18:02:42 #startmeeting networking_policy 18:02:43 Meeting started Thu Oct 5 18:02:42 2017 UTC and is due to finish in 60 minutes. The chair is SumitNaiksatam. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:02:44 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:02:45 tbachman: hi 18:02:46 The meeting name has been set to 'networking_policy' 18:02:48 SumitNaiksatam: hi! 18:02:51 sorry I’m late 18:03:19 #topic Zuul v3 migration 18:03:24 tbachman: i am too :-) 18:03:29 lol 18:03:46 * tbachman notes that can be interpreted two different ways…. picks the more favorable one ;) 18:03:57 so for now i think things are back to fairly normal, with the rollback to v2 18:03:59 tbachman: lol 18:04:14 even the devstack job is working 18:04:19 SumitNaiksatam: nic! 18:04:21 nice! 18:04:42 i did not see any new updates on when the zuul v3 migration will be attempted again 18:05:02 i think its a couple of weeks out? 18:05:35 they mentioned a few weeks, yes 18:05:38 meanwhile, just wanted to bring it up in the meeting so that we can be prepared better the next time around 18:05:46 SumitNaiksatam: Is there anything we should do before then, other than merging anything that is ready? 18:05:50 annakk: yeah, i recall seeing that in one of the emails 18:06:01 rkukura: exactly, my question too :-) 18:06:23 part of the issue is that we dont know what to expect :-) 18:06:44 my understanding from the earlier iteration was that the jobs were getting migrated automatically 18:06:57 so i at least expected our standard py27 job to work 18:07:04 but that didnt happen either 18:07:27 i guess we just need to watch out for any more emails that are sent out on this 18:07:54 ok 18:09:08 i didn't read the emails close enough to understand what went wrong.. 18:09:24 annakk: i am not sure the emails had those details 18:09:38 but i think some job definitions also did not work 18:09:56 in our case we were seeing a wierd issue with the devstack install 18:10:14 as a matter of being good citizens, i think we should not recheck too many patches when the migration happens again 18:10:26 we should ideally recheck on one patch GBP patch 18:10:35 good point 18:10:35 for that purpose i used to have this patch: 18:10:41 #link https://review.openstack.org/#/c/208180/ 18:10:45 i will rebase it now 18:11:01 (its easy to rebase this from the web gerrit) 18:11:08 and then we can try rechecking this patch 18:11:21 because this patch does not add any new code 18:11:31 so if it passes then we can start dealing with other pending patches 18:11:49 this is just a suggestion obviously 18:12:09 last week we desperately needed a few patches to be merged 18:12:21 so i was guilty of frequent recheck 18:12:26 *rechecks 18:12:44 but i will try to avoid in the future 18:12:58 #topic Open Discussion 18:13:10 i think most of the pending patches right now are driver specific 18:13:11 * tbachman notes at some point we should go through the gate logs and look for exceptions 18:13:26 tbachman: oh yeah good point 18:13:42 we need to get back to that activity 18:13:44 I think I’ve seen a few, which I’m wondering whether they are delaying tests 18:13:55 oh :-( 18:14:09 also some more deprecation warnings, in preparation for pike, queens.... 18:14:17 tbachman: right 18:14:27 also annakk’s earlier question on moving to pike 18:14:32 ack 18:14:43 would be good to be on par with upstream 18:14:48 i think we had pushed the discussion to this week 18:16:05 on our side we need pike in order to start doing some service chaining (relevant services are only available in pike..) 18:16:07 i think it would make sense to clean up the deprecation warnings and other exceptions we see in the logs 18:16:15 annakk: yes sure 18:16:47 I'll continue looking at the exceptions 18:16:53 annakk: great thanks 18:17:06 its good to know that you need pike urgently 18:17:18 in which case we should prioritze accordingly 18:17:34 (i guess we should have been doing it anyway) 18:17:51 should we split up the work in terms of cleaning up the logs? 18:18:29 I'll look at the remaining SubnetInUse 18:18:35 annakk: ok good 18:18:49 i will look for others and coordinate with rkukura and tbachman 18:19:12 OK, I had seen a couple I might attack, but don’t remember which 18:19:14 SumitNaiksatam: will be glad to help 18:19:20 rkukura: tbachman great thanks 18:19:32 i think there are 2 or 3 which we need to address right away 18:19:39 they complete inundate the logs 18:19:43 *completely 18:20:03 there is one more UI patch that i still need to review 18:20:20 #link https://review.openstack.org/#/c/497197/ 18:20:27 but it has a merge conflict 18:20:31 i will reach out to the author 18:20:53 will need one more reviewer, as with others this has to be tested to ensure that it works 18:21:27 anything else we need to discuss today? 18:21:53 not from me.. 18:21:54 alrighty, thanks all for joining! 18:22:02 SumitNaiksatam: thanks! 18:22:03 rkukura: tbachman annakk: bye! 18:22:05 thanks! 18:22:06 bye! 18:22:09 #endmeeting