Resource allocation on the grid: the GRAPPA approach

GRID of heterogeneous computing units is increasingly used to perform distributed computing operations where local resources are globally shared to perform some function. Resource management for highly heterogeneous and dynamically evolving networks where it is difficult to predict what resources are or will be available at some point in time constitute one of the issues. The GRAPPA approach provides the solution for the above difficulty and considers them as different behaviors (or configurations) of the same underlying system.