Running Large-Scale Applications on Cluster Grids
International Journal of High Performance Computing Applications
A transparent framework for hierarchical master-slave grid computing
Euro-Par'06 Proceedings of the CoreGRID 2006, UNICORE Summit 2006, Petascale Computational Biology and Bioinformatics conference on Parallel processing
An economy-driven mapping heuristic for hierarchical master-slave applications in grid systems
IPDPS'06 Proceedings of the 20th international conference on Parallel and distributed processing
Hi-index | 0.01 |
The widespread diffusion of metasystems and grid environments makes it necessary to employ programming models able to well exploit a high, variable number of distributed heterogeneous resources. Many software frameworks designed for Grid computing do not address this problem. They only allow the use of existing programming libraries based on explicit message-passing communication models, often not suitable to manage the variability of a Grid. In this paper we present the customization of a component-based middleware for metacomputing, HiMM (Hierarchical Metacomputer Middleware), in order to support distributed programming based on the Active Object model provided by ProActive. This way a meta-system can be efficiently and transparently programmed by unifying the asynchronousremote method invocation model and the reflection provided by meta-objects.