Soru

Zorluk: ZorDesigning Network Architecture and Hybrid Connectivity

An energy enterprise collects telemetry from millions of smart grid sensors. They are establishing hybrid connectivity between their primary on-premises data center and Google Cloud. The workload requires a sustained throughput of 12 Gbps with a strict 99.99% availability SLA. Additionally, three isolated workload environments (Production, Analytics, and Staging) residing in distinct VPC networks must reach on-premises resources. A draft proposal suggests deploying an HA VPN gateway to a central Transit VPC and connecting the spoke VPCs via VPC Network Peering so they can route through the Transit VPC to on-premises. Which architectural modification should a Cloud Architect implement to meet all throughput, availability SLA, and multi-VPC connectivity requirements?

  1. Provision Dedicated Interconnect with redundant 10 Gbps circuits across two metropolitan locations, and replace the Transit-Spoke VPC peering model with a Shared VPC architecture where the Host VPC contains the Interconnect VLAN attachments and Service Projects house the workload subnets.Cevap
  2. B
    Upgrade the on-premises connection to Dedicated Interconnect connected to the Transit VPC, while retaining the VPC Network Peering topology between the Transit VPC and spoke VPCs with custom route exports enabled.
  3. C
    Scale the HA VPN architecture by adding five parallel HA VPN tunnels to the Transit VPC to accommodate 12 Gbps, and configure Cloud VPN gateways inside each spoke VPC to peer directly with the on-premises VPN gateways.
  4. D
    Deploy Partner Interconnect with Layer 2 connections in a single metropolitan location, and use Cloud Router with static routing to distribute traffic across the peered spoke VPCs.

Cevap

Provision Dedicated Interconnect with redundant circuits across two metropolitan locations, and implement a Shared VPC topology where the Host VPC manages the VLAN attachments and Service Projects host the workload environments.
To satisfy a continuous 12 Gbps throughput requirement and a 99.99% availability SLA, Google Cloud mandates Dedicated Interconnect deployed across two metropolitan locations with dual attachments in each metro. Furthermore, because VPC Network Peering does not support transitive routing, substituting the peered transit network with a Shared VPC structure enables workload subnets in service projects to natively send traffic over the Host VPC's Interconnect VLAN attachments.

Adım Adım Çözüm

1
Analyze hybrid connectivity throughput and SLA constraints.
The requirement specifies 12 Gbps sustained bandwidth and a 99.99% SLA. HA VPN caps at 3 Gbps per tunnel and is unsuitable for primary 12 Gbps traffic. Dedicated Interconnect with 10 Gbps links deployed across 2 metros (4 connections total across 2 edge availability domains per metro) is required to meet the 99.99% SLA.
Google Cloud requires dual-metro Dedicated Interconnect topology to guarantee a 99.99% availability SLA for hybrid connectivity.
2
Evaluate multi-VPC routing topology constraints.
VPC Network Peering is non-transitive. Spoke VPCs peered to a central Transit VPC cannot route traffic through the Transit VPC's hybrid connections (Interconnect or VPN) to reach on-premises networks.
Directing traffic from multiple spoke VPCs over a central hybrid connection via VPC Peering fails due to GCP network protocol boundaries.
3
Select the correct topology modification.
By transitioning to a Shared VPC model, the Interconnect attachments reside in the Host VPC, while subnets defined in the Host VPC are shared directly with Service Projects (Production, Analytics, Staging). Workloads use the Host VPC's routing table and Interconnect seamlessly.
Shared VPC eliminates the need for transitive peering by giving service projects native subnet access to the Host VPC's hybrid connections.

Anahtar Kavram

Designing High-Availability Dedicated Interconnect and Shared VPC Topologies
Tahmini Süre:2m 0s
Bu soruyu puanla