| *** tjgresha has quit IRC | 00:09 | |
| *** rlandy|rover has quit IRC | 00:44 | |
| *** rfolco has quit IRC | 00:52 | |
| *** sgw has quit IRC | 01:01 | |
| *** armstrongs has joined #zuul | 01:04 | |
| *** armstrongs has quit IRC | 01:13 | |
| *** sgw has joined #zuul | 01:18 | |
| *** adso78 has quit IRC | 01:27 | |
| *** jamesmcarthur has joined #zuul | 02:06 | |
| *** michael-beaver has quit IRC | 02:12 | |
| *** igordc has quit IRC | 02:18 | |
| *** bhavikdbavishi has joined #zuul | 02:35 | |
| *** bhavikdbavishi1 has joined #zuul | 03:18 | |
| *** bhavikdbavishi has quit IRC | 03:20 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:20 | |
| *** bjackman has joined #zuul | 03:34 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: use-buildset-registry: Vendor pytoml and remarshal https://review.opendev.org/695050 | 04:46 |
|---|---|---|
| *** threestrands has joined #zuul | 05:18 | |
| *** pcaruana has joined #zuul | 05:25 | |
| openstackgerrit | Ian Wienand proposed zuul/nodepool master: Dockerfile: add DEBUG environment flag https://review.opendev.org/694845 | 05:32 |
| openstackgerrit | Ian Wienand proposed zuul/nodepool master: [wip] move openstack testing to use containerised daemon https://review.opendev.org/693464 | 05:32 |
| *** themroc has joined #zuul | 07:23 | |
| *** bjackman has quit IRC | 07:32 | |
| *** tosky has joined #zuul | 07:35 | |
| *** threestrands has quit IRC | 07:58 | |
| *** avass has joined #zuul | 08:23 | |
| *** jpena|off is now known as jpena | 08:49 | |
| avass | I'm not sure if I'm blind because I'm tired or something else is going on here but this: http://paste.openstack.org/show/786386/ is producing a strange error for the weekly pipeline. | 08:57 |
| avass | With this error: http://paste.openstack.org/show/786388/ | 08:57 |
| avass | which is strange because I can't find an reference to a 'weekly-pass-release-configuration-parse' in any of our repos | 08:58 |
| avass | the check/gate jobs are working perfectly fine | 08:59 |
| avass | any suggestions on how to proceed? | 09:04 |
| *** rfolco has joined #zuul | 09:05 | |
| *** jamesmcarthur has quit IRC | 09:07 | |
| *** rfolco has quit IRC | 09:09 | |
| *** saneax has joined #zuul | 09:12 | |
| *** tosky has quit IRC | 09:13 | |
| *** tosky has joined #zuul | 09:14 | |
| *** jamesmcarthur has joined #zuul | 09:15 | |
| *** themr0c has joined #zuul | 09:56 | |
| *** themroc has quit IRC | 09:58 | |
| *** NBorg has joined #zuul | 10:18 | |
| *** mhu has joined #zuul | 10:27 | |
| *** bhavikdbavishi has quit IRC | 10:28 | |
| NBorg | Hello. We want to run ansible-lint through tox. Is there a package we can require for the linter to not throw an error when we use zuul_return in our playbooks/roles? | 10:31 |
| *** mhu has quit IRC | 10:34 | |
| *** mhu has joined #zuul | 10:43 | |
| *** bhavikdbavishi has joined #zuul | 10:50 | |
| *** swest has joined #zuul | 10:54 | |
| *** bhavikdbavishi1 has joined #zuul | 10:57 | |
| *** bhavikdbavishi has quit IRC | 10:58 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 10:58 | |
| *** swest has quit IRC | 11:03 | |
| *** swest has joined #zuul | 11:04 | |
| *** PrinzElvis has quit IRC | 11:08 | |
| NBorg | Never mind. It was just zuul. Ee had an old path to ansible action plugins. | 11:42 |
| NBorg | We* | 11:42 |
| *** rlandy has joined #zuul | 11:55 | |
| *** rlandy is now known as rlandy|rover | 11:55 | |
| *** NBorg has quit IRC | 12:03 | |
| *** PrinzElvis has joined #zuul | 12:15 | |
| *** pcaruana has quit IRC | 12:24 | |
| *** jpena is now known as jpena|lunch | 12:29 | |
| *** swest has quit IRC | 12:31 | |
| *** pcaruana has joined #zuul | 12:50 | |
| fungi | avass: unless there's a dependency declared from weekly-test-pass-release-configuration-verify or one of its parent jobs on a weekly-pass-release-configuration-parse job on some branch of some repository (whether or not that's defined) then the only other guess would be it used to exist and a zuul bug causing it to cling to outdated configuration for some reason | 12:56 |
