15:01:10 <mgoddard> #startmeeting kolla
15:01:11 <openstack> Meeting started Wed Jan 22 15:01:10 2020 UTC and is due to finish in 60 minutes.  The chair is mgoddard. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:12 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:14 <openstack> The meeting name has been set to 'kolla'
15:01:26 <scottsol> o/
15:01:38 <hrw> o\
15:01:42 <mgoddard> mgoddard mnasiadka hrw egonzalez yoctozepto rafaelweingartne cosmicsound
15:01:44 <osmanlicilegi> o/
15:01:48 <yoctozepto> o/
15:03:05 <mgoddard> #topic agenda
15:03:08 <mgoddard> * Roll-call
15:03:10 <mgoddard> * Announcements
15:03:12 <mgoddard> * Review action items from last meeting
15:03:14 <mgoddard> * CI status
15:03:16 <mgoddard> * Ussuri release planning
15:03:18 <mgoddard> * CI scenarios: https://etherpad.openstack.org/p/KollaAnsibleScenarios
15:03:20 <mgoddard> #topic announcements
15:03:34 <mgoddard> I don't have any
15:03:37 <mgoddard> anyone else?
15:04:28 <hrw> Linaro US added
15:04:33 <hrw> will be more arm64 nodes soon
15:04:46 <mgoddard> #topic Review action items from last meeting
15:04:59 <mgoddard> look forward to more arms hrw
15:05:08 <mgoddard> yoctozepto add kolla train deploy guide
15:05:10 <mgoddard> yoctozepto add project-deploy-guide to contributor release guide
15:05:12 <mgoddard> mnasiadka to look at using the master branch of ceph, as used in the ceph-container project, as a way to get ceph libraries for CentOS 8
15:05:14 <mgoddard> yoctozepto to map which services are covered by approved CI scenarios
15:05:18 <mgoddard> yoctozepto claims to have done his
15:05:42 * yoctozepto done my actions
15:05:49 <mgoddard> mnasiadka: did you do yours?
15:06:01 * yoctozepto means yoctozepto did yoctozepto's actions
15:06:02 <mnasiadka> Yeah, it works locally - need to push the change
15:06:21 <mnasiadka> mgoddard: but if it's on top of your changes - then I will most probably re-push yours as well, right?
15:06:35 <cosmicsound> O/
15:06:57 <mgoddard> mnasiadka: not necessarily. Check out my changes, cherry pick yours on top, git review
15:07:15 <mgoddard> or rebase if you prefer
15:07:22 <mnasiadka> mgoddard: right, will do that
15:07:31 <mnasiadka> I have no preference to touch your changes - they look good ;)
15:07:46 <mgoddard> they're great
15:07:54 <mgoddard> #topic CI status
15:08:03 <mnasiadka> mgoddard: yoctozepto was complaining about openstacklight I remember :D
15:08:08 <mgoddard> Are we green?
15:08:23 <mgoddard> if only there were an openstacklight :)
15:08:24 <yoctozepto> openstacklight ftw
15:08:29 <hrw> bifrost got disabled so we should be green
15:08:34 <mnasiadka> we are green-ish - I see more and more occurences of fernet problem on multinode jobs
15:08:43 <yoctozepto> mgoddard, mnasiadka: in my chat client you are green, yeah
15:08:46 <openstackgerrit> LEDUC Florian proposed openstack/kolla-ansible master: Use iSCSI multipath for libvirt  https://review.opendev.org/700910
15:08:48 <mnasiadka> but maybe that's a coincidence :)
15:08:56 <yoctozepto> hrw is red though
15:09:08 <yoctozepto> and I'm blue
15:09:15 <mnasiadka> in mine hrw is green, so it's balanced
15:09:16 <openstackgerrit> Merged openstack/kolla stable/train: Fixes swift-object-expirer for Debian and Ubuntu binary images.  https://review.opendev.org/703492
15:09:18 <openstackgerrit> Merged openstack/kolla stable/stein: Fixes swift-object-expirer for Debian and Ubuntu binary images.  https://review.opendev.org/703493
15:09:19 <mnasiadka> and you're orange yoctozepto
15:09:30 <yoctozepto> mnasiadka: not true, me blue
15:10:04 <hrw> https://i.imgur.com/3eJ9AAr.png shows colours in my client
15:10:06 <yoctozepto> pretty green, bifrost still not buildable, but revert is awaiting in case it is
15:10:09 <mgoddard> thanks everyone for sharing their colours
15:10:25 <mgoddard> I'll add a topic for us to do this each week
15:10:33 <mgoddard> #topic Ussuri release planning
15:10:39 <yoctozepto> mgoddard: I like the idea
15:11:06 <mgoddard> How are we looking for Ussuri?
