| damiandabrowski | noonedeadpunk: your fix looks good at first glance, but do you know why my patch fixed the issue in 2022 but the problem appeared again now? | 08:53 |
|---|---|---|
| noonedeadpunk[e] | [@damiandabrowski:matrix.org](https://matrix.to/#/@damiandabrowski:matrix.org) have no idea, but your message that glusterd starts before network-online sounds wrong... As how glusterd can technically start without networking at all? | 09:14 |
| noonedeadpunk[e] | Wait part is totally valid though | 09:14 |
| noonedeadpunk[e] | But now looking at unit files, race conditions kinda make sense... | 09:15 |
| noonedeadpunk[e] | *looking at unit files now | 09:15 |
| noonedeadpunk[e] | I mean, it could be correct at a time, but I meant that unit files configs were wrong back then | 09:20 |
| noonedeadpunk[e] | Crap, I need to formulate my thoughts better.... | 09:23 |
| noonedeadpunk[e] | So what I meant, is that 2y ago there could be a packaging issue which defined systemd unit files in a weird way, trying to start glusterd before network is up, which could be a different race condition issue | 09:24 |
| noonedeadpunk[e] | But it made mount working | 09:24 |
| damiandabrowski | ahhh okok | 10:05 |
| opendevreview | Marcus Klein proposed openstack/openstack-ansible-ops master: fix Grafana database configuration and deb822 repositories https://review.opendev.org/c/openstack/openstack-ansible-ops/+/964766 | 11:10 |
| kleini | Where can I find openstack.osa.tools.set_haproxy_backends_state? | 11:36 |
| noonedeadpunk | kleini: this is smth we moved just weeks ago, so it's gonna be for Flamingo | 11:37 |
| noonedeadpunk | but it;s here: https://opendev.org/openstack/openstack-ansible-plugins/src/branch/master/playbooks/tools/set_haproxy_backends_state.yml | 11:37 |
| noonedeadpunk | for ops on 2025.1 and earlier, it's here: https://opendev.org/openstack/openstack-ansible-ops/src/branch/stable/2025.1/ansible_tools/playbooks/set-haproxy-backends-state.yml | 11:38 |
| kleini | Ah, need to use stable/2025.1 of openstack-ansible-ops. | 11:39 |
| noonedeadpunk | I tried my best to make it less confusing... | 11:42 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Upgrade setuptools to version 75.8.2 https://review.opendev.org/c/openstack/openstack-ansible/+/964773 | 12:27 |
| noonedeadpunk | jrosser: do you recall how we are configuring PIP these days? | 12:49 |
| noonedeadpunk | I'm just looking on ways to define https://pip.pypa.io/en/stable/cli/pip/#cmdoption-trusted-host and really can't recall this part | 12:49 |
| noonedeadpunk | or we just don't? | 12:50 |
| jrosser | this? https://opendev.org/openstack/ansible-role-python_venv_build/src/branch/master/defaults/main.yml#L86 | 12:51 |
| jrosser | or more generally? | 12:51 |
| noonedeadpunk | oh, huh | 12:51 |
| noonedeadpunk | so why this is not applied: https://zuul.opendev.org/t/openstack/build/bc361b2188874a8d9f6f9b337a98fbff/log/job-output.txt#9089-9098 | 12:53 |
| noonedeadpunk | ok, thanks | 12:53 |
| noonedeadpunk | I think this is whatr I was looking for | 12:53 |
| noonedeadpunk | ah, because on metal we're not building wheels? | 12:55 |
| noonedeadpunk | but I think we still cache u-c | 12:57 |
| noonedeadpunk | ok, thanks a ton for pointing to that. I was looking in a very random places before | 12:59 |
| jrosser | codesearch++ | 13:00 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/ansible-role-python_venv_build master: Define trusted-host even when wheels are not built https://review.opendev.org/c/openstack/ansible-role-python_venv_build/+/964786 | 14:07 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Bump global requirements https://review.opendev.org/c/openstack/openstack-ansible/+/951288 | 14:07 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!