| openstackgerrit | Merged zuul/zuul-jobs master: Fix the checking helm_values_file definition https://review.opendev.org/722516 | 00:40 |
|---|---|---|
| *** swest has quit IRC | 01:09 | |
| ianw | sigh, somehow the sibling container build now looks broken ... https://zuul.opendev.org/t/openstack/build/09f147450ed64da984d6f1e0a99def6c/log/job-output.txt#736 | 01:19 |
| ianw | i'm thinking it has to do with 721514 but not sure how | 01:20 |
| openstackgerrit | Ian Wienand proposed zuul/nodepool master: Add nodepool as required project to siblings image build https://review.opendev.org/722556 | 01:24 |
| *** swest has joined #zuul | 01:24 | |
| *** rlandy has quit IRC | 02:12 | |
| openstackgerrit | Ian Wienand proposed zuul/zuul master: Add intermediate flag for jobs https://review.opendev.org/722200 | 02:52 |
| *** bhavikdbavishi has joined #zuul | 03:02 | |
| *** bhavikdbavishi1 has joined #zuul | 03:24 | |
| *** bhavikdbavishi has quit IRC | 03:26 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 03:26 | |
| *** bhavikdbavishi has quit IRC | 04:09 | |
| *** ysandeep|away is now known as ysandeep | 04:10 | |
| *** bhavikdbavishi has joined #zuul | 04:22 | |
| *** evrardjp has quit IRC | 04:35 | |
| *** evrardjp has joined #zuul | 04:35 | |
| *** ysandeep is now known as ysandeep|brb | 04:58 | |
| *** ysandeep|brb is now known as ysandeep | 05:29 | |
| *** dpawlik has joined #zuul | 06:03 | |
| *** bhavikdbavishi has quit IRC | 06:27 | |
| *** yolanda has joined #zuul | 06:28 | |
| *** bhavikdbavishi has joined #zuul | 06:37 | |
| *** jcapitao has joined #zuul | 07:09 | |
| *** rpittau|afk is now known as rpittau | 07:17 | |
| *** jpena|off is now known as jpena | 07:20 | |
| *** tosky has joined #zuul | 07:33 | |
| *** bhavikdbavishi has quit IRC | 07:51 | |
| *** ysandeep is now known as ysandeep|lunch | 07:57 | |
| *** bhavikdbavishi has joined #zuul | 08:06 | |
| *** ysandeep|lunch is now known as ysandeep | 09:08 | |
| openstackgerrit | Albin Vass proposed zuul/zuul master: WIP: Add commit id to Change for mqtt reporter https://review.opendev.org/722478 | 09:09 |
| tobiash | zuul-maint: looks like the line mapping might be off by one: https://review.opendev.org/#/c/722478/1/zuul/driver/mqtt/mqttreporter.py@44 | 09:13 |
| tobiash | zuul-maint: never mind, the linter itself reported exactly this line, so it's correct | 09:15 |
| openstackgerrit | Fabien Boucher proposed zuul/zuul-jobs master: Remove hardcoded use of /home/zuul https://review.opendev.org/722649 | 09:33 |
| openstackgerrit | Andy Ladjadj proposed zuul/zuul master: convert utc format to local timezone with format YYYY-MM-DD HH:mm:ss https://review.opendev.org/722653 | 09:59 |
| openstackgerrit | Andy Ladjadj proposed zuul/zuul master: [web] convert utc format to local timezone with format YYYY-MM-DD HH:mm:ss https://review.opendev.org/722653 | 09:59 |
| openstackgerrit | Andy Ladjadj proposed zuul/zuul master: [web] convert utc format to local timezone with format YYYY-MM-DD HH:mm:ss https://review.opendev.org/722653 | 10:02 |
| openstackgerrit | Andy Ladjadj proposed zuul/zuul master: [web] convert utc format to local timezone with format YYYY-MM-DD HH:mm:ss https://review.opendev.org/722653 | 10:05 |
| *** yoctozepto has quit IRC | 10:09 | |
| *** rpittau is now known as rpittau|bbl | 10:12 | |
| tobias-urdin | in a non-modified centos 8 cloud image neither py2 or py3 is installed by default, is there any way to run ansible tasks before zuul launches the ansible "setup" part (i.e runAnsibleSetup method)? doesn't look like it | 10:16 |
| tobias-urdin | meaning i need to have a customized image with preinstalled python, i cannot inject a ansible "raw" task somewhere | 10:16 |
| tobiash | tobias-urdin: that's correct atm | 10:19 |
| tobiash | for this use case it would be needed to skip the setup part in zuul | 10:20 |
| tobias-urdin | there is no way to skip setup as of now? | 10:20 |
| tobiash | tobias-urdin: no, someone would need to implement this | 10:21 |
| tobias-urdin | thanks, then i'll work around the problem for now :) | 10:21 |
| tobiash | I guess it would be a 'skip-ansible-setup' flag on the job | 10:21 |
| *** bhavikdbavishi has quit IRC | 10:23 | |
| openstackgerrit | Andy Ladjadj proposed zuul/zuul master: [web] convert utc format to local timezone with format YYYY-MM-DD HH:mm:ss https://review.opendev.org/722653 | 10:44 |
