| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-04-07 | |||
| 18:32:07 | mriedem | efried: thorst: we don't translate log messages anymore | |
| 18:32:12 | mriedem | so you can remove the _LI usage in there | |
| 18:32:15 | openstackgerrit | Vladik Romanovsky proposed openstack/nova master: Support HTTP CONNECT in websocket proxy https://review.openstack.org/454226 | |
| 18:34:25 | efried | mriedem Roger that. | |
| 18:37:38 | nicodemus_ | hello again... I'm having some trouble with a few instances in a Mitaka deploy. In short, a few instances were migrated from one compute node to another, but then a DB issue happened and I had to restore the DB to a point prior to the migration. | |
| 18:38:41 | nicodemus_ | An instance in compute "A" is running but has no networking setup, so I modified the instances table to reflect that it's in compute A (DB said it was on compute B) | |
| 18:39:23 | nicodemus_ | So, I'm trying to migrate it from compute A to compute B but have an "InstanceActionNotFound" error on nova-api logs :( | |
| 18:40:11 | nicodemus_ | this is the full trace: http://paste.openstack.org/show/605774/ | |
| 18:40:28 | nicodemus_ | has anyone come across such issue? | |
| 18:45:32 | sdague | mriedem: btw, just to see, I inverted the livesnapshot flag - https://review.openstack.org/#/c/454323/ | |
| 18:45:46 | sdague | 2 different tempest runs failed on snapshotting paused images | |
| 18:45:59 | sdague | so... I think that remains a can of worms | |
| 18:47:45 | mriedem | those could just be latent bugs | |
| 18:47:58 | mriedem | we likely don't have a check in the libvirt driver to not live snapshot a paused guest | |
| 18:48:17 | mriedem | could probably just open a bug for that | |
| 18:48:57 | mriedem | yeah we get the power state here https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L1541 | |
| 18:49:05 | mriedem | and determine if it's live snapshot here https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L1552 | |
| 18:49:07 | mriedem | but don't check the state | |
| 18:49:18 | mriedem | here we check the state https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L1571 | |
| 18:49:20 | mriedem | later | |
| 18:49:59 | mriedem | https://github.com/openstack/nova/blob/master/nova/compute/power_state.py#L34 | |
| 18:50:10 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: spawn/delete #1: no-ops https://review.openstack.org/438119 | |
| 18:50:11 | efried | mriedem ^^ back atcha. Thanks! | |
| 18:50:31 | efried | thorst esberglu adreznec ^^ FYI and +1 | |
| 18:50:49 | openstackgerrit | Anusha Unnam proposed openstack/nova master: Add description to policies in security_groups.py https://review.openstack.org/452309 | |
| 18:53:29 | mriedem | sdague: guessing maybe something with this? https://github.com/openstack/nova/blob/master/nova/virt/libvirt/driver.py#L1661 | |
| 18:53:37 | mriedem | pause vs suspend wrt qemu confuses me though | |
| 18:55:22 | mriedem | http://logs.openstack.org/23/454323/1/check/gate-tempest-dsvm-neutron-full-ubuntu-xenial/07a0fe4/logs/screen-n-cpu.txt.gz?level=TRACE#_2017-04-06_20_40_25_285 | |
| 18:55:29 | mriedem | trying to snapshot a non-running instance: (state: 3 expected: 1) | |
| 18:56:24 | mriedem | libvirtError: Domain not found: no domain with matching uuid 'f2e92033-2fc7-44c4-8dda-70ce75620071' (instance-00000014) | |
| 18:57:24 | sdague | mriedem: well, it works some of the time | |
| 18:58:00 | mriedem | sdague: i don't know what that means | |
| 18:58:08 | mriedem | you haven't rechecked this | |
| 18:58:13 | mriedem | you mean it passes in some jobs? | |
| 18:59:34 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: spawn/destroy #2: functional https://review.openstack.org/438598 | |
| 18:59:47 | mriedem | efried: doh | |
| 18:59:51 | mriedem | i was in the middle of that one | |
| 19:00:04 | efried | mriedem Well, figured I should get rid of the log translations. | |
| 19:00:22 | efried | Sorry about that. | |
| 19:00:24 | openstackgerrit | Dan Smith proposed openstack/nova master: Make discover_hosts only query for unmapped ComputeNode records https://review.openstack.org/427903 | |
| 19:00:25 | openstackgerrit | Dan Smith proposed openstack/nova master: Add ComputeNode.mapped field https://review.openstack.org/427901 | |
| 19:00:25 | openstackgerrit | Dan Smith proposed openstack/nova master: Add ability to query for ComputeNodes by their mapped value https://review.openstack.org/427902 | |
| 19:00:53 | sdague | mriedem: it does pass some jobs | |
| 19:02:03 | mriedem | sdague: do those jobs run that test? | |
| 19:02:11 | sdague | yes | |
