*** jamesmcarthur has joined #zuul | 00:08 | |
*** jamesmcarthur has quit IRC | 00:19 | |
mordred | corvus: we don't have dockerhub tags because at the moment doing so would be manual ... and hard - because we don't tend to tag tip of master, and we don't have a good way to map from a git commit to a particular older image IIRC | 00:24 |
---|---|---|
mordred | corvus: but it's been a hectic last month for me, so I could easily be wrong or misremembering | 00:25 |
corvus | mordred: i agree; with our current state, we would need to have post jobs build docker images. that's totally doable, it's just that no one has typed in the words to make that happen | 00:26 |
corvus | er, tag jobs | 00:26 |
corvus | er, release jobs | 00:26 |
corvus | promote for tip; release for tags. | 00:27 |
corvus | which is how docs are set up for zuul-sphinx (and will be for all the projects when we move them to the zuul tenant) | 00:28 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration roles https://review.opendev.org/668767 | 00:33 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration roles https://review.opendev.org/668767 | 00:34 |
mordred | corvus: ah - yes - nod | 00:38 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base integration roles https://review.opendev.org/668061 | 00:49 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration roles https://review.opendev.org/668767 | 00:49 |
*** panda has quit IRC | 01:03 | |
*** panda has joined #zuul | 01:05 | |
*** swest has quit IRC | 01:46 | |
*** sgw has quit IRC | 01:51 | |
*** swest has joined #zuul | 02:01 | |
*** bhavikdbavishi has joined #zuul | 02:34 | |
*** rlandy has quit IRC | 02:34 | |
*** bhavikdbavishi1 has joined #zuul | 02:36 | |
*** bhavikdbavishi has quit IRC | 02:38 | |
*** bhavikdbavishi1 is now known as bhavikdbavishi | 02:38 | |
*** rfolco has quit IRC | 02:47 | |
*** jamesmcarthur has joined #zuul | 04:08 | |
*** jamesmcarthur has quit IRC | 04:24 | |
*** bolg has joined #zuul | 04:43 | |
*** saneax has joined #zuul | 04:52 | |
*** pcaruana has joined #zuul | 05:06 | |
*** saneax has quit IRC | 05:14 | |
*** wxy-xiyuan has quit IRC | 05:18 | |
*** bolg has quit IRC | 05:33 | |
*** igordc has quit IRC | 05:41 | |
*** snapiri has quit IRC | 06:26 | |
*** snapiri has joined #zuul | 06:26 | |
*** hwangbo has quit IRC | 06:38 | |
*** themroc has joined #zuul | 07:01 | |
*** tosky has joined #zuul | 07:20 | |
*** toabctl has joined #zuul | 07:30 | |
*** hashar has joined #zuul | 07:35 | |
*** jamesmcarthur has joined #zuul | 08:00 | |
*** saneax has joined #zuul | 08:05 | |
*** jamesmcarthur has quit IRC | 08:06 | |
*** sshnaidm|afk is now known as sshnaidm|ruck | 09:19 | |
*** tosky has quit IRC | 09:39 | |
*** tosky has joined #zuul | 09:40 | |
*** bhavikdbavishi has quit IRC | 09:42 | |
*** hashar has quit IRC | 09:46 | |
*** dmsimard9 has joined #zuul | 10:08 | |
*** dmsimard has quit IRC | 10:08 | |
*** saneax has quit IRC | 10:50 | |
*** bkorren has joined #zuul | 11:11 | |
bkorren | hi there, I'm trying to use 'zuul_return' from a job to customize the results URL Zuul posts to Gerrit - and its not working for me, can someone help me to figure out how to debug it? | 11:13 |
*** saneax has joined #zuul | 11:17 | |
mordred | bkorren: in your zuul, is there also a zuul_return in the base job that might be running later and overriding what your job is returning? | 11:31 |
mordred | bkorren: (also, the yaml you sent in the rh email might be useful to eyes out here - mind if I put it in to paste.openstack.org to share it? or you can) | 11:32 |
AJaeger | mordred: good morning, want to review a zuul-jobs reorg, please? https://review.opendev.org/#/c/668763/ | 11:35 |
mordred | AJaeger: always | 11:35 |
bkorren | mordred, you can put it there, sure. | 11:35 |
AJaeger | thanks, mordred | 11:35 |
mordred | AJaeger: wow. that works? neat | 11:36 |
bkorren | mordred, could it be overriden bu the post-run playbook in the base job? I'm using the default one,... | 11:38 |
mordred | http://paste.openstack.org/show/753809/ <-- this is the zuul_return invocation from bkorren | 11:38 |
mordred | bkorren: yes - the post-run in the base job will run after the post-run in your job's playbook. we should see what corvus thinks when he gets up about how to maybe address that | 11:41 |
