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 designing a multi-region Shared VPC connected to on-premises via Dedicated Interconnect. Your IPAM and subnet architecture must satisfy these constraints:
10.0.0.0/8, leaving only 10.200.0.0/16 free in RFC 1918; VMs run Docker with default bridge networking.Which subnet and IP address allocation strategy should you implement?
/29 Micro-Subnets DoThis design places each workload team in an independent VPC linked to a hub via VPC Network Peering, attempts to reuse 10.200.0.0/16 across spokes and Private Services Access (PSA), and mimics legacy physical VLANs by creating a /29 subnet for each microservice.
10.200.0.0/16 across peered spokes or overlapping 10.200.0.0/21 for PSA with active primary subnets causes immediate peering and subnet creation failures./29 subnet provides only 8 IPv4 addresses, and 4 addresses are reserved and unusable in every primary subnet range, leaving only 4 usable IPs per microservice—insufficient for GKE nodes, internal load balancers, or scaling./29 subnets adds administrative overhead without routing benefits and complicates hub-and-spoke non-transitivity.Teams operated isolated floating VPCs without VPC Network Peering—exposing discrete endpoints via Private Service Connect where overlapping 10.200.0.0/16 CIDRs are permitted—and used appropriately sized subnets rather than /29 blocks.
172.0.0.0/10, GKE-Managed Shared VPC Subnets, and Source-Tag Rules AttemptThis approach tries to create broad 172.0.0.0/10 regional ranges, delegate secondary IP range creation to the GKE control plane from within service projects, and isolate Pod alias IP traffic using network source tags in firewall rules.
172.0.0.0/10 because a single subnet range cannot span both an RFC 1918 private block (172.16.0.0/12) and privately used public IP (PUPI) space. Even if split, 172.16.0.0/12 is unavailable due to partner/on-premises reservations.roles/compute.networkUser permissions, while host project administrators retain sole control over creating subnets and secondary ranges.Clusters were deployed in standalone projects (where GKE can auto-manage secondary ranges) using valid non-spanning CIDRs, and firewall policies used CIDR source ranges instead of source tags to filter Pod traffic.
This strategy combines a custom mode Shared VPC with a hierarchical IP address management (IPAM) plan that separates scarce RFC 1918 primary infrastructure space from RFC 6598 (100.64.0.0/10) secondary ranges used for GKE Pod alias IPs. In a Shared VPC, a host project centralizes subnets, routes, and firewalls while service projects deploy workloads into authorized subnets.
10.128.0.0/9 block, which would collide with on-premises 10.0.0.0/8 networks. Carving primary ranges and the global /21 Private Services Access (PSA) block solely from 10.200.0.0/16 prevents Dedicated Interconnect BGP conflicts and avoids 172.17.0.0/16, which is used internally by Docker's default bridge network.10.200.0.0/16 ensures predictable expansion to six regions. Offloading high-density Pod ranges to 100.64.0.0/10 preserves scarce RFC 1918 space while sizing Pods generously upfront—crucial because secondary IPv4 ranges do not support CIDR expansion after creation.roles/compute.networkUser at the subnet level enforces least-privilege regional boundaries per service project. Pre-creating secondary ranges in the host project satisfies Shared VPC GKE prerequisites, and using CIDR source/destination ranges in firewall rules properly filters alias IPs where source tags fail.10.200.0.0/16 can be expanded later if node or VM counts exceed initial estimates./21 PSA range lets managed services like Cloud SQL automatically allocate regional /24 subnets without overlapping VM subnets.Unlike auto mode networks, overlapping spoke VPCs, or undersized secondary ranges, this architecture maximizes limited RFC 1918 capacity, accommodates immutable secondary Pod CIDRs, and enforces strict regional and workload isolation across teams.
172.17.0.0/16 Allocation DoConverting an auto mode VPC network to custom mode stops Google Cloud from automatically provisioning subnets in newly launched regions, but it preserves all existing predefined subnets inside 10.128.0.0/9. Adding 172.17.0.0/16 attempts to reclaim extra private space for Docker hosts, while sizing secondary ranges at /24 assumes they can grow dynamically later.
10.128.0.0/9 subnets overlaps directly with the organization's existing 10.0.0.0/8 on-premises footprint, causing route collisions and breaking hybrid connectivity over Dedicated Interconnect.172.17.0.0/16 to a VPC subnet conflicts with the default Docker bridge network configured inside the VM guest operating system, leading to broken container or VPC routing./24 with a plan to expand them fails because CIDR expansion is not supported for secondary IPv4 ranges. Exhausting a /24 range requires creating new clusters, recreating node pools with lower max-pods-per-node, or configuring discontiguous multi-Pod CIDRs./20 ranges within 10.128.0.0/9, preventing VPC Network Peering between auto mode VPCs and colliding with 10.0.0.0/8 enterprise networks.172.17.0.0/16 inside the OS require excluding that CIDR from all primary and secondary VPC ranges.The organization had zero 10.128.0.0/9 overlap on-premises, did not use default Docker bridge networking (172.17.0.0/16), and pre-sized its secondary Pod ranges large enough without relying on CIDR expansion.
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