A classification approach of information needs for anti-pattern detection in microservice applications


Microservice applications are on the rise to become one of the most popular architectural styles. They offer many advantages over monolithic applications, such as being easily scalable and maintainable as well as enabling teams to work autonomously. Nevertheless, it is crucial to properly implement microservice architectures to avoid common drawbacks. Frequently used solutions that impair the quality of an application, often referred to as anti-patterns, and the detection of such in software architectures are a well-known problem in the domain of software engineering. However, in contrast to monolithic architectures, not much work has been done to improve the detection of anti-patterns in microservice architectures. In this thesis we propose a structure and classification approach of information needs for anti-pattern detection. In this context, the concepts of information requirements and information sources are defined and explained on the basis of an exemplary microservice application. Additionally, a catalog which proposes information requirements of 20 anti-patterns is presented. Building up on this catalog, methods to satisfy these information requirements are proposed. These insights are used to present an overview of the relationship between information requirements and information sources. Overall, the concepts form a basis for further refinement and evaluation of automated detection methods of anti-patterns in microservice applications.


Project information



Thesis for degree:



Tim Voigt