15:02:10 <mgoddard> #startmeeting kolla 15:02:11 <opendevmeet> Meeting started Wed Jun 9 15:02:10 2021 UTC and is due to finish in 60 minutes. The chair is mgoddard. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:12 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:02:15 <opendevmeet> The meeting name has been set to 'kolla' 15:02:18 <mgoddard> #topic rollcall 15:02:33 <mgoddard> \o 15:03:05 <yoctozepto> o/ 15:03:24 <headphoneJames> o/ 15:04:24 <priteau> ~o~ 15:06:11 <mgoddard> #topic agenda 15:06:17 <mgoddard> * Roll-call 15:06:19 <mgoddard> * Agenda 15:06:21 <mgoddard> * Announcements 15:06:23 <mgoddard> ** Kolla release process has changed 15:06:25 <mgoddard> * Review action items from the last meeting 15:06:27 <mgoddard> * CI status 15:06:29 <mgoddard> * Wallaby release planning 15:06:31 <mgoddard> ** libvirt/OVMF bug 15:06:33 <mgoddard> ** default kolla-build to network_mode == host (https://review.opendev.org/c/openstack/kolla/+/795089) 15:06:35 <mgoddard> * Xena cycle planning 15:06:37 <mgoddard> ** master branch life cycle https://review.opendev.org/c/openstack/kolla/+/795320 15:06:39 <mgoddard> * Open discussion 15:06:41 <mgoddard> #topic announcements 15:06:43 <mgoddard> #info Kolla release process has changed 15:06:46 <mgoddard> #link https://review.opendev.org/c/openstack/kolla/+/795320 15:06:58 <mgoddard> ^ proposed documentation for new release process 15:07:14 <mgoddard> Any other announcements? 15:07:43 <yoctozepto> nope 15:08:02 <mgoddard> #topic Review action items from the last meeting 15:08:22 <mgoddard> mgoddard email openstack-discuss about quay.io credentials 15:08:24 <mgoddard> mgoddard document new release process 15:08:26 <mgoddard> start following new release process 15:08:28 <mgoddard> mgoddard follow up about documentation improvements 15:08:33 <mgoddard> 1. nope 15:08:36 <mgoddard> 2. yep 15:08:48 <mgoddard> 3. not yet, still time 15:08:55 <mgoddard> 4. yep 15:09:05 <mgoddard> #action mgoddard email openstack-discuss about quay.io credentials 15:09:05 <yoctozepto> congrats 15:09:23 <mgoddard> #topic CI status 15:10:17 <mgoddard> Kolla broken 15:10:36 <mgoddard> We need https://review.opendev.org/c/openstack/bifrost/+/795508 (and a backport of it) to land before tomorrow 15:10:58 <mgoddard> I just pinged again in #openstack-ironic 15:11:54 <mgoddard> everything else happy I think 15:12:18 <priteau> I am seeing some failures on kayobe stable/ussuri, need to confirm if it random or repeatable 15:12:29 <mgoddard> hmm 15:12:33 <mgoddard> what kind of failures? 15:13:26 <priteau> While installing bifrost: Could not find the requested service mariadb: host 15:13:44 <mgoddard> hmm 15:13:50 <priteau> The other one was out of disk space 15:13:57 <mgoddard> haven't seen that 15:14:03 <mgoddard> keep us posted! 15:14:06 <priteau> recheck currently running, with errors again, will investigate 15:14:20 <mgoddard> #topic Wallaby release planning / libvirt/OVMF bug 15:14:35 <mgoddard> I think this is currently the only Wallaby release blocker 15:14:48 <mgoddard> which is good and bad 15:15:12 <yoctozepto> we are expecting a bump to libvirt 7.0 15:15:16 <yoctozepto> it's going to be fun ride 15:15:17 <priteau> The latest update is that the centos-release-advanced-virtualization package in c8s is likely going to be updated later today 15:15:27 <yoctozepto> that indeed 15:16:06 <mgoddard> so why did we hit this before if we had an older version? 