Constrained Energy-Cost-Aware Workflow Scheduling for Cloud Environment

Nowadays, cloud computing providers offer to their users the computing resources that are capable of operating on the CPU frequency in between minimum and maximum values. This gives to the users a big number of Virtual Machine(VM) configurations to choose from when planning for the execution of their applications. On the user side, higher CPU frequency incurs a high monetary cost, on the providers side higher CPU frequency incurs high energy consumption. The reduction rate of cost and the reduction rate of energy differ from each other. A big challenge that arises is how to select the proper CPU frequency that can lead to energy-cost-efficient configuration and strike a good balance between energy-cost and deadline. In this paper, an algorithm that achieves energy-cost-aware VM provisioning by selecting different CPU frequencies for each VM in order to execute workflow within a deadline is presented.