Recently a great deal of attention has been paid to the design of hierarchical shared memory cluster system. Cluster computing has made hierarchical computing systems increasingly common as target environment for large-scale scientific computations. This paper proposes hierarchical shared memory cluster architecture with load balancing and fault tolerance. Hierarchies of shared memory and caches structure the architecture. The hierarchical load balancing approach focuses on reducing the redistribution cost. The fault tolerant model is adopted to build highly available clusters in hierarchical shared memory clusters. Performance analysis and results reveal that hierarchical shared memory clusters performs much better creating a reliable hierarchical network cluster system with high scalability.

ABSTRACT

Abstract-In this paper, an architecture with a load balancing model and a fault tolerant model for virtual shared memory clusters is proposed. The centralized dynamic load balancing model uses manager worker concept with a virtualized compute server. The virtual server can be expanded "on the fly" for load balancing by adding virtual machines temporarily. The fault tolerant model controlled by a virtual machine monitor describes the checkpointing based recovery methods. The performance evaluation results show that the proposed system achieves significant speedup in terms of execution time and checkpoint time.Index Terms-Virtual cluster, virtual machine, virtual server, load balancing, fault tolerance, check pointing and recovery.

In recent past, many types of shared memory cluster computing interconnection systems have been proposed. Each of these systems has its own advantages and limitations. With the increase in system size of the cluster interconnection systems, the comparative analysis of their various performance measures becomes quite inevitable. The cluster architecture, load balancing, and fault tolerance are some of the important aspects, which need to be addressed. The comparison needs to be made in order to choose the best one for a particular application. In this paper, a detailed comparative study on four important and different classes of shared memory cluster architectures has been made. The systems taken up for the purpose of the study are shared memory clusters, hierarchical shared memory clusters, distributed shared memory clusters, and the virtual distributed shared memory clusters. These clusters are analyzed and compared on the basis of the architecture, load balancing, and fault tolerance aspects. The results of comparison are reported.

No abstract