15:00:15 <mgoddard> #startmeeting kolla 15:00:15 <opendevmeet> Meeting started Wed Jun 16 15:00:15 2021 UTC and is due to finish in 60 minutes. The chair is mgoddard. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:15 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:15 <opendevmeet> The meeting name has been set to 'kolla' 15:00:30 <mgoddard> #topic rollcall 15:01:03 <mgoddard> 8< 8< 8< 15:03:22 <parallax> o/ 15:03:29 <priteau> \o 15:03:39 <yoctozepto> \o/ 15:03:43 <hrw> ]o[ 15:04:23 <mgoddard> #topic agenda 15:04:28 <yoctozepto> (mgoddard: yup, but it could be threaded) 15:04:33 <mgoddard> * Roll-call 15:04:35 <mgoddard> * Agenda 15:04:37 <mgoddard> * Announcements 15:04:39 <mgoddard> * Review action items from the last meeting 15:04:41 <mgoddard> * CI status 15:04:43 <mgoddard> * Wallaby release planning 15:04:45 <mgoddard> ** libvirt/OVMF bug 15:04:47 <mgoddard> * Xena cycle planning 15:04:49 <mgoddard> ** master branch life cycle https://review.opendev.org/c/openstack/kolla/+/795320 15:04:51 <mgoddard> ** Testing log coverage (and more?) https://review.opendev.org/c/openstack/kolla-ansible/+/795512 15:04:53 <mgoddard> * Open discussion 15:05:39 <mgoddard> #topic announcements 15:05:48 <mgoddard> None from me 15:06:26 <yoctozepto> none 15:06:26 <mgoddard> #topic 15:06:29 <hrw> nope 15:06:29 <mgoddard> #topic Review action items from the last meeting 15:06:35 <mgoddard> mgoddard email openstack-discuss about quay.io credentials 15:06:37 <mgoddard> mnasiadka perform R-17 release tasks 15:06:45 <mgoddard> mgoddard: nope 15:06:49 <mgoddard> mnasiadka: started 15:07:07 <mgoddard> I'm dropping that one. It's on a personal TODO list, just low priority 15:07:12 <yoctozepto> ok 15:07:17 <mgoddard> #topic CI status 15:07:46 <yoctozepto> centos8 broke masakari-monitors in train 15:07:50 <yoctozepto> otherwise pretty green 15:07:52 <mgoddard> kolla: [train] libvirt-python dep broke masakari-monitors build on c8 15:07:58 <yoctozepto> with the random failure on neutron upgrades 15:08:26 <yoctozepto> re masakari-monitors and c8 15:08:29 <yoctozepto> we have a bug report 15:08:36 <yoctozepto> that it fails up to victoria 15:08:41 <yoctozepto> fails to build* 15:08:48 <yoctozepto> but CI confirms only for train 15:08:59 <yoctozepto> I did not have time to dive in 15:09:00 <yoctozepto> but priteau did 15:09:14 <priteau> I said I would try a build in victoria but lost access to my env so haven't completed it 15:09:22 <yoctozepto> https://review.opendev.org/c/openstack/kolla/+/796387 15:09:27 <yoctozepto> no problem 15:09:34 <yoctozepto> I wanted to mention the workaround ^^ 15:10:26 <mgoddard> doesn't seem to work 15:11:07 <yoctozepto> ? 15:11:08 <priteau> Are you saying this because of Zuul -1? The centos jobs passed 15:11:09 <yoctozepto> it built 15:11:14 <mgoddard> not debian 15:11:33 <yoctozepto> debian's patch is broken on centos8 not passing 15:11:43 <yoctozepto> https://review.opendev.org/c/openstack/kolla/+/795874 15:11:50 <yoctozepto> and so is final bump 15:11:55 <yoctozepto> and so is our em release 15:12:21 <mgoddard> so should we combine those patches and merge 15:12:23 <mgoddard> ? 15:12:56 <yoctozepto> debian is non-voting 15:13:08 <mgoddard> the workaround doesn't seem any worse than the existing code 15:13:09 <yoctozepto> I guess we could go with the workaround 15:13:17 <yoctozepto> and bury train 15:13:25 <yoctozepto> then worry about ussuri+ 15:13:54 <mgoddard> we might get pressure to tag our EM soon 15:13:59 <yoctozepto> indeed 15:14:01 <mgoddard> don't want to get caught with our pants down 15:14:03 <yoctozepto> I'm rechecking priteau's workaround 15:14:16 <mgoddard> ok 15:14:21 <yoctozepto> priteau to lift w-1 15:14:30 <yoctozepto> and /me rebasing other patches 15:14:33 <mgoddard> k 15:14:58 <priteau> I'll recheck too because of the POST_FAILURE 15:15:03 <opendevreview> Radosław Piliszek proposed openstack/kolla stable/train: Enable Debian updates repo https://review.opendev.org/c/openstack/kolla/+/795874 15:15:05 <priteau> You already did 15:15:12 <opendevreview> Radosław Piliszek proposed openstack/kolla stable/train: Bump final OpenStack versions for Train https://review.opendev.org/c/openstack/kolla/+/795258 15:15:17 <yoctozepto> fingers crossed 15:15:22 <yoctozepto> priteau: yeah 15:15:28 <yoctozepto> priteau: just need w0 from you 15:15:31 <priteau> done 15:15:36 <mgoddard> ok 15:15:37 <yoctozepto> thx 15:15:51 <yoctozepto> mgoddard: cr+2w+1? 