15:16:59 <yoctozepto> that version had a bug 15:17:03 <yoctozepto> well, has 15:17:08 <priteau> All versions of libvirt are impacted, it's only fixed in 7.4.0 or older versions which include a backport 15:17:15 <mgoddard> ok 15:17:15 <yoctozepto> ++ 15:17:27 <priteau> Currently, only 6.0.0 in AppStream has the backport 15:17:36 <priteau> But not 6.6.0 from advanced-virt 15:17:50 <yoctozepto> btw 15:17:57 <yoctozepto> it's interesting we will get libvirt 7 15:18:02 <yoctozepto> while still on wallaby 15:18:05 <yoctozepto> just due to using stream 15:18:18 <yoctozepto> centos *scream* 15:18:43 <priteau> I wonder if it will affect TripleO in the same way? 15:19:02 <yoctozepto> no idea, tripleo is switching to an entirely independent model of release 15:19:31 <yoctozepto> (which I kinda like because it means they autobackport features to stable release of openstack) 15:19:44 <mgoddard> are we expecting any specific issues from the upgrade to libvirt 7? 15:19:52 <yoctozepto> we don't know really 15:21:04 <mgoddard> well, seems that we just need to sit tight for a little longer 15:21:06 <priteau> If we want to be proactive we could set up the c8s adv-virt repo and test 15:21:46 <yoctozepto> well, i'm guessing we will find out tomorrow, no? 15:21:46 <mgoddard> or we just let the stream wash over us 15:22:12 <priteau> probably soon enough 15:22:14 <mgoddard> s/stream/firehose/ 15:23:12 <mgoddard> let's move on, mere fish cannot stop the stream 15:23:28 <mgoddard> #topic Wallaby / default kolla-build to network_mode == host (https://review.opendev.org/c/openstack/kolla/+/795089) 15:23:35 <mgoddard> this one is mnasiadka 15:23:45 <mgoddard> but I think he's away 15:23:56 <mgoddard> still, should we do it? 15:24:12 <yoctozepto> +2 15:24:18 <yoctozepto> w8ing for mgoddard 15:24:39 <mgoddard> it kind of makes sense from a practical perspective, but should we make assumptions about the build host? 15:24:47 <yoctozepto> it will also fix for ipv6-only folks 15:24:53 <mgoddard> it may not even have had the bootstrap-servers treatment 15:25:06 <yoctozepto> host networking should always work 15:25:11 <yoctozepto> everything else is host networking 15:25:15 <yoctozepto> why make kolla different 15:26:48 <mgoddard> fine 15:27:38 <mgoddard> approved 15:27:48 <mgoddard> backport to W 15:27:53 <yoctozepto> ++ 15:28:12 <mgoddard> #topic Xena cycle planning / master branch life cycle 15:28:17 <mgoddard> #link https://review.opendev.org/c/openstack/kolla/+/795320 15:28:57 <mgoddard> That is the proposed new release process which should capture all the undocumented steps and now includes kayobe 15:29:27 <mgoddard> https://storage.bhs.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/zuul_opendev_logs_207/795320/1/check/openstack-tox-docs/20719e5/docs/contributor/release-management.html 15:29:29 <yoctozepto> I have yet to review it, but will 15:29:56 <mgoddard> it makes it quite obvious why we take a long time to release 15:30:27 <yoctozepto> kind of 15:30:29 <mgoddard> if anyone has ideas to simplify it, come forward 15:32:11 <mgoddard> Anyway, comments are welcome 15:32:25 <mgoddard> #topic Open discussion 15:32:33 <mgoddard> Does anyone have anything else to cover? 