DiffECN: Differential ECN Marking for Datacenter Networks
ECN marking has been integrated into datacenter switches to enable high-throughput and low-latency transport. We observe that current marking schemes are coarse-grained: they blindly mark all flows when congestion occurs, causing large flows to occupy undeserved bandwidth and preventing newly arriving small flows from finishing quickly. In this paper, we propose DiffECN, a differential marking strategy that marks only the flows that are the culprits of congestion and protects the remaining flows from being limited. We have implemented it in the Barefoot Tofino switch and performed extensive evaluations via both physical testbed and large-scale simulations. The results show that DiffECN can restrain flows responsible for congestion successfully while providing desirable network performance. For instance, compared to the legacy way of ECN marking, DiffECN achieves up to 32.5% (40.1%) lower average (99th percentile) flow completion time (FCT) for small flows while delivering similar FCT for large flows under production workloads.
Authors
- Xiangyu Gao (ORCID: https://orcid.org/0000-0002-7552-6192)
- Zhuotao Liu (ORCID: https://orcid.org/0000-0002-7532-0434)
- Hanlin Huang (ORCID: https://orcid.org/0000-0002-1972-2156)
- Xinle Du (ORCID: https://orcid.org/0000-0001-8918-9580)
- Tong Li (ORCID: https://orcid.org/0000-0002-6805-9565)
- Ke Xu (ORCID: https://orcid.org/0000-0003-2587-8517)
Institutions
- Huawei Technologies (China) (CN)
- Renmin University of China (CN)
- Tsinghua University (CN)
Publication Details
- Journal
- IEEE Transactions on Networking
- Published
- 2024-10-17
- DOI
- https://doi.org/10.1109/tnet.2024.3477511
- Citations
- 4
- Primary Topic
- Cloud Computing and Resource Management
- Type
- article
- Field-Weighted Citation Impact
- 3.12
Funders
- National Natural Science Foundation of China
- China National Funds for Distinguished Young Scientists
- National Key Research and Development Program of China