Servers need to have high availability and redundancy. A few methods and technologies allow that.

RAID10 (Redundant Array of Independent Disks) is a method of writing data to many disks for high performance.

1 - means mirroring, writing data to more than one disk at the same time. One fails, another survives.

0 - means striping, data is chunked into multiple disks, allowing it to be accessed simultaneously. This is not redundancy because you need data from all disks.

0 is performed before 1. Data is 'striped' (like a zebra) and then 'mirrored'. RAID10 typically uses at least 4 servers and storage capacity is halved. Four servers of 60GB each equals to only 120GB.

@spiceworks

RAID10 makes data highly available but it is not backup. Corrupt data is copied to both sides of the cluster. DR still needs to be implemented for backups. DR copies data to separate disks at different points in time to allow rollbacks to before the corruption happened.

Read about RAID10 here.