logs | stats
   July 10, 2009  
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31

[00:06:33] *** jieryn-w has joined #hudson
[00:12:43] <rtyler> maxb: IIRC that's in the node configuration page
[00:12:50] <rtyler> Environment Variables at the bottom IIRC
[00:14:11] <maxb> oh, how odd, it's hidden behind a checkbox :-)
[00:14:12] <maxb> thanks
[00:14:56] <maxb> hm
[00:15:04] <maxb> it would appear to not work for MAVEN_OPTS
[00:15:17] <rtyler> surely there are global maven opts available
[00:15:21] * rtyler thought those were around there as well
[00:16:32] <maxb> it would seem not
[00:16:39] <rtyler> well that's too bad
[00:16:49] * rtyler suggests whining to kohsuke or something similar :OP
[00:17:36] <kohsuke> Global MAVEN_OPTS would be no different from setting an environment variable
[00:27:31] *** notme has joined #hudson
[00:35:58] *** waz has quit IRC
[00:41:55] <rtyler> kohsuke: forgive me for asking this offlist
[00:42:05] <kohsuke> ga
[00:42:06] <rtyler> but how do you accomplish pre-commit testing without innate knowledge of the VCS?
[00:42:33] <kohsuke> We obviously need some things to be done by SCM, for example merge and commit.
[00:42:37] <rtyler> and did abayer_ just say he wants concurrent build support?
[00:42:49] <abayer_> For non-build jobs, actually. =)
[00:42:52] <rtyler> after he dinged me so much about it?
[00:42:52] <kohsuke> But different approaches require different levels of support.
[00:42:59] <rtyler> :D
[00:43:25] <kohsuke> Yes, Andrew said he wants concurrent builds.
[00:43:33] * rtyler grins
[00:43:38] <kohsuke> And did you know that we now have a branch for that?
[00:43:39] <abayer_> Not for compilation/testing, though.
[00:43:56] <rtyler> OH WOW
[00:43:57] * rtyler grins
[00:44:28] <rtyler> kohsuke: branch name?
[00:44:37] <kohsuke> concurrent-builds. I haven't committed my local changes yet
[00:44:41] <rtyler> ah
[00:44:42] <rtyler> <3
[00:45:01] <kohsuke> And that's just a beginning --- this will likely be a non-trivial change.
[00:45:12] <rtyler> plz release before pre-tested ;)
[00:45:32] <kohsuke> I really need some time chunk to concentrate on this.
[00:45:47] <kohsuke> Can Slide inc. please buy my time to do this ?
[00:45:49] <kohsuke> :-)
[00:46:09] <rtyler> kohsuke: I'm still having trouble getting Slide to fucking commit to CI in general right now
[00:46:36] <rtyler> majority of the push comes from admc and I, we've not yet been able to codify CI and unit/integration testing as part of the development workflow
[00:46:41] * rtyler cries quietly to himself
[00:46:43] *** admc has joined #hudson
[00:46:53] <rtyler> admc: branch: concurrent-builds
[00:46:59] <admc> irssi had gone insane
[00:46:59] <rtyler> admc: guess what's going on in there ;)
[00:47:21] <admc> I can't even guess, some magic?
[00:47:43] <rtyler> jobs running concurrently yo
[00:47:55] <rtyler> read: my crazy external job hacks --> true support in core
[00:47:58] <rtyler> admc: are you on dev@?
[00:48:33] <admc> yea
[00:49:08] <rtyler> admc: catch up on the "Pre-tested commit feature" thread
[00:49:21] <rtyler> admc: also http://wiki.hudson-ci.org/display/HUDSON/Designing+pre-tested+commit
[00:51:46] *** BigAllan has quit IRC
[00:53:04] <admc> rtyler: looks like this feature made itself important without you even being too naggy :)
[00:55:27] <rtyler> admc: kohsuke wants slide to buy some consulting time for it :P
[00:55:38] <rtyler> I pointed out our pisspoor commitment to CI >_<
[00:57:04] <admc> yeah, I wish that was a bit higher priority
[00:57:05] <admc> :/
[00:57:22] *** notme is now known as waz
[01:00:36] <mrooney> anyone know if when the locks and latches plugin is holding up one build because two actually started at the same time, if that counts against the build timer time?
