| *** jamesmcarthur has quit IRC | 00:09 | |
| *** mattw4 has quit IRC | 00:12 | |
| *** jamesmcarthur has joined #zuul | 00:16 | |
| *** jamesmcarthur has quit IRC | 00:38 | |
| *** igordc has quit IRC | 01:00 | |
| *** spsurya has joined #zuul | 01:17 | |
| *** altlogbot_0 has quit IRC | 01:37 | |
| *** altlogbot_1 has joined #zuul | 01:39 | |
| *** jamesmcarthur has joined #zuul | 01:44 | |
| *** jamesmcarthur has quit IRC | 01:49 | |
| *** tributarian has quit IRC | 01:51 | |
| *** bhavikdbavishi has joined #zuul | 01:52 | |
| *** rlandy|bbl is now known as rlandy | 01:55 | |
| *** rlandy has quit IRC | 01:57 | |
| *** bhavikdbavishi1 has joined #zuul | 02:04 | |
| *** bhavikdbavishi has quit IRC | 02:06 | |
| *** bhavikdbavishi1 is now known as bhavikdbavishi | 02:06 | |
| *** weshay_dentist is now known as weshay | 02:13 | |
| *** jamesmcarthur has joined #zuul | 02:18 | |
| *** bhavikdbavishi has quit IRC | 02:49 | |
| *** jamesmcarthur has quit IRC | 02:53 | |
| *** jamesmcarthur has joined #zuul | 02:54 | |
| *** jamesmcarthur has quit IRC | 02:59 | |
| *** jamesmcarthur has joined #zuul | 03:04 | |
| *** jamesmcarthur has quit IRC | 03:17 | |
| *** yolanda has quit IRC | 03:21 | |
| *** yolanda has joined #zuul | 03:22 | |
| *** bhavikdbavishi has joined #zuul | 03:35 | |
| *** jamesmcarthur has joined #zuul | 03:40 | |
| *** bjackman has joined #zuul | 04:21 | |
| *** jamesmcarthur has quit IRC | 04:40 | |
| *** jank has joined #zuul | 04:46 | |
| *** jhesketh has joined #zuul | 04:54 | |
| *** jamesmcarthur has joined #zuul | 05:09 | |
| *** jamesmcarthur has quit IRC | 05:15 | |
| *** sshnaidm is now known as sshnaidm|afk | 05:19 | |
| *** jamesmcarthur has joined #zuul | 06:11 | |
| *** jamesmcarthur has quit IRC | 06:16 | |
| *** jamesmcarthur has joined #zuul | 06:46 | |
| *** jamesmcarthur has quit IRC | 06:51 | |
| *** jamesmcarthur has joined #zuul | 07:00 | |
| *** jamesmcarthur has quit IRC | 07:05 | |
| *** themroc has joined #zuul | 07:08 | |
| *** tosky has joined #zuul | 07:24 | |
| *** jamesmcarthur has joined #zuul | 07:39 | |
| *** jamesmcarthur has quit IRC | 07:43 | |
| *** gtema has joined #zuul | 07:56 | |
| *** jamesmcarthur has joined #zuul | 08:08 | |
| *** electrofelix has joined #zuul | 08:11 | |
| *** jamesmcarthur has quit IRC | 08:12 | |
| *** jamesmcarthur has joined #zuul | 08:14 | |
| *** gtema has quit IRC | 08:14 | |
| *** gtema has joined #zuul | 08:14 | |
| *** jamesmcarthur has quit IRC | 08:20 | |
| *** gtema has quit IRC | 08:24 | |
| *** jamesmcarthur has joined #zuul | 08:44 | |
| *** panda|pubholiday is now known as panda | 08:49 | |
| openstackgerrit | Felix Schmidt proposed zuul/zuul master: Provide buildset.uuid in /builds API result https://review.opendev.org/674759 | 08:58 |
|---|---|---|
| flaper87 | is there a way to define tenant wide variables? | 08:59 |
| tobiash | flaper87: no, currently not | 09:04 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Fix uneven distribution of jobs https://review.opendev.org/674762 | 09:14 |
| tobiash | corvus: I just noticed that the latest zuul update broke the even distribution of jobs over the executors. This is the fix ^ | 09:15 |
| tobiash | mordred: ^ | 09:16 |
| *** jamesmcarthur has quit IRC | 09:18 | |
| tobiash | appearently my gear refactoring broke it :( | 09:20 |
| flaper87 | tobiash: bummer, thanks for the info | 09:21 |
| *** AJaeger has quit IRC | 09:38 | |
| *** AJaeger has joined #zuul | 09:45 | |
| *** bhavikdbavishi has quit IRC | 09:48 | |
| *** shrutic3 has joined #zuul | 09:51 | |
| shrutic3 | Hello guys, | 09:51 |
| shrutic3 | I have a query regarding zuul | 09:52 |
| shrutic3 | My particular pipeline did not run yesterday night | 09:52 |
| shrutic3 | Can you help me where I can start with? | 09:53 |
| shrutic3 | expecting some help? | 09:55 |
| *** jamesmcarthur has joined #zuul | 09:55 | |
| *** jamesmcarthur has quit IRC | 09:59 | |
| *** mgoddard has quit IRC | 10:03 | |
| *** mgoddard has joined #zuul | 10:07 | |
| shrutic3 | ? | 10:09 |
