westonpace commented on PR #33886: URL: https://github.com/apache/arrow/pull/33886#issuecomment-1406789364
Looks like I'll have to tackle both issues after all :) I broke the scalar test into multiple tests. Some kernel tests were still pretty long so as most of those tests run on a large combination of types so I reduced the combinations when testing with valgrind. The scanner test was using too much data. I turns out that sorting an array with 2Mi rows is pretty slow when running in valgrind. The CSV tests were slow because they were testing many combinations of compression and writing CSV files (even 50B CSV files) is very slow under valgrind (the conversion to string in particular). I figure the various compressions are independently tested already and so removed those when running under valgrind. The hash tests were simply running too many iterations. -- 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]
