How HTTP works

August 23, 2026 · in Cisco U

The application layer protocol underneath the whole web. Client-server: the browser asks, the web server answers, always in request-response pairs. Ports and transport basics are in common ports.

URL anatomy

https://www.cisco.com/c/en/us/index.html

Protocol, https: what language client and server speak.
Server, www.cisco.com: domain name of the host to contact (DNS resolves it).
Path, /c/en/us/index.html: relative location of the file on the server’s filesystem.

Browsers speak more than HTTP: https, file (local browsing), mailto, ftp. HTTPS is HTTP plus an encryption layer, for anything needing confidentiality (e-commerce being the classic case).

Transport underneath

HTTP assumes a reliable transport, so in practice that means TCP on port 80, with HTTPS on TCP 443.

The spec permits UDP, and modern HTTP/3 actually does run over UDP (QUIC). But for CCNA purposes: HTTP means TCP.

Stateless, not connectionless

HTTP is stateless: the server keeps no client info between requests, and each request stands alone.

The course equates stateless with connectionless, but the two are different properties. The transport connection (TCP) is very much connection-oriented; it’s the application protocol that forgets you between requests.

HTTP headers (cookies live in them) can fake statefulness on top, which is how logins and shopping carts work. HTTP is also media independent: it will carry any content type both ends understand, HTML, images, video, whatever.

Request-response cycle

A sequence of request-response pairs makes an HTTP session, opened by the client. Typing the URL above triggers:

Browser sends GET /c/en/us/index.html to www.cisco.com.
Server receives, digs up the file.
Server returns an HTTP response: response code + the HTML document.
Browser parses the HTML, fires additional requests for whatever the page references (images, video, scripts), renders the finished page.

Clicking a link just repeats the cycle: the <a href="/c/en/us/training-events.html"> anchor turns into GET training-events.html, another response, another render.

Who remembers what

On the server side, the web server’s access logs record the visit. On the client side, the browser records the visit in its history and keeps a cached copy of the page.

Cache and history are the traces you leave locally while browsing.