The ?-PUC-Fit Algorithm: Optimization of the Current Allocation of Virtual Machines in Cloud Computing for Energy Efficiency
Fabio A. Pandolfo, Alcides Calsavara, Luiz A. de Paula Lima, Sediane C. L. Hernandes · 2015
The increase in the rates of energy consumption in cloud computing infrastructures, besides imposing scalability limits, raises concerns about monetary issues and damage to the environment. This paper proposes an algorithm named ε-PUC-Fit, which aims at improving energy efficiency in cloud computing environments through dynamic consolidation of virtual machines. The proposed algorithm has been evaluated through simulations based on real-world workload traces. The results show that, under certain conditions, the ε-PUC-Fit algorithm outperforms other algorithms such as the First-Fit algorithm and a variant of the Best-Fit algorithm.