16:59:42 #startmeeting FDS synch 16:59:42 Meeting started Thu Dec 8 16:59:42 2016 UTC. The chair is frankbrockners. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:59:42 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:59:42 The meeting name has been set to 'fds_synch' 16:59:50 #info Frank Brockners 17:00:21 #info Juraj Linkeš 17:00:22 #info Michal Cmarada 17:00:29 #info Sean Chandler 17:01:08 #info Vlado Lavor 17:02:25 #info Agenda for today: (a) update HA deployments (Juraj) (b) update on tap port (re-)connect issue testing (Michal) (c) update on GBP security groups adaptation to new VPP/HC API (Tomas/Michal) (d) update in L3 implementation (Michal) (e) update on ODL clustering issues (Vlado) 17:02:37 anything else to discuss? 17:03:09 we can always add things as they pop up ... 17:03:42 let's start with (a) - update on HA testing. jlinkes could you fill us in? 17:04:22 #info the issues with services not being synced up properly were resolved by configuring a ntp server 17:04:57 jlinkes: excellent news. 17:04:59 #info we are able to run functest, although the stability of the whole setup is not 100% 17:06:15 jlinkes: what is the current functest success rate - what is failing? 17:07:12 frankbrockners: sometimes functest cannot connect to the public openstack endpoint 17:08:14 frankbrockners: haven't gotten further 17:08:33 jlinkes: thanks - but looks like you made really good progress 17:08:44 right 17:08:58 let's move to (b) tap port reconnect - michal-cmarada? 17:09:22 tap port for Qrouter was tested and is now merged 17:09:37 I did some more testing 17:09:39 today 17:09:44 the issue with qr tap disconnecting was solved 17:09:45 #info tap port for Qrouter was tested and is now merged 17:09:56 I tried newer vpp - the crashes went away 17:10:14 #info the issue with qr tap disconnecting was solved, latest vpp shows no crashes anymore 17:10:27 but there is still some strange behavior - sometimes gpb deletes the qrouter tap port for no discenable reason 17:10:31 michal-cmarada, jlinkes: thanks - again great news 17:10:43 ouch 17:10:46 that is a separate issue from recreation 17:11:04 jlinkes: could you create a jira ticket for that issue? 17:11:38 sure, I'll look into it on Monday in more detail - it could be that it is a manifestation of a different issue 17:11:50 jlinkes: ok - thanks 17:11:58 and we also saw that there is some weirdness in how port are removed from bds 17:12:03 ports* 17:12:30 michal-cmarada had a good idea of asking Tyler for detailed description of how *actually* vbd works 17:12:47 sounds good 17:13:00 michal-cmarada: could you follow up with tyler over email? 17:13:11 yes, sure 17:13:17 thanks michal-cmarada 17:13:49 let's jump to (c) security groups: Update GBP implementation to new VPP/HC API 17:13:56 I believe Tomas is not here 17:14:01 no 17:14:05 michal-cmarada: do you know the status? 17:14:09 but he sent me some notes 17:14:55 today he was working on reviewing the changes for security groups. 17:15:52 but he also worked on optimalization of RPC for VPP api, because he found some errors 17:16:08 michal-cmarada: thanks 17:16:18 no problem 17:16:46 and on (d) Updates on the L3 implementation - michal-cmarada - that is again you :-) 17:16:55 yes, 17:17:48 this is still in progress, I had some problems configuring routing through HC, so I tried different versions of VPP/HC 17:18:30 Then I got some problems while running my enviro so I wasnt able to test the newest versions 17:18:48 if we need fixes in VPP we'd need to surface them fairly soon.. 17.01 API freeze is already done (was yesterday) - code freeze is coming up... 17:19:07 After these versions are tested and if the issue is still present i will create a jira ticket for this 17:19:22 thanks michal-cmarada 17:19:34 FYI - 17.01 release plan is found here: https://wiki.fd.io/view/Projects/vpp/Release_Plans/Release_Plan_17.01 17:19:58 let's move to the final agenda topic: (e) ODL clustering 17:20:17 vlavor: any progress on getting the netconf issues you saw fixed? 17:21:01 from netconf side, there is no progress, they are still working on it 17:21:46 but I was able to workaround it in GBP/VBD code 17:22:03 vlavor: great news 17:22:12 and it looks that it works very well 17:22:42 so I think this shouldn't bother us anymore 17:22:55 vlavor: VERY GOOD 17:23:26 vlavor: do you have the patch handy with your workaround? 17:23:38 vlavor: that way we can document it for the record 17:23:59 frankbrockners: I think this is it https://git.opendaylight.org/gerrit/#/c/48983/ 17:24:00 patch for VBD is already merged, GBP one is ready for review 17:24:28 yes, and this one is for vbd https://git.opendaylight.org/gerrit/#/c/48976/ 17:24:48 jlinkes, vlavor: thanks 17:24:59 that means we're done with the agenda 17:25:04 anything else to discuss? 17:25:42 ... does not sound like this is the case. 17:25:47 is the nosdn scenario done? 17:25:53 finished 17:26:06 jlinkes: it is released with a set of documented issues 17:26:18 frankbrockners: ah okay 17:27:00 but we have seanatcisco here - seanatcisco: any updates on os-nosdn-fdio-noha? 17:27:32 continued work to resolve the known issues. nothing really new to report. 17:27:42 thanks seanatcisco 17:27:43 some additional documentation work 17:27:58 any other business to discuss? 17:28:00 also Onong is creating mirror launchpad issues for those specifically relating to ML2 17:28:24 * frankbrockners is trying to finish up the meeting one more time 17:29:11 ok folks - thanks everyone - and have a great day/evening! 17:29:14 #endmeeting