I think you forgot statspack :(
--
Niall Litchfield
Oracle DBA
Audit Commission UK
"Noons" <wizofoz2k_at_yahoo.com.au> wrote in message
news:3f65bd2c$0$18592$afc38c87_at_news.optusnet.com.au...
> "Rick Denoire" <100.17706_at_germanynet.de> wrote in message
news:10dbmv4hfji3317p1jal52j81vf14vgdmf_at_4ax.com...
>
> > I have already tried to query some v$ views (don't remember which
> > now), but I was discouraged by the fact that many applications
> > (written by external contractors) use PL/SQL with lots of variables
> > and statements transform into an unreadable code inside Oracle when
> > they run.
>
> statspack.
>
> >
> > I think that one would need more than just sqlplus and plain vanilla
> > SQL statements to deal with the issue of a performance base line.
> > Perhaps some tool out there... The question of performance of an
> > Oracle DB in general is being addressed so intensive (including
> > courses about "Performance Tuning" by Oracle itself) that I would
> > assume that there is some standard way to assess the impact of tuning
> > changes in a time-effective fashion.
>
> statspack.
>
> > I mean, if I change the optimizer mode from Cost to Rule,
>
> if you do that, you won't need a tool to measure the impact...
>
> > or if I
> > change the parameter db_file_multiblock_read count from 8 to 64, or if
> > I mount the file systems with the directio option (no OS buffering),
> > there should be indicators of how successful the changes are (for most
> > applications) without having to write benchmarks scripts (I can't
> > afford spending time for that).
> >
>
> statspack.
>
> > And considering that I have almost no clue about which queries are
> > used/written by whom, and how often they run etc., I am forced to
> > consider the overall effect instead of the impact to particular
> > queries anyway.
> >
>
> did I mention statspack?
>
> --
> Cheers
> Nuno Souto
> wizofoz2k_at_yahoo.com.au.nospam
>
>
Received on Mon Sep 15 2003 - 08:51:49 CDT
