| noonedeadpunk | mornings o/ | 08:20 |
|---|---|---|
| jrosser | good morning | 09:19 |
| jrosser | interesting set of TLS failures on https://review.opendev.org/c/openstack/openstack-ansible/+/905621 | 09:24 |
| jrosser | noonedeadpunk: about the lxc config stuff from yesterday | 09:29 |
| jrosser | it is likley complicated like that for vars hierarchy reasons, a bit like we had with migrating haproxy | 09:30 |
| jrosser | when the containers are created you basically have vars available for the container hosts | 09:30 |
| jrosser | but the specific config for a particular type of container may be in group_vars/service | 09:31 |
| jrosser | and i expect that is why there is a seperate round of config done in the service playbooks | 09:31 |
| noonedeadpunk | I don't think that's the case frankly speaking. As when we create container what we do - `name: {{ inventory_hostname }}` and then delegate task to lxc host | 09:49 |
| noonedeadpunk | or well... maybe its exactly as haproxy case :D | 09:50 |
| noonedeadpunk | yeah, TLS failure is very weird... Like /constraints/upper_constraints_cached.txt should be local path? Why it fails SSL verification (or tries to do it at all) | 09:53 |
| jrosser | something starting /constraints is odd too | 09:55 |
| jrosser | https://internal-vip:repo-port/constraints/.... so maybe it doesnt print the scheme and host part in the error | 09:57 |
| noonedeadpunk | `'dest': '/openstack/venvs/utility-28.1.0.dev25/constraints.txt', 'content': '--constraint https://172.29.236.101:8181/constraints/upper_constraints_cached.txt\n'})` so in paste above it looks fine | 10:01 |
| opendevreview | Merged openstack/openstack-ansible-lxc_container_create master: Allow LXC container auto mounts to be customised https://review.opendev.org/c/openstack/openstack-ansible-lxc_container_create/+/893229 | 10:14 |
| opendevreview | Merged openstack/openstack-ansible master: Add support for extra Python packages inside Ansible runtime https://review.opendev.org/c/openstack/openstack-ansible/+/905221 | 10:22 |
| noonedeadpunk | another review needed for Zed bump: https://review.opendev.org/c/openstack/openstack-ansible/+/905344 | 10:26 |
| jrosser | done | 10:26 |
| opendevreview | Merged openstack/openstack-ansible stable/2023.2: Bump SHAs for 2023.2 (Bobcat) https://review.opendev.org/c/openstack/openstack-ansible/+/905584 | 10:34 |
| noonedeadpunk | jrosser: do you think we can make 28.1.0? | 11:12 |
| noonedeadpunk | as releases team would insist on that due to release notes we made | 11:13 |
| noonedeadpunk | or should I try to motivate why not | 11:13 |
| jrosser | noonedeadpunk: well i'm not sure - we have not tried 28.x in our lab yet as we have to finish getting everything on jammy first | 11:16 |
| jrosser | so i don't have any direct feedback from us doing an actual upgrade | 11:16 |
| signed | Hello, I am doing a AiO LXC deployment with Zun added to the base setup. However it seems that the setup-openstack playbook is hung at Execute tempest set, and has been for at least 30min, do I need it to continue or can I safely abort? | 11:17 |
| signed | "Execute tempest tests"* | 11:17 |
| noonedeadpunk | signed: well, you can abort. Most likely that Zun is having issues then if it's hung for too long | 11:24 |
| noonedeadpunk | You will have some leftovers though created by tempest and not cleaned up due to abort | 11:24 |
| signed | I don't really care about leftovers (it's a VM), but I don't know if Zun gonna be operational | 11:27 |
| noonedeadpunk | tempest jsut execute tests. so it's not affecting zun directly (except running tests towards it). So if it's not operation or not has nothing to do with cancelling the job | 11:32 |
| noonedeadpunk | But usually when tempest stuck on test, it means that it can't spawn a resource and just waits for it until hits timeout | 11:32 |
| noonedeadpunk | so potentially zun is broken for $reason | 11:33 |
| noonedeadpunk | btw, signed, do you have nested virtualization enabled for the VM? | 11:33 |
| noonedeadpunk | as that could be one of reasons | 11:34 |
| jrosser | there will be a log from tempest in the utility container which might show something helpful about why it is stuck | 11:42 |
| opendevreview | Merged openstack/openstack-ansible stable/zed: Bump SHAs for Zed https://review.opendev.org/c/openstack/openstack-ansible/+/905344 | 12:51 |
