Savvis – Sachin Kapasi – 2009

Goals

Sachin Kapasi, a MBA student at University of North Carolina, spent his summer at Savvis, a leader in outsourced Internet infrastructure services for the enterprise. Kapasi was asked to conduct a scientific experiment to quantify the energy savings that could be achieved by undertaking a virtualization strategy for data centers.

Solutions

Virtualization, in its simplest form, means consolidation of multiple application workloads, currently residing in separate under-utilized systems, onto a common platform with shared resources, and is a fundamental building-block of modern cloud computing. Consolidating workloads with virtualization technology saves energy and reduces energy costs.

For the study, Kapasi worked with Savvis’ Cloud solutions team to set up a cloud environment and tested a virtualization management algorithm on 17 servers with simulated typical workloads. The results of the study demonstrated substantial energy savings; virtualization of 17 servers resulted in a cumulative annual saving of more than 20,000 kWh.

Potential Impact

This could be used as a future framework to help data centers quantify and estimate the extent of energy savings achieved by pursuing a virtualization strategy.