Statistical Service Assurances for Applications in Utility Grid Environments

Utility grid environments o er programmatic access to information technology (IT) resources for applications. There is a substantial literature regarding admission control, resource reservation, and scheduling in support of engineering and scienti c jobs for grids. Typically each job is associated with a duration and peak requirements for resources. In this paper we focus on techniques to support admission control and advance resource reservation for applications with statistical demand characterizations. We consider business applications which require resources continuously but that have resource demands that change regularly based on factors such as time of day and day of week. We present an approach for statistically characterizing the demand pro les of business applications for utility grids. The pro les provide a grid with time varying expectations and bounds on application resource requirements. We present a method that provides statistical assurances regarding the number of resources needed to satisfy the combined requirements of many applications over time. We illustrate the feasibility of our approach with a case study that uses resource utilization information from 48 data center servers. Simulation experiments explore the sensitivity of the assurances to correlations between application resource demands and the precision of the demand pro les.