21:00:54 <mriedem> #startmeeting nova
21:00:56 <openstack> Meeting started Thu Mar  2 21:00:54 2017 UTC and is due to finish in 60 minutes.  The chair is mriedem. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:57 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:59 <openstack> The meeting name has been set to 'nova'
21:01:14 <takashin> o/
21:01:17 <edleafe> \o
21:01:19 <dansmith> o/
21:01:23 <efried> o/
21:01:31 <viks> o/
21:01:32 <raj_singh> o/
21:01:32 <melwitt> o/
21:01:35 <dtp> o/
21:02:51 <mriedem> #link agenda https://wiki.openstack.org/wiki/Meetings/Nova
21:02:59 <mriedem> let's get started
21:03:05 <mriedem> #topic release news
21:03:20 <mriedem> #link Pike release schedule: https://wiki.openstack.org/wiki/Nova/Pike_Release_Schedule
21:03:31 <mriedem> the release schedule is official
21:03:49 <mriedem> not much news there otherwise
21:03:54 <mriedem> #topic bugs
21:04:07 <mriedem> we don't have any critical bugs
21:04:28 <mriedem> we do have several newish untriaged bugs
21:05:04 <mriedem> gate status is relatively healthy
21:05:08 <mriedem> depending on who you ask
21:05:22 <mriedem> no 3rd party ci news
21:05:37 <mriedem> i don't know that vmware ever got their stuff running
21:05:45 <mriedem> but there is no one to ever ask anyway
21:06:05 <mriedem> #topic reminders
21:06:10 <mriedem> #link Pike Review Priorities etherpad template: https://etherpad.openstack.org/p/pike-nova-priorities-tracking
21:06:20 <mriedem> Subteams et al, please update, carrying over from https://etherpad.openstack.org/p/ocata-nova-priorities-tracking as appropriate.
21:06:27 <mriedem> efried got that primed this week
21:06:44 <mriedem> #info Move and/or repropose specs for Pike in the nova-specs repo.
21:06:55 <mriedem> #topic
21:06:59 <mriedem> #undo
21:07:01 <openstack> Removing item from minutes: #topic
21:07:04 <mriedem> #topic Stable branch status: https://etherpad.openstack.org/p/stable-tracker
21:07:29 <mriedem> #info Nova 14.0.4 was released for stable/newton.
21:07:36 <mriedem> #info Nova 13.1.3 was released for stable/mitaka.
21:07:41 <mriedem> that's about it
21:07:59 <mriedem> there are several changes on stable/newton with +2s
21:08:11 <mriedem> #topic subteam highlights
21:08:15 <mriedem> dansmith: cells v2?
21:08:20 <dansmith> wudup
21:08:24 <dansmith> cells, we've got 'em
21:08:36 <dansmith> we talked a bit about all the code that is up,
21:08:58 <dansmith> what happened at the ptg, and the current efforts afoot to finish out the api stuff and make devstack do that cellsy thang
21:09:18 * melwitt beams
21:09:35 * dansmith bows
21:09:42 <mriedem> is there anything that needs attention right now to move forward?
21:10:01 <dansmith> the vein popping out of the side of my head, but otherwise, no
21:10:18 <mriedem> you're still cranking on the thing to keep move operations witin the same cell?
21:10:23 <mriedem> that's WPI?
21:10:24 <mriedem> *WIP
21:10:32 <dansmith> no, that's up
21:10:46 <dansmith> no real reason for it to be WIP anymore I think
21:10:47 <mriedem> ok, i guess dump what you need reviews on in the etherpad above
21:10:56 <dansmith> the whole set is just kinda in flux as I try to get things lined up
21:11:05 <dansmith> yeah
21:11:08 <mriedem> ok
21:11:23 <mriedem> edleafe: scheduler/placement highlights?
21:11:28 <edleafe> We started, but didn't have anything to discuss early Monday after PTG, so it ended quickly and we got more coffee
21:11:54 <edleafe> Otherwise just working on traits and Jay's ironic stuff
21:11:56 * bauzas waves late
21:11:56 <mriedem> ok, anything immediate that needs review? i know jaypipes was pushing the os-traits stuff earlier today
21:12:32 <edleafe> there is this traits series starting with https://review.openstack.org/#/c/440654/
21:12:58 <edleafe> And alex_xu's spec at https://review.openstack.org/#/c/345138/
21:13:05 <edleafe> more eyes would be good
21:13:17 <dansmith> edleafe: his stuff isn't ironic
21:13:20 <dansmith> edleafe: it's for Ironic
21:13:34 <mriedem> ok, edleafe can you throw those into the review priorities etherpad too pleasE?
