Question 700
A Python client receives HTTP 403 from NX-API even though it can reach the management address over HTTPS. A scheduled deployment is pending, and the engineer must validate how the script handles this failure condition. What post-change observation gives the strongest evidence of success?
- The authorized client runs the intended read-only API call and receives the expected structured result.
- Tests with absent fields return an explicit unknown/error outcome rather than a fabricated healthy metric.
- The network change reduces packet loss and congestion while measured training-step throughput improves.
- The utility reads the intended local input under the approved Guest Shell permissions and completes the task.
Explanation
Correct Answer
A:The authorized client runs the intended read-only API call and receives the expected structured result. This result tests the original failure rather than reporting merely that the configuration command was accepted. An HTTP 403 indicates a reachable endpoint has refused the request under its authorization controls. A scheduled deployment is pending, and the engineer must validate how the script handles this failure condition.
Incorrect Answers
B:Tests with absent fields return an explicit unknown/error outcome rather than a fabricated healthy metric. This concerns on-box python exception handling, not nx-api request authorization.
C:The network change reduces packet loss and congestion while measured training-step throughput improves. This concerns ai cluster network congestion, not nx-api request authorization.
D:The utility reads the intended local input under the approved Guest Shell permissions and completes the task. This concerns guest shell isolation, not nx-api request authorization.
Question 701
A Nexus vPC secondary suspends downstream vPC member ports when the peer-link goes down but peer-keepalive still reaches the primary. The team is preparing a controlled production change and must preserve fabric reachability. What caveat is critical when interpreting a successful remediation?
- Forcing every redundant port to forward would create a bridging loop rather than improve resilience.
- LACP negotiation cannot compensate for incompatible allowed VLAN or switching-mode settings.
- A successful subscription does not guarantee every sensor or transient event is supported by the platform.
- A working keepalive is not a replacement data path for the vPC peer-link.
Explanation
Correct Answer
D:A working keepalive is not a replacement data path for the vPC peer-link. With peer-link failure and a reachable peer, the secondary suspends vPC member ports to prevent loops and dual-active forwarding. The team is preparing a controlled production change and must preserve fabric reachability.
Incorrect Answers
A:Forcing every redundant port to forward would create a bridging loop rather than improve resilience. That response is associated with rstp blocked uplink rather than the specific vpc peer-link recovery condition described here. For the stated failure, the preferred action is: Restore the peer-link and validate keepalive and vPC consistency rather than forcing both peers to forward independently. Relevant evidence: Inspect peer-link state, peer-keepalive reachability, consistency checks and suspended port reasons. The team is preparing a controlled production change and must preserve fabric reachability.
B:LACP negotiation cannot compensate for incompatible allowed VLAN or switching-mode settings. That response is associated with lacp member mismatch rather than the specific vpc peer-link recovery condition described here. For the stated failure, the preferred action is: Restore the peer-link and validate keepalive and vPC consistency rather than forcing both peers to forward independently. Relevant evidence: Inspect peer-link state, peer-keepalive reachability, consistency checks and suspended port reasons. The team is preparing a controlled production change and must preserve fabric reachability.
C:A successful subscription does not guarantee every sensor or transient event is supported by the platform. This concerns streaming telemetry health, not vpc peer-link recovery.
Question 702
Switch administrators lose access during a TACACS+ outage because there is no tested, secured local recovery path. Security engineering must restore the approved service without relaxing unrelated protection. What should the engineer collect before ruling out other causes?
- Inspect platform capability, MACsec session state, MKA status, cipher suite and protected traffic counters.
- Inspect EPG bindings, contracts, subject filters, provider-consumer association and zoning rules.
- Inspect TACACS+ reachability, AAA method lists, local credentials and access accounting.
- Inspect assigned user roles, AAA server attributes, command authorization and audit logs.
Explanation
Correct Answer
C:Inspect TACACS+ reachability, AAA method lists, local credentials and access accounting. These records help distinguish the suspected failure from a superficially similar fault before modifying production. Remote AAA improves centralized administration but must have an appropriate resilient recovery mechanism. Security engineering must restore the approved service without relaxing unrelated protection.
Incorrect Answers
A:Inspect platform capability, MACsec session state, MKA status, cipher suite and protected traffic counters. This concerns MACsec protection scope, not centralized AAA fallback.
