What Is a Database and How Does It Work

The short answer

A database is an organized collection of information stored so that it is easy to find, change and protect. Customer lists, product catalogs, order history — almost every application keeps this kind of data in a database.

Two terms are worth separating:

• Database — the data itself and its structure.
• DBMS (database management system) — the software that manages the data: PostgreSQL, MySQL, MongoDB, Redis and others.

An application does not read files on disk directly. It sends the DBMS a query — "find all orders from this customer in May" — and gets a ready answer back.

Why not just a spreadsheet

A spreadsheet works while the data is small and one person edits it. Problems start when:

• hundreds of users place orders and change stock levels at the same time;
• there are millions of records and answers are needed in a fraction of a second;
• you cannot allow a payment to go through while the order fails to be created;
• access must be restricted: a sales manager sees their own deals, an accountant sees invoices.

A DBMS exists to solve exactly these problems: concurrent access, fast lookup, integrity and permissions.

Tables, rows and columns

The most common type is the relational database. Data lives in tables that look like spreadsheet sheets but follow strict rules.

Take an online store:

• A table (customers) describes one type of object — customers.
• A row (record) is one specific customer.
• A column (field) is one property: name, email, city. Each column has a type: number, text, date.
• A primary key (id) is a unique identifier that points to exactly one row.

Orders sit in a separate orders table, and each order has a customer_id field that refers to a customer. That is a relationship between tables. Customer details are stored once instead of being copied into every order.

A CRM follows the same logic: contacts, deals and tasks tables linked by keys.

How a query works

Relational databases are queried with SQL:

SELECT name, email
FROM customers
WHERE city = 'Tashkent';

What happens inside:

1. The DBMS parses the query and checks that the table and columns exist.
2. The query planner picks the fastest way to run it: scan the whole table or use an index.
3. Data is read from memory or disk, filtered and returned to the application.

An index works like the index at the back of a book: instead of reading every row, the DBMS jumps straight to the matching ones. Without indexes, large tables become slow.

How data is stored and protected

• Data lives in files on disk, and frequently used parts are cached in RAM.
• Changes are first written to a log (write-ahead log), so the database can recover after a power failure.
• A transaction groups several actions into one: "charge the card and create the order" either completes fully or not at all.
• Backups and replication (a copy of the database on another server) protect against losing a disk or a server.

The main database families

Most projects start with a single relational database and add others only when a specific need appears — Redis for caching, for example.

Common beginner mistakes

• Putting everything into one huge table with duplicated data.
• Skipping types and constraints, so a "date" field ends up holding random text.
• Setting up backups only after the first data loss.
• Choosing a trendy database instead of the one that fits the task.