| *** hongbin has joined #openstack-release | 00:44 | |
| *** mriedem has joined #openstack-release | 00:45 | |
| mriedem | smcginnis: dhellmann: we got https://review.openstack.org/#/c/591129/ on master - we're doing an rc2 so should we backport that? | 00:45 |
|---|---|---|
| * mriedem asks in #openstack-i18n | 00:45 | |
| mriedem | nvm got my ansnwer | 01:00 |
| *** mriedem has left #openstack-release | 01:00 | |
| *** njohnston_ is now known as njohnston | 01:56 | |
| *** ekcs has quit IRC | 02:02 | |
| *** ricolin has joined #openstack-release | 02:19 | |
| *** openstack has joined #openstack-release | 02:33 | |
| *** ChanServ sets mode: +o openstack | 02:33 | |
| openstackgerrit | Hongbin Lu proposed openstack/releases master: Release Zun-ui 2.0.0 and cut stable branch https://review.openstack.org/591889 | 02:53 |
| *** bobh has quit IRC | 03:05 | |
| *** bobh has joined #openstack-release | 03:10 | |
| *** bobh has quit IRC | 03:19 | |
| *** hongbin has quit IRC | 03:23 | |
| *** hongbin has joined #openstack-release | 03:28 | |
| *** armax has quit IRC | 03:34 | |
| *** hongbin_ has joined #openstack-release | 03:48 | |
| *** hongbin has quit IRC | 03:50 | |
| *** njohnston has quit IRC | 04:02 | |
| *** hongbin_ has quit IRC | 04:14 | |
| *** samP has joined #openstack-release | 04:19 | |
| *** oanson has quit IRC | 04:33 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/releases master: Tacker: release Tacker server rocky 0.10.0rc1 https://review.openstack.org/591892 | 04:47 |
| *** oanson has joined #openstack-release | 05:00 | |
| openstackgerrit | Michael Johnson proposed openstack/releases master: Tag Rocky release of octavia-tempest-plugin https://review.openstack.org/591780 | 05:02 |
| openstackgerrit | dharmendra kushwaha proposed openstack/releases master: Tacker: release Tacker server rocky 0.10.0rc1 https://review.openstack.org/591892 | 05:05 |
| *** pcaruana has joined #openstack-release | 05:12 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/releases master: tacker-horizon rocky release 0.12.0 https://review.openstack.org/591900 | 05:42 |
| *** e0ne has joined #openstack-release | 06:16 | |
| *** alexchadin has joined #openstack-release | 06:59 | |
| openstackgerrit | Alexander Chadin proposed openstack/releases master: Add cycle-highlights to Watcher Rocky https://review.openstack.org/591325 | 07:32 |
| *** electrofelix has joined #openstack-release | 07:40 | |
| *** jpich has joined #openstack-release | 07:41 | |
| *** dkushwaha has joined #openstack-release | 08:06 | |
| *** dtantsur|afk is now known as dtantsur | 08:26 | |
| *** lennyb has joined #openstack-release | 08:37 | |
| *** dkushwaha has quit IRC | 08:40 | |
| *** alexchadin has quit IRC | 09:00 | |
| *** samueldmq_ has joined #openstack-release | 09:16 | |
| *** andreaf has quit IRC | 09:16 | |
| *** andreaf has joined #openstack-release | 09:17 | |
| *** alexchadin has joined #openstack-release | 09:17 | |
| *** dirk_ has joined #openstack-release | 09:18 | |
| *** dirk has quit IRC | 09:18 | |
| *** samueldmq has quit IRC | 09:18 | |
| *** samueldmq_ is now known as samueldmq | 09:18 | |
| *** dirk_ is now known as dirk | 09:18 | |
| *** ricolin has quit IRC | 09:29 | |
| *** dave-mccowan has joined #openstack-release | 10:12 | |
| *** olivierbourdon38 has joined #openstack-release | 10:31 | |
| *** alexchadin has quit IRC | 10:59 | |
| *** alexchadin has joined #openstack-release | 11:00 | |
| *** alexchadin has quit IRC | 11:00 | |
| *** alexchadin has joined #openstack-release | 11:01 | |
| *** alexchadin has quit IRC | 11:01 | |
| *** alexchadin has joined #openstack-release | 11:01 | |
| *** alexchadin has quit IRC | 11:02 | |
| *** ricolin has joined #openstack-release | 11:20 | |
| *** alexchadin has joined #openstack-release | 11:33 | |
| *** alexchadin has quit IRC | 11:38 | |
| *** olivierbourdon38 has quit IRC | 11:55 | |
| *** alexchadin has joined #openstack-release | 11:56 | |
| *** rosmaita has joined #openstack-release | 11:57 | |
| *** bobh has joined #openstack-release | 11:57 | |
