Using Spot Instances with On-Demand Fallback for Batch Processing
A SysOps team manages a fleet of EC2 instances used for batch processing. The workload runs daily, taking approximately 6 hours. The instances are launched via an Auto Scaling group using On-Demand instances from a custom AMI. The team has noticed that while the instances are running, the CPU utilization is moderate, but the memory usage is high. After the batch completes, the instances are terminated. The team wants to reduce costs without changing the architecture. Which solution would be MOST cost-effective?
Quick Answer
The answer is to modify the Auto Scaling group to use a mixed instances policy with a percentage of Spot Instances and a fallback to On-Demand. This is the most cost-effective solution because Spot Instances offer up to 90% discounts for fault-tolerant, interruptible workloads like batch processing, and the On-Demand fallback ensures the job completes even if Spot capacity is reclaimed. On the AWS Certified SysOps Administrator Associate SOA-C02 exam, this scenario tests your understanding of cost optimization for transient, high-memory workloads—a common trap is choosing Reserved Instances or Savings Plans, which require long-term commitments unsuitable for sporadic 6-hour daily runs. Remember that Spot Instances are ideal for stateless, batch jobs that can handle interruptions, while the mixed instances policy provides both savings and reliability. A useful memory tip: "Spot for savings, On-Demand for safety" when your batch can tolerate a fallback.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Modify the Auto Scaling group to use a mixed instances policy with a percentage of Spot Instances and a fallback to On-Demand.
Spot Instances can provide significant discounts (up to 90%) for fault-tolerant batch workloads, and the team can use a mixed instances policy with On-Demand as a fallback. Option B (memory optimized instances) would likely increase cost. Option C (reserved instances) is not suitable for short-lived, sporadic workloads. Option D (savings plan) could help but still requires committing to a consistent amount, which may not align with the varying batch size. Option A offers the most flexibility and cost savings.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Modify the Auto Scaling group to use a mixed instances policy with a percentage of Spot Instances and a fallback to On-Demand.
Why this is correct
Spot Instances are ideal for fault-tolerant workloads like batch processing, offering large discounts.
- ✗
Use a Compute Savings Plan covering the entire compute usage across the account.
Why it's wrong here
Savings Plans require a commitment; the batch workload may not have consistent enough usage to benefit optimally.
- ✗
Change the instance type to a memory-optimized family to reduce the number of instances needed.
Why it's wrong here
Memory-optimized instances are more expensive per vCPU; the issue is memory usage, but the workload is moderate CPU, so switching may increase cost.
- ✗
Purchase Reserved Instances for the expected daily usage to get a lower hourly rate.
Why it's wrong here
Reserved Instances are best for steady-state, predictable workloads, not for short daily batches that may vary in size.
Go deeper
Related to this question
About these practice questions
One of 247 original SOA-C02 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
Same concept, more angles
3 more ways this is tested on SOA-C02
These questions test the same concept from different angles. Work through them to make sure you can recognise it however the exam phrases it.
Variation 1. A company runs a batch processing job every Saturday for 3 hours. The job can be interrupted and resumed at any point. The SysOps administrator wants to minimize compute costs for this workload. Which EC2 purchasing option should the administrator use?
easy- A.On-Demand Instances
- B.Reserved Instances
- ✓ C.Spot Instances
- D.Dedicated Hosts
Why C: Spot Instances are ideal for fault-tolerant, interruptible workloads like batch processing that can be resumed. They offer significant cost savings (up to 90% compared to On-Demand) because they use spare EC2 capacity, which can be reclaimed by AWS with a 2-minute interruption notice. Since the job runs only 3 hours weekly and can be interrupted and resumed, Spot Instances minimize compute costs effectively.
Variation 2. A company runs a batch processing job every night that takes 2 hours on a single m5.xlarge EC2 instance. The job is fault-tolerant and can be interrupted. The SysOps administrator wants to reduce costs. Which solution is MOST cost-effective?
medium- A.Use an On-Demand instance and set up a CloudWatch alarm to stop it when the job completes.
- ✓ B.Use a Spot Instance with a Spot Fleet that includes a fallback to On-Demand if Spot is not available.
- C.Use a Dedicated Host to run the job.
- D.Purchase a Reserved Instance for the m5.xlarge instance.
Why B: The most cost-effective solution is to use a Spot Instance with a Spot Fleet that includes a fallback to On-Demand if Spot is not available (Option B). The job is fault-tolerant and can be interrupted, making it ideal for Spot Instances which offer significant cost savings (up to 90% compared to On-Demand). The Spot Fleet with an On-Demand fallback ensures the job completes even if Spot capacity is unavailable. Option A (On-Demand with CloudWatch alarm) does not reduce costs since On-Demand is more expensive. Option C (Dedicated Host) is costly and unnecessary for a batch job. Option D (Reserved Instance) requires a 1- or 3-year commitment and is not cost-effective for a 2-hour daily job.
Variation 3. A company runs a batch processing job on a single EC2 instance that runs for 2 hours every night. The job is fault-tolerant and can be interrupted. The SysOps administrator wants to minimize compute costs. What is the MOST cost-effective solution?
easy- A.Use an On-Demand instance to ensure the job runs every night.
- B.Purchase a Reserved Instance for 1 year to get a discount.
- C.Launch a Dedicated Host to ensure consistent performance.
- ✓ D.Use a Spot Instance that can be interrupted but is significantly cheaper.
Why D: A Spot Instance is the most cost-effective choice because the job is fault-tolerant and can be interrupted, allowing you to leverage unused AWS EC2 capacity at up to 90% discount compared to On-Demand pricing. Since the job runs for only 2 hours nightly and can handle interruptions, Spot Instances provide the lowest compute cost while meeting the workload requirements.
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This SOA-C02 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the SOA-C02 exam.