On 10/06/2014 02:47 PM, Mike Driscoll wrote:
Jean-Louis,

$ amtape Weekly1-library1 inventory
…
slot 16: unknown state barcode CA0017
And what's the label on that tape, it it loaded in /dev/st0: mt -f /dev/nst0 rewind; dd if=/dev/nst0 bs=32k count=1
[amandabackup@ca-server10 ~]$ amlabel Weekly1-library1 slot 16
Reading label...
Slot 16, label 'Weekly1-library1-CA0014', mismatch barcode between changer 
'CA0017' and tapelist file 'CA0014'

I confirmed in the library’s management interface that CA0017 is the tape 
loaded into the drive.

In the past, I delete tapelist and /var/amanda/chg-robot-dev-changer, delete 
the contents of the state folder, and then add back some tapes using amlabel.  
After this, the first few dumps work.  Am I making any mistakes and leaving 
some old, bad data behind?  Maybe I’m not cleaning up some old, bad data 
correctly.

If the tapelist is correct, you should not remove it.
Can you post the '/var/amanda/chg-robot-dev-changer' file once you get the 'mismatch barcode between changer ' error, i want to see what get corrupted.

Jean-Louis

I’ll send debug files via separate e-mail.

Mike



On Oct 3, 2014, at 4:22, Jean-Louis Martineau <[email protected]> wrote:

On 10/02/2014 09:40 PM, Mike Driscoll wrote:
Hi Jean-Louis.  I am using version 3.3.6.  I am confident that tape-device is 
correct insofar as I can complete multiple, successful dumps before the 
mismatches begin.

amtape verify looks good…

$ amtape Weekly1-library1 verify
GOOD : Drive 0 is device tape:/dev/nst0
GOOD : Drive 1 is device tape:/dev/nst1
property "TAPE-DEVICE" "0=tape:/dev/nst0" "1=tape:/dev/nst1"





On Oct 2, 2014, at 12:50, Jean-Louis Martineau <[email protected]> wrote:

Absent that, does anyone have any advice about how to avoid 
delayed/intermittent barcode mismatch problems?  Here’s a typical example of 
the error:

ERROR: Slot 16, label 'Weekly1-library1-CA0014', mismatch barcode between 
changer 'CA0017' and tapelist file 'CA0014'
Is it true the slot 16 have label Weekly1-library1-CA0014 and barcode CA0017?
Or is it the amanda database that is wrong? (changer state file)

Can you post a debug where you first see it, and was the last operation you did 
on the changer?
Can you also port the changer state file if it is corrupted?

Jean-Louis


Server check took 245.881 seconds

amanda.conf file includes:

define changer sl150-robot {
         tapedev  "chg-robot:/dev/changer"  #tape/by-id/scsi-3500104f000d21512"
         property "tape-device" "0=tape:/dev/nst0"  "1=tape:/dev/nst1”
         length 2442818848 kbytes
         filemark 1806 kbytes
         speed 74006 kps
         blocksize 32 kbytes
}
labelstr "Weekly1-library1-CA00[0-9][0-9]"
amrecover_changer "sl150-robot"
tapedev "sl150-robot"
tapetype LTO6

Reply via email to