DP-900 · 1 Describe core data concepts
Common data stores
Exam objective: Describe features of common data stores including databases
A database is a dedicated system for storing and querying records. Relational databases use tables and keys, while nonrelational databases skip the fixed schema.
A database is a dedicated system for storing and querying records, built for that job rather than just holding files.
Relational databases store structured data as tables that represent entities, such as customers or orders. Each row gets a primary key, and a related table can reference that key instead of repeating the data. That reuse of keys is called normalization, and it is what makes the model relational. Tables are queried with SQL, an ANSI standard that behaves similarly across database products.
Nonrelational databases, often called NoSQL, do not force data into that fixed table and key structure. Microsoft Learn groups them into four common types:
| Type | How it stores a record |
|---|---|
| Key-value | a unique key and a value in any format |
| Document | a key-value pair whose value is a JSON document |
| Column family | rows and columns, grouped into related column families |
| Graph | entities as nodes, with links that define relationships |
On the exam, match the scenario to the shape: a fixed schema with keys points to relational, a flexible key-value or document shape points to nonrelational.
Key points
- A database manages records directly, which is different from simply saving files to a file system.
- Relational databases store structured data in tables. A primary key identifies each row, and other tables reference that key.
- Normalization removes duplicate values, for example keeping one customer record instead of repeating the customer's details on every order.
- Nonrelational databases do not apply a relational schema. The four common types are key-value, document, column family and graph.
- A document database is a key-value database whose value is a JSON document that the engine can query directly.
Exam trap
NoSQL does not mean no SQL at all. Some nonrelational engines accept a SQL-like query language. What makes a database nonrelational is that it does not enforce a fixed relational schema and keys, not the query syntax it accepts.
Check yourself
A social platform stores millions of people and the connections between them, and needs to run queries such as finding a friend of a friend. Which type of nonrelational database fits best?
Go deeper on Microsoft Learn
Checked against Microsoft Learn on October 1, 2026.