21:00:03 #startmeeting nova 21:00:04 Meeting started Thu Feb 13 21:00:03 2020 UTC and is due to finish in 60 minutes. The chair is efried. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:00:08 The meeting name has been set to 'nova' 21:00:17 #link agenda https://wiki.openstack.org/wiki/Meetings/Nova#Agenda_for_next_meeting 21:01:01 o/ 21:01:09 o/ 21:01:09 #topic Last meeting 21:01:10 #link Minutes from last meeting: http://eavesdrop.openstack.org/meetings/nova/2020/nova.2020-02-06-14.00.html 21:01:10 * efried sean-k-mooney to do the train implemented specs business 21:01:10 Done: https://review.opendev.org/#/c/706276/ (needs review please) 21:01:10 * efried efried to ML about spec scrub day 21:01:10 Done: http://lists.openstack.org/pipermail/openstack-discuss/2020-February/012424.html 21:01:10 * efried lyarwood to curate rocky EM list 21:01:22 o/ 21:01:25 Not sure about that last one, and I imagine lyarwood is afk 21:01:32 will fup next week I guess. 21:02:12 #topic Bugs (stuck/critical) 21:02:12 No Critical bugs 21:02:15 However, 21:02:24 #link 96 new untriaged bugs (+6 since the last meeting): https://bugs.launchpad.net/nova/+bugs?search=Search&field.status=New 21:02:24 #link 21 untagged untriaged bugs (+6 since the last meeting): https://bugs.launchpad.net/nova/+bugs?field.tag=-*&field.status%3Alist=NEW 21:02:29 these numbers are rising steadily 21:02:36 #link bug triage how-to: https://wiki.openstack.org/wiki/Nova/BugTriage#Tags 21:02:37 #help need help with bug triage 21:03:00 I think mriedem said if we get to 100 untriaged bugs, a huge marshmallow man comes and wrecks shop. 21:03:58 #topic Reminders 21:03:58 SPEC FREEZE IS NOW 21:03:58 #link ussuri blueprints https://blueprints.launchpad.net/nova/ussuri 21:04:22 So we're at 30 open blueprints, of which six are still not Design:approved. 21:04:46 NUMA topology with RPs 21:04:46 #link blueprint https://blueprints.launchpad.net/nova/+spec/numa-topology-with-rps 21:04:46 #link spec Proposes NUMA topology with RPs https://review.opendev.org/552924 21:05:15 We've been back and forth on this one. I think it needs one more rev, as noted. Hopefully we can approve it in the morning. 21:05:28 nothing to be done now, since the main players are euro. 21:05:38 Unified limits 21:05:38 #link blueprint unified-limits-nova https://blueprints.launchpad.net/nova/+spec/unified-limits-nova 21:05:38 #link spec Add Unified Limits Spec https://review.opendev.org/#/c/602201/ 21:06:04 melwitt is +2. Anybody qualified to +A that? (I'm not.) dansmith? 21:06:18 I think we were hoping for alex_xu for the other approver 21:06:22 not me 21:06:49 but bauzas is also reviewing it 21:06:50 okay. 21:07:15 and stephenfin 21:07:54 k, next: 21:08:06 Support volume local cache 21:08:06 #link blueprint support-volume-local-cache https://blueprints.launchpad.net/nova/+spec/support-volume-local-cache 21:08:06 #link spec Support volume local cache https://review.opendev.org/#/c/689070/ 21:08:18 the cinder side hasn't been approved yet 21:08:27 we had a cross project meeting this morning 21:08:39 the nova people are not dead-set against it 21:08:41 unfortunately, that was at 5:30am my time, so didn't get to join 21:08:55 yeah, primary person on that is in shanghai 21:09:12 so, Liang is going to revise both cinder and nova specs 21:09:34 tonight? 21:09:37 he's usually pretty quick, and not a lot of new stuff came up in the meeting 21:09:49 i don't know if he's that quick 21:10:04 i think we will need to ask for a spec-freeze exception 21:10:14 (assuming nova does such things) 21:10:50 i think most of the cinder-side issues are under control 21:11:21 Okay. I haven't like announced how that works, and dunno if we have a doc for it, but a [nova][sfe] email to the ML seems appropriate for requesting that exception. 21:11:38 esp because the people who should agree to same are spread across time zones. 21:12:00 typically we have allowed 1 week to ask for an excetion 21:12:22 seems like the main cores on the nova side have been gibi and dansmith, with a nod by alex_xu. Are those the right people to collar IYO rosmaita? 21:12:23 and i think then wanted 3 cores to +2 or have at least 2 sign up to review it 21:12:29 sean-k-mooney: what's your assessment of the reasonableness of this getting done in U? 21:12:50 I'm definitely not going to approve it so you can just leave me out of it 21:13:07 yes, you made that clear on the spec review :) 21:13:11 i need to review it but i felt like it would be a streach to get it done 21:13:19 if the scope is limited maybe 21:13:32 move operations seam unlikely 21:13:42 right 21:13:42 well live migration 21:13:44 okay, let's discuss the feasibility via the email thread. 