From typing a URL to seeing the page

Knowledge Videos

Short animated explainers, organised by subject.

From typing a URL to seeing the page

DNS, TCP and TLS, a static frontend served from a CDN, one call to a Go API, and the final render.

Transcript

You type a U R L. The browser splits it into scheme, host and path, and checks its caches.

No luck, so it asks D N S. The resolver walks from the root, to the dot com servers, to the authoritative server.

The answer is an I P address, and every step along the way caches it.

Next, a T C P connection: the browser sends a SYN, the server answers SYN ACK, and the browser confirms with an ACK.

Then T L S agrees on encryption keys, so everything after that is private.

The request lands on a C D N edge server near you, which answers with the static H T M L straight from its cache.

That page points to C S S and Java Script bundles with hashed names, which can be cached for a year.

The browser builds the D O M, and downloads the C S S and Java Script in parallel.

The first paint shows the app shell, while the data is still on its way.

Now the Java Script calls the A P I: a fetch to the Go backend, over the same secure connection.

Go routes the request to a handler, through middleware, queries the database, and answers with J S O N.

The Java Script turns the J S O N into elements, and the browser lays out and paints the finished page.

D N S, T C P, T L S, a static shell from the edge, one call to a Go A P I, and a render. Usually in well under a second.

More in this series