If I recall correctly, one of my fellow Oakies (James Morle, I think it was, but don't quote me on that) did some testing in this area, and came up with a figure of around 10% or 15% overhead, running single instance on RAC binaries vs. non-RAC binaries. -Mark On Sat, 2006-12-30 at 11:46 +0530, Sudheendra Vijayakumar wrote: > Wonder how much of a difference would it make if you truned RAC off on > the same binaries. Any thoughts? > > On 12/26/06, rjamya <rjamya@xxxxxxxxx> wrote: > > IMHO non-rac benchmark should be done on non-rac binaries. > > > > Rjamya > > > > > > -- //www.freelists.org/webpage/oracle-l