There Is No Cloud

You’re on your phone. You tap a YouTube link. Two seconds later, a video from a server in California is playing on your screen in Lagos. You didn’t think about it. You never do. But somewhere between your thumb and that video, your request travelled through cables on the ocean floor, got chopped into pieces, routed through machines in Europe, reassembled on a computer you’ll never see, and streamed back to you — all in under two seconds.

Most people call this “the cloud.” There is no cloud. There’s just someone else’s computer, connected to yours by wires you’ll never touch. Let’s meet the wires.

I need to break something to you gently: the cloud is a lie. A useful lie, but a lie.

When someone says “it’s in the cloud,” they mean: it’s on a computer I don’t own, in a building I’ve never visited, connected to me by cables I’ll never see. That’s it. That’s the whole thing. AWS, Google Cloud, Azure — these are just companies with very large buildings full of very ordinary computers.

Physically, the internet is:

Undersea cables — over 500 of them, laid on the ocean floor, connecting continents. They carry roughly 99% of all international data. Yes, really. Your WhatsApp messages to your cousin in London are physically travelling through a cable at the bottom of the Atlantic.
Data centres — warehouse-sized buildings full of servers, in places like Lagos, London, Frankfurt, Virginia.
Routers and switches — the traffic cops of the network, passing data from one point to the next.
Cell towers and satellites — the last-mile delivery.
Your device — the final destination.

Internet — The global system of interconnected computer networks, together with the physical links between them, that carries data between devices.

In plain terms: “The roads. The web is just one vehicle that drives on them.”

Analogy (Nigerian edition): Think of the internet as the national road network. The undersea cables are the trunk roads connecting states and countries. The backbone routers are the major highways. Your ISP’s network is the local road to your street. Your router is the gate to your compound. And your phone is the room where the actual work happens.

Every website you load is data physically travelling across all of that. In under a second. Every time.

The internet is not the web (and this trips up everyone)

Here’s something most tutorials skip, and most beginners get wrong for years:

The internet and the web are not the same thing.

The internet is the physical network — the cables, routers, protocols, and machines. It’s the road.
The web is a service that runs on top of the internet — websites, URLs, HTTP, browsers. It’s the vehicles.

Web — Short for World Wide Web: a system of interlinked hypertext documents and other resources, addressed by URLs and transferred over HTTP, that runs as a service on top of the internet.

In plain terms: “The vehicles. One way of using the roads, not the roads themselves.”

Other services also run on the internet: email, video calls, online gaming, file transfers, WhatsApp, Zoom. None of those are “the web.” They all just use the internet.

The internet existed before the web. It was built in the late 1960s. The web was invented in 1989 by a man at CERN who gave it away for free. We’ll meet him properly in a minute.

Hold onto this distinction. It’ll make you sound like you know what you’re talking about in any interview.

Chopping the book into pages

Now the first really clever idea: nothing on the internet is ever sent as one piece.

When you send a WhatsApp message, load a YouTube video, or open a website — your data is chopped into small chunks called packets. Each packet is typically a few hundred to about 1,500 bytes.

Each packet carries two things:

A slice of the data — a small piece of your message, image, or video.
A header — metadata saying where it came from, where it’s going, which packet number it is, and how to reassemble.

Packet — A small, independently addressed unit of data — typically a few hundred to about 1,500 bytes — carrying a slice of a message plus a header describing where it came from, where it is going, and how to reassemble it.

In plain terms: “One page torn out of a book, in its own envelope, posted on its own.”

Analogy: Imagine you want to send a 500-page book from Lagos to Kano. You can’t put it in one envelope. So you tear the book into 500 individual pages. You put each page in its own envelope. You write on each envelope: “From: You, Lagos. To: Friend, Kano. Page 1 of 500.” Then you post them all — and crucially, they don’t travel together. Some go by road. Some go by air. Some take a longer route. They arrive out of order. Some get lost. Your friend in Kano receives them all, sorts them by page number, and reassembles the book.

That’s exactly how the internet works. Every message. Every video. Every website.

Why bother doing it this way? Because packets are resilient. If one route is blocked, packets take another. If a packet is lost, only that one packet is re-sent — not the whole message. Different packets can travel in parallel, using the entire network efficiently.

This wasn’t an accident. This design was deliberate. Read on.

Every device has a home

Every device connected to the internet has an IP address — a unique number that identifies it, like a house address.

There are two versions in use today:

IPv4 — the old system. Looks like 192.168.1.1 — four numbers from 0 to 255, separated by dots. It allows for about 4.3 billion unique addresses. We ran out of them in 2011.

