Non-relational databases ¶

Non-relational databases, a.k.a. NoSQL (NotOnlySQL), are databases using a different approach to data storage and manipulation compared to relational databases ✨.

Here are some majors NoSQL databases models/stores:

• Key-value 🔑: a key-value system
• Document 📂: data stored as documents
• Columnar 🗄️: using a key-column (values) system
• Graph (GDB) 🕸️: nodes and edges

👉 NoSQL means that SQL databases are not the only databases.

Each NoSQL database model has its pro and cons, but overall, NoSQL databases' pros and cons are:

• 🌱 They were designed to handle large amounts of data (scalable)
• ✍️ They are flexible (no schema, semi- or completely unstructured)
• 🔥 Data may be duplicated or inconsistent (quantity over quality)
• 🚀 They have high performances (required by low-latency apps)
• 📒 They have limited query features (relatively to SQL/...)

NoSQL databases became quite popular around 2010, and they are used by a lot of actors like Google, Microsoft, Facebook, and Twitter.

Key-value

• Redis

Document

• MongoDB
• RavenDB
• RethinkDB
• Apache CouchDB

Columnar

• Apache Cassandra
• Apache HBase

Graph

• Neo4j

• BASE (SQL ACID)
• Wide-column store (subset of columnar)

Graph Query Language (GQL)

• SPARQL