15:04:31 <smcginnis> #startmeeting releaseteam
15:04:31 <openstack> Meeting started Fri Feb 16 15:04:31 2018 UTC and is due to finish in 60 minutes.  The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:04:32 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:04:34 <openstack> The meeting name has been set to 'releaseteam'
15:04:45 <smcginnis> Ping list: dhellmann, dims, fungi, tonyb, lbragstad, ttx, armstrong
15:04:54 <dhellmann> o/
15:05:06 <smcginnis> #link https://etherpad.openstack.org/p/queens-relmgt-tracking Agenda/Tracking
15:06:34 <smcginnis> Guess it's a small group today. :)
15:06:45 <smcginnis> #topic RC status
15:07:00 <smcginnis> We have a few teams that have put up RC2 patches.
15:07:12 <smcginnis> A few already went through, and a few are up with -W indicating more to come.
15:07:26 <smcginnis> I've also heard from a few other groups that they are planning on early next week for another.
15:07:47 <smcginnis> Anyone aware of any issues with teams prepping an RC2|3?
15:08:39 <fungi> just tripleo getting surprised by stable/queens branches, but that's not really rc-specific
15:08:59 <smcginnis> Yeah... that was an issue.
15:09:09 <fungi> more them not realizing they had some deliverables in cycle-with-intermediary which missed the branching deadline
15:09:13 <smcginnis> Thanks to infra for helping sort out the cleanup.
15:09:31 <smcginnis> They are at least aware now and plan on changing their model for rocky.
15:10:03 <smcginnis> That was the only issue I saw with us forcing branching.
15:10:05 <dhellmann> maybe that's also an indicator for us that we shouldn't create branches for folks?
15:10:33 <smcginnis> Well, that would be ideal.
15:10:37 <dhellmann> we should think through the ramifications of a policy that says if they don't create a branch, they just don't have one.
15:10:47 <fungi> the flip side is that things start breaking for some projects after requirements unfreezes
15:10:53 <smcginnis> I thought we did though, and we determined we kind of had to.
15:11:12 <dhellmann> we did because there were already releases
15:11:29 <dhellmann> so maybe during that period we are more strict about requiring them to indicate they are ready to branch
15:12:07 <dhellmann> I'd have to go through the logs again to find the reasoning I came up with, but I thought it was tied to them having releases that weren't going to be part of the branch or something
15:12:27 <dhellmann> or maybe I'm mixing up 2 issues
15:12:32 <smcginnis> I've worked ahead on countdown emails for Rocky, and I've tried to mention several times that the release team does not want to be the ones initiating any kind of release or branching.
15:12:36 <fungi> if memory serves, one concern in the still-syncing-reqs era is that if we branch things like libs from a point after the requirements unfreeze and they've merged a sync from master, we can end up with incompatible reqs between stable branches of different deliverables
15:12:44 <dhellmann> fungi : ah, yes, that's it
15:12:52 <dhellmann> so, when we *stop* syncing requirements we can stop forcing branches
15:13:13 <dhellmann> I'm going to try to help prometheanfire and dirk with that this cycle
15:13:15 <fungi> probably so. or at least we can try and see what else we've forgotten makes that a problem ;)
15:13:25 <dhellmann> yes, what breaks in a new way :-)
15:15:26 <smcginnis> OK, on to bitrot.
15:15:40 <smcginnis> Andrea has some patches up to update process in the new zuulv3 world.
15:15:58 <fungi> i haven't looked at them... hopefully simpler than bulk-copying the stable/pike jobs?
15:16:02 <smcginnis> Which is much appreciated, because I looked at the previous examples and was lost. :)
15:16:21 <dhellmann> fungi : I think we end up needing 1 more patch but they didn't look that complicated
15:16:22 <smcginnis> fungi: I think so. They are open in tabs for me, but what I saw yesterday didn't look too bad.
15:16:44 <dhellmann> props to andreaf for putting them together so quickly
15:17:02 <fungi> duplicating the old legacy stable jobs would have been relatively straightforward, however it would have been a lot of patches and also carrying technical debt forward
15:17:16 <fungi> so i'm glad he had time to put that together
15:17:41 <smcginnis> Very glad and much appreciated.
15:17:57 <fungi> also the e-mail he sent to the ml (and corvus's followup) were a great recap
15:18:14 <dhellmann> yes, we discussed that and he's going to try to send out regular updates
15:18:28 <smcginnis> Anything else we should discuss about RC status?
15:18:55 <smcginnis> I feel like overall we are in pretty good shape now.
15:18:55 <fungi> #link http://lists.openstack.org/pipermail/openstack-dev/2018-February/127453.html Migration of Tempest / Grenade jobs to Zuul v3 native
15:18:56 <dhellmann> it looks like the neutron rc2 is ready for review
15:19:30 <smcginnis> dhellmann: Do you want to do Friday releases? I was thinking it might be better to wait until Monday to give time for any additional things to pop up.