| *** jcapitao is now known as jcapitao_lunch | 10:59 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 11:18 |
| reiterative | I'm trying to run up Zuul Quickstart for local testing, but version on master fails in executor with a FileNotFoundError for /var/run/zuul/executor.pid | 11:25 |
| reiterative | Has anyone else had problems running quickstart? I'm running Debian stretch. | 11:26 |
| openstackgerrit | Paul Albertella proposed zuul/zuul-jobs master: Add Bazel build and ensure roles https://review.opendev.org/693513 | 11:27 |
| *** jpena is now known as jpena|lunch | 11:28 | |
| *** sassyn has quit IRC | 11:34 | |
| *** bhavikdbavishi has joined #zuul | 11:36 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 11:41 |
| avass | I don't really like that the tox roles default to using venv as a target test environment | 11:45 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 11:53 |
| *** sgw has quit IRC | 12:08 | |
| *** rpittau|bbl is now known as rpittau | 12:13 | |
| *** jcapitao_lunch is now known as jcapitao | 12:20 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 12:27 |
| openstackgerrit | Paul Albertella proposed zuul/zuul-jobs master: Add Bazel build and ensure roles https://review.opendev.org/693513 | 12:31 |
| *** rlandy has joined #zuul | 12:31 | |
| *** rlandy is now known as rlandy|mtg | 12:35 | |
| *** jpena|lunch is now known as jpena | 12:38 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 12:42 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 12:48 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 12:53 |
| *** Goneri has joined #zuul | 13:08 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: cabal-test: add initial haskell job https://review.opendev.org/721735 | 13:12 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: hlint: add haskell source code suggestions job https://review.opendev.org/722309 | 13:12 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 13:24 |
| *** rlandy|mtg is now known as rlandy | 13:34 | |
| *** ysandeep is now known as ysandeep|afk | 13:50 | |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Support multi-arch image builds with docker buildx https://review.opendev.org/722339 | 13:51 |
| corvus | reiterative: the quickstart runs in docker, so the host os shouldn't matter (much). and we run that on every change to zuul, so it should be very well tested. you're following these instructions? https://zuul-ci.org/docs/zuul/tutorials/quick-start.html | 13:55 |
| *** bhavikdbavishi has quit IRC | 13:56 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 13:58 |
| fungi | avass: can you elaborate on the problem you're having with venv in the tox roles? what would you prefer? | 14:02 |
| fungi | i'm confused by your comment about "target test environment" since tox generally uses virtualenv not venv (though the latest releases of virtualenv are supposed to be almost identical to stdlib venv environments) | 14:03 |
| fungi | there is a tox-venv plugin which will tell tox to use stdlib venv instead of virtualenv, but i didn't think we were installing that in the role | 14:04 |
| corvus | fungi: i think he means the "tox environment" named "venv"; see https://review.opendev.org/721796 | 14:04 |
| fungi | ohhh | 14:04 |
| avass | fungi, corvus: yeah, instead of checking if the variable is undefined I have to set it to an empty string and check if tox_envlist is empty | 14:05 |
| fungi | i see, the problem is that the tox_envlist rolevar has a default of "venv" so launches the [testenv:venv] defined in the tox.ini if you don't pass any tox_envlist value | 14:06 |
| avass | yeah | 14:06 |
| fungi | makes sense, i agree that's probably an old openstackism leaking into our defaults | 14:07 |
| avass | it's a bit counter intuitive that the tox job runs venv by default instead of working the same way as if you would run 'tox' on the commandline | 14:07 |
| fungi | for openstack jobs since ages we used a fallback generic testenv:venv to run arbitrary commands in a virtualenv | 14:07 |
| avass | fungi: that's what I was guessing, so I don't want to break it :) | 14:08 |
| fungi | and i agree, that's somewhat inflexible and a questionable default choice | 14:08 |
| openstackgerrit | Merged zuul/nodepool master: Add nodepool as required project to siblings image build https://review.opendev.org/722556 | 14:13 |
