from 300 to 60, it will go from 50 hours to 10 hours
[23-Jul-2009 13:34:57] <mrayzenoss> it will affect any that aren't custom
[23-Jul-2009 13:35:16] <mrayzenoss> so in the Data source you can define custom RRD CREATE commands as well
[23-Jul-2009 13:35:35] <blocky> so the net-snmp has 4 graphs
[23-Jul-2009 13:36:30] <blocky> the command is about 20 lines
[23-Jul-2009 13:37:57] <mrayzenoss> it's probably just inheriting the default rrd create defined on the collector
[23-Jul-2009 13:38:24] <mrayzenoss> look at http://yourhost:58080/zport/dmd/Monitors/Performance/localhost/editPerformanceConf
[23-Jul-2009 13:39:46] <mrayzenoss> you have to change your host and port
[23-Jul-2009 13:40:40] <blocky> ah cool
[23-Jul-2009 13:40:53] <blocky> so unless i change it all the graphs derive from that command?
[23-Jul-2009 13:40:58] <mrayzenoss> yes
[23-Jul-2009 13:43:13] <blocky> so do i change all those 600s to 3000
[23-Jul-2009 13:44:16] <mrayzenoss> yeah, assuming you want to keep the same length of storage at your new collection rate
[23-Jul-2009 13:44:27] <mrayzenoss> so your files will also be 5X larger
[23-Jul-2009 13:44:55] <mrayzenoss> and you'll be writing 5X the number of data points, so keep that in mind if IO starts to bog down
[23-Jul-2009 13:45:11] <blocky> right now im only monitoring 2 machines
[23-Jul-2009 13:45:17] <blocky> so hopefully my zen box can keep up
[23-Jul-2009 13:45:19] <mrayzenoss> that should be OK
[23-Jul-2009 13:46:00] <blocky> so ill save the config options, then delete all the rrd files?
[23-Jul-2009 13:46:06] <blocky> then remodel perhaps?
[23-Jul-2009 13:46:37] <mrayzenoss> I think you have to restart zenhub, I could be wrong
[23-Jul-2009 13:46:50] <blocky> okay ill restart zenoss
[23-Jul-2009 13:51:47] <blocky> my graphs are back but they say nan
[23-Jul-2009 13:51:54] <blocky> i guess i have to wait for polling
[23-Jul-2009 13:51:58] <mrayzenoss> they take a couple minutes to get data
[23-Jul-2009 13:52:24] <rmatte> that patch worked nicely
[23-Jul-2009 13:52:35] <rmatte> I think we're still going to hold off for 2.4.3 though
[23-Jul-2009 13:53:06] <mrayzenoss> in theory that will be the last release before King Crab
[23-Jul-2009 13:53:26] <rmatte> yeh, so I think we'll wait for that, then just do all of our boxes at once
[23-Jul-2009 13:56:19] <twm1010> King Crab is 2.5?
[23-Jul-2009 13:56:42] <mrayzenoss> yeah, marketing may change the version number but it's the next major release
[23-Jul-2009 13:56:53] * mrayzenoss is a fan of year.month
[23-Jul-2009 13:57:13] <twm1010> ah, thinking of a jump to 3.0?
[23-Jul-2009 13:57:31] <mrayzenoss> yeah, my guess is it'll still be 2.5
[23-Jul-2009 13:57:43] <mrayzenoss> perhaps Stone Crab will be 3.0, since it should be on Python 2.6
[23-Jul-2009 13:57:54] <twm1010> Does the tracker have a list of what WILL be in king crab?
[23-Jul-2009 13:58:06] <twm1010> Kinda curious to know what to look forward to
[23-Jul-2009 13:58:20] <mrayzenoss> http://dev.zenoss.org/trac/report has our bugs
[23-Jul-2009 13:58:26] <mrayzenoss> in the various reports
[23-Jul-2009 13:59:03] <mrayzenoss> hmm.. I don't see one that looks for the King Crab milestone
[23-Jul-2009 13:59:48] <rmatte> yeh, doesn't appear to be one
[23-Jul-2009 14:00:09] <mrayzenoss> it's Trac, so you can build custom reports, let me see if I can put together a King Crab one for you
[23-Jul-2009 14:00:35] <rmatte> cool
[23-Jul-2009 14:00:50] <mrayzenoss> http://dev.zenoss.org/trac/query?status=assigned&status=backlog&status=closed&status=new&status=verification&group=patch_state&order=status&col=id&col=summary&col=status&col=owner&col=priority&col=milestone&col=component&col=changetime&milestone=King_Crab&report=6
[23-Jul-2009 14:01:14] <rmatte> that was quick
[23-Jul-2009 14:01:57] <mrayzenoss> just removed the Maintenance Target from 2.4.3 and added the Milestone "King Crab"
[23-Jul-2009 14:02:09] <rmatte> cool
[23-Jul-2009 14:02:21] <mrayzenoss> I'm sure there are a lot of dupes in there, but there are 1000 or so currently tagged for King Crab
[23-Jul-2009 14:02:26] <twm1010> yeah i basically made my way to that, looking through all the "enhancements" right now.
[23-Jul-2009 14:02:40] <twm1010> automatic device classification would be pretty sweet
[23-Jul-2009 14:03:06] <rmatte> yeh, it would indeed
[23-Jul-2009 14:03:08] <twm1010> It wouldn't be that hard to gather a list of common sysObjectID's and use that to put things where they need to go.
[23-Jul-2009 14:03:27] <rmatte> and make the list editable by the user
[23-Jul-2009 14:03:30] <rmatte> like the products list
[23-Jul-2009 14:03:47] <twm1010> In SolarWinds I group my devices by "vendor" which I'm sure is by that OID...
[23-Jul-2009 14:04:45] <mrayzenoss> also, a lot of the tickets for King Crab were fixed in the 2.4.x branch
[23-Jul-2009 14:05:44] <rmatte> thanks goodness, someone logged this: http://dev.zenoss.org/trac/ticket/287
[23-Jul-2009 14:05:57] <rmatte> If that gets fixed in King Crab, booyeh
[23-Jul-2009 14:05:59] * chudler scrolls up.  For a second there I thought mray said Stone Crab would target python 3.0 ;-)
[23-Jul-2009 14:06:07] <rmatte> plus we might be able to stop having to monitor ports using thresholds
[23-Jul-2009 14:06:51] <mrayzenoss> well, a big disclaimer on those tickets
[23-Jul-2009 14:06:59] <mrayzenoss> that's what's currently assigned to King Crab
[23-Jul-2009 14:07:10] <rmatte> Opened 3 years ago
[23-Jul-2009 14:07:15] <rmatte> this one has been open for 3 years
[23-Jul-2009 14:07:36] <mrayzenoss> when we get closer, open defects will get reassessed and possibly moved to the next release
[23-Jul-2009 14:07:44] <rmatte>  Changed 3 months ago by bbibeault  ¶
[23-Jul-2009 14:07:44] <rmatte>     * status changed from backlog to assigned
[23-Jul-2009 14:07:44] <rmatte>     * owner changed from edahl to cluther
[23-Jul-2009 14:07:57] <mrayzenoss> we don't ship with any open Priority 1s
[23-Jul-2009 14:08:09] <rmatte> Ah, this is a pri 2
[23-Jul-2009 14:09:11] <mrayzenoss> about the time we release the beta, the open tickets that won't get addressed will start getting weeded out
[23-Jul-2009 14:12:29] <rmatte> mrayzenoss: http://dev.zenoss.org/trac/ticket/3218 - They just need to add an option that executes that function that we're using as a workaround for the Cisco and HP Mibs ZenPacks.
[23-Jul-2009 14:12:47] <rmatte> You could tip jstevens off about it
[23-Jul-2009 14:13:00] <mrayzenoss> heh, he hasn't worked here since December
[23-Jul-2009 14:13:04] <rmatte> ah
[23-Jul-2009 14:13:09] <rmatte> maybe that should be reassigned then
[23-Jul-2009 14:13:38] <rmatte> but yeh, that's be an easy fix/close
[23-Jul-2009 14:13:40] <mrayzenoss> yeah, I'll bring that up as a suggestion for the next defect review
[23-Jul-2009 14:13:54] <mrayzenoss> to unassign them from people who don't work here anymore
[23-Jul-2009 14:14:00] <rmatte> yeh
[23-Jul-2009 14:16:04] <mrayzenoss> rmatte: since you have a login I believe you can add notes
[23-Jul-2009 14:17:55] <rmatte> yeh, I'll add a note
[23-Jul-2009 14:18:00] <mrayzenoss> thanks
[23-Jul-2009 14:22:41] <rmatte> done
[23-Jul-2009 14:27:00] <rmatte> They could just add an ajax popup that says something like "You are about to remove selected ZenPacks" then have a checkbox below it that is checked by default and reads "Remove ZenPack contents as well"
[23-Jul-2009 14:29:51] <rmatte> there, I just added that as a note as well
[23-Jul-2009 15:04:33] <dberger> if I add zProperties through packZProperties, is there any way to remove them?
[23-Jul-2009 15:05:35] <dberger> I'm developing a ZenPack that has had zProperties change a few times and has been installed more than once, but the old zProperties stay in the system
[23-Jul-2009 15:12:43] <rmatte> dberger: http://www.zenoss.com/community/wiki/tips-and-tricks/ZendmdDeleteLocalDeviceProperty/
[23-Jul-2009 15:12:56] <rmatte> use a modified version of that code in zendmd
[23-Jul-2009 15:12:59] <rmatte> then commit()
[23-Jul-2009 15:14:05] <rmatte> actually, that's not what you want
[23-Jul-2009 15:14:16] <rmatte> (After reading the code, that's just to remove local copies)
[23-Jul-2009 15:14:23] <rmatte> hmmm
[23-Jul-2009 15:16:22] <rmatte> how did you create the zProperties in the first place?
[23-Jul-2009 15:17:16] <dberger> in the zenpack class
[23-Jul-2009 15:17:26] <dberger> specified packZProperties
[23-Jul-2009 15:17:49] <rmatte> ah
[23-Jul-2009 15:19:06] <rmatte> apparently you can creat one in zendmd by doing:
[23-Jul-2009 15:19:07] <rmatte> >>> ch = devices
[23-Jul-2009 15:19:07] <rmatte> >>> ch._setProperty('zTest', 'Test', type='string')
[23-Jul-2009 15:19:07] <rmatte> >>> commit()
[23-Jul-2009 15:19:20] <rmatte> so maybe you can do something like unsetProperty?
[23-Jul-2009 15:22:01] <dberger> maybe
[23-Jul-2009 15:22:30] <rmatte> it's not unset
[23-Jul-2009 15:22:34] <rmatte> I'll try to figure out what it is
[23-Jul-2009 15:24:38] <dberger> _delProperty
[23-Jul-2009 15:24:40] <dberger> I think
[23-Jul-2009 15:25:01] <rmatte> ch._delProperty('zTest') just worked for me
[23-Jul-2009 15:25:04] <rmatte> so yeh, that's it
[23-Jul-2009 15:25:10] <dberger> awesome
[23-Jul-2009 15:25:11] <rmatte> I was trying it with too many arguments before
[23-Jul-2009 15:29:22] <dberger> uggg, now I'm getting a DisconnectedError on commit
[23-Jul-2009 15:29:47] <rmatte> really?  weird, I didn't get anything like that
[23-Jul-2009 15:29:55] <rmatte> I was able to add and remove the zProperty no problem
[23-Jul-2009 15:30:40] <dberger> restarted zendmd, it works
[23-Jul-2009 15:30:48] <rmatte> good stuff
[23-Jul-2009 15:31:35] <rmatte> what kind of ZenPack are you making?
[23-Jul-2009 15:33:15] <dberger> a really basic Asterisk monitor
[23-Jul-2009 15:33:37] <mrayzenoss> feel free to submit it, the existing one is flagged as "broken" and it's unmaintained
[23-Jul-2009 15:33:38] <dberger> zProperties are not necessarily the right way to go, but I got started and they seemed to work
[23-Jul-2009 15:33:40] <rmatte> ah, cool
[23-Jul-2009 16:10:25] <jb> hey matt
[23-Jul-2009 16:10:33] <jb> is there a newer aix snmpd pack?
[23-Jul-2009 16:10:39] <jb> than what is posted on the website currently
[23-Jul-2009 16:11:01] <rmatte> jb: doubt it
[23-Jul-2009 16:11:13] <jb> k.. need to talk to rocket then
[23-Jul-2009 16:11:25] <mrayzenoss> yeah, if you have any patches or fixes let him know
[23-Jul-2009 16:11:41] <jb> ok
[23-Jul-2009 16:39:01] <mrayzenoss> Tip of the Month: Supporting Large Linux and Unix Filesystems http://blog.zenoss.com/2009/07/23/tip-of-the-month-supporting-large-linux-and-unix-filesystems/
[23-Jul-2009 16:45:12] <jb> argh.. zen is running super slow on this VM
[23-Jul-2009 16:45:15] <jb> and i have no idea why
[23-Jul-2009 16:45:27] <jb> mrayzenoss: did you test rocket's AIX pack at all?
[23-Jul-2009 16:46:15] <mrayzenoss> just smoketesting
[23-Jul-2009 16:46:21] <mrayzenoss> I don't have any AIX boxes myself
[23-Jul-2009 16:46:23] <jb> you didnt notice it made the GUI slow did you?
[23-Jul-2009 16:46:26] <jb> k
[23-Jul-2009 16:47:40] <mrayzenoss> not really
[23-Jul-2009 16:58:40] <rmatte> what's slow?
[23-Jul-2009 16:58:52] <rmatte> do you notice it more with certain pages?
[23-Jul-2009 17:05:46] <rmatte> mrayzenoss: I need a definitive answer on this. How does the SSH monitoring work? Does Zenoss keep an ssh connection open to the server at all times, or does it connect each time it collects data, then disconnects?
[23-Jul-2009 17:06:10] <rmatte> I've heard mixed theories on it
[23-Jul-2009 17:06:44] <mrayzenoss> I will go ask the developer and get the definitive answer
[23-Jul-2009 17:06:51] <rmatte> thanks
[23-Jul-2009 17:07:08] <rmatte> also, the ssh account doesn't need to be root or anything obviously, just a regular user with no special perms?
[23-Jul-2009 17:08:06] <mrayzenoss> correct
[23-Jul-2009 17:09:48] <rmatte> cool
[23-Jul-2009 17:09:57] <mrayzenoss> the modeling closes the session, monitoring leaves it open
[23-Jul-2009 17:10:05] <rmatte> perfect
[23-Jul-2009 17:10:22] <rmatte> the guys here were concerned that it logged in each time, which would flood the access logs
[23-Jul-2009 17:10:32] <mrayzenoss> so we reuse the connection for each command call
[23-Jul-2009 17:10:39] <rmatte> cool
[23-Jul-2009 17:10:53] <mrayzenoss> yeah, there was a ticket about stale connections that was closed recently
[23-Jul-2009 17:10:57] <rmatte> There would probably be more overhead and lag if you reauthed the session each time
[23-Jul-2009 17:11:04] <rmatte> yeah, I remember seeing that ticket
[23-Jul-2009 17:11:06] <rmatte> afk for a sec
[23-Jul-2009 17:14:43] <mrayzenoss> jb: was this useful? http://forums.zenoss.com/viewtopic.php?p=37026#37026
[23-Jul-2009 17:20:50] <rmatte> that's a good tip
[23-Jul-2009 17:20:53] * rmatte bookmarks
[23-Jul-2009 17:22:36] <mrayzenoss> I added it to the notes on the wiki for the AIX ZenPack
[23-Jul-2009 17:24:16] <Blehk> In the docs it's stated that to get snmpd communicating with zenoss properly you just have to set the snmpd rocommunity to public in the snmpd.conf, however, I'm still not receiving a positive flag on the zenoss side. (Ubuntu) Is there a way to find out what's going on/wrong?
[23-Jul-2009 17:25:18] <rmatte> I don't see it on the AIX SNMP page?
[23-Jul-2009 17:25:33] <rmatte> oh, you added it to the wiki, not the actual zenpack page?
[23-Jul-2009 17:25:45] <mrayzenoss> trac-zenpacks/wiki/ZenPackscommunityAIX
[23-Jul-2009 17:26:06] <mrayzenoss> obviously I've done a poor job of evangelizing those pages
[23-Jul-2009 17:26:15] <rmatte> hehe
[23-Jul-2009 17:26:32] <mrayzenoss> the new community website will condense them into a single page
[23-Jul-2009 17:27:06] <mrayzenoss> Blehk: let me take a quick look at my Debian snmpd.conf
[23-Jul-2009 17:27:14] <rmatte> Blehk: you need to make sure that your rocommunity has all permissions
[23-Jul-2009 17:27:30] <rmatte> Blehk: like: access MyROGroup ""      any       noauth    exact  all    none   none
[23-Jul-2009 17:28:14] <Blehk> And I would change this... where?
[23-Jul-2009 17:28:37] <rmatte> let me pastebin our conf
[23-Jul-2009 17:28:48] <Blehk> snmpd.conf just has the community. Would snmp.conf have the rest of the details? Actually, that'd be best/easier.
[23-Jul-2009 17:28:58] <rmatte> no, it's all done in snmpd.conf
[23-Jul-2009 17:29:39] <mrayzenoss> Blehk: here's mine minus comments: http://pastebin.com/d5abfce74
[23-Jul-2009 17:30:04] <Blehk> cool, thanks.
[23-Jul-2009 17:31:11] <rmatte> http://pastebin.com/m510ce721
[23-Jul-2009 17:31:37] <rmatte> mycommunitystring is where we put our community string, and the trapsink lines at the top need to be changed to the IP of your Zenoss server so that it traps properly
[23-Jul-2009 17:31:45] <rmatte> we're using Ubuntu Server as well
[23-Jul-2009 17:32:00] <mrayzenoss> even better
[23-Jul-2009 17:32:12] <rmatte> the syslocation and syscontact should be changed as well
[23-Jul-2009 17:32:48] <mrayzenoss> heh, you can see I gave mine a lot of attention
[23-Jul-2009 17:32:49] <rmatte> by the way, the snmpconf program is pretty handy for generating snmp config files, that's what we used
[23-Jul-2009 17:33:19] <rmatte> yeh, but if it works it works
[23-Jul-2009 17:33:20] <rmatte> hehe
[23-Jul-2009 17:36:30] <rmatte> mrayzenoss: you should probably throw a link up to that wiki page from the zenpack page
[23-Jul-2009 17:36:38] <rmatte> since that's the first thing people will see, less digging
[23-Jul-2009 17:36:57] <rmatte> "For further info on configuring this ZenPack see: Wiki"
[23-Jul-2009 17:37:07] <mrayzenoss> rmatte: yeah, they're on most of the pages listed as "Community ZenPack Repository"
[23-Jul-2009 17:37:19] <rmatte> oh I see
[23-Jul-2009 17:37:24] <mrayzenoss> what would be better?  "Wiki page for AIX SNMP ZenPack"?
[23-Jul-2009 17:37:25] <rmatte> not very descriptive
[23-Jul-2009 17:37:45] <rmatte> Additional details for AIX SNMP ZenPack
[23-Jul-2009 17:37:47] <rmatte> probably
[23-Jul-2009 17:38:06] <mrayzenoss> hmmm... time to go update 90 pages
[23-Jul-2009 17:38:36] <rmatte> because quite honestly, I always ignored that link since it makes it sound like a link to an svn tree or something
[23-Jul-2009 17:38:57] <mrayzenoss> yeah... I've been meaning to make it better, I just haven't
[23-Jul-2009 17:39:02] <rmatte> hehe
[23-Jul-2009 17:39:04] <mrayzenoss> ahh intertia, my enemy
[23-Jul-2009 17:39:16] <rmatte> time constraints govern all
[23-Jul-2009 17:43:10] <rmatte> anyways, I'm out of here, been at work way too long for today
[23-Jul-2009 17:43:10] <rmatte> be back tomorrow
[23-Jul-2009 17:51:39] <mrayzenoss> later all
[23-Jul-2009 19:36:49] <ironpaw> hi
[23-Jul-2009 19:37:10] <ironpaw> anyone awake?
[23-Jul-2009 19:37:50] <ironpaw> got a few questions in regards to migrating zenoss data and upgrading
[23-Jul-2009 19:38:04] <ironpaw> well.. i've upgraded to latest 2.4.2 from 2.2.3 by following the upgrade path
[23-Jul-2009 19:38:36] <ironpaw> went from 2.2.3-> 2.2.4 then 2.2.4 to 2.3.3 then 2.3.3 to 2.4.2
[23-Jul-2009 19:39:06] <ironpaw> i have tar'd up the /usr/local/zenoss prior to starting the process so I'm hoping the db files/content are still there but just need to be "migrated"
[24-Jul-2009 00:00:46] [disconnected at Fri Jul 24 00:00:46 2009]
[24-Jul-2009 00:00:46] [connected at Fri Jul 24 00:00:46 2009]
[24-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[24-Jul-2009 01:29:59] <ironpaw> anyone up?
[24-Jul-2009 04:38:21] <Dieterbe_> ironpaw: yes. hi
[24-Jul-2009 04:45:26] <Dieterbe_> anyone knows how to get all possible functions for the 'device' object in event transform?
[24-Jul-2009 04:58:55] <Dieterbe_> i could not find that in the 'events' chapter of the official documentation, neither in the developers guide
[24-Jul-2009 06:57:45] <jmp242> Dieterbe_: I believe you'd have to look in the code - there aren't any complete lists I'm aware of
[24-Jul-2009 06:59:18] <Dieterbe_> thanks jmp242 . do you maybe know if i can do 'import re' in the event transform hook or is the re lib already imported?
[24-Jul-2009 07:05:05] Dieterbe_ is now known as Dieterbe
[24-Jul-2009 07:25:44] <jmp242> I generally import anything I'm going to use in the transforms, and the examples seem to do so as well
[24-Jul-2009 07:41:40] <Dieterbe> examples? i only found one example in the docs
[24-Jul-2009 07:58:18] <jmp242> For event transforms? Have you looked at the Community FAQ and Wiki?
[24-Jul-2009 07:58:48] <jmp242> trac-zenpacks/wiki/CommunityFAQGeneral
[24-Jul-2009 08:06:12] <Dieterbe> aha cool. hadn't seen that page yet
[24-Jul-2009 08:06:22] <Dieterbe> in the official docs there was only one brief example
[24-Jul-2009 08:08:39] <jmp242> Yea, I put it up to deal with consolidating some of the questions asked a bunch
[24-Jul-2009 08:10:06] <Dieterbe> nice nice
[24-Jul-2009 08:18:07] <twm1010> morning
[24-Jul-2009 08:18:20] <twm1010> jmp242: that last comment has me intrigued, are you referring to a FAQ?
[24-Jul-2009 08:22:34] <jmp242> yes?
[24-Jul-2009 08:23:46] <twm1010> url?
[24-Jul-2009 08:24:40] <jmp242> trac-zenpacks/wiki/CommunityFAQGeneral
[24-Jul-2009 08:24:59] <jmp242> I missed when you joined today
[24-Jul-2009 08:25:07] <jmp242> I had just posted the URL and was confused lol
[24-Jul-2009 08:39:26] <twm1010> Can I recommend a section on the difference between a source, stack, and RPM install
[24-Jul-2009 08:39:34] <twm1010> Linux newbies have NO idea what the difference is.
[24-Jul-2009 08:40:14] <bobrog> tim
[24-Jul-2009 08:41:45] <jmp242> Sure
[24-Jul-2009 08:42:03] <jmp242> I'll look into adding that under new users
[24-Jul-2009 08:48:56] <TBKDan> I'm looking to remodel a whole class of devices, so I don't want to do it manually via the GUI. Is there a method in zendmd to remodel a device?
[24-Jul-2009 08:51:08] <mrayzenoss> morning
[24-Jul-2009 08:51:16] <TBKDan> Morning
[24-Jul-2009 08:51:35] <TBKDan> Just going to re-ask now that you're in here
[24-Jul-2009 08:51:38] <TBKDan> I'm looking to remodel a whole class of devices, so I don't want to do it manually via the GUI. Is there a method in zendmd to remodel a device?
[24-Jul-2009 08:54:10] <TBKDan> Aha! Just found the zenmodeler command.  nvm!
[24-Jul-2009 08:54:53] <jmp242> twm1010: ok, I've added the section
[24-Jul-2009 08:56:06] <mrayzenoss> TBKDan: zenmodeler run --path=/Server/Linux
[24-Jul-2009 08:57:00] <TBKDan> mrayzenoss: Yup, that's running as we speak (well, for Windows) Thanks!
[24-Jul-2009 08:57:29] <rmatte> TBKDan: http://pastebin.com/m3c2fdf59 <--- Save that as "model.sh" in /home/zenoss, then create a file called devices.txt or whatever and put the device names that you want to model in the file, then "chmod +x model.sh" to make it executable and "./model.sh devices.txt" to model what devices are listed in the file.
[24-Jul-2009 08:57:45] <rmatte> that works too I guess
[24-Jul-2009 08:58:14] <rmatte> (I basically wrote this just because I wanted to model 1 device at a time since a lot of them were over slow links on this one particular zenoss box)
[24-Jul-2009 08:59:03] <TBKDan> Ah. All these devices are on the local LAN and I finally figured out why it wasn't picking up the interface info.
[24-Jul-2009 08:59:14] <rmatte> I see
[24-Jul-2009 08:59:24] <rmatte> bad snmp config?
[24-Jul-2009 09:00:45] <TBKDan> No, a zProperty needs to be explained a bit I had snmpMonitor turned off because I didn't want to be alerted every so often when the snmp service didn't respond in time for some reason. What I didn't know was that by doing that, it just completely ignored snmp altogether, not just the snmp service itself. So now I've changed my alerting rules to ignore snmp and put snmpMonitor back on.
[24-Jul-2009 09:00:58] <rmatte> that threshold of low swap for linux devices is so useless, it kicks off all the time unless you disable it
[24-Jul-2009 09:01:26] <rmatte> TBKDan: just increase the snmp timeout value
[24-Jul-2009 09:01:33] <rmatte> in zProperties for /Devices
[24-Jul-2009 09:02:28] <TBKDan> Eh, easy enough to just completely ignore them. I use WMI for the majority of my monitoring anyway, so it's not critical that I have it up and running.
[24-Jul-2009 09:02:44] <rmatte> ah
[24-Jul-2009 09:02:48] <rmatte> I use all SNMP for mine
[24-Jul-2009 09:03:11] <rmatte> we worked with WMI early on, but the WMI performance monitor pack is not very good
[24-Jul-2009 09:03:19] <rmatte> (I'm using core though)
[24-Jul-2009 09:03:23] <TBKDan> So am I
[24-Jul-2009 09:03:26] <rmatte> ah
[24-Jul-2009 09:03:37] <rmatte> I assume you just installed dastrup's packs?
[24-Jul-2009 09:03:47] <TBKDan> One sec, let me see which one I have
[24-Jul-2009 09:04:08] <TBKDan> Yup.
[24-Jul-2009 09:04:18] <TBKDan> I remember one didn't work very well, but then the next one I found worked like a champ.
[24-Jul-2009 09:04:25] <rmatte> yeh, I don't like how it spits out the username and password when the command fails on that pack
[24-Jul-2009 09:04:33] <rmatte> that really needs to be converted to a proper collector plugin
[24-Jul-2009 09:04:45] <TBKDan> I didn't want to install snmp informant on every friggin box, so I preferred wmi
[24-Jul-2009 09:04:57] <rmatte> I found a way to not have to use informant
[24-Jul-2009 09:05:12] <TBKDan> I'm pretty much the only one who watches zenoss, so meh
[24-Jul-2009 09:05:18] <rmatte> I wrote some scripts to collect the info and used them as command based datapoints, works like a charm
[24-Jul-2009 09:06:19] <rmatte> I'm actually working to convert that to a ZenPack right now (though it does need major improvement eventually)
[24-Jul-2009 09:06:48] <rmatte> I'm monitoring 62 windows servers on one of my boxes without any issues
[24-Jul-2009 09:07:19] <TBKDan> Cool.  It provides all the stats like disk and memory performance?
[24-Jul-2009 09:07:29] <TBKDan> (Another reason I wanted WMI)
[24-Jul-2009 09:07:36] <rmatte> I don't track disk IO, but it provides memory and CPU
[24-Jul-2009 09:07:59] <rmatte> I also had it modified to do swap as well but then we had a SAN failure and I lost the modifications, though I could redo it pretty easily
[24-Jul-2009 09:08:39] <TBKDan> Ugh. Out of curiosity, what was the SAN make and what happened? (Looking to purchase a SAN soon)
[24-Jul-2009 09:09:45] <twm1010> rmatte: im modifying one of ejulson's scripts in the WMI zenpack to get statistics from terminal servers
[24-Jul-2009 09:09:57] <rmatte> HP, one of the controller(fiber) cards failed (for some unknown reason), and it didn't fail over to the secondary. When we got it back up and running we discovered that none of our VMs would boot, looked like it corrupted the .vmdk files
[24-Jul-2009 09:10:12] <rmatte> twm1010: cool
[24-Jul-2009 09:10:16] <TBKDan> Ouch
[24-Jul-2009 09:10:44] <twm1010> actually have it working already, im just confused on where under $ZENHOME I should put this perl script
[24-Jul-2009 09:10:45] <rmatte> and the latest backups we had at the time were from a week ago, but now we're taking daily clones of the VMs and saving them to local disk on the blades
[24-Jul-2009 09:11:02] <rmatte> and we're going to be implementing tape backups fairly soon
[24-Jul-2009 09:11:16] <twm1010> What kind of SAN was that?
[24-Jul-2009 09:11:23] <rmatte> twm1010: are you looking to package it in to a ZenPack?
[24-Jul-2009 09:11:25] <TBKDan> Do you just use custom scripts for the clones?
[24-Jul-2009 09:11:42] <rmatte> TBKDan: We have it scheduled in VMWare
[24-Jul-2009 09:11:50] <rmatte> since we have the enterprise version of it
[24-Jul-2009 09:11:56] <twm1010> rmatte: I might... it will be a simple one, just a script and a template
[24-Jul-2009 09:12:26] <twm1010> Though, I might just make a ton of WMI scripts/templates. I'm working on one that pulls Exchange metrics from a 2003 exchange box
[24-Jul-2009 09:12:43] <rmatte> twm1010: one sec, finding where I saved the info on doing that
[24-Jul-2009 09:12:51] <rmatte> twm1010: it's fairly simple, one sec
[24-Jul-2009 09:13:08] <rmatte> aha
[24-Jul-2009 09:13:27] <rmatte> ok, so he's what you do... create a ZenPack
[24-Jul-2009 09:13:47] <twm1010> did that.
