19:00:00 <tonyb> #startmeeting releaseteam
19:00:01 <openstack> Meeting started Thu May 16 19:00:00 2019 UTC and is due to finish in 60 minutes.  The chair is tonyb. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:00:05 <openstack> The meeting name has been set to 'releaseteam'
19:00:14 <smcginnis> o/
19:00:17 <tonyb> ttx dhellmann diablo_rojo hberaud evrardjp fungi armstrong: Release team meeting is starting
19:00:22 <ttx> o/
19:00:34 <openstackgerrit> Sean McGinnis proposed openstack/releases master: [monasca] final releases for pike  https://review.opendev.org/652854
19:00:42 * tonyb will point out that the start time was "Thu May 16 19:00:00 2019 UTC" #nailedit!
19:00:45 <smcginnis> Tests took 30 seconds too long.
19:00:53 <smcginnis> Haha, nice!
19:01:15 <tonyb> smcginnis: gotta take the wins ;P
19:01:20 <evrardjp> o/ (but commuting back home so that I can have my laptop)
19:01:25 <diablo_rojo_phon> o/
19:01:48 <dhellmann> o/
19:02:11 <tonyb> ttx: has (as always) set us up with a good agenda
19:03:34 <tonyb> okay let's start
19:03:52 <tonyb> #topic Confirm task assignments from https://etherpad.openstack.org/p/relmgmt-train-ptg
19:04:10 <ttx> So there were a fez open question marks on last week assignment exercise
19:04:22 <ttx> few*
19:04:32 <tonyb> Yup we have evrardjp here today so ...
19:04:43 * smcginnis wants an official release team fez
19:04:44 <ttx> On the process changes we had to complete by train-1...
19:04:47 <diablo_rojo_phon> So we can assign him a bunch of stuff.
19:04:56 <ttx> Do stable autoreleases around milestones: evrardjp?
19:05:02 <ttx> Track release liaisons in deliverable files: diablo_rojo?
19:05:04 * diablo_rojo_phon agrees with smcginnis's fez idea
19:05:12 <diablo_rojo_phon> Yeah I can take that one.
19:05:26 <tonyb> I was thinking he can let us know his capacaity but sure we can just give it all to him ;P
19:06:03 <evrardjp> haha
19:06:18 <ttx> evrardjp: are you ok with driving that change?
19:06:21 <evrardjp> I am not yet sure if I can take this, tbh.
19:06:31 <evrardjp> s/yet//
19:06:42 <ttx> do we have anyone else who could take it? needs to be completed soon enough
19:06:56 <evrardjp> that's the problem for me
19:06:59 <ttx> I can't, my travel schedule in crazy in june
19:07:25 <tonyb> I can do the scripting side of things but the hard part is looking at each deliverable and making the call on if it needs to be released
19:07:34 <smcginnis> I'm hesitant to sign up for anything because I'm not sure how much time I am going to have to devote to things soon, but I'm willing to help someone else.
19:07:36 <tonyb> ttx: I did notcie that ;P
19:07:40 <ttx> t-1 is in 3 weeks
19:07:51 <ttx> I mean it's also crazy in May
19:08:10 <smcginnis> tonyb: I had looked at making the scripting interactive where it could print out the changes, then you y/n whether anything requires a release.
19:08:10 <ttx> come to think of it, July is not really better :|
19:08:23 <diablo_rojo_phon> ....I could maybe take it. With help.
19:08:28 <diablo_rojo_phon> My June doesn't suck.
19:08:33 <diablo_rojo_phon> Just the rest of this month.
19:08:41 <evrardjp> lucky you
19:08:52 <tonyb> I think you're traveling too much :/
19:08:54 <ttx> hmm, do we NEED to have it done by train-1?
19:09:03 <tonyb> ttx: I was just going to ask that
19:09:15 <ttx> I said t1 because we kinda wanted to do that milestone in autorelease
19:09:32 <fungi> not really on-topic, but tonyb mentioned a great meeting agenda just didn't link to it (and the irc-meetings data doesn't include one either)... can haz link?
