15:00:58 <noonedeadpunk> #startmeeting openstack_ansible_meeting
15:00:58 <opendevmeet> Meeting started Tue Mar 10 15:00:58 2026 UTC and is due to finish in 60 minutes.  The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:58 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:58 <opendevmeet> The meeting name has been set to 'openstack_ansible_meeting'
15:01:05 <noonedeadpunk> #topic rollcall
15:01:08 <noonedeadpunk> o/ hey there
15:01:31 <anfimovir> hello.
15:02:54 <jrosser> o/ hello
15:04:46 <noonedeadpunk> #topic office hours
15:04:59 <noonedeadpunk> we had a really productive week on merging and reviewing things
15:05:06 <noonedeadpunk> and a great progress has been made
15:05:20 <noonedeadpunk> so thanks everyone who took time for reviews
15:05:36 <jrosser> there is a lot of backports being made that i'm not sure are bugfixes vs features
15:06:15 <noonedeadpunk> yes, we should not be really doing backports of features
15:07:41 <jrosser> maybe i should not be +2 on this for example https://review.opendev.org/c/openstack/openstack-ansible-os_keystone/+/978323
15:08:16 <jrosser> but it is a really good reason for having good commit messages to explain why a change is needed, i.e to fix a breakage or just general improvement or feature
15:08:33 <noonedeadpunk> eh, this is my bad...
15:08:57 <noonedeadpunk> anfimovir: do you have any insight if it makes issues in stable branches?
15:09:14 <noonedeadpunk> as I can some vague memories about having some discussion around it
15:09:36 <anfimovir> jrosser, you right, but this change really help with setup SSO in production.
15:09:46 <noonedeadpunk> I _think_ it was bringing some troubles
15:10:18 <jrosser> yes so if it really is a bugfix then the commit message should explain what is being fixed
15:10:26 <noonedeadpunk> but I didn't write good commit msg as I don't have a EL env with SSO
15:10:27 <jrosser> or even better link to a launchpad bug
15:11:06 <jrosser> anyway just one example - there are quite some where I am not totally sure of the intention of the backport
15:11:12 <noonedeadpunk> so was not sure about much details, except this seems to be a reasonable thing to do
15:11:27 <jrosser> so i am hesitating to review when i am not sure
15:12:27 <opendevreview> Merged openstack/openstack-ansible-ops master: Bump prometheus.prometheus to 0.28.0  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/970925
15:12:50 <noonedeadpunk> I will go through backports and review them today
15:12:57 <anfimovir> noonedeadpunk, without this change need manually clear old repo, we not add in task it.
15:13:26 <anfimovir> But, jrosser right, details we not added, sorry((
15:13:43 <noonedeadpunk> I also should have added a release note....
15:13:53 <jrosser> anfimovir: no problem - just makes a little more tricky to review so all details are useful
15:13:56 * noonedeadpunk finally need to get more organized
15:14:36 <jrosser> and when in 2 years we look at the code again git blame can give context on why we did something
15:14:51 <noonedeadpunk> ++
15:15:30 <noonedeadpunk> so, we have couple of big topics hanging around
15:15:33 <noonedeadpunk> #link https://review.opendev.org/q/topic:%22osa/k8s-mcapi%22
15:15:54 <noonedeadpunk> refactoring of magnum cluster api deployment and it's promotion from ops
15:16:25 <noonedeadpunk> after discussion yesterday with andrewbonney, I've started work on reviving the fork and getting versions up to date
15:16:30 <noonedeadpunk> #link https://github.com/adriacloud/ansible-collection-kubernetes/pull/8
15:16:44 <andrewbonney> excellent, thanks
15:17:21 <noonedeadpunk> There's still work to do about ansible 2.19 and potentially pulling some important features
15:17:21 <andrewbonney> I just added a note to the etherpad about the docker-image-py dependency as I think it is needed on the deploy host
15:17:33 <noonedeadpunk> but also wanna do some clean-up of things that are not needed in our context
15:17:48 <jrosser> i think we have a hook now to allow extra packages to be put in the deploy host venv
15:17:52 <jrosser> hook/var
15:17:59 <noonedeadpunk> yep, we do
15:18:15 <noonedeadpunk> I think it's a file in openstack_deploy actually...
15:18:26 <noonedeadpunk> but maybe also an env var to set path to ut
15:18:28 <noonedeadpunk> *it
15:18:46 <andrewbonney> If anyone can find it that'll solve my problem :) I couldn't spot anything myself
15:19:35 <noonedeadpunk> #link https://docs.openstack.org/openstack-ansible/latest/reference/configuration/extending-osa.html#installing-extra-python-packages-inside-ansible-virtualenv
15:20:05 <noonedeadpunk> this?
