21:00:33 <melwitt> #startmeeting nova
21:00:34 <openstack> Meeting started Thu Jan  3 21:00:33 2019 UTC and is due to finish in 60 minutes.  The chair is melwitt. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:00:35 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
21:00:37 <openstack> The meeting name has been set to 'nova'
21:00:49 <takashin> o/
21:00:50 <melwitt> welcome everyone, happy new year
21:00:59 <mriedem> o/
21:01:32 <melwitt> #topic https://etherpad.openstack.org/p/nova-stein-blueprint-spec-freeze
21:01:34 <efried> o/
21:01:35 <melwitt> #undo
21:01:36 <openstack> Removing item from minutes: #topic https://etherpad.openstack.org/p/nova-stein-blueprint-spec-freeze
21:01:51 * melwitt sighs
21:01:52 <melwitt> #topic Release News
21:01:58 <melwitt> #link Stein release schedule: https://wiki.openstack.org/wiki/Nova/Stein_Release_Schedule
21:02:15 <melwitt> s-2 spec freeze is next week Jan 10
21:02:42 <melwitt> we're tracking specs to focus review on in the last stretch on an etherpad
21:02:51 <melwitt> #link spec freeze tracking pad: https://etherpad.openstack.org/p/nova-stein-blueprint-spec-freeze
21:03:31 <melwitt> we've collected a list of specs that are close to approval or are small/simple
21:04:06 <melwitt> please take a look at the list and add comments if you know of other specs that are close that are not yet on the list
21:04:31 <melwitt> #link Stein runway etherpad: https://etherpad.openstack.org/p/nova-runways-stein
21:04:48 <melwitt> note: I haven't updated runways yet since returning from break
21:04:56 <melwitt> #link runway #1: https://blueprints.launchpad.net/nova/+spec/bandwidth-resource-provider (gibi) (current patch series below supports server create, delete and re-schedule cases and reject some intentionally unsupported cases) [END 2018-12-24] https://review.openstack.org/#/c/625941
21:04:57 <patchbot> patch 625941 - nova - Create RequestGroup from neutron port - 3 patch sets
21:05:03 <melwitt> #link runway #2: https://blueprints.launchpad.net/nova/+spec/local-disk-serial-numbers (mdbooth) [END 2019-01-10] (extra week because of holiday week) https://review.openstack.org/#/c/530786
21:05:04 <patchbot> patch 530786 - nova - Expose volume host type and path independent of li... - 10 patch sets
21:05:20 <mriedem> the first runway is overdue,
21:05:23 <mriedem> discussion in the ML
21:05:40 <mriedem> 3rd slot is open for the configurable vol attachment bp of yours
21:06:01 <mriedem> and likely could put Kevin_Zheng's per-cell metadata API in there too, it's a single hcange
21:06:02 <mriedem> *change
21:06:09 <mriedem> the down cell API change is languishing
21:06:18 <mriedem> so while that's in the queue, it's not really ready for a runway slot
21:06:55 <melwitt> ok, I see the note there now. ok, I'll update runways after the meeting. thanks
21:07:47 <melwitt> I think that's all I have for release news. does anyone have anything to add or any questions?
21:08:19 <melwitt> ok
21:08:29 <melwitt> #topic Bugs (stuck/critical)
21:08:43 <melwitt> no Critical bugs
21:08:50 <melwitt> #link 61 new untriaged bugs (up 3 since the last meeting): https://bugs.launchpad.net/nova/+bugs?search=Search&field.status=New
21:09:25 <melwitt> thanks mriedem for doing triage
21:09:52 <melwitt> #link 13 untagged untriaged bugs (up 3 since the last meeting): https://bugs.launchpad.net/nova/+bugs?field.tag=-*&field.status%3Alist=NEW
21:09:55 <melwitt> as usual, please help out with triage if you can
21:09:56 <melwitt> #link bug triage how-to: https://wiki.openstack.org/wiki/Nova/BugTriage#Tags
21:10:10 <melwitt> Gate status
21:10:16 <melwitt> #link check queue gate status http://status.openstack.org/elastic-recheck/index.html
21:10:47 <melwitt> I'm not sure how the gate's doing of late but I've noticed rechecks in my gerrit notifications
21:11:19 <mriedem> the nova-lvm job is broken but i'm working on a fix,
21:11:32 <mriedem> otherwise there isn't a good sample the last 2 weeks to know if there are any major problems
21:12:08 * melwitt nods
21:12:17 <melwitt> 3rd party CI
21:12:37 <melwitt> #link 3rd party CI status http://ciwatch.mmedvede.net/project?project=nova&time=7+days
21:12:57 <sean-k-mooney1> on that note im working on a replacement for the nfv ci
21:13:05 <melwitt> looks like xenserver CI is doing bad
21:13:13 <sean-k-mooney1> i hope to have someting running in the next week or two
21:13:18 <melwitt> sean-k-mooney1: ah, cool. that's good to hear
21:13:22 <melwitt> thanks
21:13:53 <melwitt> PCI CI looks ~50% fail
21:14:11 <melwitt> vmware CI also failing it appears
21:15:01 <melwitt> anything else on bugs or CI before we move on?