19:09:34 <tonyb> it was a preety soft "lets do it at this time" rather than a hard "it must be done"
19:09:35 <ttx> but I guess we could also not do anything for t1
19:09:54 <evrardjp> We can work on it this cycle and put that 'in prod' for next ? ;)
19:10:00 <tonyb> fungi: sorry
19:10:04 <tonyb> #link https://etherpad.openstack.org/p/train-relmgt-tracking
19:10:08 <fungi> thanks!
19:10:08 <smcginnis> s/train-1/as soon as possible/
19:10:10 <ttx> evrardjp: we ship to prod here
19:10:12 <tonyb> Line 134'ish
19:10:32 <evrardjp> :)
19:10:40 <ttx> evrardjp: could you take it if it was a t2 thing?
19:10:59 <ttx> (July 24)
19:11:12 <evrardjp> probably the same as t1
19:11:29 <evrardjp> so I would say no.
19:11:47 <evrardjp> but after it looks better
19:11:58 <evrardjp> "theoretically "
19:12:06 <ttx> smcginnis: could you mentor diablo_rojo through it ?
19:12:10 <evrardjp> sorry :/
19:12:55 <tonyb> evrardjp: It's all good.
19:13:16 <tonyb> evrardjp: You can make it up to us later ;P
19:13:34 <evrardjp> :)
19:13:35 <smcginnis> ttx: Sure, I can help.
19:13:57 <tonyb> smcginnis: thanks
19:14:03 <ttx> OK so I'll move that task on the PTG doc to t2 and make it diablo_rojo mentored by smcginnis
19:14:40 <evrardjp> thx ttx
19:14:42 <diablo_rojo_phon> If we did it for T1 what's the actual date on that?
19:14:52 <tonyb> that'll mean we can use the liasion stuff diablo_rojo_phon is doing/will do
19:14:58 <ttx> before June 6
19:15:37 <ttx> I mean you can always finish before
19:15:38 <diablo_rojo_phon> smcginnis if you're free the fee days leading up to that I think we can have it done then.
19:15:46 <diablo_rojo_phon> *few
19:16:12 <smcginnis> We can also start with a crawl, walk, run approach and start semi-manual/semi-automated and work towards complete(r) automation later.
19:16:14 * tonyb likes smcginnis idea to add '--interactive' to new-release which will help with this
19:16:18 <ttx> We also had a must-do improvement
19:16:26 <ttx> Add a tool script for producing the list of intermediary deliverables that have not released
19:16:38 <ttx> I'll do that one if nobody else wants it
19:16:42 <evrardjp> tonyb : agreed
19:17:00 <smcginnis> That one is basically a wrapper around the existing commands to make the task easier, right?
19:17:01 <tonyb> ttx: I think it's yours ;P
19:17:08 <ttx> smcginnis: I guess
19:17:14 <smcginnis> Going once, going twice, sold to ttx
19:17:19 <ttx> haven't looked :)
19:17:46 <evrardjp> smcginnis: I like your style
19:17:52 <smcginnis> ;)
19:18:07 <ttx> ok, I have nothing more on this topic
19:18:12 <dhellmann> tonyb , smcginnis : regarding the interactive mode, harlowja built a tool like that in the repo already (don't remember the name, but look at the list of cli apps in setup.cfg)
19:18:34 <dhellmann> I think it's out of date, but it should be possible to update it
19:18:40 <dhellmann> or at least crib from it to update new-release
19:18:42 <tonyb> dhellmann: Yup I do recall that
19:18:43 <smcginnis> That would be great to have a starting point.
19:19:10 <tonyb> interactive-release ?
19:19:19 <dhellmann> probably
19:19:26 <smcginnis> Huh, would you look at that. :)
19:19:53 <smcginnis> Might be worth doing an inventory of all our commands and tools and get that all in once place with short descriptions.
19:20:02 <smcginnis> s/once/one/
19:20:06 <dhellmann> I thought we had that in the readme, too, but maybe not
19:20:14 <tonyb> It's in the readme
19:20:15 <dhellmann> that's probably out of date, too
19:20:25 <smcginnis> I think we do for the tools we want others to use. Maybe not.
