Project Nessie

About Project Nessie

A transactional catalog for data lakes with git-like semantics. Nessie works with Apache Iceberg tables to provide multi-table transactions, branching, tagging, and time-travel queries across your data lake.

Key Features

• 1Git-inspired catalog for Iceberg and Delta Lake tables on object storage
• 2Branching and tagging enables safe multi-table schema experiments
• 3ACID commits group changes to multiple Iceberg tables into one atomic transaction
• 4SQL catalog integration: Spark, Flink, and Dremio read Nessie as a catalog
• 5REST API and Python client for programmatic catalog management

How Python Data Engineers Use Project Nessie

Python data engineers configure PySpark to use Project Nessie as the Iceberg catalog — enabling table branching within Spark jobs. An engineer creates a Nessie branch, runs a PySpark transformation that modifies multiple Iceberg tables, validates the results, then merges the branch to main — providing atomic multi-table updates with full rollback capability.