| *** ianychoi has joined #zuul | 00:46 | |
| *** hamalq has quit IRC | 01:00 | |
| *** fsvsbs has quit IRC | 01:11 | |
| *** rlandy|rover|bbl is now known as rlandy|rover | 01:53 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-operator master: Flake8 cleanups https://review.opendev.org/c/zuul/zuul-operator/+/786349 | 01:54 |
|---|---|---|
| *** aluria has quit IRC | 01:56 | |
| *** aluria has joined #zuul | 02:00 | |
| *** rlandy|rover has quit IRC | 02:26 | |
| *** evrardjp has quit IRC | 02:33 | |
| *** evrardjp has joined #zuul | 02:33 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Keep jobgraphs frozen across reconfiguration https://review.opendev.org/c/zuul/zuul/+/785536 | 02:39 |
| *** bhavikdbavishi has joined #zuul | 03:10 | |
| *** bhavikdbavishi1 has joined #zuul | 03:43 | |
| *** bhavikdbavishi has quit IRC | 03:44 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:44 | |
| *** ykarel has joined #zuul | 03:45 | |
| *** bhavikdbavishi has quit IRC | 04:21 | |
| *** vishalmanchanda has joined #zuul | 04:23 | |
| *** bhavikdbavishi has joined #zuul | 04:52 | |
| *** bhavikdbavishi has quit IRC | 04:56 | |
| *** bhavikdbavishi has joined #zuul | 05:07 | |
| *** sshnaidm|pto has quit IRC | 05:47 | |
| *** sshnaidm|pto has joined #zuul | 05:48 | |
| *** ykarel has quit IRC | 06:00 | |
| *** ykarel has joined #zuul | 06:00 | |
| *** saneax has joined #zuul | 06:31 | |
| *** icey has quit IRC | 06:34 | |
| *** icey has joined #zuul | 06:35 | |
| *** bhavikdbavishi has quit IRC | 06:35 | |
| *** saneax has quit IRC | 06:36 | |
| *** jcapitao has joined #zuul | 06:37 | |
| *** jcapitao has quit IRC | 06:37 | |
| *** jcapitao has joined #zuul | 06:39 | |
| *** bhavikdbavishi has joined #zuul | 06:40 | |
| *** ykarel_ has joined #zuul | 06:49 | |
| *** reiterative has quit IRC | 06:49 | |
| *** reiterative has joined #zuul | 06:50 | |
| *** ykarel has quit IRC | 06:51 | |
| *** jpena|off is now known as jpena | 06:55 | |
| *** bhavikdbavishi has quit IRC | 06:57 | |
| *** bhavikdbavishi has joined #zuul | 07:11 | |
| *** bhavikdbavishi has quit IRC | 07:25 | |
| *** bhavikdbavishi has joined #zuul | 07:26 | |
| *** ykarel_ is now known as ykarel | 07:30 | |
| *** bhavikdbavishi1 has joined #zuul | 07:44 | |
| *** rpittau|afk is now known as rpittau | 07:45 | |
| *** bhavikdbavishi has quit IRC | 07:46 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 07:46 | |
| *** tosky has joined #zuul | 07:46 | |
| *** nils has joined #zuul | 08:04 | |
| *** ykarel is now known as ykarel|lunch | 08:33 | |
| *** vishalmanchanda has quit IRC | 09:27 | |
| *** saneax has joined #zuul | 09:29 | |
| *** ykarel|lunch is now known as ykarel | 09:38 | |
| *** dpawlik9 has quit IRC | 09:46 | |
| *** vishalmanchanda has joined #zuul | 09:48 | |
| *** jcapitao has quit IRC | 09:53 | |
| *** dpawlik4 has joined #zuul | 09:59 | |
| *** jcapitao has joined #zuul | 10:01 | |
| *** bhavikdbavishi has quit IRC | 10:07 | |
| *** jcapitao is now known as jcapitao_lunch | 10:28 | |
| *** ykarel_ has joined #zuul | 10:29 | |
| *** ykarel has quit IRC | 10:32 | |
| *** ykarel_ has quit IRC | 10:44 | |
| *** ykarel_ has joined #zuul | 10:45 | |
| *** ykarel_ is now known as ykarel | 11:15 | |
| *** jpena is now known as jpena|lunch | 11:31 | |
| *** rlandy has joined #zuul | 11:46 | |
| *** rlandy is now known as rlandy|rover | 11:46 | |
| *** ykarel_ has joined #zuul | 11:54 | |
| *** ykarel has quit IRC | 11:56 | |
| *** rlandy|rover is now known as rlandy|rover|mtg | 12:01 | |
| *** ykarel_ is now known as ykarel | 12:02 | |
| *** jcapitao_lunch is now known as jcapitao | 12:03 | |
| *** bhavikdbavishi has joined #zuul | 12:29 | |
| *** bhavikdbavishi1 has joined #zuul | 12:32 | |
