| *** Goneri has quit IRC | 00:06 | |
| mordred | tristanC: but - there is a username/password yes? if so - I can't figure out how to configure buildkitd.toml to use it to talk to the registry mirror (or anything for that matter - I can't figure out how to do a "login" with buildkitd | 00:08 |
|---|---|---|
| tristanC | mordred: it should yes, at least according to https://opendev.org/zuul/zuul-jobs/src/branch/master/roles/run-buildset-registry/tasks/main.yaml#L61-L68 | 00:11 |
| openstackgerrit | Monty Taylor proposed zuul/zuul master: Install skopeo in container images https://review.opendev.org/724344 | 00:12 |
| mordred | ianw: ^^ | 00:12 |
| mordred | tristanC: awesome. well - I'll have to figure out how auth works tomorrow then :( | 00:12 |
| tristanC | though i have to say this whole container process in zuul-jobs is quite under documented and i find it difficult to grasp | 00:12 |
| corvus | tristanC: which part is under-documented exactly? | 00:13 |
| corvus | tristanC: because i think they are among the most documented parts of zuul jobs | 00:13 |
| tristanC | corvus: for example the answer to the question mordred was asking :) | 00:14 |
| corvus | tristanC: there's an abstract overview of the system here: https://zuul-ci.org/docs/zuul-jobs/docker-image.html | 00:14 |
| corvus | tristanC: and specific documentation of all the inputs and outputs of all the roles and jobs here: https://zuul-ci.org/docs/zuul-jobs/container-roles.html | 00:14 |
| *** rlandy has quit IRC | 00:14 | |
| corvus | tristanC: i really don't know what else we could do. maybe it's too much documentation? | 00:15 |
| tristanC | corvus: yes there is a lots of documentation, but that is also a complicated process | 00:15 |
| mordred | it is - it's a complex use case it's solving | 00:15 |
| corvus | yeah | 00:15 |
| mordred | it is literally solved nowhere else on the planet | 00:15 |
| corvus | and we've never used buildx yet | 00:16 |
| mordred | so there are some, you know, complex concepts :) | 00:16 |
| mordred | https://zuul-ci.org/docs/zuul-jobs/container-roles.html#rolevar-run-buildset-registry.buildset_registry.username is the answer to my earlier question - I as totally being lazy asking here | 00:16 |
| corvus | mordred: whew. thanks :) | 00:16 |
| mordred | corvus: but - what I was saying to tristanC - I have no idea how to tell buildkitd how to use that | 00:16 |
| mordred | if you want to see something that's underdocumented - see the docs for buildkitd.toml | 00:17 |
| mordred | THAT is under documented | 00:17 |
| tristanC | and it's built on top of container technology that also need to be understood | 00:17 |
| mordred | https://github.com/moby/buildkit/blob/master/docs/buildkitd.toml.md is it | 00:17 |
| mordred | *headdesk* | 00:18 |
| tristanC | corvus: unfortunately i don't know either what else we could do. Perhaps we could breakdown the documentation into the 4 tutorial/how-to/discussion/reference categories? | 00:18 |
| corvus | tristanC: why? | 00:19 |
| corvus | tristanC: what problem are we solving? | 00:19 |
| corvus | tristanC: that mordred didn't read the documentation that was there? | 00:19 |
| corvus | tristanC: that didn't happen because they weren't organized right | 00:19 |
| corvus | tristanC: that just happened because mordred didn't feel like reading the docs | 00:20 |
| corvus | that happens | 00:20 |
| corvus | we don't need to re-org the docs everytime a developer doesn't read them | 00:20 |
| mordred | well - I think _adding_ a blog-style howto wouldn't be bad - but I think we're stil learning things about usage patterns now that we have the tool. that wouldn't help *me* in this case at all, because what I'm having an issue with is internals | 00:20 |
| mordred | and thankfully - those internals are mostly hidden most of the time | 00:20 |
| corvus | mordred: howto what? | 00:20 |
| mordred | in fact, thte fact that they're so hidden is great! but it means I have to page things in when I want to work on internals | 00:21 |
| corvus | mordred: yeah, it's not easy. docker is never going to make it easy. | 00:21 |
| mordred | corvus: how to set up some of the more complicated job relationships that we've gone through a couple of iterations to get right over in system-config. I feel like at some point a blog-post retrospective on that would be nice | 00:21 |
| mordred | because I feel like we learned some lessons on useful patterns while doing it | 00:22 |
| tristanC | corvus: it seems like this problem is kind of solved for opendev, but it seems quite difficult to replicate. | 00:22 |
| corvus | mordred: yeah. i don't think we've even completely figured them out yet. | 00:22 |
| mordred | corvus: exactly | 00:22 |
| corvus | tristanC: ah, i see what you're getting at. let me make a suggestion: | 00:22 |
| mordred | that's wy I don't think it's time to write that yet | 00:22 |
| corvus | tristanC: these jobs are not directly consumable from zuul-jobs; they need an implementation to be provided by the zuul administrator. so we can't make it super-easy for end users, but we can do this: | 00:23 |
| corvus | tristanC: this documentation tells a zuul site admin how to set the system up: https://zuul-ci.org/docs/zuul-jobs/docker-image.html | 00:23 |
| corvus | tristanC: and then they can document how to use that system for their end users. opendev does that here: https://docs.opendev.org/opendev/base-jobs/latest/docker-image.html | 00:24 |
| corvus | tristanC: anyone of course is welcome to copy and adapt that last bit of documentation for their own use | 00:24 |
| corvus | tristanC: and yeah, once we have a better handle on how we should recommend people actually use these, we can probably consider that the reference documentation and add a blogpost/howto/tutorial that's more like "add this job. type "requires:foo" here. etc." | 00:25 |
| mordred | yeah | 00:27 |
| mordred | we're still actively learning lessons and getting stuff wrong | 00:27 |
| corvus | (but note that opendev documentation is pretty close to that -- it does say what jobs to put in what kinds of repositories) | 00:27 |
| mordred | corvus: ianw taught me something re: https://review.opendev.org/#/c/724344/ - and it is now more simple | 00:31 |
| mordred | corvus: that thing only works in latest ubuntu and latest debian - but we're using those here, so it's ok | 00:31 |
| tristanC | corvus: i didn't knew opendev base-jobs had such a docker-image documentation, it's quite handy. Shouldn't it be referenced from zuul-jobs or zuul-registry? | 00:40 |
| *** guillaumec has quit IRC | 01:07 | |
| *** swest has quit IRC | 01:57 | |
| *** swest has joined #zuul | 02:12 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: [wip] xenial pip install https://review.opendev.org/724788 | 02:24 |
| *** bhavikdbavishi has joined #zuul | 02:51 | |
| *** bhavikdbavishi1 has joined #zuul | 03:32 | |
| *** bhavikdbavishi has quit IRC | 03:34 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:34 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: [wip] xenial pip install https://review.opendev.org/724788 | 03:37 |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: [wip] xenial pip install https://review.opendev.org/724788 | 03:44 |
| *** evrardjp has quit IRC | 04:35 | |
| *** evrardjp has joined #zuul | 04:36 | |
| *** bhavikdbavishi has quit IRC | 04:51 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: [wip] xenial pip install https://review.opendev.org/724788 | 05:18 |
| *** zxiiro has quit IRC | 05:23 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: [wip] xenial pip install https://review.opendev.org/724788 | 05:24 |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: [wip] xenial pip install https://review.opendev.org/724788 | 05:30 |
| openstackgerrit | Ian Wienand proposed zuul/zuul-jobs master: [wip] xenial pip install https://review.opendev.org/724788 | 05:36 |
| *** sgw has quit IRC | 06:00 | |
| *** bhavikdbavishi has joined #zuul | 06:03 | |
| *** bhavikdbavishi has quit IRC | 06:46 | |
| masterpe | Good morning, I see that there is a gitlab driver for zuul: https://opendev.org/zuul/zuul/src/branch/master/zuul/driver/gitlab but is there also a manual how to configure it? | 07:05 |
| *** yolanda has quit IRC | 07:38 | |
| *** bhavikdbavishi has joined #zuul | 07:47 | |
| *** yolanda has joined #zuul | 07:51 | |
| *** bhavikdbavishi has quit IRC | 08:02 | |
| *** bhavikdbavishi has joined #zuul | 08:02 | |
| *** bhavikdbavishi1 has joined #zuul | 08:37 | |
| *** bhavikdbavishi has quit IRC | 08:39 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 08:39 | |
| AJaeger | avass: could you check for your changes to go from role install- to ensure-* whether those need backporting, please? I backported the magnum one now | 09:17 |
| avass | AJaeger: to release branches? sure | 09:23 |
| AJaeger | avass: yes, to stable ones like the magnum ones I did | 09:25 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Control log archive and user preservation with vars https://review.opendev.org/701381 | 09:25 |
| avass | AJager: I'll see what I can do, got the day off today so I have all the time in the world :) | 09:26 |
| AJaeger | avass: I think the airship ones might need backporting as well, openstack and opendev looks ok | 09:26 |
| AJaeger | avass: enjoy ;) | 09:26 |
| AJaeger | and no urgency, just a reminder :) | 09:26 |
