Cisco CCNP Enterprise 350-401 ENCOR YANG Controller APIs and REST Responses Practice Test

 

Topic 20 covers yang, controller apis and rest responses for the Cisco Certified Specialist – Enterprise Core certification. These original practice questions apply the verified 350-401 objectives to practical decisions and troubleshooting. Select one answer unless a fixed number is requested. For broader preparation, visit the Cisco 350-401 ENCOR Exam Dumps page. Each option includes an explanation of the relevant behavior and scenario constraints.

Question 1

An engineer is designing an automation client. The device exposes configuration through RESTCONF, but the team is debating whether YANG is the API transport. Which statement is accurate?

1. YANG is an XML-only RPC transport equivalent to NETCONF.
2. YANG is a Cisco-specific CLI parser that translates every command into JSON.
3. YANG defines the modeled data structure; RESTCONF exposes it over HTTP.
4. YANG replaces HTTPS by encrypting the modeled data before transport.
5. YANG is the session token format used by controller REST APIs.

Correct Answer: C

 

Correct Answer

Answer C is correct because YANG describes data schema and semantics, while RESTCONF provides an HTTP-based protocol for accessing YANG-modeled resources.

Incorrect Answers

Answer A is incorrect because YANG can model data used by multiple interfaces; NETCONF is the protocol and is not interchangeable with the model language.

Answer B is incorrect because YANG is a standards-based modeling language and is not simply a proprietary CLI parser.

Answer D is incorrect because YANG is a modeling language and does not provide transport encryption or replace HTTPS.

Answer E is incorrect because Controller authentication tokens are separate from YANG data modeling and are not YANG schemas.

 

Question 2

A YANG module models sites. Each site can contain multiple devices, and every device entry has a unique serial number. Which TWO schema constructs best represent these requirements? Choose TWO.

1. A container can group the child nodes that belong to a site.
2. A leaf must contain all devices as one comma-separated string.
3. A notification is required to create each device entry in the configuration tree.
4. A list can represent repeated device entries under the site.
5. An identity statement must replace the site hierarchy because containers cannot have children.
6. An RPC must be used because repeated data cannot be modeled as configuration.

Correct Answers: A, D

 

Correct Answers

Answer A is correct because A container is suitable for grouping related child data under a single structural node.

Answer D is correct because A YANG list represents repeated entries and can define keys that identify individual entries.

Incorrect Answers

Answer B is incorrect because A leaf represents a single value and should not replace a repeated structured list in this design.

Answer C is incorrect because Notifications report events and are not the required structural construct for repeated configured entries.

Answer E is incorrect because Containers are specifically used to hold child nodes, so this claim reverses their basic structural role.

Answer F is incorrect because YANG lists can model repeated configuration or state data without requiring an RPC merely because entries repeat.

 

Question 3

A YANG snippet declares `leaf mtu { type uint16; }`. A client reads the schema before constructing a payload. What should the client infer?

1. `mtu` must be transported with NETCONF and cannot appear through RESTCONF.
2. `mtu` is automatically a list key even if no list declares it as a key.
3. `mtu` is a single modeled value whose schema type is an unsigned 16-bit integer.
4. `mtu` is operational-only because every numeric YANG leaf is read-only.
5. `mtu` is a controller task identifier because uint16 values are reserved for tasks.

Correct Answer: C

 

Correct Answer

Answer C is correct because A leaf represents one value, and the declared uint16 type constrains the modeled value to that data type.

Incorrect Answers

Answer A is incorrect because The YANG model can be exposed by supported model-driven interfaces, including RESTCONF where available.

Answer B is incorrect because A leaf becomes a list key only when the enclosing list explicitly names it in the key statement.

Answer D is incorrect because Numeric type does not determine configuration status; config versus state depends on the schema context and statements.

Answer E is incorrect because The type only constrains the leaf value and does not assign controller-specific task semantics.

 

Question 4

A YANG list named `interface` uses `key “name”`. The payload contains two entries with names `GigabitEthernet1` and `GigabitEthernet2`. Which TWO conclusions are valid? Choose TWO.

