15:05:21 #startmeeting releaseteam 15:05:22 Meeting started Fri Mar 3 15:05:21 2017 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:05:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:05:25 The meeting name has been set to 'releaseteam' 15:05:40 o/ 15:05:42 meeting agenda at 15:05:44 #link https://etherpad.openstack.org/p/pike-relmgt-tracking 15:06:16 dhellmann: did you have some elaborate ping list I should copy ? 15:06:55 #topic PTG 15:07:11 So last week we were at PTG 15:07:14 courtesy ping: dhellmann, dims, fungi, sigmavirus 15:07:15 I posted a summary at... 15:07:41 #link http://lists.openstack.org/pipermail/openstack-dev/2017-March/113121.html 15:07:46 courtesy ping: tonyb, stevemar 15:08:12 Is there anything significant I missed ? 15:08:52 Copying decision on "Being more relaxed about independent library releases during the freeze period" to pike plan 15:09:05 * stevemar is lurking a teeny bit 15:09:25 that covered everything big that I remembered 15:09:42 dhellmann: is there anything TODO linked to that decision about indep releases ? 15:10:02 (copied to https://etherpad.openstack.org/p/pike-relmgt-plan as Process Changes /10) 15:10:02 we may end up with more reno features if the python-dev team decides to adopt it, but otherwise I agree it's probably not going to change much aside from bug fixes 15:10:42 ttx: we should write down what we said wherever we describe the freeze, so we remember the change and so those teams know what to expect 15:10:53 ack, will check PROCESS and copy it there 15:10:54 I don't know if that's in our README or in the schedule 15:11:59 ok, anything else related to PTG ? 15:12:31 #topic pbr / reno release postmortem and leftover actions 15:12:50 so the pbr and reno releases did, as supsected, create a bit of a ripple 15:12:54 nothing from me 15:13:02 well, it was good to see you all again :-) 15:13:15 was wondering what there was left to do there to clean up 15:13:21 * ttx opens links 15:13:22 I think all of the reno issues are fixed 15:13:56 wow Rico has been busy 15:13:58 dims will correct me if I'm wrong, but I think the pbr situation is stable with no wide ranging issues and fixes up for any individual projects affected 15:14:11 dhellmann : yep 15:14:25 yeah, this is a good example of where mechanically generated patches are a good thing 15:14:49 we have a few infra projects that still need some changes merged to drop old pbr<2 caps but nothing that's used as deps of other projects at this point 15:15:01 ttx: I think yaql is a case where we need a release because of pbr, but they're not "official" yet so we weren't doing it through the automation 15:15:07 do we expect similar pain next time we upgrade pbr ? Or will the work to uncap it now pay dividends in future ? 15:15:13 dims, do you remember if yaql already released? 15:15:36 fungi : at least one of those was blocking updates in the requirements bot job 15:15:51 I mean the job that runs to update constraints in openstack/requirements 15:15:52 I guess it's generally good to do such releases in "calm" weeks 15:16:03 dhellmann: still, or you mean yesterday? 15:16:06 ttx: uncapping now should avoid the pain in the future 15:16:21 or at least limit it :) 15:16:25 dhellmann : yep, yaql is out 15:16:29 fungi : yesterday for sure. I haven't looked today. Did you do any releases? 15:16:35 * dhellmann looks for the project name 15:16:42 at least one (requestsexceptions) 15:17:09 that may have been wednesday actually 15:17:10 requestsexceptions 15:17:26 ttx: there will be other pains, but maybe not this one 15:17:30 yeah, requestsexceptions got a new release on wednesday with the uncapped pbr dep 15:17:32 dims : cool, thanks 15:17:36 fungi : great, thanks 15:17:47 ok, I think this is outside our control now and up for projects to fix 15:17:54 yes 15:17:56 anything else on that topic ? 15:17:58 u-c has been updated for all the problem projects 15:18:25 nice side effect, we have moved up sphinx too 15:18:36 thanks to dims and tonyb for pushing those fixes through 15:18:40 indeed 15:18:48 that's it from me 15:18:48 #topic Priority reviews 15:19:03 dims: could you rereview https://review.openstack.org/#/c/439669/1 ? 