16:00:29 <gthiemonge> #startmeeting Octavia
16:00:29 <opendevmeet> Meeting started Wed Apr 13 16:00:29 2022 UTC and is due to finish in 60 minutes.  The chair is gthiemonge. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:29 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:29 <opendevmeet> The meeting name has been set to 'octavia'
16:00:31 <gthiemonge> Hi!
16:00:40 <tweining> hi
16:00:45 <johnsom> o/
16:02:09 <gthiemonge> #topic Announcements
16:02:16 <gthiemonge> * New CI status etherpad
16:02:36 <rm_work> Hi
16:02:39 <gthiemonge> as discussed during the PTG, I started building an etherpad that collects all the (not so) "random" CI failures
16:02:48 <gthiemonge> #link https://etherpad.opendev.org/p/octavia-ci-failures
16:03:11 <gthiemonge> if you notice new issues (Octavia issues) please fill the etherpad or ping me on IRC
16:03:37 <johnsom> That second one makes sense. The first one I don't fully understand where the SG update is getting dropped
16:04:23 <gthiemonge> johnsom: an update on the 2nd one: it happens after a listener update (PUT), nearly 10 second after its creation, so this is still weird
16:04:38 <gthiemonge> I have DNM patch that tries to add debug for both cases
16:05:01 <gthiemonge> I reproduced twice the "0 listener reported"
16:06:03 <johnsom> At least in the wild it shouldn't impact anything as the next heartbeat should clear the situation
16:06:14 <johnsom> Just some wonky log messages I suspect
16:06:38 <johnsom> Oh, nevermind, it shows up in the operating status
16:06:46 <gthiemonge> yeah, one workaround would be to ignore the ERROR while wiating for the ONLINE status
16:08:08 <gthiemonge> I will update the test to ignore the ERROR status, at least it will show if the operating status switches back to ONLINE after a few seconds
16:08:35 <gthiemonge> in the current test, if the status is ERROR the listener is deleted immediatly
16:09:09 <gthiemonge> About CI, a gentle reminder about the recheck policy:
16:09:22 <gthiemonge> don't recheck without adding a reason to the message
16:09:38 <gthiemonge> for instance "recheck octavia-v2-dsvm-scenario Amphora reports 0 listeners when 1 expected"
16:09:49 <gthiemonge> or "recheck octavia-v2-dsvm-noop-api timeout", etc...
16:10:01 <johnsom> Well, don't recheck until the devstack/git issue is resolved either. Every devstack based job on Ubuntu is broken right now.
16:10:08 <gthiemonge> If you need help to understand why a job failed, ping me ;-)
16:10:39 * johnsom raises his hand for help
16:10:41 <johnsom> grin
16:11:08 <gthiemonge> you broke devstack, you have to fix it yourself /s
16:11:16 <johnsom> lol
16:11:33 <johnsom> Octavia first saw it, so it must be an Octavia bug
16:11:57 <gthiemonge> :D
16:12:13 <gthiemonge> * Zed Release Schedule
16:12:23 <tweining> johnsom: is that why you mainly work on Designate? ;)
16:12:51 <gthiemonge> We had a great PTG last week, thank you all!
16:13:02 <gthiemonge> Just a reminder, you can follow the Zed release cycle on this page:
16:13:07 * johnsom thinks tweining picks things up quickly....
16:13:09 <gthiemonge> #link https://releases.openstack.org/zed/schedule.html
16:13:15 <gthiemonge> lol
16:14:26 <rm_work> So what I forgot to bring up in the PTG was this spec we talked about: https://review.opendev.org/c/openstack/octavia/+/831619
16:14:47 <rm_work> Hoping to get this done Zed
16:15:35 <gthiemonge> rm_work: ok, thanks, I'll take a look
16:16:20 <gthiemonge> any other announcements?
16:19:11 <gthiemonge> nop
16:19:18 <gthiemonge> #topic Octavia CI status
16:19:59 <gthiemonge> Well, I added this new weekly topic about the CI
16:20:11 <johnsom> I noticed the scoped tokens job isn't running outside of the tempest tests. Is that right?
16:20:21 <gthiemonge> to talk about patches that would fix/improve our CI (we have a lot of open review in o-t-p)
16:20:46 <gthiemonge> do you mean it's not in the projects.yaml file in Octavia?
