Database Basics

A NoSQL database is a type of database which uses different data models than the traditional relational model used with SQL.

A key-value NoSQL database uses individual records consisting of:

• Name, which is the key
• Data, which is the value

The data is assigned to a name when stored.

A document-based NoSQL database stores objects which contain data in standard encodings such as JSON, YAML, or XML.

A graph-based NoSQL database connects many data objects (nodes) together through relationships (edges).

This type of database combines the querying of relational databases with NoSQL flexibility and scalability.

A column-oriented NoSQL database groups data into columns instead of using the rows of a traditional database.

The columns can consist of different sizes and stored data types.

In software engineering, databases are systems that store, modify, and access collections of information electronically.

Databases can store a variety of data types including strings, numeric data, booleans, dates and times, and files of various extensions.

A database management system (DBMS) is software designed to manage a database. Database management systems allow developers to communicate with a database, via code or a graphical user interface, to perform database operations and administrative actions such as user management.

Relational databases represent data in terms of its relationship to other data. These types of databases organize data into tables and establish relationships between them. The most common types of relationships between tables are: “one-to-one”, “one-to-many”, and “many-to-many”.

Let’s use books to illustrate examples of database relationships:

• The relationship between a book and it’s ISBN number can be represented as a “one-to-one” relationship. Each book has one ISBN number, and each ISBN number belongs to one book.
• The relationship between publisher and book titles can be represented as a “one-to-many” relationship. Each publisher has published many books but each book can only be released by one publisher.
• The relationship between book titles and authors can be represented as “many-to-many”. Each author may have written many books and each book can have multiple authors.

A database schema is a blueprint that outlines how a database’s data will be structured and organized. It is always a good idea to have a database schema in place before actually inserting data.

Relational databases typically follow predefined schemas. Developers outline the tables and their associated relationships before inserting any data.

Two considerable disadvantages of working with relational databases are:

• They can be costly to set up, scale, and maintain.
• The organization of data into tables consumes a great deal of space which can negatively impact cost and performance.

Considerable advantages of NoSQL databases are:

• They are relatively inexpensive to maintain.
• The flexible nature of their schema makes them highly adaptable to an application’s changing needs.

Due to their flexible, unstructured nature, NoSQL databases can make accuracy a challenge to maintain.