I have been using apache airflow for months.
and also have experiences with GCP composer, AWS data pipeline, and Glue which is managed services
In Airflow, I know that DAGs and tasks are written in Python, not using GUI. UI for Airflow is not for building dags and tasks. However, there are many pipeline solutions(ex. AWS data pipeline, Glue, etc.) that have features for building those dags and tasks with UI like drag n drop or something else with minimum coding.
Can someone explain why those capabilties are not needed in Airflow?
Thanks
Advantages of configuration as code:
- Automation and standardisation
- Versioning of changes
- Traceability of changes
- Coding assistance and validation