Home / Workshop / Anatomy of a page request

Anatomy of a page request

What happens when this page loads, measured on your own visit.

Address

At first your device only knows the name stehl.ing. A DNS resolver, usually your internet provider's, translates it into the server's IP address. The answer is cached, so the next visit skips this step.

Lookup time
–
Server IP address
–

Connection

Your device opens a TCP connection to the server. Both sides then agree on a key via TLS; from here on everything is encrypted, which is what the https in the address bar stands for. Only then is the page requested and transferred.

Protocol
–

Server

The request reaches a server at Hetzner. Traefik accepts it, decrypts it and passes it on to a Docker container, where the Caddy web server delivers the page. Along the way the server sees which IP address the request came from.

Server
–
Server response time
–
Your IP address
–
Your network operator
–

Way back

The response travels in packets through a series of routers to you. Traceroute makes these routers visible: it sends packets with a limited lifetime (TTL). Each router subtracts one; when it reaches zero, the router drops the packet and reports back to the sender. With TTL 1, 2, 3 … one router after another answers.

This measures from the server to you, i.e. the way back. The way there may use different routers, as internet routing is often asymmetric.

Your IP address is only used for this measurement and is not stored.

IP geodata by DB-IP (CC BY 4.0), as of –. Places and operators are estimates from this database.