14:00:03 <bauzas> #startmeeting nova
14:00:07 <openstack> Meeting started Thu Jun  1 14:00:03 2017 UTC and is due to finish in 60 minutes.  The chair is bauzas. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:08 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:10 <openstack> The meeting name has been set to 'nova'
14:00:12 <bauzas> howdy
14:00:14 <takashin> o/
14:00:15 <edleafe> \o
14:00:22 <dansmith> o.
14:00:28 <bauzas> mriedem can't make all the meeting, so chairing if you agree
14:00:30 <alex_xu> o/
14:00:40 <cdent> o/
14:00:48 <bauzas> #link https://wiki.openstack.org/wiki/Meetings/Nova
14:01:12 <bauzas> fortunately, he did filled the agenda, so I can still be lazy :)
14:01:24 * edleafe wonders about dansmith's tiny arm
14:01:41 <dansmith> edleafe: it's my shoulder
14:01:47 <dansmith> or something
14:01:52 * edleafe wonders about dansmith's tiny shoulder
14:01:55 <dansmith> heh
14:02:03 <sfinucan> o/
14:02:14 <bauzas> let's start
14:02:14 <efried> o/
14:02:21 <bauzas> #topic Release news
14:02:26 <bauzas> #link Pike release schedule: https://wiki.openstack.org/wiki/Nova/Pike_Release_Schedule
14:02:32 <bauzas> #info Next upcoming milestone: Jun 8: p-2 milestone (1 week)
14:02:37 <bauzas> #info Blueprints: 72 targeted, 68 approved, 15 completed, 4 not started
14:03:16 <bauzas> given pike-2 is in one week, mriedem had a thing to make sure we don't diverge too much
14:03:19 <bauzas> #info Blueprints which are blocked, not started or do not have an  owner after the p-2 milestone will likely be deferred to Queens.
14:03:23 <bauzas> comments ?
14:03:33 <bauzas> folks not scary about that?
14:03:38 <bauzas> personnally, I'm fine with
14:04:24 <bauzas> in other words, if you haven't started yet to work on your implementation, please consider it as urgent for your week :)
14:04:36 <bauzas> I take it as a silent nodding
14:04:53 <bauzas> #topic Bugs
14:05:09 <bauzas> #link Critical bugs https://bugs.launchpad.net/nova/+bugs?search=Search&field.importance=Critical&field.status=New&field.status=Incomplete&field.status=Confirmed&field.status=Triaged&field.status=In+Progress
14:05:20 <bauzas> none I heard
14:05:36 <bauzas> #help Need help with bug triage; there are 96 (down 8 from last week) new untriaged bugs as of today (June 1).
14:06:03 <bauzas> and I truly blame myself for the above ^
14:06:17 <bauzas> throw e-tomatoes at me if you will
14:06:59 <bauzas> so the gate
14:07:05 <bauzas> #link check queue gate status http://status.openstack.org/elastic-recheck/index.html
14:07:54 <bauzas> since I was off last week, can someone point me if there were CI failures that required investigation ?
14:08:29 <bauzas> I can see a couple of rechecks along when reviewing, but I'd say the situation is steady, correct?
14:08:48 <sfinucan> afaik, yeah
14:08:57 <bauzas> cool
14:09:06 <bauzas> 3rd-party CI is more concerning me
14:09:13 <bauzas> #link 3rd party CI status http://ci-watch.tintri.com/project?project=nova&time=7+days
14:09:33 <bauzas> mriedem wrote an email 2 days ago about VMware NSX CI
14:10:01 <bauzas> AFAICS, some other 3rd-party CIs are having problems
14:10:35 <bauzas> and if you see at the latter, you'll see not a good shape
14:10:46 <bauzas> #info XenServer CI is having some known issues with Tempest SSH  validation enabled by default; has been removed from nova-ci voting  group for now.
14:10:52 <sfinucan> I would assume a lot of that is the setuptools issue, no?
14:10:57 <bauzas> probably
14:11:19 <efried> Link please?
14:11:35 <bauzas> efried: link for what ? mriedem's email ?
14:11:42 <efried> "the setuptools issue"
14:11:53 <sfinucan> efried: -openstackstatus/#openstack-nova- NOTICE: There is a known issue with setuptools 36.0.0 and errors about the "six" package. For current details see https://github.com/pypa/setuptools/issues/1042 and monitor #openstack-infra
14:11:58 <sfinucan> from the channel this morning
14:12:08 <efried> Thanks sfinucan - missed it.
