Course Code: CSE-5110
Course Title: Big Data Analytics
Topic: NoSQL
(Not Only SQL)
Created By: Shadia Afrin
Agenda
01 Definition of NoSQL
02 Differences between SQL and NoSQL
03 Why & When to Use NoSQL?
04 Types of NoSQL Databases
NoSQL - CAP Theorem
05 Popular NoSQL Tools
06 Use Cases of NoSQL in Big Data
07
Conclusion
10
08 Advantages
09 Limitations of NoSQL
Definition of NoSQL
The NoSQL system or "Not Only SQL" is essentially a
database that is made specifically for unstructured and semi-
structured data in very large quantities. Unlike Conventional
Relational Databases, where data are organized into tables
using predefined schemas. NoSQL allows flexible models to
be organized and horizontally scalable.
Differences between SQL and NoSQL
Feature SQL (Relational DB)
NoSQL (Non-
Relational DB)
Data Model Structured, Tabular
Flexible (Documents, Key-Value,
Graphs)
Scalability Vertical Scaling Horizontal Scaling
Schema Predefined Dynamic & Schema-less
ACID Support Strong Limited or Eventual Consistency
Best For Transactional applications Big data, real-time analytics
Examples MySQL, PostgreSQL, Oracle MongoDB, Cassandra, Redis
Why & When to Use NoSQL?
 No Universal Query Language: Unlike SQL in relational databases, each
NoSQL database has its own unique query approach.
 Flexible Data Model: NoSQL supports unstructured or semi-structured data
without requiring a fixed schema.
 Scalable and High-Performance: Designed for horizontal scaling and optimized
for fast read/write with large datasets.
 Distributed Architecture: Built for high availability and fault tolerance in
distributed environments.
 High scalability, flexible schema for unstructured data, real-time performance,
efficient distributed processing, and relaxed ACID requirements favoring
availability and speed.
Types of NoSQL Databases
Type Description Examples
Document
Stores data in
JSON-like
documents
MongoDB,
CouchDB
Key-Value Simple key-value
pairs
Redis, DynamoDB
Column-Family Column-based
storage
Cassandra, HBase
Graph Node and edge-
based relationships
Neo4j, JanusGraph
Popular NoSQL Tools
 MongoDB – Aggregation, indexing, flexible schema .
 Cassandra – Write-heavy systems, high scalability.
 Redis – In-memory, real-time analytics, caching .
 HBase – Based on Hadoop HDFS, real-time access.
 Neo4j – Relationship data, Cypher query language.
 DynamoDB – AWS managed NoSQL service.
Use Cases of NoSQL in Big Data
 Real-time analytics (Redis, MongoDB)
 Time-series data (Cassandra, InfluxDB)
 Social network analysis (Neo4j)
 Product catalogs & e-commerce (MongoDB, DynamoDB)
 IoT and sensor data (HBase, DynamoDB)
NoSQL - CAP Theorem
Advantages of NoSQL in Big Data Analytics
 Highly Scalable: Supports horizontal scaling across distributed systems.
 Flexible Schema: Schema-less design allows easy handling of unstructured/semi-structured data.
 High Performance: Fast read/write operations suitable for large-scale environments.
 Supports Diverse Data Types: Handles documents, key-value pairs, columnar, and graph data.
 Real-Time Processing: Enables real-time analytics and low-latency data access.
 High Availability: Built-in replication and failover ensure continuous data access.
 Integration Ready: Seamlessly connects with Big Data tools like Hadoop, Spark, and Kafka.
 Cloud Friendly: Many NoSQL tools offer fully managed, scalable cloud solutions.
Limitations of NoSQL
 Lack of Standardization
– No common query language like SQL; syntax varies by database.
 Limited Complex Query Support
– Poor handling of joins and multi-table operations.
 Eventual Consistency
– Prioritizes availability over consistency (CAP theorem trade-off).
 Steeper Learning Curve
– Requires learning unique data models and APIs (e.g., CQL, Cypher).
 Limited Transaction Support
– Often lacks full ACID compliance; unsuitable for critical systems.
 Tool Maturity Issues
– Less mature than RDBMS; fewer admin and optimization tools.
 Data Duplication Risk
– Schema flexibility may lead to redundancy and higher storage needs.
Conclusion
 NoSQL databases are flexible, scalable and high-performance
alternative to relational databases.
 Make modern applications such as real-time analytics, big data
processing and web applications more suitable for maintaining complex
requirements.
 Unfortunately, with their benefits come several trade-offs, including a
lack of ACID (atomicity, consistency, isolation, durability) compliance
and more complex management.
THANK YOU
Insert the Subtitle of Your Presentation

NoSQL in Big Data Analytics Tools .pptx

  • 1.
