高可用服务器故障管理板的设计与实现

发布时间：2018-01-10 12:20

本文关键词：高可用服务器故障管理板的设计与实现　出处：《哈尔滨工业大学》2012年硕士论文　论文类型：学位论文

【摘要】：服务器已经被广泛应用于电信、银行、航天、气象、军事等领域。在这些领域中，服务器作为关键设备承担着主要的工作任务。但是无论是在航天、军事等国防领域还是在电信、银行等民用领域，决定这些行业选择服务器的一个重要因素就是服务器的可用性。如何保证服务器的可用性，保障其持续工作的能力已经成为人们的重要研究课题。服务器的管理是保证服务器可用性的关键，为实现服务器的有效管理，设计了高可用服务器故障管理平台。本文结合故障管理的一般过程，分析了自主计算的相关概念和管理模型，，讨论了故障管理中需要解决的问题，并对故障监控、故障诊断和故障恢复等技术进行了研究。在此基础上，论文针对具有分布式结构的高可用目标系统设计了故障管理方案，采用双层架构实现服务器故障管理平台对服务器进行管理。管理平台由本地管理模块LMM和全局管理模块GMM组成。LMM对应于每个节点，负责直接对计算节点的状态进行监控；全局有两个GMM，负责对LMM收集的状态信息进行分析、诊断，然后对故障进行恢复。本文设计实现了服务器故障管理的硬件平台，并在其基础上设计了服务器故障管理的软件框架，结合硬件平台，实现了服务器状态信息的采集功能，包括从带内和带外两方面角度对服务器的状态进行采集；研究了故障定位和故障诊断的过程，实现了故障诊断的基本功能；制定了部分故障的管理策略；提供了用户远程管理的接口。最后本文对管理平台的功能做了简单的测试。
[Abstract]:The server has been widely used in the fields of telecom , banking , aerospace , meteorology , military , etc . In these fields , the server plays the main task as a key equipment . However , it is an important factor to determine the server availability in the fields of defense , such as space , military and other civilian areas . How to guarantee the availability of the server and guarantee the ability of its continuing work has become an important research topic . The management of the server is the key to ensure the availability of the server . In order to realize the effective management of the server , a high - availability server fault management platform is designed . Based on the general process of fault management , the paper analyzes the relevant concepts and management models of the self - calculation , discusses the problems that need to be solved in the fault management , and controls the fault monitoring , fault diagnosis and fault recovery . The management platform is composed of the local management module LMM and the global management module gmm . The management platform is composed of the local management module LMM and the global management module gmm . The management platform is composed of a local management module LMM and a global management module gmm . The management platform is composed of a local management module LMM and a global management module gmm . In this paper , the hardware platform of server fault management is designed , and the software framework of server fault management is designed on the basis of it , and the collection function of server state information is realized by combining hardware platform . The fault location and fault diagnosis process is researched , the basic function of fault diagnosis is realized , and the interface of remote management is provided . Finally , the function of management platform is tested .

【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP368.5

【参考文献】