17:00:15 #startmeeting qa 17:00:15 Meeting started Thu Jul 20 17:00:15 2017 UTC and is due to finish in 60 minutes. The chair is andreaf. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:16 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 17:00:18 The meeting name has been set to 'qa' 17:00:25 hello folks 17:00:25 o/ 17:00:54 #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Agenda_for_July_20th_2017_.281700_UTC.29 today's agenda 17:01:34 o/ 17:01:35 who else is here for the meeting today? 17:01:37 o/ 17:01:43 me :) 17:01:48 \o 17:01:54 oomichi, mtreinish, afazekas: around? 17:02:08 \o 17:02:11 o/ 17:03:05 ok let's start - welcome everyone 17:03:35 There are no action items from last meeting to be reviewed 17:03:36 #topic PTG 17:04:07 I just wanted to mention that I set-up an etherpad to start collecting ideas about how to best spend out face to face time at the PTG 17:04:14 #link https://etherpad.openstack.org/p/qa-queens-ptg PTG etherpad 17:04:35 if you think you are going to be there and/or have a topic to be discussed please enter it in there 17:04:46 even if you're not sure you're going to be in Denver 17:05:39 also I'm looking for volunteers for the QA help ours 17:06:16 andreaf, What exactly these help ours means ? 17:06:24 afazekas 17:07:34 afazekas: the proposal is that during the first two days WG will hold their sessions and in parallel horizontal teams will be available in a room to answer questions / solve problems for people from other teams 17:07:41 afazekas: there was a thread in the ML 17:07:57 o/ 17:08:15 afazekas: my plan is to rotate our presence in the room so that we can join WG (or SIG?) sessions when relevant 17:08:50 afazekas: since we have a community goal in queens specific to tempest plugins it would be good to provide help to whoever needs it on that 17:09:01 andreaf, ok, If I am going I can help 17:09:34 afazekas: cool, thanks! 17:09:44 afazekas: I hope you can be there 17:10:33 ok anything else on PTG? 17:11:34 #topic Gate statbility 17:11:40 this week chandakumar was signed up for https://etherpad.openstack.org/p/pike-gate-issue-categotisation - but he's not around 17:12:58 I don't have anything special to mention on stability today 17:13:34 Just one note about the puppet jobs in Tempest - they are non-voting but we should check them on a regularly 17:13:46 And not break them like I did with the orchestration client patch 17:14:24 EmilienM was kind enough to add the testr_html report to the test results, so they test results are easier to get as well now 17:14:30 o/ 17:14:38 andreaf: not merged yet 17:14:45 but merged very soon if not this week 17:15:34 thanks for that, it is helpful 17:16:02 it was good feedback, thank *you* for the idea 17:16:46 ok - anything else on gate stability? 17:17:34 #topic Spec reviews 17:17:35 #link https://review.openstack.org/#/q/status:open+project:openstack/qa-specs,n,z 17:17:38 I don't have anything today on spec reviews 17:18:07 andreaf: i reviewed your spec as best i understood it 17:18:36 But we will probably want to discuss some of them at the PTG, so it would be good to review them in advance 17:18:45 felipemonteiro: cool - the resources one? 17:18:48 yes 17:19:30 the multiple rbac policy i wrote still hasn't been reviewed and i probably won't attend the ptg, so would be nice to get some feedback on it before queens 17:19:34 oomichi took over that spec about a cycle ago, so it has been a bit out of my radar 17:20:29 felipemonteiro: right, I didn't get around reviewing it yet, I will try to do so tomorrow 17:20:55 andreaf: thanks, it's fairly straightforward, nothing revolutionary 17:21:05 felipemonteiro: ok 17:21:35 anything else on spec reviews? 17:22:30 #topic Tempest 17:23:03 prateek around? 17:23:34 looking for update on bug triage... 17:24:31 on Tempest side mtreinish is working on cred providers as stable interfaces, I'm working on test.py , clients.py and others as stable interfaces, gmann is working on swift client as stable interface 17:24:40 #link https://etherpad.openstack.org/p/tempest-test-module-stable some of the stable interface work 17:25:34 I would really like to close that work during Pike so comments and reviews on any of that would be very welcome 17:26:09 this is marked as top priority for Pike in our prio list, and it has high impact on the community since if we have more stable interfaces we are less likely to break the gate for folks 17:26:27 anything else on Tempest? 17:27:34 #topic Patrole 17:27:45 Anything on Patrole? 17:28:46 just updating documentation, striving to make some interfaces stable 17:29:34 felipemonteiro_ : nice 17:29:41 the policy in code goal for queens will have impact on Patrole - is that something we can address after the PTG ? 