15:02:44 #startmeeting Mistral 15:02:45 Meeting started Mon Dec 11 15:02:44 2017 UTC and is due to finish in 60 minutes. The chair is rakhmerov. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:46 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:02:48 The meeting name has been set to 'mistral' 15:02:51 hi 15:02:54 o/ 15:03:02 Hey! 15:03:25 hello 15:03:59 ooh, 4 people ) 15:04:01 not so bad 15:04:07 hi guys 15:04:36 I wonder for how long we haven't had any meetings 15:04:56 d0ugal: btw, did you have meetings in the last 2 weeks? 15:05:08 rakhmerov: yup 15:05:11 ok 15:05:29 then maybe can you quickly update? 15:05:37 mostly myself I guess.. 15:05:45 if there was anything important 15:06:00 I know about some CI problems and bugs that you pointed to last week 15:06:04 rakhmerov: nothing important that I remeber, but I'd have to go and dig up the logs. 15:06:22 ok ) 15:07:07 I have a couple topics to discuss with you but let's give quick updates maybe 15:07:09 as usually 15:07:17 #topic Current status 15:07:58 working on a yaql blueprint to add executions to the exported functions 15:08:02 my status: mostly doing reviews after my vacation, still have a few patches to review in my queue 15:08:13 apetrich: how is that going? 15:08:16 WIP here https://review.openstack.org/#/c/525926 15:08:18 close? 15:08:35 rakhmerov, missing the date range filters. Should have that ready for reviewing today 15:08:49 I want to get that in before we freeze freeze 15:08:56 o/ sorry late 15:08:57 yep 15:09:03 bobh: hi, that's ok ) 15:09:08 we just started 15:09:17 apetrich: ok, cool 15:09:20 very cool 15:09:31 bobh: if you have some updates, please share 15:09:46 I have been working on tempest failures and the refactor of the mistral context is now starting to pass CI. 15:10:09 I'm also planning with my management what to do next since the most critical phase is finally over for us 15:10:20 working on the task on-clause change, need to refine the on-clause spec definition 15:10:27 d0ugal: yep, I think we're close to merge what's needed 15:10:39 bobh: ok 15:10:44 One more review needed on https://review.openstack.org/#/c/506185/ 15:11:01 status: rerun execution still waiting on reviews 506652, 506653 and I have been working on porting ansible actions into mistral-extras 15:11:24 d0ugal: btw, when I was reviewing your patches I found myself on the thought that now we assume that our SecurityContext always reflects Keystone stuff 15:11:28 zuul v3 native job definitions are almost complete. 15:11:36 whereas, in fact, we can also work with Keycloack 15:11:52 toure: it is probably a good idea to remove [WIP] from reviews that are ready, people don't tend to look at them otherwise 15:12:03 d0ugal ack just did 15:12:04 :) 15:12:16 d0ugal: I understand it's a big change and not sure if we can adapt it for both but may be you've already thought about it 15:12:17 toure: oh, cool. I'll take a look then. 15:12:18 I totally spaced on the description 15:12:35 #action rakhmerov: review https://review.openstack.org/#/c/506185/ 15:12:47 rakhmerov: I did think it was a bit strange, maybe we should have called it "context.keystone"? 15:13:09 rakhmerov: but yeah, changing it now would be super painful. 15:13:15 yep 15:13:17 I know 15:13:48 hm... don't know maybe we could have some very generic base class for security context for now with just one implementation for Keystone 15:13:56 True 15:14:05 but I know it's super painful 15:14:31 rakhmerov: I really want to get this work merged, then we can look into refactoring/improving after? 15:14:41 absolutely, yes 15:14:43 I agree 15:15:02 let's merge it and during Q-3 we can potentially make some adjustements 15:15:07 (theoretically) 15:15:08 Thanks 15:15:11 :) 15:15:49 toure: hey, I thought 506652, 506653 are still WIPs, no? 15:15:56 or they are ready to be reviewed? 15:16:01 if so, I'll do it asap 15:16:34 rakhmerov they are ready for review I forgot to push the change for the commit message to remove WUP 15:16:37 **WIP 15:16:47 ooh, that explains a lot ;) 15:16:49 please do 15:16:51 :) 15:16:53 done 15:17:06 very few people will review until you remove WIP ) 15:17:16 ok, good 15:17:32 thrash: question to you. You said "zuul v3 native job definitions are almost complete" 15:17:42 rakhmerov: yes. 15:17:43 so what exactly are you working on there? 15:17:52 sorry, I'm just not in the loop honestly 15:17:57 been away for a while 15:18:10 or maybe you can point to something so that I could learn 15:18:11 https://review.openstack.org/509428 15:18:25 rakhmerov: the "legacy" jobs, which is what we have now, use a lot of shell, and don't take advantage of the zuul v3 stuff... Like being able to inherit from other jobs. 15:18:44 ooh, legacy jobs 15:18:55 ok 15:19:09 I'm asking because, for some reason, I thought it was all done already 15:19:09 moving the job definitions into the repo was the first step. This gets rid of the legacy stuff, and uses far less boilerplate. 15:19:11 a while ago 15:19:24 ok, I see 15:19:26 that was just the move into the repo. This is to make them better. :D 15:19:41 so this is supposed to the the final step towards the full migration, right? 