Self-Configuration in Parallel Processing

Loosely-coupled environments, such as clusters and grids are increasingly popular platforms for parallel processing. These systems present highly dynamic environments, in which many sources of variability affect the run-time efficiency of tasks. This paper proposes an adaptive strategy for the run-time deployment of tasks of parallel applications in loosely-coupled systems; to continuously maintain efficiency despite the environmental variability. The approach centres around policy-based scheduling which is informed by contextual and environmental inputs such as the round-trip communication time between nodes and their processing performance. To demonstrate the effectiveness of the adaptive approach, a self configuring and self-optimising parallel application is presented