Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2022-02-08
15:04:02 sean-k-mooney its only 2 lines so i woudl leave it in the driver change
15:04:25 sean-k-mooney chateaulav: so ya if you could make it 3 patches that woudl be nice
15:05:08 chateaulav sean-k-mooney: yeah that can work. is it usually best to create new patches overall and eventually abandon this once all are marged? or have this patch be one of the 3
15:05:10 sean-k-mooney objects/notifications, driver/secheduler/docs and zuul-job/roles
15:05:48 sean-k-mooney am no i woudl keep this one adn split out the rest form it
15:05:54 chateaulav sounds good
15:06:45 sean-k-mooney thanks ill try an review it this week before or after the split
15:08:52 frickler headsup: I just approved devstack switching Fedora from 34 to 35, you will want to check your depending jobs (not sure whether there are still any)
15:09:29 sean-k-mooney frickler: we have an experimental job i think
15:09:39 sean-k-mooney but we do not use fedroa in the gate normally
15:11:20 bauzas reminder : nova meeting in 48 mins here at #openstack-nova
15:12:27 frickler it will use py3.10, so might discover some interesting things like it did for horizon
15:13:20 sean-k-mooney frickler: well we shoudl not need fedora to test that we shoudl be able to run the unit and functets already right
15:13:30 sean-k-mooney but i agree that will break thigns proably
15:13:53 sean-k-mooney frickler: part of the issue is soem of the things depercated in 3.6 are removed in 3.10
15:13:57 frickler I don't think there are other tests on py3.10 yet, still no pkgs for focal last time I checked
15:14:07 sean-k-mooney and 3.11 will similar remove more things
15:26:56 gibi stephenfin: if you are around do you have an insight how to fix this sphinx issue in a good way?
15:27:09 gibi stephenfin: https://zuul.opendev.org/t/openstack/build/6b956593572d4d1c81ac65e46475f24e/log/job-output.txt the nova-specs doc gen fails with duplicet ref
15:27:48 gibi stephenfin: it seems that _`[1]` type of refs are become global
15:28:16 gibi stephenfin: and we have spec docs that are independently using those and causing the doc build to break now
15:28:34 gibi if I have to guess it is caused by this feature from sphinx 4.4.0 https://github.com/sphinx-doc/sphinx/issues/9993
15:31:33 stephenfin gibi: Good find on the RFE. That's a bug IMO
15:31:53 stephenfin Let me open a report. I think I'm a maintainer on Sphinx somehow :-D
15:32:01 gibi stephenfin: thank you :)
15:32:15 gibi I will try to pin back the sphinx version in nova-specs then
15:36:17 gibi stephenfin: I confirmed that with Sphinx 4.3.2 the problem is not reproduceable
15:40:44 opendevreview Balazs Gibizer proposed openstack/nova-specs master: Avoid Sphinx 4.4.0 to fix specs doc build https://review.opendev.org/c/openstack/nova-specs/+/828368
15:45:09 opendevreview Jonathan Race proposed openstack/nova master: object/notification for Adds Pick guest CPU architecture based on host arch in libvirt driver support https://review.opendev.org/c/openstack/nova/+/828369
15:45:57 gmann stephenfin: sean-k-mooney gibi py3.6 is there for Yoga testing but for sure in Z cycle it will be dropped from testing runtime.
15:46:24 gibi gmann: sounds good. thanks for confirming
15:48:14 bauzas gmann: any update so far about the Z naming ?
15:48:26 bauzas I know it's in the hands of the legal folks
15:48:44 gmann bauzas: yeah, TC votes are done. it is under trademark checks from foundation which will be done in 1-2 weeks
15:49:05 bauzas gmann: can we already know what was the TC votes ?
