ETERNUS SF Recovery Manager for Oracle EE 13.0.1 User's Guide - Solaris(TM) Operating System - |
Contents
![]() ![]() |
This chapter explains the backup function of Recovery Manager for Oracle.
Recovery Manager for Oracle provides the following backup functions:
This section explains the procedure for online backup of the database.
The function backs up files online using the "/opt/FJSVrmfo/bin/rmfobackup" command.
The command is executed as root from the target database server.
# /opt/FJSVrmfo/bin/rmfobackup -b DB_NAME |
Option |
Explanation |
-b |
Execute the online backup operation |
Operand |
Explanation |
DB_NAME |
Specify the database name to be backed up online |
= 0: Normal ending
> 0: Abnormal ending
Example: Performs backup for "database name: orcl" online
# /opt/FJSVrmfo/bin/rmfobackup -b orcl |
This section explains the procedure for removing archived redo log files no longer required.
The function removes the archived redo log files that have exceeded the number of days specified in the configuration (the number of backup generation managements (BK_GEN)) using the"/opt/FJSVrmfo/bin/rmfobackup" command.
The command is executed as root from the target database server.
Ensure this procedure is followed, as the next backup operation may end in error if archived redo log files are removed using the rm command.
# /opt/FJSVrmfo/bin/rmfobackup -a DB_NAME |
Option |
Explanation |
-a |
Remove expired archived redo log files |
Operand |
Explanation |
DB_NAME |
Specify the name of the database where the expired archived redo log files are to be removed |
= 0: Normal ending
> 0: Abnormal ending
Example: Remove the expired archived redo log files of "database name: orcl"
# /opt/FJSVrmfo/bin/rmfobackup -a orcl |
The function is configured to wait for completion of the OPC physical copy after executing the online backup operation when OPC has been selected as the data file backup method (BK_TYPE).
To wait for completion of the OPC physical copy, execute the " /opt/FJSVrmfo/bin/rmfobackup" command.
The command is executed as root from the target database server.
# /opt/FJSVrmfo/bin/rmfobackup -w DB_NAME |
Option |
Explanation |
-w |
Wait for completion of the OPC physical copy |
Operand |
Explanation |
DB_NAME |
Specify the database name which will wait for completion of the OPC physical copy |
= 0: Normal ending
> 0: Abnormal ending
Example: Waits for completion of the OPC physical copy of "database name: orcl"
# /opt/FJSVrmfo/bin/rmfobackup -w orcl |
Confirm from the following information that the backup command has completed.
When the backup command ends normally, the return value will be 0.
An error has been generated when a value other than 0 is returned.
The following message is output to syslog if the online backup operation (rmfobackup -b) has ended in error.
Example: An online backup error message (rmfobackup -b)
Mar 22 11:22:59 rmfo200 FSP_RMGRFO_backup: [ID 702911 user.error] 'Abnormal End: 00074: The backup operation of the database (ora920) ended abnormally.' |
The following warning message is output in the cluster operation (RAC) if some of the nodes have been terminated.
Note: If the user.warning message is inhibited in the syslog message processing, the warning message will not be output.
Example: Warning message output from the online backup operation (rmfobackup -b)
Mar 29 15:34:41 rac001 FSP_RMGRFO_backup: [ID 702911 user.warning] 'WARNING: The instance(rac002) is not activated. Continue the process.' |
The function outputs detailed information to a log file at the time of backup command execution.
Log files are created under the "/opt/FJSVrmfo/log/backup/DB_NAME" directory.
The log file name is as follows:
bup1_YYYYMMDDHHMISS.log |
Online backup execution log and execution log for waiting for completion OPC physical copy |
bup2_YYYYMMDDHHMISS.log |
execution log for removing the expired archived redo log files |
YYYYMMDDHHMISS : Time of execution (year/month/day/hour/minute/second)
Refer to the log file if the backup command ended in error, remove the cause of the error, and re-execute the backup operation immediately.
Command execution log files that have exceeded the backup execution log storage period (LOG_GEN) are deleted at the time of backup operation.
This section explains general backup operations using Recover Manager for Oracle.
Perform the backup operation as follows:
Backup online the target database disk to another disk, as directed in Performing Online Backup.
Execute as directed in Waiting for Completion of the OPC Physical Copy if the OPC backup method has been selected, and files are to be backed up to a tape, after completion of the OPC physical copy.
Recovery Manager for Oracle does not provide a function that backs up data to a tape.
By backing up data to tapes, risks may be dispersed.
To backup data in file system images from a server other than the target database server, always unmount the backup volume after completing the backup operation. If mounted in duplication, the file system will be destroyed.
To backup to a tape from a server other than the target database server, backing up in raw images is recommended.
The following volumes require tape backup:
For details on these volumes, refer to the examples of disk resource arrangement in Disk Resources Allocation.
After completing the tape backup, remove archived redo log files that have exceeded the number of generations being managed in the target database as directed in the section, Removing Expired Archived Log File
Examples of the Configuration file are shown below.:
An example of a configuration file is below.:
Conditions:
DB_NAME=orcl |
change_on_install |
Not required
BK_TYPE=1 |
/orcl/bk_dbf1|/dev/dsk/c5t16d20s0|1| |
/orcl/bk_ctl|/dev/dsk/c5t16d22s0|1| |
Not required
|/orcl/bk_arch|/dev/dsk/c5t16d23s0|1| |
R_DB_NAME=r_orcl |
An example of cluster operation (HA) is shown as follows:
Conditions:
DB_NAME=ha10g |
change_on_install |
hasvr001|ha10g|ha10g_app1| |
BK_TYPE=2 |
/ha10g/dbf1|/dev/sfdsk/ha10g/dsk/dbf1:c2t16d133|/dev/dsk/c2t16d172s0|/dev/dsk/c2t16d173s0|| |
/ha10g_bk/ctl|/dev/dsk/c2t16d172s4|1| |
Not required
|/ha10g_bk/arch|/dev/dsk/c2t16d174s7|1| |
R_DB_NAME=r_ha10g LSNR_NAME=LISTENER_R_HA10G PORT=1526 R_TAR_MP=/ha10g/r_ha10g R_TAR_DEV=/dev/dsk/c2t16d172s7 R_BK_TYPE=2 R_BK_MP=NONE R_BK_DEV=/dev/dsk/c2t16d173s7 |
An example of cluster operation (RAC) is shown as follows:
Conditions:
DB_NAME=rmfo |
change_on_install |
rac001|rmfo01|app_rmfo01| |
BK_TYPE=3 |
SYSTEM|/dev/sfdsk/rac10g/dsk/system:c3t16d3|/dev/dsk/c3t16d13s0|/dev/dsk/c3t16d15s0|/dev/dsk/c3t16d17s0| |
/rmfo_bk/ctl|/dev/dsk/c3t16d24s3|1| |
fjh001|/rac10g/arch1|/dev/sfdsk/rac10g/dsk/arch1|4| |
rac001|/rmfo_bk/arch1|/dev/dsk/c3t16d19s0|1| |
R_DB_NAME=r_rac10g |
Contents
![]() ![]() |