[Airflow][kubernetes_pod_operator] TypeError: 'NoneType' 类型的参数不可迭代

Question

When I try to run this task, I get the following error:

from airflow.contrib.operators.kubernetes_pod_operator import KubernetesPodOperator 
from airflow import DAG
from datetime import datetime, timedelta

default_args = {
    "owner": "airflow",
    "depends_on_past": False,
    "start_date": datetime(2015, 6, 1),
    "email": ["airflow@airflow.com"],
    "email_on_failure": False,
    "email_on_retry": False,
    "retries": 1,
    "retry_delay": timedelta(minutes=5),
}

dag = DAG("kubernetes", default_args=default_args, schedule_interval=None)

k = KubernetesPodOperator(
    namespace='kubernetes',
    image="ubuntu:16.04",
    cmds=["bash", "-cx"],
    arguments=["echo", "10", "echo pwd"],
    labels={"foo": "bar"},
    name="airflow-test-pod",
    is_delete_pod_operator=True,
    in_cluster=True,
    task_id="task-two",
    get_logs=True,
    dag=dag)

Error:

  File "/usr/local/lib/python3.7/site-packages/kubernetes/config/kube_config.py", line 491, in safe_get
    key in self.value):
TypeError: argument of type 'NoneType' is not iterable

What am I doing wrong? I'm using puckel/airflow and the correct dependencies. <https://github.com/puckel/docker-airflow> I need to edit something in airflow.cfg? I don't know where to search for this.

Answer 1

It seems, you don't have config_file parameter set, so KubernetesPodOperator fall back to its default value, which probably doesn't exist as well.

My suggestion would be to add "config_file=/path/to/kube_config.yaml". In the following file you also provide your credentials/tokens.