Erasure coding is a data-protection method that breaks data into fragments, computes parity fragments, and spreads them across many drives or nodes so the original can be rebuilt even if several fail. It delivers durability comparable to replication while using far less capacity overhead. In large object and scale-out storage systems it is the standard way to protect petabytes cost-effectively against drive and node failures.
Category: IT & Compute Infrastructure · Subcategory: Storage & Virtualization · Part of the AIDC Technical Terminology reference.