Tuesday, 2024-08-06

*** elodilles_pto is now known as elodilles05:51
fricklermost of the infra-prod stuff finally worked fine tonight, I'm checking the meetpad log now https://zuul.opendev.org/t/openstack/buildset/058cef6980624f6899ee7855d78e60e207:23
fricklertonyb: looks like meetpad is broken, not sure if that is the cause or the result of the playbook failure https://paste.opendev.org/show/brMSLYlVNO9K5BztRAZr/ if you have time to take a look that would be nice07:30
tonybfrickler: I can look in a bit.  I'm currently celebrating a friend's birthday07:57
tonybfrickler: that looks like the podman issue with restart vs stop/start.07:58
tonybdoing the latter will fix it.07:58
fricklerhmm, just doing "docker-compose up -d" gives the same error. is it safe to do "docker-compose down" first? also don't IRC while partying ;)08:34
tonybyeah the containers are already nonfunctional so doing the docker-compose down is fine.09:18
tonybthen docker-compose up -d should work09:18
fricklerok, seems to be working again09:22
opendevreviewDong Zhang proposed zuul/zuul-jobs master: Print instance type in emit-job-header role  https://review.opendev.org/c/zuul/zuul-jobs/+/92575410:36
fungiopenssh 9.8p1 entered debian/sid today, stops compiling dsa key support entirely (can't be turned on in non-default configuration any longer)14:39
fungiparticularly interesting is that sshd "will now block client addresses that repeatedly fail authentication, repeatedly connect without ever completing authentication or that crash the server"14:45
fungithat's an interesting scope increase for the server, doing stuff that you'd usually do with e.g. fail2ban14:45
*** ykarel is now known as ykarel|away15:29
fungitonyb: not sure if you saw the response over the weekend/early your monday, but Ramereth[m] increased our quota (thanks Ramereth[m]!!!), so should be room there to work on the mirror server replacement there now16:10
JayFfungi: that's super exciting tbh16:23
fungiJayF: the new opensshd features? i agree16:23
JayFabsolutely16:24
JayFit always seemed weird that it wasn't a first-class thing16:24
fungii'm looking forward to seeing how it goes on my personal servers16:24
JayFyeah, this is making me want to pull it down on my public-facing gentoo box that runs ssh16:24
fungii haven't looked into the details yet, but knowing the openbsd/openssh devs there are almost certainly a slew of tunables for those new behaviors16:26
JayFoh yeah, part of why I'm excited is I trust them to do it well16:27
JayF9.8 is in gentoo already too, score16:27
JayFupgrading to non-stable opensshd, what could go wrong /s16:27
fungithere's some irony in the fact that i won't have it on my openbsd machines until openbsd 7.6 releases in a couple of months16:28
JayFtime to install openssh in a gentoo prefix on your openbsd box for maximum recursive irony16:28
fungiwell, i mean, i could build 9.8 from -next more easily than that ;)16:30
JayFdoes obsd have something like ports from fbsd?16:33
fungiyes16:33
JayFI haven't run an openbsd machine in 20 years16:33
fungithere is a ports tree, but also autobuilt packages and i opt for the latter16:34
JayFand I'm pretty sure configuring CARP + PF on that machine was easier than it is on any modern OS today 16:34
fungibut openssh is in the main source tree for openbsd anyway, not in ports16:34
JayFthat's more just me showing my ignorance about how bsd arranges packages :)16:34
fungiyeah, typical approach for the *bsds is that the base operating system (kernel and userland) are in one giant source tree that all gets built together to make the operating system, and then anything that isn't officially part of the os may have "ports" (essentially build recipes/makefiles) and also "packages" (pre-built bundles of things from the ports tree)16:38
JayFI think my assumption was that sshd wouldn't be core-system level stuff16:40
JayFbut again, I think that's linux-brain :) distribute ALL THE THINGS! 16:40
fungiyeah, it may not be part of the core operating system in other *bsds, but in openbsd's case they're the developers of openssh16:43
opendevreviewJeremy Stanley proposed opendev/system-config master: Stop null-routing mail for edge-computing-owner  https://review.opendev.org/c/opendev/system-config/+/92579017:29
opendevreviewAndy Wu proposed openstack/project-config master: Mirror cinder-nfs charms to GitHub  https://review.opendev.org/c/openstack/project-config/+/92579117:37
tonybfungi: I did see that, just haven't tested it yet to confirm19:05
tonybAs we're skipping the meeting, I'll drop this https://review.opendev.org/q/topic:%22noble-mirror%22 and ask for reviews19:06
fungithanks tonyb!19:12
funginoonedeadpunk: we can catch up whenever you're available, but prior discussions about rocky mirroring involved NeilHanlon and mnasiadka. the topic hasn't come up that i can see since the end of 2023, but we've been engaged in cleaning up some older mirrored content (fedora, centos stream 8, soon ubuntu xenial) so likely either have sufficient room now or will in the near future19:19
opendevreviewAndy Wu proposed openstack/project-config master: Mirror cinder-nfs charms to GitHub  https://review.opendev.org/c/openstack/project-config/+/92579119:23
opendevreviewMerged openstack/project-config master: taas: Stop gating for tap-as-a-service-dashboard  https://review.opendev.org/c/openstack/project-config/+/92381619:38
opendevreviewMerged openstack/project-config master: Mirror cinder-nfs charms to GitHub  https://review.opendev.org/c/openstack/project-config/+/92579119:38
noonedeadpunkI'm a bit /o\ right now with things but eventually would be interested in having rocky mirrors as well. Will catch up with NeilHanlon as well once he's back as well20:31
noonedeadpunkbtw, fwiw there's some issue with mirrors in OVH: https://zuul.opendev.org/t/openstack/build/e74fc63cc250464d87075381e075fa7f20:33

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