Data models are a programmatic and standards-based way of configuring and collecting operational data of a network device, replacing the process of manual configuration. Using Data models, Cisco IOS XR operating system supports the automating of configurations that belong to multiple routers across the network. Data models are written in a standard, industry-defined language, which can define a new configuration and state an existing configuration on a network.

Traditional CLI-based configurations, are proprietary, cumbersome, and highly text-based. Managing automated operations on a large network using CLIs is a challenge.

Cisco IOS XR supports the YANG data modeling language. YANG can be used with the Network Configuration Protocol (Netconf) or with gRPC (google-defined Remote Procedure Calls) to automate programmable network operations. Data models allow administrators to customize settings easily and automatically, without wasting time on manual configuration.

To get started with using data models, see the Obtain Data Models section in Cisco IOS XR Programmability Configuration Guide for the NCS 5000 Series Router.

Streaming telemetry lets users direct data to a configured receiver for analysis and troubleshooting purposes in order to maintain the health of the network. This is achieved by leveraging the capabilities of machine-to-machine communication.

Traditionally, organizations used the pull model to collect data, where a client pulls data from network elements. This pull model, however, does not scale when there is more than one network management station in the network. These traditional techniques do not cater to all the underlying information of the router, and they require manual intervention.

Tuning a network based on real-time data is crucial for seamless operation of the network. Instead of a pull model, using a push model to continuously stream data out of the network enhances the operational performance and reduces the troubleshooting time. Data can be pushed out at intervals determined by the administrator, at a cadence as low as 10 seconds. Using sophisticated algorithms, a back-end server can then analyze data received from the Cisco IOS XR operating system. The data can be encoded in JavaScript Object Notation (JSON) or Google Protocol Buffers (GPB). This analysis enables back-end management systems to measure and even predict control-plane and data-plane trends.