qlemaire22 edited a comment on issue #6124: [AIRFLOW-5501] in_cluster default value in KubernetesPodOperator overwrites configuration URL: https://github.com/apache/airflow/pull/6124#issuecomment-534082214 I did the rebase, but there are still some errors and Travis is still sad. After some investigation, I realised that my changes were modifying the behavior of some tests. For example this test in [`tests/minikube/test_kubernetes_pod_operator.py`](https://github.com/apache/airflow/blob/master/tests/minikube/test_kubernetes_pod_operator.py): ``` def test_working_pod(self): k = KubernetesPodOperator( namespace='default', image="ubuntu:16.04", cmds=["bash", "-cx"], arguments=["echo 10"], labels={"foo": "bar"}, name="test", task_id="task" ) k.execute(None) actual_pod = self.api_client.sanitize_for_serialization(k.pod) self.assertEqual(self.expected_pod, actual_pod) ``` It does not specify the `in_cluster` argument. Therefore, the old behavior was to use `in_cluster=False` which was the default value. Now, it is using the value defined in the config which is `in_cluster=True` [(link to the config)](https://github.com/apache/airflow/blob/master/scripts/ci/kubernetes/kube/templates/configmaps.template.yaml). Therefore, the test is now different and it is the case for every failing test. Shall I modify the failing tests to include the `in_cluster=False` argument to come back to the old behavior or shall I try to solve the `Service host/port is not set` error which seems to be caused by some missing environment variables?
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
