GCP Professional Cloud Architect Practice Question

Your company operates two on-premises data centers in Chicago and Dallas that host latency-sensitive ERP systems. You are designing a hybrid network to migrate the application layer to a new VPC in us-central1 while the database tier remains on-premises for the next 12 months. The application will exchange a steady 12 Gbps (peaking at 18 Gbps) of traffic with both data centers and must maintain sub-5 ms round-trip latency to Chicago. The network team does not want to lease new rack space or purchase additional routing hardware, but they can order managed circuits from their existing telco, which is a Google-approved connectivity provider in both metros. The CFO requests the most cost-efficient solution that still delivers private connectivity and 99.9% availability across the two locations. Which connectivity design best meets the requirements?

Deploy two 100 Gbps Dedicated Interconnect circuits in a single Chicago colocation facility and advertise all on-premises prefixes over BGP to maximize bandwidth and minimize latency.
Establish four HA VPN tunnels (two per data center) over diverse internet links and use Cloud Router dynamic routing to load-balance traffic across tunnels.
Order Partner Interconnect from the telco with two independent 10 Gbps VLAN attachments in each of the Chicago and Dallas partner edge locations, connect them to separate Cloud Routers, and configure active/active BGP for automatic failover.
Provision one 50 Mbps Partner Interconnect VLAN attachment in each data center and rely on Cloud NAT for additional capacity during traffic spikes.

Report Issue

Answer Description

Partner Interconnect lets you obtain private connectivity to Google's network through a supported service provider without placing your own equipment in a colocation facility, avoiding the capital expense of Dedicated Interconnect. Two 10 Gbps VLAN attachments in separate Chicago and Dallas partner locations provide the required 18 Gbps peak throughput (each attachment supports up to 10 Gbps) and, when combined with separate physical circuits and Cloud Router BGP sessions, satisfy Google's High Availability (HA) configuration, which offers a 99.9% availability SLA per location. Using a single metro or lower-bandwidth attachments would violate the redundancy or bandwidth requirements. Dedicated Interconnect would exceed capacity but impose colocation costs the team wants to avoid. Cloud VPN cannot guarantee sub-5 ms latency because traffic traverses the public internet and each tunnel's throughput ceiling is lower, so it does not meet the performance objective. Thus, the dual-location Partner Interconnect with 10 Gbps attachments is the optimal, cost-effective choice.

Ask Bash

Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.