
Correlated failures The theory behind the error correction in RAID assumes that failures of drives are independent. Given these assumptions it is possible to calculate how often they can fail and to arrange the array to make data loss arbitrarily improbable. In practice, the drives are often the same ages, with similar wear. Since many [...]
















