Availability-Aware Multi-Objective Cluster Allocation Optimization in Energy-Efficient Datacenters

Changing workload structures in data centers due to network virtualization puts much pressure on data center networks. This is especially true for service-oriented applications that have a time-bounded response. For simplicity of administration, network virtualization consolidates computer resources and network operations into a single software-based administrative unit called a virtual network. As more people send requests to the server at once, it becomes slower. Buying a new server is too costly. Therefore, we build a virtual system to make it easier to manage this issue. As the number of cloud apps deployed in datacenters grows, this tendency will continue. There are several physical machines (PM) in a data center that are linked together via switches. Virtualization technology is often used to share PM hardware resources to provide scalable and elastic computing capabilities. This virtual cluster includes several virtual machines, each of which uses PMs' resources as needed. As more applications are installed, the energy consumption of a data center increases. The assignment is referred to as virtual cluster allocation. This paper discusses cloud computing’s virtual networking, cluster building, and virtualized data center architecture and automation. VMs and virtual clusters are used to demonstrate dynamically organized group, grid, and cloud architectures.