19:01:02 #startmeeting 19:01:03 Meeting started Tue Feb 28 19:01:02 2012 UTC. The chair is mtaylor. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:01:04 Useful Commands: #action #agreed #help #info #idea #link #topic. 19:01:28 #topic CI Discussion 19:01:39 hey everybody. anybody around? 19:01:51 Me, sort of 19:01:51 o/ 19:01:59 o/ - but I might have to run off... 19:02:06 Oh crap, I do have to run off. Nevermind 19:02:50 heckj: anything you need/want from us before you go? 19:04:09 guess not 19:04:36 jeblair: you had a couple of things you thought we should bring up today, yeah? 19:04:53 yeah, people keep asking about stackforge 19:05:00 good call. 19:05:01 ++ 19:05:04 #topic stackforge 19:05:34 if anybody has been lurking, we've (and by we, I mean LinuxJedi|cell) been working on spinning up a second copy of the OpenStack Infrastructure 19:06:01 the intent is for it to be a place for satellite projects or other things openstack related but which have no intent or hope of become core to go should they wish 19:08:06 Is the intent to let people run it themselves or will "we" run it? 19:08:16 For whatever value of "we" makes sense. 19:10:03 well, both 19:10:10 we're also working on getting more and more stuff in to puppet 19:10:33 so that, other than having root on the machine to diagnose odd failures, most of "running it" should be doable through code reviews 19:10:42 (same will apply to the official openstack one as well) 19:11:05 and actually - the goal is for both systems to work from the same set of puppet modules, so that work done on one applies to both 19:11:27 Ok, cool. 19:11:59 we've been batting around ideas as well for how to manage the actual jenkins job content via puppet too, so that we can get out of the business of manually editing jobs in web forms and then forgetting some of them 19:12:23 that'll be WAY easier to do sensibly if we can get folks to buy in to http://wiki.openstack.org/ProjectTestingInterface 19:13:02 since then we can have a set of vetted job description templates that apply to $PROJECT ... and then a list of projects - you know - modularization and code reuse! 19:14:38 ok. I think that's the gist of that... 19:14:45 jeblair: you wanna say anything about gerrit trigger? 19:14:53 ya 19:14:58 #topic gerrit trigger plugin 19:15:10 so i upgraded jenkins and all the modules on jenkins.o.o last night 19:15:47 and when i went to install the new version of the gerrit trigger plugin, jenkins wasn't happy with it and wouldn't load it 19:15:58 so i rolled that part back, the rest of the upgrade is still in place 19:16:10 k 19:16:24 today i'm setting up jenkins-dev.openstack.org so we can have a machine that's much closer to production to test out upgrades on 19:16:30 any idea what wasn't happy? or do you need the new jenkins dev server to fully figure that out? 19:16:33 ++ 19:16:33 (side effect: better puppet modules) 19:16:57 yep. because of course i ran this module on a simple jenkins test server many times without seeing that error 19:17:23 so it could be something triggered by a different version of jenkins or another plugin 19:17:29 or something about upgrading the module 19:17:38 so i'll work on replicating and fixing that 19:17:56 but once that's done, we should have a gerrit trigger plugin that behaves a little more correctly 19:18:06 w00t! 19:18:12 i've been working on upstreaming our changes 19:18:24 and that code review process has pointed out some things we could be doing better 19:18:45 so the new version should avoid some race conditions we have now 19:19:13 and i'm also adding a ref-updated trigger, so that we can run post-merge jobs triggered directly from gerrit 19:19:49 and finally, we have a job in jenkins that builds the plugin .hpi file we're going to use 19:19:57 (looking for link) 19:20:11 #link https://jenkins.openstack.org/job/gerrit-trigger-plugin-package/ 19:20:25 so once we get this working, we can point people right at the build we're using 19:20:40 i think that's about it 19:20:45 awesome 19:21:25 mildly related thought to the two previous things - but LinuxJedi|cell has installed gerrit 2.2.2 on stackforge, which lets us run innodb 19:21:53 once that's been tested there, we can roll that out to openstack, and then have a similar job with artifacts for gerrit as well 19:22:01 * mtaylor is happy about the idea of that 19:22:11 Me too :) 19:22:27 yeah, we should also remember to move gerrit to a new server 19:24:54 ++ 19:25:27 I'll work on improving the puppet gerrit modules for that at some point 19:26:31 sweet 19:26:37 #topic open discussion 19:26:40 anybody else? 19:27:19 * soren has nothing useful to say 19:27:20 meetbot is nearly done but on hold 19:27:26 awesome 19:27:28 * LinuxJedi (back home now) 19:31:55 bcwaldon: hello 19:32:18 hey! it's anotherjesse! 19:32:35 hi mr taylor 19:32:42 is there a first jesse? 19:33:02 LinuxJedi: yes, but he is a bit of punk 19:33:11 lol :) 19:33:15 unless anotherjesse has anything for us, I think we're done with this meeting 19:33:28 (anybody have any food? I'm hungry!) 19:33:36 * LinuxJedi has birthday cake 19:33:57 lots of little ones with edible lego bricks on 19:34:08 * LinuxJedi give mtaylor a cake 19:34:19 mmm. tasty 19:34:27 #endmeeting