| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: | 02:25 | |
| - [zuul/zuul] 860033: Add access-rules configuration and documentation https://review.opendev.org/c/zuul/zuul/+/860033 | ||
| - [zuul/zuul] 860034: Set Access-Control-Allow-Origin headers in check_auth tool https://review.opendev.org/c/zuul/zuul/+/860034 | ||
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: | 02:25 | |
| - [zuul/zuul] 860035: Support authz for read-only web access https://review.opendev.org/c/zuul/zuul/+/860035 | ||
| - [zuul/zuul] 860987: Change merge mode default based on driver https://review.opendev.org/c/zuul/zuul/+/860987 | ||
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] 861668: Add rebase-merge merge mode https://review.opendev.org/c/zuul/zuul/+/861668 | 02:25 | |
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/nodepool] 811016: Expose slot number from metastatic driver https://review.opendev.org/c/zuul/nodepool/+/811016 | 03:29 | |
| -@gerrit:opendev.org- Simon Westphahl proposed: [zuul/zuul] 862782: wip: Reuse existing pipeline state during reconfigs https://review.opendev.org/c/zuul/zuul/+/862782 | 07:52 | |
| -@gerrit:opendev.org- Simon Westphahl proposed: [zuul/zuul] 862794: Don't refresh the pipeline state during reconfig https://review.opendev.org/c/zuul/zuul/+/862794 | 10:27 | |
| -@gerrit:opendev.org- Simon Westphahl proposed: [zuul/zuul] 862794: Don't refresh the pipeline state during reconfig https://review.opendev.org/c/zuul/zuul/+/862794 | 12:36 | |
| @clarkb:matrix.org | One upside to dropping zuul support for old ansibles is the size of our executor images has dropped quite a bit | 15:51 |
|---|---|---|
| @jim:acmegating.com | yeah batteries included installs of ansible are not small | 15:52 |
| @jim:acmegating.com | i expect ansible 7 to be out very soon | 15:53 |
| -@gerrit:opendev.org- Thomas Cardonne proposed on behalf of Guillaume Chauvel: [zuul/zuul] 735476: zuul_stream: close streamer when task is skipped https://review.opendev.org/c/zuul/zuul/+/735476 | 16:04 | |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 862829: Add a zkprofile scheduler debug command https://review.opendev.org/c/zuul/zuul/+/862829 | 16:26 | |
| @clarkb:matrix.org | ianw: I left a review on that zuul-sphinx change. Slightly worried we may allow actual circular imports now, but I think CI would catch that through the more explosive fireworks that would result in? | 17:29 |
| @fungicide:matrix.org | > <@clarkb:matrix.org> ianw: I left a review on that zuul-sphinx change. Slightly worried we may allow actual circular imports now, but I think CI would catch that through the more explosive fireworks that would result in? | 18:45 |
| circular imports are only problematic if one of the modules involved calls into another within the import cycle though, right? | ||
| @fungicide:matrix.org | e.g. modules with import side-effects | 18:45 |
| @clarkb:matrix.org | fungi: these are rst docs though so I'm not complete sure | 18:47 |
| @clarkb:matrix.org | I suspect that it may copy content in a circular fashion until you run out memory or something | 18:47 |
| @fungicide:matrix.org | oh, sorry! i thought you meant circular python module imports | 18:58 |
| @jim:acmegating.com | https://review.opendev.org/862215 is the change in question | 19:10 |
| @iwienand:matrix.org | > <@fungicide:matrix.org> circular imports are only problematic if one of the modules involved calls into another within the import cycle though, right? | 19:59 |
| i guess if it really were a circular import it would hang. the bug just got a comment https://sourceforge.net/p/docutils/bugs/459/ that basically says "your directive should update the include log". i guess the problem with that is that really from their (docutils) side they've introduced a breaking change -- but OTOH i don't think there's any sort of documentation or api clearly expressed either | ||
| @clarkb:matrix.org | update it to clear entries like we are doing or update it to add more info than we are/were adding? | 20:01 |
| @iwienand:matrix.org | Clark: not sure :) the thing is we use ```insert_input``` which just so happens to include input that uses ```.. include::```. if we have to do extra things around that, it feels a bit like a api breaking change https://sourceforge.net/p/docutils/bugs/459/#61e7/9cf8 | 20:37 |
| -@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed on behalf of Simon Westphahl: [zuul/zuul] 862794: Don't refresh the pipeline state during reconfig https://review.opendev.org/c/zuul/zuul/+/862794 | 22:24 | |
| -@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] 862193: Fix implied branch matchers and override-checkout https://review.opendev.org/c/zuul/zuul/+/862193 | 23:39 | |
| @michael_kelly_anet:matrix.org | corvus: Questions regarding your comments on https://review.opendev.org/c/zuul/zuul-operator/+/861279/3 - 'running from the CLI': are folks running the operator from outside of a Kubernetes cluster or something? | 23:40 |
| Based on my experimentation, running in cluster it seems likely that the config is *already* loaded when we get to this point as the exec is able to run without issue. | ||
| @michael_kelly_anet:matrix.org | I can put some cycles into a test around this next week, but it might be good to merge a fix before the test as the operator is otherwise straight up broken. | 23:41 |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/zuul] 862622: Add Python 3.11 testing https://review.opendev.org/c/zuul/zuul/+/862622 | 23:50 | |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/nodepool] 862630: Add Python 3.11 testing https://review.opendev.org/c/zuul/nodepool/+/862630 | 23:54 | |
| @clarkb:matrix.org | ianw: ^ we may need to consider the sibling job there. But I think pushing forward on that if there are incompatibilities is a good thing and I'm happy to help. | 23:54 |
| @jim:acmegating.com | Michael Kelly: yes, the operator can be run outside the cluster; that is the best way to develop on it since there's no image build cycle necessary. i'd really like to have the test with the fix to verify it. ideally we would have included that in the original implementation, but now that we've discovered a hole in testing, we should add it. | 23:55 |
| @iwienand:matrix.org | Clark: i guess the problem might be if it needs to build too much and runs too long? | 23:55 |
| @michael_kelly_anet:matrix.org | Fascinating. | 23:55 |
| @michael_kelly_anet:matrix.org | I had no expectation of running this outside of an actual cluster :) | 23:55 |
| @michael_kelly_anet:matrix.org | My concern with the test is that I have to spend cycles to get a feel for how the existing testing works, that's all. | 23:56 |
| @michael_kelly_anet:matrix.org | Sounds like maybe the soln looks like wrapping the original call? Or maybe being aware of where we're running | 23:56 |
| @jim:acmegating.com | Michael Kelly: i suspect that may be the idea behind the other change i linked | 23:57 |
| @michael_kelly_anet:matrix.org | Yea, it makes more sense now, of course. | 23:57 |
| @clarkb:matrix.org | > <@iwienand:matrix.org> Clark: i guess the problem might be if it needs to build too much and runs too long? | 23:58 |
| ya or if dib or glean don't do python3.11 yet | ||
| @clarkb:matrix.org | we'll see what testing shows as far as build times as that is a concern generally on arm64 | 23:58 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!