Dynamic load balancing using buffer management in distributed database environment

Abstract

Distributed real time database processing is still one of the most typical problems in high data computing. Complex systems have to execute millions of data accesses in each and every hours with hundreds of thousands process to executes some tens of millions of data information. This paper has focus on process with a large number of processing instances, or servers, respectively. In this paper, we proposed a dynamic Load balancing algorithm used in distributed database environment. Our basic thinking is to provide a considerable request response time and transaction per unit time instead of distributing the all workload among all servers equally. Next, new algorithm is designed and evaluated using analytical models.