13:04:26 #startmeeting releaseteam 13:04:26 Meeting started Mon Sep 28 13:04:26 2015 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:04:27 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:04:30 The meeting name has been set to 'releaseteam' 13:04:52 Agenda @ https://wiki.openstack.org/wiki/Release_Cycle_Management#Meeting 13:05:11 First item, two governance reviews requesting addition to release:managed 13:05:29 https://review.openstack.org/#/c/226933/ is for Octavia 13:05:43 a project under Neutron 13:06:07 https://review.openstack.org/#/c/227741/ for gnocchiclient 13:06:50 the gnocchiclient one is easy, since I think that was just an oversight from when the server project was added 13:06:54 I've already RC+1 that 13:07:02 except the server is not release:managed 13:07:17 it's independent 13:07:19 oh, I thought it was? 13:07:27 I don't think we want to mix and match 13:07:37 so that would be a release:independent + release:managed client lib, which would make it quite unique 13:07:52 yeah, those are suppose to be mutually exclusive 13:08:03 if gnocchi moves to release:cycle* then it surely makes sense 13:09:29 for octavia... feels a bit weird to jump to release:cycle-with-milestones and release:managed in one go 13:09:47 I'd prefer they prove that they can follow a milestone model first before having us onto them 13:09:49 what did they do this cycle? 13:09:56 independent 13:10:09 do we know if they did milestones? 13:10:21 they just did one tag 5 days ago 13:10:28 so it's pretty rawa 13:10:30 raw 13:10:42 ok, I agree, let's get them onto the milestone system before we agree to manage them 13:11:01 You going to comment on the reviews ? Or should I ? 13:11:02 I've left feedback on the gnocchiclient request, do you want to respond to the octavia request? 13:11:22 I'll leave it to you 13:11:25 ok 13:11:36 I can +1 you if it gets tough 13:11:58 #topic Liberty release coordination 13:12:13 #link https://etherpad.openstack.org/p/liberty-final-release 13:12:24 We are at release week -2 13:12:44 We caught up with the timeline over the last week 13:13:13 This week is supposedly calm 13:13:30 We should create the requirements branch and end the freeze, maybe today ? 13:13:43 let me look at the open reviews... 13:14:03 We may or may not have to disable the cross-check job 13:14:15 We may start issuing early RC2s 13:14:29 it looks like we don't have any requirements changes that we need, so we could go ahead and branch 13:14:31 and we may have a Swift liberty release 13:14:44 yes, we expect that by the end of the week 13:14:51 I expect the bulk of RC2s to come next week when we can finalize the translations imports 13:15:19 makes sense 13:15:26 let me have a quick look at the cross-check thing, see if we need to disable anything 13:15:31 * dhellmann makes a note to document this release schedule in the project-team-guide 13:16:46 I feel like we need to at least revert https://review.openstack.org/#/c/217300/2/zuul/layout.yaml 13:17:21 I think you're right, but see lifeless' comment in https://etherpad.openstack.org/p/liberty-end-of-release-requirements-checklist 13:17:33 yeah, that's what makes me pause 13:17:50 we added tests, and now it seems those would be testing the same thing twice, so I think we can remove them, but his comment makes me doubt my reasoning 13:18:26 we have a branch override defined there, and we want it to use stable/liberty instead 13:18:49 Let's wait a bit 13:19:05 I'll ask clarkb and you can ask lifeless 13:19:10 ok 13:19:22 no need to rush it, not eaxactly that much pressure to add deps yet 13:19:50 yeah 13:19:59 we've had a few requests, but no one complained too loudly about having to wait 13:20:10 #action ttx to see with clarkb what (if anything) to do to disable the extra cross-check job 13:20:47 #action dhellmann to check with lifeless what he meant re: disabling crosscheck 13:21:15 #agreed wait until Tuesday morning to create teh requirements stable/liberty branch 13:21:55 Otherwise the only big thing that emerged last week release-wide is the Cinder / NetAPP lib licensing issue 13:22:07 I think it's under control now, but that may require a rush for a RC2 13:22:19 since we don't want to keep that up for long 13:22:45 yeah, that was a surprise 13:22:49 I'll be syncing with PTLs tomorrow re: RC2s 13:23:08 do you want me to assist with that? 13:23:09 see which need an urgent one and which can wait until the end of translations 13:23:28 dhellmann: sure, we'll see who I didn't get to by the end of the day 13:23:32 k 13:23:51 dhellmann: anything on your side ? 13:23:59 if you have some notes for the questions to make sure I ask, please drop them in https://etherpad.openstack.org/p/liberty-final-release 13:24:25 sure. I'll probably formulate the questions as I go, but I'll note them down there 13:25:01 I'm going to spend some time this week pulling together my thoughts about moving off of the strictly synchronized milestone release model, so we can discuss it some before the summit 13:25:02 #topic Mitaka release schedule 13:25:17 #link https://wiki.openstack.org/wiki/Mitaka_Release_Schedule 13:25:35 I'll communicate that today, unless you have a last-minute change for it 13:26:19 I think I'm going to add a column with R-n countdown so we can document the general end-of-release week highlights in those terms, but that won't change the actual schedule 13:26:30 ack, good idea 13:26:46 makes teh whole thing a bit less experimental 13:27:22 #action ttx to communicate proposed Mitaka release schedule 13:27:28 #topic Release management session content brainstorming 13:27:35 yeah, and having more of this written down is going to be more important, since I don't plan to do as much hand-holding next cycle :-) 13:27:41 #link https://etherpad.openstack.org/p/mitaka-release-management 13:29:05 I think the two themes are good. Worried we are packing a bit too much into that fishbowl, though 13:29:25 hmm 13:29:36 yeah, we might be 13:29:45 how long are the sessions, 40 minutes? 13:29:50 FOr the last point you just added, I don't expect we'll have enough of them in the room to be able to discuss that 13:29:53 yeah, 40min 13:30:14 maybe move it to the work room so that we prepare docs work 13:31:05 The goal of that workroom is to come up with a document like https://etherpad.openstack.org/p/liberty-release-mgmt to outline the work we plan to do during the cycle 13:31:24 if we have time left we can start working on it, but it's more to come up with a clear Mitaka plan 13:31:36 ++ 13:31:50 I found that page very useful to try to organize and keep up with what we palnned to do, so I epxect you'll like it too :) 13:32:06 except we kind of built it along the way, and it's easier to capture while it's hot 13:32:10 yeah, I did, too 13:32:36 so if we can go out of Tokyo with it already compiled, that would really help bootstrap the work afterwards 13:32:41 this is normally the sort of thing a project team would use blueprints for, but in our case the etherpad worked really well for keeping notes as we went 13:33:18 We can use the "contributors meetup" to do the heavy lifting on the tooling refactorig maybe 13:33:28 I like that 13:33:43 I have a bunch of the python rewrite started but not submitted for review 13:34:04 I'm not sure those patches are in a good state for reviews, but I can take a look at them and clean them up before the summit 13:35:29 #topic Open discussion 13:35:40 OK, that is all I had for the meeting 13:35:46 while I'm thinking of it, I'm probably taking the day of wed for my birthday 13:35:58 great 13:36:22 We could discuss the talk sometimes in the next 2 hours ? 13:36:42 Like agree on the outline, split the work 13:36:54 whenever you can in the next 2 hours I mean 13:37:08 that's a good idea. I need a few minutes, but could do it more or less right after this meeting if you want 13:37:40 ok, let's say in 23 min, I have a few pings I need to answer 13:37:49 ok, sounds good 13:38:00 ok, let's close this 13:38:04 #endmeeting