Query optimization in distributed databases through load balancing

Abstract

Abstract : As technology advances, computing environments composed of large numbers of workstations and mainframes connected by a high bandwidth local area network become attractive. These systems typically support a large number of application classes, and the heterogeneity of their workload, coupled with the decentralization of the systems, can lead to load imbalances across the network. This work attempts to study the benefits of load balancing strategies in the context of a particular application, distributed database systems. It was felt that, by focusing on a specific area, the problem would become more tractable. The choice of database management systems can be justified not only by their intrinsic importance, but also by the adaptability of load balancing strategies to query optimization algorithms. In order to determine whether load balancing strategies can indeed be adapted to current optimizers with a moderate amount of effort and to see whether the resulting performance benefits are sizable, both benchmarking and simulation experiments were carried out.