Nearly two-thirds of IT and facilities personnel consider their data center energy efficiency "average" or worse -- and their development and test environments might be the biggest cause of that, according to a survey conducted by Cassatt.

More than a quarter of survey respondents said that greater than 60 percent of their development and test servers are idle during off-peak hours. There is some good news, though: 62 percent are working on a data center energy-efficiency project now or expect to within the next year, according to the "Cassatt 2008 Data Center Energy Efficiency Survey." And, contrary to conventional wisdom, 59 percent would consider turning off computers that are idle.

As expected, virtualization ranks highest on the energy-efficiency project list, with 69 percent of respondents pursuing a server consolidation/virtualization strategy, and nearly 49 percent pursuing storage consolidation/virtualization. But, while nearly half of the companies (46 percent) say they need a payback on energy efficiency projects in less than two years, organizations are primarily pursuing consolidation, which is frequently a longer-term project.

More than half of respondents recognize the importance of using more efficient equipment, but only a quarter (24 percent) have plans in the works to improve the efficient operation of that equipment with approaches like active power management software to shut off unused servers. Even though the survey shows server power management to be a missed opportunity for many organizations, a number are looking to complement their long-term energy-efficiency projects with innovative techniques designed to deliver compelling, short-term benefits.

"We conducted the 'Cassatt 2008 Data Center Energy Efficiency Survey' to learn more about the extent to which companies are pursuing energy-efficiency initiatives and the rationale behind the ones they choose," said Bill Coleman, chairman and CEO of Cassatt Corp. "Many of the findings were expected, such as those that emphasize the data center power crunch, the popularity of virtualization as a potential solution, and the massive waste in development and test environments.

"Less expected," Coleman continued, "and very problematic for the industry, are the findings that show that many companies simply don't measure their power consumption at all, or do so at a very superficial level. If you can't measure it, as they say, you can't manage it. And it may be that companies are fixing only part of the problem with initiatives based on incomplete information. While organizations are showing a willingness to try some new ideas, many are still ignoring simpler solutions that could help them with energy efficiency almost immediately."

Justifying Turning Off Servers to Save Energy
Are companies that are engaging in complex, long-term, costly server consolidation projects based on virtualization also thinking about simple steps, like turning their unused servers off? According to the Cassatt survey, some are -- which goes against conventional wisdom in IT operations today. In fact, 59 percent said that, yes, they could justify turning off servers for some period of time. When asked how many hours each day a server needs to be idle to justify turning it off, 11 percent say one to three hours, 20 percent say four to five hours, nearly 16 percent say five to seven hours, and nearly 11 percent say eight to 10 hours. Forty-one percent say they cannot justify turning a server off.

While a solid majority of respondents are willing to consider turning off servers they aren't using, only 24 percent are pursuing server power-management software as an energy-efficiency strategy today. On the face of it, a solution like server power management is a simple strategy, like turning off the lights when leaving a room. What's holding them back? Companies cite a range of reasons, including "impact of turning off idle machines on application availability" (45 percent), "application stability when shut down and restarted" (42 percent), "impact on physical reliability" (36 percent), "difficulty determining ROI for a power-management solution" (28 percent), and "lack of integration into existing systems management tool" (26 percent).