NetHint: White-Box Networking for Multi-Tenant Data Centers
Jingrong Chen, Hong Zhang, Wei Zhang, Liang Luo, Jeffrey S. Chase, Ion Stoica, Danyang Zhuo
摘要
A cloud provider today provides its network resources to its tenants as a black box, such that cloud tenants have little knowledge of the underlying network characteristics. Meanwhile, data-intensive applications have increasingly migrated to the cloud, and these applications have both the ability and the incentive to adapt their data transfer schedules based on the cloud network characteristics. We find that the black-box networking abstraction and the adaptiveness of data-intensive applications together create a mismatch, leading to sub-optimal application performance.
This paper explores a white-box approach to resolving this mismatch. We propose NetHint, an interactive mechanism between a cloud tenant and a cloud provider to jointly enhance application performance. With NetHint, the provider provides a hint -an indirect indication of the underlying network characteristics (e.g., link-layer network topologies for a tenant's virtual machines, number of co-locating tenants, network bandwidth utilization), and the tenant's applications then adapt their transfer schedules accordingly. The NetHint design provides abundant network information for cloud tenants to compute their optimal transfer schedules, while introducing little overhead for the cloud provider to collect and expose this information. Evaluation results show that NetHint improves the average performance of allreduce completion time, broadcast completion time, and MapReduce shuffle completion time by 2.7×, 1.5×, and 1.2×, respectively.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
引用它的顶会 Paper6
- MCCS: A Service-based Approach to Collective Communication for Multi-Tenant CloudYongji Wu, Yechen Xu, Jingrong Chen, Zhaodong Wang 等SIGCOMM 2024 · 被引用 15 次
- eNetSTL: Towards an In-kernel Library for High-Performance eBPF-based Network FunctionsBin Yang, Dian Shen, Junxue Zhang, Hanlin Yang 等EuroSys 2025 · 被引用 9 次
- Phantora: Maximizing Code Reuse in Simulation-based Machine Learning System Performance EstimationJianxing Qin, Jingrong Chen, Xinhao Kong, Yongji Wu 等NSDI 2026 · 被引用 5 次
- Decouple and Decompose: Scaling Resource Allocation with DeDeZhiying Xu, Minlan Yu, Francis Y. YanOSDI 2025 · 被引用 5 次
- Preventing Network Bottlenecks: Accelerating Datacenter Services with Hotspot-Aware Placement for Compute and StorageHamid Hajabdolali Bazzaz, Yingjie Bi, Weiwu Pang, Minlan Yu 等NSDI 2025 · 被引用 3 次
它引用的顶会 Paper5
- Sirius: A Flat Datacenter Network with Nanosecond Optical SwitchingHitesh Ballani, Paolo Costa, Raphael Behrendt, Daniel Cletheroe 等SIGCOMM 2020 · 被引用 204 次
- Is Big Data Performance Reproducible in Modern Cloud Networks?Alexandru Uta, Alexandru Custura, Dmitry Duplyakin, Ivo Jimenez 等NSDI 2020 · 被引用 74 次
- Twenty Years After: Hierarchical Core-Stateless Fair QueueingZhuolong Yu, Jingfeng Wu, Vladimir Braverman, Ion Stoica 等NSDI 2021 · 被引用 45 次
- NetKernel: Making Network Stack Part of the Virtualized InfrastructureZhixiong Niu, Hong Xu, Peng Cheng, Qiang Su 等USENIX ATC 2020 · 被引用 24 次
- Scaling Distributed Machine Learning with In-Network AggregationAmedeo Sapio, Marco Canini, Chen-Yu Ho, Jacob Nelson 等NSDI 2021
相关 Paper
- Skyplane: Optimizing Transfer Cost and Throughput Using Cloud-Aware OverlaysParas Jain, Sam Kumar, Sarah Wooders, Shishir G. Patil 等NSDI 2023 · 被引用 66 次
- Workload Intelligence: Workload-Aware IaaS abstraction for Cloud EfficiencyLexiang Huang, Anjaly Parayil, Jue Zhang, Xiaoting Qin 等SC 2025 · 被引用 1 次
- CoINT2: A Heuristic Coordinator for Responsive Receive-Side Network I/O Virtualization in Overcommitment CloudXu Huan, Jian Li, Haibing GuanINFOCOM 2025 · 被引用 1 次
- Optimizing Network Provisioning through CooperationHarsha Sharma, Parth Thakkar, Sagar Bharadwaj, Ranjita Bhagwan 等NSDI 2022 · 被引用 4 次
- TINA: A Fair Inter-datacenter Transmission Mechanism with Deadline GuaranteeXiaodong Dong, Wenxin Li, Xiaobo Zhou, Keqiu Li 等INFOCOM 2020 · 被引用 11 次
