| *** jkilpatr has joined #zuul | 00:00 | |
| *** lennyb has joined #zuul | 00:02 | |
| *** JasonCL has quit IRC | 00:04 | |
| *** JasonCL has joined #zuul | 00:20 | |
| *** JasonCL has quit IRC | 00:25 | |
| jlk | hahahahahahaha | 00:25 |
|---|---|---|
| jlk | jeblair: so.... github3.py is in the middle of a refactor, that they want to finish before doing a new release. | 00:25 |
| jlk | my life is porting fucking code around. | 00:26 |
| jeblair | jlk: wow. maybe i've been watching too much jeremy clarkson, but could you use some heavy machinery to move it around...? | 00:28 |
| jlk | hah | 00:28 |
| jlk | I picked the wrong week to have a "dry week" | 00:29 |
| jeblair | jlk: you have a feel for how long this might be? like... a couple of weeks... or they're 2 years into a 6 week project? | 00:29 |
| jlk | more like the latter. https://github.com/sigmavirus24/github3.py/issues/670 about 1 year into a 6 week project | 00:30 |
| jlk | that ran out of steam. Partially done. | 00:30 |
| jlk | I've picked up one redesign to get a feel for it, as for how long it'd take to finish the rest. I've been on it most of a day and I'm not done... | 00:30 |
| *** shunde has joined #zuul | 00:34 | |
| shunde | Hi, if I have a list of nodes to run jobs, can I supply a static list to zuulv3 and not use nodepool? | 00:35 |
| shunde | n.b. I know which node a job should run on, and the node is ready | 00:37 |
| clarkb | shunde: https://review.openstack.org/#/c/468624/ is the current idea. Basically nodepool would register static nodes and broker them via zookeeper | 00:37 |
| clarkb | so zuul speaks the same language regardless of how nodes are provisioned | 00:37 |
| jeblair | mordred: zuul-web restarted, should have the cors headers now | 00:43 |
| shunde | clarkb: thanks, would that be in the next nodepool release? is there a document about how to configure nodepool for zuulv3? | 00:46 |
| clarkb | shunde: yes that is a required feature for the release, so it will be there and willcome with docs | 00:49 |
| *** jkilpatr has quit IRC | 00:58 | |
| *** JasonCL has joined #zuul | 01:10 | |
| fungi | rcarrillocruz: dire as it is, net neutrality isn't entirely done and dusted yet. now state governors are banding together to overturn the fcc decision, so... stay tuned? | 01:13 |
| *** JasonCL has quit IRC | 01:26 | |
| jlk | oooh! movement. Things are getting merged now on github3.py, and I was granted write access. | 01:38 |
| mrhillsman | NICE! | 01:56 |
| mrhillsman | some feedback on the zuulv3 updated dashboard - builds tab does not have pagination so after a certain number of jobs, cannot scroll or visit next page of builds, and is it possible to have an option to hide or by default hide base jobs | 01:58 |
| *** JasonCL has joined #zuul | 02:30 | |
| *** harlowja has joined #zuul | 02:41 | |
| *** harlowja has quit IRC | 02:42 | |
| *** harlowja has joined #zuul | 02:43 | |
| *** shunde has quit IRC | 02:44 | |
| *** harlowja has quit IRC | 03:15 | |
| *** jappleii__ has joined #zuul | 03:17 | |
| *** jappleii__ has quit IRC | 03:19 | |
| *** jappleii__ has joined #zuul | 03:19 | |
| *** threestrands_ has quit IRC | 03:20 | |
| *** jappleii__ has quit IRC | 03:56 | |
| *** dmellado has quit IRC | 04:27 | |
| *** dmellado has joined #zuul | 04:35 | |
| *** harlowja has joined #zuul | 04:36 | |
| *** harlowja has quit IRC | 04:45 | |
| *** harlowja has joined #zuul | 05:21 | |
| *** harlowja has quit IRC | 05:22 | |
| *** harlowja has joined #zuul | 05:24 | |
| *** harlowja has quit IRC | 05:29 | |
| dmsimard | Huh, shippable's CI is based on Ansible too apparently: http://blog.shippable.com/how-to-create-on-demand-test-environments | 05:37 |
| *** harlowja has joined #zuul | 05:38 | |
| *** harlowja has quit IRC | 06:32 | |
| openstackgerrit | dongwenjuan proposed openstack-infra/zuul master: use local gearman for merger service https://review.openstack.org/528162 | 06:39 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Update javascript jobs to take npm_command variable https://review.openstack.org/527770 | 06:49 |
| *** nguyentrihai has quit IRC | 08:02 | |
| *** xinliang has quit IRC | 08:32 | |
