Monday, 2026-03-02

opendevreviewStephen Finucane proposed openstack/nova master: tests: Filter out (more) eventlet deprecation warnings  https://review.opendev.org/c/openstack/nova/+/97468611:26
opendevreviewStephen Finucane proposed openstack/nova master: console: Fix type error  https://review.opendev.org/c/openstack/nova/+/97468711:26
opendevreviewStephen Finucane proposed openstack/nova master: privsep: Remove dead code  https://review.opendev.org/c/openstack/nova/+/97711111:26
opendevreviewStephen Finucane proposed openstack/nova master: Add ruff-check  https://review.opendev.org/c/openstack/nova/+/97444111:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Add hints to nova.cmd  https://review.opendev.org/c/openstack/nova/+/70565711:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Add hints to nova.conf  https://review.opendev.org/c/openstack/nova/+/97468811:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Add hints to nova.console  https://review.opendev.org/c/openstack/nova/+/97468911:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Add hints to remaining top-level modules  https://review.opendev.org/c/openstack/nova/+/70565811:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Add hints to nova.virt, nova.virt.libvirt  https://review.opendev.org/c/openstack/nova/+/97422011:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Correct import issues  https://review.opendev.org/c/openstack/nova/+/97422111:26
opendevreviewStephen Finucane proposed openstack/nova master: mypy: Enable incremental checks  https://review.opendev.org/c/openstack/nova/+/97422211:26
opendevreviewStephen Finucane proposed openstack/nova master: mypy: Disallow incomplete defs  https://review.opendev.org/c/openstack/nova/+/97469011:26
opendevreviewStephen Finucane proposed openstack/nova master: mypy: Disallow untyped defs (where possible)  https://review.opendev.org/c/openstack/nova/+/97472511:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Add hints to nova.privsep  https://review.opendev.org/c/openstack/nova/+/97711211:26
opendevreviewStephen Finucane proposed openstack/nova master: typing: Add hints to nova.policy  https://review.opendev.org/c/openstack/nova/+/97715311:26
opendevreviewStephen Finucane proposed openstack/nova master: libvirt: Fix logging type error  https://review.opendev.org/c/openstack/nova/+/97843211:26
opendevreviewStephen Finucane proposed openstack/nova master: utils: Pass string to lock  https://review.opendev.org/c/openstack/nova/+/97843311:26
opendevreviewStephen Finucane proposed openstack/nova master: privsep: Pass strings, not ints  https://review.opendev.org/c/openstack/nova/+/97843411:26
opendevreviewStephen Finucane proposed openstack/nova master: privsep: Correct arg order  https://review.opendev.org/c/openstack/nova/+/97843511:26
opendevreviewStephen Finucane proposed openstack/nova master: Remove more dead code  https://review.opendev.org/c/openstack/nova/+/97843611:26
elodillessean-k-mooney: about novaclient release model change you raised during friday evening: yeah, so a patch should be created in openstack/releases repository where we move the novaclient from the deliverables/hibiscus -> deliverables/_independent (note that python-novaclient was release for gazpacho already)11:55
elodillessean-k-mooney: and we can do this after 2026.1 Gazpacho release, after we have created deliverables/hibiscus o:)11:56
elodillessean-k-mooney: i forgot about this one after the Antelope PTG :/11:57
elodillessean-k-mooney: one comment though: once we have moved python-novaclient to 'independent' we won't be able to do stable releases. in general this would not be a problem. BUT.... if there is any important bug fix and the master branch is dropped old py3 version (or other similar) support then we won't be able to get that fixed on stable branches. so this is something to be keep in mind if we want 12:00
elodillesto move a deliverable to cycle independent model12:00
elodillessean-k-mooney: i'm saying this as we had in the past some issues when multiple oslo project moved to independent model and caused issues when fixes could not be added and released. so oslo projects moved back from cycle independent to cycle based release model because of this.12:02
sean-k-mooneyelodilles: so whe ahve dicussed entirly frezing novaclient and discontiuing even bug fix release12:03
sean-k-mooneythe only reaosn it still exist is to provide a brindge for serivice like heat or neutron to move to the sdk12:03
