RoCC

Taheri, Parvin; Menikkumbura, Danushka; Vanini, Erico; Fahmy, Sonia; Eugster, Patrick; Edsall, Tom

doi:10.1145/3386367.3431316

Cited by 38 publications

(2 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We propose a new switch-driven congestion control solution for RDMA-based datacenter networks, RoCC (Robust Congestion Control), that: (i) computes a fair rate using a classic proportional integral (PI) controller [22], (ii) signals that fair rate to the sources via ICMP (internet control message protocol), and (iii) auto-tunes the control parameters to ensure stability and responsiveness. This paper extends our previous work [23] with emphasis on the hardware feasibility of the solution and its comparison to some recentlyproposed solutions.…”

Section: D) Contributionsmentioning

confidence: 62%

Congestion Control for Datacenter Networks: A Control-Theoretic Approach

Menikkumbura

Taheri

Vanini

et al. 2023

IEEE Trans. Parallel Distrib. Syst.

Self Cite

View full text Add to dashboard Cite

In this paper, we present RoCC, a robust congestion control approach for datacenter networks based on RDMA. RoCC leverages switch queue size as an input to a PI controller, which computes the fair data rate of flows in the queue. The PI parameters are self-tuning to guarantee stability, rapid convergence, and fair and near-optimal throughput in a wide range of congestion scenarios. Our simulation and DPDK implementation results show that RoCC can achieve up to 7× reduction in PFC frames generated under high load levels, compared to DCQCN. At the same time, RoCC can achieve 1.7 − 4.5× and 1.4 − 3.9× lower tail latency for long flows and 2.1−7× and 3.5−8.2× lower tail latency for short flows, compared to DCQCN and HPCC, respectively. We also find that RoCC does not require PFC. The functional components of RoCC can be efficiently implemented in P4 and FPGA-based switch hardware.

show abstract

Section: D) Contributionsmentioning

confidence: 62%

Congestion Control for Datacenter Networks: A Control-Theoretic Approach

Menikkumbura

Taheri

Vanini

et al. 2023

IEEE Trans. Parallel Distrib. Syst.

Self Cite

View full text Add to dashboard Cite

show abstract

“…To effectively handle frequent incast scenarios, some explorations have been made on single-path congestion control schemes. Some schemes [4,[9][10][11] endeavor to control low queuing delays and avoid PFC [12] problems [13][14][15]. TIMELY [16] utilizes the increment between two consecutive observed Round-Trip Times (RTTs) as a congestion signal.…”

Section: Introductionmentioning

confidence: 99%

MPTD: optimizing multi-path transport with dynamic target delay in datacenters

Li,

Wang,

Huang

et al. 2024

Cluster Comput

View full text Add to dashboard Cite

In contemporary datacenter networks, various applications experience the challenge of incast occurrences. Data-intensive applications, particularly in distributed scenarios such as distributed deep learning, frequently encounter incast. While single-path transport protocols are widely deployed in datacenter networks, the available bandwidth is often severely underutilized. MP-RDMA, as a state-of-the-art multi-path transport protocol, has been implemented; however, it faces performance issues during large-scale incast communications. MP-RDMA employs insufficient congestion feedback and fixed rate adjustment for congestion management, making the scheme hardly scalable. In this paper, we introduce MPTD, an RTT-based multi-path transport protocol. MPTD encompasses two primary aspects: (1) a novel congestion control algorithm based on the setting of target delay; (2) an optimized packet scheduling mechanism. Specifically, the receiver dynamically updates the target delay in real-time based on network conditions, and the sender adjusts the sending rate accordingly. A categorized implementation for paths is added to achieve a more accurate identification of path characteristics. Our evaluation demonstrates that, in comparison to MP-RDMA, MPTD can achieve speedups of 1.35x/1.40x in flow completion time for two traffic patterns.

show abstract