16:00:30 <gibi> #startmeeting nova
16:00:34 <openstack> Meeting started Thu Mar 11 16:00:30 2021 UTC and is due to finish in 60 minutes.  The chair is gibi. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:35 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:37 <openstack> The meeting name has been set to 'nova'
16:01:18 <dansmith> o/
16:01:37 <stephenfin> o/
16:02:03 <gibi> o/
16:02:15 <artom> ~o~
16:02:25 <gibi> #topic Bugs (stuck/critical)
16:02:30 <gibi> No Critical bugs
16:02:35 <gibi> #link 18 new untriaged bugs (+5 since the last meeting): #link https://bugs.launchpad.net/nova/+bugs?search=Search&field.status=New
16:02:43 <lyarwood> \o
16:02:51 <gibi> I've closed ~80 bugs opened about fault injections
16:03:07 <gibi> it was fun
16:03:23 <artom> :O
16:03:24 <gibi> is there any bug we need to talk about here?
16:03:52 <gibi> (btw this was the first fault injection bug https://bugs.launchpad.net/nova/+bug/1918340 )
16:03:53 <openstack> Launchpad bug 1918340 in OpenStack Compute (nova) "Fault Injection #1 - improve unit test effectiveness" [Undecided,Incomplete]
16:04:26 <stephenfin> fun
16:04:35 <stephenfin> gibi: sorry you had to go through that :)
16:04:56 <stephenfin> we can spend time on the bug backlog next week, I guess
16:05:03 <gibi> sure
16:05:05 <gmann> o/
16:05:10 <dansmith> niice
16:05:26 <gibi> if no needy bug then moving on
16:05:26 <gibi> #topic Gate status
16:05:34 <gibi> gate is in a shape as expected around FF
16:05:56 <dansmith> pretty decent, considering actually
16:06:03 <gibi> please try to recheck only things that are either bugs or has a chance to be approved before EOB today
16:06:16 <gibi> s/bugs/bugfixes/
16:06:43 <gibi> #topic Feature status
16:06:58 <gibi> let meg copy a list from the agenda
16:07:03 <gibi> #link https://blueprints.launchpad.net/nova/+spec/nova-support-webvnc-with-password-anthentication : in merge conflict with negative review so I assume it is deferred from Wallaby
16:07:11 <gibi> #link https://blueprints.launchpad.net/nova/+spec/compact-db-migrations-wallaby : patches has been approved, implementation going through the gate
16:07:19 <gibi> #link https://blueprints.launchpad.net/nova/+spec/modernize-os-hypervisors-api : merged, the policy part is deferred out from Wallaby
16:07:20 * bauzas is around
16:07:28 <gibi> #link https://blueprints.launchpad.net/nova/+spec/libvirt-default-machine-type : merged, except the integration testing. That can still land after FF.
16:07:38 <gibi> #link https://blueprints.launchpad.net/nova/+spec/pci-socket-affinity : expected to be approved today
16:07:44 <gibi> #link https://blueprints.launchpad.net/nova/+spec/port-scoped-sriov-numa-affinity : expected to be approved today
16:07:49 <gibi> #link https://blueprints.launchpad.net/nova/+spec/libvirt-vdpa-support : still open
16:07:53 <gibi> #link https://blueprints.launchpad.net/nova/+spec/allow-secure-boot-for-qemu-kvm-guests : still open
16:08:04 <stephenfin> for nova-support-webvnc-with-password-anthentication, we're deferring
16:08:05 <bauzas> the secure boot series got +Wd
16:08:05 <gibi> the rest is either merged or in a shape that means it is deferred
16:08:21 <gibi> bauzas: thanks that is good news
16:08:33 <gibi> stephenfin: ack
16:08:34 <stephenfin> Melwitt and I took a close look at it and there were some serious issues with usability that meant we weren't happy merging it
16:08:42 <bauzas> we still have a few patches up, but those are FUPs that can merge after FF
16:08:56 <gibi> bauzas: sure
16:09:23 <gibi> is there any other bp that I'm not listed up there but the author still thinks it has a chance to land?
16:09:30 <stephenfin> regarding https://blueprints.launchpad.net/nova/+spec/libvirt-vdpa-support I've been going through that with Sean and helping them reorder things so they're fit to merge. I've gone through it with a fine comb now and am happy with all but a few things
16:09:56 <stephenfin> problem is you need some PCI knowhow to really review, hence why it's got so few eyes I suspect
16:10:26 <gibi> I can take  a look at one series before I time out for today
16:10:39 <gibi> so if nothing else pops up then I will look at the vdpa
16:11:22 <stephenfin> that would be great
16:11:30 <gibi> one more thing I not listed above is the smartnic support. I finished with the review today but I had a lot of findings so I assume that is also deferred
16:12:05 <gibi> anything else about features?
