14:00:47 <whoami-rajat> #startmeeting cinder 14:00:47 <opendevmeet> Meeting started Wed Apr 13 14:00:47 2022 UTC and is due to finish in 60 minutes. The chair is whoami-rajat. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:47 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:47 <opendevmeet> The meeting name has been set to 'cinder' 14:00:55 <whoami-rajat> #topic roll call 14:00:57 <simondodsley> hi 14:00:58 <enriquetaso> hi 14:01:00 <jungleboyj> o/ 14:01:12 <eharney> hi 14:01:14 <geguileo> hi! o/ 14:02:07 <tosky> hi 14:02:10 <whoami-rajat> let's wait a couple of minutes for other people to join 14:02:18 <nahimsouza[m]> o/ 14:02:26 <rosmaita> o/ 14:03:28 <andrebeltrami[m]> o/ 14:04:08 <whoami-rajat> ok, let's get started! 14:04:15 <whoami-rajat> Welcome everyone to the first cinder meeting of Zed development cycle 14:04:21 <whoami-rajat> #link https://etherpad.openstack.org/p/cinder-zed-meetings 14:04:27 <jungleboyj> \o/ 14:04:40 <whoami-rajat> good to see everyone here shortly after the PTG :) 14:04:51 <felipe_rodrigues> o/ 14:04:55 <whoami-rajat> we've a lot of announcements, so let's start 14:04:57 <whoami-rajat> #topic announcements 14:05:06 <whoami-rajat> first, Add IRC nick to courtesy reminder 14:05:17 <whoami-rajat> I've already added the IRC nicks from the yoga meetings etherpad but if there are new people that would like to add their names, they can add it to the courtesy reminder list on the etherpad on L#33 14:05:52 <whoami-rajat> I will remind about the meeting in the #openstack-cinder channel few minutes before the meeting starts here 14:06:01 <whoami-rajat> so make sure, you've joined both the channels 14:06:23 <whoami-rajat> moving on 14:06:26 <whoami-rajat> next announcements, stable/victoria moving to EM 14:06:45 <whoami-rajat> so there's a mail from Elod (release PTL) that stable/victoria will be moving to EM soon 14:06:48 <whoami-rajat> stable/victoria is going to EM on 27th April, 2022, so if there are any patches you would like to get released with the last victoria release, start backporting them now 14:07:03 <whoami-rajat> #link http://lists.openstack.org/pipermail/openstack-discuss/2022-April/028013.html 14:07:21 <whoami-rajat> more about this we will discuss in the next announcement 14:07:47 <whoami-rajat> next, Release liaison 14:08:23 <whoami-rajat> so some of you might know, I've been managing the stable releases for the last 3-4 cycles but since now I've taken up the role of PTL, if anyone is interesting in doing stable releases, they can contact me or Brian and we can guide you through the process 14:08:50 <whoami-rajat> It has to be done every couple of months and if we have enough changes to release so not a big burden, and also you don't have to be a core reviewer to do it 14:09:11 <whoami-rajat> This glance doc is a good starting point for the role and responsibilities http://docs.qstack.com.cn/glance/latest/contributor/release-cpl.html 14:09:28 <whoami-rajat> we can think about adding this to our documentation as well 14:10:07 <whoami-rajat> next, Bug squad deputy 14:10:32 <whoami-rajat> enriquetaso has been handling the bug triaging and reporting and she has been doing an excellent work! 14:10:45 <whoami-rajat> If she would like to continue, it would be great but also if she wants to move on from the responsibility, we can discuss if anyone else would like to take that role 14:11:07 <whoami-rajat> for now, i will leave the decision to enriquetaso 14:11:23 <enriquetaso> i haven't though of that :P 14:11:23 <whoami-rajat> next, Release schedule 14:11:46 <whoami-rajat> sure, you can