| Large scale distributed computational environments, such as computational grids, are growing in popularity for solving large and complex problems in many fields including science, engineering, medicine and commerce. A major function of grid systems is managing the allocation of shared resources (e.g. computational capacity, storage space and files) between grid users. In the recent past, pay-for-use systems have been developed and deployed for solving large computational problems which have lead to the development of new resource management strategies, some based on economic models from the real world.;The efficiency and practicality of a simple grid resource management system are evaluated with and without the use of my options-like framework. Resource allocation efficiency is estimated for three different scenarios (resource trading for immediate use by consumers, resource trading with advanced reservation and resource trading involving options). These three scenarios are measured and compared to each other, most importantly for two extreme situations: high resource demand with low supply of resources and high resource supply with low demand.;Results show that the system supports improved resource utilization and improved price certainty for both resource providers and consumers.;In pay-for-use systems, demand for computational resources may not be uniform. There are certainly seasonal changes that need to be considered and, in some cases, unpredictable demand changes may also occur. These affect the price of resource use. Prices for resources may also change due to other factors (e.g. when a new generation of processors are released and deployed, resource providers may be motivated to increase prices). Existing economic models used in grid systems to allocate available resources do not provide a way to hedge against price increases/decreases for resources. In this research, I introduce an options-like contract framework for use in economy based grid resource management systems. The use of such a contract allows users to hedge against price increases for resources and service providers to hedge against price decreases. |
