15:00:41 <dhellmann> #startmeeting releaseteam
15:00:42 <openstack> Meeting started Fri Feb 17 15:00:41 2017 UTC and is due to finish in 60 minutes.  The chair is dhellmann. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:43 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:45 <ttx> o/
15:00:45 <openstack> The meeting name has been set to 'releaseteam'
15:01:10 <dhellmann> courtesy ping: ttx, dims, fungi, stevemar, sigmavirus
15:01:23 <fungi> last fillup before ocata!
15:01:30 * dhellmann puts on ominous music
15:01:32 <dims> yay
15:01:35 <dims> :)
15:01:36 <dhellmann> this is week R-1
15:01:38 <sigmavirus> o/
15:01:46 <dhellmann> our agenda is in the etherpad as usual
15:01:48 <dhellmann> #link https://etherpad.openstack.org/p/ocata-relmgt-tracking
15:02:16 <dhellmann> #topic last minute final release candidates
15:02:40 <dhellmann> there's an updated report, hot off of the presses
15:02:41 <dhellmann> #link http://paste.openstack.org/show/599421
15:03:12 <dhellmann> we're expecting an update from the senlin team to deal with their long list
15:03:16 <dhellmann> most of the other projects look short
15:03:24 <ttx> yeah, except senlin
15:03:28 <ttx> but that is in progress
15:03:31 <dhellmann> right
15:03:43 <dims> we ignore the ones with UPPER_CONSTRAINTS_FILE and .gitreview?
15:03:50 <dhellmann> so I think it should be safe to say we're done except for senlin
15:04:01 <dhellmann> dims : yes, those don't need to be in a release, really
15:04:07 <dims> right
15:04:09 <ttx> yes, I'm fine granting senlin a delay since they are the only ones
15:04:21 <ttx> not as if everything was oi fire
15:04:23 <ttx> on
15:04:49 <dhellmann> ok. I expect to be leaving the hotel shortly after our meeting to drive home, which will take several hours. if one of you catches the senlin update, go ahead and process it. otherwise I'll do it when I reach home.
15:04:58 <ttx> ack
15:05:08 <dims> freezer seems to have some stuff
15:05:13 <dims> ack dhellmann
15:05:17 <dhellmann> #action dhellmann send email declaring release candidates frozen except for very critical issues
15:05:52 <dhellmann> dims : at this point unless they complain I think we'll just handle those in a point release
15:06:08 <dims> sounds good
15:06:52 <dhellmann> #topic Missing stable branches
15:07:02 <dhellmann> the only branch we're missing is python-magnumclient, right?
15:07:34 <dhellmann> hmm, magnum too?
15:07:40 <ttx> right. I was just wondering if cuttinh the stable branch post-release won't actually give us even more trouble than cutting it now
15:07:45 <dhellmann> $ .tox/venv/bin/list-deliverables --no-stable-branch
15:07:45 <dhellmann> magnum
15:07:45 <dhellmann> python-magnumclient
15:07:45 <dhellmann> tripleo-quickstart
15:07:45 <dhellmann> tripleo-validations
15:07:56 <ttx> I thought the magnum branch was cut, checking
15:08:03 <dhellmann> maybe I have an out of date repo
15:08:37 <ttx> no, they missed the stable branch request
15:08:41 <dhellmann> no, I don't see it in the magnum repo
15:08:45 <dhellmann> yeah
15:08:46 <ttx> https://review.openstack.org/#/c/435026/3
15:09:04 <dhellmann> oh, are they the ones with the cve?
15:09:09 <dhellmann> I think that's it
15:09:10 <ttx> I'm fine forcing it
15:09:14 <ttx> ah.
15:09:21 <dhellmann> I told them it would be safe to hold off
15:09:24 <ttx> ok
15:09:31 <dhellmann> if they are not already experiencing gate issues, there shouldn't be any harm in waiting
15:09:47 <dhellmann> packagers may be confused, but we should get a branch set up as soon as the cve fix lands
15:09:58 <dims> strigazi : around?
