Friday, 2026-07-03

opendevreviewKamil Sambor proposed openstack/nova master: Add concurrency_backend config option for Nova services  https://review.opendev.org/c/openstack/nova/+/99376007:21
* gibi sends love and support to the gerrit servers 09:13
gibitkajinam: do I understand correctly that if you use stateless firmware (like in SNP) then you cannot reboot the VM from inside the guest? At least I see that the VM stops normally, printing that it will be rebooted but it never starts up. I assume it is connected to the statelessness of things but I can be mistaken09:56
gibithis is the last thing visible in the guest09:56
gibi[   33.971858] reboot: Restarting system09:56
gibihm, qemu logs10:00
gibi2026-07-03T09:53:41.222388Z qemu-kvm: cpus are not resettable, terminating10:00
gibihm it is not the firmware it is SEV-ES / SNP that has this limitation10:07
gibiI'm filing a bug to document it10:07
gibihttps://www.redhat.com/en/blog/confidential-guest-reset-qemu-hypervisor-design-choices-and-approach  10:13
gibiso it is just a qemu limitation that is being fixed10:13
*** ykarel_ is now known as ykarel10:16
opendevreviewKamil Sambor proposed openstack/nova master: Add concurrency_backend config option for Nova services  https://review.opendev.org/c/openstack/nova/+/99376010:21
opendevreviewRodolfo Alonso proposed openstack/os-vif master: ovs: Pass ``multi_queue`` inside ``ifr`` dict for tuntap devices  https://review.opendev.org/c/openstack/os-vif/+/99593310:27
opendevreviewRodolfo Alonso proposed openstack/os-vif master: ovs: Pass ``multi_queue`` inside ``ifr`` dict for tuntap devices  https://review.opendev.org/c/openstack/os-vif/+/99593310:34
ralonsohsean-k-mooney, hi! just if you didn't see the conversation in the neutron channel10:42
ralonsohI created https://bugs.launchpad.net/os-vif/+bug/215902410:42
ralonsohand pushed this: https://review.opendev.org/c/openstack/os-vif/+/99593310:42
tkajinamgibi, oh, thanks for that. I'll check that later11:21
sean-k-mooneyralonsoh: i have not read it back yet but my clinet is there ill reply in the neutorn channel11:22
tkajinamwe likely need to block the operation properly at api layer instead of putting the instance in that hung state ...11:22
gibitkajinam: it does not hung, it stops cleanly. and can be started via the API 11:28
gibibut it cannot be fully restarted from the guest11:28
gibihttps://bugs.launchpad.net/nova/+bug/215902311:29
sean-k-mooneygibi: that os-vif patch above form ralonsoh  is a nice easy one to review if you have time11:29
sean-k-mooneygibi: also thts fun, it sound like perhaps a qemu bug11:30
sean-k-mooneyyou shoudl be able to reboot form inside the guest as that doe not halt the domain11:31
opendevreviewJoan Gilabert proposed openstack/nova master: Report OWNER_NOVA trait on resource providers  https://review.opendev.org/c/openstack/nova/+/99429911:33
opendevreviewJoan Gilabert proposed openstack/nova master: Add libvirt driver MDEV ARQ support for Cyborg  https://review.opendev.org/c/openstack/nova/+/99584011:33
opendevreviewAshish Gupta proposed openstack/placement master: tests: Support shared file-backed SQLite in placement fixtures  https://review.opendev.org/c/openstack/placement/+/99310611:42
opendevreviewAshish Gupta proposed openstack/nova master: tests: use file-backed Placement SQLite in functional threading  https://review.opendev.org/c/openstack/nova/+/99258111:46
opendevreviewAshish Gupta proposed openstack/placement master: tests: Support shared file-backed SQLite in placement fixtures  https://review.opendev.org/c/openstack/placement/+/99310612:06
tkajinamgibi, ahh ok.12:53
tkajinamnow I remember I saw that behavior during my old testing and just documenting that limitation would be enough now12:54
sean-k-mooneytkajinam: it should be pretty proment12:56
sean-k-mooneyits pretty problematic as it a faily core fucntioantliy12:56
sean-k-mooneyit means that you can use normal widnows update flows or just reboot normaly for kernel upgrades ectra12:57
gibisean-k-mooney: it is a qemu limitation that is being lifted in qemu 1112:57
sean-k-mooneyso it defintly need to get reported to qemu or amd as a issue to be adressed not jsut a footnote in our docs12:57
sean-k-mooneygibi: ah cool12:57
gibithey way how qemu did the reset of the vcpu did not work for ES and SEV instances12:57