1. The key converts operational state into configuration data.
2. The key forces the list to contain exactly two entries.
3. The key selects which HTTP method RESTCONF must use for every operation.
4. The key encrypts the interface name when the payload is sent.
5. The `name` value identifies an individual entry within that list.
6. Two entries cannot use the same key value within the same list instance.

Correct Answers: E, F

 

Correct Answers

Answer E is correct because A YANG list key supplies the identifying value used to distinguish individual list entries.

Answer F is correct because List keys are intended to uniquely identify entries in that list, so duplicate key values would violate the modeled identity.

Incorrect Answers

Answer A is incorrect because Keying does not determine whether modeled nodes are configuration or operational state.

Answer B is incorrect because A key identifies entries but does not impose the list’s cardinality unless separate schema constraints do so.

Answer C is incorrect because HTTP method selection depends on the intended RESTCONF operation, not merely on the presence of a YANG key.

Answer D is incorrect because A list key is a schema identity mechanism and provides no encryption by itself.

 

Question 5

A monitoring system needs the configured interface description and the live packet counter for the same interface. Which TWO observations correctly separate configuration from operational state? Choose TWO.

1. Both values must be configuration because they appear in the same YANG module.
2. Both values must be operational because a controller retrieves them with GET.
3. The description becomes operational-only once a device has rebooted.
4. The configured description represents intended configuration data.
5. The live packet counter represents operational state produced by current device activity.
6. The packet counter becomes configuration when the interface is administratively enabled.

Correct Answers: D, E

 

Correct Answers

Answer D is correct because A configured description is persistent intended state that an operator or automation system sets on the device.

Answer E is correct because Packet counters are observed runtime state rather than desired configuration values.

Incorrect Answers

Answer A is incorrect because A single YANG model can include both configuration and operational state nodes, so module membership alone does not decide this.

Answer B is incorrect because Retrieval method does not convert configured state into operational-only data; GET can read modeled configuration as well.

Answer C is incorrect because A persistent configured description remains configuration state after reboot when the configuration is retained.

Answer F is incorrect because Administrative state does not turn a runtime traffic counter into desired configuration data.

 

Question 6

An automation pipeline rejects a candidate payload before sending it because a required list key is missing. What benefit of using the YANG schema is being applied?

1. The client is proving the requested network design is operationally safe.
2. The client is bypassing device authorization because the schema approved the payload.
3. The client is discovering a new transport protocol automatically.
4. Validate the payload structure against the YANG model before sending it.
5. The client is confirming that every platform implements the same vendor-native module.

Correct Answer: D

 

Correct Answer

Answer D is correct because Schema validation can detect required structural elements and type constraints before a request reaches the device.

Incorrect Answers

Answer A is incorrect because A schema can validate representation and constraints but cannot prove that a technically valid change is appropriate for the live network.

Answer B is incorrect because A structurally valid payload still requires authentication and authorization on the target service.

Answer C is incorrect because Schema validation does not discover or replace the protocol used to reach the device.

Answer E is incorrect because Local schema validation does not establish that every target device supports identical modules or revisions.

 

Question 7

A YANG leaf for a percentage permits values from 0 through 100. An application tries to submit 125. Which statement best describes the schema’s role?

1. The modeled constraint can reject 125 as outside the allowed value range.
2. The schema must accept 125 because JSON numbers have no upper bound.
3. The schema encrypts 125 so the device can decide whether it is safe.
4. The schema automatically converts 125 to 100 before configuration.
5. The schema changes the request into an asynchronous task because the value is invalid.

Correct Answer: A

 

Correct Answer

Answer A is correct because YANG type and range constraints can restrict values that are structurally acceptable for a modeled leaf.

Incorrect Answers

Answer B is incorrect because JSON representation does not override a narrower semantic range defined by the YANG schema.

Answer C is incorrect because YANG constraints describe data and do not encrypt request values.

Answer D is incorrect because Validation constraints generally reject invalid values rather than silently changing the operator’s requested value.

