Getting Started with Apache Airflow for Qubinode Navigator

Documentation status

Validation: IN PROGRESS – Based on deploy-qubinode-with-airflow.sh and current Airflow sidecar configuration.

Last reviewed: 2025-11-21

Community: If you use this tutorial to bring up Airflow and run Qubinode DAGs, please share fixes or enhancements via Contributing to docs.

This tutorial shows how to:

Deploy Qubinode Navigator together with Apache Airflow and nginx using deploy-qubinode-with-airflow.sh.
Access the Airflow UI through nginx.
Locate and run Qubinode-related DAGs.

1. Prerequisites

A host where Qubinode Navigator can run (RHEL 9/10, CentOS Stream 9/10, Rocky Linux 9, etc.).
Container runtime and networking set up per your platform guide.
Platform-specific inventory and environment files prepared as described in:
- Deploying on Hetzner
- Deploying on Red Hat Product Demo System
Ports 80/443 available for nginx.

If you havent done a base deployment yet, start with:

Deploy To Production

2. Run `deploy-qubinode-with-airflow.sh`

On the prepared host:

cd /root/qubinode_navigator    # or your clone path

./deploy-qubinode-with-airflow.sh

The script will:

Run the base Qubinode deployment (if needed).
Deploy Apache Airflow using the existing configuration.
Configure nginx as a reverse proxy in front of Airflow and the AI Assistant.
Close direct ports (e.g., 8888/8080) and expose only 80/443.

At the end, you should see output similar to:

✓ Direct ports 8888/8080 closed
✓ Access only through nginx (port 80/443)
✓ Ready for SSL/TLS configuration

📚 Next Steps:
   1. Access Airflow UI and enable example DAGs
   2. Test kcli and virsh operators
   3. Chat with AI Assistant for guidance

If the script fails, review the logs it prints and update this tutorial with any repeatable troubleshooting steps you discover.

3. Access the Airflow UI

Once deployment completes successfully:

Determine the host IP:
```
hostname -I | awk '{print $1}'
```
In a browser, open:
- http://YOUR_HOST_IP/ (served by nginx as the Airflow UI).
Log in with the credentials configured for Airflow (often admin / admin by default e change immediately for production).

nginx is configured to:

Proxy / to the Airflow webserver.
Proxy /ai/ (or similar) to the AI Assistant API (depending on your nginx config in deploy-qubinode-with-airflow.sh).

4. Locate and Run Qubinode DAGs

In the Airflow UI:

Go to DAGs view.
Look for DAGs related to Qubinode Navigator, for example (exact names may vary by version):
- qubinode_simple_deploy
- multi_cloud_deploy
- RAG/AI‑related workflows documented in:
  - Airflow Integration Guide
  - DAG Deployment Workflows

To run a DAG:

Unpause it if it is paused.
Use the Play button to trigger a run.
Monitor the tasks in the Graph or Tree view and inspect logs when needed.

5. Integrate with the AI Assistant

With Airflow and the AI Assistant running behind nginx:

You can use the AI Assistant to:
- Trigger DAGs via natural language requests.
- Ask for status of specific workflows.
- Query documentation or RAG‑powered answers about deployment issues.

6. Next Steps and Contributions

If you adopt additional DAGs or customize workflows, consider contributing:
- New example DAGs.
- Small sections here describing your common patterns.
If you run into repeated issues (e.g., firewall, SELinux, SSL termination), add a Troubleshooting section to this tutorial so future users benefit from your findings.