Question #865
A media streaming company needs to optimize the cost of its AWS workloads. The company uses a dedicated account for its production streaming application and a separate account for its data processing application. The streaming application runs on Amazon EC2 instances that must remain available 24/7 throughout the year. The data processing application analyzes content metadata stored in Amazon S3 and can tolerate interruptions. Which solution meets these requirements MOST cost-effectively?
Purchase a Compute Savings Plan for the streaming application instances. Use On-Demand Instances for the data processing application.
Purchase a Compute Savings Plan for the streaming application instances. Use Spot Instances for the data processing application.
Use Spot Instances for both the streaming and data processing applications. Deploy AWS Service Catalog to provision discounted resources.
Use On-Demand Instances for the streaming application. Use Spot Instances for the data processing application. Deploy AWS Service Catalog to provision discounted resources.
Explanation
Answer B is correct because:
1. Compute Savings Plans are ideal for the streaming application (24/7 uptime), offering significant discounts compared to On-Demand pricing while ensuring availability.
2. Spot Instances are optimal for the data processing application (tolerates interruptions), reducing costs by up to 90% compared to On-Demand.
Why other options are incorrect:
- A: On-Demand Instances for data processing are more expensive than Spot.
- C: Spot Instances are unsuitable for the streaming application due to potential interruptions.
- D: On-Demand Instances for streaming are costlier than Savings Plans.
Key Points:
- Use Savings Plans/Reserved Instances for steady-state workloads.
- Use Spot Instances for fault-tolerant or flexible workloads.
- AWS Service Catalog (options C/D) is unrelated to cost optimization in this scenario.
Answer
The correct answer is: B