Abstract: With the rapid development and mature of hardware techniques, software fault tolerance become the main reason of system dependability. However, fault tolerance system is hard to be developed directly. As well all know the development of system greatly benefits from the middleware for its convenient development environment and transparent communication mechanisms. So the paper puts forward a new kind of software fault tolerant structure based on middleware, and forms a complete solution project including some key techniques of fault tolerance, such as redundancy, fault detection and recovery, finally makes out its reliable degree which can verify the feasibility and reliability of this system.

Key words: Middleware, Fault tolerant technique, Fault detection, Reliability

摘要： 随着硬件容错技术的成熟，软件容错技术成为提高系统可靠性的热点问题。直接开发容错应用是非常困难的，鉴于中间件为应用系统提供了良好的开发环境，该文研究和设计了一个基于中间件的容错系统模型，提出了一种新的节点容错结构构造方法，为解决冗余、失效检测和恢复等容错的关键技术问题形成了一套较完整的解决方案。采用马尔科夫过程求出系统的可靠度，验证了系统设计的合理性和可靠性。

关键词: 中间件, 容错技术, 失效检测, 可靠性

CLC Number: 

• TP311