Performance Testing for Kubernetes Autoscaling
Published in
3 min readJan 5, 2020
We have been running applications on K8s for a year or so now but have only recently spent the time to truly leverage autoscaling. Previously we were autoscaling on CPU (75%) but our applications were never near this limit and we were running far too many instances.
CPU is often said to be a bad metric for autoscaling but why is this the case? It was only when people became aware of…