Abstract:
In this work the simulator of computer cluster and its management system which considers its topology, communication latency in data transmition and multiprocessor nodes is described. It is used for investigation of parallel jobs scheduling algorithms on computer cluster. The simulating scheme and model of cluster, its computational load model were developed, job scheduling algorithms were presented, and also the system of criteria and metrics for their comparison is described.
Keywords:a simulator of computer cluster and its management system, jobs scheduling, high performance computings, a simulation modeling.