Scalable Fine-grained Path Control in Software Defined Networks

Abstract

The OpenFlow-based SDN is widely studied to better network performance through planning fine-grained paths. However, being designed to configure path hop-by-hop, it faces the scalability issue that both the flow table overhead and path setup delay are unacceptable for large-scale networks. In this paper, we propose PACO, a framework based on Source Routing to address that problem through quickly pushing paths into the packet header at network edges and pre-installing few rules at the network core. The straightforward implementation of SR is inefficient as it would incur too many path labels; other efficient approaches would sacrifice path flexibility (e.g., DEFO). To implement SR efficiently and flexibly, PACO presents each path as a concatenation of pathlets and introduces algorithms to compute pathlets and concatenate paths with minimum path labels. Our extensive simulations confirm the scalability of PACO as it saves the flow table overhead up to 94% compared with OpenFlow-SDN solutions and show that PACO outperforms SR-SDN solutions by supporting more than 40% paths with few label overhead.