IPv4 — The first widely deployed version of the Internet Protocol: 32-bit addresses written as four decimal numbers between 0 and 255, giving about 4.3 billion possible addresses — a pool that ran out in 2011.

In plain terms: “The old numbering system, and it is full.”

IPv6 — the new system. Looks like 2001:0db8:85a3:0000:0000:8a2e:0370:7334. Much longer, uses hexadecimal. Provides so many addresses that every grain of sand on Earth could have one — and you’d still have room left over.

IPv6 — The current version of the Internet Protocol: 128-bit addresses written in hexadecimal, providing an address supply large enough that exhaustion is no longer a practical concern.

In plain terms: “The new numbering system, with room to spare.”

IP address — A numerical address assigned to a device on a network so that packets can be routed to it.

In plain terms: “A house address, for packets.”

Analogy: An IP address is like the address on an envelope. Without it, the postal system has no idea where to send your letter. Without an IP, packets have nowhere to go.

But here’s a wrinkle: IP addresses are not permanent. Your home IP changes. Your phone’s IP changes as you move between cell towers. So how does a packet find you?

Two kinds of IP exist in your daily life:

Public IP — what the outside world sees. Assigned by your internet provider. Like the address of your entire street.
Private IP — what devices use inside your home network (like 192.168.1.5). Like apartment numbers inside a building.

Router — A device that forwards packets between networks, choosing the path each one takes toward its destination.

In plain terms: “The traffic cop at the junction between networks.”

Your router translates between the two. That translation process has a name — NAT (Network Address Translation) — but you don’t need the details today. Just know the concept exists, and that it’s why your laptop, phone, and TV can all be online at once without the outside world confusing them.

NAT — Network Address Translation: the technique of rewriting the source or destination address in packets as they pass through, letting many private devices share a single public IP address.

In plain terms: “One street address outside, many apartment numbers inside.”

The dispatch rider and the tracker

Now we get to the two protocols that make the entire internet work: TCP and IP. They’re always mentioned together — “TCP/IP” — but they do very different jobs.

IP — Internet Protocol

IP’s job is simple: get the packet to the right address. That’s it. IP is fast, dumb, and doesn’t care. It receives a packet, reads the destination address, and sends it on its way. It doesn’t check if the packet arrives. It doesn’t reorder them. It doesn’t call you back to say “hey, that one got lost.” It’s the dispatch rider who drops the package at the address and zooms off.

TCP — Transmission Control Protocol

TCP is IP’s responsible older sibling. It sits on top of IP and adds the guarantees:

Handshake — before sending data, TCP confirms the connection: “Can you hear me?” “Yes, I can hear you.” “Great, let’s talk.”
Acknowledgement — every packet received is confirmed: “Got packet 1.” “Got packet 2.”
Retransmission — if a packet is lost, TCP re-sends it.
Reordering — if packets arrive out of order, TCP puts them back in sequence.
Flow control — if the receiver is overwhelmed, TCP slows down.

TCP — Transmission Control Protocol: a transport protocol that provides reliable, ordered delivery of a stream of bytes over a connection, by acknowledging received packets, re-sending lost ones, and putting out-of-order arrivals back in sequence.

In plain terms: “The tracker who signs for every parcel and chases the ones that go missing.”

Analogy: If IP is the dispatch rider, TCP is the WhatsApp thread between you and the recipient. Every package is confirmed. Missing ones are re-sent. Everything is accounted for.

And then there’s UDP

Sometimes you don’t need TCP’s guarantees. Live video calls, online games, and live streams use UDP instead — a faster protocol that makes zero promises.

If a frame drops during a video call, you don’t want the whole call to freeze while it re-sends. You want it to keep going. That’s UDP.

	TCP	UDP
Connection	Yes (handshake)	No
Guarantees delivery	Yes	No
Reorders packets	Yes	No
Speed	Slower	Faster
Used for	Websites, emails, file transfers	Video calls, games, live streams

UDP — User Datagram Protocol: a transport protocol that sends datagrams with no handshake, no acknowledgement, and no re-sending, trading reliability for speed.

In plain terms: “Posting a letter and walking away. Fast, cheap, no receipt.”

Four layers, one delivery

The internet is built in layers. Each layer does one job and hands off to the next. Think of it like sending a parcel through a courier company.

Layer	Job	Analogy
Application	HTTP, FTP, SMTP — the “language” being spoken	What you wrote in the letter
Transport	TCP/UDP — chop into packets, guarantee delivery	The courier’s tracking system
Internet	IP — address the packet and route it	The address on the envelope
Network Access	Physical wires, WiFi, Ethernet	The road, the bike, the plane

Each layer only cares about its own job. The application layer doesn’t know or care whether packets travel by undersea cable or satellite. The physical layer doesn’t know or care what’s inside the packet.

