AWS Pricing Models: SAA-C03 Study Guide
In the AWS Certified Solutions Architect – Associate (SAA-C03) exam, understanding Cost Management and Optimization is critical. You aren’t just building for performance; you are building for the best value. Choosing the wrong pricing model can lead to architectural failure due to budget overruns.
The Analogy: Choosing a Commute
Think of AWS pricing models like choosing how to get to work:
- On-Demand: Taking a Taxi. You pay exactly for the trip you take, no commitment, but it’s the most expensive per mile.
- Savings Plans / Reserved Instances: A Monthly Transit Pass. You commit to paying for a certain amount of travel upfront or over time for a significant discount.
- Spot Instances: Standby Tickets. You get a massive discount on a flight, but if a full-fare passenger shows up, you might get bumped off the plane.
1. Core Pricing Models
On-Demand Instances
Pay for compute capacity by the hour or second depending on which instances you run. No long-term commitments or upfront payments are needed.
- Best for: Short-term, unpredictable workloads that cannot be interrupted.
- Exam Scenario: “A company is launching a new application with unknown traffic patterns.”
Savings Plans
A flexible pricing model that offers low prices on EC2, Lambda, and Fargate usage, in exchange for a commitment to a consistent amount of usage (measured in $/hour) for a 1 or 3-year term.
- Compute Savings Plans: Most flexible; applies to EC2 regardless of family, region, or OS. Also covers Fargate and Lambda.
- EC2 Instance Savings Plans: Commit to a specific instance family in a region (e.g., M5 in US-East-1) for deeper discounts.
Reserved Instances (RI)
Provide a significant discount (up to 72%) compared to On-Demand pricing. RIs are defined by Instance Type, Region, and Tenancy.
- Standard RI: Best for steady-state usage.
- Convertible RI: Allows you to change the RI attributes as long as the exchange results in the creation of RIs of equal or greater value.
- Scheduled RI: (Legacy) For workloads that run on a recurring schedule (e.g., every Thursday night).
Spot Instances
Take advantage of unused EC2 capacity in the AWS cloud. Spot Instances are available at up to a 90% discount compared to On-Demand prices.
- The Catch: AWS can reclaim the instance with a 2-minute notification if they need the capacity back.
- Best for: Stateless, fault-tolerant, or flexible applications (Batch processing, CI/CD, Big Data).
2. Comparison Table
| Model | Discount | Commitment | Flexibility | Best Use Case |
|---|---|---|---|---|
| On-Demand | 0% (Base) | None | Highest | New apps, Spiky traffic |
| Savings Plans | Up to 72% | 1 or 3 Years | High (Compute) | Steady state, Lambda/Fargate |
| Reserved Instances | Up to 72% | 1 or 3 Years | Medium | Steady state EC2, RDS |
| Spot Instances | Up to 90% | None | Low (Interruption) | Batch jobs, Data analysis |
3. Decision Matrix: If–Then Guide
- If the workload is fault-tolerant and requires massive scale at lowest cost → Choose Spot Instances.
- If the workload is steady-state and you use Fargate or Lambda → Choose Compute Savings Plans.
- If you need to reserve capacity in a specific Availability Zone → Choose Zonal Reserved Instances.
- If you have strict compliance/license requirements (BYOL) → Choose Dedicated Hosts.
Exam Tips and Gotchas
- Golden Nugget 1: Spot Instances are never the answer for long-running, stateful databases unless the question explicitly mentions a fault-tolerant distributed DB.
- Golden Nugget 2: Savings Plans do not provide a capacity reservation. They are a billing discount only. If you need a guarantee that the instance will be available, use On-Demand Capacity Reservations.
- Golden Nugget 3: Convertible RIs are better than Standard RIs when you anticipate changing your instance type (e.g., moving from C4 to C5) during the 3-year term.
- Golden Nugget 4: To maximize savings for a global company using multiple accounts, use Consolidated Billing in AWS Organizations to share RI/Savings Plan discounts across all accounts.
Topics covered:
Summary of key subtopics covered in this guide:
- On-Demand vs. Spot vs. Reserved vs. Savings Plans
- Standard vs. Convertible Reserved Instances
- Compute vs. EC2 Instance Savings Plans
- Dedicated Hosts vs. Dedicated Instances
- Capacity Reservations vs. Billing Discounts
- Cost-optimization tools: Cost Explorer and Compute Optimizer
Infographic: AWS Cost Architecture
AWS Organizations: Use Consolidated Billing to aggregate usage and reach volume discount tiers faster.
Cost Explorer: Visualize and forecast your spend. Identify underutilized RIs.
Compute Optimizer: Uses Machine Learning to recommend optimal instance types for your workload to avoid over-provisioning.
Auto Scaling: Combine Purchase Options. Use On-Demand for baseline and Spot for scaling peaks.
Scenario: A Big Data processing pipeline using EMR.
Solution: Use On-Demand for the Master Node (stability) and Spot Instances for Core/Task nodes (cost savings on fault-tolerant tasks).