两级RAID高可靠分布式iSCSI网络存储系统
发布时间：2018-08-17 11:43
【摘要】：全球经济已进入后工业革命时代,即信息革命时代。信息数据作为这场革命重要的载体,具有越来越重要的意义。企业和个人在日常经济生活中产生了大量的信息数据,这些数据都需要保存在储存系统当中。随着数据量的爆炸增长,存储系统的不断增大,系统越大,多块磁盘同时损坏的可能性也越大,系统的可靠性显得越加重要。美国911恐怖事件,日本海啸等大型人为自然灾害,让人们更加深刻地认识到灾难对数据的毁灭性损害,更加重视系统的容灾能力。目前存储系统的编码布局方案多集中于单冗余或二冗余存储,为满足日益增长的需求,研究多冗余网络存储系统显得十分必要。 本文着重于存储系统可靠性,提出了一种新型的基于iSCSI两级RAID分布式高可靠存储系统,此系统具有极好的可靠性,亦可避免地震等灾害对数据造成的不可恢复损害。本文阐述了主流RAID技术的优缺点,对目前主流的NAS、SAN和IP-SAN技术进行了分析,特别是IP-SAN中的iSCSI技术。 主要工作分为以下两个方面： 1.提出了基于新的存储结构的系统,两级RAID高可靠分布式iSCSI存储系统,重点建立了基于马尔可夫过程的数学模型。模型复杂,状态多,根据两级RAID系统的特性,利用Matlab符号计算,计算得出系统可靠性,论证了两级RAID系统具有极高的数据可靠性,极大优于以往的系统。 2.利用Linux实现了实验系统的搭建,利用IOMeter测试工具,测试了两级RAID分布式高可靠存储系统的数据读写性能,测试了系统并发性,并通过实验分析条带单元、子节点带宽、条带组合对系统性能的影响,最后阐述了两级RAID系统可应用领域。与其他系统相比,该系统兼具良好读写性能,高可靠性和分布式异地存储三个优点,具有高可靠性与高容灾能力。
[Abstract]:The global economy has entered the post-industrial revolution era, that is, the information revolution era. As an important carrier of this revolution, information data has more and more important significance. Enterprises and individuals generate a lot of information data in their daily economic life, which need to be stored in the storage system. With the explosion of data and the increasing of storage system, the larger the system is, the more likely it is that multiple disks will be damaged at the same time. The reliability of the system becomes more and more important. The September 11th terrorist event in the United States, the Japanese tsunami and other large-scale man-made natural disasters have made people more deeply aware of the devastating damage to the data caused by the disaster and pay more attention to the disaster tolerance ability of the system. At present, the coding layout of storage system is mostly focused on single or double redundant storage. In order to meet the increasing demand, it is necessary to study the multi-redundancy network storage system. This paper focuses on the reliability of storage system, and proposes a new distributed high reliability storage system based on iSCSI two-level RAID. This system has excellent reliability and can avoid the unrecoverable damage caused by earthquake and other disasters to data. This paper describes the advantages and disadvantages of the mainstream RAID technology, and analyzes the current mainstream NASSAN and IP-SAN technologies, especially the iSCSI technology in IP-SAN. The main work is divided into the following two aspects: 1. Based on a new storage structure, a two-level RAID high reliability distributed iSCSI storage system is proposed, and a mathematical model based on Markov process is established. The model is complex and has many states. According to the characteristics of the two-stage RAID system, the system reliability is calculated by using the Matlab symbol calculation. It is proved that the two-stage RAID system has extremely high data reliability, which is much better than the previous system. 2. The experiment system is built with Linux, and the data reading and writing performance of two-level RAID distributed high reliability storage system is tested by using IOMeter test tool. The system co-generation is tested, and the band unit and bandwidth of sub-nodes are analyzed by experiment. The effect of strip combination on system performance is discussed. Finally, the application field of two-stage RAID system is described. Compared with other systems, the system has three advantages, such as good read and write performance, high reliability and distributed remote storage, and has high reliability and disaster tolerance.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP333
本文编号：2187528
[Abstract]:The global economy has entered the post-industrial revolution era, that is, the information revolution era. As an important carrier of this revolution, information data has more and more important significance. Enterprises and individuals generate a lot of information data in their daily economic life, which need to be stored in the storage system. With the explosion of data and the increasing of storage system, the larger the system is, the more likely it is that multiple disks will be damaged at the same time. The reliability of the system becomes more and more important. The September 11th terrorist event in the United States, the Japanese tsunami and other large-scale man-made natural disasters have made people more deeply aware of the devastating damage to the data caused by the disaster and pay more attention to the disaster tolerance ability of the system. At present, the coding layout of storage system is mostly focused on single or double redundant storage. In order to meet the increasing demand, it is necessary to study the multi-redundancy network storage system. This paper focuses on the reliability of storage system, and proposes a new distributed high reliability storage system based on iSCSI two-level RAID. This system has excellent reliability and can avoid the unrecoverable damage caused by earthquake and other disasters to data. This paper describes the advantages and disadvantages of the mainstream RAID technology, and analyzes the current mainstream NASSAN and IP-SAN technologies, especially the iSCSI technology in IP-SAN. The main work is divided into the following two aspects: 1. Based on a new storage structure, a two-level RAID high reliability distributed iSCSI storage system is proposed, and a mathematical model based on Markov process is established. The model is complex and has many states. According to the characteristics of the two-stage RAID system, the system reliability is calculated by using the Matlab symbol calculation. It is proved that the two-stage RAID system has extremely high data reliability, which is much better than the previous system. 2. The experiment system is built with Linux, and the data reading and writing performance of two-level RAID distributed high reliability storage system is tested by using IOMeter test tool. The system co-generation is tested, and the band unit and bandwidth of sub-nodes are analyzed by experiment. The effect of strip combination on system performance is discussed. Finally, the application field of two-stage RAID system is described. Compared with other systems, the system has three advantages, such as good read and write performance, high reliability and distributed remote storage, and has high reliability and disaster tolerance.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP333
【参考文献】
相关期刊论文 前5条
1 刘卫平;蔡皖东;;一种基于iSCSI的网络RAID设计与实现[J];计算机工程与应用;2006年14期
2 彭铮良;IPSec网络安全体系结构与VPN的安全性[J];上海微型计算机;2000年10期
3 张本宏;陆阳;韩江洪;魏臻;;“二乘二取二”冗余系统的可靠性和安全性分析[J];系统仿真学报;2009年01期
4 蔡皖东;基于SAN的高可用性网络存储解决方案[J];小型微型计算机系统;2001年03期
5 任劲,谢长生,李为;iSCSI协议及其Linux下的实现[J];小型微型计算机系统;2003年07期
相关博士学位论文 前1条
1 张晓;基于存储区域网络的数据可靠性技术研究[D];西北工业大学;2006年
相关硕士学位论文 前1条
1 陈大恒;基于iSCSI协议的IP SAN网络存储技术研究[D];国防科学技术大学;2006年
，本文编号：2187528
本文链接：https://www.wllwen.com/kejilunwen/jisuanjikexuelunwen/2187528.html
