NOSQL

TOPICS

• What is NoSql?
• Types of Nosql
• What is Schema?
• Traditional Sql vs NoSql
• Vertical Scalability
• Horizondal Scalability

What is NoSQL?

NoSQL stands for "Not Only SQL," non-relational databases designed to store, retrieve, and manage large volumes of unstructured, semi-structured, or flexible structured data.

NoSQL databases are designed to handle:

• Unstructured or semi-structured data
• Massive volumes of data
• High-speed, high-throughput applications

NoSQL databases do not require a fixed schema and typically do not use SQL for querying.

---

Types of NoSQL Databases

There are four main types of NoSQL databases, each suited to different use cases:

1. Document-Based Database
2. Key-value Stores
3. Graph Databases
4. Column-family stores

1. Document-Based Databases (e.g., MongoDB, CouchDB)

• Store data in JSON-like documents.
• Each document is a self-contained unit of data.
• Ideal for: Content management systems, user profiles, product catalogs.

Example:

{
  "name": "Alice",
  "email": "alice@example.com",
  "orders": [
    { "id": 1, "item": "Book", "price": 15 },
    { "id": 2, "item": "Pen", "price": 2 }
  ]
}

---

2. Key-Value Stores (e.g., Redis, DynamoDB)

• Data stored as key-value pairs (like a big dictionary or hash map).
• Extremely fast for read/write operations.
• Ideal for: caching, session storage, real-time apps.

Example:

"session123": "{user: 'John', isLoggedIn: true}"

---

3. Graph Databases (e.g., Neo4j, ArangoDB)

• Use nodes (entities), edges (relationships), and properties to represent and store data.
• Nodes: Entities like people, places, and things.
• Edges: Relationships between nodes (e.g., "friend of", "purchased").
• Properties: Both nodes and edges can have key-value data.
• Ideal for: Social networks, recommendation engines, fraud detection.

Example:

Alice --friend--> Bob --follows--> Charlie

---

4. Column-Family Stores (e.g., Apache Cassandra, HBase)

• Store data in columns instead of rows.
• Designed for fast reading of large volumes of data.
• Ideal for: logging, IoT, and time-series data.

Example:

yamlCopyEditRow: 123
  Name: Alice
  Age: 30
  Email: alice@example.com

UserID	Personal Info (column family)	Preferences (column family)
1	name: Alice, age: 30	language: English, theme: dark
2	name: Bob	language: Spanish

---

What Is a Schema in Databases?

In traditional SQL (relational) databases like MySQL, PostgreSQL, or Oracle, a schema defines the structure of your data before you store it. This includes:

• Tables
• Columns (with data types)
• Relationships (foreign keys)
• Constraints (like NOT NULL, UNIQUE, etc.)

For example:

CREATE TABLE users (
    id INT PRIMARY KEY,
    name VARCHAR(100),
    email VARCHAR(100),
    age INT
);

This schema says every user must have:

• an integer id,
• a name,
• an email,
• and, optionally, an age.

You must follow this structure when adding data. If your data doesn’t match this schema, the database will reject it.

---

Traditional SQL vs. NoSQL

Feature	SQL (Relational DB)	NoSQL (Non-relational DB)
Data model	Tables with rows and columns	Documents, key-values, graphs, etc.
Schema	Fixed and predefined	Flexible or schema-less
Joins	Supported	Usually not supported or limited
Scaling	Vertical (scale up)	Horizontal (scale out)
Query Language	SQL	Varies (JSON queries, API calls, etc.)
Best for	Structured, relational data	Unstructured or rapidly changing data

---

Vertical Scalability (Scale Up)

• Adding more power (CPU, RAM, SSD) to one server.
• Common in SQL databases.
• Has physical limits (you can’t upgrade forever).
• Gets expensive as upgrades are high-end hardware.

Example: You upgrade a machine from 16GB to 64GB RAM to handle more database queries.

---

Horizontal Scalability (Scale Out)

• Add more servers/nodes instead of upgrading one.
• NoSQL databases are designed to work on clusters of machines.
• They can distribute data across many machines.
• Helps in handling huge amounts of data or high user loads.

Example: You have 10 servers, and your NoSQL database automatically stores part of the data on each one.