22:00:41 #startmeeting qa 22:00:42 Meeting started Thu Apr 16 22:00:41 2015 UTC and is due to finish in 60 minutes. The chair is mtreinish. Information about MeetBot at http://wiki.debian.org/MeetBot. 22:00:43 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 22:00:45 The meeting name has been set to 'qa' 22:00:52 hi, who's here today? 22:00:59 hi 22:01:02 #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Proposed_Agenda_for_April_16th_2015_.282200_UTC.29 22:01:04 hi 22:01:07 ^^^ Today's agenda 22:01:48 dtroyer, jogo: around? 22:02:01 mtreinish: sort of 22:02:10 * andreaf lurks around 22:02:27 ok, lets get started 22:02:36 it'll probably be a pretty quick meeting i guess 22:02:48 #topic Summit Session Brainstorm 22:02:58 so I guess I left this on the agenda 22:03:10 although I really don't have anything to add on last week 22:03:19 oh wait, yes I do 22:03:30 ttx sent the numbers to the ML 22:04:10 we have 5 fishbowl sessions 22:04:18 4 work room slots 22:04:34 and a full day contributor meetup (shared with infra and relmgt) 22:05:04 #link http://lists.openstack.org/pipermail/openstack-dev/2015-April/061167.html 22:05:29 I'll put that on the brainstorming etherpad 22:05:35 #link https://etherpad.openstack.org/p/liberty-qa-summit-topics 22:06:05 other than that I didn't have anything else to talk about summit topics 22:06:09 does anyone have anything else? 22:07:23 ok, I'll take that as a no :) 22:07:28 lets move on 22:07:45 #topic Spec Reviews 22:08:09 does anyone have an open spec review to discuss? 22:08:19 mtreinish: #link https://review.openstack.org/#/c/173334/ 22:08:22 #link https://review.openstack.org/#/q/status:open+project:openstack/qa-specs,n,z 22:08:46 andreaf: heh, failed sphinx :) 22:09:50 mtreinish: this is something that I've been thinking about for a while, a YAML to describe test resources, which could also be used to provision them, all before testing - to solve various issues, like IDs in configuration, API calls for testscenarios, multiple guest images... 22:10:00 mtreinish: yes I know didn't have time to fix that 22:10:14 ok, yeah we've talked about doing that before 22:10:15 mtreinish: well I wouldn't mind a first impression comment from folks 22:10:20 I think it's probably a good idea 22:10:27 I'll take a look at the spec soon 22:11:08 I wrote some config docs this week so I know how crazy some of this stuff is 22:11:49 (not the proposal, the current state of providing images and stuff to tempest) 22:12:53 mtreinish: ok I will try to fix the spec to pass check 22:13:09 cool, thanks 22:13:20 ok does anyone else have a spec proposal to discuss? 22:14:20 ok, then let's move on 22:14:26 #topic Blueprints 22:14:41 does anyone have an in progress bps to discuss? 22:15:25 latest patches for ssh-auth-strategy: #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/ssh-auth-strategy,n,z 22:16:24 cool, I'll take a look tomorrow 22:16:40 although it looks like there is a -1 on the base patch 22:16:45 and regarding https://blueprints.launchpad.net/tempest/+spec/keystone-v3-jobs, keystone folks started working on a v3 only devstack - which will enable progress when ready 22:17:16 mtreinish: yes from afazekas 22:17:32 mtreinish: it's about not catching exceptions at tearDown 22:17:37 tearDownClass 22:17:47 apart from NotFound 22:17:50 ah ok 22:18:53 I have an update this week too 22:19:02 on test-accounts-continued for tempest 22:19:25 the network support for test accounts just landed 22:19:33 so all that's left are a few cleanup patches 22:19:40 mtreinish: ^_^ 22:19:58 and the script to help automate setting up an accounts file 22:20:20 that tool will also let us start having experimental jobs for the accounts path 22:20:43 so I can finally abandon https://review.openstack.org/158511 and https://review.openstack.org/173114 22:21:15 mtreinish: cool. it will be great to have gate job for account file. 22:21:57 gmann: yeah it will. I it also raises the question about whether we should use accounts or tenant isolation for our gating jobs 22:22:08 and which one makes a better default 22:22:32 mtreinish: is there any plan to move non-isolated and new account job in check/gate pipeline instead of experimental 22:23:16 gmann: well we can't gate on non-isolated because if there is no isolation it would have to be serial 22:23:21 which is too slow 22:23:41 but I think we can make one with the accounts file check or gate 22:23:54 ohkk 22:24:02 make sense 22:24:06 mtreinish, gmann: we could even mix the two in a single job 22:24:50 I also think in L we should start considering how we're going to deprecate the use of the identity config options for user and alt_user 22:24:50 mtreinish, gmann: each of them as benefits and potential to highlight issues 22:25:22 andreaf: yea, fully agree 22:25:38 mtreinish: yes we should deprecated those somehow 22:26:08 mtreinish: we should move away from having test resources in tempest.conf at all I think - only tempest behavioural settings 22:26:18 features on/off and so 22:26:20 andreaf: well I think after we tag tempest-4 we can start emitting a warning on their use 22:26:35 and send a ml post about it 22:27:05 andreaf: yeah that's probably a fair approach 22:27:32 mtreinish: we can also produce a bit of code that builds a YAML file from those options if found, and drop the code path that supports them in conf 22:27:41 luckily most of the config file is already that way, it's just a few things like images and flavors that are in config 22:29:08 andreaf: oh, you mean drop the NonLockingAccounts provider and just put a shim in place 22:29:12 yeah that's an idea too 22:29:50 ok, does anyone have anything else to discuss on bps? 22:31:06 let's move on then 22:31:12 #topic Devstack 22:31:24 I don't think dtroyer is around 22:31:36 so does anyone else have anything to discuss about devstack today? 