Job Migration Policies for Grid Environment

Abstract

For efficient resource management in Grid, the resources overloading must be prevented, which can be obtained by efficient Load Balancing and Job Migration (JM) mechanisms. JM addresses the problem of resource overloading through various techniques in a Grid environment. In this paper, JM model has been designed. The proposed JM model offers three policies based on the scheduling, checkpointing and replication for Grid environment decisions. The JM problem has been formulated as an optimization problem and resource scheduling algorithm minimizes the average finish time and makespan. For all the three proposed techniques, JM policies have been designed and the policy rules have been specified in Extensible Markup Language schema for decision making. Each of the above techniques has been executed according to its specific condition at run time. Finally, the performances of the proposed model, algorithms and techniques have been rigorously examined over the GridSim simulator using various parameters, such as makespan, hit-ratio etc. Experimental results establish the superiority of the proposed techniques over the existing techniques.