15:32:57 <yoctozepto> not really 15:33:03 <yoctozepto> need to mention we should release train em 15:33:06 <yoctozepto> and have it off us 15:33:53 <headphoneJames> At what point do we discuss new features proposals for Xena? 15:34:49 <yoctozepto> you can now 15:34:50 <mgoddard> headphoneJames: anyone can add to the meeting agenda 15:34:59 <mgoddard> headphoneJames: or open discussion is fine 15:35:24 <yoctozepto> open discussion is really: first come, first served 15:36:44 <headphoneJames> A feature I would like to explore is having a streamlined way to setup and enable secure RBAC for services in Xena that support it 15:37:06 <mgoddard> Sounds like a good idea 15:37:15 <mgoddard> What does it entail for each service? 15:37:23 <yoctozepto> I agree 15:37:46 <mgoddard> is it the enforce_scope setting? 15:38:13 <headphoneJames> I'm not sure yet. I've started looking at what being implemented in tripleo 15:38:35 <mgoddard> good plan 15:39:29 <mgoddard> there is this one for keystone: https://review.opendev.org/c/openstack/kolla-ansible/+/692179/ 15:39:34 <headphoneJames> but yes, things like keystone::policy::enforce_scope: True 15:40:18 <mgoddard> at the moment we get a lot of warnings in keystone about scope 15:40:43 <mgoddard> a good first step would be to change to use scoped tokens, without necessarily enforcing them 15:41:08 <mgoddard> the above patch does both 15:41:25 <headphoneJames> that is an excellent patch to know about 15:41:34 <opendevreview> Pierre Riteau proposed openstack/kayobe-config-dev stable/train: Synchronise with latest kayobe for Train https://review.opendev.org/c/openstack/kayobe-config-dev/+/795578 15:42:52 <headphoneJames> I will probably pick it up 15:43:04 <mgoddard> great 15:44:03 <headphoneJames> also, the first half of the let's encrypt patch is ready for review: https://review.opendev.org/c/openstack/kolla-ansible/+/741340 15:44:30 <mgoddard> I would also suggest starting a thread on openstack-discuss, inviting other deployment projects and the secure RBAC team to collaborate 15:44:57 <headphoneJames> mgoddard: good idea 15:45:00 <opendevreview> Pierre Riteau proposed openstack/kayobe-config-dev stable/victoria: Sync configs with kayobe @ 0a8f1fb67fb4cab7029cdf615f3280e975bcd1bf https://review.opendev.org/c/openstack/kayobe-config-dev/+/795580 15:46:18 <mgoddard> you might tag as [kolla][keystone][tripleo][openstack-ansible] 15:47:21 <priteau> puppet? 15:47:31 <yoctozepto> and charms 15:47:40 <priteau> need a [deploy] tag 15:47:44 <yoctozepto> indeed 15:49:21 <mnasiadka> sorry for being late 15:49:25 <mnasiadka> thanks for approving my patch :) 15:50:00 <mgoddard> you are lucky we didn't abandon it :p 15:50:21 <yoctozepto> LIKE YOU DID US 15:50:28 <yoctozepto> ;d 15:51:01 <mgoddard> anything else to cover today? 15:51:34 <mgoddard> oh, hold on 15:51:52 <mgoddard> we didn't discuss the release activities for this week 15:51:52 <opendevreview> Merged openstack/kayobe stable/wallaby: Add kayobe overcloud service prechecks command https://review.opendev.org/c/openstack/kayobe/+/795462 15:52:00 <opendevreview> Merged openstack/kayobe stable/wallaby: docs: improve automated seed development info https://review.opendev.org/c/openstack/kayobe/+/795463 15:52:27 <mgoddard> R-17: Switch source images to current release 15:53:23 <mgoddard> would anyone like to do this week's release tasks? 15:54:33 <mnasiadka> I can do that 15:55:26 <mgoddard> thanks 15:55:38 <mgoddard> #action mnasiadka perform R-17 release tasks 15:55:59 <mgoddard> the kayobe one merged already 15:56:10 <mgoddard> let's wrap up then 15:56:13 <mgoddard> Thanks all 15:56:18 <mgoddard> #endmeeting