sean-k-mooneyit might make snes for use to jsut go do that for those services next release12:03
sean-k-mooneythe same way neuton is moving nova to the sdk this cycle12:04
elodillessean-k-mooney: can we set novaclient as retired maybe next cycle?12:05
sean-k-mooneyi was thinking more for 2027.112:05
elodillesmaybe that would be the most straightforward way12:05
elodillessean-k-mooney: i see, 2027.1 also sounds OK to me12:06
sean-k-mooneywe still have services using it so if we set that as a goal for 2027.1 that give 2 relases to complete the move but we shoudl discuss that with the wider team12:06
sean-k-mooneywatcher had it as a priorty this cycle12:07
elodillessean-k-mooney: this cycle the clients are already released, so i think we cannot do anything about it12:07
sean-k-mooneyya we cant12:07
sean-k-mooneywhich is fine12:07
elodillesACK12:08
sean-k-mooneyi actully need to add "supprotign watcher in the sdk" to the ptg adjenda there and speak to stephen et al about the right way to do that so whtat we can also deprecate adn or remove the watcherclient python binding in a similar time frame12:09
gokhanhi folks, I am currently testing OpenStack Epoxy with Masakari for instance high availability. In my lab environment, I am using Ceph RBD as the backend for Nova ephemeral disks and for cinder disks. I am performing "hard failure" tests by physically pulling the power plug of a compute node to trigger Masakari’s evacuation process. While Masakari successfully detects the failure and starts the instances on another node, I am facing sever12:21
gokhane filesystem corruption on almost all evacuated VMs.Upon recovery on the new node, VMs drop to emergency mode with blk_update_request: I/O error and journal corruption. I suspect network=writeback is the primary culprit during a hard power-off because volatile buffers are lost before being flushed to the Ceph OSDs. Is it a common consensus to avoid network=writeback when using Masakari or any HA driver that expects survival after a hard node crash? 12:21
gokhanWhat is the recommended disk_cachemodes for Ceph RBD in production environments where power-loss-induced HA is a requirement? Is network=none the only safe path? Would switching to virtio-scsi (with hw_scsi_model and hw_disk_bus) significantly improve the "flush" command reliability during sudden failures compared to standard virtio-blk?12:21
sean-k-mooneygokhan: so i asume masakari is using evacuate. its a hard requiremt of using the evacuate api that the vm must be stopped on the orginal host and the admin or masikari in this case is requried to verify that before calling evacuate. by fencing i assume you mean that the phsyical host runing the vm is power off with ipmi/redfhist/pdu12:23
sean-k-mooneynetwork=writeback still commit all guest flushes to the backing store i.e. ceph12:24
sean-k-mooneyso if your guest os is properly flushing write then there shoudl not be any currption12:25
sean-k-mooneyor at least chanign to to none for the cache mode woudl not alter the flush behavior12:25
gokhanthanks sean-k-mooney , Regarding your point on fencing: Yes, Masakari is indeed using the Evacuate API, and I am simulating a hard failure by physically pulling the power plug of the compute node (simulating a sudden power outage or a motherboard failure where no graceful shutdown or IPMI action is possible from the host side).12:27
sean-k-mooneythe production recomendation for ceph is  hw_disk_bus=scsi hw_scsi_model=virtio-scsi and cash mode set to writeback12:27
sean-k-mooneythat is the recommendation form the ceph comunity12:27
sean-k-mooneyhttps://docs.ceph.com/en/latest/rbd/rbd-openstack/#image-properties https://docs.ceph.com/en/latest/rbd/rbd-openstack/#configuring-nova12:33
sean-k-mooneyby the way there si not ceph level rbd caching as well12:33
sean-k-mooneyi.e. at the ceph.conf level12:34
sean-k-mooneyi think the current recommendaions are to weriteback at the qemu level and enabling rbd cachign with ` rbd cache writethrough until flush = true` at the rbd/ceph level12:36
sean-k-mooneywith that sadi the removed the expiclty specificatgion of the livbirt cache_mode form there doc at some poitn12:36
sean-k-mooneyhttps://docs.ceph.com/en/mimic/rbd/rbd-openstack/#id212:37
sean-k-mooneygokhan: one of hte main reasonis for usein virtio-scsi in the past was for discard/trim supprot because that was not previously implmented wehn useing virtio-blk12:38
