| fungi | does this urllib3 exception look familiar to anyone? https://zuul.opendev.org/t/openstack/build/9ac4f82d7d45469fa71be79b054067fa | 00:22 |
|---|---|---|
| fungi | too-old requests on too-new python? | 00:22 |
| fungi | oh, this might be docker client related too | 00:23 |
| ianw | fungi: is this the same problem as https://review.opendev.org/c/opendev/system-config/+/882171 | 00:25 |
| tonyb | So WRT to decrecating Fedora. We have removed the f37 stuff from mirrors because it was unused, now in parallel we can remove the f36 content from mirrors and disable/delete jobs using f36. Once the second part of that is done we can remove the fedora diskimages and any lingering content is that correct? | 01:03 |
| opendevreview | melanie witt proposed openstack/project-config master: launchpad: Use repo name to match project + series https://review.opendev.org/c/openstack/project-config/+/883603 | 01:13 |
| ianw | tonyb: first we'll want to make sure we stop existing fedora jobs using the mirrors | 01:13 |
| ianw | the particularly susceptible one is zuul-jobs i'd say | 01:14 |
| ianw | we don't have to stop fedora testing there, just stop it pulling from the opendev mirrors | 01:15 |
| ianw | it should be as simple as pulling out the fedora path in the mirror setup. if it *is* that simple remains to be seen :) | 01:16 |
| tonyb | ianw: I guess I assumed, incorrectly, that the diskimages/jobs used our mirrors in preference to the distro specific ones. | 01:41 |
| tonyb | SO to stop the jobds from using our mirrors we need to update/remove https://opendev.org/openstack/diskimage-builder/src/branch/master/roles/dib-setup-gate-mirrors/tasks/main.yaml#L68 and rebuiuld the images? or something else? | 01:42 |
| opendevreview | Tony Breeds proposed opendev/system-config master: [docs] Use RST url link syntax to improve layout https://review.opendev.org/c/opendev/system-config/+/883617 | 02:43 |
| ianw | tonyb: umm, we'd want to pull out ... let me find it ... | 02:59 |
| ianw | https://opendev.org/zuul/zuul-jobs/src/branch/master/roles/configure-mirrors/tasks/mirror/Fedora.yaml | 03:00 |
| tonyb | ianw: Thanks. I'll go read up on that. | 03:01 |
| ianw | DIB should leave the images pointing to generic mirrors. Some builds have been setup to build against the opendev mirrors, with that role. But I think only ubuntu-ish ones | 03:09 |
| ianw | *should* :) | 03:09 |
| tonyb | Yup that makes way more sense. | 03:10 |
| * tonyb prepares to fire off a change | 03:11 | |
| ianw | centos definitely doesn't, because it's been a persistent problem that we build from upstream, but then something has happened to the mirror we clone from so the opendev.org mirrors are behind, which ends up unhappy | 03:11 |
| ianw | s/unhappy/with unhappy centos images/ | 03:11 |
| tonyb | Yeah I can see how that'd suck | 03:13 |
| ianw | a big clue is usually the wheel builds. if the centos build fails, the first thing to check is that. what happens is the booted kernel has moved on from the old kernel-headers package in our mirror, and so building openafs fails | 03:14 |
| ianw | of course, if we can figure out kafs that might be a thing of the past :) | 03:15 |
| opendevreview | Merged opendev/system-config master: [docs] Use RST url link syntax to improve layout https://review.opendev.org/c/opendev/system-config/+/883617 | 03:16 |
| tonyb | ianw: Good to know. | 03:32 |
| tonyb | Okay so I'm probably overthinking it but I don't think we can just disable https://opendev.org/zuul/zuul-jobs/src/branch/master/roles/configure-mirrors/tasks/mirror/Fedora.yaml because zuul-jobs is use by everyone not just opendev right? | 03:35 |
| tonyb | So I'm thinking I shoudl add a ${DISTRO}_mirror_enabled : True variable and then in the opendev/base/pre.yaml we'd set that to Fedora_mirror_enabled to False | 03:36 |
| ianw | i think the theory is that without that, we should just be using upstream mirrors | 03:45 |
| ianw | so we should be able to pull it out transparently | 03:47 |
| tonyb | but if another zuul user is running on Fedora and has a Fedora mirror wont I alter their runs if I just pull it out of zuul-jobs ? | 03:50 |
| tonyb | I did start by stating I'm probably overthnking it | 03:51 |
| ianw | no this is a good point | 03:54 |
| ianw | if we're going to deprecate fedora from zuul-jobs, by our own policies it does require a notice and grace period | 03:55 |
| tonyb | I didn't think we were deprecating fedora in zuul, I thought we were deprecating them in opendev, which may also require a notice and grace period. | 03:56 |
| tonyb | that's why I was thinking that I'd create a way to just not do it on Fedora for opendev | 03:57 |
| ianw | right, but if the nodes go away in opendev and zuul isn't testing on fedora, then it's as good as broken | 03:57 |
