Chapter 2: Solution Architecture

In this chapter, we will review the broad landscape of Azure, looking at it through the eyes of a solution architect. More specifically, we will cover the following topics:

• The solution architecture map
• Zooming in on the different workload types
• Zooming in on containerization
• Solution architecture use case

This chapter should give you the keys to designing any type of solution with Azure. After reading this chapter, you should be able to find a workable solution, whatever use case you are confronted with. However, you still need to refer to the other maps of this book in order to dig deeper into additional in-depth information.

Technical requirements

Admittedly, pure architecture is not typically hands-on. However, by the end of the chapter, we will have walked you through the design of a reference architecture and its corresponding code implementation. To open the physical reference architecture file and test the code, you will need the following:

• Visual Studio 2019 to run and debug the provided .NET Core program locally. It makes use of a local Azure Storage emulator and the Azure Functions runtime.
• Fiddler, Postman, or any HTTP tool you want, to send HTTP requests to our program.
• Microsoft Visio to open the diagrams. We also provide the corresponding PNG files.

The full code files are available at https://github.com/PacktPublishing/The-Azure-Cloud-Native-Architecture-Mapbook/tree/master/Chapter02.

The CiA videos for this book can be viewed at: http://bit.ly/3pp9vIH

The solution architecture map

The purpose of the solution architecture map is to help solution architects find their way in Azure. We defined the duties of a solution architect in the previous chapter, which is typically an architect who assembles the different building blocks and services of a solution while considering the non-functional requirements. Solution architects engage with their specialized peers, who are often application, infrastructure, and security architects.

This solution architecture map, illustrated in Figure 2.1, does not alone regroup all the Azure services, nor does it cover all the possible use cases, but it can be used as a source of inspiration. Depending on the extent to which you want to practice solution architecture, we suggest that you carefully read every chapter of this book, to be able to consider the end-to-end aspects of a solution:

Figure 2.1 – The solution architecture map

Important note

To see the full...

Zooming in on the different workload types

In the following sections, we will take a closer look at some typical use cases and cross-cutting concerns that solution architects deal with (we refer to these workload types as categories). Separating the map into smaller sections makes it easier to digest and understand. Let's begin by looking at Systems of Engagement (SoE).

Understanding systems of engagement

This category regroups all the services that belong to the frontend layer, such as user interfaces, mobile apps, and every channel that helps engage with first and third parties:

Figure 2.2 – The SoE category

In Figure 2.2, the SoE category map includes four top-level groups:

• REAL-TIME WEB
• MOBILE
• UI
• MISCELLANEOUS

The miscellaneous group includes CDN and Media Services. Azure Content Delivery Network (CDN) helps you reduce your load times for more responsive customer interactions. If you are working with rich...

Zooming in on containerization

Containers are everywhere and on everyone's lips! In the following sections, we will explore Azure's container offering.

The Azure platform supports different flavors, which range from single-container support to full orchestrators. The solution architecture map already describes the different high-level use cases. Therefore, let's zoom deeper with a richer map that specifically targets containers (see Figure 2.16):

Figure 2.16 – Zoom in on containers

Microservices are one of the top use cases for running container orchestrators, such as AKS. Service Fabric Mesh has been designed at its core to deal with microservice architectures, by providing both stateless and stateful services. However, over the past 2 years, the adoption of Kubernetes worldwide has grown so fast that Microsoft's focus has now shifted to AKS. To bring statefulness (and more) to services in your AKS cluster, you can leverage...

Solution architecture use case

In the following sections, we will focus on a concrete use case (description follows). Our objective is to help you build a reference architecture, by using the map as your Azure compass to find the relevant options for a given business scenario.

Looking at a business scenario

Since we decided to zoom in a little more on containerization in this chapter, we will demonstrate one possible usage of containers in a workflow-like scenario.

For our example, we will consider the following requirements:

Contoso needs a configurable workflow tool that allows you to orchestrate multiple resource-intensive tasks. Each task must launch large datasets to perform in-memory calculations. For some reason, the datasets cannot be chunked into smaller pieces, which means that memory contention could quickly become an issue under a high load. A single task may take between a few minutes to an hour to complete. Workflows are completely unattended (no human interaction...

Summary

In this chapter, we described the solution architecture map and its different classification categories, which are SoE, SoR, SoI, and systems of interaction (IPaaS). This categorization, commonly used in architecture, makes it easier to dispatch services. We provided explanations and extra-focused maps, which helped further refine the alternatives.

We also emphasized the importance of cross-cutting concerns that apply to every solution, and we discussed which concerns should be considered by solution architects. Remember that it might be too challenging to address all of the concerns on day one of your cloud journey. It is, therefore, interesting to think of different maturity levels, and how we would put them on a roadmap to manage our stakeholders' expectations.

Next, we highlighted the containerization components, with a focused map that depicted the container landscape of Azure. We also considered other dimensions, such as cost, complexity, and the level of residual...