Known Issues and Limitations: Version 3.2.x FlashCopy® Manager for Microsoft Exchange & Version 6.4.x Data Protection for Microsoft Exchange

Preventive Service Planning


Abstract

This document details the Known Issues and Limitations for FlashCopy Manager® for Microsoft Exchange Server V3.2.x and Data Protection for Microsoft Exchange Server V6.4.x

Content

The following components share the same known issues and limitations. The components are:

  • FlashCopy Manager for Microsoft Exchange V3.2 --> which is a part of the IBM Tivoli Storage FlashCopy Manager V3.2 product
  • Data Protection for Microsoft Exchange V6.4 --> which is a part of the IBM Tivoli Storage Manager for Mail V6.4 product.

This document describes the Known Issues and Limitations applicable to:
  • FlashCopy Manager for Microsoft Exchange V3.2.x
  • Data Protection for Microsoft Exchange 6.4.x



This document is divided into linked sections for ease of navigation. You may use the links below to jump to the desired section of the document.


General Issues and Limitations
Exchange 2013 Specific Issues and Limitations
GUI Issues and Limitations
Mailbox Restore Issues and Limitations
VSS Issues and Limitations
National Language Support (NLS), and Other Issues When Operating in a Non-English Environment
Exchange Server 2007 Continuous Cluster Replication Information
Exchange Server 2010 Database Availability Group (DAG) Backups



General Issues and Limitations



  • Multiple Problems Found When Using Microsoft Exchange 2010 SP3 and MAPI/CDO 6.5.8309.0

    After applying:
      • Microsoft Exchange 2010 Service Pack 3, or
      • Microsoft Exchange Messaging API (MAPI) Client and Collaboration Data Objects (MAPI/CDO) download version 6.5.8309.0
    Data Protection for Microsoft Exchange Server and FlashCopy Manager for Microsoft Exchange Server will not work properly.

    Additional details for both limitations can be found in APAR IC90654 here: http://www.ibm.com/support/docview.wss?uid=swg1IC90654

    Workaround: Update to either Data Protection for Microsoft Exchange Server 6.4.1 or FlashCopy Manager for Microsoft Exchange Server 3.2.1.

  • Data Protection for Microsoft Exchange and FlashCopy Manager for Microsoft Exchange fail in Windows Management 3.0 environments

    After installing Windows Management Framework 3.0, backup, restore, and query operations fail with an error like:
      The WinRM client cannot process the request. The WinRM client tried to use Negotiate authentication mechanism, but the destination computer (EXCHDB01) returned an 'access denied' error. Change the configuration to allow Negotiate authentication mechanism to be used or specify one of the authentication mechanisms supported by the server. To use Kerberos, specify the local computer name as the remote destination. Also verify that the client computer and the destination computer are joined to a domain. To use Basic, specify the local computer name as the remote destination, specify Basic authentication and provide user name and password. Possible authentication mechanisms reported by server:

      For more information, see the about_Remote_Troubleshooting Help topic.
    According to this Microsoft blog article:

    http://blogs.technet.com/b/exchange/archive/2012/12/14/windows-management-framework-3-0-on-exchange-2007-and-exchange-2010.aspx

    Windows Management Framework 3.0 is not supported in Exchange 2007 or 2010 environments. Since the release of that article, Microsoft added support for the coexistence of Windows Management Framework 3.0 with Microsoft Exchange Server 2010 Service Pack 3.

    Though not directly used by Microsoft Exchange Server 2010, coexistence with Windows Management Framework 3.0 is supported by Microsoft Exchange Server 2010 Service Pack 3. However, Windows Management Framework 3.0 is not supported by Data Protection for Microsoft Exchange Server or FlashCopy Manager for Microsoft Exchange Server at this time.

    Workaround: Update to either Data Protection for Microsoft Exchange Server 6.4.1 or FlashCopy Manager for Microsoft Exchange Server 3.2.1.

    (IC89531) (IC90654)

  • When using Data Protection for Microsoft Exchange Server, or FlashCopy Manager for Microsoft Exchange Server, the database name or the storage group (sg) name can not end with 14 digits surrounded by parentheses characters.

    For example:

    Database or
    Storage Group Name
    Supported? Reason
    DB(20120409153304) not supported Database name ends with 14 digits surrounded by parenthesis
    DB20120409153304 supported Database name ends with 14 digits but they are not surrounded by parenthesis
    DB(2012040915330) supported Database name is surrounded by parenthesis but only has 13 digits within them
    DB(abcdefghijklm1) supported Database name ends with 13 characters and only 1 digit surrounded by parenthesis

    (IC79445)

  • Remotely Mounted backups are not removed from the storage device after manual deletion or expiration

    Data Protection for Microsoft Exchange Server and FlashCopy Manager for Microsoft Exchange backups can be mounted using the Remote Mount feature. When one of these remotely mounted backups is manually deleted, or expires based on policy retention definitions, it will be deleted from the Tivoli Storage FlashCopy Manager inventory. However, the actual snapshot located on the storage device will remain.

