| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2025-09-16 | |||
| 16:33:55 | Uggla | Thanks for joining this meeting, a bit shorter than usual. See you next week. | |
| 16:34:05 | Uggla | #endmeeting | |
| 16:34:05 | opendevmeet | Meeting ended Tue Sep 16 16:34:05 2025 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | |
| 16:34:05 | opendevmeet | Minutes: https://meetings.opendev.org/meetings/nova/2025/nova.2025-09-16-16.00.html | |
| 16:34:05 | opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/nova/2025/nova.2025-09-16-16.00.txt | |
| 16:34:05 | opendevmeet | Log: https://meetings.opendev.org/meetings/nova/2025/nova.2025-09-16-16.00.log.html | |
| 16:34:26 | r-taketn | Thank you | |
| 16:34:34 | bauzas | thanks ! | |
| 16:34:36 | elodilles | thanks o/ | |
| 16:34:38 | Uggla | tkajinam no I was not but received a slack from my manager at the same time. ;) | |
| 16:35:04 | tkajinam | Uggla, :-D | |
| 16:37:57 | opendevreview | Takashi Kajinami proposed openstack/nova master: Drop explicit setting of iommu for AMD SEV support https://review.opendev.org/c/openstack/nova/+/909635 | |
| 17:01:25 | opendevreview | Takashi Kajinami proposed openstack/nova master: AMD SEV: Stop explicitly enabling emulated iommu for virtio devices https://review.opendev.org/c/openstack/nova/+/909635 | |
| 17:14:16 | opendevreview | Takashi Kajinami proposed openstack/nova master: AMD SEV: Stop explicitly enabling emulated iommu for virtio devices https://review.opendev.org/c/openstack/nova/+/909635 | |
| 17:22:11 | opendevreview | Takashi Kajinami proposed openstack/nova master: AMD SEV: omit iommu='on' for virtio devices https://review.opendev.org/c/openstack/nova/+/909635 | |
| #openstack-nova - 2025-09-17 | |||
| 02:42:39 | opendevreview | Takashi Kajinami proposed openstack/nova master: AMD SEV: omit iommu='on' for virtio devices https://review.opendev.org/c/openstack/nova/+/909635 | |
| 07:31:56 | jlejeune | gibi: hello, can i have review for https://review.opendev.org/c/openstack/nova/+/960349 and https://review.opendev.org/c/openstack/nova/+/901655 please ? :) | |
| 08:44:01 | ianwatsonit | Hey nova team I have a few PRs that are pending review wondering if anyone would be able to take a look? | |
| 08:44:24 | ianwatsonit | https://review.opendev.org/c/openstack/nova/+/955690 | https://review.opendev.org/c/openstack/nova/+/955686 | https://review.opendev.org/c/openstack/nova/+/955687 | https://review.opendev.org/c/openstack/nova/+/955685 | |
| 09:35:56 | opendevreview | Kamil Sambor proposed openstack/nova master: Switch nova-conductor to use ThreadPoolExecutor https://review.opendev.org/c/openstack/nova/+/957088 | |
| 10:13:16 | opendevreview | Merged openstack/nova stable/2025.1: libvirt: Disable VMCoreInfo device for SEV-encrypted instances https://review.opendev.org/c/openstack/nova/+/958874 | |
| 10:41:54 | opendevreview | Jonas Schäfer proposed openstack/nova master: Preserve vTPM state between power off and power on https://review.opendev.org/c/openstack/nova/+/955657 | |
| 10:57:53 | opendevreview | Stephen Finucane proposed openstack/nova master: api: Correct version for rebuild response https://review.opendev.org/c/openstack/nova/+/958862 | |
| 10:57:53 | opendevreview | Stephen Finucane proposed openstack/nova master: api: Remove unnecessary action method prefix https://review.opendev.org/c/openstack/nova/+/958863 | |
| 10:57:54 | opendevreview | Stephen Finucane proposed openstack/nova master: api: Deprecate os-volumes_boot API https://review.opendev.org/c/openstack/nova/+/958864 | |
| 10:59:16 | opendevreview | Stephen Finucane proposed openstack/nova master: api: Deprecate os-volumes_boot API https://review.opendev.org/c/openstack/nova/+/958864 | |
| 11:00:46 | opendevreview | Merged openstack/nova master: Adds regression test for bug LP#2044235 https://review.opendev.org/c/openstack/nova/+/960349 | |
| 11:10:58 | opendevreview | Stephen Finucane proposed openstack/nova master: api: Remove unnecessary action method prefix https://review.opendev.org/c/openstack/nova/+/958863 | |
| 11:10:58 | opendevreview | Stephen Finucane proposed openstack/nova master: api: Deprecate os-volumes_boot API https://review.opendev.org/c/openstack/nova/+/958864 | |
| 12:23:48 | gibi | can I get some stable reviews on https://review.opendev.org/q/topic:%22bug/2115905%22 ? | |
| 12:24:52 | opendevreview | Takashi Kajinami proposed openstack/nova stable/2024.2: libvirt: Disable VMCoreInfo device for SEV-encrypted instances https://review.opendev.org/c/openstack/nova/+/959913 | |
