You tap a link, and a page appears. It takes less than a second, and it feels like magic — but it is not magic. It is a chain of about half a dozen systems, each doing one simple job, handing your request along like a relay race. The whole thing was designed in the 1970s to survive a nuclear war, and it still runs on the same basic idea today.
This is the journey of a single click, from your finger to a server on the other side of the planet and back again.
Step 1: DNS — the internet's phone book
Computers do not find each other by name. They find each other by IP address — a string of numbers like 142.250.80.46 that identifies one machine on the network. But humans are terrible at remembering numbers, so we use names like techtoday.space instead.
DNS (the Domain Name System) is the phone book that translates names into numbers. When you click a link, your device asks a DNS resolver — usually run by your internet provider or a service like Google or Cloudflare — "what is the IP address for this name?" We have a full guide to how DNS works if you want the deeper version.
If DNS fails, nothing else can start. That is why a DNS outage can make the whole internet feel broken even though your connection is fine.
Step 2: your request leaves home
Your device now knows the server's address, so it builds a request: "please send me this page." But your device is not connected directly to the internet. Your request first goes to your router, the small box in your home, which forwards it to your internet provider.
From there it enters the provider's network, and the real journey begins. The request does not travel as one piece. It is chopped into packets — small chunks of data, usually around 1,500 bytes each, each stamped with its destination address and a sequence number so the other end can put them back in order.
Splitting data into packets is the internet's central trick. It means a huge file and a tiny message can share the same cables, and if one packet is lost or delayed, only that packet needs to be resent — not the whole thing.
Step 3: routers — the traffic controllers
Packets hop from one router to another, each one reading the destination address and passing the packet to the next router that gets it closer. A typical click crosses 10 to 20 routers, and the path can change from one packet to the next, depending on congestion.
Routers decide where to send packets using a system called BGP (Border Gateway Protocol) — essentially, networks constantly telling each other which destinations they can reach and how quickly. When a cable breaks or a network goes down, BGP routes around it. That self-healing design is why the internet keeps working through failures that would cripple a centrally planned network.
Step 4: under the ocean
If the website you want lives on another continent, your packets probably travel under the sea. More than 99% of intercontinental internet traffic runs through submarine fibre-optic cables lying on the ocean floor — not satellites, which are far too slow and low-capacity for the bulk of traffic. We covered the full story of how undersea internet cables work separately.
Inside those cables, your data travels as pulses of light through strands of glass thinner than a hair. Light in fibre moves at roughly 200,000 kilometres per second — about two-thirds of the speed of light in a vacuum, because glass slows it down. Even so, a signal can cross the Atlantic in about 60 milliseconds.
There are hundreds of these cables, some longer than 10,000 kilometres, and laying or repairing one requires specialised ships. When a cable is damaged — by an earthquake or a ship's anchor — traffic automatically reroutes through other cables, and most users never notice.
Step 5: the server answers
Your packets finally arrive at a data centre, a warehouse full of computers whose only job is to answer requests like yours. The server reads your request, finds the page, and sends it back — again as packets, again hopping across routers and cables.
But there is a shortcut most big sites use: a CDN (Content Delivery Network). A CDN keeps copies of popular pages on servers all over the world, so your request is often answered by a server in your own country — or even your own city — instead of one an ocean away. That is a large part of why pages feel instant: much of the content was already waiting nearby.
Step 6: your browser builds the page
When the packets arrive, your browser reassembles them in order and starts building the page: text, then images, then styling and interactive bits. A single page often triggers dozens of further requests, each repeating the whole journey in miniature. The full round trip typically takes a few hundred milliseconds — and the speed of light sets a hard floor, which is why a site hosted on the far side of the world always feels slightly slower than one hosted nearby.
What each piece does
| Piece | Job | Plain-English version |
|---|---|---|
| DNS | Translates names to numbers | The phone book |
| Router | Forwards packets toward their destination | A traffic controller at every junction |
| Packet | A numbered chunk of your data | An envelope in a relay race |
| BGP | Networks telling each other the fastest routes | The map every router shares |
| Submarine cable | Carries light pulses between continents | The motorway under the sea |
| CDN | Serves copies of content from nearby servers | A local warehouse for popular pages |
| Server | Stores the website and answers requests | The shop that fills your order |
| Browser | Reassembles packets and draws the page | The builder that turns parts into a page |
What can go wrong
When the internet "stops working", one of these steps is usually at fault:
- DNS is down — names cannot be resolved, so nothing loads even though your connection is fine.
- Your Wi-Fi is congested — too many devices, or a neighbour on the same channel, slows the first hop out of your home.
- A cable or router fails — traffic reroutes, but the detour adds delay.
- The server is overloaded — too many requests at once, and the site slows down or refuses connections.
If other sites load fine, the problem is almost certainly the one site.
Why it still works this way
The internet's design is more than 50 years old, and it survives because each piece is simple and replaceable. DNS can be swapped for a faster resolver. Copper can be swapped for fibre. A broken cable can be routed around. No single company owns the network, and no single failure can take it down — which is exactly what its designers wanted.
The next time a page appears in a blink, remember what just happened: a phone-book lookup, a chopped-up request, twenty routers, a beam of light under an ocean, and a warehouse of computers — all in less time than it takes to blink twice.
Explore the Internet & Cloud series
- How Does Streaming Video Actually Work? The Chunks, Codecs and CDNs Behind Play
- Fibre Broadband Explained: Why Light Through Glass Beats Cable
- How Email Actually Works: The 50-Year-Old System the Internet Runs On
- How Undersea Internet Cables Carry 99% of the World's Data
- How DNS Works: The Internet's Phone Book
- What Is Edge Computing and Why It Matters
- How Cloud Computing Powers Modern Apps
Takeaway checklist
- DNS turns a website's name into a numbered address — it is the first step of every click.
- Packets are small numbered chunks; splitting data this way makes the network resilient and shareable.
- Routers pass packets along, rerouting around failures using BGP.
- Submarine fibre cables carry over 99% of intercontinental traffic as pulses of light.
- CDNs keep copies of popular content near you, which is why pages feel instant.
- The whole trip usually takes a few hundred milliseconds — and the speed of light is the ultimate limit.
- When things break, the fault is usually DNS, your Wi-Fi, a congested route, or the server itself.
