At 20:09 +0300 9/14/02, Savaidis wrote:
>I have about 9.000 records of companies I want to put in a database . I want
>to put it in a MySQL sever on the net that I'm not sure it is quite fast (It
>is a very cheap one :).
9,000 rows is nothing. In any case, you could easily test this out
by just trying it.
>So I'm thinking if it is a good idea or a stupid one to split it in 10
>tables as are the geographical areas (the search will be on geografical
>basis) to speed up the searches.
>But I believe tables are all placed almost as one inside the database,
>separated by some kind of delimiters, so there is no difference actually if
>I'll do this. They use the same buffers, same login etc.
>Does anyone knows these things better?
>The only benefit will be to get easier (smaller) backups from the data
>dumps? Or will be worst consuming more resources? Should I check something
>on server configuration?
>The 10 tables will be -more or less- equal but with no equal load. The
>maximum/minimum load to tables could be 10-20/1.
>Please is important to me.