14:00:10 #startmeeting tripleo 14:00:11 #topic agenda 14:00:11 * Review past action items 14:00:11 * One off agenda items 14:00:11 * Squad status 14:00:11 * Bugs & Blueprints 14:00:11 * Projects releases or stable backports 14:00:12 * Specs 14:00:12 Meeting started Tue Jun 25 14:00:10 2019 UTC and is due to finish in 60 minutes. The chair is mwhahaha. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:12 * open discussion 14:00:13 Anyone can use the #link, #action and #info commands, not just the moderatorǃ 14:00:13 Hi everyone! who is around today? 14:00:13 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:15 The meeting name has been set to 'tripleo' 14:00:18 o/ 14:00:19 o/ 14:00:24 «o/ 14:00:34 o/ 14:00:36 o/ 14:00:57 o/ 14:01:02 o/ 14:01:44 hellow 14:01:46 \o/ 14:02:46 o/ 14:02:56 o/ 14:03:18 we've got an assortment of people and not much on the agenda so let's go 14:03:22 #topic review past action items 14:03:23 None. 14:04:21 #topic one off agenda items 14:04:21 #link https://etherpad.openstack.org/p/tripleo-meeting-items 14:04:39 currently the agenda is empty. anyone have something they want to bring up? 14:06:31 sounds like no 14:06:45 moving 14:06:51 #topic Squad status 14:06:51 ci 14:06:51 #link https://etherpad.openstack.org/p/tripleo-ci-squad-meeting 14:06:51 upgrade 14:06:51 #link https://etherpad.openstack.org/p/tripleo-upgrade-squad-status 14:06:52 edge 14:06:52 #link https://etherpad.openstack.org/p/tripleo-edge-squad-status 14:06:53 integration 14:06:53 #link https://etherpad.openstack.org/p/tripleo-integration-squad-status 14:06:54 validations 14:06:54 #link https://etherpad.openstack.org/p/tripleo-validations-squad-status 14:06:55 networking 14:06:55 #link https://etherpad.openstack.org/p/tripleo-networking-squad-status 14:07:04 cloudnull: is there a status etherpad or something for the transition squad? 14:07:29 there is, let me grab a link 14:07:47 #link https://etherpad.openstack.org/p/tripleo-ansible-agenda 14:08:09 that will likely change / update once we get the storyboard bits going. 14:08:21 k sure let me know 14:08:36 we ran into a couple issues with storyboard itself which are being resolved within the core project 14:09:09 sounds about right 14:09:33 example - 14:09:34 #link https://review.opendev.org/#/c/667275/ 14:09:41 but things are moving along 14:11:10 with that pending, I've still been working off of the trello board - https://trello.com/b/tEjJlT3k/tripleo-ansible-content-import - though I'll be closing that down as soon as possible 14:11:37 sounds good, thanks 14:11:43 any other status related items? 14:11:44 we should still adopt the use of standardized openstack-tox-molecule template (even if we just use it for parenting) 14:12:54 is there a reason not to? 14:15:11 right now we cant in tripleo-ansible, that job requires docker and we have jobs which are not running in docker due to them needing access to the host. I commented on that in https://review.opendev.org/#/c/667171 14:17:36 we are using the underlying parent molecule job "tox-molecule" 14:18:14 cloudnull: that's good! 14:19:14 cool anything else? 14:19:19 still the tox.ini connfig does not scale up to more runs. People should be able to add new molecule scenarios without having to define new jobs, or to edit tox.ini file. 14:20:50 one thing that I observed over the weekend was that ``pytest --collect-only` was failing to run with some exceptions caused by molecule/tests/ being recognizeded as unittest. anyway, i it is an easy fix. 14:21:19 once done, we should be able to list all scenarios with that command, and run individual one the same way we do with normal unittests. 14:23:53 one thing that i am not sure about is about triggering tests only when we modify the roles, based on my own experience a role can broke over night without being touched, only because some external things were updated (os images, ansible version, repo content,...). Probably we want to assure that we always run all tests at least on gate. for check we can optimize. 14:24:16 cloudnull: what do you think? 14:24:50 maybe? I think we get that with the integrated gates. 