fluid-work IRC Logs-2010-05-06

Owned by Fluid IRC Log Bot
Last updated: May 06, 2010Version comment
11 min read

[02:11:05 EDT(-0400)] * joost (~joost@dfki-113.dfki.uni-kl.de) has joined #fluid-work
[05:39:01 EDT(-0400)] * denbuzze (~anonymous@ginger.caret.cam.ac.uk) has joined #fluid-work
[06:39:11 EDT(-0400)] * JASIGLogBot (~PircBot@jasig.Princeton.EDU) has joined #fluid-work
[06:39:11 EDT(-0400)] * Topic is 'This channel is logged – for details see: http://wiki.fluidproject.org/display/fluid/IRC+Channel' set by jessm on 2009-11-04 08:30:00 EST(-0500)
[08:15:33 EDT(-0400)] * joost (~joost@dfki-113.dfki.uni-kl.de) has joined #fluid-work
[08:34:44 EDT(-0400)] * Justin_o (~Justin@142.150.154.171) has joined #fluid-work
[08:38:36 EDT(-0400)] * anastasiac (~team@142.150.154.193) has joined #fluid-work
[08:48:37 EDT(-0400)] * michelled (~michelled@142.150.154.141) has joined #fluid-work
[08:49:33 EDT(-0400)] * christianv (~anonymous@ginger.caret.cam.ac.uk) has joined #fluid-work
[08:52:53 EDT(-0400)] * kasper_ (~kasper@187.132.250.167) has joined #fluid-work
[09:00:40 EDT(-0400)] * colinclark (~colin@bas2-toronto09-1176131031.dsl.bell.ca) has joined #fluid-work
[09:16:21 EDT(-0400)] * athena (~athena@adsl-99-62-112-41.dsl.wlfrct.sbcglobal.net) has joined #fluid-work
[09:17:57 EDT(-0400)] * yura_ (~yura@142.150.154.114) has joined #fluid-work
[09:19:02 EDT(-0400)] <Justin_o> colinclark: i'm cloning your hg infusion repo now... i think i'm going to try to break out just the Infusion 1.2 tag for use in Decapod
[09:19:08 EDT(-0400)] <Justin_o> as a subrepo
[09:19:14 EDT(-0400)] <colinclark> ok
[09:19:17 EDT(-0400)] <colinclark> cool
[09:19:35 EDT(-0400)] <colinclark> can't wait to see how it goes
[09:19:49 EDT(-0400)] * denbuzze (~anonymous@ginger.caret.cam.ac.uk) has joined #fluid-work
[09:22:49 EDT(-0400)] <Justin_o> colinclark: (smile) just have to wait for hg to finish pulling down the 280mb infusion repo
[09:22:55 EDT(-0400)] <colinclark> ouch
[09:22:57 EDT(-0400)] <colinclark> that big, huh?
[09:23:04 EDT(-0400)] <colinclark> Use ssh, it was a lot faster for me
[09:23:37 EDT(-0400)] <Justin_o> oh really... okay... i'll try that
[09:29:36 EDT(-0400)] <colinclark> Justin_o: Also, I found plugging into your wire helped a lot, too
[09:29:44 EDT(-0400)] <colinclark> Much faster than wireless
[09:33:53 EDT(-0400)] <Justin_o> that makes sense... i'm using my mini for the conversion stuff which is plugged into the network.. I just have to set up an ssh key though
[09:44:04 EDT(-0400)] * clown (~clown@142.150.154.101) has joined #fluid-work
[10:20:50 EDT(-0400)] * colinclark (~colin@142.150.154.101) has joined #fluid-work
[11:19:15 EDT(-0400)] <Justin_o> kasper_: Hello, just wanted to let you know that I have a some time to lend a hand if you need some help getting a daily build up and running for cspace
[11:25:35 EDT(-0400)] <kasper_> hey justin_o, thanks a lot
[11:26:21 EDT(-0400)] <kasper_> justin_o: ok, so a quick sum-up of what i've done so far
[11:26:54 EDT(-0400)] <kasper_> we got the bamboo server up and running, and I've set up a service-layer build task, and a service-layer deploy task
[11:27:57 EDT(-0400)] <kasper_> and all day yesterday (and this morning) I've had an intense fist fight with jboss/maven/cargo, as to be able to start/stop the two jboss servers from within bamboo
[11:28:07 EDT(-0400)] * anastasiac (~team@142.150.154.193) has joined #fluid-work
[11:28:21 EDT(-0400)] <Justin_o> kasper_: those are all for the service layer?