| *** jpena|off is now known as jpena | 08:43 | |
| *** xinliang has joined #zuul | 08:44 | |
| *** hashar has joined #zuul | 08:55 | |
| kklimonda | Hi, I started seeing this traceback a lot recently: http://paste.openstack.org/show/629030/ | 09:26 |
| kklimonda | my working theory is that zuul-merger gets stuck trying to update repository, and somehow job gets aborted? cancelled? | 09:27 |
| kklimonda | and zuul-scheduler is just killing reconfiguration, as all jobs have to be processed in order | 09:27 |
| kklimonda | I see zuul-merger getting stuck during git fetch/pull - there seems to be a bug in gitpython library, where _timeout code that is supposed to kill git process and all its children is not killing grandchildren, and leaving git process in defunct state | 09:28 |
| kklimonda | but I'm not sure how could that bubble up to zuul-scheduler.. | 09:29 |
| *** electrofelix has joined #zuul | 09:53 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-jobs master: Fix npm-run-test job https://review.openstack.org/528224 | 10:32 |
| *** baiyi has quit IRC | 10:34 | |
| *** dmellado has quit IRC | 10:39 | |
| *** dmellado has joined #zuul | 10:44 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-jobs master: Fix npm-run-test job https://review.openstack.org/528224 | 11:03 |
| *** JasonCL has quit IRC | 11:26 | |
| *** JasonCL has joined #zuul | 11:27 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Fix npm-run-test job https://review.openstack.org/528224 | 11:34 |
| *** nguyentrihai has joined #zuul | 11:38 | |
| *** jkilpatr has joined #zuul | 11:38 | |
| Wei_Liu | hi | 11:39 |
| Wei_Liu | hello | 11:39 |
| *** JasonCL has quit IRC | 11:40 | |
| *** JasonCL has joined #zuul | 11:40 | |
| Wei_Liu | I found that the when zuul execute ansible playbook, it setup a ansible lib, which cause shell command like "helm serve &" terminated in 10 seconds, can I remove this ansible lib? | 11:41 |
| *** JasonCL has quit IRC | 11:42 | |
| Wei_Liu | tristanC: hello | 11:42 |
| *** JasonCL has joined #zuul | 11:43 | |
| Wei_Liu | tristanC: Could the zuul execute ansible jobs without /var/lib/zuul/ansible/zuul/ansible/library? | 11:44 |
| *** JasonCL has quit IRC | 11:49 | |
| *** JasonCL has joined #zuul | 11:49 | |
| *** jkilpatr has quit IRC | 11:51 | |
| *** jkilpatr has joined #zuul | 11:52 | |
| *** JasonCL has quit IRC | 12:05 | |
| *** JasonCL has joined #zuul | 12:06 | |
| *** jkilpatr has quit IRC | 12:07 | |
| Wei_Liu | hello, I found that the when zuul execute ansible playbook, it setup a ansible lib, which cause shell command like "helm serve &" terminated in 10 seconds, can I remove this ansible lib? | 12:08 |
| *** JasonCL has quit IRC | 12:11 | |
| *** nguyentrihai has quit IRC | 12:12 | |
| *** nguyentrihai has joined #zuul | 12:13 | |
| *** nguyentrihai has quit IRC | 12:18 | |
| *** nguyentrihai has joined #zuul | 12:19 | |
| Shrews | Wei_Liu: no. zuul installs some modified versions of plugins, callbacks and library modules. | 12:20 |
| *** JasonCL has joined #zuul | 12:22 | |
| Wei_Liu | Shrews: How can I execute shell like "helm serve &" in ansible task to make it alive? | 12:24 |
| openstackgerrit | Akihiro Motoki proposed openstack-infra/zuul-jobs master: Fix npm-run-test job https://review.openstack.org/528244 | 12:25 |
| openstackgerrit | Jens Harbott (frickler) proposed openstack-infra/zuul-jobs master: Fix npm-run-test job once more https://review.openstack.org/528245 | 12:26 |
| Shrews | Wei_Liu: are you using the command module to start it? That is one we've modified to capture logging. It may not work for that use case. Maybe using 'shell' will work for you? | 12:28 |
| Wei_Liu | Shrews: I used "shell", it didn't work | 12:29 |
| Wei_Liu | Shrews: I try to copy a shell script file to node and execute it, hope it works | 12:31 |
| Shrews | jeblair: mordred: anything about background processes in job playbooks that might confuse zuul? | 12:34 |
| *** yolanda__ has joined #zuul | 12:36 | |
| *** flepied_ has joined #zuul | 12:43 | |
| *** flepied has quit IRC | 12:47 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Fix npm-run-test job https://review.openstack.org/528244 | 12:52 |
