Observability Engineering, 2nd Edition

Chapter 4. Getting Started with Instrumentation

Now that we’re on the same page about overarching theories and the origins of observability as a practice, we’ll get into concrete details about technical fundamentals and implementations. We will begin by examining the fundamentals of instrumentation.

There’s an old business adage that roughly says, “you can’t improve what you don’t measure.” That idea rings true for the systems you build with a high degree of observability. Because there are so many different functional aspects of modern application systems, in practice, we must collect an increasingly vast amount of data in order to fully understand them.

As you’ll see in this section of the book, generating and collecting telemetry data is the foundational cornerstone of an observable system. Given a sufficient volume of telemetry data to work from, it’s possible to understand any given state of your system—no matter how novel or bizarre. However, before we start talking about data analysis, we first need to dig into what a “sufficient volume of telemetry data” means. Let’s dive into what it takes to properly instrument your systems.

This chapter provides an introduction to instrumentation, and it offers considerations when generating telemetry and regarding where that telemetry is sent. As a software engineer responsible for ensuring application reliability, these considerations must be essential to your workflows.

We also briefly introduce concepts that all software engineers ...