On Sun, Sep 10, 2006 at 10:48:08PM -0700, Stephen Carville wrote:
> I've been using Amanda for about five years now.  The only problems I've 
> ever had are because a single dumps has to fit on a single tape.  I 
> usually solved this by using tar, client side compression, and breaking 
> up the partitions into smaller chunks.  However, I seem to have reached 
> a limit.
> 
> One of my Oracle RMAN backups now stands at 109G.  Even with compression 
> this requires 44G to 45G of space and will not fit on an AIT-2 tape. 
> The backups are dumped into a single directory and I don't know of any 
> way to split a directory with no subdirectories into two dumps.  If 
> there is, I'd appreciate knowing how.  Any other sugesstions are welcome.
> 

Stephen,
what are we talking about in terms of files in a single directory
that total nearly 110GB?  Is it one or two huge files?  A hundred,
maybe a thousand, or more?  What are the naming schemes?  Always
the same or always predictable?  Are there name extensions for
file types?  If the numbers are reasonable you could create a
set of DLEs using include statements and a catch-all using
exclude statements.

For example, (assumptions include reasonable number of files,
say < 1000 but I'm not certain what is reasonable, most with names
beginning with lowercase letters but a few with digits or uppercase
and an ascii charset :)

  host Ora-a2g /path/to/RMAN/backup/dir {
        your_current_dumptype
        include file ./[a-g]*
  }
  host Ora-h2s /path/to/RMAN/backup/dir {
        your_current_dumptype
        include file ./[h-s]*
  }
  host Ora-t2z /path/to/RMAN/backup/dir {
        your_current_dumptype
        include file ./[t-z]*
  }
  host Ora-rest /path/to/RMAN/backup/dir {
        your_current_dumptype
        exclude file ./[a-z]*
  }

-- 
Jon H. LaBadie                  jon AT jgcomp DOT com
 JG Computing
 4455 Province Line Road        (609) 252-0159
 Princeton, NJ  08540-4322      (609) 683-7220 (fax)