16:00:03 <gibi> #startmeeting nova 16:00:03 <openstack> Meeting started Thu Jun 25 16:00:03 2020 UTC and is due to finish in 60 minutes. The chair is gibi. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:04 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:06 <openstack> The meeting name has been set to 'nova' 16:00:39 <gibi> o/ 16:00:41 <dansmith> o/ 16:00:56 <gmann> o/ 16:01:23 <gibi> #topic Bugs (stuck/critical) 16:01:30 <stephenfin> o/ 16:01:32 <gibi> no critical bug 16:01:40 <gibi> #link 27 new untriaged bugs (+3 since the last meeting): https://bugs.launchpad.net/nova/+bugs?search=Search&field.status=New 16:01:52 <gibi> I need help with triaging 16:01:58 <bauzas> \o 16:02:05 <gmann> we have one bug for focal distro 16:02:22 * bauzas is tempted to say "I can" but I'll be on and off next week due to internal stuff 16:02:25 <gmann> #link https://bugs.launchpad.net/nova/+bug/1882521 16:02:25 <openstack> Launchpad bug 1882521 in OpenStack Compute (nova) "Failing device detachments on Focal" [Undecided,New] 16:02:49 <gibi> gmann: do you need some help with that bug? 16:02:56 <gmann> this is dependency of migrating the CI/CD to focal 16:03:12 <gmann> i checked the cinder logs and all flow but could not spot the issue 16:03:26 <gmann> some more eyes will be helpful 16:03:49 <gmann> issue happening while test cleanup try to detach the volume and failing 16:05:03 <gibi> did you tried to involve cinde folks into that investigation? 16:05:18 <gibi> I can try to look at it more closely tomorrow / next week 16:05:37 <gmann> not yet, cinder seems finishing the detachment but i can ask their help 16:05:44 <gmann> thanks 16:06:27 <gibi> any other bug that needs attention? 16:07:15 <gibi> #topic Runways 16:07:20 <gibi> Etherpad is up for Victoria #link https://etherpad.opendev.org/p/nova-runways-victoria 16:07:25 <gibi> Both bp/rbd-glance-multistore and bp/use-pcpu-and-vcpu-in-one-instance were flushed from the slots as the patches has been approved (and the slots are expired) 16:07:29 <akhil-g> yeah just have a look at this bug https://bugs.launchpad.net/nova/+bug/1884068 16:07:29 <openstack> Launchpad bug 1884068 in OpenStack Compute (nova) "Instance stuck in build state when some/one compute node is unreachable" [Undecided,New] 16:08:22 <gibi> akhil-g: looking at the logs you posted I don't see any logs from your build request 16:09:43 <akhil-g> yeah but I find only those logs from my setup 16:10:04 <gibi> akhil-g: I will circe back to that bug tomorrow but I suspect some connectivity issue between the nova services. 16:11:01 <gibi> going back to the runways 16:11:10 <gibi> so I flushed the slots 16:11:20 <gibi> I think we made good progress on the patches in the slots 16:11:28 <gibi> There is one item in the queue #link https://review.opendev.org/#/q/topic:bp/add-emulated-virtual-tpm 16:11:46 <gibi> stephenfin: do you have time in the next two weeks to iterate on feedback about it 16:11:48 <gibi> ? 16:11:53 <stephenfin> I do 16:11:55 <gibi> coo 16:11:56 <gibi> l 16:12:17 <gibi> is there at least one core (besides me) who willing to look at that series in the next two weeks? 16:12:59 <stephenfin> I was hoping bauzas would be able to chip in, given the area of the code it's touching 16:13:12 <stephenfin> and lyarwood, who should be back next week 16:13:26 <gibi> lyarwood is back next week? awesome 16:13:36 <stephenfin> iirc, yup! 16:14:06 <gibi> OK I will rely on bauzas and lyarwood then 16:14:26 <bauzas> okay, I'll volunteer 16:14:30 <gibi> #action gibi to put bp/add-emulated-virtual-tpm to a runway slot 16:15:03 <gibi> If anybody has any ready patch series / feature then do not hesitate to put it into the queue, we still have two free slots 16:16:07 <gibi> #topic Release Planning 16:16:11 <sean-k-mooney> gibi: maybe the patches form aarents 16:16:26 <gibi> aarents: when you read this, please add them to the queue ^^ 16:16:38 <sean-k-mooney> but