Paul Murray

Java Development on Ulitzer Writing meaningful Java benchmarks is a tricky business. It's well known that the Java Virtual Machine's just in time (JIT) compilation process means that running an application for a few seconds won't let you predict the performance of the application over hours or days of uptime. In spite of this, developers often rely on micro-benchmarks to set performance SLAs for their applications. Micro-benchmarks test some small, discrete component of an application. They're usually written in an effort to benchmark a component considered critical to the app's overall performance. Here's a typical example, summing all the numbers from one to a specified limit: long accumulatedTotal(int limit) { long result=0L; for (int i=1; i<=limit; i++) { result += i; } return result; } How long does this method take to execute for different values of limit? On m... (more)