Earlier  
Posted Nick Remark
#openstack-nova - 2019-02-19
17:05:04 mriedem those could be stable branches, checking logstash
17:05:57 smcginnis Grenade on the first one, so maybe?
17:06:04 mriedem sean-k-mooney: you hit a nova bug http://logs.openstack.org/15/620115/23/check/tempest-full/fcabb4b/controller/logs/screen-n-cpu.txt.gz?level=TRACE#_Feb_19_14_49_49_277325
17:06:17 mriedem http://status.openstack.org/elastic-recheck/#1770211
17:06:25 mriedem WARNING nova.virt.block_device [None req-2c3d004a-05e7-43d3-af5e-ccac75bed5fd tempest-AttachVolumeNegativeTest-1375016298 tempest-AttachVolumeNegativeTest-1375016298] [instance: 4b50efcf-0d26-4b01-91f0-1b8b1e9eef0f] Guest refused to detach volume aea37477-6c0c-4727-8092-da3b3d2f7c5c: DeviceDetachFailed: Device detach failed for vdb: Unable to detach the device from the live config.
17:06:45 mriedem i remember cfriesen asking about this still being an issue the other day
17:07:08 mriedem smcginnis: yeah looks like it's hitting on grenade so it's just a stable branch issue with devstack at this poin,
17:07:09 mriedem *point
17:07:10 dansmith mriedem: I saw a ping from you fly by from last night.. can you link me again so I don't have to dig out mah logs?
17:07:16 mriedem i'll work on backporting those devstack / etcd changes
17:07:33 mriedem dansmith: https://review.openstack.org/#/c/635315/1
17:10:20 dansmith mriedem: nice
17:11:18 cfriesen mriedem: smcginnis: I was hitting this:https://bugs.launchpad.net/nova/+bug/1815949
17:11:19 openstack Launchpad bug 1815949 in OpenStack Compute (nova) "missing special-case libvirt exception during device detach" [Medium,Confirmed]
17:11:39 smcginnis Hmm, that's a new one to me.
17:12:02 sean-k-mooney ok ya that makes sense. im pretty sure there are other voluem attachemtn failures like https://bugs.launchpad.net/cinder/+bug/1814916 too which are unrelated to tooz
17:12:03 mriedem cfriesen: ok that looks different than the 'unable to detach from transient / live config'
17:12:04 openstack Launchpad bug 1814916 in Cinder "MessagingTimeout in cinder-api during volume attachment update due to VolumeAttachmentNotFound in cinder-volume" [Medium,Confirmed]
17:12:19 mriedem yeah i don't know what's up with that volume attachment not found one
17:12:31 mriedem the api finds the record and then c-vol blows up b/c it's gone from the db, so clearly a race somewhere
17:12:57 cfriesen mriedem: smcginnis: but after fixing that I'm hitting the "Guest refused to detach" issue. I'm guessing it's related to stalls we're seeing in the guest, but still digging into causes.
17:12:57 mriedem probably just need someone to trace requests in a log to see when the PUT request is made and when the DELETE request is made
17:14:26 openstackgerrit Adam Spiers proposed openstack/nova master: WIP: Convert driver supported capabilities to compute node provider traits https://review.openstack.org/538498
17:15:11 aspiers mriedem, efried: I've just finished the first version of the tests which were missing for capabilities to traits: https://review.openstack.org/#/c/538498/12/nova/tests/functional/test_servers.py
17:15:28 mriedem etcd changes for devstack backported https://review.openstack.org/#/q/topic:bug/1810526+(status:open+OR+status:merged)
17:15:43 aspiers mriedem, efried: I'm sure you'll find plenty of things to improve, but hopefully this is a good base
17:16:21 aspiers not sure if I went overkill on the tests for standard vs. custom traits, and capabilities vs. update_provider_tree
17:22:42 aspiers mriedem: BTW I kept you as the author and added myself via Co-Authored-By - was that the right thing to do?
17:22:51 aspiers not sure what the convention is here
17:24:18 mriedem aspiers: yup
17:24:28 mriedem more testing > less testing
17:24:50 aspiers mriedem: also, should I put "related to blueprint expose-host-capabilities" in the commit message?
17:32:48 mriedem hmm, that's not approved https://blueprints.launchpad.net/nova/+spec/expose-host-capabilities but the change is essentially trying to solve that same problem
17:32:50 mriedem so sure
17:33:26 mriedem the spec for that bp predates traits https://review.openstack.org/#/c/222200/
17:33:35 mriedem so it was trying to solve it with something different
17:43:32 sean-k-mooney adrianc: o/ are you wokring on a new version of https://review.openstack.org/#/c/620115/
17:44:05 sean-k-mooney adrianc: if not i might start adressing stephenfin nits after i get something to eat.
17:54:27 openstackgerrit Balazs Gibizer proposed openstack/nova master: Refactor _heal_allocations_for_instance to make place for port healing https://review.openstack.org/637953
17:54:28 openstackgerrit Balazs Gibizer proposed openstack/nova master: nova-manage: heal port allocations https://review.openstack.org/637955
17:54:28 openstackgerrit Balazs Gibizer proposed openstack/nova master: Refactor _heal_allocations_for_instance (2) https://review.openstack.org/637954
18:09:22 melwitt mriedem: omg, thank you. I was completely baffled by the 0'ed out instance uuid and could not imagine how that could happen
18:12:20 melwitt enforce request context project via policy instead of hard-coded
18:12:20 melwitt alex_xu: by unrelated, I meant that project_only=True isn't related to the instance target. the only thing that project_only=True does is enforce the request context project, not the instance project. so I was thinking that changing to use instance target is an unrelated change from removing project_only=True. if we remove project_only=True without changing instance target, we get the same behavior as today except that users choose to
18:16:36 efried aspiers: Still around?
