First , make sure you are not vulnerable to SQL Injection. It sounds very likely you are. Get your supplier to sort that ASAP. To fix that...

Date data is NOT stored in a particular format.

Please see this link http://www.karaszi.com/SQLServer/info_datetime.asp

Please bear in mind that we have no knowledge of your data or the relationships within it , so to be presented with massive execution plans are be able...

@Gail , thanks for clarifying. didnt seem right ...

@sam-3 , the most helpful thing you could do is slowly strip the query down. Once it is at the simplest...

Use of scalar UDF's is not a good idea either

http://sqlblogcasts.com/blogs/sqlandthelike/archive/2009/10/15/udf-overhead-a-simple-example.aspx

GilaMonster (6/28/2010)

That's a correct estimate.

7 rows (estimated) * 36876 (number of executions) = 258132...

Sure about that Gail ?

Difference in 2005 / 2008 Perhaps ?

This profiler trace screen shot is 2005.

Doesnt Look like the update statistics has invalidated the plan.

You sure about the stats ?

The table STOREROOM has an estimated row count of 7 but actual of 250,000+.

Update the statistics and issue sp_recompile "STOREROOM", to force the plan out...

I know *we* are on the same page here. Personally i dislike the word physical as that would imply that on the hardware, page 2 would follow page 1...

Please post execution plan and DDL as per http://www.sqlservercentral.com/articles/SQLServerCentral/66909/

Hi ,

The consensus of opinion is that newid() provides a pretty good randomness factor.

So if you select top(100) * from <yourtable> order by newid() , you will get 100 random...

Logically , i think you mean Jeff 😉

Can you show us the code you have tried so far , and where you are having difficulties?

Quick , get...

At least it didnt all depend on a dodgy decision. Outclassed in pretty much every dept.