Forum Discussion

Travis_Bilingto's avatar
18 years ago

Tape drive dissapears after reboot

Running BE 11 on W2K3 server.  Tape drive is an IBM VXA-3 on its own controller card by itself.  Everytime that we simply reboot(maybe due to an automatic update), the drive dissapears from BE, but if we actually shut the system down and power back up again the drive is fine.  Please help.

5 Replies

  • Is it external?  If yes,  go to a shorter cable.   I have seen 6 foot cables not allow the signal to reset the device, when the controller was restarted.   It only resets the SCSI bus on off/on.  
  • Hi Travis,
     
    Check in the System Event log, to see if there are any errors being generated by the SCSI controller or the tape hardware.
     
    Greg
  • Event log will show you what you already know.   The causes of restarting and server and no reconnecting are a few things
     
    My first SCSI drive was 85 MB SCSI-I.  I have used SCSI scanners and tape drives, since they came out. 
     
    SCSI device and controller did not reset initialization.   This can be a bad device, bad terminator, bad cable, or the device signal strength is not returning the reset to the controller or vice versa properly.   I had a Sony replacement AIT external drive do this on a 6 foot cable.   It was LVD and should work fine on a 6 foot cable.  Sony recommended a 3 foot cable because the device had signal issues.   I returned the device for my money back because it violated SCSI standards for cable length and usage. 
     
    Run the manufacturer tools for quick and long tape test and other diagnostics to check for faulty drive.
     
    Replace cable and terminator to see if its the issue.  I have seen two terminators go bad in 20 years of SCSI.  I have seen them fail on hard drives internally.   Active termination (on the cable) is better than passive termination on devices.  
     
    It could be a flaky SCSI controller causing it.  I had a few dual channel cards with a bad channel in 20 years.  Most were older SE cards.   Usually new cards don't fail as easily.   Most of the time, upgrading the firmware or reflashing the firmware fixes the controller.  
     
    Upgrading the tape firmware or reflashing it fixes tape device issues.   I have seen this with my Exabyte drives.  It is like the firmware gets corrupted slightly.   Reflashing it can refresh the device.   They also slip in fixes for SCSI reset and other SCSI related issues. 
     
    On HP Proliants, I have seen a BIOS rev upgrade needed for compatibility on a new SCSI controller or device.  I don't usually see this on other servers or my Tyan Thunder boards.   Using HP drivers and firmware and only HP tape drives helps with compatibility and reduce SCSI issues.   My 39160 would have issues in the ML370G2 even though it was same phyiscal cards as their SCSI controller but without HP firmware.    I heard similar things with Dell and their controllers. 
     
     
  • Thanks for the replies guys.
     
    This is an internal drive terminated on the cable.  It is the only device on the chain.
     
    There are no error in the eventlogs for SCSI or tape devices. 
     
    I will check the firmware on the drive as well.
  • Did you mix SE and LVD devices on that cable?   Do you have an active terminator on that cable or is last device the terminator?   Do other devices on that cable disappear?