Building an evolutionary infrastructure using pipelines
Building a system on the cloud with infrastructure as code tools gets messy as it grows, especially when you have multiple teams working on it. How do you maintain multiple environments consistently? How do you minimize the chances that a change to one part of the system will break something else? Can you use automated testing for your infrastructure?
As with any code base, organizing your infrastructure code sensibly, writing automated tests, and automating the delivery of changes across environments can not only help you to deliver changes more quickly, but also more reliably. It should involve less effort, less stress, and fewer failures, so you can spend more time on higher value activities.