Network Administration: Disaster Recovery Testing

Any disaster recovery plan is incomplete unless you test it to see whether it works. Testing doesn’t mean that you should burn your building down one day to see how long it takes you to get back up and running. You should, though, periodically simulate a disaster in order to prove to yourself and your staff that you can recover.

The most basic type of disaster recovery drill is a simple test of your network backup procedures. You should periodically attempt to restore key files from your backup tapes just to make sure that you can. You achieve several benefits by restoring files on a regular basis:

  • Tapes are unreliable. The only way to be sure that your tapes are working is to periodically restore files from them.

  • Backup programs are confusing to configure. I’ve seen people run backup jobs for years that don’t include all the data they think they’re backing up. Only when disaster strikes and they need to recover a key file do they discover that the file isn’t included in the backup.

  • Restoring files can be a little confusing, especially when you use a combination of normal and incremental or differential backups. Add to that the pressure of having the head of the company watching over your shoulder while you try to recover a lost file.

    If you regularly conduct file restore drills, you’ll familiarize yourself with the restore features of your backup software in a low-pressure situation. Then, you can easily restore files for real when the pressure’s on.

You can also conduct walkthroughs of more serious disaster scenarios. For example, you can set aside a day to walk through moving your entire staff to an alternate location. You can double-check that all the backup equipment, documents, and data are available as planned.

If something is missing, it’s better to find out now rather than while the fire department is still putting water on the last remaining hot spots in what used to be your office.