    Workaround: Ensure that remotely mounted backup images are unmounted prior to manual deletion or expiration. (33605) (33605-321)

  • The LogEstimate value from tdpexc.cfg file is ignored. The default value of 0 will be used

    Workaround: Specify the /LOGESTimate value on the tdpsqlc command line.
    (30654) (30654-410)

  • If a stand-alone Data Protection for Microsoft Exchange Server package is installed on a system where Tivoli Storage FlashCopy Manager is already installed, and the Data Protection for Microsoft Exchange Server component is also already installed at the same level as the stand-alone package, the Data Protection license enablement file will not be installed.


    Workaround: Manually copy the file excclient.lic from the installation media to the TDPexchange folder. By default, this folder is located at:
      c:\program files\tivoli\tsm\TDPexchange

    (77952) (85882) (IC86907) (IC82568)

  • Under rare circumstances, a backup in a FlashCopy Manager stand-alone environment (e.g. no TSM integration) can fail in such a way that future stand alone backups are not possible without implementation of the recommended circumvention.

    The problem can only occur during the initial stand-alone server backup when the DSMAGENT is abnormally terminated while the backup is running. When this occurs, a saved copy of the local FlashCopy Manager database is not available for recovery because the previous backup never completed. This causes subsequent backups to fail because the database files were left in a corrupt state by the previous backup.

    This issue can only occur during the initial backup.

    Workaround: Stop the DSMAGENT service, delete all local FlashCopy Manager database files in the FlashCopy Manager staging directory, and then restart the DSMAGENT agent service. (85829)

  • When a backup process will include multiple components, such as DB1, DB2, and DB3, the behavior for mounting local snapshots differs when the snapshots are mounted remotely.

    Example backup scenario with multiple components via the command line:

      tdpexcc backup db1,db2,db3 ....


    or via the MMC GUI interface:
      Select multiple components (rows)
      Press the "Backup" button

    When the backup is made as described in the examples above, and a local mount is done, the only component(s) that you will see at the mount point are the ones that were specified on the mount command.

    In contrast, when a remote mount is done you will see all the components, specified as a part of the backup, at the mount point.

    For example:

    If an Exchange backup was created like:
      tdpexcc backup db1,db2,db3 ...

    and then after the backup completed, a remote mount command was issued on a server other than the one where the backup was created:
      tdpexcc mount backup db2=c:\mnt /remotecomputer=server2 /remotecomputeruser=user /remotecomputerpassword=passwd

    you would see directories for db1, db2 and db3 at the mount point c:\mnt rather than just db2.
    (85502)

  • When running the stand-alone or TSM configuration wizard, the wizard may fail during the Configuration step. Specifically, the Configuring Services task may show a "Failed" status.

    Workaround: Use the Re-Run button to perform the operation again. (88813)

  • When using the /PITDATE and /PITTIME options to remote mount a backup via the command line interface, these two options will be ignored and the active backup will be mounted instead.

    Workaround: Use the <object-id> positional parameter to specify the specific inactive backup to mount. (87250)

  • The "object-id" parameter is missing from the fcmcli help for the mount command.

    Workaround: The fcmcli mount command syntax, including the "object-id" parameter, is documented in the Installation and Users Guide. (87252)

  • The Exchange restore options allow both of the following settings to be true at the same time:

    "Replay Restored AND Current Logs"
    "Replay Restored Logs ONLY"

    when in fact these are conflicting options. Therefore, an error message will be displayed.

    Workaround: Do not set "Replay Restored AND Current Logs" and "Replay Restored Logs ONLY" at the same time. (86935)

  • When performing legacy backup operations, the BUFFERSIze option value must be an exact multiple of 8 or the following message is seen:

    "ACN5798E MS Exchange API HRESEBACKUPREADFILE() failed with HRESULT: 0xc7ff1004"

    Workaround: Specify a BUFFERSIze that is an exact multiple of 8. (47167).

  • Through the use of repeated restore attempts, it is possible for a storage group to become unrecoverable with the current transaction log even though the Exchange recovery process is told not to replay the current transaction log. For example, when using these options:

    - Command line option: /RECOVER=APPLYRESToredlogs
    - GUI check box: "Replay Restored Logs ONLY"
    - ESEUTIL option: /t

    Workaround: Erase all current transaction log files for the storage group being restored and retry the operation. This is a Microsoft Exchange Server issue.

