16:00:24 <smcginnis> #startmeeting releaseteam
16:00:25 <openstack> Meeting started Thu Oct 17 16:00:24 2019 UTC and is due to finish in 60 minutes.  The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:26 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:27 <smcginnis> Ping - ttx dhellmann diablo_rojo hberaud evrardjp armstrong tonyb
16:00:28 <openstack> The meeting name has been set to 'releaseteam'
16:00:30 <ttx> o/
16:00:32 <hberaud> o/
16:00:46 <fungi> all aboard!
16:00:54 <smcginnis> Probably a small group of us this week, but we can wait a bit in case anyone else joins.
16:01:25 <smcginnis> #link https://etherpad.openstack.org/p/train-relmgt-tracking Agenda
16:01:50 <smcginnis> OK...
16:01:59 <smcginnis> #topic Postmortem analysis
16:02:21 <ttx> anythign that we should fix before next round?
16:02:25 <smcginnis> 🚂🚂🚂🚂🚂🚂🚂🚂🚂🚂
16:02:27 <fungi> yesterday went really smoothly, for what that's worth ;)
16:02:42 <smcginnis> Yeah, I think things went pretty well overall.
16:02:57 <smcginnis> Nothing major I can think of that we need to fix.
16:04:24 <ttx> ok moving on then
16:04:52 <smcginnis> If anyone ever has any suggestions, feel free to make suggestions at any time.
16:05:14 <smcginnis> #topic Queens extended maintenance transition
16:05:36 <smcginnis> According to our schedule, it looks like next Friday we are due to transition Queens to EM.
16:05:59 <smcginnis> Elod was great enough to send out a post to openstack-discuss to alert everyone of this.
16:06:09 <smcginnis> Last call for getting any official releases out there.
16:06:58 <smcginnis> We've gotten one stable/queens release since then. We'll see if any others trickle through.
16:07:04 <openstackgerrit> Eric Fried proposed openstack/releases master: Release os-traits 1.1.0  https://review.opendev.org/689206
16:07:20 <smcginnis> The output from list_unreleased shows mostly non-functionality commits for most repos.
16:08:17 <smcginnis> Not sure if there's much else to do or say related to that for now. I'm not planning on proposing any releases for teams with this.
16:08:22 <smcginnis> Though I guess... should we?
16:08:31 <smcginnis> We've kind of set that precendence now.
16:09:15 <ttx> hmm we did, but for the current cycle
16:09:23 <ttx> For stable, I don;t think it is our role
16:09:31 <smcginnis> That works for me.
16:09:56 <smcginnis> I'm honestly a little worried about the number of things we've put on this group vs. the individual teams to get things done.
16:10:28 <ttx> I think it's workable with good tooling
16:10:36 <ttx> But yes I would draw the line at release
16:11:01 <smcginnis> #topic Ussuri tracking pad and ping list update
16:11:16 <smcginnis> #link https://etherpad.openstack.org/p/ussuri-relmgt-tracking
16:11:42 <smcginnis> Today's wraps up this etherpads usefulness and we now move on to the fresh new sparkly ussuri one.
16:11:52 <ttx> yes jump
16:11:58 <ttx> I copied the last lines
16:12:21 <smcginnis> One thing I wanted to point out there is I've cleared out the ping list. So if anyone likes getting those pings at meeting start, please readd your nick to the list on line 12.
16:13:18 <smcginnis> We should probably also review the Review Days.
16:13:30 <fungi> i've added a reminder there that i'm clear to push up the change to swap the release keys now
16:13:32 <smcginnis> Though that might be too depressing. :]
16:13:38 <smcginnis> fungi: Thanks!
16:14:05 <smcginnis> I think for now we can keep the review days as is, but just speak up if you want to make any changes there.
16:14:42 <ttx> smcginnis: did you want to send a personal email to liaisons for Ussuri?
16:14:54 <smcginnis> Just watching the etherpad. I agree ttx, we can likely skip a lot of those emails.
16:14:57 <ttx> The "please pay attention to [release] on openstack-discuss" email
16:15:04 <smcginnis> I should.
16:15:33 <smcginnis> It was a bit of a pain last time I did it with my email providers throttling me since it looked like I was sending out mass spam.
16:15:41 <ttx> Should probably be sent early next week
16:15:54 <ttx> Or we can make it the first countdown email
16:15:58 <ttx> your call
16:16:07 <smcginnis> But given how the last couple cycles came to a close with a couple teams seemingly clueless about where we were in the cycle and what the expectations were, I probably should try to avoid that.
16:16:26 <ttx> I prefer to send all countdown emails to the list
16:16:39 <ttx> but don;t want to do email overload either
16:17:04 <smcginnis> Yes, countdowns I think still should go to the list. Just we had the item to reach out directly to PTLs and liaisons at the start of the cycle, and I stopped doing that.
16:17:20 <smcginnis> Not sure if Tony did it for Train, but I'm guessing probably not.
16:17:37 <smcginnis> OK, on to the task list...
16:18:17 <ttx> Who takes the fixing missing things task?
16:18:20 <smcginnis> So, the "fixing missing..." stuff.
16:18:27 <smcginnis> What do we want to do there?
