Hi all

I have a table with 98,000 records and my user need to input a name of a part of the name and retrieve all the similars to this name.

I'm using Select * from data!table where "%input_name%" like table!name, i found that the where like command is not optimizable, then someone suggest my phdbase, this is a lot of money foor me and I think that is not around.

Ok I found a document in MSDN that talk about this problem:

I try to set SYS(3050,1) to 1,000,000 and SYS(3050,2) to 10,000,000 and the results were good:

for example: (column (a) No changes to SYS(3050) (b) changed sys(3050))

Running the query: the time for get the results:

1 a) 41.84 Sec b) 15.75 Sec
2 a) 93.42 Sec b) 17.23 Sec
3 a) 70.59 Sec b) 13.81 Sec
4 a) 72.87 Sec b) 15.67 Sec
5 a) 72.02 Sec b) 12.12 Sec

6,7,8 and next runs: a) 60.00 Sec b) 10 - 12 Sec.

Notes:

If i use SET OPTIMIZE ON or OFF the results never change.

When I don't change the SYS(3050) values and I minimize my application and then restore it the answer time is to slow, 1 minute for the application get the focus.

And if I change the values of SYS(3050) the minimize/restore time for the application is inmediatly.



PLEASE SOME COMMENTS ABOUT THIS,
CHANGING THE VALUES OF THE SYS(3050) WILL AFFECT MY APPLICATION ?
HOW COULD I DETECT OR DETERMINE THE BEST VALUES FOR THE MEMORY SETTINGS ?

I WILL APRECIATE ANY COMMENT, AND SORRY FOR MY ENGLISH

THANS AND REGARDS

GB