| *** tobiash has quit IRC | 03:11 | |
| *** tobiash has joined #zuul | 03:15 | |
| *** jiapei has quit IRC | 04:12 | |
| *** ianychoi has quit IRC | 04:19 | |
| openstackgerrit | Simon Westphahl proposed openstack-infra/nodepool master: Implement liveness check for static nodes https://review.openstack.org/601513 | 05:44 |
|---|---|---|
| swest | Shrews: thanks for the hint! Using checkReusableNode() looks indeed a little cleaner ^ | 05:47 |
| *** threestrands has joined #zuul | 06:13 | |
| *** pcaruana has joined #zuul | 06:23 | |
| *** nilashishc has joined #zuul | 06:46 | |
| *** hashar has joined #zuul | 07:32 | |
| *** hashar has quit IRC | 07:37 | |
| *** hashar has joined #zuul | 07:37 | |
| *** jpena|off is now known as jpena | 07:44 | |
| mhu | hello, this change has a +2, could it get a +3 please? https://review.openstack.org/#/c/589854/ | 08:30 |
| *** electrofelix has joined #zuul | 08:51 | |
| *** hashar is now known as hasharAway | 09:49 | |
| *** luizbag has joined #zuul | 10:43 | |
| xbend | Hi. Are there any plans/roadmap/blueprint for zuul dashboard and zuul API development? Right now I am interested if there are any plans for extending current API for management purposes like stopping/starting jobs. | 11:25 |
| SotK | ooi, has anyone had success running zuul on Windows at all? | 11:34 |
| mhu | xbend, I've given it a try: https://review.openstack.org/#/c/576907/ it's tenant-scoped and uses JWT | 11:37 |
| mhu | it covers enqueue, dequeue, autohold | 11:37 |
| mhu | it's just a PoC and it needs further discussion | 11:38 |
| xbend | mhu, thanks. Ill take a look. | 11:42 |
| *** jpena is now known as jpena|lunch | 11:44 | |
| *** nilashis_ has joined #zuul | 11:57 | |
| *** nilashishc has quit IRC | 11:57 | |
| *** luizbag has quit IRC | 12:13 | |
| *** rlandy has joined #zuul | 12:34 | |
| *** nilashis_ has quit IRC | 12:42 | |
| *** jpena|lunch is now known as jpena | 12:49 | |
| *** nilashishc has joined #zuul | 12:54 | |
| *** samccann has joined #zuul | 12:56 | |
| *** elyezer has left #zuul | 13:02 | |
| SpamapS | oy.. I think the python:alpine image might be broken | 13:07 |
| SpamapS | they pushed a new one 9 hours ago and now my pbrx image builds can't compile things | 13:07 |
| Shrews | SpamapS: uh oh | 13:08 |
| Shrews | hrm, yep. seems to have broken our pbrx jobs too: http://logs.openstack.org/13/601513/3/check/pbrx-build-container-images/898f732/job-output.txt.gz#_2018-09-12_05_48_30_654192 | 13:09 |
| Shrews | mordred: suggestions? ^^ | 13:10 |
| *** dkranz has joined #zuul | 13:20 | |
| *** threestrands has quit IRC | 13:24 | |
| dmsimard | If zuul-web is in a private network or behind a firewall, I suppose that makes it impossible for github to send webhooks to it ? Is there another way to hook into events ? | 13:32 |
| *** dkranz has quit IRC | 13:43 | |
| *** hasharAway has quit IRC | 13:44 | |
| clarkb | SotK: I think geard in particular requires a select alternative that doesnt run on windows. The server itself likely wont run there. However tobiash has jobs running on windoes test nodes with zuul | 13:58 |
| clarkb | dmsimard: I believe that is the only way for github to tell external applications state has changed | 13:58 |
| rcarrillocruz | dmsimard: http://www.ultrahook.com/ , there are other tools around iirc | 14:00 |
| dmsimard | clarkb: hmm, so I guess we'd need some sort of proxy/buffer ? like github -> proxy -> buffer/bus (not zuul-web) <- zuul-scheduler | 14:00 |
