d9d84f814c
We've traditionally used org.apache.activemq.artemis.utils.Base64 for Base64 encoding/decoding. This implementation is based on public domain code from http://iharder.net/base64. In Java 8 java.util.Base64 was introduced. I assumed we hadn't switched to this implementation for performance reasons so I created a simple JMH-based test to compare the two implementations and it appears to me that java.util.Base64 is significantly faster than our current implementation. Using the JDK's class will simplify our code and improve performance. Also, it should be 100% backwards compatible since Base64 encoding/decoding is standardized. |
||
|---|---|---|
| .. | ||
| src/main/java/org/apache/activemq/artemis/tests/performance/jmh | ||
| README.md | ||
| pom.xml | ||
README.md
Apache Artemix JMH Benchmarks
This module contains optional JMH performance tests.
Note that this module is an optional part of the overall project build and does not deploy anything, due to its use of JMH which is not permissively licensed. The module must be built directly.
Building the benchmarks
The benchmarks are maven built and involve some code generation for the JMH part. As such it is required that you rebuild upon changing the code.
mvn clean install
Running the benchmarks: General
It is recommended that you consider some basic benchmarking practices before running benchmarks:
- Use a quiet machine with enough CPUs to run the number of threads you mean to run.
- Set the CPU freq to avoid variance due to turbo boost/heating.
- Use an OS tool such as taskset to pin the threads in the topology you mean to measure.
Running the JMH Benchmarks
To run all JMH benchmarks:
java -jar target/benchmark.jar
To list available benchmarks:
java -jar target/benchmark.jar -l
Some JMH help:
java -jar target/benchmark.jar -h
Example
To run a benchmark on a single thread (tg 1) with gc profiling use:
java -jar target/benchmark.jar -prof gc -tg 1