16:00:44 #startmeeting vswitchperf 16:00:44 Meeting started Wed Nov 27 16:00:44 2019 UTC. The chair is SRao. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:44 Useful Commands: #action #agreed #help #info #idea #link #topic. 16:00:44 The meeting name has been set to 'vswitchperf' 16:00:59 #topic Development Update. 16:04:42 #info: VSPERF Update patch is merged. Now, we have VSPERF with latest version of OVS, VPP, DPDK and Qemu. 16:05:21 #info: Received reviews from Shailesh and AL for VSPERF containers. 16:06:46 #action Add Alex and Ancuta as reviewers. 16:09:22 #info Created version-1 daemonset for the necessary setup 16:09:32 #topic CN Benchmarking 16:09:43 #info Created version-1 daemonset for the necessary setup 16:09:53 #info Working on the automation part. 16:15:43 #action Alex to update wiki with daemonset progression - single pod, dual-interface, 2-pods with each dual-interface, single pod with 3-interfaces. 16:16:16 #info https://wiki.opnfv.org/display/vsperf/VSPERF-Container+Networking+Benchmarking+and+Testing 16:17:47 #info vsperf-br0 16:18:56 #action Sridhar to send alex the steps to run vsperf with "OVS" running without TGEN. 16:28:11 #topic Ixia Live-Results 16:33:20 #info Ancuta shared the current statistics generated - Port level and the traffic level. 16:35:03 #info Ancuta submit the changes 16:35:19 #action Sridhar to share with ancuta the path to write the files. 16:37:43 #topic Kubecon update 16:37:55 #info Attended sessions Kubernetes Networking. 16:38:24 #info Multiple interfaces is still an open problem - there are not of issues yet to be addressed. 16:39:14 #info Attempts for multiple interfaces: celium and Calico. Get it working with celium is very difficult, even the authors are not sure about the approach. 16:39:41 #info Calico: Its not clear if single instance of calico is sufficient. 16:40:26 #info Most the networking solutions used in production are for single interfaces, and very few with dual-interface. 16:40:45 #info No Demo on two data + 1 Mgmt interfaces. 16:41:16 #info Most of the demos are targeting enterprises and very few for telcos. 16:41:48 #action Create/Update this information in Wiki. 16:42:40 #topic Container Metrics 16:43:35 #info Parth is working on Prometheus based solution 16:43:44 #info We need to collect all the related metrics. 16:44:06 #info Yes. It is necessary (else there will be no receiver). 16:44:28 #info But, you can write to files while you are testing. 16:45:25 #info Update Results container with (a) Prometheus (b) Grafana dashboard for container metrics. 16:45:46 #info (c) Connect Grafana with prometheus data. 16:47:02 #endmeeting