15:49:09 gmann bauzas: may be by next week we might get name. but let's see how fast legal team is ready
15:49:18 gmann bauzas: yeah, it is open
15:49:33 gmann https://civs1.civs.us/cgi-bin/control.pl?id=E_693bd8bbe63ca52f&key=9ec9708dc7e8a814&akey=b13149621594beda
15:49:55 gmann sorry, let me grab result one
15:50:09 gmann https://civs1.civs.us/cgi-bin/results.pl?id=E_693bd8bbe63ca52f
15:50:11 gmann ^^ this one
15:51:37 gmann In the 'show details' button at bottom you can see 'Ballot report' too
15:52:29 bauzas gmann: thanks
15:58:51 opendevreview Jonathan Race proposed openstack/nova master: zuul-job for Adds Pick guest CPU architecture based on host arch in libvirt driver support https://review.opendev.org/c/openstack/nova/+/828372
16:00:42 bauzas #startmeeting nova
16:00:42 opendevmeet Meeting started Tue Feb 8 16:00:42 2022 UTC and is due to finish in 60 minutes. The chair is bauzas. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:42 opendevmeet Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:42 opendevmeet The meeting name has been set to 'nova'
16:00:51 bauzas #link https://wiki.openstack.org/wiki/Meetings/Nova#Agenda_for_next_meeting
16:00:56 gibi o/
16:00:59 dansmith o/
16:01:02 gmann o/
16:01:02 chateaulav \o
16:01:15 bauzas sorry I was late, someone pinged me at house
16:01:23 elodilles o/
16:02:19 bauzas ok, let's start
16:02:25 bauzas #topic Bugs (stuck/critical)
16:02:30 bauzas #info One Critical bug
16:02:35 bauzas #link https://bugs.launchpad.net/nova/+bug/1959899 this is the critical bug for nova-next
16:02:45 bauzas do we want to discuss this one ?
16:03:17 gibi my last info
16:03:34 gibi that artom and sean-k-mooney figured out that this is a q35 machine type related issue
16:03:46 gibi then we turned off the test on master to unblock the gate
16:03:59 gibi I don't know if there is any fix in the works
16:04:16 bauzas oh shit
16:05:20 bauzas ok, so maybe we should put it to High ?
16:05:50 gibi yeah we can decrease it to High as the gate is not blocked now
16:05:52 gibi as we skip
16:05:56 bauzas as this test is executed in other jobs
16:06:06 sean-k-mooney we dont have a fix yet
16:06:21 sean-k-mooney artom speculated it might be related to a state tracking but with qemu
16:06:27 bauzas I'm only afraid of the fact it means we would have problems when upgrading
16:06:31 sean-k-mooney that is still pending
16:07:05 bauzas sean-k-mooney: at least this is blocking q35 to be the next machine type, right ?
16:07:13 artom The only data point I have is that it passed without q35... once
16:07:34 artom But it also passed intermittently *with* q35 on other patches
16:07:35 bauzas ah, this is different then
16:07:39 sean-k-mooney yes so q35 and pc use difffernt hotplug code paths
16:07:42 artom So... more testing needed?
16:07:50 bauzas I guess
16:07:57 bauzas anyway, let's punt the bug report to High
16:08:07 sean-k-mooney bauzas: we als think the way the device is presented in teh guest might change based on q35 or pc
16:08:09 bauzas but we need to fix this bug
16:08:21 sean-k-mooney so we shoudl keep looking into it and try and fix it
16:08:29 sean-k-mooney yep
16:08:37 sean-k-mooney it may not be fixable in nova
16:08:47 sean-k-mooney but we have nto figure that out yet
16:09:02 opendevreview Alexey Stupnikov proposed openstack/nova master: Fix clean-up process for aborted queued live migrations https://review.opendev.org/c/openstack/nova/+/828374
16:09:09 gibi artom: this test runs in tempest-integrated-compute job with pc machine type
16:09:19 gibi artom: and it appears to be green there
16:10:03 bauzas could we create another job for testing ?
16:10:18 bauzas non-voting, of course
16:10:20 sean-k-mooney this is the qemu bug we think might be relevent https://bugzilla.redhat.com/show_bug.cgi?id=2007129
16:11:17 sean-k-mooney bauzas: i htink we have enough test coverage with teh integrated job
16:11:29 gibi agree ^
16:11:31 bauzas OK
16:11:33 sean-k-mooney so i think we can move on for now and just monitor/investigate
16:11:41 bauzas cool then, we can moveo n
16:11:45 bauzas move on, even
16:11:57 bauzas #link https://bugs.launchpad.net/nova/+bugs?search=Search&field.status=New 39 new untriaged bugs (+0 since the last meeting)

Earlier   Later