| ianw | s/zuul/zuul-jobs/ | 03:58 |
| tonyb | Ahhhh right now I see the angle | 03:58 |
| tonyb | it's more circular than I appreciated | 03:59 |
| ianw | i think your inclination towards configure-mirrors having a flag to actually ignore a distro is probably right. that means no change to zuul-jobs extant behaviour, but then we can switch it off in https://opendev.org/opendev/base-jobs/src/commit/3fc688b08dbe2ff41a75f051f53b4929dd35800f/playbooks/base/pre.yaml#L63 | 04:03 |
| tonyb | ianw: Okay. I'll write that up. We can then look at the case/value of Fedora images in OpenDev, without mirroring. | 04:05 |
| *** dmellado90 is now known as dmellado9 | 05:09 | |
| *** amoralej|off is now known as amoralej | 06:20 | |
| *** timburke_ is now known as timburke | 06:46 | |
| opendevreview | Merged openstack/project-config master: grafana: Update kolla dashboards with proper job names https://review.opendev.org/c/openstack/project-config/+/883234 | 07:15 |
| *** amoralej is now known as amoralej|lunch | 12:34 | |
| opendevreview | Jeremy Stanley proposed opendev/gerritlib master: Pin python requests when installing docker-compose https://review.opendev.org/c/opendev/gerritlib/+/883664 | 12:55 |
| *** amoralej|lunch is now known as amoralej | 13:06 | |
| opendevreview | Jeremy Stanley proposed opendev/gerritlib master: Pin python requests when installing docker-compose https://review.opendev.org/c/opendev/gerritlib/+/883664 | 13:10 |
| opendevreview | Merged openstack/project-config master: Add cinder-nfs charm to OpenStack charms https://review.opendev.org/c/openstack/project-config/+/879469 | 13:14 |
| opendevreview | Jeremy Stanley proposed opendev/gerritlib master: Pin python requests when installing docker-compose https://review.opendev.org/c/opendev/gerritlib/+/883664 | 13:43 |
| opendevreview | Jeremy Stanley proposed opendev/jeepyb master: update_bug.py: Check project along with series https://review.opendev.org/c/opendev/jeepyb/+/883597 | 14:12 |
| opendevreview | Tobias Rydberg proposed opendev/irc-meetings master: Changing meeting time to 0700 UTC over summer. https://review.opendev.org/c/opendev/irc-meetings/+/883675 | 14:17 |
| opendevreview | Merged opendev/irc-meetings master: Changing meeting time to 0700 UTC over summer. https://review.opendev.org/c/opendev/irc-meetings/+/883675 | 14:41 |
| clarkb | no complaints from certcheck today for wiki so that all looks good and done now | 15:35 |
| fungi | yep, thanks! | 16:00 |
| *** timburke_ is now known as timburke | 16:48 | |
| *** amoralej is now known as amoralej|off | 16:55 | |
| *** melwitt_ is now known as melwitt | 17:11 | |
| opendevreview | Jeremy Stanley proposed opendev/gerritlib master: Switch from tox to nox https://review.opendev.org/c/opendev/gerritlib/+/871847 | 17:14 |
| opendevreview | Jeremy Stanley proposed opendev/jeepyb master: Switch from tox to nox https://review.opendev.org/c/opendev/jeepyb/+/871653 | 17:16 |
| clarkb | fungi: I approved the gerritlib docker-compose requests fix. Not sure how we want to handle tox -> nox I suspect the two of us may be the ones who acre the most so if we want to do it we can just go for it? | 18:34 |
| fungi | yeah, i just wanted to make sure it was still working and then i was going to approve | 18:38 |
| clarkb | ++ | 18:38 |
| fungi | this was mostly in service of being able to merge melwitt's bug update script fix in jeepyb | 18:38 |
| fungi | https://review.opendev.org/883597 if you're interested in another very tiny review | 18:41 |
| opendevreview | Merged opendev/gerritlib master: Pin python requests when installing docker-compose https://review.opendev.org/c/opendev/gerritlib/+/883664 | 18:42 |
| opendevreview | Merged opendev/gerritlib master: Switch from tox to nox https://review.opendev.org/c/opendev/gerritlib/+/871847 | 18:49 |
| opendevreview | Merged opendev/jeepyb master: Raise an error if acl pushes fail https://review.opendev.org/c/opendev/jeepyb/+/869873 | 19:27 |
| Clark[m] | fungi: I had forgotten about 869873. I thought there was a reason to not merge it, but now I can't remember and doesn't seem it got added to the change. Maybe it was just the read only issue I pushed a fix for | 19:36 |
| Clark[m] | Anyway probably fine to just watch it and make sure it does what we want | 19:37 |
| fungi | yep, i planned to keep a close eye on the next acl additions/changes | 19:38 |
| opendevreview | Merged opendev/jeepyb master: Switch from tox to nox https://review.opendev.org/c/opendev/jeepyb/+/871653 | 19:54 |
| opendevreview | Merged opendev/jeepyb master: update_bug.py: Check project along with series https://review.opendev.org/c/opendev/jeepyb/+/883597 | 20:44 |
| Clark[m] | Gerrit 3.8 has just released | 23:12 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!