"spacedog" <s1m0nc00k_at_hotmail.com> wrote in message
news:1188900539.104563.43980_at_g4g2000hsf.googlegroups.com...
> Hi,
>
> I am looking for the most efficient way to perform a daily search of
> 15,000 customers against 60M transactions in the system, using a
> complicated close matching set of rules. Any matches will then be
> recorded for future use.
>
> The transactions have a number of indexes which have been put on the
> table to help optimize the search.
>
> The current approach suggested is to take batches of the customers and
> fire off multiple searches. This doesn't seem to me like a very
> efficient way of doing things as it looks like we'll be doing lots of
> repeat gets from disk. The indexes are about 15 to 20 GB in size so
> its doubtful that we'll be able to store them in memory.
>
> Any ideas would be much appreciated.
>
> Thanks
> Simon
>
Do you mean you expect to find about 15,000
customers from the 60M transactions, or you
have a total of about 15,000 customers (with an
average of 4,000 transaction each) and you want
to find the N customers whose transactions match
some complex rules.
--
Regards
Jonathan Lewis
http://jonathanlewis.wordpress.com
Author: Cost Based Oracle: Fundamentals
http://www.jlcomp.demon.co.uk/cbo_book/ind_book.html
The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html
Received on Wed Sep 05 2007 - 00:31:08 CDT
