Arthur Cole spoke with Firooz Ghanbarzadeh, director of technology services and solutions, CDW Corp.

Cole: Recent research indicates that many large and small enterprises recognize both the vital importance of disaster recovery and business continuity and the inadequacy of their own policies and systems. What do you think needs to happen to push DR higher on the enterprise agenda?
Ghanbarzadeh: Unfortunately, many businesses don’t think deeply about disaster recovery planning until they’ve actually experienced a disaster or have had a close call. However, the question that could precede this one is: “How many enterprises believe they have a disaster recovery plan, but really don’t?” Many have only enough elements of DR in place to kid themselves that they’re ready.

Ultimately, DR planning and systems are insurance, and every organization should consider its tolerance for risk to decide how important DR is to the resilience of their business. In a nutshell – how lucky do they feel? What are the probabilities of data or IT system loss, and what would be the financial impact on them, considering their specific business?

The best way for IT to focus more attention on DR is to conduct regular testing of the systems in place, just as other types of organizations conduct drills and rehearsals for disasters. That helps to find gaps and either fix them or provide the basis to escalate issues for management attention. It’s fine to talk in generalities about the value of DR planning, but regular testing and management review puts it in your own organization’s context.

Cole: How exactly does virtualization help improve business continuity operations?
Ghanbarzadeh: There’s no question: Virtualization provides for quicker recovery from a disruptive event because it allows replication of the production environment at lesser cost, and it is simply faster to boot or reboot a virtualized server than to cold boot a hardware server. Plus, the reboot can be done remotely.

Furthermore, if the situation requires creation of a server, that is much faster to accomplish in a virtualized environment. In any circumstance where time to recovery is critical, virtualization offers a distinct advantage.

Cole: What other critical elements are needed for a workable DR environment?
Ghanbarzadeh: First, it’s important to recognize that technology is the enabler for disaster recovery, but the most essential ingredient is “human software” – a solid business plan and process in place, where everyone understands when and how to activate the DR facilities and then switch back to the primary systems. All of the redundancies an IT department can build will be pointless unless the team knows what to do with them.

Beyond the business plan, critical technology elements include provision for:

Power – Secure enough backup to at least enable a smooth shutdown or fail-over, and in some cases enough alternate source to maintain critical systems throughout an event.

Communication links – Be certain that you have adequately separated, redundant physical cable connections from your carrier or carriers. Just having multiple carriers may not guarantee that, because some use common routes in various locations.

Data center and infrastructure – Provide replication of core systems and data, with physical separation that will ensure availability if the primary systems are destroyed or disabled.

User connection – Many DR plans overlook dislocation or isolation of system users – e.g., branch offices, remotely located individuals or even separate buildings within one campus – from otherwise undisturbed core systems. If critical to the business, provide alternate locations or means of access in case a disruption cuts them off from their primary access.