| *** bhavikdbavishi has joined #zuul | 01:45 | |
| *** bhavikdbavishi has quit IRC | 01:52 | |
| *** ianychoi has quit IRC | 02:25 | |
| *** ianychoi has joined #zuul | 02:26 | |
| *** homeski_ has joined #zuul | 03:38 | |
| *** homeski_ is now known as homeski | 03:41 | |
| homeski | I installed zuul from source, and everything works now except the builds API ... | 03:49 |
|---|---|---|
| homeski | I can `curl myhost.com/api/tenant/my-tenant/info` and get a JSON response back | 03:50 |
| homeski | but `curl myhost.com/api/tenant/my-tenant/builds` returns a 404 | 03:50 |
| homeski | `404: Tenant homero does not exist` | 03:51 |
| homeski | ** sorry `404: Tenant my-tenant does not exist` | 03:51 |
| homeski | `curl myhost.com/api/tenant/my-tenant/jobs` returns successfully | 03:52 |
| homeski | I've tried creating multiple tenants, changing names ... but for whatever reason `.../builds` always returns a 404 and "tenant does not exist" | 03:53 |
| homeski | any ideas on what is going wrong? | 03:53 |
| homeski | everything seems to work (labels, jobs, console-stream, info, etc) besides`/builds` | 03:56 |
| tristanC | homeski: iirc the builds api is only available when there is an sql connection, did you configured one? | 03:58 |
| homeski | nope... | 03:59 |
| homeski | I don't see anything in docs specifically about that | 04:00 |
| homeski | any tips on setting that up? | 04:00 |
| homeski | just setup a `[connection sql]` in zuul.conf, then restart services? | 04:02 |
| tristanC | homeski: yes, such as: https://opendev.org/zuul/zuul/src/branch/master/doc/source/admin/examples/etc_zuul/zuul.conf#L28 | 04:04 |
| homeski | does a reporter need to be configured as well? as talked about here: https://zuul-ci.org/docs/zuul/admin/drivers/sql.html | 04:05 |
| tristanC | homeski: yes, zuul doesn't store build data automatically, pipelines need the sql connection in succes/failure reporters list | 04:07 |
| tristanC | such as: https://opendev.org/zuul/zuul/src/branch/master/doc/source/admin/examples/zuul-config/zuul.d/pipelines.yaml#L20 | 04:08 |
| homeski | I see now. Thanks for the info | 04:09 |
| tristanC | homeski: you're welcome :) | 04:09 |
| *** bhavikdbavishi has joined #zuul | 04:26 | |
| *** sshnaidm|off has quit IRC | 05:01 | |
| openstackgerrit | Joshua Hesketh proposed zuul/zuul master: Expose date time as facts https://review.opendev.org/664674 | 05:26 |
| openstackgerrit | Joshua Hesketh proposed zuul/zuul master: Expose date time as facts https://review.opendev.org/664674 | 06:10 |
| *** saneax has joined #zuul | 06:18 | |
| *** gtema has joined #zuul | 06:56 | |
| *** jpena|off is now known as jpena | 07:52 | |
| openstackgerrit | Fabien Boucher proposed zuul/zuul master: Pagure driver - https://pagure.io/pagure/ https://review.opendev.org/604404 | 07:58 |
| *** sshnaidm has joined #zuul | 08:08 | |
| *** hashar has joined #zuul | 08:11 | |
| *** sshnaidm has quit IRC | 08:14 | |
| *** sshnaidm has joined #zuul | 08:17 | |
| *** flepied has joined #zuul | 08:47 | |
| *** pcaruana has joined #zuul | 08:52 | |
| *** sileht has joined #zuul | 09:24 | |
| tobias-urdin | there is a 3.8.2 release on pypi but can't see a 3.8.2 git tag? | 09:42 |
| *** gtema has quit IRC | 10:19 | |
| *** gtema has joined #zuul | 10:19 | |
| openstackgerrit | Guillaume Vincent proposed zuul/zuul master: Return store object instead of a function https://review.opendev.org/665642 | 10:59 |
| *** jpena is now known as jpena|lunch | 11:15 | |
| *** panda is now known as panda|lunch | 11:23 | |
| *** hashar has quit IRC | 11:27 | |
| fungi | tobias-urdin: where do you see a 3.8.2 release on pypi? | 11:49 |
| fungi | https://pypi.org/project/zuul/ says zuul 3.8.1 | 11:50 |
