Continual importance of backups
RAID Recovery - Continual importance of backups
Almost all the RAID levels use striping which distributes data into small pieces across multiple drives in the array. It might improve performance but can be risky for data integrity for obvious reason. For example, if a file goes on a striped array, partial part of each file is bound to be lost besides identifying what exactly has been lost. A fault tolerant array with parity protection, to some extent, eliminates risk in failure of single disk, but RAID 0 cannot. As arrays have minimum hardware reliability than individual drives, some risks on RAID 0 are more comprehensive than users expect.
Mirroring, either single or with other techniques, ensures protection against data loss, especially hardware failure. There are multiple factors behind data loss.
- Unprecedented hard drive failure: The simultaneous failure of two mirrored drives is most common phenomenon. The chances become high when they are modelled alike.
- Support hardware failure: The other hardware that runs the entire array has the obvious chances to fail due to failure of power supplies or motherboard which may increase difficulties because of overheating. All these can cause outright failure of the array or gradual corruption in some extreme situations.
- Physical damage: Due to negligence or sabotage hardware can easily be damaged physically. This will harm all the hard drives in the array.
- Software problems: Software can contain bugs these bugs can potentially harm data.
- Viruses: A virus can destroy individual file or entire disk volume at the system level.
- Human errors: It is the most prominent reason of data loss because RAID cannot help against unintended removal of data.
It hasn’t covered all the commonly visible reasons of data loss but some other factors work behind that should be taken care of.
