| pabelanger | been trying to use ansible-runner over the last few days, kinda wonder how hard it would be to have it run a zuul playbook | 00:07 |
|---|---|---|
| pabelanger | so far, using it to boot openstack server, via CLI, then run raw play to install ansible dependencies, then other playbooks take over | 00:08 |
| pabelanger | it also support brwap, but haven't tested that step yet | 00:08 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: executor: add log_stream_port and log_stream_file settings https://review.openstack.org/535538 | 00:15 |
| mordred | tristanC: you don't need multiple pipeline. if you made a pipeline, say "manual" that was triggered by web/amqp. then had a section that maps message content to job name - then someone could send some payload like {'project': 'foo/bar', 'job': 'tox-py35'} and only trigger the tox-py35 job defined in the manual pipeline for the foo/bar project. (waving hands about actual payload structure of course) | 00:28 |
| tristanC | mordred: yes, but the problem remains that those job needs the message content. In our case, that would be a release number or artifact url shared through amqp, and we would like to run test with that information | 00:33 |
| tristanC | mordred: i mean, we would like zuul to run test automatically, without having us manually create fake review with the amqp message content... | 00:46 |
| openstackgerrit | Merged openstack-infra/zuul master: scheduler: add job's variants to the rpc job_list method https://review.openstack.org/573473 | 01:11 |
| *** jamesmcarthur has joined #zuul | 01:23 | |
| *** jamesmcarthur has quit IRC | 01:49 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: executor: add zuul_release_nodeset Ansible action https://review.openstack.org/639002 | 02:05 |
| *** jamesmcarthur has joined #zuul | 02:25 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Fix Editable Git install package name parsing https://review.openstack.org/633049 | 02:49 |
| *** jamesmcarthur has quit IRC | 03:25 | |
| *** jamesmcarthur has joined #zuul | 03:28 | |
| *** rlandy has quit IRC | 03:28 | |
| *** bjackman_ has joined #zuul | 03:32 | |
| *** jamesmcarthur has quit IRC | 03:36 | |
| *** bjackman_ has quit IRC | 03:37 | |
| *** jamesmcarthur has joined #zuul | 03:40 | |
| *** raukadah is now known as chandankumar | 03:41 | |
| *** josefwells_ has quit IRC | 03:43 | |
| *** jamesmcarthur has quit IRC | 03:45 | |
| *** bhavikdbavishi has joined #zuul | 03:46 | |
| *** bjackman_ has joined #zuul | 03:59 | |
| *** sdake has joined #zuul | 04:06 | |
| *** jamesmcarthur has joined #zuul | 04:07 | |
| *** sdake has quit IRC | 04:10 | |
| *** sdake has joined #zuul | 04:11 | |
| *** sdake has quit IRC | 04:13 | |
| *** sdake has joined #zuul | 04:14 | |
| *** jamesmcarthur has quit IRC | 04:23 | |
| *** sdake has quit IRC | 04:23 | |
| tobiash | corvus: I think as this feature is still very new I'd vote for merging and mail. Further adding the !unsafe tag already broke our job post processing (for all jobs) so I think that breaks backwards compatibility in any case. | 05:18 |
| *** bhavikdbavishi has quit IRC | 05:35 | |
| *** saneax has joined #zuul | 05:36 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: Add API endpoint to get frozen jobs https://review.openstack.org/607077 | 06:33 |
| tristanC | jhesketh: about the freeze-jobs/frozen-job REST url, I think we have an issue with branch containing a '/', the route_map won't work... | 06:45 |
| tristanC | jhesketh: how do you feel about making branch a query string argument, optional, with a default to 'master'? | 06:46 |
| jhesketh | tristanC: ah, good catch | 06:58 |
| jhesketh | we should be able to have it in the path in the same way we do projects (which can also have slashes) | 06:59 |
