21:00:09 <mriedem> #startmeeting nova 21:00:10 <openstack> Meeting started Thu Sep 29 21:00:09 2016 UTC and is due to finish in 60 minutes. The chair is mriedem. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:11 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:00:14 <openstack> The meeting name has been set to 'nova' 21:00:19 <tonyb> \o 21:00:21 <takashin> o/ 21:00:21 <diana_> o/ 21:00:21 <dansmith> o/ 21:00:25 <alaski> o/ 21:00:25 <edleafe> \o 21:00:25 <raj_singh> o/ 21:00:27 <mikal> Hi 21:00:59 <mriedem> #link agenda https://wiki.openstack.org/wiki/Meetings/Nova#Agenda_for_next_meeting 21:01:10 <mriedem> #topic Release News 21:01:14 <lbeliveau> o/ 21:01:18 <mriedem> #link Newton release schedule: https://wiki.openstack.org/wiki/Nova/Newton_Release_Schedule 21:01:38 <mriedem> #info newton rc2 tag patch: https://review.openstack.org/379806 21:01:42 * Vek waves 21:01:53 <mriedem> so rc2 should be out just in time 21:01:58 <mriedem> #link Draft Ocata release schedule: https://wiki.openstack.org/wiki/Nova/Ocata_Release_Schedule 21:02:02 * edleafe waves back to Vek 21:02:07 <mriedem> #info Specs are open for Ocata so get things re-proposed if they didn't land in Newton. 21:02:20 <mriedem> questions about the release? 21:02:41 <mriedem> #topic bugs 21:02:54 <mriedem> #link check queue gate status http://status.openstack.org/elastic-recheck/index.html 21:03:11 <mriedem> there are 2 bugs that have been bugging us in the gate, one merged earlier today 21:03:15 <mriedem> the other is a ceph one 21:03:26 <mriedem> http://status.openstack.org//elastic-recheck/index.html#1627220 21:03:53 <mriedem> although now it looks like that's dropping off for some reason 21:04:11 <mriedem> anyway 21:04:25 <mriedem> #link 3rd party CI status http://ci-watch.tintri.com/project?project=nova&time=7+days 21:04:44 <mriedem> intel nfv ci had some issues earlier in the week with a pypi mirror, but i think that's been sorted out 21:04:51 <mriedem> or at least not -1ing things 21:05:02 <mriedem> vmware has continued to have issues 21:05:23 <mriedem> and i posted a request in the mailing list to get the recheck instructions in the vmware nsx ci comment back on the patch when it fails 21:05:29 <mriedem> i think that is the 3rd or 4th time i've asked for that 21:05:32 <bauzas> nova nova nova hop ? 21:05:32 <bauzas> https://www.youtube.com/watch?v=E9wWBjnaEck 21:05:32 * bauzas feels alone in the dark :) 21:05:53 <tonyb> mriedem: Are we getting close to asking infra to suspend the account? 21:05:54 <mriedem> umm, ok? 21:06:02 <mriedem> tonyb: not yet 21:06:09 <mriedem> tonyb: it's not voting 21:06:15 <tonyb> mriedem: is more patient than I am ;P 21:06:23 <mriedem> tonyb: it's just annoying that it's not working when i want to +2 a vmware patch 21:06:59 <tonyb> mriedem: Yeah, I have a few in that boat aswell 21:07:05 <mriedem> we don't have any critical bugs 21:07:14 <mriedem> i just did a pass through triage, there are some NFV looking bugs in there 21:07:17 <mriedem> which are over my head to triage 21:07:26 <mriedem> so looking for help from lbeliveau, sfinucan and moshele 21:07:48 <mriedem> questions about bugs? 21:07:56 <bauzas> mmm, I'm 7 min ahead... weirdo 21:08:03 <mriedem> #topic Summit planning 21:08:11 <mriedem> #link Etherpad for Ocata summit topics: https://etherpad.openstack.org/p/ocata-nova-summit-ideas 21:08:19 <mriedem> i've started to try and filter through that list 21:08:23 <mriedem> and spec out a rough schedule 21:08:28 <mriedem> we have more slots than we need really i think 21:08:54 <mriedem> will try to finalize the proposed schedule before next week's meeting for feedback 21:09:03 <mriedem> #info Cross project summit session ideas are due by 10/1: http://lists.openstack.org/pipermail/openstack-dev/2016-September/103657.html 21:09:08 <bauzas> anyway, \o 21:09:11 <mriedem> ^ There are proposals for (1) nova/cinder API workflow redo, (2) discoverable API capabilities, (3) cross-service communications re: tokens, (4) scaling review teams (among others). 