The backup script should be redirected to the Log and saved so as to find useful information in case of an error.
Optimization related to RMAN backup
To back up data, follow these steps:
1. read data from the disk
2. process data blocks in memory
3. Write Data to disk or tape
The preceding read/write operations can be completed synchronously or asynchronously. In synchronous I/O operations, only one IO operation is allowed at a time, but in asynchronous I/O operations, multiple I/O operations are allowed at a time. Therefore, backup and recovery optimization is mainly focused on the following aspects:
1. Improve synchronous or asynchronous I/O operations
In an operating system that supports asynchronous operations, you can set TAPE_AYSNCH_IO, DISK_ASYNCH_IO, and BACKUP_TYPE_IO_SLAVES to support asynchronous operations and improve the write capability.
2. Improve disk read capability
After the backup command, you can set DISKRATIO to ensure that data is read from multiple disks and continuous data streams are ensured.
3. Set the buffer and parameter values correctly.
Set LARGE_POOL_SIZE to enable continuous buffer pools for backup. You can increase the utilization of the buffer zone by setting DB_FIL_DIRECT_IO_COUNT. If you use tape backup, you can also set BACKUP_TYPE_IO_SLAVES to improve the write capability of the tape.
4. Parallel backup
Multiple channels can be opened for parallel backup and recovery
. Back up the RMAN Database
RMAN databases also need to be backed up, but they are small and do not change frequently. Therefore, after each RMAN backup, you can use the following script to back up the RMAN database.
EXP pafile = exprman. SQL
Exprman. SQL is
USERID = RMAN/RMAN
Buffer= 32768
OWNER = RMAN
FILE = RMAN. DMP
ROWS = Y
GRANTS = Y
COMPRESS = Y
CONSISTENT = Y