| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-03-08 | |||
| 11:38:25 | sean-k-mooney | ha i could use the old microverion yes | |
| 11:39:03 | bauzas | AFAIR, we still need to use the force flag or just using a specific aggregate for forcing to use a compute | |
| 11:39:06 | sean-k-mooney | ya ok ill do that and ill see if we can de evacuate with osc while ime at it | |
| 11:39:14 | bauzas | sean-k-mooney: thanks | |
| 11:43:59 | openstackgerrit | Adam Spiers proposed openstack/nova master: Use assertXmlEqual() helper for all XML comparison tests https://review.openstack.org/641852 | |
| 11:44:35 | aspiers | stephenfin: https://review.openstack.org/#/c/641852/2..3/nova/test.py | |
| 11:48:12 | sean-k-mooney | dumb question but --os-compute-api-version is used to set the microverion right | |
| 11:48:31 | sean-k-mooney | wew dont use the term micro version in the help text | |
| 11:52:18 | openstackgerrit | sean mooney proposed openstack/nova master: update gate test for removal of force evacuate https://review.openstack.org/641986 | |
| 12:20:39 | kashyap | mdbooth: On the off-chance you have time today: https://review.openstack.org/#/c/641981/ | |
| 12:20:54 | kashyap | FWIW, the code comment I wrote in there, I got it reviewed by Kevin Wolf of QEMU. | |
| 12:25:20 | shilpasd | smcginnis: Hi, need review for openstack/releases patch for masakariclient https://review.openstack.org/#/c/641988/1 | |
| 12:25:43 | shilpasd | dtantsur: Hi, need review for openstack/releases patch for masakariclient https://review.openstack.org/#/c/641988/1 | |
| 12:28:52 | dtantsur | shilpasd: hi, why do you want a review from me on that? I have no relationship to either releases or masakari (or nova) | |
| 12:30:03 | shilpasd | dtantsur: ohh, thanks for confirmation, please ignore the request | |
| 12:31:52 | openstackgerrit | Adam Spiers proposed openstack/nova-specs master: Re-approve AMD SEV support for Train https://review.openstack.org/641994 | |
| 12:33:28 | kashyap | aspiers: s/Stein/Train/ in the commit message :-) | |
| 12:35:30 | bauzas | sean-k-mooney: yup, that's the field to use | |
| 12:39:19 | aspiers | kashyap: doh :) | |
| 12:39:42 | openstackgerrit | Adam Spiers proposed openstack/nova-specs master: Re-approve AMD SEV support for Train https://review.openstack.org/641994 | |
| 12:48:17 | shilpasd | smcginnis: Thank you for workflow +1 for masakariclient https://review.openstack.org/#/c/641988/1 | |
| 12:51:23 | smcginnis | shilpasd: No problem. Probably best not to use the nova channel to discuss release activities for masakari though. ;) | |
| 12:52:31 | shilpasd | smcginnis: ok, will take a note of it | |
| 13:11:47 | cdent | gibi: is the rest of the bandwidth resource provider expected to land or is it being kicked to train? | |
| 13:12:38 | sean-k-mooney | cdent: on monday i think there were only 3 patches left for stien | |
| 13:12:44 | sean-k-mooney | the rest was going to be in train | |
| 13:12:52 | sean-k-mooney | i dont know if we got those 3 in | |
| 13:13:44 | sean-k-mooney | https://review.openstack.org/#/c/636360/ | |
| 13:13:50 | cdent | thanks sean-k-mooney. trying to make some notes about it in the pupdate | |
| 13:13:55 | sean-k-mooney | cdent: ^ that was the patch that enabled the feature i think] | |
| 13:14:51 | sean-k-mooney | so if i was following correctly we have landed enough for spwaning vms with bandwith request | |
| 13:15:05 | sean-k-mooney | all move opertaions are puntted to train | |
| 13:16:03 | gibi | cdent: the heal_allocation part is train (but will be backportad as that is a nova-manage extension). starts here: https://review.openstack.org/#/c/637953 | |
| 13:16:24 | cdent | thanks gibi | |
| 13:16:40 | sean-k-mooney | gibi: we are going to make this as compelte right https://blueprints.launchpad.net/nova/+spec/bandwidth-resource-provider | |
| 13:16:41 | gibi | cdent: there is two bug like things https://review.openstack.org/#/c/639608/ | |
| 13:16:48 | gibi | cdent: https://review.openstack.org/#/c/638711/ | |
| 13:16:54 | sean-k-mooney | and add a new spec/blueprint for the final bits for train? | |
| 13:17:08 | gibi | sean-k-mooney: I'm not sure, the original spec talked about serve move operations which was not finished (not even started) in Stein | |
| 13:17:29 | gibi | sean-k-mooney: I have to talk to melwitt and mriedem about the administration | |
| 13:17:39 | sean-k-mooney | sure | |
| 13:17:54 | gibi | cdent: and there is test coverage extension and refactor here https://review.openstack.org/#/c/641679/ | |
| 13:18:03 | sean-k-mooney | but enough has lannded that wyou can boot a vm with a min bandwith request now? | |
| 13:18:04 | gibi | sean-k-mooney: I'm OK in both ways | |
| 13:18:19 | gibi | sean-k-mooney: yeas, boot, delete, and detach port works | |
| 13:18:27 | cdent | \o/ | |
