Balancing Workload and Recovery Load on Distributed Fault-Tolerant VOD Systems

Abstract

This letter presents a new code sequence for dispatching playback jobs among distributed video-on-demand servers. The dispatch accordingly has the property of evenly distributing workload among all servers and balancing recovery load on surviving servers if one of the servers fails. In this letter we give an O((n - 1)!) algorithm for efficiently finding a dispatch sequence for n servers.