Abstract
Grid computing plays an important role in solving large-scale computational problems in a high performance computing environment. Scheduling of tasks to efficient and best suitable resource is one of the most challenging phase in grid computing systems. Grid environment reveals several challenges in efficient scheduling of complex applications because of its heterogeneity, dynamic behavior and shared resources. Scheduling of independent tasks in grid computing is dealt by a number of heuristic algorithms. This study proposes a new heuristic algorithm for mapping independent tasks in a grid environment to be assigned optimally among the available machines in a grid computing system. Due to the multi-objective nature of the grid scheduling problem, several performance measures and optimization criteria can be assumed to determine the quality of a given schedule. The metrics used here include makespan and resource utilization. This algorithm provides effective resource utilization by reducing machine idle time and minimizes makespan. This algorithm also balances load among the grid resources and produce high resource utilization with low computational complexity. The proposed algorithm is compared with other popular heuristics for performance measures.
Highlights
Grid computing allows to use remote resources in a high performance computing environment for solving large scale computational problems
Programs define schedule for assigning tasks to available machines and calculates makespan, resource utilization, average resource utilization, flow time and fitness value based on the ETC matrix supplied to it
Better results are produced in terms of makespan, resource utilization and fitness value compared to other algorithms
Summary
Grid computing allows to use remote resources in a high performance computing environment for solving large scale computational problems. Resources in a grid environment may be homogeneous or heterogeneous. Grid environment reveals several challenges in efficient scheduling of complex applications because of its heterogeneity, dynamic behavior and shared resources. In the case of static scheduling, information related to all resources and tasks in the Grid as application is assumed to be known earlier by the time the application is scheduled. In the case of dynamic scheduling, the basic idea is to perform task allocation as the application executes. Both static and dynamic scheduling is widely adopted in Grid computing. It’s hard to include load balance as metric to obtain stable and efficient scheduling algorithm
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
