Tuesday, 2026-08-25

opendevreviewMerged openstack/nova master: doc: Add releases information for SEV-ES and SEV-SNP  https://review.opendev.org/c/openstack/nova/+/100216400:10
opendevreviewMerged openstack/nova master: weights: add regression test for JSON log TypeError (bug 2028518)  https://review.opendev.org/c/openstack/nova/+/98902900:10
sean-k-mooneymelwitt: ya i had to special case get_endpoint_data in my patch as weel https://review.opendev.org/c/openstack/nova/+/994579/4/nova/accelerator/cyborg.py#14800:11
sean-k-mooney.get_endpoint() does not incldue the microversion info00:12
sean-k-mooneyso you need to extract the info via the session00:12
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Use a per-instance cephx identity for CephFS shares  https://review.opendev.org/c/openstack/nova/+/100190000:45
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Extract share management from ComputeManager  https://review.opendev.org/c/openstack/nova/+/99934900:45
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Support cold migration with virtiofs shares  https://review.opendev.org/c/openstack/nova/+/98963300:45
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Reconcile stale Manila share access rules  https://review.opendev.org/c/openstack/nova/+/100191600:45
opendevreviewMerged openstack/nova master: Fix PEP-765 syntax warning  https://review.opendev.org/c/openstack/nova/+/98863601:05
opendevreviewMerged openstack/nova master: Drop description for ancient horizon  https://review.opendev.org/c/openstack/nova/+/98941901:06
opendevreviewMerged openstack/nova master: Update contributor guide for 2026.1 Gazpacho  https://review.opendev.org/c/openstack/nova/+/96189601:06
opendevreviewMerged openstack/nova master: Update contributor guide for 2026.2 Hibiscus  https://review.opendev.org/c/openstack/nova/+/99787001:06
opendevreviewMerged openstack/nova master: Filter out deleted instances when looking for build timouts  https://review.opendev.org/c/openstack/nova/+/88012501:31
opendevreviewGhanshyam Maan proposed openstack/nova master: Fix graceful shutdown functional unstable tests  https://review.opendev.org/c/openstack/nova/+/100222302:27
opendevreviewGhanshyam Maan proposed openstack/nova master: Fix graceful shutdown functional unstable tests  https://review.opendev.org/c/openstack/nova/+/100222302:30
gmaangibi: melwitt ^^ this is possible fix for the graceful shutdown unstable tests. We can see how it goes, I am not closing bug in case there are other race exist there.02:31
gmaanI added it on top of other fixes02:31
opendevreviewClif Houck proposed openstack/nova master: perf(ironic): eliminate O(N²) ProviderTree deepcopy at startup  https://review.opendev.org/c/openstack/nova/+/98067603:13
opendevreviewClif Houck proposed openstack/nova master: Parallelize per-node resource updates  https://review.opendev.org/c/openstack/nova/+/98067903:13
opendevreviewGhanshyam Maan proposed openstack/nova master: Add functional test for cleanup_host cancelling queued live migrations  https://review.opendev.org/c/openstack/nova/+/98647703:40
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Reconcile stale Manila share access rules  https://review.opendev.org/c/openstack/nova/+/100191605:02
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Use a per-instance cephx identity for CephFS shares  https://review.opendev.org/c/openstack/nova/+/100190005:47
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Extract share management from ComputeManager  https://review.opendev.org/c/openstack/nova/+/99934905:47
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Support cold migration with virtiofs shares  https://review.opendev.org/c/openstack/nova/+/98963305:47
opendevreviewGoutham Pacha Ravi proposed openstack/nova master: Reconcile stale Manila share access rules  https://review.opendev.org/c/openstack/nova/+/100191605:47
*** benj_5 is now known as benj_07:20
*** nicolairuckel_ is now known as nicolairuckel08:00
gibiI'm pretty sure that some of the instability is due to threading allows switching between tasks at any point of execution and then we have the CI slowdown that forces more switches resulting in more race conditions materializing now08:43
gibiI ran a lot of functional-threading tests locally and reasonable amount of it in CI before we made the functional-py313-threading voting and it was all shiny 3 weeks ago. Now I have multiple bugs filed on new race conditions08:44
gibigmaan: https://review.opendev.org/c/openstack/nova/+/1002223 needs more tweaking according to CI 08:55
