pFabric

TLDR

pFabric is a minimalistic datacenter transport design that achieves near‑theoretically optimal flow completion times for short flows while reducing average completion time for long flows. pFabric decouples flow scheduling from rate control, using per‑flow priority numbers and small buffers for priority‑based scheduling, and starts at line rate, throttling only under persistent loss. Simulations show that pFabric delivers near‑optimal flow completion times at the 99th percentile for short flows and improves average completion time for long flows, validating its simple design.

Abstract

In this paper we present pFabric, a minimalistic datacenter transport design that provides near theoretically optimal flow completion times even at the 99th percentile for short flows, while still minimizing average flow completion time for long flows. Moreover, pFabric delivers this performance with a very simple design that is based on a key conceptual insight: datacenter transport should decouple flow scheduling from rate control. For flow scheduling, packets carry a single priority number set independently by each flow; switches have very small buffers and implement a very simple priority-based scheduling/dropping mechanism. Rate control is also correspondingly simpler; flows start at line rate and throttle back only under high and persistent packet loss. We provide theoretical intuition and show via extensive simulations that the combination of these two simple mechanisms is sufficient to provide near-optimal performance.