opendevreview | Michal Arbet proposed openstack/kolla-ansible master: Add TLS support for MariaDB Connections https://review.opendev.org/c/openstack/kolla-ansible/+/933924 | 00:37 |
---|---|---|
opendevreview | Michal Arbet proposed openstack/kolla-ansible master: Enhance TLS configuration and certificate handling in Kolla https://review.opendev.org/c/openstack/kolla-ansible/+/937108 | 00:37 |
kevko | shermanm[m]: yep, go ahead ...ask | 00:41 |
opendevreview | Michal Nasiadka proposed openstack/kolla stable/2024.1: CI: Drop upgrade jobs from 2023.2 https://review.opendev.org/c/openstack/kolla/+/949837 | 05:02 |
opendevreview | Michal Nasiadka proposed openstack/kolla-ansible stable/2024.1: CI: Drop upgrade jobs from 2023.2 https://review.opendev.org/c/openstack/kolla-ansible/+/949838 | 05:05 |
opendevreview | Michal Nasiadka proposed openstack/kolla stable/2024.1: Revert "mariadb: pin to 10.11.11" https://review.opendev.org/c/openstack/kolla/+/949839 | 05:07 |
opendevreview | Michal Nasiadka proposed openstack/kolla stable/2024.2: Revert "mariadb: pin to 10.11.11" https://review.opendev.org/c/openstack/kolla/+/949840 | 05:08 |
opendevreview | Michal Nasiadka proposed openstack/kolla master: Revert "mariadb: pin to 10.11.11" https://review.opendev.org/c/openstack/kolla/+/949841 | 05:08 |
opendevreview | Merged openstack/kolla stable/2024.1: CI: Drop upgrade jobs from 2023.2 https://review.opendev.org/c/openstack/kolla/+/949837 | 05:23 |
mnasiadka | frickler: I'm afraid something is wrong with our complicated Zuul config in kolla-ansible - I just removed jobs and it's running not only the tox ones - https://review.opendev.org/949838 | 05:24 |
frickler | mnasiadka: it may be that a change in the project definition triggers everything that is in there | 05:27 |
mnasiadka | well, kolla one doesn't - but yeah, it only references k-a jobs | 05:27 |
mnasiadka | so maybe that's true | 05:28 |
mnasiadka | I'm still not really convinced we need to test hashicorp vault passwords plugin ;-) | 05:28 |
mnasiadka | (and that's why we have vars in separate ,,base job'') | 05:28 |
frickler | hmm, we might be using that plugin soon, so I'm hesitating to say "just drop it" ;) | 05:30 |
mnasiadka | well, we could rework the CI to use OpenBao at least :) | 05:34 |
mnasiadka | Trying to sort of ,,fix'' the RP votes removal ability (when somebody that gave RP is away) - https://review.opendev.org/c/openstack/project-config/+/949842 | 05:35 |
mnasiadka | I'm not sure it's the correct way to do it, but let's see :) | 05:35 |
mnasiadka | So it's official, py3.9 is out through the door, and we still don't have CentOS Stream 10 DIB support, OpenDev nodes, and so on to start adapting CS10/Rocky10 (when it's out) | 05:52 |
mnasiadka | So I think for now we have to live with stable/2025.1 in master images for a while, to be able to test and backport patches from master | 05:55 |
frickler | devstack is trying to run with newer python on cs9, not sure if that would be an option? I don't like having to block progress for debuntu just because rhel is mad | 05:56 |
frickler | mnasiadka: different topic, can someone make a patch to stop these? https://zuul.opendev.org/t/openstack/buildset/dfa6a645702a4f6c882c5c827be47e65 | 06:45 |
mnasiadka | sure, will do | 06:47 |
mnasiadka | frickler: I can try if it's a quick stunt to use newer python - but I'm not going to work many hours on it if it's not :) | 06:48 |
frickler | mnasiadka: but I assume you'll also block simply dropping rocky testing on master, even temporarily? | 06:49 |
frickler | jammy looks to have some mirror issue https://zuul.opendev.org/t/openstack/build/48cc986d7c0b4282bbd4ecff3d42c0e4 | 06:49 |
mnasiadka | Yeah, we could drop rocky/centos testing on master | 06:51 |
