Cluster-Based Hybrid Scheduling Mechanisms for Workflow Applications on the Grid

Thanks to advances in wide-area network technologies and the decreasing cost of computing resources, Grid computing is now a viable computational paradigm. However, many aspects of successfully using the Grid remain research topics. Among them, we identify scheduling of workflow applications as a key problem. While there is much work on static scheduling approaches for workflow applications in parallel environments, little work has been done on a real-world Grid environment. In this paper, we launch four model workflow applications with different configurations on a multi-cluster Grid testbed. By observing the applications' performance, we propose a new cluster-based hybrid scheduling mechanism that dynamically executes a top-down static scheduling algorithm using the real-time feedback from the execution monitor. Our experimental results show that this approach achieves the best performance among all the scheduling approaches we implemented on both reserved resources and those with external loads.