Being Prepared for Disaster

Performing regular backups is only part of a good disaster recovery plan. You must also test your plans and be prepared for when disaster does strike. We generally recommend the following practices:

Using recovery servers A computer that you have prepared to take the place of a failed Exchange server is called a recovery server (sometimes referred to as a standby server or cold server ). A recovery server would have Windows Server 2003 and Exchange Server 2003 already installed. It should also have the Exchange database files loaded. At the least, the database files should be loaded on media that would facilitate loading on that server. Some organizations create a batch file that stops the Exchange services on the primary server and copies the relevant files to the standby server. These files could also be copied to removable media and then uploaded to the standby server when needed. Recovery servers are also useful for restoring single mailboxes.

Validating your backups It is important to validate that each backup occurs without errors. Check the Windows Backup Log and the Windows Event Log to ensure your backup has completed as scheduled.

Documenting backups It is important to document your backup strategy and provide step-by-step instructions describing how to use backups to restore data. You should also keep track of old backup logs in case you need to refer to them.

Running fire drills You should schedule regular fire drills in which administrators are required to restore a failed server from backup under different scenarios. This process accomplishes two things. First, it helps train administrators on how to perform restore procedures. Second, it helps verify that your backup media and plans really work.