Graph neural networks for detecting anomalies in scientific workflows

Jin, Hongwei; Raghavan, Krishnan; Papadimitriou, George; Wang, Cong; Mandal, Anirban; Kiran, Mariam; Deelman, Ewa; Balaprakash, Prasanna

doi:10.1177/10943420231172140

Graph neural networks for detecting anomalies in scientific workflows

Journal Article · 29 May 2023 · International Journal of High Performance Computing Applications

DOI:https://doi.org/10.1177/10943420231172140· OSTI ID:1975863

^[1]; Raghavan, Krishnan ^[1]; ^[2]; Wang, Cong ^[3]; Mandal, Anirban ^[3]; Kiran, Mariam ^[4]; ^[2]; Balaprakash, Prasanna ^[5]

Argonne National Laboratory, Lemont, IL, USA
University of Southern California, Los Angeles, CA, USA
Renaissance Computing Institute (RENCI), Chapel Hill, NC, USA
Energy Sciences Network (ESnet), Berkeley, CA, USA
Oak Ridge National Laboratory, Oak Ridge, TN, USA

Identifying and addressing anomalies in complex, distributed systems can be challenging for reliable execution of scientific workflows. We model these workflows as directed acyclic graphs (DAGs), where the nodes and edges of the DAGs represent jobs and their dependencies, respectively. We develop graph neural networks (GNNs) to learn patterns in the DAGs and to detect anomalies at the node (job) and graph (workflow) levels. We investigate workflow-specific GNN models that are trained on a particular workflow and workflow-agnostic GNN models that are trained across the workflows. Our GNN models, which incorporate both individual job features and topological information from the workflow, show improved accuracy and efficiency compared to conventional learning methods for detecting anomalies. While joint trained with multiple scientific workflows, our GNN models reached an accuracy more than 80% for workflow level and 75% for job level anomalies. In addition, we illustrate the importance of hyperparameter tuning method in our study that can significantly improve the metric(s) measure of evaluating the GNN models. Finally, we integrate explainable GNN methods to provide insights on job features in the workflow that cause an anomaly.

View Journal Article

Sponsoring Organization:: USDOE

Grant/Contract Number:: SC0022328

OSTI ID:: 1975863

Alternate ID(s):: OSTI ID: 2404548

Journal Information:: International Journal of High Performance Computing Applications, Journal Name: International Journal of High Performance Computing Applications Journal Issue: 3-4 Vol. 37; ISSN 1094-3420

Publisher:: SAGE PublicationsCopyright Statement

Country of Publication:: United States

Language:: English

References (22)

