00:02:57 #startmeeting qa 00:02:58 Meeting started Thu Nov 14 00:02:57 2019 UTC and is due to finish in 60 minutes. The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 00:02:59 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 00:03:01 The meeting name has been set to 'qa' 00:03:04 who all here today ? 00:03:23 I am, hi o/ 00:03:23 hi 00:04:16 hi, let's start 00:04:22 #link https://wiki.openstack.org/wiki/Meetings/QATeamMeeting#Agenda_for_next_Office_hours 00:04:27 agenda ^^ 00:05:03 #topic Announcement and Action Item (Optional) 00:05:17 we had PTG last week which will talk about in next section 00:05:34 #topic OpenStack Events Updates and Planning 00:06:15 feedback forum notes are captured in this etherpad #link https://etherpad.openstack.org/p/PVG-forum-qa-ops-user-feedback 00:06:32 PTG #link https://etherpad.openstack.org/p/shanghai-ptg-qa 00:06:54 I will summarize the discussions in ML by this week. 00:07:08 ++ 00:07:13 Thanks! 00:08:33 #topic Sub Teams highlights (Sub Teams means individual projects under QA program) 00:08:49 Tempest 00:09:29 #link https://review.opendev.org/#/q/project:openstack/tempest+status:open 00:09:32 open review 00:10:00 kopecmartin: your cleanup cli fix is merged? 00:10:26 gmann: yeah, the one regarding admin manager yes, but we wrote others :) 00:10:43 kopecmartin: ok, we can review those, link ? 00:10:47 for example this one if you can take. a look https://review.opendev.org/#/c/692795/ Delete network quotas 00:11:29 kopecmartin: can we delete the quota ? 00:11:42 it might reset the env quota. 00:12:09 hmm, we do the same with volume and nova quotas 00:12:38 is it ? 00:13:39 it should be safe because cleanup will touch the project quotas only if the project will be deleted as well .. for example if the project is in saved_state.json, the quotas will remain with the project 00:14:28 what's interesting with quotas is, that if the project is deleted but quotas not, the quotas will remain in the database with no chance of removing it .. only by accessing the db directly 00:15:04 yeah, project level quotas which is going to delete is ok. but if that is system level quota (which might not be the case now but with unified limit things it can be) are something Tempest should nto touch 00:15:09 anyway I will check that 00:15:35 gmann: ok, that's all I'm asking :) thanks! 00:15:44 sure. 00:15:53 Patrole 00:15:53 i have also this patch for cleanup: https://review.opendev.org/#/c/694064/ 00:15:57 sorry 00:16:32 kopecmartin: thanks. i am worried about unit tests for cleanup if we can do or add that in gate job 00:16:54 kopecmartin: idea is to run cleanup CLI in some of the gate job and check if everything work fine. 00:17:12 ok, i'll have a look 00:17:38 it's true that cleanup is covered only by unit tests, no gate jobs :/ 00:17:50 kopecmartin: thanks. because unit test night not be applicable or complex for many cleanup changes but if we run on some job then we can make sure we are not breaking anything 00:18:05 kopecmartin: thanks. I will add your name for this items :) 00:18:22 gmann: ok .. i have one last patch i'd like to discuss 00:18:26 sure 00:18:27 this one https://review.opendev.org/#/c/692375/ 00:18:53 it's a remake of another patch (per afazekas suggestion), see comments 00:19:51 this one ? #link https://review.opendev.org/#/c/615558 00:19:58 gmann: yes 00:20:25 both are same or different ? 00:20:40 both are trying to solve the same problem 00:20:49 first one using a new option 00:21:24 ok, what is the issue it solve 00:21:36 if you can elaborate and then we can take that to review 00:22:19 the issue is testing radosgw by swift tests 00:22:32 the endpoint is slightly different in that case 00:22:50 and the patches are trying to make tempest to realize that 00:23:35 the info should be summarised in the releasenote: https://review.opendev.org/#/c/692375/1/releasenotes/notes/Add-new-filter-to-tempest.lib.auth.py-dc4225766bd2e0ff.yaml 00:23:35 'Ceph's RadosGW has a 'swift' prefix in its API ' from bug 00:23:47 eys 00:23:48 yes 00:24:10 is it not customize API and not upstream API ? i am not sure API url path is different per backend for swift 00:24:33 I remember of this bug and not sure how Tempest can support such API 00:25:28 I would like to get feedback from swift team (mattoliverau or tim) to answer on having different path for capability API. 00:25:29 well radosgw (in my understanding) should 'mock' swift in a ceph deployment .. therefore it would make sense to test it by swift tests 00:25:50 sure, i can ping them 00:25:57 kopecmartin: but Tempest is not the right place to tests radosgw 00:26:41 Tempest needs to run on APi endpoints registered in keystone. it can be anything but API path cannot be changed from service side 00:26:52 i will comment on those review 00:27:15 noted, i'll deep more into that, thanks for the feedback! 00:27:34 *dig more 00:27:38 any other review or fix you would like to bring ? 00:27:49 gmann: no, this is all for now :) 00:27:54 ok, thakns 00:27:58 let's move next 00:27:59 patrole 00:28:44 no much updates for this only things is we discussed in PTG about it to add in keystone and neutron gate. 00:28:52 any other sub team updates ? 00:29:25 nothing from my side 00:29:36 #topic Bug Triage 00:29:41 did you have any time to discuss the grenade job patches during the PTG? 00:29:44 uh, too late 00:29:58 #link https://etherpad.openstack.org/p/openstack_qa_tempest_2019_bug_review 00:30:41 we discussed to start the triage rotation. I will update the same in my PTG summary email 00:31:03 tosky: yeah, we did as part of community goal. moving grenade job is priority before that goal is started. 00:31:13 in parallel, I'd say 00:31:23 I will plan to have a common review with zuul team on that 00:31:45 tosky: goal is something we need to discuss on tc channel, it is accepted as pre-approved goal for V cycle 00:31:51 ok 00:32:14 not U? Ok, let's discuss it in a more appropriate time-slot 00:32:15 but grenade zuulv3 is priority irrespective of goal things 00:32:19 tosky: yeah 00:33:09 #topic Critical Reviews 00:33:28 any more critical review than what we discussed? 00:33:49 not from my side 00:34:06 nope 00:34:40 ok.if nothing else we will close today office hour. 00:34:49 sure, thanks! 00:35:01 thanks all for joining. 00:35:04 thank you 00:35:06 #endmeeting