Question

I have a fairly horrible query using 2 “IN” statements that needs running on our database. Firstly the schema (Simplified for this example):

CREATE TABLE [dbo].[SystemUser]
(
    [SystemUserID] [int] IDENTITY(1,1) NOT NULL,
    [FirstName] [nvarchar](50) NULL,
    [Surname] [nvarchar](50) NULL
    CONSTRAINT [PK_ApplicationUser] PRIMARY KEY CLUSTERED 
    (
        [SystemUserID] ASC
    )
)
GO

CREATE TABLE [dbo].[Group]
(
    [GroupID] [int] IDENTITY(1,1) NOT NULL,
    [Name] [nvarchar](50) NULL
    CONSTRAINT [PK_Group] PRIMARY KEY CLUSTERED 
    (
        [GroupID] ASC
    )
)
GO

CREATE TABLE [dbo].[GroupMembership]
(
    [SystemUserID] [int] NOT NULL,
    [GroupID] [int] NOT NULL
    CONSTRAINT [PK_GroupMembership] PRIMARY KEY CLUSTERED 
    (
        [SystemUserID] ASC,
        [GroupID] ASC
    )
)
GO

What I want to do is find all “SystemUser” records that match a list of SystemUserIDs that do NOT have membership to a “Group” that is in a list of GroupIDs.

So 2 seperate lists of IDs beng compared in one query. The fastest way I can think of doing this currently is below:

SELECT SU.SystemUserID
FROM [dbo].[SystemUser] SU 
LEFT JOIN
(
    SELECT GM.SystemUserID
    FROM [dbo].[GroupMembership] GM
    WHERE GM.GroupID IN
    (
        1, 7, 8, 10, 32
    )
) GM ON GM.SystemUserID = SU.SystemUserID
WHERE SU.SystemUserID IN
(
    10, 61, 80, 93, 98
)
AND GM.SystemUserID IS NULL /* Not matched */

Is there anything I’m missing; would a “WHERE NOT EXISTS” check be more efficient? Or can you think of a better way of processing and filtering by the two lists?

Answer 1

Assuming SQL Server 2005 or later,

SELECT SU.SystemUserID
FROM [dbo].[SystemUser] SU 
WHERE SU.SystemUserID IN
(
    10, 61, 80, 93, 98
)
EXCEPT
SELECT GM.SystemUserID
FROM [dbo].[GroupMembership] GM
WHERE GM.GroupID IN
(
    1, 7, 8, 10, 32
)