DOP-C02 Question 169
Select 3You are managing an e-commerce application running on AWS that experiences fluctuating traffic throughout the day. The application is deployed behind an Application Load Balancer (ALB) and uses Amazon EC2 instances in an Auto Scaling group. To ensure optimal cost efficiency and performance, which metrics should you use to scale the EC2 instances effectively?
- A
Average CPU utilization of the EC2 instances
- B
Request count per target on the ALB
- C
Disk I/O operations of the EC2 instances
- D
Memory usage of the EC2 instances using a custom CloudWatch metric
- E
Network In/Out traffic of the EC2 instances
Show answer and explanation
Correct answers: A, B, D
Explanation
To scale services effectively, you need to choose metrics that directly correlate with the performance and load characteristics of your application. In this scenario, CPU utilization, request count per target (as it reflects incoming traffic to the ALB), and memory usage (via a custom metric) are the most appropriate metrics. Disk I/O and network traffic may provide insights into resource usage but are not as directly relevant for scaling in this particular case.
- A. Correct.
CPU utilization is a common metric for scaling because it directly correlates with the load on the EC2 instances. High CPU utilization indicates the need for more instances.
- B. Correct.
Request count per target on the ALB is a key metric that reflects the incoming traffic load on the instances. This is particularly relevant for web applications handling HTTP(S) requests.
- C. Incorrect.
Disk I/O operations are not typically used for scaling decisions in this scenario because they do not directly correlate with the e-commerce application's incoming traffic or load patterns.
- D. Correct.
Memory usage is often a critical factor for scaling certain applications. Since AWS does not provide memory metrics natively, using a custom CloudWatch metric for memory usage enables more precise scaling decisions.
- E. Incorrect.
Network In/Out traffic can be useful in some use cases, particularly for applications with high data transfer requirements, but it is less directly tied to the performance of an e-commerce application in this context.