Apache Airflow has become the de facto standard for Data Orchestration. However, throughout the years and versions, it accumulated a set of nuances and bugs which can hinder its production usage.
This package aims at creating a centralized source of tools to overcome some of the limitations found in Airflow and to extend some of its features.
Feature | Description |
---|---|
Correct Task (not DAG) catchup | Provide each Task the ability to catchup / correct previous failures |
Retry task combinations / dependencies | Allow Task combinations to be marked for Retry (for instance, a Run Task and the corresponding Sensor Task ) |
Fail task combinations / dependencies | Allow Task combinations to be marked as Fail (for instance, a Run Task and the corresponding Sensor Task ) |
Retrieve upstream TaskInstances |
Get a set of upstream TaskInstance s either by name or by depth. \n Note: Used for task combination / dependency retry and failure |
Generate Azure Data Factory triggers | Create Azure Data Factory Runs with the corresponding dependencies and appropriate Grouped Retries / Failures |