With autoscaling, you no longer have to constantly run your application on the resources needed for peak demand. Instead, you can run on just one small server when you are seeing low traffic, and then scale up as traffic increases.
Moreover, if you run a bunch of microservices in your cluster, you can take advantage of service-level autoscaling. Since all services run on a shared set of instances, the cluster's compute power can be distributed proportional to the amount of resources a service requires. This allows you to scale individual services, without having to necessarily scale up the entire cluster.
If you use Breeze, we will automatically configure autoscaling for both your individual services, as well as your cluster (scaling of the underlying machines). This gives you maximum flexibility, at minimum effort.
As with everything else managed by Breeze, autoscaling is easy to manually override through the standard Kubernetes API, and Breeze will detect your manual overrides and respect those changes.