[01:00:47] * rtyler lulz at locks
[01:00:58] <mrooney> well, I do it for the lulz
[01:01:05] <mrooney> I am trying to figure out if these intermittent aborted builds are because of that
[01:01:09] <rtyler> kohsuke: I think you can acheive what this wiki page describes with a DVCS fairly easily
[01:01:36] <kohsuke> SCM part of it is not too hard, yes.
[01:01:51] <abayer_> Kohsuke, I'll reply about the problems with ClearCase later tonight.
[01:01:55] <rtyler> kohsuke: Dev --[push]--> Hudson repo...[test]...[success]...[merge]--[push]--> primary repo
[01:02:06] <mrooney> rtyler: yeah that's a good point
[01:02:20] <mrooney> so it wouldn't really need much underlying changes in Hudson in that case
[01:02:22] <abayer_> Basically, yo uneed more information than just branch names.
[01:02:42] *** decriptor is now known as decriptor|out
[01:02:48] <rtyler> mrooney: yeah, but you kind of leave the merging up to an automated process, whcih is kind of lame
[01:03:36] <mrooney> rtyler: well the dev does that once they are happy and pushes, so Hudson then builds trunk
[01:05:23] <mrooney> rtyler: was there ever a sweet Hudson hackathon decided upon?
[01:05:48] <rtyler> oh right
[01:05:54] <rtyler> hey kohsuke
[01:06:02] <rtyler> want to havea a hackathon next weekend in SF?
[01:06:06] <rtyler> and meetup
[01:06:22] <abayer_> Works for me.
[01:06:24] <kohsuke> sure
[01:06:49] <mrooney> excellent!
[01:08:06] <kohsuke> rtyler: will you be hosting a place?
[01:08:16] <kohsuke> I mean hosting a meet up?
[01:08:23] <rtyler> kohsuke: yeah
[01:09:24] <kohsuke> cool
[01:09:55] <rtyler> alright, let me get final confirmation from my boss, and then I'll email dev@ and users@
[01:10:06] <kohsuke> thanks
[01:13:21] *** Lewisham has joined #hudson
[01:13:29] <Lewisham> Hi there
[01:13:31] <rtyler> ahoy
[01:14:41] <Lewisham> I have a newbie problem: I'm trying to write a plugin that extends Recorder, so I can write a JDepend plugin. I'm using the code from Crap4J as a guideline, but I can't get my Plugin to appear in the post-build list of options. I'm feeling completely lost :/
[01:15:05] <kohsuke> Are you an Eclipse user?
[01:16:24] <Lewisham> yes
[01:16:48] <kohsuke> Maybe your Eclipse is failing to create the necessary files for discovering your extensions
[01:16:55] <kohsuke> Can you try "mvn clean hpi:run" and see if that helps?
[01:16:58] <Lewisham> I used Maven to check out hpi:create, and I've been building on that
[01:17:01] <Lewisham> I'll have a go
[01:17:29] <kohsuke> This only needs to happen once when you define a new extension point --- you can make all the other usual changes from Eclipse like you normally do
[01:21:01] <Lewisham> no love :(
[01:21:29] <kohsuke> Where did you put your @Extension?
[01:21:34] <Lewisham> http://hudson.pastebin.com/m24d2c587
[01:21:40] <Lewisham> In PluginImpl.java
[01:22:04] <kohsuke> Ah, you need to put it on your DescriptorImpl, too
[01:22:09] <Lewisham> ohhh
[01:22:12] <kohsuke> In fact I don't think you need a PluginImpl class at all.
[01:22:29] <kohsuke> at least as it stands now, unless you are thinking about extending it.
[01:23:22] <Lewisham> I probably don't, but I was just building off of the hpi:create archetype to build my confidence up :)
[01:23:28] <kohsuke> yes
[01:23:31] <kohsuke> I understand
[01:23:38] <kohsuke> Are you already a committer?
[01:23:49] <Lewisham> no, I'm a complete newbie
[01:24:01] <Lewisham> I started on this project a couple of days ago for some PhD work I'm doing
[01:24:06] <kohsuke> You should be a committer, so that you can push your plugin to Hudson's subversion repository.
