*** spsurya has joined #zuul | 01:29 | |
*** noorul has joined #zuul | 02:06 | |
noorul | Hi | 02:06 |
---|---|---|
noorul | I am facing a strange issue | 02:06 |
noorul | console log show different tasks getting executed compared to executor log | 02:06 |
*** jhesketh has joined #zuul | 02:29 | |
*** noorul has quit IRC | 02:32 | |
*** noorul has joined #zuul | 02:42 | |
*** noorul has quit IRC | 02:44 | |
*** noorul has joined #zuul | 03:00 | |
*** noorul has quit IRC | 03:06 | |
*** noorul has joined #zuul | 03:09 | |
*** noorul has quit IRC | 03:15 | |
*** noorul has joined #zuul | 03:30 | |
*** jamesmcarthur has joined #zuul | 03:48 | |
*** jamesmcarthur has quit IRC | 04:06 | |
*** jamesmcarthur has joined #zuul | 04:09 | |
*** jamesmcarthur has quit IRC | 04:14 | |
*** jamesmcarthur has joined #zuul | 04:23 | |
*** noorul has quit IRC | 04:25 | |
*** noorul has joined #zuul | 04:45 | |
SpamapS | noorul: I think you'll need to provide pastes and / or reproducers. That sounds very serious, but I've certainly never seen it. | 04:50 |
*** jamesmcarthur has quit IRC | 04:56 | |
noorul | SpamapS: I found out the problem. I had two Zuul instances running | 05:06 |
noorul | SpamapS: once docker based and another using "from scratch" doc | 05:06 |
noorul | SpamapS: Thank you | 05:06 |
noorul | SpamapS: s/once/one | 05:07 |
SpamapS | OHH ow | 05:18 |
SpamapS | noorul: glad you resolved. :) | 05:18 |
*** noorul has quit IRC | 05:20 | |
*** jamesmcarthur has joined #zuul | 05:25 | |
*** noorul has joined #zuul | 05:28 | |
SpamapS | Hrm.. seems like nodepool min-ready doesn't scale down very well. I have 6 AZ's, spread across 2 providers. I ony need min-ready: 3 .. but because the 3 ready nodes happened in region 1, reqs served by the region 2 provider still boot on-demand nodes. :-/ | 05:29 |
*** jamesmcarthur has quit IRC | 05:33 | |
*** noorul has quit IRC | 05:49 | |
*** wxy-xiyuan has joined #zuul | 06:13 | |
*** jamesmcarthur has joined #zuul | 06:29 | |
*** jamesmcarthur has quit IRC | 06:34 | |
*** themroc has joined #zuul | 06:55 | |
*** noorul has joined #zuul | 07:15 | |
*** jangutter has joined #zuul | 07:21 | |
*** jamesmcarthur has joined #zuul | 07:30 | |
*** tosky_ has joined #zuul | 07:34 | |
*** jamesmcarthur has quit IRC | 07:35 | |
*** jpena|off is now known as jpena | 07:45 | |
*** pcaruana has joined #zuul | 07:49 | |
*** hashar has joined #zuul | 07:56 | |
*** arxcruz is now known as arxcruz_pto | 08:19 | |
*** panda|rover|off is now known as panda|rover | 08:23 | |
*** avass has joined #zuul | 08:30 | |
*** jamesmcarthur has joined #zuul | 08:31 | |
*** tosky_ is now known as tosky | 08:32 | |
*** gtema_ has joined #zuul | 08:32 | |
*** jamesmcarthur has quit IRC | 08:36 | |
*** jank has joined #zuul | 09:11 | |
*** jamesmcarthur has joined #zuul | 09:33 | |
*** jamesmcarthur has quit IRC | 09:44 | |
openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: add-build-sshkey: add centos/rhel-8 support https://review.opendev.org/674092 | 09:48 |
*** noorul has quit IRC | 09:53 | |
*** noorul has joined #zuul | 10:26 | |
*** jamesmcarthur has joined #zuul | 10:29 | |
*** jamesmcarthur has quit IRC | 10:34 | |
*** hashar has quit IRC | 10:38 | |
openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: add-build-sshkey: add centos/rhel-8 support https://review.opendev.org/674092 | 10:41 |
