Mining Scientific Workflows for Anomalous Data Transfers

TL;DR

This paper introduces X-FLASH, a machine learning tool that improves detection of network anomalies in scientific workflows, significantly enhancing accuracy through hyperparameter tuning and data mining techniques.

Contribution

The paper presents a novel anomaly detection method combining XGBoost with a sequential optimizer, achieving up to 40% improvement over existing approaches.

Findings

X-FLASH outperforms previous methods in F-measure, G-score, and recall.

Hyperparameter tuning significantly enhances detection performance.

The approach requires fewer evaluations, making it efficient.

Abstract

Modern scientific workflows are data-driven and are often executed on distributed, heterogeneous, high-performance computing infrastructures. Anomalies and failures in the workflow execution cause loss of scientific productivity and inefficient use of the infrastructure. Hence, detecting, diagnosing, and mitigating these anomalies are immensely important for reliable and performant scientific workflows. Since these workflows rely heavily on high-performance network transfers that require strict QoS constraints, accurately detecting anomalous network performance is crucial to ensure reliable and efficient workflow execution. To address this challenge, we have developed X-FLASH, a network anomaly detection tool for faulty TCP workflow transfers. X-FLASH incorporates novel hyperparameter tuning and data mining approaches for improving the performance of the machine learning algorithms to accurately classify the anomalous TCP packets. X-FLASH leverages XGBoost as an ensemble model and couples XGBoost with a sequential optimizer, FLASH, borrowed from search-based Software Engineering to learn the optimal model parameters. X-FLASH found configurations that outperformed the existing approach up to 28\%, 29\%, and 40\% relatively for F-measure, G-score, and recall in less than 30 evaluations. From (1) large improvement and (2) simple tuning, we recommend future research to have additional tuning study as a new standard, at least in the area of scientific workflow anomaly detection.