Use case 1: Web applications

Let’s start with the first example we used a few chapters ago to see autoscaling in action with a Monte Carlo PI simulation app. This time, to see Karpenter in action, we’ll send more traffic than before to the application, and we’ll scale based on the throughput (the number of requests per second). Moreover, as said before, we’re going to use tools such as k6 to automate the load tests. k6 is a tool designed to help you simulate how applications perform under heavy usage by generating load and tracking responses.

The information you get back will be more relevant in subsequent labs, but we want to start using it now to make the load test consistent.

The purpose of this first use case is to see KEDA adding more pods as the requests per second go up (I’ll dive deeper into this topic in this lab). Karpenter then launches only the nodes needed at that precise moment, and that will continue to happen until the load...