本文關(guān)鍵詞： SAN 以太網(wǎng)交換機 SDN　出處：《電子科技大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：在多元化的大數(shù)據(jù)網(wǎng)絡(luò)時代,人們構(gòu)建了由通信結(jié)構(gòu)、管理層、存儲部件和計算機系統(tǒng)組成的存儲區(qū)域網(wǎng)路(SAN)來解決網(wǎng)絡(luò)中的大數(shù)據(jù)傳輸和存儲的問題。但由于現(xiàn)有的網(wǎng)絡(luò)中同時存在以太網(wǎng)和光纖兩種傳輸模式,它們之間的互聯(lián)在產(chǎn)生冗余的同時也極大的提高了網(wǎng)絡(luò)的故障率,因此,如何將存儲網(wǎng)絡(luò)和以太網(wǎng)融合起來就成了新的發(fā)展趨勢�，F(xiàn)今的以太網(wǎng)與存儲網(wǎng)絡(luò)的融合主要有以太網(wǎng)和iSCSI的融合、基于以太網(wǎng)的NAS和基于以太網(wǎng)和光纖的SANs這三種,在這三種中,由于光纖網(wǎng)絡(luò)具有高速、無丟包和低延遲的優(yōu)勢,因此SANs和以太網(wǎng)的融合被看做是以太網(wǎng)存儲的最佳選擇。軟件定義網(wǎng)絡(luò)(SDN)提出了在控制層面上的抽象技術(shù),通過OpenFlow將網(wǎng)絡(luò)中的設(shè)備控制和數(shù)據(jù)分離開,通過軟件來定義邏輯上的網(wǎng)絡(luò)拓?fù)?從而實現(xiàn)了對網(wǎng)絡(luò)流量的靈活控制,同時這種控制層面上的抽象拓展了網(wǎng)絡(luò)管理和設(shè)計上的思路,使我們在對整個網(wǎng)絡(luò)的設(shè)計和管理時具有很大的靈活性。這為本課題的研究提供了充足的理論基礎(chǔ)。鑒于本課題的研究基礎(chǔ)在于現(xiàn)有的以太網(wǎng)和光纖網(wǎng)絡(luò),因此我們不僅要了解SAN和以太網(wǎng),熟悉他們的工作模式和技術(shù)特點,還需要從以太網(wǎng)中選擇出我們需要的部分并加以修改來應(yīng)用到改進(jìn)的SAN中,同時,還需要分析現(xiàn)有的交換機來提取出有用的部分進(jìn)行融合,因此本課題的核心工作就在于如何利用現(xiàn)有的規(guī)范和標(biāo)準(zhǔn),融合出一種新的處理模式,將原有的設(shè)備高效的融合起來,成為一種新型的整體。本文首先對本課題的背景、任務(wù)、研究意義和理論基礎(chǔ)做了系統(tǒng)性介紹,然后將現(xiàn)有普通存儲技術(shù)、以太網(wǎng)與融合SAN網(wǎng)絡(luò)的以太網(wǎng)技術(shù)做了詳細(xì)的對比分析得出融合SAN網(wǎng)絡(luò)的以太網(wǎng)技術(shù)的優(yōu)點,確定了要基于SDN的集中式架構(gòu)來實現(xiàn)SAN和以太網(wǎng)的融合,接著針對這一方案進(jìn)行了框架設(shè)計、流程設(shè)計和軟件設(shè)計,并對流程中的每一個重要模塊進(jìn)行了單獨的設(shè)計與實現(xiàn),然后進(jìn)行了流程整合和可靠性分析以及執(zhí)行效率的最優(yōu)化。最后,根據(jù)這一技術(shù)的特點提出了合理的測試方案和測試環(huán)境,并進(jìn)行了相應(yīng)的測試,根據(jù)測試結(jié)果進(jìn)行了整個方案的實用性分析和性能分析。
[Abstract]:In the era of diversified big data network, people have constructed a communication structure, management, The storage LAN composed of storage components and computer systems (SAN) is used to solve the transmission and storage problems of big data in the network. However, since there are both Ethernet and fiber transmission modes in the existing network, The interconnection between them not only produces redundancy, but also greatly increases the failure rate of the network. How to integrate storage network and Ethernet has become a new development trend. Nowadays, the fusion of Ethernet and storage network mainly includes Ethernet and iSCSI, Ethernet based NAS and Ethernet and fiber optic SANs. Among these three kinds of networks, because of the advantages of high speed, no packet loss and low delay, the integration of SANs and Ethernet is considered as the best choice for Ethernet storage. The device control in the network is separated from the data by OpenFlow, the logical network topology is defined by software, and the flexible control of network traffic is realized. At the same time, the abstraction at the control level expands the thinking of network management and design. It gives us great flexibility in designing and managing the whole network. This provides a sufficient theoretical basis for the research of this topic. Since the research foundation of this subject lies in the existing Ethernet and fiber optic networks, So not only do we need to learn about SAN and Ethernet, we need to be familiar with their working patterns and technical features, but we also need to select and modify the parts we need from Ethernet to apply to the improved SAN, and at the same time, Also need to analyze the existing switch to extract useful parts for fusion, so the core work of this topic is how to use the existing norms and standards, fuse a new processing mode, the original equipment will be highly efficient integration. This paper first introduces the background, task, research significance and theoretical basis of this topic systematically, and then introduces the existing common storage technology. Ethernet and the Ethernet technology of SAN network are compared and analyzed in detail. The advantages of the ethernet technology of integrating SAN network are obtained. It is determined that the integration of SAN and Ethernet should be realized based on the centralized architecture of SDN. Then the framework design, process design and software design are carried out, and each important module in the process is designed and implemented separately. Then the process integration, reliability analysis and the optimization of execution efficiency are carried out. Finally, according to the characteristics of this technology, a reasonable test scheme and test environment are put forward, and the corresponding tests are carried out. The practicability and performance of the whole scheme are analyzed according to the test results.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：TP393.11
，

本文編號：1549992