15:00:16 #startmeeting kolla 15:00:16 Meeting started Wed Feb 9 15:00:16 2022 UTC and is due to finish in 60 minutes. The chair is mnasiadka. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:16 The meeting name has been set to 'kolla' 15:00:24 #topic rollcall 15:00:40 o/ 15:02:08 \o 15:02:22 o/ 15:02:25 \o 15:02:59 o/ 15:05:44 #topic agenda 15:05:44 * Announcements 15:05:44 * Review action items from the last meeting 15:05:44 * CI status 15:05:44 * Release tasks 15:05:45 * Current cycle planning 15:05:45 * Additional agenda (from whiteboard) 15:05:47 * Open discussion 15:05:48 #topic Announcements 15:06:08 FYI - The combined PTL and TC elections have started. 15:06:27 I've already put my name in the Kolla subdir for Z - if anybody else wants to run - feel free ;-) 15:06:48 mnasiadka go go go o/ 15:07:06 #topic Review action items from the last meeting 15:08:01 (anybody) to triage security bugs and update them with resolution plan (if needed) 15:08:01 mnasiadka post a patch for docs - standard topics that should be discussed over PTG and then revisited in mid-cycle 15:08:01 mnasiadka to triage security bugs and update them with resolution plan (if needed) 15:08:01 hrw to discuss with pynacl upstream to release binary wheel of 1.4.0 for aarch64 15:08:01 mnasiadka to propose a reno and change in docs claiming that CentOS 8 is EOL and we don't support it anymore on train/ussuri/victoria - and users are on their own... + stop any CI related to CentOS 8 (non-stream) 15:08:03 yoctozepto to remove debian master binary images from container registry 15:08:09 well, not anybody but me 15:08:25 but yes, triaging sec bugs is nearly done, will update them tomorrow with closure 15:08:37 I haven't posted patch for docs, promise to really do it this week 15:08:53 hrw is not here, I have no clue what's with pynacl and aarch 15:09:24 CentOS 8 patch series is done, allegedly centos-release-ceph-nautilus package is fixed, so I will pursue removal of the temporary RDO repo 15:09:27 I removed the images 15:09:33 fantastic 15:09:40 #action mnasiadka post a patch for docs - standard topics that should be discussed over PTG and then revisited in mid-cycle 15:09:48 #action mnasiadka to triage security bugs and update them with resolution plan (if needed) 15:09:58 that's about it I guess 15:10:02 and hrw's 15:10:06 ah right 15:10:12 #action hrw to discuss with pynacl upstream to release binary wheel of 1.4.0 for aarch64 15:10:18 #topic CI Status 15:10:55 I think it looks green-ish - not counting the Ceph jobs which I need to put some efforts in, seems Ceph devs like to break setups like ours (with public Ceph network over a vxlan fabric) 15:11:27 Merged openstack/kolla stable/ussuri: Drop CentOS Linux 8 image builds https://review.opendev.org/c/openstack/kolla/+/828403 15:11:46 Is there anything else that needs attention in CI? 15:11:51 what's wrong with vxlan fabric? 15:11:53 Can we update the status from RED to GREEN now? 15:12:00 yes, let's do it 15:12:14 yoctozepto: they assume the entry in /etc/hosts should point to your public Ceph cluster network... 15:12:42 Anyway, let's not spend time on this, need put some minutes in finding out how to fix it and get over with it. 15:12:46 oh my 15:13:20 #topic Release tasks 15:13:42 So, we had some R-8 tasks - hrw was on them, I think only Ubuntu had some issues? 15:13:53 yeah, with horizon 15:14:19 https://review.opendev.org/c/openstack/kolla/+/826488 15:14:23 seems that's merged 15:15:12 R-2 is feature freeze - Mar 14 - Mar 18 15:15:16 as a reminder 15:15:56 #topic Current cycle planning 15:17:01 So, looking at priorities - RMQ 3.9 needs a second review https://review.opendev.org/c/openstack/kolla/+/827656 15:17:48 The Ceph upstream repos seems to be a conflict now - because Kayobe has removed EPEL support and mgoddard's libvirt on host feature needs to install some Ceph packages for libvirt... 15:18:02 TL;DR - Ceph upstream repos need EPEL 15:18:27 And I'm not happy about it - at least not much. 15:18:57 meh 15:19:03 let's just drop centos and get over it 15:19:28 that's also a possibility 15:20:07 Opendev nodes with Rocky Linux 8 are coming soon, so we might want to discuss later how much CI would we like to run there as part of the RL8 support in Yoga. 15:20:48 But let's be realistic, we can't kill of CS8 before we move on to Rocky Linux 9 or something similar ;-) 15:20:53 *kill off 15:21:15 no, we can - move to debian like planned, or ubuntu if you like to have a company backing your system 15:22:24 Of course. 15:22:41 So - what about Podman? Should we discuss it now, or in the ,,additional topics'' slot? 15:22:52 either way works for me 15:23:10 we're not dropping centos stream 8 15:23:45 Michal Nasiadka proposed openstack/kolla master: rabbitmq: update to 3.9 https://review.opendev.org/c/openstack/kolla/+/827656 15:24:13 halomiva, hinermar: ready? 