16:00:12 #startmeeting cinder 16:00:14 Meeting started Wed Jun 20 16:00:12 2018 UTC and is due to finish in 60 minutes. The chair is smcginnis. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:15 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:17 The meeting name has been set to 'cinder' 16:00:19 #topic roll call 16:00:25 #link https://etherpad.openstack.org/p/cinder-queens-meeting-agendas 16:01:01 o/ 16:01:16 hey 16:01:31 ping DuncanT diablo_rojo, diablo_rojo_phon, rajinir tbarron xyang xyang1 e0ne gouthamr thingee erlon tpsilva patrickeast tommylikehu eharney geguileo smcginnis lhx_ lhx__ aspiers jgriffith moshele hwalsh felipemonteiro lpetrut lseki _alastor_ 16:01:42 hi 16:01:46 Hi 16:01:56 hi! o/ 16:02:23 #topic Announcements 16:02:36 #info No meeting Wednesday 7/4 due to US Independence Day. 16:02:53 jungleboyj let me know he plans on cancelling the meeting for the US holiday week. 16:03:08 Many US-based folks are on vacation around that time, so probably good. 16:03:22 sounds like a good idea 16:03:28 o/ 16:03:33 We can always discuss anything important that comes up in the cinder channel, but hopefully it is a quiet week. 16:03:57 That's all I really had for announcements. Anyone know if I'm missing anything important to share? 16:04:17 #topic Rocky Priorities Review 16:04:24 what happened with the agenda? I had something there, not finding it now 16:04:27 #link https://etherpad.openstack.org/p/cinder-spec-review-tracking 16:04:38 erlon_: I still see one with your name on it. 16:05:05 smcginnis, ok, might be missing something here 16:05:18 erlon_: Feel free to add something if it's missing. 16:05:25 sure 16:05:42 In addition to the cycle priorities linked above, there's also the bugsmash event happening in Beijing. 16:05:45 #link https://etherpad.openstack.org/p/OpenSource-Hackathon-Rocky-Beijing-Bugs-List 16:05:55 Participants are putting links to their reviews there. 16:06:23 If we can help review those and make that event productive, I'm sure it would be appreciated and would help us knock out a few bugs. 16:06:45 #topic Check-in on HA development progress 16:06:51 * smcginnis shines spotlight on geguileo 16:06:54 :) 16:06:57 o/ 16:07:03 crickets, crickets... 16:07:04 Any updates on that topic geguileo? 16:07:08 Hah! 16:07:19 no, sorry 16:07:26 will try to finish the document this week 16:07:35 I forget, are there any patches up related to that? 16:07:49 there's one that's being reviewed 16:08:02 https://review.openstack.org/#/c/571242/ 16:08:02 patch 571242 - cinder - WIP: Write high availability devref 16:08:14 OK, perfect. 16:08:27 Feel free to review and give feedback even though it's still WIP ^^ 16:08:46 will try to give it another push to move it out of WIP 16:08:59 geguileo: Great, thanks for your work on that. 16:09:32 #topic Allowing drivers to save metadata from driver side 16:09:45 erlon_, walshh: This is yours I believe. 16:09:50 Yes 16:09:58 #link https://review.openstack.org/#/c/557448/ 16:09:59 patch 557448 - cinder - VMAX driver - Add VMAX specific metadata to volume 16:10:04 #link https://review.openstack.org/#/c/567473 16:10:05 patch 567473 - cinder - Cloning image fails results duplicate cache entry (MERGED) 16:10:07 yep, walshh wanna drive it? 16:10:16 There is no way to update model for extend, attach and detach 16:10:33 there is no model_update returned from those operations 16:10:40 I could maybe see extend, but what is needed for attach/detach? 16:10:55 vmax metadate 16:10:59 metadata* 16:11:09 storage groups, masking views etc 16:11:27 walshh, so you would need updates on all volume operations 16:11:33 isn't that something that can be queried when needed? 16:11:37 That doesn't really seem like stuff that should be kept in the cinder database. 16:11:50 smcginnis: +1 16:11:50 its not a case of having to 16:12:00 but it would be great if we could 16:12:10 walshh, I really want convinced that we should pput that in the database 16:12:35 walshh: is that to avoid a couple of extra calls on other driver operations? 16:12:42 it was an initiative led by QE 16:12:42 walshh, most of it or all are only debug information right? 16:12:49 If there's a good reason to be able to, we can probably add model updates for those. But it doesn't seem like it should be necessary. 16:12:54 they found it difficult tracing the path of the volume 16:12:55 walshh, QE? 16:13:04 our test group 16:13:30 walshh: but that should traced using the req ids, right? 16:13:37 walshh, may be printing more debug message on the driver side? 16:14:05 this feature would make it easier for them 16:14:12 one stop shop so to speak 16:14:31 it would be great if we could persist it but I guess we could just log it also 16:15:07 walshh, yeap, given the changes that would require from Cinder side, it would be the way Id recommend to go 16:15:40 If it's for the test group, maybe they can apply a local patch to enable things for their testing needs. 16:15:47 Otherwise, I don't think Cinder is the place for that. 16:16:12 smcginnis: +1 16:16:29 walshh: I think Sean's idea could be easily implemented 16:16:35 as a "debug" driver kind of thing 16:16:56 if we cant do a direct update then I concede, I dont expect an interface change 16:17:28 walshh: you "can" do a direct change, though we shouldn't allow it ;-) 16:17:47 :) 16:17:48 yes, it works quite well :-) 16:18:12 walshh: Is it OK if I move on then? 16:18:33 I will just log the information instead 16:18:56 Thanks, that would be the preference for this type of thing. 16:19:12 sure....thanks for your time on this 16:19:17 walshh: If they are interested in a debug/test patch, feel free to ask in the channel if any questions come up. 16:19:20 walshh: Thanks! 16:19:27 #topic Storyboard Update 16:19:44 So background for anyone that missed prior discussions... 16:20:00 There's more talk of moving off of launchpad to StoryBoard. 16:20:22 It being something we kind of control and have the ability to customize and link in with a common auth provider. 16:20:32 And hopefully a little easier to interact with. 16:20:45 A test migration has been done: 16:20:48 #link https://storyboard-dev.openstack.org/#!/project_group/55 16:21:09 Please poke around and see how things work if you have not interacted with SB yet. 16:21:19 Jay plans on trying a few things out and documenting usage. 16:21:37 There are things like being able to create Trello/Kanban style boards to track work. 16:21:59 That looks interesting, but I think at least at first it will take some fiddling to get things set up how we might want it. 16:22:46 smcginnis: +1 16:23:02 It definitely has a bit of a learning curve, but I've heard once you start to understand the model it's kind of nice. 16:23:50 There are also definitely some bugs with SB yet, so if you see anything wrong, please do file bugs against any issues for openstack/storyboard in https://storyboard.openstack.org/ 16:24:39 Not too much to say about it other than that at this point. 16:24:59 Hopefully jungleboyj will have some things documented with it and we can get more details in the near future. 16:26:08 If anyone does have questions about its usage, I think you can probably ask in #openstack-dev or #openstack-infra to get some help for now. 16:26:16 And I think there is some online documentation. 16:27:02 #topic Open Discussion 16:27:10 Anything else to talk about today? 16:27:56 crickets, crickets... 16:28:00 :) 16:28:06 OK, we can wrap up early then. 16:28:10 Thanks everyone! 16:28:16 smcginnis: thanks for leading the meeting today :-) 16:28:25 +1 16:28:36 I think you're stuck with me for one or two more, then Jay should be back. :D 16:28:49 #endmeeting