| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-11-30 | |||
| 19:33:53 | jaypipes | efried: rebased and fixed conflicts up to ^ | |
| 19:34:44 | jaypipes | efried: I stopped there because I don't know which of alex and tetsuro's add func tests patches are duplicated and which are real. | |
| 19:35:00 | efried | jaypipes ack | |
| 19:56:55 | melwitt | mriedem: turns out your test just always passes locally for me (on top of my celldatabases patch) even without the periodic task fix. so something must be afoot if you're getting a different result locally than me | |
| 19:58:47 | mriedem | melwitt: at some point i rebased out your change since it wasn't fixing stuff, that's when i started writing that workaround in start_service | |
| 19:59:13 | melwitt | oh, okay | |
| 20:01:19 | melwitt | if you still had my change when you fixed the wrong assert you mentioned earlier and it didn't pass, then that's really unexpected | |
| 20:04:30 | mriedem | unrelated, but if you use gerrit-dash-creator for your nova review dashboard, it's been fixed for zuul https://github.com/openstack/gerrit-dash-creator/blob/master/dashboards/nova.dash | |
| 20:04:42 | mriedem | i was wondering why my dashboard was showing me changes with a -2 as ready for final +2 | |
| 20:05:09 | openstackgerrit | Jackie Truong proposed openstack/python-novaclient master: Microversion 2.57 - Add trusted_image_certificates https://review.openstack.org/500396 | |
| 20:05:37 | melwitt | but I need to re-create my dashboard right? my old dashboard doesn't work anymore | |
| 20:05:56 | mriedem | correct | |
| 20:06:06 | mriedem | you have to git clone gerrit-dash-creator | |
| 20:06:15 | mriedem | and then: tox -e venv -- gerrit-dash-creator dashboards/nova.dash | |
| 20:06:15 | melwitt | okay, thanks | |
| 20:06:29 | mriedem | i put the url in a shortener and then put that in my gerrit settings | |
| 20:07:33 | mriedem | efried: seen this https://review.openstack.org/#/c/521764/ ? | |
| 20:07:36 | melwitt | sweet. good tips | |
| 20:18:19 | openstackgerrit | Mike Lowe proposed openstack/nova master: live-mig: keep disk device address same https://review.openstack.org/518022 | |
| 20:28:05 | mriedem | dansmith: melwitt: cellsv2 thingamajig https://review.openstack.org/#/c/521837/5 | |
| 20:28:10 | mriedem | ux stuff | |
| 20:32:33 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Update description of Rebuild in server_concepts.rst https://review.openstack.org/522086 | |
| 20:41:20 | openstackgerrit | Matt Riedemann proposed openstack/nova master: [placement]Enhance doc for placement allocation list https://review.openstack.org/511715 | |
| 20:49:55 | openstackgerrit | Merged openstack/os-traits master: Updated from global requirements https://review.openstack.org/520477 | |
| 20:50:55 | efried | mriedem Nope, hadn't seen it. Thanks for the pointer. | |
| 20:52:53 | melwitt | mriedem: yeah, seems like it would be uncool of us to remove 'nova-manage host list' before we have another way to list hosts in a cell, even though it's deprecated | |
| 20:54:38 | melwitt | I was thinking discover_hosts had a --dry-run mode that just shows the hosts. but looks like not | |
| 20:54:45 | efried | jaypipes Care to eyeball https://review.openstack.org/#/c/524263/ ? Looks like we may want to backport it (mriedem?) | |
| 20:54:52 | mriedem | discover_hosts is idempotent so that's the fallback | |
| 20:54:56 | melwitt | although I guess a dry run would only show unmapped hosts | |
| 20:55:00 | melwitt | or does it show all? | |
| 20:55:09 | mriedem | efried: yes i think we probably want to backport that | |
| 20:55:15 | mriedem | i marked the bug as affecting stable | |
| 20:55:31 | efried | mriedem Should we tag the bug with placement, or is that only for stuff on the server side of placement? | |
| 20:55:42 | melwitt | I wouldn't be surprised if ppl would like to be able to see what it's going to map before it maps it | |
| 20:55:49 | mriedem | efried: not sure if that matters | |
| 20:55:55 | efried | ight | |
| 21:02:11 | rwellum | Hi folks. I'm seeing an issue with Pike nova, I assume it's my setup - but wondering if I could get a few pointers what to debug? This is running on kolla containers I should add. https://www.irccloud.com/pastebin/WH8k1XV3/ - looks like a database sync issue? | |
| 21:02:54 | mriedem | rwellum: yes, | |
| 21:03:04 | mriedem | it's looking in the cell0 database for the compute_nodes table, | |
| 21:03:11 | mriedem | so you must not have run 'nova-manage db sync' yet | |
| 21:03:20 | mriedem | after creating the nova_cell0 database | |
| 21:06:45 | rwellum | And that should be run on presumably one of the nova containers | |
| 21:08:03 | mriedem | shrug, i've never used kolla | |
| 21:08:15 | mriedem | i think they are in #openstack-kolla | |
| 21:09:09 | rwellum | Yeah I'm there :) Just not sure who should execute the sync. | |
| 21:09:18 | rwellum | But thanks this is really helpful | |