[01:24:26] <kohsuke> Let me know your java.net ID so that I can make you a committer.
[01:24:41] <Lewisham> my java.net ID is Lewisham
[01:24:57] <kohsuke> I don't think java.net ID allowed upper case.
[01:25:00] <Lewisham> oh
[01:25:05] <Lewisham> then it's lewisham ;)
[01:25:06] <kohsuke> Can you double check what your ID is?
[01:25:24] <Lewisham> yeah, logged in as "lewisham"
[01:25:24] <kohsuke> OK. It should be set in 30 secs or so.
[01:25:34] <Lewisham> awesome, thanks kohsuke :)
[01:25:44] <kohsuke> Looking forward to seeing the code in the repository
[01:25:51] <kohsuke> There's a wiki page that talks about hosting on hudson.dev.java.net.
[01:25:59] <kohsuke> please take a look when you can.
[01:26:22] <Lewisham> yeah, I've taken a look at that page. I figured I'd contact you guys when I had something that works
[01:26:28] <Lewisham> I'm new to Maven and Hudson *gulp*
[01:26:34] <Lewisham> so it's a good thing there's a nice long summer to go!
[01:26:47] <kohsuke> :-)
[01:27:53] <Lewisham> ah, that worked perfectly, thanks kohsuke
[01:27:59] <kohsuke> good
[01:28:11] <Lewisham> one other question I had: is there a good way to find directory paths?
[01:28:16] <kohsuke> I should probably have some annotation processor to detect that
[01:28:26] <kohsuke> Not sure if I understand your question.
[01:28:52] <Lewisham> JDepend generates an intermediate file, and I'd rather it went somewhere more clever than the plugin directory, just in case it can't get cleaned up
[01:29:07] <Lewisham> but I was wondering if Hudson had a tmp directory or something that I could reference
[01:29:54] <kohsuke> FilePath.createTempDir?
[01:30:19] <kohsuke> Or maybe even just File.createTempFile
[01:30:46] <Lewisham> oh
[01:30:47] <Lewisham> genius~!
[01:31:12] <Lewisham> it's been a long time since I've used Java after wandering the Perl wilderness for employment :)
[01:41:32] *** waz has quit IRC
[01:57:01] *** waz has joined #hudson
[02:00:27] <admc> Lewisham: good on ya -- someone has to keep perl alive :)
[02:00:47] <admc> I was actually surprised at OSBridge how many people were using perl and R
[02:02:25] <Lewisham> Perl isn't my personal choice of hurting myself in dynamic languages, I'm a Python guy, but the boss was of the Unix/bare feet crowd of 20 years ago, so the idea of anything else horrified him
[02:02:29] <Lewisham> "That sounds like a regex problem"
[02:02:45] <Lewisham> "How do I get this Apache file to work?"
[02:02:49] <calculus> I like R
[02:02:54] <Lewisham> "Sounds like a regex problem"
[02:03:00] <Lewisham> "The coffee machine doesn't work."
[02:03:04] <Lewisham> "Have you tried regex?"
[02:09:02] <admc> hahaha
[02:09:09] <admc> awesome, ive met some pretty rad perl guys at conferences
[02:09:17] <admc> dusty backpacks, 2 foot long white beards
[02:09:24] <admc> you can pick em out of a crowd
[02:09:34] <rtyler> do they eat their toenails like RMS?
[02:09:38] * rtyler gags
[02:10:22] <admc> not sure if thats a perl requirement or not
[02:22:49] <Lewisham> they actually hand them out at the door at PerlConf
[02:22:53] <Lewisham> you're turned away if you don't chew
[02:23:05] *** wsmoak has quit IRC
[02:24:39] *** wsmoak has joined #hudson
[02:27:17] *** wsmoak has quit IRC
[03:05:47] *** abayer_ has quit IRC
[03:06:01] *** abayer has joined #hudson
[03:18:41] <Lewisham> I have another question! :) JDepend requires that it's given the location of the data when it is called. (ie. it takes a directory path to the source in the constructor) It looks like I have to do this with FilePath, but I'm finding the class a bit confusing
[03:33:48] <mrooney> Interesting, Hudson seems to be eating my javascript in an iframe?
