A company runs an internal web app on EC2 instances behind an Application Load Balancer. The instances are in an Auto Scaling group that spans only one Availability Zone. The SysOps administrator must make the application resilient to AZ failures. Which change should be made to achieve high availability?

The correct answer is to modify the Auto Scaling group to launch instances in a second Availability Zone within the same AWS Region. This directly addresses the requirement for resilience against Availability Zone (AZ) failures. By distributing instances across multiple AZs, the application can continue to operate even if one AZ experiences an outage. Raising the maximum or minimum instance count alone does not solve the AZ failure problem if all instances are still in a single AZ. While these settings are important for scaling and performance, they don't provide AZ-level fault tolerance. Launching instances in an Availability Zone in a different AWS Region would provide disaster recovery capabilities, but it's an over-engineered solution for simply making the application resilient to AZ failures within its current region. Cross-region deployments introduce higher latency and increased complexity, which are not necessary for the stated goal of AZ resilience.