RELIABILITY DRIVEN LOAD BALANCING FOR TASK ALLOCATION IN DISTRIBUTED SYSTEMS

Abstract

Distributed computer systems are increasingly being employed for critical applications, such as aircraft control, industrial process control and banking systems. Maximizing performance has been the conventional objective in the allocation of tasks for such systems. Inherently, distributed systems are more complex than centralized systems. The added complexity could increase the potential for s, stem failures. Some work has been done in the past in allocating task to distributed systems, considering load balancing and performance (eg,completion time) as the objective function. This, however, does not consider reliability of the system. For distributed systems, 3ve define system reliability as the probability that the system can run an entire task successfully. When the system's hardware configuration is fixed, the system reliability is mainly dependent on sofhvare design. This dissertation work deals with a method to determine an allocation that maximize the reliability of the hetrogeneous distributed system and at the same time attempt to achieve optimization by balancing load over processors. The algorithms discussed here are implemented and tested for various problems. From result it can be seen that proposed heuristics provides optimal or nearly optimal solutions in most of the cases, thus maximizing reliability and achieve load balancing both. We present a number of simulation results to prove the efficacy of our scheme.