| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Use first utility host as default host for openstack_resources https://review.opendev.org/c/openstack/openstack-ansible/+/945085 | 11:51 |
|---|---|---|
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Move left playbooks to collection https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/945089 | 12:18 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Move remaining playbooks to collection https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/945089 | 12:18 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Move remaining playbooks to collection https://review.opendev.org/c/openstack/openstack-ansible/+/945090 | 12:21 |
| harun | hi, are default cluster operations for magnum service such as changing cluster size, updating node group supported in the vexxhost cluster API driver? I tried to replace nodegroup but it didn't work | 12:50 |
| harun | for instance, how can i change min - max node count in the cluster | 12:51 |
| harun | or master count | 12:51 |
| noonedeadpunk | I think that did work for us at lest | 12:53 |
| noonedeadpunk | But it would all depend on the version if the driver used I guess | 12:54 |
| harun | my driver version is 0.23.0 right now | 12:54 |
| noonedeadpunk | the one thing with resizing the cluster, is that we were going below minimum limit - it kind of allowed to do so, and marked all nodes as taint | 12:54 |
| noonedeadpunk | but were not actually deleting them | 12:55 |
| noonedeadpunk | I was testing this with v0.24.0 or smth | 12:55 |
| noonedeadpunk | but there were some breaking changes between these 2 | 12:55 |
| noonedeadpunk | like requirement to set `[nova_client]/api_version: '2.15'` | 12:56 |
| noonedeadpunk | in magnum.conf | 12:56 |
| harun | i checked this but these lines are not in my magnum.conf | 13:00 |
| harun | i need to upgrade my vexxhost driver version to resize or upgrade cluster, right? | 13:02 |
| noonedeadpunk | um, I'm really not sure, though I can say that rtesize did work for one of my environments recently | 13:10 |
| noonedeadpunk | but also - it was even pre 0.23 version - I just double checked | 13:10 |
| noonedeadpunk | it was using a fork with backported bugfixes | 13:11 |
| noonedeadpunk | https://paste.openstack.org/show/bPI5BEleznJlBi6Ru8Yq/ | 13:12 |
| noonedeadpunk | and that was on 2024.1 | 13:12 |
| noonedeadpunk | I did not check recent environments on 2024.2 that thorougly though - ones that are using 0.24.0 | 13:13 |
| noonedeadpunk | but in general these oeprations are expected to work | 13:13 |
| noonedeadpunk | I think I can try to check thoug | 13:14 |
| noonedeadpunk | but also - I never managed to get proxy service wqorking.... | 13:18 |
| harun | thank you for your responses, what magnum version do you suggest for 0.29.0 vexxhost cluster driver? i'll try to upgrade sir | 13:23 |
| noonedeadpunk | I never run that. It's hugely refactored and half-re-written into rust | 13:23 |
| noonedeadpunk | so no idea about this one | 13:24 |
| noonedeadpunk | we've tried to update it here for test: https://review.opendev.org/c/openstack/openstack-ansible-ops/+/944136 | 13:24 |
| noonedeadpunk | but -ENOTIME to check in more details | 13:24 |
| noonedeadpunk | and now I'm also thinking about checking on the capi-helm driver as well and what would it take to add it;'s support as well | 13:25 |
| noonedeadpunk | eh, it seems that cluster creation stuck in this other region on third worker creation or smth.... | 13:33 |
| noonedeadpunk | could be quota thing :D | 13:36 |
| harun | Why was the code refactored, code optimized, execution time sped up, etc.? | 13:37 |
| noonedeadpunk | why do you ask me? | 13:38 |
| harun | I was just wondering :d | 13:38 |
| noonedeadpunk | When I tried to figure out same - it didn't work out well | 13:38 |
| harun | okey sir, thank you, i'll try to figure out | 13:39 |
| noonedeadpunk | some context could be seen here: https://www.linkedin.com/posts/vexxhost_enhancing-kubernetes-clusters-with-intelligent-activity-7307400378252238850-XZy5 | 13:39 |
| noonedeadpunk | but there was actually an answer to the question after all :) | 13:40 |
| harun | i am honestly surprised it was refactored to rust, i hope it is a good decision. | 13:43 |
| noonedeadpunk | ¯\(◉◡◔)/¯ | 13:44 |
| noonedeadpunk | having 3 languages in stack also feels owherwhelming to me, but not for me to judge I guess | 13:45 |
| noonedeadpunk | But kinda why I want to check and investigate https://opendev.org/openstack/magnum-capi-helm as alternative. | 13:46 |
| noonedeadpunk | so with 0.24.2 I was able to resize the cluster | 14:15 |
| noonedeadpunk | on 2024.2 without custom forks | 14:15 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible master: Switch OpenStack-Ansible to track 2025.1 https://review.opendev.org/c/openstack/openstack-ansible/+/945115 | 15:11 |
| noonedeadpunk | I think we need to do couple of very important bugfixes for 2025.1 | 15:12 |
| noonedeadpunk | 1. Fix race condition for OVN with OS upgrades | 15:13 |
| noonedeadpunk | 2. rabbitmq re-boostrap process and extra check for quorum queus | 15:25 |
| noonedeadpunk | and smth else was on my mind, but I completely forgot what it was... | 15:26 |
| noonedeadpunk | would be great to recieve another review on https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944970 to unblock heat/magnum roles | 16:22 |
| noonedeadpunk | thanks! | 17:13 |
| mgariepy | you are welcome. | 17:13 |
| opendevreview | Merged openstack/openstack-ansible-plugins master: Omit passing the project when no project scope is needed https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944970 | 19:51 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!