A Guide to Document Databases
Document Databases Guide: Benefits and Use Cases
Among the many database options available to developers, document databases make it easier for developers to store and query data by using the same document-model format they use in their application code. It uses state-of-the-art design methods to create a more agile way to store and retrieve data. It’s an improvement over the traditional relational database (RDB) design used in the early days of computing, before the real brunt of the “big data revolution.”
In this article, you will learn about document databases, how they work, some common use cases, and some popular document databases that you can use in your own projects.
What is a document database?
A document database or document-oriented database is a new type of noSQL database structure. Along with other types of noSQL database designs, a document database represents the use of a type of object-oriented engineering in database technologies, or some similar innovation that improves how data is used. The data is typically stored in a JavaScript Object Notation or “JSON” format.
Experts have described this JSON format as a “readable” way to store key-value pairs and create the noSQL design that leads to certain data storage and retrieval benefits.
Compared to relational databases, noSQL databases, like document databases, have abundant use cases in enterprise and administrative data governance strategies. While methodologies vary a bit, the document database aims to accomplish the objective of using the same data model that developers use in their application code.
When to use document databases - use cases
Many businesses use a document database to keep structured or semi-structured data in an easy-to-retrieve environment. As we’ll talk about later, they may also use these formats to make it easier to input raw data into a more structured format. Queries that use the noSQL design allow users to get specific results based on the agile nature of the document database storage setup.
For example, companies may use a document database to store customer identifiers and other customer data. They may retrieve that customer data for predictive analytics, as part of the sales funnel operation, or to use in standalone customer relationship management software systems or other business purposes. In any case, the document database is the central holder of information for this and other types of business intelligence.
Document databases can also be used for inventory management. The key-value pairs and data structures would be linked to physical products in a warehouse, and a company would retrieve this information to move product or otherwise handle inventory. If the document database is open source they can be used for edge computing workloads like collecting sensor data from IoT devices and then sending data from the edge back to a cloud instance for analysis.
As a third example, a document database could be used for product development, in which key-value pairs and a data storage system identify product characteristics and attributes, helping teams bring products to market.
Because of their flexible data model, document databases can also be extended with different types of indexes to support features like full text search that is normally done by dedicated search engine databases.
Benefits of NoSQL document databases
Experts have identified various benefits to using a new document database or noSQL design. First of all, these databases often provide faster development cycles. Another benefit of document databases or noSQL databases is better querying. As mentioned above, these systems make data retrieval more efficient in various ways.
Document databases can also accommodate rich data structures. To understand this, it’s helpful to get a good understanding of JSON format and what that entails.
Beyond these, one of the other key benefits of NoSQL document databases in terms of querying is that some of these noSQL types of designs can scale horizontally.
In general, “horizontal scaling” refers to creating more independent modules to tackle parts of a complex data task. For instance, horizontal scaling in hardware means adding more machines, but vertical scaling simply means adding more power to one machine. In database management, horizontal scaling involves splitting data sets into multiple tables, objects, or modules, so that querying can be done better or more efficiently. There are various tradeoffs in database performance when it comes to sharding which are described in CAP theorem.
Another benefit of noSQL databases involves more accessible development. An intuitive way to understand this is that developers feel most comfortable with technologies that are transparent to them and ones that they understand. Since many developers are trained on new NoSQL database designs, they are accustomed to working through these structures.
JSON format and data types
Javascript Object Notation or JSON format is often discussed in relation to the semantic web or new mapping technologies. Parties like the IETF or the W3C that work on innovating the web often talk about how JSON can help facilitate a “read-write-function” web that can do more than just passively contain data.
But JSON is also useful in the database world, as evidenced by its use in a document database design.
The JSON system has several main data types:
- Number
- String
- Boolean
- Array
- Object
- Null (pointer)
The object data type is made up of name-value pairs that show the characteristics and attributes of a digital or virtual object. In programming, data objects allow for certain types of coding outcomes related to manipulating the objects in question, rather than working through a linear codebase (as was the custom with early programming languages like BASIC and FORTRAN.)
So with that in mind, JSON brings a profound sense of object-oriented programming to database design.
For example, the JSON format might include name-value pairs for strings with a customer’s first and last name and/or middle initial. It might include Boolean attributes to show whether a customer has purchased a given product. It might have an array that describes a sequence of held data for some customer record or other setups for using these data types to record relevant information for an enterprise.
