In the Oracle environment,
“restore” means to replace a damaged or missing file with a backup;
“recover” means to synchronize the file with the rest of the database by use of archivelogs.
■Recovery Structures and Processes
・controlfile
CREATE CONTROLFILE
・online redo logfile
ALTER DATABASE CLEAR LOGFILE GROUP # command (where # is the number of the group with a damaged member)
ALTER DATABASE CLEAR UNARCHIVED LOGFILE GROUP #
・datafile
A damaged datafile requires use of backups and archivelogs.
■Instance recovery process
To open a database, all the controlfile copies, at least one member of each online
logfile group, and all the online datafiles must be present and synchronized.If,during a startup, SMON finds that this is not the case, the startup will not complete.
①If a controlfile copy is damaged or missing, the startup will stop in NOMOUNT mode.A message is written out to the alert log detailing which copy (or copies) of the controlfile is damaged.
②Assuming that the controlfiles are fine, SMON proceeds to open the database. During this phase, it checks the headers of all the online datafiles.If any are missing or damaged, appropriate error messages are written out to the alert log, and the database remains in mount mode.
③If all the online files are present and not damaged, but one or more of them are not synchronized, SMON will attempt to synchronize them by using the online redo logs
■Recovery from Media Failure
①Recovery from Loss of a Multiplexed Controlfile
Just replace it with a surviving copy of the controlfile.
②Recovery from Loss of a Multiplexed Online Redo Log File
For open recovery, use the ALTER DATABASE CLEAR LOGFILE command to delete
the existing files (or at least, those that still exist) and create new ones
③Recovery from Loss of Datafiles
阅读(365) | 评论(0) | 转发(0) |
