Archives of the TeradataForum

Message Posted: Wed, 26 Sep 2007 @ 20:57:48 GMT

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

Being that the tables are not defined the same, the merge into probably is not occurring and it is spooling the source table. You have a few options, and I will highlight three.

1) increase your spool to cover your spool and skewing needs - Sr. DBA as yourself should have privileges.

2) insert/select batches to (by date range) into the new table. Taking the largest chunk first since it will be inserting to an empty table (still staying under your spool limit). Then take smaller manageable chunks of data. The PPI should help the insert performance greatly as the block inserts should be within the ranged partitions.

3) fastexport the source table using named pipes access module and mload/fload it into the target. You may have to chunk this too as FastExporting the entire table has overhead that won't pass your spool limits either.

Kyle Prescott
Database Administrator
Unum-IODS

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