| avass | and checking for an empty string is hard for some reason.. :( | 14:18 |
| avass | I really don't understand why that fails | 14:22 |
| *** zxiiro has joined #zuul | 14:31 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 14:34 |
| openstackgerrit | Merged zuul/zuul-jobs master: Remove hardcoded use of /home/zuul https://review.opendev.org/722649 | 14:34 |
| openstackgerrit | Merged zuul/zuul-jobs master: cabal-test: add initial haskell job https://review.opendev.org/721735 | 14:35 |
| *** sgw has joined #zuul | 14:43 | |
| *** wxy has quit IRC | 14:45 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 14:46 |
| *** harrymichal has joined #zuul | 14:47 | |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Support multi-arch image builds with docker buildx https://review.opendev.org/722339 | 14:48 |
| openstackgerrit | Monty Taylor proposed zuul/zuul-jobs master: Support multi-arch image builds with docker buildx https://review.opendev.org/722339 | 14:50 |
| openstackgerrit | Merged zuul/zuul-operator master: Increase scheduler wait timeout and improve logs collection https://review.opendev.org/718162 | 14:51 |
| openstackgerrit | Merged zuul/zuul-operator master: Add schema validation error message https://review.opendev.org/718999 | 14:51 |
| reiterative | corvus Yes, I'm following the instructions and had it working with an older iteration. Just pruned my Docker containers and cloned a fresh copy to make sure - same problem. | 14:52 |
| reiterative | Here's the log output: http://paste.openstack.org/show/792670/ | 14:52 |
| openstackgerrit | Merged zuul/zuul-operator master: Add a zuul-ensure-database-passwords role https://review.opendev.org/717880 | 14:54 |
| corvus | reiterative: do you have any modifications to the docker-compose file? bind-mounting any extra volumes in, or user id settings? | 14:56 |
| tristanC | corvus: reiterative: shouldn't the container entrypoint be using `-f` command line argument, resulting in the run file to *not* be created? | 14:56 |
| corvus | tristanC: i believe that is correct -- it should use "-f" and according to the code, that should bypass the lockfile | 14:58 |
| avass | corvus, tristanC, reiterative: I get the same error running docker-compose up -d | 14:58 |
| avass | on ubuntu-bionic | 14:58 |
| tristanC | corvus: hmm, looking at the zuul.cmd.ZuulDaemonApp class, the main function seems to use pid_file_module.TimeoutPIDLockFile before the nodaemon check | 15:00 |
| corvus | tristanC: but that shouldn't aquire it, and reiterative's traceback included the context manager | 15:00 |
| openstackgerrit | Merged zuul/zuul-operator master: Add zuul-registry deployment https://review.opendev.org/710650 | 15:01 |
| corvus | i have run 'docker-compose down -v' and am running 'docker-compose pull' now. | 15:02 |
| reiterative | tristanC Interesting. When I first tried running the latest version, I got a different error stating that -f was not a valid option for zuul-scheduler... | 15:03 |
| reiterative | I wonder if there are some old Docker containers hanging around on my system to confuse things? | 15:04 |
| corvus | reiterative: you can run 'docker-compose pull' to update them all | 15:05 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 15:05 |
| reiterative | corvus Just trying that now (I'm still quite new to Docker) | 15:06 |
| *** dpawlik has quit IRC | 15:10 | |
| corvus | reiterative: i can't reproduce with the following steps: update to latest zuul checkout (to get the latest docker-compose file) -- sha should be ccddc37fa04dc66ab29b9bc97cc63b0665e64430; then 'docker-compose down -v' (to remove old containers and volumes); 'docker-compose pull' to pull updated images; then 'docker-compose up -d' | 15:11 |
| corvus | avass: ^ | 15:11 |
| corvus | so maybe try that process to make sure everything is in sync | 15:11 |
| avass | corvus: it worked for me after pulling to make sure I had the latest containers | 15:11 |
| avass | must have had something old on this machine | 15:11 |
| avass | *images | 15:12 |
| reiterative | corvus tristanC avass Thanks very much for your help - that all seems to be working now! | 15:16 |
| mordred | *phew* | 15:23 |
| avass | am I missing something stupid here or is this behaving strange? https://review.opendev.org/#/c/721796/13 | 15:26 |
| avass | I really can't figure out why it behaves like that | 15:26 |
| corvus | avass: i'm guessing you need something like != "" / | 15:34 |