| clarkb | why not just open firewall from github to zuul web? | 14:01 |
| clarkb | I believe github publishes their list of source addrs | 14:01 |
| dmsimard | huh, I didn't realize that: https://api.github.com/meta | 14:01 |
| SotK | clarkb: interesting, do you happen to know which version of windows those nodes are? | 14:07 |
| clarkb | SotK: I dont, but do know it uses a non ssh ansible connection. tobiash should have all that info though | 14:12 |
| *** samccann has quit IRC | 14:22 | |
| *** annabelleB has joined #zuul | 14:22 | |
| *** samccann has joined #zuul | 14:22 | |
| *** nilashishc has quit IRC | 14:26 | |
| *** jiapei has joined #zuul | 14:27 | |
| *** sshnaidm|afk has quit IRC | 14:40 | |
| *** sshnaidm|afk has joined #zuul | 14:46 | |
| *** annabelleB has quit IRC | 14:48 | |
| openstackgerrit | Benoît Bayszczak proposed openstack-infra/zuul master: replace dict.update by a dict merge in zuul_return https://review.openstack.org/602054 | 14:53 |
| mordred | Shrews, SpamapS: oh goodie! | 15:05 |
| SpamapS | Have to guess that alpine pushed a new image since nothing has changed in the docker repo for python | 15:19 |
| SpamapS | https://github.com/docker-library/python | 15:19 |
| SpamapS | what's really really really really annoying is it looks like there's no existing tag for the previous working images | 15:19 |
| SotK | clarkb: thanks, I'll wait until he is around then | 15:20 |
| SpamapS | geard won't work on windows or os x | 15:24 |
| *** nilashishc has joined #zuul | 15:30 | |
| *** bbayszczak has joined #zuul | 15:33 | |
| *** bbayszczak has left #zuul | 15:33 | |
| *** bbayszczak has joined #zuul | 15:35 | |
| bbayszczak | pabelanger: Hi, following the issue I had using zuul_return yesterday, I found something which looks like a bug in the zuul_return behavior. You can find the review here: https://review.openstack.org/#/c/602054/1 | 15:44 |
| *** nilashis_ has joined #zuul | 15:45 | |
| openstackgerrit | Steve Kowalik proposed openstack-infra/zuul master: Correctly show remaining time as unknown https://review.openstack.org/601462 | 15:46 |
| *** nilashi__ has joined #zuul | 15:47 | |
| *** nilashishc has quit IRC | 15:47 | |
| pabelanger | bbayszczak: sure, I can dive more into it shortly, do you mind adding a test case to expose the issue you are seeing? | 15:50 |
| *** nilashis_ has quit IRC | 15:50 | |
| bbayszczak | sure, thanks | 15:54 |
| *** pcaruana has quit IRC | 15:55 | |
| *** xbend has quit IRC | 15:56 | |
| *** hashar has joined #zuul | 16:01 | |
| *** annabelleB has joined #zuul | 16:05 | |
| *** bhavikdbavishi has joined #zuul | 16:11 | |
| *** nilashishc has joined #zuul | 16:12 | |
| *** nilashi__ has quit IRC | 16:15 | |
| *** sshnaidm|afk is now known as sshnaidm|ptg | 16:23 | |
| *** dkehn has quit IRC | 16:29 | |
| *** dkehn has joined #zuul | 16:34 | |
| *** nilashishc has quit IRC | 16:35 | |
| tobiash | SotK: we run jobs on windows with the winrm connection and client certs | 16:35 |
| tobiash | Our current windows nodes are server 2012 | 16:35 |
| tobiash | And we use nodepool to boot them | 16:36 |
| *** nchakrab has joined #zuul | 16:37 | |
| *** nchakrab has quit IRC | 16:39 | |
| *** bbayszczak has quit IRC | 16:44 | |
| *** pcaruana has joined #zuul | 16:51 | |
| *** jpena is now known as jpena|off | 16:52 | |
