| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2026-08-31 | |||
| 16:15:22 | elodilles | add here ^^^ | |
| 16:17:06 | Uggla | thx elodilles | |
| 16:17:40 | Uggla | skipping vmware topic fwiesel is still on PTO I guess | |
| 16:18:39 | Uggla | #topic Kamil's news about eventlet removal | |
| 16:19:09 | Uggla | sambork is not available today. But he left this info about eventlet: btw only update regarding eventlet is perf testing, so we're done there.Scaling down to 9 scheduler/conductor pods gave native threading a lower, peak-free, steady-state memory footprint (still marginally above eventlet), while CPU usage and timing stayed on par with eventlet | |
| 16:19:10 | Uggla | and occasionally beat it thanks to native low-level thread scheduling. Setting executor_thread_pool_size to 32 for both the scheduler and conductor kept the queue depth at ~1, making those two the only default config changes required | |
| 16:19:38 | gibi | + Last week we worked on stabilizing the functional-py313-threading job | |
| 16:20:17 | Uggla | oh gibi, I thought you were not part of the meeting. My bad. | |
| 16:20:35 | tkajinam | I like hearing such "the new model works better than the old model" news :-) | |
| 16:20:51 | gibi | I think we eventually need to reopen https://bugs.launchpad.net/nova/+bug/2165149 as it might still happne | |
| 16:21:03 | gibi | also I saw TIMED_OUTs from this job | |
| 16:21:08 | gibi | that needs to be investiaget | |
| 16:21:11 | gibi | investigated | |
| 16:21:31 | Uggla | tkajinam ++ | |
| 16:21:31 | tkajinam | gibi, is that what you discussed with sean today ? | |
| 16:22:17 | gibi | yepp | |
| 16:22:27 | gibi | or mostly we discussed the DB locked issue | |
| 16:22:32 | gibi | not the TIMED_OUT one | |
| 16:22:37 | tkajinam | ok | |
| 16:22:46 | gibi | I have some local work on the former | |
| 16:22:59 | gibi | but it is hard to reproduce the problem lcoally | |
| 16:23:02 | gibi | locally | |
| 16:24:36 | Uggla | anything else ? | |
| 16:25:11 | Uggla | #topic Confidential computing, status of SNP and TDX features | |
| 16:25:32 | Uggla | Not sure there are new stuffs on that topic. | |
| 16:25:54 | antia | one question regarding TDX | |
| 16:26:01 | antia | Should I make a patch for Glance? | |
| 16:26:18 | tkajinam | antia, oh yes we have to update metadef | |
| 16:26:32 | tkajinam | antia, I think you can find my change to add sev-snp and follow it | |
| 16:26:48 | tkajinam | or I can do that if you want | |
| 16:26:59 | antia | I can fix it after this | |
| 16:27:03 | tkajinam | https://review.opendev.org/c/openstack/glance/+/995179 | |
| 16:27:19 | antia | great! | |
| 16:27:33 | tkajinam | you don't have to care a new release note file but modify the file added by that change | |
| 16:28:12 | tkajinam | no news about sev-snp either. these features are in master so please try these and give your feedback :-) | |
| 16:30:35 | Uggla | ok can we move on ? | |
| 16:31:05 | Uggla | #topic Nova using openstack sdk for neutron | |
| 16:31:18 | Uggla | lajoskatona something you'd like to sharE ? | |
| 16:31:30 | lajoskatona | small things: SDK release is out | |
| 16:31:49 | lajoskatona | so I can update the patch series to have fields=[....] like filtering with SDK also | |
| 16:32:05 | lajoskatona | after the meeting I hope I can push some of the updated patches | |
| 16:32:13 | lajoskatona | that's it for this topic from me | |
| 16:32:22 | Uggla | thx lajoskatona | |
| 16:32:51 | Uggla | #topic Bug scrubbing | |
| 16:32:59 | Uggla | #info down to 68 (-1). | |
| 16:33:05 | Uggla | #link https://etherpad.opendev.org/p/nova-bug-triage-roster | |
| 16:33:11 | Uggla | #link https://truc.uggla.fr/ to follow the trend. | |
| 16:33:17 | Uggla | #info I need to skip this week. | |
| 16:33:29 | Uggla | #info Next meeting (next week): [public] Upstream bug triage. Wednesday, Septembre 09th · 15:30 – 16:00 UTC. Video call link: meet.google.com/zjr-rxus-hzj | |
| 16:33:37 | Uggla | #topic Open discussion | |
| 16:33:45 | gibi | o/ one quick question | |
| 16:33:50 | Uggla | no topic in the agenda for today. | |
| 16:34:01 | Uggla | gibi please go ahead | |
| 16:34:02 | gibi | when you said 18:13 < Uggla> virtiofs cold migration needs focus too. I focused on testing it last week. | |
