Duration: 2 hours (Lecture: 2 hours, No Lab)

Syllabus Topics:

• Lecture:
  • Introduction to NoSQL Database
  • Features of NoSQL Database
  • Structured vs. Semi-structured and Unstructured Data
  • Difference between RDBMS and NoSQL Databases
  • CAP Theorem
  • BASE Model
  • Categories of NoSQL Databases: Key-Value Store, Document Store, Column-Oriented, Graph

Lecture Notes (In-Depth)

1. Introduction to NoSQL Database

• Definition: NoSQL (Not Only SQL) databases are non-relational databases designed to handle large-scale, diverse, and unstructured data, offering flexibility, scalability, and performance for modern applications.
• Purpose:
  • Address limitations of traditional RDBMS in handling big data, high traffic, and varied data types.
  • Support distributed architectures for cloud-based and web-scale applications.
• Key Characteristics:
  • Schema-less or flexible schema: No fixed structure, allowing dynamic data models.
  • Horizontal scaling: Scale out by adding servers (sharding) rather than upgrading hardware (vertical scaling).
  • High performance for specific workloads (e.g., key-value lookups, document queries).
  • Support for diverse data types (e.g., JSON, XML, graphs).
• Examples:
  • MongoDB (Document Store), Redis (Key-Value Store), Cassandra (Column-Oriented), Neo4j (Graph).

2. Features of NoSQL Database

• Scalability:
  • Horizontal scaling across distributed nodes, ideal for cloud environments.
  • Example: Add more servers to handle increased traffic in MongoDB or Cassandra.
• Flexible Schema:
  • No fixed schema; documents or records can have varying structures.
  • Example: A MongoDB collection can store documents with different fields without predefined columns.
• High Availability:
  • Replication and partitioning ensure data availability even during node failures.
  • Example: Cassandra’s multi-node replication ensures no single point of failure.
• Performance:
  • Optimized for specific query patterns (e.g., key lookups in Redis, document retrieval in MongoDB).
  • Often faster than RDBMS for certain workloads due to denormalized data and indexing.
• Distributed Architecture:
  • Data is spread across multiple nodes for load balancing and fault tolerance.
  • Example: Sharding in MongoDB distributes data across servers.
• Support for Big Data:
  • Handles large volumes of data (terabytes to petabytes) and high-velocity data (e.g., real-time analytics).
• Key Notes:
  • NoSQL databases sacrifice some relational features (e.g., complex joins, strict ACID compliance) for scalability and flexibility.
  • Best suited for applications like social media, IoT, e-commerce, and real-time analytics.

3. Structured vs. Semi-structured and Unstructured Data

• Structured Data:
  • Highly organized, fits into fixed schemas (e.g., tables with rows and columns).
  • Example: RDBMS tables like employees with defined columns (emp_id, name, salary).
  • Characteristics: Fixed format, easily queried with SQL, used in traditional business applications.