Certain investigations on grid job Scheduling

Abstract

Grid computing aims to find idle resources that are available newlineworldwide across the varied administrative domains to solve data and newlinecomputational intensive problems The problems are divided into several newlineindependent tasks which are concurrently executed by the Grid resources The newlineconcurrent execution of parallel tasks reasonably reduces the overall job newlinecompletion time makespan in almost all the applications There is a need newlinefor efficient global scheduling algorithm to schedule the tasks to the newlineresources newlineA grid is a distributed and heterogeneous environment newlineHeterogeneous environment involves dynamic arrival of tasks where the newlinetasks and resources are from different administrative domains It has to newlinemanage multiple technologies This makes it necessary to have right newlinemanagement tools to find a suitable resource for the submitted task and newlinecontrol the execution which requires considerable effort Meta Scheduler newlinewhich uses scheduling algorithms manages jobs among the available sites in newlinethe grid environment Although there are many methods available for Metascheduler newlinethere is still a need to improve the scheduling techniques to avoid newlinesystem performance degradation newlineIn order to tide over these problems there is a need to optimize the newlineefficiency of the Meta Scheduler in terms of makespan and resource newlineutilization by providing more enhanced techniques newline