Your web tier runs on an EC2 Auto Scaling group behind an Application Load Balancer (ALB). You currently deploy both the ALB and the Auto Scaling group in only two Availability Zones (AZs). One AZ fails. What is the best configuration change to improve resilience?

Reduce health check timeouts so instances are replaced sooner in the failed AZ.

Health check tuning can change how quickly instances are marked unhealthy and replaced, but it does not increase usable capacity in the remaining AZ. If the remaining AZ cannot sustain load, outages can still occur.

Enable instance scale-in protection to stop the ASG from terminating unhealthy instances.

Scale-in protection primarily affects scale-in actions (for example, terminating instances due to decreasing desired capacity). It does not solve the capacity-loss problem during an AZ failure, and it can interfere with replacing instances based on unhealthy status and replacement policies.

Switch the ALB to an internal Network Load Balancer (NLB) to avoid cross-AZ traffic.

Changing from ALB to NLB does not provide additional AZ fault tolerance for the compute capacity placement. Cross-AZ traffic considerations are separate from handling an AZ outage.

What does this SAA-C03 question test?

CIDR notation defines the prefix length.

What is the correct answer to this question?

The correct answer is: Add a third Availability Zone so the ALB and Auto Scaling group span at least three AZs. — The most important resilience lever for an ALB + Auto Scaling group is to distribute both the load balancer and compute capacity across multiple AZs. With only two AZs, losing one AZ can leave you with reduced capacity (potentially not enough to serve requests). Adding a third AZ—by configuring the ALB with subnets in three AZs and configuring the ASG with subnets in three AZs—ensures that if one AZ fails, the remaining AZs can continue to serve traffic and maintain availability. A focuses on detection/replacement timing within the failed AZ, but the core issue is insufficient capacity across AZs after one fails. C can hinder proper replacement and does not create additional capacity in other AZs. D addresses traffic flow characteristics, not AZ-level fault tolerance of the deployed resources.

What should I do if I get this SAA-C03 question wrong?

Then try more questions from the same exam bank and focus on understanding why the wrong options are tempting.