15:19:45 <dhellmann> hmm, good point
15:20:07 <dhellmann> it's just an rc, so I don't know if it's better to be cautious or to let it out early to give people time to try it
15:20:17 <dhellmann> I don't know how many people actually try the release candidates these days
15:20:17 <smcginnis> Either is fine with me really, just thinking at this point, we're getting close so it might be worth waiting a bit.
15:20:23 <smcginnis> True.
15:20:38 <smcginnis> I wonder how we could get feedback on that.
15:20:55 <smcginnis> We do a lot of things to help packagers and downstream folks, but never really hear much from them.
15:21:04 <smcginnis> Unless something goes wrong of course. :)
15:21:51 <dhellmann> the first 3 components only have tox settings updates and a g-r sync, so I'm not sure those are worth worrying about. After the meeting I'll read through the rest to see what else they have and we can talk about it.
15:22:14 <smcginnis> Sounds good.
15:22:17 <dhellmann> smcginnis : we could send "tracer bullets" by building bad packages at random intervals ;-)
15:22:24 <smcginnis> Hah!
15:22:51 <smcginnis> #topic Team status next week through PTG
15:22:54 <fungi> i know at least some of the packagers do it in their "free" time and probably have more bandwidth on weekends than weekdays
15:23:12 <smcginnis> I should be around all next week, but I leave Friday afternoon for Dublin IIRC.
15:23:29 <fungi> same for me, though i leave saturday
15:23:43 <dhellmann> I'll be around all week and leave Sat as well
15:24:03 <smcginnis> Good, sounds like we can have the meeting as usual then.
15:24:12 <fungi> wfm
15:24:31 <smcginnis> One thing with the PTG - I had to reschedule the team photo.
15:24:44 <dhellmann> oh, yeah, what time slot did we end up in?
15:24:58 <smcginnis> Still right after the requirements team. Now I'm scrambling to find the spreadsheet...
15:25:26 <dhellmann> #link https://docs.google.com/spreadsheets/d/1J2MRdVQzSyakz9HgTHfwYPe49PaoTypX66eNURsopQY/edit?usp=sharing
15:25:42 <smcginnis> Thanks!
15:25:43 <dhellmann> #info Tuesday 2:50 - 3:00
15:26:10 <smcginnis> We will need to meet in the lobby I think, and get escorted over.
15:27:11 <smcginnis> We do have a planning etherpad for the PTG:
15:27:13 <smcginnis> #link https://etherpad.openstack.org/p/relmgt-ptg-rocky
15:27:19 <smcginnis> Nothing really on there yet.
15:27:57 <smcginnis> So please add anyting you want to make sure we cover while there.
15:28:22 <smcginnis> dhellmann: Was the plan to meet on Thursday for our team work session?
15:28:32 <fungi> looks like dst in dublin doesn't start until march 25, so we'll be on utc while we're there... convenient!
15:28:44 <smcginnis> Oh good.
15:28:46 <dhellmann> we need to do the release on wednesday, so we could meet after that. or have the work session thursday if that's more convenient.
15:28:48 <fungi> for once i don't need to tz-convert all my appointments ;)
15:28:53 <smcginnis> :)
15:29:21 <smcginnis> dhellmann: Since it's a small group, I think we can probably play it by ear.
15:29:27 <dhellmann> yeah
15:29:54 <smcginnis> As far as processing the release - just grab a corner somewhere on Wednesday and work through it?
15:30:26 <dhellmann> yeah, just like we did in denver
15:30:36 <smcginnis> ++
15:30:40 <dhellmann> we should have the patch up before then to give people time to +1 it
15:30:54 <dhellmann> and then there are just a few steps to follow, iirc
15:31:08 <smcginnis> So we can do that. Then if nothing urgent we can just continue on with the rest. Or we can break and regroup on Thursday.
15:31:23 <dhellmann> ++
15:31:39 <smcginnis> We can put up the patch late next week or early PTG week if the ripples haven't settled yet.
15:31:49 <dhellmann> we should verify the timing of the press release with someone from the foundation, but we can do that early in the week
15:32:15 <smcginnis> Anne joined us last time, right. Denver is a bit of a blur for me.
15:32:44 <fungi> i can ask anne if she'll be around and up for it
15:33:12 <smcginnis> #action fungi to check with Anne on Foundation involvement for final Queens release
15:33:12 <dhellmann> it would be good to have her feedback on the highlights thing at least, though it may be early to tell
15:33:14 <fungi> doing that now (she may not be awake yet though)
15:33:34 <smcginnis> Yeah, bit early for the west coast.
15:33:48 <smcginnis> I think there may be some cycle highlights she might want to update.
15:34:03 <smcginnis> A couple teams went a little overboard from what I think our original vision for this was.
15:34:08 <smcginnis> Which is a good problem to have.
15:34:17 <smcginnis> But might be good to get her feedback on what is all out there now.
15:35:34 <dhellmann> ++
15:35:56 <smcginnis> I guess that's all I have. Anything else we should cover in meeting?
15:36:31 <dhellmann> nothing from me
15:36:58 <smcginnis> OK, let's wrap it up then. Thanks everyone.
15:37:11 <smcginnis> #endmeeting