In all this, NoSQL document databases transcend the traditional relational database design. In a typical relational database, data was stored according to its location in a static table. NoSQL databases are different, however, with data held according to its place in an object-oriented model. All the little tags or pieces of data that would be in rows and columns of a static database table are in object identifiers instead and in the JSON format mentioned above.
Document database data model
Essentially, while creating a document-oriented database, engineers write a JSON-formatted script that describes the object and how it is stored and handled. In other words, instead of entering data into a table that looks like a Microsoft Excel spreadsheet, the schema for the database object will include lines that look like code or script, with the identifiers written in JSON format; for example, “name” or Boolean “T.”
Then there’s a question of how to use NoSQL document databases to a company’s advantage. When data is entered into a noSQL environment (in appropriate use cases), engineers can use complex queries to pull out data for business intelligence.
As an example of good-use cases, database administrators can build queries that pull out information on all customers over a certain age, or all customers of a certain gender or location area. Or in inventory management, they may pull up all products that have proximity to a given location, or all products that are not obsolete or discontinued.
The way developers build queries determines the search result. The way the database is built determines how easy it will be to retrieve the data, and how well systems will scale under pressure or demand.
Why use NoSQL document databases?
As mentioned above with use cases, companies often use document databases because of the agile nature of their data handling capabilities. However, there’s another major benefit to using document databases or other noSQL designs. It has to do with the use of raw or unstructured data. Many companies face challenges with raw or unstructured data and funneling it into a database design. For instance, suppose a company gets letters it wants to mine for data, and they’re stored in digital format. The letters themselves don’t have those static relational database tables built in. Instead, they have a narrative where the identifiers are hidden in the text. Therefore, to get that important data and use it, the company must have some consistent, universal way to mine it from a letter and get it into a context where it can be queried. Typically, you can’t query something out of a letter format.
Over time, the ability to take raw, unstructured data and add structure to it became a valuable part of what database engineers do. So when experts say a noSQL design “facilitates rich data structures,” they mean, in part, that it puts data in a situation or environment where people can conduct richer queries. On the other hand, you could describe it as a system in which the JSON format is simply what makes the data more structured than it otherwise would be.
A new document database design in a JSON format or similar object format helps facilitate the mining of raw information, aggregating it, and generally governing it according to business principles. Another way to think about this is that having a database in a new JSON format may make it easier for companies to upgrade over time.
One of the biggest challenges in working with data systems is the chore of manual data entry or other manual data work. If data is in a static traditional database table, there are only certain use cases that work and ways to retrieve that data. Migrating the data from a legacy system to a new one might take a lot of labor-intensive manual data entry work.
By contrast, when the data is in a new noSQL format, there will be different ways to retrieve it or migrate it without manual data entry.
Document database examples
AWS DynamoDB
Amazon DocumentDB is a Database as a Service(DBaas) provided by Amazon Web service that supports document data structures that allows you to store and query rich documents in your application. Amazon DocumentDB has some compatibility with MongoDB version 3.6 and version 4.0.
MongoDB
MongoDB is a document-oriented NoSQL database used for high-volume data storage. MongoDB uses collections and documents, unlike traditional relational databases that use tables and rows. Documents consist of key-value pairs, the basic unit of data in MongoDB, whereas collections contain sets of documents and functions, the equivalent of relational database tables.
Couchbase
Couchbase Server is an open-source, distributed multi-model NoSQL document-oriented database software package optimized for interactive applications. These applications may serve many concurrent users by creating, storing, retrieving, aggregating, manipulating, and presenting data.
FAQs
What are some other examples of a noSQL database?
In addition to a document database, some other examples of noSQL database design include a column database and a key-value database setup.
What is JSON used for?
Although JSON is used for databases, it is also used to create the new semantic web and other types of mapping systems.
Why should a company use document databases?
Better querying and more efficient storage are some benefits of using a document database design with good enterprise use cases. Companies may also use them to conform to modern best practices in terms of data storage and data governance.
Are document databases a type of relational database?
Typically, document databases are not relational databases, but a new type of noSQL database design. A relational database refers to a database in which data is held according to its place in a structured table with rows and columns. Document databases, on the other hand, adopt a JSON or object format and data method model.
What is the difference between document databases and data warehouses?
A document database facilitates storing and query data by using the same document-model format that developers use in their application code. A data warehouse is a data management system that supports business intelligence, and helps you make more insightful decisions about your business.
