BUG: Arithmetic Overflow When Convert Real Value

• Microsoft SQL Server versions 4.2 and 4.2a for OS/2

SYMPTOMS

Using the CONVERT command to convert a valid real value from a column which allows nulls to a varchar or char large enough to hold the value causes arithmetic overflow or returns of zero.

For example, the following script :

   create table t1
   (col1 float NULL)
   go
   insert t1
   values (4.5)
   go
   select convert( varchar(10), col1)
   from t1
   go

   Msg 232, Level 16, State 2:
   Arithmetic overflow error for type varchar, value = 0.000000
   Arithmetic overflow occurred.

NOTE: This is not a problem when the column does not allow nulls.

CAUSE

SQL Server incorrectly handles the conversion of a real column which allows nulls to varchar or char.

WORKAROUND

Convert the real value to float or real before converting to varchar or char datatypes.

For the above example, use the following query to generate the correct result:

   select convert( varchar(10), convert(float, col1))
   from t1

STATUS

Microsoft has confirmed this to be a problem in SQL Server version 4.2 for OS/2. We are researching this problem and will post new information here in the Microsoft Knowledge Base as it becomes available.