Thursday, 2026-06-04

opendevreviewMerged openstack/nova master: Avoid reading 4GiB of reason text from RFB  https://review.opendev.org/c/openstack/nova/+/99122001:42
opendevreviewMerged openstack/nova stable/2025.1: [CI][stable-only] nova-grenade-multinode fix  https://review.opendev.org/c/openstack/nova/+/98957903:18
*** benj_5 is now known as benj_04:45
opendevreviewMerged openstack/nova master: Fix swap disk creation skipped on NFS during cold migration  https://review.opendev.org/c/openstack/nova/+/98854707:54
UgglaReminder: Today is spec review day09:04
opendevreviewKamil Sambor proposed openstack/nova master: Change nova-alt-configurations job  https://review.opendev.org/c/openstack/nova/+/98317909:48
opendevreviewMerged openstack/nova-specs master: Intel TDX support in libvirt driver  https://review.opendev.org/c/openstack/nova-specs/+/97960810:07
gibiUggla: both SNP and TDX specs are approved now. So we made good progress :)10:19
opendevreviewMerged openstack/nova-specs master: libvirt: AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova-specs/+/98337610:24
gibibauzas: dansmith: I reviewed the bidirectional RPC liveness spec and I have one RPC compat questions that one of you probably can answer easily https://review.opendev.org/c/openstack/nova-specs/+/98438410:48
bauzasgibi: I can try to look at this afternoon 10:49
gibithanks10:55
* bauzas will have to invoke his very old chamanic knowledge about ServiceGroups and RPC compat :)10:57
opendevreviewdalekseev proposed openstack/nova master: Restrict machine type check to QEMU instances  https://review.opendev.org/c/openstack/nova/+/99113712:10
opendevreviewTakashi Kajinami proposed openstack/osc-placement master: Update support python versions  https://review.opendev.org/c/openstack/osc-placement/+/99160312:18
gibisambork, Uggla filed a new CLI UX bug for eventlet + py3.14 https://bugs.launchpad.net/nova/+bug/215524512:24
Ugglagibi ok12:27
Ugglaand cool for the CC spec12:28
sambork@gibi, ack, I can take a look at that. At first glance, it looks like a similar issue to the one you described yesterday with oslo and pickling12:28
gibiyeah someting wrong happens when the python interpreter exits12:35
gibiit only uggly but causing no issue as far as I see12:35
opendevreviewKamil Sambor proposed openstack/nova master: Restore 'fork' start method for daemon mode on Python 3.14  https://review.opendev.org/c/openstack/nova/+/98781512:35
opendevreviewKamil Sambor proposed openstack/nova master: [dnm] testing python 3.14 with restored fork  https://review.opendev.org/c/openstack/nova/+/98782112:35
Ugglagibi, if you have time to review 471815: Expose vlan trunking in metadata/configdrive | https://review.opendev.org/c/openstack/nova-specs/+/47181512:50
Ugglasorry to bother all of you with this one, but I would like not to miss it another time.12:51
opendevreviewThibaut Démaret proposed openstack/nova master: libvirt: add disk rotation_rate support for local disks  https://review.opendev.org/c/openstack/nova/+/97969313:15
Ugglagmaan, there is this new bug entered by jlejeune that might interest you: https://bugs.launchpad.net/nova/+bug/215515313:25
opendevreviewChungWon Lee proposed openstack/nova-specs master: Spec for reduce failed builds value by time  https://review.opendev.org/c/openstack/nova-specs/+/99163613:31
gibiUggla: I feel like https://bugs.launchpad.net/nova/+bug/2155153 is a duplicate of an older bug of mine. Looking for it...13:33
Ugglagibi, :) I was just thinnking about it13:33
gibifound https://bugs.launchpad.net/nova/+bug/2147554 and did the paperwork13:35
opendevreviewChungWon Lee proposed openstack/nova-specs master: Spec for reduce failed builds value by time  https://review.opendev.org/c/openstack/nova-specs/+/99163613:37
opendevreviewribaudr proposed openstack/nova master: Add reproducer for bug #2115798  https://review.opendev.org/c/openstack/nova/+/99164213:46
opendevreviewribaudr proposed openstack/nova master: Fix race condition in security group add/remove  https://review.opendev.org/c/openstack/nova/+/99164313:46
opendevreviewribaudr proposed openstack/nova master: Fix misleading description of assisted volume snapshots  https://review.opendev.org/c/openstack/nova/+/99164913:57
sean-k-mooneybauzas: is this on your review list https://review.opendev.org/c/openstack/nova-specs/+/99025914:02
bauzasno14:02
sean-k-mooneyi woudl assume if we are goint to do the virtio fs cold migrtion of host attach stuff14:02
sean-k-mooneywe would want that to be ocmplete first14:02
bauzasI won't have time for reviewing the implementation14:02
sean-k-mooneyack14:03
sean-k-mooneyit feel strang not to adress the hugepage/file backed memory pain point with manilla share attachment so that the feature "just works" with all vms14:04
sean-k-mooneybefore inveting in the lifecyle operations but ok the memfd work is much smaller then the move opertions14:04
sean-k-mooneyor hot attahce/detach14:04
sean-k-mooneytis technically not a hard requriement 14:05
sean-k-mooneybut it would also make testing this in ci simpler14:05
sean-k-mooneyanyway im happe to review the memfd sutff if someone else is14:05
opendevreviewKamil Sambor proposed openstack/nova master: Fix resize error notifications using wrong event type  https://review.opendev.org/c/openstack/nova/+/99165714:05
sean-k-mooneyi probaly wont have tiem for the other virtio fs features however14:05