| avass | I also saw that we're using 'with_items' instead of 'loop' on some tasks, checking if it's possible to use loop_control on those and if not, maybe we should avoid 'with_' type loops entirely? | 09:27 |
| AJaeger | sorry, can't help with that one... | 09:28 |
| avass | :) | 09:28 |
| AJaeger | mnaser: ok to +2A now https://review.opendev.org/724132 ? | 09:29 |
| *** tosky has joined #zuul | 09:30 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Do not keep owner when pulling zuul-output https://review.opendev.org/701381 | 09:39 |
| *** yolanda has quit IRC | 09:43 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Check for loop_control in with_ type loops https://review.opendev.org/724810 | 09:48 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-repositories: Add role https://review.opendev.org/717507 | 09:49 |
| zbr | afaik, there is are no plans from ansible team to deprecate with_ --- they wanted it at some point until they realised loop implementation made very common use practices much harder | 09:50 |
| zbr | avass: am curious what are the reasons behind dropping the use of item/with_items? | 09:51 |
| avass | zbr: yeah, looks like you can use loop_control on with_ type loops so it's fine | 09:51 |
| avass | zbr: https://zuul-ci.org/docs/zuul-jobs/policy.html#ansible-loops-in-roles | 09:52 |
| zbr | yes you can, loop can do anything with_* do, but at the cost of making much harder to read the code | 09:52 |
| avass | zbr: So that ^ should also be applied to with_ type loops, not just 'loop' | 09:52 |
| avass | AJaeger: can you update the with_items to use loop control? :) | 09:53 |
| zbr | afaik, use of loops with includes and blocks was never supported anyway | 09:53 |
| avass | zbr: you can't loop over blocks but you can loop over includes, so in what way is it not supported? | 09:54 |
| avass | not sure what you mean | 09:55 |
| zbr | what if you have a block inside the role.... | 09:55 |
| zbr | avass: i am not saying this change is a mistake, just trying to gather facts, if true it means that in general roles should use unique vars in their loops? | 09:56 |
| zbr | i bet 99% ore roles over written do not follow this practice | 09:57 |
| avass | zbr: I guess you usually don't need unique loop vars, but the roles in zuul-jobs are supposed to be easily imported to other projects | 09:57 |
| zbr | if that is really needed, it sounds like a cool new feature to add to ansible-lint | 09:57 |
| avass | zbr: but I guess collections would need unique loop vars as well | 09:58 |
| avass | zbr: so yeah, maybe all roles should avoid using the default loop_var, but if we're thinking on that scope maybe there just shouldn't be a default loop var? | 09:59 |
| avass | would be hard to deprecate that though | 09:59 |
| zbr | close to impossible | 10:00 |
| zbr | in fact I did like the implicit loop var, make code easier to read, less boilerplate | 10:00 |
| reiterative | avass Thanks, I thought that was the case. Just wasn't sure about the significance of the Workflow label.. | 10:03 |
| avass | zbr: yeah I like it too, but as soon as you want to loop over the role it can cause problems | 10:06 |
| avass | reiterative: no problem! | 10:06 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-repositories: Add role https://review.opendev.org/717507 | 10:12 |
| avass | AJaeger: thank you! | 10:13 |
| AJaeger | avass: further updates needed... | 10:15 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-repositories: Add role https://review.opendev.org/717507 | 10:18 |
| AJaeger | avass: is that one ok now? ^ | 10:18 |
| AJaeger | quick first ok would be fine - then I'll rebase the rest on top of it... | 10:19 |
| AJaeger | I meant "quick review" | 10:21 |
| * AJaeger rebases now everyting on master... | 10:22 | |
| avass | AJaeger: I'll take a look | 10:28 |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul master: Add convenience Makefile https://review.opendev.org/723837 | 10:28 |
| AJaeger | avass: please give me ten minutes to finish rebasing | 10:28 |
| *** bhavikdbavishi has quit IRC | 10:29 | |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-repositories: Add role https://review.opendev.org/717507 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-repositories https://review.opendev.org/717508 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-repositories https://review.opendev.org/717509 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-repositories https://review.opendev.org/717510 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure-repositories https://review.opendev.org/717511 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-repositories https://review.opendev.org/717512 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-repositories https://review.opendev.org/717513 | 10:31 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 10:33 |
| avass | zbr: there's no difference between with_item vs loop right? | 10:33 |
| zbr | what do you mean by difference? | 10:34 |
| AJaeger | so, everything rebased and updated - let's see whether it still works ;) | 10:34 |
| zbr | loops are newer, so they may have problems with really old ansible versions | 10:34 |
| avass | zbr: looks like they're used the same way at least :) | 10:34 |
| avass | so I'm doing a simple with_items -> loop in a lot of places | 10:34 |
| avass | AJaeger: I'll review once I see a +1 verified then :) | 10:35 |
| AJaeger | avass: no urgency - this was lingering several weeks ;) | 10:37 |
| AJaeger | I'll take a longer break now.. | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-repositories: Add role https://review.opendev.org/717507 | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-repositories https://review.opendev.org/717508 | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-repositories https://review.opendev.org/717509 | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-repositories https://review.opendev.org/717510 | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure-repositories https://review.opendev.org/717511 | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-repositories https://review.opendev.org/717512 | 10:38 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-repositories https://review.opendev.org/717513 | 10:38 |
| AJaeger | argh, forgot to push latest revision - now good... | 10:39 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 10:39 |
| zbr | interesting, cannot build zuul docs on mac, `zuul-manage-ansible -h` produces a segmentation fault | 10:40 |
| zbr | i wonder if it would not be useful to add a BSD distro nodepool image to use for some tests | 10:42 |
| zbr | oh fb-re2 joys, compiles and installs but re2.compile() chokes. | 10:59 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-repositories: Add role https://review.opendev.org/717507 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-repositories https://review.opendev.org/717508 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-repositories https://review.opendev.org/717509 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-repositories https://review.opendev.org/717510 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure-repositories https://review.opendev.org/717511 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-repositories https://review.opendev.org/717512 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-repositories https://review.opendev.org/717513 | 11:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 11:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 11:09 |
| AJaeger | mnaser: please review my changes ^ | 11:14 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Check for loop_control in with_ type loops https://review.opendev.org/724810 | 11:17 |
| masterpe | I'm trying to configure zuul, I see in executor.log the following message | 11:18 |
| avass | big change, hope I don't break something this time | 11:18 |
| masterpe | fatal: could not read Username for 'https://xxxxxxx': No such device or address' | 11:18 |
| masterpe | at first I had in /etc/zuul/zuul.conf in section [merger] the wrong git_user_name, I correct that, But I still see the message | 11:19 |
| avass | masterpe: can you send a snippet of the logs in a paste like: http://paste.openstack.org/ ? | 11:20 |
| avass | easier to see what could be wrong if I know the context :) | 11:20 |
| avass | masterpe: oh is this a github issue? | 11:22 |
| masterpe | http://paste.openstack.org/show/792982/ | 11:24 |
| masterpe | I use gitlab as driver | 11:24 |
| avass | hmm, I don't know how well the gitlab driver is supported yet and I haven't used it | 11:27 |
| *** bhavikdbavishi has joined #zuul | 11:31 | |
| avass | masterpe: what does your zuul.conf look like? | 11:32 |
| masterpe | I have the username in the merger section and gitlab | 11:33 |
| masterpe | I'm currently away will look into it an hour, when I'm back | 11:34 |
| avass | masterpe: looking at the source I guess you set either cloneurl, baseurl or server in the gitlab connection with username and password? | 11:36 |
| mgoddard | Hi, I'm trying to use Ansible's reboot module in a Zuul but it's giving me an ERROR. Anyone tried it before? | 11:38 |
| mgoddard | https://a1110360d0aa5b1915ef-de0170061a4c33caf35d62b404ee94ad.ssl.cf5.rackcdn.com/724819/2/check/kayobe-overcloud-host-configure-centos8/d17c0af/job-output.txt | 11:38 |
