【摘要】：基于TCP/IP體系結(jié)構(gòu)的現(xiàn)行互聯(lián)網(wǎng)絡(luò)，在靈活性、可擴(kuò)展性和可管理性方面，表現(xiàn)出難以融入新技術(shù)和新業(yè)務(wù)的趨勢(shì)，呈現(xiàn)一種結(jié)構(gòu)僵化的特征。為了支持未來(lái)網(wǎng)絡(luò)創(chuàng)新，解決現(xiàn)行網(wǎng)絡(luò)存在的弊端，迫切地需要真實(shí)的網(wǎng)絡(luò)環(huán)境來(lái)測(cè)試驗(yàn)證未來(lái)網(wǎng)絡(luò)體系架構(gòu)和協(xié)議，研究者提出了網(wǎng)絡(luò)虛擬化的思想，并且得到未來(lái)網(wǎng)絡(luò)領(lǐng)域研究的廣泛關(guān)注。在網(wǎng)絡(luò)虛擬化的研究中，虛擬資源管理是一個(gè)關(guān)鍵問(wèn)題。 本文分析了NetFPGA的虛擬化交換機(jī)開(kāi)發(fā)與實(shí)現(xiàn)技術(shù)，在此基礎(chǔ)上，將XenServer虛擬化軟件與NetFPGA硬件相結(jié)合，，在XenServer DDK環(huán)境中開(kāi)發(fā)NetFPGA的驅(qū)動(dòng)模塊，并將驅(qū)動(dòng)模塊移植到XenServer Domain0中，構(gòu)建了4×1Gb/s的多端口物理交換機(jī)。論文進(jìn)一步通過(guò)橋接手段實(shí)現(xiàn)虛擬交換機(jī)與物理交換機(jī)的端口連接，構(gòu)建了基于XenServer的虛擬化物理交換機(jī)。通過(guò)兩個(gè)邏輯獨(dú)立的虛擬交換機(jī)的分配和部署實(shí)驗(yàn)，驗(yàn)證了虛擬化交換機(jī)的有效性。為了使虛擬交換機(jī)提供更好的網(wǎng)絡(luò)性能，論文設(shè)計(jì)了資源管理方案，該方案監(jiān)控交換機(jī)的物理資源和虛擬交換機(jī)資源使用情況，并與集中式資源管理節(jié)點(diǎn)協(xié)作，完成資源的查看和監(jiān)控功能。最后，針對(duì)虛擬交換機(jī)競(jìng)爭(zhēng)占用物理交換機(jī)帶寬資源的問(wèn)題，通過(guò)測(cè)試驗(yàn)證了XenServer下通用帶寬管理工具的可用性。
[Abstract]:The current Internet based on TCP/IP architecture shows the trend that it is difficult to integrate new technology and new business in flexibility, scalability and manageability, and presents a feature of rigid structure. In order to support the future network innovation and solve the drawbacks of the current network, the real network environment is urgently needed to test and verify the future network architecture and protocols. Researchers put forward the idea of network virtualization. And get the future network domain research widespread attention. Virtual resource management is a key issue in the research of network virtualization. This paper analyzes the development and implementation technology of NetFPGA virtualization switch. On this basis, combining XenServer virtualization software with NetFPGA hardware, we develop the driver module of NetFPGA in XenServer DDK environment, and transplant the driver module to XenServer Domain0. A 4 脳 1Gb/s multiport physical switch is constructed. In this paper, the port connection between virtual switch and physical switch is realized by bridging method, and a virtual physical switch based on XenServer is constructed. The effectiveness of the virtualization switch is verified by the allocation and deployment experiments of two logically independent virtual switches. In order to make the virtual switch provide better network performance, this paper designs a resource management scheme, which monitors the use of physical resources and virtual switch resources of the switch, and collaborates with the centralized resource management node. Complete resource viewing and monitoring functions. Finally, aiming at the problem of virtual switch competing to occupy the physical switch bandwidth resources, the availability of the universal bandwidth management tool under XenServer is verified by testing.
【學(xué)位授予單位】：南京郵電大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP393.05

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 謝高崗;張玉軍;李振宇;孫毅;謝應(yīng)科;李忠誠(chéng);劉韻潔;;未來(lái)互聯(lián)網(wǎng)體系結(jié)構(gòu)研究綜述[J];計(jì)算機(jī)學(xué)報(bào);2012年06期

本文編號(hào)：2343001