Answer E is incorrect because Invalid modeled input is a validation issue and does not inherently create a controller task.

 

Question 8

A multi-vendor automation team has an industry-defined interface model and a Cisco-native model available for the same device family. Which assessment is most accurate?

1. The industry-defined model is guaranteed to expose every Cisco-specific feature on every release.
2. The two models are interchangeable if their leaf names look similar.
3. Shared models improve portability; native models can expose vendor-specific capabilities.
4. The Cisco-native model cannot be used by NETCONF because only standards-defined models are transportable.
5. Selecting a model determines whether SSH or HTTPS is encrypted.

Correct Answer: C

 

Correct Answer

Answer C is correct because Shared models can reduce vendor coupling, whereas native models can represent features that are specific to a vendor or platform.

Incorrect Answers

Answer A is incorrect because A common model cannot be assumed to cover every vendor extension or release-specific capability.

Answer B is incorrect because Similar names do not prove equivalent namespaces, semantics, constraints, revisions, or platform support.

Answer D is incorrect because NETCONF and RESTCONF can expose vendor-native YANG models when the device advertises and supports them.

Answer E is incorrect because Model choice and transport security are separate concerns controlled by the interface and session configuration.

 

Question 9

An application was tested against one IOS XE release. Before using a YANG path on a newly upgraded switch, what should the application verify first?

1. The application can ping the management address.
2. The device advertises the required YANG module and compatible revision or capability.
3. The path worked on the earlier release, which proves it remains supported.
4. The REST status code for a different endpoint is 200.
5. The switch has the same hostname as the lab device.

Correct Answer: B

 

Correct Answer

Answer B is correct because Capability discovery verifies what models the actual device and release expose instead of assuming prior support remains unchanged.

Incorrect Answers

Answer A is incorrect because IP reachability is necessary for access but does not prove the required model is advertised.

Answer C is incorrect because Model support and revisions can change, so historical success is not sufficient evidence for the upgraded target.

Answer D is incorrect because Success on an unrelated resource does not establish capability for the intended YANG module and path.

Answer E is incorrect because Hostname equality has no direct relationship to support for a specific YANG module or revision.

 

Question 10

A payload passes YANG validation and would set an interface MTU to 9000. The adjacent device supports only 1500 bytes and the change would break the link’s intended traffic. What conclusion should the automation system make?

1500. The model must reject 9000 on every platform because Ethernet MTU is always 1500.
1501. The controller should ignore the peer because operational checks occur only after deployment.
1502. A valid YANG value cannot cause a forwarding problem.
1503. Schema validity does not prove the change is operationally appropriate for the topology.
1504. YANG validation guarantees the peer will adopt the same MTU automatically.

Correct Answer: D

 

Correct Answer

Answer D is correct because The value can satisfy the model while still violating an external network requirement such as peer compatibility.

Incorrect Answers

Answer A is incorrect because Platforms can support larger MTUs, and the relevant issue here is the stated peer requirement rather than a universal fixed value.

Answer B is incorrect because Prechange policy and topology validation can be necessary even when the payload itself is schema-valid.

Answer C is incorrect because A syntactically and semantically valid local setting can still be wrong for the broader network design.

Answer E is incorrect because The schema validates the local modeled request and does not coordinate configuration on an unrelated neighbor.

 

Question 11

A script needs the serial number, management IP address, and platform identifier for devices already known to Catalyst Center. Which API category should it query first?

1. Task status resources only.
2. Assurance health resources only.
3. Issue-remediation resources only.
4. Device inventory/network-device resources.
5. Template deployment resources only.

Correct Answer: D

 

Correct Answer

Answer D is correct because Inventory APIs are designed to return attributes about managed devices such as identity and management addressing.

Incorrect Answers

Answer A is incorrect because Task APIs report asynchronous operation state and do not replace device inventory resources.

Answer B is incorrect because Health APIs focus on observed condition and performance rather than being the primary inventory source for device identity.

Answer C is incorrect because Issue resources describe detected problems and recommendations rather than serving as the authoritative device inventory query.

