An extensible data-driven approach for evaluating the quality of microservice architectures.

Microservice architecture (MSA) is defined as an architectural style where the software system is developed as a suite of small services, each running in its own process and communicating with lightweight mechanisms. The benefits of MSA are many, ranging from an increase in development productivity, to better business-IT alignment, agility, scalability, and technology flexibility. The high degree of microservices distribution and decoupling is, however, imposing a number of relevant challenges from an architectural perspective. In this context, measuring, controlling, and keeping a satisfactory level of quality of the system architecture is of paramount importance. In this paper we propose an approach for the specification, aggregation, and evaluation of software quality attributes for the architecture of microservice-based systems. The proposed approach allows developers to (i) produce architecture models of the system, either manually or automatically via recovering techniques, (ii) contribute to an ecosystem of well-specified and automatically-computable software quality attributes for MSAs, and (iii) continuously measure and evaluate the architecture of their systems by (re-)using the software quality attributes defined in the ecosystem. The approach is implemented by using Model-Driven Engineering techniques. The current implementation of the approach has been validated by assessing the maintainability of a third-party, publicly available benchmark system.