What We Know About Software Dependability in DevOps - A Tertiary Study

• Álan Júnior da Cruz Andrade UNIRIO
• Ednilson Veloso UNIRIO
• Gleison Santos UNIRIO

Resumo

Background: DevOps is viewed as an alternative approach to achieving high-quality software products. Dependability is recognized as a crucial aspect of software product quality. Existing literature highlights the lack of established standards, models, or methods for evaluating product quality within the DevOps paradigm. This emphasizes the need for further research to investigate the impact of DevOps on software quality attributes, particularly in relation to dependability.Objective: Our objective is to evaluate the scope of research on dependability in DevOps and identify what is known about this context by relating DevOps practices with dependability attributes. Method: We conducted a tertiary study to enhance the understanding of dependability in the context of DevOps. Results: We found 13 secondary studies that address dependability in DevOps. Within these studies, we identified 16 DevOps practices that have an impact on dependability and 12 attributes that are affected by DevOps practices. Additionally, we identified 6 measures related to dependability in the context of DevOps. Among the DevOps practices, the most commonly reported ones that impact dependability are Automation Practices, including deployment, testing, and infrastructure automation, as well as Cloud Computing Implementation. Conclusions: The results show that DevOps practices contribute to improve software dependability, mainly due to the impacts of these practices on dependability attributes. However, even though the literature reports some measures related to dependability, there is still a gap in understanding how organizations can assess dependability in DevOps.

Palavras-chave: Software Product Quality, DevOps, Dependability