| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Usenet -> c.d.o.server -> Re: OMLETv4 The Ultimate Visual Real Time Oracle Monitoring Tool
"omlet v4" <amjadd_at_uop.edu.jo> wrote in message
news:604b7892.0406160311.227d65e3_at_posting.google.com...
> > Well - you know what - I have had cause to log a number of calls with
Oracle
> > support over the years. A significant number have been related to
setting
> > CURSOR_SHARING to other than the default. This was rather my point. If
you
> > follow your installation instructions for your product, you will a)
change
> > the performance characteristics of the database that you are looking
at - a
> > very bad thing indeed for a 'monitoring' tool. and b) make your clients
apps
> > more likely to fail totally.
>
> I might at somepoint decide to listen to fools; but the product does
> not require this parameter set to anything.
One last attempt to address the issue that you seem to have been ignoring. You explicitly tell people to set this parameter in your read me. You agree that it will affect your database adversely, you say that you don't need to set the parameter. Why tell people to set it then. Your documentation on this point in full states
<quote>
5- Set the following three init.ora parameters to:
cursor_sharing = force
session_cached_cursors = 200
timed_statistics = true
example: alter system set cursor_sharing = force;
alter system set timed_statistics = true;
or add the above three parameters to init.ora and bounce the
instance.
</quote>
Perhaps you intend for people not to read the documentation? If something is optional - say so.
> However once you turn
> timed_stats on
> you are querying lots of x$ tables and this is done using views,
> select statements, etc. and you would reduce overhead. Not much
> really; just in the range of few points 2%-5%. OMLET paints pretty
> picture for you using SQL and that involves overhead.
>
> But I guess you already know that! The original message I posted for
> "leading experts" I really did not mean YOU; Pinky.
That's fair enough since I never have, nor doubt I ever will, claim to be a 'leading expert'. I don't do the cult of names thing.
On the other hand I note that you say
"Certain key ratios can be used to measure efficiency of database usage. For example, the buffer cache miss ratio indicates the frequency with which required blocks are not found in the buffer cache and must be fetched from disk. A low value indicates an inefficient use of the SGA and potential for performance improvements." which is nice and vague, but also rather hopeless as a tuning measure.
> > In what may well be a rather vain attempt to be constructive I'd suggest
> > that
>
> I thought you would list those as well:
>
> <quote>
> "
> 1) Visit Google and hunt around for advice.
>
> 2) Take nothing on face value: test it yourself to destruction.
>
> 3) Lurk here for a month or two, and try and pick up on the sort of
> questions being asked, and ask yourself how *you* would answer them.
> Then see what answers actually come through, and compare. When you're
> feeling brave, post some of your answers and see how they are taken by
> people.
>
> 4) Buy books. Anything with the names Jonathan Lewis, or Thomas Kyte
> on the cover are *extremely* good bets for accurate advice. O'Reilly
> are good. As are Apress (used to be Wrox)
>
> 5) Avoid anything that mentions OCP (Oracle Certified Professional).
> The qualification is not worth a damn, and leads you straight into
> myth territory.
> "
> </quote>
It isn't bad advice - but equally it wasn't mine.
> I wonder what Oracle or OCPs feel about your accusations. Let me get
> this, you publish all this on your web site and you are a PINKY
> angel!?
Hey, I'm an OCP too, I've acquired a number of pieces of paper and stuff over the years. OCP is one of the least valuable of them.
>
> >
> > 1. You rethink your documentation and requirements.
> > 2. You think about what effect being publicly abusive and unhelpful has
on
>
> I am just follwing you lead; PINKY! You are not worth your PINK case.
Where was I publicly abusive?
> > your target market.
>
> newsjunkies are not my market! I mean newsjunkies who lurk here or
> there for a month or two! and again not you PINKY. I solicited
> "leading experts" help and review which I think I got none.
I read your original email as publicity for your product. That was possibly misplaced - there being a newsgroup for that sort of thing - but fair enough. If you didn't mean to publicize the product here - usenet readers not being your target market, but instead solicited advice I can only remain bemused by your reaction to the criticism here.
-- Niall Litchfield Oracle DBA http://www.niall.litchfield.dial.pipex.comReceived on Wed Jun 16 2004 - 15:23:54 CDT
 |
 |