| *** bobh has quit IRC | 12:03 | |
| *** bobh has joined #openstack-release | 12:09 | |
| *** bobh has quit IRC | 12:13 | |
| *** gcb_ has joined #openstack-release | 12:19 | |
| *** alexchadin has quit IRC | 12:23 | |
| *** alexchadin has joined #openstack-release | 12:30 | |
| *** bobh has joined #openstack-release | 12:32 | |
| openstackgerrit | Merged openstack/releases master: Cut stable/rocky branch for devstack-plugin-container https://review.openstack.org/591804 | 12:32 |
| *** pkovar has joined #openstack-release | 12:35 | |
| openstackgerrit | Merged openstack/releases master: Release osc-lib 1.11.1 (Rocky) https://review.openstack.org/591556 | 12:37 |
| *** ykarel has joined #openstack-release | 12:37 | |
| *** ykarel is now known as ykarel|away | 12:49 | |
| dhellmann | melwitt : ah, good | 12:55 |
| smcginnis | It is a bit confusing that translations are going into master (stein) and not rocky as we are finalizing that release. | 12:55 |
| dhellmann | we should probably talk with the i18n team about that. I imagine it has something to do with how easy it is to configure zanata | 12:56 |
| *** ykarel|away has quit IRC | 13:17 | |
| *** luizbag has joined #openstack-release | 13:29 | |
| *** luizbag has quit IRC | 13:30 | |
| dhellmann | smcginnis : are we all clear for tagging and branching? It looks like we have 2-3 we could send through. | 13:44 |
| dhellmann | I'm assuming we want to hold off on the requirements branch for now | 13:44 |
| *** annabelleB has joined #openstack-release | 13:55 | |
| smcginnis | dhellmann: Yep, I believe we are clear to go. Let's check with Matt when he actually wants to proceed with requirements. | 14:08 |
| openstackgerrit | Merged openstack/releases master: Add cycle-highlights to Watcher Rocky https://review.openstack.org/591325 | 14:20 |
| annabelleB | dhellmann smcginnis: Instead of the usual “marketing preview” for Rocky, we were thinking of just having one big community meeting open to anyone—might try to get a few PTLs to talk about their work too. Unfortunately 8am on the 30th is the most available time. Any reason that that’s a terrible time? | 14:29 |
| smcginnis | annabelleB: What time zone? That would really be the only challenge I see. No time is going to work well for everyone. | 14:32 |
| annabelleB | smcginnis: whoops, sorry, yes, 8am pacific | 14:32 |
| *** alexchadin has quit IRC | 14:32 | |
| *** armax has joined #openstack-release | 14:49 | |
| *** d0ugal has joined #openstack-release | 14:58 | |
| *** annabelleB has quit IRC | 14:59 | |
| *** annabelleB has joined #openstack-release | 15:07 | |
| *** rpioso|afk is now known as rpioso | 15:08 | |
| *** masayukig has joined #openstack-release | 15:10 | |
| e0ne | hi team. does anybody know when we'll have stable/rocky in the openstack/requirements project? | 15:24 |
| e0ne | horizon needs if for rc2 | 15:26 |
| prometheanfire | I'll need to update the sha for the reqs branch for the osc-lib update, hope there are no others :P | 15:34 |
| dhellmann | e0ne : why does horizon need a branch of requirements for rc2? | 15:37 |
| dhellmann | that doesn't sound right | 15:37 |
| e0ne | dhellmann: sorry, looks like my mistake. I'll confirm it a bit later today once CI will report on a backport | 15:37 |
| prometheanfire | k, ya, that does sound odd | 15:37 |
| dhellmann | smcginnis , prometheanfire : we've had some confusion this cycle with the constraints update patches that we create when we open branches. I wonder if we want to wait to create those until we have the requirements branch? it would require separate scripting but we should just have to rearrange some of the pieces we already have | 15:38 |
| dhellmann | e0ne : ok, let us know | 15:38 |
| smcginnis | I was wondering about that too. There does seem to be more confusion than before on that. | 15:40 |
| smcginnis | Even though the commit message says exactly what to expect. | 15:40 |
| prometheanfire | dhellmann: I +2'd one of those updates without realizing we didn't branch first :P | 15:40 |
| dhellmann | smcginnis : who reads commit messages? ;-) | 15:40 |
