Microbenchmarking with JMH Sean Sullivan Portland Java User Group March 14, 2023

Java Microbenchmark Harness “JMH is a Java harness for building, running, and analysing nano/micro/milli/macro benchmarks written in Java and other languages targeting the JVM.”

OSS projects that use JMH ● Netty ● Log4j2 ● Spring Framework ● Apache Cassandra ● Apache Kafka ● Netflix Atlas ● Netflix Spectator ● Netflix Zuul

NOTE: JMH is not intended to be used in the same way as a typical testing library such as JUnit. Simply adding the jmh-core jar file to your build is not enough to be able to run benchmarks.

“the key to using JMH is enabling the annotation- or bytecode-processors to generate the synthetic benchmark code”

“Running benchmarks from the IDE is generally not recommended due to generally uncontrolled environment in which the benchmarks run.”

Zuul: HeadersBenchmark

Zuul: HeadersBenchmark

Zuul: HeadersBenchmark

Zuul: HeadersBenchmark

Zuul: build.gradle jmh-gradle-plugin

Zuul: HeadersBenchmark ./gradlew --no-daemon clean :zuul-core:jmh

Learning JMH

Learning JMH

Learning JMH

microbenchmarks-java ● CompressionBenchmark ● CounterBenchmark ● HttpHeadersBenchmark ● Log4j2ClockBenchmark ● RandomBenchmark ● UuidBenchmark

CompressionBenchmark ● Goal: measure throughput of Gzip compression ● Goal: measure throughput of Brotli compression ● Goal: evaluate Brotli compression parameters

CompressionBenchmark public enum CompressionType ● GZIP ● BROTLI_0 ● BROTLI_4 ● BROTLI_11

CompressionBenchmark

CompressionBenchmark

CompressionBenchmark # JMH version: 1.36 # VM version: JDK 17.0.6, OpenJDK 64-Bit Server VM, 17.0.6+10-LTS # Warmup: 2 iterations, 10 s each # Measurement: 5 iterations, 10 s each # Threads: 1 thread, will synchronize iterations # Benchmark mode: Throughput, ops/time

CompressionBenchmark

The End