An Expected Value Prediction of Game Theory for Pre-copy over Cloud Computing

Abstract

Virtual Machine (VM) in a cloud environment has become popular. An operating system (OS) can be installed on a VM platform which emulates a real machine in a cloud environment. In such a setup, the physical host can be taken offline for maintenance and the service continues. This approach relies on Pre-copy live migration technique. With this technique, one of the challenges is how to ensure that the existing memory pages can be securely migrated to the alternative host without wasting unnecessary migrations; having minimum migrations will reduce the total migration time. In this paper, the authors propose a method which uses the expected value of Game Theory over the Gilbert-Elliot (GE) model to predict memory migration. Furthermore, with this method, it is envisaged when the service of the original physical host is halted it can reduce the delay when VM migrates to an alternative host. By the simulations, it is observed that using the Game Theory model is faster than the method in the GE model to obtain the “CUT-OFF” point in varied memory states.