SAP-C02 Question 519
Single answerA company runs a batch processing application on Amazon EC2 instances. The application workload is predictable and requires significant compute capacity for 6 hours daily. The company wants to minimize costs while ensuring the application can complete its tasks within the required timeframe. What is the MOST cost-effective solution?
- A
Use On-Demand Instances for the workload and terminate them after 6 hours daily.
- B
Use Reserved Instances for the workload to guarantee capacity and reduce costs.
- C
Use Spot Instances with a defined duration (Spot Blocks) for the workload.
- D
Use Auto Scaling with Spot Instances and configure a scaling policy to launch instances only during the 6-hour workload period.
Show answer and explanation
Correct answer: D
Explanation
The company can achieve the most cost-effective solution by leveraging Spot Instances, which offer substantial discounts compared to On-Demand pricing. By using Auto Scaling with Spot Instances and a scaling policy, the company can ensure that the required compute capacity is available during the 6-hour workload period while minimizing costs. This solution also provides flexibility to adjust resources dynamically based on demand.
- A. Incorrect.
Using On-Demand Instances ensures flexibility but is not the most cost-effective option for predictable workloads. On-Demand pricing is higher compared to other purchasing models.
- B. Incorrect.
Reserved Instances provide cost savings but are best suited for workloads that run consistently throughout the year, not for predictable, time-bound workloads like this one.
- C. Incorrect.
Spot Instances with defined duration (Spot Blocks) provide cost savings, but they are limited to a maximum duration of 6 hours at a time. While this could work, Spot Blocks are generally more expensive than regular Spot Instances and do not support scaling dynamically.
- D. Correct.
Using Auto Scaling with Spot Instances is the most cost-effective solution. Spot Instances provide significant cost savings, and Auto Scaling ensures that the required instances are launched during the 6-hour workload period. This approach also provides flexibility to scale based on demand.