Troubleshoot data recovery issues


Updated: May 13, 2016

Applies To: System Center 2012 SP1 - Data Protection Manager, System Center 2012 - Data Protection Manager, System Center 2012 R2 Data Protection Manager

This topic provides troubleshooting guidance for general data recovery issues.

General issues

Issue Possible Cause Details
The Recovery Wizard Summary page does not display the details of the recoverable items. If you selected data to restore at the item level, the Recovery Wizard Summary page will not display the details of the recoverable items. Item details will not appear on the Recovery Wizard Summary page for the following:

- Windows SharePoint Services sites and items
- Exchange Server mailboxes

The Summary page will display recovery details such as recovery points, media, and notifications.
The recovery job fails. If you perform a recovery at the same time a disk-to-tape (D2T) short-term protection job is scheduled to run, the recovery will fail. DPM will cancel a D2T short-term protection job if it is running at the same time you are performing a recovery.

You can complete the steps in the Recovery Wizard; however, the recovery job will fail. You must either cancel the D2T short-term protection job or wait for it to complete before performing the recovery.
Starting recovery for a parent data source which is referential from external tape does not show any recovery job in the Administrator Console. This happens when the parent data source does not have any filespec associated with it. Select the referenced data source explicitly for recovery.

Tape library recovery

Issue Possible Cause Resolution
Tape read errors occur when attempting to recover a file server volume from tape. File server data source recovery from tape will fail if there are corrupt files on the tape. If tape read errors occur, do the following:

1. Select the contents of the volume for recovery rather than the entire volume. For example, select all of the files and folders on <drive letter>:\ instead of selecting the entire <drive letter>:\ <volume>.
2. Retry the recovery.
     DPM will attempt to skip over the corrupt files on tape. Caution: Because of serious tape corruption, DPM may not be able to find the files that are not corrupt. Therefore, DPM may not succeed in recovering the selected individual files.

SQL Server recovery

Issue Possible Cause Resolution
Unable to recover from secondary server because replica is in inconsistent state. If the database you are trying to recover was collocated, and protection for one of them was stopped on the primary server and its data deleted, then you cannot recover the deleted database from the replica on the secondary. Run a consistency check on the replica volume.
The database recovery fails. If you recover a SQL Server database by selecting the latest recovery point and then perform another recovery using the latest recovery point, the database recovery will fail. When you recover a SQL Server database using the latest recovery point, DPM performs the recovery using the SQL Server logs that are available on the protected server. After you recover the database, you cannot perform another recovery using the latest recovery point until you perform a synchronization with consistency check.

However, you can recover to a recovery point listed just prior to the latest recovery point. In this case, you will recover the database to the latest available recovery point on the DPM server; however, the logs on the SQL Server will not be utilized. This is because the logs on the SQL Server were already used as part of the latest point recovery that you performed earlier.

Exchange recovery

Issue Possible Cause Resolution
Recovery fails with internal code 0x8099A52 The name of the Exchange Server database starts with a space. Rename the database and reprotect it.
Recovery fails when trying to mount or dismount the Exchange Server database. If you rename an Exchange Server storage group after protecting it, when DPM performs a recovery operation, the recovery will fail when trying to mount or dismount the database. The caption of the Exchange Server storage group is used to mount and dismount the Exchange Server databases. If you rename the storage group, the caption stored in the protected object will be different from the actual caption and the mount and dismount operations will fail. If this occurs, we recommend that you do the following:

- Rename the storage group back to the original storage group name.
- When you detect a name change, update the caption name in the protected object.
- You can use the component name to identify a storage group to mount and dismount the databases under it instead of the caption name.
ID 104: An unexpected error occurred during job execution. Exchange Server 2007 requires credentials of an Exchange Server administrator for dismounting or mounting the database.

The DPM replication agent requires Local System account privileges, and therefore, recovery fails when trying to recover data on an Exchange Server 2007 server that is configured for clustered continuous replication (CCR).
Perform the following steps to provide the user with the Exchange Server Administrator privileges:

1. From the command prompt, type dcomcnfg, and then click OK.
2. On the Component Services screen, expand Component Services, expand Computers, expand My Computer, and then click DCOM Config.
3. Right-click ExchangeCmdletsWrapper, and then click Properties.
4. In the Properties dialog box, on the Identity tab, select This user.
5. Provide the credentials for the account that has Exchange Server Administrator privileges, and then click OK.
Recovery to the latest Programmable Interval Timer (PIT) fails. If the embedded database (.edb) and the streaming (.stm) files are renamed and then the databases are not dismounted and then remounted, the recovery fails. Dismount and remount the databases, and then retry the recovery. If you cannot dismount and remount the databases, restore the files and then copy them with a clean shutdown. Then move the files to the original location and mount the database.
Mailbox recovery fails with a catastrophic error when recovering to an alternate Exchange Server 2003 database. The restore flag was not set when attempting to recover a mailbox to an alternate Exchange Server 2003 database. After the recovery fails, go to the Exchange Server Management Console, set the recovery flag, and then mount the database.
The Recover-RecoverableItem cmdlet fails specifying that the specified recovery options were not valid. If you specify the active server name when recovering a storage group or database for Exchange Server Clustered Continuous Replication (CCR) through the command-line interface (CLI), the recovery fails. The –TargetServer attribute should be ResourceGroupName.ClusterServerName.DomainName.
eseutil execution fails with error code 0xfffff764 in the Application log. This error occurs if the .stm and .edb files that are associated with a database are not in the same folder. When this occurs, the database is not brought to a Clean Shutdown state. To bring the database to a Clean Shutdown state, move the .stm file to the same folder as the .edb file, and then run eseutil with the appropriate parameters.

