PowerProtect Data Manager 19.15 Microsoft SQL Server User Guide

Backups fail when credentials include a backslash character (\)

When you enter credentials that include a backslash character (\) for an application agent in the PowerProtect Data Manager UI, the backups fail.

For example, when you enter a password for the operating system or database user that includes the backslash character, subsequent backups fail with the following error message:

This error might occur when updating the password for a storage unit.

To resolve this issue, type \\ (double backslash) instead of \ (single backslash) when you enter the credentials for an application agent in the PowerProtect Data Manager user interface.

Application-aware backup failure after interactive session using service account

A Microsoft SQL Server application-aware full or log backup might fail after an administrator uses a service account to complete an interactive session on the Microsoft SQL Server virtual machine. The administrator uses the service account to log in to the Microsoft SQL Server virtual machine, completes an interactive session, and then logs out. If the administrator then runs an application-aware full or log backup, the backup might fail with the following message in the log file:

Illegal operation attempted on a registry key that has been marked for deletion.
COM: Intermittent error 800703fa: Illegal Operation attempted on a registry key
Other COM related errors.

As a workaround, complete the following steps:

1. Ensure that you do not log in to the Microsoft SQL Server virtual machine for an interactive session by using a service account.

2. Disable the related Windows User Profile service:

   1. Open the Group Policy Editor gpedit.msc on the affected server.
   2. Open the UserProfiles folder by selecting Computer Configuration > Administrative Templates > System > UserProfiles.
   3. Change the Do not forcefully unload the user registry at user logoff setting to Enabled.

Application-aware log backup failure for AAG database on secondary replica with Microsoft SQL Server 2016 or 2017

With Microsoft SQL Server 2016 or 2017, when you perform an application-aware log backup for an AAG database on a secondary replica, the log backup might fail with the following error message:

Log backup for database "<database_name>" on a secondary replica failed because a synchronization point could not be established on the primary database. Either locks could not be acquired on the primary database or the database is not operating as part of the availability replica. Check the database status in the SQL Server error log of the server instance that is hosting the current primary replica. If the primary database is participating in the availability group, retry the operation. (Error: 35294)

As a workaround:

1. Install the latest cumulative update on the Microsoft SQL Server 2016 or 2017 system:

   • For Microsoft SQL Server 2016, install the cumulative update version 13.0.2216.0.
   • For Microsoft SQL Server 2017, install the cumulative update version 14.0.3456.2.
2. Rerun the application-aware log backup for the AAG database on the secondary replica.

Application-aware backup failure in multiple VLAN environment

In a Microsoft SQL Server environment of multiple virtual networks (VLANs) where each application client has access to only one network interface, an application-aware backup fails with the following error message:

PowerProtect agent for Microsoft Applications: Unable to update persisted settings for PowerProtect Data Manager host '<IP_address>': Exception 'Timeout was reached' occurred while sending Certificate Signing Request.

A Microsoft SQL Server application-aware backup is not supported in this type of multiple virtual network environment.

In a multiple virtual network environment, ensure that the Microsoft SQL Server application-aware host has network connectivity to both the management network and data network. The "Virtual networks (VLANs)" section in the PowerProtect Data Manager Administrator Guide provides more details.

Application-aware backup failure with persisted settings message

An application-aware virtual machine backup performed through a protection policy might fail with the following error message, where xx.xx.xx.xx is the PowerProtect Data Manager IP address:

PowerProtect agent for Microsoft Applications: Unable to update persisted settings for PowerProtect Data Manager host 'xx.xx.xx.xx': Unable to log in to PowerProtect Data Manager host 'xx.xx.xx.xx'. Received the HTTP code 400 and response 'Invalid FQDN'.

As a workaround, when the Microsoft SQL Server host is a member of a workgroup but is assigned an IP address that is DNS resolvable by PowerProtect Data Manager, if the DNS resolution is only available for FQDN, ensure that the Microsoft SQL Server host adds the same domain suffix to the Advanced TCP/IP parameters DNS settings.

In the System Properties dialog box:

1. On the Computer Name tab, click Change and then click More.
2. In Primary DNS suffix of this computer field, specify the PowerProtect Data Manager resolvable FQDN of the Microsoft SQL Server virtual machine domain name.
3. Click OK.

