本文選題：云計(jì)算 + OpenStack��； 參考：《吉林大學(xué)》2014年碩士論文

【摘要】：隨著信息科技的發(fā)展，云計(jì)算被廣泛應(yīng)用到商業(yè)中，它將軟件、基礎(chǔ)設(shè)施以及平臺(tái)作為服務(wù)通過(guò)網(wǎng)絡(luò)提供給用戶，從而使用戶節(jié)約了管理軟件和硬件資源的費(fèi)用。虛擬化是實(shí)現(xiàn)云計(jì)算的主要技術(shù)之一，虛擬化技術(shù)能夠提供比實(shí)際物理機(jī)器運(yùn)行更多的虛擬機(jī)，實(shí)現(xiàn)資源的有效利用，虛擬機(jī)的管理和部署對(duì)于虛擬化具有很大意義。虛擬機(jī)具有靈活性，能夠在不中斷服務(wù)的情況下在物理機(jī)上進(jìn)行遷移。虛擬機(jī)的遷移實(shí)現(xiàn)系統(tǒng)負(fù)載均衡和優(yōu)化電源管理。 由于云計(jì)算中的資源龐大，需要對(duì)云計(jì)算中的資源進(jìn)行有效地管理，各大公司推出了自己的云管理平臺(tái)，在各自的云平臺(tái)上實(shí)現(xiàn)對(duì)資源進(jìn)行管理。2010年RackSpace公司提出了OpenStack開(kāi)放云管理平臺(tái)，OpenStack具有簡(jiǎn)單部署、大規(guī)�？缮炜s的特點(diǎn)，受到了更多公司的支持。在OpenStack云平臺(tái)上Nova利用API接口實(shí)現(xiàn)對(duì)虛擬機(jī)的監(jiān)控和管理。 本文首先介紹了云計(jì)算以及虛擬化技術(shù)，重點(diǎn)研究了OpenStack云平臺(tái)如何實(shí)現(xiàn)對(duì)資源和虛擬機(jī)的管理，，以及在OpenStack云平臺(tái)上虛擬機(jī)遷移的方式和原理。在分析了現(xiàn)有的虛擬機(jī)遷移技術(shù)的基礎(chǔ)上，結(jié)合OpenStack云平臺(tái)的特點(diǎn)，提出了在OpenStack云平臺(tái)上多個(gè)虛擬機(jī)同時(shí)遷移的方法。由于具有共享存儲(chǔ)的虛擬機(jī)集群在遷移過(guò)程中，內(nèi)存中存在大量的重復(fù)數(shù)據(jù)，傳輸這些數(shù)據(jù)占用大量的網(wǎng)絡(luò)帶寬，使用重復(fù)數(shù)據(jù)刪除算法對(duì)內(nèi)存數(shù)據(jù)進(jìn)行處理，使得重復(fù)數(shù)據(jù)在遷移的過(guò)程只傳輸一次。虛擬機(jī)集群遷移可以有效的提高資源的利用率，同時(shí)可以降低集群消耗。 通過(guò)在OpenStack云平臺(tái)上進(jìn)行驗(yàn)證，相對(duì)于直接拷貝內(nèi)存和基于內(nèi)存壓縮的虛擬機(jī)集群遷移的方法，此方法有效的減少了網(wǎng)絡(luò)數(shù)據(jù)傳輸量以及虛擬機(jī)遷移的停機(jī)時(shí)間和總遷移時(shí)間，并且在隨著遷移的虛擬機(jī)的數(shù)量增多時(shí)，網(wǎng)絡(luò)的數(shù)據(jù)傳輸量并沒(méi)有顯著的增加，實(shí)現(xiàn)虛擬機(jī)的高效遷移。
[Abstract]:With the development of information technology, cloud computing is widely used in business. It provides software, infrastructure and platform as services to users through the network, which saves users the cost of managing software and hardware resources. Virtualization is one of the main technologies to realize cloud computing. The virtualization technology can provide more virtual machines than the actual physical machines and realize the efficient utilization of resources. The management and deployment of virtual machines have great significance for virtualization. The virtual machine has the flexibility to migrate on the physical machine without interruption of service. The migration of virtual machine realizes system load balance and power management optimization. Due to the huge resources in cloud computing, the resources in cloud computing need to be managed effectively, so each company has launched its own cloud management platform. In 2010, RackSpace put forward OpenStack Open Cloud Management platform (OpenStack), which has the characteristics of simple deployment, large-scale scalability, and has been supported by more companies. Nova uses API interface to monitor and manage virtual machine on OpenStack cloud platform. This paper first introduces cloud computing and virtualization technology, focusing on how OpenStack cloud platform to manage resources and virtual machines, and how to migrate virtual machines on OpenStack cloud platform. Based on the analysis of the existing migration techniques of virtual machines and the characteristics of OpenStack cloud platform, a method of simultaneous migration of multiple virtual machines on OpenStack cloud platform is proposed. Because there is a lot of duplicate data in memory in the migration process of virtual machine cluster with shared storage, the transmission of these data takes up a lot of network bandwidth, so the repeated data deletion algorithm is used to process the memory data. Causes duplicate data to be transmitted only once during the migration process. Virtual machine cluster migration can effectively improve the utilization of resources and reduce cluster consumption. Compared with the methods of directly copying memory and migration of virtual machine cluster based on memory compression, this method can effectively reduce the amount of network data transfer and the downtime and total migration time of virtual machine migration. And with the number of migrated virtual machines increasing, the network data transfer volume does not significantly increase, to achieve efficient migration of virtual machines.
【學(xué)位授予單位】：吉林大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP393.07;TP311.52

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 張彬彬;羅英偉;汪小林;王振林;孫逸峰;陳昊罡;許卓群;李曉明;;虛擬機(jī)全系統(tǒng)在線遷移[J];電子學(xué)報(bào);2009年04期

2 王仕象;;關(guān)于虛擬機(jī)遷移技術(shù)的應(yīng)用研究[J];科技傳播;2012年20期

相關(guān)重要報(bào)紙文章 前1條

1 中國(guó)移動(dòng)通信集團(tuán)設(shè)計(jì)院有限公司 史彤;[N];計(jì)算機(jī)世界;2013年

相關(guān)博士學(xué)位論文 前1條

1 趙佳;虛擬機(jī)動(dòng)態(tài)遷移的關(guān)鍵問(wèn)題研究[D];吉林大學(xué);2013年

本文編號(hào)：2098863