| corvus | ? | 15:34 |
| avass | coruvs: I believe I tried that earlier, but got an error with something like "should not compare to an empty string" | 15:37 |
| avass | coruvs: "if tox_envlist" should return false for tox_envlist = "" as well | 15:38 |
| avass | corvus: let's see what happens anyway | 15:39 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 15:40 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 15:41 |
| *** harrymichal has quit IRC | 15:42 | |
| corvus | avass: for these things, sometimes i'll set up a simple playbook for testing, just using "vars:" and "debug" tasks | 15:42 |
| avass | corvus: I should probably do that, was messing around with a jinja2 parser but I guess the versions don't match or something | 15:44 |
| *** jcapitao has quit IRC | 15:47 | |
| corvus | avass: something like http://paste.openstack.org/show/792675/ | 15:48 |
| corvus | that repros the error, and this seems to work? when: tox_envlist != '' | 15:48 |
| avass | corvus: yep, just tried that and it works | 15:48 |
| *** ysandeep|afk is now known as ysandeep|away | 15:49 | |
| avass | what ansible version were you using? | 15:49 |
| avass | corvus: I'm confused to why the first debug works and the second one doesn't: http://paste.openstack.org/show/792678/ | 15:55 |
| corvus | avass: 2.8.8 | 15:55 |
| corvus | avass: maybe the second form it just tries to evaluate the substitution, where as in the first form it implicitly treats it as a boolean? | 15:56 |
| avass | corvus: aah, yeah that could be it | 15:57 |
| avass | corvus: also seing a deprecation warning for that | 15:57 |
| avass | corvus: ah yeah, I'm getting a linting error for comparing to an empty string, guess I'll just disable that | 16:04 |
| corvus | avass: could you just use the "when: tox_envlist is defined and tox_envlist" form? | 16:05 |
| avass | corvus: I could, the list form feels easier to read, but that could just be me | 16:07 |
| corvus | avass: i agree, but it seems better than a lint ignore :) | 16:07 |
| avass | corvus: tried using the recommended |bool filter but I ended up with a wrong 'false' value instead: https://github.com/ansible/ansible/issues/25893 | 16:09 |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 16:10 |
| avass | let's if this works | 16:10 |
| *** harrymichal has joined #zuul | 16:10 | |
| *** bhavikdbavishi has joined #zuul | 16:14 | |
| avass | I like how the solution is to use "some_var|default(False, boolean=True)|bool" instead of comparing to an empty string because that's evil for some reason | 16:14 |
| *** bhavikdbavishi1 has joined #zuul | 16:17 | |
| *** bhavikdbavishi has quit IRC | 16:18 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 16:18 | |
| openstackgerrit | Albin Vass proposed zuul/zuul-jobs master: tox: allow running default envlist in tox https://review.opendev.org/721796 | 16:23 |
| avass | corvus: that format also didn't work :) | 16:30 |
| avass | corvus: it doesn't error but it seems like it evaluates to false instead | 16:31 |
| avass | corvus: wait, nevermind I was looking at an old report | 16:33 |
| *** evrardjp has quit IRC | 16:35 | |
| *** evrardjp has joined #zuul | 16:35 | |
| *** rpittau is now known as rpittau|afk | 16:41 | |
| *** harrymichal has quit IRC | 16:42 | |
| *** harrymichal has joined #zuul | 16:43 | |
| *** jpena is now known as jpena|off | 16:48 | |
| avass | corvus: looks like it works, thanks! | 17:03 |
| *** bhavikdbavishi has quit IRC | 17:09 | |
| *** rlandy is now known as rlandy|brb | 17:14 | |
| *** logan- has quit IRC | 17:30 | |
| *** logan- has joined #zuul | 17:42 | |
| *** mhu has quit IRC | 18:05 | |
| *** tjgresha__ has quit IRC | 18:06 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Add serial pipeline manager https://review.opendev.org/722981 | 18:10 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Make fake test Gerrit merger more realistic https://review.opendev.org/722982 | 18:10 |
| corvus | mordred: ^ i think maybe that's the way to go. | 18:10 |
| *** yolanda has quit IRC | 18:11 | |
| corvus | i expect the first change to pass tests, and i think it's an adequate simulation with the caveats. the second change is a follup which is going to affect almost every unit test, so i have no idea if it's feasible. | 18:11 |
| corvus | i consider it an optional improvement | 18:11 |