take your time! 14:12:00 <whoami-rajat> I've proposed the Zed release schedule, you can check the patch and html rendering on the etherpad 14:12:18 <whoami-rajat> #link https://review.opendev.org/c/openstack/releases/+/837494 14:12:30 <whoami-rajat> #link https://58140b717b0fd2108807-1424e0f18ed1d65bb7e7c00bb059b2d8.ssl.cf2.rackcdn.com/837494/1/check/openstack-tox-docs/6b018f1/docs/zed/schedule.html 14:12:42 <rosmaita> the only thing i noticed is that monday of the week of milestone 2 is a holiday in the USA 14:13:01 <rosmaita> or maybe not 14:13:15 <rosmaita> the lack of grid lines on the HTML schedule may be confusing me 14:13:18 <geguileo> ooooooh, ninja release then? 14:13:44 <whoami-rajat> oh ok, but i guess we've time till friday for any deadline so only monday holiday should be fine? 14:13:46 <rosmaita> i take it back, m-2 is week of 14 july 14:14:02 <jungleboyj> Right, it is the week after the 4th of July. 14:14:18 <whoami-rajat> also Brian has a suggestion to include mid cycle events in the schedule which i think would be useful for us, I will add them and see if the release team complaints 14:14:31 <whoami-rajat> great, no conflicts! 14:14:33 <jungleboyj> That is a great idea! 14:15:06 <geguileo> +1 to include all events, mid cycle, ptg, everything 14:15:31 <whoami-rajat> sounds good, i will add them 14:15:38 <whoami-rajat> so check out the release schedule and add comments regarding date conflicts and if we should move things around 14:16:00 <whoami-rajat> next announcement, April Festival of XS Reviews rescheduled 14:16:19 <whoami-rajat> so we've Good Friday holiday coming up this week for most of the people 14:16:27 <jungleboyj> ++ 14:16:34 <whoami-rajat> but this is also the 3rd week of April i.e. Festival of XS reviews week 14:16:46 <whoami-rajat> so we're planning to move it to next or next to next friday 14:16:58 <whoami-rajat> (i was surprised to know we've 5 fridays in April) 14:17:22 <whoami-rajat> the best way to decide is to get the team votes in for it 14:17:30 <enriquetaso> 29th? 14:17:39 <enriquetaso> +1 14:17:51 <whoami-rajat> Lets vote for festival of XS reviews date 14:17:51 <whoami-rajat> options are: 14:17:51 <whoami-rajat> A 22nd April, 2022 14:17:51 <whoami-rajat> B 29th April, 2022 14:17:51 <whoami-rajat> C either one of the above 14:17:52 <whoami-rajat> D need more options 14:17:58 <whoami-rajat> #startvote time for the festival of XS reviews ? A, B, C, D 14:17:58 <opendevmeet> Begin voting on: time for the festival of XS reviews ? Valid vote options are A, B, C, D. 14:17:58 <opendevmeet> Vote using '#vote OPTION'. Only your last vote counts. 14:18:09 <enriquetaso> #vote C 14:18:25 <whoami-rajat> #vote C 14:18:28 <rosmaita> #vote C 14:18:35 <eharney> #vote C 14:19:09 <sfernand> #vote C 14:19:17 <jungleboyj> #vote C 14:20:09 <whoami-rajat> let's wait 2 more minutes for everyone to get their votes in 14:20:39 <lucasmoliveira059> #vote C 14:20:43 <geguileo> #vote C 14:20:48 <ecsantos[m]> #vote C 14:21:03 <rfluisa_> #vote C 14:21:40 <rosmaita> whoami-rajat: looks like you will get to choose! 14:21:45 <whoami-rajat> ok, i guess we already have majority 14:21:46 <whoami-rajat> #endvote 14:21:46 <opendevmeet> Voted on "time for the festival of XS reviews ?" Results are 14:21:46 <opendevmeet> C (10): geguileo, rosmaita, lucasmoliveira059, rfluisa_, whoami-rajat, sfernand, eharney, enriquetaso, ecsantos[m], jungleboyj 14:22:09 <whoami-rajat> rosmaita, :D yeah, voting wasn't much helpful here 14:22:24 <rosmaita> well, at least we know everyone can make it, whichever day you pick 14:22:50 <whoami-rajat> so i propose 29th April since we're still too close to PTG and people might not like attending another 2 hour meeting this early 14:23:01 <rosmaita> works for me 14:23:06 <jungleboyj> Makes sense. 