| prometheanfire | so maybe ordering is hard | 15:40 |
| prometheanfire | dhellmann: not I | 15:40 |
| prometheanfire | who needs it | 15:40 |
| * dhellmann doesns't always | 15:40 | |
| dhellmann | I'm not sure what we would use as a trigger | 15:40 |
| dhellmann | maybe a special case in the branch script | 15:41 |
| *** annabelleB has quit IRC | 15:41 | |
| dhellmann | it would end up having to clone all of the deliverables for the series at once, which might cause problems with job timeouts | 15:41 |
| dhellmann | the alternative would be to run the script by hand, since that wouldn't have the same timeout problem | 15:41 |
| openstackgerrit | Merged openstack/releases master: Tag Rocky release of octavia-tempest-plugin https://review.openstack.org/591780 | 15:41 |
| dhellmann | that goes against our automation push, though | 15:42 |
| dhellmann | ooo, another alternative would be to propose the requirements branch patch and WIP it and then somehow pass that URL to the script that creates all of the constraint update patches so it could be included in a depends-on | 15:43 |
| dhellmann | that's a bit complicated, though | 15:43 |
| prometheanfire | that'd have a proper dependency at least | 15:46 |
| dhellmann | yeah | 15:46 |
| dhellmann | the timeout thing is a real concern, though, so we would want to test that somehow to see how long it would take | 15:47 |
| smcginnis | Could we add a "release" job for the requirements repo that would check if a new branch was created and have that trigger it? | 15:49 |
| dhellmann | hmm | 15:50 |
| dhellmann | that's a good question | 15:50 |
| dhellmann | I don't know if we can trigger on branch creation | 15:50 |
| dhellmann | fungi, corvus : can we have a job triggered by zuul when a branch is created? ^^ | 15:50 |
| dhellmann | it's a ref, so I would think so? | 15:51 |
| *** annabelleB has joined #openstack-release | 15:51 | |
| corvus | dhellmann: yes | 15:52 |
| corvus | (more to come) | 15:52 |
| dhellmann | do we have any examples we could look at? | 15:52 |
| *** FracKen has joined #openstack-release | 15:56 | |
| corvus | dhellmann: branch creation events should trigger the post pipeline. you can test for whether it's a created event in the job by checking if https://zuul-ci.org/docs/zuul/user/jobs.html#var-zuul.oldrev is defined. | 15:56 |
| dhellmann | corvus : ok, so our job would run when something merges and when a branch is created, and we could test for that variable to decide if we need to actually do any work | 15:58 |
| corvus | dhellmann: yep | 15:59 |
| dhellmann | that should do what we want, thanks! | 15:59 |
| dhellmann | prometheanfire , smcginnis : I made some notes in https://storyboard.openstack.org/#!/story/2003446 | 16:03 |
| smcginnis | Great! | 16:04 |
| dhellmann | there was some other work going on at one point to change the way that constraint list is managed/published so we might want to tie in with that | 16:04 |
| dhellmann | tonyb was doing something but I don't remember the details | 16:05 |
| prometheanfire | dhellmann: I'm thinking of migrating reqs to storyboard now that we only have one bug and just deal with anything else as it comes up | 16:07 |
| dhellmann | sounds good to me | 16:07 |
| *** pkovar has quit IRC | 16:09 | |
| *** jpich has quit IRC | 16:25 | |
| melwitt | smcginnis: re: translations, we were told that the translations are indeed not for rocky and that we are not supposed to backport them. so I think I've not been understanding what the past "rc2 translations" were about | 16:34 |
| smcginnis | melwitt: I think there have been times where repos have gotten translations. It's just been one of the reasons to do an RC2 release. | 16:35 |
| dhellmann | we used to translate API messages | 16:41 |
| dhellmann | maybe we're not doing that any more, if they are focused on the documentation and release notes | 16:42 |
| *** annabelleB has quit IRC | 16:43 | |
| *** annabelleB has joined #openstack-release | 16:47 | |
| *** e0ne has quit IRC | 16:47 | |
