Overview

Airflow 2.1, released on May 22, 2021, introduces cross-DAG dependencies and a calendar view to better visualize scheduled runs.

Main Features

Cross-DAG dependencies

The ExternalTaskSensor and cross-DAG dependencies let you create pipelines where one DAG waits for a task in another DAG to complete before running.

python 
from airflow import DAG
from airflow.operators.python import PythonOperator
from airflow.sensors.external_task import ExternalTaskSensor
from datetime import datetime

with DAG('consumer_dag', start_date=datetime(2021, 1, 1)) as dag:
    # Wait for the producer DAG to finish
    wait = ExternalTaskSensor(
        task_id='wait_for_extraction',
        external_dag_id='producer_dag',
        external_task_id='extract_data',
        timeout=3600,
    )

    process = PythonOperator(
        task_id='process',
        python_callable=lambda: print('Processing'),
    )

    wait >> process

Calendar view

The new calendar view in the web interface displays DAG runs on a monthly calendar, making it easy to identify failures and execution trends.

python 
from airflow import DAG
from airflow.operators.python import PythonOperator
from datetime import datetime, timedelta

# Daily DAG visible in the calendar view
with DAG(
    'daily_report',
    start_date=datetime(2021, 1, 1),
    schedule_interval='@daily',
    catchup=False,
) as dag:
    generate = PythonOperator(
        task_id='generate_report',
        python_callable=lambda: print('Report generated'),
    )

Sources