NilsBevaart
Programmer
My application, connected to a SQL2000 database returns time-outs. On analysis of the query's, I found that the execution plan is far from optimal. The estimated rowcount was 10 and the actual rowcount was 15000.
My auto-update-statistics option is on. I found that the last update of the statistics was at the time I started a manual index defragmantation.
I manualy updated my statistics. This helped in determining a better plan (using 2 indexes), but still my estimated rowcount is far off. The estimated rowcount is 500 instead of the 15000.
I updated the statistics for this index again, using a fullscan, but that still doesn't help.
How can this happen? What can I do to solve the problem for now, and prevent this from happening again?
Nils Bevaart
My auto-update-statistics option is on. I found that the last update of the statistics was at the time I started a manual index defragmantation.
I manualy updated my statistics. This helped in determining a better plan (using 2 indexes), but still my estimated rowcount is far off. The estimated rowcount is 500 instead of the 15000.
I updated the statistics for this index again, using a fullscan, but that still doesn't help.
How can this happen? What can I do to solve the problem for now, and prevent this from happening again?
Nils Bevaart