| jeblair | adam_g: i want to say that since in v2, the merge conflict detection happens before the jobs are launched, that ensures that A and C will end up in the queue before we release the jobs. but in v3, the conflict detection happens after the jobs are launched, which means when you release it, B will fail, C will restart, and if that happens after A merges, you'll end up with the output you pasted (C tested without A -- but only because A was ... | 00:01 |
|---|---|---|
| jeblair | ... already merged) | 00:01 |
| jeblair | adam_g: i am not 100% sure about that, but if so, doing the stepwise release of the -merge jobs like in the old change would probably preserve the behavior (A would still be running when C is re-enqueued after B fails out, so you should see A and C together in the history) | 00:02 |
| adam_g | jeblair: ok that all makes some sense. it looks like the merge conflict detection isn't happening ATM without the little change in that patch to the launcher | 00:03 |
| adam_g | i couldn't get the stepwise releases to work as they did in v2 but ill play with what im looking at in test_parallel_changes. | 00:03 |
| jeblair | adam_g: yeah, i agree that fix is probably necessary. re stepwise -- since you're switching to holding the jobs in the launch_server, you'll need to use the launch_server release, which i think just "self.release('.*-merge')" will do. | 00:06 |
| jeblair | adam_g: er, looks like we're preferring self.launch_server.release | 00:07 |
| adam_g | jeblair: yep | 00:07 |
| adam_g | jeblair: OK! the stepwise releases helped there | 00:11 |
| jeblair | yay! | 00:11 |
| adam_g | i may keep the patch as-is without them, though. im not sure it provides much value at this point if its a non-standard behavior | 00:13 |
| openstackgerrit | Adam Gandelman proposed openstack-infra/zuul: Re-enable test_build_configuration_conflict https://review.openstack.org/409376 | 00:20 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul: WIP organize connections into drivers https://review.openstack.org/408849 | 00:24 |
| openstackgerrit | Jerry Zhao proposed openstack-infra/zuul: add check if an event matches the connection which triggered it. https://review.openstack.org/409397 | 03:17 |
| *** harlowja_ has quit IRC | 03:36 | |
| *** jeblair_ has joined #zuul | 03:37 | |
| *** pleia2_ has joined #zuul | 03:37 | |
| *** pleia2 has quit IRC | 03:38 | |
| *** jeblair has quit IRC | 03:38 | |
| *** hogepodge has quit IRC | 05:06 | |
| *** saneax-_-|AFK is now known as saneax | 06:20 | |
| *** saneax is now known as saneax-_-|AFK | 07:01 | |
| *** saneax-_-|AFK is now known as saneax | 07:02 | |
| *** pleia2_ has quit IRC | 07:28 | |
| *** pleia2 has joined #zuul | 07:29 | |
| *** jeblair_ has quit IRC | 08:22 | |
| *** jeblair has joined #zuul | 08:23 | |
| *** saneax is now known as saneax-_-|AFK | 08:52 | |
| *** Cibo_ has quit IRC | 11:32 | |
| *** Cibo_ has joined #zuul | 11:41 | |
| *** pleia2 has quit IRC | 12:51 | |
| *** cinerama has quit IRC | 12:51 | |
| *** mmedvede has quit IRC | 12:51 | |
| *** pleia2 has joined #zuul | 12:51 | |
| *** cinerama has joined #zuul | 12:51 | |
| *** mmedvede has joined #zuul | 12:51 | |
| dmsimard | jeblair: I'm curious to know how Zuul is planned to be licensed. Is it going to stick with Apache v2 and keep the callbacks/plugins GPLv3 ? | 14:20 |
| dmsimard | fungi: ^ | 14:21 |
| mordred | dmsimard: that is my understanding of the current plan, yes | 15:28 |
| dmsimard | Ok, I'll align ara in a similar fashion .. callbacks and plugins were created with an apache v2 header from inception due to ignorance | 15:29 |
| *** bhavik has joined #zuul | 16:12 | |
| openstackgerrit | Merged openstack-infra/zuul: Re-enable requirement vote tests https://review.openstack.org/401061 | 17:45 |
| pabelanger | jeblair: Shrews: Now that we are running feature/zuulv3 in production for nodepool what is the plan for the feature/zuulv3 branch? Are we thinking of merging to master, then dropping it, and propose patches directly to master? Or keep it alive for a little longer? | 17:53 |
| *** bhavik has quit IRC | 18:25 | |
| mordred | pabelanger: I think the plan is to let it burn in for just a bit, then merge it back to master (switching puppet to run from master) and start working on the node-launcher bits in the zuulv3 branch | 18:52 |
| *** jesusaur has joined #zuul | 19:24 | |
| *** jeblair has quit IRC | 21:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!