14:03:38 <mnasiadka> #startmeeting kolla
14:03:38 <opendevmeet> Meeting started Wed Sep  7 14:03:38 2022 UTC and is due to finish in 60 minutes.  The chair is mnasiadka. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:03:38 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:03:38 <opendevmeet> The meeting name has been set to 'kolla'
14:03:43 <mnasiadka> #topic rollcall
14:03:44 <mnasiadka> o/
14:03:51 <mnasiadka> tibeer[m]: RamonaBeermann[m] mgoddard now ;)
14:03:59 <mnasiadka> Sorry for being late
14:04:06 <frickler> \o
14:04:32 <tibeer[m]> what do you mean with "now"?
14:04:54 <mgoddard> \o
14:05:08 <mnasiadka> tibeer[m]: the meeting was not started before, so your attendance would not count ;)
14:05:18 <RamonaBeermann[m]> ok o/
14:05:27 <tibeer[m]> ah, so doing "o/" triggers a counter or something?
14:05:52 <mnasiadka> well, writing anything in rollcall
14:06:15 <tibeer[m]> got it. thanks!
14:06:24 <mnasiadka> #topic agenda
14:06:24 <mnasiadka> * CI status
14:06:24 <mnasiadka> * Release tasks
14:06:24 <mnasiadka> * Regular stable releases (first meeting in a month)
14:06:24 <mnasiadka> * Current cycle planning
14:06:25 <mnasiadka> * Additional agenda (from whiteboard)
14:06:25 <mnasiadka> * Open discussion
14:06:27 <mnasiadka> #topic CI Status
14:06:46 <mnasiadka> I didn't see any problematic CI failures, so I assume it's green
14:07:10 <mnasiadka> although whiteboard says it's red
14:07:36 <mnasiadka> but zuul says mainly green
14:07:39 <mnasiadka> I'll update whiteboard to green
14:08:04 <mnasiadka> #topic Release tasks
14:08:53 <mnasiadka> It's R-4, two weeks to feature freeze
14:09:26 <mnasiadka> On 1st Sep we did new point releases, after backporting Docker-py fixes, good
14:09:36 <mnasiadka> #topic Current cycle planning
14:10:07 <frickler> we need another set of stable releases for kolla now with the CVE fix I think
14:10:18 <mnasiadka> Right
14:10:25 <mnasiadka> frickler: will you take care of it?
14:10:35 <frickler> yoctozepto said he will do, will coordinate
14:10:57 <mnasiadka> ah ok
14:11:04 <mnasiadka> good that it's covered
14:11:10 <mnasiadka> So, current cycle
14:11:11 <yoctozepto> o/
14:11:17 <yoctozepto> I confirm
14:11:36 <mnasiadka> https://review.opendev.org/c/openstack/kolla-ansible/+/839715 - this is hopefully ready after latest tenks followup
14:11:58 <yoctozepto> ok, will check
14:11:59 <mnasiadka> and then we could merge cs9 support in Kolla (given it's ready)
14:12:09 <yoctozepto> to check as well
14:12:11 <mnasiadka> and then I'll push RL9 support based on CS9 support
14:12:21 <mnasiadka> And we might have something working finally
14:13:16 <mnasiadka> Elasticsearch is problematic from what I understand, they realised they need to ramp up their repo a bit to support EL9
14:13:25 <mnasiadka> But we should be moving to OpenSearch anyway
14:13:57 <mnasiadka> parallax is working on this, I'll try to help him - but I doubt we'll make the feature freeze date
14:14:26 <mnasiadka> How are the other patches? Has anybody looked at Let's Encrypt or systemd/podman?
14:14:55 <frickler> I looked at systemd today, just some minor comments
14:16:25 <yoctozepto> we can do ffe for opensearch if it is possible at all
14:16:38 <yoctozepto> or else let's backport support
14:16:40 <mnasiadka> yoctozepto: yeah, probably it will end up like that
14:17:15 <mnasiadka> so, the systemd patch - I know it's not ideal timeframe to merge it, but... do we want to wait another cycle?
14:18:07 <frickler> I'd be ok with merging it, better some progress than none
14:18:21 <yoctozepto> I agree
14:18:29 <frickler> and it is some kind of feature on its own
14:18:57 <yoctozepto> otherwise it is not going to happen
14:19:19 <mnasiadka> yeah, I'm sure we're going to revisit that and fix it on the way, but we need to start somewhere
14:19:34 <mnasiadka> has anybody tested that on a local environment?