18:20:31 aspiers efried: just got back
18:20:37 efried aspiers: Sorry that took a little longer than expected (interrupted by internal work, among other things), but I got it figured out. Now I have to go around and fix all the test cases that call into the shimmed API.
18:20:46 aspiers efried: nice work :)
18:20:54 efried WIP incoming, lemme finish writing up the commit message.
18:20:56 aspiers I bet whatever you changed would have taken me a week
18:21:26 aspiers mriedem: OK thanks, I'll do that
18:21:32 openstackgerrit Eric Fried proposed openstack/nova master: WIP: De-shim PlacementFixture https://review.openstack.org/637959
18:21:41 efried aspiers: ^
18:21:43 aspiers mriedem, melwitt: is it OK to reference 2 blueprints in a single commit message?
18:21:46 aspiers efried: ack
18:22:05 efried I originally wrote it on top of your patch, to make sure your test case worked, and it did.
18:22:15 efried But this version will be lacking your delete helper.
18:22:27 efried Once I've got it polished I'll rebase it under yours.
18:22:34 efried or you can
18:22:40 aspiers Sure, I can do that
18:22:48 aspiers I have some other tweaks to go in mine anyway
18:23:09 melwitt aspiers: I don't see why not. doing that would just put the link on both lp whiteboard AFAIK
18:23:19 aspiers not to mention fixes for the glaring errors you and mriedem are about to uncover ;-)
18:23:31 aspiers melwitt: cool thanks, that's what I expected
18:25:19 aspiers woohoo, my version switcher patch landed on https://docs.openstack.org/
18:25:36 aspiers hopefully it doesn't break anything
18:25:43 aspiers s/version/release/
18:40:24 openstackgerrit Boris Bobrov proposed openstack/nova master: Fallback to the first element if no nodes match https://review.openstack.org/637965
19:19:54 openstackgerrit Merged openstack/nova master: api-ref: warn about changing/unsetting AZ name with instances https://review.openstack.org/635315
19:29:41 openstackgerrit melanie witt proposed openstack/nova master: Add online data migration for populating user_id https://review.openstack.org/633351
19:29:41 openstackgerrit melanie witt proposed openstack/nova master: Add user_id field to InstanceMapping https://review.openstack.org/633350
19:35:24 openstackgerrit Merged openstack/nova master: api-ref: mention policy defaults for aggregates https://review.openstack.org/635317
19:37:10 efried mriedem: easy +A if you please https://review.openstack.org/#/c/637657/
19:45:43 mriedem +W
19:50:11 efried thanks mriedem
19:57:42 efried aspiers: I had to give up. Your fix is gateward.
20:28:34 openstackgerrit Merged openstack/nova master: Fix fake DELETE in PlacementFixture https://review.openstack.org/637657
20:42:25 openstackgerrit Merged openstack/os-vif master: Change python3.5 job to python3.7 job on Stein+ https://review.openstack.org/610068
21:07:56 mriedem rom error)
21:07:56 mriedem i wonder how many operators are trying to recover an instance from a failed resize confirm/revert where the instance is in ERROR status and they fixed something in the compute and just want to retry the confirm/revert operation, but can't b/c the API doesn't allow those operations on an instance in ERROR status. seems weird that the os-resetStatus server action API doesn't allow you to put an instance in 'resized' status agai
21:08:10 mriedem dansmith: ever thunk on that?^
21:08:48 dansmith well, it would be pretty bad if you put it into resized status when it wasn't,
21:08:58 dansmith and it'd be hard to tell when and when not to do something like that
21:09:29 mriedem if you got through revert_resize on the dest but finish_revert_resize fails on the source, you can maybe hard reboot your way out of it
21:09:41 openstackgerrit Boris Bobrov proposed openstack/nova master: Don't reference unassigned variable in xml matchers in test https://review.openstack.org/637965
21:09:42 mriedem if revert_resize on the dest fails, you're kind of stuck there now
21:09:48 mriedem anywho
21:10:06 dansmith right but is it realistic for an operator to know that it failed on finish_revert_resize?
21:10:09 dansmith even if they fix something,
21:10:28 dansmith they could reset into resized state from another place where they think it's likely to help and further break stuff
21:10:42 dansmith I mean, I know what you're saying, but.. it's a dangerous tool
21:10:56 mriedem sure, and i wouldn't expect most operators to know that unless they were pretty knowledgeable on low level nova
21:11:06 mriedem which their time is probably just better spent on rebuilding the vm
21:11:16 artom And, is it needed? I mean, if they have data they want to keep on a volume, they can just attach that to a new instance with the correct flavor?
21:11:18 mriedem and writing an apology to the user
21:11:51 mriedem artom: i'm thinking more about local storage, no volumes
21:12:08 artom So actually, is a resize destructive in that respect?
21:12:09 mriedem granted your data should be stored on a persistent volume
21:12:31 mriedem no, the local disks are left intact on the source host when you resize
21:12:46 mriedem they are destroyed when you confirm the resize to keep the instance on the dest host

Earlier   Later