Wednesday, 2022-03-16

*** dviroel|ruck is now known as dviroel|ruck|Afk00:05
*** dviroel|ruck|Afk is now known as dviroel|ruck11:16
*** dasm|off is now known as dasm13:43
rosmaita#startmeeting cinder14:00
opendevmeetMeeting started Wed Mar 16 14:00:18 2022 UTC and is due to finish in 60 minutes.  The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot.14:00
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.14:00
opendevmeetThe meeting name has been set to 'cinder'14:00
toskyo/14:00
simondodsleyhi14:00
whoami-rajatHi14:00
eharneyhi14:00
hemnayough14:00
geguileohi! o/14:00
yuvalhey14:01
rosmaitahello everyone14:01
rosmaita#link https://etherpad.opendev.org/p/cinder-yoga-meetings14:01
jungleboyjo/14:02
rosmaitalight agenda today, so plenty of time for open discussion later14:02
rosmaita#topic announcements14:02
rosmaitaFestival of XS Reviews on Friday, 1400-1600 UTC14:02
rosmaitayou can find info about the festival here: http://eavesdrop.openstack.org/#Cinder_Festival_of_XS_Reviews14:03
fabiooliveirao/14:03
rosmaitawhich reminds me that i didn't send out a reminder yet14:03
rosmaita#action rosmaita festival of reviews reminder to ML14:03
whoami-rajatWe've a national festival on friday (Holi) so won't be able to attend...14:03
rosmaitaok, that's a good excuse14:03
whoami-rajat(probably anyone from India won't be around on friday)14:04
whoami-rajat:)14:04
jungleboyjrosmaita:  I am not sure if I will be able to attend.  My son has a college visit.14:04
rosmaitaok, that's also a good excuse, though not as good as Rajat's14:04
rosmaita:D14:04
jungleboyj:-)14:04
rosmaitaenriquetaso also won't be there14:05
rosmaitai have the time blocked off, so i am planning to be there14:05
rosmaitanext item14:05
rosmaitafinal cinder RC must be released by Thursday next week14:05
rosmaitanot much more to say about that, i think everyone knows what it means14:06
rosmaitaRC-1 was released on Monday14:06
rosmaita(that's "cinder release candidate #1"14:06
rosmaitathe stable/yoga branch has now been cut14:06
rosmaitabugfixes must merge to master and then be backported to stable/yoga14:06
jungleboyj++14:07
rosmaitathat's the case for everything (cinder, os-brick, cinderclient) other than cinderlib14:07
rosmaitacinderlib master remains yoga development until it has its yoga release at zed R-1514:07
rosmaitawhich reminds me, i could use some eyes on https://review.opendev.org/c/openstack/cinderlib/+/83378114:07
simondodsleyso things we only want in Zed can now be put up in master, correct?14:08
rosmaitathat is correct14:08
simondodsleycool14:08
rosmaitaso at this point, only release-critical changes will be merged to stable/yoga14:09
rosmaitaso let's take a quick look at what we have14:09
rosmaita#link https://etherpad.opendev.org/p/cinder-yoga-release-critical-changes14:09
rosmaitaso, first the good news14:09
rosmaitathe transition-to-alembic patches were all merged14:09
hemnaI hope that works when upgrading a real deployment :)14:10
rosmaitawe don't have any db migrations for yoga, but probably will soon14:10
rosmaitaso, https://docs.openstack.org/cinder/latest/contributor/database-migrations.html14:10
rosmaitathat explains how to use the new stuff14:10
rosmaitahemna: me too, i'll be putting up a patch very early in Zed to do a minor migration so we can make sure it works as expected14:11
rosmaitawe have the advantage that nova is about 1.5 cycles ahead of us on this, and the bugs have been mostly shaken out14:11
rosmaitaanyway, i want to thank stephenfin for working on this for cinder14:12
rosmaitawe will declare Stephen Finucane a "Hero of Cinder" at the PTG14:12
jungleboyj++14:12
whoami-rajatalso there's a new grenade job that checks migration from tock -> tock (or it's tick->tick) but it checks skipping one intermediate release14:12
rosmaitaas i mention on the etherpad, stephenfin has also posted a series of patches so that cinder will no longer use a deprecated oslo.db feature14:12
rosmaitawe will aim to merge those into Z quickly after the final RC14:13
rosmaitaok, now the not-so-good news14:13
rosmaitathere is still some work ongoing for known issues in Xena14:13
