Component level risk assessment in Grids: A probablistic risk model and experimentation

The current approaches to manage risk in Grid computing are a major step towards the provision of Quality of Service (QoS) to the end-user. However these approaches are based on node or machine level Assessment. As a node may contain CPU(s), storage devices, connections for communication and software resource, a node failure may actually be a failure of any of these components. This paper proposes a probabilistic risk model at the component level; the probabilistic risk model encompasses series and parallel model(s). Our approach towards risk assessment is aimed at a granularity level of individual components as compared to previous efforts at node level. The benefits of this probabilistic approach is the provision of a detailed risk assessment to the Grid resource provider leading to risk aware scheduling and an efficient usage of resources. Grid failure data was analyzed and experimentation was conducted based the proposed risk model. The results of the experiments provide detailed risk information at component level for the nodes required in the SLA (Service Level Agreement).