|
|
Archives of the TeradataForum
Message Posted: Tue, 25 May 2004 @ 15:43:38 GMT
Subj: | | Re: Identifying row which caused 2802 Error (Duplicate Row) |
|
From: | | Judge, James A |
Without thinking about it much, and assuming the target table is not size prohibitive (as in you can't create a 2nd image), you might try
creating a new copy of the target table that you INSERT-select into rather than the UPDATE. Assuming the 2802 is caused by join(s) results that
you are not expecting, and your current target table has to have a NUPI, you could create this new table copy with an IDENTITY column for
uniqueness (otherwise the duplicate rows should be discarded) and look for PK (or a combination of columns that you know are unique) distinct
values that have a COUNT(*) GT 1. Don't necessarily have to build a table to do this; probably can create a derived table query from this
exercise.
| |