| shrutic3 | ? | 10:09 |
| shrutic3 | ? | 10:09 |
| shrutic3 | ? | 10:09 |
| shrutic3 | HELP MEEEEEEEEEEEEEEEEEE!!!!!!!!!!!!!!111 | 10:09 |
| *** rfolco|ruck is now known as rfolco|doctor | 10:33 | |
| *** daniel3 is now known as daniel2 | 10:36 | |
| *** shrutic3 has quit IRC | 10:49 | |
| *** jamesmcarthur has joined #zuul | 10:55 | |
| *** jamesmcarthur has quit IRC | 11:00 | |
| *** jank has quit IRC | 11:08 | |
| *** jamesmcarthur has joined #zuul | 11:31 | |
| *** jank has joined #zuul | 11:32 | |
| *** jamesmcarthur has quit IRC | 11:35 | |
| *** jamesmcarthur has joined #zuul | 11:35 | |
| *** jank has quit IRC | 11:42 | |
| *** rlandy has joined #zuul | 11:46 | |
| *** panda is now known as panda|lunch | 11:49 | |
| *** jank has joined #zuul | 11:58 | |
| *** zbr is now known as zbr|lunch | 12:01 | |
| *** jamesmcarthur has quit IRC | 12:20 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Fix uneven distribution of jobs https://review.opendev.org/674762 | 12:27 |
| *** zbr|lunch is now known as zbr | 12:38 | |
| *** sshnaidm|afk is now known as sshnaidm | 12:58 | |
| *** jank has quit IRC | 13:01 | |
| *** bjackman has quit IRC | 13:02 | |
| *** jank has joined #zuul | 13:06 | |
| *** panda|lunch is now known as panda | 13:12 | |
| *** jamesmcarthur has joined #zuul | 13:17 | |
| tobiash | corvus, mordred: I rolled this out into production and it works ^ | 13:30 |
| *** jeliu_ has joined #zuul | 13:33 | |
| *** olaph has quit IRC | 13:34 | |
| pabelanger | does anybody know sigmavirus24 ? re: https://github.com/sigmavirus24/github3.py/issues/951 would like to see if we can get a new release | 13:36 |
| pabelanger | of github3.py | 13:36 |
| pabelanger | https://review.opendev.org/673304/ switch all of our jobs in zuul tenant to ansible 2.8, if people would like to review | 13:37 |
| *** jank has quit IRC | 14:06 | |
| *** altlogbot_1 has quit IRC | 14:12 | |
| *** altlogbot_3 has joined #zuul | 14:15 | |
| *** rfolco|doctor is now known as rfolco | 14:19 | |
| *** rfolco is now known as rfolco|ruck | 14:19 | |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 14:28 |
| corvus | tobiash: ah thanks, i was starting to suspect something was wrong with that | 14:30 |
| corvus | flaper87: there are *site* wide variables | 14:30 |
| *** tributarian has joined #zuul | 14:30 | |
| corvus | flaper87: https://zuul-ci.org/docs/zuul/user/jobs.html#user-jobs-sitewide-variables | 14:30 |
| *** michael-beaver has joined #zuul | 14:30 | |
| corvus | flaper87: they apply to all tenants | 14:31 |
| *** AshBullock has joined #zuul | 14:37 | |
| tristanC | jeliu_: fwiw, you can find some convenient jsonpath selector in these playbooks: https://softwarefactory-project.io/cgit/config/tree/playbooks/openshift/deploy-project.yaml#n47 | 14:37 |
| tristanC | jeliu_: and how to use add_host with the resulting pod to run assertion in containers: https://softwarefactory-project.io/cgit/software-factory/sf-ci/tree/roles/health-check/openshift/tasks/demo_project_zuul_configuration.yml#n44 | 14:38 |
| clarkb | pabelanger: jlk can help with that iirc | 14:40 |
| AshBullock | Hey all, does anyone know if the "pipeline.trigger.<github source>.comment" that can be used as a trigger to a job is made available to the job itself as a var? | 14:41 |
| clarkb | corvus: tobiash does it make senseto drop the default workerclass to force us to supply what we need? | 14:41 |
| tobiash | clarkb: the executor should be the only special case afaik | 14:45 |
| jeliu_ | tristanC: thanks! I'll check it out! | 14:45 |
| *** themroc has quit IRC | 14:58 | |
| *** andreykurilin_ has joined #zuul | 14:58 | |
| *** AshBullock has quit IRC | 15:00 | |
| corvus | AshBullock: i don't believe that either the regex or the match is made available, but i believe the whole commit message is: https://zuul-ci.org/docs/zuul/user/jobs.html#var-zuul.message | 15:00 |
| andreykurilin_ | hi folks! can someone help me to find more info about https://zuul-ci.org/docs/nodepool/configuration.html#attr-diskimages.elements ? | 15:03 |