[24-Jul-2009 09:13:57] <rmatte> k, what's it called?
[24-Jul-2009 09:14:15] <twm1010> ZenPacks.community.terminalserviceswmi
[24-Jul-2009 09:15:10] <rmatte> ok, navigate to $ZENHOME/ZenPacks/ZenPacks.community.terminalserviceswmi/ZenPacks/community/terminalserviceswmi/libexec (I believe will be the path)
[24-Jul-2009 09:15:50] <rmatte> you might have to actually create that folder if it's not already there
[24-Jul-2009 09:16:16] <rmatte> then drop your script in it
[24-Jul-2009 09:17:25] <twm1010> ah ha... nice...
[24-Jul-2009 09:17:27] <rmatte> Then for the command based datasource make the command
[24-Jul-2009 09:17:29] <rmatte> ${here/ZenPackManager/packs/ZenPacks.community.terminalserviceswmi/path}/libexec/whatever.pl
[24-Jul-2009 09:17:40] <rmatte> and put whatever arguments at the end
[24-Jul-2009 09:17:46] <rmatte> like: ${here/ZenPackManager/packs/ZenPacks.community.terminalserviceswmi/path}/libexec/whatever.pl ${dev/manageIp} ${dev/zSnmpCommunity} ${here/mount}
[24-Jul-2009 09:17:49] <rmatte> or whatever
[24-Jul-2009 09:18:06] <twm1010> sweet, in this ${dev/zWinUser}
[24-Jul-2009 09:18:16] <rmatte> yeh, throw that at the end
[24-Jul-2009 09:18:29] <twm1010> I gotta install ejulson's WMI zenpack to see how he did the command though to properly escape the domain/user style
[24-Jul-2009 09:18:36] <rmatte> yeh
[24-Jul-2009 09:18:52] <twm1010> i remember in my first go round he had to change the zenpack to work cleanly.
[24-Jul-2009 09:19:06] <rmatte> but yeh, you're basically telling Zenoss to execute the command from the ZenPack's directory, which is quite clean for installing/removing zenpacks
[24-Jul-2009 09:19:19] <twm1010> Yeah, rather than full path.
[24-Jul-2009 09:19:35] <rmatte> exactly, it also keeps stuff out of the Products directory
[24-Jul-2009 09:19:43] <rmatte> which is only really intended for daemons/plugins
[24-Jul-2009 09:19:59] <twm1010> Excellent, I should have the script for Exchange done later, some people might find that handy
[24-Jul-2009 09:20:15] <rmatte> If it's good then submit it
[24-Jul-2009 09:20:20] <rmatte> always room for more community packs
[24-Jul-2009 09:21:02] <twm1010> It would be nice though, to have some way to automatically enumerate the information stores and their status
[24-Jul-2009 09:21:11] <twm1010> seems like that would be its own tab there.
[24-Jul-2009 09:21:40] <jb> that would be cool
[24-Jul-2009 09:22:01] <rmatte> that would be cool
[24-Jul-2009 09:22:22] <rmatte> If you're looking at doing UI stuff you can always look at rocket's AIX pack for inspiration
[24-Jul-2009 09:22:34] <jb> yeah
[24-Jul-2009 09:22:52] <jb> where is rocket?!
[24-Jul-2009 09:23:02] <rmatte> I'm going to eventually grab that and try to make something WMI based out of it if I can
[24-Jul-2009 09:25:39] <twm1010> If a dinky little program like OpManager can do this, I have faith Zenoss can do it
[24-Jul-2009 09:27:53] <twm1010> I suppose everyone will have different information store setups, so it will be hard to setup templates that go with that
[24-Jul-2009 09:28:12] <twm1010> I could probably do MTA stacks, send queues, etc, real generic like
[24-Jul-2009 09:28:21] <jb> hmm
[24-Jul-2009 09:28:26] <jb> i'll show you what i graph for exchange
[24-Jul-2009 09:29:20] <jb> Exchange - Mailbox Operations
[24-Jul-2009 09:29:24] <jb> Exchange - RPC Operations
[24-Jul-2009 09:29:28] <jb> Exchange - Store CPU Utilization
[24-Jul-2009 09:29:32] <jb> SMTP Server - Local Queue Length
[24-Jul-2009 09:29:36] <jb> Exchange - Client RPC Operations
[24-Jul-2009 09:29:41] <jb> SMTP Server - Message Delivery Rate
[24-Jul-2009 09:29:44] <jb> Exchange - Queues
[24-Jul-2009 09:29:58] <jb> Exchange - Active User Connections
[24-Jul-2009 09:30:03] <jb> Exchange - SMTP Message Delivery
[24-Jul-2009 09:30:11] <jb> Exchange - Local and Remote Delivery
[24-Jul-2009 09:30:15] <jb> Exchange - Client RPC Latency
[24-Jul-2009 09:30:23] <jb> an IS tab would be cool though
[24-Jul-2009 09:32:45] <twm1010> jb, that's from enterprise?
[24-Jul-2009 09:32:53] <jb> yeh i use zenwinperf
[24-Jul-2009 09:32:58] <jb> the enterprise zenpack came with a few of those
[24-Jul-2009 09:33:05] <jb> but i added additional datasources and graphs
[24-Jul-2009 09:47:37] <rmatte> hmmm, well that's kind of anoying
[24-Jul-2009 09:47:52] <rmatte> if I go to /Devices and select my interface templates there is no option to add them to a ZenPack
[24-Jul-2009 09:48:00] <rmatte> and I certainly don't want to add all of /Devices to one
[24-Jul-2009 09:48:55] <rmatte> guess I'll have to add the items then delete everything except for those
[24-Jul-2009 09:49:05] <rmatte> actually, I don't even think that'll work, hmmm
[24-Jul-2009 09:50:51] <rmatte> well, I'm going to make a device class called /Interfaces and just pack them in there, then move them out when installed
[24-Jul-2009 09:52:10] <twm1010>
[24-Jul-2009 09:52:22] <rmatte> pretty annoying though
[24-Jul-2009 09:52:37] <twm1010> might perhaps be on purpose
[24-Jul-2009 09:52:48] <twm1010> imagine a zenpack overwriting someone's ethernetCSmacd?
[24-Jul-2009 09:53:03] <rmatte> well, I guess
[24-Jul-2009 09:53:19] <rmatte> that's why ZenPacks need to be tested before putting them up for download though
[24-Jul-2009 09:53:21] <rmatte>
[24-Jul-2009 10:16:47] <dberger> is there a vmware appliance with zenoss pre-installed?
[24-Jul-2009 10:43:55] <rmatte> Ok, for the "Aggregate Reports" report, I only see 1 out of the 5 graphs, any idea why?
[24-Jul-2009 10:44:14] <rmatte> They appear to be generated via form plugin located in Zope, which I can't even look at apparently
[24-Jul-2009 10:44:59] <jmp242> dberger: there is, but I don't really recommend it
[24-Jul-2009 10:48:25] <rmatte> nice, I got my scripts packaged up in to my SNMP performance monitor ZenPack, woot
[24-Jul-2009 10:49:04] <rmatte> ah damn, just noticed I'm missing something in the templates
[24-Jul-2009 10:52:31] <dberger> jmp242, why not?
[24-Jul-2009 11:03:06] <jmp242> Well - from the forums it doesn't seem to work as well as a clean install
[24-Jul-2009 11:03:21] <jmp242> and it uses an odd rpath linux, which there is limited info about that I've found.
[24-Jul-2009 11:03:28] <jmp242> I'd recommend using RHEL5.3
[24-Jul-2009 11:03:38] <jmp242> and just installing the needed RPMs or use the Stack installer
[24-Jul-2009 11:03:58] <dberger> hmmm
[24-Jul-2009 11:04:09] <dberger> interesting
[24-Jul-2009 11:05:25] <dberger> see we have production install, and I have a dev install on my local machine, but we don't really have a home for a test install
[24-Jul-2009 11:06:40] <dberger> we kind of wanted a vm so that we could take snapshots and then make sure the zenpacks we're writing install correctly
[24-Jul-2009 11:08:51] <rmatte> dberger: you can always just build a vm from scratch
[24-Jul-2009 11:08:56] <rmatte> that's what I did for my lab box
[24-Jul-2009 11:12:46] <dberger> yeah, I was kind of coming around to that line of thinking
[24-Jul-2009 11:16:37] <jmp242> Sorry dberger, I did mean just do an EL or OS of choice install in your VM
[24-Jul-2009 11:28:57] <rmatte> http://forums.zenoss.com/viewtopic.php?p=37277#37277
[24-Jul-2009 11:29:09] <rmatte> If anyone has an answer to that please let me know
[24-Jul-2009 11:35:03] <rmatte> mrayzenoss: could you take a look at this please: http://forums.zenoss.com/viewtopic.php?p=37277#37277
[24-Jul-2009 11:46:58] <rmatte> Here's what I have so far, so to go one month back it's: 1 Month:
[24-Jul-2009 11:47:04] <rmatte> http://lhin01.novanoc.com:8180/zport/RenderServer/plugin?name=upperLower&arg=--units-exponent%3D6&title=Network%20Output%2FInput&label=Mbs&upper=ifOutOctets&lower=ifInOctets&upperLabel=Out&lowerLabel=In%20&rpn=%2C8%2C*&width=500&height=140&miny=&maxy=&devices=.*&start=-1m&u=1248453478911
[24-Jul-2009 11:47:19] <rmatte> the value at the end seems to be the one to modify
[24-Jul-2009 11:47:28] <rmatte> now I just need to figure out how to put an end value as well
[24-Jul-2009 11:47:53] <rmatte> and also, how those values are being determined
[24-Jul-2009 11:49:24] <gwb235> earlier this week somebody posted about the updated Netapp community zenpack: http://www.zenoss.com/community/projects/zenpacks/netapp
[24-Jul-2009 11:50:53] <gwb235> after creating the /Mibs/Storage subfolder, trying to install the new pack errors with:
[24-Jul-2009 11:51:02] <gwb235> WARNING:zen.AddToPack:Unable to find context path /zport/dmd/Mibs/Storage/mibs (line 2082 ?) for NETWORK-APPLIANCE-MIB WARNING:zen.AddToPack:Not committing any changes ERROR: zenpack command failed. Reason: Unable to create object using the following attributes: * id: /zport/dmd/Mibs/Storage/mibs/NETWORK-APPLIANCE-MIB
[24-Jul-2009 11:51:35] <gwb235> ?
[24-Jul-2009 11:53:28] <rmatte> nevermind, we figured it out finally
[24-Jul-2009 11:53:44] mrayzenoss1 is now known as mrayzenoss
[24-Jul-2009 12:00:43] <gwb235> nevermind, I deleted /Mibs/Storage, re-created, and install works fine now
[24-Jul-2009 12:08:36] <rmatte> oh
[24-Jul-2009 13:23:36] <dollarbang> hello folks, anyone tried dumping the device list under 2.4.1?
[24-Jul-2009 13:29:15] <dberger> anyone know how to change the --allow-hosts on a zenpack install?
[24-Jul-2009 13:41:24] <rmatte> let me guess... brocade?
[24-Jul-2009 13:41:50] <rmatte> (the zenpack you're trying to install)
[24-Jul-2009 13:44:32] <twm1010> Anyone here use custom properties?
[24-Jul-2009 13:52:18] <rmatte> like custom zProperties?
[24-Jul-2009 13:52:51] <twm1010> well, let's say you wanted to add a custom field to the edit tab of a device
[24-Jul-2009 13:53:23] <twm1010> perhaps i should say custom attribute
[24-Jul-2009 13:53:41] <twm1010> so you can put in things like, the purchase date of equipment, service contract #, etc.
[24-Jul-2009 13:54:47] <twm1010> The custom thing seems to cut it, but it'd be nice if things like that were right there with the rest
[24-Jul-2009 13:54:59] <twm1010> i just found a forum article on how to add things like that to the status page, very cool
[24-Jul-2009 13:55:11] <rmatte> I think more what you're looking for is if you go to a device properties page, and go to More -> Custom
[24-Jul-2009 13:55:26] <rmatte> possibly
[24-Jul-2009 13:55:28] <rmatte> afk for a sec
[24-Jul-2009 13:55:31] <twm1010> K
[24-Jul-2009 13:57:47] <rmatte> I know we were looking at possibly doing something similar for stuff like device contacts and what not
[24-Jul-2009 13:57:56] <rmatte> but I never really got very in depth with it
[24-Jul-2009 13:59:24] <rmatte> apparently the properties in Custom Properties are defined in some "Custom Schema" according to the admin guide
[24-Jul-2009 14:01:27] <twm1010> Yeah, I see that... I just didnt' know if it was real entrenched or what.
[24-Jul-2009 14:01:39] <twm1010> But, I'd like to get away from these stupid spreadsheets we keep everything in.
[24-Jul-2009 14:01:52] <twm1010> Solarwinds lets you extend as much as you like
[24-Jul-2009 14:01:53] <jb> whats the forum article?
[24-Jul-2009 14:02:12] <twm1010> http://forums.zenoss.com/viewtopic.php?t=5805
[24-Jul-2009 14:02:16] <jb> thanks
[24-Jul-2009 14:02:17] <rmatte> yeh, we're moving towards using a central sql database to store everything now
[24-Jul-2009 14:02:27] <jb> yeh we store everything in another sql database
[24-Jul-2009 14:03:01] <twm1010> we make all sorts of good decisions here.
[24-Jul-2009 14:03:29] <twm1010> for instance, it blows my mind that we have an excel spreadsheet with superuser passwords in it, to a lot of different things
[24-Jul-2009 14:03:30] <rmatte> twm1010: every IT company is like that in my experience.
[24-Jul-2009 14:04:35] <twm1010> I suppose... eh, I like it all integrated. Did I tell you this is my second time doing this? I had a good bit of this information in a Zenoss 2.1 VM 18 months ago. Then we put in SolarWinds, now I'm rebuilding.
[24-Jul-2009 14:04:53] <rmatte> You're basically looking at adding something like the "Comments" field?
[24-Jul-2009 14:05:10] <twm1010> Yep, just want to keep it clean, I could probably come up with 10 off the top of my head that would be handy.
[24-Jul-2009 14:05:27] <twm1010> for instance, a link to the management card would be damn skippy
[24-Jul-2009 14:05:43] <dollarbang> Can you execute the Zenpatch with a stack Debian Stack install?
[24-Jul-2009 14:05:51] <twm1010> dollarbang: pretty sure
[24-Jul-2009 14:05:53] <mrayzenoss> dollarbang: yes
[24-Jul-2009 14:06:23] <rmatte> twm1010: there's already a "links" section
[24-Jul-2009 14:06:48] <twm1010> how is that populated?
[24-Jul-2009 14:07:06] <rmatte> that's the part that I forget, but I know there's a way to populate that
[24-Jul-2009 14:07:08] <mrayzenoss> zProperty, you can put whatever you want in there.  HTML, Javascript
[24-Jul-2009 14:07:09] <rmatte> I remember seeing it
[24-Jul-2009 14:07:17] <rmatte> oh right, zProperty
[24-Jul-2009 14:07:20] <dollarbang> mrayzenoss and twm1010: thanks, now to find that commands doc
[24-Jul-2009 14:07:32] <mrayzenoss> and it has references to the device
[24-Jul-2009 14:08:05] <mrayzenoss> example:    <a href="http://yourwiki/Devices?id=${here/id}">Wiki Page</a>
[24-Jul-2009 14:08:06] <twm1010> ah, very good, that's how the fella did it with the Netware boxes
[24-Jul-2009 14:08:38] <twm1010> <a href='http://${here/manageIp}:8008' target='_'>Novell Remote Manager: ${here/id}</a>
[24-Jul-2009 14:09:46] <rmatte> twm1010: you could use a mixture of links and comments
[24-Jul-2009 14:10:45] <mrayzenoss> You could also put a big block of HTML with Javascript with links to lots of things (consoles, wikis, VNC, ticket tracking)
[24-Jul-2009 14:29:22] <rmatte> It would be nice if Zenoss had some sort of device records component where you could add fields from the UI (for contacts, login details, etc...) and then pull from
[24-Jul-2009 14:30:24] <rmatte> It would add value to it since users wouldn't have to go out and setup a completely separate database to manage the data
[24-Jul-2009 14:32:35] <gwb2351> we've thought about adding a "tab" to the devices bits that would allow us to query into our (existing) CMDB and pull that information
[24-Jul-2009 14:32:53] <rmatte> gwb2351: that would be pretty neat
[24-Jul-2009 14:54:25] <jmp242> So I'm already amazed and happy with zLinks + iFrames
[24-Jul-2009 14:54:42] <jmp242> I'm wondering though, I don't know js so I can't say how to do this
[24-Jul-2009 14:54:59] <jmp242> but I have OCSNG and GLPI that are really cool to link to the entries for the device in Zenoss
[24-Jul-2009 14:55:05] <jmp242> as iframes on the status page.
[24-Jul-2009 14:55:49] <jmp242> But they don't all use the same ID
[24-Jul-2009 14:56:02] <jmp242> I'm wondering if there'd be a way to do a lookup based on DNS name
[24-Jul-2009 14:57:22] <mrayzenoss> so neither the id nor the manageIp will work?
[24-Jul-2009 15:03:20] <bigegor> mrayzenoss: how to rename rrdtemplate from ZenPack Install script?
[24-Jul-2009 15:03:21] <jmp242> Hmmm
[24-Jul-2009 15:03:31] <jmp242> I wonder if manageIP would work
[24-Jul-2009 15:03:54] <mrayzenoss> King Crab is going to introduce "title" I believe
[24-Jul-2009 15:04:35] <mrayzenoss> to help distinguish duplicate IP and IDs (like in virtualization)
[24-Jul-2009 15:10:07] <mrayzenoss> bigegor: digging
[24-Jul-2009 15:16:03] <mrayzenoss> bigegor: take a look at http://www.zenoss.com/community/docs/zenoss-dev-guide/2.4.2/ch03s05.html#zenpacks_migrating
[24-Jul-2009 15:16:34] <mrayzenoss> bigegor: ZenJMX has examples of the migration being used between versions, you could use this to make changes between versions
[24-Jul-2009 15:17:20] <mrayzenoss> http://dev.zenoss.com/svn/trunk/zenpacks/ZenPacks.zenoss.ZenJMX/ZenPacks/zenoss/ZenJMX/migrate/
[24-Jul-2009 15:39:11] <rmatte> hehe, GLPI is very french
[24-Jul-2009 15:39:37] <rmatte> looking at some screenshots and they have "Non" in a dropdown instead of No
[24-Jul-2009 15:39:49] <rmatte> Reparation instead of repaired
[24-Jul-2009 15:41:35] <rmatte> (I speak french, just commenting)
[24-Jul-2009 15:45:24] <rmatte> ah, this is more for a LAN Admin type job
[24-Jul-2009 15:45:35] <rmatte> computer/software inventory with technician ticketing
[24-Jul-2009 16:10:29] <kd4zay> hi rmatte
[24-Jul-2009 16:26:21] <rmatte> hi
[24-Jul-2009 16:33:02] <rmatte> well, I'm out of here, eugh, it's raining, hopefully I don't get soaked
[24-Jul-2009 16:33:11] <rmatte> have a good weekend gents
[24-Jul-2009 17:23:23] <kd4zay> anyone here setup a sonicwall with zenoss ? i have add it as a device and enabled snmp but im not getting the correct metrics back.
[24-Jul-2009 17:30:07] <mrayzenoss> later everyone
[24-Jul-2009 19:00:24] bzed_ is now known as bzed
[25-Jul-2009 00:00:46] [disconnected at Sat Jul 25 00:00:46 2009]
[25-Jul-2009 00:00:46] [connected at Sat Jul 25 00:00:46 2009]
[25-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[25-Jul-2009 05:16:56] <Kisharcos>  hali
[25-Jul-2009 18:48:10] bzed_ is now known as bzed
[25-Jul-2009 18:49:16] <Diddi> alright, I'm having a hard time understanding how events are connected to devices... I know have an event in event class /perf/snmp saying 'error reading value for millivolts on <hostname>, bad oid...'. Where did I define that zenoss should check for millivolt? how do I delete the check?
[25-Jul-2009 18:50:00] <Diddi> i'm reading the admin manual aswell.. though I'm not sure how templates and zProperties act yet..
[25-Jul-2009 18:59:22] <sergeymasushko> Diddi: you should open the device and look for templates which are assigned to it.
[25-Jul-2009 18:59:43] <Diddi> Device is the only one
[25-Jul-2009 18:59:55] <Diddi> in /Devices/Server/Linux
[25-Jul-2009 19:00:51] <Diddi> I had a little try with the Dell zenpack, but it's now uninstalled and I can't find anything related to it anymore
[25-Jul-2009 19:00:52] <sergeymasushko> ok. open device page... click on the triangle -> more -> templates
[25-Jul-2009 19:02:33] <Diddi> alright
[25-Jul-2009 19:04:00] <Diddi> the only thing listed there is "Device", with definition path "/Devices/Server/Linux"
[25-Jul-2009 19:05:03] <sergeymasushko> open that template and look for OID which you have seen in the event.
[25-Jul-2009 19:05:52] <Diddi> not there
[25-Jul-2009 19:06:08] <sergeymasushko> dunno
[25-Jul-2009 19:07:28] <sergeymasushko> btw, the event was created before you uninstalled zenpack?
[25-Jul-2009 19:07:49] <Diddi> ah, perhaps it's not automagically cleared..
[25-Jul-2009 19:07:57] <Diddi> and, yes, to your question
[25-Jul-2009 19:07:59] <sergeymasushko> probably
[25-Jul-2009 19:08:21] <sergeymasushko> then just move the event to the history
[25-Jul-2009 19:08:31] <Diddi> yeah
[25-Jul-2009 19:08:41] <sergeymasushko> or wait some time and the event should be aged...
[25-Jul-2009 19:09:01] <sergeymasushko> by default aging time is set to 4 hours.
[25-Jul-2009 19:09:25] <Diddi> ah, i'll wait and see then
[25-Jul-2009 19:09:46] <sergeymasushko> one thing...
[25-Jul-2009 19:09:55] <sergeymasushko> what severity of the event?
[25-Jul-2009 19:10:02] <Diddi> debug
[25-Jul-2009 19:10:31] <sergeymasushko> than it will be aged....
[25-Jul-2009 19:10:48] <Diddi> ok
[25-Jul-2009 19:11:58] <sergeymasushko> there is an option 'age events with severity ... and higher' (something like it)
[25-Jul-2009 19:12:46] <Diddi> ah, yeah
[25-Jul-2009 19:13:22] <sergeymasushko> btw
[25-Jul-2009 19:13:34] <sergeymasushko> can you help me with one thing...?
[25-Jul-2009 19:13:55] <Diddi> uh, maybe? (:
[25-Jul-2009 19:14:17] <sergeymasushko> can you open device template and tell what there defined in the cpu threshold ... I have broken my template
[25-Jul-2009 19:14:41] <Diddi> heh
[25-Jul-2009 19:15:42] <Diddi> low_cpu_idle?
[25-Jul-2009 19:16:06] <sergeymasushko> yes
[25-Jul-2009 19:16:33] <Diddi> says, min value: 2
[25-Jul-2009 19:16:34] <Diddi>
[25-Jul-2009 19:16:46] <Diddi> ssCpuIdle_ssCpuIdle as datapoint
[25-Jul-2009 19:17:29] <Diddi>  /Perf/CPU as event class, warning as severity
[25-Jul-2009 19:17:33] <Diddi> pretty much it
[25-Jul-2009 19:18:04] <sergeymasushko> ssCpuRawIdle 1.3.6.1.4.1.2021.11.53.0 ???
[25-Jul-2009 19:18:40] <Diddi> I don't have that datapoint defined
[25-Jul-2009 19:18:53] <Diddi>  ssCpuIdle   1.3.6.1.4.1.2021.11.11.0
[25-Jul-2009 19:19:04] <sergeymasushko> oh. ok. thank you.
[25-Jul-2009 19:19:08] <Diddi> (:
[25-Jul-2009 19:19:10] <Diddi> no probs
[25-Jul-2009 19:28:00] <Diddi> is it not possible to have more than one template on one device?
[25-Jul-2009 19:30:01] <Diddi> ah, it's inherited..
[25-Jul-2009 19:30:55] <sergeymasushko> ? what does it maen? I'm not a native speaker
[25-Jul-2009 19:32:39] <Diddi> in a tree of device classes, the lower classes have the templates defined in the higher classes
[25-Jul-2009 19:33:00] <Diddi> something like that
[25-Jul-2009 19:33:20] <sergeymasushko> yes
[25-Jul-2009 22:46:24] <cmdln> hello
[25-Jul-2009 22:48:30] <cmdln> I keep getting errors with the httpmonitor zenpack
[26-Jul-2009 00:00:46] [disconnected at Sun Jul 26 00:00:46 2009]
[26-Jul-2009 00:00:46] [connected at Sun Jul 26 00:00:46 2009]
[26-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[26-Jul-2009 10:21:51] <Diddi> I'm still trying to figure out how events works... I'm currently having an even "SNMP agent down" (as it should be), but where is this event defined? Where do I say that this event should appear for a certain device?
[26-Jul-2009 10:22:15] <Diddi> I've read the admin manual, but I think I might be overseeing something
[26-Jul-2009 12:40:59] <Diddi> Is it possible to look for a process on a device without letting the "Model device" discover the process first?
[26-Jul-2009 12:41:40] <Diddi> I currently have ntpd in /Processes, but only devices recently remodeled will be monitored for that process...
[26-Jul-2009 12:42:16] <Diddi> and I also don't want to monitor ntpd on all device where it's found
[26-Jul-2009 12:42:56] <Diddi> the manual says "Now you are monitoring this process, so after a remodel (which you can do manually or it occurs at 6 hour intervals), it will show every device (occurrence) where this process is running. As such, the process is now being monitored wherever it occurs.", but isn't it possible to make manual changes to this?
[26-Jul-2009 13:54:03] <cmdln> hello
[26-Jul-2009 13:54:28] <cmdln> I am trying to monitor my linksys router with ddwrt loaded.
[26-Jul-2009 13:58:18] <cmdln> 1.3.6.1.2.1.2.2.1.10.6 appears to be the OID for inbound traffic on my wan interface, I added a new datasource with that oid and made a new graph with that datasource as a graph point but the graph never gets displayed. I do get a result when I test the datasouce against the router.
[26-Jul-2009 16:18:08] <otakup0pe> heh i saw that too cmdln
[26-Jul-2009 16:18:32] <otakup0pe> side note ; what is the syntax to refer to things extracted from the regex in a mapping ?
[26-Jul-2009 16:18:44] <otakup0pe> i want to build a new evt.summary i've done this in the past but it was years ago
[26-Jul-2009 18:29:21] <daleamon> mrayzenoss... quick question since you were so helpful the last time I dropped in...
[26-Jul-2009 18:29:40] <daleamon> Why would the netmap show as empty after a discovery finds 57 hosts?
[26-Jul-2009 20:28:26] <SilverFox> so, if I see via tcpdump/ngrep that a trap is reaching the zenoss server, but I don't see it in event console, where might I find it?
[26-Jul-2009 21:26:49] <SilverFo1> what might cause this:
[26-Jul-2009 21:26:50] <SilverFo1> (1, "Can't create/write to file '/usr/local/zenoss/mysql/data/#sql_7511_0.MYI' (Errcode: 13)")
[26-Jul-2009 21:27:01] SilverFo1 is now known as SilverFox
[26-Jul-2009 23:25:44] <daleamon> No answer yet... anyone have an idea why the Network diagram does not show the results of a discover that picked up 57 workstations and other hosts?
[27-Jul-2009 00:00:46] [disconnected at Mon Jul 27 00:00:46 2009]
[27-Jul-2009 00:00:46] [connected at Mon Jul 27 00:00:46 2009]
[27-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[27-Jul-2009 08:38:02] <ckrough> anyone know if it's really necessary to 'recreate' an RRD after changing the MIN or MAX value? I see other people recreate after making the change, but the documentation for rrdtune doesnt say it's necessary
[27-Jul-2009 08:39:21] <jb> anyone here monitoring several windows boxes that has noticed window's snmpd crashing?
[27-Jul-2009 08:41:19] <ckrough> running snmp-informant?
[27-Jul-2009 08:41:40] <jb> yeah
[27-Jul-2009 08:41:44] <jb> it just started
[27-Jul-2009 08:57:05] <rmatte> ckrough: max and min value for the datapoint itself?
[27-Jul-2009 09:03:24] <ckrough> yeah
[27-Jul-2009 09:13:59] <mrayzenoss> Greetings and salutations, hopefully a good weekend was had by all
[27-Jul-2009 09:18:06] <ckrough> morning mrayzenoss
[27-Jul-2009 09:22:33] <jb> hmm.. did the behavior of zenperfsnmp change in 2.4? I am getting threshold alerts that zenperfsnmp cycles are taking ~300 seconds, but when I manually run "zenperfsnmp run -v10" it takes only a few seconds.. and it doesn't even look like its grabbing SNMP data from all hosts.
[27-Jul-2009 09:24:06] <mrayzenoss> that doesn't sound familiar to me
[27-Jul-2009 09:24:17] <jb> INFO:zen.zenperfsnmp:******** Cycle completed ********
[27-Jul-2009 09:24:17] <jb> INFO:zen.zenperfsnmp:Sent 0 OID requests
[27-Jul-2009 09:24:17] <jb> INFO:zen.zenperfsnmp:Queried 0 devices
[27-Jul-2009 09:24:20] <jb> isnt that odd?
[27-Jul-2009 09:27:03] <mrayzenoss> mine does that too, I believe if you run zenperfsnmp when the daemon is still running, it will try to grab anything from zenhub that is in the queue
[27-Jul-2009 09:27:12] <mrayzenoss> but since the other daemon is running, the queue is empty
[27-Jul-2009 09:27:13] <jb> ah
[27-Jul-2009 09:27:24] <jb> i'm just trying to determine whats causing it to take so long to run
[27-Jul-2009 09:28:05] <mrayzenoss> yeah, shutting down the other daemon and running it with "zenperfsnmp run -v10" is the way to go
[27-Jul-2009 09:28:24] <jb> gotcha
[27-Jul-2009 09:31:27] <mrayzenoss> is it just 1 device that's having problems?
