A Context-aware Task Scheduling in Ad-hoc Desktop Grids

Abstract

Abstract To harvest idle, unused computational resources in networked environments, researchers have proposed different architectures for desktop grid infrastructure. In this paper, we present one such infrastructure, called the Mini-Grid Framework for resource management in ad hoc grids using market-based scheduling and context-based resource and application modeling. The framework proposes peer-to-peer architecture that supports several futures: decentralized task distribution, small scale ad hoc grid formation, and symmetric resource. Furthermore, users can model and specify non-performance based parameters that influence resource allocation. Different types of resources can provide similar capabilities but with varying degrees of quality of service. Hence, the resource capabilities are required to be presented in such a way that resource providers can evaluate their capabilities against the requested capabilities for task execution, and the resource consumers can find optimal resources.