cde-submit-migration 0.14.0

Lightweight spark-based migration runner for executing SQL migrations on Spark catalog tables in Cloudera Data Engineering (CDE).

cde-submit-migration

Lightweight Spark-based migration runner for executing SQL migrations on Spark catalog tables in Cloudera Data Engineering (CDE).

This repository contains a small CLI and a Spark job that can apply SQL migration files (single .sql file or a zipped folder containing multiple .sql files) to a Spark catalog. It was created to make it easy to run migrations either locally (for testing) or by submitting the job through CDE's cde spark submit command.

Highlights

• Run a single SQL migration file or a folder (zipped automatically) containing multiple ordered .sql files.
• Local mode (uses a local path-based Iceberg catalog) for development and testing.
• Integration with Cloudera Data Engineering (cde spark submit) for production runs.

Requirements

• Python 3.11 or later
• pyspark==3.5.1

Install

uv tool install cde-submit-migration

pip install cde-submit-migration

After installation you should have following console command available: - submit-cde-migration to perform migration on CDE environment, - submit-local-migration to perform migration in local environment (for testing purposes)

To upgrade package to the latest version use following commands:

uv tool upgrade cde-submit-migration 

pip install --upgrade cde-submit-migration

Usage

submit-cde-migration --path path/to/migrations --db target_database --profile my-cde-profile

submit-local-migration --path path/to/migrations --db target_database

Notes on arguments:

• --path — path to a single .sql file or a folder containing .sql files. When passing a folder in non-local mode, the folder is copied and zipped before submission.
• --db — target database name to use as the current catalog database. If the database does not exist, it will be created.
• --profile — optional CDE configuration profile name used when calling cde spark submit.

Project structure

• src/cde_submit_migration/main.py — CLI entrypoint that either runs locally or packages and submits to CDE.
• src/cde_submit_migration/migration.py — Spark job that executes SQL migration(s).
• pyproject.toml — project metadata and dependencies.

Troubleshooting

• If cde spark submit is not in your PATH, the CDE mode will fail. Make sure the cde CLI is installed and configured in the environment you use to run submit-cde-migration.
• If you see Spark/Java version incompatibilities when running locally, verify your local Spark/PySpark installation matches the expected runtime (the project pins pyspark==3.5.1).