[27-Jul-2009 09:31:31] <mrayzenoss> or multiple?
[27-Jul-2009 09:55:10] <rmatte> http://dev.zenoss.org/trac/ticket/5366
[27-Jul-2009 09:56:56] <mrayzenoss> that'll come up in tomorrow's defect review
[27-Jul-2009 09:56:58] <mrayzenoss> sounds legit to me
[27-Jul-2009 09:57:09] <mrayzenoss> but I imagine it won't be in the 2.4.x series
[27-Jul-2009 09:57:36] <rmatte> cool
[27-Jul-2009 09:57:58] <rmatte> as long as it gets fixed at some point
[27-Jul-2009 09:58:57] <mrayzenoss> speaking of things getting fixed at some point... Egor submitted a WMI Data Source ZenPack which handles the passwords in the logs issue
[27-Jul-2009 09:59:20] <rmatte> nice
[27-Jul-2009 09:59:42] <rmatte> is it just a framework to build WMI ZenPacks on, or does it actually provide performance data?
[27-Jul-2009 09:59:54] <mrayzenoss> he sent me 3 ZenPacks on Friday
[27-Jul-2009 09:59:59] <mrayzenoss> 1 for the WMI data source
[27-Jul-2009 10:00:07] <mrayzenoss> 1 for WMI Perf data for Windows boxes
[27-Jul-2009 10:00:11] <mrayzenoss> and 1 for IIS6
[27-Jul-2009 10:00:19] <rmatte> I'll have an SNMP Performance Monitor ZenPack for you some time in the near future
[27-Jul-2009 10:00:27] <mrayzenoss> the WMI Perf one is meant to eventually replace the need for SNMP Informant
[27-Jul-2009 10:00:28] <rmatte> I just need to fix it up a bit
[27-Jul-2009 10:00:36] <jb> i need to figure out whats causing my windows snmpd.exe's to crash
[27-Jul-2009 10:00:41] <jb> anybody seen that?
[27-Jul-2009 10:00:58] <rmatte> My ZenPack replaces the need for SNMP informant, but does it via SNMP
[27-Jul-2009 10:01:08] <mrayzenoss> cool
[27-Jul-2009 10:01:15] <rmatte> jb: you have many crashing?
[27-Jul-2009 10:01:16] <mrayzenoss> it would be nice to have a pure SNMP and a pure WMI in Core
[27-Jul-2009 10:01:26] <jb> i had 5 crash over the weekend
[27-Jul-2009 10:01:29] <rmatte> any recent patching?
[27-Jul-2009 10:01:33] <jb> nope
[27-Jul-2009 10:01:42] <mrayzenoss> all same OS version and hotpatch version?
[27-Jul-2009 10:01:46] <rmatte> Have you checked event logs?
[27-Jul-2009 10:01:46] <mrayzenoss> err... service pack
[27-Jul-2009 10:01:51] <jb> havent verified that yet
[27-Jul-2009 10:01:57] <jb> yeah i have checked logs
[27-Jul-2009 10:01:59] <jb> nothing informative
[27-Jul-2009 10:02:08] <rmatte> hmmm
[27-Jul-2009 10:02:15] <rmatte> how many times have they crashed, just the once?
[27-Jul-2009 10:02:25] <jb> 2 or 3 times now
[27-Jul-2009 10:02:38] <rmatte> we actually had 3 crash at once the other day, but after a reboot it's been good since
[27-Jul-2009 10:05:20] <rmatte> but yeh, there must be a commonality between those servers
[27-Jul-2009 10:06:15] <jb> yeah i'll dig in and try to find it
[27-Jul-2009 10:07:53] <rmatte> mrayzenossP: the main issue with my ZenPack is that it's capable of monitoring and thresholding against each individual cpu/core, but since there's no way to make templates dynamic, I've had to create different groups, like /Server/Windows/1 Processor, 2 Processors, 4 Processors, etc... You need to do an snmpwalk to find out how many processors SNMP is seeing on a given server then move it to the correct group. It works quite well.
[27-Jul-2009 10:08:17] <rmatte> What I may do though for a public release is only monitor based on the Total amount of CPU usage and only have 1 group
[27-Jul-2009 10:09:05] <rmatte> Do Egor's WMI ZenPacks use a collector plugin, or command based datapoints?
[27-Jul-2009 10:10:41] <mrayzenoss> rmatte: here's an example of using the model to decide which template to bind to: trac-zenpacks/browser/zenpacks/ZenPacks.community.HPSIMMonitor/ZenPacks/community/HPSIMMonitor/modeler/plugins/community/snmp/HPDeviceTemplateMap.py
[27-Jul-2009 10:10:56] <mrayzenoss> somewhat dynamic
[27-Jul-2009 10:12:05] <mrayzenoss> Egor's ZenPack adds WMI as a datasource option and creates a new zenperfwmi daemon that leverages the built-in stuff in Zenoss
[27-Jul-2009 10:12:28] <mrayzenoss> so performance should be much better than WMI calls via wmic
[27-Jul-2009 10:12:50] <cain22> howdy all
[27-Jul-2009 10:13:24] <jb> nice
[27-Jul-2009 10:13:32] <jb> so its pretty much just like zenwinperf?
[27-Jul-2009 10:13:50] <mrayzenoss> jb: not exactly
[27-Jul-2009 10:14:04] <mrayzenoss> zenwinperf uses the perfmon interface, zenperfwmi uses the WMI interface
[27-Jul-2009 10:14:21] <mrayzenoss> but they get to the same data
[27-Jul-2009 10:14:33] <jb> yeah
[27-Jul-2009 10:14:38] <mrayzenoss> that's how it was explained to me, I'm gonna dig into it a bit more
[27-Jul-2009 10:15:41] <rmatte> I still need to convert mine to a collector plugin
[27-Jul-2009 10:15:46] <rmatte> it'd make it much more efficient
[27-Jul-2009 10:16:10] <rmatte> I have the basic method on how to collect the data done, and it works perfectly, but it's done in bash scripts right now
[27-Jul-2009 10:16:40] <rmatte> Maybe Egor can help me out with it or something, I'll have to discuss it with him next time I see him around
[27-Jul-2009 10:20:40] <rmatte> I'm glad to see that he's getting in to the more advanced WMI stuff though
[27-Jul-2009 10:30:20] <cain22> anyone receive an error when they if they try to set a winservice to false?
[27-Jul-2009 10:30:37] <cain22> i get a BadRequest: The property port does not exist
[27-Jul-2009 10:49:40] <mrayzenoss> don't recall if I mentioned these here, but there are 3 new ZenPacks
[27-Jul-2009 10:49:45] <mrayzenoss> Fedora Linux http://www.zenoss.com/community/projects/zenpacks/fedora
[27-Jul-2009 10:49:56] <mrayzenoss> MGE UPS: http://www.zenoss.com/community/projects/zenpacks/mge-ups
[27-Jul-2009 10:50:06] <mrayzenoss> Ubuntu Linux: http://www.zenoss.com/community/projects/zenpacks/ubuntu
[27-Jul-2009 11:16:11] <ckrough> How do you guys do most of your REST calls in-house? urllib2 or xmlrpc?
[27-Jul-2009 11:19:11] <daleamon> Seems more lively today... can anyone tell me why the network map would not show the 57 workstations etc found by discovery?
[27-Jul-2009 11:20:02] <jb> were their routing tables disovered properly?
[27-Jul-2009 11:28:33] <sacpinball> mrayzenoss i got a quick question for you, which device class should i put an IBM BladeCenter Chasis under?
[27-Jul-2009 11:30:37] <mrayzenoss> hmmm... you could add a new organizer
[27-Jul-2009 11:30:51] <mrayzenoss> perhaps /Devices/Hardware?
[27-Jul-2009 11:31:05] <mrayzenoss> or /Devices/Rack?
[27-Jul-2009 11:31:17] <mrayzenoss> surprised I haven't been asked that before... good question
[27-Jul-2009 11:35:49] <daleamon> jb: no idea. Most of them are on the same LAN. Not much to discover from a bunch of windoze office workstations.
[27-Jul-2009 11:38:00] <jb> check in the OS tab of the devices
[27-Jul-2009 11:38:17] <jb> the network map won't work unless your devices have a fully disovered layer 3 routing table
[27-Jul-2009 11:40:18] <daleamon> This is my first attempt to roll this out... how do I manage that? snmp from the gateway?
[27-Jul-2009 11:45:55] <rmatte> daleamon: do an snmpwalk on a workstation and see if you see route info
[27-Jul-2009 11:46:26] <rmatte> or check the device in Zenoss, you should see route info listed at the bottom on the OS tab
[27-Jul-2009 11:56:07] <daleamon> I picked a random Windoz box and it shows no interfaces, no routes, nada. Do I have to do this manually for all 57?
[27-Jul-2009 11:56:15] <daleamon> ie, set up the info for them?
[27-Jul-2009 12:05:32] <cain22> do all devices have to have a route in the table for network map to work correctly
[27-Jul-2009 12:05:40] <cain22> including switches?
[27-Jul-2009 12:06:42] <rmatte> daleamon: you need to make sure that the SNMP agent is actually installed on all 57, and configured with a community string
[27-Jul-2009 12:06:52] <rmatte> also make sure it's configured to accept SNMP requests from your Zenoss box
[27-Jul-2009 12:07:08] <rmatte> otherwise Zenoss has no way of grabbing any info off any of those devices
[27-Jul-2009 12:08:07] <rmatte> cain22: you need as many routes as possible, I don't think it applies to switches unless they are router/switch combos
[27-Jul-2009 12:08:59] <al_x> what does it mean when i see this in my snmpd log
[27-Jul-2009 12:09:01] <al_x> Did not find 'mgmt' in module RFC1155-SMI (/usr/share/snmp/mibs/rfc-1213.mib
[27-Jul-2009 12:09:10] <al_x> it happened sometime yesterday and snmpd stopped responding
[27-Jul-2009 12:09:37] <cain22> we have different locations which are connected by switches...and they don't show up on the map as being connected
[27-Jul-2009 12:12:56] <rmatte> cain22: you're sure they all modeled correctly?
[27-Jul-2009 12:14:03] <rmatte> al_x: not sure, seeing what I can dig up
[27-Jul-2009 12:14:26] <rmatte> does /usr/share/snmp/mibs/rfc-1213.mib even exist on that system?
[27-Jul-2009 12:16:32] <rmatte> al_x: it sounds like something is messed up with your mibs in /usr/share/snmp/mibs/
[27-Jul-2009 12:17:41] <rmatte> rmatte@lab01:/usr/share/snmp/mibs$ cat RFC1213-MIB.txt | grep mgmt
[27-Jul-2009 12:17:42] <rmatte>         mgmt, NetworkAddress, IpAddress, Counter, Gauge,
[27-Jul-2009 12:17:42] <rmatte> mib-2      OBJECT IDENTIFIER ::= { mgmt 1 }
[27-Jul-2009 12:17:42] <rmatte>                 netmgmt(3),     -- management protocol
[27-Jul-2009 12:18:26] <rmatte> try "cat /usr/share/snmp/mibs/rfc-1213.mib | grep mgmt
[27-Jul-2009 12:18:40] <rmatte> (mine are txt files for whatever reason)
[27-Jul-2009 12:32:02] <daleamon> I'm looking at Settings->Commands->snmpwalk... where do I find zSnmpCommunity and other parameters for snmp?
[27-Jul-2009 12:36:54] <ckrough> most snmp stuff is in the zproperties of the class or device
[27-Jul-2009 12:37:26] <ckrough> a la: https://your_server/zport/dmd/Devices/Network/zPropertyEdit
[27-Jul-2009 12:47:56] <al_x> rmatte: thanks for the info
[27-Jul-2009 12:48:03] <al_x> i wonder what caused it to stop working suddenly on sunday
[27-Jul-2009 12:48:58] <daleamon> if 'your_server' = 'machine running zenoss' I can't get to that because that port is redirected elsewhere. In any case, all I want to do is check and be able to modify the community and password string zenoss is using for discovery
[27-Jul-2009 12:50:46] <ckrough> daleamon: try it without https, the trailing part of the URL is an example of one of the places where you can set that info
[27-Jul-2009 12:51:06] <ckrough> http://your_server:8080/zport/dmd/Devices/Network/zPropertyEdit
[27-Jul-2009 12:51:30] <rmatte> al_x: does anyone else administer that box?
[27-Jul-2009 12:51:32] <daleamon> ah, that should be reachable :-)
[27-Jul-2009 12:52:00] <daleamon> not yet. Just installing it, not even in real service yet.
[27-Jul-2009 12:52:00] <ckrough> ya, sorry, the box I coped from uses https proxying for zen
[27-Jul-2009 12:52:40] <al_x> heh
[27-Jul-2009 12:52:45] <al_x> lots of people
[27-Jul-2009 12:52:59] <al_x> ill ask around
[27-Jul-2009 12:53:57] <daleamon> Thanks. That showed me where it is: Classes_Devices->zProperties gets me there.
[27-Jul-2009 12:56:33] <al_x> the last thing in /var/log/messages was this syslogd 1.4.1: restart.
[27-Jul-2009 12:56:45] <al_x> does that mean the system was rebooted?
[27-Jul-2009 12:56:56] <jb> logs rotated?
[27-Jul-2009 13:00:23] <rmatte> yeh, I think that means it rebooted
[27-Jul-2009 13:00:54] <rmatte> oh no wait, that is log rotation
[27-Jul-2009 13:01:02] <ckrough> it would mean that syslogd restarted, which could have been done by logrotate
[27-Jul-2009 13:01:06] <jb> right.
[27-Jul-2009 13:01:11] <ckrough> whats the 'uptime'
[27-Jul-2009 13:09:28] <konfoo> does zenoss use the /ping command rule to ping hosts, or is it referenced somewhere else?
[27-Jul-2009 13:09:50] <konfoo> i need to 'ping' via some other mechanism since i have icmp blocked on all production hosts
[27-Jul-2009 13:09:57] <konfoo> i.e. snmpwalk -v2c -cpublic ${device/manageIp} .1.3.6.1.2.1.1.5
[27-Jul-2009 13:10:34] <rmatte> it uses a daemon (zenping)
[27-Jul-2009 13:10:42] <mrayzenoss> yes, Zenoss uses ping to determine whether or not a device is reachable
[27-Jul-2009 13:10:54] <mrayzenoss> so the other daemons will check with zenping to determine if something is available
[27-Jul-2009 13:11:17] <mrayzenoss> you'll have to tweak the alerts of SNMP down to get the same effect
[27-Jul-2009 13:11:56] <konfoo> so do i have to go in and edit the zenping command code, or is it configurable via the etc/zenping.conf file somehow?
[27-Jul-2009 13:13:08] <mrayzenoss> what do you want to do exactly?
[27-Jul-2009 13:13:17] <mrayzenoss> you can turn off pinging with the zProperty
[27-Jul-2009 13:13:29] <rmatte> yeh, setup a transform for "SNMP Agent Down" that says something like "ip <bleh> is down" instead, then you need to go in to zProperties and disable ping monitoring in zProperties (zPingMonitorIgnore)
[27-Jul-2009 13:13:41] <konfoo> ah ok
[27-Jul-2009 13:13:45] <mrayzenoss> zPingMonitorIgnore = True
[27-Jul-2009 13:13:54] <rmatte> or you can just assume that "SNMP Agent Down" == Device is down, and just not do a transform
[27-Jul-2009 13:13:57] <rmatte> really up to you
[27-Jul-2009 13:14:36] <konfoo> nod
[27-Jul-2009 13:14:39] <konfoo> thanks for the help
[27-Jul-2009 13:14:39] <mrayzenoss> yeah, the only difference is if you're using multiple collectors to monitor something, you'll get multiple alerts
[27-Jul-2009 13:15:08] <mrayzenoss> instead of the big "Device is unreachable" from ZenPing which suppresses the other events
[27-Jul-2009 13:15:51] <konfoo> i guess i could always block icmp on the border and enable it on the production boxes
[27-Jul-2009 13:16:05] <rmatte> that'd probably be the best way of doing things
[27-Jul-2009 13:16:10] <konfoo> just trying to figure out if zenoss would run more efficiently with it enabled
[27-Jul-2009 13:16:16] <konfoo> and that seems to be what youre inferring
[27-Jul-2009 13:16:27] <rmatte> yeh, it would
[27-Jul-2009 13:16:29] <mrayzenoss> yeah, ping is used to track if devices are available or not
[27-Jul-2009 13:16:37] <mrayzenoss> but there are work-arounds
[27-Jul-2009 13:16:51] <rmatte> ping is good too because it checks every minute or so, where as snmp only kicks off every 5 minutes or so
[27-Jul-2009 13:17:40] <konfoo> ah ok
[27-Jul-2009 13:17:55] <jb> not to mention snmp can occasionaly crash
[27-Jul-2009 13:18:03] <jb> (smmpd)
[27-Jul-2009 13:18:17] <jb> i'd rather rely on ICMP
[27-Jul-2009 13:18:45] <konfoo> is there any way to log ping failures to the zensyslog file?
[27-Jul-2009 13:18:51] <konfoo> so i can pull those into splunk
[27-Jul-2009 13:20:06] <konfoo> i cranked up the verbose level but didnt notice anything relevant going into the file
[27-Jul-2009 13:20:22] <mrayzenoss> zenping logs to $ZENHOME/logs/zenping.log
[27-Jul-2009 13:20:54] <konfoo> ah so it does
[27-Jul-2009 13:22:04] <konfoo> well thats that sorted then
[27-Jul-2009 13:22:13] <mrayzenoss> you can look at changing the logpath for the zenping daemon (and other daemons) under Settings->Daemons->Edit Config
[27-Jul-2009 13:22:38] <mrayzenoss> I haven't modified that, but I assume it can redirect to a syslog if permissions are sufficient
[27-Jul-2009 13:23:15] <konfoo> oh thats fine.. splunk can pull from any flat file as long as you have a location and a regex
[27-Jul-2009 13:23:30] <mrayzenoss> the logseverity can be tweaked there as well
[27-Jul-2009 13:34:59] <ckrough> mrayzenoss: got any good urllib based REST examples in your magic hat? Looking for something to compare to what Im trying
[27-Jul-2009 13:37:04] <mrayzenoss> ckrough asking around, I know the dashboard applet uses xml-rpc
[27-Jul-2009 13:38:14] <ckrough> Im trying to go through a list of templates and configure rrdmin and rrdmax using urllib2 to make the requests, but it doesnt seem to be working.
[27-Jul-2009 13:38:30] <ckrough> (not using https, saw that tidbit in the docs)
[27-Jul-2009 13:40:22] <ckrough> ah, its the redirection. my http://...8080 requests are being redirected to the login page
[27-Jul-2009 13:40:56] <mrayzenoss> hmmm... there's a bug with getRRDValue via REST: http://dev.zenoss.org/trac/ticket/5360
[27-Jul-2009 13:42:57] <ckrough> thanks. we use that functionality, good to know. We havent upgraded to 2.4x yet so it hasnt hit us
[27-Jul-2009 13:48:36] <cain22> does zenoss map layer 2 switches?
[27-Jul-2009 13:49:22] <jb> no
[27-Jul-2009 13:49:37] <jb> the network map isn't layer2 capable..
[27-Jul-2009 13:49:39] <jb> if thats what you mean
[27-Jul-2009 14:05:10] <cain22> right, is that going to be changed with the next update, or anytime soon?
[27-Jul-2009 14:05:22] <jb> probably not
[27-Jul-2009 14:07:40] <mrayzenoss> Product Management is aware of it, since everyone asks for it, but it's not in the next release
[27-Jul-2009 14:09:20] <mrayzenoss> we've got a fairly small dev team and a huge wishlist
[27-Jul-2009 14:12:25] <rmatte> hmmm, I'm trying to find that trac ticket from the other day, about adding the ability to remove packs without removing content via the UI
[27-Jul-2009 14:13:01] <rmatte> found it
[27-Jul-2009 15:08:23] <al_x> if I run an snmpwalk from within zenoss should it use the snmp port specified in the local copy of that zproperty
[27-Jul-2009 15:10:41] <rmatte> how would I go about aging an event from zendmd?
[27-Jul-2009 15:10:47] <rmatte> or from some sort of REST call?
[27-Jul-2009 15:40:47] <rmatte> mrayzenoss: do you know of a way to move events to history from outside of Zenoss?
[27-Jul-2009 15:41:04] <rmatte> I found the SQL command: DELETE FROM status  where evid in ('7f0000013693398bff5b81c');
[27-Jul-2009 15:41:14] <rmatte> but that seems like it would outright delete the event instead of moving it
[27-Jul-2009 15:41:48] <mrayzenoss> I know the desktop applet uses xml-rpc to do it
[27-Jul-2009 15:41:53] <mrayzenoss> but you can do it with SQL as well
[27-Jul-2009 15:42:08] <rmatte> ok
[27-Jul-2009 15:43:30] <rmatte> got an example of the xml-rpc?
[27-Jul-2009 15:45:04] <rmatte> In the dev guide I see the xml-rpc code to send an event
[27-Jul-2009 15:45:09] <rmatte> but not to move an event to history
[27-Jul-2009 15:45:37] <mrayzenoss> http://zapplet.svn.sourceforge.net/viewvc/zapplet/trunk/zapplet/event.py?view=markup&pathrev=14
[27-Jul-2009 15:46:10] <rmatte>   139 def historify(self, async=None):
[27-Jul-2009 15:46:10] <rmatte>   140 xmlrpc = self.__class__._get_xmlrpc_proxy(self.server, "/zport/dmd/ZenEventManager")
[27-Jul-2009 15:46:10] <rmatte>   141 return xmlrpc.manage_deleteEvents((self.evid,))
[27-Jul-2009 15:46:11] <rmatte> ?
[27-Jul-2009 15:46:43] <mrayzenoss> still figuring it out myself
[27-Jul-2009 15:47:02] <rmatte> that seems to be the function to do it, hmmm
[27-Jul-2009 15:47:54] <rmatte> yeh, I think that's it
[27-Jul-2009 15:48:06] <rmatte> since there's also a function in there for acknowledging events which is similar
[27-Jul-2009 15:48:54] <rmatte> thanks
[27-Jul-2009 15:50:45] <Disconnect> ugh. just did an update (2.3.3 to 2.4.2) and it seems to have lost most/all of the snmp communities. getting hundreds of pages but if i hit a device and 'reset community' 'model' it resets and works.
[27-Jul-2009 15:51:31] <rmatte> Disconnect: you monitor mostly Linux servers?
[27-Jul-2009 15:52:33] <Disconnect> yah
[27-Jul-2009 15:53:08] <rmatte> Disconnect: check their zProperties to see if there's a local zSnmpCommunity value applied to them
[27-Jul-2009 15:53:20] <rmatte> If there is, then you've just confirmed what I've suspected to be a bug
[27-Jul-2009 15:54:25] <Disconnect> its got a blank one, with snmpcommunities populated (correctly at first glance)
[27-Jul-2009 15:54:39] <rmatte> but there's no local property?
[27-Jul-2009 15:54:45] <rmatte> like if you scroll right to the bottom?
[27-Jul-2009 15:55:01] <rmatte> do you see it listed in the local properties dropdown?
[27-Jul-2009 15:55:07] <Disconnect> zero local properties
[27-Jul-2009 15:55:12] <rmatte> ok
[27-Jul-2009 15:55:21] <rmatte> then it's not what I was thinking then
[27-Jul-2009 15:55:28] <Disconnect> (unrelated: that dropdown prolly should have a minimum size of some sort)
[27-Jul-2009 15:56:09] <al_x> do I want to use command monitoring if all I want to know is if SSH is responding?
[27-Jul-2009 15:56:57] <rmatte> al_x: You want IP Services monitoring
[27-Jul-2009 15:57:22] <rmatte> add "ssh" to the IP Services section on the OS tab
[27-Jul-2009 15:58:00] <al_x> per host or to the device template
[27-Jul-2009 15:58:26] <rmatte> Dropdown -> Add IP Service, start typing ssh in the input field, if you don't see a dropdown appear with services to select, click Cancel, reopen it again, start typing ssh again, you should see the dropdown, select ssh
[27-Jul-2009 15:58:35] <rmatte> do you need it on multiple hosts?
[27-Jul-2009 15:59:00] <al_x> ideally yea
[27-Jul-2009 15:59:06] <al_x> but i only have a few hosts at the moment
[27-Jul-2009 15:59:37] <rmatte> If so, click on Services on the left, then IpService, then Privileged, then search for ssh, click on it, go to the edit tab, change Monitor from False to True
[27-Jul-2009 15:59:49] <rmatte> then remodel any devices and it'll automatically monitor it provided it detects it
[27-Jul-2009 16:00:11] <al_x> do I have to add the send string and expect regex myself?
[27-Jul-2009 16:00:18] <rmatte> it's already set
[27-Jul-2009 16:00:36] <rmatte> Zenoss has tons of pre-made regex values for IP Services
[27-Jul-2009 16:00:41] <Disconnect> rmatte: ok i just tried a different host -- turns out hte one I tested on first was actually having snmp issues (sigh). but i found one that had the community local property
[27-Jul-2009 16:01:00] <rmatte> Disconnect: you'll probably notice a lot of them having the local property
[27-Jul-2009 16:01:03] <Disconnect> (it was correct, however)
[27-Jul-2009 16:01:14] <rmatte> Disconnect: Ah, strange then
[27-Jul-2009 16:01:17] <al_x> ok cool
[27-Jul-2009 16:01:26] <al_x> i remodelled but its not in the ip services list
[27-Jul-2009 16:01:32] <rmatte> Disconnect: try deleting the local property anyways, see if it makes a difference
[27-Jul-2009 16:01:49] <rmatte> al_x: you need to push changes to the collector so that it knows it should be monitoring it
[27-Jul-2009 16:02:01] <rmatte> just click on Devices on the left
[27-Jul-2009 16:02:06] <rmatte> then do Manage -> Push changes
[27-Jul-2009 16:02:09] <rmatte> then try again
[27-Jul-2009 16:02:24] <rmatte> I'm afk for a few
[27-Jul-2009 16:02:28] <Disconnect> still times out during the model
[27-Jul-2009 16:04:46] <Disconnect> oh for.. ok i give up for now. boss decided to be exciting and move eth0 around on this particular example. (so snmpd isn't listening on the right interface..)
[27-Jul-2009 16:08:41] <al_x> hmm i tried that it's still disappearing
[27-Jul-2009 17:15:43] <Diddi> Hi! how do I tell a device that eth0 speed should be at 100Mb/s instead of 10Mb/s? I see in <hostname>/os/eth0 that zenoss detects 10Mb/s interface, but I can't edit it
[27-Jul-2009 17:17:27] <mrayzenoss> that would be what the OS is reporting
[27-Jul-2009 17:18:04] <Diddi> interesting..
[27-Jul-2009 17:19:13] <mrayzenoss> or at least, the SNMP or SSH daemon
[27-Jul-2009 17:19:20] <mrayzenoss> what kind of device is it?
[27-Jul-2009 17:19:55] <Diddi> gotta be the snmpd... I found that only debian boxes reports 10Mb/s. ArchLinux boxes reports 100Mb/s
[27-Jul-2009 17:20:54] <mrayzenoss> hmm.. my Debian box is being monitored via SSH
[27-Jul-2009 17:21:44] <mrayzenoss> moving it over to SNMP to see what it models
[27-Jul-2009 17:25:00] <Diddi> I'm a bit new to snmpd on linux... debian and Arch tend to report different things... although they are mostly the same... e.g. the speed of NIC, running processes... Arch doesn't have the hrSWRunPath MIB while debian does.. making zenoss think that processes aren't up and running on my Arch box
[27-Jul-2009 17:25:33] <Diddi> not sure if it's a configuration option or actual build-differencies
[27-Jul-2009 19:27:42] <johns55> hey guys I'm coming from zenoss version 2.1.x and now on 2.4 for one of my monitoring servers, may I ask what you guys did with the History tab?
[27-Jul-2009 19:28:56] <johns55> ok nm I see it now
[27-Jul-2009 19:28:58] <johns55> its under events
[27-Jul-2009 20:10:14] <rmatte> yup
[27-Jul-2009 20:10:36] <rmatte> If you go to the event console there's a link to it in the top right
[27-Jul-2009 20:10:57] <rmatte> ...annnnd he already left
[27-Jul-2009 20:19:38] <Diddi> (:
[28-Jul-2009 00:00:46] [disconnected at Tue Jul 28 00:00:46 2009]
[28-Jul-2009 00:00:46] [connected at Tue Jul 28 00:00:46 2009]
[28-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[28-Jul-2009 02:28:18] <draytm01> hi all
[28-Jul-2009 02:28:30] <draytm01> i've managed to get my zopedb into a mess
[28-Jul-2009 02:30:14] <draytm01> >>> d = find("lb.dev.local")
[28-Jul-2009 02:30:14] <draytm01> 2009-07-28 08:29:04 WARNING ZODB Could not import class 'Zxtm' from module 'ZenPacks.dev.Zxtm'
[28-Jul-2009 02:30:14] <draytm01> >>> d
[28-Jul-2009 02:30:14] <draytm01> <Zxtm at /zport/dmd/Devices/Server/Linux/ZXTM/devices/broken>
[28-Jul-2009 02:30:46] <draytm01> i can't find any reference to 'broken' on the forums
[28-Jul-2009 02:33:52] <draytm01> so, erm, any ideas? I've tried reindex() (gives 'AttributeError: index_object'), as well as http://www.zenoss.com/Members/cluther/fix-a-broken-devicesearch-or-componentsearch-catalog/
[28-Jul-2009 02:35:20] <draytm01> it looks like something's in a real state so can i somehow zap the whole zope db? this is a dev instance so it doesn't matter (though if i could delete this device and fix it all it'd be better)
[28-Jul-2009 03:05:20] <Weetos> hey there
[28-Jul-2009 03:05:21] <Weetos> I'm having trouble to fetch the processes of a debian server (hrsystemprocesses via snmp)
[28-Jul-2009 03:05:42] <Weetos> I can't put my finger on it as it works fine with some other debian server (same OS, same template)
[28-Jul-2009 03:06:05] <Weetos> when I use snmpwalk, it gives "No more variables left in this MIB View"
[28-Jul-2009 03:06:35] <Weetos> I'm using the same snmpd config
[28-Jul-2009 03:06:57] <Weetos> Has anyone experienced the same issue ?