| *** jpena|lunch is now known as jpena | 12:33 | |
| *** bhavikdbavishi has quit IRC | 12:33 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 12:33 | |
| *** rlandy|rover|mtg is now known as rlandy|rover | 12:42 | |
| *** bhavikdbavishi has quit IRC | 12:43 | |
| lyr | Hi there | 12:49 |
| lyr | Gitta migrate secret between two zuul, how can I do that ? | 12:49 |
| corvus | lyr: re-encrypt it for the other zuul, or if you are migrating the whole zuul system and you manage both, then copy the secret keys from the scheduler | 13:30 |
| lyr | I was willing to decrypt on the old & encrypt on the new (to avoid looking everywhere sfor the original secrets), but I can't get this decrypt script https://review.opendev.org/c/zuul/zuul/+/639771 to work | 13:58 |
| openstackgerrit | Andy Ladjadj proposed zuul/zuul master: [reporter][elasticsearch] ensure the @timestamp as export with UTC format in case of system has a different timezone https://review.opendev.org/c/zuul/zuul/+/786444 | 14:04 |
| openstackgerrit | Andy Ladjadj proposed zuul/zuul master: [reporter][elasticsearch] ensure the @timestamp as export with UTC format in case of system has a different timezone https://review.opendev.org/c/zuul/zuul/+/786444 | 14:08 |
| openstackgerrit | Felix Edel proposed zuul/zuul master: Switch to ZooKeeper backed build result events https://review.opendev.org/c/zuul/zuul/+/782939 | 14:25 |
| *** avass has quit IRC | 14:42 | |
| *** avass has joined #zuul | 14:43 | |
| avass | when onboarding an organization I noticed that they've structured their repos like <product>, <product>/<component-1> and <product>/<component-2> and I have a feeling that's going to cause some problems in zuul | 14:45 |
| avass | best case zuul is probably going to delete merger/executor cache a lot, worst case we're gonna be seeing a lot of MERGE_FAILURE or executor errors when setup up the workspace :/ | 14:46 |
| fungi | are the projects in gerrit? github? something else? | 14:47 |
| fungi | just curious how the hosting backend handles that | 14:47 |
| fungi | seems like it would similarly conflict in gerrit | 14:48 |
| avass | it's gerrit | 14:48 |
| fungi | any idea how that looks on the gerrit server's filesystem? | 14:48 |
| avass | no clue since we're not hosting any gerrit instance ourselves | 14:48 |
| avass | I would expect other tools like golang to have problems with that as well | 14:48 |
| fungi | yeah | 14:49 |
| corvus | avass: zuul uses golang-style repo layouts, so that will definitely cause issues, though it's not clear to me right now how significant those issues will be | 14:49 |
| fungi | ahh, nevermind, gerrit's using bare repos (at least in our deployment), so even though git repositories are in a common file tree, there would be a review_site/git/<product>.git and a review_site/git/<product> directory and they wouldn't conflict | 14:50 |
| avass | me neither, but a nested git repos shows up as a dirty state so that's why I'm guessing that a "parent" repo could cause problems by doing a git clean | 14:50 |
| corvus | avass: is moving product -> product/main or something an option? that would almost certainly be best | 14:50 |
| avass | corvus: I'm checking that with them right now but I have a feeling that they could have a lot of tools dependent on that structure | 14:51 |
| avass | fungi: ah right of course | 14:51 |
| corvus | there's two things to consider from zuul's pov: the mergers and what zuul checks out for a job in the workspace. it's probably not too difficult to make the mergers safe for this. but i have a really hard time thinking that we would be able to actually articulate what a workspace checkout should look like in that case | 14:53 |