bkorren | mordred, got a build report here - https://ovirt-staging.softwarefactory-project.io/logs/56/256/15/check-patch/trigger-gate-job/bf0d24c/ara-report/ | 11:41 |
bkorren | mordred, actually I'm not seeing a call to zuul_return in the post playbook | 11:42 |
mordred | yeah - then that's definitely what's going on. you probably won't - because it tends to run after the logs have been uploaded | 11:43 |
mordred | what we might want to do is return the log_url as an artifact_url ... but I'm still on first coffee so not having *awesome* thinking power yet :) | 11:45 |
openstackgerrit | Merged zuul/zuul-jobs master: Shard project-pipeline config https://review.opendev.org/668763 | 11:46 |
bkorren | mordred, I see, I'll have a look at the playbook/job source code if I'm missing something | 11:50 |
AJaeger | mordred: zuul is full of surprises ;) | 11:56 |
openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: Add base integration roles https://review.opendev.org/668061 | 11:56 |
openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: Add multi-node integration roles https://review.opendev.org/668767 | 11:56 |
*** cosss_ has quit IRC | 12:06 | |
*** bhavikdbavishi has joined #zuul | 12:13 | |
openstackgerrit | Andreas Jaeger proposed zuul/zuul-jobs master: Add multi-node integration roles https://review.opendev.org/668767 | 12:28 |
*** rlandy has joined #zuul | 12:33 | |
*** dmsimard9 is now known as dmsimard | 12:43 | |
bkorren | mordred, found the culprit - its the `uload-logs` role called from the `post.yaml` playbook - unfortunately there does not seem to be an easy way to make it just upload logs without also setting the the `log_url` | 13:00 |
*** bhavikdbavishi1 has joined #zuul | 13:00 | |
bkorren | mordred, I can customize `post.yaml` in my tenant to make that happen but I'm not sure that would be smart in the long run | 13:01 |
*** bhavikdbavishi has quit IRC | 13:02 | |
*** bhavikdbavishi1 is now known as bhavikdbavishi | 13:02 | |
*** jeliu_ has joined #zuul | 13:40 | |
*** bhavikdbavishi has quit IRC | 13:41 | |
ofosos | Got a problem with nodepool. The docs read: | 13:47 |
ofosos | `if Nodepool is running inside Kubernetes, this setting and the kube/config file may be omitted` w.r.t. Kubernetes provider | 13:48 |
ofosos | But the source doesn't tolerate omitting this setting AFAICT https://github.com/openstack-infra/nodepool/blob/master/nodepool/driver/kubernetes/config.py | 13:48 |
tobiash | corvus: shall we land 661880? The full stack looks to be +2 now | 13:52 |
tobiash | (cleanup-phase) | 13:52 |
corvus | bkorren, mordred: i have an idea, but there are a couple of small things missing from zuul -- what you should be able to do is to return that as an *artifact* rather than a log_url. then, when you look at the build page in zuul, it should show you a link to the artifact url. The things missing from zuul are: the build page doesn't show artifact urls yet (the api does, so it's just a little JS), and, this | 13:54 |
corvus | will be a lot smoother when zuul reports the link to the build page rather than the log_url. | 13:54 |
corvus | tobiash: done, thx! | 13:55 |
tobiash | :) | 13:55 |
corvus | ofosos: flaper87 just added that feature, maybe he can clarify | 13:56 |
*** sgw has joined #zuul | 13:57 | |
corvus | bkorren: in the mean time, you could use "success-url" and "failure-url" to override the link that zuul reports for the job: https://zuul-ci.org/docs/zuul/user/config.html#attr-job.success-url | 13:58 |
*** rfolco has joined #zuul | 14:01 | |
bkorren | corvus, yeah, but that would make it static for the whole job, while I need it to be different in each run | 14:02 |
bkorren | corvus, I'm working on a solution though - I'm changing post.yaml a little so I could tell it not to set log_url and let the value I set through | 14:03 |
corvus | bkorren: ah, i see you're templating values not available in job config | 14:03 |
corvus | bkorren: interested in hacking a little javascript? | 14:04 |
bkorren | corvus, mordred : anyway, thanks for your help so far, I need to leave now; good day | 14:04 |
bkorren | corvus, I thought about doing a JS redirect; but that makes me feel dirty ;) | 14:04 |