| jhesketh | I think we just need {branch:.*} but I'd have to consult cherrypy's docs on that | 06:59 |
| tristanC | jhesketh: it doesn't seems possible to parse org/project/check/stable/branch/job safely | 07:03 |
| tristanC | unless we use known separator like /{pipline}/branch/{branch:.*} | 07:03 |
| jhesketh | hmm, what if we used something like: /api/tenant/{tenant}/project/{project:.*}/branch/{branch:.*} ? | 07:05 |
| jhesketh | it's ugly, but at least consistent with the rest | 07:05 |
| jhesketh | woops, should be /api/tenant/{tenant}/project/{project:.*}/branch/{branch:.*}/freeze_jobs | 07:06 |
| jhesketh | s/-/_ even | 07:06 |
| jhesketh | (sigh, other way around, but you get it) | 07:07 |
| tristanC | i don't mind either way | 07:08 |
| tristanC | what about pipeline, should we put before freeze-jobs ? | 07:08 |
| jhesketh | hmm, we don't have a well defined pattern for routes.. I'm kinda thinking maybe it should go after tenant/before project | 07:13 |
| jhesketh | /api/tenant/{tenant}/pipeline/{pipeline}/project/{project:.*}/branch/{branch:.*}/freeze-jobs | 07:14 |
| tristanC | alright, let me propose this one then | 07:14 |
| tristanC | jhesketh: and how about we rename frozen-jobs with freeze-job, so that we can have .../freeze-job/{job} ? | 07:16 |
| jhesketh | yep, I think I need to do a tidy in 607078 after you've done that | 07:16 |
| jhesketh | right, but the patterns use "...{tenant}/projects" to list all, and "...{tenant}/project/{project}" to select one | 07:17 |
| jhesketh | so I think freeze-jobs is still correct at the higher level | 07:17 |
| tristanC | jhesketh: i'm on it, i already moved the redact_secrets param in the next patch too | 07:17 |
| jhesketh | :thumbsup: | 07:18 |
| *** logan- has quit IRC | 07:35 | |
| *** logan- has joined #zuul | 07:37 | |
| *** gtema has joined #zuul | 07:39 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: Add API endpoint to get frozen jobs https://review.openstack.org/607077 | 07:44 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: Get executor job params https://review.openstack.org/607078 | 07:44 |
| tristanC | jhesketh: i'd like to start working on zuul-runner unit-tests, should we move the current Runner class method to something like a zuul/executor/runner.py module? | 07:58 |
| *** quiquell is now known as quiquell|brb | 08:03 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: Add depends-on support to frozen jobs API https://review.openstack.org/639022 | 08:06 |
| *** bhavikdbavishi has joined #zuul | 08:17 | |
| *** quiquell|brb is now known as quiquell | 08:23 | |
| *** pcaruana has joined #zuul | 08:35 | |
| *** electrofelix has joined #zuul | 08:37 | |
| jhesketh | tristanC: yep, that sounds good to me.. I also wonder if we should move some of zuul/executor/common into there too, but that's just semantics | 08:41 |
| *** jpena|off is now known as jpena | 08:42 | |
| *** hashar has joined #zuul | 08:51 | |
| *** bhavikdbavishi has quit IRC | 08:54 | |
| *** bhavikdbavishi has joined #zuul | 10:04 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: Separate out executor server from runner https://review.openstack.org/607079 | 10:14 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: zuul-runner: implement prep-workspace https://review.openstack.org/607082 | 10:14 |
| tristanC | jhesketh: 607082 now includes a test for the prep_workspace procedure :) | 10:14 |
| jhesketh | tristanC: awesome, thanks :-) | 10:24 |
| tristanC | SpamapS: tobiash: mordred: corvus: it would really help if we could land the first 3 patches of the zuul-runner serie: 607077, 607078 and 607079 | 10:34 |
| jhesketh | tristanC: you've just been uploading individual changes, do you want to rebase the rest of the series with the above fixes? | 10:42 |
