Syntax - MultiLoad

Teradata® MultiLoad Reference

Product
MultiLoad
Release Number
17.00
Published
June 2020
Language
English (United States)
Last Update
2020-06-18
dita:mapPath
aim1544831946660.ditamap
dita:ditavalPath
gyk1507317446489.ditaval
dita:id
B035-2409
lifecycle
previous
Product Category
Teradata Tools and Utilities


where the following is true:

DO INSERT FOR
An upsert may be implemented by subsequent UPDATE and INSERT statements for:
  • ROWS
  • MISSING UPDATE ROWS
label
The unique name of the label that is used for the immediately following set of one or more INSERT, UPDATE, or DELETE statements
The label name must obey the same construction rules as Teradata SQL column names.
The label name in the APPLY clause of an IMPORT command can be referenced.
MARK
IGNORE
Teradata MultiLoad either places (MARK) or does not place (IGNORE) rows in the tname4 error table for the type of entry specified:
  • DUPLICATE
  • DUPLICATE INSERT
  • DUPLICATE UPDATE
  • MISSING
  • MISSING UPDATE
  • MISSING DELETE
MARK/IGNORE DUPLICATE ROWS has no effect if the table is a multiset table (which allows duplicate rows).
IGNORE DUPLICATE ROWS has no effect if the table has a unique primary index. Since a duplicate row implies a uniqueness violation in this case, the row is logged to the uniqueness violation table.
In the case of an upsert operation, both the insert and update portions must fail for an error to be recorded. In this case, the mark rows for the missing update operations then have nulls for the target table columns.
If either INSERT or UPDATE with DUPLICATE is specified, then the MARK or IGNORE specification applies to both insert and update operations.
Similarly, if either UPDATE or DELETE with MISSING is not specified, then the MARK or IGNORE specification applies to both update and delete operations.
MARK is the default for all actions except MISSING UPDATE for an upsert operation.