corvus | bkorren: i was wondering if you had some time to help with the js display of artifacts -- that's going to be the long-term solution to this, and it's pretty close | 14:05 |
corvus | bkorren: anyway, if you do, let us know when you get back -- i don't mean to delay you :) | 14:06 |
*** bkorren has quit IRC | 14:09 | |
AJaeger | corvus: FYI, I updated your changes 668061 and 668767, the first is ready to merge now - the second has a failure in the job. | 14:13 |
corvus | AJaeger: thanks, just looking at that -- the second isn't actually testing with multiple nodes yet -- i think my script is going to have to be more complicated. it's interesting that only the persistent-firewall job failed -- maybe the rest just short-circuit because there's no "switch" node? | 14:14 |
AJaeger | corvus: ah! | 14:16 |
AJaeger | really interesting ;( | 14:16 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration roles https://review.opendev.org/668767 | 14:23 |
corvus | AJaeger: okay. that's gonna burn some nodes. :) | 14:24 |
AJaeger | yeah ;( | 14:24 |
corvus | 54 | 14:24 |
AJaeger | wow, new record? | 14:25 |
corvus | i think we've used more on some system-config changes (like in the 80s?) | 14:25 |
corvus | but the good news is that in the future, it shouldn't use any more than it did previously, possibly fewer, as long as only one role is being updated (the typical case) | 14:25 |
mordred | corvus: that's amazing | 14:26 |
corvus | this is only because we're adding a bunch at once | 14:26 |
AJaeger | corvus: agreed | 14:26 |
corvus | also, if we don't like this approach, we can go back to testing all these roles in 2 jobs... i mostly wanted to see whether separating them as job-per-role was cleaner (made it more obvious that each was tested) | 14:28 |
corvus | but we will still have jobs that test multiple roles (for instance, the registry job tests a bunch of roles, because they all work together in sequence) | 14:29 |
mordred | corvus: so far I like the separation | 14:31 |
openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: Create tasks that spin up kubernetes and run the operator https://review.opendev.org/668029 | 14:31 |
corvus | oh, i forgot to parent these to base-minimal, so one more rev on both patches | 14:32 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base integration roles https://review.opendev.org/668061 | 14:33 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration roles https://review.opendev.org/668767 | 14:33 |
openstackgerrit | Merged zuul/zuul master: model: add cleanup-run to the job configuration https://review.opendev.org/661880 | 14:43 |
AJaeger | tristanC, I see no new documentation on https://review.opendev.org/#/c/661880/ . could you add some, please? | 14:47 |
corvus | AJaeger: it's a followup | 14:49 |
corvus | 3 patch series, and all 3 are approved | 14:49 |
corvus | https://review.opendev.org/662147 is docs | 14:49 |
corvus | (we held the series until everything was ready) | 14:50 |
*** igordc has joined #zuul | 14:50 | |
openstackgerrit | Merged zuul/zuul master: executor: run cleanup playbook on stop https://review.opendev.org/661881 | 14:51 |
AJaeger | corvus: good! Sorry, tristanC for missing that | 14:51 |
corvus | i'd rather too many people check for docs than not enough :) | 14:52 |
AJaeger | ;) | 14:54 |
AJaeger | corvus: https://review.opendev.org/668061 is failing with post_failures now, the rebase to base-minimal broke some jobs | 14:59 |
openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: Create tasks that spin up kubernetes and run the operator https://review.opendev.org/668029 | 15:02 |
*** mattw4 has joined #zuul | 15:06 | |
*** mattw4 has quit IRC | 15:09 | |
*** mattw4 has joined #zuul | 15:09 | |
*** mattw4 has quit IRC | 15:11 | |
*** mattw4 has joined #zuul | 15:12 | |
openstackgerrit | Merged zuul/zuul master: docs: add cleanup-run documentation https://review.opendev.org/662147 | 15:16 |
corvus | SpamapS: ^ cleanup landed! | 15:16 |
*** jamesmcarthur has joined #zuul | 15:21 | |
*** jamesmcarthur has quit IRC | 15:21 | |
*** jamesmcarthur has joined #zuul | 15:22 | |
SpamapS | ssssaaahhhweeeet | 15:24 |
SpamapS | corvus:all my dreams are coming true | 15:24 |
AJaeger | ;) | 15:25 |
SpamapS | now I just need EC2 nodepool builder and zuul pushes. | 15:25 |