| 12:25:44 | opendevreview | Takashi Kajinami proposed openstack/nova stable/2024.1: libvirt: Disable VMCoreInfo device for SEV-encrypted instances https://review.opendev.org/c/openstack/nova/+/959914 | |
| 15:00:26 | MichielPiscaer[m] | Hi, I'm searching for some documentation on about live-migration? I would like to know what decision tree it uses to determine the destination host when I let "Openstack" choose the host. | |
| 15:05:06 | sean-k-mooney | MichielPiscaer[m]: so there isnt a descion tree as such | |
| 15:05:42 | sean-k-mooney | the saem proceess that is used to select a hsot when intially creat the vm is performed by the schdluer with the only excption being the current hsot is excluded | |
| 15:06:20 | sean-k-mooney | MichielPiscaer[m]: so https://docs.openstack.org/nova/latest/admin/scheduling.html is the correct entry point in our docs to review | |
| 15:07:01 | sean-k-mooney | effectivly when you ask nova to migrate a vm, the api calls the conductor which then calls the schdluer to select a destination | |
| 15:07:55 | sean-k-mooney | the schduler use the info in teh reqeust spec form the rpc call to first call the placement service for a set of allcaiton candate based on reqcoue and trait request computeed form the flavor and image | |
| 15:08:24 | sean-k-mooney | the the filters run over the candiate host aplying difent policies liek numa affinity | |
| 15:08:46 | sean-k-mooney | finally teh weigher sort the restulant list of hosts | |
| 15:09:05 | sean-k-mooney | with default config we will then take the first host form the list | |
| 15:09:23 | sean-k-mooney | but you can apply a ranadomn suffel over the top n candiates via config | |
| 15:09:45 | sean-k-mooney | when you pass a host its mostly the same wrokflow the only real change is we only consier that host as a candiate | |
| 15:10:39 | sean-k-mooney | MichielPiscaer[m]: so there is no finite descison tree baked in in code. it depend on the filter, weghers adn cofnig option you define | |
| 15:29:14 | opendevreview | Takashi Kajinami proposed openstack/nova stable/2024.1: libvirt: Disable VMCoreInfo device for SEV-encrypted instances https://review.opendev.org/c/openstack/nova/+/959914 | |
| 15:31:38 | MichielPiscaer[m] | Thanks for the answer | |
| 15:52:31 | opendevreview | Merged openstack/nova stable/2025.1: hypervisors: Optimize uptime retrieval for better performance https://review.opendev.org/c/openstack/nova/+/960409 | |
| 21:10:03 | opendevreview | Merged openstack/nova stable/2024.2: libvirt: Disable VMCoreInfo device for SEV-encrypted instances https://review.opendev.org/c/openstack/nova/+/959913 | |
| 22:28:18 | opendevreview | Merged openstack/nova master: nova-conductor puts instance in error state https://review.opendev.org/c/openstack/nova/+/901655 | |
| #openstack-nova - 2025-09-18 | |||
| 00:45:50 | opendevreview | Merged openstack/nova stable/2024.2: hypervisors: Optimize uptime retrieval for better performance https://review.opendev.org/c/openstack/nova/+/960410 | |
| 01:08:47 | opendevreview | melanie witt proposed openstack/nova-specs master: Re-propose vTPM live migration https://review.opendev.org/c/openstack/nova-specs/+/961564 | |
| 06:29:45 | opendevreview | yangjianfeng proposed openstack/nova master: Support reusing existing port binding during instance live-migration https://review.opendev.org/c/openstack/nova/+/961593 | |
| 06:50:52 | opendevreview | yangjianfeng proposed openstack/nova master: Support reusing existing port binding during instance live-migration https://review.opendev.org/c/openstack/nova/+/961593 | |
| 07:42:10 | opendevreview | Julien LE JEUNE proposed openstack/nova stable/2025.1: Adds regression test for bug LP#2044235 https://review.opendev.org/c/openstack/nova/+/961601 | |
| 07:43:48 | opendevreview | Julien LE JEUNE proposed openstack/nova stable/2024.2: Adds regression test for bug LP#2044235 https://review.opendev.org/c/openstack/nova/+/961602 | |
| 07:47:10 | opendevreview | Julien LE JEUNE proposed openstack/nova stable/2024.1: Adds regression test for bug LP#2044235 https://review.opendev.org/c/openstack/nova/+/961603 | |
| 08:49:39 | opendevreview | Merged openstack/nova stable/2024.1: hypervisors: Optimize uptime retrieval for better performance https://review.opendev.org/c/openstack/nova/+/960411 | |
| 10:12:52 | opendevreview | Sylvain Bauza proposed openstack/nova master: Regression test for ImagePropsWeigher due to missing context https://review.opendev.org/c/openstack/nova/+/961615 | |
