Channels

Info

Credits

  • cheeser
  • joed
  • kinabalu
  • lunk

2010-03-08

#opennms 2010-03-09

2010-03-10

Nick Message Date
<markw78> when I select * from alarms I have like 12 of them (maybe 5 unique) but I only see 1 in my dashboard...? am I missing something? [00:51]
<markw78> alarm type maybe... [00:53]
<markw78> or because those are resolved alarms... and this one I see red still isn't [00:55]
*** poldii :myrmidia.tng.de joined the channel [00:57]
*** diq :70-36-142-191.dsl.dynamic.sonic.net joined the channel [01:48]
*** ramo102 :voip.gfdsa.org joined the channel [01:55]
*** Jamminjud :donobi-24-161.ra.donobi.net joined the channel [02:18]
*** derjohn_mob :213.238.45.2 joined the channel [02:27]
*** k-tr :DSL01.83.171.188.239.ip-pool.NEFkom.net joined the channel [02:30]
*** k-tr :DSL01.83.171.188.239.ip-pool.NEFkom.net joined the channel [02:33]
*** _indigo :212.218.65.209 joined the channel [02:46]
*** santoroj :91.215.166.4 joined the channel [02:46]
*** rmis :213.172.226.74 joined the channel [03:09]
*** ceu :195.60.129.16 joined the channel [03:25]
*** ml_cms113 :tmo-105-125.customers.d1-online.com joined the channel [05:02]
*** ilreds :gateway/web/freenode/x-wykhmxtajubzrjda joined the channel [05:02]
<ilreds> hi to all [05:03]
<rmis> hi ilreds [05:31]
*** ramo102 :voip.gfdsa.org joined the channel [05:32]
*** robinbowes :84-51-160-74.pambow882.adsl.metronet.co.uk joined the channel [05:54]
*** cpm :pdpc/supporter/active/cpm joined the channel [05:57]
*_snd has sold two opennmii installs today [06:27]
*** _indigo :2001:4cd8:1:0:21f:5bff:fef1:ed55 joined the channel [06:35]
<sfb> Nicely done my friend! [06:42]
<_snd> well, they called me and said "we need this", so i guess cred goes to the people that made this stuff [06:43]
*** agalue :200.109.15.87 joined the channel [06:44]
<_snd> the commies are baaack [06:45]
*** steak__ :fire.perspectix.com joined the channel [07:15]
<steak__> hello [07:15]
<steak__> a question... I have been looking through the OpenNMS documentation but I found no information re modifying the list of assets categories... is that possible at all? [07:16]
*** nmsmanjohn :nat/ibm/x-nqjrmayvjriigfqt joined the channel [07:45]
<cpm> _snd! [07:52]
*cpm needs commie stuff [07:52]
<nmsmanjohn> morning all [08:19]
<_indigo> steak__: it´s hard coded in $OPENNMS_SRC/opennms-webapp/src/main/Asset.java [08:19]
<RangerRick> making that more variable is a long-time desired feature that no one's had time to work on :) [08:21]
<steak__> ok [08:21]
<RangerRick> the goal is to eventuallly allow arbitrary tags on elements (nodes, interfaces, etc.) [08:22]
<steak__> I was thinking something like that, after browsing around a lot :P [08:22]
<RangerRick> but we haven't come up with an implementation of such yet :) [08:22]
<steak__> instead the other fields that show up in the asset editor are stored in the pg database, right? [08:22]
<RangerRick> yes [08:22]
<steak__> ok then maybe I can try to reuse a field there [08:23]
<steak__> 'cause I have quite a mixed environment [08:23]
<steak__> real server, vm servers, on multiple locations, et [08:23]
<steak__> etc etc [08:23]
<_snd> cpm: howdy [08:41]
<_snd> RangerRick: you're a developementer of there here contraption called opennms? [08:42]
<RangerRick> I've been known to break a line-of-code or two [08:44]
<_snd> RangerRick: to get my git to pull what is the 1.7 bleeding edge, what is the right concoction and encatation of a git command line? [08:45]
<RangerRick> if you're not in 'master' already, then "git checkout master" [08:46]
<RangerRick> then a "git pull" will grab the latest and greatest and apply it to your current branch [08:47]
<_snd> ah, i was just reading the cvs list and figured there was some distinction betweenmaster and rc/1.7.blah? [08:47]
<RangerRick> yes and no... rc/1.7.10 is a subset of master at this point [08:48]
<RangerRick> we've stopped having two regular branches for unstable development, and instead make a "release candidate" branch the week before release [08:48]
<_snd> ah k [08:48]
<RangerRick> nothing's allowed to go into it without a bug and being cherry-picked from master, no direct commits [08:48]
<RangerRick> 1.7.11 development continues on in master [08:48]
<_snd> ah ok [08:48]
<RangerRick> _snd: see: http://www.opennms.org/wiki/Development_Methodology_Rework_Specification [08:49]
<_snd> and as soon as 1.7.10 is out the door then nothing gets applied to it? [08:49]
<_snd> ~methodology is http://www.opennms.org/wiki/Development_Methodology_Rework_Specification [08:49]
<isaacwang> Okay, _snd. [08:49]
<RangerRick> the automation parts of that specification have not been done yet, but we're trying to follow the rest :) [08:49]
<RangerRick> _snd: well, we'll still make a tag, but yeah, the rc branch becomes dead at that point [08:49]
<_snd> k [08:50]
<RangerRick> might keep it in case we need a -2 of the .deb package because of a packaging error, or somesuch, but it will be a dead end as soon as release basically [08:50]
<_snd> so then tracking rc/1.6.10 and rc/1.7.10 makes sense, and when they are released to track 10+1? [08:50]
<RangerRick> tracking the rc branches doesn't make sense, no [08:51]
<RangerRick> just track master if your'e doing new development [08:51]
<RangerRick> if the rc branch is made, it's probably too late for your code to get integrated anyways [08:51]
*** darkneko :81-64-227-21.rev.numericable.fr joined the channel [08:51]
<RangerRick> and it'll have to wait a month :) [08:51]
<_snd> im not touching code, i just want to figure whats sensible to track to have whats similar to old 1.6-testing and 1.7-testing :) [08:51]
<RangerRick> ah, I see... [08:51]
<_snd> so i know what to build and use for various thins :) [08:51]
<RangerRick> well, "master" is the same as testing [08:51]
<_snd> i figuered by now you knew i hardly knew a true fro ma false? :p [08:52]
<RangerRick> the rc is like "1.6" or "1.7" used to be... hand-merged patches from 1.6-testing or master [08:52]
<_snd> k [08:53]
<_snd> i think i got it :) [08:53]
*** darkneko parted the channel [08:54]
<sfb> RangerRick: Where's fjay's jdbc collector/poller? [08:58]
<sfb> I'd like to check it out and work on it. [08:58]
<markw781> ooo [08:58]
<fjay> sfb: my poller should be in the repo [08:58]
<sfb> But in a branch or..? [08:59]