21:09:32 <mriedem> so one of the reason nova will probably have more slots than we need is because we might have some things covered in the wednesday xp sessions 21:09:45 <mriedem> with overlap / follow up in the nova track 21:09:54 <mriedem> same with neutron, i think they are going to have us over for a chat on wednesday 21:10:02 <mriedem> ^ is my impersonation of someone british 21:10:13 <tonyb> mriedem: :) 21:10:37 <Vek> heh. 21:10:47 <mikal> Chats don't normally involve physical harm 21:10:59 <mriedem> questions on summit planning? 21:11:05 <cdent> mikal: when was the last time your were in the uk? 21:11:27 <cdent> it's all about physical harm these days 21:11:45 <tonyb> mriedem: chats about football most certainly do 21:12:13 <bauzas2> \o (waves super late because of lagging) 21:12:30 <mriedem> mikal: we're actually quite chummy with neutron these days 21:12:30 <mriedem> it's those cinder bastards i can't stand! :P 21:12:30 <mriedem> #topic Reminders 21:12:30 <mriedem> #help https://wiki.openstack.org/wiki/Nova/BugTriage#Weekly_bug_skimming_duty Volunteers for 1 week of bug skimming duty? 21:12:30 <mriedem> looks like ^ isn't really getting anyone signed up 21:12:52 <mriedem> bauzas: bauzas2: yes we know you're here 21:13:00 <mriedem> put the drink down 21:13:12 <mriedem> #topic Stable branch status: https://etherpad.openstack.org/p/stable-tracker 21:13:19 <mriedem> #info We need to do a review push on stable/liberty and stable/mitaka and do a release of those. There is a security bug fix in stable/liberty and several unreleased commits in mitaka. 21:13:26 * jroll steals bauzas2's drink 21:13:47 <mriedem> i've already got a patch up for a liberty 12.0.5 release 21:13:59 <mriedem> that's probably going to be the last one for liberty before EOL 21:14:17 <bauzas2> mriedem: heh, I wish I would, but that's probably my IRC bouncer which drunk 21:14:25 <mriedem> we need to make a review pass through stable/mitaka https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/mitaka,n,z 21:14:35 <mriedem> and then a release next week 21:14:37 <dansmith> mriedem: I shall do some of that when we're done here 21:14:38 <mriedem> tonyb scolded me for having too many unreleased commits 21:14:46 <mriedem> dansmith: thanks 21:14:48 <tonyb> lol 21:14:56 <tonyb> "scolded" 21:15:02 <mriedem> he was really terrible 21:15:10 <mriedem> i didn't get a single piece of cake 21:15:22 <mriedem> #topic subteam highlights 21:15:26 <mriedem> alaski: cells! 21:15:44 <alaski> there was a bug found around reschedules which was fixed yesterday 21:15:53 <alaski> and fixed in stable/newton today 21:16:08 <alaski> the meeting this week was mostly about getting organized for specs and summit 21:16:24 <alaski> I'm going to put together an etherpad with what's left to get to multiple cells 21:16:33 <alaski> I think that's it 21:16:41 <mriedem> ok 21:16:49 <mriedem> edleafe: jaypipes: scheduler meeting? 21:16:53 <mriedem> cdent: ^ 21:16:54 <edleafe> Just discussed some of the work for Ocata that was just starting up 21:17:02 <edleafe> Also, how we might divide it up better 21:17:19 <edleafe> We also threw around ideas for topics to discuss at the summit 21:17:26 <edleafe> I added them to the etherpad 21:17:30 <edleafe> that's it 21:17:41 <mriedem> ok 21:17:52 <mriedem> we have 2 slots penciled in for scheduler/placement 21:18:00 <mriedem> and 2 slots for cellsv2 things 21:18:04 <mriedem> tbd 21:18:19 <mriedem> anyone around from the live migration meeting? 21:18:22 <mriedem> tdurakov: ^? 21:18:38 <mriedem> i know we talked about testing out pkoniszewski's patch to run a live-migration + grenade multinode job 21:18:47 <mriedem> i've got a todo to write a d-g wip patch for that 21:19:15 <mriedem> and there was talk about some specs, like live migrate rescued instances, and that lead to a ML thread from mdbooth about canonical libvirt domain xml 21:19:36 <raj_singh> mriedem: I have one for grenade LM job https://review.openstack.org/#/c/378015/ 21:20:04 <mriedem> #link patch to test live migration + grenade https://review.openstack.org/#/c/378015/ 21:20:07 <mriedem> thanks 21:20:31 <mriedem> ok, API meeting 21:20:33 <mriedem> alex_xu: ? 