Partitioning might help in this situation. What I'd really recommend, as a full solution, is hire someone (a contractor) to come in, look at the database, run some performance...

That code will work. It's the second solution I mentioned in my prior post. The problem with it is that it gets slower and slower as the table...

Well, you can copy-and-paste from Management Studio to Excel, if that helps.

Otherwise, SSIS is really good at exporting query results to Excel.

Am I mistaken, or did I answer this same post in another forum?

Take a look at DDL triggers in Books Online. They should be able to do what you want.

Check the two execution plans. That's where the difference will be.

I use the Master Proc-Sub Procs idea a lot. Makes for more modular code, much easier to document and maintain.

Are the two tables supposed to be exact duplicates? (If so, be sure you really need two tables.)

If so, then you might just truncate table B and insert everything...

You're welcome.

Not sure what you mean by starting over from 1 for each property. Do you mean you have multiple tables that all need this type of handling? If...

I would definitely NOT require cursors. Could be done as a pretty standard "upsert" (update/insert). Since it's conditional on the insert into the other table, it can either...

Not sure from just this little bit, but a place to start is that it will never get to F if the test for P comes up true. Case...

Can't you just have the proc return a recordset/output param with the XML data in it? Wouldn't that be easier?

Declare a varchar variable, assign the query to that, then use that in the parameter. It doesn't like building the query string in the parameter.

Is it possible for you to add an identity column to the table, without breaking the rest of the code?

If so, then an insert trigger could update the column based...