Getting a distinct value across 2 union sql server tables...

Question

I know this is a silly question, but Im a newb

I'm trying to get all distinct values across 2 tables using a union... The idea is to get a count of all unique values in the columnA column without repeats so that I can get a summation of all columns that contain a unique columnA... sounds simple eh?

THis is what I tried (sql server express 2008)

select count(Distinct ColumnA) from ( select Distinct ColumnA as ColumnA from tableX where x = y union select Distinct ColumnA as ColumnA from tableY where y=z )

Answer 1

SELECT COUNT(distinct tmp.ColumnA) FROM ( (SELECT ColumnA FROM TableX WHERE x=y) 
UNION (SELECT ColumnA FROM TableY WHERE y=z) ) as tmp

The extra distincts on TableX and TableY aren't necessary; they'll get stripped in the tmp.ColumnA clause. Declaring a temporary table should eliminate the ambiguity that might've prevented your query from executing.

Answer 2

SELECT COUNT(*)
FROM
(
SELECT DISTINCT ColumnA From TableX WHERE x = y
UNION
SELECT DISTINCT ColumnA From TableY WHERE y = z
) t

Using a "UNION" will not return duplicates. If you used "UNION ALL" then duplicate ColumnA values from each table WOULD be return.