Wednesday, 2026-01-21

*** mhen_ is now known as mhen02:46
opendevreviewmengxiangzhi proposed openstack/cinder master: Support zte vstorage driver  https://review.opendev.org/c/openstack/cinder/+/96920705:56
opendevreviewHarsh Ailani proposed openstack/cinder master: [IBM SVf Driver] Add basic support for Policy-Based Replication (PBR)  https://review.opendev.org/c/openstack/cinder/+/96544209:05
opendevreviewNimesh Desai proposed openstack/cinder master: [IBM SVf Driver] Define meaningful names for error constants  https://review.opendev.org/c/openstack/cinder/+/95968312:02
rosmaitaCinder meeting in #openstack-cinder at 1400 UTC14:02
rosmaitajungleboyj rosmaita smcginnis tosky whoami-rajat m5z e0ne geguileo eharney jbernard hemna fabiooliveira yuval tobias-urdin adiare happystacker dosaboy hillpd msaravan sp-bmilanov14:02
rosmaitaLuzi sfernand simondodsley  zaubea nileshthathagar flelain wizardbit agalica lutimura kaisers14:02
Luzio/14:02
simondodsleyo/14:02
harsho/14:02
rosmaita#startmeeting cinder14:02
opendevmeetMeeting started Wed Jan 21 14:02:43 2026 UTC and is due to finish in 60 minutes.  The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot.14:02
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.14:02
opendevmeetThe meeting name has been set to 'cinder'14:02
erlon\o14:02
sp-bmilanovo/14:02
mheno/14:02
VolodymyrBoiko[m]o/14:02
RanjithRethinamo/14:03
sfernandhi14:04
rosmaitahello everyone14:04
rosmaitajbernard can't be here today, but i will do the best i can14:04
simondodsleyI'm sure you will crush it...14:05
rosmaitawell, i will crush something14:05
rosmaitaok, let's get started14:06
rosmaita#topic announcements14:06
rosmaitawe had a Festival of Reviews this past Friday14:06
rosmaitaattendance was good, so thanks to everyone who attended14:06
rosmaitaand people who didn't, the next one will be coming up the first week of February14:06
rosmaita#link https://meetings.opendev.org/#Cinder_Festival_of_Reviews14:07
rosmaita^^ contains a ICS file so you can get it on your calendar and never miss it14:08
rosmaitain other news14:09
rosmaitai think the gates are mostly behaving (probably should not have said that out loud)14:09
rosmaitathere were some issues with the python 3.13 jobs being moved to debian-trixie nodes, but i think those are mostly sorted out14:09
rosmaitawe still don't have python 3.14 jobs running, but those are optional14:10
rosmaitai need to redo my patch for those14:11
rosmaitathe one thing i can share is that even when the jobs run, they fail ... we have some py 3.14 incompatible code14:11
rosmaitabut we don't need to address that today14:11
rosmaitaand finally, i think Jon said last week that the midcycle will be the last week of February?14:12
rosmaitaand, i think he said first week14:12
simondodsleylet's have 214:13
rosmaita:D14:13
rosmaitawell, looks like February ... i will remind him to send something out to the ML14:14
rosmaita#action jbernard send email about midcycle to the ML14:14
rosmaitathat's all i've got ... anyone have any announcements?14:14
erlonrosmaita: did you get any of the conversation regarding why they're moving the jobs from ubuntu?14:14
simondodsleydoes this mean that 3rd party CIs should move to debian as well?14:15
kaiserso/ (sry, late)14:15
rosmaitagimme a sec to figure out an answer to both questions14:16
rosmaitaok, so at the beginning of the cycle, the TC decides what stable distros are considered "supported"14:17
rosmaita#link https://governance.openstack.org/tc/reference/runtimes/2026.1.html14:17
rosmaitagazpacho ^^14:17
rosmaitaso we have Ubuntu 24.04 and Debian 1314:17
rosmaitai believe the default python on 24.04 is 3.1214:17
rosmaitaand the default on Debian 13 is 3.1314:17
rosmaitaso, it's easier to run the 3.13 jobs on Debian because the python you want is "right there" without having to do anything14:18
erlonAhh, right makes sense14:18
rosmaitayeah, so we are still running ubuntu jobs for the other python versions we are supporting in gazpacho14:19
rosmaitaso we are still testing with both ubuntu and debian14:20
simondodsleyif we move to debain does that fix the novnc packaging bug 14:20
rosmaitaand it's optional to test with centos stream 10 or rocky linux 1014:20
simondodsleyhttps://bugs.launchpad.net/neutron/+bug/210959214:21
yuvalI want to bring up this patch which we discussed last week: https://review.opendev.org/c/openstack/os-brick/+/95759914:22
simondodsleyThe devstack fix for this bug does not sit well in SoftwareFactory and has broken our CI (unless someone using SF has figured out a fix)14:22
rosmaitasimondodsley: maybe? zigo is really good about keeping stuff updated, so i imagine that debian is distributing the fixed novnc package14:22
yuvalI updated the commit14:22
zigoI'm for gating on Python 3.14... :)14:23
zigoWhat is it about NoVNC ?14:23
simondodsleySee the bug above14:23
