Apache Iceberg

Noun · Development

Definitions

  1. An open table format for large analytic datasets that brings ACID transactions, schema evolution, time travel, and partition evolution to data lakes. Works with Spark, Trino, Flink, and DuckDB. Stores metadata separately from data files, enabling fast query planning on petabyte-scale tables.

    In plain English: A file format for big data that adds database-like features (transactions, time travel) to data stored in a data lake.

    Example: "Iceberg's time travel lets us query last week's snapshot of the table — invaluable when someone accidentally deletes rows."

Related Terms