| opendevreview | Ian Wienand proposed openstack/project-config master: Pause Fedora 34 builds https://review.opendev.org/c/openstack/project-config/+/817318 | 00:10 |
|---|---|---|
| opendevreview | Merged openstack/project-config master: Pause Fedora 34 builds https://review.opendev.org/c/openstack/project-config/+/817318 | 00:24 |
| EmilienM | fungi: hey, when moving a repo under opendev, is it possible to maintain a mirror in its original namespace (e.g. foo/bar, moving to opendev/bar, but we want to maintain the foo/bar updated from opendev). | 01:12 |
| fungi | EmilienM: yep, basically just need to give the mirroring job a zuul secret of some account credentials which have write access to wherever it's going to mirror | 01:13 |
| EmilienM | fungi: cool. Bonus question: it mirror tags as well? | 01:15 |
| fungi | yep | 01:17 |
| fungi | EmilienM: https://github.com/airshipit/airshipctl/tags is an example | 01:18 |
| EmilienM | ok, thank you | 01:18 |
| fungi | in fact, the job should be able to mirror to any writeable git remote | 01:18 |
| EmilienM | fungi: ok a last one then I'll stop :D | 01:18 |
| fungi | doesn't have to be github | 01:18 |
| EmilienM | gophercloud currently uses openlab, and they use a fork of zuul jobs, which aren't quite how they are now in opendev. So the migration wouldn't be just a switch of zuul but also a rewrite of the jobs. This is not a big deal as we don't have that many jobs, but for a smooth transition, I wish we had a fork of the repo already under opendev (or anywhere where zuul opendev can | 01:20 |
| EmilienM | control), so we can already work on the new zuul layout, and when we're ready we can make the switch and rebase the fork. | 01:20 |
| EmilienM | I don't know if I explained it well but I think you got the idea | 01:20 |
| fungi | makes sense, yeah | 01:21 |
| EmilienM | how would it be possible? (in a nutshell) | 01:22 |
| fungi | import the codebase, then work on a change to rewrite the jobs, merge it once it passes | 01:23 |
| fungi | any jobs which would prevent merging a change should run speculatively on the change which alters them anyway | 01:24 |
| EmilienM | ok, and once we make the switch, we would cherry-pick what merged into original repo, in the new repo, and then enable mirroring? | 01:25 |
| fungi | you'd probably want to not merge anything new to github once it's imported to opendev, so that as soon as mirroring is set up it'll be fast-forwardable for anyone with existing github remotes | 01:26 |
| fungi | though if you manually push the same commits to github which merge to opendev in the meantime that would work | 01:28 |
| EmilienM | I see | 01:28 |
| EmilienM | fungi: we would need to be super careful; the repo has 400+ forks and is used by 5000+ libraries on github :D | 01:28 |
| fungi | yeah, i don't expect they'd see any difference, the commits/tags already there will still be there once mirroring | 01:29 |
| fungi | though you'll likely want to clean up any unwanted branches before importing | 01:30 |
| EmilienM | yeah right | 01:31 |
| EmilienM | "you'd probably want to not merge anything new to github once it's imported to opendev" > that's my point, I think this will have to happen. Because there will be some time between the import & having working zuul jobs | 01:34 |
| EmilienM | but we'll manually push the missing commits into gerrit | 01:35 |
| fungi | yeah, but if you keep merging things to github and also pushing them as changes to gerrit, the commits won't be the same since you'll need to merge the job config fix change before those other changes can merge, so once mirroring gets set up you'll wind up making a non-fat-forwardable overwrite to what's in github | 01:37 |
| fungi | er, non-fast-forwardable | 01:37 |
| fungi | the job rewrites can't be prioritized? | 01:37 |
| EmilienM | fungi: it'll probably be me doing it and I can't tell how much time it'll take | 01:41 |
| EmilienM | i hope not long, but let's say a week (in case) | 01:41 |
| EmilienM | well it's a week where we can't merge PRs :/ | 01:41 |
| EmilienM | we don't have that many PRs, maybe 3 a week? | 01:41 |
