14:00:16 #startmeeting cinder 14:00:16 Meeting started Wed Sep 7 14:00:16 2022 UTC and is due to finish in 60 minutes. The chair is whoami-rajat. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:16 The meeting name has been set to 'cinder' 14:00:24 Hello 14:00:25 o/ 14:00:25 #topic roll call 14:00:35 o/ 14:00:40 oops lol 14:00:53 o/ 14:00:57 o/ 14:01:09 o/ 14:01:20 hi 14:01:22 hi! o/ 14:01:49 #link https://etherpad.openstack.org/p/cinder-zed-meetings 14:02:08 hi 14:02:11 hi 14:02:45 hi 14:03:44 hello everyone 14:03:46 good turnout today 14:03:50 let's get started 14:03:57 #topic announcements 14:04:07 first, oslo.db 12.1.0 issue 14:04:13 #link https://lists.openstack.org/pipermail/openstack-discuss/2022-September/030317.html 14:04:50 so there was an issue with oslo.db new release, due to which a lot of project jobs failed on the requirement gate 14:05:28 Stephen worked on fixing all the gates and today everthing was fixed as I last checked 14:05:55 in any case, Cinder wasn't affected, probably because of Stephen's work on our db patches (Thanks to him) 14:06:27 so tl;dr we are going forward with oslo.db 12.1.0 release and not reverting anything there 14:06:38 next, Operator hour 14:06:45 #link https://lists.openstack.org/pipermail/openstack-discuss/2022-September/030301.html 14:07:00 There was an email regarding an operator hour to be included in our PTG 14:07:05 which seems like a very good idea 14:07:21 so every project has to book 1 hour slot for project-operator discussion 14:07:31 I think there are already placeholders in the PTG schedule page 14:07:49 #link https://ptg.opendev.org/ptg.html 14:08:11 so we will be having PTG from tuesday to friday, 1400-1700 UTC 14:08:29 sorry 1300-1700UTC 14:08:42 sounds a good idea 14:08:51 s/sounds/sounds like/ 14:08:59 so we've a slot everyday at 1300 or 1400 UTC, but not sure which day would be best to discuss 14:09:10 should we discuss it early in the PTG or towards the end? 14:09:38 possibly early, so we can follow up if necessary 14:10:07 that was my thought as well, also we can discuss topics with operator feedback in mind 14:10:13 operators appear to be very excited to be invited to join the project teams, by the way 14:10:45 that's a good thing 14:11:06 That is good! 14:11:06 they like to get into the light occasionally 14:11:18 \o/ 14:11:41 so if everyone agrees early is good, I will try to book a slot on tuesday/wednesday 14:11:51 +1 to early 14:12:34 great, will followup with infra team then 14:12:45 Makes sense. 14:13:08 #action whoami-rajat to followup to book slot for operator hour early in the PTG 14:13:16 okay moving on then 14:13:22 next, Feature freeze extended deadline this week - 09 September 2022 14:13:37 so we extended the feature freeze deadline last week to this week 14:14:01 I will elaborate on this in the upcoming topic but this is just a reminder 14:14:28 so everyone can get to reviewing things and authors to work on quick feedback 14:14:41 next, RC1 - 16 September 2022 14:14:50 so after we have sorted the features, we will focus on bug fixes 14:15:13 there are quite a few I've been procrastinating for the final week 14:15:27 but yeah, bug fixes would be our focus for next week 14:16:07 so anything else anyone would like to share related to announcements? 14:16:57 guess not 14:17:01 so let's move to topics 14:17:06 #topic Features update 14:17:17 #link https://etherpad.opendev.org/p/cinder-zed-features 14:17:28 so wanted to quickly discuss the current status of all the features proposed 14:17:34 first, Reset state robustification 14:17:40 #link https://review.opendev.org/q/topic:bp%252Freset-state-robustification 14:18:11 the patches are ready for review 14:18:26 honestly I don't have much idea about the area of code and would take a lot of time to review those 14:18:29 yes base patch already merged as well 14:19:09 any takers that can help get this merged this week? 14:19:52 also TusharTgite , if it's not critical for Zed, would you be OK to move it to next cycle? 