| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-01-31 | |||
| 15:03:51 | sfinucan | derekh: Hmm, much as I hate to say it, could you move that patch /after/ the bug patch? | |
| 15:04:31 | sfinucan | I realize that might mean using Mox, but I'm not sure if we can merge a Mox-mock conversion patch first now...? | |
| 15:04:55 | sfinucan | Let's see what johnthetubaguy and/or dansmith think first though | |
| 15:05:45 | johnthetubaguy | sfinucan: I was eyeing that one up, I can try squash it | |
| 15:05:50 | johnthetubaguy | dansmith: one sec, checking | |
| 15:06:32 | johnthetubaguy | dansmith: https://review.openstack.org/#/c/427009/ | |
| 15:06:50 | dansmith | johnthetubaguy: ah okay | |
| 15:09:55 | mdbooth | dansmith: I've had 2 people call me on the N, N-1 upgrade thing in 2 days now: https://review.openstack.org/#/c/423392/ | |
| 15:10:24 | mdbooth | dansmith: Can you confirm I am/am not over-assuming there? | |
| 15:11:02 | dansmith | mdbooth: I think they want you to use versions to enforce the behavior and not just assume right? | |
| 15:17:15 | mriedem | sean-k-mooney: jaypipes: sfinucan: i'm going to put up a releases patch to cut stable/ocata for os-vif today unless you are aware of anything we have to get merged before that happens | |
| 15:17:32 | sfinucan | mriedem: Nothing on my end | |
| 15:18:32 | openstackgerrit | Maciej Szankin proposed openstack/nova: conf: Resolve TODOs in 'database' https://review.openstack.org/393695 | |
| 15:23:26 | sean-k-mooney | mriedem: i dont think anything else needs to be added before the release | |
| 15:23:43 | jaypipes | mriedem: go for it. | |
| 15:27:15 | mriedem | ok, thanks. will do after morning meetings. | |
| 15:28:17 | mdbooth | mriedem: Do we have a concept of a minimum supported service version? | |
| 15:40:20 | mdbooth | johnthetubaguy: Can we hold off on the libvirt+xen thing temporarily? Concerned about adding a quick hack for something that has been broken a long time. Quick hacks tend to last a long time, and are very expensive. Looking now. | |
| 15:41:05 | mdbooth | If this is causing a problem in the gate, could we disable the test again? | |
| 15:43:13 | johnthetubaguy | mdbooth: we could, just seems like a big gain not to have | |
| 15:44:04 | mdbooth | johnthetubaguy: I'm not definitely saying we shouldn't take the patch, just that I'd like to look first at why we need it, especially as ftersin was also concerned about it. | |
| 15:44:35 | johnthetubaguy | mdbooth: you got the link again? | |
| 15:44:40 | mdbooth | Problem is it's adding a new foo(x) returns y normally, but z in this other bizarre situation which nobody will every think of. | |
| 15:44:54 | mdbooth | https://review.openstack.org/#/c/402318/1/nova/virt/libvirt/blockinfo.py | |
| 15:44:55 | johnthetubaguy | I am just cruzing all bugs with a +2 already | |
| 15:45:17 | mdbooth | We have lots of these, and they're just edge-case bug factories. | |
| 15:47:34 | mdbooth | johnthetubaguy: My immediate thought, btw, is your suggestion to fix it during spawn | |
| 16:03:08 | openstackgerrit | Lee Yarwood proposed openstack/nova: libvirt: Limit destroying disks during cleanup to spawn https://review.openstack.org/427267 | |
| 16:05:27 | lyarwood | dansmith, mdbooth ; ^ I'd appreciate your thoughts on https://review.openstack.org/427267 if you have time today | |
| 16:05:48 | mdbooth | lyarwood: It just set my spidey sense tingling :) Was about to look. | |
| 16:06:30 | lyarwood | mdbooth: thanks | |
| 16:08:52 | derekh | sfinucan: sorry was on a call, so do I need to swap the patches around? | |
