Could anyone give me some feedback on the following situation :

Suppose you need to create a process that makes invoices. But you have
to check if there still some stock left, and you also have to check if
the articles need export licences and then if there still are import
licences for that type of article and than if there's still enough quota
on those licences.
If you use quota of any of those licences, you have to fill up a
database with the amounts per licence.

My question is : When an invoice can contain up to thousands of records
to process, wouldn't it be an enormous task for a database server to
complete ?  I'm new to client-server computing.  Nowadays this runs in a
CLIPPER environment and it's pretty fast.  Normally in a C/S environment
you would expect things to go even faster, but I have my doubts about

In Clipper you can jump directly to a certain record, but if a database
server has to process SQL statements every time a specific record is
needed, I think the process will slow down.

What do YOU think about this ?