JZUS - Journal of Zhejiang University SCIENCE

Frontiers of Information Technology & Electronic Engineering 2024 Vol.25 No.10 P.1337-1352

HSDBA: a hierarchical and scalable dynamic bandwidth allocation for programmable data planes

Author(s): Dengyu RAN, Xiao CHEN, Lei SONG
Affiliation(s): National Network New Media Engineering Research Center, Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, China; more
Corresponding email(s): randy@dsp.ac.cn, xxchen@dsp.ac.cn, songl@dsp.ac.cn
Key Words: Dynamic bandwidth allocation, Software-defined networking, Programmable data plane, Protocol-oblivious forwarding switch (POFSwitch)

Share this article to： More <<< Previous Article \|Next Article >>>

Abstract: dynamic bandwidth allocation (DBA) is a fundamental challenge in the realm of networking. The rapid, accurate, and fair allocation of bandwidth is crucial for network service providers to fulfill service-level agreements, alleviate link congestion, and devise strategies to counter network attacks. However, existing bandwidth allocation algorithms operate mainly on the control plane of the software-defined networking paradigm, which can lead to considerable probing overhead and convergence latency. Moreover, contemporary network architectures necessitate a hierarchical bandwidth allocation system that addresses latency requirements. We introduce a fine-grained, hierarchical, and scalable DBA algorithm, i.e., the HSDBA algorithm, implemented on the programmable data plane. This algorithm reduces network overhead and latency between the data plane and the controller, and it is proficient in dynamically adding and removing network configurations. We investigate the practicality of HSDBA using protocol-oblivious forwarding switches. Experimental results show that HSDBA achieves fair bandwidth allocation and isolation guarantee within approximately 25 packets. It boasts a convergence speed 0.5 times higher than that of the most recent algorithm, namely, approximate hierarchical allocation of bandwidth (AHAB); meanwhile, it maintains a bandwidth enforcement accuracy of 98.1%.

HSDBA：一种面向可编程数据平面的分层可扩展动态带宽分配方法

冉登宇^1,2，陈晓^1,2，宋磊^1,2
¹中国科学院声学研究所，国家网络新媒体工程技术研究中心，中国北京市，100190
²中国科学院大学电子电气与通信工程学院，中国北京市，100049
摘要：动态带宽分配（DBA）是网络中一项经典问题。快速、准确和公平的带宽分配对于网络服务提供商的服务等级保障（SLA）、链路拥塞缓解和网络攻击应对具有重要意义。然而，现有的带宽分配算法主要在软件定义网络（SDN）范式的控制平面中实现，可能导致较大的探测开销和收敛延迟。此外，当代网络架构需要一个能满足延迟要求的分层带宽分配系统。本文提出HSDBA，这是一种完全在可编程数据面实现的细粒度、可扩展的动态带宽分配方案，消除了数据面与控制器的网络开销和延迟，并能应对随时到来的配置节点加入和退出。本文在协议无关转发软件交换机上探索了HSDBA的可行性。实验结果表明，HSDBA在接收到大约25个数据包内实现带宽的公平分配和隔离性保障。算法收敛速度比最新的近似分层带宽分配算法（AHAB）快0.5倍，并且带宽限制准确率达到98.1%。

关键词：动态带宽分配；软件定义网络；可编程数据平面；协议无关转发交换机（POFSwitch）

Reference

[1]Alcoz AG, Dietmüller A, Vanbever L, 2020. SP-PIFO: approximating push-in first-out behaviors using strict-priority queues. Proc 17^th USENIX Conf on Networked Systems Design and Implementation, p.59-76.

[2]Al-Fares M, Radhakrishnan S, Raghavan B, et al., 2010. Hedera: dynamic flow scheduling for data center networks. Proc 7^th USENIX Conf on Networked Systems Design and Implementation, Article 19.

[3]Barach D, Linguaglossa L, Marion D, et al., 2018. High-speed software data plane via vectorized packet processing. IEEE Commun Mag, 56(12):97-103.

