A Tabu Search Algorithm for Object Replication in Distributed Web Server Systems

摘要

Data replication has been considered a promising technique for improving performance of a large distributed web server (DWS) system. One of the key issues in the design of DWS is determining the optimal number and placement of objects (e.g. HTML documents) on the web servers. This paper specializes the basic tabu search algorithm for object replication problem that not only determines the number of replicas but also their placement in a web server system to minimize a cost function subject to storage and server load constraints. The performance of the proposed algorithm is evaluated and compared with four other algorithms through a simulation study and results are reported. The simulation results demonstrate that the proposed tabu search is an effective and superior algorithm to solve the object replication problem as compared to many other proposed algorithms.