21:15:03 <cdent> some problem with firewall interactions
21:15:10 <cdent> "is being worked on"
21:15:52 <melwitt> ok, that's good to know. thanks
21:16:14 <melwitt> #topic Reminders
21:16:24 <melwitt> #link Stein Subteam Patches n Bugs: https://etherpad.openstack.org/p/stein-nova-subteam-tracking
21:17:27 <melwitt> I put a trivial bugs section at the top of this etherpad awhile back ^ fyi if anyone has +1ed any trivial bugs and can put them in the list
21:17:51 <melwitt> I had one but it's gone, so I need to check on that
21:17:52 <melwitt> #link Spec freeze/non-priority blueprint freeze is s-2 January 10, 2019
21:18:13 <melwitt> that's all I have for reminders, anyone else have anything for reminders?
21:18:52 <melwitt> #topic Stable branch status
21:18:59 <melwitt> #link stable/rocky: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/rocky,n,z
21:19:09 <melwitt> #link stable/queens: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/queens,n,z
21:19:20 <melwitt> #link stable/pike: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/pike,n,z
21:19:39 <melwitt> we did a release of rocky, queens, and pike just before the holiday break
21:20:03 <melwitt> but there's still more to review, so please feel free to review some more :) I'm going to go through some
21:20:24 <melwitt> anything else for stable branches before we move on to subteams?
21:20:38 <melwitt> #topic Subteam Highlights
21:20:54 <melwitt> efried: anything to highlight for scheduler?
21:21:34 <cdent> efried was gone for most of december
21:21:40 <mriedem> planning an "aggressive checkpoint" meeting to discuss extracted placement the week of jan 14
21:21:50 <cdent> main thing I'd highlight is what mriedem just said
21:22:05 <efried> yeah, I don't have anything. Gonna miss next week's sched mtg too. Hopefully back in the swing after that.
21:22:40 <melwitt> right, cool. thanks
21:22:41 <cdent> I'll be writing a big catch up placement update tomorrow which will try to get everything back into view
21:23:17 <melwitt> ok, nice
21:23:46 <melwitt> ok, moving on to stuck reviews
21:23:49 <melwitt> #topic Stuck Reviews
21:23:54 <melwitt> we actually have one this week
21:24:08 <melwitt> per-instance live migration timeout/action http://lists.openstack.org/pipermail/openstack-discuss/2018-December/001112.html
21:24:25 <mriedem> that might be old
21:24:28 <mriedem> but yeah
21:24:43 <mriedem> i've pung dan but would like other feedback given the deadline
21:25:15 <melwitt> ok. I'll take a look at it and if anyone else can read the thread and weigh in on the review, it would be helpful
21:25:45 <mriedem> also,
21:25:46 <mriedem> not really stuck,
21:25:57 <mriedem> but there is an open question on https://review.openstack.org/#/c/612531/ - the per instance serial number spec,
21:25:58 <patchbot> patch 612531 - nova-specs - Per-instance serial number (spec) - 1 patch set
21:26:01 <mriedem> i think i've summarized at the bottom
21:26:15 <mriedem> tl;dr should we add a config option choice in addition to the image/flavor meta
21:27:09 <mriedem> gibi's bw provider stuff is kind of stuck too...
21:27:14 <mriedem> details in the ML thread
21:27:27 <mriedem> the big question on that one is about discoverability
21:27:35 <mriedem> since he doesn't have a microversion proposed in his series
21:28:01 <melwitt> oh, interesting. I hadn't caught up to the latest on the thread yet
21:28:03 <mriedem> i don't know how others feel about a microversion for that feature
21:28:22 <sean-k-mooney1> mriedem: it strictly is not changing an nova api correct
21:28:31 <mriedem> it's a behavior change
21:28:43 <melwitt> I would have thought we'd have a microversion that means "new ability now available"
21:28:53 <melwitt> otherwise how would you know when you can start doing it
21:28:58 <mriedem> it's a question of, does this cloud honor ports with qos on them
21:29:18 <sean-k-mooney1> yes have wew had micro version bumps for pure behavioral change in the past. i think it would be fine but im not sure what we have done before
21:29:24 <cfriesen> if we specify an old microversion would we turn off the new behaviour?