  • The password option in the Tivoli Storage Manager client options file (dsm.opt, by default) is ignored. This is working as designed. You must use PASSWORDACCESS GENERATE if you do not want to be prompted for a password.

    Workaround: Set PASSWORDACCESS to a value of GENERATE.


Exchange 2013 Specific Issues and Limitations

  • The VSS Instant Restore process may fail with the following message:
      ANS1050E Before performing the VSS Instant Restore TSM cannot lock the volume 'D:'

    in environments using Microsoft Exchange 2013 SP1 and when the files to restore will be restored to a drive where either of the following Microsoft Exchange services:
    • Microsoft Exchange Diagnostic
    • Microsoft Exchange Health Manager

    write to the directory $volume:\MonitoringDiagnosticsLogs. This prevents the VSS Instant Restore function from locking the volume for the restore operation.

    Workaround: Until a change or fix from Microsoft is available, temporarily disable the 2 services specified above when it is necessary to perform a VSS Instant Restore operation to a volume where the MonitoringDiagnosticsLogs directory exists. After completion of the VSS Instant restore process, restart the 2 services.

  • Exchange 2013 CU1 is supported. CU2 is recommended.

  • Exchange 2013 CU1 reports that Exchange is ready for a restore operation even when it is not yet completely ready. As a result, an Exchange 2013 restore operations may fail. The problem does not occur in Exchange 2013 CU2 environments. Therefore, to work around this issue, a 180 second delay was implemented in 6.4.1. With the 6.4.1.1 interim fix level, the delay will be disabled for Exchange 2013 CU2 environments.

  • In mixed Exchange 2010 / Exchange 2013 environments, the administrator account that you use to perform the mailbox restore must reside on an Exchange server at the same level as the mailbox that you are trying to restore. For example, if you wish to restore an Exchange 2013 mailbox, you must log on as an administrative user whose mailbox is located on an Exchange 2013 database.


  • The Mailbox Restore filter BODY/ALLCONTENT doesn't work with Exchange 2013. This is a Microsoft issue and ticket has been opened with Microsoft.

  • Mailbox Restore on Exchange 2013 pops up an "Enter Password" dialog asking for Domain, User Name, and Password. This problem might be caused by changing Client Access Server settings (e.g. local server is a CAS server before, but the CAS role has been removed now). In such case, open regedit and delete RpcHttpProxyMap_TSM key in HKEY_CURRENT_USER->Software->Microsoft->Windows NT->CurrentVersion->Windows Message Subsystem

  • When using mailbox restore or the mailbox restore browser with Exchange Server 2013, the "Message Body Text" and "All Content" filters do not work as expected.

    Problem Verification: Using mailbox restore browser, add a filter to search for some text that you know exists in a message body. Depending on the type of message, the mailbox restore browser may not return any results.

    Additional Information: The MAPI search capability is handled by the Microsoft Exchange FAST/Content Indexing. As such, the mail content cannot be searched unless it has been indexed. A problem with the search capability was fixed in Microsoft Exchange 2013 CU2. However, there is an additional Microsoft limitation that prevents mail content searching immediately after an RDB is created but before Exchange FAST/Content Indexing processes have completed.

    Workaround: If possible, wait several hours for the Content Indexing process to complete before attempting to perform a mailbox restore using filters. Otherwise, use another filter, or restore the entire mailbox to a PST file and then use Microsoft Exchange to search the PST file.
    (36410)

  • When performing an Instant Restore of an Exchange 2013 database, the Instant Restore operation fails with error code 402. The following information appears in the logs:

    tdpexc.log
    ------------------
    03/26/2013 08:58:17 <FCMtest1 : VSS : full : 03/22/2013 17:20:46>
    03/26/2013 08:58:22 Restoring 'FCMtest1' via volume-level copy from snapshot(s). This process may take
    some time. Please wait.
    03/26/2013 08:58:53 FMV1050E Before performing the VSS Instant Restore FCM cannot lock the volume 'J:'.
    03/26/2013 08:58:55 FMX0519E The VSS operation failed with rc = 402.
    03/26/2013 08:58:57 FMX0151E Errors occurred while processing the request.


    dsmerror.log
    -------------------
    03/28/2013 23:16:46 ANS5250E An unexpected error was encountered.
    TSM function name : psFileLock
    TSM function : psFileLock(): Unable to lock volume: '2560', try = '10', error = '5'
    TSM return code : 164
    TSM file : ..\..\common\winnt\psfile.cpp (3107)
    03/28/2013 23:16:46 ANS1050E Before performing the VSS Instant Restore TSM cannot lock the volume '\\centori\h$'.
    03/28/2013 23:16:46 ANS5250E An unexpected error was encountered.
    TSM function name : UncfgVssVolumesBeforeSnapRestore
    TSM function : psPrepareVolForSnapRestore() failed for volume '\\centori\h$'
    TSM return code : 164
    TSM file : ..\..\common\ba\restsnap.cpp (2978)
    03/28/2013 23:16:46 ANS1050E Before performing the VSS Instant Restore TSM cannot lock the volume 'H:'.