| jhesketh | or just get them merged and then worry about it | 10:43 |
| tristanC | jhesketh: it's getting difficult to know what should be the rebase order, and the follow-up changes are missing tests. I would prefer we get the first major refactor merged first | 10:50 |
| jhesketh | that makes sense | 10:50 |
| tristanC | note that zuul-runner execute worked with the current changes | 10:50 |
| mordred | tristanC: +2 from me on those three, looking good. I believe corvus would like to get tobiash and SpamapS to review this stack too | 10:52 |
| mordred | tristanC: oh - duh. you already pinged them :) | 10:52 |
| * mordred is awake too early | 10:52 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: Get executor job params https://review.openstack.org/607078 | 10:54 |
| jhesketh | tristanC: looks like there was a regression in your rebase which I just fixed with ^. Also fixed some code that was duplicated in the common.py | 10:55 |
| *** bhavikdbavishi has quit IRC | 10:56 | |
| *** bhavikdbavishi has joined #zuul | 10:56 | |
| tristanC | jhesketh: tobiash and corvus found that https://review.openstack.org/#/c/607078/12..13/zuul/model.py was wrongly part of 607077, but it's needed by the freeze-job api of 607078 | 11:04 |
| jhesketh | oh, that's my mistake getting confused in the commit series sorry! | 11:05 |
| jhesketh | (gimme a second to fix) | 11:05 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: Get executor job params https://review.openstack.org/607078 | 11:08 |
| jhesketh | mordred: sorry to stuff your reviews around with ^ | 11:08 |
| *** bhavikdbavishi1 has joined #zuul | 11:12 | |
| *** bhavikdbavishi has quit IRC | 11:16 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 11:16 | |
| *** bjackman_ has quit IRC | 11:23 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: Get executor job params https://review.openstack.org/607078 | 11:35 |
| *** jamesmcarthur has joined #zuul | 11:44 | |
| jkt | tristanC: FYI, it looks like the runc driver is too eager to "stick" to node requests to which it doesn't have a quota right now | 11:46 |
| *** bhavikdbavishi has quit IRC | 11:46 | |
| jkt | tristanC: I have it instantiated with four pools (i.e., four runc launchers), I have two jobs running right now, each on just one node, there's one more pending request | 11:47 |
| jkt | and that remaining request looks like it's stuck somehow assigned to a pool which is currently busy even though there are two more free pools | 11:47 |
| *** jpena is now known as jpena|lunch | 11:57 | |
| tristanC | jkt: iirc that a limitation of the nodepool design, once a request is accepted by a launcher, it would wait until there is quota | 11:57 |
| *** quiquell is now known as quiquell|lunch | 12:04 | |
| *** rfolco|rover has quit IRC | 12:11 | |
| *** rfolco|rover has joined #zuul | 12:11 | |
| jkt | tristanC: :(, okay | 12:12 |
| *** quiquell|lunch is now known as quiquell | 12:15 | |
| tristanC | jkt: or maybe there is an issue, i don't remember the full workflow. Though it seems like the launcher service should try to assign other pool if the current one is busy | 12:25 |
| tobiash | tristanC: not sure if I can review it today because of workshop but probably tomorrow | 12:25 |
| *** sanjayu_ has joined #zuul | 12:26 | |
| *** rf0lc0 has joined #zuul | 12:26 | |
| *** shachar has joined #zuul | 12:26 | |
| tristanC | tobiash: thanks, that would be really appreciated | 12:26 |
| *** logan_ has joined #zuul | 12:27 | |
| *** logan_ is now known as Guest11647 | 12:28 | |
| *** dmellado_ has joined #zuul | 12:28 | |
| *** mhu has quit IRC | 12:30 | |
| *** gouthamr has quit IRC | 12:30 | |
| *** electrofelix has quit IRC | 12:30 | |
| *** pwhalen has quit IRC | 12:30 | |
| *** Shrews has quit IRC | 12:30 | |
