Wednesday, 2020-02-05

*** Goneri has joined #zuul00:05
*** mattw4 has quit IRC00:06
*** jamesmcarthur has joined #zuul00:45
*** jamesmcarthur has quit IRC01:00
*** jamesmcarthur has joined #zuul01:10
tristanCcorvus: that is indeed a useful feature, allowed-labels is hard to use to prevent all-but-one tenant from using a label01:11
*** michael-beaver has quit IRC01:13
*** jamesmcarthur has quit IRC01:22
*** jamesmcarthur has joined #zuul01:45
*** Goneri has quit IRC01:47
*** igordc has joined #zuul01:52
*** jamesmcarthur has quit IRC02:22
*** bhavikdbavishi has joined #zuul02:56
openstackgerritMerged zuul/project-config master: Migrate to upload-afs-roots role  https://review.opendev.org/70537202:59
*** bhavikdbavishi1 has joined #zuul02:59
*** bhavikdbavishi has quit IRC03:00
*** bhavikdbavishi1 is now known as bhavikdbavishi03:00
*** rlandy has quit IRC03:03
*** raukadah is now known as chkumar|rover04:22
*** igordc has quit IRC04:59
ianwfungi: so we don't lose track of the zuul release tarball  issues you raised, i've filed https://storyboard.openstack.org/#!/story/2007261 with the updates discussed in #openstack-infra05:06
*** bolg has joined #zuul05:06
ianwi've assigned it to you for lack of anyone else05:07
ianwmordred/corvus : ^ you will probably have some thoughts on the issues in there too.  basically release job only releases to pypi; it's unclear if that is the overall intention05:08
*** evrardjp has quit IRC05:33
*** evrardjp has joined #zuul05:34
openstackgerritMerged zuul/zuul-jobs master: Use unique loop vars to avoid conflicts  https://review.opendev.org/70533705:43
openstackgerritMerged zuul/zuul-jobs master: Link the policy to the CONTRIBUTING.rst file  https://review.opendev.org/70448405:51
*** felixedel has joined #zuul06:24
openstackgerritJan Kubovy proposed zuul/zuul master: WIP: Store unparsed branch config in Zookeeper  https://review.opendev.org/70571606:36
*** bolg has quit IRC07:34
openstackgerritFelix Schmidt proposed zuul/zuul master: Report retried builds via sql reporter.  https://review.opendev.org/63350107:52
openstackgerritFelix Schmidt proposed zuul/zuul master: Store information about gate resets in MQTT and SQL reporter  https://review.opendev.org/69667007:52
openstackgerritTobias Henkel proposed zuul/zuul master: Evaluate CODEOWNERS settings during canMerge check  https://review.opendev.org/64455708:12
*** sugaar has joined #zuul08:14
*** hashar has joined #zuul08:27
openstackgerritSimon Westphahl proposed zuul/zuul master: Spec for allowing circular dependencies  https://review.opendev.org/64330908:36
openstackgerritSimon Westphahl proposed zuul/zuul master: Add optional support for circular dependencies  https://review.opendev.org/68535408:36
*** jpena|off is now known as jpena08:53
*** zxiiro has quit IRC09:06
*** bhavikdbavishi has quit IRC09:19
openstackgerritSimon Westphahl proposed zuul/zuul master: Spec for allowing circular dependencies  https://review.opendev.org/64330910:04
*** bolg has joined #zuul10:14
openstackgerritChandan Kumar (raukadah) proposed zuul/zuul-jobs master: Added bzip2 package for CentOS 8  https://review.opendev.org/70591110:18
*** hashar has quit IRC10:41
*** felixedel has quit IRC11:21
*** felixedel has joined #zuul11:24
*** wxy-xiyuan has quit IRC11:36
*** pcaruana has quit IRC11:37
*** pcaruana has joined #zuul11:50
*** tosky has joined #zuul12:12
*** jpena is now known as jpena|lunch12:33
fungithanks ianw! i may not be able to get to it this week, but suspect we can copy logic from existing jobs for it12:49
*** bhavikdbavishi has joined #zuul12:52
*** bhavikdbavishi1 has joined #zuul12:55
*** bhavikdbavishi has quit IRC12:57
*** bhavikdbavishi1 is now known as bhavikdbavishi12:57
*** rlandy has joined #zuul13:01
*** nhicher has quit IRC13:05
*** nhicher has joined #zuul13:06
*** smcginnis|FOSDEM is now known as smcginnis13:19
*** jamesmcarthur has joined #zuul13:20
mordredianw: I don't think it was intentional, other than we didn't have an appropriate destination at the time IIRC.13:23
*** bhavikdbavishi has quit IRC13:29
*** electrofelix has joined #zuul13:29
*** bhavikdbavishi has joined #zuul13:29
*** jpena|lunch is now known as jpena13:35
*** jamesmcarthur has quit IRC13:37
fungiyep, and then some months later we forgot the job was incomplete and put it into production13:44
