After compute (EC2), the second major pillar of AWS is storage. And the star service is S3. It’s one of the most used services in the world, easy to get started with, and has countless applications. Let’s start with its three basic concepts: buckets, objects, and keys.

S3 stands for Simple Storage Service. It’s a service for storing files (which we call “objects” here) in a virtually unlimited, highly durable way, accessible from anywhere over the internet.

This type of storage is called object storage. It’s different from your computer’s disk:

• It does not have real folders or a traditional file system.
• You store complete files and retrieve them completely by their “name.”
• It’s ideal for images, videos, backups, website files, logs, data… almost any file.

Important: S3 is not for installing an operating system or for a database to constantly write to it (that’s what EBS, the EC2 disk, is for). S3 is for storing and retrieving files, not for using it as a server’s disk.

S3 is based on three ideas. Here’s a simple analogy:

Filing Cabinet Analogy: A bucket is like a filing cabinet. Each object is a document you store inside. The key is the label that identifies each document so you can find it.

A bucket is the container where you store your objects. Features:

• It has a globally unique name: there can’t be two buckets with the same name in all of AWS, in any account on the planet. (That’s why they often have prefixes like your company name.)
• It belongs to a specific region (remember Chapter 3): you choose where it physically lives.
• It can contain a virtually unlimited number of objects.

Example: A bucket named fotos-empresa-acme-2026. The name is unique worldwide and lives, for example, in the Ireland region.

An object is each file you store: a photo, a PDF, a video, a backup… Each object includes:

• The data itself (the file’s content).
• Metadata (information about the file: type, date, tags…).
• A key that identifies it (we’ll see this now).

An object can be very small (a few bytes) or huge (up to several terabytes).

The key is the full and unique name of the object within the bucket. It’s like a file path:

Here’s the trick that confuses many: “folders” in S3 don’t really exist. What you see as folders is actually part of the key’s name. The key 2026/junio/vacaciones/playa.jpg is a single text string; the slashes / just make the console display it as if they were folders, for easier navigation.

Why it matters: S3 has a flat structure (all objects are at the same level, identified by their key). The appearance of folders is just visual. This is different from your computer’s disk, where folders are real.

Each object has a unique address that combines the bucket and the key. Conceptually:

With the right permissions, you can upload, download, or delete that object from the web console, from the command line (AWS CLI), from code (with an SDK), or via API.

• Almost infinite scale: you don’t worry about size; store what you need.
• Highly durable: S3 is designed for 99.999999999% durability (the famous “eleven nines”). In practice, it’s extremely unlikely to lose an object, because AWS automatically keeps redundant copies.
• Pay as you go: you pay for what you store and the operations you perform, with no need to reserve anything in advance.
• Accessible from anywhere over the internet (with the right permissions).
• Foundation for many architectures: static websites, data lakes, backups, content distribution… we’ll see this in upcoming subchapters and chapters.

Real example: A mobile app stores its users’ profile photos in S3. Every time someone uploads a photo, an object is created with a key like usuarios/12345/perfil.jpg. The app can have millions of users without the team ever worrying about “disk space.”

• S3 is object (file) storage: ideal for images, videos, backups, data… not for using it as a server’s disk.
• Bucket: the container (globally unique name, lives in a region).
• Object: each stored file (data + metadata + key).
• Key: the object’s unique identifier. “Folders” are just a visual appearance; the real structure is flat.
• S3 stands out for its almost infinite scale and its “eleven nines” durability.

In the next subchapter, we’ll look at S3’s storage classes (Standard, IA, Glacier…), which let you adjust cost and access speed depending on how often you use each file.