An exact optimization tool for market-oriented grid middleware

Abstract

With the fast growth of grid service demands, an efficient resource management mechanism is required. Business-oriented algorithms were shown to be adequate to resolve such a problem. Business models mainly consider the problem of computing, storage and network resources virtualization. This paper investigates the problem of resource allocation under flexible traffic in market-oriented grids. We consider a discrete time approach for resource allocation under flexible scheduled traffic in the context of layer-2 Virtual Privates Networks (VPN). We aim to exploit end-users time requirements in order to provide a fair management algorithm in terms of gain for the service provider. For that purpose, we introduce a weighted cost function enabling a service differentiation relying on time constraints disparity. We propose an exact approach based on an Integer Linear Programming (ILP) formulation to solve this problem. Simulation results outline the impact of using flexible traffic on the system performance and the gain of the service provider.