Chapter 5
Speed Up Your Systems

So there you are, deep in the midst of a task to analyze your website traffic logs. You were hoping to have finished it yesterday, so your heart leaps as you realize you finally have everything tested and working. Now it’s time to kick off the task on the full dataset, and wait. And wait. And wait. How long will it take to finish? A day? A week? Until the heat death of the universe? After about an hour, things are looking pretty bleak.

Now, Clojure is fast. One of the language’s guiding principles is that it should be useful wherever Java is useful, including use cases that require high performance, like your log analysis work. As with any other programming language, there are times when we’ll need a bag of tricks to make our code as fast as possible. Sometimes, performance hacks like type hints and other Java interop artifacts result in uglier code. But with macros, we can hide the noise and maintain the beauty of our code. You’ll see how in this chapter, and you’ll also see how you can entirely avoid runtime costs in some situations with macros. But first you need to make sure you optimize the right things by benchmarking.