21:14:03 for now, we'll assume it will need an exception, unless it miraculously gets approved tomorrow. 21:14:09 ok, i'll work with Liang to get something the ML early next week 21:14:11 thanks! 21:14:13 thanks for joining rosmaita. 21:14:22 next: 21:14:23 Support re-configure deleted_on_termination in server 21:14:23 #link blueprint destroy-instance-with-datavolume https://blueprints.launchpad.net/nova/+spec/destroy-instance-with-datavolume 21:14:23 #link spec Support re-configure deleted_on_termination in server https://review.opendev.org/#/c/580336/ 21:14:52 I can't tell whether brinzhang is still working this, but I'll assume if it hasn't had action by tomorrow it's also deferred. 21:15:33 anyone have insight? 21:15:38 If not, I'll move on... 21:15:52 noVNC requires password authentication 21:15:52 #link blueprint nova-support-webvnc-with-password-anthentication https://blueprints.launchpad.net/nova/+spec/nova-support-webvnc-with-password-anthentication 21:15:52 #link spec Proposal for a safer noVNC console with password authentication https://review.opendev.org/#/c/623120/20 21:16:16 ditto 21:16:24 anyone have insight? 21:16:35 Last one: 21:16:36 nova-audit 21:16:36 #link blueprint nova-audit https://blueprints.launchpad.net/nova/+spec/nova-audit 21:16:36 #link Add nova-audit spec https://review.opendev.org/#/c/693226/ 21:16:48 I have a little 21:16:59 on the vnc one? 21:17:03 the use case around the vnc password feature isn't described in much detail on the spec, which raised a lot of questions 21:17:20 i think the vnc password spec has merrit 21:17:30 no one's opposed to the general idea though and there is some valid use case, but the approach needs to be adjusted before we'd want to approve it 21:17:33 the usecase is mirroring the behavior of a vnc server to a client 21:17:55 Okay, great that it has merit and seems doable, but it's going to need someone to own it and drive it. 21:18:01 where you have to enter a vnc password before the terminal of the guest is exposed 21:18:16 alex_xu had a suggestion for the design that I prefer and the spec has not yet been updated to reflect that. the author said they would update it though if I understood their last reply correctly 21:18:22 I left a comment that it needs to be revised asap. Otherwise I'll defer. 21:18:25 okay. 21:18:35 ya makes sense 21:18:57 back to nova-audit: 21:18:57 melwitt has taken ownership, but needs to get support from cores. 21:18:57 melwitt: comments? 21:19:25 I sent email to some cores to see if there's any interest in their review of it, we'll see if anyone reviews it by EOD tomorrow 21:19:37 wfm 21:19:49 any specs or blueprints I missed? 21:20:18 i missed the start of the meeting did you cover the mixed cpu and numa specs 21:20:38 if so ill just read scollback 21:20:41 mixed CPU was approved a couple hours ago. NUMA I mentioned, yes. 21:21:08 ok cool 21:21:21 So after today's scrub (where I already deferred several that were/seemed dead) we're at 21:21:21 30 total. 24 are Design:Approved, of which 7 are Implemented 21:21:58 I've been making noise about trying to reduce the scope to a number we actually think we have a good chance of getting done in U. 21:22:20 I don't remember exactly, but I want to say the number 25 was flying around the ML at some point. 21:22:26 so, we're really not too far off. 21:22:52 I'm going to try to crystallize some kind of process/criteria/whatever and kick something off next week. 21:23:15 In the meantime, if you have thoughts on this, please brain dump in 21:23:15 #link etherpad for working the scrub https://etherpad.openstack.org/p/nova-ussuri-planning 21:23:46 Any questions or other subtopics under specs, blueprints, freeze, process thereon, etc? 21:24:12 okay, moving on. 21:24:13 #topic PTG/Summit planning 21:24:13 Please mark **attendance** and topics on 21:24:13 #link PTG etherpad https://etherpad.openstack.org/p/nova-victoria-ptg 21:24:27 ok, am to highlight two that you were involved with, with the changes happening at intel will you/intel be able to complete the vtpm and provider yaml spects. 