Before you can begin to use MongoDB properly, it's important to understand its architectural underpinnings. Key to this understanding is what is often referred to as NoSQL. As the rubric implies, a NoSQL database does not use SQL language. More importantly, this means that a NoSQL database is free from the legacy, two-dimensional restrictions that handcuff the traditional RDBMS.

There is no formal definition of NoSQL, but there are certain common characteristics. First of all, a NoSQL database does not adhere to the relational model. This is evident in that MongoDB has no fixed schema, no tables, columns, or rows. A second consideration is that NoSQL databases are driven by the needs of big data. Accordingly, NoSQL databases tend to be scalable as well as distributed. This aspect is visible in the MongoDB feature called sharding, where shards of the database...

In order fully to understand how to use MongoDB, it is important to discuss the terms used to describe MongoDB data structures: documents, fields, and collections. The following table summarizes these three terms by way of illustration. In this table, we also draw an equivalence between MongoDB and RDBMS (Relational Database Management Systems):

MongoDB	RDBMS	Illustration
Document	Row
Field	Column
Collection	Table

In the preceding example, we see a document that represents information about a member whose name is Smith. Within the document are fields, the example highlighted being cost. Documents that are stored together for a single logical purpose represent a collection. A MongoDB database consists of one or more collections.

So far, if you are used to traditional SQL-based database systems, you will no doubt be pleased to learn there are direct equivalences between basic data structures, such as fields, documents, and so forth. When it comes to data modeling, on the other hand, a radical departure in thought is needed. We will begin this discussion by comparing a MongoDB reference with an embedded document.

Using references (https://docs.mongodb.com/manual/reference/database-references/), it is possible to create a series of related collections in order to establish a normalized (https://docs.mongodb.com/manual/core/data-model-design/#normalized-data-models) data model. In the following diagram, a normalized...

It is now time to put into practice how to create a MongoDB database and collection. The Chapter 3, Using the MongoDB Shell, will show you how to perform operations using the MongoDB shell. For the purposes of this introductory chapter we will use MongoDB Compass (https://www.mongodb.com/products/compass), which is a free graphical utility you can use to perform administration.

To create a MongoDB database and start a collection using MongoDB Compass, proceed as follows:

1. Open MongoDB Compass using the link as appropriate to your operating system. In this illustration we are using Ubuntu Linux, so we launch using the GUI.
2. Click on the plus sign (+) to create a database:
   

3. When the Create Database dialog box appears, enter the name of the database and the name of the first collection. When first creating a database, you cannot have one...

In this chapter, you learned:

• The architectural underpinnings of NoSQL databases and MongoDB, in particular
• MongoDB terms, including databases, collections, documents, and fields
• Data-modeling considerations, document design, and how to avoid SQL-based thinking
• How to create a MongoDB database and start a collection using MongoDB Compass

The next chapter will cover how to use the MongoDB shell to manage the database, including basic CRUD operations.