Archives of the TeradataForum

Message Posted: Tue, 17 Sep 2013 @ 22:17:31 GMT

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

<-- Anonymously Posted: Tuesday, September 17, 2013 16:05 -->

Under the covers the pack is rejected and the records retried until only the actual duplicates wind up in the reject table.

STREAM should handle deadlock and retry, but yes it is possible. Look at using SERIALIZE ON a subset of the PI to minimize this.

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