Thursday, 2026-09-10

simondodsleygibi: Is there any chance that this patch (https://review.opendev.org/c/openstack/nova/+/973750) be relooked at again. It's a really simple fix for a big issue. Already have one +206:33
gibisimondodsley: I've approved it. I hope I did not risk RC1 with it. But I'm in a good mood so I taking my chances here. Btw next time feel free to ping me eariler if I reviewed someting before and never showed up re-reviewing the patch07:32
bauzasgibi: would you agree with gmaan's comment on https://review.opendev.org/c/openstack/nova/+/1003386 that we go for a known  bug reno instead of fixing it ?09:15
bauzasI'm cool with that and if you look at the irc logs we discussed a nice alternative09:15
gibilet me check09:53
gibiI'v read the review comments. I'm OK to keep as a known bug for now. I agree that it is not fatal thing to have this task leaked as it only slows down the shutdown, does not prevent it09:59
* Uggla has written about the nova-reviewers group situation --> https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/VDH5NVO6S4XDZDYEK32GZPT2444U2OJD/09:59
gibiUggla: thanks I will jump there 10:01
gibibut I also need to read the tc channel logs to get the full context 10:02
bauzasgiven the lack of consensus yet, I'd propose to defer the application of nova-reviewers until a consequent discussion can happen (before PTG or in it) for at least being all of up agreeing for find some policy how to use this new group10:42
sean-k-mooneyls12:23
sean-k-mooney:)12:23
sean-k-mooneygibi: gmaan: https://review.opendev.org/c/openstack/nova/+/1004687/1 and https://review.opendev.org/c/openstack/nova/+/966479/5 update nova next to ubuntu 26.04 and nova-alt-config to to debian 13, the latter was ment ot be done in 2026.1 so it woudl be nice if we can proceed with both this cycle, can ye take a look when ye have time12:25
*** erlon6 is now known as erlon12:43
gibiUggla: thanks for the email, I responded there and commented on the https://review.opendev.org/c/openstack/nova/+/986141 as well12:51
gibiwe need to start figure out a the way forward12:52
*** chandank` is now known as chandankumar`13:02
opendevreviewKamil Sambor proposed openstack/nova master: Fix pickling error that forces oslo.service to fall back to fork  https://review.opendev.org/c/openstack/nova/+/100095613:04
gablanhi gibi! I'm backporting the graceful shutdown work from master to stable/2026.1 downstream, and I hit the TODO you left in Host.cleanup() in `nova/virt/libvirt/host.py` about `stopping _event_handler` and `the _conn_event_thread` loop. do you have anything planned for it, or can I propose a patch for it? 13:06
gibigablan: check master I think gmaan recent changes there are addressing it13:17
gibisean-k-mooney: are you sure this is save to do on the day of RC1? If yes then we can land it but I need you to react and revert if something breaks on the gate after landing13:19
gibi/save/safe/13:19
gibicc gmaan, Uggla 13:19
sean-k-mooneyso i think so but im ok to wait until after rc1 is cut and backport if reqruied.13:20
sean-k-mooneyi coudl also flip the order13:20
sean-k-mooneythe debian 12->13 change for alt config13:20
sean-k-mooneyshoudl be safe because our ceph jobs have been on debina 13 all cycle13:20
sean-k-mooneywe just didnt get aroudn to movaing nova-alt-conig last cycel or eailar becasue i forgot the patch was still open/in merge conflict13:21
gibiwhen gmaan is up we can make the call I think13:21
gibiI will be around for longer today to accomodate that (and due to other unrelated reasons)13:22
sean-k-mooneyunless we have a strong reason not to i woudl perfer to have some py3.14 testign on stable 2026.2 but im ok if perpoel prefer not too13:22
sean-k-mooneyack works for me13:22
dansmithmy reasoning would be that we didn't go all of the cycle with that testing on, so flipping it right now seems much riskier than necessary13:39
bauzasis it me or the level of changes we plan to deliver so close to RC1 is high today ?14:02
bauzasas a reminder, we can make RC2 if there are some regressions that need to be fixed asap14:03
bauzasbut on the graceful shutdown case, I think we agreed yesterday to just amend the existing relnotes with a known bug file so people would know about the existing and decide whether to opt-in or not14:04
opendevreviewribaudr proposed openstack/nova master: Add Hibiscus prelude section  https://review.opendev.org/c/openstack/nova/+/100492714:05
