16:00:17 <gibi> #startmeeting nova
16:00:18 <opendevmeet> Meeting started Tue Jun  8 16:00:17 2021 UTC and is due to finish in 60 minutes.  The chair is gibi. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:19 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:21 <opendevmeet> The meeting name has been set to 'nova'
16:00:28 <bauzas> \o
16:00:32 <gibi> o/
16:00:42 <elodilles> o/
16:00:51 * kashyap waves
16:01:24 <gmann> o/
16:02:33 <gibi> I'm just back from a long weekend so I might not be up to date with things
16:02:38 <gibi> anyhow
16:02:41 <gibi> lets get started
16:02:56 <gibi> I do a bit of resuffle for todays agenda
16:03:03 <gibi> as kashyap has a topic but he needs to leave early
16:03:16 <gibi> (kashyap) https://blueprints.launchpad.net/nova/+spec/add-bochs-display-device
16:03:29 <gibi> kashyap: are you looking for bp approval?
16:03:37 <kashyap> gibi: Hi
16:03:45 <kashyap> gibi: Yeah, that too
16:05:12 <kashyap> It's fairly mechanical; and the rationale is: the QEMU folks suggest this as a more effecive alternative to 'qxl' (one of the other display devs) in the long term
16:06:07 <gibi> so it is a new value for an existing flavor extra_spec and image meta?
16:06:46 <kashyap> gibi: Yep; mostly for image meta
16:07:04 <bauzas> uupgrade impact, maybe ?
16:07:21 <kashyap> gibi: Here's the WIP patch: https://review.opendev.org/c/openstack/nova/+/794796
16:07:35 <bauzas> what happens for existing instances ?
16:07:57 <gibi> bauzas: I think we still support all the existing display devices
16:08:06 <kashyap> bauzas: What gibi said
16:08:12 <stephenfin> o/
16:08:13 <gibi> bauzas: so old instances use whathever they was booted with
16:08:14 <bauzas> ok, so we are not changing the default value ?
16:08:14 <kashyap> We're not deprecating any existing devices upstream
16:08:19 <kashyap> bauzas: No, we're not :)
16:08:23 <bauzas> ack
16:08:39 <kashyap> bauzas: Simply making another option that is safer and simpler possible.
16:08:39 <gibi> as of upgrade, kashyap do you need any changes on the compute service?
16:08:49 <bauzas> so it's just a new option then
16:09:07 <kashyap> gibi: No; I don't see any for that. It's simply a mechanical yet-another-possible-option
16:09:07 <bauzas> and then just a documentation question
16:09:10 <gibi> as far as I see the WIP patch does not change the compute manager or the virt driver
16:09:14 <kashyap> (Yes, docs; of course)
16:09:20 <gibi> kashyap: cool
16:09:23 <kashyap> gibi: Yep; you notice it well :)
16:09:37 <bauzas> okay, so it looks to me a non-spec BP
16:09:37 <gibi> so the old computes will support the new option out of the box
16:09:39 <gibi> that is cool
16:09:55 <bauzas> yup
16:10:03 <gibi> any objection to to approve this as a specless bp?
16:10:34 <kashyap> (We satisfy thea min_required_libvirt version too -- the current in-tree version is above what's required.  So we're good there too.)
16:10:39 * dansmith stumbles in late
16:11:09 * gibi lets stephenfin and dansmith to read back and possibly object against the bp approval
16:11:27 <stephenfin> no issues from me - I've already started reviewing the patches :)
16:11:32 <dansmith> for the bochs display?
16:11:35 <gibi> dansmith: yes
16:11:38 <gibi> stephenfin: ack
16:11:43 <dansmith> does it not require a new console proxy?
16:11:47 <dansmith> or is this just guest-facing?
16:12:14 <dansmith> ah, it's a guest device, I see
16:12:18 <dansmith> yeah, no objection
16:12:42 <gibi> cool
16:12:50 <gibi> then I will approve the bp after the meeting
16:13:01 <gibi> kashyap: is that all about this topic?
16:13:22 <kashyap> gibi: That's all on it
16:13:25 <gibi> kashyap: thanks
16:13:31 <gibi> then lets go back to the normal agenda
16:13:40 <gibi> #topic Bugs (stuck/critical)
16:13:48 <gibi> no critical bugs
16:13:48 <kashyap> dansmith: Yes, you parsed it right
16:14:10 <gibi> 10 new untriage bugs (+1 since the last meeting)
16:14:27 <gibi> I guess we have limited inflow as summer hits
16:14:39 <gibi> any specific bugs we need to talk about today?
16:16:02 <gibi> #topic Gate status
16:16:08 <gibi> placement periodic jobs are green
16:16:24 <gibi> as I was on PTO in the last two days I'm not up to date about the gate status
16:16:33 <gibi> is there any gate issue?
16:17:01 <dansmith> there was a stable one,
16:17:04 <dansmith> I think melwitt is on it tho
16:17:13 <bauzas> yup
16:18:05 <gibi> is it the one where we install too new Ceph version on stable?
16:18:23 <elodilles> yes (and i linked it in the stable summary :))
16:18:26 <gibi> #link http://lists.openstack.org/pipermail/openstack-discuss/2021-June/022900.html
16:18:30 <gibi> cool
16:18:43 <gibi> anything else about the gate?
16:20:11 <gibi> #topic Release Planning
16:20:15 <gibi> Milestone 2 is 15 of July
16:20:33 <gibi> we will freeze our Xena specs at that point
16:20:53 <gibi> I will propose a spec review day a week before that
16:21:19 <bauzas> lgtm
16:21:42 <gibi> anything else about the release or other deadlines?
