Scribd
Upload
2 views1 page

NoSQL Overview - Key-Value Stores & Graph Databases

NoSQL databases were developed to address the limitations of relational databases in handling large-scale, fast-changing, or loosely-structured data. Key-value stores provide a simple way to store data as unique keys paired with values, while graph databases focus on representing data as nodes and edges to efficiently traverse relationships. Overall, NoSQL offers scalability, flexible schemas, and high performance, making it suitable for various applications like social networks and e-commerce.

Uploaded by

codexone1one
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
2 views1 page

NoSQL Overview - Key-Value Stores & Graph Databases

NoSQL databases were developed to address the limitations of relational databases in handling large-scale, fast-changing, or loosely-structured data. Key-value stores provide a simple way to store data as unique keys paired with values, while graph databases focus on representing data as nodes and edges to efficiently traverse relationships. Overall, NoSQL offers scalability, flexible schemas, and high performance, making it suitable for various applications like social networks and e-commerce.

Uploaded by

codexone1one
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

NoSQL Databases — Quick Overview

Database System Theory — Week 16 (Elmasri Ch. 24)

This is a quick, plain-language overview of three related ideas: why NoSQL databases exist, what key-value
stores are, and what graph databases are.

1. Why NoSQL? (The Motivation)


Relational databases (tables, rows, fixed schemas) work great for structured data, but they start to struggle with
very large-scale, fast-changing, or loosely-structured data — think social media feeds, sensor data, or
e-commerce catalogs with millions of users. NoSQL ("Not Only SQL") databases were built to handle exactly
these situations.
●
Scalability: NoSQL systems are designed to scale out across many cheap servers (horizontal scaling), rather
than scaling up one big expensive server.
●
Flexible schema: data doesn't need a fixed table structure decided in advance — useful when data shapes
vary or change often.
●
High performance: optimized for very fast reads/writes at massive scale, sometimes by relaxing strict
consistency guarantees (trading some consistency for speed and availability).
●
Specialized data shapes: some data (e.g., networks of relationships, simple lookups) maps more naturally onto
non-relational models.

2. Key-Value Stores
The simplest type of NoSQL database. Think of it like a giant dictionary or hash map: every piece of data is
stored as a key (a unique identifier) paired with a value (the actual data, which can be almost anything — a
string, a number, a JSON object, a file). The database doesn't care what's inside the value; it just stores and
retrieves it by key, extremely fast.

Key Value

user:1024 { name: "Ali", cart: ["itemA", "itemB"] }

session:9f3a { loggedIn: true, expires: "18:00" }

Good for: caching, session storage, shopping carts, user preferences — anything that's looked up by a simple ID
and doesn't need complex queries. Examples: Redis, Amazon DynamoDB, Riak.

3. Graph Databases
Graph databases store data as nodes (entities, e.g. a person, a place, a product) connected by edges
(relationships, e.g. "FRIENDS_WITH", "PURCHASED", "LOCATED_IN"). Both nodes and edges can hold their
own properties (e.g., a "FRIENDS_WITH" edge might have a "since: 2021" property). The model is built
specifically to make it fast to traverse relationships, something relational databases get slow at once you need
many chained joins.
Example: Alice → FRIENDS_WITH → Bob → WORKS_AT → Acme Corp. A graph database can quickly answer
questions like "who are Alice's friends-of-friends?" or "what's the shortest connection path between two people?"
— queries that would require many expensive joins in a relational database.
Good for: social networks, recommendation engines, fraud detection, knowledge graphs, network/route
mapping. Examples: Neo4j, Amazon Neptune.

In short: NoSQL trades some of the structure and strict consistency of relational databases for scale, flexibility,
and speed — key-value stores optimize for simple fast lookups, while graph databases optimize for exploring
connected relationships.

SSUET — Database System Theory Page 1

You might also like