Prototyping and Evaluation of Virtual Cache Server Management Function for Distributed Web System

We develop a distributed Web system that adjusts the number of virtual cache servers in the Cloud according to load of them to keep responsiveness and reduce running costs. A method of monitoring load of Web servers, and a algorithm to adjust the number of virtual cache servers are discussed in previous researches. This paper describes examination of time for bootup and shutdown virtual cache server, and prototyping a virtual cache server management function that boots up and shuts down virtual cache server using libvirt which provides common API for managing virtual platforms. From results of experiments, we confirm that this function is possible to boot up and shut down virtual cache servers according to load.