發(fā)布時(shí)間：2018-08-28 17:54

【摘要】：生存性是信息系統(tǒng)的重要指標(biāo)之一，它描述了系統(tǒng)在遭受攻擊、出現(xiàn)故障或意外事故后，即使部分組件或服務(wù)失效、系統(tǒng)不再完整，仍然能夠及時(shí)完成其任務(wù)或提供關(guān)鍵服務(wù)的能力。在越來越多的信息系統(tǒng)被遷移到云環(huán)境下的情況下，信息系統(tǒng)生存性的研究也迎來了新的問題與條件。 提升分布式信息系統(tǒng)生存性的方式是多種多樣的。本論文關(guān)注于分布式信息系統(tǒng)的生存性的評估和優(yōu)化，提出了一種部署于IaaS云環(huán)境的分布式信息系統(tǒng)的生存性量化分析框架。本論文的主要研究方向是：在一個(gè)有限的包括若干物理設(shè)備和部署了分布式信息系統(tǒng)的虛擬機(jī)的環(huán)境內(nèi)，分析物理設(shè)備的可靠性，依據(jù)分布式信息系統(tǒng)不同組件間的邏輯關(guān)系得到各個(gè)組件對其他組件的貢獻(xiàn)值，并與組件本身的價(jià)值合并得到組件的性能，通過對這些數(shù)據(jù)的分析，量化評估一種部署方案下該分布式信息系統(tǒng)的生存性。然后，生存性量化分析框架可以進(jìn)一步得出該分布式信息系統(tǒng)在限定環(huán)境下的近似最優(yōu)生存性，以及與之對應(yīng)的能達(dá)到近似最優(yōu)生存性的部署方案。最后，在IaaS云中執(zhí)行由該部署方案生成的運(yùn)維指令，將信息系統(tǒng)按新的部署方案部署，使之獲得近似最優(yōu)的生存性。 本論文的主要貢獻(xiàn)包括：首先，，提出了一種量化評估部署于IaaS云環(huán)境的分布式信息系統(tǒng)的生存性的框架；其次，在該框架下，提供了分析計(jì)算近似最優(yōu)生存性的部署方案的方法；最后，應(yīng)用多種算法進(jìn)行了實(shí)驗(yàn)，驗(yàn)證了生存性量化分析框架的有效性和應(yīng)用到實(shí)際場景中的效果。
[Abstract]:Survivability is one of the most important indicators of information systems. It describes that the system is no longer complete after an attack, failure or accident, even if some components or services fail. The ability to complete its mandate or provide critical services in a timely manner. With more and more information systems being migrated to the cloud environment, the research on the survivability of information systems is confronted with new problems and conditions. There are many ways to improve the survivability of distributed information systems. This paper focuses on the survivability evaluation and optimization of distributed information systems, and proposes a quantitative survivability analysis framework for distributed information systems deployed in the IaaS cloud environment. The main research direction of this thesis is to analyze the reliability of physical devices in a limited environment, which includes a number of physical devices and a virtual machine that has deployed distributed information systems. According to the logical relationship between the different components of the distributed information system, the contribution value of each component to other components is obtained, and the performance of the component is obtained by combining with the value of the component itself. Quantificationally evaluate the survivability of the distributed information system under a deployment scheme. Then, the framework of survivability quantification analysis can further obtain the approximate optimal survivability of the distributed information system in a limited environment, and the corresponding deployment scheme which can achieve the approximate optimal survivability. Finally, the operation instructions generated by the deployment scheme are executed in the IaaS cloud, and the information system is deployed according to the new deployment scheme to obtain approximately optimal survivability. The main contributions of this thesis are as follows: firstly, a framework to quantitatively evaluate the survivability of distributed information systems deployed in IaaS cloud environment is proposed. A method to analyze and calculate the approximately optimal survivability deployment scheme is provided. Finally, experiments are carried out with various algorithms to verify the validity of the survivability quantitative analysis framework and the effectiveness of its application to actual scenarios.
【學(xué)位授予單位】：中山大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2013
【分類號】：TP309

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 郭亞軍;動(dòng)態(tài)綜合評價(jià)的二次加權(quán)法[J];東北大學(xué)學(xué)報(bào);1995年05期

2 鮑鳴;戴躍偉;孔建壽;鄒云;;冗余系統(tǒng)的生存能力分析與評估[J];控制與決策;2009年12期

本文編號：2210129