Before understanding what the cloud solves, you need to understand how networked computing used to work (and still does). Almost everything you use daily—a website, a mobile app, email—relies on a very simple idea called the client-server model.

There is a program that requests things (the client) and another program that responds to those requests (the server).

The client and the server are two roles, not necessarily two different machines. But in practice, they are usually on different computers connected by a network.

Imagine a restaurant:

• You (the client) sit down and order a dish from the waiter.
• The kitchen (the server) receives the order, prepares the dish, and delivers it to you.
• You don’t need to know how the kitchen is organized, what stoves they use, or where they buy the ingredients. You just order and receive.

In computing, it works exactly the same way:

• Your browser (client) requests a page: “give me https://store.com/products”.
• The web server receives the request, looks up the data, generates the HTML, and sends it back to you.
• You don’t know (nor need to know) if there’s one or a hundred computers behind it.

When you open a website, this happens, simplified:

1. You type a URL.
2. Your browser makes a request using the HTTP protocol.
3. The server processes that request (sometimes queries a database).
4. The server returns a response with the content.
5. Your browser renders it on the screen.

When you open WhatsApp Web:

Your laptop does not store the entire conversation history of all users in the world: it only requests the part you’re interested in. The heavy lifting is done by the server.

For decades, setting up a service on the internet meant buying or renting a physical computer (a server) and putting it in a room with:

• Air conditioning so it doesn’t overheat.
• Redundant internet connection.
• Power systems with batteries in case the power goes out.
• Someone to watch over it and fix it when it breaks.

This room is called a datacenter. A company that wanted its website “always on” had to take care of all this itself. This is called having on-premise infrastructure (on your own premises), and we’ll look at it in detail in subchapter 1.3.

The client-server model does not disappear with the cloud: it’s the foundation of everything. What changes with the cloud is who provides and maintains the server.

• Before: you bought, set up, and maintained the physical server.
• With the cloud: you rent the server (or just the capacity you need) from a provider like AWS, and they take care of the hardware, the room, the air conditioning, and the electricity.

In other words: the “server” role still exists, but it’s no longer your physical problem. That’s exactly what we’re going to explore in the rest of the book.

• The client requests and the server responds: it’s the foundation of the internet.
• Client and server are roles, usually on different machines connected by a network.
• A traditional physical server requires space, energy, cooling, and maintenance.
• The cloud does not eliminate the client-server model: it just changes who takes care of the server.

In the next subchapter, we’ll look at the specific problems this traditional approach had and that gave rise to the cloud.