DevNet Professional key terms

A data modeling language defined in RFC 7950 that describes the structure, types, and constraints of configuration and state data exposed by a network device, typically consumed over NETCONF or RESTCONF.

An IETF protocol that uses XML-encoded remote procedure calls over a secure SSH transport to retrieve, edit, and validate configuration data on network devices, organized into datastores such as running and candidate.

An HTTP-based protocol defined in RFC 8040 that maps YANG-modeled data to REST resources, using standard methods such as GET, POST, PUT, PATCH, and DELETE plus media types like application/yang-data+json.

The gRPC Network Management Interface, an OpenConfig-defined service that uses protobuf-encoded RPCs to get, set, subscribe to, and stream telemetry from network device data trees.

A vendor-neutral initiative producing common YANG data models and gNMI specifications so that the same automation code can manage devices from multiple manufacturers with consistent paths and semantics.

A NETCONF configuration store where edits are staged and validated before being committed to the running configuration, enabling atomic changes and rollback if a commit fails.

A NETCONF or device feature that applies a configuration change temporarily and automatically reverts it unless the operator explicitly confirms within a timeout, protecting against lockout from bad changes.

A property of an operation where repeating it produces the same end state as running it once, allowing automation scripts to be safely retried after failures without duplicating configuration.

An automation approach where the operator describes the intended end state of the network and tooling computes and applies the necessary changes, as opposed to issuing step-by-step imperative commands.

The divergence between a device's actual running configuration and the intended or version-controlled baseline, often caused by manual changes, which automation should detect and remediate.

A push-based monitoring method in which a device streams structured, YANG-modeled operational data to a collector at a configured interval or on change, replacing periodic SNMP polling.

Cisco's intent-based network controller that exposes northbound REST APIs for inventory, site design, device provisioning, policy, and assurance, authenticated with token-based access via the API.

A RESTful HTTPS interface for managing Cisco Meraki organizations, networks, and devices, authenticated with an API key header and rate-limited per organization, returning JSON responses.

Cisco Application Centric Infrastructure, a policy-driven data center fabric whose APIC controller offers a REST API and object model for tenants, endpoint groups, contracts, and filters.

The Application Policy Infrastructure Controller that manages an ACI fabric, storing policy as a hierarchical managed object tree accessible through a REST API with JSON or XML payloads.

A Cisco Nexus switch interface offering CLI over HTTP or HTTPS, plus REST and JSON-RPC methods, letting automation send show and configuration commands and receive structured responses.

A Cisco toolset for exploring YANG models, building and testing NETCONF, RESTCONF, and gNMI requests, and generating code, useful for validating model paths before writing automation.

Cisco's SD-WAN management plane component that provides a REST API for configuring templates, policies, and devices across the overlay, with authentication and versioned endpoints.

An HTTP callback registered with a platform so that the platform sends event notifications as POST requests to a listener URL when a specified condition occurs.

A platform control that caps how many API requests a client may make in a time window, requiring automation to handle HTTP 429 responses with backoff and retry logic.

An interface style using HTTP methods on resource URLs with stateless requests, where status codes and representations such as JSON convey results to the calling automation.

A secret token passed in a request header or parameter to identify and authorize an automation client, typically simpler than OAuth but requiring secure storage and rotation.

A time-limited credential issued by an authorization server after client authentication, included as a bearer token in API requests and refreshed before expiry to maintain access.

A retry strategy that increases the wait between attempts, often with random jitter, to avoid overwhelming a service or API that is throttling or temporarily failing.

Code that reattempts a failed operation a bounded number of times for transient errors, distinguishing retryable conditions from permanent failures and respecting idempotency.

An automated test that exercises a small isolated piece of code, such as a parser or payload builder, using mocks or fixtures so it runs quickly without touching real network devices.

An automated sequence that builds, tests, and deploys code on each change, providing consistent validation and repeatable delivery of network automation artifacts.

A version control practice where changes are developed on separate branches and merged through review, enabling parallel work and controlled promotion of automation code.

A request to merge changes from one branch into another, enabling peer review, automated checks, and discussion before code reaches the main branch.

Managing network and compute resources through versioned, machine-readable definition files rather than manual configuration, so environments can be recreated and reviewed consistently.

An agentless automation tool that runs YAML playbooks over SSH or API connections, using modules and inventories to configure devices and enforce desired state.

A YAML file in Ansible that defines an ordered set of plays, tasks, and variables applied to groups of hosts from an inventory.

A structured list of managed devices with connection details and grouping variables that automation tools use to target tasks at the right hosts.

A parameterized text file, often using Jinja2 syntax, that renders device configuration from variables so consistent configurations can be generated for many devices.

The authoritative repository of intended network state, such as a database or version-controlled data file, from which automation generates configuration and detects divergence.

The target configuration and operational condition defined for a device or service, against which automation compares actual state and computes corrective actions.

Restoring a previously known-good configuration or software version after a failed change, often triggered automatically by validation failures or manually by an operator.

Prepared input data or a recorded device response stored with tests so results are deterministic and no live network connection is required during execution.

A stand-in object that simulates a dependency such as an API client, returning canned responses so code can be tested in isolation from real systems.