Search

The necessity and significance of improving the energy efficiency of cloud implementations have increased due to the rapid growth and proliferation of cloud computing services around the world. Virtual machines (VMs) comprise the backend of most, if not all, cloud computing services. Several VMs are often consolidated on a physical machine to efficiently utilize its resources. In this paper, we take into account the cooling and network structure of the datacenter host ing the physical machines when consolidating the VMs so that fewer racks and routers are employed, without compromising the service-level agreements; consequently, idle routing and cooling equipment can be turned off in order to reduce the energy consumption. Our experimental results on four benchmarks show that the proposed techniques improve energy consumption of servers, network equipment, and cooling systems by 2.5%, 18.8%, and 28.2% respectively compared to state of the art, resulting in a total of 14.7% energy improvement on average in the entire datacenter. [ABSTRACT FROM AUTHOR]