Abstract
The microservice architecture is claimed to satisfy ongoing software development demands, such as resilience, flexibility, and velocity. However, developing applications based on microservices also brings some drawbacks, such as the increased software operational complexity. Recent studies have also pointed out the lack of methods to prevent problems related to the maintainability of these solutions. Disregarding established design principles during the software evolution may lead to the so-called architectural erosion, which can end up in a condition of unfeasible maintenance. As microservices can be considered a new architecture style, there are few initiatives to monitoring the evolution of software microservice-based architectures. In this paper, we introduce the SYMBIOTE method for monitoring the coupling evolution of microservice-based systems. More specifically, this method collects coupling metrics during runtime (staging or production environments) and monitors them throughout software evolution. The longitudinal analysis of the collected measures allows detecting an upward trend in coupling metrics that could represent signs of architectural degradation. To develop the proposed method, we performed an experimental analysis of the coupling metrics behavior using artificially generated data. The results of these experiment revealed the metrics behavior in different scenarios, providing insights to develop the analysis method for the identification of architectural degradation. We evaluated the SYMBIOTE method in a real-case open source project called Spinnaker. The results obtained in this evaluation show the relationship between architectural changes and upward trends in coupling metrics for most of the analyzed release intervals. Therefore, the first version of SYMBIOTE has shown potential to detect signs of architectural degradation during the evolution of microservice-based architectures.
Highlights
Companies have established the use of the microservice architectural style in software development
The decision to start with minor release 6 was based on the fact that we identified in the release changelog that minor release 7 was the only one that had a change in the number of services and we needed to analyze if the method was capable of identifying this in our analysis
In this first analysis, the SYMBIOTE method issued an alert for indications of architectural degradation, as the metrics Service Importance Distribution (SID), Average number of directly connected services (ADCS), and Service Coupling Factor (SCF) indicate an upward trend in the metrics values throughout all 17 releases
Summary
Companies have established the use of the microservice architectural style in software development. Scaling agile processes and continuous deployment are among the primary motivations for its use [1] Microservices have as their main characteristic the breaking down of the software into small and independently deployable services that communicate through lightweight mechanisms [2]. The authors mention the growing demand for scalability and new practices for software development, such as continuous integration and continuous deployment, containerization, and cloud technologies. In this context, constant software changes may impact software architecture. The second part of this definition is a motivation for our work since the coupling is a relevant aspect regarding well-established design principles and maintainability [11]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
