DOP-C02 Question 276
Select 3Your company runs a critical e-commerce application on AWS that experiences unpredictable traffic spikes during flash sales. The architecture includes an ECS cluster with Fargate tasks for the application, an RDS database for transactional data, and DynamoDB for session storage. You must ensure that the application scales efficiently to handle the traffic spikes with minimal manual intervention. Which of the following configurations will help achieve this goal?
- A
Configure an ECS Service Auto Scaling policy for the Fargate tasks to scale based on CPU utilization.
- B
Enable RDS storage auto scaling to accommodate increased database demands during traffic spikes.
- C
Set up DynamoDB Auto Scaling to adjust read and write capacity units based on demand.
- D
Use an EC2 Auto Scaling group with custom scaling policies to manage the ECS Fargate tasks.
- E
Deploy the Kubernetes Cluster Autoscaler (KCA) to manage ECS task scaling.
Show answer and explanation
Correct answers: A, B, C
Explanation
To handle unpredictable traffic spikes efficiently, it is crucial to implement auto scaling mechanisms specific to each service in your architecture. ECS Service Auto Scaling ensures Fargate tasks scale based on demand, RDS storage auto scaling automatically handles database storage needs, and DynamoDB Auto Scaling optimizes throughput for session storage. Each of these services has native scaling features that align with the requirements of the scenario, while unrelated solutions like EC2 Auto Scaling or Kubernetes Cluster Autoscaler are not applicable for this use case.
- A. Correct.
Correct: ECS Service Auto Scaling allows Fargate tasks to scale dynamically based on metrics like CPU or memory utilization, ensuring adequate resources during traffic spikes.
- B. Correct.
Correct: Enabling RDS storage auto scaling automatically adjusts the storage capacity of your database, preventing downtime or performance issues during high traffic periods.
- C. Correct.
Correct: DynamoDB Auto Scaling adjusts the provisioned read and write capacity units based on traffic patterns, ensuring smooth application performance during sudden spikes.
- D. Incorrect.
Incorrect: Fargate tasks are serverless and do not run on EC2 instances, so an EC2 Auto Scaling group is not applicable for managing them.
- E. Incorrect.
Incorrect: The Kubernetes Cluster Autoscaler (KCA) is specific to Amazon EKS and Kubernetes clusters, not ECS. ECS has its own scaling mechanisms.