What are the main differences between relational and non-relational data storage systems?

Relational DBMSs (e.g., MySQL, PostgreSQL) store data in tables with a clearly defined schema (rows and columns). They use SQL for queries and support transactions, relationships between tables (via keys), ensuring data integrity.

Non-relational DBMSs (NoSQL, e.g., MongoDB, Cassandra) can store data in various formats: documents, key-value, graphs, columns. They are more flexible in data structure, often scale horizontally, and are better suited for large volumes of unstructured data.

Main differences:

• Data structure: fixed schema (relational) vs flexible/no schema (non-relational)
• Query language: SQL vs various APIs/query languages
• Scalability: vertical (relational) vs horizontal (non-relational)
• Transactions: ACID (relational) vs often BASE (non-relational)

Example: in a relational database, a users table with columns id, name, email; in a document database — a collection of documents, where each document is a JSON with arbitrary fields.