Implementing elasticity with Amazon Auto Scaling
One of the most amazing services on AWS is the ability to automatically scale your workloads when demand increases and then scale back in when demand drops. This service is offered as part of various core technologies—for example, computing services such as EC2 and database services such as DynamoDB.
Automatic scaling in response to a particular condition such as an increase in demand (for example, when average CPU utilization across your fleet of EC2 instances goes above a threshold such as 70%) can help provision additional capacity when it is required most. However, you are not stuck with the new size of your fleet. You can configure Auto Scaling so that if demand drops below a specific threshold value, it will terminate EC2 instances and therefore save on costs. Let's look at Auto Scaling for EC2 instances in detail next.
Auto Scaling is a regional service, and you can scale across AZs within a given Region, allowing...
