Tuesday, 2025-08-26

*** __ministry is now known as Guest2500502:05
opendevreviewGhanshyam proposed openstack/nova master: Add service role in Nova policy  https://review.opendev.org/c/openstack/nova/+/95757802:41
opendevreviewGhanshyam proposed openstack/nova master: Add service role in Nova policy  https://review.opendev.org/c/openstack/nova/+/95757803:11
gmaansean-k-mooney: ^^ updated service role change - https://review.opendev.org/q/topic:%22bp/policy-service-role-default%22+status:open03:12
gmaansean-k-mooney: I updated the tempest tests also. it seems neutron, cinder (for server external event and volume swap) still needs to assign admin role also to service user because nova need to perform the background task like get server/volume of other project which service user do not have access too. basically service user will have:03:14
gmaan- service role: for nova to check the policy if request is from service or not. ( API policy will be default to 'role:service' only (admin also for this cycle) )03:14
gmaan- admin role: to perform the actual operation03:14
gmaani will check gate result if anything failing but for safer side I am testing cinder, neutron, ironic also depends-no on nova change03:15
gmaancomplete set with testing changes are https://review.opendev.org/q/topic:%22bp/policy-service-role-default%22+status:open03:15
*** __ministry is now known as Guest2501405:44
opendevreviewKamil Sambor proposed openstack/nova master: Switch nova-conductor to use ThreadPoolExecutor  https://review.opendev.org/c/openstack/nova/+/95708809:19
ralonsohHi folks, quick question, if you know how to do it and if it is possible. Can Nova serve the metadata compressed? Context: https://bugs.launchpad.net/neutron/+bug/212072309:43
opendevreviewPavlo Shchelokovskyy proposed openstack/nova master: Fix the help for hw:cpu_policy flavor extra spec  https://review.opendev.org/c/openstack/nova/+/95852409:47
sean-k-mooneyralonsoh: not that im aware of, althoguh technically if your runing the wsgi applcaition under apache or similar it can add compressions10:51
sean-k-mooneyralonsoh: we do not have code to compress it in nova itslef10:51
sean-k-mooneybut apache or nginx coudl do it 10:51
sean-k-mooneyif it did that would be refected in headers in the reponce10:52
sean-k-mooneyif they are going via a loadblancer like haproxy that could also do compression tecnially btu i dont think that would be typical.10:53
sean-k-mooneyralonsoh: my advice wodl be do not treat it as text but a binary string10:54
ralonsohsean-k-mooney, yeah, the problem is the metadata proxy in neutron10:54
ralonsohwe are doing exactly that: treating the http content as text10:54
elodillessean-k-mooney: i've looked at the watcher release patch and i have a question there: https://review.opendev.org/c/openstack/releases/+/95846811:13
sean-k-mooneyelodilles: replied its technically a bugfix bump not a patch bump but i can elevate it to a Feature/Minor bump if you like11:19
sean-k-mooneythe external api behavior should remain effectivly the same11:19
sean-k-mooneyi added translation code to make swap effectivly be migrate so even if you request swap it will now use migrate and be semanticly the same effect11:20
sean-k-mooneyelodilles: but i dont have a stong prefernce here so if you would like me to respin i can do that quickly just let me know either way11:20
elodillessean-k-mooney: neither have i :)11:21
sean-k-mooneyshall i update to minor? it will take like 5 mins11:22
sean-k-mooneyotherwise ill go back to code reivew :)11:22
elodillessean-k-mooney: i've upgraded to +2 then, but feel free to edit if you changed your mind o:)11:24
sean-k-mooneywe have elevated other secuirty patches to feature bumps in the past os for visiblity im torn. bugfix patches everyone shoudl basiclly take sometime folks are less inclined to take minor bumps but i still have the change open in emacs so let me quickly edit it and ill update teh commit message too11:27
priteauHello. Sorry to bother you again, we're still looking for approval for these stable backports: https://review.opendev.org/c/openstack/nova/+/945098 and https://review.opendev.org/c/openstack/nova/+/94509911:29
sean-k-mooneyelodilles: updated11:33
sean-k-mooneypriteau: sure no worries i keep opening those and not finsihing revieing them ill do it now11:33
elodillessean-k-mooney: +2'd, thanks!11:37
sean-k-mooneypriteau: elodilles  i have +2w'd the 2024.2 version and +2'd the 2024.1 version of those so once the former merges elod can upgrade there +1 on the latter and merge them11:39
elodillesACK, i'll take an eye on it11:42
sean-k-mooneygeneral quetion for folks. how woudl people feel about have ing "meta traits" taht are reported if an only iff all the cpu feature flags for x86-64 version profile are met11:47
sean-k-mooneyhttps://en.wikipedia.org/w/index.php?title=X86-64&section=20#Microarchitecture_levels11:47
sean-k-mooneybasiclly the libvirt driver would addtionally report x86-64-v2 if all of the cpu feature flags requried for that were actully supported11:48
