Wednesday, 2024-10-02

opendevreviewTakashi Kajinami proposed openstack/releases master: Puppet OpenStack 2024.2 release  https://review.opendev.org/c/openstack/releases/+/93111402:43
opendevreviewTakashi Kajinami proposed openstack/releases master: Puppet OpenStack 2024.2 release  https://review.opendev.org/c/openstack/releases/+/93111402:45
opendevreviewTakashi Kajinami proposed openstack/releases master: Puppet OpenStack 2024.2 release  https://review.opendev.org/c/openstack/releases/+/93111403:45
opendevreviewTakashi Kajinami proposed openstack/releases master: Puppet OpenStack 2024.2 release  https://review.opendev.org/c/openstack/releases/+/93111404:38
opendevreviewTakashi Kajinami proposed openstack/releases master: Puppet OpenStack 2024.2 release  https://review.opendev.org/c/openstack/releases/+/93111405:38
ttxIt's release day!07:38
elodilles~o~07:46
opendevreviewJoel Capitao proposed openstack/releases master: oslo.reports 3.4.1  https://review.opendev.org/c/openstack/releases/+/93114007:48
elodillesno new reno links, so we don't need to rebase the release patch \o/07:59
ttxyay08:20
elodillesthe semaphore patch has landed yesterday ( https://review.opendev.org/c/openstack/openstack-doc-tools/+/931029 ) so we are good with that too \o/09:31
elodillesbtw, i've proposed this patch, with freshly generated static sitemap, but i guess this is OK to be merged slightly after the release: https://review.opendev.org/c/openstack/openstack-doc-tools/+/93102909:33
ttxyeah, let's wait09:34
ttxReady to push the big button in 20min09:34
opendevreviewElod Illes proposed openstack/releases master: Mark 2024.2 Dalmatian as released  https://review.opendev.org/c/openstack/releases/+/93115309:45
elodillesi almost forgot that one ^^^09:45
elodillesttx: please review it, so that we can merge this later09:46
elodillesso far ' All Systems Operational ' @  https://status.python.org/09:47
elodillesT-509:55
fungii'm awake (for some definition thereof)09:57
elodillesgood morning fungi o/09:57
elodillesstatus.python.org still reports things as operational09:58
elodillesas i see zuul is also healthy09:58
elodillesmostly busy with kolla patches :)09:59
ttxok09:59
elodillesttx: are you around to push the button?09:59
elodilleshttps://review.opendev.org/c/openstack/releases/+/93075210:00
ttxbutton pushed10:00
elodilles\o/10:00
zigoDoes this releases EVERYTHING at once ?!?10:01
elodilles'status.python.org'10:01
elodilleszigo: yes and no: these are the cycle-with-rc release model deliverables, and the post-release jobs will do the processing somewhat parallel but it will take some time10:02
elodilless/''/10:03
elodillesStarting gate jobs.10:03
zigoelodilles: Right, but then I can expect everything in a few minutes then... :P10:03
elodillesa bit more than a few minutes, but yes :]10:04
fungizigo: "all at once" over the course of about 4-5 hours10:17
zigoOh ok. So I can go have lunch ... :P10:18
fungiit's mostly automated aside from a few manual checks that the automation didn't miss anything, but it takes a while to complete yes10:18
fungizigo: yes, lunch and more. the goal is to have it wrapped up and final documentation updates merged by 15:00 utc10:20
fungitags will probably be present in all repositories by 13z, with things like tarballs and wheels generated/uploaded by 14z10:22
fungipossibly sooner, but that's a conservative estimate10:23
ttxReleasing Everything everywhere all at once10:24
elodilles:)10:24
ttxOK, pausing for lunch10:38
opendevreviewMerged openstack/releases master: 2024.2 Dalmatian final releases for cycle-with-rc projects  https://review.opendev.org/c/openstack/releases/+/93075210:39
elodillesttx: Bon Appétit! (will do the same soon)10:40
elodillesand the patch has merged! \o/10:40
elodilleslet's see how the release jobs starts to work!10:40
fungitags are appearing now10:49
elodillesyepp. release queue is also increasing10:50
fungitagging seems to be done, release pipeline has 51 items now10:54
fungirelease pipeline topped out at 65 entries i think10:56
elodilles:-o10:57
fungiunfortunately all those kolla jobs had already sapped the available zuul capacity starting around 08:45 utc, so the release and tag jobs are having to wait for quota to free up in order to get new nodes assigned: https://grafana.opendev.org/d/21a6e53ea4/zuul-status10:59
elodillesi see some kolla patches in the check queue that have 15-20 mins left, so maybe we will have more nodes then o:)11:04