[11:28:37 EDT(-0400)] <kasper_> it looks like it's working now (when running it from console), but haven't tested on bamboo yet
[11:28:49 EDT(-0400)] <Justin_o> kasper_: okay...
[11:29:05 EDT(-0400)] <Justin_o> can you send me a link to the docs that describes how to build cspace
[11:29:17 EDT(-0400)] <kasper_> justin_o: yup, sec
[11:29:40 EDT(-0400)] <kasper_> http://wiki.collectionspace.org/display/collectionspace/Release+0.6+Quick+Installation+Instructions
[11:30:37 EDT(-0400)] <kasper_> Justin_o: oh, actually, those are not describing the process that bamboo has to go through precisely
[11:31:08 EDT(-0400)] <kasper_> there are the added steps of getting the service layer and UI layer sources and compiling them
[11:31:32 EDT(-0400)] <Justin_o> right
[11:33:25 EDT(-0400)] <kasper_> So i guess the process would be something like this:
[11:34:30 EDT(-0400)] <kasper_> (1) stop jboss servers, (2) update/build service layer, (3) deploy service layer, (4) start jboss servers, (5) test service layer
[11:35:21 EDT(-0400)] <kasper_> (6) update/build app layer, (7) stop jboss server, (8) deploy app layer (9) start jboss server, (10) test app layer
[11:35:52 EDT(-0400)] <kasper_> (11) (15): same as (6)(10) with the UI
[11:37:02 EDT(-0400)] <kasper_> ... actually i'm not sure if jboss needs to be stopped/started between ui and app deploys.. asking now
[11:38:19 EDT(-0400)] <kasper_> justin_o: ok step 7 and 9 can be skipped
[11:39:35 EDT(-0400)] <kasper_> justin_o: and 12 and 14
[11:41:19 EDT(-0400)] * EricDalquist (~dalquist@2607:f388:e:0:221:9bff:fe37:e768) has joined #fluid-work
[11:43:28 EDT(-0400)] <Justin_o> kasper_: okay... .what do you mean by steps 5 and 10
[11:44:38 EDT(-0400)] <kasper_> Justin_o: the service layer has a set of unit tests, that can be run using ant or maven
[11:45:12 EDT(-0400)] <kasper_> justin_o: similarly the app layer has unit tests testing the app layer it self, and also a few tests using the service layer
[11:46:13 EDT(-0400)] <Justin_o> kasper_: i'm guessing there is an maven and ant install/build scripts that are run.. are the unit tests run by these or triggered seperately
[11:47:18 EDT(-0400)] <kasper_> justin_o: triggered separately (by a seperate target, that it)
[11:48:15 EDT(-0400)] <Justin_o> kasper_: okay...
[11:48:28 EDT(-0400)] <kasper_> s/it/is/
[11:48:35 EDT(-0400)] <Justin_o> so you currently have step 1 - 5 potentially working
[11:48:52 EDT(-0400)] <kasper_> justin_o: in theory yes
[11:49:24 EDT(-0400)] <kasper_> justin_o: and in practice i have step 1 working
[11:49:47 EDT(-0400)] <Justin_o> kasper_: (smile) okay... so you are writing a script to take care of this?
[11:52:29 EDT(-0400)] <kasper_> Justin_o: That is one thing i I haven't quite decided on yet... Whether to make a bamboo task for each step (and maybe even several for each step), or to script my way out of it
[11:53:31 EDT(-0400)] <kasper_> I guess tasks like editing the two files (re. install instructions), I would do a small bash script
[11:54:06 EDT(-0400)] <Justin_o> kasper_: probably i think... which version of bamboo are you using and do you have a link to the cspace instance of it
[11:55:04 EDT(-0400)] <kasper_> justin_o: we are using version 2.5
[11:55:20 EDT(-0400)] <kasper_> and what do you mean by link to cspace?
