Agent Based Service Discovery in Computational Grids
Grid computing aims to develop the computational infrastructure to enable pervasive, dependable, consistent and inexpensive access to advanced computational and data capabilities. This talk will describe how a a de-centralised approach to resource management and discovery, based on a community of interacting software agents can be used to support service discovery and advertising on Computational Grids. Each agent either represents a user application, a resource, or a MatchMaking service. The proposed approach can support dynamic registration of resources and user tasks, facilitating the establishment of dynamic clusters. Resource capability and task requirements are described using an object based data model, enabling new types of devices or new features in existing devices to be identified. Co-ordination between agents is supported through the concept of "utility" -- whereby agents select tasks which maximises their local utility. The matchmaking and discovery mechanism is tied in to this utility idea -- and tries to improve discovery for tasks that lead to a higher local utility for each agent.