Virtual private servers (VPSs) such as the Amazon EC2 are widely used throughout the world. It is necessary to use VPSs properly in terms of the users' usage patterns and the frequency of use. However, it is often only the cost that receives attention. Therefore, a VPS may not serve the user's purpose. Our proposed system evaluates a certain constant index so that the user can select the appropriate VPS. Moreover, automation is accomplished from verification to evaluation to rapidly and adequately verify the various VPSs that will be routinely offered in the future. We proposed the use of the system at Shizuoka University where hundreds of VPS were introduced. We constructed the system for an analytic hierarchy process using the score of each benchmark. With this system, we can easily propose the best server to clients.