Auto-scaling is a feature of cloud computing that enables users to automatically adjust the computing resources allocated to their applications based on demand. Auto-scaling allows users to maintain optimal application performance, while at the same time minimizing costs by only using the resources needed to handle the current workload.

In AWS, auto-scaling is achieved through the use of Auto Scaling Groups, which are collections of EC2 instances that are automatically launched and terminated based on changes in demand. Auto Scaling Groups monitor application load and adjust the number of EC2 instances in the group to match the current demand. When demand increases, Auto Scaling Groups automatically launch additional EC2 instances to handle the load, and when demand decreases, they terminate unnecessary instances to save costs.

Auto-scaling in AWS is highly configurable and can be based on a variety of metrics such as CPU utilization, network traffic, and custom metrics. Users can set up scaling policies that dictate how and when to scale, including minimum and maximum instance limits, and thresholds for scaling up and down.

Auto-scaling is a powerful feature of cloud computing that allows users to ensure that their applications are always available and responsive, while at the same time minimizing costs by only using the resources they need. It is an essential tool for building highly scalable and reliable applications in the cloud.