| *** amoralej|off is now known as amoralej | 07:40 | |
| *** dviroel|afk is now known as dviroel | 11:43 | |
| opendevreview | Dmitriy Rabotyagov proposed openstack/releases master: Last release of OSA Wallaby https://review.opendev.org/c/openstack/releases/+/864262 | 11:57 |
|---|---|---|
| *** dviroel is now known as dviroel|doc_appt | 12:45 | |
| fungi | elodilles: regarding 864261, do you have any more detail on why the release announcement jobs need to install projects' bindep lists? what "garbage content" arises in them if you don't? | 12:50 |
| elodilles | fungi: there was a comment in the code that when building sdist some deliverables installs extra packages and that installation output would end up in the mail. let me search for the exact comment | 13:04 |
| fungi | mainly wondering why the job would need to install the deliverable at all | 13:04 |
| fungi | is it for generating the release notes? | 13:07 |
| fungi | nevermind, the announcements don't include release notes, those are git logs | 13:07 |
| fungi | yeah, just looking at a few of the announcements, i don't see anything in there which should need the package installed to generate. that's all coming from other sources of data, right? | 13:08 |
| fungi | that's the main reason i'm asking, it seems like the release announcements job could be fixed by making it just not do unnecessary steps | 13:15 |
| fungi | ahh, okay some of them do generate and include release notes. so the question becomes, should we have to install a project in order to generate its release notes? | 13:21 |
| fungi | and the answer is a definite "no" | 13:23 |
| fungi | i have a completely separate reno install on my workstation, and can run `reno report` in the git repo of any project and get release notes generated that way, without installing the project at all | 13:23 |
| fungi | so i doubt the installation is for supporting release notes inclusion | 13:24 |
| fungi | i see announce.sh runs `python setup.py --name` and `python setup.py --description` in the project... that doesn't require installation but we should replace that mechanism with something else anyway since setuptools has deprecated directly calling setup.py on the command line | 13:34 |
| elodilles | fungi: sorry, i was distracted with a downstream meeting :S so here is the comment: https://opendev.org/openstack/releases/src/branch/master/tools/announce.sh#L172-L175 | 13:40 |
| elodilles | but as i see you found the same code part | 13:42 |
| *** amoralej is now known as amoralej|lunch | 13:47 | |
| elodilles | i agree by the way: if the installation can be avoided then let's avoid it | 13:47 |
| fungi | looks like we could just remove https://opendev.org/openstack/openstack-zuul-jobs/src/branch/master/playbooks/release/pre.yaml#L3-L5 | 13:48 |
| fungi | at least my rudimentary local tests confirm that calling setup.py --name/--description doesn't install anything. but if we find places where it does try to, then we should probably look for a different solution than calling setup.py anyway | 13:49 |
| elodilles | but in that case (according to the comment I linked) we should refactor the name and description retrieval from setup.py | 13:50 |
| elodilles | yepp | 13:51 |
| elodilles | so you suggest to drop the 'announce-release' job duplication patch i did and simply remove the bindep install part? I can accept that if it really works | 13:52 |
| fungi | and yeah, going through the entire log of an announcement i find no evidence the project being released is ever actually installed | 13:55 |
| fungi | elodilles: yes, looking at the git history for that job, we were installing the project's bindep list when we first migrated it from zuul v2 job definitions, so it's quite likely that's just unnecessary cruft | 13:56 |
| elodilles | fungi: ack, then i'll create the patch in a minute :) | 13:56 |
| elodilles | but let's start the meeting now! | 14:00 |
| elodilles | #startmeeting releaseteam | 14:00 |
