Autoscaling
Autoscaling automatically adds or removes capacity in response to demand, so you pay for resources when they are needed and release them when they are not.
Horizontal autoscaling changes the number of instances or containers based on metrics like CPU, request rate, or queue depth; scheduled scaling adjusts capacity on a known daily or weekly pattern. Done well, it is a primary defense against paying for idle capacity.
Autoscaling pairs naturally with spot capacity for the variable portion of a workload, keeping a smaller reserved or on-demand baseline for steady demand.
Related terms
Put the theory to work
Estimate and compare cloud costs across AWS, Azure, GCP, DigitalOcean, Hetzner, and Vultr with the free SpendArk calculator — no signup required.
Open the cloud cost calculator →Part of the SpendArk cloud cost & FinOps glossary.