You should not benchmark instances, but the applications. If you think that setting a parameter to certain value will help you, then execute the application before and after setting the parameter and record the difference. You missed the whole point of the "method R". The moral of Cary's and Lawson's book is exactly that: you do not measure performance of an instance, because you don't have criteria to evaluate such performance. You measure performance of an application. The performance of applications is measured in decibels. The louder the end users scream, the worse application is performing. Performance of an instance is measured with BCHR. -- Mladen Gogala Oracle DBA email:mladeng@xxxxxxxxx Ext: 9787 > -----Original Message----- > From: ryan.gaffuri@xxxxxxxxxxx [mailto:ryan.gaffuri@xxxxxxxxxxx] > Sent: Thursday, August 19, 2004 11:07 AM > To: oracle-l@xxxxxxxxxxxxx > Subject: ((RE)): db_file_mutliblock_read_count and physical IO > > > Let's say I determined that I Had a physical IO problem. How > would I benchmark to see if db_file_multiblock_read_count > reduced this problem? Would I do a 10046 trace and compare > the number of 'db file scattered reads'? > I know I asked a similiar question yesterday, but I am > looking for a little more. What else would you do to > benchmark this to see if this parameter change had any affect > on physical IO? ---------------------------------------------------------------- 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 //www.freelists.org/archives/oracle-l/ FAQ is at //www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------