rosmaitazigo: i'm working on it (the py 3.14 gating): https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/96851014:23
zigorosmaita: Great, thanks!14:24
rosmaitaok, anything else before we move on?14:24
zigoFYI, I remember I fixed novnc.14:24
zigonovnc (1:1.6.0-2) unstable; urgency=medium14:25
zigo  * Kill python3-novnc: novnc doesn't have any Python things for a long time14:25
zigo    already, it's just javascript.14:25
zigo -- Thomas Goirand <zigo@debian.org>  Wed, 30 Apr 2025 14:51:43 +020014:25
zigoUbuntu may not have the fix though.14:25
simondodsleyso if i switch our SF CI to use Debian we may be able to come back to life???14:26
zigoIndeed, it's not in 24.04, only in Trixie.14:26
zigosimondodsley: At least, novnc 1:1.6.0-2 in Trixie has the issue adressed, and no python3-novnc package.14:26
simondodsleyzigo: thanks.14:27
simondodsleyDoes anyone have a nodepool `diskimages` definiton for a trixie image that will work for devstack?14:27
zigorosmaita: One more thing: I'm receiving no review on my patch: https://review.opendev.org/c/openstack/cinder/+/97171014:28
zigoWe're using it in production already, and we're very happy of it...14:28
rosmaitasimondodsley: i think ask in #openstack-infra, they have debian-trixie nodes available, not sure where the definitions are, though14:29
rosmaitazigo: noted14:30
rosmaitaok, moving on ... 2 topics: encryption regression and continuation of os-brick config opts discussion14:30
rosmaita#topic image encryption14:30
rosmaitamhen: that's you14:30
mhenplease have a look at the last two sections of https://etherpad.opendev.org/p/2026-01-image-encryption-client-regression14:31
mhenyou can skip the lengthy preface at the top for now, its just context14:31
mhenit's quite minor and easy to fix I think just want to know which way of going about it would be preferred when making the client patch14:33
RanjithRethinamCan someone pls review/approve this PR, https://review.opendev.org/c/openstack/cinder/+/971739 all review comments are addressed and CI is passing14:35
rosmaitathanks, mhen.  i was just looking at the cinder API for this: https://docs.openstack.org/api-ref/block-storage/v3/#upload-volume-to-image14:35
rosmaitajust wanted to verify that this is an osc issue, not something we (cinder) caused14:36
mhencontainer_format is optional as per API but the client *always* sends it with an opinionated default although unnecessary14:36
rosmaitathe container_format is optional, so i would vote for osc not sending anything for container_format if the user has not specified one14:36
rosmaitait's a fine line between being opinionated and stupid!14:37
rosmaitawhile you're fixing that, maybe you could make sure they're not sending opinionated content for other unspecified options?14:38
rosmaitathough i guess that should be a separate patch14:38
mhenyes, for this at least we have a reason (the change introduced by the image encryption)14:39
mhendunno if the osc folks would be happy with removing other stuff14:39
rosmaitai was thinking of the 'force' flag, they send it all the time for some requests (forget which one) and it causes issues with microversion 3.6614:40
rosmaitaanyway, thank you for presenting both the issue and the solution!14:41
rosmaitaanything else on this?14:41
mhenno14:41
mhenthanks14:41
mhenI will try to go that route as discussed14:41
RanjithRethinamhttps://review.opendev.org/c/openstack/cinder/+/971739 can this get some attention pls14:41
rosmaitagreat! moving on14:41
rosmaita#topic continue discussion of os-brick patch from last week14:41
rosmaitayuval: that's you14:42
yuvalI am here, anything left not answered?14:42
rosmaita#link https://review.opendev.org/c/openstack/os-brick/+/95759914:42
yuvalthanks14:42
rosmaitai will have a minor comment about the release note14:43
rosmaitamight as well say it here as a reminder to everyone14:43
rosmaitadon't forget that the release notes are handled by automation that will re-arrange them14:43
rosmaitaso even though you have a single note, the sections will be far apart (most likely)14:44
rosmaitaso in yuval's case, in the "upgrade" section, he refers to the options but doesn't say which ones14:44
rosmaitabecause they're right there in the previous bullet point14:44
rosmaitabut when the release notes are all rendered for the release, they will be far apart14:44
rosmaitaso sometimes you have to be  a bit repetitive in your release notes14:45
yuvalI get you14:45
yuvalok will handle14:45
yuvalthank you14:45
rosmaitasounds good, i notice that erlon gave you a +1 and he's careful, so i imagine that your commit message update is good14:46
rosmaita(i will read after the meeting)14:46
rosmaitaok, so mainly this was a reminder that the patch was updated as requested, and the author is available here if there are questions14:46
rosmaitathanks!14:46
