cancel
Showing results for 
Search instead for 
Did you mean: 

Tapes drive goes AVR to Down

Vamsi1
Level 5

Team,

In my environment, Master server hosted on Solaris which no media servers.

2 tape libraries were connect one is Qutnam i500 & i40. Frequently I get an issue with the i40. Current issue is the i40 tape drivers are one is in AVR mode and other is down. I have logged in to the library console and all looks good.

Currently I couldn’t restart the backup services as the backups are Full backups are running. 

Can some one help with the restarting the services can the Tape drives bring up and also I would like to know why the only i40 tape library is getting issue, Need guidance on it.

1 ACCEPTED SOLUTION

Accepted Solutions

RiaanBadenhorst
Level 6
Partner    VIP    Accredited Certified

Check you messages log, there should be evidence of communication to the robot being lost (drives going into AVR). There might some SCSI issues.

View solution in original post

7 REPLIES 7

RiaanBadenhorst
Level 6
Partner    VIP    Accredited Certified

Check you messages log, there should be evidence of communication to the robot being lost (drives going into AVR). There might some SCSI issues.

Marianne
Level 6
Partner    VIP    Accredited Certified
It seems the OS is losing connection to the library. As per Riaan's post, check /var/adm/messages for device-related errors. Verify that OS can see both robots with cfgadm. See if devices respond to scsi_commands with commands such as scan and sgscan.

sdo
Moderator
Moderator
Partner    VIP    Certified

You say that frequently you have an issue with the i40.  This implies that frequently you are able to temporarily resolve the issue.  If you can tell us what is that you do (the process / procedure / steps / commands / actions / choices) when you resolve the issue, then I think this will help steer us to a hopefully fairly quick solution.

Vamsi1
Level 5

Thank you very much all for the responses. 

Previous when ever the tape drives go down we will perform the reboot of the tape library, By that most of the time Drives will up or will perform the netbackup services restart.

But in yesterday case, After a lot of troubleshooting I found the issue after reviweing the /var/adm/messages file.

We have 30 Slots in the library with 2 tape drives. When a restore is in progress one tape got picked by teh tape drive. In the same time local IT though the slot was empty and he loaded a scratch tape into the empty slot.

Once the restore is completed the drive cant release the tape as the slot is already occupied. That leaded the tape drive to go offline.

 

RiaanBadenhorst
Level 6
Partner    VIP    Accredited Certified

That is a common problem :)

Marianne
Level 6
Partner    VIP    Accredited Certified

Tape handling by the local IT team needs to be changed:

They should NOT open the library door to add or remove tapes.

Opening the robot door will effectively put robot in a 'DOWN' state, and drives will go to AVR (no robot) state in NBU.

Putting tapes in slots that actually belong to tapes in tape drives will result in drives going DOWN when the tapes need to be unloaded and returned to their slots.

Local IT team must be taught and instructed to ONLY use the CAP/MAP of the robot to load and unload tapes. 
Robot will stay UP and tapes will put into slots known by robot as empty (i.e. not home location for tapes in drives).

Vamsi1
Level 5

Yep I agree. I have already diccussed with the Local IT team and informed the same.