14:01:06 <rosmaita> #startmeeting cinder 14:01:07 <openstack> Meeting started Wed Oct 7 14:01:06 2020 UTC and is due to finish in 60 minutes. The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:10 <openstack> The meeting name has been set to 'cinder' 14:01:13 <enriquetaso> \o/ 14:01:17 <jungleboyj> o/ 14:01:18 <rajinir> hi 14:01:19 <e0ne> hi 14:01:21 <enriquetaso> hi 14:01:23 <sfernand> Hi 14:01:23 <eharney> hi 14:01:26 <pots> \o 14:01:27 <whoami-rajat___> Hi 14:01:28 <walshh_> hi 14:01:35 <geguileo> hi! o/ 14:01:36 <smcginnis> o/ 14:01:39 <jungleboyj> pots ! 14:02:03 <rosmaita> #link https://etherpad.openstack.org/p/cinder-victoria-meetings 14:02:09 <rosmaita> good turnout 14:02:18 <rosmaita> a lot on the agenda today, but it's mostly announcements 14:02:32 <TusharTgite> hi 14:02:41 <rosmaita> #topic announcements 14:03:14 <rosmaita> ok, next week starts wallaby! 14:03:28 <rosmaita> so we will have a new agenda: https://etherpad.opendev.org/p/cinder-wallaby-meetings 14:03:36 <rosmaita> please update your links 14:03:59 <rosmaita> also, the stable releases happened last week 14:04:19 <rosmaita> this is a reminder about the upgrade-to-ussuri bug that was fixed: 14:04:29 <rosmaita> #link http://lists.openstack.org/pipermail/openstack-discuss/2020-October/017722.html 14:04:55 <rosmaita> ok, next item is that we are going to have a cinder type-checking hackfest 14:05:03 <rosmaita> here's my proposal 14:05:14 <rosmaita> we celebrate the Victoria release by holding it 1300-1600 UTC Wednesday 14 October and skipping the cinder weekly meeting 14:05:35 <rosmaita> if that's ok with everyone, i won't do a poll 14:05:40 <enriquetaso> ++ 14:05:59 <rosmaita> so it's a 3 hour block right around the cinder meeting time 14:06:14 <whoami-rajat___> +1 14:06:17 <lseki_> +1 14:06:18 <rosmaita> i do have a question, though 14:06:29 <rosmaita> do we want to do it in a video conference, or just on IRC? 14:07:22 <eharney> screen sharing may be handy 14:07:57 <rosmaita> ok, let's plan to at least start in video 14:08:03 <smcginnis> https://2.bp.blogspot.com/-YQfGEJjE0D0/Wn_t96fSqYI/AAAAAAAAMgQ/olcY9HDTbc4vhkdfoTbB6HrGr0CVls6vgCLcBGAs/s1600/why-not-both-animated-gif-7.gif 14:08:04 <TusharTgite> +1 14:08:26 <jungleboyj> :-) 14:08:34 <rosmaita> next question ... we usually use bluejeans, but i happen to know that hasn't been working for eharney recently 14:09:00 <rosmaita> since you'd most likely do the screen sharing, what technology do you suggest eharney? 14:09:28 <smcginnis> There's the OpenDev Jitsi if we want to try that again. 14:09:57 <rosmaita> maybe we should try that 14:10:04 <rosmaita> we don't need to record this 14:10:14 <eharney> trying jitsi might be good, i can make bluejeans work again if i have to (i assume) 14:11:08 <jungleboyj> eharney: How are you working at RH without BlueJeans working? ;-) 14:11:18 <rosmaita> we are flexible! 14:11:22 <eharney> using my phone like a caveman 14:11:33 <jungleboyj> :-) 14:11:46 <rosmaita> that's true, we are only flexible about the meetings that eharney himself runs 14:11:50 <rosmaita> everything else is bluejeans 14:12:57 <rosmaita> sorry about the delay here, i am trying to connect to meetpad 14:13:29 <rosmaita> ok, let's go for meetpad/opendev jitsi ... i will send out connection info on the ML 14:13:40 <rosmaita> and also put it in the agenda etherpad 14:14:05 <rosmaita> next item is that the TC election is happening now 14:14:13 <rosmaita> vote before 2020-10-13 23:45 UTC 14:14:22 <rosmaita> you should have an email with a customized link 14:14:27 <rosmaita> so i can't give you one here 14:14:36 <jungleboyj> ++ 14:14:43 <rosmaita> but if you haven't seen the email, search in your spam folder 14:14:48 * eharney is still looking for his email to vote 14:14:52 <rosmaita> subject: Poll: OpenStack Wallaby Cycle Technical Committee Election Poll 14:15:01 <rosmaita> from: Jeremy Stanley (CIVS poll supervisor) <civs@cs.cornell.edu> 14:15:23 <rosmaita> and if you don't see anything, i don't think fungi will mind if you ping him directly 14:16:00 <rosmaita> my email is dated yesterday evening NY time 14:16:40 <rosmaita> ok, now comes a list of upcoming events 14:16:53 <rosmaita> so, next week: 14:16:54 <rosmaita> victoria release and type-checking hackfest 14:17:02 <rosmaita> week after that is summit/forum 14:17:26 <rosmaita> anyone here doing a summit presentation, please put a link on the agenda when you have a few minutes 14:17:48 <rosmaita> we have 2 forum sessions where it would be good for as many cinder people who