| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-12-13 | |||
| 09:42:53 | openstack | Launchpad bug 1856296 in kolla-ansible "upgrade to Train might fail due to mariadb row format" [Critical,Triaged] | |
| 09:42:53 | yoctozepto | I assigned https://bugs.launchpad.net/nova/+bug/1856296 to nova, mostly to let you know, not sure if you can do too much about it, other than letting users know they might be impacted on upgrades | |
| 09:54:11 | openstackgerrit | Guo Jingyu proposed openstack/nova master: Make scheduling more debuggable https://review.opendev.org/698421 | |
| 11:25:30 | openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Drop tables for removed services, features https://review.opendev.org/698706 | |
| 11:25:31 | openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Remove unused ec2 DB APIs https://review.opendev.org/698880 | |
| 11:25:31 | openstackgerrit | Stephen Finucane proposed openstack/nova master: WIP: db: Remove placement tables https://review.opendev.org/698707 | |
| 11:25:32 | openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Remove unused models https://review.opendev.org/698881 | |
| 11:58:01 | gibi | stephenfin: hi! what do you think about https://review.opendev.org/#/c/696574/10/nova/objects/instance.py@1239 ? shall I move this whole thing back to utils? | |
| 11:59:16 | stephenfin | gibi: I guess if there's nowhere better, you've no choice :) | |
| 11:59:27 | stephenfin | nova/compute/utils.py though, rather than nova/utils.py | |
| 12:07:27 | gibi | stephenfin: ack. moving | |
| 12:42:26 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Revert "(Temporarily) readd bare support for py27" https://review.opendev.org/698890 | |
| 12:42:45 | stephenfin | tosky: we're good to merge that now, right? ^ | |
| 12:43:18 | sean-k-mooney | that drops support again since m1 was yesterday? | |
| 12:43:36 | stephenfin | officially, yeah | |
| 12:43:45 | stephenfin | it prevents you installing nova under python 2.7 | |
| 12:43:51 | stephenfin | pip will spit the dummy out | |
| 12:43:53 | sean-k-mooney | oh i know what the patch does | |
| 12:44:07 | sean-k-mooney | i was just checking m1 was yesterday | |
| 12:44:13 | sean-k-mooney | i was 99% sure it was | |
| 12:44:47 | sean-k-mooney | also that is such an irish saying | |
| 12:49:00 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Remove remaining Python 2.7-only dependencies https://review.opendev.org/698895 | |
| 12:49:00 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Revert "(Temporarily) readd bare support for py27" https://review.opendev.org/698890 | |
| 12:49:31 | sean-k-mooney | stephenfin: abandoned by typeing module fixes you flake 8 changes already cover that right | |
| 12:49:51 | stephenfin | yup | |
| 12:50:12 | sean-k-mooney | actully i really dont like the if False | |
| 12:50:44 | sean-k-mooney | can you just convert the comments in the native form and import them normally | |
| 12:50:57 | sean-k-mooney | or import them normally and add the #noqa | |
| 12:51:06 | sean-k-mooney | to silence the unused import thing | |
| 12:51:14 | stephenfin | Once we merge https://review.opendev.org/698890. yup | |
| 12:51:21 | stephenfin | Feel free to -1 to remind me :) | |
| 12:51:38 | sean-k-mooney | oh your right py27 | |
| 12:51:47 | sean-k-mooney | ya we should merge that :) | |
| 12:52:03 | stephenfin | and now, I'm nearly an hour late for lunch, but at least X-cell resize is reviewed finally. bbiab | |
| 12:54:51 | openstackgerrit | jichenjc proposed openstack/nova master: libvirt: avoid cpu check at s390x arch https://review.opendev.org/696228 | |
| 13:06:06 | tosky | stephenfin: uhm, maybe | |
| 13:21:23 | tobiash | hi, I'd like to revive discussion on this nova-spec which was proposed for rocky and abandoned: https://review.opendev.org/572805. What is the preferred way of doing this? Shall I create a new change or ask a nova-core to hit the restore button? | |
| 13:23:02 | gibi | tobiash: restored the spec. Please update it (e.g. move it to the ussuri dir at least) | |
| 13:23:19 | tobiash | gibi: thank you very much! | |
| 13:24:09 | sean-k-mooney | tobiash: you will need to add the new feature liason field at a minium to get it to pass the gate checks | |
| 13:24:21 | sean-k-mooney | there is a description of that in the readme | |
| 13:24:43 | tobiash | sean-k-mooney: thanks for the hint | |
| 13:24:45 | sean-k-mooney | there may be some other template changes too but i think that is the main one since rocky | |
| 13:26:11 | openstackgerrit | Eric Xie proposed openstack/nova master: Fix error when live-migrating instance cause of gic version https://review.opendev.org/698905 | |
| 13:38:40 | openstackgerrit | Tobias Henkel proposed openstack/nova-specs master: Add spec for downloading images via RBD https://review.opendev.org/572805 | |
| 13:54:58 | openstackgerrit | Matthew Booth proposed openstack/nova master: Don't retry database connections from command line applications https://review.opendev.org/666892 | |
| 14:01:46 | openstack | Launchpad bug 1856311 in OpenStack Compute (nova) "server_external_events response status is always 'completed' instead of event status which requested as failed" [Undecided,New] | |
