Airflow Integration

Apache Airflow is Zingle's optional orchestration integration. Connect Airflow to wire data models to DAGs and deep-link from Zingle into the Airflow UI for monitoring.

---

What you need

Requirement	Details
Airflow base URL	The root URL of your Airflow instance (e.g., https://airflow.yourcompany.com)
Credentials	Username and password, or API token for authentication

---

Connecting Airflow

1. Navigate to Connections

   Open Connections from the sidebar and scroll to the Airflow Connection section.

2. Enter connection details

   Click Connect Airflow and provide:

   • A display name for this connection
   • The Airflow base URL
   • Authentication credentials

3. Validate connectivity

   Zingle tests the connection by calling the Airflow API. If the instance responds successfully, the connection is saved.

   Credentials are stored behind a secret reference — the connection config only holds the metadata needed to communicate with Airflow.

---

Connection status

Once connected, the Airflow card shows:

Field	Description
Name	Display name you provided
Base URL	The Airflow instance URL
Status	Connection health (active/error)

You can Disconnect at any time.

---

What Airflow enables

Feature	How Airflow is used
Observe in Airflow	Deep-link from any pipeline in the Data Pipelines list directly to its Airflow DAG
DAG generation	Future: generate or update DAGs tied to Zingle models
Schedule orchestration	Use Airflow as the scheduler for pipeline execution

---

Troubleshooting

AIRFLOW_TIMEOUT

Zingle couldn't get a response from the Airflow API. Check:

• The Airflow instance is running and accessible
• The base URL is correct (including protocol and port)
• Network policies allow connections from Zingle

AIRFLOW_CONNECTION_FAILED

The Airflow URL or path is invalid, or the server is unreachable. Steps to fix:

• Verify the URL in a browser or via curl
• Check for reverse proxies or load balancers that might be blocking requests
• Ensure authentication credentials are correct