*** NBorg has joined #zuul | 11:02 | |
*** jamesmcarthur has joined #zuul | 11:14 | |
*** hashar has joined #zuul | 11:23 | |
*** jamesmcarthur has quit IRC | 11:24 | |
*** jamesmcarthur has joined #zuul | 11:25 | |
*** sean-k-mooney has joined #zuul | 11:30 | |
sean-k-mooney | dumb question, is it posible to define a nodeset that can use "label-a or label-b" instead of just a single label? | 11:31 |
sean-k-mooney | its not a blocker if i cant do this currently but im about to start creating a gate jobs that reuiqres nested virt and i would like to create a nodeset that works on vexhost or fortnoble in the furture using there provider specific lables rather then the standard ones | 11:33 |
*** jamesmcarthur has quit IRC | 11:35 | |
*** jpena is now known as jpena|lunch | 11:39 | |
zbr | is there a part of zuul I can use to parse Depends-On(s) in a given commit message in order to build a dependency graph? | 11:44 |
zbr | i got the commit from gerrit api, now I want to know on what else it depends on. the format used by depends-on is quite flexible,... | 11:45 |
*** rfolco has joined #zuul | 12:04 | |
*** gtema_ has quit IRC | 12:07 | |
*** jamesmcarthur has joined #zuul | 12:09 | |
*** panda|rover is now known as panda|rover|eat | 12:13 | |
*** jamesmcarthur has quit IRC | 12:13 | |
*** NBorg has quit IRC | 12:21 | |
*** jpena|lunch is now known as jpena | 12:32 | |
*** panda|rover|eat is now known as panda|rover | 12:36 | |
*** jamesmcarthur has joined #zuul | 12:45 | |
*** nBorg has joined #zuul | 12:49 | |
*** bhavikdbavishi has joined #zuul | 12:53 | |
*** bhavikdbavishi has quit IRC | 13:10 | |
fbo | tobiash: hi, for the CODEOWNERS patch, did you tried to use this endpoint https://api.github.com/repos/morucci/repopo1/pulls/30/requested_reviewers ? it lists the pending requested reviews | 13:38 |
*** jamesmcarthur has quit IRC | 13:49 | |
*** hashar has quit IRC | 14:06 | |
*** noorul has quit IRC | 14:14 | |
*** armstrongs has joined #zuul | 14:26 | |
armstrongs | Following on from my question yesterday about attaching artifacts to a build. I am currently producing an artifact on my nodepool vm, synchronizing (rsyncing) it back to the executor and synchronizing( rsyncing) to the log server. | 14:30 |
armstrongs | his works great for vms, but when I'm using the kubernetes nodepool container instead of the nodepool vm what mechanism can I use to sync the file back to the executor? As zuul doesn't allow me to sync files not on the executor filesystem? | 14:31 |
*** themroc has quit IRC | 14:47 | |
*** noorul has joined #zuul | 14:48 | |
*** themroc has joined #zuul | 14:48 | |
*** jangutter has quit IRC | 14:51 | |
*** noorul has quit IRC | 14:52 | |
*** noorul has joined #zuul | 14:58 | |
*** noorul has quit IRC | 15:00 | |
*** themroc has quit IRC | 15:06 | |
*** jank has quit IRC | 15:15 | |
*** jangutter has joined #zuul | 15:16 | |
SpamapS | sean-k-mooney: labels can be the same label, but different configs, across different providers. | 15:19 |
sean-k-mooney | they can yes | 15:19 |
SpamapS | sean-k-mooney: they're meant to be high level abstractions. It's not "xlarge-rax-nested-virt" it's just "xlarge-mytests" and you config the label the way it needs to be to run your tests in whichever provider. | 15:20 |
sean-k-mooney | yes i know. i may add some myself later | 15:20 |
sean-k-mooney | but basically i was wondoing if i could pass a list and say give me a node iwth any of these labels | 15:21 |