*** jamesmcarthur has joined #zuul13:47
*** saneax has quit IRC13:49
*** zxiiro has joined #zuul13:53
openstackgerritFelix Schmidt proposed zuul/zuul master: Implement github checks API  https://review.opendev.org/70516813:55
*** pcaruana has quit IRC14:17
*** felixedel has quit IRC14:20
openstackgerritAlfredo Moralejo proposed zuul/zuul-jobs master: Add CentOS8 High Availability repository  https://review.opendev.org/70597314:45
openstackgerritAlfredo Moralejo proposed zuul/zuul-jobs master: Add CentOS8 High Availability repository  https://review.opendev.org/70597314:47
*** swest has quit IRC14:47
*** swest has joined #zuul14:48
*** bhavikdbavishi1 has joined #zuul14:48
*** swest has quit IRC14:48
*** bhavikdbavishi has quit IRC14:49
*** bhavikdbavishi1 is now known as bhavikdbavishi14:49
*** swest has joined #zuul14:50
*** sshnaidm|afk is now known as sshnaidm14:51
*** swest has quit IRC14:55
openstackgerritJames E. Blair proposed zuul/zuul-jobs master: Add index_links option to zuul manifest  https://review.opendev.org/70558015:19
openstackgerritMonty Taylor proposed zuul/zuul-jobs master: Google logs: Link to index.html files  https://review.opendev.org/70558615:21
mordredcorvus: ^^ rebased the followup change so I could +A it15:22
*** bolg has quit IRC15:23
corvusmordred: thanks!15:24
corvusi think i'll carry over ianw and tobiash's +2 since the fix wasn't substantial15:24
mordred++15:25
*** chkumar|rover is now known as raukadah15:27
corvusadding in the slowest test output has made the failing test fall outside the display error in the job summary window (see https://zuul.opendev.org/t/zuul/build/2a1e75da64804bf8a7e728837f952a47 )15:27
corvusshould we expand that window to copy over a few more lines?15:27
clarkbcorvus: maybe we can do a grep -i failure|error -A 30 -B 30 type of thing?15:28
mordredcorvus: yeah- I think so - I frequently find that it's just slightly too short - which may also be a result of having slowest on in other openstack repos15:28
clarkbto try and get it closer to the relevant data15:28
corvusclarkb: oh yeah -- maybe do that backwards from the end and show the first hit15:28
clarkbcorvus: ++15:29
mordred++15:29
openstackgerritMerged zuul/zuul-jobs master: Add upload-logs-gcs role  https://review.opendev.org/70371115:34
openstackgerritJames E. Blair proposed zuul/zuul master: Add disallowed-labels tenant option  https://review.opendev.org/70585615:36
*** ianychoi_ is now known as ianychoi15:40
openstackgerritMerged zuul/zuul-jobs master: Add index_links option to zuul manifest  https://review.opendev.org/70558015:40
openstackgerritMerged zuul/zuul-jobs master: Google logs: Link to index.html files  https://review.opendev.org/70558615:44
*** pcaruana has joined #zuul15:49
corvusthe new pipeline proposed by ttx for openstack may be interesting to folks: https://review.opendev.org/705990  (it basically will be used for quick jobs on the zuul executor to do a process approval check and vote in a dedicated gerrit review label)15:59
*** Goneri has joined #zuul16:02
mordredcorvus: that's cool16:03
ttxkinda curious to see if that will work in practice :)16:04
mordredttx: I like it as a concept!16:04
*** openstackstatus has joined #zuul16:06
*** ChanServ sets mode: +v openstackstatus16:06
fungiif nothing else, it'll be an interesting experiment16:07
ttxWe might extend that job so that it would subscribe the PTL/liaison to the change if they are missing16:07
corvusyeah, that should work; you might need an extra gerrit account credential to authonticate to do that16:09
ttxdoes comment-added also trigger on patchset-added?16:09
ttxlike when a patchset is added, I /think/ it also posts a comment so one should include the other16:09
ttx(in order to catch that the request is actually coming from the PTL so does not need further approval in comment)16:10
fungittx: opendev's check pipeline trigger lists them out separately at least: https://opendev.org/openstack/project-config/src/branch/master/zuul.d/pipelines.yaml#L19-L3016:20
ttxok, we'll see if there are corner cases to catch.16:21
fungiso we enqueue on patchset-created, change-restored, comment-added with a recheck pattern, and comment-added with an approval vote16:21
fungifor the openstack tenant anyway16:21
fungithe zuul tenant is slightly simpler what with lacking clean-check enforcement: https://opendev.org/zuul/project-config/src/branch/master/zuul.d/pipelines.yaml#L16-L2116:23
