14:00:33 #startmeeting cinder 14:00:33 Meeting started Wed Sep 22 14:00:33 2021 UTC and is due to finish in 60 minutes. The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:33 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:33 The meeting name has been set to 'cinder' 14:00:39 #topic roll call 14:00:41 Hi 14:00:58 hi 14:00:59 hi 14:00:59 hi 14:01:01 Hi 14:01:26 Hi 14:02:09 hi 14:02:09 hi 14:02:29 hi 14:02:38 hello everyone 14:02:42 let's get started 14:02:51 #link https://etherpad.opendev.org/p/cinder-xena-meetings 14:03:01 #topic announcements 14:03:10 we will be doing an RC-2 14:03:33 there are some requirements changes that we needed to make, and some release critical bugs 14:03:45 we'll go over those in a few minutes 14:03:55 i'll just continue with the announcements 14:04:10 hi 14:04:11 next week is the final Release Candidate 14:04:23 traditionally on Thursday 14:04:54 we'll talk about whether we should do an RC-2 quickly and then possibly another, or just do an RC-2 14:05:05 next announcement: 14:05:17 next week's meeting is the last of the month, so will be our usual video meeting 14:05:52 next, the Outreachy proposal deadline is 29 September 14:06:06 #link http://lists.openstack.org/pipermail/openstack-discuss/2021-September/024992.html 14:06:17 you can read ^^ to see what that's about 14:06:46 i don't know whether i will put in a proposal or not 14:07:14 it's a good mentoring opportunity, if you would like to do that 14:07:25 and might get some work done on cinder 14:07:46 let me know if you are interested, and i will work with you to get a proposal together 14:08:01 but think about it fast, because the proposal is due in one week 14:08:18 finally, the Yoga PTG is the week of 18 October 14:08:28 #link https://etherpad.opendev.org/p/yoga-ptg-cinder-planning 14:08:33 add topics to ^^ 14:08:43 also, info is on the etherpad for registering for the PTG 14:09:01 it's free, but the foundation asks you to register 14:09:22 i guess so that they can remind you about the code of conduct 14:09:33 i think that's all 14:10:03 #topic RC-2 14:10:09 #link https://etherpad.opendev.org/p/cinder-xena-final-patches-for-RC 14:10:15 ok, please open ^^ 14:10:49 since the stable/xena branch has been cut, everything must merge first to master and then be backported to stable/xena 14:11:21 there are a few requirements changes we have to make 14:11:36 namely, raising the minimum in requirements.txt 14:12:07 i think everything has merged to master at this point, but we need to get teh backports merged too 14:12:22 so stable cores, please note the etherpad and keep an eye on it 14:13:05 after we found a few minimum versions of some requirements that broke tests, i put together a quick lower-constraints kind of job 14:13:16 #link https://review.opendev.org/c/openstack/cinder/+/810131 14:13:26 ^^ if anyone wants to check my work 14:13:51 all the tests passed with the current xena backports and the generated lower constraints file 14:14:17 so, everything seems to work with the current requirements minima (after the patches on lines 1-18 on the etherpad merge) 14:14:35 but, i noticed a few things that might be a bit out of whack: 14:14:46 #link https://paste.opendev.org/raw/809501/ 14:14:56 so, please open ^^ so we can discuss 14:15:32 btw, i did not check all our requirements against current upper constraints, just some that i thought might be worth looking at 14:16:08 you can see that for pretty much all the oslo stuff, there are only minor version changes, so i don't think we need to worry about them 14:16:19 the exceptions are oslo.db and PyMySQL 14:16:45 i think our PyMySQL req is only for tests so it's not very interesting 14:16:46 here's the change log for oslo.db: 14:16:49 #link https://paste.opendev.org/raw/809503/ 14:17:48 there doesn't look like a lot in there, but i wonder whether we should raise the min to the xena release of oslo.db ? 