<markw781> jdbc collection would make me cheese :D [08:59]
<fjay> master [08:59]
<markw781> I wanna learn some java :( [08:59]
<sfb> It's KILLING me not having JDBC collection [08:59]
<markw781> we could do a LOT with it as well [08:59]
<sfb> And writing stupid vbscripts to dump stats to a web server for HttpCollector [08:59]
<markw781> ew [08:59]
<fjay> one could extend the poller to do some collection alla the memcached poller [08:59]
<markw781> more so then knowing some java, I wish I had dev skillz :( [08:59]
<markw781> ie. whats a method? whats a function? how does it all fit together... type of stuff [09:00]
<sfb> Or I could use the poller as the basic logic for the collector. (; [09:00]
<sfb> markw781: Only one way to get started. Head first. [09:00]
<sfb> (= [09:00]
<markw781> yah [09:01]
<fjay> the collector should be easy peasy [09:01]
<fjay> in theory :) [09:01]
<markw781> yah [09:01]
<fjay> but.. you have to deal w some issue that you would not have to deal w/ in terms of snmp [09:01]
<fjay> connection pooling and some other stuff [09:01]
<sfb> fjay: The hard part for WMI was putting the initial framework together. So once the poller polled I just had to struggle through the Collector interfaces. It was relatively simple once you figured out what you needed to implement. [09:02]
<fjay> ahh [09:02]
<fjay> well then it should be pretty easy to accomplish [09:02]
<sfb> brb tho meeting [09:02]
<sfb> I'm going to hunt down mhuot and make him help me get Eclipse and git set up on my laptop. [09:02]
<markw781> lol [09:02]
<fjay> sfb: i can help you w/ things, asthat would be an amazing feature to get [09:04]
<fjay> and can give you any help you need [09:04]
<markw78> I want to learn some java so I can write a poller for our app... right now we have soemone write an app to test our ap, then post to a website TRUE/FALSE if its up or down, then monitor that [09:10]
<RangerRick> you should come to my talk at the UCE, I'll be going through writing a simple monitor [09:14]
<markw78> uce? [09:14]
<markw78> my problem is tho I'm not a dev, most of your talk would go right over my head lol [09:14]
<RangerRick> http://www.adventuresinoss.com/?p=1257 [09:14]
<markw78> ahh nice [09:15]
<markw78> also the E part makes it unrealistic for me hehe [09:15]
*** nrml :204.145.64.58 joined the channel [09:24]
*** sartin :88-106-137-217.dynamic.dsl.as9105.com joined the channel [09:25]
*** plarsen :ip65-46-125-186.z125-46-65.customer.algx.net joined the channel [09:26]
<_snd> iDad! [09:30]
<nrml> anyone know if you can use the short hostname of a server in notifications.xml? %nodelabel% seems to only give me the fqdn...? [09:30]
<markw78> %nodelabel% should get you whatever its named in the DB / node list, no? [09:33]
<nrml> correct [09:33]
<nrml> and %nodelebel% in our case is generated from the fqdn [09:33]
<markw78> I don't think opennms knows abut anything other then the fqdn as far as the name of any node, not 100% sure tho [09:34]
<nrml> im trying to shrink notifications so the email subject is more readable from mobile email clients [09:34]
<markw78> rename the nodelabels [09:34]
<markw78> drop the domain off them... though it may revert after the next rescan (I don't know, I provision all my nodes) [09:34]
<nrml> yeah, CTO likes the full fqdn there [09:35]
*nrml shrugs [09:35]
<nrml> figured it'd likely not be possible [09:35]
<markw78> this pathOutage alarm won't go away in the dashboard =/ [09:40]
<sartin> _snd!: Booked my flight to Frankfurt today. [09:41]
<_snd> congrats :) [09:41]
<_snd> wünderbar! [09:41]
<_snd> sartin: when do you arrive? [09:54]
<sfb> fjay: Excellent. [09:56]
<sfb> fjay: I'm cleaning up my laptop and getting things like git installed. (; [09:56]
<sfb> fjay: I think there was an OGP email within the last couple weeks with how-to-get-crap-checked-out-and-building instructions. [09:57]
<RangerRick> the final version is at http://www.opennms.org/wiki/Eclipse_and_OpenNMS [10:06]
<markw78> that mouth picture is a little disturbing [10:10]
*** brozow :cpe-071-065-197-116.nc.res.rr.com joined the channel [10:13]
<_snd> bowso :) [10:13]
<sartin> _snd: thursday night. I have a rather full week that week. [10:31]
<_snd> ah k [10:32]
*** mhuot :pdpc/supporter/active/mhuot joined the channel [10:32]
<_snd> but dont tell mhuot [10:33]
*** _indigo_ :212.218.65.172 joined the channel [10:38]
*** m4rzito :unaffiliated/m4rzito joined the channel [10:39]
<m4rzito> hi [10:42]
<_snd> bon journo [10:43]
<sfb> RangerRick: Thank you [10:54]
<m4rzito> que tal _snd? [10:54]
<_snd> que? [10:55]
*** m4rzito parted the channel [10:56]
*** _indigo_ parted the channel [10:56]
<sfb> RangerRick: ping [10:58]
<RangerRick> sfb: pong [11:01]
<RangerRick> well, that was easy [11:04]
*RangerRick goes away again [11:05]
<RangerRick> ;) [11:05]
<_snd> hehehe [11:05]
<RangerRick> so by popular demand, I actually put all the 1.7.10-specific bugs closed in the 1.7.10 release notes [11:05]
<RangerRick> although most of the bugs were 1.6 and 1.7, so they went in the 1.6.10 notes [11:06]
<sfb> RangerRick: How do I get fjay's jdbc branch in git? [11:07]
<RangerRick> oh, he never merged it? um... generally "git checkout -b mybranchname origin/hisbranchname" [11:07]
<RangerRick> for some reason I thought he'd merged it to 1.7 at some point [11:07]
<sfb> I have no idea? [11:07]
<sfb> fjay: Where the heck is your code? [11:07]
<sfb> RangerRick: Also "Browse Source Code" on onms.org points to http://opennms.svn.sourceforge.net/viewvc/opennms/ [11:08]
<sfb> On the left hand bar. [11:08]
<RangerRick> heh, guess we need to fix that then :) [11:08]
<sfb> (= [11:08]
<sfb> Have you considered putting FCK Editor in to onms.org (assuming we're nowhere close to Confluence...) [11:09]
<RangerRick> I thought it was ./opennms-services/src/main/java/org/opennms/netmgt/poller/monitors/JDBCQueryMonitor.java [11:09]
<RangerRick> seeing how I've never heard of FCK Editor, no, I have not considered it [11:09]
<sfb> That looks right [11:09]
<sfb> The other one was the store procedure monitor. [11:09]
<sfb> FCK is MediaWiki's rich text editor of choice iirc [11:09]
<RangerRick> http://www.opennms.org/wiki/JDBCQueryMonitor [11:10]
<RangerRick> then why isn't it part of mediawiki? :) [11:10]
<sfb> It's not theirs? License? Because wiki purists think we should all punnish ourselves with the awful syntax? [11:11]
