airflow-provider-ibm-db2

Apache Airflow provider to connect to IBM Db2 using ibm_db_dbi (native) with a fallback to pyodbc.
Includes basic Hook and Operators, an example DAG, and unit tests.

Status: Initial MVP — designed to evolve with your feedback and contributions.

✨ Features

• Db2Hook with driver auto-detection: ibm_db_dbi → pyodbc (fallback).
• Db2SqlOperator to execute parameterized SQL (based on SQLExecuteQueryOperator).
• Db2StoredProcedureOperator to invoke CALL schema.proc(?,?).
• Db2CheckOperator for data quality checks (counts/values).
• bulk_load() via SYSPROC.ADMIN_CMD('LOAD/IMPORT ...') (if the user has privileges).
• Example DAG and tests with pytest.

⚡ Installation (editable)

pip install -e .[dev]

🔧 Airflow Connection Setup

Create a Connection with ID db2_default:

• Conn Type: Db2 (free text) or Generic

• Host: hostname | Port: 50000

• Schema: DBNAME | Login: user | Password: ******

• Optional Extra (JSON):

{
  "ssl": true,
  "currentSchema": "DB2ADMIN",
  "securityMechanism": "13"
}

🚀 Quick Start

Check out the example DAG:
src/airflow_provider_ibm_db2/example_dags/example_db2_dag.py

🗺️ Short-term Roadmap

• Transfer operators (Db2 → Parquet, Db2 → Postgres)
• Sensible defaults for isolation level and retries
• Compatibility matrix (DB2 LUW 11.1/11.5, Python 3.9–3.12, Airflow ≥2.7)

📜 License

MIT