Archives of the TeradataForum
Message Posted: Thu, 03 Feb 2005 @ 14:55:03 GMT
<-- Anonymously Posted: Thursday, February 03, 2005 09:51 -->
Thanks Dieter, that was helpful.
We have a process that creates the volatile table (VT) and then inserts and updates the VT. We ran the process with and without 4 SI's on the VT and found no major difference in run times. The explain plan had changed a little but the SI's were not being used. We found it interesting that the run times were not longer due to the extra I/O required to write to the 4 subtables. We determined there was no advantage of using this mysterious feature. Who knows, it might cause an error in the next release.
Thanks again for your input.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 28 Jun 2020|