Tuesday, 2022-04-26

-@gerrit:opendev.org- Ian Wienand proposed: [zuul/zuul] 838971: Add unit test for zuul_json text result parsing https://review.opendev.org/c/zuul/zuul/+/83897100:08
@q:fricklercloud.dehow is log-streaming going to be handled in unrestricted ansible? I noticed that already now I don't get log streaming if I use e.g. ansible.builtin.shell instead of just shell06:07
@avass:vassast.org> <@q:fricklercloud.de> how is log-streaming going to be handled in unrestricted ansible? I noticed that already now I don't get log streaming if I use e.g. ansible.builtin.shell instead of just shell06:33
I guess that's because the callback only checks for `command` or `shell`: vm000059.int.autoheim.net
@avass:vassast.org> <@q:fricklercloud.de> how is log-streaming going to be handled in unrestricted ansible? I noticed that already now I don't get log streaming if I use e.g. ansible.builtin.shell instead of just shell06:34
* I guess that's because the callback only checks for `command` or `shell`: https://opendev.org/zuul/zuul/src/branch/master/zuul/ansible/base/callback/zuul_stream.py#L264
-@gerrit:opendev.org- Guillaume Chauvel proposed: [zuul/zuul] 839309: TEST: Cherry-pick tests & piece of code from change 762887 https://review.opendev.org/c/zuul/zuul/+/83930907:38
-@gerrit:opendev.org- Guillaume Chauvel proposed: [zuul/zuul] 839309: TEST: Cherry-pick tests & piece of code from change 762887 https://review.opendev.org/c/zuul/zuul/+/83930907:58
-@gerrit:opendev.org- Oleksandr Kozachenko proposed: [zuul/zuul-jobs] 838919: Add promote-container-image role https://review.opendev.org/c/zuul/zuul-jobs/+/83891908:56
-@gerrit:opendev.org- Simon Westphahl proposed: [zuul/zuul] 839321: Don't try to update deps of non-change refs https://review.opendev.org/c/zuul/zuul/+/83932109:14
-@gerrit:opendev.org- Guillaume Chauvel proposed: [zuul/zuul] 838208: Check quick-start result through zuul-web https://review.opendev.org/c/zuul/zuul/+/83820809:30
-@gerrit:opendev.org- Oleksandr Kozachenko proposed: [zuul/zuul-jobs] 838919: Add promote-container-image role https://review.opendev.org/c/zuul/zuul-jobs/+/83891909:35
@cdesai:matrix.orghello everyone, I've been lurking here for a while.10:51
I'm looking at setting up zuul for usage with gerrit, and had a couple questions
@cdesai:matrix.org1. Is there any official / recommended ansible setup to install zuul itself?10:51
@cdesai:matrix.org2. is there a concept of firing off one-off builds directly from zuul (or does it have to be triggered through a git commit always?)10:51
@cdesai:matrix.orgThanks!10:51
-@gerrit:opendev.org- Dong Zhang proposed: [zuul/zuul] 839370: WIP: Fix a bug that re-enqueue may skip jobs unexpectly https://review.opendev.org/c/zuul/zuul/+/83937012:23
-@gerrit:opendev.org- Dong Zhang proposed: [zuul/zuul] 839370: WIP: Fix a bug that re-enqueue may skip jobs unexpectly https://review.opendev.org/c/zuul/zuul/+/83937012:53
@tristanc_:matrix.orgcde: Hello, welcome! To get started it is recommended to go through https://zuul-ci.org/docs/zuul/latest/tutorials/quick-start.html13:55
@tristanc_:matrix.orgcde: and to run one-off builds, you will have to create a commit with a project pipeline config13:55
@cdesai:matrix.org> <@tristanc_:matrix.org> cde: Hello, welcome! To get started it is recommended to go through https://zuul-ci.org/docs/zuul/latest/tutorials/quick-start.html13:59
thanks. I did that, got it running as well. If there's no ansible example I guess I'll just end up writing it on my own based on the docker compose files
@jim:acmegating.comcde: pabelanger wrote https://opendev.org/windmill/ansible-role-zuul -- it's not official, and i'm not sure it's up to date.  it also doesn't use containers, which i would highly recommend.14:02
@jim:acmegating.comcde: https://opendev.org/opendev/system-config/src/branch/master/playbooks/roles is how opendev runs zuul's own zuul.  it does use containers.14:03
@jim:acmegating.comcde: one of those might be a good starting point for developing your own.14:03
@jim:acmegating.comcde: (note that opendev's zuul roles are not intended for direct public re-use -- feel free to fork them and modify to your heart's content -- opendev doesn't really want to maintain general roles)14:05
@tristanc_:matrix.orgthe software-factory project maintains general roles that uses podman: https://softwarefactory-project.io/cgit/software-factory/sf-config/tree/ansible/roles/sf-zuul/14:10