gibidue to encrypted registers12:57
gibiafaik12:57
sean-k-mooneyint13?12:58
gibihttps://www.redhat.com/en/blog/confidential-guest-reset-qemu-hypervisor-design-choices-and-approach12:58
gibimore details are here ^^12:58
gibiI haven't tested qemu 11 as it is only available in rawhide at the moment and the machine I have does not like fedora :)12:59
sean-k-mooneyah no i was thinking in INT 1912:59
sean-k-mooneyanyway cool13:00
gibi:)13:00
sean-k-mooneyif its going away then ya just add an important:: note in the docs is fine13:01
sean-k-mooneyqemu 11 isnec quite out yet in distos13:02
gibiyeah13:02
gibitoo new13:02
gibihm, I have it in sid, but that would be even harder to put on this lab machine13:03
sean-k-mooneytkajinam: i would ask that we also put this in the release not for the feature if that ok13:03
sean-k-mooneygibi: you can enable the virt preview repo13:03
sean-k-mooneyare you on c10s13:04
sean-k-mooneyhttps://copr.fedorainfracloud.org/coprs/g/virtmaint-sig/virt-preview/13:04
sean-k-mooneyhtat has qemu 2:11.0.0-11 13:04
sean-k-mooneyyou just need to add https://copr.fedorainfracloud.org/coprs/g/virtmaint-sig/virt-preview/repo/centos-stream-10/group_virtmaint-sig-virt-preview-centos-stream-10.repo13:05
sean-k-mooneyto your repo config13:05
gibiooh OK, I can try that. The lab machine is on centos10 yes13:06
sean-k-mooneywith that said i woudl finish any ohter testing you want to do first before doing that13:06
sean-k-mooneyit normally does not breakthing but it can13:06
gibiyeah I'm mostly done with the exploratory testing13:06
sean-k-mooneythere was a patch probaly litrally decades ago to add a flag to enabel the virt preview repo in devstack but i dotn think that ever landed13:08
sean-k-mooneyi remember "dicusssing" that wehn trying to get dpdk suprpot into openstack13:08
*** mhen_ is now known as mhen13:26
opendevreviewTakashi Kajinami proposed openstack/nova master: Return 400 for invalid properties for locked memory  https://review.opendev.org/c/openstack/nova/+/99581313:30
sean-k-mooneytkajinam: locked memory is not ment ot requried hugepages13:33
tdemaretHi core team, I need help to understant, what I must doing to validate my spec and validate my code : my pr is here, and I think, I don t understand something : https://review.opendev.org/c/openstack/nova-specs/+/979696 - https://review.opendev.org/c/openstack/nova/+/979693 ( I see the merge conflict, but it is a 3 times, I think, I need to validate my spec before I fix it ? )13:33
sean-k-mooneytkajinam: unless im conflatin gthat with realtime13:34
sean-k-mooneytkajinam: i know i added that so that vdpa would work13:35
opendevreviewTakashi Kajinami proposed openstack/nova master: Return 400 for invalid properties for locked memory  https://review.opendev.org/c/openstack/nova/+/99581313:35
sean-k-mooneywithout needing realtime13:35
sean-k-mooneybut i tough it was inteded to work with file backed memory and small pages as well13:35
sean-k-mooneyi woudl have to dig back into the history13:35
sean-k-mooneytkajinam: actully no your right https://github.com/openstack/nova-specs/blob/b855fe1b5d192aefa238cd9465fc0ee60568b2d1/specs/zed/implemented/libvirt-viommu-device.rst?plain=1#L6113:38
sean-k-mooneyi intneed it to be set only when hugepages to avodi needign to mark the vm as realtime and foce cpu pinning13:38
tkajinamsean-k-mooney, ah, ok. honestly I found that requirement a bit strange first (because locked memory should work without huge pages afaik) but now it's clearer13:39
sean-k-mooneywe could make it work withough hugepages13:39
tkajinamyeah13:39
sean-k-mooneybut it was to work around a bug in qemu13:39
sean-k-mooneywell actully maybe libvirt, qemu or libvirt did not allcoate addtion locked memroy for pci passhtough or vdpa automaticly13:40
sean-k-mooneyso we had to mark the entire mroy as locked as a workaround13:40
sean-k-mooneylater version fo libvirt do allocate addtioanl locked memeory for passthough devices but we still have the option13:41
tkajinamok13:43
sean-k-mooneyits still usefull in general but lower level then we woudl have prefered to expose13:43
sean-k-mooneyi.e. as an operator you really should not need to think about that when creating flavors13:43
