2010YOUY01 commented on PR #22626: URL: https://github.com/apache/datafusion/pull/22626#issuecomment-4585536018
Thank you — this is an important problem to solve. One concern is that some SLTs intentionally target memory-intensive queries and may adjust the memory pool limit themselves, while this PR assumes a preconfigured memory limit when the SLT runner starts. To address this, how about the following? - Set a hard-coded small memory limit for regular SLTs, with the assumption that most test cases should run with limited memory. Tests that require more memory should be moved to a dedicated memory-intensive test suite. - Apply this memory-accounting check only to that dedicated memory-intensive test suite, where the memory assumptions are explicit. (but they might still adjust the memory limit within the same `slt`, from query to query, it's still a challenge how to deal with such cases) -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
