multiple column histograms in Yukon

Hi,
I have a simple query (sql 2000):
select * from table where feld1 = 'X' and feld2 <> 'Y'
Both fields are not very selective, but together they are
very selective and will find only 4 rows out of 1700000
rows. I have an multiple column index on both fields but
the optimizer chooses always a table scan. I heard Yukon
will handle such situations much better (multiple column
histograms) but I tried the version I got from PDC and
YUKON is still choosing a table scan too.
Any hints?
Andreas
Austria> the optimizer chooses always a table scan. I heard Yukon
> will handle such situations much better (multiple column
> histograms) but I tried the version I got from PDC and
> YUKON is still choosing a table scan too.
I wouldn't try to judge the delivery of all of Yukon's enhancements based
solely on the PDC build. I'm sure many of the optimizations are
forthcoming.
--
Aaron Bertrand
SQL Server MVP
http://www.aspfaq.com/