Rename the lockbox folder in C:\Program Files\DPSAPPS\common on the Microsoft SQL Server client, and retry the virtual machine configuration job.

Application-aware backup failure with msagentcon message

An application-aware backup might fail in the PowerProtect Data Manager UI, with the following error message in the msagentcon log file:

msagentcon PowerProtect Data Manager info Unable to retrieve lockbox password :'Error encountered while retrieving information from the lockbox 'C:\Program Files\DPSAPPS\common\lockbox\agents.clb' for the device host '10.125.196.91': The Lockbox contains no entries. The drive that contains the lockbox might be full, causing possible corruption of the lockbox. Try the operation again after you reconfigure the lockbox. 

To resolve this issue, free up disk space on the C:\ drive and reconfigure the policy for the Microsoft SQL Server asset so that the lockbox is updated.

Application-aware backup failure with ABV0030 error

An application-aware virtual machine backup might fail with the following error message:

ABV0030: VM Direct engine … is unable to back up the virtual machine … on vCenter … because the vCenter snapshot was unsuccessful. 

To resolve this issue, confirm the following, and then retry the backup:

• The vCenter server is not busy
• There is enough free space on the datastore to perform the operation
• If this is a quiesced backup, the virtual machine is not busy

If this does not resolve the issue, perform a manual quiesced snapshot of the virtual machine and look in the datastore location of the virtual machine for a VSS_manifest_zip file. If there is no file, contact VMware support. If there is a file, contact Customer Support.

Memory or paging issues during Microsoft SQL Server backups

When the Microsoft application agent is integrated with the PowerProtect Data Manager, Microsoft SQL Server backups that are initiated by a protection policy might fail with the following types of error messages:

ddbmsqlsv: BackupVirtualDeviceSet::SetBufferParms: Request large buffers failure on backup device 'EMC#6f1e7adf-678f-4ceb-88fb-248211218add'. Operating system error 0x8007000e (Not enough storage is available to complete this operation.).

ddbmsqlsv SYSTEM critical Unable to allocate memory: Not enough space
ddbmsqlsv PowerProtect Data Manager critical BackupVirtualDeviceSet::SetBufferParms: Request large buffers failure on backup device 'EMC#77002120-0a31-4f53-98b1-3311f997e8ac'. Operating system error 0x800705af (The paging file is too small for this operation to complete.).

To prevent or resolve these backup issues, use any of the following procedures:

• Increase the RAM size on the Microsoft SQL Server host.

• Increase the virtual paging value (swap size) on the Microsoft SQL Server host by completing the following steps:

  1. Select Control Panel > System and Security > System.
  2. Select Advanced system settings to display the system properties of the Microsoft SQL Server host.
  3. In the System Properties dialog box, select the Advanced tab.
  4. Under Performance, select Settings.
  5. In the Performance Options dialog box, select the Advanced tab.
  6. Under Virtual memory, select Change.

  7. In the Virtual Memory dialog box:

     1. Clear the default setting, Automatically manage paging file size for all drives.
     2. Select the drive for which you want to change the paging file size.
     3. Select Custom size.
     4. Set the Initial size to the same value as the Currently allocated size.
     5. Set the Maximum size required for the paging file, based on the recommended size for the Microsoft SQL Server host system. To ensure system stability, Windows restricts the paging file size to three times the value of the installed RAM.

• On the Infrastructure > Assets page in the PowerProtect Data Manager UI, reduce the stripe value for all databases. Enable multi-stream backups for Microsoft SQL Server protection policy provides details about how to set the stripe value.

  NOTE:If the stripe value is decreased to 1, full and differential backups can take longer to complete, depending on the size of the database.
• If the backup schedule results in overlapping backups that run at the same time, such as concurrent FULL and LOG backups, modify the backup schedule to prevent the overlapping backups.

• If possible, create multiple protection policies with fewer databases that have non-overlapping schedules.

  NOTE:A Microsoft SQL Server database asset may be included in only one protection policy.

Missing duration value for centralized backup

A centralized Microsoft SQL Server backup succeeds, but the backup job duration at the asset level appears as 0 ms on the Step Log tab and Details tab in the PowerProtect Data Manager UI.

The Step Log tab and Details tab appear in the Job ID Summary window, which you can access by selecting Jobs > Protection Jobs and clicking the job ID next to the job name.

As a workaround, view the correct backup duration in the Duration column in the left pane that has a grid format.

