14:02:00 #startmeeting Nova Live Migration 14:02:01 Meeting started Tue Dec 6 14:02:00 2016 UTC and is due to finish in 60 minutes. The chair is tdurakov. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:02:02 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:02:04 The meeting name has been set to 'nova_live_migration' 14:02:14 hi folks 14:02:19 o/ 14:02:24 agenda https://wiki.openstack.org/wiki/Meetings/NovaLiveMigration 14:02:28 o/ 14:02:35 let's start 14:02:43 #topic CI 14:02:48 o/ 14:02:58 so, serial console hook for live migration 14:03:11 markus_z: what options do you want to propose? 14:03:49 I have 5 possibilities identified for https://review.openstack.org/#/c/347471/39 14:04:11 ocata->ocata is working, newton->ocata is *not* working 14:04:11 markus_z: share them with us, please 14:04:19 proposal 1: let the grenade job read: nova/nova/tests/live_migration/hooks/run_tests_grenade.sh 14:04:19 right 14:04:24 proposal 2: have a "ignore-test-cases"-regex env variable only written by the grenade job 14:04:29 proposal 3: have a check in tempest which checks the oldest release 14:04:33 proposal 4: wait for the backports 14:04:37 proposal 5: wait for next release cut-off to merge the new tests 14:04:49 That's the copy&paste of what I prepared 14:04:58 ok 14:05:45 tbh, I like an idea proposed by pkoniszewski 14:05:45 The "issue" is that both jobs "normal l-m" + "grenade l-m" trigger the same "run_tests.sh". 14:05:58 Which one was that again? 14:05:59 to identify that it's a grenade job and skip tests 14:06:22 once we backport all patches required - just remove a check 14:06:46 uh, even better, I like that 14:07:01 yeah, +1 to that 14:07:10 It would be then just a conditional check in run_tests.sh ? 14:07:15 or we can just push backports forward 14:07:26 I'd not wait for backports 14:07:44 I mean there are several patches on review for serial console 14:07:58 does it really matter when the serial console hook will be merged? 14:08:00 and it would be great them to be tested asap 14:08:23 i mean, do we need to have it in a week or two? 14:08:24 pkoniszewski: Not necessarily, but that problem will come again, I think 14:09:23 if we decide to add check to skip these tests for grenade, there are 2 variables: BRANCH_OVERRIDE and DEVSTACK_GATE_GRENADE that will help to identify that we are in grenade job 14:09:23 tonyb is already a little hesitant to the backports, they are not easy stuff 14:09:47 they aren't, that's true 14:10:19 so, i'm okay with the proposal and i'm gonna support it 14:10:29 I'll prepare test to identify that that theory will work 14:10:33 +1 14:10:40 if so will update markus_z patch 14:10:49 thanks a lot 14:11:43 it's not a big deal 14:12:09 if it won't work we need another workaround, but let's check it first 14:12:41 markus_z, pkoniszewski thanks for pushing that! 14:13:21 #topic bugs 14:13:38 last week faced wih a bug https://bugs.launchpad.net/nova/+bug/1647451 14:13:38 Launchpad bug 1647451 in OpenStack Compute (nova) "Post live migration step could fail due to auth errors" [Medium,In progress] - Assigned to Timofey Durakov (tdurakov) 14:13:51 review is very welcome 14:14:37 and fyi, there is a bp https://blueprints.launchpad.net/nova/+spec/use-service-tokens for the same things 14:15:02 but I think it's worth to fix a bug anyway, because we need it to be backported 14:15:26 thoughts on that? 14:15:39 johnthetubaguy: ^ 14:20:39 do we have anything to discuss? 14:22:25 markus_z: https://review.openstack.org/#/c/407532/ 14:22:46 looks like there is nothing else for today, thanks for participating 14:23:00 #endmeeting