| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-08-23 | |||
| 12:04:35 | gmann_ | alex_xu: i added this item in PTG etherpad also - https://etherpad.openstack.org/p/nova-ptg-queens | |
| 12:23:04 | alex_xu | gmann_: thanks! | |
| 12:39:32 | openstackgerrit | Viktor Varga proposed openstack/nova master: Remove mox from unit/conductor/test_conductor.py (1) https://review.openstack.org/496691 | |
| 12:44:22 | cdent | (gibi-core)++ | |
| 12:45:19 | alex_xu | nova api meeting is in 15 mins at #openstack-meeting-4 | |
| 12:45:32 | bhagyashri_s | jaypipes, melwitt: Hi, I want to discuss regarding bfv case, I have done some testing on the stable/pike and what i observed is even if I boot the instance from volume then in that case it allocates the disk_gb as the flavor root_gb in both compute_nodes and allocations table but after applying the patch [1]: https://review.openstack.org/#/c/428481/ on stable/pike then it makes the correct allocations. So why these two patche [1][2] are | |
| 12:45:32 | bhagyashri_s | ignored in the pike. [1]: https://review.openstack.org/#/c/428481/ [2]: https://review.openstack.org/#/c/428505/ | |
| 12:47:55 | edleafe | stephenfin: sure. Would it be easier for me to revise it and push a new version? | |
| 12:49:38 | stephenfin | edleafe: Probably would, yeah. cdent just left a comment saying he'd take a look too. Feel free to take ownership of the whole thing if you'd like to - I just care about the resulting doc :) | |
| 12:49:44 | jaypipes | bhagyashri_s: you can propose them as backports to stable/pike. they just were not highest priority during the last two weeks | |
| 12:50:03 | cdent | edleafe: if you’ve got time/brain you may take point | |
| 12:51:00 | edleafe | cdent: well, since it was my poorly-worded email that started all this... :) | |
| 12:51:09 | edleafe | I'll take a crack at it | |
| 12:51:26 | stephenfin | jaypipes: Talking of backports, could you stick this on your review backlog? https://review.openstack.org/#/c/496605/ Something I'd like to get into Pike before final | |
| 12:52:03 | bhagyashri_s | jaypipes: ok I can. Thank you :) | |
| 12:53:51 | cdent | thanks edleafe | |
| 12:55:09 | jaypipes | stephenfin: I'm not core on stable :) | |
| 12:56:03 | stephenfin | jaypipes: Good thing it's on master (I hope :)) Master first, then backport | |
| 12:56:15 | stephenfin | Only if you've time though, of course | |
| 12:58:46 | stephenfin | thanks edleafe, cdent :) | |
| 13:00:26 | alex_xu | nova api meeting is running at #openstack-meeting-4 | |
| 13:04:56 | jaypipes | stephenfin: done | |
| 13:05:12 | stephenfin | jaypipes: Ta :) | |
| 13:05:15 | mriedem | jaypipes: alex_xu: stephenfin: oomichi: we've got a series of fixes needed for RC2 starting here: https://review.openstack.org/#/c/496031/ - they all have a +2 already | |
| 13:05:30 | stephenfin | mriedem: Sure, I'll take a look in 2 | |
| 13:12:30 | jaypipes | stephenfin: tá fáilte romhat | |
| 13:13:29 | openstackgerrit | Stephen Finucane proposed openstack/nova master: conf: Deprecate 'keymap' options https://review.openstack.org/483994 | |
| 13:15:02 | stephenfin | :D | |
| 13:20:29 | jaypipes | mriedem: thx. my eyes are now bleeding. | |
| 13:27:01 | zen369 | how do i check what openstack release i'm running? | |
| 13:32:18 | jaypipes | mriedem: conductor/tasks/live_migrate.py is code that runs in the cell-specific conductor, right? not the superconductor... | |
| 13:33:10 | dansmith | jaypipes: what does it matter? | |
| 13:33:25 | dansmith | jaypipes: however, when we have cross-cell migrations it would have to run on superconductor, and I expect it does today | |
| 13:33:52 | jaypipes | dansmith: oh hey dan :) for some reason I thought you were on vacay... | |
| 13:34:00 | jaypipes | dansmith: I was just checking that this code: https://review.openstack.org/#/c/496419/2/nova/conductor/tasks/live_migrate.py | |
| 13:34:10 | jaypipes | dansmith: would indeed be possible to run in the cell conductor | |
| 13:34:33 | jaypipes | dansmith: since I thought instance mapping information was in the API db. | |
| 13:34:54 | jaypipes | dansmith: just wanted to double check that those code paths were kosher for the cell conductor. | |
| 13:35:08 | dansmith | no, cell conductor can't look things up in api db | |
| 13:35:27 | dansmith | all of our moves should be happening in superconductor | |
| 13:36:23 | jaypipes | dansmith: ok, so is my cause for concern grounded in good reason? | |
| 13:36:35 | openstackgerrit | Merged openstack/nova master: Fix quobyte test_validate_volume_no_mtab_entry https://review.openstack.org/495485 | |
| 13:36:49 | dansmith | jaypipes: your concern is that it can't run in cell conductor right? | |
| 13:37:02 | jaypipes | dansmith: sorry, it's also the end of the day for me so I very well might just be tired :) | |
| 13:37:08 | jaypipes | dansmith: yeah, that is my concern | |
| 13:37:27 | dansmith | jaypipes: and I'm telling you it is running in superconductor, so I don't think there's a problem | |
| 13:37:46 | jaypipes | dansmith: k, got it. I just wanted to double-check, that's all. | |
| 13:39:00 | dansmith | also, the way devstack in the gate runs, it should fail if he's doing something that requires api db access | |
| 13:39:51 | jaypipes | oh, good to know we have the safeguard in place. nice. ++ | |
| 13:40:57 | jaypipes | mriedem: k, that series +W'd | |