TLOG backup issue with mirrored database configuration

When you use a mirrored database configuration with Microsoft SQL Server 2012 or later, a TLOG backup is promoted to a full backup and the following type of error message appears:

Unable to determine the status of mirror partner backups for database DBA_TEST. Promoting the backup level full database backup.

With Microsoft SQL Server 2012 or later, use of an Always On availability group (AAG) is recommended. A mirrored database configuration is not supported with Microsoft SQL Server 2012 or later.

Self-service T-SQL backups with an I/O error message

When the Microsoft application agent is integrated with the PowerProtect Data Manager, a successful self-service Transact-SQL (T-SQL) backup might display the following error message:

IO error: Permission denied

To prevent this error message during self-service T-SQL backups, ensure that the Microsoft SQL Server instance service runs as the OS/Microsoft SQL Server user, instead of NT SERVICE\<instance_name>.

Expired imported backup copies are not deleted

PowerProtect Data Manager does not delete any Microsoft SQL Server backup copies that are imported but not associated with a protection policy, even when the backup copies are expired.

When you try to delete such copies by using the ddbmexptool or msagentadmin tool, you receive the following error message:

Write and delete operations are not supported on the storage unit '<storage_unit_pathname>' because the storage unit is managed by the PowerProtect Data Manager system.

To delete the expired backup copies, you can use the following procedures from the PowerProtect Data Manager UI or Microsoft SQL Server host:

• From the PowerProtect Data Manager UI, select the expired backup copies and delete the backup copies by clicking the Delete button.

• From the Microsoft SQL Server host, perform the following steps:

  1. Open the <installation_path>\Settings\.app.settings file for editing. The file content is in XML format, for example:

     <?xml version="1.0" encoding="utf-8"?>
     <?xml version="1.0" encoding="utf-8"?>
     <settings>
     <credentials>
     <credential>
     <ddHostName>10.31.140.153</ddHostName>
     <devicePath>sql_rc-5b2bbffe443d-6451a</devicePath>
     <ddUser>sql_rc-5b2bbffe443d-6451a</ddUser>
     <type>PROTECTION</type>
     <protocol>DDBOOST</protocol>
     <encryptionOverWire>false</encryptionOverWire>
     <retentionLock>false</retentionLock>
     <retentionPeriod>86400</retentionPeriod>
     <configName>sql_rc</configName>
     <sources>
     <name>INST1_SQL2012::dell1</name>
     <name>INST1_SQL2012::dell2</name>
     <name>MSSQLSERVER::dell_1</name>
     <name>MSSQLSERVER::dell_2</name>
     </sources>
     </credential>
     </credentials>
     </settings>
     

  2. In the <installation_path>\Settings\.app.settings file, locate the section between the <credential> and </credential> tags that contains the <ddHostName>, <devicePath>, <ddUser>, and <type> tags for the DD or MTree where the backup copies are generated and imported to PowerProtect Data Manager.

  3. Delete that section in the <installation_path>\Settings\.app.settings file.

     CAUTION:Delete only the tags <ddHostName>, </ddHostName>, <devicePath>, </devicePath>, <ddUser>, </ddUser>, <type>, and </type> and the information between those tags. Do not delete any other tags or information in the file, which contain the DD details that the PowerProtect Data Manager generates.

Backup failure during FCI cluster failover

While the failover of a Microsoft SQL Server Failover Cluster Instances (FCI) cluster is in process, Microsoft SQL Server backups that are initiated by a protection policy fail.

As a workaround, wait until the failover process is complete and then perform the Microsoft SQL Server backup.

Microsoft SQL Server databases skipped during virtual machine transaction log backup

If a transaction log backup is not appropriate for a database, the database is automatically skipped. Databases are skipped for the reasons outlined in the following table.

All skipped databases are backed up as part of the next full backup. Also, a skipped database does not result in a failure.

The only instance in which a transaction log backup job would potentially fail is if all Microsoft SQL Server instance databases failed to be backed up or were skipped.

Microsoft SQL Server application-aware backup error about disk.EnableUUID variable

A Microsoft SQL Server application-aware virtual machine backup succeeds but displays the following error when the disk.EnableUUID variable for the virtual machine is set to TRUE:

To resolve this issue, set the disk.EnableUUID variable to TRUE and then reboot the virtual machine.