Table of Contents

Parallel Job Scheduling — A Status Report.- Scheduling on the Top 50 Machines.- Parallel Computer Workload Modeling with Markov Chains.- Enhancements to the Decision Process of the Self-Tuning dynP Scheduler.- Reconfigurable Gang Scheduling Algorithm.- Time-Critical Scheduling on a Well Utilised HPC System at ECMWF Using Loadleveler with Resource Reservation.- Inferring the Topology and Traffic Load of Parallel Programs Running in a Virtual Machine Environment.- Multi-toroidal Interconnects: Using Additional Communication Links to Improve Utilization of Parallel Computers.- Costs and Benefits of Load Sharing in the Computational Grid.- Workload Characteristics of a Multi-cluster Supercomputer.- A Dynamic Co-allocation Service in Multicluster Systems.- Exploiting Replication and Data Reuse to Efficiently Schedule Data-Intensive Applications on Grids.- Performance Implications of Failures in Large-Scale Cluster Scheduling.- Are User Runtime Estimates Inherently Inaccurate?.- Improving Speedup and Response Times by Replicating Parallel Programs on a SNOW.- LOMARC — Lookahead Matchmaking for Multi-resource Coscheduling.