<RangerRick> isn't the whole point of wiki markup that you don't need a big fancy WYSIWYG editor? [11:11]
<RangerRick> I mean, I guess I can add it, just not real keen on maintaining/keeping up with plugins if I can help it :P [11:12]
<RangerRick> mediawiki is already a manually-maintained thing, which sucks [11:12]
<sfb> Confluence. [11:12]
<sfb> That's the correct solution to this problem. (; [11:12]
<sfb> Okay this is fjay's code. [11:13]
<RangerRick> well yeah, when I have time :) [11:13]
<sfb> It's in 1.7? [11:13]
<RangerRick> yeah, it has been for a few releases, I believe [11:13]
<sfb> git clone git://opennms.git.sourceforge.net/gitroot/opennms/opennms [11:13]
<sfb> git checkout -b 1.7 origin/1.7 [11:13]
<sfb> I'm thinking of continuing his work on the collector side. [11:13]
<RangerRick> although there's no 1.7 branch anymore [11:13]
<RangerRick> a la the development methodology thing I pointed to earlier :) [11:14]
<RangerRick> master *is* 17 [11:14]
<RangerRick> but yeah, that would be cool [11:14]
<sfb> Oh [11:14]
<sfb> I have to look further down in the page [11:14]
<RangerRick> git clone foo [11:14]
<RangerRick> cd foo [11:14]
<sfb> onms 1.7 and later. (; [11:14]
<RangerRick> git branch my-feature [11:14]
<RangerRick> git checkout my-feature [11:14]
<RangerRick> etc... [11:14]
<sfb> Is that what I should do? [11:15]
<RangerRick> then git checkout master; git merge --squash my-feature [11:15]
<RangerRick> git push [11:15]
<RangerRick> that's the general workflow [11:15]
<RangerRick> just clone, you're automatically on master [11:15]
<RangerRick> make a local branch and work on it 'till the feature's done [11:15]
<sfb> Jesus, now I'm thoroughly confused. [11:15]
<sfb> Is there a git2svn. (; [11:15]
<RangerRick> when it is, merge --squash it back to master, and push it upstream [11:15]
<sfb> I have to learn dvcs someday I suppose. [11:15]
<RangerRick> ah, I need to update the docs [11:16]
<RangerRick> gimme a second and then the wiki will have a good walkthrough [11:16]
<sfb> Can I work against master directly? [11:16]
<sfb> I mean if I'm just tweaking WMI or making a minor feature change? [11:17]
<RangerRick> if you're only going to do one commit, yeah... but branches are so cheap, there's no reason not to do one [11:17]
<sfb> Sorry to users who could care less about this, btw... [11:17]
<RangerRick> then "git merge --squash" turns it all into one nice little commit on master wherey ou can say "I added this nice feature, blah blah" [11:18]
<sfb> K [11:18]
<sfb> mhuot must not be at work today. [11:19]
<sfb> Slacker. [11:19]
<sfb> I wanted to talk to him abuot GSoC [11:19]
<RangerRick> OK, sfb, http://www.opennms.org/wiki/Developing_with_Git is updated with the proper stuff, and the "Recommended Workflow" is probably what you're looking for [11:22]
<sfb> Thanks! [11:26]
<sfb> I'll take a look at what I'm putting into WMI and see if it makes sense to branch. It may. [11:26]
<sfb> Changing namespaces may entail a minor API tweak... [11:26]
<RangerRick> wow: http://www.myfoxorlando.com/dpps/news/dpgoh-megan-mariah-barnes-crashes-while-shaving-bikini-area-fc-20100308_6447120 [11:26]
*** f0urtyfive :75.150.130.121 joined the channel [11:27]
<markw78> why hasn't this alarm gone green? :/ [11:28]
<sfb> hahah [11:28]
<markw78> I ack'd it [11:28]
<sfb> ACK! [11:28]
<sfb> remotes/origin/master is master? [11:29]
<_snd> its remotely the master or your origin [11:29]
<RangerRick> it's master on the opennms git repo [11:29]
<sfb> I have to clone first tho right? [11:29]
<RangerRick> your own master is based on it, I just like that syntax 'cause it's easier to track upstream [11:30]
<RangerRick> yeah, start from the top [11:30]
<RangerRick> there's some suggested setup stuff that's a good idea [11:30]
<sfb> Okay. [11:30]
<sfb> Soon I'm going to have to learn Mercurial too since that's what Google Code uses.. [11:31]
*sfb also installs TortoiseGit [11:32]
*** lavermil :206.169.25.100 joined the channel [11:33]
<markw78> ah, theres no way to clean path outage alarms [11:33]
<markw78> cause theres no reduction key [11:33]
<markw78> who decided to make that an alarm :P [11:33]
<markw78> and acknowledge isn't good enough, have to clear it [11:33]
<lavermil> got one better for you markw78, wouldn't it be better to have the outage path defined for cricitcal path down like it does, but when it comes up not to send service/node up for everything beyond the critical path? [11:36]
<markw78> yah [11:39]
<markw78> does it? [11:39]
<markw78> I haven't noticed that [11:40]
<markw78> but I guess it makes sense since theres no way to have a reduction key for it I guess [11:40]
<markw78> the dashboard is nice, but the data on index.jsp is more useful :/ [11:40]
<sfb> RangerRick: git config push.default tracking [11:56]
<sfb> RangerRick: Is that necessary? [11:56]
<RangerRick> it simplifies things, but not necessary, no [11:57]
*** steak__ :fire.perspectix.com joined the channel [11:57]
<sfb> So I should do it. (= [12:00]
<sfb> So here goes [12:13]
<sfb> First new branch. [12:13]
<RangerRick> WAIT! [12:14]
<RangerRick> no, just kidding... go for it [12:14]
*** Lin :unaffiliated/lincity joined the channel [12:17]
<sfb> haha [12:19]
<lavermil> markw78: the dashboard is completely useless for my purpose since we really need the red light green light setup. [12:26]
<lavermil> and I need an audible alarm for an outage, which is easily done with index.jsp [12:26]
<lavermil> ;) [12:26]
*** davsharyan33 parted the channel [12:43]
*** davsharyan33 :195.88.67.37 joined the channel [12:49]
*** _indigo :ip-94-79-151-40.unitymediagroup.de joined the channel [12:50]
<lavermil> is it possible to configure a per service retry poll time period if the prior poll fails? So if I poll via ICMP every 30 seconds, if I fail to ping a ICMP service on a node, I may want to have ICMP retry interval be 1 second instead of the standard 30 seconds. [13:12]
<lavermil> eww...Any update on the status of Acknowledgement Daemon. That looks really interesting. [13:19]
<RangerRick> mmmm... lamb saag [13:23]
<lavermil> RangerRick: I think itis your webpage raccoonfink.com that I saw the post on. Any update on that wonderful looking daemon? [13:23]
<RangerRick> I was under the impression ackd works... http://www.opennms.org/wiki/Ackd [13:23]