Answer E is incorrect because Template APIs manage configuration intent and deployment workflows rather than simply retrieving current inventory attributes.

 

Question 12

An operations dashboard already knows a switch’s inventory identity and now needs its current network-health condition. Which API family best matches the new requirement?

1. Task-count retrieval only.
2. Authentication-token generation only.
3. Inventory creation only.
4. Configuration template creation only.
5. Catalyst Center health or assurance APIs.

Correct Answer: E

 

Correct Answer

Answer E is correct because Health and assurance resources expose observed network, site, device, and client condition rather than static identity alone.

Incorrect Answers

Answer A is incorrect because Task counts describe asynchronous platform work and do not represent the switch’s network-health condition.

Answer B is incorrect because Authentication is required for access but does not itself return the requested network-health measurements.

Answer C is incorrect because Inventory can identify the device but does not substitute for health-oriented assurance data.

Answer D is incorrect because Templates describe intended configuration actions rather than the current observed health state.

 

Question 13

A Catalyst Center POST request to deploy configuration returns a task identifier immediately. The script receives no evidence yet that the device changed. What should the script infer?

1. The request initiated controller work, but completion must be verified through the returned task workflow.
2. The device configuration is complete because every successful POST is synchronous.
3. The client should resubmit the same change repeatedly until no task ID is returned.
4. The request failed because a successful operation never returns a task identifier.
5. The task identifier is the new device serial number.

Correct Answer: A

 

Correct Answer

Answer A is correct because An asynchronous task identifier indicates that the operation has been accepted for processing, not that downstream device work has already succeeded.

Incorrect Answers

Answer B is incorrect because Controller operations can be asynchronous, so request acceptance alone does not prove device-side completion.

Answer C is incorrect because Blind resubmission can duplicate work; the appropriate action is to follow the task status and result details.

Answer D is incorrect because Catalyst Center uses asynchronous tasks for operations, making a task identifier expected for supported workflows.

Answer E is incorrect because Task IDs identify platform work and should not be confused with managed-device inventory identifiers.

 

Question 14

A Catalyst Center operation returns task ID `abc123` and status `PENDING`. Which TWO actions are appropriate for a robust client? Choose TWO.

1. Inspect task detail or result-location information when the task completes or fails.
2. Treat `PENDING` as proof that the network change succeeded.
3. Generate a new authentication token for every poll because task IDs expire immediately.
4. Delete the managed device from inventory before reading the task.
5. Poll or query the task by its identifier until it reaches a terminal state.
6. Assume an HTTP 200 from the task query means the underlying deployment succeeded.

Correct Answers: A, E

 

Correct Answers

Answer A is correct because Task metadata can point to additional result information that explains the outcome of the asynchronous operation.

Answer E is correct because The task resource exists specifically to expose progress and terminal SUCCESS or FAILURE state for asynchronous operations.

Incorrect Answers

Answer B is incorrect because Pending means the task has not yet reached a terminal success or failure state.

Answer C is incorrect because Authentication renewal follows token lifetime and policy; the task identifier itself does not require a new token for each query.

Answer D is incorrect because Inventory deletion is unrelated to determining the status of the submitted asynchronous operation.

Answer F is incorrect because A successful retrieval can return a task whose own status is PENDING or FAILURE, so payload state still must be interpreted.

 

Question 15

A workflow must discover devices, confirm current health, deploy a configuration intent, and prove the deployment completed. Which THREE API stages are necessary? Choose THREE.

1. Submit the configuration operation and then follow its task status to a terminal result.
2. Query device inventory to identify the managed targets.
3. Use only a ping to the Catalyst Center appliance because API state is unnecessary.
4. Replace the controller API with a YANG list key because keys perform deployment.
5. Interpret a generated token as proof that the target configuration was applied.
6. Query health or assurance data to capture the stated prechange condition.

Correct Answers: A, B, F

 

Correct Answers

Answer A is correct because Deployment intent and asynchronous task verification together establish whether the requested controller work completed.

Answer B is correct because Inventory establishes the target devices and identifiers used by later workflow steps.