| *** jpena is now known as jpena|lunch | 12:55 | |
| tobiash | Wei_Liu: you need to daemonize it | 12:55 |
| tobiash | for that you could use e.g. http://manpages.ubuntu.com/manpages/precise/de/man8/start-stop-daemon.8.html | 12:55 |
| tobiash | or install it as a systemd unit and start the service | 12:56 |
| tobiash | depends on your use case | 12:56 |
| tobiash | I used start-stop-daemon for starting a resource monitoring tool on the start of the build | 12:57 |
| Wei_Liu | tobiash: cool, thanks. I will try it. | 12:57 |
| tobiash | Wei_Liu: example: http://paste.openstack.org/show/629047/ | 13:00 |
| *** dkranz has joined #zuul | 13:19 | |
| *** JasonCL has quit IRC | 13:19 | |
| openstackgerrit | Fabien Boucher proposed openstack-infra/zuul feature/zuulv3: Git driver https://review.openstack.org/525614 | 13:33 |
| fbo_ | Hi, regarding gerrit/github connections there is a maintainCache method defined but it seems it is never called. maintainConnectionCache call in the scheduler is commented. Is it normal ? | 13:51 |
| *** jpena|lunch is now known as jpena | 13:58 | |
| dmsimard | rcarrillocruz: oh wow your ansible-container zuul stuff is actually pretty cool | 13:59 |
| gundalow | \o/ | 14:00 |
| dmsimard | rcarrillocruz: I think some of those bits could be made easier if you used the packaged software factory repo | 14:01 |
| dmsimard | everything from zuul to nodepool and zookeeper is there: http://38.145.34.53/kojifiles/repos/sf-2.7-el7-release/source/SRPMS/ | 14:02 |
| dmsimard | well, here actually: http://38.145.34.53/kojifiles/repos/sf-2.7-el7-release/Mash/ | 14:02 |
| tobiash | fbo_: this method is a relict of v2 and broken with v3 | 14:27 |
| tobiash | fbo_: see https://review.openstack.org/#/c/502121/ and the change it reverts | 14:28 |
| tristanC | dmsimard: rcarrillocruz: actually to install sf zuul package, you better do: yum install http://softwarefactory-project.io/repos/sf-release-2.7.rpm && yum install rh-python35-zuul zookeeper-lite | 14:29 |
| fbo_ | tobiash: alright so it means the cache of changes for the gerrit driver will grow until the scheduler is restarted | 14:31 |
| tobiash | fbo_: currently yes | 14:45 |
| fbo_ | tobiash: thanks for the explanation, I think we can keep the maintainCache method in git driver too then it should work when re-activated later. | 14:47 |
| *** openstackgerrit has quit IRC | 14:48 | |
| tobiash | fbo_: sounds ok | 14:52 |
| *** openstackgerrit has joined #zuul | 14:57 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Use yarn and webpack to manage status and streaming https://review.openstack.org/487538 | 14:57 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add babel transpiling enabling use of ES6 features https://review.openstack.org/528295 | 14:57 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add StandardJS linting and analysis https://review.openstack.org/528296 | 14:57 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Fix StandardJS warnings and turn them to errors https://review.openstack.org/528297 | 14:57 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add bundle analysis to the lint target https://review.openstack.org/528298 | 14:57 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Revert "Revert "Add sphinx_python variable to sphinx role and job"" https://review.openstack.org/526666 | 14:59 |
| *** myoung is now known as myoung|training | 15:00 | |
| jeblair | kklimonda: what version of gitpython are you using? | 15:05 |
| clarkb | Wei_Liu: Shrews & is a shell construct. command likely will enver work with that even with vanilla ansible as command isn't going to expand shell things | 15:06 |
| jeblair | kklimonda: however, i suspect that even if the timeout isn't working quite right, it's working well enough to finish the gearman job and return control to zuul, but the scheduler has no error handling there. so we probably need to handle that case and perhaps retry. | 15:06 |
| kklimonda | @jeblair I believe I have your temporary fork installed - I have the commit from the tip | 15:07 |
| jeblair | kklimonda: ok. both issues are probably worth looking into (better cleanup on timeout kill, and error handling on job return), but i think they're separate. | 15:08 |
