Loading...

Recover Replicated Data - How To

Topics How To  
Table of Contents

Automatically Mount a Recovery Point Using CDR

Before You Begin

Required Capability: See Capabilities and Permitted Actions

To configure a Replication Set so that its Recovery Points are automatically mounted:

  1. From the CommCell Browser, right-click a Replication Set for which at least one Replication Pair has already been created, and select Properties.
  2. From the Replication Options tab of the Replication Set Properties screen, select the Auto-Mount snapshots after Recovery Point Creation option.
  3. Click Enter Mountpoints.
  4. In the Enter Mount Points dialog box, click the Mount Point column for each Destination Path and do one of the following:
    • For Windows, do one of the following for each snapshot to be mounted:

      Specify a drive letter (e.g., G:) to which the snapshot will be mounted. Ensure that you specify an unused drive letter.

      Type a mount point name (e.g., G:\mountpoint1) to which the snapshot will be mounted.

    • For UNIX, type the path to the symbolic link that points to the directory where the snapshot is mounted.
  5. Click OK.

Manually Mount or Share the Snapshots of an Existing Recovery Point

Before You Begin

Required Capability: See Capabilities and Permitted Actions

To mount or share the snapshots of an existing Recovery Point, on Windows:

  1. In the CommCell Browser, perform a Browse operation for one of the following:
    • Replication Set
    • CDR Agent
    • Client
    • Application iDataAgent (Exchange, SQL, Oracle)
  2. Do one of the following, depending on the object you browsed:
    • From the Recovery Points dialog box, select a Recovery Point and click View Snapshots.

      In the Snapshots dialog box, select a snapshot and click Mount.

    • In the Browse dialog box, right-click the snapshot to be mounted, and select Mount Snapshot.
  3. Do one of the following:
    • specify a drive letter (e.g., G:) to which to mount the Recovery Point. Ensure that you specify an unused drive letter.
    • type a mount point name (e.g., X:\mountpoint1) for the Recovery Point.

    For Windows 2003, Recovery Point snapshots created with VSS, cannot be mounted on a mount point (e.g. X:\mountpoint1), this needs to be mounted to a drive letter (e.g., G:).

  4. If you want to create a share, do one of the following, depending on which object you browsed:
    • In the Snapshots dialog box, select a mounted snapshot and click Share.
    • In the Browse dialog box, right-click the mounted snapshot to be shared, and select Create Network Share.

    In the Enter Share Name dialog box, type a name for the share and select Make share writeable option. Click OK.

  5. Close the Snapshots dialog box or Browse window.

Automatically Mount Snapshots for Fan-In Recovery Point

Before you Begin

  • Review How to Setup Fan-In Configuration
  • You can configure CDR so that the snapshots that comprise the latest Fan-In Recovery Point will be automatically mounted when it is created. This will affect all Replication Sets that use the computer as a Fan-In destination. Alternately, once Recovery Points have been created, you can select an existing Recovery Point and mount its snapshots.

Required Capability: See Capabilities and Permitted Actions

To configure CDR to automatically mount snapshots for Fan-In Recovery Points:

  1. From the CommCell Browser, right-click ContinuousDataReplicator on the destination client, and select Properties.
  2. From the Fan-In tab, select the Recovery Points tab, and select the Auto-Mount snapshots after Recovery Point Creation option.
  3. Click Enter Mountpoints.
  4. In the Enter Mount Points dialog box, click the Mount Point column for each Destination Path and specify a drive letter (e.g., G:) to which the snapshot will be mounted.

    • Ensure that you specify an unused drive letter.
    • Do not specify mount points for a UNC path. If you specify the mount point for a recovery point on a UNC path, this mount point is ignored when a recovery point is created on a UNC path.
    • When a Recovery Point snapshots created with ONTAP snap engine and Auto-Mount snapshots after Recovery Point Creation option is selected, share of snapshot will be created.

  5. Click OK.

Manually Mount or Share the Snapshots of an Existing Fan-In Recovery Point

Before you Begin

Required Capability: See Capabilities and Permitted Actions

To mount the snapshots of an existing Fan-In Recovery Point, on Windows:

  1. From the CommCell Browser, right-click ContinuousDataReplicator on the destination client, and select Browse Destination Volumes.
  2. In the Browse Destination Volumes dialog box, right-click a Recovery Point, and select Mount.
  3. Specify a drive letter (e.g., G:) to which to mount the Recovery Point. Ensure that you specify an unused drive letter.
  4. To share a snapshot on ONTAP, do the following, depending on which object you browsed:
    • In the Browse window, right-click the mounted snapshot to be shared, and select Create Network Share.
    • In the Enter Share Name dialog box, type a name for the share.
    • Select Make share writeable option. Click OK.
  5. Close the Browse window.