tkajinamI think we could consider lifting the limitation after some quick testings, but IMO we should fix the status code separately in a short term (because 403 doesn't match the actual cause imho)13:44
tkajinamyeah13:44
sean-k-mooneyoh i was about to ask if you got a 40913:44
sean-k-mooneyya 403 is wrong 13:44
sean-k-mooney409 or 400 are fine13:44
tkajinamI'm changing it to 40013:44
sean-k-mooneywell it should be the ame as the other flavor iamge conflicts13:45
tkajinamother flavor image conflicts return 400 now so it's consistent with that behavior13:46
tkajinamtdemaret, I'm unsure what your expect by "validate" but in general we expect that spec is reviewed and approved (you can push actual change for reference) and then actual implementations are reviewed according to the spec13:46
tkajinamtafkamax, I saw several comments still open and unanswered so the first thing you need may be to respond to these.13:47
opendevreviewThibaut Démaret proposed openstack/nova master: libvirt: add disk rotation_rate support for local disks  https://review.opendev.org/c/openstack/nova/+/97969313:47
tkajinamtafkamax, oops. sorry that's a wrong mention13:47
sean-k-mooneytdemaret: yes unforctuetly you spec did not get much visablity before the deadlien beacus you had abandoned it13:47
sean-k-mooneyso it not approved for the current release and as such the code cannot proceed13:48
sean-k-mooneytdemaret: if you want to advance that for teh 2027.1 cycle the first stpe would be to socialise the freature and repoorpse the spec for the next release13:48
opendevreviewTakashi Kajinami proposed openstack/nova master: Validate image properties for stateless firmware at API layer  https://review.opendev.org/c/openstack/nova/+/99583513:49
sean-k-mooneytdemaret: by socailise the feature i mean bring it up in teh  nova irc meeint or the ptg 13:49
sean-k-mooneyoverall the idea is a good one and i think it fits in teh scope of nova13:50
sean-k-mooneybut procedually its too late in teh cycle since nova had its spec freeze a few weeks ago.the default for other project would have been yesterdady as well13:50
sean-k-mooneythe feature design freeze for most teams is milesotne two which was yesterday13:51
tdemaret@sean-k-mooney : Ok it is new for me, I need to socialise to integrate a new spec in the new release. Ok I understand more the workflow. The next PTG is in Oktober ? 13:52
sean-k-mooneyteam are not shifting to completing the already approved work before feature freeze at the end of august, the branches will re open for new feature work for next cycle in mid september13:52
sean-k-mooneytdemaret: yes but you dont need to wait that long13:53
sean-k-mooneywe ussually create the spec directory for the next release a few weeks after milestone 213:53
sean-k-mooneyso you can actully just do that now but i wauld start by bring this up in teh next irc meeting on monday13:54
tdemaretOk and I must introduce my feature at this time when the directory has been created ?13:54
tdemaret@sean-k-mooney : thanks, I am happy for this feedback13:55
sean-k-mooneytdemaret: so  if you want to get visablity it helps, its not stricly requried but unless your already an active contibutor and have reached out to the team its very easy for your spec to go unnotice13:55
ykarelclarkb, melwitt sean-k-mooney i have added summary in https://review.opendev.org/c/zuul/zuul-jobs/+/959393 for the linux bridge/geneve move and current impact and required changes/options to move forward. please do check when you get chance13:56
sean-k-mooneytdemaret: i know the process can seem a bit much for a small change but the realtity is when we add a feature we effectivly need to support that aproximatly forever13:57
opendevreviewTakashi Kajinami proposed openstack/nova master: Validate image properties for stateless firmware at API layer  https://review.opendev.org/c/openstack/nova/+/99583513:58
tdemaretsean-k-mooney: thanks, I try to answer on few question I have, and I understand more the process. Thanks for the manual13:58
opendevreviewTakashi Kajinami proposed openstack/nova master: compute: Validate SEV/SEV-ES support at start up  https://review.opendev.org/c/openstack/nova/+/99434214:01