| kklimonda | you are most likely correct, the kill goes through after all, just leaving some processes around | 15:10 |
| clarkb | Shrews: jeblair: so before shrews takes off on vacation, did we want to do anything more for that nodepool thing I patched? maybe drop the exception handling from poll() and just do it in the handler cleanup? | 15:10 |
| *** harlowja has joined #zuul | 15:11 | |
| *** harlowja has quit IRC | 15:13 | |
| *** JasonCL has joined #zuul | 15:14 | |
| Shrews | clarkb: I think what you have is perfect. We *probably* don't need the handling in poll() now, but it certainly doesn't hurt to leave it | 15:14 |
| Shrews | clarkb: I'm only around for a couple of more hours, so poke my brain now if you need to :) | 15:15 |
| clarkb | Shrews: ok, just wanted to make sure there wasn't some piece we needed to change then would forget about | 15:15 |
| clarkb | (I'm fine with it as is too) | 15:15 |
| *** ricky_at_rh_madr has joined #zuul | 15:15 | |
| Shrews | clarkb: I do think we should restart the launchers to grab it, but we've also merged a new feature with the quota stuff, so that complicates it a bit | 15:16 |
| jeblair | clarkb: yeah. i'm less itchy about the 'unecessary' exception handler now that we think it shouldn't be used. i could go either way. | 15:17 |
| *** JasonCL has quit IRC | 15:18 | |
| *** JasonCL has joined #zuul | 15:18 | |
| *** electrofelix has quit IRC | 15:19 | |
| clarkb | Wei_Liu: Shrews as for why & doesn't work with shell either the reason is & background a process that is still parented to the current shell. My hunch is that once that shell goes away at the end of the module execution the shell cleans up its children and it doesn't reparent to init. So tobiash is correct you will need to daemonize it | 15:19 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Correctly use project name in getGitUrl https://review.openstack.org/528309 | 15:38 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Initialize github auth in getGitUrl if needed https://review.openstack.org/528310 | 15:38 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Use configured github server in app mode https://review.openstack.org/528311 | 15:38 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix github app authentication in _getNeededByFromPR https://review.openstack.org/528312 | 15:38 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Add the project to GithubUser https://review.openstack.org/528313 | 15:38 |
| tobiash | jlk, SpamapS: this is a first stack to make github apps working on github enterprise ^^ | 15:39 |
| tobiash | with that I had working gating for a first test pull request | 15:39 |
| tobiash | note that this is without any tests yet | 15:44 |
| mordred | tobiash: https://review.openstack.org/#/c/528311/ - +2 with a comment | 15:49 |
| tobiash | mordred: good point | 15:53 |
| tobiash | Had a quick hack/deploy/tryout cycle and cut the changes into pieces afterwards | 15:54 |
| tobiash | So I'm sure there's room for improvement ;) | 15:54 |
| clarkb | tobiash: did you see that shrews thought he had tracked down that nodepool issue yesterday? | 15:55 |
| tobiash | clarkb: yes | 15:55 |
| *** flepied has joined #zuul | 16:11 | |
| *** flepied_ has quit IRC | 16:12 | |
| SpamapS | https://engineering.instagram.com/let-your-code-type-hint-itself-introducing-open-source-monkeytype-a855c7284881 | 16:23 |
| *** hashar has quit IRC | 16:35 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Correctly use project name in getGitUrl https://review.openstack.org/528309 | 16:36 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-jobs master: npm-test needs xvfb etc https://review.openstack.org/528329 | 16:40 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-jobs master: npm-test needs xvfb etc https://review.openstack.org/528329 | 16:42 |
| openstackgerrit | James E. Blair proposed openstack-infra/zone-zuul-ci.org master: Initial commit https://review.openstack.org/528331 | 16:42 |
| jlk | tobiash: woo, I'll look later today. I'm deep in some github3.py refactoring | 16:47 |
