File Partitioning and Record Placement in Attribute-Based File Organizations.

Abstract

The position occupied by a record of the data base on secondary storage can affect performance in a variety of ways. Record placement and file organization interact with one another. A model was developed for certain techniques that use the physical device characteristics and the logical file content to optimize retrieval efficiency and precision. Methods were reviewed for partitioning the file into disjoint groups of records(called clusters) such that in most cases an access to the file will involve a small number of clusters. A record placement technique that preserves these clusters was developed, followed by a search algorithm which, when this record placement policy is followed, gives improved performance in the areas of precision and efficiency. Finally, application of the techniques to existing systems was considered.

Document Details

Document Type
Technical Report
Publication Date
Jul 10, 1975
Accession Number
ADA012937

Entities

People

  • Edwin J. Mccauley
  • Frank A. Manola

Organizations

  • United States Naval Research Laboratory

Tags

DTIC Thesaurus Topics

  • Algorithms
  • Databases
  • Efficiency
  • Precision

Readers

  • Computer Science.
  • Parallel and Distributed Computing.
  • Systems Analysis and Design