14:20:03 we can prioritize it that way early next cycle 14:21:05 * whoami-rajat hears crickets 14:21:17 i think it's useful work but not critical for zed 14:21:20 it definitely going to take lot of time to review 14:21:37 reviewing this is supposed to be on my list, but i haven't looked at it in a bit 14:22:25 if other feature r on priority then ok but could yo please try before relase as we also merge small patches after freez as well 14:23:23 so this affects the core cinder code, other features are mostly driver features which doesn't require considering all cases that could cause problems in cinder 14:24:32 so I think it would be realistic to move it to next cycle and eharney is also interested in taking a look at these 14:24:39 ok but can we make it a priority from start of next cycle then 14:24:55 yes, will do that 14:25:13 shouldn't https://review.opendev.org/c/openstack/cinder/+/804035 have a release note? 14:25:37 I've been dealing with deadlines since the beginning so it's my bad, will try to prioritize things better next cycle 14:25:52 enriquetaso: I think so 14:26:29 i'm not sure we want a release note in each of the patches 14:26:57 maybe one at/after the last one merged during a release 14:27:06 I haven't looked at the whole series but one releasenote describing all the reset state changes should be good 14:27:14 or as eharney said 14:27:20 sure 14:27:31 ++ 14:28:19 cool, so we cna move to next feature 14:28:24 NetApp ONTAP: Implement support to ONTAP REST API 14:28:32 #link https://review.opendev.org/q/topic:bp%252Fnetapp-ontap-rest-api-client 14:29:03 NetApp team is moving their driver->backend communication from ZAPI to REST 14:29:12 so there are a bunch of changes to support it 14:29:22 I've taken a look at first two patches and currently both have +2s 14:29:47 mostly it's addition of new interface and tests, haven't noticed any change in the driver as of now (but might be in later patches) 14:30:24 they're quite big in terms of LOC but impact on cinder is minimal there so should be good for a quick look over the code 14:30:56 we tried to get changes separated in smaller patches so review attention can be splited, tho it will be a bit more difficult to merge since patches have a order 14:31:38 thank you for your reviews, whoami-rajat :) 14:31:49 thanks whoami-rajat ! :D 14:31:54 sfernand, yeah i see one patch was abandoned and the order doesn't show up correctly, maybe rebase the top changes after the first two changes merge 14:32:09 no problem, thanks felipe_rodrigues sfernand for your reviews on other changes 14:32:31 this was a very good example of you can do reviews on other patches to receive reviews 14:32:51 \o/ 14:33:05 as originally stated by rosmaita :) 14:33:06 :) 14:33:24 we are working on the comments provided, all changes will be submitted by the end of the day 14:33:36 great 14:33:54 next is another feature by netapp, NetApp NFS: Clone image using copy file operation 14:34:08 I've left a -1 since we're not deprecating the config option in the patch 14:34:12 else it looks good 14:34:31 actually, we are deprecating in a different patch: https://review.opendev.org/c/openstack/cinder/+/847733 14:34:39 they should be merged together during this release 14:34:51 yep 14:34:58 this patch that deprecates the old approach is pretty simple: just deprecate the config option 14:35:00 should we refer it in the commit message? 14:35:45 hmm, yeah they should've been linked somehow by a dependency chain but mentioning in the commit message also sounds good 14:36:11 ack 14:36:52 moving on, IBM SVf: Support for volumegroup functioanlity 14:36:57 #link https://review.opendev.org/c/openstack/cinder/+/852364 14:37:14 the CI hasn't reported and author said they're working on getting it configured 14:37:20 else the changes look reasonable 14:37:50 will upgrade my vote once the CI is functioning again 14:38:04 next, Infinidat: support for revert to snapshot operation and Infinidat: support for manage/unmanage API 14:38:18 there are 2 features by Infinidat team but my issue currently is 14:38:27 they've rebased them on top of a bunch of bug fixes 14:38:56 which is really not good if they want to merge their changes in time 14:39:24 I've blocked one bugfix on which rosmaita and geguileo also has shared their comments 14:39:31 is someone from Infinidat team here to discuss it? 14:40:10 what is the balance between multiple patches and different features/bugs 14:40:44 we've the feature deadline this week but bugs can merge next week as well 14:41:19 I mean if I have 2 bugs - should I do a patch for each? 14:41:33 or merge them at 1 patch 14:41:45 yes 14:41:52 you should have 2 patches 14:41:57 yuval__, if they're related you can do it in one patch but for different issues different patches are good 14:42:16 ok 14:42:17 it is much easier to review and to revert if necessary 14:42:34 yep 14:42:35 whoami-rajat: ++ 14:42:46 looks like