Most businesses have some sort of RAID system in place, and in recent years, personal computer users have adopted the technology to achieve better fault tolerance and faster data access speeds. RAID technology improves on traditional hard drive storage significantly, especially in terms of dependability--many configurations can sustain more than one hard drive failure without losing data. This dependability comes at a fairly low cost, as hard drives are relatively inexpensive media.
However, RAID technology does not provide perfect protection from data loss, and some users put too much trust in the improved physical fault tolerance. If you use RAID or if you are considering a new RAID system for your home or office, remember the limitations of the technology.
Fault Tolerance and RAID Systems - While RAID technically stands for Redundant Array of Independent (or Inexpensive) Disks, some RAID configurations are not actually redundant. RAID 0 stripes data across several disks, but provides no fault tolerance. This leads to some confusion, and many personal computer users believe that data on RAID 0 has protection from a hard drive failure. When a hard drive failure occurs, these users lose data.
Even on tolerant systems, multiple media failures can occur. When a drive fails on a RAID 5, you need to replace the damaged member and rebuild the array. The rebuild process can put a tremendous strain on the other drives, occasionally causing a second hard drive failure and pushing the system past its point of redundancy.
Many businesses add fault tolerance by building nested RAID arrays that use several RAID levels. While problematic media failures are rare on these systems, controller failures can lead to parity damage and data loss in some circumstances. Some users also build nested RAIDs without adding hot spares. A hard drive failure can leave a nested RAID running in a degraded state, which puts working stress on the other members.
Failed Rebuilds - As mentioned above, RAID rebuilds can result in multiple hard drive failures. They can also cause data loss if the system rebuilds improperly. Failed rebuilds can occur due to power surges, blackouts and other electrical events.
The RAID administrator might also carry out rebuild procedures improperly. Professional data recovery services regularly receive RAID arrays that were accidentally reinitialized and extensively overwritten during rebuilds.
Corruption and Other Issues - Even if you use proper rebuild procedures, you may encounter file corruption due to bad sectors on one of the healthy members of the RAID. Most RAID systems have built-in features that protect against corruption, but it is still a common problem on rebuilt systems.
Data recovery engineers need to treat corrupt or damaged files using specialized utilities designed for striped and mirrored data, and you cannot reliably treat file corruption without these specialized utilities. Dozens of other logical issues can cause RAID data loss; operating system failures, software issues and malware attacks are all common sources of file damage.
The bottom line is that RAID systems require regular backup. No RAID is completely safe from logical data loss. While you can certainly use RAID as a part of your data loss prevention plan, you should always keep a secondary and tertiary copy of important files. Never use a RAID without a proper backup, and never keep all of your data storage devices in a single physical location.
Make a disaster recovery plan that maximizes system availability while limiting your chances of permanent file loss through incremental and full backups. With proper planning, you can use RAID technology to improve your storage capabilities while effectively minimizing your risks.