Computing over Networks: An Illustrated Example
Bernd Bruegge, Hirofumi NISHIKAWA, Peter Steenkiste · 2005
monitor both the application and the environment is critical. For this reason we developed monitoring tools With the advances in high-speed networking, partitioning based on BEE [4], a monitoring kernel for distributed applications over a group of computer systems is environments that allows users to trace important events becoming an attractive way of exploiting parallelism. in their application efficiently. Programming general multicomputers is however very challenging: nodes are typically heterogeneous and shared In Section 2 we first give an overview of the with other users, making the availability of computing application that is used to illustrate the problems of cycles on the nodes and communication bandwidth on the network computing. In Sections 3 and 4 we describe a network unpredictable. This environment often requires static and dynamic partitioning of the problem on Nectar. users to use a programming model based on dynamic load We finally illustrate how the two implementati...