| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-07-20 | |||
| 15:28:59 | openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/os-vif master: Read datapath_type from VIF object https://review.openstack.org/474914 | |
| 15:28:59 | sdague | I'm going to figure out test failures first | |
| 15:29:00 | jaypipes | mriedem: nope | |
| 15:29:11 | mriedem | ok i'll get the req update in and then release it | |
| 15:29:17 | jaypipes | mriedem: danke | |
| 15:33:14 | mriedem | sdague: another thing to consider, report a bug for this request id logging thing and then we can tag it pike-rc-potential for tracking | |
| 15:33:18 | mriedem | i'm starting to tag things for the rc | |
| 15:35:11 | jaypipes | sean-k-mooney: ralonsoh's https://review.openstack.org/#/c/474914/ is good to go. feel free to +2 if agree. | |
| 15:35:19 | openstackgerrit | OpenStack Proposal Bot proposed openstack/nova master: Updated from global requirements https://review.openstack.org/485634 | |
| 15:35:41 | ralonsoh | jaypipes, sean-k-moone: thanks! | |
| 15:35:53 | mriedem | jaypipes: sean-k-mooney: ralonsoh: does ^ need to be in an os-vif release for pike? | |
| 15:35:57 | mriedem | because we released os-vif yesterday | |
| 15:36:11 | jaypipes | mriedem: I don't believe so, no. | |
| 15:36:25 | mriedem | ok, it's also dependent on a neutron change with -1s and test failures | |
| 15:36:37 | jaypipes | mriedem: right. | |
| 15:36:59 | jaypipes | mriedem: os-vif was good yesterday when it was released. | |
| 15:37:01 | ralonsoh | mriedem, jaypipes: well, this patch solves a bug for OVS userspace | |
| 15:37:21 | mriedem | ralonsoh: at this point it's not going to get released for pike | |
| 15:37:28 | mriedem | unless it's later backported | |
| 15:37:31 | ralonsoh | mriedem; no, not now | |
| 15:37:51 | sdague | mriedem: sure | |
| 15:38:25 | jaypipes | ralonsoh: as mriedem said, there is a dependent neutron patch that would need to be merged for that os-vif patch to even be workable though. | |
| 15:38:26 | sean-k-mooney | mriedem: need no nice to have definetly | |
| 15:39:07 | jaypipes | ralonsoh: well, I shouldn't say "even be workable". I meant "even be useful". | |
| 15:39:35 | ralonsoh | jaypipes: yes, this bug needed changes in 4 projects. I don't think is going to land in Pike | |
| 15:39:48 | jaypipes | cdent: hey, just to repeat what I said on that bug report... really excellent investigative work on that sorted(sharing_providers.items()..) thing. | |
| 15:40:08 | jaypipes | cdent: I had spent hours looking at the SQL generated going to myself "there's nothing wrong with this SQL... :(" | |
| 15:40:18 | cdent | thanks jaypipes, it was really because of gibi | |
| 15:40:21 | jaypipes | cdent: and your comment sparked a renewed look at the JOIN order. | |
| 15:40:31 | cdent | he had pasted up things to look at and started wondering “why is that different" | |
| 15:40:37 | sean-k-mooney | mriedem: it would be good to backport https://review.openstack.org/#/c/474914 all the way to ocata but not sure if we can due to ovo chagnes | |
| 15:40:47 | cdent | also having the gabbi test (which makes things easier for me to toy with) helped | |
| 15:40:48 | jaypipes | well, thank you cdent and gibi both :) | |
| 15:41:23 | jaypipes | I have a feeling that little patch may fix some of the odd random NoValidHosts failures we were seeing intermittently. | |
| 15:41:58 | jaypipes | since it produces an incorrect false negative on providers meeting requested resources when including shared resources. | |
| 15:43:22 | cdent | jaypipes: you’ve seen today’s bug that gibi found, yes? | |
| 15:43:33 | cdent | I’ve now become expectant of him finding one every day | |
| 15:43:52 | jaypipes | cdent: :( | |
| 15:44:02 | jaypipes | cdent: you mean the "init alloc_reqs earler" one?> | |
| 15:44:05 | cdent | yeah | |
| 15:44:25 | cdent | It’s not :( it’s :). Find them now when they are fresh and new before they go hurt someone. | |
| 15:44:41 | mriedem | final os-traits release https://review.openstack.org/485722 | |
| 15:44:43 | mriedem | bauzas: ^ | |
| 15:45:07 | bauzas | mriedem: looking | |
| 15:47:08 | gibi | cdent, jaypipes: I guess I will fail on the bug / day expectation as I will be out tomorrow | |
| 15:47:24 | cdent | I will try to stand in for you | |
| 15:47:28 | jaypipes | gibi: that's no excuse. we expect at least one bug a day now. | |
| 15:47:41 | gibi | cdent, jaypipes: however there is one more for today https://bugs.launchpad.net/nova/+bug/1705487 | |
| 15:47:42 | openstack | Launchpad bug 1705487 in OpenStack Compute (nova) "placement logs an ERROR when PUT /allocation result in an invalid inventory" [Undecided,New] | |
| 15:48:18 | gibi | cdent, jaypipes: I'm hesitant if this is just a logging thing or something a bit deeper | |
| 15:48:26 | cdent | gibi: yay! more bugly goodness. | |
| 15:48:40 | jaypipes | gibi: looking ... | |
| 15:49:38 | gibi | jaypipes, cdent: If I'm not mistaken then the exception is coming from here https://github.com/openstack/nova/blob/master/nova/objects/resource_provider.py#L1607 | |