19:20:32 <dhellmann> man, the old release team was really a bunch of slackers
19:20:36 <smcginnis> Yeah, likely could use a refresh at any rate.
19:20:41 <smcginnis> Hah!
19:20:46 <evrardjp> dhellmann: haha
19:20:48 <ttx> especially the last 3 PTLs
19:20:50 <tonyb> Okay I'll do that
19:21:30 <tonyb> at least those PTLs could send email on time .. this new PTL ... what a schmuck!
19:21:38 <smcginnis> Hah!!
19:21:40 <ttx> now onto pike-em fun
19:21:49 <dhellmann> email on time or meeting on time, pick one
19:22:01 <tonyb> hehe
19:22:08 <evrardjp> :)
19:22:09 <smcginnis> tonyb: To be fair, I think it took me a few weeks to get into the habit.
19:22:33 <tonyb> smcginnis: I'll take it
19:22:43 <tonyb> #topic pike-em
19:23:00 <smcginnis> Monasca should be ready to go now. Then we just need to update the pike-em tagging patch.
19:23:14 * ttx fetches a drink
19:23:14 <diablo_rojo_phon> I emailed the four PTLs/liaisons.
19:23:22 <smcginnis> Awesome!
19:23:37 <smcginnis> Yeah, a few of them have shown more attention the last few days, so that has helped.
19:24:01 <tonyb> Looks like tripleo is released we just need the pike-em tag
19:24:04 <smcginnis> Tricircle is stuck pending an update for the sphinx requirements that's blocking the README fixes needed.
19:24:09 <diablo_rojo_phon> Glad it want wasted effort :)
19:24:28 <smcginnis> I think I just pushed tripleo a little earlier.
19:24:43 <evrardjp> yup saw that :) thanks smcginnis
19:25:04 <smcginnis> We have monasca (ready), tricircle, and puppet.
19:25:23 <tonyb> so how much longer do we give tricircle or puppet?
19:25:51 <tonyb> and heat is a whole other question
19:26:14 <smcginnis> A couple heat and ansible issues that are bigger.
19:26:18 <tonyb> we did say today we'd just tag the existing releases as pike-em and declare it done
19:26:27 <evrardjp> ansible ?
19:26:32 <smcginnis> diablo_rojo_phon: Is that what you communicated? ^^
19:26:36 <diablo_rojo_phon> Yeah.
19:26:49 <diablo_rojo_phon> I said if there was no action by today we'd just tag them
19:26:51 <evrardjp> something new I should be aware ?
19:27:05 <tonyb> evrardjp: https://etherpad.openstack.org/p/train-relmgt-tracking Line 118
19:27:09 <smcginnis> evrardjp: Links to failures in the etherpad, but os_molteniron had a releasenotes job failure and python-pankoclient had a tagging failure.
19:27:44 <smcginnis> The python-pankoclient failure is a bit odd.
19:28:35 <tonyb> Umm yeah that is odd
19:28:36 <evrardjp> I see, it's probably with the fact we discussed retiring this repo
19:28:46 <evrardjp> but it isn't retired yet
19:29:02 <smcginnis> So maybe both of those are safe to ignore and we can just do the final taggin?
19:29:06 <smcginnis> g
19:29:23 <smcginnis> Err, wait. I think these were on the final taggin.
19:29:29 <evrardjp> I can't judge for pankoclient
19:29:30 <smcginnis> Gah, can't spell ing.
19:29:51 <tonyb> smcginnis: yeah I think at least pankoclient was the pike-em tag
19:30:13 <tonyb> smcginnis: and the tag is in that repo
19:30:13 <fungi> were there any failures associated with the zuul artifacts issue yesterday which need to be rerun? we had to dequeue some refs, i think tripleo-ui at a minimum
19:30:28 <smcginnis> fungi: None I've seen.
19:30:37 <smcginnis> os_molteniron has the pike-em tag too.
19:30:39 <smcginnis> So I guess we're good.
19:30:50 <tonyb> fungi: like what smcginnis said ;P
19:31:12 <smcginnis> So Heat's the big question mark at this point.
