CheHax scribbled on 3/31/05 2:26 AM:

> Records in my table can be filtered by regional code, and I suggested
> we could split the table into several smaller ones, but this solution
> wasn't accepted. What do you think of it ?
> 

from the swish-e.org home page:

"Swish-e is ideally suited for collections of a million documents or smaller."

You could probably split up your millions of records into smaller subsets and 
create an index for each subset, then search all the indexes (or groups as 
needed). This would not require splitting your table; just a subset of rows per 
index. I think folks on this list have done M+ records fairly successfully.

-- 
Peter Karman  .  http://peknet.com/  .  peter(at)not-real.peknet.com