| *** jlvillal has quit IRC | 12:30 | |
| *** Shrews_ has joined #zuul | 12:30 | |
| *** dmellado has quit IRC | 12:30 | |
| *** rfolco|rover has quit IRC | 12:30 | |
| *** logan- has quit IRC | 12:30 | |
| *** saneax has quit IRC | 12:30 | |
| *** snapiri has quit IRC | 12:30 | |
| *** openstackgerrit has quit IRC | 12:30 | |
| *** dmellado_ is now known as dmellado | 12:30 | |
| *** Guest11647 is now known as logan- | 12:31 | |
| *** jlvillal has joined #zuul | 12:31 | |
| *** gouthamr has joined #zuul | 12:33 | |
| *** rf0lc0 is now known as rfolco|rover | 12:33 | |
| *** jamesmcarthur has quit IRC | 12:34 | |
| *** jamesmcarthur has joined #zuul | 12:34 | |
| *** electrofelix has joined #zuul | 12:35 | |
| jkt | I'm playing with artifact promotion, I'm using a pipeline called "promote" which is triggered by gerrit's change-merged | 12:37 |
| jkt | it looks that artifacts which were built by some jobs in "check" are not available to another job running in "promote", is that expected? | 12:38 |
| jkt | (I understand that there are drawbacks, and that I should be using "gate", but I'm trying to save on some builds here) | 12:39 |
| *** jamesmcarthur has quit IRC | 12:52 | |
| *** jamesmcarthur has joined #zuul | 12:57 | |
| *** sdake has joined #zuul | 13:14 | |
| *** pwhalen has joined #zuul | 13:26 | |
| *** jpena|lunch is now known as jpena | 13:33 | |
| *** rlandy has joined #zuul | 13:33 | |
| *** jamesmcarthur has quit IRC | 13:35 | |
| *** jamesmcarthur has joined #zuul | 13:39 | |
| *** jamesmcarthur has quit IRC | 13:40 | |
| *** jamesmcarthur has joined #zuul | 13:40 | |
| *** sdake has quit IRC | 14:21 | |
| *** sdake has joined #zuul | 14:34 | |
| *** sdake has quit IRC | 14:55 | |
| *** sdake has joined #zuul | 15:00 | |
| *** badboy has quit IRC | 15:05 | |
| *** jamesmcarthur has quit IRC | 15:11 | |
| *** jamesmcarthur has joined #zuul | 15:11 | |
| *** jamesmcarthur has quit IRC | 15:15 | |
| *** jamesmcarthur has joined #zuul | 15:27 | |
| *** jamesmcarthur has quit IRC | 15:30 | |
| *** sanjayu_ has quit IRC | 15:34 | |
| *** sanjayu_ has joined #zuul | 15:35 | |
| *** jamesmcarthur has joined #zuul | 15:39 | |
| *** bhavikdbavishi has joined #zuul | 15:49 | |
| *** sdake has quit IRC | 15:59 | |
| *** sdake has joined #zuul | 16:01 | |
| *** sdake has quit IRC | 16:02 | |
| arxcruz | fungi: hi, any chance you take a look at https://review.openstack.org/#/c/607077 ? already have a +2 from mordred :D | 16:09 |
| arxcruz | this is my dream become true for zuul :D | 16:09 |
| *** sdake has joined #zuul | 16:18 | |
| *** sdake has quit IRC | 16:25 | |
| *** sdake has joined #zuul | 16:27 | |
| *** quiquell is now known as quiquell|off | 16:28 | |
| clarkb | There is a thread on the openstack mailing list that points out zuul job dependencies don't necessarily do what users expect when jobs are skipped due to file rules (potentially other rules too) | 16:29 |
| clarkb | https://review.openstack.org/#/c/568543 illustrates this if you look at the zuul comments | 16:29 |
| clarkb | I'm not sure that is necessarily a bug, but it may be a use case of this feature that zuul should consider | 16:29 |
| *** pcaruana has quit IRC | 16:32 | |
| *** sdake has quit IRC | 16:35 | |
| *** sdake has joined #zuul | 16:36 | |
| *** sdake has quit IRC | 16:40 | |
| *** sdake_ has joined #zuul | 16:41 | |
| corvus | clarkb: i think accomodating that would be very dangerous. i think it would be more surprising for a job to be silently dropped from the graph in that way. the desired behavior can be accomplished easily and safely by simply being explicit about the file matchers on both jobs. | 16:44 |
| *** sdake_ has quit IRC | 16:45 | |
| clarkb | corvus: I think the issue is the job is dropped when tehy want it to run? | 16:46 |
