Late last week, amanda started to consistently produce reports such as
the following:

----- Forwarded message from backup <[EMAIL PROTECTED]> -----

*** THE DUMPS DID NOT FINISH PROPERLY!

*** A TAPE ERROR OCCURRED: [[writing filemark: Input/output error]].
*** PERFORMED ALL DUMPS TO HOLDING DISK.

THESE DUMPS WERE TO DISK.  Flush them onto tape Daily01 or a new tape.
Tonight's dumps should go onto 1 tape: Daily02.

FAILURE AND STRANGE DUMP SUMMARY:
  nova       hda1 lev 0 FAILED [disk hda1 offline on nova?]
  cat        md1 lev 0 FAILED [out of tape]
  taper: FATAL syncpipe_get: w: unexpected EOF
  bradley    md10 lev 1 FAILED [no more holding disk space]
  cat        md5 lev 1 FAILED [can't dump no-hold disk in degraded mode]
  peterbilt  //server/f$ lev 0 FAILED [can't dump no-hold disk in degraded mode]


STATISTICS:
                          Total       Full      Daily
                        --------   --------   --------
Dump Time (hrs:min)        0:12       0:00       0:00   (0:12 start)
Output Size (meg)        5584.3      569.8     5014.5
Original Size (meg)     14467.9     2202.4    12265.5
Avg Compressed Size (%)    38.6       25.9       40.9
Tape Used (%)              27.3        2.8       24.5   (level:#disks ...)
Filesystems Dumped           15          7          8   (1:8)
Avg Dump Rate (k/s)       460.3     1032.8      433.0
Avg Tp Write Rate (k/s)     --         --         -- 

DUMP SUMMARY:
                                      DUMPER STATS                  TAPER STATS
HOSTNAME  DISK           L  ORIG-KB   OUT-KB COMP%  MMM:SS   KB/s  MMM:SS   KB/s
-------------------------- -------------------------------------- --------------
bradley   md10           1   FAILED --------------------------------------------
bradley   md6            1    69220    29568  42.7    0:50  592.6    N/A    N/A
...

----- End forwarded message -----

So, first off, the big question:  Why is amanda first claiming that it
ran out of tape, then saying that only 27.3% of the tape was used?

Second, the taper stats read "N/A    N/A" for all drives that were dumped.
This would seem to indicate that nothing was actually written to the
tape, wouldn't it?  And, based on that, am I correct to infer that it
would be better to amflush last night's dumps onto the same tape instead
of advancing to the next one as instructed by the report?

And, finally, how do I actually debug and fix this?

Thanks, and more information is available on request, of course, just
tell me what would be useful.

Reply via email to