21:29:45 <mriedem> we did a microversion for volume multiattach which was more or less a behavior change
21:29:54 <mriedem> before^ you could attach a multiattach volume to exactly one server
21:30:10 <mriedem> but w/o a microversion, how would you know when you can try to attach it to multiple servers?
21:30:52 <sean-k-mooney1> in that case it think it woudl be resonable to have micro version which leave teh question of have hapens if you use an old microvirion
21:30:54 <mriedem> cfriesen: well one way to do that is if it's an old microversion and the port has resource requests, you fail
21:31:05 <mriedem> ^ is what we do for multiattach volumes
21:31:42 <mriedem> https://github.com/openstack/nova/blob/master/nova/compute/api.py#L4067
21:31:45 <cfriesen> what would happen now if you have resource requests?
21:32:06 <cfriesen> because I'm thinking the "old microversion" case should behave like it would now
21:32:23 <sean-k-mooney1> ya that makes sense since we  dont support this when the qos policy is applied to the network currently.
21:32:51 <sean-k-mooney1> cfriesen: well before this change set nova would have had no code to read teh resouce requestr so it would have been ignored
21:33:05 <mriedem> right today those kinds of ports are just ignored
21:33:21 <mriedem> so is my qos not being honored because nova is old? or something failed? or i screwed up?
21:33:22 <mriedem> it's a confusing ux
21:34:09 <cfriesen> agreed, so does it make more sense to fix up the ux or to preserve current behaviour?
21:34:25 <cfriesen> with the current microversion on new code I mean
21:34:35 <sean-k-mooney1> mriedem: well neutron will still enforce it as best it can nova will just not make the allocation in placement to prevent the oversubsription. i think people are generally infavor of a micro version however
21:35:18 <melwitt> yeah, it seems like not using a microversion and rejecting resource requests when older microversion specified would make things needlessly random from a user perspective. i.e. just ignoring resource requests if microversion < required
21:35:33 <mriedem> i've asked the questoin in the ML so thoughts there are probably better than here given the number of people not here
21:35:59 <melwitt> yeah, good point. let's get on the thread
21:36:55 <melwitt> ok, so we have the live migration review to weigh in on, the per instance serial number spec, and the bandwidth aware scheduling thread
21:37:16 <melwitt> anything else before we move on to open discussion?
21:37:26 <melwitt> #topic Open discussion
21:37:37 <melwitt> some spec review requests in here
21:37:58 <melwitt> persistent-memory specs review https://review.openstack.org/#/c/601596/ Review revealed two resource tracking approaches. Need further discussion on community preference.
21:37:59 <patchbot> patch 601596 - nova-specs - support virtual persistent memory - 6 patch sets
21:38:20 <melwitt> I'm seeing active review today on that one
21:38:46 <melwitt> cpu-model-selection spec review https://review.openstack.org/#/c/620959
21:38:47 <patchbot> patch 620959 - nova-specs - Select cpu model from a list of cpu models (MERGED) - 9 patch sets
21:38:57 <melwitt> oh, maybe these are old
21:38:58 <mriedem> those are also old
21:38:59 <mriedem> yeah
21:39:29 <melwitt> ok, well last one since I already linked those
21:39:39 <melwitt> flavor-extra-spec-image-property-validation spec review https://review.openstack.org/#/c/618542/
21:39:40 <patchbot> patch 618542 - nova-specs - Flavor Extra Spec and Image Properties Validation - 6 patch sets
21:39:52 <melwitt> that one has one +2 already
21:40:07 <melwitt> that's all we have on the agenda
21:40:14 <mriedem> i forgot to mention during stuck/stable,
21:40:22 <mriedem> but other stable cores should probably be aware of https://review.openstack.org/#/c/625880/ as well
21:40:23 <patchbot> patch 625880 - nova (stable/rocky) - Fail to live migration if instance has a NUMA topo... - 1 patch set
21:40:33 <mriedem> it's stephenfin's change to disallow numa-aware live migration by default
21:40:41 <melwitt> yeah, I saw that one
21:40:53 <mriedem> since that's a behavior change, it's controversial for stable,
21:41:02 <mriedem> but tends to be a 'don't break by default' kind of fix
21:41:05 <mriedem> like a cve
21:41:09 <mriedem> tonyb had a nice summary
21:41:16 <mriedem> dansmith: ^
21:41:26 <melwitt> aye
21:41:46 <mriedem> sounds like it's on me to approve, but wanted to get other people looking
21:43:20 <melwitt> ok, I'll add my review to it. but I'm thinking it's ok based on what tonyb explained
21:44:04 <melwitt> anyone have anything else for open discussion before we wrap up?
21:44:29 <melwitt> ok, thanks everyone
21:44:33 <melwitt> #endmeeting