On Nov 5, 2005, at 8:00 AM, Jones, Eric J wrote:

...Currently the UNIX folks are issuing 22 restores and it happens
that all
the data is on a single tape which means it has to be loaded 22 times
and unloaded 22 times. ...

Well, that depends upon your MOUNTRetention value.  Depending upon
general environmental factors, you may be able to boost this to cover
latency if restorals are frequent enough to warrant the action. And,
depending upon volume of data, you may be able to keep a reasonable
amount on disk, rather than tape, via a File devclass or Disk type
with caching. Given that all that data was on one tape, it seems like
having a good amount of disk in your storage pool hierarchy could
cover a lot of restorals.

   Richard Sims