A Novel Approach for Effective Dynamic Load Balancing by using Tendensive Weight
K. Shahu Chatrapati, K. Purnachand, Y. Ranjith Kumar, A. Vinaya Babu, Asst . Proffessor, M. Tech Student · 2008
Clusters have emerged as a primary and cost- effective infrastructure for parallel applications, including communication-intensive applications that transfer a large amount of data among nodes of a cluster via interco nnection networks. Efficient resource usage is a key to achi eving better performance in cluster systems. Previously, most research in this area has focused on balancing the load of each node to use the resources of an entire system more effectively. Memory utilization has a key role in t he implementation of a high performance cluster. However determining the load without considering the memory utilization will degrade the cluster performance. I n this paper, a load metric has been proposed, termed as Tendensive Weight , in order to balance the cluster dynamically. This metric considers both CPU and IO utilization includ ing memory exploitation. Simulation results shows that the system incurs 10% shorter execution time than the conventional approach using number of effective tas ks.