| *** dkehn has quit IRC | 17:09 | |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: web: add /{tenant}/projects and /{tenant}/project/{project} routes https://review.openstack.org/550979 | 17:09 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: web: add /{tenant}/pipelines route https://review.openstack.org/541521 | 17:09 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: web: add /{tenant}/labels route https://review.openstack.org/553979 | 17:09 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: web: add /{tenant}/nodes route https://review.openstack.org/553998 | 17:09 |
| openstackgerrit | Joshua Hesketh proposed openstack-infra/zuul master: scheduler: add job's parent name to the rpc job_list method https://review.openstack.org/573473 | 17:09 |
| *** annabelleB has quit IRC | 17:22 | |
| *** dkehn has joined #zuul | 17:24 | |
| pabelanger | tobiash: corvus: q: how does zuul mark a job finished for the starting_builds sensor? We are seeing http://softwarefactory-project.io/paste/show/T9ZpRDtjgU6EiVsO5S90/ and I'm trying to debug why we have too many starting builds. | 17:25 |
| pabelanger | we have a bottleneck some place, but not sure where to look | 17:26 |
| *** annabelleB has joined #zuul | 17:26 | |
| tobiash | pabelanger: not at computer atm, but I think just before the first ansible run | 17:27 |
| tobiash | pabelanger: do you have cache and workdir on the same filesystem? | 17:28 |
| corvus | tobiash, pabelanger: yes, right before the first pre playbook. after the ansible setup step. | 17:28 |
| pabelanger | can't remember, checking | 17:28 |
| pabelanger | tobiash: oh, right. yes, single filesystem | 17:29 |
| tobiash | If not, that can produce a big overhead because git cannot use hardlinks | 17:29 |
| corvus | good -- they should be on the same system so git uses hard links | 17:29 |
| tobiash | Maybe just big repos? | 17:29 |
| pabelanger | yah, I need to remember if they are still backed with ceph, I think we switched them to SSD | 17:29 |
| pabelanger | yah, they are large repos, upstream 3pci jobs | 17:30 |
| pabelanger | for tripleo | 17:30 |
| tobiash | How many per job? | 17:30 |
| pabelanger | would need to check, but 10 or so | 17:30 |
| tobiash | I guess if you have large and many repos per job your only solution is more executors | 17:31 |
| tobiash | Or faster disks | 17:31 |
| pabelanger | I did notice http://paste.openstack.org/show/729973/ this morning, but haven't looked yet why the large delay. I believe it was because the executor was unregistered because of govenor | 17:32 |
| pabelanger | tobiash: yah, more executors is first suggestion too | 17:32 |
| pabelanger | we have 4 right now | 17:32 |
| pabelanger | with 4vcpu and 8gb | 17:33 |
| pabelanger | going to see if we can bump to 8vcpu | 17:33 |
| tobiash | We have 8 and run max around 20 jobs per executor | 17:33 |
| tobiash | Which is less than openstack but our executors are ceph backed | 17:34 |
| pabelanger | we get ~15 jobs today | 17:34 |
| pabelanger | corvus: is https://git.zuul-ci.org/cgit/zuul/tree/zuul/executor/server.py#n892 what I am looking for, which indicates starting_builds has finished? | 17:37 |
| corvus | pabelanger: https://git.zuul-ci.org/cgit/zuul/tree/zuul/executor/server.py#n1084 is when a build starts | 17:38 |
| pabelanger | thanks | 17:38 |
| *** dkehn has quit IRC | 17:44 | |
| *** dkehn has joined #zuul | 17:46 | |
