Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-08-15
16:57:32 artom dansmith, mriedem, hey, you guys were the main reviewers of NUMA LM in Stein, will have bandwidth to do the same in Train, or should I be sucking up to other cores? ;)
16:57:40 artom *will you have
16:58:08 dansmith I will try
16:58:26 artom dansmith, ack, thanks :)
16:58:48 openstackgerrit sean mooney proposed openstack/nova master: Libvirt: add support for vPMU configuration. https://review.opendev.org/671338
17:01:52 mriedem artom: not today
17:02:05 mriedem artom: you should also beat down the starlingx people's door,
17:02:12 artom mriedem, I wasn't really expecting today :)
17:02:14 mriedem because at the ptg and last summit they were saying this was the big ticket thing they needed
17:02:15 artom I meant in general
17:02:28 mriedem so if this is so critical for starlingx they should have people reviewing and testing it
17:02:30 artom mriedem, yeah, I emailed the ML, and dtroyer and Gerry personally
17:02:40 artom Radio silence :/
17:03:00 mriedem cfriesen is probably the best starlingx person to review the series but he's been out of nova for awhile now
17:03:08 sean-k-mooney well that and sriov numa policies was on the list too.
17:03:25 sean-k-mooney i think we almost have everything this cycle?
17:03:35 artom Gerry Kopec (not upstream as far as I can tell, though he did leave some reviews) tested it
17:03:41 artom And found some problems
17:03:50 artom Was hoping he'd step up again
17:04:03 artom Radio silence so far. Which is weird, because Dean actually emailed me first
17:04:06 artom Asking how it was going
17:05:12 sean-k-mooney artom: i need to set up a second node but when im done testign a sriov thing i plan to test your numa live migration code with stephens cpu pinning stuff and my sriov migration code and report back
17:05:26 sean-k-mooney eta on that is sometime next week
17:06:06 artom sean-k-mooney, my series has conflicts with stephenfin's, so not sure how you'll test those together
17:06:12 artom Well, resolve manually, I suppose
17:06:33 mriedem cpu pinning stuff == PCPU?
17:06:33 sean-k-mooney ya i was thinking of doign that but if there are a lot ill test them seperatly
17:06:38 sean-k-mooney yes
17:06:57 sean-k-mooney i would also like to test gibis bandwith stuff
17:07:01 mriedem if it were me, i'd test artom's numa live migration stuff w/o trying to wrangle the PCPU stuff in
17:07:05 mriedem jesus
17:07:09 mriedem so yo'ud like to actually not get anything done :)
17:07:36 sean-k-mooney well ill test them seperatly first
17:07:45 artom He'll become a master at git rebasing and conflict resolution though :)
17:08:00 sean-k-mooney but i want to actully test those 4 features in 1 vm in train
17:08:27 sean-k-mooney i guess it better to wait untill some/all have merged to do that
17:09:09 mriedem well i applaud your efforts for at least trying to test this stuff
17:09:11 mriedem since no one else is
17:09:21 mriedem together i mean
17:09:56 sean-k-mooney well people will want to combine those so i want to test them before i get the bug report
17:10:10 mriedem don't worry, the bug report will be in 18+ months
17:10:20 mriedem you might not even be working on openstack at that point
17:10:33 sean-k-mooney ya.... if we are lucky and people deploy train by then
17:11:27 artom I still don't think testing them together before they're even merged is a good idea
17:11:29 sean-k-mooney TIL we still have at least 1 customer with a suppor exception for liberty until 2021 ...
17:11:33 artom Laudable, but not good ;)
17:11:49 prometheanfire nova still has an issue with the latest libvirt https://storage.gra1.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/logs_16/675816/5/check/cross-nova-py36/a6c71f8/testr_results.html.gz
17:11:55 sean-k-mooney artom: well in that case ill just test them 1 by one
17:12:06 artom But yeah, I totally get mriedem's gripe about the RH dev/QE lag
17:12:10 openstackgerrit Adam Spiers proposed openstack/nova master: Fix use of mock.patch with new_callable=PropertyMock https://review.opendev.org/676738
17:12:52 aspiers efried: ^^^ hope that's right, still trying to wrap my head around some of this stuff
17:13:08 artom It's been half-assedly brought up internally, but it takes a lot of work (literally - force over a distance) for a long time to change that
17:13:24 mriedem artom: well it's not really only RH, it's any distro that EOLs an old series and customers are forced to move up
17:13:41 aspiers any other mock experts wanna review that?
