15:00:07 #startmeeting heat 15:00:07 Meeting started Wed Dec 7 15:00:07 2016 UTC and is due to finish in 60 minutes. The chair is ramishra. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:09 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:12 The meeting name has been set to 'heat' 15:00:15 #topic roll call 15:00:21 howdy 15:00:24 Yop 15:01:14 o/ 15:01:31 hey guys, thanks for joining. 15:02:46 Not sure if we have any more joining. 15:03:35 ok, let's move on:) 15:03:56 #topic adding items to agenda 15:04:07 #link https://wiki.openstack.org/wiki/Meetings/HeatAgenda#Agenda_.282016-12-37_1500_UTC.29 15:05:00 #topic ocata-2 status 15:05:20 #link https://launchpad.net/heat/+milestone/ocata-2 15:05:58 milestone tagging is next week, too soon;) 15:06:38 I critical and 2 high priority bugs. 15:07:31 wow, this cycle is really going quick 15:07:37 ramishra, https://bugs.launchpad.net/heat/+bug/1578854 is still relevant? 15:07:37 Launchpad bug 1578854 in heat "resource-list makes SQL calls for every resource which is a nested stack" [High,In progress] - Assigned to Steve Baker (steve-stevebaker) 15:07:55 There is no opened patches against it, I think we can close it 15:08:28 therve: yeah, I did not check that. 15:08:55 cwolferh: can we land all patches for bug 1524013? 15:08:55 bug 1524013 in heat "Arbitrarily-large resource properties should be allowed in events" [Medium,In progress] https://launchpad.net/bugs/1524013 - Assigned to Crag Wolfe (cwolfe) 15:09:52 ramishra, the patches should be fine. the only thing i'm uncertain of is this is not writing data to the old place (which would be easy to change) 15:10:20 for the sake of a rolling upgrade 15:10:37 And there are some bugs, which are assigned to some but no patches. 15:10:46 I think we should reassign and fix them. 15:11:28 cwolferh: I'm wondering if we can get them reviewed in time for o-2. 15:11:46 I don't see anything critical though 15:12:41 ok, we can move them to o-3;) 15:12:50 ramishra, also, i reordered them yesterday, so 2 more refactors before real changes 15:13:38 cwolferh, The first one in the queue is rough to review 15:13:46 time for me to review that series I guess 15:15:27 I think we'll try and review as many as we can:) 15:16:16 Yeah some patches are really stalled in there :/ 15:16:37 I wanted to discuss about the rolling upgrade thing, that's the next topic. 15:16:42 should we move on? 15:16:46 Sure 15:16:56 #topic rolling upgrade 15:17:06 #link https://review.openstack.org/#/c/407989/ 15:17:23 I pushed this draft spec today. more to discuss about it, 15:18:15 I'm not sure if anyone has seen the glance specs and reviews on this. 15:18:28 #link https://review.openstack.org/#/c/331740/ 15:18:53 they have also recorded a good demo. #link https://www.youtube.com/watch?v=Z4iwJRlPqOw 15:19:00 I like that plan, but I've never deployed heat in production, so... 15:19:13 That's using triggers 15:19:19 It'd be nice to have Rackspace guys feedback on it 15:20:08 therve: yeah, I will add all cores to the review. 15:20:31 also any feed back on the glance approach for db migrations would also be good too. 15:21:04 db upgrades 15:21:58 I'd like to avoid triggers, but if it comes down to that, so be it 15:22:36 We will wait and see how it works for glance, before going that way;) 15:23:56 triggers make me very nervous 15:24:21 zaneb, +1000 15:24:30 the only justification for that is if writing to the old and new place results in too much duplication 15:24:46 like maybe the case with resource properties data 15:25:14 though i would also prefer to avoid triggers 15:25:27 cwolferh: my only concern is over a period of time the db layer code may be unreadable if we start writing to both old and new location. 15:25:47 I don't like triggers though. 15:25:53 can we just migrate the properties data now and try not to screw up anything that big again? ;) 15:26:16 sounds good to me :-) 15:27:05 ramishra: I was kinda hoping ovo would isolate the compat code so it didn't become unmanageable. it hasn't really turned out that way though :/ 15:27:44 ramishra, i don't think it is too terrible (we only read from one place but potentially write to 2), but i see your point 15:28:44 yeah, we can probably manage the db changes carefully, to avoid triggers. 