200-901 — frequently asked questions

Basic sends the Base64-encoded username-password on every request, and Digest sends a hash derived from the credential, so both deliver a credential-based secret to the API service.

Yes. The authorization server authenticates the administrator and issues an access token whose scope represents authorization, so one mechanism covers both.

No. Detecting logical errors is one outcome, but the main purpose is the broader team effort of identifying and reviewing errors, sharing knowledge and enforcing coding standards.

It spreads knowledge among team members, keeps the codebase consistent with coding standards, and surfaces design, security and maintainability concerns before the code is merged.

Standard controller architectures expose northbound and southbound APIs; eastbound and westbound interfaces are not part of the usual controller model, so that description is inaccurate.

The controller's programmatic APIs abstract many devices behind one interface, so a developer expresses intent once instead of configuring each device with traditional protocols.

Edge computing deliberately distributes processing away from a central data centre, so centralisation is the opposite of the architecture's purpose and cannot be one of its benefits.

Spreading compute across many edge nodes places processing near the data sources, which lowers latency and keeps parts of the system working even if one node or a central link fails.

The Meraki dashboard is accessed over HTTPS, which uses TCP port 443, and Meraki devices also use 443 for their secure cloud connection.

8080 is an alternate HTTP port sometimes used by local admin pages; the Meraki dashboard is a cloud service published over standard HTTPS on 443, not 8080.

It decouples the subject from its observers: the subject notifies observers through an abstract interface, so either side can change or be reused independently.

Yes, but that automatic notification is the mechanism; the architectural advantage the pattern provides is the loose coupling between subject and observers.

Cisco Modeling Labs (CML) is an on-premises simulator that builds a topology and runs virtual Cisco and non-Cisco nodes using real Cisco images.

CML is a simulation tool you run locally to build your own topology, while DevNet Sandbox provides remote, pre-built lab environments hosted by Cisco.

Finesse exposes a Phone Books resource for agents and supervisors to read a team's phone books at runtime; it does not provision or update the records, which AXL does.

AXL (Administrative XML) is a SOAP-based provisioning API for Cisco Unified Communications Manager that performs create, read, update and delete operations on configuration objects.

The request contains no device ID or serial-number parameter that would narrow the result, so the DNA Center API returns the entire device inventory with serial numbers.

No. It only issues a read request to list devices; provisioning would require a create call that writes new configuration, which the script does not make.

The developer opened a TCP socket to the remote server with Telnet, which shows the port is reachable, so a simple firewall block on 443 cannot explain the failure.

They show the network path and the transport connection are fine for this client, which shifts the fault to the application layer where the HTTPS exchange is blocked.

Configuration is a management-plane task; the resulting behaviour — classification, marking, queuing and scheduling — is enforced by the data plane on every forwarded packet.

The control plane runs routing and signalling protocols and builds the forwarding table; it decides where traffic goes, not how each packet is treated.

A controller models devices in an abstracted, vendor-neutral way, so it can only present the settings its data model and drivers implement, not every native option.

Yes — many devices expose REST or NETCONF APIs directly. Option C is false because controller platforms are themselves REST-driven, not because devices lack APIs.

Each device exposes its own model, plus shared standards such as IETF and OpenConfig. Model-driven programmability standardizes the API, not one model for the entire network.

The core model-driven protocols are NETCONF and RESTCONF (and gRPC), carrying YANG-modelled data in JSON or XML over SSH, TLS or HTTP.