[
https://issues.apache.org/jira/browse/AIRFLOW-3378?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16698638#comment-16698638
]
ASF GitHub Bot commented on AIRFLOW-3378:
-----------------------------------------
Fokko closed pull request #4218: [AIRFLOW-3378] KubernetesPodOperator does not
delete on timeout failure
URL: https://github.com/apache/incubator-airflow/pull/4218
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/airflow/contrib/operators/kubernetes_pod_operator.py
b/airflow/contrib/operators/kubernetes_pod_operator.py
index 99c6da11b3..f6c1f9d45e 100644
--- a/airflow/contrib/operators/kubernetes_pod_operator.py
+++ b/airflow/contrib/operators/kubernetes_pod_operator.py
@@ -124,13 +124,14 @@ def execute(self, context):
launcher = pod_launcher.PodLauncher(kube_client=client,
extract_xcom=self.xcom_push)
- (final_state, result) = launcher.run_pod(
- pod,
- startup_timeout=self.startup_timeout_seconds,
- get_logs=self.get_logs)
-
- if self.is_delete_operator_pod:
- launcher.delete_pod(pod)
+ try:
+ (final_state, result) = launcher.run_pod(
+ pod,
+ startup_timeout=self.startup_timeout_seconds,
+ get_logs=self.get_logs)
+ finally:
+ if self.is_delete_operator_pod:
+ launcher.delete_pod(pod)
if final_state != State.SUCCESS:
raise AirflowException(
diff --git a/tests/contrib/minikube/test_kubernetes_pod_operator.py
b/tests/contrib/minikube/test_kubernetes_pod_operator.py
index f808a2f47f..e1dfce28c4 100644
--- a/tests/contrib/minikube/test_kubernetes_pod_operator.py
+++ b/tests/contrib/minikube/test_kubernetes_pod_operator.py
@@ -107,6 +107,27 @@ def test_image_pull_secrets_correctly_set(self,
client_mock, launcher_mock):
self.assertEqual(launcher_mock.call_args[0][0].image_pull_secrets,
fake_pull_secrets)
+ @mock.patch("airflow.contrib.kubernetes.pod_launcher.PodLauncher.run_pod")
+
@mock.patch("airflow.contrib.kubernetes.pod_launcher.PodLauncher.delete_pod")
+ @mock.patch("airflow.contrib.kubernetes.kube_client.get_kube_client")
+ def test_pod_delete_even_on_launcher_error(self, client_mock,
delete_pod_mock, run_pod_mock):
+ k = KubernetesPodOperator(
+ namespace='default',
+ image="ubuntu:16.04",
+ cmds=["bash", "-cx"],
+ arguments=["echo 10"],
+ labels={"foo": "bar"},
+ name="test",
+ task_id="task",
+ in_cluster=False,
+ cluster_context='default',
+ is_delete_operator_pod=True
+ )
+ run_pod_mock.side_effect = AirflowException('fake failure')
+ with self.assertRaises(AirflowException):
+ k.execute(None)
+ delete_pod_mock.assert_called()
+
@staticmethod
def test_working_pod():
k = KubernetesPodOperator(
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
> KubernetesPodOperator does not delete on timeout failure
> --------------------------------------------------------
>
> Key: AIRFLOW-3378
> URL: https://issues.apache.org/jira/browse/AIRFLOW-3378
> Project: Apache Airflow
> Issue Type: Bug
> Components: kubernetes
> Affects Versions: 1.10.1
> Reporter: Victor
> Assignee: Victor
> Priority: Major
> Fix For: 2.0.0
>
>
> I am runnig airflow 1.10.1rc2 on kubernetes with the LocalExecutior and
> KubernetesPodOperator to run pods.
>
> When a failure happens because the pod can't be created (see logs below), the
> whole operator fails and if is_delete_operator_pod is set to True, the pod is
> NOT deleted as expected.
