Scheduling and load balancing

Abstract

Scheduling and load balancing are fundamental issues for the success of parallel, distributed, and network-based computer systems. Even the most powerful high performance computing environment requires proper support of such techniques in order to fulfill the users expectations in terms of performance and efficiency. For this reason, many past and ongoing research projects investigate new and improved techniques for optimized scheduling decisions and load distribution. At the same time, current trends in computer architecture and system environments, such as cluster computing and computational grids, represent ever more challenges for computer scientists and system administrators. The variety and heterogeneity of systems, and the possible distribution of resources across long distances, often require dedicated techniques for each resource constellation as well as dynamic adaptation to the changes of the execution environment.