Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-12-10
10:54:00 lyarwood elod: that I'm not a fan of tbh
10:55:42 elod lyarwood: ok, good, thx!
11:40:06 openstackgerrit Lucian Petrut proposed openstack/nova master: Fix unshelved image ref https://review.opendev.org/c/openstack/nova/+/765924
11:58:03 openstackgerrit Lee Yarwood proposed openstack/nova-specs master: DNM/WIP Remove device from attach_volume request response https://review.opendev.org/c/openstack/nova-specs/+/765097
12:00:25 gibi lyarwood: I will start looking into the qos failure on the master as melwitt said we propably hit that on stable after we fix the context targeting issue
12:01:05 gibi hm, first, I will confirm that the stable failure after the context targeting fix is the same as the master failure
12:12:49 gibi OK, I confirm that after https://review.opendev.org/c/openstack/nova/+/766364 on stable we hit the same issue as visible on msater
12:12:52 gibi master
12:13:33 openstack Launchpad bug 1907522 in OpenStack Compute (nova) "test_*_with_qos_min_bw_allocation fails in the nova-multi-cell job with: nova.exception.MigrationPreCheckError: Migration pre-check error: Failed to create port bindings for host <host>" [Undecided,New]
12:13:33 gibi so on master we need to fix https://bugs.launchpad.net/nova/+bug/1907522 and then backport the fix to stable on top of https://review.opendev.org/c/openstack/nova/+/766364 which is only needed on stable
12:13:41 gibi this will be fun
13:00:06 gibi lyarwood, melwitt : I'm affraid that we missed the qos support for cross cell resize. I think qos and cross cell work was done in parallel.
13:00:58 gibi as the migration uses different conductor tasks and ComputeManager methods we simply missing the qos code from the cross cell path
13:07:27 openstackgerrit Illes Elod proposed openstack/nova stable/train: DNM: tox constraints test https://review.opendev.org/c/openstack/nova/+/766463
13:38:06 openstackgerrit Balazs Gibizer proposed openstack/nova master: Disable qos resize tempest test for nova-multi-cell job https://review.opendev.org/c/openstack/nova/+/766471
13:38:35 gibi lyarwood, melwitt: ^^ this is my way forward on the master and this needs to be backported to stable
13:44:14 bauzas I agree
13:44:19 bauzas we need to revert the tests
13:44:29 openstackgerrit Balazs Gibizer proposed openstack/nova master: Disable qos resize tempest test for nova-multi-cell job https://review.opendev.org/c/openstack/nova/+/766471
13:45:02 gibi bauzas: the test works fine with same cell resize so I would not revert the whole test
13:47:25 bauzas yeah, by reverting, I meant disabling the faulty ones
13:47:30 bauzas not the whole patch
13:47:37 bauzas anyway +2d with comments
13:49:01 gibi let me quicly fix those nits
13:52:56 openstackgerrit Balazs Gibizer proposed openstack/nova master: Disable qos resize tempest test for nova-multi-cell job https://review.opendev.org/c/openstack/nova/+/766471
13:53:01 gibi bauzas: ^^
13:54:03 bauzas ack
14:05:55 elod lyarwood: I think the bandit patch needs to be merged starting from pike, and then to newer branches, branch by branch... due to grenade job. (which would be unnecessary if bandit 1.6.3 was yanked, but that might not happen soon...)
14:08:27 elod lyarwood: also, I'm experimenting with the 'general' lower-constraint fix ( https://review.opendev.org/q/topic:"constrain-tox-install" ) but it might be better to try to fix the lower-constraints.txt instead, if it does not bring up infinite package version bumps there...
14:10:20 openstackgerrit Balazs Gibizer proposed openstack/nova stable/victoria: Disable qos resize tempest test for nova-multi-cell job https://review.opendev.org/c/openstack/nova/+/766474
14:23:59 gibi elod: lyarwood: now I two possible way forward for stable. a) disable nova-multi-cell job b) disable the qos tests in the nova-multi-cell job
14:24:50 lyarwood hey sorry just back from lunch
14:25:03 lyarwood elod: ack I can work my way forwards from pike if it's going to be easier
14:25:38 lyarwood gibi: https://review.opendev.org/c/openstack/nova/+/765749 is alreadying in the gate for victoria
14:25:46 lyarwood gibi: making the entire job non-voting
14:26:53 gibi lyarwood: OK to merge that. Then to re-enable the job we need https://review.opendev.org/c/openstack/nova/+/766474
14:27:41 gibi then the question what to do https://review.opendev.org/c/openstack/nova/+/766364/
14:29:41 lyarwood gibi: https://review.opendev.org/c/openstack/nova/+/765749 isn't actually in the gate yet so we could yank it out now
14:29:45 elod lyarwood: I've just checked that there's no lower-constraint job @ pike, so bandit patch could be merged easier (fingers crossed) :]
14:30:27 gibi lyarwood: keep it as it is closer to merge
14:30:28 lyarwood elod: cool, I'll try to make a start shortly
14:30:39 lyarwood gibi: kk
14:30:46 gibi lyarwood: then I will try to reenable the gate with https://review.opendev.org/c/openstack/nova/+/766474
14:31:38 gibi regarding https://review.opendev.org/c/openstack/nova/+/766364/ I have to figure out if we need this to avoid leaking something instead of clearly rolling back the migration
14:40:09 gmann brinzhang: ah that is good question, I do not think we need to do any improvement on deprecated APIs. We have not done any update on deprecated API in past too. policy changes were on policy side only so we thought of keeping consistency. But I will discuss it in today meeting to get more opinion.