| 13:41:51 | mriedem | jaypipes: thanks | |
| 13:42:24 | stephenfin | sean-k-mooney: Remind me - we can't guarantee that the portbinding extension is enabled, right? https://review.openstack.org/#/c/357726/ | |
| 13:42:29 | stephenfin | i.e. it's still optional | |
| 13:43:57 | stephenfin | wait - after reading it, it's clear we want to enforce this. nvm me :) | |
| 13:44:42 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Assume neutron port_binding extensions enabled https://review.openstack.org/357726 | |
| 13:47:38 | mriedem | stephenfin: the assumption is that no one is *not* enabling the portbinding extension | |
| 13:47:57 | stephenfin | mriedem: Yeah, I read the reno. That's a fair assumption | |
| 13:48:44 | mriedem | like i said in my comment, that's a thing we won't know about until queens-eol when someone says, "i just upgraded to queens and now it's saying i need the port binding extension but we don't use that b/c $some_reason" | |
| 13:49:07 | mriedem | "because GLUON!" | |
| 13:49:31 | mriedem | nvm, if you're using gluon you're not using this code anyway | |
| 13:50:43 | mriedem | dansmith: i will commence the backporting | |
| 13:50:49 | dansmith | cool | |
| 13:52:16 | auggy | ratailor stephenfin Ugh, I did a bunch of work on bug #1538011 with mtreinish and came to the conclusion that the added complexity was not worth the small gain. But others may have different opinions | |
| 13:52:18 | openstack | bug 1538011 in OpenStack Compute (nova) "Metadata keys are not case sensitive" [Medium,In progress] https://launchpad.net/bugs/1538011 - Assigned to Rajesh Tailor (ratailor) | |
| 13:55:37 | auggy | But yeah, I'm not doing much in the way of openstack these days :) I'm doing more "Meta" openstack. Data science stuff around developing better ways to describe open source community activity. | |
| 13:56:05 | openstackgerrit | Ed Leafe proposed openstack/nova master: docs: Document the scheduler workflow https://review.openstack.org/475810 | |
| 13:56:10 | edleafe | stephenfin: cdent: updated ^^ | |
| 13:56:44 | stephenfin | auggy: Aye, I recall looking at the same bug and quickly dropping it. ratailor: you've been warned | |
| 13:58:17 | auggy | ratailor stephenfin I'll update it later today with the work mtreinish and I did. I abandoned a lot of patches haha | |
| 14:05:08 | auggy | Oh it looks like the patches and blueprint are linked already | |
| 14:05:13 | openstackgerrit | Matt Riedemann proposed openstack/nova master: api-ref: fix key_name note formatting https://review.openstack.org/496718 | |
| 14:05:15 | dansmith | mriedem: do you know if this person is on irc? https://review.openstack.org/#/c/496202/ | |
| 14:06:31 | mriedem | dansmith: don't think so | |
| 14:06:37 | mriedem | gibi would know but he's on vacation | |
| 14:06:44 | mriedem | ildikov might know | |
| 14:06:53 | dansmith | okay I wonder if I should just update that patch myself | |
| 14:07:33 | mriedem | is it an rc2 thing? | |
| 14:08:09 | ildikov | dansmith: mriedem: don't think so, but will try to double check | |
| 14:08:18 | dansmith | mriedem: it's another thing that is broken and needs fixing right? | |
| 14:08:48 | dansmith | I actually thought this was in the middle and a fix was on top, but I guess this is yet another broken bit | |
| 14:09:41 | mriedem | i'd have to dig into it, it's not clear what the issue is from the commit message | |
| 14:10:24 | dansmith | well you can see the assertions commented out of the test: https://review.openstack.org/#/c/496202/1/nova/tests/functional/test_servers.py | |
| 14:10:46 | dansmith | I'm not really sure why this is force specific though | |
| 14:10:50 | mriedem | dansmith: is it basically the same thing you already fixed, but different place? | |
| 14:11:02 | dansmith | maybe I should re-run this on top of everything else we have fixed recently and make sure | |
| 14:11:46 | mriedem | yeah bc you're doing it in _post_live_migration _post_live_migration | |
| 14:11:48 | mriedem | oops | |
| 14:11:51 | mriedem | https://review.openstack.org/#/c/496032/4/nova/compute/manager.py | |
| 14:12:47 | mriedem | i need to see where that live_migration_force_complete is called from | |
| 14:13:18 | mriedem | oh i guess the api | |
| 14:13:45 | mriedem | well, that just calls through to the driver to do post-copy right? so the other thread that's waiting for live migration to complete should then see it complete and eventually get to _post_live_migration, which is where your fix is | |
| 14:13:52 | mriedem | so i think your fix probably already handles this | |
| 14:16:33 | dansmith | yeah test fails now, so let me swap the assertions and make sure they're all good | |
| 14:18:39 | dansmith | yeah, it's all good now | |
| 14:18:55 | dansmith | so I'll fix it up to swap the assertions and fix up the epic long sleep thing | |
| 14:23:54 | openstackgerrit | Dan Smith proposed openstack/nova master: Add functional migrate force_complete test https://review.openstack.org/496202 | |
| 14:23:58 | dansmith | mriedem: ^ | |
| 14:24:06 | dansmith | so you want to wait on that since it's just a test I assume? | |
| 14:26:15 | mriedem | yeah i don't think we need to backport and hold up rc2 for a test | |
| 14:29:33 | mriedem | backports are all up too now https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:stable/pike+topic:bug/1712008 | |