No such thing as a “free launch”? Systematic benchmarking of containers
Bibliographic record
Abstract
Docker containers have recently become an extremely popular means of building robust, modular systems. Essentially, many architectures leverage lightweight virtualization to manage micro-services. Surprisingly however, there are very few studies revealing the overheads, such as starting new containers in orchestration systems, such as Kubernetes. Though traditional Virtual Machines (VMs) can take on the order of minutes to launch, containers are much faster and the launch times can be on the order of seconds. These overheads typically considered to be negligible compared with the benefits of container-based systems, however, are the predictable? Our work investigates these costs in a systematic study within a private cloud platform. The evaluation outlines a process for studies of this kind. Our results confirm that launch times of VMs are in the range of minutes, whereas containers typically only take seconds. However, these results also show that launch times for new containers do not always scale linearly. Specifically, we identify our system organized by Minikube, a tool that eases local deployment of Kubernetes, introduces a penalty on launch times once the number of containers exceeds 80% of the maximum number of pods available for the cluster. This work demonstrates the presence of unexpected overheads and the need for our proposed systematic infrastructure for testing deployments of containerized services at scale.
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 | 0.011 | 0.052 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.002 | 0.005 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| 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".