ttxback11:05
fungisome are running now11:13
fungiadjutant should have been uploaded11:13
fungiyeah https://pypi.org/project/python-adjutant/ looks good11:14
elodillesyepp, even nova and neutron are there on pypi11:15
fungithe tag pipeline is nearly empty already (releasenotes jobs)11:19
fungirelease pipeline count is starting to fall now that release announcement and constraint update proposal jobs are completing11:20
elodilles~o~11:22
fungithe node request backlog is nearly cleared so most jobs should be running by now11:27
fungiall release and tag jobs have completed, i see no reported failures11:43
ttxthat was... fast11:44
* ttx checks status11:44
ttxrunning missing-releases11:45
elodilles++11:46
ttxrunning into some issue with the script, AttributeError: module 'wheel.bdist_wheel' has no attribute 'safer_name'11:48
ttxinvestigating11:48
fungicould be due to distutils changes in recent setuptools?11:50
fungiwheel has apparently deprecated wheel.bdist_wheel in the most recent release: https://wheel.readthedocs.io/en/stable/news.html11:52
ttxyeah, bdist_wheel should no longer be consumed from wheel11:52
ttxI can workaround it quickly using _bdist_wheel instead11:52
ttxbut needs a more thorough fix11:52
ttxadding a note about that in postmortem11:54
elodillesACK11:55
elodilles(by using older setuptools the script works, so yepp...)11:56
ttxrun in progress, no issue so far11:58
funginot bad at all for being only 2 hours in12:00
ttxwondering if https://review.opendev.org/c/openstack/releases/+/929126 did not result in significantly faster turnaround12:02
ttxalright, no issue reporter in missing-releases12:05
fungithat would have sped up release announcement jobs, but any others?12:05
ttxelodilles: I think you can remove your W-1 from https://review.opendev.org/c/openstack/openstack-manuals/+/93033212:05
elodillesttx: done12:06
ttxany reason to hold on approving that one?12:06
fungiyeah, keep in mind that we've seen it take more than an hour to get the openstack-manuals change in and published, so getting started on that with plenty of time to spare would be good12:06
elodilles(so bdist_wheel is deprecated in 'wheel' package: https://github.com/pypa/wheel/commit/0add7d6c5fe6c03d7bec22d7d18f51d8a519862c )12:06
ttxI'm fine approving it now12:07
elodillesi'm OK with that, too12:07
ttx(I have a call in 53min)12:07
ttxAlright, approving12:07
elodillesttx: i don't have +2+W rights on openstack-manuals anyway o:)12:08
elodillesthanks o/12:08
fungizuul estimates about another 17 minutes for the promote job to complete, and then afs releases the read-only replica approximately every 5 minutes, after which time the docs site should reflect it12:37
elodilles++12:39
elodilles(meanwhile i've prepared the 2 announce mails content: https://etherpad.opendev.org/p/relmgmt-weekly-emails will send it around 15:00 UTC if everything goes fine)12:40
fungiawesome12:43
fungii'm on hand to approve openstack-announce posts when the time comes12:43
elodilles+1 thanks in advance12:45
fungipromote upload just finished, 13:00 vos release should publish it12:58
fungi2024-10-02 13:00:01,420 release DEBUG    Running: ssh -T -i /root/.ssh/id_vos_release vos_release@afs01.dfw.openstack.org -- vos release docs13:00
fungi2024-10-02 13:00:40,558 release INFO     Release of docs successful13:01
elodilles\o/13:01
fungihttps://docs.openstack.org/ is now redirecting to https://docs.openstack.org/2024.2/13:02
fungiwhich seems to have the expected content, e.g. "Currently viewing 2024.2 which is the current supported release." and "Documentation for 2024.2 (October 2024)"13:03
elodillesyepp, things look good there13:03
fungithe release notes link also correctly goes to https://releases.openstack.org/dalmatian/index.html13:04
fungiis https://review.opendev.org/c/openstack/releases/+/931153 the last thing before sending the announcement e-mails?13:07
elodillesthat is the next step, yes, and then we have '    Remind Foundation staff to switch openstack.org 2024.2 Dalmatian messaging on (ttx)' task :)13:08
elodillesand then the announcement e-mails13:09
ttxyeah, let's wait maybe 50min to flip the switch on the openstack.org website13:23
ttxlike ask them to do it around 14utc13:23
elodilles+113:24
ttxOK well it's done now13:28
ttxLet's prepare to approve https://review.opendev.org/c/openstack/releases/+/93115313:29