14:19:53 <frickler> not yet, but I plan to do that
14:20:19 <mnasiadka> ok, so let's try to test that locally, and then merge if no objections
14:21:13 <mnasiadka> yoctozepto: do we have any idea how to make test-core-openstack script better than just barfing with NoneType here and there?
14:21:16 <mgoddard> I would prefer not to merge it if we can't be confident in it
14:22:25 <mnasiadka> Well, that's why I would feel safer if somebody can test that locally
14:22:46 <mgoddard> it needs a thorough test & review
14:23:16 <mnasiadka> second thing is to get it to be run on other jobs than the basic ones
14:23:41 <mnasiadka> and on some EL clone, currently it's only Debian and Ubuntu
14:24:30 <mnasiadka> ok, we all have our reservations - let's see how it evolves
14:26:15 <mnasiadka> mgoddard: on the priorities for this cycle is splitting out baremetal role - do we have any additional changes to review?
14:27:27 <mgoddard> there is one in kayobe, https://review.opendev.org/c/openstack/kayobe/+/824998
14:27:39 <mgoddard> looks like I need to resolve a merge conflict, but it passed last time
14:28:23 <mgoddard> just updating the list of patches in the whiteboard for the collection
14:28:27 <yoctozepto> mnasiadka: barfing? where?
14:29:05 <mnasiadka> yoctozepto: like here https://zuul.opendev.org/t/openstack/build/0db989bb04f94519a9d3e22a3fd23672
14:29:12 <mnasiadka> NoneType: None
14:29:15 <mnasiadka> we could be a bit better :)
14:29:36 <mnasiadka> (of course you can find the cause in the script output, but well)
14:29:57 <mnasiadka> And the ubuntu-ovn-upgrade job probably needs a bigger timeout in waiting for VM to be active, because it often fails on this
14:30:40 <mnasiadka> mgoddard: ok, then I probably should look at Kayobe reviews, I haven't done them for some time
14:33:36 <mnasiadka> ok then
14:33:44 <mnasiadka> #topic Open discussion
14:33:51 <mnasiadka> Anybody anything?
14:35:10 <opendevreview> Mark Goddard proposed openstack/kayobe master: Stop using kolla-ansible bootstrap-servers  https://review.opendev.org/c/openstack/kayobe/+/824998
14:35:25 <frickler> did anyone try to upgrade ironic from xena to yoga yet?
14:35:44 <frickler> seems the old -pxe container needs to be stopped/removed manually
14:35:52 <yoctozepto> mnasiadka: it is not the script that barfs, it's zuul's ansible
14:36:40 <yoctozepto> frickler: I remember adding code handling this for upgrades
14:36:51 <yoctozepto> have you used the upgrade command?
14:37:19 <frickler> well no, I use osism, maybe that's the issue then
14:37:55 <yoctozepto> could be
14:38:11 <frickler> I'll do some digging, thx for the pointer
14:38:12 <mnasiadka> Does anybody have time to review https://review.opendev.org/c/openstack/kolla-ansible/+/847769? It would help with the order for OVN upgrades (and deploys)
14:38:56 <yoctozepto> I have a few patches pending to merge, reviews welcome
14:40:14 <mnasiadka> hmm, did we forget about cinder active-active for rbd? :)
14:40:40 <yoctozepto> may be
14:41:10 <mnasiadka> seems it's not that critical after all
14:41:36 <yoctozepto> seems so
14:42:17 <mnasiadka> yoctozepto: so the barfing is the effect of newer Ansible in Zuul?
14:42:51 <yoctozepto> mnasiadka: to my knowledge - yes
14:43:15 <mnasiadka> anyway, I'll extend the timeout if I'll see ovn upgrade job failing once again
14:43:43 <tibeer[m]> Is this related? https://review.opendev.org/c/openstack/kolla-ansible/+/856239
14:43:57 <tibeer[m]> If so, this is quite important for us.
14:44:26 <mnasiadka> I think yoctozepto's comment is clear, we need to understand the use case.
14:45:14 <mnasiadka> describe it in the commit message, so we understand it
14:45:30 <tibeer[m]> Thanks
14:45:55 <mnasiadka> ok, I guess that's it for today
14:45:58 <mnasiadka> Thanks for joining!
14:46:01 <mnasiadka> #endmeeting