硬件支持的多虚拟机数据交换及动态带宽分配方法

李龙飞; 王剑峰; 刘欢; 史阳春

Chinese

您当前的位置：

首页 >

文章列表页 >

硬件支持的多虚拟机数据交换及动态带宽分配方法

更新时间：2026-09-04

- 硬件支持的多虚拟机数据交换及动态带宽分配方法
- Hardware supported methods of multi-VM switching and dynamic bandwidth allocation
- 微电子学与计算机 2020年37卷第1期页码：93-98
- 作者机构：
  
  西安微电子技术研究所集成电路设计部,陕西,西安,710065
- 作者简介：
- 基金信息：
  
  装发部型谱系列项目 1407XJ0900
- DOI：
  中图分类号：
- 纸质出版：2020
- 稿件说明：
移动端阅览
李龙飞, 王剑峰, 刘欢, 史阳春. 硬件支持的多虚拟机数据交换及动态带宽分配方法[J]. 微电子学与计算机, 2020, 37(1): 93-98.

LI Long-fei, WANG Jian-feng, LIU Huan, et al. Hardware supported methods of multi-VM switching and dynamic bandwidth allocation[J]. 2020, 37(1): 93-98.
李龙飞, 王剑峰, 刘欢, 史阳春. 硬件支持的多虚拟机数据交换及动态带宽分配方法[J]. 微电子学与计算机, 2020, 37(1): 93-98. DOI：

LI Long-fei, WANG Jian-feng, LIU Huan, et al. Hardware supported methods of multi-VM switching and dynamic bandwidth allocation[J]. 2020, 37(1): 93-98. DOI：

摘要

针对同一宿主计算机上多虚拟机之间数据交换开销大且带宽分配不灵活的问题，提出了一种硬件支持的多虚拟机数据交换及动态带宽分配方法，并进行建模和实验.该方法釆用IO虚拟化的思想，面向以太网控制器的硬件架构进行改进与优化.通过对虚拟机的发送数据进行解析，同时扩展发送引擎对接收BD环的访问权限，实现数据从发送引擎直接向目的虚拟机接收队列的交换过程；通过对虚拟机中接收队列的数据信息进行统计与分析，实现对各个虚拟机的带宽进行动态分配与调整.以自主研发的千兆以太网控制器为原型搭建测试平台进行实验.结果表明，本文提出的方法不仅减小了多虚拟机之间数据交换和带宽分配的CPU开销，而且对以太网控制器和虚拟机管理程序均保持了兼容.

Abstract

Aiming at the problem of high data exchange overhead and inflexible bandwidth allocation between multi- virtual machines(VMs) on the same host

hardware supported methods of multi-VMs switching and dynamic bandwidth allocation are proposed. Based on the concept of IO virtualization

this method focuses on improving and optimizing the hardware architecture of Ethernet controller. By parsing the sending data of VM and extending the access authority of the sending engine to the receiving BD ring

the switching process from the sending engine to the destination VM is realized directly. Besides

through statistics and analysis of data of receiving queue in VM

the bandwidth of each VM can be dynamically allocated and adjusted. The experimental results show that the proposed methods not only reduce the CPU overhead of data exchange and bandwidth allocation between multi-VMs

but also maintains compatibility with Ethernet controllers and hypervisors.

关键词

Keywords

references

DONG Y, YANG X, LI J, et al. High performance network virtualization with SR-IOV[J]. Journal of Parallel and Distributed Computing, 2012, 72(11): 1471-1480. DOI: 1471

AZODOLMOLKY S, WIEDER P, YAHYAPOUR R. Cloud computing networking: challenges and opportunities for innovations[J]. IEEE Communications Magazine, 2013, 51(7): 54-62. DOI: 54

ADAMS K, AGESEN O. A comparison of software and hardware techniques for x86 virtualization[J]. ACM SIGOPS Operating Systems Review, 2006, 40(5): 2-13. DOI: 2

SHEA R, LIU J. Network interface virtualization: challenges and solutions[J]. IEEE Network, 2012, 26(5): 28-34. DOI: 28

GUAN H, MA R, LI J. Workload-aware credit scheduler for improving network I/O performance in virtualization environment[J]. IEEE transactions on cloud computing, 2014, 2(2): 130-142. DOI: 130

GUMASTE A, DAS T, KHANDWALA K, et al. Network Hardware Virtualization for Application Provisioning in Core Networks[J]. IEEE Communications Magazine, 2017, 55(2):152-159. DOI: 152

LIAO G, BHUYAN L. Performance measurement of an integrated nic architecture with 10gbe[C]// 17th IEEE Symposium on High Performance Interconnects. New York, NY, USA, IEEE, 2009: 52-59.

叶可江, 吴朝晖, 姜晓红, 等.虚拟化云计算平台的能耗管理[J].计算机学报, 2012, 35(6):1262-1285. DOI: 1262

CHERKASOVA L, GARDNER R. Measuring CPU Overhead for I/O Processing in the Xen Virtual Machine Monitor[C]//USENIX Annual Technical Conference, General Track, 2005: 50.

郭军, 闫永明, 马安香, 等.云环境下基于冷点虚拟机迁移的热点消除方法[J].清华大学学报(自然科学版), 2016(11):1232-1236. DOI: 1232

张洪素.基于Xen的虚拟交换技术的研究与设计[D].天津: 天津大学, 2012.

FEI Z, LIU G, FU X, et al. A Survey on Virtual Machine Migration: Challenges, Techniques and Open Issues[J]. IEEE Communications Surveys % Tutorials, 2018(99):1. DOI: 1

金顺福, 郝闪闪, 王宝帅.融合双速率和工作休眠的虚拟机调度策略及参数优化[J].通信学报, 2017, 38(12):10-20. DOI: 10

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

基于MSS-YOLO-MPHands融合架构的装配手部动作识别研究

基于监控指标关联性的微服务异常检测

基于改进RT-DETR的航拍小目标检测算法

注意力机制与改进密集卷积引导的信息隐藏模型

一种用于极低压boost系统的高效率电荷泵