<lavermil> looks nice. so this should be in 1.8 woot!! [13:25]
<lavermil> looks promising. now make it easy to configure using the GUI [13:25]
<lavermil> LOL [13:27]
<lavermil> http://www.raccoonfink.com/2009/04/09/many-faces-of-ranger-rick-animation.gif [13:27]
<RangerRick> =) [13:28]
<lavermil> good laugh [13:28]
<fjay> _indigo: i am going to give the report stuff a go now [13:29]
<_indigo> fjay: thanks i wish you luck ;) [13:29]
<lavermil> _indigo: someone was saying you might have had to produce a layer 2 outage report from the data contained in ONMS. Is this true? [13:30]
<_indigo> lavermil: sorry at the moment only layer-3 outages. [13:31]
<lavermil> darn [13:31]
<lavermil> currently a task of mine. :( [13:31]
<_indigo> lavermil: but we have a lot of stuff in the database, i´m ready to explore the possibilities :) [13:32]
<_indigo> lavermil: the snmp interface poller should help in that case a lot [13:32]
<lavermil> I was thinking about creating a snmpinterfaceoutage table [13:32]
<_indigo> lavermil: create snmp-poller an outage in the outage-table? [13:35]
<lavermil> ?? [13:40]
<_indigo> lavermil: i didn´t played very much with the snmp poller in the new unstable. [13:44]
<_indigo> lavermil: snmp interface poller sorry [13:44]
<fjay> _indigo: so I see the jfree chart error [13:47]
<_indigo> fjay: ok that sounds good maybe :) [13:48]
<fjay> Caused by: java.lang.NoClassDefFoundError: org/jfree/chart/StandardChartTheme [13:48]
<_indigo> have you upgraded to jfree-chart to 1.0.12 or leave it on 1.0.9 [13:48]
<fjay> its on 1.0.9 [13:49]
<fjay> let me try to upgrade it... [13:49]
<_indigo> fjay: thx that would be great [13:50]
<lavermil> _indigo: I haven't looked at it at all but was told it keeps track of interface changes, but not in respect to outages. [13:51]
<_indigo> lavermil: imho the only way to detect layer-2 outages? [13:52]
<lavermil> _indigo: that is already available in some ways. the only piece missing is polling and that is added with snmp poller. [14:00]
<lavermil> snmptraps for interface down/up is already kept in the events table. [14:00]
*** ml_cms113 :tmo-099-41.customers.d1-online.com joined the channel [14:04]
*** derjohn_mob :c189248.adsl.hansenet.de joined the channel [14:13]
*** mayday_jay :noon.nas.net joined the channel [14:40]
<_indigo> lavermil: mhh some statistics about layer-2 outages should no problem but we have no direct association between the link_down and link_up event like in outages. [14:52]
<_indigo> lavermil: that makes some things harder [14:52]
<fjay> _indigo: it looks like the report is running here... other than me not having SMTP working. [15:01]
*** mayday-jay :noon.nas.net joined the channel [15:02]
<_indigo> fjay: that would be nice is it safed in share/reports? [15:04]
<_indigo> so it looks like the only problem is the jfree-chart lib 1.0.9-1.0.12 issue [15:04]
<fjay> yup [15:04]
<_snd> y0 [15:04]
*** nmsmanjoh1 :nat/ibm/x-dmwibnvdaepppxek joined the channel [15:17]
<lavermil> _indigo: correct. if the SNMP poller could be tied then it would be true. no different than how a IP service is marked as down. [15:21]
*** mayday_jay :noon.nas.net joined the channel [15:21]
<_indigo> fjay: ok i give it a new chance on my machine ;) [15:24]
<fjay> lavermil: that can be done w/ the snmp stuff in 1.7 [15:24]
<fjay> lavermil: it has the notion of admin/oper status [15:24]
<fjay> if your interface is admin up and oper down I belive that could be considered an outage [15:24]
<_indigo> fjay: thats a good hint, the layer-2 outage report should then be possible. [15:25]
<markw78> fjay you did the jdbc sproc poller right? [15:25]
<lavermil> fjay: tarus mentioned that would be a place to start looking but said it would never be considered an outage. [15:26]
<fjay> markw78: no.. just the JDBCQuery stuff [15:27]
<fjay> lavermil: not in the traditional sense, no [15:27]
<fjay> lavermil: because there is no service tied to it. [15:27]
<fjay> lavermil: you could build a fake service around it.. perhaps [15:28]
*** nmsmanjoh1 parted the channel [15:28]
<fjay> as internal events about the state changes are generated [15:28]
<lavermil> fjay: wouldn't that request having a service for each OID that ties to a interface? [15:28]
<fjay> no [15:28]
<fjay> there is prob a way around that [15:28]
<fjay> but the easy way to do it would be to poll the status of the port (the oper status?) [15:29]
<fjay> and add 48+ services [15:29]
<fjay> and that would get you info when things break [15:29]
<fjay> the thing is the snmp layer2 stuff, from my understanding, does not generate an outage, just events of state change... maybe report off of those events? [15:30]
<fjay> and not off outages [15:30]
<lavermil> 48+ services are SNMP services tied to an OID for the ifindex...or how would you do this dynamically without writting a custom module for 1.7 [15:30]
<fjay> yeah tied to the oid [15:30]
<fjay> or possibly tied to the interface name [15:30]
<_indigo> fjay: ok if it helps i can figure this stuff out for reporting stuff in 1.7.x, have a ci...o 2960 at work, where i can play with. [15:33]
<lavermil> fjay: are you thinking this coudl be done with configuration changes or require custom piece of code? [15:33]
<fjay> lavermil: just config changes [15:33]
<fjay> i think [15:33]
<fjay> i know the snmpplugin and snmpmonitor have been improved [15:34]
<_indigo> fjay: i´ve seen the snmp-interface poller has pretty nice filter and rules for interface types, alias etc. [15:34]
<fjay> _indigo: what do you mean? [15:34]
<lavermil> how would you add the serivce tied to each OID for each ifindex dynamically for every host and keep it up to date? [15:34]
<fjay> _indigo: yeah, its just that it doesnt generate outage records [15:34]
<fjay> lavermil: add it to capsd and to pollerd and let capsd determine if the interface is there or not? [15:34]
<fjay> manually provision those nodes perhaps? [15:35]
<_indigo> fjay: ok I´ll give it a try the result will be more than nothing ;) [15:35]
<lavermil> if it populates a table like the regular outages table then it is easy to mine data from it. [15:35]
<fjay> _indigo: :) i didnt have any problems running the report after i upgraded jfreehart [15:35]
<fjay> lavermil: I think it just keeps current status, not state changes. [15:35]
<markw78> SOA SOAP Service Router - HTTP GET NOT Supported [15:37]
