Performance Benchmarking Guide with JMH

To benchmark our performance-critical code, we use JMH (Java Microbenchmark Harness). Follow these steps to set up and execute the benchmarks:

Step 1: Download and Prepare the Benchmark Dataset

• Download the Dataset: Download a benchmark dataset from the provided link. Benchmark Dataset
• Extract the Dataset: After downloading, extract the dataset into the directory edux/benchmark-data/.
  • Important: Do not rename the folder after extraction.

Step 2: Run JMH Benchmarks

A complete run can take more than 30 min.

• Execute Benchmarks Using Gradle:
  • Use the Gradle command to run the JMH benchmarks. This can typically be done with a command like ./gradlew jmh in your terminal or command prompt, depending on your project's Gradle setup.

Step 3: Analyze the Results

• JSON Results File:
  • After the benchmarks have been run, a JSON file with the results will be created.
  • This file can typically be found in the edux\lib\benchmark-results.json directory, or the directory specified in your JMH configuration.

Step 4: Visualizing Benchmark Results

• Use a Visualization Tool:
  • You can use tools like JMH Visualizer to upload and visually analyze the JSON results file.
  • This helps in better understanding the performance metrics and identifying any bottlenecks or areas for improvement.

---

By following these steps, you can effectively benchmark and analyze the performance of your critical code segments using JMH.