*** hashar has joined #zuul | 15:30 | |
mordred | SpamapS: zuul push solves the universe | 15:38 |
mordred | well - except that github doesn't let apps push to protected branches, IIRC - so that'll need to be sorted out | 15:39 |
*** jangutter has quit IRC | 15:42 | |
*** rfolco has quit IRC | 15:47 | |
*** jeliu_ has quit IRC | 16:05 | |
SpamapS | mordred:well if you wanted to solve the universe you should have put that in the requirements! | 16:06 |
SpamapS | For the apps-can't-push thing.. I think the right thing to do is to yell at jlk | 16:06 |
mordred | SpamapS: you mean nudge him with lovingkindness I'm sure | 16:08 |
SpamapS | that's not what yell means? | 16:08 |
SpamapS | this explains so much | 16:08 |
*** hwangbo has joined #zuul | 16:25 | |
*** jeliu_ has joined #zuul | 16:31 | |
openstackgerrit | Kevin Carter (cloudnull) proposed zuul/zuul-jobs master: Add python3 detection to the sphinx role https://review.opendev.org/668954 | 16:33 |
*** jeliu_ has quit IRC | 16:35 | |
*** jeliu_ has joined #zuul | 16:38 | |
openstackgerrit | Kevin Carter (cloudnull) proposed zuul/zuul-jobs master: Add python3 detection to the sphinx role https://review.opendev.org/668954 | 16:43 |
openstackgerrit | Kevin Carter (cloudnull) proposed zuul/zuul-jobs master: Add python3 detection to the sphinx role https://review.opendev.org/668954 | 16:44 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base integration roles https://review.opendev.org/668061 | 16:47 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration job https://review.opendev.org/668767 | 16:47 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add a script to make platform-specific versions of jobs https://review.opendev.org/668955 | 16:47 |
*** hashar has quit IRC | 16:48 | |
corvus | AJaeger, mordred: there's some amount of interaction between the roles in those test jobs, so i think it's going to be easiest to stick with them as two jobs rather than job-per-role. but i think we can look at moving some of those roles out separately later -- probably starting with fetch-zuul-cloner (it should no longer need to be run under base-minimal, and i doubt anything relies on it). | 16:49 |
corvus | AJaeger, mordred: i did update the multinode roles so they don't depend on base-minimal or the playbook from the other job anymore -- i don't think that should be necessary. so at least that's a simpler job. | 16:49 |
corvus | (but even for it, at least 2 of its roles interact) | 16:50 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base role integration jobs https://review.opendev.org/668061 | 16:52 |
openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: Create tasks that spin up kubernetes and run the operator https://review.opendev.org/668029 | 16:52 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration jobs https://review.opendev.org/668767 | 16:52 |
*** jamesmcarthur has quit IRC | 16:56 | |
*** jamesmcarthur has joined #zuul | 16:57 | |
openstackgerrit | Kevin Carter (cloudnull) proposed zuul/zuul-jobs master: Add python3 detection to the sphinx role https://review.opendev.org/668954 | 16:59 |
*** jamesmcarthur has quit IRC | 17:02 | |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add test job for ensure-sphinx https://review.opendev.org/668958 | 17:03 |
*** openstackgerrit has quit IRC | 17:04 | |
*** openstackgerrit has joined #zuul | 17:13 | |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add a script to make platform-specific versions of jobs https://review.opendev.org/668955 | 17:13 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base role integration jobs https://review.opendev.org/668061 | 17:13 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration jobs https://review.opendev.org/668767 | 17:13 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add test job for ensure-sphinx https://review.opendev.org/668958 | 17:20 |
*** jamesmcarthur has joined #zuul | 17:28 | |
*** hwangbo has quit IRC | 17:44 | |
*** jamesmcarthur has quit IRC | 17:44 | |
*** hwangbo has joined #zuul | 17:44 | |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base role integration jobs https://review.opendev.org/668061 | 17:44 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration jobs https://review.opendev.org/668767 | 17:44 |
*** igordc has quit IRC | 17:45 | |
*** sshnaidm|ruck is now known as sshnaidm|afk | 17:53 | |
*** rfolco has joined #zuul | 17:56 | |
*** buumi has joined #zuul | 18:04 | |
buumi | Hey, i was reading briefly about Zuul, do you think it would help with our situation. We've project A and B. If specific files are changed in Project A, our integration tests will fail. When Project A submodule in repository of Project B is updated, tests should pass again. Would Zuul CI make it possible to do so that submodule update of Project B | 18:09 |
