Data Center IT Equipment Energy Assessment Tools , Current State of Commercial Tools, Proposal for a Future Set of Assessment Tools

Date Published
06/2012
Publication Type
Report
Author
LBL Report Number
LBNL-6252E
Abstract

Information Technology (IT) equipment (servers, storage, and network) in data centers represents a large component of the total data center energy use. When the data center facility is efficiently designed and operated, the IT equipment energy becomes dominant. Majority of work in the area of energy efficiency has been focused on the data center infrastructure. Metrics are still developing when it comes to effectively defining and measuring IT equipment energy consumption.

This research project, which was conducted during the Summer and Fall of 2011, investigated some commercially available assessment tools with a focus on IT equipment to see if such tools could round out the DC Pro tool suite. In this research, the assessment capabilities of the various tools were compiled to help make "non-biased" information available to the public. This research should not be considered to be exhaustive on all existing vendor tools although a number of vendors were contacted. Large IT equipment OEM's like IBM and Dell provide their proprietary internal automated software which does not work on any other IT equipment. However, found two companies with products that showed promise in performing automated assessments for IT equipment from different OEM vendors.

This report documents the research and provides a list of software products reviewed, contacts and websites, product details, discussions with specific companies, a set of recommendations, and next steps. As a result of this research, a simple 3-level approach to an IT assessment tool is proposed along with an example of an assessment using a simple IT equipment data collection tool (Level 1, spreadsheet). The tool has been reviewed with the Green Grid and LBNL staff. The initial feedback has been positive although further refinement to the tool will be necessary.

Proposed next steps include a field trial of at least two vendors' software in two different data centers with an objective to prove the concept, ascertain the extent of energy and computational assessment, ease of installation and opportunities for continuous improvement. Based on the discussions, field trials (or case studies) are proposed with two vendors – JouleX (expected to be completed in 2012) and Sentilla.

Year of Publication
2012
Institution
Lawrence Berkeley National Laboratory
City
Berkeley
Refereed Designation
Unknown
Keywords
Download citation