Report: Almost 30% of physical servers sitting idle in data centers


A new report released this week has made the startling find that more than 10 million physical servers deployed in data centers around the world are sitting idle, not being used.

The discovery was made by researchers from the IT consulting firm Anthesis Group, Stanford University and TSO Logic, a provider of data center analytics tools, and refers to the idle servers as being “comatose”, Data Center Knowledge reported.

Jonathan Koomey, research fellow at Stanford’s Steyer-Taylor Center for Energy Policy and Finance, and co-author of the report, said a general lack of management oversight was the main reason why around 30 percent of data center servers have not delivered information or computing services in the last six months.

The report estimates that idle IT infrastructure assets are worth around $30 billion, and recommends companies work quickly to eliminate management silos within their organizations. It blames a lack of communication between IT teams is the main reason for this manageme oversight.

“There really needs to be one boss, one team, and one budget,” Koomey told Data Center Knowledge. “We need to change the way the data center is managed.”

If organizations fail to adopt a more centralized approach to IT management, they will find it more difficult to justify the acquisition of new servers, especially when Infrastructure-as-a-Service companies like AWS already offer access to servers as and when organizations need them.

Aaron Rallo, CEO of TSO Logic, said the lack of access to analytics tools was one of the main reasons behind so many servers sitting idle. This lack of access means IT staff have insufficient knowledge about the resources their applications are consuming.

A second problem is that most IT organizations end up overprovisioning their server capacity. Servers are usually paid for with capital budgets, which mean they’re often acquired years before they’re actually needed. However, business projections rarely live up to expectations when compared to the actual capacity IT requires, Rallo said.

To remedy the situation, Rallo says IT organizations must rationalize their application portfolios so they can reduce the number of servers sitting idle and make savings in terms of the number of application licenses they have running on servers that are not actually being used.

Above all else, the data center needs to be thought of as a strategic asset that should be optimized as far as possible, said Koomey. This means mapping out the usage of every server to a specific business process it facilitates.

Of course, that’s easier said than done, and not all organizations will have the capacity to do so. But in the event that organizations make no effort at all to better manage their IT infrastructure, they could end up losing control of their data centers altogether, Koomey said.

Photo Credit: Tom Raftery via Compfight cc