00:01:10 #startmeeting qa 00:01:10 Meeting started Thu Dec 5 00:01:10 2019 UTC and is due to finish in 60 minutes. The chair is gmann. Information about MeetBot at http://wiki.debian.org/MeetBot. 00:01:11 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 00:01:13 The meeting name has been set to 'qa' 00:01:19 who all here today? 00:01:27 hi o/ 00:03:35 kopecmartin: i think two of us. I will skip agenda and discuss the critical review + your topic you want to discuss 00:03:49 gmann: ok, sure 00:03:54 irc nickname:- soniya29 Review pending 00:03:56 1. https://review.opendev.org/#/c/696068/ 00:03:56 2. https://review.opendev.org/#/c/696031/ 00:03:56 3. https://review.opendev.org/#/c/682835/ 00:04:10 soniya listed these review in list 00:04:43 yeah, i saw them, they are fine by me 00:05:14 great, you reviewed all of them. thanks. I have opened those to do review right after office hour 00:05:32 kopecmartin: from your side ? any review 00:05:41 ok, please also this one https://review.opendev.org/#/c/682964/ 00:06:33 and this one: https://review.opendev.org/#/c/696223/ .. i was triaging some review today .. so 00:06:40 well from my side 00:06:54 i'd like to bring this to the attention again https://review.opendev.org/#/c/692375/ 00:07:29 there's just a small proposal for different naming by afazekas 00:07:39 anything else i should change in that one? 00:08:05 for 682964- get_tenant_network() should be able to provide network for server. 00:09:01 yes, that's correct 00:09:15 the review is more about adding a default network (using that method) when none is specified 00:11:22 but the test class specifying the network is same way as base class does - from tenant_network = cls.get_tenant_network() only 00:13:01 i mean cls.default_networks from test class is doing the same what base class and common create_server used to fetch the network 00:14:22 hmm, i need to look closer with fresh mind tomorrow 00:17:03 kopecmartin: i commented on review, only creating the network resource should solve the issue - cls.set_network_resources(network=True, subnet=True) 00:17:30 gmann: great, thanks 00:18:13 basically test require network just need to mention the network resource requirement via cls.set_network_resources() and rest all will be taken care by base class 00:19:19 kopecmartin: i think i missed to comment on bug and patch for 'new filter to auth.py' i remember we discussed in some ofifce hour 00:19:26 i will recall that and comment 00:19:45 gmann: thanks! 00:20:05 one more update from my side, regarding the testing of tempest cleanup in gates 00:20:11 ok 00:20:26 i have finally reserved some time and i will propose a change within few days 00:20:33 perfect. 00:20:43 which job you are targeting ? 00:20:48 basically I'm thinking about a new role which can be included in a job by setting a parameter let's say 00:20:59 or you are adding in devstack tempest run logic 00:21:40 i don't know yet, gmann any recommendation? i would create a classic role, like tempest run is 00:21:42 new role? because 'member' not able to run cleanup ? 00:21:55 ohk, ansible role 00:22:04 yeah, sorry, i meant ansible role :) 00:22:52 as you know cleanup is not so stable so i will suggest to try on tempest-full-py3 via role as you mentioned. and if all stable for few months or in ussuri then we will enable that in devstack logic 00:23:16 that's sounds reasonable 00:23:21 +1 on role like tempest run 00:23:43 that will be easy to migrate to jobs and later in base devstack job 00:23:50 devstack-tempest job 00:24:32 but it will just simply run init-saved-state and then after test run tempest cleanup, that's all, i mean there is no point to add some more complex check if tempest cleanup deleted something and if it was to meant deleted, right? 00:24:53 we want to just run it so that we can see it won't fail on something stupid 00:26:04 as first yes. and logging the saved-state file will be helpful to 1. fix the resource leak tests 2. to verify that cleanup deleting the right thngs 00:26:17 any extra verification on that is next step if required 00:26:50 good, all set then 00:27:11 cool. thanks for working on that 00:27:20 as you are here and missed the PTG, i summarized the summary for QA related discussions- http://lists.openstack.org/pipermail/openstack-discuss/2019-November/011112.html 00:27:37 and this etherpad is for priority for us in ussuri - https://etherpad.openstack.org/p/qa-ussuri-priority 00:28:16 cool, thanks for the update 00:28:17 if you would like to help on any of those feel free to add your name 00:28:31 sure 00:28:40 thanks 00:29:19 last thing on office hour time, there is request from one member from India sonia to change the office hour so that folks from India can join 00:30:07 I think we need to change the time, i will try to propose the new time soon but not sure it will accommodate all of us or not. 00:30:40 ok, i don't have a problem with that, it's 1:30am in my time :D 00:30:51 yeah, it will be problematic to find a right time for everyone 00:32:03 kopecmartin: yeah, let's see. i will try to have some better time for you 00:32:11 1.30 AM is not good :) 00:32:38 it still could be worse :D 00:32:53 heheh 00:32:55 anything else for today ? 00:33:03 not from my side 00:33:31 ok. thanks for joining. let's close then 00:33:44 #endmeeting