Try 10 focused AWS SAA-C03 questions on Design Cost-Optimized Architectures, with explanations, then continue with IT Mastery.
Open the matching IT Mastery practice page for timed mocks, topic drills, progress tracking, explanations, and full practice.
| Field | Detail |
|---|---|
| Exam route | AWS SAA-C03 |
| Topic area | Design Cost-Optimized Architectures |
| Blueprint weight | 20% |
| Page purpose | Focused sample questions before returning to mixed practice |
Use this page to isolate Design Cost-Optimized Architectures for AWS SAA-C03. Work through the 10 questions first, then review the explanations and return to mixed practice in IT Mastery.
| Pass | What to do | What to record |
|---|---|---|
| First attempt | Answer without checking the explanation first. | The fact, rule, calculation, or judgment point that controlled your answer. |
| Review | Read the explanation even when you were correct. | Why the best answer is stronger than the closest distractor. |
| Repair | Repeat only missed or uncertain items after a short break. | The pattern behind misses, not the answer letter. |
| Transfer | Return to mixed practice once the topic feels stable. | Whether the same skill holds up when the topic is no longer obvious. |
Blueprint context: 20% of the practice outline. A focused topic score can overstate readiness if you recognize the pattern too quickly, so use it as repair work before timed mixed sets.
These questions are original IT Mastery practice items aligned to this topic area. They are designed for self-assessment and are not official exam questions.
Topic: Design Cost-Optimized Architectures
Which of the following describe valid throttling techniques on AWS that help limit unnecessary data transfer while preserving required traffic? (Select THREE.)
Options:
A. Disable Amazon CloudFront caching by setting all cache-control headers to no-store so that every user request is sent to the origin, making overuse easier to detect and control.
B. Configure AWS WAF rate-based rules on an Application Load Balancer or Amazon CloudFront to automatically block IP addresses that exceed a safe request rate for non-essential paths.
C. Implement client-side exponential backoff with a cap on maximum retries so failing calls do not continuously resend traffic across the network.
D. Use larger EC2 instances with higher network bandwidth so bulk transfers finish faster and therefore incur fewer data transfer charges.
E. Rely on security groups, which automatically slow down traffic when bandwidth usage becomes high, to throttle excess requests.
F. Apply Amazon API Gateway usage plans and per-API key throttling to rate limit non-critical clients before requests reach backend services.
Correct answers: B, C and F
Explanation: Throttling strategies help control the rate of requests or retries so that non-critical or abusive traffic does not consume unnecessary bandwidth or drive up data transfer costs. On AWS, throttling is typically implemented at managed service boundaries (such as API Gateway, CloudFront, or an ALB with AWS WAF) and in client behavior (retry policies), rather than relying on low-level network constructs.
Effective cost-optimized designs distinguish between critical and non-critical traffic. They allow required workloads to proceed while applying rate limits or backoff to test users, partner integrations, scrapers, or failing clients to avoid wasteful data transfer. The correct options all describe such rate limiting or retry-control mechanisms. The incorrect options either misunderstand how a service works or would actually increase data transfer usage rather than reduce it.
Topic: Design Cost-Optimized Architectures
A company is testing four alternative load balancing designs for a new HTTP web application. The application requires path-based routing and a 95th percentile latency under 40 ms. The company wants the most cost-optimized design that still meets these requirements.
Based on the following exhibit, which design should the solutions architect choose?
Exhibit:
| Design | Supports path-based routing? | 95th percentile latency (ms) | Estimated monthly cost |
|---|---|---|---|
| A | Yes | 35 | USD 1,200 |
| B | Yes | 45 | USD 900 |
| C | No | 20 | USD 600 |
| D | Yes | 30 | USD 1,000 |
Options:
A. Implement Design C from the exhibit.
B. Implement Design B from the exhibit.
C. Implement Design D from the exhibit.
D. Implement Design A from the exhibit.
Best answer: C
Explanation: The stem states two hard requirements: the design must support path-based routing and must achieve a 95th percentile latency under 40 ms. Among the designs that meet both, the company wants the lowest estimated monthly cost.
