Airflow

Use HttpSensor to pause an Airflow DAG until a website is available

Pause an Airflow DAG until an HTTP endpoint returns 200 OK

Bartosz Mikulski 20 Nov 2020 – 1 min read

What should I do if I wanted to pause an Airflow DAG until one of my articles gets publishes?

Let’s assume that I have already scheduled publication of the text you are reading right now, and I know the article URL. It is going to be https://mikulskibartosz.name/use-httpsensor-to-wait-for-website

Now, I have to define an HTTP connection in my Airflow connections settings. The connection should point to the root address of my website: https://mikulskibartosz.name.

After that, I add a HttpSensor to my DAG. It is going to use the connection defined earlier and wait until the resource is available:

sensor = HttpSensor(
    task_id='http_sensor_check',
    http_conn_id='mikulskibartosz_name',
    endpoint='use-httpsensor-to-wait-for-website',
    poke_interval=60,
    dag=dag)

The Airflow DAG will wait until the website returns a status code, which indicates success. When you look at this sensor’s code, you will see that it returns False only when the failure was due to 404 Not Found status. In a case of any other failure, it propagates the exception to Airflow.

Older post

Use HttpSensor to pause an Airflow DAG until a website is available

How to add an EMR step in Airflow and wait until it finishes running

How to select a random sample of rows using Athena

Use HttpSensor to pause an Airflow DAG until a website is available

How to add an EMR step in Airflow and wait until it finishes running

How to select a random sample of rows using Athena

Related Posts

Anomaly detection in Airflow DAG using Prophet library

How to restart a stuck Airflow DAG

Why does the DayOfWeekSensor exist in Airflow?