| fungi | (which corresponds with the most recent tag) | 11:50 |
| tobias-urdin | ah sry, missed pre-release so it's probably from master | 11:53 |
| fungi | yeah, those are auto-numbered dev snapshots of the master branch state | 12:00 |
| *** rfolco has joined #zuul | 12:01 | |
| *** rlandy has joined #zuul | 12:14 | |
| *** rlandy is now known as rlandy|ruck | 12:15 | |
| *** panda|lunch is now known as panda | 12:25 | |
| *** rlandy|ruck is now known as rlandy|ruck|mtg | 12:29 | |
| *** hashar has joined #zuul | 12:38 | |
| *** lennyb has quit IRC | 12:40 | |
| *** jpena|lunch is now known as jpena | 12:49 | |
| *** rlandy|ruck|mtg is now known as rlandy|ruck | 12:54 | |
| *** armstrongs has joined #zuul | 12:58 | |
| armstrongs | hi im trying to setup ara 1.0.1 with zuul but it doesn't seem that ara-wsgi-sqlite binary is packaged with it. Is it not ready to use with ara-wsgi-sqlite? | 13:00 |
| armstrongs | I had it working on 0.16.4 | 13:00 |
| *** gtema has quit IRC | 13:09 | |
| fungi | armstrongs: we've also pinned our installation to the last release before 1.0.0, as far as i know there's no middleware or static deployment method for ara 1.x | 13:10 |
| fungi | at least not yet, but i think dmsimard was working on something | 13:10 |
| armstrongs | ah ok thanks for the info | 13:10 |
| *** lennyb has joined #zuul | 13:13 | |
| *** pcaruana has quit IRC | 13:26 | |
| *** pcaruana|afk| has joined #zuul | 13:26 | |
| *** bhavikdbavishi has quit IRC | 13:33 | |
| openstackgerrit | Mark Meyer proposed zuul/zuul master: Extend event reporting https://review.opendev.org/662134 | 13:40 |
| *** pcaruana has joined #zuul | 13:48 | |
| *** pcaruana|afk| has quit IRC | 13:51 | |
| dmsimard | armstrongs: we have a WIP patch for implementing a similar sqlite middleware for 1.0 | 13:59 |
| dmsimard | so it's in progress :) | 14:01 |
| *** hashar_ has joined #zuul | 14:03 | |
| *** hashar_ has quit IRC | 14:03 | |
| armstrongs | dismard: sounds great looking forward to it :) | 14:07 |
| openstackgerrit | Guillaume Vincent proposed zuul/zuul master: Return store object instead of a function https://review.opendev.org/665642 | 14:28 |
| *** hashar has quit IRC | 14:29 | |
| *** jamesmcarthur has joined #zuul | 14:44 | |
| *** pcaruana has quit IRC | 14:47 | |
| *** saneax has quit IRC | 14:56 | |
| pabelanger | does autohold work with regex? didn't see anything in docs, figured I ask before diving into code | 14:57 |
| pabelanger | doesn't look like it for job name | 15:02 |
| pabelanger | tobiash: I look a quick look at reviews last night, looks pretty sane to me. I'm not sure how to best test this however, have any ideas? | 15:07 |
| pabelanger | I was thinking we could run a zuul-test instance setup to github, with new github app maybe on gtest org that opendev has | 15:08 |
| corvus | pabelanger: the unit test coverage gives me enough confidence to restart the opendev system with it. but perhaps tobiash has already run it on his? | 15:11 |
| Shrews | pabelanger: i don't think there is any regex support on any of the autohold fields | 15:11 |
| fbo | corvus: Hi, should we retry a W+1 on the Pagure Driver patch https://review.opendev.org/604404/ ? | 15:12 |
| pabelanger | corvus: okay, cool. That works for me, if you are good :) | 15:12 |
| pabelanger | Shrews: I _think_ refs can | 15:12 |
| *** hashar has joined #zuul | 15:13 | |
| Shrews | pabelanger: oh, you're right. there is a re.match() on ref | 15:13 |
| pabelanger | corvus: tobiash: just checking moring ERRORs in scheduler, noticed http://paste.openstack.org/show/753097/ for the first time. | 15:14 |
