EUVIMEDEUVIMED entdecken
Crossref

Scalable, Low-Latency, and Hi-Precision Congestion Control in RDMA Datacenter Networks

Sun Xu, Bodong Yan, Yangming Zhao, Jianchun Liu, Hongli Xu

IEEE Transactions on Networking · 2026 · 34 · 5725-5739

DOI: 10.1109/ton.2026.3704115 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Networking, 2026. Verfasst von Sun Xu und 4 weiteren Autor:innen. Gefördert von: National Natural Science Foundation of China, Quantum Science and Technology-National Science and Technology Major Project, Fundamental and Interdisciplinary Disciplines Breakthrough Plan of the Ministry of Education of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Congestion Control (CC) plays a vital role in deploying lossless datacenter networks based on Remote Direct Memory Access (RDMA). A high-performance CC scheme should provide low-latency and precise feedback to congestion events without incurring a large system overhead. However, no existing CC schemes achieved all these features simultaneously. In this paper, we propose SLHCC, a Scalable, Low-latency, and Hi-precision Congestion Control scheme for RDMA datacenter networks. SLHCC uses out-of-band signaling to notify the network status and hence a packet sender can detect congestion events within an RTT. In addition, SLHCC adjusts the packet sending rate by taking into consideration all queues along the entire path that a packet has gone through. Accordingly, it provides a more precise CC compared with existing schemes especially when there are multiple bottlenecks in the networks. Furthermore, by tuning the frequency of network telemetry, SLHCC introduces little additional traffic into networks. We build the SLHCC prototype on a real testbed carrying NVIDIA BlueField-3 NICs and AGM39D FPGAs. Both testbed experiments and extensive simulations show that SLHCC can reduce the Flow Completion Time (FCT) slow down and reduce the buffer usage (i.e., reduce the queue lengths) by up to 62.5% and 58%, respectively, compared with HPCC. Even compared with the latest sub-RTT CC scheme, i.e., Bolt, SLHCC can reduce the FCT slow down and the buffer usage by up to 13.2% and 39%, respectively.

Förderung

  • National Natural Science Foundation of China
  • Quantum Science and Technology-National Science and Technology Major Project
  • Fundamental and Interdisciplinary Disciplines Breakthrough Plan of the Ministry of Education of China
  • “111 Center”

Zitieren

APA

Xu, S., Yan, B., Zhao, Y., Liu, J., & Xu, H. (2026). Scalable, Low-Latency, and Hi-Precision Congestion Control in RDMA Datacenter Networks. IEEE Transactions on Networking, 34, 5725-5739. https://doi.org/10.1109/ton.2026.3704115

Vancouver

Xu S, Yan B, Zhao Y, Liu J, Xu H. Scalable, Low-Latency, and Hi-Precision Congestion Control in RDMA Datacenter Networks. IEEE Transactions on Networking. 2026;34:5725-5739. doi:10.1109/ton.2026.3704115

BibTeX
@article{Xu2026,
  author = {Xu, Sun and Yan, Bodong and Zhao, Yangming and Liu, Jianchun and Xu, Hongli},
  title = {Scalable, Low-Latency, and Hi-Precision Congestion Control in RDMA Datacenter Networks},
  journal = {IEEE Transactions on Networking},
  year = {2026},
  volume = {34},
  pages = {5725--5739},
  doi = {10.1109/ton.2026.3704115}
}
Kostenlos starten →