Working in partnership with Stanford University and TSO Logic, a provider of data center analytics tools, IT consulting firm Anthesis Group released a report this week that suggests there are 10 million physical servers deployed inside data centers around the world that are currently not actually being used. The report refers to these idle servers as “comatose.”
Citing a general lack of management oversight in the data center, Jonathan Koomey, research fellow at Stanford's Steyer-Taylor Center for Energy Policy and Finance who co-authored the report, says roughly 30 percent of the servers deployed worldwide have not delivered information or computing services in the last six months.
To address what the report estimates to be about $30 billion in IT infrastructure assets sitting idle, Koomey says IT organizations need to aggressively eliminate all the management silos inside their organization. Lack of communication between teams results in deployment of IT infrastructure that ultimately gets wasted.
“There really needs to be one boss, one team, and one budget,” says Koomey. “We need to change the way the data center is managed.”
In the absence of that centralized approach to IT management, he contends, it will only become more difficult for IT organizations to justify acquiring servers when the business can rely on Infrastructure-as-a-Service providers that offer access to servers as needed.
TSO Logic CEO Aaron Rallo says the primary reason so much idle server capacity exists is that IT organizations generally lack access to analytics tools. As a result, they don’t have much visibility into what resources are actually being consumed by applications.
Worse yet, based on the projects provided by the business, Rallo says, most IT organizations wind up overprovisioning server capacity. Because they usually pay for servers using capital budgets, they generally acquire servers years in advance of anticipated need. The projections by the business, however, rarely live up to expectations in terms of IT capacity required.
Rallo says that first and foremost IT organizations need to rationalize their application portfolios in a way that will not only reduce the number of servers they might have sitting idle, but also generate additional savings in terms of the number of application licenses they have running on servers that are not actually being used.
All things considered, Koomey says, organizations need to think of their data center as a strategic asset that needs to be optimized to the fullest extent possible. To achieve that IT organizations need to map out every usage of a server to a specific business process it enables.
The degree to which any organization can actually accomplish that goal will naturally vary. But in the absence of any effort being made at all, Koomey says, it’s now only a matter of time before the IT organization winds up losing control of their data centers altogether.