Extracting Candidates of Microservices from Monolithic Application Code
Manabu Kamimura, Keisuke Yano, Tomomi Hatano, Akihiko Matsuo · 2018
Technology that facilitates rapid modification of existing business applications is necessary and it has been reported that making the system more adaptable to change is the strongest driver for legacy system modernization. There has been considerable interest in service-oriented architectures or microservices which enables the system to be quickly changed. Refactoring and, in particular, re-modularization operations can be performed to repair the design of a software system. Various approaches have been proposed to support developers during the re-modularization of a software system. The common problem in these efforts is to identify from monolithic applications the candidates of microservices, i.e., the programs or data that can be turned into cohesive, standalone services; this is a tedious manual effort that requires analyzing many dimensions of software architecture views and often heavily relies on the experience and expertise of the expert performing the extraction. To solve this problem, we developed a method that identifies the candidates of microservices from the source code by using software clustering algorithm SArF with the relation of "program groups" and "data" which we defined. Our method also visualizes the extracted candidates to show the relationship between extracted candidates and the whole structure. The candidates and visualization help the developers to capture the overview of the whole system and facilitated a dialogue with customers. We report two case studies to evaluate our results in which we applied our method to an open source application and an industrial application with our results reviewed by developers.