Wednesday, 2025-04-02

clarkbinfra-root the centos stream and ubuntu afs volumes are close to their quota limits. centos stream is closer than ubuntu so I don't think this will affect the release tomorrow. But maybe after the release tomorrow we can bump quotas?00:01
tonybclarkb: Sounds fair, I can use this as motivation to finally setup my AFS credentials00:05
opendevreviewMerged openstack/diskimage-builder master: CI: Add a tool for displaying CPU flags and QEMU version  https://review.opendev.org/c/openstack/diskimage-builder/+/93783605:24
fricklerinfra-root: looks like tox no longer thinks our pyenv-built python 3.13 matches its "py313" spec :-( https://zuul.opendev.org/t/openstack/build/5a97cfb50a754c8ea9e76bfa1cd5aac7/console06:23
opendevreviewDmitriy Rabotyagov proposed openstack/diskimage-builder master: Add support for building Fedora 40  https://review.opendev.org/c/openstack/diskimage-builder/+/92210907:14
opendevreviewLukas Kranz proposed zuul/zuul-jobs master: emit-job-header: add instance and region information  https://review.opendev.org/c/zuul/zuul-jobs/+/94603507:27
opendevreviewLukas Kranz proposed zuul/zuul-jobs master: emit-job-header: add instance and region information  https://review.opendev.org/c/zuul/zuul-jobs/+/94603508:53
opendevreviewBartosz Bezak proposed opendev/irc-meetings master: kolla: Move meeting one hour backwards (DST)  https://review.opendev.org/c/opendev/irc-meetings/+/94612609:53
fungifrickler: known issue for months, https://review.opendev.org/940158 should fix it if anyone wants to review that10:00
fricklerfascinating. +310:23
opendevreviewMerged zuul/zuul-jobs master: ensure-python: Skip "t" versions in pyenv  https://review.opendev.org/c/zuul/zuul-jobs/+/94015810:32
opendevreviewDmitriy Rabotyagov proposed openstack/diskimage-builder master: Add support for building Fedora 40, 41  https://review.opendev.org/c/openstack/diskimage-builder/+/92210911:31
opendevreviewMerged opendev/system-config master: Update .well-known/matrix/client for Element Call  https://review.opendev.org/c/opendev/system-config/+/94582614:02
opendevreviewMerged openstack/project-config master: Revert "Temporarily remove release docs semaphores"  https://review.opendev.org/c/openstack/project-config/+/94588514:33
opendevreviewDmitriy Rabotyagov proposed openstack/diskimage-builder master: Add support for building Fedora 40  https://review.opendev.org/c/openstack/diskimage-builder/+/92210914:55
opendevreviewDmitriy Rabotyagov proposed openstack/diskimage-builder master: Add support for building Fedora 40  https://review.opendev.org/c/openstack/diskimage-builder/+/92210914:56
fungiinfra-prod-service-gitea failed deploying for the element well-known file update14:56
fungiTASK [gitea : Run docker-compose pull] fatal: [gitea12.opendev.org]: FAILED! => ERROR: for memcached  Head "https://quay.io/v2/opendevmirror/memcached/manifests/latest": received unexpected HTTP status: 502 Bad Gateway14:58
clarkbI expect that means 09-11 are updated and 12-14 are not since we only updated one backend at a time serially with ansible14:58
fungilooks like it only hit that on gitea1214:58
fungiyes14:59
clarkbI suspect its safe to reenqueue the buildset as long as nothing else has merged that would conflict ( and I don't see anything)14:59
fungiyeah, doing that now14:59
clarkbhttps://status.redhat.com/15:00
clarkbmaybe we want to wait for that to be resolved first?15:00
clarkbI don't think it hurts to try it again and see what happens we'd just be in the same position we're in now worst case15:00
fungioh, oof15:00
fungiwell, i had already enqueued it by the time you pointed that out15:01
fungimaybe we'll get lucky15:01
clarkbya not a big deal. If it fails again we can wait for the status page to look ahppy before reenqueing again15:01
clarkband just avoid merging anything to system config until then15:01
fungibut if not, yeah i'll not reenqueue again until the all-clear15:01
clarkbI have confirmed gitea09 serves the new content15:02
fungicool15:02
clarkbso once the job is happy we should be done15:02
*** dmellado0755393737 is now known as dmellado15:03
fungiheading out for a quick lunch while we wait for quay to re-happy itself, bbiab15:04
clarkbhttps://zuul.opendev.org/t/openstack/buildset/4d4318f0ddc44d7baee76bc63514c214 it succeeded this time15:09
clarkbI'll check each server serves the new content15:09
clarkbI had to hard refresh gitea13's file because I had cached the old value so there may be some cache delay but otherwise this lgtm15:10
