| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Mailing Lists -> Oracle-L -> Re: 10g query
In the past 6 months, I have found "SQL*Net" messages as the most common
indicator of performance problems. Chatty java, incorrect app design,
RAID-5 performance issues have all been presented as SQL*Net messages. If
I had ignored them as 'idle', I would have wasted significant amounts of
time on low-impact tuning activities. I just read a note on Jonathan
Lewis's site about the 'rdbms ipc' message indicating a too large buffer
cache. IMHO, the only way to classify an event as idle or non-idle is to
know the application and understand how it communicates with the db.
As for the jet airplane analogy, let's keep that to ourselves and not give the lawn care companies any ideas. Those &*#$& portable leaf blowers are noisy enough!
Daniel
Cary Millsap wrote:
> > "The majority of the idle events should be ignored when tuning,
> because they do not indicate the nature of the performance bottleneck.
> Some idle events can be useful in indicating what the bottleneck is
> not." citing SQL*Net message from client as an example of the latter.
>
> Heh, heh, heh--Cool. It's better than it used to be, but saying only
> that "some idle events can be useful in indicating what the bottleneck
> is not" is a little bit like saying "a jet airplane can be useful in
> clearing leaves from a field."
>
> Saying that the majority of idle events "do not indicate the nature of
> the performance bottleneck" is--forgive me--preposterous.
>
> Cary Millsap
> Hotsos Enterprises, Ltd.
> http://www.hotsos.com
> * Nullius in verba *
>
-- Archives are at http://www.freelists.org/archives/oracle-l/ FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------Received on Tue Feb 24 2004 - 12:24:32 CST
 |
 |