Steve -

Thanks for the reply. It's reassuring that you're not suggesting we should necessarily be looking at more elaborate partitioning schemes than that. There is one main database, and...

Rajesh -

I don't think I understand - my script showed the two cases -

- Create a table from scratch, *don't* insert a new record, execute a DELETE for...

Oh I see - that's pretty ugly - frist reaction was it was a bug, but it seems to be documented behaviour from BOL:

The current identity value is set to...

Your original answer was the posted immediately *before* the post raising the error message...

I guess all the "WHERE" keywords in the LEFT JOINs should be AND:

LEFT JOIN org_lvl4 l4 on l4.Lvl4 = o.NodeName AND l4.Org_Lvl4_ID = @l4

LEFT JOIN org_lvl3 l3 on l3.Lvl3 =...

Isn't the value you specify for the RESEED notionally the current existing value - so you just need to reseed to zero? (i.e. it increments before it uses it, rather...

This article describes getting a list of DTS jobs, stored in each of the three possibilities of SQL Server, the repository, and in files:

http://support.microsoft.com/kb/241249

I guess in general, there might be...

I think 2039 is the correct build number for SP4. I installed it recently, and the install itself seemed to go okay, both on the test server, and on the...

That's an interesting idea, that it might having something to do with the query plan, and maybe it gets implicitly recompiled when using QA for some reason? I guess next...

Not sure what you mean by "local or remote" - they're not running on the actual SQL server box, but they're all on machine with low latency network links via...

> So I would have to disagree that the problem is just ADO.

Let me amplify on the evidence for this. Another instance of this same type of issue was the...

Implemented from the formula for the sample correlation coefficient on

http://en.wikipedia.org/wiki/Correlation (I'm assuming you want the linear

correlation coefficient).

Table called "data", columns "x" and "y":

select (count(*) * sum(x...

Another approach that might be more suitable in some circumstances is to use SET IDENTITY_INSERT <table> ON.

You would then need to copy the record with the new ID, and delete...

Something like:

SELECT

  CASE WHEN [Employee ID] = (SELECT MIN([Employee ID]) FROM emp e

                                  WHERE e.[Department ID] = emp.[Department ID])

  THEN CAST([Department ID] AS varchar(10))

  ELSE '' 

  END AS [Department],

 ...