| 摘要 |
<P>PROBLEM TO BE SOLVED: To eliminate downtime of an on-line operation when a system is reconstructed by integrating, as a new standby node, an old operation node into the system that is temporarily separated from the system due to occurrence of a trouble. <P>SOLUTION: A DB server (21') of an old operation node (20') corrects a recovery log stored in a recovery log storage section (22b') by using a difference log stored in a difference log storage section (32c'). To explain specifically, a duplication controller (21b') and a DBMS (21a') correct content of a recovery log file according to result obtained by comparing a difference log file stored in the difference log storage section (32c') and the recovery log file stored in the recovery log storage section (22b') with each other. <P>COPYRIGHT: (C)2013,JPO&INPIT |