Unshare the Snapshots of a Recovery Point

Before You Begin

Required Capability: See Capabilities and Permitted Actions

To unshare the snapshots of a Recovery Point (On Windows only):

  1. In the CommCell Browser, perform a Browse operation for one of the following:
    • Replication Set
    • CDR Agent
    • Client
    • Application iDataAgent (Exchange, SQL, Oracle)
  2. Do one of the following, depending on which object you browsed:
    • From the Recovery Points dialog box, select a Recovery Point and click View Snapshots.

      In the Snapshots dialog box, select a shared snapshot and click Unshare.

    • In the Browse dialog box, right-click a mounted snapshot, and select Delete Network Share.
  3. Close the Snapshots dialog box or Browse dialog box.

Unmount the Snapshots of a Recovery Point

Before You Begin

Required Capability: See Capabilities and Permitted Actions

To unmount the snapshots of a Recovery Point:

  1. In the CommCell Browser, perform a Browse operation for one of the following:
    • Replication Set
    • CDR Agent
    • Client
    • Application iDataAgent (Exchange, SQL, Oracle)
  2. Do one of the following, depending on which object you browsed:
    • From the Recovery Points dialog box, select a Recovery Point and click View Snapshots.

      In the Snapshots dialog box, select a mounted snapshot and click Unmount.

    • In the Browse dialog box, right-click a mounted snapshot, and select Unmount Snapshot.
  3. Close the Snapshots dialog box or Browse dialog box.

Unmount a Fan-In Recovery Point Using CDR

Before You Begin

Required Capability: See Capabilities and Permitted Actions

To unmount the snapshots of a Fan-In Recovery Point:

  1. From the CommCell Browser, right-click ContinuousDataReplicator on the destination client, and select Browse Destination Volumes.
  2. In the Browse Destination Volumes dialog box, right-click a Recovery Point, and select Unmount.
  3. Close the Browse Destination Volumes dialog box.

Copyback File System Data from a Recovery Point or the Live Copy

Before You Begin:

Required Capability: See Capabilities and Permitted Actions

To Copyback file system data from a Recovery Point or the Live Copy:

  1. In the CommCell Browser, perform a Browse operation for one of the following:
    • Replication Set
    • CDR Agent
    • Client
  2. Do one of the following, depending on which object you browsed:
    • In the Recovery Points dialog box, select the Live Copy and/or Recovery Point(s) to be copied back, and click Copyback.
    • In the Browse dialog box, select the Live Copy and/or Recovery Point(s) to be copied back, and click Recover all Selected...
  3. In the Copyback dialog box, select a Recovery Host, the host to which the data will be copied back.
  4. Specify a Recovery Path for the live data and/or Recovery Point(s) you want to copy back. The Recovery Path must be larger than the data being copied back. To assign a Recovery Path, do one of the following:
    • double-click the path listed in the Recovery Path column and type a path on the selected host
    • click the Browse 'Recovery Host' button and browse to a destination path on the selected host
  5. Optionally, select Overwrite existing data under 'Recovery Path'.
  6. Click OK to copy back the data. The system displays a progress bar as it copies back the data. You can track the progress of the recover operation from the Job Controller dialog box.

Copyback File System Data from a Fan-In Recovery Point

Before You Begin

Required Capability: See Capabilities and Permitted Actions

To Copyback file system data from a Fan-In Recovery Point:

  1. From the CommCell Browser, right-click ContinuousDataReplicator on the destination client, and select Browse Destination Volumes.
  2. In the Browse Destination Volumes dialog box, right-click a Recovery Point, and select Copyback.
  3. In the Copyback dialog box, select a Recovery Host, the host to which the data will be copied back.
  4. Specify a Recovery Path for the Recovery Point(s) you want to copy back. The Recovery Path must be larger than the data being copied back. To assign a Recovery Path, do one of the following:
    • double-click the path listed in the Recovery Path column and type a path on the selected host
    • click the Browse 'Recovery Host' button and browse to a destination path on the selected host
  5. Optionally, select Overwrite existing data under 'Recovery Path'.
  6. Click OK to copy back the data. The system displays a progress bar as it copies back the data. You can track the progress of the recover operation from the Job Controller dialog box.

Copyback Exchange Data from a Consistent Recovery Point

Before You Begin

  • While it is possible to copy back Exchange data from the Live Copy, this data is not in a consistent state and thus application integrity is not ensured.

