A DYNAMIC PROCESSOR ALLOCATION POLICY FOR MULTIPROGRAMMED SHARED-MEMORY MULTIPROCESSORS

We propose and evaluate empirically the performance of a dynamic proce ssor-scheduling policy for multiprogrammed shared-memory multiprocesso rs. The policy is dynamic in that it reallocates processors from one p arallel job to another based on the currently realized parallelism of those jobs. The policy is suitable for implementation in production sy stems in that: It interacts well with very efficient user-level thread packages, leaving to them many low-level thread operations that do no t require kernel intervention. It deals with thread blocking due to us er I/O and page faults. It ensures fairness in delivering resources to jobs. Its performance, measured in terms of average job response time , is superior to that of previously proposed schedulers, including tho se implemented in existing systems. It provides good performance to ve ry short, sequential (e.g., interactive) requests. We have evaluated o ur scheduler and compared it to alternatives using a set of prototype implementations running on a Sequent Symmetry multiprocessor. Using a number of parallel applications with distinct qualitative behaviors, w e have both evaluated the policies according to the major criterion of overall performance and examined a number of more general policy issu es, including the advantage of ''space sharing'' over ''time sharing'' the processors of a multiprocessor, and the importance of cooperation between the kernel and the application in reallocating processors bet ween jobs. We have also compared the policies according to other crite ria important in real implementations, in particular, fairness and res ponse time to short, sequential requests. We conclude that a combinati on of performance and implementation considerations makes a compelling case for our dynamic scheduling policy.