16:01:03 <evrardjp> #startmeeting openstack_ansible_meeting
16:01:04 <openstack> Meeting started Tue Apr 10 16:01:03 2018 UTC and is due to finish in 60 minutes.  The chair is evrardjp. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:05 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:07 <openstack> The meeting name has been set to 'openstack_ansible_meeting'
16:01:28 <hwoarang> o/
16:01:29 <evrardjp> #topic "What happened since last meeting"
16:01:35 <evrardjp> evrardjp reminds we have specs that need some eyes.
16:01:40 <evrardjp> evrardjp says the newton EOL is in process, and only the integrated/tests repo are to EOL now.
16:01:42 <spotz> o/
16:01:55 <odyssey4me> o/
16:01:56 <evrardjp> evrardjp reminded the etherpad for vancouver forum: https://etherpad.openstack.org/p/YVR-openstack-ansible-brainstorming
16:01:56 <d34dh0r53> o/
16:02:00 <evrardjp> evrardjp said the next release will be queried according to plans, the 13 April.
16:02:03 <jmccrory> o/
16:02:04 <evrardjp> odyssey4me suggested to use novnc instead of spice by default, or to have more than one method at the same time.
16:02:11 <evrardjp> odyssey4me suggested changes to the tempest role: http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/latest.log.html#t2018-04-10T12:19:33 (see also the conversation above this link for more context)
16:02:15 <prometheanfire> o/
16:02:16 <evrardjp> odyssey4me and hwoarang discussed the distribution package vs source install (http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/latest.log.html#t2018-04-10T09:35:54 )
16:02:21 <evrardjp> Two people asked for multi-region OSA deploys this week, and evrardjp proposed to add some user stories about it (http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/latest.log.html#t2018-04-10T09:13:45) using logan- 's example (http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/latest.log.html#t2018-04-10T11:44:11)
16:02:44 <evrardjp> I let you a few minutes to process this, then let's move to bug triage
16:02:58 <evrardjp> it also allows ppl to join :)
16:03:13 <evrardjp> If you think I forgot to highlight something, please put it on the wiki
16:04:39 <evrardjp> ok
16:04:49 <evrardjp> I think it's time to continue
16:04:57 <evrardjp> #topic Bug triage
16:04:59 <openstackgerrit> Merged openstack/openstack-ansible stable/queens: repo-use.yml: Scope host groups properly  https://review.openstack.org/559726
16:05:10 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1762483
16:05:11 <openstack> Launchpad bug 1762483 in openstack-ansible "openstack-ansible rabbitmq-install.yml -e 'rabbitmq_upgrade=true' does not upgrade rabbitmq" [Undecided,New]
16:06:22 <hwoarang> makes sense
16:06:22 <evrardjp> that looks legit
16:06:31 <evrardjp> since we moved suse/centos to not use latest
16:06:46 <evrardjp> that's bad :p
16:07:01 <evrardjp> That's also why I said we should do it on case by case in each role
16:07:07 <evrardjp> but that thing got merged anyway
16:07:10 <hwoarang> it's ok you only want that on upgrades
16:07:36 <evrardjp> we can change the upgrade docs/code
16:07:44 <evrardjp> let's mark it as confirmed and medium?
16:07:50 <odyssey4me> yeah
16:08:30 <odyssey4me> I always thought that it made better sense to use present and to have packages upgraded through a different playbook which you can target as you wish.
16:09:21 <evrardjp> ok
16:09:35 <evrardjp> let's do it this way then
16:09:42 <odyssey4me> Then we can use it for minor/major upgrades too.
16:10:01 <evrardjp> I'll mark it as high instead
16:10:08 <odyssey4me> although in that role I suppose just tying the package state to whether the upgrade var is set or not will work as a workaround
16:10:17 <odyssey4me> and will be back portable
16:10:18 <evrardjp> odyssey4me: we can also use a flip switch to load some variables only when upgrades
16:11:04 <evrardjp> I'd rather do it in the role, because the role knows
16:11:08 <evrardjp> and we set a convention
16:11:14 <odyssey4me> yeah
16:11:20 <evrardjp> like "in_upgrade: False" by default in the role
16:11:30 <odyssey4me> anyway, I'd say high because it impacts upgrades quite seriously
16:11:34 <evrardjp> and if someone does is_upgrade: True ->boom!
