19:01:28 <mtaylor> #startmeeting
19:01:29 <openstack> Meeting started Tue May 29 19:01:28 2012 UTC.  The chair is mtaylor. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:01:30 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:01:59 <LinuxJedi> o/
19:02:06 * LinuxJedi appears to be the only one :)
19:02:25 <mtaylor> LinuxJedi: we'll give folks a few to straggle in
19:02:35 <clarkb> I am here
19:02:58 * clarkb read mtaylor's "aroo" in Nixon from futurama's voice
19:03:08 <LinuxJedi> hehe :)
19:03:21 <mtaylor> well...
19:03:31 <mtaylor> LinuxJedi: how about you tell us whatcha been doing?
19:03:43 <LinuxJedi> mtaylor: sure
19:04:05 <soren> o/
19:04:06 <LinuxJedi> so, lots of little things this week
19:04:24 <LinuxJedi> Stackforge now has a Gerritbot running on #stackforge-dev (after requests from heat)
19:05:04 <LinuxJedi> Jenkins Jobs has lots of minor updates and on more Openstack projects (I'm sure mtaylor will speak about that).  I have done documentation updates and blogged about it
19:05:49 <jeblair> o/
19:05:52 <LinuxJedi> We had an issue with pep8 where some new rules in a new release broke lots of projects during a release.  A slight modification to tox.ini worked around it for now
19:06:15 <LinuxJedi> All Stackforge jobs now run on the new Jenkins jobs filler too
19:06:34 <LinuxJedi> Stackforge's disk filled again due to my error, shouldn't happen again.
19:06:54 <LinuxJedi> And the openstack-ci-puppet tree is now tested with pyflakes
19:07:10 <LinuxJedi> and about a billion little things.
19:07:13 <LinuxJedi> any questions?
19:07:18 <mtaylor> it's been a week :)
19:07:59 <clarkb> do we expect openstack to honor pep8 1.1 at some point? I suppose we don't really worry about it too much?
19:08:10 <LinuxJedi> clarkb: yes
19:08:21 <LinuxJedi> I don't have more of an answer than that though
19:08:23 <mtaylor> several of the projects have fixed themselves alread
19:08:42 <notmyname> LinuxJedi: is the gerrit hook fixed from this weekend? what was up with that?
19:08:55 <mtaylor> the underlying build bug is that the pep8 version specified in tools/test-requires is not the thing pulling in the pep8 for the test...
19:08:57 <notmyname> mtaylor: swift is 1.1 compliant now :-)
19:08:59 <mtaylor> so there was a version mismatch
19:09:00 <LinuxJedi> notmyname: ah, yes, so...
19:09:09 <mtaylor> we're working on a better solution for that systemically
19:09:26 <mtaylor> unfortunately, there is a setuptools/pip disconnect which is really frustrating
19:09:29 <LinuxJedi> gerrit trigger broke over the weekend.  It appears to be due to a stuck thread in Jenkins.  A restart fixed it
19:09:40 <LinuxJedi> but we are working on making such things less likely
19:09:49 <mtaylor> in the future, jeblair's work should obviate that problem moving forward
19:09:53 <mtaylor> what LinuxJedi said
19:09:55 <mtaylor> :)
19:10:20 <LinuxJedi> just one of the billion things I fixed this week ;)
19:10:32 <mtaylor> yah
19:10:51 <mtaylor> as a follow up from me on one of LinuxJedi's things ... all of the openstack projects are on jenkins job filler now
19:11:25 <mtaylor> so things like coverage or docs jobs that break at this point are things we should treat as bugs in the project to be fixed there
19:11:55 <mtaylor> which means I think we should be able to annoy people on failures in non check-* or gate-* jobs
19:12:03 <LinuxJedi> blog post at http://www.linuxjedi.co.uk/2012/05/openstacks-jenkins-job-filler.html by the way for people who want to learn more about it
19:12:30 <mtaylor> cool. thanks LinuxJedi
19:12:41 <mtaylor> clarkb: how's life in your world?
19:12:42 <jeblair> #link http://www.linuxjedi.co.uk/2012/05/openstacks-jenkins-job-filler.html
19:12:45 <annegentle> will this jenkins job filter fix the sphinx doc builds (autodoc python modules problems)?
19:13:22 <annegentle> P.S. nice job on the docs at http://ci.openstack.org/jenkins_jobs.html
19:13:24 <mtaylor> annegentle: no, but we will be building all of the docs jobs the same now, so there's one less variable to track down
19:13:32 <annegentle> +1
19:13:38 <clarkb> So late Friday I thought my gerrit patches were all ready and good to go when I realized that I had been working off of master which is old.
