For Apache Airflow, How can I pass the parameters when manually trigger DAG via CLI?

I use Airflow to manage ETL tasks execution and schedule. A DAG has been created and it works fine. But is it possible to pass parameters when manually trigger the dag via cli.

For example: My DAG runs every day at 01:30, and processes data for yesterday(time range from 01:30 yesterday to 01:30 today). There might be some issues with the data source. I need to re-process those data (manually specify the time range).

So can I create such an airflow DAG, when it's scheduled, that the default time range is from 01:30 yesterday to 01:30 today. Then if anything wrong with the data source, I need to manually trigger the DAG and manually pass the time range as parameters.

As I know airflow test has -tp that can pass params to the task. But this is only for testing a specific task. and airflow trigger_dag doesn't have -tp option. So is there any way to tigger_dag and pass parameters to the DAG, and then the Operator can read these parameters?

Thanks!

Solution

You can pass parameters from the CLI using --conf '{"key":"value"}' and then use it in the DAG file as "{{ dag_run.conf["key"] }}" in templated field.

CLI:

airflow trigger_dag 'example_dag_conf' -r 'run_id' --conf '{"message":"value"}'

DAG File:

args = {
    'start_date': datetime.utcnow(),
    'owner': 'airflow',
}

dag = DAG(
    dag_id='example_dag_conf',
    default_args=args,
    schedule_interval=None,
)

def run_this_func(ds, **kwargs):
    print("Remotely received value of {} for key=message".
          format(kwargs['dag_run'].conf['message']))


run_this = PythonOperator(
    task_id='run_this',
    provide_context=True,
    python_callable=run_this_func,
    dag=dag,
)

# You can also access the DagRun object in templates
bash_task = BashOperator(
    task_id="bash_task",
    bash_command='echo "Here is the message: '
                 '{{ dag_run.conf["message"] if dag_run else "" }}" ',
    dag=dag,
)

PSA for new airflow users: it's worth making the switch to --conf. Params are a nice pattern, but they are an architectural dead end because they don't appear widely supported for programmatic use. Meanwhile, --conf is exposed in services like google Cloud Composer's trigger dag run.