6Automation & Programmability

• 6.1Network automation: traditional vs. controller-based

  Covers the consistency, speed, and scale that network automation brings to operations; the split between the control plane (which decides where traffic goes) and the data plane (which actually forwards it); the difference between traditional (distributed) networking, where each device decides on its own, and controller-based networking, where a central controller consolidates those decisions; and using AI/ML to spot anomalies from operational data—all as practical judgment.

• 6.2SDN architecture and the fabric

  Covers the physical underlay and the logical overlay, the fabric that binds the two, Cisco's campus automation SD-Access, the northbound API connecting the controller to applications and the east-west API connecting controllers to one another, and the cloud management dashboard Cisco Catalyst Center (formerly DNA Center) that handles automation and assurance.

• 6.3REST APIs and data formats

  Covers, from a practical angle, the nature of REST APIs (stateless, running over HTTP) that underpin operating controllers and devices from automation; the correspondence between CRUD operations and HTTP methods (POST/GET/PUT/DELETE); the HTTP status codes (200/201/4xx/5xx) that indicate success or failure; and reading JSON (the {} object, [] array, keys and values, and data types)—the main vehicle of data exchange.

• 6.4Configuration management tools (Ansible, Terraform, Puppet/Chef)

  Covers the idea of Infrastructure as Code (IaC) managing configuration as code; the meaning of idempotency; Ansible, which is agentless, works over SSH, is push-based, and is written in YAML; Terraform, which builds infrastructure declaratively; and the positioning of the agent-resident, pull-based Puppet/Chef—all as tool-selection judgment.