Data Centers

Incorporate problem applications in your DR plan

This column describes the problems inherent in some applications making it more difficult to plan for business continuity. Here are some things to watch out for.

How well can your organization deal with an emergency? Automatically sign up for our free Disaster Recovery newsletter, delivered each Tuesday, and make sure you're prepared for the next catastrophe.

When developing a business continuity plan, most organizations typically start with reviewing their applications and deciding how the company can best protect them. If the primary goal is to immediately resume operations at another location, organizations may soon discover that some systems just don't easily lend themselves to protection for one reason or another.

One reason for this could be due to licensing issues. If licensing only allows you to install software on one physical machine, this can create barriers to high availability (HA) solutions. While it can be technically possible to circumvent these protections, many software solutions' design prohibits such rule-breaking via different methods.

Some systems don't allow installation on more than a single server. While companies can often buy multiple software licenses, this approach can quickly become cost-prohibitive.

Other times, networking itself causes problems for system protection, a type of problem I've addressed in previous columns. For example, if an application only works when using a specific IP address, and you can't make that IP address appear at another data center, you're stuck with only one system that can run the software.

The same goes for solutions that require intense server-specific configuration outside the networking arena. You may find yourself forced to reconfigure a server at the time of failover.

As we've discussed in prior columns, many applications can pose incredible difficulties when it comes to either accessing data or resuming operation in another location. Protection via tape backup is generally an option, but keep in mind that it doesn't allow for immediate failover.

First, there's the issue of the time it takes to restore the data from the tape. In addition, you need to reinstall or reconfigure the applications in question. The end result is that you can't get the systems up and running immediately using only a tape backup—but you can bring them back up eventually.

So what can you do in these situations? Your best bet is to create backup hardware in your DR data center and configure it as closely as you can to the solution in the primary data center. Make sure you keep copies of the installation media for your applications in the DR data center, preferably in some form of fire-proof media safe.

In the same safe, you should also keep copies of the tapes that correspond to your recovery point objective. For example, if you can only afford to lose 24 hours of data, you need to keep a copy of last night's tape in the safe—not an easy feat but definitely possible.

By keeping your DR systems "warm," keeping the data close at hand, and being ready to restore, you can bring back even an "unprotectable" system (from an immediate availability standpoint) within a reasonable timeframe after a major disaster.

Editor's Picks

Free Newsletters, In your Inbox