can to attend 14:18:01 <rosmaita> cinder forum session: monday 1230 UTC: https://www.openstack.org/summit/2020/summit-schedule/events/24748/cinder-user-messages 14:18:12 <rosmaita> cinder forum session: wednesday 1500 UTC: https://www.openstack.org/summit/2020/summit-schedule/events/24749/cinder-third-party-ci-whats-in-it-for-you 14:18:30 <rosmaita> and don't forget to register for the summit if you haven't yet 14:18:47 <rosmaita> ok, the week after the summit is the Wallaby PTG 14:18:52 <rosmaita> that's the week of 26 october 14:19:05 <rosmaita> put topics on the planning etherpad: 14:19:15 <rosmaita> #link https://etherpad.opendev.org/p/wallaby-ptg-cinder-planning 14:19:33 <rosmaita> and block off the cinder times on your calendar: 14:19:43 <rosmaita> tuesday-friday at 1300-1600 UTC 14:20:09 <rosmaita> and that reminds me, did we want to do a "happy hour"? 14:20:30 <e0ne> +1 14:20:46 <jungleboyj> +1 14:20:53 <lseki_> +1 14:21:23 <rosmaita> ok, looks like there's a groundswell of support 14:21:34 <whoami-rajat___> +1 (i think we thought of doing at start of PTG this time) 14:21:44 <rosmaita> whoami-rajat___: right you are 14:21:54 <whoami-rajat___> for better introduction 14:22:17 <rosmaita> so is tuesday at 1500 UTC good for everyone? 14:22:33 <e0ne> +1 14:22:34 <rosmaita> that would be tuesday 27 october 14:22:49 <enriquetaso> +1 14:23:29 <sfernand> +1 14:23:34 <lseki_> +1 14:23:38 <TusharTgite> +1 14:23:53 * lseki_ adds beer to his shopping list 14:23:57 <jungleboyj> I can make that work. 14:24:06 <enriquetaso> lseki_++ 14:24:10 <rosmaita> ok, i will propose that on the ML and if conflicts arise we can work it out in the planning etherpad 14:24:44 <rosmaita> so that's alll the events over the next 3 weeks 14:24:51 <rosmaita> for The Future: 14:24:59 <rosmaita> 30 november: Wallaby Milestone 1 14:25:07 <rosmaita> 14 december: Cinder Spec Freeze 14:25:23 <rosmaita> M-1 seems alarmingly close, if you ask me 14:25:40 <rosmaita> ok, that's all the announcements 14:26:02 <rosmaita> just a reminder: anyone with a summit preso, please put it on the agenda etherpad 14:26:26 <rosmaita> #topic RC-2 14:26:39 <rosmaita> we will be doing a second release candidate 14:26:48 <rosmaita> must be done tomorrow 14:26:59 <rosmaita> #link https://etherpad.opendev.org/p/cinder-victoria-release-critical-patches 14:27:28 <rosmaita> basically there are two patches whose backport needs reviewing 14:27:42 <rosmaita> #link https://review.opendev.org/#/c/756358/ 14:27:52 <rosmaita> #link https://review.opendev.org/#/c/756357/ 14:28:21 <rosmaita> the gates were really slow the past few days, plus one of the tempest storage jobs was unstable 14:28:42 <rosmaita> so the situation with those patches is that they don't have a +1 from Zuul 14:28:43 <jungleboyj> Looks like smcginnis got them already? 14:29:14 <rosmaita> ok, nevermind 14:29:18 <smcginnis> Feel free to take a look if you have time though. 14:29:51 <rosmaita> thanks smcginnis 14:29:55 <pots> thanks smcginnis and rosmaita for the help with https://review.opendev.org/#/c/756357/ 14:30:27 <rosmaita> well, if the gates cooperate, we should be ready for RC-2 tomorrow 14:30:47 * jungleboyj makes a sacrifice to the gate gods. 14:31:22 <smcginnis> 113 in check queue. I remember when deadline days would be well over 300+. 14:32:11 <rosmaita> check back in a few hours! 14:32:19 <smcginnis> Haha, right. ;) 14:32:34 <e0ne> smcginnis: it's a usual number for this time of the day 14:33:26 <rosmaita> the patches in the check seem to sit longer than usual, or maybe it's a "watched pot never boils" kind of thing 14:33:49 <rosmaita> ok, last scheduled topic 14:34:03 <rosmaita> #topic HP Primera backend situation 14:34:18 <rosmaita> this was raised by https://review.opendev.org/#/c/752934/ 14:34:19 <raghavendrat> hi 14:34:24 <rosmaita> hello 14:34:39 <rosmaita> so about 1 year ago, hpe added support for the Primera backend 14:34:57 <rosmaita> just for iscsi, i think? 14:35:06 <rosmaita> and this patch adds it for FC 14:35:14 <fungi> rosmaita: indeed, i don't mind direct irc pings or e-mails 14:35:18 <raghavendrat> last year, it was only for FC. now its for iSCSI 14:35:28 <rosmaita> ok, sorry, i had it backwards 14:35:36 <smcginnis> Our docs aren't super clear, but we do state "For example, if your company has two volume drivers in Cinder and they both use ISCSI and FibreChannel, you would need to have a CI that tests against four backends and reports the results for each backend, for every Cinder upstream patch" 14:36:03 <smcginnis> We should clarify that to also be clear that even if it is the same driver, all supported backends that that driver supports also need to be running tests. 