摘要服务器虚拟化技术通过运用虚拟化的技术充分发挥服务器的硬件性能，在IT领域被广泛应用，VM（虚拟机）在虚拟化系统中的可用性问题愈加突出，其中如何降低系统损失，提高系统可靠性成为人们面对的主要问题。软件自愈技术是一种预防、延缓软件失效，提高系统可用性的有效途径。软件自愈分析模型作为自愈技术研究的核心内容之一，为如何实施自愈提供了决策依据。采用随机回报网，建立了三种不同自愈策略的分析模型，分别是Cold-VM自愈、Warm-VM自愈、Migrate-VM自愈，从系统稳态可用性、事务丢失数量方面对这三种模型进行评估。数值仿真实验表明：预拷贝下的实时VM迁移优于停机拷贝下的实时VM迁移，自愈技术优劣顺序依赖于当前参数设置，并不是固定不变。62972

毕业论文关键词  虚拟机 软件自愈 实时VM迁移 随机回报网 分析模型

毕业设计说明书（论文）外文摘要

Title     Modeling and Availability Analysis of Live VM    migration by SRN                    

Abstract Server virtualization technology gives full play to the performance of the hardware of the server through the use of virtualization technology.As server virtualization is used in a number of IT systems, the unavailability of virtual machines (VM) on server virtualized systems becomes a significant concern,and how to reduce system losses and improve system reliability become the main problems faced by the people.Software rejuvenation is an effective way to postpone or prevent failures caused by software aging and improve the system availability.Software rejuvenation analytical models,as one of the cores of rejuvenation technology research,provide a decision-making basis for implementing the rejuvenation.In this paper we build using stochastic reward nets the analytic models with three different rejuvenation policies,the Cold-VM rejuvenation,the Warm-VM rejuvenation,the Migrate-VM rejuvenation.We evaluate three models from the aspect of the system steady-state availability and the expected number of transactions lost.The results of numberical simulation experiments show: ‘pre-copy’ live VM migration is encouraged rather than pure ‘stop-and-copy’ migration.The relative order of these rejuvenation technology might be changed by a change in parameter.

Keywords  Virtual machine Software rejuvenation Live VM migration Stochastic reward net Analytic model

1 绪论 1

1.1 论文的背景和意义 1

1.2 软件自愈技术的研究现状 3

1.3 本文研究工作 4

1.4 论文结构 5

2 自愈技术 6

2.1 虚拟化系统 6

2.2 三种自愈技术 7

3 系统建模 10

3.1 SPNP及其优化 10

3.2 Cold-VM自愈模型 15

3.3 Warm-VM自愈模型 18

3.4 Migrate-VM自愈模型 20

4 模型分析 24

4.1 实验设置 24

4.2 分析方法和工具 25

4.3 稳态可用性 32

4.4 事务丢失数量 33

5 自愈技术对比分析 35

5.1 Cold-VM自愈技术VS Warm-VM自愈技术