Data integrity is important. A RAID array is a hardware and software combination that helps to ensure that your important data is never lost. Originally, they were hard to configure and expensive and thus were only for businesses. However, most desktop computer manufacturers have started to make creating an array on your home computer quick and easy.
The exact process on how to create an array differs depending on what type of array you want/need, what hardware you have, and, what kind of computer you have. I’m going to delve into the basics, first by describing the different types of arrays you can choose from and then I’ll share steps you’ll need to take, with links to more in-depth instructions. Finally, since no hardware/software configuration is infallible, I’ll give some information on how to recover an array, with links to more in-depth instructions.
What is a RAID Array?
A RAID array is a “Redundant Array of Independent Disks” that utilizes multiple disks to mirror your data. What this means is that instead of your data residing on just one disk and only being written to one disk, it resides and is written to multiple disks.
How many disks depends on the type of array in use. There are several types of RAID array that are used: RAID 0, RAID 1, and up to RAID 6. Some of these types can be combined for greater efficiency, speed, data safety and ease of use. I’m going to talk about the ones I find most common for home computers: RAID 0 and RAID1 RAID 5.
The Different Types of RAID Array
RAID 0 has less of a focus on making sure you “never” lose your data and more on how fast you can access that data. How it works is: With a RAID 0 setup, files get broken up into smaller blocks, with alternating blocks being written to different disks. Computer geeks like me call this striping because the data is written in stripes. It must be mentioned that this type of array provides little to no protection against data loss in the event of a drive failure.
That last sentence is why RAID 1 came into being. In this type of array, you still have multiple drives, but instead of the data being written in blocks across multiple drives to enhance access speed, the data is mirrored across all the drives equally. A great illustration of this is if you hold a mirror and look into a mirror the images appear to go on into infinity. RAID 1 arrays require two identical drives in order to work properly. RAID 1 is a great way to inexpensively ensure data integrity.
What if you want both increased drive speed and data redundancy? Easy. You choose RAID 5. Describing this type of array gets pretty geeky, so bear with me. Using this format, your data is striped and mirrored across the disks in the array. There are also little pieces of seemingly pointless pieces of information written to the drives that are called parity markers. This parity data allows the controller to reconstruct lost data in the event of a drive failure using a specialized mathematical algorithm. However, the parity data typically uses about a third of the available drive space, so, if you have a one terabyte array (meaning two one terabyte drives), you’ll only have about 600 gigabytes of usable space to store your data. Additionally, unlike the other 0 and 1 arrays, you need at least three identical drives.
Data Integrity: How to Create a RAID Array
Obviously you need at least two identically sized (preferably same maker and model number) drives. After that, things get tricky. Newer desktop computers come with software for RAID controllers built into what’s known as the BIOS (Basic Input Output System — the set of commands that tells the computer how to use input and output devices). REMEMBER! You must configure the BIOS for RAID BEFORE you install the operating system!
You could also choose to install your operating system on a single internal drive and then use an external RAID controller to create your array. I’ve been using the HighPoint RocketRAID controller (and its predecessors) quite successfully for a few years for customers with computers that don’t have RAID support built-in for a few reasons. First, it’s not very expensive. Second, it’s compatible with every system I’ve tried it in, and third, it’s easy to use and maintain and gives me a few array types to choose from.
Typically, configuring the array is simply a matter of enabling the controller in BIOS and then selecting the drives to use. For a more in-depth description of the different types of arrays and how to set them up, I highly recommend this PCWorld article and this Hardware Secrets piece. They both go into more detail than I have space for.
How to Recover a RAID 0 Array
This is a rather lengthy process, so I’m going to mention the ways a RAID 0 array can fail and then link you to in-depth descriptions on how to recover them. The most common failure point in an array is one of the drives. The other is euphemistically called “non-drive-related RAID failures.” These include failures induced by operator error, software failure, or controller hardware failure. I use this site when I work on a RAID array using RAID 0. DiskInternals also has a program that allows a single use trial that works very well and gives you multiple options for recovering your data.
RAID 1 Recovery
The RAID 1 array is the easiest to recover because of that parity data I mentioned earlier. The controller itself can use the parity markers to rebuild your data once you replace the failed drive. However, if the failure is in the controller itself, you’re most likely going to need to replace the controller and maybe even rebuild the array.