15:12:00 <yoctozepto> always could be better, we have c8 pending still
15:12:31 <mgoddard> yes
15:12:41 <yoctozepto> mgoddard: https://review.opendev.org/#/c/701010/ <- reminding - if you were so kind as to review post-merge :-)
15:12:42 <patchbot> patch 701010 - kolla-ansible - Fix multiple issues with MariaDB handling (MERGED) - 19 patch sets
15:12:46 <mnasiadka> well, with Ceph working I guess we could start properly testing
15:12:49 <mnasiadka> and phase-out c7 in master
15:12:54 <mgoddard> yoctozepto: will do
15:12:57 <hrw> infra images are issue as it works on local but not CI
15:12:59 <mgoddard> overload
15:13:09 <yoctozepto> overload here too
15:13:38 <osmanlicilegi> https://review.opendev.org/#/c/694799/ is ready for reviews. just waiting the tripleo ci.
15:13:38 <patchbot> patch 694799 - kolla - Upgrade RabbitMQ to 3.8.2 - 24 patch sets
15:13:41 <hrw> I work on non-kolla mostly recently
15:13:49 <mgoddard> Reminder on ussuri schedule
15:14:10 <mgoddard> Kolla feature freeze: Apr 27 - May 01
15:14:11 <mgoddard> Kolla RC1: May 11 - May 15
15:14:13 <mgoddard> Kolla final release: Jun 1 - Jun 5
15:14:24 <osmanlicilegi> https://review.opendev.org/#/c/698694/ will be ready for reviews this weekend
15:14:24 <patchbot> patch 698694 - kolla-ansible - Adds Docker image pruning command - 12 patch sets
15:14:30 <hrw> mgoddard: so 20 March is time to mark 'next release' patches?
15:14:48 <mgoddard> hrw: 20 march?
15:15:00 <hrw> April
15:15:12 <mgoddard> ah, yeah
15:15:26 <mgoddard> 3 months
15:15:33 <yoctozepto> we've got that much time?
15:15:34 <mgoddard> sounds like a lot but will be here soon
15:15:51 <yoctozepto> let's slack off from now on
15:15:57 <yoctozepto> free popcorn for slackers
15:16:29 <mnasiadka> well, a lot of time
15:16:35 <hrw> "a lot"
15:16:42 <mgoddard> also lots still to do :)
15:16:43 <yoctozepto> "of time"
15:16:54 <mnasiadka> but for example merging OVN properly means we need to do C8 first and get back to real ''master'' on centos jobs
15:17:01 <mgoddard> not many features completed so far
15:17:09 <yoctozepto> true
15:17:13 <osmanlicilegi> yes
15:17:36 <mgoddard> I have local backing to push on C8, so hopefully we'll see it move
15:18:03 <mnasiadka> yeah, let's try to focus on merging that part
15:18:11 <mgoddard> just have customer work to do too
15:19:11 <mgoddard> yoctozepto: are you blocking https://review.opendev.org/#/c/699413/ on openstacklight?