    The problem occurs because the Microsoft Exchange Search Host Controller service has a open handle to the volume. If the database is in a DAG configuration, the Microsoft Exchange Replication service may also have a handle open to the volume.

    Workaround: Before attempting an Instant Restore on Exchange 2013, stop both the Microsoft Exchange Search Host Controller service and the Microsoft Exchange Replication service.
    (28263)


GUI Issues and Limitations

  • After changing the name of a database within a storage group, a command-line legacy restore fails but a GUI legacy restore works. This is because the command-line client checks each and every database against the current Exchange server whereas the GUI does not. The error message on the command-line client is:

    ACN5067E None of the storage groups or databases entered exist. Or, if they do exist, and you are using Exchange 2000 or Exchange 2003, the databases have not been dismounted.

    Workaround: Dismount all databases manually and use the GUI to perform the legacy restore operation when this occurs. This is a known problem currently under investigation (24658).

  • Extraneous "ActiveInactive" phrase appear in output of "QUERY TSM * /detail" command. Also, Backup Object State field is empty.

    Workaround: None. (77878)


Mailbox Restore Issues and Limitations

  • Multiple Problems Found When Using Microsoft Exchange 2010 SP3 and MAPI/CDO 6.5.8309.0

    After applying:
    • Microsoft Exchange 2010 Service Pack 3, or
    • Microsoft Exchange Messaging API (MAPI) Client and Collaboration Data Objects (MAPI/CDO) download version 6.5.8309.0
    Data Protection for Microsoft Exchange Server and FlashCopy Manager for Microsoft Exchange Server will not work properly.

    Additional details for both limitations can be found in APAR IC90654 here: http://www.ibm.com/support/docview.wss?uid=swg1IC90654

    Workaround: Update to either Data Protection for Microsoft Exchange Server 6.4.1 or FlashCopy Manager for Microsoft Exchange Server 3.2.1.

  • When performing a restore with the Mailbox Restore Browser, the "Total Restore Time" displayed at the completion of the operation is inaccurate. This problem can occur when several restore operations are run in succession.

    Workaround: None
    (30530)

  • Performing a restore of a large number of mailboxes (3000+) to an alternative mailbox may fail. This can happen when:

    1. The Destination mailbox exceeds the maximum defined size.
      Note: The Default size is 2GB.
    2. The MAPI folder connection exceeds the defined maximum.
    In some environments, rather than a restore failure, the MMC GUI may hang after the restore. (85406)

  • When performing a mailbox restore operation, the mailbox restore browser cannot restore deleted mailboxes that have a backslash "\" character in the name.


    Workaround: Perform the mailbox restore operation via the command line. (87146)

  • During batch mailbox restore operations, the MMC GUI communicates with the tdpmapi.exe process to perform the actual restore. When restoring a large number of mailboxes at a time, the tdpmapi.exe backend process might crash. In low memory environments, this crash might occur with as few as six mailboxes. In normal memory environments, it might take up to 100 mailboxes at a time to reproduce the problem.


    Workaround: Restore fewer mailboxes at a time. Monitor the mailbox restore operation and determine which mailboxes still need to be restored, and restore them separately. (77141)

  • Unable to restore mailboxes that contain parentheses in the mailbox alias.

    Workaround: Use the Mailbox Restore Browser GUI or Mailbox Restore command-line interface to restore these mailboxes instead. (77440)

  • MAILBOXFILTER by time/date only works with default formats

    The command line for individual mailbox restore supports multiple date and time formats.

    Date Formats:
    1 MM/DD/YYYY
    2 DD-MM-YYYY
    3 YYYY-MM-DD
    4 DD.MM.YYYY
    5 YYYY.MM.DD

    Time Formats
    1 HH:MM:SS
    2 HH,MM,SS
    3 HH.MM.SS
    4 HH:MM:SSA/P

    During an mailbox restore operation if a non-default date and/or time format is set (Non-default is any date and time format other than 1), and the /MAILBOXFILTER=STARTDATETIME option is used to specify a start and end date to restore the individual mailbox. The stat and end date will be ignored and all mail for the individual mail box will be restored. (65191)

  • In order to perform Mailbox Restore (IMR) operations on Exchange Server 2010, the Exchange Server that you are performing the mailbox restore on can not have a recovery database defined. IBM is working with Microsoft to resolve this issue This is a known Microsoft problem with reference number 110051477034945. Until that problem is resolved, you will see the following error message when performing a mailbox restore when a recovery database already exists:

    ACN6032E Cannot restore mailbox because there is an existing recovery storage group or database <database name> on the Exchange Server.

