| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-04-06 | |||
| 16:40:51 | mriedem | oh | |
| 16:40:51 | mriedem | i see | |
| 16:41:05 | mriedem | let me get out my red pen | |
| 16:41:13 | dansmith | mriedem: I haven't gotten to the actual subnode part of my devstack patch to know if I need that again or not | |
| 16:41:16 | dansmith | mriedem: I think I will | |
| 16:41:35 | dansmith | mriedem: if it's bothering you I can resubmit it under my name so you can abandon | |
| 16:41:36 | dansmith | I've long since given up trying to keep my dashboard clean | |
| 16:41:39 | ildikov | mriedem: cool, I'll fix it now, tnx | |
| 16:44:27 | mriedem | dansmith: it's fine, i just clean up every few weeks | |
| 16:45:35 | mdnadeem | ildikov, dansmith , mriedem , Hello guys, i have a query | |
| 16:45:42 | mdnadeem | https://github.com/openstack/nova/blob/stable/newton/nova/scheduler/filters/aggregate_image_properties_isolation.py#L42 | |
| 16:46:07 | mdnadeem | ^^ It returns only core image property | |
| 16:46:16 | mriedem | yes | |
| 16:46:21 | mriedem | your out of tree image properties won't work | |
| 16:46:40 | mdnadeem | I have set some custom property to my image , however i cant get it there | |
| 16:46:46 | mriedem | because of https://github.com/openstack/nova/blob/stable/newton/nova/objects/image_meta.py#L213 | |
| 16:47:13 | mdnadeem | mriedem, any way to use custom image property ? | |
| 16:47:14 | mriedem | https://github.com/openstack/nova/blob/stable/newton/nova/objects/image_meta.py#L478 | |
| 16:47:20 | mriedem | mdnadeem: upstream it | |
| 16:47:27 | mriedem | or hack that object and live with the version changes | |
| 16:48:52 | mdnadeem | mriedem, is this a bug ? | |
| 16:49:01 | mriedem | mdnadeem: no, it's by design | |
| 16:49:22 | mriedem | a wild west of custom image properties is a barrier to interoperability | |
| 16:49:32 | mriedem | flavor extra specs is still a wild west if that helps | |
| 16:49:57 | mdnadeem | mriedem, I have a requirement to use custom image property, should i propose a fix upstream for the same? | |
| 16:50:16 | mriedem | mdnadeem: if you have an image property that makes sense to upstream then i encourage you to do so | |
| 16:50:29 | mriedem | because chances are someone else has a similar use case | |
| 16:50:42 | mdnadeem | mriedem, ok, Thanks a lot :) | |
| 16:50:45 | mriedem | yw | |
| 16:50:57 | mdnadeem | mriedem++ | |
| 16:52:39 | mriedem | dansmith: i'm brain isn't working on why api/scheduler/conductor show up in cell0 https://review.openstack.org/#/c/442162/19/nova/compute/api.py@4433 - i thought those would only show up if we start the services? | |
| 16:52:47 | mriedem | s/i'm/my | |
| 16:54:49 | mriedem | https://github.com/openstack/nova/blob/8d9bf947a4c8654a30e016a5d95d9bec55447506/nova/service.py#L154 | |
| 16:56:26 | openstackgerrit | Ildiko Vancsa proposed openstack/nova master: Deprecate the Cinder API v2 support https://review.openstack.org/444253 | |
| 17:00:47 | dansmith | mriedem: they're going to show up in whatever db is configured in their config file | |
| 17:01:06 | dansmith | mriedem: which in my devstack patch, and in real deploys after we're past this hump, will be cell0 I expect | |
| 17:01:07 | mriedem | right, which means you f'ed up and started services with a nova.conf pointing at cell0 right? | |
| 17:01:33 | mriedem | hmm | |
| 17:01:35 | dansmith | I mean, you could just choose a random cell to have them show up in, but I dunno why you would | |
| 17:01:50 | mriedem | yeah i guess that's a good point | |
| 17:01:52 | mriedem | they aren't in the api db | |
| 17:01:58 | mriedem | they wouldn't be in a random cell db | |
| 17:02:00 | dansmith | right | |
| 17:02:05 | dansmith | cell0 seems like a fine place for them to be | |
| 17:02:07 | mriedem | right now they are the nova db b/c we have 1 cell | |
| 17:02:09 | dansmith | right | |
| 17:02:10 | mriedem | ok | |
| 17:02:20 | mriedem | and nova-compute goes in nova1, nova2, etc et | |
| 17:02:27 | dansmith | right never any computes in cell0, | |
| 17:02:35 | mriedem | i'll pay you $5 for that topology diagram | |
| 17:02:36 | dansmith | but the api level services seem appropriate | |
| 17:02:38 | dansmith | hah | |
| 17:02:57 | mriedem | my friend abe lincoln would love to see it | |
| 17:03:40 | dansmith | I have a shitty one sketched out for my talk in boston, but it's more op-focused with pictures of computers and stuff | |
| 17:04:01 | dansmith | doesn't really have this level of detail | |
| 17:04:33 | mriedem | you could do two pictures :) | |
| 17:04:38 | mriedem | high level cells topology | |
| 17:04:43 | mriedem | and then what's inside a cell | |