From the exhibit:
First, filter by functionality and performance. Designs A and D have path-based routing and latency below 40 ms. Design B fails latency, and Design C fails the functional requirement. Among A and D, Design D has a lower monthly cost (USD 1,000 vs. USD 1,200). Therefore, Design D is the most cost-optimized load balancing design that still meets all stated requirements.
This reflects a key cost-optimization pattern: never choose the absolute cheapest option if it violates functional or performance requirements; instead, minimize cost within the subset of designs that satisfy constraints.
Topic: Design Cost-Optimized Architectures
Which statement about data transfer costs for AWS load balancers is correct and should be considered when designing a cost-optimized architecture using cross-zone load balancing?
Options:
A. Network Load Balancers include cross-zone data transfer at no additional charge, but Application Load Balancers incur per-GB charges for cross-zone traffic between Availability Zones.
B. Gateway Load Balancers and Application Load Balancers both charge per-GB for cross-zone data transfer, but Network Load Balancers do not support cross-zone load balancing.
C. All Elastic Load Balancing types charge the same data transfer rate for cross-zone and same-Availability Zone traffic, so cross-zone load balancing has no additional cost impact.
D. Application Load Balancers include cross-zone data transfer at no additional charge, while Network Load Balancers and Gateway Load Balancers incur cross-zone data transfer charges when traffic is sent across Availability Zones.
Best answer: D
Explanation: When designing cost-optimized network architectures with Elastic Load Balancing, it is important to understand how cross-zone load balancing is priced for each load balancer type.
For Application Load Balancers (ALB), AWS pricing documentation states that there is no additional cross-zone data transfer charge. You still pay for the ALB usage and for data transferred out of the AWS Region as usual, but enabling cross-zone load balancing on an ALB does not add extra inter-AZ data transfer fees.
For Network Load Balancers (NLB) and Gateway Load Balancers (GWLB), cross-zone load balancing does incur inter-Availability Zone (inter-AZ) data transfer charges whenever the load balancer node in one AZ forwards traffic to a target in another AZ. This means that for NLB and GWLB, enabling cross-zone load balancing can materially increase data transfer costs if a significant portion of traffic crosses AZ boundaries.
Therefore, the correct factual statement is that ALBs include cross-zone data transfer at no extra charge, while NLBs and GWLBs incur cross-zone inter-AZ data transfer costs. This distinction is key when comparing architectures that use ALB versus NLB or GWLB under heavy, multi-AZ traffic patterns.
Topic: Design Cost-Optimized Architectures
Which of the following statements about designing cost-optimized data retention and archival policies on AWS are true? (Select TWO.)
Options:
A. To control costs, it is best practice to retain Amazon RDS automated backups indefinitely so that any past database state can be restored on demand.
B. Using Amazon S3 Lifecycle policies to transition older, rarely accessed objects to archival classes such as S3 Glacier Deep Archive and then permanently delete them when no longer required helps minimize long-term storage cost.
C. Because S3 Glacier retrievals incur additional charges, it is more cost-effective to keep all historical log files in S3 Standard for at least 10 years instead of expiring or archiving them.
D. Enabling Time to Live (TTL) on DynamoDB tables that store expiring data (such as session or event data) automatically removes old items and prevents unbounded storage growth.
E. S3 Intelligent-Tiering automatically moves any S3 Standard bucket to the lowest-cost Glacier storage class after a year, so separate lifecycle policies and deletion schedules are usually unnecessary.
Correct answers: B and D
Explanation: Cost-optimized data retention on AWS means keeping data only as long as it has business, legal, or compliance value, and using the lowest-cost storage tier that still meets access and durability requirements. AWS provides mechanisms like S3 Lifecycle policies and DynamoDB TTL to automate data movement and deletion according to retention rules.
For object storage, S3 Lifecycle policies can transition data between storage classes (for example, from S3 Standard to S3 Glacier Deep Archive) as it ages, and can eventually expire objects entirely. This prevents large, rarely used datasets from remaining in high-cost storage indefinitely.
For NoSQL workloads, DynamoDB TTL allows you to mark items with an expiration time. DynamoDB then automatically deletes expired items in the background, which is ideal for data such as sessions, events, and temporary records that lose value after a specific time. This avoids unbounded table growth and unnecessary ongoing storage charges.
