Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Mailing Lists -> Oracle-L -> Re: library cache hit ratio > 100%

Re: library cache hit ratio > 100%

From: James Howerton <jhowerton_at_uabmc.edu>
Date: Mon, 14 Mar 2005 16:10:51 -0600
Message-Id: <s235b79d.058@uabmc.edu>


Thanks, I forgot about the bugs in 8i x$...

If hit ratios were paradise then this would be the fastest db in the land(almost).

The application vendor finally certified on 9.2.0.X so I get to drag this database up to only two years behind.;-0

...JIM...
>>> Mladen Gogala <mgogala_at_allegientsystems.com> 3/14/05 3:24:14 PM
>>>

James Howerton wrote:

>DBA's,
>
>Has anyone seen library cache hit ratio > 100% ??? This is a very
busy
>production database, it is usually < 97 % ( I don't tune by hit
ratios
>I'm just curious).
>
>
>SQL> select SUM(PINS)/(SUM(PINS)+SUM(RELOADS))*100
> from v$librarycache;
>
>
>Library Hit Ratio
>--------------------------
> 100.009799
>
>1 row selected.
>
>
>

This is a well known bug from 8i: all quantities in X$ tables are 32-bit
and when the database is
busy enough, those quantities can turn negative. That is how you get negative reloads. That was
limiting usefulness of statspack on 8i, too.Don't worry, be happy, you're living in the database tuner's
paradise: BCHR>100%. What else can you ask for?

-- 
Mladen Gogala
Oracle DBA
Ext. 121


--
http://www.freelists.org/webpage/oracle-l
--
http://www.freelists.org/webpage/oracle-l
Received on Mon Mar 14 2005 - 17:14:34 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US