Autoscaling

Two metric recommendations merge into the larger count: seven from CPU and twelve from average queue backlog, in the article’s arithmetic model.

The Horizontal Pod Autoscaler depends on metric semantics

A HorizontalPodAutoscaler turns configured metrics into replica recommendations. Choosing the metric and its aggregation is therefore part of application design. Increasing replicas should reduce the measured pressure or increase useful throughput; otherwise the controller may spend more money without relieving the bottleneck. Set explicit minimum and maximum counts, but do not confuse a cost bound with a performance guarantee. ...

A high recommendation at 240 seconds holds six replicas through a lower recommendation at 400 seconds; after that recommendation expires, the model drops to two at 541 seconds.

The horizontal autoscaler scales on demand

The HorizontalPodAutoscaler adjusts a scalable workload toward configured metric targets. Its output is a desired replica count, not a guarantee that additional Pods are already serving traffic. Scheduling, image startup, initialization, readiness, and available node capacity determine how quickly that recommendation becomes useful capacity. Design the control loop together with those delays. ...