| 13:19:00 | sean-k-mooney | cool :) i need to go test that i guess but that is good to see | |
| 13:20:05 | sean-k-mooney | gibi: we dont have any tempest tests for this right? | |
| 13:20:30 | sean-k-mooney | we can praobly add them to the numa/nfv test discussion at the ptg | |
| 13:26:47 | openstackgerrit | Surya Seetharaman proposed openstack/nova master: Delete orphaned RS and IM records of archived instances https://review.openstack.org/560042 | |
| 13:30:55 | openstackgerrit | Kashyap Chamarthy proposed openstack/nova master: libvirt: Rework 'EBUSY' (SIGKILL) error handling code path https://review.openstack.org/639091 | |
| 13:34:29 | gibi | sean-k-mooney: there is a tempest test proposed https://review.openstack.org/#/c/629253/ | |
| 13:35:15 | gibi | sean-k-mooney: today I finally able too make an env with real SRIOV so I tested that the SRIOV part of the feature works | |
| 13:35:24 | gibi | so far so good | |
| 13:39:55 | mriedem | i heard the gate is broken? | |
| 13:42:12 | mriedem | ah shit i see | |
| 13:42:28 | sean-k-mooney | mriedem: i proposed a patch for the --force issue | |
| 13:42:32 | mriedem | link? | |
| 13:42:39 | mriedem | bauzas: i bit my own ass http://logs.openstack.org/50/633350/16/check/nova-live-migration/86f6551/job-output.txt.gz#_2019-03-08_09_20_42_237644 | |
| 13:42:39 | sean-k-mooney | https://review.openstack.org/#/c/641986/1 | |
| 13:42:48 | sean-k-mooney | wrong link | |
| 13:42:56 | sean-k-mooney | oh no never mind | |
| 13:43:18 | sean-k-mooney | ya that was the failure | |
| 13:43:26 | bauzas | mriedem: yup, sean-k-mooney discovered it | |
| 13:43:51 | mriedem | is there a bug? | |
| 13:44:04 | sean-k-mooney | well technically one of the sriov live migration ptachs discovered it when it was running its gate job | |
| 13:44:17 | sean-k-mooney | no but i can open one quickly | |
| 13:44:24 | sean-k-mooney | then respin the patch | |
| 13:45:12 | mriedem | i'm opening one | |
| 13:45:15 | mriedem | just for e-r | |
| 13:45:45 | sean-k-mooney | ok ill stop creating one so. | |
| 13:46:08 | sean-k-mooney | i also looked and osc still does not have evacuate support | |
| 13:46:23 | sean-k-mooney | do we want to go add it? | |
| 13:46:29 | openstackgerrit | Matt Riedemann proposed openstack/nova master: update gate test for removal of force evacuate https://review.openstack.org/641986 | |
| 13:47:20 | mriedem | yes osc should have evacuate support at some point | |
| 13:47:37 | sean-k-mooney | e.g. i will be on a plane for a few hours tomorrow should i go play with trying to add it | |
| 13:49:56 | sean-k-mooney | ill pull down osc anyway just to have it. does anyone know if they swapped to useing the sdk yet? | |
| 13:53:26 | mriedem | not for nova | |
| 13:53:30 | mriedem | but some things yes, like neutron i think | |
| 13:57:03 | sean-k-mooney | mriedem: by the way what does e-r stand for. | |
| 13:57:09 | sean-k-mooney | oh elastic recheck | |
| 13:57:27 | mriedem | yeah | |
| 14:08:46 | sean-k-mooney | out of interest if i was to add evacuate to osc, would people object to me calling it "recreate" e.g. openstack server recreate or somethign closer to what it actully does? | |
| 14:09:36 | edleafe | sean-k-mooney: generally the osc names are supposed to be saner than the names in the python-*client libraries | |
| 14:10:36 | sean-k-mooney | ya what i would personally prefer is 3 commands | |
| 14:10:50 | sean-k-mooney | openstack server recreate e.g. nova evacuate | |
| 14:11:04 | sean-k-mooney | openstack host evacuate e.g. nova host-evecuate | |
| 14:11:26 | sean-k-mooney | and opentack host evacuate --live fo nova host-evacuate-live | |
| 14:12:54 | sean-k-mooney | or actully it could be openstack host migrate and openstack host migrate --live | |
| 14:13:53 | sean-k-mooney | in anycase i have never modifed osc before so ill jsut start with porting nova evacuate and see how things go | |
| 14:14:04 | bauzas | sean-k-mooney: openstack host evacuate doesn't exist from an API PoV | |
| 14:14:11 | bauzas | it's just a client thingy | |
| 14:14:15 | edleafe | yeah, I'm not sure which terms would make the most sense. But yes, names that actually convey what the call does are preferred. :) | |
| 14:14:15 | sean-k-mooney | bauzas: yes i know | |
| 14:14:20 | bauzas | so, no to be an OSC CLI command | |
| 14:15:13 | sean-k-mooney | bauzas: is that a rule that osc cant have things that are not directly supported by the api | |
| 14:15:38 | bauzas | good question | |
| 14:15:57 | sean-k-mooney | but ok i was going to look at those after so no worries ill just focus on nova evacuate first ot figure out how things work | |
| 14:15:58 | bauzas | but honestly, I hate the host-evacuate method | |
| 14:17:23 | sean-k-mooney | ya i have personally neverr used it but i have used the horizon one in the past but mainly for testing | |