| corvus | clarkb: no, the issue is that zuul refuses to drop the job | 16:46 |
| *** sdake has joined #zuul | 16:46 | |
| corvus | it's right there in the error message: Unable to freeze job graph: Job tripleo-ci-centos-7-scenario001-standalone depends on openstack-tox-py27 which was not run. | 16:46 |
| clarkb | right I think their intention is that tripleo-ci-centos-7-scenario001-standalone would run | 16:47 |
| *** hashar has quit IRC | 16:47 | |
| corvus | they've told zuul "always run tripleo-ci-centos-7-scenario001-standalone, but only do it after openstack-tox-py27 finishes. and, by the way, don't run openstack-tox-py27" | 16:47 |
| clarkb | yup | 16:47 |
| corvus | zuul rightly says "that's nonsense" | 16:47 |
| clarkb | for the general case zuul gets it right, particularly if you use this feature to set up something like a docker registry. But they aren't doing the job dep tree to set up dependencies instead they want to optomize which jobs run and only run things if other things pass (and not running because unnecessary is sort of a weird "pass") | 16:48 |
| corvus | clarkb: yeah, what they're trying to do is a bad idea, and we've told them so | 16:48 |
| clarkb | gotcha | 16:48 |
| corvus | i mean, heck | 16:49 |
| corvus | the *commit* message for that change links to a message i wrote to the mailing list saying "don't do the thing you're talking about doing" | 16:49 |
| corvus | what kind of a supporting citation is that anyway? | 16:49 |
| *** sdake has quit IRC | 16:50 | |
| SpamapS | tristanC: I'll take a look at zuul-runner soon. Wasn't aware anybody was working on it. | 16:51 |
| *** sdake has joined #zuul | 16:51 | |
| corvus | SpamapS, tobiash, mordred: since the zuul-runner stuff is a significant change to the executor, as well as to the audience of zuul (we would, for the first time, be considering a zuul binary an end-user application), i'd like to make sure we (as authors of the most significant parts of the executor) buy into it, so i'm asking all of us (and jhesketh too, though he's partial-author to the changes, so is | 16:58 |
| corvus | somewhat implicit) to +2 those before we merge. See my comment on https://review.openstack.org/607078 | 16:58 |
| corvus | i also would like, apparently, to write run-on sentences with lots of commas. | 16:59 |
| *** sdake has quit IRC | 17:00 | |
| *** sdake_ has joined #zuul | 17:00 | |
| corvus | tobiash: https://review.openstack.org/607077 is ready for your re-review | 17:07 |
| *** openstackgerrit has joined #zuul | 17:12 | |
| openstackgerrit | Tobias Urdin proposed openstack-infra/zuul-jobs master: Use PDK to build puppet module https://review.openstack.org/627534 | 17:12 |
| *** sshnaidm is now known as sshnaidm|afk | 17:17 | |
| pabelanger | Subject: [Important] diskimage-builder dependency yum-utils removed from Rawhide | 17:18 |
| pabelanger | just seen that in inbox | 17:18 |
| pabelanger | which means, DIB on fedora won't be able to build fedora / centos any more, since we use yum-utils for that | 17:18 |
| clarkb | pabelanger: I'm guessing yum is gone and needs dnf? | 17:18 |
| pabelanger | yah | 17:19 |
| pabelanger | I think there is a dnf-utils | 17:19 |
| pabelanger | but haven't really looked | 17:19 |
| pabelanger | ianw: ^if you didn't see | 17:19 |
| pabelanger | i believe we are still okay building them from debuntu | 17:20 |
| SpamapS | ugh, my global site variable of ansible_python_interpreter=/usr/bin/python3 has now backfired on me... | 17:21 |
| SpamapS | because now I can't use stock centos7 images :-P | 17:21 |
| * SpamapS glares at Ansible | 17:21 | |
| pabelanger | SpamapS: Yah, I ran into that issue on ansible-network last week too | 17:22 |
