| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Mailing Lists -> Oracle-L -> RE: rm RULE based optimizer != GOOD IDEA
Wait a minute. You say that they *do* regression testing? Are you sure?
Have you seen the list of bugs?
Maybe they only do *gression* testing - you know - a half-assed job.
-----Original Message-----
From: Lex de Haan [mailto:lex.de.haan_at_naturaljoin.nl]
Sent: Friday, April 22, 2005 12:34 PM
To: Ethan.Post_at_ps.net; 'Pete Sharman'; 'Christian Antognini'
Cc: oracle-l_at_freelists.org
Subject: RE: rm RULE based optimizer != GOOD IDEA
it is about maintaining the full code path behind the RBO, keeping it in all
regression tests, checking it over and over again on all platforms and all
releases -- and last but not least, produce bug fixes if they don't do the
testing properly.
and indeed, the SQL Tuning Advisor is able to call the CBO in Tuning mode,
and
come back with a *statement specific* bunch of optimization goodies, aka SQL
profiles.
RBO is dead.
kind regards,
Lex.
-----Original Message-----
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org]
On
Behalf Of Post, Ethan
If there really is significant QA/Testing/Development that has to be done to
maintain the RBO then I understand but I would like to know the "truth"
about
the matter. If the truth is that maintaining the RULE hint isn't all that
more
complicated than maintaining the USE_HASH hint then I would like to see it
stay.
Recall, my argument is that Oracle only maintain the hint at the SQL
statement
level and that it does not need to be supported.
--
http://www.freelists.org/webpage/oracle-l
--
http://www.freelists.org/webpage/oracle-l
Received on Fri Apr 22 2005 - 13:31:15 CDT
 |
 |