[11:57:47 EDT(-0400)] <Justin_o> kasper_: sorry... meant a link to your use of bamboo
[11:57:55 EDT(-0400)] <Justin_o> i'm guessing there is some sort of web interface
[11:58:02 EDT(-0400)] <Justin_o> or are you working on it locally at the moment
[11:58:17 EDT(-0400)] <kasper_> oh, sorry: qa.collectionspace.org:8085
[11:59:08 EDT(-0400)] <kasper_> ... we actually also have a bamboo.collectionspace.org .. but the plan is to remove that, once the qa-bamboo is up and running
[11:59:23 EDT(-0400)] <Justin_o> kasper_: okay.. thanks
[12:00:05 EDT(-0400)] <kasper_> justin_o: the project i've created for this is called "nightly build" bottom of list
[12:02:36 EDT(-0400)] <kasper_> justin_o: got standup meeting now for about 15 minutes
[12:02:46 EDT(-0400)] <kasper_> will ping you afterwards
[12:02:52 EDT(-0400)] <Justin_o> kasper_: sure..
[12:27:40 EDT(-0400)] * jimeng (~jimeng@141.212.237.217) has joined #fluid-work
[12:48:14 EDT(-0400)] <kasper_> justin_o: ok, back now
[12:48:18 EDT(-0400)] <kasper_> ... sorry it took so long
[12:49:19 EDT(-0400)] <Justin_o> kasper_: no problem... i'm just finishing up creating myself a virtual machine of ubuntu to try make a local instance of bamboo for myself to work with
[12:49:38 EDT(-0400)] <kasper_> ah nice
[12:49:46 EDT(-0400)] <kasper_> do you use bamboo on fluid?
[12:49:58 EDT(-0400)] <Justin_o> no we use continuum
[12:50:11 EDT(-0400)] <kasper_> justin_o: ok, never tried that before
[12:50:29 EDT(-0400)] <kasper_> (then, I hadn't tried bamboo before either)
[12:50:47 EDT(-0400)] <kasper_> justin_o: By the way, In terms of the community/committers, do you have a particular set of guidelines, as to avoid braking it too often?
[12:51:02 EDT(-0400)] <Justin_o> kasper_: it's open source, part of apache... i haven't done much with bamboo either... i looked at a bit a few months ago when i was looking at ci servers
[12:51:13 EDT(-0400)] <Justin_o> kasper_: well.. the goal is to never brake the build
[12:51:24 EDT(-0400)] <Justin_o> if it happens it has to be fixed asap
[12:52:16 EDT(-0400)] <kasper_> ok
[12:52:18 EDT(-0400)] <Justin_o> kasper_: Once the daily build is up and running, I think you should start by letting everyone know that the daily build should always be working..
[12:52:32 EDT(-0400)] <Justin_o> and that it isn't accecptable to break it...
[12:52:37 EDT(-0400)] <kasper_> justin_o; yeah, that makes sense
[12:52:46 EDT(-0400)] <Justin_o> of course this will happen... so make sure that they know it is a priority to restore it
[12:52:58 EDT(-0400)] <kasper_> yeah
[12:53:02 EDT(-0400)] <Justin_o> you may also want to suggest that people take a look at it everyday if they can
[12:53:26 EDT(-0400)] <Justin_o> This should help the people working on various layers also know what is going on in another layer
[12:53:43 EDT(-0400)] <Justin_o> i'm not sure how closely everyone works on cspace though...
[12:55:21 EDT(-0400)] <kasper_> justin_o: it depends on what people are working on, i thnk.
