An Efficient Approach for Data Placement in Distributed Systems

The problem of allocating data fragments to the most appropriate sites of a distributed database systems is a difficult one to solve, however, most available approaches rely on heuristic techniques usually based on mathematical programming and formulations. The approach adopted in this paper contributes in determining best possible allocation of a data fragment in a distributed environment based on the fragment access patterns and the cost of moving data fragments from one site to the other.