| *** ricolin has quit IRC | 16:50 | |
| *** openstackstatus has joined #openstack-release | 16:58 | |
| *** ChanServ sets mode: +v openstackstatus | 16:58 | |
| *** annabelleB has quit IRC | 17:10 | |
| edmondsw | several of the "Update UPPER_CONSTRAINTS_FILE for stable/rocky" commits have already been merged, despite the commit message clearly saying they should not yet | 17:29 |
| edmondsw | could these be proposed later next time, or have a -W until openstack/requirements branches? | 17:30 |
| *** annabelleB has joined #openstack-release | 17:32 | |
| openstackgerrit | Matthew Thode proposed openstack/releases master: Branch requirements https://review.openstack.org/591823 | 17:35 |
| *** dtantsur is now known as dtantsur|afk | 17:40 | |
| smcginnis | edmondsw: We were talking about that earlier actually. | 17:44 |
| smcginnis | edmondsw: Seems not many people read commit messages. :) | 17:44 |
| edmondsw | seems like it :) | 17:44 |
| edmondsw | I can go back and read the history if you've already discussed | 17:44 |
| smcginnis | edmondsw: We might be able to delay having those proposed and have the actual branching of requirements trigger the creation of them. | 17:44 |
| edmondsw | +1 | 17:45 |
| smcginnis | It's not an issue in the gate since that passes in a variable, but it can cause problems running tests locally. | 17:45 |
| *** e0ne has joined #openstack-release | 17:56 | |
| *** ekcs has joined #openstack-release | 18:05 | |
| *** ykarel has joined #openstack-release | 18:11 | |
| tonyb | dhellmann, smcginnis: Yup IIUC the plan for the constraints publishing will pretty much remove that confusion | 18:14 |
| smcginnis | tonyb: What is the plan? | 18:15 |
| tonyb | if it had been done then today we'd have valid constraints files releases.o.o/constraints/{rocky,stein}/upper | 18:16 |
| tonyb | So we can always just update it and the state of the requirements repo doesn't matter | 18:16 |
| smcginnis | So it won't be dependent on branches in the repo? | 18:16 |
| tonyb | smcginnis: It will and it wont. for and stable branch in requiremenst we'll publish that file "eg constraints/queens/upper" but on master we'll publish both current and current+1 (rocky, stein) | 18:18 |
| tonyb | once we branch we'll switch master to stein and once it's known "train" | 18:18 |
| smcginnis | Ah, OK. That sounds like it should ease the transition between release cycles then. | 18:19 |
| tonyb | it's a little more work for the requirements team but less on projects and the cross over time is well defined | 18:19 |
| tonyb | smcginnis: It was supposed to happen by now but I didn't do the needed project management so .... | 18:20 |
| *** electrofelix has quit IRC | 18:25 | |
| dhellmann | tonyb , smcginnis : so we'll be doing the constraints switch on master instead of the stable branch? | 18:35 |
| tonyb | dhellmann: the realease tooling wont change. | 18:36 |
| dhellmann | that is, we'd need a patch to change the URL in the tox.ini still, but on master | 18:36 |
| dhellmann | ? | 18:36 |
| dhellmann | or is tox going to somehow figure out the URL in a way that doesn't require patches? | 18:36 |
| tonyb | dhellmann: It's just that *if* the work had been done we'd have a releases.o.o/constraints/rocky/upper file already | 18:37 |
| dhellmann | yeah, I'm trying to understand how the client side of it will work when we go from stein to tuppence | 18:37 |
| dhellmann | do we still edit tox.ini each cycle? | 18:38 |
| tonyb | dhellmann: Yup we still need the edit | 18:38 |
| dhellmann | ok | 18:38 |
| tonyb | dhellmann: we're just more flexible on when we do it | 18:38 |
| dhellmann | ah, right, I see how that helps now | 18:38 |
| dhellmann | because we would have both stein and tuppence published at the same time, that constraints URL patch could land right away | 18:39 |
| dhellmann | in the stable branch | 18:39 |
| tonyb | dhellmann: most projects will still do it as a result of the release tooling but it's possible some (like perhaps tripleo) will do it themselves | 18:39 |
| dhellmann | yeah | 18:39 |
