Evelyn,

In addition to what Dana sad, make sure that tables have indexes on all fields used to join tables: checkno, agycode, brcode and amount. See Using Rushmore to Speed Data Access in the help for details.


>hi!
>
>any help on how i can speed up my "matching procedure"? it usually takes 30 mins to finish the procedure. here's my SQL:
>
>SELECT * FROM xa a WHERE EXISTS (SELECT checkno FROM xm b WHERE a.checkno=b.checkno .AND. b.agycode=a.agycode .AND. b.brcode=a.brcode .AND. a.amount=b.amount) INTO TABLE M
>
>SELECT * FROM xa a WHERE EXISTS (SELECT checkno FROM xm b WHERE a.checkno=b.checkno .AND. b.agycode=a.agycode .AND. b.brcode=a.brcode .AND. a.amount!=b.amount) INTO TABLE Q
>
>
>for each table i have at least 20,000 records. and the execution of that SQL takes the least of 30 mins. you bet it is really slow!
>
>i need to have both the matched record in table XA and XM into table M. meaning if the query found a match, it will copy the matched record from XA and XM to M. both tables have the same structure so if i execute advancepack's SQL, it will give me fields like checkno_a, checkno_b.. etc. all in one record.
>
>i have indexes for both tables.
>
>
>thanks