Archives of the TeradataForum
Message Posted: Tue, 18 Jan 2005 @ 19:30:10 GMT
I might understand what you mean but the formal answer to the question is that they are pretty much always used. PI contain the information about physical location of the data, and sooner or later the SQL query will need that data. Efficient queries will get to the PI faster than the slow ones, and they will be more precise with their PI ranges. Perhaps your question is how to measure such efficiency?
There could be situations when the data comes from dependent objects such as join indexes. Perhaps you want to single out such cases?
Anyway, you probably need to be more specific.
|Copyright 2016 - All Rights Reserved|
|Last Modified: 28 Jun 2020|