| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2026-04-16 | |||
| 14:34:27 | stephenfin | sean-k-mooney: the dev-env doc references fake compute. That's not a thing any more, right? | |
| 14:34:37 | sean-k-mooney | yes it is | |
| 14:34:51 | sean-k-mooney | we are using them to test teh eventlet removeal with rally | |
| 14:35:11 | stephenfin | Oh, okay. I thought the fake driver was only used for functional tests | |
| 14:35:24 | sean-k-mooney | nope it can be used as a "real" driver | |
| 14:35:48 | sean-k-mooney | typeicly without networkign but there is technially a fake neturon driver too | |
| 14:36:40 | sean-k-mooney | https://github.com/openstack/devstack/blob/master/lib/nova_plugins/hypervisor-fake still works more or less | |
| 14:41:29 | stephenfin | TIL | |
| 14:58:07 | opendevreview | Stephen Finucane proposed openstack/nova master: docs: Update development-environment guide https://review.opendev.org/c/openstack/nova/+/984956 | |
| 14:58:13 | stephenfin | bauzas: sean-k-mooney: ^ | |
| 14:58:31 | stephenfin | dansmith: you might tell me if the (minimal) macOS instructions are correct (I've no way to verify) | |
| 14:59:57 | dansmith | stephenfin: ack, joining a call but will look async | |
| 15:00:04 | stephenfin | ty 🙏 | |
| 16:31:09 | Uggla | gibi, fyi regarding bug https://bugs.launchpad.net/nova/+bug/2145783 found in the spec : While it would possible to support defining different resource_class names for different VFs under the same parent PF, this is considered bad practice and unnecessary complexity. Such configuration will be rejected. | |
| 16:31:10 | Uggla | https://specs.openstack.org/openstack/nova-specs/specs/zed/approved/pci-device-tracking-in-placement.html | |
| 16:45:20 | gibi | Uggla: that sounds good at least it was outscope due to complexity not due to infeasibility | |
| 16:46:17 | gibi | the bad pratice holds for NICs but it seems it does not hold for GPUs | |
| 16:46:34 | gibi | so I'm OK to get a new spec that adds this capability | |
| 16:46:34 | Uggla | gibi, yes we might revisit that as to me it makes sens for GPUs (was not the case at that moment). | |
| 16:48:13 | Uggla | gibi, I'd like first an answer from the reporter. Maybe he does not need it. | |
| 16:48:54 | gibi | yeah make sense | |
| 16:49:43 | Uggla | gibi, do you think it is something to share at the PTG ? (assuming I'll find a slot) | |
| 16:51:53 | opendevreview | Stephen Finucane proposed openstack/nova master: docs: Update development-environment guide https://review.opendev.org/c/openstack/nova/+/984956 | |
| 16:53:40 | opendevreview | Stephen Finucane proposed openstack/nova master: Fix functional tests and mypy on macOS https://review.opendev.org/c/openstack/nova/+/937727 | |
| 16:53:42 | gibi | Uggla: Why not. Put it at the end of the Agenda like "best effort topic if we have time" | |
| 16:54:14 | gibi | and either we touch it next week or not. If not then we can bring it up on the weekly meeting when the spec is raised | |
| 16:54:27 | Uggla | gibi 👍 | |
| 16:54:55 | dansmith | stephenfin: oh I was just running unit tests anyway, which is why I don't see those | |
| 16:55:43 | stephenfin | Ah, okay. I wonder why the author of the patch didn't see (and fix) those other issues. Anyway, not related as you say | |
| 16:56:40 | dansmith | well, one of mine is because libcrypt is not available (which also fails on oslo).. I assume that should be a skip though, I just haven't cared enough since it's one failure and easy to ignore | |
| 16:58:21 | dansmith | I usually run unit/functional in my fedora vm, I just like to be able to run targeted single-test things natively sometimes so I don't really care or notice that a few tests fail for whatever reason | |
| 17:00:09 | dansmith | stephenfin: the other failure is " AttributeError: module 'socket' has no attribute 'AF_PACKET'" | |
| 17:00:17 | dansmith | which I assume is because AF_PACKET is a linux/netlink thing that isn't portable | |
| 17:00:24 | dansmith | so again a skip seems reasonable | |
| 17:02:48 | dansmith | (not netlink, just not POSIX) | |
| 17:02:49 | stephenfin | looks like it https://github.com/python/typeshed/blob/bd18cc640cba4dadc51ac9a5157b79ff6adf5de3/stdlib/_socket.pyi#L478 | |