| 16:09:09 | sfinucan | derekh: No, I think johnthetubaguy is just going to squash them | |
| 16:09:21 | derekh | sfinucan: ok, cool | |
| 16:10:22 | mdbooth | lyarwood: Oof, so a failed reboot would delete the instance? | |
| 16:10:56 | lyarwood | mdbooth: well, failure to plug the vifs during a reboot would delete any local files, yeah | |
| 16:11:06 | mdbooth | Eesh, that's baaaad | |
| 16:13:24 | mdbooth | lyarwood: Compare this change for context: https://review.openstack.org/#/c/389593/2/nova/virt/libvirt/driver.py | |
| 16:13:35 | dansmith | lyarwood: oof, that's from me huh? | |
| 16:13:59 | dansmith | although it's just because I'm calling cleanup where we weren't before | |
| 16:14:10 | mdbooth | I'm -1 on ^^^ because it superficially probably passes the cases we can think of right now, but it depends on the phase of the moon. | |
| 16:14:24 | mdbooth | If we took your change and ^^^, rbd disks would be deleted anyway | |
| 16:15:26 | mdbooth | That's just a comment on why I'm against that sort of thing in general, btw, and I want us to start fixing root causes. | |
| 16:15:51 | dansmith | mdbooth: is this demonstrably worse than what we have today? because I don't think it is, and the bug is pretty bad... | |
| 16:15:58 | mdbooth | At first glance, I think your patch effectively specifies the caller's intent. | |
| 16:16:31 | mdbooth | dansmith: I'm not being clear, I think the *other* patch is bad, and lyarwood's patch is an example of why. | |
| 16:17:47 | mdbooth | dansmith: Sorry, was internally also trying to justify being against https://review.openstack.org/#/c/402318/ | |
| 16:19:48 | lyarwood | dansmith: yeah AFAICT we just didn't make it to cleanup before your patch | |
| 16:20:20 | dansmith | so, to restate, I think we should merge lyarwood's patch | |
| 16:20:50 | lyarwood | cool thanks | |
| 16:20:51 | mdbooth | dansmith: Yeah. Reviewing now. | |
| 16:34:53 | mriedem | sdague: did you see this? http://lists.openstack.org/pipermail/openstack-dev/2017-January/111137.html | |
| 16:35:07 | mriedem | sdague: i was wondering if the new devstack-tools stuff somehow screwed up local.conf for multinode grenade? | |
| 16:35:43 | sdague | mriedem: it's possible, though I thought I had that covered | |
| 16:36:22 | mriedem | we know d-g is setting it http://logs.openstack.org/61/417961/40/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/8c363cd/console.html.gz#_2017-01-27_23_16_09_271784 | |
| 16:36:26 | mriedem | echo 'export LIVE_MIGRATION_AVAILABLE=True' | |
| 16:36:26 | sdague | mriedem: where is that job definition | |
| 16:37:13 | sdague | is that coming from project-config or d-g? | |
| 16:37:13 | mriedem | sdague: https://github.com/openstack-infra/project-config/blob/master/jenkins/jobs/devstack-gate.yaml#L1674 | |
| 16:37:22 | mriedem | echo 'export LIVE_MIGRATION_AVAILABLE=True' is from d- | |
| 16:37:24 | mriedem | *d-g | |
| 16:37:40 | mriedem | sdague: https://github.com/openstack-infra/devstack-gate/blob/master/devstack-vm-gate.sh#L542 | |
| 16:38:09 | mriedem | sdague: the weird thing i noticed was, | |
| 16:38:12 | mriedem | looking at the local.conf http://logs.openstack.org/61/417961/40/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/8c363cd/logs/new/local.conf.txt.gz | |
| 16:38:20 | mriedem | NOVA_ALLOW_MOVE_TO_SAME_HOST=False is in there | |
| 16:38:25 | mriedem | from https://github.com/openstack-infra/devstack-gate/blob/master/devstack-vm-gate.sh#L541 | |
| 16:38:34 | mriedem | but export LIVE_MIGRATION_AVAILABLE=True isn't | |