-@gerrit:opendev.org- Dong Zhang proposed: [zuul/zuul] 839370: WIP: Fix a bug that re-enqueue may skip jobs unexpectly https://review.opendev.org/c/zuul/zuul/+/83937014:32
-@gerrit:opendev.org- Dong Zhang proposed: [zuul/zuul] 839370: WIP: Fix a bug that re-enqueue may skip jobs unexpectly https://review.opendev.org/c/zuul/zuul/+/83937015:00
@cdesai:matrix.orgthank you so much!15:43
-@gerrit:opendev.org- James E. Blair https://matrix.to/#/@jim:acmegating.com proposed: [zuul/zuul] 839423: Fix reporting certain config syntax errors https://review.opendev.org/c/zuul/zuul/+/83942318:19
@jpew:matrix.orgAre the nodepool builds broken?19:03
@clarkb:matrix.orgjpew: I don't think so. OpenDev is building Jammy images as of last night relative to me using a very recent docker image build19:07
@clarkb:matrix.orgcan you be more specific about the problem you are having?19:07
@jpew:matrix.orgClark: sorry, more specifically the gate and check pipelines for nodepool are failing with an aws error19:08
@clarkb:matrix.orgthat almost looks like ti is making a real request again aws19:09
@clarkb:matrix.orgthe aws fakes may need updating to match some boto updates?19:09
@clarkb:matrix.orgI'm not super familiar with how that works19:09
@jpew:matrix.orgI'm not either 19:10
-@gerrit:opendev.org- Ian Wienand proposed: [zuul/nodepool] 839431: [wip] try to find boto3 error https://review.opendev.org/c/zuul/nodepool/+/83943119:20
-@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul] 839423: Fix reporting certain config syntax errors https://review.opendev.org/c/zuul/zuul/+/83942319:49
@iwienand:matrix.orgyeah there has been a flurry of activity in boto3 with several releases in the last few days20:04
@iwienand:matrix.orgi think it is just a override needed but trying to bisect which one from the changes20:04
-@gerrit:opendev.org- Ian Wienand proposed: [zuul/nodepool] 839431: [wip] pin moto to 3.1.5 https://review.opendev.org/c/zuul/nodepool/+/83943120:32
@iwienand:matrix.orgok, it seems it's actually a "moto" release20:35
-@gerrit:opendev.org- Zuul merged on behalf of James E. Blair https://matrix.to/#/@jim:acmegating.com: [zuul/zuul-jobs] 837416: Add per-build WinRM cert generation https://review.opendev.org/c/zuul/zuul-jobs/+/83741621:52
@clarkb:matrix.orgCouple of things that OpenDev wanted to bring up. First is frickler would like a nodepool release to ensure nodepool has a release that can build jammy images.21:53
@clarkb:matrix.orgSecond is that we've discovered that zuul-jobs' configure-mirrors sets up Debian (but not Ubuntu) with deb-src entries in sources.list config files21:53
@clarkb:matrix.orgOpenDev is looking at removing our source mirroring (to reduce mirror size). I'm wondering if anyone is using deb-src from the Debian configs or if we can just drop it. If we need to be careful with backward compat we should be able to wrap it in a flag that enables it by default21:54
@jpew:matrix.orgI'd very much appreciate https://review.opendev.org/c/zuul/nodepool/+/839226 and https://review.opendev.org/c/zuul/nodepool/+/833099 in a nodepool release.... I'm not sure if I'm missing some part of the development workflow to get them gated?21:55
@clarkb:matrix.orgjpew: for the first one we need passing tests, but I assume that is the moto thing ianw is tracking down so should be possible. For the second I'm not sure why tristanC didn't approve it21:55
@clarkb:matrix.orgwe can also do multiple releases in a short period of time21:56
@clarkb:matrix.orgtristanC: is there a specific concern that kept you from approving that change? Its definitely the sort of change that more eyeballs on wouldn't be the worst thing to ensure we aren't missing something with how the quota accounting works21:56
@iwienand:matrix.org(yes just pulling some source and reverting backwards to see if i can pinpoint the change)21:56
@jpew:matrix.orgClark: Ya, that's fine. It had just been sitting a while so I didn't know what was going on :)21:57
@tristanc_:matrix.orgClark: I was not entirely sure if the previous questions on that change are fully resolved21:59
@clarkb:matrix.orgtristanC: I think they were by splitting the behavior behind the config option. But maybe tobiash can review it and double check that assumption22:00
