Airflow and TaskFlow API first steps: 'NoneType' object is not iterable

Question

I am trying to understand how TaskFlow API works and I am starting from this basic DAG:

def _push_argument(ti):
    ti.xcom_push(key='p1', value={'param1': 'uno', 'param2': 'due'})

def _print_data(param1, param2):
    print(param1)
    print(param2)
  

with DAG(
    dag_id='test_airflow_api_dag',
    schedule_interval="@daily",
    start_date=pendulum.datetime(2022, 1, 1, tz="UTC"),
    catchup=False,
    default_args=default_args,
    tags=["taskapi", "test"],
) as dag:
    create_output = PythonOperator(
        task_id="create_output",
        python_callable=_push_argument
    ).output
    bash_operator_task = PythonOperator(
        task_id="bash_operator_task",
        python_callable=_print_data,
        op_kwargs=create_output
    )

I am following a tutorial. The dag, in any case does no work. I receive this error:

[2022-05-31, 12:02:49 UTC] {taskinstance.py:1718} ERROR - Task failed with exception
Traceback (most recent call last):
  File "/home/airflow/.local/lib/python3.7/site-packages/airflow/models/taskinstance.py", line 1334, in _run_raw_task
    self._execute_task_with_callbacks(context)
  File "/home/airflow/.local/lib/python3.7/site-packages/airflow/models/taskinstance.py", line 1460, in _execute_task_with_callbacks
    result = self._execute_task(context, self.task)
  File "/home/airflow/.local/lib/python3.7/site-packages/airflow/models/taskinstance.py", line 1516, in _execute_task
    result = execute_callable(context=context)
  File "/home/airflow/.local/lib/python3.7/site-packages/airflow/operators/python.py", line 169, in execute
    context.update(self.op_kwargs)
  File "/usr/local/lib/python3.7/_collections_abc.py", line 846, in update
    for key, value in other:
TypeError: 'NoneType' object is not iterable

What am I wronging? I am following the tutorial at the moment and the code seems the same.

score 0 · Answer 1 · answered May 31 '22 at 12:57

I changed the code in this way:

def _push_argument(ti):
    lista =  ["stringa","stringa2"]
    return {
        'uno': 1,
        'due': 2,
        'tre': lista
    }

def _print_data(uno, due, tre):
    print(uno)
    print(due)
    print(tre)



with DAG(
    dag_id='test_airflow_api_dag',
    schedule_interval="@daily",
    start_date=pendulum.datetime(2022, 1, 1, tz="UTC"),
    catchup=False,
    default_args=default_args,
    tags=["taskapi", "test"],
) as dag:
    create_output = PythonOperator(
        task_id="create_output",
        python_callable=_push_argument
    )
    bash_operator_task = PythonOperator(
        task_id="bash_operator_task",
        python_callable=_print_data,
        op_kwargs=create_output.output
    )

I removed:

ti.xcom_push(key='p1', value={'param1': 'uno', 'param2': 'due'})

and I put a return. In this way worked.

Airflow and TaskFlow API first steps: 'NoneType' object is not iterable

1 Answers1