Fintech workloads and few other modern work loads require very fine grained control (down to a couple of milliseconds) over packet pacing and at the same time be able to do this for millions of flows; all requiring different granularity of flow pacing.

In this talk, Anjali S Jain, Vadivel Kannappa, Joshua Hay, Robert Hathway, Magnus Karlsson, Chihjen Chang, Naren Mididaddi, Priyalee Kushwaha and Milena Olech go into implementation details of how to achieve these requirements.

Details: https://netdevconf.info/0x19/32

cheers, jamal