Archives of the TeradataForum
Message Posted: Wed, 31 Oct 2001 @ 16:35:14 GMT
Dave made some good points but missed one important one.
There is extra logic for many queries if the query has to deal with nulls. Sometimes this can be a few extra steps in a query plan for any joins, sub-queries, etc...
Also, I would specifically check on the issue with using columns containing nulls in sub-queries. I believe that this may still be a bad choice. I've not heard anything to change this. You can use a correlated sub-query with a not exist to get around the issue, but this may mean changing your SQL.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|