15:10:20 <ttx> so, back to magnumclient... I wonder if cutting Friday is not worse than cutting now
15:10:40 <ttx> but I may miss part of the picture
15:10:43 <dhellmann> ttx: how so?
15:11:00 <ttx> we'll be missing a branch, and if we cut one anyway...
15:11:08 <ttx> what's the point in delaying ?
15:11:14 <dhellmann> heat, mistral, and rally use python-magnumclient
15:11:22 <dhellmann> we've declared a library release freeze until after the final release
15:11:36 <dhellmann> in order to not introduce instability
15:11:53 <dims> create branch from last release, right?
15:12:03 <ttx> ok, that would trigger a minimums bump, I see
15:12:09 <ttx> ok, ignore me
15:12:14 <dhellmann> well, they have a new release requested so I would expect to branch from that
15:12:41 <dhellmann> it wouldn't change the mins, but it would update constraints
15:12:42 <ttx> alternatively we could say that what we have on release day is what will be in stable
15:13:07 <dhellmann> so their new release would be a pike release?
15:13:25 <ttx> do we want to set a precedent that cutting stable branches 5 weeks after the deadline for libraries is ok ?
15:13:28 <dhellmann> that's 4 weeks late, so I could see doing that. We should talk to the magnum team about it, I guess.
15:13:41 <dhellmann> not especially
15:14:20 <ttx> I mean, I'm fine with making an exception and all, but we should make sure they understand that we have two options here
15:14:45 <ttx> and that they put themselves in a painful situation
15:15:08 <dhellmann> option 1 is to approve the new release and branch from that before the final
15:15:10 <ttx> Normally we shoudl have cut the stable branch from what we had
15:15:15 <dhellmann> option 2 is to do option 1 after the final
15:15:16 <dims> ttx : dhellmann : they did not know release process at all until a week or so ago.
15:15:31 <dhellmann> option 3 is to branch from the most current release before the final
15:15:33 <dims> option 3? cut branch from last release?
15:15:34 <dhellmann> right?
15:15:43 <dims> yep i vote for #3
15:15:48 <dhellmann> dims : I no longer accept that excuse.
15:16:19 <dims> dhellmann : i don't either, so i'd vote for #3
15:16:21 <ttx> dhellmann: I think we should have done #3 at library freeze time
15:16:35 <ttx> since we didn't, I'm open to #1 and #2
15:16:39 <dims> they will have to wait till past release and manage cherry picks
15:16:41 * dhellmann is pleased to find others being even more hardline than himself
15:17:06 <dhellmann> ok, I will send an email outlining those options and the reasons and indicating that we prefer option 3
15:17:16 <ttx> I'm not necessarily for a hard line here -- I just want to make sure they understand it's an exceptional situation
15:17:19 <dhellmann> #action dhellmann send email to address magnum client release
15:17:54 <dhellmann> yeah, I was trying to be a little flexible this time around because of the shortened cycle, but 4 weeks is pretty late no matter what the schedule
15:18:04 <ttx> I'm fine with #1 or #2 as long as the magnum team understands that we should normally do #3
15:18:21 <ttx> i.e. we are cutting them some slack
15:18:41 <dhellmann> I think I'd rank them 2, 3, 1
15:19:18 <ttx> basically I don't want them to think it's OK, or them to expect we'll do #2 or #1 next cycle again
15:19:20 <dhellmann> maybe 3, 2, 1
15:19:24 * fungi wonders if the ptl didn't get the e-mail telling them to follow the weekly release schedule annoncements, or didn't heed the advice, or didn't understand
15:19:33 <ttx> but I'm OK to make an exceptions this time
15:19:43 <ttx> since they are the only ones
15:19:51 <dims> fungi : i requested PTL to add tags to his filter
15:20:04 <dhellmann> fungi : yeah. They've had a long time to ask, though, and the schedule is not a secret.
15:20:15 <dhellmann> we've had several new ptls and liaisons this cycle
15:20:17 <fungi> honestly surprised there's just one... doing pretty well with the new process i guess!
15:20:22 <ttx> dhellmann: do you agree that normally we should have done #3 a few weeks ago ?
