RAID allows you to combine multiple physical hard drives into a single logical hard drive. This allows you to mirror your data across two hard drives, ensuring you always have your important data stored in multiple places.
RAID stands for “redundant array of independent disks,” although there’s a type of RAID that provides no redundancy and only increases performance.
RAID isn’t just a single way of combining disks. There are multiple RAID levels that provide different levels of performance and redundancy. All RAID levels have one thing in common: they combine multiple physical disks into a single logical disk that is presented to the operating system.
- RAID 0: Unlike other RAID levels, RAID 0 provides no redundancy. However, RAID 0 allows you to increase performance using multiple disks. When you use RAID 0, data your computer writes to a hard disk is split across two (or more) hard drives evenly. For example, if your computer writes a 100MB file, 50MB will be written to one hard drive and 50MB will be written to the other hard drive. When the computer needs to read the file back, it can read 50MB from one hard drive and 50MB from the other hard drive at the same time — this will be faster than reading 100MB from a single hard drive. However, if any of the hard drives in the RAID array dies, you’ll lose your data. When you use RAID 0, your multiple disks appear to be a larger and faster hard disk — but they’re much more fragile.
- RAID 1: In RAID 1, two disks are configured to mirror each other. When your computer writes 100MB of data to its disks, it will write the same 100MB to both hard disks. Each disk contains a complete copy of the data. This ensures that, if one of the disks ever fails, you will always have a complete, up-to-date copy of your data.
- RAID 2, 3, and 4: These RAID levels are little-used and often considered obsolete.
- RAID 5: To use RAID 5, you will need at least three disks. RAID 5 uses striping to divide data across all hard drives, with additional parity data divided across all disks. If one of the hard drives dies, you won’t lose any of your data. RAID 5 offers data redundancy with less storage cost than RAID 1 — for example, if you had four 1TB hard drives, you could create two separate RAID 1 arrays (1TB each for a total of 2TB storage space) or a single RAID 5 array with 3TB of storage space.
- RAID 6: RAID 6 is similar to RAID 5, but adds an additional parity block, writing two parity blocks for each bit of data striped across the disks. You lose storage capacity, but RAID 6 provides additional protection from data loss. For example, if two hard drives die in a RAID 5 configuration, you’ll lose your data. If two hard drives die in a RAID 6 configuration, you’ll still have all your data.
- RAID 10: Also known as RAID 1+0, RAID 10 divides data between primary disks and mirrors this data to secondary disks. In this way, it attempts to provide the advantages of RAID 0 (dividing data across multiple disks for a performance increase) with the advantages of RAID 1 (redundancy).
There are also other, non-standard RAID levels.
RAID is generally used on servers, mainframes, and other computer systems where having redundantly stored data is important. RAID isn’t used on desktop computers and laptops as often, but many computers ship with RAID controllers. If you wanted to, you could likely set up a RAID 1 configuration with two drives to ensure your data is mirrored across two drives.
When using RAID, you can either use “hardware RAID” or “software RAID.” With hardware RAID, a hardware device in your computer does all the RAID work. For example, if you had hardware RAID and set up two disks to function in a RAID 1 configuration, the hardware RAID controller would present the two disks to your operating system as a single disk. All the work of the RAID — mirroring the data, dividing it across hard disks, and so on — is handled by the hardware RAID controller. Your operating system would have no idea you were actually using a RAID.
With software RAID, the work is handled by the operating system. For example, you can create a software RAID while installing Linux on your computer — the Linux kernel knows about the RAID and will do the work itself without any special hardware necessary. You can also create a software RAID in Windows.
To configure a hardware RAID, you’ll need to use the software that controls the RAID controller — this may be accessed through a computer’s BIOS. You should check your hardware RAID controller’s documentation for exact steps if you’re doing this.
Popular operating systems have technologies that function similarly to RAID. Windows 8 introduced Storage Spaces. Linux has the logical volume manager, or LVM. Both technologies allow you to group several physical disks into a single logical disk to mirror your data for redundancy or pool your disks’ storage, making it available as a single disk without providing redundancy.
These technologies may seem a bit complicated, but they’re actually ways of simplifying things. Once you’ve set up the appropriate RAID, your data will be automatically stored across multiple hard disks so you don’t have to worry about losing it. Your software doesn’t need to even know that the RAID exists.
Chris Hoffman is a technology writer and all-around computer geek. He's as at home using the Linux terminal as he is digging into the Windows registry. Connect with him on Google+.
- Published 05/7/13