|
|
Archives of the TeradataForum
Message Posted: Thu, 22 Oct 2009 @ 16:27:07 GMT
Subj: | | Re: Why store duplicate rows in a table |
|
From: | | Leslie, Quincy |
What is the status of the presence of columns to store an eff date (date data captured), a date/time stamp of the call made, or a date and
number of calls (gen number) on that date, with a rollover with each new date?
If ideally all of those columns were present, then the index(es) can be configured to allow appropriate uniqueness. In additiona, a call
duration sum can occur for the calling and/or called number, rolled up per date of call, or in total or both. Time stamp will probably need to be
quite granular.
I know that in the case of cell phones (at one time), many multiple servers processed calls. Do you need to have the identity of the server(s)
involved in the call?
| |