Which StorageGRID feature provides data durability for large unstructured datasets?

• A. Erasure coding
• B. Object replication
• C. Data tiering
• D. Snapshots

Answer: (A)

Durability for large unstructured datasets in StorageGRID is best achieved with erasure coding. It encodes each object into a set of data and parity shards that are distributed across multiple storage nodes. If some shards are lost due to hardware failures, the remaining shards are sufficient to reconstruct the original object. This provides very high durability while using far less additional storage than keeping full duplicates of every object, which is essential for large datasets. Data tiering moves data between storage classes but doesn’t by itself increase durability. Snapshots are point-in-time copies and don’t protect against ongoing hardware failures across a large dataset. Object replication is durable but requires much more storage, making it less scalable for vast unstructured data.