| *** mhen_ is now known as mhen | 02:26 | |
| *** ykarel__ is now known as ykarel | 03:59 | |
| opendevreview | minwoo seo proposed openstack/nova master: Add support for target_zone in live-migration API https://review.opendev.org/c/openstack/nova/+/976085 | 06:23 |
|---|---|---|
| opendevreview | Kamil Sambor proposed openstack/nova master: Replace eventlet with standard threading in novncproxy https://review.opendev.org/c/openstack/nova/+/976089 | 08:28 |
| gibi | o/ As far as I see the gate is still blocked until we land https://review.opendev.org/c/openstack/openstacksdk/+/976008 | 08:29 |
| gibi | more details are in https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/5HHEYPZA6VIORX2XLBZGNMM2EVX2LR65/ | 08:44 |
| opendevreview | Masanori Kuroha proposed openstack/nova-specs master: Clarify trait removal behavior https://review.opendev.org/c/openstack/nova-specs/+/976092 | 08:45 |
| opendevreview | Masanori Kuroha proposed openstack/nova-specs master: Clarify trait removal behavior https://review.opendev.org/c/openstack/nova-specs/+/976092 | 08:47 |
| mkuroha | gibi: Sorry for the repeated reminders. I’ve also updated the spec to match the code. I’d appreciate it if you could review these patches when you have time. https://review.opendev.org/c/openstack/nova/+/948304 (spec: https://review.opendev.org/c/openstack/nova-specs/+/976092) | 08:52 |
| gibi | mkuroha: I will try but I cannot promise I get to it this week | 08:53 |
| opendevreview | minwoo seo proposed openstack/nova master: Add support for target_zone in live-migration API https://review.opendev.org/c/openstack/nova/+/976085 | 08:54 |
| mkuroha | gibi: Sure, thanks for letting me know. | 08:59 |
| opendevreview | Balazs Gibizer proposed openstack/nova master: Set openstacksdk-functional-devstack non voting https://review.opendev.org/c/openstack/nova/+/976095 | 09:18 |
| gibi | bauzas: sean-k-mooney[m]: ^^ I think due to the multiple blockage in the sdk gate we need to make our sdk job nonvoting for a whiel | 09:18 |
| bauzas | good call, just one question, I guess we can't skip the failing tests ? | 09:19 |
| opendevreview | Balazs Gibizer proposed openstack/nova master: Revert "Set openstacksdk-functional-devstack non voting" https://review.opendev.org/c/openstack/nova/+/976096 | 09:20 |
| gibi | bauzas: I don't know how to do that in our zuul config. Skipping it directly in the sdk repo is not possible due to the multiple breakage in their gate | 09:20 |
| bauzas | my bad, this isn't a tempest job, we can't just configure which tests to skip | 09:23 |
| bauzas | yeah, that's just a tox target which is exposed, there is nothing configurable https://github.com/openstack/openstacksdk/blob/master/zuul.d/functional-jobs.yaml#L51 | 09:27 |
| opendevreview | Taketani Ryo proposed openstack/nova master: mem-enc: stop using _get_mem_encryption_config() for SEV checks https://review.opendev.org/c/openstack/nova/+/967970 | 09:46 |
| opendevreview | Taketani Ryo proposed openstack/nova master: mem-enc: refactor memory encryption trait logic for extensiblity https://review.opendev.org/c/openstack/nova/+/967971 | 09:46 |
| opendevreview | Taketani Ryo proposed openstack/nova master: mem-enc: make RP creation independent of specific encryption models https://review.opendev.org/c/openstack/nova/+/967972 | 09:46 |
| opendevreview | Taketani Ryo proposed openstack/nova master: mem-enc: refactor _guest_configure_mem_encryption() for extensibility https://review.opendev.org/c/openstack/nova/+/967973 | 09:46 |
| opendevreview | Taketani Ryo proposed openstack/nova master: mem-enc: adjust requirement checks for mem_encryption guests https://review.opendev.org/c/openstack/nova/+/967974 | 09:46 |
| opendevreview | Taketani Ryo proposed openstack/nova master: mem-enc: introduce a check between mem_encryption and locked_memory https://review.opendev.org/c/openstack/nova/+/971300 | 09:46 |
