Towards Generic Failure Prediction Models in Large-Scale Distributed Computing Systems
Bibliographic record
Abstract
The increasing complexity of Distributed Computing (DC) systems necessitates advanced failure prediction techniques to ensure reliability and efficiency. However, the hierarchical architecture of DC systems presents a significant challenge in developing generic failure prediction models owing to the trade-off between prediction accuracy and cost. This study addresses this challenge by proposing a comprehensive methodology for developing and evaluating hierarchical failure prediction models in DC systems. We used the Grid5000 failure dataset to conduct an in-depth exploratory data analysis and statistical examination of the system failure characteristics, providing insights into complex failure patterns. Our approach involves developing, evaluating, and comparing ML-based models across different hierarchical levels (systems, sites, and clusters) using logistic regression, random forest, and XGBoost. These models can predict key metrics, such as time between failures (TBF), time to return/repair (TTR), and failing node identification (NFI). After intensive experiments, XGBoost demonstrated robust predictive capabilities, achieving 66-100% accuracy for TBF, TTR, and NFI across different levels in the DC system hierarchy. In addition, we introduce a hierarchical DC failure prediction architecture to enhance system reliability and cost efficiency. We also demonstrate how service providers can optimise resource utilisation and costs by tailoring service reliability through hierarchical failure prediction, leading to superior cost-effective system level predictions or highly reliable cluster level predictions.
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 | 0.004 | 0.010 |
| 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 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 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 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".