| noonedeadpunk | #startmeeting openstack_ansible_meeting | 15:00 |
| opendevmeet | Meeting started Tue Jan 16 15:00:22 2024 UTC and is due to finish in 60 minutes. The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:00 |
| opendevmeet | The meeting name has been set to 'openstack_ansible_meeting' | 15:00 |
| noonedeadpunk | #topic rollcall | 15:00 |
| NeilHanlon | o/ | 15:00 |
| noonedeadpunk | o/ | 15:00 |
| NeilHanlon | morning | 15:00 |
| NeilHanlon | afternoon, i guess ;) | 15:00 |
| noonedeadpunk | heya :) | 15:00 |
| noonedeadpunk | it's slightly closer to evening hehe | 15:00 |
| noonedeadpunk | 4pm | 15:01 |
| NeilHanlon | ah, yeah | 15:01 |
| NeilHanlon | we've got lightly falling snow outside right now... not looking forward to cleaning it up later | 15:01 |
| noonedeadpunk | when I asked locals when they saw snow last time, they answered that it was in 2016 | 15:02 |
| noonedeadpunk | so I kinda get nostalgic about snow | 15:03 |
| noonedeadpunk | #topic office hours | 15:03 |
| noonedeadpunk | We had relatively calm week I'd say | 15:03 |
| noonedeadpunk | as a result - no new bugs. And this allowed to merge bumps and propose new point releases | 15:04 |
| noonedeadpunk | except 2023.2, which I proposed as bugfix, rest are minor releases | 15:04 |
| noonedeadpunk | Also I see that reviews started for openstack_resources role: https://review.opendev.org/q/topic:%22osa/openstack_resources%22 | 15:06 |
| noonedeadpunk | Also interestingly, update of ansible version results with quite weird/bad failures: https://review.opendev.org/c/openstack/openstack-ansible/+/905619 | 15:07 |
| noonedeadpunk | with "Could not find a module for {{hostvars['aio1_repo_container-b7922f21']['ansible_facts']['pkg_mgr']}}." during python_venv_build | 15:08 |
| noonedeadpunk | and fails here o_O https://opendev.org/openstack/ansible-role-python_venv_build/src/branch/master/tasks/python_venv_wheel_build.yml#L19-L32 | 15:09 |
| noonedeadpunk | have close to no idea why it fails only for lxc | 15:10 |
| noonedeadpunk | Another "interesting" topic changes to dynamic_inventory logic: https://review.opendev.org/c/openstack/openstack-ansible/+/905438/2 | 15:11 |
| noonedeadpunk | basically there're 2 goals: 1 - name LXC containers in a same way as their hostname as today we have mess with underscores/dashes in name | 15:12 |
| NeilHanlon | weird... | 15:13 |
| noonedeadpunk | and second to get rid of ANSIBLE_TRANSFORM_INVALID_GROUP_CHARS: https://opendev.org/openstack/openstack-ansible/src/branch/master/scripts/openstack-ansible.rc#L60 | 15:14 |
| noonedeadpunk | yeah, ansible failure is just crazy.... | 15:14 |
| noonedeadpunk | and what's more interesting, it's only for LXC | 15:14 |
| noonedeadpunk | so potentially somehow related to connection plugin changes... | 15:15 |
| NeilHanlon | yeah, could be... | 15:18 |
| damiandabrowski | hey! sorry, I'm late again... | 15:20 |
| noonedeadpunk | no worries | 15:20 |
| jrosser | o/ hello | 15:25 |
| NeilHanlon | heya damiandabrowski, jrosser | 15:25 |
| noonedeadpunk | but eventually I have nothing to add for today. | 15:31 |
| noonedeadpunk | shortly will try to have a look at pki role regarding vault integration | 15:31 |
| noonedeadpunk | but likely not this week | 15:31 |
| jrosser | nothing much from me, i work a lot on the capi patches at the moment to try to make a CI job | 15:41 |
| noonedeadpunk | ++ | 15:41 |
| noonedeadpunk | yeah, that is really nice thing to have | 15:41 |
| jrosser | need to get something to pass and merge it quick | 15:41 |
| noonedeadpunk | I'm pretty much looking forward to it as well | 15:41 |
| jrosser | as there are a lot of moving parts and we need CI to give some assurances about bumping versions of collections etc | 15:42 |
| noonedeadpunk | Interesting part, if vexxhost collections do have some "fixed" requirements for dependencies | 15:43 |
| noonedeadpunk | Hopefully, they will keep reasonable compatability for these https://github.com/jrosser/ansible-collection-kubernetes/blob/main/galaxy.yml#L11-L13 | 15:45 |
| noonedeadpunk | As recently we had some hassle with bein unable to resolve ansible requirements between osa and another collection | 15:46 |
| jrosser | ah well git blame says i fixed those :) | 15:49 |
| jrosser | becasue of exactly that reason | 15:49 |