mnasiadka | And I want to finally drop publishing to docker hub | 06:56 |
opendevreview | Pierre Riteau proposed openstack/kolla-ansible stable/2025.1: Fix unconditional use of become in post-deploy https://review.opendev.org/c/openstack/kolla-ansible/+/949850 | 07:41 |
opendevreview | Verification of a change to openstack/kolla-ansible stable/2024.1 failed: CI: Drop upgrade jobs from 2023.2 https://review.opendev.org/c/openstack/kolla-ansible/+/949838 | 08:12 |
opendevreview | Will Szumski proposed openstack/kayobe master: CI: Test kayobe_control_host_become https://review.opendev.org/c/openstack/kayobe/+/949519 | 08:43 |
opendevreview | Will Szumski proposed openstack/kayobe master: CI: Improve readabiity of ansible output https://review.opendev.org/c/openstack/kayobe/+/949856 | 08:43 |
opendevreview | Verification of a change to openstack/kolla-ansible stable/2024.1 failed: CI: Drop upgrade jobs from 2023.2 https://review.opendev.org/c/openstack/kolla-ansible/+/949838 | 09:19 |
opendevreview | Maksim Malchuk proposed openstack/kayobe stable/2024.2: Add the ability to override the system's default repositories https://review.opendev.org/c/openstack/kayobe/+/949858 | 09:22 |
opendevreview | Maksim Malchuk proposed openstack/kayobe stable/2024.1: Add the ability to override the system's default repositories https://review.opendev.org/c/openstack/kayobe/+/949859 | 09:24 |
mnasiadka | frickler, kevko : Would like to merge https://review.opendev.org/c/openstack/kolla-ansible/+/940428 (and probably backport it to 2025.1) - to get a more meaningful handler notification message | 09:38 |
frickler | commented | 10:12 |
kevko | mnasiadka: also commented | 11:01 |
PrzemekK | Could You fix octavia.conf [oslo_messaging_rabbit] heartbeat_in_pthread = False Deprecated: Option "heartbeat_in_pthread" from group "oslo_messaging_rabbit" is deprecated for removal (The option is related to Eventlet which will be removed. In addition this has never worked as expected with services using eventlet for core service framework.). https://docs.openstack.org/octavia/2024.2/configuration/configref.html | 11:02 |
PrzemekK | same with WARNING keystonemiddleware.auth_token [-] AuthToken middleware is set with keystone_authtoken.service_token_roles_required set to False. This is backwards compatible but deprecated behaviour. Please set this to True. | 11:04 |
kevko | mnasiadka: I found a very elegant way to handle the optional creation of a remote runtime venv for Kolla if the user chooses it (the idea is that in group_vars, virtualenv and ansible_python_interpreter are set to the path inside the virtualenv).. However, the upgrade job is throwing a wrench into this — it checks out from the previous branch | 11:05 |
kevko | where this obviously doesn’t work, because bootstrap-servers fails. So I rewrote the wrapper for bootstrap_servers so that in this case it always keeps the default auto_silent. That means the bootstrap uses the system Python, creates the venv there, and then subsequent Kolla-Ansible commands just use it. In any case, this is safe... Could we | 11:05 |
kevko | merge it? | 11:05 |
kevko | mnasiadka: https://review.opendev.org/q/topic:%22bootstrap-servers-system-interpreter%22 | 11:06 |
kevko | mnasiadka: second step will be add venv creation into ansible-collection-kolla for stable branches + master | 11:06 |
mnasiadka | kevko: I think we need to make sure in a-c-k that non-venv setups will work as well | 11:07 |
kevko | mnasiadka: and last step ..add interpreter to group_vars/all.yml where it will by auto_silent or venv/bin/python if virtualenv will be defined and length > 0 | 11:07 |
kevko | mnasiadka: yes of course ...it is working for both | 11:07 |
