An Efficient Scheduling of HPC Applications on Geographically Distributed Cloud Data Centers

Cloud computing provides a flexible infrastructure for IT industries to run their High Performance Computing (HPC) applications. Cloud providers deliver such computing infrastructures through a set of data centers called a cloud federation. The data centers of a cloud federation are usually distributed over the world. The profit of cloud providers strongly depends on the cost of energy consumption. As the data centers are located in various corners of the world, the cost of energy consumption and the amount of CO2 emission in different data centers varies significantly. Therefore, a proper allocation of HPC applications in such systems can result in a decrease of CO2 emission and a substantial increase of the providers’ profit. Reduction of CO2 emission also mitigates the destructive environmental impacts. In this paper, the problem of scheduling HPC applications on a geographically distributed cloud federation is scrutinized. To address the problem, we propose a two-level scheduler which is able to reach a good compromise between CO2 emission and the profit of cloud provider. The scheduler should also satisfy all HPC applications’ deadline and memory constraints. Simulation results based on a real intensive workload indicate that the proposed scheduler reduces the CO2 emission by 17 % while at the same time it improves the provider’s profit by 9 % on average.