Ugglabauzas, gibi just to let you know I have updated the highlight with latest input from gmann/bauzas https://review.opendev.org/c/openstack/releases/+/1004465 also reflect that change to the prelude patch https://review.opendev.org/c/openstack/nova/+/100492714:08
bauzasack14:09
Ugglaand sorry it comes later than I what I expected.14:09
dansmithbauzas: what else is proposed? we agreed to punt on the graceful shutdown one for sure14:10
bauzasdansmith: I maybe misunderstood the above discussion, I thought two different things were discussed, the one about graceful shutdown leaking threads and the other one being a nova-next change14:12
dansmithbauzas: oh yeah, the shutdown one is punted, and I agree, the nova-next change should wait.. I thought you meant something else14:13
*** erlon3 is now known as erlon14:14
gibiUggla: ack14:17
gibibauzas: this needs second code (max microversion) https://review.opendev.org/c/openstack/nova/+/100487214:37
gibis/code/core/14:37
bauzasoh I forgot that one14:37
gibiUggla: bauzas: what should we do with translation? https://review.opendev.org/c/openstack/nova/+/98054414:39
bauzaswe ship them by a single approver14:39
gibiOK then I approve it14:40
bauzasI did it14:40
gibiohh you did it good14:40
gibithanks14:40
bauzasnw14:40
bauzasand fwiw the i18n team works on a change from zanata14:40
bauzasI dunno where they are14:40
* gibi looks around definitely not here :)14:41
opendevreviewGabin Lanore proposed openstack/nova master: libvirt: stop the event dispatch loops on cleanup  https://review.opendev.org/c/openstack/nova/+/100502214:48
gablangibi: thanks! Checked both of gmaan's patches 1003386 is the task-registry leak, 1001572 is the executor ownership refactor. Neither seems to touch Host.cleanup(), so I pushed a patch for it :D14:50
gibigablan: thanks cc gmaan ^^ 14:51
opendevreviewClif Houck proposed openstack/nova stable/2025.1: perf(ironic): eliminate O(N²) ProviderTree deepcopy at startup  https://review.opendev.org/c/openstack/nova/+/100507315:27
opendevreviewGrzegorz Grasza proposed openstack/nova master: Mask password-like content in exception notifications  https://review.opendev.org/c/openstack/nova/+/100507415:48
opendevreviewMerged openstack/nova master: Mark 2.104 as maximum API version for 2026.2 Hibiscus  https://review.opendev.org/c/openstack/nova/+/100487216:09
opendevreviewMerged openstack/nova master: Imported Translations from Zanata  https://review.opendev.org/c/openstack/nova/+/98054416:09
gibiUggla: I've approved the prelude as the highlights laned. Are we done? :) 16:14
gibis/laned/landed/16:14
gibiahh we are not doen16:15
gibidone16:15
gibihttps://review.opendev.org/c/openstack/nova/+/1004430 and https://review.opendev.org/c/openstack/nova/+/1004424 are still need to land to have RC116:15
gibibut then every patch from the tracking etherpad are approved now16:16
gibiso we only need to wait for the gate16:16
gibisean-k-mooney: I guess the discussion about about the job changes went into the direction that we should not risk it now16:16
sean-k-mooneydid i miss that16:17
sean-k-mooneywere there comemtns on the patches16:17
sean-k-mooneybecause i tought we were landing the other way16:17
sean-k-mooneythat htis was low risk16:17
gibiI think bauzas and dansmith discussed above16:18
sean-k-mooneyah ok i was in a watcher irc meetign when we were chatting about it so i had swapped back to that16:18
sean-k-mooneythe debian change i feel really shoudl be done this cycel becase it was propsed 10 months ago and was defered lat cycel for similar reasons16:19
sean-k-mooneythe ubuntu 24.04->26.04 change for nova next is more a nice ot have16:19
sean-k-mooneyanyway i guess we can do them right after we cut rc116:20
zigoHi team !16:20
zigoI get it for the 2 nova patches above that needs to land, but anything preventing placement to be released ?16:20
sean-k-mooneyim not directly aware of anything pending on that side beyond perhaps a release not prelude16:21
sean-k-mooneybut im not sure that is strictly requried16:21
Ugglagibi, thanks I think we are done. We just need to wait the latest SHA to put it into the RC1 release16:21
sean-k-mooneywe didnt really have any placmeent feature so the normal release notes i sproably enough16:21
sean-k-mooneygibi: anyway so looking at the service version patchs they are now approved so i think thats all good16:22