19:13:39 <LinuxJedi> annegentle: thx :)
19:13:58 * mtaylor apologizes to clarkb for that
19:14:06 <clarkb> I spend the majority of my time last week getting those going
19:14:38 <clarkb> after some quick cherry picking this morning I think those patches are good to go again
19:14:53 <clarkb> (they have been reapplied to the openstack/2.4 branch)
19:15:15 <clarkb> I lucked out with that as I don't think Shrews had an easy time reapplying the commits to the newer branch
19:15:59 <clarkb> I think that is it for me. No exciting fires were put out
19:16:05 <mtaylor> cool. thanks!
19:16:30 <mtaylor> (speaking of, I kinda broke the gerrit gating jobs this weekend, which I have suitably chastised for - hopefully they're better now)
19:17:15 <LinuxJedi> between me and mtaylor we broke all jobs for a very short time on Sunday, but no jobs were run in this time :)
19:17:25 <mtaylor> :)
19:17:38 <LinuxJedi> apart from Stackforge but I don't think anyone noticed
19:17:51 <mtaylor> jeblair: anything on the make-things-parallel front?
19:18:32 <jeblair> yep
19:18:47 <clarkb> looks like jenkins just +1ed my most recent push to gerrit so I think it is working again
19:18:47 <jeblair> gerrit-git-prep now supports testing dependent changes
19:19:22 <jeblair> so you can tell it "merge change A, B, then C and test"
19:19:49 <jeblair> which is a pre-requisite for speculative execution of jenkins gating tests
19:19:55 <notmyname> (oh, and thanks LinuxJedi for helping me merge things during a holiday weekend)
19:20:11 <LinuxJedi> notmyname: wasn't a holiday here in the UK ;)
19:20:42 <notmyname> LinuxJedi: it was still Sunday pm central time. very late for you either way :-)
19:20:46 <mtaylor> clarkb: w00t!
19:20:51 <jeblair> that's in now, but not being used; when we switch to the new parallelized gating system, we'll start using that.
19:20:54 <jeblair> that's it.
19:21:03 <mtaylor> jeblair: I'm excited by that
19:21:18 <mtaylor> Shrews: you wanna punch me in the kidney about work in progress?
19:21:37 <Shrews> mtaylor: emphatically yes
19:22:00 * LinuxJedi thinks Shrews doesn't want the Java mug I bought for him :/
19:22:24 <Shrews> plumbing appears to have changed rather significantly in 2.4, so i need to re-do a lot of that. but at least this time i know where to look for things
19:22:37 <Shrews> so yay
19:22:55 <mtaylor> that's good
19:23:16 <mtaylor> (for folks following at home, same problem as clarkb -- I forgot to tell clarkb and Shrews about how we're tracking gerrit upstream...)
19:23:56 <LinuxJedi> so ignore the random bullet holes in mtaylor's torso
19:24:10 <mtaylor> yeah
19:24:30 <mtaylor> I guess from me, other than breaking things - I've been working on getting jclouds unittest builders running
19:24:56 <mtaylor> which would be easy - except for the setup time it takes to launch and configure a build slave
19:25:16 <mtaylor> I was going DEEP trying to get it sorted, but it seems I may have left a zero off of the timeout - so I'm re-testing :)
19:26:15 <mtaylor> so hopefully that'll be done this week ... as will proper support for maven jobs from jenkins job filler
19:26:37 <mtaylor> anybody else got anything?
19:26:54 <LinuxJedi> mtaylor: do you want me or you to do maven stuff?
19:27:55 <clarkb> mtaylor: you mentioned that some of our changes to gerrit in 2.3 would not be applied to 2.4
19:28:19 <mtaylor> LinuxJedi: I'll poke at it ... the current maven job thing was my doing :)
19:28:23 <mtaylor> LinuxJedi: unless you get bored :)
19:28:41 <clarkb> and do we have a plan for upgrading gerrit to 2.4?
19:28:47 <LinuxJedi> mtaylor: I have a few things I was planning to do first, if you haven't started by Thursday I can take over
19:29:28 <mtaylor> clarkb: yeah - I was working on upstreaming the require-short-subject patch
19:29:57 <mtaylor> and it looks like there's a better way to do it coming up
19:30:29 <mtaylor> so I'm just going to wait on that (also, nobody's actually using it)
19:35:37 <mtaylor> ok. I think that does it
19:35:58 <mtaylor> thanks everybody!
19:36:01 <mtaylor> #endmeeting