Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Mailing Lists -> Oracle-L -> Re: v$access + latch free(library cache) BUG???
Thanks for your time Steve.
My question is why in other database this is not happening? Moreover,
any access on v$access or desc on this view causing all the session to
wait on latch free, I had to kill the session which was doing access
or desc on this view and after few second, conditions come under
control.
The strange thing is that desc causing more cpu to be used than select
* from v$access.
Thanks for your time again.
On 2/28/06, Steve Adams <steve.adams_at_ixora.com.au> wrote:
> Hello Syed,
>
> Any access to V$ACCESS on an instance with 1000+ connections creates at
> least short term contention for the library cache latches. That is not a
> bug. It is a consequence of all the work that is needed to instantiate
> the view. With your system already suffering contention for another
> latch, it is not strange that any reference to V$ACCESS would induce
> severe library cache latch contention. In short, I don't think that this
> is a bug.
>
> @ Regards,
> @ Steve Adams
> @ http://www.ixora.com.au/ - For DBAs
> @ http://www.christianity.net.au/ - For all
>
>
> The Human Fly wrote:
> > Hello list,
> >
> > We have faced a very very strange problem when accessing or desc v$access table.
> >
> > Database : Oracle 9i Rel.2 (9205)
> > OS : AIX 5.3 JFS2 Filesystem.
> > CPU = 30+
> > No. Connections = 1000+
> >
> > We are facing problem with latch free (cache buffer chains), when the
> > problem occurs, cpu becomes 0% ideal and we kill some process from
> > which transaction are coming.
> >
> > I just desc v$access from the sql prompt, and strangly, all the
> > session are started waiting on 'latch free' and this time its was
> > library cache and when I kill the session which was desc v$access,
> > then, everything comes normal.
> >
> > It was not just an accident, I can still reproduce the same error.
> > Whatever action I do on v$access, this is causing. If this was a bug,
> > we have same OS server, same oracle version for another database,
> > where this doesn't happens. The only differene between these two
> > databases are just no. of cpus.
> >
> > It is really drawing me crazy. Does anyone has experience the same?
> >
> > --
> > Best Regards,
> > Syed Jaffar Hussain
> [snip]
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>
-- Best Regards, Syed Jaffar Hussain 8i,9i & 10g, OCP DBA Banque Saudi Fransi, Saudi Arabia http://jaffardba.blogspot.com/ ---------------------------------------------------------------------------------- "Winners don't do different things. They do things differently." -- http://www.freelists.org/webpage/oracle-lReceived on Tue Feb 28 2006 - 01:44:19 CST
![]() |
![]() |