What I meant was that instead of using V$SESSTAT 'db block changes' as your
starting point, you should use V$SESSSTAT 'redo size' if you are
interested in
Redo generation.
At 08:39 PM Tuesday, Jurijs Velikanovs wrote:
> V$SESSTAT 'redo size' !!! statement level statistics !!! ?
> >My thoughts so far: > >- Can easily spot the session which generating the most of the changes > >(V$SESSTAT db block changes).
Regards, Jurijs
On 2/21/06, Hemant K Chitale <hkchital@xxxxxxxxxxxxxx> wrote: > > V$SESSTAT 'redo size' > > At 08:16 PM Tuesday, Jurijs Velikanovs wrote: > >Hello Everyone, > > > >Some times there is need to troubleshoot systems/modules with huge > >amount of REDO generated. > > > >My thoughts so far: > >- Can easily spot the session which generating the most of the changes > >(V$SESSTAT db block changes). > >- Can find the object (V$SEGMENT_STATISTICS db block changes) > >- Can try to find any DML which referencing the object (the session > >have as an opened cursor). But it is still just guessing. > > > Hemant K Chitale > http://web.singnet.com.sg/~hkchital > > >
-- Jurijs +44 7738 013090 (GMT) ============================================ http://otn.oracle.com/ocm/jvelikanovs.html
Hemant K Chitale http://web.singnet.com.sg/~hkchital
-- //www.freelists.org/webpage/oracle-l