sean-k-mooneythat gap has been resolved in the last 2-3 years in modern qemu releases12:38
gokhansean-k-mooney, there is no cache on ceph environment 12:41
gokhanı will try with new configs 12:41
sean-k-mooneyrbd cachign is apprenly enabled by default reading the ceph docs12:41
gokhansean-k-mooney, do we need to disable it or keep as is 12:52
sean-k-mooneyhonestly this is not something we test or can really advise on. you might get more userful feedback form the ops mailing list or operator hour session12:53
sean-k-mooneythe best i can do is point you at the offical docs from teh ceph comnuity above12:53
sean-k-mooneywell that and https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L668-L70012:54
sean-k-mooneya flush form the guest is flushed to ceph in all cases expct unsafe12:55
sean-k-mooneyyou absolutely shoudl no use unsafe unless you hate your guests data12:55
sean-k-mooneyi woudl read that comemtn that explians the differnt modes in detail and decive which fits your performance and correctness needs12:57
sean-k-mooneya guest shoudl never rely on writs being commited without flushing. if they do they have a bug in there software 12:57
sean-k-mooneywritethough and directsync will inject extra fsync on every wite with the performance penaly that implies 12:58
gokhandisk_cachemodes=file=none,block=none,network=none13:07
gokhan  is also not worked forthis scenario. Now I am trying with recommended settings 13:07
gokhansean-k-mooney, I find the problem. ceph is locked disk from previous host. 13:13
gokhan rbd lock list cinder-volumes/volume-4263698a-ad00-472c-b4c8-62591664a8e113:16
gokhanThere is 1 exclusive lock on this image.13:16
gokhanLocker           ID                    Address13:16
gokhanclient.95989750  auto 125705438402288  10.x.x.x:0/119302136913:16
gokhansean-k-mooney, I also saw tour response to mail list https://lists.openstack.org/pipermail/openstack-discuss/2023-February/032256.html13:18
*** kaisers is now known as Guest414714:28
*** Guest4147 is now known as kaisers14:28
opendevreviewribaudr proposed openstack/nova master: FUP Add HW_PCI_LIVE_MIGRATABLE trait to PCI resource providers  https://review.opendev.org/c/openstack/nova/+/97731014:44
opendevreviewribaudr proposed openstack/nova master: Tidy up pci self.flags() calls in SR-IOV functional tests  https://review.opendev.org/c/openstack/nova/+/97847914:44
opendevreviewMax proposed openstack/nova master: fix: delete attachments after rescheduling delete  https://review.opendev.org/c/openstack/nova/+/97483214:48
opendevreviewTakashi Kajinami proposed openstack/nova master: Drop wrong strict assertion of quota class set id  https://review.opendev.org/c/openstack/nova/+/97849415:10
opendevreviewTakashi Kajinami proposed openstack/nova master: Drop wrong strict assertion of quota class class set id  https://review.opendev.org/c/openstack/nova/+/97849415:12
tkajinam^^^ this is strange but it's the behavior novaclient has been testing...15:12
opendevreviewTakashi Kajinami proposed openstack/nova master: Fix wrong strict assertion of quota class class set id  https://review.opendev.org/c/openstack/nova/+/97849415:13
opendevreviewTakashi Kajinami proposed openstack/python-novaclient master: DNM: testing  https://review.opendev.org/c/openstack/python-novaclient/+/97853215:13
UgglaNova meeting in ~35mn15:25
opendevreviewTakashi Kajinami proposed openstack/nova master: Fix wrong strict assertion of quota class class set id  https://review.opendev.org/c/openstack/nova/+/97849415:52
tkajinamhmmm so there is an issue with zuul now and jobs may be delayed for some time.15:54
tkajinamI mean it take some time until zuul actually starts processing notifications from gerrit15:54
tkajinam(seeing the discussion in OpenDev room in matrix 15:55
Uggla#startmeeting nova16:00
opendevmeetMeeting started Mon Mar  2 16:00:47 2026 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
UgglaHello everyone16:00
elodilleso/16:01
bauzaso/16:01
kaiserso/16:01
sean-k-mooneyo/ is here but distracted16:01
tkajinamo/16:01
gmaano/16:01
lajoskatonao/16:02
UgglaLet's start16:03
Uggla#topic Bugs (stuck/critical) 16:03
fwieselo/16:03
Uggla#info No Critical bug16:03
tkajinamhttps://bugs.launchpad.net/nova/+bug/214305716:03
Ugglaoh16:03
dansmitho/16:03
tkajinamthere is one I reported 1 hour ago (I changed its importance a few minutes ago)16:03
tkajinamwhich is blocking the novaclient CI16:03
tkajinamI've proposed a potential fix but zuul is now slow and is not starting jobs timely...16:04