| *** rlandy|brb is now known as rlandy | 18:12 | |
| corvus | tristanC, tobiash: ^ feedback on a new serial pipeline manager feature welcome | 18:19 |
| tristanC | corvus: nice, i'll have a look shortly | 18:21 |
| mordred | corvus: "one queue per project" | 18:35 |
| mordred | corvus: how does that interact with jobs triggered by both opendev/system-config and openstack/project-config | 18:36 |
| tristanC | corvus: at a first look serial looks like a combinaison of check/gate/promote | 18:39 |
| avass | mordred: can't you still configure multiple projects to share a queue for a pipelie? I would expect that to be the default behaviour | 18:44 |
| avass | mordred: oh, maybe that's specific to the dependent pipeline | 18:45 |
| mordred | avass: it's specific to this new pipeline manager proposal | 18:53 |
| corvus | mordred: that's a very good question -- but for that specific example, opendev/system-config and openstack/project-config are going to be in different tenants. but we could address the issue for opendev/system-config and opendev/project-config, which will be in the same tenant. | 18:55 |
| corvus | mordred: in that case, maybe we need a queue attachment like dependent | 18:55 |
| avass | mordred: yeah, had to look up how that works | 18:56 |
| mordred | corvus: yeah - opendev/system-config and opendev/project-config is definitely the issue | 18:56 |
| corvus | if we do that, we need to base the work on https://review.opendev.org/720182 | 18:56 |
| dmsimard | mordred: o/ re: https://opendev.org/zuul/zuul/src/branch/master/zuul/executor/server.py#L2019-L2027 | 18:56 |
| corvus | mordred: or, if not base, at least make sure it's forward-compatible with that | 18:56 |
| dmsimard | in avass' https://review.opendev.org/#/c/717260/, I thought we could take that opportunity to make it generic | 18:57 |
| dmsimard | but it's for configuring callback_plugins vs callback_whitelist | 18:58 |
| mordred | corvus: yeah. | 18:58 |
| dmsimard | can an operator set custom ansible action plugin and library paths ? it should work the same | 18:59 |
| corvus | dmsimard: zuul executors maintain their own virtualenvs for each ansible version, and i think any callback needs to be installed in those | 19:00 |
| corvus | so i think we can handle this by... | 19:00 |
| avass | corvus: right | 19:01 |
| corvus | dmsimard, avass: making this generic: https://opendev.org/zuul/zuul/src/branch/master/zuul/lib/ansible.py#L240 | 19:01 |
| corvus | so the process becomes: 1) tell zuul-manage-ansible to install the callback plugin; 2) add a new config option that says to add the path for these additional callbacks plugins. | 19:02 |
| corvus | for example for ara: | 19:03 |
| corvus | 1) ANSIBLE_EXTRA_PACKAGES=ara zuul-manage-ansible | 19:03 |
| corvus | 2) ansible_extra_callback_modules=ara | 19:03 |
| corvus | is that enough to get us to running "python -m ara.setup.callback_plugins" to get the path? | 19:04 |
| avass | corvus: seems like we could make the 'ara.setup.callback_plugins' configurable | 19:04 |
| corvus | and does that work for any module? | 19:04 |
| corvus | or is that ara-specific? | 19:04 |
| avass | corvus: I'll have to do some research on that :) | 19:05 |
| dmsimard | ara.setup.callback_plugins is a shortcut to printing the path where ara's callback_plugins directory is | 19:05 |
| dmsimard | because it can change depending on where it's installed and hard coding is meh | 19:05 |
| corvus | right -- that's what we need to do to make this generic for any plugin | 19:05 |
| corvus | to put this another way: it's not possible for the user to actually specify the path to the callback plugins. so we need a way to automate the executor finding out the right callback_plugins path for any plugin the user names | 19:06 |
| avass | corvus: how about we add an ansible_extra_packages in zuul.conf, that way pywinrm can be installed as well | 19:06 |
| corvus | avass: that feature already exists, but it's an env variable: https://zuul-ci.org/docs/zuul/howtos/installation.html#ansible | 19:07 |
| avass | corvus: guess I missed that | 19:08 |
| corvus | also, does collections help out? (ie, does it standardize anything about callback plugin installation paths?) | 19:08 |
| corvus | mordred: ^? | 19:08 |
| corvus | sorry, i have to grab lunch now, biab. | 19:09 |
| avass | corvus: thanks, have to make a quick update to our scripts then :) | 19:09 |
