Introduction to Autoscaling
Talkat KubeCon + CloudNativeCon North America 2019 · on Tuesday, 2019-11-19· at 14:25· Room 28ABCDE - San Diego Convention Center, Upper Level·
summary
Explains how Kubernetes autoscaling actually works: Horizontal Pod Autoscaler adjusts replicas from CPU or custom metrics toward a target utilization (examples: set target <65% for a 50% two‑minute spike), Vertical Pod Autoscaler runs in three modes—recommendation, update on new pods, and auto that recreates pods to change requests—and Cluster Autoscaler adds/removes nodes via scheduling simulation and compaction, raising cluster requested utilization roughly 10% (from ~50% to ~60%).
AI-written summary · about
chapters (7)
- 0:00Introduction
- 2:35Horizontal Pot Autoscaling
- 8:42Vertical Pot Autoscaling
- 15:40Cluster Autoscaling
- 21:14VPA
- 23:17Cluster Autoscaler
- 24:53Questions