Ugglaok I have seen you message earlier about this.16:04
tkajinamyeah16:05
Ugglathe bug sounds like something introduced with openapi.16:05
Ugglaam I right ?16:05
tkajinamyup16:06
tkajinamthe error is caused by schema validation of responses16:06
sean-k-mooneywell that or its now showign a bug in the client16:06
tkajinamwhich was recently enabled at the last phase of openapi series16:06
sean-k-mooneyya so respocne validation is only enabeld in ci16:06
gmaanis not the test should be fixed in novaclient?16:06
sean-k-mooneyit not inteneded for production use16:06
sean-k-mooneyand ya i am wonderign if its a novaclinet issue or not16:07
gmaanif API doc says, default is the only supported16:07
tkajinambut only API doc says that. that is the problem.16:07
gmaanfake-class-2-1 ->  'default'16:07
tkajinamI mean API doc says only default is supported but it's not actually validated in API16:07
sean-k-mooneytkajinam: that is the actual api contract however16:07
tkajinamI'm ok with "fixing" novaclient tests but in that case we should also fix nova to return an appropriate error to client for non-default id16:08
tkajinamcurrently it returns 400 with response schema validation error16:08
sean-k-mooneyim alittel confused because i see tempst/lib/cli in that path16:09
sean-k-mooneyare these novaclient tempest tests?16:09
sean-k-mooney"/home/zuul/src/opendev.org/openstack/python-novaclient/.tox/functional/lib/python3.12/site-packages/tempest/lib/cli/base.py"16:09
gmaannovaclient but they use tempest base test for cli16:10
sean-k-mooneyack16:10
gmaananyways, let's continue on gerrit, I wll comment there16:10
tkajinamok16:10
Ugglagood thx gmaan16:10
opendevreviewTakashi Kajinami proposed openstack/nova master: Fix wrong strict assertion of quota class set id  https://review.opendev.org/c/openstack/nova/+/97849416:11
Uggla#topic Gate status 16:11
Uggla#link https://bugs.launchpad.net/nova/+bugs?field.tag=gate-failure Nova gate bugs 16:11
Uggla#link https://etherpad.opendev.org/p/nova-ci-failures-minimal16:11
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 status16:11
Uggla#info Please look at the gate failures and file a bug report with the gate-failure tag.16:11
Uggla#info Please try to provide a meaningful comment when you recheck16:11
UgglaAnything else except what we discussed before ?16:12
Ugglaseems not so moving on.16:13
Uggla#topic Release Planning16:13
Uggla#link https://releases.openstack.org/gazpacho/schedule.html16:13
Uggla#info Nova deadlines are set in the above schedule16:13
Ugglainfo PTG etherpad for 2026.1 is available: https://etherpad.opendev.org/p/nova-2026.1-ptg16:13
Uggla#info Last week was Feature Freeze. (Thursday 20260226) Thanks for all the reviews and merges to close open features in this cycle.16:13
Uggla#topic Review priorities 16:15
tkajinammaybe it's time to prepare etherpad for 2026."2" ?16:15
Ugglatkajinam, yep it is part of post FF tasks.16:16
tkajinamUggla, ok :-)16:16
Uggla#link https://etherpad.opendev.org/p/nova-2026.1-status16:16
UgglaStarting here: https://etherpad.opendev.org/p/nova-2026.1-status#L70 there are interesting bugs to improve live migration. So I would like cores to review them.16:17
UgglaCan we also look at https://review.opendev.org/c/openstack/nova/+/972601: return error about external network to the user on build failure from Cardoe. It would be great to get special care for that one, because, unless I'm wrong, we missed it in the previous cycle, and it's important for Cardoe.16:17
kaisersQuick question related to reviews: I've a small change up to drop warnings on the libvirt volume Quobyte driver, it was recently set back to supported in Cinder. Does such a change need any additional background? It is not really a bugfix but also not a new feature, so i am not sure if should do anything or just let it sit. 16:18
cardoeThanks Uggla. We just got a number of bug reports about this from users. It was 4 by my last count so clearly folks are using nova+ironic out there.16:18
kaiserssry, didn't want to fire in between16:19
Ugglakaisers can you provide the link of the patch, I will add it to the list16:20
kaisersyes, thnx. https://review.opendev.org/c/openstack/nova/+/97730016:20
Uggladoes someone wants to add something else ?16:22
r-taketnUggla: I’ve read the irclogs and understood your decision that we have extra weeks regarding sev refactor series. Thank you for your confirmation.16:22