[12:56:01 EDT(-0400)] <kasper_> justin_o: we are trying a few things to make people better updated on each teams progress
[12:56:46 EDT(-0400)] <kasper_> amonst other, by creating JIRAs for each user story, and hopefully this nightly build will help too
[13:00:20 EDT(-0400)] <Justin_o> kasper_: yes... sounds like things are coming together
[13:00:47 EDT(-0400)] <kasper_> justin_o: yeah... getting closer, but i still feel i'm very far behind
[13:01:52 EDT(-0400)] <kasper_> desperately trying to catch up on the test plans so we're set for 0.7 in 1.5 weeks
[13:02:21 EDT(-0400)] <kasper_> and really anxious to get this nighly build up and running, as well as getting some automated acceptance testing going
[13:03:04 EDT(-0400)] <Justin_o> kasper_: i know what you mean, that's a lot of work
[13:03:58 EDT(-0400)] <kasper_> yeah (smile) .... but at least I fell like i'm catching up a bit, rather than falling farther behind (smile)
[13:04:36 EDT(-0400)] <Justin_o> (smile) that's good...
[13:12:17 EDT(-0400)] <kasper_> justin_o: oh, completely unrelated: In terms of performance and speed testing - is that something you do, and are there any particular tools you have a positive experience with?
[13:13:57 EDT(-0400)] <Justin_o> kasper_: that's not something we've had to do much of yet... so i don't have a lot of opinions... i'm assuming you are referring to load testing... JMeter
[13:14:02 EDT(-0400)] <kasper_> only things that comes to mind for me is firebug (for speed of downloading files, responsetimes) and maybe some sort of jboss and general server monitor, for checking serverload
[13:14:11 EDT(-0400)] <Justin_o> sorrry... meant to say you could use something like JMeter
[13:14:41 EDT(-0400)] <kasper_> justin_o: ok cool.. never bumped into that before.. will take a look
[13:21:59 EDT(-0400)] * anastasiac (~team@142.150.154.193) has joined #fluid-work
[13:23:11 EDT(-0400)] * joost (~joost@pD9E26FE0.dip.t-dialin.net) has joined #fluid-work
[13:23:18 EDT(-0400)] * joost (~joost@pD9E26FE0.dip.t-dialin.net) has left #fluid-work
[13:43:34 EDT(-0400)] * colinclark (~colin@142.150.154.101) has joined #fluid-work
[14:31:42 EDT(-0400)] <Justin_o> kasper_: I just finished setting up a local version of bamboo for myself... how are things going for you?
[14:32:42 EDT(-0400)] <kasper_> justin_o: alright i guess.. but very slow too
[14:33:00 EDT(-0400)] <kasper_> just created a task for stopping the nuxeo server
[14:33:22 EDT(-0400)] <kasper_> (btw: we use too jboss instances, a nuxeo and a cspace)
[14:33:30 EDT(-0400)] <kasper_> but it fails for some reason
[14:33:44 EDT(-0400)] <kasper_> it's the damn cargo tool messing with my head again
[14:34:27 EDT(-0400)] <kasper_> justin_o: I'm by the way sorry for not taking more advantage of you offering your help
[14:34:54 EDT(-0400)] <Justin_o> kasper_: np problem.. i'm still getting things set up here on my end so that i'll be able to help you better (smile)
[14:35:10 EDT(-0400)] <kasper_> justin_o ok, great
[14:35:38 EDT(-0400)] <Justin_o> on that note, the get and install section of the quick install page on the wiki, is that stuff done once and then its done... or do things there change from release to release
[14:37:48 EDT(-0400)] <kasper_> justin_o: they change from release to release
[14:38:23 EDT(-0400)] <Justin_o> what are they for exactly
[14:38:43 EDT(-0400)] <kasper_> justin_o: But actually i think you would be ok, by just downloading the newest war/xml files
[14:39:13 EDT(-0400)] <kasper_> well.. the tar.gz files are for installing the jboss servers
[14:39:37 EDT(-0400)] <kasper_> the war files and xml are the latest UI and app layer files
[14:40:35 EDT(-0400)] <kasper_> but we're having a discussion right now on our list about packing everything in a bit tar.gz file, so you would just have to download, unzip, change host-name in two files, and run a database script
[14:43:22 EDT(-0400)] <Justin_o> kasper_: okay... thanks... that makes sense
[14:43:43 EDT(-0400)] <Justin_o> so the jboss files don't change from release to relesae.. .it is just the war and xml ones that might
[14:43:46 EDT(-0400)] <Justin_o> ?
