The request path
Between typing the address and seeing the content there are stages with different responsibilities.
The path between entering an address and receiving a response: name resolution, connection, edge, origin, application and browser processing.
Delivery flow
From discovery to useful content
-
Discover The name needs to find the network
DNS and routing begin the path before the connection.
-
Deliver Edge and origin serve different roles
The edge receives the connection and forwards the work that is needed.
-
Present The browser still needs to process
HTML, styles, fonts, images and scripts compose the experience.
Name, route and connection come first
The browser resolves the domain, establishes the connection and negotiates transport before requesting content. Network distance, connection reuse and resolver state influence this beginning.
Each stage has its own measurement. Adding everything up as server time hides the difference between a network delay and an application delay.
Edge, origin and application have different jobs
The edge receives the public connection and may answer with an available resource or forward the request. The origin performs work that depends on the application and its data.
The edge may answer in a few milliseconds while the origin takes seconds to build the same page. The measurements are independent and are read in the order of the path.
Receiving bytes does not complete the experience
After the initial response, the browser parses markup, fetches resources and calculates presentation. Resource count, size and order change when content appears and when the page responds to interaction.
Delivery is therefore part of the infrastructure: compression, resource order and image size change the result as much as processing time on the server.