| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2021-04-01 | |||
| 09:17:07 | kashyap | Also, a gentle reminder: many virt and kernel bugs get first fixed in Fedora | |
| 09:17:09 | sean-k-mooney | i have hit a few other things like that that im hoping to avoid | |
| 09:17:28 | sean-k-mooney | kashyap: yep but i very realy hit those | |
| 09:17:33 | stephenfin | kashyap: You'd swear you worked for Red Hat or something 0:) | |
| 09:17:43 | kashyap | stephenfin: LOL, it's not about Red Hat, really :) | |
| 09:17:56 | sean-k-mooney | kashyap: what i have hit is selinux being unhappy with me on fedora alot | |
| 09:17:57 | kashyap | stephenfin: I'm speaking with my upstream hat, really :) | |
| 09:18:25 | lyarwood | sean-k-mooney: there's a command and a shiny website for that problem | |
| 09:18:40 | lyarwood | makes someone cry however so you might not want to try it | |
| 09:18:51 | sean-k-mooney | set it to permissive mode | |
| 09:19:01 | kashyap | stephenfin: Also, I closely w/ the virt upstreams and downstreams to the point that I feel part of those teams too :) | |
| 09:19:24 | lyarwood | sean-k-mooney: https://stopdisablingselinux.com/ | |
| 09:19:30 | kashyap | sean-k-mooney: It's not about bugs, per se. Even new virt features first land in Fedora, BTW :) | |
| 09:19:36 | kashyap | And kernel, of course | |
| 09:19:51 | lyarwood | but yeah selinux can be a royal PITA | |
| 09:20:15 | sean-k-mooney | lyarwood: well i have tried using the command to updat the policy that you get prometed with when there is a failure by the way | |
| 09:20:56 | kashyap | lyarwood: Go tell that to Red Hat It's part of the value prop ;-) | |
| 09:21:17 | sean-k-mooney | kashyap: sure but i dont want bleeding edge for my infra i was new but stable which is why i normally go with debiab/ubuntu distros with the mainline kernel | |
| 09:21:33 | lyarwood | it can add value and still be a PITA to use ;) | |
| 09:21:43 | sean-k-mooney | centos with some tweeks i coudl live with | |
| 09:21:54 | sean-k-mooney | fedora not so much for anything i want to not have to maintain too much | |
| 09:22:01 | kashyap | lyarwood: Heh, sure | |
| 09:22:48 | kashyap | lyarwood: It took me and RHT virt team several weeks to debug a crazy VMs+container interaction bug. | |
| 09:23:02 | sean-k-mooney | kashyap: lyarwood by the way the main thing that anowyed me about selinux recently is it does not allow you to use iso form your home directoy to create vms in virt-manager | |
| 09:23:06 | kashyap | lyarwood: Here's my summary notes, for your "bedtime SELinux reading": https://kashyapc.fedorapeople.org/SELinux_libvirt_and_QEMU_in_a_container.html | |
| 09:23:50 | sean-k-mooney | i rememebr you working on that ya | |
| 09:24:28 | lyarwood | sean-k-mooney: yeah I've updated policykit to launch domains as my normal user without sudo but I still copy disks into /var/lib/libvirt/images etc | |
| 09:25:07 | sean-k-mooney | to be fair apparmor is not nessisarly any eaiser to debug or work with so i think any security system is going to be complex by nature of all the context they need to work and that you need to understand them | |
| 09:27:34 | sean-k-mooney | its those kind of things that are unfreidly to new comers and make distos feel unpolished however | |
| 09:27:56 | sean-k-mooney | i kind of get why you might not want libvirt or qemu being able to read/write your home directory | |
| 09:28:42 | sean-k-mooney | but since i can add it as a location in virt manger and select the image i would expect it to work or give me an error when i try to add the storage pool location | |
| 09:29:07 | lyarwood | remind me again, where are the irc chat logs for this channel? | |
| 09:29:22 | sean-k-mooney | ill get it one sec | |
| 09:29:33 | sean-k-mooney | http://eavesdrop.openstack.org/irclogs/%23openstack-nova/ | |
| 09:29:56 | lyarwood | ta | |
| 09:44:31 | sean-k-mooney | gibi: shall i prepare the patch to move the implemtned specs to the correct directory? have the blueprints been updated in launchpad? | |
| 09:55:12 | sean-k-mooney | im mostly done i need to fix a few files names too wehre the file name did not match the blueprint name | |
| 10:06:00 | sean-k-mooney | lyarwood: mind if i create a second blueprint for the linux part of the ephemeral sotrage encyption | |
| 10:06:18 | sean-k-mooney | lyarwood: the spec repo tooling assume the file name of the spec matches the repo | |
| 10:06:23 | sean-k-mooney | *blueprint | |
| 10:06:32 | sean-k-mooney | but we have 2 specs for the saem bluepinrt | |
| 10:06:59 | sean-k-mooney | i can also just move or in this case not move that one by hand | |
| 10:07:22 | sean-k-mooney | which ever you prefer | |
| 10:07:57 | lyarwood | sean-k-mooney: yeah I don't mind having a second blueprint for the libvirt part | |
| 10:08:11 | lyarwood | sean-k-mooney: keeps the bean counters happy :) | |
| 10:08:25 | sean-k-mooney | ok ill creeated and make one depend on the the other | |
| 10:08:40 | sean-k-mooney | * create it | |
| 10:09:49 | lyarwood | cool TIL you can do that with blueprints | |