| tonyb | dhellmann: Yup. That's how it's supposed to work ;P | 18:39 |
| dhellmann | I think we should probably write all of this out to make sure the details are clear | 18:39 |
| dhellmann | I guess we don't really have specs for the requirements and release work, so maybe an email thread | 18:40 |
| tonyb | dhellmann: we had one wheer we came up with the design | 18:40 |
| dhellmann | oh, good | 18:40 |
| tonyb | dhellmann: it's just close to a year ago now :/ | 18:40 |
| dhellmann | I guess I should go find that | 18:40 |
| dhellmann | because I'm left wondering how master ends up getting the right name | 18:40 |
| dhellmann | or is master always just "master"? | 18:41 |
| tonyb | dhellmann: It could probably stand a summary ;p | 18:41 |
| dhellmann | I could use a refresher, at least | 18:41 |
| dhellmann | maybe if we can get that together by the ptg we can find someone to help you with the work | 18:41 |
| tonyb | So today master will publish to both rocky and stein. | 18:42 |
| tonyb | once we branch it drops master | 18:42 |
| dhellmann | what's listed in the tox.ini on master? | 18:42 |
| *** annabelleB has quit IRC | 18:42 | |
| tonyb | when we know what t will we we add that | 18:42 |
| dhellmann | hmm | 18:42 |
| tonyb | ... oh and it's also always master | 18:42 |
| dhellmann | ah, ok | 18:42 |
| dhellmann | so we still only have to make changes to the URL on the stable branch | 18:43 |
| dhellmann | but the stable branch name will work even before the repo is branched | 18:43 |
| dhellmann | so on master tox points to a master URL and on the stable branch it points to master and we change it to the branch name | 18:43 |
| tonyb | dhellmann: Yup from the release side the process doesn't chnage (but the URL will) it's just that the target URL will exist regardless of the branch existing in requirements | 18:43 |
| dhellmann | we could roll all of these changes into a single patch on the stable branch if that's the case | 18:43 |
| dhellmann | .gitreview, tox.ini, etc. | 18:43 |
| dhellmann | I wish we could avoid having to patch at all, though | 18:44 |
| tonyb | dhellmann: Yup, at some point we resisted doign it in a single patch I don't recall why but times chnage ... | 18:44 |
| dhellmann | we can't even rely on having those .gitreview patches landed quickly, though, and I don't really see another way to find the branch name accurately | 18:45 |
| dhellmann | well for now we need .gitreview landed ASAP and we have to wait for the constraints URL update, so that's why they are separate patches today | 18:45 |
| tonyb | dhellmann: fro tox.ini can can avoid the patch but it in turn needs a helper in install_cmd | 18:45 |
| dhellmann | if we eliminate the need to wait to land the constraints update we can combine them | 18:45 |
| dhellmann | yeah | 18:45 |
| tonyb | dhellmann: Ahh I understand now | 18:46 |
| dhellmann | but any helper would need to do something to determine the branch, and you can't just look at git for that because locally my branch might be called 'backport-bug-fix' or something and what stable branch is that? | 18:46 |
| dhellmann | so we could read .gitreview to find the branch, but only if the patch to update .gitreview is approved quickly after being proposed | 18:46 |
| dhellmann | which in the past they have been, but this cycle I've run into quite a few that weren't | 18:47 |
| tonyb | dhellmann: my vauge plan was to use the release data and the version from pbr | 18:47 |
| dhellmann | I actually had to rewrite some of the goal tool scripts to be explicit about the branch where they were proposing patches to deal with that | 18:47 |
| dhellmann | hmm | 18:47 |
| dhellmann | I wonder if we were able to shift things so that we needed to land patches on master instead of the stable branch if that would mean they would be landed faster | 18:48 |
| dhellmann | like if we put stein in the URL on master, and then when we branch for tuppence we submit the patches on master to use tuppence | 18:48 |
| dhellmann | more people review patches on master... | 18:48 |