[4]Bennett JCR, Zhang H, 1996. WF/sup 2/Q: worst-case fair weighted fair queueing. Proc IEEE INFOCOM, p.120-128.

[5]Bennett JCR, Zhang H, 1997. Hierarchical packet fair queueing algorithms. IEEE/ACM Trans Netw, 5(5):675-689.

[6]Benson T, Anand A, Akella A, et al., 2011. MicroTE: fine grained traffic engineering for data centers. Proc 7^th Conf on Emerging Networking Experiments and Technologies, Article 8.

[7]Cardwell N, Cheng Y, Gunn CS, et al., 2017. BBR: congestion-based congestion control. Commun ACM, 60(2):58-66.

[8]Cascone C, Bonelli N, Bianchi L, et al., 2017. Towards approximate fair bandwidth sharing via dynamic priority queuing. IEEE Int Symp on Local and Metropolitan Area Networks, p.1-6.

[9]Chen L, Li BC, Li B, 2016. Barrier-aware max-min fair bandwidth sharing and path selection in datacenter networks. IEEE Int Conf on Cloud Engineering, p.151-160.

[10]Curtis AR, Kim W, Yalagandula P, 2011. Mahout: low-overhead datacenter traffic management using end-host-based elephant detection. Proc IEEE INFOCOM, p.1629-1637.

[11]Dalton M, Schultz D, Adriaens J, et al., 2018. Andromeda: performance, isolation, and velocity at scale in cloud network virtualization. Proc 15^th USENIX Conf on Networked Systems Design and Implementation, p.373-387.

[12]Devera M, 2003. Linux Hierarchical Token Bucket. http://luxik.cdi.cz/devik/qos/htb/ [Accessed on Aug. 31, 2023].

[13]Floyd S, Jacobson V, 1995. Link-sharing and resource management models for packet networks. IEEE/ACM Trans Netw, 3(4):365-386.

[14]Guo CX, 2001. SRR: an O(1) time complexity packet scheduler for flows in multi-service packet networks. ACM SIGCOMM Comput Commun Rev, 31(4):211-222.

[15]Hong CY, Kandula S, Mahajan R, et al., 2013. Achieving high utilization with software-driven WAN. Proc ACM SIGCOMM, p.15-26.

[16]Hu YX, Li D, Sun PH, et al., 2020. Polymorphic smart network: an open, flexible and universal architecture for future heterogeneous networks. IEEE Trans Netw Sci Eng, 7(4):2515-2525.

[17]Jain S, Kumar A, Mandal S, et al., 2013. B4: experience with a globally-deployed software defined WAN. ACM SIGCOMM Comput Commun Rev, 43(4):3-14.

[18]Jing LN, Chen X, Wang JL, 2021. Design and implementation of programmable data plane supporting multiple data types. Electronics, 10(21):2639.

[19]Jing LN, Wang JL, Chen X, 2022. MSSA: constant time state search through multi-scope state area. Appl Sci, 12(2):559.

[20]Kumar A, Jain S, Naik U, et al., 2015. BwE: flexible, hierarchical bandwidth allocation for WAN distributed computing. Proc ACM Conf on Special Interest Group on Data Communication, p.1-14.

[21]Lee SSW, Chan KY, 2019. A traffic meter based on a multicolor marker for bandwidth guarantee and priority differentiation in SDN virtual networks. IEEE Trans Netw Serv Manag, 16(3):1046-1058.

[22]Li ZY, Hu YX, Tian L, et al., 2023. Packet rank-aware active queue management for programmable flow scheduling. Comput Netw, 225:109632.

[23]Luangsomboon N, Liebeherr J, 2021. HLS: a packet scheduler for hierarchical fairness. IEEE 29^th Int Conf on Network Protocols, p.1-11.

[24]MacDavid R, Chen XQ, Rexford J, 2023. Scalable real-time bandwidth fairness in switches. IEEE Conf on Computer Communications, p.1-10.

[25]Noormohammadpour M, Raghavendra CS, 2018. Datacenter traffic control: understanding techniques and tradeoffs. IEEE Commun Surv Tutor, 20(2):1492-1525.

