Howdy all,

I'm using Amanda with a Dell PV120T tape library, which has a DTL 4000 
drive capable of writing 20GB uncompressed data to each tape.  I have a 
directory that's currently tipping the scales at 30GB, and is only going 
to get larger.  Normally, I would look into splitting this backup apart 
by subdirectory, but that's not going to work here:

* There are currently 409 subdirectories beneath the parent.

* The structure of these subdirectories is mandated by the application, 
so I can't create some sort of top-level "grouping" directories.

* Directories will be created, deleted, or renamed.

A ridiculously simple solution to this whole problem would be using 
wildcards in the disklist file, e.g.,:

  localhost /export/data/[a-mA-M]* dumptype
  localhost /export/data/[n-zN-Z]* dumptype

But of course this doesn't work :(.

One possible solution I've considered is running a script prior to 
running amdump that would generate a disklist file automatically.  Can 
anyone suggest any alternatives?

Thanks,

-- Lars