15:20:52 <dhellmann> ttx: yes. like I said, i was trying to be a bit flexible because of the compressed schedule, to make it easier for patch releases if there was a  need.
15:21:08 <dhellmann> but I wouldn't do that for a longer cycle, like pike will be
15:21:28 <dhellmann> I believe we've already updated the process doc to indicate that we expect branches from all projects around the third milestone
15:21:46 <ttx> OK, I propose we do (2) after explaining that normally we should have just branched from what we had at freeze time
15:21:50 <dhellmann> yes, b922b6dfebf0771f44d2a1a880fc377cdb1f48a6
15:22:04 <dims> i can go with that ttx dhellmann (#2)
15:22:14 <ttx> i.e. "won't work for Pike"
15:22:14 <dhellmann> ok, option 2 is going to be potentially disruptive to those 3 other teams
15:22:33 <ttx> more than #1 ?
15:22:48 <ttx> (that was my original question)
15:22:52 <dhellmann> waiting until after the final postpones the disruption until after the ptg
15:23:10 <dhellmann> basically, we're in the period where teams are supposed to not have to deal with firefights, and this may introduce one
15:23:22 <dhellmann> I guess we could always revert the constraint update
15:23:43 <dims> dhellmann : i can do a test with the affected projects before i let the u-c in
15:23:51 <dhellmann> dims : ack, ok
15:24:07 <dhellmann> well, if we're going to do 2, do we still need to send an email?
15:24:31 <dims> dhellmann : y we should publicize the plan
15:24:33 <ttx> I think we should make a point to explain it's an exception and not the rule
15:24:41 <dhellmann> ok
15:24:52 <dhellmann> #agreed we will make an exception for python-magnumclient and release and branch today
15:25:08 <dhellmann> we need to introduce the branch instructions into that patch, or a follow-up
15:25:43 <strigazi> dims yes
15:25:48 <ttx> hmm #2 is after final
15:26:00 <ttx> "2 is to do option 1 after the final"
15:26:03 <dhellmann> oops
15:26:04 <dhellmann> #undo
15:26:04 <openstack> Removing item from minutes: #agreed we will make an exception for python-magnumclient and release and branch today
15:26:14 <ttx> so we may be not talking about the same thing
15:26:20 <dhellmann> yes, let's stop using option numbers
15:26:27 <dims> :)
15:26:41 <dhellmann> do we want to tag and branch today or next friday
15:26:47 <ttx> dhellmann: with "branch now" would you say that affected projects should roll a new RC to catch the constrinat update ?
15:26:59 <dhellmann> ttx: constraint updates don't go into the releases, so they don't need to do that
15:27:20 <dhellmann> we've been asking the requirements team to pre-approve constraint updates
15:27:26 <ttx> same with "branch after release", won't trigger a bunch of point releases ?
15:27:30 <dhellmann> tonyb, prometheanfire: ping?
15:27:52 <dhellmann> ttx: no, we don't update the minimum version so there is nothing to sync into the projects
15:28:04 <dhellmann> the constraints file is used from the global location
15:28:11 <ttx> hmm, sounds safe to me then
15:28:22 <dhellmann> so is that a vote for today?
15:28:43 <ttx> I'm fine with today or next week. Small preference for today and a clean slate
15:28:50 <dhellmann> dims: ?
15:28:52 <dims> so make the new release with a branch, then test u-c change with affected projects
15:29:07 <dims> +1
15:29:17 <dhellmann> dims : ok. will you have time to do that testing?
15:29:20 <dims> yep
15:29:32 <dhellmann> cool, do you want to take the action to process the release, then?
15:29:49 <dims> dhellmann : sounds good, please give me an action
15:29:56 <dhellmann> #agreed we will make an exception for python-magnumclient and release and branch today
15:30:17 <dims> strigazi : can you edit the magnumclient request and add request for creating the stable branch please?
