Reactive System: A Better Architecture for Modern Software

Reading Time: 3 minutes

Today, most real-time software are built as a reactive system. However, they have evolved over time. A few years back, software were way different from what they are today. They were different not only in terms of features and functionalities they offer but also in the way they were designed. For instance, earlier software dealt with limited data which was mostly at rest- meaning it did not change very rapidly. In addition to that, they had maintenance windows for which it could be shut down wholly or partially.

However, with the advancement in time users’ expectations of software have changed. They no longer expect software or service to be unavailable. For instance, imagine what would happen to a developer if GitHub, Google, or any other tool or service is down. It can bring a halt to the entire workflow. Secondly, software now deals with petabytes of data that is not at rest anymore. It changes very rapidly and software applications must quickly adapt to consume the most recent data. How long it takes to serve the request is also an important consideration in determining the usability of the software. Modern software solutions should meet these expectations. A Reactive Architecture can help software developers to design and develop such software.

What is Reactive Architecture?

A reactive system is an architectural style that lets the system react to changes in its environment. It allows multiple individual applications to blend together to form a complete system. It brings together the set of principles detailed in the Reactive Manifesto. These principles are vital to make the system capable of responding to users’ queries in a timely fashion and continue to do so in case of failure or overload.

The Reactive Manifesto is authored by Jonas Boner, Dave Farley, Roland Kuhn and Martin Thompson

Principles of Reactive System

A reactive system follows four basic principles. They are:

Image showing four Reactive Principles and relationship between them.

Message Driven:

Message Driven implies that all components interact with each other through messages. These messages are asynchronous in nature. The asynchronous communication creates a temporal boundary between the components, making them decoupled in space and time. This decoupling forms the basis of other reactive principles.


Elasticity enables the system to scale horizontally and is the result of space decoupling. It means that the individual component(s) of the system should have the ability to efficiently deal with peak load. An elastic component is auto-scalable in response to changes in load. In other words, when the system experiences a peak load, it launches new instances and later gets rid of them.


Resilience means that the software should be able to respond to a failure. In response to failure, it should not completely break down. It should continue to perform to the best of its ability even if a component is failing.


Being responsive is the most important aspect of the software. Software is responsive if it can timely respond to users’ requests. This should hold true even in case of failure, breakdown, or high load. Responsiveness is a combined outcome of elasticity and resiliency of a system.

How Reactive Principles Build a Reactive System?

Responsiveness is the cornerstone of usability. Hence, the aim is to create software that is highly responsive. The message-driven principle allows a system to be elastic and resilient as it enables the system to handle non-blocking failures. Asynchronous interactions with messages allow a system to handle failure locally because components do not actively wait for a response. Hence, the failure of one component does not affect another component. Due to elasticity, the extreme load conditions are managed and the system remains stable under high load. With these two features, the system becomes capable of responding to changes in load and component failure. By tuning itself accordingly, the system continues to perform to the best of its ability. Such a system effectively meets the ideal expectations of the user.

Further Reading:

Official Reactive Manifesto: