Medium changer not getting detected.

[shaz1987]

Hi All,

Need help. I am configuring two new media servers, the tape library is scalar i6000 with HP LTO5 tape drives. 

I have done the zoning of the tape drives with the media servers and also installed the HP drivers.From the device manager, I am able to see the tape drives but not the medium changer. Can anyone to understand what I missed out?

[mph999]

Is it detected at all, even if an unknown device ?

[mph999]

How does the robot connect to the server, is it a separate cable from the drives, or does it share one of the drive connections. 

If it has its own FC, can you see it's WWN using the HBA utility on the server, e.g. HBANYwhere (Emulex) other Sansurfer (Qlogic) depending on the brand of the HBA.

[sdo]

Try the command:

> "C:\Program Files\Veritas\Volmgr\bin\scan.exe" -changer

...if nothing is shown, the you may have missed presenting the robot / medium change to one of your backup hosts, i.e. typically one of the hosts that you also present the tape drives too.

.

Also, see if there are any tips / similarities here:

https://www-secure.symantec.com/connect/forums/sso-and-fc-io-blade-configuration-question

[Marianne]

You say '2 new media servers'. Does this mean that there are more media servers? And that the robot control is on another media server? If you have zoned tape drives to new media servers, there is no need for them to see the robot as well. Simply select the robot control host and the new media server in the Device config wizard.

[mph999]

Good point Marianne ...

If the robot is seen (controlled) by an existing media server then te other media severs, that only have drives talk to the 'robot control host' when then need tapes moved.

In more detail.

robot control host runs tldcd process (for TLD type robot).  This process talks directly to the robot.

The other media servers (with drives) only run the tldd process, which talks to tldcd over the network on the robot control host.  (THe RCH also runs tldd for the same reason).

[shaz1987]

Thanks for your reply guys.

There is no separate cable connectivity for robot.One of the drives has been marked as Primary Control path.

As of now, Netbackup is now discovering the drives but they are all coming up as standalone. 

[sdo]

There may be some pointers in this post:

https://www-secure.symantec.com/connect/forums/sso-and-fc-io-blade-configuration-question

[mph999 · Accepted solution]

OK, thx - so it works via a control path, and you see the drive that is marked as the control path so that makes it sound like some issue on the library.

I would suggest loading up the HBA utility (hbanywhere/ sansurfer depending on HBA brand) and using that to see what can be seen 'down the fibre channel'.

If it's not seen, then we know it's not being presented from the library correctly.

The HBA utility works at the scsi layer, so kinda below the OS.  I'm not sure if there is a windows coand to do the same thing.

[Marianne]

No reply in 7 days....

Have you checked 'scan -changer' on all media servers?

Can you see "Unknown Medium Changer" in Windows Device Manager on the media servers?

Have you downloaded and installed HBA utility as per Martin's suggestion?

[areznik · Accepted solution]

Scalar i2k/i6k has an internal zoning tool. You have to go in there and assign the robot control to the drive or IO blade you want it to be seen under. 

Then when you're on your host you will see the drive come up as <wwn>:Lun0 and the robot as <wwn>:Lun1; if you're using IO blades you will have up to 5 luns (depending on how many drives you have on that blade, I think 4 is max) and one of them will be the bot. 

Hope that helps.