13:00:20 #startmeeting kolla 13:00:20 Meeting started Wed Jun 5 13:00:20 2024 UTC and is due to finish in 60 minutes. The chair is bbezak. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:20 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:20 The meeting name has been set to 'kolla' 13:00:24 #topic rollcall 13:00:28 o/ 13:00:37 o/ 13:00:38 o/ 13:00:42 \o 13:00:44 o/ 13:01:12 o/ 13:01:32 0/ 13:02:57 #topic agenda 13:02:57 * Roll-call 13:02:57 * Agenda 13:02:57 * Announcements 13:02:57 * Review action items from the last meeting 13:02:58 * CI status 13:02:58 * Release tasks 13:03:00 * Regular stable releases (first meeting in a month) 13:03:00 * Current cycle planning 13:03:02 * Additional agenda (from whiteboard) 13:03:02 * Open discussion 13:03:47 #topic CI status 13:04:06 periodics jobs are failing 13:04:25 possible fix - https://review.opendev.org/c/openstack/kolla/+/921032. I'm working on bumping community.docker in CI 13:04:49 test change of periodics jobs will follow after the meeting 13:05:27 I think we are all green in kayobe land 13:05:31 today we fixed bobcat and antelope for kolla 13:06:12 great 13:06:34 I think except periodics we're OK 13:06:36 why do the periodic jobs need another fix? I don't understand the comment in your commit message 13:07:08 https://opendev.org/openstack/kolla/src/commit/968711579fbe73316227bd2b1993c5656adad9f4/tests/playbooks/publish.yml#L8 this explictly tells that it's using docker sdk (I guess they mean docker-py?) from kolla venv 13:07:20 publish.yml is the only place where we're using community.docker collection 13:07:32 and apparently it also needs to be bumped 13:07:44 and publish.yml is being used only in periodics 13:08:04 docker_host_info: 13:08:05 for example 13:08:51 okay, unfortunate 13:09:13 maybe creating a dedicated venv for the publish job might be a better solution though? 13:09:39 maybe 13:10:00 as ansible version (and default community.docker) is coming from opendev base jobs 13:10:38 I'll look into it 13:10:51 great 13:11:18 #topic Release tasks 13:11:34 how we're doing on Kayobe release jovial[m] ? 13:12:19 Just trying to merge the last few patches and then I was going to propose a release candidate 13:13:30 ok, let me know if you need to push some changes forward 13:13:37 Any further patches people would like to see in the release? 13:13:48 bbezak, will do - thanks 13:15:11 I think we should just merge what we can and do a release at this point 13:15:32 Right, we are cutting it fine already :) 13:15:40 :) 13:15:51 #topic Regular stable releases (first meeting in a month) 13:16:02 I'd say let's move stable release when we have kayobe as well 13:16:21 as we recently released kolla/k-a 13:17:10 #topic Current cycle planning 13:17:19 Will Szumski proposed openstack/kayobe master: Add proxy support to Vagrantfile https://review.opendev.org/c/openstack/kayobe/+/703666 13:17:59 I think we should add Dalmatian release and move the priorities there 13:18:02 I'll look into that 13:18:13 (on the whiteboard) 13:20:42 #topic Additional agenda (from whiteboard) 13:22:10 speaking of federation bug - I somehow don't have this problem in antelope deployment, however I didn't dig into that why is that 13:22:21 https://etherpad.opendev.org/p/KollaWhiteBoard#L72 13:22:34 I added a comment in the change 13:24:24 okay, thanks, I just wanted to get some reviews, I didn't really reproduce it locally (no federation available locally on my laptop), but it seemed like an important bugfix :) 13:24:55 and as far as I could tell the fixed syntax looks more correct to me, I think I even consulted the haproxy docs for that 13:25:52 yeah it maybe makes sense to update the syntax too 13:26:14 maybe some ansible/jinja shenanigans 13:27:19 there are two comments of r-krcek 13:27:28 https://etherpad.opendev.org/p/KollaWhiteBoard#L74 13:28:22 I guess those didn't merge because backportvoting was not done when workflow was set to +1 13:28:33 I hoped me doing backport +1 would suffice 13:28:42 the first one only needed another W+1 after backport flag was set 13:28:45 maybe toggle that w+1 switch again 13:29:49 that trove thing, does this need a backport? I really like the new switch, honestly. 13:31:38 https://review.opendev.org/c/openstack/kolla-ansible/+/920714 is a big change, might need some more discussion. but I only took a quick look just now 13:32:24 Definitely, I'd be glad for any feedback/discussion .) 13:33:40 maybe I can squeeze that in my ever growing list, I'm on the reviewers list already anyway. 13:33:56 I added some more reviewers 13:34:48 Matúš Jenča proposed openstack/kolla-ansible master: Add certificates for RabbitMQ internode TLS https://review.opendev.org/c/openstack/kolla-ansible/+/921380 13:35:44 the bug description and the commit message don't really seem to well connected and might need an update as well, but I just took a quick look 13:35:58 and yes, I read both bug reports. 13:37:05 my initial reaction is, this should be split into at least two parts: the refactoring (converting scripts into playbooks) and the bugfixing (removing all datadirs) just going by the commit message and the size of the patch 13:37:27 unless there is a very good reason why both must be done at the same time, of course. 13:38:03 Matúš Jenča proposed openstack/kolla-ansible master: Add certificates for RabbitMQ internode TLS https://review.opendev.org/c/openstack/kolla-ansible/+/921380 13:38:09 SvenKieske: Since the list of containers that will be destroyed is taken from the config, the playbooks wont remove any containers that were not deployed directly by kolla-anisble. 13:38:28 mhm, is that a bug, or a feature? ;) 13:38:37 I see, okay, I will split it into multiple. I dont belive there is a direct need to do it at the same time 13:39:12 Huh, did not think of if was a feature :O 13:39:33 I at least know of some deployments who run totally separate playbooks (with distinct inventory files) for kolla and "other stuff". I know that's not good, but that's what people seem to do.. 13:39:57 so there might very well be containers not deployed by k-a that you surely don't want k-a to delete :) 13:40:10 I like the move to use playbooks - finally we will be able to use --tags :) 13:40:14 as long as there is an option to turn that feature/bug off, I guess it's fine 13:40:52 ftr: I also like the general idea and the idea to "own" all containers on kolla deployed hosts, but I guess the history of deployments needs some escape switches :) 13:41:34 we've also got some ihalomi changes on whiteboard 13:41:35 I guess we should write the rest directly into gerrit for preservation of review history :) 13:41:44 https://etherpad.opendev.org/p/KollaWhiteBoard#L78 13:44:25 the most important is to merge the first patchset https://review.opendev.org/c/openstack/ansible-collection-kolla/+/916258 13:44:50 so debian tests will pass on the docker worker refactor patchset 13:46:41 ok I'll try to look into that 13:47:34 and some changes of mhiner as well 13:48:05 yes, both ready for review 13:49:01 ok, let's try to review them in following week 13:49:05 commas one is pretty simple :) 13:50:14 #topic Open discussion 13:51:04 anything to discuss? 13:51:15 next PTG dates announced: 2024-10-21 - -25 https://openinfra.dev/ptg/ 13:51:22 just if ppl like to plan early 13:51:35 thx! 13:51:48 thanks for info 13:52:06 only 4 months to go 13:52:29 Matúš Jenča proposed openstack/kolla-ansible master: Add support for RabbitMQ internode tls https://review.opendev.org/c/openstack/kolla-ansible/+/921381 13:52:31 :) 13:52:52 ok let's finish for today then 13:52:53 #endmeeting