# Autoscaling

A step of the [Cloud application](https://gradus.kazet.blog/tree/cloud-application) tree on Gradus, in the Serving & compute area: HTTP, workers, scaling.

Page: https://gradus.kazet.blog/tree/cloud-application/step/autoscaling

Scale instances on real signals such as queue depth or request latency, with sane floors and ceilings.

Usually comes after: [Container orchestration](https://gradus.kazet.blog/tree/cloud-application/step/container-orchestration).

## Prompt for your AI assistant

```
Configure horizontal autoscaling for the application deployment based on CPU and a custom metric such as queue depth or request latency, with minimum and maximum replica counts.
```

Every step of the tree: https://gradus.kazet.blog/tree/cloud-application/llms.txt