| noonedeadpunk | ThiagoCMC: yes, but today we're using ansible-core 2.15 for 2 releases in a row already. which would extremely tighten Ansible version to ubuntu specific, which may not match with what EL is providing | 08:12 |
|---|---|---|
| noonedeadpunk | and I guess that's why we love virtualenvs :D | 08:12 |
| ThiagoCMC | noonedeadpunk, sounds good! :-P | 08:28 |
| ThiagoCMC | Morning! lol | 08:29 |
| noonedeadpunk | o/ | 08:29 |
| jrosser | o/ hello | 08:48 |
| jrosser | noonedeadpunk: i was just looking at what we need to do regarding removing xenial based jobs for when those node types are removed | 09:00 |
| jrosser | and surprisingly, we still have pike/queens branches in openstack-ansible-tests | 09:01 |
| noonedeadpunk | ah, yes, I pinged Elod some time ago about that | 09:02 |
| noonedeadpunk | these are not onl repos | 09:02 |
| noonedeadpunk | there were some missing from releases deliverables | 09:03 |
| jrosser | it might be much simpler to remove those old branches rather than try to clean up the old jobs | 09:21 |
| noonedeadpunk | I actually can't recall how the discussion eneded... But sounded like it should be infra to clean up them manually or smth | 09:26 |
| jrosser | from the ML `We plan to remove Ubuntu Xenial test nodes after CentOS 7 is removed if you want to get a head start on that cleanup too.` | 09:42 |
| jrosser | so i just took a look and i think that the issue is that the easy part is to remove the job definitions | 09:42 |
| jrosser | but then that potentially creates * zuul errors in any places that still reference those jobs | 09:42 |
| jrosser | and those are pretty difficult to find | 09:42 |
| noonedeadpunk | so, I guess we need to come up with etherpad of things that needs to be cleaned-out manually and ask infra-root to drop them | 09:44 |
| noonedeadpunk | in terms of old branches that are not part of releases process and were created manually back in the days as well | 09:44 |
| jrosser | yeah, i feel it would be simpler to clean up ancient branches totally, rather than trying a ton of patches to remove jobs / fix zuul errors | 09:44 |
| noonedeadpunk | sure, no reason for that | 09:44 |
| noonedeadpunk | and they needs to be cleaned up anyway | 09:45 |
| jrosser | ok cool - i'll leave these alone then, i was considering trying to fix up the jobs but it was 8-O amount of mess | 09:45 |
| noonedeadpunk | jrosser: there;s some issues after physnet was merged: https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/912481? | 09:57 |
| noonedeadpunk | yeah, jobs part totally not worth of time | 09:58 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Use ansible_facts['processor_vcpus'] instead of fact variable https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/912481 | 09:59 |
| jrosser | so we have `physical_interface_mappings = physnet1:br-vlan,physnet1:eth12` there | 10:00 |
| noonedeadpunk | yeah, and I guess you can't have same physnet for multiple things? | 10:04 |
| noonedeadpunk | on a flat networking? | 10:04 |
| noonedeadpunk | or well.... | 10:05 |
| noonedeadpunk | I'm confused now I guess | 10:05 |
| jrosser | andrewbonney: do you recall anything about this ^ ? | 10:06 |
| jrosser | ah well it maps a label to a physical interface | 10:06 |
| jrosser | so the same label cannot be used twice | 10:06 |
| jrosser | in the ovn case the ml2 config file is more helpfully structured https://08bbd9f3626d3defcfd4-239a29fa7add7ba7ba8e1040fbce5f75.ssl.cf1.rackcdn.com/912481/2/check/openstack-ansible-deploy-aio_metal-ubuntu-jammy/f6a033e/logs/etc/host/neutron/plugins/ml2/ml2_conf.ini.txt | 10:12 |
| jrosser | but i don't know if this same thing is applicable to linuxbridge as well | 10:12 |
| * jrosser makes AIO | 10:25 | |
| opendevreview | Merged openstack/openstack-ansible master: Remove obsolete nova_force_config_drive variable from docs https://review.opendev.org/c/openstack/openstack-ansible/+/912486 | 10:27 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Restart OVN on certificate changes https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/912762 | 11:54 |
| fungi | jrosser: noonedeadpunk: ideally you'd just amend your acl(s) to grant branch delete permission to some restricted subset of the team and then you'll be able to take care of those | 12:48 |
| fungi | or if they're under release management then the release team generally does it | 12:49 |
| noonedeadpunk | fungi: well, they're under release management except some old branches for some repos | 12:52 |
| noonedeadpunk | which sounds quite random at the moment | 12:52 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible master: Fix physical network mapping for linuxbridge https://review.opendev.org/c/openstack/openstack-ansible/+/912768 | 13:15 |
| mgariepy | humm . https://mail.openvswitch.org/pipermail/ovs-discuss/2024-March/052994.html | 13:16 |
| noonedeadpunk | yeah | 13:23 |
| noonedeadpunk | Neil has sent that right on the release. | 13:23 |
| noonedeadpunk | what a ridiculous thing I've found today... https://paste.openstack.org/show/bwPuruDU6950IG8UbN9t/ | 13:27 |
| mgariepy | fun | 13:28 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-os_neutron master: Use ansible_facts['processor_vcpus'] instead of fact variable https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/912481 | 13:38 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Use ansible_facts['processor_vcpus'] instead of fact variable https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/912481 | 13:41 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron master: Restart OVN on certificate changes https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/912762 | 18:04 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!