sean-k-mooneybauzas: for context on the ubuntu change i proght it up at FF and i was asked to wait to RC1 which is why i put it up yesterday16:25
sean-k-mooneywe can wait again but i woudl like to get that testign in place early in 2027.116:26
sean-k-mooneythe base testing of nova on 26.04 has been runing in devstack for several months now which is why i was not expecting this to be contoversial16:27
sean-k-mooneyhttps://meetings.opendev.org/meetings/nova/2026/nova.2026-08-24-16.00.log.html#l-21716:29
gmaansean-k-mooney: gibi: same here, i would not prefer to touch the testing bit at this stage, let's have master rc-1 cut and then we can do those for 2027.116:30
gmaangibi: gablan thanks, will check16:30
gmaansean-k-mooney: for py3.14 testing on 2026.2 you mean unit/functional test or tempest job?16:30
sean-k-mooneyi ment moving nova-next to ubuntu 26.04 to test with 3.14 in tempest16:31
sean-k-mooneythat what i propsed a few weeks ago in the irc meeting16:31
sean-k-mooneyspecifcly for advance testing in 2026.216:31
sean-k-mooneygibi asked i do it after FFs and i asked to do it before RC116:32
sean-k-mooneyso i proposed the patches this week to do that16:32
gmaanyeah i know we discussed it on IRC to do it after FF but seeing gate unstability over last month, i agree to do it after rc-1. 16:33
sean-k-mooneyok anyway here is the devstack patch to make it voting there https://review.opendev.org/c/openstack/devstack/+/1004684?tab=change-view-tab-header-zuul-results-summary16:33
gmaanI am not against of adding it on 2020.2 but let's do it once stable/2026.2 exist so that it would not block current master dev/release16:33
sean-k-mooneyif you have not seen that as well16:33
sean-k-mooneywell once we cut RC1 master will be 2027.1 and it shoudl be voting there anyway in the next week or 2 as our default shoudl be chaning16:34
sean-k-mooneybut ya lets focus on getting RC1 out for nova and placement and come back to this next week16:35
gmaanyeah, we can backport that based on hoiw it run on stable/2026.216:36
sean-k-mooneyack16:36
gmaansean-k-mooney: but from testing runtime, i think we will keep 24.04 and 26.04 noth voting on 2027.116:36
sean-k-mooneyno16:36
gmaanhttps://governance.openstack.org/tc/reference/runtimes/2027.1.html#additional-testing-for-a-smooth-upgrade16:37
gmaanyes16:37
sean-k-mooneywe will keep 24.04 for upgrades16:37
sean-k-mooneybut 26.04 shoudl be voting and our new default16:37
gmaanbecause its 2027.1 where we are bumping the distro testing16:37
gmaanyes and at least one job for 24.0416:37
gmaanand voting16:37
sean-k-mooneyya so proably the skiplevel greande job16:38
sean-k-mooneywith the normal greade job using 26.0416:38
sean-k-mooneybut we can select one fo the others16:38
gmaanyeah, that only. so upgrade path from 2026.2 (on ubuntu 24.04) -> 2027.1 (ubuntu 26.04 ) is covered16:38
gmaanski level job is always there for SLURP and NON-SLURP by default unless project opt-out on non-SLURP release16:38
sean-k-mooneyyep i think the skip level job is often the best for that smoth upgrade path16:39
gmaanyeah16:39
sean-k-mooneyok i need to check that all the rc/stable branches for the cybrog and watcher repos that should exsit have been created so ping if anything comes up else ill be doing that for the next while16:40
sean-k-mooneygmaan: actully for the tempest plugins those will happen as a group correct?16:42
sean-k-mooneyi guess they are not typiclly branched16:43
sean-k-mooneyso that probaly fine but they will be tagged16:43
gmaansean-k-mooney: yeah, there will be only final release for tempest plugin and no stable branch16:43
sean-k-mooneyack perfect16:43
gmaanand jobs/ new master twick will happen usually at same time16:44
gibiUggla: sean-k-mooney: zigo: the placement RC1 release patch is not up to date with the tip of master16:44
gmaanfor devstack/greande branches, yes it wait for all projects stable exist and when i get release go ahead16:44
gmaan*release team go ahead16:44
sean-k-mooneygibi: ah is it missing the packagign fix form earlier in the week16:45
gmaans/i/QA PTL, hoping it will not be me all the time :)16:45
gibithe placement unreleased release notes looks OK to me I don't think we need a prelude there as it is sort16:45
gibiUggla: are you still around or should I go ahead and update the placement release patch?16:46
