Catalog and detection techniques of microservice anti-patterns and bad smells: A tertiary study

Abstract

Background: Various works investigated microservice anti-patterns and bad smells in the past few years. We identified seven secondary publications that summarize these, but they have little overlap in purpose and often use different terms to describe the identified anti-patterns and smells. Objective: This work catalogs recurring bad design practices known as anti-patterns and bad smells for microservice architectures, and provides a classification into categories as well as methods for detecting these practices. Method: We conducted a systematic literature review in the form of a tertiary study targeting secondary studies identifying poor design practices for microservices. Results: We provide a comprehensive catalog of 58 disjoint anti-patterns, grouped into five categories, which we derived from 203 originally identified anti-patterns for microservices. Conclusion: The results provide a reference to microservice developers to design better-quality systems and researchers who aim to detect system quality based on anti-patterns. It also serves as an anti-pattern catalog for development-aiding tools, which are not currently available for microservice system development but could mitigate quality degradation throughout system evolution.