Names, and the protocol on top

Two application-layer protocols carry almost everything you do.

DNS

Resolving www.example.com with nothing cached walks down a hierarchy:

1. 1Root servers: "I do not know it, but the .com servers are here."
2. 2`.com`: "Nor do I, but example.com's servers are here."
3. 3example.com's own server: "It is 93.184.216.34."

The resolver starts at the root because root addresses are the only ones it knows without asking: they ship as a hardcoded file.

In practice almost every step is cached, which is why this full walk is rare and why a DNS change takes hours to spread. Each record carries a TTL saying how long it may be cached, and that TTL is the delay.

HTTP

A request is a method, a path and headers:

GET /problems HTTP/1.1
Host: leetfoundation.com
Accept: text/html

A response is a status, headers and a body:

HTTP/1.1 200 OK
Content-Type: text/html
Content-Length: 4271

HTTP is stateless. Each request stands alone and the server remembers nothing between them. Cookies exist to work around exactly that: the server sends one, the browser returns it, and the server uses it to look up what it chose to remember.