| 摘要 |
PURPOSE:To manage with less backup data necessary for recovery, by using an up-to-date check point log written on a disk for recovery for writing to a disk for operation when a fault occurs to the disk for operation. CONSTITUTION:When the operation of the system which uses an electronic computer 1a starts, and every time transaction is performed at time intervals specified by a system control table 2a or to a specific frequency, a check point is provided and a file which is an object of recovery is logged on the disk 5a for recovery at the check point. If a fault occurs, an up-to-date check point is searched for in the table 2a to read an up-to-date check point log out of the disk 5a; and it is written on the disk 4a for operation and the contents of the disk 4a are returned to the up-to-date check point. Then, only the transaction for file update from the up-to-date log corresponding to the up-to-date check point is executed. Further, if a fault occurs to the disk 5a, logging to the disk 5b is carried out to allow the disk 4a to recover. |
