The only way is to make sure you never have to restore the whole 100 GB.
 Or put more everything in including ADSM servers, network, etc.

I would expect that if one were to design the backup better, then perhaps
the restore can be done more quickly.  For instance, don't backup all of
the data to a single filespace in a single storage pool.  Rather split it
up so you can multi-thread the restore.  Depending on where the restore
bottleneck is, this might help.  The other thing to do is to run the ADSM
server on Exchange server and backup and restore the stuff directly.  Take
the network out of the game.

We really don't have any problems that sufficient money can't fix.  But we
do need to determine what is a "reasonable" timeframe and cost
appropriately.  If we lose big dollars for every minute we're down, then we
should be willing to spend big dollars to make sure we stay up.
 Conversely, if there isn't any real value in the system, i.e., we don't
lose any money if isn't available, then we obviously won't be willing to
spend big money to restore it.  There is a happy medium someplace.

All things within reason is my advice.

Kelly Lipp
Storage Solutions Specialists, Inc.
lipp AT storsol DOT com
www.storsol.com
(719) 531-5926