This article explores the way the microservice architecture can be used together with Java EE using the new KumuluzEE framework.
It expands on the benefits and drawbacks compared to the monolithic architecture that’s popular in Java EE. It shows how to quickly and simply develop two microservices with standard Java EE using KumuluzEE. You can find the examples produced in this article on GitHub under the name
The industry standard approach for deploying Java EE applications is packing all components into single EAR/WAR archive and deploying the archive on an application server. Although this approach has several advantages, particularly from the ease-of-development perspective, it leads to monolithic architecture, makes applications difficult to maintain, and - particularly important - makes such applications more difficult and sometimes impossible to scale to meet today’s real world demands, especially in PaaS (cloud) environments.
Microservice architecture addresses these shortcomings by decomposing an application into a set of microservices. Each microservice has well-defined functionalities and an interface to communicate with other microservices (such as REST, WSDL, or if needed even RMI). Most often, microservices are stateless.
Instead of packing all microservices into a single archive (EAR/WAR), each microservice is developed and deployed independently of each other. This brings several advantages, such as:
- With microservices, applications are more flexible;
- Every microservice can be developed independently of other microservices, which simplifies lifecycle and change management, makes it easier to use different technologies or upgrade to newer versions;
- Makes it easier to adopt new technologies for parts of an application;
- Makes it much more efficient to scale applications in PaaS and Docker-like environments.