I. Hernandez and M. Cole (UK)
Grid computing, dynamic scheduling, parallel processing, list scheduling,heterogeneous computing,DAG scheduling
We consider the problem of scheduling parallel applications, rep resented by directed acyclic graphs (DAGs), onto Grid style re source pools. The core issues are that the availability and perfor mance of grid resources, which are already by their nature het erogeneous, can be expected to vary dynamically, even during the course of an execution. Typical scheduling methods in the liter ature partially address this issue because they consider static het erogenous computing environments (i.e. heterogeneous resources are dedicated and unchanging over time). This paper presents the Grid Task Positioning GTP scheduling method, which addresses the problem by allowing rescheduling of an executing applica tion in response to significant variations in resource characteris tics. GTP considers the impact of partial completion of tasks and task migration. We compare the performance of GTP with that of the well-known, and static, Heterogeneous Earliest Finish Time (HEFT) algorithm.
Important Links:
Go Back