| mgoddard | https://review.opendev.org/724819 | 11:38 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Check for loop_control in with_ type loops https://review.opendev.org/724810 | 11:39 |
| avass | mgoddard: if you check the json logs it gives you more info | 11:40 |
| *** lseki has quit IRC | 11:40 | |
| avass | mgoddard: http://paste.openstack.org/show/792984/ | 11:41 |
| avass | mgoddard: first guess is that you need to use 'become:' and execute it as root | 11:42 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Check for loop_control in with_ type loops https://review.opendev.org/724810 | 11:45 |
| *** Goneri has joined #zuul | 12:00 | |
| mgoddard | thanks avass | 12:05 |
| avass | mgoddard: no problem :) | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-repositories: Add role https://review.opendev.org/717507 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-repositories https://review.opendev.org/717508 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-repositories https://review.opendev.org/717509 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-repositories https://review.opendev.org/717510 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure-repositories https://review.opendev.org/717511 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-repositories https://review.opendev.org/717512 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-repositories https://review.opendev.org/717513 | 12:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 12:09 |
| *** saneax has joined #zuul | 12:12 | |
| fungi | masterpe: fbo|off has been working on the gitlab driver. the main reason it's undocumented still is that it's experimental and missing critical feature parity with other source connection drivers (you can't use it for project gating yet, as i understand), but he did have a working poc of it doing advisory testing on a gitlab deployment (or maybe it was gitlab.com) since late last year | 12:21 |
| masterpe | @av | 12:21 |
| masterpe | avass: http://paste.openstack.org/show/792985/ is the zuul config | 12:21 |
| avass | masterpe: ah yeah, looking at the source it seems the username/password needs to be in the server/baseurl/cloneurl | 12:22 |
| fungi | aha, he had it working with gitlab.com apparently... https://gitlab.com/fabien.dot.boucher/demo-zuul/merge_requests/7#note_244782493 | 12:24 |
| avass | fungi: nice! I think gitlab is something we're going to have in the future :) | 12:24 |
| *** rlandy has joined #zuul | 12:24 | |
| avass | fungi: or, I mean we have it, be we probably want to be able to gate those projects with zuul | 12:24 |
| fungi | for reference, these are the changes which implemented what's there so far: https://review.opendev.org/#/q/project:zuul/zuul+topic:gitlab | 12:27 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 12:34 |
| masterpe | avass: using username:password in the server config works, but it is not secure, there is also visible in the dashboard | 12:36 |
| avass | masterpe: you can use %()s to interpolate from environment variables in zuul.conf | 12:37 |
| avass | let me link something | 12:37 |
| avass | masterpe: Oh, or yo mean the web dashboard? | 12:37 |
| masterpe | yes | 12:37 |
| avass | masterpe: I think that shows the canonical_url, what happens if you set that value? | 12:38 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 12:38 |
| fungi | but yeah, for safety's sake that value should probably move to a separate option similar to https://zuul-ci.org/docs/zuul/reference/drivers/github.html#attr-%3Cgithub%20connection%3E.webhook_token | 12:41 |
| fungi | it would be better if leaving out the canonical_url inadvertently exposed a password | 12:42 |
| fungi | er, better if it *didn't* i mean | 12:42 |
| * fungi tries to get more coffee into his veins | 12:42 | |
| avass | masterpe: yeah, canonical_url defaults to server if it's not set | 12:42 |
| avass | fungi: yeah | 12:43 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-package-repositories: Add role https://review.opendev.org/717507 | 12:46 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-package-repositories https://review.opendev.org/717508 | 12:46 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-package-repositories https://review.opendev.org/717509 | 12:46 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 12:46 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-package-repositories https://review.opendev.org/717510 | 12:46 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure--package-repositories https://review.opendev.org/717511 | 12:46 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-package-repositories https://review.opendev.org/717512 | 12:46 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-package-repositories https://review.opendev.org/717513 | 12:46 |
| masterpe | had to use the cloneurl to solve this | 12:47 |
| avass | this is ready: https://review.opendev.org/#/c/724810/ :) | 12:48 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 12:49 |
| avass | masterpe: yeah that looks like a better option :) | 12:49 |
| AJaeger | if we use a role ensure-nodejs from ensure-yarn, do we add a file-matcher on ensure-yarn for ensure-nodejs as well? See https://review.opendev.org/#/c/717817/9/zuul-tests.d/js-roles-jobs.yaml | 12:53 |
| * AJaeger thinks we need to update the stack a bit more for this, but wants feedback from zuul-jobs experts on this ^ | 12:53 | |
| * AJaeger found the answer, we do this in other case. Sorry, another change set is coming... | 12:54 | |
| avass | AJaeger: I'd say yes | 12:55 |
| avass | AJaeger: I think it's more important that we don't break something than we run more tests | 12:55 |
| avass | :) | 12:55 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write buildkitd.toml in use-buildset-registry https://review.opendev.org/724837 | 12:58 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-package-repositories https://review.opendev.org/717508 | 13:01 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-package-repositories https://review.opendev.org/717509 | 13:01 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 13:01 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-package-repositories https://review.opendev.org/717510 | 13:01 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure--package-repositories https://review.opendev.org/717511 | 13:01 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-package-repositories https://review.opendev.org/717512 | 13:01 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-package-repositories https://review.opendev.org/717513 | 13:01 |
| AJaeger | avass: agreed, and found we do it elsewhere -so updated again... | 13:01 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 13:03 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write buildkitd.toml in use-buildset-registry https://review.opendev.org/724837 | 13:03 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 13:05 |
| mordred | AJaeger: comment on that last one | 13:05 |
| mordred | AJaeger: on 717513 | 13:05 |
| AJaeger | mordred: only on the last one? Good ;) thanks | 13:06 |
| avass | mordred: heh, tristanC already commented that on an earlier patchset ;) | 13:07 |
| mordred | haven't looked at the others yet :) | 13:07 |
| mordred | haha | 13:07 |
| AJaeger | I wanted to do that - let me check where I screwed up | 13:07 |
| mordred | AJaeger, avass: so - that has pointed out to me I've need to go update our use of ppa:projectatomic ... | 13:08 |
| mordred | turns out we want to use it for xenial but for bionic and focal we want to use kubic | 13:08 |
| mordred | \o/ | 13:09 |
| AJaeger | ah, there were two places where it needed removing, I only touched one - done now. | 13:09 |
| mordred | sweet | 13:09 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-package-repositories https://review.opendev.org/717513 | 13:09 |
| AJaeger | mordred: so, want to make a change and I rebase this stack - or help merging this stack? ;) | 13:12 |
| mordred | AJaeger: oh - I'll totally do it after that stack :) | 13:12 |
| mordred | AJaeger: I have way too many balls juggling right now | 13:13 |
| avass | I know that feeling | 13:13 |
| * AJaeger knows it as well... | 13:13 | |
| AJaeger | anybody an idea why Zuul complains now on https://review.opendev.org/#/c/717513 about "an invalid configuration". Everything else in the stack looks sane | 13:15 |
| corvus | mordred: with https://review.opendev.org/724344 approved do you want to abandon https://review.opendev.org/717613 ? :) | 13:16 |
| avass | AJaeger: I think that's a bug | 13:16 |
| AJaeger | tobiash, tristanC, want to review the ensure-repositories stack again? It starts at https://review.opendev.org/#/c/717507/ and I just updated heavily | 13:17 |
| corvus | the memory use on opendev is much better; enough that i think we can say that it looks like a memory leak in latest cherrypy | 13:17 |
| AJaeger | avass: ok, let's wait... | 13:17 |
| AJaeger | corvus: great | 13:18 |
| corvus | webknjaz: hi, i suspect there is a memory leak in cherrypy 18.6.0 | 13:18 |
| avass | corvus, mordred, AJaeger: update to loop-var linting so we catch with_ type loops is ready: https://review.opendev.org/#/c/724810/ | 13:19 |
| tristanC | AJaeger: thanks! | 13:21 |
| AJaeger | avass: wow, will review later. | 13:21 |
| avass | AJaeger: yeah, might have to update it later if any loops get merged before it :) | 13:22 |
