An efficient fault tolerant hypercubic p2p grid resource management system

Abstract

Grid Resource Management is one of the major components of a grid system, which identifies resource requirements, matches, allocates, schedules and monitors resources, for utilizing them efficiently. Peer-to-Peer(P2P) and Grids are started as two different distributed computing philosophies: Grid computing represents a more enterprise-wide or multi-institutional orchestration of relatively homogeneous and powerful distributed computing resources, to automate and optimize the execution time of consuming tasks, whereas P2P networks lay an emphasis on the discovery and sharing of resources on the edges of a heterogeneous network. The Set Pair Analysis (SPA) method was applied to the task scheduling methodology in the HPGRID, Mainly three variants of scheduling algorithm was proposed SPA based Double Min-Min, SPA based Min-Mean and Bounded Min-Mean which gave better makespan and balances the load. Quality of Service (QoS) is mainly considered for matching of resources in a non-deterministically shared heterogeneous environment. A novel approach for the QoS guided task scheduling algorithm called QoS guided P2P Grid Metascheduler (QPGM) has been proposed, using the Min-Min with Mean Computation algorithm has been designed, which gives improved performance on load balancing and generates an optimal schedule so as to complete the tasks in a minimum time and to utilize the resources efficiently. The simulation results demonstrate that the QPGM task scheduling algorithm can get better effect for a large scale optimization problem. Node failure and link failure may occur dynamically in grid system, provided that the total number of faulty nodes at any time is the node connectivity n of the HPGRID, whereas in the hypercube it is less than the node-connectivity n of the n-cube. In the HPGRID of dimension n, the strategy tolerates up to n encountered faults in its routes, which are free from deadlock and livelock. newline newline newline