SAA-C03 Question 430
Single answerA startup is building a web application that experiences variable traffic patterns. During peak hours, the application requires additional compute resources, but during off-peak hours, the usage drops significantly. The company wants to minimize costs while ensuring the application is highly available and can scale automatically with demand. Which compute solution should the startup use?
- A
Launch On-Demand EC2 instances with Auto Scaling enabled.
- B
Use EC2 Spot Instances with Auto Scaling enabled.
- C
Deploy the application on AWS Elastic Beanstalk with an EC2 Auto Scaling group using Reserved Instances.
- D
Use AWS Fargate with an Application Load Balancer for the application.
Show answer and explanation
Correct answer: B
Explanation
EC2 Spot Instances with Auto Scaling enabled provide the best balance of cost optimization and the ability to scale with variable traffic. Spot Instances are ideal for fault-tolerant workloads and can significantly reduce costs compared to On-Demand Instances. Auto Scaling ensures that the application can handle peak traffic while minimizing resource usage during off-peak periods.
- A. Incorrect.
On-Demand EC2 instances provide flexibility but are not cost-optimized for variable traffic patterns since they charge at a fixed hourly rate regardless of demand. While Auto Scaling ensures availability, this option is not the most cost-effective solution.
- B. Correct.
EC2 Spot Instances are highly cost-optimized, offering up to 90% savings compared to On-Demand pricing. When combined with Auto Scaling, they can handle the application's variable traffic while minimizing costs. However, Spot Instances can be interrupted, so they are ideal for fault-tolerant and stateless workloads.
- C. Incorrect.
Reserved Instances provide a cost-saving option for predictable, steady-state workloads but are not cost-effective for variable traffic as they require upfront or long-term commitments. Elastic Beanstalk with Reserved Instances would not address the startup's need for cost optimization during off-peak periods.
- D. Incorrect.
AWS Fargate is a serverless compute option for running containers and can scale automatically with demand. However, it is typically more expensive than Spot Instances for workloads with variable traffic patterns, making it less cost-optimized in this scenario.