A System for Remote Multi-Program Checkpointing and Error Recovery

• Main Authors: Sheng-Kuei Wang, 王聖貴
• Other Authors: 蔡智強
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/99811736148925090070

碩士 === 國立中興大學 === 電機工程學系所 === 95 === The checkpoint fault-tolerant technique has wildly discussed in many research studies, especially at communications and databases designing and implementing. The basic conception of checkpoint is to preserve the state and data of the target running process at the moment in order to recover the state and data when the process has failed. The checkpoint and error recovery can be implemented on software or hardware, and cooperated with both is also usual way. However, implementing checkpoint and error recovery techniques in software fault-tolerance almost choose Unix/Linux OS as the implementing platform, in the other way select the Microsoft Windows OS as the implementing platform is comparably much less than common. This paper is implementing the checkpoint and rollback-recovery mechanism in software manner on Windows OS platform. The implementation can be divided into two parts. One is the computing of the checkpoint and error recovery, and another is the intercepting system API. I also implement multi-program checkpoint and periodical checkpoint in this paper. In multi-program checkpoint, the checkpoint information should be saved as program related file; in periodical checkpoint, a sample interface is provided by the remote console and the user use it that choose different checkpoint intervals for different application programs. Finally by TCP/IP networking and remote checkpointing server interaction work pattern, the main goal is to construct a Client/Server checkpoint foundation model in Windows networking environment.