A virtual distributed computing system

Cost of parallel and closely coupled architectures and the complex issues involved in parallel computing have led to the alteration in the state of computing. With the advent of high performance workstations and high speed networking, the computational power of workstations have become attractive and indispensable for parallel computing. Utilizing the idle CPU cycles of a remote processor, if the jobs cannot be processed locally, is the current trend in research.;Performance analysis of the system has been done using a number of hypothetical applications and one real application (in this case matrix multiplication). Hypothetical applications provide a flexibility for tuning the environment to test VDCS. VDCS has responded positively to parallel matrix multiplication. Using load balancing the average speedup achieved for different applications was 66% of its theoretical level, on a small network of workstations, while fault tolerance has proved to provide a reliable environment.;The main objective of this study is to transform a network of workstations into a Virtual Distributed Computing System (VDCS). Such a system has two components; load balancing and fault tolerance. Load balancing in a distributed computing system improves the performance of the system to a substantial amount while improving the job response times. Defining the load metrics, determining the load on individual hosts, detecting the load imbalance and load balancing the system are the issues to be dealt in load balancing. Providing backup for each application, detecting host failures, restoring the failed applications are the issues involved in fault tolerance.