Web Operations

Chapter 3. Infrastructure and Application Metrics

John Allspaw

Matt Massie

IN ANY SIZE WEB OPERATION, gathering metrics is about as important as plugging servers into a network. For infrastructures that are growing, it's even more important. Imagine designing a car without any of the dials or indicators in front of the driver. Now paint the windshield black. That's what it's like to run a web operation without metrics.

This chapter is about gathering metrics regarding your backing infrastructure, and the material presented here will overlap with other chapters in this book about user-facing metrics and monitoring. We're going to discuss not just the types of metrics you should gather and keep an eye on, but also what you can do with them to better enable you to react to various issues. We'll also go over what makes a great metrics collection system.

Gathering, storing, and displaying metrics for a growing web application is quite a feat, but there are a number of great tools that can do most (if not all) of the heavy lifting for you. The differences between the most popular ones are items for a blog post, not for this chapter. The important thing is that whatever tool you choose or build, it should be something you can depend on. You should be able to treat your metrics like a microphone—always on and always recording what is happening with your infrastructure.

One thing I want to note is that there should be a distinct difference in your mind between metrics collection and monitoring ...