In recent years, Amazon Web Services has experienced rapid growth to become a leading provider of networking, computing and storage solutions. With a 7-year head start in the cloud computing industry and over 20 years of development, AWS is a well-established and trusted provider.
With over 300 services and a customisable pricing model, Amazon Web Services (AWS) is an ideal choice for any business looking to leverage cloud computing for improved efficiency. With such a wide range of options, it is understandable to question the differences between AWS’ SQL and NoSQL databases.
A database is defined, and its significance is discussed.
In order to start, let’s keep things straightforward. Data that can be requested, modified and manipulated which is held in a computer system is known as a database. A database, like a library, requires structure in order to make the most efficient use of its resources. Locating a particular book is far simpler and quicker if the shelves are organised alphabetically instead of being scattered around the room.
A database is a vital component of any programme, as it stores the data necessary for its functioning. In a metaphorical sense, the central processing unit (CPU) can be thought of as the brain and the programme as the cognitive processes it performs. The database, then, can be considered analogous to the memory banks we use to store, recall and update data from the world around us.
Programmes often require a database in order to store and retrieve data. Just as a person with global amnesia cannot recall their past or create new memories, an application cannot store data without a database.
RDMS, or Relational Data Modeling System
In 1970, Edgar Frank Codd first proposed the concept of relational databases (SQL) in his work, “A Relational Model of Data for Large Shared Data Banks”. This theory suggests that data can be stored in tables and linked together.
A table is a data structure, similar to an Excel spreadsheet, which consists of rows and columns. Each cell holds a different piece of information, such as a person’s name, address, email address, age, gender, and so on. Every row of the table contains information on a single individual.
It is possible to join tables together. For example, if there is a table containing transactions, it is possible to link each entry to the individual who completed the transaction. This allows us to have a single record that keeps track of the user’s history, regardless of how many transactions they have completed.
While there is a fair amount of hassle, the primary goal of relational databases is to provide information in a structured, manageable format.
Databases that don’t rely on a relational model
Since their introduction, relational databases have been a widely used approach for software development, though they are far from perfect. This is because it can be difficult to represent real-world data in a tabular format when it is complex, large and unstructured.
Unlike traditional relational databases, NoSQL databases do not use tables to structure their data. One common approach is to store data as key-value pairs or graphs, as this provides greater flexibility and scalability with fewer constraints and structures.
The increasing demand for data storage has seen a surge in the use of non-relational databases in the field of data science. Whilst such solutions are yet to reach the level of popularity seen in SQL-based systems, it is clear that they are becoming increasingly valuable as data collection continues to rise.
AWS Databases
Amazon Web Services (AWS) offers a range of solutions, from basic hosting to advanced data science applications, with comprehensive support for both SQL and NoSQL databases.
Amazon Relational Database Service (RDS) provides an efficient solution for creating, managing and scaling relational databases. The service supports PostgreSQL, MySQL, MariaDB, Oracle Database and Microsoft SQL Server, along with Amazon’s own relational database service, Aurora.
RDS offers a range of tools, accessible either through the dashboard or API, that assist in automating and streamlining operations such as automated failover, backups, disaster recovery, monitoring, access management and performance enhancement.
Aurora is a purpose-built relational database service designed for cloud computing. It boasts 5x the speed of MySQL, full compatibility with MySQL and PostgreSQL, and a far more cost-effective solution than comparable services.
If you opt for Aurora Serverless, you can enjoy the benefit of scalability being managed automatically. Capacity can be started, stopped, increased or decreased depending on your requirements. There are numerous alternatives available, but AWS encourages customers to try and adopt Aurora as their preferred relational solution.
Finally, before we consider any NoSQL solutions, it is important to discuss Amazon Redshift, a data warehouse designed for conducting analysis on relational databases. You do not need to collect data separately in order to analyse it; instead, the analytical tools are incorporated into the service.
Amazon offers more than a dozen non-relational database solutions, such as Elasticache for Memcaches and Elasticache for Redis for managing in-memory databases, Amazon Neptune for managing graph databases, TimeStream for managing time series, and QLDB for managing ledgers. However, we shall focus on their two main products: DocumentDB and DynamoDB.
DocumentDB is an ideal choice for large-scale JSON data handling. It is specifically designed to handle this type of workload, providing seamless integration with Amazon Web Services and an ability to process millions of requests per second without any performance degradation.
DocumentDB’s compatibility with MongoDB is an important advantage. By switching a project’s database endpoint from MongoDB to DocumentDB, users are able to utilise the same API without needing to make any changes to their existing code.
Amazon have developed their own key-value and document database system, DynamoDB. This system has the advantage of scaling and shrinking according to demand, similar to other cloud services, as it does not require dedicated servers.
It is noteworthy that DynamoDB global tables replicate data across multiple AWS Regions, enabling users to access their data from the most geographically convenient server, regardless of their current geographic location.
What service should I select?
If you are starting from scratch and envisage utilising AWS for the foreseeable future, it is recommended that you opt for one of their options. Typically, these are more cost-effective and offer greater support than alternatives.
Services such as RDS and DocumentDB are the ideal choice if you are looking to migrate your database to the cloud or guarantee maximum compatibility. Whichever you choose, you will be making a sound decision.
