I’ve seen many creative Microsoft Exchange Server backup and recovery methods that companies use to protect Exchange data. Some of these Exchange Server backup techniques worked well, while others presented disaster recovery challenges. Finding best practices for Exchange Server backup is easy, so here are five Exchange Server worst practices that you may not have considered.

1. Not backing up Exchange Server

Not having an Exchange Server backup not only leaves the obvious potential for data loss, but transaction log files continuously accumulate data until the backup process purges them. If you aren’t running regular Exchange Server backups, there’s a good chance that your transaction logs will eventually fill up their available storage volume.

2.Backing up Exchange Server to a file

There is nothing wrong with backing up Exchange Server to a file instead of tape as long as you have a removable backup on hand that you can store off site in case of fire, hurricane or other disaster. However, backing up Exchange Server to a file is a worst practice, depending on how you do it

For example, I did a consulting job for an organization that was having problems with their Exchange server. When asked about their backup practices, they assured me that they backed up their Exchange server religiously. What I didn’t discover until later, though, was that they were backing up to a file rather than to a tape. While this in and of itself isn’t a problem, they were storing the backup file on the same volume as their Exchange information store. Had that volume failed, they would have lost both Exchange information store and backup data.

If you’re going to back up Exchange Server to a file, it’s best to write the file to removable media or to a network server. Be sure that you don’t rob the network of bandwidth when you create the backup

3. Depending on alternative Exchange Server backup methods

I absolutely love some of the alternative backup techniques that are out there, such as disk-to-disk backups and continuous replication. Although these backup methods offer benefits that were unheard of a few years ago, they shouldn’t be used as complete substitutes for tape backup.

The biggest problem with disk-based backup solutions is that if the building were to catch fire, you would lose the backup Exchange server as well as the server that it was protecting. In addition, some of these backup solutions don’t let you perform a point-in-time recovery. This means that if your Exchange organization were attacked by a virus, you may not be able to restore the Exchange server to the state that it existed prior to the infection.

4. Relying solely on NTBACKUP

Using NTBACKUP is fine. However, NTBACKUP has a bug that makes it impossible to restore extremely large backups.

Although Microsoft doesn’t document this bug, there are countless posts about it on the Internet.

5. Not testing Exchange Server backups

I had a client who had a rather large Exchange Server deployment. As you would expect from an enterprise-class organization, the company had the best backup software and hardware that money could buy. At some point, minor corruption occurred within the Exchange information store. This corruption continued over time until it finally caused the database to fail.

By the time the Exchange Server failure occurred, the corruption was too extensive to repair. The only option was to restore a backup. Unfortunately, the organization had been backing up corrupt data for months and didn’t know it. It was impossible to restore any of the backups. Had this organization occasionally tested its Exchange backups, the organization could have found the problem early and prevented significant data loss.