[03:48:01] <mrooney> kohsuke: so I'm using DirectoryBrowserSupport to serve an html file, which includes another file in an iframe
[03:48:38] <mrooney> the top level file uses js with no problem, but the iframe has js which doesn't seem to work in hudson, I put an alert in to debug but don't get it
[03:48:40] <mrooney> however, if I make that same directory web accessible via apache it works fine
[03:48:59] <mrooney> I wonder if I am doing something wrong
[04:16:02] *** admc has quit IRC
[05:58:55] *** davidstrauss has joined #hudson
[06:05:01] *** Lewisham has quit IRC
[06:22:59] *** glass-eye has quit IRC
[06:28:01] *** jodell has quit IRC
[06:42:18] *** Lewisham has joined #hudson
[06:43:48] *** jodell has joined #hudson
[06:46:06] *** waz has quit IRC
[07:35:40] *** Lewisham_ has joined #hudson
[07:51:04] *** Lewisham has quit IRC
[08:07:22] *** Lewisham_ has left #hudson
[08:20:01] *** admc has joined #hudson
[08:28:02] *** madsdyd has joined #hudson
[08:34:57] *** akostadinov has joined #hudson
[08:40:12] *** admc_ has joined #hudson
[08:42:29] *** admc has quit IRC
[08:45:12] <madsdyd> Hi there. Anyone knows if hudson sometimes decides to do some nifty cleanout in the workspace directories?
[08:47:15] *** admc has joined #hudson
[08:47:19] *** admc_ has quit IRC
[08:54:32] *** khmarbaise has joined #hudson
[09:06:26] *** echelog-1_ has joined #hudson
[09:06:58] *** echelog-1_ is now known as echelog-1
[09:31:37] *** jodell has quit IRC
[09:42:34] * madsdyd answers himself: Yes it does, apparently it thinks it needs to be done at 8:30 am...
[10:00:30] *** BigAllan has joined #hudson
[10:04:09] *** akostadinov has quit IRC
[10:17:56] *** timp_ has joined #hudson
[10:28:20] *** madsdyd is now known as madsdyd_work
[10:31:30] *** akostadinov has joined #hudson
[10:31:32] *** dvrzalik has joined #hudson
[10:33:37] *** akostadinov has quit IRC
[10:39:54] *** akostadinov has joined #hudson
[11:24:37] *** akostadinov has quit IRC
[11:31:15] *** dirkD has quit IRC
[11:32:41] *** akostadinov has joined #hudson
[12:11:48] *** luca_perfetti has joined #hudson
[12:57:32] *** akostadinov has quit IRC
[13:18:59] <madsdyd_work> Does anyone know if I can get a buildtrigger to trigger another project, even if this project failed?
[13:31:52] *** madsdyd_work has quit IRC
[14:06:38] *** akostadinov has joined #hudson
[14:12:44] *** velo has joined #hudson
[14:34:51] *** jdolan_ has joined #hudson
[15:10:53] *** khmarbaise_ has joined #hudson
[15:11:23] *** khmarbaise has quit IRC
[15:13:42] *** abayer_ has joined #hudson
[15:16:51] *** khmarbaise_ has quit IRC
[15:28:16] *** madsdyd has joined #hudson
[15:31:16] *** abayer has quit IRC
[15:33:19] *** wsmoak has joined #hudson
[15:34:42] *** waz has joined #hudson
[15:36:35] *** jbenlinlee has quit IRC
[16:21:39] *** karihre has quit IRC
[16:31:10] *** jodell has joined #hudson
[16:33:03] *** davidstrauss has quit IRC
[16:34:39] *** akostadinov has quit IRC
[16:44:30] *** madsdyd has quit IRC
[16:47:57] *** luca_perfetti has quit IRC
[16:49:23] *** davidstrauss has joined #hudson
[16:55:12] *** khmarbaise has joined #hudson
[17:10:06] *** decriptor|out is now known as decriptor
[17:15:36] *** brent_ has joined #hudson
[17:35:18] *** dvrzalik has quit IRC
[17:36:43] *** jodell has quit IRC
[17:57:39] *** keshureddyp has quit IRC
[17:57:59] *** keshureddyp has joined #hudson
[18:12:40] *** madsdyd has joined #hudson
[18:20:45] *** jodell has joined #hudson
[18:43:17] *** timp_ has quit IRC
[19:25:01] *** BigAllan has quit IRC
[19:44:58] <jieryn-w> uhh. i upgrade from 1.310 to 1.314 and nearly all my jobs are gone
[19:45:12] * jieryn-w screams painfuully
[19:48:38] *** decriptor is now known as decriptor|lunch
[19:49:46] <mrooney> oh no!
