Huawei HCIP-5G-Core (H35-651) Certification Sample Questions

01. A new slice, S-NSSAI X, is configured on the AMF and added to the subscriptions of a customer's devices. The gNBs in TA-1 were updated to support X and completed NG Setup again. The gNBs in TA-2 still support only the eMBB slice. The devices request both X and eMBB when they register.
What happens to devices that register in TA-2?
a) X is left out of their Allowed NSSAI, as TA-2 gNBs never reported it.
b) X is allowed at registration, but the SMF rejects every session request on it from TA-2.
c) Their registration is rejected until the TA-2 gNBs are updated.
d) X is allowed, because the AMF and the subscriptions both support it in every tracking area it serves.

02. An operator adds an enterprise slice, S-NSSAI X, and configures it on the gNBs, the AMF and the slice's SMF and UPF. Test UE-A requests X and receives it in its Allowed NSSAI. Test UE-B, in the same cell with the same device model and settings, also requests X, but its Allowed NSSAI contains only the default eMBB slice.
What should the engineer change?
a) Add X to the AMF's supported slices for this tracking area, so that the AMF can accept it.
b) Add X to the subscribed S-NSSAIs in UE-B's record on the UDM.
c) Add X to UE-B's Configured NSSAI.
d) Register X in the NF profile of the slice's SMF, so that the AMF can discover it for UE-B.

03. The two redundant instances of a UDM VNFC, deployed so that one can take over if the other fails, were both placed on the same compute host. When that host lost power, both instances went down and the UDM service stopped until the host came back.
Which placement rule should the VNFD have applied to these instances?
a) A larger flavor, so each instance has more vCPU.
b) Affinity, so the two instances share one host and fail over to each other with the least latency.
c) Anti-affinity, so the two instances are always placed on different compute hosts.
d) Volumes on shared storage reachable from every host.

04. A new SMF is added beside two existing SMFs for the internet DNN. A discovery query for that S-NSSAI and DNN returns all three profiles, and test sessions placed on the new SMF by hand succeed. A week later it carries under 2% of new sessions, while the other two share the rest.
What should the engineer check first?
a) The heartbeat interval between the new SMF and the NRF.
b) The S-NSSAI and DNN listed in the new SMF's profile.
c) The N4 associations between the new SMF and its UPFs.
d) The priority and capacity values in the new SMF's NF profile.

05. An operator is launching a new S-NSSAI for a logistics customer in two tracking areas. The team has confirmed that the customer's subscriptions list the S-NSSAI and that the SMF and UPF NF profiles registered at the NRF include it. The AMF uses the NSSF for slice selection.
Which two further items must be in place before a UE in those areas can use the slice?
(Choose two.)
a) The customer's devices use a new DNN created only for this slice.
b) The serving AMF is configured to support the S-NSSAI.
c) The MME in the EPC holds the S-NSSAI, so that the slice also exists for LTE access in those areas.
d) The UPF has a separate N3 address for each S-NSSAI.
e) The NSSF allows the S-NSSAI in both of the launch tracking areas.

06. In one cell, a test UE's registration never finishes. The AMF trace shows authentication and NAS security completing, subscription data being retrieved from the UDM and a Registration Accept being sent toward the gNB. No Registration Complete ever arrives from the UE, and the same subscriber registers normally in neighboring cells.
Where should the engineer look next?
a) At the UDM, whose subscription data for this subscriber may be incomplete.
b) At the AUSF, since authentication may have to be repeated before the UE can answer.
c) At the AMF's served TAI list, which may omit the tracking area of this cell.
d) At that cell's gNB and its radio link to the UE.

07. A trace of a UE's initial registration shows it arriving at AMF-1. AMF-1 retrieves the UE's subscription and queries the NSSF, and then the Registration Request is rerouted through the gNB to AMF-3 in another AMF set, which completes the registration. AMF-1 has no active alarms and its load is low.
What does the trace indicate?
a) AMF-1 rejected the UE for overload, and the gNB then selected another AMF on its own.
b) The UE's 5G-GUTI identified AMF-3, which AMF-1 does not own, so it forwarded the request.
c) AMF-1 cannot serve the slices the UE is allowed, so the UE was reallocated to an AMF set that serves them.
d) Authentication failed on AMF-1, and AMF-3 was chosen to retry it with a different AUSF.

08. An operator wants the PCF on the UPCF to apply a service area restriction to a group of fixed wireless subscribers, so that they are served only in their home tracking areas. Their subscription data carries no restriction. After rollout, the subscribers still register anywhere. A trace shows SM policy associations on N7 with the expected session decisions, and no N15 exchange for these UEs at registration.
Which change applies the restriction?
a) Add the service area restriction to the SM policy decisions that the PCF sends the SMF over N7.
b) Enable the AMF to create an AM policy association with the PCF at registration.
c) Provision the restriction on the gNBs so NG Setup delivers it.
d) Configure the SMF to pass its policy decision on to the AMF over N11 after each session.

09. UEs on DNN internet receive addresses from a private pool, 10.64.0.0/14. Sessions establish, the data-network router holds a route for the pool back to the UPF's N6 address, and pings from a UE to a server on the operator's own service network succeed. Pings from the same UE to public internet server addresses get no reply.
What must the engineer add?
a) NAT from the pool to public addresses, at or beyond N6.
b) DNS server addresses for the DNN, so that the UEs can resolve internet names.
c) A route for 10.64.0.0/14 toward the UPF's N6 address on the data-network router.
d) A larger Session-AMBR for internet sessions, so that the replies are not dropped or delayed at the UPF.

10. A UNC VNF package is onboarded and usable at a new site, and the NFVO grants the instantiation. The VNFM then asks the VIM for resources, and the request fails. The VIM reports that compute capacity is available, but that an external network named in the VNFD's connection points does not exist in the tenant.
What must be done before the instantiation is retried?
a) Re-onboard the VNF package, so the NFVO validates the VNFD again against its catalog.
b) Heal the partly created VNF.
c) Add compute hosts to the tenant, so the VIM has room for the VNF's virtual machines.
d) Create the network the VNFD's connection point calls for in the VIM tenant.