NoSQL Databases

Introduction to NoSQL databases

• Overview of the NoSQL database [1], [2], [3], [4], [5]

  NoSQL (not only SQL) databases, or non-relational databases, have been gaining popularity in recent years due to their flexibility, scalability, and ability to handle big data. Unlike traditional relational databases, NoSQL databases are not based on the relational model and do not use SQL as their primary query language. They store data in a variety of ways, such as key-value pairs, document-based models, graph databases, and column-family stores, and so on.

• History of NoSQL databases

  The term “NoSQL” was first used in 1998 to describe a lightweight, open-source relational database management system called “Strozzi NoSQL.” The term was later popularized by Johan Oskarsson in 2009 to describe a new generation of non-relational databases that were designed to handle large volumes of unstructured data. Since then, NoSQL databases have become increasingly popular due to their ability to scale horizontally, handle big data, and provide high availability.

  The old meaning of NoSQL was “No support for SQL”, but now it means “Not only SQL”. This is because NoSQL databases can support SQL-like query languages and other query languages that are specific to the database.

• Characteristics of NoSQL databases

  These characteristics are not for every NoSQL databases. They are just commonly seen for most NoSQL databases.

  • Non-Relational: NoSQL databases do not use the traditional relational model for storing data. Instead, they use alternative data models such as document, key-value, column-family, or graph-based models.

  • Horizontal Scalability: NoSQL databases are designed to scale horizontally by adding more commodity servers to the system. This makes it easy to handle large amounts of data and high levels of traffic.

  • High Availability: NoSQL databases are designed to be highly available by replicating data across multiple nodes in the system. This ensures that the system can continue to operate even if some nodes fail.

  • Distributed Architecture: NoSQL databases are designed to run on distributed architectures, which allows them to handle large amounts of data and high levels of traffic. Data is typically partitioned across multiple nodes in the system to enable parallel processing.

  • Flexible Schema: NoSQL databases have a flexible schema that allows for schemaless data modeling. This means that data can be added or removed from the database without the need for a predefined schema.

  • Performance: NoSQL databases are optimized for high performance and low latency. They achieve this by using memory-based data structures, caching, and distributed processing.

• Advantages over traditional relational databases

  Comparing NoSQL databases with relational databases, NoSQL databases offer several advantages, including the ability to handle unstructured data, horizontal scalability, and high availability. They are designed to handle large volumes of data, which can be stored across multiple nodes in a distributed environment. In contrast, traditional relational databases are optimized for structured data, and they are not as flexible when it comes to handling unstructured or semi-structured data.

• Drawbacks compared to relational databases

  There are also some drawbacks to using NoSQL databases, such as a lack of support for complex transactions, difficulty in integrating with legacy systems, and limited query capabilities. The data integrity model for NoSQL databases are usually not as strong as that for relational databases. [6] Therefore, it is essential to evaluate the pros and cons of NoSQL databases before deciding whether to use them.

SQL vs NoSQL

Features	SQL	NoSQL
Model	Relational	Non-relational
Schema	Rigid	Flexible
Query language	SQL	DSL
Scalability	Vertical	Horizontal
Transactions	ACID	BASE to ACID
Integrity	Strong	Eventual to strong

• Future of NoSQL databases

  The future of NoSQL databases looks promising, as they are well-suited for handling big data, which is becoming increasingly important in many industries. With the growth of cloud computing and the adoption of distributed computing, NoSQL databases are expected to become even more popular in the coming years.

Types of NoSQL databases

NoSQL databases can be classified into different categories based on their data models, architecture, and storage systems. One of the most common classifications of NoSQL databases is based on their data models. In this classification, there are four main types of NoSQL databases: document-based, key-value, column-family, and graph databases. Each type of NoSQL database has its own unique features, advantages, and limitations, making it important to choose the right type of database for a particular use case.

