Resilience-Aware Elastic Scaling for Cloud-Native Online DL Training on Multi-Tenant GPU Clusters
Summary: WeFlex elastically leases idle GPUs from online DL jobs to other tenants, reclaiming them as demand rises. Its interruption-mitigation pipeline, topology-aware orchestration, and right-of-return leasing preserve training continuity at 10K+ GPU scale. (summarized by gpt-5.6-luna on Jul 24 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Qianhao Wu (Tencent)
- 2. Jiazhi Jiang (Beijing Normal University)
- 3. Guohui Ling (Tencent)
- 4. Yue Pang (Tencent)
BibTeX Citation
@article{wu_vldb26,
title = {{Resilience-Aware Elastic Scaling for Cloud-Native Online DL Training on Multi-Tenant GPU Clusters}},
author = {Wu, Qianhao and Jiang, Jiazhi and Ling, Guohui and Pang, Yue},
journal = {PVLDB},
series = {{VLDB} '26},
volume = {19},
number = {3},
pages = {375--387},
doi = {10.14778/3778092.3778099},
url = {https://doi.org/10.14778/3778092.3778099},
year = {2026}
}Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 3 of 3 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 3,950 | MagicScaler: Uncertainty-aware, Predictive Autoscaling | 2023 | VLDB | 6.9980354e-05 |
| 5,821 | StreamOps: Cloud-Native Runtime Management for Streaming Services in ByteDance | 2023 | VLDB | 6.07598e-05 |
| 10,695 | Oceanus: Enable SLO-Aware Vertical Autoscaling for Cloud-Native Streaming Services in Tencent | 2025 | SIGMOD | 5.093636e-05 |
Previous
Page 1 / 1
Next