Overview
AI, Data Science & Cloud Certificates from Google, IBM & Meta — 40% Off
One plan covers every Professional Certificate on Coursera. 40% off Coursera Plus Annual.
Unlock All Certificates
Learn how to effectively debug hardware offload processing in a 25-minute technical talk that explores debugging techniques for applications using rte_flow API. Discover methods to overcome limited visibility challenges in both partial and full hardware offloads, where packets are either preprocessed before CPU handling or processed entirely through internal switching. Master the use of existing rte_flow API concepts including MARK, META, COUNT, and group miss actions to enhance debugging capabilities in complex offloading pipelines. Gain practical insights into troubleshooting performance issues that arise from offloaded pipeline design, particularly when dealing with multiple flow groups and varied packet routing paths through the flow pipeline.
Syllabus
Debug Functional and Performance Issues in Rte_flow - Dariusz Sosnowski, NVIDIA Corp.
Taught by
DPDK Project