| corvus | like, i'm not sure we would be okay saying that zuul should dirty one git repo checkout with another | 14:53 |
| mordred | yeah | 14:54 |
| corvus | (and it's possible that it could be extremely difficult/impossible to actually get zuul to do that even if we wanted to) | 14:54 |
| mordred | maybe a (mostly very hidden) option to have workspace checkouts append a .git suffix? | 14:55 |
| fungi | also gerrit could probably similarly break if you had projects named foo and foo.git/bar, so i don't think working around it by suffixing the repository names is necessarily a solution | 14:55 |
| mordred | (I can't even begin to imagine all the things that would break though) | 14:55 |
| mordred | fungi: good point | 14:55 |
| fungi | yeah, probably the only real way to address it would be to make the directory tree something like a sha256 hash of each repository name rather than the actual name | 14:56 |
| corvus | off the top of my head, if we had to make this work, i think about the only thing we could do is sort of completely punt and do something like ^ | 14:56 |
| fungi | and at least then you're down to sha2-256 hash collisions | 14:56 |
| mordred | what about an explicit aliasing? | 14:56 |
| corvus | basically, rename the repo | 14:56 |
| corvus | (in the zuul checkout) | 14:56 |
| mordred | so that someone could configure zuul with a real-name to checkout-name mapping for a project | 14:57 |
| fungi | the git prep role would also have to keep that structure on job nodes/workspaces | 14:57 |
| fungi | which means jobs get more complicated since you need a lookup method to sha256sum the name | 14:58 |
| mordred | then for that org, they could add a config saying project: project/main (for instance) | 14:58 |
| mordred | and there would only be the one impacted repo | 14:58 |
| corvus | mordred: yep. then if they want to dirty up the repo, add a pre-playbook to move it in place. | 14:58 |
| mordred | yup | 14:58 |
| mordred | and if they're fine with that in their conent - no worries | 14:59 |
| mordred | I think it would want to be a mapping configured in zuul.yaml so that basically everywhere zuul is talking about that project it's talkign about it differently than gerrit is | 14:59 |
| corvus | ++ | 14:59 |
| avass | fungi: that could be solved by setting zuul.project.src_dir accordingly no? | 15:00 |
| fungi | we do have similar "translate this branch name to this other branch name" options, so i suppose that would be somewhat consistent with how zuul does branch aliasing | 15:00 |
| mordred | ++ | 15:00 |
| fungi | avass: it could, but would require that jobs leverage that variable (maybe not the worst thing, but some folks do bake repository names into job definitions) | 15:01 |
| fungi | the checksum solution would be non-backward-compatible with existing job definitions | 15:02 |
| clarkb | me thinking out loud here: from a user experience perspective I think putting the burden on people who want to do weird esoteric things and otherwise making life easy is important. If we did aliasing or mapping of repos it would be good to not apply it globally imo | 15:03 |
| corvus | clarkb: yes, i agree; that is the assumption i'm working from | 15:04 |
| fungi | i suppose if the checksum method were desirable, it could have a compatibility method in the git prep role to put the directories in the "legacy" readable directory names | 15:04 |
| fungi | clarkb: yep, that's how the branch aliasing works now too | 15:04 |
| fungi | so i agree, there's precedent for mordred's suggestion | 15:05 |
| corvus | i don't think we should do anything with checksums in the workspace checkout. it's a useful way to think about the problem (renaming the repo). it might be part of how we solve the merger side of the problem. | 15:05 |
