Bridging the gap between IT and facilities

There was a time when an IT admin would hardly give a second thought to slapping a new server into a rack in the data center. There was no need for him or her to fret over whether there was enough power or cooling to support the new machine; that was something for facilities to worry about.

But things have changed. With data centers facing a shortage of floor space, limited energy supplies, and demands from on high to reduce operating costs and carbon emissions, IT admins have had to broaden their purview beyond simply ensuring that servers are serving, storage gear is storing, and the network is blazing. IT is now expected to track energy efficiency at granular and broad levels, including monitoring energy consumption and even heeding climate, humidity, and power availability.

[ InfoWorld is currently accepting nominations for the 2010 InfoWorld Green 15 awards. |  The Green Grid recently announced new data center tools. ]

The good news is vendors continue to roll out tools for monitoring and managing energy consumption, energy efficiency, system performance (both IT and infrastructure), and the like. Unfortunately, these tools still tend to fall into one of two silos, facilities or IT, but that gap is closing. The highly sought single pane of glass for monitoring all aspects of the data center may not be too far off.

One recent entry into the data center planning, monitoring, and management space is the latest version of APC’s InfraStruXure Management Software, designed to provide a unified view of physical infrastructure in the data center, such as generators and air handlers. The software suite equips admins with tools to plan changes to the data center, as well as track service quality, energy use, and the associated costs of essentially anything connected to the network. It performs this feat primarily via sensors and meters — which are increasingly becoming standard in data center gear.

InfraStruXure comprises several modules, including InfraStruXure Energy Efficiency, InfraStruXure Operations, InfraStruXure Central 6.0, InfraStruXure Capacity 6.0, and InfraStruXure Mobile 6.0, each geared toward a particular function.

One of the key features of the package is its ability to calculate PUE on an ongoing basis. Granted, calculating PUE isn’t entirely difficult once you’ve set up meters to gather all the necessary data from your assorted sources, but it’s the go-to metric for data center operators (or their bosses) to assess the overall efficiency of their facilities.

Whereas a CXO might be more interested in the daily or hourly PUE report, anyone charged with overseeing the data center on a day-to-day basis will likely be more interested in InfraStruXure’s mapping and inventory-management functionality. Effectively, these features can provide an admin with a real-time view of the overall health and energy efficiency, from a multifacility perspective (if you have many data centers) down to a granular level, such as the state of a PDU.

Morover, an admin can set up automatic alerts to respond to certain conditions, such as a rise in temperature or a backup power failure. The system also offers user-defined reports and graphs.

Data center operators contemplating changes to the data center might find use in the predictive simulation features of the Capacity application, which can determine how a new set of servers would affect power availability, for example, or which machines would be affected were a specific PDU to fail. That sort of information can help an admin determine the best location of new physical infrastructure or IT gear, rather than having to learn a costly lesson only after the equipment is installed.

You may have observed by now InfraStruXure’s primarily facilities-oriented slant. It can tell you how much power a rack or a server is drawing, but it won’t tell you what’s running on the server or its utilization level. In other words, it doesn’t supply the full data center picture. That’s not a knock at APC; for one thing, the InfraStruXure does integrate with IT management packages such as Tivoli and Microsoft System Center. It also comes with a full software developer kit (SDK), allowing custom integration with other systems.

Still, data center operators are missing the comprehensive tool that seamlessly unites infrastructure and IT data behind a single pane of glass. Imagine, for example, being able to drill down and see not only how much power a server is using, what infrastructure it relies on, and its temperature, but also what applications it’s running, what IT systems (storage, databases) it’s connected with, and how efficiently it’s doing its work (measured in performance per watt). Adding in that kind of application mapping and server-utilization data would provide a far richer picture of a data center’s overall health and efficiency.

A comprehensive suite of tools for monitoring and managing the entire data center ecosystem (to borrow from The Green Grid’s vernacular) can’t be too far off. More IT and facilities equipment are now coming equipped with meters for delivering data such as power draw and temperature. More IT and facilities companies (IBM and APC, for example) are working together on data center offerings. The next logical step, though, is an out-of-the-box data center monitoring and management solution, one that doesn’t treat IT and facilities as separate silos.

This story, “Bridging the gap between IT and facilities,” was originally published at InfoWorld.com. Follow the latest developments in green IT at InfoWorld.com.