16:11:40 <evrardjp> odyssey4me: agreed!
16:11:44 <evrardjp> ok done deal
16:12:21 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1762459
16:12:22 <openstack> Launchpad bug 1762459 in openstack-ansible "inventory-manage.py broken on 17.0.x" [Undecided,New]
16:12:25 <evrardjp> I guess that one is solved?
16:12:36 <odyssey4me> yeah, should be
16:12:41 <evrardjp> odyssey4me: it merged back in Q hasn't it?
16:12:52 <evrardjp> ok will mark it as fix commited
16:12:56 <odyssey4me> one more review for the pike backport: https://review.openstack.org/#/q/I6055df3da794439fb65ad2244445135dafb1cecf
16:13:21 <evrardjp> next
16:13:24 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1762421
16:13:25 <openstack> Launchpad bug 1762421 in openstack-ansible "openstack-ansible-galera_server role needlessly removes and reinstalls MariaDB-server" [Undecided,New]
16:13:36 <evrardjp> odyssey4me: thanks! :)
16:14:04 <odyssey4me> yeah, apparently this one's back - someone's going to have to confirm and figure out a fix
16:14:13 <odyssey4me> wasn't mnaser looking at it during the PTG?
16:14:19 <evrardjp> omg we should test this
16:14:28 <evrardjp> I have started to work on idempotency testing
16:14:34 <evrardjp> taking the patch from mhayden
16:14:40 <evrardjp> but we should do it for galera_server
16:14:56 <evrardjp> by running galera_client in the middle
16:14:58 <evrardjp> :)
16:15:34 <hwoarang> yeah the prob is that the MariaDB-server from mariadb repos defines 'mariadb-server' symbol
16:15:41 <hwoarang> which collides with the one from the RDO repos
16:15:52 <hwoarang> so you can't really tell if you are re moving the RDO package or the MariaDB package
16:16:05 <hwoarang> ideally we shouldn't be removing anything and let dnf or yum figure out what to do
16:16:11 <hwoarang> there must be a magic switch for that
16:16:11 <evrardjp> hwoarang: I agree there
16:16:34 <evrardjp> let's not compile mariadb from sources.
16:16:36 <odyssey4me> yeah, the package removal was only done because that was recommended by mariadb in their knowledgebase for the 5.5 -> 10 upgrade
16:16:36 <evrardjp> :D
16:16:44 <evrardjp> maybe we can remove it now?
16:16:51 <hwoarang> we could try
16:17:00 <odyssey4me> now that we're beyond that, and their docs say to just install over the top of it, we should be able to just do that
16:17:28 <odyssey4me> it'd make the install/upgrade process quite a bit simpler too
16:18:16 <evrardjp> yeah I think we are good to do that
16:18:24 <evrardjp> I will push up a patch tomorrow
16:18:32 <evrardjp> confirmed high?
16:18:49 <evrardjp> not critical
16:18:54 <odyssey4me> yeah
16:18:59 <evrardjp> but it's very disruptive to uninstall it
16:19:01 <evrardjp> :p
16:19:22 <evrardjp> if someone else wants to fix it, very welcome, I already have so much on my plate.
16:19:24 <evrardjp> :D
16:19:25 <odyssey4me> well, it'll only be doing it to one server at a time - which is fine in a cluster, but not when you only have one ;)
16:19:33 <evrardjp> next
16:19:41 <evrardjp> odyssey4me: haha true
16:19:43 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1762051
16:19:44 <openstack> Launchpad bug 1762051 in openstack-ansible "no logrotation on in /var/log/gnocchi (in gnocchi container)" [Undecided,New]
16:19:56 <evrardjp> that looks legit
16:20:00 <evrardjp> I never did it
16:20:02 <evrardjp> :p
16:20:31 <odyssey4me> we should perhaps use ansible-role-logrotate for all our log rotation
16:20:54 <odyssey4me> https://github.com/openstack/ansible-role-logrotate
16:20:56 <odyssey4me> very simple role
16:21:27 <evrardjp> I thought it was here: https://github.com/openstack/openstack-ansible-rsyslog_client/blob/master/templates/logrotate.j2
16:21:32 <evrardjp> yeah maybe we should
16:22:29 <evrardjp> confirmed medium?
