Autoscaling and Cost Optimization on Kubernetes - From 0 to 100

Explore autoscaling and cost optimization techniques for Kubernetes in this comprehensive conference talk. Learn how to effectively scale applications and infrastructure dynamically based on demand while optimizing costs across various cluster environments. Discover the practical implementation of Cluster Autoscaler (CA), Horizontal Pod Autoscaler (HPA), and Vertical Pod Autoscaler (VPA) through real-world examples from Skyscanner. Delve into advanced features, including custom metrics for HPA, CA expanders, scaling from zero, and node group balancing. Gain insights into handling challenging scaling scenarios and leveraging community tools to enhance autoscaling capabilities. Master the art of maintaining performance and availability while optimizing resource utilization in Kubernetes environments.

Intro
Agenda
Overview
Autoscaling HPA
Matrix Types
Vpa recommender
Cluster Autoscaler
Scaled down process
Expanders
Common CA problems
GPU Optimizations
Other Community Projects