How the internet works: what happens between pressing Enter and a loaded page
3. 9. 2026
When you type an address into your browser and press Enter, a journey begins that takes a fraction of a second. First, your address is translated into a number that tells the network where to turn. The request then flies through your provider and the backbone networks all the way to the computer where the page lives – the server. The server sends the page back in small pieces, and the browser puts it together and displays it. Here we go through the whole journey step by step, clearly and without jargon.
What happens when you press Enter
Imagine you type uvtnet.cz into your browser and press Enter. At that moment the browser has no idea where the page is – it only knows its name. So first it has to ask where to turn. The whole journey has just a few steps and usually happens faster than you can blink.
First, the translation. The website's name is translated into a number the network understands (that is what the next section is about). Then, finding the route. Your request travels from your router to your provider and from there along the main routes to the computer where the page lives. Finally, the reply. The server sends the page back – not all at once, but in small pieces – and the browser puts the page together from them and displays it. And because one page is made up of many parts (text, images, styles), this exchange may well happen many times in a row.
You do not have to deal with any of this – it all happens by itself. But when you know what goes on under the bonnet, you will better understand why something sometimes fails to load and where the snag might be.
Why is it so fast? Apart from the fact that data flies through the network very quickly, the browser remembers a lot of things. When you return to a page, it already has part of it stored (in what is called the cache) and does not have to download it again from the other end of the route. That is why familiar websites open more quickly than ones you are seeing for the first time.
Why every device has its own number
For two devices on a network to send each other data, they have to find each other. That is what an IP address is for – a number that works much like a postal address. Without it, the network would not know whom to deliver a parcel of data to. Every phone, laptop and server has such a number.
It is good to know that a public address and a home address are not the same thing. Inside your home, devices have their own private numbers that only apply among them. To the outside internet, however, the whole household usually appears under a single public address assigned to it by the provider. It is a bit like the number of a block of flats: inside the flats you call each other by your first names, but the postman finds the building by a single address.
There is only a limited supply of these old-type public numbers (they are called IPv4) in the world, and in some places there are no new ones left. That is why the world is gradually moving to the newer and much roomier IPv6 addressing. Nothing changes for you – it is more like extending the numbering of a street when the old numbers run out.
Who hands out the home numbers? Your router. The moment you connect a new phone or laptop to it, the router automatically assigns it a free home number so that it does not clash with the others. You do not have to set anything up – that is why it is enough to “connect the device to the Wi-Fi” and it works straight away.
DNS: the internet's directory of names
People remember names; the network works with numbers. DNS is the service that translates between the two – it works like the internet's phone book. You type a name (uvtnet.cz), DNS finds the server's number for it, and only then does the connection set off on its way. Without DNS you would have to remember long numbers instead of website names.
When DNS is not working, it typically shows up as a page not loading at all even though “the internet is working” – it simply has nowhere to turn. That is exactly why the quality and security of the translation matter; we have devoted a separate article on DNS resolvers to it, where you can read how smart translation can also protect you against fraudulent websites.
Where does the translation take place? Your device asks a DNS server – often the one run by your provider, but it can be a different one. The network remembers the answer for a while so that it does not have to keep asking about the same name over and over again. That is another reason why the same website loads a little faster the second time.
Why data is sent in pieces
This is perhaps the nicest “aha” moment of the whole internet. Data is not sent in one piece but split into small chunks – packets. Each packet gets a sequence number and a destination address, sets off on its own, and on the other side the packets are put back together into the original whole in the right order. This splitting and reassembling is handled by a family of rules called TCP/IP – an agreed “language” that networks use to talk to each other.
Why so complicated? Because it makes the internet resilient. If one packet gets lost on the way, only that one is resent, not the whole file. And if part of the route goes down somewhere, the packets find another one – the network has plenty of connections between any two points.
Thanks to this, a connection survives even a fault that you often do not even notice. When a video “freezes” for a moment and then starts again, you have usually just seen this mechanism in action: a few packets arrived late or by a different route.
To make sure nothing is missing, the network ticks off the pieces it has received. If one does not arrive, the other side asks for it again – and only when they are all there are they put together into a complete whole. This check is exactly why an email reaches you complete even over a slow connection and never just half of it; the network would rather wait a moment than send half.
From you all the way to the backbone network
The data's journey has layers that fit inside each other like Russian dolls. Closest to you is your home network – the router and the laptop, phone or TV connected to it. It connects to the provider's network, which brings the connection all the way to you. And that in turn connects to the backbone network – the main routes linking providers and entire countries with each other.
The heart of your home layer is the router. We looked at which router you need for which connection, and why it matters more than it seems, in the article which router you need for your connection. For this explanation, it is enough to know that the router is the gateway between your home network and the rest of the world.
Where the internet physically runs
“The cloud” sounds airy, but the reality is very tangible. The internet runs on physical infrastructure: cables in the ground, optical fibres, wireless links and satellites. When you open a website, the data really does travel through a piece of glass or copper – it does not vanish into some “cloud”.
Today's most modern backbone is optical fibre – thin glass fibres that carry light instead of electricity. They can carry an enormous amount of data and are the foundation of what is called fast internet; we write more about them in the article on fibre internet. Between continents, data then travels along undersea cables laid on the ocean floor – yes, the vast majority of intercontinental traffic runs along the seabed, not via satellites. Where a cable does not pay off, wireless links take over, and in remote places satellites.
And the websites themselves? They “live” on servers – powerful computers running in large data centres. These are real buildings full of equipment, cooling and backup power. When you open a page, you ultimately connect to exactly such a computer – it may just as well be standing on the other side of the world.
Who owns the internet
Nobody. There is no owner or headquarters that owns the internet or switches it on. The internet is a network of networks – tens of thousands of separate networks that have connected voluntarily because it pays off for them. Each provider runs its own network and agrees with the others to let each other's traffic through. This interconnection is called peering and often takes place at shared interconnection points; the main Czech one is called NIX (Neutral Internet Exchange), and it is where the networks of Czech operators meet.
To hold it all together, there are a few international organisations that look after the rules, standards and registries – for example making sure that address numbers never overlap anywhere. None of them runs the internet, though; rather, they oversee rules that have been agreed jointly. We at ÚVT are one of the providers that build and interconnect such networks in the Czech Republic. Incidentally, how the internet in the Czech Republic grew from a hobby of enthusiasts into what it is today is nicely mapped out in the article the history of the internet in the Czech Republic.
This very openness is the whole magic of the internet: because everyone sticks to the same agreed rules (protocols), anyone new can join the network and immediately talk to everyone else. They do not need to ask anyone for permission – it is enough to follow the rules. Thanks to this, the internet keeps growing without anyone having to run it from one place.
Wi-Fi is not the same as the internet
This is the most common layman's misconception of all, and it is worth clearing up. Wi-Fi is just the wireless connection over the last few metres between your device and the router – it replaces a cable running through your flat. The internet is the whole network behind the router, that huge journey we described above. That is why Wi-Fi can work perfectly and yet the internet “is not working” – for example when the provider has an outage. And the other way round: you can also connect by cable, with no Wi-Fi at all.
If you want to get a better grasp of this – what kinds of wireless networks exist and how they differ – we have covered it in the article what Wi-Fi actually is.