| corvus | andreykurilin_: yes! https://docs.openstack.org/diskimage-builder/latest/elements.html | 15:06 |
| corvus | we should link to that | 15:06 |
| corvus | also https://docs.openstack.org/diskimage-builder/latest/developer/developing_elements.html | 15:06 |
| corvus | (the first describes built-in elements, the second how to make your own) | 15:07 |
| andreykurilin_ | corvus: great. thanks! | 15:08 |
| andreykurilin_ | corvus: one more question. Am I right that for adding python3 package to centos7 image, I need to create 2 elements ( one for `centos-release-scl` package in package-installs.json and another with actual python packages) or right a fully custom element with one script inside install.d to execute several operations with yum ? | 15:16 |
| pabelanger | both should work | 15:20 |
| *** jank has joined #zuul | 15:21 | |
| clarkb | dib also has support for installing packages via the install-packages element or whatever it is called (so you may be able to list those two packages in one of those config files and that be sufficient) | 15:22 |
| openstackgerrit | Merged zuul/zuul master: Fix uneven distribution of jobs https://review.opendev.org/674762 | 15:23 |
| andreykurilin_ | thanks! | 15:23 |
| *** igordc has joined #zuul | 15:48 | |
| openstackgerrit | Paul Belanger proposed zuul/zuul master: WIP: Support Ansible 2.9 https://review.opendev.org/674854 | 15:49 |
| pabelanger | tobiash: with beta1 of ansible less then a month away, figure we can start some testing with zuul^ | 15:49 |
| pabelanger | beta1 2.9 that is | 15:49 |
| tobiash | pabelanger: cool :) | 15:49 |
| pabelanger | we also likely want to start talking about supporting collections with zuul. As that is the new hotness for ansible | 15:50 |
| clarkb | Zuul's zuul tenant on opendev has been updated to default to ansible 2.8 (I think we actually have to wait for a puppet pulse to update it so it may not quite have ahppened yet) | 15:53 |
| pabelanger | ++ | 15:53 |
| fungi | pabelanger: do you have a reference for a nice summary of what collections is? | 15:54 |
| fungi | i keep seeing folks mention it | 15:54 |
| fungi | closest i see is https://galaxy.ansible.com/docs/contributing/creating_collections.html | 15:55 |
| fungi | which i guess isn't too bad. couple sentences summarizing followed by a technical specification of the format | 15:55 |
| fungi | there had previously been discussion of zuul being able to consume roles directly from galaxy, though i don't recall if that was theoretical or had been roughed in already | 15:57 |
| fungi | and i'm curious how collections relates to (or obsoletes) that | 15:57 |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 16:02 |
| *** AshBullock has joined #zuul | 16:03 | |
| mordred | fungi: not 100% sure- but it seems like collections very well might - and also might make it possible to share playbooks in addition to roles | 16:08 |
| mordred | fungi: so - in some ways it's probably good that we havent' implemented roles-from-galaxy yet - and could potentially just add support for collections | 16:08 |
| mordred | but also with the namespacing it's still likely to be something we need to at least think about even for the zuul-roles case | 16:09 |
| mordred | and I don't thnik anyone has done that thinking yet | 16:09 |
| *** mattw4 has joined #zuul | 16:09 | |
| *** AshBullock has quit IRC | 16:09 | |
| fungi | so in theory job playbooks could be packaged as collections? | 16:10 |
| mordred | MAYBE | 16:11 |
| *** igordc has quit IRC | 16:12 | |
| mordred | my understanding is that part of the motivation behind collections is to be able to share things that aren't roles | 16:12 |
| mordred | so MAYBE that makes sense | 16:12 |
| mordred | fungi: I think I've been sort of waiting for the dust to settle and to then figure out what's up with it in general | 16:13 |
| *** mattw4 has quit IRC | 16:19 | |
| *** mattw4 has joined #zuul | 16:20 | |
| pabelanger | fungi: https://docs.ansible.com/ansible/devel/dev_guide/collections_tech_preview.html is another | 16:21 |
