Re: Building a header table

Hi Jim,
I think it's not only confusing, but also will be hard to maintain. You are talking about three tables design, but at some point between those tables you CONTACT_DATE (or whatever 'key' you assign to it) will be non-unique anyway. So, I would either index this column in your large table or try to partition it.
Hope it helps.

Sergey
Jim <jim_esti_at_hotmail.com> wrote in message news:f0e3dc0b.0106191401.25026e6a_at_posting.google.com...
> Hello,
>
> How would I go about building a header table that would have the
> following:
>
> I have a large table with these columns:
> PR_KEY, ID_NUM, PHONE_NUM, CONTACT_DATE
>
> This header table should have a unique primary key that points to
> another table
> It should also have a CONTACT_DATE field that, and a time stamp field.
>
> Now this primary key from header table will correspond with another
> table - that other table contains a PK field (that matches the Primary
> Key of the header table) and two foreign keys (namely the ID_NUM, &
> PHONE_NUM) - the ID_NUM, & PHONE_NUM should be indexed.
>
> The header table will contain a list of distinctive CONTACT_DATEs from
> the larger table. So when I ask for all CONTACT_DATEs = 28-MAY-2001
> it will go to the other table and quickly pull all the records from
> the Large Table.
>
> Is there anything specific I would need to do to the Large table?
>
> Basically what I am trying to do here is differentiate the records of
> the large table based upon the CONTACT_DATE.
>
> Confused? I am.
> Any ideas?
Received on Sat Jul 21 2001 - 23:56:04 CEST