kevko | mnasiadka: with this changes user will define virtualenv: "some_path" ...or don't define ..and it's working as now ... | 11:08 |
kevko | mnasiadka: point is that it's automatic and elegant | 11:08 |
kevko | mnasiadka: It would simply be controlled by the virtualenv variable. | 11:08 |
mnasiadka | nice | 11:09 |
opendevreview | Marek Buch proposed openstack/kolla-ansible master: Add documentation for Kolla Ansible CLI command completion https://review.opendev.org/c/openstack/kolla-ansible/+/948409 | 11:14 |
opendevreview | Verification of a change to openstack/kolla stable/2024.2 failed: Revert "mariadb: pin to 10.11.11" https://review.opendev.org/c/openstack/kolla/+/949840 | 11:16 |
frickler | meh, fluentd mirror issues | 11:19 |
kevko | mnasiadka: point is that now I need to strictly set -e ansible_python_interpreter=auto (or auto_silent) for bootstrapping ...so i can continue ...actually it's default value of ansible -> https://github.com/ansible/ansible/blob/7ac74ab5917de0397a2cd2e10ef37c13fa97d62b/lib/ansible/config/base.yml#L1627-L1629 ...so can we merge :) ? | 11:19 |
sean-k-mooney | hi folks have ye ever seen ModuleNotFoundError: No module named '_dbus_bindings' | 11:27 |
sean-k-mooney | im getting that if i do a pull | 11:27 |
sean-k-mooney | it might be related to the version of python on the destination node or i might need to update the version of dbus pythn but im not sure where the venv would be on the destioantion node assumeing that what kolla is using | 11:29 |
kevko | sean-k-mooney: do you use venv for kolla on target hosts or host python ? | 11:31 |
sean-k-mooney | hum good point i have a venv on the node im running kolla from but its proably using host python on the dest | 11:32 |
sean-k-mooney | so im considering nuking and reisntalling nayway | 11:32 |
sean-k-mooney | the node that is havign isseus is actully installed with debian testign which was proably not a good choice gvien the other node si runing ubuntu | 11:33 |
kevko | sean-k-mooney: ansible -> tar module's python code -> transfer to host -> ansible check which interpreter is -> if not defined it's auto -> means target host python -> means check your python3-dbus | 11:33 |
sean-k-mooney | ack ill try updatign that first and see if thats the issue | 11:34 |
kevko | sean-k-mooney: you can create venv or target hock | 11:34 |
kevko | *on | 11:34 |
kevko | sean-k-mooney: and let kolla-ansible bootstrap venv ... | 11:34 |
kevko | sean-k-mooney: actually this is what i am trying to move on above :) ^^ | 11:34 |
sean-k-mooney | oh the -e ansible_python_interpreter=auto stuff | 11:35 |
kevko | sean-k-mooney: this will cause ansible will use libraries from the system ..if your dbus is too new on testing ... it's causeing your error ..probably | 11:38 |
kevko | sean-k-mooney: but if you create virtualenv in /your/path/for/venv ...and set virtualenv: /your/path/for/venv and ansible_python_interpreter: /your/path/for/venv/bin/python to globals ...or -e FOO=BAR ....you can run bootstrap-servers ...which bootstrap your venv with right versions ... | 11:39 |
kevko | sean-k-mooney: and everything then will be executed against this venv | 11:40 |
kevko | sean-k-mooney: this should work .. | 11:40 |
kevko | sean-k-mooney: Sorry, maybe I'm explaining it poorly - does it make sense? :) | 11:41 |
sean-k-mooney | ok i know what the problem is | 11:41 |
kevko | sean-k-mooney: what ? | 11:41 |
sean-k-mooney | ansibel is trying to use 3.12 but this is debian testing and default python is now 3.13 | 11:42 |
sean-k-mooney | and dbus is aviabale in 3.13 but not 3.12 to import | 11:42 |
opendevreview | Merged openstack/kolla-ansible stable/2024.1: CI: Drop upgrade jobs from 2023.2 https://review.opendev.org/c/openstack/kolla-ansible/+/949838 | 11:43 |