rosmaitajust a reminder to everyone, our release note docs (how to write) are here: https://docs.openstack.org/cinder/latest/contributor/releasenotes.html14:48
rosmaitaand if you want to see the actual release notes, https://docs.openstack.org/releasenotes/cinder/index.html14:48
rosmaitaand, finally, you can always check to see what your note looks like in html by either running tox locally, or looking at the results of the build-openstack-releasenotes job on your patch14:49
rosmaitaso, not to pick on yuval, but you can see his current note here: https://fe6d6d8568001e0da1b7-c29773bdeee4530a738751d9e026e2a7.ssl.cf2.rackcdn.com/openstack/ddcd6ef273e0474d9f8c5c89c95514d1/docs/unreleased.html14:50
rosmaitaok, that's all the topics on the agenda14:51
rosmaitai think RanjithRethinam has frantically been trying to get peoples' attention14:51
kaisersI'll fall in line: Would be great to have some fb on a small revert: https://review.opendev.org/c/openstack/cinder/+/967672 14:53
seunghunleeHello, can I ask some answers for the latests comments on #link https://review.opendev.org/c/openstack/cinder/+/82037514:53
rosmaitakaisers: you may need to rebase to make sure you pick up the change that will let py313 job pass14:54
opendevreviewTakashi Kajinami proposed openstack/cinder-specs master: Replace obsolete PCRE packages  https://review.opendev.org/c/openstack/cinder-specs/+/97407114:55
kaisersrosmaita will do, thnx for the pointer14:55
sanketh_shettyRequesting core reviewers to look into - https://review.opendev.org/c/openstack/cinder/+/973908. It a very small regression fix.14:55
rosmaitakaisers: it's the change to bindep.txt from https://review.opendev.org/c/openstack/cinder/+/970038 that you need14:56
kaisersack14:56
rosmaitaeharney: when you have some time, seunghunlee has a question about your question on https://review.opendev.org/c/openstack/cinder/+/82037514:58
eharneyyeah i need to followup there - IMO we probably don't want that patch because i think you can achieve the desired behavior by just setting up ceph correctly14:59
rosmaitaseunghunlee: if you do a 'git blame' on the rbd driver, look at the section where the client version is checked ... the commit message may have some info for you14:59
rosmaita(at least that's my memory)15:00
rosmaitasanketh_shetty: still looking at your patch, but could you please wrap the text in the release note?  right now line 4 is a single line, please break it up into a series of shorter lines15:02
rosmaitai think you can do our standard 79-char line length15:03
rosmaitaok, we have gone over time ... thanks everyone for attending!15:03
rosmaita#endmeeting15:04
opendevmeetMeeting ended Wed Jan 21 15:04:05 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)15:04
opendevmeetMinutes:        https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-01-21-14.02.html15:04
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-01-21-14.02.txt15:04
opendevmeetLog:            https://meetings.opendev.org/meetings/cinder/2026/cinder.2026-01-21-14.02.log.html15:04
seunghunleerosmita: Got it thank you. eharney: but does that ceph setup allows creating flatten volume from image?15:06
eharneyseunghunlee: if you enable image format v2 by setting minimum client compat to mimic, you do not have to flatten the image to remove the glance image that a vol was cloned from15:06
eharneyi need to add a note to the cinder docs about this probably15:06
seunghunleeeharney: hmm okay. Could you add a note as a comment on the patch as well? For the record.15:10
seunghunleeThanks for your time by the way15:10
mhencan somebody take a look at https://bugs.launchpad.net/cinder/+bug/2138806 and tell me if this is a user/validation error (slashes actually not allowed) or a genuine bug in Cinder or swiftclient?15:17
opendevreviewHarsh Ailani proposed openstack/cinder master: [IBM SVf Driver] Add basic support for Policy-Based Replication (PBR)  https://review.opendev.org/c/openstack/cinder/+/96544215:30
erloneharney: hey, can you give another check in the 512e/4k disk geometry patch? https://review.opendev.org/c/openstack/cinder/+/65828315:39
opendevreviewMerged openstack/cinder master: [tests-only] update for oslo.concurrency 7.3.0  https://review.opendev.org/c/openstack/cinder/+/97369317:04
opendevreviewDmitriy Rabotyagov proposed openstack/cinder master: Allow operators to control Ceph pool selection for backups  https://review.opendev.org/c/openstack/cinder/+/95867119:38
opendevreviewEric Harney proposed openstack/cinder master: WIP: API: Allow newlines in snapshot descriptions  https://review.opendev.org/c/openstack/cinder/+/92691421:48
opendevreviewEric Harney proposed openstack/cinder-tempest-plugin master: WIP: Test that snapshot descriptions can contain newlines  https://review.opendev.org/c/openstack/cinder-tempest-plugin/+/92691321:56

Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!