Answer F is correct because Health resources provide the current observed condition required by the workflow before deployment.

Incorrect Answers

Answer C is incorrect because Reachability to the controller does not provide device inventory, health evidence, or deployment completion state.

Answer D is incorrect because YANG keys identify modeled list entries and do not substitute for controller workflow operations.

Answer E is incorrect because Authentication success grants API access but says nothing about the outcome of a later device change.

 

Question 16

A central script must obtain the current list of SD-WAN edge devices known to Cisco Catalyst SD-WAN Manager. Where should the script direct the management query?

1. The local EEM history on each edge because centralized APIs cannot monitor devices.
2. The OMP data plane on every branch because inventory exists only in user traffic.
3. The underlay routing table on one transport provider because it contains all overlay identities.
4. The SD-WAN Manager REST API that exposes centralized device information.
5. The branch DHCP server because SD-WAN Manager cannot return device data.

Correct Answer: D

 

Correct Answer

Answer D is correct because SD-WAN Manager provides centralized REST operations for device monitoring, inventory, provisioning, and management.

Incorrect Answers

Answer A is incorrect because EEM history is local event information and does not replace the management-plane API’s inventory capability.

Answer B is incorrect because OMP is a control-plane protocol and edge user traffic is separate from centralized management API inventory.

Answer C is incorrect because Underlay routing does not provide authoritative centralized SD-WAN Manager inventory for all managed edges.

Answer E is incorrect because Device inventory and management information are explicit SD-WAN Manager API use cases.

 

Question 17

An SD-WAN integration needs to read tunnel-health statistics without changing policy. Which API intent best fits that requirement?

1. Install a Puppet agent on the SD-WAN Manager solely to make a REST GET possible.
2. Use a monitoring or visibility GET operation for the relevant statistics.
3. Submit a policy-edit operation because every API read must alter controller intent.
4. Send user traffic through the Manager’s API port to measure tunnel health.
5. Delete and recreate the edge device so the controller refreshes statistics.

Correct Answer: B

 

Correct Answer

Answer B is correct because Monitoring endpoints are intended to retrieve observed state without submitting a policy or configuration change.

Incorrect Answers

Answer A is incorrect because The Manager already exposes REST APIs and does not require an unrelated agent for a basic API query.

Answer C is incorrect because Reading monitoring data does not require changing centralized policy.

Answer D is incorrect because Management API traffic is separate from the overlay’s user data forwarding path.

Answer E is incorrect because Destructive inventory operations are unnecessary and inappropriate for a read-only monitoring requirement.

 

Question 18

A developer wants to automate SD-WAN configuration centrally without logging into every edge CLI. Which component should be treated as the management-plane API endpoint?

1. A Catalyst Center task URL from an unrelated controller.
2. A random transport router owned by the ISP.
3. The OMP route advertisement itself.
4. Cisco Catalyst SD-WAN Manager.
5. The edge device’s application data interface only.

Correct Answer: D

 

Correct Answer

Answer D is correct because SD-WAN Manager exposes the centralized REST API used for controlling, configuring, monitoring, and integrating the overlay.

Incorrect Answers

Answer A is incorrect because Catalyst Center and SD-WAN Manager have separate APIs; a task URL from one controller does not become the other’s endpoint.

Answer B is incorrect because The transport provider’s router is part of the underlay and is not the customer’s centralized SD-WAN management API endpoint.

Answer C is incorrect because OMP distributes overlay control information and is not an HTTP REST endpoint for external automation clients.

Answer E is incorrect because User-facing data interfaces forward traffic and are not the centralized management-plane API service.

 

Question 19

An SD-WAN Manager API operation returns an operation identifier and indicates that work is still in progress. What should the client do before declaring success?

1. Query the documented operation or task state until the controller reports a terminal result.
2. Resubmit the same change continuously so at least one copy succeeds.
3. Check only whether user traffic currently passes through one edge.
4. Assume success because the initial HTTP exchange completed.
5. Change the API request into a NETCONF `<get>` because task state cannot exist in REST APIs.

