I'd recommend breaking the query down into smaller pieces and finding which one gives you the error. Just run one column at a time, comment out the rest, till...

Multiple identical indexes have no advantages. No speed increase on selects, definite speed decrease on insert/update/delete, and it takes up more disk space too.

On the index hint thing, I...

Jeff Moden (4/7/2009)

---

GSquared (4/7/2009)

---

Jeff, the problem with the convert(int) method is that it, again, relies on the internal storage of datetime, just like the version for float.

Have you tested it...

If they don't have separate versions of the databases, then I guess they'll have to continue to query the production databases in order to build/test anything.

That's a really bad idea,...

You're welcome. Glad I could help.

Jeff, the problem with the convert(int) method is that it, again, relies on the internal storage of datetime, just like the version for float.

Have you tested it in SQL 2008...

At the top, did you change where I had "getdate()" to the date you actually want to run? Looks like 2 April for your posted query.

Try this. I can't test it (don't have the tables), so tell me if it doesn't work.

SELECT

daily.Fund,

daily.salesamount AS [Daily Sales Amount],

...

I think this is a double-post. Please don't do that.

Another possibility that occurs to me, from a bad prior experience, is that the account SQL Server is running under might end up locked out because of failed login attempts....

I can't test this, so I'm not sure if it'll work, much less whether it will be faster, slower, etc., than the current version. Can you test it out,...

There are a couple of ways you could manage that.

One would certainly be an on-insert trigger. Another would be to handle it in the insert proc. More likely,...

Have you checked logs (error log, server log) to see if there's something that happens right before that?

Certain error severities can cause SQL Server to "lock up". Take a...

mike brockington (4/7/2009)

---

While some languages are so terse that they look like a hex dump, SQL is actually one of the more verbose, yet it still manages to be one...

Instead of doing a Union between those two queries, do a join. Join them on the column that has the Fund1, Fund2, etc., in it.

Does that make sense?