yes, thank you. Not having to import '.' is good.

ok this is funny. Everyday I deal with at least one rabid pro-Microsoft nut who forgets that business see IT as nothing more...

weird.. my reply yesterday never posted.

Add the switch -r to your bcp command.

I have a header and footer text file and use dos copy to glue them together after export. The header also maps the schema and template.

Barbara

Had the same problem...

export the data in a data stream - ado or bcp work.

I had a similar type of problem. I was getting data from a large supplier that had inconsistent data layouts, despite requests otherwise. Sometimes they would even add columns in...

is this a one time deal? You could make this dynamic if not. I think this is what you want:

Insert sqltable (D_pKey,D_code1, D_code2)

select P_Key,

     CodeA1, CodeA2

     From exceltable

     where CodeA1...

My exp is often that the functions are slightly slower but usually not enough to supercede the benefits over views...

I have found functions to be a really good choice when...

This is a column header:

XML_F52E2B61-18A1-11d1-B105-00805F49916B

You can turn off headings in qa or dts export. SET NOCOUNT ON will take care of the rowcount. Neither should effect an export..but

The problem i...

good tips, thank you.

I found i had to order it by something in the second union to get them "nested".

XML mapping. Export the data as xml (for xml auto will work). The template would be an excel wkbk with xml mapping. Neither portion is hard to implement. (must be...

---

Since you're just getting the year, would the year() function perform any better? Probably a wash, but maybe worth trying in addition to other suggestions.

---

Year() simply calls Datepart - so...

I dont know if this helps to know but it is likely this that is slowing down your qry:

order by datepart(yyyy, start_date) desc

Theoretically, if the clustered index is on date...

Thanks for the great insights!

I always think the analyst role is more related to understanding the why's and hows where the programmer role is more involed with process. I think...

I believe the question to ask is where the workload is and plan according to that instead. I am assuming this is one server - so what is the filegroup...