16:00:47 #startmeeting Cinder 16:00:48 Meeting started Wed Mar 27 16:00:47 2019 UTC and is due to finish in 60 minutes. The chair is jungleboyj. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:49 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:52 The meeting name has been set to 'cinder' 16:00:56 Hi 16:00:57 hi 16:01:03 hi 16:01:04 hi 16:01:07 o/ 16:01:07 o/ 16:01:08 courtesy ping: jungleboyj diablo_rojo, diablo_rojo_phon, rajinir tbarron xyang xyang1 e0ne gouthamr erlon tpsilva ganso patrickeast tommylikehu eharney geguileo smcginnis lhx_ lhx__ aspiers jgriffith moshele hwalsh felipemonteiro lpetrut lseki _alastor_ whoami-rajat yikun rosmaita enriquetaso hemna _hemna 16:01:08 o/ 16:01:12 @! 16:01:12 <_pewp_> jungleboyj ( *՞ਊ՞*)ノ 16:01:13 hi! o/ 16:01:16 hi 16:01:17 hi 16:01:18 Hi 16:01:25 hello 16:01:37 <_alastor_> o/ 16:02:33 Hey everyone. 16:02:49 Looks like a good showing. 16:02:52 Thanks team. 16:03:03 o/ 16:03:10 Happy Wednesday. 16:04:14 So, lets get started. 16:04:18 First ... 16:04:42 As you guys noted last week, I will be PTL for another cycle. 16:04:54 Thank you for your vote of confidence? :-) 16:05:00 yay! 16:05:07 jungleboyj: could you provide the etherpad link please. 16:05:17 #link https://etherpad.openstack.org/p/cinder-stein-meetings 16:05:23 e0ne: Thanks. 16:05:31 jungleboyj: thank you for your service 16:05:32 jungleboyj, whoami-rajat: np 16:05:50 Things have been a little crazy the last month or two and I am appreciative to all who have helped keep things moving forward. 16:05:57 thanks 16:06:04 Working to schedule things better for this release so that things will go a bit smoother. 16:06:15 So, looking forward to riding the Train with you all. 16:06:39 Thanks to smcginnis for covering me last week while I was dealing with a family issue. 16:06:46 Ok, moving on ... 16:06:54 #topic announcements 16:07:08 We are in RC2 week for Stein. 16:07:23 #link https://releases.openstack.org/stein/schedule.html 16:07:49 I think we have had a couple of things backported to stable/stein. 16:08:08 Need to look and see if it is worth cutting an RC2 for. 16:08:19 rosmaita: Have any thoughts there? 16:08:39 i think we should, unless there's more coming in 16:08:50 there are a bunch of patches to master that haven't been backported 16:09:03 (some may be train stuff, though) 16:09:26 We may want to wait a little to see if any translations come in. 16:09:33 Ok. 16:09:40 good point, they came into master pretty recently 16:09:50 So, lets plan to cut an RC2 tomorrow. 16:10:09 If you have anything that you need to get to stable/stein, please propose the backports ASAP. 16:10:56 Make sense to everyone? 16:11:45 smcginnis: the stein translations will be proposed there directly, right? we don't backport the ones that came into master? 16:12:00 I will take that as a yes. Cores, please watch for any stable/stein patches. 16:12:08 jungleboyj: should i propose the filtering patch to stein? 16:12:16 rosmaita: yes, we don't need to backport translations manually 16:12:22 ty 16:12:32 whoami-rajat: Can you remind me which one that is? 16:12:44 https://review.openstack.org/#/c/641717/ 16:13:43 whoami-rajat: I am ok with that. Any objections eharney hemna_ ? 16:13:55 sounds like a good idea to me 16:14:04 Ok. whoami-rajat Please do that. 16:14:24 jungleboyj: eharney done https://review.openstack.org/#/c/648174/ 16:14:43 :-) 16:14:54 whoami-rajat: Cool. Will give it a +2 once it passes the check. 16:14:57 that cherry-pick button was a good invention 16:15:03 :) 16:15:07 jungleboyj: thanks 16:15:15 * jungleboyj keeps forgetting about that. 16:15:19 rosmaita: agreed! 16:15:35 Anything else there? 16:16:14 Ok. Moving on. 16:16:33 I now have the mid-cycle summary up. I am sorry that this took so long to get in place. 16:16:44 #link https://wiki.openstack.org/wiki/CinderSteinMidCycleSummary#Introduction' 16:16:50 #link https://wiki.openstack.org/wiki/CinderSteinMidCycleSummary#Introduction 16:17:09 I have summarized all the todos. Please take a look and see if you have a todo and take it. 16:17:16 I have started working on my todos already. 16:17:42 If there are unassigned Actions, please consider taking one. 16:18:18 Thank you to all who have helped there and attended. 16:18:27 Final announcement. 