21:25:09 e.g. may they need some help to get across the line. i can follow up after the meeting 21:25:10 ah, okay. 21:25:35 yeah, those two were Definition:Approved a while back, so they haven't been on the spec freeze radar just yet. 21:25:48 jroll has taken ownership of vTPM. 21:26:10 he will have to advocate for the Direction:Approval of that blueprint I suppose. 21:26:30 ok 21:26:47 Provider yaml is really really close to code-complete, so I think it makes sense to make it go. 21:26:56 ya 21:27:24 brb 21:27:29 I think we're waiting for dustinc to do another rev (I imagine he's primarily looking for a job rn) but if he doesn't, I'm sure we can talk somebody into doing that. 21:27:35 I would, but I want to retain my +2 power 21:27:45 but the changes needed are pretty minimal IIRC. 21:28:10 Moving on: 21:28:10 #topic Sub/related team Highlights 21:28:10 Placement (tetsuro) 21:28:56 There was noise about resurrecting can_split for the NUMA topo bp, but I think we managed to kill that. 21:29:18 There's some kind of placement API change needed for the shared disk thing, but that's not apparently going to happen in U. 21:29:25 So that leaves consumer types. melwitt, update there? 21:29:35 I'm still chasing a bug in the consumer types patches. I found what's wrong and now trying to fix (sql query stuff that I'm not good at) 21:30:01 once I fix it, they'll be ready for review again 21:30:28 melwitt: remind me, is that a dep of something in nova? 21:30:54 it is, it's something we'd need to clean up some quirks in counting quota usage from placement 21:31:07 cleanup, not a nova bp? 21:31:09 things like being able to take a "delta" of usage in the middle of a resize, etc 21:31:30 I didn't make a nova bp about it yet because it's not a thing until consumer types are available 21:31:37 its a dep for unifeid limits? 21:31:48 no, it's not a dep for unified limits 21:31:57 ok 21:32:08 it's a dep for doing things like, stop counting double flavor for a resize that has not been confirmed or reverted yet 21:32:44 sorry, I guess I'm asking: will whatever needs to get done in nova, that depends on consumer types, require a blueprint (and therefore be Victoria earliest) or not (and therefore conceivably U)? 21:32:45 ah ok 21:32:49 today we count both the old and new flavor. with consumer types, we'd know how to take the ceiling of the two flavors mid resize 21:33:08 efried: yeah, definitely won't be anything in U for nova. V at the earliest 21:33:15 okay, cool. 21:33:18 I'm just trying to get the placement side done for U 21:33:25 yeah, good plan :) 21:33:28 API (gmann) 21:33:28 This week update: http://lists.openstack.org/pipermail/openstack-discuss/2020-February/012563.html 21:33:40 As we can see in email. 4 BPs are ready/in-progress for code review. 21:33:45 Other are still on spec review side and few of them are already discussed. 21:33:48 most of those we discussed earlier. 21:33:50 yes 21:33:50 On Policy work: I did few more API policy changes and up for review and continuing the rest of the API. 21:34:03 While dong policy BP work, I am finding few bugs in policy side. 5 till now which are mainly on admin-or-owner things which is good. 21:34:11 that's all from my side. 21:34:14 thanks gmann 21:34:17 #topic Stuck Reviews 21:34:17 any? 21:34:33 #topic Open discussion 21:34:33 [efried] PTL transfer 21:35:14 It's not for sure yet, but it's looking increasingly likely that I will no longer be able to continue as PTL after March 31 at the outside. 21:35:43 So it's time to start thinking about whether YOU would be willing to take the PTL role for the remainder of the U release. 21:36:36 Does anyone want to put up their hand at this time? 21:36:39 Not a commitment, just for my information. 21:36:47 FYI next PTL election might be during March(dates are not out yet) 21:36:59 or starting april some time 21:37:03 Hm, that would be an interesting wrinkle. 21:37:13 okay, good to know. 21:37:24 So, anyone? 21:37:27 Bueller? 21:37:54 okay. 21:38:13 FYI there's not a real solid process around this, but if I can't find anyone, the TC "gets involved"... 21:38:27 Any other open topics before we move on? 21:38:34 one thing to update on stable gate: stable/rocky|queens should be passing now as stackviz issue is resolved and latest tarball is available for jobs to use. we ca recheck failing one. 21:38:43 Thanks gmann 21:39:11 Okay, thanks all. 21:39:11 o/ 21:39:11 #endmeeting