| *** annabelleB has joined #openstack-release | 18:49 | |
| tonyb | dhellmann: so I'm typing this straight into IRC so it'll have errors but I was thinking we'd just the same kind of thing we do for docs so that .../constratints/nova/16.*/ redirects to constratints/upper/pike (or whatever the correct mappign is | 18:49 |
| dhellmann | aha! | 18:49 |
| dhellmann | yes, version redirects might help | 18:49 |
| tonyb | dhellmann: switching it to that could work | 18:49 |
| dhellmann | as far as the helper, we could release that as a tool, and then tell developers to install tox and that tool | 18:50 |
| dhellmann | that would avoid the problem we had in the past where the tool was copied around a million places | 18:50 |
| dhellmann | and then diverged | 18:50 |
| dhellmann | but that also requires special tox config, and we've been moving away from that | 18:50 |
| dhellmann | so maybe let's try publishing with names and changing the names on master before we go down the tool route | 18:51 |
| tonyb | Yeah, the time I proposed it we settled on the simple branch based static process | 18:51 |
| tonyb | Yup. I'll try to get the ball rolling again | 18:53 |
| dhellmann | ++ | 19:06 |
| *** openstackgerrit has quit IRC | 19:19 | |
| *** annabelleB has quit IRC | 19:20 | |
| *** annabelleB has joined #openstack-release | 19:26 | |
| *** ykarel has quit IRC | 19:29 | |
| *** onovy has quit IRC | 19:37 | |
| *** onovy has joined #openstack-release | 19:45 | |
| *** annabelleB has quit IRC | 19:48 | |
| *** annabelleB has joined #openstack-release | 20:01 | |
| *** e0ne has quit IRC | 20:03 | |
| prometheanfire | reqs is ready to branch btw | 20:50 |
| smcginnis | prometheanfire: OK, I'll hit go on that. | 20:51 |
| prometheanfire | green means go | 20:51 |
| annabelleB | dhellmann smcginnis: I was reading through https://review.openstack.org/#/c/591778/ to see the changes, and one of the open patches is for something where (on the patch) they agreed to drop it. Do we apply the “close out your open patches close to the release” rule for cycle-intermediary projects? | 20:54 |
| smcginnis | annabelleB: Usually there are a lot of open patches for most projects. They just miss this release and hopefully get merged early in the next release cycle. | 20:56 |
| annabelleB | smcginnis: Got it! | 20:56 |
| smcginnis | annabelleB: Which one were you concerned about? | 20:56 |
| dhellmann | if the patch didn't make it we probably want to drop the highlight | 20:57 |
| smcginnis | Oh, it's something in the highlights? | 20:57 |
| annabelleB | smcginnis: https://review.openstack.org/578424 It just caught my eye since it was removing something and I went to check it out | 20:57 |
| dhellmann | annabelleB : if you want to propose a patch to drop the highlight we can get hongbin to +1 it | 20:57 |
| annabelleB | haven’t seen highlights from them so was hunting through the most recent release changes | 20:58 |
| *** openstackgerrit has joined #openstack-release | 20:59 | |
| openstackgerrit | Merged openstack/releases master: Branch requirements https://review.openstack.org/591823 | 20:59 |
| annabelleB | It’s just a docs change, but made me wonder—just generally—how much we do/don’t enforce closing open things towards the end of cycles | 21:00 |
| dhellmann | "closing open things"? | 21:01 |
| dhellmann | I think we just generally expect teams to keep their code and docs in sync as much as they can | 21:02 |
| annabelleB | seems reasonable | 21:03 |
| *** masayukig has quit IRC | 21:12 | |
| *** annabelleB has quit IRC | 22:02 | |
| *** ekcs has quit IRC | 22:08 | |
| *** masayukig has joined #openstack-release | 22:19 | |
| *** FracKen has left #openstack-release | 22:21 | |
| *** d0ugal has quit IRC | 22:33 | |
| *** d0ugal has joined #openstack-release | 22:44 | |
| *** d0ugal has quit IRC | 23:01 | |
| *** d0ugal has joined #openstack-release | 23:03 | |
| *** rpioso is now known as rpioso|afk | 23:12 | |
| *** d0ugal has quit IRC | 23:42 | |
| *** d0ugal has joined #openstack-release | 23:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!