Just to add... optimizer stats on the DR$ tables that underpin your index is not supported until v10. (I logged a tar asking that very question a few weeks back). similarly, support told me to ensure that dynamic sampling was not enabled for a session that was going to do the sync / optimize. All which (still) strikes me as odd, since if you trace a sync or optimize command, you'll often see hints in the SQL that force the CBO anyway... other issues - dbms_stats.gather_schema_stats does not know this, so it will quite happily put stats on the DR$ tables. sigh.. On 5/26/05, Mercadante, Thomas F (LABOR) <Thomas.Mercadante@xxxxxxxxxxxxxxxxx> wrote: > Well my conclusion is that I will run resync_indexes & Optimize_index as > requested by the application. I will not gather stats of any of the > CTXSYS schema objects. I always gather stats for my schema anyway, so > the local tables created when I create a context index will have their > stats gathered. >=20 > That's what I'm going to do. >=20 > -----Original Message----- > From: Marquez, Chris [mailto:cmarquez@xxxxxxxxxxxxxxxx]=3D20 > Sent: Wednesday, May 25, 2005 1:06 PM > To: Mercadante, Thomas F (LABOR); Peter.Hitchman@xxxxxxxxxxx; > oracle-l@xxxxxxxxxxxxx > Subject: RE: CTX Indexes >=20 > So what is the verdict on Context Indexes and Oracle STATS? >=20 > ||We have to analyze > >>"have to" > >>Why?...Performance? >=20 --=20 Connor McDonald =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D email: connor_mcdonald@xxxxxxxxx web: http://www.oracledba.co.uk "Semper in excremento, sole profundum qui variat" -- //www.freelists.org/webpage/oracle-l