14:23:06 <whoami-rajat> rosmaita, yep, that helped 14:23:21 <whoami-rajat> great, so 29th April it is 14:23:37 <whoami-rajat> that's all i had for announcements 14:23:41 <whoami-rajat> let's move on to topics 14:23:47 <whoami-rajat> #topic Remove legacy oslo.db enginefacade 14:23:50 <whoami-rajat> rosmaita, thats you 14:24:05 <rosmaita> thanks! 14:24:30 <rosmaita> ok, Stephen Finucane, a Hero of Cinder has proposed a bunch of patches 14:24:49 <rosmaita> to update our session usage in the sqlachemy/api.py 14:25:00 <enriquetaso> \o/ 14:25:04 <rosmaita> it helps in 2 ways 14:25:29 <rosmaita> we are using an oslo.db legacy enginefacade that the oslo project has deprecated way back and wants to remove 14:25:41 <rosmaita> and we may be the only ones left using it 14:25:55 <rosmaita> and, this change also helps us get ready for SQLAlechemy 2.0 14:26:03 <rosmaita> which is coming very soon 14:26:28 <rosmaita> stephen broke this into a bunch of patches so each review isn't too big 14:26:57 <rosmaita> also, the line of code count in gerrit is not very accurate because a large part of this is un-indenting python code 14:27:13 <rosmaita> the quick explanation of what the change is, is here: 14:27:28 <rosmaita> #link https://docs.openstack.org/oslo.db/latest/reference/api/oslo_db.sqlalchemy.html#module-oslo_db.sqlalchemy.session 14:27:46 <rosmaita> i put together a dashboard, but i don't know how useful it will be 14:27:57 <rosmaita> i thought there were about 15 patches in the patch set 14:28:05 <rosmaita> but there are actually more like 34 14:28:14 <geguileo> 34!!!! 14:28:31 <rosmaita> yes, it is almost a geguileo-sized set of patches! 14:28:45 <geguileo> it's waaaaay beyond my longest patch series lol 14:28:49 <rosmaita> actually, 34 is a bit excessive, even for geguileo 14:28:53 <rosmaita> :D 14:28:57 <geguileo> true 14:29:19 <rosmaita> yeah, so best thing is to start at the bottom, and get used to the pattern, and it gets easier as you go 14:29:21 <tosky> so wait for the devstack fix to land, merge everything, wait for explosions and fix everything in the next weeks? 14:29:41 <rosmaita> which devstack fix? 14:29:53 <whoami-rajat> I think this makes sense as we're already long overdue for the alembic work which is complete but we need to get this in to safeguard our migrations against SQLAlchemy update 14:30:20 <rosmaita> ok, my proposal is that we should not merge any DB changes before we get this set merged 14:30:32 <rosmaita> because i really don't want to review this multiple times 14:31:16 <tosky> rosmaita: devstack fix because devstack was broken by a git security fix (see openstack-discuss) 14:31:39 <rosmaita> tosky: that's what i get for not reading my email this morning 14:32:12 <whoami-rajat> I will take a look once I'm done with the PTG summary (it does take time to write) 14:32:33 <geguileo> fyi I have this patch that I need to update... https://review.opendev.org/c/openstack/cinder/+/815746 14:32:38 <rosmaita> here's the devstack bug link: 14:32:43 <rosmaita> #link https://bugs.launchpad.net/devstack/+bug/1968798 14:32:44 <geguileo> because we lost that with the alembic changes 14:33:01 <rosmaita> thanks tosky 14:34:27 <tosky> geguileo: should that be merged before or after merging the other patches? 