I'd select and group by the column from member base data, rather than from transactions, seeing as t.Buyer is allowed to be null due to the outer join.

What's the definition of the clustered index, and what are the data types of the columns that make it up?

What version of SQL is this? (run SELECT @@version)

If you read my post, I told you how it can be resolved.

Drop the clustered index on tblTime and recreate...

Your between statement is incomplete.

Where Coalesce(tblAuditMain.AuditDate,'') >= Coalesce(@BeginDate, Coalesce(tblAuditMain.AuditDate,''))

And Coalesce(tblAuditMain.AuditDate,'') <= Coalesce(@EndDate, Coalesce(tblAuditMain.AuditDate,''))

And Coalesce(tblAuditMain.AuditYear, '') = Coalesce(@AuditYear, Coalesce(tblAuditMain.AuditYear,''))

And tblAuditQuesLkup.AuditQuesType BETWEEN @AuditQuesTypeParameter -- and what?????

Group By tblAuditQuesLkup.AuditQuesNumber, tblAuditQuesLkup.AuditQuestion, tblAuditQuesType.AuditQuesAuditType,

tblAuditQuesLkup.AuditQuesType

Order By...

Please post table definitions, sample data and desired output. Read this to see the best way to post this to get quick responses.

http://www.sqlservercentral.com/articles/Best+Practices/61537/

What I've had to do before, when there were no x64 drivers, was to set up another server (x86), install SQL on there and use the x86 drivers. Not ideal,...

Is that the only error? If not, please post the results of this

DBCC CHECKDB (<Database Name>) WITH NO_INFOMSGS, ALL_ERRORMSGS

Drop the clustered index on tblTime and recreate it. That should fix...

In other words, you don't care which row you get back, you...

May I suggest, rather than trying...

If a column has only one value, then there's no use in filtering by that column, hence what's the point of indexing? Even so, a single-column index on a column...

Nice set of addresses. What are we supposed to do with them?

jcrawf02 (1/14/2010)

---

So Gail, something that I wonder occasionally (the rest of the time I wonder about everything else :-P) - if I SELECT TOP 10 and ORDER BY the clustered...

You colleague is closer to correct, but he's still wrong in one place.

Regardless of recovery model, SQL follows this method for reading/writing. (this has not changed fundamentally since SQL 2000)

Update/insert/delete....

Please post table definitions, sample data and desired output. Read this to see the best way to post this to get quick responses.

http://www.sqlservercentral.com/articles/Best+Practices/61537/

ifila (1/14/2010)

---

On the ProdDB Gail's solution would always take 8 seconds, regardless of the number of records i wanted returned.

TOP 60000 - 8 seconds all the way down...