by Greg Shields

No, really. The temperature in your data center is far, far too cold. As much as 15% too cold, in fact, according to a study reported originally in the UK publication The Register ( http://www.theregister.co.uk/2009/08/31/data_centers_run_too_cool ).

The study was done as a partnership between Intel, IBM, HP, Liebert Precision Cooling, and the Lawrence Berkeley National Lab. In it, every single data center interviewed for the study was found to be cooling their data center to a temperature below the maximum necessary for the safe operation of servers. The most optimum number: 27°C or 80.6°F . This guidance, which comes from the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) was elevated from a 2004 recommendation of 25°C or 77°F .

Interestingly, this setting is significantly higher than the temperature currently being used by every single interviewed data center. None had Computer Room Air Conditioning (CRAC) units set to higher than 74°F, with the majority dropping the thermostat down to 70°F or below. Some dropped the mercury to a bone-chilling 68°F.

A major problem with data center cooling has to do with hot and cold spots in the room. Certain parts of a cooled data center, those with the most density of power-consuming hardware, can grow to exceptional temperatures, while other unused parts freeze. Today’s CRAC units remain incapable of evening out the temperature, with air conditioning hardware often using communication languages such as Modbus while servers and networking hardware use completely different languages such as WS-MAN or IPMI. The two languages do not easily communicate with each other.

The news from this report is intended to highlight not only that data centers can turn up the thermostat a bit and save on costs but also that greater cooperation is needed between data center equipment vendors and HVAC suppliers in terms of their communication potential. The assumption here is that once servers can communicate their air conditioning needs directly to CRAC units, it grows possible to more efficiently distribute cooling to specified data center locations. The result: A substantial reduction in cost.

What should you do? You might not be ready yet to turn down the chillers to a full 80 degrees. But increasing the temperature even a small amount can result in a reduction in cost to cool. With tons of chilling requiring almost the same amount of electricity as the power to turn on your servers, this reduction in cost can be substantial, with the largest data centers seeing the largest gains. If you do, however, decide to make changes, ensure that you’re watching internal temperatures to be sure that components within your hard-spent equipment don’t go above established thresholds.

About the Author

Greg Shields is an independent author, speaker, and IT consultant, as well as a Partner and Principal Technologist with Concentrated Technology. With 15 years in information technology, Greg has developed extensive experience in systems administration, engineering, and architecture specializing in Microsoft OS, remote application, systems management, and virtualization technologies. He is a Contributing Editor and columnist for TechNet Magazine and Redmond Magazine, and serves as the Series Editor for Realtime Publishers, the world’s leading provider of high-quality content for the IT market. Greg is a highly sought-after and top-ranked speaker for both live and recorded events, and is seen regularly at conferences like TechMentor Events, Microsoft Tech Ed, VMworld, and more. He is a multiple recipient of Microsoft “Most Valuable Professional” award.