16:22:37 <odyssey4me> looks like there should be log rotation done for gnocchi: https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-gnocchi-install.yml#L61
16:22:37 <odyssey4me> it's in pike too: https://github.com/openstack/openstack-ansible/blob/stable/pike/playbooks/os-gnocchi-install.yml#L58
16:22:37 <odyssey4me> yeah
16:22:54 <odyssey4me> well, actually, not confirmed - itlooks like there is, so we need proof of a bug
16:22:59 <odyssey4me> need more info
16:24:01 <evrardjp> that's interesting
16:25:35 <evrardjp> maybe it's not named .log ?
16:26:25 <odyssey4me> it should be, if https://github.com/openstack/openstack-ansible-os_gnocchi/blob/master/templates/gnocchi-systemd-init.j2#L13 is doing what it should
16:26:43 <evrardjp> ok so no need to list it in rsyslog_client_log_files
16:26:43 <qwebirc23189> odyssey4me thaaaaaaaanks a looooooooot everything worked properly with your help and there was no network issue thanks a lot. you are my hero.
16:27:03 <odyssey4me> qwebirc23189 no worries, glad you're over that hurdle
16:27:46 <evrardjp> for me you're right it's not confirmed
16:27:49 <openstackgerrit> Andy Smith proposed openstack/openstack-ansible-tests master: (WIP) Update to use oslo.messaging services for RPC and Notify  https://review.openstack.org/559127
16:28:48 <evrardjp> asking for more details in the bug
16:28:55 <evrardjp> marking it as incomplete
16:28:58 <evrardjp> next
16:29:09 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1761785
16:29:10 <openstack> Launchpad bug 1761785 in openstack-ansible "Unbound listening on all interfaces" [Undecided,New]
16:29:49 <evrardjp> Confirmed and low
16:29:53 <evrardjp> ok for everyone?
16:29:56 <odyssey4me> yeah
16:30:10 <evrardjp> next
16:30:12 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1761722
16:30:13 <openstack> Launchpad bug 1761722 in openstack-ansible "inconsistency in documenting the use of eth12" [Undecided,New]
16:31:24 <odyssey4me> yeah, that's very likely
16:31:45 <odyssey4me> confirmed, I'd say medium given that this stuff is a reference
16:32:07 <d34dh0r53> yep
16:32:09 <ArchiFleKs> coreywright: thanks for your analysis ;)
16:32:17 <spotz> put notes in and I'll grab it
16:32:25 <jrosser> this is one where i don't feel able to write what it *should* say myself
16:32:39 <spotz> jrosser: That's why I tell them to put notes in:)
16:32:54 <odyssey4me> I couldn't clarify either :/
16:33:14 <spotz> networking is confusing.... love spotz
16:33:39 <jrosser> anway historically lots of newcomers trip over this so it could do with a tidy
16:33:46 <coreywright> ArchiFleKs, yw!  (and a quick work-around is to set `delegate_to: "{{ hostvars[random_conductor]['ansible_host'] }}"`
16:34:42 <evrardjp> if you need wiring towards a container for your bridge ...
16:34:49 <ArchiFleKs> coreywright: thanks, are planning on proposing a patch ? I'm not sure a totally understood the way you would fix it
16:34:57 <coreywright> ArchiFleKs, instead of `delegate_to: "{{ random_conductor }}"`)
16:35:02 <evrardjp> that could be simplified
16:35:10 <evrardjp> coreywright: ArchiFleKs can you speak about that later please?
