A Comparative Study of Divergence Control Algorithms

Kawaguchi, Akira; Mok, Kui W.; Pu, Calton

This paper evaluates and compares the performance of two-phase locking divergence control (2PLDC) and optimistic divergence control (ODC) algorithms using a comprehensive centralized database simulation model. We examine a system with multiclass workloads in which on-line update transactions and long-duration queries progress based on epsilon serializability (ESR). Our results demonstrate that significant performance enhancements can be achieved with a non-zero tolerable inconsistency (ϵ-spec). With sufficient ϵ-spec and limited system resources, both algorithms achieve comparable performance. However, with low resource contention, ODC performs significantly better than 2PLDC. Moreover, given a small ϵ-spec, ODC returns more accurate results on the committed queries then 2PLDC.



More About This Work

Academic Units
Computer Science
Department of Computer Science, Columbia University
Columbia University Computer Science Technical Reports, CUCS-013-94
Published Here
February 2, 2012