| pabelanger | I also noticed 2019-06-17 14:57:14,834 ERROR zuul.GithubEventConnector: [delivery: d9af4484-910e-11e9-985f-7d61fe3bb106] Failed handling status event; remote said retry after None,will retry attempt 5/5 in 60.0 seconds :( still not long enough wait time | 15:16 |
| corvus | pabelanger: was that in a sytem running with parallel support? | 15:18 |
| pabelanger | corvus: no, just the retry patch | 15:20 |
| *** jamesmcarthur_ has joined #zuul | 15:21 | |
| *** jamesmcarthur has quit IRC | 15:25 | |
| tobiash | corvus: I'm not running this yet because I first need to reimplement the parallel thing on top of it | 15:32 |
| tobiash | But I extensivly tested it using the github debug tooling that creates a standalone githubconnection object | 15:32 |
| corvus | tobiash: should we wait until you're done with that to land changes and test? | 15:38 |
| *** pcaruana has joined #zuul | 15:40 | |
| Shrews | mhu: sorry for misspelling your name in that email :) | 15:42 |
| mhu | Shrews, no worries, my girlfriend is american and that's how her side of the family calls me :) | 15:43 |
| openstackgerrit | Merged zuul/nodepool master: static: enable using a single host with different user or port https://review.opendev.org/659209 | 15:45 |
| tobiash | corvus: I think it should work and I can wait with my next update until I implemented it. So it's your choice. I think it will have quite some benefit for upstream given the current problems | 15:46 |
| *** ianychoi_ has joined #zuul | 15:48 | |
| corvus | fbo: done! :) | 15:50 |
| *** ianychoi has quit IRC | 15:52 | |
| *** gtema has joined #zuul | 15:55 | |
| corvus | clarkb, fungi: https://review.opendev.org/657102 may not be on your radar but could use a look | 16:01 |
| fungi | thanks for the heads up, checking it out now | 16:02 |
| *** mattw4 has joined #zuul | 16:03 | |
| fungi | that looks quite useful, agreed! | 16:08 |
| fbo | corvus: thanks hopefuly it will pass the gate this time :) | 16:10 |
| *** michael-beaver has joined #zuul | 16:22 | |
| *** panda is now known as panda|off | 16:30 | |
| *** igordc has joined #zuul | 16:44 | |
| openstackgerrit | Mark Meyer proposed zuul/zuul master: Extend event reporting https://review.opendev.org/662134 | 16:45 |
| *** zbr|ooo has quit IRC | 17:04 | |
| *** zbr has joined #zuul | 17:05 | |
| *** zbr is now known as zbr|ruck | 17:06 | |
| *** hashar has quit IRC | 17:13 | |
| *** jpena is now known as jpena|off | 17:14 | |
| pabelanger | tobiash: so, maybe terrible idea, but it would be neat if zuul could some how figure out the diff when a PR was created, to when zuul got the event, then emit that into statsd. There is some sort of outage going on at https://www.githubstatus.com/, and figure tracking that in grafana might be helpful to our users | 17:17 |
| pabelanger | right now, events look to be 1hr delayed | 17:17 |
| fungi | pabelanger: could it be the gh event processing queue in zuul which is backed up? | 17:21 |
| pabelanger | fungi: no, it is basically zero right now | 17:23 |
| pabelanger | events from GH just delayed today | 17:23 |
| *** sshnaidm is now known as sshnaidm|afk | 17:28 | |
| fungi | ahh | 17:31 |
| fungi | so something queuing on the gh end of things | 17:31 |
| *** chandankumar is now known as raukadah | 17:33 | |
| *** gtema has quit IRC | 17:37 | |
| pabelanger | yah, some sort of backup today | 17:43 |
| mgoddard | hello | 17:46 |
| corvus | that github queues events like that is interesting and useful knowledge | 17:46 |
| mgoddard | having an issue with some zuul config, trying to skip a job on a branch | 17:46 |
| mgoddard | https://opendev.org/openstack/kolla/src/branch/master/.zuul.d/centos.yaml#L38 | 17:46 |
| mgoddard | I've seen that pattern used elsewhere, but that periodic job still seems to run on stable/ocata | 17:47 |