| fungi | we really prefer not to import a repo "twice" since the second import involves someone manually pushing a complete overwrite, but in theory we could do it that way (create the project automatically importing the current state from github, work on a proposed change to fix the job configs, once it's passing don't approve it but overwrite the project in opendev with the latest state in | 01:45 |
| fungi | github, freeze updates in github, rebase the job fixes and approve them, then set up mirroring back to github) | 01:45 |
| fungi | oh, there might be a more lightweight way to do that | 01:52 |
| fungi | propose a change to opendev/sandbox which replaces all the master branch content with the gophercloud code and job configs, then iterate on that change until the jobs pass | 01:53 |
| fungi | clarkb: ^ do you think that might work? | 01:53 |
| opendevreview | Takashi Kajinami proposed openstack/project-config master: Retire puppet-senlin - Step 1: End project Gating https://review.opendev.org/c/openstack/project-config/+/817324 | 02:14 |
| EmilienM | fungi: oh, good idea for sandbox... | 02:16 |
| EmilienM | however it doesn't have any zuul job | 02:17 |
| EmilienM | but I guess I could add one | 02:18 |
| EmilienM | I'll give it a try. Thanks fungi , good night | 02:18 |
| fungi | yeah, the zuul job configuration would just be in your proposed change, zuul will execute it speculatively | 02:18 |
| fungi | have a good night! | 02:19 |
| opendevreview | Takashi Kajinami proposed openstack/project-config master: Retire puppet-senlin - Step 3: Remove Project https://review.opendev.org/c/openstack/project-config/+/817327 | 02:20 |
| clarkb | sorry just catching up was having dinner | 02:24 |
| clarkb | when fungi says we really prefer to not import twice I think we should taright up say that isn't part of the bargain :) | 02:26 |
| clarkb | its puts a lot of work on us as admins and there really isn't any reason to ever do that | 02:26 |
| clarkb | and yes I think we've suggested the sandbox thing to others before | 02:27 |
| EmilienM | ++ | 02:27 |
| clarkb | It won't be efficient with on disk caches but should be sufficient for the exercise | 02:27 |
| EmilienM | trying something here https://review.opendev.org/c/opendev/sandbox/+/817332 | 02:34 |
| EmilienM | but it seems like the job I'm creating is being dequeued | 02:34 |
| EmilienM | nevermind it seems to run now | 02:34 |
| clarkb | that may just be related to the multi zuul scheduler setup we're running now. One thing that can happen is it may flip flop between each of their views on the status page but they should eventually agree with each other | 02:36 |
| EmilienM | ack | 02:37 |
| opendevreview | Ian Wienand proposed openstack/project-config master: Set debian-stretch to min-ready: 0 https://review.opendev.org/c/openstack/project-config/+/817338 | 04:11 |
| opendevreview | Ian Wienand proposed openstack/project-config master: Remove debian-stretch nodes and builds https://review.opendev.org/c/openstack/project-config/+/817339 | 04:11 |
| *** ysandeep|out is now known as ysandeep | 05:33 | |
| frickler | wow, opening that sandbox patch in firefox makes my computer want to go up in flames | 06:36 |
| *** amoralej|off is now known as amoralej | 07:57 | |
| *** ysandeep is now known as ysandeep|lunch | 08:41 | |
| *** akahat|rover is now known as akahat|lunch | 08:44 | |
| *** ykarel is now known as ykarel|lunch | 08:51 | |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Initial project commit https://review.opendev.org/c/openstack/ci-log-processing/+/815604 | 09:06 |
| *** akahat|lunch is now known as akahat|rover | 09:13 | |
| *** ysandeep|lunch is now known as ysandeep | 09:23 | |
| *** ykarel|lunch is now known as ykarel | 09:57 | |
| *** melwitt is now known as Guest5508 | 10:12 | |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: DNM Add ansible role that is configuring logscraper https://review.opendev.org/c/openstack/ci-log-processing/+/817050 | 10:14 |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Add ansible role that is configuring logscraper https://review.opendev.org/c/openstack/ci-log-processing/+/817050 | 10:18 |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Initial project commit https://review.opendev.org/c/openstack/ci-log-processing/+/815604 | 10:23 |