16:35:28 <evrardjp> Let's mark it as confirmed and medium
16:35:41 <evrardjp> and we should remove the documentation from this file, it's confusing everyone
16:35:50 <evrardjp> let's put that into the documentation with examples
16:35:54 <evrardjp> that can be more explanatory
16:36:05 <evrardjp> more text than just comments
16:36:20 <evrardjp> let's move on shall we?
16:36:27 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1761718
16:36:28 <openstack> Launchpad bug 1761718 in openstack-ansible "container networks docs not updated for bare metal l3 agent deploy" [Undecided,New]
16:36:41 <evrardjp> confirmed medium
16:37:00 <odyssey4me> yeah
16:37:03 <evrardjp> cloudnull: could you check on both bugs ?
16:37:10 <evrardjp> kill 2 birds one stone?
16:37:28 <evrardjp> just explaining better what we did with the hyper converged?
16:38:01 <evrardjp> I'm assigning it to you anyway :p
16:38:03 <evrardjp> ok
16:38:05 <evrardjp> let's move on
16:38:08 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1761550
16:38:09 <openstack> Launchpad bug 1761550 in openstack-ansible "'Reload systemd deamon' handle not being called in neutron" [Undecided,New]
16:39:11 <evrardjp> that is bad
16:39:27 <odyssey4me> if it's newton, then there's nothing we can do about it
16:40:07 <evrardjp> P and O can be fixed though
16:40:23 <evrardjp> if it's a problem in P and O
16:40:54 <odyssey4me> yeah, ocata onwards does the daemon reload when it stops/starts the service
16:41:00 <d34dh0r53> I'm reading that it's a problem in >p, but that could be me
16:41:04 <odyssey4me> newton does not
16:41:09 <evrardjp> looks like this sha is N
16:41:11 <evrardjp> 84ed68a4a57713c8a66aada1028dbbf8ffa1eb00
16:41:28 <d34dh0r53> ohh
16:41:31 <d34dh0r53> you're right
16:41:31 <spotz> Anyway to get permission to patch?
16:41:46 <odyssey4me> nah, for RPC we'll have to fork and fix in the fork
16:41:47 <evrardjp> to patch what?
16:41:53 <spotz> newton
16:41:56 <evrardjp> no
16:42:00 <odyssey4me> newton in the openstack namespace is gone
16:42:24 <evrardjp> so
16:42:33 <evrardjp> for me that's not clear if that's above N
16:42:40 <odyssey4me> I don't think it is.
16:42:42 <evrardjp> ok
16:43:24 <evrardjp> ok won't fix then
16:43:25 <odyssey4me> the notifies in https://github.com/openstack/openstack-ansible-os_neutron/blob/newton-eol/tasks/neutron_init_systemd.yml look fine though
16:43:35 <evrardjp> next
16:43:37 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1760878
16:43:38 <openstack> Launchpad bug 1760878 in openstack-ansible "galera_server: Reveals galera passwords in task: "Create galera users"" [Undecided,New]
16:43:46 <evrardjp> wow
16:43:48 <evrardjp> that's bad
16:43:51 <evrardjp> critical
16:43:58 <evrardjp> I will do it tomorrow
16:44:19 <odyssey4me> hmm, I though I saw a patch already for that
16:44:47 <evrardjp> I will see
16:44:51 <evrardjp> no worries I handle it
16:45:04 <odyssey4me> https://review.openstack.org/#/q/Ia170e4f25e35984e68b4cd47c3d964a034236d11
16:45:10 <evrardjp> next
16:45:10 <d34dh0r53> confirmed
16:45:11 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1759896
16:45:13 <openstack> Launchpad bug 1759896 in openstack-ansible "Placing certificate on keystone containers does not work" [Undecided,New]
16:45:32 <evrardjp> odyssey4me: not the same task
16:45:37 <evrardjp> mysql_user is the one
16:45:41 <evrardjp> but yeah
16:45:42 <odyssey4me> ah ok
16:45:52 <evrardjp> will do it
16:45:55 <evrardjp> let's move on
16:45:57 <evrardjp> 15 minutes
16:45:59 <evrardjp> :)
16:46:28 <odyssey4me> yeah, ssl on the back-end doesn't really work
16:46:33 <odyssey4me> confirmed, medium
16:46:36 <evrardjp> ok
16:46:38 <odyssey4me> I do wish someone would work on that
16:46:45 <evrardjp> anyone to look at it?