15:19:12 <patchbot> patch 699413 - kolla - CentOS 8: base and openstack-base images - 10 patch sets
15:19:56 <yoctozepto> mgoddard: nah, I didn't have time for a legit review
15:20:03 <mgoddard> ok
15:20:50 <yoctozepto> I believe it is fine when you do it and hrw approves but I give only sincere CR+s
15:21:39 <mgoddard> looking at our priorities, the top one with patches to review is internal TLS
15:22:24 <yoctozepto> unfortunately
15:22:40 <yoctozepto> :/
15:22:54 <mgoddard> I think we should also push through ceph-ansible
15:23:02 <mgoddard> as it allows us to remove ceph support
15:23:16 <mgoddard> (for CI this is)
15:23:30 <mgoddard> we don't want to remove ceph too late in the cycle
15:24:18 <mnasiadka> Yeah, I think I addressed both mgoddard's and yoctozepto's comments on this https://review.opendev.org/#/c/685032/ (dependency for ceph-ansible CI) - if that's ok to go, I can focus on doing the same with ceph-ansible change
15:24:18 <patchbot> patch 685032 - kolla-ansible - External Ceph: keys as variables - 34 patch sets
15:24:53 <mgoddard> added to queue
15:25:14 <mgoddard> CI scenarios: https://etherpad.openstack.org/p/KollaAnsibleScenarios
15:25:17 <mgoddard> #topic CI scenarios: https://etherpad.openstack.org/p/KollaAnsibleScenarios
15:25:39 <yoctozepto> oh, so I get to discuss it :-)
15:25:42 <mgoddard> yoctozepto: what's changed?
15:25:56 <yoctozepto> mgoddard: scenario descriptions
15:26:34 <yoctozepto> I did this in parts so narration style differs
15:29:56 <mloza> hi, how can I failover external and internal VIP to another node?
15:30:40 <yoctozepto> mloza: not currently possible (and we are in a meeting so let's discuss it later :-) )
15:33:15 <mgoddard> just adding notes to secondary roles where they are tested
15:33:44 <yoctozepto> mgoddard: yeah, thanks
15:33:46 <yoctozepto> makes sense
15:34:11 <mnasiadka> while we are at CI - I see bifrost - do we plan to support it until the end of the world? Or when (and if) kayobe moves out to ,,bikolla'' then we could deprecate it?
15:34:51 <mgoddard> mnasiadka: we could consider deprecation if/when kayobe stops requiring it
15:34:52 <yoctozepto> bikolla?
15:35:24 <yoctozepto> mgoddard: so you are planning to use k-a to deploy bifrost-like env for deployment?
15:35:36 <mgoddard> yes
15:35:48 <mgoddard> PoC: https://github.com/markgoddard/bikolla
15:35:49 <yoctozepto> mgoddard: I wanted to propose precisely that on Monday
15:35:53 <mnasiadka> mgoddard: just asking, it breaks every now and then :)
15:36:31 <mgoddard> ok, let's make some decisions
15:36:32 <yoctozepto> and no users except for kayobe I believe
15:36:40 <yoctozepto> decisions, decisions
15:36:43 <mgoddard> I propose we graduate zun scenario to one we want to keep
15:36:49 <mgoddard> it covers a few things
15:36:59 <openstackgerrit> Merged openstack/kolla stable/stein: CI: Fix symlinks to failed logs  https://review.opendev.org/703486
15:37:02 <mgoddard> we could throw more in if it fits but it's a good starting point
15:37:08 <yoctozepto> and nfv?
15:37:19 <yoctozepto> we need to extedn the tests by sfc in there
15:37:26 <yoctozepto> some folk was complaining it broke
15:37:28 <mgoddard> nfv already keeping
15:37:40 <yoctozepto> ah, sorry, me blind
15:37:53 <mgoddard> #vote keep zun scenario/
15:38:04 <yoctozepto> #vote +1
15:38:05 <yoctozepto> ?
15:38:14 <yoctozepto> #startvote ?
15:38:15 <openstack> Only the meeting chair may start a vote.
15:38:21 <mgoddard> #startvote keep zun scenario?
15:38:23 <openstack> Begin voting on: keep zun scenario? Valid vote options are Yes, No.
15:38:24 <openstack> Vote using '#vote OPTION'. Only your last vote counts.