*** jamesmcarthur has joined #zuul | 15:25 | |
*** jamesmcarthur has quit IRC | 15:31 | |
*** hashar has joined #zuul | 15:39 | |
*** mgoddard has quit IRC | 15:41 | |
*** bhavikdbavishi has joined #zuul | 15:43 | |
*** mgoddard has joined #zuul | 15:43 | |
*** zbr is now known as zbr|ooo | 15:52 | |
*** jamesmcarthur has joined #zuul | 16:02 | |
*** jamesmcarthur has quit IRC | 16:06 | |
*** spsurya has quit IRC | 16:44 | |
*** jamesmcarthur has joined #zuul | 16:49 | |
*** jamesmcarthur has quit IRC | 17:20 | |
*** hashar has quit IRC | 17:21 | |
*** jpena is now known as jpena|off | 17:25 | |
openstackgerrit | Clark Boylan proposed zuul/zuul-website master: Add Zuul FAQ page https://review.opendev.org/679670 | 17:27 |
clarkb | This change adds an FAQ page to the zuul website. As zuul has gotten more attention it seems like we are getting enough common questions that we can go ahead and start answering them on the website | 17:28 |
clarkb | if others agree I left the actual questions and answers blank for now and figured we could hash those out in review? | 17:28 |
*** jamesmcarthur has joined #zuul | 17:29 | |
clarkb | Things like: What code review systems are supported, where can nodepool provision test nodes, and how Zuul is different than Jenkins seem like common questions | 17:29 |
fungi | sounds like a great start, thanks!!! | 17:30 |
*** mgoddard has quit IRC | 17:32 | |
*** mgoddard has joined #zuul | 17:35 | |
*** mgoddard has quit IRC | 17:40 | |
*** zbr|ooo is now known as zbr | 17:53 | |
*** armstrongs has quit IRC | 17:53 | |
*** mgoddard has joined #zuul | 17:53 | |
*** jamesmcarthur has quit IRC | 17:57 | |
*** themroc has joined #zuul | 18:15 | |
*** pcaruana has quit IRC | 18:19 | |
*** pcaruana has joined #zuul | 18:19 | |
*** rfolco is now known as rfolco|doctor | 18:28 | |
sean-k-mooney | clarkb: fungi i asked this prviously but since ye are about. i assume it would be non trival to extend nodeset so a node can have a list of lables instead of just one. | 18:30 |
clarkb | sean-k-mooney: you should instead create a label that covers your cases | 18:31 |
sean-k-mooney | ya i could in some cases. the use case being that you have a set of providers that provide different node types and i want to normalise it | 18:31 |
clarkb | ya the normalization should happen in the nodepool label config | 18:31 |
sean-k-mooney | e.g create a nodeset that can run on vexhost or FN | 18:31 |
sean-k-mooney | but not the other providers | 18:32 |
clarkb | create a label that both of those providers provide | 18:32 |
sean-k-mooney | ya. thats what i assume dwas the desired approch | 18:32 |
sean-k-mooney | just wanted to check | 18:32 |
sean-k-mooney | i dont need it right now but it would be nice to have a nested virt lable at some point eventhough i know we shoudl not rely on nest vert in general. | 18:33 |
sean-k-mooney | out of interest how were the html docs in https://review.opendev.org/#/c/679670/ created | 18:34 |
sean-k-mooney | did you generate them by hand or use a tool like sphinx | 18:35 |
clarkb | I think https://html5up.net/ was used to bootstrap then we've edited by hand from there | 18:35 |
clarkb | that particular change was done by hand | 18:36 |
sean-k-mooney | ok. has the idea of moving it to rst at somepoint come up. it just looks strange to reivew the html files | 18:36 |
sean-k-mooney | i mean yes we can view the rendered form https://openstack.fortnebula.com:13808/v1/AUTH_e8fd161dc34c421a979a9e6421f823e9/logs_70/679670/1/check/zuul-website-build/53822b4/html/docs/faq.html | 18:37 |
clarkb | I don't think there is any intention of using sphinx for the website (sphinx generates the docs though) | 18:37 |