16:18:36 Please help with the Denver PTG planning. 16:18:47 #link https://etherpad.openstack.org/p/cinder-train-ptg-planning 16:19:20 I have populated some topics on there based on what I found in the Actions from the mid-cycle. 16:19:44 The more topics we can get proposed, the more productive a time we will have. 16:20:06 Hopefully people saw that the strawman schedule for the PTG has been announced. 16:20:32 i managed to miss that, i need to adjust my email filters 16:21:07 jungleboyj: i think eharney assigned the untyped volumes spec to me? 16:21:24 I was just looking for the schedule 16:22:19 #link http://lists.openstack.org/pipermail/openstack-discuss/2019-March/003955.html 16:22:52 https://usercontent.irccloud-cdn.com/file/AC7oMdJ5/image.png 16:23:04 So, we have Thursday afternoon through Saturday. 16:24:00 Who all is planning to attend at this point in time? 16:24:01 #link https://www.openstack.org/ptg/#tab_schedule 16:24:03 I will be there. 16:24:07 +1 16:24:23 +1 16:24:42 +1 16:24:48 yes 16:25:07 Ok, that is a good start. 16:25:12 I think geguileo as well? 16:25:18 no, unfortunately 16:25:31 Ok, so hopefully he can attach remotely at least. 16:25:33 jungleboyj: nop, I'll just attend the summit 16:25:53 Oh, Summit only? Ok. 16:26:01 jungleboyj: it was my best friend's wedding during the PTG :-( 16:26:15 Oh. Well that is fun. 16:26:25 but Cinder is also your best friend 16:26:33 :) 16:26:37 * jungleboyj laughs 16:26:38 you can give him a tarball as a wedding present 16:26:38 eharney: but it's not getting married ;-P 16:26:46 geguileo: Cinder loves you. 16:26:56 rofl 16:26:57 rosmaita: ++ 16:27:11 rosmaita: That is wonderfully nerdy. 16:27:27 and probably more useful than a toaster? 16:27:32 Ok, so we will either chat during the Summit or something like that. 16:27:46 We will need to get together for dinner before you leave. 16:27:50 As a team. 16:27:59 I would love that :-) 16:28:04 Sounds good. 16:28:28 Any other discussion around the PTG? 16:29:02 Ok. 16:29:04 I’ll be at the summit too 16:29:14 xyang: Oh cool. Will be good to see you. 16:30:15 whoami-rajat: Will you be able to look at the untyped volumes discussion? 16:31:14 jungleboyj: yes. 16:31:25 Ok. Good. Thank you. 16:31:33 Ok. Moving on for real now. 16:31:54 #topic Stein cinderclient update 16:31:59 rosmaita: 16:32:18 #link https://review.openstack.org/#/q/topic:stein-ketchup+AND+project:openstack/python-cinderclient 16:32:23 we had a small snafu with where the cinderclient stable/stein branch was cut 16:32:36 anyway, thanks to everyone who has reviewed the patches 16:32:39 only 2 left 16:32:52 #link https://review.openstack.org/645583 16:33:02 that one i think should go in, has 1 +2 16:33:27 the other one i thought might cause backport problems if we left it out, but now i think it's probably ok to skip: 16:33:37 #link https://review.openstack.org/645564 16:33:59 Ok, I will approve the first one unless someone has a major concern. 16:34:15 I think the objectives are overly picky. 16:34:35 so we've basically caught stable/stein up with master at this point 16:34:35 *objections* 16:35:27 Ooops, both Sean and I +Wd. 16:35:32 Well, we made our statement. 16:35:38 rosmaita: Great. Thank you! 16:35:39 Super approval! 16:35:43 my ask is that if people agree that https://review.openstack.org/645564 is not necessary, please -2 it and i can abandon it 16:35:49 dah dah daaaah! 16:36:10 rosmaita: Yeah, I think that one can be abandoned. It's useful, but less so on a stable branch. 16:36:25 cool 16:36:28 that's all from me 16:36:34 smcginnis: ++ 16:36:54 rosmaita: Besides, I don't really trust the guy that wrote it. :) 16:37:05 :D 16:37:18 He he. 16:38:00 rosmaita: So once the last couple of things are merged you will do a new release? 16:38:18 i think i have to wait until after the coordinate Stein release? 16:38:32 rosmaita: Yeah, that sounds right. 16:38:36 or did we want to propose an exception and update global requirements? 16:39:14 I don't think there's anything release critical for stein. So we can just do a stable release once things open up again. 16:39:20 agree 16:39:20 it seems ok to wait to me 16:39:22 smcginnis: ++ 16:40:34 Ok. So, last topic. 16:40:46 #topic Consistency Groups removal 16:40:48 e0ne: 16:40:58 thanks, jungleboyj! 