>
> Logs:
> {noformat}
> *** Reading local file:
> /airflow/logs/demo/datapipe/2018-11-21T08:29:57.206456+00:00/1.log
> [2018-11-21 08:30:02,027] {models.py:1361} INFO - Dependencies all met for
> <TaskInstance: demo.datapipe 2018-11-21T08:29:57.206456+00:00 [queued]>
> [2018-11-21 08:30:02,040] {models.py:1361} INFO - Dependencies all met for
> <TaskInstance: demo.datapipe 2018-11-21T08:29:57.206456+00:00 [queued]>
> [2018-11-21 08:30:02,041] {models.py:1573} INFO -
> --------------------------------------------------------------------------------
> Starting attempt 1 of 1
> --------------------------------------------------------------------------------
> [2018-11-21 08:30:02,066] {models.py:1595} INFO - Executing
> <Task(KubernetesPodOperator): datapipe> on 2018-11-21T08:29:57.206456+00:00
> [2018-11-21 08:30:02,067] {base_task_runner.py:118} INFO - Running: ['bash',
> '-c', 'airflow run demo datapipe 2018-11-21T08:29:57.206456+00:00 --job_id 4
> --raw -sd /usr/local/airflow/dags_volume/..data/demo.py --cfg_path
> /tmp/tmp5qa5qyhs']
> [2018-11-21 08:30:03,065] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe [2018-11-21 08:30:03,065] {settings.py:174} INFO -
> setting.configure_orm(): Using pool settings. pool_size=5, pool_recycle=1800
> [2018-11-21 08:30:03,734] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe [2018-11-21 08:30:03,732] {__init__.py:51} INFO - Using executor
> LocalExecutor
> [2018-11-21 08:30:04,038] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe [2018-11-21 08:30:04,038] {models.py:271} INFO - Filling up the
> DagBag from /usr/local/airflow/dags_volume/..data/demo.py
> [2018-11-21 08:30:04,590] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe [2018-11-21 08:30:04,590] {cli.py:484} INFO - Running <TaskInstance:
> demo.datapipe 2018-11-21T08:29:57.206456+00:00 [running]> on host
> infra-airflow-6d78c56489-r9trl
> [2018-11-21 08:30:04,684] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:04,676] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:05,698] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:05,698] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:06,706] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:06,706] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:07,715] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:07,714] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:08,723] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:08,723] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:09,730] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:09,730] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:10,736] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:10,736] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:11,742] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:11,742] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:12,748] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:12,748] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:13,754] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:13,754] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:14,764] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:14,764] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:15,769] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:15,769] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:16,777] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:16,777] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:17,783] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:17,783] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:18,790] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:18,790] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:19,796] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:19,796] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:20,802] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:20,802] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:21,808] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:21,808] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:22,815] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:22,815] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:23,822] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:23,822] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:24,833] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:24,833] {pod_launcher.py:121} INFO - Event: datapipe-a5591f0c had an
> event of type Pending
> [2018-11-21 08:30:24,841] {models.py:1760} ERROR - Pod Launching failed: Pod
> took too long to start
> Traceback (most recent call last):
> File
> "/usr/local/lib/python3.6/site-packages/airflow/contrib/operators/kubernetes_pod_operator.py",
> line 123, in execute
> get_logs=self.get_logs)
> File
> "/usr/local/lib/python3.6/site-packages/airflow/contrib/kubernetes/pod_launcher.py",
> line 86, in run_pod
> raise AirflowException("Pod took too long to start")
> airflow.exceptions.AirflowException: Pod took too long to start
> During handling of the above exception, another exception occurred:
> Traceback (most recent call last):
> File "/usr/local/lib/python3.6/site-packages/airflow/models.py", line 1659,
> in _run_raw_task
> result = task_copy.execute(context=context)
> File
> "/usr/local/lib/python3.6/site-packages/airflow/contrib/operators/kubernetes_pod_operator.py",
> line 135, in execute
> raise AirflowException('Pod Launching failed: {error}'.format(error=ex))
> airflow.exceptions.AirflowException: Pod Launching failed: Pod took too long
> to start
> [2018-11-21 08:30:24,844] {models.py:1791} INFO - Marking task as FAILED.
> [2018-11-21 08:30:25,001] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:25,001] {email.py:127} INFO - Sent an alert email to
> ['victor.n...@xxx.com']
> [2018-11-21 08:30:26,646] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe Traceback (most recent call last):
> [2018-11-21 08:30:26,646] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File
> "/usr/local/lib/python3.6/site-packages/airflow/contrib/operators/kubernetes_pod_operator.py",
> line 123, in execute
> [2018-11-21 08:30:26,646] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe get_logs=self.get_logs)
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File
> "/usr/local/lib/python3.6/site-packages/airflow/contrib/kubernetes/pod_launcher.py",
> line 86, in run_pod
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe raise AirflowException("Pod took too long to start")
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe airflow.exceptions.AirflowException: Pod took too long to start
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe During handling of the above exception, another exception occurred:
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe Traceback (most recent call last):
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File "/usr/local/bin/airflow", line 32, in <module>
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe args.func(args)
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File
> "/usr/local/lib/python3.6/site-packages/airflow/utils/cli.py", line 74, in
> wrapper
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe return f(*args, **kwargs)
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File "/usr/local/lib/python3.6/site-packages/airflow/bin/cli.py",
> line 490, in run
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe _run(args, dag, ti)
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File "/usr/local/lib/python3.6/site-packages/airflow/bin/cli.py",
> line 406, in _run
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe pool=args.pool,
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File "/usr/local/lib/python3.6/site-packages/airflow/utils/db.py",
> line 74, in wrapper
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe return func(*args, **kwargs)
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File "/usr/local/lib/python3.6/site-packages/airflow/models.py",
> line 1659, in _run_raw_task
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe result = task_copy.execute(context=context)
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe File
> "/usr/local/lib/python3.6/site-packages/airflow/contrib/operators/kubernetes_pod_operator.py",
> line 135, in execute
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe raise AirflowException('Pod Launching failed:
> {error}'.format(error=ex))
> [2018-11-21 08:30:26,647] {base_task_runner.py:101} INFO - Job 4: Subtask
> datapipe airflow.exceptions.AirflowException: Pod Launching failed: Pod took
> too long to start
> [2018-11-21 08:30:27,054] {logging_mixin.py:95} INFO - [2018-11-21
> 08:30:27,053] {jobs.py:2627} INFO - Task exited with return code 1
> {noformat}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
