Databases architecture design

Azure Database for MariaDB
Azure Database Migration service
Azure Database for MySQL
Azure Database for PostgreSQL

This article provides an overview of the Azure database solutions described in Azure Architecture Center.

Diagram that contrasts relational database management system and big data solutions.

Azure Database solutions include traditional relational database management systems (RDBMS and OLTP), big data and analytics workloads (including OLAP), and NoSQL workloads.

RDBMS workloads include online transaction processing (OLTP) and online analytical processing (OLAP). Data from multiple sources in the organization can be consolidated into a data warehouse. You can use an extract, transform, and load (ETL) or extract, load and transform (ELT) process to move and transform the source data. For more information on RDBMS Databases, see Explore Relational Databases in Azure.

A big data architecture is designed to handle the ingestion, processing, and analysis of large or complex data. Big data solutions typically involve a large amount of relational and nonrelational data, which traditional RDBMS systems aren't well suited to store. These typically involve solutions such as Data Lakes, Delta Lakes, and lakehouses. See more information in Analytics Architecture Design.

NoSQL databases are interchangeably referred to as "nonrelational", "NoSQL DBs", or "non-SQL" to highlight the fact that they can handle huge volumes of rapidly changing, unstructured data. They do not store data in tables, rows, and columns, like (SQL) databases. For more information on No SQL DBs Databases, see What are NoSQL Databases? and NoSQL Data.

This article provides resources to learn about Azure databases. It outlines paths to implement the architectures that meet your needs, and best practices to keep in mind as you design your solutions.

There are many architectures for you to draw from to address your database needs. We also provide solution ideas for you to build on, which include links to all the components you need.

Learn about databases on Azure

As you start thinking about possible architectures for your solution, it's vital that you choose the correct data store. If you're new to databases on Azure, the best place to start is Microsoft Learn. This free online platform provides videos and tutorials for hands-on learning. Microsoft Learn offers learning paths that are based on your job role, such as developer or data analyst.

You can start with a general description of the different databases in Azure and their use. You can also browse Azure data modules and Choose a data storage approach in Azure to get an understanding of your choices in Azure data solutions and learn why some solutions are recommended in specific scenarios.

Here are some Learn modules you might find useful:

Path to production

To find options helpful for dealing with relational data, consider these resources:

  • To learn about resources for gathering data from multiple sources and how to and apply data transformations within the data pipelines, see Analytics in Azure.
  • To learn about Online analytical processing (OLAP), which organizes large business databases and supports complex analysis, see Online analytical processing.
  • Online transaction processing systems record business interactions as they occur. For more information, see Online transaction processing (OLTP).

A non-relational database doesn't use the tabular schema of rows and columns. For more information, see Non-relational data and NoSQL.

To learn about data lakes, which hold a large amount of data in its native, raw format, see Data lakes.

A big data architecture can handle ingestion, processing, and analysis of data that is too large or too complex for traditional database systems.

A hybrid cloud is an IT environment that combines public cloud and on-premises datacenters. For more information, see Extend on-premises data solutions to the cloud or consider Azure Arc combined with Azure databases.

Azure Cosmos DB is a fully managed NoSQL database service for modern app development. For more information, see Azure Cosmos DB resource model.

To learn about the options for transferring data to and from Azure, see Transfer data to and from Azure.

Best practices

Review these best practices when designing your solutions.

Best practices Description
Data management patterns Data management is the key element of cloud applications. It influences most quality attributes.
Transactional Outbox pattern with Azure Cosmos DB Learn how to use the Transactional Outbox pattern for reliable messaging and guaranteed delivery of events.
Distribute your data globally with Azure Cosmos DB To achieve low latency and high availability, some applications need to be deployed in datacenters that are close to their users.
Security in Azure Cosmos DB Security best practices help prevent, detect, and respond to database breaches.
Continuous backup with point-in-time restore in Azure Cosmos DB Learn about Azure Cosmos DB point-in-time restore feature.
Achieve high availability with Azure Cosmos DB Azure Cosmos DB provides multiple features and configuration options to achieve high availability.
High availability for Azure SQL Database and SQL Managed Instance The database shouldn't be a single point of failure in your architecture.

Technology choices

There are many options for technologies to use with Azure Databases. These articles help you choose the best technologies for your needs.

Stay current with databases

Refer to Azure updates to keep current with Azure Databases technology.

These architectures use database technologies.

Here are some other resources:

Example solutions

These solution ideas are some of the example approaches that you can adapt to your needs.

Similar database products

If you're familiar with Amazon Web Services (AWS) or Google Cloud, refer to the following comparisons: