Oracle's documentation indicates that the most ideal JMH use case is to use a Maven project that is dependent on the application's JAR files. They further recommend that microbenchmarking take place via the command-line and not from within an Integrated Development Environment (IDE), as that could impact the results.
To use JMH, we will use bytecode processors (annotations) to generate the benchmark code. We use Maven archetypes to enable JMH.
In order to test the JMH, we require an IDE with support for Maven and Java 9. If you do not yet have Java ...