cw0306-lee[m]Hi, Uggla! I tried to write spec for reducing failed_builds value. https://review.opendev.org/c/openstack/nova-specs/+/99163614:06
Ugglacw0306-lee[m] cool thanks putting it on the list14:54
sean-k-mooneycw0306-lee[m]: i provcided some feedback on wordign ectra15:20
gouthamrsean-k-mooney: Uggla: my 0.02 agreement for your sentiment.. i'm okay continuing to work on hot attach/live migration with virtiofs and let them slip into I. Would really like the memfd and cold migration enablement to land in hibiscus15:20
gouthamrand can help push this as much as possible :) 15:21
sean-k-mooneygouthamr: dont think tis stricly a bloceker for any of that work15:21
sean-k-mooneybut if we are going ot invest int impvoing the ux i think the memfd work si certally one of the more valuabel aspects15:21
sean-k-mooneynot that i really have an issue with any of the virtiofs enchment in genreal15:22
sean-k-mooneywe can test with file backed memory or hugepages in ci today if we need too but for the maniall share move operaiton we will need to do that and have tempest testing 15:23
sean-k-mooneyfunctional testing will prove out most of the flow but i woudl really like to see tempest actuly do the moves as well15:23
gouthamryes.. am working on that part15:23
gouthamrnot memfd, but working on the cold migration tempest tests right now, i see memfd as making things a whole lot easier15:24
gibibauzas: thanks for confirming my RPC compat issue in the bidirectional RPC hearthbeat spec 15:24
sean-k-mooneyUggla: how are we currently testing manilla shares in our jobs15:25
sean-k-mooneyi dont see manilla enabeld in any of them15:26
sean-k-mooneythat defintly somethign we will have to corect if it is a gap today before building on more features15:27
Ugglasean-k-mooney, there are tests against devstack in the sdk.15:44
sean-k-mooneyoh ok so its the sdk job that givign use oru current test coverage15:51
sean-k-mooneywe probaly shoudl fix that as that pretty light and add new temptet test for this into one of our jobs15:51
sean-k-mooneybut ok that better then nothing15:52
gouthamrUggla: sean-k-mooney: We'll have a test job running against manila repos soon.. https://review.opendev.org/c/openstack/manila-tempest-plugin/+/98488415:56
gouthamri can add that non-voting to nova so we can see things work on the cold migration code changes15:56
sean-k-mooneygouthamr: we can figure out where to run it but we can probaly add it or a subset of its test into one of our exsting jobs15:57
gouthamrack, especially the multinode one, because i don't want to reinvent the setup into a new job elsewhere15:57
sean-k-mooneywell i was thinkign nova-alt-config15:57
* gouthamr can look15:57
sean-k-mooneybut we have a couple of multi node ones we coudl use15:57
sean-k-mooneydo we need ceph?15:58
sean-k-mooneyor will it work with lvm15:58
gouthamryes, either works15:58
sean-k-mooneythen ya we can look at the job run time and eitehr add a new one or extend one of our shorter jobs15:58
gouthamrwe'll cover both in manila land.. but we can pick one for nova15:58
sean-k-mooneynova-live-migration and the ceph version are pretty small15:58
gouthamrack15:59
sean-k-mooneyif the test are short to run and we dont have issues with memory when deploying another serivce we coudl have ti in re then one15:59
sean-k-mooneywe norally try to keep our josb below 2 hours if we can16:00
sean-k-mooneythe live migration jobs are about an hour at the moment16:00
sean-k-mooneyso those woudl be fine. the alt config job is a littel longer but its still below 90 mins typically16:01
sean-k-mooneyspeaking of job lenght we may want os see fi some of the things tested in nova-next shoudl move to some of the other jobs16:03
sean-k-mooneythats our longest job by far and some of those "next" itmes shoud proably be the default by now16:04
gmaanUggla: ack, that was same bug which gibi reported. it is when service is stopped before RPC is completely initialized. I planned to propose the fix but got distracted, I will propose fix today.16:13
gmaanlet me do that before i forget again16:14
Ugglagmaan 👍16:16
gibi:)16:36
*** EugenMayer4401808 is now known as EugenMayer44018016:44
opendevreviewGhanshyam Maan proposed openstack/nova master: Fix AttributeError in Service.stop() during early startup shutdown  https://review.opendev.org/c/openstack/nova/+/99170217:40
opendevreviewRodolfo Alonso proposed openstack/os-vif master: Remove workaround for oslo.i18n ``translate`` method  https://review.opendev.org/c/openstack/os-vif/+/99172018:51
opendevreviewRodolfo Alonso proposed openstack/os-vif master: Make ``PyPyroute2._lookup_interface`` private  https://review.opendev.org/c/openstack/os-vif/+/99129218:52
*** elodilles is now known as elodilles_OoO20:17
gmaangibi: Uggla graceful shutdown bug fix https://review.opendev.org/c/openstack/nova/+/99170220:56
opendevreviewGhanshyam Maan proposed openstack/nova master: Add functional tests for compute service graceful shutdown  https://review.opendev.org/c/openstack/nova/+/99174321:25
gmaansean-k-mooney: ^^ as you asked about the graceful shutdown functional tests. It is easy to create the scenario with event set/wait and timeouts21:26
opendevreviewMerged openstack/os-vif master: Remove workaround for oslo.i18n ``translate`` method  https://review.opendev.org/c/openstack/os-vif/+/99172023:21

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