Skip to Main Content
The work described in this paper suggests that throughput gains are possible for some work loads if dispatching is controlled by a self-adjusting algorithm thatakes into accounthe changing characteristics of tasks both singly and as a group. Minimum benefits can be anticipated when the characteristics of the tasks are homogeneous; if all tasks are heavily Iporiented or all tasks are heavily CPU-oriented, no throughput improvement can be anticipated. In fact, additional overhead will be incurred handling time intervals and manipulating task queues. Under these conditions, the only benefit is the cyclic dispatching of all CPU-oriented tasks, which prevents any single task from monopolizing use of the CPU. Maximum benefits can be anticipated for a mixture of heavily CPU-oriented and heavily Iporiented tasks. It is under these conditions that the algorithm can strive for maximum CPU-I/O overlap.
Note: The Institute of Electrical and Electronics Engineers, Incorporated is distributing this Article with permission of the International Business Machines Corporation (IBM) who is the exclusive owner. The recipient of this Article may not assign, sublicense, lease, rent or otherwise transfer, reproduce, prepare derivative works, publicly display or perform, or distribute the Article.