A topology-aware load balancing algorithm for P2P systems

Abstract

One of the challenges of P2P systems is to perform load balancing efficiently. A distributed hash table (DHT) abstraction, heterogeneous nodes, and non uniform distribution of objects cause load imbalance in structured P2P overlay networks. Several solutions are suggested to solve this problem but they have some restrictions. They assume the homogeneous capabilities of nodes, unawareness of the link latency during transferring load and imposing logical structures to collect and reassign load. This paper presents a distributed load balancing algorithm with topology awareness using the concept of virtual servers. In our proposed approach, each node collects neighborhood load information from physically close nodes and reassigns virtual servers to overlay nodes according to topology of underlying network. Consequently, it provides rapid convergence on load balance and reduces the load transfer cost. Moreover, our parametric algorithm increases the quality of load balancing among close nodes of overlay and also provides a new tradeoff between the quality of load balancing and load transfer cost among all overlay nodes. Our simulations show that our approach reduces the load transfer cost and saves a great network bandwidth.