Question

I am trying to delete some data from a table variable using a SELECT query. The following code works perfectly:

DECLARE @sgID int
SET @sgID = 1234

DELETE FROM
    @tbl_users 
WHERE
    (userID NOT IN (
            SELECT
                userID 
            FROM
                [SGTable] 
            WHERE
                (sgID = @sgID)
            ))

I’m trying to speed up this query, and read that the following approach may be better. However, when I use the following code – ALL records are deleted from the table variable.

DELETE
    tmp 
FROM 
    @tbl_Users tmp INNER JOIN
    [SGTable] sgu ON sgu.userID = tmp.userID 
WHERE
    (sgu.sgID <> @sgID)

I (obviously incorrectly) assumed that these two queries did the same thing (delete all userIDs in the table variable where the userID is not found in the sub-query). Can anyone please offer some advice on making the second query work, as it is obviously easier to read and maintain?

Answer 1

Is it possible that a userID can be associated with more than one sgID in SGTable?

If so, then you’re deleting users in your sgID (@sgID) because they are also associated with another sgID.

You might prefer:

DELETE tmp 
FROM @tbl_Users tmp 
LEFT OUTER JOIN [SGTable] sgu ON sgu.userID = tmp.userID AND sgu.sgID = @sgID
WHERE sgu.sgID IS NULL