sean-k-mooneysame for v3 and v411:48
sean-k-mooneythat would mean if you had a rhel 10 iamge for example that can only run on v3 you coudl do a singel required trait x86-64-v311:49
sean-k-mooneyinstead of listing all of the v3 ones indeviually11:49
sean-k-mooneyyou could go one step futher and supprot defining meta traits in placement directly so you would not need to report them but coudl request them and it do the translation internally11:51
sean-k-mooneythat could be very useful eventually but i am not sure i want to go that far right now11:52
sean-k-mooneyhow that woudl work would be creatign a new api to define a meta traits as a union of other triaits (possible even a set of requried/forbiden triat instead of a plain union of required) and placement woudl just pre process the allocation candiate string and replace any meta taits with the expanded set11:53
sean-k-mooneythat woudl allow admins to defien there own but im not sure we need that level of flexablity11:54
opendevreviewMerged openstack/nova stable/2024.2: Reproduce bug/2098496  https://review.opendev.org/c/openstack/nova/+/94509812:12
opendevreviewStephen Finucane proposed openstack/nova master: api: Separate volume, snapshot and volume attachments  https://review.opendev.org/c/openstack/nova/+/95234712:55
opendevreviewStephen Finucane proposed openstack/nova master: tests: Use valid UUIDs for cinder resources  https://review.opendev.org/c/openstack/nova/+/95293512:55
opendevreviewStephen Finucane proposed openstack/nova master: api: Only apply "soft" additionalProperties validation to requests  https://review.opendev.org/c/openstack/nova/+/95293612:55
stephenfinsean-k-mooney: Those functional tests are fixed now, if you're game to re-apply your +2/+W? ^13:00
sean-k-mooneydone ill let gmaan hit https://review.opendev.org/c/openstack/nova/+/952347/9 but the next 2-3?  have +2w13:02
opendevreviewFlorian proposed openstack/nova master: Add check for PCIe devices attach limit for volume and ports  https://review.opendev.org/c/openstack/nova/+/95558413:02
sean-k-mooneyas we dicussed offline once that batch is in we can see if you need to rebase the rest or if they will just apply13:03
opendevreviewFlorian proposed openstack/nova master: Add check for PCIe devices attach limit for volume and ports  https://review.opendev.org/c/openstack/nova/+/95558413:57
opendevreviewFlorian proposed openstack/nova master: Add check for PCIe devices attach limit for volume and ports  https://review.opendev.org/c/openstack/nova/+/95558414:12
bauzasdansmith: can you just +2/+W that very old Flamingo patch that modifies the support envelope for our non-SLURP release ? https://review.opendev.org/c/openstack/nova/+/94401814:30
bauzasthis is too way overdue14:30
dansmithbauzas: I thought we were sticking with two releases always to make sure we always support skip-level?14:32
bauzasnope, we always set the minimum object version to N-1 if we are on a non-SLURP as we don't support skip-level for ops14:33
dansmith...but we run grenate-skip-level-always14:33
dansmithI'm not sure why that didn't fail though (logs are gone)14:33
bauzasindeed, but that's again not our support envelope14:33
bauzasskip-level changes the option and bypasses that value then14:33
dansmithdoes it?14:33
bauzas(IIRC)14:34
bauzasyeah, IIRC we have a flag that says "I gonna skip the min version"14:34
bauzasfor something we called FFU way long before14:34
bauzasthe beast is gone but the flag remains14:34
dansmithif so I don't know where that is14:34
bauzashttps://review.opendev.org/c/openstack/nova/+/913890 was the one we merged for Dalmatian14:35
bauzasNOVA_ENABLE_UPGRADE_WORKAROUND is the tempest flag if I'm not mistaken14:36
dansmithyeah, I see, but .. I'm not sure that's right14:36
bauzashttps://opendev.org/openstack/grenade/src/branch/master/projects/60_nova/upgrade.sh#L7714:38
dansmithI'm not sure skip-level-always actually does a multinode and leaves a node behind.. from the code I don't see how this would not break a skip level14:38
bauzasthat's the thing we did from Antelope and older14:38
bauzasskip-level was turning the object version check off14:38
bauzasbut we always said that our support envelope was N-1 14:39
bauzasfor non-SLURP14:39
dansmithokay, I see yeah14:39
bauzasI even wrote that big fat comment in the object 14:39
bauzassome day around early october after we branch RC1, we would keep the value of that object14:40
bauzasas we would continue to support Epoxy14:40
dansmithyes there's a big comment, but it doesn't cover the skip-level detail14:40
bauzaseven if master is Gazpatcho-based14:40
dansmithwhile I was grepping through the tree, my shell was reminding me that I have run some of the same greps in the last year, so clearly I was doing the same thing last time ;)14:41
bauzasI can add some extra paragraph that would mention that that value is not read by skip-level14:41
bauzasas skip-level turns on [workarounds]/disable_compute_service_check_for_ffu14:41