| 14:01:46 | mriedem | efried: have you seen this? https://bugs.launchpad.net/nova/+bug/1856311 | |
| 14:03:49 | efried | mriedem: nope, /me clicks | |
| 14:04:01 | efried | what, now that I'm a server-external-events expert? | |
| 14:04:18 | mriedem | yes you own it now | |
| 14:04:26 | mriedem | but it looks like it's not what i thought they were complaining about | |
| 14:04:27 | efried | I haven't looked at the novaclient shim at all yet | |
| 14:05:00 | sean-k-mooney | is that valid? i would have assuemd that status faild for network-vif-plugged ment the pluggin failed | |
| 14:05:46 | sean-k-mooney | and code 200 ment the event was sucessfully recieved? | |
| 14:06:04 | sean-k-mooney | i dont actully know if that is how it works | |
| 14:06:40 | sean-k-mooney | oh sorry read that wrong | |
| 14:09:00 | mriedem | https://github.com/openstack/nova/blob/e6f742544432d6066f1fba4666580919eb7859bd/nova/api/openstack/compute/server_external_events.py#L133 | |
| 14:09:02 | mriedem | it's working as designed | |
| 14:09:12 | mriedem | completed == we got the event and we're sending it on for processing, thanks for calling | |
| 14:09:15 | sean-k-mooney | ya i think so too was just readin the api ref | |
| 14:09:22 | mriedem | dansmith: correct ^? | |
| 14:09:41 | mriedem | laura did rub cbd oil all over my lower back this morning though so i might just be like super duper high | |
| 14:11:13 | sean-k-mooney | mellow maybe cbd does not have the more interesing sideeffects | |
| 14:12:41 | mriedem | heh, i know, but it's a ruse i like to play on my wife like i have no idea and i'm 85 | |
| 14:18:55 | efried | I'm supposed to be vacationing starting next week... | |
| 14:18:55 | efried | https://review.opendev.org/#/c/686804/ | |
| 14:18:55 | efried | gibi, stephenfin: Will y'all have a chance to look at the vTPM spec today? | |
| 14:19:56 | gibi | efried: sorry. I will not have the chance today | |
| 14:20:20 | gibi | efried: or at least I cannot promise | |
| 14:20:30 | efried | okay. | |
| 14:22:07 | stephenfin | Afraid I'm in the same boat. Stuck with internal stuff :( | |
| 14:22:59 | efried | Okay. | |
| 14:24:29 | sean-k-mooney | if i get this backport done soon ill try to look at it again that said i think its going to be pretty close jsut reading your comments | |
| 14:26:22 | efried | thanks sean-k-mooney | |
| 14:31:56 | dansmith | mriedem: completed in the return from the api means "got it dude" yes.. but status in the actual event body is the external service telling nova (i.e. compute) if something worked or not | |
| 14:32:24 | mriedem | right | |
| 14:32:32 | mriedem | hey nova, this event failed; ok, ack | |
| 14:32:46 | dansmith | yup | |
| 14:37:19 | mriedem | stephenfin: thanks for hitting all of those cross-cell resize changes | |
| 14:45:08 | stephenfin | mriedem: np | |
| 14:45:50 | stephenfin | It was pretty much just functional tests once I got past the remaining conductor tasks. Need to go find someone else now :) | |
| 15:02:41 | openstackgerrit | Eric Fried proposed openstack/nova master: DRY: Build ImageMetaPropsPayload from ImageMetaProps https://review.opendev.org/698803 | |
| 15:10:25 | KeithMnemonic | dansmith: Thanks for pushing that patch through. Do you think you could please also do that last cherry pick in that set? https://review.opendev.org/#/c/690735/2 | |
| 15:11:06 | dansmith | are we still doing stuff for pike? | |
| 15:11:18 | dansmith | mriedem: | |
| 15:13:02 | mriedem | it's in EM (as is queens) so we can if people have a need, | |
| 15:13:19 | mriedem | i know suse cares about it and so does ericsson since they have products based on pike | |
| 15:13:37 | mriedem | and as you know rhosp cares about queens (or at least the internal CI/CD stuff at RHOSP cares about queens) | |
| 15:13:48 | dansmith | yeah, already did the queens one | |
| 15:14:25 | dansmith | EM means "we can throw anything in there anyone cares about without inspecting it or worrying about breakage" right? | |
| 15:14:49 | mriedem | heh, not really | |
| 15:15:03 | mriedem | it just means it's there if people want a common place to backport, we can merge things, but we don't have to | |
| 15:15:13 | dansmith | KeithMnemonic: can you please +1 those patches with "I stake my reputation on the fitness of this patch as I have tested it fully" ? | |
| 15:15:31 | mriedem | https://docs.openstack.org/project-team-guide/stable-branches.html#extended-maintenance | |
| 15:15:56 | dansmith | mriedem: right, I thought we said EM would be like a dumping ground, where we wouldn't try to keep tests working, etc | |
| 15:16:07 | mriedem | "There is no statement about the level of testing and upgrades from Extended Maintenance are not supported within the Community." | |
| 15:16:20 | mriedem | i wouldn't say we don't try, | |
| 15:16:24 | mriedem | but we don't bend over backwards | |
| 15:16:48 | mriedem | if tests are failing for some reason and there is a trivial way to fix, sure, but if things are chronically busted and no one is working on it or it's a lot of work then it's fine to EOL | |