Class Central is learner-supported. When you buy through links on our site, we may earn an affiliate commission.

YouTube

Unified Representor with Large Scale Ports for Network Management

DPDK Project via YouTube

Overview

Learn about a groundbreaking approach to managing large-scale network ports in a technical presentation from NVIDIA Semiconductor. Explore how the traditional model of using individual software representors for each VF/SF port faces scalability challenges when dealing with hundreds of VMs and thousands of Kubernetes pods. Discover a new solution that implements a unified representor system through a proxy port's Rx/Tx queue and single switch representor for handling miss traffic. Understand how this innovative approach addresses memory consumption issues, eliminates inefficient queue polling, reduces cache misses, and significantly improves initialization times. Examine the evolution from the RTE_ETH_DEV_CAPA_RXQ_SHARE capability to this more streamlined architecture that efficiently manages thousands of ports while optimizing CPU cycle usage in switchdev mode DPDK applications.

Syllabus

Unified Representor with Large Scale Ports - Suanming Mou, NVIDIA Semiconductor

Taught by

DPDK Project

Reviews

Start your review of Unified Representor with Large Scale Ports for Network Management

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.