| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2020-04-14 | |||
| 23:17:18 | melwitt | gmann: hah, nice | |
| 23:18:26 | gmann | for nodepool case, can it be done via server metadata ? | |
| 23:21:50 | ianw | gmann: we're really just testing that the value in the config file for nodepool made it through to the booted nova instance -- the actual data is not used in the test | |
| 23:23:18 | gmann | ohk i thought it as just tagging the things. | |
| 23:56:41 | openstackgerrit | Merged openstack/nova master: Introduce scope_types in flavor extra spec policy https://review.opendev.org/719375 | |
| #openstack-nova - 2020-04-15 | |||
| 00:14:36 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add test coverage of existing remaining servers policies https://review.opendev.org/720104 | |
| 00:30:06 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in remaining servers Policies https://review.opendev.org/720106 | |
| 00:34:32 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in servers Policies https://review.opendev.org/662968 | |
| 00:34:43 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add new default roles in servers policies https://review.opendev.org/662971 | |
| 00:34:54 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add test coverage of existing server attributes policies https://review.opendev.org/719726 | |
| 00:35:05 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in servers attributes Policies https://review.opendev.org/719729 | |
| 00:44:34 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add new default roles in servers attributes policies https://review.opendev.org/719730 | |
| 00:45:44 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add test coverage of existing remaining servers policies https://review.opendev.org/720104 | |
| 00:45:54 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Introduce scope_types in remaining servers Policies https://review.opendev.org/720106 | |
| 01:05:03 | openstackgerrit | Brin Zhang proposed openstack/nova-specs master: Re-proposes 'Proposal for a safer remote console with password authentication' https://review.opendev.org/719720 | |
| 01:16:12 | brinzhang_ | gibi, sean-k-mooney: I agree with sean-k-mooney, I think his understand is correct, if we failed to bind, that just leak arq db record in Cyborg, I will check again with Sundar, once I am confirmed, I will respin you again. | |
| 01:34:58 | openstackgerrit | zhufl proposed openstack/nova master: Add missing white spaces between words in log messages https://review.opendev.org/720112 | |
| 01:58:18 | brinzhang_ | gibi, sean-k-monney, dansmith: another thing maybe need to concern, please check, see below: | |
| 01:58:21 | brinzhang_ | While we build_instances() we should cleanup arqs if the _create_and_bind_arqs() failed. | |
| 01:58:30 | brinzhang_ | cyborgclinet), right? | |
| 01:58:30 | brinzhang_ | In _cleanup_when_reschedule_fails()[1] we clean the arqs if _create_and_bind_arqs() failed, but I think we just only clean it at L847 if the _create_and_bind_arqs() failed [3]. If we add "compute_utils.delete_arqs_if_needed(context, instance)"[2] in _cleanup_when_reschedule_fails()[1], once we call _cleanup_when_reschedule_fails() it will be cleanup arqs with dp name in instance's flavor, I think that will increase Nova's call to Cyborg (by | |
| 01:58:43 | brinzhang_ | [1]https://opendev.org/openstack/nova/src/branch/master/nova/conductor/manager.py#L582 | |
| 01:58:51 | brinzhang_ | [2]https://opendev.org/openstack/nova/src/branch/master/nova/conductor/manager.py#L598 | |
| 01:58:57 | brinzhang_ | [3]https://opendev.org/openstack/nova/src/branch/master/nova/conductor/manager.py#L848-L849 | |
| 01:59:16 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add new default roles in remaining servers policies https://review.opendev.org/720116 | |
| 02:00:24 | brinzhang_ | IMO, if I am correct, I think we should move "compute_utils.delete_arqs_if_needed(context, instance)"[2] to the right place, i.e., when _create_and_bind_arqs() failed | |
| 02:05:00 | gmann | johnthetubaguy: i replied on server policies comments and finished the remaining one. please check. my worried is to allow host attr for project that is why i kept them system-admin for now | |
| 02:20:21 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Fix follow up comments on policy work https://review.opendev.org/717835 | |
| 02:23:37 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Use oslo policy flag to disable default change warning instead of all https://review.opendev.org/717884 | |
| 03:10:20 | openstackgerrit | Merged openstack/nova master: Introduce scope_types in quota set Policies https://review.opendev.org/719317 | |
| 03:44:48 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Fix server actions to be system and project scoped https://review.opendev.org/720123 | |
| 05:56:03 | openstackgerrit | Merged openstack/nova master: Add new default roles in server group policies https://review.opendev.org/717175 | |
| 06:35:39 | openstackgerrit | Merged openstack/nova master: Add new default roles in quota sets policies https://review.opendev.org/719361 | |
| 06:45:13 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add doc for policy new defaults https://review.opendev.org/720129 | |
| 07:05:38 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Fix server actions to be system and project scoped https://review.opendev.org/720123 | |
| 07:25:21 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Add new default roles in remaining servers policies https://review.opendev.org/720116 | |
