An elastic capacity planning method in cloud

With the development of the cloud computing technology, more and more cloud platforms support elastic buying patterns of computing resources. Today, customers can adjust the amount of resources bought from cloud providers according to the changes of application needs. It is important to efficiently determine a capacity plan according to a number of computing instances provided in the market with different capacities and prices. In this paper, we propose an algorithm to generate a resource provisioning scheme which both satisfies the application needs and minimizes the cost for customers who get computing resources from the cloud platform. The algorithm is based on branch and bound method. And the algorithm is verified to be both cost and time efficient.