<markw78> :( [15:37]
<lavermil> I love this SQL. I need to convert this to a simple report. handy to know what kind of events you are generating the most. [15:37]
<lavermil> select eventuei, count(eventuei) from events group by eventuei order by count DESC; [15:37]
<markw78> so much for http poller :/ [15:37]
<_indigo> fjay: thats a good message, should i create a bugzilla enhancement with the morning-report stuff or is there a better location? [15:38]
<fjay> yeah, if you want to create one for the jfreechart issue you can.. i can see if there is time to get it into 1.7.10 [15:39]
<_indigo> fjay: thx for support [15:40]
<RangerRick> fjay: well, technically I'm building the first package right now [15:40]
<_indigo> fjay: create the bugzilla entries [15:40]
<RangerRick> it's not tagged yet ;) [15:40]
<RangerRick> (I hit a deb bug, so starting over anyways) [15:41]
<fjay> well this is a simple change from jfreechart 1.0.9 to 1.0.13 [15:41]
<markw78> someone asked about a SOAP poller back in 07 lol [15:41]
<RangerRick> yeah, that should be sane enough; push it to master and I'll cherry-pick it [15:41]
<fjay> i am running tests now to make sure it doesnt blow anything up [15:41]
*** isaacwang :carboy.networkgeek.org joined the channel [15:52]
<sartin> _indigo: so you got your report into the release? [15:59]
<_indigo> fjay: helped me to get this stuff running, we had some jasperreport snafus with jfree-chart lib versions but we see a light at the end of the tunnel ;) [16:01]
<sartin> _indigo: that's great! [16:01]
*** diq :shiva.mochimedia.net joined the channel [16:02]
<fjay> _indigo: we might have to punt on this fix till the next release... it might have broken something else. [16:02]
<sartin> _indigo: I wonder if you could get the start and finish times as parameters to the report? [16:02]
<_indigo> fjay: i guess the charts should be well tested, i guess, don´t know if there is another part of opennms which use jfree-chart [16:03]
<_indigo> sartin: what do you mean exactly the current report playground has no parameters at the moment [16:04]
<sartin> the report API stuff I'm working on can do that. [16:06]
*** Prim3 :64.61.40.162 joined the channel [16:06]
<sartin> So imagine you have a couple of shifts, day and night, and you want a report for each... [16:07]
<sartin> or you want to run a report for yesterday, because you forgot to schedule it. [16:07]
<sartin> I'm not familiar with the way you get JasperReports to spit out a map of it's parameters then feed it back at runtime, but I'm sure it can. [16:09]
<fjay> _indigo: so I don't think that fix is going ot make it into 1.7.10 [16:09]
<_indigo> sartin: that sounds pretty nice. can´t wait to get this stuff running on your api ;) [16:09]
<fjay> sartin: its pretty simple to do that... we can setup a call sometime to chat? I am working on a new branch to let your webflow stuff 'talk' to reportd [16:10]
<sartin> yeah, well, hours in the day and everything. Pushing to have something to demo at the conference. [16:10]
<_indigo> sartin: you have to put a hashmap to the runAndRender method the rest is a job in the report we´ll got it ;) [16:10]
<sartin> fjay: yeah, that would be great. [16:10]
<ml_cms113> Hi, building from source fails because jfreechart 1.0.13 not available in maven repos [16:12]
<ml_cms113> [INFO] skip non existing resourceDirectory /source/opennms/dependencies/opennms-jasper/src/main/resources [16:12]
<ml_cms113> Downloading: http://repo.opennms.org/maven2/jfree/jfreechart/1.0.13/jfreechart-1.0.13.pom [16:12]
<ml_cms113> [INFO] Unable to find resource 'jfree:jfreechart:pom:1.0.13' in repository opennms-repo (http://repo.opennms.org/maven2) [16:12]
<ml_cms113> Downloading: http://jasperreports.sourceforge.net/maven2/jfree/jfreechart/1.0.13/jfreechart-1.0.13.pom [16:12]
<ml_cms113> [INFO] Unable to find resource 'jfree:jfreechart:pom:1.0.13' in repository jasperreports (http://jasperreports.sourceforge.net/maven2) [16:12]
<ml_cms113> Downloading: http://repo1.maven.org/maven2/jfree/jfreechart/1.0.13/jfreechart-1.0.13.pom [16:12]
<_indigo> fjay: btw. fjay whats your theme on the conference? i would like do the things around the reporting things, configuration reportd and how to create reports with ireport and opennms. [16:12]
<fjay> _indigo: mine is a followup to what RangerRick is talking about really... [16:13]
<sartin> _indigo: exactly. The webflow grabs the parameters, collects your input, then spits them back out into the report at runtime. [16:13]
<fjay> showing how to write a poller... and use the BSFmonitor [16:13]
<sartin> _indigo: my slot content depends on how far the three of us get before then ;-) [16:13]
<_indigo> :) [16:14]
<Prim3> hey all, was wondering if anyone could help me out with a notification problem. I had notifications setup to send me an email and text message, but recently switched service providers from att to verizon. I've changed the pager email from xxxXXXxxxx@txt.att.net to xxxXXXxxxx@vtext.com and now i'm getting an error stating invalid address. Previously everything was working fine. [16:14]
<Prim3> Also phone # hasn't changed, same one. [16:14]
<_indigo> sartin: ok so we should run in no conflicts in the conference workshops :) [16:14]
<sartin> would be nice if I could talk a little about running reports, and you could follow up on writing them. [16:16]
<_indigo> i guess it was a crazy but good decision to run in two ways for reporting implementation :) [16:16]
<sartin> _indigo hours in the day. My work is all spare time, and fjay had deadlines to meet. [16:16]
<sartin> Also I was trying to tidy up some existing code. [16:17]
<sartin> so we made a tactical decision to run separately and try to converge later. [16:17]
<_indigo> sartin: yes would be nice if you can talk about behind the scenes reporting api. [16:20]
<sartin> _indigo: That's the plan. [16:21]
<_indigo> sartin: yes the fast forward with reportd and a long term perspective is pretty pretty nice. [16:22]
<fjay> _indigo: you were running .12 of jfreechart right? and that was giving you that weird 'null' ? [16:23]
<_indigo> fjay: yes [16:23]
<fjay> and it works fine w/ .13? (it did for me.. what about you?) [16:23]
<_indigo> checked out the current git and bring all up to the newest version [16:24]
<fjay> it won't build right w/ .13 (tl1d and the webapp complain) [16:24]
