FIX: A SELECT query unexpectedly returns an incorrect number of rows in SQL Server 2005 Mobile Edition and in SQL Server 2005 Compact Edition

View products that this article applies to.

SYMPTOMS

When you run a SELECT query in Microsoft SQL Server 2005 Mobile Edition and in Microsoft SQL Server 2005 Compact Edition, the query unexpectedly returns an incorrect number of rows. This problem occurs if the following conditions are true:

The SELECT query contains an inner join.
One of the two tables that are joined has an index on the joined column. The other table has no index on any of its columns.

For example, you run the following query in SQL Server 2005 Mobile Edition:

select * from T1, T2 where T1.Col = T2.Col1 order by T1.Col

Note In this example, table T1 has an index on column Col. Table T2 does not have an index on column Col1.

Back to the top

CAUSE

This problem occurs because the query optimizer does not discard the previous plan completely. The query optimizer considers multiple query plans before the query optimizer decides on the best plan to execute a query. In some cases, the query optimizer can find a plan that uses an index to evaluate the condition. However, the query optimizer may later find a better plan. In this case, the query optimizer discards the previous plan to use the better plan.

Back to the top

RESOLUTION

To resolve this problem, use one of the following methods:

Create an index on the joined column on the second table.
Delete the index of the joined column on the first table.

Note In the example in the "Symptoms" section, the joined column on the second table is T2.Col1. The joined column on the first table is T1.Col .

Back to the top

STATUS

Microsoft has confirmed that this is a bug in the Microsoft products that are listed in the "Applies to" section.

Back to the top

MORE INFORMATION

Steps to reproduce the problem

Run the following statements against a database in SQL Server 2005 Mobile Edition:

CREATE TABLE T1 (COL INT);
GO;
CREATE INDEX T1_IDX ON T1 (COL);
GO;
INSERT INTO T1 VALUES (1);
INSERT INTO T1 VALUES (2);
INSERT INTO T1 VALUES (3);
GO;
CREATE TABLE T2 (COL1 INT, COL2 INT);
GO;
INSERT INTO T2 VALUES (1,1);
INSERT INTO T2 VALUES (2,2);
INSERT INTO T2 VALUES (3,3);
GO;

Run the following statement:
select * from T1, T2 where T1.Col = T2.Col1 order by T1.Col
You experience the problem that is described in the "Symptoms" section.

Back to the top