How do NoSQL databases differ from relational ones?

Main differences between NoSQL databases and relational databases:

1. Data structure:

   • Relational databases (SQL) store data in tables with a fixed schema (rows and columns).
   • NoSQL databases can store data in various formats: documents (JSON), key-value, graphs, column stores.

2. Schema:

   • SQL schemas are strictly defined and require migrations when changes are made.
   • NoSQL schemas are often schema-less, allowing flexible addition of new fields without structural changes.

3. Scalability:

   • Relational databases typically scale vertically (adding resources to a single server).
   • NoSQL databases are designed for horizontal scaling (adding new nodes).

4. Transactions and consistency:

   • SQL supports ACID transactions, ensuring strong consistency.
   • Many NoSQL databases sacrifice strict consistency for performance and scalability, supporting eventual consistency.

5. Query language:

   • SQL uses a standardized query language.
   • NoSQL databases have different APIs and query languages, often simpler but less versatile.

Thus, the choice between SQL and NoSQL depends on requirements for data structure, scalability, and consistency.