| 17:03:00 | dansmith | yah | |
| 17:08:02 | opendevreview | Dan Smith proposed openstack/nova master: Skip some unit tests on macOS https://review.opendev.org/c/openstack/nova/+/985014 | |
| 17:08:11 | dansmith | stephenfin: ^ | |
| 17:08:50 | stephenfin | +2 | |
| 17:09:15 | opendevreview | Merged openstack/nova master: Reproduce bug #2108974: keypairs lost during cross-cell resize https://review.opendev.org/c/openstack/nova/+/981557 | |
| 20:07:44 | sean-k-mooney | Uggla: you can defien diffent resouce class names per VF today | |
| 20:08:01 | sean-k-mooney | you jsut have to use the address to specify it | |
| 20:08:38 | sean-k-mooney | i.e. in the devspec you can match on the vf adress and spcify which resouce class to report | |
| 20:09:05 | sean-k-mooney | you can use the fact that the adress can have bash globs or a regex to apply that to sets of VFs too | |
| 20:09:48 | sean-k-mooney | or at least you shoudl be able to unless we scoped that out when we did the implemeiont | |
| 20:10:53 | sean-k-mooney | https://docs.openstack.org/nova/latest/configuration/config.html#pci.device_spec | |
| 20:11:04 | sean-k-mooney | the config options were desigend to enabeld that usecase | |
| 20:11:22 | sean-k-mooney | but i dont know if we are properly handling this in rhe resouce tracker | |
| 20:11:50 | sean-k-mooney | we shoudl be but the pci stat pools need to prperly handel that grouping | |
| 20:12:40 | sean-k-mooney | it hilgy likely we never tested it and it broken as a result but its totally expressable today vai our config options | |
| 20:15:05 | sean-k-mooney | Uggla: i have not had time to add pci grouping to the nova ptg but if we wanted to talk about different VF in the same pf having diffent rsouce classes for gpus usecases ti woudl make sense to talk about it toghetehr with nova pci grouping | |
| 20:15:16 | sean-k-mooney | i had ot parly implemnt that for that feature anyway | |
| 20:15:37 | sean-k-mooney | since i want to supprot creating groups fo VFs from teh same or diffent PFs | |
| 20:19:04 | sean-k-mooney | i was mainly testing with takign 64 VF that i was subdeviding into groups | |
| 20:19:52 | sean-k-mooney | a groups of 8 and groups of 4 would have 2 diffent resocues classes even if they are effectivly the same device | |
| 20:21:06 | sean-k-mooney | as long as all host that report CUSTOM_PCI_GROUP_NICX8 provide the same VF type for that i.e. so they are interchangable form the schdluer point of view | |
| 20:21:45 | sean-k-mooney | its ok to report other VF form the same card with a differnt type. | |
| 20:21:58 | sean-k-mooney | *differnt resouce class. | |
| 20:23:40 | sean-k-mooney | if you had a gpu that you partioned into 3 Mig device 1 that used half the card an 2 that used 1/4 each | |
| 20:24:20 | sean-k-mooney | the correct way to model that in placment via nova ro cyborg would be 1 RP for the PF with an inveotyr of 2 of one resocue class and 1 of a diffent one | |
| 20:26:36 | sean-k-mooney | being very hand wavy that is the eventual plan for how to do that in cybrog but not right away. | |
| 21:22:11 | opendevreview | Merged openstack/nova master: Fix keypairs lost during cross-cell resize https://review.opendev.org/c/openstack/nova/+/981558 | |
| 21:22:28 | opendevreview | Merged openstack/nova master: Skip some unit tests on macOS https://review.opendev.org/c/openstack/nova/+/985014 | |
| #openstack-nova - 2026-04-17 | |||
| 00:27:47 | opendevreview | ChungWon Lee proposed openstack/nova master: Add regression test for bug #2134375 https://review.opendev.org/c/openstack/nova/+/984900 | |
| 02:58:11 | opendevreview | Sam Morrison proposed openstack/nova master: Filter out deleted instances when looking for build timouts https://review.opendev.org/c/openstack/nova/+/880125 | |
| 06:57:50 | opendevreview | chanyeol yoon proposed openstack/nova master: Fix infinite WARNING loop in _reclaim_queued_deletes https://review.opendev.org/c/openstack/nova/+/985068 | |
| 07:55:13 | opendevreview | Kamil Sambor proposed openstack/nova master: Enable threading mode for proxy services https://review.opendev.org/c/openstack/nova/+/976089 | |
