It would be a very nice addition (and a practical one), to have a dedicated chapter in the Starter Kit for Horizontal Pod Autoscaling (HPA) and Vertical Pod Autoscaling or VPA (in the future). What I suggest for now, is a generic name for the chapter, such as 09-autoscaling-application-workloads, and have HPAs discussed alongside metrics-server.
Then, we can also include Prometheus into the mix as an example, to get the required metrics for HPA (via prometheus-adapter). We already have it in place, and maybe most of people do as well.
In addition, HPA examples are required (and tooling), to simulate the load. Then, we can observe how the system behaves and how HPAs respond to external load, the outcome being scaling up or down application workloads.
Description
It would be a very nice addition (and a practical one), to have a dedicated chapter in the Starter Kit for
Horizontal Pod Autoscaling(HPA) andVertical Pod Autoscalingor VPA (in the future). What I suggest for now, is a generic name for the chapter, such as09-autoscaling-application-workloads, and have HPAs discussed alongside metrics-server.Then, we can also include
Prometheusinto the mix as an example, to get the required metrics forHPA(via prometheus-adapter). We already have it in place, and maybe most of people do as well.In addition, HPA examples are required (and tooling), to simulate the load. Then, we can observe how the system behaves and how HPAs respond to external load, the outcome being scaling up or down application workloads.