that can be done async of the meeting 16:17:30 <gibi> OK, going back to the release planning 16:17:32 <gibi> 5 weeks until Victoria M2 16:17:38 <gibi> The Victoria community-wide goals are finalized: #link http://lists.openstack.org/pipermail/openstack-discuss/2020-June/015459.html 16:17:45 <gibi> We need volunteers for: 16:17:49 <gibi> Switch legacy Zuul jobs to native 16:17:54 <gibi> Migrate CI/CD jobs to new Ubuntu LTS Focal 16:18:42 <gmann> I can help onfocal migration which will be more of testing the things 16:18:53 <gibi> gmann: thans 16:18:55 <gibi> thanks 16:19:03 <gmann> and after that devstack and tempest jobs will be moved to focal which automatically switch nova zuulv3 jobs to focal 16:19:30 <gmann> but as per goal agreement, legacy jobs will need to migrate to zuulv3. legacy jobs will not be moved to focal 16:19:39 <sean-k-mooney> gmann: what legacy zuul jobs do we still have at this point 16:19:49 <gmann> i can try on zuulv3 migration but can't commit fully. 16:20:03 <gmann> sean-k-mooney: nova-live-migration and nova-greande job i thin 16:20:05 <gmann> thinks 16:20:24 <sean-k-mooney> so greande should be trivial now that there is a greade native job 16:20:32 <gmann> grenade will be easy now as grenade zuulv3 base job is backported till train 16:20:38 <sean-k-mooney> i think lyarwood had patches for the live migartion job 16:20:39 <gmann> yeah 16:21:19 <gmann> I will check that patch state 16:21:57 <sean-k-mooney> so grenade should be v3 https://review.opendev.org/#/c/704364/ 16:22:19 <gibi> lyarwood's patch #link https://review.opendev.org/#/c/711604/ 16:22:24 <gmann> that is different, i mean nova-grenade-multinode job 16:22:33 <gmann> gibi: yeah this one 16:22:57 <gmann> i think all these can go in one shot as all legacy multinode jobs derived from nova-legacy-base job 16:23:21 <sean-k-mooney> yes proably 16:23:30 <gibi> OK. I see some movement around these topic, that is good. Let me know if you need my help 16:23:44 <gibi> any other release related topic? 16:24:24 <sean-k-mooney> do we need to do release for libs for m1 16:24:35 <gibi> sean-k-mooney: those were proposed automatically 16:24:43 <sean-k-mooney> gibi: ok 16:24:56 <gibi> 16:14:25 <gibi> #link https://review.opendev.org/#/c/735716/ 16:24:57 <gibi> 16:14:31 <gibi> #link https://review.opendev.org/#/c/735690/ 16:25:19 <gibi> both merged 16:25:53 <gibi> OK moving forward 16:25:53 <sean-k-mooney> yep i didnt see the os-vif one but there was also no pending critial patch to hold it so its fine 16:26:10 <gibi> sean-k-mooney: I checked and seemed good 16:26:21 <gibi> I will ping you next time 16:26:32 <sean-k-mooney> its fine i trust you :) 16:26:39 <gibi> :) 16:26:42 <gibi> #topic Stable Branches 16:26:59 <gibi> any stable news? 16:27:05 <gmann> I have some updated on stable gate status 16:27:11 <gibi> sure, go 16:27:17 <gmann> stable/train gate is up. 'virtualenv not found' and neutron-greande-multinode issue is fixed in legacy base jobs and with grenade backport. feel free to recheck. 16:27:42 <gmann> stable/stein gate need this to merge to be green - https://review.opendev.org/#/c/737332/ 16:27:54 <elod> i've +W'd the stein patch 16:28:00 <gmann> and stable/rocky and less are all blocked till devstack fix - #link https://review.opendev.org/#/c/735615/ 16:28:04 <gmann> elod: thanks, 16:28:04 <elod> so it's will be ok, too 16:28:48 <gibi> thank you all working on unblocking the gate on stable 16:28:49 <gmann> one strange thing hapepning on pep8 cherrypick checks - https://review.opendev.org/#/c/728057/ 16:29:16 <elod> gmann: that's not strange :) 16:29:28 <elod> I guess that's the new check 16:29:31 <gmann> though all previous backport are merged but script still fail. melwitt and I were discussing it yesterday and it ran locally fine but on gate it is not 16:29:42 <elod> and patch was not merged when it ran 16:29:52 <gmann> elod: yeah new checks is not working fine