Minimizing the response time of executing a join between fragmented relations in a distributed database system

  • Authors:

  • D. J. Reid

  • Affiliations:

  • Information Technology Division Defence Science and Technology Organisation P.O. Box 1500, Salisbury SA 5108, Australia

  • Venue:
  • Mathematical and Computer Modelling: An International Journal
  • Year:
  • 1997

Quantified Score

Hi-index 0.98

Visualization

Abstract

The problem of describing minimal response time execution strategies in evaluating the join of several fragmented database relations, is considered. The consequential optimization problem assumes the convenient form of a min-max integer program. With further attention, various generalizations are realized that also include the performance objective of total execution cost. Tables of data logically conforming to the relational model of information are, at the physical level, frequently divided into numerous pieces. These fragments are found disseminated amongst the various sites of a distributed database system, with each one possibly replicated at any number of separate facilities. A submission demanding the amalgamation of many such relations is resolved by joining together their sets of component fragments in an appropriate manner, as defined by complicated patterns of overlapping attribute values. The final result is realized by then concatenating the products of these computations. This process is to be performed under the supervision of the database management system in such a way as to minimize the time taken, as perceived by the user who issued the request. These developments are based upon earlier investigations [1-5] that consider only the alternative optimization goal of minimal execution cost. With this in mind, several different different approaches may be taken to realize distinct hybrid models that give due regard to both measures of join query performance.