*** santoroj1 :genkt-050-056.t-mobile.co.uk joined the channel [16:29]
<santoroj1> join /#solaris [16:30]
*** santoroj1 parted the channel [16:30]
<_indigo> uhm ? [16:35]
<_indigo> join /#solaris? [16:35]
<RangerRick> his / moved quite a bit [16:37]
<_indigo> ah ok, thought the hole time what he mean with this question [16:37]
<phyce> grah, solaris sucks, and now I have proof :) [16:40]
*** brozow :adsl-074-239-161-165.sip.rmo.bellsouth.net joined the channel [16:54]
<_snd> bowso :) [16:54]
*** sartin :88-106-137-217.dynamic.dsl.as9105.com joined the channel [17:10]
*** ml_cms113 :tmo-043-152.customers.d1-online.com joined the channel [17:24]
*** f0urtyfive :75.150.130.121 joined the channel [17:27]
<diq> I'd just like to state how much ass xfs_copy kicks in comparison to rsync [17:27]
<diq> that is all [17:27]
*** mhuot :pdpc/supporter/active/mhuot joined the channel [17:27]
<diq> 1.6.10 not released yet? [17:57]
<RangerRick> rpms are indexing now, source is on the web site [17:57]
<RangerRick> debs are up [17:57]
<RangerRick> still working on the 1.7.10 stuff, and the standalone installers [17:57]
<fjay> RangerRick: i think i fixed the problem.. it looks like 1.0.13 doesnt have the jcommon stuff w/ it.. or its not a dependancy or something [18:02]
<fjay> adding that to my pom.. and building now.. [18:03]
<RangerRick> ok, well, it'll be in 1.7.11 :) [18:03]
<fjay> yup! [18:03]
*** jimmah :ubuntu/member/pak33m joined the channel [18:07]
*** pak33m :ubuntu/member/pak33m joined the channel [18:08]
<RangerRick> alright, debian packages all done [18:21]
<RangerRick> just working through jar installer and rpms [18:21]
<lavermil> ok. how can ONMS get ifindexes totally confused [18:29]
<lavermil> it thinks two IP address belong to the same ifindex [18:29]
<lavermil> I verified the ifindex on the router and it is not like it shows. UGH. [18:29]
<eshelbk> it happens. there are a variety of fun things that can get ugly in the db.. have to make queries to find them when you run across them [18:30]
*** pak33m :ubuntu/member/pak33m joined the channel [18:31]
<lavermil> better yet the DB is correct the WebUI is wrong. [18:32]
<lavermil> WTF. [18:32]
<lavermil> DB = device, but webUI is wrong. [18:32]
<lavermil> UGH. maybe logout and login...??? [18:32]
<eshelbk> that is beyond my powers [18:32]
*** brozow :cpe-071-065-197-116.nc.res.rr.com joined the channel [18:32]
<eshelbk> if I want to make a fairly large collection change, can I do something like SET ipinterface.issnmpprimary='C' where (conditionX and issnmpprimary='N'), and collections will automagically begin? [18:33]
<fjay> kinda sorta maybe [18:33]
<fjay> depends what you wanna do [18:33]
<eshelbk> i need to set physical serial ports to be collected if they aren't, so i can get error collection [18:33]
<lavermil> how are the webapp going to to know to start collection on the changes you make? [18:34]
<eshelbk> that's what i didn't know [18:34]
<fjay> it prob sends a signal to onms the daemon [18:34]
<lavermil> fjay: if you make a bulk change in the DB the onms daemons get a signal from the DB?.?.? [18:34]
<fjay> no [18:35]
<eshelbk> something would have to read the db [18:35]
<lavermil> ok. [18:35]
<fjay> best bet is to stop opennms.. make your change, start opennms [18:35]
<lavermil> I thought there was something magical there. hehe [18:35]
<fjay> and the better idea is to back it up [18:35]
<fjay> first [18:35]
<eshelbk> backups are for wimps! [18:35]
<fjay> agreed [18:35]
<fjay> :) [18:35]
<lavermil> I hate how ONMS pickups up Null0,Sit0,etc interfaces [18:35]
<lavermil> is there a filter to prevent those from showing up? [18:36]
<eshelbk> and six bazillion vlan interfaces as Secondary [18:36]
<eshelbk> much h8 [18:36]
<fjay> not that I am aware of, but I want to say 1.8 will let you do that. [18:36]
<fjay> i always just turn data collection on for every interface via the collectd configuration rather than in the DB [18:37]
<eshelbk> i have 150,000 interfaces :( [18:37]
<fjay> yeah... that might not play so nice :) [18:37]
<eshelbk> i want to build something to turn them on intelligently and automagically, but i'll do it by hand for now [18:37]
<fjay> you could make a package maybe.. w/ a filter rule to wilde card stuff.. [18:37]
<eshelbk> we're starting to tag ifAlias in a certain way and i'll probably key off that [18:38]
<fjay> yeah.. you could also hange the asset info and specify it in there.. [18:38]
<eshelbk> though turning on collection for everything would be an interesting test for the SAN.. [18:39]
<fjay> or cause you to get kicked in the face by the SAN admins :) [18:40]
<eshelbk> i've been watching our 9513 fabric switches- opennms and our security group's logging server are getting beat by other boxes now [18:41]
<lavermil> vote for this [18:41]
<lavermil> http://bugzilla.opennms.org/show_bug.cgi?id=3626 [18:41]
<eshelbk> hehe [18:42]
<eshelbk> my 10 votes are precious! [18:42]
*lavermil sobs [18:42]
<fjay> well the thing is [18:43]
<fjay> you could maybe [18:43]
<fjay> get outages [18:43]
<fjay> by umm [18:43]
<fjay> making services for each interface in your env... [18:43]
<eshelbk> hah [18:43]
<fjay> ie FastEthernet0/0 [18:43]
<fjay> etc [18:43]
<fjay> not like every FastEthernet0/0 [18:43]
<fjay> but just one [18:43]
<lavermil> Huh? [18:44]
<fjay> go with me here [18:44]
<lavermil> Create a service for an oid that points to Fa0/0, and then another one for fa0/1 [18:44]
<fjay> well not based on oid but based on the ifname [18:44]
<lavermil> that would be great if all OIDs point to the same thing for Fa0/0. [18:44]
<fjay> yeah [18:44]
<lavermil> yeah the ifname taken from the oid. [18:44]
<fjay> which is why you use ifname [18:44]
<fjay> no no.. you can create a service based on the ifname [18:45]
<fjay> not the oid [18:45]
<lavermil> fjay: I am all for this, if it will work. [18:45]
<fjay> it should work [18:45]
<lavermil> so I need to create something in capsd and poller? [18:45]
<fjay> its very similar to how the diskusagemonitor works [18:45]
<fjay> yes [18:45]
<lavermil> the only way I can see doing this and it is by doing a snmpPlugin and use the right oid and expect the vbvalue of Fa0/0..etc [18:46]
<fjay> i know the capsd snmp plugin can look up things in a table [18:47]
<fjay> thus.. the interface name [18:47]
<fjay> and i 'think' the pollerd monitor can do the same [18:47]
