RAID(Redundant Array of Independent Disks)

Common levels include RAID 0 (striping, no redundancy), RAID 1 (mirroring), RAID 5 (single-parity), RAID 6 (double-parity) and RAID 10 (mirrored stripes). RAID protects against hardware failure and improves uptime, but it is not a backup: it won't help against accidental deletion, ransomware or fire. Use our RAID calculator to see usable capacity and fault tolerance.

The key trade-off is redundancy versus usable capacity: more parity means you can survive more simultaneous drive failures but sacrifice more space. With today's large drives, rebuild (resilver) times after a failure can stretch to many hours or days, during which a second failure would be catastrophic — which is why double parity (RAID 6 / RAID-Z2) is recommended for larger arrays rather than single parity. On ZFS-based systems, the equivalent RAID-Z levels also add block-level checksums that detect and repair silent corruption, something traditional hardware RAID cannot do.