| mgoddard | http://zuul.openstack.org/builds?job_name=kolla-publish-centos-source&branch=stable%2Focata | 17:47 |
| fungi | mgoddard: could it be because of https://opendev.org/openstack/kolla/src/branch/stable/ocata/.zuul.yaml#L49 | 17:50 |
| mgoddard | fungi: good question | 17:51 |
| fungi | generally you should either control jobs in branched zuul configuration or only in master, but having them defined in both places gets confusing | 17:51 |
| fungi | i can't remember whether explicit vs implicit branch filters take precedence | 17:52 |
| corvus | neither, they both win :) | 17:52 |
| fungi | aha. so as long as at least one of them says it runs, then it runs | 17:52 |
| corvus | yep | 17:52 |
| corvus | between the two, i'd recommend controlling things like that in the branches themselves rather than master | 17:53 |
| corvus | so i'd drop https://opendev.org/openstack/kolla/src/branch/master/.zuul.d/centos.yaml#L35-L40 as well as https://opendev.org/openstack/kolla/src/branch/stable/ocata/.zuul.yaml#L44 | 17:54 |
| mgoddard | so I could remove the filter from master and add it to stable/ocata | 17:54 |
| corvus | nothing says "don't run this job" like not having anything that says "run this job" :) | 17:54 |
| corvus | mgoddard: no need for a filter on stable/ocata | 17:54 |
| corvus | mgoddard: as long as nothing on stable/ocata says to run the job, then it won't run | 17:54 |
| *** hashar has joined #zuul | 17:56 | |
| mgoddard | I think my mental model of zuul config is incomplete | 17:56 |
| *** jamesmcarthur_ has quit IRC | 17:57 | |
| mgoddard | we do want to run periodic jobs for pike,queens,rocky,stein,master | 17:57 |
| corvus | mgoddard: i'd recommend adding the jobs you want to .zuul.yaml (or equivalent) on each of those branches | 17:57 |
| mgoddard | ideally we'd not touch the zuul config when we branch, but is that going to be a problem? | 17:57 |
| fungi | normally it's designed so that when you branch your old branches keep running what they were running at the time they were branched, unless you update the branch | 17:58 |
| mgoddard | right | 17:58 |
| fungi | so in theory the kolla-publish-centos-source job was running for master when stable/ocata was branched from it | 17:58 |
| mgoddard | corvus: you said to remove https://opendev.org/openstack/kolla/src/branch/master/.zuul.d/centos.yaml#L35-L40 Is that not going to break our periodic jobs? | 17:59 |
| fungi | if you want stable/ocata to stop running kolla-publish-centos-source in its periodic pipeline after branching, then you would edit the .zuul.yaml file in the stable/ocata branch removing the kolla-publish-centos-source job from its periodic pipeline definition | 17:59 |
| mgoddard | ok. I'll give that a shot | 18:00 |
| mgoddard | and maybe read the zuul docs again :) | 18:00 |
| corvus | mgoddard: since stable/pike still has the job listed (just like stable/ocata), it will continue running the job | 18:00 |
| fungi | but for stable/pike of course | 18:00 |
| corvus | right | 18:00 |
| mgoddard | great, thanks for the help | 18:01 |
| fungi | the branch the job should operate on is inferred from the zuul config if that branch has zuul configuration, otherwise the configuration on the master branch is used. jobs defined on a branch assume they are scoped to that branch unless you override that via further configuration | 18:01 |
| corvus | mgoddard: this is probably the most relevant bit of docs if you want that refresher: https://zuul-ci.org/docs/zuul/user/config.html#attr-job.branches | 18:02 |
| *** rlandy|ruck is now known as rlandy|roverish | 18:02 | |
| corvus | (especially the part at the end) | 18:02 |
