Patterns for extending Puppet coverage
Let's analyze in detail what to do when we have to introduce Puppet on an existing infrastructure.
Some of these activities are needed also when working on a brand new project or when we want to refactor our existing Puppet configurations; consider them as a reference that you can adapt to your own needs and scenario.
The obvious assumption is that Puppet deployment on something that is already working and serving services has to be gradual. We have to understand the current situation, the expected goals, what is quick or easy to do, where and what to manage with Puppet, and our operational priorities.
Raising the bar, step by step
A gradual approach involves a step-by-step process that needs informed decisions, reiterations, and verification of what is done, evaluating each case according to its uniqueness.
We can extend our Puppet coverage using the following two axes:
Vertically: Working on all the existing nodes managed by Puppet and adding, step by...