14:02:26 #startmeeting releaseteam 14:02:27 Meeting started Fri Jul 15 14:02:26 2016 UTC and is due to finish in 60 minutes. The chair is dhellmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:02:28 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:02:30 The meeting name has been set to 'releaseteam' 14:02:40 our agenda today is under R-12 of https://etherpad.openstack.org/p/newton-relmgt-tracking 14:02:57 courtesy ping: ttx, dims, lifeless, tonyb, stevemar, fungi 14:03:17 o/ 14:03:41 hi, dims 14:03:55 today should be a relatively quick meeting 14:03:58 #topic n-2 milestone status 14:03:59 yep 14:04:02 best kind 14:04:08 we have a couple of deliverables without tags 14:04:20 other than telemetry, I think all of them won't actually get released this cycle 14:04:27 I've sent email to the list asking for confirmation of that 14:04:44 #link http://lists.openstack.org/pipermail/openstack-dev/2016-July/099335.html 14:04:46 yep saw a response for the trove-image-builder 14:05:14 jd__ has posted a review for the telemetry releases, so once that passes validation we should go ahead and tag it 14:05:25 do we have to adjust tags in governance repo? 14:05:26 we'll see what the other ptls say about their items 14:05:29 ack dhellmann 14:05:47 * jd__ hurrah 14:05:54 yeah, I think after ttx and I get back from europython we should talk about moving those other deliverables to independent to avoid confusion at the end of the cycle 14:06:20 dims : can you handle the telemetry tags today? 14:06:54 dhellmann : let me look for the review 14:07:00 yes i can 14:07:48 ok, thanks 14:07:48 found it https://review.openstack.org/#/c/342862/ 14:08:04 oh, yeah, sorry, I had that link in the etherpad 14:08:37 i should have looked there first :) 14:08:51 most of the projects were on time, and I processed a late rush last night, so it's good that we only have one straggler. folks seem to be getting the idea of following the schedule. :-) 14:09:11 :) 14:09:26 #topic release automation update 14:09:38 I started working on a job to run release_from_yaml.sh 14:09:46 #link https://review.openstack.org/#/c/341058/ 14:10:33 fungi : have you had a chance to look at that? 14:10:34 the release bot account exists and is on the release team now 14:10:41 #link https://review.openstack.org/#/admin/groups/11,members 14:10:57 and i have a couple puppet changes proposed to get its account details available on the signing node 14:11:00 fungi : ah nice 14:11:07 #link https://review.openstack.org/#/q/topic:artifact-signing 14:11:26 those should be prerequisites for 341058 but looking closer at it this morning now that those are out of teh way 14:11:38 ok, cool 14:12:03 are we going to bring the bot's keys to a signing party? :) 14:12:14 I was particularly worried about whether I was setting up the local git repo properly, since there are a few ways to do it and the right way depends on the job queue to some degree 14:12:21 dims : it's on the todo list, yes 14:12:38 dims : we're waiting for the infra team to sign it, then we can have a virtual party 14:12:47 yay! 14:12:54 * stevemar sneaks in 14:12:54 with virtual cake! 14:13:06 dims: also as soon as i get another core review on the documentation for our key management, i'll get a majority of the infra root sysadmins to sign it 14:13:10 minecraft based! 14:13:16 #link https://review.openstack.org/339248 14:13:24 thanks fungi 14:13:24 fungi : so are we blocked on anything other than review time? it sounds like not, but I want to make sure... 14:13:41 dhellmann: i _think_ all the pieces are up for review at this point, correct 14:13:47 ok, great, thank you 14:13:56 is there anything else to mention this week? 14:14:01 reviewers will hopefully let us know if not, as will testing 14:14:12 not that i know of 14:14:35 yeah, if that job merges next week I'll test it the monday following when I'm back on my usual schedule 14:14:46 or maybe next friday, if I'm not a zombie 14:14:47 sounds good 14:14:57 #topic zuul-cloner issue 14:15:16 yesterday we discovered an issue with the way we were cloning repos for the validation job on openstack/releases 14:15:36 some assumptions I made about zuul-cloner and our environment proved incorrect 14:15:59 fungi and jeblair helped debug and we have fixes in place now, but if you're running any of our tools locally you want to update to the latest versions 14:16:16 I have not applied the fix to the release-tools repo, yet, I just realized 14:16:57 ouch ok. i do pull down latest for project-config, release-tools and releases before i cut anything 14:17:03 ok, good 14:17:28 the tl;dr is that we weren't actually cloning enough of the repo to get all of the tags not on the master branch (or whatever branch was being cloned) 14:17:49 i see 14:17:51 see https://review.openstack.org/342396 for the fix in project-config 14:18:04 clarkb also has a related fix to get the additional refs cached on our workers (we thought we were doing that but it turns out we had a bug there too) 14:18:14 ok, that's good to know 14:18:48 #link https://review.openstack.org/342461 14:18:53 dims : the symptoms were weird validation errors where the tag was recognized as already existing, but then not found when the sha match test ran 14:18:54 ah 14:19:15 the difference in behavior is because "is there such a tag" is done by querying cgit and the other looks at the local repo, and they were out of sync 14:19:59 so after much head-scratching fungi and jeblair helped figure out the minimal set of update commands to fix that 14:20:25 git tries to be cautious about fetching unnecessary refs 14:20:37 so only fetches tags for objects in the history of tracked branches 14:20:51 makes sense 14:20:57 this is a weird need, for these tools 14:20:57 this becomes especially odd in situations like eol releases whose branches have been deleted 14:21:12 we need a lot more history than most jobs 14:21:17 yep 14:21:20 thanks fungi and jeblair ! 14:21:25 and now ye shall have them 14:21:29 woot 14:21:44 the scary thing is how this has seemed to be working in the past 14:21:47 lucky I guess 14:22:03 let's move on and see if we can finish under 30 min 14:22:06 #topic priority reviews 14:22:17 we have some tooling reviews up and some validation reviews in openstack/releases 14:22:20 those need attention 14:22:34 we have a bunch of constraint updates open still, but since it's friday maybe those should wait for next week 14:22:53 I just submitted a change to fix neutron's deliverable files so they match what the governance repo has listed 14:22:58 #link https://review.openstack.org/#/c/342859/ 14:23:08 is there anything else folks are aware of that needs attention? 14:23:30 nothing from me dhellmann. 14:23:39 ok cool 14:23:46 #topic open discussion 14:23:54 is there anything else to raise this week? 14:24:40 not from me anyway 14:24:44 nothing from me 14:24:45 just that the storyboard api docs have some examples now: http://docs.openstack.org/infra/storyboard/webapi/v1.html 14:24:56 oh, nice, thanks anteaya 14:25:01 I'll put that on my reading list 14:25:16 so if release team folks are willing to take a look to see if that addresses release team needs from a task tracker I'm looking for feedback 14:25:18 thanks 14:25:41 the question I am asking a small group of folks right now is, if we were to migrate tomorrow, could you get work done? 14:25:47 #link http://docs.openstack.org/infra/storyboard/webapi/v1.html 14:26:04 if something is blocking your current workflow or is missing I'd like to discuss it 14:26:06 and thanks 14:26:09 yep, I'll look them over -- probably late next week -- and see how they look 14:26:19 just talking to specific folks right now 14:26:23 dhellmann: thank you 14:26:39 ok, if there's nothing else, we can call the meeting early... 14:26:49 sounds good dhellmann 14:26:59 ok, have a good weekend everyone! 14:27:02 #endmeeting