Performance Issues in Distributed Query Processing

  • Authors:
  • C. Liu;C. Yu

  • Affiliations:
  • -;-

  • Venue:
  • IEEE Transactions on Parallel and Distributed Systems
  • Year:
  • 1993

Quantified Score

Hi-index 0.00

Visualization

Abstract

The authors discuss various performance issues in distributed query processing. Theyvalidate and evaluate the performance of the local reduction (LR) the fragment andreplicate strategy (FRS) and the partition and replicate strategy (PRS) optimizationalgorithms. The experimental results reveal that the choices made by these algorithmsconcerning which local operations should be performed, which relation should remainfragmented or which relation should be partitioned are valid. It is shown usingexperimental results that various parameters, such as the number of processing sites,partitioning speed relative to join speed, and sizes of the join relations, affect theperformance of PRS significantly. It is also shown that the response times of queryexecution are affected significantly by the degree of site autonomy, interferences amongprocesses, interface with the local database management systems (DBMSs) andcommunications facilities. Pipeline strategies for processing queries in an environmentwhere relations are fragmented are studied.