sean-k-mooneyoh that is there it just missing the tox changes16:47
sean-k-mooneythose are nice ot hvae if can bump them quickly16:47
gibiI can 16:47
gibiso I start doing it16:48
sean-k-mooneyi think its low risk and we might as well inculde them16:48
gibihttps://review.opendev.org/c/openstack/releases/+/1004525/1..216:52
gibiupdated16:52
gibibut we need Uggla's or bauzas' approval a release liasons16:52
sean-k-mooneywell the current version is not approved16:53
sean-k-mooneyadn the release team only merge them tomorow if late16:54
sean-k-mooneyso updatign now wont really impact there appoval16:54
sean-k-mooneyi commented on the patch that the current sha is ok since the tox chagne are nice to have16:54
sean-k-mooneywe can alwasy backport them if needed16:54
gibiI pushed the change :)16:55
sean-k-mooneyi guess the irc bot is not flexibel enough to tell use when nova related release patches are pushed16:56
gibiyeah that would be too much :)16:56
sean-k-mooneythat looks good to me16:58
opendevreviewGhanshyam Maan proposed openstack/nova master: Update release notes with known bug of live migration task leak  https://review.opendev.org/c/openstack/nova/+/100509317:09
opendevreviewGhanshyam Maan proposed openstack/nova master: Update release notes with known bug of live migration task leak  https://review.opendev.org/c/openstack/nova/+/100338617:10
*** mhen_ is now known as mhen17:11
opendevreviewGhanshyam Maan proposed openstack/nova master: Update release notes with known bug of live migration task leak  https://review.opendev.org/c/openstack/nova/+/100338617:11
gmaandansmith: gibi bauzas updated release notes to list that task tracking bug as known bug https://review.opendev.org/c/openstack/nova/+/100338617:12
gmaangablan: ^^ as you are backporting the graceful shutdown in downstream, this is known bug. 17:13
gmaangablan: btw, good to see interest in graceful shutdown feature, let us know how it works and especially any feedback on default value of graceful shutdown timeout configuration https://docs.openstack.org/nova/latest/admin/graceful-shutdown.html#configuration17:15
opendevreviewMerged openstack/nova master: Add Hibiscus prelude section  https://review.opendev.org/c/openstack/nova/+/100492717:28
dansmithgmaan: see my comment on the reno17:33
dansmithif you agree I can ninja the updated version, and if not, I'll +W17:33
gmaandansmith: checked, yes that is correct. let me update it17:35
opendevreviewGhanshyam Maan proposed openstack/nova master: Update release notes with known bug of live migration task leak  https://review.opendev.org/c/openstack/nova/+/100338617:37
gmaandansmith: ^^ done17:38
dansmithgot it17:38
opendevreviewMerged openstack/nova master: Update release notes with known bug of live migration task leak  https://review.opendev.org/c/openstack/nova/+/100338618:54
opendevreviewClif Houck proposed openstack/nova stable/2026.1: Parallelize per-node resource updates  https://review.opendev.org/c/openstack/nova/+/100511519:45
clifman I'm going to have to rework that backport patch again for 2025.2 since it seems like at that point in time futurist.GreenThreadPoolExecutor was used instead of nova.utils.create_executor19:49
JayFNo guardrail on the eventlet migration cliff :( don't fall (I know this is maybe unbelivable but pun not intended)19:53
clifI guess it wasn't as bad as I expected20:11
opendevreviewClif Houck proposed openstack/nova stable/2025.2: Parallelize per-node resource updates  https://review.opendev.org/c/openstack/nova/+/100511820:21
opendevreviewClif Houck proposed openstack/nova stable/2025.1: Parallelize per-node resource updates  https://review.opendev.org/c/openstack/nova/+/100511920:23
sean-k-mooneyclif: we have been slowly cenrtalisaing all the executor stuff release to release and putting it in a corner20:29
sean-k-mooneyso ya i woudl expect some rework to be needed there20:29
clifyeah I understand, the main hurdle was initially adapting to the change that happened recently with thread_pool_executor20:30
clifgoing backward has been relatively easy20:30
clifand the initial adaptation wasn't that bad in hindsight either, just unexpected20:31
sean-k-mooneyat least your doing it the "right way" release by release, its much much harder to skip im my expirnce20:35
sean-k-mooneyanyway im calling it a day o/20:36

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!