| mgoddard | that makes sense now you say it. I think there was a mention of branchless config at some point that confused me, but google suggests thats more for projects without branches (of course) | 18:03 |
| mgoddard | that does look helpful, I'll give it a read. Thanks | 18:03 |
| *** mattw4 has quit IRC | 18:07 | |
| *** mattw4 has joined #zuul | 18:07 | |
| *** mattw4 has quit IRC | 18:12 | |
| *** gtema has joined #zuul | 18:36 | |
| *** gtema has quit IRC | 18:42 | |
| *** pcaruana has quit IRC | 18:44 | |
| *** jamesmcarthur has joined #zuul | 18:51 | |
| *** hashar has quit IRC | 18:53 | |
| *** pcaruana has joined #zuul | 19:17 | |
| *** pcaruana has quit IRC | 19:27 | |
| *** rlandy|roverish is now known as rlandy|roverish| | 19:28 | |
| *** rlandy|roverish| is now known as rlandy|rover|brb | 19:29 | |
| *** jamesmcarthur has quit IRC | 19:42 | |
| *** armstrongs has quit IRC | 19:44 | |
| *** rlandy|rover|brb is now known as rlandy|rover | 19:50 | |
| *** mattw4 has joined #zuul | 19:50 | |
| *** jamesmcarthur has joined #zuul | 20:15 | |
| corvus | tobiash: question on 665505 | 20:28 |
| *** jamesmcarthur has quit IRC | 20:28 | |
| tobiash | corvus: responded | 20:33 |
| tobiash | I guess I should just remove that comment | 20:33 |
| corvus | tobiash: okay, if 403 always has the needed header i agree it's okay, and it's only a confusing comment :) | 20:34 |
| corvus | tobiash: probably worth removing the comment though, considering how confusing it could be | 20:35 |
| tobiash | yeah, I'll remove it | 20:35 |
| tobiash | just a sec | 20:35 |
| corvus | tobiash: that whole stack looks great :) | 20:36 |
| tobiash | thanks :) | 20:36 |
| *** jamesmcarthur has joined #zuul | 20:37 | |
| tobiash | shall I push the header removal as followup or change 665505? | 20:37 |
| corvus | tobiash: change 505, i'll re+2 | 20:37 |
| tobiash | k | 20:37 |
| *** hashar has joined #zuul | 20:38 | |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Rework GitHub rate limit handling https://review.opendev.org/665505 | 20:39 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Switch getPullBySha to using the search api https://review.opendev.org/665469 | 20:40 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Add central retry handling for github requests https://review.opendev.org/665506 | 20:40 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Cleanup specialized retry in _process_event https://review.opendev.org/665507 | 20:40 |
| openstackgerrit | Tobias Henkel proposed zuul/zuul master: Eliminate two github requests per _updateChange https://review.opendev.org/665511 | 20:40 |
| *** hashar has quit IRC | 20:40 | |
| *** hashar has joined #zuul | 20:40 | |
| fungi | aprice just noticed this in her twitter feed: https://twitter.com/gchenuet/status/1140719769349476354 | 20:40 |
| fungi | guilhermesp: that's really amazing!!! | 20:41 |
| *** jamesmcarthur has quit IRC | 20:41 | |
| corvus | nice :) | 20:42 |
| *** jamesmcarthur has joined #zuul | 20:42 | |
| fungi | guilhermesp: er, sorry, tab completion failed me | 20:42 |
| tobiash | cool :) | 20:42 |
| fungi | >4k pipelines and >40k virtual machines? | 20:44 |
| fungi | no idea what size deployments the rest of you have, but that's waaay beyond the size of the one i help run ;) | 20:45 |
| fungi | i wonder if they've seen our scheduler memory leak | 20:45 |
| *** jamesmcarthur has quit IRC | 21:01 | |
| *** mattw4 has quit IRC | 21:04 | |
| *** mattw4 has joined #zuul | 21:04 | |
| *** jamesmcarthur has joined #zuul | 21:15 | |
| *** rfolco has quit IRC | 21:16 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Pagure driver - https://pagure.io/pagure/ https://review.opendev.org/604404 | 21:18 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Make git repo leak check advisory in TestExecutor https://review.opendev.org/665764 | 21:18 |
