AN ANALYSIS OF SCHEDULES FOR PERFORMING MULTI-PAGE REQUESTS

In this paper, we address the problem of efficiently reading a set of pages from a file which is kept on some cylinders of a magnetic disk. It is assumed that several pages can be read from disk in a single mul ti-page request without interruption by other requests. First, a simpl e algorithm is presented for computing a schedule how the required pag es are read from disk. Then, the expected cost of the schedules is ana lyzed by using a pure analytical model. In addition to the disk geomet ry, the analysis takes into account the three major cost components (s eek time, rotational delay and transfer time) which occur when pages a re retrieved from magnetic disk. The derived cost function depends on two parameters: the number of required pages and the degree of cluster ing of the underlying file. The cost function demonstrates that signif icant performance improvements can be achieved by using multi-page req uests when the required pages are read according to a well-computed sc hedule. In addition to the response time of a query exclusively perfor med in the system, we also examine the impact of multi-page requests o n the average response time when multiple queries are concurrently per formed at a time. Results of an experimental performance comparison de monstrate that the average response time can considerably be improved on a highly loaded system by several factors when multi-page requests are used in comparison to reading the required pages one at a time. Co pyright (C) 1996 Elsevier Science Ltd