how to configure the redundant NAT gateway ?

Since the end of 2025, you have an additional option with the Amazon VPC Regional NAT Gateway, described and introduced in the following link by AWS:

• https://aws.amazon.com/blogs/networking-and-content-delivery/introducing-amazon-vpc-regional-nat-gateway/

For redundant NAT gateways in AWS, you don't need to configure explicit failover mechanisms as you would with NAT instances. The redundancy is achieved through proper subnet and route table configuration rather than failover scripts.

Here's how to configure redundant NAT gateways:

1. Create NAT gateways in different Availability Zones (AZs) - as shown in your diagram with NAT Gateway A and NAT Gateway B in separate AZs.

2. Configure your route tables for each private subnet to point to the NAT gateway in the same AZ:

• Private subnet in AZ-A: Set the route for 0.0.0.0/0 to point to NAT Gateway A
• Private subnet in AZ-B: Set the route for 0.0.0.0/0 to point to NAT Gateway B

This configuration doesn't provide automatic failover between NAT gateways. Instead, it provides AZ-level redundancy. If an entire AZ fails (including its NAT gateway), only resources in that AZ are affected. Resources in the other AZ continue to function normally using their own NAT gateway.

For true automatic failover between NAT gateways, you would need to implement a more complex solution involving monitoring and route table updates. However, the standard best practice is to design your application to be resilient to AZ failures by distributing workloads across multiple AZs, each with its own NAT gateway, as shown in your diagram.
Sources
Nat gateway and internet gateway | AWS re:Post
Redundant AWS Site-to-Site VPN connections for failover - AWS Site-to-Site VPN