<lavermil> that would be a completely valid solution. [18:47]
<fjay> yeah.. complex to manage.. but [18:48]
<lavermil> pollerd meaning capsd? [18:48]
<lavermil> nm [18:48]
<lavermil> blah [18:48]
<fjay> worst case, an interface monitor would need to be created [18:48]
<fjay> which is not that hard to do [18:48]
<lavermil> what do you mean an interface monitor? [18:48]
<fjay> something for the pollerd side of things to look up interface status by name [18:48]
<lavermil> any idea where it would be documented on how to get capsd/pollerd to look at the table and match against the particular table? [18:49]
<fjay> the wiki? [18:49]
<lavermil> fjay: are you talking abotu new code for hte interface monitor bit? [18:49]
<fjay> yes [18:49]
<fjay> something might have to be written to do that if the default snmp monitor does not do it [18:50]
<lavermil> I don't know a like of java so i am kind of screwed. :( [18:50]
<fjay> come to the users conf.... i'll show you how to do it .. and or use it as my example for my workshop :) [18:50]
<eshelbk> beware, he likes to use the word "extend" a lot [18:50]
<eshelbk> and hetalksfast [18:50]
<fjay> its the speed [18:50]
<fjay> people are not this skinny unless speed is involved :) [18:51]
<lavermil> I would have to be able to know the index of the interface fa0/0 to be able to pull the oper status [18:51]
<fjay> yeah [18:51]
<fjay> and... the default snmp monitor 'might' do that [18:51]
<lavermil> not going to be possible via snmp plugin bit with out additional code or very static info. [18:51]
<fjay> yeah it will [18:51]
<lavermil> for cisco ifindex persist I can hope things don't change. [18:52]
<lavermil> hehe [18:52]
<fjay> i think something in 1.7 can do it [18:53]
<eshelbk> antonio's monitor has it worked out. what is it missing to populate outages? [18:53]
<fjay> antonio's code is not a monitor in the service up/down sense [18:53]
<lavermil> I will dig on that [18:53]
<eshelbk> it is a check/notify? [18:53]
<lavermil> fjay: any thought why the webui might report the wrong ifindex. [18:53]
<fjay> eshelbk: it just checks, permists and sends events [18:53]
<eshelbk> gotcha [18:54]
<fjay> it does not generate outages [18:54]
<lavermil> eshelbk: antonio's monitor....what is that? [18:54]
<eshelbk> the snmp interface monitor [18:55]
<lavermil> aahh [18:55]
<eshelbk> or poller, however it's worded [18:55]
<fjay> lavermil: give me like ~5 min [18:55]
<fjay> and i'll look [18:55]
<lavermil> ok. [18:55]
<eshelbk> it seems like your service would have to be a combo of ifadmin/ifoper, you should be able to a table of oid.ifindex off snmpinterface [18:56]
<fjay> so the swrun monitor is kinda what you need.. but... [18:57]
<fjay> for the interface [18:57]
<fjay> so yeah.. something would have to be written [18:57]
<lavermil> :( [18:57]
<fjay> it wouldnt be hard [18:57]
<lavermil> damn if I had only taken up java and went to school [18:57]
<eshelbk> something checks for an ip of 0.0.0.0 or null for a filter so you aren't duplicating ip monitoring [18:57]
<lavermil> if it were perl or php I would be in good shape. hehe [18:57]
<fjay> you might want to duplicate it? [18:58]
<eshelbk> if it were those, it wouldn't scale as well :) [18:58]
<fjay> *shrug* [18:58]
<fjay> you could write one in python [18:58]
<fjay> if i wasnt sick i would get you pointed in the right direction [18:59]
<fjay> but looking at code makes my brain hurt [18:59]
<eshelbk> get away from the glowy screen! [18:59]
<fjay> and i a trying to debug an issue [18:59]
<fjay> w/ reportd [18:59]
<lavermil> thx anyhow. [19:00]
<fjay> you really just need to check 2 oids and those oids need to be based on the ifindex that you lookup based on interface name [19:00]
<fjay> you need admin up and oper down [19:00]
<lavermil> really just oper [19:01]
<lavermil> but ideally both is good practice [19:01]
<eshelbk> true, i don't remember and admindown/operup [19:01]
<markw781> you making progress on your interface outage report? [19:01]
<fjay> well.. more than just oper [19:01]
<fjay> that way a port that shouldnt be up... [19:02]
<fjay> is not monitored [19:02]
<fjay> or am i wrong? [19:02]
<eshelbk> ah right, that's what is is [19:02]
<lavermil> markw781: I am able to write a simple perl script to generate one, but look at ways to have the right data when I need one. [19:02]
<lavermil> fjay: correct, admin is important for that reason for sure. [19:03]
<markw781> ah [19:05]
<lavermil> markw781: and most of what I want I can hack together via perl and procmail, but that is really HACK [19:07]
<lavermil> fjay: I can't see a way to use snmp_monitor to create a service unless I had a very static set of ifindexes. if the ifindexes never changed then I wouldn't be worried. [19:09]
<lavermil> now if ONMS had the option to call an external script for a monitor then I could certainly do this really easy [19:09]
<eshelbk> lavermil: that groovy script works pretty well for IP devices and it wasn't too hard to install and get running.. i don't know that much and it didn't seem hard to follow and modify to work for my devices/box [19:11]
<lavermil> eshelbk: is there a groovy plugin that can be used for this sort of thing? [19:11]
<eshelbk> i had to install the groovy rpm and download/copy the iText jar into the opennms/lib dir [19:12]
<eshelbk> you just run it as a script and it goes an interrogates the outages db and yanks data out of jrbs [19:12]
<eshelbk> counts up downtime over total time and calculates uptime [19:13]
<lavermil> I guess it is not in the yum tree [19:14]
<eshelbk> nope, i had to go to codehaus [19:14]
<lavermil> was there a repo available for codehaus [19:14]
<eshelbk> i don't remember seeing one, but i wasn't looking hard- groovy.codehaus.org [19:15]
<lavermil> just got there [19:15]
<lavermil> checking around [19:15]
<lavermil> not really sure what maven is but i always hear it mentioned with eclipse [19:15]
<eshelbk> the script does text formatting in the pdf with iText-5.0.1.jar, so you need to grab that [19:17]
<lavermil> iText 5.0.1? [19:18]
<lavermil> from iText PDF.com [19:18]
<eshelbk> yeah, it's a GPL'd PDF jar [19:18]
<eshelbk> otherwise, i learned not to use a dash in my script file name and set things up from the onms groovy wiki [19:19]
<lavermil> where did the iText need to be copied to? [19:19]
