CAREER: DESPOT: Enhanced Dynamic Process Management for Beowulf Clusters on the Grid

Kansas State University, Manhattan KS

Investigators

Abstract

EIA-0092839 Daniel Andresen Kansas State University CAREER DESPOT: Enhanced Dynamic Process Management for Beowulf Clusters on the Grid Current systems for managing workload on clusters of workstations, particularly those available for Linuxbased (Beowulf) cluster, are typically based on traditional process-based, coarse-grained parallel and distributed programming. In addition, most systems do not address the need for dynamic process migration based on differing phases of computation. As we have outlined in the proposal, the DESPOT project will first build a sophisticated thread-level resource-monitoring system for computational, and network resources. This information will be used in an intelligent scheduling system to perform adaptive process/thread migration within the cluster.