Sunday, April 17, 2011

Tutorial 4: What is RAID?? Seriously...

That's how I felt when I first started learning about them. Turns out, they're not as bad as they seem at first. RAID stands for "Redundant Array of Inexpensive Drives". This is basically a way of  combining the storage across multiple drives to offer a single virtual device. RAID can provide the following benefits: Data redundancy (protects important data from failures), Fault Tolerance (provides a more reliable system), Increased Capacity (combines several disks together into one), Increased Performance (subjective to the type of RAID is used, overall the technology is the better option). These are the main types of RAID...

RAID 0: Also known as "Striping", the read/write operations are executed from the various drives simultaneously. Meaning, your data is split among different disks for storage. The total amount of storage available for use is the sum of all disk's memory. This translates into faster writing/reading times because of the simple fact that it can execute from several disks at once, as supposed to one for a big file. RAID 0, is the fastest and most efficient type of RAID. However, it does not offer any fault tolerance, simply because there is no extra disk where you may backup your data. This is a great choice for someone that backs up their data on an external hard drive often. Below we have RAID 0, striping.

Credit to: www.Iomega.com tutorial team.


RAID 1: Also called, Mirroring, is a two disk system that allows redundancy. The structure offers redundancy (as supposed to RAID 0) because, the data is written fully on BOTH disks. This is great for environments where, cost is not an issue but data loss is a big concern. Although, cost is 50% higher than Striping (you need double GB to store your data, twice), it's great when handling critical data. You may also add what is called a "Hot Spare" disk which, is an extra disk that is only used when one of the others (with data) fails. In the case of said disk failure, "Hot Spare" disk copies all the data from the "good" disk to create a second copy. When the failed disk is exchanged, it becomes the new "Hot Spare". Here is what RAID 1 looks like...
Credit to: www.Iomega.com tutorial team.
RAID 5: Is a redundant system, which is similar to striping but distributes parity among the drives. RAID 5 requires three or more drives as it spreads data blocks across all disks except for one, "Hot Spare". This system has relatively fast reads similar to striping. however, due to having to compute parity, writing times are a bit slower than previous RAID. In addition to fast read times, this provides data redundancy. If a single disk fails, due to parity, "how spare" can easily reconstruct the data lost. Here is what it may look like..
Credit to: www.Iomega.Com Tutorial Team
So now you have the main idea of what RAID is what their advantages are. If you're in the market for a RAID system, I would advice you to do extensive research on which is more appropriate to your environment. Here are some websites I find useful: Addonics RAID quick tutorial & Publib Boulder's RAID Journal.


No comments:

Post a Comment