| opendevreview | Max proposed openstack/nova master: Add regression test for bug #2137366 https://review.opendev.org/c/openstack/nova/+/972044 | 10:03 |
| opendevreview | Max proposed openstack/nova master: fix: orphaned attachments during migration delete https://review.opendev.org/c/openstack/nova/+/972045 | 10:03 |
| MaxLamprecht[m] | o/ After working more and more with the attachment stuff, I noticed that we always check for the persistent libvirt config after detaching disks from the live config. This doesn't make a functional difference but the warning log statement is really confusing. https://review.opendev.org/c/openstack/nova/+/975651 Would be great if someone could review this | 10:10 |
| MaxLamprecht[m] | I noticed also that the legacy fallback of the libvirt-dev-alias change(merged since 2024.1) is kinda broken. We crash in the fallback path because of an info log statement which uses an uninitialized variable - https://review.opendev.org/c/openstack/nova/+/973103 | 10:15 |
| opendevreview | Max proposed openstack/nova master: Add regression test for bug #2139135 https://review.opendev.org/c/openstack/nova/+/974831 | 10:42 |
| opendevreview | Max proposed openstack/nova master: fix: delete attachments after rescheduling delete https://review.opendev.org/c/openstack/nova/+/974832 | 10:42 |
| *** ubuntu is now known as Guest2023 | 10:52 | |
| *** Guest2023 is now known as kgube | 10:52 | |
| opendevreview | Tobias Urdin proposed openstack/nova master: Add a timeout for `ceph df` command https://review.opendev.org/c/openstack/nova/+/975927 | 11:21 |
| opendevreview | Lajos Katona proposed openstack/nova master: Add regression test to repoduce bug 2140537 https://review.opendev.org/c/openstack/nova/+/975832 | 12:13 |
| opendevreview | Julien LE JEUNE proposed openstack/nova master: Add mechanism to manage snapshot during nc init https://review.opendev.org/c/openstack/nova/+/873062 | 12:30 |
| opendevreview | Julien LE JEUNE proposed openstack/nova master: Add regression test for interrupted snapshot on compute restart https://review.opendev.org/c/openstack/nova/+/976110 | 12:30 |
| opendevreview | Lajos Katona proposed openstack/nova master: WIP: Fix for bug 2140537 https://review.opendev.org/c/openstack/nova/+/975934 | 12:55 |
| opendevreview | minwoo seo proposed openstack/nova master: Add support for target_zone in live-migration API https://review.opendev.org/c/openstack/nova/+/976085 | 12:59 |
| opendevreview | Tobias Urdin proposed openstack/nova master: Add a timeout for `ceph df` command https://review.opendev.org/c/openstack/nova/+/975927 | 13:21 |
| tobias-urdin | sean-k-mooney[m]: ^updated, see if that's better | 13:25 |
| opendevreview | Tobias Urdin proposed openstack/nova master: Add a timeout for `ceph df` command https://review.opendev.org/c/openstack/nova/+/975927 | 13:26 |
| opendevreview | Kamil Sambor proposed openstack/nova master: Rename deadline parameter to more accurate timeout https://review.opendev.org/c/openstack/nova/+/957411 | 14:01 |
| sean-k-mooney | tobias-urdin: sure ill take a look. the setuptools 82 issue is going to have to be resolved to unblock the gate | 14:07 |
| sean-k-mooney | in nova we are likely less impacted as we can just turn off horizon if its not already disabled | 14:07 |
| *** haleyb|out is now known as haleyb | 14:24 | |
| tobias-urdin | sean-k-mooney: ack, thanks! | 14:25 |
| tobias-urdin | let me give it another try, i think i misunderstood what you meant first time | 14:30 |
| opendevreview | Tobias Urdin proposed openstack/nova master: Add a timeout for `ceph df` command https://review.opendev.org/c/openstack/nova/+/975927 | 14:31 |
| sean-k-mooney | my initall feedback was unclear. normally we would not use except Exception but for the reason of avoiding the eventlet import i think that its valid to deviate as long as we have a comment to say why | 14:31 |
| tobias-urdin | sean-k-mooney: i see, sounds good | 14:32 |