Ugglar-taketn yep.16:22
Ugglar-taketn, please continue the effort on that topic if you can. That will be good to have this for next cycle AMD-SNP and TDX.16:23
r-taketnThank you. I will continue this series during next two weeks. I’ve updated the code and comments. Please review them if cores have time. 16:25
Ugglayep at least gibi told me he will follow up on your serie16:26
gibir-taketn: I'm planning to spend time with your refactor tomorrow of the day after16:26
Ugglagibi 👍16:26
r-taketngibi: Thank you!16:26
Ugglacan we move on ?16:27
r-taketnThat’s all from me. Thanks16:27
Ugglar-taketn your welcome.16:27
Uggla#topic OpenAPI 16:28
Uggla#link: https://review.opendev.org/q/topic:%22openapi%22+(project:openstack/nova+OR+project:openstack/placement)+-status:merged+-status:abandoned16:28
Uggla#info still 0 remaining16:28
Uggla#info This topic is closed. A big thank you to stephenfin, sean-k-mooney, gmaan and all people involved with this.16:28
Uggla\o/16:28
gibicongrats16:29
Uggla#topic Stable Branches 16:29
gmaantkajinam: on that topic, I commented on gerrit, your change is ok for me but let's fix the test as this bug fix and we can discuss further about how better we can handle the error16:30
gmaanhttps://review.opendev.org/c/openstack/nova/+/97849416:30
* Uggla passing the mic to elodilles16:30
elodillesthx Uggla 16:30
elodilles#info stable gates should be in good state16:30
elodilles#info stable/2026.1 branch cut for libraries: https://review.opendev.org/c/openstack/releases/+/97839716:30
elodillesthanks for the review Uggla :)16:30
elodilles#info stable branch status / gate failures tracking etherpad: https://etherpad.opendev.org/p/nova-stable-branch-ci16:31
* elodilles gives back the mic to Uggla 16:31
Ugglathx elodilles16:32
Uggla#topic vmwareapi 3rd-party CI efforts Highlights16:32
* Uggla giving the mic to fwiesel16:32
Ugglaseems fwiesel is not available so moving on.16:34
Uggla#topic Gibi's news about eventlet removal16:34
gibio/16:34
* Uggla passing the mic to gibi16:34
gibinova-compute now can run in native threading mode (default is still eventlet)16:35
bauzas\o/16:35
dansmithyay16:35
gibiand nova-next runs nova-compute and the rest with native threading16:35
Uggla\o/16:35
gmaan\o/16:35
dansmithI apologize for being so absent on those reviews, but is there a reno asking operators to test and provide feedback?16:35
fwieselUggla: Sorry. was distracted. no updates 16:35
Ugglafwiesel no worries16:36
gibidansmith: there is reno about the change and nova logs at startup to asking for testing first in pre-prod16:36
gibidansmith: we can emphesize the feedback part in the prelude16:36
Uggladansmith, gibi, I will try to do it in prelude and highlights16:36
dansmithack, sounds good.. this, probably more than any previous thing really needs some concerted feedback I think16:37
gmaan++ for prelude16:37
gmaangibi: I am changing to run graceful shutdown job in both mode (eventlet mode as periodic), so that we will have coverage for both https://review.opendev.org/c/openstack/nova/+/97829216:37
dansmithprelude seems like a good place16:37
gibigmaan: cool16:37
gibidansmith: I agree.16:37
gibiwe punted novncproxy to H but there is some WIP patch up already from Kamil16:38
gibiI have a set of unit test patches that we can still land becore RC1 https://review.opendev.org/c/openstack/nova/+/97006916:38
gibiand I have a long list of TODOs for cleanups about executors that I will prepare patches for for early H16:39
gibiI still plan to hold the eventlet sync this week but maybe skip some calls during the RC period16:40
gibijust to restrat in full force once master is open of H16:40
gibithat is it from me16:40
* gibi passes the mic back16:40
gibiohh one more thing16:41
gibithat fwiesel is around16:41
gibifwiesel: could you help me debugging the vmware CI with native threading16:41
gibihttps://review.opendev.org/c/openstack/nova/+/97346816:41
opendevreviewDoug Goldstein proposed openstack/nova master: return error about external network to the user on build failure  https://review.opendev.org/c/openstack/nova/+/97260116:41
gibisee the comments from me in ^^16:41
gibiI mean in https://review.opendev.org/c/openstack/nova/+/97346816:42
gibithe image upload seems to be problematic16:42
gibibut snapshot works well with the libvirt driver so this is vmware driver specific16:42
gibiso I need your help16:42
Ugglafwiesel ?16:44
gibifwiesel: it is OK if you just read back and we can discuss outside of the meeting as well16:44