@jpew:matrix.org> tristanC: I think they were by splitting the behavior behind the config option. But maybe tobiash can review it and double check that assumption22:01
Yes that was the intent
@iwienand:matrix.orggit bisect has led to first bad commit as -> https://github.com/spulec/moto/commit/752eee1941de3214eea01d9649f33456aedf097c22:02
@iwienand:matrix.orgthe error originates from https://opendev.org/zuul/nodepool/src/branch/master/nodepool/tests/unit/test_driver_aws.py#L49022:06
@jim:acmegating.comjpewClark tristanC that seems like a very strange change... would it make more sense to just turn quota handling off in nodepool?22:08
@clarkb:matrix.orgianw something in https://github.com/spulec/moto/commit/752eee1941de3214eea01d9649f33456aedf097c#diff-9dba1b3c572ca75c4c0f539c78301a8c8300e8b42825d4e4286b95eb99827944L225-L262 ?22:08
@jim:acmegating.comjpew: like, it sounds like what you really want is for nodepool to ignore quota22:09
@jpew:matrix.orgcorvus: We have to do that to (via `ignore-provider-quota`), but even then a quota error from openstack itself makes the node allocation completely fail22:09
@jpew:matrix.orgOr rather nodepool gives up on a quota failure from openstack, and we want it to keep trying so that we can reactively bump up the quota for the project22:10
@jpew:matrix.orgIt needs the "pressure" of a pending node allocation to know that the project needs more quota22:10
@jim:acmegating.comjpew: okay, i'm sold on the need for it, but i do think that change needs another revision; i left some comments.22:15
@jim:acmegating.comClark: frickler i expect to make a nodepool 6.0 release with the zuul 6.0 release, which I hope will be this week.  I trust that would be okay?  (incidentally since the specific feature that frickler wants in the release is the ability to build jammy, that probably should have gotten a release note)22:16
@jpew:matrix.orgcorvus: excellent! I'll get that fixed up tomorrow. Thanks22:17
@clarkb:matrix.orgcorvus: release note makes sense. I guess its somewhat of a side effect and largely to do with the images themselves. But ++ since people are interested in releases for it22:17
@clarkb:matrix.orgI'll let frickler push that up as frickler was far more involved in getting all of those pieces together22:18
@jim:acmegating.comClark: i don't think we need it for this -- just noting that if we didn't already have a release lined up with a bunch of other stuff, that would be a blocker in that case.22:18
@jim:acmegating.com(but since we have a pile of features/release notes already and the totality warrants a 6.0, it's not necessary (but still not unwelcome) now)22:19
@clarkb:matrix.orgcorvus: for the zuul-jobs thing any sense for whether or not we should notify people with a two week notice email or just hide the change behind a toggle var?22:19
@jim:acmegating.comClark: remind me the zuul-jobs thing you're talking about?22:19
@clarkb:matrix.orgcorvus: Debian (but not Ubuntu) adds deb-src entries to its source.list stuff configured by zuul-jobs configure-mirrors role. OpenDev is looking at dropping source packages to slim down our mirrors. Doing that would break our current debian jobs when they try to apt-get update and find a mirror with no deb-src22:20
@clarkb:matrix.orgI think it is unlikely that anyone acutally uses deb-src packages in their CI jobs so wonder if we can just remove those lines (can announce and give people time to prep or say that doesn't work for them). Or we can remove those lines if an option to configure-mirrors is set then opendev would set that option22:21
@jim:acmegating.comClark: i feel like that's probably a good candidate for a flag22:21
@jim:acmegating.comClark: i agree with you about the relative probability, but it seems like it's easy enough to add the flag there and avoid breakage?22:22
@clarkb:matrix.orgya that works for me22:22
@jim:acmegating.comClark: could add the flag and change the default after a 2 week notice?22:22
@clarkb:matrix.orgoh thats a good idea.22:22
@clarkb:matrix.orgok I'll put that on my list for tomorrow morning to add a flag22:24
@jim:acmegating.comzuul-maint: how does this look for a zuul release (i would make it tomorrow if no issues come up): commit 77524b359cf427bca16d2a3339be9c1976755bc8 (HEAD -> master, tag: 5.2.5, origin/master, refs/changes/23/839423/1)22:25

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!