Golang Microservices: To Go or Not to Go

Microservices may not be the fastest way to build software, but in the long run, they are a smart choice. More cost-effective and agile than monolithic architectures, microservices can be developed and tested simultaneously and almost independently by small teams. Microservices architecture also enables easier deployment and troubleshooting. That's not to mention that microservices are the right choice for scale-up projects.

At MadAppGang, we actively apply microservice architecture for various projects (for example, Evergen), and we know that to get the most out of it, you need to choose the right technology stack. In light of our experience, we'd like to share some thoughts about why Golang microservices could be a better choice for your project compared with those built with Java, Python, or other languages. 

Let's first examine what microservices architecture is, how it differs from monolithic architecture, and which projects can benefit from each. We will then discuss the main criteria for choosing a microservices language, and look at whether creating microservices in Golang is worth the effort.

O’Reilly’s Microservices Adoption in 2020 report highlights the increased popularity of microservices and the successes of companies that adopted this architecture. According to the study — which is based on a survey of 1,500 software engineers, technical architects, and decision-makers — 77% of businesses have adopted microservices and 92% of these reported a high level of satisfaction with their choice. 

In 2021, the trend for microservices is thriving. But are microservices right for your project? This question can be answered by explaining what microservices are and how they are different from monolithic architecture:

• Microservices or microservice architecture is a software development approach that allows the structuring of an application as a collection of small services around a business domain. In other words, the software is composed of small independent services that communicate through well-defined application programming interfaces (APIs). 

• Monolithic architecture is the complete opposite of microservices. With this method, the entire software comprises a single piece that is designed to be self-contained. All the components of a monolithic program are interconnected and interdependent.

Source: AEM Tutorials for Beginners

Monolithic architecture is a common method of building apps. It enables relatively easy implementation, development, and software testing. So, if you choose this method, you'll get the final product faster and for a better price. Monolithic architecture is a good choice for building a proof of concept or for companies at a founding stage. 

However, when your application requires advanced scalability, microservices architecture is better. If you choose a monolithic app, you’ll need to alter the whole stack to make a slight change. This is time-consuming, complicated, and risky as a single point of failure can crash the entire program. A microservices architecture means the ability to change any function independently — and avoid all risk and extra expenses.

Comparison of monolithic and microservice architectures. Source: IBM

Once you choose microservices for your project, the next step is to select a technology that allows you to get the most out of this architecture. In theory, any programming language is suitable. However, to avoid certain issues and maximise microservices benefits, the technology you use should meet the following requirements:

• Modelled around a business domain
• High readability
• Decentralised components
• Supports continuous integration 
• Supports automation
• Independent deployment 

In addition, you should also take into account the following tips on choosing a suitable microservices language:

• A language that simplifies coding is a better choice. Microservice developers try to avoid C languages as they're quite complicated.

• Consider speed. It's best to pick a language that responds quickly to requests and is able to handle a large amount of traffic. 

Python, Java, Ruby and Go all qualify as each boasts a good level of readability and compatibility, while offering frameworks for the development of microservices architecture. However, more and more companies are choosing Golang over other languages for microservices development.

For example, The Economist chose Go language for microservice development when moving from a monolith to a microservice architecture; primarily because of its excellent alignment with the platform's architecture. Plus, Go's syntax is relatively simple, so it was easy to start writing functional code right away. A quick win for the team during a period of transition. 

There's no better marriage than Go and microservices. The Golang language meets and exceeds all the requirements outlined in the previous section. Here are some other reasons your business will benefit from building microservices with Golang:

Go is a simple language that’s easy to learn, probably the greatest characteristic of a programming language in terms of business processes. Golang's syntax is relatively small, and it has been in use for a long time without major changes. As there’s no need to learn new paradigms or syntax, code written with Go is transparent and easy to understand. 

This makes Go an ideal choice for backends, microservices, cloud-computing systems, and more. Additionally, its characteristics help mitigate the drawbacks of microservices: writing communication layers, handling service errors, and dealing with distributed architecture. 

A Golang microservices architecture ensures developers don't waste much time trying to read someone else's code and will get up to speed faster. Even if you need to change your development team, you will still be able to maintain and update your software hassle-free. 

Golang compiles faster than most other languages, saving time for developers and money for business owners. These are some of the reasons: 

• It has a faster compiler compared to other languages such as C or C++.
• Directed acyclic graphs and the absence of cycles make it possible to compile packages separately and in parallel. 
• A compiled Go package's object file contains information about the package itself and its dependencies. To compile a package, the compiler must read an object file for each import, but it doesn't need to examine any other files.

Go’s runtime performance is also excellent. Here are several benchmarks to illustrate that in many cases Golang outperforms popular programming languages such as Java and Python:

Go vs Java and Go vs Python benchmarks. Source: Benchmarks Game

As Golang matures and the community grows, more useful add-ons, libraries, frameworks, and toolkits for various purposes are appearing. Of course, there are instruments that help build Go microservices as well: 

• Go Micro – The most popular remote procedure call (RPC) framework today. It provides load balancing, synchronous and asynchronous communication, message encoding, service discovery, and RPC client/server packages. Go Micro allows developers to build a three-layered stack architecture as follows:

And importantly, Go Micro’s Sidecar features mean the ability to integrate services written in other languages. This allows you to mix and match technologies when building your Golang microservice architecture.

• Go Kit – A library that provides solutions for most operational and infrastructure concerns. It has packages for service discovery, authentication, tracing, transport, metrics, and more. If you use Go Kit, your Go microservice architecture will look like this:

• Gizmo – A microservice toolkit that was shared by the New York Times on GitHub. It provides tools to put together server and pubsub daemons, including higher-level building components such as config and pubsub packages.

• Kite – The Kite framework has RPC client and server packages. While useful, developers don't tend to use it much. 

The problem is that created services are automatically registered with Kontrol, a service discovery system written in Kite. It means that Kite's microservices are well integrated into Kite's environment. But customising the Kite microservice if you need to connect it to another service discovery system takes time. 

Microservices architecture is a good fit for large projects that will be scaled in the future. If that’s the case with your project, go for it, just remember that you'll need the right technology stack to get the most out of this type of architecture. The Go language is an excellent choice for this purpose. Not only does it meet all the criteria for a microservices stack, but it offers huge business advantages.

Golang microservices are highly readable, hence, easy to maintain. If you need to make changes to your team over time, it won't put the entire business at risk. Besides, there is a range of other benefits that Go provides. At MadAppGang, we've been using Golang for a long time and suggest it to most of our customers. And we'll probably suggest it to you, too. Contact us, and let's discuss what architecture and language fit your idea and project best.