15:01:27 #startmeeting ceilometer 15:01:28 Meeting started Thu Oct 15 15:01:27 2015 UTC and is due to finish in 60 minutes. The chair is gordc. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:29 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:01:32 The meeting name has been set to 'ceilometer' 15:01:34 o/ 15:01:39 o/ 15:01:40 o/ 15:01:46 o/ 15:01:47 o/ 15:02:17 o/ 15:02:32 #topic design summit 15:02:46 ok. so i put up the schedule for design summit. 15:02:51 https://mitakadesignsummit.sched.org/overview/type/ceilometer 15:03:03 o/ 15:03:04 please let me know if you have any conflicts and i'll try to address them. 15:03:41 you can also find the list of etherpads for each session here: https://wiki.openstack.org/wiki/Design_Summit/Mitaka/Etherpads 15:03:50 #link https://wiki.openstack.org/wiki/Design_Summit/Mitaka/Etherpads 15:04:20 i think one item to point out (something llu-laptop reminded me of) 15:04:51 as i previously mentioned, we have oppportunity to test ceilometer on 1000 node environment 15:05:19 i was thinking it'd be good to have some ideas how/what to deploy on this environment to test. 15:05:34 we can use the contributor's meetup to discuss that 15:05:48 or do we want a dedicated session? 15:05:56 gordc: Ceilo with the old back ends or Gnocchi? 15:06:32 ildikov: nadya(mirantis) are currently testing mongodb i believe 15:06:35 but they aren't sharding. 15:06:37 gordc: 1000 node without installed OpenStack ? 15:06:48 gordc: we can use the contributors meetup; i assume this is for the OSIC cluster (intel-Rackspace) 15:06:52 gordc: ok 15:06:56 KrishR: correct. 15:07:05 nadya: not sure what you just asked 15:07:06 KrishR: yes 15:07:38 ildikov: but yeah, i think the idea is to come up with what we want to test. 15:08:18 gordc: I asked Gnocchi as we plan to move to that API soon I guess, etc. so it would be nice to see what it can handle 15:08:19 just bringing it up now so you have a few days/weeks to think about it. 15:08:30 gordc: +1 15:08:54 gordc: you have hardware or OpenStack env? you asked 'how deply on this env', what we need to deploy? 15:09:03 ildikov: yep. or we can do multiple storage as well. just difference scenarios. 15:09:26 gordc: if that's an option that would be even better 15:09:40 nadya: intel-rax have a openstack env they are building 15:09:51 nadya: we can set up the OSIC cluster with whatever versions of ceilometer/gnocchi you want 15:10:03 we just need to decide how to implement the ceilometer/aodh/gnocchi parts 15:10:11 gordc, KrishR: ok, I see 15:11:22 that's all i had on design summit. 15:11:26 gordc: yeap, I was thinking about that too, I mean having all three components in a setup, etc. 15:11:44 ildikov: yeah, that's a good idea. 15:12:21 just a reminder, the etherpads are live, so if you are leading/participating in one of the sessions. feel free to edit the etherpads 15:12:54 i currently just copy/pasted stuff in. i'm going to use "i'm sick" excuse to justify laziness 15:13:15 * cdent gives gordc a cookie 15:13:35 cookie cures nothing. but i'll take it. 15:13:50 anyone else have items to add? 15:14:54 #topic RDO needs/wants this fix (on liberty) to mongodb replica set handling: https://review.openstack.org/#/c/234254 15:15:05 yeah that was me 15:15:30 apparently there was no test coverage on that chunk and when rdo test day tried to use ?replicaset for ha in a ceilo deploy it didn't work 15:15:43 they can workaround by going back to the deprecated method of using a conf setting 15:15:53 but that's not got any shelf life 15:15:55 cdent: did you want a rc3? 15:16:06 so yeah the question is: what are the chances of an rc3 15:16:13 because i think ttx just released stable/liberty this morning 15:16:33 i'll need to check with ttx/dhellmann post meeting. 15:16:52 but i don't think we can spin up another release candidate 15:16:55 note that it is not a backport, which sort of weirds it up 15:17:31 I guess from rdo's standpoint it doesn't really matter if it gets in a release, just that it will be release eventually, and they can manage the patch until then 15:17:43 cdent: is this related to sileht's patch? i vaguely remember it being changed 15:17:46 so having it merge asap, even if not released, would be nice 15:17:59 cdent: kk. 15:17:59 his patch for o.messaging? 