SEZ9 commented on issue #13248: URL: https://github.com/apache/seatunnel/issues/13248#issuecomment-6844241432
Thanks for the detailed report and reproduction, and for opening #13249 with the `PythonTestImage` approach (deriving from each engine image, preserving platform and USER, installing Python only when missing, propagating build failures, and caching per source image within the test JVM). Preparing Python ahead of container startup rather than installing from mirrors at test runtime sounds like the right direction, and keeping the enablement gates, executable allowlists and engine startup commands unchanged is appreciated. Noted that #13249 depends on #13247 for the Java 17 G1 cleanup check and that #13247 should be merged first; we'll keep that ordering in mind during review. One point for the PR discussion: as you mention, the hosted failures show runtime package-resolution errors but the reproduction does not establish their underlying cause. If you have a link to a failing hosted run, adding it to the PR would help reviewers confirm whether this change addresses those failures as well. Let's continue the review on #13249. <!-- streview-comment:2320 --> -- 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]