| mordred | corvus: wow - I did that twice? | 13:22 |
| AJaeger | mordred: a wise man said recently "I have way too many balls juggling right now" ;) | 13:23 |
| avass | haha | 13:24 |
| AJaeger | corvus: you had a -1 earlier on https://review.opendev.org/#/c/717817/ - could you review again, please? | 13:24 |
| mordred | AJaeger: hahaha | 13:24 |
| AJaeger | argh, stack still broken - fixing RedHat installs.. | 13:26 |
| avass | corvus: if we don't want to keep the owner of unarchived or synchronized files between executor and remote, how about we add a linting rule for that too? | 13:27 |
| mordred | corvus: latest buildkit patch didn't work - but looking at the registry logs, it's showing containerd as the UA - so maybe I need to do https://github.com/containerd/cri/blob/master/docs/registry.md | 13:28 |
| corvus | avass: have we fixed that yet and proven it everywhere? | 13:28 |
| avass | corvus: no :) | 13:29 |
| corvus | avass: we should probably make sure the pattern is well established, proven, and always necessary before adding a linting rule; and it should probably be in the policy docs too :) | 13:29 |
| AJaeger | https://zuul.opendev.org/t/zuul/build/050ceb250c7f4c878ef40d3f4523b614 fails - could anybody tell me what I did wrong in https://review.opendev.org/#/c/717507/11/roles/ensure-package-repositories/tasks/RedHat.yaml , please? | 13:30 |
| AJaeger | corvus: thx | 13:30 |
| avass | corvus: yeah I was thinking of writing some docs for that when I start working on it :) | 13:30 |
| corvus | mordred: do we not do that one already? (jeez, how many of these are there?) | 13:30 |
| avass | AJaeger: left a comment | 13:32 |
| corvus | mordred: ok, i don't see it. so i guess we should add that | 13:32 |
| mordred | corvus: right? | 13:33 |
| AJaeger | thanks, avass | 13:34 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-package-repositories: Add role https://review.opendev.org/717507 | 13:36 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-package-repositories https://review.opendev.org/717508 | 13:36 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-package-repositories https://review.opendev.org/717509 | 13:36 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 13:36 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-package-repositories https://review.opendev.org/717510 | 13:36 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure--package-repositories https://review.opendev.org/717511 | 13:36 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-package-repositories https://review.opendev.org/717512 | 13:36 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-package-repositories https://review.opendev.org/717513 | 13:36 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: use-buildset-registry: fix modify_registries_conf library idempotency https://review.opendev.org/724840 | 13:36 |
| mordred | corvus: I don't _know_ that that config will be the one | 13:36 |
| mordred | corvus: it's also possible we need to restart buildkitd in the container | 13:37 |
| avass | corvus: but since the synchronize owner/group parameter tries to set the owner to user on the remote, I'd say it's necessary to make sure it works for every system. Unless we want to force everyone to create a user on the executor for every user they have on a remote node :) | 13:37 |
| corvus | mordred: it looks like we should be able to add the prefix needed to support multiple registries; so this should enable more than docker.io. | 13:37 |
| corvus | avass: i'm pretty sure you're right; i just don't want the linting rules to get ahead of the practice :) | 13:37 |
| mordred | corvus: which - I'm going to try now on my test node, because I'm not sure if that's even possible if it's the only process in the container - we might have to build our own image and tell buildx to use it | 13:37 |
| mordred | corvus: heh. yeah - killing buildkitd sure does kill the container - but docker start sure does re-start it :) | 13:39 |
| avass | corvus: sure, I'll push a change for it to get it ready, but we can wait with merging it | 13:39 |
| mordred | corvus: actually - before I go writing containerd.toml - I'm going to toss in a container restart - I think that might be required to pick up the config.json now that I think of it | 13:40 |
| corvus | mordred: https://github.com/containerd/containerd/blob/master/.zuul.yaml | 13:40 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: use-buildset-registry: do not update ca when not necessary https://review.opendev.org/724841 | 13:40 |
| corvus | tristanC, fbo|off: ^ fyi containerd repo has a .zuul.yaml from openlab | 13:41 |
| tristanC | corvus: that is nice :-) | 13:42 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 13:43 |
| avass | AJaeger: left another comment :) | 13:44 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write buildkitd.toml in use-buildset-registry https://review.opendev.org/724837 | 13:44 |
| tristanC | corvus: containers/libpod doesn't have a zuul.yaml, but rdoproject's zuul check each PR | 13:44 |
| *** sanjayu_ has joined #zuul | 13:44 | |
| fungi | https://github.com/theopenlab/ | 13:45 |
| corvus | tristanC: that's some nice cross-community collaboration going on :) | 13:45 |
| fungi | interestingly, they have a zuul-jobs fork from the defunct openstack-infra/zuul-jobs github mirror | 13:45 |
| fungi | https://openlabtesting.org/ seems to have an expired letsencrypt cert | 13:46 |
| fungi | several weeks expired | 13:46 |
| *** saneax has quit IRC | 13:47 | |
| masterpe | How can I figerout what to do with the pipeline in the config repo with the gitlab driver? | 13:47 |
| corvus | masterpe: there isn't complete documentation since it's not finished yet, but a lot of it should be similar to github: https://zuul-ci.org/docs/zuul/reference/drivers/github.html#reference-pipelines-configuration | 13:48 |
| fungi | looks like openlab's zuul instance is still up and running, at least: http://status.openlabtesting.org/ | 13:49 |
| *** bhavikdbavishi has quit IRC | 13:49 | |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-package-repositories: Add role https://review.opendev.org/717507 | 13:49 |
| *** guillaumec has joined #zuul | 13:49 | |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-package-repositories https://review.opendev.org/717508 | 13:49 |
| AJaeger | avass: thanks | 13:49 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-package-repositories https://review.opendev.org/717509 | 13:49 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 13:49 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-package-repositories https://review.opendev.org/717510 | 13:49 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-podman: refactor to use ensure--package-repositories https://review.opendev.org/717511 | 13:49 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-docker: refactor to use ensure-package-repositories https://review.opendev.org/717512 | 13:49 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: cleanup: move tests to use ensure-package-repositories https://review.opendev.org/717513 | 13:49 |
| corvus | masterpe: /win 11 | 13:50 |
| corvus | oops | 13:50 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-operator master: Add CA certificate to the registry tls secret https://review.opendev.org/724844 | 13:51 |
| *** bhavikdbavishi has joined #zuul | 13:51 | |
| corvus | mordred: i just double checked and didn't see a way to reload the config without restarting containerd | 13:55 |
| AJaeger | ensure-package-repositories stack loooks passing now. please review https://review.opendev.org/#/c/717507 and https://review.opendev.org/717513 so that we can merge this stack | 13:56 |
| * AJaeger enjoys tea now... | 13:56 | |
| mordred | corvus: I tossed in a docker restart - let's see what that does for us | 14:01 |
| *** bhavikdbavishi has quit IRC | 14:05 | |
| mordred | corvus: quick question ... | 14:07 |
| mordred | corvus: reading this log: https://zuul.opendev.org/t/zuul/build/8d355b2bd85a4a5c8a8ae8e3bf348a9f/log/builder/docker/buildset_registry.txt#124 | 14:08 |
| masterpe | If I use the example pipeline on https://zuul-ci.org/docs/zuul/reference/drivers/github.html#reference-pipelines-configuration I get a couple of failues: http://paste.openstack.org/show/792988/ @fbo can post an example? | 14:09 |
| mordred | corvus: it looks like we might be authenticating? but looking for things that aren't there? | 14:09 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 14:18 |
| openstackgerrit | Merged zuul/zuul master: Install skopeo in container images https://review.opendev.org/724344 | 14:19 |
| *** sanjayu__ has joined #zuul | 14:19 | |
| *** sanjayu_ has quit IRC | 14:22 | |
| AJaeger | yeah,stack starting at https://review.opendev.org/717513 passes fully - please review | 14:23 |
| AJaeger | https://review.opendev.org/#/c/717507 is the start of the ensure-package-repositories stack - 717513 is the end | 14:24 |
| AJaeger | those two are the only ones that need two +2s | 14:25 |
| *** lseki has joined #zuul | 14:30 | |
| AJaeger | tristanC, avass, in case you want to review again, please ^ | 14:33 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 14:33 |
| AJaeger | oh, there are more changes that need a second +2 in the stack... | 14:35 |
| AJaeger | mordred: thanks for some reviews | 14:35 |
