Airflow ᴮᴱᵀᴬ

This page covers Sifflet's Airflow integration, which allows you to see Airflow metadata in Sifflet. To trigger Sifflet operations from within your Airflow DAGs, you can find all of Sifflet's custom operators and their purposes here.

You can leverage our full data stack approach to obtain a complete view of your data pipelines, from your orchestrator to the data warehouse and your BI tool.

By integrating Airflow with Sifflet, you will have at a glance:

• Various metadata about your DAGs, such as "Last/Next Execution Date" and "Last Updated Date"
• The latest status of your DAG runs. It allows you to see directly whether a failure happened recently.

To integrate Airflow with Sifflet, these are the steps to follow:

1. Create a dedicated read-only user
2. Connect to Sifflet

📘

Supported Airflow versions

We currently support any self-hosted Airflow instance - version 2.0.0+ - and are working on cloud-managed instances (MWAA on AWS, etc.)

1. Create a read-only user

You can create a dedicated Sifflet user with a "Viewer" role.
Please choose a "User Name" (for instance, "sifflet_user") and a secure password. Store them carefully as you will need them when configuring the connection in Sifflet later.

2. Connect to Sifflet

Add an Airflow secret

To create the Airflow secret, follow the below steps:

• In "Integration" --> tab "Secrets", create a new secret.
• In the "Secret" area, copy-paste the below text and replace it with the correct username and password previously created in step 1:

{
  "user": "<username>",
  "password": "<password>"
}

Create a new Airflow integration

To connect Airflow with Sifflet, you will need three items:

• Connection details:
  • Host: You can add the entire URL. For instance, if your URL is http://xxxxx.yy, your Host value would be http://xxxxx.yy.
  • Port: The port used to interact with Airflow's REST API. By default, this is 8080.
• Secret: corresponds to the username and password you previously chose.
• Frequency: determines how often the information is refreshed.

You can also refer to this page on adding a data source in Sifflet.