Application of GSE Methodology in Distributed Computing

Abstract

Distributed computing enables organizations to fully utilize their existing heterogeneous computing resources to meet their heavy-duty computing needs without buying expensive supercomputers and servers. When developing applications it is therefore useful to design systems that can be effectively partitioned and distributed across the existing heterogeneous networks. Such applications need to posses the necessary qualities of a distributed system such as high cohesiveness and loose coupling. Three important aspects that should be considered in producing distributed systems are, partitioning, interfacing and granularity. This paper describes the application of a novel design method called the Genetic Software Engineering (GSE) methodology that naturally enables these features to be designed into the system.