A task cluster size determination in heterogeneous distributed systems

There are many task scheduling algorithms for minimizing the schedule length in distributed systems. Especially, task clustering heuristics are well known as an approach for deciding the number of required processors. However, the actual number may become very large depending on characteristics of the input job and each processor´s capability. Thus, currently there is no theoretical criterion for achieving processor utilization. In this paper, we propose a method for deriving the lower bound of the cluster (assignment unit) size in heterogeneous distributed systems for limiting the number of processors. From results of experimental simulations, we discuss the applicability of the proposal to obtain better processor utilization.