Forum Discussion

Dharsana's avatar
Dharsana
Level 4
10 years ago
Solved

Drive frequently going down

Hi,

Master and media server are same - OS windows server 2008 R2 standard,Netbackup 7.1.0.3.

One of the drive is going down 3 to 4 times/day but it is working when it is up. Drives are used for duplication.

Job details shows 84 or 86 error code and Event viewer shows "TLD(0) [6112] Unload requested, but no unload, drive 2 (device 1)" error.

And it also freezes the tapes (even new scratch tapes)

Steps carried out to fix it.

Replaced the drive (even the hardware vendor said the drive is fine, I requested him to change the drive as we got hardware errors)

Configured the drive

Updated the drivers

Restarted the Master server.

Removed the tapes which was frozen.

 

But even after that drive is going down again and again. Need help in finding the root cause of this issue.

 

7.1.x and Earlier
Backup and Recovery
Managing Backup Devices
NetBackup
Troubleshooting
Windows Server (2003-2008)
  • Marianne's avatar
    Marianne
    10 years ago

    Not sure why we haven't seen this post previously....

    This error looks like device mapping mismatch:

    Unload requested, but no unload, drive 2 (device 1)" error. 

    If there is no Persistent Binding between OS and HBA, then the OS device names may change when the server is rebooted.

    So, my suggestion:

    Find out which make/model hba in the server, then download the management GUI from hba vendor site (e.g. SanSurfer for Qlogic).
    Use the tool to config Persistent Binding, then delete all tape drives in NBU.
    Use the Device Config Wizard to re-add devices.

    Use robtest to test mount/dismount of tapes.

    Robtest commands that can be used to test the SCSI functionality of a robot 
    http://www.symantec.com/docs/TECH83129

  • Marianne's avatar
    Marianne
    Level 6
    10 years ago

    Not sure why we haven't seen this post previously....

    This error looks like device mapping mismatch:

    Unload requested, but no unload, drive 2 (device 1)" error. 

    If there is no Persistent Binding between OS and HBA, then the OS device names may change when the server is rebooted.

    So, my suggestion:

    Find out which make/model hba in the server, then download the management GUI from hba vendor site (e.g. SanSurfer for Qlogic).
    Use the tool to config Persistent Binding, then delete all tape drives in NBU.
    Use the Device Config Wizard to re-add devices.

    Use robtest to test mount/dismount of tapes.

    Robtest commands that can be used to test the SCSI functionality of a robot 
    http://www.symantec.com/docs/TECH83129

  • Michael_G_Ander's avatar
    Michael_G_Ander
    Level 6
    10 years ago

    Think it might be worth to look at if there are anything else is accessing the tape drive. In some Windows versions the internal driver probe function could cause things like this.

    Another thing can be the rewind time if you are using some of the bigger tapes

    Does the robot log show anything at the time of the error ? Have had a robot that had an intermittent unload issue on the top drive. In that case the log showed that the unload on that drive always took multiple tries, where the was one or two on the other drives.

    Lastly a running process can prevent a unload, have seen this with bptm not being closed down before the unload command was recieved by the drive. I guess that would show in the bptm and the volmgr/debug logs, here my guess would be tpreq

    If not the drive, could be something like a cable or sfp ?

    Hope this helps