B:Inspect EPG bindings, contracts, subject filters, provider-consumer association and zoning rules. That response is associated with aci inter-epg contract rather than the specific centralized aaa fallback condition described here. For the stated failure, the preferred action is: Configure and test the approved AAA method order and protected break-glass account following change policy. Relevant evidence: Inspect TACACS+ reachability, AAA method lists, local credentials and access accounting. Security engineering must restore the approved service without relaxing unrelated protection.
D:Inspect assigned user roles, AAA server attributes, command authorization and audit logs. That response is associated with nx-os administrative rbac rather than the specific centralized aaa fallback condition described here. For the stated failure, the preferred action is: Configure and test the approved AAA method order and protected break-glass account following change policy. Relevant evidence: Inspect TACACS+ reachability, AAA method lists, local credentials and access accounting. Security engineering must restore the approved service without relaxing unrelated protection.
Question 703
An orchestration service retries a configuration POST after timeout and produces duplicate resources on a Nexus management controller. The automation change must pass a preproduction test and leave a verifiable audit trail. Why is the targeted engineering response justified in this situation?
- An HTTP 403 indicates a reachable endpoint has refused the request under its authorization controls.
- Distributed AI jobs can be bottlenecked by data-center fabric transport when many nodes exchange bursts of data.
- Retrying non-idempotent create operations without state checks can duplicate resources after ambiguous timeouts.
- Automation must distinguish unavailable telemetry from a real numeric zero to avoid false success reports.
Explanation
Correct Answer
C:Retrying non-idempotent create operations without state checks can duplicate resources after ambiguous timeouts. The automation change must pass a preproduction test and leave a verifiable audit trail.
Incorrect Answers
A:An HTTP 403 indicates a reachable endpoint has refused the request under its authorization controls. That response is associated with nx-api request authorization rather than the specific rest api retries and idempotence condition described here. For the stated failure, the preferred action is: Check API operation semantics and implement safe read-before-create or idempotent reconciliation behavior. Relevant evidence: Inspect API logs, resource identifiers, HTTP response history and current running object state. The automation change must pass a preproduction test and leave a verifiable audit trail.
B:Distributed AI jobs can be bottlenecked by data-center fabric transport when many nodes exchange bursts of data. This concerns ai cluster network congestion, not rest api retries and idempotence.
D:Automation must distinguish unavailable telemetry from a real numeric zero to avoid false success reports. That response is associated with on-box python exception handling rather than the specific rest api retries and idempotence condition described here. For the stated failure, the preferred action is: Check API operation semantics and implement safe read-before-create or idempotent reconciliation behavior. Relevant evidence: Inspect API logs, resource identifiers, HTTP response history and current running object state. The automation change must pass a preproduction test and leave a verifiable audit trail.
Question 704
A SAN supervisor EPLD change is scheduled during peak transaction load without assessing module reload requirements. A storage change board requires a traceable diagnosis before allowing further fabric changes. Which observations would best validate the initial diagnosis?
- Inspect upstream NPIV configuration, NPV status, uplink VSAN, FLOGI database and server WWPNs.
- Inspect alias database, recorded target WWPNs, zones using the alias and change approvals.
- Inspect FCNS database, FLOGI records, target WWPN, VSAN association and port transition logs.
- Inspect MDS platform release notes, EPLD package compatibility, reload dependencies and alternate paths.
Explanation
Correct Answer
D:Inspect MDS platform release notes, EPLD package compatibility, reload dependencies and alternate paths. These records help distinguish the suspected failure from a superficially similar fault before modifying production. An EPLD update may reset affected SAN hardware even if a previous NX-OS patch required no outage. A storage change board requires a traceable diagnosis before allowing further fabric changes.
Incorrect Answers
A:Inspect upstream NPIV configuration, NPV status, uplink VSAN, FLOGI database and server WWPNs. This concerns npv uplink fabric services, not mds epld update impact.
B:Inspect alias database, recorded target WWPNs, zones using the alias and change approvals. This concerns device alias maintenance, not mds epld update impact.
C:Inspect FCNS database, FLOGI records, target WWPN, VSAN association and port transition logs. This concerns fibre channel name-server entries, not mds epld update impact.
Question 705
A UCS blade shows unexplained throttling during compute peaks despite normal guest operating-system logs. Operations is validating the host before approving production workload placement. What should the engineer collect before ruling out other causes?
- Inspect capture-point support, vNIC policy, frame timestamps, queue counters and retransmission patterns.
