Understanding the difference between microservices and monolithic architectures can significantly impact an organization’s agility, scalability, and bottom line. Beyond adding flexibility, the right architecture can also enhance system reliability, a factor that directly correlates with customer satisfaction and retention. While it’s true that the choice between microservices and monoliths may seem like a technical decision, it has far-reaching implications that extend to business strategy and customer experience. The architecture you choose can influence the ability to compete in a fast-paced market. Companies that can adapt quickly to changing conditions often have the upper hand.

Microservices vs. Monoliths – Which is right?

Monolithic architecture

A monolithic application is a single, unified unit where all the components are tightly integrated. This architecture offers the benefit of better performance due to components being housed within a single unit. Monolithic applications generally execute tasks more quickly, as they don’t require network calls to communicate between different parts of the application. 

In a traditional monolithic platform-centric or suite-based approach, pricing is relatively well defined for core capabilities. However, it can become incredibly costly to customize and extend to meet the exact requirements. Some estimates place the TCO at 300 to 400 % of the initial equipment’s purchase price. A study published in the IEEE 18th International Symposium on Computational Intelligence and Informatics (CINTI) found that monolithic architecture showed better performance in throughput by 6% when compared to microservices architecture in concurrency testing.

The lack of flexibility in monolithic architectures makes it a challenge to add new features or scale specific components independently. Any change, no matter how minor, necessitates the modification and redeployment of the entire application, which can be both time-consuming and risky. Monolith architecture is usually preferred due to its simplicity and speed. However, the main issue is redeploying the entire application every time a new update was released. These difficulties were the primary motivation for switching from a monolithic to a more efficient architecture. 

Microservices architecture

Contrastingly, a microservices-based application is composed of smaller, independently deployable services, each responsible for a specific piece of functionality. According to a survey of 354 enterprises, 63 % of companies are currently using microservices architectures. Over 60 % reported that their engineering division has adopted or plans to adopt a microservices architecture to achieve faster time to market for new products and services. 

One of the most compelling advantages of this architecture is its scalability. Because each microservice operates independently, you can scale individual services based on demand, making the system as a whole more responsive and cost-effective. However, managing a microservices-based application requires a more sophisticated infrastructure and toolset, including service discovery, load balancing, and inter-service communication mechanisms, to name a few.

In a report by O’Reilly, 61 % of companies have implemented microservices over the past five years, with a 55% “complete success” rate. Statistics show that 68% of companies are already using microservices in production and development. This includes 36% of large companies, 50% of medium companies, and 44% of small companies.

Main challenges expected to face by companies with microservices architectures include lack of visibility into end-to-end business processes that span multiple microservices (59%) , error handling issues at the boundary of two or more microservices (50%), and communication between teams (46%).

Impact on businesses and customers

For applications with straightforward requirements, a monolithic architecture is usually more cost-effective, simpler to develop, deploy, and manage, making it a suitable choice for businesses with limited resources. On the other hand, microservices shine in scenarios involving large, complex applications; offering greater fault tolerance and inherently more adaptable. Companies with the resources to manage the additional complexity will find microservices to be a more fitting solution.

Leveraging a microservices-based architecture can result in a far more efficient use of code and underlying infrastructure. It is not uncommon to experience significant cost savings by as much as 50 % by reducing the amount of infrastructure required to run a given application. For instance, if you have an existing eCommerce application built as a monolith and needs to integrate with a shipping API, instead of implementing the shipping API communication logic within the monolithic application, you can build a separate microservice that handles the shipping API requests. This provides a degree of separation between current monolithic and additional logic.

The power of the hybrid architecture

The debate between microservices and monolithic architectures often presents them as mutually exclusive choices. However, the reality is more nuanced. Companies can, in fact, leverage the strengths of both architectures to create a more scalable, and adaptable system.

This hybrid approach can offer a strategic advantage, allowing companies to capitalize on the strengths of both microservices and monolithic architectures. This enables organizations to build systems that are scalable, reliable, and adaptable, thereby aligning more closely with business objectives and customer expectations.

Improved scalability

One of the most compelling reasons to consider a hybrid approach is scalability. Microservices excel in this area, allowing individual components to be scaled independently to meet demand. Monolithic architectures, conversely, require the entire application to be scaled, which can be both inefficient and costly. By integrating microservices into a predominantly monolithic application, organizations can achieve targeted scalability, thereby optimizing resource allocation and cost.

