FrankYang0529 opened a new pull request, #76279: URL: https://github.com/apache/airflow/pull/76279
### Why The stash step compresses the file on one CPU core. It uses zlib level 6 on GitHub-hosted runners and `tar -z` on RunsOn runners. This compression makes the stash step slow. The CI image had the same problem. Since #73997, the CI image build first compresses its tar file with `zstd -6 -T0`. The build then stashes the file with `compression-level: '0'`, which turns off the compression in the stash step. So the CI stash step is now much faster than the PROD stash step. ### How This PR makes the same change for the PROD image. The PROD image build now first compresses its tar file with `zstd -6 -T0`. The build then stashes the file with `compression-level: '0'`. The compressed file keeps its `.tar` name, and the stash key stays the same. `breeze prod-image load` only accepts a `prod-image-save-*-<python>.tar` file. `docker image load` detects zstd from the file content. So Breeze and the jobs that load the image need no change. --- ##### Was generative AI tooling used to co-author this PR? <!-- If generative AI tooling has been used in the process of authoring this PR, please change below checkbox to `[X]` followed by the name of the tool, uncomment the "Generated-by". --> - [X] Yes - Claude Code <!-- Generated-by: [Tool Name] following [the guidelines](https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst#gen-ai-assisted-contributions) --> -- 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]