bauzaslemme then amend my patch and add that comment to prevent another brain cycle next time :)14:42
dansmithyeah just mentioning the config option might be enough14:43
opendevreviewSylvain Bauza proposed openstack/nova master: Update min support for Flamingo  https://review.opendev.org/c/openstack/nova/+/94401814:52
bauzasdansmith: can you just reapply your +W ? 14:53
dansmithdone14:54
bauzasthanks14:55
dansmithgibi: I don't have anything proposed but unmerged for precache15:00
gibiack15:00
gibiI will test a bit manually the eventlet removal change then15:00
dansmith_probably_ because we can't externally validate it, we could only make sure it didn't fail the call or something15:01
sean-k-mooneyi was talking to gmann about this at one point15:01
gibidansmith: ahh good point. Maybe whitebox then...15:01
sean-k-mooneyabout maybe adding a whitebox test to ssh in and assert the base image is created after caching15:01
dansmithyeah15:01
sean-k-mooneyi dont know if they wrote that test yest hwoever but let me look15:01
gibiI don't see anything in whitebox either 15:02
sean-k-mooneyno i don tthink they  started on it15:03
sean-k-mooneywe were chating about it in the context of dicussiong what ci coverage we had in general when they joined the team15:03
gibiOK. manual testing then... I anyhow wanted to test two different caching running in paralle15:03
dansmiththat might be hard unless you have a large image or artificially slow it down15:04
sean-k-mooneyit would be nice to add a test for it to whitebox eventually15:04
gibisean-k-mooney: I agree15:04
gibidansmith: yepp I will do one or the other :)15:04
sean-k-mooneyso gmaan recently added a test for booting form an iso to validate that https://review.opendev.org/c/openstack/whitebox-tempest-plugin/+/955950/4/whitebox_tempest_plugin/api/compute/test_iso_image.py15:05
sean-k-mooneyso we already have a custom image file  in ci we could use 15:05
sean-k-mooneyi.e. we can create a new imag e in the test in clance using the iso 15:06
sean-k-mooneyand the pre cache that to an aggreate15:06
gibicool15:10
gmaangibi: sean-k-mooney this is where I added Core-14.0.iso image in job and devstack download it and tempest plugin can play with that image from file locattion - https://github.com/openstack/whitebox-tempest-plugin/blob/master/.zuul.yaml#L72C206-L72C21915:11
gibithanks15:13
sean-k-mooneywe could also have used any of the sandard images like cirros but since whitebox now expect an iso at least to run that test we might as well use it esplciy since we are intenallysuing tiny core which as the name implies is small15:14
sean-k-mooneyi want to revive the idea of running a subset of whitebox in nova's check jobs at somepoint by the way15:15
gmaanyeah, that is not too large but we can add larger image here15:15
sean-k-mooneywe can skip all the ahrdware specfic stuff but obvioulsy but this and the iso but check woudl be nice to have15:15
gmaanonly tricky part of ssh into server booted from iso image. But i do not think pre-cache image test need that15:15
sean-k-mooneywe are also not doing that in whitebox today for the iso case15:16
gmaansean-k-mooney: good idea, I can add that or maybe we can run a separate peroidic job as full whitebox plugin15:16
gmaanI mean running iso image test in nova15:16
sean-k-mooneyi spoke to melwitt breifly about changing nova-ovs-hybird plug in to nova-alt-config adn having it test more things15:17
sean-k-mooneyso rather then adding a new job or continuing to extend nova-next i woudl put it in that job15:17
sean-k-mooneywe can chat about that in more detail after FF ro at teh ptg whenever we have tiem to plan that out15:18
sean-k-mooneybut i was chatting to her about enableing nova on nfs and cinder nfs in that job15:18
sean-k-mooneyso we have coverage of both goign forward15:18
sean-k-mooneyso nova-alt-config woud test spice, ml2/ovs, nfs and whitebox things15:19
sean-k-mooneyon debian 1315:19
sean-k-mooneyso it will also test python 3.1315:20
sean-k-mooneythat is how i woudl like to evolve that job in anyrate15:20
gmaansure15:26
sean-k-mooneygmaan: when you have time would you mind reappoving https://review.opendev.org/c/openstack/nova/+/95234715:32
gmaansean-k-mooney: yeah, I opened it, will do after meeting15:32
sean-k-mooneyack thanks15:32
UgglaNova meeting in ~30mn15:35
gmaanUggla: openinfra board finance committee at same time. so I might not be or might be late in today meeting15:44
Ugglagmaan ok, just a question about https://review.opendev.org/c/openstack/nova/+/957578 do you think it is something that could land before FF ?15:46
gmaanUggla: I hope so, we are very close to that. I am testing it (forgot to recheck my testing change yesterday). I think sean-k-mooney is planning to review it (once I ping that it is working)15:47
UgglaIt will be great !15:48
opendevreviewStephen Finucane proposed openstack/nova-specs master: Fix URL for v2.0 API deprecation spec  https://review.opendev.org/c/openstack/nova-specs/+/95855215:48