14:34:40 <rosmaita> tosky: i was about to ask that 14:34:59 <geguileo> tosky: there are probably no conflicts between the 2 things 14:35:16 <geguileo> and if there are, we should merge the others first 14:35:33 <geguileo> because resolving a merge conflict in a 34 patch series would kill our CI 14:35:39 <tosky> geguileo: but could it be used to check if the other patches introduce regressions? 14:35:46 <rosmaita> i guess my question is, what is a good way to get these all reviewed? 14:36:03 <geguileo> tosky: not really, the check is for new features and changes to the actual DB model, which the new code should not be touching 14:36:04 <rosmaita> i am committed to reviewing all of them, but i am insane 14:37:35 <whoami-rajat> i see geguileo has a series of patches where some may conflict with stephen's chain https://review.opendev.org/q/topic:die-quota-die-base 14:37:55 <geguileo> whoami-rajat: those can wait 14:38:08 <geguileo> Since I have to make additional changes 14:38:17 <geguileo> I'll just rebase them on top of his 34 patches 14:38:36 <whoami-rajat> ok, so let's review stephen's patches as priority (at least as much as we can get in from them) 14:38:46 <whoami-rajat> and other DB changes can be reviewed/merged later 14:39:19 <rosmaita> let's revisit this next week, if there's not a lot of review action, i will set up a spreadsheet or something and we can "volunteer" specific people to review subsets of patches 14:39:43 <whoami-rajat> that sounds like a good idea 14:39:56 <rosmaita> that's all from me, unless people think 1 week is too long to wait 14:40:23 <whoami-rajat> i guess they can review changes anyway! 14:40:30 <whoami-rajat> so let's move on to next topic 14:40:33 <rosmaita> yep 14:40:34 <whoami-rajat> #topic Disable c-bak by default in CI jobs 14:40:43 <whoami-rajat> rosmaita, that's you again 14:40:56 <rosmaita> #link https://review.opendev.org/c/openstack/devstack/+/837630 14:41:03 <tosky> (sorry) I thought it was disabled long ago, we explicitly enable it in a few cinder jobs 14:41:27 <rosmaita> apparently it's enabled by default in devstack 14:42:06 <rosmaita> i guess as long as we have test coverage, it doesnt' need to be enabled by default? 14:42:44 <whoami-rajat> there's a comment by Dan saying it's causing OOM failures frequently in ceph multistore job (i assume that's in glance) 14:42:54 <rosmaita> has anyone had time to look at David Caro's patch: https://review.opendev.org/c/openstack/cinder/+/834527 14:43:05 <tosky> well, they can also disable it in the specific jobs which shows resource issues 14:43:37 <eharney> rosmaita: i looked over it a bit, it makes sense conceptually 14:45:44 <rosmaita> i don't have any more to say, just wanted people to be aware 14:46:01 <rosmaita> clarkb put a -W on the devstack patch 14:46:03 <whoami-rajat> so should c-bak be disabled/enabled by default? what is our take on this 14:46:28 <eharney> by default in CI jobs, you mean 14:46:43 <rosmaita> right 14:46:50 <whoami-rajat> yes 14:46:55 <tosky> whoami-rajat: for the record, it was disabled some time ago 14:47:40 <tosky> by default; I don't remember where and when it was reenabled, maybe when the signle integrated job was split into more integrated job (integrated-storage in our case) 14:48:12 <whoami-rajat> ok, i need to followup on the history of it then 14:48:15 <rosmaita> well, i'll leave a comment that we are looking at a possible fix 14:48:21 <eharney> i don't think it matters too much whether it's on in jobs by default, but it does mean we have to go make sure this doesn't reduce test coverage for us, or if we need to change our jobs 14:48:45 <whoami-rajat> eharney +1 14:48:49 <rosmaita> i think clark -W'd because the logs showed that somethign