gibiI assigned two bugs to myself and started looking for a root cause / fix https://bugs.launchpad.net/nova/+bug/2164606 and https://bugs.launchpad.net/nova/+bug/2164893 they are more frequent than the rest of non-graceful hits09:04
gibibtw I can trigger some new race conditions locally if I artificially restric the amount of cpu and io I give to tox via09:51
gibisystemd-run --user --scope -p CPUQuota=400% -p IOReadBandwidthMax="/ 2M" tox -e functional-py313-threading09:51
opendevreviewBalazs Gibizer proposed openstack/nova master: Stabilize TestInstanceNotificationSample.test_instance_action  https://review.opendev.org/c/openstack/nova/+/100224609:55
gibiOK that is one fix for the many ^^09:56
opendevreviewTakashi Kajinami proposed openstack/nova master: Remove unused default argument  https://review.opendev.org/c/openstack/nova/+/92207310:07
sean-k-mooney'compute.exception' is a notification?11:16
sean-k-mooneyis that like any uncuaght excpetion on the compaute agent11:16
sean-k-mooneyor something we send at a specific time11:16
sean-k-mooneyi mean that definetly a think that exists and we are waitign for it in rebuild filaers https://github.com/openstack/nova/blob/9aa9a54e04fdfec9a8775bf90488558783179d4b/nova/tests/functional/test_servers_resource_request.py#L2230-L223211:19
sean-k-mooneyso the pattern looks right11:19
gibisean-k-mooney: yepp compute.exception is a catch all notification12:03
sean-k-mooneyi +2'd with a note about doing it for our only other *_with_error helper as well12:04
sean-k-mooneyi assume we may have a similar race for share detach as share attach12:05
gibiyepp you are right and I'm updating the patch12:06
sean-k-mooneyi dont know if it has a simialr assertion however hence +2, if you think its needed ill happilry re review12:06
sean-k-mooneycool i didnt really want ot dig into that to confirm12:06
opendevreviewBalazs Gibizer proposed openstack/nova master: Stabilize TestInstanceNotificationSample.test_instance_action  https://review.opendev.org/c/openstack/nova/+/100224612:15
gibisean-k-mooney: fixed it up ^^12:15
sean-k-mooneyoh it was listing them but only asserting 212:32
sean-k-mooneyfun12:32
gibi:)12:32
opendevreviewBalazs Gibizer proposed openstack/nova master: Stabilize TestInstanceNotificationSampleWithMultipleCompute  https://review.opendev.org/c/openstack/nova/+/100228113:06
gibianother fix ^^13:06
gibiI'm taking https://bugs.launchpad.net/nova/+bug/2164888 as well now13:10
clifanyone familiar with a mypy failure around "ExceptionGroup" it complains the name is not defined despite it being a builtin...13:25
clifalso are nova tox functional tests able to be run locally?13:31
sean-k-mooneyan excpetion group is not a type13:31
sean-k-mooneyits a collection of types13:32
sean-k-mooneyas far as im aware at least13:32
sean-k-mooneybut no i ahve not seen that13:32
clifwdym, it is a type isn't it? https://docs.python.org/3/library/exceptions.html#ExceptionGroup13:32
sean-k-mooneyhum ok added in 3.1113:33
sean-k-mooneyit wasnt a type before this is the first release we have been abot to use 3.11 features13:33
sean-k-mooneybefore expction groups were jsut a tuple of possible expctions13:33
clifit seems to work fine based on local testing but mypy blows up on it in tox pep813:34
sean-k-mooneywhy are you trying to define it in a type signigure?13:34
sean-k-mooney3.11 was for derive not the full feature 13:35
clifI'm not sure what you mean, I'm not using it as a type signature but as a raised exception due to stephenfin's review comment13:40
sean-k-mooneyi would not use it here for example https://review.opendev.org/c/openstack/nova/+/980679/12/nova/compute/manager.py#1157213:40
sean-k-mooneywhich comment13:40
clifhttps://review.opendev.org/c/openstack/nova/+/980679/comment/a6f02680_f3646440/13:41
stephenfinclif: Got the review link to hand? I can take a look13:41
stephenfinnvm :)13:41
sean-k-mooneyhttps://review.opendev.org/c/openstack/nova/+/980679/comment/a6f02680_f3646440/13:41
sean-k-mooneyso i dont think we shoudl be raiseing an execption groupu there13:41
sean-k-mooneyi intentlaly was raisign only the first excption13:41
sean-k-mooneywe coudl do that but we cant assuem the srounding code will handle the group properly13:42
stephenfinclif: look at tool.mypy.python_version in pyproject.toml :(13:42
clifstephenfin: aha :(13:43
clifis it fine to bump that to 3.11?13:43
stephenfinnot the first itme I've been bitten by that https://github.com/python/mypy/issues/1934913:43
sean-k-mooneyah 3.10 so it was a 3.11 related chagne13:43
stephenfinclif: yeah, it should have been done when we bumped project.requires-python but obviously got missed13:44
