SDN Controlled Local Re-routing to Reduce Congestion in Cloud Data Center

Abstract

Data center networks require densely interconnected topologies to provide high bandwidth for various cloud computing services. It is required to fully utilize the bandwidth resource in such networks with varying traffic patterns. Conventional tree network topologies and routing mechanisms with single shortest path routing will cause congestion along oversubscribed links. Thus, path selection in a load-balanced way is needed to alleviate congestion and improve application performance. Multi-path routing algorithms can distribute traffic over diverse paths optimally than simple solutions like ECMP. They, however, require considerable computations, i.e. Frequently updating the path cost tree and dynamically optimizing path selections, thus they do not adapt well for large scale data center networks. In this paper, we propose a local rerouting mechanism in software defined networking (SDN) based Data Center networks to effectively manage congestion in the event of link congestion or failure. Unlike the works that deal with congestion by notifying the source which then reacts by rate adjustment or rerouting flows from source, our rerouting approach will re-forward flows (at the point of congestion or one hop before) to other possible paths based on our flow classification scheme. We demonstrate the effectiveness of our proposal through the simulation results on various topologies.