- RAID provides data protection solutions for Linux storage as it provide a reliable, redundant means of protecting vital data on the system.
- Is aims to improve disk performance and provide data redundancy.
- Through RAID information is spread among various physical hard drives using technique such as : disk striping and disk mirroring to achieve redundancy, lower latency and higher bandwidth for reading or writing, and recover ability from hard-disk crashes.
- There are two types of RAID
- Software based
- It offer the cheapest possible solution that abstracts multiple devices and provide a single virtual device.
- Its performance depends on server CPU and load.
- Hardware based
- This manage RAID subsystem and present to host only .
- Most dedicated server use hardware based RAID.
- Software based
- There are six RAID levels that can be used.
- RAID-0 :- It is most efficient array type which result in higer data throughput but offers no fault tolerance. It requires at least 2 disks and uses method call stripping that write across both drives.
- RAID-1 :- It provide redundancy, faster performance no data loss and fault tolerance. It requires at least 2 disks and uses method called mirroring that write data in both drives. Each drive is exact mirror of other drive and can be used if either drive fails.
- RAID-2 :- It uses hamming error correction codes which facilitates drive which do not have built in error detection. Now days this is not in much use.
- RAID-3 :- It does not allow multiple I/O operations to run simultaneously. It is used in single user environment.
- RAID-4 :- it does not support multiple simultaneous operations. It strips data across block level across several drives. It have very good reading speed but slow writing speed.
- RAID-5 :- It used in multi-user environments. It required at least 3 to 5 disks. It can write faster but reading speed is slower.