14:00:13 <n0ano> #startmeeting nova-scheduler
14:00:14 <openstack> Meeting started Mon Oct 12 14:00:13 2015 UTC and is due to finish in 60 minutes.  The chair is n0ano. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:15 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:17 <openstack> The meeting name has been set to 'nova_scheduler'
14:00:22 <n0ano> anyone here to talk about the scheduler?
14:00:34 <PaulMurray> o/
14:01:46 <edleafe> o/
14:02:29 <n0ano> people are a little slow to join today
14:02:44 <edleafe> I wasn't sure if this was canceled or not
14:03:15 <n0ano> edleafe, I believe we said yes this week, not next (next week being too close to the summit), this is our last chance
14:03:58 <n0ano> bauzas, should be joining soon, he's on the nova channel
14:04:17 <bauzas> \o
14:04:36 <n0ano> welcome all
14:04:43 <n0ano> #topic Mitaka planning
14:05:32 <n0ano> I guess the only thing for today is to talk about the summit session, bauzas has proposed one on APIs and I've done two, one on flavor and a kind of umbrella one on the scheduler in general
14:05:51 <n0ano> have we all looked at the summit proposals?
14:05:53 <bauzas> n0ano: so the outcome is that johnthetubaguy and the other nova-drivers held a discussion
14:06:07 <n0ano> #link https://docs.google.com/spreadsheets/d/1MZVwxv8t6sM15Kct5i7yOS-8-NCvqsW742s5Y_LDyjg/edit#gid=816860825
14:06:18 <bauzas> the outcome is there
14:06:20 <bauzas> #link https://etherpad.openstack.org/p/mitaka-nova-summit-suggestions
14:06:28 <bauzas> (at the bottom, scroll down)
14:06:51 <bauzas> that's still draft atm
14:07:09 <bauzas> you'll see there is one slot pending some feedback from jaypipes
14:07:33 <n0ano> so is the one on Wed. 11:15 your suggestion?
14:07:43 <bauzas> not exactly
14:08:04 <bauzas> the API discussion is pretty large
14:08:19 <bauzas> I mean, it's not really worth a 40-min slot
14:08:34 <n0ano> yeah, what I was thinking, not near enough time
14:08:34 <bauzas> but we should somehow discuss that somewhere
14:09:00 <bauzas> that's pretty close to the Resource modeling session
14:09:20 <bauzas> which should discuss about how we model our resources
14:09:23 <bauzas> anyway
14:09:30 <bauzas> I don't want to overthink about that
14:09:36 <edleafe> is the API slot just scheduler API, or all Nova API?
14:09:42 <bauzas> Nova API
14:09:46 <edleafe> ah, ok
14:10:07 <bauzas> Resouce Providers and Resource Modeling are for the moment the sched slots we have
14:10:27 <bauzas> plus one extra on 12:05 that needs to be confirmed
14:11:07 <n0ano> I really hope we can confirm the 12:05 one, I really want to have a scheduler session with everyone there
14:11:54 <bauzas> n0ano: the resource modeling is tighly integrated with scheduling, y'know :)
14:12:11 <bauzas> n0ano: but agreed, there is at least one thing I'd like to discuss
14:12:17 <n0ano> bauzas, agreed, but that's not the only thing
14:12:24 <bauzas> https://review.openstack.org/#/c/191914/
14:12:32 <bauzas> is the thing I'd like to address
14:12:53 <bauzas> n0ano: we can leave some sched bits to Friday, if you feel
14:13:07 <bauzas> n0ano: speaking of the priorities for Mitaka
14:14:21 <n0ano> well, there's also the firestorm that edleafe has created with his new sched desing, I think we should make time for that also
14:15:00 <edleafe> after the midcycle, I don't think that they will want to spend time on that
14:15:38 <bauzas> n0ano: well, you probably wonder why I haven't replied
14:15:44 <bauzas> and my answer is https://review.openstack.org/#/c/191914/
14:15:49 <n0ano> I don't know, given the email activity this seems to me to be something perfect for a face to face meeting
14:16:26 <n0ano> bauzas, might want to point that out on the mailing list
14:16:40 <bauzas> well, it's not secrete
14:16:44 <bauzas> dammit
14:16:49 * bauzas needs coffee
14:17:21 <n0ano> bauzas, at mid-afternoon you need coffee :-)
14:17:38 <bauzas> yup, 4.15pm here
14:17:45 <edleafe> n0ano: I have an IV of coffee set up at my desk :)
14:17:51 <edleafe> mainlining it
14:17:55 <bauzas> \o/
14:18:00 <bauzas> 4th in a row
14:18:06 <bauzas> anyway
14:18:16 <bauzas> I guess we pretty have good slots
14:18:31 <bauzas> like I said, 2 sessions dedicated on resource tracking
14:18:50 <edleafe> Yeah, those will be important
14:18:52 <n0ano> yeah, it's looking good to me, I think we're in OK shape
14:19:15 <bauzas> the 3rd one could be indeed focused on scalability
14:19:45 <bauzas> given we have ideas but not yet a consensus
14:20:03 <bauzas> moving on ?
