Which of the following is TRUE about RAID 5 ?

RAID levels differ in how they combine striping, mirroring, and parity to trade off performance, capacity, and fault tolerance. RAID 5 specifically uses block-level striping with parity computed via XOR, and that parity is rotated across every disk in the array rather than fixed to one disk.

In a RAID 5 array of N disks, each stripe writes data blocks to N-1 disks and a parity block to the remaining disk; which physical disk holds the parity block rotates from one stripe to the next, so over time every disk carries both data and parity blocks roughly equally. This design needs a minimum of three disks, and if any single disk is lost the array reconstructs the missing data by XOR-ing the surviving data and parity blocks -- so RAID 5 tolerates exactly one simultaneous disk failure.

Each of the other statements instead describes a different RAID configuration:

• A dedicated parity disk (all parity on one fixed drive) is the architecture used by RAID 3/RAID 4; that fixed drive becomes a write bottleneck, which RAID 5's rotation avoids.

• Zero fault tolerance (any single disk loss destroys the array) describes RAID 0, which stripes data with no parity or mirror copy at all.

• A two-disk minimum corresponds to RAID 1 (mirroring) or a two-disk RAID 0 stripe; a parity-based scheme needs at least three disks so that stripe capacity remains for parity alongside data.

So the statement describing parity spread across every disk in the array is the one that matches RAID 5's actual design.