    Workaround: Remove the existing recovery database and perform the mailbox restore operation again. (60522)

  • Mailbox Restore limitation between Exchange 2003 and Exchange 2010

    After you move mailboxes from Exchange 2003 to Exchange 2010, you will no longer be able to perform mailbox restore operations using the backups taken from Exchange 2003. This restriction exists because Exchange 2010 changes the mailbox information that is stored in Active Directory.

    Workaround: Restore the appropriate Exchange 2003 databases into the Recovery Storage Group. Then, use ExMerge to copy the messages to the desired location.

  • Mailbox not accessible via MAPI (MAPI_E_FAILONEPROVIDER)

    There are multiple reasons for this error:
    • Mailbox does not exist
    • Mailbox has never been sent to or logged into
    • Mailbox database is not mounted
    • Insufficient privilege to access mailbox (see Installation & User's Guide publication)
    • Exchange server where mailbox database is located is not available or down
    • Microsoft Information Store service is not running
    • Microsoft Exchange System Attendant service is not running

  • Creating Profile (MAPI_E_NETWORK_ERROR) reported on Windows 2008

    This problem occurs during the following circumstances:

    1. Existing Exchange 2003 environment hosted completely on a 32-bit Windows Server 2003.
    2. An Exchange 2007 server hosted on Windows Server 2008 is added into this Exchange organization.
    3. Installed the Microsoft Exchange Server MAPI Client and Collaboration Data Objects 1.2.1 download to provide client MAPI support on the Exchange 2007 server.
    4. Installed Data Protection for Exchange 6.1.0 on the Exchange 2007 Server and it always fails when creating the MAPI profile for the mailbox on the local Exchange 2007 Server. It fails when calling IMsgServiceAdmin::ConfigureMsgService and returns MAPI_E_NETWORK_ERROR.

    Relevant Microsoft MAPI BLOG:http://blogs.msdn.com/stephen_griffin/archive/2008/08/15/mapi-and-ipv6.aspx

    The workaround is to disable IPv6 on the Exchange Server:

    In the Network Connections folder, obtain properties for all of your connections and adapters and clear the check box next to the Internet Protocol version 6 (TCP/IPv6) component in the list under "This connection uses the following items". This method disables IPv6 on your LAN interfaces and connections, but does not disable IPv6 on tunnel interfaces or the IPv6 loopback interface. Add the following registry value (DWORD type) set to 0xFFFFFFFF:


    HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip6\Parameters\DisabledComponents

    This method disables IPv6 on all your LAN interfaces, connections, and tunnel interfaces but does not disable the IPv6 loopback interface. You must restart the computer for this registry value to take effect. Opened Microsoft Support Case SRX0809196000563 on (09/19/08)

  • Logging on to MAPI (MAPI_E_LOGON_FAILED)

    The two most common reasons for this error are:
    1. User running mailbox restore has insufficient privilege to access mailboxes through MAPI.
    2. User running mailbox restore does not have a live accessible Exchange mailbox in the domain.

    The user account requirements for running Mailbox Restores from FlashCopy Manager or Data Protection for Exchange require that the user has sufficient privilege as described below:

    Exchange 2010:
    - user account with membership in the Organization Management group.

    Exchange Server 2007:
    - user account with membership in the Exchange Organization Administrators group.

  • The MAPI Spooler (mapisp32.exe) could not be started

    Missing or Damaged File (mapisp32.exe)

    A missing Mapisp32.exe file can cause this problem.
    The Mapisp32.exe file might have been removed or renamed by your anti-virus program.

    Exchange Server 2007 and 2010:
    Uninstall / Reinstall MAPI CDO 1.2.1 package.

