=============================== #opnfv-vswitchperf: vswitchperf =============================== Meeting started by SRao at 15:03:25 UTC. The full logs are available at http://ircbot.wl.linuxfoundation.org/meetings/opnfv-vswitchperf/2020/vswitchperf/opnfv-vswitchperf-vswitchperf.2020-09-09-15.03.log.html . Meeting summary --------------- * Jerma Release Planning Actions. (SRao, 15:03:45) * Release planning Wiki Page is Ready. Set of Issues related to Jerma is Created under VSPERF. RelReq 6 and 9 are also created. (SRao, 15:09:25) * Unable to link the issues created under VSPERF (for jerma) to The RelReq-6 and RelReq-9. (SRao, 15:09:59) * ACTION: Send an email to David - explaining the completed activities, and asking for ToDoS. (SRao, 15:20:01) * Openstack-VSPERF (SRao, 15:23:09) * Current Status: We can deploy Traffic Generators in different topologies. WIP: VSPERF Config File Creations. Issue: Instance Public-Network reachability. (SRao, 15:24:14) * vSwitch in Containers - PB. (SRao, 15:25:41) * We don't have any performance numbers on running vSwitch in a container. (SRao, 15:26:03) * In CNTT with Airship Deployments, on all compute Nodes, with vswitch is running in a container. (SRao, 15:26:27) * Do we need to do any performance testing for this usecase? Bob: Consider NUMA assignment, PMD Assignments (SRao, 15:34:38) * Normal-Mode: Isolate CPUs used for VSWITCH (2), PMDs (3-6), VNFs. We should be able to replicate the same behavior. (SRao, 15:36:04) * Can we run container and force with CPU allocation to it? Possibly YES. (SRao, 15:36:36) * - How it affects the automation - Setting up OVS with all physical and virtual ports, running VNF, configuring VNF and running the test. (SRao, 15:38:06) * CNB-K8S (SRao, 15:44:47) * Relook - B2B Results. Update OVS Throughput with New Results. (SRao, 16:04:15) Meeting ended at 16:04:20 UTC. People present (lines said) --------------------------- * SRao (17) * collab-meetbot (3) Generated by `MeetBot`_ 0.1.4