was using c-back 14:49:06 <rosmaita> ok, thanks, we can see what happens 14:49:22 <rosmaita> in the meantime, would be good to look at https://review.opendev.org/c/openstack/cinder/+/834527 14:49:37 <eharney> i'm not sure if the 834527 patch would even help if it's in Ceph jobs, would need to see what those are running exactly 14:50:02 <eharney> in general, if people are having OOM issues, we need bug reports 14:50:05 <rosmaita> that's a good point 14:51:37 <whoami-rajat> rosmaita, so i guess we've enough points to hold the patch as of now until we accurately know if it's a good idea or not 14:51:48 <rosmaita> sounds good to me 14:51:49 <whoami-rajat> as eharney pointed out, it shouldn't reduce our test coverage 14:51:57 <whoami-rajat> great 14:52:06 <whoami-rajat> let's move on to next topic then 14:52:13 <whoami-rajat> #topic vote on NetApp proposal to backport ONTAP REST API support 14:52:47 <whoami-rajat> so people who have attended the PTG on Thursday might be familiar with NetApp's proposal about changing their ZAPI calls to REST API calls 14:53:16 <whoami-rajat> their ask is to allow the change to be backported to the active stable branches 14:53:23 <geguileo> whoami-rajat: I don't feel good voting for this without e0ne being present (major opposition to doing it) 14:53:41 <sfernand> geguileo: ++ 14:53:44 <jungleboyj> ++ 14:53:50 <whoami-rajat> oh, i didn't notice, that's a good point by geguileo 14:53:52 <sfernand> I think we can recap today 14:54:00 <sfernand> And do the votes next week 14:54:08 <rosmaita> i'm on board with e0ne, i think 14:54:19 <whoami-rajat> sfernand, that also sounds good, would you like to provide a summary of it? 14:54:37 <sfernand> Yes give me a minute to type 14:54:46 <jungleboyj> Makes sense. 14:54:55 <whoami-rajat> sure 14:56:33 <sfernand> Ok, so during the ptg we discussed the possibility of backporting our old xml based calls to rest, the reason for this is because zapis will be deprecated on 2023, which means customers wouldn’t be able to upgrade ontap arrays in any older version then zed (if me merge in zed) 14:57:02 <sfernand> So the debate was around if that should be considered a new feature 14:57:19 <sfernand> Or a Bug we are trying to antecipate 14:57:58 <whoami-rajat> #link https://etherpad.opendev.org/p/zed-ptg-cinder#L431 14:57:59 <sfernand> Since it is going to directly affect rhosp deployments which currently relies on train.. 14:58:20 <sfernand> During the discussion we agreed that there is no chance for train 14:58:47 <sfernand> But that could be considered for the versions we still provide new releases 14:59:07 <sfernand> Like yoga, xena, wallaby 14:59:59 <sfernand> Our idea is to backport to the active stable branches and make it possible for rhosp 17 15:00:08 <simondodsley> The next RHOSP release will be based on Wallaby so that would be the oldest version you would have to backport to (if allowed...) 15:00:24 <whoami-rajat> that's a good summary and people can also follow the PTG discussion and recording to get more familiar with it, netapp had a nice presentation as well 15:00:40 <whoami-rajat> we're out of time 15:00:44 <sfernand> Yep, and that would be enough since we could at least advice our customers to upgrade 15:00:47 <simondodsley> Although I believe RH have already frozen 17.0 15:01:13 <whoami-rajat> the last topic is by ganso requesting reviews for his backports, so check them out on the etherpad 15:01:19 <whoami-rajat> we can continue discussion next week 15:01:36 <ganso> yes please review 15:01:45 <whoami-rajat> thanks all! 15:01:47 <whoami-rajat> #endmeeting