[28-Jul-2009 03:55:13] <Weetos> Nervermind
[28-Jul-2009 03:56:44] <Weetos> Fixed it by modifying snmpd.conf so zenoss can access all instead of just the system subtree
[28-Jul-2009 04:00:48] <rhett> hm
[28-Jul-2009 07:32:02] <draytm01> quiet here, huh
[28-Jul-2009 07:32:27] <draytm01> anyone able to help debug my extending device problem
[28-Jul-2009 07:33:22] <draytm01> (not *that* extending device -- i get plenty of offers of chemical assistance there)
[28-Jul-2009 08:30:14] <ckrough> anyone doing authenticated REST calls through an https proxy?
[28-Jul-2009 09:13:38] <ckrough> rmatte: congrats zen master
[28-Jul-2009 09:33:29] <rmatte> ckrough: thanks
[28-Jul-2009 09:35:18] <Dieterbe> Hi, i want to make an event transform to check whether a network interface on another host is down, and if so, supress the event. i already do something like: server = device.findDevice("<ip goes here>"). but now, how do I check whether the specific interface is up or down?
[28-Jul-2009 09:35:21] <rmatte> mrayzenoss: noticed you have this list too: http://www.zenoss.com/community/zen-of-zenoss/the-zenoss-masters/?searchterm=documentation
[28-Jul-2009 09:35:56] <rmatte> Dieterbe: hmmm, good question
[28-Jul-2009 09:36:09] <mrayzenoss> rmatte: I better update that!
[28-Jul-2009 09:36:18] <rmatte> hehe
[28-Jul-2009 09:36:43] <Dieterbe> hmm shit i just noticed findDevice () needs some kind of "id", and not the api
[28-Jul-2009 09:37:04] <rmatte> you can just do find('<ip>')
[28-Jul-2009 09:37:17] <Dieterbe> aren't these api's documented anywhere? like all functions you can use to find devices, how you can check the status of all things that are monitored on it, etc
[28-Jul-2009 09:37:33] <rmatte> Dieterbe: I have to go afk for about 5 mins, but I'll have some stuff for you when I get back
[28-Jul-2009 09:37:59] <Dieterbe> ok thanks
[28-Jul-2009 09:39:55] <dollarbang2> anyone know about a plugin to zenoss called netflow?
[28-Jul-2009 09:40:53] <mrayzenoss> dollarbang2: there's this for Enterprise: http://www.zenoss.com/product/plixer-scrutinizer
[28-Jul-2009 09:41:18] <mrayzenoss> but you could probably put something together with NFSen or some other open source netflow product
[28-Jul-2009 09:41:34] <mrayzenoss> using zlinks and a bit of elbow grease
[28-Jul-2009 09:41:57] <dollarbang2> mrayzenoss: thank you, will look for it. Plenty of grease available.
[28-Jul-2009 09:42:53] <mrayzenoss> dollarbang2: feel free to document what you do and I'll send people your way
[28-Jul-2009 09:44:52] <dollarbang2> mrayzenoss:documentation? I'm a flyboy, if it works, it's production...Knowing this place, there will be a Wiki developed.
[28-Jul-2009 09:45:07] <rmatte> Dieterbe: ok back, I just need to dig through some stuff to find what you need...
[28-Jul-2009 09:45:25] <Dieterbe> also , "Intelligently suppress "downstream" tests and alerts based on layer 3 network dependencies" does not work for me. my monitoring box accesses the target hosts through a vpn. if the vpn interface on the monitoring host goes down, i get a lot of events/alerts about all the unreachable hosts/services
[28-Jul-2009 09:45:31] <Dieterbe> rmatte: thanks man
[28-Jul-2009 09:45:55] <mrayzenoss> dollarbang2: the Zenoss/MindTouch mashup might be useful as an example and a wiki when you need one: https://sourceforge.net/projects/zenossdeki/
[28-Jul-2009 09:46:33] <rmatte> Dieterbe: I'm fairly sure that it'll be something similar to the code in http://www.zenoss.com/community/wiki/tips-and-tricks/ZendmdPlayingWithInterfaces/
[28-Jul-2009 09:47:27] <rmatte> Dieterbe: just need to figure out how to specify the interface, the other issue is that Zenoss only updates operational status of ports on remodelling, so for active monitoring you need to use a threshold...
[28-Jul-2009 09:47:39] <rmatte> hmmm
[28-Jul-2009 09:48:47] <Dieterbe> rmatte: what? so if you remodel every 24hour, and the interface goes down in between remodelings. are you saying zenoss won't know until it's remodeled again?
[28-Jul-2009 09:49:33] <rmatte> Dieterbe: it'll know based on ping monitoring, but that's not operational status. To monitor true operational status via SNMP you need: http://www.zenoss.com/Members/cluther/polling-interface-status/
[28-Jul-2009 09:50:35] <rmatte> Dieterbe: I believe there has already been a performance enhancement request submitted to build this type of functionality right in to Zenoss, no idea when that's going to actually be implemented though
[28-Jul-2009 09:51:02] <rmatte> Dieterbe: the threshold method works very well, but it makes it more difficult to do what you want to do
[28-Jul-2009 09:54:36] <rmatte> Dieterbe: actually, I'm not even entirely sure that what you want to do is even possible at the moment, since you need to be able to assess the status of an interface via something like: d = find('192.168.0.1') ... d.os.interface() (then you need the function to check the status, which I'm not sure of.)
[28-Jul-2009 09:54:55] <rmatte> But that status doesn't get properly updated on the fly
[28-Jul-2009 09:55:12] <rmatte> Maybe Matt has something he can dig up?
[28-Jul-2009 09:59:06] <rmatte> you could possibly do it in reverse... like, implement the threshold operational status monitoring, then add something to the transform for the threshold where if device = whatever and if component = <port name> then clear any events from whichever device, then put a delay of a minute or two on notifications. That might work.
[28-Jul-2009 09:59:43] <rmatte> That's really the only way that I can think of doing it, and the code would probably be simpler
[28-Jul-2009 09:59:49] <mrayzenoss> Dieterbe: if you want to do event suppression on one device based on the status of another device, take a look at this: http://blog.zenoss.com/2009/05/27/tip-of-the-month-event-suppression/
[28-Jul-2009 10:00:05] <rmatte> he wants to do it based on the status of a port though
[28-Jul-2009 10:01:00] <mrayzenoss> well, if you modify the second example to check the status of a port on the dc1_router as opposed to the getPingStatus, that should work
[28-Jul-2009 10:01:20] <rmatte> yeh, but the problem is that the port status doesn't get properly updated unless modeled
[28-Jul-2009 10:01:31] <rmatte> hence the need for the threshold to monitor port operational status
[28-Jul-2009 10:02:28] <rmatte> so he'd basically have to do it backwards, apply something to the threshold transform that kicks off a function to clear tickets for a device when it sees a specific port on a specific device as being down
[28-Jul-2009 10:02:45] <rmatte> and then put a delay for notification that's a bit longer than his polling cycle, not sure how well it'd work
[28-Jul-2009 10:02:52] <rmatte> but it's really the only solution I can think of
[28-Jul-2009 10:03:07] <Dieterbe> mrayzenoss: but where do i find all these functions? i couldn't find an API for this. how do i check a certain interface or service check?
[28-Jul-2009 10:03:34] <rmatte> There are APIs listed in the dev guide, but I've noticed that it's missing quite a few
[28-Jul-2009 10:04:09] <rmatte> aha
[28-Jul-2009 10:04:10] <rmatte> http://www.zenoss.com/community/docs/zenoss-api-docs/2.2/
[28-Jul-2009 10:04:14] <rmatte> that might help
[28-Jul-2009 10:04:20] <mrayzenoss> http://www.zenoss.com/community/docs/zenoss-api-docs/2.4.0/
[28-Jul-2009 10:04:24] <mrayzenoss> 2.4
[28-Jul-2009 10:04:54] <rmatte> ah, I clicked the wrong link
[28-Jul-2009 10:06:17] <rmatte> hmmm, even that doesn't seem to have quite what you're looking for
[28-Jul-2009 10:06:27] <Dieterbe> indeed
[28-Jul-2009 10:07:02] <Dieterbe> eg ZenEvents.Event which i assume is the one to look at for the 'evt' object , it doesn't really list what i'm looking for (it doesn't even list the things i already know of)
[28-Jul-2009 10:08:42] <rmatte> what are you looking for for evt?
[28-Jul-2009 10:09:10] <rmatte> evt is pretty easy, it's basically anything you can see on the details tab
[28-Jul-2009 10:09:23] <rmatte> details window, rather, for an event
[28-Jul-2009 10:10:09] <rmatte> evt.summary, evt.eventClassKey, evt.evid, and so on
[28-Jul-2009 10:13:38] <Dieterbe> ok but my point was just that i cannot find what i'm looking for but let me reread everything you guys said
[28-Jul-2009 10:13:48] <rmatte> k
[28-Jul-2009 10:13:58] <rmatte> and yeh, I agree that the documentation is lacking in some areas
[28-Jul-2009 10:18:14] <Dieterbe> and the thing is, i've looked at the administrator guide, extended monitoring guide etc. but the stuff you guys are talking about seems like a lot "deeper into it". there seems to be a gap between zenoss newbies and people who know it well
[28-Jul-2009 10:18:50] <rmatte> Most of the stuff I've personally learned has been from various Wiki postings that I've dug up, and from chatting with people in here.
[28-Jul-2009 10:19:31] <rmatte> Have you read through the developer's guide at all?
[28-Jul-2009 10:20:41] <Dieterbe> not really. i thought the developers guide was meant for people who want to work on zenoss' code
[28-Jul-2009 10:20:57] <rmatte> actually, the dev guide is more geared towards database manipulation, adding, removing, etc...
[28-Jul-2009 10:21:16] <mrayzenoss> yeah, it's a bit of a hacking things together with Zenoss guide
[28-Jul-2009 10:21:31] <rmatte> mrayzenoss: is there some piece of documentation that explains common zendmd and transform functions?
[28-Jul-2009 10:21:53] <rmatte> like, what if I wanted to determine the status of a specific interface via zendmd, is there a reference document that would allow me to figure that out?
[28-Jul-2009 10:22:04] <mrayzenoss> that would be nice
[28-Jul-2009 10:22:12] <rmatte> If not, it might be worthwhile to have someone like Chet whip something up, explaining the more common functions?
[28-Jul-2009 10:22:15] <mrayzenoss> http://www.zenoss.com/Members/netdata/zendmd-tips-and-tricks/
[28-Jul-2009 10:22:18] <rmatte> or anyone who knows them well for that matter
[28-Jul-2009 10:22:46] <rmatte> mrayzenoss: that really only shows 2 tricks hehe
[28-Jul-2009 10:22:47] <mrayzenoss> I think when we launch the new Community site we'll do a zendmd page similar to the Transforms one
[28-Jul-2009 10:22:59] * ckrough wants this
[28-Jul-2009 10:23:02] <mrayzenoss> http://www.zenoss.com/Members/shaskell/useful-zendmd-scripts/
[28-Jul-2009 10:23:15] <mrayzenoss> http://www.zenoss.com/Members/netdata/use-a-zendmd-script-as-standalone-program/
[28-Jul-2009 10:23:25] <mrayzenoss> http://www.zenoss.com/community/docs/howtos/mass-adding-locations-using-zendmd
[28-Jul-2009 10:23:35] <mrayzenoss> they're scattered throughout the wiki (which sucks)
[28-Jul-2009 10:23:49] <rmatte> the thing is, the functions that I use in everyday transforms are not documented anywhere it seems in any kind of standard format. they are just things that I've picked up along the way
[28-Jul-2009 10:24:28] <rmatte> there needs to be something like "this is the os.interfaces() function to fetch interface objects, here is all the stuff you can do with that function, here are the related sub-functions
[28-Jul-2009 10:25:06] <mrayzenoss> ckrough: BTW, Chet and Kells both said they use XML-RPC almost exclusively
[28-Jul-2009 10:25:52] <mrayzenoss> Kells sent me a bunch of zendmd scripts I need to post somewhere
[28-Jul-2009 10:26:36] <rmatte> Here's a zendmd script to disable monitoring on all interfaces on all devices:
[28-Jul-2009 10:26:37] <rmatte> >>> for d in dmd.Devices.getSubDevices():
[28-Jul-2009 10:26:37] <rmatte> ...     for interface in d.os.interfaces():
[28-Jul-2009 10:26:37] <rmatte> ...         interface.monitor = False
[28-Jul-2009 10:27:08] <DanB> hey guys, I am pretty new with zenoss and need to develop something around it, so please bear with me; I need to collect some data over snmp and display it as some sort of statistics in a custom portlet on the dashboard. Can u tell me what's the best way to get the data, considering is not always performance numbers (can be chars to - eg: your top 3 friends names). I should mention that I already have performance graphs running fine, but the statistics is someth
[28-Jul-2009 10:27:16] <rmatte> actually I just found the property that I was looking for
[28-Jul-2009 10:27:26] <rmatte> interface.operStatus
[28-Jul-2009 10:28:33] <mrayzenoss> DanB: there are 3 ZenPacks that may be useful to look at
[28-Jul-2009 10:28:41] <mrayzenoss> Show Graph Portlet http://www.zenoss.com/community/projects/zenpacks/graph-portlet
[28-Jul-2009 10:28:43] <ckrough> mrayzenoss: thanks
[28-Jul-2009 10:28:51] <mrayzenoss> Site Window Portlet: http://www.zenoss.com/community/projects/zenpacks/sitewindow
[28-Jul-2009 10:29:03] <mrayzenoss> Reports Porlet: http://www.zenoss.com/community/projects/zenpacks/reports
[28-Jul-2009 10:29:26] <mrayzenoss> DanB: if you're doing graphs, the graph portlet is supposed to put a perf graph on your dashboard
[28-Jul-2009 10:29:44] <DanB> mrayzenoss: perfect, a good start then
[28-Jul-2009 10:30:10] <DanB> mrayzenoss: is there a way to retrieve the data directly from my snmp daemon, without storing it into zenoss?
[28-Jul-2009 10:30:21] <DanB> I don't really need to have graphs on this part
[28-Jul-2009 10:30:25] <DanB> just tables
[28-Jul-2009 10:30:59] <DanB> or otherwise where should I store it considering it can be chars?
[28-Jul-2009 10:31:02] <mrayzenoss> DanB: You can embed just about anything in the portlet, so you could make Python calls straight to the snmp daemon I guess
[28-Jul-2009 10:32:00] <mrayzenoss> I can't think of an example of that right now, but there's probably a better way to do it
[28-Jul-2009 10:32:52] <mrayzenoss> you could add a command data source that writes out your table and the snmp values to a file, then include it with the Site Window Portlet
[28-Jul-2009 10:32:53] <DanB> I was thinking about a dirty way of doing it: encode the data into an event and decode it in a portlet
[28-Jul-2009 10:33:09] <mrayzenoss> yeah, both are dirty solutions
[28-Jul-2009 10:33:20] <mrayzenoss> I gotta go get ready for a demo, be back later
[28-Jul-2009 10:33:21] <DanB> is there some documentation about writing portlets around?
[28-Jul-2009 10:33:27] <mrayzenoss> Dev guide has a bit
[28-Jul-2009 10:33:29] <rmatte> Dieterbe: here's a rough idea of what you're looking for (not 100% sure it'd work as a transform since it's a zendmd script): http://pastebin.com/m3341edef
[28-Jul-2009 10:33:36] <mrayzenoss> the source is available for all those portlets
[28-Jul-2009 10:33:45] <DanB> mrayzenoss: thanks for your time, good luck with the demo
[28-Jul-2009 10:34:00] <mrayzenoss> DanB: trac-zenpacks/browser/zenpacks
[28-Jul-2009 10:34:23] <rmatte> Dieterbe: it basically scans through the interfaces on the device with ip 192.168.0.1 looking for one with interface name of FastEthernet0/1, when it finds it it checks if it's operationally down, if it is it does whatever specified on the next line
[28-Jul-2009 10:34:24] * mrayzenoss away
[28-Jul-2009 10:34:43] <Dieterbe> rmatte: thanks man
[28-Jul-2009 10:36:04] <rmatte> Dieterbe: no problem, but I'm still concerned that the interface operational status won't reflect the actual up to date status of the port, though I just had an interesting idea... I might be able to adjust the transform for that threshold monitoring solution so that it actually updates the operational status in Zenoss to reflect the status of the port when it sees it go down or up
[28-Jul-2009 10:36:15] <rmatte> Dieterbe: I'll have to work on it and see if I can pull it off
[28-Jul-2009 10:40:12] <Dieterbe> that brings me to something else. how can you actually test your transforms? i know you can submit dummy events, but how can you easily debug your code? can you do logging? what happens if you call functions that don't exist? where do you see that?
[28-Jul-2009 10:40:43] <rmatte> Well, it's just python code so you can test it in zendmd
[28-Jul-2009 10:41:10] <rmatte> you need to set the variables yourself though, instead of evt.summary I set evtsummary
[28-Jul-2009 10:41:15] <rmatte> same with any other variables
[28-Jul-2009 10:41:58] <rmatte> beyond that, you really just need to wait for an event to come in to be 100% sure that it works, also, zendmd will highlight your transform code in red if there are ay syntax errors
[28-Jul-2009 10:42:09] <rmatte> erm, Zenoss will, rather
[28-Jul-2009 10:42:11] <rmatte> in the UI
[28-Jul-2009 10:43:14] <Dieterbe> does that include referencing non-existing functions/attributes?
[28-Jul-2009 10:44:05] <rmatte> yes
[28-Jul-2009 10:44:15] <Dieterbe> nice
[28-Jul-2009 10:44:38] <rmatte> as I recall anyways, it's usually pretty good about detecting any sorts of problems
[28-Jul-2009 10:44:50] <rmatte> there's still a margin for error obviously, but it's fairly slim
[28-Jul-2009 10:47:17] <rmatte> I wonder if something like this would work: http://pastebin.com/m145ab746
[28-Jul-2009 10:47:36] <rmatte> basically tracks down the port and sets the operational status to inactive when it detects that the port went down
[28-Jul-2009 10:48:37] <rmatte> then I'd just need to set it up to put it back to active when it sees it come back up
[28-Jul-2009 10:49:29] <rmatte> I wish Chet were around so I could run the code by him
[28-Jul-2009 10:53:25] <Dieterbe> rmatte: if p.match(interface.getInterfaceName()): --> probably needs to be if interface.getInterfaceName() == evt.component ? otherwise you could have problems if you have many ifaces like eth1 and eth12
[28-Jul-2009 10:54:54] <rmatte> good point
[28-Jul-2009 10:55:14] <rmatte> actually, I have a better function I can use, just need to dig it up
[28-Jul-2009 10:57:40] <rmatte> if re.match(evt.component, interface.getInterfaceName()):
[28-Jul-2009 10:57:43] <rmatte> that should work
[28-Jul-2009 11:01:22] <rmatte> http://pastebin.com/m3b35c197
[28-Jul-2009 11:01:28] <rmatte> oops
[28-Jul-2009 11:01:40] <rmatte> need to change the operstatus setting for when it comes back up, forgot to change the number
[28-Jul-2009 11:01:48] <rmatte> otherwise, that should technically work, I'd have to test it
[28-Jul-2009 11:02:05] <rmatte> hmmm, is it operstatus 1 or 0?
[28-Jul-2009 11:02:10] * rmatte scratches his head
[28-Jul-2009 11:02:59] <Dieterbe> isn't that still affected by the problem i just described?
[28-Jul-2009 11:03:15] <rmatte> no, it's re.match
[28-Jul-2009 11:03:23] <rmatte> so it only accepts an exact match to the name
[28-Jul-2009 11:03:32] <rmatte> or wait
[28-Jul-2009 11:03:39] <rmatte> maybe I'm looking for re.string or something
[28-Jul-2009 11:03:49] <rmatte> been a while since I've used these functions
[28-Jul-2009 11:04:20] <rmatte> oh, at the start of the string
[28-Jul-2009 11:04:22] <rmatte> hmmm
[28-Jul-2009 11:04:32] <rmatte> yeh, you're right
[28-Jul-2009 11:05:04] <rmatte> and I was thinking of re.search as an alternative, which is even worse lol
[28-Jul-2009 11:05:17] <rmatte> guess if interface.getInterfaceName() == evt.component it is
[28-Jul-2009 11:06:06] <rmatte> now I just need to know what operStatus to set it to when it comes back up
[28-Jul-2009 11:06:31] <rmatte> I think it's 1, but I'm not sure
[28-Jul-2009 11:06:36] <Dieterbe> i have to go. i hope to talk to you later
[28-Jul-2009 11:06:44] <Dieterbe> you've been very helpful
[28-Jul-2009 11:06:54] <rmatte> I'll be around, take care
[28-Jul-2009 11:12:57] <rmatte> yeh, it is 1, ok...
[28-Jul-2009 11:13:20] <rmatte> voila: http://pastebin.com/m4175bfdd
[28-Jul-2009 11:15:49] <rmatte> mrayzenoss: I'm going to test that code, if it works properly, think you could include it in http://www.zenoss.com/Members/cluther/polling-interface-status/?
[28-Jul-2009 11:20:28] <rmatte> If it does work I might just package it up as a ZenPack.  Interface Polling ZenPack.
[28-Jul-2009 11:22:37] <rmatte> so far so good, it saved in to my lab box without complaining about syntax
[28-Jul-2009 11:25:46] <ckrough> what does it do
[28-Jul-2009 11:26:24] <ckrough> err, what triggers the 'threshold' warning
[28-Jul-2009 11:26:44] <rmatte> Read this: http://www.zenoss.com/Members/cluther/polling-interface-status/
[28-Jul-2009 11:26:48] <rmatte> it explains the setup
[28-Jul-2009 11:27:05] <rmatte> it's basically for active operational status polling for interfaces via SNMP (since Zenoss doesn't do it on it's own)
[28-Jul-2009 11:27:14] <ckrough> gotcha
[28-Jul-2009 11:27:29] <rmatte> I'm just modifying the transform so that it actually updates the interface status on the OS tab when an interface goes down or comes back up
[28-Jul-2009 11:27:38] <ckrough> nice nice
[28-Jul-2009 11:27:39] <rmatte> that way the OS tab is actually useful for checking port status
[28-Jul-2009 11:28:11] <ckrough> rather than waiting for a model cycle. makes sense. I get this info via snmptraps, but dont update the int statuses
[28-Jul-2009 11:28:41] <rmatte> yeh, traps are alright, but I prefer to poll since that way you're only getting alerts for ports that you actually want to monitor
[28-Jul-2009 11:28:49] <rmatte> traps work fine if you want to monitor everything
[28-Jul-2009 11:29:01] <rmatte> but on a 48 port switch I don't necessarily care about the 36 or so user ports on it
[28-Jul-2009 11:29:12] <rmatte> don't want to get an alert every time someone turns their computer off
[28-Jul-2009 11:29:14] <rmatte>
[28-Jul-2009 11:29:56] <rmatte> If this works it'll update status on ports, but only on ports you've specifically set to monitor
[28-Jul-2009 11:30:14] <rmatte> Less taxing
[28-Jul-2009 11:30:15] <ckrough> yeah, agreed. actually I lied, not using traps, I use a syslog-ng filter to forward LINK-UPDOWNS for interesting devices (all ports) to zensyslog for alerting
[28-Jul-2009 11:30:24] <rmatte> ahhhh
[28-Jul-2009 11:30:49] <rmatte> yeh, we just started implementing syslog-ng between systems and Zenoss
[28-Jul-2009 11:31:04] <rmatte> offers some nice flexibility to filter or even redirect syslogs
[28-Jul-2009 11:31:26] <ckrough> yeah. roll Splunk into the mix and it's a great troubleshooting tool
[28-Jul-2009 11:31:36] <rmatte> yup
[28-Jul-2009 11:33:13] <rmatte> now I just need to find a device that I can actually take a port down on, apparently we don't have anything hooked up in the back right now...
[28-Jul-2009 11:33:21] <rmatte> afk for a few, time to go raid the lab
[28-Jul-2009 11:35:34] <rmatte> hmmm, nothing good in there
[28-Jul-2009 11:35:55] <rmatte> I might have to do this with a virtual system
[28-Jul-2009 12:13:14] <rmatte> ok, so I just tested this: http://pastebin.com/f26f3112f and it doesn't work for some reason, need to figure out why...
[28-Jul-2009 12:48:09] <konfoo> is anyone using the xmppbot with googletalk?
[28-Jul-2009 12:57:32] <mrayzenoss> chudler is the author...
[28-Jul-2009 13:40:37] <rmatte> mrayzenoss: when a transform executes can it be assumed that the imports here: http://www.zenoss.com/Members/netdata/use-a-zendmd-script-as-standalone-program/ are already done?
[28-Jul-2009 13:41:10] <rmatte> mrayzenoss: I'm trying to incorporate some zendmd code in to a transform
[28-Jul-2009 13:44:23] <mrayzenoss> that's probably a safe assumption, I'd look at zendmd to see what it does
[28-Jul-2009 13:44:28] <mrayzenoss> http://dev.zenoss.org/trac/browser/trunk/bin/zendmd
[28-Jul-2009 13:44:59] <mrayzenoss> http://dev.zenoss.org/trac/browser/trunk/Products/ZenModel/zendmd.py
[28-Jul-2009 13:45:02] <rmatte> I'm trying to figure out why this code won't work: http://pastebin.com/f26f3112f
[28-Jul-2009 13:45:31] <rmatte> when I put it as a transform it's like it doesn't even run at all
[28-Jul-2009 13:45:43] <rmatte> doesn't transform the event summary or anything
[28-Jul-2009 13:48:04] <rmatte> just moved the setting of the d variable in to the if statement: http://pastebin.com/f24a1c11d
[28-Jul-2009 13:48:09] <rmatte> I'll see if that fixes it...
[28-Jul-2009 13:51:20] <mrayzenoss> I've got a long meeting to head to, I'll check back afterwards
[28-Jul-2009 14:03:06] <rmatte> well, now it's running the transform, but the zendmd part isn't working properly, the operational status of the port hasn't changed, hmmm
[28-Jul-2009 14:03:31] <rmatte> I had to add the zendmd import statements to the top of the transform to get it to work properly
[28-Jul-2009 14:04:19] <rmatte> Ok, so now there's something wrong with this code...
[28-Jul-2009 14:04:20] <rmatte>         for interface in d.os.interface():
[28-Jul-2009 14:04:20] <rmatte>             if interface.getInterfaceName() == evt.component:
[28-Jul-2009 14:04:20] <rmatte>                 interface.operStatus = 2
[28-Jul-2009 14:05:49] <rocket> rmatte: where is evt.component getting set?
[28-Jul-2009 14:06:38] <rmatte> rocket: that's the part that's puzzling, it's an alert from a threshold and it just gets set on it's own
[28-Jul-2009 14:07:12] <rmatte> ok, here's the code I'm currently testing:
[28-Jul-2009 14:07:12] <rocket> rmatte: what is the problem you are experiencing?
[28-Jul-2009 14:07:13] <rmatte> http://pastebin.com/m1207fee
[28-Jul-2009 14:07:51] <rmatte> I'm trying to make the transform for this: http://www.zenoss.com/Members/cluther/polling-interface-status/ update the interface status on the OS tab when it detects the interface is up or down
[28-Jul-2009 14:08:44] <rocket> rmatte: ok I think I see what the problem is with your code
[28-Jul-2009 14:09:03] <rmatte> I thought you might since you're much better at my in python
[28-Jul-2009 14:09:11] <rocket> rmatte: though I havent looked at the code so I am going off of memorty
[28-Jul-2009 14:09:12] <rmatte> better than me*
[28-Jul-2009 14:09:20] <rmatte> k
[28-Jul-2009 14:09:40] <rocket> I believe evt.component is the grouping eg interface
[28-Jul-2009 14:09:57] <rocket> I dont think it give the individual name of the nic for example
[28-Jul-2009 14:10:02] <rocket> so they dont match up
[28-Jul-2009 14:10:24] <rmatte> evt.component ends up coming out like: Intel(R) PRO/1000 MT Desktop Adapter #2
[28-Jul-2009 14:10:33] <rocket> ok it does .. hrmm
[28-Jul-2009 14:10:38] <rmatte> (that's the interface I'm testing on)
[28-Jul-2009 14:10:50] <rocket> can you do a print for interface.getInterfaceName for me
[28-Jul-2009 14:10:59] <rocket> and let me know what that one comes out as?
[28-Jul-2009 14:11:02] <rmatte> sure, one second
[28-Jul-2009 14:12:57] <rmatte> hmmm, can't seem to do it with zendmd
[28-Jul-2009 14:13:24] <rmatte> >>> for interface in d.os.interface():
[28-Jul-2009 14:13:24] <rmatte> ...  print interface.getInterfaceName()
[28-Jul-2009 14:13:24] <rmatte> ...
[28-Jul-2009 14:13:24] <rmatte> Traceback (most recent call last):
[28-Jul-2009 14:13:24] <rmatte>   File "<console>", line 1, in ?
[28-Jul-2009 14:13:25] <rmatte> AttributeError: interface
[28-Jul-2009 14:13:36] <rmatte> Oh, and I did d = dmd.Devices.findDevice('NOTT-DDW-RMAT') before
[28-Jul-2009 14:13:52] <rmatte> I tried without "print" too
[28-Jul-2009 14:13:55] <rmatte> same thing
[28-Jul-2009 14:14:40] <rocket> hrmm ok
[28-Jul-2009 14:14:49] <rmatte> so is that not the right property then?