Correct Answer: A

 

Correct Answer

Answer A is correct because An in-progress response explicitly means the requested work has not yet reached a confirmed terminal outcome.

Incorrect Answers

Answer B is incorrect because Repeated submissions can create duplicate or conflicting work and do not replace following the original operation state.

Answer C is incorrect because Existing forwarding can continue independently and does not prove a new management operation completed.

Answer D is incorrect because A completed HTTP exchange can carry an in-progress application state and therefore does not prove the requested controller operation succeeded.

Answer E is incorrect because REST management platforms can expose asynchronous task or operation state; switching protocols is not required by the presence of a task identifier.

 

Question 20

A branch continues forwarding existing application traffic while the administrator’s SD-WAN Manager API session expires. Which explanation is most accurate?

1. An expired API session proves the SD-WAN control plane is down.
2. All edge data must traverse the administrator’s REST session, so forwarding should have stopped.
3. The management API session and edge data forwarding are separate functions, so losing the API session does not by itself stop established forwarding.
4. The branch must be unmanaged because managed edges cannot forward without a live API client.
5. The API cookie is the packet-encryption key for every branch tunnel.

Correct Answer: C

 

Correct Answer

Answer C is correct because The REST session serves centralized management while edge devices perform data-plane forwarding using their current operational state.

Incorrect Answers

Answer A is incorrect because The administrator’s management-session expiry does not establish controller or overlay control-plane failure.

Answer B is incorrect because User traffic does not flow through the external management client’s REST session.

Answer D is incorrect because Managed edges do not require an external automation client to remain continuously connected for every forwarded packet.

Answer E is incorrect because API authentication state is distinct from the keys and protocols used by SD-WAN data-plane tunnels.

 

Question 21

A Catalyst Center GET request for tasks returns HTTP 200 and a JSON response containing the requested task records. What does the 200 indicate?

1. The client is authorized to perform all other API operations.
2. The controller has finished every asynchronous network operation.
3. Every returned task necessarily has status SUCCESS.
4. The request created a new task because 200 always means creation.
5. The HTTP request was successfully processed and the response payload can now be interpreted for task state.

Correct Answer: E

 

Correct Answer

Answer E is correct because The documented endpoint uses 200 for a successful request, while the task objects inside the payload carry their own operational status.

Incorrect Answers

Answer A is incorrect because Successful access to one endpoint does not grant unlimited authorization across unrelated resources.

Answer B is incorrect because A successful read can return PENDING, FAILURE, or SUCCESS task records.

Answer C is incorrect because HTTP success for reading tasks does not force each task’s application status to be successful.

Answer D is incorrect because This is a GET retrieval, and 200 indicates successful retrieval rather than creation of a new task.

 

Question 22

A controller API responds that a submitted operation has been accepted for asynchronous processing and provides a task identifier, but the task is not finished. Which TWO conclusions are correct? Choose TWO.

1. The controller must return the final device configuration in the same response.
2. The task identifier should be used as an authentication token.
3. The response proves every target device was reachable during execution.
4. Acceptance is not equivalent to successful completion of the underlying network change.
5. The client should report the device change as completed immediately.
6. The client should retain the task identifier for later status lookup.

Correct Answers: D, F

 

Correct Answers

Answer D is correct because An accepted asynchronous request can later succeed or fail, so terminal task state still matters.

Answer F is correct because The task identifier is the correlation value needed to query the asynchronous operation after submission.

Incorrect Answers

Answer A is incorrect because Asynchronous workflows intentionally separate initial submission from later result retrieval.

Answer B is incorrect because Task IDs correlate work and are not credentials that authorize later API access.

Answer C is incorrect because Acceptance occurs before all downstream execution evidence is necessarily available.

Answer E is incorrect because Doing so would confuse request acceptance with the later outcome of asynchronous work.

 

Question 23

A Catalyst Center request receives HTTP 400 with an error indicating that a required query value has invalid syntax. What is the best interpretation?

