Cost based data center selection policy for large scale networks

Cloud computing is the technology where, IT infrastructure and applications are provided as a service using data center to the end user as pay per use model. Cloud infrastructure has distributed applications that can be deployed in data centers situated at different geographic locations. These data centers are overloaded as more number of client applications is serviced in the same location. Distributed application impacts its performance for users that are far from the data center. Applications of different users may have different configuration, composition and deployment requirements. Measuring such user application workload, performance of various resources is tough to achieve. We propose a service broker policy to reduce overloading of the data center by redirecting the user requests to the next neighboring data center which is in off peak time resulting cost effectiveness. A quantitative performance estimate may guide the service provider in making decisions for right level of resource. In this paper, we consider the location of cloud infrastructures, application clients between data centers so as to optimize the waiting time and cost to the end users. The aim of this paper is to briefly discuss about various existing service broker policies and new proposed policy to minimize the cost and processing time of data center.