rosmaitathe good thing is that i don't think any of them are destabilizing14:14
rosmaitathey are fairly localized14:14
rosmaitaso, we will consider them release-critical14:14
rosmaitaother than that, if anyone finds a bug that looks like it needs to be addressed before release, please put it on the etherpad so we can be aware14:15
rosmaitaand decide whether it really is release-critical or not14:15
rosmaitaany questions?14:15
rosmaitaok, next item14:17
rosmaitaZed virtual PTG 4-8 April14:17
rosmaita#link https://etherpad.opendev.org/p/zed-ptg-cinder-planning14:17
rosmaitausual deal, you need to register even though it's free14:17
rosmaitawhoami-rajat is doing the organization this time14:18
whoami-rajato/14:18
rosmaitaif you look at the etherpad, there are a bunch of driver topics that i added14:18
rosmaitait's looking like maybe we'll have a drivers' day14:18
whoami-rajatah, they weren't there the last time i checked14:19
rosmaitaso if you are a driver maintainer, and have something we should discuss, please add it 14:19
whoami-rajatlooks like a drivers' day is needed14:19
whoami-rajat(it could also be called Brian's day)14:19
rosmaitabtw, i will be happy to let anyone else who's interested take those over14:19
rosmaitathe PTG is coming up fast, so please spend a bit of time thinking about it14:20
geguileodriver's day++14:20
rosmaitahemna: i didn't see your recomputing available space topic on there yet14:20
jungleboyjOooh, that sounds like an exciting one.14:21
rosmaitathat doesn't necessarily have to be on drivers day, though i imagine there may be some reporting issues that it would be good to get input on14:22
simondodsleyI won't be able to attend, but I'm going to add a topic for discussion if that is ok.14:22
geguileojungleboyj: was that a sarcastic comment?  XD14:22
geguileosimondodsley: nvme related?14:22
rosmaitasimondodsley: that's fine14:22
jungleboyjNo ... I have a feeling that it will be somewhat contentious.  :-)14:22
rosmaitathough you'll need to get someone to agree to lead the discussion for you14:22
geguileojungleboyj: probably :-)14:22
jungleboyjMight be interesting.14:23
rosmaitasince we're not in a hurry today, i will pause to give people time to look at the etherpad and add comments14:23
simondodsleyI'll add the topic and if someone want s to lead it that would be great, but if not it can be dropped for another session14:24
rosmaita#link https://etherpad.opendev.org/p/zed-ptg-cinder-planning14:24
geguileoI'll probably add a couple of topics:  Current nvme-of connector issues, drivers reporting their storage_protocol, quota update14:25
rosmaitageguileo: ++14:25
whoami-rajatIf you've added a topic and have a preferred time and won't be attending the whole PTG, you can add the time (UTC) suited for your TZ (only in the available time slot)14:27
whoami-rajatgeguileo++14:27
geguileosimondodsley: are there leftovers on ALL successful runs or just some?14:29
simondodsleyi'd have to check, but it feels like all14:29
rosmaitajust a reminder about the PTG schedule for cinder: the Cinder team will meet in the 13:00-17:00 UTC blocks on Tuesday 5th April through Friday 8th April 2022.14:29
rosmaitawow, we do have a lot of topics for drivers' day!14:31
geguileosimondodsley: that'd be the best case scenario, because going into the system we can tell the test that left it behind14:31
rosmaitaok, the etherpad is available 24 hours a day, so please add items as they occur to you14:31
geguileoand the new field that tells us if it's a temporary resource may help as well14:31
rosmaitaok, final announcement14:32
rosmaitareminder about the cinder resource count survey14:32
rosmaita#link http://lists.openstack.org/pipermail/openstack-discuss/2022-March/027660.html14:32
rosmaitawe are trying to get geguileo some data to determine what direction his quotas improvement plans should take14:33
rosmaitaso please tell your customers14:33
rosmaitaand if you run a cloud, please fill it out yourself!14:33
geguileoplease, pretty please14:33
rosmaitathat's all from me, anyone else have an announcement?14:33
rosmaitaactually, if sofia were here, she would remind us about outreachy14:34
