Grid Service Level Agreements Using Financial Risk Analysis Techniques

Abstract

Grid computing continues to hold promise for the high-availability of a wide range of computational systems and techniques. It is suggested that Grids will attain greater acceptance by a larger audience of commercial end-users if binding Service Level Agreements (SLAs) are provided. We discuss Grid commoditization, and in particular the use of Grid technologies for certain kinds of financial risk analysis where both data and computation requirements can be substantial. The nature of such analysis, and the need for it to run to completion, suggests the need to guarantee availability and capability in the underlying Grid infrastructure. This further suggests that it is necessary to be able to evaluate the infrastructure in relation both to historic analysis and to the needs of a specific analysis. Our aim, then, becomes one of predicting availability and capability, essentially introducing risk analysis for Grids. Prediction, quantification of risk, and consideration of liability in case of failure, are considered essential for the future provision of Grid Economics – specifically, relating to the provision of SLAs through resource brokers, and comparable to markets in other commodities – but perhaps also more widely applicable to the configuration and management of related architectures such as those of Peer-to-Peer (P2P) and Cloud Computing systems. The authors explore and evaluate some of the factors required for the automatic construction of SLAs, with broad consideration for Financial Risk and the potential formulation of a Grid Economy as a commodity market, which may in future involve the trading and hedging of risk, options, futures and structured products.