| sean-k-mooney | tobias-urdin: just as genera feedback you likely shoudl have some kind of unit test to ensure the timeout is passed but lest see what ci says | 14:33 |
| sean-k-mooney | what is the behvior of update aviabel resouce when we time out with this fix? | 14:35 |
| *** bogdando_ is now known as bogdando | 14:35 | |
| sean-k-mooney | do we just fail and not update teh capsity | 14:35 |
| sean-k-mooney | we are going to raise exception.StorageError instead of hanging so im expecting that to cuase the update_aviable_resouces perodic to abort | 14:37 |
| sean-k-mooney | but the compute agent to still be alive in general | 14:37 |
| sean-k-mooney | which is net better but obviouly if operators dont review the log message then nova will silently be broken if you try to intereact with ceph | 14:38 |
| sean-k-mooney | i.e. snapshots and new vm creation will fail but adding or removign a prot will work. | 14:39 |
| tobias-urdin | now that you mention it, i get a vague memory that's why i've backed off patching it before, that i didn't know what the best outcome was; broken and hung forever or started broken | 14:44 |
| tobias-urdin | not such a big deal for new compute nodes, as we start them with service record disabled, and on existing, well, if it's broken it's broken and has to be found like any similar problems | 14:46 |
| tobias-urdin | (i.e we have enable_new_services=false configured, not that it's the default) | 14:46 |
| sean-k-mooney | well we could also have nova shutdown seperatly at least if we are getting a exception we have the opertuntiy do define a more correct behavior | 14:53 |
| sean-k-mooney | so it might not close the but but its a partial fix that moves us one step closer | 14:53 |
| nicolairuckel | sean-k-mooney: Do you know when you would have time to test my NVRAM patch? | 14:55 |
| sean-k-mooney | nicolairuckel: let me do it now. i need to look at come watcher and cybrog stuff so my nova capastcity for the rest fo the cycle is limited | 14:56 |
| sean-k-mooney | nicolairuckel: last time i looked it was in merge conflict | 14:56 |
| sean-k-mooney | have you updated it | 14:56 |
| nicolairuckel | I resolved some merge conflicts at some point but let me check if there are new ones | 14:56 |
| sean-k-mooney | https://review.opendev.org/c/openstack/nova/+/959682 | 14:57 |
| sean-k-mooney | it needs to be rebased again | 14:57 |
| sean-k-mooney | it went into merge conflcit about a week ago | 14:57 |
| sean-k-mooney | i do have a env where i can test this so if you update it today ill pull it in an test it out and add my +2 | 14:58 |
| sean-k-mooney | i was morely happy with it the last time i looked so i think its ready to go | 14:59 |
| *** ykarel_ is now known as ykarel | 15:00 | |
| *** sambork_ is now known as sambork | 15:02 | |
| nicolairuckel | okay, let me check how to resolve that | 15:02 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 15:15 |
| nicolairuckel | I broke something while rebasing so i need to look into that first. I'm not sure if I will be able to manage that today. Depends on when your today ends. | 15:25 |
| sean-k-mooney | ack not for about anyter 4-5 hours | 15:33 |
| sean-k-mooney | maybe 3-4 | 15:33 |
| sean-k-mooney | it depend on how much i get done today | 15:33 |
| Uggla | Reminder: nova upstream meeting in ~20mn | 15:38 |
| nicolairuckel | Is there a way to get notifications when there are merge conflicts? That's the second or third time I completely missed that. | 15:40 |
| sean-k-mooney | nicolairuckel: not that im aware of | 15:46 |
| sean-k-mooney | you might be able to enabel it in gerrit but for many of us that would be too verbose | 15:47 |
| sean-k-mooney | you can detect it in the api so you could write a simple bash script with curl ot list your reviews that are in merge conflcit or create your own dash baord in gerrit to show that | 15:48 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 15:53 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 15:59 |