19:31:18 <tonyb> yeah
19:31:21 <smcginnis> But also maybe something we can just ignore and move on.
19:31:32 <tonyb> we can fix it but that kinda means a new release
19:32:00 <tonyb> I kinda think just declare pike-em done ... with some issues and aim for better with queens
19:32:13 <smcginnis> I'm fine with that.
19:32:28 <tonyb> which *should* be better if we do the automatic releases we disussed
19:32:32 <smcginnis> We'll hit this with queens too, so either we need to get them to do some work, or we can just expect it and ignore it again.
19:32:42 <ttx> ++
19:32:44 <dhellmann> what's the story with heat? is there something blocking the name change?
19:32:49 <tonyb> less bitrot anyway, and less delta from the last release
19:33:19 <smcginnis> dhellmann: They changed from heat to openstack-heat in later (sooner?) branches, but didn't go back to pike and queens with those updates.
19:33:29 <dhellmann> yeah, so they just need to fix that, right?
19:33:32 <tonyb> dhellmann: I don't think it's blocked just hasn't been done
19:33:38 <dhellmann> ok
19:33:39 <tonyb> dhellmann: Yup
19:33:41 <smcginnis> That would make things a little easier.
19:33:44 <dhellmann> as long as it's just a matter of not doing it
19:33:57 <dhellmann> tell them there won't be any new releases from any branch until that's done :-)
19:34:05 <tonyb> dhellmann: ;P
19:34:57 <diablo_rojo_phon> Haha :)
19:35:57 <dhellmann> or, you know, make the validation job report that failure
19:36:01 <dhellmann> do what you have to
19:37:16 <evrardjp> that doesn't sounds like a problem a little socialization wouldn't fix
19:37:29 <tonyb> dhellmann: validation is already failing
19:38:08 <dhellmann> tonyb : yeah, I was making a joke about having it fail on master, too
19:38:13 <dhellmann> "all branches must be publishable"
19:38:26 <tonyb> evrardjp: I think the only tension really is our desire to get the pike-em stuff off the plate and the (probably small) delays that fixinx it in heat will take
19:38:54 <evrardjp> makes sense
19:38:55 <tonyb> the heat team is alsmost certainly okay with it and has been pretty responsive
19:39:03 <tonyb> dhellmann: ahh sorry I missed it ;P
19:40:53 <tonyb> Let's just fix queens and declare pike done, as per the email diablo_rojo_phon sent
19:41:15 <tonyb> it's a little unfair to heat but then we can put pike behind us
19:41:25 <smcginnis> ++
19:41:33 <diablo_rojo_phon> Works for me.
19:42:01 <tonyb> okay moving on
19:42:02 <smcginnis> So just to make sure I have the plan, drop any outstanding final releases (other than Monasca that is ready to go) and update the pike-em patches to just use the last successfully release versions.
19:42:06 <smcginnis> Right?
19:42:13 <tonyb> #topic Do we want to add per-projects links for Feature freeze
19:42:18 <tonyb> smcginnis: correct
19:42:38 <tonyb> so on $topic
19:42:43 <smcginnis> I raised this one on a few reviews of project specific deadlines.
19:42:57 <smcginnis> A few have added $project Feature Freeze on the overall feature freeze week.
19:43:12 <smcginnis> So.. a bit redundant IMO as that applies to everyone already.
19:43:31 <tonyb> yup
19:43:53 <ttx> I'd say unless they have a specific thing it's not worth mentioning each
19:43:54 <diablo_rojo_phon> I think if it's different than the official feature freeze (like being earlier)  it would be fine.
19:44:05 <diablo_rojo_phon> But it doesn't make sense to restate what is already there.
19:44:10 <smcginnis> Yeah, earlier is good and should definitely be called out.
19:44:21 <smcginnis> So everyone OK if I put up a patch to remove those?
19:44:40 <diablo_rojo_phon> Fine by me.
19:44:50 <smcginnis> We could even still leave the text descriptions in under the project section, just not have them linked from the weekly table.