| opendevmeet | Meeting started Fri Nov 11 14:00:30 2022 UTC and is due to finish in 60 minutes. The chair is elodilles. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:00 |
| opendevmeet | The meeting name has been set to 'releaseteam' | 14:00 |
| elodilles | Ping list: hberaud armstrong elodilles | 14:00 |
| elodilles | #link https://etherpad.opendev.org/p/antelope-relmgt-tracking | 14:00 |
| ttx | o/ | 14:01 |
| elodilles | o/ | 14:01 |
| elodilles | we are at line 79 | 14:01 |
| elodilles | hberaud is on PTO, just as a reminder | 14:03 |
| elodilles | so let's start | 14:03 |
| elodilles | #topic Review task completion | 14:04 |
| elodilles | the 1st 2 tasks are related to Wallaby's Extended Maintenance transition | 14:04 |
| elodilles | almost every project have transitioned, i think ironic and swift patches can be merged (no PTL-approved though) | 14:05 |
| ttx | iirc there was a few leftovers yes | 14:05 |
| elodilles | https://review.opendev.org/q/topic:wallaby-em+is:open | 14:06 |
| elodilles | puppet openstack + openstack ansible want usually to transition last | 14:06 |
| ttx | how aggressively do we need to process the leftovers? | 14:07 |
| ttx | Can we wait a bit more? | 14:07 |
| elodilles | the final release for OSA is propsed, so that can be hopefully updated and merged | 14:07 |
| elodilles | ttx: we don't need to aggressively push those, so we can wait | 14:08 |
| ttx | ok | 14:08 |
| ttx | let's place a remnidner in the following weeks to reassess? | 14:08 |
| elodilles | ttx: em transition is not that strict but it is better not to postpone it too long | 14:08 |
| elodilles | ttx: that's OK for me | 14:09 |
| ttx | like in two weeks maybe | 14:09 |
| ttx | hmm skipped meeting, make that 3 weeks | 14:09 |
| elodilles | i'm optimistic and think that we can close the transition next week, anyway :) | 14:09 |
| ttx | added | 14:10 |
| elodilles | (if the puppet openstack patch is sorted out... fingers crossed...) | 14:10 |
| elodilles | ttx: thanks :) | 14:10 |
| elodilles | so let's move on to the next task: | 14:11 |
| elodilles | #info Review cycle-trailing projects to check which haven’t released yet (hberaud) | 14:11 |
| elodilles | mail was sent: | 14:11 |
| elodilles | https://lists.openstack.org/pipermail/openstack-discuss/2022-November/031095.html | 14:11 |
| elodilles | hberaud is not here, but anything to add here? | 14:12 |
| ttx | nono | 14:14 |
| elodilles | double negation? is that a yes? :) | 14:14 |
| ttx | nono | 14:14 |
| elodilles | anyway, move on :) | 14:14 |
| elodilles | #info move the announce job to focal for series older than antelope and keep jammy on master (elod) | 14:14 |
| elodilles | now we have 3 solution for this :) | 14:15 |
| elodilles | hberaud started to clean up the old bindeps: | 14:15 |
| elodilles | https://review.opendev.org/q/topic:drop-python-dev-from-bindep | 14:15 |
| elodilles | i've created the patch we agreed last meeting: | 14:15 |
| elodilles | https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/864261 | 14:16 |
| elodilles | and fungi suggested an even better fix: https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/864267 | 14:16 |
| fungi | i don't think the split job proposed in 864261 would have worked anyway. that job is triggered in the release pipeline for the project with the tag, but you're trying to filter on deliverables directories from the releases repo instead | 14:16 |
| elodilles | to simply remove the bindep installation | 14:16 |
| elodilles | fungi: oh, i see. then definitely the latter fix is the better | 14:17 |
| fungi | if you want to run different tag-triggered jobs on different platforms per branch, you need some way of guessing the branch associated with the tag | 14:17 |
| fungi | which is nontrivial | 14:18 |
| elodilles | i see | 14:18 |
| *** amoralej|lunch is now known as amoralej | 14:19 | |
| elodilles | fungi: i'll update the bindep removal patch based on your comment and then we are good to go | 14:19 |
