RAID

One disk is a single point of failure: when it dies, your data dies with it, and one disk's speed caps how fast you can read. RAID — Redundant Array of Independent Disks — is the idea of ganging several cheap disks together and making them behave like one better disk: faster, bigger, more reliable, or some combination. The trick is to spread data across the disks cleverly, sometimes with extra redundant copies or parity, so that the array survives a disk failure and can serve multiple disks at once. Think of it as a team of mules sharing a load, with a spare or two in case one stumbles.

RAID comes in numbered levels with different trade-offs. RAID 0 (striping) splits data across disks for speed and capacity but adds no redundancy — lose any disk and you lose everything; it is the dangerous one. RAID 1 (mirroring) keeps an identical copy on a second disk: it survives a failure and reads fast, but you pay double the disks for the same capacity. RAID 5 stripes data plus one block of parity across the disks so that the array can rebuild any single lost disk from the others — good capacity efficiency, survives one failure, but writes are slower (parity must be recomputed) and rebuilding a failed disk is stressful and slow. RAID 6 adds a second parity to survive two simultaneous failures.

Why it matters and where beginners go wrong: RAID buys availability (the system keeps running through a disk failure), not a backup. RAID does not protect against the file you deleted by accident, the ransomware that encrypted everything, or the building that burned down — those hit all disks at once, and a deletion is faithfully mirrored. The famous slogan is 'RAID is not a backup'. RAID is the storage embodiment of dependability through redundancy: it turns a likely fault (one disk failing) into a non-event, but you still need real backups for the failures redundancy cannot catch.