Question

Difficulty: HardNetworking and Connectivity Design

An enterprise is designing a new multi-VPC AWS architecture in the `us-east-1` Region. The architecture consists of a production VPC (`vpc-prod-us-1` with CIDR 10.100.0.0/1610.100.0.0/16) and a shared services VPC (`vpc-shared-us-1` with CIDR 10.200.0.0/1610.200.0.0/16). Both VPCs are deployed across two Availability Zones (`us-east-1a` and `us-east-1b`). The enterprise needs to connect this AWS environment to an on-premises datacenter using an AWS Direct Connect (DX) connection.

The network design must satisfy the following requirements:
- Enable secure, private bidirectional communication between the on-premises datacenter and both VPCs.
- Enable the production VPC to resolve hostnames in the on-premises DNS zone (`corp.internal`), and enable on-premises hosts to resolve hostnames in a Route 53 Private Hosted Zone (PHZ) for `shared.aws` associated with the shared services VPC.
- Enable high availability for outbound internet traffic from private subnets in `vpc-prod-us-1` such that an outage in a single Availability Zone does not impact internet egress from the remaining Availability Zone.
- Maintain routing isolation between `vpc-prod-us-1` and `vpc-shared-us-1` for VPC-to-VPC traffic, allowing traffic only to and from the datacenter.

Which of the following designs meets these requirements with the least operational complexity?

  1. A
    Deploy an AWS Direct Connect Gateway (DXGW) and associate it directly with Virtual Private Gateways (VGWs) attached to `vpc-prod-us-1` and `vpc-shared-us-1` using Private Virtual Interfaces (Private VIFs). To achieve VPC-to-VPC routing isolation on AWS, configure the on-premises router to hair-pin and transitively route traffic between the two VPCs. Associate the Route 53 Private Hosted Zone (PHZ) for `shared.aws` with both VPCs. In `vpc-shared-us-1`, deploy Route 53 Resolver Inbound and Outbound Endpoints across both Availability Zones, and configure a forwarding rule for `corp.internal` associated with both VPCs. In `vpc-prod-us-1`, deploy a NAT Gateway in a public subnet in `us-east-1a` and another in `us-east-1b` for redundant egress.
  2. B
    Deploy an AWS Transit Gateway (`tgw-us-east-1`) and associate it with an AWS Direct Connect Gateway (DXGW) using a Transit Virtual Interface (Transit VIF). Attach both VPCs to the Transit Gateway and configure separate Transit Gateway route tables to isolate VPC-to-VPC traffic. Associate the Route 53 PHZ for `shared.aws` only with `vpc-shared-us-1`. In `vpc-shared-us-1`, deploy Route 53 Resolver Inbound and Outbound Endpoints across both Availability Zones, and configure a Route 53 Resolver forwarding rule for `corp.internal` associated with both VPCs. In `vpc-prod-us-1`, deploy a single NAT Gateway in a public subnet in `us-east-1a`, and configure the private subnet route tables in both `us-east-1a` and `us-east-1b` to route all internet-bound traffic to this single NAT Gateway.
  3. Deploy an AWS Transit Gateway (`tgw-us-east-1`) and associate it with an AWS Direct Connect Gateway (DXGW) using a Transit Virtual Interface (Transit VIF). Attach `vpc-prod-us-1` and `vpc-shared-us-1` to the Transit Gateway. Configure separate Transit Gateway route tables for the VPC attachments and the DXGW association, ensuring VPC-to-VPC route propagation is disabled to maintain isolation. Associate the Route 53 Private Hosted Zone (PHZ) for `shared.aws` with both `vpc-prod-us-1` and `vpc-shared-us-1`. In `vpc-shared-us-1`, deploy Route 53 Resolver Inbound and Outbound Endpoints across both Availability Zones. Configure a Route 53 Resolver forwarding rule for `corp.internal` pointing to on-premises DNS servers, and associate this rule with both VPCs. In `vpc-prod-us-1`, deploy a NAT Gateway in a public subnet in `us-east-1a` and another in `us-east-1b`, and configure the private subnet route tables to route internet-bound traffic to the local NAT Gateway in their respective Availability Zone.Answer
  4. D
    Deploy an AWS Transit Gateway (`tgw-us-east-1`) and associate it with an AWS Direct Connect Gateway (DXGW) using a Transit Virtual Interface (Transit VIF). Attach both VPCs to the Transit Gateway and configure separate Transit Gateway route tables to isolate VPC-to-VPC traffic. Associate the Route 53 PHZ for `shared.aws` only with `vpc-shared-us-1`. In `vpc-shared-us-1`, deploy Route 53 Resolver Inbound and Outbound Endpoints across both Availability Zones. Configure a Route 53 Resolver forwarding rule for `corp.internal` pointing to on-premises DNS servers, and associate this rule with both VPCs. In `vpc-prod-us-1`, deploy a NAT Gateway in a public subnet in `us-east-1a` and another in `us-east-1b`, and configure the private subnet route tables to route internet-bound traffic to the local NAT Gateway in their respective Availability Zone.

Answer

Deploy an AWS Transit Gateway associated with a Direct Connect Gateway and configure separate Transit Gateway route tables to prevent VPC-to-VPC routing. Associate the Route 53 Private Hosted Zone for `shared.aws` with both VPCs, deploy Route 53 Resolver Endpoints in the shared services VPC, and associate the forwarding rule for `corp.internal` with both VPCs. Deploy a NAT Gateway in each Availability Zone of the production VPC to ensure high availability for outbound internet traffic.
The correct design uses AWS Transit Gateway to centralize Direct Connect traffic and implements separate Transit Gateway route tables to enforce routing isolation between the production and shared services VPCs. To ensure resilient internet egress in the production VPC, a NAT Gateway is deployed in each Availability Zone. For DNS resolution, Route 53 Resolver Outbound Endpoints are shared by associating the forwarding rule for the on-premises domain with both VPCs, and the Private Hosted Zone for `shared.aws` is explicitly associated with both VPCs to allow direct cross-VPC resolution.

Step-by-Step Solution

1
Configure hybrid connectivity with routing isolation.
Deploy AWS Transit Gateway, associate it with a Direct Connect Gateway using a Transit VIF, and attach both VPCs. Create separate Transit Gateway route tables to disable VPC-to-VPC route propagation while allowing traffic to/from the on-premises datacenter.
This establishes Direct Connect access to both VPCs while enforcing routing isolation on AWS without relying on unsupported on-premises hair-pinning.
2
Design multi-VPC DNS resolution.
Deploy Route 53 Resolver Inbound and Outbound Endpoints in the shared services VPC. Configure a forwarding rule for `corp.internal` pointing to the on-premises DNS servers and associate it with both VPCs. Associate the Private Hosted Zone for `shared.aws` with both the production and shared services VPCs.
This allows both VPCs to resolve the on-premises domain and allows the production VPC to resolve the private hosted zone directly, which is a requirement for multi-VPC DNS access.
3
Implement highly available internet egress.
Deploy two NAT Gateways in the production VPC (one in each Availability Zone) and update the private subnet route tables to route outbound internet traffic through their local NAT Gateway.
This design avoids a single point of failure and ensures that an outage in one Availability Zone does not impact internet egress for the remaining zone.

Key Concept

Designing a resilient hybrid network architecture with routing isolation, multi-VPC private DNS resolution, and zone-redundant internet egress.
Estimated Time:3m 0s
Rate this question