buumi | would depend on change to Project A and both commits would be merged if integration tests pass after submodule update? | 18:09 |
AJaeger | mordred: want to review https://review.opendev.org/#/c/668955 for zuul-jobs, please? | 18:11 |
clarkb | buumi: yes with one current behavior requirement (simply because the other thing hasn't yet been implemented) zuul does a strict sequencing of changes. This means if A is queued before B and passes it will merge regardless of B's result. But B if queued after A will only merge if B works with A's changes | 18:11 |
clarkb | there has been discussion of doing atomic sets of merges across projects but that work hasn't been done yet. So if you don't need that you should be in a good spot | 18:13 |
buumi | Ok that's fine. And if A fails but later queued B will pass, it will then merge both A and B? | 18:13 |
clarkb | buumi: as long as B properly expresses its dependency on A zuul will require A pass and merge first yes | 18:13 |
clarkb | there are dependency tags you can set in commit messages to express that | 18:13 |
SpamapS | buumi:one caveat: submodules do not work well with Zuul | 18:14 |
SpamapS | I've been down that road before, and it was extremely painful. | 18:14 |
clarkb | SpamapS: thats largely a matter of updating the base jobs though right? | 18:14 |
clarkb | (I mean I agree its painful, but should also be doable) | 18:14 |
buumi | Cool, i was worried if B fails to merge if A first fails or it just merges B, but that sounds fine. Need to experiment with it then | 18:15 |
SpamapS | clarkb: In theory yes. submodules are just a hidden config file with hashes. But in practice, this has been a fail every time I touched it. | 18:15 |
clarkb | SpamapS: I think that has more to do with submodules than with zuul :) | 18:15 |
buumi | We also had in the past rebase-submit strategy which i guess would cause some issues with submodules since commit SHA might change, but luckily now that's not an issue anymore | 18:16 |
SpamapS | What did work, was just assembling a directory from two git working copies, with no submodules involved. | 18:16 |
SpamapS | buumi:I'd highly recommend that you do some basic tests. Submodules are rigid in the way they relate to git hashes, and Zuul is pretty opinionated about how it wants to assemble your git repos. | 18:17 |
SpamapS | (And, IMO, just abandon submodules. They're probably not realy helping you) | 18:18 |
buumi | Ok, that's something to test then, thanks for help both of you! | 18:19 |
clarkb | ya I think the bit that is missing is updating the submodule hashes on related projects based on their hashes in the adjacent zuul managed repos | 18:19 |
SpamapS | git clone A ; git clone B ; ln -s ../A B/A | 18:19 |
SpamapS | far more predictable | 18:19 |
SpamapS | clarkb:yeah, we had tried to spike on that a berzillion years ago (pirate cloud days) but IIRC there were enough problems we just started ripping submodules out and doing symlinks. | 18:20 |
clarkb | SpamapS: were you doing it at the job level or in the zuul mergers? | 18:20 |
SpamapS | job | 18:20 |
*** mattw4 has quit IRC | 18:21 | |
clarkb | hrm I wouldn't expect that to be too difficult since zuul doesn't say "this is the sha you should have" so you could make a quick update based on the on disk repo state | 18:21 |
clarkb | was much harder in the past with v2 when zuul said "here is your sha(s)" | 18:21 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base role integration jobs https://review.opendev.org/668061 | 18:21 |
