I concur -- it does little good to point a finger and laugh if a
mistake has been made. Better to bring attention to the error and any
corrective action that needs to be taken.
--
David Fitzjarrell
Oracle Certified DBA
In article <39FB0A70.1F8D0740_at_fmr.com>,
Karen <karen.abgarian_at_fmr.com> wrote:
> The original poster did, of course, a mistake sizing his shared pool.
But I see
> little
> honour in making a laughin'-stock out of him just because you happen
to know
> more about this particular subject.
>
> Regards
> Karen Abgarian
>
> "Howard J. Rogers" wrote:
>
> > "sowhat" <sowhat_at_yahoo.com> wrote in message
> > news:39FA60E5.54A04997_at_yahoo.com...
> > > Understood, and point well received.
> > > However, I wonder if they were really serious....looks like they
posted
twice,
> > > with the second in a joking manner.
> > > Don't see how 35 MB of redo generation would cause the need for
235MB of
redo
> > > log buffer :-). WOW.
> > >
> > > You were a little cruel, if they weren't joking tho :-)
> > >
> >
> > Cruel is as cruel does.
> >
> > And no, I haven't a clue what that means, so it must be profound
(yikes,
> > I've just slipped into developer-mode!!)
> >
> > Stop It!!
> >
> > Ah. That's better. I'm back in DBA mode, so...
> >
> > If I offended anyone with my pathetic attempts at cruelty, er... I
mean,
> > humour, I apologise. Profusely.
> >
> > <Uncross fingers from behind back>
> >
> > Will that do?
> >
> > ;-)
> > HJR
> >
> > > "Howard J. Rogers" wrote:
> > >
> > > > It wasn't the SGA per se that had me in stitches, but the size
of the
Redo
> > > > Log Buffer. As I said, the biggest I've ever seen was 10Mb or
so, and
even
> > > > that was too big... and that was on one hell of a busy database.
> > > >
> > > > Based on the figures supplied (and I warrant my maths may be
dodgy), but
his
> > > > Shared Pool is actually smaller than his Log Buffer. If that's
properly
> > > > tuned, I'll chew my left arm off. Slowly.
> > > >
> > > > Regards
> > > > HJR
> > > > --
> > >
> > > ------------------------------------------------------------------
--------
> > > > Opinions expressed are my own, and not those of Oracle
Corporation
> > > > Oracle DBA Resources:
http://www.geocities.com/howardjr2000
> > >
> > > ------------------------------------------------------------------
--------
> > > >
> > > > "sowhat" <sowhat_at_yahoo.com> wrote in message
> > > > news:39FA066E.6412CC4C_at_yahoo.com...
> > > > > Altho one could always reply, why have 8X8 machines for Oracle
Database
Servers
> > > > > if you are only going to use 100MB for the SGA.
> > > > > I've actually seen situations with larger SGA's before --
pretty heavy
databases
> > > > > tho.
> > > > >
> > > > > "Howard J. Rogers" wrote:
> > > > >
> > > > > > "Yo" <Yo_at_ibm.net> wrote in message
news:39F9D226.6659C26B_at_ibm.net...
> > > > > > > Our SGA is sized:
> > > > > > >
> > > > > > > Total System Global Area 937877008 bytes
> > > > > > > Database Buffers 524288000 bytes
> > > > > > > Redo Buffers 235945984 bytes
> > > > > > >
> > > > > >
> > > > > > Thanks for this. I haven't had such a laugh for ages.
> > > > > >
> > > > > > Oh -but, you were being serious? You mean, you honestly
have 937Mb
SGA?
In
> > > > > > addition to a 525Mb Database Buffer Cache? Impressive
stuff if so.
> > > > > >
> > > > > > But the funniest thing has to be that Redo Log Buffer. All
235Mb of
it!
> > > > > >
> > > > > > Hilarious. This I must use on my next DBA course.... I
think 10Mb
is
the
> > > > > > biggest I've ever seen in my life before, and even that was
too big.
> > > > > >
> > > > > > Cheers, I enjoyed it.
> > > > > > HJR
> > > > > >
> > > > > > > This is for about the same size database -- generating
about 35 MB
of
redo
> > > > > > > every 15 minutes -- only one user, but a very busy user
logging
network
> > > > > > > events from a app.
> > > > > > > Yours looks a little small....but can't say for sure.
> > > > > > > I've never dealt with an SGA smaller than, say 500MB.
> > > > > > >
> > > > > > >
> > > > > > > Yo wrote:
> > > > > > >
> > > > > > > > Hi, just started a new job.
> > > > > > > > Running Oracle 8 on Solaris 2.6.
> > > > > > > > 8X8 E4500 with Fibre Channel disk arrray attached.
> > > > > > > > Redo Generation on one of the main larger (100+ GB)
Databases is
about
> > > > > > > > 50MB every 10-15 minutes.
> > > > > > > > Typically anywhere from 100-200 user connections
performing
various
> > > > > > > > transactions -- mix between read and write.
> > > > > > > > Also some batch processes running that pull data into
other
databases
> > > > > > > > from this particular instance.
> > > > > > > >
> > > > > > > > Boss told me there was some need for some performance
tuning.
> > > > > > > >
> > > > > > > > I've been poking around for a couple of days, and right
off the
bat
I
> > > > > > > > noticed something odd....
> > > > > > > >
> > > > > > > > This is the size of the SGA, of which I have never been
involved
with a
> > > > > > > > Database SGA sized so small.
> > > > > > > >
> > > > > > > > Total System Global Area
116170256 bytes
> > > > > > > > Fixed Size
48656 bytes
> > > > > > > > Variable Size
45326336 bytes
> > > > > > > > Database Buffers
65536000 bytes
> > > > > > > > Redo Buffers
5259264 bytes
> > > > > > > >
> > > > > > > > Opinions?
> > > > > > >
> > > > >
> > >
>
>
Sent via Deja.com http://www.deja.com/
Before you buy.
Received on Sat Oct 28 2000 - 18:17:48 CDT
