智能网络存储系统（INSS）存储管理技术研究

发布时间：2018-03-12 06:37

本文选题：智能网络存储系统　切入点：INSS　出处：《华南理工大学》2012年硕士论文　论文类型：学位论文

【摘要】：进入21世纪以来，随着网络存储技术的飞速发展，存储设备和存储网络的管理问题日益成为存储行业研究的热点。为了解决异构存储系统之间互操作性差、管理困难、费用高昂等问题，全球存储工业协会（SNIA）于2002年提出了标准化的存储管理规范（SMI-S），主要目的是为存储设备和存储管理软件之间提供标准化的通信方式，实现异构存储设备的统一管理，提高存储管理的灵活性和可靠性，降低管理成本。本文在认真分析了当前主流网络存储系统结构和存储管理技术的基础上，重点对课题组提出的新型智能网络存储系统（IntelligentNetworkStorageSystem，INSS）及其存储管理技术进行了详细的研究与设计。INSS综合了SAN、NAS和DAS的优点，通过智能化算法和虚拟化技术来解决存储容量扩展、单点失效和负载均衡等一系列目前在主流网络存储系统中普遍存在的问题，它具有高性能、高可靠性和智能化等特点，能够很好地满足用户对数据存取简单、高效等需求。为了更好地管理INSS，本文结合当前主流存储管理技术设计了INSS存储管理方案，初步实现了对INSS的监控与管理，提高了管理效率。以下是本文所做的主要研究工作： 1、分析和研究了与INSS存储管理相关的SMI-S规范、Windows管理规范（WMI）、基于网络的企业管理（WBEM）和公共信息模型（CIM）等存储管理规范。 2、根据INSS系统结构和心跳信息运行的不同位置，，提出了三层INSS心跳网络的系统结构，并设计了一种适用于INSS的自适应心跳算法，该算法通过阶段性检测心跳服务等级动态调整心跳周期，有效减少了网络存储管理的通信量，提高了INSS的高可用性。 3、在INSS自适应心跳算法的基础上，对INSS的负载均衡中的几个关键问题进行了研究，提出了一种分层动态负载均衡调度策略，为负载均衡算法提供了负载信息和任务调度支持。 4、对设计和实现的INSS存储管理系统进行了功能测试，并对测试结果进行了相应的性能分析。测试结果证明，本文提出的存储管理方法能够实时监控INSS存储设备和存储网络，较好地提高了INSS的高可用性和存储性能。
[Abstract]:Since 21th century, with the rapid development of network storage technology, the management of storage devices and storage networks has become a hot issue in storage industry. In order to solve the problem of poor interoperability between heterogeneous storage systems, management is difficult. In 2002, the Global Association of Storage Industries (Sanya) proposed a standardized storage management specification, SMI-SU, aimed at providing standardized communication between storage devices and storage management software. The uniform management of heterogeneous storage devices is realized, the flexibility and reliability of storage management are improved, and the management cost is reduced. Based on the analysis of the current mainstream network storage system structure and storage management technology, A new intelligent network storage system named Intelligent Network Storage system (INSS) and its storage management technology are studied and designed in detail. INSS integrates the advantages of SANNAS and DAS, and solves the problem of storage capacity expansion through intelligent algorithm and virtualization technology. Single point failure and load balancing are a series of common problems in the mainstream network storage system. It has the characteristics of high performance, high reliability and intelligence, and can satisfy the user's simple access to data. In order to manage INSS better, this paper designs a storage management scheme of INSS combined with the current mainstream storage management technology, which realizes the monitoring and management of INSS, and improves the management efficiency. The following is the main research work of this paper:. 1. This paper analyzes and studies the storage management specifications of INSS, such as SMI-S, Web-based Enterprise Management (INSS) and Common Information Model (INSS), and so on. 2. According to the structure of INSS system and the different position of heartbeat information, the system structure of three-layer INSS heartbeat network is proposed, and an adaptive heartbeat algorithm suitable for INSS is designed. The algorithm dynamically adjusts the heartbeat cycle by detecting the heartbeat service level, which effectively reduces the traffic of network storage management and improves the high availability of INSS. 3. Based on the INSS adaptive heartbeat algorithm, several key problems in INSS load balancing are studied, and a hierarchical dynamic load balancing scheduling strategy is proposed, which provides load information and task scheduling support for the load balancing algorithm. 4. The function of the INSS storage management system designed and implemented is tested, and the corresponding performance of the test results is analyzed. The test results show that the proposed storage management method can monitor the INSS storage devices and storage networks in real time. The high availability and storage performance of INSS are improved.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP333

【参考文献】