Re: v$sql - executions vs loads vs invalidations

From: GG <grzegorzof_at_interia.pl>
Date: Tue, 01 May 2012 15:04:30 +0200
Message-ID: <4F9FDF5E.3000504_at_interia.pl>



W dniu 2012-05-01 14:37, Gerry Miller pisze:
> Hi GG
>
> These cursors don't have a high version count; there is only one of
> each, but it has been loaded thousands of times.
>
> and it said that ALL of the ORA_04031 errors occurred in subpool 2.
> Until then I was sceptical about subpool imbalance being the cause but
> now I am not so sure.

Subpool imbalance maybe caused by bug as well, You're saying there is a huge number of non-shared cursors even when setting cursor_sharing to force .
I've got feeling we are missing something important here . As a workaround You can use single subpool BUT first consult that with Oracle support and dont blame me if any other issues appear . || alter system set "_kghdsidx_count"=1 scope=spfile;|| - This requires a database restart, and may cause some performance impact as latch waits may increase.

Proper way for troubleshooting such issues is when You find out which allocations makes imbalance and do some MOS search for related bugs or kick some developers a.. . Regards
GregG

--
http://www.freelists.org/webpage/oracle-l
Received on Tue May 01 2012 - 08:04:30 CDT

Original text of this message