14:17:55 to remind you what this means: 14:18:12 we are testing (usually) with whatever is in upper constraints 14:18:29 so all the recent jobs are using oslo.db 11.0.0 14:18:33 for the past few weeks 14:19:05 but if the min in requrements is 8.4.0, then we are saying that it's ok to package cinder with oslo.db 8.4.0 14:19:26 and that may be fine, all our unit tests pass with that version 14:19:33 we should raise it IMO 14:19:59 eharney: ty, i agree 14:20:02 anyone else? 14:20:05 +1 14:20:11 thanks ivan 14:20:19 ok, i will put up a patch 14:20:39 #action rosmaita patch to raise oslo.db min in requirements to 11.0.0 14:20:49 ok, the other one is PyMySQL 14:20:55 the changelog is here: 14:21:09 #link https://github.com/PyMySQL/PyMySQL/blob/main/CHANGELOG.md 14:21:36 it looks to me like the major changes are to remove py2.7 and py3.5 support 14:22:10 and stop supporting MySQL 5.5 (which we talked about last week!) 14:23:36 i think we should probably raise this one too ... any thoughts? 14:24:33 i'm not sure we should depend on it directly 14:24:35 see https://opendev.org/openstack/oslo.db/src/branch/master/setup.cfg#L25 14:24:45 what version of mysql is available in official repos of supported OSes? 14:25:07 e0ne: not sure 14:26:06 eharney: you have a good point, we should probably let oslo.db set the version 14:29:25 ok, so for pymysql i propose to do nothing 14:29:42 not as big a deal, it's in test-requrements whereas oslo.db is in real requirements.txt 14:29:55 sounds good 14:30:03 ok, thanks 14:30:09 on to some release critical bugs 14:30:24 line 23 in https://etherpad.opendev.org/p/cinder-xena-final-patches-for-RC 14:31:03 this was reported for nova, but since we have a lot of the same changes for the sqlalchemy-migrate -> alembic transition, it probably applies to us too 14:31:35 #link https://review.opendev.org/c/openstack/cinder/+/810288 14:31:35 #link https://review.opendev.org/c/openstack/cinder/+/810000 14:31:42 the two patches for master look ok, one is in recheck 14:31:45 thanks enriquetaso 14:32:12 so i think we need to get these reviewed, merged, and backported befor RC2 14:32:54 next one is the cgroupv2 thing: https://bugs.launchpad.net/cinder/+bug/1942203 14:33:09 i will try to get a patch up for that one in the next day or two 14:33:28 i'm not sure how many people actually use the throttling 14:33:43 but it would be better to have it fixed for the release than not 14:34:07 next one is a bug that maybe requires a "known issue" release note 14:34:17 #link https://bugs.launchpad.net/cinder/+bug/1909207 14:35:13 i think at one point we were thinking this might not be a bug? 14:35:47 1909207 i think this issues isn't a rbd problem so far 14:36:20 this probably does need more investigation 14:36:33 it was on my list at one point but clearly i lost track of it 14:36:54 maybe we can ask for the backend xuanyandong is using 14:37:24 we can just try to reproduce it 14:37:46 and investigate if this is a thing that can happen 14:37:51 does anyone have time to look at this before the next meeting? 14:39:00 we can always have in "known issues" that this has been reported and may affect some backends 14:39:25 sounds good, I'm going PTO but I can try to reproduce it for next meeting. 14:39:40 merging a release note is pretty fast, so we can decide next week 14:39:44 thanks enriquetaso 14:40:09 since enriquetaso will be on pto, though, if anyone else is interested in this, would be good to help out 14:40:46 ok, we have another known issue that we discussed at the midcycle 14:40:58 #link https://wiki.openstack.org/wiki/CinderXenaMidCycleSummary#what_to_do_about_qcow2-v2_support_being_yanked_from_various_linux_distros 14:41:10 there was an action item to look into this 14:41:33 what i found out is that we do look at whether the volume being managed is actually a qcow2 14:41:44 but we only reject it if it has a backing file 14:41:51 (because that's a security issue) 14:42:26 i will write up a release note patch and bug people to review it 14:42:55 as far as issues for the xena release, that's all i'm aware of 14:43:02 anyone here run into anything? 