sean-k-mooney | this is how i have defied my hosts. https://paste.opendev.org/show/bss7aDAcItrvWUKFPFWr/ i can likely overried the interpreter for the lake bed node to use 3.13 | 11:44 |
sean-k-mooney | althogu that might require a newer ansibel but ill find out | 11:45 |
kevko | sean-k-mooney: yeah, but you still can create target venv from 3.12 :) | 11:45 |
kevko | sean-k-mooney: as i said | 11:45 |
sean-k-mooney | kevko: none of this is urgent by the way i just wanted to update my home cluster which has been offlien for a while since i replaced my home router | 11:45 |
sean-k-mooney | and yes thats an option too. | 11:46 |
opendevreview | Verification of a change to openstack/kolla master failed: Revert "mariadb: pin to 10.11.11" https://review.opendev.org/c/openstack/kolla/+/949841 | 11:47 |
sean-k-mooney | kevko: ok ya just settign it to 3.13 fixed it | 11:49 |
sean-k-mooney | i guess the default behavior in ansibel is to default to the saem version you are runnign anaible on or somethign like that | 11:49 |
kevko | sean-k-mooney: ansible_python_interpreter is set to "auto" | 11:53 |
sean-k-mooney | sure but i woudl expect auto to be /usr/bin/python3 on the dest host | 11:53 |
kevko | sean-k-mooney: which means ..find in path ...typically returns /usr/bin/python3 | 11:53 |
kevko | sean-k-mooney: and what python3 is on yours ? i would expect 3.13 | 11:54 |
sean-k-mooney | it is 3.13 but it used 3.12 | 11:55 |
sean-k-mooney | i think this is a wered interaction where python3 is differnt on the host im runnign ansible from vs each of the host in the inventory | 11:55 |
kevko | sean-k-mooney: and what is /usr/bin/python ? | 11:55 |
sean-k-mooney | it does not exist | 11:55 |
sean-k-mooney | on the debian host at least | 11:56 |
kevko | ansible your_host -m setup -a 'filter=ansible_python_interpreter' | 11:56 |
kevko | interisting | 11:56 |
kevko | sean-k-mooney: no problem, i know it's not urgent ...but i am working actually on this with Ubuntu 24.04 comming as replacement for 22.04 ... etc etc | 11:57 |
sean-k-mooney | welll im defintly not running a supproted config just so your aware of that | 11:58 |
kevko | sean-k-mooney: got it :) | 11:58 |
opendevreview | Verification of a change to openstack/kolla stable/2024.2 failed: Revert "mariadb: pin to 10.11.11" https://review.opendev.org/c/openstack/kolla/+/949840 | 12:00 |
kevko | sean-k-mooney: Some time ago I asked you about a problem with pre-copy that gets stuck in an infinite loop ...live migration, and after the timeout, the 'pause' action doesn’t happen even though it’s set... Have you happened to look into it since then? | 12:01 |
sean-k-mooney | ok so im actully not mixing operating systems i must have reisllaed the contoler at some poitn. so im usign debian testign on both host but on the contoler i do not have python 3.13 yes only 3.12 and on the comptue i have both 3.12 and 3.13 | 12:01 |
sean-k-mooney | oh no i just looked at the wrong file so in using 24.04 as the contoler and the host where im runing kolla ansible and then debian testing for the compute node | 12:05 |
sean-k-mooney | https://termbin.com/kp4y | 12:06 |
sean-k-mooney | it looks like it does nto auto discover a interperter on the debian host | 12:06 |
sean-k-mooney | oh that when i have hard coded it | 12:08 |
sean-k-mooney | if i filter to just python we can see it discovers 3.12 https://termbin.com/gkf6 | 12:09 |
sean-k-mooney | vs if i hardcode it https://termbin.com/uk4q | 12:09 |
sean-k-mooney | that was to 3.13 exiplcity but we can see the same if i just use py3 https://termbin.com/3dv8 | 12:10 |
sean-k-mooney | my guess is the auto discovery has some bias | 12:11 |
sean-k-mooney | either to usign the same verison on all hosts or to defaultign to the same verison as ansible is runnign under, or the newest python ansibel claims supprot for | 12:11 |
