Hi,

Amanda hasnt been acting as it should.  See the output of a failed dump below:

STATISTICS:
                          Total       Full      Incr.
                        --------   --------   --------
Estimate Time (hrs:min)    0:04
Run Time (hrs:min)         3:07
Dump Time (hrs:min)        5:03       5:03       0:00
Output Size (meg)      618864.1   618864.1        0.0
Original Size (meg)    618864.1   618864.1        0.0
Avg Compressed Size (%)     --         --         -- 
Filesystems Dumped           61         61          0
Avg Dump Rate (k/s)     34866.3    34866.3        -- 

Tape Time (hrs:min)        3:03       3:03       0:00
Tape Size (meg)        446721.9   446721.9        0.0
Tape Used (%)             111.0      111.0        0.0
Filesystems Taped            60         60          0

Chunks Taped                 74         74          0
Avg Tp Write Rate (k/s) 41721.8    41721.8        -- 

USAGE BY TAPE:
  Label            Time      Size      %    Nb    Nc
  AmaTor-040       3:05 462842042K  112.3    59    60
  AmaTor-041       0:00     2912K    0.0     0     1
  AmaTor-043       0:00    48576K    0.0     0     1
  AmaTor-044       0:00     3296K    0.0     0     1
  AmaTor-045       0:00     3680K    0.0     0     1
  AmaTor-046       0:00     7840K    0.0     0     1
  AmaTor-047       0:00    10848K    0.0     0     1
  AmaTor-048       0:00    10048K    0.0     0     1
  AmaTor-049       0:00     3680K    0.0     0     1
  AmaTor-050       0:00     3296K    0.0     0     1
  AmaTor-051       0:00     3296K    0.0     0     1
  AmaTor-052       0:00     3296K    0.0     0     1
  AmaTor-053       0:00     3296K    0.0     0     1
  AmaTor-054       0:00     3296K    0.0     0     1
  AmaTor-055       0:00     3456K    0.0     1     1

Why is it writing all that data to the first tape, then failing to write more 
than 3-10mb on the rest of the tapes?

We have never observed this behaviour before.  Our DLE's are sizes ranging from 
1gb to 200gb, and we have been fine doing as such up until now.

We used root-tar on our DLE, taperalgo largestfit and dumporder "TTTTTT" - as 
from the amanda wiki.  We also tried reverting back to taperalgo first and 
dumporder "sssS" to see if this wuold make a difference and we still saw these 
errors.  

The only thing i can see from the amdump is:
driver: interface-state time 17584.842 if default: free 98976
driver: hdisk-state time 17584.842 hdisk 0: free 1420454186 dumpers 1
driver: result time 17584.842 from taper: NEW-TAPE 01-00121 AmaTor-054
Got EIO on /dev/nst0, assuming end of tape.
Error writing final filemark: Input/output error
Error writing final filemark: Input/output error

what does this suggest? the tapes need to be rewound? im doing so for our next 
(attempted) dump. im gonna go back to using'span' in our DLE too, as we never 
had this tape writing problem previously.

Please let me know if you need more info, 

thanks.

+----------------------------------------------------------------------
|This was sent by rory AT mrxfx DOT com via Backup Central.
|Forward SPAM to abuse AT backupcentral DOT com.
+----------------------------------------------------------------------