Microservices architecture vs. service-oriented architecture (SOA)

What developer teams need to understand about the complexities of microservices, which use small autonomous modules of code and API protocols, and SOA, which has reusable software components with a middleware communication tool.

December 1, 2021

In today’s world of mobile apps and responsive web functions, agile and responsive code is a must. But developer teams often face a tough question of the best way to develop and deploy this: whether to use a microservices architecture or service-oriented architecture (SOA)?

While there are many similarities in these two approaches, there are some glaring key differences that are crucial to keep in mind when deciding on what’s right for you.

What is a service-oriented architecture (SOA)?

The concept of a Service-oriented architecture (SOA) emerged as an approach to software development for applications in the late 1990s and gained popularity in the mid-2000s.  SOA is a way to manage dependencies between systems by improving interoperability. This infrastructure approach breaks the application down into reusable software components, or services, which are tied to business domain and business logic.

Each component is a small, discrete unit of software (with separate code, but usually sharing a relational database) designed for a specific business functionality, such as checking a customer’s credit, calculating a monthly loan payment, or processing a mortgage application. Each software component can range in size, from small application services to large enterprise services. The communication can involve either straightforward data passing, or two or more services coordinating on some agreed-upon activity. The software components can also be published in a registry, which stores contracts from service providers and provides them to interested service consumers. This lets developer teams find and reuse them to build new applications or business processes.

The messaging itself for SOA architectures needs to be flexible to enable so much communication. The communication mechanism can be implemented with technologies such as Web Services, such as: a Web Service Definition Language (WSDL), which is a standard tag structure based on xml (extensible markup language), or SOAP (Simple Object Access Protocol); messaging such as ActiveMQ, JMS, RabbitMQ; or RESTful HTTP, with Representational state transfer (REST). A middleware tool like an enterprise service bus (ESB) can act as a managed message system, to provide routing, data transformation, translation upon a service consumer's request and event interpretation.

Benefits and challenges of service-oriented architecture (SOA)

With SOA, advantages include how separate developer teams can work on each individual service in a different code language, so that complex applications can be delivered and updated quickly, frequently, and reliably. In addition, application developers don’t have to redevelop or duplicate existing functionality when they’re coding. Overall, the system can be relatively easily deployed, maintained, tested, and managed with tools.

Some challenges include management, testing, and security issues. Because the services within an SOA can be so diverse, communication among them can raise thousands of messages across a network in several directions. Managing these messages can be complicated. In addition, security issues can arise because services share the same data set.

What is Microservices Architecture?

In comparison to SOA, microservices are an even more modern approach to development. Microservices architecture is a popular software system these days in which an application’s services are built as small autonomous modules according to business capabilities (the end use). It’s usually cloud-based.

These independent services all have bounded contexts and function separately from each other, but they’re loosely coupled, and they communicate with other services via an API. The common protocols that microservices use are Representational State Transfers (REST) and/or simple messaging such as Java Messaging Service (JMS).

They’re often managed and deployed in container environments, like a Docker container. Many microservices are also usually built with a continuous delivery mode so developers and DevOps teams can build and deploy features consistently, with agility and scalability. Many companies from Google to Amazon to Netflix to Twitter to PayPal to Microsoft are employing microservices architecture.

Benefits and challenges of Microservices Architecture

With a microservices architecture, applications are split into a collection of smaller different services, so advantages include how they are easier to build and maintain for developer teams. Teams can use different programming languages, databases and software environments for the microservices. Developers also stay more productive by working with microservices because they’re not duplicating work. In addition, because so many teams work on each service, each can have an authentication point for users to make sure the right developer is working on the right microservice.

The reliance on APIs also keeps data in each module secure and confidential. APIs as a communication protocol are relatively easy to manage.

In addition, this approach, when combined with testing, helps easily identify errors in code or runtime. For example, if one microservice takes too much memory, has a higher processor load, and/or lags in speed, testing can identify the problematic service.

However, a disadvantage of microservices can be that so many microservices can lead to higher complexity. That can lead to increased latency and performance issues. In addition, testing can be difficult and/or a lengthy process because testing must be completed at the microservice unit-level, at the communication interface level, and across the entire microservices system.

Microservices vs. SOA: what are the key differences?

If SOA and microservices architecture sound similar, that’s because they are. They’re both service-oriented architectures, designed to solve the problems that come with the single-application build approach of monolithic applications by decoupling functionalities within it. Monolithic architecture is designed as a single unit, that has traditionally been the standard in development.

Both microservices and SOA services allow for automation in processes that can make teams more agile and productive compared to monolithic architecture.

However, key differences remain. As Martin Fowler, a pioneer in the world of microservices, once said we should think about SOA as a superset of microservices.

SOA tends to be for larger and more diverse environments because services are different sizes and messaging protocols can be different, but integrated via the enterprise service bus. Microservices tend to be used for smaller environments such as mobile and responsive applications. SOA is considered more “coarse-grained” and microservices are more finer grained with more granularity.

In terms of new deployments or code changes, because microservices have such small modules, each can be modified without affecting other modules. But SOA, because some services have interdependencies, you might have to also upgrade the other services when performing a general upgrade.

In terms of communication, SOA services rely on communication and data storage outside the services. However, microservices communicate APIs, which are language-agnostic protocols, typically over a network, and feature many remote calls. This means microservices typically have faster communication than SOA.

SOA and microservices vs. monolithic architecture

SOA and microservices are considered to be more modern approaches than monolithic architecture. They’re considered to be ways to “future-proof” applications because new modules can be added and distinct modules can be updated, as opposed to monolithic structures which need to be updated as a whole.

Although it might be simpler to develop, because monolithic architecture is a large piece of code, with tightly coupled and ordered components, it can be harder to upgrade and scale.

So which architectural style is right for your business?

One software architecture is not necessarily better than the other. It’s crucial for businesses to choose the right architecture pattern for their business' use case and for their DevOps strategy.

Developer teams can collaborate on both SOA and microservices architecture well. SOA’s prioritization of service reusability is helpful, while microservices architecture’s relaxed governance allows teams to develop their own workflow and methodology for each microservice.

SOA is great for large, complex enterprise applications. Microservices architecture is ideal for smaller and well-defined web-based applications, such as mobile and web apps, where developers have greater control in application development.

Businesses must reflect on their unique needs and the problems that they’re trying to solve in order to pick the type of software architecture that suits them and their use case best.

What's driving urgency for IDPs?