[26]Pfaff B, Pettit J, Koponen T, et al., 2015. The design and implementation of Open vSwitch. Proc 12^th USENIX Conf on Networked Systems Design and Implementation, p.117-130.

[27]Ramabhadran S, Pasquale J, 2003. Stratified round robin: a low complexity packet scheduler with bandwidth fairness and bounded delay. Proc Conf on Applications, Technologies, Architectures, and Protocols for Computer Communications, p.239-250.

[28]Saeed A, Zhao YM, Dukkipati N, et al., 2019. Eiffel: efficient and flexible software packet scheduling. Proc 16^th USENIX Conf on Networked Systems Design and Implementation, p.17-32.

[29]Sharma NK, Liu M, Atreya K, et al., 2018. Approximating fair queueing on reconfigurable switches. Proc 15^th USENIX Conf on Networked Systems Design and Implementation, p.1-16.

[30]Sharma NK, Zhao CXY, Liu M, et al., 2020. Programmable calendar queues for high-speed packet scheduling. Proc 17^th USENIX Conf on Networked Systems Design and Implementation, p.685-699.

[31]Shieh A, Kandula S, Greenberg A, et al., 2011. Sharing the data center network. Proc 8^th USENIX Conf on Networked Systems Design and Implementation, p.309-322.

[32]Shrivastav V, 2019. Fast, scalable, and programmable packet scheduler in hardware. Proc ACM Special Interest Group on Data Communication, p.367-379.

[33]Sivaraman A, Subramanian S, Alizadeh M, et al., 2016. Programmable packet scheduling at line rate. Proc ACM SIGCOMM, p.44-57.

[34]Stoica I, Shenker S, Zhang H, 1998. Core-stateless fair queueing: achieving approximately fair bandwidth allocations in high speed networks. Proc ACM SIGCOMM, p.118-130.

[35]Thapeta VS, Shinde K, Malekpourshahraki M, et al., 2021. Nimble: scalable TCP-friendly programmable in-network rate-limiting. Proc ACM SIGCOMM Symp on SDN Research, p.27-40.

[36]Wang JL, Jing LN, Chen X, et al., 2022. Programmable data processing method and system design for polymorphic network. J Commun, 43(4):14-25.

[37]Wu XY, Wang Z, Wang WT, et al., 2023. Augmented queue: a scalable in-network abstraction for data center network sharing. Proc ACM SIGCOMM, p.305-318.

[38]Xia WF, Zhao P, Wen YG, et al., 2017. A survey on data center networking (DCN): infrastructure and operations. IEEE Commun Surv Tutor, 19(1):640-656.

[39]Xylomenos G, Ververidis CN, Siris VA, et al., 2014. A survey of information-centric networking research. IEEE Commun Surv Tutor, 16(2):1024-1049.

[40]Yang Y, Jiang HY, Wu YL, et al., 2021. C2QoS: CPU-cycle based network QoS strategy in vSwitch of public cloud. IFIP/IEEE Int Symp on Integrated Network Management, p.438-444.

[41]Yu JZ, Wang XZ, 2014. POFSwitch v1.0. https://github.com/ProtocolObliviousForwarding/pofswitch [Accessed on Aug. 31, 2023].

[42]Yu LC, Sonchack J, Liu V, 2022. Cebinae: scalable in-network fairness augmentation. Proc ACM SIGCOMM, p.219-232.

[43]Yu ZL, Hu CH, Wu JF, et al., 2021a. Programmable packet scheduling with a single queue. Proc ACM SIGCOMM, p.179-193.

[44]Yu ZL, Wu JF, Braverman V, et al., 2021b. Twenty years after: hierarchical core-stateless fair queueing. Proc 18^th USENIX Symp on Networked Systems Design and Implementation, p.29-45.

Open peer comments: Debate/Discuss/Question/Opinion

<1>

Similar articles

- Go to

HSDBA：一种面向可编程数据平面的分层可扩展动态带宽分配方法

Darkslateblue:Affiliate; Royal Blue:Author; Turquoise:Article

Reference