| 07:37:10 | bauzas | good morning Nova | |
| 07:38:52 | gibi | bauzas: good morning | |
| 07:39:52 | bauzas | gibi: good news, now that I'm done with the last series I had to work on, I'll do upstream bug triage today \o/ | |
| 07:40:37 | bauzas | it's been a while | |
| 07:42:18 | bauzas | gibi: stephenfin: sean-k-mooney: lyarwood: since you reviewed the vGPU multiple types series, you could be interested in that quite easy series for fixing cold migration issues for vGPUs | |
| 07:42:45 | bauzas | tl;dr: that's just adding a new argument to resize virt methods and just calling a new method | |
| 07:42:59 | bauzas | the main review is just for the functional test | |
| 07:43:08 | gibi | bauzas: \o/ | |
| 07:43:43 | gibi | bauzas: hit me with the link | |
| 07:44:01 | bauzas | gibi: https://review.opendev.org/#/c/712741 and below | |
| 07:44:47 | gibi | thanks | |
| 07:44:55 | bauzas | gibi: thanks for you :p | |
| 07:45:18 | openstack | Launchpad bug 1868033 in OpenStack Compute (nova) "Booting instance with pci_device fails during rocky->stein live upgrade" [High,New] | |
| 07:45:18 | gibi | bauzas: I'm looking at https://bugs.launchpad.net/nova/+bug/1868033 so you can skip that | |
| 07:45:27 | bauzas | gibi: ack | |
| 07:48:44 | bauzas | 103 bugs... :( | |
| 07:54:12 | gibi | and I already triaged couple of yesterday :) | |
| 07:54:23 | gibi | at least we won't be bored | |
| 07:59:22 | bauzas | gibi: yeah, my main worries go to find some potential bugs that are Ussuri regressions | |
| 08:10:10 | openstack | Launchpad bug 1777608 in OpenStack Compute (nova) "Nova compute calls plug_vifs unnecessarily for ironic nodes in init_host" [Medium,New] | |
| 08:10:10 | bauzas | gibi: heh, from what I can see, nothing really changes from my day-to-day internal bug triage https://bugs.launchpad.net/nova/+bug/1777608 | |
| 08:10:22 | bauzas | gibi: sorry if you feel getting trampled by OSP :p | |
| 08:11:48 | gibi | :) | |
| 08:18:34 | bauzas | lyarwood: looks a valid bug, do you confirm ? | |
| 08:18:38 | bauzas | whoops | |
| 08:18:42 | openstack | Launchpad bug 1871861 in OpenStack Compute (nova) "Share the snapshot of volume-backed instance to other tenant, but can not build instance with this image." [Undecided,New] | |
| 08:18:42 | bauzas | lyarwood: https://bugs.launchpad.net/nova/+bug/1871861 | |
| 08:23:27 | nightmare_unreal | sean-k-mooney: I saw your comments/review here : https://review.opendev.org/#/c/715395/ . I have added my draft, what do you suggest ? | |
| 08:23:37 | nightmare_unreal | Thanks | |
| 08:23:43 | lyarwood | bauzas: looking | |
| 08:23:58 | bauzas | lyarwood: tl;dr: glance allowed to share an image | |
| 08:24:18 | bauzas | lyarwood: another tenant can take this image to build an volume-baked instance | |
| 08:24:30 | bauzas | lyarwood: but this fails miserably | |
| 08:24:44 | lyarwood | bauzas: yeah the issue is it's not an image, it's just a metadata store pointing to the volume snapshot that's in the other tenant still | |
| 08:25:08 | bauzas | but then a cinder bug, correct ? | |
| 08:25:24 | bauzas | we just get the metadata from cinder, nope ? | |
| 08:25:37 | lyarwood | bauzas: yeah it's a cinder feature bug | |
| 08:25:58 | bauzas | lyarwood: thanks, I'm done with this bug | |
| 08:25:58 | lyarwood | well, between glance and cinder at least | |
| 08:26:14 | bauzas | we're talking about upstream bug triage | |
| 08:26:23 | bauzas | I just want the cinder folks to look at it | |
| 08:26:39 | lyarwood | tbh there might be a way of forcing the data to actually end up in glance | |
| 08:26:46 | lyarwood | ack yeah | |
| 08:26:55 | lyarwood | I'll take a look quickly but I'll leave it to them if I can't find anything | |
| 08:34:09 | johnthetubaguy | lyarwood: do cinder do volume snapshot sharing between tenants? | |
| 08:34:33 | johnthetubaguy | (or get glance to copy it, I guess) | |
| 08:36:43 | bauzas | johnthetubaguy: I feel this bug will potentially require a cross-project effort in order to get fixed | |
| 08:36:52 | bauzas | (but there could be duplicates) | |
| 08:36:55 | johnthetubaguy | bauzas: sounds like it | |
| 08:37:19 | johnthetubaguy | I mean, I actually don't think nova should do anything here, but I would say that :P | |
| 08:37:31 | lyarwood | johnthetubaguy: yeah I was thinking you could force the upload into glance to workaround | |
| 08:38:02 | johnthetubaguy | lyarwood: I was thinking glance forces an upload into cinder for the additional project, but either way I guess | |
| 08:38:57 | lyarwood | that could also work, at least c-api/c-vol can offload things to the backends that way | |
| 08:41:29 | bauzas | johnthetubaguy: feel free to add nova as a related project to the bug then ;) | |
| 08:41:47 | bauzas | johnthetubaguy: I only marked cinder, but obviously glance and maybe nova could be added | |
| 08:58:54 | openstack | Launchpad bug 1871583 in OpenStack Compute (nova) "Migrate the instance, the destination host is not specified. When the "destination host" == " source host" , will set the instance's state error." [Undecided,New] - Assigned to Qiu Fossen (fossen123) | |
| 08:58:54 | bauzas | gibi: I've seen you started to look at https://bugs.launchpad.net/nova/+bug/1871583 | |
| 08:59:44 | bauzas | gibi: I'm confused with the exception | |
| 09:01:03 | bauzas | gibi: so, libvirt doesn't support resizing to the same host | |
| 09:01:56 | bauzas | hence the exception | |
| 09:02:12 | bauzas | anyway, nevermind | |
| 09:02:18 | bauzas | I think what to triage | |
| 09:02:22 | bauzas | what/how | |
| 09:06:48 | slaweq | gibi: hi | |