15:30:22 <dhellmann> #action dims test the python-magnumclient u-c change with heat, mistral, rally
15:30:43 <dims> strigazi : https://review.openstack.org/#/c/435241/
15:30:55 <dhellmann> dims: thanks for handling that
15:31:01 <strigazi> dims ok
15:31:07 <dims> thanks strigazi
15:31:32 <dhellmann> moving on then
15:31:39 <dhellmann> #topic Release notes update patch
15:31:50 <dhellmann> as ttx points out in the agenda notes, it would be good to merge this now
15:31:59 <dhellmann> there will be more updates, and those can happen early next week
15:32:17 <ttx> conflicts with:
15:32:23 <ttx> https://review.openstack.org/#/c/435282/2
15:32:31 <ttx> https://review.openstack.org/#/c/435456/1
15:32:40 <dhellmann> I'll prep a patch and we can get it merged before the final release next week
15:32:50 <ttx> both need a refresh anyway so it's fine
15:32:53 <dhellmann> we expect an update to the senlin rc2 patch so that should work itself out
15:33:00 <ttx> so merge now
15:33:03 <dhellmann> and yep, same for the other
15:33:24 <dhellmann> ttx: do you want to do the honors there?
15:33:33 <ttx> sure
15:33:36 <dhellmann> thanks
15:33:41 <dhellmann> #topic Coordinating for final release work next Wednesday
15:33:52 <dhellmann> ttx, fungi, and I will all be in ATL next week
15:34:02 <dims> apologies folks, i won't be there to celebrate in person
15:34:08 <dhellmann> I proposed that we meet at or right after breakfast Wednesday to do the merging work
15:34:09 <ttx> dims: aw
15:34:12 <dhellmann> we'll also be here in channel
15:34:18 <dhellmann> dims: we'll definitely miss you!
15:34:19 <dims> dhellmann : cool!
15:34:40 <dhellmann> loquacities has asked me to ping the docs folks so they can update the landing page when we're ready
15:35:05 <dhellmann> so when we have a firm time for when we're meeting, I will let her know so they can be present, too
15:35:31 <dhellmann> I figure we can set that time on monday or tuesday, when we have seen what the morning rush at the restaurants in the hotel looks like
15:35:36 <dhellmann> thoughts?
15:35:56 <fungi> i'm happy to join you all on wednesday since infra stuff will be officially over by then
15:36:10 <ttx> ack
15:36:17 <dhellmann> fungi : excellent, thanks
15:36:32 <ttx> we might be able to secure a release war room
15:36:40 <dhellmann> next up then
15:36:41 <dhellmann> #topic Pike release schedule final approval
15:36:43 <dhellmann> #undo
15:36:44 <dhellmann> yeah, I wasn't sure how early we would need to start
15:36:44 <openstack> Removing item from minutes: #topic Pike release schedule final approval
15:36:47 <dhellmann> it took a bit longer than expected last time
15:37:05 <dhellmann> if we need to be done by 11:00, I would rather be done early than rushing
15:37:06 <ttx> we can discuss that on-site
15:37:08 <dhellmann> yep
15:37:34 <dhellmann> I intend to have all of the patches we need lined up by monday morning
15:37:45 <dhellmann> #topic Pike release schedule final approval
15:37:54 <dhellmann> I think we're ready to approve the details we have now
15:38:05 <ttx> yes, will add PTg data once the date is 100%
15:38:08 <ttx> date*
15:38:09 <dhellmann> I do agree with Brian that as soon as we have the next ptg dates we'll ad those
15:38:09 <dhellmann> yep
15:38:20 <ttx> don't want people to start organizing
15:38:33 <ttx> probably next week we'll have the date set in stone
15:38:41 <dhellmann> I'll go ahead and approve that patch now
15:38:45 <ttx> ack
15:39:18 <ttx> we might want to set Pike "Under development" too
15:39:34 <dhellmann> is there already a patch for that?
15:39:45 <ttx> I'll prepare that (moving Ocata to "pre-release freeze" as well)
15:39:50 <fungi> would be nice if maybe we had it settled in time for the ptg feedback session on thursday?
