Managing Quality Assurance Challenges of DevOps through Analytics

Abstract

DevOps is an intermarriage between developmental practices and operational modalities. The methodology employs the practices of continuous integration and delivery and places the deployment pipeline as the main requirement to automate, deliver and operate software in a robust way, without compromising on the quality in the software development process. Over time, many systems and tools have been developed to implement the deployment pipeline and support the continuous delivery process. A pipeline splits the process of software delivery into various stages. Each stage is designed to verify the quality of new features from a new perspective to attest to the functionality and prevent either small or big errors from impacting the end users. The pipeline must provide a response and feedback loop to the concerned team and provide a window into the flow of changes that takes place. However, there is no clear rule to define what goes into a pipeline. This paper reviews the challenges of quality assurance of DevOps and provides tentative recommendations to deal with quality issues. Our proposed pipeline with analytic features is expected to provide accurate metrics on a real-time basis.