Forum Discussion

trs06's avatar
trs06
Level 5
13 years ago

Windows File Server errors with Status 13

Windows 2008 server with client 7.1.0.4.  Master Server (REHL 6.2) and Media Servers (Windows 2008) are 7.1.0.4.

The above client has been problematic for weeks.  I seldom get a successful backup though I have from time to time.  Frequently I get Detailed Status messages saying that a "data is invalid" yet the file seems perfectly valid and opens as expected.  See the below excerpt from the Detailed Status:

"01/20/2013 14:39:02 - Error bpbrm (pid=10936) from client utlp5153: ERR - failure reading file: G:\DEPT\Pxx\PJD\1 - NETWORK\Docs\Product Marketing and Specs\Nortel\Nortel Enterprise Network Management System Requirements.pdf (WIN32 13: The data is invalid. )
01/20/2013 14:39:32 - Error bpbrm (pid=10936) from client utlp5153: ERR - failure reading file: G:\DEPT\Pxx\PJD\1 - NETWORK\Docs\Product Marketing and Specs\Nortel\SR1004 - Product Brief.pdf (WIN32 13: The data is invalid. )
01/20/2013 14:40:02 - Error bpbrm (pid=10936) from client utlp5153: ERR - failure reading file: G:\DEPT\Pxx\PJD\1 - NETWORK\Docs\Product Marketing and Specs\Nortel\SR3120 - Product Brief.pdf (WIN32 13: The data is invalid. )
01/20/2013 14:40:32 - Error bpbrm (pid=10936) from client utlp5153: ERR - failure reading file: G:\DEPT\Pxx\PJD\1 - NETWORK\Docs\Product Marketing and Specs\Proofpoint\DS Enterprise Privacy Suite.pdf (WIN32 13: The data is invalid. )
01/20/2013 14:47:49 - Error bptm (pid=7248) socket operation failed - 10054 (at child.c.1294)
01/20/2013 14:47:49 - Error bpbrm (pid=10936) socket read failed, An existing connection was forcibly closed by the remote host.  (10054)
01/20/2013 14:47:49 - Error bptm (pid=7248) unable to perform read from client socket, connection may have been broken
01/20/2013 14:48:10 - end writing; write time: 18:47:49"

It isn't always even the same files or directories.  I have been unable to ascertain any sort of correlation between any of the failures.  This job ran for 21:48 hours and backed up ~ 473 GB before disconnecting.  It will backup anywhere from a few hunderd giga bytes to almost completing the backup with ~ 800 GB and run anywhere from a few hours to 30 hours then die.
 

7 Replies

  • We have verified that the file system is NTFS.  The server is a VMWare VM with a Netbackup Client installed.

  • hi,

    are you getting these failurs only for G Drive?

    did you fine any errors in Window event viewer?

    disk check might be requires?

     

    please enable the bpbkar (in clients) and bpbrm(in media server) logs in the  with Verbose=5 and re-run the backup and provide the logs.

     

  • You forgot to tell us which OS on client? There are specific TNs for W2003 and W2008.

    Is NBU Client Service starting with Domain account with full access to all files/folders on the File server?

    Have you tried to break down large data drives into smaller streams with 'Allow Multiple data streams selected in policy attributes?

    e.g.

    NEW_STREAM
    G:\folder\folder1
    G:\folder\folder2
    G:\folder\folder3
    NEW_STREAM
    G:\folder\folder4
    G:\folder\folder5
    G:\folder\folder6
    NEW_STREAM
    G:\folder\folder7
    G:\folder\folder8
    G:\folder\folder9
    NEW_STREAM
    G:\folder\folder10
    ....
    ......

     

    If MPX in schedule and STU is set to 4 and Max Jobs per Client is set to 4, you will get 4 simultaneous streams for G-drive, writing to same tape drive, resulting in much faster overall backup.

  • I will get the log files.  Utilities like chkdsk are impractical for us since this is a production environment, the server drives are presented by a VMAXe Storage Array and the drive has to be offline to run chkdsk and I think it would take an excessive time to chkdsk a TB drive.

    My first sentence stated the OSs of the client, master and media servers:

    "Windows 2008 server with client 7.1.0.4.  Master Server (REHL 6.2) and Media Servers (Windows 2008) are 7.1.0.4."  [I meant RHEL 6.2]

    Multiplexing is 1 and Max Jobs per policy is not checked.  The target is a DataDomain dd670 and is an Open Storage Technology (OST) device that dedups all backups and is set to 64.  So you are saying I can simply set Multiplexing to 4 to get the desired increase?

    This same client has a policy to backup F:\ drive and All other drives with policy exclusions for G:\ and F:\.  The All Drives and the F:\ always run perfectly with Status 0.  The All drives is ~243 GB and the F:\ is ~ 100 GB.  The G:\ is the big one with over 900 GB for our weekly full backup that I try to do when traffic and backup are less busy.

  • I am saying that you will get 4 simutaneous streams if:

    • MPX in schedule is set to 4
    • MPX in STU is set to 4 (or more)
    • Max Jobs per Client is set to 4 (in Master server Host Properties -> Global attributes)
    • Backup selection broken up into 4 or more streams
  • The log files have other backups mixed in with them from the same server and backing up the same directories and files.  I believe this will make it very hard to read and analyze.  I am attaching them but will try and find a time to run the backup I'm after without the other information interspersed in it.

    Also on this morning's backup of this server I got the same error as in the below but all of my backup versions are 7.1.0.4, client, media server, master server.

    https://www-secure.symantec.com/connect/forums/backup-skipping-system-files-wrn-cant-open-file-system-stateautomated-system-recoverybcdbcd-w

  • Do you have Anti Virus installed on your client(s) - if so have you excluded all NetBackup processes from scanning - bpfis and bpbkar32 need to be excluded at least

    The original log you posted indicates a timeout which could be a locked file caused by AV or worse still Application Access Protection (or what ever it is call - a McAfee thing that causes real problems for backups)

    Check you AV logs to see if it is just locking files stopping NetBackup from backing them up - if files were corrupt you would normally get a slightly different win32 error.

    Worth a check - and if it is the case your backups will go much faster afterwards!

    Hope this helps