Monitor power to reach higher levels of efficiency
Thursday, Jun 27th 2013

In recent years there has been an increased focus on efficiency in the data center, and this trend is not likely to slow down anytime soon, supporting the idea that IT managers must monitor power in order to operate at an optimal level. ZDNet recently reported that Singapore data centers are considerably power inefficient, and offered up some insight from Gay Chi Sen, director of data center infrastructure management solutions for Schneider Electric Japan and Greater China.

Sen explained that areas like Singapore with warmer climates generally tend to use more power to cool down their data centers. In order to reach higher levels of efficiency, data center managers can utilize temperature monitoring tools to focus efforts on reaching more efficient cooling first.

Power fluctuations
Patrick Donovan, senior research analyst with Schneider Electric's Data Center Science Center, explored the commonly overlooked problem of dynamic power variations in IT equipment in network and server rooms as well as data centers as a whole.

Donovan explained that the total electrical power consumed by IT equipment in data centers was relatively stable historically, but new designs for server processors include energy management capabilities that can result in substantial power consumption fluctuations. Roughly 20 years ago, server power variation was largely separate from computational loads, with the few fluctuations caused mainly by disk drive spin-up and fans, Donovan noted. However, today's processing equipment has additional power management capabilities such as the ability to change the frequency of the clocks, adjust voltage magnitude and move virtual loads. So while power variation hovered around 5 percent two decades ago, today's servers can have a power variation of anywhere from 45 to 106 percent.

Monitoring to fix inefficiencies
According to Donovan, this results in numerous issues, including overheating, branch circuit overload and loss of redundancy. However, with tools that allow IT managers to monitor power and server temperature, an organization can make significant strides toward reducing fluctuations in consumption and reaching efficiency.

"Typically, servers operate at light computational loads, with actual power draw amounting to less than the server's potential maximum power draw capabilities," Donovan wrote. "However, because many data center and network managers can be unaware of this power use discrepancy, they often plug more servers than are necessary into a single branch circuit. This in turn creates the potential for possible circuit overloads, as the branch circuit rating can be exceeded by the total maximum server power consumption."

Donovan also explained that when servers are simultaneously subjected to heavy loads, circuit overloads will occur. When branch circuits are overloaded, the entire circuit can be tripped and power shut off to computing equipment. Furthermore, since this happens as a result of heavy loads, power outages at this time can have significant negative effects on business.

Temperature monitoring to counter overheating
One of the other problems Donovan discussed was overheating. Servers in general consume power and release it as heat, but when there are large variances in consumption because of workloads, the heat released from IT equipment also rises.

"As such, sudden fluctuations in power consumption can cause dangerous increases in heat production, creating heat spots," Donovan wrote. "While data center cooling systems are put in place to regulate overall temperature, they may not be designed to handle specific, localized hot spots caused by increases in power consumption. As temperature rises, equipment is likely to shut down or act abnormally."

A server room monitor is one tool that could prove effective in resolving the problem of overheating. Other than focusing on cooling power consumption, however, businesses can take on tasks like removing unutilized servers from the data center and monitor to identify available capacities that are not yet being fully utilized.