westonpace commented on issue #36783: URL: https://github.com/apache/arrow/issues/36783#issuecomment-1650329566
The latest plot is somewhat fascinating. It seems performance has partially recovered, but not all at once. It has recovered in a very smooth (practically linear) curve. While this is aesthetically pleasing it is baffling as to what might be the cause.  There are also other benchmarks (not tpc-h) which are exhibiting regressions on this machine that haven't been detected yet such as https://conbench.ursa.dev/benchmark-results/064c00b67d8c7b298000439798b81c6a/ or https://conbench.ursa.dev/benchmark-results/064c00bbf4567f3380006cb89abd317c/ (this one is only testing integer parsing and not very complex) Given all these touch points I do not believe this to be a change in the Arrow code itself. It could perhaps be tied to brew, mac CI, or just the hardware. -- 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]