22:32:45 heh, I'll that as a no 22:32:53 so let's move on 22:33:02 #topic Grenade 22:33:23 so this week sdague has been continuing on his refactor to make grenade more modular 22:33:27 and support external plugins 22:34:01 I think he's pretty close to wrapping it up, which is awesome 22:34:30 all the patches are here: 22:34:32 #link https://review.openstack.org/#/q/status:open+project:openstack-dev/grenade+branch:master+topic:modular_grenade,n,z 22:35:13 I didn't really have anything else to add on grenade 22:35:16 mtreinish: how about having ext plugin for tempest too? 22:35:19 jogo: did you have anything else? 22:35:42 mtreinish: nope 22:35:52 gmann: so tempest is a bit different, because it just uses unit test so it already is pretty modular 22:36:00 gmann: that's why we did tempest-lib 22:36:26 mtreinish: i meand ext plugin for devstack 22:36:29 so there was an externally consumable interfaces for people to use so they can write tempest like tests 22:36:55 gmann: oh, so write a plugin for devstack to configure tempest? 22:37:00 yes 22:37:11 I'm not sure we really need that because that already lives in the devstack repo 22:37:19 like ec2API repo have 22:37:59 gmann: sure, but isnt that what devstack's lib/tempest already does 22:38:16 I mean I guess we could refactor that to live in the tempest repo, but I'm not sure where the value is 22:38:44 but maybe I'm missing something 22:39:24 mtreinish: yes, thats why m also not sure about plan for ext plugin is for new repo only or moving existing one also. 22:40:28 gmann: well that's what http://git.openstack.org/cgit/openstack-dev/devstack/tree/FUTURE.rst covers there is a set of things which will stay in devstack 22:40:39 some core services which includes tempest 22:40:52 humm 22:40:58 but the rest which operate at a higher layer will be converted to plugins 22:41:16 that's what I started to cover in: https://review.openstack.org/152683 22:41:49 I see 22:43:14 ok is there anything else on grenade? 22:44:37 ok lets move on then 22:44:40 #topic Bugs 22:44:58 gmann: anything on bugs this week? 22:45:05 oh I had the rotation didn't I 22:45:12 oops, I forgot... 22:45:22 mtreinish: yea :) 22:45:51 and it looks like we're back in double digits because of that 22:46:01 I'll go through the open bugs tonight or tomorrow then 22:46:20 gmann: do you want to make another ML about another round of triage volunteers? 22:46:36 mtreinish: yea, I will add dates on etherpad and send ML 22:46:56 cool 22:47:06 were there any particular bugs people wanted to discuss? 22:48:39 ok then let's move on 22:48:42 #topic Critical Reviews 22:49:04 does anyone have any open reviews they'd like to get extra eyes on? 22:49:07 #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/resource-cleanup,n,z 22:49:26 the cleanup after resource-cleanup 22:50:02 ok, I'll try to look at the most recent rev again soon 22:50:25 #link https://review.openstack.org/#/c/172507 22:50:36 that's the start of a longish branch for some config docs 22:51:01 I'd appreciate it if people could review them and let me know what I screwed up 22:51:25 #link https://review.openstack.org/#/c/162875/ 22:51:41 feel free to add on the negative comments even if there are some, I'll try to do batch updates instead of reply to each individual review one by one 22:52:43 gmann: how close is that refactor to being finished? 22:52:59 after 162875, last patch i will update for schema cleanup 22:53:05 ok, cool 22:53:24 only 2 lest #link https://review.openstack.org/#/q/status:open+project:openstack/tempest+branch:master+topic:bp/rearrange-nova-response-schemas,n,z 22:53:40 *left 22:54:06 ok are there any other reviews to bring up? 22:55:17 ok then let's open the floor 22:55:21 #topic Open Discussion 22:55:31 in the last 5 min does anyone have anything else to bring up? 22:56:01 I was chatting with Stuart today 22:56:15 from the glance team - sorry I don't know his irc nick 22:56:26 about testing of python-glanceclient 22:57:01 atm most python clients are dsvm-src-(master|juno|icehouse) tests 22:57:02 oh moving the cli tests? 22:57:17 because someone needs to do that for glance still... 22:57:25 I know 22:57:36 andreaf: yeah to test the clients are working against any version of glance 22:57:53 because you should be able to use new glance client against an old cloud 22:58:02 although it's only really tested via devstack setup 22:58:06 right 22:58:22 but that just means that glance needs to start working on functional client tests 22:58:24 but once we move CLI tests away from tempest we won't be testing that anymore 22:58:25 which is a real gap 22:58:56 andreaf: sure, but the migrated cli jobs are still dsvm, so they can setup stable jobs for them too 22:59:19 if that's what they're looking for 22:59:29 yes - only they need to setup juno and icehouse jobs for those 22:59:32 andreaf: we can take this to -qa since we're at ~30secs 22:59:37 and I havent seen anyone doing it 22:59:42 not that I know of 22:59:58 but I think the first prio was to just get them spun up 23:00:12 tbh, from what I can tell most people don't really think about the python-*clients much 23:00:17 anyway we're at time 23:00:21 thanks everyone 23:00:22 #endmeeting