13:00:06 <bbezak> #startmeeting kolla 13:00:06 <opendevmeet> Meeting started Wed Apr 16 13:00:06 2025 UTC and is due to finish in 60 minutes. The chair is bbezak. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:06 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:06 <opendevmeet> The meeting name has been set to 'kolla' 13:00:14 <bbezak> #topic rollcall 13:00:41 <enick_764> o7 13:00:44 <r-krcek> o/ 13:00:49 <amir58118> o/ 13:01:00 <MattCrees[m]> o/ 13:01:06 <frickler> \o 13:01:41 <msalari> \o/ 13:01:54 <bbezak> #topic agenda 13:02:25 <bbezak_2> * Roll-call * Agenda * Announcements * Review action items from the last meeting * CI status * Release tasks * Regular stable releases (first meeting in a month) * Current cycle planning * Additional agenda (from whiteboard) * Open discussion 13:03:03 <bbezak> I'm getting worse and worse and worse at agenda bit :) 13:03:29 <bbezak> #topic CI status 13:03:47 <bbezak> frickler: found a ironic CI 2024.1 bug that we need to fix 13:04:07 <bbezak> I can take a look at that 13:04:33 <MattCrees[m]> Kayobe CI is red on multiple fronts. I'm trying to get fixes to the rocky jobs through, keep needing rechecks to unrelated errors in 2023.2 though https://review.opendev.org/c/openstack/kayobe/+/947127 13:04:47 <MattCrees[m]> Once that's in, I should be able to fix the other issue: rabbitmq for master 13:05:08 <bbezak> in Kayobe that is? 13:05:09 <bbezak> (rabbitmq) 13:05:12 <frickler> 2023.2 will go eol in two weeks, maybe don't put too much effort into that 13:05:13 <MattCrees[m]> Yes 13:05:34 <MattCrees[m]> It needs the queue migrations in upgrade jobs, I have a patch up already too 13:05:40 <bbezak> ok 13:05:47 <bbezak> good 13:05:59 <bbezak> #topic Release tasks 13:06:12 <bbezak> we're finishing rabbitmq 4 stuff 13:06:19 <bbezak> and we still need promv3 13:06:28 <bbezak> which is on my radar 13:06:50 <frickler> what about the container migration stuff? I still didn't get to a proper review for that 13:06:59 <frickler> so docker->podman 13:07:13 <frickler> (not sure if the other direction is needed/supported as well) 13:07:37 <r-krcek> in theory it should work, but I dont think we are testing for this 13:08:03 <frickler> #link https://review.opendev.org/c/openstack/kolla-ansible/+/836941 in particular 13:08:57 <bbezak> I'm wondering if we should flag that as experimental? 13:09:19 <r-krcek> It does take up quite a bit of resources... 13:09:31 <bbezak> commit message says it goes both ways 13:09:33 <bbezak> For example Docker to Podman and vice versa. 13:09:56 <frickler> yes, likely make all of these jobs experimental 13:10:24 <frickler> afaict from a quick look the job only do d->p, too, but I'd think that's fine? 13:10:55 <r-krcek> Yeah, the CI only does d->p as of now 13:11:21 <bbezak> yes. jobs as well. but whole flow in the docs could be marked as such imho. as probably it is not heavily tested (maybe) 13:12:15 <frickler> since the migration is an optional feature, maybe we could do with some less thorough reviews 13:12:36 <frickler> and then have the option for people to do some field testing? ;) 13:12:56 <bbezak> ok. but let's flag that in the docs :) 13:13:03 <frickler> ack 13:13:26 <frickler> I'll try to review this way until next week 13:13:42 <bbezak> makes sense. let's continue in gerrit then. thank you 13:14:48 <bbezak> if somebody feels that some change NEEDS to be merged in Epoxy, let us know. we can flag it as +2 review priority. BTW we've got too much +1 RP :) 13:15:09 <bbezak> I did review some of them and merged some as well 13:15:35 <opendevreview> vahideh proposed openstack/kolla-ansible stable/2024.2: sysctl: Safely handle missing IPv6 when disabled in kernel https://review.opendev.org/c/openstack/kolla-ansible/+/947474 