| *** ricky_at_rh_madr has quit IRC | 16:58 | |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Initialize github auth in getGitUrl if needed https://review.openstack.org/528310 | 17:02 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Use configured github server in app mode https://review.openstack.org/528311 | 17:02 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Use yarn and webpack to manage status and streaming https://review.openstack.org/487538 | 17:02 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add babel transpiling enabling use of ES6 features https://review.openstack.org/528295 | 17:02 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add StandardJS linting and analysis https://review.openstack.org/528296 | 17:03 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Fix StandardJS warnings and turn them to errors https://review.openstack.org/528297 | 17:03 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add bundle analysis to the lint target https://review.openstack.org/528298 | 17:03 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix github app authentication in _getNeededByFromPR https://review.openstack.org/528312 | 17:05 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Add the project to GithubUser https://review.openstack.org/528313 | 17:05 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-jobs master: npm-test needs xvfb etc https://review.openstack.org/528329 | 17:07 |
| *** jpena is now known as jpena|off | 17:08 | |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Install yarn if needed in javascript jobs https://review.openstack.org/527771 | 17:11 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Use configured github server in app mode https://review.openstack.org/528311 | 17:16 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Fix github app authentication in _getNeededByFromPR https://review.openstack.org/528312 | 17:16 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Add the project to GithubUser https://review.openstack.org/528313 | 17:16 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul feature/zuulv3: Use python3 for docs publication https://review.openstack.org/525718 | 17:18 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul feature/zuulv3: Update sphinx jobs to use python3 https://review.openstack.org/525690 | 17:31 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-jobs master: npm-test needs xvfb etc https://review.openstack.org/528329 | 17:54 |
| tobiash | jeblair: found a doc typo in https://review.openstack.org/#/c/527805/2 | 17:55 |
| tobiash | lgtm othereise | 17:55 |
| tobiash | otherwise | 17:55 |
| jeblair | tobiash: ok if i fix that in a followup to preserve the +2 ? | 17:57 |
| tobiash | fine for me | 17:57 |
| * mordred was just goign to suggest the same thing | 17:58 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul feature/zuulv3: Fix doc typo https://review.openstack.org/528361 | 17:58 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Update sphinx jobs to use python3 https://review.openstack.org/525690 | 18:14 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add implied-branches pragma directive https://review.openstack.org/527805 | 18:14 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Add debug project-pipeline option https://review.openstack.org/527828 | 18:14 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: npm-test needs xvfb etc https://review.openstack.org/528329 | 18:22 |
| openstackgerrit | Merged openstack-infra/zuul feature/zuulv3: Fix doc typo https://review.openstack.org/528361 | 18:30 |
| openstackgerrit | Tobias Henkel proposed openstack-infra/zuul feature/zuulv3: Add support for protected jobs https://review.openstack.org/522985 | 18:31 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add babel transpiling enabling use of ES6 features https://review.openstack.org/528295 | 18:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add StandardJS linting and analysis https://review.openstack.org/528296 | 18:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Fix StandardJS warnings and turn them to errors https://review.openstack.org/528297 | 18:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add bundle analysis to the lint target https://review.openstack.org/528298 | 18:32 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-jobs master: translations: Check for translation-source first https://review.openstack.org/528369 | 18:56 |
