Reproduce “The 'as Code' Activities: Development Anti-patterns for Infrastructure as Code” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Akond Rahman, Effat Farhana, Laurie Williams
This paper empirically identifies five development anti-patterns in Infrastructure as Code (IaC) scripts, such as 'Many Cooks Spoil' and 'Silos,' which correlate with increased defect rates, using a mixed-methods study of 2,138 Puppet scripts. It also demonstrates that models based on these development activities can predict IaC defects more effectively than code-level metrics.
Key insights
- Five distinct development anti-patterns, including 'Many Cooks Spoil' (many developers on a script) and 'Unfocused Contribution' (developers modifying many scripts), are consistently linked to higher defect rates in IaC.
- Practitioner perceptions regarding defect-prone activities sometimes diverge from quantitative empirical evidence, highlighting the importance of data-driven insights.
- Development activity metrics prove to be effective features for predicting IaC defects, outperforming traditional code-level and commit-message-based approaches.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.