[28-Jul-2009 14:15:12] <rmatte> I got it from here: http://www.zenoss.com/community/wiki/tips-and-tricks/ZendmdPlayingWithInterfaces/
[28-Jul-2009 14:15:40] <rocket> missing an s
[28-Jul-2009 14:15:47] <rocket> d.os.interfaces()
[28-Jul-2009 14:15:53] <rocket> not d.os.interface()
[28-Jul-2009 14:15:54] <rmatte> oh, so I am
[28-Jul-2009 14:15:57] <rmatte> bah
[28-Jul-2009 14:17:31] <rmatte> That's probably what the problem was with my transform
[28-Jul-2009 14:17:32] <rmatte> >>> d = dmd.Devices.findDevice('NOTT-DDW-RMAT')
[28-Jul-2009 14:17:32] <rmatte> >>> for interface in d.os.interfaces():
[28-Jul-2009 14:17:32] <rmatte> ...  print interface.getInterfaceName()
[28-Jul-2009 14:17:32] <rmatte> ...
[28-Jul-2009 14:17:32] <rmatte> Intel(R) PRO/1000 MT Desktop Adapter
[28-Jul-2009 14:17:33] <rmatte> Intel(R) PRO/1000 MT Desktop Adapter #2
[28-Jul-2009 14:18:13] <rmatte> ooooh, it seems to have possibly worked
[28-Jul-2009 14:18:32] <rmatte> can't believe it was just a typo lol
[28-Jul-2009 14:18:46] <rocket> it usually is .. believe it or not
[28-Jul-2009 14:18:50] <rmatte> time to re-enable the interface and see if that part works too
[28-Jul-2009 14:20:33] <rmatte> thanks for the help, this transform is going to make Zenoss actually display proper up to date interface operational status for any ports being monitored
[28-Jul-2009 14:20:35] <rmatte> going to be nice
[28-Jul-2009 14:21:30] <rmatte> just need to wait for these events to clear to confirm that it's 100% working...
[28-Jul-2009 14:24:03] <rmatte> hmmm, taking a long time to clear
[28-Jul-2009 14:25:30] <rmatte> bah, now I'm wondering why no clear event is coming in now that the interface is back up...
[28-Jul-2009 14:30:33] <rmatte> just restarted Zenoss, let's try this again...
[28-Jul-2009 14:33:43] <rmatte> now no threshold events are coming in at all for operational status, what the heck
[28-Jul-2009 14:39:14] <rmatte> I completely removed the transform and still nothing
[28-Jul-2009 14:39:16] <rmatte> what the heck
[28-Jul-2009 14:41:15] <dollarbang> has anyone done a "ZenDeviceDump" under 2.3 in then a "ZenDeviceLoad" under 2.4?
[28-Jul-2009 14:48:56] <rmatte> dollarbang, can't say that I have, I've just done straight upgrades
[28-Jul-2009 15:11:57] <rmatte> well, this transform half works, it does mark the port as operationally down, but it also seems to somehow prevent further threshold alerts from coming in
[28-Jul-2009 15:12:23] <rmatte> including clear events
[28-Jul-2009 15:13:11] <rocket> rmatte: maybe because its operationally down? not sure but perhaps zenoss looks at that status as deems because it is set to that status it no longer sends events
[28-Jul-2009 15:13:43] <rocket> rmatte: are you setting the right status .. eg operdown might not be the correct one to set
[28-Jul-2009 15:13:58] <rmatte> it looks right
[28-Jul-2009 15:14:37] <rmatte> although you may be right, the port is technically administratively down right now
[28-Jul-2009 15:15:43] <rmatte> I don't get why that'd make a difference though, Zenoss should still try to poll the port
[28-Jul-2009 15:15:47] <rmatte> since it's set to be monitored
[28-Jul-2009 15:16:54] <rocket> there is goofy code trying to be smart and skip stuff when the port is intentionally down
[28-Jul-2009 15:17:08] <rmatte> hmmmm
[28-Jul-2009 15:17:33] <rocket> again thats from memory of me digging around so I cant remember exactly what its doing without looking again
[28-Jul-2009 15:17:49] <rmatte> that's kind of dumb though, if a port is down it should still poll it as long as it's set to monitor it
[28-Jul-2009 15:18:23] <rmatte> what if you remodel and the port shows as down after the remodel but you want to monitor it, it won't monitor?
[28-Jul-2009 15:18:26] <rmatte> seems kind of weird
[28-Jul-2009 15:18:32] <rocket> I would agree there .. except in the case of a cisco for example if the port is down administratively it doesnt make sense to monitor a port the admin has turned off from cisco
[28-Jul-2009 15:18:49] <rocket> rmatte: I would guess it wont monitor
[28-Jul-2009 15:18:51] <rmatte> yeh, but I'm not setting it admin down, I'm setting it operationally down
[28-Jul-2009 15:18:54] <rmatte> which is different
[28-Jul-2009 15:19:17] <rocket> rmatte: I hope you are correct and I am wrong
[28-Jul-2009 15:19:32] <rmatte> now the port is active on the device, so it should technically poll it, see that it's up, and send a clear event (resulting in the transform setting port status back to active)
[28-Jul-2009 15:20:14] <rmatte> if it doesn't then that's quite bad design
[28-Jul-2009 15:20:35] <rmatte> but you may be right, I'll set the port status back to active and see if that's the case
[28-Jul-2009 15:23:46] <rmatte> ok, I set it back to active, let's see if that's the case...
[28-Jul-2009 15:26:32] <rmatte> well, I received a threshold of operational status restored message, which should have been transformed but didn't
[28-Jul-2009 15:27:22] <rmatte> oh, because it came in with a capital T at the start for some reason
[28-Jul-2009 15:27:25] <rmatte> lol
[28-Jul-2009 15:27:34] <rmatte> bizarre
[28-Jul-2009 15:47:05] <rmatte> well, I just stumbled on a major issue
[28-Jul-2009 15:47:31] <rmatte> time to log a trac...
[28-Jul-2009 16:01:06] <rmatte> http://dev.zenoss.org/trac/ticket/5376
[28-Jul-2009 16:12:05] HKhan is now known as Hamzah
[28-Jul-2009 16:22:20] <rmatte> ?
[28-Jul-2009 16:27:46] <MACscr> how do I reset my admin password? No idea why its not currently working
[28-Jul-2009 16:28:05] <mrayzenoss> zenpassword
[28-Jul-2009 16:28:17] <MACscr> I have to say, zenoss is a bitch to use in general, but its at least a step up from configuring nagios
[28-Jul-2009 16:28:37] <mrayzenoss> MACscr: awww, where's the love? 
[28-Jul-2009 16:28:48] <mrayzenoss> it's getting better all the time
[28-Jul-2009 16:29:07] <MACscr> mrayzenoss: sorry, I just personally hate how a lot of the UI was designed and of course the use of zope, ewwwwww
[28-Jul-2009 16:29:48] <rmatte> MACscr: it just takes time, you'll get used to it, once you know where everything is it's quite quick to use
[28-Jul-2009 16:31:05] <MACscr> mrayzenoss: so what were you talking about when you said zenpassword
[28-Jul-2009 16:31:06] <mrayzenoss> I will agree that the UI is not intuitive
[28-Jul-2009 16:31:27] <rmatte> Yeh, the UI is getting a makeover anyways
[28-Jul-2009 16:32:19] <mrayzenoss> there's a command in $ZENHOME/bin called zenpass for resetting the password
[28-Jul-2009 16:32:30] <mrayzenoss> not zenpassword as I previously wrote
[28-Jul-2009 16:38:32] <mrayzenoss> a note to people who care, IE7 doesn't work with Medium security settings
[28-Jul-2009 16:40:44] <MACscr> I think my $ZENHOME path is ganked. How do I reset that again?
[28-Jul-2009 16:41:07] <mrayzenoss> you set that in the .bash_profile or .profile of the zenoss account
[28-Jul-2009 16:41:12] <MACscr> (note, I installed zenoss months ago, but really haven't had a chance to use it until now, so I have forgotten a lot)
[28-Jul-2009 16:41:30] <mrayzenoss> export ZENHOME=/usr/local/zenoss
[28-Jul-2009 16:41:41] <mrayzenoss> if you're root, just 'su - zenoss'
[28-Jul-2009 16:41:57] <mrayzenoss> the zenoss account is probably configured correctly depending on how you installed
[28-Jul-2009 16:44:57] <rmatte> actually, they are set by /usr/local/zenoss/scripts/setenv.sh
[28-Jul-2009 16:45:31] <rmatte> and my /home/zenoss/.bashrc file has ". /usr/local/zenoss/scripts/setenv.sh" at the bottom of it
[28-Jul-2009 16:46:12] <rmatte> but that's from stack installer
[28-Jul-2009 16:47:27] <rmatte> speaking of which, it was listed twice, probably from the last upgrade I did
[28-Jul-2009 16:47:55] <rmatte> I could write a better method of populating that line in to the file that actually checks if the line already exists, if you want
[28-Jul-2009 16:48:00] <rmatte> I assume it's done in bash
[28-Jul-2009 16:51:02] <mrayzenoss> hmm... I'm not sure where that comes from.
[28-Jul-2009 16:51:23] <mrayzenoss> BitNami provides the stack installers, they may own those scripts
[28-Jul-2009 16:51:33] <rmatte> ah
[28-Jul-2009 16:51:34] <mrayzenoss> I assume they're in our source tree somewhere
[28-Jul-2009 16:52:51] <rmatte> you should try to get listed on the BitNami site
[28-Jul-2009 16:52:57] <rmatte> they have a bunch of other projects listed
[28-Jul-2009 16:53:58] <mrayzenoss> yeah, I'm not sure how that side of it works. I mean, they provide those stacks free I believe
[28-Jul-2009 16:54:06] <mrayzenoss> we have to pay them to work on our stuff
[28-Jul-2009 16:54:22] <mrayzenoss> but I'll go ask
[28-Jul-2009 16:54:41] <rmatte> cool
[28-Jul-2009 16:54:51] <syntaxcollector> I LOVE COOKIES!
[28-Jul-2009 16:57:33] <MACscr> how do I simple agknowledge an alert so that it returns the status to green?
[28-Jul-2009 16:57:49] <rmatte> yikes, could Asterisk have picked any more eye-bleeding colours for their site?
[28-Jul-2009 16:58:10] <rmatte> MACscr: move the event to history
[28-Jul-2009 16:58:15] <MACscr> oh, I see, that drop down error thing is lame =P
[28-Jul-2009 16:58:27] <rmatte> drop down error?
[28-Jul-2009 16:58:41] <MACscr> er, arrow I mean
[28-Jul-2009 16:58:55] <rmatte> well now that you know it's there, they are everywhere
[28-Jul-2009 16:58:57] * MACscr not a fan of drop downs
[28-Jul-2009 16:59:01] <MACscr> yep
[28-Jul-2009 16:59:05] <rmatte> I kind of like it actually
[28-Jul-2009 16:59:13] <rmatte> takes all those links off the screen
[28-Jul-2009 16:59:16] <rmatte> to avoid clutter
[28-Jul-2009 17:03:20] <mrayzenoss> it gets better with the new UI
[28-Jul-2009 17:03:36] <mrayzenoss> which we'll put out a public demo probably this week for the new event console
[28-Jul-2009 17:38:25] <al_x> hmm
[28-Jul-2009 17:38:37] <al_x> my temperature sensors for my hp proliant server disappeared
[28-Jul-2009 17:38:45] <al_x> the template is still bound
[28-Jul-2009 17:39:59] <al_x> ... or is it
[28-Jul-2009 17:52:02] <al_x> does anyone know what this means? Unlinked OID in HOST-RESOURCES-TYPES
[29-Jul-2009 00:00:46] [disconnected at Wed Jul 29 00:00:46 2009]
[29-Jul-2009 00:00:46] [connected at Wed Jul 29 00:00:46 2009]
[29-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[29-Jul-2009 07:26:01] <m_pahlevanadeh> hi
[29-Jul-2009 07:26:15] <m_pahlevanadeh> i'm  newbie
[29-Jul-2009 07:27:13] <m_pahlevanadeh> i just installed on debian vi dpkg -i zenoss-stack....
[29-Jul-2009 07:27:36] <m_pahlevanadeh> Now, i don't know how to access its web interface
[29-Jul-2009 07:27:43] <m_pahlevanadeh> which URL?
[29-Jul-2009 07:28:56] <ckrough> http://your_server:8080/
[29-Jul-2009 07:29:28] <m_pahlevanadeh> thx
[29-Jul-2009 07:29:39] <ckrough> np
[29-Jul-2009 07:29:41] <ckrough> http://www.zenoss.com/community/docs/installation-guide/2.4.2/
[29-Jul-2009 07:29:44] <ckrough> that should help
[29-Jul-2009 07:30:02] <ckrough> lots of good stuff here:http://www.zenoss.com/community/docs
[29-Jul-2009 07:30:09] <ckrough> and here: http://www.zenoss.com/community/docs/howtos
[29-Jul-2009 08:43:11] <ckrough> something has gone screwey with the stylesheets on zenoss.com... or FF3.5.1
[29-Jul-2009 08:48:57] <jb> looks ok for me
[29-Jul-2009 08:53:08] <rmatte> good morning Matt
[29-Jul-2009 08:53:29] <mrayzenoss> good morning
[29-Jul-2009 08:59:38] <ckrough> jb: if you click through to something like Zenoss Developers Guide from http://www.zenoss.com/community/docs, are the fonts much bigger than usual?
[29-Jul-2009 08:59:42] <ckrough> morning matt
[29-Jul-2009 09:00:43] <mrayzenoss> ckrough: hmmm... the fonts are a little larger
[29-Jul-2009 09:01:29] <ckrough> no biggy, just noticed it recently
[29-Jul-2009 09:02:25] <mrayzenoss> I think the template was updated with the 2.4.2 refresh on the docs
[29-Jul-2009 09:02:29] <ckrough> it seems to happen on a lot of stuff in the community docs, but its not consistent. 2.3.3 docs dont do it. Maybe its isolated to 2.4
[29-Jul-2009 09:03:13] <mrayzenoss> we fixed a bunch of HTML rendering issues with the 2.4.2 rev on the docs, and introduced HTML versions of the Install Guide and Extended Monitoring Guides
[29-Jul-2009 09:03:26] * ckrough loves html versions
[29-Jul-2009 09:03:34] <ckrough> PDF is overused/abused
[29-Jul-2009 09:03:38] <mrayzenoss> going forward, everything will be in HTML and PDF except the Getting Started Guide
[29-Jul-2009 09:03:45] <mrayzenoss> I like search better in PDF
[29-Jul-2009 09:03:51] <mrayzenoss> but HTML is better for linking
[29-Jul-2009 09:05:01] <chudler> speaking of guides/docs. One of my cow-orkers pointed out to me that the HA section of some docs has been removed. I also couldn't find it anymore. Should I read into that?
[29-Jul-2009 09:06:01] <mrayzenoss> I think Professional Services is rewriting it
[29-Jul-2009 09:06:15] <mrayzenoss> they didn't follow their own instructions, so we told them to rewrtie them
[29-Jul-2009 09:06:40] <ckrough> lol. inter-department drama, you guys *are* growing
[29-Jul-2009 09:07:00] <chudler> thats cool. I think it wasn't generally available anyways. I followed the guide nearly to the letter and it worked out pretty good, FWIW.
[29-Jul-2009 09:07:14] <chudler> That is, I am not sure the Core docs ever had it, did it?
[29-Jul-2009 09:07:36] <mrayzenoss> well, we've tried to make the docs reflect how we use the product, the Install Guide is getting some more additions as well
[29-Jul-2009 09:07:58] <mrayzenoss> I don't believe they were in the Core docs, but they were based off of one of the 2 or 3 techniques in the Community
[29-Jul-2009 09:08:21] <ckrough> Documentation has been getting better at an exponential rate. It used to be one of my peeves with Zenoss, but any more I can find what I need fairly quickly. Thanks for the effort!
[29-Jul-2009 09:08:42] <mrayzenoss> I think the HA stuff was based on this: http://www.zenoss.com/Members/netdata/create-a-highavailable-zenoss/
[29-Jul-2009 09:08:54] <mrayzenoss> and netdata is a customer, so he and Chet probably wrote that together
[29-Jul-2009 09:09:49] <mrayzenoss> so there is or will be an Enterprise doc for HA
[29-Jul-2009 09:10:00] <chudler> sweet.  Thanks for the info
[29-Jul-2009 09:10:12] <mrayzenoss> and I'm working on migrating the community content to the new website now
[29-Jul-2009 09:10:25] <mrayzenoss> so nothing will be lost, and hopefully it will be easier to find and maintain
[29-Jul-2009 09:14:04] <ckrough> mrayzenoss: is this site run in Plone?
[29-Jul-2009 09:14:15] <mrayzenoss> yeah, we're investigating alternatives right now
[29-Jul-2009 09:14:26] <ckrough> we use it internally here
[29-Jul-2009 09:14:44] <jb> expressionengine
[29-Jul-2009 09:14:47] <mrayzenoss> The Community and WWW website are going to be run on different platforms
[29-Jul-2009 09:15:26] <ckrough> you guys were messing with that .. whats it called... mindmeld?
[29-Jul-2009 09:15:33] <mrayzenoss> MindTouch
[29-Jul-2009 09:15:46] <ckrough> Ill have to play with it sometime in the lab
[29-Jul-2009 09:16:01] <ckrough> plone is a beast, but we have so much documentation in there that migrating out is scary
[29-Jul-2009 09:18:26] <mrayzenoss> well, we're migrating out
[29-Jul-2009 09:18:46] <mrayzenoss> and I'll be sorting through all the content that has been stuffed into every random location
[29-Jul-2009 09:18:58] <jb> good.. its sort of confusing right now
[29-Jul-2009 09:19:02] <mrayzenoss> agreed
[29-Jul-2009 09:19:36] <mrayzenoss> It's nice to see how well the community has done with such a crappy website and wiki, I expect big things from the new platform
[29-Jul-2009 09:29:56] <ckrough> can I make configuration changes to to the system through xmlrpc? so, rather than sending an evt to ZenEventManager, am I also able to send a user definition to ZenUsers?
[29-Jul-2009 09:30:14] * ckrough trying to wrap my mind around how zendmd interacts with xmlrpc
[29-Jul-2009 09:58:21] <iprone> Hello, I was curious to know of anyone here has setup a distributed collector for Core? I have followed the http://www.zenoss.com/Members/fdeckert/how-to-install-distributed-collectors/ documentation, all seemed fine, but as soon as I added the new collector and set the Render URL on the main Zenoss instance, I noticed that a lot of stuff went red, and then I went over to the distributed collector and I could see it trying to che
[29-Jul-2009 09:59:34] <rmatte> iprone: I've only tried it once, and none of the RRDs were populating so I went back to separate instances of Zenoss, which is actually working out better for us anyways
[29-Jul-2009 10:01:31] <iprone> rmatte: I was hoping I wouldn't have to use a separate instance, ultimately though I'll need to have a single dashboard per the user requirements.
[29-Jul-2009 10:01:40] <rmatte> ah
[29-Jul-2009 10:01:54] <rmatte> then figuring that out is really your best bet
[29-Jul-2009 10:02:14] <rmatte> I'll have to play around with it sometime and see if I can write up a better guide to get it working
[29-Jul-2009 10:02:35] <rmatte> which reminds me, I still have a LDAP guide to write when I have a second
[29-Jul-2009 10:08:28] <mrayzenoss> improvements to http://www.zenoss.com/community/docs/howtos/how-to-authenticate-via-ldap/ ?
[29-Jul-2009 10:16:49] <rmatte> yeh, remember that forum posting I did?
[29-Jul-2009 10:16:56] <rmatte> it'll basically be a dressed up version of that
[29-Jul-2009 10:17:31] <rmatte> http://forums.zenoss.com/viewtopic.php?t=9652&highlight=ldap
[29-Jul-2009 10:19:07] <rmatte> it's just going to go a little more in to detail in terms of setting up the roles
[29-Jul-2009 10:56:31] <mrayzenoss> anyone have Citrix NetScalers?  http://forums.zenoss.com/viewtopic.php?t=10492
[29-Jul-2009 10:59:59] <Grizmawe> Good afternoon - Does the zenmodeler run according to a schedule and remodel all devices or is it only a manual process?
[29-Jul-2009 11:01:15] <mrayzenoss> Grizmawe: if you go to Collectors->localhost you'll see the setting "Modeler Cycle Interval (mins)"
[29-Jul-2009 11:01:32] <mrayzenoss> by default, every 12 hours
[29-Jul-2009 11:02:00] <mrayzenoss> you can lock devices to prevent remodeling and/or turn this off if you like
[29-Jul-2009 11:04:01] <rmatte> do you "turn it off" by setting it to 0?
[29-Jul-2009 11:04:33] <rmatte> I actually just configured my boxes to prevent the process from running unless called upon
[29-Jul-2009 11:05:10] <Grizmawe> mrayzenoss: many thanks - can I tell it to remodel everything in a class (/servers/windows) now or only wait for the next interval. is there a parameter I can pass to zenmodeler to tell it to scan the class. I know i can do zenmodeler start but how can i tell it i only need to remodel a certain class
[29-Jul-2009 11:06:13] <mrayzenoss> rmatte: I assume 0 is off
[29-Jul-2009 11:06:16] <rmatte> yeh there is, I think it's "zenmodeler run --class=/whatever/whatever
[29-Jul-2009 11:06:33] <rmatte> mrayzenoss: I just followed the guide http://blog.zenoss.com/2009/01/27/tip-of-the-month-running-only-some-daemons/ to disable the daemon.
[29-Jul-2009 11:06:51] <mrayzenoss> rmatte: yeah, that works too
[29-Jul-2009 11:07:06] <ckrough> jsonGetDeviceNames is fast. happy happy
[29-Jul-2009 11:07:24] <mrayzenoss> zenmodeler run --path=/Server/Linux
[29-Jul-2009 11:08:14] <rmatte> yeh, I just noticed it was path hehe
[29-Jul-2009 11:08:47] <rmatte> http://www.zenoss.com/community/docs/zenoss-guide/2.2.4/ch24s06.html
[29-Jul-2009 11:08:53] <rmatte> daemon command options
[29-Jul-2009 11:11:59] <Grizmawe> many thanks to both of you
[29-Jul-2009 11:53:22] <Disconnect> hmm. anyone wanna help me figure out this snmp mess? if i do a 'reset community' and model, it works (usually) for a while, then starts to fail again. snmpwalk works, can't figure it out (and its severely impacting my goal of "make zenoss the authoritative monitor, and make it mostly/all green" ..)
[29-Jul-2009 11:53:32] <Disconnect> on existing, up, running, functional hosts
[29-Jul-2009 12:20:16] <jb> anybody have any ideas on how to pipe postfix data into zen?
[29-Jul-2009 12:22:07] <Disconnect> "data"? i have a nagios mailq monitor that I'm using with zenoss w/o much trouble
[29-Jul-2009 12:22:34] <jb> hmm. i more interested in number of messages sent/recieved
[29-Jul-2009 12:22:37] <jb> per minute/etc
[29-Jul-2009 12:26:03] <rmatte> jb: as long as you can write a script to collect the data you're looking for, you can use the script as a command based datapoint
[29-Jul-2009 12:26:20] <jb> yeah.. trying to figure out how to collect the data
[29-Jul-2009 12:26:42] <jb> or collect it locally on the postfix box, and pipe it thru net-snmp
[29-Jul-2009 12:26:49] <jb> which is how I do BIND..
[29-Jul-2009 12:26:52] <rmatte> can't help you there
[29-Jul-2009 12:27:15] <rmatte> yeh, you'd probably have to do something like that
[29-Jul-2009 12:33:18] <jb> hmm ok.. think i may have figured something out
[29-Jul-2009 12:33:22] <jb> if I can get the graphs working correctly
[29-Jul-2009 12:33:25] <jb> i'll let you know
[29-Jul-2009 12:40:37] <rmatte> cool
[29-Jul-2009 12:46:11] <al_x> how can i ensure i have hp insight manager mib
[29-Jul-2009 12:49:00] <rmatte> download and install it using zenmib
[29-Jul-2009 12:49:30] <rmatte> just put it in a folder and do: zenmib run <mib file name>
[29-Jul-2009 12:49:40] <rmatte> as zenoss user
[29-Jul-2009 12:51:03] <mrayzenoss> I prefer the MIB Browser ZenPack, it allows you to upload MIBs via the browser
[29-Jul-2009 12:51:12] <mrayzenoss> http://www.zenoss.com/community/projects/zenpacks/mib-browser
[29-Jul-2009 12:52:37] <rmatte> that works well for single mibs files, yeh
[29-Jul-2009 12:53:42] <rmatte> if we're receiving traps like: enterprises.2.24.565.13.123.1.21.1 or whatever, I'm assuming it's because we're missing Mibs?
[29-Jul-2009 12:53:51] <rmatte> or is that what the traps are supposed to look like?
[29-Jul-2009 12:54:19] <rmatte> I remember when we didn't have the Cisco mibs they just came in as straight OIDs
[29-Jul-2009 12:54:30] <rmatte> so that's why I'm wondering if those are only partially getting translated or something
[29-Jul-2009 12:54:55] <rmatte> we're able to transform them fine since there are things that come in on the details tab, but it just gets a bit confusing
[29-Jul-2009 12:54:56] <mrayzenoss> hmmm... not sure
[29-Jul-2009 12:55:15] <rmatte> we're primarily seeing it with windows boxes
[29-Jul-2009 12:58:55] <rmatte> Maybe we are missing Mibs.  I'll have to look in to it tomorrow.
[29-Jul-2009 13:12:25] <al_x> sorry i guess this isn't the place to ask
[29-Jul-2009 13:12:42] <al_x> but im getting snmpd errors about unlinked oid
[29-Jul-2009 13:12:49] <al_x> and undefined identifier
[29-Jul-2009 13:12:50] <konfoo> anyone seen this issue on 2.4 - add a device, set zproperties snmp monitoring to false, and it still continues to throw snmp scan errors?
[29-Jul-2009 13:13:08] <al_x> and my hp proliant pack temp sensor has disappeared from the hardware page
[29-Jul-2009 13:14:54] <jb> konfoo: have you tried remodeling?
[29-Jul-2009 13:15:10] <Disconnect> konfoo: i've got some devices in that pile. they're sitting on the sideline until i can figure out why it keeps losing track of the ones that -do- have snmp
[29-Jul-2009 13:17:03] <konfoo> ah.. think i may have found the issue.. lets see what happens on rescan
[29-Jul-2009 13:17:22] <konfoo> nope, still scanning snmp
[29-Jul-2009 13:18:02] <konfoo> remodeled, waiting for rescan
[29-Jul-2009 13:19:12] <jb> if I have an OID that provides a rolling total (counter?).. what are my options for having zen graph?
[29-Jul-2009 13:19:18] <jb> would I have to use some custom graph definitions?
[29-Jul-2009 13:19:29] <jb> i want to graph something like "messages per minute"
[29-Jul-2009 13:20:10] <rmatte> konfoo: push changes to the collector after the remodel as well
[29-Jul-2009 13:20:20] <rmatte> Manage -> Push Changes
[29-Jul-2009 13:21:10] <rmatte> jb: you want to graph messages per minute with a 5 minute polling cycle?
[29-Jul-2009 13:21:23] <jb> actually..i guess i would need to do a per 5 minuts
[29-Jul-2009 13:21:29] <rmatte> yeh
[29-Jul-2009 13:21:30] <jb> with BIND.. i have the same dillema
[29-Jul-2009 13:21:53] <jb> so I use an RPN of "300,*"
[29-Jul-2009 13:21:58] <jb> i guess this is the only way to do it
[29-Jul-2009 13:27:21] <konfoo> hmm remodeling and pushing the changes didnt help.. still monitoring snmp
[29-Jul-2009 13:27:30] <konfoo> the other device i added works fine however
[29-Jul-2009 13:28:16] <rmatte> what's the issue again?
[29-Jul-2009 13:28:39] <konfoo> device set to ping only (no snmp) still gets scanned with snmp and generates an error
[29-Jul-2009 13:39:07] <konfoo> hmm restarting zenoss seems to have fixed it
[29-Jul-2009 13:41:42] <jb> have you guys found that google indexes the zenoss forums/lists better than the "search" function on the forums?
[29-Jul-2009 13:42:17] <konfoo> it does for most forums
[29-Jul-2009 13:46:08] <Disconnect> except for it's own android lists
[29-Jul-2009 13:46:29] <rmatte> konfoo: you're sure you set the right zProperties setting for that device?
[29-Jul-2009 13:48:04] <konfoo> rmatte: 100%
[29-Jul-2009 13:48:51] <rmatte> last thing I can think of is to go in to zendmd as the zenoss user then type:
[29-Jul-2009 13:48:54] <rmatte> reindex()
[29-Jul-2009 13:48:56] <rmatte> commit()
[29-Jul-2009 13:49:03] <rmatte> that will probably fix it
[29-Jul-2009 13:49:06] <konfoo> thanks, noted
[29-Jul-2009 13:51:49] <jb> anybody used net-snmp'd extend?
[29-Jul-2009 14:07:13] <m3hm3t> hiya
[29-Jul-2009 14:08:50] <m3hm3t> whereis zenoss-plugins-*.rpm located?
[29-Jul-2009 14:12:07] <rmatte> m3hm3t: http://sourceforge.net/projects/zenoss/files/
[29-Jul-2009 14:12:12] <rmatte> under zenplugins
[29-Jul-2009 14:14:29] <straterra> heh
[29-Jul-2009 14:14:50] <straterra> the CPU graph in Zenoss for one of my Linux machines shows 0% idle..but 0% system/user/wait too
[29-Jul-2009 14:14:57] <straterra> top on the machine shows 99% idle
[29-Jul-2009 14:15:20] <rmatte> that's a problem with netsnmp then, not Zenoss
[29-Jul-2009 14:15:38] <rmatte> if you snmpwalk the cpu values they will all reflect what's shown in Zenoss
[29-Jul-2009 14:16:12] <straterra> If only I could get in to the cacti installation
[29-Jul-2009 14:21:43] <straterra> rmatte: I found out what it is..it's snmpd eating up all the CPU when zenoss gets info from it
[29-Jul-2009 14:23:02] <rmatte> I see
[29-Jul-2009 14:23:11] <rmatte> that shouldn't happen though
[29-Jul-2009 14:23:27] <rmatte> it's just polling a few values, if that's eating up all of the CPU that's quite the issue
[29-Jul-2009 14:23:52] <straterra> Well..that's what I saw when I did an snmpwalk
[29-Jul-2009 14:35:47] <m3hm3t> am i too dump or is it not possible to remove a device?
[29-Jul-2009 14:35:49] <m3hm3t> thanks rmatte
[29-Jul-2009 14:49:15] <straterra> rmatte: what method should I use to query the server and see what idle % its giving?