| Uggla | #startmeeting nova | 16:02 |
| opendevmeet | Meeting started Mon Feb 9 16:02:42 2026 UTC and is due to finish in 60 minutes. The chair is Uggla. Information about MeetBot at http://wiki.debian.org/MeetBot. | 16:02 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 16:02 |
| opendevmeet | The meeting name has been set to 'nova' | 16:02 |
| Uggla | Hello everyone | 16:02 |
| lajoskatona | o/ | 16:02 |
| bauzas | o/ | 16:02 |
| elodilles | o/ | 16:02 |
| obre | o/ | 16:02 |
| gibi | o/ | 16:02 |
| tkajinam | o/ | 16:03 |
| MaxLamprecht[m] | o/ | 16:03 |
| gmaan | o/ | 16:03 |
| Uggla | Let's go. | 16:03 |
| sambork | o/ | 16:03 |
| Uggla | #topic Bugs (stuck/critical) | 16:04 |
| Uggla | #info No Critical bug | 16:04 |
| fwiesel | o/ | 16:04 |
| Uggla | #topic Gate status | 16:04 |
| Uggla | #link https://bugs.launchpad.net/nova/+bugs?field.tag=gate-failure Nova gate bugs | 16:04 |
| Uggla | #link https://etherpad.opendev.org/p/nova-ci-failures-minimal | 16:04 |
| 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 status | 16:04 |
| Uggla | #info Please look at the gate failures and file a bug report with the gate-failure tag. | 16:05 |
| tkajinam | I think many of you are aware of this, but multiple jobs are broken by setuptools 82.0.0 which removed pkg_resources | 16:05 |
| Uggla | #info Please try to provide a meaningful comment when you recheck | 16:05 |
| gibi | tkajinam: yepp | 16:05 |
| tkajinam | one typical one is devstack job, which fails to install horizon, due to xstatic packages using pkg_resources for namespace packages | 16:05 |
| gibi | and openstacksdk job also broken | 16:05 |
| tkajinam | yeah | 16:05 |
| gmaan | also, sdk job. do we want to wait for sdk test fix merge or stop job running on nova gate until it is fixed? | 16:05 |
| gmaan | in devstack we need to stop running it to land setuptools fixes | 16:06 |
| gibi | I proposed to make sdk job non voting here https://review.opendev.org/c/openstack/nova/+/976095 | 16:06 |
| gibi | but it fails on setuptools :D | 16:06 |
| sean-k-mooney | for nova we can turn off horizon as a workaround | 16:06 |
| gmaan | ohk, i see now. | 16:06 |
| sean-k-mooney | we should not need it in any of our jobs | 16:06 |
| sean-k-mooney | even the vnc tests dont use it | 16:06 |
| gmaan | let me review the devstack and grenade fixes | 16:07 |
| sean-k-mooney | horizon shoudl be the only thing useing the xstatic packages that are impacated | 16:07 |
| elodilles | yepp, grenades are failing, but there are some fixes on the gate: https://review.opendev.org/q/Icf07fa246b3532f939d20e3cbea32cc0b7b439ed | 16:07 |
| gmaan | sean-k-mooney: I saw you msg about neutron does not use microversion but this is backward imcomtible change and can break user. | 16:07 |
| gmaan | anyways we do not need to discuss it here but just my observartion | 16:07 |
| elodilles | though they might fail due to the horizon issue as well :/ | 16:08 |
| tkajinam | yeah that's a good first try, though there can be legacy packages which still use pkg_resources | 16:08 |
| gibi | the devstask / grenade fixes needs to land revers order 2024.2 landed 2025.1 going in | 16:08 |
| sean-k-mooney | gmaan: it is but they dont provide stable api garentees | 16:08 |
| gmaan | gibi: yeah | 16:08 |
| sean-k-mooney | at least not to the same degree that service with microversions do | 16:08 |
| tkajinam | gmaan, yeah the update indicates tenant_id field in API response will be replaced :-( | 16:08 |
| sean-k-mooney | so im not sayign they should keep it | 16:08 |
| sean-k-mooney | tenant-id | 16:08 |
| gmaan | that is unnecessary breaking change without any gain/value | 16:09 |
| sean-k-mooney | but just that they dont make the same commit ment to api stablity | 16:09 |
| sean-k-mooney | i don not disagree :) | 16:09 |