| SpamapS | Has anyone come up with a valid work-around for using stock images? | 17:22 |
| pabelanger | still stuck on python2 | 17:22 |
| clarkb | SpamapS: dmsimard suggested we do it yesterday, but I think there are other problems for openstack's install (namely rax's particular networking) so not sure | 17:23 |
| SpamapS | The problem I have is that AFAICT, there's no other level at which I can correct the problem than image build. | 17:23 |
| pabelanger | https://review.openstack.org/637338/ seems to be a way to setup python-path on nodepool side, but haven't reviewed yet | 17:23 |
| SpamapS | Yeah that looks like a good plan, though that's actually not the ansible_python_interpreter problem I have. | 17:25 |
| SpamapS | The one I have is that Ubuntu 18.04+ has no /usr/bin/python2, and CentOS7 has no /usr/bin/python3 :-P | 17:25 |
| SpamapS | I guess label python path works. | 17:25 |
| corvus | SpamapS, tobiash, mordred: reading further down the zuul-runner stack, i now realize i made my cut-off for extra reviews too deep -- we should all be reviewing the parent of that one as well: https://review.openstack.org/607077 | 17:26 |
| pabelanger | let me double check how we are doing network appliances | 17:26 |
| pabelanger | there are some without python installed | 17:26 |
| SpamapS | If there was a way to tell nodepool or ansible to do a raw task that normalizes things, that would work. | 17:26 |
| SpamapS | Otherwise I have to take on the extra weight of building images. | 17:27 |
| corvus | SpamapS: the nodepool thing should work, right? | 17:27 |
| SpamapS | corvus: I think so yeah, I'm looking now. | 17:27 |
| * SpamapS was just getting used to running patch free.. ;) | 17:27 | |
| corvus | i think that's the best solution, because we're basically just providing correct information about the images to ansible | 17:27 |
| *** gtema has quit IRC | 17:28 | |
| pabelanger | Hmm, I thought ricky had a patch up to modify runAnsibleSetup not to use -m setup, then first base jobs could be raw | 17:30 |
| SpamapS | pabelanger: yeah that would make a lot of sense too | 17:32 |
| pabelanger | yah, in our case, we are hoping to boot vendor network VMs, liked I say some without python. So we use raw in some case, or different connection plugin. However, I know we have some challenges because zuul blacklists some of the ones we need | 17:34 |
| SpamapS | pabelanger: seems like something we could make configurable per-tenant. | 17:44 |
| SpamapS | You know.. ever since the build page started showing fail results.. | 17:45 |
| SpamapS | I basically never want to see the logs. | 17:45 |
| corvus | mordred: can you review https://review.openstack.org/637813 ? | 17:45 |
| SpamapS | The logs are for going deeper, but I almost always get what I want from the json interpretation. | 17:45 |
| SpamapS | We should sweep through the UI and flip things from log url to build url. | 17:46 |
| corvus | SpamapS: i agree, let's see if we can convince mordred, who had some reservations last time this came up. | 17:46 |
| SpamapS | Like, in the comment... I find myself avoiding clicking the comment links and instead clicking the status link (which I have set to /builds?change={change}) so I can start from builds. | 17:47 |
| corvus | (this seems backwards to me -- i would expect me to be the one with reservations and mordred to be gung-ho) | 17:47 |
| SpamapS | Exception proves the rule? | 17:48 |
| corvus | SpamapS, mordred: or, more constructively, i think we do all want to get to that point, what are the steps we need to do to get there? and, i guess, let's write them down this time :) | 17:48 |
| SpamapS | crazy talk | 17:52 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Request child nodes of paused jobs at higher priority https://review.openstack.org/638796 | 17:56 |