[29-Jul-2009 14:56:07] <ckrough> hmm, two of my sites allow me to do REST calls with urllib2 using HTTPPasswordMgrWithDefaultRealm() and HTTPS, 3 dont
[29-Jul-2009 14:56:25] <ckrough> LDAP configs are all the same, http redirects all the same
[29-Jul-2009 14:56:38] * ckrough is at a loss
[29-Jul-2009 14:56:47] <m3hm3t> zTransportPreference --- isn't where it should be
[29-Jul-2009 15:52:29] <straterra> rmatte: Yeeeaaah..I think its that machine. I don't know what exactly is making its snmp stuff messed up..but it is. Cacti's graphs are all NAN
[29-Jul-2009 15:54:38] <ckrough> try rebuilding your snmpd.conf? maybe you have something bad in there
[29-Jul-2009 15:55:01] <straterra> Nothing has changed in it
[29-Jul-2009 16:14:58] <konfoo> anyone here using snare to send events to zensyslog?
[29-Jul-2009 16:15:11] <konfoo> (windows snare, that is)
[29-Jul-2009 17:18:55] <mrayzenoss> I know I've heard of folks using it
[29-Jul-2009 17:19:12] <mrayzenoss> I think jmp242 is who told me about it
[30-Jul-2009 00:00:46] [disconnected at Thu Jul 30 00:00:46 2009]
[30-Jul-2009 00:00:46] [connected at Thu Jul 30 00:00:46 2009]
[30-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[30-Jul-2009 04:09:51] <Dieterbe> Hi, when i add a http service check to a device, and then the remodel the device, the check disappears. why could this be?
[30-Jul-2009 04:14:14] <sergeymasushko> Dieterbe: try to lock the service
[30-Jul-2009 04:20:42] <Dieterbe> thanks, that seems to work
[30-Jul-2009 04:20:49] <Dieterbe> though i don't really understand the logic
[30-Jul-2009 06:28:13] <Dieterbe> anyone an idea what i should put in the 'send string' part for a http check?
[30-Jul-2009 07:23:32] <Disconnect> hmm. here's a host that claims snmp agent down since minutes after I fixed this yesterday, but the agent is fine and snmpwalk with the community (from the zenoss box) works fine.. I've got a ton like this. any suggestions? ("forget zenoss" is getting higher on the list but nagios is just so BAD...)
[30-Jul-2009 07:37:46] <Disconnect> sigh. model the device, 'snmp timeout'. zenmodeler from the commandline works fine. lets see if it sticks.
[30-Jul-2009 07:43:39] <Disconnect> any way to do 'reset community' from the commandline? at least I can speed this process up that way
[30-Jul-2009 08:30:46] <Disconnect> hmm. at what point does it make sense to just wipe it out and start over?
[30-Jul-2009 08:33:48] <Dieterbe> i want to monitor a few "applications", basically just http but on a specific url and with specific return stuff. oddly enough when i do add ip service -> http, this check always returns OK even for paths that don't exist (eg send string: GET /dontexist HTTP/1.0), and when i try to create a new service with the idea of later applying that service to a device or device class, i cannot specify the attributes such as request uri
[30-Jul-2009 08:44:12] <Disconnect> Dieterbe: there is a way to do that, lemme find it. (i've got a very similar test)
[30-Jul-2009 08:45:05] <Disconnect> HttpMonitor template
[30-Jul-2009 08:45:05] <Dieterbe> thanks
[30-Jul-2009 08:45:14] <Dieterbe> hmm i have that installed
[30-Jul-2009 08:46:05] <Dieterbe> please go on
[30-Jul-2009 08:46:13] <Disconnect> yah hang on pulling it up again
[30-Jul-2009 08:46:29] <Disconnect> on the new template page, add a data source
[30-Jul-2009 08:46:39] <Disconnect> of type httpmonitor
[30-Jul-2009 08:47:45] <Disconnect> there is a url and regex - set the regex to whatever you want to find (eg "<!-- Render Complete -->" or whatever) not sure what it does for status though
[30-Jul-2009 08:48:29] <Dieterbe> Disconnect: you mean: go to the device -> more->templates?
[30-Jul-2009 08:48:52] <Disconnect> Dieterbe: go to the class (eg Server/Linux/Webserver) and hit 'templates' then create a new template for it
[30-Jul-2009 08:49:29] * Disconnect has an issue with thresholds there though - min value "" max value 5 (5 second warning) and it fires on correct subsecond responses "oh no its fast, quick someone help!"
[30-Jul-2009 08:50:46] <Disconnect> sigh. speaking of.. looks like the data gathering for that broke when I upgraded zenoss.
[30-Jul-2009 08:50:53] <Dieterbe> heh
[30-Jul-2009 08:50:57] <Disconnect> thats Bad since I have to present those charts tomorrow
[30-Jul-2009 08:50:58] <Dieterbe> i'll try to do this
[30-Jul-2009 08:51:59] <jb> hrm ok looks like I got my postfix stuff working
[30-Jul-2009 08:55:07] <Dieterbe> Disconnect: so i created the template, then went to a device and bound it. but now i can't find the statuscheck or anything on the devices page
[30-Jul-2009 08:55:21] <ckrough> anyone know of a REST or XMLRPC accessible method for getting a device's performance collector?
[30-Jul-2009 08:55:21] <Disconnect> template has the datapoint?
[30-Jul-2009 08:55:35] <Disconnect> (er, data source)
[30-Jul-2009 08:55:53] <Dieterbe> yes it has
[30-Jul-2009 08:56:52] <Disconnect> model the device
[30-Jul-2009 08:58:31] <Disconnect> (ugh. clicking on a device from the event console should not do a search if there is an exact match. "server1" leads me to "server1" "server11" "server12".."server110"..:(
[30-Jul-2009 08:59:18] <Dieterbe> Disconnect: still don't see it. not on the "main" page and not on the OS tab
[30-Jul-2009 08:59:25] <Disconnect> try 'performance'
[30-Jul-2009 09:01:20] <Dieterbe> Disconnect: hmm there are graphs (not working but whatever), but i actually don't really need graphs. i just want alerts/events
[30-Jul-2009 09:01:29] <Disconnect> the graphs will work once it populates.
[30-Jul-2009 09:01:31] <Disconnect> give it a sec
[30-Jul-2009 09:02:34] <Disconnect> (well, a few minutes. I don't know of a way to trigger an immediate retest of a host ..)
[30-Jul-2009 09:24:49] hackeron_ is now known as hackeron
[30-Jul-2009 09:25:37] <TBKDan> is there a way to manually remove a zenpack? The HttpMonitor zenpack errors out when I try to uninstall, and the Cisco MIBs zenpack has been "uninstalling" for over 30 minutes now.
[30-Jul-2009 09:39:38] <Disconnect> where do I look to find out why a perf graph is blank?
[30-Jul-2009 09:39:55] <TBKDan> How long has it been since the device was modeled?
[30-Jul-2009 09:40:15] <rmatte> TBKDan: the Cisco MIBs zenpack has been what for over 30 mins now?
[30-Jul-2009 09:40:20] <rmatte> (sorry, it got cut off)
[30-Jul-2009 09:40:36] <TBKDan> rmatte: Uninstalling, or so it claims.
[30-Jul-2009 09:40:48] <rmatte> yes, that ZenPack takes a very long time to uninstall
[30-Jul-2009 09:40:54] <rmatte> just like it takes a long time to install
[30-Jul-2009 09:41:04] <TBKDan> I don't remember it taking 30 mins to install >_<
[30-Jul-2009 09:41:09] <Disconnect> TBKDan: its a virtual device, so no modelling. it was working monday, then it just stopped for reasons unknown
[30-Jul-2009 09:41:20] <rmatte> it takes 5-7 hours to install, provided it installed properly
[30-Jul-2009 09:41:22] <TBKDan> Disconnect: Ah, ok.
[30-Jul-2009 09:41:29] <TBKDan> rmatte: ........
[30-Jul-2009 09:41:32] <rmatte> It states this right on the ZenPack page
[30-Jul-2009 09:42:07] <rmatte> "There are a very large number of MIBs included in this ZenPack (1128), so it is advised that you install with the command-line rather than the UI so it does not time-out. It may still take 4 or 5 hours to process."
[30-Jul-2009 09:42:10] <TBKDan> ... Missed that
[30-Jul-2009 09:42:11] <ckrough> anyone know how to get a devices 'getDevicePath()' via REST call? I can't pass that method to the device without already having the device's path... catch 22
[30-Jul-2009 09:42:38] <rmatte> took as long as 7 hours one time for me, but that was on a fairly slow server
[30-Jul-2009 09:43:01] <TBKDan> rmatte: Well, how about being unable to remove the HttpMonitor zenpack?
[30-Jul-2009 09:43:41] <rmatte> TBKDan: Have you tried restarting Zenoss, then removing it?
[30-Jul-2009 09:43:57] <TBKDan> rmatte: Yup.  Also tried removing var/*.zec, re-installing then removing... no go.
[30-Jul-2009 09:44:17] <rmatte> what's it saying when you try to remove it?
[30-Jul-2009 09:44:36] <TBKDan> rmatte: Is it ok to try to remove it while the Cisco MIB pack is in the removal process?
[30-Jul-2009 09:44:58] <rmatte> I'm not sure, you may have to wait
[30-Jul-2009 09:45:08] <rmatte> why are you removing the Cisco Mibs pack just out of curiosity?
[30-Jul-2009 09:45:41] <TBKDan> rmatte: Ever since I installed it, certain pages take a REALLY long time to load, especially the zenpack listing.
[30-Jul-2009 09:45:58] <rmatte> Yeh, thought so, you didn't read the ZenPack page at all did you? lol
[30-Jul-2009 09:45:59] <rmatte> Having this ZenPack installed appears to slow down the UI. A work-around has been found to delete the ZenPack and leave the MIBs installed. To do this, start zendmd and enter the following:
[30-Jul-2009 09:46:00] <rmatte> >>> dmd.ZenPackManager.packs._delObject('ZenPacks.Nova.Cisco.Mibs')
[30-Jul-2009 09:46:00] <rmatte> >>> commit()
[30-Jul-2009 09:46:00] <rmatte> Performing these steps will eliminate the performance issues, while leaving the MIBs installed.
[30-Jul-2009 09:46:13] <rmatte>
[30-Jul-2009 09:46:37] <TBKDan> haha
[30-Jul-2009 09:46:48] <TBKDan> Bit late for that
[30-Jul-2009 09:46:56] <rmatte> yeh, seems to be
[30-Jul-2009 09:47:07] <rmatte> actually, if you cancel the removal right now it shouldn't affect the pack
[30-Jul-2009 09:47:18] <rmatte> Zenoss has to do all these checks before removal, which is what takes so long
[30-Jul-2009 09:47:24] <TBKDan> I can re-install it later, but i was wondering if it had something to do with why I couldn't remove the HttpMonitor
[30-Jul-2009 09:47:26] <rmatte> the actual removal just takes a couple minutes at the end
[30-Jul-2009 09:47:36] <rmatte> I doubt that it does
[30-Jul-2009 09:47:50] <rmatte> but perhaps trying to remove 2 packs simultaneously is the issue
[30-Jul-2009 09:48:00] <TBKDan> I wasn't trying to remove them at the same time
[30-Jul-2009 09:48:27] <TBKDan> I had been trying to remove the http monitor, and when that was failing, I decided to try the Cisco removal. That's where I've been for the past ~40 minutes
[30-Jul-2009 09:48:42] <rmatte> I'm just saying that if you're in the process of removing one ZenPack, there's a good chance that you won't be able to remove another at the same time
[30-Jul-2009 09:48:46] <rmatte> ah
[30-Jul-2009 09:49:06] <rmatte> well, once the Cisco removal is done (or if you decide to cancel it), then I can help you out with removing the other one
[30-Jul-2009 09:49:39] <TBKDan> heh... if it lets me cancel it is a better choice of words
[30-Jul-2009 09:49:54] <rmatte> well, are you removing from commandline or UI?
[30-Jul-2009 09:49:57] <TBKDan> Throwing a KeyboardInterrupt exception it looks like.  But doesn't do anything
[30-Jul-2009 09:49:58] <TBKDan> CLI
[30-Jul-2009 09:50:06] <rmatte> hit ctrl-c multiple times
[30-Jul-2009 09:50:09] <rmatte> it'll eventually cancel
[30-Jul-2009 09:50:13] <rmatte> (the joys of python)
[30-Jul-2009 09:50:15] <TBKDan> Tried lol
[30-Jul-2009 09:50:15] <rmatte>
[30-Jul-2009 09:50:23] <TBKDan> There!
[30-Jul-2009 09:50:28] <rmatte> you need to hit it like 10 times or something
[30-Jul-2009 09:50:28] <TBKDan> 10+ times FTW!
[30-Jul-2009 09:50:28] <rmatte> hehe
[30-Jul-2009 09:50:31] <TBKDan> lol
[30-Jul-2009 09:50:44] <rmatte> go in to Zenoss and check to make sure that you still have 1128 Cisco Mibs
[30-Jul-2009 09:50:47] <rmatte> if you do then you're fine
[30-Jul-2009 09:51:00] <rmatte> then you can use that zendmd command (which only takes about 4 mins to run)
[30-Jul-2009 09:51:04] <TBKDan> The exception was being thrown in the "beforeDelete()" method, so probably... lemme check.
[30-Jul-2009 09:51:30] <rmatte> and then I can help you out with removing the http pack
[30-Jul-2009 09:51:54] <TBKDan> Say if I run this zendmd command, what if I really do wish to delete that zenpack in the future? Just re-install and un-install?
[30-Jul-2009 09:52:28] <ckrough> mrayzenoss: you around? got any good xmlrpc examples of calling methods? anything more detailed than the sendevt in the dev guide?
[30-Jul-2009 09:52:28] <rmatte> no, all the ZenPack is is that Cisco folder under Mibs
[30-Jul-2009 09:52:37] <rmatte> so just select that organizer and delete it
[30-Jul-2009 09:52:39] <rmatte> and you're done
[30-Jul-2009 09:52:43] <TBKDan> Awesome.
[30-Jul-2009 09:53:04] <rmatte> The ZenPack is really just to make installing the Mibs easier since it's a total pain to do it by hand
[30-Jul-2009 09:53:30] <TBKDan> Ok. Well I'm still waiting for the ZenPack page to come up to see if it's still listed in there.
[30-Jul-2009 09:53:32] <TBKDan> Yup
[30-Jul-2009 09:53:55] <TBKDan> WARNING OFS.Application Duplicate Product name
[30-Jul-2009 09:53:55] <TBKDan> After loading Product 'Five' from '/opt/zenoss/Products',
[30-Jul-2009 09:53:55] <TBKDan> I skipped the one in '/opt/zenoss/lib/python/Products'.
[30-Jul-2009 09:54:03] <TBKDan> (While starting ZenDMD)
[30-Jul-2009 09:54:05] <rmatte> yeh, that response issue with zProperties, the ZenPack page, and other pages will dissapear, just execute that zendmd command
[30-Jul-2009 09:54:22] <rmatte> that's fine, it does that to me too, some bug from upgrading from 2.3.3 to 2.4.x
[30-Jul-2009 09:54:26] <mrayzenoss> ckrough: I've got a couple that Kells sent me, I'll throw them up on a wiki page
[30-Jul-2009 09:54:27] <rmatte> just ignore that
[30-Jul-2009 09:54:33] <TBKDan> Will do.
[30-Jul-2009 09:54:56] <ckrough> mrayzenoss: your the man
[30-Jul-2009 09:54:57] <TBKDan> Still waiting for dmd to load........
[30-Jul-2009 09:55:11] <rmatte> should only take like 30 seconds
[30-Jul-2009 09:55:14] <rmatte> or less
[30-Jul-2009 09:55:28] <rmatte> if it takes longer then something is most likely wrong
[30-Jul-2009 09:55:41] <TBKDan> Just ran the command.
[30-Jul-2009 09:55:43] <TBKDan> Well
[30-Jul-2009 09:55:45] <TBKDan> Running
[30-Jul-2009 09:56:06] <rmatte> yeh, the delObject command will take like 4-5 mins, then the commit will take like 2 or 3
[30-Jul-2009 09:56:59] * TBKDan twiddles his thumbs
[30-Jul-2009 10:00:20] <rmatte> the response times should increase right away after the commit, you don't need to restart Zenoss or anything
[30-Jul-2009 10:00:42] <TBKDan> Well right now going to the device list is generating an error. i hope that resolves once the commit finishes....
[30-Jul-2009 10:01:20] <rmatte> yeh, wait for the commit to finish
[30-Jul-2009 10:01:24] <TBKDan> Done
[30-Jul-2009 10:01:29] <rmatte> now check it
[30-Jul-2009 10:01:34] <TBKDan> Still an error
[30-Jul-2009 10:01:37] <rmatte> what error?
[30-Jul-2009 10:02:10] <rmatte> (I never saw any error when I did it, and I did it to 5 different Zenoss servers running either 2.3.3 or 2.4.2
[30-Jul-2009 10:02:43] <jb> just posted a howto on how I used zenoss to graph postfix stats (using net-snmpd's extend) to the forums if anyone is interested..
[30-Jul-2009 10:02:59] <rmatte> jb: cool, I'll take a look in a few
[30-Jul-2009 10:03:04] <jb> k
[30-Jul-2009 10:03:11] <TBKDan> rmatte: http://pastebin.com/d4faae4cd
[30-Jul-2009 10:03:14] <jb> let me know if it makes sense
[30-Jul-2009 10:03:19] <rmatte> will do
[30-Jul-2009 10:03:31] <rmatte> just need to get TBKDan sorted out first
[30-Jul-2009 10:04:02] <rmatte> TBKDan: ok, well, restart Zenoss for starters and see if that fixes it
[30-Jul-2009 10:04:04] <TBKDan> I'm gonna remove the .zec's and restart zenoss...
[30-Jul-2009 10:04:08] <rmatte> if not, then we'll try something else
[30-Jul-2009 10:04:21] <rmatte> k
[30-Jul-2009 10:04:51] <TBKDan> Device list!
[30-Jul-2009 10:04:52] <TBKDan> lol
[30-Jul-2009 10:05:02] <rmatte> so it works?
[30-Jul-2009 10:05:04] <TBKDan> Zenpack listing came up much faster
[30-Jul-2009 10:05:07] <rmatte> good
[30-Jul-2009 10:05:08] <TBKDan> Everything seems to be in order
[30-Jul-2009 10:05:14] <rmatte> ok, now for the Http pack
[30-Jul-2009 10:05:20] <rmatte> try removing it and pastebin the error
[30-Jul-2009 10:05:40] <TBKDan> Is there a way to get more verbose output via CLI removal?
[30-Jul-2009 10:05:45] <TBKDan> I didn't see anything in the --help
[30-Jul-2009 10:05:59] <rmatte> put -v10 at the end of the command
[30-Jul-2009 10:06:19] <rmatte> it'll display the highest verbosity
[30-Jul-2009 10:06:20] <TBKDan> From the Web GUI: http://pastebin.com/d486973d8
[30-Jul-2009 10:06:54] <mrayzenoss> try restarting zopectl and zenhub, then try again
[30-Jul-2009 10:07:04] <rmatte> he just restarted all of Zenoss
[30-Jul-2009 10:07:15] <TBKDan> Bit more info on the CLI -v10, one sec
[30-Jul-2009 10:07:20] <rmatte> k
[30-Jul-2009 10:07:43] <mrayzenoss> ckrough: here are some more zendmd scripts, some use xml-rpc http://www.zenoss.com/community/wiki/tips-and-tricks/KellsZendmdScripts
[30-Jul-2009 10:08:10] <TBKDan> rmatte: http://pastebin.com/d6176e75d
[30-Jul-2009 10:09:23] <rmatte> TBKDan: ok, we're going to try a series of steps and hopefully these will work...
[30-Jul-2009 10:09:35] <TBKDan> Okey dokey
[30-Jul-2009 10:09:42] <rmatte> first off, go back in to zendmd and execute a reindex() then a commit() and try removing it again
[30-Jul-2009 10:10:00] <TBKDan> ...
[30-Jul-2009 10:10:02] <TBKDan> starting zendmd
[30-Jul-2009 10:10:07] <TBKDan> More errors than before
[30-Jul-2009 10:10:21] <rmatte> pastebin them please
[30-Jul-2009 10:10:36] <TBKDan> http://pastebin.com/m71c157c2
[30-Jul-2009 10:10:50] <edwardam> has anyone ever looking into integrating ZenOSS with OpenID ?
[30-Jul-2009 10:11:01] <rmatte> ok, in zendmd do a reindex and a commit
[30-Jul-2009 10:11:09] <rmatte> it sounds like your database is out of wack
[30-Jul-2009 10:11:28] <TBKDan> Is reindex() the only thing to run checks/repairs onit?
[30-Jul-2009 10:11:29] <TBKDan> on it*
[30-Jul-2009 10:11:36] <rmatte> yeh
[30-Jul-2009 10:11:55] <rmatte> well, it reindexes the database to make sure that everything is in check
[30-Jul-2009 10:12:03] <rmatte> try it out, hopefully it'll resolve the issue
[30-Jul-2009 10:12:08] <TBKDan> Ok.  It's doing it's thing
[30-Jul-2009 10:12:11] <TBKDan> Crap
[30-Jul-2009 10:12:13] <rmatte> k cool
[30-Jul-2009 10:12:14] <TBKDan> More errors
[30-Jul-2009 10:12:14] <rmatte> ?
[30-Jul-2009 10:12:19] <rmatte> pastebin please
[30-Jul-2009 10:12:41] <TBKDan> http://pastebin.com/m592f0bd1
[30-Jul-2009 10:13:13] <rmatte> ok, as zenoss user: zopectl restart && zenhub restart
[30-Jul-2009 10:13:18] <rmatte> then open zendmd again
[30-Jul-2009 10:13:39] <TBKDan> Working... (fyi, I've been doing this all as zenoss user)
[30-Jul-2009 10:13:50] <TBKDan> Can't start zendmd now
[30-Jul-2009 10:13:59] <rmatte> eugh
[30-Jul-2009 10:14:18] <TBKDan> http://pastebin.com/m4dad1ba7
[30-Jul-2009 10:14:23] <rmatte> "raise ValueError("corrupted record, data")", that doesn't look good
[30-Jul-2009 10:14:58] <TBKDan> But that's with the cache, correct?
[30-Jul-2009 10:15:27] <rmatte> correct
[30-Jul-2009 10:15:49] <rmatte> the zeo cache
[30-Jul-2009 10:16:00] <mrayzenoss> edwardam: I haven't seen that integration yet
[30-Jul-2009 10:16:29] <edwardam> mrayzenoss: anyone at zenoss interested in doing it?
[30-Jul-2009 10:16:37] <mrayzenoss> edwardam: I've seen this, which reportedly works with Zenoss http://plone.org/products/webserverauth
[30-Jul-2009 10:16:42] <TBKDan> Restarting zenoss again...
[30-Jul-2009 10:16:46] <rmatte> TBKDan: kill off any zeo cache files and restart again I guess
[30-Jul-2009 10:16:50] <mrayzenoss> edwardam: not something that has come up much
[30-Jul-2009 10:16:51] <TBKDan> Yup, just what I did
[30-Jul-2009 10:17:03] <rmatte> TBKDan: hopefully it's not the actual zope database that's become corrupt
[30-Jul-2009 10:17:05] <edwardam> mrayzenoss: yeah, that's not the same thing at all. :-)
[30-Jul-2009 10:17:16] <mrayzenoss> edwardam: no, I know
[30-Jul-2009 10:17:25] <mrayzenoss> edwardam: I'm just saying that's the only other auth thing out there I'm aware of
[30-Jul-2009 10:17:30] <TBKDan> rmatte:
[30-Jul-2009 10:17:35] <TBKDan> http://pastebin.com/m6539a77
[30-Jul-2009 10:17:40] <TBKDan> Friggin c&p fail.
[30-Jul-2009 10:17:43] <mrayzenoss> edwardam: we're still working on getting LDAP/AD to work right
[30-Jul-2009 10:17:46] <TBKDan> At least I can get back to zendmd...
[30-Jul-2009 10:18:02] <edwardam> mrayzenoss: IMO that's a dead end. ;-) But probably important to your enterprise customers.
[30-Jul-2009 10:18:27] <rmatte> TBKDan: ok, give me a few minutes, need to research this....
[30-Jul-2009 10:18:30] <edwardam> mrayzenoss: think this would work: http://pypi.python.org/pypi/plone.openid ?
[30-Jul-2009 10:18:38] <TBKDan> Ok, Thanks for all your help rmatte!
[30-Jul-2009 10:18:52] <edwardam> does zenoss use PAS ? probably not.
[30-Jul-2009 10:19:31] <mrayzenoss> edwardam: I'm running it by Ian
[30-Jul-2009 10:20:23] <edwardam> atm I'm very torn between totally diving in with zenoss ... or just using collectd and a home rolled web layer on top of it.
[30-Jul-2009 10:20:29] <rmatte> TBKDan: close zendmd and open it again, same error?
[30-Jul-2009 10:20:31] <mrayzenoss> edwardam: he said it should work with SVN trunk, since we've moved to Zope 2.11.2 there
[30-Jul-2009 10:20:40] <mrayzenoss> but not the current product, since it's on 2.8
[30-Jul-2009 10:20:44] <TBKDan> rmatte: ... No. lol
[30-Jul-2009 10:20:49] * edwardam is soooooooo out of the zope loop. :-(
[30-Jul-2009 10:20:58] <rmatte> TBKDan: so it loaded ok this time?
[30-Jul-2009 10:20:59] <edwardam> mrayzenoss: I c .
[30-Jul-2009 10:21:04] <mrayzenoss> edwardam: well, we were in the 2.8 dark ages
[30-Jul-2009 10:21:05] <jmp242> Anyone know what this means?
[30-Jul-2009 10:21:06] <TBKDan> rmatte: Just the 'Five' Warning.
[30-Jul-2009 10:21:14] <rmatte> TBKDan: yeh, that's normal, ok...
[30-Jul-2009 10:21:16] <TBKDan> rmatte: Try a reindex again?
[30-Jul-2009 10:21:16] <jmp242> 2009-07-30 11:15:34 ERROR zen.Events: Error transforming EventClassInst snmp_authenticationFailure (Event instance has no attribute '1.3.6.1.4.1.9.2.1.5.0')
[30-Jul-2009 10:21:21] <rmatte> TBKDan: reindex and commit please
[30-Jul-2009 10:21:22] <rmatte>
[30-Jul-2009 10:21:32] <jmp242> I'm getting ZenHub heartbeat failures rather suddenly
[30-Jul-2009 10:21:37] <mrayzenoss> edwardam: so zope 2.11.2 will be in the Octoberish release and zope 2.12 will be in the Aprilish release
[30-Jul-2009 10:21:46] <TBKDan> rmatte: Immediate error on reindex...
[30-Jul-2009 10:21:56] <rmatte> TBKDan: eugh, pastebin please
[30-Jul-2009 10:21:59] <rmatte> lol
[30-Jul-2009 10:22:08] <TBKDan> rmatte: http://pastebin.com/m48ba8d02
[30-Jul-2009 10:22:09] <edwardam> mrayzenoss: I used/developed on zope since before 2.X ... Stopped around 2.6 or so.
[30-Jul-2009 10:23:18] <mrayzenoss> jb: saw your write-up, did you happen to see this before you wrote it? http://netilium.org/~mad/technotes/postfixstat/
[30-Jul-2009 10:24:01] <rmatte> hmmm, MemoryError, strange
[30-Jul-2009 10:24:22] <TBKDan> Yeah... the VM still has 100+MB of RAM free, plus almost all swap.
[30-Jul-2009 10:24:39] <rmatte> TBKDan: restart zendmd and try a reindex again, see what happens
[30-Jul-2009 10:25:14] <TBKDan> rmatte: heh, immediate error out the the corrupted cache again.
[30-Jul-2009 10:25:21] <TBKDan> with the*
[30-Jul-2009 10:25:26] <rmatte> restart again
[30-Jul-2009 10:25:34] <rmatte> I mean zendmd
[30-Jul-2009 10:25:41] <TBKDan> .. zenoss is now restarting
[30-Jul-2009 10:25:42] <rmatte> fire it up again, the errors seems to be changing each time
[30-Jul-2009 10:25:51] <rmatte> ah, ok
[30-Jul-2009 10:26:12] <straterra> Hmm..For some reason, a linux bridge only shows up as 10 meg..I'm getting lots of pages about bandwidth usage on it
[30-Jul-2009 10:26:46] <rmatte> straterra: probably the way it's configured?
[30-Jul-2009 10:27:03] <TBKDan> rmatte: Ok, I go tthe original error the first time, now I'm just getting the 'Five' warning. Three dmd restarts later, same result. Reindex?
[30-Jul-2009 10:27:06] <TBKDan> got the*
[30-Jul-2009 10:27:16] <straterra> rmatte: The bridge?
[30-Jul-2009 10:27:29] <rmatte> we can try a reindex but I have a feeling that it's going to die out again
[30-Jul-2009 10:27:31] <rmatte> try it anyways
[30-Jul-2009 10:27:38] <rmatte> if it does we'll need to take a different approach
[30-Jul-2009 10:27:41] <jb> mrayzenoss: hmmm
[30-Jul-2009 10:27:48] <jb> haha
[30-Jul-2009 10:27:49] <TBKDan> Yup, errored out on corrupted cache rmatte
[30-Jul-2009 10:28:09] <rmatte> TBKDan: are you able to run the command that you used to remove the Cisco pack, but do it for the httpmonitor pack instead?
[30-Jul-2009 10:28:11] <straterra> Well..can I turn off notifications for that interface?
[30-Jul-2009 10:28:12] <jb> mrayzenoss: no, I didn't
[30-Jul-2009 10:28:15] <TBKDan> rmatte: http://pastebin.com/m62d34a54
[30-Jul-2009 10:28:26] <straterra> ah, nvm..found it
[30-Jul-2009 10:28:33] <TBKDan> Lemme try
[30-Jul-2009 10:28:46] <rmatte> it's complaining about different corrupted records each time, very strange
[30-Jul-2009 10:29:14] <TBKDan> rmatte: I ran the command no errors, but commit() threw a memory error
[30-Jul-2009 10:29:25] <TBKDan> rmatte:
[30-Jul-2009 10:29:29] <TBKDan> http://pastebin.com/m15918c17
[30-Jul-2009 10:29:32] <jmp242> why would I get zenhub log entries if heartbeats are failing?