| gmaan | anyways, will check on devstack/greande one during meeting and I am +w on gibi change to make sdk job non voting | 16:09 |
| gibi | also as an extra neutron with oslo.service master is failing to start services in 50% of the time it does not directly effect us as we run with release oslo version but local devstacks are affected if using oslo from git | 16:09 |
| gibi | what a Monday | 16:10 |
| tkajinam | gibi, I guess it's because of non-serializable service instance, I guess ? | 16:10 |
| elodilles | :) | 16:10 |
| gibi | tkajinam: yepp | 16:10 |
| tkajinam | ok | 16:10 |
| elodilles | gibi: at least we are not at the final release rush right now ;) | 16:10 |
| gibi | https://bugs.launchpad.net/neutron/+bug/2141304 | 16:10 |
| gmaan | gibi: tkajinam due to recent change of spawn things or it is other? | 16:11 |
| gibi | gmaan: probably the spawn one | 16:11 |
| tkajinam | gmaan, yeah that's caused by recent update to use spawn-safe context | 16:11 |
| gibi | but I have no confirmation yet | 16:11 |
| gmaan | but that is not released yet, | 16:11 |
| gmaan | I will check that, I think fallback should work there to fork | 16:11 |
| gibi | yes, hence not affecting our gate, but affecting local devstacks with olso.service from git | 16:11 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 16:11 |
| gmaan | gibi: ack | 16:12 |
| tkajinam | gmaan, https://review.opendev.org/c/openstack/oslo.service/+/965208 | 16:12 |
| gmaan | yeah | 16:12 |
| tkajinam | it's interesting that we didn't catch that error in the devstack job in oslo.service, which should apply that change | 16:12 |
| tkajinam | but that's yet another topic | 16:12 |
| gibi | if you want more, then the openstacksdk fix for the neutron change also blocked on a manila failre too https://review.opendev.org/c/openstack/openstacksdk/+/976008 | 16:13 |
| gmaan | yeah, its hard to run all type of configuration there but I will add some jobs in periodic so that we can notice | 16:13 |
| gibi | not just on the setuptools one | 16:13 |
| gmaan | gibi: tkajinam I will propose the oslo.service fix soon | 16:14 |
| tkajinam | gmaan, thanks | 16:14 |
| gmaan | gibi: yes, what a Monday :) | 16:14 |
| tkajinam | though that oslo.service fix may be blocked by the setuptools mess ... | 16:14 |
| gmaan | ack | 16:15 |
| gmaan | I think it shjould be ok as we run only master tempest-full job but I will check | 16:15 |
| tkajinam | I thought some of the master jobs are also impacted but I could be wrong | 16:16 |
| Uggla | ok thx tkajinam. | 16:18 |
| Uggla | anything else ? or we can move on ? | 16:18 |
| tkajinam | I think we can move on. please watch CI status carefully before triggering mass recheck | 16:19 |
| Uggla | tkajinam ++ | 16:20 |
| Uggla | #topic Release Planning | 16:20 |
| Uggla | #link https://releases.openstack.org/gazpacho/schedule.html | 16:20 |
| Uggla | #info Nova deadlines are set in the above schedule | 16:20 |
| Uggla | #info 3 weeks before feature freeze | 16:21 |
| Uggla | #info PTG etherpad for 2026.1 is available: https://etherpad.opendev.org/p/nova-2026.1-ptg | 16:21 |
| Uggla | #topic Review priorities | 16:21 |
| Uggla | #link https://etherpad.opendev.org/p/nova-2026.1-status | 16:21 |
| Uggla | #info I have updated Launchpad and the above doc. Please ping me if you spot something missing. | 16:21 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 16:21 |
| Uggla | btw I have also introduced the bugs we discussed last time. More or less at the bottom of the doc. | 16:22 |
| Uggla | #topic OpenAPI | 16:22 |
| Uggla | #link: https://review.opendev.org/q/topic:%22openapi%22+(project:openstack/nova+OR+project:openstack/placement)+-status:merged+-status:abandoned | 16:22 |
| Uggla | #info still 11 remaining atm. | 16:23 |
| sean-k-mooney | i think they were just blocked by ci issue mostly | 16:23 |
| sean-k-mooney | but it woudl be nice to land them this/next week | 16:23 |
