14:00:19 <mriedem> #startmeeting nova
14:00:20 <openstack> Meeting started Thu Aug 24 14:00:19 2017 UTC and is due to finish in 60 minutes.  The chair is mriedem. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:21 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:23 <openstack> The meeting name has been set to 'nova'
14:00:38 <efried> \o
14:00:42 <dansmith> o/
14:00:43 <stephenfin> o/
14:01:14 <edleafe> \o
14:01:49 <mriedem> alright i'll assume there are others lingering
14:01:54 <cdent> o/
14:01:55 <mriedem> #link agenda https://wiki.openstack.org/wiki/Meetings/Nova
14:02:02 <mriedem> #topic release news
14:02:09 <mriedem> #link Pike release schedule: https://wiki.openstack.org/wiki/Nova/Pike_Release_Schedule
14:02:14 <mriedem> #info August 24 (today) is Pike RC2
14:02:22 <mriedem> #link Pike RC TODO etherpad: https://etherpad.openstack.org/p/nova-pike-release-candidate-todo
14:02:27 <mriedem> there are a few things in there
14:02:43 <mriedem> dansmith: https://review.openstack.org/#/c/496997/ is one backport
14:02:51 <dansmith> ack
14:02:52 <mriedem> https://review.openstack.org/#/c/497237/ is another
14:02:56 <mriedem> those are the two from yesterday
14:03:12 <mriedem> stephenfin also backported the nova-manage cell command deprecation patch https://review.openstack.org/#/c/497166/
14:03:24 <mriedem> as i said yesterday, i don't feel a need to push this in, but don't care too much either way
14:03:28 <dansmith> I thought you wanted to wait on that?
14:03:39 <dansmith> yeah, our rc2 list is pretty long at the moment
14:04:06 * alex_xu waves late
14:04:08 <mriedem> the other known one that will be coming is alex_xu's patch for removing allocations on reschedule https://review.openstack.org/#/c/496995/
14:04:23 <dansmith> yeah will look at that when we're donehere
14:04:25 <mriedem> which has me going back and forth not on the reschedule part, but the abort part
14:04:43 <mriedem> also, ^ doesn't handle the case that we reschedule during migrate/resize, which is a similar bug
14:05:00 <mriedem> i think that's a fairly easy fix on it's own, which i could push on top of alex's after we're done
14:05:19 <cdent> Is https://bugs.launchpad.net/nova/+bug/1712816 agreed as rc potential? recently tagged as such
14:05:20 <openstack> Launchpad bug 1712816 in OpenStack Compute (nova) "reporting incorrect allocations for "DISK_GB" in case of volume-backed instances" [Undecided,New]
14:06:21 <mriedem> i think that's the same person that was asking about https://review.openstack.org/#/c/428481/
14:06:58 <cdent> yeah, it links to that
14:07:31 <mriedem> so, isn't this a latent issue already where we've always incorrectly claimed disk during boot from volume?
14:07:32 <mriedem> dansmith: ^?
14:07:42 <dansmith> mriedem: yes it is
14:07:48 <mriedem> ok, then not rc2
14:07:58 * cdent removes the tag
14:08:52 <mriedem> we could probably also duplicate the bug
14:09:04 * kashyap is lingering around, but forgot to wave hi
14:09:46 <mriedem> ok so we'll sort out alex's fix after the meeting, i'll start on the resize + reschedule fix after the meeting, and then hopefully that's it for rc2
14:10:03 <mriedem> #info Re-propose specs for Queens.
14:10:10 <mriedem> i know people are already doing this
14:10:25 <mriedem> oh which reminds me
14:10:46 <mriedem> melwitt has a nova-specs patch to add an 'upgrade impact' section to the queens template, which would probably be good to get in early before people start writing new specs for queens https://review.openstack.org/#/c/456756/
14:10:59 <cdent> that’s a good idea
14:11:09 <dansmith> we didn't already have that section?