jssfrUggla, can I queue up https://review.opendev.org/c/openstack/nova/+/955657 and https://review.opendev.org/c/openstack/nova/+/621646 for the "further discussion" section?15:53
Ugglajssfr, yes I think so. 15:54
jssfrIn case I don't make it (it's been a busy day... and I might get pulled away): the vTPM patch has been updated and is awaiting review. Would be nice to get it into 2025.2.15:54
jssfrthe UEFI patch has *not* been updated, but we'd like to hear whether it's realistic to get it into 2025.2 *if* we were to update it this week and if so, until when we would have to be done.15:54
Ugglajssfr, I think it is a bug, so there is less constraints, bugs are not fully concerned by FF.15:55
jssfroh, that's good to know15:56
Uggla#startmeeting nova16:00
opendevmeetMeeting started Tue Aug 26 16:00:31 2025 UTC and is due to finish in 60 minutes.  The chair is Uggla. Information about MeetBot at http://wiki.debian.org/MeetBot.16:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.16:00
opendevmeetThe meeting name has been set to 'nova'16:00
elodilleso/16:00
dansmitho/16:00
UgglaHello everyone16:00
gibio/16:02
UgglaLet's start.16:02
Uggla#topic Bugs (stuck/critical) 16:02
Uggla#info No Critical bug16:02
Uggla#topic Gate status 16:02
Uggla#link https://bugs.launchpad.net/nova/+bugs?field.tag=gate-failure Nova gate bugs 16:02
Uggla#link https://etherpad.opendev.org/p/nova-ci-failures-minimal16:03
tkajinamo/16:03
Uggla#link https://zuul.openstack.org/builds?project=openstack%2Fnova&project=openstack%2Fplacement&branch=stable%2F*&branch=master&pipeline=periodic-weekly&skip=0 Nova&Placement periodic jobs status16:03
Uggla#info Please look at the gate failures and file a bug report with the gate-failure tag.16:03
Uggla#info Please try to provide a meaningful comment when you recheck16:03
bauzas o/16:03
Ugglaskipping next point as gmaan, is in another meeting.16:04
Ugglatopic Release Planning 16:04
Uggla#link https://releases.openstack.org/flamingo/schedule.html16:04
Uggla#info Nova deadlines are set in the above schedule16:04
Uggla#info Feature freeze is Thursday.16:04
Uggla#topic Review priorities16:05
Uggla#link https://etherpad.opendev.org/p/nova-2025.2-status16:05
gibiUggla: Am I correct that FF is on Thurday EOD, whathever apporoved can still rebased / rechecked after to land? Do we plan any FF exception process?16:05
UgglaFirst I need to thank bauzas, who helped me to better highlight status in the board and sync correctly within launchpad. 16:05
bauzaswell, I should have given you some help way before16:06
Ugglagibi, I think so. I don't think exception will be needed but we might discuss it here.16:07
UgglaBefore that here are the review priorities we identified:16:07
Uggla#link Amd Sev serie (tkajinam) need review https://review.opendev.org/q/topic:%22bp/amd-sev-es-libvirt-support%2216:07
Uggla#link https://blueprints.launchpad.net/nova/+spec/copy-applied-provider-yaml https://review.opendev.org/c/openstack/nova/+/95757816:07
Uggla#link https://blueprints.launchpad.net/nova/+spec/libvirt-migrate-parallel https://review.opendev.org/c/openstack/nova/+/95066716:08
Uggla#link https://blueprints.launchpad.net/nova/+spec/policy-service-role-default one patch to review https://review.opendev.org/c/openstack/nova/+/95757816:08
UgglaI discussed latest one with gmaan and that might land if we review it. (only one patch)16:09
Ugglabauzas and I are currently reviewing Amd sev series.16:10
sean-k-mooneyo/16:10
gibiUggla: I'm currently try to squeez in https://review.opendev.org/c/openstack/nova/+/957088 before FF16:11
gibithe rest of the eventlet series can wait post FF / G16:11
bauzasmy only concern is that AMD SEV-ES series requires a reshape16:11
Ugglagibi, thanks that was one of the question I had.16:12
dansmithgibi: you said you were fine with the content on that one, but...no unit tests?16:12
bauzasthis requires a lot of caution16:12
dansmithare those coming?16:12
sean-k-mooneyUggla: techihnaly  teh parla clodu migration chnag eis for 2026.116:12
bauzasthe what ?16:12
sean-k-mooneyor did we approve it late for this cycle16:12
gibidansmith: it is a refactor like the scatter gather change16:12
gibidansmith: I don't see what extra unit test we need16:13
bauzaswell, yes and no16:14
dansmithokay I guess I thought something has to change with those mechanics, but I'll go look at what is testing this to see16:14
bauzasthere is a refactor but there is also a slight behavioral change16:14
dansmiththe lock obviously isn't verified, but if it's being run over then maybe that's okay16:15
gibithe change is the lock and I want to verify that locally. If you need a functional test to ensure that lock is valid then sure we can do that too16:17
gibiI don't see any other change in logic16:17
bauzasI guess having such an invasive change without any single line of nova/tests made me afraid and I missed the point of the refactor16:18
gibiinvasive? you mean a lock?16:19
gibiI'm confused16:19