| noonedeadpunk | ugh, crap, ok, they did excatly what I was afraid of then | 15:50 |
| jrosser | originally there was specific versions called for | 15:50 |
| * noonedeadpunk no idea why one would do that | 15:50 | |
| noonedeadpunk | or well. | 15:50 |
| jrosser | and i changed it to be >= | 15:50 |
| noonedeadpunk | I know the reason but no idea how this can be used in a real world | 15:50 |
| noonedeadpunk | Like galaxy.yml should be treated as pyproject.toml (setup.txt) | 15:51 |
| noonedeadpunk | where you define minimal requirements, but then you have some requirements.yml with specific ones | 15:51 |
| noonedeadpunk | or dunno... | 15:54 |
| noonedeadpunk | #endmeeting | 16:00 |
| opendevmeet | Meeting ended Tue Jan 16 16:00:54 2024 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 16:00 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2024/openstack_ansible_meeting.2024-01-16-15.00.html | 16:00 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/openstack_ansible_meeting/2024/openstack_ansible_meeting.2024-01-16-15.00.txt | 16:00 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2024/openstack_ansible_meeting.2024-01-16-15.00.log.html | 16:00 |
| NeilHanlon | thanks for running the meeting noonedeadpunk | 16:01 |
| noonedeadpunk | my pleasure :) | 16:01 |
| spatel | I have a question about how to migrate VMs from openstack cloud to another openstack cloud? | 16:51 |
| spatel | what is the best and less painful way to move VMs or vm's volumes | 16:52 |
| noonedeadpunk | spatel: have you checked https://github.com/os-migrate/os-migrate already? | 17:02 |
| spatel | reading | 17:05 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Run ceph_client when cinder uses Ceph https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/905741 | 17:26 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_nova master: Run ceph_client when cinder uses Ceph https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/905741 | 17:29 |
| spatel | noonedeadpunk did you use it before? | 17:45 |
| spatel | looks good | 17:46 |
| admin1 | spatel, if its linux, rsync is the best option forward :) | 17:58 |
| admin1 | it puts the customer in control, no downtimes ( based on workload ) | 17:59 |
| admin1 | else its a snapshot, download image, upload image, create vm from image | 17:59 |
| spatel | rsync ? | 17:59 |
| spatel | between two ceph ? | 18:00 |
| spatel | How to to volume migration same way, snapshot.. move...upload | 18:00 |
| admin1 | what is the end consumer ? what kind of data is it ? | 18:01 |
| admin1 | mostly its linux, volume mounted -- something that can rsync with identical setup | 18:01 |
| admin1 | look it as 1 laptop running linux with a usb disk mounted .. u create another latpop with same linux, mount an empty volume and rsync from 1st -> 2nd the data | 18:02 |
| admin1 | this method puts the end-user in control of the situation, how fast./slow he wants to take, what he wants to backup, when he wants the failover to be done | 18:02 |
| admin1 | vs if you are a provider and you commit saying "i am going to take care of it " , and then you might face issues or delays and the client is also not happy with downtime, the way its moved/migrated | 18:03 |
| admin1 | mine is not a technical answer .. its more an operational appraoch | 18:03 |
| spatel | We don't know what is running inside VM.. end goal is to move entire VM not matter what kind of data running inside. | 18:04 |
| spatel | I will try with taking snapshot of VM and download snapshot and move it to next new cloud | 18:04 |
| noonedeadpunk | spatel: really long time ago | 18:05 |
| admin1 | you can snapshot and export and import that in diff ceph | 18:05 |
| noonedeadpunk | when it wasn't collection yet, jsut set of roles/modules | 18:05 |
| noonedeadpunk | so it's nice to see it's healthy and evolving | 18:05 |
| spatel | This doc is written by Redhat and its redhat tool | 18:22 |
| spatel | without doing experiment its not good to do in production :) | 18:23 |
| spatel | noonedeadpunk very strange.. when I take snapshot its bytes are 0 ? | 18:24 |
| spatel | Did you see this issue? | 18:24 |
| noonedeadpunk | I _think_ it might depend on the backend | 18:29 |
| noonedeadpunk | as for some it counts difference from original volume as size (at least on backend side) | 18:29 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!