Abstract:
The power consumption of physical and virtual machines is a major challenge for small to large scale cloud computing data centers. Tow states of the virtual machine, active state if selected by cloudlet and idle state if not selected. Two typed of works applied which are existing work and proposed work. In existing work non-power aware data center,the power consumption used time-shared data center thatdiscarded cloudlet file size to selected VMs. To performance enhanced of power consumption in existing work applied power-aware data centerin proposed work used intelligent distribution of cloudlets that according cloudlet file size organized cloudlet on five range to selected VMs. The result shows in existing work that five VMs selected by five cloudlets, which consumed all powerof the host.VM1 difference in existing work to proposed work by selected cloudlet1 with file size 300 to consumed 1000 second in existing work and selected three cloudlets which are cloudlet1 with file size 300, cloudlet2 with file size 400 and cloudlet3 with file size 500 to consumed 2999.99 second in proposed work, sothere is positive relationship of VM execution time and number of cloudlet that selected by VM. That positive relationship affected positively to power consumption of VM, but in the same time, the VM2 and VM3 not selected by cloudlet stayed inidle state near to zero power consumption and zero execution time in proposed work to reduced power consumption of host from 7.134 KW to 6.562 kW with difference 0.572 kw. VM4 and VM5 selected by cloudlet4 and cloudlet5 in two work there is no change similar to VM2 and VM3. Performance enhancement ofVMs power consumption reduced the cost to increased lifetimeandreducedcarbon footprints to make environment-friendly.
