Field Guide · concept

Also known as: Redundant Array of Independent Disks

RAID (Redundant Array of Independent Disks) combines multiple physical drives into one logical unit to gain redundancy, performance, or both — so data can survive a disk failure or be read and written faster.1

RAID 0 · stripe A1 A3 A5 A2 A4 A6 fast · no redundancy RAID 1 · mirror A1 A2 A3 A1 A2 A3 identical copy RAID 5 · parity A1 B1 P A2 P C1 P B2 C2 rebuild any one disk RAID guards against a drive failure — it is not a backup
Three building blocks, mixed into the numbered levels. RAID 0 stripes for speed but loses everything if a disk dies; RAID 1 keeps a full mirror; RAID 5 rotates parity so any single disk can be rebuilt. None of them replaces a backup — they don't protect against deletion, corruption, or disaster.

Overview

RAID uses three basic techniques, mixed in different levels. Striping spreads data across drives for speed (RAID 0, no redundancy). Mirroring keeps identical copies so either drive can fail (RAID 1). Parity stores recovery information that lets the array rebuild a lost drive (RAID 5 tolerates one failure, RAID 6 tolerates two). Combined levels like RAID 10 mirror and stripe together. A key caveat: RAID protects against drive failure, not against deletion, corruption, or disaster — it is not a backup.

Where it fits

RAID is the foundation of reliable data storage in network-attached storage, servers, and the data center, and it underpins high availability at the disk level. Built from ordinary hard disk drives or SSDs, a small mirror is a cheap way to keep a GopherTrunk archive of decoded calls alive through a single disk failure.

Sources

  1. RAID — Wikipedia, on RAID levels and techniques. ↩

See also