15:15:54 <mgoddard> the train is ready to go 15:15:56 <yoctozepto> all the patches 15:16:08 <mgoddard> I'll wait for zuul 15:16:21 <mgoddard> #topic Wallaby release planning / libvirt/OVMF bug 15:16:37 <yoctozepto> mgoddard ok 15:16:40 <mgoddard> How are we looking with this one? 15:16:44 <yoctozepto> priteau also knows 15:17:29 <priteau> So, we appear to have a fix because c8 AV imported libvirt 7.0, although I would think it should *not* work 15:17:45 <priteau> But anyway, c8s is due to fix their AV repo package today 15:17:53 <priteau> Which will give us libvirt 7.4 15:18:12 <priteau> So I am refreshing http://mirror.centos.org/centos/8-stream/extras/x86_64/os/Packages/ multiple times, but still no updated package 15:18:53 <priteau> hughesjr said by wednesday cob us time 15:18:56 <mgoddard> So we continue to wait 15:19:08 <priteau> unfortunately yes 15:19:14 <mgoddard> hopefully tomorrow 15:19:44 <mgoddard> #topic Xena cycle planning / master branch life cycle 15:19:44 <priteau> If someone wants to proactively test libvirt 7.4, we could just update the repo file after the package has been installed 15:19:57 <mgoddard> #link https://review.opendev.org/c/openstack/kolla/+/795320 15:20:15 <mgoddard> Not much to say on this one 15:20:21 <mgoddard> Except please review 15:20:52 <mgoddard> #topic Testing log coverage (and more?) 15:20:58 <mgoddard> #link https://review.opendev.org/c/openstack/kolla-ansible/+/795512 15:21:55 <mgoddard> This is a patch that I've been iterating on recently, that could improve our log coverage 15:22:24 <mgoddard> We currently have no real test coverage of our logging 15:22:28 <yoctozepto> yeah, we need something in that direction 15:22:29 <mgoddard> This should help 15:23:24 <mgoddard> Basic idea is, we have an output plugin in fluentd image that writes a tab separated values file containing various labels for each log message 15:23:57 <mgoddard> Then in kolla-ansible we have a new test that ingests that TSV file and verifies that everything is as expected 15:24:35 <mgoddard> The test is implemented using py.test / testinfra 15:24:41 <mgoddard> #link https://review.opendev.org/c/openstack/kolla-ansible/+/795512/13/tests/test-fluentd.py 15:25:03 <yoctozepto> this obviously tests only whether we log at least one message 15:25:06 <mgoddard> I think we could really tighten up our testing by introducing this 15:25:18 <mgoddard> yoctozepto: correct, currently 15:25:22 <yoctozepto> yeah 15:25:36 <yoctozepto> I think someone worked on improving the fluentd integration 15:25:38 <yoctozepto> re parsing 15:25:52 <yoctozepto> so that we don't lose messages or get weird entries 15:25:59 <mgoddard> right 15:26:24 <mgoddard> we could do more in the output plugin to validate each fluentd event, currently it is quite minimal 15:26:37 <mgoddard> e.g. does it have at least Hostname, programname, log_level set 15:27:12 <mgoddard> I wanted to get the moving parts in place then it can be iterated on 15:27:37 <yoctozepto> great, I am all in 15:27:40 <yoctozepto> will review for sure 15:27:42 <mgoddard> excellent 15:28:19 <mgoddard> I think the existing tests/check-* scripts could be added to this py.test 15:28:32 <mgoddard> then we get results in one place 15:28:46 <mgoddard> and zuul and expose them as artifacts 15:29:19 <mgoddard> we're using this approach in kayobe already: https://opendev.org/openstack/kayobe/src/branch/master/playbooks/kayobe-overcloud-host-configure-base/tests/test_overcloud_host_configure.py 15:29:50 <yoctozepto> ++ 15:31:00 <mgoddard> Any other comments/suggestions? 15:32:09 <parallax> Looking good :) 15:33:15 <mgoddard> cool 15:33:17 <mgoddard> #topic Open discussion 15:33:28 <mgoddard> Does anyone have anything else to discuss today? 15:34:55 <headphoneJames> I'm reviving https://review.opendev.org/c/openstack/kolla-ansible/+/692179 15:34:55 <hrw> nope 15:35:38 <mgoddard> that's good headphoneJames 15:35:52 <mgoddard> are you still planning to email openstack-discuss? 15:36:28 <headphoneJames> Yes. Would like to get this one at least working first 15:36:59 <mgoddard> ok 15:39:51 <opendevreview> Mark Goddard proposed openstack/kolla master: docs: Update release management for Xena process https://review.opendev.org/c/openstack/kolla/+/795320 15:41:33 <mgoddard> Let's wrap up there then 15:41:36 <mgoddard> Thanks al 15:41:38 <opendevreview> Mark Goddard proposed openstack/kolla master: docs: Update release management for Xena process https://review.opendev.org/c/openstack/kolla/+/795320 15:41:40 <mgoddard> *all 15:41:43 <mgoddard> #endmeeting