atrbgithub opened a new pull request, #61839:
URL: https://github.com/apache/airflow/pull/61839

   Closes: https://github.com/apache/airflow/issues/57553
   
   This commit ensures that completed pods are eventually cleaned up.
   
   <!--
    Licensed to the Apache Software Foundation (ASF) under one
    or more contributor license agreements.  See the NOTICE file
    distributed with this work for additional information
    regarding copyright ownership.  The ASF licenses this file
    to you under the Apache License, Version 2.0 (the
    "License"); you may not use this file except in compliance
    with the License.  You may obtain a copy of the License at
   
      http://www.apache.org/licenses/LICENSE-2.0
   
    Unless required by applicable law or agreed to in writing,
    software distributed under the License is distributed on an
    "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
    KIND, either express or implied.  See the License for the
    specific language governing permissions and limitations
    under the License.
    -->
   
   <!--
   Thank you for contributing! Please make sure that your code changes
   are covered with tests. And in case of new features or big changes
   remember to adjust the documentation.
   
   Feel free to ping committers for the review!
   
   In case of an existing issue, reference it using one of the following:
   
   closes: #ISSUE
   related: #ISSUE
   
   How to write a good git commit message:
   http://chris.beams.io/posts/git-commit/
   -->
   
   
   
   
   <!-- Please keep an empty line above the dashes. -->
   ---
   
   Previously, in say `2.9.3`, `self.job.executor.try_adopt_task_instances` was 
always called here:
   
   
https://github.com/apache/airflow/blob/81845de9d95a733b4eb7826aaabe23ba9813eba3/airflow/jobs/scheduler_job_runner.py#L1641
   
   It was called unconditionally, even if it found no TaskInstances to adopt. 
   
   This meant that in the kubernetes executor, we would always call this line:
   
   
https://github.com/apache/airflow/blob/81845de9d95a733b4eb7826aaabe23ba9813eba3/airflow/providers/cncf/kubernetes/executors/kubernetes_executor.py#L601
   
   This was triggered at startup by calling `adopt_or_reset_orphaned_tasks`, 
here:
   
   
https://github.com/apache/airflow/blob/81845de9d95a733b4eb7826aaabe23ba9813eba3/airflow/jobs/scheduler_job_runner.py#L928
   
   It was also then called frequently, configurable with 
`orphaned_tasks_check_interval`. 
   
   The result of this is that if the query that is run to detect adoptable 
tasks does not find any tasks to adopt, we no longer make a call to 
`_adopt_completed_pods`, and as a result completed pods are left hanging 
around. This happens when an old scheduler instance is stopped and a new one 
takes its place. 
   
   This PR partially restores the old behaviour (frequent calls to 
`_adopt_completed_pods`).
   
   Tests pass with:
   
   ```
   breeze run pytest providers/cncf/kubernetes/tests/unit -v
   ```
   
   The issue can be replicated in the following way using Aiflow 3.1.7:
   
   1. Spin up a 3.1.7 cluster locally with unmodified code.
   2. Spin up tasks via the executor
   3. Stop the scheduler before they finished.
   4. Pods go to a completed state
   5. Start the scheduler
   6. Pods are never cleaned up
   
   With this patch, shortly after starting up the scheduler the completed pods 
will be cleaned up. 


-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to