Understanding cost optimization automation and policies

Before we address how to control and monitor costs, we need to know the details of how our costs are distributed across business applications.

In large enterprises, it's pretty normal to have a few core applications that spend about 20% more than others and waste a little bit of compute or memory resources: core applications need to be instantly available, thus we need more contingency compared to the non-critical ones, despite every possible horizontal scaling architecture you can realize.

So, the first thing to know when you migrate an application from on-premises to a cloud infrastructure is how the applications work, their flow, dependencies, and so on, and identify all the resources involved in that application, to better concentrate at a later stage on the component's optimization.

Important Note:

If you don't know how the application is composed and how it works, it's quite unlikely you will...