SOA-C02 Question 294
Select 3Your company has several EC2 instances running non-critical batch processing tasks. These instances are running 24/7, but you notice that the tasks only run for a few hours each day. As a SysOps Administrator, what cost optimization strategies should you implement to reduce the EC2 costs while ensuring the tasks are completed successfully?
- A
Use Spot Instances to run the batch processing tasks.
- B
Schedule the EC2 instances to start and stop during the hours the tasks are active using AWS Instance Scheduler.
- C
Migrate the workload to Reserved Instances to lock in a lower hourly rate.
- D
Leverage AWS Auto Scaling to automatically adjust the number of instances based on demand.
- E
Move the batch processing tasks to AWS Lambda to take advantage of serverless pricing.
Show answer and explanation
Correct answers: A, B, E
Explanation
To optimize costs for non-critical batch processing tasks that run sporadically, using Spot Instances and scheduling instance start/stop times are effective strategies to reduce costs. Additionally, migrating the workload to AWS Lambda can further save costs by eliminating the need for EC2 instances entirely. Reserved Instances and Auto Scaling are better suited for other types of workloads and do not align well with the described scenario.
- A. Correct.
Spot Instances are a cost-effective solution for non-critical workloads such as batch processing, as they offer significant cost savings compared to On-Demand Instances. However, these instances can be interrupted, so they are suitable only if interruptions are tolerable or the tasks can be resumed.
- B. Correct.
Using AWS Instance Scheduler to stop and start EC2 instances based on task schedules ensures that you only pay for the time the instances are active, which reduces costs.
- C. Incorrect.
Reserved Instances provide cost savings on predictable workloads with consistent usage patterns. However, they are not ideal for workloads with sporadic usage, such as batch processing that only runs for a few hours a day.
- D. Incorrect.
While Auto Scaling is useful for dynamic workloads with fluctuating demand, it may not be the best fit for batch processing tasks with predictable schedules. Auto Scaling doesn't directly address the need to reduce costs for tasks with sporadic, predictable usage patterns.
- E. Correct.
AWS Lambda is a serverless solution that charges only for the compute time consumed. It is an excellent choice for running infrequent, short-lived tasks like batch processing, enabling further cost optimization.