- Inspect adapter health, fault codes, service-profile inventory, event chronology and firmware state.
- Inspect effective vNIC templates, VLAN object assignments, LAN uplink trunk allowances and fault logs.
- Inspect chassis health, power budgets, fan alarms, temperature sensors and affected blade performance.
Explanation
Correct Answer
D:Inspect chassis health, power budgets, fan alarms, temperature sensors and affected blade performance. These records help distinguish the suspected failure from a superficially similar fault before modifying production. Physical power or cooling constraints can throttle compute even when guest metrics show no software error. Operations is validating the host before approving production workload placement.
Incorrect Answers
A:Inspect capture-point support, vNIC policy, frame timestamps, queue counters and retransmission patterns. That response is associated with compute traffic span rather than the specific chassis power and thermal monitoring condition described here. For the stated failure, the preferred action is: Inspect chassis power, fan and thermal faults before changing application performance settings. Relevant evidence: Inspect chassis health, power budgets, fan alarms, temperature sensors and affected blade performance. Operations is validating the host before approving production workload placement.
B:Inspect adapter health, fault codes, service-profile inventory, event chronology and firmware state. That response is associated with intersight alert triage rather than the specific chassis power and thermal monitoring condition described here. For the stated failure, the preferred action is: Inspect chassis power, fan and thermal faults before changing application performance settings. Relevant evidence: Inspect chassis health, power budgets, fan alarms, temperature sensors and affected blade performance. Operations is validating the host before approving production workload placement.
C:Inspect effective vNIC templates, VLAN object assignments, LAN uplink trunk allowances and fault logs. This concerns UCS VLAN pool and template, not chassis power and thermal monitoring.
Question 706
A SAN supervisor EPLD change is scheduled during peak transaction load without assessing module reload requirements. During the next path-failover exercise, the SAN team must confirm the same dependency on the affected VSAN. What should the engineer collect before ruling out other causes?
- Inspect current/target images, supported upgrade matrix, module health and connected-host multipathing.
- Inspect export paths, permitted clients, protocol version, mount request and server logs.
- Inspect trunk allowed lists, operational VSANs, fcdomain state and VSAN-specific fabric links.
- Inspect MDS platform release notes, EPLD package compatibility, reload dependencies and alternate paths.
Explanation
Correct Answer
D:Inspect MDS platform release notes, EPLD package compatibility, reload dependencies and alternate paths. These records help distinguish the suspected failure from a superficially similar fault before modifying production. An EPLD update may reset affected SAN hardware even if a previous NX-OS patch required no outage. During the next path-failover exercise, the SAN team must confirm the same dependency on the affected VSAN.
Incorrect Answers
A:Inspect current/target images, supported upgrade matrix, module health and connected-host multipathing. This concerns mds nondisruptive upgrade planning, not mds epld update impact.
B:Inspect export paths, permitted clients, protocol version, mount request and server logs. This concerns nfs export permissions, not mds epld update impact.
C:Inspect trunk allowed lists, operational VSANs, fcdomain state and VSAN-specific fabric links. This concerns VSAN trunk mismatch, not mds epld update impact.
Question 707
A SAN supervisor EPLD change is scheduled during peak transaction load without assessing module reload requirements. A storage change board requires a traceable diagnosis before allowing further fabric changes. Why is the targeted engineering response justified in this situation?
- An initiator must complete fabric login and receive an FCID before normal fabric services can identify it.
- An EPLD update may reset affected SAN hardware even if a previous NX-OS patch required no outage.
- Fibre Channel port channels bundle compatible physical links; inconsistent member configuration can suspend interfaces.
- Topology, fabric events and time-series indicators reveal path problems hidden by simple reachability polling.
Explanation
Correct Answer
B:An EPLD update may reset affected SAN hardware even if a previous NX-OS patch required no outage. A storage change board requires a traceable diagnosis before allowing further fabric changes.
Incorrect Answers
A:An initiator must complete fabric login and receive an FCID before normal fabric services can identify it. This concerns fibre channel fabric login, not mds epld update impact.
C:Fibre Channel port channels bundle compatible physical links; inconsistent member configuration can suspend interfaces. This concerns fibre channel port-channel consistency, not mds epld update impact.
D:Topology, fabric events and time-series indicators reveal path problems hidden by simple reachability polling. This concerns nexus dashboard SAN monitoring, not mds epld update impact.
