Re: how to extract duplicates from a hugh database quickly

Did you try such select

select MAX(D), V1, V2 from <Table> group by V1, V2;

It must return exactly what you want, the only problem may be not so fast. But you may try to make experiment with different indexes.

"DCCoolBreeze" <DelawareKD_at_starpower.net> a écrit dans le message news: 9r4kem$huu$1_at_bob.news.rcn.net...
> I am working with a table that contains 100,000,000 records. I want to
> quickly filter out all of the duplicate records (w/o checking the date)
and
> then display the maximum date for the duplicates. Example
>
> rec1 1/1/01 fieldValue1 fieldValue2
> rec x 12/30/00 fieldValue1 fieldValue2
> ...
>
> Create a recordset listing the latest (relative to the date) duplicate
> records quickly.
>
> Any ideas
>
>
Received on Sat Oct 27 2001 - 08:16:57 CDT