L LearnViaNotes @learnvianotes · November 3, 2026 · NoSQL (Not Only SQL) databases are built for horizontal scalability, unstructured schemas, and high write throughput, departing from the rigid ACID constraints of relational systems.
Traditional RDBMS scales vertically (faster CPU, more RAM), which has physical scaling limits and high hardware costs. NoSQL is designed to scale horizontally (adding cheap server nodes in a distributed cluster). Dynamic Schemas: Allows storing unstructured document records (JSON) without running migrations. Low Latency: Uses simple query patterns (e.g. key-value hash lookups) to bypass slow relational joins.
Dimension SQL Databases (RDBMS) NoSQL Databases Schema Strict, predefined tabular schema. Changes require DDL migration locks. Dynamic / Schema-less. Records can store custom dynamic columns. Scaling Vertical (Scale Up). Horizontal (Scale Out). Shards data partitions across multiple nodes. Joins Supported natively. Executed horizontally. Not supported. Data is denormalized in documents to avoid joins. Transactions Strict ACID properties. Absolute consistency. BASE model. Prioritizes availability; eventual consistency.
The 4 Types of NoSQL Databases 1. Document Database
Stores records as documents (JSON/BSON). Documents can nest objects and arrays inside themselves.
Use Cases: E-commerce catalogs, Content Management Systems (CMS), user profile configurations.
MongoDB · CouchDB
2. Key-Value Store
Stores simple key-to-value associations. Extremely fast because it behaves as a giant hash table in RAM.
Use Cases: Real-time caching, user session states, global leaderboards.
Redis · Memcached
3. Column-Family Store
Stores columns of data together on disk, rather than rows. Highly optimized for mass writes across columns.
Use Cases: IoT sensor metrics, time-series telemetry data, event logs.
Apache Cassandra · HBase
4. Graph Database
Stores data nodes and directed edges (relationships) natively, bypassing join layers for deep networks.
Use Cases: Social graph analysis, fraud detection, recommendation engines.
Neo4j · Amazon Neptune
Query Comparison: SQL vs MongoDB Action SQL Database MongoDB (NoSQL Document) Insert Row INSERT INTO Users (Name, Age) VALUES ('Alice', 25);db.users.insertOne({ name: "Alice", age: 25 });Select rows SELECT * FROM Users WHERE Age > 21;db.users.find({ age: { $gt: 21 } });Update rows UPDATE Users SET Age=26 WHERE Name='Alice';db.users.updateMany({ name: "Alice" }, { $set: { age: 26 } });
Write notes like this one
Cards for every idea, quizzes and flashcards, read aloud — on the web and your phone.
Something wrong with this note? Open it to report