Organization of clustered files for consecutive retrieval

Abstract

This paper studies the problem of storing single-level and multilevel clustered files. Necessary and sufficient conditions for a single-level clustered file to have the consecutive retrieval property (CRP) are developed. A linear time algorithm to test the CRP for a given clustered file and to identify the proper arrangement of objects, if CRP exists, is presented. For the single-level clustered files that do not have CRP, it is shown that the problem of identifying a storage organization with minimum redundancy is NP-complete. Consequently, an efficient heuristic algorithm to generate a good storage organization for such files is developed. Furthermore, it is shown that, for certain types of multilevel clustered files, there exists a storage organization such that the objects in each cluster, for all clusters in each level of the clustering, appear in consecutive locations.