1. The server timed out while waiting for an upstream device.
2. The controller accepted the operation and created an asynchronous task.
3. The resource does not exist on the controller.
4. The client should correct the malformed request rather than treat the response as a controller outage.
5. The client’s credentials were valid but lacked permission for the resource.

Correct Answer: D

 

Correct Answer

Answer D is correct because The documented 400 response indicates a request the server could not understand, such as invalid syntax.

Incorrect Answers

Answer A is incorrect because The stated 400 response identifies a client-request problem rather than a gateway or service timeout.

Answer B is incorrect because A malformed request is not evidence that the requested work was accepted for execution.

Answer C is incorrect because A missing resource is distinct from a request whose syntax or parameters are invalid.

Answer E is incorrect because A forbidden response aligns with an authorization problem rather than the stated malformed-request error.

 

Question 24

Two API calls fail differently. Call A returns 401 because its credentials are missing. Call B returns 403 after the controller recognizes the credentials but denies the requested operation. Which TWO interpretations are correct? Choose TWO.

403. Call A is an authorization success because 401 is lower than 403.
404. Call A proves the resource exists while Call B proves it does not.
405. Call B should always be fixed by sending the same credentials again unchanged.
406. Both responses mean the controller is unavailable.
407. Call A must establish valid authentication before the protected resource can be accessed.
408. Call B is authenticated but lacks authorization for that requested operation or resource.

Correct Answers: E, F

 

Correct Answers

Answer E is correct because The documented 401 condition indicates missing or invalid client authentication credentials.

Answer F is correct because The documented 403 condition means the server recognizes the credentials but refuses the action due to authorization.

Incorrect Answers

Answer A is incorrect because HTTP status numbers are categorical response codes, not a numeric scale of permission success.

Answer B is incorrect because Authentication and authorization failures do not establish whether the target resource exists.

Answer C is incorrect because Recognized credentials can remain insufficient; the relevant permission or role must be addressed.

Answer D is incorrect because Both are access-control responses from a reachable controller rather than evidence of service unavailability.

 

Question 25

A GET request for `/tasks/does-not-exist` reaches Catalyst Center and returns 404 for that identifier. Other API requests to the controller succeed. What should the client conclude first?

1. The credentials are definitely invalid.
2. The task must have status FAILURE.
3. The controller is powered off because every 404 means the server is unreachable.
4. The client should change the API to UDP because HTTP cannot address task resources.
5. The specific requested resource or identifier was not found; this does not by itself indicate a controller outage.

Correct Answer: E

 

Correct Answer

Answer E is correct because The documented 404 condition concerns a resource that does not exist, while successful calls prove the controller remains reachable.

Incorrect Answers

Answer A is incorrect because Invalid or missing credentials are handled separately from a resource-not-found condition.

Answer B is incorrect because A nonexistent task resource cannot be equated with an existing task whose application status is FAILURE.

Answer C is incorrect because A server must respond to produce this HTTP status, and other successful requests further contradict an outage.

Answer D is incorrect because The issue is the requested identifier, not the use of HTTP as the controller API protocol.

 

Question 26

A deployment response contains `{“taskId”:”7f9″}`. The API documentation states that deployments are asynchronous. What should the client do with `7f9`?

1. Use it to query the corresponding task or operation until a terminal outcome is available.
2. Use it as the next API authentication token.
3. Convert it to a YANG list key for every device resource.
4. Discard it because asynchronous workflows cannot be queried after submission.
5. Treat it as the serial number of the first deployed switch.

Correct Answer: A

 

Correct Answer

Answer A is correct because The identifier correlates the submitted asynchronous work with later status and result retrieval.

Incorrect Answers

Answer B is incorrect because Authentication tokens and task identifiers have different purposes and security semantics.

Answer C is incorrect because A controller task ID does not automatically become a key in unrelated YANG resource models.

Answer D is incorrect because The identifier is valuable precisely because it supports subsequent task-state lookup.

Answer E is incorrect because A task identifier represents controller work and is not a managed-device hardware identifier.

 

Question 27

A filtered inventory GET returns HTTP 200 and `{“response”:[]}`. The filter is valid and no devices currently match it. What is the best interpretation?

