Sooner or later, your equipment will fail, an thus result in a loss of data. Implementing fault tolerance is a good way of preventing data loss.
Last week I made a post about backups regarding data loss. This week it’s all about RAID systems.
RAID is an acronym for redundant array of independent disks. It allows servers to have more than one hard drive so that if the main hard drive fails, the system keeps working. Here is a quick glimpse of some array levels:
|RAID 0||On this level, the data is distributed across multiple disks, which gives you more speed. But doesn’t offer fault tolerance. A minimum of two disks is required.|
|RAID 1||This level introduces fault tolerance as it mirrors the contents of the disks. This means that 50% of your disk space is for mirroring the other 50%. If the main drive fails, the system can still run on the backup drive. You’ll need a minimum of two disks here, too.|
|RAID 3, 4||Three or more disks are combined with the data split across them. This RAID also uses one disk to store parity information. If a disk fails, that is only partial loss of data. And the parity info+ the data that isn’t lost, allow to restore the data.|
|RAID 5||Similar to RAID 3. But the parity is distributed across the drive array. This way you don’t use up a whole disk for the parity.|
|RAID 6||Combines four or more disks on a way that protects data against the loss of any two disks. It’s basically an additional parity block to RAID 5. However the parity block is distributed across the drive array.|
|RAID 1+0||A mirrored data set, which is then striped. You’ll need four drives for this one: two mirrored drives to hold half of the striped data, plus another two mirrored drives for the other half of the data.|
|RAID 0+1||The opposite of RAID1+0: Two mirrored drives replicate the data on the RAID 0 array.|
However, it is still considered to make a backup of the drives.