By contrast, retaining everything forever “just in case”—such as keeping RDS backups indefinitely or never expiring S3 data—tends to increase cost without clear benefit and can even increase operational risk and complexity. Similarly, misunderstanding how features like S3 Intelligent-Tiering work can lead to incorrect assumptions that retention and archival happen automatically without explicit configuration.
Topic: Design Cost-Optimized Architectures
Which of the following statements about horizontal and vertical scaling on AWS are true? (Select THREE.)
Options:
A. Vertical scaling is often simpler to implement for legacy monolithic applications that cannot easily be distributed across multiple instances.
B. For workloads with highly variable or unpredictable traffic, horizontal scaling with Auto Scaling is usually more cost-efficient than permanently running a very large instance.
C. Vertical scaling is generally unconstrained on AWS; you can keep increasing instance size without practical limits.
D. Horizontal scaling typically improves availability because the workload can continue to run if a single instance fails.
E. For any production workload requiring high availability, vertical scaling on a single large instance is preferred over horizontal scaling on multiple smaller instances.
F. Horizontal scaling always increases both performance and cost compared to vertical scaling, because more instances are required.
Correct answers: A, B and D
Explanation: Horizontal scaling means adding or removing instances to change total capacity, typically behind a load balancer or as part of an Auto Scaling group. This approach spreads load across multiple instances and AZs, improving availability and enabling elastic capacity that tracks demand.
Vertical scaling means changing the size of an individual instance (for example, moving from t3.medium to t3.large). This is often simpler for legacy or monolithic applications that are not designed to run on multiple instances. However, it has hard limits (maximum instance sizes) and usually provides less fault tolerance because it still relies on a single node.
From a cost perspective, horizontal scaling pairs well with elastic workloads. Auto Scaling can add instances only when traffic increases and remove them when traffic drops, so you pay proportionally to demand instead of running a large instance at full capacity all the time. Vertical scaling can be cost-effective for steady, predictable workloads that fit comfortably within a single instance and where simplicity is a priority.
In AWS, the Well-Architected approach for high availability is to use horizontal scaling across multiple smaller instances spread across multiple Availability Zones, combined with monitoring and automatic scaling policies. Vertical scaling is still useful, but mainly as a simple way to increase capacity for workloads that cannot easily be distributed or as a one-time right-sizing step.
Topic: Design Cost-Optimized Architectures
Which of the following statements about using AWS cost management tools to analyze network-related spend are true? (Select THREE.)
Options:
A. In Cost Explorer, you can use the built-in Data transfer usage type group to review historical data transfer spend and to forecast future data transfer charges based on past trends.
B. Enabling VPC Flow Logs automatically adds per-flow cost information so you can directly see how much you paid for each connection without using any other billing tools.
C. The AWS Cost and Usage Report (CUR) can be configured to deliver hourly line items to Amazon S3, which you can query to identify the specific resources responsible for the largest data transfer charges.
D. To analyze network spend, you must first enable AWS Budgets; otherwise, network-related costs are excluded from both Cost Explorer and the Cost and Usage Report.
E. In Cost Explorer, you can group costs by Usage Type to separate internet data transfer, inter-AZ data transfer, and cross-Region data transfer charges.
Correct answers: A, C and E
Explanation: Network-related charges can represent a large portion of an AWS bill, especially for workloads with heavy data transfer, cross-AZ traffic, or cross-Region replication. AWS provides multiple cost management tools to understand and optimize this spend.
Cost Explorer is a visual tool that lets you filter and group costs by dimensions such as Service, Usage Type, and Usage Type Group (for example, Data transfer). Grouping by Usage Type is particularly helpful for distinguishing between types of data transfer, such as internet egress, inter-AZ traffic, and cross-Region transfers. Cost Explorer also offers forecasting based on historical data, which you can apply specifically to data transfer usage.
The AWS Cost and Usage Report (CUR) is a highly detailed dataset delivered to Amazon S3. It can be configured with hourly (or daily) granularity and includes line items for each usage dimension, often down to individual resources. By querying CUR (for example, with Amazon Athena), you can identify exactly which instances, load balancers, or other resources account for the majority of your data transfer costs.
