STORAGE ALLOCATION FOR ACCESS PATH MINIMIZATION IN NETWORK STRUCTURED DATA BASES.

Johann P. Malmquist, Ehud Gudes, Edward L. Robertson

Research output: Contribution to journalConference articlepeer-review

1 Scopus citations

Abstract

One problem in implementing a network (CODASYL) structured data base is the allocation of record occurrences to physical storage blocks. This study presents a heuristic algorithm for performing this allocation for a stable network structured data base, in order to minimize the number of interblock references. A network data base is mapped into a weighted directed graph such that record occurrences are mapped into nodes and set occurrences are mapped into edges. Nodes are partitioned attempting to minimize the total weight of those edges which connect nodes in distinct partition classes. The optimality criteria for this algorithm is the total weight of connecting edges (between distinct partition classes). The constraints are the given size of the partition classes (i. e. physical block sizes) and a packing factor of at least 1/2. The efficiency of the algorithm is considered, and the cost is shown to be quite affordable for a stable data base.

Original languageEnglish
Pages (from-to)319-340
Number of pages22
JournalProceedings of The Western Snow Conference
StatePublished - 1 Jan 1978
EventProc of the Int Conf on Databases, Improving Usability and Responsiveness - Haifa, Isr
Duration: 2 Aug 19783 Aug 1978

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Transportation
  • Atmospheric Science

Fingerprint

Dive into the research topics of 'STORAGE ALLOCATION FOR ACCESS PATH MINIMIZATION IN NETWORK STRUCTURED DATA BASES.'. Together they form a unique fingerprint.

Cite this