| fungi | thanks! | 16:22 |
| pabelanger | one example of a collection, which people here will likey understand, is moving all of the openstack modules from ansible/ansible into an ansible.clouds / ansible.openstack collection. Then the removal of that functionality from ansible/ansible. | 16:24 |
| pabelanger | but as mordred said, everything is a namespace now | 16:24 |
| *** tosky has quit IRC | 16:25 | |
| pabelanger | however, on executor side, they do need to be installed into a specific folder layout, without using - (hyphen) in the path | 16:25 |
| fungi | at least until they relent on the hyphen thing | 16:26 |
| pabelanger | we likely need to dig more into why that is, from googling a little there is a way to load that using importlib, which I _think_ what is being used | 16:27 |
| pabelanger | does anybody else have issues loading SUCCESS logs from https://zuul.opendev.org/t/zuul/status | 16:27 |
| pabelanger | actually, let me check plugins first | 16:27 |
| fungi | Not Found (404) http://zuul.opendev.org/api/tenant/zuul/build/f4193a32a9224fa59c047c95a1a75306 | 16:28 |
| pabelanger | http://zuul.opendev.org/t/zuul/build/a8b62d805eda43ff91dc91fdc4527ef2 | 16:28 |
| pabelanger | I get 404 | 16:28 |
| pabelanger | yah | 16:28 |
| pabelanger | Uncaught (in promise) TypeError: Cannot read property 'log_url' of undefined | 16:28 |
| pabelanger | that is from console in chrome | 16:28 |
| mordred | I get a 404 | 16:29 |
| corvus | yeah, that's a side effect of the way we're overriding the success-url to test out reporting the build page | 16:30 |
| corvus | it'll work once the build finishes and it reports | 16:30 |
| corvus | but before then, the build data haven't been written to the sql database | 16:31 |
| *** jamesmcarthur has quit IRC | 16:31 | |
| corvus | we'll need to figure out what we want to do about that when we throw the switch for reporting the build page... | 16:31 |
| fungi | aha, makes sense | 16:31 |
| fungi | so it's only doing that for builds whose buildset has not reported | 16:31 |
| corvus | do we change how sql reporting happens? (if so, that might have to be a larger structural change -- the one where we move the sql reporter to be a built-in requirement instead of a reporter) | 16:32 |
| corvus | or do we return the log_url until the buildset reports | 16:32 |
| fungi | i think the latter is probably fine | 16:32 |
| corvus | option 1 is the long-term plan i think, but it's a big change, option 2 may be the better stepping stone | 16:33 |
| pabelanger | yah, having access to logs from status UI is helpful, expecially if the buildset takes a long time to report. So, +1 for stepping stone | 16:34 |
| mordred | ++ | 16:37 |
| openstackgerrit | Paul Belanger proposed zuul/zuul master: Remove support for ansible 2.5 https://review.opendev.org/650431 | 16:44 |
| openstackgerrit | Paul Belanger proposed zuul/zuul master: WIP: Remove support for ansible 2.5 https://review.opendev.org/650431 | 16:58 |
| corvus | cool, that shouldn't be a problem when we do this for real, but we won't be able to simulate that right now | 17:00 |
| openstackgerrit | Paul Belanger proposed zuul/zuul master: WIP: Support Ansible 2.9 https://review.opendev.org/674854 | 17:00 |
| openstackgerrit | Sagi Shnaidman proposed zuul/zuul-jobs master: Don't install centos repos on RHEL https://review.opendev.org/674572 | 17:28 |
| *** chandankumar is now known as raukadah | 17:30 | |
| *** sshnaidm is now known as sshnaidm|afk | 17:31 | |
| corvus | fbo: i don't see a release note for the pagure driver; can you retroactively add one? :) | 17:31 |
| corvus | fbo: you can figure out which tag it ended up in and add it to an existing reno file for that tag | 17:32 |
| *** spsurya has quit IRC | 17:52 | |
| *** bhavikdbavishi has joined #zuul | 17:58 | |
| openstackgerrit | Paul Belanger proposed zuul/zuul master: WIP: Remove support for ansible 2.5 https://review.opendev.org/650431 | 18:01 |
| openstackgerrit | Paul Belanger proposed zuul/zuul master: WIP: Support Ansible 2.9 https://review.opendev.org/674854 | 18:01 |
