RAID-3 is one of the lesser RAID levels. While it isn't as obscure as RAID-4, or proprietary,
This isn't quite true. While RAID-3 has some serious drawbacks for general storage applications, it offers a fast, efficient method of storage in some specialized areas.
RAID-3 combines striping and parity, but it puts all the parity information on a separate parity disk. The result is an array that is especially poor at random writes and only so-so on random reads. This means RAID-3 is poorly suited for transactional applications.
Where RAID-3 shines is in sequential reads and, to a lesser extent, in sequential writes. In fact, it's actually faster than RAID-5 on sequential reads and writes, and nearly as fast as RAID-0 (striping), with the advantage of data protection. This means that in applications where large files comprise all of the workload, RAID-3 is a good choice.
Not surprisingly, RAID-3 is most common in applications like streaming media, graphics and video editing, in which the files are very large and the storage medium must keep up with system speed.
A RAID-3 array tolerates the loss of any single drive and most modern RAID-3 systems support hot spares and automatic rebuilding. While performance doesn't degrade much in the event of a single-disk failure, rebuilding the array can take hours, as the blocks are checked and the bad ones reconstructed.
RAID-3 requires at least three disks (two for data and one for parity) and at least a medium level controller.
Generally speaking, any application that reads and writes large files sequentially is a good candidate for RAID-3.
For more information:
Topics: RAID technology
Tip: RAID-10: Two great RAIDs that go great together
About the author: Rick Cook has been writing about mass storage since the days when the term meant an 80 K floppy disk. The computers he learned on used ferrite cores and magnetic drums. For the last twenty years, he has been a freelance writer specializing in storage and other computer issues.
This was first published in April 2005