Microservices are a commonly used architectural style targeting independent development, deployment, and release of services, as well as supporting polyglot capabilities and rapid release strategies. This depends on the presence of certain software architecture qualities. A number of architecture patterns and best practices that support the required qualities have been proposed in the literature, but usually in isolation of one another. Additionally, in real-world systems, assessing conformance to these patterns and practices and detecting possible violations is a significant challenge. For small-scale systems of a few services, a manual assessment and violation detection by an expert is probably both accurate and sufficient. However, for industrial-scale systems of several hundred or more services, manual assessment and violation detection is laborious and likely leads to inaccurate results. Furthermore, manual assessment is impractical for rapidly evolving and frequently released system architectures. In this work we examine a subset of microservice-relevant patterns, and propose a method for the semi-automatic detection and resolution of conformance violations. Our aim is to assist the software architect by providing a set of possible fix options and generating models of "fixed" architectures.

Top