15:08:43 #startmeeting bgpvpn 15:08:44 Meeting started Tue Jul 5 15:08:43 2016 UTC and is due to finish in 60 minutes. The chair is tmorin. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:08:45 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:08:49 The meeting name has been set to 'bgpvpn' 15:09:13 hi everyone, doude, michchap, pcarver, timirnich, steve_ruan 15:09:28 hi 15:10:14 wimdeclerq told me he would not attend 15:10:33 I think this will be a short meeting 15:10:42 hi 15:10:47 hi 15:10:54 #topic work in progress 15:10:59 a few things 15:11:09 the horizon tenant GUI work is in progress 15:11:24 #link https://review.openstack.org/#/c/335091/ 15:11:37 not much more to say about it 15:12:06 work is in progress as well to fix a bug happening with the opencontrail driver 15:12:22 #link https://bugs.launchpad.net/bgpvpn/+bug/1596962 15:12:22 Launchpad bug 1596962 in networking-bgpvpn "opencontrail driver issue with a particular network association" [Medium,In progress] - Assigned to Thomas Monguillon (thomas-monguillon) 15:12:30 different people have hit this bug 15:12:40 the fix is simple and I'll merge it soon 15:13:14 except that our unit test job is currently broken (let's discuss this later) 15:13:30 the VNI change is in progress (steve_ruan) 15:14:08 we still need to reflect in the API doc which additional constraints we need to put on associations to avoid conflicts between BGPVPN VNIs and VNIs used by Neutron L2/VXLAN 15:15:14 steve_ruan: I have seen your last comment, but I still need to think a bit more what we need to specify for BGPVPN VNIs and Router associations 15:15:15 yes, the constraints need more consideration 15:15:30 thanks 15:16:18 #topic open discussion 15:16:25 do we have anything more to discuss today ? 15:16:32 #undo 15:16:34 Removing item from minutes: 15:16:36 my college find another bug https://bugs.launchpad.net/bgpvpn/+bug/1599055 15:16:36 Launchpad bug 1599055 in networking-bgpvpn "tox unit test failed without testresources package" [High,In progress] - Assigned to Henry Dong (henry-dong) 15:16:37 ah yes, of course 15:16:46 our gate is broken 15:16:47 it blocked the tox 15:17:01 I had a look at is as well 15:17:12 I don't get where the issue is 15:17:49 it's very strange that someone remove "testresources>=0.2.4" 15:17:58 from test_requirement 15:18:05 where/when was it removed ? 15:18:31 need dig out 15:19:34 I would have expected testresources to be installed in the test venv because neutron is installed there, but maybe only neutron/requirements.txt deps are installed 15:21:09 identifying the change that caused the failure would be very useful 15:21:20 have you discussed with net-sfc folks about that yet ? 15:23:37 sorry, the root cause is not that someone change test_requirement.text 15:23:52 but add testresource can fix it 15:24:32 I found since March 2016, test_requirement.text is not changed 15:25:31 someone rang? 15:26:27 :) 15:26:57 cross IRC room discussion 15:27:22 the problem with missing testresources is that you need to include it in test-requirements.txt 15:27:22 steve_ruan: feel free to read #openstack-neutron logs 15:27:34 but... that's not likely going to be enough 15:27:44 regXboi: did you identify why it wasn't necessary before ? 15:28:18 tmorin: I wasn't in a search mode, I was in a fix mode 15:28:25 .. or more generally why much more is now needed 15:28:29 regXboi: ok 15:29:01 oh well *that* I know 15:29:18 first, the testresources is an update from openstack bot 15:29:27 but I suspect neutron was forking and changed 15:29:44 but as far as the OVS stuff is concerned, neutron switched to the native ovs interface 15:29:56 and looking at neutron-vpnaas, it's not prepared for that change 15:30:08 so that's where I'd look next 15:30:39 test-requirements has included testresources since January, so this is not recent, right ? 15:31:27 well, for neutron it has, neutron-vpnaas is another matter - I can't speak to networking-sfc 15:31:36 er wait I said that wrong (too many projects) 15:31:39 regXboi: understood about the ofctl/native switch, it seems to be a likely cause why importing testresources now fails 15:31:58 networking-sfc didn't have it, I can't speak to networking-bgpvpn 15:32:16 it doesn't have it either 15:33:24 then I'd add look at adding it and seeing what breaks next 15:33:31 yes 15:34:21 now i've not dug beyond just adding that 15:34:34 I wanted to know why we now had to include it explicitly, but I think I will just assume that one of the requirements or test-requirements of net-bgpvpn was declaring it as a requirement 15:36:13 somebody in #openstack-neutron might have a better idea 15:36:23 as to the chapter and verse of why 15:36:54 yep, let's follow-up there 15:37:08 http://logs.openstack.org/41/337541/1/check/gate-networking-bgpvpn-python27/21135f0/console.html#_2016-07-05_09_39_56_293009 16:00:25 hongbin: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 16:01:12 hongbin: the bgbvpn meeting has not ended 16:01:33 It looks they forgot to end the meeting 16:01:38 shardy: ping 16:01:45 can you end it yourself? 16:01:57 #endmeeting 16:02:08 No, doesn't seem to work 16:02:13 hmm 16:02:18 I think shardy started it 16:05:04 I pinged on tripleo 16:05:50 hongbin: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 16:06:05 .......... 16:06:25 tmorin ^^ the bgpvpn meeting is still running 16:06:26 we need to keep trying to end it. 16:06:30 #endmeeting 16:06:56 wops 16:06:59 #endmeeting 16:07:09 #endmeeting