16:18:38 <smcginnis> Or maybe, who wants to take the action to figure out what to do there?
16:18:39 <ttx> Investigate why they show up, if they should show up etc
16:18:44 <ttx> I can take it
16:18:51 <smcginnis> Thanks!
16:19:18 <smcginnis> fungi: And the assumption is it is something you would do, but I'll put your nick on the signing key line item.
16:19:45 <fungi> thanks, yes it is
16:19:51 <smcginnis> I guess I can take the queens EM transition one. Unless anyone else wants it.
16:20:27 <smcginnis> I added some scripting to that last time, but I think I still need to address breaking it out into smaller chunks by team or something so it doesn't time out.
16:20:57 <smcginnis> And ttx has signed up for the email content again. And again, thank you for that.
16:21:25 <smcginnis> I think that's it.
16:21:34 <smcginnis> We've covered everything else in https://releases.openstack.org/reference/process.html#before-ptg-after-closing-previous-release
16:22:16 <smcginnis> #topic Next week's meeting
16:22:21 <ttx> Alright. Not sure we need a meeting next week
16:22:33 <ttx> I think we can vet email content asynchronously
16:22:33 <smcginnis> I agree.
16:22:56 <smcginnis> Email content and anything that pops up with Queens EM we can just discuss in channel as needed.
16:22:56 <ttx> I don;t expect any task on the week after to review
16:23:18 <smcginnis> Hopefully a quiet couple of weeks.
16:23:22 * smcginnis knocks on wood
16:23:23 <ttx> that is the idea
16:23:29 <smcginnis> Then the Summit
16:23:58 <smcginnis> #topic Open Floor
16:24:07 <smcginnis> Anything else to discuss?
16:24:17 <ttx> nothing from me
16:24:26 <ttx> Oh there is
16:24:29 <ttx> the Ironic thing
16:24:34 <smcginnis> Too late! :P
16:24:37 * ttx checks thread status
16:24:39 <smcginnis> What's the Ironic thing?
16:24:41 <smcginnis> Oh!
16:24:48 <smcginnis> Yeah, we should discuss that.
16:24:57 <ttx> "Ironic Train release can be broken due to entry in driver-requirements.txt"
16:25:08 <ttx> I think we need to provide guidance on that thread
16:25:28 <fungi> i was curious about that. the indicated line is a >= isn't it?
16:25:39 <ttx> basically the driver-requirements spec for sushy is a bit too wide
16:25:40 <fungi> or did i misread? i probably misread
16:25:48 <ttx> >=1.9.0
16:26:02 <ttx> while some drivers can be broken if not 2.0.0 is used
16:26:14 <ttx> that was my understanding
16:26:44 <smcginnis> Looks like I've deleted my local copy already.
16:26:47 <smcginnis> #link http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010148.html
16:26:48 <fungi> yeah, so if someone tries to install with an older version the driver requirements list claims to support, they'll be broken. however it's technically not wrong to say it requires versions >=1.9.0 in a literal sense
16:27:13 <ttx> It's more of a stable question, but in that circumstance I would be fine with a bump in driver-req
16:27:26 <fungi> it just happens to not work with some versions >=1.9.0 but the versions it does work with are allowed by that expression
16:27:35 <smcginnis> Seems fair. But I wonder how that impacts distros.
16:27:40 <ttx> fungi: yeah... it's a bit fuzzy territory
16:28:10 <fungi> so it's not preventing a working installation
16:28:20 <fungi> it just happens to provide inaccurate guidance
16:28:27 <smcginnis> U-C caps it at 2.0.0 https://opendev.org/openstack/requirements/src/branch/stable/train/upper-constraints.txt#L431
16:28:56 <smcginnis> Maybe better to have prometheanfire and/or other packagers respond to that?
16:29:15 <ttx> yeah, I just did not want it to be ignored
16:29:25 <ttx> I don't have a dog in that race
16:29:40 <smcginnis> As far as stable requirements goes, we can't really lower the upper-constraints.
16:29:59 <smcginnis> But yeah, the driver-requirements doesn't force things to be broken.
16:30:07 <smcginnis> It "works"
16:30:50 <smcginnis> Kind of along these same lines, I wonder about https://review.opendev.org/#/c/686845/
16:31:18 <smcginnis> I wonder if it would have been better to keep that as a train release.
16:31:34 <smcginnis> But looks like this should work, then they just need to backport and do a bugfix release.
16:32:43 <smcginnis> OK, I don't have anything else on this or other topics.
16:32:45 <ttx> hmm maybe just post an answer asking for opinions from stable and requirements + saying that >=1.9.0 is not technically wrong
16:33:38 <smcginnis> Can someone else respond since I've already deleted my local copy? :]
16:33:46 <ttx> re: ovsdbapp, meh, whatever works for them :)
16:33:50 <smcginnis> Yeah
16:34:40 <ttx> Oh well, let's wait for prometheanfire to comment on it
16:34:56 * ttx needs to run
16:34:58 <smcginnis> Works for me. ;)
16:35:05 <smcginnis> I guess that's all anyway.
16:35:08 <smcginnis> Thanks everyone!
16:35:16 <smcginnis> #endmeeting