big table or several littles tables ?

[olivier]

Valentina 1.90 - RB 5.5.3
cache 20 Mo.

Hi Ruslan and list,

I have a big db with a table of 800 Mb consisted of street names, and a 
table was joined of 45 Mb for zip and cities.

I have the possibility of cutting the table of streets in 5 or 10. (The 
purpose for the customer is to look for street names thus always in a 
particular zip / city).

To accelerate the searches and optimize the cache memory, it is better 
to keep a big table of streets or to have several of it?

I think naturally that it is better there to have ten of 80 Mb than one 
of 800 Mb but I would like to be sure of it.

The table of streets will be only in mode reading.
Thus only the speed of sorting / search is here important.
If we have to make 1000 searches very quickly on street names, it is 
faster to use an only big table or to juggle between 10 small different 
tables?

Thank you very much for your help

olivier