| clarkb | apparently zuul is on the openstack ops meetup agenda, https://etherpad.openstack.org/p/BER19-OPS-DUDES | 18:00 |
| *** jpena is now known as jpena|off | 18:15 | |
| *** chandankumar is now known as raukadah | 18:16 | |
| *** panda|ruck is now known as panda|ruck|off | 18:28 | |
| *** sdake_ has quit IRC | 18:30 | |
| *** sdake has joined #zuul | 18:33 | |
| *** sdake has quit IRC | 18:35 | |
| *** sdake has joined #zuul | 18:35 | |
| *** sdake has quit IRC | 18:46 | |
| fungi | i saw! along with a david bowie reference apparently | 18:46 |
| *** rfolco|rover has quit IRC | 18:46 | |
| fungi | (or could also have been a mott the hoople reference i suppose) | 18:47 |
| *** sdake has joined #zuul | 18:50 | |
| *** sdake_ has joined #zuul | 18:55 | |
| *** sdake has quit IRC | 18:55 | |
| *** sdake_ has quit IRC | 18:56 | |
| *** sdake has joined #zuul | 18:58 | |
| *** sanjayu_ has quit IRC | 18:59 | |
| *** sdake has quit IRC | 19:00 | |
| *** sdake_ has joined #zuul | 19:02 | |
| *** bhavikdbavishi has quit IRC | 19:02 | |
| *** sdake_ has quit IRC | 19:06 | |
| *** sdake has joined #zuul | 19:06 | |
| *** sdake has quit IRC | 19:10 | |
| *** rfolco has joined #zuul | 19:12 | |
| *** sdake has joined #zuul | 19:12 | |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Support requests for specific providers https://review.openstack.org/639418 | 19:20 |
| *** sdake has quit IRC | 19:20 | |
| *** gouthamr has quit IRC | 19:20 | |
| *** gouthamr_ has joined #zuul | 19:20 | |
| *** sdake_ has joined #zuul | 19:21 | |
| *** sdake_ has quit IRC | 19:25 | |
| *** sdake has joined #zuul | 19:26 | |
| *** rfolco is now known as rfolco|rover | 19:28 | |
| *** sdake has quit IRC | 19:35 | |
| *** sdake_ has joined #zuul | 19:36 | |
| *** sdake_ has quit IRC | 19:45 | |
| *** sdake has joined #zuul | 19:47 | |
| ianw | pabelanger: yeah, and we have the inverse of no dnf on debuntu afaics, and then no zypper on bionic+ too ... | 19:54 |
| SpamapS | Cannot see zypper without thinking of Dr. Evil | 19:54 |
| *** sdake has quit IRC | 19:55 | |
| SpamapS | https://imgur.com/gallery/vc3SAyt | 19:55 |
| corvus | SpamapS: you have successfully propogated a meme | 19:56 |
| *** sdake has joined #zuul | 19:56 | |
| SpamapS | corvus: #WheresMyCookie | 19:57 |
| *** sdake has quit IRC | 20:03 | |
| mordred | corvus, SpamapS: my main objection to switching to the build page (although I agree with SpamapS that what I want to see is the build page) is that it has the potential to make getting to the console log be an additional click "click to build, click to logs dir, click console log file" | 20:05 |
| mordred | which is why I'd sort of been wanting to have the build page have the rest of the json parsing so that "get to console log" could be "click to expand console log to see more than just the log snipped" | 20:06 |
| mordred | BUT | 20:06 |
| mordred | maybe that's perfect in the way of good | 20:06 |
| mordred | and like SpamapS says, the error chunk in the build page is probably what I want to see from the console log most of the time anyway | 20:07 |
| mordred | so maybe let's do it already | 20:07 |
| *** jamesmcarthur has quit IRC | 20:15 | |
| *** jamesmcarthur has joined #zuul | 20:16 | |
| *** jamesmcarthur has quit IRC | 20:20 | |
| *** delhage has quit IRC | 20:27 | |
| *** delhage has joined #zuul | 20:28 | |
| SpamapS | mordred: yeah I'm finding the error chunk is 90% of what I want, and so I always want to start there.. even if I then want to click to error logs. | 20:31 |
| openstackgerrit | Merged openstack-infra/zuul master: Log to job output when running Ansible setup https://review.openstack.org/637813 | 20:39 |
| corvus | tobiash: do you want to update https://review.openstack.org/638801 with a release note? | 20:40 |