16:12:46 <stephenfin> the other thing I had on my list was the change to enable 'openstack server --hostname'
16:12:56 <stephenfin> it's small, but it's had no attention so I think we should defer
16:13:00 <stephenfin> and merge early in Xena
16:13:06 <gibi> yeah, I had no time to dig into that
16:13:12 <gibi> sorry
16:13:16 <dansmith> is that just an osc thing,
16:13:18 <stephenfin> All good. I left it late
16:13:19 <dansmith> or requires an api?
16:13:24 <stephenfin> Nah, microversion change
16:13:42 <dansmith> okay
16:13:46 <stephenfin> just a new field in the request, but any microversion change brings a lot of boilerplate changes with it
16:13:54 <stephenfin> *server create request
16:13:55 <dansmith> sure sure,
16:14:03 <dansmith> just wasn't sure if we had it already and just needed to pass it or not
16:14:28 <stephenfin> not yet, no
16:14:32 <stephenfin> change is here for context https://review.opendev.org/c/openstack/nova/+/778550
16:14:48 <stephenfin> (ignore the fail - it's irrelevant)
16:15:02 <stephenfin> but like I said, I'm happy to defer. This was really a nice-to-have
16:15:10 <dansmith> yup. I was just curious
16:15:18 <stephenfin> ack
16:15:21 <gibi> anything else?
16:15:36 <stephenfin> not from me
16:15:41 <gibi> #topic Release Planning
16:15:48 <gibi> so we have Feature Freeze today EOB
16:16:03 <gibi> FFE process posted on the ML #link http://lists.openstack.org/pipermail/openstack-discuss/2021-March/020772.html
16:16:11 <gibi> but I not expecting FFE requests
16:16:49 <gibi> So from tomorrow lets focus on high priority bugs and patches that preparing the RC1
16:16:54 <gibi> RC1 is in two weeks
16:17:15 <gibi> (and do the rechecks for the already approved things_)
16:17:19 <dansmith> I think we're looking for another rev of the rpc bump
16:17:35 <gibi> I can put my eyes on it when needed
16:18:27 <gibi> release todos are tracked on the etherpad #link https://etherpad.opendev.org/p/nova-wallaby-rc-potential
16:18:54 <gibi> like cycle highlights, reno prelude and other necessities
16:19:37 <gibi> also please use wallaby-rc-potential tag on bugs that you consider warranting a new RC if fixed
16:19:55 <artom> Currently there are none, FWIW
16:19:58 <artom> (Yey?)
16:20:13 <gibi> because we are awesome and everything works :)
16:20:37 <gibi> any other release related topic to discuss
16:20:38 <gibi> ?
16:20:41 * dansmith hopes those words tase good later
16:20:44 <dansmith> *taste
16:21:01 <gibi> I'm not naive I'm sarcastic :)
16:21:30 <dansmith> heh I know :P
16:21:36 <gibi> :)
16:21:47 <gibi> oh one more thing the TC and PTL nominations are closed this week
16:22:04 <gibi> as thing looks you stuck with me as the PTL for the next 6 months
16:22:34 <gmann> \o/
16:22:52 <gibi> #topic PTG planning
16:22:59 <gibi> PTG date had been announced: April 19-23, 2021
16:23:05 <gibi> Use the PTG etherpad to collect topics: #link https://etherpad.opendev.org/p/nova-xena-ptg
16:23:18 <gibi> Please fill the doodle about the timing of the nova slots before end of next week #link https://doodle.com/poll/ib2eu3c4346iqii3
16:23:36 <gibi> as I have to "book" those virtual roomns
16:23:57 <gibi> any questions about the PTG?
16:24:12 <lyarwood> gibi: did you have a deadline in mind for topics?
16:24:51 <gibi> I will do etherpad reorg before the PTG week
16:25:13 <gibi> so that is sort-of deadline. anything comes after might end up at the end of the pad
16:25:33 <lyarwood> yeah ack that's fine, just something to keep in mind as stuff comes in from downstream ;)
16:25:55 <gibi> but if you have topics that needs more than the nova team (e.g. cinder, neutron, cyborg) then please prioritize those as we need to find a common time with the other teams then
16:26:07 <lyarwood> yup understood
16:26:40 <gibi> anything else about the PTG?
16:27:09 <gibi> #topic Stable Branches
16:27:20 <gibi> copying elod's update from the agenda
16:27:23 <gibi> lyarwood found issue with ceph plugin, that seems to block now *train* and older branches -- fix: https://review.opendev.org/c/openstack/nova/+/780014
16:27:27 <gibi> victoria + ussury should be OK
16:27:32 <gibi> elod prepares victoria + ussuri + train(?) release patches tomorrow, March 12th
16:27:35 <gibi> EOM
16:27:51 <gibi> thanks elod
16:27:55 <lyarwood> FWIW I had nothing urgent to land ahead of the releases elod
16:28:12 <elod> lyarwood: ack
16:28:53 <gibi> if nothing else then I skipping over the libvirt subteam topic as bauzas told me that nothing to report there
16:29:08 <gibi> #topic Open discussion
16:29:12 <gibi> nothing on the agenda
16:29:19 <gibi> is there anything we need to discuss?
