| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-10-24 | |||
| 08:47:16 | bauzas | jianghuaw: I'm about to write a change for adding the new conf opt | |
| 08:47:23 | jianghuaw | hi bauzas | |
| 08:47:28 | bauzas | jianghuaw: did you already uploaded this ? | |
| 08:47:42 | jianghuaw | which new conf opt? | |
| 08:47:57 | bauzas | jianghuaw: enabled_vgpu_types | |
| 08:48:21 | jianghuaw | https://review.openstack.org/#/c/512580/ | |
| 08:48:30 | jianghuaw | bauzas, naichuan is working on it. | |
| 08:48:56 | jianghuaw | it will be appreciated if you can help to review it. | |
| 08:51:05 | bauzas | jianghuaw: honestly, I'll even upload a new patchset | |
| 08:51:20 | bauzas | jianghuaw: because I disagree with using a separate group | |
| 08:51:39 | jianghuaw | What's your preference? | |
| 08:51:55 | jianghuaw | using a single option in [Default]? | |
| 08:53:21 | bauzas | jianghuaw: no, we have a [compute] section | |
| 08:53:42 | bauzas | jianghuaw: lemme upload my change | |
| 08:54:03 | jianghuaw | Ah, I see. I also agree with to put that in [compute]. | |
| 08:54:47 | jianghuaw | yes, go ahead to upload your change. | |
| 08:54:48 | jianghuaw | thanks. | |
| 08:55:38 | openstackgerrit | jichenjc proposed openstack/nova master: Add create inventories doc for placement https://review.openstack.org/512215 | |
| 08:58:22 | openstackgerrit | jichenjc proposed openstack/nova master: [Placement] Add create inventories doc for placement https://review.openstack.org/512215 | |
| 09:00:56 | openstackgerrit | jichenjc proposed openstack/nova master: [placement]Enhance doc for placement allocation list https://review.openstack.org/511715 | |
| 09:02:20 | openstackgerrit | jichenjc proposed openstack/nova master: Add quiesce and unquiesce in support matrix https://review.openstack.org/482390 | |
| 09:07:29 | Dinesh_Bhor | bauzas: Hi, do you know when 'Chris Friesen' will be available for discussion on IRC. I want to discuss this spec: https://review.openstack.org/#/c/512990/ | |
| 09:08:11 | bauzas | Dinesh_Bhor: he's located in Alberta, Canada AFAIK, so I guess a late US TZ (probably UTC-6) | |
| 09:08:24 | bauzas | MST or something | |
| 09:09:40 | Dinesh_Bhor | bauzas: ohh, okay. I will try to catch him. If possible could you please give your feedback for the same spec? | |
| 09:12:44 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: VGPU_support: add enabled white list https://review.openstack.org/512580 | |
| 09:12:53 | bauzas | jianghuaw: ^ | |
| 09:13:47 | jianghuaw | bauzas, cool. | |
| 09:14:25 | bauzas | Dinesh_Bhor: well, we can surely discuss on the spec for sure, but I guess you probably don't know we're past Spec Approval deadline :( | |
| 09:16:31 | Dinesh_Bhor | bauzas: okay, thank you for information | |
| 09:30:33 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Fix missing marker functions https://review.openstack.org/514579 | |
| 10:27:17 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: VGPU: Define vgpu resource class https://review.openstack.org/511180 | |
| 10:27:47 | bauzas | jianghuaw: I just ordered the changes ^ | |
| 10:28:52 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: VGPU_support: add enabled white list https://review.openstack.org/512580 | |
| 10:28:52 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: VGPU: Define vgpu resource class https://review.openstack.org/511180 | |
| 10:29:48 | jianghuaw | bauzas, saw it:-) Thanks. | |
| 11:34:29 | efried | bauzas jianghuaw - can we talk about the appropriate conf section for this VGPU opt? | |
| 11:38:05 | openstackgerrit | Merged openstack/nova master: Add test so we remember why CUSTOM_ prefix added https://review.openstack.org/513001 | |
| 11:42:14 | bauzas | efried: sure, I just want to avoid a nitpick here :) | |
| 11:43:20 | efried | bauzas I'm just thinking we'll potentially end up with any number of these device/virtual-device options, and it would make sense to have a separate [device] section rather than lumping them in [compute]. | |
| 11:43:44 | efried | bauzas Conceptually, [device] is a slightly more generic superset of what [pci] does today. | |
| 11:44:27 | bauzas | efried: probably a good idea, but IMHO in a later change | |
| 11:44:45 | efried | bauzas Whyzat? Then we have to do a deprecation and stuff. | |
| 11:45:04 | bauzas | efried: oh, you meant for VGPUs ? | |
| 11:45:17 | efried | yes | |
| 11:45:22 | bauzas | efried: well, I'm not opposed but I would like to make sure someone does that for PCI too | |
| 11:46:12 | bauzas | efried: let's provide your thoughts in the series, and see if we have a consensus here | |
| 11:46:15 | efried | bauzas Yes, long term everything from [pci] should move to [devices]. I suppose we could put this VGPU stuff into [pci] for now - but I don't see any reason not to start as we mean to go on. | |
