Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-02-19
15:13:25 aspiers or traits could have an owner associated explicitly
15:14:06 aspiers so then creation and deletion would always both one-way pushes from provider tree to placement
15:15:20 efried aspiers: Have a gander at this: https://github.com/openstack/nova/blob/master/nova/tests/functional/libvirt/test_report_cpu_traits.py
15:15:48 kashyap stephenfin: What do you mean by a "UT"?
15:15:49 efried aspiers: Yes, we discussed those things a bit at a PTG or two, decided on this.
15:15:57 stephenfin kashyap: Unit Test
15:16:05 kashyap Ah, thanks. :-)
15:16:05 stephenfin I'd assume, anyway
15:16:27 kashyap stephenfin: Yes, you win, because the `wtf` tool actually detects the acronym
15:16:45 kashyap $> rpm -qf `which wtf`
15:16:46 kashyap bsd-games-2.17-57.fc28.x86_64
15:17:03 kashyap It's super-tiny RPM. And comes with an acryonym file
15:18:25 aspiers efried: I found this post-Dublin conversation http://eavesdrop.openstack.org/irclogs/%23openstack-nova/%23openstack-nova.2018-03-12.log.html#t2018-03-12T16:09:22
15:19:13 aspiers I was reading through to try to catch up on a year's worth of thinking (not to mention several decades of nova experience...)
15:19:41 aspiers it was linked from https://review.openstack.org/#/c/552122/4
15:20:01 aspiers efried: but OK, for now I'll just do reset()
15:20:51 efried aspiers: Yeah, that's the right thing. Not sure what to do about the other. Maybe we just leave it alone and hope it doesn't become a problem. Operators always do what they're supposed to, right?
15:21:35 aspiers haha yeah sure, always ;-)
15:28:36 aspiers efried: it works!
15:30:08 efried woot!
15:31:07 aspiers It's been an unexpectedly rough 24 hours on this. Took me many hours yesterday trawling through with pdb to find this microversion issue https://review.openstack.org/#/c/637657/
15:31:39 aspiers Once things go wrong in the middleware that's a whole can of voodoo worms which I'm not used to opening up
15:33:22 efried holy unnecessary shim, Batman
15:33:40 efried I remember last time I encountered these fake_* methods I griped about how they should just diaf.
15:33:41 aspiers X-)
15:35:02 efried aspiers: in fact, I may demand that we do that instead of the fix you've proposed.
15:36:00 aspiers That would make me sad, because it would introduce a big new dependency to my SEV work
15:36:36 efried aspiers: It wouldn't be a big thang; just needs someone to do it. And since I'm the one being difficult...
15:36:40 aspiers I'm trying to do the right thing by landing the capabilities-to-traits feature and then implementing the SEV trait on top of it
15:37:06 aspiers rather than just tweaking the libvirt driver's u_p_t()
15:37:48 efried okay, I hear you.
15:37:50 aspiers The latter would be the easy way out, but I wanted to give something useful upstream before asking for "our" feature work to be accepted
15:37:54 efried Give me half an hour?
15:38:06 aspiers Oh wow, well sure, if you're that fast :)
15:38:23 jangutter kashyap: wtf's pretty nifty. It's got dry, crud, rest and a bunch of others that are tricky to google.
15:38:24 efried aspiers: I remember thinking it would be pretty easy. 25 minutes of that is going to be refreshing myself on the code.
15:38:27 aspiers It sounded to my inexperienced ears like it would take days
15:38:39 efried aspiers: What test were you running when you hit this?
15:38:48 kashyap jangutter: Yeah.
15:38:52 aspiers efried: the patch which sits on top of it
15:39:02 aspiers efried: https://review.openstack.org/#/c/538498/10
15:39:16 aspiers but let me upload a working PS 11
15:39:30 aspiers which will even pass PEP8 *shock* :)
15:39:37 openstackgerrit sean mooney proposed openstack/nova master: libvirt: auto detach/attach sriov ports on migration https://review.openstack.org/629589
15:39:46 efried aspiers: test_servers presumably?
15:39:48 aspiers turns out that tox -e flake8 pretends to do something useful but actually checks nothing at all ...
15:39:52 aspiers correct
15:40:06 aspiers stestr --test-path=./nova/tests/functional run -n nova.tests.functional.test_servers.TraitsTrackingTests.test_resource_provider_traits
15:40:16 aspiers if you want the quicker version
15:40:24 efried okay, thanks.
15:40:48 kashyap stephenfin: I'm still a bit squirmy about this whole muddling with s390x, and its console devices. Because both "sclp" and "sclplm" are for different purposes, per s390x maintainers. And I don't have access to a machine to test it all
15:40:54 efried aspiers: btw, if you install tox 3.1.1 (uh, which I guess I forced you to do anyway), tox -e functional -- test_resource_provider_traits should be almost as quick.
15:40:58 kashyap So I was just poking the code to see until it poked me back...
