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.
An enterprise is deploying three VPCs in us-central1 with custom route exchange enabled (workload-vpc peers with both transit-vpc and partner-vpc, which are not peered with each other):
transit-vpc: 4 Cloud Routers learning 180 unique local dynamic routes; DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP limit is 400.partner-vpc: 1 Cloud Router learning 140 unique local dynamic routes; DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP limit is 300.workload-vpc: 1 Cloud Router learning 30 unique local dynamic routes; limit is 300; hosts a VM with 120 alias IP ranges on nic0 (workload-vpc) and nic1 in dmz-vpc.Assuming all 350 dynamic route prefixes are unique, what happens when both peering connections are active?
workload-vpc retains all 30 local dynamic routes, and Google Cloud unpredictably drops excess peering dynamic routes imported from transit-vpc and partner-vpc.Every VPC network connected via VPC Network Peering forms its own peering group, consisting of the VPC network itself and all directly connected peer VPC networks. The DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP quota caps the aggregate number of unique IPv4 and IPv6 dynamic routes—both local to the VPC and imported from direct peers—programmed within each region of that VPC.
workload-vpc: Because no higher per-network dynamic route quota applies, the effective limit uses the workload-vpc quota (300) compared against the minimum quota across its direct peers (min(400, 300) = 300). Taking the greater of those two values yields max(300, 300) = 300.us-central1, workload-vpc's peering group aggregates 30 local routes plus 180 routes from transit-vpc and 140 routes from partner-vpc, totaling 350 unique dynamic routes—exceeding the 300 effective limit.transit-vpc uses 4 Cloud Routers in us-central1 (within the hard limit of 5 per VPC network and region). The VM in workload-vpc uses 1 NIC per VPC network (nic0 in workload-vpc, nic1 in dmz-vpc) and 120 alias IP ranges on nic0 (within the 150 per-NIC hard limit).transit-vpc (180 + 30 = 210) and partner-vpc (140 + 30 = 170) remain well below their respective limits.workload-vpc itself (all 30 local routes remain intact).transit-vpc or partner-vpc are omitted.To avoid unpredictable route loss as multi-VPC topologies scale, architects should request a quota increase for DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP across workload-vpc and partner-vpc, summarize BGP prefixes, or migrate to Network Connectivity Center (NCC) VPC spokes, which eliminate peering group quota aggregation.
workload-vpc, dropping 50 lowest-ranked routes including local dynamic routes.Cloud Router uses a deterministic route order when a VPC network exceeds its unique Cloud Router dynamic route prefixes quotas—specifically the from-own-region quota or, in global dynamic routing mode, the from-other-regions quota. When those Cloud Router quotas are exhausted, Google Cloud deterministically selects up to the quota limit and drops the remaining lower-ranked learned prefixes consistently across regions.
This scenario does not breach Cloud Router's per-region local prefix quotas (workload-vpc only learns 30 local dynamic routes in us-central1). Instead, the bottleneck occurs at the VPC Network Peering boundary (DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP). When a peering group's dynamic route limit is exceeded, Google Cloud explicitly protects local dynamic routes and does not use Cloud Router's deterministic route order to evaluate local and peering routes in a shared pool.
DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP is exceeded; all 30 local routes in workload-vpc stay programmed.from-own-region and from-other-regions) govern routes learned by Cloud Routers inside a single VPC, whereas peering group quotas govern the combined routing table capacity across directly peered VPCs.Deterministic route dropping would apply if a single VPC network's Cloud Routers learned more unique prefixes than allowed by its from-own-region quota (for example, exceeding 250 unique prefixes in one region) or from-other-regions quota in global dynamic routing mode, without involving peering group route limits.
workload-vpc with a QUOTA_EXCEEDED error because aggregating 350 dynamic routes exceeds its 300-route effective peering group limit.QUOTA_EXCEEDED enforcement isIn Google Cloud, Cloud Quotas typically enforce hard boundaries at resource creation time. When an API call or CLI command attempts to allocate a resource—such as a new subnet IP range, static route, or Cloud Router—beyond the configured project or per-network quota, Google Cloud immediately rejects the operation and returns a QUOTA_EXCEEDED error.
While workload-vpc does indeed exceed its 300-route effective limit for DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP (aggregating 30 + 180 + 140 = 350 unique dynamic routes), VPC Network Peering does not block peering creation when dynamic route limits are exceeded. Unlike static quota checks on created objects, dynamic routes are learned asynchronously over Border Gateway Protocol (BGP) and imported dynamically across peering boundaries.
ACTIVE, workloads in workload-vpc can experience intermittent or partial reachability to remote destinations whose routes were dropped.DYNAMIC_ROUTES_PER_REGION_PER_PEERING_GROUP has been exceeded.This behavior would occur if the limit being exceeded were a static creation-time quota—such as Peerings per VPC network (PEERINGS_PER_NETWORK), Subnetwork ranges per peering group, or Static routes per peering group—where Google Cloud synchronously validates the total resource count and rejects the peering request or new resource creation with a QUOTA_EXCEEDED error.
workload-vpc inherits an effective limit of 400 from transit-vpc so all 350 routes are programmed, but deployment fails because workload-vpc's VM exceeds per-interface alias IP limits and transit-vpc exceeds regional Cloud Router limits.Per-peering group quotas calculate an effective limit for each VPC network based on its own quota and the quotas of all its directly connected peers. Meanwhile, hard system limits on compute instances and Cloud Routers enforce fixed boundaries—such as the maximum number of Cloud Routers per VPC network and region, network interfaces per VPC network, and alias IP ranges per network interface—that cannot be bypassed.
First, workload-vpc does not inherit an effective limit of 400 from transit-vpc. The effective limit algorithm takes the minimum quota value across all direct peers before comparing it with the local VPC's quota. Because workload-vpc is peered with both transit-vpc (400) and partner-vpc (300), the minimum across its direct peers is min(400, 300) = 300, keeping workload-vpc's effective limit at max(300, 300) = 300.
Second, neither the VM in workload-vpc nor the Cloud Routers in transit-vpc violate any hard system limits.
300).transit-vpc's 4 Cloud Routers in us-central1 are within limits.nic0 is valid.nic0 in workload-vpc and nic1 in dmz-vpc satisfies this requirement.The effective limit would rise to 400 only if both transit-vpc and partner-vpc had quotas of at least 400. Likewise, deployment would fail on hard limits only if transit-vpc attempted to create a 6th Cloud Router in us-central1, the VM assigned more than 150 alias IP ranges to nic0, or both nic0 and nic1 attached to workload-vpc.
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