| elodilles | * bindep task removal patch... | 14:19 |
| fungi | well, then we're ready to do a release-test release and make sure my supposition isn't incorrect | 14:19 |
| elodilles | fungi: good idea, i'll do a release-test release | 14:20 |
| elodilles | anyhow, i think we are on track with this | 14:21 |
| elodilles | and this was the last task for the week | 14:21 |
| elodilles | can we move on to the next topic? | 14:22 |
| fungi | seems like it | 14:22 |
| elodilles | #topic Assign next week tasks | 14:23 |
| elodilles | hberaud already added his name for some tasks | 14:23 |
| elodilles | and all taken as i see | 14:24 |
| elodilles | thanks! | 14:24 |
| elodilles | #topic Review countdown email | 14:24 |
| elodilles | https://etherpad.opendev.org/p/relmgmt-weekly-emails | 14:24 |
| elodilles | please review ^^^ | 14:24 |
| *** blarnath is now known as d34dh0r53 | 14:26 | |
| ttx | added final release date so that Antelope-1 is not too lonely | 14:26 |
| elodilles | ++ | 14:27 |
| ttx | otherwise lgtm | 14:27 |
| elodilles | cool, thanks! | 14:27 |
| elodilles | #topic Open Discussion | 14:27 |
| elodilles | anything to raise? | 14:27 |
| ttx | nothing here | 14:28 |
| elodilles | neither here | 14:29 |
| elodilles | so let's cut this meeting short then | 14:29 |
| fungi | nothing from me | 14:29 |
| elodilles | thanks for participating && next meetings chair will be hberaud | 14:30 |
| elodilles | as he volunteered last week :) | 14:30 |
| elodilles | #endmeeting | 14:30 |
| opendevmeet | Meeting ended Fri Nov 11 14:30:42 2022 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:30 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/releaseteam/2022/releaseteam.2022-11-11-14.00.html | 14:30 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/releaseteam/2022/releaseteam.2022-11-11-14.00.txt | 14:30 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/releaseteam/2022/releaseteam.2022-11-11-14.00.log.html | 14:30 |
| *** dviroel|doc_appt is now known as dviroel | 14:49 | |
| *** dviroel is now known as dviroel\LUNCH | 15:12 | |
| *** dviroel\LUNCH is now known as dviroel|lunch | 15:12 | |
| fungi | elodilles: the bindep removal worked, if you want to test | 15:25 |
| fungi | er, s/worked/merged/ | 15:25 |
| opendevreview | Elod Illes proposed openstack/releases master: release-test test release https://review.opendev.org/c/openstack/releases/+/864295 | 15:37 |
| elodilles | fungi: ack, let's see ^^^ | 15:38 |
| *** dviroel|lunch is now known as dviroel | 16:08 | |
| *** marios is now known as marios|out | 16:15 | |
| elodilles | oh, rsync: connection unexpectedly closed :S let's recheck | 16:44 |
| opendevreview | Merged openstack/releases master: release-test test release https://review.opendev.org/c/openstack/releases/+/864295 | 16:58 |
| *** amoralej is now known as amoralej|off | 17:21 | |
| elodilles | the announce job succeeded: https://zuul.opendev.org/t/openstack/build/9ecded74d74b4134ac74652ea97ad4bb | 17:46 |
| elodilles | (though there were no reno generated as there were no entry for the release...) | 17:47 |
| opendevreview | Merged openstack/releases master: [nova] Proposing EOL of Queens, Rocky and Stein https://review.opendev.org/c/openstack/releases/+/862520 | 18:01 |
| elodilles | so this seems to be OK, but maybe we should not approve only 1 release at a time, as soon as there will be a 'real' release patch, to check that the announce works well | 18:05 |
| fungi | yeah, makes sense | 18:07 |
| fungi | at least we can see it's not thoroughly broken anyway | 18:07 |
| elodilles | (...we *should* approve only 1 release patch at a time... to add some sense to my sentence :P) | 18:17 |
| opendevreview | Carlos Eduardo proposed openstack/releases master: [manila][zed] Add project tracking schedule https://review.opendev.org/c/openstack/releases/+/864313 | 21:20 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!