| fungi | also we'd need to change how project keys are stored right? (or does zk solve that?) | 15:06 |
| corvus | fungi: i think keys are okay (both before and after zk), but worth double checking | 15:06 |
| corvus | just because they're like "/connection/project/0.key" and "/connection/project/component/0.key" | 15:07 |
| fungi | oh, yeah you'd end up with foo/keys... and foo/bar/keys... coexisting in a nested tree, but not a problem unless we had projects named like the key files themselves | 15:07 |
| avass | looks like renaming their repos is not an option :/ | 15:07 |
| corvus | yep. basically exactly like gerrit | 15:07 |
| openstackgerrit | James E. Blair proposed zuul/zuul-operator master: Flake8 cleanups https://review.opendev.org/c/zuul/zuul-operator/+/786349 | 15:29 |
| *** saneax has quit IRC | 15:34 | |
| *** xarragon has joined #zuul | 15:46 | |
| *** ykarel has quit IRC | 15:47 | |
| *** bhavikdbavishi has joined #zuul | 15:48 | |
| *** bhavikdbavishi has quit IRC | 15:53 | |
| *** hamalq has joined #zuul | 16:00 | |
| *** hamalq has quit IRC | 16:02 | |
| *** hamalq has joined #zuul | 16:02 | |
| *** jpena is now known as jpena|off | 16:05 | |
| *** jcapitao has quit IRC | 16:06 | |
| *** rpittau is now known as rpittau|afk | 16:09 | |
| clarkb | corvus: looks like testing failed on the end of the consistent buildset revs stack. I think you mentioned the last two needed to be squashed but I would expect the last one to pass testing in that case | 16:17 |
| clarkb | oh I see tobiash mentions linter issues /me should look closer :) | 16:17 |
| corvus | clarkb: yeah, it's just linting; do you want to go ahead and +/-1 that and then i'll fix and squash? | 16:27 |
| clarkb | corvus: ya I'm running some gerrit account cleanups then will dive into proper review after | 16:29 |
| corvus | cool, no rush | 16:29 |
| *** nils has quit IRC | 17:10 | |
| corvus | clarkb, ianw: the gerrit zuul plugin can be enabled/disabled in the gui: https://imgur.com/a/ElHmyy5 | 17:29 |
| corvus | (2nd picture is just to show the tooltip) | 17:30 |
| clarkb | corvus: ya that is a frontend to the project acl file I think (the contributor agreement stuff above goes in the file too) | 17:31 |
| corvus | yep; not *everything* can be changed through the gui, so i thought it notable that this can be | 17:58 |
| openstackgerrit | Albin Vass proposed zuul/nodepool master: WIP: Digitalocean driver https://review.opendev.org/c/zuul/nodepool/+/759559 | 18:03 |
| clarkb | corvus: ok I left a couple of comments on that stack | 18:05 |
| clarkb | corvus: oh ya | 18:05 |
| *** vishalmanchanda has quit IRC | 18:27 | |
| *** holser has quit IRC | 18:47 | |
| corvus | clarkb: thx, replied. | 19:07 |
| corvus | clarkb: regarding your comment about the test; do you have a test for that? | 19:07 |
| clarkb | corvus: https://review.opendev.org/c/zuul/zuul/+/784142/1/tests/unit/test_merger_repo.py is what I put together to test that but I'm not sure how valid it is with the alternative appraoch we have agreed on | 19:10 |
| *** hamalq has quit IRC | 19:17 | |
| *** rlandy|rover is now known as rlandy|rover|mtg | 19:28 | |
| *** holser has joined #zuul | 19:28 | |
| *** hamalq has joined #zuul | 19:31 | |
| clarkb | corvus: your responses made sense to me thanks | 19:40 |
| corvus | clarkb: thanks, i'll look into adding that test | 19:44 |
| *** rlandy|rover|mtg is now known as rlandy|rover | 19:56 | |
| *** rlandy|rover is now known as rlandy|rover|bbl | 22:33 | |
| *** tosky has quit IRC | 22:34 | |
| openstackgerrit | James E. Blair proposed zuul/nodepool master: Log decline reason at info https://review.opendev.org/c/zuul/nodepool/+/786513 | 23:25 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!