| janno | We use openstack-ansbile with ceph and two custom pools - one for nvme and one for hdd, called volumes_nvme and volumes_hdd. During upgrades we observed that the auth caps had been removed for those pools. Is this a known issue? Is there any way avoid that? | 09:36 |
|---|---|---|
| janno | We are still on Train and had that issue during upgrades from Stein to Train, within Train versions and during a test upgrade to Ussuri. | 09:36 |
| noonedeadpunk | janno: hey | 11:30 |
| noonedeadpunk | are you using ceph-ansible for managing ceph? | 11:31 |
| noonedeadpunk | also I'd guess that that's kind of "integrated" with osa setup rather then standalone one? | 11:31 |
| noonedeadpunk | Eventually,I've dealt with issue like that no later then these weekends when configuring the new cluster | 11:32 |
| noonedeadpunk | but on Dalmatian :) but I'd assume that the issue is the same | 11:33 |
| noonedeadpunk | so right now the variable is named openstack_keys and I had to overrideit as a whole to define proper caps | 11:34 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible/src/branch/master/inventory/group_vars/all/ceph.yml#L55 | 11:34 |
| noonedeadpunk | But now it's part of the internal playbook as this funcctionality was dropped from ceph-ansible lately | 11:35 |
| noonedeadpunk | so varible naming could be different back then | 11:35 |
| noonedeadpunk | janno: ^ | 11:35 |
| noonedeadpunk | ok,the variable is the same, just defined in a different place: https://github.com/ceph/ceph-ansible/blob/stable-4.0/roles/ceph-defaults/defaults/main.yml#L720-L725 | 11:42 |
| mgariepy | stable4.0 is quite old. | 11:45 |
| mgariepy | stable-8.0 is reef, 9.0 is squid. | 11:45 |
| mgariepy | 4.0 is nautilus | 11:47 |
| janno | noonedeadpunk: Yes, it's osa with ceph-ansible. | 11:53 |
| noonedeadpunk | mgariepy: well, theys said train - 4/0 what we had for it | 11:55 |
| janno | yep :) | 11:55 |
| noonedeadpunk | janno: so I think you need to configure caps there for them to persist ceph-ansible runtime | 11:56 |
| janno | noonedeadpunk: I'll give it a try with overriding openstack_keys | 11:56 |
| janno | noonedeadpunk: Thank you | 11:58 |
| mgariepy | ha ok missed that part of the conversation :) | 11:59 |
| mgariepy | sorry, i'll go back to work ;p haha | 11:59 |
| noonedeadpunk | as eventually any configuration project assumes that all configuration is made by it and tries to reset the state to expected one :) | 11:59 |
| noonedeadpunk | lol, my working week is totally ruined anyway.... | 12:00 |
| * noonedeadpunk just recieved UHK 80 so learning how to type from scratch | 12:00 | |
| noonedeadpunk | cloudnull was promoting UHK quite heavily back in the days | 12:01 |
| * noonedeadpunk already faced a bug and applied experimental firmware lol | 12:02 | |
| mgariepy | 400usd that's a pricy keyboard | 12:02 |
| noonedeadpunk | It's somehow even more expensive in EU despite it's produced in EU... | 12:03 |
| jrosser | keyboard for extreme productivity eh | 12:03 |
| mgariepy | seems to be a great gadget tho. | 12:03 |
| mgariepy | not sure my wife would approuve lol | 12:04 |
| noonedeadpunk | Well... I was whining for couple of years, so mine just gave up | 12:04 |
| janno | wild | 12:04 |
| mgariepy | which keys did you get? | 12:05 |
| noonedeadpunk | red box (linear) | 12:05 |
| mgariepy | i have a mechanical kb with brown keys at work and blue one at home. | 12:05 |
| noonedeadpunk | was really not sure about them or silent ones though | 12:05 |
| noonedeadpunk | how brown are? | 12:05 |
| mgariepy | i do prefer the blue clicky ones from home ;D haha | 12:05 |
| noonedeadpunk | or well, which one you prefer? | 12:06 |
| noonedeadpunk | ph | 12:06 |
| noonedeadpunk | *oh | 12:06 |
| mgariepy | but when i'm in a meeting i really need to mute my mic so they think i'm listening haha | 12:06 |
| noonedeadpunk | I bet if I'd tacke clicky, my wife will throw itr out of the window despite the price | 12:06 |
| mgariepy | haha | 12:07 |
| noonedeadpunk | like day2 | 12:07 |
| noonedeadpunk | as she though my old membrane used to be too loud | 12:07 |
| noonedeadpunk | jrosser: what are we doing with https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/939899? is it time to drop functional tests or still worth fixing them? | 12:57 |
| jrosser | noonedeadpunk: so - the patch which adds the most basic molecule tests to the plugins repo basically just checks that setup: module or something works | 13:08 |
| jrosser | a missing step is to port the actual functional tests that we used to have into molecue | 13:08 |
| jrosser | we could knowingly merge 939899 with the functional tests removed as we can see that they pass now | 13:10 |
| jrosser | but then we have no real coverage and should make priority to re-introduce actual tests in molecue | 13:10 |
| alvinstarr | I tried re-running the playbooks after a change and now they are failing on "systemctl reload-or-restart var-www-repo.mount" in the repo-container. | 13:21 |
| alvinstarr | In a couple of tests I tried it seems like systemctl reload-or-restart will fail on a mounted filesystem | 13:21 |
| jrosser | are you sure they actually fail | 13:21 |
| jrosser | as some of those tasks use an ansible rescue: | 13:21 |
