[Bug-tar] GNU tar 1.26: multi-volume continuation issue

[peter.chiu]

Hello,

 

This is not really a bug, but a request for help solve a problem we have

in using tar to read a multi-volume set of LTO5 tapes.

 

We have received a few sets of LTO5 tapes, knowing it was written using the command at source site:

 

   tar --create -v --multi-volume --new-volume-script=/etc/tape/nexttape.sh \

   --totals --blocking-factor 2048 --file=/dev/nst0 -C /disk  /disk/foo /disk/bar /disk/bazz

 

 

We have a SUSE Enterprise Server 11 Sp2 host connected with a Quantum Superloader3 8-slots and 1 LTO5 tape drive.

The tape drive is set up (as default) in random mode, allowing the host to query and maneuver the tapes.

 

Having placed the first volume tape (tape 3 with barcode JM036L5) into the drive, I used the following commands

 

mt –f /dev/nst0 setblk 0

 

time tar -b 2048 -xv -M -k -F /csupport/script/nexttape \

  -C /group_workspaces/cems/globalalbedo/JMP036L5 -f /dev/nst0 1>>/tmp/JMP036L5.

 

where my nexttape contains:

. /etc/profile

PATH=$PATH:/usr/sbin

if [ "${ThisLog}" == "" ]

then

  ThisLog=/tmp/nexttape.log

fi

echo "`date`-`hostname -s` unload current tape ..." >> $ThisLog

mtx next

mt -f /dev/nst0 setblk 0

echo "`date`-`hostname -s` loaded in the next tape:" >> $ThisLog

mtx status >> $ThisLog

 

The above command started off reading the 1^st volume, but fails with this erro:

frost:/csupport/script # time tar -b 2048 -xv -M -k -F /csupport/script/nexttape

  -C /group_workspaces/cems/globalalbedo/JMP036L5 -f /dev/nst0 1>>/tmp/JMP036L5.

lis2

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: /dev/nst0: Cannot read: Device or resource busy

tar: Too many errors, quitting

tar: Error is not recoverable: exiting now

 

real    185m56.158s

user    0m21.693s

sys     10m43.264s

 

While the log in /tmp/nexttape.log shows:

frost:/csupport/script # less /tmp/nexttape.log

Thu Jan 31 22:59:07 GMT 2013-frost unload current tape ...

Thu Jan 31 23:02:35 GMT 2013-frost loaded in the next tape:

  Storage Changer /dev/changer:1 Drives, 8 Slots ( 0 Import/Export )

Data Transfer Element 0:Full (Storage Element 4 Loaded):VolumeTag = JMP037L5   

                    

      Storage Element 1:Full :VolumeTag=JMP034L5                       

      Storage Element 2:Full :VolumeTag=JMP035L5                       

      Storage Element 3:Full :VolumeTag=JMP036L5                       

      Storage Element 4:Empty

      Storage Element 5:Full :VolumeTag=JMP038L5                       

      Storage Element 6:Full :VolumeTag=JMP039L5                       

      Storage Element 7:Full :VolumeTag=JMP040L5                       

      Storage Element 8:Full :VolumeTag=JMP041L5                       

 

That means tape 4 has been loaded, but some how,

the tar command fails to continue with the above error.

 

Does anyone have any idea of what could be the cause?

 

Many thanks.

 

Regards,

 

Peter Chiu

 

STFC Rutherford Appleton Laboratory

UK

 

--
Scanned by iCritical.