| pabelanger | zuul-maint: jobs in zuul tenant are now using ansible 2.8.0, I don't see any issue. However, we should upgrade ansible to 2.8.3 (on executors), as that is the latest version. I'll look to see why that isn't happening | 18:03 |
| *** igordc has joined #zuul | 18:08 | |
| *** bhavikdbavishi has quit IRC | 18:15 | |
| *** bhavikdbavishi has joined #zuul | 18:15 | |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 18:15 |
| *** andreykurilin_ has quit IRC | 18:17 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: WIP: render console in js with listview https://review.opendev.org/674663 | 18:19 |
| *** igordc has quit IRC | 18:25 | |
| fungi | pabelanger: none of our executors have ever upgraded their ansible environments that i've been able to tell. they're all dated contemporary with when they were first added | 18:26 |
| fungi | i think we're expecting something to periodically run `zuul ansible-manage` and maybe that was never added? | 18:27 |
| pabelanger | fungi: makes sense, for zuul.a.c we explicitly upgrade ansible when each new dot release happens | 18:28 |
| *** jangutter has quit IRC | 18:42 | |
| openstackgerrit | Paul Belanger proposed zuul/zuul master: WIP: Support Ansible 2.9 https://review.opendev.org/674854 | 18:42 |
| *** igordc has joined #zuul | 18:46 | |
| *** bhavikdbavishi has quit IRC | 18:55 | |
| *** jamesmcarthur has joined #zuul | 18:57 | |
| *** jamesmcarthur has quit IRC | 19:08 | |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 19:36 |
| *** jamesmcarthur has joined #zuul | 19:44 | |
| *** jamesmcarthur has quit IRC | 19:54 | |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 20:00 |
| corvus | mordred, clarkb, tristanC: i could use some user testing feedback on https://logs.opendev.org/63/674663/3/check/zuul-build-dashboard/f08b592/npm/html | 20:01 |
| corvus | can you load that up, navigate to a build page and then the "console" tab and let me know what you think? | 20:01 |
| mordred | corvus: I like the trusted lable | 20:02 |
| mordred | and that the pre playbooks are hidden by default | 20:02 |
| clarkb | corvus: my first thought is I would collapse more of the details. Maybe have run open up to plays and tasks but don't expand the task output? | 20:02 |
| clarkb | For example in https://logs.opendev.org/63/674663/3/check/zuul-build-dashboard/f08b592/npm/html/build/2df498812eb54f45bc77a18bd175ee02#console "Run tox without tests" should be collapsed I think | 20:03 |
| clarkb | THat gives you a nice overview of what was done without being overwhelming then you can expand from there to see what changed or why things were skipped etc | 20:03 |
| clarkb | I like the details button for the super indepth data too | 20:04 |
| clarkb | permalinks to those details popups would be a good improvement too | 20:04 |
| mordred | also - maybe drop the word playbook (or something)? I say that because the primary salient info about which playbook is at the end, and on the one I opened most of the playbook names are cutoff with ... | 20:04 |
| mordred | (or a better solution) | 20:05 |
| mordred | corvus: I went to a zuul-build-dashboard job - and while everything is there in the details popup, if I keep scrolling the text it goes away | 20:06 |
| mordred | https://i.imgur.com/pe6UDht.png | 20:07 |
| corvus | how did you get to a zuul-build dashboard job? | 20:07 |
| corvus | they don't run in the openstack tenant | 20:07 |
| mordred | oh - sorry - haha. that is the dashboard from zuul-build-dashboard :) | 20:08 |
| corvus | bstinson: what's the build uud you're looking at? | 20:08 |
| corvus | mordred: ^ | 20:08 |
| corvus | sorry bstinson | 20:08 |
| corvus | ce574...? | 20:08 |
| mordred | corvus: infra-puppet-apply-4-ubuntu-xenial | 20:08 |
| mordred | ce574f85a30344c18f043050502b9f23 | 20:09 |
| corvus | mordred: fwiw, that works fine for me in firefox | 20:09 |
| mordred | awesome | 20:09 |
| corvus | maybe chrome has issues with that much data? | 20:10 |