ttxand send announcements emails maybe 30minutes ahead of deadline13:29
ttxPing me if you want me to proofread stuff13:30
elodillesACK. ttx feel free to +2+W the patch (if looks fine) as it was prepared by me o:)13:30
elodillesttx: these will be the mails https://etherpad.opendev.org/p/relmgmt-weekly-emails13:30
ttxelodilles: ok will +a if you post another +2 on https://review.opendev.org/c/openstack/releases/+/93115313:31
elodillesttx: +2'd13:33
ttx+a'd13:35
elodilles++13:37
ttxemails look good, let's hold a bit before sending them.. maybe send around 14utc?13:38
ttxor 14:30utc13:38
elodillesboth works for me :)13:38
elodillesfirst i have to send the mail to openstack-announce then fungi approves it, then i can send the 2nd to openstack-discuss13:39
fungiyep13:40
fungisounds good13:40
fungii like that we've got a bit of breathing room this time, for a change13:40
fungiit's been a while since we didn't have some sort of last-moment scramble13:41
fungiso i wouldn't take it as an indication that we should start later next time13:41
elodillesyeah, it depends many things, how many patches are on the gate, pypi and other systems health, etc13:43
elodillesso we can stick to starting @ 10:00 UTC13:44
fungiyeah, i think that's worked out pretty well13:44
elodillesyes13:45
fungii'm going to take the opportunity for a quick shower, brb13:53
elodilles++13:54
opendevreviewMerged openstack/releases master: Mark 2024.2 Dalmatian as released  https://review.opendev.org/c/openstack/releases/+/93115313:58
elodillesmerged!13:58
elodilleslet's see when it will take on effect on releases.o.o13:59
Clark[m]The jobs are queued/waiting in zuul to publish that then it's on the 5 minute afs release timer I think14:09
elodilles++14:10
elodillesnow the publish-tox-docs-releases job has started14:18
fungiyep14:19
fungialso i'm back14:19
ttxback too14:21
elodillespublish job is still running14:23
elodillesfinished14:24
fungi14:25 vos release should have made it available on the site14:25
ttxIt's on14:25
ttxClear for takeoff14:25
fungiyep, dalmatian now shows as maintained14:25
elodillesyepp, 2024.2 Dalmatian 14:25
elodillesMaintained14:25
elodilles~o~14:26
ttxTime to send some emails!14:26
elodillesACK14:26
fungii'm ready to approve the one to openstack-announce when it shows up in the moderation queue14:26
elodillesfungi: e-mail is sent14:28
fungiand it's accepted for delivery to the list14:28
Clark[m]I received it14:29
elodilles\o/14:29
ttxI tooted it14:30
fungitoot toot!14:31
elodillesand there goes the other e-mail!14:31
ttxBoth received14:34
ttxCongrats everyoen on a smooth release day14:35
zigoIndeed! :P14:36
elodillesyes-yes, congrats everyone! \o/14:36
fungigreat work everyone!14:36
clarkbplenty of time to review the SLURP upgrade path diagram I pushed up :)14:42
clarkbI still don't think it is urgent but from discussions yesterday do think it would be helpful to have something liek that14:43
elodillesand there it is: https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_91c/931066/1/check/openstack-tox-docs/91c43bc/docs/#release-series14:44
ttxIt's "Dalmatian" not "Dalmation" but otherwise I like it14:47
clarkbttx: I'll get that fixed14:48
opendevreviewClark Boylan proposed openstack/releases master: Render a graph showing the SLURP and not SLURP upgrade paths  https://review.opendev.org/c/openstack/releases/+/93106614:49
zigoclarkb: You wrote A, C, B, with caracal before bobcat ...14:57
zigoOr am I missing something?14:58
clarkbzigo: correct. That affects the rendering if I listed them in alpha order then bobcat and dalmatian get rendered above the slurp path15:00
clarkbthe graph order is defined by the edges ( -> lines ) and it should rendering in the correct order15:00
fungiwhich would still be a valid diagram of it, just not as appealingly arranged15:00
clarkbthose names are also arbitrary I could've called them 1,2,3,4,5 and then labeled them appropriately I suppose15:01
fungiin dealing with the blockdiag/seqdiag issues, you've become our expert on sphinx's graphviz extension. now everyone will be sending diagramming requests to you directly ;)15:01
clarkbmy only protip is to use https://dreampuf.github.io/GraphvizOnline/ to test and figure things uot quickly15:02
clarkbits much quicker than rerunning tox over and over or figuring ou the right command line15:02