| 09:31:01 | opendevreview | Kamil Sambor proposed openstack/nova master: Test nova CLI commands with native threading https://review.opendev.org/c/openstack/nova/+/984036 | |
| 09:43:38 | Uggla | sean-k-mooney, as discussed I will prepare the pci grouping topic because that's an internal high prio topic. | |
| 09:44:38 | sean-k-mooney | thanks i have not really had tiem to propsoe it bvut that partly becasue i dont really knwo what if any the open questiosn are | |
| 09:44:53 | sean-k-mooney | the spec has obvioulsy been up for a while and its jsut a refinement of the previoulsy appoved spec | |
| 09:45:35 | sean-k-mooney | so i dont really know what need ptg discussions vs what just need gerrit reivew | |
| 09:45:55 | bauzas | Uggla: sean-k-mooney: IMHO we should be discussing this in the cyborg-nova x-p session | |
| 09:46:02 | sean-k-mooney | nope | |
| 09:46:16 | sean-k-mooney | thi is not a cyborg related topic | |
| 09:46:50 | sean-k-mooney | this is a pure nova proposal and even if cyborg may have simialr functionaltiy in the future it is indepent of that | |
| 09:46:59 | gibi | it is hard to say what are the open question before loading context on the PCI grouping topic. I'm doing that loading now | |
| 09:47:16 | sean-k-mooney | gibi: thanks ++ | |
| 09:48:06 | sean-k-mooney | this is what im suggesting for the nova-cybrog topics https://etherpad.opendev.org/p/cyborg-2026.2-ptg#L294 | |
| 09:48:15 | sean-k-mooney | but honestly i feel like that already too much | |
| 09:48:33 | gibi | Uggla: on a related note, I found hints in the pci/stats code that also assumes that VFs under the same parent are equivalent https://github.com/openstack/nova/blob/525158521281cf643a9942522788efd15ba12d6b/nova/pci/stats.py#L164-L169 so this code need to be looked at when we want to remove the restriction of one RC for all VFs under the same PF | |
| 09:49:10 | Uggla | sean-k-mooney, Could you send me the link to the spec if you have it handy? | |
| 09:49:22 | sean-k-mooney | sure ill go grab it | |
| 09:49:57 | sean-k-mooney | https://review.opendev.org/c/openstack/nova-specs/+/975635 | |
| 09:50:17 | Uggla | I agree the cyborg content seems already too large. So definitively need a session for PCI grouping. | |
| 09:50:26 | sean-k-mooney | weired in my brian i pushed that in like december not febuary | |
| 09:51:06 | sean-k-mooney | i guess because i sated on it just before i went on pto and then picked it up agan breifly in late jan early febuary | |
| 09:51:19 | Uggla | sean-k-mooney thx | |
| 09:52:00 | sean-k-mooney | the other spec https://review.opendev.org/c/openstack/nova-specs/+/967515 will be taken over by joan thaat is the first cyborg sepcifc spec for nova for this cycle | |
| 09:52:32 | sean-k-mooney | im not sure if we will have another | |
| 09:52:44 | sean-k-mooney | it kind of depend on which feature/integration we priorites | |
| 09:55:29 | Uggla | Also, I’m likely lacking some knowledge on PCI grouping, but I’m counting on all of you to help me out. <3 | |
| 09:57:43 | bauzas | stephenfin: thanks a lot for your proposals ! Sorry I preferred to have some bootstrapped files but sure I'll review your patch and accept it once I doublecheck about the differences between the markdown file and the restructured one | |
| 09:57:44 | sean-k-mooney | in essence its just a way to say "provision these togeher and schdule them as a group" that the core idea, today you can have man pci alias request in a single flavor but they alwas provide you 1 device per aplis, pci grouping is just a way to say thses specific set of device are member of a group and i would like 1 of that type of group in a concise way | |
| 09:58:14 | sean-k-mooney | everything beyond that and the usecase that enabels are the details but that is the 10,000 foot view | |
| 09:59:59 | stephenfin | bauzas: Thanks. Can I get core on that too, please? I think sean-k-mooney is after it also | |
| 10:00:31 | sean-k-mooney | yes i would like to activly be inovled in the development fo that repo | |
| 10:00:34 | stephenfin | or better, just add various `*-core` projects for now, which we can extend with non-cores (ops, other end-users) over time | |
| 10:00:50 | stephenfin | *`*-core` groups from projects | |