    Course Code: CSE-5110 CourseTitle: Big Data Analytics
  • 2.
    Topic: NoSQL (Not OnlySQL) Created By: Shadia Afrin
  • 3.
    Agenda 01 Definition ofNoSQL 02 Differences between SQL and NoSQL 03 Why & When to Use NoSQL? 04 Types of NoSQL Databases NoSQL - CAP Theorem 05 Popular NoSQL Tools 06 Use Cases of NoSQL in Big Data 07 Conclusion 10 08 Advantages 09 Limitations of NoSQL
  • 4.
    Definition of NoSQL TheNoSQL system or "Not Only SQL" is essentially a database that is made specifically for unstructured and semi- structured data in very large quantities. Unlike Conventional Relational Databases, where data are organized into tables using predefined schemas. NoSQL allows flexible models to be organized and horizontally scalable.
  • 5.
    Differences between SQLand NoSQL Feature SQL (Relational DB) NoSQL (Non- Relational DB) Data Model Structured, Tabular Flexible (Documents, Key-Value, Graphs) Scalability Vertical Scaling Horizontal Scaling Schema Predefined Dynamic & Schema-less ACID Support Strong Limited or Eventual Consistency Best For Transactional applications Big data, real-time analytics Examples MySQL, PostgreSQL, Oracle MongoDB, Cassandra, Redis
  • 6.
    Why & Whento Use NoSQL?  No Universal Query Language: Unlike SQL in relational databases, each NoSQL database has its own unique query approach.  Flexible Data Model: NoSQL supports unstructured or semi-structured data without requiring a fixed schema.  Scalable and High-Performance: Designed for horizontal scaling and optimized for fast read/write with large datasets.  Distributed Architecture: Built for high availability and fault tolerance in distributed environments.  High scalability, flexible schema for unstructured data, real-time performance, efficient distributed processing, and relaxed ACID requirements favoring availability and speed.
  • 7.
    Types of NoSQLDatabases Type Description Examples Document Stores data in JSON-like documents MongoDB, CouchDB Key-Value Simple key-value pairs Redis, DynamoDB Column-Family Column-based storage Cassandra, HBase Graph Node and edge- based relationships Neo4j, JanusGraph
  • 8.
    Popular NoSQL Tools MongoDB – Aggregation, indexing, flexible schema .  Cassandra – Write-heavy systems, high scalability.  Redis – In-memory, real-time analytics, caching .  HBase – Based on Hadoop HDFS, real-time access.  Neo4j – Relationship data, Cypher query language.  DynamoDB – AWS managed NoSQL service.
  • 9.
    Use Cases ofNoSQL in Big Data  Real-time analytics (Redis, MongoDB)  Time-series data (Cassandra, InfluxDB)  Social network analysis (Neo4j)  Product catalogs & e-commerce (MongoDB, DynamoDB)  IoT and sensor data (HBase, DynamoDB)
  • 10.
    NoSQL - CAPTheorem
  • 11.
    Advantages of NoSQLin Big Data Analytics  Highly Scalable: Supports horizontal scaling across distributed systems.  Flexible Schema: Schema-less design allows easy handling of unstructured/semi-structured data.  High Performance: Fast read/write operations suitable for large-scale environments.  Supports Diverse Data Types: Handles documents, key-value pairs, columnar, and graph data.  Real-Time Processing: Enables real-time analytics and low-latency data access.  High Availability: Built-in replication and failover ensure continuous data access.  Integration Ready: Seamlessly connects with Big Data tools like Hadoop, Spark, and Kafka.  Cloud Friendly: Many NoSQL tools offer fully managed, scalable cloud solutions.
  • 12.
    Limitations of NoSQL Lack of Standardization – No common query language like SQL; syntax varies by database.  Limited Complex Query Support – Poor handling of joins and multi-table operations.  Eventual Consistency – Prioritizes availability over consistency (CAP theorem trade-off).  Steeper Learning Curve – Requires learning unique data models and APIs (e.g., CQL, Cypher).  Limited Transaction Support – Often lacks full ACID compliance; unsuitable for critical systems.  Tool Maturity Issues – Less mature than RDBMS; fewer admin and optimization tools.  Data Duplication Risk – Schema flexibility may lead to redundancy and higher storage needs.
  • 13.
    Conclusion  NoSQL databasesare flexible, scalable and high-performance alternative to relational databases.  Make modern applications such as real-time analytics, big data processing and web applications more suitable for maintaining complex requirements.  Unfortunately, with their benefits come several trade-offs, including a lack of ACID (atomicity, consistency, isolation, durability) compliance and more complex management.
  • 14.
    THANK YOU Insert theSubtitle of Your Presentation