14:00:00 <mnasiadka> #startmeeting kolla
14:00:00 <opendevmeet> Meeting started Wed Oct 26 14:00:00 2022 UTC and is due to finish in 60 minutes.  The chair is mnasiadka. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:00 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:00 <opendevmeet> The meeting name has been set to 'kolla'
14:00:04 <mnasiadka> #topic rollcall
14:00:05 <mnasiadka> o/
14:00:08 <ebbex> o/
14:00:21 <jhorstmann> o/
14:00:43 <ramona-beermann[m]> o/
14:01:11 <frickler> \o
14:01:44 <bbezak> \o
14:02:03 <TimBeermann[m]> o/
14:02:14 <ravlew> o7
14:02:34 <mnasiadka> #topic agenda
14:02:34 <mnasiadka> * CI status
14:02:34 <mnasiadka> * Release tasks
14:02:34 <mnasiadka> * Regular stable releases (first meeting in a month)
14:02:34 <mnasiadka> * Current cycle planning
14:02:35 <mnasiadka> * Additional agenda (from whiteboard)
14:02:35 <mnasiadka> * Open discussion
14:02:38 <mnasiadka> #topic CI status
14:02:53 <mnasiadka> I think we have some gate breakage due to EPEL 9
14:03:18 <frickler> yes, do we want to wait or make stuff non-voting?
14:04:23 <mnasiadka> well, probably most of the patches would be happier if we would really build and publish - and most of them are rocky9 related
14:04:45 <mnasiadka> is it going to fix itself, or is someone proposing to change the mirror in the config?
14:05:44 <frickler> not sure yet, this has happened before, but I'm not involved in any redhatty things
14:06:20 <frickler> finding a stable, reliable mirror might be helpful
14:06:29 <mnasiadka> ok, I'll check on #opendev after the meeting
14:06:37 <mnasiadka> #topic Release tasks
14:07:42 <mnasiadka> So, basically before we branch and release an rc1 - for kolla/kolla-ansible we need OpenSearch merged (still in progress, I'll add some CI to it), for kayobe we need Rocky Linux 9 patch - and probably OpenSearch followup - in progress, bbezak working on it
14:07:47 <mnasiadka> bbezak: any progress?
14:08:33 <bbezak> been caught up in customer work this week, will progress later this week. Got some input from ebbex
14:08:45 <mnasiadka> great
14:09:10 <mnasiadka> do we have any other patches that need to be merged in Zed?
14:09:43 <mnasiadka> I know we assumed systemd, but now that we have review velocity problems - it might be complicated
14:10:43 <mnasiadka> ok, seems none
14:10:52 <mnasiadka> #topic Current cycle planning
14:10:59 <mnasiadka> Current probably means Antelope
14:11:13 <mnasiadka> I'll tidy up PTG notes this week and send out a summary on the ML
14:11:23 <mnasiadka> I'll also populate whiteboard with priorities for this cycle
14:11:23 <opendevreview> Ivan Halomi proposed openstack/ansible-collection-kolla master: Adding podman support  https://review.opendev.org/c/openstack/ansible-collection-kolla/+/852240
14:12:05 <mnasiadka> #topic Additional agenda (from whiteboard)
14:12:19 <mnasiadka> (jhorstmann) adding grafana loki (https://grafana.com/oss/loki/) as an additional log store to kolla/kolla-ansible
14:12:36 <mnasiadka> additional details on https://etherpad.opendev.org/p/KollaWhiteBoard#L53
14:13:13 <jhorstmann> yes, we have been looking at grafana loki for our log management and would like to add it to our stack, which we deploy with kolla
14:13:53 <jhorstmann> so, the question is wether kolla would be interested in having support for loki upstream?
14:14:23 <mnasiadka> I think yes, given we find reviewers to review and guide you towards merging this
14:14:32 <mnasiadka> What is the storage backend Loki would need?
