15:05:19 #startmeeting nova_scheduler 15:05:20 Meeting started Tue May 5 15:05:19 2015 UTC and is due to finish in 60 minutes. The chair is bauzas. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:05:21 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:05:23 The meeting name has been set to 'nova_scheduler' 15:05:31 o/ 15:05:35 \o 15:05:43 aloha, sorry for having missed out the start 15:06:02 no worries - on a call for another meeting now anyway 15:06:32 let's wait for a few other peers - unless I just bored them 15:08:11 tic tac 15:08:23 * alaski lurking 15:08:31 tic tock 15:08:46 * edleafe thinks of mints when I see 'tic tac' 15:09:03 :) 15:09:21 lots of peers there 15:09:33 ok, starting but it should be quick 15:09:42 since I expect no real debate :) 15:09:47 #topic Liberty specs 15:10:11 #link https://wiki.openstack.org/wiki/Gantt/liberty 15:10:40 so this page ^ is a little out-to-date, I'll discuss with n0ano to see how we can update it 15:11:00 most of the BPs we target for Lib are in 15:11:01 #link https://etherpad.openstack.org/p/liberty-nova-summit-ideas 15:11:19 https://blueprints.launchpad.net/nova/+spec/allocation-ratio-to-resource-tracker => https://review.openstack.org/173252 (sbauza)https://blueprints.launchpad.net/nova/+spec/request-spec-object => https://review.openstack.org/173316 (sbauza)https://blueprints.launchpad.net/nova/+spec/resource-objectshttps://blueprints.launchpad.net/nova/+spec/make-resource-tracker-use-objectshttps://blueprints.launchpad.net/nova/+spec/shared-s 15:11:28 oh man, sorry for the wrong paste 15:13:14 #info if people have BPs they want to include in the scheduler effort, please add them in the etherpad (L98) so we can review that before the Summit 15:14:14 nothing really in progress from my POV, I'm tighly following https://review.openstack.org/#/c/76234/ as the first step for RequestSpec object 15:14:41 will discuss with danpb to see when he plans to upload a new update 15:15:32 okay, anything else to mention about specs ? 15:16:05 (silence) 15:16:28 nope 15:16:39 ok, moving on 15:16:47 #topic Vancouver design summit 15:17:12 nothing really important from my POV again, just has to work on a spec before the summit 15:17:30 https://blueprints.launchpad.net/nova/+spec/shared-state-scheduler is not having a spec yet 15:17:59 I remember n0ano was interested in discussing on cross-project sessions 15:18:36 bauzas: I feel a lot will come after the discussions 15:18:46 Still too much up in the air now 15:18:54 agreed 15:19:04 just looking at https://docs.google.com/spreadsheets/d/1vCTZBJKCMZ2xBhglnuK3ciKo3E8UMFo5S5lmIAYMCSE/edit to see if anything related to us 15:19:36 Line 28 15:19:55 oh 15:19:57 rightr 15:20:01 let's see the outcome then 15:20:19 I honestly doubt it would be accepted, but that's good anyway for discussing 15:20:33 I see that as a refinement of the whose request spec object/ claim object pattern 15:20:46 s/whose/whole 15:21:36 edleafe: not sure I follow you 15:21:46 edleafe: oh 15:22:06 edleafe: well, yeah, it's related to if a resource needs to be claimed 15:22:06 bauzas: how to design the objects to represent any type of resource 15:22:34 edleafe: but IMHO all of that is related to https://blueprints.launchpad.net/nova/+spec/resource-objects at some point 15:22:48 without having that BP implemented, that sounds a little premature IMHO 15:23:18 anyway 15:23:21 bauzas: I think it will make it more robust so that it requires less refactoring later 15:23:35 edleafe: that's just chicken-and-egg 15:23:42 bauzas: not really 15:23:53 edleafe: let's work on implementing bp/resource-objects, we could refine later 15:24:07 anyway 15:24:14 understanding other types of resources means less refinement needed later 15:24:19 let's move on unless someone has a thought ? 15:24:48 edleafe: the spec is approved, so there are little options to change the direction 15:25:13 but I don't want to debate much about that spec, it was enough done during Kilo :D 15:25:17 specs can be amended and re-approved 15:25:20 :) 15:25:53 edleafe: well, the situation is clear : we need to implement that spec - and possibly amend it 15:26:02 yep 15:26:03 edleafe: but at least, we need to implement it 15:26:21 anyway for anyway 15:26:24 let's move on 15:26:29 #topic open discussion 15:27:39 let them speak or forever hold, should I say ? 15:27:57 * edleafe hears crickets 15:28:38 * bauzas hears that too 15:28:41 #endmeeting