| With the communication and the rapid development of computer technology, more and more computer systems used to provide a variety of timely and reliable service, how to ensure that a computer system for reliable, stable and lasting is the key problem to be solved, it requires the system with redundancy and fault tolerance.In this paper, factors that affect system availability based on the analysis, taking into account the system cost and flexibility of enterprise application and program factors and designed based on the primary disk array free from the structure of Fault-tolerant hardware platform running the software.Firstly, an overview of the development of fault-tolerant computer systems, and domestic and international situation, on the main issues are briefly introduced. Secondly, the hot standby model and the main technical work were reviewed to compare the two hot standby way to achieve: Shared storage and data synchronization. Third, the paper gives the main technical Fault-tolerant software, one of it is single system image and fault diagnosis and recovery technology. Fourthly, according to the work of hot standby models and technical analysis, the design will be divided into the heartbeat detection module Fault-tolerant software and resource management module in two parts. Heartbeat detection module including the node fault detection, system services fault detection, resources fault detection while maintaining the double connection. Resource management module to achieve master / slave switch, resources state information and so on. The system is two-node cluster system as the goal, but also has good scalability and portability. Finally, the realization and description of testing the basic functions of Fault-tolerant software, and video switching for the double application experiments. The results show that the system can maintain stable operation of the basic functions of the core business of enterprises, to meet the design objectives. |
PDF Full Text Request