403. The client is unauthorized because an empty array is equivalent to 403.
404. The controller necessarily failed because every successful GET must contain at least one object.
405. The resource path is malformed because an empty result is equivalent to 400.
406. The first matching device was deleted during transport.
407. The request succeeded and produced an empty result set for the supplied filter.

Correct Answer: E

 

Correct Answer

Answer E is correct because A successful response with an empty array can legitimately mean that zero resources match the request.

Incorrect Answers

Answer A is incorrect because Authorization failures use their own status and should not be inferred merely from an empty successful collection.

Answer B is incorrect because An empty collection is a valid successful result when no entries satisfy the filter.

Answer C is incorrect because The stated 200 status and valid filter indicate the request was understood and completed.

Answer D is incorrect because The payload supplies no evidence of a deletion; it simply reports no matching resources at query time.

 

Question 28

A RESTCONF PATCH fails and the returned error identifies the modeled path `/interfaces/interface[name=’Gi1′]/mtu` as violating a schema constraint. Which response detail is most useful for correction?

1. The HTTP server banner by itself.
2. The modeled error path and constraint information identifying the failing `mtu` node.
3. The controller’s unrelated device-inventory count.
4. The last successful OSPF neighbor state.
5. The display order of options in the automation tool.

Correct Answer: B

 

Correct Answer

Answer B is correct because The path directs the client to the exact modeled node whose value or operation violates the server’s YANG-based validation.

Incorrect Answers

Answer A is incorrect because Server software identity does not identify which requested modeled value violated the schema constraint.

Answer C is incorrect because Inventory totals do not explain why the specific RESTCONF modeled path failed validation.

Answer D is incorrect because Routing state is unrelated to the stated model-validation error for the requested MTU node.

Answer E is incorrect because Presentation order has no bearing on RESTCONF validation of a YANG resource path.

 

Question 29

A controller returns HTTP 409 and states that the target resource is already being edited by another workflow. What should the client infer?

1. The client’s credentials are missing and must be regenerated.
2. The controller is offline and cannot return application errors.
3. The client should ignore the conflict and report successful completion.
4. The requested resource can never exist on this controller.
5. The request conflicts with the resource’s current state; retry or reconciliation may be appropriate after the competing edit is resolved.

Correct Answer: E

 

Correct Answer

Answer E is correct because The documented conflict response describes a resource state that prevents the requested operation, such as an edit conflict.

Incorrect Answers

Answer A is incorrect because Missing or invalid authentication is represented separately from a resource-state conflict.

Answer B is incorrect because Receiving a structured conflict response shows that the controller is reachable and processing the request.

Answer C is incorrect because A conflict means the requested operation has not been established as successful and requires resolution or retry logic.

Answer D is incorrect because A conflict often implies the resource does exist but is currently in a state that blocks the operation.

 

Question 30

An API returns 100 devices per page. The first response contains 100 devices and metadata showing 240 total matches with the next offset at 101. Which TWO actions are necessary to build a complete result? Choose TWO.

1. Request the subsequent pages using the documented offset or pagination mechanism.
2. Stop after the first page because later offsets are only for write operations.
3. Treat the total value of 240 as a task identifier.
4. Change the HTTP method to DELETE to reveal the remaining resources.
5. Combine results while preventing duplicate processing if a page is retried.
6. Assume the first 100 devices are the complete inventory because HTTP 200 means complete pagination.

Correct Answers: A, E

 

Correct Answers

Answer A is correct because The first page covers only part of the 240 matching resources, so additional pages are required.

Answer E is correct because A robust client should aggregate the paginated result and handle retries without counting the same resource twice.

Incorrect Answers

Answer B is incorrect because Offsets and page limits are used specifically to retrieve subsequent portions of large result sets.

Answer C is incorrect because The total is result metadata describing match count, not asynchronous work correlation.

Answer D is incorrect because Deletion is unrelated to navigating a paginated read and would be destructive.

Answer F is incorrect because HTTP success applies to the individual page and does not override pagination metadata showing more matches.