Beep Beep Planning Backups

Beep! Beep! Planning Backups

One of the most important tasks that you can perform is to plan your backup. Some of the best backup systems and methods can still fail, so it's smart to always test them first.

Store backup tapes off-site

It's not uncommon for companies to store their backup tapes in a fire-proof safe usually stored in the computer room where the fire might occur. These safes are great for one thing, and one thing only: paper. They're designed to withstand enough heat so paper does not burn. This heat, however, is more than sufficient to melt any tapes that may be stored in them. For this reason, smarter companies store their tapes off-site on a daily, weekly, or monthly basis, depending on how often data changes. Storing data off-site ensures data availability in case of a disaster, such as a fire or a flood. If such a disaster were to take place, a new system could be installed and data could be restored.

There are companies that provide off-site data vaulting for a fee. The tapes are picked up periodically and transferred to an off-site safe facility for vaulting. If needed, customers can call to get the required tapes for a restore. Such companies might have different service contracts. Although it's more costly to have this service 24 hours and 7 days a week, spending the extra funds to ensure access to tapes might be justifiable, depending on the operation and how critical the data is.

Document your hardware and its settings

It's especially important to document your hardware and its setting because it saves time that would be spent guessing and trying different configurations to get the system back up and ready for data to be restored from tape. Document all important settings, such as the size of volumes or drives , the type of file system used, and whether the file systems are File Allocation Table (FAT) or NT File System (NTFS). If you've been using any kind of fault tolerance on your disk subsystem, such as mirroring or some type of Redundant Array of Inexpensive Disks (RAID), it notes these details. Generally, accurate and comprehensive documentation ensures fast and reliable recovery.

You should also have a list of the part numbers of all the components installed in your systems. Include a list of all the vendors and your contacts as well. If the components in your system are from one of the large server manufacturers, such as Compaq or Hewlett Packard, they might loan you a system to recover your data. Finally, have all your device drivers readily available.

Practice disaster recovery for your system

It's wise to practice disaster recovery periodically to ensure its success whenever it's required after a real disaster. In large companies, it's normal practice and, in fact, mandated by management. Whereas it's easy to make this statement, actually implementing this practice is harder than you may think. The problem is that you can't attempt to recover your data on a production system (which is a system already used for another function). If you do and the backup is corrupt, your live data will become corrupt as well. Many organizations purchase a second system with the same specifications as the production system as a backup. These systems tend to get "borrowed" for other pilot projects and come into production.

Windows Server 2003 has a new feature that makes disaster recovery much easier. This feature is similar to a third-party solution in which a total system recovery is possible with a few diskettes. Search for "disaster recovery" in the Windows 2003 Help and Support Center for more information.