rm5248 commented on code in PR #539: URL: https://github.com/apache/logging-log4cxx/pull/539#discussion_r2350441430
########## src/site/markdown/performance.md: ########## @@ -65,3 +59,46 @@ The throughput tests may be built by specifying `BUILD_THROUGHPUT` with CMake when building Log4cxx. The benckmark tests require Google's [Benchmark](https://github.com/google/benchmark) library and may be built by specifying `BUILD_BENCHMARK_CHECKS` with CMake when building Log4cxx. + +The following table shows some timing tests using Google's benchmarking library. + + g++ (Ubuntu 11.4.0-1ubuntu1~22.04.2) 11.4.0 + Run on (8 X 2328.61 MHz CPU s) + CPU Caches: + L1 Data 32 KiB (x4) + L1 Instruction 32 KiB (x4) + L2 Unified 256 KiB (x4) + L3 Unified 6144 KiB (x1) + Load Average: 0.07, 0.03, 0.01 + +| Benchmark | Time | CPU | Iterations | Review Comment: that sounds better at least. I've come up with the following statement, is this correct? If so, there's probably a better way to phrase it. > The 'Iterations' column shows how many benchmark iterations were done during the test period. The test period is shown in the 'Time' column, so the number of operations during a set time is given by the formula `iterations / time`. > Using the above forumla, we can calculate how many log statements can be executed in a second. Note that this is best-case scenario, where messages are not being output to an appender(such as `stderr` or a file). I can come up with some graphs later, the few minutes I just spent on it does show some trends. Most notably the contention on multithreading logging causes the performance to drop down a lot. -- 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: notifications-unsubscr...@logging.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