[20:00:15] <jieryn-w> i went into .hudson/jobs and i see them still existing
[20:15:32] *** alexlod_ has joined #hudson
[20:18:11] <jieryn-w> http://dpaste.com/65584/
[20:19:45] <jieryn-w> trying to restart tomcat, maybe the Windows mentality would work
[20:26:01] <jieryn-w> something is really fragged with hudson+antlr
[20:26:16] <jieryn-w> restart of tomcat didn't solve it completely, but i now have about half my jobs before hudson dies parsing config.xml
[20:26:27] <jieryn-w> yaay!! ... reverting downt o 1.310
[20:26:29] <jieryn-w> hopefully i won't die
[20:29:59] <jieryn-w> ...and reverting to 1.310 didn't work.. crap on me
[20:31:13] *** alexlod has quit IRC
[20:37:38] *** jdolan_ has quit IRC
[20:46:58] <mrooney> jieryn-w: you might ping kohsuke
[20:47:30] <jieryn-w> kohsuke: ping
[20:47:36] <kohsuke> look at the log output from Hudson
[20:47:42] <jieryn-w> i pasted it to a bin
[20:48:26] <jieryn-w> another restart of tomcat yields a new error.. yaay
[20:48:35] <jieryn-w> http://dpaste.com/65594/
[20:49:07] <kohsuke> There appears to be some sort of war file corruption
[20:49:26] <kohsuke> The directory where your war file gets expanded might have been compromised.
[20:49:34] <kohsuke> I suggest you remove those and let Tomcat redeploy your war
[20:50:00] <jieryn-w> i'm using tomcat manager app's form post to deploy the war from my local machine
[20:50:06] <jieryn-w> is there an md5 i can run on the hudson.war?
[20:50:31] <kohsuke> I don't think it's a corrupted war file --- compressed files are very hard to corrupt without you noticing it
[20:51:02] <jieryn-w> ok, so i'll bring down tomcat and then rm -rf hudson in the webapp dir
[20:52:19] <kohsuke> yeah
[20:55:14] *** Lewisham has joined #hudson
[21:00:31] <Lewisham> hey guys, is there a way to ensure that my plugin is able to operate on data on the local machine? JDepend requires the data to be local to wherever it is executed (ie. it can only handle a local path given to it), and it seems that is not guaranteed with Hudson. Because it also outputs as a file that has to be parsed and read back in (ick) I can't easily offload the processing and get the results back again. Any ideas?
[21:01:13] <kohsuke> Distributed computing always have two options: either move your data, or move your code
[21:01:50] <kohsuke> You can still do both --- copy all the inputs to the master and then let JDepend work on it, or let JDepend run on slave and bring the result back to the master
[21:02:05] <kohsuke> The latter should be the first default choice to avoid doing too much on the master.
[21:02:09] <kohsuke> But it's your call.
[21:02:46] <rtyler> mrooney: RSVP KTHXBAI
[21:02:54] *** jodell_ has joined #hudson
[21:09:09] *** admc_ has joined #hudson
[21:09:18] <jieryn-w> ok - once again i am getting Caused by: java.lang.ClassNotFoundException: antlr.ParserSharedInputState
[21:11:23] <Lewisham> kohsuke: I think it would be easier to implement the former because JDepend is pretty icky... how would I go about copying all the data back to the master? Something with the FilePath from build.getProject().getWorkspace()?
[21:11:46] <kohsuke> See various methods on FilePath. it hides most of the remote operations
[21:12:51] *** meonkeys has joined #hudson
[21:13:46] <meonkeys> hi there, I'm seeing a strange issue in a Hudson job where I run a script that restarts Tomcat--once the Hudson job completes, Tomcat shuts down.