15:19:46 Yes 15:19:49 ok 15:19:52 good 15:19:55 thanks 15:20:15 rakhmerov: for reference... https://docs.openstack.org/infra/manual/zuulv3.html#legacy-job-migration-details 15:20:24 ok 15:20:37 thank you 15:20:44 then let's move on 15:20:54 I'd like to discuss the following topic 15:21:02 #topic Do we need weekly meetings now? 15:21:08 :) 15:21:33 rakhmerov: http://eavesdrop.openstack.org/meetings/mistral/2017/mistral.2017-12-04-15.00.log.html#l-76 15:21:43 From last week :) see the 4th last line. 15:21:48 the reason I'd like to discuss it is because 1) I know some of the teams don't have these meetings anymore 2) in the last few months we skipped a lot of them 15:22:21 aah 15:22:37 d0ugal: what are these "office hours" exactly? 15:22:42 not sure I fully understand 15:23:07 rakhmerov: I'm not sure exactly. But I think we just advertise somewhere when we are normally available in #openstack-mistral and encourage people to come and speak to us then 15:23:13 so instead of telling them to come to the meeting 15:23:23 For example this is how the TC does it: https://governance.openstack.org/tc/#office-hours 15:23:34 ok 15:23:36 let me see.. 15:23:36 keystone does as well I believe 15:23:42 yup 15:23:50 yeah, I've seen that for keystone a lot 15:24:26 hah, interesting.. 15:24:37 that looks good to me 15:24:58 I'm not sure which days/times we pick. 15:25:10 but I guess we can work that out with a poll or something 15:25:10 I just thought that maybe we need some known time windows when we could overlap ourselves, team members I think 15:25:23 to sync up and discuss relevant topics 15:25:31 Right, I think we can use the office hours for that. 15:26:13 why does it have to be "#openstack-tc"? 15:26:19 can it be just our channel? 15:26:27 rakhmerov: It doesn't - that is just where the TC have their office :) 15:26:31 it should be our own channel 15:26:38 ok 15:26:40 yep 15:27:03 and I'm just concerned with 1 hour duration 15:27:14 I don't understand why it's just 1 hour 15:27:23 but I guess it's up to us, right? 15:27:24 We can make it longer 15:27:28 yes, ok 15:27:40 I don't think it is a gurantee, it just means "we will normall be around during there times" 15:27:43 or, that is how I read it 15:27:50 I see 15:28:00 I guess the TC is a bit different, because I assume they are usually busy doing other things 15:28:14 well, sounds good to me. In addition, I'd like to have a clear picture of when we could talk ourselves 15:28:18 somehow 15:28:33 it's pretty easy for folks whose time zones are close 15:28:42 like me, d0ugal and apetrich 15:28:54 but harder with North America, for example 15:28:54 :) 15:29:04 aye 15:29:33 so office hours is mostly for non team members to talk to team members? 15:29:39 I guess we can propose a number of different blocks and add/remove/change them over time. 15:29:52 or they can also be used as a replacement for this kind of meetings? 15:29:55 I think office hours is for anyone that wants to talk, within the team or anyone else. 15:30:02 but I'm not sure. 15:30:05 ok 15:30:19 I think they are just the same as meetings, but without any structure and if there is nothing to talk about nothing happens 15:30:32 I see 15:30:42 so, yes, I don't think we would need the meeting 15:31:07 so are there objections from anyone else? 15:31:19 I also think we don't need these regular meetings 15:31:36 pretty often they are almost empty (very few people) 15:31:47 I believe we need some flexibility here 15:32:00 bobh, toure, apetrich, thrash? 15:32:02 +1 15:32:13 +1 15:32:20 +1 15:32:25 ok 15:32:53 then I can learn more about this "office hours" concept and try to apply to us 15:33:22 I guess we can communicate time slots over ML and/or some voting system 15:33:32 Sounds good. 15:33:46 sounds good 15:33:53 ok 15:34:08 #action rakhmerov: learn about "office hours" and propose time slots 15:34:56 the other topic I wanted to discuss was already explained (ongoing CI work) 15:35:03 so I don't have much else 15:35:47 I'm planning to do some house keeping in launchpad this week and update statuses of the most important tickets 15:35:48 +1 15:36:02 and pay attention to things that require it 15:36:31 bobh: like event notifications (mostly finished my review but was distracted today) 15:36:41 :) 15:36:49 anything else guys? 15:36:50 :-) 15:37:22 one more thing, I know it's pretty early but please try to plan your participation in the PTG 15:37:41 trying :-) 15:37:51 I already said "yes" to the question "Is Mistral team going to be at the PTG?" 15:37:52 :) 15:38:05 great 15:38:39 ok 15:38:47 so ending the meeting? 15:39:00 Nothing else from me 15:39:20 and I'd like to ask you very much: please go over all of your patches once in a while and make sure to address comments 15:39:26 and update their statuses 15:39:32 make them visible please 15:39:36 it helps a lot 15:40:10 ok, thanks to everyone! 15:40:20 have a good week ) 15:40:24 bye 15:40:34 Thanks! 15:40:38 have a good one 15:40:43 bye 15:40:47 #endmeeting