It is 9.2.0.6.=20
Not sure if it's relevant, but this is a RAC environment, but with
only 1 node been started. The second node is shutdown.
On 4/27/05, Alfonso Le=F3n <aleon68_at_gmail.com> wrote:
> which version do you have?
> there is a bug on 9.2.0.5 and below
>=20
>=20
> On 4/27/05, Christo Kutrovsky <kutrovsky.oracle_at_gmail.com> wrote:
> > Hello All,
> >
> > We have a rather bizare situation.
> >
> > We have a call to a procedure which has multiple children.
> >
> > The call looks like this:
> >
> > call package.procedure(:1,:2)
> >
> > As you can see, it uses 2 parameters which are bound via bind variables=
.
> >
> > We run this concurently in 12 sessions, and during the time this runs
> > we would have 400+ of the "call package.procdure" in the shared pool.
> >
> > We fail to understand why this call would not be shared. We examined
> > v$sql_shared_cursor and the only columns marked with Y are:
> >
> > OPTIMIZER_MISMATCH
> > STATS_ROW_MISMATCH
> >
> > The first one i can understand, even thow it's a mystery too , we're
> > not changing anything.
> >
> > The second one however is just not right. How can a call to a
> > procedure be invalidated by STATS ?
> >
> > And we would have 400+ versions of this sql, but only 50-ish rows in
> > v$sql_shared_cursor.
> >
> > And why would we have 400+ versions of this exact same procedure call.
> >
> > FYI:=3D20
> > cursor_sharing is FORCE
> > 9i database
> >
> > --=3D20
> > Christo Kutrovsky
> > Database/System Administrator
> > The Pythian Group
> > --
> > http://www.freelists.org/webpage/oracle-l
> >
>=20
> --
> Alfonso Leon
>=20
--=20
Christo Kutrovsky
Database/System Administrator
The Pythian Group
--
http://www.freelists.org/webpage/oracle-l
Received on Wed Apr 27 2005 - 12:58:44 CDT
