论文部分内容阅读
针对国内目前的容错系统研究现状，选择多机系统中的特例－双机系统作为研究对象，在以一致性断点为最基本的容错机制的基础上，提出了能够获得断点释放算法，详细讨论了这种算法的基本策略及其改进方案．通过分析故障进程的影响，考察了相关进程的发散效应的特点，根据相关进程之间的通信关系来获得断点的最小集合，并用模拟的方法对该算法进行了验证及性能分析．证明了该算法的确能够有效地消除在断点保留过程中所形成的无效数据，释放存储空间．
Aiming at the current research status of fault-tolerant systems in China, this paper chooses the special case-double-machine system in multi-machine system as the research object. Based on the consistent fault breakup as the most basic fault tolerant mechanism, a breakpoint release algorithm The basic strategy of this algorithm and its improvement are discussed. By analyzing the influence of the fault process, the characteristics of the divergent effect of the related processes are investigated. The minimum set of breakpoints is obtained according to the communication relationship among the related processes. The algorithm is verified and the performance is analyzed by simulation. It is proved that the algorithm can effectively eliminate the invalid data formed during the reservation of the breakpoint and release the storage space.