[30-Jul-2009 10:29:35] <TBKDan> Friggin FF...
[30-Jul-2009 10:29:38] <jb> mrayzenoss: my method is a bit different, I suppose..
[30-Jul-2009 10:30:51] <rmatte> TBKDan: yeh, this is really not looking good, you're sure you're completely clearing the cache files each time?
[30-Jul-2009 10:31:04] <TBKDan> rmatte: rm *.zec ?
[30-Jul-2009 10:31:44] <TBKDan> rmatte: http://pastebin.com/m2491aa04
[30-Jul-2009 10:31:45] <rmatte> stop zenoss, delete all .zec files, then start zenoss
[30-Jul-2009 10:32:00] <rmatte> the file should be in $ZENHOME/var
[30-Jul-2009 10:32:21] <TBKDan> Done.
[30-Jul-2009 10:32:45] <rmatte> I don't even have .zec files though
[30-Jul-2009 10:32:58] <rmatte> what version of Zenoss are you running?
[30-Jul-2009 10:33:06] <TBKDan> 2.4.2
[30-Jul-2009 10:33:13] <rmatte> hmmm
[30-Jul-2009 10:33:36] <rmatte> zenoss@lab01:/usr/local/zenoss$ find . -name "*.zec"
[30-Jul-2009 10:33:37] <rmatte> find: ./mysql/data/events: Permission denied
[30-Jul-2009 10:33:37] <rmatte> zenoss@lab01:/usr/local/zenoss$
[30-Jul-2009 10:33:38] <rmatte> nothing
[30-Jul-2009 10:33:53] <rmatte> you installed zenoss from rpm or something though right?
[30-Jul-2009 10:34:04] <TBKDan> Correct
[30-Jul-2009 10:34:16] <rmatte> ah, not sure why that would make a difference, but ok
[30-Jul-2009 10:34:40] <rmatte> so deleting the .zec file from /var and restarting, what's the status now?
[30-Jul-2009 10:34:56] <TBKDan> rmatte: When starting up zendmd, still that error
[30-Jul-2009 10:35:11] <TBKDan> rmatte: http://pastebin.com/m3265b8c7
[30-Jul-2009 10:35:54] <rmatte> eugh
[30-Jul-2009 10:36:00] <rmatte> I'm running out of ideas
[30-Jul-2009 10:36:44] <TBKDan> ... device list generates a "Unable to communicate with server" and list is blank
[30-Jul-2009 10:37:05] <rmatte> do a zopectl restart
[30-Jul-2009 10:37:09] <rmatte> then check the device list again
[30-Jul-2009 10:37:19] <TBKDan> Just did a whole zenoss restart
[30-Jul-2009 10:37:21] <rmatte> (is this a production box?)
[30-Jul-2009 10:37:25] <TBKDan> Yes.
[30-Jul-2009 10:37:28] <rmatte> yikes
[30-Jul-2009 10:37:47] <TBKDan> Nogo on the device list
[30-Jul-2009 10:37:51] <rmatte> eugh
[30-Jul-2009 10:38:03] <rmatte> I've never seen a zope db problem like this before, insane
[30-Jul-2009 10:38:22] <TBKDan> holy crap.  The backup that "timed out" yesterday actually looks like it finished.
[30-Jul-2009 10:38:25] <rmatte> do you have any backups of that box?
[30-Jul-2009 10:38:28] <TBKDan> haha
[30-Jul-2009 10:38:39] <rmatte> if you do, we can grab the zope db file from a backup and throw it on there
[30-Jul-2009 10:38:41] <TBKDan> it's sitting in that list, so I'm hoping it's good...
[30-Jul-2009 10:39:06] <TBKDan> Ok, I'll brb, I have something else that just bumped this... one sec
[30-Jul-2009 10:39:41] <rmatte> you need to copy $ZENHOME/var/Data.fs from the backup to the same spot on the prod box
[30-Jul-2009 10:39:54] <rmatte> stop zenoss before you do
[30-Jul-2009 10:40:07] <rmatte> then delete the .zec and start it up, and hopefully we'll have a working database again
[30-Jul-2009 10:40:26] <rmatte> copy over the Data.fs.index file as well
[30-Jul-2009 10:40:48] <jb> mrayzenoss: looks like there are a few postfix-snmp hackjobs..
[30-Jul-2009 10:42:12] <rmatte> you can also just restore the entire box if it's a full backup, up to you
[30-Jul-2009 10:44:08] <TBKDan> Ok, back for a bit
[30-Jul-2009 10:44:15] <rmatte> k
[30-Jul-2009 10:46:07] <TBKDan> Do the Backups from the Backups tab actually back up the Data.fs...?
[30-Jul-2009 10:46:27] <rmatte> I'm not sure, I don't use that system for backups
[30-Jul-2009 10:46:45] <TBKDan> -rw-r--r-- 1 zenoss zenoss 365298006 Jul 28 16:47 2009-07-28-20-46-56.fs
[30-Jul-2009 10:46:47] <TBKDan> Maybe....?
[30-Jul-2009 10:46:47] <rmatte> it should
[30-Jul-2009 10:46:51] <rmatte> then you need to use zenrestore
[30-Jul-2009 10:47:09] <TBKDan> Should zenoss be running for a zenrestore?
[30-Jul-2009 10:47:16] <rmatte> no
[30-Jul-2009 10:47:47] <TBKDan> Zenrestore is running
[30-Jul-2009 10:47:51] <rmatte> k
[30-Jul-2009 10:47:57] <rmatte> cross your fingers hehe
[30-Jul-2009 10:48:04] <TBKDan> My hands are crossed lol
[30-Jul-2009 10:48:11] <rmatte> what are you using for hosting vms?
[30-Jul-2009 10:48:14] <TBKDan> (Mostly because it's friggin freezing in this room)
[30-Jul-2009 10:48:19] <TBKDan> VMWare ESX
[30-Jul-2009 10:48:24] <rmatte> free edition?
[30-Jul-2009 10:48:37] <TBKDan> Well, using ESXi for this particular host, licensed though
[30-Jul-2009 10:48:45] <TBKDan> Most of our ESXi boxes are the free
[30-Jul-2009 10:48:53] <TBKDan> For now at least
[30-Jul-2009 10:49:02] <rmatte> ok, if it's licensed does that give you ability to clone vms on the fly?
[30-Jul-2009 10:49:21] <TBKDan> It would appear so
[30-Jul-2009 10:49:30] <TBKDan> (Using it through VC)
[30-Jul-2009 10:49:33] <rmatte> you should setup schedules cloning of that vm
[30-Jul-2009 10:49:35] <TBKDan> zenrestore finished
[30-Jul-2009 10:49:37] <rmatte> scheduled*
[30-Jul-2009 10:49:47] <rmatte> we clone all of our zenoss boxes at midnight every night
[30-Jul-2009 10:49:48] <TBKDan> Yeah, I need to get a better backup strategy for our linux boxes
[30-Jul-2009 10:49:55] <rmatte> that way we always have a 24 hour copy to restore from
[30-Jul-2009 10:49:59] <TBKDan> Clone or snapshot?
[30-Jul-2009 10:50:05] <rmatte> clone
[30-Jul-2009 10:50:09] <rmatte> snapshot is different
[30-Jul-2009 10:50:09] <TBKDan> Why not just snapshot?
[30-Jul-2009 10:50:22] <rmatte> because a snapshot stops the machine from running while it's being taken
[30-Jul-2009 10:50:31] <rmatte> basically freezes it up
[30-Jul-2009 10:50:41] <rmatte> the clones don't interfere with operation
[30-Jul-2009 10:51:06] <TBKDan> Does it freeze it up if you don't snapshot the running memory? (Haven't tried)
[30-Jul-2009 10:51:19] <rmatte> also, we clone them from the SAN to the local disks on each server, so that the clones are stored in a separate location
[30-Jul-2009 10:51:30] <rmatte> TBKDan: I believe it still does
[30-Jul-2009 10:51:48] <TBKDan> Hm.  Well when we snapshot, it's usually only unresponsive for a few seconds at most.
[30-Jul-2009 10:52:01] <TBKDan> Anyway, zenrestore is done
[30-Jul-2009 10:52:05] <rmatte> cool
[30-Jul-2009 10:52:10] <TBKDan> Anything else to do before I fire up zenoss?
[30-Jul-2009 10:52:12] <rmatte> fire it up and hopefully it'll be fixed
[30-Jul-2009 10:52:19] <rmatte> not that I'm aware of
[30-Jul-2009 10:52:31] <TBKDan> *drumroll*
[30-Jul-2009 10:53:16] <TBKDan> zendmd came up without that error
[30-Jul-2009 10:53:20] <TBKDan> Should I try a re-index?
[30-Jul-2009 10:53:24] <rmatte> ok, login to Zenoss
[30-Jul-2009 10:53:28] <rmatte> can you see the devices?
[30-Jul-2009 10:53:30] <TBKDan> Yup
[30-Jul-2009 10:53:32] <rmatte> no, don't reindex yet
[30-Jul-2009 10:53:37] <TBKDan> I see all devices
[30-Jul-2009 10:53:39] <rmatte> ok, load up the zenpacks page
[30-Jul-2009 10:53:44] <rmatte> are the cisco and httpmon listed?
[30-Jul-2009 10:53:53] <TBKDan> It's taking forever, so probably will be
[30-Jul-2009 10:53:54] <TBKDan> lol
[30-Jul-2009 10:54:00] <rmatte> lol, k
[30-Jul-2009 10:54:16] <rmatte> ok, so yeh, we want to reindex then
[30-Jul-2009 10:54:20] <rmatte> for starters
[30-Jul-2009 10:54:40] <rmatte> (good thing you have that backup, if we screw things up again we can restore)
[30-Jul-2009 10:54:46] <TBKDan> Yup
[30-Jul-2009 10:54:56] <rmatte> though hopefully we can fix this cleanly
[30-Jul-2009 10:55:00] <TBKDan> And man, our external monitoring service (to make sure the internal monitor stays up ) is going crazy
[30-Jul-2009 10:55:08] <rmatte> lol
[30-Jul-2009 10:55:36] <TBKDan> Still waiting for the zenpack page
[30-Jul-2009 10:55:41] <TBKDan> Should I wait for that to load
[30-Jul-2009 10:55:42] <TBKDan> Nvm
[30-Jul-2009 10:55:43] <TBKDan> It's up
[30-Jul-2009 10:55:48] <rmatte> hehe
[30-Jul-2009 10:55:51] <TBKDan> Only cisco mibs, no http monitor
[30-Jul-2009 10:56:01] <rmatte> ok, that's kind of a good thing hehe
[30-Jul-2009 10:56:04] <TBKDan> Hmm... i might not have tried to putz with the monitor at this point yet
[30-Jul-2009 10:56:06] <rmatte> reindex and commit for starters
[30-Jul-2009 10:56:12] <rmatte> if that goes cleanly, then we'll proceed
[30-Jul-2009 10:56:26] <TBKDan> Re-indexing...
[30-Jul-2009 10:57:12] <TBKDan> brb
[30-Jul-2009 10:57:15] <rmatte> k
[30-Jul-2009 10:59:30] <TBKDan> Back, commit time
[30-Jul-2009 10:59:37] <TBKDan> Done
[30-Jul-2009 10:59:59] <TBKDan> Re-loaded zendmd, no errors
[30-Jul-2009 11:00:05] <rmatte> good
[30-Jul-2009 11:00:12] <rmatte> ok, now execute the zendmd command for the cisco pack
[30-Jul-2009 11:00:16] <rmatte> and commit
[30-Jul-2009 11:00:43] <rmatte> was the httpmonitor zenpack installed during the time of that backup, or was that something you just tried today?
[30-Jul-2009 11:01:12] <TBKDan> Backup was from Tues night, I might have tried the http monitor yesterday
[30-Jul-2009 11:01:25] <TBKDan> I thought I had tried it earlier, but apparently I was mistaken
[30-Jul-2009 11:01:32] <rmatte> ok, so the backup shouldn't contain any httpmon stuff then
[30-Jul-2009 11:01:40] <rmatte> which is good, because now we can try to install it cleanly
[30-Jul-2009 11:01:48] <TBKDan> Well
[30-Jul-2009 11:01:53] <TBKDan> It didn't show up in the zenpack list
[30-Jul-2009 11:02:01] <TBKDan> But it shows up in $ZENHOME/ZenPacks....
[30-Jul-2009 11:02:08] <rmatte> ah
[30-Jul-2009 11:02:20] <TBKDan> (Still removing Cisco zenpack)
[30-Jul-2009 11:02:28] <rmatte> no biggie, hopefully we can get it installed properly after we finish with the cisco thing
[30-Jul-2009 11:04:05] <rmatte> brb, just grabbing some water...
[30-Jul-2009 11:04:44] <TBKDan> Gah, need to help with something else...
[30-Jul-2009 11:04:52] <TBKDan> (Still removing)
[30-Jul-2009 11:05:35] <rmatte> k
[30-Jul-2009 11:06:57] <TBKDan> committing...
[30-Jul-2009 11:09:03] <sergeymasushko> Is there a way to monitor ran processes on a server and draw his load graphs? e.g. I'd like to know how much percentage of CPU eaten by spamassasin ?
[30-Jul-2009 11:10:12] <rmatte> sergeymasushko: yeh, as long as you're monitoring the process via snmp it'll automatically draw those graphs for you
[30-Jul-2009 11:10:33] <rmatte> sergeymasushko: just monitor the process then click on the process name and you should see graphs
[30-Jul-2009 11:11:13] <sergeymasushko> oh... it's true... thanks
[30-Jul-2009 11:11:16] <rmatte> np
[30-Jul-2009 11:13:57] <TBKDan> rmatte: Done.
[30-Jul-2009 11:14:24] <rmatte> TBKDan: ok, now try installing the httpmonitor ZenPack again (feel free to make a backup of Zenoss beforehand if you'd like)
[30-Jul-2009 11:14:43] <rmatte> pastebin any errors you see
[30-Jul-2009 11:33:36] <sergeymasushko> I've opened template 'OSProcess' (Monitors for OSProcess object) and see there is no OIDs defined in the datasources... How zenoss draw RRD in this case?... Actually the problem is that zenoss show empty CPU load graphs for the processes... (mem and count are OK).
[30-Jul-2009 11:34:17] <rmatte> it grabs the values from a collector plugin or something
[30-Jul-2009 11:34:20] <rmatte> it doesn't hit specific oids
[30-Jul-2009 11:34:42] <rmatte> is the cpu graph showing up at all?
[30-Jul-2009 11:34:50] <sergeymasushko> ok... how can I check why it doesn't show cpu load?
[30-Jul-2009 11:35:06] <sergeymasushko> yes it showed but it is blank
[30-Jul-2009 11:35:06] <rmatte> first off, does the graph show up at all?
[30-Jul-2009 11:35:15] <rmatte> blank as in "nan" ?
[30-Jul-2009 11:35:29] <rmatte> or is there an error saying the RRD hasn't been created?
[30-Jul-2009 11:35:34] <rmatte> or are all the values 0?
[30-Jul-2009 11:35:38] <sergeymasushko> 0.00
[30-Jul-2009 11:35:56] <rmatte> how long have you been monitoring that particular process?
[30-Jul-2009 11:36:15] <sergeymasushko> few months...
[30-Jul-2009 11:36:34] <jb> is it using 0% really?
[30-Jul-2009 11:36:37] <sergeymasushko> I see data in the mem and count graphs correctly
[30-Jul-2009 11:36:50] <rmatte> sergeymasushko: zoom out on the graph, has it ever displayed more than 0?
[30-Jul-2009 11:37:26] <sergeymasushko> jb: no... it couldn't =))) it's spamassasin
[30-Jul-2009 11:38:18] <sergeymasushko> rmatte: max 0.04
[30-Jul-2009 11:38:29] <jb> oh good point
[30-Jul-2009 11:40:47] <rmatte> sergeymasushko: if you do a "top", what's it's utilization?
[30-Jul-2009 11:42:48] <al_x> anybody know where I can find documentation on configuring the HP Proliant DL server to send out temperature info via SNMP
[30-Jul-2009 11:43:02] <jb> do you have insight installed?
[30-Jul-2009 11:43:40] <al_x> no, just the hp-snmp-agents rpm
[30-Jul-2009 11:43:48] <rmatte> al_x: http://www.zenoss.com/community/projects/zenpacks/hptemperaturesensors
[30-Jul-2009 11:43:52] <jb> you will need to install insight
[30-Jul-2009 11:43:59] <al_x> rmatte: i have that installed
[30-Jul-2009 11:46:07] <al_x> wow insight is 371 mb?
[30-Jul-2009 11:46:22] <al_x> im pretty sure i had the temp sensor working already without installing insight...
[30-Jul-2009 11:47:43] <al_x> but then the oids disappeared from snmpwalk
[30-Jul-2009 11:56:04] <al_x> does that make any sense?
[30-Jul-2009 12:03:22] <rmatte> al_x: not really, but you might try rebooting the server, sometimes the microsoft snmp service goes wonky
[30-Jul-2009 12:04:46] <rmatte> though I'm pretty sure you need insight
[30-Jul-2009 12:12:42] <TBKDan> rmatte: Back (Sorry, had a crapload of things come up at once)
[30-Jul-2009 12:15:39] <TBKDan> Well, I installed the http monitor and now I see graphs, still waiting on data for them though.
[30-Jul-2009 12:18:44] <rmatte> no problem
[30-Jul-2009 12:18:50] <rmatte> right on
[30-Jul-2009 12:19:03] <rmatte> data takes about 3 polling cycles
[30-Jul-2009 12:19:14] <TBKDan> data!
[30-Jul-2009 12:19:15] <TBKDan> lol
[30-Jul-2009 12:19:20] <rmatte> If you don't see data in a bit I can help you with the config, since I've worked with it before
[30-Jul-2009 12:19:24] <rmatte> so it's working?
[30-Jul-2009 12:19:31] <TBKDan> it would appear that it is
[30-Jul-2009 12:19:36] <rmatte> perfect
[30-Jul-2009 12:19:37] <TBKDan> Let me try taking it down now to see
[30-Jul-2009 12:19:42] <rmatte> k
[30-Jul-2009 12:24:28] <TBKDan> So, I set the timeout to 15 and cycle time to 60, it took 3+ minutes longer than the normal zenoss port check to say that the http connection timed out
[30-Jul-2009 12:25:12] <rmatte> hmmm
[30-Jul-2009 12:25:19] <rmatte> ours was pretty fast as I recall
[30-Jul-2009 12:25:42] <rmatte> then again you should really use a combination of both checks to be 100%
[30-Jul-2009 12:25:45] <TBKDan> Still only one count too, even though it's past the 75 seconds it should have took...
[30-Jul-2009 12:25:56] <TBKDan> Yes, but if I want to monitor the content...
[30-Jul-2009 12:26:12] <rmatte> yeh, with regex, hmmm
[30-Jul-2009 12:26:15] <TBKDan> Maybe it's still using the default values?  Lemme try a re-model...
[30-Jul-2009 12:26:15] <rmatte> let me check our settings
[30-Jul-2009 12:26:20] <rmatte> k
[30-Jul-2009 12:27:15] <TBKDan> Ok, 2 count with first being 13:23, last being 13:26. Let's see how long the next one takes (just remodeled)
[30-Jul-2009 12:28:04] <TBKDan> Ok, that might be better.  Got the third @ 13:27:42
[30-Jul-2009 12:28:52] <TBKDan> Sweet
[30-Jul-2009 12:28:57] <TBKDan> Remodel did it
[30-Jul-2009 12:34:52] <TBKDan> So i think we're good to go now o.O
[30-Jul-2009 12:34:59] <TBKDan> Stuff seems to be working as expected from what I can tell lol
[30-Jul-2009 12:35:01] <brice> After I've imported the latest Dell MIBs in, is there anyway to associate the traps oids with events?
[30-Jul-2009 12:37:27] <TBKDan> Just gotta play with the regex expressions now
[30-Jul-2009 12:38:41] <rmatte> brice: they shouldn't be coming in with OIDs, they should automatically be translated if you have the right Mibs installed
[30-Jul-2009 12:42:34] <kisielk> okay, I *must* fix this quota monitoring problem
[30-Jul-2009 12:42:42] <kisielk> I need some suggestions as to how to tackle this
[30-Jul-2009 12:43:12] <kisielk> basically, on our main file servers we use the project quotas feature of XFS, which can be reported through xfs_quota
[30-Jul-2009 12:43:13] <brice> ahh nm...i guess i'd need openmanage on my dell server for it to send out snmp info about the cpu temp, etc
[30-Jul-2009 12:43:29] <kisielk> I have an SNMP module that exposes the information from xfs_quota
[30-Jul-2009 12:43:35] <kisielk> what I want to do
[30-Jul-2009 12:43:55] <kisielk> is track how much space is used by each project, what the current soft and hard quotas are, and have alerts when the quotas are reached
[30-Jul-2009 12:43:59] <kisielk> how do I do this?
[30-Jul-2009 12:58:00] <rmatte> kisielk: mrayzenoss has shown me some way to dynamically set quota values at one point but I've lost the bookmark, let me see if I can dig it up
[30-Jul-2009 13:00:24] <rmatte> kisielk: http://www.thetechchef.com/zenoss-snmp-min-max-update/
[30-Jul-2009 13:00:32] <rmatte> that's something along the lines of what you want
[30-Jul-2009 13:00:41] <rmatte> for the quota threshold portion anyways
[30-Jul-2009 13:01:43] <jb> rmatte: did you have a chance to skim my thread?
[30-Jul-2009 13:02:20] <rmatte> jb: no I haven't yet, mind passing the link?
[30-Jul-2009 13:02:26] <rmatte> I'll take a look since I have a second now
[30-Jul-2009 13:03:04] <jb> http://forums.zenoss.com/viewtopic.php?t=10503
[30-Jul-2009 13:04:38] <ckrough> jb: nice work
[30-Jul-2009 13:04:54] <jb> thanks.. let me know if something doesn't make sense
[30-Jul-2009 13:05:00] <jb> i use the same method to monitor BIND data
[30-Jul-2009 13:05:59] <TBKDan> Is there a way to correlate events such that if the *entire* machine is down, it won't alert me for every component on that machine that is down? (The IP is down, obviously the web server is down as well)
[30-Jul-2009 13:06:03] <kisielk> rmatte: okay, so that's one piece of it
[30-Jul-2009 13:07:58] <jb> TBKDan: give <jb> thanks.. let me know if something doesn't make sense
[30-Jul-2009 13:07:59] <jb> oops
[30-Jul-2009 13:08:08] <jb> TBKDan: give this a try.. http://blog.zenoss.com/2009/05/27/tip-of-the-month-event-suppression/
[30-Jul-2009 13:15:32] <rmatte> kisielk: beyond that, you'll need to find some way of passing the rest of the info over, I'm not sure exactly how you'd do it. Can it be polled with straight SNMP?
[30-Jul-2009 13:15:49] <TBKDan> Back
[30-Jul-2009 13:16:21] <kisielk> rmatte: yeah. Our SNMP module reports each property as an array
[30-Jul-2009 13:16:30] <kisielk> so there's an array of names, array of disk usage, array of soft quotas, etc
[30-Jul-2009 13:16:43] <kisielk> ideally Zenoss would do the following:
[30-Jul-2009 13:16:52] <TBKDan> jb: Thanks, I'll see how that works.
[30-Jul-2009 13:16:58] <kisielk> contact the server, get a list of projects
[30-Jul-2009 13:17:15] <kisielk> for each project, monitor the disk space usage, and automatically set thresholds based on the quotas
[30-Jul-2009 13:17:26] <rmatte> jb: that's pretty cool, I'm going to bookmark that for future use/reference
[30-Jul-2009 13:17:58] <kisielk> that way I get historical trends and alerting
[30-Jul-2009 13:18:14] <kisielk> and ideally I wouldn't have to add datapoints manually each time we add a new project
[30-Jul-2009 13:18:57] <rmatte> kisielk: the problem is that I don't think there's a way to dynamically add datapoints. If there is, it'd save me a TON of time with a lot of the stuff that I've setup
[30-Jul-2009 13:19:15] <kisielk> yeah, that's what sucks
[30-Jul-2009 13:19:19] <rmatte> kisielk: but to my knowledge, there isn't, so you'll probably have to do it by hand
[30-Jul-2009 13:19:26] <kisielk> although, they seem to do it dynamically for the ethernet interfaces
[30-Jul-2009 13:19:28] <kisielk> and filesystems
[30-Jul-2009 13:19:30] <jb> rmatte: thanks..
[30-Jul-2009 13:19:31] <rmatte> hopefully they everhaul the template system and make stuff like that possible eventually
[30-Jul-2009 13:20:32] <rmatte> dynamic datapoints, and the ability to have child templates inherit properties/values from parent templates would make them much more powerful
[30-Jul-2009 13:20:59] <kisielk> well, I wouldn't mind writing a Zenpack for this
[30-Jul-2009 13:21:08] <kisielk> if I just knew where to do it
[30-Jul-2009 13:21:18] <kisielk> I'm fairly proficient in Python, but the API is confusing as hell
[30-Jul-2009 13:23:35] <rmatte> yeh, well, the API is mostly Zope, and Zope makes eyes bleed
[30-Jul-2009 15:19:07] <rmatte> kisielk: for interfaces, disk, and all that, it's populated by a collector plugin, you'd have to look at some of the code for those to figure it out
[30-Jul-2009 15:52:07] <brice> hmm...for almost everything i try to get a performance graph on says it's missing an RDD file
[30-Jul-2009 15:55:51] <brice> er RRD*
[30-Jul-2009 15:56:47] <ckrough> brice: if your just getting started, did you give it a bit of time?
[30-Jul-2009 15:57:07] <ckrough> brice: rrdtool needs to get through 3 polling cycles (usually 300seconds) before it can generate any graph data
[30-Jul-2009 15:57:55] * ckrough goes home
[30-Jul-2009 15:57:58] ckrough is now known as ckrough_away
[30-Jul-2009 15:58:18] <brice> ahh
[30-Jul-2009 15:58:23] <brice> I didn't give it time...will do that
[30-Jul-2009 16:30:05] <mrayzenoss> a demo of the new Event Console is available: http://blog.zenoss.com/2009/07/30/zenoss-king-crab-event-console-preview/
[30-Jul-2009 16:33:34] <nzle> what do I need to configure so the gmap portlet shows my links. it shows my locations with a color representing if hosts are down there, but no lines between locations.
[30-Jul-2009 16:36:01] <nzle> ?
[30-Jul-2009 16:37:32] <mrayzenoss> the lines will be drawn between the locations if the network topology is complete
[30-Jul-2009 16:37:45] <mrayzenoss> so your routes need to have the link to the other location
[30-Jul-2009 16:38:11] <mrayzenoss> the easiest way to test it would be to use the Network Map
[30-Jul-2009 16:38:40] <mrayzenoss> if there's a route that crosses locations, it should show up on the map properly
[30-Jul-2009 16:39:05] <mrayzenoss> and you can add and delete routes on the OS tab
[30-Jul-2009 16:40:08] <rmatte> nzle: Click on Networks on the left, then go to the zProperties tab, make sure that zDrawMapLinks is set to True
[30-Jul-2009 16:43:12] <rmatte> the examples events for the new event console start sounding like fortunes from fortune cookies if you read them long enough
[30-Jul-2009 16:43:20] <mrayzenoss> heh
[30-Jul-2009 16:43:59] <rmatte> the nicest thing is the scrolling, it's not all wonky like the old one
[30-Jul-2009 16:44:48] <mrayzenoss> you can tell Ian was having fun: "I'm having trouble coming up with more event summaries."
[30-Jul-2009 16:44:54] <rmatte> yeh lol
[30-Jul-2009 16:47:14] <rmatte> Fortunes: "Zenoss sees a career change in your future", "If you act now, you will avoid significant expense" & "Something is horribly wrong"
[30-Jul-2009 16:47:23] <rmatte> I like how you can shift-select events
[30-Jul-2009 16:47:35] <rmatte> what's the "Suppress" option?
[30-Jul-2009 16:47:42] <rmatte> a quick way to suppress events?
[30-Jul-2009 16:48:11] <mrayzenoss> hmmm... Ian just added that
[30-Jul-2009 16:49:21] <rmatte> lol, is that "Demo Screens" section showing stock prices?
[30-Jul-2009 16:50:47] <rmatte> ooooh, the new popup dialogs are secksy
[30-Jul-2009 16:51:00] <rmatte> translucent and everything
[30-Jul-2009 16:53:48] <nzle> rmatte, zDrawmaps is on, but I just noticed that the Zenoss Networks isn't showing anyting at all. Thats about having discovered routes for everything starting with the zenoss server, right?
[30-Jul-2009 16:54:10] <rmatte> correct, it needs to have discovered routes between objects to be able to draw lines
[30-Jul-2009 16:54:42] <rmatte> let's say you only add servers in and no network equipment, it's not going to understand how each server connects to the other servers
[30-Jul-2009 16:54:46] <rmatte> you need to get as much in as possible
[30-Jul-2009 16:54:54] <mrayzenoss> "Suppress" is supposed to send you to the zProperties for that Event class, where you can either drop it or send it straight to history or change the severity... not sure what the planned behavior will be. Post it on the thread and I'll get Brandon to flesh it out better
[30-Jul-2009 16:55:15] <rmatte> k
[30-Jul-2009 16:55:32] <mrayzenoss> but I'm not sure if the plan is to go to the zProperites page or if it will be done via a floating div
[30-Jul-2009 16:56:42] <rmatte> just posted it
[30-Jul-2009 16:57:01] <rmatte> it'd be cool if it was done with a floating div, saves time
[30-Jul-2009 16:57:10] <mrayzenoss> yeah
[30-Jul-2009 16:57:36] <rmatte> you should try to design as much instant access to information without leaving the current page as possible
[30-Jul-2009 16:57:53] <rmatte> gives the user more of the feeling that they are using an actual application rather than a website
[30-Jul-2009 16:58:19] <mrayzenoss> yeah, that's the goal
[30-Jul-2009 16:58:28] <rmatte> It is definitely looking impressive so far though, the graphics design for the new UI is awesome
[30-Jul-2009 16:59:01] <rmatte> Is the core UI going to be the same colours and such, or are you following the same approach of skinning it differently?
