Research On Adaptive Grid Workflow Scheduling Based On Domain Trust Mechanism

With the development of grid technology and cloud computing, more and more complex applications can be solved by using grid workflow in grid platform. Only through the traditional way of security authentication can't solve the issue caused by the dynamic and autonomy of resource. This paper proposes a workflow scheduling algorithm to ensure operating efficiency. Based on dynamic assessing the trust of each domain, we propose a method to produce trust-based scheduling on domain and adaptive dynamic grid workflow. According to the parallel character and task dependency of grid workflow, we develop an efficient and reliable abstract grid workflow model. When the system is running, this method can adaptively select the appropriate physical service to complete the abstract model by real-time adjustment.Usually we set workflow modeling use DAG, Petri nets, etc. Because the grid's characteristics is performing concurrent tasks well, thus particularly suitable for solving computation-intensive and data-intensive applications. So the application in grid fits the characteristics of DAG model, using the DAG model we can reasonable distribute task group and calculate different priority of tasks.For determining the reliability of different resources, trust mechanism is introduced but not simply adds to the trust property in the workflow schedule. Since domain structure is the most structure of the grid, we use domain trust mechanism and dynamic adjust the trust of domains during operation, according to the domain trust to choice the different scheduling methods to ensure efficient operation of the system.Experimental results show that the method can not only guarantee the stable of the system, but also reduce the time and cost in completing tasks. And it also can adaptively select the appropriate way to complete the task when some nodes fail accidentally.