| 19:02:19 | sdague | it's passing another tempest run | |
| 19:02:19 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Change livesnapshot to true by default https://review.openstack.org/454323 | |
| 19:02:38 | mriedem | which one? the live migration job? | |
| 19:03:20 | sdague | mriedem: oh, you know what, maybe it isn't | |
| 19:03:21 | openstackgerrit | Eric Fried proposed openstack/nova master: PowerVM Driver: power_on/off and reboot https://review.openstack.org/427380 | |
| 19:03:24 | mriedem | sdague: right | |
| 19:03:29 | mriedem | so i pushed up the paused check | |
| 19:03:32 | mriedem | let's see what happens | |
| 19:04:39 | sdague | mriedem: ok, so the flake8-import-order completely is returning the wrong answers during my inbuffer file checking. Can I delete this thing. | |
| 19:05:09 | mriedem | i don't know what that means either | |
| 19:05:18 | mriedem | wrong answers, as in false negatives? | |
| 19:06:12 | sdague | yes | |
| 19:06:31 | sdague | it keeps saying nova needs to be before some other thing that it shouldn't be | |
| 19:06:35 | openstackgerrit | Sean Dague proposed openstack/nova master: remove flake8-import-order https://review.openstack.org/454853 | |
| 19:06:36 | openstackgerrit | Sean Dague proposed openstack/nova master: remove log translation tags from nova.cells https://review.openstack.org/454854 | |
| 19:08:04 | mriedem | do you have this change? https://github.com/openstack/nova/commit/07a042ba2c3355b4481ee976894834e819dc9710 | |
| 19:08:18 | sdague | yes, it's way after that change | |
| 19:08:28 | sdague | I don't know why it's giving me the entirely wrong answers, but it is | |
| 19:09:26 | openstackgerrit | Anusha Unnam proposed openstack/nova master: fix admin context scenarios for cinderclient https://review.openstack.org/384799 | |
| 19:17:02 | sdague | mriedem: https://dl.dropboxusercontent.com/1/view/wrhl4v895bc5rdr/Apps/Shutter/screenshot_368.png | |
| 19:17:15 | dansmith | MY EYES | |
| 19:17:20 | sdague | dansmith: :P | |
| 19:18:59 | mriedem | it's probably the import nova.conf | |
| 19:19:23 | mriedem | can't distinguish between from nova.* and import nova.*? | |
| 19:19:45 | mriedem | efried: comments in the 2nd patch | |
| 19:21:08 | sdague | mriedem: right, I don't know why it does it here, and not via tox runs | |
| 19:21:40 | cdent | sdague: is your in buffer stuff running out of the same venv? | |
| 19:21:46 | sdague | cdent: yep | |
| 19:21:49 | cdent | weird | |
| 19:22:29 | sdague | but, the whole point of doing pep8 things was to catch real errors and not make folks have to debug their local envs | |
| 19:22:36 | sdague | so I really want to get rid of this thing | |
| 19:22:49 | sdague | especially because we actually implemented that in hacking years ago, and deleted it | |
| 19:22:51 | cdent | yeah, how did it manage to show up? It's relatively new, yeah? | |
| 19:22:56 | sdague | it's new | |
| 19:25:42 | sdague | mriedem: it's not in test-requirements | |
| 19:25:44 | mriedem | sdague: cdent: while you're both here, | |
| 19:25:56 | mriedem | nova-api under httpd in devstack | |
| 19:26:11 | cdent | ah yes | |
| 19:26:20 | mriedem | sdague: i guess you're right https://github.com/openstack/nova/commit/a5d4adbd60291adc2eaaefdf76186edf74523ece | |
| 19:26:28 | mriedem | sdague: cdent: so at the ptg you guys said you'd help with that | |
| 19:26:28 | cdent | that's on our lists | |
| 19:26:36 | mriedem | my ass is grass come p1 | |
| 19:26:47 | mriedem | EmilienM is going to be all over me | |
| 19:27:00 | openstackgerrit | ayoung proposed openstack/nova master: Admin API Policy enforcement contingent on is_admin_project https://review.openstack.org/384148 | |
| 19:27:07 | EmilienM | I'm not here next week | |
| 19:27:11 | cdent | I can probalby do it next weeek then | |
| 19:27:12 | mriedem | whew | |
| 19:27:16 | sdague | mriedem: the first thing that actually should be done is get placement under uwsgi | |
| 19:27:24 | cdent | can probably do that too | |
| 19:27:38 | cdent | I'm fairly unscheduled for a very short while | |
| 19:27:47 | mriedem | i'm asking b/c i think this came up in the api subteam meeting a couple of weeks ago and i wasn't sure if there was a plan | |
| 19:28:03 | cdent | i'm not aware of a plan, but I think we can probably make one quickly, and make it so | |
| 19:28:03 | mriedem | and we still have that service version bug for the api | |
| 19:28:07 | sdague | cdent: we should really do this under uwsgi and not apache so that the systemd stuff is more useful | |
| 19:28:20 | mriedem | systemd systemd systemd | |