14:44:05 enriquetaso: that rbd trash bug might be a candidate 14:44:48 #link https://review.opendev.org/c/openstack/cinder/+/810402 14:45:19 yes 14:45:26 I think we may want ^ merged before RC1 14:45:46 well, RC2, anyway 14:46:03 let's discuss at the bug meeting, then 14:46:14 sure 14:46:30 #topic deprecations 14:46:45 I forgot to bring up the RateLimitingMiddleware deprecation last week, but since we're doing an RC-2, we still have time to decide for Xena 14:47:03 i sent an email to the ML aimed at operators 14:47:10 the response was ... nothing 14:48:01 So, it would be nice to assume that means it isn't being used. 14:48:32 my feeling is that since nova doesn't have it any more, anyone interested in this is using a different solution for nova, and probably that same solution for any other services that need rate limiting 14:49:11 ++ 14:49:15 That makes sense to me. 14:49:17 i think eharney was skeptical about removing this, anyone else? 14:50:10 one thing that did come up is that sfinucane had a docs patch up for this that's been sitting for a while 14:50:26 which we don't need if we deprecate the rate limiting middleware 14:50:51 but made me think that we are missing some docs patches in the Festival of Reviews 14:51:19 so we may want to start emphasizing that 14:51:44 anyway, i guess my ask is that if we are going to deprecate this, we need to do it this week 14:51:57 so please leave comments on the patch: 14:52:04 https://review.opendev.org/c/openstack/cinder/+/807469 14:52:37 #topic stable release update 14:52:40 whoami-rajat: that's you 14:52:46 thanks 14:53:22 so there's a patch up for victoria release, which i had to revise since it had some patches that doesn't exist in wallaby 14:53:47 and in Ussuri, there was one patch left which is approved and in gate, so I've proposed the ussuri release patch as well 14:54:09 victoria 14:54:10 #link https://review.opendev.org/c/openstack/releases/+/809892 14:54:17 ussuri 14:54:20 #link https://review.opendev.org/c/openstack/releases/+/810442 14:54:35 since train is EM, we won't be doing any release for it 14:54:57 thanks everyone who helped with the reviews and made these releases possible! 14:55:00 for victoria, you mean there were patches that are in stable/wallaby, but were not included in the most recent wallaby release? 14:55:21 so you couldn't include them in the victoria release, either? 14:55:31 yes, (I always word it wrong), they are merged in wallaby but not there in recent wallaby release 18.1.0 14:55:47 ok, cool, just wanted to be clear 14:56:17 I've recently updated the hash for the victoria release patch so would be great if you can take a look 14:56:30 ok, will do 14:56:34 that's all from me 14:56:35 thanks 14:56:48 thanks whoami-rajat 14:56:52 fwiw I think we are using the rate limit middleware 14:57:08 hemna: ok, that is good to know 14:57:14 please leave a comment on that review 14:57:28 sorry about the late response, too dang busy to follow everything 14:57:28 #topic os-brick bug ? 14:57:36 simondodsley: that's you 14:57:53 thanks - we can take this to the cinder channel due to time constraints if you want. 14:58:05 hemna: np, glad you brought i up because i was inclined to deprecate and remove 14:58:24 simondodsley: that's a good idea, since the bug squad meeting is in cinder channel in 2 minutes! 14:58:32 #topic open discussion 14:58:44 90 seconds for anything on anyone's mind 14:58:54 i started poking at volume manage + encryption re: the bug earlier -- think i already found an issue in there, will probably write up a new, specific, bug report 14:59:16 maybe we can use the recording of the bug meeting for the os-brick bug so keep track somewhere simondodsley 14:59:20 eharney: sounds good, ty 15:00:13 well, that's all we have time for 15:00:15 thanks everyone 15:00:23 don't forget, video meeting next week 15:00:34 Thanks! 15:00:45 see you next week 15:00:51 ttfn 15:00:58 #endmeeting