Are You Prepared?

Organizations often build secondary data centers in other cities, typically far enough away to escape damage in the event of disaster and with the ability to operate from different power grids and water sources. “In many years of doing this, I think the biggest change is that disaster recovery centers have evolved into full-blown secondary data and production centers,” Doty says. “More companies find it cost effective to have that second center where they can replicate data, which also allows home-based workers to get to it in the event they can’t access data from headquarters during a disaster.”

The traditional disaster recovery strategy had everyone converge on one location, Doty says, but workers today have “more connectivity from home or mobile sites than they’ve ever had before, and they can go to those sites if a headquarters building is damaged or power is out.”

Bult says he spends an increasing amount of time in his consulting work helping companies create redundant paths for telecommunications. “The ideal is to have separate fiber paths down separate streets, hopefully to separate central offices that then come in separate points of the building, and the only place they have in common is the data center,” he says. “Companies sometimes put in circuits from multiple carriers, but if they all come down the same bundled fiber line, and it comes in only one place in the building, that can be a problem if, say, a backhoe operator in an alley mistakenly cuts that line.”

Good disaster recovery planning also should focus on keeping individual network applications running, Lawson says. “Many times, what people think is an entire network going down is actually a problem with a particular application on that network like an e-mail server, customer relationship software, or instant messaging” he says. “Good disaster recovery is plotted around keeping each of those individual building blocks available to users.”

Once disaster recovery systems are in place, it’s important they be tested—a step some companies overlook in attempts to save money, Lawson says.

The VOIP Question

The growing use of voice over Internet protocol technology raises another important issue around protecting phone communication in the event of problems. VOIP, increasingly popular because of its low cost and portability, runs over broadband connections and is dependent on a computer’s electrical connection. Unlike a traditional phone, if the power goes out so does VOIP service. Some use generators as a backup, and smaller businesses typically depend on cell phones as option B in the event of VOIP outage.

But larger companies using true IP telephony systems require “fail over” systems in the event they lose VOIP connectivity. “That usually means having a couple of copper analog connections from a telecommunications company in each and every location if, for some reason, VOIP goes away,” Lawson says.