| @avass:vassast.org | corvus: Clark just tested 816176 and it fixes the capability issues in our openshift environment so reviews would be very appreciated :) | 07:51 |
|---|---|---|
| -@gerrit:opendev.org- Artem Goncharov proposed: [zuul/zuul] 813136: Check blocking_discussions_resolved in gitlab driver https://review.opendev.org/c/zuul/zuul/+/813136 | 12:15 | |
| -@gerrit:opendev.org- Felix Edel proposed: | 13:48 | |
| - [zuul/zuul] 814996: Make the ConfigLoader work independently of the Scheduler https://review.opendev.org/c/zuul/zuul/+/814996 | ||
| - [zuul/zuul] 816361: Instantiate configloader in zuul-web https://review.opendev.org/c/zuul/zuul/+/816361 | ||
| - [zuul/zuul] 816362: WIP: Implement job freezing API in zuul-web https://review.opendev.org/c/zuul/zuul/+/816362 | ||
| @jim:acmegating.com | Albin Vass: +2 with comment, thanks! | 17:26 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 816381: Keep consistent types on QueueItem ahead/behind attrs https://review.opendev.org/c/zuul/zuul/+/816381 | 17:36 | |
| @jim:acmegating.com | fungi, Clark, tobiash, swest, felixedel: ^ that should address the other issue we discovered over the weekend | 17:37 |
| @clarkb:matrix.org | corvus: responded with a question | 17:55 |
| @clarkb:matrix.org | er +2'd with a question | 17:55 |
| -@gerrit:opendev.org- Andre Aranha proposed: [zuul/zuul-jobs] 816385: Add enable-fips as false to unittest job https://review.opendev.org/c/zuul/zuul-jobs/+/816385 | 17:59 | |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/nodepool] 816389: Fix openshift dep for arm64 docker image builds https://review.opendev.org/c/zuul/nodepool/+/816389 | 18:28 | |
| @clarkb:matrix.org | Zuulians ^ I discovered that issue just now. I think it worked before because it was able to find an openshift 0.8.6 wheel on the opendev wheel mirror for buster. But bullseye doesn't have any openshift wheels. Bumping the dep up is a good thing anyway since we can't install with the oldest openshift library available | 18:30 |
| @clarkb:matrix.org | The planned improvments for image building will address this more properly. But this should be safe and correct for now as well | 18:30 |
| @clarkb:matrix.org | I'm also happy to lower the lower bound if anyone knows what an appropriate version is | 18:30 |
| @jim:acmegating.com | Clark: thx, responded | 18:33 |
| @clarkb:matrix.org | corvus: thanks it didn't seem critical but also thought it might've been overlooked so wanted to call it out. If that was intention (seems to be) then should be fine | 18:33 |
| @jim:acmegating.com | Clark: yep. i totally flipped that coin before uploading. :) | 18:34 |
| @jim:acmegating.com | Clark: will the arm64 change make nodepool image builds slow? | 18:38 |
| @clarkb:matrix.org | corvus: that is the one risk. I don't think it will beacuse I don't think openshift has any non python building requirements to make a wheel | 18:39 |
| @jim:acmegating.com | ok, i guess we'll know for sure when the image build job finishes :) | 18:39 |
| @clarkb:matrix.org | corvus: it should be a trivial wheel build. But the image build job for that change should give us confirmation of that (we can examine the build log) | 18:40 |
| @bridgefan:matrix.org | Sorry to interrupt (one day I will hopefully answer more questions than I ask), but does zuul have a way to deal with jobs that need 2 branches of the same repo? | 18:42 |
| @jim:acmegating.com | bridgefan: you can just check them out in the job | 18:43 |
| @jim:acmegating.com | bridgefan: the repo state zuul prepares has the correct speculative future state of every branch in the repo | 18:44 |
| @clarkb:matrix.org | bridgefan: it does. By default zuul will curate the state for all the branches of the repo. Then in your job you can check out the appropriate branch at different times. Openstack does this for upgrade testing. Check out the old branch and install. Run tests. Then check out the new branch and upgrade. Run tests. | 18:44 |
| @jim:acmegating.com | so if you have a change on 'branch1' in the queue and then a change on 'branch2' behind it, both changes will be in the repo on disk. | 18:45 |
| -@gerrit:opendev.org- Zuul merged on behalf of Albin Vass: [zuul/zuul] 816176: Drop ambient capabilities when running bwrap https://review.opendev.org/c/zuul/zuul/+/816176 | 18:46 | |
| @bridgefan:matrix.org | ok thanks | 18:50 |
| @jim:acmegating.com | Clark, fungi, tobiash, swest, felixedel: i restarted opendev earlier today to try to catch the changecache bug. i caught a changecache bug, but not the same one as before (but it is possible they have the same underlying cause). the issue was caused by pruning the change cache during a reconfiguration event. i suspect it gathered the list of relevant changes when far fewer items (or perhaps no items!) were in the queues. so we'll either need to lock that with mutexes, or perhaps just have the apscheduler timers put events into the management queue and have that act on a management event (so it happens in the main thread) | 21:34 |
| @jim:acmegating.com | i think fixing that is the next step, then we can see if we see the "other" bug again | 21:35 |
| @jim:acmegating.com | (the new change cache logging was helpful, and not exceedingly verbose, so i think it'll be good to keep that) | 21:35 |
| @clarkb:matrix.org | yay for being able to track down the bug we hit with the new logging | 21:57 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 816413: Fix change cache relevance detection for refs https://review.opendev.org/c/zuul/zuul/+/816413 | 22:03 | |
| @jim:acmegating.com | we have not seen that ^ bug, but i noticed it while i was looking into the ones we did see | 22:04 |
| @clarkb:matrix.org | corvus: left a nit on that I'll let you decide if you want to approve or fix it (tobiahs already +2'd) | 22:45 |
| @clarkb:matrix.org | I think https://zuul.opendev.org/t/zuul/build/1adef12661804ffb89edb98fe576f160/log/job-output.txt#5178-5183 shows it took 6 seconds to make the arm64 wheel for openshift | 22:49 |
| @clarkb:matrix.org | I'll approve that change once check testing passes as that should be a non issue | 22:49 |
| @jim:acmegating.com | Clark: thanks. i agree it's a little weird, but it was intentional, so i went ahead and +wd | 23:20 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!