Archives of the TeradataForum
Message Posted: Sat, 25 Jul 2015 @ 07:55:56 GMT
Forum - There is this table A with 2.9 B records and it has 25 columns. Now I need to insert the records from table A into another table B with a slightly different structure after finding the MAX (one of VARCHAR columns) of table A based on 4 columns of table A which are not part of PI. Now when I am trying to use MAX(COLUMN 10) OVER (PARTITION BY COL 3, COL 4, COL 5, COL 6) , the INSERT is running for more than an hour. I believe this is due to the fact that the 4 columns (3,4,5, AND 6) on which I am partitioning on is not part of the PI. Is there anything that I can try to bring down the execution time or is the only way to have those 4 columns part of the index?
Please let me know, if my question is not clear.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 27 Dec 2016|