| alvinstarr | running the command manually results in a fail from systemctl | 13:22 |
| jrosser | https://opendev.org/openstack/ansible-role-systemd_mount/src/branch/master/tasks/systemd_mounts.yml#L75-L102 | 13:23 |
| jrosser | ansible is very unhelpful when it reports the status in a block/rescue | 13:24 |
| jrosser | as the first task will fail, be reported as failed, and then the rescue block will execute conditionally on the failure of the preceeding task | 13:24 |
| jrosser | so i guess i'm asking is the playbook actually failing and exiting with a failure | 13:25 |
| jrosser | or is ti that there is a failed task, thats expected as part of a block/rescue | 13:25 |
| noonedeadpunk | alvinstarr: so the question is, if you have any failed hosts in playbook results or all shown as OK | 13:25 |
| alvinstarr | My bad. | 13:26 |
| alvinstarr | I logged the output and searching for FAILED that was the first one that kicked out | 13:26 |
| noonedeadpunk | as that is what determines if it was actually failing or not | 13:26 |
| noonedeadpunk | jrosser: well, I really love to drop test repo this cycle tbh | 13:26 |
| jrosser | absolutely that would be great | 13:26 |
| jrosser | i think we are nearly there | 13:26 |
| jrosser | except i have been just -ETIME^2 | 13:27 |
| noonedeadpunk | btw about heat being broken: https://bugs.launchpad.net/openstack-ansible/+bug/2103512 | 13:27 |
| jrosser | oh system scope vs project scope? | 13:27 |
| jrosser | we just had something like that here with placement | 13:27 |
| alvinstarr | The failed= in the summary also matched the repo-container. | 13:29 |
| alvinstarr | but now it is different. | 13:29 |
| alvinstarr | It makes me crazy when I do the same thing and get different results. | 13:29 |
| noonedeadpunk | I think so, yes | 13:29 |
| jrosser | if you find something thats not idempotent then that really is a bug | 13:29 |
| noonedeadpunk | will try to reproduce it | 13:29 |
| alvinstarr | My next dumb question. | 13:33 |
| alvinstarr | I am installing stable/2023.2. | 13:33 |
| alvinstarr | Can I grab 2024.2 and run an upgrade? | 13:33 |
| noonedeadpunk | nope | 13:42 |
| noonedeadpunk | you can either upgrade between slurp releazses (2023.1->2024.1->2025.1) | 13:43 |
| noonedeadpunk | or if you on non-SLURP then upgrade only to slurp | 13:43 |
| noonedeadpunk | so 2023.2 can go only to 2024.1 | 13:43 |
| alvinstarr | can 2024.1 go to 2024.2? | 13:49 |
| noonedeadpunk | yes | 13:49 |
| noonedeadpunk | slurp can go to next slurp or non-slurp | 13:49 |
| noonedeadpunk | non-slurp can go only toi the next sluyrp | 13:50 |
| noonedeadpunk | https://releases.openstack.org/ for the reference which are slurp ) | 13:50 |
| alvinstarr | Thanks. | 13:52 |
| noonedeadpunk | #startmeeting openstack_ansible_meeting | 15:07 |
| opendevmeet | Meeting started Tue Mar 18 15:07:53 2025 UTC and is due to finish in 60 minutes. The chair is noonedeadpunk. Information about MeetBot at http://wiki.debian.org/MeetBot. | 15:07 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 15:07 |
| opendevmeet | The meeting name has been set to 'openstack_ansible_meeting' | 15:07 |
| noonedeadpunk | #topic rollcall | 15:07 |
| noonedeadpunk | o/ | 15:08 |
| noonedeadpunk | hey, sorry for being late | 15:08 |
| noonedeadpunk | I'm suuuper slow today | 15:08 |
| noonedeadpunk | #topic office hours | 15:10 |
| noonedeadpunk | So with plugins repo being unblocked with molecule, I'd suggest go towards covering the gap for molecule tests and fully dropping old functional tests this cvycle | 15:13 |
| noonedeadpunk | and do not spend time effort on trying to fix them | 15:14 |
| noonedeadpunk | I guess the main challenge there is to use "self" as an actual collection version? | 15:16 |
| noonedeadpunk | So we kind of need to install it from dir I guess | 15:16 |
| noonedeadpunk | I somehow wonder how ssh connection plugin works though... | 15:18 |
| noonedeadpunk | but would really love to work on that this week | 15:19 |
| noonedeadpunk | and I guess then re-adding testing to lxc_cvontainer_create and etc can be done later | 15:20 |
| noonedeadpunk | as plugins so fart is the only one left witrh active functional jobs | 15:20 |
| noonedeadpunk | https://codesearch.openstack.org/?q=openstack-ansible-role-jobs&i=nope&literal=nope&files=&excludeFiles=&repos= | 15:21 |
| noonedeadpunk | another thing which we should take care of - broken heat role | 15:21 |
| noonedeadpunk | #link https://bugs.launchpad.net/openstack-ansible/+bug/2103512 | 15:22 |
| opendevreview | Merged openstack/openstack-ansible-os_nova stable/2024.2: Fix ansible `difference()` filter use https://review.opendev.org/c/openstack/openstack-ansible-os_nova/+/944640 | 15:23 |
| noonedeadpunk | I also have not pushed any cycle-highlights for this release, as seems we are not gonna have any | 15:32 |
| noonedeadpunk | as what we did was mainly stability and technical debt | 15:32 |
| noonedeadpunk | while being quite a progress, means nothing to end users | 15:33 |
| noonedeadpunk | #endmeeting | 15:46 |
