H. Karatza: a Comparison of Load Sharing & Job Scheduling in a Network of Workstations

This paper examines load sharing and job scheduling in a network of workstations (NOW). Along with traditional methods of load sharing and job scheduling, it also examines methods referred to as epoch load sharing and epoch scheduling respectively. Epoch load sharing evenly distributes the load among workstations with job migration that occurs only at the end of predefined intervals. The time interval between successive load sharing events is called an epoch. The objective is to reduce the number of times that global system information is needed to make allocation decisions, while at the same time achieving good overall performance and fairness of individual job service. In the epoch scheduling case, processor queues are rearranged only at the end of epochs. The objective is to find an epoch that minimizes the number of queue rearrangements, while being fair and producing good system performance. Simulation is used to measure performance issues associated with load sharing and job scheduling.