16:29:28 <dansmith> I have a thing
16:29:51 <gibi> go
16:30:07 <dansmith> so, I want to do that funny thing and ask a glance question in the nova meeting.. BUT I have a good reason :)
16:30:23 <dansmith> is anyone here familiar with the glance metadef api, how it may be used, and by who?
16:30:33 <dansmith> the reason I ask is that it seems mostly there to collect nova-related metadata
16:31:16 <dansmith> as was noted earlier on the list, there are multiple low-grade security problems with the API, and the glance team needs some info from people that actually use it in order to know how and when to fix it
16:31:37 <dansmith> since it's nova metadata (i.e. image metadata that affects instances) related, I thought maybe someone here might know
16:32:13 <stephenfin> dansmith: sean isn't here, weirdly, but they talk about it a lot
16:32:19 <gibi> besides some common image metadata usage I did not see metadef usage in our downstream bug tracker
16:32:41 <dansmith> gibi: ack, that's good to know, not in ours either :)
16:32:55 <dansmith> stephenfin: ah yeah I thought sean might be a candidate, but didn't notice he wasn't in here
16:32:58 <stephenfin> they wanted to use it for the extra spec validation stuff we did in microversion 2.87 but we found it was mega out-of-date and the definitions were really hard to define/extend
16:33:12 <stephenfin> xml-- iitv
16:33:14 <stephenfin> *iirc
16:33:43 <dansmith> so they thought maybe it would be useful, but decided not to try, is that right?
16:33:44 <dansmith> not.. are using it, or were using it, or..
16:34:10 <stephenfin> yeah, correct. I added that microversion and just went with our own schema format
16:34:19 <dansmith> it seems to be a "feature" that was "designed" and dropped on the community all in a short period, and then the originators are all long gone,
16:34:31 <stephenfin> yeah, that's my impression too
16:34:32 <dansmith> so it's hard to know if anyone actually uses it, whether it serves any purpose anymore, etc
16:34:35 <stephenfin> I think Horizon might use it too
16:34:51 <dansmith> yeah, horizon and heat are the two things we're wondering about
16:34:56 <stephenfin> when you're creating the guest I see some "CIM" metadata and I think that's coming from the Glance API
16:34:57 <dansmith> horizon was one of the original goals, I think, but..
16:35:16 <dansmith> okay there are some CIM things in there by default, so maybe that's right
16:35:51 <stephenfin> if it were an option, deprecating it in Glance might make more sense. nova already validates image metadata and flavor extra specs for you
16:35:56 <dansmith> by default, it's completely wide open, any user can create anything they want, and as much as they want (until the DB is full), and there isn't proper tenancy protection
16:36:08 <dansmith> deprecating is an option,
16:36:21 <dansmith> we're recommending disabling it in policy by default and unless you want or need it for something
16:36:31 <dansmith> but we want to know if we should fix or deprecate
16:36:43 <stephenfin> and we (nova) could add an API if people wanted a way to retrieve information about the metadata we support (extra specs and image meta props) ahead of time, though I suspect that's a very small use case
16:36:52 <dansmith> I think the goal is to make things like horizon able to offer up common metadata, not so much for the validation part
16:37:25 <dansmith> yeah, basically a serialized ImageMetadataProperties representation I guess
16:37:38 <stephenfin> yeah
16:37:50 <stephenfin> nova at least will ignore everything else
16:37:58 <stephenfin> though nova isn't the only thing talking to glance of course
16:38:05 <dansmith> aight, so gibi you just checked your downstream bug tracker in response to the OSSN I guess to see if your customers were using it?
16:38:13 <dansmith> or at least, reporting issues with it I guess
16:38:17 <gibi> yes
16:38:24 <stephenfin> (I say of course but I'm not actually sure if anyone else talks to Glance)
16:38:25 <gibi> nothing really popped up
16:38:29 <dansmith> and are you planning to just push out policy to disable?
16:38:55 <gibi> I don't have the infor about that yet
16:38:57 <gibi> but I can ask around
16:39:07 <dansmith> okay
16:39:07 <gibi> I would disable with policy yes
16:39:45 <dansmith> okay, well, we're asking ops to come to ptg with use cases or data about it,
16:40:05 <dansmith> not sure if we'll get any, but if you have anything that comes from asking that you want me to point out there, let me know
16:40:26 <dansmith> and I'll poke sean after this
16:40:32 <dansmith> that's all, thanks for the discussion
16:40:33 <gibi> sounds good
16:40:45 <gibi> anything else for today?
16:41:55 <gibi> then thanks for joining today. Happy FF folks! :)
16:42:07 <stephenfin> thanks! o/
16:42:12 <gibi> #endmeeting