wjddn279 opened a new pull request, #58365: URL: https://github.com/apache/airflow/pull/58365
related: https://github.com/apache/airflow/discussions/58143 ### Body As discussed (not confirmed yet), this resolves the issue of sudden memory usage spikes in worker processes when using LocalExecutor. Memory increases due to unnecessary copying of read-only shared memory through COW caused by gc. By applying gc.freeze and moving existing objects to the permanent generation, we prevent COW from occurring. When using fork mode, we create many worker processes at once to minimize gc.freeze and unfreeze calls. When using spawn mode, we maintain the existing approach to ensure stability. ### Benchmark #### memory usage Comparison of per-process memory usage in LocalExecutor before and after applying this PR. Measured in the same environment running 500 tasks per minute for 12 hours. | AS-IS | TO-BE | |-------|-------| | [smem_as-is.txt](https://github.com/user-attachments/files/23567938/smem_as-is.txt) | [smem_to-be.txt](https://github.com/user-attachments/files/23567942/smem_to-be.txt) | | <img width="400" alt="AS-IS" src="https://github.com/user-attachments/assets/53e6122a-5b39-46c6-a862-b635d53d3993"; /> | <img width="400" alt="TO-BE" src="https://github.com/user-attachments/assets/9796032f-6d8d-4234-91eb-e2aa186381ba"; /> | #### gc.freeze / unfreeze performance (elapsed time) We measured the elapsed time of gc.freeze and gc.unfreeze for each scheduler loop iteration. Most operations took microseconds, confirming virtually no impact. The actual operation is a very lightweight process that simply marks objects in the current generation as permanent generation without any memory copying.https://github.com/python/cpython/pull/3705/files ``` [airflow.jobs.scheduler_job_runner.SchedulerJobRunner] loc=scheduler_job_runner.py:1375 2025-11-15T05:13:36.911715Z [info ] freeze: 18.83 μs [airflow.jobs.scheduler_job_runner.SchedulerJobRunner] loc=scheduler_job_runner.py:1366 2025-11-15T05:13:36.912238Z [info ] freeze_cpu: 10.08 μs [airflow.jobs.scheduler_job_runner.SchedulerJobRunner] loc=scheduler_job_runner.py:1367 2025-11-15T05:13:36.912517Z [info ] unfreeze: 10.92 μs [airflow.jobs.scheduler_job_runner.SchedulerJobRunner] loc=scheduler_job_runner.py:1374 2025-11-15T05:13:36.913268Z [info ] unfreeze_cpu: 6.87 μs ``` <!-- 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. --> --- **^ Add meaningful description above** Read the **[Pull Request Guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#pull-request-guidelines)** for more information. In case of fundamental code changes, an Airflow Improvement Proposal ([AIP](https://cwiki.apache.org/confluence/display/AIRFLOW/Airflow+Improvement+Proposals)) is needed. In case of a new dependency, check compliance with the [ASF 3rd Party License Policy](https://www.apache.org/legal/resolved.html#category-x). In case of backwards incompatible changes please leave a note in a newsfragment file, named `{pr_number}.significant.rst` or `{issue_number}.significant.rst`, in [airflow-core/newsfragments](https://github.com/apache/airflow/tree/main/airflow-core/newsfragments). -- 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]
