| @clarkb:matrix.org | corvus: I'm a derp I was testing with python 3.8 but the deprecation doesn't happen until 3.10. Rerunning against 3.10 (I should install 3.11 already) allows me to check locally as well. Thank you for checking too. I'll get a kazoo PR up momentarily | 00:23 |
|---|---|---|
| @clarkb:matrix.org | https://github.com/python-zk/kazoo/pull/706 | 00:29 |
| -@gerrit:opendev.org- Zuul merged on behalf of Ian Wienand: [zuul/zuul-jobs] 872371: linters-requirements : update Ansible to 2.12, ansible-lint <6.12.0 https://review.opendev.org/c/zuul/zuul-jobs/+/872371 | 00:34 | |
| @iwienand:matrix.org | what are we testing with zuul-jobs tox-py27? https://zuul.opendev.org/t/zuul/build/91fb3562c8204422a2266275e0dca2d5/console | 00:43 |
| -@gerrit:opendev.org- Zuul merged on behalf of Ian Wienand: [zuul/zuul-jobs] 872489: ansible-lint: Ignore some command-instead-of-module warnings https://review.opendev.org/c/zuul/zuul-jobs/+/872489 | 00:44 | |
| @clarkb:matrix.org | ianw: some deployments of zuul may still run Ansible under 2.7 (on the remote side) | 00:48 |
| @clarkb:matrix.org | I think we're checking things generally function under that python version to support those users? | 00:48 |
| -@gerrit:opendev.org- Ian Wienand proposed: | 00:54 | |
| - [zuul/zuul-jobs] 872365: zuul-jobs-test-registry-docker-* : update to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872365 | ||
| - [zuul/zuul-jobs] 872490: ansible-lint: fix a bunch of command-instead-of-shell errors https://review.opendev.org/c/zuul/zuul-jobs/+/872490 | ||
| - [zuul/zuul-jobs] 872491: ansible-lint: add names to blocks/includes, etc. https://review.opendev.org/c/zuul/zuul-jobs/+/872491 | ||
| - [zuul/zuul-jobs] 872492: ansible-lint: ignore use of mkdir https://review.opendev.org/c/zuul/zuul-jobs/+/872492 | ||
| - [zuul/zuul-jobs] 872493: ansible-lint: use pipefail https://review.opendev.org/c/zuul/zuul-jobs/+/872493 | ||
| - [zuul/zuul-jobs] 872494: ansible-lint: ignore latest git pull https://review.opendev.org/c/zuul/zuul-jobs/+/872494 | ||
| - [zuul/zuul-jobs] 872495: ansible-lint: uncap https://review.opendev.org/c/zuul/zuul-jobs/+/872495 | ||
| - [zuul/zuul-jobs] 872258: build-docker-image: fix change prefix https://review.opendev.org/c/zuul/zuul-jobs/+/872258 | ||
| - [zuul/zuul-jobs] 872375: container-roles-jobs: Update tests to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872375 | ||
| @iwienand:matrix.org | yeah some libraries might, i guess. console streaming etc. but i wonder if stuff that failed in that job like File "/home/zuul/src/opendev.org/zuul/zuul-jobs/roles/test-upload-logs-swift/library/test_zuul_swift_upload.py", line 462, in side_effect can run on py27 | 00:55 |
| @iwienand:matrix.org | anyway, see how that reorged stack goes | 00:56 |
| @iwienand:matrix.org | well maybe it's not py27 | 01:04 |
| @iwienand:matrix.org | ``` | 01:04 |
| File "/home/zuul/src/opendev.org/zuul/zuul-jobs/roles/upload-logs-base/library/test_index.py", line 473, in side_effect | ||
| raise requests.exceptions.RequestException( | ||
| requests.exceptions.RequestException: Failed for a reason | ||
| ``` | ||
| @iwienand:matrix.org | i'm glad it failed for a reason, and not for no reason | 01:04 |
| @iwienand:matrix.org | i think that's a red herring | 01:35 |
| @iwienand:matrix.org | here's a big part of the issue | 01:35 |
| @iwienand:matrix.org | https://packages.ubuntu.com/jammy/skopeo | 01:35 |
| @iwienand:matrix.org | Package: skopeo (1.4.1+ds1-1) [universe] | 01:36 |
| @iwienand:matrix.org | another ancient version in the LTS | 01:36 |
| -@gerrit:opendev.org- Ian Wienand proposed: | 03:25 | |
| - [zuul/zuul-jobs] 872365: zuul-jobs-test-registry-docker-* : update to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872365 | ||
| - [zuul/zuul-jobs] 872617: ensure-skopeo: add install from upstream option https://review.opendev.org/c/zuul/zuul-jobs/+/872617 | ||
| -@gerrit:opendev.org- Ian Wienand proposed: | 03:37 | |
| - [zuul/zuul-jobs] 872617: ensure-skopeo: add install from upstream option https://review.opendev.org/c/zuul/zuul-jobs/+/872617 | ||
| - [zuul/zuul-jobs] 872365: zuul-jobs-test-registry-docker-* : update to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872365 | ||
| -@gerrit:opendev.org- Ian Wienand proposed: | 03:41 | |
| - [zuul/zuul-jobs] 872617: ensure-skopeo: add install from upstream option https://review.opendev.org/c/zuul/zuul-jobs/+/872617 | ||
| - [zuul/zuul-jobs] 872365: zuul-jobs-test-registry-docker-* : update to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872365 | ||
| -@gerrit:opendev.org- Ian Wienand proposed: | 03:45 | |