[30-Jul-2009 17:02:01] <mrayzenoss> there will be a different color scheme for Core
[30-Jul-2009 17:02:06] <mrayzenoss> like there currently is
[30-Jul-2009 17:02:15] <rmatte> cool
[30-Jul-2009 17:02:17] <brice> Is it possible to put devices in different classes depending on certain values returned from SNMP during a discovery on a subnet?
[30-Jul-2009 17:02:36] <Disconnect> that'd be entirely too cool
[30-Jul-2009 17:02:45] <Disconnect> :_
[30-Jul-2009 17:02:48] <Disconnect> er,
[30-Jul-2009 17:02:54] <rmatte> there's no current built in way to do it, you'd have to write something
[30-Jul-2009 17:02:55] <mrayzenoss> here's an example of binding different templates: trac-zenpacks/browser/zenpacks/ZenPacks.community.HPSIMMonitor/ZenPacks/community/HPSIMMonitor/modeler/plugins/community/snmp/HPDeviceTemplateMap.py
[30-Jul-2009 17:03:18] <mrayzenoss> chooses which template to bind by what shows up in modeling
[30-Jul-2009 17:03:40] <mrayzenoss> not a frequently used technique, but it can work
[30-Jul-2009 17:04:04] <brice> nice, thanks
[30-Jul-2009 17:04:31] <mrayzenoss> hmm... Jane Curry had an example in one of her papers too
[30-Jul-2009 17:05:19] <mrayzenoss> look at this one: http://www.skills-1st.co.uk/papers/jane/auto_disco_paper.pdf
[30-Jul-2009 17:06:54] <rmatte> Jane's papers are amazing by the way, you should grab all of them and give them a read
[30-Jul-2009 17:07:11] <brice> thanks
[30-Jul-2009 17:07:47] <mrayzenoss> she's doing a Zenoss Event Management Workshop: http://www.zenoss.com/news/archive/zenossevent.2009-07-10.3131703299
[30-Jul-2009 17:08:34] <rmatte> yeh, I had a read through her site about a month ago, I can't afford to fly over to the UK though
[30-Jul-2009 17:08:51] <mrayzenoss> we're talking to Ohio Linux Fest about another free training day
[30-Jul-2009 17:09:00] <rmatte> nice
[30-Jul-2009 17:09:00] <mrayzenoss> and we're planning on one at LISA
[30-Jul-2009 17:11:01] <rmatte> you know about fossevents.org I assume?
[30-Jul-2009 17:13:54] <mrayzenoss> Yeah, we keep pretty busy as it is
[30-Jul-2009 17:13:57] <rmatte> LISA should be fun
[30-Jul-2009 17:14:09] <rmatte> I need to get myself a passport
[30-Jul-2009 17:16:43] <mrayzenoss> for the rest of the year we'll be at Atlanta Linux Fest, Red Hat Summit, Ohio LinuxFest, Central Pennsylvania Open Source Conference, LISA, FUDCON and the TX Python UnConference for Open Source events
[30-Jul-2009 17:17:13] <mrayzenoss> off the top of my head
[30-Jul-2009 17:17:41] <mrayzenoss> and Sales goes to other events
[30-Jul-2009 17:17:44] <mrayzenoss> like VMworld
[30-Jul-2009 17:24:54] <rmatte> cool
[30-Jul-2009 17:31:33] <rmatte> wow
[30-Jul-2009 17:31:35] <rmatte> http://www.forbes.com/2009/07/28/hackers-iphone-apple-technology-security-hackers.html
[30-Jul-2009 17:48:43] <konfoo> any screenshots of the new ui design
[30-Jul-2009 17:48:44] <konfoo> ?
[30-Jul-2009 17:49:14] <mrayzenoss> konfoo: http://public-demo.zenoss.com/evconsole
[30-Jul-2009 17:49:22] <konfoo> thx
[30-Jul-2009 17:49:36] <mrayzenoss> that will be in the next release, minus the top navigation
[30-Jul-2009 17:49:47] <konfoo> ah very nice
[30-Jul-2009 17:49:50] <mrayzenoss> replacing the current event console
[30-Jul-2009 17:50:02] <konfoo> good to see the icons are smaller allowing for more info to be packed onto the screen
[30-Jul-2009 17:51:06] <mrayzenoss> also finally fixed the color-blind issue
[30-Jul-2009 18:00:27] <kisielk> is there a way to make templates through zendmd?
[31-Jul-2009 00:00:46] [disconnected at Fri Jul 31 00:00:46 2009]
[31-Jul-2009 00:00:46] [connected at Fri Jul 31 00:00:46 2009]
[31-Jul-2009 00:00:56] [zenoss-logger (logger bot) has joined #zenoss]
[31-Jul-2009 07:15:37] ckrough_away is now known as ckrough
[31-Jul-2009 08:10:05] <brice> If I have a device being detected as something wrong...how can I correct it so that it effects all other devices of the same make/model?
[31-Jul-2009 08:16:29] <ckrough> Can you rephrase that? Im not following.
[31-Jul-2009 08:17:33] <brice> I have 5 Dell servers that are the same model...they're getting detected as a R200 when they're a 2950s. Where does it get the R200 from? I'd like to change it to 2950
[31-Jul-2009 08:20:26] <ckrough> ah. let me take a look at mine
[31-Jul-2009 08:20:54] <brice> thanks
[31-Jul-2009 08:25:20] <ckrough> brice: Im not sure, bit I think the DellDeviceMap zCollector plugin probably makes that call
[31-Jul-2009 08:27:58] <brice> that's what I'm using
[31-Jul-2009 08:27:59] <brice> hmmm
[31-Jul-2009 08:28:36] <ckrough> it looks like DellDeviceMap gets that info from snmp, which should be returning the model number that OpenManage claims
[31-Jul-2009 08:30:05] <ckrough> brice: snmpget one of the boxes for .1.3.6.1.4.1.674.10892.1.300.10.1.9.1
[31-Jul-2009 08:30:11] <ckrough> SNMPv2-SMI::enterprises.674.10892.1.300.10.1.9.1 = STRING: "PowerEdge 2950"
[31-Jul-2009 08:30:13] <ckrough> my 2950
[31-Jul-2009 08:30:50] <brice> odd lol
[31-Jul-2009 08:30:51] <brice> R200
[31-Jul-2009 08:31:03] <brice> maybe it is an R200 and i just forgot that this one box was
[31-Jul-2009 08:31:07] <ckrough> lol
[31-Jul-2009 08:31:08] <brice> we bought so many 2950s heh
[31-Jul-2009 08:31:26] <ckrough> zenoss:1, brice:0
[31-Jul-2009 08:31:29] <brice> lol yep
[31-Jul-2009 08:31:35] <brice> thanks for the help though
[31-Jul-2009 08:31:39] <ckrough> yw
[31-Jul-2009 09:41:39] <ckrough> lol/clear
[31-Jul-2009 09:41:47] <ckrough> wow. double wrong window.
[31-Jul-2009 09:42:40] <malbon> hi folks, the new mock up on the event view doesn't work on IE7, it looks horrid
[31-Jul-2009 09:43:00] <mrayzenoss> oh yeah?
[31-Jul-2009 09:43:09] <mrayzenoss> __ian__: you get that?
[31-Jul-2009 09:43:29] <__ian__> Yeah, i didn't test in IE for the last week.
[31-Jul-2009 09:43:43] <ckrough> Im getting the impression that this new interface will have keyboard shortcuts, is that the case?
[31-Jul-2009 09:43:56] <__ian__> Is it more than just the status icons repeating?
[31-Jul-2009 09:43:59] <__ian__> ckrough: yes.
[31-Jul-2009 09:44:05] <ckrough> __ian__: sweet
[31-Jul-2009 09:44:31] <ckrough> oooh... keyboard shortcuts for saved event console configs.. that would be useful
[31-Jul-2009 09:44:40] <__ian__> Currently in place we have minimal ones. Enter to activate event detail, esc to collapse it, and arrow keys (and jk, for vimmers) scroll the table
[31-Jul-2009 09:46:48] <chudler> LOVE the "log" feature :-)
[31-Jul-2009 09:47:19] <__ian__> Yeah, we were pretty pleased to have an opportunity to bring that and custom event views up to the front. They're totally hidden now.
[31-Jul-2009 09:47:22] <ckrough> log feature?
[31-Jul-2009 09:47:42] <ckrough> oh yeah, that. awesome
[31-Jul-2009 09:48:05] <ckrough> we actually use that frequently in the current release, bringing it to the front will be great
[31-Jul-2009 09:48:23] <malbon> It puts ticks all over the page, want a screenshot?
[31-Jul-2009 09:48:30] <__ian__> I've got one.
[31-Jul-2009 09:48:36] <malbon> ok cool.
[31-Jul-2009 09:49:21] <malbon> looks nice in firefox, are more screens going to be shown in the same way?
[31-Jul-2009 09:49:59] <__ian__> When we get to them, yeah. Only the event console will be in King Crab tho.
[31-Jul-2009 09:50:07] <__ian__> And all the navigation at the top of the screen won't be included.
[31-Jul-2009 09:50:26] <__ian__> Not in this release, that is.
[31-Jul-2009 09:50:38] <malbon> aha, shame, but understandable.
[31-Jul-2009 09:50:41] * ckrough loves the nav example, saves my precious screen space
[31-Jul-2009 09:51:34] <__ian__> Yeah, it'd be sweet, but we'd have to skin and rearrange the entire app.
[31-Jul-2009 09:52:01] <__ian__> We have many more beautiful screens planned that will ease that, but it'll be a somewhat gradual process.
[31-Jul-2009 09:52:30] <malbon> love some of the test data. it'd be very cool if it could be reskinned as we have users complaining about the look/feel at present.
[31-Jul-2009 09:52:48] <ckrough> my guys just want fast fast fast
[31-Jul-2009 09:53:04] <ckrough> pretty only counts for a few days
[31-Jul-2009 09:53:30] <__ian__> it'll be much faster as well.
[31-Jul-2009 09:54:07] <malbon> we got fast after chet hexed our system with an LDAP cache, it's quite fast now.
[31-Jul-2009 09:54:44] <ckrough> yeah, were running that as well, I wasnt involved. I assume it's a chetification
[31-Jul-2009 09:55:56] <malbon> anyhow hometime, enjoy all.
[31-Jul-2009 09:56:08] <ckrough> enjoy
[31-Jul-2009 09:58:41] <__ian__> So, the only thing that's a bit slow in the new UI at the moment is the collapse/expand of the detail panel, and only in firefox. in IE and Safari it's immediate.
[31-Jul-2009 09:58:53] <__ian__> I'll be working on that.
[31-Jul-2009 10:02:21] <ckrough> __ian__: check out zillow.com, they use a similar design, and not well
[31-Jul-2009 10:16:16] <kisielk> what log feature?
[31-Jul-2009 10:21:06] <ckrough> you can log comments to an event
[31-Jul-2009 10:21:18] <ckrough> its kinda buried in the current release
[31-Jul-2009 11:56:15] <cwj> is there a means to copy data sources from one performance template to another? nothing is jumping out at me in the UI or the documentation.
[31-Jul-2009 12:07:16] <nzle> I have a dlink switch that is continuously throwing 'snmp agent down' events despite the fact that the snmp community is set, the snmp version is set, and I am able to snmpwalk it, and the interfaces are detected during model. What am I missing here?
[31-Jul-2009 12:12:07] <mrayzenoss> cwj: you can copy templates from beneath your current position in the hierarchy by going to the page menu and selecting "More->All Templates"
[31-Jul-2009 12:12:29] <mrayzenoss> cwj: so for instance if you do this at /Devices you can see every single template in the system this way
[31-Jul-2009 12:12:44] <mrayzenoss> nzle: perhaps the snmp collections are timing out?
[31-Jul-2009 12:13:15] <mrayzenoss> bbiab
[31-Jul-2009 12:13:45] <venturaville> mrayzenoss: odd question ... in the custom device reports ... is the query field something that is basically a long if statement or can it be a series of statements where only the last statement returns a boolean?
[31-Jul-2009 12:14:21] <cwj> mrayzenoss: that sounds like what i need. thanks very much.
[31-Jul-2009 12:21:03] <mrayzenoss> venturaville: take a look at trac-zenpacks/browser/zenpacks/ZenPacks.sodonnell.MACReport/ZenPacks/sodonnell/MACReport/reports/Device_Reports/MAC_Addresses.rpt for an example
[31-Jul-2009 12:27:48] <nzle> mray:  how would I determine if thats the case?  modeling is going ok...
[31-Jul-2009 12:29:18] <nzle> If I snmpwalk without a -Cc option it stops midway with an 'Oids not increasing' error. Could this be the problem, and how/where would I deal with that.
[31-Jul-2009 12:30:36] <nzle> Oops, mray is gone.  Anyone else??
[31-Jul-2009 12:33:48] <nzle> mray:  how would I determine if thats the case?  modeling is going ok...
[31-Jul-2009 12:33:55] <nzle> If I snmpwalk without a -Cc option it stops midway with an 'Oids not increasing' error. Could this be the problem, and how/where would I deal with that.
[31-Jul-2009 12:39:01] <zahna> on my zenoss box, should i run my web daemon as the web user or the zenoss user?
[31-Jul-2009 12:39:23] <zahna> is either way preferrable?
[31-Jul-2009 12:39:30] <mrayzenoss> nzle: so a standard snmpwalk times out?
[31-Jul-2009 12:39:45] <mrayzenoss> zahna: zenoss works best if everything runs under the 'zenoss' user
[31-Jul-2009 12:39:57] <mrayzenoss> zahna: there's no web daemon, there's zopectl
[31-Jul-2009 12:41:12] <zahna> mrayzenoss: you can probably tell i didn't actually set up my zenoss box then, eh?
[31-Jul-2009 12:41:30] <mrayzenoss> zahna: had my suspicions
[31-Jul-2009 12:41:39] <mrayzenoss> zahna: did you inherit an install or you're planning one?
[31-Jul-2009 12:41:51] <zahna> another admin here set it up and i need to get familiar with it
[31-Jul-2009 12:42:03] <zahna> we've set up a test box to evaluate it
[31-Jul-2009 12:42:16] <zahna> we're looking at migrating from nagios/cacti
[31-Jul-2009 12:42:18] <nzle> It shows lots of data but stops when it gets to some mibs that include ip addresses, and gives an error 'Oids not increasing'
[31-Jul-2009 12:43:49] <mrayzenoss> nzle: nothing on the zenoss forums, but google found this: http://www.linuxquestions.org/questions/linux-wireless-networking-41/snmpwalk-error-oid-not-increasing-159725/
[31-Jul-2009 12:44:08] <mrayzenoss> zahna: your zenoss install is probably under /usr/local/zenoss and is owned by the 'zenoss' user
[31-Jul-2009 12:44:17] <nzle> looks like this after several hundred lines: UDP-MIB::udpLocalAddress.10.10.30.254.64518 = IpAddress: 10.10.30.254
[31-Jul-2009 12:44:17] <nzle> UDP-MIB::udpLocalAddress.10.10.30.254.64515 = IpAddress: 10.10.30.254
[31-Jul-2009 12:44:17] <nzle> Error: OID not increasing: UDP-MIB::udpLocalAddress.10.10.30.254.64518
[31-Jul-2009 12:44:17] <nzle>  >= UDP-MIB::udpLocalAddress.10.10.30.254.64515
[31-Jul-2009 12:44:18] <zahna> apparently, it's in /opt
[31-Jul-2009 12:44:24] <mrayzenoss> zahna: that's fine
[31-Jul-2009 12:44:40] <zahna> but, yeah, owned by zenoss
[31-Jul-2009 12:44:40] <mrayzenoss> zahna: as root, just 'su - zenoss' and use the zenoss account to manage it
[31-Jul-2009 12:45:01] <mrayzenoss> zahna: so you can do things like 'zenoss status' and 'zenoss restart' and whatnot
[31-Jul-2009 12:45:17] <mrayzenoss> zahna: it may have an entry in /etc/init.d/ as well
[31-Jul-2009 12:45:28] <zahna> i'm sure this is in the docs, i'll RTFM
[31-Jul-2009 12:45:36] <mrayzenoss> if it's running, it will be on http://yourhost:8080
[31-Jul-2009 12:45:39] <zahna> i don't want to waste your time
[31-Jul-2009 12:45:59] <mrayzenoss> http://www.zenoss.com/community/docs/ and check out the "Getting Started With Zenoss" guide
[31-Jul-2009 12:46:20] <mrayzenoss> zahna: and we'll have a free webinar next Tuesday: http://forms.zenoss.com/getting-started
[31-Jul-2009 12:46:29] <zahna> oh yeah?  excellent.
[31-Jul-2009 12:46:44] <mrayzenoss> yeah, I'll be hosting, so not so excellent ;p
[31-Jul-2009 12:47:20] <zahna> heh, if you outline best practices, that would be great
[31-Jul-2009 12:47:51] <mrayzenoss> well, it's mostly a walkthrough the system, directed by Q&A
[31-Jul-2009 12:48:03] <mrayzenoss> but if you have specific questions we'll hit them
[31-Jul-2009 12:48:03] <zahna> that's awesome too
[31-Jul-2009 12:48:28] <mrayzenoss> and I'm here most days
[31-Jul-2009 12:48:50] <zahna> perfect
[31-Jul-2009 12:49:54] <zahna> btw, are there any plans to bring a zenoss package to non-x86 platforms, like say, debian sparc64?
[31-Jul-2009 12:50:07] <mrayzenoss> yeah, you can build from source
[31-Jul-2009 12:50:13] <mrayzenoss> I run Debian PPC
[31-Jul-2009 12:50:26] <mrayzenoss> and eventually we'll get into the distros
[31-Jul-2009 12:50:31] <zahna> i'm sure i can, but was hoping for a convenient prebuilt package
[31-Jul-2009 12:50:36] <mrayzenoss> someday
[31-Jul-2009 12:50:38] <nzle> mray, that link you gave is consistent with what I found. I can do an snmpwalk if I give the -Cc option. Is there a way to do this in zenoss since this device apparently needs it?
[31-Jul-2009 12:51:14] <mrayzenoss> nzle: I'll ask around, but we don't use walks, we pull oids or ranges
[31-Jul-2009 12:51:24] <zahna> are build dependencies listed in the docs?
[31-Jul-2009 12:51:37] <mrayzenoss> zahna: sadly, we package up most of them
[31-Jul-2009 12:51:48] <zahna>
[31-Jul-2009 12:51:48] <mrayzenoss> zahna: which isn't good practice
[31-Jul-2009 12:51:55] <mrayzenoss> but we're making progress
[31-Jul-2009 12:52:17] <zahna> zenoss does make an appliance, right?
[31-Jul-2009 12:52:21] <mrayzenoss> yes
[31-Jul-2009 12:52:46] <mrayzenoss> it's VMware, reportedly can be converted to Xen (if you care)
[31-Jul-2009 12:53:04] <zahna> oh, so no hardware?  i thought there was a 1u box that could be thrown in a rack
[31-Jul-2009 12:53:27] <mrayzenoss> I think we dropped the hardware option, lemme check
[31-Jul-2009 12:53:56] <nzle> I dont care if the walk works or not. My problem is that the device is showing snmp agent down despite the fact that its not. Modeling it works, and the snmpwalk user command works with the mod mentioned. which would indicate that the agent is not down.
[31-Jul-2009 12:54:35] <konfoo> is it just me or is wmi monitoring a complete clusterfuck
[31-Jul-2009 12:55:55] <cgibbons> it's a little less than complete in the Core version of the product
[31-Jul-2009 12:57:03] <konfoo> little less?
[31-Jul-2009 12:57:15] <konfoo> my point was that wmi in general... bothers me
[31-Jul-2009 12:57:44] <rmatte> WMI is known to be much better in the Enterprise version
[31-Jul-2009 12:57:45] <cgibbons> how come?
[31-Jul-2009 12:57:49] <rmatte> the core version has limited support
[31-Jul-2009 12:58:03] <rmatte> though there are 3 new WMI community ZenPacks coming out at some point
[31-Jul-2009 12:58:11] <mrayzenoss> zahna: the hardware appliance wasn't super popular, but will is available through Sales if someone really wants it
[31-Jul-2009 12:58:45] <konfoo> the number of ports and required admin account amongst other things
[31-Jul-2009 12:59:32] <konfoo> anyone here using dell openmanage on remote boxes?
[31-Jul-2009 12:59:33] <cgibbons> You can configure it without an Admin account but I don't find it to be worth the effort.
[31-Jul-2009 12:59:41] <konfoo> cgibbons: nod
[31-Jul-2009 13:00:00] <cgibbons> Ports are easy, you can ratchet that down to 1 or 2 if you really want.
[31-Jul-2009 13:00:19] <konfoo> im considering using snare instead
[31-Jul-2009 13:00:22] <cgibbons> I assume you're putting devices behind firewalls where it matters?
[31-Jul-2009 13:00:38] <konfoo> yeah naturally and blocking with ACLs on the borders
[31-Jul-2009 13:00:54] <mrayzenoss> konfoo: most of the Windows issues in Core are issues in Enterprise
[31-Jul-2009 13:01:04] <brice> If the OS Make is returning Unknown on a model of a Win2k3 R2 box...anyway I can set it to Microsoft without the model resetting it? I'm guessing the OS Make OID is defined somewhere...
[31-Jul-2009 13:01:08] <mrayzenoss> as far as configuring the end machine
[31-Jul-2009 13:01:17] <cgibbons> have you seen the using DCOM w/ firewalls doc from the wizards in Redmond?
[31-Jul-2009 13:01:44] <konfoo> cgibbons: no but i have it ratcheted down and working through a firewall
[31-Jul-2009 13:02:19] <cgibbons> it pretty much just gives you the rules it wants and some registry settings you can configure to limit which ports the windows boxes will give out to DCOM clients
[31-Jul-2009 13:02:58] <konfoo> nod i have a range of 10 ports for dcom connects assigned
[31-Jul-2009 13:03:27] <cgibbons> i haven't personally tried it, but you might be able to configure it to just stick with 135 the whole time
[31-Jul-2009 13:04:17] <konfoo> im using 135 and 4000-4009. works fine for querying
[31-Jul-2009 13:04:32] <konfoo> win2k8 x64 doesnt seem to play nice though
[31-Jul-2009 13:04:40] <cgibbons> huh
[31-Jul-2009 13:04:44] <konfoo> i assume its because the wmi packs are querying *32*
[31-Jul-2009 13:04:53] <rmatte> hmmm, one of my monitors just randomly shut off, grrr
[31-Jul-2009 13:05:04] <cgibbons> do you have access to it real quick? you can get some easy debug data that might shed some light.
[31-Jul-2009 13:05:21] <konfoo> sure
[31-Jul-2009 13:05:41] <cgibbons> try this from the zenoss box: wmic -d99 -U 'domain\user' //yourhost 'select name from win32_computersystem'
[31-Jul-2009 13:05:51] <cgibbons> it'll prompt for password and then give you a lot of really noisy low level debug data.
[31-Jul-2009 13:06:04] <cgibbons> look for "STRINGBINDING" in all of that data, if it gets taht far.
[31-Jul-2009 13:08:29] <konfoo> got a bunch of those
[31-Jul-2009 13:09:13] <cgibbons> does it get further past that at all? what you should see in the stringbindings is a list of IP addresses / hostnames and port numbers in the brackets. those should match the port range you configured.
[31-Jul-2009 13:09:13] <konfoo> and then a wmi error 0x80041003
[31-Jul-2009 13:09:28] <mrayzenoss> zahna: for my PPC build, I mostly just follow these instructions: http://dev.zenoss.org/trac/browser/trunk/inst/docs/INSTALL_DEB-ETCH.txt and replace the SVN instructions to pull from the latest stable
[31-Jul-2009 13:10:11] <cgibbons> what kind of user / domain configuration does that host have? is it in a domain?
[31-Jul-2009 13:10:53] <zahna> mrayzenoss: yow, do run that in a production environment?
[31-Jul-2009 13:11:02] <konfoo> standalone, no domain
[31-Jul-2009 13:11:09] <mrayzenoss> zahna: well, I run it at home
[31-Jul-2009 13:11:18] <cgibbons> ok cool, when you get into domains there are some more rules about the account permissions
[31-Jul-2009 13:11:24] <zahna> oh, ok.  that seems better for home.
[31-Jul-2009 13:11:39] <mrayzenoss> zahna: but there are a good number of FreeBSD users who run production environments from source
[31-Jul-2009 13:11:55] <zahna> well, it's probably already in ports and that's a different story
[31-Jul-2009 13:12:07] <mrayzenoss> zahna: nope, not in ports yet
[31-Jul-2009 13:12:16] <zahna> oh really?
[31-Jul-2009 13:12:29] <zahna> that surprises me
[31-Jul-2009 13:12:36] <mrayzenoss> yeah, as soon as we fix our build for recent toolchains, we'll be in ports
[31-Jul-2009 13:12:43] <konfoo> cgibbons: no firewall issues on this box since im bound to an internal nic
[31-Jul-2009 13:12:53] <mrayzenoss> our source build only works on RHEL and Debian as far as I know
[31-Jul-2009 13:12:58] <mrayzenoss> without patching
[31-Jul-2009 13:13:16] <mrayzenoss> everything else has newer tools that break our build
[31-Jul-2009 13:13:27] <mrayzenoss> we've got tickets for it
[31-Jul-2009 13:13:31] <cgibbons> interesting. we haven't had any isues with w2k8 64bit here. now how about service pack version? there were some pretty bad server-side memory leaks in vista & w2k8 until SP2 cameout.
[31-Jul-2009 13:14:01] <konfoo> this test box is running sp1
[31-Jul-2009 13:14:52] <cgibbons> okay, there is a hotfix you can install for SP1 that fixes the memory leak. that error code is kind of a generic server-side failure, so it might be worth a reboot to see if it goes away (icky) if you haven't done that yet. do you get that error code right away, or does it feel kinda like a timeout pause?
[31-Jul-2009 13:15:21] <cgibbons> they seem to create this memory leak every OS release, it's annoying (XP had it)
[31-Jul-2009 13:19:33] <konfoo> oh this 0x80041003 looks like a heap of fun to troubleshoot
[31-Jul-2009 13:21:46] <cgibbons> yep
[31-Jul-2009 13:22:08] <cgibbons> some quick things I'd do in any of those scenarios is see if you can do the same WMI query with WBEMTEST on the box itself, and then move to another Windows box talking to it, etc.
[31-Jul-2009 14:04:29] <brice> Could someone point me to a document about how to create a tab under a particular device class and have it display a few specific values pulled via SNMP?
[31-Jul-2009 14:19:42] <rmatte> read the developer's guide, it's pretty much the only documentation for stuff like that
[31-Jul-2009 14:20:00] <rmatte> and use the AIX ZenPack if you need something to review since it makes use of that sort of thing
[31-Jul-2009 14:28:04] <brice> thanks
[31-Jul-2009 14:28:16] <rmatte> np
[31-Jul-2009 14:41:51] <brice> an `snmpwalk -v1 -c mycommunity hostname` should return everything available, right?
[31-Jul-2009 14:53:01] <nzle> Is there a command line I can do to perform a snmp collection on a host to see if/how its breaking down? Not a walk, the actual collection that zenoss does.
[31-Jul-2009 14:54:52] <cgibbons> zenperfsnmp -d yourdevice -v10
[31-Jul-2009 14:55:02] <cgibbons> see if that gives you what you need
[31-Jul-2009 15:35:19] <brice> i think zenperfsnmp is the daemon, no?
[31-Jul-2009 15:35:45] <mrayzenoss> yeah, but you can do one-off calls to it as well.
[31-Jul-2009 15:36:09] <brice> ahh the above syntax is kicking back to me the usage
[31-Jul-2009 15:36:37] <mrayzenoss> try "zenperfsnmp run -v10 -d yourdevice"
[31-Jul-2009 15:36:45] <brice> ah
[31-Jul-2009 15:37:04] <brice> ty
[31-Jul-2009 16:04:08] <victor-> for some reason today zenoss isn't letting me login
[31-Jul-2009 16:04:57] <mrayzenoss> zenpass will allow you to reset you password
[31-Jul-2009 16:08:49] <victor-> It's like something has broken on the system: http://pastebin.com/m20e4e73c
[31-Jul-2009 16:09:21] <mrayzenoss> you run that as the 'zenoss' user?
[31-Jul-2009 16:10:35] <victor-> same errors when i do
[31-Jul-2009 16:12:24] <mrayzenoss> is zenoss running?
[31-Jul-2009 16:12:46] <victor-> yeah, definitely.  i get failures on the web login prompt.
[31-Jul-2009 16:13:20] <victor-> i'm curious about why the built-in python is getting a 'command not found' running 'import ...'
[31-Jul-2009 16:15:51] <mrayzenoss> is the $PYTHONPATH set?
[31-Jul-2009 16:16:22] <victor-> /usr/local/zenoss/python/lib/python2.4/site-packages:/usr/local/zenoss/common/lib:/usr/local/zenoss/zenoss/lib/python:/usr/local/zenoss/python/lib:
[31-Jul-2009 16:20:21] <victor-> ok, i did a /etc/init.d/zenoss-stack stop, but found a zenoss process still running and it was still responding via http.
[31-Jul-2009 16:21:47] <mrayzenoss> odd, kill it and do do a zenoss-stack start, then retry
[31-Jul-2009 17:29:54] <konfoo> anyone know the deal with 'oid .1.3.6.1.4.1.9600.1.1.1.1.15 is bad' errors? .. the oid works fine with snmpwalk
[31-Jul-2009 17:40:31] <mrayzenoss> I haven't finished this yet, but here's a sneak preview of the WMI Data Source: http://www.zenoss.com/community/projects/zenpacks/wmidatasource
[31-Jul-2009 17:41:17] <mrayzenoss> there's a Windows Perf and an IIS6 Perf ZenPacks that go with it, publishing all 3 hopefully tonight
[31-Jul-2009 17:41:22] <mrayzenoss> but now I gotta go
[31-Jul-2009 19:45:41] <konfoo> ahh i have finally made wmi my bitch
[01-Aug-2009 00:00:46] [disconnected at Sat Aug  1 00:00:46 2009]