RAID-controller and levels
Explanation of the RAID-Controllers and the specific RAID-Levels
RAID, (redundant array of independent disks) is a storage technology that combines multiple disk drives components into a logical unit.
Data is distributed across the drives in one of several ways called "RAID levels", depending on what level of redundancy and performance (via parallel communication) is required.
The operatation of a RAID system requires a minimum of two hard drives. The hard drives operate together and are joined in an array, that viewed from at least one aspekt is more efficient than a single hard drive.
With RAID systems you can gain following advantages:
- Increase the reliability (redundancy)
- Increase the transfer rate (performance)
- Setup of large logical drives
- Hard drive replacement and increasement of the system storage capacity during operation
The hard drive interaction is specified by the RAID levels.The most common RAID levels are RAID-0, RAID-1 and RAID-5.
RAID with hardware controller
Professional hardware RAID implementations contain a seperate embedded CPU, use a large additional cache and therefore offer highest transfer rates and reduce the main CPU load.
The addatitional buildt-in hardware RAID-Controller in a PC distributes the actual data of the RAID array to the controller conected hard drives.
Requires 2 hard drives
Provides half storage, data redudancy and normal transfer rates by distributing the same data onto both hard drives at the same time, virtually mirrored. Every hard drive contains identical data.
In case of a (complete) failure of one hard drive getting defective, the RAID controller can reconstruct the data of the missing hard drive. The defective hard drive is simply replaced by an identical one and the controller recovers the data automatically by itself.
By design RAID-1 is recommended for applications where high transfer rates is not required but data integrity is.
For example: By assigning 2 hard drives with 1TB of storage each you obtain 1TB of storage within your RAID-1 system.
Requires 2 hard drives
Provides double storage and improved transfer rates by dividing the data into associated blocks of the same size that are arranged virtually like a zipper, one block to drive 1, next block to drive 2. Thereby the data is accesed parallel by all involved drives via stripping.
In case of a (complete) failure of one hard drive getting defective, the RAID controller can not reconstruct the data of the missing hard drive without the destroyed hard drive.
By design RAID-0 is only recommended for applications with high transfer rates required, where data integrity is not required or managed by suitable data backup solution.
For example: By assigning 2 hard drives with 1TB of storage each you obtain 2TB of storage within your RAID-0 system.
Requires 3 hard drives
Provides improved transfer rates in reading, redundancy as well as more storage by dividing the data into associated blocks of the same size, like RAID-0. These are arranged virtually like a zipper onto drive 1 and drive 2 and the parity data of drive 1+2 onto drive 3. The next blocks are written onto drive 2 and drive 3 and the parity data of drive 2+3 onto drive 1 and so on.
In case of a (complete) failure of one hard drive getting defective, the RAID controller can reconstruct the data of the missing hard drive with the parity data stored on the remaining working hard drives.
By design RAID-5 is recommended for applications with higher transfer rates and data integrity is required.
For example: By assigning 3 hard drives with 1TB of storage each you obtain 2TB of storage within your RAID-5 system, 1TB is reservered for the parity data and can't be used.