Scheduling Jobs in a Heterogeneous System with Unreliable Network Links by DLT

Abstract

During the last decade, the use of parallel and distributed systems has become more common. In these systems, a huge chunk of data or computation is distributed among many systems in order to obtain better performance. Dividing data is one of the challenges in this type of systems. Divisible Load Theory (DLT) is proposed method for scheduling data distribution in parallel or distributed systems. In many researches carried out in this field, it was assumed that all processors are reliable for grid system but that is not always true in real systems. Although unreliable processors were investigated in some studies, unreliable or unreliable link connections have not been addressed before this research. Disconnection in network links is a common happening in all network topologies specially in wireless connections. In this investigation, we assume that we have two static parameters for processors in a unreliable system, probability of disconnection in network links and required time for repairing disconnection. Then, we present a closed-form formula for scheduling tasks in this type of systems. Experiments show this scheduling method has better performance than applying the proposed methods for reliable systems on an unreliable system.