Please note: This PhD seminar will take place in DC 2564 and online.
Nimmi Rashinika Weeraddana, PhD candidate
David R. Cheriton School of Computer Science
Supervisor: Professor Shane McIntosh
DevOps is a key element in sustaining the quality and efficiency of software development. Yet, the effectiveness of DevOps methodologies extends beyond just technological expertise. It is greatly affected by the manner in which teams handle and engage with DevOps artefacts. Grasping the intricacies of code ownership and contribution patterns within DevOps artefacts is vital for refining strategies and ensuring they deliver their full potential. There are two main strategies to manage DevOps artefacts as suggested in prior work: (1) all project developers need to contribute to DevOps artefacts, and (2) a dedicated group of developers needs to be authoring DevOps artefacts.
To analyze which strategy works best for Open-Source Software (OSS) projects, we conduct an empirical analysis on a dataset of 892,193 CircleCI builds spanning 1,689 OSS projects. We employ a two-pronged approach to our study. First, we investigate the impact of chronological code ownership of DevOps artefacts on the outcome of a CI build on a build level. Second, we study the impact of the Skewness of DevOps contributions on the success rate of CI builds at the project level. Our findings reveal that, in general, larger chronological ownership and higher Skewness values of DevOps contributions are related to more successful build outcomes and higher rates of successful build outcomes, respectively. We further find that projects with low Skewness values could have high build success rates when the number of developers in the project is relatively small. Thus, our results suggest that while larger software organizations are better off having dedicated DevOps developers, smaller organizations would benefit from having all developers involved in DevOps.
To attend this PhD seminar in person, please go to DC 2564. You can also attend virtually on Zoom.