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 hosts an internal financial application in europe-west1 and us-central1 behind regional internal Application Load Balancers in a single VPC network. On-premises clients connect over Cloud VPN tunnels in both regions and resolve app.corp.internal through a Cloud DNS private zone using an inbound server policy.
You must meet the following requirements:
europe-west1 fail health checks, European traffic never fails over to us-central1.v1 and 5% to a canary v2 while preserving Cloud DNS health checks.Which configuration should you implement?
weightedBackendServices in each regional URL map to split traffic 95% to v1 and 5% to v2. In the Cloud DNS private zone, create an A record set for app.corp.internal with a geolocation (GEO) routing policy and geofencing enabled, referencing each load balancer's forwarding rule name with health checking enabled.Combining a Cloud DNS geolocation (GEO) routing policy with geofencing and Application Load Balancer URL map traffic splitting separates geographic DNS steering from fine-grained Layer 7 canary routing. Cloud DNS directs clients to the nearest regional virtual IP (VIP) while keeping traffic strictly within regional boundaries, and each regional internal Application Load Balancer uses weightedBackendServices in its URL map to distribute requests across v1 and v2 backends.
europe-west1 VPN tunnel resolve to the europe-west1 load balancer, while US queries resolve to us-central1.--enable-geo-fencing) overrides the default GEO failover behavior. Even if all backends in europe-west1 fail health checks, Cloud DNS does not fail over to us-central1; instead, it continues returning the europe-west1 VIP unaltered.weightedBackendServices (95 for v1 and 5 for v2) inside each regional URL map splits HTTP(S) traffic at the proxy layer without violating Cloud DNS's restriction against combining GEO and weighted round-robin (WRR) policies. Enabling global access on the regional internal load balancers and referencing their forwarding rule names (rather than raw IP addresses) ensures Cloud DNS health checking monitors the load balancers properly.WRR policies.v1 and v2 share the same regional frontend VIP, TLS configuration, and routing rules.Cloud DNS does not allow combining GEO and WRR routing policies on a single resource record set. Offloading the 95%/5% split to weightedBackendServices in the regional URL maps while keeping GEO with geofencing and forwarding-rule health checks in Cloud DNS satisfies every routing, compliance, and canary requirement cleanly.
v1 and v2 regional internal Application Load Balancers with global access enabled in each region. In the Cloud DNS private zone, configure a FAILOVER routing policy for app.corp.internal using the v1 forwarding rules as primary targets, a GEO backup policy with geofencing enabled for v2, and a 0.05 backup trickle ratio.A Cloud DNS FAILOVER routing policy implements an active-passive high-availability configuration for internal VPC resources. During normal operation, Cloud DNS returns the IP addresses from the active (primary) set; only when all primary targets fail health checks does Cloud DNS shift traffic to the backup set, which uses a GEO policy. An optional trickle traffic ratio (between 0 and 1, such as 0.05 or 0.1) continuously sends a small fraction of DNS queries to the backup targets to keep them primed.
v1) traffic: In a FAILOVER policy, the primary target set (--routing-policy-primary-data) does not perform geolocation matching. Cloud DNS returns all healthy primary VIPs regardless of whether a client query originated in europe-west1 or us-central1, sending European clients to US v1 load balancers and vice versa.v1 targets are not geo-routed or geo-fenced, European traffic can be directed to us-central1 during normal operation, immediately breaching data residency rules.0.05 trickle ratio sends 5% of all global DNS queries to the backup GEO set (v2), rather than reliably splitting 5% of regional HTTP(S) requests at the load balancer.FAILOVER policies are intended for disaster recovery where a primary region serves 100% of steady-state traffic (minus any trickle ratio) and a secondary GEO set acts as a cold or warm standby.backup set when every forwarding rule in the active set is unhealthy.backup set in a FAILOVER policy supports GEO mapping.The organization required an active-passive disaster recovery topology—for example, serving 100% of steady-state traffic from a single primary region while sending a 5% trickle to backup GEO regions to verify readiness—rather than active-active regional proximity routing with strict data residency.
v1 and v2 regional internal Application Load Balancers with global access enabled in each region. In the Cloud DNS private zone, configure a single A record set for app.corp.internal that combines a geolocation (GEO) routing policy with geofencing and a weighted round-robin (WRR) routing policy assigning weights of 95 to v1 and 5 to v2 per region using their forwarding rule names.This approach attempts to perform both geographic proximity routing and weighted canary traffic splitting directly inside a single Cloud DNS resource record set by nesting or combining GEO and WRR routing policies across separate v1 and v2 load balancer forwarding rules.
GEO) routing policy with a weighted round-robin (WRR) routing policy on the same resource record set. Only one routing policy type can be applied to a record set at a time (with the sole exception of a FAILOVER policy that uses a GEO policy for its backup set).GEO and WRR policy structures is rejected by the Cloud DNS API and gcloud CLI.WRR policies are supported, DNS resolvers cache responses for the duration of the record's Time to Live (TTL), which can cause uneven traffic distribution across clients compared to Layer 7 load balancer traffic splitting.routingPolicy field supports wrr, geo, or primaryBackup (FAILOVER), but never geo and wrr together.WRR accepts weights from 0 to 1000 (inclusive) and distributes DNS responses proportionally across healthy weighted targets.GEO policy item can list multiple IP addresses or forwarding rules for a region, but Cloud DNS returns all healthy targets configured for that region rather than applying custom weights among them.Wait—this configuration is never valid in Cloud DNS as written. However, a standalone WRR routing policy (without GEO or regional data residency requirements) would be appropriate if all backends served a single region or global pool and you only needed simple DNS-level canary traffic distribution between v1 and v2 endpoints.
When an architecture requires both geographic steering (with or without geofencing) and weighted canary rollouts, the standard pattern is to handle regional steering at the DNS layer with GEO and perform weighted traffic splitting at the Application Load Balancer layer using weightedBackendServices in the URL map.
weightedBackendServices in each regional URL map to split traffic 95% to v1 and 5% to v2. In the Cloud DNS private zone, create an A record set for app.corp.internal with a GEO routing policy (geofencing disabled) that specifies each forwarding rule's internal IP address with health checking enabled.This option correctly delegates the 95%/5% canary split to weightedBackendServices in each regional internal Application Load Balancer's URL map and enables global access, but it disables geofencing and references raw IP addresses instead of forwarding rule names in the Cloud DNS GEO record set.
GEO policies), if all endpoints in europe-west1 fail health checks, Cloud DNS automatically fails over European DNS queries to the next closest healthy region—us-central1—violating the strict data residency requirement.gcloud or the API for internal load balancers, you must specify the forwarding rule name (or structured internal load balancer target) rather than a bare IPv4 address. Specifying only the raw IP address prevents Cloud DNS from associating internal health check signals with the record.GEO failover behavior: Without --enable-geo-fencing, Cloud DNS prioritizes availability over regional isolation by failing over across geographies whenever a region's health-checked targets become unhealthy.Data residency was not required (making cross-region failover desirable during outages) and the Cloud DNS record referenced the forwarding rule names rather than raw IP addresses so health checks would actually function.
A common misconception is that deploying regional inbound server policy entry points or regional load balancers alone prevents cross-region DNS failover. In reality, Cloud DNS GEO policies always fail over to the next nearest region during an outage unless geofencing is explicitly enabled.
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