stephenfinjust remember to run `tox -e mypy` locally before submitting since mypy _might_ raise new warnings about e.g. things that were newly deprecated in Python 3.1113:45
stephenfinsean-k-mooney: ExceptionGroup is new in 3.11, yup13:45
sean-k-mooneywill an expction group match an excpt block that only matches on the encloded exeptions?13:45
sean-k-mooneyand will it properly handel all of them or just one13:45
stephenfinI believe so, but there's a very easy way to test it ;)13:45
opendevreviewClif Houck proposed openstack/nova master: Bump python_version for mypy to 3.11  https://review.opendev.org/c/openstack/nova/+/100229013:45
sean-k-mooneystephenfin: right so the issue i have is your chanign the behvior of the algorthim not just changing the typing13:46
sean-k-mooneyagain it was intentlly only rasing the first excption not all of them13:46
sean-k-mooneyit was loging all of them13:46
stephenfinSorry, what algorithm was changed?13:49
sean-k-mooneyit now raising all the expctions as a group instead of just the first one13:50
sean-k-mooneyin the non paralle mode it would have failed after the first expction13:51
sean-k-mooneyso the intent was to keep the external behvior the same 13:51
sean-k-mooneywe can rasie all of them as a group but that a slightly diffenernt semantic13:51
stephenfin(a) that desire isn't captured anywhere so I'm not sure if that was the intent13:52
sean-k-mooneythat was part of the code i wote13:52
sean-k-mooneybefore clif took it over so that was the intent13:52
stephenfinand (b) the behavior has already changed since in the non-parallel mode, only N/M operations would have run (where the Nth operation raised the exception) whereas now M operations will run13:52
stephenfinif operation N raises FooException and N+X raises BarException, raising only FooException is lossy and makes things harder to debug. My suggesion was to try to avoid that13:54
sean-k-mooneyboth can work but im not sure what we get by raisign all of them13:54
sean-k-mooneystephenfin: that why all of them were logged13:55
sean-k-mooneyfor debugging13:55
sean-k-mooneystephenfin: we intend to backport this13:55
sean-k-mooneyso that why im qustioning using a 3.11 feature13:55
stephenfinokay, backporting does change the equation. I wasn't aware of that13:55
sean-k-mooneywell we intned to backprot he prevsiou patch this wone was deabateable13:55
sean-k-mooneywe can modify it on backport if requried but ya the hope ofr the O(n2) patch at least was to backport to all stable branches13:57
sean-k-mooneythe parallisze i know operator will want it backported but im not sure it fully fits stabel policy13:58
stephenfinokay, if that's the plan then yeah, we'd want to go back to just raising the first exception13:58
sean-k-mooneyso that patch may be propsoedn abut not mergd13:58
sean-k-mooney*may be proposed but not merged13:59
sean-k-mooneywe can do both we could add a patch to swap to excption groups as a followup13:59
stephenfinclif: ☝️ apologies for the slight waste of your time there 🙇14:00
sean-k-mooneybut if we can keep this 3.10 compatiable it woudl be nice14:00
sean-k-mooneystephenfin: since your about https://review.opendev.org/c/openstack/nova/+/1002246 gibi puhsed htat to stablise a functional test mind taking a look14:03
clifok, yea I can revert the ExceptionGroup change14:11
clifrelatedly, some functional tests are failing on the parallel per-node resource update change, for instance: nova.tests.functional.libvirt.test_pci_in_placement.PlacementPCIInventoryReportingTests.test_device_reconfiguration_with_allocations_config_change_stop fails because the exception raised has changed14:12
clifany advice on that?14:14
clifwell now that I think about it some the failures may be related to the exceptiongroup change14:19
clifah yea it was the exceptiongroup changing how exceptions got raised, makes sense14:34
opendevreviewClif Houck proposed openstack/nova master: Parallelize per-node resource updates  https://review.opendev.org/c/openstack/nova/+/98067914:39
gmaangibi: ack, will check15:05
gibigmaan: thanks15:19
gmaanI saw this again, I think this hit twice, I will work on this today https://bugs.launchpad.net/nova/+bug/216488015:23
gibigmaan: thanks15:26
melwittthank you gmaan for working on these 15:29
melwittsean-k-mooney: gotcha, I see. yeah. really hope we can switch to sdk rather than add those one-off ... the patch is not too ugly (I hope!)15:31
