Programmable Ethernet Switches and Their Applications

TL;DR

This paper explores how programmable Ethernet switches can be dynamically controlled to optimize network performance, reliability, and QoS, enabling advanced applications in metropolitan, storage, and cluster networks.

Contribution

It introduces a novel approach to utilize Ethernet switch features as dynamic control mechanisms for enhanced network management and performance.

Findings

Achieved multi-fold throughput gains in MEN architecture

Demonstrated sub-second failure recovery time

Enabled advanced network control features

Abstract

Modern Ethernet switches support many advanced features beyond route learning and packet forwarding such as VLAN tagging, IGMP snooping, rate limiting, and status monitoring, which can be controlled through a programmatic interface. Traditionally, these features are mostly used to statically configure a network. This paper proposes to apply them as dynamic control mechanisms to maximize physical network link resources, to minimize failure recovery time, to enforce QoS requirements, and to support link-layer multicast without broadcasting. With these advanced programmable control mechanisms, standard Ethernet switches can be used as effective building blocks for metropolitan-area Ethernet networks (MEN), storage-area networks (SAN), and computation cluster interconnects. We demonstrate the usefulness of this new level of control over Ethernet switches with a MEN architecture that features multi-fold throughput gains and sub-second failure recovery time.