15:29:16 Anyway, that's not in the plan for this cycle, triggers or not;) 15:30:04 If we can get the grenade tests setup, then we can probably ask for the tag:) 15:31:16 We don't have more topic to discuss today. 15:31:48 I had this topic for v2 api, we discussed it a little last meeting. 15:32:08 #topic Heat API V2 15:32:51 zaneb, cwolferh, you guys may put your opinion on this. 15:33:19 We thought that we don't have enough bandwidth to work on this in the last meeting. 15:33:29 and it's not super important. 15:33:30 * zaneb probably didn't read the minutes from last week 15:33:59 I think Qiming had some concerns with the v1 api 15:34:04 there's definitely a long list of ugly stuff we'd like to fix 15:34:15 though he does not seem to around. 15:34:33 otoh, moving to a new API version is painful, as proven by... just about every other OpenStack project at this point 15:34:56 i also need to read back on that, v2 would be a nice clean slate... sounds like it ought to be uncontroversial :-) 15:35:10 zaneb: Are the current issues listed somewhere? 15:35:18 yes 15:35:23 definitely 15:35:30 somewhere. 15:35:36 :) 15:36:42 https://wiki.openstack.org/wiki/Heat/Blueprints/V2API 15:37:19 #link https://wiki.openstack.org/wiki/Heat/Blueprints/V2API 15:38:08 Ah, I think some guys are interested to start working on it. 15:38:36 Some of them are new contributors, may be someone can help create a spec and guide them? 15:38:52 I wished they worked on performance instead, but whatever :) 15:40:30 yeah, performance stuff is not for everyone;) limited experts:) 15:41:13 Anyway, I don't think there can be much done in this cycle. 15:41:59 ok, let's move on. 15:42:10 #topic open discussion 15:42:19 Anything else we want to discuss? 15:42:53 already discussed the resource properties patches, if people review that that would be awesome :-) 15:43:12 Yeah review review reviews 15:43:28 :-) 15:44:34 I also wanted some of the gabbi api test patches to be reviewed too. 15:45:11 ok, I think we are done. should we move back to heat? 15:45:17 reviews on https://review.openstack.org/#/c/395124/ would be great 15:45:21 zaneb, Should I create a spec for CustomResource? 15:45:26 Or whatever we called it 15:45:33 therve: yes, by all means 15:46:18 zaneb, That somehow brings back the discussion about Mistral::WorkflowExecution 15:47:22 I've thought a bit about Mistral::WorkflowExecution 15:47:38 I think I would be ok with it if it worked a bit like SoftwareComponent 15:48:06 i.e. you can specify an action for create, update, delete, and there's a way to trigger replacement 15:48:43 like CustomResource that doesn't mean people will use it properly. but at least they have a chance 15:48:50 * therve nods 15:50:04 therve, zaneb: I had registered this bp https://blueprints.launchpad.net/heat/+spec/implement-custom-resource 15:50:09 long time back 15:50:35 Somewhat inline with the AWS implementation. 15:50:39 Indeed :) 15:50:55 Replace marconi by zaqar and that's the idea 15:51:11 lol, that *is* a long time back :D 15:52:18 yeah, that time it was outrightly rejected;) 15:53:48 was it? by whom? 15:54:27 zaneb: I don't remember exactly. But, in some discussions I found not buyers for it. 15:54:36 tbh it was probably too early to be really useful back then. now all the other pieces are in place 15:54:36 no buyers 15:54:48 yeah, I agree 15:55:41 http://lists.openstack.org/pipermail/openstack-dev/2014-April/032609.html :) 15:56:12 therve: wow:) 15:57:21 happy to see that it turns out I have been consistent :) 15:58:48 oh, no I haven't 15:59:07 I said user hooks ought to be enough for anybody 15:59:27 they aren't 16:00:46 ramishra: time's up 16:00:53 I'm happy that we're back to it:) Though I knew very little heat that time to participate in a discussion. 16:00:56 yep 16:01:03 thanks all for joining. 16:01:09 #endmeeting heat