| *** smyers has joined #zuul | 18:57 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Fix StandardJS warnings and turn them to errors https://review.openstack.org/528297 | 19:15 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add bundle analysis to the lint target https://review.openstack.org/528298 | 19:15 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Fix source_url handling for jobs view https://review.openstack.org/528373 | 19:15 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Make bundle of build web content https://review.openstack.org/528374 | 19:15 |
| mordred | jeblair: fwiw, the stack above is *close* to being ready for review - I'm still seeing a difference in behavior between gate and my laptop that I'm trying to track down | 19:16 |
| mordred | namely - http://localhost/builds.html?source_url=http:%2F%2Fzuulv3.openstack.org works fine but http://logs.openstack.org/98/528298/3/check/zuul-npm-build/615cdb5/web/builds.html?source_url=http:%2F%2Fzuulv3.openstack.org is empty | 19:17 |
| mordred | so - you know - once I've got that sorted, it should be good and stuff | 19:19 |
| jeblair | mordred: tenant=web ? | 19:19 |
| jeblair | mordred: http://zuulv3.openstack.org/builds.json?tenant=web | 19:19 |
| mordred | ah - nod. | 19:19 |
| mordred | jeblair: thanks! that makes sense - will work on fixing | 19:20 |
| jeblair | yay i helped! | 19:20 |
| mordred | jeblair: have I mentioned how much I like the whole "publish the web app to the build logs and point it to the live api service" thing? | 19:21 |
| jeblair | i'm a fan | 19:22 |
| openstackgerrit | James E. Blair proposed openstack-infra/zone-zuul-ci.org master: Initial commit https://review.openstack.org/528331 | 19:32 |
| *** smyers has quit IRC | 19:38 | |
| mmedvede | jeblair: I am out next week in case puppet-openstackci zuulv3 comes up then. I have been updating https://storyboard.openstack.org/#!/story/2001367 with related patches. | 19:47 |
| jeblair | mmedvede: oh thanks! if we work on that, we'll probably just update your patches then, if that's okay. | 19:48 |
| jeblair | mmedvede: are you out through the rest of the year? | 19:48 |
| mmedvede | jeblair: of course feel free to update! I hope the direction I took it is not too much from what you had in mind | 19:49 |
| mmedvede | jeblair: I am out until the day after christmas | 19:49 |
| mmedvede | I am trying to deploy my little zuulv3 allinone node as I go. So far struggling with apache2 rewrites for zuul-web | 19:50 |
| *** hashar has joined #zuul | 19:53 | |
| jeblair | mmedvede: yeah, that's still in flux. it'll be better soon | 19:57 |
| *** JasonCL has quit IRC | 19:58 | |
| *** JasonCL has joined #zuul | 20:01 | |
| *** JasonCL has quit IRC | 20:06 | |
| *** JasonCL has joined #zuul | 20:19 | |
| *** jkilpatr has joined #zuul | 20:21 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Fix source_url handling for jobs view https://review.openstack.org/528373 | 20:22 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Fix StandardJS warnings and turn them to errors https://review.openstack.org/528297 | 20:22 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add bundle analysis to the lint target https://review.openstack.org/528298 | 20:22 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Make bundle of build web content https://review.openstack.org/528374 | 20:22 |
| *** flepied has quit IRC | 20:35 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Make bundle of build web content https://review.openstack.org/528374 | 20:38 |
| *** openstackgerrit has quit IRC | 20:48 | |
| *** flepied has joined #zuul | 20:51 | |
| *** openstackgerrit has joined #zuul | 20:52 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Add bundle analysis to the lint target https://review.openstack.org/528298 | 20:52 |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Make bundle of build web content https://review.openstack.org/528374 | 20:52 |