21:13:37 <mriedem> *please
21:13:40 <edleafe> mriedem: k
21:13:46 <mriedem> thanks
21:13:47 <edleafe> dansmith: alanis would disagree
21:13:56 <dansmith> oh, alanis
21:13:59 <mriedem> tdurakov: around for live migration?
21:15:06 <mriedem> probably unrelated to live migration, but a bug that came up today that's failing a lot on the multinode job, https://bugs.launchpad.net/nova/+bug/1669468
21:15:06 <openstack> Launchpad bug 1669468 in OpenStack Compute (nova) "tempest.api.compute.servers.test_novnc.NoVNCConsoleTestJSON.test_novnc fails intermittently in neutron multinode nv job" [Undecided,Confirmed]
21:15:18 <mriedem> it's a pretty high failure rate, on the multinode job
21:15:28 <mriedem> so if anyone knows anything about novnc and that test, help please
21:16:07 <mriedem> api, i don't think sdague or alex_xu are around
21:16:15 <mriedem> there are a flurry of specs for api
21:16:43 <mriedem> sriov/pci, i don't think anyone is around for that
21:16:58 <mriedem> also, lbeliveau is no longer working on openstack, we found out last week
21:17:00 <mriedem> just fyi
21:17:16 <mriedem> notifications updates
21:17:24 <mriedem> A short recap from the PTG is on the ML http://lists.openstack.org/pipermail/openstack-dev/2017-March/113098.html
21:17:30 <mriedem> Patches for the high prio searchlight notification bp is ready for the cores to look at https://review.openstack.org/#/q/topic:bp/additional-notification-fields-for-searchlight
21:17:39 <mriedem> Fix for the hight prio bug https://bugs.launchpad.net/nova/+bug/1665263 is ready for the cores to look at https://review.openstack.org/#/c/437222/
21:17:39 <openstack> Launchpad bug 1665263 in OpenStack Compute (nova) "instance.delete notification is missing for unscheduled instance" [High,In progress] - Assigned to Balazs Gibizer (balazs-gibizer)
21:17:54 <mriedem> that last one was a regression in ocata so we'll need to get it backported
21:18:00 <mriedem> i'll put it in the review etherpad
21:18:07 <mriedem> powervm - efried?
21:18:12 <efried> Yup
21:18:24 <efried> We have four change sets ready for review by the broader community.
21:18:38 <efried> Per PTG guidance, we've split up the spawn/delete into four smaller chunks.
21:18:54 <efried> Three of those are included in the aforementioned four-ready-for-review.
21:18:59 <mriedem> efried: i saw the base change which doesn't do any operations was passing the powervm ci - how does that work when it doesn't create a server? is it just testing that you can stack the thing but no tempest?
21:19:40 <efried> mriedem Actually, tempest is running.  The tests that we have whitelisted run the spawn and make sure that doesn't error, but they make no effort to, like, log into the thing or whatever.
21:20:25 <efried> So yeah, we could theoretically merge these one at a time, which we didn't think we were going to be able to do.
21:20:44 <efried> Though #2 is where we actually start creating the LPAR.
21:20:44 <mriedem> ok
21:20:53 <mriedem> delicious lpars
21:21:04 <mriedem> finally,
21:21:07 <efried> As for our CI, it's been running very smoothly.
21:21:08 <mriedem> does thorst have a kid yet?
21:21:16 <efried> almost, but not quite.
21:21:22 <mriedem> heh, ok
21:21:27 <efried> Unless you count nova-powervm.  That's his baby.
21:21:35 <mriedem> fair enough
21:21:47 <mriedem> i added cinder to the subteam highlights since we have a weekly meeting for it (for the last year now)
21:22:07 <mriedem> highlights from today were lyarwood has patches up to add uuid and attachment_id to the bdm table, we're going to get those in,
21:22:22 <mriedem> and scottda has a change to make nova support cinder v3 which we need to get in, but requires a cinderclient release first
21:22:26 <mriedem> which i think jungleboyj was going to propose
21:22:45 <mriedem> plus reviewing johnthetubaguy's new cinder api spec for nova
21:22:52 <mriedem> #topic stuck reviews
21:22:59 <mriedem> there was nothing on the agenda - anything to bring up?