| 15:50:15 | gibi | jaypipes: can be that it is connected to the fact that I'm using shared custom resource | |
| 15:50:53 | jaypipes | gibi: do you have a test case for that newest bug? I'm not entirely sure how an invalid resource class can be allocated by the scheduler, but perhaps it could. | |
| 15:51:18 | gibi | jaypipes: not yet unfortunately and I have to leave soon. | |
| 15:51:40 | jaypipes | gibi: ah, I see now... it's a race condition that we should retry on instead of logging an exception I guess | |
| 15:52:31 | gibi | jaypipes: I can try to craft the test case on Monday | |
| 15:52:49 | jaypipes | gibi: cool. enjoy your long weekend! :) | |
| 15:53:11 | gibi | jaypipes: thanks! | |
| 16:00:54 | openstackgerrit | Jay Pipes proposed openstack/nova master: Remove improper LOG.exception() calls in placement https://review.openstack.org/485726 | |
| 16:00:56 | jaypipes | gibi: fixed already :) ^^ | |
| 16:16:32 | mriedem | +2 | |
| 16:19:46 | openstackgerrit | Stephen Finucane proposed openstack/nova master: DNM: Try to figure out what the tenant is returning https://review.openstack.org/485733 | |
| 16:20:59 | mriedem | would be nice to get some eyes on this https://review.openstack.org/#/c/446671/ | |
| 16:21:24 | mriedem | we've got 3 patches left for the cinder attachment stuff | |
| 16:21:30 | mriedem | and this is probably the least complicated | |
| 16:22:43 | dfisher | mriedem: is this for multi-attach? | |
| 16:23:05 | mriedem | it's for using the new cinder 3.27 attachments api which is a prereq for multiattach | |
| 16:23:09 | openstackgerrit | Matthew Booth proposed openstack/nova master: Split Compute.errors_out_migration into a separate contextmanager https://review.openstack.org/485734 | |
| 16:23:47 | dfisher | mriedem: i assume with all this changing, https://bugs.launchpad.net/nova/+bug/1527623 may get another look? | |
| 16:23:48 | openstack | Launchpad bug 1527623 in OpenStack Compute (nova) "Nova might orphan volumes when it's racing to delete a volume-backed instance" [Medium,In progress] - Assigned to ChangBo Guo(gcb) (glongwave) | |
| 16:24:31 | mriedem | w/o loading a bunch of context into my head right now i can't say | |
| 16:25:04 | openstackgerrit | Ildiko Vancsa proposed openstack/nova master: Implement new attach Cinder flow https://review.openstack.org/330285 | |
| 16:27:01 | oorgeron | Hi folks, I'd like to get https://review.openstack.org/#/c/446643/ reviewed. Anyone around to help? | |
| 16:27:59 | mriedem | oorgeron: that's not something we need to spend time reviewing when we're 7 days from the pike feature freeze | |
| 16:28:01 | mriedem | and is not a bug fix | |
| 16:28:24 | dansmith | +1 | |
| 16:28:42 | oorgeron | so you want me to shift it to blueprint or rfe? | |
| 16:28:50 | mriedem | it's already got a blueprint tied to it | |
| 16:28:59 | mriedem | so bring it up when we start talking about blueprints for queens | |
| 16:29:02 | mriedem | in a few weeks please | |
| 16:29:43 | openstackgerrit | Sean Dague proposed openstack/nova master: WIP: request_log addition for running under uwsgi https://review.openstack.org/485602 | |
| 16:31:54 | oorgeron | mriedem: I'll hold off on it for now. | |
| 16:32:09 | mriedem | thanks | |
| 16:32:32 | mriedem | if there are going to be oracle people at the ptg in denver, i think it would also be good to bring this up as a topic in a cross-project session | |
| 16:32:35 | mriedem | since this is more than just nova | |
| 16:32:39 | mriedem | and all projects should be on the same page | |
| 16:32:58 | mriedem | otherwise you get 50% of the projects accepting this and others don't and you're stuck | |
| 16:33:02 | mriedem | oorgeron: ^ | |
| 16:33:07 | oorgeron | mriedem: it's already in other projects and most don't require any changes beyond having oslo.db 4.24 or above. | |
| 16:33:25 | jaypipes | oorgeron: wish I had been aware of https://review.openstack.org/#/c/427970/ :( | |
| 16:33:47 | mriedem | this is what i'm talking about | |
| 16:33:58 | mriedem | poorly communicated crosss-project effort | |
| 16:42:16 | TheJulia | mriedem: following up from last night re ironic bfv, we have achieved a successful boot, and teardown, at least as far as nova is concerned. Our test vm never booted back up for disk cleaning. hshiina has volunteered to take a stab at fixing the unit tests since the tooling doesn't make creating a wrapped object easy. | |
| 16:42:49 | mriedem | ildikov: comments in https://review.openstack.org/#/c/446671/ | |
| 16:42:59 | mriedem | ildikov: only thing i worry about is the VolumeUnattached case | |
| 16:43:23 | mriedem | TheJulia: can't just use bdm._bdm_obj.is_volume? | |
| 16:43:56 | mriedem | the unit test should just pass in a block_device_info['mappings'] = [DriverVolumeBlockDevice(BlockDeviceMapping)] | |
| 16:44:15 | TheJulia | Well, if that works, then that should be a stupidly quick fix | |