Algorithm of Performance Prediction by Resource Sharing in Distributed Database
Resource allocation is one of the main issues in solving database applications where resources and data fragments are distributed geographically. The query of each use case assigned into resources to solve parallel computing problems and avoid remote data access. Hence system resources have to be allocated to handle workload and minimize the cost of computing and maximize the utility of resources. In this paper, it is propose an algorithm for optimal allocation strategy that minimizes the cost of computation by predict the performance. The overall goal is to minimize the cost of allocated resources usage in distributed database system during early stages. The authors propose game theoretic approach for finding the optimum allocation strategy which determines the performance during the early stages of life cycle.