Performance Evaluation of a Load Sharing System on a Cluster of Workstations

摘要

This paper presents a methodology to evaluate the performance of load allocating algorithms using queuing network facilities. First, we develop a generic model of load sharing systems. Then, a specific model is derived from the generic model that simulates a real load sharing system. This phase consists of successive manipulations of algorithms parameters until reaching satisfied approximations. Once the suitable parameters have been tuned, which configure the specific model, it is easy to evaluate the performance of the load allocating algorithms for various applications and hardware configurations.