SpamapS | the mergers were doing fine delivering the state of the repos. But that was one problem in particular. If somebody was submitting an update to the submodules.. what do we do? Because we don't want to test with the hashes they just submitted, but with the git state Zuul prepared. | 18:21 |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add multi-node integration jobs https://review.opendev.org/668767 | 18:21 |
clarkb | SpamapS: ah I think the problem there is/was github :P Gerrit will auto update those shas for you | 18:22 |
SpamapS | so once we started ignoring the hashes.. we wondered "what are these even doing for us?" | 18:22 |
clarkb | so you can test the future and never manually update the hashes | 18:22 |
SpamapS | no it was on Gerrit | 18:22 |
SpamapS | these were not Gerrit auto submodules | 18:22 |
SpamapS | (those had other problems ;) | 18:22 |
SpamapS | anyway, I'm pretty sour on submodules. It's a misfeature of git, IMO. | 18:23 |
clarkb | yup I agree and I think finally many others are starting to see it tht way too (golang doesn't use submodules for vendoring anymore aiui for example) | 18:24 |
*** buumi has quit IRC | 18:25 | |
openstackgerrit | Kevin Carter (cloudnull) proposed zuul/zuul-jobs master: Update the sphinx python package sets https://review.opendev.org/668954 | 18:28 |
SpamapS | clarkb:yeah, golang has almost gotten to sane dependency management now. ;) | 18:32 |
*** hashar has joined #zuul | 18:46 | |
*** themroc has quit IRC | 18:49 | |
*** mattw4 has joined #zuul | 18:54 | |
*** jeliu_ has quit IRC | 18:54 | |
*** jeliu_ has joined #zuul | 19:00 | |
*** jeliu_ has quit IRC | 19:05 | |
*** igordc has joined #zuul | 19:05 | |
*** mattw4 has quit IRC | 19:30 | |
*** igordc has quit IRC | 19:31 | |
*** mattw4 has joined #zuul | 19:41 | |
openstackgerrit | James E. Blair proposed zuul/zuul-jobs master: Add base role integration jobs https://review.opendev.org/668061 | 19:45 |
*** mattw4 has quit IRC | 19:57 | |
*** mattw4 has joined #zuul | 19:57 | |
openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: Create tasks that spin up kubernetes and run the operator https://review.opendev.org/668029 | 20:40 |
* fungi is not here at the moment, but wholeheartedly agrees there's little practical benefit to submodules over just having multiple git repos and some instruction to users to clone or symlink repo b inside repo a | 20:41 | |
*** igordc has joined #zuul | 21:14 | |
*** pcaruana has quit IRC | 21:16 | |
*** rfolco has quit IRC | 21:22 | |
portdirect | hey - is there is spec of proposal re `zuul/zuul-operator` | 21:43 |
portdirect | sound interesting from the name, im starting to look seriously at deploying zuul in k8s | 21:44 |
clarkb | portdirect: https://review.opendev.org/#/c/659180/ | 21:44 |
portdirect | clarkb: thanks dude! will read now :D | 21:44 |
portdirect | at 1st pass this looks really nice - totally makes sense to reuse ansible operator here | 21:49 |
portdirect | my timeline means I'm gonna have to forge ahead with a `dumb` helm chart for now, but will certainly try and contribute in any way i can to this effort | 21:49 |
*** rlandy is now known as rlandy|bbl | 21:58 | |
SpamapS | portdirect: several of us have gone the "dumb" route (not helm, but.. one-off non-reusable k8s deployment) | 22:33 |
SpamapS | Looking forward to the day when I replace my janky scripts with an operator. :) | 22:34 |
portdirect | yeah - it really seems like an event driven application like zuul would benefit from the approach | 22:34 |
portdirect | do you have an example of a k8s deployment of zuul atm, that i could use for 'ahem' inspiration? | 22:35 |
portdirect | in the meantime i was planning on just k8s'izing these: https://opendev.org/zuul/zuul/src/branch/master/doc/source/admin/examples/docker-compose.yaml | 22:37 |
*** tosky has quit IRC | 22:38 | |
SpamapS | portdirect: Mine is very environment specific, so sharing it wouldn't be productive | 23:02 |
portdirect | Gotcha, thanks SpamapS | 23:04 |
*** mattw4 has quit IRC | 23:21 | |
openstackgerrit | James E. Blair proposed zuul/zuul master: Fix multi-tenant caching of extra config files https://review.opendev.org/669008 | 23:34 |
openstackgerrit | James E. Blair proposed zuul/zuul master: Fix multi-tenant caching of extra config files https://review.opendev.org/669008 | 23:37 |
*** saneax has quit IRC | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!