Cloud-native Software Architecture Evaluation

Cloud-native is used as a term in industry and academia for describing how software should be built in order to run in cloud environments properly. The usage and popularity of the term leads to the desire to build web-based enterprise applications in a cloud-native way to benefit from the advantages of modern cloud environments. However, due to the complexity of modern cloud environments, cloud-native spans a broad range of aspects from the design of an application's structure and communication to the selected deployment technologies and infrastructure. Thus, understanding thoroughly what a cloud-native application design means and deciding how a specific application can benefit from cloud-native patterns and characteristics, is challenging for software architects and developers. This work presents the formulation and validation of a hierarchical quality model for cloud-native software architectures focusing on design time aspects. Cloud-native characteristics are formulated as factors for the quality model and linked to quality aspects through impact relationships. An impact describes how a quality aspect is influenced, if a factor is present in a system. Together with the quality model, a modeling approach for cloud-native software architectures and the Clounaq tool was developed. The Clounaq tool offers software architects a model-based method for designing and evaluating software architectures based on the quality model. For the evaluation approach, architectural measures quantify the factors of the quality model. Software architects and developers can thus better understand consequences of cloud-native characteristics and evaluate for new or existing applications, how the application architecture design can benefit from cloud-native characteristics. The resulting quality model captures the topic of cloud-native software architectures from a quality-focused point of view. It explains how applications can be built in a cloud-native way to achieve certain benefits. Through the characterization in the form of a quality model, software architects can evaluate and prioritize which cloud-native characteristics are relevant and applicable to a certain application at hand. The presented modeling approach demonstrates how core aspects of cloud-native software architectures can be represented in a formal textual and a graphical notation. The Clounaq tool includes implemented calculations for the catalog of architectural measures aligned with the modeling approach and can thus be used by practitioners and researchers for future work on cloud-native software architectures.