Required Capability: See Capabilities and Permitted Actions

To Copyback Exchange data from a Consistent Recovery Point:

  1. For Outlook 2003, change the profile for any mailboxes being restored which have "Cached Mode" enabled, to DISABLE CACHED MODE. "Cached mode" can be enabled again after the Copyback operation completes successfully.
  2. Abort the Replication Pairs for the source volume to which you will copy back Exchange data. This avoids slow copyback operation on the source volume.
  3. Dismount the Exchange Stores for the Storage Group to which you will copy back point-in-time data.
  4. Delete or rename all Exchange transaction logs on the Production Exchange Server for the Storage Group being restored.
  5. In the CommCell Browser, perform a Browse operation for one of the following:
    • Replication Set
    • CDR Agent
    • Client
    • Application iDataAgent (Exchange)
  6. Do one of the following, depending on which object you browsed:
    • In the Recovery Points dialog box, select the Consistent Recovery Point(s) to be copied back, and click Copyback.
    • In the Browse dialog box, select the Consistent Recovery Point(s) to be copied back, and click Recover all Selected...
  7. In the Copyback dialog box, select a Recovery Host, the host to which the data will be copied back.
  8. Specify a Recovery Path for the data you want to copy back. The Recovery Path must be larger than the data being copied back. To assign a Recovery Path, do one of the following:
    • double-click the path listed in the Recovery Path column and type a path on the selected host
    • click the Browse 'Recovery Host' button and browse to a destination path on the selected host
  9. Select Overwrite existing data under 'Recovery Path'.
  10. Click OK to copy back the data. The system displays a progress bar as it copies back the data. You can track the progress of the recover operation from the Job Controller dialog box.
  11. Mount the Exchange Stores.
  12. For Outlook 2003, the profile for any mailbox(es) changed to DISABLE CACHED MODE can be changed back to Cached Mode. The user will be able to see the recovered data instead of the cached copy.

Copyback SQL Data from a Consistent Recovery Point

Before You Begin

  • Review Copyback
  • This procedure assumes that you used Add App to create your Replication Pairs. For more information, refer to Add or Edit a Replication Pair and Application Integration.
  • While it is possible to copy back SQL data from the Live Copy, this data is not in a consistent state and thus application integrity is not ensured.

Required Capability: See Capabilities and Permitted Actions

To Copyback SQL data from a Consistent Recovery Point:

  1. Abort the Replication Pairs for the source volume to which you will copy back SQL data. This avoids slow copyback operation on the source volume.
  2. Detach the SQL database(s) for which you will copy back point-in-time SQL data.
  3. In the CommCell Browser, perform a Browse operation for one of the following:
    • Replication Set
    • CDR Agent
    • Client
    • Application iDataAgent (SQL)
  4. Do one of the following, depending on which object you browsed:
    • In the Recovery Points dialog box, select the Consistent Recovery Point(s) to be copied back, and click Copyback.
    • In the Browse dialog box, select the Consistent Recovery Point(s) to be copied back, and click Recover all Selected...
  5. In the Copyback dialog box, select a Recovery Host, the host to which the data will be copied back.
  6. Specify a Recovery Path for the data you want to copy back. The Recovery Path must be larger than the data being copied back. To assign a Recovery Path, do one of the following:
    • double-click the path listed in the Recovery Path column and type a path on the selected host
    • click the Browse 'Recovery Host' button and browse to a destination path on the selected host
  7. Select Overwrite existing data under 'Recovery Path'.
  8. Click OK to copy back the data. The system displays a progress bar as it copies back the data. You can track the progress of the recover operation from the Job Controller dialog box.
  9. Attach your SQL database(s).

Copyback Oracle Data from a Consistent Recovery Point

Before You Begin

  • Review Copyback
  • Review Using ContinuousDataReplicator with Oracle
  • This procedure assumes that you used Add App to create your Replication Pairs. For more information, refer to Add or Edit a Replication Pair and Application Integration.
  • While it is possible to copy back Oracle data from the Live Copy, this data is not in a consistent state and thus application integrity is not ensured.
  • Each Consistent Recovery Point for a given Oracle database comprises two snapshots, which will be referred to in the procedures as "first snapshot" and "second snapshot" according to the following definition of content:
    • The first snapshot includes all Oracle data volumes.
    • The second snapshot includes all Oracle log volumes and backup control file for the database at the time the Consistent Recovery Point was created.