16:46:54 <odyssey4me> I think it'll be easier once the fastrouter is in though
16:47:01 <evrardjp> ok next
16:47:05 <evrardjp> #link #link https://bugs.launchpad.net/openstack-ansible/+bug/1758144
16:47:06 <openstack> Launchpad bug 1758144 in openstack-ansible "resolv.conf in containers set too late" [Undecided,New] - Assigned to Logan V (loganv)
16:47:06 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1758144
16:47:21 <evrardjp> it's assigned to logan let's move on
16:47:38 <d34dh0r53> should it be set to confirmed?
16:47:39 <evrardjp> except if logan- is there?
16:47:45 <evrardjp> d34dh0r53: logan- has to confirm
16:47:52 <evrardjp> We discussed this last week
16:47:56 <evrardjp> next
16:47:58 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1755821
16:48:00 <openstack> Launchpad bug 1755821 in openstack-ansible "config_template fails to parse template if it contains a comment with leading spaces" [High,New] - Assigned to Kevin Carter (kevin-carter)
16:48:06 <d34dh0r53> ahh, ok
16:48:22 <evrardjp> that's still in new state because we wanted to make sure cloudnull hasn't forgotten
16:48:29 <evrardjp> I guess he has, so let's move on
16:48:38 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1743032
16:48:39 <openstack> Launchpad bug 1743032 in openstack-ansible "Galera cluster maintenance in OpenStack-Ansible" [Undecided,New] - Assigned to Kevin Carter (kevin-carter)
16:48:44 <evrardjp> same
16:48:52 <evrardjp> next
16:48:54 <evrardjp> #link https://bugs.launchpad.net/openstack-ansible/+bug/1737827
16:48:56 <openstack> Launchpad bug 1737827 in openstack-ansible "(ceph-client): setting 'nova_ceph_client' results in deployment where volumes can't be attached to VMs" [Undecided,New] - Assigned to Andy McCrae (andrew-mccrae)
16:49:32 <evrardjp> andymccr: ?
16:49:58 <andymccr> ahh i was waiting on response from the bug filer
16:50:06 <evrardjp> ok
16:50:15 <evrardjp> marking it incomplet
16:50:17 <andymccr> ok cool
16:50:23 <andymccr> i'll take another look at it though
16:50:30 <evrardjp> unassigning
16:50:36 <evrardjp> ok you can if you want :)
16:50:38 <evrardjp> thanks
16:50:44 <evrardjp> and thanks everyone for the triage
16:50:50 <evrardjp> #topic open discussion
16:50:54 <evrardjp> because we have extra time
16:51:00 <evrardjp> open discussion
16:51:04 <evrardjp> I have a topic!
16:51:08 <evrardjp> 234 BUGS
16:51:14 <evrardjp> we've beaten our record.
16:51:26 <evrardjp> nothing to be proud off.
16:51:29 <spotz> that sounds bad:(
16:51:39 <evrardjp> it is
16:51:55 <hwoarang> :/
16:54:14 <evrardjp> I have nothing else to say. We have so many bugs, so many things pending. I am glad we are bringing new features and all, but at some point we need to look back and make sure our project is reliable and the experience is good. Not everyone is an expert that can quickly fix things.
16:54:29 <d34dh0r53> agreed
16:54:32 <odyssey4me> yep, agreed
16:55:48 <evrardjp> ok
16:55:51 <evrardjp> that's all I had to say
16:56:23 <evrardjp> anyone has anything to say ?
16:56:26 <openstackgerrit> Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/queens: Add container/service clean up for ironic/trove  https://review.openstack.org/559995
16:56:34 <odyssey4me> patches speak louder than words ;)
16:57:27 <evrardjp> indeed
16:57:31 <evrardjp> thanks everyone!
16:57:38 <evrardjp> closing up
16:57:41 <evrardjp> #endmeeting