15:38:32 <mgoddard> #Vote Yes
15:38:39 <yoctozepto> #vote Yes
15:38:44 <osmanlicilegi> #vote yes
15:38:45 <mnasiadka> #vote Yes
15:38:56 <osmanlicilegi> #vote Yes
15:39:05 <mgoddard> #Vote No
15:39:07 <mgoddard> #Vote Yes
15:39:10 <yoctozepto> proposing to ditch cinder-lvm in favor of Zun
15:39:16 <mgoddard> +1
15:39:30 <mgoddard> #endvote
15:39:31 <openstack> Voted on "keep zun scenario?" Results are
15:39:45 <yoctozepto> Results are
15:39:52 <yoctozepto> good for them then
15:40:28 <mnasiadka> lol
15:40:38 <mnasiadka> I think it's better to vote on etherpad :)
15:41:01 <yoctozepto> let's go with "any objections?"
15:41:57 <mnasiadka> and wait 30 minutes? lol :)
15:42:03 <mgoddard> Added a section on scenarios to drop
15:42:11 <mgoddard> including ceph, ceph-upgrade, cinder-lvm
15:43:25 <mgoddard> proposing to keep mariadb
15:44:13 <hrw> gave us more than enough issues to be kept
15:44:41 <yoctozepto> hrw: mariadb is for testing mariadb actions
15:45:24 <yoctozepto> folks, /me leaving
15:45:28 <yoctozepto> buh-bye
15:45:29 <yoctozepto> have fun
15:45:33 <hrw> ignore me - in two meetings at same time
15:45:43 <mgoddard> k
15:45:56 <mgoddard> keeping mariadb
15:51:35 <mnasiadka> mgoddard: I guess we have understanding around main parts
15:53:59 <mgoddard> proposed an 'alternative compute' scenario: qinling?
15:54:01 <mgoddard> senlin
15:54:03 <mgoddard> masakari
15:54:05 <mgoddard> neutron+ovn
15:56:16 <mgoddard> I think we have lost interest
15:56:23 <mgoddard> anyway, I made some proposals
15:56:49 <mgoddard> #topic Open discussion
15:56:55 <mgoddard> Anyone have anything today?
15:57:29 <osmanlicilegi> mgoddard: please see https://bugs.launchpad.net/kolla-ansible/+bug/1860440
15:57:29 <openstack> Launchpad bug 1860440 in kolla-ansible ussuri "watcher fails because of missing /etc/timezone" [Medium,New] - Assigned to Dincer Celik (osmanlicilegi)
15:58:31 <mgoddard> osmanlicilegi: I'm sure I've seen similar bug reports previously
15:58:44 <hrw> and we had some patches for it
15:58:46 <osmanlicilegi> Its ok to introduce /etc/timezone to master but I'd like to ask for backporting. should I go with only fixing watcher?
15:59:25 <mgoddard> https://bugs.launchpad.net/kolla-ansible/+bug/1821592
15:59:25 <openstack> Launchpad bug 1821592 in kolla-ansible "Kolla Watcher Service - Unable to determine the name of the local timezone" [Undecided,In progress] - Assigned to Marek Grudzinski (ivve)
15:59:48 <mgoddard> Marked yours as a duplicate
16:00:27 <hrw> can we just add it as field in globals.yml and then deploy properly?
16:00:30 <osmanlicilegi> ah, blind me
16:01:05 <mgoddard> hrw: which field?
16:01:22 <mgoddard> default_extra_volumes?
16:01:34 <osmanlicilegi> iirc stein does not support default_extra_volumes
16:01:40 <mgoddard> true
16:01:49 <mgoddard> I think we can make it part of the container config
16:02:24 <mgoddard> I don't mind if it's done for every container or just watcher
16:02:29 <hrw> mgoddard: cloud_timezone
16:02:49 <osmanlicilegi> mgoddard: seems only watcher
16:02:50 <mgoddard> hrw: we just mount the file from the host
16:02:57 <hrw> mgoddard: ok
16:03:57 <mgoddard> Anything else?
16:04:04 <osmanlicilegi> nope
16:04:23 <hrw> nope
16:04:33 * hrw back -> cirros
16:04:36 <mgoddard> thanks all, have a nice day :)
16:04:38 <mgoddard> #endmeeting