"tq" <tomm_quinn_at_yahoo.com> wrote in message
news:1118096159.564048.176910_at_o13g2000cwo.googlegroups.com...
> Thanks for the feedback. But if I hinted it to process the large table
> first, would it make a difference from a performance perspective? (I'm
> assuming just 2 tables, simple join, and no filtering off data during
> the full table scans.)
>
If you made Oracle use the larger data set first, i.e. for
the "build table" as Oracle calls it, then the performance
would probably be worse. In the first case, because you
may now be creating a hash table that won't fit into memory,
so you have to dump some data to disk and re-read it; in
the second case because there would be more data to
copy into memory, so the CPU resource usage would probably
go up. It's possible that the CPU change would be insignificant,
though.
--
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk/faq/ind_faq.html
The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/seminar.html
Public Appearances - schedule updated April 5th 2005
Received on Tue Jun 07 2005 - 01:43:44 CDT
