Indian Journal of Science and Technology

Abstract

The key idea of this paper is to present a novel method for Re-fragmentation in a Distributed Database System that ensures that the Database System is adaptive to the changes in access frequency patterns. Conventionally, fragmentation is done to reduce network transfer cost and communication costs. But with time, changing access patterns at the sites defies the very purpose of fragmentation. In this paper, we have presented an algorithm to maximize and maintain the efficiency of fragmentation in a Distributed Database System. This algorithm takes into account the change in the access patterns and queries; evaluates and quantifies them and performs re-fragmentation periodically according to the updated access frequency patterns. This helps in sustaining the efficiency of fragmentation irrespective of the changing access patterns. The technique used for studying the efficiency of the new algorithm was to employ the new algorithm on a sample Distributed Database system and compare the same with the existing conventional algorithm. The performance of a Distributed Database system is inversely proportional to its communication cost and execution time of queries1,2. On analysis of the presented algorithm, we see that the time complexity of this algorithm is linear. Also the total Network Transfer Cost is considerably lesser than the conventional fragmentation algorithm for Distributed Database. The algorithm presented in the paper is applicable to all Distributed Database systems, which have changing access frequency patterns.

Keyword: Access Frequency, LAF Matrix, NLAF Matrix, Network Transfer Cost, Optimization Re-Fragmentation