An operational workload runs on Amazon EC2 instances in an Auto Scaling group. The VPC spans two Availability Zones (AZs), with one subnet in each AZ targeted by the Auto Scaling group. The VPC is connected to on-premises infrastructure and that connectivity must not be disrupted. The Auto Scaling group maximum is 20 instances. The current IPv4 addressing is: VPC 10.0.0.0/23, AZ1 subnet 10.0.0.0/24, AZ2 subnet 10.0.1.0/24. A third AZ is now available in the Region. You must add the new AZ for the Auto Scaling group without adding IPv4 address space and without service downtime. Which approach satisfies these constraints?

The correct approach ensures no downtime and no new IPv4 address space. By shifting traffic to AZ2, AZ1's subnet can be deleted and recreated with a smaller CIDR (10.0.0.0/25). The Auto Scaling group then targets this new AZ1 subnet. Once instances are healthy, traffic shifts to AZ1, allowing AZ2's subnet to be deleted. The remaining address space from the original AZ1 (10.0.0.128/25) is used for a new AZ2 subnet. Finally, the original AZ2's address space (10.0.1.0/24) is split, with 10.0.1.0/25 used for a new AZ3 subnet. This phased approach maintains connectivity and instance availability throughout the process. Incorrect options: Terminating instances in an AZ and recreating subnets simultaneously would cause downtime. Creating a new VPC would disrupt connectivity to on-premises infrastructure, violating a constraint. Updating a subnet to reduce its address space is not possible directly; subnets must be deleted and recreated to change their CIDR.