Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2022-08-29
10:55:36 sean-k-mooney i was the util function
10:55:42 sean-k-mooney you moved it to the object
10:56:41 gibi ahh yes
10:56:42 gibi true
10:56:55 sean-k-mooney https://review.opendev.org/c/openstack/nova/+/853835/9
10:57:09 sean-k-mooney i was just clenaing up the nova repo i had the DNM change checked out in
10:57:33 sean-k-mooney im going to start backporting the first vdpa patch
10:57:47 sean-k-mooney although i think i need to backport one of your patches first
10:58:09 sean-k-mooney the compute manger restart patch
10:59:20 gibi yeah on the libvirt tests fixture we overlap a bit
10:59:25 gibi hopefully not much
10:59:37 sean-k-mooney well im just using it for one test
10:59:41 sean-k-mooney the evacuate test
10:59:51 sean-k-mooney so i just need to backport it and then rebase on top
11:00:11 sean-k-mooney ah found it https://review.opendev.org/c/openstack/nova/+/850510
11:00:44 sean-k-mooney ok so im going to try backporting that and then ill cherry pick my patch
11:05:03 gibi now I'm done with the func test part of the rebuild bfv as well. I found nothing serious so I think we can land the series this week
11:05:21 gibi next up is the manila series...
11:05:47 gibi but I need to get food first
11:05:50 sean-k-mooney has the userdata patch been updated
11:06:24 sean-k-mooney if that is not updated today we should change the order or perhaps one of us can adress the nits
11:06:51 gibi sean-k-mooney: it was updated during the weekend
11:07:12 sean-k-mooney ah ok ill review that when im done with this then
11:07:13 gibi there is some failing unit tests but the overal logic now looks good to me
11:11:28 sean-k-mooney api sample tests or generic unit tests
11:20:43 sean-k-mooney oh ... this might also need some of the double mockign seriese
11:32:58 sean-k-mooney :)
11:33:09 stephenfin sean-k-mooney: gibi: continuing it here, I think they need a PTL to ack the patch against openstack/release
11:33:11 sean-k-mooney stephenfin: are there any others i should look at or jsut nova client
11:33:22 sean-k-mooney stephenfin: no they dont
11:33:36 stephenfin I'd ask elodilles directly but he's on PTO
11:34:00 sean-k-mooney it ptl or whoever is listed here https://github.com/openstack/releases/blob/master/data/release_liaisons.yaml#L93-L96
11:35:08 stephenfin ah
11:35:10 stephenfin sweet
11:35:17 stephenfin then I guess you can do it :)
11:35:38 sean-k-mooney im just looking at the changes between the current sha and the old one
11:38:01 sean-k-mooney stephenfin: we might end up doing another release if we land some of the inflight changes
11:38:11 sean-k-mooney but i have no issue with doing this release now
11:38:14 stephenfin that's okay, I think?
