14:00:08 <edleafe> #startmeeting nova_scheduler
14:00:09 <openstack> Meeting started Mon Sep 19 14:00:08 2016 UTC and is due to finish in 60 minutes.  The chair is edleafe. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:10 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:12 <openstack> The meeting name has been set to 'nova_scheduler'
14:00:16 <edleafe> Anyone around?
14:00:20 <_gryf> o/
14:00:29 <Yingxin> \o
14:01:14 <alex_xu> o/
14:01:17 <edleafe> Looked pretty empty in the #openstack-nova channel this morning, so let's give people a few more minutes
14:01:38 <jaypipes> o/
14:03:24 <edleafe> well, let's get started
14:03:26 <edleafe> #topic Specs and Reviews
14:03:37 <edleafe> Master is now open for Ocata
14:03:44 <edleafe> as we cut RC1 last week
14:03:48 * bauzas waves late
14:03:57 <edleafe> Is there anything we need to be looking at now?
14:04:10 * edleafe realizes we are still working on Newton, too
14:04:53 <edleafe> ok, moving on
14:04:56 <edleafe> #topic Bugs
14:05:10 <edleafe> Any release-blocking bugs to discuss?
14:05:20 <bauzas> nothing I know
14:05:27 <edleafe> bauzas: ok, thanks
14:05:32 <Yingxin> https://launchpad.net/bugs/1622538
14:05:33 <openstack> Launchpad bug 1622538 in OpenStack Compute (nova) "Wrong "can_host" field of compute node resource providers" [Medium,In progress] - Assigned to Yingxin (cyx1231st)
14:05:56 <Yingxin> not sure why it is -2ed..
14:06:01 <edleafe> Yingxin: is that potentially release-blocking?
14:06:07 <Yingxin> no
14:06:08 <bauzas> Yingxin: that's something we agreed to not merge by Newton
14:06:52 <Yingxin> bauzas: dan said probably not even in ocata
14:07:11 <edleafe> Yingxin: also, dansmith's comment suggests that we may be chaning the whole 'can_host' approach
14:07:30 <edleafe> changing
14:07:41 <edleafe> let
14:07:44 <edleafe> oops
14:07:48 <edleafe> let's move on
14:07:53 <edleafe> #topic Opens
14:08:00 <edleafe> We have two open topics
14:08:08 <edleafe> First is summit discussion topics
14:08:21 <edleafe> Any thoughts on this yet?
14:09:39 <edleafe> OK, let's push that to a later meeting
14:09:50 <edleafe> Next is workload sharing
14:09:58 <edleafe> I wrote a thing with some ideas:
14:10:01 <edleafe> #link https://blog.leafe.com/pair-development/
14:10:24 <edleafe> cdent isn't around, but he gave me very positive feedback on the notion
14:10:38 <bauzas> I tend to disagree :)
14:10:39 <edleafe> I'd like to have people think about it
14:10:48 <edleafe> We can discuss in more depth at the summit
14:10:57 <edleafe> bauzas: I'm shocked! :)
14:11:16 <bauzas> yeah, I'm not surprised you'd be shocked :p
14:11:38 <bauzas> that's something I'd like rather to discuss during the retrospective
14:11:42 <bauzas> but tl;dr
14:12:10 <jaypipes> edleafe: I like the idea of being able to semi-continuously make progress on something, for sure.
14:12:22 <edleafe> bauzas: I added it to the retrospective page
14:12:22 <jaypipes> edleafe: I hate pair programming, but this isn't that :)
14:12:25 <edleafe> #link https://etherpad.openstack.org/p/nova-newton-retrospective
14:12:32 <bauzas> I think the problem is not related to a methodology, rather that the placement effort was difficult to have all the changes being merged for only one cycle
14:12:49 <bauzas> for example, we don't have this problem with cellsv2 IMHO
14:12:49 <edleafe> jaypipes: yeah, I rarely found anyone who matched my focus
14:13:11 <edleafe> bauzas: but this isn't new this cycle
14:13:19 <jaypipes> bauzas: what particular problem are you talking about?
14:13:30 <jaypipes> bauzas: that cellsv2 doesn't have?
14:13:34 <alex_xu> when we rush for freeze, people help me when I sleep a lot, I got a lot of help on update the patch in the night, and catch up the deadline :)
14:13:55 <bauzas> jaypipes: I just think that all the needed changes for Newton were just too huge for only one cycle
14:14:18 <edleafe> alex_xu: there is a definite advantage in different timezones, but there should ideally be enough overlap for 1:1 communication
14:14:28 <bauzas> but again, it seems to be a retrospective discussion
14:14:37 <alex_xu> yea
14:14:43 <jaypipes> bauzas: I'm not a fan of long release cycles (or releases at all, frankly)
14:14:54 <bauzas> jaypipes: agreed
14:15:08 <bauzas> jaypipes: I just think that 6 months for what we wanted to merge was too shorty
14:15:09 <edleafe> jaypipes: that ship sailed long ago. I never wanted releases like we have
14:15:22 <edleafe> but that's way off-topic
14:15:26 <bauzas> given of course the other priorities and reviews
14:15:52 <edleafe> so let's all contribute our experiences to the Newton retrospective page
14:15:58 <bauzas> ++
14:16:02 <jaypipes> bauzas: I think we could have gotten a lot more done in Newton had I had the time to focus on placement API and not been pulled away for endless management bullshit ;)
14:16:23 <bauzas> agreee to disagree :)
14:16:29 <edleafe> jaypipes: that's definitely part of the problem
14:16:39 <jaypipes> edleafe: yeah :(
14:16:41 <edleafe> critical path management
14:16:53 <jaypipes> also, summer vacations always play a part in things this time of year
14:17:08 <edleafe> jaypipes: winter holidays, too
14:17:23 <edleafe> having more coverage will help alleviate that, IMO
14:18:00 <edleafe> Anyway, it's something for us to think about
14:18:05 <edleafe> and unless anyone has any other open topics...
14:18:42 <edleafe> #endmeeting