A Hybrid Fault Tolerant Scheduler for Computational Grid Environment

Abstract

A computational grid is an environment for achieving better performance and throughput by pooling geographically distributed heterogeneous resources dynamically depending on their availability, capability, performance, and cost and user quality of self-service requirement. Fault tolerant grid scheduling is a significant concern for computational grid systems. The handling of failures can happen either before or after scheduling tasks on grid resources. Generally, there are two approaches used for the handling of failures namely, post-active fault-tolerant approach and pro-active fault tolerant approach. Recently, a fault tolerant scheduler for grids proposed uses pro-active approach in selecting resources by computing scheduling indicator. However, this study did not considered failure of node while the task is being executed. Thus in our study, we incorporates post-active fault tolerant approach to the exiting study, i.e., migrating of task to another node in the event of failure of node while the task is being executed. We constructed a hybrid fault tolerant grid scheduler using GridSim 4.2 toolkit. We demonstrated that our proposed fault tolerant scheduler shows better results in terms of success rate in comparison with the existing fault tolerant scheduler.

An Adaptive Fault Tolerant Scheduling System for Desktop Grid

“…During scheduling the proposed scheduler will use this scheduling indicator to generate scheduling decisions. A most recent scheduling system presented by R.M.R Kovvur et al [20] incorporated rescheduling on account of resource failures during execution of tasks.…”