Question 708
A blade's vHBA is configured but the upstream SAN fabric never sees its WWPN login. The issue reappears on a second managed host, so the team is checking its effective profiles and hardware state. Which intervention most directly addresses the stated condition?
- Review UCS QoS classes and upstream priority handling across the entire network path.
- Assign WWPNs from a nonoverlapping pool and validate effective per-fabric vHBA identities.
- Stage and review the template relationship and scope before applying an impactful shared-policy change.
- Validate the UCS SAN uplink, vHBA fabric pinning and VSAN assignment before modifying LUN masks.
Explanation
Correct Answer
D:Validate the UCS SAN uplink, vHBA fabric pinning and VSAN assignment before modifying LUN masks. A blade vHBA needs an operational fabric path and correct VSAN association before a fabric login can occur. The issue reappears on a second managed host, so the team is checking its effective profiles and hardware state.
Incorrect Answers
A:Review UCS QoS classes and upstream priority handling across the entire network path. This concerns UCS network qos policy, not SAN connectivity from blade chassis.
B:Assign WWPNs from a nonoverlapping pool and validate effective per-fabric vHBA identities. This concerns wwpn pool allocation, not SAN connectivity from blade chassis.
C:Stage and review the template relationship and scope before applying an impactful shared-policy change. This concerns updating service-profile templates, not SAN connectivity from blade chassis.
Question 709
Updating a UCS updating-template appears to change vNIC settings on more blades than intended. Operations is validating the host before approving production workload placement. What should the operator avoid assuming from the observed state?
- Successful one-time installation does not ensure the persistent boot policy has been corrected.
- A template update must not be assumed to affect just the blade presently open in the UI.
- Mirroring an inactive or wrong vNIC path cannot prove what happened to the actual workload traffic.
- Manual association to an incompatible blade may bypass placement intent without satisfying hardware needs.
Explanation
Correct Answer
B:A template update must not be assumed to affect just the blade presently open in the UI. Updating templates can propagate modifications to associated service profiles; change impact depends on associations. Operations is validating the host before approving production workload placement.
Incorrect Answers
A:Successful one-time installation does not ensure the persistent boot policy has been corrected. That response is associated with boot order and removable media rather than the specific updating service-profile templates condition described here. For the stated failure, the preferred action is: Stage and review the template relationship and scope before applying an impactful shared-policy change. Relevant evidence: Inspect parent templates, derived profiles, policy inheritance and pending association changes. Operations is validating the host before approving production workload placement.
C:Mirroring an inactive or wrong vNIC path cannot prove what happened to the actual workload traffic. This concerns compute traffic span, not updating service-profile templates.
D:Manual association to an incompatible blade may bypass placement intent without satisfying hardware needs. That response is associated with server pool qualification rather than the specific updating service-profile templates condition described here. For the stated failure, the preferred action is: Stage and review the template relationship and scope before applying an impactful shared-policy change. Relevant evidence: Inspect parent templates, derived profiles, policy inheritance and pending association changes. Operations is validating the host before approving production workload placement.
Question 710
An automated UCS association binds a high-memory workload to a blade that lacks the required adapter capability. The issue reappears on a second managed host, so the team is checking its effective profiles and hardware state. What is the best next implementation or troubleshooting step?
- Use a staged maintenance sequence with workload evacuation, redundancy validation and documented rollback criteria.
- Validate the service-profile SAN boot target WWPN/LUN settings and underlying zoning/masking path.
- Correct server-pool qualifications and service-profile placement criteria for the workload's hardware dependencies.
- Stage and review the template relationship and scope before applying an impactful shared-policy change.
Explanation
Correct Answer
C:Correct server-pool qualifications and service-profile placement criteria for the workload's hardware dependencies. Pool qualification policies can restrict association based on compatible server hardware characteristics. The issue reappears on a second managed host, so the team is checking its effective profiles and hardware state.
Incorrect Answers
A:Use a staged maintenance sequence with workload evacuation, redundancy validation and documented rollback criteria. This concerns UCS firmware change sequencing, not server pool qualification.
B:Validate the service-profile SAN boot target WWPN/LUN settings and underlying zoning/masking path. This concerns SAN boot path policy, not server pool qualification.
D:Stage and review the template relationship and scope before applying an impactful shared-policy change. This concerns updating service-profile templates, not server pool qualification.
Implementing and Operating Cisco Data Center Core Technologies (DCCOR) (350-601) Practice Tests