The most common area of disagreement between IT and FM is "how reliable is reliable enough," says Rick Schuknecht, vice president of the Uptime Institute. A data center can be classified as Tier 1, 2, 3, or 4 under the Uptime Institute's Tier Classification System; the level depends on the amount of redundancy that is in place to ensure that computers will remain operational during data maintenance operations or if there is a system failure. While a Tier 4 facility may be most desirable from one perspective, it may be too expensive and even unnecessary from another.
For example, universities have time to do data maintenance, Spinazzola says, and do not need to have their systems running all the time. It is important for everyone to "buy into the common goal," he says.
One solution for increasing reliability without going to a higher tier is by using two data centers which house the same information. "If IT and FM can work together and do a cost analysis, they might find it is cheaper to do it this way," he says.
One roadblock to coordination is when IT and FM are not reporting to the same boss. "If you have one decision maker, life is a lot easier," Spinazzola says.
Schuknecht would add the security function to that group. "We have long advocated that they have one common boss — that they work for the same organization," he says. The Uptime Institute calls that the integrated management model. "Everyone reports to the data center manager, and that person reports vertically up the chain of command," he says. "When they don't work together they have competing priorities, rules, regulations and urgencies, and that equates to money and risk."
More and more, the trend is toward an IT infrastructure group where engineers who would typically be on the facility side are working under one umbrella with IT, Schlattman says. "Every day we are getting closer to the two working together as the norm. It is becoming an integrative process much more than it was five years ago," says Schlattman. "The more people rely on technology, the stronger the relationship will be between IT and facilities."
In cases where IT and FM do not report to the same boss, Spinazzola says, weekly or monthly auditing and tracking done jointly by IT and FM is advisable — for example, by checking to ensure that the company is not exceeding the number of kilowatts that should be put on each computer rack. "I see this on a regular basis," he says, "where they will do auditing of power cooling capacity per rack and area to make sure that they are sticking to the design criteria. Some buildings have power monitoring built into the cabinet where they can get real-time reports."
By establishing mechanisms to share information, companies can avoid unpleasant surprises. It's still all too common for facilities not to know that IT is bringing in a new server until the day before it arrives. Under a more cooperative culture, facilities would receive such news several weeks before the event so that it could plan accordingly.
"We ask a company if they have ever sat down and enumerated the job tasks that each group does in relation to the data center, and the answer is usually no," Schuknecht says. "We say that you ought to do that and get rid of duplication." One example: a roles and responsibilities matrix for bringing a piece of IT equipment in the door. The matrix identifies who owns the process, who leads it, who participates in it, and who supports it.
It's important to do job-task inventories and compare them across groups, he adds, and to compare policies, protocols, processes, and procedures. The data center should have written procedures, so that the same task is done in the same way every time. "A set of processes and documents reduces risks and brings groups together," he says.
Cooperation in the Data Center Can Reduce Costs, Increase Reliability
Managing Common Disagreements Between FM and IT