| fungi | have you tried exploring the relevant jobs with the jobs interface in the zuul web dashboard? | 12:56 |
| *** avass has quit IRC | 13:19 | |
| *** jpena|lunch is now known as jpena | 13:36 | |
| *** AJaeger has quit IRC | 13:58 | |
| *** AJaeger has joined #zuul | 14:06 | |
| *** saneax has quit IRC | 14:11 | |
| *** PrinzElvis has quit IRC | 14:38 | |
| *** jamesmcarthur has quit IRC | 14:49 | |
| *** avass has joined #zuul | 14:52 | |
| avass | fungi: I'm getting the 'Something went wrong.' screen for taht job in the dashboard | 14:54 |
| avass | even after doing a full-reconfigure | 14:55 |
| corvus | avass: were there any lines before http://paste.openstack.org/show/786388/ in the log? it may have indicated whether it was going to run that job or not. also, is that logging at debug level? | 14:56 |
| corvus | avass: another thing to check is that there may be a files matcher or something on test-pass-release-configuration-parse which is the parent of the job that's not being run. it could also cause it not to run. | 14:57 |
| *** jamesmcarthur has joined #zuul | 14:57 | |
| avass | corvus: I'll check | 14:58 |
| avass | corvus: but I saw a discussion about jobs in timed pipelines not triggering if they have any files specified a couple of days ago so that's why they are separated | 14:59 |
| corvus | avass: ah yeah, then presumably that one shouldn't. the other thing to check would be branches then. | 15:00 |
| avass | ah yeah, it's also because of the branches. we're still running zuul2 on master but working on zuul3 on that feature branch | 15:01 |
| corvus | avass: that'll do it. every job in the inheritance hierarchy will need a variant on the feature/zuul3 branch in order for the job to run. so probably just adding a 1-line job def for "name: test-pass-release-configuration-parse" with "branches: feature/zuul3" would do it. | 15:04 |
| corvus | avass: er, strike that | 15:04 |
| avass | :) | 15:05 |
| corvus | avass: just add "branches: [master, feature/zuul3]" to the existing def of test-pass-release-configuration-parse" | 15:05 |
| avass | ah, but we want to run the weekly on master but the check/gate jobs on the branch. that's why they look like that | 15:05 |
| avass | looks like I got a different error from the scheduler after running full-reconfigure actually so that could be it. | 15:07 |
| corvus | avass: is that config snippet you sent from .zuul.yaml in the feature/zuul3 branch of an untrusted project? | 15:10 |
| avass | it's from master in an untrusted project | 15:12 |
| corvus | avass: it also has a feature/zuul3 branch? | 15:13 |
| avass | yep | 15:13 |
| avass | fixed the new error and the old one is back | 15:13 |
| corvus | avass: here's a suggestion: http://paste.openstack.org/show/786423/ | 15:14 |
| corvus | er, 1 sec. | 15:15 |
| corvus | avass: http://paste.openstack.org/show/786424/ | 15:15 |
| corvus | this is a little tricky since you're wanting to avoid check/gate on master. that turns of implied branch matchers, so that all the job definitions apply to every branch you use them. then on the feature/zuul3 branch, we add the check/gate jobs. then in the config project we set up the weekly jobs to run on master. | 15:16 |
| corvus | avass: let me explore one more idea | 15:17 |
| corvus | (since now i know you're okay with putting zuul config in the master branch) | 15:17 |
| avass | yeah, as long as we don't have anything trigger by changes to master really | 15:18 |
| corvus | avass: this should do the same with no config-project: http://paste.openstack.org/show/786428/ | 15:20 |
| corvus | avass: you can also switch the branches like this: http://paste.openstack.org/show/786426/ | 15:21 |
| corvus | depending on whichever is more convenient | 15:21 |
| corvus | then, when you're ready to engage gating on master, copy the job defs to the other branch, then drop the "pragma". | 15:21 |