| frickler | EmilienM: if you need help with transforming or debugging jobs let me know | 10:55 |
| *** rlandy is now known as rlandy|ruck | 11:10 | |
| EmilienM | frickler: thx | 11:12 |
| *** dviroel|out is now known as dviroel | 11:16 | |
| *** jcapitao is now known as jcapitao_lunch | 11:36 | |
| rlandy|ruck | fungi: clarkb: hello ... it looks like the tripleo gate is stuck | 12:09 |
| rlandy|ruck | and others | 12:09 |
| rlandy|ruck | can you unstick those for us? | 12:10 |
| rlandy|ruck | tripleo gate is waiting for 15 hours | 12:10 |
| *** soniya29|ruck is now known as soniya29|ruck|afk | 12:23 | |
| frickler | rlandy|ruck: see backlog, waiting for corvus to have a deeper look | 12:28 |
| rlandy|ruck | frickler: thanks | 12:28 |
| frickler | rlandy|ruck: oh, actually the other discussion was in #opendev | 12:28 |
| rlandy|ruck | joining there | 12:29 |
| *** ysandeep is now known as ysandeep|afk | 12:32 | |
| dpawlik | clarkb, fungi: Hey, I can move log-gearman-client,worker to ci-log-processing repo, right? | 12:38 |
| *** jcapitao_lunch is now known as jcapitao | 12:55 | |
| *** ysandeep|afk is now known as ysandeep | 13:08 | |
| *** jpena|off is now known as jpena | 13:21 | |
| *** amoralej is now known as amoralej|lunch | 13:22 | |
| *** soniya29|ruck|afk is now known as soniya29|ruck | 13:33 | |
| *** rlandy|ruck is now known as rlandy|ruck|mtg | 14:01 | |
| *** amoralej|lunch is now known as amoralej | 14:06 | |
| fungi | dpawlik: yeah, if you want to reuse it, forking a copy of that script into the new repo seems like a reasonable way to go about it, then we'll be able to retire the old puppet-log_processor repo | 14:20 |
| fungi | retire it eventually once we turn off our servers, i mean | 14:20 |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Add Log Processor module https://review.opendev.org/c/openstack/ci-log-processing/+/817436 | 14:31 |
| dpawlik | fungi: allright. Proposed a change ^^ | 14:32 |
| dpawlik | fungi: is there any problem with nodeset definition? I got error "Nodeset ubuntu-bionic-2-node already defined" | 14:33 |
| fungi | dpawlik: it's related to the current multi-scheduler work in zuul, we hope to restart with https://review.opendev.org/817328 in place shortly to correct it | 14:34 |
| dpawlik | ack | 14:34 |
| zul | Hey this one should have 2+/1+ https://review.opendev.org/c/starlingx/upstream/+/815691 but it hasnt merged yet | 14:40 |
| fungi | zul: we've got a problem with stuck queues at the moment, likely related to the multi-scheduler work, so things aren't getting enqueued but investigation is underway and hopefully we'll be able to clear the situation shortly (whether that will entail clearing the current state though, i'm unsure yet) | 14:43 |
| zul | cool thanks for letting me know | 14:43 |
| fungi | zul: there's a restart underway now, but any of your changes which aren't automatically reenqueued after that will likely need to be rechecked to get testing going for them | 14:56 |
| zul | ack thanks | 14:58 |
| *** ykarel is now known as ykarel|away | 15:12 | |
| *** soniya29|ruck is now known as soniya29|ruck|dinner | 15:38 | |
| *** akahat|rover is now known as akahat|lunch | 16:03 | |
| *** akahat|lunch is now known as akahat|dinner | 16:03 | |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Initial project commit https://review.opendev.org/c/openstack/ci-log-processing/+/815604 | 16:05 |
| *** soniya29|ruck|dinner is now known as soniya29|ruck | 16:21 | |
| opendevreview | daniel.pawlik proposed openstack/ci-log-processing master: Initial project commit https://review.opendev.org/c/openstack/ci-log-processing/+/815604 | 16:46 |
| *** ysandeep is now known as ysandeep|out | 16:51 | |
| *** akahat|dinner is now known as akahat|rover | 16:52 | |
| *** soniya29|ruck is now known as soniya29|ruck|out | 17:01 | |
| *** rlandy|ruck|mtg is now known as rlandy|ruck | 17:09 | |
| *** jpena is now known as jpena|off | 17:34 | |
| *** rlandy is now known as rlandy|ruck | 17:40 | |
| *** amoralej is now known as amoralej|off | 18:23 | |
| *** dviroel is now known as dviroel|out | 21:15 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!