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.
A regional internal Application Load Balancer in us-central1 fronts three backend VMs (web-1, web-2, and web-3) in 10.10.1.0/24, using a proxy-only subnet (10.10.2.0/24). Following a security update, VPC ingress firewall rules allow TCP port 8080 from client subnet 10.20.0.0/16 and health checks from 35.191.0.0/16 and 130.211.0.0/22 for targets with the app-backend tag, after which clients in 10.20.0.0/16 (us-central1) lose connectivity.
You run a Connectivity Test from a client VM in 10.20.0.0/16 to the load balancer forwarding rule on TCP 8080, revealing:
web-1 and web-2 ends in Drop at the ingress firewall evaluation.web-1 and web-2 but are dropped by the implied deny ingress rule for web-3.Which remediation steps should you take based on this Connectivity Tests analysis?
A regional internal Application Load Balancer operates as an Envoy-based HTTP(S) reverse proxy. Client connections terminate at managed proxies allocated in the region's proxy-only subnet (10.10.2.0/24), and those proxies open new connections to the backend VMs using source IP addresses from 10.10.2.0/24. Meanwhile, network tags bind individual VMs to target-filtered VPC firewall rules.
Drop state on web-1 and web-2: Because backend VMs receive traffic from the Envoy proxies rather than directly from 10.20.0.0/16, updating the VPC ingress firewall rule on TCP port 8080 to allow source range 10.10.2.0/24 permits proxied client requests to reach the backends.web-3: Applying the app-backend network tag to web-3 eliminates the inconsistent backend configuration, allowing both Google Cloud health check probes (35.191.0.0/16 and 130.211.0.0/22) and proxy traffic to match the allow rules instead of hitting the implied deny ingress rule.8080 solely to the load balancer's proxy-only subnet rather than exposing backends directly to client IP ranges.web-3 to a healthy state across all backend instances.X-Forwarded-For) while keeping VPC firewall boundaries tight.This remediation directly addresses both root causes identified in the Connectivity Tests trace—source IP translation by the Layer 7 proxies and a missing network tag on web-3—without introducing invalid static routes or misapplying Layer 4 passthrough load balancer settings.
VPC firewall rule priorities (0 to 65535, where lower numbers evaluate first) control precedence when multiple rules match a packet. Reconfiguring a load balancer health check changes the protocol and port probed on backend instances (such as switching from application port 8080 to port 80).
10.20.0.0/16 allow rule to 100 has no effect because the implied deny ingress rule already runs at priority 65535. Proxied packets arriving at web-1 and web-2 originate from 10.10.2.0/24, so they never match a rule scoped to 10.20.0.0/16 at any priority.web-3's missing tag: Changing the health check to port 80 breaks probing if the application listens on 8080 and leaves web-3 blocked by the implied deny ingress rule because it still lacks the app-backend target tag.deny ingress rule sits at priority 65535, so any matching allow rule with priority 0–65534 already supersedes it unless shadowed by a higher-priority explicit deny rule.10.10.2.0/24), not the client subnet.A higher-priority explicit deny rule (for example, at priority 500) were shadowing an allow rule for 10.10.2.0/24 at priority 1000, and web-3 were already tagged app-backend while serving traffic on port 80.
Global access, local routes installed by the Google guest agent, and legacy health check IP ranges (209.85.152.0/22 and 209.85.204.0/22) apply to passthrough Network Load Balancers (Layer 4) or cross-region client access topologies. On passthrough load balancers, packets arrive at backend VMs with the destination IP address still set to the load balancer's forwarding rule IP.
web-1 and web-2: An internal Application Load Balancer terminates connections at Envoy proxies in 10.10.2.0/24 and forwards requests to each backend's internal IP address. Verifying local routes or enabling global access (especially when clients are already in us-central1) does not allow 10.10.2.0/24 through the ingress firewall.web-3 health checks: Legacy Network Load Balancer ranges (209.85.152.0/22 and 209.85.204.0/22) are not used by internal Application Load Balancers, and web-3 remains untagged and blocked by the implied deny ingress rule.35.191.0.0/16 and 130.211.0.0/22, not legacy target-pool NLB ranges.table local are unnecessary for proxy-based load balancers because backend traffic targets each VM's primary interface IP.The deployment used a regional internal passthrough Network Load Balancer with clients located in another region, missing guest-agent local routes on backends, and legacy Network Load Balancer health checks rather than an internal Application Load Balancer.
A custom static route directs outbound VPC traffic toward a specific next hop—such as an internal passthrough Network Load Balancer forwarding rule (--next-hop-ilb)—often used to steer packets through virtual network appliances. VPC egress firewall rules evaluate outbound packets leaving a VM interface rather than inbound packets entering a backend.
10.20.0.0/16.web-1, web-2, and web-3—not a return-path routing failure or egress block.web-3's missing app-backend tag.The architecture required steering non-subnet traffic (such as 0.0.0.0/0) through an internal passthrough Network Load Balancer next hop for inspection appliances, and the Connectivity Tests trace indicated a missing return route or egress block.
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