SharePoint recovery

Issue Possible Cause Resolution
Microsoft Office SharePoint Server 2007 – Unable to see recovered list item. The item may still be in the site or site administrator’s Recycle Bin. We recommend that you recover the item from the site or site administrator’s Recycle Bin, if it is the same version of the file that you are looking for, rather than recover it from the replica. Alternatively, clear both the Recycle Bins in SharePoint and then retry the recovery.
Microsoft Office SharePoint Server 2010 – Recovering list item does not recover its permissions. The item already exists in the list, or is in either the user’s Recycle Bin or the site administrator’s Recycle Bin. If you are recovering an item that already exists in either the user’s or the site administrator’s Recycle Bin, DPM will restore a duplicate item and apply the parent folder’s permissions. However, if the item has been deleted from all Recycle Bins, then it will be restored along with the permissions.
Recovery points are not displayed for sites or documents in DPM Administrator Console in the Recovery task area. (If you double-click the content databases, the sites do not appear.) DPM backs up the databases and then performs a catalog task to get a list of all the URLs in the content databases. By default, this process runs three hours after the first scheduled backup job of the day. If the catalog task has not completed, you will not be able to expand the content database in the DPM Recovery UI. Run the catalog task.
After recovering a Windows SharePoint Services farm, you cannot attach a content database. If you recover a Windows SharePoint Services farm that had a detached database prior to recovery, you cannot attach the database after the recovery if the front-end Web server was different than the database DPM used for the recovery. To resolve this issue, do the following:

- Delete the config cache on the front-end Web server that was not used for the recovery.
The cache is a set of XML files located in the following folder: <drive letter>:\Documents and Settings\All Users\Application Data\Microsoft\SharePoint\Config\<config database ID>. Caution: Do not delete the folder. Delete only the XML files in the folder.
- The cache will automatically rebuild, enabling you to attach the database.
Recovery of the Windows SharePoint Services farm fails. If a Quick Fix Engineering (QFE) is installed after the Windows SharePoint Services farm is protected, the recovery will fail. To resolve this issue, do the following:

1. Recover the Windows SharePoint Services farm by using the Recovery Wizard.
2. On the front-end Web server, from the command prompt, type psconfig -cmd upgrade -wait -force -inplace b2b.
Consistency check fails with "Data source not found". If you perform a recovery after the content database is deleted from SQL Server Management Studio, and not through the SharePoint Console, the consistency check will fail. You must stop protection, and then use the New Protection Group Wizard to reprotect the Windows SharePoint Services farm.
Permissions or ACLs for the document or list item have not been restored. The document or list item is being restored to an alternative URL within the same farm. DPM recovers the permissions or ACLs only if the item is being restored to the original location.
Recovering a SharePoint content database to its original location fails. The recovery point is located on the tape and the database has been removed from the SharePoint farm. Recover the database by using the “Recover to any SQL Instance” option.

End user recovery

Issue Possible Cause Resolution
End-user recovery (EUR) configuration fails on Windows Server 2008. Try either one of the following solutions:

- Copy DPMADSchemaExtension.exe from DPM install folder to the active directory and run it from there.
- In Server Manager, click Add roles. Select Active Directory Domain Services role and install it. Launch DPM again and configure EUR.
If you enable end-user recovery functionality on your DPM server and then try to recover an earlier version of a file or folder from your computer, your computer automatically tries to recover the file or folder from the DPMserver. This issue occurs if the computer is secured by DPM or if there are no local shadow copies of the file or folder on the computer. To resolve this issue, you must download Microsoft Knowledge Base article 903234 to update the Timewarp client on the client computer. To download article 903234, see An update is available to the way that the Shadow Copy Client accesses shadow copies in Windows Server 2003 and in Windows XP.

Recovery from workgroups and untrusted domains

Issue Possible Cause Resolution
Protection agent refresh fails after system state recovery of protected computer. Password set for protection was updated after the system state of the protected machine was backed up. Reset the password.

For more information, see Updating Password for Workgroup or Untrusted Computers.

System recovery

Issue Possible Cause Resolution
BMR and System State backups inaccessible If your computer’s localization settings are set to Czech, then the period (.) that DPM adds at the end of directory name where the backups are stored maybe the issue. Rename the directory without the period (.) at the end.