Design and comparative analysis of DSS queries in distributed environment

摘要

Query processing and its optimization is one of the major aspects of distributed database system. The research has exposed that the design of a query and its execution technique plays an important role in the optimization of a query. There are two major categories of distributed queries in distributed database system known as Decision Support system (DSS) queries and Online Transaction Processing (OLTP) queries. In this paper the prime focus is on design and analysis of DSS queries. The selected set of DSS queries are simulated by using exhaustive enumerative technique and genetic approach under serial and parallel processing environment. The simulation results show that an exhaustive enumeration approach provides optimized solution but takes huge time for complex DSS queries (Hours, Days, Month or even Years), hence it is infeasible to implement this approach for optimizing a set of DSS queries. On the other hand genetic algorithms optimize DSS Queries very quickly but show loss in accuracy and quality of solution as compare to exhaustive enumerative approach. Further the parallel execution of the different sub operations of a DSS query significantly reduces the total cost of system resources.