| 11:46:30 | bauzas | efried: please, no | |
| 11:46:36 | efried | bauzas I did leave comments on the change you just uploaded, at PS9 | |
| 11:46:47 | bauzas | efried: I mean, PCI != vGPUs | |
| 11:46:51 | bauzas | don't open the wounds :p | |
| 11:46:56 | efried | Totally | |
| 11:46:58 | openstackgerrit | Merged openstack/nova stable/pike: Account for compute.metrics.update in legacy notification whitelist https://review.openstack.org/509439 | |
| 11:47:22 | bauzas | honestly, I don't really where it lands, I just want to make sure people agree on it | |
| 11:47:37 | bauzas | really *care | |
| 11:47:39 | bauzas | meh | |
| 11:47:52 | efried | bauzas Yeah, I'd like Jay to weigh in on this. | |
| 11:48:12 | bauzas | also, I did made a compromise, I had to add that option before we're going to use it | |
| 11:48:20 | efried | saw that. | |
| 11:48:25 | openstackgerrit | Merged openstack/nova master: Add alternate hosts https://review.openstack.org/486215 | |
| 11:48:27 | efried | Not sure why | |
| 11:48:35 | openstackgerrit | Merged openstack/nova stable/pike: Add functional migrate force_complete test https://review.openstack.org/509924 | |
| 11:48:41 | efried | but don't see any reason to object | |
| 11:48:44 | bauzas | I thought oslo.config was somehow having an "experimental" or "next" attribute for the opt, but no | |
| 11:48:48 | openstackgerrit | Merged openstack/nova stable/pike: Add functional for live migrate delete https://review.openstack.org/509925 | |
| 11:48:55 | bauzas | efried: because libvirt and xen will both use it | |
| 11:49:07 | bauzas | of course we could just have it in one set | |
| 11:49:14 | bauzas | and wait the set to be merged | |
| 11:49:24 | bauzas | but that could take much longer | |
| 11:49:37 | efried | bauzas Naw, you can just base both of those changes on this one. | |
| 11:49:51 | bauzas | efried: that's correct, hence my commit msgh | |
| 11:49:57 | bauzas | anyway | |
| 11:50:00 | bauzas | let's wait for reviews | |
| 11:50:15 | bauzas | I'm just writing the yippikay libvirt side | |
| 11:50:18 | efried | There's nothing stopping you from having branches in a series. | |
| 11:50:31 | efried | My ksa-adapter series has several | |
| 11:50:36 | efried | :) | |
| 11:54:37 | efried | bauzas Okay, left review. Thanks for the talk. | |
| 12:02:02 | efried | johnthetubaguy yt? | |
| 12:24:21 | openstackgerrit | Sean Dague proposed openstack/nova master: Move kpartx calls to privsep. https://review.openstack.org/500354 | |
| 12:49:51 | openstackgerrit | Merged openstack/nova stable/pike: Remove dest node allocations during live migration rollback https://review.openstack.org/509926 | |
| 12:57:46 | cdent | efried: did you resolve your resources= thing satisfactorily? | |
| 12:58:04 | efried | cdent Sorry, which resources= thing was that? | |
| 12:58:18 | cdent | [t 2wmI] | |
| 12:58:19 | purplerbot | <efried> Anyone know if we're supposed to handle queryparams with multiple values in the placement API? [2017-10-23 20:04:11.392701] [n 2wmI] | |
| 12:58:20 | efried | cdent Oh, whether you could specify a key multiple times in a querystring? | |
| 12:58:36 | efried | Yeah, the answer was "no", but edleafe thought you would remember why it was decided that way. | |
| 12:58:41 | cdent | the defacto standards with query params is that multiple of the same param should work, but it is not something we really do in nova | |
| 12:59:06 | mriedem | only for sort key/dir i think | |
| 12:59:47 | efried | cdent Well, at least for GET /allocation_candidates and GET /resource_providers?resources= it doesn't work. | |
| 12:59:50 | efried | You'll only get the last one. | |
| 13:00:28 | cdent | yes, that’s a) intentional b) because of the way it is coded c) a fairly arbitrary choice | |
| 13:01:05 | efried | Okay, wfm | |
| 13:01:07 | cdent | the confusion being avoided is whether a list is created by commas or multiple params, and since most people didn’t feel comfy with multiple params, commas was chosen, exclusively | |
| 13:01:50 | openstackgerrit | Merged openstack/nova master: Remove duplicate error info https://review.openstack.org/510719 | |
| 13:02:39 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: CLI for allocations https://review.openstack.org/457534 | |
| 13:02:39 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: CLI for inventories https://review.openstack.org/457533 | |
| 13:02:40 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: CLI for aggregates (v1.1) https://review.openstack.org/505643 | |
| 13:02:40 | openstackgerrit | Andrey Volkov proposed openstack/osc-placement master: CLI for usages https://review.openstack.org/457535 | |