16:41:15 I just want to sync with the team on this topic 16:41:17 Welcome. :-) 16:41:34 are we going to remove CGs support in Train release? 16:42:01 CGs that aren't through the generic groups API, i assume 16:42:10 ? 16:42:20 eharney: exactly 16:42:29 So, the way we left it at the mid-cycle was that smcginnis was going to investigate and see if it was a possibility with the goal of dropping it in Train. 16:42:47 I think smcginnis took that todo. 16:43:30 so it looks like it's safe to remove it from horizon 16:43:46 horizon supports generic groups now 16:44:14 e0ne: isn't it better to remove it from cinder first? 16:44:27 whoami-rajat: sure 16:44:55 e0ne: Yeah, I think it is probably best to make sure we remove it from Cinder and then remove it from Horizon. 16:44:56 horizon now follows the same deprecation policy and projects 16:45:09 jungleboyj: +1 16:45:58 I thought we had already disabled the CG APIs 16:46:10 I did not get a chance to dig through that more. 16:46:39 Oh, we did mark it deprecated in the api-ref at least: https://developer.openstack.org/api-ref/block-storage/v3/index.html#consistency-groups-deprecated 16:46:50 So I think it makes sense for Horizon to stop using them. 16:47:20 Ok, so that means we can remove them in Train then? 16:47:33 afair, it's deprecated, but still enabled 16:48:06 The problem is microversions. We can bump a microversion that drops them, but would still need to have it there for older MVs. 16:48:46 smcginnis: :( 16:49:18 That sounds kind of yucky. 16:49:32 smcginnis: isn't there a way to remove a feature if no more used in MVs? 16:49:58 but CGs are in API extensions. we don't nave microversions for them 16:49:58 The only thing that gains us is - if we do what was discussed at the Forum in Vancouver - someday we could decide that microversion 3.87 (example) has been stable for awhile and make a v4 API starting from that point. 16:50:31 whoami-rajat: I suppose for older versions we can keep the API endpoints there and just have them return some 400 error. 16:50:35 if we have to keep the CG APIs around, can they not be mapped into the generic group code? (i feel like we considered this at one point) 16:50:46 Or maybe something like a 301 with reference to the new API? 16:51:16 eharney: I do feel like we considered that. 16:51:37 Maybe we did. 16:51:59 Just noticing, other than a small change for the update call, this was a contrib API. So maybe we can drop it. 16:52:00 So, given that the API is deprecated it seems like Horizon could remove it. Anyone disagree there? 16:52:17 Since those contrib APIs were not "official". 16:52:26 :-) 16:52:56 yes, no sense in horizon supporting a deprecated API 16:52:59 smcginnis: :) 16:53:22 is it really a contrib API..? 16:53:36 e0ne: i think horizon completely supports volume groups in project and admin panel now? 16:53:40 Like quotas and a few other core things. 16:53:50 So, lets land on that agreement and work on solving this problem in Denver? 16:53:56 I did verify, the cg API does call through to the group code. 16:54:05 whoami-rajat: yes, since Stein 16:54:11 But either way, I agree Horizon should stop using the deprecated APIs. 16:54:15 https://github.com/openstack/cinder/blob/master/cinder/api/contrib/consistencygroups.py 16:54:39 https://github.com/openstack/cinder/blob/master/cinder/api/contrib/consistencygroups.py#L83 16:55:43 5 minutes 16:55:56 https://github.com/openstack/cinder/blob/master/cinder/api/v3/consistencygroups.py doesn't look very contrib-y, but i'm not sure 16:56:03 smcginnis: great! 16:56:17 eharney: Yeah, that kind of uncontrib'd it I guess. 16:56:43 So, lets get it out of Horizon. 16:56:52 anyway, it's deprecated and uses generic groups api inside. we can drop it from horizon now :) 16:57:08 And decide what to do in the future. 16:57:13 e0ne: ++ 16:57:58 e0ne: good call. +1 to remove call to cg api 16:58:00 Ok, anything else we need to cover in the last couple of minutes. 16:59:40 I guess not. 16:59:51 Thanks to everyone joining. 16:59:54 Thanks! 17:00:00 see you next week 17:00:06 Have a good rest of your week and thanks for helping get Stein out! 17:00:12 quick reminder to stable cores: https://review.openstack.org/648174 17:00:19 #endmeeting