| *** mattw4 has quit IRC | 21:22 | |
| *** mattw4 has joined #zuul | 21:22 | |
| openstackgerrit | Merged zuul/zuul master: Show retry attempts in UI https://review.opendev.org/660960 | 21:30 |
| *** hashar has quit IRC | 21:45 | |
| *** sgw has joined #zuul | 22:07 | |
| *** ChanServ has quit IRC | 22:09 | |
| openstackgerrit | Merged zuul/zuul master: Report job depedencies via MQTT reporter https://review.opendev.org/660940 | 22:09 |
| *** ChanServ has joined #zuul | 22:14 | |
| *** niven.freenode.net sets mode: +o ChanServ | 22:14 | |
| *** shachar has quit IRC | 22:17 | |
| *** shachar has joined #zuul | 22:20 | |
| *** jamesmcarthur has quit IRC | 22:22 | |
| *** jamesmcarthur has joined #zuul | 22:22 | |
| clarkb | fungi ours does 20k jobs ish on 20kish VMs | 22:34 |
| clarkb | so we run more jobs on fewer machines | 22:35 |
| fungi | yeah, they didn't say how many jobs (unless that's what they meant by "pipelines") | 22:35 |
| clarkb | yes its 4k pipeline jobs as I read it | 22:36 |
| clarkb | not 4k pipelines | 22:36 |
| fungi | i took "4k+ pipelines" to mean they had configured >4k different actual pipelines in zuul | 22:36 |
| fungi | so maybe i just assumed a more accurate phrasing than they meant | 22:36 |
| clarkb | its "4k pipelines jobs" in the tweet | 22:36 |
| fungi | oh, so it is. i guess it's just a plurality error in their phrasing | 22:37 |
| fungi | i read it as "4k+ pipelines, jobs ran on 40k+ vms" | 22:38 |
| fungi | so thought that was 4k pipeline configurations and they had enough quota to run 40k virtual machines concurrently | 22:39 |
| fungi | i guess that's the problem with trying to cram information into sms | 22:40 |
| corvus | it's still big whether it's 1/2 opendev or 2x opendev :) | 22:40 |
| clarkb | Ya just wanted to point out our number is similarly large? | 22:41 |
| clarkb | er :) | 22:41 |
| corvus | also, tomorrow we should see if we can find someone to tweet about how *small* their zuul is :) | 22:41 |
| openstackgerrit | Merged zuul/zuul master: Add more test coverage on using python-path https://review.opendev.org/659812 | 22:42 |
| corvus | (so we don't get typecast as "ci for really big projects") | 22:43 |
| corvus | the nuance i'd like to convey is, "ci that will grow with your project" | 22:44 |
| clarkb | ++ | 22:46 |
| fungi | i'm tempted to set up my own zuul for a very small project of mine... but it's equally tempting to just move it into opendev | 22:50 |
| SpamapS | corvus:my zuul is pretty medium/small | 22:53 |
| corvus | SpamapS: :) | 22:59 |
| *** gfhellma has joined #zuul | 23:03 | |
| *** tjgresha has quit IRC | 23:09 | |
| SpamapS | By what measure do we size Zuul's? | 23:13 |
| SpamapS | We run about 100 jobs per day, and have a pool of 24 AWS EC2 nodes of varying sizes that we spin up on demand. | 23:14 |
| corvus | really good question -- total projects, total jobs, change rate, build rate, total workers, worker rate are all reasonable things i've seen. | 23:15 |
| *** gfhellma has quit IRC | 23:21 | |
| *** rlandy|rover has quit IRC | 23:22 | |
| fungi | sounds like it's time for SpamapS to write a blog post ;) | 23:29 |
| dmsimard | clarkb: 20k jobs or 20k patchsets ? Last time I looked (albeit it was a while ago) there was around ~300k job results per sql reporter | 23:32 |
| dmsimard | (per month) | 23:33 |
| corvus | the number of jobs launched is greater than the number of jobs reported too | 23:34 |
| dmsimard | Because of retries ? | 23:34 |
| corvus | mostly speculation | 23:34 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!