15:24:21 mgoddard: sadly~ 15:24:30 yep 15:24:38 Adding podman: Docker systemd patchset https://review.opendev.org/c/openstack/kolla-ansible/+/816724 15:24:38 dbus testing issues 15:24:38 requirements issues 15:24:38 adding abstract class 15:25:36 so, what are the issues, and are we able to fix them before mid March? 15:26:17 regarding the dbus problems, the dbus package is dependent on system packages so it can't be trivially installed on virtual environment 15:26:59 hmm 15:27:03 these issues are only for testing, deployment works fine 15:27:05 should not it be mocked out? 15:27:17 unit tests should not use system dbus 15:27:17 +1 15:29:45 I should have asked sooner because I have already finished unit test for it, which works in Python 3.6 but not in 3.9 as you can see in the Zuul report 15:29:45 anybody home? :D 15:29:50 Merged openstack/kolla-ansible master: [docs] Rewrite operating-kolla upgrade-wise https://review.opendev.org/c/openstack/kolla-ansible/+/809445 15:29:58 hmm 15:31:48 the problem seems to be somewhere in the difference of the python versions 15:32:12 ModuleNotFoundError: No module named '_dbus_bindings' 15:32:24 File "/usr/lib/python3/dist-packages/dbus/types.py", line 8, in 15:32:24 from _dbus_bindings import ( 15:34:15 yes, when I looked up that error, people said it vas caused by changing the python version 15:34:50 could be some C module 15:35:50 yeah, it is 15:35:53 or because it needs to match the system python version which has dbus? 15:36:10 it's compiled for py3.6 and tried to be used with py3.9 15:36:15 we can't really do that 15:36:20 it seems it's a CPython binding library 15:36:48 The 3.9 CI unit test job runs on Ubuntu? 15:37:00 yeah 15:37:44 Merged openstack/kayobe master: Remove deprecated kolla_tls_cert variable https://review.opendev.org/c/openstack/kayobe/+/821607 15:37:48 Merged openstack/kayobe master: Update documentation about CentOS support https://review.opendev.org/c/openstack/kayobe/+/824141 15:38:08 then it won't work for 3.9 15:38:41 yeah, I guess we are testing at the wrong level 15:39:34 I'd suggest we do a python mock of the dbus object, and do a more end to end test in the integration jobs 15:39:48 Verification of a change to openstack/kayobe master failed: Support building multiple disk images https://review.opendev.org/c/openstack/kayobe/+/812516 15:39:52 Verification of a change to openstack/kayobe master failed: Support elements repositories for overcloud host images https://review.opendev.org/c/openstack/kayobe/+/821413 15:41:18 Sounds like a good plan - any other issues to tackle for Podman support? 15:41:35 btw, why can't we just install https://pypi.org/project/dbus-python/ 15:41:40 inside the venv 15:41:43 what fails? 15:42:28 yea, what should be the next step? adding an abstract class or adding podman with that class in one commit? 15:44:05 yoctozepto: it seems that it can't see some system packages and fails at that 15:44:57 hinermar: did you try installing the dev package for dbus? 15:45:41 yeah, it was already present 15:45:55 odd 15:46:22 halomiva: re next step, I don't mind. I expect there will be a few itertions, and it may be easier to put it all in one patch to begin with 15:46:33 hinermar: let me edit your patch, ok? 15:47:08 how do i do that? :D 15:47:54 yoctozepto: or run unit tests locally? 15:48:31 so we can continue patching this one ? https://review.opendev.org/c/openstack/kolla-ansible/+/799229 15:48:48 mgoddard: yeah, it works for me with the dev pkg so I assume it's a CI issue 15:49:06 hinermar: just don't mind when I do ;-) 15:49:22 ahh, that's alright then 15:49:40 halomiva: yes, rebase that onto systemd 15:50:00 Radosław Piliszek proposed openstack/kolla-ansible master: Systemd container control https://review.opendev.org/c/openstack/kolla-ansible/+/816724 15:51:17 Ok, anything else to add, or can we move on? 15:51:33 just a quick help 15:52:05 how do I add requirement in openstack/requirements? 15:52:20 based on this error from Zuul: ERROR: Requirement '{Requirement(package='python-dbusmock', location='', specifiers='>=0.25', markers='', comment='# GPLv3', extras=frozenset())}' not in openstack/requirements 15:53:04 Radosław Piliszek proposed openstack/kolla-ansible master: Systemd container control https://review.opendev.org/c/openstack/kolla-ansible/+/816724 15:53:27 Radosław Piliszek proposed openstack/kolla-ansible master: Systemd container control https://review.opendev.org/c/openstack/kolla-ansible/+/816724 15:54:54 hinermar: ahm, you need to propose it against the openstack/requirements repo 15:56:08 okay, thanks 15:56:14 hinermar: could just add it to the blacklist: https://opendev.org/openstack/requirements/src/branch/master/blacklist.txt 15:56:47 yeah, best there 15:57:14 alright, that's all from us 15:57:52 ok then, it's two minutes until the end 15:58:07 yoctozepto: we'll talk about your topic next week, is that ok? 15:58:16 that's ok 15:58:18 no hurry 15:58:27 Ok, then let's finish for today. 15:58:31 thanks mnasiadka 15:58:37 have a nice evening 15:58:37 Thanks for attending and see you next week :) 15:58:39 #endmeeting