17:30:23 we already support policy in code. there's an intermittent period between the time when it's in policy.json and in code where the policies won't be discoverably by patrole, that's it to my knowledge 17:30:57 because requires updating setup.cfg with the appropriate entry point for oslo policy generator (and patrole) to discover policies 17:31:29 we don't support policy.yaml though...patrole actually needs more work in supporting custom policy files 17:33:34 felipemonteiro_: updating setup.cfg on patrole requires all services to have policy in code? 17:34:18 er, here's an example: #link https://github.com/openstack/nova/blob/79f6920eac135ab67476f75cc1d281e295cf0c84/setup.cfg#L39 17:34:28 keystone does the same, and murano 17:35:27 so if glance is making the transition to policy in code, they usually update their setup.cfg as the last step of the transition. until that's done though, patrole won't find the policies because they're not in the policy.json and not yet discoverable in the setup.cfg. 17:35:34 felipemonteiro_: heh ok so the change is on the service side, fine 17:37:34 felipemonteiro_: I see - so how do you deal with that for patrole CI? do you disable tests until the work is done? 17:38:43 we skip them to avoid breaking the gate. it's very unfortunate (we skip ~15 tests right now) but when keystone was doing the transition in p-1 we were skipping like 80. then as soon as the setup.cfg change was done in keystone, we went back to our normal # of skips... we have a CONF opt to make it fail the test though which is false in the gates. 17:39:55 felipemonteiro_: got it - thanks for bearing with my questions :) 17:40:04 np, ptl has to know these things :) 17:40:22 anything else on Patrole? 17:41:06 not from me 17:41:34 #topic Critical Reviews 17:41:45 Any review that requires urgent attention? 17:42:39 #link https://review.openstack.org/#/q/topic:cred_providers_clients 17:42:51 does that include one that affects tempest.lib? 17:43:34 felipemonteiro_: yes 17:43:34 it's the migration of credential providers to tempest.lib 17:43:46 #link https://review.openstack.org/#/c/483115/ 17:43:57 ^ i had in mind that, not sure if it's "critical" though 17:45:34 thanks 17:45:35 anything else? 17:45:52 #topic Open Discussion 17:46:25 Hey, I'm just using moment to ask for some review, I'm intern in the Outreachy program and I've submitted this patch #link https://review.openstack.org/#/c/472482/ fixing the ldap plugin on devstack and creating a ldap domain, could someone who understand devstack take a look at it? 17:46:32 What happened to my topic??? 17:47:01 Removal of core reviewers from openstack-dev/hacking (jlvillal) 17:47:34 #link https://review.openstack.org/#/c/472482/ 17:47:42 lwanderley: thanks I will have a look 17:47:54 jlvillal: doh sorry refresh problem - I see it now 17:48:00 heh :) 17:48:06 #topic Removal of core reviewers from openstack-dev/hacking 17:48:20 jlvillal: the floor is yours 17:48:35 Well I think we should remove those two core reviewers. 17:48:48 No reviews in over a year for one and almost a year for the other 17:49:02 And it would be nice if more cores were added or current cores reviewed more. 17:49:34 jlvillal: I have no objection to removing those two since they are not active anymore 17:50:00 speaking of which, there's a patrole core that can be removed too, david purcell 17:50:05 More reviewers would be nice. I haven't had any reviews on my patch from a core: https://review.openstack.org/#/c/447252/ 17:50:35 jlvillal: what's the amount of traffic on the project? do we have many open reviews? 17:50:53 Fairly low I think 17:51:07 https://review.openstack.org/#/q/project:openstack-dev/hacking+status:open 17:51:15 Five open reviews. So very small :) 17:52:48 jlvillal: I think we can handle this load on demand - i.e. if there is a patch that needs reviews ping in the QA room / ask for reviews in the meeting 17:53:08 andreaf: Okay. Thanks 17:53:14 Reviews please for: https://review.openstack.org/#/c/447252/ :) 17:53:34 jlvillal: more core reviewers are always welcome of course 17:53:44 jlvillal: are you volunteering to ramp up on it? 17:54:01 #link https://review.openstack.org/#/c/447252/ 17:54:12 andreaf: Uh... Maybe. Ironic core is keeping me pretty busy. 17:54:30 andreaf: But I'll try to review more of the patches. I see a couple which I haven't reviewed yet. 17:55:02 jlvillal: heh I bet it does keep you busy :) 17:55:22 andreaf: Yep. Especially since we lost four cores in the last six months :( 17:56:15 I'm done and happy to move to Open Discussion 17:56:37 ok 17:56:39 #topic Open Discussion 17:56:46 3 minutes left 17:57:34 I was wondering if anyone had negative / positive stories to share about office hours in other projects 17:57:42 but I guess there's little time left now 17:58:39 I'm starting to looking into those as a potential alternative to meetings for QA but I have no direct experience with them yet 17:59:11 andreaf, I suggest talk with the Keystone team, they are doing office hours for a while 17:59:34 raildo: will do, thank you! 17:59:46 andreaf, yw 17:59:56 thanks everyone for joining today 18:00:04 #endmeeting