[21:14:23] *** jodell has quit IRC
[21:14:28] <kohsuke> See http://wiki.hudson-ci.org/display/HUDSON/Spawning+processes+from+build
[21:14:30] <meonkeys> There is a warning: "Process leaked file descriptors. See ... for more information"
[21:14:33] <meonkeys> yeah, that one. :)
[21:15:55] <meonkeys> so, what is the effect of this leaked file descriptor problem? Could that potentially cause Tomcat, fired off from within the control script, to receive a SIGINT or SIGTERM or something?
[21:16:02] *** admc__ has joined #hudson
[21:16:08] *** admc has quit IRC
[21:16:33] <meonkeys> (by the way, the script uses jsvc to control Tomcat)
[21:16:57] <Lewisham> kohsuke: so something like getWorkspace().copyTo(new FilePath("/tmp")); would dump it onto the master?
[21:17:20] <Lewisham> then use the new FilePath to get a URI...
[21:17:24] <Lewisham> *should work*
[21:17:59] *** davidstrauss has quit IRC
[21:20:04] *** madsdyd has quit IRC
[21:21:14] <meonkeys> ah, maybe I need to force the JVM fired up by jsvc to be headless? This Hudson job of mine uses xvnc...
[21:24:21] *** madsdyd has joined #hudson
[21:28:35] <jieryn-w> kohsuke: i verified that the hudson.war i upload via manager and the one i upload to the server via sftp are identical
[21:28:48] <jieryn-w> (and let tomcat deploy the .war from local file system)
[21:28:58] <jieryn-w> so, something else is causing trouble
[21:29:28] *** brent_ has quit IRC
[21:32:19] *** admc_ has quit IRC
[21:40:14] *** ldsjohn316 has quit IRC
[21:40:21] *** ldsjohn has joined #hudson
[21:47:06] *** abayer_ has quit IRC
[21:47:19] *** abayer has joined #hudson
[21:48:46] <jieryn-w> hudson/WEB-INF/libs/antlr-2.7.6.jar definitely contains the class that is choking on ClassNotFoundException ...
[21:49:39] *** ed_mann has joined #hudson
[21:49:40] <Lewisham> maven q: How do I update Maven so that it builds with a newer Hudson? The archetype comes down with 1.300, which is too old. Do I need to edit the pom.xml?
[21:50:02] <Lewisham> *for a plugin
[21:51:14] *** decriptor|lunch is now known as decriptor
[21:51:15] <Lewisham> oh, I think I just add the newer version as a dependency
[21:51:19] <Lewisham> easy!
[21:51:56] *** alexlod_ has quit IRC
[21:53:49] <Lewisham> should I always be trying to build against the latest stable release, or the earliest one that will support the plugin?
[22:02:25] *** jdolan_ has joined #hudson
[22:09:50] *** khmarbaise_ has joined #hudson
[22:16:02] *** khmarbaise has quit IRC
[22:16:16] *** admc__ has quit IRC
[22:16:25] <Lewisham> kohsuke: I got the workspace copying to a temp directory, running JDepend, then cleaning up afterwards. It's not pretty, but it works (I'm agile, you see? ;) )! Thanks a lot for the help!
[22:18:45] *** BigAllan has joined #hudson
[22:20:16] *** alexlod has joined #hudson
[22:23:19] *** admc has joined #hudson
[22:28:46] <calculus> Lewisham: I saw a /tmp in the code your posted earlier... not all machines have a /tmp (e.g. Windows)
[22:29:02] <Lewisham> yeah, that was some pseudo-code
[22:29:08] <calculus> oh, ok
[22:29:17] <Lewisham> I'm using System.getProperty("java.io.tmpdir")
[22:29:53] <Lewisham> and it's guarded so I only do the copy if the data is remote
[22:30:08] <Lewisham> so it's still not perfect (really the computation should be off-loaded) but it gets things done
[22:31:05] <calculus> I wasn't aware of that property... maybe that is what File.createTempFile uses
[22:32:07] <Lewisham> I believe it is
[22:32:16] <Lewisham> that property goes to some ridiculous dir in OS X 10.5
[22:32:30] <Lewisham> today its: /var/folders/K+/K+YwCJIDEbSqfl4pxVGgg++++TI/-Tmp-
[22:32:56] <Lewisham> I don't know if they're timestamps, or hashes or what, but Apple's Java implementation is aware of it
[22:33:31] <calculus> lovely
[22:43:31] <jieryn-w> i removed all the plugins except for maven/ssh/svn plugins which seem like system ones
[22:43:38] <jieryn-w> and now i have about a dozen jobs .... still missing many
[22:44:23] <jieryn-w> other than removing .hudson/plugins/*.hpi and corresponding unpacked dir, what must i do to fully remove a plugin from hudson?
