Hello...
Im having a hard time finetuning my queries. Im using SQL2000 Enterprise
with SP3.
I've a table with approximately 500.000.000 rows in and for some strange
reason I get different queryplans according to similar where clauses, take a
look at this:

The Table1 has 2 indexes:
nonclustered index idx1 at founddate
nonclustered index idx2 at founddate, traveldate, travelagent

select founddate, traveldate, costprice, fareclass, travelagent
from Table1
where founddate = 2500
and traveldate = 2600
and travelagent in('xxx', 'yyy')

Then the query optimizer chooses the right index = idx2 using index seek

But when I change the query to:
select founddate, traveldate, costprice, fareclass, travelagent
from Table1
where founddate = 2500
and traveldate = 2600
and travelagent in('xxx')

It suddenly chooses idx1 with index seek and them uses a scan for the rest
of the values.

If I then just change the last where clause to and travelagent in('xxx',
'xxx') and then jumps back and uses idx2 again...

Any suggestions?

Clustered Index Not Working
Select distinct problems
SELECT Question: Get each date when I have start und end date
DESPERATE - Need help getting a SQL stored procedure output parameter to my .NET app
Cast/Convert Query help