| corvus | the "project" stanzas, when they appear in an untrusted-project, always apply only to the branch they appear in. i think that's the clearest way of saying "do this on this branch". you just check the project stanza in the .zuul.yaml on the branch in question, and you'll see what's running on that branch. | 15:24 |
| corvus | (the pragma that turns off implied-branch-matchers just means that it no longer matters which branch the jobs are defined in, they will run the same way on every branch and only need to be defined once. it's useful in this case since you don't need branch variants, but once you start having zuul manage more branches, it'll probably get in the way and you should go back to the normal case of per-branch job | 15:26 |
| corvus | variants) | 15:26 |
| corvus | https://zuul-ci.org/docs/zuul/user/config.html#attr-pragma.implied-branch-matchers | 15:26 |
| avass | wait, so does job 'A' on branch feature/1 with 'branches: master' even though the job itself is not on master? | 15:28 |
| avass | because I just found that missing job and it's on the feature branch. | 15:29 |
| corvus | avass: can you clarify the question? | 15:29 |
| corvus | (that question was missing at least one important word :) | 15:30 |
| avass | job 'A' trigger.. :) | 15:30 |
| AJaeger | corvus: see also https://opendev.org/openstack/glance/src/branch/stable/train/.zuul.yaml#L291-L304 - this is a problem, isn't it? | 15:30 |
| AJaeger | So, that needs some education with the glance folks... | 15:30 |
| avass | so the job 'weekly-test-pass-release-configuration-verify' exists on feature/zuul3 with 'branches: master', 'depedencies: [weekly-pass-release-configuration-parse]' (which is wrongly configured) | 15:32 |
| corvus | AJaeger: that may actually be the right thing to do in that case -- assuming there's a copy of that on the master branch, that will all match there, but it won't match anywhere else, therefore, things only run on master. | 15:33 |
| AJaeger | corvus: it's on all branches. So, I wonder: IF they remove a job from master branch but it's still on stable/train - will it run? | 15:34 |
| corvus | avass: yes, if you set explicit branch matchers, they override anything else. so you can have a job on feature/zuul3 that matches the master branch (but you couldn't cause it to run on the master branch except by adding it to a project stanza on master, but if you did that, it would). | 15:34 |
| corvus | AJaeger: it will not, because the project stanza on train won't match master. | 15:34 |
| AJaeger | ah | 15:35 |
| AJaeger | thanks, corvus | 15:35 |
| corvus | AJaeger: (basically, just like jobs have a "branches" matcher, project stanzas do as well, but you can't change it, it's always simply the branch where the project stanza is defined) | 15:35 |
| avass | corvus: that's probably what happened then because the jobs have the same name but different dependencies, one of them being wrongly configured. | 15:35 |
| corvus | AJaeger: (except in config projects, where there are no branch matchers on project stanzas) | 15:36 |
| corvus | AJaeger: (so a project stanza in a config project matches all branches -- to do branch selection, you would then have to put branch matchers on the jobs themselves in the project-pipeline config) | 15:36 |
| AJaeger | corvus: the config-project I understand, the others is the fun part ;) | 15:37 |
| corvus | AJaeger: it sounds complex when i write it out, but it's usually pretty intuitive: probably the most intuitive part: if you want to see what jobs run on a given branch, just look at the "project" stanza in that branch. :) | 15:39 |
| avass | corvus: gotta go soon so I'll have to take a closer look at your suggestions tomorrow. thanks! | 15:39 |
| corvus | sorry about that sentence, i'm still jetlagged, but you get the idea. | 15:39 |
| corvus | avass: hope it helps! let me know if you have more questions. | 15:40 |
| avass | corvus: anyway, there's gertty, boartty and presentty. any plan on a zuultty? :) | 15:40 |
| AJaeger | corvus: ;) | 15:40 |
| corvus | avass: :) my favorite interface to zuul is gerrit itself... but i think someone actually did write a tool that renders the status page in the console | 15:41 |