16:20:51 <johnsom> Right
16:21:17 <gthiemonge> Yeah we need to add it to the long list :D
16:21:37 <johnsom> Or pick a job to merge it with
16:21:55 <gthiemonge> I'm fine if we add a new job in octavia
16:22:01 <gthiemonge> we have the same gap with the c9s job
16:22:08 <gthiemonge> it's only in octavia-tempest-plugin
16:22:41 <gthiemonge> and about that, there was a recent update in c9s
16:23:10 <gthiemonge> the ssh client in c9s now excludes rsa ssh keys (like the key we generate in the devstack plugin and we use in octavia-tempest-plugin)
16:23:23 <gthiemonge> I have an update for the job:
16:23:28 <gthiemonge> #link https://review.opendev.org/c/openstack/octavia-tempest-plugin/+/837597
16:24:11 <johnsom> Should we prune the ussuri jobs?
16:24:37 <gthiemonge> yeah we should
16:24:52 <gthiemonge> and we still have the spare pool jobs for master
16:24:59 <gthiemonge> (I have an open patch to remove it)
16:25:23 <gthiemonge> but I think we cannot move forward until devstack is fixed
16:25:35 <johnsom> Right
16:27:17 <gthiemonge> there are also some patches to (re)enable act-stdby jobs as voting jobs in octavia and tempest
16:27:37 <gthiemonge> I will create a list of patches for the next meeting
16:30:18 <gthiemonge> #topic Brief progress reports / bugs needing review
16:31:10 <gthiemonge> I wanted to talk about a new patch about a SELinux issue (something I introduced a few months ago), but I saw that johnsom and tweining already commented in the review, so this is fine
16:31:14 <gthiemonge> but FYI
16:31:19 <gthiemonge> #link https://review.opendev.org/c/openstack/octavia/+/837721
16:32:08 <gthiemonge> and FYI we have a lot of backports in review:
16:32:19 <johnsom> I have two patches looking for a second +2:
16:32:19 <johnsom> https://review.opendev.org/c/openstack/octavia/+/831062
16:32:19 <johnsom> https://review.opendev.org/c/openstack/octavia/+/831087
16:32:19 <johnsom> Quick and easy reviews....
16:32:20 <gthiemonge> #link https://review.opendev.org/q/(project:openstack/octavia+OR+project:openstack/octavia-lib+OR+project:openstack/octavia-dashboard+OR+project:openstack/octavia-tempest-plugin+OR+project:openstack/python-octaviaclient)+status:open+branch:%255Estable/.*
16:32:39 <johnsom> Dang, forgot the link...
16:33:06 <tweining> we discussed a few changes that need review during PTG. See https://etherpad.opendev.org/p/zed-ptg-octavia
16:33:25 <tweining> btw. it seems the bot picks up links without #link too
16:33:34 <johnsom> Oh, does it now?
16:34:14 <tweining> yes, I saw it in the last notes that it picked up one of my links and I never use the tag
16:35:04 <gthiemonge> probably some smart regex ;-)
16:35:36 <johnsom> Neat
16:35:58 <gthiemonge> tweining: are you talking about the failover stop threshold patch?
16:36:18 <tweining> yes, for instance.
16:36:56 <tweining> first of all we need to agree about the new state
16:37:23 <tweining> see https://review.opendev.org/c/openstack/octavia-lib/+/835049
16:37:46 <gthiemonge> so I'm fine with the new state
16:39:22 <tweining> btw, I am currently working on tempest tests for that change
16:39:57 <johnsom> I think I commented on the main patch. The topics don't align, so I would have to track it down
16:40:42 <tweining> for reference: main patch is https://review.opendev.org/c/openstack/octavia/+/656811
16:41:24 <johnsom> Oh, looks like I didn't finish, I have pending comments
16:41:56 <gthiemonge> added a comment
16:42:52 <tweining> thanks
16:45:22 <gthiemonge> #topic Open Discussion
16:45:26 <gthiemonge> anything else folks?
16:45:48 <tweining> not from my side
16:46:22 <johnsom> Nothing else here
16:46:41 <gthiemonge> ok! thank you all!
16:46:43 <gthiemonge> #endmeeting