noone is here from Inifinidat team so we can move on 14:42:51 overall I agree, I reviewed some patches which were a mess and I was wondering if I should comment about it 14:43:20 I've left a comment for them to separate out the features, let's see if it gets updated 14:43:34 yuval__, would be good to leave comments there stating it 14:43:43 ok 14:43:55 but there's no proper rule about it, just convention we follow to make the process easier 14:44:34 next, Remove legacyfacade 14:44:46 #link https://review.opendev.org/q/topic:remove-legacyfacade+project:openstack/cinder+is:open 14:45:08 there are still changes left for our DB work (they never end) 14:45:33 I'm not very positive about getting them in during the last minute, also since it affects DB 14:45:44 I would probably move and retarget them for next cycle 14:46:13 That sounds safest if they changes aren't urgent. 14:46:48 is that really a feature? maybe we could try get them merged by the end of next week if not 14:46:57 doesn't look urgent since cinder is compatible with new oslo.db release, which was the biggest concern with the legacy engine facade thing IIRC 14:47:28 sfernand, can't really say a feature but I'm skeptical about getting them in so close to the releae 14:47:31 release 14:47:53 got it 14:47:59 rosmaita, what do you think about it? (you've already reviewed the last series) 14:48:43 i was hoping we could get this out of our lives 14:48:58 but it is getting awful close 14:49:34 those remaining changes look mostly harmless, though 14:49:59 okay, we can retrospect them after we're done with the features 14:50:27 and if not feasible, we can target them next cycle M-1 14:50:52 sounds good 14:50:56 cool 14:51:02 next, DEMC PowerMax: Unisphere 10 support 14:51:10 it has 3 -1's and no update since 20 days 14:51:20 so thinking about deprioritizing this and moving to next cycle 14:51:44 and final one is New Quotas System 14:52:04 which geguileo being so busy throughout the cycle is OK to continue the work next cycle 14:52:25 whoami-rajat: yeah, sorry about that, but with the latest spec changes 14:52:34 the code got more complicated than I anticipated 14:52:50 and the UT fixes delayed my work on quotas 14:52:51 :-( 14:53:08 geguileo, no problem, you were already too occupied that this would've been impossible to complete 14:53:13 can't even remember how many things geguileo worked on this cycle 14:53:33 and that's a reminder to the UT fixes proposed by geguileo 14:53:35 #link https://review.opendev.org/q/topic:rbd-native-treads 14:53:42 this is one thing we need to target for next week 14:54:05 as everyone is seeing failures (me as well) in unit tests 14:54:11 s/failures/timeouts 14:54:37 something urgent to get fixed before the release 14:55:14 anyway, that's all from my side, sorry for taking the whole meeting time but this was kind of urgent 14:55:51 we still have some time for open discussion 14:55:57 #topic open discussion 14:56:07 i've got something 14:56:29 sure 14:56:29 the everyone-must-use-openstackclient-cli is rearing its ugly head again 14:56:42 may become a community goal for Antelope 14:56:42 :-) 14:56:59 here is the latest situation, they've added a lot of support: 14:57:05 #link https://docs.openstack.org/python-openstackclient/latest/cli/decoder.html#cinder-cli 14:57:17 but there are still some weirdnesses 14:57:29 i think mostly around volume/snapshot manage/unmanage 14:57:52 and it's not clear how microversion support works 14:58:27 yeah we will need to prioritize that next cycle since everyone is headed that way and we might be the only ones left blocking the effort ... 14:58:38 ++ 14:58:47 yeah, apparently even glance is getting on the osc train 14:59:19 yeah, but looking at the glance gaps, the OSC support is mostly empty 14:59:19 Wow. Well, I guess that is good we are all going that direction. 14:59:27 so might take them more than a cycle to get there 14:59:39 i guess my main ask here is if anyone is fundamentally philosophically opposed to osc replacing the cinder cli 15:00:16 I think the main issue in the past was the lack of parity. Looks like they are getting much closer. 15:01:03 that's all from me 15:01:10 thanks rosmaita for bringing this up 15:01:27 we need a closer look at the OSC equivalent to find any issues 15:01:34 time check 15:01:34 but something we should prioritize 15:01:41 and we're overtime 15:01:44 thanks everyone for joining 15:01:53 focus on feature reviews! 15:01:55 #endmeeting