We develop and examine job migration policies by considering effective usage of global memory in addition to CPU load sharing in distributed systems. When a node is identified fo...
Clusters of workstations are increasingly being viewed as a cost-e ective alternative to parallel supercomputers. However, resource management and scheduling on workstations clust...
Abdur Chowdhury, Lisa D. Nicklas, Sanjeev Setia, E...
The interest among a geographically distributed user base to mine massive collections of scientific data propels the need for efficient data dissemination solutions. An optimal dat...
The task of balancing dynamically generated work load occurs in a wide range of parallel and distributed applications. Diffusion based schemes, which belong to the class of neares...
—This paper reviews the characteristics of overlay networks and defines effective relay nodes that can improve the performance of interactive real-time applications. A heuristic ...