| fungi | "czuul" was it? | 15:41 |
| fungi | not sure if it kept up pace with v3 development though | 15:41 |
| corvus | fungi: yeah that's it. it's probably not far out of date, the status page didn't change much | 15:41 |
| corvus | https://github.com/harlowja/gerrit_view/blob/master/scripts/czuul | 15:42 |
| fungi | you beat me to that link by seconds! ;) | 15:42 |
| avass | nice, I'll take a look at it :) | 15:44 |
| avass | I'm gonna have a quck demo of zuul3 here at Volvo Cars for a couple of people using presentty. Just thought that it was a bit strange that there wasn't a tty for zuul itself ;) | 15:47 |
| fungi | well, there's always `wget -Qo- ... | flamel -` | 15:48 |
| avass | haha | 15:48 |
| corvus | czuul is what introduced me to urwid (the toolkit that gertty/presentty use) | 15:48 |
| fungi | speaking of which, i thank SpamapS in my head daily for flamel | 15:48 |
| fungi | i should thank him more out loud too | 15:48 |
| fungi | SpamapS: thanks again for making flamel! | 15:48 |
| dmsimard | flamel ? | 15:48 |
| SpamapS | fungi: 🙇 (bow) | 15:49 |
| dmsimard | SpamapS, fungi: link ? curious :) | 15:49 |
| SpamapS | pip install flamel | 15:49 |
| SpamapS | I use it daily. :) | 15:49 |
| fungi | me too | 15:50 |
| fungi | https://github.com/SpamapS/flamel | 15:50 |
| SpamapS | patches accepted: https://github.com/SpamapS/flamel | 15:50 |
| SpamapS | I should add a cheeky README | 15:50 |
| *** rlandy|rover is now known as rlandy|rover|mtg | 15:50 | |
| fungi | something something alchemy something | 15:50 |
| SpamapS | The name is a play on words with YAML and Nicholas Flamel, who claimed to be able to turn lead in to gold via alchemy. ;) | 15:50 |
| dmsimard | not exactly clear what it does at first glance -- loads stdin and pretty prints it ? | 15:50 |
| corvus | dmsimard: the project description says: | 15:51 |
| corvus | Flamel flips yaml to json and vice versa | 15:51 |
| fungi | translates back and forth between json and yaml | 15:51 |
| dmsimard | ohhhh | 15:51 |
| fungi | the implementation is fairly simple, but also very useful | 15:51 |
| dmsimard | neat | 15:51 |
| avass | gotta go, I'll check in later tonight to update my change in zuul-jobs | 15:52 |
| avass | (please have a look at it if you have time) | 15:52 |
| *** avass has quit IRC | 15:53 | |
| SpamapS | fungi: and many thanks to mordred for making the mapping keys appear in the same order as they appear in the input. | 15:54 |
| fungi | ahh, yep | 15:59 |
| fungi | though in python3.7+ that may happen by default anyway | 16:00 |
| *** rlandy|rover|mtg is now known as rlandy|rover | 16:01 | |
| SpamapS | fungi: I'm not so sure, that would require the json and YAML modules to keep the order in what they produce, which is the magic that mordred did. | 16:08 |
| SpamapS | But yeah probably. :) | 16:09 |
| SpamapS | anyway, thanks for the bump, I added a README. :) | 16:09 |
| fungi | right, i'm surmising they they both internally use python dicts and so would inherit the change to deterministic dict ordering anyway | 16:10 |
| fungi | my primary use is (per above) to grab unreadable rest api json vomit and shove it through stdin on a pipe to get pretty yaml out the other end | 16:12 |
| fungi | so the autodetection is especially useful there | 16:12 |
| SpamapS | same | 16:13 |
| SpamapS | also I use it to pipe yaml into jq sometimes. :) | 16:13 |
| fungi | oh, yep, that can also be handy. done that a few times myself | 16:13 |
| fungi | also, great readme! | 16:14 |
| AJaeger | corvus: in the context of jobs and branch matches, there's also http://lists.openstack.org/pipermail/openstack-discuss/2019-November/010990.html - hope my initial email was correct | 16:18 |
| *** gmann has joined #zuul | 16:21 | |
| corvus | AJaeger: your advice looks good to me | 16:23 |
| AJaeger | thanks for confirming | 16:27 |
| mordred | SpamapS: my pleasure - but also, I second the thankfulness to you for having created it in the first place. I use the crap out of it | 16:32 |
