You're working with 200k lines, it won't take 1ms to run the query .

can you send the code of the udf to see...

Dim MyCmd As ADODB.Command

Dim MyParam As ADODB.Parameter

MyCmd.Execute exec_sp_biginsert

Same problem... it seems like a flawed design. It can be done with a self join.

You'll have ignore those types when you generate the query. And no, there's not a better query to get those results.

You can't seriously tell me you read the article :

Recompiles due to interleaving DDL and DML operations

Stored procedures will recompile is that the developer has place interleaving Data Definition...

Give yourself a chance... 2 days at most . 2 hours if you're smart .

Are you sure you read the code .

You should always qualify all objects... this is just one of the problems you'll meet .

Why are you executing a sp instead of...

Are you talking about sysdepends error/comment?

If so you can just ignore it, assuming you retest every proc to make sure it works fine

There's just no point in checking for error in a select, it'll either be a fatal error that you can't catch or something like division by 0 error which shouldn't...

Convert the udf into a single master query and it'll solve your performance problem. It'll still be somewhat slow but it will take only a few minutes at worst.

On top of my head :

IF Object_id('fnSplit_Set') > 0

DROP FUNCTION dbo.fnSplit_Set

GO

IF Object_id('Numbers') > 0

DROP TABLE dbo.Numbers

GO

CREATE TABLE dbo.Numbers (PkNumber int identity(1,1),...

HTH.

How about a openquery/openrowset article from noeld (sorry but he just nows too much to pass that up).

Did you read this???

Optimizing Stored Procedures To Avoid Recompiles

if you use a temp table, there's gonna be a recompile, there's just no way around that.