| pabelanger | tobiash: corvus: okay, what it looks like a bunch of git operations now on zuul-executors. I think playbooks have changed in rdo, and have added a bunch more require-projects, along with slow disks, jobs are just taking much longer to start | 17:52 |
| *** sshnaidm|ptg has quit IRC | 17:58 | |
| *** bhavikdbavishi has quit IRC | 18:01 | |
| *** dkehn has quit IRC | 18:01 | |
| *** annabelleB has quit IRC | 18:20 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul master: Use tox-docs directly https://review.openstack.org/602131 | 18:21 |
| *** dkehn has joined #zuul | 18:24 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Autohold on timeout https://review.openstack.org/602134 | 18:26 |
| *** pcaruana has quit IRC | 18:47 | |
| *** sshnaidm|ptg has joined #zuul | 19:33 | |
| *** annabelleB has joined #zuul | 19:40 | |
| *** pcaruana has joined #zuul | 19:58 | |
| openstackgerrit | Steve Kowalik proposed openstack/pbrx master: Correct small typo https://review.openstack.org/602151 | 20:14 |
| *** pcaruana has quit IRC | 20:20 | |
| *** annabelleB has quit IRC | 20:22 | |
| *** annabelleB has joined #zuul | 20:25 | |
| *** annabelleB has quit IRC | 20:26 | |
| *** annabelleB has joined #zuul | 20:31 | |
| *** samccann has quit IRC | 20:56 | |
| *** annabelleB has quit IRC | 21:20 | |
| *** annabelleB has joined #zuul | 21:20 | |
| *** annabelleB has quit IRC | 21:25 | |
| *** hashar has quit IRC | 21:29 | |
| openstackgerrit | Steve Kowalik proposed openstack/pbrx master: Revert "Don't store apk indexes in cache" https://review.openstack.org/602172 | 21:35 |
| *** ianychoi has joined #zuul | 21:35 | |
| *** annabelleB has joined #zuul | 21:36 | |
| *** annabelleB has quit IRC | 21:40 | |
| *** annabelleB has joined #zuul | 21:43 | |
| *** annabelleB has quit IRC | 22:00 | |
| *** annabelleB has joined #zuul | 22:04 | |
| *** electrofelix has quit IRC | 22:29 | |
| openstackgerrit | Merged openstack/pbrx master: Revert "Don't store apk indexes in cache" https://review.openstack.org/602172 | 22:44 |
| openstackgerrit | Steve Kowalik proposed openstack/pbrx master: Correct small typo https://review.openstack.org/602151 | 22:47 |
| openstackgerrit | Steve Kowalik proposed openstack-infra/zuul master: Correct Last-Modified header for status API https://review.openstack.org/601832 | 22:48 |
| *** annabelleB has quit IRC | 22:58 | |
| dkehn | We are running Zuul 2.5.2 and are seeing traceback regularly of mode.py errors, pertaining to the Ref obj, has no number, Ref.equals - PullRequest has no Ref obj? beside updating Zuul to 3.2 is that anything I can do to remedy these issues? | 23:04 |
| clarkb | dkehn: it would probably be helpful to have the verbatim traceback posted via a paste service | 23:04 |
| dkehn | clarkb: ack | 23:05 |
| clarkb | as far as correcting that Ithink that can happen if you try to apply ref updated events to pipelines that operate on changes | 23:05 |
| clarkb | you cna't mix those | 23:05 |
| dkehn | clarkb: thanks | 23:06 |
| dkehn | clarkb: is this more of mis-config issue | 23:06 |
| clarkb | I think it may be | 23:11 |
| clarkb | if you mix ref-updated as a trigger event with requirements on a pipeline that come from changes | 23:11 |
| *** rlandy is now known as rlandy|bbl | 23:27 | |
| dmsimard | dkehn: zuul v3 is really good, we'd be happy to help if you have questions upgrading :) | 23:32 |
| *** sshnaidm|ptg has quit IRC | 23:48 | |
| *** sshnaidm|ptg has joined #zuul | 23:51 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!