Teradata Database Message 7547 - 7547 - Analytics Database - Teradata Vantage

Message

Target row updated by multiple source rows.

Explanation

In present system we do not support to update the target row with multiple source rows. Before implementing Merge Update the join updates were processed in two steps, Merge Delete and Merge Insert. Since, we used to delete all the rows, we end up generating the duplicate rows to insert. In that scenario, we return a different error. This type of Merge Updates (Join Updates) are not allowed. For tables T1(a1,b1,c1), T2(a2,b2,c2), consider the update update t1 from t2 set c1=c2 where a1=b2 and b2 \< 10; If there are more than one row satisfying this condition, we are not sure which row needs to be used for this update. In this scenario, we return this error message. Note that with the introduction of the UDT indexing feature we also return this error when an update operation is performed on a table with a hash index if: a) the hash index is defined in one or more columns and one of the column is a structured udt column. b) the ordering function defined for the structured udt column returns null.

Generated By

AMP subsystem

For Whom

User

Notes

None

Remedy

See that source rows are unique before updating. If the update operation is performed on a table with a hash index and the hash index includes a structured udt column, verify that the ordering function does not return null.