| 21:09:32 | openstackgerrit | Merged openstack/os-vif master: Updated from global requirements https://review.openstack.org/520478 | |
| 21:16:10 | jaypipes | efried: I'm kind of annoyed right now... I can't figure out for the life of me why https://review.openstack.org/#/c/384807/77//COMMIT_MSG changed back from 1.14 to 1.13... | |
| 21:18:03 | efried | jaypipes Probably you had PS74 locally. edleafe published PS75 with that being the only change. | |
| 21:18:31 | jaypipes | efried: but I pulled a fresh branch before I did anything. | |
| 21:18:42 | openstackgerrit | Jay Pipes proposed openstack/nova master: placement: adds REST API for nested providers https://review.openstack.org/384807 | |
| 21:18:42 | openstackgerrit | Jay Pipes proposed openstack/nova master: placement: update client to set parent provider https://review.openstack.org/385693 | |
| 21:18:42 | jaypipes | efried: whatevs, I've fixed it.. | |
| 21:18:46 | efried | jaypipes ack | |
| 21:19:33 | efried | jaypipes +1. Ship it! | |
| 21:19:40 | rwellum | mriedem: that indeed fixed the database sync error! However I still am getting 'No valid host was found. There are not enough hosts available' | |
| 21:19:58 | mriedem | rwellum: have you created cell1 and discovered hosts? | |
| 21:20:07 | mriedem | rwellum: i thought kolla automated all of this? | |
| 21:20:54 | dansmith | I thought kolla was just containers of stuff | |
| 21:23:42 | mriedem | i thought there was an ansible part to kolla for actually deploying openstack in those containers | |
| 21:23:52 | rwellum | Yeah kolla is just the images and some config built into the image. I'm orchestrating with kolla-kubernetes. Works with ocata btw. | |
| 21:23:55 | mriedem | clearly this isn't the place to ask about wtf kolla is or does :) | |
| 21:24:30 | mriedem | i remember the kolla guys asking about the cellsv2 discover_hosts stuff in pike | |
| 21:24:34 | rwellum | Yeah my bad. More wanted to know where to start with with a generic error like the above. | |
| 21:24:36 | mriedem | so that is automated somewhere in the kolla stadium | |
| 21:24:57 | rwellum | I'll try to find who did it. There's only a few of them. | |
| 21:25:06 | mriedem | http://git.openstack.org/cgit/openstack/kolla-ansible/ ? | |
| 21:25:46 | mriedem | http://git.openstack.org/cgit/openstack/kolla-ansible/tree/ansible/roles/nova/tasks/simple_cell_setup.yml ? | |
| 21:26:17 | mriedem | https://review.openstack.org/#/c/456920/ is the change i remember | |
| 21:26:37 | mriedem | i think | |
| 21:26:42 | openstackgerrit | melanie witt proposed openstack/nova master: Improve the CellDatabases test fixture and usage https://review.openstack.org/508432 | |
| 21:26:51 | mriedem | eduardo is definitely the person to ask | |
| 21:26:53 | mriedem | rwellum: ^ | |
| 21:27:28 | rwellum | ty mriedem | |
| 21:33:27 | rwellum | mriedem: our code (not ansible - kubernetes) is calling: "nova-manage cell_v2 simple_cell_setup" | |
| 21:35:14 | mriedem | rwellum: probably want to read through this then https://docs.openstack.org/nova/latest/user/cells.html#step-by-step-for-common-use-cases | |
| 21:37:11 | rwellum | perfect thanks. | |
| 21:49:18 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Handle spawning error on unshelving https://review.openstack.org/378009 | |
| 21:53:48 | mriedem | dansmith: +2 on the generic lister split | |
| 21:54:42 | dansmith | mriedem: thanks, did you comment on the migration stuff or should I? | |
| 21:54:44 | mriedem | hmm, the bottom patch failed tests http://logs.openstack.org/53/524253/3/check/openstack-tox-py27/540f78a/testr_results.html.gz | |
| 21:54:48 | mriedem | but those look unrelated | |
| 21:54:52 | mriedem | dansmith: i haven't yet | |
| 21:55:00 | dansmith | I rechecked because yeah | |
| 21:56:31 | dansmith | okay I commented on the one you suggested rolling back to PS21 | |
| 21:58:58 | mriedem | thanks | |
| 22:26:58 | mriedem | efried: with the ksa adapter stuff, just to check my understanding, valid_interfaces defaults to (internal, public) and iterates through them when trying to find the endpoint? | |
| 22:27:08 | efried | mriedem yes | |
| 22:27:10 | mriedem | cool | |
| 22:27:21 | mriedem | was making sure i was explaining that correctly here https://review.openstack.org/#/c/513243/2 | |
| 22:29:41 | efried | mriedem lgtm | |
| 22:30:58 | mriedem | wordsmithing that a bit | |
| 22:31:00 | mriedem | in the commit message | |
| 22:31:08 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Downgrade log for keystone verify client fail https://review.openstack.org/513243 | |
| 22:31:15 | mriedem | efried: see if ^ makes sense | |
| 22:32:38 | mriedem | hmm now that i think about it, | |
| 22:32:43 | efried | mriedem Hum, I don't think "like there isn't a keystone v3 endpoint" is right - that would have been covered by the previous except | |
| 22:32:45 | mriedem | your change should go on top of chen's | |
| 22:33:02 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Use ksa adapter for keystone conf & requests https://review.openstack.org/507693 | |