Cisco DevNet Course in Pune — Network Automation & Programmability Training

Python is the language of network automation — not because it is the only option, but because the networking community has standardised on it to a remarkable degree. Every major networking automation library (Netmiko, NAPALM, Nornir), every network vendor's SDK (Cisco Cobra, Cisco Meraki Python SDK), and every infrastructure automation framework (Ansible, Terraform providers) uses Python as its primary interface. Learning Python in the context of networking automation is fundamentally different from learning Python for web development or data science — the data structures you encounter (nested dictionaries representing device configurations, JSON from REST API responses), the error handling patterns you need (what to do when an SSH connection to a device times out), and the way you structure programs (iterating over inventories of devices rather than processing rows in a dataset) are all specific to the networking use case.

The module starts with Python fundamentals through a networking lens: variables, strings, lists, and dictionaries with examples taken from network configuration data rather than abstract exercises. File I/O for reading device inventories from CSV and JSON files. Regular expressions for parsing CLI output — extracting interface status, IP addresses, and routing table entries from the text output of show commands. Error handling with try-except for the connection failures, authentication errors, and timeout conditions that real network automation encounters regularly. Functions for building reusable automation components. The Netmiko library — the most widely used Python library for SSH-based Cisco device automation — is introduced with connection setup, command execution, configuration sending, and output capture. Students finish this module having built their first real automation tool: a Python script that connects to a list of routers, collects interface status from each, and generates a formatted report.

REST APIs have transformed how network devices and controllers are managed. Rather than SSH-ing into a device and scraping text output from CLI commands — a brittle approach that breaks every time Cisco changes a show command's output format — REST APIs provide structured, machine-readable data in JSON or XML format via standard HTTP requests. Every modern Cisco platform exposes REST APIs: IOS-XE devices, DNA Center, vManage, Meraki Dashboard, ACI APIC, ISE — all of them. Understanding REST API concepts is therefore the most foundational automation skill a modern network engineer can have.

REST API concepts are covered from first principles: HTTP methods (GET for retrieval, POST for creation, PUT/PATCH for modification, DELETE for removal), HTTP status codes (200 OK, 201 Created, 400 Bad Request, 401 Unauthorized, 404 Not Found, 500 Server Error), request headers (Content-Type: application/json, Accept, Authorization), and the URL structure of REST API endpoints. JSON — the data format returned by virtually every network REST API — is covered with Python's json library for parsing and constructing JSON payloads. The Requests library in Python makes HTTP API calls straightforward: making authenticated GET requests to retrieve device configuration, parsing the JSON response into Python dictionaries, extracting specific values, and making POST requests with JSON payloads to modify configuration. API authentication methods — basic auth, token-based auth, and OAuth — are covered with practical examples from Cisco APIs. Postman is used for interactive API exploration before moving to Python automation.

NETCONF and RESTCONF are the protocols that represent the networking industry's move away from ad-hoc CLI scraping towards structured, model-driven configuration management. They are significant exam topics for both DevNet Associate and DevNet Professional, and they are increasingly the right answer for any serious network automation project involving Cisco IOS-XE devices — because they provide atomic configuration transactions (the entire change succeeds or the entire change rolls back), candidate configuration (you can prepare a configuration change and validate it before committing), and structured data that does not change format when Cisco releases a new IOS-XE version.

YANG (Yet Another Next Generation) data models are the schema that define what NETCONF and RESTCONF can configure on a device. Every configurable element — an interface, a routing protocol instance, an ACL entry — has a corresponding YANG node with defined types, constraints, and relationships. Understanding the YANG tree structure, how to navigate it using tools like pyang and the YANG Catalog, and how to construct the correct XPath queries or JSON keys to address specific configuration elements is the core skill of model-driven automation. NETCONF uses XML over SSH to retrieve and modify device configuration using the YANG tree structure. RESTCONF provides the same YANG-modelled configuration access over HTTP with JSON payloads — making it more accessible for Python automation. Both protocols support the running datastore (current active config), the candidate datastore (pending changes), and configuration validation before commit.

Ansible has become the dominant tool for network configuration management at organisations that need to manage configuration across hundreds of devices consistently and reproducibly. Unlike Python scripts (which require programming knowledge to write and maintain), Ansible playbooks are structured YAML files that express the desired state of network configurations in a format that is readable by both engineers and non-engineers. The ability to take a configuration requirement from a security team — "all access switches must have DHCP snooping enabled on all access VLANs" — and implement it as an Ansible playbook that applies the change across 200 switches in 10 minutes is one of the most immediately valuable skills this module produces.

