The report refers to these idle servers as "comatose." These particular servers have not delivered information or computing services in the last six months.
The number is taken from a 2012 IDC survey that found 35 million physical servers deployed worldwide. The study found 30% were idle, which translated to approximately 10 million.
The problem isn't technological, it's a general lack of management oversight in the data center, according to a report by Jonathan Koomey, research fellow at Stanford’s Steyer-Taylor Center for Energy Policy and Finance and Jon Taylor, a partner at Anthesis Group, who co-authored the report.
"In the twenty first century, every company is an IT company, but too many enterprises settle for vast inefficiencies in their IT infrastructure," the two wrote in their report. "The existence of so many comatose servers is a clear indication that the ways IT resources in enterprises are designed, built, provisioned, and operated need to change. The needed changes are not primarily technical, but revolve instead around management practices, information flows, and incentives." The findings support a previous report from the Uptime Institute, which also found that around 30% of servers are unused. The 10 million estimated idle servers translates into at least $30 billion in data center capital sitting idle, assuming an average server cost of $3,000, and that does not even take in to consideration infrastructure capital costs and operating costs.
The report says data centers often have excess server capacity because identifying unused or over provisioned hardware is difficult with conventional tools. Simply measuring CPU and memory usage isn’t enough to determine if that server qualifies as comatose. Data center operators need to look at upstream traffic or user access information per server from central IT management, virtualization and workload distribution systems, the report said. In publishing the results, Koomey said removing idle servers would result in "gigawatt-scale reductions in global IT load, the displaced power use from which could then support new IT loads that actually deliver business value. That’s a result that everyone should cheer."
The report was done over the last several months and may not reflect the impact of Windows Server 2003 migration, which has been ongoing for a few months and will continue well into next year. The migration is forcing companies to perform complete inventories of their assets, which may help them discover comatose servers.
And since they've been in deployment since 2003, many of these servers might be the offenders; IT firms might have moved apps off the old server but never shut it down. Firms moving off 32-bit Server 2003 to 64-bit Server 2012 are likely doing significant server consolidation and virtualization, so when it's over, the company has less physical hardware running at higher utilization. Hopefully a similar study such as this will be performed in another year.
The study can be read here .