| 11:57:34 | opendevreview | Pavlo Shchelokovskyy proposed openstack/nova master: Mention socket in pci numa affinity extra spec help https://review.opendev.org/c/openstack/nova/+/961634 | |
| 13:23:21 | opendevreview | Josephine Seifert proposed openstack/nova stable/2024.2: Add hw_mem_encryption_model image property https://review.opendev.org/c/openstack/nova/+/961655 | |
| 14:03:55 | bauzas | dansmith: ping, if you are free of meetings :) | |
| 14:04:01 | dansmith | bauzas: just a sec | |
| 14:04:04 | bauzas | cool | |
| 14:05:58 | dansmith | bauzas: so, fill_faults() and fill_metadata() are operations on the list object, just like .save() on an instance or anything else | |
| 14:06:09 | dansmith | they expect to use the context that the object maintains | |
| 14:06:31 | bauzas | I see, those are not getters | |
| 14:06:33 | dansmith | since you're sort of yanking InstanceList into existence (and not getting one as the result of a DB query), you don't have a context set on the object itself | |
| 14:06:49 | dansmith | it's not wrong, it's just not the intended workflow | |
| 14:07:07 | dansmith | so just create the list like insts = InstanceList(context, ...) and then it will have it internally | |
| 14:07:27 | bauzas | sure, but then I wonder which context to pass | |
| 14:07:34 | dansmith | but you need to use an admin context, not the one for the current boot request, since that one won't be able to read all the instances you're going to survey | |
| 14:07:44 | bauzas | context for others, we speak about https://bugs.launchpad.net/nova/+bug/2125052 | |
| 14:08:12 | bauzas | yup, I figured it, so I probably need to instantiate a fully new admin context | |
| 14:08:38 | bauzas | we have the user request context stored in the requestspec, but as you said, that can't work | |
| 14:09:11 | bauzas | okay, I think I have everything I need | |
| 14:09:18 | bauzas | I'll write the bugfix | |
| 14:09:45 | dansmith | yup | |
| 14:10:07 | dansmith | I also definitely think it makes sense to have this enabled in nova-next, if not more of the weighers | |
| 14:10:17 | bauzas | dansmith: fancy question, is it better to hydrate a list of instances by passing the context in InstanceList creator or do exactly liek https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/servers.py#L339 ? | |
| 14:10:21 | dansmith | even if they're simple, if we're not running them we're not testing them | |
| 14:10:51 | dansmith | pass it on create, IIRC.. that other usage is likely very old (or misguided) | |
| 14:11:47 | bauzas | dansmith: I don't disagree but if so, we need to find the right weigher configs in order to have predictable and easy-to-understand scheduling | |
| 14:12:48 | dansmith | I surely hope that the weighers won't change the behavior in a way that makes the tests not work, especially on single-node jobs :) | |
| 14:31:55 | opendevreview | Sylvain Bauza proposed openstack/nova master: Regression test for ImagePropsWeigher due to missing context https://review.opendev.org/c/openstack/nova/+/961615 | |
| 14:31:55 | opendevreview | Sylvain Bauza proposed openstack/nova master: Add admin context when filling metadata on ImagePropsWeigher https://review.opendev.org/c/openstack/nova/+/961673 | |
| 14:47:11 | bauzas | gibi: dansmith: given you still have context about the bug (oh oh oh, pun intended), reviews would be appreciated before we all forget https://review.opendev.org/c/openstack/nova/+/961673 | |
| 14:47:36 | dansmith | already did? | |
| 14:53:51 | bauzas | well, that's a one-liner addition, right? | |
| 14:54:14 | bauzas | the regression suite was written this morning | |
| 14:54:32 | dansmith | what? | |
| 14:56:37 | bauzas | the functional test I meant | |
| 14:57:04 | dansmith | I'm not sure what you're saying | |
| 14:57:24 | dansmith | I already +2d the regression and left a comment on the other one | |
| 15:03:27 | bauzas | let me rephrase: my morning, I investigated the cause of the bug and I tried to reproduce it using a functional test, which I uploaded | |
| 15:03:58 | bauzas | that's why it only took me 10 mins to upload the fix itself | |
| 15:04:19 | bauzas | (I was just replying to to 'already did ?') | |
| 15:04:27 | dansmith | weren't you asking for review? | |
| 15:04:43 | dansmith | that's what I was responding to: "already did...review" | |
| 15:09:06 | bauzas | oh sorry about the misunderstanding then :) | |
| 15:18:30 | gibi | I'm +2 on the first patch and the second looks good too I will +2 once Dan's comment is fixed | |