You type a web address and press Enter

Typing an address triggers nothing on the network yet. The browser first checks whether the host is a search term or a real address, then breaks it into the parts it needs: how to connect, which machine to reach, and which resource to ask for.

Computers route by number, not by name. A resolver is asked for the IP address behind the hostname. If the answer is cached from a recent request, this is nearly instant; otherwise the resolver walks a hierarchy of name servers to find it.

The browser and server exchange messages to agree that both are ready to talk, then negotiate encryption. Certificates are checked so your browser knows it is talking to the real server and not an impostor. This round trip is often the largest single cost of loading a page.

Each packet carries a destination and hops from network to network. Routers only need to know the next sensible step, not the whole path. Physical distance matters: light in fibre takes real time to cross an ocean.

A load balancer picks a healthy server. That server may query a database, call other services, or assemble a page from cached fragments. The work itself might take microseconds; queueing and coordination usually dominate.

The browser builds a document tree, works out how much space everything needs, and paints it. JavaScript can change all of this after the first paint, which is why pages often shift as they load.