Apache Iceberg: Open Table Format For Large Analytic Datasets

**Apache Iceberg** is an open table format for huge analytic datasets stored on object storage (S3, GCS, Azure Blob). It tracks **snapshots**, supports **schema evolution**, and exposes **hidden partitioning** so query engines skip irrelevant files.

Engines including Spark, Trino, Flink, and DuckDB integrate Iceberg through catalog APIs. Metadata files (manifests) live beside Parquet or ORC data files.

Iceberg competes in the same problem space as **Delta Lake** and **Apache Hudi**; choice often follows your compute engine and existing cloud catalog (Glue, Polaris, Nessie).

Read `iceberg.apache.org` specifications for snapshot isolation and expire snapshots operations. Operational tasks (compaction, orphan removal) need scheduled jobs.