| - [zuul/zuul-jobs] 872617: ensure-skopeo: add install from upstream option https://review.opendev.org/c/zuul/zuul-jobs/+/872617 | ||
| - [zuul/zuul-jobs] 872365: zuul-jobs-test-registry-docker-* : update to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872365 | ||
| -@gerrit:opendev.org- Ian Wienand proposed: | 04:06 | |
| - [zuul/zuul-jobs] 872617: ensure-skopeo: add install from upstream option https://review.opendev.org/c/zuul/zuul-jobs/+/872617 | ||
| - [zuul/zuul-jobs] 872365: zuul-jobs-test-registry-docker-* : update to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872365 | ||
| - [zuul/zuul-jobs] 872490: ansible-lint: fix a bunch of command-instead-of-shell errors https://review.opendev.org/c/zuul/zuul-jobs/+/872490 | ||
| - [zuul/zuul-jobs] 872491: ansible-lint: add names to blocks/includes, etc. https://review.opendev.org/c/zuul/zuul-jobs/+/872491 | ||
| - [zuul/zuul-jobs] 872492: ansible-lint: ignore use of mkdir https://review.opendev.org/c/zuul/zuul-jobs/+/872492 | ||
| - [zuul/zuul-jobs] 872493: ansible-lint: use pipefail https://review.opendev.org/c/zuul/zuul-jobs/+/872493 | ||
| - [zuul/zuul-jobs] 872494: ansible-lint: ignore latest git pull https://review.opendev.org/c/zuul/zuul-jobs/+/872494 | ||
| - [zuul/zuul-jobs] 872495: ansible-lint: uncap https://review.opendev.org/c/zuul/zuul-jobs/+/872495 | ||
| - [zuul/zuul-jobs] 872258: build-docker-image: fix change prefix https://review.opendev.org/c/zuul/zuul-jobs/+/872258 | ||
| - [zuul/zuul-jobs] 872375: container-roles-jobs: Update tests to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872375 | ||
| @iwienand:matrix.org | there's definitely some sort of race in the zuul-jobs tox runs | 04:25 |
| @iwienand:matrix.org | the ```requests.exceptions.RequestException: Failed for a reason``` stuff is a red-herring; that's injected errors | 04:26 |
| -@gerrit:opendev.org- Ian Wienand proposed: | 04:37 | |
| - [zuul/zuul-jobs] 872493: ansible-lint: use pipefail https://review.opendev.org/c/zuul/zuul-jobs/+/872493 | ||
| - [zuul/zuul-jobs] 872494: ansible-lint: ignore latest git pull https://review.opendev.org/c/zuul/zuul-jobs/+/872494 | ||
| - [zuul/zuul-jobs] 872495: ansible-lint: uncap https://review.opendev.org/c/zuul/zuul-jobs/+/872495 | ||
| - [zuul/zuul-jobs] 872258: build-docker-image: fix change prefix https://review.opendev.org/c/zuul/zuul-jobs/+/872258 | ||
| - [zuul/zuul-jobs] 872375: container-roles-jobs: Update tests to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872375 | ||
| @iwienand:matrix.org | corvus: Clark : all but the last change to merge of https://review.opendev.org/q/topic:ansible-lint-req-2.12 is green now. the last change to merge is one that updates a bigger bunch of the registry jobs to jammy -- i'll look into that on monday. most are reviewed, but there's a new skopeo build and some reorganising. | 07:40 |
| @mhuin:matrix.org | > <@jkt_:matrix.org> so I just got biten in my rear by a typo in my commit message, I said `Depend-on` instead of `Depends-on`. What would be the easiest way of writing a super-simple job which lints the commit message? I guess I still need a VM for that, and that I cannot "just" run something on the executor, right? | 07:56 |
| here's an alternative if you host your code on gerrit: I think it's possible to set custom pre-commit hooks in addition to the default one that adds a ChangeId to the commit message. you could add a lint check there | ||
| -@gerrit:opendev.org- Zuul merged on behalf of Simon Westphahl: [zuul/zuul] 871107: Periodically cleanup leaked pipeline state https://review.opendev.org/c/zuul/zuul/+/871107 | 09:09 | |
| -@gerrit:opendev.org- Matthieu Huin https://matrix.to/#/@mhuin:matrix.org proposed: [zuul/zuul] 835452: Test zuul-client dequeue-all https://review.opendev.org/c/zuul/zuul/+/835452 | 09:39 | |
| @jkt_:matrix.org | mhu: yeah, perhaps a client-side thingy will be enough after all | 11:05 |
| -@gerrit:opendev.org- Zuul merged on behalf of Simon Westphahl: [zuul/zuul] 871108: Cleanup deleted pipelines and and event queues https://review.opendev.org/c/zuul/zuul/+/871108 | 13:56 | |
| @elpell:matrix.org | Hi! I just had a change (A) which zuul failed to schedule in gate with the following messages: | 16:26 |
| Failed to enqueue changes ahead of A | ||
| This was caused by a broken relation chain between change A and B (change B had been merged earlier). However, zuul never reported back to gerrit that it had failed (after 20 minutes we rebased and it was eventually picked up by gate). So I was wondering if this behavior is expected and if so why? And if not, that would be great to know as well of course! | ||
