Date Added: Nov 2011
Few Service providers provide compute intensive and data intensive services over web platform; where in applications can be deployed on demand. These service providers usually employ machine virtualization for providing cost effective solution. At the time of infrastructure purchase, one may opt for a particular instance, assuming that this will satisfy the computational needs. Whereas considering the case when no of service request increases, the more robust instance may be needed. In this paper, the authors present a mechanism which provides optimal instance allocation satisfying the computational needs. Results show that this approach utilizes the infrastructure in more optimal manner.