[15:01:21 EDT(-0400)] * anastasiac (~team@142.150.154.193) has joined #fluid-work
[15:08:19 EDT(-0400)] <kasper_> oh, sorry.. didn't see you question justin_o... I'm actually a bit unsure about that... I'll ask just ask the service team... but I dont think it changes
[15:08:49 EDT(-0400)] <Justin_o> kasper_: thanks
[15:15:33 EDT(-0400)] <Justin_o> kasper_: do you know why those two war files and the xml file are on an ftp server and not with the source code for example?
[15:18:21 EDT(-0400)] <kasper_> justin_o: .. not really understanding the question .. The war files can be generated from an application/UI build/deploy (respectively), but this requires getting the source code, and running the build/deploy command .. to avoid these steps on install, the files are simply retrievable via ftp. . does that answer your question?
[15:18:41 EDT(-0400)] <kasper_> justin_o: the source code downloaded in the install instructions is only the source from the service layer
[15:18:51 EDT(-0400)] <kasper_> s/from/of/
[15:19:05 EDT(-0400)] <Justin_o> kasper_: okay... got that now... thanks
[15:20:11 EDT(-0400)] <kasper_> justin_o: so, just an update: looks like i got steps 1-4 running: stop the two jboss servers, build and deploy service layer
[15:20:24 EDT(-0400)] <kasper_> http://qa.collectionspace.org:8085/browse/NIGHTLY
[15:20:51 EDT(-0400)] <kasper_> I've set it up so the NIGHTLY-NUXSTOP task runs every 24 hours
[15:21:46 EDT(-0400)] <kasper_> the NIGHTLY-CSPSTOP runs everytime NIGHTLY-NUXSTOP is done (NIGHTLY-CSPSTOP being the task that stops cspace-jboss server)
[15:22:26 EDT(-0400)] <kasper_> the NIGHTLY-SVCBLD (build service layer) runs when NIGHTLY-CSPSTOP is done
[15:22:57 EDT(-0400)] <kasper_> and finally NIGHTLY-SVCDPL (deploy service layer) runs when NIGHTLY-SVCBLD is done
[15:24:27 EDT(-0400)] <Justin_o> kasper_: that sounds good
[15:24:29 EDT(-0400)] <kasper_> justin_o: re the jboss question – The JBoss tar is usually stable but can change from release to release. Usually our Ant "deploy" target updates whatever is needed, but sometimes we need to change something directly in the JBoss installation. Probably best to check with Sanjay in reference to the current release.
[15:24:57 EDT(-0400)] <Justin_o> can you send me the info on how you created those task so i can follow along...
[15:25:02 EDT(-0400)] <kasper_> "Probably best to check with Sanjay in reference to the current release." – ignore that .. was just copy pasting Richard Millets answer (smile)
[15:25:25 EDT(-0400)] <Justin_o> (smile)
[15:33:23 EDT(-0400)] <kasper_> justin_o: gah.. just ran into an annoying issue i have been having a bad feeling about
[15:33:25 EDT(-0400)] <kasper_> http://qa.collectionspace.org:8085/browse/NIGHTLY-NUXSTART/log
[15:33:50 EDT(-0400)] <kasper_> I'm starting the NUXEO server using the cargo tool, via maven
[15:34:06 EDT(-0400)] <kasper_> It gives you control over starting/stopping jboss, which is all good
[15:34:22 EDT(-0400)] <kasper_> problem is that it by default kills jboss after mvn is done
[15:34:41 EDT(-0400)] <kasper_> this is of course not viable, so i pass an option to continue untill you hit CTRL+C
[15:35:09 EDT(-0400)] <kasper_> problem is now that bamboo doesn't realize that it is done, when it stalls on CTRL+C
[15:35:49 EDT(-0400)] <Justin_o> kasper_: so it times out?
[15:36:24 EDT(-0400)] <kasper_> justin_o: it hasn't yet (smile) .. but the problem is, that the next plan is dependent on the server actually starting
[15:36:35 EDT(-0400)] <Justin_o> ah i see
[15:37:32 EDT(-0400)] <Justin_o> why are you using maven to start the server? Is it part of the current maven script that is used to deploy the service layer?
[15:38:57 EDT(-0400)] <kasper_> Justin_o: nope .. actually I'm completely new to maven (and hate it already) ... I followed an advice from our tech-list to look into the cargo tool..
