Choosing Your Data Storage in the Cloud

Azure, like other clouds, offers a wide range of storage services. The first approach that we may consider is defining a scalable set of virtual machines hosted in the cloud where we can implement our custom solutions. For instance, we can create a SQL Server cluster on our cloud-hosted virtual machines to increase reliability and computational power. However, usually, custom architectures are not the optimal solution and do not take full advantage of the opportunities offered by cloud infrastructure. Scalability, fast setup, focus on the business, and security are some of the criteria that you might consider while deciding on your data storage on Azure. To help you with this, many of the platform-as-a-service (PaaS) options for storing data can be a great solution.

Therefore, this chapter will not discuss such custom architectures but will focus mainly on the various PaaS storage offerings that are available in the cloud and on Azure. These...

Technical requirements

This chapter requires that you have the following:

• Visual Studio 2022 free Community edition or better, with all the database tool components installed.
• A free Azure account. The Creating an Azure account subsection in Chapter 1, Understanding the Importance of Software Architecture, explains how to create one.
• For a better development experience, we advise that you also install the local emulator of Cosmos DB, which can be found at https://aka.ms/cosmosdb-emulator.
• You can find the sample code for the chapter at https://github.com/PacktPublishing/Software-Architecture-with-C-Sharp-12-and-.NET-8-4E.

Understanding the different repositories for different purposes

This section describes the functionalities that are offered by the most popular data storage techniques. We will mainly focus on the functional requirements they are able to satisfy. Performance and scaling-out features will be analyzed in the next section, which is dedicated to comparing relational and NoSQL databases.

In Azure, the various offerings can be found by typing product names into the search bar at the top of all Azure portal pages.

The following subsections describe the various kinds of databases that we can use in our C# projects.

Relational databases

These databases are the most common and studied type of storage. They guarantee a high level of service and store an immeasurable amount of data. Dozens of applications have been designed to store data in this kind of database, and we can find them in banks, stores, industries, and so on. When you store data in a relational database, the basic...

Choosing between SQL and NoSQL document-oriented databases

As a software architect, you may consider some aspects of SQL and NoSQL databases to decide the best storage option for you. In many cases, both will be needed. The key point here will surely be how organized your data is and how big the database will become.

In the previous section, we stated that NoSQL document-oriented databases should be preferred by you, as a software architect, when data has almost no predefined structure. They not only keep variable attributes close to their owners, but they also keep some related objects close since they allow related objects to be nested inside properties and collections.

Unstructured data can be represented in relational databases if variable properties of a tuple (t) can be placed in a connected table containing the property name, property value, and the external key of t. However, the problem in this scenario is performance. In fact, property values that belong to a single...

Azure Cosmos DB – an opportunity to manage a multi-continental database

Azure Cosmos DB is Azure’s main NoSQL offering. Azure Cosmos DB has its own interface that is a subset of SQL, but it can be configured with a MongoDB interface, a Table API, or a Cassandra API. It can also be configured as a graph data model that can be queried with Gremlin.

You can find further details about Cosmos DB in the official documentation: https://docs.microsoft.com/en-us/azure/cosmos-db/.

Cosmos DB allows replication for fault tolerance and read scale-out, and replicas can be distributed geographically to optimize communication performance. Moreover, you can specify which data center all the replicas are placed in. The user also has the option to write-enable all the replicas so that writes are immediately available in the geographical area where they are done. Write scale-up is achieved with sharding, which the user can configure by defining which properties to use...

Summary

In this chapter, we looked at the main storage options available in Azure and learned when to use them. Then, we compared relational and NoSQL databases. We pointed out that relational databases offer automatic consistency checking and transaction isolation, but NoSQL databases are cheaper and offer better performance, especially when distributed writes form a high percentage of the average workload.

Then, we described Azure’s main NoSQL option, Cosmos DB, and explained how to configure it and how to connect with a client.

Finally, we learned how to interact with Cosmos DB with Entity Framework Core. Here, we learned how to decide whether to use relational or NoSQL databases for all families of data involved in an application. So, you can choose the kind of data storage that ensures the best compromise between data coherence, speed, and parallel access to data in each of your applications.

In the next chapter, we will learn all about how to interact with...

Questions

1. Is Redis a valid alternative to relational databases?
2. Are NoSQL databases a valid alternative to relational databases?
3. What operation is more difficult to scale out in relational databases?
4. What is the main weakness of NoSQL databases? What is their main advantage?
5. Can you list all Cosmos DB consistency levels?
6. Can we use auto-increment integer keys with Cosmos DB?
7. Which Entity Framework configuration method is used to store an entity inside its related father document?
8. Can nested collections be searched efficiently with Cosmos DB?

Further reading

• The following is a reference to the Gremlin language, which is supported by Cosmos DB: http://tinkerpop.apache.org/docs/current/reference/#graph-traversal-steps.
• The following is a general description of the Cosmos DB graph data model: https://docs.microsoft.com/en-us/azure/cosmos-db/graph-introduction.
• Details on how to use Cosmos DB’s official .NET client can be found at https://docs.microsoft.com/en-us/azure/cosmos-db/sql-api-dotnetcore-get-started. A good introduction to the MvcControlsToolkit.Business.DocumentDB NuGet package that we mentioned in this chapter is the Fast Azure Cosmos DB Development with the DocumentDB Package article contained in Issue 34 of DNCMagazine. This can be downloaded from https://www.dotnetcurry.com/microsoft-azure/aspnet-core-cosmos-db-documentdb.

Learn more on Discord

To join the Discord community for this book – where you can share feedback, ask questions to the authors, and learn about...