| 17:04:58 | mriedem | showing the superconductor thing will be important | |
| 17:09:36 | dansmith | yeah | |
| 17:09:47 | mriedem | done with https://review.openstack.org/#/c/442162/ btw | |
| 17:09:52 | mriedem | gonna get lunch | |
| 17:24:14 | melwitt | jroll: do you know how ironic nodes get registered as resource providers in placement? I'm not finding anything in the code so far | |
| 17:24:46 | jroll | melwitt: they're each presented as a node to the resource tracker, which does the registering | |
| 17:25:13 | melwitt | jroll: okay, thanks. I see the stuff in resource tracker but didn't see the connection with the ironic driver | |
| 17:25:19 | jroll | melwitt: https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L754 | |
| 17:25:31 | jroll | that returns CUSTOM_FOO or whatever from ironic | |
| 17:25:35 | jroll | ironic driver* | |
| 17:25:40 | melwitt | ah, thank you | |
| 17:25:48 | jroll | no problem | |
| 17:32:31 | openstackgerrit | Dan Smith proposed openstack/nova master: Teach HostAPI about cells https://review.openstack.org/442162 | |
| 17:32:32 | openstackgerrit | Dan Smith proposed openstack/nova master: Make scheduler target cells to get compute node instance info https://review.openstack.org/439891 | |
| 17:32:33 | openstackgerrit | Dan Smith proposed openstack/nova master: Sort CellMappingList.get_all() for safety https://review.openstack.org/443174 | |
| 17:32:34 | openstackgerrit | Dan Smith proposed openstack/nova master: Add get_by_instance_uuids() to InstanceMappingList https://review.openstack.org/443292 | |
| 17:32:35 | openstackgerrit | Dan Smith proposed openstack/nova master: Clean up ClientRouter debt https://review.openstack.org/444487 | |
| 17:32:36 | openstackgerrit | Dan Smith proposed openstack/nova master: Make server_groups determine deleted-ness from InstanceMappingList https://review.openstack.org/443293 | |
| 17:32:37 | openstackgerrit | Dan Smith proposed openstack/nova master: Remove Mitaka-era service version check https://review.openstack.org/442861 | |
| 17:32:39 | openstackgerrit | Dan Smith proposed openstack/nova master: Add workaround to disable group policy check upcall https://review.openstack.org/442736 | |
| 17:32:56 | dansmith | mriedem: bottom patch got hit by the mirror thing so I rebased and re-pushed.. so it needs a re-=W | |
| 17:33:00 | dansmith | er, re-+W | |
| 17:57:17 | mriedem | dansmith: got it, so the first 5 are ready | |
| 17:57:33 | mriedem | looks like typo city in https://review.openstack.org/#/c/442736/ | |
| 17:58:04 | dansmith | yeah | |
| 17:58:05 | dansmith | thanks | |
| 17:58:57 | cfriesen | does anyone know why we default to not returning the userdata when showing server details to a normal user, but show it for an admin user? | |
| 18:03:16 | mriedem | nope | |
| 18:04:31 | EmilienM | mriedem: hey, I think I'll abandon https://review.openstack.org/#/c/453338/ if nobody approves it, it needed to land before you tag next week but since're approaching next week, I can we can wait now for the tag... | |
| 18:05:13 | mriedem | wfm | |
| 18:11:30 | openstackgerrit | Dan Smith proposed openstack/nova master: Sort CellMappingList.get_all() for safety https://review.openstack.org/443174 | |
| 18:11:31 | openstackgerrit | Dan Smith proposed openstack/nova master: Clean up ClientRouter debt https://review.openstack.org/444487 | |
| 18:11:33 | openstackgerrit | Dan Smith proposed openstack/nova master: Add workaround to disable group policy check upcall https://review.openstack.org/442736 | |
| 18:14:35 | cdent | melwitt: do I recall that you were going to adjust the project/user association in placement to use usages? | |
| 18:15:05 | melwitt | cdent: yes, I had intended to propose an amendment to the spec yesterday but didn't get to it yet | |
| 18:15:24 | cdent | ah, cool, that would explain why I can't find it :) thanks | |
| 18:15:35 | melwitt | heh, sorry about that | |
| 18:16:14 | cdent | no problem, you're the one who is making with the good works | |
| 18:51:32 | openstackgerrit | Sean Dague proposed openstack/nova master: Remove the extension check for os-networks in servers API https://review.openstack.org/450837 | |
| 18:51:56 | openstackgerrit | Sean Dague proposed openstack/nova master: Remove the json-schema extension point of server create https://review.openstack.org/451194 | |
| 18:52:12 | openstackgerrit | Sean Dague proposed openstack/nova master: Remove the stevedore extension point for server create https://review.openstack.org/451195 | |
| 18:54:42 | sdague | mriedem: you want in on - https://review.openstack.org/#/c/448882 before I approve (nova cert removal) | |