[15:39:27 EDT(-0400)] <kasper_> the thing is we need to know if the server is properly started
[15:39:46 EDT(-0400)] <kasper_> using cargo should be able to tell you that – and fail if it doesn't properly start
[15:39:58 EDT(-0400)] <kasper_> ... which i figured would play nicely with bamboo
[15:40:27 EDT(-0400)] <kasper_> ... The alternative, i figured, would be some shell scripting, starting the server, and attempting to grep on success message
[15:41:12 EDT(-0400)] <kasper_> justin_o: but any alternatives would be very welcome
[15:41:31 EDT(-0400)] <Justin_o> kasper_: i was thinking about the shell script approach too, as this maven tool does not seem to be working
[15:42:25 EDT(-0400)] <kasper_> justin_o: so the issues would be to have the shell script return a success or error return code on successful/failed jboss start
[15:42:53 EDT(-0400)] <kasper_> oh, yeah.. there is the issues of stopping jboss too
[15:43:24 EDT(-0400)] <kasper_> but i guess that could be handled with a kill command, and maybe a ps or something to find the pid of jboss
[15:48:53 EDT(-0400)] <Justin_o> kasper_: i think you could just call the shutdown script
[15:49:43 EDT(-0400)] <kasper_> lol... there is a shutdown script?
[15:50:07 EDT(-0400)] <kasper_> ... jesus... read the f**ing manual kasper!!!
[15:51:10 EDT(-0400)] <Justin_o> (smile) in jboss-4.2.3.GA that i downloaded earlier there is both a run.sh (also bat and jar) and shutdown.sh (also bat and jar)
[15:52:06 EDT(-0400)] <kasper_> justin_o: so would the following make sense to you:
[15:53:15 EDT(-0400)] <kasper_> Create a script to: (1) start jboss in background (2) grep for successfull/fail start message in log (3) on successfull/fail message, return the relevant error code
[15:55:11 EDT(-0400)] <Justin_o> kasper_: that sounds right
[15:56:05 EDT(-0400)] <kasper_> justin_o: ok, I'll get started on that
[15:56:18 EDT(-0400)] <jamon> what are you guys trying to do kasper_, Justin_o?
[15:56:56 EDT(-0400)] <Justin_o> jamon: script to start a jboss server and verify that it is running
[15:57:10 EDT(-0400)] <Justin_o> for a build server that will be used by collectionspace
[15:58:04 EDT(-0400)] <jamon> ah gotcha
[15:58:18 EDT(-0400)] <jamon> try using only catalina.sh instead of the various run/shutdown etc. scripts
[15:58:45 EDT(-0400)] <jamon> they all just hook into it, and catalina.sh explicitly describes how to run in foreground or as a forked background process
[16:02:31 EDT(-0400)] <Justin_o> jamon: in my install of jboss i can't seem to find the catalina.sh script... do you know where it is locatec
[16:06:47 EDT(-0400)] <Justin_o> kasper_: i'm going to be heading home now... we can talk about all this tomorrow too...
[16:07:20 EDT(-0400)] <kasper_> justin_o: ok, thanks a lot for your help
[16:08:11 EDT(-0400)] <Justin_o> kasper_: np..
[16:08:17 EDT(-0400)] <kasper_> jamon: where is the catalina.sh script located?
[16:08:58 EDT(-0400)] * Justin_o (~Justin@142.150.154.171) has left #fluid-work
[16:09:07 EDT(-0400)] <jamon> my mistake guys, catalina, that's specific to tomcat
[16:11:14 EDT(-0400)] <kasper_> jamon: ah ok. Thanks anyways
[16:11:35 EDT(-0400)] <jamon> sorry about that
[17:17:18 EDT(-0400)] * anastasiac (~team@142.150.154.193) has left #fluid-work
[19:49:59 EDT(-0400)] * colinclark (~colin@bas2-toronto09-1176131031.dsl.bell.ca) has joined #fluid-work
[20:44:27 EDT(-0400)] * athena7 (~athena@adsl-99-62-112-41.dsl.wlfrct.sbcglobal.net) has joined #fluid-work