Why Apache Iceberg Is Not “Just Another Table Format”

Apache Iceberg is a table format specification that sits between file formats (Parquet, ORC) and compute engines, solving critical failures of the Hive era: slow query planning via directory listing, silent schema corruption from position-based column tracking, and lack of safe concurrent writes. Iceberg replaces directory-based tracking with a versioned metadata tree, enabling file-level statistics for query pruning, ACID transactions via immutable snapshots with atomic commits, and metadata-only schema evolution using permanent column IDs. Unlike a traditional warehouse, Iceberg delivers warehouse-grade guarantees on open, decoupled object storage — but shifts operational responsibility (compaction, snapshot expiration) to the user. Major adoption signals include Snowflake achieving full parity with native tables, AWS S3 Tables with built-in Iceberg support, and companies like LinkedIn and Airbnb reporting significant performance and cost gains.