Ugglafwiesel will probably answer later, moving on to the next topic16:45
Uggla#topic Nova using openstack sdk for neutron16:46
* Uggla passing the mic to lajoskatona16:46
lajoskatonao/ thanks16:46
lajoskatonashort feedback: no progress since last meeting16:47
* Uggla shame on me I have still not provided my comments.16:47
lajoskatonaI plan to update the floating IP and port reated changes and clean them (the one for ports actually is red so have to work more on that I think)16:47
gmaanwhat's the plan on this, are we ok to merge those in rc period or wait till H?16:47
lajoskatonano worries, I think now everybody busy with the releae and later with RCs16:48
dansmithI assume these are not without risk right?16:48
gmaanyeah, there is risk16:48
Ugglagmaan, to my mind there is no urgency. And it is more something for H16:48
lajoskatonayes, I would wait till beginning of H to have time for testing in CI16:48
lajoskatona+116:49
gmaank, bcz I planned to review it but missed the FF so I can start but wait for H to open for merging 16:49
Ugglalajoskatona I think we are good.16:50
Ugglamoving to next topic16:50
Uggla#topic Bug scrubbing 16:50
Uggla#info up to 212 (-4)16:50
UgglaI'll continue to focus on that this week. 16:51
UgglaToday because we are a bit short of time, I would like to discuss only this bug: https://bugs.launchpad.net/nova/+bug/2141325 opened by gibi. 16:51
gibio/16:51
gibiso in short16:52
gibinova assumes during live migration prep that a single VM does not have multiple tap devs with the same MAC16:52
gibibut neutron allows setting MAC on port objects16:52
gibiso one can set up a VM with two ports in different neworks with the same MAC16:52
gibiand live migration will fail16:53
gibiquestion: do we want to fix it, or do we want to state that please don't try to do this16:53
dansmithis the limitation a nova thing or a libvirt/qemu thing?16:53
gibion nova, how we identifies tap devs across the two XMLs16:54
dansmithi.e. is it just us using the MAC as a primary key for mapping ports?16:54
gibijust us16:54
dansmithack16:54
gibiI rather not try to touch that XML gen code16:54
gibias manually managing MAC in the cloud is sort of an edge case in my head 16:54
dansmithmeanin you prefer we not fix this?16:55
gibiyepp I prefer documenting this as a known limitation16:55
bauzascould we at least fix it by doc ?16:55
bauzasah too late :)16:55
dansmithI think documenting it as a limitation now is obviously better than having it fail,16:56
dansmithbut it is also probably indicative of us using something as a PK that shouldn't be, and I also think there are probably legit reasons for doing this related to HA16:56
dansmithIIRC, we had a todo to make the NICs identifiable  by alias like the disks work I did,16:57
gibiI tried to poke the downstream reporter about their use case doing this but so far I got nothing16:57
dansmithwhich is probably what we _should_ be using instead of MAC right?16:57
bauzascould we fail early with a good self-explaned exception ?16:58
gibiyeah libvirt has aliases. I haven't checked if it fits to our live migration XML manipulation code to use that16:58
gibibauzas: today we fail here which is probably late https://github.com/openstack/nova/blob/264e868d4931595140260c0f655a10b525be38f7/nova/virt/libvirt/migration.py#L496-L50016:59
gibisorry not there16:59
gibithat is the place of the root cause16:59
gibiwe fail at16:59
bauzasI'm thinking of when creating the instance 16:59
gibi  File "/usr/lib64/python3.9/site-packages/libvirt.py", line 2225, in migrateToURI316:59
gibibauzas: that would open a can of worms, 1) pre-existing instance with duplicated MAC 2) interface attach17:00
bauzasif you ask for two ports, can we quick get the MAC addresses already or do we need another roundtrip ? (please, I don't want this if so)17:00
gibiif failing early is the goal then we should fail pre-live migration17:00
bauzasyup, then yes 17:01
dansmithwell,17:01
bauzaseither in the conductor check or later in compute pre-flighyt17:01
dansmithI think if we're going to say it's not allowed, we should prevent it at instance spawn/interface attach right?17:01
dansmithbecause in this case, a user can do something that prevents an admin from live migrating an instance17:01
dansmithwhich is not a cool thing to find out later17:01
bauzasthen it should be when we attach a port ?17:02
dansmithso if we say "not allowed" it probably needs to be enforced much earlier17:02
