A tested backup strategy should be an integral part of a healthy data security system.
Computers and computer components can and do fail; multiple disk failures in RAID configurations, data center catastrophes (no matter how small) and virus infections can take down your system and corrupt critical data.
There is usually no warning before it's too late. Suddenly data is unrecoverable.
Data loss can be costly and may impact productivity. There is a large amount of investment into the data on a user hard drive. Some data value is obvious, such as company financial data or employee records, which would be disastrous to lose. Consider the man-hours reinstalling the base software (operating system) on your computer and reconfiguring your environment to suit your preferences again. In some instances this could easily cost you several days of productivity. In extreme cases, some data may not be recoverable.
Implementing a backup plan will often turn several days of lost productivity and weeks of reorganizing into an hour of restoring your disk image. Depending on the importance and value of your data, storing a regular backup securely off site may be prudent, in case of a disaster or catastrophe.
Backups can be local to tape or other media, or remote to a data backup service provider.
It is especially important to backup data which is held on storage systems or RAID arrays, as these tend to be used by many people in a corporate environment. Due to the large hard drive capacity on a storage system, some lost data may be impossible to recover, unless a backup is performed on a regular basis.
Setting up a RAID array is no substitute for a regular backup. Depending on the RAID configuration, data may still be lost if one or more drives in the RAID array fail.
The real purpose of a RAID array is not data backup at all. RAID was created to increase the speed of accessing files (data throughput), and increase the logical size of a "hard drive". Given that hard disk drives are mechanical devices that could fail at any time, RAID evolved to include some built in redundancy (e.g. RAID 1 and 5). Redundancy provides the ability to survive the loss of a single hard drive, and replace the failed drive without a system crash. In theory this is good, but does not provide for data viability in the event of two drive failures, or two drive data corruption. DO NOT confuse hard drive redundancy with a true backup.