Ansible inventory for network devices is covered with both static inventory files (defining device groups by function — core switches, distribution switches, branch routers) and dynamic inventory from CMDB or network management systems. Ansible network connection types are covered with the important distinction between network_cli (SSH-based, runs show commands and sends config commands like a human would), NETCONF (model-driven, uses ncclient over NETCONF), and HTTPAPI (REST API-based, used for DNA Center, vManage, and other platforms). Cisco IOS modules — ios_command, ios_config, ios_facts, ios_interfaces, ios_vlans — are the primary tools for IOS device automation and are practised extensively. The idempotency principle — running the same playbook twice should produce the same result without errors — is demonstrated and its importance for safe network automation explained. Error handling in playbooks, using handlers for conditional tasks (restart a service only if its configuration changed), and roles for organising complex automation code into reusable components are covered for the DevNet Professional exam domain.

The real power of network automation comes when you combine Python and REST API skills with platform-specific APIs that expose higher-level abstractions. Rather than configuring individual switches, the DNA Center API lets you express intent — "onboard this new site with these VLANs" — and the platform handles the underlying device configuration. Rather than SSH-ing into 500 vEdge routers one by one, the vManage REST API lets you push template changes to all 500 simultaneously from a Python script. This module covers three of the most important Cisco platform APIs in depth.

Cisco DNA Center API is one of the most comprehensive network management APIs available. It provides endpoints for device inventory (discover and list all managed network devices), topology visualisation (retrieve the current network topology as a data structure), command runner (execute show commands on devices programmatically and retrieve structured output), configuration templates (push template-based configurations to devices), network health (retrieve real-time health scores for devices, clients, and sites), and intent API (high-level operations like adding a new site or deploying a VLAN to a set of switches). Cisco SD-WAN vManage REST API covers device inventory, template management, monitoring endpoints (retrieving BFD session data and OMP routes programmatically), and policy management. Meraki Dashboard API — Cisco's cloud-managed networking platform — provides RESTful APIs for every configuration and monitoring operation: retrieving organisation inventory, configuring VLANs and firewall rules, monitoring client connections, and generating network analytics reports.

The most mature network automation practices in leading Indian IT organisations treat network configuration as code — stored in version control, tested automatically, and deployed through pipelines that require review and approval before any change reaches production. This approach, borrowed from software engineering and applied to infrastructure, is what the DevNet Professional certification targets and what the most progressive networking teams are building towards. This module introduces the tools and practices that make it possible.

Terraform for network infrastructure provides an infrastructure-as-code approach to network configuration: writing HCL (HashiCorp Configuration Language) resource definitions that declare the desired state of network objects (Meraki MX firewall rules, ACI tenant policies, SD-WAN templates), running terraform plan to see what changes will be made before applying them, and running terraform apply to make those changes atomically. Terraform state management — how Terraform tracks what it has created and compares current state to desired state — is covered with the implications for how to safely manage shared infrastructure. Git workflows for network automation code — branches, pull requests, code review for network changes — are practised with the cultural change they represent: treating a network configuration change the same way a development team treats a code change, with peer review and automated testing before deployment. GitHub Actions CI/CD pipelines for network automation — automatically running Ansible playbooks or Python scripts when a pull request is merged to the main branch — complete the picture of a mature network automation practice.

The final module has two components that run in parallel during the last two weeks: exam preparation for the DevNet Associate (DEVASC 200-901) certification, and a capstone automation project that demonstrates the complete automation workflow from requirements to production deployment.

DevNet Associate exam preparation covers all six exam domains: software development and design (Python fundamentals, data structures, design patterns), understanding and using APIs (REST, RESTCONF, NETCONF, YANG), Cisco platforms and development (DNA Center, Meraki, ACI, SD-WAN API coverage), application deployment and security (containers with Docker, basic CI/CD concepts, API security), infrastructure and automation (Ansible, Terraform, Git), and network fundamentals (covering the networking knowledge the exam assumes). Domain-by-domain practice question analysis identifies weak areas for targeted revision. The capstone project takes a realistic automation requirement — for example, a daily compliance audit that checks 50 switches for DHCP snooping configuration, reports non-compliant devices, and automatically applies the remediation — and builds it as a complete, production-quality automation solution with Python, Ansible or Terraform for the core logic, Git for version control, documentation, and a simple CI/CD pipeline for the deployment process. The capstone is portfolio-ready and demonstrates exactly the skills that DevNet and network automation employers look for.