dansmithbut again, I feel like this is probably something we can/should fix (even if later) with the alias stuff like we did for disks and planned for nics17:02
bauzasright17:03
gibiOK, if preventing it needs a bunch of code then maybe fixing it is comparable complexity (I was on the side of doccing it)17:03
dansmiththe disk stuff has fallen out of my brain a bit, but seems like we could use the port uuid as the alias and then we'd be able to match up the port and interface in that code, right?17:03
dansmithgibi: well, I think documenting it also documents "how to prevent your operator from moving your instance around" :)17:04
gibidansmith: right port uuid is unique in openstack, nic alias is unique in the VM17:04
dansmithyeah17:04
gibiOK I will summarize this discussion back to the bug17:05
gibiI got enough feedback that a simple doc patch will not be enough :)17:05
gibithanks 17:05
dansmithsorry gibi :/17:06
Ugglagibi thanks to you explaining this.17:06
UgglaWe are already overtime, so time to close for this week.17:06
cw0306-lee[m]Hi! I have a question.17:06
gibidansmith: no worries :)17:07
Ugglacw0306-lee[m] last minute question, please go ahead17:07
cw0306-lee[m]I tried to fix bug and added it to etherpad proposed bugfixes, will they be reviewed in order?17:07
Ugglacw0306-lee[m] we will try. There is no specific order though, really depend on prio and reviewer bandwith.17:08
cw0306-lee[m]Uggla: Thank you!17:09
Ugglareadiness too17:09
Ugglaok so now time to close, thanks for joining this meeting. Have a nice day/evening and see you next week. 17:10
Uggla#endmeeting17:10
opendevmeetMeeting ended Mon Mar  2 17:10:29 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)17:10
opendevmeetMinutes:        https://meetings.opendev.org/meetings/nova/2026/nova.2026-03-02-16.00.html17:10
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/nova/2026/nova.2026-03-02-16.00.txt17:10
opendevmeetLog:            https://meetings.opendev.org/meetings/nova/2026/nova.2026-03-02-16.00.log.html17:10
gibithanks Uggla 17:10
elodillesthanks o/17:10
lajoskatonao/17:11
Ugglagibi thanks to you, we have one more bug triaged !17:11
tkajinamgmaan, regarding that novaclient tests update, do you know any stestr magic to avoid specific tests from being executed concurrently ?17:12
tkajinamhttps://opendev.org/openstack/python-novaclient/src/branch/master/novaclient/tests/functional/v2/test_quota_classes.py#L79-L10717:12
tkajinamthe problem is that there are few tests accessing quota-class and if we use only the default one then we should refactor these to avoid conflicts among these17:13
tkajinamI'll look into that tomorrow but the test update looks tricky for me now17:15
sean-k-mooneytkajinam: for what its worth even in queens we have notes stating that only the default classs is supproted by nova17:18
sean-k-mooneyhttps://docs.openstack.org/nova/queens/admin/quotas.html#to-update-quota-values-for-an-existing-project17:18
sean-k-mooneyhttps://that.guru/blog/quotas-in-openstack/#quota-classes looking at the history the supprot for this was ripped out in ocata and there was never eally any intree supprot17:21
gmaantkajinam: for quota concurrent tests, I think you can use LockFixture like tempest is doing https://github.com/search?q=repo%3Aopenstack%2Ftempest%20compute_quotas&type=code17:21
gmaandoing it via stestr also possible but that will be more complicated of grouping them and running in serial mode17:23
sean-k-mooneyya porting the serial decorator is one approch17:25
*** mtreinish_ is now known as mtreinish17:30
opendevreviewDominik proposed openstack/nova-specs master: Repurpose NUMA Topology with Resource Providers for 2026.2  https://review.opendev.org/c/openstack/nova-specs/+/97857017:39
*** mtreinish_ is now known as mtreinish17:53
Zhan[m]Hi friends, I have a quick question. Does Nova allow launching/live-migrating VMs to a host with nova-compute disabled? For example, if I want to do a QC before putting the host back to service, I can launch a VM on there with elevated privileges (e.g., using admin credential), make sure everything works, and then finally enable nova-compute.18:29
sean-k-mooneyZhan[m]: no. disbaling a compute node exist to prevent schdulign to the host18:31
sean-k-mooneywe enforce this with a frobiden trait via a placment prefilter18:31
sean-k-mooneyhttps://github.com/openstack/nova/blob/master/nova/scheduler/request_filter.py#L243-L25518:33