21:23:19 <mriedem> #topic open discussion
21:23:28 <mriedem> a couple of things
21:23:33 <mriedem> (mriedem): Spec freeze? Non-priority blueprint freeze?
21:23:39 <mriedem> In Newton and Ocata we had spec freeze on the first milestone. Should we still do that for Pike?
21:23:54 <mriedem> get ready
21:23:56 <mriedem> In Newton we had a non-priority feature freeze between n-1 and n-2. In  Ocata we just had the feature freeze at o-3 because of the short  schedule. What should we do for Pike? Same as Newton? Non-priority  feature freeze at p-2? Or just feature freeze at p-3 for all, priority  or not? We have fewer core reviewers so I personally don't want to cut  off the majority of blueprints too early in the cycle.
21:24:19 <cuhler> We responded to Veritas Nova volume driver blueprint and spec review comments. Can it be approved for Pike?
21:24:23 <mriedem> so i was thinking spec freeze at the first milestone
21:24:25 <cuhler> blueprint: https://blueprints.launchpad.net/nova/+spec/veritas-hyperscale-libvirt-volume-driver volume driver spec: https://review.openstack.org/#/c/430488/ os brick connector spec: https://review.openstack.org/#/c/436670/
21:24:56 <mriedem> i was thinking no non-priority feature freeze, and just have the general FF at p-3
21:25:12 <bauzas> How many open specs do we have atm ?
21:25:15 <mriedem> we don't have to decide here, i'll send my thoughts to the ML
21:25:36 <bauzas> B/c
21:25:42 <mriedem> bauzas: i count 5 pages
21:25:43 <mriedem> in gerrit
21:26:03 <mriedem> 25 per page
21:26:05 <dansmith> mriedem: not having a prio freeze because we have fewer cores is a nice tip of the hat I think, acknowledging the issue I think
21:26:07 <mriedem> so ~125 open specs
21:26:13 <dansmith> it won't fix anything, but might ease the pressure a bit
21:26:24 <bauzas> b/c it would mean only one milestone for merging all of them then
21:26:39 <mriedem> bauzas: we aren't going to merge all of them
21:26:41 <mriedem> we never do
21:26:44 <bauzas> So yeah I agree with you
21:26:47 <mriedem> we'll have enough to do with just re-approvals
21:27:09 <mriedem> so i'll propose spec freeze on p-1 and FF on p-3 to the ML
21:27:12 <mriedem> and see who freaks out
21:27:25 <bauzas> I just want to explain that its better to have more milestones for merging implementations than specs IMHO
21:27:35 <mriedem> that's what this gives us
21:27:50 <mriedem> p-1 is 4/13 btw
21:27:54 <mriedem> so we have 6 weeks for specs
21:28:05 <mriedem> i think that's more than enough time
21:28:08 <mriedem> to overcommit
21:28:10 <mriedem> and underdeliver
21:28:15 <dtp> lol
21:28:18 <bauzas> Agreed
21:28:34 <mriedem> next thing i had was,
21:28:35 <mriedem> (mriedem): The docs team is looking for a liaison from the Nova  team to help coordinate things, like getting Nova specific changes into  the install guide (like placement and cells v2 from Ocata). Is anyone  interested in that role?
21:28:49 <mriedem> i posted to the ML and there is already some interest from at least 2 people
21:28:54 <mriedem> one was a private reply
21:29:36 <cuhler> Sorry to interrupt earlier. We responded to Veritas Nova volume driver blueprint and spec review comments. Can it be approved for Pike?
21:29:47 <mriedem> so, cuhler - i can look at your spec again, it's totally contingent on getting a connector into os-brick first,
21:29:54 <mriedem> which will depend on getting your volume driver, with CI, into cinder
21:30:17 <cuhler> Spec approval is contingent on getting the os-brick connector code in?
21:30:22 <mriedem> no
21:30:26 <mriedem> code approval in nova
21:30:39 <mriedem> i expect the spec is straight-forward
21:30:48 <mriedem> we generally don't even need a spec for volume drivers
21:30:53 <mriedem> since they are just a pass through to os-brick
21:31:01 <cuhler> i.e. we need os-brick spec approval. yes it's straightforward.
21:31:03 <mriedem> but i'll look at the nova spec again
21:31:08 <cuhler> Thx
21:31:12 <mriedem> anyone else for anything else?
21:31:24 <mriedem> ok thanks everyone
21:31:28 <mriedem> #endmeeting