You can define the autoscaling feature of many services as follows:
Autoscaling is a feature that allows a service, a machine, or an application to automatically scale up or out based on predefined parameters, like CPU utilization, memory used, or artificial factors, like throughput units, or worker utilization.
In general, you can describe autoscaling as follows:
The preceding diagram can be described as follows:
- A resource accepts incoming requests as normal
- Simultaneously there is an entity that monitors a resource—it checks it against the scaling rules and decides whether a scaling operation is required
- An entity takes a decision ...