| avass | AJaeger: I'll look :) | 14:35 |
| AJaeger | thx | 14:36 |
| AJaeger | avass: once that stack is merged, I'll review others - including your loop change | 14:37 |
| corvus | mordred: yeah https://zuul.opendev.org/t/zuul/build/8d355b2bd85a4a5c8a8ae8e3bf348a9f/log/builder/docker/buildset_registry.txt#134 is a 404 -- zuul-registry doesn't implement HEAD on a manifest | 14:37 |
| mordred | corvus: so - we are now authenticating, but there's a missing verb in zuul-registry? | 14:40 |
| mordred | (just making sure I'm getting it all right in my brain head) | 14:41 |
| corvus | mordred: yep, i'm looking at adding it | 14:41 |
| mordred | cool | 14:42 |
| avass | AJaeger: everything looks good to me | 14:42 |
| AJaeger | thanks, avass | 14:42 |
| avass | AJaeger: and it didn't break the loop linting :) | 14:43 |
| AJaeger | great! | 14:44 |
| *** sanjayu__ has quit IRC | 14:46 | |
| AJaeger | corvus, https://zuul.opendev.org/t/zuul/status hows me that https://review.opendev.org/717510 is failing since "it has an invalid configuration" - but it passed checks | 14:46 |
| AJaeger | that looks like a bug - isn't it? | 14:47 |
| openstackgerrit | James E. Blair proposed zuul/zuul-registry master: Support HEAD on manifests https://review.opendev.org/724859 | 14:47 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 14:48 |
| corvus | mordred: 724859 | 14:48 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 14:50 |
| mordred | corvus: depends-on inception | 14:50 |
| *** zxiiro has joined #zuul | 14:51 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 14:51 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: Do not keep owner when pulling zuul-output https://review.opendev.org/701381 | 14:51 |
| AJaeger | avass: just addressed the nits ^ | 14:51 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 14:52 |
| avass | AJaeger: thanks! | 14:53 |
| AJaeger | corvus: you had a -1 on 701381 earlier, want to review again? | 14:54 |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-package-repositories: Add role https://review.opendev.org/717507 | 14:57 |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-nodejs: refactor to use ensure-package-repositories https://review.opendev.org/717508 | 14:57 |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-yarn: refactor to use ensure-package-repositories https://review.opendev.org/717509 | 14:57 |
| openstackgerrit | Merged zuul/zuul-jobs master: install-yarn: always install https://review.opendev.org/717375 | 14:57 |
| mordred | corvus: interestingly - https://review.opendev.org/#/c/724837/ - also fails on the podman job | 14:57 |
| AJaeger | corvus: https://review.opendev.org/717510 has "unknown configuration error" | 14:58 |
| AJaeger | zuul-maint, should I recheck or is there something you want to investigate? | 14:59 |
| avass | AJaeger: recheck :) | 15:01 |
| avass | AJaeger: I think it's known, but no one has figured out why it happens yet | 15:02 |
| AJaeger | thx - will do in a few minutes | 15:02 |
| avass | AJaeger: except that it happens whenever you have big stacks like that | 15:02 |
| corvus | AJaeger: 701381 lgtm; it's tested so should be safe to land, but it'll be good to keep an eye on opendev jobs when it does | 15:03 |
| zbr | avass: regarding ansilbe-lint, I just added a RFP at https://github.com/ansible/ansible-lint/issues/755 | 15:03 |
| AJaeger | corvus: I'll merge slowly one stack after the other... | 15:03 |
| corvus | mordred, clarkb: can one of you also look at https://review.opendev.org/701381 ? ( AJaeger, avass: let's get one more set of eyes on that) | 15:04 |
| AJaeger | corvus: ok | 15:05 |
| *** rlandy is now known as rlandy|biab | 15:05 | |
| avass | corvus: sure | 15:05 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 15:08 |
| openstackgerrit | Sorin Sbarnea (zbr) proposed zuul/zuul-jobs master: fetch-sphinx-tarbal: allow to follow symlinks https://review.opendev.org/724868 | 15:13 |
| avass | zbr: interesting | 15:14 |
| clarkb | corvus: avass AJaeger I've approved it | 15:14 |
| AJaeger | thanks, clarkb | 15:15 |
| zbr | avass: see https://dashboard.zuul.ansible.com/t/ansible/build/e959ce3203dc495cabe9bef7caaf4040/console | 15:15 |
| *** sgw has joined #zuul | 15:21 | |
| mordred | corvus: the podman failure is now showing up in more places | 15:21 |
| mordred | corvus: https://zuul.opendev.org/t/zuul/build/8b5cbca02d89483e9714e9e8348980d5 | 15:21 |
| mordred | corvus: https://zuul.opendev.org/t/zuul/build/8b5cbca02d89483e9714e9e8348980d5/log/builder/podman/buildset_registry.txt#220 | 15:21 |
| *** sanjayu__ has joined #zuul | 15:24 | |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-kubernetes: refactor to use ensure-package-repositories https://review.opendev.org/717510 | 15:28 |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-podman: refactor to use ensure--package-repositories https://review.opendev.org/717511 | 15:28 |
| tobiash | mordred: did that multi arch change land could that have something to do with those failures? | 15:29 |
| *** dpawlik has joined #zuul | 15:30 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: haskell-stack-test: add haskell tool stack test https://review.opendev.org/723263 | 15:35 |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-docker: refactor to use ensure-package-repositories https://review.opendev.org/717512 | 15:36 |
| AJaeger | I had to resolve a merge conflict on https://review.opendev.org/717817 - could I get reviews again, please? That is part of the ensure-package-repositories zuul-jobs stack | 15:38 |
| openstackgerrit | Merged zuul/zuul-jobs master: cleanup: move tests to use ensure-package-repositories https://review.opendev.org/717513 | 15:39 |
| openstackgerrit | Merged zuul/zuul-jobs master: Add Bazel build and ensure roles https://review.opendev.org/693513 | 15:39 |
| openstackgerrit | Merged zuul/zuul-jobs master: Do not keep owner when pulling zuul-output https://review.opendev.org/701381 | 15:39 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 15:41 |
| zbr | corvus: mordred: any chance to do https://review.opendev.org/#/c/724868/ ? | 15:43 |
| *** bhavikdbavishi has joined #zuul | 15:45 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 15:48 |
| avass | corvus: alright, I think we can keep the discussion in that ^ change now :) | 15:48 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 15:49 |
| mordred | tobiash: nope, still working on it | 15:50 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Add ansible-lint rule to check owner and group is not preserved https://review.opendev.org/724855 | 15:53 |
| AJaeger | avass: want to review https://review.opendev.org/723263 again after the rename of directories, please? | 15:55 |
| avass | AJaeger: sure | 15:56 |
| AJaeger | thx | 15:56 |
| AJaeger | clarkb, mordred, corvus, avass has a stack up to change the "ignore_errors" we use, could one of you review the stack starting at https://review.opendev.org/#/c/723640/ , please? | 15:57 |
| *** bhavikdbavishi has quit IRC | 16:01 | |
| avass | I'll be back in an hour, going for a walk :) | 16:03 |
| clarkb | AJaeger: I feel like I don't know enough about ansible to properly say if the old or new code is preferred there | 16:04 |
| openstackgerrit | Merged zuul/zuul-jobs master: Check for loop_control in with_ type loops https://review.opendev.org/724810 | 16:06 |
| AJaeger | clarkb: thx for checking | 16:06 |
| *** bhavikdbavishi has joined #zuul | 16:06 | |
| *** dpawlik has quit IRC | 16:08 | |
| openstackgerrit | Merged zuul/zuul-jobs master: fetch-sphinx-tarbal: allow to follow symlinks https://review.opendev.org/724868 | 16:08 |
| openstackgerrit | Merged zuul/zuul-jobs master: haskell-stack-test: add haskell tool stack test https://review.opendev.org/723263 | 16:09 |
| AJaeger | avass: enjoy the walk. could you double check later your loop_control change. I would expect that "item" is unused now but it's not, like in test-playbooks/multinode/persistent-firewall.yaml | 16:11 |
| tristanC | AJaeger: avass: thank you for reviewing the new haskell jobs! | 16:28 |
| AJaeger | tristanC: you're welcome! | 16:30 |
| AJaeger | tristanC: do you have time to review https://review.opendev.org/#/c/717817/ after merge-conflict, please? and the stack starting at https://review.opendev.org/#/c/723640 ? | 16:31 |
| * AJaeger is currently spring-cleaning zuul-jobs | 16:31 | |
| AJaeger | mnaser: shall we merge https://review.opendev.org/#/c/724132/ now? | 16:32 |
| *** rlandy|biab is now known as rlandy | 16:33 | |
| *** evrardjp has quit IRC | 16:35 | |
| *** evrardjp has joined #zuul | 16:36 | |
| avass | AJaeger: the policy is only for roles since they can be imported and used in other projects, it's fine to use the default loop_var 'item' in playbooks and tests | 16:58 |
| avass | AJaeger: that's why i only check tasks: https://opendev.org/zuul/zuul-jobs/src/branch/master/.rules/ZuulJobsNamespaceLoopVar.py#L19 | 16:59 |
| openstackgerrit | Merged zuul/zuul-jobs master: tox: Use 'block: ... always: ...' instead of ignore_errors https://review.opendev.org/723640 | 17:01 |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-sphinx: use failed_when: false instead of ignore_errors: true https://review.opendev.org/723642 | 17:01 |
| clarkb | tobiash: are you still building your own zuul images? | 17:03 |
| tobiash | clarkb: yes | 17:04 |