| mordred | http://logs.openstack.org/98/528298/5/check/zuul-npm-build/7b101c2/web/status.html?source_url=http://zuulv3.openstack.org http://logs.openstack.org/98/528298/5/check/zuul-npm-build/7b101c2/web/builds.html?source_url=http:%2F%2Fzuulv3.openstack.org and http://logs.openstack.org/98/528298/5/check/zuul-npm-build/7b101c2/web/jobs.html?source_url=http:%2F%2Fzuulv3.openstack.org all work - and | 20:54 |
| mordred | http://logs.openstack.org/98/528298/5/check/nodejs-npm-run-lint/991ef87/npm/reports/bundle.html shows a nice little report on bundle contents | 20:54 |
| mordred | jeblair: ^^ | 20:54 |
| mordred | jeblair: also - the webpack flow is generating "map" files for the minified results - so if you open developer console in a web browser and get error messages or use the debugger or anything, it'll link you to readable source code, rather than minified junk | 20:55 |
| mordred | so http://logs.openstack.org/98/528298/5/check/zuul-npm-build/7b101c2/web/status.bundle.js is the minified bundle, but next to it is http://logs.openstack.org/98/528298/5/check/zuul-npm-build/7b101c2/web/status.bundle.js.map - which has the source and mapping info but is only loaded when a browser asks for it | 20:58 |
| *** JasonCL has quit IRC | 21:06 | |
| *** smyers has joined #zuul | 21:06 | |
| *** smyers has quit IRC | 21:09 | |
| *** smyers has joined #zuul | 21:12 | |
| *** openstackgerrit has quit IRC | 21:18 | |
| jeblair | mordred: sweet! thanks! | 21:23 |
| jeblair | mordred: i'm thinking i'll jump into that stack for real first thing next week | 21:23 |
| mordred | jeblair: I was just thinking about writing a mailing list "so there's this stack - some suggestions on how to approach it" message | 21:23 |
| mordred | jeblair: tires to kick, hoods to look under, teeth to count, etc | 21:24 |
| jeblair | mordred: cool. i'll probably read that next week if you do. :) i'm too friday brained to start something like that now, but it sounds like it would be helpful. | 21:24 |
| jeblair | mordred: may also help you remember what you were thinking in your fevered delirium. ;) | 21:24 |
| mordred | jeblair: oh - yah - definitely don't try to start in on it on a friday unless your assessment of it is going to be limited to "the status page in the build logs looks correct" | 21:25 |
| mordred | jeblair: heh. there's no chance of that :) | 21:25 |
| *** smyers has quit IRC | 21:29 | |
| *** smyers has joined #zuul | 21:33 | |
| *** dkranz has quit IRC | 21:38 | |
| dmsimard | I was pulling data from RDO's Zuul SQL reporter to get some minimal metrics -- I did some of it manually but in case you're curious: https://i.imgur.com/r276T9j.png | 21:42 |
| dmsimard | It dwarfs the numbers from zuulv3.o.o :) | 21:45 |
| jeblair | dmsimard: which numbers? | 21:46 |
| *** JasonCL has joined #zuul | 21:47 | |
| dmsimard | jeblair: the amount of jobs | 21:47 |
| dmsimard | jeblair: as in, zuul's RDO is pretty small in comparison | 21:47 |
| mordred | dmsimard: 21k jobs in a month? I think I remember back when we only ran 21k jobs / month :) | 21:47 |
| jeblair | dmsimard: oh i think i misunderstood the direction of dwarfing :) | 21:47 |
| dmsimard | mordred: yeah, and that growth pattern you see is TripleO moving their jobs off of RH1 as third party from review.rdo | 21:48 |
| mordred | dmsimard: I like the rise in average compute time per job from august to november | 21:48 |
| dmsimard | mordred: tripleo explains the raise in avg job time, yes :) | 21:48 |
| mordred | dmsimard: tripleo. the cause of and solution to all life's problems. | 21:49 |
| mordred | dmsimard: wait - I think that's supposed to be s/tripleo/alcohol/ | 21:49 |
| dmsimard | jeblair, mordred: do you mind if I get those same numbers from zuulv3.o.o ? It's a fairly simple query: SELECT job_name,result,start_time,end_time,timediff(end_time,start_time) as delta FROM zuul_build WHERE start_time BETWEEN '2017-08-01 00:00:00' AND '2017-08-31 23:59:59'; | 21:50 |
| *** dkranz has joined #zuul | 21:51 | |
| mordred | dmsimard: sure, I don't think that's problematic | 21:52 |
