14:00:02 #startmeeting cinder 14:00:02 Meeting started Wed Aug 30 14:00:02 2023 UTC and is due to finish in 60 minutes. The chair is whoami-rajat. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:02 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:02 The meeting name has been set to 'cinder' 14:00:05 #topic roll call 14:00:07 o/ 14:00:12 o/ 14:00:15 o/ 14:00:37 o/ 14:00:40 o/ 14:00:41 o/ 14:00:58 o/ 14:01:03 o/ 14:01:10 o/ 14:01:16 o/ 14:02:12 o/ 14:02:14 #link https://etherpad.opendev.org/p/cinder-bobcat-meetings 14:02:19 Hi 14:03:03 hello everyone 14:03:06 let's get started 14:03:12 #topic announcements 14:03:24 first, Nominated myself for Cinder PTL during 2024.1 Cycle 14:03:32 #link https://lists.openstack.org/pipermail/openstack-discuss/2023-August/034902.html 14:03:54 :+1 14:03:55 \o/ 14:04:01 o/ 14:04:19 I've nominated for the cinder PTL for 2024.1 (Caracal) cycle and if no one else nominates themselves, I will be your PTL again for next cycle :) 14:05:11 good to see people happy after 3 cycles :D 14:05:20 that does add some motivation for the next cycle 14:05:38 whoami-rajat: Thank you for continuing! 14:06:16 :) 14:06:25 ok moving on 14:06:28 next, os-brick 6.4.0 released 14:06:34 #link https://review.opendev.org/c/openstack/releases/+/892213 14:06:39 #link https://pypi.org/project/os-brick/6.4.0/ 14:06:49 last week was non-client library release 14:07:04 so os-brick was part of that release cycle 14:07:20 the patch was merged on monday but proposed last week so technically we were on time 14:07:32 anyway, os-brick is released with it's latest version 6.4.0 14:08:10 finally we have some upcoming deadlines 14:08:16 Bobcat-3 milestone (feature freeze): August 31st (R-5 week) 14:08:16 RC1 deadline: September 14th (R-3 week) 14:08:17 Final RC deadline: September 28th (R-1 week) 14:08:17 Final 2023.2 Bobcat release: October 4th, 2023 14:08:30 We are going to discuss the feature freeze as a topic 14:08:52 that were all the announcements, I couldn't find any other thing of interest on the mailing list 14:09:03 but i remember we had a gate issue with tox that rosmaita was referring to 14:09:18 which seems to be fixed now, but if you would like to highlight it? rosmaita 14:09:47 there was an issue with a dependency of tox, got fixed by a new release of that dependency and a new release of tox 14:09:55 so 4.11.0 should be good 14:10:09 thanks! 14:10:20 so if you see any gate failures due to tox, you can recheck and it should be good 14:12:38 also I'm working on the mid cycle summary, it should be up today or tomorrow, I will update on the mailing list 14:14:04 and a reminder that if you haven't registered for the virtual PTG in October, here is the link 14:14:06 #link https://openinfra.dev/ptg/ 14:15:08 that's all for the announcements 14:15:19 anyone has anything else? 14:15:25 for the announcements i mean 14:17:36 ok, let's move to topics 14:17:44 #topic Feature Discussion 14:17:52 #link https://etherpad.opendev.org/p/cinder-2023.2-bobcat-features 14:18:15 it would have been easier in the video meeting that was planned for today but since we met last week for midcycle, I thought of skipping this week's video meeting 14:18:37 anyway, let's get started with discussing the features one by one 14:18:39 first, Add the os-extend_volume_completion volume action 14:18:48 #link https://review.opendev.org/c/openstack/cinder/+/873557 14:19:13 The reason I've held reviews on this is that cinder calls Nova to extend the volume (in case of FS drivers) 14:19:22 last time i checked this morning, nova side changes haven't merged yet 14:19:37 so I'm skeptical of merging the cinder changes since the extend operation might start failing 14:19:54 is the author of this feature around? 14:20:21 i think he just disconnected ... 14:21:39 here is the nova patch 14:21:41 #link https://review.opendev.org/c/openstack/nova/+/873560 14:22:16 sorry, i disconnected just at the wrong moment 14:22:25 kgube, no worries 14:22:42 kgube, I was pointing out the dependency on nova for this feature which is still not merged 14:22:53 so the nova change is actually dependent on the first cinder change 14:23:06 the cinder change is multi-part 14:23:56 the first one just adds the new volume action 14:24:15 then there is the nova change that depends on that 14:24:56 and then there is the second cinder change that relies on the new behavior of nova: https://review.opendev.org/c/openstack/nova/+/873560 14:25:34 this is reflected in the gerrit dependencies 14:26:17 and Nova actually held back reviews because the first cinder change was still open :/ 14:28:00 ok, thanks for clarifying that 14:28:15 So I guess I should have tried to communicate this more clearly 14:28:54 kgube, another question is, if we merge the first change, will it impact the current workflow we have? I haven't gone through the code yet so not sure if it's making any functional changes in the extend workflow 14:30:12 It won't impact the current workflow, as far as I can tell 14:31:01 but it adds dead code for now 14:31:16 ok, a related question is, how flexible is nova on granting an FFE for the nova changes? 