A Java-Based Model for I/O Scheduling in Tertiary Storage Subsystems

Abstract

Modern I/O subsystems include large storage servers that are configured to include multiple on-line and off-line storage media and to deal with a large number of requests with unpredictable access patterns. The problem of minimizing the cost of accessing data stored in all media is critical to the performance of the system. Given the large storage requirements of modern applications, tertiary storage subsystems have become a crucial component of modern large-scale storage servers. This paper’s contribution refers to the performance improvement of tertiary storage systems by the use of specific scheduling algorithms implemented under Java programming environment. The paper presents a general model to be applied to different tertiary storage topologies under various computer platforms. Magnetic tapes and tape libraries models are examined by the Java-based simulation model in order to increase functionality and interactivity. The Java-based model is validated and experimented by three most widely used scheduling algorithms, SORT, SLF, and SCAN policies. Comparative analysis of the scheduling policies and discussion of the most beneficial scheduling policy for each model are provided.