Question

Difficulty: HardDesigning Network Architecture and Hybrid Connectivity

A global media broadcasting company is designing hybrid connectivity between its primary on-premises processing center and Google Cloud. The architecture requires a continuous, private network connection with a minimum baseline bandwidth of 20 Gbps to support uncompressed video rendering workloads. Applications are distributed across two distinct Google Cloud projects: a networking project housing core connectivity and a workload project housing render nodes in a separate VPC network. The network team needs to allow workloads in the rendering VPC to communicate directly with on-premises storage arrays without exposing traffic to the public internet while maintaining administrative separation. Which network architecture design meets these requirements?

  1. A
    Provision a Dedicated Interconnect deployment with dual 10 Gbps VLAN attachments in the networking VPC, establish VPC Network Peering between the networking VPC and rendering VPC, and enable custom route export on the Cloud Router to forward on-premises traffic transitively.
  2. Provision a Shared VPC topology using the networking project as the Host Project and the rendering project as a Service Project, attaching Dedicated Interconnect VLAN attachments to the Host VPC to provide direct hybrid access to service project subnets.Answer
  3. C
    Deploy multiple HA VPN gateways with equal-cost multi-path (ECMP) routing between on-premises and the rendering VPC, while utilizing VPC Network Peering strictly for inter-VPC project communication.
  4. D
    Establish VPC Network Peering between the networking VPC and rendering VPC, and deploy Cloud VPN gateways inside the peered VPC path with dynamic BGP routing to enable transit routing from on-premises.

Answer

Provision a Shared VPC topology using the networking project as the Host Project and the rendering project as a Service Project, attaching Dedicated Interconnect VLAN attachments to the Host VPC to provide direct hybrid access to service project subnets.
The correct solution uses Shared VPC with Dedicated Interconnect. Dedicated Interconnect satisfies the 20 Gbps throughput demand. Using Shared VPC ensures that rendering project workloads reside in subnets belonging to the Host VPC, allowing them to communicate directly with on-premises networks over Cloud Router and VLAN attachments without hitting VPC Peering non-transitivity restrictions.

Step-by-Step Solution

1
Analyze bandwidth and connection type requirements
A 20 Gbps bandwidth requirement exceeds the recommended design envelope for Cloud VPN tunnels and necessitates Dedicated Interconnect (10 Gbps or 100 Gbps physical links).
Dedicated Interconnect guarantees high-throughput, enterprise-grade SLA connectivity.
2
Evaluate cross-project network architecture options for transit limitations
VPC Network Peering is non-transitive; routes learned from Cloud Router via Dedicated Interconnect in VPC A cannot be advertised or transitively used by workloads in peered VPC B.
To grant separate project workloads access to on-premises resources via the same Interconnect, Shared VPC must be used so that workloads reside on subnets of the same overarching VPC network.
3
Select the compliant solution
Designing a Shared VPC with the Host VPC managing Dedicated Interconnect attachments satisfies administrative separation while granting service project workloads direct connectivity.
Shared VPC eliminates peering transitivity constraints and meets all performance and security criteria.

Key Concept

Shared VPC vs VPC Peering for Hybrid Interconnect Connectivity
Estimated Time:2m 0s
Rate this question