14:12:14 <sfinucan> (y)
14:12:23 <bauzas> FWIW, the email I mentioned http://lists.openstack.org/pipermail/openstack-dev/2017-May/117667.html
14:12:49 <bauzas> anyway, let's see if things improve next week, or we could possibly need to reach stakeholders
14:13:06 <bauzas> #topic Reminders
14:13:12 <bauzas> #link Pike Review Priorities etherpad: https://etherpad.openstack.org/p/pike-nova-priorities-tracking
14:13:42 <bauzas> I can see some of the sections of the etherpad relevant, some not
14:13:58 <bauzas> I beg subteams to make sure the above is up to date
14:14:32 <bauzas> #topic Stable branches status
14:14:43 <bauzas> #link stable/ocata: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/ocata,n,z
14:14:52 <bauzas> #info Ocata 15.0.5 has been released.
14:14:59 <bauzas> #link stable/newton: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/newton,n,z
14:15:05 <bauzas> #info Newton 14.0.7 has been released.
14:15:22 <bauzas> I know we have a couple of bugfixes worth releasing
14:15:48 <bauzas> I'll try to review the above asap
14:16:33 <bauzas> moving on
14:16:50 <bauzas> #topic Subteam highlights
14:17:06 <bauzas> dansmith, you had meeting on last Wed, right ?
14:17:16 <dansmith> you mean yesterday?
14:17:25 <bauzas> I am
14:17:38 <dansmith> yes, we talked about a few bug fixes that are in the backport queue,
14:17:43 <dansmith> and outstanding patches like quotas and the last couple for multicell devstack
14:17:54 <dansmith> that's about it, moving along well
14:17:55 <bauzas> I guess quotas is the hot plate ?
14:18:03 <dansmith> yes
14:18:06 <bauzas> okay
14:18:19 <bauzas> and AFAIK, cells-aware BP is pretty all done, correct?
14:18:28 <bauzas> except the devstack side
14:18:56 <dansmith> I have a few more patches that I need to collect
14:19:04 <bauzas> okay
14:19:12 <dansmith> so no, but most of the ones that are up are merged or merged last night yeah
14:19:25 <bauzas> cool
14:19:45 <bauzas> so, we had no scheduler meeting on Monday given Memorial Day
14:19:52 <bauzas> passing tho
14:20:25 <bauzas> can someone tell me if we had an API meeting yesterday?
14:20:34 <alex_xu> yes, but very short
14:20:59 <alex_xu> the spec for quota-class API bug is up https://review.openstack.org/468619
14:21:00 <patchbot> patch 468619 - nova-specs - Fix to return server groups quota in the quota-cla...
14:21:11 <alex_xu> that is all
14:22:01 <bauzas> alex_xu: well, it's a spec, right?
14:22:09 <mriedem> it's a spec for a bug fix
14:22:14 <mriedem> regression in v2.1
14:22:15 <bauzas> ah ok
14:22:18 <mriedem> for the quota class sets api
14:22:18 <alex_xu> bauzas: yea, a API bug, but it need to be fix by microversion
14:22:23 <bauzas> okay
14:22:33 <bauzas> so not impacted by the spec freeze then
14:22:36 <mriedem> no
14:22:38 <bauzas> mriedem: thanks for the explanation
14:22:45 <mriedem> i'm back so watch out
14:22:53 <bauzas> hah
14:23:08 * bauzas hides the nasty things because daddy is back
14:23:16 <dansmith> umm
14:23:33 <bauzas> so, gibi said no notifications meeting
14:23:38 <bauzas> moving on quickly
14:23:54 <bauzas> aaaaand, perfect timing
14:23:58 <bauzas> Cinder, mriedem ?