| mordred | dmsimard: if you fleshed it out a little, it might be useful to calculate average build time both including and not including aborted jobs | 21:52 |
| dmsimard | don't get me started on implementing a full fledged dashboard/table on the new zuul dashboard :P | 21:54 |
| pabelanger | dmsimard: you should be able to setup a query in http://graphite.openstack.org/ with statsd | 21:55 |
| pabelanger | http://grafana.openstack.org/dashboard/db/zuul-status today renders jobs launched an hour | 21:55 |
| dmsimard | oh that's right, graphite is probably more convenient than SQL | 21:55 |
| dmsimard | we don't have graphite in RDO (yet) | 21:55 |
| pabelanger | would likely update that to over 30d | 21:55 |
| dmsimard | pabelanger: yeah, the numbers I was looking for was totals | 21:55 |
| dmsimard | I can probably work something out for totals | 21:56 |
| pabelanger | think you can get totals from graphite | 21:56 |
| pabelanger | cannot remember the function | 21:56 |
| dmsimard | I could look it up | 21:57 |
| pabelanger | if math is right, in November we did ~260k jobs | 22:01 |
| pabelanger | summarize(sumSeries(stats_counts.zuul.tenant.openstack.pipeline.*.all_jobs), '1d') | 22:02 |
| pabelanger | select date range | 22:02 |
| pabelanger | in graphite.o.o | 22:02 |
| dmsimard | zuul_build shows 325865 for november | 22:02 |
| dmsimard | maybe there's a distinction between the data you pulled up and mine -- i.e, executor builds vs buildsets? or something | 22:03 |
| dmsimard | clarkb: how did you pull up the job numbers/compute time back in denver ? | 22:11 |
| clarkb | dmsimard: it was a scan of the nodepool logs using a script that jeblair gave me | 22:19 |
| clarkb | dmsimard: I don't think it will work anymore due to the new nodepool stuff for zuulv3 | 22:19 |
| jeblair | yeah, though i think we could probably recreate that with sql queries (similar to what dmsimard has started) | 22:20 |
| clarkb | oh right with the sql reporter we should have all that info | 22:24 |
| clarkb | job name, start stop time, do we have number of nodes requested? | 22:24 |
| clarkb | number of nodes requested was actually the tricky bit last time because there are a non trivial number of multinode jobs now | 22:24 |
| dmsimard | clarkb: hmm let me see | 22:24 |
| dmsimard | clarkb: not in zuul afaict, maybe we can cross reference with the nodepool database if the data is available there http://paste.openstack.org/raw/629091/ | 22:27 |
| *** myoung|training is now known as myoung | 22:32 | |
| dmsimard | clarkb: there's no date data in the nodepool database so hard to correlate | 22:33 |
| clarkb | dmsimard: ya you might need to build a nodeset to job name mapping out of the config itself maybe | 22:34 |
| dmsimard | oh, wait, there is -- 1495191423 | 22:34 |
| jeblair | clarkb: yep. that's a hole. we should figure out how to put info like that in the sql db. | 22:34 |
| dmsimard | state_time | 22:34 |
| jeblair | clarkb: but we should anticipate differently sized nodes, so a simple node count isn't really sufficent. we probably need a list of labels. so that's either an arbitrarily long string (could go with that and be best effort), or another table + join. | 22:35 |
| dmsimard | jeblair: that data might not be necessary if there's a way to correlate nodes from nodepool to zuul ? like if zuul passed a buildset uuid and nodepool recorded that X and Y nodes were assigned to that buildset | 22:35 |
| jeblair | or we could aggregate: node_count, node_cpu_count, node_ram, etc -- if we change nodepool to give that data back to zuul | 22:36 |
| jeblair | dmsimard: nodepool's database is ephemeral, so you'd have to correlate with logs | 22:36 |
| dmsimard | jeblair: is it ? I mean, it's mysql | 22:36 |
| jeblair | dmsimard: in any version it's ephemeral, nodes are deleted from db when they are gone. in v3 it's zk not mysql. | 22:37 |
| dmsimard | oh, right | 22:37 |
| clarkb | jeblair: good point | 22:38 |
| *** hashar has quit IRC | 22:43 | |
| *** openstackgerrit has joined #zuul | 23:19 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul feature/zuulv3: Remove use strict https://review.openstack.org/528437 | 23:19 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