| 10:10:32 | sean-k-mooney | yep we rarely do and i have not don it in a very long time so trying to rememebr how | |
| 10:11:36 | gibi | sean-k-mooney: thanks for preparing that patch I forgot | |
| 10:12:06 | sean-k-mooney | no worries i did it last time i think and i have done it at least one other time so i dont mind the script does most of the work | |
| 10:12:28 | sean-k-mooney | just fixing the warnign form the inccorect file names now | |
| 10:14:29 | gibi | thanks | |
| 10:19:47 | sean-k-mooney | lyarwood: there https://blueprints.launchpad.net/nova/+spec/ephemeral-encryption-libvirt you can see the depency tree at the bottom | |
| 10:20:49 | sean-k-mooney | we dont ofter use that feature but when doing cross project stuff we proably shoudl use it more | |
| 10:39:36 | openstackgerrit | sean mooney proposed openstack/nova-specs master: Move the implemented specs for the wallaby release https://review.opendev.org/c/openstack/nova-specs/+/784369 | |
| 10:41:17 | sean-k-mooney | gibi: so assuming the blueprint status is correcct that should be correct | |
| 10:41:33 | sean-k-mooney | looking at the list we had a fairly good completion ratio | |
| 10:49:48 | gibi | yepp we had ~70% | |
| 11:21:20 | artom | Wait, how the hell did Zuul not yet run on a patch I proposed last evening? | |
| 11:21:46 | gibi | artom: there was some outage | |
| 11:21:48 | sean-k-mooney | zuul restart | |
| 11:22:03 | gibi | do a recheck to queue it | |
| 11:22:05 | sean-k-mooney | all pending jobs got lost so just recheck | |
| 11:22:09 | artom | Ah | |
| 11:22:59 | sean-k-mooney | it was mention on th emailing list but i recall was " something something ooo ...." | |
| 11:23:03 | sean-k-mooney | then i stopped caring | |
| 11:26:09 | artom | Lulz | |
| 11:35:20 | openstackgerrit | Merged openstack/nova-specs master: Move the implemented specs for the wallaby release https://review.opendev.org/c/openstack/nova-specs/+/784369 | |
| 11:55:21 | openstackgerrit | Merged openstack/nova master: docs: Rewrite 'nova-manage db' docs to use proper directives https://review.opendev.org/c/openstack/nova/+/677509 | |
| 11:56:25 | openstackgerrit | Merged openstack/nova master: docs: Rewrite 'nova-manage api_db', 'cell_v2' docs https://review.opendev.org/c/openstack/nova/+/726192 | |
| 11:57:15 | openstackgerrit | Merged openstack/nova master: docs: Rewrite 'nova-manage placement' docs https://review.opendev.org/c/openstack/nova/+/726193 | |
| 11:58:16 | openstackgerrit | Merged openstack/nova master: docs: Rewrite 'nova-manage libvirt' docs https://review.opendev.org/c/openstack/nova/+/781504 | |
| 12:38:17 | gibi | cores: my today's doctor's appointment got pushed back to 17:50 CEST which means I cannot chair today's nova meeting. I updated the agenda on the wiki, and I might be able to follow the meeting. Who will volunteer to chair the meeting? | |
| 12:55:03 | gibi | lyarwood, bauzas, stephenfin, ^^ any volunteers :) | |
| 12:55:15 | bauzas | gibi: i can do | |
| 12:55:24 | gibi | bauzas: thanks, it is sold to you then :) | |
| 12:55:34 | bauzas | 18:00 CEST, right? | |
| 12:55:39 | gibi | 18:00 CEST yes | |
| 12:55:43 | bauzas | kk | |
| 12:56:01 | bauzas | gibi: any agenda already ? | |
| 12:56:39 | gibi | bauzas: the agenda is up to date on the wiki https://wiki.openstack.org/wiki/Meetings/Nova#Agenda_for_next_meeting | |
| 13:02:11 | bauzas | gibi: all good, thanks | |
| 13:02:50 | bauzas | gibi: fwiw, as schools are closing next week, I'll take 2 days off on next Thursday and Friday | |
| 13:03:02 | gibi | bauzas: ack, | |
| 13:03:04 | bauzas | (and Monday is a bank holiday) | |
| 13:03:31 | bauzas | then, children will be on vacations for 2 weeks and then back to schools (hopefully) | |
| 13:10:44 | stephenfin | gibi: If you've a chance, could you take a look at https://review.opendev.org/c/openstack/nova/+/726195/ and the other few patches in that series today? Would be nice to close that out | |
| 13:14:09 | gibi | stephenfin: will try | |
| 13:14:23 | stephenfin | good enough for me. Thanks | |
| 13:46:25 | yonglihe | gibi, alex, xinren repropose the spec, hope you have time slot for that. btw, code alse rebased and addressed comments: Repropose smartnic support spec https://review.opendev.org/c/openstack/nova-specs/+/783632 | |
| 13:46:35 | gibi | yonglihe: ack | |
| 13:47:32 | yonglihe | thanks, i gonna run, have a good day. | |
| 13:49:55 | gmann | sean-k-mooney: lyarwood gibi basically if new return error code is new to that API (not from existing one) then we need microversion bump. and these are general return error code for all APis 400, 403, 404 and 415 | |
| 13:51:00 | gmann | sean-k-mooney: lyarwood gibi which APIs is this? I think most of server APIs have 409 exist return code | |
| 13:55:11 | gibi | os-services/{id}/force_doewn PUT | |
| 13:59:27 | gmann | yeah 409 is not there https://github.com/openstack/nova/blob/50fdbc752a9ca9c31488140ef2997ed59d861a41/nova/api/openstack/compute/services.py#L358 | |
| 14:02:28 | gibi | yepp | |
| 14:02:43 | gibi | I accept 400 as a compromise for backportability | |
| 14:03:17 | gmann | +1 | |