The capacity is shown in the following diagram:
Running a system such as Kubernetes means that you can respond to additional demand for your services literally within the time it takes for your applications to start up. This process can even become automated with tools such as the Horizontal Pod Autoscaler (which we we will discuss in Chapter 8, Sorry My App Ate the Cluster).
When we couple this flexibility with the ability for us to launch new EC2 instances at will, capacity planning is much less involved than it might have been in the past. Kubernetes and AWS allow us to build applications that only consume the amount of resources that they need to be using at any given time. Rather than anticipating demand for our application and pre-committing to use resources, we can react to the usage requirements of our applications. Kubernetes finally allows us to deliver one...