Prepare and test your skills

Prepare and test your skills

Free sample question with answers and detailed explanation below. The Professional Cloud Network Engineer practice bank has an additional 2,437 practice questions.Try 10 free realistic questions: instant score, no credit card, and no account required.
Your organization is deploying a multi-region workload across us-central1 and us-east4 in a single VPC network (prod-vpc). All Compute Engine VMs use private RFC 1918 IP addresses without external IPs. Your architecture must meet these requirements:
ORD and IAD), with automatic cross-region failover.Which network configuration should you implement?
prod-vpc in custom subnet mode with regional dynamic routing. Provision four Dedicated Interconnect connections across both edge availability domains in ORD and IAD, attach four VLAN attachments to two Cloud Routers (us-central1 and us-east4), and deploy an auto-scaling Cloud NAT gateway on each Cloud Router.Regional dynamic routing restricts each Cloud Router to learning and advertising dynamic routes strictly within its own Google Cloud region. In this mode, a Cloud Router only advertises VPC subnets that co-reside in the same region and only programs BGP-learned on-premises routes for VM instances located in that local region.
ORD metro fail, VMs in us-central1 cannot reach on-premises networks via the healthy us-east4 Cloud Router because regional routing prevents learned routes in us-east4 from being programmed into us-central1.The organization only required a 99.9% single-region SLA with strict regional isolation where workloads in us-central1 and us-east4 never needed to fail over across regions, or if every region had four Dedicated Interconnect connections in a single-region 99.99% metro pair.
Many engineers assume regional dynamic routing is required to keep traffic local during normal operations. In reality, global dynamic routing already preserves regional affinity automatically by adding a cross-region penalty metric, giving you both local preference and seamless cross-region failover.
prod-vpc in custom subnet mode with non-overlapping subnets from a planned supernet and global dynamic routing. Provision four Dedicated Interconnect connections across both edge availability domains in ORD and IAD with four VLAN attachments across two Cloud Routers (us-central1 and us-east4), and deploy an auto-scaling Cloud NAT gateway with dynamic port allocation on each region's Cloud Router.Global dynamic routing is a VPC-level routing mode that allows Cloud Routers in any region to advertise all VPC subnets and program BGP-learned dynamic routes across all regions of the VPC. Cloud NAT is a Google-managed, distributed, regional network address translation service that provides private VMs without external IPs secure, scalable outbound internet connectivity.
ORD and IAD), spanning both edge availability domains (zone1 and zone2) in each metro, with four VLAN attachments distributed across us-central1 and us-east4 satisfies all physical and logical SLA prerequisites. Global dynamic routing ensures that if all links in one region fail, traffic automatically reroutes via the surviving region.200 + RTT in milliseconds) to remote subnet advertisements and learned routes. Consequently, steady-state traffic always prefers the local region's interconnect, maintaining low latency and path symmetry.Unlike regional dynamic routing (which breaks cross-region failover), a single-region Cloud NAT gateway (which leaves the second region without egress), or under-provisioned static NAT IPs and dual interconnects (which fail the 99.99% SLA and risk port exhaustion), this design aligns every layer with Google Cloud's global control plane and regional data plane architecture.
prod-vpc in custom subnet mode with global dynamic routing. Provision four Dedicated Interconnect connections across both edge availability domains in ORD and IAD with four VLAN attachments across two Cloud Routers (us-central1 and us-east4), and deploy a single auto-scaling Cloud NAT gateway with multiple IP addresses on the us-central1 Cloud Router to handle egress for both regions.A Cloud NAT gateway is a regional managed resource bound to a Cloud Router in a specific Google Cloud region. It performs source network address translation (SNAT) exclusively for VM instances and subnets residing within that same region so they can initiate outbound internet connections without external IP addresses.
us-east4: Because Cloud NAT does not operate across regions, a gateway attached to the us-central1 Cloud Router cannot translate or forward outbound internet traffic for VMs located in us-east4.prod-vpc enables cross-region BGP hybrid routing, not cross-region Cloud NAT sharing.0.0.0.0/0 next-hop routes across regions.Private VMs requiring outbound internet access were deployed exclusively in us-central1, while us-east4 only served as an isolated hybrid transit region or hosted workloads that did not require internet egress.
Because VPC networks and global dynamic routing operate globally, architects sometimes mistakenly assume Cloud NAT is also global. Unlike VPC firewall policies or global load balancing, Cloud NAT is strictly regional and must be deployed in every region where private VMs need internet access.
prod-vpc in custom subnet mode with global dynamic routing. Provision two Dedicated Interconnect connections (one in ORD and one in IAD) connected to two Cloud Routers (us-central1 and us-east4), and deploy a Cloud NAT gateway in each region configured with a single manual static external IP address.Deploying one Dedicated Interconnect connection per metro (ORD and IAD) provides geographic separation across two locations, while a Cloud NAT gateway with a single manually assigned static IP address translates outbound traffic from private VMs through one fixed external IP per region.
ORD and IAD can break hybrid connectivity completely; a 99.99% SLA requires at least four connections (two per metro across both zone1 and zone2 edge availability domains).1024–65535) per destination IP and port tuple. Under peak concurrency—especially when many VMs connect to the same third-party API endpoint—static single-IP allocation exhausts available ports and drops new connections.zone1 and zone2 within the same metro, not between separate metros.ORD immediately forces all us-central1 traffic across the backbone to us-east4, degrading latency instead of failing over locally within ORD.The organization did not require a 99.99% SLA (accepting reduced resilience against simultaneous metro maintenance) and experienced low, predictable outbound connection concurrency that would never approach single-IP port limits.
A common misconception is that deploying two links in two different metros equals four-nines availability. For 99.99% SLA compliance, local edge availability domain diversity (zone1 and zone2) inside each metro is mandatory alongside metro diversity, and auto-scaling NAT IPs with dynamic port allocation is essential for high-throughput egress.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated