An Improvement on Algorithm of Grid-Workflow Based on QoS

【Author】 ZHANG Yun-feng, GE WeiDepartment of Computer Science, Northwest University, Xi’an 710069, Shaanxi, China

【摘要】 <正> With the emergence of grid computing, new challenges have arisen in workflow tasks scheduling. The goal of grid-workflow task scheduling is to achieve high system throughput and to match the application needs with the available computing resources. This matching of resources in a non-deterministically share heterogeneous environment leads to concerns on quality of service (QoS). Grid concept is presented in this paper, coupled with the QoS requirement of workflow task and an improved algorithm-ILGSS algorithm, has been brought out. The complexity of the improved scheduling algorithm has been analyzed. The experiment results show that the improved algorithm can lead to significant performance gain in various applications. An important research domain-adaptive workflow transaction in grid computing environment , has been explored and a new solution for the scheduling of distribute workflow has been bring forward in grid environment.更多还原

【Abstract】 With the emergence of grid computing, new challenges have arisen in workflow tasks scheduling. The goal of grid-workflow task scheduling is to achieve high system throughput and to match the application needs with the available computing resources. This matching of resources in a non-deterministically share heterogeneous environment leads to concerns on quality of service (QoS). Grid concept is presented in this paper, coupled with the QoS requirement of workflow task and an improved algorithm-ILGSS algorithm, has been brought out. The complexity of the improved scheduling algorithm has been analyzed. The experiment results show that the improved algorithm can lead to significant performance gain in various applications. An important research domain-adaptive workflow transaction in grid computing environment , has been explored and a new solution for the scheduling of distribute workflow has been bring forward in grid environment.更多还原