15:05:57 <noonedeadpunk> #startmeeting openstack_ansible_meeting
15:05:57 <opendevmeet> Meeting started Tue Sep 16 15:05:57 2025 UTC and is due to finish in 60 minutes.  The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:05:57 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:05:57 <opendevmeet> The meeting name has been set to 'openstack_ansible_meeting'
15:06:07 <noonedeadpunk> #topic rollcall
15:06:25 <noonedeadpunk> d a m i a n d a b r o w s k i is away this week
15:06:52 <noonedeadpunk> courtesy ping jrosser NeilHanlon_
15:07:07 <jrosser> o/
15:07:09 <jrosser> hello
15:07:18 <jrosser> and thanks, thats helpful
15:07:30 * jrosser was just about to head to the computer room :)
15:11:40 <noonedeadpunk> :D
15:11:45 <noonedeadpunk> #topic office hours
15:12:13 <noonedeadpunk> quite a while ago folks asked to be stopped pinged, so yeah
15:12:18 <noonedeadpunk> ok
15:12:58 <noonedeadpunk> So in EPEL systemd-networkd went to testing, which means that in around 10 day it become available from there
15:13:28 <noonedeadpunk> and this should enable us to get some progress with Rocky 10 support
15:13:46 <noonedeadpunk> at the same time there are discussions about also adding Alma Linux images to CI
15:14:19 <noonedeadpunk> on the positive note - it does not require more modern architecture alike to CentOS and Rocky linux, so it can be spawned on older hardware
15:14:42 <noonedeadpunk> thus I was thinking on adding like... 1 lxc job or smth like that on Alma?
15:14:54 <noonedeadpunk> (once it become available)
15:15:07 <noonedeadpunk> as I was asked already a couple of times about alma support
15:15:55 <jrosser> we can probably start with the infra jobs and whatever
15:16:48 <noonedeadpunk> yeah. I'd guess it should be just the same as rocky... but well
15:17:16 <noonedeadpunk> on the same topic the question is - do we wanna make 31.1.0 tag now with just CentOS being backported?
15:17:26 <noonedeadpunk> or wait for networkd and EPEL
15:18:11 <jrosser> i can not really remember totally where we got to
15:18:40 <noonedeadpunk> so CentOS 10 Stream has couple of CI jobs on Epoxy
15:18:54 <noonedeadpunk> so pretty much things are fixed right now
15:19:29 <noonedeadpunk> the problem is, that systemd-networkd is coming from a SIG, which means that systemd as a whole is majorly upgraded in the process
15:20:06 <jrosser> hmm yes and thats not great
15:20:27 <noonedeadpunk> https://opendev.org/openstack/ansible-role-systemd_networkd/src/branch/master/vars/redhat-10.yml#L40-L51
15:20:28 <jrosser> well centos 10 choice is not great anyway but thats a different matter.....
15:20:36 <noonedeadpunk> so it would be hard to rollback from that once released
15:20:44 <noonedeadpunk> right
15:25:57 <jrosser> maybe i would wait
15:26:14 <jrosser> as thats not a particularly good thing to put into the .1 release
15:26:16 <noonedeadpunk> ++ yeah, I'm also inclined towards this
15:26:24 <noonedeadpunk> given that "due date" is known
15:26:29 <jrosser> but i don't know what we do if there is some other unexpdected delay
15:26:51 <noonedeadpunk> like absence of openvswitch ?:)
15:28:22 <jrosser> well right
15:30:11 <noonedeadpunk> ok, so next thing is MariaDB 11.8
15:30:32 <noonedeadpunk> kolla has decided to stick with having 11.4 due to the mess with charsets and collations
15:31:13 <noonedeadpunk> at the same time, I believe that this change does the trick so far
15:31:16 <noonedeadpunk> #link https://review.opendev.org/c/openstack/openstack-ansible-galera_server/+/959075
15:41:49 <noonedeadpunk> I think I've answered Sean's comment
15:42:15 <noonedeadpunk> and keeping the existing behavior still makes most sense to me
15:45:21 <noonedeadpunk> do we have anything else to discuss, except having just tons of work tbd and zero time for any of that?
15:45:32 <jrosser> ^ this :(
15:45:44 <jrosser> we need to more the pki stuff forward i guess
15:45:51 <jrosser> and try to deal with the loads of reviews
15:46:14 <jrosser> i've been trying to deal with those sort of one type at a time
15:46:18 <noonedeadpunk> "Without negative feedback" page is just huge
15:46:43 <noonedeadpunk> I was actually thinking, if we should add some rule for reviews, aroud trivial patches
15:47:05 <noonedeadpunk> like - syntax fixes, adding spaces, fixing typos
15:47:19 <noonedeadpunk> and land that kind of stuff with a single vote
15:47:24 <noonedeadpunk> (even on master)
15:47:45 <noonedeadpunk> I'm not sure there're a lot of examples...
15:47:51 <jrosser> perhaps, though you did spot some good things like make in the tox.ini
15:48:29 <noonedeadpunk> I was talking more about things like this: https://review.opendev.org/c/openstack/openstack-ansible/+/960674/5/etc/openstack_deploy/openstack_user_config.yml.aio.j2
15:48:41 <jrosser> yeah
15:48:44 <noonedeadpunk> or that https://review.opendev.org/c/openstack/openstack-ansible/+/960421/6/doc/source/index.rst
15:49:19 <noonedeadpunk> and I think I just went on and merged some path with just a typo in a task name....
15:50:12 <noonedeadpunk> but dunno
15:52:38 <noonedeadpunk> I also am planning to look into merging our contributors documentation into a single place
15:53:15 <noonedeadpunk> and then we are having https://review.opendev.org/c/openstack/openstack-ansible/+/949497 which is in a merge conflict with plenty of things
15:53:29 <noonedeadpunk> So we really need to decide if we wanna do that, or just abandon it
15:55:16 <opendevreview> Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_cinder master: Use loop label for deb822_repository  https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/959030
15:58:20 <noonedeadpunk> #endmeeting