*** jamesmcarthur has quit IRC16:24
*** jamesmcarthur has joined #zuul16:38
*** jamesmcarthur_ has joined #zuul16:41
*** jamesmcarthur has quit IRC16:44
*** jamesmcarthur_ has quit IRC16:55
*** jamesmcarthur has joined #zuul16:58
tobiashcorvus: I added a question on 705856, what do you think?17:03
corvustobiash: ++.  followup ok?17:04
tobiashsure :)17:04
*** jamesmcarthur has quit IRC17:05
*** zxiiro has quit IRC17:07
*** zxiiro has joined #zuul17:08
*** jamesmcarthur has joined #zuul17:27
*** evrardjp has quit IRC17:33
*** evrardjp has joined #zuul17:34
*** igordc has joined #zuul17:34
openstackgerritTobias Henkel proposed zuul/zuul master: Add spec for scale out scheduler  https://review.opendev.org/62147917:49
sshnaidmis there a good way to apply "irrelevant_files" to whole repo, no matter what jobs run there and which project templates are used?17:51
dmsimardsshnaidm: https://opendev.org/openstack/project-config/src/branch/master/zuul.d/projects.yaml#L13-L1817:52
dmsimardah, I guess irrelevant_files isn't used at the project level17:53
tristanCdmsimard: iiuc, sshnaidm wants to apply an attribute to all the jobs except one of a pipeline17:53
*** jpena is now known as jpena|off17:53
sshnaidmright ^17:53
pabelangeryou could write a base job for that project, and parent everything in the project to that. And base has irrelevant_files17:55
dmsimardsshnaidm: ^ what he said but it looks like you're already doing that: https://github.com/openstack/tripleo-ci/blob/master/zuul.d/base.yaml#L6-L6517:55
sshnaidmpabelanger, I use mostly jobs and project templates from other repos17:55
*** jamesmcarthur has quit IRC17:56
sshnaidmdmsimard, hmm.. actually this is the way, thanks17:56
sshnaidmif zuul supported two parents, I could create a "mixin" and attach to specific base job :)17:58
dmsimardsshnaidm: closest I could think of would be to have a job which has the base job as parent, define some other things in that child and then have jobs which parent to the child17:59
dmsimardbase -> base_with_more_stuff <- actual-job17:59
sshnaidmdmsimard, yeah17:59
tristanCsshnaidm: actually zuul support multi parenting using job variant, but iirc all variants needs to be defined in a single repo, you can't mix in outside of the repo defining the job18:00
tristanCiirc, that results in a job using 2 parents:  job: { name: "my-job", parent: "parent1" }    and   job: { name: "my-job", parent: "parent2" }18:02
*** johnsom has quit IRC18:03
sshnaidmtristanC, and second won't just overwrite the first one?18:03
*** johnsom has joined #zuul18:03
tristanCsshnaidm: i don't think, let me test this out in your collect-logs template18:03
*** jamesmcarthur has joined #zuul18:03
sshnaidmtristanC, interesting.. does somebody use it? wondering if there are real-life examples18:04
tristanCcorvus: is using variant to use multiple parents still used in opendev somewhere?18:05
corvustristanC: i'm not sure; i think we used it in system-config then refactored something and stopped.18:06
tristanCsshnaidm: iirc, that should work: https://review.opendev.org/706071 DNM: test parent mixin18:07
corvusnote that yaml anchors may also be useful here (they let you repeat data blocks)18:08
*** mattw4 has joined #zuul18:09
tristanCcorvus: iirc, sshnaidm would like this rule to be also applied in the ppc18:11
mordredcorvus: do we not still use multi-parent for container jobs?18:12
tristanCe.g. where extra jobs can be added locally, and iiuc yaml anchor, their scope stops at a single file18:12
corvusyeah, you could add "irrelevant-files: *my_files" to every item in the ppc18:12
* mordred also can't remember18:12
corvusmordred: i have a vague memory of changing the inheritance hierarchy in such a way it didn't make sense any more.  i think as part of the promote work.18:13
tristanCcorvus: could *my_files be available when the ppc is defined in another repository from the anchor?18:13
mordredcorvus: oh yeah - I think you're right18:13
corvustristanC: no, anchors are only scoped to a file18:13
sshnaidmtristanC, couldn't you do just that? http://paste.openstack.org/show/789173/18:13
sshnaidmtristanC, yeah, yaml anchors works only in the file they're defined..18:14
corvussshnaidm: i don't think you will convince the maintainers of the openstack-tox-linters job to merge that change18:14
sshnaidmcorvus, I mean in different repo18:15
tristanCsshnaidm: iirc job variant can only be defined in a single project, e.g. you can't declare a job `openstack-tox-linters` in both ansible-role-collect-jobs and openstack-zuul-jobs18:15