17:13:48 mriedem which is why we get all the RH queens bug reports bubblin gup now
17:14:23 artom mriedem, are there other significant distros besides OSP at this point? Honest question
17:14:36 cdent efried: I assume you probably know this already and we're merely following protocol: but I didn't write that "single pass instance info fetch in host manager". My ressurection of it was solely based on the profiling benefits. So a) I'll have to dig deeper to respond to the comment, b) the comments make that digging easier.
17:14:44 artom Commercial, I mean
17:14:45 mriedem canonical? VIO? suse? idk ask their devs
17:14:47 sean-k-mooney i think ubuntu still has more deplopyment then osp
17:14:52 efried aspiers: lgtm, thank you.
17:15:00 artom Ah, SUSE, right
17:15:08 artom (With apologies to aspiers)
17:15:08 mriedem artom: you can ask aspiers how suse is ramping up for rolling out rocky with SEV forks :)
17:15:31 sean-k-mooney and canonical openstack
17:15:32 aspiers artom: what are you apologising for? :)
17:15:45 artom aspiers, forgetting you exist, essentially ;)
17:15:49 aspiers haha
17:15:55 aspiers easily done
17:16:06 aspiers mriedem: yeah sadly that's probably gonna happen
17:16:33 efried cdent: ack. If we want to land this pretty much as-is, we just need unit/func test coverage in this patch. The rest can be FUPed.
17:16:34 aspiers Not 100% sure though. We have a policy of trying to stay as close to upstream as humanly possible, even on stable branches.
17:16:58 aspiers artom: let me know if I can help with anything from the distro angle
17:17:07 sean-k-mooney so do we but PMs have a different opipion on that when there are $$$$
17:17:22 artom aspiers, thank you :) Though we were just discussing dev/QE lag
17:17:36 artom aspiers, and how the RH QE is lagging behind upstream
17:17:55 artom So we'll do an RFE, and like a year later, we find out if never worked
17:18:07 artom (Only slightly exaggerating)
17:18:12 aspiers artom: well at least you do QE on upstream :)
17:18:21 artom aspiers, we don't, hence the lag
17:18:34 aspiers I mean eventually
17:18:40 artom By the time stuff's in our builds, it's been months, if not years
17:18:45 aspiers lag implies it happens eventually :)
17:18:53 aspiers If it never happens, that's infinite lag
17:19:09 aspiers although maybe I kinda just contradicted myself
17:20:04 openstackgerrit Matt Riedemann proposed openstack/nova master: API microversion 2.76: Add 'power-update' external event https://review.opendev.org/645611
17:20:09 mriedem dansmith: ^
17:35:11 dansmith mriedem: looks good
18:03:57 melwitt argh why does it do "An auth plugin is required to determine endpoint URL" when I'm using the PlacementFixture
18:04:17 efried melwitt: You need to use the sdk fixture :)
18:04:19 melwitt this is what happens when I don't use ServersTestBase
18:04:39 melwitt oh. ok
18:04:53 efried (as well as PlacementFixture, if what you're trying to do involves real placement calls)
18:05:32 melwitt the sdk fixture is in test.TestCase
18:06:15 efried melwitt: are you getting the error locally or in zuul?
18:06:49 efried locally => upgrade your ksa and sdk in your virtualenv
18:06:49 efried zuul => rebase and recheck, since the u-c bump landed overnight.
18:06:58 melwitt locally. I'm trying to change from deriving from ServersTestCase to just TestCase
18:07:07 melwitt ok
18:07:39 efried yeah, I bet it's because ksa and/or sdk in your venv is downlevel OR those are latest but you don't have the new fixture you merged last night.

Earlier   Later