This is the genius of the design. Each layer can be replaced or upgraded without breaking the others. That’s why WiFi, 4G, 5G, and fibre — all different physical layers — can all carry the same websites. The upper layers don’t care.

Built to survive a war

You asked for backstory. This one’s worth it.

1969 — ARPANET. The US Department of Defense’s research agency (ARPA) wanted a communication network that could survive a nuclear attack. If one node was destroyed, the network should keep working. The solution: decentralise everything. No single point of failure. Packets can route around damage.

ARPANET — The packet-switching research network funded by the US Department of Defense’s Advanced Research Projects Agency and switched on in 1969 — the direct ancestor of the internet.

In plain terms: “The first network built on the assumption that parts of it might be destroyed.”

This is why the internet has no centre. It wasn’t designed to be efficient. It was designed to be unbreakable.

1974 — TCP is born. Vint Cerf and Bob Kahn publish a paper describing TCP. Later, TCP is split into TCP and IP. Vint Cerf is now called “one of the fathers of the internet.”

1983 — The flag day. On January 1, 1983, ARPANET officially switches from an older protocol to TCP/IP. This is considered the birthday of the modern internet. Everyone had to switch on the same day. It was chaos. It worked.

1989–1991 — The Web is invented. Tim Berners-Lee, a British scientist at CERN, invents the World Wide Web — a system of hyperlinked documents on top of the internet. He builds HTML, HTTP, URLs, and the first browser. Then he gives it away for free. No patent. No licence. He could have been a trillionaire. He chose not to be.

The internet already existed. The web was built on top of it. That’s the distinction I promised at the start.

What you’d say in the room

If an interviewer asks “Can you explain how the internet works?” — here’s a 60-second answer that puts you above most juniors:

“The internet is a global network of computers connected by physical infrastructure — undersea cables, routers, and data centres. When you send data, it’s broken into packets, each with a destination IP address. These packets travel independently across the network, potentially taking different routes. TCP ensures they all arrive, in order, by acknowledging each one and re-sending lost packets. IP handles addressing and routing. The Web — websites, HTTP, URLs — is just one service running on top of this. Email, video calls, and games are others. The whole thing is designed in layers, so each layer can evolve independently.”

That’s the answer. Learn it. Say it out loud until it’s yours.

Say it like a professional

The definitions from today, in one reviewable block. Say each one out loud. If you stumble, go back to that section.

Internet — The global system of interconnected computer networks, together with the physical links between them, that carries data between devices.
Web — Short for World Wide Web: a system of interlinked hypertext documents and other resources, addressed by URLs and transferred over HTTP, that runs as a service on top of the internet.
Packet — A small, independently addressed unit of data — typically a few hundred to about 1,500 bytes — carrying a slice of a message plus a header describing where it came from, where it is going, and how to reassemble it.
IP address — A numerical address assigned to a device on a network so that packets can be routed to it.
IPv4 — The first widely deployed version of the Internet Protocol: 32-bit addresses written as four decimal numbers between 0 and 255, giving about 4.3 billion possible addresses — a pool that ran out in 2011.
IPv6 — The current version of the Internet Protocol: 128-bit addresses written in hexadecimal, providing an address supply large enough that exhaustion is no longer a practical concern.
TCP — Transmission Control Protocol: a transport protocol that provides reliable, ordered delivery of a stream of bytes over a connection, by acknowledging received packets, re-sending lost ones, and putting out-of-order arrivals back in sequence.
UDP — User Datagram Protocol: a transport protocol that sends datagrams with no handshake, no acknowledgement, and no re-sending, trading reliability for speed.
Router — A device that forwards packets between networks, choosing the path each one takes toward its destination.
NAT — Network Address Translation: the technique of rewriting the source or destination address in packets as they pass through, letting many private devices share a single public IP address.
ARPANET — The packet-switching research network funded by the US Department of Defense’s Advanced Research Projects Agency and switched on in 1969 — the direct ancestor of the internet.

Before you close this tab

Say these answers out loud, in your own words, without looking:

What’s the difference between the internet and the web?
What is a packet, and why is data broken into them?
What does IP do? What does TCP do?
What’s the difference between TCP and UDP?
Why was the internet designed to have no central point?
Who invented the Web, and when?

If you can answer all six without peeking — you’ve nailed Day 1.

If you can’t, read the section you’re fuzzy on again. No rush. This is the foundation. Everything else sits on top of it.

Tomorrow, we meet the two roles every computer on the internet plays: the client and the server. It turns out those aren’t specific machines — they’re roles. And you play both, every day, without realising it.