| 16:34:10 | gibi | what do you mean? | |
| 16:34:33 | gibi | is there some post FF stuff to land there? | |
| 16:34:43 | dansmith | an FFE was requested | |
| 16:35:02 | dansmith | I thought that it was mostly good to go though after Uggla found and gouthamr fixed (right?) a ceph secret thing | |
| 16:35:21 | gibi | ohh i missed the FFE | |
| 16:35:23 | dansmith | still trying to catch up on the traffic since I was out but that was what I saw | |
| 16:36:08 | sean-k-mooney | not sure if the micvoverison issue has been adressed. tl;dr we shoudl not have one for the migration as there is no api chagne | |
| 16:36:21 | Uggla | yes absolutely a strange behavior with cepfs and migration that happens if you migrate vm depending of the order they created the secret. | |
| 16:36:24 | dansmith | I commented on that | |
| 16:36:34 | dansmith | I assume if we decide to drop it then it'll be trivial to do that | |
| 16:36:55 | dansmith | so I wasn't really including it in "needs focus" :) | |
| 16:37:25 | elodilles | (the FFE mail: https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/GIYUU732R5VB6BNQONWPTHR6QF5S2FVT/ ) | |
| 16:37:35 | sean-k-mooney | ack, i have been off since tursday evening so didnt follow up im back tomorrow | |
| 16:38:09 | sean-k-mooney | that was my only real concern | |
| 16:38:23 | gibi | cool, thanks for the summaries. :) | |
| 16:39:27 | Uggla | I'm testing it again, to ensure the bug I found last week is corrected. | |
| 16:40:54 | Uggla | I have also a couple of things remaining to test. But currently the feature looks quite solid. | |
| 16:42:15 | Uggla | anything else you'd like to discuss ? | |
| 16:43:48 | Uggla | if not I think we are done. | |
| 16:44:17 | Uggla | Thanks for joining this meeting. Have a nice day/evening. | |
| 16:44:30 | Uggla | #endmeeting | |
| 16:44:30 | opendevmeet | Meeting ended Mon Aug 31 16:44:30 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | |
| 16:44:30 | opendevmeet | Minutes: https://meetings.opendev.org/meetings/nova/2026/nova.2026-08-31-16.02.html | |
| 16:44:30 | opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/nova/2026/nova.2026-08-31-16.02.txt | |
| 16:44:30 | opendevmeet | Log: https://meetings.opendev.org/meetings/nova/2026/nova.2026-08-31-16.02.log.html | |
| 16:44:32 | tkajinam | thanks. | |
| 16:44:33 | elodilles | thanks Uggla o/ | |
| 16:44:40 | gibi | thanks Uggla | |
| 16:45:43 | lajoskatona | o/ | |
| 16:49:54 | tkajinam | just in case you have a few moments there are a few libvirt driver cleanups with one +2 . https://review.opendev.org/c/openstack/nova/+/997818 https://review.opendev.org/c/openstack/nova/+/995160 and https://review.opendev.org/c/openstack/nova/+/997878 | |
| 16:54:18 | gmaan | gibi: this is the further refactoring (as dansmith commented on the original change) of thread pool executors. dansmith has initial review on that, waiting for you to have a look and I can update it further | |
| 16:57:49 | gibi | gmaan: which one? :) | |
| 16:58:22 | gibi | I guess https://review.opendev.org/c/openstack/nova/+/1001572 | |
| 16:58:33 | gmaan | gibi: yeah this one :) https://review.opendev.org/c/openstack/nova/+/1001572 | |
| 16:58:35 | gibi | I will try to get to it tomorrow | |
| 16:58:36 | gibi | :) | |
| 16:58:39 | gmaan | thanks | |
| 16:59:51 | gmaan | monday morning, need more coffee | |
| 17:43:20 | gouthamr | dansmith: welcome back | |
| 17:47:34 | gouthamr | sean-k-mooney: dansmith : ack on the review comments; I’ll get to them as soon as I get to my laptop | |
| 18:12:02 | gmaan | sean-k-mooney: as you pinged me on this, I think returning 400 is more suitable here, returning 409 need microversion https://review.opendev.org/c/openstack/nova/+/946223 | |
| 18:12:07 | gmaan | i commented on it | |
| 18:27:22 | sean-k-mooney | i would not really agree that all action apis are indivgual interfaces | |
| 18:27:35 | sean-k-mooney | but i dont have a stong perfence | |
| 18:28:18 | sean-k-mooney | its a single endpoing form my persecitie and the return code shoudl not really vary based on the request body once the request is valid | |
| 18:28:37 | sean-k-mooney | but ya either is fine if you prefer the 400 that works for me | |