08:30:44 #startmeeting Yardstick Work Meeting 08:30:44 Meeting started Mon Feb 6 08:30:44 2017 UTC. The chair is kubi001. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:30:44 Useful Commands: #action #agreed #help #info #idea #link #topic. 08:30:44 The meeting name has been set to 'yardstick_work_meeting' 08:30:54 #topic Roll call 08:30:57 #info kubi 08:30:59 #info Jing 08:31:02 #info Ross 08:31:03 #info Jack 08:31:11 #info Rex 08:31:33 #info Kanglin 08:31:35 rbbratta: hi, thanks for attending the meeting with a late time :) 08:31:50 no problem 08:32:37 well, actually I'm interested in getting Jenkins fixed 08:33:09 yes, i've added a topic about it 08:33:15 rbbratta: Ok, Let 's make it as first topic 08:33:39 #topic Yardstick Jenkins job fail fix 08:34:22 I think we need to merge https://gerrit.opnfv.org/gerrit/#/c/27479/ to fix the python 3 unittests 08:35:02 rbbratta: got it 08:35:32 and then I think we need to merge https://gerrit.opnfv.org/gerrit/#/c/28091/ to fix AMQP failures 08:35:48 #infplease review the patch and merge it ASAP https://gerrit.opnfv.org/gerrit/#/c/27479/ , https://gerrit.opnfv.org/gerrit/#/c/28091/ 08:35:58 #info please review the patch and merge it ASAP https://gerrit.opnfv.org/gerrit/#/c/27479/ , https://gerrit.opnfv.org/gerrit/#/c/28091/ 08:36:12 and then recheck everything and see what else fails. 08:36:24 rbbratta: thanks for fixing the jenkins issue 08:36:38 have seen these patches, will review them ASAP 08:36:57 i think this patched can be merged quickly to check ci 08:36:58 is there a way to re-trigger all failed jobs? 08:37:21 anyway, those two fixes should hopefully unblock everything 08:38:04 but also note that Python3 unittests are enabled now, thanks to releng enabling tox. 08:39:01 understood 08:39:01 that's all I wanted to say on the topic. 08:39:40 does it rely on jenkins server to have python3-dev? 08:39:52 #action kubi will re-trigger all fail jobs 08:39:58 since we've already enable py3 test 08:40:14 that was fixed, added back in apt-get install python3-dev before tox 08:41:45 so it's fixed in releng or in the jenkins server? 08:42:01 fixed in releng 08:42:31 #link https://gerrit.opnfv.org/gerrit/#/c/28087/ 08:42:34 thanks, i see 08:42:48 I don't know where the Jenkins slaves are configured, I created Jira ticket, we could ask releng about the slave setup. 08:42:56 but for now we just sudo apt-get install 08:43:21 https://jira.opnfv.org/browse/RELENG-190 08:45:17 rbbratta: I think that would be OK for now 08:46:01 i think each patch should rebase to pass jenkins verify 08:46:26 since patch after tox.ini should be rebased 08:47:12 or the verify job cant find tox.ini file 08:47:32 ah yes, need to be rebased on top of job with tox.ini 08:47:42 Mingjiang: good point 08:48:00 Another thing; KVMFORNFV just added two scenarios in our test suite, once the verify is fixed we can review and merge the patch asap. 08:48:01 08:48:36 #info all patch should be rebased for the tox.ini 08:48:44 It is kinda weird to include tox.ini in the project, but dovetail and others include tox.ini as well. 08:49:39 here is the link of kvm patch 08:49:46 #link: https://gerrit.opnfv.org/gerrit/#/c/28085/ 08:50:41 #info KVMFORNFV just added two scenarios in yardstick test suites, please help to review #link: https://gerrit.opnfv.org/gerrit/#/c/28085/ 08:50:47 rbbratta: can you explan about it? 08:51:01 why tox.ini should not be included 08:52:00 https://github.com/openstack/nova 08:52:24 well, in theory someone could submit a review that disables a tox check and Jenkins would then +1 the commit. 08:52:25 It seems that nova put tox.ini in project. 08:53:04 we would have to check in code-review that nobody is modifying tox.ini incorrectly. 08:53:28 but I guess it doesn't matter since run_tests.sh is also in the project. 08:53:31 rbbratta: I see. 08:54:05 rbbratta: good point 08:54:38 #info since tox.ini were included in project, we would have to check in code-review that nobody is modifying tox.ini incorrectly 08:55:29 But where should we put it? It seems in project is the only choise. 08:55:44 got it, thanks : ) 08:56:26 got it. learned a lot 08:56:45 in theory we could embed it in releng somehow, like with a < But if we produce it dynamicly, how to run it locally? 08:58:27 that's true too, you'd lose that capability. 08:59:18 or we could include a sample, but then modify releng to use a specific generated file echo | tox or something 09:00:14 So if anyone modify it and submit a patch, we should merge it more carefully. 09:00:53 yes, just check for tox.ini changes 09:02:59 I think that would be a good solution. 09:04:09 #info we could include a sample tox.ini in the project, but then modify releng to use a specific generated file echo | tox or something 09:04:41 OK. It seems that we are already over the time, any more comments on this topic? 09:05:31 #topic others 09:05:50 any others topic? 09:06:01 Kanglin: how about you? 09:07:19 It seems that no more topic for toady 09:07:26 OK, thanks all for today 09:07:31 kubi001: nothing 09:07:53 Kanglin: will you attend this week HA meeting? 09:08:32 kubi001: yes, qiujuan and me will attend the meeting , discuss HA test cases on D release. 09:08:58 Kanglin: great, thanks a lot. hope you have a good discussion. 09:09:18 kubi001: thanks : ) 09:09:23 Reminder , Milestone6 should be Feb. 17 09:09:39 i can also join the meeting on Wednesday 09:09:50 we should implement the test case before MS6 09:10:36 so there's not much of time left.. 09:10:55 https://wiki.opnfv.org/display/SWREL/Release+Milestone+Description 09:11:50 got it. 09:12:01 Kanglin: Thanks a lot 09:12:06 Thanks all for today 09:12:11 #endmeeting