14:00:44 #startmeeting Fortnightly Open vSwitch for NFV meeting (January 30th 2017) 14:00:44 Meeting started Mon Jan 30 14:00:44 2017 UTC. The chair is mdgray_. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:44 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:00:44 The meeting name has been set to 'fortnightly_open_vswitch_for_nfv_meeting__january_30th_2017_' 14:00:53 #topic Roll Call 14:00:59 #info Mark Gray 14:01:01 #info Billy O'Mahony 14:01:03 #info Tom Herbert 14:02:20 #topic Agenda Items 14:02:34 #info MS5 (Milestone 5 - Danube Release) 14:02:37 #info Chunghan Lee 14:02:47 #info Priority Data Path 14:03:13 #topic MS5 14:03:39 #info Main questions follow: 14:03:47 https://www.irccloud.com/pastebin/WRDhO5u3/ 14:04:38 #info Fuel has the following status: 14:05:09 #info Ruijing has developed a plugin 14:06:18 #info However, the preferred path at the moment is the built-in OVS-DPDK version. 14:06:33 #info The scenario has not been updated to reflect this. 14:08:43 #info The current list of scenarios for danube is defined here https://wiki.opnfv.org/display/SWREL/Danube+Scenario+Status 14:10:11 #info There is no Fuel-OVS scenario here. There is a kvm-ovs scenario 14:10:25 #info Mark to check the status of the KVM-OVS scenario 14:10:32 #info Apex 14:14:21 #info There are a number of Apex OVS scenarios owned by Tom. However, there is no associated Jenkins job. 14:28:19 #topic Priority Data Path 14:31:17 #info Billy has implemented a patch but it is showing strange results 14:31:52 #undo 14:31:52 Removing item from minutes: 14:32:22 #info Billy has implemented a patch but it was showing packet loss for packets that were prioritized. 14:33:15 #info Billy redid the tests with P2P test deployment which did not show packet loss but did not change the latency profile. 14:34:35 #undo 14:34:35 Removing item from minutes: 14:34:59 #info Billy redid the tests with P2P test deployment which did not show large packet loss but did not change the latency profile. 14:41:58 #info Further work is required. Suggestion to split hi-priority and lo-priority traffic into separate cores to see if the latency profile improves. 14:42:16 #info Link is currently saturated. Suggestion to down clock the CPU. 14:46:52 #info Any suggestions for Billy, please contact him 14:46:57 #endmeeting