An Efficient Periodic-Request-Grouping Technique for Reduced Seek Time in Disk Array-based Video-on-Demand Server

摘要

Disk throughput is significantly affected by scheduling of disk I/O requests. In the case of VoD (Video-on-Demand) servers, disk throughput is directly related to the number of user requests that can be served simultaneously. In this paper, we propose an efficient periodic request grouping scheme for disk array-based VoD servers to improve the disk throughput. To reduce the disk seek time that are needed in processing the periodic I/O requests, this scheme groups the periodic requests that access adjacent regions into one, and arranges the groups in a pre-determined order. We show that by using this scheme, we can reduce the average disk bandwidth required by a stream and serve more periodic requests simultaneously than existing schemes. We also propose an adaptation technique that conforms the proposed scheme to access pattern changes of user requests. We performed simulation studies to evaluate the performance of the proposed scheme.