Before you shut down an underutilized data server, you better take some extra steps to monitor, test and analyze whether it’s truly a non-productive system. According to a white paper by Viridity, entitled “Non-productive Servers in the Data Center,” it is imperative for data center operators to identify and eliminate the underemployed servers and find a practical method of deploying proper data center management.

It may come as little surprise that 30 percent of servers in the data center workplace are actually not useful and fail to comply with business protocol. The easy way out may be to shut down the server, but this can be extremely problematic in several ways, and even debilitate the careers of those involved. According to the white paper, there are several reasons servers get to the point of becoming completely idle and non-useful as there are many challenges associated with the upkeep and management of these servers.

Often companies have no method of predicting application-usage profiles, thus systems are forced to run with little use. In addition, servers are left running when new applications replace old ones – or those old applications retire. Even further, businesses are unaware of how their applications are performing or the impact they have on the data center environment due to limitations in tracking IT assets, space, power and cooling of the servers.

These non-productive servers cause quite the challenge as data center operators have limited insight into the usefulness or status of the applications running. According to the white paper, many data center operators follow the data center’s Hippocratic Oath “First do no harm, then take no chances.” With a challenge of juggling the maintenance of service and conserving energy, operators reduce power and cooling systems, or replace AC units or UPS’s or more efficient methods – but this has proven both long-lasting and requires a pretty hefty investment in hardware.

So, how is it done? An inventory and monitoring system, using Microsoft (News - Alert) Excel and Visio, can poll the server load to find non-productive servers. Or, an agent living on the server can put them in several sleep stages. However, according to the white paper, the best method would be to “discover, monitor and analyze servers for function, and then to virtualize the operating system and redeploy or decommission less-useful servers,” thus forcing power usage to plummet and providing more room for other business applications.

The white paper provides data center operators with several benefits in following a procedure in proper data center management, including:

·Shutting down serves reduces energy waste, and boosts data center value

·Eliminating unused servers allows of a quick fix, and guarantees savings

·Improving power usage effectiveness by reducing wasted power and cooling

·Reducing the total number of assets to be managed

·Eliminating “hot spots”

·Trimming an organization’s carbon footprint

·Making room for more advanced technology

“We propose continuous, non-intrusive monitoring of the servers in   the data center over extended periods,” the white paper concludes. “A system that sets up attributes that identify IT components and maps their business needs could help determine whether or not a server is a redeployment candidate, a consolidation candidate, or a useful—albeit limited—part of the system.”

Viridity Software provides innovative data center optimization and energy management software.

Tammy Wolf is a TMCnet copy editor. Previously she was assistant to the editor at The Darien Times, a weekly newspaper in Darien, Conn., where she edited submissions, did page layout and design and helped manage the newspaper's website. To read more of her articles, please visit her columnist page.