in gate, in the patch i mentioned should pass the check but faiking 16:29:54 <gmann> failing 16:30:17 <elod> ok, thanks for the notice 16:30:33 <gibi> I hope dansmith can look at it as he was the author of that check 16:30:36 <elod> I'll try to figure out what's the problem there 16:30:41 <gmann> i am suspecting rebase needs as check patch is merged after stable/train was proposed to zuul merging those at pep8 job runtime is causing some strange behaviour 16:30:55 <sean-k-mooney> gmann: i need to update the commit mesage 16:30:59 <gmann> sean-k-mooney: can you rebase this and we will see if that run fine 16:31:03 <sean-k-mooney> and fix the hashes i think 16:31:12 <sean-k-mooney> yep i can 16:31:14 <gmann> sean-k-mooney: no, hash are all fine:) 16:31:31 <gmann> i will rebase can give us more clarity 16:31:48 <sean-k-mooney> just did it in the ui 16:31:56 <gmann> cool, thanks. 16:32:06 <gibi> anything else for stable? 16:32:06 <gmann> that's all from my side. 16:32:58 * bauzas has to leave \o 16:33:01 <dansmith> that failure isn't finding the commit hashes in the commit message 16:33:03 <gibi> bauzas: thanks 16:33:17 <dansmith> it's not complaining that they're wrong, it's not seeing them for some reason 16:33:36 <gmann> dansmith: yeah, and if i ran locally it find correctly. 16:33:54 <gmann> let's see after rebase 16:33:58 <dansmith> okay 16:34:03 <gibi> moving forward 16:34:05 <gibi> #topic Sub/related team Highlights 16:34:11 <gibi> API (gmann) 16:34:32 <gmann> one thing to share as progress is unified limit 16:34:34 <gmann> #link https://review.opendev.org/#/q/topic:bp/unified-limits-nova+(status:open+OR+status:merged) 16:35:06 <gmann> review is going good as gibi stephenfin melwitt and i also reviewed yesterday 16:35:06 <gibi> gmann: is it reasy for a runway slot? 16:35:14 <gibi> ready 16:35:27 <gmann> i am waiting for johnthetubaguy if he can update the series 16:35:31 <stephenfin> I think we need johnthetubaguy to be around 16:35:35 <gmann> yeah 16:35:47 <gmann> once he update then we can add in runway 16:35:59 <gibi> OK, cool 16:36:27 <gibi> Libvirt (bauzas) 16:36:44 <gibi> he did not mentioned any news to me on #openstack-nova 16:37:25 <gibi> #topic Stuck Reviews 16:37:32 <gibi> nothing on the agenda 16:37:37 <gibi> anything to bring up here? 16:38:01 <gibi> #topic Open discussion 16:38:12 <gibi> vinay_m: your turn 16:38:55 <vinay_m> https://bugs.launchpad.net/nova/+bug/1759839 16:38:55 <openstack> Launchpad bug 1759839 in OpenStack Compute (nova) "Documentation Error for Placement API Port Number" [Low,Triaged] - Assigned to vinay harsha mitta (vinay7) 16:39:51 <gibi> vinay_m: do you need help with that bug? 16:40:16 <stephenfin> oh, I looked at this 16:40:21 <stephenfin> context is (I suspect) https://review.opendev.org/731141 16:40:35 <vinay_m> https://bugs.launchpad.net/nova/+bug/1741329 .. this bug specifies quite opp to above mention bug 16:40:35 <openstack> Launchpad bug 1741329 in OpenStack Compute (nova) "Install and configure controller node for openSUSE and SUSE Linux Enterprise in nova" [Medium,Fix released] - Assigned to Andreas Jaeger (jaegerandi) 16:41:18 <stephenfin> so it seems the SUSE packages are using port 8780 for placement, while everything else uses 8778 16:41:34 <vinay_m> yes 16:41:34 <sean-k-mooney> what is the default in the code 16:41:57 <stephenfin> No idea. Not sure if there is one 16:42:42 <stephenfin> TripleO deploys to 8778 as does DevStack 16:43:00 <gibi> if our suse doc aligned with the suse package then I guess this is simply an invalid bug. Or am I missing something? 