14:41:50 elod lyarwood: btw, if you are busy, I can do the bandit backports till pike
14:42:17 elod just let me know :)
14:44:10 lyarwood elod: if you have time then please go ahead and I can review
14:44:42 gmann brinzhang: but server sec group is not deprecated API -https://docs.openstack.org/api-ref/compute/?expanded=list-security-groups-by-server-detail#servers-security-groups-servers-os-security-groups
14:45:10 elod lyarwood: ok, doing it now
14:45:52 gmann brinzhang: ah just saw that we did mention about not to change deprecated APIs for tenant-id - https://specs.openstack.org/openstack/nova-specs/specs/wallaby/approved/remove-tenant-id.html
14:46:24 gmann brinzhang: so we are good on that and let's target only non-deprecated APIs only
14:49:12 gmann brinzhang: also our all deprecated APIs are capped with corresponding microversion and they are 404 for higher than deprecated microversion so technically we cannot do any change in those API which include the microversion bump.
14:49:55 openstackgerrit Illes Elod proposed openstack/nova stable/stein: [stable-only] Cap bandit to 1.6.2 https://review.opendev.org/c/openstack/nova/+/766487
15:00:52 openstackgerrit Illes Elod proposed openstack/nova stable/stein: [stable-only] Cap bandit to 1.6.2 https://review.opendev.org/c/openstack/nova/+/766487
15:03:47 openstackgerrit Illes Elod proposed openstack/nova stable/rocky: [stable-only] Cap bandit to 1.6.2 https://review.opendev.org/c/openstack/nova/+/766492
15:11:14 lyarwood elod: Bernard just replied on the ML about the cross project issue with the bandit cap
15:11:32 lyarwood elod: and thinking about it more an overall cap in requirements might be the only way to unblock things here
15:11:49 lyarwood elod: but landing from stable/pike as you suggested for the project specific changes
15:11:56 lyarwood elod: so grenade works etc
15:12:53 lyarwood `ERROR Failed to update project None in 3s` oh cool a new zuul error
15:13:29 openstackgerrit Illes Elod proposed openstack/nova stable/pike: [stable-only] Cap bandit to 1.6.2 https://review.opendev.org/c/openstack/nova/+/766497
15:14:30 lyarwood https://github.com/PyCQA/bandit/issues/663 that said it looks like they might yank 1.6.3 after all
15:23:53 gibi lyarwood: this is not the first time I saw that zuul ERROR
15:24:09 gibi lyarwood: at the previous time it was around a gerrit restart so I assmue that caused it
15:24:33 lyarwood ah kk
15:24:54 bauzas ergh, call me stupid
15:25:00 bauzas requested_networks can be None
15:26:39 gibi lyarwood: pinged infra about it
15:28:00 lyarwood gibi: ah thanks, I was just going to recheck tbh
15:28:10 gibi lyarwood: sure,
15:29:27 gibi lyarwood: infra increased some git cloning timeouts in the zuul executors
15:29:32 gibi recently
15:30:22 gibi they expect that it helps
15:35:09 elod lyarwood: if they would yank bandit 1.6.3 that would be the best :)
15:35:53 elod lyarwood: until that, let's see if we can merge the bandit cap patches, as it might take some time
15:36:06 lyarwood elod: ack lets keep trying
15:36:27 lyarwood elod: I'm just worried that we will need to land multiple caps across multiple projects to allow the gate to work again
15:37:08 elod lyarwood: yes, that's possible :S
17:00:12 gibi lyarwood, melwitt: I rechecked https://review.opendev.org/c/openstack/nova/+/766471 (master unblock patch) as it bounced due to zuul " ERROR Failed to update project None in"
17:01:07 gibi I have to leave in ~15 minutes so I hope melwitt can take over rechecking it if needed
17:12:14 openstackgerrit Balazs Gibizer proposed openstack/nova master: DNM https://review.opendev.org/c/openstack/nova/+/766509
17:13:52 openstackgerrit Balazs Gibizer proposed openstack/nova stable/victoria: DNM https://review.opendev.org/c/openstack/nova/+/766511
17:15:42 lyarwood gibi: ack thanks I'll also be around for a while to recheck when required
17:16:04 gibi lyarwood: thanks I will check back too before I go to sleep but now dinner time
17:16:07 gibi o/
17:17:32 gibi lyarwood: ahh, it is already got errors from zuul on https://zuul.opendev.org/t/openstack/status
17:17:39 gibi :/
17:17:54 gibi so I think we have to wait for infra
17:22:23 lyarwood ack np enjoy dinner \o
17:29:20 bauzas sean-k-mooney: around ?
17:29:38 bauzas sean-k-mooney: I found why my functest is not working, I need to mock the neutron extension for segments
17:30:24 sean-k-mooney bauzas: ah yes you do
17:30:40 sean-k-mooney i think i have an example of that
17:30:58 sean-k-mooney in one of teh regression test i wrote
17:32:13 sean-k-mooney bauzas: https://github.com/openstack/nova/commit/71bc6fc9b89535679252ffe5a737eddad60e4102#diff-4763b3c03cb51e2efd60e19a11af59d4bfb76fbca93c953dfd9201eb5187ee88
17:33:11 sean-k-mooney bauzas: you can just do https://github.com/openstack/nova/blob/master/nova/tests/functional/regressions/test_bug_1888395.py#L38-L55
17:33:54 sean-k-mooney bauzas: you could also update the fixture
17:34:18 sean-k-mooney bauzas: but im not sure we want it there by default
17:35:14 sean-k-mooney https://github.com/openstack/nova/blob/master/nova/tests/fixtures.py#L1891-L1905

Earlier   Later