Relaxing constraints in stateful network data plane design

TL;DR

This paper demonstrates that relaxing conservative assumptions about traffic in stateful network data plane design enables more complex functions to be implemented at line rate without performance loss.

Contribution

It challenges existing assumptions by showing that more realistic traffic models allow for less constrained, more capable stateful data plane implementations.

Findings

Relaxed design constraints enable more complex functions.

Larger average packet size improves data plane programmability.

No performance degradation observed with relaxed assumptions.

Abstract

Supporting programmable states in the data plane of a forwarding element, e.g., a switch or a NIC, has recently attracted the interest of the research community, which is now looking for the right abstraction to enable the programming of stateful network functions in hardware at line rate. We challenge the conservative assumptions of state-of-the-art abstractions in this field, e.g. always assuming minimum size packets arriving back-to-back. Using trace-based simulations we show that by making more realistic assumptions on the traffic characteristics, e.g. larger average packet size, we can relax the design constraints that currently limit the set of functions that can be implemented at line rate, allowing for more complex functions, with no harm for performance.