It Works (only) on My Machine: A Study on Reproducibility Smells in Ansible Scripts
Bibliographic record
Abstract
Infrastructure as Code (IaC) automates the creation, configuration, management, and monitoring of computing infrastructure through code. One of the key principles that IaC promises is repeatability and reproducibility. However, certain programming practices in IaC platforms, especially those that allow imperative configuration, such as Ansible, hinder reproducibility in IaC scripts. This study, first, identifies such programming practices that we refer to as reproducibility smells by conducting a comprehensive multi-vocal literature review and propose a first-ever validated catalog of reproducibility smells for IaC scripts. We implement a tool viz. Reduse to identify reproducibility smells in Ansible scripts. Furthermore, we conduct an empirical study to reveal the proliferation of reproducibility smells in open-source projects and explore correlation and fine-grained co-occurrence relationships among them. We observe that broken dependency chain smell occurs the most in approximately $71 \%$ tasks that we analyzed. Our analysis uncovers significant positive correlations between specific reproducibility smells, implying that repositories with one such smell tend to exhibit others. Moreover, the co-occurrence analysis reveals smell pairs that show a high tendency of co-occurrence at the task granularity. With the developed tool Reduse, DevOps engineers can identify and rectify reproducibility issues before becoming part of the production system. Software engineering researchers can use the smells catalog proposed first in this study and can utilize Reduse in empirical studies exploring various facets of reproducibility.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.013 | 0.083 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.002 | 0.002 |
| Scholarly communication | 0.003 | 0.005 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".