RE: Wrong Results - Bad Execution Plan

Are you referring to the bug discussed in note  406966.1?

----
Matt Adams - GE Consumer and Industrial
Database Administration
It will make sense as soon as you stop thinking logically
and start thinking oracle-ly.  - Jim Droppa


-----Original Message-----
From: oracle-l-bounce@xxxxxxxxxxxxx
[mailto:oracle-l-bounce@xxxxxxxxxxxxx]On Behalf Of Henry Poras
Sent: Monday, November 19, 2007 11:25 AM
To: oracle-l@xxxxxxxxxxxxx
Subject: Re: Wrong Results - Bad Execution Plan


I'll post it if and when I get it. If this is an existing bug I haven't
tracked it down yet. If it is new, Oracle hasn't assigned a bug# to it (they
are still playing the 'send me your poor, your tired database and I'll see
what I can do' game.). 

Henry


> -----Original Message-----
> From: Bob Carlin [mailto:smeghead.rimmer@xxxxxxxxx] 
> Sent: Saturday, November 17, 2007 12:01 PM
> To: henry@xxxxxxxxxxxxxxx
> Cc: oracle-l@xxxxxxxxxxxxx
> Subject: Re: Wrong Results - Bad Execution Plan
> 
> 
> Henry,
> Do you have the bug number for this issue?
> 
> Henry Poras wrote:
> > Oracle can, on occasion, generate an execution plan which 
> returns incorrect
> > results.
> >
> > In our 10.2.0.3 database we have two SQL queries which 
> should give identical
> > result sets, but don't. Outside of a filter which does not 
> apply, the
> > queries are logically identical. They do, however, resolve 
> to different
> > execution plans, one of which is correct and the other 
> which is WRONG. 
> >

...

> >
> > For the "OUTER" query, the execution plan should read
> > VIEW
> >    FILTER
> >
> > and not
> >
> > FILTER
> >    VIEW
> >
> > This is the bug!!!
> >
> > The final filter in "OUTER" filters on a.1 for the MAX 
> clause, not b.1 as is
> > correct, since a.1 is the only value for 1 passed through the view. 
> > Henry
> >

--
http://www.freelists.org/webpage/oracle-l


--
http://www.freelists.org/webpage/oracle-l

Other related posts: