A company wants a more reliable architecture for its application. The application currently uses one Amazon RDS DB instance and two manually provisioned Amazon EC2 web servers in a single Availability Zone. An employee recently deleted the DB instance and the application was unavailable for 24 hours. What should a solutions architect do to maximize the reliability of the application's infrastructure?

The correct option maximizes reliability by addressing single points of failure and human error. Making the RDS DB instance Multi-AZ ensures high availability and automatic failover to a standby replica in another AZ if the primary fails. Enabling deletion protection prevents accidental deletion of the database. Placing EC2 instances behind an Application Load Balancer (ALB) distributes traffic and improves fault tolerance. Running EC2 instances in an Auto Scaling group across multiple AZs ensures instances are automatically replaced if they fail and provides elasticity. The other options are incorrect because: Deleting an EC2 instance reduces availability, and termination protection doesn't prevent application-level failures. Adding an API Gateway and Lambda for DB writes increases complexity and doesn't directly address the reliability of the existing EC2 or RDS setup, nor does it prevent accidental deletion of the original DB. Using Spot Instances is cost-effective but less reliable for critical, continuously running workloads as they can be interrupted.