clarkbquay says the fix is implemented as of 15:19 UTC and they are monitoring. SO I think we're good to land other system-config changes that rely on quay if we like as well15:23
opendevreviewMichael Johnson proposed openstack/diskimage-builder master: Remove deprecated datetime.utcnow()  https://review.opendev.org/c/openstack/diskimage-builder/+/94600716:14
fungiokay, back and catching up16:31
fungilooks like my reenqueue of 945826,1 succeeded anyway16:31
fungiah, you also noted that16:32
clarkbyup I think everything should be caught up and happy now16:33
opendevreviewJeremy Stanley proposed opendev/yaml2ical master: Add pytz to requirements  https://review.opendev.org/c/opendev/yaml2ical/+/94617917:37
opendevreviewJeremy Stanley proposed opendev/yaml2ical master: Remove unused extras  https://review.opendev.org/c/opendev/yaml2ical/+/91973617:45
opendevreviewMerged opendev/yaml2ical master: Add pytz to requirements  https://review.opendev.org/c/opendev/yaml2ical/+/94617917:47
opendevreviewMerged opendev/yaml2ical master: Remove unused extras  https://review.opendev.org/c/opendev/yaml2ical/+/91973617:59
clarkbI've been trying to keep an eye on gitea response times today as historically the day of an opesntack release has been a high load situation for gitea. So far all of my checks have looked good18:05
clarkblooking at grafana I don't think we're seeing any abnormal traffic though18:06
clarkbgitea13 continues to be our most used backend18:06
clarkbjust the luck of the ip addr draw and NAT  Iguess18:06
opendevreviewMichael Johnson proposed openstack/diskimage-builder master: Remove deprecated datetime.utcnow()  https://review.opendev.org/c/openstack/diskimage-builder/+/94600721:04
clarkbjohnsom: ^ one small thing on that but I like this approach much better21:06
johnsomclarkb ack21:06
johnsomArgh, I always forget that double stack.21:07
fungialso there's a thread on discuss.p.o about un-deprecating it21:08
opendevreviewMichael Johnson proposed openstack/diskimage-builder master: Remove deprecated datetime.utcnow()  https://review.opendev.org/c/openstack/diskimage-builder/+/94600721:09
fungihttps://discuss.python.org/t/why-is-datetime-utcnow-deprecated/8686821:09
clarkbfwiw I think in this case using the new form is fine and we can sidestep the entire discussion21:10
fungiseems some of the core devs aren't thrilled with the decision to deprecate it in the first place21:10
clarkbI agree deprecating something beacuse the timezone wasn't explicit when its literally utcnow() seems weird21:10
clarkbfungi: ya the arguemtn to deprecate was that datetime methods should be timezone explicit. Then they deprecated the one thing that is timezone explicit already21:10
fungithe suggestion to document the pitfalls it has makes more sense to me21:11
opendevreviewJeremy Stanley proposed opendev/bindep master: Fix authors/maintainers format in pyproject.toml  https://review.opendev.org/c/opendev/bindep/+/94621822:37
opendevreviewJeremy Stanley proposed opendev/engagement master: Fix authors/maintainers format in pyproject.toml  https://review.opendev.org/c/opendev/engagement/+/94621922:37
clarkbfungi: can you not do {name = foo,\n email = bar} in ^22:40
clarkbhttps://github.com/toml-lang/toml/pull/904 looks like it depends on your version of toml22:42
clarkbI don't know how that factors into pip and setuptools and so on so maybe it is best to leave this as is22:42
cardoeIs there a reason to use opendev-build-docker-image vs opendev-build-container-image ?23:05
cardoeTrying to refactor that loci repo I mentioned a few weeks back. It's uploading to docker hub and to quay23:05
cardoeI'll want to get another quay org setup cause we're just stuffing things into airshipit which seems wrong.23:05
tonybcardoe: I don't think so.  The former is opinionated about using docker tools specifically whereas the latter is more general and probably podman23:06
tonybcardoe: I think in general I'd pick opendev-build-container-image23:07
tonybcardoe: Hmm unless you have a need to run the container build infra on <= Jammy ... I think podman needs >= Noble23:08
cardoeThanks.23:11
cardoeIt looks like I can override the backend to docker if I need23:11
tonybYup.  Good luck23:17
tonybfungi: I left a question, but basically LGTM feel free to +A both 94621[89] if I'm wrong23:21
tonybfungi: Also Sorry I wasn't around for the realease, it seems I did very bad math and it wasn't 2100 (local) it was 0200 :/23:25
clarkbtonyb: the good news is everything went really smoothly23:32

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