15:39:50 <dhellmann> I think preparing that is part of the tasks I was supposed to do today
15:39:54 <dhellmann> ttx: thanks
15:40:11 <dhellmann> fungi : yes, good point
15:40:53 <ttx> fungi: ++
15:41:05 <dhellmann> right then, let's talk about next week
15:41:07 <dhellmann> #topic PTG planning
15:41:22 <dhellmann> ttx: this one is yours
15:41:42 <ttx> there is an attendee list for the room, if you already know you'll be somewhere else part of the time, good to indicate that there
15:42:24 <dhellmann> yeah, it's going to be a crazy 2 days for me and then fairly quiet after wednesday morning
15:42:27 <ttx> We mentioned we could do a release retrospective
15:42:43 <ttx> i think we should schedule it in the large fishbowl room
15:42:51 <ttx> and Thursday or Friday
15:42:53 <dhellmann> you asked about scheduling for that, it might be better to do it some time thursday or friday morning
15:43:11 <dhellmann> I know some teams/people are leaving friday so thursday may give us more liaisons
15:43:12 <ttx> Let's do Friday morning
15:43:18 <dhellmann> either way
15:43:20 <ttx> ah, good point
15:43:33 * ttx looks up ethercalc
15:43:45 <dhellmann> https://ethercalc.openstack.org/Pike-PTG-Discussion-Rooms
15:44:07 <ttx> 2:30pm Thu ?
15:44:31 <fungi> wfm
15:44:44 <dhellmann> we might want to avoid that horizon item, in case they have input
15:44:59 <dhellmann> the keystone team is on top of things, but we had npm-related issues with some stuff this time around
15:45:00 <ttx> dhellmann: that's wednesday though
15:45:08 * dhellmann can't read
15:45:09 <dhellmann> ok
15:45:22 <ttx> ok deal
15:45:36 <dhellmann> #agreed retrospective thursday 2:30 in south captial room
15:45:47 <ttx> Quick reminder that requirements/stable/Relmgt will have a team dinner on Tuesday night
15:46:06 <ttx> hopefully not poisoning the whole team at the critical time
15:46:12 <fungi> hah
15:46:19 <ttx> RSVP @ https://framadate.org/XIbbQnbxSKRPW1yK
15:46:22 <fungi> that could lead to a terrible release
15:46:24 <dhellmann> no seafood :-)
15:46:32 <ttx> I'll probably book over the weekend while there are still options
15:46:32 <dims> :)
15:46:47 <ttx> I was thinking Alma Cocina
15:47:00 <ttx> which is close by
15:47:17 <dhellmann> wfm. I've never been there.
15:47:32 <ttx> Alright that is all from me
15:47:42 <ttx> oh, unless you hae other suggestions
15:47:52 <ttx> dhellmann: you know the city way better than I do :)
15:48:08 <dhellmann> I live far enough away that I don't actually know it that well any more
15:48:13 <ttx> feels like if we want to avoid shuttling people around town there aren't that many options
15:48:25 <dhellmann> yeah
15:48:57 <dhellmann> we're not going to want to deal with going to far by car anyway, due to weekday traffic
15:49:10 <dhellmann> although by dinner it may be better
15:49:28 <dhellmann> it looks like that's the last formal topic on the agenda
15:49:31 <dhellmann> #topic open discussion
15:49:38 <dhellmann> does anyone have anything else to bring up?
15:49:53 <ttx> nope
15:50:06 <ttx> https://review.openstack.org/435500 sets Pike to "under development"
15:50:18 * dhellmann opens in the background
15:50:32 <ttx> dhellmann: when do you get in ? Late on Sunday I assume ?
15:50:41 <dhellmann> if that's it, I think we can close early
15:50:51 <dhellmann> ttx: I'm driving in, so I can come in any time after lunch
15:51:00 <dims> safe travels everyone!
15:51:03 <fungi> about the only venues i'm familiar with in atlanta are in little five points, which is a bit of a haul (that area has a pretty decent nightclub/bar scene though)
15:51:21 <dhellmann> fungi : ++
15:51:25 <fungi> see everyone next week!
15:51:35 <dhellmann> see you next week!
15:51:42 <dhellmann> #endmeeting