一种多租户云的内部网络共享策略

郑鹏飞 1; 2; 尤佳莉 1; 王劲林 1; 曾学文 1

doi:10.7652/xjtuxb201408010

您当前的位置：

首页 >

文章列表页 >

一种多租户云的内部网络共享策略

更新时间：2025-07-09

- 一种多租户云的内部网络共享策略
- An Intra-Network Sharing Policy for Multi-Tenant Clouds
- 西安交通大学学报 2014年48卷第8期页码：54-59
- 作者机构：
  
  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,北京,100190
  2. 中国科学院大学,北京,100049
- 作者简介：
- 基金信息：
  
  : 2014-03-04。作者简介: 郑鹏飞(1987—),男,博士生;尤佳莉(通信作者),女,副研究员。基金项目: 国家高技术研究发展计划资助项目(2011AA01A102);中国科学院战略性先导科技专项资助项目(XDA06010301);国家科技支撑计划资助项目(2012BAH02B03);中国科学院声学所前瞻课题资助项目(Y154211601)
- DOI：10.7652/xjtuxb201408010
  中图分类号： TP393
- 网络首发：2014-08-10，
  
  纸质出版：2014
- 稿件说明：
移动端阅览
郑鹏飞 1, 2, 尤佳莉 1, 等. 一种多租户云的内部网络共享策略[J]. 西安交通大学学报, 2014,48(8):54-59.

An Intra-Network Sharing Policy for Multi-Tenant Clouds[J]. 2014, 48(8): 54-59.
郑鹏飞 1, 2, 尤佳莉 1, 等. 一种多租户云的内部网络共享策略[J]. 西安交通大学学报, 2014,48(8):54-59. DOI： 10.7652/xjtuxb201408010.

An Intra-Network Sharing Policy for Multi-Tenant Clouds[J]. 2014, 48(8): 54-59. DOI： 10.7652/xjtuxb201408010.

摘要

针对现有多租户云无法保障租户网络性能的问题

提出一种内部网络共享策略(FairNet)。该策略平衡了最小保证、公平分配、最大利用3个目标; 为虚拟机增加了保证带宽和最大带宽2个定量描述网络性能的新参数

作为计算虚拟机和租户权重的基础; 在虚拟机监视器中实现了限速器、队列和调度器; 确保主机上每个虚拟机都能分配到一个最小带宽

且剩余带宽被虚拟机按需共享; 综合考虑租户在链路发送端和接收端的带宽需求来计算租户权重

并按照权重的比例分配链路带宽。在NS-3仿真器上进行了实验

结果表明:FairNet在不同的通信模式下均能比其他算法更公平地分配链路带宽

有效消除了发送端和接收端不平衡的影响

能够确保虚拟机在主机上的最小带宽

以远比PS-N小的计算代价取得仅次于PS-N的网络公平性。

Abstract

An intra-network sharing policy

FairNet

is proposed to solve the problem that existing multi-tenant clouds cannot guarantee network performance for tenants. The proposed policy gives a balance among three objectives

namely the minimum guarantee

the equitable allocation

and the maximum utilization. Two new parameters

guaranteed bandwidth and bandwidth cap

are introduced to quantitatively describe the network performance of a virtual machine

and are used as the bases of calculating the weights of virtual machines and tenants. Speed limiters

queues

and schedulers are realized in virtual machine monitors. Every virtual machine in a host is able to obtain a minimum bandwidth

and the idle bandwidth is shared by virtual machines with demands. The weights of tenants are calculated by comprehensively considering the bandwidth demand of tenants in sending side and receiving side

and the bandwidth is allocated according to the proportion of weights. Experimental results on the NS-3 simulator show that the FairNet allocates the bandwidth of links more fairly than other algorithms in different communication patterns

and effectively eliminates the influence caused by the unbalance between the sending side and the receiving side. The minimum bandwidth of virtual machines is guaranteed in hosts

and the calculation cost of the FairNet paid is much smaller than that of the PS-N

though the network fairness of the FairNet is little worse than that of the PS-N.

关键词

Keywords

references

LAM V T, RADHAKRISHNAN S, PAN Rong, et al. Netshare and stochastic netshare: predictable bandwidth allocation for data centers[J]. ACM SIGCOMM Computer Communication Review, 2012, 42(3): 5-11.

BALLANI H, COSTA P, KARAGIANNIS T, et al. Towards predictable datacenter networks[J]. ACM SIGCOMM Computer Communication Review, 2011, 41(4): 242-253.

GUO Chuanxiong, LU Guohan, WANG H J, et al. SecondNet: a data center network virtualization architecture with bandwidth guarantees[C]∥Proceedings of the 6th International Conference on Emerging Networking Experiments and Technologies. New York, USA: ACM, 2010: 1-12.

XIE Di, DING Ning, HU Y C, et al. The only constant is change: incorporating time-varying network reservations in data centers[J]. ACM SIGCOMM Computer Communication Review, 2012, 42(4): 199-210.

PAREKH A K, GALLAGER R G. A generalized processor sharing approach to flow control in integrated services networks: the single-node case[J]. IEEE/ACM Transactions on Networking, 1993, 1(3): 344-357.

DEMERS A, KESHAV S, SHENKER S. Analysis and simulation of a fair queueing algorithm[J]. ACM SIGCOMM Computer Communication Review, 1989, 19(4): 1-12.

BENNETT J C R, ZHANG Hui. WF2Q: worst-case fair weighted fair queueing[C]∥Proceedings of the 15th Annual Joint Conference of the IEEE Computer and Communications Societies Conference on the Conference on Computer Communications. Washington, DC, USA: IEEE Computer Society, 1996: 120-128.

SHIEH A, KANDULA S, GREENBERG A, et al. Sharing the data center network[C]∥Proceedings of the 8th USENIX Conference on Networked Systems Design and Implementation. Berkeley, CA, USA: USENIX Association, 2011: 23-23.

KABBANI A, ALIZADEH M, YASUDA M, et al. AF-QCN: approximate fairness with quantized congestion notification for multi-tenanted data centers[C]∥Proceedings of the 18th Annual Symposium on High Performance Interconnects. Washington, DC, USA: IEEE Computer Society, 2010: 58-65.

RODRIGUES H, SANTOS J R, TURNER Y, et al. Gatekeeper: supporting bandwidth guarantees for multi-tenant datacenter networks[C]∥Proceedings of the 3rd Conference on I/O Virtualization. Berkeley, CA, USA: USENIX Association, 2011: 6.

POPA L, KUMAR G, CHOWDHURY M, et al. FairCloud: sharing the network in cloud computing[J]. ACM SIGCOMM Computer Communication Review, 2012, 42(4): 187-198.

DEVIK M D A. HTB home[EB/OL].(2003-07-12)[2014-02-19]. http:∥luxik.cdi.cz/～devik/qos/htb/.

FLOYD S, JACOBSON V. Link-sharing and resource management models for packet networks[J]. IEEE/ACM Transactions on Networking, 1995, 3(4): 365-386.

NS-3[EB/OL].(2013-05-12)[2014-02-19]. http:∥www.nsnam.org/.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

多P2P覆盖网络的带宽分配方法

云计算环境下的大规模图状数据处理任务调度算法

云环境下基于多属性层次分析的虚拟机部署与调度策略

云计算环境中的组合文档模型及其访问控制方案

一种短作业环境下的延迟调度算法