| corvus | clarkb, mordred: those are all good ideas, thanks; i'll try them out | 20:10 |
| mordred | corvus: maybe? I don't see anything in the console - and the popup loads fine. it's reproducible if I re-navigate to the same one | 20:11 |
| mordred | corvus: also - based on be being confused a second ago - maybe putting job name on the page somewhere- like maybe in the line with the build uuid? sometimes my brainhole forgets what it's doing | 20:12 |
| mordred | yeah - I can scroll to the bottom of the details panel | 20:13 |
| corvus | clarkb, mordred: i also had 2 ideas about reducing the nesting depth: 1) concatenate all the plays together so that a playbook expands to a single list of tasks. 2) don't nest host-tasks under task (do what ara does and repeat the task for each host). without seeing either of those, do you imagine that either one or both of them might be interesting enough to try out? | 20:13 |
| corvus | (or is the current amount of nesting okay / helpful?) | 20:13 |
| corvus | (i will note, we should totally name more of the plays in our jobs :) | 20:14 |
| clarkb | corvus: if you do the first one I Think we still need some way of specifying where the task originated from | 20:14 |
| clarkb | but I do think that might help people undersatnd what the job did better (because it is a mostly linear run through of tasks) | 20:14 |
| *** jamesmcarthur has joined #zuul | 20:14 | |
| corvus | clarkb: it'd still be under a playbook | 20:15 |
| clarkb | oh I see play vs playbook | 20:15 |
| corvus | ya | 20:15 |
| clarkb | ya I think I like that | 20:15 |
| *** jamesmcarthur has quit IRC | 20:15 | |
| clarkb | for 2) I expect most jobs will result in roughly the same result since they are single node? | 20:15 |
| corvus | mordred: that page works for me in chromium too | 20:15 |
| mordred | corvus: the first sounds the most intriguing - but I'm not super sure | 20:16 |
| corvus | clarkb: yeah, for single node jobs the task list would look the same and we would just lose a "redundant" level of nesting (modulo some slight weirdness where we run things on localhost). for multinode jobs, on plays that run on "all", we'll see every task twice. | 20:17 |
| *** jamesmcarthur has joined #zuul | 20:17 | |
| clarkb | corvus: maybe do like stdout does and have hostname: taskname when listing them like that? | 20:18 |
| corvus | clarkb: yeah | 20:18 |
| clarkb | I think ara does that with columnar data? | 20:18 |
| corvus | yep | 20:18 |
| corvus | sounds like both might be worth a look, with priority on the first one. they're both moderate changes so i wanted to at least get a "maybe vs. no way" signal before i did anything. | 20:19 |
| *** jamesmcarthur has quit IRC | 20:19 | |
| *** hwangbo has quit IRC | 20:20 | |
| corvus | also, btw, the expansion behavior is slightly different on a failing build -- it always tries to expand down to the output level of a failed task (regardless of which playbook it's in) | 20:21 |
| clarkb | I think that is good | 20:21 |
| mordred | ++ | 20:22 |
| corvus | i'm going to start with a cleanup and squash of the current patch to make it mergable, then start building on it with experiments for the other things we talked about | 20:22 |
| mordred | corvus: cool | 20:23 |
| clarkb | sounds good | 20:23 |
| mordred | corvus: fwiw - I went and found a failing job to look at it (looks good) and I still have the blank issue int he output | 20:23 |
| mordred | so we probably want to figure that out at some point | 20:23 |
| mordred | I don't have any good ideas yet - I'll keep poking at things in my browser to see what I can learn | 20:24 |
| tristanC | corvus: that's well done! though it seems like a re-implementation of the ara interface, perhaps we could share the react component with the one being developped in the ara-web project? | 20:24 |
| corvus | mordred: yeah, i think you're going to have to take the lead on that at least until we find a way to reproduce it | 20:25 |
| *** jamesmcarthur has joined #zuul | 20:25 | |
| mordred | corvus: yeah | 20:25 |
