RAID 6 is also important to consider as denser, higher capacity disk drives are used, such as 1 Terabyte (TB) devices. Because these latest drives have so much capacity, the time to rebuild the data after a failure is greatly increased over less dense, lower capacity drives. The risk then becomes that there is a higher likelihood that while one drive is rebuilding after a failure, a second could conceivably fail during the long rebuild process, and all data be lost. RAID 6 eliminates this risk.
RAID 6 will thus grow in importance in the future, but is secondary to RAID 5 for physical security applications today.
RAID 10 is a newer standard, which like both RAID 0 and 6 was not one of the original RAID techniques. As in RAID 1, mirrors are created, and then additionally a RAID 0 stripe is applied to all of the drives, linking them together.
RAID 10 is very fast, and offers good drive failure tolerance, but requires double the storage to create the mirror as without. Thus, it is also twice as expensive per TB as RAID 5. While found in some large capacity, high performance IT systems, and useful for very large video surveillance storage requirements, RAID 10 is not common or required in most smaller implementations today.
Like RAID 6, RAID 10 will grow in importance in the future as larger storage deployments become more common for physical security implementations.
This simple Powepoint file/chart demonstrates how the different RAID categories write the data [download the RAID Chart]
There Are More?
While these are the basic RAID techniques, there are other RAID levels that individual IT manufacturers sometimes promote. As well, there are some newer, double-digit levels that are being adopted by the largest IT systems for speed and fault tolerance or to address scalability issues. Since they are often vendor-specific and largely not appropriate for most of today's physical security applications, these additional RAID levels (like vendor-specific RAID 7) will not be discussed here.
So now you know that RAID isn't all that complicated a concept to understand, and a couple of its techniques offer real advantages to the physical security industry for video surveillance and other data storage needs.
Today, many DVRs do not offer RAID protection, and as a result non-redundant systems may leave your recorded data open to complete loss. Fortunately this can be easily avoided. A subset of available DVRs offer RAID 1 and the best also feature RAID 5. For most physical security installations today, RAID 1 offers great protection and performance, but is often too expensive to deploy effectively.
For most video surveillance deployments, RAID 5 is usually the best choice, with its strong mix of affordable protection and good performance.
If your DVR vendor doesn't offer RAID 5, or you need additional capacity but don't want to replace those expensive DVRs you just paid for, you may want to consider adding an inexpensive external IP storage solution that will increase both reliability and performance of all of your video surveillance systems. Today, there are external IP solutions which can be connected in just a few minutes to popular brands of DVRs without noticeable disruption, and will continue to support any mix of analog and digital cameras as before.
Finally, ask your vendor about the type of drives that they use in their DVR systems. To lessen the frequency of disk drive failures, the best storage solutions use a higher grade disk drive than is typically found in PCs and some DVRs. These "server grade" drives often cost slightly more, but are required to pass more stringent manufacturer tests than the drives that are found in most PCs and many DVRs.
IP offers many advantages over tape and the proprietary, closed systems of our industry's past. IP can be less costly, yet much more capable. By using well designed IP systems with RAID, and ensuring that you are using server-grade drives within redundant storage systems, you can enjoy all of the benefits of IP, worry free.