Energy-Conscious Cloud Computing Adopting DVFS and State-Switching for Workflow Applications

Reducing energy consumption without scarifying service quality is important for cloud computing. Efficient scheduling algorithms HEFT-D and HEFT-DS based on frequency-scaling and state-switching techniques are proposed. Our scheduling algorithms use the fact that the hosts employing a lower frequency or entering a sleeping state may consume less energy without leading to a longer makespan. Experimental results have shown that our algorithms maintain the performance as good as that of HEFT while the energy consumption is reduced.