Decode the Workload: Training Deep Learning Models for Efficient Compute Cluster Representation
Monitoring cluster activity at higher sampling fidelity enhances job differentiation, while Graph Neural (GNN) assisted autoencoders higher reconstruction error for anomalous jobs.
Authors
- Malachi Schram (ORCID: https://orcid.org/0000-0002-3475-2871)
- Bryan Hess (ORCID: https://orcid.org/0000-0003-3391-7045)
- Kishansingh Rajput (ORCID: https://orcid.org/0000-0002-4430-9937)
- Mark A. Jones (ORCID: https://orcid.org/0000-0002-4243-9367)
- Steven T. Moore (ORCID: https://orcid.org/0000-0003-3995-4912)
- Helen McSpadden
- Laura Hild
- Zhenyu Dai (ORCID: https://orcid.org/0000-0002-2951-2910)
- Ahmed Mohammed
Institutions
- Thomas Jefferson National Accelerator Facility (US)
Publication Details
- Published
- 2024-10-01
- DOI
- https://doi.org/10.2172/2489884
- Primary Topic
- Advanced Clustering Algorithms Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- U.S. Department of Energy
- Office of Science
- Nuclear Physics