Can you post the actual execution plan together will both table definition including any index definition if they exist? (I'd expect no index on the tables since you mention a...

I would either use a varchar(8000) variable and a comma delimited list (or any delimiter that will never ever be part of the parameter) and split it in the sproc...

😀

... until you test both versions against a larger XML file...

Your preference may change 😉

Duplicate post. No replies please.

Original post can be found here

Hoola-HIPAA-Hippo as requested 🙂

Use your original query????

You could use a cte with ROW_NUMBER() OVER (PARTITION BY Case_Id ORDER BY DateTime DESC) and then query for row=1.

phood (3/3/2011)

---

Based on your forum etiquette I did a hack job on this post so I deleted what was in here and will post again in a bit when I...

Since I don't know where the blank is coming from (presumably because Field1 is defined as CHAR(5), but it's also possible that the value of field2 starts with a blank)...

You could modify your current script to return the query or any other information before the actual statement.

Something like

SELECT 'query: SELECT * FROM sys.columns'

SELECT * FROM sys.columns

would you please provide the sample data you'd like to concatenate?

Something like

SELECT 'abc','def'

Also, please include your current approach.

Please provide table definition and ready to use sample data as described in the first link in my signature. Also, please provide your expected result based on those samples together...

The issue is the

CONVERT(VARCHAR,DATEADD(DAY,2, @fec),112),

part of your GROUP BY statement.

Since there is no column reference, it'll throw the error you see.

You should get the expected result set result by either...

Craig Farrell (3/2/2011)

---

...

By Jove! You've forgotten about taking care of your SOX while you're at it! In dealing with a rainforest, you'll catch some nasty things...