tdemaretsean-k-mooney: I have one comment, I don t understand, how I can change something in my spec file. I am a little bit confused/unclear about how to apply it, or what is expected. I have the code example, but I don t have the spec file. I just make a check on the comment and I verify my implementation to be sure, it is ok or I must make something in the spec file. Sorry it is14:05
tdemaretreally the newby question... 14:05
sean-k-mooneyyou mena your asking how to update the review for the spec?14:06
tdemaretyes14:08
opendevreviewTakashi Kajinami proposed openstack/nova master: compute: Validate SEV/SEV-ES support at start up  https://review.opendev.org/c/openstack/nova/+/99434214:11
tdemaretReally, I don't understand, what I need to change : https://review.opendev.org/c/openstack/nova-specs/+/979696?tab=comments , really I think, I must change the mindset and read more spec file to understand the template and informations are mandatory.14:14
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Ignore SEV-ES when SEV-SNP is detected  https://review.opendev.org/c/openstack/nova/+/99434314:15
opendevreviewDan Smith proposed openstack/nova stable/2026.1: Fix mutating global config in websocket proxy  https://review.opendev.org/c/openstack/nova/+/99595614:29
opendevreviewDan Smith proposed openstack/nova stable/2025.2: Fix mutating global config in websocket proxy  https://review.opendev.org/c/openstack/nova/+/99595714:30
opendevreviewDan Smith proposed openstack/nova stable/2025.1: Fix mutating global config in websocket proxy  https://review.opendev.org/c/openstack/nova/+/99595814:31
opendevreviewThibaut Démaret proposed openstack/nova master: libvirt: add disk rotation_rate support for local disks  https://review.opendev.org/c/openstack/nova/+/97969314:57
gibisean-k-mooney: I tried qemu 11 from virt preview but it is crashing15:04
gibi2026-07-03T15:00:54.247846Z qemu-system-x86_64: SNP_LAUNCH_UPDATE ret=-5 fw_error=22 'Invalid parameter'15:04
gibi2026-07-03T15:00:54.247920Z qemu-system-x86_64: SEV-SNP: CPUID validation failed for function 0x80000021, index: 0x0, provided: eax:0x18000065, ebx: 0x00000000, ecx: 0x00000006, edx: 0x00000000, expected: eax:0x18000065, ebx: 0x00000000, ecx: 0x00000000, edx: 0x0000000015:04
gibi2026-07-03T15:00:54.247933Z qemu-system-x86_64: SEV-SNP: failed update CPUID page15:04
gibi2026-07-03 15:00:54.454+0000: shutting down, reason=crashed15:04
gibiOK cpu_model_extra_flags = -tsa-sq-no, -tsa-l1-no15:13
gibisolve sthat15:13
gibisolves that15:13
gibiaand reboot works from within the SNP guest now15:14
sean-k-mooneyoh cool15:15
gibiupdated the bug 15:15
gibiand dropping for the weekend15:16
sean-k-mooneyhave you tried cpu-mode=host-passthough15:16
gibio/15:16
gibiit is with host-passthrough15:16
sean-k-mooneyah thats a sperte bug so15:16
sean-k-mooneygibi: enjoy the weekend15:16
gibithanks. same to you too15:17
*** dviroel is now known as dviroel_lunch15:32
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Detect AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova/+/99476416:01
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Detect AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova/+/99476416:17
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Support launchSecurity element for SEV-SNP  https://review.opendev.org/c/openstack/nova/+/99492016:17
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Omit optional fields for SEV/SEV-ES  https://review.opendev.org/c/openstack/nova/+/99492116:17
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Drop sev capability check in launch instance flow  https://review.opendev.org/c/openstack/nova/+/99492916:21
opendevreviewTakashi Kajinami proposed openstack/nova master: AMD SEV: omit iommu='on' for virtio devices  https://review.opendev.org/c/openstack/nova/+/90963516:22
opendevreviewTakashi Kajinami proposed openstack/nova master: libvirt: Launch instances with SEV-SNP memory encryption  https://review.opendev.org/c/openstack/nova/+/99493016:40
opendevreviewTakashi Kajinami proposed openstack/nova master: Update documentations for AMD SEV-SNP support  https://review.opendev.org/c/openstack/nova/+/99509016:40
opendevreviewTakashi Kajinami proposed openstack/nova master: Remove [libvirt] num_memory_encrypted_guests  https://review.opendev.org/c/openstack/nova/+/99512016:41
*** dviroel_lunch is now known as dviroel16:52

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