| sean-k-mooney | before we get to FF | 16:23 |
| Uggla | I think so too. | 16:23 |
| Uggla | sean-k-mooney ++ | 16:23 |
| * gibi__ lost network connectivity | 16:23 | |
| elodilles | gibi: still Monday? :-o | 16:24 |
| gibi__ | Seem so very Monday | 16:24 |
| Uggla | #topic Stable Branches | 16:24 |
| * Uggla giving the mic to elodilles | 16:24 | |
| elodilles | thx Uggla | 16:25 |
| elodilles | #info stable gates are impacted with the "No module named 'pkg_resources'" issue | 16:25 |
| elodilles | so the same stuff that we discussed previously | 16:25 |
| elodilles | fixes are already proposed by stephenfin: https://review.opendev.org/q/Icf07fa246b3532f939d20e3cbea32cc0b7b439ed | 16:25 |
| elodilles | #info stable branch status / gate failures tracking etherpad: https://etherpad.opendev.org/p/nova-stable-branch-ci | 16:25 |
| elodilles | that's all from me | 16:25 |
| elodilles | Uggla: back to you | 16:25 |
| Uggla | thx elodilles | 16:26 |
| Uggla | #topic vmwareapi 3rd-party CI efforts Highlights | 16:26 |
| * Uggla giving the mic to fwiesel | 16:26 | |
| fwiesel | #info CI tries to run change with dependencies including `Depends-On` | 16:27 |
| fwiesel | So, I changed the CI that it will parse the commit message, and if you have changes that depend on one another (i.e. oslo.vmware) they can be tested together | 16:27 |
| fwiesel | #info CI will run the tests with eventlet disabled in the nova-compute process, when the topic is eventlet-removal | 16:28 |
| fwiesel | I hope, that should make it easier to test that and not having me in the loop. | 16:29 |
| gibi__ | Thanks | 16:29 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 16:29 |
| fwiesel | Also, I think we have a general regression in the instance creation, specifically the image upload. I haven't looked into it in detail, but the tests seem to fail even without eventlet disable with a timeout on the image transfer into vsphere | 16:30 |
| fwiesel | I'll look into that next | 16:30 |
| fwiesel | That would it be from my side. | 16:30 |
| fwiesel | Uggla: back to you, unless there are questions | 16:31 |
| Uggla | super fwiesel thx for the hard work | 16:31 |
| Uggla | #topic Gibi's news about eventlet removal | 16:32 |
| gibi__ | o/ | 16:32 |
| * Uggla giving the mic to gibi | 16:32 | |
| gibi__ | So the compute deadlock bug showed by the vmware testing (and reproduced with libvirt) needs decision on fix direction | 16:33 |
| gibi__ | Alternatives with pocs are in https://etherpad.opendev.org/p/eventlet-executor-semaphore-limit | 16:34 |
| gibi__ | I also saw Kamil pushed a WIP patch for novncproxy | 16:35 |
| gibi__ | But I havent looked at it yet | 16:35 |
| gibi__ | That is it from me | 16:35 |
| * gibi__ mic drop | 16:35 | |
| Uggla | #topic Nova using openstack sdk for neutron | 16:36 |
| * Uggla giving the mic to lajoskatona if he wants to share something | 16:37 | |
| lajoskatona | o/ thanks | 16:37 |
| lajoskatona | Yes, I updated the 1st patch (https://review.opendev.org/c/openstack/nova/+/928022 ) based on gibi's comments | 16:37 |
| lajoskatona | that's it basically, thanks for the review :-) | 16:38 |
| gibi__ | Thanks Lajos | 16:38 |
| Uggla | thanks lajoskatona | 16:38 |
| Uggla | btw I have started reviewing the series, but I have not pushed comments yet. Probably tomorrow | 16:39 |
| lajoskatona | Uggla: thanks | 16:39 |
| Uggla | #topic Open discussion | 16:40 |
| Uggla | First topic from obre, about AMX support | 16:40 |
| obre | Yeah | 16:40 |
| Uggla | obre, please go ahead | 16:40 |
| obre | In Nova there is already functionality where compute-nodes are reporting their cpu capabilities as standard traits. These traits can then be used in flavors so that certain flavors provides a more feature-rich IA than other flavors. When that flavor is scheduled the compute-driver picks the first CPU-model in a list of models that fullfills all the traits from the flavor. And | 16:40 |
