摘要：Since the grid is emerging all over the internet, there comes a need for optimal job scheduling and load balancing approaches that takes various characteristics of the grid environment into consideration. Our principal seek is to present a novel fault tolerant approach for job scheduling and load balancing. Our approach called as NovelAlg_FLB mainly focuses on the architecture of the grid, bandwidth of the network, delay caused by communication, availability of resources, heterogeneity of computer resources and other features of jobs. This approach culminates the strength of various other algorithms such as load based and cluster based algorithms. The primary aim of this research is to schedule jobs in such a way that the job’s task delivers their response in minimum time. Also, this approach ensures that the computing nodes are optimally used. The secondary aim of this approach is to reduce complexity and the reduction in the total number of additional communication lines due to load balancing. We have used a java based simulation toolkit called grid GridSim to conduct the simulation of the proposed approach. Experimental results prove that our approach works extremely well in a huge and complex grid environment
