Archives of the TeradataForum

Message Posted: Tue, 11 Jul 2006 @ 16:42:36 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

<-- Anonymously Posted: Tuesday, July 11, 2006 12:08 -->

David,

Your join index has COU_ID and ACC_ID from both tables and your primary index definition does not tell the system which set of the columns would you be using to distribute the join index result.

If this is a select statement with order by of PI columns, you would get a error message for that. Could that be your problem with the JI?

Also you have join a not null column to a null able column. That might bring in problem later on since you can perform comparison with null values.

     DEVNAD_T.NAD
           ACC_ID CHAR(9) CHARACTER SET LATIN NOT CASESPECIFIC COMPRESS '  ',

     DEVNAD_T."ACCOUNT"
           ACC_ID CHAR(9) CHARACTER SET LATIN CASESPECIFIC NOT NULL,

     CREATE JOIN INDEX DEVNAD_T.ACCOUNT_JI2,
          FALLBACK AS
          SELECT  a1.COU_ID, a1.ACC_ID, a1.ACG_ID, n1.NAD_ID, n1.NAM_NM,
                  n1.COU_ID,n1.ACC_ID, n1.ACG_ID
          FROM      DEVNAD_T."ACCOUNT" a1 JOIN DEVNAD_T.NAD n1
          ON               (
               a1.COU_ID = n1.COU_ID AND
               a1.ACC_ID = n1.ACC_ID AND
               a1.ACG_ID = n1.ACG_ID
                           )
          PRIMARY INDEX (COU_ID, ACC_ID);

Thanks

	<Prev	Next>		<<First	<Prev	Next>	Last>>