Operational logs like VPC Flow Logs are complementary: they show traffic behavior (who is talking to whom and how much), but they do not contain cost figures. To attach dollar amounts to that traffic, you must still use billing tools such as Cost Explorer and CUR. AWS Budgets, while useful for alerts and governance, does not control whether network costs appear in billing data; network charges are always present in Cost Explorer and CUR when they occur.
Topic: Design Cost-Optimized Architectures
Which of the following statements about Amazon S3 Requester Pays buckets is INCORRECT?
Options:
A. Objects in a Requester Pays bucket can be accessed anonymously over the internet if the bucket policy allows public reads, and AWS bills requesters based on their IP address.
B. S3 Requester Pays is useful for shared datasets where many external AWS accounts download objects, allowing each consumer to pay its own data transfer and request costs.
C. To successfully read from a Requester Pays bucket, client applications must explicitly indicate that the requester agrees to pay, for example by including the appropriate Requester Pays header or parameter in the request.
D. In a Requester Pays bucket, the bucket owner continues to pay for object storage, while requesters pay for data transfer and request charges when they access the data.
Best answer: A
Explanation: Amazon S3 Requester Pays buckets shift certain costs from the bucket owner to the data consumer. The bucket owner continues to pay for storage, but the requester pays for data transfer and request charges when accessing objects. This model is useful for shared datasets where many different AWS accounts download the data.
However, Requester Pays has important constraints. It does not support anonymous access. Every requester must be authenticated with valid AWS credentials, and the billing is tied to the requester’s AWS account, not to an IP address. In addition, clients must explicitly acknowledge that they are willing to pay, typically by including a specific Requester Pays header or parameter in the request.
The incorrect statement suggests that anonymous public access is possible and that AWS charges requesters based on their IP address. Both of these claims contradict the documented behavior of S3 Requester Pays buckets.
Topic: Design Cost-Optimized Architectures
Which THREE statements about using Amazon Route 53 to design cost-optimized network architectures are true? (Select THREE.)
Options:
A. Latency-based routing in Route 53 lets a single DNS name route users to the Region with the best latency, eliminating the need for custom geo-IP or latency-probing code in the application.
B. Replacing the default AmazonProvidedDNS in a VPC with Route 53 Resolver endpoints will always lower DNS query costs for internal workloads.
C. Route 53 failover routing with health checks can shift traffic between primary and DR endpoints, avoiding the need to build and operate a custom health-check and failover system.
D. Running custom health-check agents on EC2 instances is typically more cost-effective than using managed Route 53 health checks, because Route 53 health checks require you to provision and maintain additional compute resources.
E. Route 53 can create health checks against on-premises or third-party web endpoints, enabling low-cost DNS failover for hybrid workloads.
Correct answers: A, C and E
Explanation: Amazon Route 53 provides managed DNS and health-check features that can replace custom-built solutions, often reducing both direct infrastructure spend and ongoing operational effort.
Using Route 53 failover routing with health checks lets you implement DNS-level failover between a primary and a DR endpoint. Because the health checks and failover logic are managed, you avoid building your own heartbeat system, monitoring daemons, or scripts, and you do not need extra EC2 instances dedicated to monitoring.
Latency-based routing is another Route 53 feature that simplifies global architectures. With one DNS name, Route 53 can direct users to the Region that offers the lowest latency, based on AWS measurements. This avoids writing and maintaining custom geolocation or latency-testing code in your application or at the edge.
Route 53 health checks can target any publicly reachable endpoint, including on-premises or third-party services. This enables cost-effective hybrid failover: you can use DNS to shift traffic away from a failing data center or SaaS endpoint without deploying and operating a separate monitoring platform.
By contrast, the default AmazonProvidedDNS inside a VPC is already included at no extra charge, and adding custom Route 53 Resolver endpoints incurs additional hourly costs. Likewise, implementing your own health-checking layer on EC2 typically costs more in infrastructure and operations than using managed Route 53 health checks.
Topic: Design Cost-Optimized Architectures
A startup runs an order-tracking microservice on AWS. Data is 20GB of key-value records queried by order ID. Traffic is unpredictable with 20x spikes. The company wants to minimize operational effort and overall database cost, with point-in-time recovery for the last 7 days in a single Region. Which approaches should the solutions architect AVOID? (Select TWO.)
