Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-08-20
14:46:37 sean-k-mooney stephenfin: ill address your comments on the deprecation patch after all of that is done ^
14:48:22 mriedem melwitt: so abandon https://review.opendev.org/#/c/675218/ at this point yeah?
14:49:18 aspiers alex_xu: does https://review.opendev.org/#/c/638680/46/nova/tests/functional/libvirt/test_report_cpu_traits.py@92 make sense now?
14:50:57 mriedem melwitt: actually https://review.opendev.org/#/c/507486/67/nova/db/sqlalchemy/api.py@5572
14:52:07 sean-k-mooney aspiers im changeing how some of that works by the way.
14:52:08 mriedem meta = MetaData(get_engine(context=context, use_slave=True))
14:52:08 mriedem heh yeah in PS65 you had this:
14:52:28 aspiers sean-k-mooney: some of which?
14:52:40 sean-k-mooney how the libvirt driver reports traits
14:52:52 aspiers sean-k-mooney: I'm just reviewing your patch now
14:53:38 aspiers sean-k-mooney: you mean https://review.opendev.org/#/c/666915/8/nova/virt/libvirt/driver.py ?
14:54:19 sean-k-mooney that and the general helper function i have intoduced to remove most of the bolier plate
14:57:34 sean-k-mooney aspiers: hopefully the chagne wont conflict or wont confict too badly just wanted you to be aware
14:57:50 aspiers sean-k-mooney: I don't think it will conflict because the SEV trait is provided via a capability
14:58:27 sean-k-mooney ok
14:58:38 aspiers fingers crossed anyway, because you are just ahead of me on the runway X-D
15:01:17 sean-k-mooney sure i am vauguly aware that the new version of lxml broke things
15:02:36 sean-k-mooney i had thought that there was a larger patch that capped 4 or 5 libs floating around somewhere
15:05:01 sean-k-mooney efried: this is the ml thread http://lists.openstack.org/pipermail/openstack-discuss/2019-August/008562.html
15:05:55 efried yes
15:06:25 sean-k-mooney i think the bug present in how we compare xmls as equal or somthing like that
15:06:45 sean-k-mooney so i dont think it breaks the libvirt dirver but does break our unit tests
15:06:56 efried sean-k-mooney: in short, we either need to see if we can fix this in nova, or blacklist the lxml version(s) and open bugs against them and eventually uncap if/when they fix etc etc.
15:07:39 sean-k-mooney we likely shoudl be able to fix it in nova. ill see if i can repoduce this later today.
15:08:06 efried thanks sean-k-mooney. It just looks like attr ordering. We could probably resolve it in XMLMatcher, I would think.
15:08:16 efried I'm actually kind of surprised that that's not already taken into account there.
15:08:27 efried or maybe those tests aren't using XMLMatcher.
15:08:27 sean-k-mooney current upper-constratis is set to 4.3.4? so we dont hit this
15:08:48 efried sean-k-mooney: that's the point, prometheanfire is trying to get u-c to the latest published versions and running into this.
15:09:19 sean-k-mooney yep i know but we simply would not uncapp them in Train if that was not fixed by m3
15:09:32 sean-k-mooney we should be able to fix it if its not too pervasive before then
15:10:20 sean-k-mooney it could be that it broke the XMLMatcher i just vaguly recall talking to prometheanfire about this so need to look at the actull error first to load context
15:12:42 prometheanfire ya, think we talked about it, wasn't it an ordering thing?
15:12:49 sean-k-mooney yep
15:12:54 sean-k-mooney i think so
15:13:19 sean-k-mooney once i reporduce it it will likely be a simple fix
15:13:33 sean-k-mooney either use the matcher if its using it already or fix the matcher
15:13:53 sean-k-mooney we created the xmlmatcher to ignore order
15:14:17 sean-k-mooney unlike other markup formate whitespace in xml and order is technically signifcant
15:14:22 prometheanfire just made https://review.opendev.org/677480 for you to submit reviews against to test (if you want)
15:15:01 sean-k-mooney ya i saw i can just add a depens on to make sure it passes
15:15:06 efried prometheanfire: Any particular temest test I should look at for the websockify thing?
15:15:10 efried tempest*
15:15:19 prometheanfire I'm not sure for that one
15:15:27 efried so I can pare down the jobs and conserve resources
15:15:32 efried okay, I'll just run 'em all then.
15:15:40 openstackgerrit Eric Fried proposed openstack/nova master: DNM: Test websockify 0.9.0 https://review.opendev.org/677483
15:15:50 prometheanfire kk, if you want to wait for https://review.opendev.org/677479 you can
15:16:09 openstackgerrit Eric Fried proposed openstack/nova master: DNM: Test websockify 0.9.0 https://review.opendev.org/677483
15:16:29 efried oh, is that going to run something tempestuous?
15:16:43 efried I made --^ depend on it
15:17:08 sean-k-mooney efried: i am going to loudly ignore that ....