16:43:34 <stephenfin> Yes, https://bugs.launchpad.net/nova/+bug/1759839 looks invalid 16:43:34 <openstack> Launchpad bug 1759839 in OpenStack Compute (nova) "Documentation Error for Placement API Port Number" [Low,Triaged] - Assigned to vinay harsha mitta (vinay7) 16:43:48 <stephenfin> and https://bugs.launchpad.net/nova/+bug/1741329 *is* valid 16:43:48 <openstack> Launchpad bug 1741329 in OpenStack Compute (nova) "Install and configure controller node for openSUSE and SUSE Linux Enterprise in nova" [Medium,Fix released] - Assigned to Andreas Jaeger (jaegerandi) 16:43:52 <stephenfin> IMO 16:44:20 <gibi> stephenfin: I agre 16:44:21 <gibi> e 16:44:24 <sean-k-mooney> so placement was only ever deployed via wsgi 16:44:44 <sean-k-mooney> so unlike the nova service it has nver had a python wrapper that bound to a defualt port 16:44:55 <sean-k-mooney> so ya i tend to agree with stephenfin 16:45:20 <gibi> vinay_m: is it OK for you if I mark https://bugs.launchpad.net/nova/+bug/1759839 invalid? 16:45:20 <openstack> Launchpad bug 1759839 in OpenStack Compute (nova) "Documentation Error for Placement API Port Number" [Low,Triaged] - Assigned to vinay harsha mitta (vinay7) 16:46:47 <vinay_m> ok gibi that is what doubt was whether that bug was invalid or not 16:47:01 <gibi> OK cool 16:47:20 <gibi> I will mark it after the meeting 16:47:31 <gibi> I have one more topic 16:47:33 <gibi> (gibi): VMware 3pp CI status: Based on #link https://review.opendev.org/#/c/734114/ the 3pp CI is green now. I will continue monitoring the situation through the V cycle. 16:48:02 <gibi> as far as I see there is real effort now making the CI green 16:48:53 <gibi> so I think if the greenness remain during the cycle, then I will propose an undeprecation patch 16:48:57 <sean-k-mooney> fyi the placemnt docs say it can be on a different port or now port https://github.com/openstack/placement/blame/master/doc/source/install/shared/endpoints.rst#L56 16:49:25 <sean-k-mooney> gibi: ya that is nice to see 16:50:05 <sean-k-mooney> by the way http://ci-watch.tintri.com/ seams to be dead and i think the other one is too. 16:50:20 <sean-k-mooney> if i was to host a copy of it would that be useful to people 16:50:27 <gibi> this opens for me http://ciwatch.mmedvede.net/project?project=nova 16:50:46 <sean-k-mooney> oh that was dead the last time i hit it but if its working again cool 16:51:34 <akhil-g> https://bugs.launchpad.net/nova/+bug/1801714 This bug I think that openstack-nova is not looking for availability of the compute node so when it comes to scheduling and deployment there it gets stuck 16:51:34 <openstack> Launchpad bug 1801714 in OpenStack Compute (nova) "Rebuild instance is stuck in rebuilding state when hosting Compute is powered off" [Medium,Confirmed] - Assigned to Akhil Gudise (akhil-g) 16:52:29 <akhil-g> because host is not available so its stuck in rebuild state 16:52:54 <gibi> akhil-g: as I said above we would need more logs. Something is clearly missing 16:53:29 <akhil-g> yeah but this is another bug gibi 16:54:10 <gibi> akhil-g: ahh 16:54:19 <gibi> akhil-g: is this the same deployment? 16:55:00 <gibi> ahh I see I was even able to reproduce it 16:55:09 <akhil-g> yeah I've used same setups 16:56:38 <gibi> so you suggestion is to add a check in the compute-api and reject the rebuild if the compute service is down? 16:57:59 <sean-k-mooney> is rebuild a cast? 16:58:01 <akhil-g> not exactly but during scheduling we can eliminate unavailable hosts? 16:58:10 <sean-k-mooney> if it was an rpc call it would timeout right and fail? 16:58:27 <sean-k-mooney> akhil-g: which host is down 16:58:32 <sean-k-mooney> the one the vm is on? 16:58:39 <sean-k-mooney> akhil-g: rebuild is not a move action 16:58:53 <sean-k-mooney> so we only go to the schulder to validate the image is compatiable 16:58:59 <sean-k-mooney> we dont ever move the vm to another host 16:59:11 <gibi> we have one minute left 16:59:12 <sean-k-mooney> thats the main difference between evacuate and rebuild 16:59:45 <gibi> please continue discussing it on #openstack-nova 16:59:50 <akhil-g> ok sorry I mentioned it for the one which I've reported earlier 16:59:57 <sean-k-mooney> sure :) 17:00:04 <gibi> time is uo 17:00:05 <gibi> up 17:00:07 <gibi> #endmeeting