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 operates a multi-tenant GKE Standard cluster (v1.36) with GKE Dataplane V2 and Workload Identity Federation for GKE enabled, using no third-party CNI add-ons.
You must enforce these controls across namespaces:
audit-vault namespace from all cluster traffic.NetworkPolicy resources.payments Pods to only reach backend Pods on TCP port 8443, query DNS in kube-system, and access the GKE metadata server.Which policy configuration should you implement?
Admin-tier ClusterNetworkPolicy with Deny rules isolating audit-vault and a Baseline-tier ClusterNetworkPolicy with a Deny rule for cluster-wide ingress; in payments, configure an egress NetworkPolicy using ipBlock.cidr for the backend Service ClusterIP and Pod range with protocol: TCP (omitting port), allowing TCP/UDP port 53 to kube-system, and allowing 169.254.169.254/32 on TCP ports 80 and 8080.This option correctly structures the Admin and Baseline ClusterNetworkPolicy tiers and targets the right metadata server endpoint, but it configures the payments egress NetworkPolicy using ipBlock.cidr for cluster workloads and omits the port field alongside protocol: TCP.
ipBlock.cidr for Pod or Service IPs: Under GKE Dataplane V2, NetworkPolicy manifests cannot use Pod IP addresses or Service ClusterIP addresses in the ipBlock.cidr field. Internal cluster workloads must be selected using podSelector and namespaceSelector labels.ports.port field: GKE Dataplane V2 does not permit specifying a protocol without a corresponding port number in a NetworkPolicy manifest.backend access strictly to TCP port 8443.ipBlock is intended for external CIDR blocks outside cluster workloads.ports that defines protocol (such as TCP, UDP, or SCTP) must explicitly specify a port.If the payments egress policy replaced ipBlock.cidr with namespaceSelector and podSelector targeting the backend Pods and explicitly specified port: 8443 under protocol: TCP, it would be fully valid on GKE Dataplane V2.
Admin-tier ClusterNetworkPolicy with Deny rules isolating audit-vault and a Baseline-tier ClusterNetworkPolicy with a Deny rule for cluster-wide ingress; in payments, configure an egress NetworkPolicy using namespaceSelector and podSelector for backend on TCP port 8443, allowing TCP/UDP port 53 to kube-system, and allowing 169.254.169.252/32 on TCP ports 987 and 988.This configuration correctly applies the three-tier policy hierarchy using Admin and Baseline ClusterNetworkPolicy resources alongside label-based NetworkPolicy egress rules, but it targets the wrong GKE metadata server IP address and ports for a Dataplane V2 cluster.
169.254.169.254/32 on ports 80 and 8080. Because a NetworkPolicy specifying egress rules implicitly denies all unmatched outbound traffic, allowing only 169.254.169.252/32 on ports 987 and 988 blocks Pods in payments from retrieving IAM credentials.169.254.169.252/32 (ports 987/988): Used by the GKE metadata server on standard dataplane (non-Dataplane V2) clusters running GKE 1.21.0-gke.1000 and later.127.0.0.1/32 (ports 987/988): Used by legacy clusters earlier than 1.21.0-gke.1000.169.254.169.254/32 (ports 80/8080): Required specifically when GKE Dataplane V2 is active.This metadata server rule would be correct if the cluster were using the legacy dataplane with the Calico network policy plugin on GKE 1.21.0-gke.1000 or later instead of GKE Dataplane V2 (note that ClusterNetworkPolicy itself also requires Dataplane V2).
Admin-tier ClusterNetworkPolicy isolating audit-vault and a lower-priority Admin-tier ClusterNetworkPolicy with a Deny rule for cluster-wide ingress; in payments, configure an egress NetworkPolicy using namespaceSelector and podSelector for backend on TCP port 8443, allowing TCP/UDP port 53 to kube-system, and allowing 169.254.169.254/32 on TCP ports 80 and 8080.This option places both the audit-vault isolation policy and the cluster-wide default Deny ingress policy inside the Admin tier of ClusterNetworkPolicy, attempting to rely on numeric priority ordering (0 to 1000) to separate them.
NetworkPolicy tier: In the GKE policy evaluation pipeline, all Admin-tier ClusterNetworkPolicy rules execute before any namespace-scoped NetworkPolicy resource is evaluated.Deny: When a packet matches a Deny rule in the Admin tier—even a lower-priority one—evaluation immediately terminates and the traffic is dropped. Application teams cannot selectively permit ingress using standard NetworkPolicy manifests.Admin tier -> NetworkPolicy tier -> Baseline tier -> Default implicit allow.ClusterNetworkPolicy objects within the same tier; a lower-priority Admin policy still outranks every standard NetworkPolicy.Pass vs. Baseline: To let namespace NetworkPolicy rules apply, administrators either use an explicit Pass action in the Admin tier for targeted traffic or place default-deny rules in the Baseline tier.This configuration would be correct if the requirement were to enforce a permanent, non-overridable cluster-wide ingress blockade, or if the second policy used tier: Baseline instead of tier: Admin.
Admin-tier ClusterNetworkPolicy with Deny rules isolating audit-vault and a Baseline-tier ClusterNetworkPolicy with a Deny rule for cluster-wide ingress; in payments, configure an egress NetworkPolicy using namespaceSelector and podSelector for backend on TCP port 8443, allowing TCP/UDP port 53 to kube-system, and allowing 169.254.169.254/32 on TCP ports 80 and 8080.This solution combines Kubernetes-native multi-tier ClusterNetworkPolicy resources with namespace-scoped NetworkPolicy rules enforced directly by GKE Dataplane V2's eBPF data path (built on Cilium) without requiring Calico or external CNI add-ons.
audit-vault): Placing Deny rules in the Admin tier of a ClusterNetworkPolicy gives them the highest evaluation precedence. Any matching ingress or egress packet immediately stops evaluation and is dropped before reaching namespace policies.Deny ingress rule in the Baseline tier establishes cluster-wide zero-trust behavior at the lowest precedence tier. Standard NetworkPolicy resources sit in the middle NetworkPolicy tier, allowing application teams to explicitly permit authorized ingress traffic.payments egress: Using namespaceSelector and podSelector with an explicit protocol: TCP and port: 8443 adheres to Dataplane V2 enforcement rules, which require label selectors for workloads and explicit port numbers whenever a protocol is specified.53 to kube-system enables service discovery, while permitting 169.254.169.254/32 on TCP ports 80 and 8080 matches the exact GKE metadata server endpoint used on Dataplane V2 clusters.iptables chains or kube-proxy.Admin), developer application rules (NetworkPolicy), and fallback defaults (Baseline).This design leverages GKE Dataplane V2's native support for both ClusterNetworkPolicy and NetworkPolicy while respecting Dataplane V2-specific metadata server endpoints and selector syntax.
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