  • MAPI Problems after Outlook is installed on Exchange Server

    Exchange and Outlook MAPI implementations are incompatible and Outlook should never be installed on an Exchange Server running IMR. In the event that this occurs, the following procedure might reconcile the issue:

    1. Uninstall Outlook from Exchange Server
    2. Reinstall MAPI CDO 1.2.1 package.
    3. Remove any existing TSMPROFILE profiles (this will not be necessary for later drivers): MFCMAPI Instructions:
      i. Bring up MFCMAPI, select the "Profile" menu.
      ii. Select "Show Profiles"
      iii. Delete all profiles that begin with "TSMPROFILE"
      iv. Select profile to delete and either:
        a. right click on selected profile and press "Delete Profile"
        b. from "Actions" menu select "Delete Profile"
      v. Exit MFCMAPI

  • MAPI Library Initialization Failure (MAPI_E_CALL_FAILED)

    Initialization of MAPI libraries fails as onset, usually indicative of a problem with the MAPI CDO 1.2.1 installation. Example trace output from:

    tdpmapi TESTMAPI /tracefile=<trace filename>

    ... ( 332): MAPIHandler::initialize(), current directory is <C:\Program Files\Tivoli\TSM\TDPExchange>
    ... ( 344): MAPIInitialize error: MAPI_E_CALL_FAILED
    ... ( 350): MAPIHandler::initialize(), reset current directory to <C:\Program Files\Tivoli\TSM\TDPExchange>
    ... ( 363): Exit MAPIHandler::initialize(), hr = MAPI_E_CALL_FAILED

    If installed, uninstall and reinstall the latest download from this Microsoft Web site:
    http://www.microsoft.com/downloads/details.aspx?FamilyID=E17E7F31-079A-43A9-BFF2-0A110307611E&amp;displaylang=en

  • Creating Profile ConfigureMsgService Error: 0x81002746

    Undocumented MAPI error that is related to the restoremailbox user having access to their mailbox.
    This problem is most likely because the user does not have a live Exchange mailbox.

    Troubleshooting steps:
    1. Check the user has sufficient privilege in the domain.
    2. Check that the user has a live Exchange mailbox and has access to it.
    3. If these do not solve the problem, try explicitly granting full access for the user to their mailbox.

    Microsoft link on granting full access to a mailbox for a user:
    http://technet.microsoft.com/en-us/library/aa996343.aspx

    Reference from the internet on this problem:

    Only a mail-enabled user account can create MAPI profiles. The "Local System account" is not mail-enabled and therefore cannot create MAPI profiles. In addition, a MAPI profile is associated to the user that creates it. So even if there was a profile with the required name, it cannot be found if the services are run by the "Local System account".

  • MAPI_E_NO_ACCESS: The user does not have permission's.

    Try again running mailbox restore again with from an user account that meets the following:

    Exchange 2010:
    - user account with membership in the Organization Management group.

    Exchange Server 2007:
    - user account with membership in the Exchange Organization Administrators group.


VSS Issues and Limitations

  • The following message:
      FMV9350E (RC6055) Backup operation failed due to volume mixture. Please refer to explanation section for additional information.

    should be:
      FMV9350E (RC6055) Backup operation failed due to volume mixture. Please refer to the explanation in the Tivoli Storage FlashCopy Manager Messages Guide for additional information.
    (87040)

  • Metro/ Global Mirror configurations are not supported with Tivoli Storage FlashCopy Manager components running on the Windows platform.

  • On Windows Server 2008 R2 x64 the DSMAGENT code crashes if the TSM Client Acceptor service stops while performing a VSS INSTANT RESTORE operation with the following error:

    FMV1017E (RC-50) Session rejected: TCP/IP connection failure

    For example, dsmcrash.log contains the following:

    2011/07/21 17:44:27
    IBM Tivoli Storage Manager
    Version: 6.3.0.42 110707C
    Build date: Thu Jul 07 23:18:06 2011
    dsmagent.exe caused exception C0000005 (EXCEPTION_ACCESS_VIOLATION) at 000000003F1885DE

    Register dump:
    Crash dump successfully written to file 'C:\PROGRA~1\Tivoli\tsm\baclient\dsmcrash.dmp'

    Stack Trace:
    0x000000003F1885DE (0x00B38690 0x00000001 0x00000000 0x3F2FB99C) dsmagent.exe, cXMLiterator::cXMLiterator()+810190 bytes
    0x000000003F188A65 (0x00B0DEC8 0x00000000 0x00000000 0x00000000) dsmagent.exe, cXMLiterator::cXMLiterator()+811349 bytes
    0x000000003F2FB686 (0x3F7F3078 0x00000000 0x000003E8 0x00000000) dsmagent.exe, cXMLiterator::cXMLiterator()+2329974 bytes
    0x000000003F2FAE53 (0x00000001 0x00000001 0x00BEFEB0 0x00B0DE40) dsmagent.exe, cXMLiterator::cXMLiterator()+2327875 bytes
    0x000000003F3057E5 (0x00B13740 0x01593C60 0x00002449 0x00B0DE40) dsmagent.exe, cXMLiterator::cXMLiterator()+2371285 bytes
    0x000000003F0C5932 (0x00330640 0x3F74C910 0x015A0954 0x00000000) dsmagent.exe, cXMLiterator::cXMLiterator()+12322 bytes
    0x000000003F0C178A (0x00000000 0x00000000 0x00000000 0x00000000) dsmagent.exe
    0x0000000076E7652D (0x00000000 0x00000000 0x00000000 0x00000000) kernel32.dll, BaseThreadInitThunk()+13 bytes
    0x0000000076FAC521 (0x00000000 0x00000000 0x00000000 0x00000000) ntdll.dll, RtlUserThreadStart()+33 bytes

