SOA-C02 Question 323
Single answerAn organization runs a web application on an Amazon EC2 instance that has a t2.micro instance type. Over time, they notice performance degradation during periods of high traffic. Amazon CloudWatch metrics indicate that the CPU utilization consistently exceeds 90% during peak hours. As a SysOps Administrator, what is the most appropriate action to take to address the performance issue?
- A
Upgrade the instance type to a t3.medium.
- B
Enable Auto Scaling for the application and add more t2.micro instances.
- C
Migrate the application to AWS Lambda to handle the traffic spikes.
- D
Upgrade the instance type to an m5.large.
Show answer and explanation
Correct answer: A
Explanation
The web application is experiencing high CPU utilization, indicating that the current instance type (t2.micro) cannot handle the workload effectively during peak traffic. Upgrading to a t3.medium instance offers better baseline performance and burst capacity within the same instance family, making it the most cost-effective and scalable choice for addressing the performance issue. More drastic changes, such as refactoring to AWS Lambda or over-provisioning with a larger instance type (m5.large), are unnecessary at this stage.
- A. Correct.
Upgrading the instance to a t3.medium is a cost-effective solution that provides better performance than a t2.micro and includes burstable performance. This is a good first step in addressing the issue without over-provisioning.
- B. Incorrect.
Enabling Auto Scaling and adding more t2.micro instances might work, but it would be more complex to manage and could still face performance limitations due to the instance type's CPU credits. This is not the most efficient initial recommendation.
- C. Incorrect.
Migrating the application to AWS Lambda could be an option for handling spikes, but it would require significant refactoring of the application. This is not the most immediate or appropriate solution to address the current performance issue.
- D. Incorrect.
Upgrading the instance to an m5.large provides significantly more resources, but it may be over-provisioned and unnecessarily expensive for the application's current needs.