15:18:03 no 15:18:28 no. i think he touched replicaset stuff but i could just be saying sileht because he touches everything 15:18:32 the messaging thing is another topic we probably ought to work out: does that cause release crazy? 15:18:37 it's true he does 15:18:45 I think he is an aggregate 15:19:08 :) 15:19:20 cdent: lol :) 15:19:21 obviously too late for a rc3 15:19:29 since the release is out 15:19:37 ttx: thanks for confirmation. 15:20:14 cdent: yeah so i guess we'll get it in next stable/liberty 15:20:16 yesterday was the last day for respins. Today is final tagging day. FYI it takes 7 hours to go through everything 15:20:48 gordc: you can issue a point release early next week if necessary 15:20:56 gordc: not a problem, do we happen to know when we'll be able to merge that so it is present on the branch? 15:20:57 ttx: i got 200emails from you this morning. that's why i knew it was done 15:21:21 ttx: sounds good. i'll sync up with you folks when we want next point release. 15:22:07 cdent: i think once ttx tags everything we can merge it in. 15:22:14 roger 15:22:23 i don't really have a problem with it not being in master (since it's not needed) 15:22:50 I'm sure it's not the first cacse it happens 15:22:51 good 15:23:48 cdent: you're required to provide an envelope when you add stuff to stable that aren't backports. 15:23:56 you decide what is in the envelope. 15:24:16 for you, I say tissues 15:24:26 gordc: I guess cookies :) 15:24:43 both are good options given current state. 15:25:20 #topic meeting next week? 15:25:25 no 15:25:29 hahaha 15:25:38 cdent: that was fast :) 15:25:56 i won't be available either as i'll be travelling 15:26:01 gordc: if we sorted out everything for the design summit, which seems to be the case, I'm fine if we cancel next week 15:26:19 have a great rip, folks! 15:26:24 *trip 15:26:30 do we want to tentatively schedule one but cancel if nothing comes up next few days? 15:26:39 nadya: thanks! 15:26:41 nadya: thanks! 15:27:08 ildikov: do you want to lead it next week if we do have one? 15:27:32 i think if nothing gets raised in mailing list by next tuesday, we can cancel. 15:27:39 gordc: I'm available, so I can do that if needed 15:27:46 ildikov: great. 15:27:57 should need a mail thread to inform this cancelation ? 15:28:08 i'm pretty sure it'll be cancelled since i know quite a few people will be away 15:28:37 gordc: yeap, I think it's ok if we wait until Tuesday as you said and then send a mail 15:28:38 liusheng: i can send an email if/when we officially cancel. 15:29:17 about the meeting, forgive me if I missed the email, any conclusion about the alternative meeting time? 15:29:49 I think it didn't reach a conclusion 15:29:58 llu-laptop: that was just an internal query i had. 15:30:26 llu-laptop: is there a specific time that works better for APAC? 15:31:40 i'm getting used to this time, but I can't speak on behalf of others 15:31:48 llu-laptop: ack 15:31:58 llu-laptop: i'll add a thread regarding this 15:32:20 i think ultimately we still want to make the ML play a more important role 15:32:30 gordc: agreed 15:32:36 +1 15:33:16 #action gordc send email regarding modes of communication. 15:33:40 #topic open discussion 15:34:09 do we need to talk about dealing with o.messaging or is that under control? 15:34:15 (especially the liberty parts) 15:34:37 cdent: i got the requirements patch to block 2.6.1 approved 15:34:50 i take it it needs to be backported to liberty as well 15:34:56 presumably 15:35:21 cdent: kk. i'll get that looked at as well. 15:35:21 it will need may "recheck" latter :) 15:35:31 s/may/many/ 15:35:45 liusheng: yeah, definitely 15:36:12 is there a backport for liberty already? 15:36:35 I don't know. 15:36:50 liusheng, you know? 15:37:03 cdent: no 15:37:46 ok. i'll be offline for a meeting next hour or so. if there's no backport when i get back, i'll do it and raise it 15:38:07 we'll need to wait for all the tags to be finished/merged though. 15:39:41 all's good? final countdown. 15:40:43 thanks folks. 15:40:51 see you in Tokyo. 15:40:52 thanks 15:40:56 * cdent waves 15:40:57 see you in Tokyo! 15:41:03 #endmeeting