    Workaround: Restart the TSM Client Acceptor service, and perform again the VSS INSTANT RESTORE operation. (72561)

  • On Exchange Server 2010, when performing a VSS restore of a database that has been rebuilt or recreated (has a new database GUID value), you must specify the restore operation using the "Restore Into" options. For the GUI, that consists of clicking the right-mouse button, while the mouse is over the selected backup being restored and then choosing the database that matches the database being restored. For the command-line interface, that consists of using the /INTODB=original-database-name" option, where "original-database-name" matches the database being restored. The list of conditions that require using the "Restore Into" option when performing VSS restore operations are as follows:

    - Restoring a backup of a database that has been recreated since the VSS backup was done
    - Restoring a backup of a database that has been relocated since the VSS backup was done
    - Restoring a backup of a database that has been renamed since the VSS backup was done
    - Restoring a backup of a database into a Recovery database
    - Restoring a backup of a database into an alternate database

  • When performing VSS restore operation, if you are restoring a VSS backup of a storage group that contains databases that do not currently exist in the live storage group, the files for the old databases will still be restored but you will not see any message indicating that there is a problem. This is a harmless situation.

    Workaround: If you do not want the additional data to be restored, you can perform a partial VSS restore operation and only select the desired databases that you would like to be restored.

  • Performing VSS Instant Restore operations with the Exchange Server running in a Microsoft Cluster Server environment requires the following:

    - Microsoft hotfix KB 919117

    If you do not have this hotfix installed, you should make sure to choose either the GUI option to "Disable VSS Instant Restore" or the CLI option "/INSTANTRestore=NO". Failure to have the hotfix installed or to disable the VSS Instant Restore operations can result in restore processing errors.

  • When using IBM System Storage SAN Volume Controller v6.1.0 and/or IBM Storwize® V7000 disk systems, VSS Backups and Restores may take longer to complete than with earlier versions of IBM System Storage SAN Volume Controller. This is resolved by applying the appropriate SVC 6.1.x fix when available. This problem is tracked via SVC APAR IC73044. (65377)

  • During VSS backup operations the following error may be recorded in the Windows event log

    Device IO Control, generic_floppy_drive, Incorrect function

    The event log entry may look similar to this:

    DeviceIoControl(\\?\fdc#generic_floppy_drive#6&amp;2bc13940&amp;0&amp;0#{53f5630d-b6bf-11d0-94f2-00a0c91efb8b} - 0000000000000308,0x00560000,0000000000000000,0,00000000002BD110,4096,[0]), Incorrect function.

    This error recorded in the Windows event log does NOT indicate the backup operation failed and can be ignored. (62535)

  • Failure when trying to snapshot the volume where the IBM System Storage VSS hardware provider is installed

    If you attempt to perform a VSS snapshot of the volume where the IBM System Storage VSS hardware provider software is installed, the snapshot will fail if the snapshot set contains a DS6000, DS8000, or SVC LUN. To work around this issue, move the IBM System Storage VSS hardware provider installation to a volume not being used in snapshot operations. This is a known problem with the IBM System Storage VSS hardware provider that is currently under investigation.

  • Failure in VSS Provider communications can result in premature removal of VSS LOCAL backups. If you are performing VSS LOCAL backup operations and connection problems occur with the VSS Provider, Windows is not able to see the existing snapshots managed by that VSS Provider during query operations. If this happens, TSM may remove the inventory of those backups on the TSM Server. IBM is working with Microsoft to resolve this issue This is a known Microsoft bug with reference number Windows 8 bug 5583. Microsoft has documented the issue in the following KB:
    http://support.microsoft.com/kb/2025800

    Workaround: Before issuing TSM VSS commands, make sure VSS Provider communications is working correctly. (49369_1 and Microsoft reference number 109090353881466)

  • When running command-line interface (CLI) operations, if the screen is locked through an operation like "Quick Edit Mode" or "Cntrl-S", the VSS operation may fail.