14:14:36 <frickler> I heard about loki some time ago and it sounds interesting to me. I didn't look closer yet, but I sure would be interested in looking at it in more detail
14:15:29 <mnasiadka> Looking at https://grafana.com/docs/loki/latest/operations/storage/ - doesn't ring a bell which one would be easy to add to k/k-a
14:15:42 <jhorstmann> there are a lot of different backends (https://grafana.com/docs/loki/latest/storage/), but we thought about starting just with local storage
14:16:11 <jhorstmann> overriding the config to use a different backend should be straight forward. e.g. an existing object store
14:16:15 <mnasiadka> I think we need to make sure local storage would work for HA
14:16:30 <jhorstmann> that might be a problem
14:17:03 <jhorstmann> loki does not have a builtin replication mechanism and relies on a distributed storage backend for HA
14:17:15 <mnasiadka> That means we're left with either Object Storage or adding Apache Cassandra
14:17:59 <frickler> would be interesting to know whether "S3" also means it works on RGW
14:18:31 <jhorstmann> frickler: I have not tested that, but I think it should
14:18:39 <frickler> but for starting, local storage would be fine IMO
14:19:06 <mnasiadka> on a single node - well for starting probably yes, and then we can sort out the implementation issues
14:19:36 <mnasiadka> anyway, jhorstmann can you add something about it on https://etherpad.opendev.org/p/kolla-antelope-ptg#L215 - I'll add it to cycle priorities with your name on it - and we'll see how it goes
14:20:19 <jhorstmann> sure, will do
14:20:22 <mnasiadka> thanks
14:20:43 <mnasiadka> Another topic
14:20:44 <mnasiadka> (mnasiadka) Review velocity
14:21:12 <mnasiadka> I think with Radek away (or spending minimum amount of time on reviews) - we're going to get soon in a spot that we would not like to be
14:21:58 <mnasiadka> currently me and hrw are raising k/k-a patches - that need a third person to merge them - especially those patches that are needed for Zed release
14:23:16 <frickler> I can do some, but I'm really not comfortable approving rocky patches
14:23:48 <frickler> I can ask cberendt whether he can spare some time, but I expect him to be similarly reluctant on that topic
14:23:59 <mnasiadka> I understand, and we're not going to be able to reach the same velocity as before soon
14:24:22 <frickler> so the question then might be whether we should agree to single-approve certain things
14:24:49 <mnasiadka> Short term probably we should - for Rocky/CentOS based things
14:25:36 <mnasiadka> in the meantime I'll try to bring more SHPC colleagues to start doing reviews often - so in a couple of months time we could get more core reviewers - but unfortunately from the same company as me
14:25:41 <frickler> and longer term if no cores interested in rocky turn up, that might be a reason to consider deprecating it again
14:25:54 <frickler> we cannot maintain things without reviewers
14:27:05 <mnasiadka> Well, StackHPC promised to do CentOS/Rocky in Kolla/Kolla-Ansible - so I think we need to step up and have more core reviewers in a longer term
14:27:41 <mnasiadka> But with that direction, we'll need to raise exception on the single-company-commit rule for CentOS/Rocky
14:27:46 <mnasiadka> is that fine?
14:27:59 <frickler> fine for me
14:28:32 <mnasiadka> ok then, we have some path forward
14:28:43 <mnasiadka> thanks
14:28:48 <mnasiadka> #topic Open discussion
14:28:53 <mnasiadka> Do we have any open points for today?
14:29:21 <frickler> just mentioning that the discussion about the next ptg has started
14:29:33 <frickler> #link https://openinfrafoundation.formstack.com/forms/openinfra_ptg_2023
14:29:46 <mnasiadka> Ah right, I saw the discussion around having PTG colocated with Summit
14:30:05 <mnasiadka> I think still for Kolla projects - we would prefer to have that remotely in the usual timeframe
14:30:16 * frickler won't travel anywhere, so it is virtual or nothing for me
14:30:57 <frickler> I also added a comment about extending the timeframe to reduce overlap. if the TC can do, why should other projects not follow their lead?
14:31:56 <frickler> but maybe I'm also just a weird unicorn with all my involvement
14:32:37 <mnasiadka> Might be, I don't have an opinion - but having a PTG in June just before the vacation season seems like loosing some time
14:32:49 <ramona-beermann[m]> then we are two unicorns :)
14:33:37 <mnasiadka> Anyway, let's see how the discussion evolves - and fill out the form for those interested
14:34:34 <mnasiadka> I don't see any other topics being raised - so let's finish for today
14:34:37 <mnasiadka> Thanks for coming!
14:34:39 <mnasiadka> #endmeeting