15:19:17 you had an objection that was overruled 15:19:29 I'd like to add that one asap 15:19:36 +2A 15:20:13 ok cool, any other priority review ? 15:20:55 ttx : we need to deal with https://review.openstack.org/#/q/status:open+branch:master+topic:move-tools-to-releases 15:21:01 yes we do 15:21:08 https://review.openstack.org/#/q/status:open+topic:move-tools-to-releases 15:21:13 maybenot on a Friday ? 15:21:25 first thing Monday ? 15:21:27 yeah, maybe early next week 15:21:28 ttx : first thing monday is good 15:21:37 it would be good to do those before the milestone, though, so we have some releases to test with 15:21:52 Let's merge them once we are all in on Monday 15:21:55 * dhellmann is experiencing lag 15:22:43 #topic Review initial PTL email contents 15:22:49 I posted a draft at 15:22:57 #link https://etherpad.openstack.org/p/As13M6IsSz 15:23:06 I did a number of smallish changes 15:23:19 like extra focus on liaisons, encouraging PTLs to nominate one, doesn't have to be core 15:23:25 ++ 15:23:29 Also tweaked IRC commitments to daily presence during milestone weeks 15:23:36 since we were so far from that last cycle 15:23:42 sounds like areasonable bar to set 15:23:57 yep 15:24:11 dhellmann: your guidelines included continuous presence the week before, though 15:24:25 so I wanted to check if "milestoen week" sounds enough to you 15:24:35 I think that's the bare minimum 15:24:38 yeah, that may have been excessive 15:24:51 by setting the bat too high, some people just don't jump at all 15:24:54 bar* 15:25:23 Tried (unsuccessfully) to trim it down to less words 15:25:31 IIRC, the milestones worked out ok, it was just the RCs where we were having trouble for Ocata 15:25:38 anyway, will send shortly after meeting 15:25:50 so if you have comments let me know 15:25:58 so the week of the deadline should be good enough coverage 15:26:10 #topic Review tracking tasks / reminder email contents up to Pike-1 15:26:24 although it would be nice if the liaison was there all the time 15:26:25 esp. for non-milestone projects 15:26:31 I tried to write up tasks / meeting skips / countdown email contents up to P1+1 15:26:51 i.e. R-19 15:27:13 ttx: it looks like you copied ocata notes in without changing ocata->pike in a couple of places 15:27:17 If you see anything missing, just add it 15:27:25 yes 15:28:02 I'll start sending weekly emails next week 15:28:10 once the pTLs received the reminder 15:28:25 I was thinking sending them Friday EU morning 15:28:31 sounds good. I sent the last one for ocata yesterday 15:28:33 ttx : dhellmann : we should make it a priority to attract fresh blood to the team. starting with a call for liaisons and updating the list of liaisons 15:28:43 since dhellmann did send them late Thursday US 15:29:21 dims: the email to PTLs encourages that already, and will be cced to list 15:29:22 and somehow getting an ack from ptl that they are watching the tags we use 15:29:41 dims: yes, an ack is required from that email as well 15:29:48 ++ttx 15:30:03 OK, that is all I had 15:30:21 #topic Open Discussion 15:30:23 ttx: that timing sounds good, it's sort of why I tried to do them thursday afternoon 15:30:34 never release pbr? or hacking? :) 15:30:48 * dhellmann wonders if the lag is freenode or his bouncer 15:31:45 i'm curious, with respect to the stewardship working group discussion on release workflow steps and positive acknowledgement from expected participants, has anybody played around yet with trying to use the current state of storyboard to that end? maybe as a board with lanes for each release stage or something? 15:31:56 I have a couple of PTO periods planned over the next couple of months. I've added them to the tracking pad. 15:31:59 dhellmann: mosh and weechat indicate relative lags to me 15:32:24 fungi : we've discussed it, but haven't actually tried it 15:32:34 my weechat isn't claiming a lag, but it could be a problem with one or more specific freenode servers and just not affecting the one through which i'm connecting 15:33:37 Alright, if nothing else, let's close early. Still a lot to do before calling it a week 15:33:45 ++ 15:33:50 sure, wfm 15:34:01 sounds good 15:34:13 #endmeeting