A Large-scale Data Set and an Empirical Study of Docker Images Hosted on Docker Hub
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Docker is currently one of the most popular containerization solutions. Previous work investigated various characteristics of the Docker ecosystem, but has mainly focused on Dockerfiles from GitHub, limiting the type of questions that can be asked, and did not investigate evolution aspects. In this paper, we create a recent and more comprehensive data set by collecting data from Docker Hub, GitHub, and Bitbucket. Our data set contains information about 3,364,529 Docker images and 378,615 git repositories behind them. Using this data set, we conduct a large-scale empirical study with four research questions where we reproduce previously explored characteristics (e.g., popular languages and base images), investigate new characteristics such as image tagging practices, and study evolution trends. Our results demonstrate the maturity of the Docker ecosystem: we find more reliance on ready-to-use language and application base images as opposed to yet-to-be-configured OS images, a downward trend of Docker image sizes demonstrating the adoption of best practices of keeping images small, and a declining trend in the number of smells in Dockerfiles suggesting a general improvement in quality. On the downside, we find an upward trend in using obsolete OS base images, posing security risks, and find problematic usages of the latest tag, including version lagging. Overall, our results bring good news such as more developers following best practices, but they also indicate the need to build tools and infrastructure embracing new trends and addressing potential issues.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it