> 
> This seems to be a veeeeeeerrrrry old config.
> It still uses parameter "diskdir" which is deprecated for
> over 8 years now.  Moreover in 2.5.0 is does't exist anymore.
> Also the comments about infofile being a database filename (which
> is was in Amanda 2.3 or so -- it is a directory now), seems to
> indicate that.
> 
> On the other hand, you are using "displayunit" which exists only
> in very recent releases :-)
> 
> That said, you should inspect the debug files on the client
> and see at which time each of those "double" backups is created.
> And how that is possible???
> Also in the server log file "amdump.X", you should find more info
> where those double backup images are coming from.

Actually I just copied the script that comes with the Debian Sarge packages 
(that is 2.5.0)
and made the changes I needed. I've updated that, thanks.

On my dump logs reagards one of the DLE I have
pondering acag.ess.co.at:/var/www... next_level0 2 last_level 1 (not due for a 
full dump, picking an incr level)
   pick: size 3257520 level 1 days 1 (thresh 20480K, 1 days)
   pick: next size 51740... BUMPED
  curr level 2 size 18826 total size 21552813 total_lev0 20827585 
balanced-lev0size 49104002

then
INITIAL SCHEDULE (size 97290001):
...
  acag.ess.co.at /var/www pri 1 lev 2 size 18826
...

the it gets promoted:
no try oldg.ess.co.at:/BACKUP/scripts 2 4 2 = -11
   promote: moving acag.ess.co.at:/var/www up, total_lev0 118832423, total_size 
121480639

and dumped:
DUMP acag.ess.co.at fffffeff9ffeffff07 /var/www 20060701 1 0 1970:1:1:0:0:0 
2130192 787 2 2006:6:28:22:2:5 18826 6

and gets to tape, so I don't see why in the report I receive by mail have
acag.ess.co. /var/www    0    3438    2123   61.8   18:25 1968.6   6:36 5487.8
acag.ess.co. /var/www      MISSING -------------------------------------------

and on amstatus have 
acag.ess.co.at:/var/www        0     2123m finished (5:52:13)

Any more ideas whould be helpfull.

Luis