>Is it true about *SM hiting a dog-leg in the performance curve after the
>data base has exceeded 30 gigs????

Joe - I recall the question of database size coming up maybe a year and
      a half ago, and that multiple customers responded that they have
databases larger than 30 GB and were doing okay.  I think the size you
can support depended a lot upon configuration: fast disk, abundant
memory for caching/buffering, a fast CPU, and spreading over multiple
volumes to benefit from multiple threads.
    Much of the issue boils down to how much one is willing to spend on
various types of solutions.  You pay a premium for the latest hardware,
versus perhaps splitting into multiple servers over older, less
expensive hardware as one approach.  The legacy single-threaded approach
of various server processes of course does not help the situation.
Having an Expire Inventory run all day is awfully painful.

  Richard Sims, BU