A PREDICTION BASED MULTI- PHASES LIVE MIGRATION APPROACH TO MINIMIZE THE NUMBER OF TRANSFERRED PAGES, IN CLOUD COMPUTING ENVIRONMENT

Cloud computing provides a cost-effective and dynamic scalable services to the user on the rent basic with the help of internet. These services can be used at anytime and anywhere in the world. Virtualization is the prime technology in the cloud which enables efficient resource management and isolation for various big data applications. As cloud provide services to the multiple user and cloud resources are share by the multiple user to increase the resource utilization. In order to increase the resource utilization provider creates the number of virtual machine into a single physical machine which may create a situation where some virtual machines are overloaded. In this case virtual machine is transferred to the other physical machine known as virtual machine migration. Several steps are involved in the migration process. These steps are source physical selection, virtual machine selection, target physical machine selection and selecting method for transferring the virtual machine data. Performance of the virtual machine is mainly depends on the method which is used to transfer the virtual machine data. Pre-copy, post-copy and stop and copy are the methods which are used for sending the data from one physic machine to another physical machine. This paper present an pre copy approach which minimize the total migration time and down time by minimizing the number of transferred pages. Main concept which is use to reduce the number of transferred pages is to avoid the pages which is highly modified. CloudSim simulator is use to implement our approach.