cancel
Showing results for 
Search instead for 
Did you mean: 

Tape drive - scsi id issue

Peaches_K
Level 3

Hi,

I get a pending request for a tape drive often, but sometimes Netbackup is able to mount tape and backup is running on that tape drive. The issue is intermittent. 

For that drive, when no backup is running, robtest shows :

drive 21 (addr 278) access = 1 Contains Cartridge = no
SCSI ID from drive 21 is 20

The drive 20 has backups running well. Its serial number is showing up in scan and tpautoconf -t output. 

I need help on how to proceed with this. 

Backup environment : Windows 2008 master, Netbackup 7.7.2. Twenty SSO tape drives shared among 4 media servers (one of them is a 5220) , master is robotic host.

4 REPLIES 4

Marianne
Moderator
Moderator
Partner    VIP    Accredited Certified
Why do you feel that the scsi id is causing the issue here?
If
scsi id 0 = drive position 1
scsi id 1 = drive position 2
Then
scsi id 20 = drive position 21
Right?

When you run robtest for any reason, this utility controls the robot and no mounts or dismounts can take place.
Rather check drive assignment with commands such as these on the master:
vmdareq
vmoprcmd
nbrbutil -dump



Hi Marianne, 

It is only drive 21 which shows that when I do a robtest and I have not seen that before. 

I am not knowing if this is an issue that needs to be addressed. I have come across multiple pending requests which I am unable to assign and the job fails when I deny the request. 

We have some long duplicate jobs running and when I deny a request, the job fails after running for more than one day. 

drive 8 (addr 265) access = 1 Contains Cartridge = no
drive 9 (addr 266) access = 1 Contains Cartridge = no
drive 10 (addr 267) access = 1 Contains Cartridge = no
drive 11 (addr 268) access = 1 Contains Cartridge = no
drive 12 (addr 269) access = 1 Contains Cartridge = no
drive 13 (addr 270) access = 1 Contains Cartridge = no
drive 14 (addr 271) access = 1 Contains Cartridge = no
drive 15 (addr 272) access = 1 Contains Cartridge = no
drive 16 (addr 273) access = 1 Contains Cartridge = no
drive 17 (addr 274) access = 1 Contains Cartridge = no
drive 18 (addr 275) access = 1 Contains Cartridge = no
drive 19 (addr 276) access = 1 Contains Cartridge = no
<< Press return to continue, or q and return to stop >>

drive 20 (addr 277) access = 1 Contains Cartridge = no
drive 21 (addr 278) access = 1 Contains Cartridge = no
SCSI ID from drive 21 is 20
drive 22 (addr 279) access = 1 Contains Cartridge = no

 

Thanks.

Havn't that particular issue before, but pend is often cause by a process locking the tape drive. Unless you can identify the process, the solution for me has been to reboot/IPL the tape drive.

Being on windows, there a some services and internal checks that should be disabled, which was described in windows performance tuning technote. Sorry don't have the current link.

Besides the described stuff, I have see things like tape monitor programs cause issue similar to this

Another thing that is very important when runing SSO is the SCSI reservations.

The standard questions: Have you checked: 1) What has changed. 2) The manual 3) If there are any tech notes or VOX posts regarding the issue

SplashMasterson
Level 4
Certified

Assuming this has worked in the past, the things I do when we have SCSI issues:

  • In Windows Device Monitor make sure all dives are preset and update drivers if necessary
  • Log on to the library console and verify there are no errors
  • Check physical connections
  • Scan the tape drive using the vendors support tools (for example HP offers the HP Tape Tool Utility)
  • If you're usins a switch, check the switch logs, zoning, etc
  • Send tape and library logs to vendor if possible

If nothing comes up I assume there is something in the OS like a "stale SCSI connection" and I do a full reboot

  • Reboot Switch (wait for it to come up)
  • Reboot Library (wait for it to come up and fully initilize)
    • Do an inventory to make sure it's connected correctly
  • Reboot Master (wait fot it to come up)
  • Reboot all Media Servers
  • Verify you can see all the drives on Master and Media Servers after reboot

If you still have issues

  • Delete all the drives on the Master Server (take note of the density of the drives before deleting. Example, HCART3)
  • Reboot Master and Media Servers again
  • Configure Storage Devices again
    • Start with the Master Server/Robot Control Host only
    • Monitor progress
      • Does the status of the fist scan show all drives?
      • Do all drives say "Configured" under the review?
      • Since you're using SSO do all drives say "(shared)"?
      • Since you deleted all of the drives make sure the drive Density is correct before applying configurations. I feel it is easier to do it here than after the configuration
    • If all these checks were successful Configure Storage Devices on all the Media Servers next

If you still have issues try replacing the cable and/or then the drive. Engage support.