Infrastructure as Code

Chapter 13. Workflow for the Infrastructure Team

Managing infrastructure as code is a radically different way of working for most IT operations people. This chapter aims to explain how an IT operations team gets their work done in this kind of environment, with some guidance for how to make it work well.

The big shift is away from working directly on servers and infrastructures, to working on them indirectly. An infrastructure engineer can no longer just log onto a server to make a change. Instead, they make changes to the tools and definitions, and then allow the change management pipeline to roll the changes out to the server.

This can be frustrating at first. It feels like a slower, more complicated way to do something simple. But putting the right tooling and workflows into place means that people spend less time on routine, repetitive tasks. Instead, they focus their energies on handling exceptions and problems, and on making improvements and changes that will make exceptions and problems less common.

A good infrastructure-as-code workflow makes it easy to work indirectly on servers this way. But it can be challenging to get the workflow right and keep it that way. If the tooling and processes aren’t easy, then team members will often find they need to jump onto a server or configuration UI and make a quick change outside the automation tools. It might be necessary to do this in the early days of moving to infrastructure as code, but it should be a priority to address whatever ...