• Document-based databases (MongoDB, Couchbase)

  Document-based databases store data in a document-oriented model, where each document represents a record in the database. This model allows for flexible and dynamic schema design, making it suitable for storing and retrieving complex and unstructured data. Examples of document-based databases include MongoDB, and Couchbase.

• Key-value databases (Redis, Amazon DynamoDB)

  Key-value databases store data in a simple key-value format, where each record is identified by a unique key. This model is suitable for storing and retrieving simple data types and is known for its high scalability and performance. Redis and Amazon DynamoDB are two popular examples of a key-value database.

• Column-family databases (Apache Cassandra, Apache HBase, Google BigTable)

  Column-family databases store data in a column-oriented format, where data is organized in column families or column groups. This model is suitable for storing and retrieving large amounts of structured and semi-structured data, and is known for its high scalability and fault-tolerance. Examples of column-family databases include Apache Cassandra and HBase.

• Graph databases (Neo4j, OrientDB)

  Graph databases store data in a graph or network structure, where data entities are represented as nodes or vertices, and their relationships are represented as edges or links. This model is suitable for storing and querying complex data relationships, making it popular for applications such as social networks, recommendation systems, and fraud detection. Examples of graph databases include Neo4j and OrientDB.

Use cases of NoSQL databases

• Overview of NoSQL databases use cases [7]

  NoSQL databases are becoming increasingly popular due to their scalability, flexibility, and performance advantages over traditional relational databases. These databases are used for a variety of purposes, and each type of NoSQL database has its own specific use cases.

• Use cases for document-based NoSQL databases

  Document-based NoSQL databases are commonly used for storing and managing unstructured or semi-structured data such as JSON or XML documents. They are particularly useful for web applications that require dynamic and scalable data models. Examples of use cases for document-based NoSQL databases include content management, e-commerce, and social media platforms. Popular document-based NoSQL databases include MongoDB, and Couchbase.

• Use cases for key-value NoSQL databases

  Key-value NoSQL databases are ideal for managing large volumes of unstructured or semi-structured data, such as sensor data or log files. They are designed to handle large amounts of data and provide fast read and write access. Examples of use cases for key-value NoSQL databases include caching, session management, and real-time analytics. Popular key-value NoSQL databases include Redis and Amazon DynamoDB.

• Use cases for column-family NoSQL databases

  Column-family NoSQL databases are designed for handling large volumes of structured data. They are commonly used in large-scale distributed systems where data needs to be partitioned across multiple nodes. Examples of use cases for column-family NoSQL databases include analytics, content management, and real-time bidding systems. Popular column-family NoSQL databases include Apache Cassandra and HBase.

• Use cases for graph NoSQL database

  Graph NoSQL databases are designed to manage complex relationships between data points, making them particularly useful for social networking, fraud detection, and recommendation systems. Graph databases use nodes and edges to represent data relationships and can perform complex queries quickly and efficiently. Popular graph NoSQL databases include Neo4j and OrientDB.

References

Note

You can search the title in Google Scholar to find the pdf file.

References

[1]

“The NoSQL Know-It-All Compendium”, kdnuggets.com, link

[2]

“NoSQL for Beginners”, kdnuggets.com, link

[3]

Khasawneh, Tariq N., Mahmoud H. AL-Sahlee, and Ali A. Safia. “Sql, newsql, and nosql databases: A comparative survey.” 2020 11th International Conference on Information and Communication Systems (ICICS). IEEE, 2020.

[4]

RUNCEANU, Adrian, et al. “SHORT REVIEW OF SOME NOSQL DATABASES.” Fiability & Durability/Fiabilitate si Durabilitate 1 (2022).

[5]

Gessert, Felix, et al. “NoSQL database systems: a survey and decision guidance.” Computer Science-Research and Development 32 (2017): 353-365.

[6]

“ACID vs. BASE: Comparison of Database Transaction Models”, phoenixnap.com, link

[7]

Lourenço, João Ricardo, et al. “Choosing the right NoSQL database for the job: a quality attribute evaluation.” Journal of Big Data 2.1 (2015): 1-26.