Database Structure: Server → Table

A big company does not keep all its data in one giant table. Data is arranged in a hierarchy, like a library: the building (server) has many sections (databases), each section has shelves (schemas), and each shelf holds books (tables). Knowing this hierarchy helps you read error messages, write full object names, and talk confidently with database admins.

Definition

Database hierarchy: a server holds many databases; a database groups related schemas; a schema is a named container (namespace) that groups related tables and other objects; a table holds the actual rows of data.

SERVER
├── DATABASE: IPL_Stats
│   ├── schema: matches   → tables: innings, balls
│   └── schema: players   → tables: players, teams
└── DATABASE: IPL_Fans
    ├── schema: tickets   → tables: orders, seats
    └── schema: app       → tables: users, sessions

Each level explained

Why schemas? They keep things organised and secure. The ticketing team can be given permission on the tickets schema only, without touching the app schema. Two schemas can even contain tables with the same name (tickets.orders and app.orders) without a clash.

Note: In SQL Server every database has a default schema called dbo (database owner). If you create a table without naming a schema, it goes into dbo.

Fully qualified names

Because of the hierarchy, a table can be referred to by a longer, unambiguous name:

-- table only (uses your current database and default schema)
SELECT * FROM players;

-- schema.table
SELECT * FROM dbo.players;

-- database.schema.table
SELECT * FROM IPL_Stats.dbo.players;

The pattern is server.database.schema.table (the server part is only used for linked servers). Most of the time you write just the table name, after choosing the database with USE IPL_Stats;.

How tables are stored on disk

The table you see in a query result is a logical view. Physically, SQL Server saves everything in database files on the disk:

Inside a data file the rows are saved in small blocks called pages (8 KB each in SQL Server). When you run a query, the DBMS reads the needed pages from disk into memory and shows you a neat table. You never open the .mdf file yourself — the DBMS does that for you.

Note: Strictly speaking the files belong to the database, not to one table — one IPL_Stats.mdf file holds all tables of that database.

A single value at one row-and-column position — for example 'Virat Kohli' in the name column of row 1 — is called a cell (or field value).

Common mistakes

• ❌ Treating "schema" as a synonym for "database". ✅ In SQL Server a schema lives inside a database. (In MySQL, though, the words schema and database mean the same thing.)
• ❌ Getting "Invalid object name 'players'" and assuming the table is missing. ✅ Check you are in the right database (USE IPL_Stats;) and the right schema (matches.innings vs dbo.innings).

SQL Server vs MySQL / PostgreSQL

Interview tip

Q: What is a schema in SQL Server and why use one? A: A schema is a named container inside a database that groups tables, views and other objects. It helps organise objects by area (sales, HR, tickets), lets you grant permissions at schema level, and allows the same object name in different schemas. The default schema is dbo.