| *** gouthamr_ is now known as gouthamr | 20:42 | |
| *** takamatsu_ has quit IRC | 20:48 | |
| *** takamatsu_ has joined #zuul | 20:50 | |
| *** spsurya has quit IRC | 21:03 | |
| *** jamesmcarthur has joined #zuul | 21:11 | |
| *** manjeets_ has joined #zuul | 21:22 | |
| *** manjeets has quit IRC | 21:24 | |
| *** zbr|out has quit IRC | 21:27 | |
| ianw | pabelanger / SpamapS: i guess in the end, it's all just wrappers around .rpm files, which is all just wrappers around a cpio file ... | 21:28 |
| SpamapS | ianw: which is just wrappers around unix filesystem assumptions | 21:31 |
| *** manjeets_ is now known as manjeets | 21:35 | |
| *** sdake has joined #zuul | 22:28 | |
| daniel2 | I finally was able to get nodepool to build images. That was a really educational experience for me. Was able to at least upgrade to 0.5.0, now I need to learn ansible. | 22:29 |
| *** sdake has quit IRC | 22:30 | |
| *** sdake has joined #zuul | 22:33 | |
| *** sdake has quit IRC | 22:40 | |
| *** sdake_ has joined #zuul | 22:40 | |
| *** sdake_ has quit IRC | 22:45 | |
| openstackgerrit | James E. Blair proposed openstack-infra/nodepool master: Support requests for specific providers https://review.openstack.org/639418 | 22:47 |
| *** sdake has joined #zuul | 22:48 | |
| corvus | Shrews_, mordred, clarkb, fungi, tobiash: ^ that should get us paused-job node-provider affinity | 22:48 |
| corvus | well, the nodepool part at least | 22:48 |
| *** jamesmcarthur has quit IRC | 22:49 | |
| jkt | wow wow wow | 22:54 |
| jkt | so, urlib.parse.urlparse converts hostname to lowercase | 22:54 |
| jkt | which means that my gerrit_connection.canonical_hostname won't match | 22:55 |
| jkt | that looks like a good explanation for my multitenancy issue | 22:55 |
| *** sdake has quit IRC | 22:56 | |
| *** sdake has joined #zuul | 22:58 | |
| *** sdake has quit IRC | 23:00 | |
| *** jamesmcarthur has joined #zuul | 23:02 | |
| *** sdake has joined #zuul | 23:02 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Request dependent job nodes from paused parent provider https://review.openstack.org/639467 | 23:04 |
| *** sdake has quit IRC | 23:06 | |
| *** sdake_ has joined #zuul | 23:06 | |
| *** jamesmcarthur has quit IRC | 23:09 | |
| *** sdake_ has quit IRC | 23:10 | |
| tobiash | corvus: cool :) | 23:11 |
| *** sdake has joined #zuul | 23:11 | |
| tobiash | will review tomorrow | 23:12 |
| tobiash | And yes, will add a release note to the base64 change | 23:13 |
| *** sanjayu_ has joined #zuul | 23:19 | |
| *** sdake has quit IRC | 23:20 | |
| *** sdake has joined #zuul | 23:23 | |
| *** sdake has quit IRC | 23:25 | |
| *** sdake has joined #zuul | 23:27 | |
| *** sdake has quit IRC | 23:30 | |
| *** sdake has joined #zuul | 23:32 | |
| *** sdake has quit IRC | 23:35 | |
| *** sdake has joined #zuul | 23:36 | |
| *** sdake has quit IRC | 23:37 | |
| *** sdake has joined #zuul | 23:38 | |
| *** sdake has quit IRC | 23:40 | |
| *** sdake has joined #zuul | 23:42 | |
| *** sdake has quit IRC | 23:50 | |
| *** jesusaur has quit IRC | 23:51 | |
| *** sdake has joined #zuul | 23:52 | |
| jkt | when returning artifacts, what is the purpose of artifact.name? Is that related to the name given in job.requires and job.provides? | 23:54 |
| jkt | I'm asking because I have six providing jobs and six consumig jobs, I see that different artifacts are stored in SQL for the providing jobs and that their results as reported by API contain correct artifacts | 23:55 |
| jkt | but when my consuming jobs are executed, they all get artifact URL which was created by the first provider | 23:55 |
| *** sdake has quit IRC | 23:55 | |
| *** sdake has joined #zuul | 23:56 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!