gibiI thought the goal of the refactor is not to change behavior so if the existing test does not need to be touched that is actaully good16:19
gibias it proves no behavior change happened16:20
dansmithgibi: I guess I expected to see at least something running this code in both modes, but perhaps I'm jumping the gun there. I know we don't have that for everything we're thread-ifying, but this is also sort of a self-contained example16:21
sean-k-mooneyso it would be nice to have the zuul change to enable it in the condutore in the patch16:21
bauzasgibi: I agree with the fact that until some env var is set, nothing changes16:21
sean-k-mooneybut i dont consider that an invasive change either for what its worth16:22
gibidansmith: yeah we did not transformed the unit test to both modes as part of the changes but we did it separately. 16:22
bauzasbut yeah, I'd have appreciated some guideline to temper my guts16:22
sean-k-mooneyits adding a lock that shoudl techenically have been there even under eventlet + swapting the executor we use16:22
dansmithunlike general thread spawning, this is a self-contained create/destroy of a single executor pool, which means we could test both modes here and now which is kinda what I was expecting16:22
dansmithbut that's fine16:23
gibidansmith: I will check my unit test enalbing series if it enables the related tests already...16:23
gibiwith threading in a separate tox target16:23
sean-k-mooneywhat was that job called by the way16:24
sean-k-mooneyi tough we already enabled that in ci16:24
sean-k-mooneyis that still pending?16:24
gibinova-tox-py312-threading16:24
gibiit is pending16:25
gibihttps://review.opendev.org/c/openstack/nova/+/955791/916:25
sean-k-mooneyah ok16:25
sean-k-mooneysorry i tought that was earlier in the series16:25
sean-k-mooneyhttps://review.opendev.org/c/openstack/nova/+/953475/20 that one right16:25
Ugglaassuming gibi is working on the eventlet refactor any chance to land something in the above list ? Can we define some priorities/owners about them ?16:25
sean-k-mooneythere are a number of the ci/testing changes that woudl be nice to also land16:26
gibidansmith: I see multiple image cache unit test passing in https://4d426b9917a9f7962308-d3c6167007a6d8477c4cec901604a1c7.ssl.cf1.rackcdn.com/openstack/4beb6e364cb74aaeb2b8ff0b475d966f/testr_results.html with threading. So if we move the refactor under the unit test series we can verify the change 16:26
sean-k-mooneybut i wonder if we woudl be oke with some of those happening post FF16:26
dansmithgibi: ack16:27
gibisean-k-mooney: https://review.opendev.org/c/openstack/nova/+/953475/20 adds the job, but later patches enables more and more tests16:27
sean-k-mooneythe 6 patches including and up to https://review.opendev.org/c/openstack/nova/+/955791/9 are strech goals for me16:27
gibiyeah the unit test enablement was planned post FF as streach for Flamingo16:27
gibibut if we block the nova-conductor refactor on it then I will at least make an arrangement to have test results16:28
gibiI mean unit test results with image cache tests running in threading mode16:28
gibido you want me to do that?16:28
bauzasI agree with the statement that I feel that the patches above n-api seem more important for now16:29
opendevreviewMax proposed openstack/nova master: db: add indexes for source/dest migrations  https://review.opendev.org/c/openstack/nova/+/95855616:29
gibibauzas: so you say lets prioritize the unit test enablement over the nova-conductor work.16:30
gibi?16:30
gibi(I have no problem with it if that is what the team wants)16:30
bauzasfor me, at least, I'd prefer reviewing those patches16:31
sean-k-mooneyi was going to suggest rebasign the 6 packages on top of the conductor one16:31
bauzasand I think those would be more beneficial for the rest of the work we have to do16:31
bauzassean-k-mooney: you mean, on the bottom ?16:32
sean-k-mooneyno16:32
gibisean-k-mooney: yeah that what I meant by arranging to get test results16:32
sean-k-mooneybecause i disagree that havign the ablity to run the condcutor in treading mdoe is less valuable16:32
gibisambork: Are you around? ^^ :)16:33
sean-k-mooneyif we have the conductor patch we can test it and ask operator to test it and report bugs16:33
bauzasif wre're not able to run it in CI, what's the benefit ?16:33
sean-k-mooneywe can trivially16:33
sean-k-mooneyits liktrally a 1 line change to the zuul job16:33
sean-k-mooneyto set the var in devstack to enabel threading mode16:33
bauzasbut we won't execute that code, right?16:33
gibisean-k-mooney: I think what bauzas imply is that we don't have tempest coverage for the changed code path16:33
sean-k-mooneyok that should nto block the code change16:34
bauzasI mean, the zuul config change will be meaningless16:34
bauzasif we don't have zuul jobs that call image caching16:34
sean-k-mooneyits wont be it makes sure everythign else in the conductor works16:34
sean-k-mooneyi.e. that we are not implcitly relying on eventlet indrectly16:34
bauzasthat is a good point16:34
bauzaswe could have missed other eventlet dependencies16:35