15:17:36 sean-k-mooney thats such a dad joke
15:17:37 prometheanfire it does :D
15:17:45 efried sean-k-mooney: non sequitur? Assume we're talking about aspiers's python-rather-than-bash thing now?
15:18:00 openstackgerrit melanie witt proposed openstack/nova master: nova-manage db archive_deleted_rows is not multi-cell aware https://review.opendev.org/507486
15:18:01 openstackgerrit melanie witt proposed openstack/nova master: Verify archive_deleted_rows --all-cells in post test hook https://review.opendev.org/672840
15:19:25 openstackgerrit Merged openstack/nova master: Update api-ref for 2.75 to add config_drive in server update response https://review.opendev.org/677324
15:19:25 sean-k-mooney i assume the horrible pun between tempestuous and tempest was intentionally because if it wasnt that even worse some how
15:19:42 prometheanfire we do include a general tempest test in our test suite
15:20:43 openstackgerrit Matt Riedemann proposed openstack/nova master: Add functional recreate test for bug 1823370 https://review.opendev.org/650424
15:20:44 openstackgerrit Matt Riedemann proposed openstack/nova master: WIP: Use RequestSpec.requested_destination when rebuilding with new image https://review.opendev.org/650376
15:20:44 openstackgerrit Matt Riedemann proposed openstack/nova master: Restrict RequestSpec to cell when evacuating https://review.opendev.org/650429
15:20:44 openstack bug 1823370 in OpenStack Compute (nova) "Evacuations are not restricted to the source cell during scheduling" [Medium,In progress] https://launchpad.net/bugs/1823370 - Assigned to Matt Riedemann (mriedem)
15:21:17 sean-k-mooney prometheanfire: did you have a patch that put all the constratis fixes into one by the way or were they just listed in the email
15:23:37 prometheanfire no, generally the global-reqs nightly update will include it
15:23:45 prometheanfire that's where I see it generally
15:25:40 sean-k-mooney ok i might have been looking at an automated patch
15:31:12 efried stephenfin: Where was https://review.opendev.org/#/c/677444/1/doc/source/cli/nova-manage.rst rendering before? I don't see it anywhere in https://docs.openstack.org/nova/latest/cli/nova-manage.html
15:31:41 stephenfin efried: https://review.opendev.org/#/c/677443/
15:32:04 efried o
15:32:19 stephenfin I didn't want that patch to block the docs patch
15:32:35 stephenfin plus the docs one might be worth backporting
15:33:08 openstackgerrit Matt Riedemann proposed openstack/python-novaclient master: API microversion 2.76: Add 'power-update' external event https://review.opendev.org/666792
15:33:26 efried stephenfin: So the CLI remote_debug options... no longer work in that top patch?
15:34:07 stephenfin efried: For nova-manage, nova-policy and nova-status, but they never actually worked there
15:34:14 stephenfin Because there's no "service" to debug
15:35:09 efried okay, that makes sense. So why bother documenting them in that middle patch?
15:35:20 efried Or should we at least say "These don't work, don't try to use them"?
15:35:37 stephenfin I wanted to be complete, but perhaps I should just drop them
15:35:49 stephenfin Dropping them is probably the wiser option, actually
15:36:01 stephenfin (never documenting them, that is)
15:37:02 efried ++
15:42:56 openstackgerrit Stephen Finucane proposed openstack/nova master: config: Explicitly register 'remote_debug' CLI opts https://review.opendev.org/677444
15:42:56 openstackgerrit Stephen Finucane proposed openstack/nova master: docs: Document global options for nova-manage https://review.opendev.org/677443
15:43:03 stephenfin donezo
15:43:28 efried stephenfin: we may have crossed in the mail, there was another comment on the middle one
15:44:22 stephenfin efried: Hmm, there shouldn't be another '--version' parameter any more. I killed that
15:44:29 openstackgerrit Ghanshyam Mann proposed openstack/nova master: Add new default roles in os-services API policies https://review.opendev.org/648480
15:45:07 stephenfin looks like there's more cleanup to do \o/
15:45:13 stephenfin and red pen
15:47:07 mriedem mnaser: are you working around this in stein right now https://review.opendev.org/#/q/topic:bug/1839560+branch:stable/stein or is getting that in pretty critical for you?
15:48:18 stephenfin sean-k-mooney: One easy follow-up on https://review.opendev.org/#/c/671338/ but it's +W'd
15:51:05 sean-k-mooney ya i get a littel confused where that works and where it does not. e.g. does not in release notes but does in docs but not in someting else i tried
15:52:32 stephenfin sean-k-mooney: Inside 'doc/source', it works. Everywhere else, it does not
15:53:07 stephenfin That's because 'doc', 'releasenotes', 'api-ref' and 'api-guide' are all separate doc builds with their own indexes
15:54:11 sean-k-mooney it was the api ref i think i tried to use it in
15:54:14 sean-k-mooney ok so just docs
15:54:20 stephenfin yeah, no go there
16:26:48 openstackgerrit Stephen Finucane proposed openstack/nova master: docs: Document global options for nova-manage https://review.opendev.org/677443

Earlier   Later