| clarkb | we are seeing memory leaks in opendev and I'm trying to compile a list of changes and realized that in addition to code changes we switched from running on ubuntu xenial hosts to running in zuul built containres which changed python3 versions and libc (and possibly other system libs) | 17:04 |
| corvus | clarkb: i wonder if scheduler memory use corresponds with the leading_errors bug? | 17:05 |
| clarkb | corvus: I'm not familiar with that bug | 17:05 |
| corvus | i was just looking for log entries about the incident that AJaeger mentioned earlier and since you mentioned scheduler memory growing around the same time i thought maybe they're related | 17:05 |
| corvus | clarkb: it's your pet bug | 17:05 |
| corvus | AttributeError: 'NoneType' object has no attribute 'loading_errors' | 17:05 |
| corvus | 2020-05-01 14:42:29,373 DEBUG zuul.Pipeline.zuul.gate: [e: b1ce40a9ea0d47b096e694914c8d9a37] <QueueItem 0x7f99c5cb8b50 for <Change 0x7f99db37d810 zuul/zuul-jobs 717510,9> in gate> is a failing item because ['it has an invalid configuration'] | 17:05 |
| clarkb | oh that one | 17:05 |
| corvus | clarkb: that one^ | 17:05 |
| corvus | clarkb: 2020-05-01 13:09:57,081 is first instance today | 17:06 |
| tobiash | clarkb: at least we had a memleak with py37 a year ago and reverted to py36 and that's what we're running tipl now | 17:08 |
| corvus | clarkb: hrm, not a strong correlation there, but there is a stronger correlation at 14:42:29,373 today | 17:08 |
| tobiash | s/tipl/till | 17:08 |
| clarkb | corvus: and according to cacti thats about when it started increasing the rate of growth | 17:09 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 17:10 |
| mordred | corvus: ^^ hopefully that will work with requires: lines | 17:10 |
| mordred | clarkb, corvus, tobiash: we can also switch the base image to py3.8 at some point if we want to compare/try that | 17:11 |
| clarkb | mordred: or 3.6? | 17:11 |
| mordred | clarkb: well, we'd have to start building 3.6 python-base images - but yes, we could do that too | 17:11 |
| clarkb | note the bug corvus is talking about has been observed on 3.5 and 3.7 now so doesn't seem version specific | 17:12 |
| mordred | totally- it's possible the leak we're seeing is related to the bug | 17:12 |
| webknjaz | corvus: AFAIR the only thing added in the last version is support for UTF-8 filenames in the content-disposition header. Open a ticket if you have a reproducer | 17:12 |
| mordred | but it's also possible there is a bug in 3.7 :) | 17:12 |
| mordred | many things to try | 17:12 |
| clarkb | mordred: yup | 17:13 |
| mordred | clarkb: I think I'd suggest if we start trying different python versions we try 3.8 first - since that would be a roll forward - and if that doesn't have it we can try 3.6 | 17:13 |
| clarkb | mordred: ++ | 17:13 |
| corvus | webknjaz: i don't yet; it's just an indication at this point. thought i'd give you a heads up. | 17:13 |
| mordred | but - I think poking at the loading_errors bug first before we start doing python version roulette | 17:13 |
| corvus | webknjaz: could be a false alarm, but in case someone else reports it, it's a data point | 17:13 |
| clarkb | on the loading_errors bug I dug into ita bit and think I wandered down an invalid path. Might be good to have fresh eyes on it? | 17:14 |
| webknjaz | 👍 | 17:15 |
| clarkb | this mightalso be a case where we need more logging | 17:15 |
| clarkb | we can catch the exception and log some state? | 17:15 |
| openstackgerrit | Merged zuul/zuul-jobs master: ensure-yarn: run ensure-nodejs before https://review.opendev.org/717817 | 17:24 |
| mordred | corvus: 2020-05-01 17:19:44,960 INFO cherrypy.access.140474922906896: ::ffff:172.17.0.1 - - [01/May/2020:17:19:44] "HEAD /v2/upstream/image/manifests/latest HTTP/1.1" 404 767 "" "containerd/1.3.0+unknown" | 17:28 |
| mordred | corvus: still getting 404s | 17:28 |
| mordred | corvus: perhaps since we're testing buildset registry we're not using buildset-registry for zuul-registry itself? | 17:29 |
| corvus | mordred: it should be plumb completely through -- there's actually quite a bit of code to make sure that the test buildset registry uses an image from the real buildset registry | 17:29 |
| corvus | mordred: it's probably worth a quick look at the logs to verify that's working (maybe checking image shas or pull logs or whatever), then we should probably confirm the functionality with a local test | 17:31 |
| mordred | nod | 17:31 |
| corvus | mordred: i'm still combing through loading_error logs, but can pitch in on that in a few mins | 17:31 |
| mordred | corvus: in the meantime, there's a pep8 error - I can push up a fix for it if you want | 17:31 |
| corvus | mordred: ++ | 17:31 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-registry master: Support HEAD on manifests https://review.opendev.org/724859 | 17:32 |
| mordred | clarkb: if you get a sec ^^ | 17:32 |
| mordred | corvus: I'm going to attempt to discover the concept known as "sandwich" | 17:32 |
| mordred | corvus: so maybe let's sync in a bit | 17:32 |
| corvus | mordred: oh wait -- i *may* be thinking of the jobs in zuul-registry repo | 17:32 |
| corvus | mordred: it may be worth a look at that -- perhaps they work with speculative images but the ones in zuul-jobs don't? i don't know, it's all worth double checking. | 17:33 |
| corvus | mordred: (also, maybe we should add a buildx job to zuul-jobs) | 17:33 |
| corvus | grr | 17:33 |
| corvus | mordred: (also, maybe we should add a buildx job to zuul-registry) | 17:33 |
| mordred | corvus: yes to bopth of those things | 17:33 |
| mordred | I'll see what I can see | 17:34 |
| AJaeger | avass: ah, thanks! | 17:39 |
| corvus | clarkb: theory: 2 changes in queue; first is a regular change, second has a dynamic config update; launch merge jobs for both; the second merge job arrives before the first; item.layout on the first change is not primed, but we process the second item anyway. | 17:40 |
| AJaeger | avass: what about roles/ensure-bazel/tasks/install-bazel.yaml ? | 17:40 |
| clarkb | corvus: ooh | 17:41 |
| AJaeger | avass: and roles/configure-mirrors/handlers/main.yaml | 17:41 |
| corvus | clarkb: that should be provable in a unit test. if that's so, that would explain the exception; not sure if the exception can explain the leak | 17:41 |
| avass | AJaeger: I'm not sure why it's not reacting to install-bazel, have to take a look at that later | 17:42 |
| avass | AJaeger: but yeah ,should probably check handlers as well | 17:43 |
| avass | AJaeger: maybe we should have tests for the custom linter rules :) | 17:44 |
| AJaeger | maybe :) | 17:45 |
| avass | AJaeger: oooh, for some reason ansible-lint thinks install-bazel is a playbook | 17:46 |
| avass | I guess that's a bug in ansible-lint | 17:47 |
| zbr | the detection logic is flawed in many places, what can you expect when yaml is so popular? :) | 17:47 |
| zbr | try to help it using ignore patterns | 17:48 |
| avass | but you would think checking if it's a playbook would be easy, just check if "- hosts:" exist or not | 17:48 |
| zbr | it is far more complex than this, it needs to detect: playbooks, tasks, vars, roles,.... | 17:48 |
| avass | ah wait, no it detects is as a task if i give it the role path instead of the yaml file directly | 17:49 |
| avass | or, a tasks list | 17:49 |
| AJaeger | corvus, mordred, we reverted earlier https://review.opendev.org/#/c/723524/ and avass has fixed it in https://review.opendev.org/#/c/724132/ - want to merge again? | 17:49 |
| zbr | until recently we did not had any guidelines for repository layout, but luckly collections helped. | 17:49 |
| zbr | i am considering proposing a change to support only collection layout. | 17:50 |
| corvus | clarkb: i was looking into why this error doesn't happen over and over, and it turns out the situation can correct itself, but zuul has already annotated the item with the config error list, so it's stuck thinking it's an error until the next time it's enqueued (or nnfi changes) | 17:52 |
| corvus | as typified by these 2 lines: 2020-05-01 14:42:32,645 INFO zuul.Pipeline.zuul.gate: [e: b1ce40a9ea0d47b096e694914c8d9a37] Configuration syntax error not related to change context. Error won't be reported. | 17:53 |
| corvus | 2020-05-01 14:42:32,646 DEBUG zuul.Pipeline.zuul.gate: [e: b1ce40a9ea0d47b096e694914c8d9a37] <QueueItem 0x7f99c5cb8b50 for <Change 0x7f99db37d810 zuul/zuul-jobs 717510,9> in gate> is a failing item because ['it has an invalid configuration'] | 17:53 |
| corvus | they're kind of contradictory when you think about it :) | 17:53 |
| avass | AJaeger: good job finding that, I'll try to figure out what's happening with install-bazel :) | 17:53 |
| clarkb | corvus: that first message is what led me astray I think | 17:53 |
| corvus | ("will not be reported"..."has in invalid configuration") | 17:53 |
| clarkb | corvus: it had me thinking the error was elsewhere so we just had to ignore it | 17:53 |
| AJaeger | avass: thanks - but no urgency for that one, it's not a false positive ;) | 17:54 |
| AJaeger | corvus: and re-approving the change it merged. So, somehow Zuul was confused | 17:55 |
| corvus | AJaeger: that re-enqueued it, so it started with a clean slate | 17:55 |
| AJaeger | there was no invalid configuratoin - just Zuul thought with 8 changes enqueued there is one | 17:55 |
| AJaeger | corvus: yes | 17:55 |
