Jigsaw: A High-Utilization, Interference-Free Job Scheduler for Fat-Tree Clusters
Staci A. Smith, David K. Lowenthal
2021Year
4Citations
Abstract
Jobs on HPC clusters can suffer significant performance degradation due to inter-job network interference. Approaches to mitigating this interference primarily focus on reactive routing schemes. A better approach---in that it completely eliminates inter-job interference---is to implement scheduling policies that proactively enforce network isolation for every job. However, existing schedulers that allocate isolated partitions lead to lowered system utilization, which creates a barrier to adoption.
Ask about this paper
Ask your agent about it.
Lune has read the top-tier papers around this one, so every answer names the papers it rests on.
Related papers
- Pushing Performance Isolation Boundaries into Application with pBoxYigong Hu, Gongqi Huang, Peng HuangSOSP 2023 · 2 citations
- SATORI: Efficient and Fair Resource Partitioning by Sacrificing Short-Term Benefits for Long-Term Gains*Rohan Basu Roy, Tirthak Patel, Devesh TiwariISCA 2021 · 34 citations
- SchedInspector: A Batch Job Scheduling Inspector Using Reinforcement LearningDi Zhang, Dong Dai, Bing XieHPDC 2022 · 26 citations
- MonkeyTree: Near-Minimal Congestion for Multi-tenant Training via MigrationAnton A. Zabreyko, Weiyang Wang, Manya GhobadiSIGCOMM 2026
- DRLPart: A Deep Reinforcement Learning Framework for Optimally Efficient and Robust Resource Partitioning on Commodity ServersRuobing Chen, Jinping Wu, Haosen Shi, Yusen Li et al.HPDC 2021 · 22 citations
