| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2020-04-09 | |||
| 08:43:48 | gibi | brinzhang: looking at the change they are separate by unchanged lines so I think git will manage to merge it | |
| 08:45:47 | bauzas | brinzhang: gibi: sorry a bit distracted due to the kids :) | |
| 08:46:28 | brinzhang | gibi: yes, it's ok, I tested in local used: | |
| 08:46:34 | brinzhang | git review -d 701430 | |
| 08:46:39 | brinzhang | git review -x 718357 | |
| 08:47:19 | brinzhang | I think I should simulate a scenario to check "git review -R", thanks gibi, bauzas | |
| 08:47:32 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in servers Policies https://review.opendev.org/662968 | |
| 08:47:57 | bauzas | brinzhang: since recent revisions, Gerrit now tells you whether it would create a merge conflict | |
| 08:48:02 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in servers Policies https://review.opendev.org/662968 | |
| 08:48:07 | bauzas | I mean, the Gerrit UI | |
| 08:48:59 | gibi | I never used git review. I like the full controll of git push <remote> <what>:refs/for/<branch>/topic | |
| 08:49:18 | gibi | I mean <topic> | |
| 08:49:24 | brinzhang_ | bauzas: yeah, I know, but that until the patch merged then we can know whether conflict from gerrit | |
| 08:49:45 | huaqiang | :qa | |
| 08:49:47 | huaqiang | exit | |
| 08:49:51 | huaqiang | :( | |
| 08:50:19 | bauzas | brinzhang: no, you can test on your local git repo | |
| 08:50:22 | brinzhang_ | ah, I was dropped, I think I missed what did you said above, gibi | |
| 08:50:42 | gibi | I never used git review. I like the full controll of git push <remote> | |
| 08:50:45 | gibi | <what>:refs/for/<branch>/topic | |
| 08:50:48 | bauzas | brinzhang: you have a local git remote on your git repo that's named 'gerrit' | |
| 08:50:54 | bauzas | git remote -v | grep gerrit | |
| 08:51:14 | brinzhang_ | bauzas: yeah, I know, it's in the project's git/config | |
| 08:51:24 | bauzas | correct | |
| 08:51:31 | bauzas | so you can basically test what you want | |
| 08:51:48 | brinzhang_ | I mean, Gerrit cannot know the merged conflict, if the patch doesnot merge | |
| 08:52:21 | brinzhang_ | in local I am easy to know which is conflict | |
| 08:52:30 | brinzhang_ | s/am/can | |
| 08:52:53 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add new default roles in keypairs policies https://review.opendev.org/718619 | |
| 08:53:04 | bauzas | maybe I misunderstand you | |
| 08:53:34 | bauzas | you mean that you can't predict that Zuul will tell you a merge conflict if there are concurrent patches coming by, if so yes I agree | |
| 08:53:40 | brinzhang_ | gibi: yes, git push gave by gerrit form patch's "Download" tag | |
| 08:54:13 | brinzhang_ | bauzas: yes, that's what I want to say | |
| 08:54:37 | bauzas | ... and nobody can :-) | |
| 08:55:01 | bauzas | that's what happens when you merge N changes at once without testing them each one by one | |
| 08:55:14 | brinzhang_ | bauzas, gibi: haha, thanks, I will check my question from sandbox test. | |
| 08:55:36 | brinzhang_ | bauzas: yes, agree | |
| 08:58:17 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Pass the actual target in keypairs policy https://review.opendev.org/718621 | |
| 08:58:38 | bauzas | gibi: okay so I'm confused | |
| 08:58:47 | bauzas | gibi: about https://review.opendev.org/#/c/718461/ | |
| 08:59:01 | bauzas | tbh, I dropped attention yesterday evening about how things were going | |
| 08:59:59 | gibi | bauzas: we found yesterday that keys like trait:CUSTOM_* is not accepted | |
| 09:00:13 | bauzas | that, I understood :) | |
| 09:00:21 | gibi | and also found out that placement is more strict about trait names and resource class names than nova | |
| 09:00:44 | gibi | I think this is the two change in that patch | |
| 09:01:59 | kashyap | sean-k-mooney: stephenfin: Can you remind me again: Nova doesn't yet support vCPU hot unplug, yeah? | |
| 09:02:08 | openstackgerrit | Kevin Zhao proposed openstack/nova master: fix scsi disk unit number of the attaching volume when cdrom bus is scsi https://review.opendev.org/712607 | |
| 09:02:22 | stephenfin | kashyap: correct | |
| 09:02:45 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add test coverage of existing keypairs policies https://review.opendev.org/718604 | |
| 09:03:03 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in keypairs https://review.opendev.org/718609 | |
| 09:03:14 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add new default roles in keypairs policies https://review.opendev.org/718619 | |
