Query optimization using clustering and Genetic Algorithm for Distributed Databases

Abstract

Query Optimization is principally a multifaceted exploration job that searches for best plan amongst the semantically equal plans that are obtained from any given query. The execution of any processing datasets essentially depends on the capability of query optimization procedure to acquire competent query processing approaches. A Distributed Database System (DDS) is a group of autonomous cooperating integrated procedure. Query at a specified place may necessitate information from distant places in a Distributed Environment. In query optimization, the cost is accompanied by every query execution plan. Cost is the summation of native cost that is I/O cost, CPU cost at every location and cost of transmitting information amongst locations. The key issue of a Query Optimization in a Distributed Database System is to obtain an effective query strategy with an efficient accuracy and minimum response time or cost to execute the given query. In this paper a novel methodology is suggested that selected the best query plan as to execute the given query employing Genetic Algorithm Strategy for Distributed Databases and a Clustering Approach within the databases so as to execute the query plan. Genetic Algorithms are extensively employed and acceptable methods for very challenging optimization problems. This proposed technique gives efficient performance in different environment. The Experimental analysis of the proposed methodology is carried out on 100 different queries distributed over 20 different sites having 8 relations in each query. This is compared with the DB2 distributed optimizers and achieved an increased reliability and high performance with respect to the optimization cost and accuracy for the queries in the distributed databases