Zhan[m]Yeah I saw that earlier too. So the only way is to enable it, launch a VM, disable it, make sure everything's correct, and then re-enable it?18:34
sean-k-mooneynormlaly i would suggest using somethign else liek a requried trait to isolate the host or teh tenat affinity filter for exmaple18:34
sean-k-mooneyi.e. test it by addign teh host to a host aggreate that can only be used by the admin project18:35
sean-k-mooneyhttps://docs.openstack.org/nova/latest/admin/aggregates.html#tenant-isolation-with-placement18:35
sean-k-mooneyso restirct it to your admin project, renebaled it, test it adn then eitehr remove it form teh aggreate or disable it again to do futher maintance18:36
Zhan[m]ahh I see, something like a "maintenance" aggregate where only admin can access18:36
sean-k-mooneyyep exactly18:36
Zhan[m]I see, thanks!18:37
sean-k-mooneyhaving a third sate enabeld,disabled and maintance woudl perhaps have been a cleaner way to do that but you can emulate that with aggreates18:38
Zhan[m]I think having a third state make sense too, but both approaches solve the same problem ultimately. Wondering if that can be something to discuss and maybe implement, or aggregate is still the recommended way to go?18:40
Zhan[m]It would be good to put this info somewhere, I can imagine others having this question too.18:43
dansmithbeing able to target disabled hosts seems reasonable to me, without requiring the aggregate thing18:47
dansmithlike, if an admin has disabled a host to prevent regular users from scheduling to it,18:47
dansmithbut has (of course) the permission to target hosts specifically either with a boot or a migration,18:48
dansmithit seems totally reasonable to ignore the disabled check for that case18:48
dansmithlike if I need to disable a failing host, then send a build to it to see if I've fixed it before I let others go there18:48
dansmithor I disable a host to prevent it from filling up so I can emergency migrate instances to it to escape a failing node18:48
Zhan[m]yeah I was actually thinking about replacing the ComputeFilter by a custom filter that will say OK to a disabled host if it's admin and state is up and destination is explicitly specified, until I found that mandatory pre-filter.18:49
dansmithyeah, the pre-filter could do the same18:49
Zhan[m]yeah, I think we can do something in nova that does "if admin (or policy-allowed-role) and if host is specified and if state it up", remove that specific pre-filter?18:53
Zhan[m]imo a third state or this approach both seem fine, just that the prior one may need more changes18:53
sean-k-mooneyso it used to be possibel before the pre-filter with --force18:58
sean-k-mooneybtu that was because that bypassed everything18:58
sean-k-mooneyinclduign the status check provide the compute node existed18:59
Zhan[m]yeah probably not a good idea...18:59
Zhan[m]Any thoughts? Maybe I should do a bug report or a spec?19:02
sean-k-mooneyproably a small spec. as its sort of an api change. it is an api change if we had a new state its techinally a policy change adn or cofnig option if we take the other approch19:03
sean-k-mooneyso not a bug19:03
Zhan[m]okie dokie, I'll work on it and list all approaches19:04
sean-k-mooneybut its not a super invaisve feature given it woudl be highly restrict to admins by default19:04
Zhan[m]definitely, if we add a policy then role will likely be admin by default too. but it's good to have it configurable as folks may create dedicated role for maintenance19:05
sean-k-mooneyya i was debating if the manager role would make sense but i dont think so. but if we have a new policy rule and defautl it to admin then operators can alwasy confirure it to the role they perfer19:07
sean-k-mooneyone thing that is not clear to me is if admin shoudl jsut alwasy bypass the check or if this shoudl be opt in either by updating the role or a config option19:08
sean-k-mooneyfor cross cell migrate we dont even allow admins to do that by default19:08
sean-k-mooneythey have ot enable it via custom policy if they want to supprot that in the cloud so this coudl be the same19:08
Zhan[m]I would advocate for including admin in the new policy (default admin too) instead of making it always bypass as it's just very clear & straight forward, people won't need to guess or look at code to know "ok admin can do this".19:15
Zhan[m]I'll put both option in the spec regardless19:15
*** erlon3 is now known as erlon19:51

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