opendevreview | Verification of a change to openstack/kolla master failed: debian: Use Epoxy extrepo https://review.opendev.org/c/openstack/kolla/+/949752 | 12:42 |
opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: Adapt notifying handlers message https://review.opendev.org/c/openstack/kolla-ansible/+/940428 | 12:56 |
opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: Adapt notifying handlers message https://review.opendev.org/c/openstack/kolla-ansible/+/940428 | 12:56 |
opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: Adapt notifying handlers message https://review.opendev.org/c/openstack/kolla-ansible/+/940428 | 12:58 |
opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: Adapt notifying handlers message https://review.opendev.org/c/openstack/kolla-ansible/+/940428 | 12:58 |
mnasiadka | Aaaand 2023.2 is gone | 13:12 |
opendevreview | Verification of a change to openstack/ansible-collection-kolla stable/2024.1 failed: Fix variable lookup for OS-specific dependencies https://review.opendev.org/c/openstack/ansible-collection-kolla/+/949652 | 13:19 |
opendevreview | Verification of a change to openstack/ansible-collection-kolla stable/2024.1 failed: Fix bootstrapping in kolla-ansible https://review.opendev.org/c/openstack/ansible-collection-kolla/+/949648 | 13:19 |
opendevreview | Michal Nasiadka proposed openstack/kolla-ansible master: Adapt notifying handlers message https://review.opendev.org/c/openstack/kolla-ansible/+/940428 | 13:39 |
kevko | +1 | 14:16 |
opendevreview | Marek Buch proposed openstack/kolla-ansible master: Add documentation for Kolla Ansible CLI command completion https://review.opendev.org/c/openstack/kolla-ansible/+/948409 | 14:35 |
opendevreview | Verification of a change to openstack/kolla master failed: debian: Use Epoxy extrepo https://review.opendev.org/c/openstack/kolla/+/949752 | 14:43 |
opendevreview | Merged openstack/kolla-ansible stable/2025.1: Fix unconditional use of become in post-deploy https://review.opendev.org/c/openstack/kolla-ansible/+/949850 | 14:52 |
opendevreview | Merged openstack/kayobe stable/2024.2: Add the ability to override the system's default repositories https://review.opendev.org/c/openstack/kayobe/+/949858 | 14:59 |
opendevreview | Merged openstack/kayobe stable/2024.1: Add the ability to override the system's default repositories https://review.opendev.org/c/openstack/kayobe/+/949859 | 15:12 |
kevko6 | nah, thelounge crashed after 10 months :( | 15:47 |
kevko6 | nah, why i am kevko6 :( | 15:47 |
kevko6 | test | 15:47 |
kevko6 | test | 15:48 |
kevko6 | test | 15:49 |
*** kevko6 is now known as kevko | 15:49 | |
kevko | test | 15:49 |
kevko | cool | 15:49 |
opendevreview | Merged openstack/ansible-collection-kolla stable/2024.1: Fix variable lookup for OS-specific dependencies https://review.opendev.org/c/openstack/ansible-collection-kolla/+/949652 | 16:58 |
opendevreview | Merged openstack/ansible-collection-kolla stable/2024.1: Fix bootstrapping in kolla-ansible https://review.opendev.org/c/openstack/ansible-collection-kolla/+/949648 | 17:17 |
opendevreview | Jake Hutchinson proposed openstack/kayobe master: Fix authentication failures to missing system scope. https://review.opendev.org/c/openstack/kayobe/+/949946 | 20:36 |
opendevreview | Jake Hutchinson proposed openstack/kayobe master: Fix authentication failures due to missing system scope. https://review.opendev.org/c/openstack/kayobe/+/949946 | 20:37 |
opendevreview | Jake Hutchinson proposed openstack/kayobe master: Fix authentication failures due to missing system scope. https://review.opendev.org/c/openstack/kayobe/+/949946 | 20:42 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!