WP#77 - Energy Efficient Data Center Storage

13 February, 2018 | White Paper

Editor:
Jay Dietrich, IBM

Contributors:
Don Goddard: Net App    
Rona Newmark: Dell (retired)
Patrick Stanko: SNIA Green Storage Technical Working Group
Herb Tanzer: SNIA Green Storage Technical Working Group
Gary Verdun: Dell
Chuck Paridon:  SNIA Green Storage Technical Working Group
Dave Thiel: SNIA Green Storage Initiative


The Green Grid Association—a consortium that works to improve IT and data center resource efficiency worldwide— analyzed metric data and configuration information from the Emerald™ analysis tool results generated by individual SNIA and Green Grid member companies and others largely for the purpose of  certifying storage products to the ENERGY STAR® Program Requirements for Data Center Storage.  The Green Grid’s goals were to evaluate the efficacy of the Emerald test suite as a tool to assess the energy efficiency of storage products and to consider its application and use in developing and setting performance/power efficiency thresholds for storage products.

The Storage Networking Industry Association (SNIA®) released Version 2.0 of the SNIA Emerald™ Power Efficiency Measurement Specification software August 12, 2013, concurrently with the publication of the ENERGY STAR Data Center Storage V1.0 Program which was released August 28, 2013. Due to this timing, manufacturers had a minimal understanding of how storage product configuration and component choices would affect the Emerald results. The Green Grid Emerald Analysis Working Group (Emerald WG) collected and blinded the Emerald results, as well as idle measurement data where it was available, from 48 machine types/models and 155 configurations. The Emerald WG analyzed this data to understand the impacts of configuration type and component selection on the Emerald performance power scores for the three workload types tested: Capacity, Sequential and Transactional. This white paper details the working group’s findings and provides storage product manufacturers, regulators, and stakeholders information on the strengths and limitations of the Emerald tool for assessing the energy efficiency of storage products. Overall, the Emerald WG concluded that due to the complexity of storage products and the importance of storage controller capabilities and storage device type, speed and capacity to the workload scores, the Emerald test results can serve as a useful indicator of a storage products performance/power capabilities but they should not be used to set thresholds in mandatory and voluntary storage system energy efficiency schemas.