SAA-C03 Design Cost-Optimized Architectures Practice Question
A media company runs a 24/7 ingestion API on EC2 behind an Application Load Balancer and a nightly transcoding job that can resume from checkpoints. The API fleet runs at roughly 65 percent CPU all day, while the batch workers sit idle most of the time. The company wants to cut compute cost without risking the API. Which two changes should they make? Select two.
⚠ Common exam trap
It's easy for candidates to confuse Savings Plans with Reserved Instances, or assume Dedicated Hosts are a cost-saving measure, when in fact they are a premium isolation feature; the key is recognizing that Spot Instances are ideal for fault-tolerant, checkpointable batch workloads, while a Compute Savings Plan covers the predictable baseline without locking into a specific instance type.
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
✓
Purchase a Compute Savings Plan for the always-on API fleet.
Option A is correct because the API fleet is an always-on, steady-state workload running 24/7 at ~65% CPU, which is exactly the profile a Compute Savings Plan discounts (up to 66% off On-Demand) while remaining flexible across instance families, sizes, Regions, and even Fargate/Lambda, so it lowers cost without changing capacity or risking the API. Option B is correct because the transcoding job is fault-tolerant and can resume from checkpoints, making it ideal for EC2 Spot Instances, which offer up to 90% off On-Demand; a Spot interruption only triggers a two-minute warning and the job simply resumes from its last checkpoint. Option C is wrong because Dedicated Hosts are for licensing/compliance or BYOL requirements and are typically more expensive, not a cost-optimization lever for a standard ALB-fronted API. Option D is wrong because buying Standard RIs and keeping batch workers running 24/7 pays for idle capacity the workload does not need, defeating the cost goal. Option E is wrong because raising the Auto Scaling minimum does not prevent Spot interruptions and would increase cost by keeping more idle workers running.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Purchase a Compute Savings Plan for the always-on API fleet.
Why this is correct
Correct. Compute Savings Plans discount steady usage across EC2 and other compute services without forcing a specific instance family. The API has predictable 24/7 demand, so a commitment fits the usage pattern and lowers cost safely.
- ✓
Move the transcoding workers to EC2 Spot Instances and checkpoint progress.
Why this is correct
Spot Instances cost far less than On-Demand but can be reclaimed with two minutes' notice. Checkpointing lets transcoding resume after interruption, so the interruption-tolerant batch workload absorbs the risk while the API fleet stays on stable capacity.
- ✗
Replace the API fleet with Dedicated Hosts to lock in lower rates.
Why it's wrong here
Dedicated Hosts bill for the entire physical server regardless of instance usage, so they cannot reduce cost for a fleet already running continuously at 65 percent CPU. They suit licensing models requiring socket or core visibility, such as bring-your-own Windows Server or Oracle licences.
- ✗
Buy Standard Reserved Instances for the batch workers and keep them running 24/7.
Why it's wrong here
Standard Reserved Instances commit to 24/7 running, yet the batch workers sit idle most of the day, so the discount applies to capacity that is not needed. Reserved Instances suit steady-state workloads running predictably around the clock, not intermittent nightly jobs.
- ✗
Increase the worker Auto Scaling minimum to prevent Spot interruptions.
Why it's wrong here
Raising the Auto Scaling minimum keeps idle workers provisioned continuously, increasing cost rather than cutting it, and does nothing to address Spot interruptions. A higher minimum suits workloads needing guaranteed baseline capacity to absorb sudden traffic spikes without scaling delay.
Quick reference
Cloud Service Model Comparison
| Model | You Manage | Provider Manages | Examples |
|---|---|---|---|
| IaaS | OS, runtime, apps, data | Hardware, hypervisor, networking | EC2, Azure VMs, GCP Compute Engine |
| PaaS | Apps and data | OS, runtime, middleware, hardware | Elastic Beanstalk, Azure App Service |
| SaaS | Data and settings only | Everything else | Microsoft 365, Salesforce, Workday |
| FaaS / Serverless | Function code only | Infra, scaling, runtime | Lambda, Azure Functions, Cloud Run |
| CaaS | Containers and apps | Kubernetes, OS, hardware | EKS, AKS, GKE |
Go deeper
Related to this question
About these practice questions
One of 935 original SAA-C03 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 →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This SAA-C03 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 SAA-C03 exam.