Do You Need A RAID?

serverrackNot the sudden attack of enemy troops, but the Redundant Array of Independent Disks (RAID) is a group of methods for storing large amounts of data redundantly and can make accessing the data very quick. Raid is a fancy term of linking a bunch of computer storage devices to form one big storage array and effectively combines all the disks into one big disk depending on the RAID configuration. The basic RAID configurations are 0, 1, 5, 6, and 10, Each RAID configuration storing data between the hard drives differently. RAID was originally developed when hard drives were much smaller with drives the size of a room only storing four Gigabytes. Over the years hard drives have gotten as big as 20 Terabytes and RAID has slowly trickled down to the consumers, but is RAID really necessary with all extra size of the newest drives? In the end, the biggest factor is the purpose of the data and the type of storage devices used.
When I refer to storage devices or hard drives, I mean any device that stores bits/bytes and is compatible with RAID configurations. In most cases, it is an HDD (Hard Disk Drive) or SSD (Solid State Drive), rarely both at one time. Both do not work well together because in order for most RAID configurations to work all of the hard drives have to be exactly the same in terms of speed and size for optimum efficiency. In more complex RAID configurations, such as RAID 5 and 6, I recommended using the same drives for the whole array. The only exception is RAID 0.
The first two types of RAID configurations are simple and straight forward. RAID 0 requires only two storage devices and links them together to form one big storage device, but if either one of the drives fail all of the information on the drive is lost and is impossible to recover, greatly increasing the read and write speeds and both disks can be utilized by storing all of the information in stripes of blocks on each hard drive. In addition, RAID 1 also only requires two storage devices and mirrors one of the drives to the other one, so if one drive were to fail the other could copy the information. The downside of RAID 1 is that half of the total storage capacity is lost because all of the data is copied twice, and the read and write speeds are not as fast as RAID 0 because it does not use the block technique.
Raid 5 and 6 are a little more complicated and harder to explain. They involve more striping of information and the use of parities. Similar to how RAID 0 stored lots of small blocks on each hard drive so it can utilize all hard drives when the processor requests it, RAID 5 does the same thing across all of the drives but also contains parities on each drive, so if one drive were to fail, it could be replaced. A parity is a datum that detects errors in drives and is used to restore the data along with data currently on the drives. The process of restoring any lost information takes hours and is not as fast as adding a new hard drive to RAID 1, but the parities take up less space than completely mirroring all of the data. RAID 5 also requires a minimum of 3 hard drives. Raid 6 is very similar to RAID 5 but has twice the amount of parity and requires a minimum of 4 hard drives, So RAID 6 can compensate for the loss of two drives but has the same long drive replacement time. When more drives are added to RAID 5 and 6 the drives speed up because the data can be spread into more stripes and the parity can be better spread out, slightly shortening the time it takes to recover a lost drive. Ultimately RAID 5 and 6 are the best compromise between speed and data safety minus the long drive recovery time.
Raid 10 is the addition of RAID 0 and 1. They require a minimum of 4 storage devices and stripes the data together across 2 of the drives as RAID 0 does. Then mirrors the first two drives to the other two hard drives, so if one of the drives failed then the information could be recovered. RAID 10 has very close read and write speeds as RAID 0, but the max potential storage is halved because half of the drives act like mirror for the other two in case one of the drives were to fail, for example four hard drives that each contain one terabyte of information named drive A, B, C and D are in a RAID, Even though the drives total capacity is four terabytes, the array could only store up to two terabytes of information, because drive A would be mirrored to C and drive B would be mirrored to D. Therefore, all of the information on any given drive would have a redundant backup on another drive, so if one were to fail it could be replaced. The drive recovery time is relatively low because all of the information is mirrored and not parodied.
Now the question of do you need a RAID? And for most people, the answer is no, unless you have lots and lots of really important data that you need saved or need access to mass amounts of data quickly. Many video editors use RAID 0 for editing because if a drive failed then the data they needed can be replicated relatively easily but may take some time due to the large files. The good thing is that most motherboards now support raid 0 and 1, and can be useful if you have a bunch of drives sitting around with no use for them, but they do not normally support RAID 5 and 6. In order to configure your disks in these RAID configuration, you would have to buy a RAID card which can cost anywhere from $500 to $1000 or more depending on the card and number of hard drives you want to connect to the card.

Lloyd, Chris. “This Is a RAID.” MaximumPC Dream Machine 2016 27 July 2010: 50-55. Print.

Photo Credit:
Sparling, Erin. “Monolith from the Front”, August 19 2007.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s