A Grid Service for Resource-to-Agent Allocation

The resource allocation problem is a classical problem in the scientific literature. Here, authors present an architectural model to allow optimum resource allocation in a grid environment. They are investigating a decentralized approach to address this problem into an inter--grid nodes context; instead they argue in favor of a centralized approach in an intra--grid node context (i.e., a cluster architecture). Furthermore, authors aim to implement integration between multi-agent and grid systems, in order to deal with the execution of agent-based applications in a grid environment. The paper focuses on a condor--based solution for resource to agent allocation in an intra--grid node context. The prototypal implementation is described, which has been integrated in grid middleware as grid service.