| corvus | AJaeger: my unconfirmed theory is that the connection with the number of changes is that more changes means more opportunity for config merge jobs to race, so the original bug is more likely to manifest | 17:56 |
| AJaeger | "race" - ah! | 17:56 |
| corvus | AJaeger: yeah see my msg at 17:40 for an explanation of the (suspected) race sequence | 17:57 |
| AJaeger | I saw it - but now I parse it correctly. thanks | 17:58 |
| corvus | it was shorthand :/ | 17:58 |
| AJaeger | it was me reading too quickly | 17:59 |
| *** bhavikdbavishi has quit IRC | 18:01 | |
| corvus | clarkb: re memory, this can cause mayn more throwaway dynamic layouts to be built (basically, one for each pass through the queue during the window where the second merge job has completed but the first has not). i see several of thos in the 13:00 hour, but only 2 near the 14:42 event. i sort of expect them to be cleaned up correctly by the gc too. i think it'd be a stretch to tie the memory leak to this | 18:01 |
| corvus | at this point. not impossible, but i think if it is, it's more likely to be some weird python gc edge case, and not a zuul logic bug. | 18:01 |
| clarkb | gotcha, we probably want to fix it either way since its a user noticeable bug but unlikely to be our memory fix | 18:02 |
| corvus | yep | 18:02 |
| corvus | i plan on writing a unit test to confirm/repo and then a fix after helping mordred. but before that, i need some exercise; biab. | 18:03 |
| corvus | also, i really apologize for my spelling today, it's atrocious | 18:03 |
| avass | AJaeger: Wow.... "include" isn't a taask | 18:11 |
| avass | AJaeger: so for includes we'd have to use "match" to match individual lines instead of "matchtask" ... | 18:11 |
| AJaeger | Interesting | 18:15 |
| avass | AJaeger: so we could to a simple check for include, include_tasks, and check if they use "item", but not for include_role, since that has a different syntax for some reason | 18:17 |
| AJaeger | one step at a time ;) | 18:24 |
| AJaeger | zuul-jobs reviewers, the stack at https://review.opendev.org/#/c/723643 and https://review.opendev.org/#/c/724132/ would be nice to get some review love, please | 18:34 |
| openstackgerrit | Monty Taylor proposed zuul/zuul master: Build zuul images with python 3.8 https://review.opendev.org/724908 | 18:40 |
| mordred | clarkb: ^^ | 18:40 |
| clarkb | +2 thanks | 18:42 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 18:51 |
| avass | AJaeger: I guess you just have to use undocumented features *shrug* | 18:51 |
| avass | AJaeger: and for some reason it's a 'task' if you match the play, I'll have to check if this is a bug or not | 18:52 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 19:04 |
| AJaeger | zbr: is that something you can help avass with, please? ^ | 19:09 |
| openstackgerrit | Clark Boylan proposed zuul/zuul-jobs master: Fix some item usage where var is zj_* now https://review.opendev.org/724916 | 19:10 |
| clarkb | AJaeger: avass ^ is a bug with loop control var renaming that was discovered by one of our users | 19:10 |
| AJaeger | +2A | 19:12 |
| avass | clarkb: thanks! | 19:16 |
| clarkb | Going to mention it here since I don't have time to fully debug it currently, but zuul-web's SIGUSR2 handling isn't completely correct. It does dump threads an all that as expected but it also stops the service | 19:25 |
| *** lseki has quit IRC | 19:35 | |
| mordred | clarkb: well that's less than awesome | 19:39 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 19:47 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 19:53 |
| *** Goneri has quit IRC | 19:55 | |
| corvus | mordred: need a hand on the registry thing? i'm all yours now | 20:01 |
| mordred | corvus: yes! | 20:01 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 20:09 |
| avass | I feel like a hypocrite using the default loop var in the tests :) | 20:10 |
| AJaeger | avass: something is wrong in stage-output, see https://zuul.opendev.org/t/openstack/build/3a76691e85dd4faeb582d30fab9f19a2/console | 20:10 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 20:10 |
| avass | checking | 20:11 |
| AJaeger | https://zuul.opendev.org/t/openstack/builds?result=POST_FAILURE shows more | 20:11 |
| AJaeger | https://review.opendev.org/#/c/724810/4/roles/stage-output/tasks/main.yaml looks broken - I don't see it yet | 20:14 |
| avass | I can't figure out what could be wrong either | 20:15 |
| AJaeger | The errors point to https://opendev.org/zuul/zuul-jobs/src/branch/master/roles/stage-output/tasks/main.yaml#L34-L46 | 20:15 |
| AJaeger | anybody has an idea? | 20:15 |
| avass | yeah, either that of the sources stat is wrong | 20:16 |
| avass | I can't see anything that could be wrong, it's just a 's/item/zj_source/' pretty much | 20:21 |
| corvus | agree | 20:21 |
| AJaeger | so, how to unbreak? Revert that change for now? | 20:22 |
| mordred | yeah - I looked at the diff from previous and it looks fine | 20:22 |
| avass | should we change it back to 'item' and ignore the linting for that file for now? | 20:22 |
| corvus | i'm trying to repro locally; give me a few mins to do that | 20:22 |
| avass | and see if that fixes it | 20:22 |
| *** iurygregory has joined #zuul | 20:23 | |
| mordred | AJaeger: if we revert - we need to find the other follow changes we've made - but let's wait for corvus | 20:23 |
| AJaeger | mordred: agreed | 20:23 |
| corvus | maybe start preparing that change in case i fail? | 20:23 |
| AJaeger | mordred: https://review.opendev.org/#/c/724916/ is not merged yet | 20:23 |
| * AJaeger removed the +A from 724916 | 20:24 | |
| mordred | AJaeger: nod | 20:24 |
| mordred | AJaeger: yeah - let's prep the revert patch - just in case | 20:24 |
| openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: Revert "Check for loop_control in with_ type loops" https://review.opendev.org/724923 | 20:24 |
| corvus | i have repro'd | 20:25 |
| AJaeger | here's a *complete* revert ^ | 20:25 |
| AJaeger | corvus: cool | 20:25 |
| corvus | fatal: [localhost]: FAILED! => {"msg": "The conditional check 'zj_source.item.value' failed. The error was: error while evaluating conditional (zj_source.item.value): 'dict object' has no attribute 'item'\n\nThe error appears to be in '/tmp/test.yaml': line 14, column 7, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n register: sources\n | 20:25 |
| corvus | - name: Set source and destination for files and folders\n ^ here\n"} | 20:25 |
| clarkb | corvus: and you ran the task "Register sources" at the beginning of the file ? | 20:26 |
| AJaeger | avass just renamed "item.item.value" to zj_source.item.value | 20:26 |
| * AJaeger is puzzled | 20:26 | |
| clarkb | I wonder if reusing the zj_source var from that loop impacts the subsequent task odly | 20:26 |
| corvus | clarkb: ya | 20:26 |
| corvus | clarkb: i just tried that -- renamed the second zj_source, same error | 20:26 |
| corvus | clarkb: good idea though, i thought :) | 20:27 |
| avass | ooh | 20:27 |
| avass | this might be stupid, but could this be it? | 20:28 |
| mordred | there are no stupid things in this context :) | 20:28 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: use zj_source.zj_source https://review.opendev.org/724925 | 20:28 |
| clarkb | if that is the fix then I think my change for the pdf thing needs updating too | 20:28 |
| AJaeger | avass: interesting theory | 20:28 |
| clarkb | it still has a foo.time | 20:29 |
| clarkb | *foo.item | 20:29 |
| corvus | that runs without error | 20:29 |
| mordred | WOW | 20:29 |
| AJaeger | then we should also double check avass' change whether there are other problems | 20:29 |
| clarkb | nice | 20:29 |
| avass | wow | 20:29 |
| AJaeger | Wow indeed | 20:29 |
| mordred | I *never* would have imagined that - but now that it's there it makes sense | 20:29 |
| corvus | yeah. mind boggled. | 20:29 |
| clarkb | someone should update my pdf change, feel free to do that | 20:29 |
| clarkb | (I need to pop out) | 20:29 |
| mordred | I mean "makes sense" | 20:29 |
| corvus | mordred: i'll have what you're having :) | 20:29 |
| mordred | I'm guessing we need to do a quick search for use of with_dict | 20:30 |
| corvus | mordred: that was with_items | 20:30 |
| AJaeger | can you take care of this, please? I need to call it a day now... Feel free to abandon my revert | 20:30 |
| mordred | it was the earlier with_dict along with loop_var that caused the inner item to be zj_source I thnjk | 20:31 |
| mordred | https://review.opendev.org/#/c/724810/4/roles/stage-output/tasks/main.yaml lines 1-8 | 20:31 |
| AJaeger | Ah! | 20:32 |
| mordred | corvus: could test that theory by changing that loop var to something else and then seeing if the second item should match it | 20:32 |
| mordred | but I think it's the register along with that with_dict that is the source of the inner .item | 20:32 |
| corvus | mordred: i did a debug on sources and it says "zj_source": { | 20:32 |
| AJaeger | So, https://review.opendev.org/#/c/724916/1/roles/fetch-sphinx-tarball/tasks/pdf.yaml looks fine, doesn't it? | 20:32 |
| corvus | "zj_source": {"key": "/var/log/syslog", "value": "logs_txt"} | 20:33 |
| mordred | yeah. that's because of the loop_var in the thing that created it | 20:33 |
| corvus | mordred: yeah, "makes sense" | 20:33 |
| mordred | corvus: "exactly" | 20:33 |
| SpamapS | isn't everything supposed to be with: now anyway? | 20:33 |