| @clarkb:matrix.org | where was the message located if it didn't report back to gerrit? | 16:30 |
| @clarkb:matrix.org | is that from the scheduler log? | 16:30 |
| @clarkb:matrix.org | I was asking because if so we did find an issue where gerrit with an http connection set up cannot properly report to old patchests | 16:58 |
| @clarkb:matrix.org | which may explain this depending on how things were enqueued | 16:59 |
| @clarkb:matrix.org | you'd need to look for reporting errors in the scheduler log | 16:59 |
| @jim:acmegating.com | in general, zuul explains its reasons for not doing things in the scheduler log | 17:01 |
| @fungicide:matrix.org | > <@clarkb:matrix.org> fungi: (or anyone else) I'm looking at startinga patch to kazoo to fix this SSLv23 deprecation warning. It looks like the things we are supposed to use is PROTOCOL_TLS_CLIENT but that enables hostname checking. We connect to our servers via IP addrs and I don't think our certs have IP addrs as altnames. Does this imply that python is basically saying you should not do ssl to something with ip addrs. Or I guess we must stick IP addrs in the altnames? | 17:07 |
| i agree that hostname checking will likely cause cert validation errors when connecting by ip address, unless ip address based connections get handled specially somehow | ||
| @clarkb:matrix.org | fungi: ya I'm pretty sure they don't beacuse python ssl forces you to pass in a hostname to check when that flag is set | 17:10 |
| @clarkb:matrix.org | fungi: anyway I think I ended up with a backward compatible change that removes deprecation warnings. Basically by using the new thing then tuning back some of its extra security requirements | 17:10 |
| @fungicide:matrix.org | sounds good to me | 17:11 |
| @fungicide:matrix.org | and yes, i expect there should be a way to include the raw ip addresses as subject altnames if it should be come necessary, though i've never tried that | 17:12 |
| @clarkb:matrix.org | yes I googled a bit and ips are valid in that field apparently | 17:14 |
| @elpell:matrix.org | > <@clarkb:matrix.org> is that from the scheduler log? | 17:14 |
| Yes it was from the scheduler logs i found what was going on! | ||
| @clarkb:matrix.org | Per Wiklund: out of curiousity what was it? | 17:15 |
| @elpell:matrix.org | > <@clarkb:matrix.org> Per Wiklund: out of curiousity what was it? | 17:18 |
| I was never able to find out sorry, the last log entries was that the enqueue cycle was empty and that it was trying the last call was an addchange. At the moment im not able to send log extracts im afraid | ||
| @clarkb:matrix.org | you have to laugh because if you don't.... It looks like podman has updated and broken too: https://zuul.opendev.org/t/zuul/build/048d996f145543d3a4ef75a673b97089 ? | 18:35 |
| @clarkb:matrix.org | I'm guessing something broke non root containers | 18:35 |
| @clarkb:matrix.org | oh this changes the underlying node type to something newer so maybe podman has just been broken on jammy with unprivileged containers? | 18:37 |
| @clarkb:matrix.org | thats at the end of the stack though and everything else looks mergeeable at this point | 18:39 |
| @clarkb:matrix.org | https://github.com/containers/podman/issues/7650 this says it is likely a combo of too old systemd and cgroupv2 | 18:44 |
| @clarkb:matrix.org | there is a workaround there which I'll try pushing up I guess | 18:45 |
| -@gerrit:opendev.org- Clark Boylan proposed on behalf of Ian Wienand: [zuul/zuul-jobs] 872375: container-roles-jobs: Update tests to jammy nodes https://review.opendev.org/c/zuul/zuul-jobs/+/872375 | 18:54 | |
| -@gerrit:opendev.org- Clark Boylan proposed: [zuul/zuul-jobs] 872698: Fix ansible-lint no-jinja-when occurence https://review.opendev.org/c/zuul/zuul-jobs/+/872698 | 21:05 | |
| @clarkb:matrix.org | That change is outside of ianw's stack so that I don't have to restack the whole thing to get tests to pass | 21:05 |
| @clarkb:matrix.org | but I think it will fix the ansible-lint issue that 872375 hit. 872375 should be fine otherwise | 21:05 |
| @clarkb:matrix.org | We have docker 23 release notes now: https://docs.docker.com/engine/release-notes/23.0/ One thing I notice is that buildx is now the default on linux. Which is surprising to me because our buildx builds failed because buildx was not installed because they split it into a separate package and didn't require it. I guess it may be the default if buildx is present | 21:08 |
| @clarkb:matrix.org | skimming the list there I don't see naything too scary. https://github.com/moby/moby/pull/42501 is interesting though (basically images will always get seccomp now?) | 21:14 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!