Plan for the worst, hope for the best: backup and disaster recovery

Computer Technology Review, March, 2004 by Morgan Edwards

To backup or not to backup, is not the question. Everyone in business today understands the importance of performing periodic backups; however, a good question to ask is if the backups being performed are with disaster prevention in mind? Disaster on a computer system can come in many forms from a corrupt file, to a failed disk, to a fire in the computer room.

Disaster Recovery: The Ultimate Goal

The wrong way to find out whether your backup strategy is providing adequate disaster recovery protection is to experience a failure and find out that you can't recover the lost or damaged files from your last backup or, in some extreme cases, any backup at all. For a small percentage of the readers of this article this will be deja vu, while for others lightening will indeed strike sometime in the future. Like death and taxes, computer failures are going to happen--with potentially disastrous results unless an adequate prevention plan is implemented and practiced religiously.

Disaster Prevention: Implement a Plan

While software and hardware lay the foundation for recovering from the myriad of computer problems that can occur in any business, a good backup plan involves much more. As a producer of backup software with thousands of users. I've heard many stories of woe that were entirely preventable. You do not have to hire a Big Eight consulting company to develop a good backup plan.

What Does it Take?

The time, effort, and expenditure required to develop and implement an effective backup strategy will always pay off in the long run. To prevent a disaster from happening, the first step is to buy and install the necessary hardware and software. The next step is to thoroughly train the operators (notice plural). If the only operator who knows the software is on vacation when your primary data disk crashes, who's going to recover it and get the system going again?

Training the operators to schedule backups and perform restores is only part of the story. Operators must be vigilant, conscientious and dedicated in their duties of administering backups and verifying their successful completion. Furthermore, keeping a log is a good idea with an entry made for each backup. Any anomalies should be noted in the log, along with detailed information about tape use and cleaning (assuming you are using tape). A tape drive can fail in many different ways, and often the process begins with a little hiccup here and there before failing completely. A log can also help sort out whether bad media could be the source of a backup failure, or maybe dirty tape heads on the drive.

So far, I've talked about basic hardware and software, and implementing a backup plan. Other elements of a good backup plan include replacing or retiring media on a systematic basis, performing sampling restores periodically, practicing off-site media storage, and documenting the entire backup and restore procedure so that a part-time or new operator would be able to perform a restore in an emergency. Training part-time operators is particularly important when emergency restores must be performed after hours or over a weekend, when the regular operator just isn't available.

The one thing to always remember is that a good backup plan should have a contingency for every type of computer problem that can occur. Plan for the worst and hope for the best!

Morgan Edwards is CEO of UltraBac Software (Bellevue, WA)

www.ultrabac.com

This is the first in a series of columns by Morgan Edwards addressing Backup and Disaster Recovery.