| 09:03:26 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Pass the actual target in keypairs policy https://review.opendev.org/718621 | |
| 09:03:46 | openstackgerrit | Stephen Finucane proposed openstack/nova master: api: Allow custom traits https://review.opendev.org/718461 | |
| 09:03:50 | kashyap | stephenfin: Thx | |
| 09:04:23 | kashyap | And we do support I/O throttling? IIRC, we do have the I/O limit settings | |
| 09:04:40 | stephenfin | yes, though I don't know how well they work since testing is light | |
| 09:04:49 | kashyap | Yep; read_bytes_sec/write_bytes_sec | |
| 09:05:06 | kashyap | stephenfin: Oh, that rings a faint bell where you were traiging a bug on it | |
| 09:05:12 | kashyap | s/on it/about it/ | |
| 09:05:32 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add new default roles in os-evacuate policies https://review.opendev.org/708237 | |
| 09:06:33 | brinzhang | stephenfin: can you explain why we add default=None, and not add for the --no-config-drive and --no-delete-on-termination? | |
| 09:06:45 | brinzhang | stephenfin: thanks for +W | |
| 09:06:51 | kashyap | Another one: no memory hotplug either, yeah? | |
| 09:07:38 | brinzhang | stephenfin: https://review.opendev.org/#/c/712651/19/novaclient/v2/shell.py@2741 and https://review.opendev.org/#/c/718349/4/novaclient/v2/shell.py@1551 | |
| 09:08:01 | stephenfin | brinzhang: It's not necessary. The default will be set by '--config-drive' if it's not specified, and will be overridden by '--no-config-drive' if necessary | |
| 09:09:05 | brinzhang | controlled by "group"? | |
| 09:10:17 | stephenfin | Yeah, group is our own invention. It's essentially an alias for dest that also implies that we should add opts sharing the same value to a mutually exclusive group. See [1] | |
| 09:10:39 | stephenfin | [1] https://review.opendev.org/#/c/718349/4/novaclient/shell.py | |
| 09:10:44 | stephenfin | kashyap: correct | |
| 09:11:22 | kashyap | stephenfin: Thx; I'm answering a doc writer's questions. Hence the questions here | |
| 09:14:55 | gibi | sh*t https://review.opendev.org/#/c/716222/ got double hit by https://bugs.launchpad.net/nova/+bug/1823251 the gate queue will reset :/ | |
| 09:14:57 | openstack | Launchpad bug 1823251 in OpenStack Compute (nova) "Spike in TestNovaMigrationsMySQL.test_walk_versions/test_innodb_tables failures since April 1 2019 on limestone-regionone" [High,Confirmed] | |
| 09:15:14 | brinzhang | stephenfin: yeah, I looked the group, got it, thanks | |
| 09:16:12 | lyarwood | gibi: time to reintroduce https://review.opendev.org/#/c/683009/ ? | |
| 09:16:58 | gibi | lyarwood: yeah. I think I will push that again | |
| 09:18:24 | bauzas | gibi: FWIW, I work tomorrow, I can fast-approve any change that was +W today | |
| 09:18:34 | bauzas | in case we got hit by the gate | |
| 09:18:56 | gibi | bauzas: thanks. I will rely on your power | |
| 09:19:19 | bauzas | gibi: the only problem is that I could miss any change, so please CC on the ones you'd like to see | |
| 09:19:23 | bauzas | CC me* | |
| 09:19:31 | bauzas | I can still ask gerrit to find me all +Wd tho | |
| 09:19:37 | bauzas | actually, nevermind, I'll work this way | |
| 09:19:39 | gibi | bauzas: also there is the novaclient release patch that needs an update when 2.87 lands https://review.opendev.org/#/c/718260/ | |
| 09:20:13 | gibi | and thank you | |
| 09:20:29 | bauzas | gibi: fill an etherpad for tracking if you feel I could loose anything | |
| 09:20:38 | gibi | bauzas: I'm on it :) | |
| 09:20:42 | bauzas | cool | |
| 09:21:47 | bauzas | stephenfin: I thought placement wasn't accepting dashes for trait names ? | |
| 09:23:18 | bauzas | shit, can't find the regex on the placement api-ref for reference | |
| 09:23:26 | stephenfin | bauzas: I thought so too, but I checked the regex and they are | |
| 09:23:30 | stephenfin | Sec, lemme drag it out | |
| 09:24:30 | stephenfin | bauzas: https://github.com/openstack/placement/blob/master/placement/schemas/common.py#L27 | |
| 09:24:31 | bauzas | stephenfin: just held a procedural -1 | |
| 09:24:41 | stephenfin | We have a similar regex somewhere in nova | |
| 09:24:43 | bauzas | stephenfin: but then it could be nova regex | |
| 09:24:50 | bauzas | yeah, I was thinking of so | |
| 09:25:09 | stephenfin | bauzas: yeah, here https://github.com/openstack/nova/blob/master/nova/scheduler/utils.py#L57 | |
| 09:25:10 | bauzas | because when I wanted to create a custom trait for nvidia-11, I got fucked up | |
| 09:25:37 | stephenfin | hyphens are perfectly fine based on those | |
| 09:26:05 | bauzas | I'm then confused | |