13:15:41 <frickler> RP+1 for me currently means: can wait until 2025.2 13:15:55 <bbezak> yeah 13:16:59 <frickler> oh my, ipv6 disabled again. cant we just agree that such a deployment is broken? :( 13:17:57 <bbezak> :) 13:18:32 <frickler> people can amend for that downstream if they really think they need to, can't they? 13:18:33 <bbezak> somehow this is still a problem in 2025 13:18:55 <frickler> https://bugs.launchpad.net/kolla-ansible/+bug/2071443 was created by will fwiw 13:19:57 <bbezak> yeah, people sometimes doing that because of other bad software 13:20:42 <bbezak> (disabling ipv6) 13:20:59 <bbezak> #topic Additional agenda (from whiteboard) 13:21:04 <bbezak> r-krcek: added 3 changes 13:21:22 <bbezak> #link https://etherpad.opendev.org/p/KollaWhiteBoard#L75 13:23:24 <bbezak> I'd say it will be for 2025.2 13:23:34 <frickler> r-krcek: is anything there important for 2025.1? 13:23:43 <bbezak> I don't have cycles to look into that. unless others feels otherwise 13:24:51 <r-krcek> fickler: 946227 I think would be good to add, but rest are just minor features. Sorry if I am being too impatient. 13:25:53 <opendevreview> vahideh proposed openstack/kolla-ansible stable/2024.2: sysctl: Safely handle missing IPv6 when disabled in kernel https://review.opendev.org/c/openstack/kolla-ansible/+/947475 13:26:47 <bbezak> it's fine r-krcek review velocity is not good, cause of small amount of reviewers ;) 13:26:58 <bbezak> I've flagged that one as RP2 13:27:43 <bbezak> #topic Open discussion 13:28:27 <r-krcek> How can I officially sign up for this one? :) https://etherpad.opendev.org/p/kolla-flamingo-ptg#L157 13:29:40 <bbezak> it is officially done 13:29:41 <msalari> Got it right? So, in the future, does the Kolla Ansible team plan to use Podman instead of Docker (And currently you're adding it as an optional feature)? If that's the case, when will the actual replacement happen? 13:29:44 <bbezak> :) 13:30:10 <frickler> r-krcek: I think this might need a bit more discussion about the actual scope 13:30:40 <bbezak> true 13:30:40 <frickler> msalari: afaict the plan for now is to support both in parallel for the foreseeable future 13:30:40 <r-krcek> I was looking in it and I was thinking the same 13:31:27 <frickler> r-krcek: so maybe wait until we do our after-ptg 2025.2 planning session 13:31:46 <frickler> which we will schedule once 2025.1 is branched and mnasiadka has some time again 13:32:22 <r-krcek> So should I look out in the mailing list for the date/time 13:32:25 <bbezak> that is a good idea. we will grab then ideas from PTG and create plan for 2025.2 13:33:08 <frickler> r-krcek: it should also be announced during one of the next IRC meetings, so watch this space ;) 13:33:13 <bbezak> yes r-krcek 13:33:21 <r-krcek> Thank you :) 13:33:52 <msalari> At our company, we're looking into bringing back the Swift role and supporting it in Kolla Ansible. Now that it's been fully removed, do you think that's feasible? And what would be the first step? 13:34:08 <bbezak> mnasiadka: will be next week I think. so it can also be discussed on next IRC meeting 13:34:39 <frickler> msalari: as mnasiadka answered on the review, we do not consider that feasible 13:35:15 <bbezak> msalari 13:35:24 <bbezak> sorry :). misclick 13:35:30 <bbezak> I don't have context frickler 13:35:46 <frickler> #link https://review.opendev.org/c/openstack/kolla-ansible/+/945791/comments/5c662ef7_d396b2ad 13:36:16 <bbezak> ok 13:37:06 <amir58118> frickler: thanks for reviewing that patch. 13:37:42 <bbezak> I think we're settled on Swift discussion 13:39:04 <msalari> ok 13:40:07 <bbezak> ok. we're in 2025.1 release focus now. we're close, but not there yet 13:40:12 <bbezak> thank you for joining 13:40:14 <bbezak> #endmeeting