stephenfinmelwitt: I'm +2 on https://review.opendev.org/c/openstack/nova/+/1002007 again. The ksa fix is also merged and should be released this week, so we could potentially remove that subclass next week. Just need sean-k-mooney to revisit that now (hint hint)16:56
sean-k-mooney[m]cool we are at the park with frya ill be home in a bit and ill take a look then17:00
melwittstephenfin: super, thank you17:22
opendevreviewStephen Finucane proposed openstack/nova master: cinder: Use keystoneauth1 Adapter for endpoint discovery  https://review.opendev.org/c/openstack/nova/+/100235917:28
sean-k-mooneystephenfin: melwitt  im pretty sure that https://review.opendev.org/c/openstack/nova/+/1002007/8/nova/accelerator/cyborg.py is incorrect18:07
sean-k-mooneystephenfin: melwitt  we do not use an admin client today to talk to cyrbog18:07
sean-k-mooneywew shoudl in the future but we dont today18:08
sean-k-mooneyand that cached client is confging it ot an admin client18:08
sean-k-mooneythe cachign itself might makes sense in the future but not today when we alwasy need to use the users token18:08
sean-k-mooneyi think the problem is that this is folow the same pattern we use to tak to ironic and manilla but the cybrog interaction is not the same as the inronic interaction18:09
melwittsean-k-mooney: hm ok, so that should be admin=False. I had thought it was but that must have been an earlier iteration18:09
sean-k-mooneyya which means we take the if brnach here https://review.opendev.org/c/openstack/nova/+/1002007/8/nova/utils.py#94018:10
sean-k-mooneynot the else18:10
sean-k-mooneybut if we take the if brnach we done need to chagne the signigure to pass in ksa_auth18:11
sean-k-mooneyif admin false branch already correctly constuct the service user token and uses the uers token form the context to constuct the sdk connection18:12
melwittargh dammit18:12
melwittsorry18:13
sean-k-mooneyon a side note i messed up my grub config on one of my dev hosts, i do not know the root password adn signle user mode didnt work so im not reinstalling an os18:13
sean-k-mooneyso its one of those days18:13
melwittI have been there18:14
sean-k-mooneyi didnt properly unstack and swift or rather the swift loopback volume we create in devstack is preventing the boot18:17
sean-k-mooneymelwitt: have you had a chance to check the auth change on a devstack by the way18:18
melwittsean-k-mooney: no, but cyborg-tempest-py3 was definitely showing when it fails18:19
sean-k-mooneyack i was plannign to test it on the host im reinstalling before trying to test the nvme serise18:20
melwitt++18:23
opendevreviewMerged openstack/nova master: Stabilize TestInstanceNotificationSample.test_instance_action  https://review.opendev.org/c/openstack/nova/+/100224618:30
sean-k-mooneymelwitt: my main concern woudl be that all the attachment end up beign owned by nova instead of the end user18:32
sean-k-mooneyi.e. the client worked but the ownership was wrong18:32
melwittyeah, that is an important point18:32
sean-k-mooneyhence why i want to check it on a devstack si i can check the db to verify the ownership18:33
melwittyeah, you are right18:36
opendevreviewGhanshyam Maan proposed openstack/nova master: Fix graceful shutdown functional unstable tests  https://review.opendev.org/c/openstack/nova/+/100222319:06
opendevreviewmelanie witt proposed openstack/nova master: Switch Cyborg client from keystoneauth1 to openstacksdk  https://review.opendev.org/c/openstack/nova/+/100200720:11
gmaanmelwitt: gibi updated this, matching the service_down_time with what  any test can take time to execute, also added the root cause of why service down is happening in between of tests in note/commit msg  https://review.opendev.org/c/openstack/nova/+/100222320:28
opendevreviewMerged openstack/nova master: Stabilize TestInstanceNotificationSampleWithMultipleCompute  https://review.opendev.org/c/openstack/nova/+/100228120:52
opendevreviewmelanie witt proposed openstack/nova master: Switch Cyborg client from keystoneauth1 to openstacksdk  https://review.opendev.org/c/openstack/nova/+/100200721:29
opendevreviewmelanie witt proposed openstack/nova master: Switch Cyborg client from keystoneauth1 to openstacksdk  https://review.opendev.org/c/openstack/nova/+/100200721:32
melwittsean-k-mooney: ok, I think https://review.opendev.org/c/openstack/nova/+/1002007 should be in the right state now. I'm leaving it in WIP mode while we test it more22:26
melwittsean-k-mooney: seems to work as intended on my devstack, created a server with the fake_driver device and the extended_accelerator_requests table has a record with project_id of the "demo" project I created the server with23:47

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