Tape Drive issue - Tape Not Mounting

[Gnana_Sekaran]

Master server Version 7.5  (Redhat Linux)

Netapp Filer

Tap Libary SL 3000

Tape library is directly attached with Filer & 5 drives are zoned to this filer .

Out of 5 ,2 drives are not working and getting error 

DecodeMount: TLD(0) drive 1, Actual status: Unable to open drive
 tldd[25870]: TLD(0) DismountTape ****** from drive 1

When we do robtest ,Could see two drives with same SCSI addr and drive index.(might casuing this issue).

Should we need to change anything in Netbackup or this has to be changed at Library end & how.

drive 1 (addr 1000) access = 0 Contains Cartridge = yes
Source address = 2000 (slot 1)

drive 1 (addr 1000) access = 0 Contains Cartridge = yes

[RamNagalla]

how about the Library end..

are you able to locate these 2 Drives and library and do mount and Dismounts from library console?, if not yet tried i would suggest you try from library end and see the results.. if it giving any error need to fix the hardware/library end issue first before anything else to check...

[Gnana_Sekaran]

Nagalla ,

Tested library end by mouting and unmounint the tape from that drives .its works fine and no errors.

[Marianne]

Where are you running robtest from?

Do you have a master or media server as robot control host?

Are devices shared with master and/or media server?
If so, are backups working using master and/or media server?

Please show us all output when running s d in robtest.

Show us output of 'scan' as well on robot control host.

[Michael_G_Ander]

Check the tape drive configuration on the netapp filer and make sure that the configuration is put into the startup script(s). rc back when I worked with netapp

[Gnana_Sekaran]

@Marianne,

There is only one master server and no media servers associated.

robot control is with the filer & I am running robtest from master server. Checked with Storage admin and all the drives are visible at filer end.

Opening abccc
MODE_SENSE complete
Enter tld commands (? returns help information)
s d
drive 1 (addr 1000) access = 0 Contains Cartridge = yes
Source address = 2000 (slot 1)
Barcode = XS***
drive 1 (addr 1000) access = 0 Contains Cartridge = yes
drive 2 (addr 1001) access = 0 Contains Cartridge = yes
Source address = 2412 (slot 413)
Barcode = SA***
drive 3 (addr 1002) access = 0 Contains Cartridge = yes
Source address = 2030 (slot 31)
Barcode = SA****
drive 4 (addr 1003) access = 0 Contains Cartridge = yes
Source address = 2064 (slot 65)
Barcode = SA****
READ_ELEMENT_STATUS complete

[Marianne]

As far as I know robtest can only be used on the robot control host.
Does this mean that the master also has access to the library?

Please send output of 'scan' and 'tpconfig -l' from the master server.

I see that there are tapes in all of the tape drives - did you mount them with robtest or were they mounted during failed backup attempts?

Can you use robtest to move all tapes back to their respective slots?

The output for drive 1 (2 entries) does not seem right - hopefully 'scan' output will help.

[WVT]

Do you have two fibre cables plugged into drive 1? Seeing two drives at element address 1000 is strange.

[jim_dalton]

>storage show tape

on netapp...check you can see each drive in turn...then try and write some data to each in turn from the filer cmdline. Use robtest to mount the tape in the drives in turn, check the drive you think its mounted in is the right one. this will check the robotics and the netapp are all correct.if it is then it has to be netbackup configuration: recreate storage unit ...jim

[Gnana_Sekaran]

@Marianne,

Below are the outputs as requested.

[root@]# scan
************************************************************
*********************** SDT_TAPE    ************************
*********************** SDT_CHANGER ************************

[root@logs]# tpconfig -l
Device Robot Drive       Robot                    Drive                 Device  Second
Type     Num Index  Type DrNum Status  Comment    Name                  Path    Device Path
robot      0    -    TLD    -       -  -          -                     mc0 (Filer ABC)
  drive    -    0  hcart    3      UP  -          HP.ULTRIUM4-SCSI.000  nrst6a
  drive    -    1  hcart    1    DOWN  -          HP.ULTRIUM4-SCSI.001  nrst1a
  drive    -    2  hcart    4      UP  -          HP.ULTRIUM4-SCSI.002  nrst0a
  drive    -    3  hcart    3    DOWN  -          HP.ULTRIUM4-SCSI.003  nrst2a
  drive    -    6  hcart    2      UP  -          HP.ULTRIUM4-SCSI.006  nrst5a

I can move tapes using robest from drive to slot and slot to drive.

@WVT ,will check with my onsite support about the fc connections .

[RamNagalla]

 drive    -    3  hcart    3    DOWN  -          HP.ULTRIUM4-SCSI.003  nrst2a

 drive    -    0  hcart    3      UP  -          HP.ULTRIUM4-SCSI.000  nrst6a

robot drive 3 is configured twice, and looks it is seeing twice.. provably you can delete once drive from netbackup and see how it goes.. meanwhile you can the physical connectivity with onsite team.

[jim_dalton]

when youve moved a tape to a drive can you write data to it ? This helps because you can map each drive , with its serial , and its nrst, and then you know for all drives the useful attributes. Then when you use netbackup to do the management  you'll be able to check that it has the same info you know is correct. Jim

[Gnana_Sekaran]

Hi All,

We worked with the Tape Library Vendor and removed all the struck tapes and upgrade the one code on the drives.

Now drives are working as expected.