Huawei HCIP-Datacom Solution Architect (H12-822) Certification Sample Questions

01. A university lab switch shows two recurring problems on its access ports. One port occasionally floods the segment with broadcast traffic from a faulty adapter, and students attach small hubs that push the switch's forwarding table toward exhaustion.
Which two mechanisms address these problems respectively?
(Choose two.)
a) Port security, which caps how many distinct source MAC addresses one access port may learn
b) MAC flapping detection, which blocks a port when an address keeps moving
c) BPDU protection, a spanning tree safeguard configured on the access ports that face end users
d) Loop detection, which disables a port that receives a frame the switch itself sent
e) Storm control, which discards broadcast frames on a port once their rate passes the configured limit

02. A campus WLAN uses a central controller. The security team requires that all wireless user traffic pass a single enforcement point before reaching the wired network, and the same SSID must be offered from access points in several buildings that sit in different wired subnets.
Which forwarding design meets both requirements, and what does it cost?
a) Tunnel forwarding, which keeps management traffic on the controller while user traffic exits locally, at the cost of duplicating policy.
b) Direct forwarding, which places user traffic onto the local wired segment at each access point, at the cost of a longer path back to the controller.
c) Tunnel forwarding, which carries user traffic to the controller for enforcement, at the cost of concentrating load and lengthening the path.
d) Direct forwarding, which extends a single broadcast domain across every building, at the cost of controller processing.

03.Huawei presents its enterprise network portfolio as a set of solutions, each addressed at a different part of an enterprise network.
Which mapping of solution to the part of the network it addresses is correct?
a) CloudCampus addresses the data center network, and CloudFabric addresses the campus network.
b) CloudCampus addresses the campus, and CloudFabric addresses the data center.
c) CloudFabric addresses the branch interconnection, and CloudCampus addresses the wide area transport network.
d) CloudWAN addresses the campus network, and SD-WAN addresses the data center fabric.

04. An egress interface uses a strict-priority queue for voice and a set of weighted queues for everything else. After a signaling change, an application begins sending a large and sustained volume of traffic that is classified into the priority queue.
Operators report that traffic in the weighted queues has almost stopped moving. What does this reveal about the scheduler?
a) Strict priority reorders packets across queues, so batch traffic is delayed rather than dropped and resumes after the burst.
b) The weighted queues carry weights that sum to less than the interface rate, so part of their share is left unclaimed.
c) The scheduler has fallen back to tail drop on the weighted queues because the priority queue exceeded its depth.
d) A strict-priority queue is served whenever it holds anything, so an unbounded volume in it starves the queues beneath.

05. A branch internet link saturates during the working day because a few users run a bulk file-sharing application, and an interactive business application on the same link suffers.
Which traffic management action at the security gateway addresses the cause?
a) Cap the file-sharing application per user and guarantee the business application a share of the link.
b) Translate the file-sharing users to a separate public address so their traffic leaves the branch by a different path.
c) Apply an attack-defense threshold to the file-sharing sessions so the gateway drops them once their rate looks anomalous.
d) Raise the bandwidth limit on the business application so it can claim more of the link when the two compete.

06. A hotel network must let every guest room reach a shared set of servers while guest rooms remain unable to reach one another, including rooms on the same floor.
Which advanced VLAN technology matches this requirement directly?
a) QinQ, with an outer tag per floor so that each floor's rooms are carried transparently across the shared distribution layer
b) One VLAN per room plus a shared server VLAN, with the gateway routing between every VLAN pair as required
c) MUX VLAN, with the servers in a principal VLAN so that rooms in subordinate VLANs reach them
d) A voice VLAN on each access port, so that server traffic is classified separately from the traffic the rooms generate

07. An open-plan office floor seats 400 people in an area that a coverage-driven calculation would serve with very few access points. Each person carries two wireless devices and video conferencing from the desk is routine.
How should the plan resolve the gap between the coverage-driven and the capacity-driven access point counts?
a) Adopt the coverage-driven count and raise transmit power, so the same access points serve more devices over a wider area.
b) Adopt the capacity-driven count and shrink the coverage area of each access point to match.
c) Adopt the coverage-driven count and rely on the higher data rates nearby clients achieve to absorb the extra devices.
d) Adopt whichever count is lower and revisit the plan after deployment, since measured demand always differs from the predicted demand.

08. A fabric carries ordinary TCP application traffic and also remote direct memory access traffic for a storage and AI training cluster.
Why does the second class of traffic require a lossless fabric while the first does not?
a) TCP hides loss by retransmitting at the application layer, whereas this traffic has no retransmission mechanism at all.
b) It runs over UDP, which carries no checksum, so a dropped frame cannot be detected anywhere along the path.
c) Its frames are larger than the forwarding pipeline can buffer, so a drop must be prevented before the frame is queued.
d) Its transport recovers from a drop far more expensively than TCP does.

09. A team familiar with OSPF is deploying IS-IS. In OSPF an area boundary falls inside a router, which may hold interfaces in two areas at once.
How does the IS-IS area model differ?
a) A router belongs entirely to one area, so the boundary falls on the link between two routers
b) A router must join the backbone area in addition to its own area, so every boundary is drawn through the backbone
c) A router places each of its interfaces in an area independently, and the boundary falls wherever two interfaces in different areas meet inside it
d) A router joins as many areas as it has adjacencies, and membership is recorded per adjacency in the database

10. An enterprise is scoping a CloudWAN-based redesign of its transport WAN. The existing network carries traffic between data centers, cloud egress points, and regional aggregation sites. It was dimensioned for steady inter-site traffic, and its paths are provisioned by hand.
Which of these is a challenge that the redesign is expected to address?
a) Manually provisioned paths cannot express a routing metric, so the network has no way to select a shortest path at all.
b) Hand provisioning cannot keep pace with traffic whose volume and direction change as workloads are placed and moved.
c) Traffic between data centers is unpredictable in size, so no dimensioning method applies to it.
d) Hand-provisioned paths cannot carry encrypted traffic, so the cloud egress points have to be redesigned first.