I'm sure we can figure something out.  What is it you need to do exactly?

Make a primary key with an autonumber column.

Then add a calculated field to show the formated invoice number :

Formula = RIGHT("000000" + CSTR(PK), 7)

This allows for 9 999 999...

Doesn't matter at all... even if he had 10 duplicates per key the update should still work :

DECLARE @A Table (PK INT NOT NULL, Col INT NOT NULL)

INSERT INTO...

I wouldn't expect it to be anything else.  You can't be that blind can ya .

Is there a trigger on that table?

Or a DDL trigger?

Data only if I remember correctly.

Someone must frame this one and keep it posted on a board somewhere... That's just too good .

Here...

Got anything more usefull to offer Collin?

This should get you started

'ALTER DATABASE [' + @Dbname+ '] MODIFY FILE (NAME = [' + @LogLogicalName + '], SIZE = ' + CONVERT(VARCHAR(20), @NewLOGSize) + 'MB)'

You get the Logical...

And for the final blow :  Which one executes the fastest??

You got me there.   I don't see it anywhere in the dts itself... Saying that I didn't scan the system tables.  Maybe it would be faster to just ask access...

Simple.  No end user is using this so performance is not absolutely required in this case.

Here 10 seconds and I'm done.

I might also remind you that almost no DBA will try...

Right click on the new package, schedule, OK.

Then open the newly created job, cut / paste the dtsrun line from the new job to the current one. Save and test.

I really like the second idea, but then again you still could have more tables than the number you taught you'd need... so you'd really be screwed on that one.

Ex...

I can already answer this one for you.  I had been using that method for a while untill I saw the exponential part in the execution plan.  So that's when...