19:45:01 <tonyb> smcginnis: yup
19:45:20 <smcginnis> OK, I'll take that action then if no one is against the idea.
19:45:55 <tonyb> smcginnis: all yours
19:46:11 <tonyb> #topic openstack-manuals owned by releases team?
19:46:17 <smcginnis> I added this one.
19:46:44 <smcginnis> Not sure if folks saw, but there's been a lot of conversation on the ML and in the -tc channel about the idea of getting rid of the docs team.
19:47:10 * tonyb missed the chatter but saw the reviews
19:47:11 <smcginnis> And one idea raised was whether it made sense to move openstack-manuals under the release team since it is now just a coordinating spot for docs.
19:47:29 <smcginnis> And there's just some procedural stuff that needs to happen around release time.
19:47:42 <smcginnis> Other than some tooling that other folks sound willing to still maintain.
19:48:05 <smcginnis> So just wanted to raise it here to get folks thinking about it, pro or con, in case that developed more.
19:48:17 <smcginnis> I'm not sure what my opinion is yet. :)
19:49:06 <tonyb> I haven't checked the review, but I more or less said openstack-manuals should be owned by a project team, realeases the least worse
19:49:42 <diablo_rojo_phon> I guess it kinda makes logical sense to live under us.
19:50:06 <ttx> If the mainenance needed is all around release cycles, then yes it makes sense
19:50:51 <tonyb> I know we have items around release time that we corordinate with them, but that isn't all of it
19:50:55 <smcginnis> I think I'd like to understand a little more of what that team currently does throughout the cycle to make sure we don't sign up for too much should the existing docs folks get pulled away.
19:51:06 <tonyb> for example I recent;y added the EM SIG to the site
19:51:28 <diablo_rojo_phon> Could make them like a sub team?
19:51:37 <tonyb> that's very "not releases", and I'd be pretty uncomfortable owning that side of things
19:51:40 <diablo_rojo_phon> smcginnis +1
19:51:53 <tonyb> so as long as we can keep some of the existing core team ...
19:52:30 <ttx> yes I feel like I'm missing details on what it would take to handle that
19:52:52 <tonyb> okay, so who wants to run that investigation?
19:53:04 * tonyb kinda looks at TC memebers ;P
19:53:12 <tonyb> oh look we have a few of those ;P
19:53:16 <ttx> I think dhellmann has a better understanding than most of what is involved
19:53:28 <ttx> and we need to make a decision soon enough
19:53:37 <smcginnis> I think I'm leaning towards rather still having a docs team and stepping in if Stephen just doesn't want to be the PTL for it.
19:53:49 <clarkb> Not to interrupt but I wanted to mention that gerrit replication has caught up so it should be safe to do release work again (thank you for your patience)
19:53:58 <ttx> \o/
19:54:00 <smcginnis> clarkb: Thanks!
19:54:05 <tonyb> clarkb: \o/
19:55:21 <tonyb> I'd rathre keep the docs team too but that doens't seem to the be the direction the docs team is going
19:56:57 <ttx> ok no time left so let's iterate on that review
19:57:04 <tonyb> okay
19:57:11 <tonyb> last item shoudl be quick
19:57:18 <tonyb> #topic meeting next week
19:57:51 <tonyb> many of y'all are in BCN next week and I assume will not be wanting to have an IRC meeting at 2000 on Thursday
19:57:53 <smcginnis> Most of us will be out.
19:57:56 <tonyb> so cancel it?
19:58:24 <tonyb> that has the benfit that I get an extra hours sleep ;P
19:58:36 <smcginnis> Then I'm in favor of cancelling. :)
19:58:41 <tonyb> \o/
19:58:59 <tonyb> y'all can have an in bar^wperson meeting instead
19:58:59 <ttx> yes cancel
19:59:07 <ttx> I'll be on a plane anyway
19:59:12 <tonyb> pfft
19:59:26 <tonyb> okay I think we're done here
19:59:33 <evrardjp> thanks everyone
19:59:40 <ttx> the sort of plane that does not have wifi
20:00:00 <tonyb> #endmeeting