Improving Disk Performance Via Latency Reduction

Due to the great disparity between the access time of main memory and the average response time of disk storage, there has always been a strong need to improve the performance of disks. Even for large computing complexes where system queueing delays may overshadow the disk service time, the disk’s basic service time is the gating factor. This paper illustrates the fact that in many of today’s environments the rotational latency and the RPS miss delay are the major contributors to a disk’s basic service time. Then, by doing a sensitivity study using a simple analytical queueing model, this paper demonstrates that a reduction in these two components (both of which are related to the rotation of disk drives) has the greatest impact in reducing the disk’s basic service time and in turn produces the greatest improvement in overall subsystem performance. While the most straightforward way to reduce latency and RPS miss penalty would be to increase the disk’s rotation speed, this paper explains some of the limitations to such an approach. Several alternatives to reducing latency and RPS miss penalty are proposed and explored here and their performance is analyzed using analytical queueing models. © 1991 IEEE