16:22:35 <gibi> #topic Stable Branches
16:22:40 <gibi> victoria, ussuri gate is broken due to nova-grenade-multinode job failure ( fix: https://review.opendev.org/q/I02b2b851a74f24816d2f782a66d94de81ee527b0 )
16:22:44 <gibi> wallaby, train..pike should be OK
16:22:51 <gibi> EOM from elodilles
16:23:03 <gibi> anything else on stable?
16:23:13 <elodilles> nothing else from me
16:23:25 <gibi> elodilles: thanks for the summary
16:23:37 <elodilles> np
16:24:10 <gibi> #topic Sub/related team Highlights
16:24:13 <gibi> Libvirt (bauzas)
16:24:17 <bauzas> nothing to report, EOM from bauzas :-)
16:24:42 <gibi> :)
16:24:47 <gibi> #topic Open discussion
16:25:12 <gibi> we agree to re-discuss the topic of moving this meeting to the the #openstack-nova channel
16:25:32 <dansmith> yes please
16:25:46 <gmann> yeah
16:25:50 <sean-k-mooney> im fine with trying it and i would not have been as late if it was there
16:26:26 <dansmith> note no traffic in -nova since the meeting started :)
16:26:40 <gmann> yeah I was typing the same :)
16:26:50 <gibi> I'm OK to give it a try
16:27:00 <bauzas> well, sometimes we discuss in the -nova channel while we have a meeting
16:27:07 <gibi> we did the Asia friendly office hours last Thursday on nova
16:27:10 <gibi> and it worked
16:27:12 <gmann> fyi, neutron also moving to neutron channel
16:27:24 <bauzas> but we could ask people to discuss in -dev if they want
16:27:43 <gibi> bauzas: true about side discussions, but I think it is even beneficial to do that _after_ the meeting
16:27:44 <dansmith> or wait a few minutes
16:27:52 <gmann> bauzas: or we can ask them to wait to finish the meeting or so
16:27:55 <gibi> :)
16:28:04 <gmann> we do same in QA.
16:28:06 <dansmith> especially if they don't realize there is a meeting because it's in another channel,
16:28:13 <dansmith> and then leave before the rest of the team is finished
16:28:38 <sean-k-mooney> ya if a side discusion betwen people in the meeting we can also just wait
16:28:38 <bauzas> gmann: well, sure, unless folks have a time around 7pm after the meeting ;)
16:29:08 <bauzas> but, meh
16:29:24 <gibi> OK then lets try it. See if it works.
16:29:31 <gibi> if not then we can switch back
16:30:16 <gmann> +1
16:30:19 <gibi> please feel free to approach me with feedback publicly or privately in the future about this
16:30:42 <gibi> I will update the wiki and the eavesdrop page with the information after this meeting
16:31:04 <gibi> and I will drop a mail to the ML
16:31:17 <gibi> any other topic for today?
16:31:22 <tosky> totally unrelated: would anyone complain if I backported https://review.opendev.org/c/openstack/nova/+/741603/ (available since victoria) to ussuri and earlier release? Just to remove a few not used legacy jobs
16:32:01 <gibi> tosky: no objection from me
16:32:15 <sean-k-mooney> from experimenatl no concern form me
16:32:22 <tosky> thanks! I will propose the patches
16:32:24 <gmann> yeah, ok for me too
16:32:29 <gibi> tosky: thanks
16:32:32 <elodilles> neither from me, since it's experimental
16:32:32 <sean-k-mooney> whould we  do a review of what in that in general
16:33:12 <gibi> sean-k-mooney: we can do that but honestly if nobody tried to fix them so far then I guess we dont need them
16:33:50 <sean-k-mooney> tempest-pg-full is that postgress
16:33:53 <elodilles> (this ran last in 2020: https://zuul.opendev.org/t/openstack/builds?job_name=legacy-grenade-dsvm-neutron-multinode-zero-downtime )
16:35:11 <gibi> sean-k-mooney: the patch only removes legacy-grenade-dsvm-neutron-multinode-zero-downtime and legacy-tempest-dsvm-nova-libvirt-kvm-apr
16:35:20 <sean-k-mooney> oh i know
16:35:40 <sean-k-mooney> i was just lookign to see if there was anything on master we shoudl remove
16:35:41 <gibi> sean-k-mooney: oh you mean review what remains
16:35:47 <sean-k-mooney> or graduate
16:35:57 <sean-k-mooney> ya
16:36:16 <gibi> sean-k-mooney: yeah I guess that make sense
16:36:18 <sean-k-mooney> i tend to avoid running experimatal beacuse of how much stuff is in it
16:36:38 <sean-k-mooney> normally i dont need to but its proably good to prune it form time to time
16:36:40 <gmann> this we can remvoe for sure as devstack does not support opensuse anymore 'tempest-full-py3-opensuse15'
16:37:03 <gmann> but yes, cleanup can be done for more jobs
16:37:29 <gibi> sean-k-mooney, gmann: cool. if you find others to be removed then feel free to add me to the review
16:37:38 <gmann> sure
16:37:46 <sean-k-mooney> well likely a new review right
16:37:53 <gibi> sean-k-mooney: yes please
16:38:18 <gibi> any other topic for today?
16:40:11 <gibi> then thanks for joining today.
16:40:17 <gibi> o/
16:40:22 <gibi> #endmeeting