Performance Evaluation of Catalog Management Schemes in Distributed Database Systems

Abstract

Catalog management schemes may affect the site autonomy, query optimization, view management and data distribution transparency. However, the performance comparison of various catalog architectures has received relatively little attention. We employ the simulation models to investigate the relative performance of six catalog management schemes-a centralized catalog, two variations of fully replicated catalogs and three variations of partitioned catalogs-in a locally distributed database system and a geographically distributed database system. We show that three variations of partitioned catalogs perform better than the centralized catalogs and fully replicated catalogs over the wide range. The performance of centralized catalogs and fully replicated catalogs with quorum consensus are the worst because of the queuing delays in several queues. Our simulation results also indicate that the performance difference among the variations of partitioned catalogs is mainly due to the recompilation rate.