Columnar Data, from the inside out Delta Lake, Explained →

Columnar / Storage and tables

Delta Lake

Delta Lake is an open table format that tracks a table through an ordered transaction log. The _delta_log folder sits beside the Parquet files. Commits add and remove file references; replaying them reconstructs a version, even while older files remain on disk. A checkpoint gives readers a shorter starting point. Databricks created Delta Lake; Spark and delta-rs produce and consume its tables.

ParquetColumns stored as bytesIceberg · Delta LakeFiles selected by a versionArrowTyped arrays in memoryDuckDB · DataFusionOperators turn values into answers
Table metadata selects files. Readers decode stored columns into arrays. Engines compute on those arrays.