14:25:53 I disagree about the scalability of the config files and would argue that running all jobs on every change would be the scalability issue. we have tooling to add items to those files and have documented what's required. 14:26:23 so it should be fairly straight forward to add things to the repo 14:27:09 #link https://docs.openstack.org/tripleo-ansible/latest/contributing.html 14:27:28 I think we can extend those docs to contain more detail, but its a good start 14:28:08 Alex Schultz proposed openstack/tripleo-ci master: Make tripleo-ci-fedora-28-standalone non voting https://review.opendev.org/666834 14:28:41 cloudnull: what will happen when we will have 30-40 roles inside? are we going to have the same number of tox environments? 14:29:07 probably 14:29:47 there's no limit to the number of environments we could have and its not like its unmanageable 14:30:02 i bet openstack-infra would be very upset of seeing this number of nodes allocated by zuul, we will be forced to wait 12h before starting these jobs. 14:30:12 no. 14:30:31 if you look at the recent changes only the jobs impacted by a given change run 14:31:31 and the current longest gate job is <15 minutes 14:32:35 i also have concerns about the number of jobs in the tripleo-ansible role 14:32:47 cloudnull: i am trying to explain the node-number-allocaton is an issue, not the length of a single molecule job (which is very short). 14:32:54 let's have a more in depth conversation about this in the metting on thursday 14:33:16 i may not be a problem right now, but could easily become one. just wanted to raise awareness for the moment. 14:34:32 #action cloudnull, zbr|ruck, mwhahaha, et al have further discussions around tripleo-ansible molecule CI structure @ next transformation meeting 14:35:11 ++ 14:35:45 alright moving on for now 14:35:53 #topic bugs & blueprints 14:35:53 #link https://launchpad.net/tripleo/+milestone/train-2 14:35:53 For Train we currently have 25 blueprints and 527 (+8) open Launchpad Bugs. 527 train-2. 109 (+6) open Storyboard bugs. 14:35:53 #link https://storyboard.openstack.org/#!/project_group/76 14:36:12 so i asked infra to nuke the tripleo-ui storyboard items but it seems like it didn't work so i'll have to follow up with them 14:36:58 any other bug/blueprint related items? 14:39:36 sounds like no, moving on 14:39:37 #topic projects releases or stable backports 14:39:47 I think we did a few stable releases last week 14:40:12 Are there any backport related issues? 14:42:20 Ok moving on 14:42:23 #topic specs 14:42:24 #link https://review.openstack.org/#/q/project:openstack/tripleo-specs+status:open 14:42:58 fyi i -2'd the various open specs. If you need them unblocked before the next cycle, please let me know 14:43:45 #topic open discussion 14:43:48 any other items? 14:46:00 there is a topic for generating pdf docs https://review.opendev.org/#/q/topic:build-pdf-docs+(status:open+OR+status:merged) 14:46:39 I'm looking for TeX experts to shoot possible warnings off. I'm planning to add something similar for tripleo repos 14:47:01 ok thanks 14:47:14 It was surprisingly low amount of warnings for the python-tripleoclient pdf :) 14:47:25 pffft docs 14:47:27 this prolly means we are great on writing docs! 14:47:42 a likely story 14:47:52 actually i think we dynamically generate them for tripleoclient 14:48:06 nice bogdando this is one of the community goals right 14:48:12 yes, we are dynamically good as well 14:48:25 "pdf docs i mean " and community goals i mean voted as summmit 14:48:31 yea 14:48:32 s/as/at 14:49:08 * bogdando also wonders why cp fails in the tox commands 14:49:24 #link https://review.opendev.org/#/c/667159/ 14:50:16 e.g. http://logs.openstack.org/59/601659/11/check/openstack-tox-docs/9edd585/job-output.txt.gz#_2019-06-25_10_28_45_710782 14:51:56 bogdando: it's quoted? or maybe the html is missing 14:52:18 it's likely the ' though 14:52:27 mwhahaha: magic 14:52:28 because /bin/cp: cannot stat 'doc/build/pdf/*.pdf': No such file or director 14:53:24 it works in the shell... 14:53:42 anyway we can follow up 14:53:42 yup works in devstack 14:53:46 :D 14:53:50 * mwhahaha trolls his way out 14:53:59 alright, anything else? 14:56:11 Sounds like no. Thanks everyone. 14:56:17 #endmeeting