sshnaidmtristanC, oh, I see18:15
sshnaidmnow it makes sense18:15
corvussshnaidm: that would change the job for everyone which you aren't allowed to do.  but you are allowed to change it for your project, just by adding irrelevant-files to the job in the ppc18:16
sshnaidmso variants are scoped to one repo18:16
*** sugaar has quit IRC18:17
corvusyou can do this: http://paste.openstack.org/show/789174/18:17
corvusand you can repeat that for every job you want to apply it to.  and you can use yaml anchors to make the repitition simpler.18:17
corvustheoretically i think maybe you could do that with a parent mixin too.  but i've never tried it.18:19
corvus(basically, a job in a ppc is an anonymous project-local variant, so generally anything you can do in a variant you can do in the ppc)18:21
corvuswe use that heavily in the zuul project; here's an example with irrelevant-files: https://opendev.org/zuul/zuul/src/branch/master/.zuul.yaml#L252-L26218:23
*** kmalloc has quit IRC18:27
*** kmalloc has joined #zuul18:27
sshnaidmcorvus, yeah, but we use templates mostly and would be nice if I had option to apply this to template (which would apply it for ALL jobs in template)18:33
corvussshnaidm: i hear you.  that doesn't exist right now, so i'm just trying to help you find something that will work18:38
sshnaidmcorvus, of course, interesting if there is an option to have it in roadmap, maybe it would be useful for others as well18:41
*** bhavikdbavishi has quit IRC18:41
fungiif you can figure out what that could look like, starting a zuul-discuss ml thread on it and/or submitting a spec could be a good way forward18:45
*** mattw4 has quit IRC18:48
*** jamesmcarthur has quit IRC18:54
*** mattw4 has joined #zuul18:56
*** electrofelix has quit IRC18:59
*** mattw4 has quit IRC19:06
*** jamesmcarthur has joined #zuul19:15
*** mattw4 has joined #zuul19:19
openstackgerritJames E. Blair proposed zuul/zuul master: Extract the watcher from git driver  https://review.opendev.org/70609219:21
*** maxamillion has quit IRC19:27
*** maxamillion has joined #zuul19:27
mordredcorvus: git add19:37
*** maxamillion has quit IRC19:37
*** kmalloc has quit IRC19:39
*** zxiiro has quit IRC19:39
*** johnsom has quit IRC19:40
corvusmordred: details20:00
openstackgerritJames E. Blair proposed zuul/zuul master: Extract the watcher from git driver  https://review.opendev.org/70609220:00
mordredcorvus: the computer should know what you mean20:01
corvusmordred: i read in the tech press that development is all done by ai at this point, so... yeah20:02
fungimy development sure is20:02
fungithere is no intelligence more artificial20:03
fungior was that alcohol intelligence?20:03
*** jamesmcarthur has quit IRC20:21
openstackgerritAndreas Jaeger proposed zuul/zuul-jobs master: Revert "Add index_links option to zuul manifest"  https://review.opendev.org/70610620:41
openstackgerritIan Wienand proposed zuul/zuul master: Handle index_links in manifest  https://review.opendev.org/70611521:03
openstackgerritJames E. Blair proposed zuul/zuul-jobs master: Move index_links to correct task in generate_zuul_manifest  https://review.opendev.org/70611621:07
*** tosky has quit IRC21:18
openstackgerritJames E. Blair proposed zuul/zuul master: Extract the watcher from git driver  https://review.opendev.org/70609221:21
openstackgerritMerged zuul/zuul-jobs master: Move index_links to correct task in generate_zuul_manifest  https://review.opendev.org/70611621:31
openstackgerritTristan Cacqueray proposed zuul/zuul master: add_host: enable using ansible_fqdn and ansible_private_key_file  https://review.opendev.org/70612421:52
openstackgerritJames E. Blair proposed zuul/zuul master: Gerrit: add polling support for refs  https://review.opendev.org/70613822:29
*** yolanda has joined #zuul22:35
openstackgerritJames E. Blair proposed zuul/zuul master: Extract allow/disallow filter into util function  https://review.opendev.org/70614422:44
openstackgerritMerged zuul/zuul-jobs master: Add CentOS8 High Availability repository  https://review.opendev.org/70597322:50
*** maxamillion has joined #zuul23:01
*** johnsom has joined #zuul23:02
*** kmalloc has joined #zuul23:02
*** zxiiro has joined #zuul23:03
*** jamesmcarthur has joined #zuul23:03
*** jamesmcarthur has quit IRC23:06
*** rlandy is now known as rlandy|bbl23:25

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!