摘要
设置检查点是容错计算机系统进行故障恢复的重要手段。因为检查点间隔选择过大或过小都将使系统性能受到影响,所以检查点间隔的适当选定是系统性能优化的一个重要指标。该文针对双机容错系统,采用检查点设置与回卷恢复的方法提出了一种系统模型,利用马尔科夫链得到了最佳检查点间隔的求解等式,通过实验证实了求解等式的正确性。
Checkpointing is one of the most important method for fault tolerant computer to recover from faults. Too big or too small checkpoint interval maybe degrade the performance of system, so proper determination of checkpoint interval can make system performance optimized. This paper pnsents a duplicated fault tolerance system with the methods of setting checkpoints and rollback recovery, and achieves an equation about the best checkpoint interval through the Markov chain. In the end, the correctness of this conclusion through experiment is testified.
出处
《计算机工程》
CAS
CSCD
北大核心
2007年第5期283-284,F0003,共3页
Computer Engineering
基金
国家自然科学基金资助项目(NSFC60473031)
湖南省自然科学基金资助项目(05JJ30116)
关键词
双机容错
回卷恢复
检查点间隔
Duplicated fault tolerance
Rollback recovery
Checkpoint interval
