From the course: Java Concurrency Troubleshooting: Latency and Throughput

Unlock the full course today

Join today to access over 22,600 courses taught by industry experts or purchase this course individually.

Microbenchmarks in Java with JMH

Microbenchmarks in Java with JMH - Java Tutorial

From the course: Java Concurrency Troubleshooting: Latency and Throughput

Start my 1-month free trial

Microbenchmarks in Java with JMH

“

- [Instructor] You've probably seen the sort of code before, check this out. On line 83, I am collecting the current system time as a variable name startTime then online 88, I'm collecting the same system time as a variable called endTime. I then use the difference between the two to measure the time it took for this code block to execute. This is the time on a tradition of checking the execution time of a particular block of code. For many different reasons, the results of this sort of latency benchmark cannot be trusted. It's hard to simulate real JVM conditions that affect the execution speed of code. You also can't properly address the effect of multithreading or the JVM warmup processes, at least until now. Micro benchmarking is benchmarking but micro, makes sense. (chuckles) The point here is instead of having to set a bulky low testing gear to benchmark your app, you can quickly and reliably find out what parts…

Contents