The Backup and DR Service Restore operation restores a backup image to the source, overwriting whatever data exists there.
System limitations and workarounds
System databases on a root partition backed up as logical volume manager (LVM) snapshots cannot be used in a traditional restore operation because the root partition cannot be unmounted. These require manual restore and recovery from a standard mount back to the same host.
To recover a volume-level database image with less downtime for users, see Mount and migrate other types of databases for instant recovery.Restoring back to the source is not supported if multiple instances share the same volume or file systems. To restore such applications, mount the image to the host and use the procedure to perform single database recovery detailed in Restore a single database from a volume-based backup image to the source.
If there are nested mount points under the production volumes being backed up, then restore and migrate operations back to the source fails because the production volumes are busy and cannot be unmounted.
To restore /backup-disaster-recovery/docs/restore-data/otherdb-restore
Restore databases from a volume-level backup image to the source
This procedure uses physical recovery of the source data area. To recover back to the source, follow these instructions:
From the App Manager Applications list, right-click the protected database and select Access. Use the Managed Backup Plan status filter to show only protected databases.
Select a snapshot image and click Restore.
Select Traditional—not mount and migrate.
If the source application is protected by a snapshot policy that has enabled database log backups, and logs are available with the image, you can use them to roll forward to a specific point in time by changing these options in the Roll Forward Time section:
- The date field contains all possible dates that the database can be rolled forward to—through the application of database transaction logs. Select which date you need the database to be rolled forward to.
- The time field contains a slider showing all possible times on the selected date that the database can be rolled forward to. If you select the latest possible date and then move the slider to the right most position, the restore job applies to all available logs. If you select the earliest possible date and move the slider to the left most position, the restore job applies no logs.
- You can specify to roll forward using either User Time or Host Time. User Time is relative to the local time of the current user. Host time is relative to the system that hosts the data to be restored.
Enable Restore With Recovery to apply recovered logs.
Click Submit.
```sh
ALTER DBSPACE IQ_SYSTEM_LOG RENAME /pitr_log_location
SET OPTION PUBLIC.IQ_POINT_IN_TIME_RECOVERY_LOGGING = 'ON'
```
Restore a single database from a volume-based backup image to the source
To restore a single Db2 or SAP ASE backup image to its source, follow these steps:
From the App Manager Applications list, right-click the protected database and select Access.
Select the latest snapshot to recover and click Mount.
In the Application Options, disable Create New Virtual Application.
In Mapping Options, provide the mount-point location.
For example, using
/mymount
mounts the database backup under this location. The log backup is mounted under/mymount_archivelog
.Click Submit.
Check the Monitor > Jobs page to see when the mount job is finished.
When the job is finished, log into the database server as root. On the server, change directory to
/act/custom_apps/<var>database type</var>/restore
.Get the
JobID
of the mount from/var/act/log/UDSAgent.log
. To find theJobID
, run the following command:grep "mount -t " /var/act/log/UDSAgent.log | grep -w "<var>mountpoint from step 4</var>"|tail -1
For example:
grep "mount -t " /var/act/log/UDSAgent.log | grep -w "/db2mnt" |tail -1 2019-11-18 23:59:19.740 GEN-INFO \[22488\] **Job_0404207** Spawning cmd: mount -t ext4 /dev/act403764_DBDump_1574101677612/act_staging_vol /db2mnt 2>&1
ARCHIVELOG_MNT
is<var>mountpoint provided in step 4</var>_archivelog
.From the target host command line as root, run the script:
IBM Db2
Script: act_db2_lvm_customdb_recovery.sh
Arguments to the script:
SOURCE_INSTANCE = <var>Db2 Instance name</var>
DB_NAME=<var>Db2 Database name to be recovered(Single)</var>
TARGET_MNT = <var>Db2 Database image mountpoint name</var>
ARCHIVELOG_MNT= <var>Archive Log backup mount point name</var>
UNTIL_TIME = <var>Recovery Time(Format: "YYYY-MM-DD-HH.MI.SS")</var>
JOBID = <var>Database mount Job name</var>
Connect to the Db2 instance and confirm that the databases are recovered and online.
db2 connect to <var>dbname</var>
db2 select db_status FROM SYSIBMADM.SNAPDB
SAP ASE
Run the script act_sybase_lvm_customdb_recovery.sh with these arguments.
./act_sybase_lvm_customdb_recovery.sh OSUSER=sybase
TARGET_SYBASE_SQLD=/home/sybase/Sybase16Home/OCS-16_0 TARGET_MNT_PNT=/sngRst
TARGET_SERVER_NAME=ASE1 TARGET_DB_USER=sa STRIPEON=4 TARGET_DBUSER_PASSWD=sybase
SRC_DBNAME=CU1 LOG_BKP_MNTPT=/sngRst_archivelog UNTIL_TIME="2019-11-07 20:31:27"
BEGIN_TIME="2019-11-07 19:31:27" JOBID="Job_2677627"
Arguments to the script
OSUSER = SAP Ase OS owner name
TARGET_SYBASE_SQLD = SAP ASE iSQL path on the target recovery host
TARGET_MNT_PNT = SAP ASE Instance image mountpoint name
TARGET_SERVER_NAME = SAP ASE data server name on the target recovery host
TARGET_DB_USER = SAP ASE Instance username on the target recovery host
TARGET_DBUSER_PASSWD = SAP ASE Instance user password on the target recovery host
SRC_DBNAME = SAP ASE Database name to be recovered (Single)
LOG_BKP_MNTPT = SAP ASE Log image mountpoint name
BEGIN_TIME= Backup begin time (Format: "YYYY-MM-DD HH24:MI:SS")
UNTIL_TIME = Point in time to recover the database (Format: "YYYY-MM-DD HH24:MI:SS")
JOBID = Database mount Job name
Connect to the SAP ASE database and verify the data.
- In the management console, access the image again and Unmount+Delete the database mount point.
Restore a file-based Full+Incremental backup image to the source
This procedure overwrites the source data. To restore the source database from a file-based backup image, follow this procedure:
From the App Manager Applications list, right-click the protected database and select Access.
Select a snapshot image and click Restore.
Select Traditional—not mount and migrate.
Use Select Items to choose one or more databases to restore.
Enable Restore With Recovery to apply all recovered logs.
Click Submit. This starts the source database physical recovery using the database's recovery API.
ALTER DBSPACE IQ_SYSTEM_LOG RENAME '/<var>pitr_log_location</var>'
SET OPTION PUBLIC.IQ_POINT_IN_TIME_RECOVERY_LOGGING = 'ON'