Better fault isolation

Microservices are designed to operate independently. If one service fails, it doesn’t bring down the entire system. Monolithic architectures are generally more susceptible to system-wide failures, as they operate as a single, unified unit. A hybrid approach can offer the best of both worlds: the fault tolerance of microservices with the efficiency of a monolith, enhancing overall system reliability.

Easier maintenance and updates

Microservices offer the advantage of being easier to update and maintain. Each service can be updated independently, reducing the risk of system-wide disruptions. Monolithic architectures, on the other hand, can make updates a complex, high-stakes operation. A hybrid approach allows for easier, more manageable updates, contributing to long-term system maintainability and agility.

Adapting to changing business requirements

Business needs are never static. Microservices offer the flexibility to quickly add or modify features, giving companies an edge in a dynamic market. Monolithic architectures, while efficient, can be rigid and slow to adapt. By employing both architectures, companies can combine the agility of microservices with the efficiency of a monolith, making it easier to adapt to market demands and customer needs.

Case studies

Amazon’s unexpected return to monoliths

Amazon was an early adopter of microservices architecture, leveraging its benefits to build and scale their vast ecosystem of services. The motivation behind Amazon’s adoption of microservices stemmed from the need for autonomous development teams, rapid deployment cycles, and enhanced fault tolerance.

An internal case study from the Prime Video team at Amazon revealed that moving from a microservices architecture to a monolith reduced their infrastructure cost by over 90 %. The requirement was for a monitoring tool to identify quality issues in “every stream viewed by customers” and therefore needed to be highly scalable, as there are “thousands of concurrent streams”. The team initially created a solution with distributed components orchestrated by AWS Step Functions, a serverless orchestration service based on state machines and tasks. However, it turned out that Step Functions was a bottleneck.

Recently Amazon made headlines by announcing their shift back to monolithic architecture after years of advocating for and practicing microservices. This move sparked considerable debate and raised questions about the suitability of both architectural patterns. The team realized that the distributed approach wasn’t bringing many benefits in their specific use case. So, they decided to re-architect the system into a monolith. The core logic of their media converter and defect detector microservices was sound. So conceptually, their high-level architecture was able to remain the same.

Netflix: Turning disaster into an opportunity for change

Netflix’s decision to switch from a monolithic to a microservices architecture was a strategic move with major improvement. As Netflix’s user base grew rapidly, the need for a system that could scale horizontally became apparent. The monolithic architecture was becoming a bottleneck, unable to handle the increasing traffic and organizational development needs.

In August 2008, Netflix experienced a major outage due to a single mistake that caused massive data corruption, leading to several days of downtime. This incident highlighted the vulnerabilities of a monolithic architecture and the need for a more resilient system.

To eliminate single points of failure and enable horizontal scalability, Netflix decided to migrate to Amazon Web Services (AWS), a public cloud platform. This move was part of a broader strategy to re-architect their systems for the cloud. By transitioning to a microservices architecture allowed Netflix to include many small teams responsible for the end-to-end development, each handling at least one microservice among hundreds already up-and-running on the platform.

Best practices for combining microservices and monolithic architectures

Navigating the complexities of software architecture is no small feat, especially when considering the integration of microservices and monoliths. However, when executed thoughtfully, this hybrid approach can yield significant benefits, marrying the scalability of microservices with the efficiency of monoliths.

Ensure compatibility between microservices and monolithic architectures

First and foremost, the microservices and monoliths within architectures must be designed to work in harmony. This involves establishing clear interfaces and communication protocols. Without this foundational compatibility, the system risks becoming a tangled web of mismatched components, leading to inefficiencies and potential points of failure.

Manage complexity

While the hybrid approach offers numerous advantages, it also introduces an additional layer of complexity. Therefore, it’s essential to have a holistic understanding of the systems. Know how each component, whether it’s a microservice or part of the monolith, fits into the larger ecosystem. A well-documented architecture can serve as a roadmap, aiding in both development and troubleshooting.

Monitor and test

Given that combining architectural styles introduces new variables and potential points of failure, rigorous monitoring and testing are imperative. Employ monitoring tools that can provide real-time analytics and alerts. Additionally, implement comprehensive testing strategies that cover unit, integration, and end-to-end tests to ensure the system’s reliability and performance.

Final thoughts 

Selecting between microservices and monolithic architectures is not a one-size-fits-all decision. Both architectural style comes with its own set of trade-offs, a thoughtful, well-informed choice that yields significant benefits in different cases. These range from cost savings and operational efficiency to enhanced customer experience. Ultimately, the goal is to align architectural strategy with business objectives, thereby maximizing both technological and operational effectiveness. 

Thomas Charneau

December 14, 2023

9 Min read