即使在進(jìn)行云遷移時存在許多大的挑戰(zhàn)（例如：安全性和可靠性）,但是實(shí)用的云計算解決方案已經(jīng)成為IT領(lǐng)域不容忽視的事實(shí)；同時,許多研究者正在接受這些挑戰(zhàn)。云計算是一種信息處理模式,在這個信息處理模式中,中央管理的計算能力被作為服務(wù)進(jìn)行交付,根據(jù)需要,通過網(wǎng)絡(luò)傳遞給各種面向用戶的設(shè)備。這些服務(wù)以基礎(chǔ)設(shè)施服務(wù)的形式、平臺服務(wù)的形式、軟件服務(wù)的形式或者網(wǎng)絡(luò)服務(wù)的形式推出。事實(shí)上,云計算已經(jīng)超出了現(xiàn)有的提供了異構(gòu)資源訪問能力的網(wǎng)格計算技術(shù)；當(dāng)資源提供者不能滿足用戶的各種要求時,用戶通常需要一個能滿足他們的特殊要求的環(huán)境。而云計算已經(jīng)被認(rèn)可為是一種能滿足用戶各種需求的解決方案,這就使得它在需求滿足方面優(yōu)于網(wǎng)格計算。云計算能滿足這些功能需要感謝的,既不是PC零件組成的大型數(shù)據(jù)中心和網(wǎng)絡(luò)服務(wù),也不是自動平衡工作量的能力,而是虛擬化。虛擬化的最簡單的形式是通過軟件將一個物理設(shè)備細(xì)分為相同的幾個離散的物理設(shè)備。雖然它們共享一臺服務(wù)器的硬件資源,但是他們的工作相互獨(dú)立沒有沖突；結(jié)果,這樣就減少使用中的硬件數(shù)量,提高資源的利用率,提高了不同應(yīng)用之間的故障和性能的隔離,緩解了虛擬機(jī)從一個主機(jī)移動到另一個主機(jī)的實(shí)時... 

【文章來源】：湖南大學(xué)湖南省 211工程院校 985工程院校 教育部直屬院校

【文章頁數(shù)】：128 頁

【學(xué)位級別】：博士

【文章目錄】：
DEDICATION
ABSTRACT
詳細(xì)中文摘要
TABLE OF CONTENTS
LIST OF FIGURES
LIST OF TABLES
CHAPTER 1:INTRODUCTION
    1.1 Background
    1.2 Virtualization technology
        1.2.1 Virtualization techniques
        1.2.2 Virtualization benefits
    1.3 Virtualization solutions
        1.3.1 Xen environment
    1.4 Problem definition
    1.5 Performance prediction modeIs
    1.6 Multi-objective optimization
    1.7 CloudSim
    1.8 Thesis structure
CHAPTER 2:LITERATURE REVIEW
    2.1 Workload managements and scheduling optimization
        2.1.1 Current technologies
        2.1.2 Academic researches
        2.1.3 Application oriented solutions
    2.2 System scaling based on migration in a virtualized environment
        2.2.1 Single objective trend
        2.2.2 Multi-objective trend
    2.3 Summary
CHAPTER 3:PROACTIVE WORKLOAD MANAGEMENT MODEL
    3.1 Introduction
    3.2 Contribution
    3.3 SMM Model
        3.3.1 Unseen sequences problem
        3.3.2 Model criteria
    3.4 Model implementation
    3.5 Workload imitation
    3.6 Experimental environment
    3.7 Configuration parameters
    3.8 Summary
CHAPTER 4:PWMM EVALUATION
    4.1 Introduction
    4.2 Evaluation in a simulation environment
        4.2.1 Virtual machine load balancing
        4.2.2 Expe rimental environment
        4.2.3 Expe rimental results
    4.3 Evaluation in a real environment
        4.3.1 Workload management methodology
        4.3.2 Experimental Environment
        4.3.3 Experimental results
        4.3.4 Discussion
    4.4 Summary
CHAPTER 5:STATIC BAYESIAN GAME BASED MULTI-OBJECTIVEGENETIC ALGORITHM
    5.1 Introduction
    5.2 Traditional MOGA
        5.2.1 Nondominated sorting
        5.2.2 Elitism mechanism
    5.3 SBG-MOGA Algorithm
        5.3.1 Game modeling
        5.3.2 Bayesian nash equilibrium
        5.3.3 Description of SBG-MOGA and convergence properties
    5.4 SBG-MOGA evaluation
    5.5 Summary
CHAPTER 6:A MULTI-OBJECTIVE VIRTUAL MACHINE MIGRATIONPOLICY
    6.1 Introduction
    6.2 Contribution
    6.3 Why SBG-MOGA
    6.4 Model description
    6.5 Objectives formulation
        6.5.1 Load Volume（LV）
        6.5.2 Power Consumption（PC）
        6.5.3 Thermal State（TS）
        6.5.4 Resource Wastage（RW）
        6.5.5 Migration Cost（MC）
    6.6 Targets Estimation
    6.7 Multi-objective evaluation
    6.8 Model implementation
        6.8.1 Algorithm analysis and applicability
    6.9 Experiments and results evaluation
    6.10 Summary
CONCLUSION
ACKNOWLEDGEMENTS
REFERENCES
APPENDIX A:SMM JAVA IMPLEMENTATION
    A.1 System prerequisites
    A.2 SMM class diagram
    A.3 SMM class breakdown
        A.3.1. Function "main"
        A.3.2. Function "probability"
        A.3.3. Function "predict"
        A.3.4. Function "variation"
        A.3.5. Function"updateHistory"
    A.4 BDcon class breakdown
APPENDIX B:LIST OF PUBICATIONS



本文編號：3002386