Are you sure TIMED_STATISTICS = TRUE?
I don’t have time right now to check V$FIXED_VIEW_DEFINITION, but check how the AVERAGE_WAIT column is calculated. Just from querying V$SYSTEM_EVENT (I have 9iR1, by the way, not R2), it looks like TIME_WAITED is still rounded to centiseconds, and TIME_WAITED_MICRO is listed in unadulterated microseconds (for systems with usec or better gettimeofday resolution). The answer you probably want is
(TIME_WAITED_MICRO/TOTAL_WAITS)/10**6 seconds
instead of AVERAGE_WAIT anyway.
Cary Millsap -----Original Message-----
No .... I should have clarified ... so far we have made changes to o_i_c_a parameter at session level only ... we haven't dared it change at system level. I read through Tim's paper ('Searching for ...') and try to came up with an appropriate value, but for my db this is the (always) scenario ... oraclei@ariel-NCS2> sys SQL*Plus: Release 9.2.0.1.0 - Production on Thu Nov 21 11:36:43 2002 Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved. Connected.
EVENT
AVERAGE_WAIT SQL>
exit So, looks like until we hit 9202 we can't use this ratio (sequential/scattered) reads... does anyone know of a equiv sql using x$ tables ?? Raj
-----Original
Message-----
Raj, I'm just
curious whether you also have reset your optimizer_index_ Thanks, Cherie
Cherie,
We are
actually stuck at a value of 32 ... don't know how it got in there,
1. As we
all know it affects cost calculations and higher the value CBO
Also as
we are watching the application and see that it is choosing wrong
We did
however change this parameter on one of the production instances, so
Bottom
line, in the instance where we made the change to 4, it not faster,
And our
db is a OLTP instance (all the way), I know these methods are no
Raj
-----Original
Message-----
Raj,
Can you
elaborate on the process that you used to determine the optimal
We are
doing some performance testing on a new generation of our product
I've read through the documentation on db_file_multiblock_read_count that I can find
on Metalink but I can't find any specifics on calculating what
What
method did you use for setting the value and then for testing for
Thanks for sharing your experience with us.
Cherie
Machler
|
Title: RE: converting to the cost based optimizer
- RE: converting to the cost based optimizer VIVEK_SHARMA
- RE: converting to the cost based optimizer Sakthi , Raj
- RE: converting to the cost based optimizer Connor McDonald
- RE: converting to the cost based optimizer Binley Lim
- RE: converting to the cost based optimizer Connor McDonald
- RE: converting to the cost based optimizer Jamadagni, Rajendra
- RE: converting to the cost based optimizer Cherie_Machler
- RE: converting to the cost based optimizer Jamadagni, Rajendra
- RE: converting to the cost based optimizer Cherie_Machler
- RE: converting to the cost based optimizer Jamadagni, Rajendra
- RE: converting to the cost based optimizer Cary Millsap
- RE: converting to the cost based optimizer Deshpande, Kirti