Autoscaling in Kubernetes by Maciej Pytel

How many replicas do I need to serve my app? How big my cluster should be? How to make sure that the bill for compute resources is minimal while still be ready to serve all the traffic that can come?

This kind of questions bother most of the application owners or cluster administrators. This demo session will show how to forget about such problems. We will do a deep dive into the mysterious of autoscaling in kubernetes and show how it can simplify life of kubernetes users.

[wehappyfewkd]

Hey, where can I configure how often the HPA scrapes the metrics from? Maciej says every 30-50 secs and that it is configurable at 15:51.