Implementing Multiprocessor Scheduling Disciplines

An important issue in multiprogrammed multiprocessor systems is the scheduling of parallel jobs. Consequently, there has been a considerable amount of analytic research in this area recently. A frequent criticism, however, is that proposed disciplines that are studied analytically are rarely ever implemented and even more rarely incorporated into commercial scheduling software. In this paper, we seek to bridge this gap by describing how at least one commercial scheduling system, namely Platform Computing's Load Sharing Facility, can be extended to support a wide variety of new scheduling disciplines. We then describe the design and implementation of a number of mul-tiprocessor scheduling disciplines, each diiering considerably in terms of the type of preemption that is assumed to be available and in terms of the exibility allowed in allocating processors. In evaluating the performance of these disciplines, we nd that preemption can signiicantly reduce overall response times, but that the performance of disciplines that must commit to allocations when a job is rst activated can be signiicantly aaected by transient loads.