It looks like a single employee can have multiple records.

1 virginia usa

1 texas usa

1 new york usa

Can an employee have more than one record in the same state?

If you can...

Can you join and compare?

SELECT

emp#

FROM Production P

INNER JOIN Staging S

ON P.emp# = S.emp#

WHERE P.Address <> S.Address OR P.country<> S.country

Once you have the identified the changed employess you...

Thanks Dan. The relevant piece of information provided here is that I will not be able create a SSIS package in VS 2008 and then publish this to my my...

You would use an update statement to set the fields to NULL

UPDATE YourTABLE

SET YourField = NULL

WHERE YourField = 'unassigned'

Don't forget your WHERE statement!

Then you can use COALESCE to return 'unassigned'...

My preferred method for this is to allow the users to insert a request record in a table.

Then I run a job continuously, say every five minutes. The job looks...

This is powerful stuff. I still haven't fully committed to going down this rabbit hole.

I like the idea about splitting the numeric and character values.

The attribute 2 or subdatatype as...

Actually the SessionID would probably just be useful for tracking the insertion time of the data so I might have driven off the track a bit there.

My model is largely...

http://www.sqlservercentral.com/articles/Database+Design/62386/

Here the article I have been looking for.

>>If you can avoid this I would. There are times things like this are necessary, but you would

typically prefer to avoid them.

How would I avoid them? Could I have...

Yes I did mix up my descriptions as you noticed. Your terms make a lot more sense.

You concerns are definitely valid. How much data? A thousand rows of current...

You could try to import each file into a one column staging table. Then write a select against the table to exclude rows, you will probably need to use LIKE....

Are you replacing a process that is based on paper forms? If so, collect examples of all the existing forms.

Your expected results and ColdCoffee's are very similar.

Your results:

Day 1st shift 2nd Shift 3rd Shift

3/1 3 1 1

3/2 2 1 0

3/3 0 0 0

Cold Coffee's:

DayNumberFirstShiftEventCountMiddleShiftEventCountLateShiftEventCount

Day 1360

Day 2360

Day 3361

Why don't you...

Actually the late, middle and early sample data was not from the original poster. I was just trying to use my intuitive powers to provide some code that might be...

All depends on the consistency of your data.

I would import into a one coumn statging table.

Then update the data. This is really ugly, but it works.

CREATE TABLE TEST

(

String ...