<eshelbk> opennms_home/lib [19:19]
<eshelbk> well anywhere, but your groovy startup has to find it somewhere [19:20]
<eshelbk> there's a load *.jar or something line [19:20]
<lavermil> what extension do groovy scripts end with? [19:20]
<eshelbk> nothing, they just need the !/bin/sh-like line at the top of the script [19:21]
<lavermil> your script had #!/usr/bin/env groovy [19:22]
<eshelbk> that's it [19:22]
<eshelbk> details, details :) [19:22]
<lavermil> did you assume the JAVA_HOME of the opennms stuff? [19:23]
<eshelbk> yeah, but i flailed a little and added some env variables too when i couldn't get things running [19:26]
<eshelbk> i ended up adding a JAVA_HOME and GROOVY_HOME to be sure, not sure if they were necessary [19:27]
<eshelbk> the rpm installed into /usr/share/groovy.. took a while to find it [19:27]
<lavermil> damn where does ONMS keep the JAVA_HOME setting....brain fart....ugh [19:28]
<eshelbk> i think that's why i added it [19:29]
<lavermil> rather jetty uses JAVA_HOME [19:31]
<eshelbk> there it is.. runjava looks for the env variable, and puts that path in onms_home/etc/java.conf [19:32]
<eshelbk> if it doesn't find the variable it tries other things [19:33]
<lavermil> where did you find that reference? [19:33]
<eshelbk> in the runjava script [19:33]
<lavermil> damn I can't it to run [19:40]
<lavermil> it keep bitching about JAVA_HOME and I have it set in my env [19:40]
<lavermil> do I need to set it in the JRE? [19:40]
<eshelbk> i didn't have to [19:46]
<eshelbk> the java binary is in my default path, not sure if that matters [19:46]
<lavermil> Exception in thread "main" java.lang.NoClassDefFoundError: org/codehaus/groovy/tools/GroovyStarter [19:47]
<lavermil> :( [19:47]
<lavermil> I guess it needs the GROOVY_HOME? [19:47]
<lavermil> eshelbk: where were you runnign this script from? your home directory? [19:49]
<eshelbk> possibly, i wasn't sure how it gets that stuff when it executes. i've got a root/.groovy/startup with JAVA_OPTS=-Dopennms.home=/opt/opennms and GROOVY_CONF=$HOME/.groovy/groovy-starter.conf in it [19:49]
<eshelbk> groovy-starter.conf is copied from the default groovy conf file and i added the opennms jar load line from the opennms groovy wiki page [19:50]
<eshelbk> i have the script in root's home dir and run it there [19:51]
<lavermil> does it have to run as root? [19:52]
<eshelbk> probably not, but then you need to be careful of your install and permissions [19:53]
*** brozow :cpe-071-065-197-116.nc.res.rr.com joined the channel [19:53]
<eshelbk> permissions on your jrbs and db [19:53]
<lavermil> org.codehaus.groovy.control.MultipleCompilationErrorsException: startup failed: [19:53]
<lavermil> /home/lvermilion/uptime_groovy: 9: unable to resolve class com.itextpdf.text.Document [19:53]
<lavermil> import com.itextpdf.text.Document; [19:54]
<isaacwang> 1.6.4, working on weather alert code. nice!!! [19:54]
<lavermil> I don't think it is loading the jar [19:54]
<eshelbk> it has to load that jar somewhere.. from the groovy-starter.conf [19:54]
<eshelbk> that's why i put it in opennms/lib, because i do a load ....*.jar [19:54]
<lavermil> nope, not with rx------- [19:54]
<eshelbk> i'd get it working as root, then back track to a user [19:55]
<lavermil> does it drop all the reports per node in the reports directory? [19:55]
<lavermil> yep [19:56]
<eshelbk> yep.. i modified my SELECT to just do a handful.. [19:56]
<lavermil> line 186 [19:56]
<eshelbk> something's wrong with the graph data. haven't looked into it [19:57]
<lavermil> BAH...damnit. gotta figure out how to comment. [19:58]
<eshelbk> but there's enough there to build it to what you would need. i will probably do more summarizing since we have a ton of nodes [19:58]
<lavermil> just like java [19:58]
<eshelbk> copy off the original :) [19:58]
<lavermil> this is for 1.7 right? [19:58]
<eshelbk> that's what mine runs on.. and jrobin. i'd like it to use rrdtool somehow but i'm not sure that is possible (since the point is java) [19:59]
<lavermil> bah, I cna use this in my other install, not this one. ugh [19:59]
<lavermil> soo clsoe. [19:59]
<lavermil> hehe [20:00]
<lavermil> brb [20:00]
<eshelbk> you can aim it at another server, and work out how to get to your jrb files remotely [20:01]
<lavermil> what is the snmpcollect piece? [20:02]
<lavermil> that seems to be my hangup [20:03]
<lavermil> not sure what you are trying to capture wtih that [20:03]
<eshelbk> it's a 1.7 thing, you could change it to issnmpprimary=(P|S|C) [20:04]
<eshelbk> i think the snmpcollect is a simpler toggle used instead of PSNC [20:05]
<lavermil> I don't see issnmpprimary in snmpinterface [20:05]
<eshelbk> it's in ip interface, so you'd have to make that query work [20:05]
<eshelbk> that's probalby the other reason for snmpcollect [20:06]
<lavermil> I am going to try it commented out and remove that check [20:06]
<eshelbk> you'll need it for graphs, it defines how to get the jrb file [20:07]
<lavermil> why? [20:07]
<lavermil> damn you script doesn't skip the Null intefaces. [20:08]
<lavermil> things that don't end up with graphs [20:08]
<eshelbk> all that is just a matter of forming your query [20:08]
<eshelbk> if i were to use it, i'd probably use an ifAlias tag and query against that for a filter [20:09]
<lavermil> eshelbk: nice wook. did it for just one node, but look pretty neat [20:16]
<lavermil> I gotta run [20:16]
<eshelbk> yeah, should be good until reportd gets in.. see ya [20:16]
<lavermil> but you gave me an idea. [20:16]
*** lavermil parted the channel [20:17]
*** brozow :cpe-071-065-197-116.nc.res.rr.com joined the channel [20:42]
*** shifter :flotsam.internetofdeath.com joined the channel [21:21]
*** mrwangma1ter :rrcs-70-62-54-110.central.biz.rr.com joined the channel [21:21]
*** jimmah :ubuntu/member/pak33m joined the channel [21:21]
*** shifter_ :flotsam.internetofdeath.com joined the channel [21:28]
<diq> did 1.6.10 not fix the http collection issue? [22:01]
*** azeem_ :pD9E1DFEA.dip0.t-ipconnect.de joined the channel [22:51]
*** robgault :64.128.239.186 joined the channel [23:49]
*** PiyushK :122.160.223.55 joined the channel [23:51]
<PiyushK> hello friends [23:54]
<PiyushK> can someone plz tell me - how can I get the cluster info (Graphs) of servers like CPU utilization of all of the servers in cluster in one graph etc.. in OpenNMS ..? [23:56]