sean-k-mooney | hugo is a really nice staic sight generator. its been years since i have created html manually so it just felt. old school? | 18:38 |
sean-k-mooney | not sure that is the right term just it wasnt what i expected | 18:38 |
pabelanger | how much work is it to make upload-logs-swift compatible with aws s3? | 18:41 |
clarkb | pabelanger: I think you'll want to write a different role called upload-logs-s3 | 18:42 |
pabelanger | that's what I feared | 18:42 |
clarkb | you may be able to reuse a fair bit of the swift roles library/ contents and update to talk to s3 instead, but I know there is a bit of swift specific stuff in there like when to retry and settings cors headers (s3 uses a different method for that) | 18:43 |
fungi | clarkb: sean-k-mooney: i think the next major uplift of the main website was possibly going to use netlify and take advantage of the gerrit backend plugin for that, but we need someone interested in driving it forward | 18:44 |
fungi | sean-k-mooney: also, some of us here are "old school" so editing and reviewing raw html isn't that foreign to us ;) | 18:45 |
pabelanger | clarkb: k, thanks for info. For now, will try to request 2nd set of credential to openstack cloud, then find time / human to work on role update | 18:45 |
sean-k-mooney | fungi: before i started working on openstack i was working on a transaltion cms written with a php frontend and c++ backend so ya i dont mind editing html by had | 18:46 |
fungi | well, i didn't mean to suggest that i don't *mind* doing it | 18:48 |
sean-k-mooney | pabelanger: ironically if we had an upload-logs-s3 you could proably use it with swift | 18:48 |
clarkb | pabelanger: I think with s3 bucket names have to be globally unique too, so a zuul role for that may need to tack on a random string to the buckets it creates to avoid collisions | 18:48 |
clarkb | pabelanger: otherwise two different zuul installs would collide each other | 18:48 |
clarkb | so ya there will be things that you have to change to use s3 | 18:48 |
sean-k-mooney | fungi: no but most people in this ecosystem proably is comforatble reading it | 18:48 |
fungi | yep ;) | 18:48 |
sean-k-mooney | even if its not our perference | 18:48 |
pabelanger | clarkb: we have that issue today with vexxhost, zuul_log_container needs to be unique per zuul install | 18:49 |
fungi | sean-k-mooney: i actually don't mind editing and reviewing html, but a higher-level abstraction would also be nice | 18:49 |
clarkb | pabelanger: I guess that is a ceph limitation? | 18:49 |
clarkb | swift doesn't have that problem aiui | 18:49 |
pabelanger | clarkb: think so, yah | 18:49 |
sean-k-mooney | i think you mentioned netlify to me in the past. i guess im not sure why you would do that vs just wrigin a job in zuul | 18:50 |
sean-k-mooney | i mean if you want a ci cd workflow for deploying the zull website zuul can provide that right | 18:50 |
clarkb | yes, netlify is more interesting as a way for people to edit websites with minimal gerrit/git interaction | 18:51 |
* pabelanger heads back into the garage | 18:51 | |
sean-k-mooney | clarkb: ah so your are looking at it as a way to allow people to extend the zuul site more simplery rather then for its build system. | 18:52 |
sean-k-mooney | e.g. to hide gerrit/git | 18:52 |
clarkb | yes | 18:52 |
clarkb | there is a gerrit driver for it that will push edits into gerrit for review | 18:52 |
clarkb | doesn't fully hide gerrit but makes interacting iwth it easierthan learning git | 18:53 |