Building a scientific workflow framework to enable real‐time machine learning and visualization Li, Feng; Song, Fengguang Concurrency and Computation: Practice and Experience, Vol. 31, Issue 16 https://doi.org/10.1002/cpe.4703	journal	June 2018
Deep Learning for Enhancing Fault Tolerant Capabilities of Scientific Workflows Singh, Alok; Altintas, Ilkay; Schram, Malachi 2018 IEEE International Conference on Big Data (Big Data) https://doi.org/10.1109/BigData.2018.8622509	conference	December 2018
Weather Radar Data Interpolation Using a Kernel-Based Lagrangian Nowcasting Technique Ruzanski, Evan; Chandrasekar, V. IEEE Transactions on Geoscience and Remote Sensing, Vol. 53, Issue 6 https://doi.org/10.1109/TGRS.2014.2368076	journal	June 2015
Identifying Execution Anomalies for Data Intensive Workflows Using Lightweight ML Techniques Wang, Cong; Papadimitriou, George; Kiran, Mariam 2020 IEEE High Performance Extreme Computing Conference (HPEC) https://doi.org/10.1109/HPEC43674.2020.9286139	conference	September 2020
Measuring and Relieving the Over-Smoothing Problem for Graph Neural Networks from the Topological View Chen, Deli; Lin, Yankai; Li, Wei Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 34, Issue 04 https://doi.org/10.1609/aaai.v34i04.5747	journal	April 2020
Anomaly Detection in Scientific Workflows using End-to-End Execution Gantt Charts and Convolutional Neural Networks Krawczuk, Patrycja; Papadimitriou, George; Nagarkar, Shubham Practice and Experience in Advanced Research Computing https://doi.org/10.1145/3437359.3465597	conference	July 2021
Detecting performance anomalies in scientific workflows using hierarchical temporal memory Rodriguez, Maria A.; Kotagiri, Ramamohanarao; Buyya, Rajkumar Future Generation Computer Systems, Vol. 88 https://doi.org/10.1016/j.future.2018.05.014	journal	November 2018
The open science grid Pordes, Ruth; Petravick, Don; Kramer, Bill Journal of Physics: Conference Series, Vol. 78 https://doi.org/10.1088/1742-6596/78/1/012057	journal	July 2007
XSEDE: Accelerating Scientific Discovery Towns, John; Cockerill, Timothy; Dahan, Maytal Computing in Science & Engineering, Vol. 16, Issue 5 https://doi.org/10.1109/MCSE.2014.80	journal	September 2014
AI4IO: A Suite of Ai-Based Tools for IO-Aware HPC Resource Management Taufer, Michela 2021 IEEE 28th International Conference on High Performance Computing, Data, and Analytics (HiPC) https://doi.org/10.1109/HiPC53243.2021.00012	conference	December 2021
ExoGENI: A Multi-Domain Infrastructure-as-a-Service Testbed Baldin, Ilya; Chase, Jeff; Xin, Yufeng The GENI Book https://doi.org/10.1007/978-3-319-33769-2_13	book	January 2016
End-to-end online performance data capture and analysis for scientific workflows Papadimitriou, George; Wang, Cong; Vahi, Karan Future Generation Computer Systems, Vol. 117 https://doi.org/10.1016/j.future.2020.11.024	journal	April 2021
Pegasus, a workflow management system for science automation Deelman, Ewa; Vahi, Karan; Juve, Gideon Future Generation Computer Systems, Vol. 46 https://doi.org/10.1016/j.future.2014.10.008	journal	May 2015
PANORAMA: An approach to performance modeling and diagnosis of extreme-scale workflows Deelman, Ewa; Carothers, Christopher; Mandal, Anirban The International Journal of High Performance Computing Applications, Vol. 31, Issue 1 https://doi.org/10.1177/1094342015594515	journal	July 2016
Anomaly detection for scientific workflow applications on networked clouds Gaikwad, Prathamesh; Mandal, Anirban; Ruth, Paul 2016 International Conference on High Performance Computing & Simulation (HPCS) https://doi.org/10.1109/HPCSim.2016.7568396	conference	July 2016
A global reference for human genetic variation Consortium, The 1000 Genomes Project; Auton, Adam; Abecasis, Gonçalo R. Nature, Vol. 526, Issue 7571, p. 68-74 https://doi.org/10.1038/nature15393	journal	January 2015
The role of machine learning in scientific workflows Deelman, Ewa; Mandal, Anirban; Jiang, Ming The International Journal of High Performance Computing Applications, Vol. 33, Issue 6 https://doi.org/10.1177/1094342019852127	journal	May 2019
Using simple PID-inspired controllers for online resilient resource management of distributed scientific workflows Ferreira da Silva, Rafael; Filgueira, Rosa; Deelman, Ewa Future Generation Computer Systems, Vol. 95 https://doi.org/10.1016/j.future.2019.01.015	journal	June 2019
DeepHyper: Asynchronous Hyperparameter Search for Deep Neural Networks Balaprakash, Prasanna; Salim, Michael; Uram, Thomas D. 2018 IEEE 25th International Conference on High Performance Computing (HiPC) https://doi.org/10.1109/HiPC.2018.00014	conference	December 2018
Training Classifiers to Identify TCP Signatures in Scientific Workflows Papadimitriou, George; Kiran, Mariam; Wang, Cong 2019 IEEE/ACM Innovating the Network for Data-Intensive Science (INDIS) https://doi.org/10.1109/INDIS49552.2019.00012	conference	November 2019
Toward a Dynamic Network-Centric Distributed Cloud Platform for Scientific Workflows: A Case Study for Adaptive Weather Sensing Lyons, Eric; Papadimitriou, George; Wang, Cong 2019 15th International Conference on eScience (eScience) https://doi.org/10.1109/eScience.2019.00015	conference	September 2019
Workflow Anomaly Detection with Graph Neural Networks Jin, Hongwei; Raghavan, Krishnan; Papadimitriou, George 2022 IEEE/ACM Workshop on Workflows in Support of Large-Scale Science (WORKS) https://doi.org/10.1109/WORKS56498.2022.00010	conference	November 2022

Graph neural networks for detecting anomalies in scientific workflows

References (22)

Similar Records

Related Subjects

Graph neural networks for detecting anomalies in scientific workflows

Citation Formats

References (22)

Similar Records

Related Subjects