A simulation-based performance analysis of gang scheduling in a distributed system

In distributed systems job scheduling is a difficult problem. In this work we study a special type of scheduling called gang scheduling under which jobs consist of a number of interacting tasks which are scheduled to run simultaneously on distinct processors. The performance of various gang scheduling schemes is studied and compared for a variety of workloads. The main objective of the processor schedulers is to achieve high overall system throughput while at the same time providing some guarantee for the performance of individual jobs in the system.