14:31:44 looks like there is only one change from nova side 14:32:02 I don't think it's a priority feature for them 14:33:45 hmm, we could grant an FFE but it would be useless until nova is also willing to do the same 14:34:01 we will be just adding a microversion and an API call that is never made 14:34:15 they told me last week that either the dependencies needs to merged asap, or they aren't going to allocate review time to it 14:34:59 ok, i wasn't aware but last week's midcycle was a good place for this discussion 14:35:29 Yeah, I couldn't make it, unfortunately 14:35:34 not denying the fact that we lack review bandwidth and that also needs to be addressed, but that's a separate topic for now 14:36:39 at this moment, even if we merge the first cinder change (given it doesn't require changes to be made), we need commitment from nova that they will merge their side of changes in this cycle 14:36:49 that's my thought on it at least 14:37:08 would like to see if the team has other comments on it 14:38:29 i think we need at least one +2 on the patch before nova will take it seriously 14:42:34 I can commit on reviewing it, since mostly it's just adding a new action 14:43:11 if bauzas is around, he can provide us the nova team's thought on the feature 14:43:52 * bauzas scrolling up 14:44:12 bauzas, thanks for joining, i will summarize the discussion 14:44:15 well, we haven't defined any FFE period 14:44:27 and we don't plan it 14:44:45 we're discussion about the feature https://review.opendev.org/c/openstack/nova/+/873560 14:45:13 the nova consensus is more about letting the feature to be merged in Cinder ideally during the Bobcat release so we can merge it quickly on Caracal on our side 14:46:12 hmm, the problem i see is that will introduce a new microversion on the cinder side in the bobcat release that is unused ... 14:46:33 maybe it's not a big issue, but i would avoid having that in our cycle highlights 14:48:07 kgube, if the feature is going to be used in Caracal by nova, would you mind if the cinder changes are also part of the same cycle? 14:49:11 whoami-rajat, I mean, its not ike we can use the feature, when only half of it is merged 14:49:33 so, no, I guess not 14:50:02 but It would be nice if we can avoid keeping it to the last minute, next cycle 14:50:08 thanks kgube 14:50:13 sure 14:50:17 so I will plan on getting this reviewed and merged by milestone-1 of Caracal 14:50:41 kgube: will you be able to repropose the feature as soon as I'm opening the dev cycle on Caracal ? 14:50:51 this won't be long, a matter of 2.5 weeks 14:51:00 bauzas: sure! 14:51:28 maybe you don't know, but Caracal feature development cycle opens as soon as we deliver Bobcat RC1 14:51:29 thanks bauzas and kgube for the discussion 14:52:02 ah, that's good to know 14:52:12 https://releases.openstack.org/bobcat/schedule.html#rc1-target-week 14:52:19 which is planned Sep 11th 14:52:45 after that, I'll shortly create the specs directory for caracal and the blueprint release tag name 14:53:14 ack, I will carry out the similar work for cinder 14:53:30 would be good if kgube can provide some reminders if we miss out on anything :) 14:53:57 whoami-rajat, I will try to be less shy about it :) 14:54:19 thanks again 14:54:22 kgube: thanks and yeah, don't feel you offense ourselves 14:54:40 thanks for your time bauzas , really appreciate it 14:54:42 we just have a reasonable amount of different priorities and somehow communication helps sorting them 14:55:40 ++ 14:56:02 so i would quickly like to go through some of the other features 14:56:14 Fujitsu Driver: Add QoS support 14:56:21 #link https://review.opendev.org/c/openstack/cinder/+/847730 14:56:38 I've already reviewed it previously, just need to go through the latest update 14:56:55 but we require another reviewer to take a look at this 14:57:17 Here I am. Previously, rajat provided some comments on it, and I have addressed those comments and passed both Zuul and our CI tests. 14:57:40 thanks for updating it inori1 , I will take a look shorty after this meeting 14:57:52 please add your name as a second reviewer here 14:57:55 #link https://etherpad.opendev.org/p/cinder-2023.2-bobcat-features#L13 14:58:15 next, NetApp ONTAP: Added support to Active/Active mode in NFS driver 14:58:24 #link https://review.opendev.org/c/openstack/cinder/+/886869 14:58:31 it has 2 +2s 14:58:37 but has a dependency on 2 other patches 14:58:40 Thank you. whoami-rajat 14:58:52 no problem 14:59:01 yes, the main patch got +2, and one of the dependent patch has zull failure and its not related. I re-triggered it now. It'll go through. 14:59:33 msaravan, ack, thanks for the update, I will take a look at the dependent patches 14:59:52 thank you Rajat. Much appreciated. 15:00:17 np 15:00:20 so we are out of time 15:00:29 can we have a possible FFE for this, if this is not closed tomorrow. 15:00:45 but I would like to provide a 1 week extension for all the remaining features 15:00:59 thank you 15:02:38 sorry i got disconnected 15:02:54 these were my last messages 15:02:56 but I would like to provide a 1 week extension for all the remaining features 15:02:58 since we have too many and too less reviews on it 15:03:02 I will send out a mail tomorrow regarding this 15:03:04 thanks everyone for joining 15:03:06 #endmeeting