| mordred | SpamapS: well - there are about 7 different "supposed to"s | 20:33 |
| mordred | :) | 20:33 |
| mordred | corvus: maybe we shoudl change the first one to be a different name just for clarity | 20:33 |
| corvus | SpamapS: maybe you're thinking of everything should be "loop" but they changed their minds on that and with_* is fine again | 20:34 |
| mordred | yeah | 20:34 |
| AJaeger | so, can we +A https://review.opendev.org/724916 ? | 20:34 |
| mordred | because it turns out loop+lookup is bonghits bad | 20:34 |
| SpamapS | oh right loop: | 20:34 |
| avass | SpamapS: yeah loop is preferred | 20:34 |
| * SpamapS may not have looked at any ansible for about 2 months. ;) | 20:34 | |
| mordred | avass: they changed their mind | 20:34 |
| corvus | avass: i would not say that at this point :) | 20:34 |
| avass | ah :) | 20:34 |
| mordred | avass: you want to update your patch? | 20:35 |
| mordred | or I can if you're busy | 20:35 |
| avass | you can do it :) | 20:35 |
| corvus | mordred: what needs updating? | 20:35 |
| avass | changing the first zj_source? | 20:35 |
| corvus | https://review.opendev.org/724925 looks fine to me | 20:35 |
| mordred | yeah - so that it's clear to readers of the file what's coming from where | 20:35 |
| corvus | ok | 20:35 |
| corvus | i'll wait for an update | 20:35 |
| tristanC | corvus: we got a first successfull image-update >> image-test >> image-promote pipeline execution. From such image description: https://softwarefactory-project.io/cgit/zuul-images-jobs/tree/example/openstacksdk.dhall , we created this zuul configuration : https://softwarefactory-project.io/cgit/zuul-images-jobs/tree/example/pipeline.yaml | 20:35 |
| AJaeger | thanks everybody! | 20:36 |
| mordred | avass: also - good idea on that :) | 20:36 |
| * AJaeger waves good night and wishes everybody a nice weekend | 20:36 | |
| mordred | AJaeger: have a good weekend! | 20:36 |
| avass | mordred: it was a gut feeling :) | 20:37 |
| avass | AJaeger: bye! | 20:38 |
| tristanC | corvus: about using the zuul-registry, we proposed those improvements: https://review.opendev.org/#/q/topic:use-buildset-registry-idempotency | 20:38 |
| corvus | tristanC: great! i've got several fires that need putting out right now, but next week when things settle down, let's work on integrating the registry roles you need with the rest of the system to maximise reuse and interoperability. also, i have ideas about how this could be done with gating if you want to have a design session to talk about it. | 20:38 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 20:38 |
| corvus | tristanC: cool, i expect those should be straightforward | 20:39 |
| corvus | avass: you're updating 724925 right? | 20:39 |
| corvus | oh, no mordred is updating it | 20:40 |
| corvus | sorry i missed that | 20:40 |
| tristanC | corvus: then looking at `upload-docker-image` and `promote-docker-image`, it seems very specific to docker and dockerhub, so i'm not sure how we could add podman and other registry support | 20:41 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Distinguish zj_source and zj_result and item https://review.opendev.org/724925 | 20:41 |
| mordred | corvus, avass, clarkb: ^^ | 20:41 |
| corvus | tristanC: the general rule is that if it says 'docker' in the name it's intentionally docker-based and we probably want a new job/role. the ones that don't say docker should work for anything. | 20:41 |
| corvus | tristanC: mostly i say that so you know what the expectation was when the role was originally written | 20:42 |
| mordred | yeah - so use-buildset-registry should handle everything. but we need an upload-container-image that's not docker specific to go along with build-container-image | 20:42 |
| mordred | yeah? | 20:42 |
| corvus | mordred: yep: https://zuul-ci.org/docs/zuul-jobs/container-roles.html has a note (cc tristanC ) | 20:43 |
| tristanC | mordred: the main issue we have with use-buildset-registry is that the registry.cert needs to be the issuer ca.crt, not the registry tls.crt | 20:43 |
| tristanC | or at least on fedora after running update-ca-trust, that didn't work with the tls.crt content | 20:43 |
| corvus | tristanC: then we should identify and fix that bug | 20:43 |
| corvus | anyway, i really need to get back to firefighting | 20:44 |
| corvus | 925 is approved | 20:44 |
| mordred | woot | 20:44 |
| tristanC | corvus: mordred: i think our initial PoC story is now completed, and i'd be happy to talk more about how to design such things | 20:45 |
| openstackgerrit | Merged zuul/zuul-jobs master: Fix some item usage where var is zj_* now https://review.opendev.org/724916 | 20:47 |
| avass | we need to fix that ^ | 20:48 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Use zj_sphinx_pdf instead of item https://review.opendev.org/724927 | 20:50 |
| avass | corvus, mordred, tristanC: ^ | 20:50 |
| openstackgerrit | Merged zuul/zuul-jobs master: Distinguish zj_source and zj_result and item https://review.opendev.org/724925 | 20:52 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 20:54 |
| *** Goneri has joined #zuul | 21:00 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 21:01 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 21:02 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 21:04 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Use zj_sphinx_pdf instead of item https://review.opendev.org/724927 | 21:06 |
| avass | mordred, corvus, tristanC: ^ had to fix linting | 21:06 |
| *** guillaumec has quit IRC | 21:11 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 21:11 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: Only run tests for ensure-bazel when it is updated https://review.opendev.org/724933 | 21:15 |
| openstackgerrit | Merged zuul/zuul master: Build zuul images with python 3.8 https://review.opendev.org/724908 | 21:17 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 21:21 |
| *** cmurphy has joined #zuul | 21:21 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 21:26 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: ansible-lint: use matchplay instead of matchtask https://review.opendev.org/724910 | 22:03 |
| mordred | clarkb: could I get you to +A https://review.opendev.org/#/c/722062/1 real quick? | 22:06 |
| *** klindgren has quit IRC | 22:07 | |
| *** mhu has quit IRC | 22:07 | |
| *** gouthamr has quit IRC | 22:07 | |
| *** evrardjp has quit IRC | 22:07 | |
| *** irclogbot_0 has quit IRC | 22:07 | |
| *** kgz has quit IRC | 22:07 | |
| *** pots has quit IRC | 22:07 | |
| *** toabctl has quit IRC | 22:07 | |
| *** toabctl has joined #zuul | 22:07 | |
| *** evrardjp has joined #zuul | 22:10 | |
| *** pots has joined #zuul | 22:10 | |
| *** klindgren has joined #zuul | 22:11 | |
| avass | mordred, clarkb, tristanC, corvus: https://review.opendev.org/#/c/724927/ | 22:11 |
| avass | zuul-jobs-maint: ^ | 22:11 |
| *** gouthamr has joined #zuul | 22:11 | |
| *** irclogbot_2 has joined #zuul | 22:12 | |
| openstackgerrit | James E. Blair proposed zuul/zuul-registry master: Add a buildkit functional test https://review.opendev.org/724942 | 22:19 |
| corvus | mordred: ^ that's a small step toward improving testing of zuul-registry | 22:19 |
| corvus | that's just using it as a standard registry; the job normally also tests using it as a buildset registry (ie, with a full mirror config) but i figure we can add that once this works | 22:20 |
| corvus | clarkb: can you +3 https://review.opendev.org/722062 ? | 22:21 |
| corvus | the HEAD change was stacked on that | 22:21 |
| mordred | corvus: thanks for the help! | 22:25 |
| mordred | I thnik we might actually be almost actually ready to almost be able to do this | 22:25 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Write a buildkitd config file pointing to buildset registry https://review.opendev.org/724757 | 22:32 |
| mordred | corvus: if that's green (once z-r change lands) - I'm going to squash all three into one change | 22:32 |
| clarkb | corvus: looking | 22:45 |
| *** rlandy has quit IRC | 22:53 | |
| openstackgerrit | Merged zuul/zuul-registry master: Support listing tags https://review.opendev.org/722062 | 23:01 |
| openstackgerrit | Merged zuul/zuul-registry master: Support HEAD on manifests https://review.opendev.org/724859 | 23:01 |
| openstackgerrit | Merged zuul/zuul-jobs master: Use zj_sphinx_pdf instead of item https://review.opendev.org/724927 | 23:05 |
| openstackgerrit | Clark Boylan proposed zuul/zuul master: Don't exit zuul-web when signals are received https://review.opendev.org/724946 | 23:17 |
| clarkb | mordred: corvus ^ I think that may fix the sigusr2 + zuul-web thing. That is mostly a reminder to myself for next week whenI can test it a bit more | 23:17 |
| clarkb | cloudnull: just noticed your tripleo container proposal and wanted to mention the Zuul docker image gating stuff thats gone on. You might want to think about incorporating that if you haven't already | 23:35 |
| clarkb | cloudnull: its somewhat orthogonal to your goals but if you are shufflign everythinganyway it might be a good time for it | 23:36 |
| *** tosky has quit IRC | 23:45 | |
| mordred | cloudnull: ++ | 23:54 |
| mordred | cloudnull: I'd be happy to help you sort out what that looks like- we've got some extra sexy container tools for you these days | 23:55 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!