The Schedule Algorithm Of Grid Workflow Based On Survivability

As Internet technology is widely applied, a variety of disciplines, especially data-intensive and computing-intensive applications, require the analytic and collaborative ability more and more highly. Therefore, grid computing technology is put forward. In grid technology, schedule algorithm is the hotspot in numerous research domains.On the basis of analyzing the recent schedule algorithm, we find that recent work mainly focuses on how to optimize task schedule to minimize the execution time of the whole workflow, ignoring malfunction of grid resource.For this reason, this paper considers grid workflow as its research subject, and introduces survivability conception into schedule algorithms based on performance and QoS.Firstly, this paper proposes a schedule algorithm which considers both performance and survivability by modifying Min-min, Max-min and Sufferage. It pursues the shortest finish time as well as survivability to avoid scheduling a task to an unreliable machine and to reduce the influence of machine malfunction on task. We use a preferable parameter to adjust ratio of these two factors when they are in conflict with each other.Secondly, we propose a heuristic schedule algorithm based on deadline and survivability, taking the latest finish time of application assigned by users as a QoS example.This paper also develops a simple grid workflow management system in order to verify the effect of the algorithms mentioned above.