14:36:20 <rosmaita> you have identified the issue 14:37:15 <raghavendrat> hhmm. currently CI runs on 3PAR backend. not running on Primera 14:37:16 <rosmaita> i thought that last year when the first patch merged, there was a primera CI test? 14:37:20 <raghavendrat> however we do perform manual testing on primera 14:37:37 <smcginnis> It needs to have automatic CI testing. 14:37:48 <raghavendrat> actually CI changed 8 months ago 14:38:03 <raghavendrat> migration from zuul v2 to v3 (or something similar) 14:38:08 <rosmaita> ok, i thought it might be something like that 14:38:17 <raghavendrat> my colleague worked on CI migration 14:40:13 <raghavendrat> regarding CI, i will provide updates on patch. any other query ? 14:41:04 <rosmaita> is hpe committed to getting CI for both 3PAR and Primera for both iscsi and FC? 14:42:53 <raghavendrat> sorry. i am not CI expert. i need to check with my colleague. 14:43:22 <rosmaita> ok 14:43:43 <rosmaita> for now, we probably don't want to merge that change 14:44:06 <raghavendrat> ok. 14:44:27 <rosmaita> let's give you some time to get the CI together, and we can revisit in a few weeks 14:44:39 <lseki_> raghavendrat: I downloaded and extracted the .tar.xz, and it's missing the cinder.conf file 14:44:39 <lseki_> it's a nice to have, although it's possible to see which back ends are enabled looking at c-vol log 14:45:30 <raghavendrat> sure. thanks 14:45:44 <rosmaita> well, a major issue is that it slows down reviewing if the info isn't easily available 14:46:22 <rosmaita> so raghavendrat it would be good for you to convey that message to your colleage 14:46:54 <rosmaita> ok, thank you 14:47:00 <rosmaita> #topic open discussion 14:50:01 <raghavendrat> sorry, some network issue. 14:50:16 <rosmaita> np 14:50:38 <rosmaita> looks like no topics for discussion? 14:50:41 <TusharTgite> hi 14:50:48 <rosmaita> hello 14:51:53 <TusharTgite> i'm planning to do a feature in the wallaby cycle so for that i have some question 14:52:00 <rosmaita> ok 14:52:25 <TusharTgite> the is a limitation in volume multi attach . Retyping an in-use volume from a multiattach-capable type to a non-multiattach-capable type, or vice-versa, is not supported. 14:52:51 <TusharTgite> so it is possible to work on this 14:54:16 <smcginnis> I seem to remember there being a very good reason not to allow that for in-use volumes. 14:54:30 <TusharTgite> if anyone can suggest some other thing apart from this if its not possible to do that will be helpfull for me 14:54:44 <whoami-rajat___> I'm not an expert on this but this might have a lot of complications when the in-use volume is actually mutliattached to multiple instances or a read/write operation is going on simultaneously 14:55:05 <smcginnis> So I guess a spec that goes through issues and how it would be addressed would be useful. 14:55:30 <lseki_> hmm I can't imagine what's the expected behavior when a multi-attached volume is retyped to a non-multiattach volume type 14:55:35 <smcginnis> It may have had something to do with nova needing to know at attach time if a volume can be multiattach. So if it's in-use, that might not be changed. 14:55:44 <rosmaita> would be good to have a discussion at the PTG if you are interested in pursuing this 14:57:03 <TusharTgite> ok, because in Victoria cycle I've only do doc patches so I'm looking for a feature item in upcoming cycle. 14:58:15 <smcginnis> The best features to implement are ones that users are actually asking for. I don't have a list or anything, but if there is a specific need for something identified, that makes it much more likely to get attention and get changes in. 14:59:09 <rosmaita> eharney has proposed a spec for "reset state robustification" 14:59:19 <rosmaita> that might be good to work on 14:59:20 <rosmaita> https://review.opendev.org/#/c/682456/ 14:59:23 <smcginnis> ++ 14:59:26 <eharney> yes 14:59:48 <smcginnis> <1 minute left 14:59:49 <rosmaita> TusharTgite: take a look at that patch and see if you're interested 14:59:57 <jungleboyj> The multi-attach would be an interesting discussion. 15:00:01 <rosmaita> we can discuss at the PTG if you need more details 15:00:03 <jungleboyj> Lots of landmines there though. 15:00:10 <rosmaita> ok, out of time 15:00:15 <rosmaita> thanks everyone! 15:00:17 <whoami-rajat___> Thanks! 15:00:20 <TusharTgite> ok i'll look into that patch 15:00:20 <rosmaita> #endmeeting