Posted by Erland Sommarskog on 04/14/06 01:01

traceable1 (tracykc@gmail.com) writes:
> I have 2 SQL databases which are the same and are giving me different
> query plans.
>...
> Both boxes have the same configuration, the only difference is that one
> of them is a cluster.

So they have the same number of CPUs?

> I have run statistics on both, and the cluster is still creating a
> bitmap and running some parallelism which the other box is not.
> Also, the the first step, the A1 box estimates the rows returned to be
> around 80K and the actual rows returned is about 40K - subtree cost =
> 248. The Acluster box estimates 400K - subtree cost=533!
> After running statistics, how can it be so off?

You could try running UPDATE STATISTICS WITH FULLSCAN on the involved
tables, to be really sure that you have factored that part out.

Also, try adding OPTION (MAXDOP 1) on the cluster. Parallelism is
sometimes good, but sometimes it's bad...



--
Erland Sommarskog, SQL Server MVP, esquel@sommarskog.se

Books Online for SQL Server 2005 at
http://www.microsoft.com/technet/prodtechnol/sql/2005/downloads/books.mspx
Books Online for SQL Server 2000 at
http://www.microsoft.com/sql/prodinfo/previousversions/books.mspx

[Back to original message]