14:20:11 <n0ano> bauzas, wfm
14:20:17 <n0ano> #topic opens
14:20:21 <n0ano> anything else?
14:21:27 <n0ano> hearing crickets
14:21:29 <bauzas> I need to redraw a few bits for the reqspec
14:21:43 <bauzas> that's something on-going in my TODO list
14:21:57 <bauzas> we also had the Lib specs proposed again for Mitaka
14:22:03 <bauzas> at least both of mine
14:22:19 <n0ano> bauzas, does proposed mean approved?
14:22:52 <bauzas> http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/check-destination-on-migrations.html http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/request-spec-object-mitaka.html and http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/persist-request-spec.html
14:23:09 <bauzas> n0ano: sorry yes
14:23:22 <bauzas> FWIW there is also http://specs.openstack.org/openstack/nova-specs/specs/mitaka/approved/host-state-level-locking.html
14:23:24 <n0ano> so sounds like we're good ther
14:23:47 <bauzas> tbc, what we agreed to continue to work on is approved
14:23:57 <bauzas> no need to get further approvals
14:24:22 <bauzas> the resource-objects BP is not yet reproposed tho
14:24:53 <edleafe> bauzas: is that jaypipes' BP?
14:24:55 <bauzas> I'm not worried about getting it fast-approved again, but I feel jaypipes is probably waiting the sched/RT sessions
14:25:00 <bauzas> edleafe: yup
14:25:02 <n0ano> sounds like we need another etherpad dashboard
14:25:11 <bauzas> n0ano: there is already one
14:25:34 <bauzas> n0ano: but given the prios are not yet identified, it's a bit premature to use it
14:26:23 <n0ano> I'm guessing the scheduler items will still be a priority so I don't see a problem with that
14:27:02 <bauzas> n0ano: it's all about respecting the process :)
14:27:10 <bauzas> anyway
14:28:00 <n0ano> indeed - anything else?
14:28:09 <edleafe> ll
14:28:13 <edleafe> doh!
14:28:22 <edleafe> 6 windows at once...
14:28:30 * n0ano waiting for coherence from edleafe :-)
14:28:48 <edleafe> you'll be waiting for a looooonnnngggg time!
14:29:10 <n0ano> and I have 12 IRC windows open :-)
14:29:34 <bauzas> we can be eventually consistent if all of us move to other tasks while someone is issuing #endmeeting
14:29:43 <edleafe> n0ano: I have at least that many, but irssi hides all but the active
14:30:03 <n0ano> anyway, I think we're done
14:30:06 <bauzas> that said, before leaving
14:30:08 <edleafe> a half-dozen ssh sessions doesn't help
14:30:13 <bauzas> no sched meeting next week for me
14:30:20 <edleafe> or any of us, right?
14:30:22 <bauzas> and I guess for all of us, right?
14:30:33 <bauzas> n0ano: you okay with notifying the ML ?
14:30:34 <n0ano> bauzas, good point, no meeting next week, save it for Tokyo
14:30:36 <edleafe> bauzas: are you going early with your family?
14:30:41 <bauzas> edleafe: indeed
14:30:48 * edleafe is jealous
14:30:52 <n0ano> bauzas, sure, I send out a notice like always the night before
14:31:00 <bauzas> n0ano: coolness
14:31:06 <edleafe> we were all supposed to go, but daughter has important school events
14:31:11 <bauzas> arf
14:31:17 <n0ano> tnx everyone, see you soon
14:31:20 <n0ano> #endmeeting