| corvus | tristanC: well, i wouldn't describe it as a reimplementation -- it's a *lot* simpler than ara and does a lot less :) | 20:25 |
| corvus | the little boxes that say "changed" and "skipped" are totally borrowed from ara's ui, i will admit :) | 20:26 |
| tristanC | begining* of a reimplementation. I mean this serve the same purpose, and it may very well become as feature-full as ara (e.g. search, task details, ...) | 20:27 |
| corvus | if there's something we can share, that's great | 20:27 |
| tristanC | iirc there was discussion about ara being published as a npm module, then zuul could yarn import ara and have access to its component. | 20:28 |
| corvus | i don't intend to do much more than what you see here, so if that happens, and it works, we can totally replace it | 20:29 |
| corvus | right now, this is little more than just translating a nested json data structure to a nested patternfly ListView. it's pretty basic. | 20:29 |
| tristanC | corvus: well, it's great you were able to implement that interface, wfm as it is now | 20:34 |
| corvus | tristanC: react is getting less and less scary :) | 20:34 |
| corvus | in working on this, i was able to see why it's useful to break up things into small components -- it turns out react doesn't even call the constructor on most of the objects in the tree because they're nested and hidden. | 20:36 |
| corvus | (so the current version has a react component for each level) | 20:36 |
| corvus | playbook/play/task/hosttask/taskoutput | 20:37 |
| tristanC | it's quite an efficient system indeed. One thing zuul-web could leverage more are PureComponent | 20:37 |
| corvus | tristanC: ooooh, yes | 20:38 |
| corvus | If your React component’s render() function renders the same result given the same props and state, you can use React.PureComponent for a performance boost in some cases. | 20:38 |
| corvus | that's us | 20:38 |
| corvus | mordred, tristanC: do either of you have a suggestion about how to fix this lint error: http://paste.openstack.org/show/755588/ | 20:44 |
| clarkb | corvus: does js(x) use the _ convention for denoting unused vars? | 20:44 |
| mordred | corvus: I think you can do the _ trick - so forEach([_, host]) | 20:45 |
| corvus | actually... https://eslint.org/docs/rules/no-unused-vars | 20:45 |
| corvus | it looks like it's just [, host] | 20:45 |
| mordred | neat! | 20:45 |
| corvus | yeah that works. yay for the linter actually having useful docs | 20:46 |
| mordred | yay! | 20:46 |
| mordred | eslint is not bad as linters go | 20:46 |
| clarkb | the executor gearman worker thing had me thinking about how we might handle that. There are apparently some tools that work at the intersection of coverage and linting and complain about code that is unexecuted | 20:47 |
| clarkb | not sure it is worth it for this one case though | 20:47 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Render console in js https://review.opendev.org/674368 | 20:49 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Experiment: Don't auto-expand past play https://review.opendev.org/674931 | 20:55 |
| corvus | clarkb: your first idea ^ | 20:55 |
| openstackgerrit | Jeff Liu proposed zuul/zuul-operator master: WIP: Add zuul-operator-functional-openshift job https://review.opendev.org/674355 | 20:56 |
| corvus | mordred: i'm having trouble seeing the 'playbook name is cut off' problem you described | 20:58 |
| corvus | oh | 20:59 |
| corvus | that's reproducible in chromium | 20:59 |
| corvus | it does not happen in ff | 20:59 |
| mordred | interesting | 20:59 |
| corvus | in ff, they just keep taking up more space if they need to | 21:00 |
| mordred | it seems like a hard 2-column in chrome | 21:00 |
| mordred | well - if I shrink the browser width, I get teh whole plabyook name and trusted on the next line | 21:01 |
| mordred | but I also get the menu up into the hamburger when I do that | 21:01 |
| corvus | ff will just push the additional info (eg, 'trusted') to the right if necessary | 21:03 |
| mordred | browsers are weird | 21:04 |
