A Novel Far Technique in Distributed Database Systems

Abstract

newlineThe corporate world finds it challenging to store data in a single location as the amount of information is growing day by day. Nowadays, businesses like Amazon, Flipkart, Facebook, and others must manage vast amounts of data. As a result, distributed database management systems are growing in popularity, and researchers are increasingly interested in implementing distributed databases. A distributed database is a group of information that, logically, is part of the same system but is dispersed over different locations on a computer network. How to distribute data to the locations connected in a network is the key challenge in using Distributed Database Systems (DDBS). Fragmentation, Allocation, and Replication (FAR) can be used to complete this time-consuming procedure, making it simple to distribute and integrate data. The process of fragmentation involves breaking up the entire database into smaller units known as fragments. There are three methods of fragmentation: hybrid fragmentation, vertical fragmentation, and horizontal fragmentation. Allocation method is a mapping that identifies which fragment should be kept at which site. Replication offers the advantages of keeping a fragment at multiple sites, hence reducing the overall access time and communication time. In this research, the K-Medoid hard clustering technique is used to achieve non-overlapping vertical fragmentation. The dissimilarity count is used as a measurement to assess the appropriateness of the created fragments. The suggested method produces the best fragmentation outcomes as compared to EMST. Allocating the fragments to suitable locations is the next phase in the architecture of a distributed database.