| 16:38:40 | mriedem | so i'm wondering if the export is screwing it up | |
| 16:40:46 | sdague | yeh, it's litterally using the export bit | |
| 16:41:25 | sdague | mriedem: you want to do a depends on https://review.openstack.org/427282 and see if that addresses things? | |
| 16:42:04 | mriedem | sure | |
| 16:42:06 | sdague | I took a strategy of doing a merge for valid looking lines, so you don't just have clobber clobber clobber going on, but that means understanding the kind of lines that can be brought over | |
| 16:43:09 | mdbooth | lyarwood: Done. | |
| 16:44:34 | sdague | mriedem: yeh - http://logs.openstack.org/61/417961/40/check/gate-grenade-dsvm-neutron-multinode-live-migration-nv/8c363cd/logs/grenade.sh.txt.gz#_2017-01-27_23_16_19_637 | |
| 16:45:07 | openstackgerrit | Matt Riedemann proposed openstack/nova: DNM: testing grenade multinode + live migration fix https://review.openstack.org/427285 | |
| 16:45:41 | mriedem | skipped LIBS_FROM_GIT since there was no valid right? | |
| 16:45:43 | lyarwood | mdbooth: thanks, I'll sort that out now | |
| 16:48:30 | sdague | mriedem: yeh | |
| 16:48:35 | sdague | because there was no value | |
| 16:48:49 | mriedem | sdague: ok. pointed out OS_NO_CACHE in the review too | |
| 16:48:55 | mriedem | otherwise lgtm | |
| 16:49:00 | mriedem | got the nova change up to test it | |
| 16:51:02 | sdague | mriedem: honestly, I'd leave the OS_NO_CACHE bit for now. I get it's probably also wrong, but it's not impacting anything right now. | |
| 16:51:13 | sdague | and, I'm not convinced it does anything useful anymore anyway | |
| 16:52:25 | mriedem | yeah hmm, looks like that's all dead code http://git.openstack.org/cgit/openstack-dev/devstack/tree/openrc#n56 | |
| 16:53:27 | mriedem | lots of stuff setting it, but nothing using it http://codesearch.openstack.org/?q=OS_NO_CACHE&i=nope&files=&repos= | |
| 16:59:44 | sdague | mriedem: yeh | |
| 17:00:04 | sdague | there were definitely a few things I noticed in d-g that were completely dead config settings | |
| 17:12:36 | openstackgerrit | Lee Yarwood proposed openstack/nova: libvirt: Limit destroying disks during cleanup to spawn https://review.openstack.org/427267 | |
| 17:14:07 | openstackgerrit | Chris Dent proposed openstack/nova: WIP: Explore functional test for the resource tracker https://review.openstack.org/427309 | |
| 17:15:04 | mgould | hi everyone | |
| 17:15:06 | mgould | I'm getting "Host is not mapped to any cell" errors in nova boot on a clean devstack on a freshly-installed baremetal machine | |
| 17:15:28 | mgould | I've previously used the same local.conf on a VM without issues | |
| 17:15:34 | mgould | any idea what I've done wrong? | |
| 17:15:34 | openstack | bug 1646896 in OpenStack Compute (nova) "System hangs when using NFS storage backend with loopback mounts" [Low,Confirmed] https://launchpad.net/bugs/1646896 | |
| 17:15:34 | openstackgerrit | Mihail proposed openstack/nova: proposed fix for bug 1646896 - System hangs when using NFS storage backend with loopback mounts, Closes-Bug: #1646896 https://review.openstack.org/422041 | |
| 17:16:07 | cdent | jaypipes, edleafe, johnthetubaguy, mriedem : that review ^ about functional test for resource provider is trying to the basis for one that tests ironic. at the moment is just does non ironic | |
| 17:16:14 | cdent | I needed to start there to get any traction | |
| 17:17:00 | jaypipes | cdent: hmm, I was right in the middle of adding a functional test to the Ironic resource classes patch. | |
| 17:17:18 | cdent | ah, good, now we'll have more | |