15:20:56 <noonedeadpunk> other then that, I was planning to take a look into this whole thing, as well as azimuth driver on weekends or smth around this time
15:21:40 <noonedeadpunk> another topic - pki/vault thing
15:21:43 <noonedeadpunk> #link https://review.opendev.org/q/topic:%22osa_hashi_vault%22
15:21:57 <noonedeadpunk> As agreed - os_cinder was reviewed and now in gates
15:21:59 <jrosser> of course docker-image-py is not in openstack requriements
15:22:10 <jrosser> so thats a bit awkward for the requriements check job
15:22:16 <noonedeadpunk> well. IF we *really* need it - we can add it there
15:22:21 <jrosser> (means we can't just have that always in the runtime venv)
15:22:37 <jrosser> right now that check never sees it as it doesnt execute in the ops repo
15:23:13 <andrewbonney> https://github.com/vexxhost/ansible-collection-kubernetes/blob/main/plugins/filter/docker_image.py
15:23:14 <andrewbonney> https://github.com/vexxhost/ansible-collection-kubernetes/blob/cc7554193155f10a89cf66ff12f2cc0c35fa459e/roles/cilium/templates/values.yml.j2#L2
15:23:21 <andrewbonney> These are the sorts of thing it's used for
15:24:37 <noonedeadpunk> eh
15:25:40 <noonedeadpunk> I wonder if there's a way to refactor this easily to remove dependency
15:26:16 <opendevreview> Merged openstack/openstack-ansible-os_nova master: Remove old task - Unset libvirt runtime options (RPM)  https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/977879
15:28:12 <noonedeadpunk> let me check what actually it's doing in more details, and maybe we can drop dependency in the fork
15:29:16 <opendevreview> Merged openstack/openstack-ansible-ops stable/2025.1: Pin setuptools via global-requirement-pins.txt  https://review.opendev.org/c/openstack/openstack-ansible-ops/+/979727
15:29:50 <noonedeadpunk> so, about PKI role
15:30:08 <noonedeadpunk> now it's up to damiandabrowski to update rest of patches to match the os_cinder
15:30:26 <noonedeadpunk> #link https://review.opendev.org/c/openstack/openstack-ansible-os_cinder/+/949427
15:32:07 <noonedeadpunk> also there was a report, that our CI is actually not respecting u-c
15:32:18 <noonedeadpunk> which, I guess  is the reason why this fails: https://review.opendev.org/c/openstack/openstack-ansible/+/977917
15:33:08 <opendevreview> Merged openstack/openstack-ansible-os_designate stable/2025.1: Pin setuptools via global-requirement-pins.txt  https://review.opendev.org/c/openstack/openstack-ansible-os_designate/+/979738
15:36:35 <opendevreview> Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Ensure inventory tests are cleaned up  https://review.opendev.org/c/openstack/openstack-ansible/+/975625
15:36:41 <opendevreview> Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Resolve invalid chars in groups  https://review.opendev.org/c/openstack/openstack-ansible/+/978568
15:38:07 <noonedeadpunk> another huge thing on the plate, is this one
15:38:13 <noonedeadpunk> #link https://review.opendev.org/q/topic:%22osa/invalid_group_chars%22
15:38:24 <noonedeadpunk> which is failing, but more or less minorly
15:38:32 <noonedeadpunk> the biggest question - why ceph job is failing
15:38:40 <noonedeadpunk> and it's failing on fetching keyrings
15:38:44 <noonedeadpunk> which is confusing....
15:39:05 <noonedeadpunk> it's possible that I have missed to update some logic for ceph though
15:39:54 <jrosser> hmm that had complicated delegate_to rules didnt it
15:41:31 <noonedeadpunk> yeah, but we had overrides for group names forever
15:41:45 <noonedeadpunk> so it should be all under our control...
15:42:00 <noonedeadpunk> ah
15:42:10 <noonedeadpunk> I probably forgot to update ceph_client role
15:46:17 <noonedeadpunk> aha
15:46:22 <noonedeadpunk> https://zuul.opendev.org/t/openstack/build/e82c4ed1f04f48f392bee725f20b5ee2/log/job-output.txt#15011
15:47:31 <opendevreview> Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Avoid using invalid group chars in playbooks  https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/978458
15:47:45 <opendevreview> Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Re-enable AIO jobs after group rename  https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/978734
15:48:43 <noonedeadpunk> I think it worked for me as I tested before 2.19....
15:48:52 <noonedeadpunk> ok, cool
15:49:05 <noonedeadpunk> final thing - has anybody played with 26.04 already?:)
15:51:02 <opendevreview> Dmitriy Rabotyagov proposed openstack/ansible-role-zookeeper master: Add molecule testing for the role  https://review.opendev.org/c/openstack/ansible-role-zookeeper/+/979649
15:53:18 <opendevreview> Merged openstack/openstack-ansible-memcached_server stable/2025.1: Pin setuptools via global-requirement-pins.txt  https://review.opendev.org/c/openstack/openstack-ansible-memcached_server/+/979747
15:54:39 <noonedeadpunk> I suppose not
15:54:46 <noonedeadpunk> we have some time for this though
15:55:13 <noonedeadpunk> anything else to discuss?
15:55:37 <anfimovir> I think no, thanks guys.
15:55:51 <opendevreview> Merged openstack/openstack-ansible-os_glance stable/2025.1: Pin setuptools via global-requirement-pins.txt  https://review.opendev.org/c/openstack/openstack-ansible-os_glance/+/979726
15:56:23 <noonedeadpunk> ok, cool, thanks all for joining!
15:56:26 <noonedeadpunk> #endmeeting