| corvus | if there is no additional info, then more space is available | 21:04 |
| corvus | so maybe we can reduce the space allocated to additional info | 21:04 |
| *** jeliu_ has quit IRC | 21:05 | |
| corvus | yeah, i think we can give it 25% and it'll work most of the time | 21:06 |
| *** igordc has quit IRC | 21:12 | |
| *** igordc has joined #zuul | 21:31 | |
| *** jamesmcarthur has quit IRC | 21:39 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Use description for playbook name https://review.opendev.org/674938 | 21:47 |
| corvus | mordred: the 25% solution didn't pan out; that's a different approach ^ | 21:48 |
| corvus | weirdly, according to the patternfly docs, the header is supposed to be *above* the description, but that doesn't seem to happen in either ff or chromium. i think it'll work either way. | 21:48 |
| corvus | that doesn't strictly depend on clarkb's change, but i thought it was easier to evaluate stacked on it, so i put it there for now | 21:49 |
| corvus | ah, i found it... it's the stacked property | 21:54 |
| clarkb | corvus: the don't expand past play change renders great. I do notice that the console tab takes a second or three to show up after pulling up a build | 22:00 |
| clarkb | I think maybe because the constructor parses the json in one go? | 22:00 |
| clarkb | and if so I guess that will be fixed by your refactoring | 22:00 |
| corvus | clarkb: actually i think you're seeing the async fetch of the console log | 22:06 |
| clarkb | ah | 22:11 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Stack listview in build console https://review.opendev.org/674942 | 22:14 |
| corvus | that's a further visual tweak to the fix for mordred's playbook length issue | 22:14 |
| *** jamesmcarthur has joined #zuul | 22:20 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Combine all plays on build page https://review.opendev.org/674945 | 22:24 |
| corvus | clarkb: that effect is going to get slightly more pronounced when we start randomly storing 1/6 of our logs in france :) | 22:26 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Flatten host tasks into tasks on build page https://review.opendev.org/674951 | 22:49 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Remove 'Details' button from build console https://review.opendev.org/674953 | 22:56 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Move hostname to additionalInfo on build page https://review.opendev.org/674954 | 22:56 |
| *** igordc has quit IRC | 22:56 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Don't auto-expand past play on build page https://review.opendev.org/674931 | 22:57 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Use description for playbook name https://review.opendev.org/674938 | 22:58 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Stack listview in build console https://review.opendev.org/674942 | 22:58 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Combine all plays on build page https://review.opendev.org/674945 | 22:58 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Flatten host tasks into tasks on build page https://review.opendev.org/674951 | 22:58 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Remove 'Details' button from build console https://review.opendev.org/674953 | 22:58 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Move hostname to additionalInfo on build page https://review.opendev.org/674954 | 22:59 |
| *** rlandy has quit IRC | 22:59 | |
| *** igordc has joined #zuul | 23:20 | |
| *** threestrands has joined #zuul | 23:24 | |
| *** mattw4 has quit IRC | 23:42 | |
| *** jamesmcarthur has quit IRC | 23:42 | |
| *** mattw4 has joined #zuul | 23:42 | |
| *** jamesmcarthur has joined #zuul | 23:47 | |
| openstackgerrit | Clark Boylan proposed zuul/zuul master: Update ara to >=0.16.5 to support ansible 2.8 https://review.opendev.org/674958 | 23:50 |
| *** jamesmcarthur_ has joined #zuul | 23:54 | |
| *** jamesmcarthur has quit IRC | 23:54 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