[22:47:38] <kohsuke> like I said, I don't think it's the war file that's corrupted. I think it's what's expanded by Tomcat that's corrupted.
[22:47:55] <jieryn-w> can i just unzip it myself then?
[22:48:06] <kohsuke> Try "java -DHUDSON_HOME=/path/to/data -jar hudson.war" to isolate the problem
[22:48:13] <jieryn-w> good idea
[22:49:22] <kohsuke> I'd also do some basic file system sanity check like available disk space, quota, etc.
[22:49:47] *** madsdyd has quit IRC
[22:50:00] <jieryn-w> i get the same result
[22:50:56] <jieryn-w> err, no
[22:51:03] <jieryn-w> now i'm taking boatloads of errors 'cuz of the plugins i removed
[23:01:04] <jieryn-w> ok, so this is catastrophic system error here
[23:01:08] *** luca__perfetti_1 has joined #hudson
[23:01:12] <jieryn-w> i guess my only recourse is to wipe .hudson and start anew
[23:05:17] *** jdolan_ has quit IRC
[23:12:10] *** meonkeys has left #hudson
[23:18:15] <keshureddyp> is any one able to make it work IRC plugin on hudson ?
[23:18:32] <rtyler> keshureddyp: the IRC plugin is currently unmaintained
[23:18:38] <keshureddyp> I see - when save configuration "-ErrorImpossible to connect to IRC server"
[23:18:48] <kohsuke> Do you see a stack trace, too?
[23:19:01] <keshureddyp> i didn't see any stack trace
[23:19:28] <keshureddyp> in logs .. but as soon as i hit configuration save button i saw that messgae
[23:20:13] <kohsuke> If you can check an issue tracker and file a bug if necessary, that would be great.
[23:20:44] <keshureddyp> sure
[23:21:00] <rtyler> I might hack on the IRC plugin next weekend as well
[23:21:03] * rtyler isn't sure
[23:21:41] *** BigAllan has quit IRC
[23:22:14] <keshureddyp>  filing bug now
[23:22:53] <rtyler> keshureddyp: I think you can "CC" somebody on a ticket, can you cc me?
[23:23:16] <kohsuke> rtyler: or you can just query the issue tracker by the 'irc' component
[23:23:18] <kohsuke> to see them all
[23:23:27] <keshureddyp> ok
[23:23:35] <rtyler> kohsuke: does a CC not result in a n email?
[23:23:41] <kohsuke> it does.
[23:24:25] <kohsuke> http://tinyurl.com/hudsonIrcBugs
[23:27:34] *** khmarbaise_ has quit IRC
[23:29:08] *** Baraius has joined #hudson
[23:30:44] <Baraius> How much control do you have over individual steps in a Hudson build?  i.e. Can you say "If this step fails, mark the build as failed but continue"  "If this step fails, abort the build and mark it as failed."
[23:31:25] *** khmarbaise has joined #hudson
[23:35:53] <keshureddyp>  auto assined to -> bruyeron (bruyeron)
[23:36:04] <rtyler> who the hell is that?
[23:36:06] * rtyler boggles
[23:36:21] <rtyler> Baraius: I don't know of any such functionality
[23:38:37] <keshureddyp> https://hudson.dev.java.net/issues/show_bug.cgi?id=4019
[23:40:17] *** luca__perfetti_1 has left #hudson
[23:44:14] *** jieryn-w has quit IRC
[23:49:37] <Baraius> Thanks rtyler.
[23:51:25] *** Baraius has quit IRC

top