Hello

We are looking into incorporating FB into a large scale application,
which currently stores data in flat files. It serves multiple
customers (up to 500) on a single Solaris SPARC machine, by spawning a
seperate process for each customers. Each of these processes will be a
single FB client, all which will connect to the same FB server (which
will run on the same machine).

Which approach is better in terms of performance: using one database
file for all of the customers, or having 500 different databases,
served by a single FB super-server process?

Thanks.