Options:
A. Use an Amazon Aurora Serverless v2 MySQL Multi-AZ cluster in a single Region with automatic backups retained for 7 days and capacity ranges sized for expected load.
B. Use an Amazon DynamoDB table with on-demand capacity and point-in-time recovery enabled.
C. Use an Amazon Aurora MySQL global database with a writer in the primary Region and a read-only secondary Region, both configured for Multi-AZ.
D. Use Amazon RDS for MySQL in a Single-AZ deployment with automated backups disabled and rely on weekly exports to Amazon S3 for recovery.
E. Use an Amazon DynamoDB table with provisioned capacity, auto scaling for read/write capacity, and point-in-time recovery enabled.
Correct answers: C and D
Explanation: The workload stores 20GB of key-value data accessed by order ID, with unpredictable 20x traffic spikes. The priorities are minimizing operational effort and overall database cost while maintaining point-in-time recovery for the last 7 days in a single Region.
Designs that introduce unnecessary Regions, excessive always-on capacity, or remove automated backup capabilities should be avoided. Cost-optimized designs for this scenario favor managed services such as DynamoDB or appropriately scaled RDS/Aurora configurations that stay within a single Region and provide point-in-time recovery.
The Aurora global database option adds an extra Region and redundant resources that the requirements do not justify, creating avoidable cost and complexity. The Single-AZ RDS instance with backups disabled fails the explicit recovery objective and undermines reliability. The DynamoDB and Aurora Serverless designs are acceptable because they align with the single-Region scope, support point-in-time recovery, and can adapt capacity to variable traffic with low operational overhead.
Topic: Design Cost-Optimized Architectures
A company runs a web application in AWS with separate production and non-production environments. Production has an availability target of 99.9% and must tolerate the loss of an Availability Zone. Non-production has a 95% target and can be offline outside business hours. The architect must reduce compute costs while meeting these targets and avoiding overengineering non-production. Which TWO proposed designs should the architect AVOID? (Select TWO.)
Options:
A. Use an Auto Scaling group of EC2 instances in two Availability Zones behind an Application Load Balancer for production, with a minimum of two On-Demand instances spread across AZs and additional Spot capacity. For non-production, use a single-AZ Auto Scaling group of Spot instances that scales to zero outside business hours.
B. Keep the current Multi-AZ production Auto Scaling group but purchase a Compute Savings Plan sized for the baseline production load. Move non-production to Spot instances with scheduled scaling so instances run only during business hours.
C. Run production services on AWS Fargate with tasks spread across two Availability Zones behind an Application Load Balancer and a minimum task count of two. Run non-production in a single Availability Zone with one small Fargate task, using a scheduled scaling policy to stop the service outside business hours.
D. Run production on a single Amazon EC2 instance in one Availability Zone with an Elastic IP address and scheduled snapshots; run non-production on a smaller single instance in the same Availability Zone. Use EC2 Auto Recovery for the production instance.
E. Run both production and non-production in identical Multi-AZ Auto Scaling groups behind an Application Load Balancer, each with a minimum of four On-Demand instances to simplify operations and ensure consistent performance.
Correct answers: D and E
Explanation: Production requires 99.9% availability and the ability to tolerate the loss of an Availability Zone, so it must use a Multi-AZ architecture with multiple instances or tasks and a load balancer. Non-production has a lower 95% availability target and may be offline outside business hours, which allows less redundant, cheaper designs such as single-AZ deployments and shutting resources down when not needed.
Well-Architected cost optimization recommends aligning architecture and spend with business requirements. Production justifies Multi-AZ and baseline capacity; non-production should avoid being built identically to production if its reliability target is lower. Using Spot instances, scheduled scaling, and right-sized baseline capacity helps control compute costs while still meeting availability goals.
The designs that should be avoided either under-provide availability for production (single instance in one AZ) or overengineer non-production by giving it production-level high availability and capacity, which wastes money with no business benefit.
Use the AWS SAA-C03 Practice Test page for the full IT Mastery route, mixed-topic practice, timed mock exams, explanations, and web/mobile app access.
Try AWS SAA-C03 on Web View AWS SAA-C03 Practice Test
Read the AWS SAA-C03 Cheat Sheet on Tech Exam Lexicon, then return to IT Mastery for timed practice.