HELP (URGENT) Need Information!

Hi,

I administering SQL server and one of or Db (size ~7 Gb) have a table with size 396 MB and 25 indexes created on this one. We working with this DB using Siebel Call center.
I have met a very strange situation: when I tried to get some data (using Siebel) my request take ~5 min. I run Profiler and catched this T-SQL statement which Siebel send to SQL server. After that, I run this T-SQL query using Query Analyser and found that Query Optimiser used wrong index. (clustered index) and try to compare every 470436 rows from the tables.
This is the statement. @P1, @P2, and @P3 just variable which siebel transfer to SQL server.

SELECT ........
FROM
dbo.S_EVT_ACT T1
LEFT OUTER JOIN dbo.S_CONTACT T2 ON T1.TARGET_PER_ID = T2.ROW_ID
WHERE
((T1.APPT_REPT_FLG != @P1 OR T1.APPT_REPT_FLG IS NULL) AND
(T1.TEMPLATE_FLG != @P2 OR T1.TEMPLATE_FLG IS NULL)) AND
(T1.SRA_DEFECT_ID = @P3)

But I found the very strange situation - when I manualy put the value '1-FRK' instead of @P3 - Query Optimizer used RIGHT index and I had a very good performance.


And addition when I tried to run the first Query on the same table but with the only 1000 rows the Query Optimiser choose right index.


Wat's on, wat happend with My Query Optimizer? Wy his decision so different from time to time?


Thank's a lot for any information about how to make Optimiser obedient.