21:20:46 <mriedem> i was at that one, we talked about some specs 21:21:00 <mriedem> johnthetubaguy is working on redoing security groups in the server create API 21:21:12 <mriedem> sdague has a spec up to start talking about discoverable capabilities from the API 21:21:33 <mriedem> we talked a bit about the standardized diagnostics API spec and how to represent BDMs 21:21:38 <mriedem> that was mostly it 21:21:55 <mriedem> oh and alex_xu has a spec up for no longer proxying image metadata, i'm +2 on that if someone wants to hit it after the meeting 21:22:06 <mriedem> https://review.openstack.org/#/c/377528/ 21:22:26 <mriedem> lbeliveau: the sriov/pci meeting was cancelled this week right? 21:22:43 <mriedem> gibi_: was there a notifications meeting this week? 21:23:08 <mriedem> ok moving on 21:23:15 <mriedem> #topic Stuck Reviews 21:23:17 <mriedem> there was nothing on the agenda 21:23:24 <lbeliveau> yeap, got canceled 21:23:27 <mriedem> ok 21:23:31 <mriedem> #topic Open discussion 21:23:34 <mriedem> there are 2 agenda iteams 21:23:36 <mriedem> *items 21:23:45 <mriedem> (dave-johnston): Blueprint for new filter. https://blueprints.launchpad.net/nova/+spec/multi-tenant-exclusion-filter . Can this be implemented without a spec? 21:24:20 <bauzas2> there was an old proposal about the above 21:24:31 <bauzas2> from belmoreira IIRC 21:24:39 <edleafe> yeah, sounds familiar 21:25:52 <mriedem> jaypipes: want to take a look at ^ too ? 21:25:54 <bauzas2> so, process-wise, I dunno, I know we agreed with having specless BPs for that but sometimes it's nice to discuss about the filter behaviour 21:26:12 <mriedem> yeah this doesn't sound obviously trivial from just reading it 21:26:17 <mriedem> i think a spec would be good 21:26:29 <bauzas2> also, a new filter would require a relnote I guess 21:26:36 <mriedem> well, you can release note w/o a spec 21:26:48 <bauzas2> sure, but I think we need to clear the behaviour 21:27:03 <bauzas2> ie. why we should rather do this or that 21:27:19 <bauzas2> so, I'm +1 on having a spec for that one 21:27:34 <bauzas2> because it's not trivial and has a long history 21:27:40 <mriedem> edleafe: feel the same? 21:28:06 <edleafe> yeah, I think that's best 21:28:22 <mriedem> ok, i'll reply in the bp to write a spec and probably try to attend the scheduler meeting to talk about it 21:28:33 <bauzas2> ack, wfm 21:28:41 <mriedem> last item 21:28:45 <mriedem> Do we need a spec for the optional placement DB blueprint? https://blueprints.launchpad.net/nova/+spec/optional-placement-database 21:28:59 <dansmith> so, given how confused it was in the begining, 21:29:03 <dansmith> I think we probably do 21:29:06 <bauzas2> +1 21:29:08 <mriedem> ok, that's why i had this on here 21:29:14 <alaski> +1 21:29:20 <mriedem> because we talked about this at the meetup but then it went all crazy i guess while i was on vacation 21:29:21 <bauzas2> that's having a very important operator's impact 21:29:22 <mriedem> and i didn't know the details 21:29:32 <jaypipes> mriedem: I can try 21:29:38 <mriedem> jaypipes: let cdent write it 21:29:44 <mriedem> or roman? 21:30:18 <mriedem> ok, well someone will do it, but i'd suggest those working on the code 21:30:21 <cdent> I'm happy to write it, if someone tells me why we need it? 21:30:23 <mriedem> anyone else have open discussion items? 21:30:25 <edleafe> cdent doesn't have much to do 21:30:35 <cdent> oh cheers edleafe :) 21:31:08 <bauzas> ooooh I'm not lagging anymore here \o/ 21:31:09 <mriedem> because long term we want to move placement things into the placement db i thought 21:31:16 <cdent> no, why we need a spec 21:31:28 <mriedem> because of the confusion with adding that in newton 21:31:46 <mriedem> i wasn't around that week though, so need someone else to fill in 21:31:46 <cdent> the confusion was over the implementation, not the need 21:31:47 <dansmith> and because it's a giant backend thing 21:31:59 <cdent> but yeah, sure, I can make a spec 21:32:02 <bauzas> what dansmith said 21:32:10 * edleafe snickers "giant backend thing" 21:32:19 <cdent> placement's got back 21:32:26 <dansmith> heh 21:32:33 <mriedem> ok anything else? 21:32:43 <dansmith> edleafe: the relation of placement to "ass" is not lost on me 21:32:59 <dansmith> mriedem: please no 21:33:03 <edleafe> dansmith: obviously 21:33:04 <mriedem> btw, newton is basically released now https://review.openstack.org/#/c/379806/ 21:33:09 <mriedem> so \o/ 21:33:15 <edleafe> \o/ 21:33:26 <mriedem> ok, that's it, thanks all 21:33:29 <mriedem> #endmeeting