| obre | all this functionality works great (imho). | 16:40 |
| obre | A couple of years ago Intel added a feature called AMX, which the libvirt-driver in nova currently does not recognize as a feature to report as a trait. I would like to expand the list of features being reported by the driver to include the amx-instructions. | 16:40 |
| obre | And as this technically is not new functionality I would kindly ask if it is possible to think of it as a wishlist-bug, and thus target the G-release. I expect to be able to both implement the changes and test it in our test-environments within this week; if that helps in deciding if its OK or not :) | 16:40 |
| *** bogdando_ is now known as bogdando | 16:42 | |
| obre | Do you have any thoughts? | 16:42 |
| sean-k-mooney | to me this si very small and could be a wishlist bug | 16:43 |
| Uggla | just something unclear, libvirt can detect it ? | 16:44 |
| sean-k-mooney | but that only makes sense if the trait is added to os-traits before the non clinet lib freeze | 16:44 |
| sean-k-mooney | Uggla: yes but nova has a map form the livbirt feature to os-trait traits | 16:44 |
| obre | Uggla: Libvirt has support for amx. | 16:44 |
| sean-k-mooney | and we only report them if we have a vlue in that mapping | 16:44 |
| sean-k-mooney | so we need a new trait in os-trati and an update to that map | 16:45 |
| sean-k-mooney | then it will "just work" | 16:45 |
| tkajinam | I wonder if specless-blueprint is preferred to a bug but that may be not very important | 16:45 |
| sean-k-mooney | provided your livbirt is new enought to detech and reprot the cpu flag | 16:45 |
| gibi__ | Lets go for it then | 16:45 |
| tkajinam | sean-k-mooney, so just extend CPU_TRAINTS_MAPPING in nova/virt/libvirt/utils.py , right ? | 16:45 |
| sean-k-mooney | tkajinam: specless bluepring and spec share the same freeze period | 16:45 |
| sean-k-mooney | tkajinam: exactly | 16:46 |
| tkajinam | ah ok | 16:46 |
| sean-k-mooney | im not against granting an expction but its also very late in the cylce for that | 16:46 |
| sean-k-mooney | we also dont backprot traits | 16:46 |
| sean-k-mooney | that why i rased the non-client lib freeze which i think is this or next week | 16:46 |
| obre | I believe its next week. | 16:47 |
| Uggla | yep | 16:47 |
| Uggla | next week | 16:47 |
| elodilles | next week, yepp | 16:47 |
| sean-k-mooney | im for adding this capablity jsu tnot sure we shoudl rush to do it this week | 16:47 |
| sean-k-mooney | so +1 in general +0 to doing it this week if other are fine with it then so am i | 16:48 |
| Uggla | +1 in general too. | 16:48 |
| obre | I guess Im not voting; but getting it this week would mean a year earlier into production for us; as the G-release is a SLURP. :) | 16:48 |
| gibi__ | Propose the patches, as they are expected to be small Im happy to review them | 16:49 |
| elodilles | obre: exactly | 16:49 |
| sean-k-mooney | well non slurps are fully supproted but if your downstream will only deploy slurps that understandable | 16:49 |
| sean-k-mooney | sound like we are tentitvly ok with trying to land it since its small | 16:50 |
| sean-k-mooney | so only queston is how to track it | 16:50 |
| Uggla | that's sound not very large, so assuming you can provide them and we can review them. I would be ok. | 16:50 |
| sean-k-mooney | the os-traits patch exits but needs a rebase i think | 16:50 |
| Uggla | Though, no promises. | 16:50 |
| sean-k-mooney | https://review.opendev.org/c/openstack/os-traits/+/868149 | 16:51 |
| opendevreview | Lin Yang proposed openstack/os-traits master: CPU: add traits for new X86 feature "AMX" https://review.opendev.org/c/openstack/os-traits/+/868149 | 16:51 |
| Uggla | ogre, is that ok for you ? | 16:51 |
| sean-k-mooney | jsut rebased it but assuming that passes all that is missing is an os-traits releease and the nova patch | 16:51 |