bauzasI then agree, 16:35
sean-k-mooneyby the way i think we could land both if we are willing to merge the unit test change up to RC116:35
sean-k-mooneymy suggesting is for gibi to rebase them so we get more results today/tomorrow16:35
bauzasagreed, I like that approach16:36
sean-k-mooneyand then we can decied if we are comfrotabel with merging the conductor changes16:36
gibiI will do the rebase today to get unit test results with threadin on image cache16:36
bauzasthanks gibi16:36
gibisambork: ^^ FYI16:36
sean-k-mooneygibi: can we add the devstack job change too if you or sambork have time16:36
gibiI will do local testing of the lock explicitly16:36
gibisean-k-mooney: I will do so 16:36
sean-k-mooney+116:36
gibiI have nothing further16:37
Ugglaasking again, assuming gibi is working on the eventlet refactor any chance to land something in the above list ? Can we define some priorities/owners about them ?16:38
bauzasabout the other things to review, I want to speak about SEV-ES16:38
bauzasbut tkajinam is not here I guess (and hope)16:38
tkajinamI am16:38
bauzaswow16:38
bauzasokayu16:39
tkajinam:-)16:39
sean-k-mooneystephenfin: and i both were interested in the sev one as well16:39
sean-k-mooneyor at least started looking at ti 16:39
dansmithUggla: land...in the next two days?16:39
sean-k-mooneyso i think taht woudl be good to try and land16:39
bauzasmy concern is that we're on a non-SLURP release16:40
sean-k-mooneybut the main problem with that serise is reviewing the reshape16:40
bauzastkajinam: I assume you know that you'll have to continue to handle the reshape case for at least two releases, right?16:40
Uggladansmith yes if possible. 16:40
Ugglaor at least try.16:40
tkajinambauzas, yes16:41
dansmiththat seems like a big ask to me,16:41
tkajinamI agree with the concern about reshape but I'm hoping that I at least get any feedback to hear how far it is from merge.16:41
dansmithbut I'm not really positioned to review a reshape16:41
tkajinamwe can discuss if that can be a subject to FFE or I'm even ok if that is eventually postponed to 2026.116:42
bauzasI feel reasonable being able to review a reshape since I wrote two of them16:42
tkajinammy main concern is that we've been saying the same for recent 2~3 cycles (if that's too much for this cycle and then put it to next)16:42
bauzasbut that's a very old knowledge part of myself I need to go with16:42
sean-k-mooneybauzas: that would help because the rest of the serise is much more straght forward16:42
bauzassean-k-mooney: I can help with providing pointers to how the reshape mechanism works16:43
bauzasfrom the very first glance I made on tkajinam's series, he perfectly managed it16:43
bauzasI mean, he's triggerring the signal, and then using the signal to reshape16:43
sean-k-mooneywell thats encuraging at least. 16:43
tkajinamI added functional tests coverage according to the existing example so I hope that helps review... if any additional scenario may be needed I can try adding it16:43
bauzastkajinam: do you have any excerpts of a real environment upgrade scenario that you tested locally ?16:44
bauzasI mean, pastebins16:44
bauzasthat would ease the review16:44
sean-k-mooneyyou basiclly mean deploy devstack with out the patch and boot an sev instance then apply the patch adn restart nova right16:45
sean-k-mooneyto see it properly migrate the exsting allcoations16:45
bauzasthat yes16:45
bauzasthat's how I tested VGPU reshapes16:45
tkajinamhmm I'll check but I'm afraid I didn't keep the test result logs, and it might take some time to spin up the env again (due to machine availability)16:46
bauzastkajinam: but I assume you tested such upgrade scenario, right?16:47
sean-k-mooneyas an asided there is an exsting bug that at least on rhel prevent sev form wroking. i aslo dont know if our internal amd host is free to do any testing16:47
bauzaslike, you had an env, you patched it, you saw the allocations moved16:47
tkajinamyes, when I wrote the series initially.16:48
bauzasexcellent16:48
bauzasand yeah, functional tests can prove it works16:48
bauzasso if you already wrote them (I haven't reviewed them yet), that's excellent16:48
bauzastkajinam: I assume you found how to do it using the existing reshape methods?16:48
tkajinamyeah16:49
tkajinamhttps://review.opendev.org/c/openstack/nova/+/921814/32/nova/tests/functional/libvirt/test_reshape.py#23316:49
bauzasperfecty16:49
tkajinamI followed the existing rehape test for mdev, IIRC16:49
bauzasOK, I'll defineitely commit myself to the review then16:49
Ugglaok shall we move on ?16:51
bauzasyup, we'll see what we can do for the other bits16:51
bauzaslike every cycle16:52
Ugglasure, what we have today is already cool.16:53
Ugglaso moving on.16:53
Uggla#topic OpenAPI 16:53
Uggla#link: https://review.opendev.org/q/topic:%22openapi%22+(project:openstack/nova+OR+project:openstack/placement)+-status:merged+-status:abandoned16:53
Uggla#info still 32 remaining atm.16:53