clarkbhttps://storage.gra.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_4c8/931066/2/check/openstack-tox-docs/4c8b284/docs/#releases-with-skip-level-upgrade-release-process-slurp is the updated preview15:15
zigoclarkb: *very* nice, thanks for writing this.15:16
zigoIt shows very clearly that this is fucked-up for Debian... :)15:16
clarkbunfortunately I don't think I can fix that with graphviz15:17
zigoYeah...15:17
zigoNever mind, my upgrade script works very well.15:17
zigo:)15:17
* zigo finished uploaded everything15:19
zigoI need to wait tomorrow to see the result on my Q/A page to see if I forgot a project or not.15:19
opendevreviewMerged openstack/releases master: Render a graph showing the SLURP and not SLURP upgrade paths  https://review.opendev.org/c/openstack/releases/+/93106615:56
fungi15:47 <opendevreview> Merged openstack/project-config master: Revert "Temporarily remove release docs semaphores"  https://review.opendev.org/c/openstack/project-config/+/93071016:04
elodillesthx o/16:10
clarkbhttps://releases.openstack.org/#releases-with-skip-level-upgrade-release-process-slurp hrm I haven't seen that before. Its like the doc build for publishing didn't try to render it at all16:22
clarkbtrying to dig up that job now to see if the artifact there does the same thing16:22
clarkbI see the bug16:25
JayFhttps://releases.openstack.org/#releases-with-skip-level-upgrade-release-process-slurp is also wrong w/r/t next stage for Dalmation; as a non-SLURP release it gets directly to EOL, not to UM16:27
JayFsimilar to how bobcat is rendered16:28
clarkbJayF: I'm not sure I understand16:29
JayFtalking about https://usercontent.irccloud-cdn.com/file/dt8YW5IN/image.png16:29
JayF(I should've dropped the anchor from the link)16:29
clarkbgot it that is unrelated to my change16:30
JayFI don't know enough about how that page is generated to have known that, but it makes sense :)16:31
* JayF sees if that's an obvious fix16:31
JayFit is, I am testing but will PR shortly16:33
clarkbif we land that fix after the fix for the graphviz thing we should be able to correct both with the same publication run16:33
clarkbI'll get a fix up for graphviz shortly16:33
JayFIt's literally a one line change, might be easier for you to pull it in16:33
JayFthan to dance with gerrit to combine 'em16:33
clarkbremote:   https://review.opendev.org/c/openstack/project-config/+/931204 Set bindep profiles for openstack release doc publication16:35
clarkbJayF: the issue is the fix for my issue is in project-config not releases16:35
clarkbalso a one line fix16:35
JayFaha16:35
clarkbalternatively I could drop the doc profile from the bindep file in releases but that is a less correct fix. We only need to install graphviz when building docs16:36
clarkbfungi: ^ fyi on 93120416:36
opendevreviewJay Faulkner proposed openstack/releases master: Next stage for Dalmation is EOL; not UM  https://review.opendev.org/c/openstack/releases/+/93120616:36
clarkbI'm also happy if we prefer to revert or just drop the profile in releases so that grpahviz is always installed. Let me know if we want a different appraoch16:40
JayFonly installing graphviz when/where it's needed is wise; it's a package that has had vulns in the past iirc16:43
clarkbfungi: repsonded to your comment but also I'm wondering if anyone else will review 931204 or if we should proceed?16:58
elodillesJayF: my bad, sorry :S thanks for the patch! +2 +PTL-Approved+1'd it17:00
JayFnot a problem, glad to be able to help even just a crumb :D 17:02
clarkbI need ot step away for a bit but I am in this room on matrix so I'll see if anything needs updating17:02
elodillesJayF: at least you noticed it quite fast, so we can fix it ASAP o:)17:04
elodillesttx: if you are around, please +2+W JayF's patch: https://review.opendev.org/c/openstack/releases/+/93120617:04
fungiclarkb: approved, thanks!17:09
elodillesyep, thanks for that, too o/17:13
clarkbhttps://review.opendev.org/c/openstack/project-config/+/931204 just merged so landing 931206 shoudl fix both issues17:25
clarkbI am in need of food so will pop out again. I'll check back to make sure the graph is resolved if 931206 or something like ti lands17:36
ttxApproved 20:29
opendevreviewMerged openstack/releases master: Next stage for Dalmation is EOL; not UM  https://review.opendev.org/c/openstack/releases/+/93120620:42

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!