    Workaround: Unlock the screen and/or hit the Enter key to resume operations. (55378)

  • If the TCPSERVERADDRESS option in the DSMAGENT options file is changed (from numeric form to name form or from name form to numeric form or from long name form to short name form or from short name form to long name form) after VSS backups stored on LOCAL devices have already been performed, it is possible that Tivoli Storage Manager meta data files stored in the ADSM.SYS directory structure could be left in an incorrect state. This will cause previous VSS backups stored on LOCAL devices to be expired prematurely due to the method in which Tivoli Storage Manager uses TCPSERVERADDRESS for reconciliation indexing.

    Workaround: Make sure TCPSERVERADDRESS is not changed once LOCAL VSS backups have been created. If this is unavoidable, and you want to retain the current local snapshot backups, contact IBM service for additional instructions.

  • Parallel VSS restore operations are not supported on Microsoft Windows 2008.

    Workaround: Perform VSS restore operations sequentially.

  • Failure when using the IBM System Storage VSS hardware provider on a system that contains GUID Partition Table (GPT) disk partitions


    When using the IBM System Storage VSS hardware provider, if you attempt to perform a VSS snapshot on a machine that contains GPT disk partitions, the snapshot will fail. To work around this issue, change to Master Boot Record (MBR) partitions or wait for the IBM System Storage VSS hardware provider to support GPT disk partitions. This is a known problem with the IBM System Storage VSS hardware provider that is currently under investigation.

  • DSMAGENT hang caused by invalid DSMAGENT options file

    VSS operations may hang due to an invalid option in the DSMAGENT options file (DSM.OPT). To work around this issue, correct the invalid option in the DSMAGENT options file, stop the DSMAGENT service, and retry the operation.

  • DSMAGENT memory leak during frequent VSS backups


    In rare situations where FlashCopy Manager for Exchange or Data Protection for Exchange is running extremely frequent VSS backups, it can cause the Tivoli Storage Manager Remote Client Agent Service (DSMAGENT) to remain loaded. If that happens, it is possible that the DSMAGENT service memory usage will continue to grow. To work around this issue, stop the DSMAGENT service in between VSS backups



National Language Support (NLS), and Other Issues When Operating in a Non-English Environment

  • In the MMC GUI, mnemonics appear in some tool tips for Japanese and other Double Byte Character Set (DBCS) languages. This happens when the same text used for the control's label is also used for the tooltip.

    Workaround: Ignore the mnemonics in tool tips (76487)

  • When performing RESTOREFILES operations with DBCS storage group names, the subdirectory names created may not contain all characters of the original storage group name. In this situation, the data restored is still valid and can be used without error.

    Workaround: None.

  • If you run FlashCopy Manager for Exchange or Data Protection for Exchange on a machine that is running in an unsupported language and locale combination, you must set the LANGUAGE option to a supported language and locale combination or you will receive an error message each time you launch FlashCopy Manager for Exchange or Data Protection for Exchange. The error messages displayed will look similar to the following:

    "ANS0101E Unable to open English message repository 'tdpeXXX.txt'"
    "ANS0102W Unable to open the message repository tdpeXXX.txt. The American English repository will be used instead"
    "ANS0113E" Message repository 'tdpeenu.txt' has an invalid control record."
    "Starting client using default English language"

    After the error messages are displayed, the application will launch and use the US English language and locale combination.

    Workaround: Set the LANGUAGE option to a supported language and locale option.


Exchange Server 2007 Continuous Cluster Replication Information


  • Transaction logs not truncated after VSS backup of a Passive CCR node on Exchange 2007 SP3

    After a VSS backup of an Exchange Cluster Continuous Replication passive node, the Exchange logs on the active node are not truncated. This can lead to the log volume running out of space, or it may cause incremental or differential backups on the passive node to fail.

    From the Microsoft knowledge doc: http://support.microsoft.com/kb/2297394/

    there may be an error similar to

    Event Type:         Error
    Event Source:        ESE
    Event ID:      214
    Description:
    MSExchangeIS (xxxx) <storage group>: The backup has stopped with error -521.

    in the application event log on the active node of the cluster. To resolve this problem, install the following update rollup:

    2279665 Description of Update Rollup 1 for Exchange Server 2007 Service Pack 3 http://support.microsoft.com/kb/2279665



Exchange Server 2010 Database Availability Group (DAG) Backups

  • Exchange DAG status information is not translated Exchange status "DAG status" is shown as English string in MMC and CLI, regardless what language pack is installed.

    Workaround: N/A (76706)

Rate this page:

(0 users)Average rating

Document information


More support for:

Tivoli Storage Manager for Mail
Data Protection for MS Exchange

Software version:

6.4, 6.4.1

Operating system(s):

Windows

Reference #:

1608297

Modified date:

2013-08-20

Translate my page

Machine Translation

Content navigation