Energy efficient virtual machine placement technique using banker algorithm in cloud data centre

Cloud computing is uses by many users trying to deploy numbers of application and virtual machine based on their needs. The demand is high and cloud provider has to manage efficiently that user don't face any downtime due to non availability of the services and downtime of the system. A virtual machine placement which works on Infrastructure as a Service (IaaS) the down layer in cloud computing provides Virtual machine (VMs) on demands, VMs are of different sizes and available in a wide range which is ready to deploy with the Operating System(OS). While placing the VMs in data centre, provider or broker has to keen care that host (server) are not overloaded and the system is out of resources. As cloud computing derives from distributed system there is high chances of deadlock while resource allocation. This work focus on deadlock avoidance by using banker algorithm, checking the whether the system is safe state or unsafe state while allocation. The proposed work is run with different overload and a VM selection algorithm to check which suite the best. The BANKER algorithm is experimented with 800 servers with 1024 Virtual Machines. From the experimental result, BANKER algorithm reduces energy in higher level.