Required Capability: See Capabilities and Permitted Actions

To Copyback Oracle data from a Consistent Recovery Point:

  1. Abort the Replication Pairs for the source volume to which you will copy back Oracle data. This avoids slow copyback operation on the source volume.
  2. Shut down the Oracle database.
  3. Copy back the following Oracle data from the second snapshot (the snapshot of any destination log volume) in the Consistent Recovery Point you have selected, to a temporary directory:
    • Backup.ctl.galaxy
    • dbconf.cfg
    • GalaxyControlFile.conf
    • init<instancename>.ora
    • SPFILE<instancename>.ora
  4. This can be done using one of the following methods:
    • Mount the Recovery Point, then copy only the files listed above to a temporary directory.
    • Perform a Copyback operation. Since this will not allow you to select which files to copy back, only use this method if you want to copy back everything (database, logs, and files listed above) - assuming the original folders are empty. Proceed as follows:
    • In the CommCell Browser, perform a Browse operation for one of the following:

      Replication Set

      DR Agent

      Client

      Application iDataAgent (Oracle)

    • Do one of the following, depending on which object you browsed:

      In the Recovery Points dialog box, select the Consistent Recovery Point(s) to be copied back, and click Copyback.

      In the Browse dialog box, select the Consistent Recovery Point(s) to be copied back, and click Recover all Selected...

    • In the Copyback dialog box, select a Recovery Host, the host to which the data will be copied back.
    • Specify a Recovery Path for the data you want to copy back. The Recovery Path must be larger than the data being copied back. To assign a Recovery Path, do one of the following:

      Double-click the path listed in the Recovery Path column and type a path on the selected host

      Click the Browse 'Recovery Host' button and browse to a destination path on the selected host

    • Select Overwrite existing data under 'Recovery Path'.
    • Click OK to copy back the data. The system displays a progress bar as it copies back the data. You can track the progress of the recover operation from the Job Controller dialog box.
  5. Create the database tracing directories (udump, bdump, cdump, etc.) for the Oracle database according to the dbconf.cfg file you copied back in Step 3. (<Snap2Dir>/archivelog_dest1/dbconf.cfg)
  6. If the source database uses spfile, copy spfile from the location you copied back to in Step 3 to the default location under ORACLE_HOME. For example:

    cp <Snap2Dir>/archivelog_dest1/spfile <ORACLE SID>.ora <default spfile file path>

  7. If the source database uses pfile, copy pfile from the location you copied back to in Step 3 to the default location under ORACLE_HOME. For example:

    cp <Snap2Dir>/archivelog_dest1/pfile <ORACLE SID>.ora <default pfile file path>

  8. Copy the backup control file from the first archive log destination (from the second snapshot) to all three control files destinations. For example:

    cp <Snap2Dir>/archivelog_dest1/backup.ctl.galaxy <controlfile1_Path from spfile>/ control01.ctl
    cp <
    Snap2Dir>/archivelog_dest1/backup.ctl.galaxy <Controlfile2_Path from spfile>/ control02.ctl
    cp <
    Snap2Dir>/archivelog_dest1/backup.ctl.galaxy <Controlfile3_Path from spfile>/ control03.ctl

    GalaxyControlFile.conf also contains the locations of the control files at the time the Consistent Recovery Point was created.

  9. If you did not perform a Copyback operation in Step 3, copy all data and log files to the exact same path as that of the original database, either by mounting the Recovery Point and copying the files, or by performing a Copyback operation. In case the same destination volume is used for database and logs, mount the first snapshot and copy all Oracle data files from the first snapshot to the path that the control file expects. For example:

    cp -a <Snap1Dir>/oradata/rman1/*.dbf /oradata/rman1/

  10. Copy all the log files for all destinations from the second snapshot to the paths that the control file expects, either by mounting the Recovery Point and copying the files, or by performing a Copyback operation. For example:

    cp -a <Snap2Dir>/archivelog_dest*/ /oralogs/rman1/

    If you have previously chosen to replicate only one (not all) Oracle instance log locations, after the logs have been copied back, copy them to the other log location (if any) on the production server. For instance:

    If you have D:\ArchiveLogs1, F:\ArchiveLogs2, H:\ArchiveLogs3 and you initially selected to replicate only D:\ArchiveLogs1, after you copyback to D:\ArchiveLogs1, copy the same files to F:\ArchiveLogs2 and H:\ArchiveLogs3. You can either copy and paste them from D:\ArchiveLogs1 or run copyback again, specifying F:\ArchiveLogs2 and H:\ArchiveLogs3 as destinations, whichever is faster in your particular environment.

  11. Set the ORACLE_SID and ORACLE_HOME environment variables.
  12. Run the following commands as an Oracle user:

    sqlplus "/ as sysdba"

    • startup mount
    • set autorecovery on;
    • recover database until cancel using backup controlfile;
    • alter database open resetlogs;
  13. Database open will succeed after recovering the database.