SAA-C03 Exam Dumps : AWS Certified Solutions Architect - Associate (SAA-C03)

Compute Savings Plansoffer the most flexible and cost-effective solution for the production instances, as they provide significant savings (up to 66%) for both EC2 and AWS Lambda usage, while allowing flexibility in the type of instance family, size, and even region. For nonproduction instances, usingOn-Demand Instancesensures you only pay for the instances when they are running, and shutting them down during off-hours further optimizes cost.

Key AWS features:

Compute Savings Plans: Provide savings based on consistent usage, making it ideal for production environments with steady load.

On-Demand Instances: Suitable for nonproduction environments that are used intermittently. Shutting them down when not in use avoids unnecessary costs.

AWS Documentation: According to AWS ' s cost optimization best practices, using a combination of Savings Plans for production and On-Demand Instances for nonproduction environments that are used sparingly results in optimal cost savings​​.

The Application Load Balancer (ALB) supports sticky sessions (session affinity) using application cookies. AWS WAF integrates natively with ALB to provide Layer 7 protection at the same endpoint.

From AWS Documentation:

“You can enable sticky sessions for your Application Load Balancer target groups to ensure that a user’s requests are consistently routed to the same target. AWS WAF integrates with Application Load Balancer to protect your web applications from common exploits.”

(Source: Elastic Load Balancing User Guide & AWS WAF Developer Guide)

Why C and E are correct:

C: ALB operates at Layer 7 (HTTP/HTTPS), supports sticky sessions, and can serve as a public endpoint.

E: AWS WAF can be directly associated with the ALB to inspect traffic and enforce rules.Together, they fulfill both the security and session affinity requirements.

Why others are incorrect:

A: Network Load Balancer doesn’t support session affinity.

B: Gateway Load Balancer is used for virtual appliances, not web applications.

D: Using EIPs bypasses load balancing and WAF integration.

The correct answer isAbecause the application is aglobal Voice over IPworkload that needslow latency,high availability, andautomated failover across AWS Regions.AWS Global Acceleratoris specifically designed to improve the availability and performance of global applications by routing user traffic over the AWS global network to the optimal healthy endpoint. It usesstatic Anycast IP addresses, which means users connect through fixed IP addresses that do not depend on client-side DNS or IP address caching behavior.

This is especially valuable for latency-sensitive applications such as Voice over IP, where rapid routing decisions and fast failover are critical. Global Accelerator continuously monitors endpoint health and can automatically direct traffic to healthy endpoints in another AWS Region if a failure occurs. This supports the requirement for cross-Region failover with minimal disruption.

Option B is incorrect becauseRoute 53 geolocation routingdirects traffic based on user location, but DNS-based routing can be affected by client-side caching and does not provide the same performance acceleration or fast failover behavior as Global Accelerator. Option C is incorrect becauseAmazon CloudFrontis primarily a content delivery network for HTTP and similar content distribution scenarios, not the preferred service for real-time Voice over IP traffic. Option D is incorrect because anApplication Load Balancerdoes not by itself provide global routing or multi-Region failover.

AWS best practices for global, latency-sensitive applications recommendAWS Global Acceleratorwhen the goal is to improve performance, use the AWS backbone network, and provide health-based failover without depending on DNS cache expiration. Therefore,AWS Global Accelerator with health checksis the best solution.