A company is designing a cloud architecture and wants to follow the Well-Architected Framework principle of 'stop guessing capacity.' Which AWS feature directly supports this principle?
Amazon EC2 Auto Scaling uses CloudWatch metrics such as CPU utilization, network traffic, or custom application metrics to automatically add or remove EC2 instances to maintain a desired performance level. With target tracking policies, you can set an average utilization target and the service continuously adjusts capacity to match actual demand, preventing both over-provisioning waste and under-provisioning slowdowns. This makes it the correct choice for automatically adjusting capacity based on demand.
Why this answer
Amazon EC2 Auto Scaling directly supports the 'stop guessing capacity' principle by automatically adjusting the number of EC2 instances in response to real-time demand using CloudWatch metrics (e.g., CPU utilization, memory). This eliminates the need to manually provision for peak loads, ensuring you only pay for what you need while maintaining performance.
Exam trap
The trap here is that candidates confuse 'stop guessing capacity' with cost optimization tools like Cost Explorer or Trusted Advisor, but the principle is specifically about dynamic scaling to match demand, not about analyzing or reducing costs after the fact.
How to eliminate wrong answers
Option A is wrong because AWS CloudFormation enables repeatable infrastructure deployments via templates, but it does not dynamically adjust capacity based on demand; it provisions static resources. Option C is wrong because AWS Trusted Advisor cost optimization checks provide recommendations to reduce costs (e.g., idle instances), but they do not automatically scale capacity to match workload changes. Option D is wrong because AWS Cost Explorer right-sizing recommendations analyze historical usage to suggest instance type changes, but they are advisory and do not provide real-time, automated scaling to handle fluctuating demand.