14:24:07 <mriedem> yeah, link is in there
14:24:10 <bauzas> #link need reviews starting here https://review.openstack.org/#/c/469579/ to use new microversion when attaching volumes
14:24:11 <patchbot> patch 469579 - nova - Use microversions for new style volume attachments
14:24:19 <bauzas> okay
14:24:24 <mriedem> i put up patches yesterday for version discovery with cinder,
14:24:33 <mriedem> which is needed to send 3.27 for the new attachment flow
14:24:48 <mriedem> pretty simple changes
14:25:02 <mriedem> jgriffith has a poc for attach with the new flow that is going to be rebased on top of that series
14:25:15 <mriedem> and steve noyes is working on the live migration conversion
14:25:24 <mriedem> oh speaking of,
14:25:34 <mriedem> https://review.openstack.org/#/c/469740/ needs to get in
14:25:35 <patchbot> patch 469740 - nova - Fix call to driver_detach in remove_volume_connection
14:25:42 <mriedem> live migration rollback for volume-attached instances was regressed in pike
14:26:04 <mriedem> i see i'm hitting the setuptools issue now though
14:26:07 <bauzas> yup
14:26:11 <mriedem> anyway, simple fix
14:26:14 <bauzas> CCd
14:26:28 <bauzas> #topic Stuck reviews
14:26:35 <bauzas> nothing in the agenda
14:26:56 <bauzas> I guess we can continue
14:27:05 <bauzas> #topic Open discussion
14:27:09 <bauzas> there was an item
14:27:17 <bauzas> mhenkel: you there ?
14:27:25 <mhenkel> yes, this was my BP on dpdk
14:27:29 <mhenkel> hi all!
14:27:38 <bauzas> the question is "Can https://blueprints.launchpad.net/nova/+spec/opencontrail-dpdk-support for a new VIF be approved as a spec-less blueprint?"
14:28:02 <mhenkel> I added more explanation to the BP today
14:28:25 <mhenkel> can you let me know if that is sufficient or if I am still missing something
14:28:28 <mriedem> i thought jaypipes had questions in there about doing stuff in os-vif?
14:28:40 <bauzas> yeah that was my question
14:28:47 <bauzas> we already have os-vif for that
14:28:47 <mriedem> sean-k-mooney and jaypipes are the authorities here
14:29:13 <mhenkel> yes, and I talked with both on it and went through the code
14:29:34 <mhenkel> from a code and integration point of view they were fine with the approach
14:29:42 <mriedem> so,
14:29:44 <mriedem> it's not in os-vif,
14:29:47 <mriedem> it's a plugin https://github.com/Juniper/contrail-nova-vif-driver right?
14:29:54 <mhenkel> that is correct
14:30:19 <sfinucan> which, tbf, is how it's supposed to be done
14:30:23 <mriedem> wow https://github.com/Juniper/contrail-nova-vif-driver/blob/master/nova_contrail_vif/contrailvif.py
14:30:27 <mriedem> ^ scares me a bit
14:30:36 <mhenkel> why is that?
14:30:51 <mriedem> monkey patching the networking api
14:31:32 <qwebirc93460> FYI.. this is the actual plugin - https://github.com/Juniper/contrail-nova-vif-driver/tree/master/vif_plug_vrouter
14:32:42 <mhenkel> correct
14:34:21 <mriedem> well i guess i don't have major objections to this,
14:34:26 <mriedem> it's all just passthrough to the plugin
14:34:45 <mriedem> i have no idea if it works of course, or will continue to work
14:34:50 <mriedem> but that's what you get if you're a plugin
14:35:09 <mhenkel> currently it works ;)
14:35:31 <mhenkel> but you are right, we need to track any interface changes
14:35:42 <mriedem> i'll talk to jay and sean about it when they are around
14:35:46 <mhenkel> but that is in the nature of plugins
14:35:53 <mhenkel> ok, thanks
14:36:18 <mhenkel> and can you let me know if the BP now contains the necessary details?
14:37:24 <bauzas> I think there was a good argument about that not specifiically tied to DPDK
14:37:47 <mriedem> mhenkel: also,
14:37:55 <mriedem> isn't "vrouter" a rather generic vif type name?
14:38:18 <mriedem> why not contrail-vrouter or something?
14:38:34 <mhenkel> mriedem: no, vrouter is OpenContrails ovs/linux bridge replacement
14:38:48 <mriedem> i get that,
14:39:01 <mriedem> but what happens when ACME SDN has a vif type of vrouter that collides with this?
14:39:07 <mriedem> ACME SDN being another vedor
14:39:08 <mriedem> *vendor
14:39:34 <mhenkel> that is probably true. I can rename it to contrail-vrouter
14:39:58 <bauzas> yup, another BP URL and naming would suit me
14:40:26 <bauzas> I guess that's all we have for today's meeting
14:40:27 <mriedem> anyway let's not do all this in the meeting
14:40:39 <mhenkel> ok
14:40:42 <bauzas> any other items to raise to the community?
14:41:09 <bauzas> again, keeping in mind that pike-2 is literally next week
14:41:58 <bauzas> if so, I can call it a day
14:42:03 <bauzas> #endmeeting