| corvus | should i make a build-podman-image role to mirror build-docker-image, or add a flag to build-docker-image to tell it to use podman commands (and then maybe start working to rename the roles)? | 16:51 |
| corvus | actually, maybe i should create 'build-container-image' and have it take a parameter, then later add upload- and promote-container-image, also with a parameter for the command system as well as target registry | 16:55 |
| *** tjgresha has joined #zuul | 17:08 | |
| *** michael-beaver has joined #zuul | 17:25 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 17:35 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add build-container-image role https://review.opendev.org/695251 | 17:35 |
| fungi | build-container-image sounds wonderfully generic, as long as "container images" are standardized (whether ietf or merely de facto) | 17:46 |
| *** igordc has joined #zuul | 18:00 | |
| *** jpena is now known as jpena|off | 18:00 | |
| *** mattw4 has joined #zuul | 18:19 | |
| *** tosky has quit IRC | 18:55 | |
| *** bhavikdbavishi has quit IRC | 18:55 | |
| *** jamesmcarthur has quit IRC | 19:14 | |
| *** avass has joined #zuul | 19:15 | |
| *** jamesmcarthur has joined #zuul | 19:27 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Adds go jobs and roles. https://review.opendev.org/691114 | 19:28 |
| *** jamesmcarthur has quit IRC | 20:20 | |
| *** jamesmcarthur has joined #zuul | 20:23 | |
| *** jamesmcarthur has quit IRC | 20:24 | |
| *** jamesmcarthur has joined #zuul | 20:26 | |
| *** michael-beaver has quit IRC | 20:45 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 20:54 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 21:22 |
| *** tosky has joined #zuul | 21:41 | |
| *** pcaruana has quit IRC | 21:42 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 21:47 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 22:04 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 22:18 |
| corvus | mordred, tristanC: i don't understand this error; it works locally: https://zuul.opendev.org/t/zuul/build/bb0eae37627c43e18ea8108a96e88ead/log/job-output.txt#966-981 | 22:41 |
| corvus | oh, it's not stdlib, i thought it was | 22:45 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: use-buildset-registry: Vendor pytoml and remarshal https://review.opendev.org/695050 | 22:48 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add build-container-image role https://review.opendev.org/695251 | 22:48 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 22:48 |
| corvus | i don't think it's used in that code path, so i excised it | 22:48 |
| *** jamesmcarthur has quit IRC | 22:57 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: use-buildset-registry: Vendor pytoml and remarshal https://review.opendev.org/695050 | 23:05 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add build-container-image role https://review.opendev.org/695251 | 23:05 |
| openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: WIP: use-buildset-registry: add podman support https://review.opendev.org/695051 | 23:05 |
| *** rlandy|rover is now known as rlandy|rover|bbl | 23:13 | |
| ianw | hrm, this is annoying ... it seems that dib when running in a container with nodepool in debug mode never exits | 23:56 |
| ianw | https://zuul.opendev.org/t/zuul/build/b5ab8b39657f41a68951fc0aa7a967ec/log/nodepool/builds/test-image-0000000002.log#7869 | 23:56 |
| ianw | it's stuck on a "wait" ... I don't think this happens in non-debug mode | 23:57 |
| ianw | i suspect something to do with open file descriptors ... ? anyone seen anything like that | 23:57 |
| ianw | it's a bit odd not running it in daemon mode (in container) ... perhaps nobody has tried this combo | 23:58 |
| clarkb | ianw: do you maybe need to do docker run -t for a tty? | 23:58 |
| ianw | yeah, i mean i wouldn't think so ... but maybe :/ | 23:59 |
| *** mattw4 has quit IRC | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!