14:00:18 #startmeeting nova_scheduler 14:00:19 Meeting started Mon Jul 2 14:00:18 2018 UTC and is due to finish in 60 minutes. The chair is cdent. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:23 The meeting name has been set to 'nova_scheduler' 14:00:37 #chair efried jaypipes edleafe bauzas 14:00:38 Current chairs: bauzas cdent edleafe efried jaypipes 14:00:39 o/ 14:00:41 o/ 14:00:51 good <<<>>> folks 14:01:31 #topic last meeting 14:01:32 #link last minutes: http://eavesdrop.openstack.org/meetings/nova_scheduler/2018/nova_scheduler.2018-06-25-14.02.html 14:01:43 any remainders from last week that people would like to revisit? 14:01:45 o/ 14:01:55 \o 14:02:16 nope 14:02:19 #topic specs and review 14:02:19 #link latest pupdate: http://lists.openstack.org/pipermail/openstack-dev/2018-June/131915.html 14:02:38 any reviews that need extra discussion or attention? 14:03:30 i'll take that as a "no" 14:03:36 #topic bugs 14:03:36 #link placement bugs: https://bugs.launchpad.net/nova/+bugs?field.tag=placement&orderby=-id 14:04:00 I've filed a new one this morning https://bugs.launchpad.net/nova/+bug/1779635 14:04:00 Launchpad bug 1779635 in OpenStack Compute (nova) "placement allows RP parent loop in PUT resource_providers/{uuid}" [Medium,Confirmed] 14:04:02 gibi reported a new bug https://bugs.launchpad.net/nova/+bug/1779635 14:04:12 #link parent loop https://bugs.launchpad.net/nova/+bug/1779635 14:04:29 gtfo, I thought we had tests for that. 14:04:39 efried: we have for RP create but not RP set 14:05:06 I've pushed a patch that reproduce the problem https://review.openstack.org/#/c/579508/ 14:05:09 Mm. And we don't allow reparenting if parent is already set. But if it wasn't set, you're saying you can set it to a loop... 14:05:17 efried: exactly 14:05:21 we had a recent bug 14:05:29 but we fixed it 14:06:02 honestlty, I didn't set it Critical because the gate was fine 14:06:05 * alex_xu waves late 14:06:54 but we need to make sure that when we use a new API, the minimum version we have is still accepting it 14:06:55 gibi: you able to work on fixing it? 14:07:21 cdent: yes, I'm working on it 14:07:26 yay! 14:07:30 any other bugs of note? 14:08:06 #topic opens 14:08:35 I added "consumer gen lifecycle" 14:08:44 because we had lengthy talk about it last week 14:09:06 #link review about con gen 404 https://review.openstack.org/#/c/579163/ 14:09:17 #link irc discussion http://eavesdrop.openstack.org/irclogs/%23openstack-placement/%23openstack-placement.2018-06-29.log.html#t2018-06-29T13:49:51 14:09:29 the irc discussions gets into a lot of meat 14:10:12 It would be useful is people who weren't in that discussion (gibi, bauzas, tetsuro) could read it and add their thoughts 14:10:23 ok will look 14:10:28 thanks 14:10:53 I've opened the link 14:11:23 I don't reckon efried, edleafe, jaypipes and I need to add any more until other folk have had a chance to think about it 14:11:30 cdent: what's the concern ? 14:11:40 by quickly look at the change, I'm fine with it 14:11:51 so, maybe a new microversion needed then ? 14:11:53 bauzas: better to read the irc log rather than me explain it, as I'm biased 14:12:05 it's more broad than that 14:12:21 there's also some useful discussion on the associated bug 14:12:56 #link consumers never get deleted: https://bugs.launchpad.net/nova/+bug/1778763 14:12:56 Launchpad bug 1778763 in OpenStack Compute (nova) "Consumers never get deleted" [Undecided,New] 14:13:05 apologies for being late. 14:13:22 I mean, from a REST point, if you ask for allocations for a consumer but then not having a consumer, it's at least a 40x 14:13:37 maybe 400 14:13:47 bauzas: you need to read the entire irc discussion or you are just repeating things we already debated there 14:13:48 That's not the core issue. Consumers don't get deleted. There's at least four different ways we could react. 14:14:18 cdent: well, sure, but you're asking me to look at like 150 lines 14:14:18 :) 14:14:22 I personally think that reading those logs and then revisiting the topic is better than rehashing things here and now 14:14:26 so I'll do it, but later :) 14:14:29 yes, I am 14:14:31 and later is good 14:14:36 there's no need to do it right now 14:14:40 ok 14:15:04 I thought you were asking for voices 14:15:08 but fine, will read it 14:15:33 I'm asking for people think about it, giving it real attention and consideration, and then speak about it, not rush 14:15:45 and preferably s/speak/write/ 14:15:52 fait :) 14:15:55 fair :) 14:16:10 any other open issues or comments on that one? 14:16:55 Would everyone like the meeting to end? 14:17:29 anyway, I need to leave you in 3 mins :) 14:17:40 silence means approval? 14:18:15 wait 14:18:22 * cdent waits 14:18:26 no, I got nothing. 14:18:40 hehe 14:18:43 yeah, I'm ok. 14:18:51 thanks everyone for coming, please read those irc logs, and the review and bug and join in the discussion there 14:18:53 #endmeeting