| opendevmeet | Meeting ended Tue Mar 18 15:46:17 2025 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 15:46 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-03-18-15.07.html | 15:46 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-03-18-15.07.txt | 15:46 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/openstack_ansible_meeting/2025/openstack_ansible_meeting.2025-03-18-15.07.log.html | 15:46 |
| jrosser | oh doh sorry - missed the meeting | 15:50 |
| noonedeadpunk | no worries :) | 15:53 |
| noonedeadpunk | jrosser: have you ever seen that? https://opendev.org/openstack/openstack-ansible-plugins/src/branch/master/plugins/filter/osa_filters.py#L35-L67 | 16:24 |
| noonedeadpunk | I wonder why we never used that | 16:25 |
| noonedeadpunk | or well, except the fact I see it for the first time in my life | 16:25 |
| jrosser | hmm no i've not seen that before | 16:26 |
| jrosser | we'd have to be careful that the plugins collection was always present | 16:26 |
| jrosser | i.e this is a bit tricky to use in the portable roles | 16:27 |
| noonedeadpunk | good point | 16:27 |
| NeilHanlon | bleh, sorry i missed this. | 16:28 |
| NeilHanlon | was even on the tab then got pulled elsewhere | 16:28 |
| noonedeadpunk | I wonder if we can propose that to community.general... | 16:28 |
| noonedeadpunk | as it looks kind of useful overall | 16:28 |
| noonedeadpunk | but still - would be trycky to use in portable roles for a while | 16:29 |
| noonedeadpunk | I wonder why we keep this? https://opendev.org/openstack/openstack-ansible-plugins/src/branch/master/plugins/strategy/tagfilter.py | 16:33 |
| noonedeadpunk | to define skippable tags through vars? | 16:35 |
| noonedeadpunk | and skip handlers even though tasks are executed? | 16:36 |
| * noonedeadpunk trying to understand if it's smth we want to keep or just drop right away | 16:37 | |
| noonedeadpunk | As ability to define skippable tags as vars sound like being handy.... But I can't come up withy a really good usecase | 16:40 |
| noonedeadpunk | As usually it's somehow vice versa | 16:41 |
| noonedeadpunk | as like add-compute.sh is pretty much set of playbooks to import with proper tags | 16:42 |
| noonedeadpunk | so if we could pass used tags as var - that can be handy | 16:43 |
| noonedeadpunk | but again - you need to do that on a playbook level? not during import_playbook | 16:44 |
| noonedeadpunk | but yeah, you can pass vars for import as well... | 16:44 |
| jrosser | not sure how much custom stuff we should have tbh | 16:45 |
| jrosser | tbh | 16:45 |
| noonedeadpunk | it's not used anywhere today | 16:45 |
| noonedeadpunk | But from time to time I hate the fact of being unable to set tags as var internally | 16:47 |
| noonedeadpunk | as smth I'd love to code in to avoid human error I have eiother to wrap in bash or do rely on ppl following instructions and docs (and they don't) | 16:48 |
| noonedeadpunk | but handlers skip is super weird there | 16:59 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Add openstack_user_config verification playbook as healthcheck https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/938980 | 17:02 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Fix ansible linting for functional tests https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944922 | 18:02 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Execute functional tests through molecule https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944923 | 18:02 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Remove tagfilter strategy from the repo https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944926 | 18:09 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Cleanup unnecessary files/templates from tests https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944927 | 18:09 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Stop running functional tests for the repo https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944929 | 18:11 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Fix ansible linting for functional tests https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944922 | 18:13 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Execute functional tests through molecule https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944923 | 18:13 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Remove tagfilter strategy from the repo https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944926 | 18:13 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Cleanup unnecessary files/templates from tests https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/944927 | 18:13 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Update ssh connection plugin paramters to match upstream https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/939899 | 18:14 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-plugins master: Update ssh connection plugin paramters to match upstream https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/939899 | 18:15 |
| opendevreview | Jonathan Rosser proposed openstack/openstack-ansible-plugins master: Automatically import ssh connection plugin options from the base class https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/939957 | 18:15 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!