| admin1 | \o | 08:37 |
|---|---|---|
| noonedeadpunk | mornings | 08:52 |
| anskiy | noonedeadpunk: hey. I saw your comment, it would take some time to explain :). I'm gonna answer on gerrit. | 08:58 |
| noonedeadpunk | sure | 08:59 |
| noonedeadpunk | probably I shouldn't have reviewed as it was still in WIP | 08:59 |
| anskiy | noonedeadpunk: well, for Zed and onwards it's final actually. I'm just trying to figure out how to fix this in our Y installation :) | 09:19 |
| damiandabrowski | I have an error while deploying rally on ma brand-new jammy AIO(i think it doesn't work since Tuesday) | 10:36 |
| damiandabrowski | https://paste.openstack.org/show/bNNKGHORS1jJuySkvWPq/ | 10:36 |
| damiandabrowski | i know there have been some discussions about broken rally yesterday, but do you have any clue why it works in our gating but doesn't work locally? | 10:36 |
| damiandabrowski | i just created jammy VM, cloned OSA master repo and executed scripts/gate-check-commit.sh | 10:37 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-haproxy_server master: Apply tags to PKI role https://review.opendev.org/c/openstack/openstack-ansible-haproxy_server/+/887263 | 10:48 |
| damiandabrowski | ok, I have some findings | 11:52 |
| damiandabrowski | I have a problem with Tempita package(dependency of sqlalchemy-migrate), full dependency chain is here: | 11:53 |
| damiandabrowski | sqlalchemy-migrate>=0.11.0->oslo.db->rally>=3.1.0->rally-openstack | 11:53 |
| damiandabrowski | When installing packages into venv we use '--pre' parameter to also include pre-release and development versions. | 11:53 |
| damiandabrowski | In this particular case, it attempts to install Tempita 0.5.3dev that fails for some reason. | 11:53 |
| damiandabrowski | Tempita 0.5.2 works fine. | 11:53 |
| damiandabrowski | example how pip behaves with and without '--pre' argument: https://paste.openstack.org/show/b74zoqZ7NTRJ8Jva3vY3/ | 11:53 |
| damiandabrowski | So I wonder: | 11:53 |
| damiandabrowski | 1. Why tempita 0.5.3dev started to fail recently. It's not a new version, it was published in 2013 O.o | 11:54 |
| damiandabrowski | 2. Why do we use '--pre' argument? I tried to install rally without it and everything went fine | 11:54 |
| damiandabrowski | 3. Why it fails on my local AIO but at the same time works fine in gating? | 11:54 |
| damiandabrowski | 2. Probably we do need it for master branch :| | 11:58 |
| noonedeadpunk | I guess you might need pre when you try to install smth by SHA? As it will be detected as .devN version? | 12:10 |
| damiandabrowski | yeah | 12:33 |
| damiandabrowski | any thoughts about 1 and 3? | 12:33 |
| noonedeadpunk | sry no idea or time to dig today :( | 13:47 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible master: Update haproxy healthcheck options https://review.opendev.org/c/openstack/openstack-ansible/+/887285 | 13:58 |
| mgariepy | CIQ update on RHEL decision. https://www.youtube.com/watch?v=MnaFeqyEXzc | 18:00 |
| noonedeadpunk | I wonder who will do this bug testing/fixing for them, given CentOS breaks sooooo much lately | 18:54 |
| mgariepy | they do take question right now ! :) | 18:54 |
| noonedeadpunk | but I guess CIQ are kinda Rocky folks? | 18:56 |
| mgariepy | yes | 18:56 |
| NeilHanlon | well, they employ me, at least noonedeadpunk. but the majority of rocky is not ciq | 22:35 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!