Types of NoSQL Databases
What problem is this solving?
Saying "NoSQL" is not enough.
Different NoSQL databases solve different problems.
A cache, a document store, a graph database, and a write-heavy distributed database are not interchangeable.
Simple definition
NoSQL is not one database type.
It is a broad category.
The main types are:
NoSQL
├── Document Database
├── Key-Value Store
├── Wide-Column Store
└── Graph Database
Each type stores and reads data differently.
Real example
| Type | Data Shape | Examples | Best For |
| Document | JSON-like documents | MongoDB, CouchDB, Firestore | Object-shaped flexible data |
| Key-Value | key → value | Redis, Memcached | Cache, sessions, simple lookup |
| Wide-Column | Distributed rows with flexible columns | Cassandra, HBase, ScyllaDB | Huge write-heavy workloads |
| Graph | Nodes and relationships | Neo4j, Amazon Neptune | Relationship traversal |
Better explanation
Ask:
What shape is my data and how will I read it?
Examples:
- User profile with nested preferences → document database
- Session lookup by ID → key-value database
- Millions of time-series writes → wide-column database
- Friends of friends recommendation → graph database
Common mistake
Do not choose a NoSQL database only because it is popular.
Choose it because its data model matches the way your application reads and writes data.
Interview Answer
If an interviewer asks:
What are the types of NoSQL databases?
You can answer:
The main NoSQL types are document databases, key-value stores, wide-column stores, and graph databases. Document databases store JSON-like documents, key-value stores provide fast lookup by key, wide-column stores handle massive distributed writes, and graph databases store nodes and relationships for traversal-heavy data.