Ugglaso this will probably not fully land in this cycle, but we made good progress.16:54
Uggla#topic Stable Branches16:54
Ugglaelodilles, please go ahead.16:54
elodilles#info stable branches (stable/2025.1 and stable/2024.*) seem to be in OK state16:55
elodilles#info nova stable versions released: 29.3.0 (2024.1 Caracal); 30.1.0 (2024.2 Dalmatian); 31.1.0 (2025.1 Epoxy)16:55
elodilleswith the sec related bug fix from last week16:55
elodillesand that's all from me for now, back to you Uggla 16:55
Ugglathx elodilles16:55
elodillesnp16:55
Ugglaskipping next point due to fwiesel pto's16:56
UgglaI also guess we can skip gibi's point as I think he already covered previously.16:57
gibiyepp nothing further for FF. I have things for post FF but we can disucss that next week16:57
UgglaYep we are at the top of hour. I know jssfr wanted to discuss, but I think we can do that next week.16:58
jssfrdid the UEFI NVRAM preservation thing also count as bugfix?16:59
sean-k-mooneyyes16:59
jssfrokay, thanks16:59
sean-k-mooneyits not merged but we said we woudl treate the nvram and tpm data the same both as bug fixes16:59
jssfr(even if not, it looks like y'all are busy enough getting all the other features over the line)16:59
jssfrthanks, then we can certainly defer to next week from my side16:59
Ugglajssfr thx.17:00
UgglaSo I'm gonna close, thanks all.17:00
Uggla#endmeeting17:00
opendevmeetMeeting ended Tue Aug 26 17:00:17 2025 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)17:00
opendevmeetMinutes:        https://meetings.opendev.org/meetings/nova/2025/nova.2025-08-26-16.00.html17:00
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/nova/2025/nova.2025-08-26-16.00.txt17:00
opendevmeetLog:            https://meetings.opendev.org/meetings/nova/2025/nova.2025-08-26-16.00.log.html17:00
sean-k-mooneyUggla: as an fyi i think we shoudl be able to merge at least the next 3 patches in the openapi seriese and possibly some more before FF17:00
sean-k-mooneyim also quitely hopeful we could land gmaan's service role change17:01
sean-k-mooneyi loooked at it today and after some back and forth on scope it looks good17:01
Ugglayep me too, at least gmaan was quite confident about it17:01
bauzastkajinam: still here ?17:01
tkajinamyes17:01
sean-k-mooneygmaan has 3 DNM change that are running in ci agains neturon cinser and ironic that i wasnt to review before upgrading to a +217:02
Ugglasean-k-mooney, it would be cool for the openapi too, even probably less a priority17:02
sean-k-mooneyi would really like to close that out htis cycle im just not sure how comnfortable people might be with a FFE for it.17:02
sean-k-mooneyi am plannign to try and compelte the review of the open patches but i have  quite a few to go17:03
bauzastkajinam: do you understand that you could have a cloud with old computes and then placement reflecting two hosts with different inventories ?17:03
sean-k-mooneyUggla: if it missies i woudl liek to keep the momentum up and continue to review/merge it after RC1 is cut17:03
Ugglasean-k-mooney +117:04
bauzastkajinam: nothing should break, since old computes fulfilling AMD SEV requests would continue to be able to allocate them17:04
bauzasnew computes would be able to do that as well17:04
tkajinambauzas, you mean that a single cluster might have old computes reporting SEV in old structure while new ones reporting new structure17:05
bauzasyup17:05
sean-k-mooneywe allow requests in the root group to be fullfiled by any RP in the tree right17:06
tkajinambauzas, yeah I understand that point.17:07
sean-k-mooneyso it shoudl be ok to come form a sub RP? or do we need to split it out for that to work?17:07
bauzasI hope, but 17:07
sean-k-mooneyi know we talked about that before i dont recall what it does today17:07
sean-k-mooneyif not we woudl ahve ot transfrom the SEV request ito a sepreate group17:07
sean-k-mooneyto suppro the mixed case17:08
bauzastkajinam: do you feel reasonably happy if I ask you amend your functional tests in https://review.opendev.org/c/openstack/nova/+/921814/32/nova/tests/functional/libvirt/test_reshape.py and provide a mixed environment ?17:08
sean-k-mooneybut i dont think that is required.17:08
bauzasI mean, ideally I'd appreciate a test that would add a mixed scenario, ie. two computes with one being upgraded while the other one is not17:09
sean-k-mooneythis is something that an upgraded schduler is capable of doing however since we dont sue resouce: directly17:09
bauzastkajinam: do you think you could write such functional test quickly?17:09
bauzasI just want to prevent any placement behavioural change I could have missed :)17:10
bauzasand we would also ensure we don't regress, particularly if someone wants to change how placement is fulfilling requests17:11
tkajinambauzas, the current functional tests covers instance creation before reshape and creation after reshape using a single RP, so that proves sev instance can be scheduled to old compute node and also new compute node17:13
bauzasit doesn't prove that instance creation can be fulfiled on a mixed env, that's the point17:13