14:11:10 <mriedem> #link mel's nova-specs queens template upgrade impact section patch https://review.openstack.org/#/c/456756/
14:11:12 <dansmith> I could have sworn
14:11:14 <mriedem> dansmith: nope
14:11:25 <edleafe> So no long-suffering pike specs can be approved? Thinking of Jay's: https://review.openstack.org/#/c/471927/
14:11:29 <mriedem> dansmith: there are some things interspersed in other sections, as i pointed out in her PS1
14:11:37 <dansmith> hmm okay
14:11:53 <mriedem> edleafe: yes that's basically docs at this point, i just haven't gotten back to it yet
14:12:23 <edleafe> mriedem: ok, cool
14:12:26 <mriedem> alright moving on
14:12:29 <mriedem> #topic bugs
14:12:31 * edleafe wipes sweat off brow
14:12:43 <mriedem> nothing marked as critical
14:12:57 <mriedem> #info 7 new untriaged bugs (up 6 from last week)
14:13:20 <mriedem> i have been going through these daily looking for rc potential stuff,
14:13:31 <mriedem> and will say, bug triage is much nicer when there are only <10 to go through
14:14:08 <mriedem> #info pike-rc-potential bugs: https://bugs.launchpad.net/nova/+bugs?field.tag=pike-rc-potential
14:14:24 <mriedem> those are the known issues,
14:14:32 <mriedem> i'll report a new bug for the resize + reschedule bug
14:14:41 <mriedem> gate status is ok
14:14:45 <mriedem> 3rd party ci - no news
14:14:51 <mriedem> #topic reminders
14:15:01 <mriedem> #link Consider signing up for the Queens PTG which is the week of September 11 in Denver, CO, USA: https://www.openstack.org/ptg
14:15:07 <mriedem> #link Pike PTG topics etherpad: https://etherpad.openstack.org/p/nova-ptg-queens
14:15:20 <mriedem> ^ is still unordered
14:15:28 <mriedem> i won't be going into that until after pike rc2 is done
14:15:38 <mriedem> #info Be a star and consider the Pike project update interview at the PTG: http://lists.openstack.org/pipermail/openstack-dev/2017-August/121369.html
14:16:13 <mriedem> #topic stable branch status
14:16:19 <mriedem> #link stable/ocata: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/ocata,n,z
14:16:24 <mriedem> #info Ocata 15.0.7 released on 8/22
14:16:34 <mriedem> thanks to all that helped push through the ocata reviews
14:16:42 <mriedem> #link stable/newton: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/newton,n,z
14:16:52 <mriedem> #info Newton 14.0.8 release request: https://review.openstack.org/#/c/496285/
14:16:55 <mriedem> ditto for newton
14:17:08 <mriedem> #topic subteam highlights
14:17:24 <mriedem> there was no cells v2 meeting this week, we'll have one next week to talk about PTG stuff
14:17:29 <mriedem> edleafe: scheduler highlights?
14:17:49 <edleafe> and thought it should be approved soon.
14:17:49 <edleafe> #link Allocation spec https://review.openstack.org/#/c/471927/
14:17:49 <edleafe> The patches for returning alternates and allocations are up for review.
14:17:52 <edleafe> #link Add alternates and allocation, series starting with https://review.openstack.org/486215/
14:17:55 <edleafe> The way this complex set of data works is troubling, so I wrote up a post to start some discussion on improving things:
14:17:58 <edleafe> #link Scheduler data issues https://blog.leafe.com/handling-unstructured-data/
14:18:01 <edleafe> Discussed the need for refactoring the placement code, as it's a lot of stuff crammed into a couple of Nova modules. Only real solution is making placement a separate repo so we can break up the monolithic modules.
14:18:14 <edleafe> Doh! Can't copy/paste
14:18:33 <edleafe> First line should have been:
14:18:34 <edleafe> Discussed Jay's spec for returning allocation requests to the scheduler , and thought it should be approved soon.
14:18:43 <edleafe> ok, now I'm done
14:19:10 <mriedem> ok
14:19:31 <mriedem> and dansmith seems to have started on making the migration record the allocation consumer during moves
14:19:40 <dansmith> yeah
14:20:13 <mriedem> alex_xu: api meeting highlights?
14:20:26 <alex_xu> we go through the queens release work items
14:20:58 <alex_xu> gmann fill the no-more-extension-in-queens stuff in the ptg etherpad, and going to prepare some sample patch
14:21:48 <alex_xu> also discussion if there is a way to run the sample tests/unittests on the latest microversion and without create another huge numbers work items which may across multiple releases
14:21:48 <alex_xu> that is all we have
14:22:15 <mriedem> ok
14:22:24 <mriedem> i thought you guys all loved multi-release mega efforts :)
14:22:34 <mriedem> that's basically all we do in nova now
14:22:46 <mriedem> there was no notification meeting this week
14:22:49 <alex_xu> heh :)
14:22:58 <mriedem> and the nova/cinder meeting is today
14:23:13 <mriedem> from what i gathered before vacation,
14:23:33 <mriedem> cinder is going to need another microversoin for a new api that nova needs to call when we finish an attachment on the host
14:23:39 <mriedem> and on detach from the host
14:24:05 <mriedem> but i'm fuzzy on details
14:24:10 <mriedem> #topic stuck reviews
14:24:18 <mriedem> there was nothing on the agenda
14:24:25 <mriedem> does anyone need to mention something here?
14:24:41 <smcginnis> mriedem: Cinder API microversion 3.44 - fyi.
14:24:53 <mriedem> smcginnis: already landed?
14:25:16 <mriedem> nvm we can talk after the meeting
14:25:20 <mriedem> #topic open discussion
14:25:21 <smcginnis> mriedem: Yes, but only in Queens unfortunately.
14:25:43 <mriedem> there is nothing on the agenda for open discussion, so feel free to discuss openly if you want
14:26:20 <mriedem> ok i guess that's it, thanks everyone
14:26:23 <mriedem> #endmeeting