11:38:24 sean-k-mooney https://review.opendev.org/c/openstack/releases/+/854362
11:38:35 sean-k-mooney ptl-approved flag is no set
11:38:45 sean-k-mooney ya its ok
11:38:53 sean-k-mooney release are cheap
11:39:16 sean-k-mooney i assume this is needed for ocs change to merge or simialr
11:39:59 sean-k-mooney i know those micorversion also had osc patches
11:42:12 sean-k-mooney stephenfin: anyway let me know if you need release for anything else in the nova deliverables or ping bauzas tomorrow
11:42:23 sean-k-mooney if im not about
11:50:00 gibi I'm not sure we have rel mgmt up yet as elodilles_pto is on pto today
11:51:09 sean-k-mooney even if we do im not sure if we have two of them
11:51:22 sean-k-mooney given its still pto season
11:51:27 gibi yeah
11:53:09 sean-k-mooney hum https://review.opendev.org/c/openstack/nova/+/851445 helped but i sill have 74 failing tests
11:53:27 sean-k-mooney AttributeError: 'VGPUTests' object has no attribute 'libvirt'
11:53:38 sean-k-mooney so the libvirt fixutre is still not quite working
11:53:41 gibi hm, that is still somehow related to the mock stuf
11:54:26 gibi have you looked at https://review.opendev.org/c/openstack/nova/+/852464
11:54:48 sean-k-mooney nope bug that looks relevent
11:55:01 sean-k-mooney *but
11:55:09 sean-k-mooney ill pull that in and see if it resolves the issue
11:55:21 sean-k-mooney i wonder if i should squash the too for backport or not
11:55:30 gibi probalby OK to squash
11:55:58 sean-k-mooney ill try adding them seperatly and see if that resovles the issue then i can squash
11:56:07 gibi ack
11:59:16 sean-k-mooney im not actully backporting the mock removal by the way
11:59:24 sean-k-mooney just the double mocking patch
11:59:32 sean-k-mooney so stable will still use mock the lib
11:59:36 sean-k-mooney but without the double mocking
12:00:09 gibi hm, I think that is OK
12:03:17 sean-k-mooney looks like ill need https://opendev.org/openstack/nova/commit/cd03bbc1c33e33872594cf002f0e7011ab8ea047 too
12:04:39 sean-k-mooney that should be ok to backport too
12:05:14 sean-k-mooney thats needed for fake_get_mac_by_pci_address
12:05:39 sean-k-mooney that bug has existied since pf support was added so backporting to wallaby is fine
12:05:40 gibi ahh yeah, that was en extension to the fixture
12:05:59 sean-k-mooney ya i could just pull that in but if this does not cause much issue ill just backport it
12:06:01 gibi but for the MAC update to work we need https://review.opendev.org/c/openstack/neutron/+/829247 on the neutron side too
12:06:08 sean-k-mooney yes
12:06:15 gibi this is getting hairy :)
12:06:17 sean-k-mooney but having the mac info wont break anything
12:06:22 gibi true
12:06:29 sean-k-mooney so we can also backport the neutron patch but its not a blocker
12:06:30 gibi then fine
12:06:35 sean-k-mooney which is why you did not have a depens on
12:06:57 sean-k-mooney if this proves unworkable ill pull in just the test fixture changes
12:07:12 gibi ack
12:07:14 gibi make sense
12:58:32 sean-k-mooney gibi: ok i have a working backport series i need to add cherrypicked lines since i pull them in via interactive rebase to get the order right but ill push them and then do that
12:58:52 gibi cool
12:59:10 opendevreview sean mooney proposed openstack/nova stable/yoga: refactor: remove duplicated logic https://review.opendev.org/c/openstack/nova/+/855022
12:59:11 opendevreview sean mooney proposed openstack/nova stable/yoga: Record SRIOV PF MAC in the binding profile https://review.opendev.org/c/openstack/nova/+/855023
12:59:12 opendevreview sean mooney proposed openstack/nova stable/yoga: Remove double mocking https://review.opendev.org/c/openstack/nova/+/855024
12:59:13 opendevreview sean mooney proposed openstack/nova stable/yoga: Remove double mocking... again https://review.opendev.org/c/openstack/nova/+/855025
12:59:14 opendevreview sean mooney proposed openstack/nova stable/yoga: Add compute restart capability for libvirt func tests https://review.opendev.org/c/openstack/nova/+/855026
12:59:39 sean-k-mooney i was tempted to pull in the vdpa change to stop using sriov
13:00:00 sean-k-mooney but im not sure if that was backportable
13:00:05 sean-k-mooney so i didnt add it
13:08:35 gibi OK I think I finished looking at the latest update of the manila series. what was fixed looks correct to me, but there is still open comment. I know Uggla working on them as we speak
13:10:13 gibi I'm going to switch back to the PCI series adding unit tests. But I will keep an eye of incoming updates on user_data, bfv rebuild and manial
13:10:24 gibi manila
13:10:26 gibi serieses
13:10:38 opendevreview sean mooney proposed openstack/nova stable/yoga: refactor: remove duplicated logic https://review.opendev.org/c/openstack/nova/+/855022

Earlier   Later