15:41:15 aspiers efried: yeah I figure that out, but I find that tox still adds unacceptable overhead
15:41:28 efried for test discovery?
15:41:30 aspiers the {posargs} trick is handy though
15:41:31 kashyap stephenfin: Anyway, so far from our code audit it is mergeable; so no need to belabour more on it now. </EOM>
15:41:38 aspiers not just test discovery
15:41:59 efried We got rid of the bytecode
15:42:19 aspiers efried: it takes ~2s on my machine to get to the point of launching stestr
15:42:44 aspiers it has to go through all the virtualenv stuff first
15:43:10 mdbooth mriedem: If you recall this one: https://review.openstack.org/#/c/626952/ . It looks like the commit message there is rubbish. I haven't been able to reproduce the underlying issue and I suspect it's actually been fixed by upgrading libraries. Haven't worked out which ones, though.
15:43:59 mdbooth That said, I still like this cleanup, and I suspect the approach is slightly more robust to similar issues. Is it worth rewriting the commit message and keeping it around?
15:44:23 mdbooth If there's no interest I'll just drop it. Or lyarwood will, anyway :)
15:45:02 mriedem i don't have a strong opinion
15:46:26 mdbooth mriedem: ack. I'll probably drop it then as it's likely to languish. Thanks.
15:46:36 mdbooth lyarwood: ^^^ Lets drop it.
15:49:35 openstackgerrit Adam Spiers proposed openstack/nova master: WIP: Convert driver supported capabilities to compute node provider traits https://review.openstack.org/538498
15:49:52 aspiers efried: there's PS 11 with a working test which includes the reset()
15:50:38 efried aspiers: The test was working for me at ps10
15:51:30 aspiers efried: yeah but it didn't test the reappearance of the trait
15:52:01 aspiers efried: https://review.openstack.org/#/c/538498/10..11/nova/tests/functional/test_servers.py@2274
15:52:18 efried okay. It doesn't really matter; I just need something that's supposed to hit that fake_delete
15:52:25 aspiers good point
16:00:53 mriedem cdent: so the nova functional tests - they use the placement fixture from extracted placement right? not the in-tree placement code. so doing https://review.openstack.org/#/c/637892/ in nova would only be for runtime if we copied that
16:02:24 cdent yes
16:02:43 mriedem ok
16:05:57 cdent mriedem: plus that osa thing that m*naser was gunna set up
16:08:03 openstackgerrit sean mooney proposed openstack/os-vif master: Change python3.5 job to python3.7 job on Stein+ https://review.openstack.org/610068
16:08:16 mriedem cdent: yeah i'm more concerned (right now anyway) on the subunit parser thing
16:11:36 bauzas efried: others, I found yet another issue with vgpu reshapes...
16:12:02 bauzas efried: others, given we need to recreate mdevs when rebooting, we need to pass allocations when init_host()
16:12:23 bauzas I wrote a bit of comment in the first change, but I'm working on a change that will do this
16:17:20 mdbooth lyarwood: https://review.openstack.org/#/c/623489/
16:17:39 mdbooth lyarwood: Just looking at the unit tests, but feel free to consider ^^^ anyway.
16:20:59 lyarwood mdbooth: ack and ack looking
16:24:41 kashyap BTW, sean-k-mooney: thanks for inadvertently reminding that this is a pending item a few weeks ago! Almost lost track of it: https://review.openstack.org/#/c/632507
16:26:07 sean-k-mooney kashyap oh the libvirt version bump ya
16:31:08 openstackgerrit Sylvain Bauza proposed openstack/nova master: libvirt: implement reshaper for vgpu https://review.openstack.org/599208
16:31:08 openstackgerrit Sylvain Bauza proposed openstack/nova master: Add functional test for libvirt vgpu reshape https://review.openstack.org/631559
16:56:46 sean-k-mooney so are we still going ahead with https://review.openstack.org/#/c/634949/ since the cinder volume test in tempest-full are continuosly failing.
16:59:35 mriedem what cinder volume tests?
17:00:39 mriedem we did do https://review.openstack.org/#/c/634970/ to hopefully help with the tooz connection errors
17:01:08 sean-k-mooney tempest.api.compute.volumes.test_attach_volume_negative.AttachVolumeNegativeTest.test_attach_attached_volume_to_same_server
17:02:08 sean-k-mooney i dont know if it sthe same two but i have notice 1-2 cinder volume attachment test filaing in a few reivews i looked at recently
17:03:09 sean-k-mooney usually either with a timeout or sometime with Invalid volue errors
17:03:16 sean-k-mooney like http://logs.openstack.org/15/620115/23/check/tempest-full/fcabb4b/testr_results.html.gz
17:03:35 smcginnis mriedem: I haven't noticed that tooz issue anymore. Have you seen any since that change?
17:04:44 mriedem smcginnis: http://status.openstack.org/elastic-recheck/#1810526
17:05:04 mriedem those could be stable branches, checking logstash

Earlier   Later