| obre | As for tracking, there is already a specless bp formulated: https://blueprints.launchpad.net/nova/+spec/add-amx-traits - Alternatively Ill guess Ill file a bug? | 16:52 |
| obre | Uggla: Im able to look at the rebase, and write the patch. | 16:52 |
| obre | (Unless my wife gives birth this week; which will get priority. But the due-date there is end of feburary :)) | 16:52 |
| Uggla | :) I guess we understand the priority here. :) | 16:53 |
| * gibi__ +2d the os-traits patch | 16:53 | |
| obre | <3 | 16:53 |
| gibi__ | https://review.opendev.org/c/openstack/os-traits/+/868149 | 16:54 |
| sean-k-mooney | ill +2w it if it passes ci | 16:54 |
| sean-k-mooney | obre: we will need to do a release and you will need to bump the min version in your nova patch | 16:55 |
| sean-k-mooney | that not hard to do but the upstream gates are kind fo broken for a number of issues that came up in teh last few days | 16:55 |
| obre | Right. So then Ill hope and pray that that works; and then write the patch? :) | 16:56 |
| sean-k-mooney | so im not sure what state teh requriements repo is in | 16:56 |
| sean-k-mooney | you can write the patch now and depen on the os-trait one | 16:56 |
| sean-k-mooney | the unit tests will fail but the tempest oens should pass | 16:56 |
| sean-k-mooney | you will just need to update it with the min verion bump after the release | 16:57 |
| obre | Ok! Then Ill do that! | 16:57 |
| Uggla | are we good with that topic ? | 16:57 |
| obre | Im happy at least :) | 16:57 |
| sean-k-mooney | yep | 16:58 |
| Uggla | There was another topic from fwiesel, but he kindly accept to defer it to next week because I have a hard stop at the top of the hour. Thx fwiesel | 16:59 |
| Uggla | so time to close except we have a last minute quick info to share. | 16:59 |
| gibi__ | Thanks Uggla | 17:00 |
| Uggla | seems not Thanks for joining this meeting. Have a nice day/evening and see you next week. | 17:00 |
| fwiesel | Thanks everyone | 17:00 |
| Uggla | #endmeeting | 17:00 |
| opendevmeet | Meeting ended Mon Feb 9 17:00:27 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 17:00 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/nova/2026/nova.2026-02-09-16.02.html | 17:00 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/nova/2026/nova.2026-02-09-16.02.txt | 17:00 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/nova/2026/nova.2026-02-09-16.02.log.html | 17:00 |
| lajoskatona | Thanks, Bye | 17:00 |
| bauzas | thanks Uggla | 17:00 |
| obre | Thanks all! | 17:00 |
| elodilles | thanks o/ | 17:00 |
| Uggla | time to go to sprint review... ;) | 17:01 |
| opendevreview | Lajos Katona proposed openstack/nova master: WIP: Fix for bug 2140537 https://review.opendev.org/c/openstack/nova/+/975934 | 17:01 |
| opendevreview | Merged openstack/os-traits master: CPU: add traits for new X86 feature "AMX" https://review.opendev.org/c/openstack/os-traits/+/868149 | 17:34 |
| sean-k-mooney | obre: ^ :) | 17:52 |
| obre | sean-k-mooney: Nice :) | 17:59 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 19:21 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 19:42 |
| opendevreview | Nicolai Ruckel proposed openstack/nova master: Preserve UEFI NVRAM variable store https://review.opendev.org/c/openstack/nova/+/959682 | 19:49 |
| opendevreview | sean mooney proposed openstack/nova master: Add regression test for unified limits resource bug https://review.opendev.org/c/openstack/nova/+/975859 | 22:37 |
| opendevreview | sean mooney proposed openstack/nova master: Fix unified limits to include all resource types https://review.opendev.org/c/openstack/nova/+/975872 | 22:37 |
| *** jonnyb3 is now known as jonnyb | 22:50 | |
| *** erlon2 is now known as erlon | 23:04 | |
| opendevreview | minwoo seo proposed openstack/nova master: Add target_zone support for migration and live-migration https://review.opendev.org/c/openstack/nova/+/976085 | 23:52 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!