bauzasoh, I see what you mean17:14
tkajinamI can create a scenario with two RPs, but do you expect that we create two servers in that case but I'm not sure if I can enforce these two scheduled to two different (old/new) compute nodes17:14
bauzastkajinam: that is doable, we have other functional tests where we fake the targets17:15
bauzasbut anyway, I see your point17:15
bauzasshall work anyway, so here I think a follow-up patch is only needed17:16
tkajinamok17:17
tkajinamI'll push follow-up to run similar scenario with two mixed nodes, but I believe that compatibility scenario is already covered by the existing scenario using one node.17:18
bauzasthanks, appreciated17:19
opendevreviewMerged openstack/nova stable/2024.2: Ignore metadata tags in pci/stats _find_pool logic  https://review.opendev.org/c/openstack/nova/+/94509917:41
opendevreviewTakashi Kajinami proposed openstack/nova master: Add functional test scenario for mixed SEV RPs  https://review.opendev.org/c/openstack/nova/+/95856217:53
tkajinambauzas, I hope this matches your expectation but lmk if you want additional steps (like creating servers before reshape)17:53
bauzastkajinam: <317:54
tkajinamI put it on the last one to avoid mass rebase but I can change the order... I noticed that I didn't add that comment to the initial functional test scenario so I'll move that part to the reshape patch if I have to respin17:54
bauzastkajinam: putting it on top is fine by me17:54
bauzasanyway, I just need to review your series for now17:55
tkajinam(the comment I'm referring to is https://review.opendev.org/c/openstack/nova/+/958562/1/nova/tests/functional/libvirt/test_reshape.py#259 )17:55
tkajinambauzas, thanks !!17:55
bauzastkajinam: one thing, we keep providing capabilities as traits to the root RP, not the children ones, right?17:59
bauzasoh nevermind, no17:59
tkajinambauzas, no we have to migrate the trait to its children, so that we can have SEV children tree as well as SEV-ES children tree18:01
tkajinamMEM_ENCRYPTION_CONTEXT for SEV-ES can't be used for SEV so we have to define the trait at children layer18:01
bauzasyup, and your functional traits test covers that, so I understood my mistake18:02
tkajinam:-)18:05
tkajinamI'm leaving soon but will check the irc logs as well as any review comments tomorrow18:05
opendevreviewBalazs Gibizer proposed openstack/nova master: Ask for pre-prod testing for native threading  https://review.opendev.org/c/openstack/nova/+/95742419:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Switch nova-conductor to use ThreadPoolExecutor  https://review.opendev.org/c/openstack/nova/+/95708819:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Allow to start unit test without eventlet  https://review.opendev.org/c/openstack/nova/+/95343619:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Run unit test with threading mode  https://review.opendev.org/c/openstack/nova/+/95347519:14
opendevreviewBalazs Gibizer proposed openstack/nova master: [test]RPC using threading or eventlet selectively  https://review.opendev.org/c/openstack/nova/+/95381519:14
opendevreviewBalazs Gibizer proposed openstack/nova master: [CI]Make nova-tox-py312-threading voting  https://review.opendev.org/c/openstack/nova/+/95579119:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Do not yield in threading mode  https://review.opendev.org/c/openstack/nova/+/95099419:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Make RBD Tpool usage conditional  https://review.opendev.org/c/openstack/nova/+/95608919:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Make libvirt Tpool proxying conditional  https://review.opendev.org/c/openstack/nova/+/95609019:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Fix ProviderTree copying with threading Lock  https://review.opendev.org/c/openstack/nova/+/95609119:14
opendevreviewBalazs Gibizer proposed openstack/nova master: [test]Further categorization of disabled unit tests  https://review.opendev.org/c/openstack/nova/+/95609219:14
opendevreviewBalazs Gibizer proposed openstack/nova master: [vncproxy]Handle ssl.wrap_socket removal in py312  https://review.opendev.org/c/openstack/nova/+/95591519:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Warn on long task wait time for executor  https://review.opendev.org/c/openstack/nova/+/95266619:14
opendevreviewBalazs Gibizer proposed openstack/nova master: Run nova-conductor in native threading mode  https://review.opendev.org/c/openstack/nova/+/95857519:14
gibiI rearranged the series ^^ so we will get unit test run with native threading with the conductor change19:17
sean-k-mooneyack ill start reviewign those now19:17
gibilocally py312-threading passing for me with the conductor changes. I will do the manual testing in devstack tomorrow19:18
sean-k-mooneygibi: bar the conductor change which i need to review im +2 on all of those up until "Do not yield in threading mode"19:31

Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!