FreeLists / oracle-l / Re: testing sql and perf

["David Green" <thump@xxxxxxxxxxxxxxxx>]

>> Statistics
>> ----------------------------------------------------------
>>        1818  recursive calls
>
> Here is one area to look at. Why 1818 recursive calls? A 10046
> trace will tell you more. In subsequent runs it is 0. IMHO,
> worry less about flushing the buffer cache and worry more about
> these calls.

On every run since this run I get 0 recursive calls.  Same code.
At the same time though, while I am trying to tune the statment related to
the stats, my question of how to tune while removing caching from the mix
remains.

> Well then don't worry about it. Your previous idea of shutting
> down/restarting would have an even worse impact.

It was'nt an idea I was suggesting nor planning on moving forward with.

> To summarize (and get around the quoting restrictions),
> 1) Address the 1818 recursive calls on the first execution.

Addressed itsle fafter the first run, which makes pinpointing what that
issue is after the first run, which leads me back to square one.  How can
I get the same result each time or differeing exec time/tkproff output
from a clean slate if you will.  This is rhetorical to a degree as you
have answered already from your point of view...

David
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to:  oracle-l-request@xxxxxxxxxxxxx
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------