Re: testing memory parameters

From: hpuxrac <johnbhurley_at_sbcglobal.net>
Date: Sun, 16 Mar 2008 14:05:20 -0700 (PDT)
Message-ID: <cf08f7a9-1950-4bd6-b6ea-acf57a4cf0cd@k13g2000hse.googlegroups.com>


On Mar 16, 1:45 pm, "mr..._at_gmail.com" <mrow..._at_gmail.com> wrote:
> Hello,
> I'm trying to test how size of the SGA parameters affects the
> database's performace.
> I'm going to test shared_pool_size, db_cache_size and db_nk_cache_size
> parameters.
> I would like to do it properly because these tests will be part of my
> master's thesis.
> I don't have ability to use real database with real users so I must
> simulate test workload.
> I am running Oracle 10g on linux.
> Actually I have written testing program in python using cx_Oracle
> module.
> Program creates several threads, each thread represents one user.
> Users connects to Oracle and execute some SQL queries.
> I compare execution time for analyzed parameter's values.
> I think that I should consider buffer and library cache hit ratios.
> Is this a good way?
>
> These two ways measures only time and physical reads.
> What other issues should I examine?
> Can you give me some hints how can I do it well?
>
> Thanks for all comments and suggestions.
>
> Adam

You are going to get a masters degree for this? Wow is all I can say.

Yet instead of actually doing research yourself to prepare your thesis topic you want ideas from a newsgroup?

Wow again.

Good luck I guess. I am guessing you will do the OCP next perhaps? Received on Sun Mar 16 2008 - 16:05:20 CDT

Original text of this message