An Optimal Task Scheduling for Cluster Systems Using Task Duplication

This paper addresses the problem of scheduling a parallel program represented by a directed acyclic task graph onto homogeneous cluster systems to minimize its execution time. Many sub-optimal algorithms have been proposed that promises acceptable running time but not optimal solutions. We propose an algorithm that guarantees optimal solutions but not acceptable running time. Experiments show that in many practical cases the proposed algorithm performs very effectively in comparison with sub-optimal algorithms while still getting the optimal solutions.