sean-k-mooney | cool so workflow woudl be edit in netlify. push to gerrit via netlify, review, merge, zuul deploy? or netlfy deploy? | 18:53 |
clarkb | zuul deploy | 18:54 |
sean-k-mooney | sound useful | 18:54 |
sean-k-mooney | i wonder if that would be a good workfolow for docs | 18:54 |
clarkb | er re bucket randomization can't be random or we'll create a new bucket for every job. Maybe hash the zuul hostname and use that as a prefix/suffix | 18:54 |
clarkb | something to namespace it | 18:55 |
*** tobiash has quit IRC | 19:47 | |
*** sshnaidm|pto has quit IRC | 19:47 | |
*** tobiash has joined #zuul | 19:48 | |
*** sshnaidm|pto has joined #zuul | 19:51 | |
*** themroc has quit IRC | 19:52 | |
*** logan- has quit IRC | 19:59 | |
*** logan- has joined #zuul | 20:03 | |
*** bhavikdbavishi has quit IRC | 20:05 | |
fungi | also netlify has a preview interface which could just link to zuul draft builds of the content, in theory anyway | 20:05 |
*** pcaruana has quit IRC | 20:36 | |
Shrews | sean-k-mooney: i agree with what the others said about node labels. But also, nodepool *does* have support for a node with multiple labels (so a node with labels A and B would match requests for either A *or* B), but this is only implemented in the static driver right now. It would have to be added to the openstack driver... in case that might be useful to you. | 20:42 |
sean-k-mooney | Shrews: dont worry about it. | 20:43 |
Shrews | done | 20:44 |
sean-k-mooney | Shrews: if i had full admin acess to the nodepool config i woudl just creat a lable on each provider | 20:44 |
sean-k-mooney | i was just wondering if there was a way to do it a a non admin project | 20:44 |
sean-k-mooney | e.g. to do it via the nova repo rather then having to do it via the project-config repo | 20:45 |
sean-k-mooney | creating a lable is trival to do and is what i originaly came up with | 20:45 |
sean-k-mooney | but i wanted to explore if there was another way before i did it | 20:45 |
Shrews | fungi: you moving inland at all??? | 20:46 |
fungi | Shrews: yeah, tomorrow is the day we batten down the hatches and move things indoors/upstairs, then wednesday is when the mandatory evacuation order goes into effect and we'll be bringing the cats on an impromptu mountain vacation | 20:55 |
Shrews | fungi: ++. good luck | 20:56 |
fungi | thanks! you too | 21:09 |
fungi | fingers crossed durham doesn't get washed away | 21:09 |
*** jamesmcarthur has joined #zuul | 21:24 | |
*** tobberydberg has quit IRC | 21:30 | |
*** fdegir has quit IRC | 21:31 | |
*** tobberydberg has joined #zuul | 21:31 | |
*** fdegir has joined #zuul | 21:32 | |
*** rfolco|doctor is now known as rfolco | 21:35 | |
mhu | o/ is it possible to use a PR for sphinx-openapi in a zuul review? see https://review.opendev.org/#/c/674257 | 22:02 |
*** jamesmcarthur has quit IRC | 22:02 | |
*** tosky_ has joined #zuul | 22:04 | |
*** tosky has quit IRC | 22:06 | |
*** tosky_ is now known as tosky | 22:09 | |
*** jamesmcarthur has joined #zuul | 22:31 | |
*** jamesmcarthur has quit IRC | 22:59 | |
*** jamesmcarthur has joined #zuul | 23:03 | |
*** jamesmcarthur has quit IRC | 23:05 | |
*** jamesmcarthur has joined #zuul | 23:05 | |
*** avass has quit IRC | 23:10 | |
*** jamesmcarthur has quit IRC | 23:23 | |
*** jamesmcarthur has joined #zuul | 23:27 | |
*** jamesmcarthur has quit IRC | 23:32 | |
*** armstrongs has joined #zuul | 23:47 | |
*** jamesmcarthur has joined #zuul | 23:56 | |
*** armstrongs has quit IRC | 23:57 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!