rosmaitai believe the deadline for proposals is early next week14:34
rosmaitai think this email has details:14:34
rosmaita#link http://lists.openstack.org/pipermail/openstack-discuss/2022-February/027239.html14:34
rosmaita#topic reminder: no "naked" rechecks14:35
rosmaitaGorka added some documentation about this:14:35
rosmaita#link https://docs.openstack.org/cinder/latest/contributor/gerrit.html#ci-job-rechecks14:36
rosmaitathe basic idea is that we want to do a bit of troubleshooting before rechecking14:36
rosmaitabecause a failure in any test that has 'volume' in the name is immediately ascribed to cinder, whether it's a cinder failure or not14:36
rosmaitai guess that's all i have to day about that14:37
whoami-rajatgeguileo, found out a typo if you would like to fix "without locking at the errors" locking -> looking14:38
* geguileo is the king of typos14:38
rosmaitathat's a good typo given that geguileo is our race conditions expert14:38
whoami-rajat:D true14:39
rosmaita#topic open discussion14:39
jungleboyj:-)14:40
eharneyi have this cool patch that just removes a bunch of dead code, if someone wants an easy review: https://review.opendev.org/c/openstack/cinder/+/80070414:43
geguileoeharney: that one has a couple of +2 already...14:43
eharneyindeed14:43
eharneylast week's meeting got it this far, maybe today's can really do it14:43
rosmaitaeharney: i think you should feel free to +W it14:44
rosmaitaor maybe jungleboyj will 14:45
geguileoeharney: why do we need this change? https://review.opendev.org/c/openstack/cinder/+/800704/3/cinder/volume/drivers/remotefs.py14:45
rosmaitalooks like he and e0ne had a race condition, both thought they were putting the first +2 on it14:45
jungleboyjrosmaita:  I can.14:45
rosmaitaty14:45
eharneygeguileo: good question14:45
eharneygeguileo: i wrote this last summer, so lemme go find out...14:45
geguileoI know how that feels14:46
jungleboyj*sad trombone*14:46
rosmaitai wonder if that's a pylint thing14:47
yuvalI can bring up an idea14:48
geguileoeharney: looks like it's irrelevant, it's merging now   XD14:48
yuvalwhat about moving out unittest from using tox to use docker14:48
yuvalour*14:49
eharneytox works pretty well14:49
rosmaitayuval: can you say some more about what you're thinking there?14:50
yuvalmy exp with tox - first was that its not working - every env I had to install it I had to debug certain bug, missing dependency libs14:50
yuvaljust a docker with the minimum req to run the unittests14:51
yuvalzuul running py36 and py39 in my machine I am running py3814:52
yuvalI would like to run exactly the same tests14:52
yuvaldocker solves that14:52
yuvalis there someone here which is also active in kolla project?14:53
whoami-rajatgeguileo, eharney i can see those variables not used anywhere so maybe signifying that those are unused (_<var> instead of __)14:53
geguileowhoami-rajat: makes sense14:53
eharneywhoami-rajat: that is the convention of what the leading underscore means, yes -- pylint etc. actually check based on this14:53
geguileothanks14:53
whoami-rajatnp14:54
whoami-rajatack14:54
rosmaitayuval: i use pyenv to give me multiple pythons beyond what's available in my operating system distro14:55
rosmaitadon't know if that would help you14:56
rosmaitaok, two-minute warning ... anyone else have something to mention?14:58
rosmaitaguess not ... thanks for attending, and don't forget about the festival of reviews on friday14:59
rosmaita#endmeeting14:59
opendevmeetMeeting ended Wed Mar 16 14:59:48 2022 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)14:59
opendevmeetMinutes:        https://meetings.opendev.org/meetings/cinder/2022/cinder.2022-03-16-14.00.html14:59
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/cinder/2022/cinder.2022-03-16-14.00.txt14:59
opendevmeetLog:            https://meetings.opendev.org/meetings/cinder/2022/cinder.2022-03-16-14.00.log.html14:59
whoami-rajatthanks!14:59
jungleboyjThank you!15:00
*** dviroel|ruck is now known as dviroel|ruck|lunch16:02
*** dviroel|ruck|lunch is now known as dviroel|ruck17:15
*** dviroel|ruck is now known as dviroel|ruck|afk20:03

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!