Threshold Scheme Approach to Balance Virtual Machines Load in Private Cloud

This paper addresses trade off problem when cloud providers attempt to balance their own objective and users quality of service evenly. Acquiring fast processing time for some specific tasks workload most likely will satisfy users. In response, cloud providers need to accommodate it by provisioning good tasks allocation policy. Cloud computing as business view, on the other hand, allows provider to manage their physical resources smartly in order to save the energy consumption for their own merit. The approach is beneficial for private Cloud provider in which the privacy can be maintained while still leveraging the potential of public Cloud. This work would like to find a middle point for the good of both. In this paper, virtual and physical machine load balancing scheme is proposed to satisfy both side. This paper contains two different allocation policies in two different level view: Tasks scheduling view and VMs remapping view.