By virtue of using the MERGE statement you have already specified the TARGET table, so no need to quality the INSERT, UPDATE or DELETE:

INSERT ([Name]

...

I'm not sure if I'm following what you want to do, but can't you simply delete the rows?

DELETE CityStates

WHERE StateID = 1

OR PartneID = 1

It was just an example of other ways to get rangom rows.

I'm not sure I understand waht you are tryign to do, but myabe this will help? SELECT

...

WHERE

1 = CASE

...

I'm not sure I fully understand, but maybe this will get you going:SELECT

*

FROM

(

SELECT *

...

I found the issue. The filter for CHARINDEX('=', Item) > 0 was being evaluated after the SUBSTRING. I changed the AttribValues CTE to this and it works:,AttribValues

AS

(

...

This doesn't seem to have any affect on the sampel data you provided but the NULL check and the LENGH seem to be out of order:,N(N)

AS

(

...

I know there are some articles out there (sorry I don't have a link), but the conclusion is that "it depends." I don't know of any hard and fast rules...

A little late to the part and it sounds like you might have it coverd. But, do the statistis show the correct row count? I've seen them get out of...

I don't know about your indexs, but applying a function to a column makes it so that predicate is not SARGable.

You might try chaning:

AND left(c.comment_code, 3) = '220'

to something like:

AND...

I realize this doesn't actually help provide a solution to your actual question, but I've been through this before, albeit non-medical. And, trust me, there is no valid solution unless...

I'm not sure if this is the case but according to Microsoft, there is no guarentee that nondeterministic scalar functions will executed for every row.

I submitted a "bug" a while...

SkyBox (5/25/2011)

---

I work with an unruly developer that always runs DTA against our production databases during prod hours. I have asked him several times not to do this and...

EdVassie (5/23/2011)

---

Row versioning and version store are different things.

Row Versioning is what happens to rows. Every change to a row changes the row version Id.

Row versioning has...

Assuming all parsed values are INTs, it's pretty easy:

;with cte(vers)

as

(

select '10.00.1818'

union all

select '10.00.1823'

union all

select '9.00.4315'

union all

select '9.00.4317'

union all

select '10.50.1600.1'

union all

select '10.50.1600'

union all

select '10.50.1600.2'

union all

select '-10.50.1600.2')

select vers

from cte

cross apply(select REVERSE(vers)) as...