| dmsimard | collections might be helpful because they are installed in a standardized location | 19:10 |
| fungi | at least once collections get standardized ;) | 19:10 |
| dmsimard | eh :D | 19:10 |
| * dmsimard looks at docs | 19:11 | |
| dmsimard | https://docs.ansible.com/ansible/latest/reference_appendices/config.html#collections-paths | 19:11 |
| mordred | corvus: yes | 19:11 |
| dmsimard | so the callback could ship inside a collection and have a common path (regardless of the path of the python interpreter) | 19:11 |
| dmsimard | but collections do not solve the problem of python dependencies that might be required | 19:11 |
| mordred | nope | 19:12 |
| mordred | but - my understanding is that you can refer to plugins by module path rather than fs path | 19:12 |
| mordred | so once you install a collection, the config file could have callback_plugin = ara.collection.callback_plugin | 19:12 |
| dmsimard | oh yeah ? | 19:12 |
| dmsimard | that could be convenient | 19:13 |
| mordred | however - I'm not sure there's a good story for per-ansible-installation collection installs - as much as there is a "install the collection on the system" | 19:13 |
| mordred | so we still might have things to figure out | 19:13 |
| dmsimard | it'd be inside the bubblewrap ? | 19:14 |
| dmsimard | I mean, with other prepared roles and stuff | 19:14 |
| mordred | yeah - we'll have to put it inside the bubblewrap - but if we do that we might have to invent a place to install/stage per-ansible install collections so we can bind-mount them into bubblewrap into the right location :) | 19:14 |
| tristanC | corvus: left a comment/question on the serial pipeline, it lgtm | 19:15 |
| avass | mordred: couldn't you just configure different collections_paths for each ansible installation? | 19:20 |
| mordred | avass: oh - duh | 19:20 |
| mordred | yes | 19:20 |
| avass | :) | 19:20 |
| * mordred goes back to staring blankly at the wall | 19:20 | |
| *** harrymichal has quit IRC | 19:26 | |
| *** harrymichal has joined #zuul | 19:26 | |
| fungi | it's almost like staring blankly at source code | 19:29 |
| avass | I get stuck doing that a bit too often | 19:29 |
| avass | I'll be back in a bit, gotta make some falafel :) | 19:38 |
| *** mgagne has joined #zuul | 19:49 | |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: hlint: add haskell source code suggestions job https://review.opendev.org/722309 | 19:52 |
| openstackgerrit | Tristan Cacqueray proposed zuul/zuul-jobs master: hlint: add haskell source code suggestions job https://review.opendev.org/722309 | 19:58 |
| *** toabctl has quit IRC | 20:00 | |
| corvus | mordred, dmsimard, avass: so maybe the "make the ara install generic for all third-party callback plugins" step should be based on collections? | 20:02 |
| corvus | that way we don't have to figure out anything for the current system | 20:03 |
| tristanC | not sure what is going on with https://review.opendev.org/722309 , it says `Patch in Merge Conflict` but it is on master without git or depends-on related change. Are opendev merger healthy? | 20:12 |
| tristanC | sorry for the noise, this is already discussed in #opendev :) | 20:15 |
| -openstackstatus- NOTICE: The Zuul project gating service is reporting new patches in merge conflict erroneously due to a configuration error, fix in progress | 20:22 | |
| avass | corvus: sounds reasonable, I guess we could make it like ANSIBLE_EXTRA_PACKAGES and enable configuring ansible-galaxy servers | 20:23 |
| *** michael-beaver has joined #zuul | 20:59 | |
| *** rf0lc0 has quit IRC | 21:36 | |
| *** cdearborn has joined #zuul | 21:39 | |
| *** rlandy has quit IRC | 22:48 | |
| *** rf0lc0 has joined #zuul | 22:55 | |
| -openstackstatus- NOTICE: the This Zuul outage was taken as an opportunity to perform an impromptu maintenance for changing our service deployment model; any merge failures received from Zuul between 19:40 and 20:20 UTC were likely in error and those changes should be rechecked; any patches uploaded between 20:55 and 22:45 UTC were missed entirely by Zuul and should also be rechecked to get fresh test results | 22:59 | |
| *** harrymichal has quit IRC | 23:13 | |
| *** harrymichal has joined #zuul | 23:13 | |
| *** tosky has quit IRC | 23:14 | |
| *** rf0lc0 has quit IRC | 23:30 | |
| *** harrymichal has quit IRC | 23:31 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!